Plant architecture is an important agronomic trait and is useful for identification of plant species. The molecular basis of plant architecture, however, is largely unknown.
Forward genetics was used to identify an Arabidopsis mutant with altered plant architecture. Using genetic and molecular approaches, we analyzed the roles of a mutated cyclophilin in the control of plant architecture.
The Arabidopsis mutant roc1 has reduced stem elongation and increased shoot branching, and the mutant phenotypes are strongly affected by temperature and photoperiod. Map-based cloning and transgenic experiments demonstrated that the roc1 mutant phenotypes are caused by a gain-of-function mutation in a cyclophilin gene, ROC1. Besides, application of the plant hormone gibberellic acid (GA) further suppresses stem elongation in the mutant. GA treatment enhances the accumulation of mutated but not of wildtype (WT) ROC1 proteins. The roc1 mutation does not seem to interfere with GA biosynthesis or signaling. GA signaling, however, antagonizes the effect of the roc1 mutation on stem elongation.
The altered plant architecture may result from the activation of an R gene by the roc1 protein. We also present a working model for the interaction between the roc1 mutation and GA signaling in regulating stem elongation.
Cyclophilins (CyPs) constitute a large family of proteins that are present in a broad range of organisms, from bacteria to humans (Galat, 1999). The first CyP, CyP A, was identified as the specific target for the immunosuppressant cyclosporin A in mammalian T-cells (Handschumacher et al., 1984). CyPs possess peptidyl-prolyl cis-trans isomerase (rotamase) activity that catalyzes the rotation of X-proline peptide bonds from a cis to a trans conformation, a rate-limiting step in the protein-folding process (Galat, 1993). As potential molecular chaperones, CyPs play important roles in a variety of cellular processes, including protein trafficking (Freskgard et al., 1992), cell division (Faure et al., 1998), transcriptional regulation (Rycyzyn & Clevenger, 2002), signal transduction (Brazin et al., 2002), and RNA processing and spliceosome assembly (Gullerova et al., 2006). In higher plants, CyPs were first identified in tomato, maize, and oilseed rape (Gasser et al., 1990). Since then, CyP genes have been discovered in many other plant species, such as rice (Buchholz et al., 1994; Kumari et al., 2009), bean (Kinoshita & Shimazaki, 1999), and wheat (Li et al., 2010). In Arabidopsis, 29 CyP genes have been identified (Romano et al., 2004). Plant CyPs exist in all subcellular compartments and participate in protection from photo-damage (Dominguez-Solis et al., 2008), assembly of photosynthesis system II (Fu et al., 2007), and plant responses to biotic (Coaker et al., 2005) and abiotic stresses (Luan et al., 1994; Chen et al., 2007). Several CyPs have been found to be involved in the control of plant growth and development. A loss-of-function mutation in the CyP40 gene causes the precocious expression of the adult vegetative trait, but does not affect flowering time in Arabidopsis (Berardini et al., 2001). Further study indicated that cyp40 mutation caused the elevated expression of miRNA165-regulated members of the SPL family of transcription factors and thereby accelerated the phase transition from juvenile to adult (Smith et al., 2009). Li et al. (2007) reported that a WD40 domain-containing CyP, CyP71, regulates gene repression and organogenesis in Arabidopsis. The cyp71 null mutant exhibits reduced apical meristem activity, delayed and abnormal lateral organ formation, and arrested root growth. CyP71 protein is localized in the nucleus and affects the amount of histone methylation at the loci of homeotic genes KNAT1 and STM, which are two key regulators of shoot meristem activity. gaid is a wheat mutant with dwarf phenotype (Li et al., 2010). This mutant overaccumulates a CyP-like protein, TaCYP20-2, and overexpression of TaCYP20-2 in wildtype (WT) wheat resulted in a similar dwarf phenotype as gaid. The tomato auxin-resistant mutant diageotropica (dgt), which carries a mutation in the LeCYP1 gene, displays a pleiotropic phenotype including a slow gravitropic response, lack of lateral roots, and reduction of apical dominance (Oh et al., 2006).
Plant architecture is an important agronomic trait and is the most useful characteristic for identifying plant species (Wang & Li, 2007). It is defined by plant height and by the numbers and spatial arrangement of shoot branches and inflorescences. Alteration of plant architecture, such as plant height and tillering number in rice, can have a dramatic effect on crop yield. Since the 1960s, introduction of dwarf wheat and rice cultivars has greatly improved plant resistance to wind and rain and has increased grain yields (Hedden, 2003). Analyses of these dwarf cultivars and some dwarf mutants of Arabidopsis, pea, and tomato have led to the discoveries of several important genes involved in the control of plant height and shoot branching. It has become clear that the plant hormone gibberellic acid (GA) and brassinosteroid are two major regulators of plant stem elongation. Mutants defective in either the biosynthetic or signaling pathways of these two hormones exhibit a dwarf phenotype (Wang & Li, 2007). In Arabidopsis and rice, GA is perceived by GA receptors (Sun, 2010; Ueguchi-Tanaka & Matsuoka, 2010). Upon binding of GA, the GA-bound receptors interact with a DELLA protein in rice (Murase et al., 2008) or five DELLA proteins in Arabidopsis (Jiang & Fu, 2007; Harberd et al., 2009; Ariizumi et al., 2010). These DELLA proteins serve as negative regulators of GA responses. In the absence of GA, they suppress downstream responses by interacting with multiple regulatory proteins such as basic helix–loop–helix (bHLH) transcription factors PHYTOCHROME INTERACTING FACTORS (Harberd et al., 2009), ALCATRAZ (Arnaud et al., 2010), and GRAS protein SCARECROW-LOKE 3 (Heo et al., 2011). The interaction between the GA-bound receptors and the DELLA proteins causes DELLA proteins to be ubiquitinated via the GID2-based SCF complex in rice (Gomi et al., 2004) or via the SLY1-based SCF complex in Arabidopsis (McGinnis et al., 2003). The ubiquitinated DELLA proteins will then be targeted to 26S proteasome for degradation. The GA-mediated degradation of DELLA proteins thus releases GA responses, which include promotion of seed germination, stem elongation, flowering, and flower and silique development. In Arabidopsis, a gain-of-function mutation in one of the DELLA proteins, GAI, renders it resistant to GA-mediated degradation and also results in a dwarf phenotype (Peng et al., 1997). The orthologs of Arabidopsis GAI in wheat (Rht) and maize (D8) were found to be the ‘Green revolution’ genes that greatly enhanced grain yield in the 1960s (Peng et al., 1999). The GA-mediated degradation of Rht protein is also suppressed in the wheat dwarf mutant gaid (Li et al., 2010). By contrast, the functional disruption of the SLR1 gene that encodes the only DELLA protein in rice causes the plants to display a constitutive GA response, resulting in a slender phenotype (Ikeda et al., 2001).
We report here that a gain-of-function mutation in a highly conserved Arabidopsis CyP gene, ROC1, causes a dramatic alteration of plant architecture. The roc1 mutation does not seem to interfere with GA biosynthesis or signaling; however, GA signaling antagonizes the effect of roc1 mutation on stem elongation. Our work provides a useful genetic material that may help with the further identification of molecular components involved in the control of plant architecture.
Materials and Methods
Plant material and growth conditions
All Arabidopsis thaliana (L.) Heynh. plants used in this study were of the Columbia-0 background, except for the mutants ga1-3 (Sun et al., 1992) and gai (Peng et al., 1997), which were of Landsberg erecta (Ler) background. Seeds were surface-sterilized with 20% (v/v) bleach for 20 min. After three washes in sterile-distilled water, seeds were sown on MS medium (Murashige & Skoog, 1962). After the seeds were stratified at 4°C for 2 d, the agar plates were placed vertically in a growth room with a photoperiod of 16 h of light and 8 h of darkness at 22–24°C. The light intensity was 100 μmol m−2 s−1. Seven days after germination, the seedlings were transferred to soil for further characterization of growth phenotypes. For the study of the effects of temperature and photoperiod, the plants were grown at a specific temperature and photoperiod as indicated.
Measurement of length and width of stem epidermal cells
For measurement of the lengths and widths of epidermal cells, epidermal strips were peeled from the first internode of the main stalk of 6-wk-old plants. The strips were stained with a solution of toluidine blue oxide (0.5 mg ml−1) for 5 min, rinsed with distilled water, and observed with a stereomicroscope (Olympus SZ61, Tokyo, Japan). Cell length and width were measured with the aid of DIGIMIZER 220.127.116.11 software (http://www.digimizer.com).
Seven-day-old seedlings grown on MS agar plates were transferred to soil and grown under conditions of constant light, under short-day (SD, 8 h of light) or long-day (LD, 16 h of light) conditions. Beginning at day 18 after seed germination, plants were sprayed every 4 d with 100 μM GA3 dissolved in 0.02% Tween 20. Control plants were sprayed with 0.02% Tween 20.
Genetic mapping of the mutant gene
The mapping population was generated by crossing the roc1 mutant to a plant of Ler ecotype. The 840 F2 progeny that displayed the mutant phenotype were selected, and DNAs from these individuals were isolated for molecular mapping. A set of simple sequence length polymorphism (SSLP) and cleaved amplified polymorphic sequence (CAPS) markers were used to map the mutant gene. The sequences of the molecular marker and their chromosomal position are listed in Supporting Information, Tables S1 and S2.
Quantitative real-time PCR analysis
Total RNAs were extracted from 8-d-old seedlings with the TIANGEN RNAprep pure plant kit (Tiangen Co., Beijing, China). The DNase-treated RNA (1 μg) was reverse-transcribed in a 20 μl reaction using reverse transcriptase (Toyobo, Osaka, Japan) according to the manufacturer's manual. The first strand of cDNA was synthesized using oligo-dT and Takara MLV-Reverse transcriptase. cDNA was amplified using SYBR Premix Ex Taq (TaKaRa, Shiga, Japan) on the Bio-Rad CFX96 real-time PCR detection system. ACTIN mRNA was used as an internal control. The primer sequences used for amplifying the ACTIN mRNA were 5′- GGTAACATTGTGCTCAGTGGTGG-3′ and 5′-AACGACCTTAATCTTCATGCTGC-3′, and those for ROC1 or roc1 mRNA were 5′-TGTTGGCGGTACCGGAAAA-3′ and 5′-GGTGTGCTTCCTCTCGAAA-3′.
Vector construction and plant transformation
The genomic sequences of the ROC1 and roc1 genes were PCR-amplified from genomic DNA of WT or roc1 plants using the primers 5′-GTCGACGCTTCCCCTTGGCCGGCTAT-3′ and 5′-ACTAGTACTAGTAACAAAGATGAAATTTGG-3′. During the amplification, SalI and SpeI restriction sites were added at the 5′ and 3′ ends of the ROC1 and roc1 genes. The amplified ROC1 and roc1 genes were cloned, separately, into the corresponding sites after the CaMV 35S promoter on the binary vector pCAMBIA1301, resulting in the constructs of 35S::ROC1 or 35S:roc1. To generate the ROC1::roc1 construct, the roc1 genomic sequence with a 2 kb upstream fragment was PCR-amplified using the primers 5′-AAGCTTGAGTGTCATATCTTACTTTTTCTTGG-3′ and 5′-GTCGACACTAGTAACAAAGATGAAATTTGG-3′. During the amplification, HindIII and Sal I sites were added to the 5′ and 3′ ends of the PCR product. The amplified fragment was used to replace the CaMV 35S promoter on the binary vector pCAMBIA1301. For the ROC1::GUS fusion construct, the same 2.0 kb ROC1 upstream sequence as in the ROC1::roc1 construct was PCR-amplified using the primers 5′-AAGCTTGAGTGTCATATCTTACTTTTTCTTGG-3′ and 5′-CCCGGGGGCGAATATCTCACAGATAAAC-3′. During the amplification, HindIII and XmaI sites were added at the 5′ and 3′ ends of the PCR fragment. The amplified fragment was then used to replace the CaMV 35S promoter on the binary vector pBI121. For the 35S::ROC1-HA and 35S::roc1-HA constructs, the ROC1 or roc1 gene was amplified using primers 5′-TCTAGAATGGCGTTCCCTAAGGTATACTTC-3′ and 5′-ACTAGTAGAGAGCTGACCACAATCGGCAAC-3′. During the amplification, the XbaI and SpeI restriction sites were added to the 5′ and 3′ ends of the PCR products. The amplified PCR fragments were cloned to the binary vector pCAMBIA1300, which had been modified to include 3X HA epitope sequences. To generate ROC1 RNAi lines, the 261 bp ROC1 genomic sequence was PCR-amplified using the primers 5′-CGACGACAAGACCCTACCATCGACGGCCAGCCCGC-3′ and 5′-GAGGAGAAGAGCCCTCCTCTCGAAATTCTCGTCCTCG-3′. This fragment was then cloned into the corresponding sites of the OZ-LIC RNAi vector (Xu et al., 2010).
Histochemical analysis of β-glucuronidase (GUS) activity
Histochemical analysis of GUS activity was performed according to Jefferson (1989).
Protein extraction and western blots
Eight-day-old Arabidopsis seedlings were ground to fine powders in liquid nitrogen. One volume of ice-cold extraction buffer (0.1 M K-acetate, 20 mM CaCl2, 2 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride, 20% glycerol (v/v), pH 5.4) was added to the powders. Samples were gently agitated on ice for 0.5 h and then centrifuged at 13 400 g at 4°C for 6 min. The supernatant was transferred to a fresh tube before use. About 30 μg of denatured protein was separated on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels. After electrophoresis, the proteins were transferred to a polyvinylidene difluoride (PVDF) membrane in transfer buffer (25 mM Tris, 43 mM glycine, 20% methanol) and subjected to western blot analysis using anti-HA antibodies as described (Wang et al., 2011).
Generation of double mutants
To generate the double mutant roc1/ga1-3, a heterozygote roc1/+ plant was crossed to a homozygote ga1-3 plant. The F2 progeny were screened for the desired double mutant using a PCR-based method. The primers used to identify the roc1/ga1-3 double mutant were 5′-CCCACTCACTTAACCTATTTTC-3′ and 5′-CGTTAATTTTATTGTATGCACG-3′, which covered the ga1-3 deletion region, and derived cleaved amplified polymorphic sequence (dCAPS) markers of 5′-GTCGACGCTTCCCCTTGGCCGGCTAT-3′ and 5-CATGAAGTTAGGGATCACACGGTGAAATCTA-3, which were designed to specifically detect the roc1 mutation. With this pair of dCAPS markers, the amplified PCR product from the roc1 mutant would produce two bands of 178 and 32 bp after XbaI digestion. The PCR product from the WT would display only one band of 210 bp because this fragment lacked an internal XbaI site. The double mutant roc1/gai was generated from the cross of a heterozygous roc1/+ plant with the homozygote gai mutant. The F2 progeny were screened for the double mutant using the primers 5′-CCAACCATGAAGAGA GATCATC-3′ and 5′-CGAACTGATTGAGAATCGCGTC-3′ for gai mutation and the same dCAPS markers for roc1 mutation as already mentioned.
Confocal laser microscopy
After the transgenic plants were grown on MS plates under long-day conditions for 14 d, they were treated with water or with 100 μM GA3 for 8 h. The root tissues were then excised with a razor blade, and green fluorescent protein (GFP) fluorescence was examined with a confocal laser scanning microscope (Zeiss 710META). Excitation and emission wavelengths of 488 and 507 nm were used to visualize the GFP signals.
Identification of an Arabidopsis mutant with altered plant architecture
An Arabidopsis mutant with reduced height and increased number of shoot branches was identified from a T-DNA activation tagging library (Koiwa et al., 2006). When the mutant was self-pollinated, its progeny segregated into three phenotypically distinct groups: WT-like plants; plants that resembled the originally identified mutant; and plants that could form a rosette but could not bolt (Fig. 1a). This suggested that the originally identified mutant was a heterozygote. The segregation ratio of these three types of plants was c. 1 : 2 : 1 (22 : 55 : 28). When the original mutant was backcrossed to the WT, half of the F1 progeny looked like the WT, and the other half was shorter and had more shoot branches than the WT. Furthermore, when the F1 plant with the mutant phenotype from the backcross was self-pollinated, its F2 progeny continued to segregate into three distinct groups in a ratio of 1 : 2 : 1. These results indicated that the mutant phenotypes were caused by a single semidominant mutation. The homozygote mutant plant was named roc1 (or roc1/roc1) because it carried a point mutation in the ROC1 gene, as described later. The roc1 mutant was backcrossed to the WT four times before further characterization. Depending on developmental stage, the heterozygote plants (roc1/+) were about one-third to one-half as tall as the WT plants. roc1/+ plants had about three times more shoot branches than the WT (6.57 ± 1.34 vs 2.22 ± 0.74 per plant; Fig. 1b). Microscopy showed that the stem epidermal cells were 50% shorter and 30% wider in roc1/+ plants than in the WT (Fig. 1c,d). There was no obvious difference in flowering time among the WT, roc1/+, and roc1/roc1plants, as measured by the time that plants required to bolt or the number of rosette leaves at the time when plants began bolting (Fig. S1). In the first 3 wk after seed germination, the morphology of roc1/+ and roc1/roc1 plants was similar to that of the WT. After growing in soil for 30 d, WT, roc1/+, and roc1/roc1 plants had the same number of rosette leaves, but the leaf shape differed between roc1/roc1 and WT plants (Fig. 1e). The roc1/roc1 plants could not bolt but could produce a few (one to three) flowers directly from the center of the rosette. The development of flower organs and production of seeds were normal in both roc1/+ and roc1/roc1 plants. All the phenotypic comparisons between the WT and the mutants described earlier were conducted under a long-day (LD) photoperiod (16 h light : 8 h dark) at 22–24°C.
roc1 phenotypes are affected by temperature and photoperiod
Environment profoundly affects plant growth and development. We first tested the effect of temperature on the growth of roc1. When grown at 16°C for 3 wk, the homozygote mutant plants (roc1/roc1) formed a rosette that was similar to that formed by the WT. After growing for 5 wk at 16°C, roc1/roc1 plants produced numerous flowers directly from the center of the rosette but did not produce any inflorescence stems. After growing for 70 d at 16°C, roc1/roc1 produced a similar number of mature siliques as the WT (106 ± 22.45 vs 120 ± 16.41 siliques per plant, n = 20; Fig. 2a,b). When roc1/roc1 plants were grown at 28°C, however, their mutant phenotypes were completely suppressed (Fig. 2c).
Next, we examined the effects of photoperiod on the mutant phenotypes. When grown under a short-day (SD) photoperiod (8 h light : 16 h dark) for 60 d at 22–24°C (in this study, all experiments were conducted at 22–24° unless indicated), the WT, roc1/+, and roc1/roc1 plants did not form inflorescence stems. In addition, these three types of plant did not differ in number, shape, and size of rosette leaves, indicating that the mutant phenotypes were suppressed under the SD condition (Fig. 3b). When grown under constant light, roc1/+ plants, but not roc1/roc1 plants, bolted but their inflorescence stems were substantially shorter than under LD conditions (Fig. 3c). Under constant light, the rosette leaves were larger for roc1/+ than for WT plants; but the morphology of roc1/roc1 plants (Fig. 3c) was similar to that under LD conditions (Fig. 3a).
In conclusion, high temperature suppresses and low temperature enhances the roc1 mutant phenotypes. The roc1 mutant phenotypes are also suppressed by low light and are enhanced by high light.
GA treatment inhibits stem elongation of the mutant plants
Because the phenotypes of roc1/+ plant resembled those of plants with defects in GA biosynthesis or GA signaling, we tested whether application of exogenous GA could rescue the mutant phenotypes. The WT, roc1/+, and noc1/roc1 plants grown under LD, SD, and constant light conditions were sprayed with 100 μM GA3 every 4 d, starting at day 18 after seed germination. Under all three photoperiod conditions, GA3 treatment of WT plants promoted stem elongation, resulting in taller plants, and caused leaves to become narrow and curled (Fig. 3). By contrast, GA3 treatment of roc1/+ plants completely inhibited stem elongation. Like the leaves of the GA3-treated WT, leaves of GA3-treated roc1/+ plants became narrow and curled. GA3 treatment of roc1/roc1 plants could not promote stem elongation either. In addition, it reduced both the size and number of rosette leaves. These results indicated that GA3 treatment could not rescue the mutant phenotypes, in terms of stem elongation, and that the roc1 mutants were not deficient in GA biosynthesis. Interestingly, treatment of the mutants with GA3 further suppressed stem elongation, suggesting an interaction between roc1 mutation and GA signaling.
roc1 is a gain-of-function mutation
Although the roc1 mutant was identified from a T-DNA activation tagging library (Koiwa et al., 2006), our genetic analysis indicated that its mutant phenotypes were not linked to T-DNA insertion (data not shown). We therefore used map-based cloning to identify the molecular lesion in roc1. A roc1/+ plant (Columbia background) was crossed to an Arabidopsis plant with a Ler background to establish a mapping population. All F1 progeny derived from this cross showed WT phenotypes. The F2 progeny from half of the self-pollinated F1 plants segregated into three distinct groups with phenotypes of WT, roc1/+, and roc1/roc1. The numbers of the plants for each group were 211, 26, and 11, respectively. These results suggested that a gene in the Ler genome suppressed roc1 phenotypes in a dominant manner. Using a series of SSLP and CAPS markers, we found two genetic loci that were tightly linked to the mutant phenotypes. Further fine mapping indicated that one locus was in a 50 kb region near the bottom of chromosome 4 (Fig. 4a). We sequenced all 12 annotated genes within this region and found a point mutation in the gene At4g38740 (ROC1), which encodes a cytosolic CyP protein (Lippuner et al., 1994; Romano et al., 2004). The ROC1 gene contains no intron and encodes a protein with a predicted size of 18 kD. The C to T mutation at its nucleotide position 173 converts a serine to a phenylalanine in a conserved region among the CyPs from yeast, humans, and plants (Fig. 4b). To further confirm that the mutated ROC1 gene was responsible for the mutant phenotypes, we introduced the mutated ROC1 (roc1) gene (including its ORF with both 5′ and 3′ untranslated regions) into WT plants under the control of a CaMV 35S promoter or ROC1's native promoter. The primary transformants carrying either construct displayed three distinct phenotypes: WT-like, roc1/+-like, and roc1/roc1-like. The number of the transformants with each phenotype is shown in Table 1. That the same construct produced transformants with different phenotypes was probably the result of a position effect that resulted in a different expression level of the roc1 gene. The recapitulation of the mutant phenotypes in WT plants with the introduced roc1 gene confirmed the authenticity of the mapping results. By contrast, when the WT ROC1 gene was expressed under a CaMV 35S promoter in WT plants, no morphological changes were observed in any transformant, although the ROC1 gene was overexpressed in these plants (Fig. S2). Because ROC1 T-DNA knockout lines are currently unavailable, we generated ROC1 RNAi lines in which the expression of the ROC1 gene was substantially reduced (Fig. S2). However, no obvious alteration of growth characteristics was found in these RNAi lines (data not shown). Taken together, these results indicate that the roc1 mutant phenotype was caused by a gain-of-function mutation in the ROC1 gene.
Table 1. Summary of the results of transgenic Arabidopsis thaliana
Number of transgenic lines in each phenotypic group
The second locus that linked to roc1 mutant phenotypes was mapped to a 140 kb interval on chromosome 1 (Fig. 4c). Interestingly, this region contains five putative disease resistance (R) genes. Among them, four (At1g58807, At1g58848, At1g59124, and At1g59218) belong to the CC-NBS-LRR type R gene family and one (At1g58602) is annotated to be an NB-LRR type R gene without a CC domain.
Regulation of ROC1 expression
To determine the expression patterns of the ROC1 gene, we extracted total RNAs from roots, leaves, stems, flowers, and siliques from WT plants and subjected the extracts to quantitative real-time PCR analysis. The results showed that ROC1 was expressed in all plant organs (Fig. 5a). This was consistent with the previously published results obtained by northern blot analysis (Lippuner et al., 1994). In the roc1 mutant, the expression level of roc1 mRNA was similar to that of ROC1 in the WT (Fig. S2). To further investigate the expression pattern of ROC1, we fused a 2.0 kb DNA sequence upstream of the ROC1 transcription start site with a GUS reporter gene and transformed this fusion construct into WT plants. Ten lines of transgenic plants were generated, and the ROC1::GUS expression pattern of a representative line is shown in Fig. 5(b–g). In roots, GUS activity was mainly detected in the root tip and vascular tissues (Fig. 5b). In cotyledons and mature rosette leaves, GUS activity was strictly limited to vascular tissues (Fig. 5b,c). ROC1::GUS was also strongly expressed in elongating inflorescence stems and leaf petioles, but was weakly expressed in older parts of these tissues (Fig. 5c–f). In reproductive organs, GUS was expressed in pollen grains, ovules, pedicles, and separating tissues of young siliques (Fig. 5g). Furthermore, real-time PCR and GUS histochemical analysis indicated that transcription of the ROC1 and roc1 genes was not affected by GA3 treatment in WT plants or roc1 mutants (Fig. S3).
Next, we examined the effect of GA3 treatment on the accumulation of WT and mutant ROC1 proteins. Because we did not have specific antibodies against ROC1 or roc1 proteins, we generated transgenic plants that produced C-terminus HA-tagged ROC1 or roc1 proteins under the control of a CaMV 35S promoter. As was the case for 35S::ROC1 plants, the transgenic plants with the 35S::ROC1-HA construct did not exhibit any obvious morphological changes. The primary transformants with 35S::roc1-HA construct, however, fell into three phenotypic groups: WT-like, roc1/+-like, and roc1/roc1-like. The progeny from the selfed roc1/+-like transformant continued to segregate into three types of plants, as described earlier for the selfed roc1/+ plant. Furthermore, when the progeny from the roc1/+-like transformant (Fig. 5h, left) was treated with GA3, they either looked like GA3-treated WT plants or their stem elongation was completely blocked (Fig. 5h, right). Molecular analysis indicated that the GA3-treated plants that behaved like the WT did not contain the 35S::roc1-HA transgene because of genetic segregation. These results further confirmed the inhibitory effect of the roc1 protein on plant stem elongation and its responsiveness to GA signals. We then compared the abundance of roc1-HA proteins among the WT-like, roc1/+-like, and roc1/roc1-like progeny segregated from the selfed roc1/+-like primary transformant. Total proteins were extracted from the rosette leaves of 1-month-old plants and were subjected to western blot using anti-HA antibodies. As expected, roc1-HA proteins were not expressed in WT-like plants. The abundance of roc1-HA proteins was higher in roc1/roc1-like plants than in roc1/+-like plants, indicating a correlation between the severity of mutant phenotype and the abundance of roc1 proteins (Fig. 5i). We then examined the effect of GA3 on the accumulation of ROC1-HA and roc1-HA proteins. Seeds from a 35S::ROC1-HA line or a 35S::roc1-HA line with roc1/+-like phenotype were germinated on MS medium. Eight days after germination, the seedlings were sprayed with 100 μM GA3 and grown for another 8 h. Total proteins were extracted from GA3-treated or untreated seedlings and were subjected to western blot using anti-HA antibodies. In 35S::roc1-HA plants, GA3 treatment caused an increased accumulation of roc1-HA protein. In 35S::ROC1-HA plants, however, GA3 treatment did not affect the abundance of ROC1-HA protein (Fig. 5j). This indicated that the point mutation in ROC1 affected its protein accumulation in a GA-dependent manner. Furthermore, our real-time PCR analyses indicated that the increase in roc1 protein accumulation was not a result of GA3 enhancement of roc1 mRNA stability, because the mRNA level of the ROC1 or roc1 gene was not affected by GA3 treatment in 35S:ROC1-HA or 35S:roc1-HA plants (Fig. S4).
Finally, we investigated the effects of temperature on the expression of mRNA and protein of ROC1 and roc1. When 14-d-old seedlings grown at 22°C were shifted to 28°C for 24 h, the mRNA levels increased 1.5- and 1.0-fold for ROC1 and roc1 genes, respectively (Fig. 5k). When the seedlings were shifted from 22 to 16°C, however, the mRNA level of ROC1 increased 20%, while that of roc1 was reduced 20%. We then examined the effect of temperature on protein stability of ROC1 and roc1. We chose a 35S::ROC1-HA and a 35S::roc1-HA transgenic line for this study. When 14-d-old seedlings of these two transgenic lines were shifted from 22 to 28°C for 24 h, the expression levels of ROC-HA and roc1-HA mRNA were enhanced 100% and 50%, respectively (Fig. S5). Because the expression of these two fused genes was directed by a constitutive 35S promoter, these results indicated that high temperature might increase the stability of these two RNA species. Accordingly, we observed an increase in the protein abundance for both ROC1-HA and roc1-HA (Fig. 5l,m). When the plants were exposed to 16°C for 24 h, mRNA levels of ROC1-HA and roc1-HA did not change (Fig. S5). Interestingly, we found a dramatic increase in the abundance of roc1-HA but not of ROC1-HA proteins, indicating that low temperature greatly increased the stability of roc1-HA proteins (Fig. 5l,m). The high accumulation of roc1-HA proteins at 16°C may explain the dramatic phenotype when the plants are grown at low temperature (Fig. 2); however, the suppression of roc1 phenotypes at 28°C may not result from a decrease in the abundance of roc1 proteins, because the abundance of roc1 proteins at 28°C was slightly increased.
Genetic interaction between roc1 and GA signaling
To investigate whether the roc1 mutation interfered with GA signaling in regulation of stem elongation, we first generated the roc1/ga1-3 double mutant through genetic cross. The Arabidopsis ga1-3 mutant contains a 5 kb deletion at the GA1 locus that encodes ent-kaurene synthase A (Sun et al., 1992). This enzyme catalyzes the first committed step in the GA biosynthetic pathway. Under LD conditions, ga1-3 seeds germinated on GA3-containing medium but the seedling did not form a normal inflorescence stem when transferred to soil (Fig. 6a). When ga1-3 seedlings were treated with GA3 every 4 d after transfer to soil, their growth characteristics resembled those of the WT. The roc1/ga1-3 double mutant, however, could not bolt even after GA3 treatment. This result further demonstrated that the roc1 mutation prevented GA-induced stem elongation. We then generated the roc1/gai double mutant through genetic cross. The Arabidopsis gai mutant harbors a GAI gene with a gain-of-function mutation that makes the GAI protein resistant to GA-mediated degradation and renders plants partially insensitive to GA (Peng et al., 1997). As a consequence, the gai mutant has reduced stem elongation. The phenotype of the roc1/gai double mutant was similar to that of the gai single mutant, indicating that gai mutation antagonized the inhibitory effect of the roc1 mutation on stem elongation (Fig. 6b).
To determine whether the roc1 mutation blocked GA-mediated degradation of DELLA proteins and caused plants to become insensitive to the GA signal, we introduced a 35S::GFP-RGA marker gene into the roc1 mutant through genetic cross. RGA is one of the five DELLA proteins in Arabidopsis. In this gene construct, RGA was fused with a GFP protein at its N-terminus and was expressed under the CaMV 35S promoter (Silverstone et al., 2001). In the absence of GA, the expression of GFP-RGA fusion proteins could be detected in the nucleus of root cells in 14-d-old seedlings of both the WT and the roc1 mutant (Fig. S6a,b). When the 35S::GFP-RGA seedlings were treated with GA3 for 8 h, the accumulation of GFP-RGA was diminished (Fig. S6c). This GA-induced degradation of GFP-RGA protein was also observed in the roc1 mutant (Fig. S6d), indicating that the roc1 mutation did not interfere with GA-mediated degradation of DELLA proteins.
In this work, we identified an Arabidopsis mutant, roc1, with altered plant architecture. The heterozygote mutant plants (roc1/+) are shorter and produce more shoot branches than the WT. The homozygote mutant plants (roc1/roc1) cannot bolt. In roc1/+ plants, stem epidermal cells are 50% shorter than in the WT, providing a cellular basis for reduced stem elongation. The roc1 mutant phenotypes are strongly affected by temperature and photoperiod. The flowering time and development of flower organs and siliques, however, are not altered in roc1/+ or roc1/roc1 plants, suggesting that the mutation does not affect reproductive development. Our genetic and molecular analyses indicate that the mutant phenotypes were caused by a gain-of-function mutation in a cyclophilin gene ROC1. ROC1 was cloned and characterized as a cytosolic CyP in Arabidopsis in 1994 (Lippuner et al., 1994). So far, the only known biological function of ROC1 is its role in a plant–pathogen interaction (Coaker et al., 2005). Specifically, when the bacterium Pseudomonas syringae infects Arabidopsis plants, it delivers an effector protein AvrRpts (a cysteine protease) into host cells. After it is delivered into host cells, AvrRpts is activated by ROC1, which binds to AvrRpts at four sites and properly folds it by peptidyl-prolyl cis/trans isomerization (Coaker et al., 2006; Aumüller et al., 2010). Activation of AvrRpts leads to the cleavage of the Arabidopsis protein RIN4, which triggers the activation of disease-resistance gene RPS2-mediated defense response. It has long been thought that plant growth and defense response are intimately connected and that there are some tradeoffs between these two processes (Heil & Baldwin, 2002). Retarded growth is sometimes exhibited by mutants with constitutive defense responses, possibly because of resource reallocation from growth to defense or the metabolic burden of defense responses on plants, or both. For example, plants with a loss-of-function mutation in the BON1 gene, which encodes a copine-like protein, exhibit precocious cell death and enhanced disease resistance at low temperatures (Jambunathan et al., 2001). The bon1 plants also have a severe dwarf phenotype (Hua et al., 2001). The disruption of the BON1 function activates a resistance (R) gene, SNC1, leading to constitutive defense responses and retarded growth (Yang & Hua, 2004). These are also the phenotypes of snc1 mutant plants that carry a gain-of-function mutation in the SNC1 gene, resulting in the activation of SNC1 (Zhang et al., 2003). Similarly, the Arabidopsis mutant cpr30 displays both dwarfism and constitutively activated defense responses (Guo et al., 2009). Our mapping of the second locus that is linked to roc1 mutant phenotypes narrowed the candidate gene to a 140 kb region on chromosome 1. In this region, we found five putative R genes. We speculate that the point mutation in the ROC1 protein may alter its enzymatic activity, resulting in the activation of one of these five R genes, thus causing plants to display a constitutive defense response. This R gene may be specific to the Col ecotype. Thus, when a roc1 plant (in the Col ecotype) was crossed to a plant of the Ler ecotype, the ortholog of this R gene in the Ler plant functions as a suppressor of the roc1 mutant phenotypes. The same scenario was previously found for SNC1-mediated plant growth and disease resistance. SNC1 is specific to the Col ecotype, and the same SNC1 sequence is absent in the Wassilewskija (Ws) ecotype (Yang & Hua, 2004). As a consequence, Ws plants with a loss-of-function mutation in the BON1 gene were unable to display the mutant phenotypes as observed for the Col plants. Previous reports have indicated that high temperature can suppress many R gene-mediated dwarf phenotypes, including those observed for bon1, snc1, and cpr30 (Zhang et al., 2003; Yang & Hua, 2004; Guo et al., 2009), and low light reduced cell death associated with disease resistance (Dietrich et al., 1994; Xiao et al., 2003). roc1 mutant phenotypes were also inhibited by high temperature and low light, providing more evidence for the hypothesis that the growth defects of the roc1 mutant result from activated defense responses. To further support this hypothesis, we must analyze the defense-related responses of roc1 plants and determine which of the five putative R genes is the target of the roc1 protein.
Alternatively, WT ROC1 protein may not play any roles in regulating plant growth and development. Instead, the mutation in the ROC1 gene may cause the roc1 protein to have a novel function in which it interacts with new targets that affect stem elongation. These new targets could include the molecular components involved in GA biosynthesis or signaling, or the proteins directly involved in the control of cell expansion. This hypothesis is supported by the fact that both ROC1-overexpressing or RNAi lines do not display altered growth phenotypes. Thus, it will interesting to determine how this mutation affects the ROC1 protein structure and its target specificity. In addition, our results indicate that the gene mutation enables roc1 proteins to respond to GA signal in terms of their protein stability, establishing a possible link between the new function of roc1 protein and GA signaling.
The plant hormone GA plays a crucial role in regulating stem elongation. Our work shows that application of exogenous GA3 cannot rescue the roc1 mutant phenotypes, indicating that the roc1 mutant phenotypes do not result from a deficiency in GA biosynthesis. In addition, the seed germination rates of roc1/+ or roc1/roc1 plants were similar to that of the WT (data not shown), further indicating that the mutant plants were not deficient in GA. An alternative explanation, that the roc1 mutation, like the overexpression of TaCYP20-2 in wheat (Li et al., 2010), makes plants insensitive to GA signaling also seems incorrect for the following reasons. First, GA3 treatment led to further suppression of stem elongation in roc1/+ plants, indicating that the mutant plants were able to sense the GA signal, even though the GA effect on growth was contrary to what was expected. Secondly, the leaf morphologies of GA3-treated roc1/+ and roc1/roc1 plants resembled that of GA3-treated WT plants, suggesting that responsiveness of leaves to GA signal in the mutant plants was not impaired. Thirdly, flowering times of the WT and the mutant plants were similar. Finally, the GA-mediated degradation of a key GA signaling component, RGA, is not affected in roc1 plants. All these results indicate that the impairment of stem elongation in roc1/+ and roc1/roc1 plants is not a result of plant insensitivity to GA signals.
To explain the inhibitory effect of GA3 on stem elongation in roc1/+ plants, we hypothesize the following model. Under normal growth conditions, the WT ROC1 proteins do not function in the control of stem elongation. The point mutation in the ROC1 gene, however, changes the conformation of its encoded protein and causes the protein (roc1) to have a new function. In this new function, roc1 protein alters the activity of new target proteins that are directly involved in the control of stem elongation (Fig. 7). The suppressive effect of roc1 protein on stem elongation is proportional to the quantity of roc1 protein expressed – that is, higher expression leads to more severe mutant phenotypes. The effects of the roc1 protein on its downstream targets, however, are antagonized by the DELLA proteins. We propose that treatment of the mutant plants with GA3 caused the degradation of DELLA proteins, which would tend to enhance stem elongation. Treatment of the mutant plants with GA3, however, also increased the abundance of roc1 proteins (perhaps as a result of the elimination of the antagonistic effect of DELLA proteins and some other unknown mechanisms), which would inhibit stem elongation. As a consequence, the phenotype of GA3-treated roc1 plants will depend on these two opposite effects on stem elongation. If the effect of enhanced roc1 accumulation is greater than that of the degradation of DELLA proteins, the plants will exhibit a further suppression of stem elongation. In the case of the roc1/gai double mutant, however, the mutated GAI protein might permanently block the activity of roc1 protein (through either direct or indirect interactions). Thus, the roc1/gai plants display only the phenotypes of gai.
Finally, we want to point out that, although the WT ROC1 protein may not play a direct role in regulating plant growth and development, the gain-of-function roc1 mutant will be useful for further identification of novel molecular components involved in the control of plant architecture. This may be achieved by final identification of the genetic suppressor of roc1 described in this study. Our molecular mapping for the second genetic locus that is linked to the roc1 phenotype implied that this suppressor is probably an ecotype-specific R gene. If this inference can be confirmed, it will help to explain how the activation of an R gene affects plant development. Alternatively, molecular and biochemical approaches, such as yeast two-hybrid assay and purification of the roc1-containing protein complex, could be used to identify the roc1 interacting proteins and thereby to gain new insights into the molecular basis of plant architecture. In addition, the introduction of mutated ROC1 proteins or the mutated form of their orthologs into crops may create new and useful cultivars for breeding.
We thank Drs Jianming Zhou (Institute of Genetics and Developmental Biology, Cinese Academy of Science) and Jianming Li (University of Michigan) for helpful discussion of the manuscript. We also thank Drs Kashchandra Raghothama and Ray Bressan (Purdue University) for providing the Arabidopsis T-DNA activation library, the Arabidopsis Biological Resource Center for providing seeds of 35S::RGA-GFP, and Dr Xiangdong Fu (Institute of Genetics and Developmental Biology, Chinese Academy of Sciences) for providing ga1-3 and gai mutant seeds. This work was supported by the Ministry of Science and Technology of China (grant no. 2009CB119100), the National Natural Science Foundation of China (grant no. 31170238), and the Ministry of Agriculture of China (grant no. 2011ZX08009-003-005).