Genetic characterization reveals no role for the reported ABA receptor, GCR2, in ABA control of seed germination and early seedling development in Arabidopsis

Authors

  • Yajun Gao,

    1. Department of Botany, University of British Columbia, Vancouver, BC V6T 1Z4 Canada
    2. College of Resources and Environment, Northwest A & F University, Yangling, Shaanxi 712100, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Qingning Zeng,

    1. Department of Botany, University of British Columbia, Vancouver, BC V6T 1Z4 Canada
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Jianjun Guo,

    1. Department of Botany, University of British Columbia, Vancouver, BC V6T 1Z4 Canada
    Search for more papers by this author
  • Jia Cheng,

    1. Department of Botany, University of British Columbia, Vancouver, BC V6T 1Z4 Canada
    2. Michael Smith Laboratories, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
    Search for more papers by this author
  • Brian E. Ellis,

    1. Michael Smith Laboratories, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
    Search for more papers by this author
  • Jin-Gui Chen

    Corresponding author
    1. Department of Botany, University of British Columbia, Vancouver, BC V6T 1Z4 Canada
    Search for more papers by this author

*(fax +1 604 822 6089; e-mail jingui@interchange.ubc.ca).

Summary

Abscisic acid (ABA) is perceived by several different types of receptors in plant cells. At the cell surface, the ABA signal is proposed to be perceived by GCR2, which mediates ABA responses in seed germination, early seedling development and stomatal movement. GCR2 was also proposed to be a seven-transmembrane (7TM) G-protein-coupled receptor (GPCR). Here we characterize GCR2 and one of its two homologs, GCR2-LIKE 1 (GCL1), in ABA-mediated seed germination and early seedling development in Arabidopsis. We show that loss-of-function mutations in GCL1 did not confer ABA insensitivity. Similarly, we did not observe ABA insensitivity in three independent gcr2 alleles. Furthermore, we generated gcr2 gcl1 double mutants and found that the double mutants still had near wild-type responses to ABA. Consistent with this, we found that the transcription of ABA marker genes was induced by ABA to levels that were comparable in wild type and gcr2 and gcl1 single and double mutants. On the other hand, the loss-of-function alleles of the sole Arabidopsis heterotrimeric G protein α subunit, GPA1, were hypersensitive to ABA in the ABA-inhibition of seed germination and early seedling development, disfavoring a genetic coupling of GCR2 by GPA1. Using multiple robust transmembrane prediction systems, GCR2 was predicted not to be a 7TM protein, a structural hallmark of GPCRs. Taken together, our results do not support the notion that GCR2 is an ABA-signaling GPCR in seed germination and early seedling development.

Introduction

The plant hormone abscisic acid (ABA) regulates diverse processes in growth and development, including seed maturation, seed dormancy, root growth, leaf senescence and flowering. ABA is also a major plant hormone involved in stress responses, best known for its role in controlling stomatal movement, a process that regulates water loss from plant tissues. ABA-response mutants have been useful tools to dissect the ABA signal transduction pathway. Analyses of these mutants revealed that the ABA signal transduction pathways involve protein kinases (e.g. ABA-activated protein kinase, AAPK), phosphatases (e.g. ABI1 and ABI2, serine/threonine phosphatases, PP2Cs) and transcription factors (e.g. ABI3, ABI4, ABI5) (Finkelstein et al., 2002; Himmelbach et al., 2003).

Biochemical analyses and the determination of the structural requirements for ABA activity suggested multiple modes of ABA perception and receptor types in plant cells (Huang et al., 2007). Early experiments using microinjected ABA (Anderson et al., 1994; Gilroy and Jones, 1994), and ABA-protein conjugates that are unable to cross the membrane (Jeannette et al., 1999; Schultz and Quatrano, 1997), suggested that ABA may be perceived by both cell surface and intracellular receptors.

Consistent with this conclusion, several different types of ABA receptors have recently been identified. FLOWERING TIME CONTROL PROTEIN A (FCA), a nuclear RNA-binding protein, was identified as an ABA receptor involved in ABA-mediated RNA metabolism, and in controlling flowering time (Razem et al., 2006). The H subunit of Mg-chelatase (CHLH), a chloroplast protein and a key component in both chlorophyll biosynthesis and plastid-to-nucleus signaling, was identified as an ABA receptor (ABAR) that mediates ABA signaling in seed germination, post-germination growth and stomatal movement (Shen et al., 2006). However, neither FCA nor CHLH is a plasma membrane-localized protein, suggesting that there are other receptors perceiving the ABA signal at the cell surface. For example, a membrane-bound, Leu-rich repeat (LRR) receptor-like kinase 1, RPK1, is a candidate receptor for perceiving the ABA signal at the plasma membrane (Osakabe et al., 2005). Very recently a long-awaited plasma membrane ABA receptor was reported by Liu et al. (2007). GCR2 was found to specifically bind ABA at physiological concentrations with expected kinetics and stereospecificity (Liu et al., 2007). Loss-of-function alleles of GCR2 were shown to be insensitive to ABA in all known major ABA responses, including seed germination, early seedling development and stomatal movement, as well as ABA-induced gene expression. GCR2 was also predicted to be a seven-transmembrane (7TM) G-protein-coupled receptor (GPCR), and was shown to physically bind to the sole heterotrimeric G-proteins α subunit (Gα), GPA1, in Arabidopsis (Liu et al., 2007). The significance of this discovery was twofold: it identified the first plasma membrane ABA receptor, and it identified the first plant GPCR together with its ligand.

GCR2 has two homologs in Arabidopsis. Here we characterize both GCR2 and one of its homologs, GCR2-LIKE 1 (GCL1), in the ABA inhibition of seed germination and early seedling development. We provide molecular and genetic evidence that neither GCR2 nor GCL1 is required for ABA responses, and that GCR2 may not be genetically coupled with GPA1 in the ABA response. We provide bioinformatic evidence that GCR2 does not possess a 7TM protein structure, a structural hallmark of all GPCRs, and that GCR2 is likely to be a plant homolog of bacterial lanthionine synthetase.

Results

GCR2 homologs in Arabidopsis

A BLAST search, using the GCR2 amino acid sequence as a template, identified two GCR2 homologs in Arabidopsis encoded by gene loci At5g65280 and At2g20770, respectively. We named the gene products of At5g65280 and At2g20770 GCL1 and GCL2, respectively (Figure 1). Alignment of the predicted amino acid sequences indicated that GCR2 is 61% similar and 44% identical to GCL1, and 79% similar and 64% identical to GCL2 (Figure 1a and 1b). The in silico data from the Genevestigator Arabidopsis thaliana microarray database (http://www.genevestigator.ethz.ch; Zimmermann et al., 2004) indicated that GCR2, GCL1 and GCL2 have overlapping expression patterns in various tissues and organs (Figure S1). We also detected GCR2 homologs in other plant species, such as rice, potato and Medicago truncatula (Figure S2).

Figure 1.

 GCR2 homologs in Arabidopsis.
(a) Protein sequence alignment of Arabidopsis GCR2 (At1g52920) and its homologs with human LANCL1 and LANCL2 proteins. The Arabidopsis genome contains two GCR2 homologs encoded by gene loci At5g65280 and At2g20770. They are designated here as GCR2-LIKE 1 (GCL1) and GCR2-LIKE 2 (GCL2), respectively. The accession numbers for human LANCL1 and LANCL2 are NP_006046 and NP_061167, respectively. Amino acids that are identical or similar are shaded with black or gray, respectively. The sequence alignment was generated by the ClustalW multiple alignment of the BioEdit Sequence Alignment Editor (http://www.mbio.ncsu.edu/BioEdit/bioedit.html).
(b) Pair-wise analysis of the identity and similarity at the amino acid level among GCR2, GCL1, GCL2, LANCL1 and LANCL2 proteins.

Is GCR2 a GPCR?

When the amino acid sequence of GCR2 was compared with those from non-plant organisms, we found that GCR2 is highly similar to human proteins LANCL1 and LANCL2 (Figure 1), both of which are human homologs of bacterial lanthionine synthetase (Bauer et al., 2000; Landlinger et al., 2006; Mayer et al., 2001). GCR2 is 56% similar and 40% identical to LANCL1, and 58% similar and 40% identical to LANCL2 at the amino acid level (Figure 1b). Significant sequence homology of GCR2 was also detected to proteins in other mammals, chicken, frogs, zebrafish, insects, fungi and bacteria (Figure S2), suggesting that GCR2 is a member of an evolutionarily conserved protein family.

LANCL1 was initially misidentified as a GPCR (Mayer et al., 1998), but this was later corrected when the authors determined that LANCL1 is not a GPCR and is in fact a LanC ortholog (Bauer et al., 2000). Biochemical studies confirmed that LANCL1 is a peripheral membrane protein, rather than a transmembrane protein (Bauer et al., 2000), and that the membrane localization of a related protein, LANCL2, is most likely to result from both myristoylation and lipid binding (Landlinger et al., 2006). These studies of human homologs of GCR2 imply that GCR2 could also be a peripheral membrane protein.

Therefore, we wanted to examine the argument that GCR2 represents an authentic GPCR as proposed by Liu et al. (2007). Liu et al. (2007) searched the Arabidopsis genome for candidate 7TM proteins and found that GCR2 was a putative GPCR, according to their search criteria. They subsequently found that GCR2 physically interacted with the sole Gα, GPA1, in Arabidopsis. GCR2 was therefore interpreted as a GPCR (Liu et al., 2007). However, when we re-examined the prediction that GCR2 has 7TM helices, the structural hallmark of GPCRs, using the dense alignment surface (DAS) transmembrane segment prediction system (http://www.sbc.su.se/~miklos/DAS) used by Liu et al. (2007), we found that GCR2 was not predicted to be a 7TM protein. When using the high stringency cut-off (DAS profile score 2.2), the software yields a prediction that GCR2 possesses only one transmembrane segment (Figure S3). When the less stringent cut-off of 1.7 is used, GCR2 is predicted to have six, but not seven, transmembrane segments (Figure S3). In our analysis we used GCR1 as a positive control, as this protein has been predicted previously to have a 7TM structure by several independent research groups (Chen et al., 2004; Josefsson and Rask, 1997; Moriyama et al., 2006; Pandey and Assmann, 2004;Plakidou-Dymock et al., 1998). We found that when the less stringent cut-off was used, GCR1 was predicted to have seven transmembrane segments (Figure S3). Because Liu et al. (2007) did not report the score threshold they used to evaluate the confidence of prediction, it is not clear how they came to the conclusion that GCR2 is a 7TM protein.

It is worth noting that the original DAS system can erroneously predict transmembrane helices with a high false-positive rate (Moller et al., 2001), which has led to the development of a new version of DAS that includes a filter for false-positive transmembrane prediction (http://mendel.imp.univie.ac.at/sat/DAS/DAS.html). A link for this new version of DAS-TMfilter prediction system is available from the same webpage of DAS. This new system predicts no transmembrane segments for GCR2, but GCR1 was still predicted to be a 7TM protein (Figure S3c,d).

To examine this more closely, we also used the TMHMM2.0 (http://www.cbs.dtu.dk/services/TMHMM) transmembrane segment prediction system, a robust algorithm reported to yield a low false-positive rate (Moller et al., 2001). Again, not a single transmembrane segment was predicted for GCR2 (Figure S3f), whereas GCR1 was predicted to be a 7TM protein with an extracellular N-terminus and intracellular C-terminus (Figure S3e). In conclusion, our bioinformatic analyses do not support the notion that GCR2 is a 7TM protein. This is consistent with the results of Moriyama et al. (2006), who used a range of protein classification methods to identify potential Arabidopsis 7TM proteins, but did not classify GCR2 as a 7TM protein, and it is also consistent with the studies of human homologs of GCR2 discussed above. Therefore, we concluded that GCR2 is unlikely to be a GPCR, because it lacks the 7TM structural hallmark of GPCRs.

Mutant alleles of GCL1

Liu et al. (2007) provided evidence that GCR2 is also a receptor for ABA. If GCR2 is not a GPCR, is it still possible that GCR2 acts as a receptor for ABA? Loss-of-function alleles of GCR2 were previously shown to have defects in perceiving the ABA signal, but the gcr2 mutant still displayed detectable ABA responses (Liu et al., 2007). One possible explanation for the weak response of gcr2 mutants to ABA was that there may be functional redundancy between GCR2 and its homologous genes in Arabidopsis. This notion is supported by the overlapping tissue/organ expression patterns of GCR2, GCL1 and GCL2 genes (Figure S1). To examine this possibility, we took a reverse-genetics approach and characterized insertional mutants for the GCR2 homologous genes, GCL1 and GCL2. Here we report on the characterization of gcl1 alleles.

By searching the SALK Institute sequence-indexed insertion mutant collection (http://signal.salk.edu/cgi-bin/tdnaexpress), we identified two independent insertional alleles of GCL1: gcl1-1 and gcl1-2 (Figure 2a). Both alleles lack full-length GCL1 transcripts, suggesting that they are likely to be loss-of-function alleles of GCL1 (Figure 2b). A truncated GCL1 transcript was also detected in the gcl1-2 allele (Figure 2b). We have yet to detect any significant morphological phenotypes of these gcl1 mutants (Figure 2c).

Figure 2.

 Mutant alleles of GCR2-LIKE 1 (GCL1).
(a) A diagram to illustrate the T-DNA or transposon insertion site in the gcl1 mutants. Gray boxes represent exons. The T-DNA and transposon inserts are not drawn to scale. LB, T-DNA left border. The arrows indicate the positions of the GCL1 gene-specific primers (FW1 or FW2 and RV) used for amplifying the GCL1 transcript.
(b) RT-PCR analysis of GCL1 transcript in gcl1 mutants. RNA was isolated from 7-day-old light-grown seedlings. GCL1 gene-specific primers, FW1 and RV, were used for amplifying the full-length GCL1 transcript; FW2 and RV were used for amplifying the partial GCL1 transcript. The expression of ACTIN2 was used as a control.
(c) gcl1 mutant plants have wild-type morphology. The plants shown are 27-days old and were grown under a 14/10-h photoperiod.
(d) Sensitivity of gcl1 mutants to abscisic acid (ABA) in the seed germination assay. The averages of three replicates ± SE are shown. ABA-insensitive mutant, abi1-1, and ABA-hypersensitive mutant, abi1-ko, were used in all ABA sensitivity assays.
(e) Sensitivity of gcl1 mutants to ABA in the early seedling development assay. The averages of three replicates ± SE are shown.

Because gcr2 mutants were reported to display defects in all major ABA responses, including seed germination, early seedling development and stomatal movement (Liu et al., 2007), we wanted to determine whether gcl1 mutants were also hyposensitive to ABA. Here we focused on the ABA inhibition of seed germination and early seedling development, scoring the percentage of seeds with radicle emergence as a measure of the ABA inhibition of seed germination, and scoring the percentage of green seedlings (green cotyledons), as a measure of ABA inhibition of early seedling development. The same assays were used by Liu et al. (2007) to assess ABA sensitivity of gcr2 mutants, and have previously been used to assay ABA sensitivity of heterotrimeric G-protein α (GPA1) and β (AGB1) subunit mutants in Arabidopsis (Pandey et al., 2006; Ullah et al., 2002).

As seed germination is affected by many intrinsic and environmental factors, for these assays we used seeds from plants grown under identical conditions, and that were harvested and stored identically. We also included in all assays well-characterized ABA-insensitive and ABA-hypersensitive mutants, in order to monitor the reliability of our seed germination and early seedling development assays. For an ABA-insensitive mutant, we used the dominant negative abi1-1 allele (Leung et al., 1997; Wu et al., 2003), and for an ABA-hypersensitive mutant, we used the loss-of-function allele of ABI1, abi1-ko (Mishra et al., 2006).

We found that, unlike the reported ABA insensitivity of gcr2 mutants (Liu et al., 2007), gcl1 mutants had wild-type sensitivity to ABA in both the ABA inhibition of seed germination (Figure 2d) and the ABA inhibition of early seedling development assays (Figure 2e). At the same time, the abi1-1 mutant was always insensitive to ABA, and the abi1-ko mutant was always hypersensitive to ABA (Figure 2d,e). Therefore, we concluded that our ABA inhibition of germination and early seedling development assays were valid, and that gcl1 mutants were unaffected in their sensitivity to ABA during seed germination and early seedling development.

gcr2 gcl1 double mutants

Because GCL1 and GCR2 are highly similar to each other at the amino acid level, and gcl1 mutants did not display responses similar to those reported for gcr2 mutants in the ABA inhibition of seed germination and early seedling development assays, we asked whether loss of GCL1 function would enhance the ABA hyposensitivity of gcr2 mutants. To test this, we generated double mutants between gcr2 and gcl1. We isolated three independent mutant alleles of GCR2, all of which were identified from the SALK T-DNA (Alonso et al., 2003) Express database. Two of these three alleles, gcr2-2 (SALK_073069) and gcr2-3 (SALK_134030) have been previously characterized by Liu et al. (2007). In the new allele, named gcr2-4 (SALK_077870), the GCR2 gene is interrupted by a T-DNA insertion within its sixth exon (Figure S4), at a similar position to that affected in the gcr2-1 allele (Liu et al., 2007). The full-length transcript of GCR2 is absent in all gcr2 alleles, but a truncated GCR2 transcript was detected in gcr2-3 and gcr2-4 alleles (Figure S4). Because gcr2-4 was the first gcr2 allele we isolated, we generated double mutants between gcl1-1 and gcr2-4. In this study, gcr2 gcl1 double mutants refer to the gcr2-4 gcl1-1 genotype. The gcr2 gcl1 double mutants displayed wild-type morphology (Figure 3a).

Figure 3.

 Double mutants of gcr2 gcl1.
(a) gcr2 gcl1 double mutants have wild-type morphology. The plants shown are 27-days old and were grown under a 14/10-h photoperiod.
(b) RT-PCR analysis of GCR2 and GCL1 transcripts in gcr2 gcl1 double mutants. RNA was isolated from 7-day-old light-grown seedlings. The expression of ACTIN2 was used as a control.
(c) Sensitivity of gcr2 gcl1 double mutants to abscisic acid (ABA) in the seed germination assay. The averages of three replicates ± SE are shown.
(d) Sensitivity of gcr2 gcl1 mutants to ABA in the early seedling development assay. The averages of three replicates ± SE are shown.

We used the same ABA-sensitivity assays described above to analyze the ABA response of gcr2 gcr1 double mutant plants, and compared this to the responses of the single mutants. For clarity, we present the data on gcr2 gcl1 double mutants in Figure 3, whereas the data on gcr2-2, gcr2-3 and gcr2-4 single mutants were presented in Figure S4. For direct comparison, we re-plotted the data of Col, gcl1-1, abi1-1 and abi1-ko from Figure 2(d,e) in Figure 3(c,d), because these assays were conducted in the same set of experiments. Surprisingly, we found that all gcr2 alleles, gcr2-2, gcr2-3 and gcr2-4, displayed near wild-type sensitivity to ABA (Figure 3 and Figure S4), as did the gcr2 gcl1 double mutants (Figure 3c,d). Our preliminary analysis of other non-ABA-related assays also did not reveal any significant phenotypes of gcr2 gcl1 double mutants (data not shown).

The expression of ABA marker genes in gcr2 and gcl1 mutants

To further analyze the ABA responses of gcr2 and gcl1 mutants, we analyzed the expression of previously characterized ABA marker genes in gcr2 and gcl1 single and double mutants. Three ABA marker genes, RD29A, KIN1 and ABI5, were used by Liu et al. (2007) to analyze the ABA responses of gcr2 mutants. That study used promoter:GUS reporter lines of these three genes to compare the expression of GUS in wild type and gcr2 mutants, and found that the GUS activity was dramatically reduced in the gcr2 mutant background upon ABA induction. Here, we used RT-PCR and quantitative real-time PCR to directly compare the transcript levels of RD29A, KIN1 and ABI5 in wild type, and gcr2 and gcl1 single and double mutants, with and without ABA induction. We found that expression of these three genes still responded to ABA induction in both gcr2 and gcl1 single and double mutants (Figure 4a). We also noticed that the transcription of RD29A, KIN1 and ABI5 was notably upregulated in the gcr2 mutant prior to ABA induction (Figure 4a). In order to quantify the difference in response to ABA between wild type and mutants, we used quantitative real-time PCR to more accurately compare the level of RD29A, KIN1 and ABI5 transcripts. ACTIN2 was used as an internal control to normalize the transcripts levels of all genes in our assays. We set the transcript levels of three genes in wild type without ABA treatment (Col-ABA) as 1.0, and calculated the relative -fold change of transcripts for these three genes in wild type with ABA treatment (Col + ABA), and in mutants with (e.g. gcr2 + ABA) or without ABA treatment (e.g. gcr2 – ABA). As shown in Figure 4(b), consistent with our in-gel RT-PCR analysis, the transcripts of these three genes in either the gcr2 or gcl1 single or double mutants were increased to similar levels as that in wild type upon ABA induction. Again, these data are in contrast to the results reported by Liu et al. (2007). Such discrepancies might result from differences in the ABA induction system used. For example, in our ABA induction assay, we used 10 μm ABA for a 2-h induction, instead of 0.3 μm ABA for a 12-h induction (Liu et al., 2007). Because all these three ABA marker genes are rapidly induced by ABA (within minutes to hours), a short time with a higher concentration of ABA induction has been widely used to assay their expression (e.g. Fujii et al., 2007; Pandey et al., 2006). The observed discrepancy may also be arise from the difference in measurement techniques, because, as discussed above, Liu et al. (2007) measured the promoter:GUS activity of these three genes, whereas we directly measured their transcript levels.

Figure 4.

 Expression of abscisic acid (ABA) marker genes in gcr2 and gcl1 single and double mutants.
(a) RT-PCR analysis of the transcript of RD29A, KIN5, ABI5, RD29B and RAB18. The expression of ACTIN2 was used as a control.
(b) Quantitative real-time PCR analysis of the transcript levels of RD29A, KIN5, ABI5, RD29B and RAB18. The expression of ACTIN2 was used as a control. All transcript levels were normalized against Col without ABA treatment (set to 1). Abbreciations: C, Col; r, gcr2-4; l, gcl1-1, rl, gcr2-4 gcl1-1.

The expression of RD29A, KIN1 and ABI5 can also be induced by stress stimuli other than ABA, whereas expression of RD29B and RAB18 has been shown to be more specifically regulated by ABA (Fujii et al., 2007; Pandey et al., 2006). We therefore extended our transcriptional analysis to these genes in order to determine whether loss-of-function alleles of GCR2 or GCL1 displayed any alterations in ABA-regulated expression of RD29B and RAB18. As was the case with RD29A, KIN1 and ABI5, we found that transcription of RD29B and RAB18 was induced by ABA to similar levels in wild type, and in gcr2 and gcl1 single and double mutants (Figure 4a,b).

Taken together, these results indicate that neither GCR2 nor GCL1 is required for normal ABA responses in the ABA inhibition of seed germination and early seedling development, or in ABA-induced gene expression.

GCR2 is not genetically coupled to GPA1

GCR2 was identified as a GPCR by Liu et al. (2007), implying that it should function in conjunction with GPA1, the sole heterotrimeric Gα in Arabidopsis. gcr2 and gpa1 mutants do not share any visible phenotypic traits (Figure 5a). To examine the putative GCR2–GPA1 relationship further, we asked whether GCR2 might be coupled with GPA1 not only physically but also genetically in the ABA response, based on the classical G-protein signaling paradigm. As GPA1 is the only Gα in Arabidopsis, if GCR2 indeed acts as a GPCR coupled by Gα, GCR2 activation by ligand binding should require a functional GPA1 protein for further downstream signal transduction. We therefore would expect to observe a similar ABA sensitivity phenotype in the gcr2 and gpa1 mutants.

Figure 5.

 Comparison of gcr2 and gpa1 mutants.
(a) Phenotype comparison between gcr2 and gpa1 mutants. The plants shown are 27-days old and were grown under a 14/10-h photoperiod.
(b) Comparison of the sensitivity of gcr2 and gpa1 mutants to abscisic acid (ABA) in the seed germination assay. The averages of three replicates ± SE are shown.
(c) Comparison of the sensitivity of gcr2 and gpa1 mutants to ABA in the early seedling development assay. The averages of three replicates ± SE are shown.
(d) Comparison of the expression of GCR2, GCL1, GCL2 and GPA1 during seed germination and early seedling development. Sterilized seeds were placed under stratification conditions for 0, 6, 12, 24 or 48 h, and were then transferred to germination conditions for 24, 48 or 72 h. The RT-PCR analysis of GCR2, GCL1, GCL2 and GPA1 transcripts is shown.

However, although gcr2 mutants displayed near wild-type sensitivity to ABA, gpa1 mutants were clearly ABA hypersensitive in seed germination and early seedling development (Figure 5b,c), as reported previously (Pandey et al., 2006; Ullah et al., 2002). These results are inconsistent with genetic coupling between GCR2 and GPA1.

The expression of ABA-induced genes in gpa1 mutants has also been described previously (Pandey et al., 2006). The transcript levels of these ABA marker genes, including ABI5, RD29A and RAB18, were found to be significantly enhanced in gpa1 mutants upon ABA induction. Taken together with the observed hypersensitivity of gpa1 mutants to ABA in the seed germination and early seedling development assays, GPA1 was interpreted as a negative modulator of ABA signaling in those processes (Assmann, 2005). The behavior of these marker genes in the gcr2 mutants, again, disfavors a model in which GCR2 (as an ABA receptor and a GPCR) is coupled with GPA1.

Finally, we wanted to examine whether GCR2 expression is correlated with seed germination. Interestingly, we found that the transcript of GCR2, GCL1 and GCL2 could be detected in stratified seeds, but 24 h after the seeds had been transferred to germination conditions, the transcript levels were dramatically reduced in each case (Figure 5d). At this time point, the radicle is penetrating through the seed coat, which marks a critical transition in seed germination. We do not know the mechanism or significance of such a dramatic reduction in the transcription of GCR2, GCL1 and GCL2 at this stage, but it is noteworthy that the level of GPA1 transcript was dramatically increased during this same stage of germination, in striking contrast to those of GCR2, GCL1 and GCL2 (Figure 5d). Our result of increase in GPA1 transcript in germinating seeds is consistent with the report by Pandey et al. (2006), who detected a dramatic increase in the levels of both GPA1 transcript and GPA1 protein in germinating seeds and young seedlings.

GCR2 is likely to be a plant homolog of bacterial lanthionine synthetase

Although our bioinformatic, genetic and molecular evidence does not support the idea that GCR2 is a GPCR, or is required for the ABA control of seed germination and early seedling development, we sought additional evidence that would shed light on the function of GCR2.

The significant similarity between the amino acid sequences of GCR2 and various prokaryotic and eukaryotic LanC proteins suggests that these proteins form an evolutionarily conserved family (Figure S2). The exact function of LANCL1 and LANCL2 is unknown. Recently, LANCL1 was identified as a glutathione binding protein in the mammalian central nervous system, and may have a regulatory role in neurodegenerative disease (Chung et al., 2007). Prokaryotic LanC enzymes catalyze the production of cyclized antimicrobial peptides (Christianson, 2006). The crystal structure of the Lactococcus lactis LanC protein, nisin cyclase (NisC), has been resolved (Li et al., 2006), and alignment of GCR2 with NisC shows that the zinc-coordinating residues of NisC, critical for cysteine cyclization, are conserved in GCR2 (Figure 6). In addition, one of the proposed catalytic bases for substrate deprotonation by NisC is also conserved in GCR2 (Figure 6). These residues, which are believed to define the enzymatic activity and thereby the identity of NisC (Li et al., 2006), are also conserved in LANCL1, LANCL2 and other NisC homologs, thus supporting the hypothesis that GCR2 is likely to be a member of the LanC protein superfamily. However, because gcr2 and gcr1 single and double mutants do not display any obvious phenotypes (Figures 2, 3 and Figure S4), the exact function of GCR2 in plants remains unclear.

Figure 6.

 Protein sequence alignment of GCR2 with nisin cyclase (NisC).
The sequence alignment was generated by the ClustalW multiple alignment of the BioEdit Sequence Alignment Editor (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). Amino acids that are identical or similar are shaded with black or gray, respectively. Helical and sheet regions are indicated according to the crystal structure of NisC (Li et al., 2006). The conserved zinc ligands are indicated by a star under the residue. The proposed catalytic base for substrate deprotonation is indicated by a triangle under the residue.

Discussion

We report here on the characterization of mutant alleles of GCR2 and one of its homologs, GCL1, in the context of ABA inhibition of seed germination and early seedling development. We provide genetic evidence that neither GCR2 nor GCL1 is required for the ABA inhibition of seed germination and early seedling development, or for ABA induction of the expression of some well-characterized ABA-regulated genes. We could find no evidence that GCR2 is genetically coupled by the sole Arabidopsis heterotrimeric Gα, GPA1. Furthermore, we provide bioinformatic evidence that GCR2 is not a predicted 7TM protein, and is therefore unlikely to be an authentic member of the GPCR family. These results are in contradiction to those reported by Liu et al. (2007).

GCR2 is not a 7TM GPCR

By using the DAS transmembrane segment prediction system, Liu et al. (2007) predicted GCR2 to be a 7TM protein, and provided evidence that GCR2 is a membrane-associated protein. However, our own bioinformatics analysis indicates that GCR2 is unlikely to be a GPCR (Figure S3). Our finding is consistent with an independent study by Johnston et al. (2007) who also did not predict GCR2 as a transmembrane protein by using multiple transmembrane prediction systems. In addition, the ease of solublization of GCR2 protein, as reported by Liu et al. (2007), also suggests that GCR2 is unlikely to be an integral membrane protein. This is a critical result, because the 7TM domain arrangement is the structural hallmark of GPCRs.

GCR2 is not genetically coupled to GPA1

Although the evidence for physical coupling of GCR2 by GPA1 appeared to be compelling (Liu et al., 2007), here we show genetically that GCR2 does not appear to be coupled by GPA1 in the context of ABA inhibition of seed germination and early seedling development. First, gcr2 mutants have wild-type (this study) or hyposensitive (reported by Liu et al., 2007) ABA sensitivity phenotypes in seed germination and early seedling development, whereas gpa1 mutants are hypersensitive to ABA. In the classical G-protein signaling paradigm, ligand-bound GPCR activates Gα, and, therefore, direct coupling of GCR2 (as a GPCR) by GPA1 (the sole Gα) typically results in a similar response of the mutants to the ligand (ABA). An opposite (reported by Liu et al., 2007) or inconsistent (this study) ABA response observed between gcr2 and gpa1 mutants is inconsistent with the GPCR–Gα coupling model.

Second, the G-protein repertoire is much simpler in plants (reviewed by Jones and Assmann, 2004; Temple and Jones, 2007). There is also only one canonical G-protein β subunit (AGB1) and two G-protein γ subunits (AGG1 and AGG2) in the genome of Arabidopsis. agb1, agg1 and agg2 mutants are also hypersensitive to ABA or high concentrations of glucose in the seed germination and early seedling development assays (Pandey et al., 2006; Trusov et al., 2007). Furthermore, a putative effector of Gα, AtPirin1, is also found to be a negative regulator of ABA signaling in the ABA inhibition of seed germination (Lapik and Kaufman, 2003). Therefore, the Arabidopsis heterotrimeric G proteins consistently behave as negative modulators for ABA signaling during seed germination and early seedling development. In light of the present evidence, it is unclear how GCR2 could function through GPA1 in these processes.

Third, the analysis of the RGS mutant in Arabidopsis also provides little support for genetic coupling of GCR2 by GPA1 (Chen et al., 2006), assuming a classical G-protein signaling pathway is operating in plant cells. There is a single gene encoding the regulator of G-protein signaling (RGS) in Arabidopsis (Chen et al., 2003). The RGS protein typically functions as a negative regulator of G-protein signaling by accelerating the intrinsic GTPase of Gα, which has been proved in plant cells (Chen et al., 2003). It has been reported that the loss-of-function alleles of the sole RGS1 gene in Arabidopsis, AtRGS1, are hyposensitive to ABA (Chen et al., 2006). It is not clear how an activator (GPCR; here GCR2) and an inhibitor (RGS; here AtRGS1) of GPA1 (Gα) signaling could share a similar mutant phenotype in the same response pathway (here the ABA response) if gcr2 mutants are hyposensitive to ABA, as reported by Liu et al. (2007).

Finally, the expression of the ABA marker gene was shown to be dramatically downregulated in gcr2 mutants by Liu et al. (2007). Although we could not detect such dramatic downregulation of these genes in gcr2 and gcl1 single or double mutants, previous experiments indicated that the transcription of these genes was upregulated by ABA in gpa1 mutants (Pandey et al., 2006). Again, these results do not support a genetic coupling of GCR2 and GPA1 in regulating ABA-induced gene expression. Furthermore, the Correlated Gene Search tool of the Platform for RIKEN Metabolomics (http://prime.psc.riken.jp) failed to detect a single gene of which expression is co-expressed with GPA1 and also co-expressed with GCR2, or vice versa (Figure S5a), indicating the lack of correlation between the biological processes associated with GCR2 and GPA1 functions. As a control, in the Correlated Gene Search tool we used the heterotrimeric G-protein β subunit (AGB1) because no GPCR has been unequivocally identified in plants. This demonstrated that a significant number of genes are co-expressed with both GPA1 and AGB1 (Figure S5b,c).

In the guard cells, gcr2 mutants were shown to be insensitive to ABA in both ABA-induced stomata closing and ABA-inhibited stomata opening (Liu et al., 2007). On the other hand, gpa1 mutants have been previously shown to be insensitive to ABA in ABA-inhibited stomatal opening but had wild-type sensitivity to ABA in ABA-induced stomata closing (Mishra et al., 2006; Wang et al., 2001). A genetic coupling of GCR2 by GPA1 in the ABA response in guard cells is worth further investigation.

GCR2 is not required for ABA response in seed germination and early seedling development

Liu et al. (2007) proposed GCR2 as an ABA receptor. We have undertaken a thorough genetic characterization of GCR2 and of one of its two homologs, GCL1, in Arabidopsis, focusing on the ABA response in seed germination and early seedling development. Contrary to expectations, we found that loss-of-function alleles of either GCR2 or GCL1 (and the double mutant) did not differ from wild type in their response to ABA in the seed germination and early seedling development assays (Figures 2, 3 and Figure S4). Consistent with this, the expression of ABA marker genes was not significantly reduced in either gcr2 or gcl1 single mutants, or the double mutant, in response to ABA (Figure 4).

One possible explanation for these observations is that the ABA insensitivity may not be manifested until the function of GCL2 is also lost. At the protein sequence level, GCR2 is more similar to GCL2 (79% similar and 64% identical) than to GCL1 (60% similar and 41% identical) (Figure 1). The other possible explanation is that GCR2 and its homolog may have distinct functions other than the mediation of ABA responses. Further studies are required to address the functional redundancy among GCR2 and its homologs.

In summary, five lines of evidence contradict the notion that GCR2 is an ABA-signaling GPCR. First, gcr2 had near wild-type sensitivity to ABA in our assays, whereas gpa1 mutants were hypersensitive to ABA. In addition, loss-of-function alleles of Arabidopsis Gβ (AGB1) and Gγ subunits (AGG1 and AGG2), and of a putative effector for GPA1, Atpirin1, were all hypersensitive to ABA or high concentrations of glucose in seed germination and early seedling development. Second, loss-of-function alleles of the sole RGS in Arabidopsis, AtRGS1, were hyposensitive to ABA. Third, GCR2 and GPA1 mutants do not share similar gene expression profiling impacts on ABA-induced genes. Fourth, GCR2 is unlikely to be a 7TM protein. Finally, GCR2 has significant sequence similarity to bacterial lanthionine synthetases, a protein family that has been shown to be neither transmembrane proteins nor GPCRs.

Because GCR2 was shown to bind both ABA and GPA1 (Liu et al., 2007), what could be the possible role of GCR2 in ABA signaling/metabolism, and what could be the relationship between GCR2 and GPA1? Although the answers for these questions are presently unknown, it is still possible that GCR2 could function as an intracellular modulator for ABA signal transduction or ABA metabolism, and/or as a modulator for G-protein signaling. However, the hunt for the first plant GPCR and its ligand, as well as the search for an authentic plasma membrane ABA receptor, is still on.

Experimental procedures

Plant materials and growth conditions

All mutants except abi1-1 are in the Arabidopsis Columbia ecotype background (Col). The abi1-1 mutant is in the Landsberg erecta (Ler) ecotype background (Leung et al., 1997; Wu et al., 2003). The gpa1-4 mutant has been reported previously (Jones et al., 2003), and the abi1-ko mutant was described by Mishra et al. (2006).

Plants were germinated and grown in 5 × 5-cm pots containing a moistened 1 : 3 mixture of Sunshine Mix #1 (Sun Gro Horticulture Canada Ltd., http://www.sungro.com) and Metro-Mix 220 (W.R. Grace & Co., http://www.grace.com) with a 14/10-h photoperiod at approximately 120 μmol m−2 sec−1 at 23°C, unless specified elsewhere.

Isolation of insertion mutants of GCL1

A T-DNA insertion mutant allele of GCL1 (At5g65280), SALK_135551, was identified from the SALK T-DNA Express database (http://signal.salk.edu/cgi-bin/tdnaexpress). The insertion was confirmed by PCR and sequencing using GCL1-specific primers (5′-ATGTCGTCGTCGGTGGAT-3′ or 5′-TTATATCTCATAACCAGGA-3′) and a T-DNA specific primer JMLB1 (5′-GGCAATCAGCTGTTGCCCGTCTCACTGGTG-3′), and the mutant allele was designated as gcl1-1. PCR genotyping and sequencing results revealed that a tandem T-DNA with two outward facing left borders (LBs) was inserted in the fourth exon of the GCL1 gene in the gcl1-1 allele (Figure 2 and Figure S6).

A transposon insertion mutant allele of GCL1, SM_3_40100, was identified through searching the Exon Trapping Insert Consortium (EXOTIC) database (http://www.jic.bbsrc.ac.uk/science/cdb/exotic/index.htm). This mutant was designated as gcl1-2. The insertion was confirmed by PCR and sequencing using GCL1-specific primers and the transposon element-specific primer (5′-TACGAATAAGAGCGTCCATTTTAGAGTGA-3′). PCR genotyping and sequencing results revealed that the transposon insertion site is in the first exon of the GCL1 gene in the gcl1-2 allele (Figure 2 and Figure S6). Loss of detectable full-length GCL1 transcript in the gcl1-1 and gcl1-2 mutant was verified by RT-PCR.

Isolation of insertion mutants of GCR2

Three independent T-DNA insertion mutant alleles of GCR2 (At1g52920) were identified from the SALK T-DNA Express database, and were obtained from the Arabidopsis Biological Resources Center (Ohio State University, Columbus, OH, USA). Two of these three alleles, gcr2-2 (SALK_073069) and gcr2-3 (SALK_134030) have been reported previously by Liu et al. (2007). The other allele, named gcr2-4 (SALK_077870), has the T-DNA inserted within its sixth exon. The insertion site in each allele was examined by PCR using GCR2-specific primers (5′-ATGCCGGAGTTTGTACCG-3′ and 5′-TTAGAGTTCATAACCTGG-3′) and a T-DNA specific primer JMLB1, and confirmed by sequencing. PCR genotyping and sequencing results revealed that a tandem T-DNA with two outward facing LBs was inserted in the GCR2 gene in each of these three alleles (Figures S4 and S7). The loss of detectable full-length GCR2 transcript in each insertion line was verified by RT-PCR.

Initially, we did not pursue gcr2-1 (SALK_041449) because the sequences associated with this allele provided by the SALK T-DNA Express database are very short, and they do not hit the GCR2 gene. Instead, this allele (SALK_041449) was annotated as an insertional allele for gene locus At1g52910, but not At1g52920 (GCR2), in the SALK T-DNA Express database. Nonetheless, Liu et al. (2007) showed that the full-length transcript of GCR2 was also undetectable in the gcr2-1 allele, and they used this allele for their germination and post-germination studies, and for generating double mutants with the loss-of-function allele of the Arabidopsis heterotrimeric Gα. Therefore, for completeness, we ordered the SALK_041449 line. Our PCR genotyping and sequencing results indicated that a T-DNA was inserted in the sixth exon of the GCR2 gene in the gcr2-1 allele (Figure S8). Our RT-PCR results indicated that the full-length GCR2 transcript was undetectable in the gcr2-1 allele (Figure S8).

Generation of the gcr2 gcl1 double mutant

The gcr2-4 gcl1-1 double mutant was generated by crossing gcr2-4 with gcl1-1, and was isolated in the F2 progeny by PCR genotyping. Subsequently, the double mutant was verified by RT-PCR analysis. For simplicity, the gcr2 gcl1 double mutant nomenclature in this report refers specifically to the gcr2-4 gcl1-1 double mutant.

ABA inhibition of seed germination and early seedling development assays

Wild-type and mutant seeds from matched lots were surface-sterilized and sown on MS/G plates consisting of half Murashige & Skoog (MS) basal medium supplemented with vitamins (Plantmedia, http://www.plantmedia.com), 1% (w/v) sucrose, 0.6% (w/v) phytoagar (Plantmedia), pH adjusted to 5.7 with 1 N KOH, and with different concentrations of ABA (0, 0.5, 1, 2, 5 and 10 μm). Imbibed seeds were cold-treated at 4°C in the dark for 2 days, and then moved to 23°C, with a 14/10-h photoperiod (120 μmol m−2 sec−1). Forty-eight hours later, the percentage of seed germination was scored. Germination was defined as an obvious protrusion of the radicle through the seed coat. Ten days later, the percentage of green seedlings was scored, based on the presence of obvious green cotyledons. Each experiment was repeated at least three times. In all assays, we used the well-characterized ABA-insensitive mutant, abi1-1 (Leung et al., 1997; Wu et al., 2003), and ABA-hypersensitive mutant, abi1-ko (Mishra et al., 2006), as controls to monitor and verify our seed germination and early seedling development assays. abi1-1 is in the Ler ectopic background, but as it is still ABA insensitive when compared with Col wild type, we used only Col as the wild-type control.

RT-PCR

For the analysis of GCR2, GCL1, GCL2 and GPA1 transcripts during seed germination, wild-type (Col) seeds were surface-sterilized and stratified at 4°C in the dark. Seeds sampled immediately after sterilization and prior to stratification were used as controls (0 time point); these had been in liquid for approximately 30 min during surface sterilization. Subsequently, seeds were sampled 6, 12, 24 and 48 h after stratification, and were directly placed into liquid nitrogen. Forty-eight hours after being placed under stratification conditions, imbibed seeds were moved to germination conditions (23°C, 14/10-h photoperiod at 120 μmol m−2 sec−1) where they were sown on MS/G plates covered with one layer of sterilized filter paper for easy sampling. Seeds/germinating seeds were sampled 24, 48 and 72 h later. Total RNA was isolated from seeds/germinating seeds using the TRIzol reagent (Invitrogen, http://www.invitrogen.com). cDNA was synthesized from 1 μg of total RNA by oligo(dT)20-primed reverse transcription, using THERMOSCRIPT RT (Invitrogen). GCR2-specific primers (5′-GGAAGATTTATCCGGAGAAGAAGAAACTGT-3′ and 5′-CAGAGCTTGTGTTGGATCATTCATGTCG-3′), GCL1-specific primers (5′-ATGTCGTCGTCGGTGGAT-3′ and 5′-TTATATCTCATAACCAGGA-3′) and GCL2-specific primers (5′-ATCCGCGAAGTTGCTCAGGA-3′ and 5′-TTAAAGCTCGTAACCTGGGAGAAG-3′) were used to amplify the transcript of these three genes. ACTIN2 (ACT2) (amplified by primers 5′-CCAGAAGGATGCATATGTTGGTGA-3′ and 5′-GAGGAGCCTCGGTAAGAAGA-3′) was used as a control in PCR reactions.

Analysis of ABA-induced gene expression

Light-grown seedlings (4.5 days old) were used for ABA-induction experiments. Seedlings grown vertically on MS/G plates were moved to MS/G liquid medium without phytoagar, and grown for another 2 h prior to ABA induction in an orbital shaking incubator. ABA was added at 10 μm for 2 h, and seedlings were immediately sampled into liquid nitrogen. Total RNA was isolated from ABA-treated and untreated seedlings, and cDNA was synthesized from each sample as described above. RD29A-specific primers (5′-CAAACAGAGGAACCACCACTCAA-3′ and 5′-CTGGTGCATCGATCACTTCAGGT-3′), KIN1-specific primers (5′-GCCCACATCTCTTCTCATCATCACTAA-3′ and 5′-TTATAACTCCCAAATTTGACCCGAATC-3′), ABI5-specific primers (5′-GAACATCCCACTAATCCTAAACCT-3′ and 5′-AGCAAACACCTGCCTGAACT-3′), RD29B-specific primers (5′-CACCAGAACTATCTCGTCCCAA-3′ and 5′-GCTTTGAGGCAACGACGTTCT-3′) and RAB18-specific primers (5′-TCCAGCAGCAGTATGACGAGTA-3′ and 5′-CCAGTTCCAAAGCCTTCAGTC-3′) were used for RT-PCR and real-time PCR analyses. ACTIN2 was used as an internal control. The quantitative real-time PCR was performed using the MJ MiniOpticon real-time PCR system (Bio-Rad, http://www.biorad.com) and IQ SYBR Green Supermix (Bio-Rad).

Acknowledgements

We thank the Arabidopsis Biological Resources Center and the Exon Trapping Insert Consortium for providing Arabidopsis gcr2, gcl1 and abi1 mutant seeds. Work in JGC’s lab is supported by grants from the Natural Sciences and Engineering Research Council of Canada, the Canada Foundation for Innovation, the British Columbia Ministry of Advanced Education and the University of British Columbia. YG is supported by a scholarship from the China Scholarship Council.

Upon request, all novel materials described in this article will be made available in a timely manner for non-commercial research purposes.

Ancillary