Heterotrimeric G protein subunits differentially respond to endoplasmic reticulum stress in Arabidopsis

Abstract

Canonical heterotrimeric G proteins in eukaryotes are major components that localize at plasma membrane and transmit extracellular stimuli into the cell. Genome of a seed plant Arabidopsis thaliana encodes at least one Gα (GPA1), one Gβ (AGB1), and 3 Gγ (AGG1, AGG2 and AGG3) subunits. The loss-of-function mutations of G protein subunit(s) cause multiple defects in development as well as biotic and abiotic stress responses. However, it remains elusive how these subunits differentially express these defects. Here, we report that Arabidopsis heterotrimeric G protein subunits differentially respond to the endoplasmic reticulum (ER) stress. An isolated homozygous mutant of AGB1, agb1-3, was more sensitive to the tunicamycin-induced ER stress compared to the wild type and the other loss-of-function mutants of G protein subunits. Moreover, ER stress responsive genes were highly expressed in the agb1-3 plant. Our results indicate that AGB1 positively contributes to ER stress tolerance in Arabidopsis.

Eukaryotic cells sense extracellular stimuli via receptor proteins localized at plasma membrane to transmit signals to intracellular components such as ion channels and enzymes.¹ Among them, the heterotrimeric guanine nucleotide-binding proteins (G proteins) complex has been most extensively studied because of its indispensability in a huge number of cellular biological processes including nerve system, visual sense, and olfactory sense. Indeed, human genome reveals the importance and diversity of G proteins complex by encoding 23 Gα, 5 Gβ, 12 Gγ subunits and more than 800 G protein-coupled receptors (GPCRs).^2-4 By contrast, in Arabidopsis, limited number of G protein components are identified to date; one Gα (GPA1; At2g26300), one Gβ (AGB1; At4g34460), and 3 Gγ (AGG1; At3g63420, AGG2; At3g22942, and AGG3; At5g20635).^5-9 This does not necessarily underestimate the importance of heterotrimeric G protein-mediated signal transduction in plants; knocking out of sole gene AGB1 causes various developmental defects such as lateral root formations and abnormal organ shapes in leaves and flowers.^10,11 Intriguingly, AGB1 is involved in unfolded protein response (UPR),^12,13 a pathway in so-called endoplasmic reticulum (ER) quality control to cope with ER stresses.^14,15. The molecular mechanism of ER quality control is evolutionarily well conserved among eukaryotic cells. Once aberrant proteins are accumulated in the ER, the cell recognizes misfolded/unfolded proteins and activates several branching pathways including UPR and ER-associated degradation (ERAD) either to restore protein folding or to eliminate aberrantly configured proteins through the ubiquitin-proteasome degradation pathway.^16-19 Thus, the ER quality control monitors conditions of ER protein folding at the ER. How plasma membrane-localized G protein function affects ER homeostasis is an open question.

Although AGB1 is involved in unfolded protein response (UPR), 2 contradictory results are reported regarding the sensitivity of the agb1 mutant to the ER stress.^12,13 Therefore, we first examined sensitivity of the agb1 plant against the ER stress caused by tunicamycin, which inhibits protein N-glycosylation pathway and consequently causes accumulation of misfolded N-glycosylated proteins in the ER. We isolated T-DNA homozygous mutant of agb1-3, and confirmed the position of T-DNA insertion in the 4^th exon (Fig. 1A). When we grew the wild-type and the agb1-3 plants on MS medium supplemented with different concentrations of tunicamycin for 2 weeks, the agb1-3 plants were obviously smaller than the wild-type plants on MS media containing 50 ng/ml tunicamycin (Fig. 1B). To statistically analyze the phenotype, we measured the weight of whole seedlings of the wild type and the agb1-3 treated with 50 ng/ml tunicamycin (Fig. 1C). The agb1-3 seedlings were significantly lighter than the wild type (P<0.0001, ****), which is consistent with the report by Chen et al. (2012).¹³ Thus, we conclude that AGB1 positively contributes to ER stress tolerance. We further examined the ER stress sensitivities of the mutants of other G protein subunits. T-DNA insertion lines of the sole Gα subunit, GPA1 and 2 out of 3 Gγ subunits were available in Arabidopsis Biological Resource Center (ABRC, OH). Homozygous mutants were isolated, and T-DNA positions were confirmed as shown in Figure 1A. Unlike the agb1-3, none of them showed significant differences compared to the wild type in the tunicamycin-induced ER stress (Fig. 1B and 1C). Since only single copy of gene has been identified for Gα and Gβ in the Arabidopsis genome, our results suggest that AGB1 unlikely functions as the heterotrimeric G protein complex for the ER stress tolerance but yet unknown Gα subunits may do. With regards to Gγ subunits, although single mutants of agg2-3 and agg3-1 showed no phenotypes, the possibility that 3 Gγ isoforms function redundantly for the ER stress tolerance cannot be ruled out. Indeed, the triple mutant of Gγ subunits, agg1 agg2 agg3, showed similar developmental defects as those in the agb1-3 single mutant.²⁰

Figure 1.

G protein β subunit is sensitive to the ER stress. (A) Schematic representation of gene structure and T-DNA position of mutants for GPA1 (At2g26300), AGB1 (At4g34460), AGG2 (At3g22942), and AGG3 (At5g20635). Gray boxes represent exons, and the positions of the T-DNA insertion are shown as triangle. The following mutant seeds were obtained from the Arabidopsis Biological Resource Center (ABRC, OH): gpa1-1 (SALK_115996), agb1-3 (SALK_061896), agg2-3 (GABI_673C04) and agg3-1 (SALK_018024). Homozygous plants were isolated by PCR-based genotyping with gene-specific primers and T-DNA-specific primers as listed in the Supplemental Table 1. (B) WT (Col-0), gpa1-1, agb1-3, agg2-3, and agg3-1 plants were grown on 1/2 MS plates containing 0, 12.5, 25, and 50 ng/ml tunicamycin (TM) for 14 days, and photographed. Dimethyl sulfoxide (DMSO) was used as mock treatment. (C) Fresh weight of the seedlings treated with 50 ng/ml tunicamycin shown in (B) was measured individually, and shown as a boxplot (n = 15). Student's t-test was performed for statistical analysis. The asterisks indicate significance between WT (Col-0) and the agb1-3 (P < 0.0001, ****).

To investigate the differential gene expression of Arabidopsis G protein subunits, we used the publicly available gene expression database (GENEVESTIGATOR; www.genevestigator.com). We first identified differential expression patterns at different stage of development among the 4 genes, GPA1, AGB1, AGG2, and AGG3 (Fig. 2A); GPA1 expression was lower in vegetative growth than seedlings or reproductive growth. This expression pattern correlated well with that of AGB1 although expression level of AGB1 was much higher than that of GPA1. The expression of 2 Gγ isoforms, AGG2 and AGG3, is relatively ubiquitous and low overall. Next, we examined tissue-specific gene expression pattern as shown by the heat map (Fig. 2B). In agreement with Figure 2A, AGB1 was expressed higher in seedlings than any other tissues, whereas GPA1 was expressed ubiquitously. Regarding AGG2 and AGG3, tissue-specific gene expression pattern was rather complementary; AGG2 was expressed higher in seedlings, shoot, and rosette but lower in raceme and flower, while AGG3 was expressed lower in seedlings, shoot, and rosette but higher in raceme and flower.

Figure 2.

Gene expression pattern of GPA1, AGB1, AGG2, and AGG3. (A) Developmental stage-specific expression patterns of GPA1, AGB1, AGG2, and AGG3. Seedlings, rosette leaves, and floral organs are sequentially marked from left to right. “HIGH,” “MEDIUM,” and “LOW” expression were calculated by microarray assay. The number of samples indicates microarray gene expression data collected by GENEVESTIGATOR (www.genevestigator.com). (B) Heat map of tissue-specific expression pattern of GPA1, AGB1, AGG2, and AGG3. Data were analyzed with GENEVESTIGATOR (www.genevestigator.com).

Our results showed that only knocking out of Gβ subunit affected ER stress tolerance. To further study this more rigorously, we performed genetic complementation of agb1-3 by stable transformation of a transgene that expresses AGB1 by its own promoter (agb1-3 ProAGB1:AGB1). We firstly confirmed a genotype and an expression level of AGB1 in the agb1-3 ProAGB1:AGB1 plant. To identify agb1-3 ProAGB1:AGB1 plants, we designed primers to distinguish between endogenous genomic AGB1 gene (“gAGB1” in Fig. 3A) and transgene of AGB1 (“Transgene” in Fig. 3A). As expected, a fragment for transgene but not genomic AGB1 was amplified for the agb1-3 ProAGB1:AGB1 plant (Fig. 3A). In addition, RT-PCR analysis revealed similar expression level of AGB1 between agb1-3 ProAGB1:AGB1 and wild-type plants (Fig. 3B), as well as no detectable transcripts of the AGB1 gene in agb1-3, an indication of agb1-3 as a null mutant. We next observed phenotype of agb1-3 ProAGB1:AGB1 to examine whether the transgene of AGB1 is functional in complementing the developmental defects reported previously.¹¹ The agb1-3 plant showed round leaf shape and flat top silique as shown in Figure 3C and D, respectively. Both organ shape phenotypes were complemented in the agb1-3 ProAGB1:AGB1 plant (Fig. 3C and D). These results indicate that transgene of AGB1 in the agb1-3 ProAGB1:AGB1 plant may be functional. We next examined the ER stress sensitivity of the agb1-3 ProAGB1:AGB1 transgenic plants. Upon tunicamycin treatment, the agb1-3 ProAGB1:AGB1 plant showed the similar extent of resistance to the wild type (Fig. 3E). This was further confirmed statistically by weighing the fresh weight of the plants (Fig. 3F). Thus, we concluded that the ER-stress phenotype in the agb1–3 is due to the loss of function of AGB1.

Figure 3.

The growth defects observed in the agb1-3 were complemented by transforming genomic AGB1 gene. (A) PCR-based genotyping of the wild type (WT), agb1-3 homozygous plants (agb1-3) and genetically complemented agb1-3 homozygous plants (agb1-3 ProAGB1:AGB1) using the specific primers listed in the Supplemental Table 1. For ProAGB1:AGB1, a 3.6 kb of the genomic sequence of AGB1 (ProAGB1:AGB1) was amplified by PCR with the primers KK228 and KK225 listed in the Supplemental Table 1. The fragment was cloned into the pENTR/D-TOPO plasmid vector (Invitrogen) to obtain pCC83, which was then recombined into the pBGW destination vector with the use of LR Clonase (Invitrogen) to obtain pCC86. The pCC86 plasmid was transformed into the agb1-3/- plants by Agrobacterium tumefaciens-mediated transformation. Sixteen transformants were selected on soil by spraying with 0.1% Basta solution. Resistant plants were genotyped and line No. 6 was used in this study. (B) RT-PCR analysis of the wild type (WT), agb1-3, and agb1-3 ProAGB1:AGB1 plants. Total RNA was extracted from 7-day-old seedlings grown on 1/2 MS media plates using RNeasy mini kit (Qiagen) with DNase treatment, and cDNA was synthesized by the Superscript III First-Strand Synthesis system (Invitrogen). The Actin (ACT) expression levels were used as a control. The primers used here are listed in the Supplemental Table 1. (C) 3-week-old plants of the wild type (WT), agb1-3, and agb1-3 ProAGB1:AGB1 grown on soil were photographed. (D) Tip of siliques in 5-week-old plants of the wild type (WT), agb1-3, and agb1-3 ProAGB1:AGB1 were photographed. The scale bar is 0.1 mm. (E) The wild type (WT), agb1-3, and agb1-3 ProAGB1:AGB1 plants were grown on 1/2 MS plates containing 0, 25, and 50 ng/ml tunicamycin (TM) for 14 days, and photographed. (F) Fresh weight of the seedlings treated with 50 ng/ml tunicamycin shown in (E) was measured individually, and shown as a boxplot (n = 25). Student's t-test was performed for statistical analysis. The asterisks indicate significance between WT and the agb1-3 (P < 0.01, **).

The UPR pathway is activated upon ER stress through the ER-membrane-bounded transcriptional factors such as bZIP17, bZIP28, and bZIP60, which sense accumulation of unfolded proteins in the ER and subsequently transmit signals to the nucleus to activate ER stress responsive genes.^21,22 Approximately three hundreds genes are upregulated upon the tunicamycin-induced ER stress in Arabidopsis.²³ To investigate whether phenotype of agb1-3 upon the tunicamycin treatment is relevant to UPR pathway, we compared expression profiles of key genes in the wild-type and the agb1-3 plants. We selected binding protein 3 (BiP3), binding protein 1 and 2 (BiP1/2), calreticulin (CRT), and calnexin (CNX) as marker genes. All genes encode ER-resident proteins that are known to be transcriptionally upregulated during the ER stress.²³ We treated the 7-day-old seedlings with tunicamycin for 0, 2, and 5 hours, and examined expression level of the aforementioned genes by quantitative RT-PCR analysis. We repeated experiments 4 times independently, and analyzed data all together (Fig. 4). Each result was also shown individually in the Supplemental Figure 1. The expression of ER stress responsive genes were similarly upregulated both in the agb1-3 and wild-type plants, and no significant differences were seen in the expression profiling for all genes tested here between the wild type and the agb1-3 (Fig. 4A, C, E and G). However, we noted that expression levels of all genes were higher in the agb1-3 compared to the wild type without tunicamycin treatment (Fig. 4B, D, F and H). These results indicate that lack of AGB1 increases expression level of ER-stress responsive genes. In conclusion, we showed that Gβ subunit is the bottleneck in the tunicamycin-induced ER stress by comparative phenotype observation of different G protein subunit mutants and functional complementation of agb1-3 mutant.

Figure 4.

The ER stress-responsive genes showed higher expression level in the agb1-3 plants. (A) Quantitative RT-PCR analysis of binding protein 3 (BiP3) transcripts in the wild type (WT) and agb1-3 plants in response to tunicamycin. Total RNA was extracted from the 7-day-old plants treated with 5 μg/ml tunicamycin at each time point (0, 2, and 5 hours for the treatment). Dimethyl sulfoxide (DMSO) was used as mock treatment. The expression of the wild type at 0h was set to 1. (B) Magnified expression level of BiP3 at time point 0 in (A). (C) Quantitative RT-PCR analysis of binding protein 1 and 2 (BiP1/2) transcripts in the wild-type (WT) and agb1-3 plants in response to tunicamycin as described in (A). (D) Magnified expression level of BiP1/2 at 0 h in (C). (E) Quantitative RT-PCR analysis of calreticulin (CRT) transcripts in the wild-type (WT) and agb1-3 plants in response to tunicamycin as described in (A). (F) Magnified expression level of CRT at 0 h in (E). (G) Quantitative RT-PCR analysis of calnexin (CNX) transcripts in the wild-type (WT) and agb1-3 plants in response to tunicamycin as described in (A). (H) Magnified expression level of CNX at 0 h in (G). Total RNA was extracted as described in Fig. 3 B. Quantitative RT-PCR analysis was performed using the 7500 Real-Time PCR systems (Applied Biosystems). The comparative threshold cycle method was used to determine relative gene expression, with the expression of ACT served as an internal control. Data are mean ±SD from 4 biological replicates, with 3 technical replicates. Student's t-test was performed for statistical analysis. The asterisks indicate significance between WT and the agb1-3 (P<0.001, ***; P<0.01, **; P<0.05, *). The primers sets for quantitative RT-PCR were listed in the Supplemental Table 1.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Yuki Nakamura (IPMB, Academia Sinica) for critical reading of the article, Kenji Inaba (IMRAM, Tohoku University) for stimulating discussion, and Chia-En Chen (IPMB, Academia Sinica) for technical assistance of molecular cloning.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.

Funding

This research was supported by the core budget provided by IPMB, Academia Sinica, and by the JSPS KAKENHI (grant no. 25440123) to K.K. and T.I. The authors declare no competing financial interests.