Introduction

Heterotrimeric G proteins are known to function as messengers involved in numerous signal transduction pathways following perception by receptors on cell surfaces. Such proteins are composed of subunits, namely α, β and γ, respectively. The Gα subunit contains a Ras-like domain with a GDP/GTP nucleotide-binding site and GTP hydrolase activity 1. The human genome contains about 23 Gα genes that belong to four subfamilies, Gαs, Gαi, Gαq and Gα12/13, according to their functional and sequence attributes. In contrast, plant genomes contain a single copy of the Gα coding gene 2. Nevertheless, recent studies in Arabidopsis have shown that plant G protein signaling is important to many fundamental processes, such as cell proliferation, hormone perception and ion-channel regulation 3, 4. In rice, the G protein α subunit was first cloned and characterized in 1995 and designated RGA1 (rice G protein α subunit 1) 5, 6. The null mutant, Daikoku d1, of the G protein α subunit displays dwarfism and small-sized seeds 7. Similar results were observed in transgenic antisense RGA plants that produce little or no mRNA of RGA 8. Investigation of dose response to gibberellic acid (GA3) revealed that the d1 mutant showed defective induction of α-amylase activity in the aleurone layer as compared to wild-type plants. In addition, RNA blot analysis showed that GA-inducible genes, OsMYB, Ca2+-ATPase and Ramy1, had a weaker response to GA in the d1 mutant than that in control plants 9. Epistatic analysis showed that slr (slender rice), caused by mutation in a crucial regulator in GA signaling, is epistatic to d1 9. Undoubtedly, this evidence revealed that GA signaling is affected in the d1 mutant. However, the precise underlying mechanism remains to be elucidated.

Recently, transgenic plants containing a constitutively active form of RGA were generated via site-directed mutagenesis at position 223, which caused the substitution of glutamine by leucine (Q223L), and showed GTP-binding activity but not GTPase activity. The transformants harboring RGA (Q223L) did not show a slender phenotype but, rather, were essentially normal cultivars, indicating that heterotrimeric G proteins do not directly regulate GA signaling in rice 10. In contrast, there have been few reports on the function of the G protein in brassinosteroid (BR) signaling pathways. OsBLE2 is a novel brassinolide-enhanced gene in rice. The transcript of OsBLE2 was found to be slightly downregulated and less sensitive to BL treatment in the d1 mutant as compared with control plants, and investigators speculated that heterotrimeric G protein might be involved in BL signaling 11. In Arabidopsis, mutants of G protein α subunit were less sensitive to brassinolide with respect to stimulated germination and inhibition of vegetative growth 12. However, little direct evidence exists about whether the G protein is involved in BR signaling in rice.

In BR signal transduction, BRI1 encodes a plasma membrane-localized leucine-rich repeat receptor-like kinase 13. BAK1, BRI1-associated receptor kinase, plays a critical role in BR signal perception via dimerization with BRI1 14, 15. In the present study, we used root elongation inhibition assay, lamina joint inclination assay and coleoptile elongation promotion assay to show that the d1 mutant was less sensitive to 24-epibrassinosteroid (24-epiBL), a synthetic brassinosteroid that plays a wide variety of roles in plant growth and development. However, d1 mutant plants did not display a canonical de-etiolated phenotype with no elongated mesocotyls, as did the d61-1 mutant (a well-known BR-insensitive rice mutant caused by OsBRI1 mutation) in the dark. In addition, 24-epiBL treatment also suppressed the expression of RGA in a dose-dependent manner. Our direct evidence supports that RGA is involved in affecting brassinosteroid response in rice. The result might be beneficial to elucidate the molecular mechanisms of G protein signaling and also provide a novel perspective to understand BR signaling in higher plants.

Materials and Methods

Plant materials and growth conditions

Seeds of wild-type rice (Oryza sativa ssp. japonica Nipponbare), d1 mutant 7 and d61-1 mutant 16 were used. The seeds were germinated at 30 °C in darkness for 2 days, and then transferred to agar plates containing half-strength Murashige and Skoog (MS) medium supplemented with different phytohormones. Seeds were cultured for 7 days at 25 ºC with a 16-h photoperiod for the physiological experiments.

Semi-quantitative RT-PCR

RNA was extracted for synthesis of first-strand cDNA as previously described 17. Total RNA of 2 μg was reverse transcribed with Superscript II reverse transcriptase, according to the manufacturer's instructions (Gibco BRL life Technologies, UK). First-strand cDNA was generated with use of an oligo (dT)18 primer. The first-strand cDNA of 5 μl was diluted to 20 μl. Diluted first-strand cDNA of 2 μl was used as a template in a subsequent PCR reaction. The abundance of actin gene was selected as an internal control. The specific primers (5′-TCA GAT GCC CAG TGA CAG GA-3′, 5′-TTG GTG ATC TCG GCA ACA GA-3′) were used to amplify actin. The primer for RGA-specific amplification was the same as that used by Ishikawa et al.5. PCR components consisted of 10 μl 2× GC buffer, 1 μl 10 mM dNTP, 0.5 μl LA Taq polymerase (Takara, Dalian, China), 1 μl of each primer (10 μM), and template first-strand cDNA in a 20 μl reaction volume. The PCR procedure consisted of the following steps: an initial denaturation (2 min, 94 °C), followed by 28 cycles of denaturation (45 s, 94 °C), annealing (45 s, 55 °C for RGA primers and 58 °C for actin primers) and primer extension (45 s, 72 °C). The final cycle was completed with a 10-min extension at 72 ºC, and then holding for 4 °C. The PCR products were run on 1.2% agarose gel. Images were analyzed by use of Bio-1D Microsoft (BIO-PROFIL, France, www.vilber.com).

Lamina inclination assay

Rice seeds were soaked in distilled water for 2 days in the dark at 30 °C after sterilization with 0.15% HgCl for 10 min, then transferred to half-strength MS medium and cultured at 30 °C for 7 days in dark. The second lamina joints were used for lamina inclination assay as described 18.

Results

Root elongation inhibition assay of d1 mutant

High concentrations of brassinolide can inhibit root growth in rice 16. To test the response of the d1 mutant to 24-epiBL, 2-day-old seedlings were treated with 0.1 nM to 10 μM 24-epiBL for 3 days to monitor the length of primary roots. Strikingly, the d1 mutant seedling exhibited a less sensitive phenotype as compared with control plants (Figure 1). The primary root growth curves of the d1 and d61-1 mutants (another well-known BR-insensitive rice mutant, caused by mutation OsBRI1, encoding brassinosteroids receptor in rice) were compared. The results showed both mutants with a similar growth tendency until up to 1 μM 24-epiBL treatment, which is distinct from wild-type plants (Figure 1). Primary root length was not greatly inhibited at 1 μM 24-epiBL in the d1 mutant as compared with the wild type. Increasing the concentration of 24-epiBL from 1 to 10 μM resulted in some inhibition of root length in d1 plants while little inhibition was observed for d61-1 (Figure 1), suggesting that heterotrimeric G protein α subunit functions in a different manner with OsBRI1 to affect BR response in rice.

Figure 1
figure 1

Analysis of sensitivity of d1 mutant to 24-epiBL by the root inhibition assay. Seedlings of the wild type, d61-1 mutant and d1 mutant were germinated on distilled water for 2 days and transferred to half-strength MS medium containing various concentrations of 24-epiBL. Roots were measured on day 5 after germination. Each data point represents the average root elongation of 15 seedlings. CK represents the mock-treated control.

Full size image

Sensitivity of d1 mutant to 24-epiBL by lamina inclination assay

Lamina inclination assay is a classical and specific method to determine the effect of BR in rice 18. Both BR-insensitive and -deficient rice mutants display erect leaves 16, 19, 20. BR-insensitive mutants d61-1 and d61-2 were found to be less sensitive to BL than the wild type on lamina inclination assay 16. To further identify the sensitivity of the d1 mutant to 24-epiBL, lamina inclination assay was adopted. The results revealed that the degree of inclination between the lamina and leaf sheath in d1 mutant plants was less than that of the control plants (Figure 2A). In d1 mutant plants, the bending degree was increased with the increase in 24-epiBL concentration, but the extent of inclination was much less than that of control plants. The leaf angles were about 165° in the wild type but just 109° in the d1 mutant at 100 nM 24-epiBL (Figure 2B), further confirming the lower sensitivity of the d1 mutant to BR.

Figure 2
figure 2

Comparison of 24-epiBL effect on etiolated lamina inclination in wild-type and d1 plants. (A) Dose-response to 24-epiBL of the second lamina joint in wild-type (WT) and d1 mutant seedlings. (B) Dose-response to 24-epiBL of the second lamina joint. Data represent the means of results from six seedlings. Error bars indicate SD, CK represents the mock-treated control.

Full size image

Coleoptile elongation response of d1 mutant to 24-epiBL

The rice coleoptile represents a feasible system to study hormonal signaling because its elongation is stopped at an early stage of germination 21. In the presence of BR, the coleoptile of wild-type plants elongated more than that in the absence of BR; even the foliage leaves cannot break out of the coleoptile at high concentrations of BR in light 16. These results suggested that BR is involved in the process of coleoptile elongation, and the extent of elongation can be considered reflect sensitivity to BR, since the length of elongation increased with increasing BR concentration. To further examine the sensitivity of the d1 mutant to BR, we compared the effect of BR on coleoptile elongation in the d1, d61-1 mutants and wild-type plants. The length of the coleoptile of control seedlings increased in the presence of 24-epiBL, whereas that of the d1 and d61-1 mutants was less sensitive to 24-epiBL (Figure 3).

Figure 3
figure 3

Sensitivity of coleoptile growth to 24-epiBL in d1. Seedlings of the wild type (WT) and d61-1 and d1 mutants were germinated on half-strength MS medium containing various concentrations of 24-epiBL. Coleoptiles were measured after germination for 7 days. CK represents the mock-treated control. Each data point represents the average coleoptile length of 10 seedlings.

Full size image

Morphology of d1 mutant in the dark

To test whether the d1 mutant is similar to other rice BR-related mutants that display canonical de-etiolated phenotypes such as unelongated mesocotyls and coleoptiles in the dark, we germinated d1, d61-1 and wild-type seeds in half-strength MS medium without any plant hormones under darkness. The rice coleoptile has been regarded as an index of skotomorphogenesis. As shown in Figure 4B, the coleoptile elongation of the d1 mutant in the dark was 3.25-fold of that in the light, while it was 3.28-fold for d1 and 4.6-fold for wild type, respectively. However, compared with the d61-1 mutant, the d1 mutant has normal elongated mesocotyls as do wild-type plants, which is different from d61-1 mutant with unelongated mesocotyls (Figure 4A). Taken together, we speculated that the d1 mutant showed non-canonical de-etiolated phenotypes in darkness.

Figure 4
figure 4

Morphology of d1 mutant in darkness. (A) The d1 mutant displays non-canonical de-etiolated phenotype as compared with d61-1, a mild OsBRI1 mutation. Triangle points to the position of the mesocotyl, and arrowhead points to the position of the first collar. (B) Comparison of coleoptile length in wild type (WT), and d1 and d61-1 mutants in dark and light.

Full size image

Response of d1 mutant to various plant hormones

In animals, heterotrimeric G protein mediates numerous extracellular environmental signals via coupled stimulus perception by G protein-coupled receptors (GPCRs) with numerous downstream effectors. Recent studies in Arabidopsis have shown that plant heterotrimeric G protein is involved in many fundamental processes such as cell proliferation, hormone perception and ion-channel regulation 1. Many hormone-response mutants usually display changed sensitivity to multiple plant hormones 22. We used root elongation assay to analyze whether other plant hormones can inhibit rice root elongation in the d1 mutant. Germinated seeds were cultured on half-strength MS medium containing ABA (abscisic acid, 1 μM), known to be involved in stress response; ACC (1-aminocyclopropane-1-carboxylate, 1 μM), a precursor for ethylene biosynthesis, JA (jasmonate, 1 μM), an important group of plant signaling molecules that play a pivotal role in the regulation of plant metabolism and defense against pathogens 23; IAA (indoleacetic acid, 1 μM); 24-epiBL (1 μM), which belongs to brassinosteroids and plays a wide variety of roles in plant growth and development; 6-BA (6-benzylaminopurine, 1 μM), a kind of cytokinin with high activity to promote cell division; and GA3 (50 μM), which play key roles in cell expansion. As shown in Figure 5, the d1 mutant showed wild-type-like response to IAA, JA and 6-BA, but was less sensitive to 24-epiBL and ACC.

Figure 5
figure 5

Inhibition of root elongation by plant hormone treatment in the d1 mutant. Sterilized seeds were germinated on half-strength MS medium containing various plant hormones. The concentrations of plant hormones were 1 μM, except for GA3, which was 50 μM. CK represents the mock-treated control. See text for details. Germination involved a 16 h photoperiod. Root lengths were measured on the 14th day after germination.

Full size image

Response of RGA expression to 24-epiBL

To determine whether the expression of RGA was affected by BR, sterilized seeds were cultured on half-strength medium containing various concentrations of 24-epiBL and expression of RGA was determined by semi-quantitative RT-PCR at the transcript level. As shown in Figure 6A, the RGA transcript was downregulated by 24-epiBL in a dose-dependant manner (Figure 6A and 6B). Furthermore, downregulation of the RGA transcript was observed in the 24-epiBL-treated d61-1 mutant (Figure 6A and 6B), suggesting that the hormone regulates the expression of RGA independently of OsBRI1.

Figure 6
figure 6

Effect of 24-epiBL on RGA expression at the steady-state level. 3-day-old seedlings of wild type (WT) and d61-1 mutant were harvested after treatment with various concentrations of 24-epiBL. CK represents the mock-treated control. (A) Images of RT-PCR products run on 1.2% agarose gel. (B) Relative expression of RGA in response to different concentrations of 24-epiBL.

Full size image

Discussion

In Arabidopsis, seeds of the gpa1-1 mutant with a null mutation of heterotrimeric G protein α subunit displayed 100-fold less sensitivity to GA. Ullah et al.12 further identified that GPA1 potentiation action of GA rather than directly coupling GA to downstream effectors leading to germination, and the potentiality was further shown to couple with BR. Rice d1 mutant also has been shown to have less sensitivity to GA as it is defective in GA3-induced α-amylase activity in the aleurone layer 9. Moreover, RNA blot analysis has shown three GA-inducible genes, OsMYB, Ca2+-ATPase and Ramy1, with a weaker response to GA in d1 mutant than in wild-type plant 9. Thus, GA signaling is definitely altered in d1 mutant. Transgenic plants containing a constitutively active form of RGA (Q223L) did not show a slender phenotype, but rather, were essentially normal cultivars, suggesting that heterotrimeric G proteins do not directly involve the GA signaling pathway in rice 10. Recently, oligonucleotide microarray analysis demonstrated that genes upregulated by GA or ABA had a higher expression level in wild type than in the d1 aleurone and genes downregulated by GA had a lower expression level in the wild type relative to the d1 aleurone 24. Further, Bethke et al.24 identified that although changes in transcript abundance were smaller in the d1 mutant than in wild type, the d1 mutation did not result in a decrease in sensitivity to GA at the level of transcription.

In the present study, we used root elongation inhibition assay to test the response of d1 mutant to 24-epiBL. Since d61-1 mutant is well-documented as a rice BRs insensitive mutant by Yamamuro et al. 16, it was used as a positive control in root elongation inhibition assay. Our data clearly showed that d1 mutant was less sensitive to 24-EBL compared to wild-type plants. At the same time, d61-1 mutant, the positive control plants, also showed insensitivity to 24-EBL. It suggested that the physiological evidence is convincible. In addition, lamina joint bending assay and coleoptile elongation analysis both further proved that d1 mutant is less sensitive to 24-EBL. Moreover, the RGA transcript was downregulated in 24-epiBL-treated seedlings in a dose-dependent manner. These results clearly show that the heterotrimeric G protein α subunit is involved in BR signal response in rice. The Arabidopsis gpa1-1 mutant, caused by mutation in the GPA1 gene (encoding the α subunit of heterotrimeric G protein), also showed reduced response to BL with respect to inhibition of vegetative growth and promotion of seed germination 12. Overall, we speculate that the involvement of the α subunit of heterotrimeric G protein in BR signal response might be a common mechanism in plants. In addition, as was shown in the det2-1 and bri1-5 mutants in Arabidopsis 12, we have found that the rice BR-related mutants d2-1 and d61-1 showed lower response to GA (our unpublished data) by the α-amylase activity analysis. Moreover, BR can rescue the germination phenotype of the severe GA-biosynthetic mutants ga1-3, ga2-1, ga3-1 and the GA-insensitive mutant sleepy1 in a concentration-dependent manner in Arabidopsis 25. Thus, we speculate that the lower responsiveness of GA in the d1 mutant might be due to cross-talk between BR and GA pathways.

In addition to our studies of altered BR responsiveness in the d1 mutant, root elongation inhibition assay revealed changed ethylene sensitivity in the d1 mutant. Actually, the transcription of ACC oxidase is increased in BR-treated lamina joints of rice 11. The ACC synthase gene family was also regulated in mung bean 26. Recently, both ethylene and BR were found to affect the elongation of cotton fibers. The inhibitory effect of BRZ (BR synthesis inhibitor) on fiber elongation can be rescued by ethylene, whereas the suppressed growth of fiber supplemented with 1-aminoethoxyvinylglycine, an ethylene synthesis inhibitor, cannot be reversed by BR, indicating that ethylene might act downstream of BR signaling 27. Taken together, these results suggest cross-talk between ethylene and BR signaling. Therefore, the changed ethylene sensitivity in the d1 mutant might be caused by altered BR responsiveness. However, whether BR is the singular plant hormone that couples with RGA and whether the reduced GA response in the d1 mutant is influenced by the BR signal need to be further investigated.

The d1 mutant is similar to the wild type in flowering time, whereas d2-1 and d61-1 (BR-deficient and -insensitive mutants, respectively) show delayed flowering time as compared with controls 16, 19. Furthermore, in our studies, the d1 mutant did not display a canonical de-etiolated phenotype, such as unelongated mesocotyls, as did d61-1. In addition, culture on media containing a high concentration of 24-epiBL (10 μM) resulted in inhibited growth of d1 mutant primary roots (Figure 1), but d61-1, the allele weaker than d61-2 and d61-4 16, 28, remained insensitive to 24-epiBL under the same condition. Moreover, the transcript level of RGA was downregulated by 24-epiBL in the d61-1 mutant (Figure 5). Overall, we speculate that RGA cannot be a direct downstream effector of OsBRI1. Recent evidence strongly supports cross-talk between GPCRs and receptor kinases in animals 29, 30, 31. Receptor kinases are also directly coupled by G proteins 32. In higher plants, the MLO gene family has been demonstrated experimentally to have a 7-transmembrane domain topology similar to GPCRs in animals, with an estimated 35 MLO genes in Arabidopsis, one might be a G protein coupled receptor 33. It would be interesting to investigate whether the action of a GPCR could functionally link the heterotrimeric G protein with BR signal response.