ABSTRACT
Cyclic Nucleotide Gated Ion Channels (CNGCs) have been firmly established as Ca2+-conducting ion channels that regulate a wide variety of physiological responses in plants. CNGC2 has been implicated in plant immunity and Ca2+ signaling due to the autoimmune phenotypes exhibited by null mutants of CNGC2. However, cngc2 mutants display additional phenotypes that are unique among autoimmune mutants, suggesting that CNGC2 has functions beyond defense and generates distinct Ca2+ signals in response to different triggers. In this study we found that cngc2 mutants showed reduced gravitropism, consistent with a defect in auxin signaling. This was mirrored in the diminished auxin response detected by the auxin reporters DR5::GUS and DII-VENUS and in a strongly impaired auxin-induced Ca2+ response. Moreover, the cngc2 mutant exhibits higher levels of the endogenous auxin indole-3-acetic acid (IAA), indicating that excess auxin in cngc2 causes its pleiotropic phenotypes. These auxin signaling defects and the autoimmunity syndrome of cngc2 could be suppressed by loss-of-function mutations in the auxin biosynthesis gene YUCCA6 (YUC6), as determined by identification of the cngc2 suppressor mutant repressor of cngc2 (rdd1) as an allele of YUC6. A loss-of-function mutation in the upstream auxin biosynthesis gene TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOPSIS (TAA1, WEAK ETHYLENE INSENSITIVE8) also suppressed the cngc2 phenotypes, further supporting the tight relationship between CNGC2 and the TAA–YUC-dependent auxin biosynthesis pathway. Taking these results together, we propose that the Ca2+ signal generated by CNGC2 is a part of the negative feedback regulation of auxin homeostasis in which CNGC2 balances cellular auxin perception by influencing auxin biosynthesis.
INTRODUCTION
Calcium (Ca2+) is a ubiquitous second messenger that orchestrates many signaling pathways in eukaryotes. A diverse range of stimuli elicit transient changes in the cytosolic free Ca2+ concentration ([Ca2+]cyt) in plants. These include developmental cues, abiotic stresses such as drought, heat, and wounding, and biotic stimuli such as interactions with pathogenic and symbiotic microorganisms (Moeder et al., 2019; Yuan et al., 2017). Each type of stimulus is thought to generate a distinctive spatio-temporal ‘Ca2+ signature’, which is decoded by the direct binding of Ca2+ to calcium sensor proteins such as calmodulins (CaM), Calcium-dependent protein kinases (CDPKs), and others (Edel et al., 2017). These sensor proteins may undergo conformational changes upon binding and interact with or phosphorylate various target substrates to regulate downstream factors (DeFalco et al., 2010; Edel et al., 2017). The Ca2+ flux is regulated by a combination of various channels, pumps, and transporters, which move Ca2+ to and from extracellular and/or intracellular Ca2+ stores to the cytoplasm (Demidchik et al., 2018). Despite the central role played by Ca2+ in plant physiological responses, the identity of plant Ca2+ channels that regulate specific [Ca2+]cyt remains elusive.
Cyclic nucleotide gated ion channels (CNGCs) function as Ca2+ channels and are linked to Ca2+ signaling in plants (DeFalco et al., 2016; Dietrich et al, 2020; Jammes et al., 2011; Zelman et al., 2012, Moeder et al., 2019). They are involved in a variety of physiological processes that are regulated by Ca2+ signaling, such as pollen tube growth, thermo-sensing, pathogen resistance, root growth, and symbiotic interactions (Brost et al., 2019; Charpentier et al., 2016; Finka et al., 2012; Moeder et al., 2011; Dietrich et al, 2020). The involvement of CNGCs in plant defense was first suggested in a study of the Arabidopsis null mutant of CNGC2, defense, no death1 (dnd1, also known as cngc2-1). cngc2 mutants exhibit autoimmune phenotypes such as stunted growth, conditional spontaneous cell death, and elevated basal levels of salicylic acid (SA), that confer enhanced resistance to various pathogens (Clough et al., 2000; Yu et al., 1998). They have a reduced ability to mount a hypersensitive response (HR), which is a form of programmed cell death (PCD) around the site of pathogen infection, often observed during effector triggered immunity (Yu et al., 1998). Two null mutants of CNGC4, HR-like lesion mimic1 (hlm1; Balagué et al., 2003) and defense, no death2 (dnd2, later referred to as cngc4; Jurkowski et al., 2004), and a gain-of-function mutant of CNGC12, constitutive expresser of pathogenesis-related genes22 (cpr22), also show alterations in defense responses (Yoshioka et al., 2006). The inhibition of HR-like spontaneous cell death in cpr22 by Ca2+ channel blockers, and the higher [Ca2+]cyt levels in cpr22 further support the notion that CNGC12 induces defense response by activating Ca2+ signalling (Urquhart et al., 2007; Moeder et al., 2011, 2019;). In contrast, the cngc2 mutant displays a reduced HR phenotype and suppression of Ca2+ signals induced by pathogen elicitors, suggesting that CNGC2 positively regulates defense (Ali et al., 2007, Tian et al., 2019). However, the autoimmune phenotype of cngc2 contradicts this notion (Moeder et al., 2011; Dietrich et al., 2020).
The cngc2 mutant is hypersensitive to elevated Ca2+ levels (Chan et al., 2003). A genome-wide transcriptional study revealed that the gene expression pattern of cngc2 resembles that of wild type under elevated Ca2+ stress (Chan et al., 2008). This observation was supported by a study reporting that CNGC2 maintains Ca2+ homeostasis by facilitating Ca2+ unloading from the vasculature into leaf cells (Wang et al., 2017). The Ca2+ hypersensitivity of cngc2 may be the cause of the autoimmune phenotypes (high SA levels, H2O2 accumulation, and cell death), since these are largely suppressed when cngc2 seedlings are grown in media with low [Ca2+] (Chan et al., 2003; Wang et al., 2017; Tian et al., 2019). cngc2 mutants are impaired in pathogen-associated molecular pattern (PAMP)-induced immunity (PTI) under Ca2+ concentrations that induce the pleiotropic phenotypes of this mutant, suggesting a complex relationship between Ca2+ concentration and immunity in this mutant (Tian et al., 2019).
In addition to these immunity phenotypes, multiple studies reported roles of CNGC2 in abiotic stress responses and development. For example, cngc2 plants display delayed flowering. This is not typical for conventional SA-accumulating autoimmune mutants, which usually exhibit early flowering (Chin et al., 2013; Fortuna et al., 2015). This finding suggests that the autoimmune phenotype observed in cngc2 may not be simply due to elevated SA levels. CNGC2 has also been suggested to play a role in CLAVATA3/CLAVATA1 (CLV3/CLV1) signaling in shoot apical meristem (SAM) maintenance (Chou et al., 2016). CLV1, a plasma membrane-localized receptor kinase, and its peptide ligand CLV3 regulate cell differentiation (Somssich et al., 2016). Most recently CNGC2 is also implicated in light stress induced Ca2+ signaling (Fichman et al., 2021). Jointly, these observations indicate that CNGC2 could be activated by a diverse set of signals, feeding into immunity, stress tolerance, and developmental outputs (Dietrich et al, 2020).
The plant hormone auxin plays a central role throughout plant development and its activity is typically associated with local accumulation that triggers a developmental response (Vanneste & Friml, 2009). The most bioactive endogenous auxin, indole-3-acetic acid (IAA), is produced from tryptophan via indole-3-pyruvate (IPA) in a two-step reaction, involving the TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOPSIS (TAA) and YUCCA (YUC) enzyme families in Arabidopsis (Mashiguchi et al., 2011; Stepanova et al., 2011; Won et al., 2011). The YUC proteins belong to the flavin-containing monooxygenase (FMO) enzyme family and play important roles in growth and development via auxin production (Cheng et al., 2006). For example, the yuc mutants and YUC over-expression lines show defects in pollen development, embryogenesis, senescence process and leaf morphology (Cao et al., 2019; Cheng et al., 2007; Kim et al., 2011). In addition, YUC genes are involved in stress responses such as heat and drought stress as many environmental stimuli converge on this pathway to modify plant growth and development (Cao et al., 2019).
To study CNGC2-mediated signal transduction, we screened for suppressors of a cngc2 null mutant and identified repressor of defense, no death1 (rdd1; Chin et al., 2013). The rdd1 mutation supresses almost all cngc2-related phenotypes, except its Ca2+ hypersensitivity, indicating that RDD1 acts downstream of CNGC2’s channel function. In this study, we identified the causal mutation of rdd1 as a loss-of-function mutation in the auxin biosynthesis gene, YUCCA6 (YUC6). Here, we propose that CNGC2 plays a role in auxin homeostasis in addition to immunity as cngc2 is defective in canonical auxin signaling and auxin-induced Ca2+ signaling, phenotypes that can also be all explained by hyper-accumulation of endogenous IAA.
RESULTS
rdd1 is a loss-of-function allele of YUC6
The CNGC2 null mutant dnd1 results from a point mutation. In this study as well as Chin et al (2013), we have used the cngc2-3 T-DNA insertion knockout line (a CNGC2 T-DNA insertion line in the Columbia background). To clone the causal locus for rdd1, rdd1 cngc2-3 plants were outcrossed with cngc2-1 (in the Wassilewskija [Ws] background) to generate a mapping population. Using 705 plants of this F2 mapping population, the causal locus for rdd1 was mapped to an approximately 800-kb region that contained 193 coding sequences (Chin et al., 2013). Using whole-genome sequencing, the causal mutation of rdd1 was narrowed down to four candidate genes within this region: AT5G24680, AT5G25590, AT5G25620, and AT5G26050. Of these, only the mutation in AT5G25620 was located in a coding region. Using qRT-PCR analysis, we found that rdd1 did not have significant alterations in the expression of the three other candidate genes (Supplemental Fig. S1). Therefore, the rdd1 mutation is likely a non-synonymous amino acid change from proline to leucine at residue 297 in the third exon of AT5G25620, which encodes the protein YUC6, a flavin-containing monooxygenase-like (FMO) protein involved in a key step in auxin biosynthesis (Fig. 1A; Mashiguchi et al., 2011; Won et al., 2011).
rdd1 is a point mutation that does not cause a premature stop codon or frame shift in the YUC6 gene. Furthermore, the rdd1 mutation was not located in or in proximity of any of the known domains of YUC6 (Fig. 1A); thus, its suppression of the cngc2 phenotype could be due to either a gain- or loss-of-function of this enzyme. To determine which of these was the case, we tested yuc6-1D, a YUC6 overexpressing activation line (Kim et al., 2007), and yucca6-3k (henceforth, yuc6), a YUC6 knockout mutant (Cheng et al, 2006), for their ability to suppress cngc2 phenotypes. First, we introduced the yuc6-1D allele into cngc2. None of the 132 F2 generation plants analyzed, resulting from a cross between yuc6-1D and cngc2 showed a rdd1-like morphological phenotype. Instead, we observed seedlings that had a smaller stature than either parental line (enhanced dwarfism), while displaying a typical yuc6-1D phenotype (such as long, narrow leaves with elongated petioles; Fig. 1B, Kim et al., 2007). Thus, we genotyped the yuc6-1D and cngc2 status of 47 F2 progenies (Supplemental Table S1). We found that these smaller plants were homozygous for cngc2 and yuc6-1D positive. Therefore, yuc6-1D does not rescue the dwarf phenotype of cngc2; rather, it enhances this phenotype (Indicated as ss in Supplemental Table S1).
By contrast, the loss-of-function mutant yuc6 suppressed the cngc2 phenotype. Indeed, the yuc6 cngc2 homozygotes in an F2 population derived from a cross between yuc6 and cngc2 were identical in phenotype to rdd1 cngc2 (Fig. 1B), indicating that rdd1 is a loss-of-function allele of YUC6. In addition, the rdd1 single mutant exhibited the identical broad leaf phenotype as yuc6, further supporting the notion that rdd1 is a loss-of-function allele of YUC6 (Fig. 1B). The cngc2-like plants were more numerous in this F2 population than would be expected if rdd1 (and therefore also yuc6) were dominant in its suppression of cngc2 (Chin et al., 2013). Therefore, we conducted additional detailed genetic analyses using two backcrossed populations of yuc6 cngc2 x cngc2 and rdd1 cngc2 x cngc2. We found that rdd1 is a semi-dominant mutation with a dosage effect (Supplemental Table S2, S3, and Supplemental Fig. S2). The difference from the previously published rdd1 segregation analysis (Chin et al., 2013) is probably due to the environmental sensitivity of the penetrance of cngc2 phenotypes. Taken together, these results show that rdd1 is a loss-of-function mutant of YUC6 and that the cngc2 phenotype depends on the dose of YUC6.
cngc2 phenotypes depend on functional YUC6
The cngc2 mutant phenotype is complex as it displays a range of seemingly unrelated phenotypes, all of which are suppressed in rdd1 cngc2, with the exception of Ca2+ hypersensitivity (Chin et al., 2013). Therefore, we analyzed the YUC6-dependence of the cngc2 phenotypes. As seen in rdd1 cngc2 (Chin et al. 2013), trypan blue staining of four-week-old plants revealed that yuc6 cngc2 plants exhibited less spontaneous cell death than cngc2 plants (Fig. 2A). Moreover, yuc6 cngc2 plants partially lost the enhanced resistance of cngc2 to the oomycete pathogen Hyaloperonospora arabidopsidis (Hpa) isolate Noco2 (Fig. 2B and C). Accordingly, increased basal levels of SA in cngc2 was suppressed to the same degree in yuc6 cngc2 as in rdd1 cngc2 (Fig. 2D). We have also analyzed the basal SA levels of the yuc6 single mutants multiple times and did not find a statistically significant alteration in these mutants, indicating that yuc6 and rdd1 epistatically suppress increased SA levels in cngc2 (Supplemental Fig. S3). Furthermore, cngc2 exhibits delayed flowering, which is suppressed in rdd1 cngc2 in an SA-independent manner (Fortuna et al., 2015). Like rdd1 cngc2, the yuc6 cngc2 plants showed suppression of the delayed flowering and even displayed earlier flowering, compared to WT (Fig. 2E). Taking these results together, we conclude that a wide range of cngc2 phenotypes, including enhanced pathogen resistance, depend on functional YUC6.
CNGC2 negatively affects TAA1/YUC6-mediated auxin biosynthesis
Since rdd1 is a loss-of-function allele of YUC6, we postulated that auxin homeostasis is mis-regulated in cngc2 mutants. In line with our expectations, cngc2 contained more endogenous IAA than the WT control, in shoots and roots in mature plants, and this was partially reversed in the rdd1 cngc2 and yuc6 cngc2 double mutants (Fig. 3). This observation was statistically significant in shoots, but not in roots; however, the same trend has been observed in roots repeatedly. Thus, we think the difference between shoot and root samples is due to technical difficulties to collect mature plant root samples. This corroborates the enhanced dwarfism of yuc6-1D cngc2 plants as yuc6-1D also hyper-accumulates auxin (Fig. 1). These results are consistent with CNGC2 negatively affecting YUC6-dependent auxin biosynthesis/homeostasis.
To discriminate between a YUC6-specific effect and a general, auxin biosynthesis-related effect, we assessed the requirement for another auxin biosynthetic gene, TAA1. The TAA1 loss-of-function allele weak ethylene insensitive 8 (wei8-1) exhibits gravitropic defects and its allelic mutant of CK-INDUCED ROOT CURLING 1 (ckrc1) additionally exhibits reduced IAA levels, due to defects in the production of the substrate of YUCCAs for IAA biosynthesis (Mashiguchi et al., 2011; Stepanova et al., 2008; Zhou et al., 2011). We generated wei8-1 cngc2 double mutants to assess the dependence of cngc2-associated phenotypes on IAA biosynthesis. These double mutants exhibited a similar suppression of cngc2 phenotypes as rdd1 cngc2 (Fig. 4A) and suppressed the elevated SA levels of cngc2 (Fig. 4B). Similar to rdd1, wei8-1 also partially rescued the delayed flowering of cngc2 (Fig. 4C, D). Taken together, these data support the hypothesis that a wide variety of cngc2 phenotypes depend on hyperactive auxin biosynthesis, rather than YUC6 function itself.
RDD1 antagonizes CNGC2-dependent auxin signaling
The rdd1 mutation suppresses almost all cngc2 mutant phenotypes (Chin et al., 2013). Morphological defects of cngc2 that differ from phenotypes of other SA-related mutants suggest alterations in hormones related to development, such as auxin. Thus, we investigated the auxin-related phenotypes of cngc2 and the rdd1 cngc2 double mutant.
We found a delayed gravitropic root tip bending response in cngc2, which was partially rescued in rdd1 cngc2 especially at earlier time points before 10 hours (Fig. 5A). We then analyzed the expression of the canonical auxin signaling output reporter, DR5 fused to GUS (DR5::GUS), as well as the Aux/IAA-based TIR1/AFB-activity sensor, DII-VENUS (Brunoud et al., 2012; Ulmasov et al., 1997). Auxin treatment elicited a strong auxin-responsive induction of DR5::GUS activity in Col wt but not in cngc2 roots (Fig. 5B). This indicates that the transcriptional auxin response is impaired in cngc2. Importantly, the rdd1 mutation nearly completely restored auxin responsiveness of DR5::GUS in rdd1 cngc2 roots (Fig. 5B), suggesting that RDD1 is involved in the control of auxin signaling by CNGC2.
Auxin activates DR5::GUS activity through the TIR1/AFB-dependent ubiquitination of Aux/IAA proteins and their subsequent proteolysis. Therefore, we monitored the activity of the TIR1/AFB co-receptor via the Aux/IAA degradation sensor DII-VENUS (Brunoud et al., 2012). Auxin treatment triggered rapid decay in the DII-VENUS signal over a 20-min period in 5-day old roots in the Col-wt. This decay was slower and dampened in the cngc2 background and was partially rescued in rdd1 cngc2 roots (Fig. 5C). Collectively, these results suggest that functional CNGC2 is required for TIR1/AFB-mediated auxin signaling. Moreover, the consistent suppression of the cngc2 auxin-insensitivity implicates RDD1 as a negative regulator of CNGC2 signalling.
CNGC2 is required for auxin-induced Ca2+ signaling in the root
A non-transcriptional branch of SCFTIR1/AFB-based auxin perception triggers Ca2+ entry into the cell in a CNGC14-dependent manner (Shih et al., 2015; Dindas et al., 2018). Given that CNGC2 is required for SCFTIR1/AFB activity in the context of transcriptional auxin signaling, we predicted a defect in auxin-induced Ca2+ signaling in cngc2. Therefore, we analyzed the auxin-induced Ca2+ response in WT, cngc2, and the rdd1 cngc2 double mutant.
The highly sensitive FRET-based Ca2+ sensor Yellow Cameleon (YC)-Nano65 (Choi et al., 2014) responded rapidly to auxin treatment at various regions of interest (ROI) along the root (Fig. 6B). Auxin treatment at the root tip of WT rapidly induced a strong peak in [Ca2+]cyt, followed by a more sustained Ca2+ response (ROI1, Fig. 6A, C). The peak response was less pronounced in more shootward ROIs (ROI2, ROI3, and ROI4), but was followed by a clear sustained Ca2+ response over at least 120 s. This observation in WT was in line with previous reports using other Ca2+ reporters (Monshausen et al., 2011; Shih et al., 2015; Waadt et al., 2017). However, the Ca2+ increase upon auxin treatment was much weaker in cngc2. This defect was largely recovered in rdd1 cngc2 (Fig. 6A, C). Application of cold water elicited identical Ca2+ signals in WT and cngc2, indicating that the impairment in IAA-induced Ca2+ signals in cngc2 is not related to a general disruption of Ca2+ signaling (Supplemental Fig. S4). Taken together, these data demonstrate a requirement for CNGC2 in auxin-mediated Ca2+ signaling that is dependent on RDD1.
DISCUSSION
In the two decades since the discovery of the autoimmunity phenotype of CNGC2 loss-of function mutants, CNGC2 has been the most intensively studied plant CNGC. However, cngc2 exhibits pleiotropic phenotypes, which are unique among conventional immunity mutants (Moeder et al., 2011, 2019), This indicates its wider physiological role beyond immunity and raises the fundamental question of whether a specific plant CNGC can generate different downstream signals depending on stimuli (Dietrich et al., 2020). To examine this question, we embarked on a suppressor screen of cngc2 and discovered an unexpected, tight connection between CNGC2 and auxin biosynthesis and signaling.
CNGC2 suppresses SCFTIR1/AFB-mediated auxin signaling
Auxin perception through the well-described SCFTIR1/AFB-based system activates transcription by proteolysis of Aux/IAA proteins (Leyser, 2018). Upon auxin treatment, the SCFTIR1/AFB receptor also activates CNGC14 via an unknown mechanism (Dindas et al., 2018; Shih et al., 2015). Here, we show that CNGC2 is required for SCFTIR1/AFB auxin signaling at the level of transcriptional regulation (DR5::GUS reporter analysis), as well as the induction of Ca2+ signals. This is reflected in the reduced root gravitropism observed in the cngc2 mutant.
Since CNGCs were suggested to form heterotetrameric channels (Chin et al., 2013; Pan et al., 2019; Tian et al., 2019) and CNGC14 plays a role in auxin-related Ca2+ signaling (Shih et al.; 2015, Dindas et al., 2018), it is possible that CNGC2 and 14 form a functional Ca2+ channel conducting auxin-induced Ca2+ influxes. Indeed, both CNGC14 and CNGC2 localize at the plasma membrane and are required for auxin-induced Ca2+ influx into the cytosol from the extracellular apoplastic Ca2+ pool (Lemtiri-chlieh & Berkowitz, 2004; Wang et al., 2017; Dindas et al. 2018; Shih et al., 2015). Moreover, the gravitropic response defects of cngc2 and cngc14 were quite similar under our experimental conditions (Supplemental Fig. S5A). However, in contrast to the pleiotropic developmental defects seen in cngc2 (Clough et al., 2000), cngc14 exhibits little to no additional developmental defects besides the reduced gravitropism and root hair phenotypes (Dindas et al., 2018; Shih et al., 2015; Brost et al., 2019). An additional obvious difference from cngc2 mutants is the absence of defense-related phenotypes in cngc14 (Brost et al., 2019; Dindas et al., 2018; Shih et al., 2015). These differences together with the limited overlap in expression patterns (Supplemental Fig. S5B) indicate that CNGC14 and CNGC2 likely have different biological functions and argues against CNGC2 being a component of the SCFTIR1/AFB-CNGC14 module and/or forming a heterotetrameric channel with CNGC14. Therefore, we favor a scenario in which CNGC2-mediated Ca2+ signals control auxin biosynthesis and the activity of SCFTIR1/AFB -auxin sensing independent from CNGC14. This also provides a straightforward explanation for the reduced auxin sensitivity of DR5::GUS and DII-VENUS auxin reporters in cngc2.
CNCC2-CNGC4 mediated Ca2+ signaling suppresses auxin biosynthesis
Whether CNGC2 can form a heterotetramer with CNGC14 remains to be seen; however, CNGC2 is known to form a functional Ca2+ channel through heteromerization with CNGC4 (Chin et al., 2013, Tian et al., 2019). The respective mutants, cngc2 (dnd1) and cngc4 (hlm1/dnd2) have identical morphological and molecular phenotypes. To date, these phenotypes have mainly been characterized with respect to their autoimmunity phenotypes (Clough et al., 2000, Balagué et al., 2003), and were recently shown to be defective in PAMP-induced Ca2+ signals (Tian et al., 2019). Here, we found that cngc2 exhibits hyper-accumulation of endogenous IAA similar to cngc4 (Kale et al., 2019). This observation indicates that a CNGC2–CNGC4 heteromeric channel generates a Ca2+ signal that suppresses IAA biosynthesis. The auxin hyperaccumulation in cngc2, and most of its pleiotropic phenotypes, could be repressed when the auxin biosynthesis gene YUC6 was mutated. Moreover, the phenotypes of cngc4 and even the double mutant cngc2 cngc4 could also be suppressed by the yuc6 allele rdd1 (Chin et al., 2013), suggesting that our findings about auxin signaling and biosynthesis in cngc2 can be extrapolated to cngc4.
In addition to its FMO function for IAA biosynthesis, YUC6 also exhibits thiol reductase activity, which plays a role in ROS homeostasis (Cha et al., 2016; Cha et al., 2015). Since hyper-accumulation of ROS is a common phenomenon associated with many lesion mimic mutants, including cngc2, it is possible that the thiol reductase activity rather than the FMO function of YUC6 is related to the suppression of cngc2 phenotypes. However, the rdd1 mutation is not located in the thiol reductase domain (Fig. 1A). Furthermore, loss-of-function of TAA1/WEI8, which functions one step upstream of YUC6 in the IAA biosynthesis pathway, also rescues cngc2 phenotypes. This strongly suggests that alterations in IAA biosynthesis, not the thiol reductase activity of YUC6, lead to the suppression of cngc2. The taa1 (wei8-1) mutation had a slightly weaker effect on cngc2 than yuc6, probably due to redundancies with TRYPTOPHAN AMINOTRANSFERASE RELATED PROTEIN 1 and 2 (TAR1, TAA2) (Stepanova et al., 2008). Taken together, the unique morphological and physiological phenotypes of cngc2 and cngc4 among lesion mimic mutants can be explained by a hyperactive TAA–YUC auxin biosynthesis pathway. Indeed, other lesion mimic mutants such as suppressor of npr1-1, constitutive1 (snc1) and constitutive expressor of PR genes 6 (cpr6) exhibit lower levels of endogenous IAA (Wang et al., 2007), further supporting this notion. Recently, auxin perception via the TRANS-MEMBRANE KINASE 4 (TMK4) receptor-like kinase, was shown to suppress auxin biosynthesis via phosphorylation of TAA1 (Wang et al., 2020). Therefore, it will be interesting to see if TMK4 acts together with CNGC2 to suppress auxin biosynthesis.
CNGC2-mediated Ca2+ signals at the nexus of immunity and development
As mentioned, the autoimmunity mutants cngc2 and cngc4 have almost identical phenotypes including elevated SA levels, which are suppressed by rdd1 (Chin et al., 2013). Thus, the identification of RDD1 as the auxin biosynthesis gene YUC6 initially led us to hypothesize that the effect of rdd1 is simply restoring the balance in the SA–auxin antagonism in cngc2. Auxin treatment promotes disease symptoms and prevents the full induction of the SA-inducible antimicrobial gene PATHOGENESIS RELATED1 (PR1), indicating that auxin antagonizes defense responses via suppression of SA signaling. Indeed, many pathogenic microorganisms manipulate plant immunity through modification of auxin signaling by their effectors and/or auxin originating from the microorganisms themselves (Kazan and Lyons, 2014; Kunkel and Harper, 2018; Mutka et al., 2013). Moreover, treatment with SA or the SA analog BTH globally suppresses transcription of auxin-related genes and various autoimmune mutants exhibit lower IAA levels and reduced sensitivity to auxin (Wang et al., 2007). However, the cpr22 autoimmune phenotype, caused by the CNGC11/12 gain-of-function mutation, which induces constitutive activation of Ca2+ signals (Moeder et al., 2019), was not suppressed by rdd1 (Fortuna et al., 2015). Furthermore, blocking SA biosynthesis by introducing the sid2 mutation reverted almost all phenotypes of cpr22 (Yoshioka et al., 2006), while the sid2 cngc2 and sid2 cngc4 double mutants retain clear morphological defects (Genger et al., 2008). Thus, simple SA–auxin antagonism is not the cause of the suppression of cngc2 and cngc4 phenotypes by rdd1.
These data also suggested that CNGC2/CNGC4 activate autoimmunity through distinct pathways, indicating a unique aspect in CNGC2/CNGC4-mediated immunity among lesion mimic mutants. Thus, CNGC2 may primarily play a role in auxin homeostasis/signaling and its observed immunity phenotypes (i.e., SA accumulation) are a consequence of this auxin-related defect. Alternatively, CNGC2 may act independently or at a pivotal point in multiple physiological processes, such as defense and development. Indeed, CNGC2 has recently been implicated in the regulation of SAM size through the CLAVATA signaling cascade and Ca2+ homeostasis (Chou et al., 2016; Wang et al., 2017) as well as PAMP-induced immunity (Tian et al., 2019). Ca2+ is a universal secondary messenger and is involved in almost all aspects of cellular signaling. It is possible that CNGC2 generates distinct Ca2+ signals (or Ca2+ signatures) depending on the stimulus and acts at the nexus of immunity and development. Such specificity could be achieved through changing channel subunit composition or by being part of a channelosome with stimuli-specific signaling components, such as receptors and decoders (Dietrich et al., 2020).
Taking our results together, we propose that plasma membrane localized CNGC2 forms a heterotetrametric channel with CNGC4 and generates Ca2+ signals that affect the TAA–YUC auxin biosynthetic pathway, likely to prevent over accumulation of IAA. CNGC4 has been reported to form a heterotetramer with CNGC2 (Chin et al., 2013, Tian et al., 2019) and cngc4 also exhibits abnormal accumulation of IAA (Kale et a., 2019). Since the loss-of-function mutants of CNGC2 are impaired in their ability to generate a Ca2+ influx upon IAA treatment, the regulation of the TAA–YUC auxin biosynthetic pathway by CNGC2/CNGC4 must be a feedback regulation for auxin homeostasis, similar to that proposed by Wang et al. (2020) for TMK4. In addition, CNGC2 function is required for TIR/AFB-mediated auxin signaling; thus, CNGC2/4 mediated Ca2+ signals may control the activity of SCFTIR1/AFB-auxin sensing directly, or via facilitating proper IAA homeostasis. In this scenario, cngc2 has abnormal accumulation of IAA resulting in the desensitization of cellular auxin signaling and morphological defects as a long-term effect. Alternatively, Ca2+ signals generated by CNGC2 may directly affect auxin signaling in addition to its biosynthesis. In any case, CNGC2 must play a role immediately after perception of auxin or associated to the perception itself. Although at this point, we cannot exclude other possibilities such as indirect effects, this model can be one plausible scenario. Further investigation of direct downstream targets of CNGC2-mediated Ca2+ influx will be necessary.
For two decades, CNGC2 has been studied intensively from an immunity point-of-view. However, publications in recent years as well as our current work strongly indicate a more diverse range of functions for CNGC2. Thus, the current study contributes to a more comprehensive understanding of CNGC2-mediated Ca2+ signaling and reveals the importance of CNGC2 beyond its role in defense.
MATERIALS AND METHODS
Plant Materials and Growth Conditions
For phenotypic analyses, Arabidopsis thaliana seeds were cold stratified at 4°C for two days prior to being grown at 23°C on Sunshine-Mix #1 (Sun Gro Horticulture, Vancouver, Canada). Plants were grown in a growth chamber with a 9-h photoperiod (9-h light/ 15-h dark) and a day/night temperature regime of 22°C/18°C. This condition prolongs the growth stage and induces clearer morphological phenotypes of the mutants. For other experiments, plants were grown on Petri dishes with half-strength Murashige and Skoog (½ MS) medium, 1% sucrose, and 0.8% (w/v) agar at pH 5.8 under ambient light conditions. Sterile media was either supplemented with IAA of the desired concentration or equal volumes of ethanol as a solvent control. Homozygous mutants were identified by PCR using primers listed in Supplementary Table S4. Since rdd1 was isolated using the T-DNA insertion allele cngc2-3 (Chin et al., 2013), we have used cngc2-3 throughout this work except for the Ca2+ imaging analysis (see FRET-based Ca2+ analysis using YC-Nano65 section).
Identification of rdd1 as an allele of YUC6
Illumina HiSeq 1500 system (Illumina, Inc., San Diego, CA, USA) at the McMaster Institute for Molecular Biology and Biotechnology (MOBIX) was used to sequence the genomic DNA extracted from leaves of 4-week old rdd1 cngc2-3 and cngc2-3 plants. DNA was extracted from 100 mg of powdered leaf tissue using the CTAB method (Thompson, 1980) with a slight modification: 200 units/ml of RNAse A and 0.4 g/ml PVP (MW 40000; Sigma-Aldrich, Canada) were added to the CTAB extraction buffer. Extracted DNA was precipitated twice with isopropanol, once with 75% EtOH and resuspended in 0.1X TE buffer supplemented with 0.2 units of RNAse A. DNA integrity was assessed using agarose gel electrophoresis and quantified with a Quant-iT PicoGreen dsDNA Assay Kit (Thermo Fisher Scientific) on a Bio-Tek Powerwave HT microplate reader. Multiplex libraries were prepared according to the manufacturer’s instructions (Nextera) and the three libraries were sequenced on one flow-cell lane in high throughput mode. Raw sequence reads were trimmed for adaptor sequences with Cutadapt software (Martin, 2011) and unpaired reads, and those of length <36 nt, were excluded from further analyses. Paired reads were aligned against the TAIR10 WT reference genome using BWA (Li & Durbin, 2009), and sequence variants were detected with the mpileup function of SAMtools (Li et al., 2009).
GUS enzymatic assay
Seedlings were grown on ½ MS, 1% sucrose, and 0.8% (w/v) agar plates for six days before being transferred to plates supplemented with 1 µM IAA, or solvent alone (ethanol at final concentration of 0.09%), for 24 hours. GUS reporter activity was analyzed at an excitation wavelength of 365 nm and emission wavelength of 455 nm, using a TECAN plate reader (every 10 min for 2 h) in the presence of 4-methylumbelliferyl glucuronide (4-MUG). GUS activity was standardized against protein concentration and data was reported as GUS activity in pmol 4-methylumbelliferone (4MU) per µg protein.
Pathogen infection
Infection with Hyaloperonospora arabidopsidis isolate Noco, which is virulent to the WT accession of Arabidopsis was performed as described previously with 5 × 105 spores per ml (Chin et al., 2013).
Analysis of endogenous salicylic acid
Endogenous SA was analyzed using the Acinetobacter sp. ADPWH lux-based biosensor as previously described (Defraia et al., 2008).
Gravitropic root bending
Five to seven-day-old seedlings were photographed every two hours between 4 and 10 hours, and then again at 24 hours, from the start of gravistimulation. The deviation in root tip angle from 90° was analyzed using the angle tool on ImageJ software (http://rsbweb.nih.gov/ij/).
Analysis of flowering transition time
Arabidopsis thaliana wildtype and mutant plants were grown on Sunshine-Mix #1 (Sun Gro Horticulture, Vancouver, Canada) in a growth chambers under 16-h photoperiod (16-h light/ 8-h dark) at 22°C/18°C. Observations were made every other day and floral transition was recorded as days taken for first bolt to form from time of sowing as described in Fortuna et al., 2015.
DII-VENUS analysis
For DII-VENUS analysis, seedlings were visualized for 20 minutes immediately after supplementation with 1 µM IAA. Confocal images were captured using a Leica TCS SP5 confocal system with an acousto-optical beam splitter (HCX PL APO CS 40x immersion oil objective; numerical aperture, 1.25), and the acousto-optic tunable filter 514 for the argon laser using the nm output, set at 20%. The detection window was set to 525 to 600 nm for YFP (Leica Microsystems). Seven-to nine-day-old seedings were stained with 1x propidium iodide and imaged at 40X magnification. Images were processed using Leica Las AF lite software.
FRET-based Ca2+ analysis using YC-Nano65
Multiple lines of Wt, dnd1, and rdd1 dnd1 carrying YC-Nano65 were generated as previously described (Choi et al., 2014). These plants were grown on the surface of a vertical agar plate with ½ MS, 1% (w/v) sucrose, and 0.5% gellan gum at pH 5.8 under ambient light conditions. The root tips were treated with 10 μl of 1 μM IAA. FRET (cpVenus) and CFP signals from YC-Nano65 were acquired using a motorized fluorescence stereo microscope (SMZ-25, Nikon) equipped with image splitting optics (W-VIEW GEMINI, Hamamatsu Photonics) and a sCMOS camera (ORCA-Flash4.0 V2, Hamamatsu Photonics) as previously described (Lenglet et al., 2017; Toyota et al., 2018). For high-resolution confocal Ca2+ analysis, the transgenic plants were grown vertically under a thin layer (approximately 2 mm) of the growth medium (½ MS, 1% (w/v) sucrose and 0.5% gellan gum at pH 5.8) on a cover glass (24 × 50 mm, Fisher Scientific) for six days at 23°C. The tip of the root was exposed by removing a small window (approximately 500 μm × 500 μm) from the gel, and 10 μl of 1 μM IAA was applied to the root tip area through this window. FRET (cpVenus) and CFP signals from YC-Nano65 were acquired using a laser scanning confocal microscope (LSM780/Elyra; Newcomb Imaging Center, Department of Botany, University of Wisconsin, Madison) as previously described (Choi et al., 2014). The cpVenus/CFP ratio was calculated using 6D imaging and Ratio & FRET plug-in modules, and the kymograph of the entire root was generated over 240 seconds (NIS-Elements AR, Nikon).
Accession Numbers
Sequence data from this article can be found in the EMBL/GenBank data libraries under accession number(s): CNGC2 (AT5G15410), RDD1/YUC6 (AT5G25620), TAA1 (AT1G70560).
Author contributions
K.Y. and W.M. conceived the project; K.Y., W.M., S.V., T.B., M.T., S.G., E.N. designed the project; S.C., K.C., A.F., M.C., S.V., M.T., W.M. and E.N. performed the experiments and analyzed the data; S.C., K.Y., W.M. and S.V. analyzed data and wrote the article.
Acknowledgements
We thank Dr. Catherine Chan for providing the Ws cngc2 seeds, Drs. Barbara Kunkel and Yunde Zhao for providing the Yucca6-1D seeds, and Drs. Thomas Berleth and Enrico Scarpella for providing the DII-VENUS and DR5::GUS transgenic seeds. We appreciate the fruitful discussion and help from Dr. Matyáš Fendrych. Genomic sequencing was done with the help of Dr. Elizabeth Weretilnyk and the McMaster Institute for Molecular Biology and Biotechnology (MOBIX). We thank Drs. Andrew S Whiteley and Zhonglin Mou for providing the ADPWH lux-based salicylic acid biosensor. This work was supported by a Discovery grant from the National Science and Engineering Research Council (NSERC) to K.Y., a graduate student fellowship from NSERC to S.C., and KAKENHI (17H05007, 18H05491) to M.T.
Footnotes
Competing interests: The authors declare that no competing interests exist.
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors is keiko.yoshioka{at}utoronto.ca.
Section on CNCC2-CNGC4 mediated Ca2+ signaling suppresses auxin biosynthesis and section on CNGC2 negatively affects TAA1/YUC6-mediated auxin biosynthesis was updated with new data.Supplemental files is also revised with new data.