RTE1 Is a Golgi-Associated and ETR1-Dependent Negative Regulator of Ethylene Responses

doi:10.1104/pp.107.104299

Journal List > Plant Physiol > v.145(1); Sep 2007

Plant Physiol. 2007 September; 145(1): 75–86.

doi: 10.1104/pp.107.104299.

PMCID: PMC1976582

RTE1 Is a Golgi-Associated and ETR1-Dependent Negative Regulator of Ethylene Responses¹^[C]^[W]

Xin Zhou, Qian Liu, Fang Xie, and Chi-Kuang Wen^*

National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200032, China

^*Corresponding author; e-mail qgwen/at/sibs.ac.cn.

Received June 18, 2007; Accepted July 16, 2007.

This article has been cited by other articles in PMC.

Abstract

Arabidopsis (Arabidopsis thaliana) RTE1 encodes a membrane protein and negatively regulates ethylene responses. Genetic and transformation studies suggest that the function of the wild-type RTE1 is primarily dependent on ETR1 and can be independent on the other receptors. Ethylene insensitivity caused by the overexpression of RTE1 is largely masked by the etr1-7 mutation, but not by any other receptor mutations. The wild-type ETR1 N terminus is sufficient to the activation of the RTE1 function and the ectopic expression of etr1(1–349) restored ethylene insensitivity conferred by 35S

gRTE1 in etr1-7. The RTE1 N terminus is not essential to the etr1-2 function and the expression of rte1(NΔ49), which has an N-terminal deletion of 49 amino acid residues, restored ethylene insensitivity in etr1-2 rte1-2. The ectopic expression of GREEN FLUORESCENT PROTEIN (GFP)-RTE1 conferred ethylene insensitivity in wild type and the GFP fusion displayed fast movement within the cytoplasm. The GFP-RTE1 and EYFP-NAG proteins colocalized and the Brefeldin A treatment caused aggregation of GFP-RTE1, suggesting RTE1 is a Golgi-associated protein. Our results suggest specificity of the RTE1 function to ETR1 and that endomembranes may play a role in the ethylene signal transduction.

Ethylene is a gaseous hormone important to plant growth and development. In Arabidopsis (Arabidopsis thaliana), the dark-grown seedling exhibits an elongated and thin hypocotyl and a long radical root. Ethylene treatment causes the so-called triple response in etiolated seedlings; that is, growth inhibition in the seedling hypocotyl and radical root, the swelling of the hypocotyl, and exaggerated tightening in the apical hook (Guzman and Ecker, 1990). For light-grown seedlings exhibiting ethylene responses, the primary root becomes short and cotyledons are small, erected, and epinastic (Xie et al., 2006). In the adult stage, ethylene responses lead to reduction in rosette size (Hua and Meyerowitz, 1998; Xie et al., 2006). The seedling triple response can be used to isolate ethylene response mutants in Arabidopsis (Guzman and Ecker, 1990) and to score for degrees of ethylene responses (Hua and Meyerowitz, 1998; Qu and Schaller, 2004; Xie et al., 2006).

There are five ethylene receptors (ETR1, ERS1, ETR2, EIN4, and ERS2) in Arabidopsis and they are structurally similar to prokaryotic two-component modules. Suc gradient fractionation suggests that ETR1 is an endoplasmic reticulum (ER)-associated protein (Chen et al., 2002). Ethylene responses are negatively regulated by ethylene receptors and the loss of multiple receptor genes additively increases degrees of ethylene responses throughout developmental stages (Hua and Meyerowitz, 1998). Arabidopsis ethylene receptors have three or four transmembrane domains, a GAF domain, and a His-kinase domain; some receptors have a receiver domain following the His-kinase domain. Ethylene binding to the transmembrane domains results in the inactivation of the receptor activity, leading to a transmitter-off state (Wang et al., 2006). Mutations that abolish ethylene binding lead to ethylene insensitivity (Schaller and Bleecker, 1995; Wang et al., 2006). His-kinase activity is not essential to the ETR1 receptor function (Wang et al., 2003) and the truncated etr1(1–349) protein, which has the transmembrane and GAF domains, is capable of the repression of ethylene responses in the presence of subfamily I receptors (Xie et al., 2006).

A chemical screen for a mutant responsive to the ethylene antagonist trans-cyclooctene identifies RESISTANT TO ANTAGONIST1 (RAN1), which encodes a copper-transporting P-type ATPase protein (Hirayama et al., 1999). The human Menkes protein (ATP9A) is a homolog of the RAN1 protein and localizes to the trans-Golgi network (Francis et al., 1998). It is believed that Arabidopsis RAN1 also localizes to the post-Golgi through which the cuprous ion is delivered to ethylene receptors (Hirayama et al., 1999). The ran1-3 allele is identified from a genetic screen for constitutive ethylene response mutants and the mutation is lethal. It is hypothesized that the loss of RAN1 results in the dysfunction of ethylene receptors because of the lack of copper cofactor for the normal functioning of ethylene receptors (Woeste and Kieber, 2000). Being localized to distinct compartments of the endomembranes, the mechanism by which RAN1 would deliver Cu(I) to ETR1 is unknown.

RTE1 encodes a putative membrane protein, conserved among higher eukaryotes, and is isolated from a suppressor screen against the dominant etr1-2 receptor mutant (Resnick et al., 2006). The tomato (Solanum lycopersicum) GR gene, an ortholog of Arabidopsis RTE1, is isolated from a dominant nonripening mutant that has elevated GR expression (Barry and Giovannoni, 2006). The overexpression of RTE1 also causes the repression of ethylene responses in wild type, but not in the loss-of-function etr1-7 mutant (Resnick et al., 2006), raising a question of whether the wild-type RTE1 function is also affected by the etr1-7 mutation. Besides, five of the Arabidopsis ethylene receptor genes are functionally redundant; whether the loss of any of the other receptor genes would also inactivate the function of RTE1 overexpression needs to be studied to further address the specificity of RTE1 to the ETR1 function.

The etr1-2 mutation (102^Ala→Thr) stabilizes the transmitter-on state II conformation, which allows ethylene binding to the etr1-2 protein but does not abolish receptor activity (Wang et al., 2006). Although rte1 suppresses etr1-2, it does not have any effect on other dominant receptor mutations, including dominant etr1 alleles, leaving it unanswered whether the function of RTE1 is specific to particular receptor mutations or to the identity of receptor genes (Resnick et al., 2006). Little is known if RTE1 would function together with other receptors.

In this study, we comprehensively analyzed genetic interactions of RTE1 and ethylene receptor genes in loss-of-function mutants and RTE1 overexpression lineages. Domains that could be important to the function of ETR1 and RTE1 were studied through the expression of truncated etr1 and rte1 variants. Subcellular localization of the RTE1 and rte1-2 proteins was shown by expressing functional GFP fusions in Arabidopsis and onion (Allium cepa) epidermal cells. Possible mechanisms by which ethylene responses are regulated by ETR1 and RTE1 are discussed.

RESULTS

The Function of RTE1 Overexpression Is Primarily Dependent on ETR1

Ethylene insensitivity conferred by the RTE1 overexpression is almost masked by the etr1-7 null mutation (Resnick et al., 2006). Arabidopsis has five ethylene receptor genes and they are genetically redundant. We investigated if the loss of the other four receptor gene would also inactivate the function of RTE1 overexpression.

The 35S promoter-driven genomic RTE1 clone (35S

gRTE1) was transformed to ers1-2 etr2-3 ein4-4 ers2-3, in which ETR1 is the only wild-type receptor gene. For three individual transformation lines examined, 35S

gRTE1 conferred ethylene insensitivity in wild type and ers1-2 etr2-3 ein4-4 ers2-3 in the seedling triple-response assay (Fig. 1, A and B). As a control, the untransformed wild-type and quadruple mutants were ethylene responsive and displayed growth inhibition in the seeding hypocotyl and primary root.

Figure 1.

Ethylene insensitivity conferred by the RTE1 overexpression is ETR1 dependent and can be independent on the other four receptor genes. Hypocotyl measurement (A) and phenotype of ethylene-treated seedlings (B). C, Phenotype of light-grown seedlings. D, (more ...)

Ethylene insensitivity conferred by 35S

gRTE1 was next examined in light-grown seedling. The ers1-2 etr2-3 ein4-4 ers2-3 mutant exhibited small cotyledons and short primary root when germinated under light (data not shown). As shown in Figure 1C, light-grown 35S

gRTE1 ers1-2 etr2-3 ein4-4 ers2-3 transformation lines exhibited normal growth and expanded cotyledons compared with wild type.

Because ethylene insensitivity conferred by the overexpression of RTE1 is substantially inactivated by etr1-7, and our results show that the inactivation of the other four receptor genes does not perturb the function of the 35S

gRTE1 transgene, ethylene insensitivity conferred by the RTE1 overexpression is primarily dependent on ETR1. On the other hand, those lines represent individual transformation events; degrees of the 35S

RTE1 function affected by the etr1-7 mutation cannot be quantified. To score effects of the inactivation of ETR1 on the function of RTE1 overexpression, the 35S

gRTE1 transgene was genetically introgressed from individual wild-type transformation lines to etr1-7. As a result, the 35S

gRTE1 transgene was expressed on the same locus in wild type and etr1-7; effects of the etr1-7 mutation on the function of RTE1 overexpression can be quantitatively evaluated.

In one cross, the resulting F1 seedlings exhibited mild ethylene responses (data not shown). In the F₂ generation, 39 individual seedlings were classified into three groups according to degrees of ethylene responses (Fig. 1D). According to this classification, we further scored 174 individual F2 seedlings, of which 61 individuals fell into the longest group, 46 into the second, and 67 into the shortest. Because the F1 parent had a reduced hypocotyl length when germinated in ethylene, we suspected that the second group would be 35S

gRTE1/−;ETR1/etr1-7. Analyzed by χ² test, this was an expected 5:4:7 segregation (Fig. 1E). These results suggest that dosages of the 35S

gRTE1 transgene and wild-type ETR1 synergistically determine degrees of ethylene insensitivity.

To further study how the copy number of 35S

gRTE1 and ETR1 would determine degrees of ethylene responses, three individual introgression lineages carrying different copy numbers of 35S

gRTE1 and ETR1 were scored for the seedling hypocotyl length (Fig. 1F). Lineages homozygous in 35S

gRTE1 and ETR1 (+/+; +/+) exhibited the longest seedling hypocotyl. The loss of one copy of 35S

gRTE1 (+/−; +/+) caused a mild reduction in the hypocotyl length. The loss of one copy of ETR1 (+/+; +/−) caused a greater reduction. Seedling hypocotyl of heterozygous lineages (+/−; +/−) was about 40% shorter than that of homozygous lineages (Fig. 1, G and H). When ETR1 was replaced by etr1-7, the hypocotyl was slightly longer than etr1-7 (P < 0.05). These results further indicate that not only is ETR1 gene but also the ETR1 dosage essential to the function of RTE1 overexpression.

The Wild-Type RTE1 Function Is ETR1 Dependent throughout Developmental Stages

Specificity of the function of RTE1 overexpression to ETR1 was reciprocally demonstrated in the quadruple and etr1-7 mutants. However, the function of RTE1 overexpression could be hypermorphic or neomorphic; roles of receptor genes in the wild-type RTE1 function will need to be examined in loss-of-function mutants. In this study, the rte1-2 mutation was introgressed to receptor mutants to examine if the inactivation of any receptor genes would result in the inactivation of the wild-type RTE1 gene. Because rte1 mutants phenocopy etr1-7 (Resnick et al., 2006), the etr1-7 mutation was also introgressed to those receptor mutants to examine if ETR1 and RTE1 would have same effects on the repression of ethylene responses.

Mutants defective in RTE1 and multiple receptor genes were scored for seedling and adult phenotypes and degrees of ethylene responses were scored by the seedling triple-response assay (Fig. 2, A and C). To eliminate endogenous ethylene production, the ethylene biosynthesis inhibitor l-α-(2-aminoethoxyvinyl) Gly (AVG) was included in the seedling triple-response assay (Fig. 2, B and D). The dark-grown etr2-3 ein4-4 displayed mild ethylene responses; the etr1-7 etr2-3 ein4-4 mutant exhibited stronger ethylene responses than rte1-2 etr2-3 ein4-4. The light-grown etr2-3 ein4-4 seedling was slightly epinastic and the primary root was shorter than that of wild type. rte1-2 etr2-3 ein4-4 and etr1-7 etr2-3 ein4-4 were smaller in cotyledons and rosette leaves than etr2-3 ein4-4 (Fig. 2, E and F). Degrees of ethylene responses enhanced by the introgression of etr1-7 to etr2-3 ein4-4 ers2-3 and ers1-2 etr2-3 ein4-4 ers2-3 were much greater than that enhanced by the introgression of rte1-2 throughout developmental stages (Fig. 2, A, B, E, and F). The ers1-2 mutant was phenotypically identical to wild type; the introgression of rte1-2 to ers1-2 resulted in enhanced ethylene responses while etr1-7 ers1-2 exhibited extreme phenotypes (Fig. 2, A, B, E, and F). These data suggest that, in a same genetic background, the loss of ETR1 will cause stronger ethylene responses than the loss of RTE1 and that both genes are a negative regulator of ethylene responses. Besides, rte1 does not simply phenocopy etr1-7 as previously described (Resnick et al., 2006).

Figure 2.

The wild-type function of RTE1 is ETR1 dependent throughout development stages. To illustrate phenotypic changes after the respective introgression of rte1-2 and etr1-7 to receptor mutants, common mutations in each set of comparison are highlighted in (more ...)

Although the inactivation of RTE1 leads to derepression in ethylene responses, the introgression of the rte1-2 mutation to etr1-7 ers2-3 did not alter degrees of ethylene responses in the seedling triple-response assay as well as in other developmental stages (Fig. 2). This result is consistent with the previous study that rte1 phenotypically resembles etr1 rte1 mutants (Resnick et al., 2006). These data suggest that the wild-type RTE1 cannot repress ethylene responses in the absence of ETR1.

The RTE1 Gene Is Expressed Ubiquitously and Has Two Transcripts

Northern analysis shows that the RTE1 transcript accumulates 2.5 h after ethylene treatment (Resnick et al., 2006). LUCIFERASE (LUC) is labile upon the reaction with its substrate, luciferine, and has been used as a real-time reporter (Xiong et al., 1999). Whole-plant expression pattern of the RTE1 gene in Arabidopsis was next examined in transformation lines expressing LUC, driven by the RTE1 promoter.

In dark-grown seedlings, fluorescence was the highest in cotyledons and weaker in hypocotyl and root (Fig. 3A). Figure 3B shows 3-week-old plants that displayed the strongest fluorescence in growing young leaves (Fig. 3C). Figure 3D shows that developing leaves, leaf veins, rachis, and flowers exhibited strong fluorescence in an adult plant. Tissues between veins in expanded leaves exhibited weaker fluorescence. For light-grown seedlings, cotyledons and developing leaves exhibited the highest fluorescence than hypocotyls and roots (Fig. 3E).

Figure 3.

The RTE1 gene is expressed throughout developmental stages and two transcripts are identified. The LUC fluorescence patterns of dark-grown seedlings (A), light-grown rosettes (C), adult plant (D), and light-grown seedlings (E) expressing the RTE1p

(more ...)

In northern analysis, the RTE1 hybridization signal appears as a doublet (Resnick et al., 2006), implying the existence of transcripts with two lengths. To further characterize the two transcripts, sequence of the 3′ end of the RTE1 transcript was examined by the 3′ RACE, and we did not identify any alternative polyadenylation site (data not shown). Examined by the 5′ RACE, the RTE1 transcript gave two major amplification signals (Fig. 3F) and the resulting cDNA fragments were sequenced. For three independent 5′ RACE and sequencing, we identified that sequence of the short fragment starts in the second intron and that the long fragment represents the full-length RTE1 transcript (Fig. 3G).

Our data show that RTE1 is globally expressed throughout developmental stages and has two transcripts.

The RTE1 N Terminus Can Be Dispensable to the Function of the Dominant etr1-2 Receptor

Although RTE genes are prevalent in higher eukaryotes, domains or motifs of known function are not found in RTE1. RTE1 shares poor similarity with other plant RTE-like proteins in the N terminus (Barry and Giovannoni, 2006; Resnick et al., 2006), implying that the N terminus could exhibit functional specificity or be not essential to the RTE1 function. Because the dominant etr1-2 activity is RTE1 dependent (Resnick et al., 2006), functional significance of the RTE1 N terminus can be examined in etr1-2 rte1-2 by expressing mutant rte1 isoforms with deletions in the N terminus.

In this study, two rte1 variants with N-terminal deletions were each expressed in etr1-2 rte1-2 and the resulting transformation lines were scored for the seedling triple response. We found that each of the overexpression of rte1(NΔ25) and rte1(NΔ49), encoding rte1 proteins with, respectively, a deletion of 25 and 49 amino acid residues in the N terminus (Fig. 4A), was able to restore ethylene insensitivity in etr1-2 rte1-2. For three individual transformation lines scored, degrees of ethylene insensitivity in 35S

rte1(NΔ25) etr1-2 rte1-2 and 35S

rte1(NΔ49) etr1-2 rte1-2 lines were comparable to that in 35S

gRTE1 etr1-2 rte1-2 lines (Fig. 4, B and C).

Figure 4.

The overexpression of the rte1(Δ25) and rte1(Δ49) variants restore ethylene insensitivity in etr1-2 rte1-2. A, Schematic structures of the wild-type RTE1 and artificially mutated rte1 transgenes. B, The seedling triple-response phenotype (more ...)

These results imply that the RTE1 N terminus may not exhibit functional specificity; rather, nearly 20% of the RTE1 N terminus is not essential to the etr1-2 function. Because the loss of the second transmembrane domain inactivates RTE1 (Resnick et al., 2006), the C terminus cannot be dispensable.

The ETR1 N Terminus Is Sufficient to the Functioning of RTE1 in the Absence of the Wild-Type ETR1

Our data show that ETR1 is essential to the RTE1 function; domains on ETR1 that could be essential to the functioning of RTE1 were next examined.

The etr1(1–609) transgene encodes a receiver domain-lacking etr1 variant and etr1(1–349) encodes an etr1 variant that has the three transmembrane domains (TMDs) and the GAF domain (Xie et al., 2006). 35S

gRTE1 etr1-7 is an introgression line obtained by genetic cross (L6; Fig. 1, G and H) and is ethylene responsive. This line was individually transformed with gETR1, etr1(1–609), and etr1(1–349), and each transgene was driven by the native ETR1 promoter (Wang et al., 2003; Xie et al., 2006). Individual lines of the resulting transformants, T:gETR1 35S

gRTE1 etr1-7(L6), T:etr1(1–609) 35S

gRTE1 etr1-7(L6), and T:etr1(1–349) 35S

gRTE1 etr1-7(L6) were subjected to the seedling triple-response assay. Because the 35S

gRTE1 transgene in these transformation lines was expressed on a same locus, degrees of ethylene response in these lines were determined by the ETR1/etr1 transgene. Germinated in dark, they were all ethylene insensitive and exhibited a long seedling hypocotyl in ethylene (Fig. 5, A and B). The hypocotyl lengths of these transformation lines were similar, indicating that etr1(1–609) and etr1(1–349) have a same effect as ETR1 on the recovery of ethylene insensitivity conferred by 35S

gRTE1.

Figure 5.

Ethylene insensitivity conferred by 35S

gRTE1 is restored by truncated etr1 variants in etr1-7. A, The seedling triple-response phenotype and hypocotyl measurement of the 35S

gRTE1 etr1-7 introgression line (B), respectively, expressing (more ...)

These results show that the ETR1 N terminus, including the three transmembrane domains and the GAF domain, is sufficient to the functioning of RTE1 while ETR1 His-kinase and receiver domains can be dispensable.

The GFP-RTE1 Fusion Protein Is Localized to the Golgi and Causes Ethylene Insensitivity

In an attempt to identify the subcellular localization of the RTE1 protein, each of the RTE1 N and C termini was fused with the GFP and ectopically expressed under the control of the native RTE1 or 35S promoter.

The RTE1-GFP fusion, of which the RTE1 C terminus was fused with GFP, failed to cause ethylene insensitivity in wild type and to rescue the rte1-2 mutation in rte1-2 etr1-2 (data not shown). The GFP-RTE1 fusion, of which GFP was fused to the N terminus of RTE1, was able to cause some degrees of ethylene insensitivity in wild type and partially rescue the rte1-2 mutation in rte1-2 etr1-2 (Fig. 6, A and B). The expression of 35S

GFP-rte1-2 failed to, respectively, restore and confer ethylene insensitivity in etr1-2 rte1-2 and wild type (data not shown).

Figure 6.

The GFP-RTE1 transformation lines are partially ethylene insensitive and the GFP-RTE1 protein localizes to the Golgi. Phenotype (A) and hypocotyl measurement (B) for seedlings expressing GFP-RTE1 in wild type (Col-0) and etr1-2 rte1-2. C, Water and ethylene (more ...)

Fluorescence patterns of 35S

GFP-RTE1- and RTE1p

GFP-RTE1-transformed wild type were similar to that of the 35S

GFP-rte1-2 transformed, except that RTE1p

GFP-RTE1 gave a weaker fluorescence signal (see Supplemental Fig. S1). These results indicate that the GFP-RTE1 fusion protein is functional and dominant, and that its subcellular localization could be sites where RTE1 exerts its function. Besides, the rte1-2 mutation may cause dysfunction of RTE1, rather than mislocalization as previously proposed (Resnick et al., 2006).

Laser scanning confocal microscopy (LSCM) shows that the green fluorescence signal of GFP-RTE1 was fast moving, suggesting that its movement is associated with cellular structures in the cytoplasm (Fig. 6C). Because RTE1 encodes a membrane protein, we next examined if RTE1 may associate with any endomembranes, including the ER and Golgi. GFP-ER, an ER-localized marker, displayed a unique reticular fluorescence pattern that is distinct to GFP-RTE1 and EYFP-NAG (see Supplemental Fig. S2), of which EYFP-NAG was a marker for the Golgi localization (Xu and Scheres, 2005). When GFP-RTE1 and the Golgi-localized EYFP-NAG were coexpressed in onion epidermal cells, both GFP-RTE1 and EYFP-NAG displayed fast movement and the fluorescence visually overlapped (Fig. 6D).

Because the RTE1 function is ETR1 dependent and ETR1 is localized to the ER, we next examined if trace amounts of RTE1 would localize to the ER. By the LSCM, a three-dimensional (3-D) fluorescence pattern was constructed. GFP-RTE1 displayed a similar 3-D pattern as EYFP-NAG and the fluorescence largely located to the Golgi. In addition, weak reticular fluorescence was also visible in small regions (Fig. 6E). This result suggests that GFP-RTE1 is mainly localized to the Golgi; very small amounts of GFP-RTE1 and EYFP-NAG may reside in the ER.

According to its size (in micrometers), fast-moving feature, and colocalization with EYFP-NAG, RTE1 appears to predominantly associate with the Golgi. Brefeldin A (BFA) is a fungal toxin and can cause the accumulation of the Golgi, the so-called BFA compartments, and the redistribution of Golgi membranes into the ER in plant cells (Nebenfuhr et al., 2002; Ritzenthaler et al., 2002; Jurgens, 2004). We next examined if the GFP-RTE1 fusion would respond to BFA treatment. Over a period of 60 min BFA treatment, the GFP-RTE1 fluorescence gradually aggregated (Fig. 6C); as a mock, the 0.5% ethanol, a solvent for BFA, did not alter the distribution of GFP-RTE1 fluorescence in Arabidopsis (Fig. 6C).

Effects of ethylene treatment on the RTE1 protein were next examined. 1-Aminocyclopropane-1-carboxylic acid (ACC) is the immediate precursor of ethylene biosynthesis and has been used to replace the ethylene treatment. For a period of 60 min ACC treatment, the movement, subcellular localization, and fluorescence intensity of the green fluorescence were not noticeably altered in seedlings expressing the 35S

GFP-RTE1 transgene (Fig. 6C). For prolonged (3 h) ACC treatments, the GFP fluorescence intensity gradually dropped; as a comparison, the water treatment gave a similar drop in the GFP fluorescence in Arabidopsis (Fig. 6F). Our data suggest that the ethylene (ACC) treatment dose not regulate RTE1 at protein levels.

DISCUSSION

etr1-7 and rte1 mutants are phenotypically identical and it is hypothesized that RTE1 is largely required for the wild-type ETR1 function (Resnick et al., 2006). Our data, however, show that the function of RTE1 is ETR1 dependent and that ETR1 has a greater effect than RTE1 on the repression of ethylene responses. This discrepancy is not mutually exclusive but could be caused by ways that the rte1 and etr1 mutant phenotypes were scored. For example, subtle effects caused by the rte1 and etr1 mutations on ethylene responses become prominent in multiple mutants but not in single mutants.

Because the inactivation of ETR1 masks the RTE1 function, it is likely that RTE1 has a regulatory role in the wild-type ETR1 activity and that the loss of RTE1 weakens, but does not nullify, the ETR1 receptor signal output. As a result, the introgression of rte1-2 to receptor mutants, except for etr1-7, leads to enhanced ethylene responses. Without ETR1, RTE1 becomes an orphan component and cannot exert repression in ethylene responses; thus, the introgression of rte1 to etr1-7 mutants had little effect on ethylene responses. In other words, the etr1-7 mutation indeed results in the inactivation of both ETR1 and RTE1 while rte1 mutations cause the inactivation of RTE1 as well as the weakening of the ETR1 activity. This hypothesis further explains why the rte1-3 ers1-3 mutant would not exhibit the extremely severe etr1 ers1 mutant phenotype.

Mutational analysis suggests that ETR1 exists in transmitter-on and transmitter-off states. The transmitter-on ETR1 can be either free of ethylene (state I) or bound with ethylene (state II). The transition of ETR1 from state II to state III (transmitter off) abolishes the ETR1 receptor activity. The etr1-2 mutation (102^{Ala → Thr}) could result in the stabilized transmitter-on state II in the etr1-2 protein but does not abolish ethylene binding (Wang et al., 2006). Because the ethylene-insensitive etr1-2 mutant becomes ethylene responsive once RTE1 is not available, RTE1 may be essential to the stabilization of the transmitter-on state II in the etr1-2 protein. The midregion of RTE1 may be sufficient to the stabilization of etr1-2 in state II because the overexpression of NΔ49rte1 restored ethylene insensitivity in etr1-2 rte1-2.

Conceivably, in wild type, endogenous RTE1 protein is not sufficient to the stabilization of ethylene-bound ETR1 in the transmitter-on states; thus, ethylene binding converts the ETR1 protein to the transmitter-off state III. When overexpressed, the RTE1 protein may exceed a threshold that would be sufficient to, constitutively, maintain the ethylene-bound ETR1 protein in an active state, leading to ethylene insensitivity. This hypothesis is in agreement with our finding that the copy number of the 35S

gRTE1 transgene and ETR1 receptor gene determines degrees of ethylene insensitivity. In the 35S

gRTE1/− lineages (+/−; +/+) the RTE1 level could be lower but still exceeds a threshold to stabilize majority of the ETR1 protein in the transmitter-on states. For lineages in which one copy of ETR1 is lacking, levels of the active ETR1 in active states are reduced, leading to substantial weakening in ethylene insensitivity even if an excessive amount of the RTE1 protein is available. When single copy of each 35S

gRTE1 and ETR1 genes are both missing (+/−; +/−), levels of transmitter-on ETR1 become even lower and degrees of ethylene insensitivity are reduced. Possibly, excessive amount of the RTE1 protein nonspecifically maintains a minor fraction of the other receptors in an active state; as a result, the 35S

gRTE1 etr1-7 lineages exhibit weak ethylene insensitivity. The dominant etr1-1 protein does not bind ethylene and thus stays in a constitutive transmitter-on state that is completely independent on RTE1. Although the RTE1 transcript accumulates upon the ethylene treatment, the amount of elevation appears modest and insufficient to alter ethylene responses. However, the minor up-regulation in the RTE1 expression could fine tune ethylene responses.

Because the expression of etr1(1–349) restored ethylene insensitivity in 35S

gRTE1 etr1-7, the ETR1 N terminus could be a target of RTE1. Previously we demonstrated that the receptor signal output mediated by the truncated etr1(1–349) is dependent on subfamily I receptors (Xie et al., 2006). Conceivably, the overexpression of RTE1 may activate the etr1(1–349) fragment which, in turn, represses ethylene responses through interaction with the subfamily I receptor ERS1. Although the cellular fragment of RTE1 and ETR1 can physically associate in vitro (C.-K. Wen, unpublished data), whether they can physically associate in vivo will need to be demonstrated. We do not exclude the possibility that a third component could be required for the activation of ETR1/etr1(1–349) by RTE1. In addition to RTE1 and RAN1, ETR1 function could be regulated by multiple factors. For example, POLARIS encodes a small peptide and negatively regulates ethylene responses, probably acting upstream of or at ETR1 (Chilley et al., 2006).

Our data suggest that RTE1 is ubiquitously expressed based on the LUC fluorescence pattern. Because the ETR1 promoter-driven LUC also shows a whole-plant expression pattern (C.-K. Wen, unpublished data), both RTE1 and ETR1 may express concurrently. These results are in agreement with the fact that the etr1-2 function is RTE1 dependent and that the etr1-2 mutation leads to whole-plant ethylene insensitivity (Resnick et al., 2006). In contrast to Arabidopsis, the overexpression of GR, a tomato RTE1 ortholog, only confers ethylene insensitivity in certain, but not all, tissues (Barry and Giovannoni, 2006). Conceivably, like RTE1, the GR function would be specific to certain ethylene receptor genes in tomato; for tissues with low expression in those GR-dependent receptor genes, elevated GR would be an orphan component and not sufficient to the repression in ethylene responses. On the other hand, GR and GRL-1 are in a same clade with RTE1 in a phylogenic analysis (Barry and Giovannoni, 2006) and they could coordinately regulate ethylene responses in different tissues together with specific ethylene receptors. We hypothesize that tomato, after a long breeding, may have evolved a unique mechanism in the regulation of ethylene responses in specific tissues through interactions between GR/GRL-1 and certain ethylene receptors.

The identification of the short RTE1 transcript implies that the RTE1 transcript could be spliced or that the RTE1 locus could have two genes. Although functional significance of the existence of two RTE1 transcripts is unknown, the formation of the short transcript does not appear to be a result of random degradation because it begins at specific site within the second intron. On the other hand, the overexpression of the RTE1 cDNA clone confers ethylene insensitivity in wild type (Resnick et al., 2006), suggesting that the short transcript would not be essential to the RTE1 function. The formation of the short transcript could represent a regulation of the RTE1 expression. Alternatively, the short transcript may exhibit independent function or could be expressed within the same locus as RTE1, as a second gene.

The subcellular localization of GFP fusions shows that both RTE1 and rte1-2 are associated with the Golgi. Computational prediction (Yuan and Teasdale, 2002) suggests that RTE1 and rte1-2 are a Golgi type II membrane protein (data not shown). This topology will allow the large RTE1 N terminus to face the cytosolic side. The TMD of Golgi type II proteins is essential to its retention to the Golgi (Saint-Jore-Dupas et al., 2004). Because the GFP-rte1-2 protein, in which the second TMD is lacking, also localizes to the Golgi, the first TMD of RTE1 would be essential to the retention to the Golgi while the second TMD may have a role in proper folding of the RTE1 protein to a functional conformation.

The Golgi complex has three functionally distinct subcompartments, the cis-Golgi, the Golgi stacks or medial Golgi, and the trans-Golgi network. Our data show that the BFA treatment caused the formation of the BFA compartment and did not result in the redistribution of GFP-RTE1 to the ER, implying that majority of the RTE1 protein is retained to the Golgi stack. RAN1 is essential to the function of wild-type ethylene receptors and believed to localize to the post-Golgi (Hirayama et al., 1999). Computational prediction (Yuan and Teasdale, 2002) also suggests that RAN1 is a protein in the post-Golgi (data not shown). Although both RAN1 and RTE1 localize to the Golgi and are involved in the regulation of ethylene responses, they may not function in a same subcompartment in the Golgi complex. Besides, the RTE1 function is specific to ETR1 while RAN1 is essential to the function of wild-type receptors; their roles in ethylene responses may be distinct.

The ETR1 receptor protein is shown to localize to the ER (Chen et al., 2002). Topological study on the melon (Cucumis melo) ERS1 receptor protein (CmERS1) suggests that the large C-terminal portion faces the cytosolic side and the N terminus locates in the ER lumen (Ma et al., 2006). Conceivably, ETR1 may exhibit a similar topology, which allows its C terminus to be exposed to interacting components. In the 3-D image, very small amounts of GFP-RTE1 was visible in the ER. Because the 3-D image was made from stacks of images, weak fluorescence signal can thus become visible. It is likely that a small fraction of RTE1 could reside in or retrograde to the ER; alternatively, a continuum could exist between the ER and Golgi (Hawes and Satiat-Jeunemaitre, 2005). In either scenario, ETR1 and a small fraction of RTE1 could colocalize. Besides, Golgi appears to attach to and moves over the ER, which would allow direct or indirect interaction between ETR1 and RTE1. The endomembranes and protein movement have been demonstrated important to signaling (Friml et al., 2004; Surpin and Raikhel, 2004; Scheres and Xu, 2006). RAN1, RTE1, and ETR1 are mainly localized to distinct compartments of the Golgi and ER; the protein/membrane transportation, via the retrograde and the anterograde traffic, between subcompartments of endomembranes could play a role in their function as well as in the regulation of ethylene signal transduction.

Ethylene plays important roles in growth and development in higher plants. Little is known about its effects on lower plants and when plants would have acquired the usage of ethylene as a plant hormone during evolution. It is shown that lower plants and some cyanobacteria can bind ethylene. Interestingly, plant species exhibit a similar ethylene-binding capacity (Wang et al., 2006). RTE1 homologs can be identified in bryophyte (C.-K. Wen, unpublished data), suggesting that regulatory roles of RTE1 in ethylene responses may have been acquired in ancestors of land vascular plants, at least, by the early Silurian (430 million years ago). On the other hand, RTE genes are only identified in higher eukaryotes while plant RTE is distant to animal RTE in phylogenic analysis (Barry and Giovannoni, 2006); we hypothesize that RTE1 may have specialized functions in plant species.

MATERIALS AND METHODS

Plant Material and Growth

rte1-2 and etr1-2 rte1-2 were from C. Chang (Resnick et al., 2006). etr1-7 ers2-3, rte1-2 etr1-7 ers2-3, rte1-2 etr2-3 ein4-4, rte1-2 etr2-3 ein4-4 ers2-3, rte1-2 ers1-2 etr2-3 ein4-4 ers2-3, etr1-7 etr2-3 ein4-4 ers2-3, and etr1-7 ers1-2 etr2-3 ein4-4 ers2-3 were obtained by genetic cross and verified by genotyping and sequencing of receptor genes and RTE1. Genotyping of the receptor genes was performed as described (Hua and Meyerowitz, 1998; Hall and Bleecker, 2003; Resnick et al., 2006; Xie et al., 2006). For the seedling triple-response assay, seeds were stratified at 4°C for 72 h and then germinated at 22°C for 80 h in the dark with or without 20 μL L⁻¹ of ethylene gas. Ethylene concentrations were measured by gas chromatography (Angilent G2070 ChemStation; Angilent Technologies). For experiments in which endogenous ethylene production was blocked, 10 μm AVG was included. Plants were grown at 16 h light/8 h dark under fluorescent light at 24°C.

Seedling Measurement and Statistics

For seedling measurement, a photo of dark-grown seedlings was taken and hypocotyl lengths were measured by VideoTesT (Moscow). Statistics for data analyses were followed as described (Xie et al., 2006) and an error rate of 0.001 (unless specified) was used for comparisons throughout this study. Difference between means (for t test and lsd) is considered significant only when the P value is smaller than 10⁻³. Measurement was represented as the 95% confidence interval of a mean.

Clones and Plant Transformation

The genomic RTE1 was obtained by PCR using the primer set gRTE1-F (5′-ATGGATCCGGTTCATTGTACCTTTCTCC-3′) and gRTE1-R (5′-ATGGATCCTCAAGTAATTATGTTCTTAAAACAGTA-3′) and the resulting DNA was cloned to the binary vector pCGN18. The truncated NΔ25-rte1 and NΔ49-rte1 clones were made by PCR described as the following. An RTE1 fragment, namely RTE1(−556) that contains the first exon and intron of RTE1, was first cloned by PCR using primer set gRTE1-F and gRTE1-556-R (5′-AGTCTAGATTCCTAATCACACAAGACAAG-3′). RTE1 fragment amplified by the primer set gRTE1Δ0-F (5′-agTCTAGAAAATGTCACGTGGAAGAGGAGT-3′) and gRTE1-R was cloned to the XbaI site in RTE1(−556), giving rise to the NΔ0RTE1 clone (no deletion). PCR fragment generated by the primer set gRTE1Δ75-F (5′-AGTCTAGAAAATGTCTATACCATCAATAATCGAA-3′) and gRTE1-R was cloned to the XbaI site of RTE1(−556), giving rise to the NΔ25rte1 clone that has a deletion of 25 amino acid residues. PCR fragment generated by the primer set gRTEΔ147-F (5′-AGTCTAGAAAATGAAATTTCCTTGCTGTATAGTT-3′) and gRTE1-R was cloned to the XbaI site of RTE1(−556), giving rise to the NΔ49rte1 clone that has a deletion of 49 amino acid residues. All these clones were verified by sequencing and then subcloned to the binary vector pCGN18. The genomic ETR1, etr1(1–609), and etr1(1–349) clones are as described (Xie et al., 2006). The EYFP-NAG clone was a gift from B. Scheres. The ER-GFP clone was from the Arabidopsis Biological Resource Center (stock no. CD3-420). Cloning of GFP-RTE1 and GFP-rte1-2 are described as the following. The RTE1 cDNA was cloned by PCR using a cDNA template (C.-K. Wen, unpublished data) and the primer set cRTE1-F (5′-ATGGATCCATGTCACGTGGAAGAGGA-3′) and cRTE1-R (5′-ATGGATCCTCAAGTAATTATGTTCTTAAAACAGTA-3′). The resulting clone was restricted by BamHI and cloned to the BamHI site on pRTL2 (that has the 35S

mGFP clone) and the resulting GFP-RTE1 cloned was then released by HindIII and subcloned to pCGN1547. The GFP-rte1-2 clone was constructed following the same procedure, except that the rte1-2 cDNA was used as the template for PCR. The RTE1 promoter, namely RTE1p, was cloned by PCR using the primer set RTE1-P-R (5′-CCAAGCTTTTAGATTCCTAATCACACAACAAGAC-3′) and RTE-P-F (5′-AGGTTCTGAATGGTTGCATGTAGAG-3′). LUC was cloned to the BamHI site following the native RTE1 promoter, giving rise to the RTE1p

LUC clone. Arabidopsis (Arabidopsis thaliana) was transformed by flower dip according to Clough and Bent (1998).

Reverse Transcription PCR and Sequencing

RNA was isolated as described (Wen et al., 1999). For the 3′ RACE, an anchored poly(A) primer was used for the first-strand cDNA synthesis. For the 5′ RACE, the gene-specific primer RTE1-825-R1 (5′-CCAGGCAAAGAGTAGTAGCGAG-3′) was used for the first-strand cDNA synthesis and ploy(G) was added to the 5′ end by terminal deoxyribonucleotide transferase. Primers RTE1-752-R2 (5′-GACGTGACCACAGCACATGGCA-3′) and RTE1-629-R3 (5′-AGACGGTTCAAACAGTTTGCAA-3′) were, respectively, used for the first and second PCR amplification. The RACE products were purified after fractionation by gel electrophoresis and subjected for sequencing. For the 5′ RACE and sequencing, three individual repeats were performed.

Imaging for LUC Expression

The fluorescence generated by LUC was captured by a cold CCD cooled by liquid nitrogen to −110°C (VersArray System, Roper Scientific). Plants to be imaged were pretreated with luciferin (1 mm in 0.01% Triton X-100) for 18 h and retreated before ethylene treatment and fluorescence imaging.

Microscopy for GFP Imaging

LSCM (Carl Zeiss LSM510 META) was performed at the Facility Center of the National Laboratory of Plant Molecular Genetics in the Shanghai Institute of Plant Physiology and Ecology and the Two-Photon Microscopy Facility (Carl Zeiss LSM510 META) in the Institute of Neuroscience of Shanghai Institutes for Biological Sciences. For the detection of the coexpression of YFP and GFP fusions, fluorescence was separated by linear unmixing to eliminate the fluorescence bleed through. Separation of the green and yellow fluorescence by the linear unmixing algorithm was confirmed; the GFP expression did not show fluorescence signal in the YFP channel and the EYFP-NAG expression did not give signal in the GFP channel (see Supplemental Fig. S3). Colocalization was visually evaluated by merging images of the GFP and YFP. Fluorescence microscopy was performed using Nikon Eclipse E400 (Nikon). Fluorescence filter sets (FITC-3540B-NQF-ZERO, TXRED-4040B-NQF-ZERO, CFP-2432A-NQF, and YFP-2427A-NQF-ZERO) were from Laser 2000 Ltd. Dark-grown seedlings were subjected to microscopy. For BFA treatment, seedlings were immersed in 10 μm of BFA (dissolved in 0.5% ethanol) and subjected to confocal microscopy; 0.5% ethanol was used as a blank control. For ACC treatment, dark-grown seedlings were immerged in 10 μm ACC (dissolved in water) and subjected to confocal microscopy. For the quantitation of the GFP fluorescence in ACC- and water-treated seedlings, images of fluorescent seedlings were taken every 60 min and the intensity was measured. For imaging of the movement of GFP fluorescence, photos were taken every 6 s. For imaging of BFA and ethanol treatments, frames were taken every 3 min.

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure S1. Fluorescence patterns of GFP-RTE1 and GFP-rte1-2 are similar.
Supplemental Figure S2. Fluorescence of ER-GFP, GFP-RTE1, and EYFP-NAG in onion epidermal cells.
Supplemental Figure S3. The separation of the GFP and EYFP signals by linear unmixing.

Supplementary Material

[Supplemental Data]

Click here to view.

Acknowledgments

We thank Drs. C. Chang for the etr1-2 rte1-2 and rte1-2 mutants and the rte1-2 cDNA clone, B. Scheres for the EYFP-NAG clone, J. Xu for advice on the separation of EYFP and GFP fluorescence, our colleagues H.X. Lin for χ² analysis and H.Q. Yang for the LUC clone, and Q. Wang for BFA treatment and confocal microscopy of the GFP-RTE1 lines. The linear unmixing algorithm was performed at the Two-Photon Microscopy Facility in the Institute of Neuroscience of Shanghai Institutes for Biological Sciences.

Notes

¹This work was supported by the Chinese Academy of Sciences, the National Natural Sciences Foundation of China (grant nos. 90408008 to C.-K. Wen and 30421001 to X.Y. Chen), and the Ministry of Science and Technology (grant no. 2002AA224021/2005AA227020 to C.-K. Wen).

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Chi-Kuang Wen (qgwen/at/sibs.ac.cn).

^[C]Some figures in this article are displayed in color online but in black and white in the print edition.

^[W]The online version of this article contains Web-only data.

www.plantphysiol.org/cgi/doi/10.1104/pp.107.104299

References

Barry CS, Giovannoni JJ (2006) From the cover: ripening in the tomato green-ripe mutant is inhibited by ectopic expression of a protein that disrupts ethylene signaling. Proc Natl Acad Sci USA 103: 7923–7928 [PubMed].
Chen YF, Randlett MD, Findell JL, Schaller GE (2002) Localization of the ethylene receptor ETR1 to the endoplasmic reticulum of Arabidopsis. J Biol Chem 277: 19861–19866 [PubMed].
Chilley PM, Casson SA, Tarkowski P, Hawkins N, Wang KLC, Hussey PJ, Beale M, Ecker JR, Sandberg GK, Lindsey K (2006) The POLARIS peptide of Arabidopsis regulates auxin transport and root growth via effects on ethylene signaling. Plant Cell 18: 3058–3072 [PubMed].
Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16: 735–743 [PubMed].
Francis MJ, Jones EE, Levy ER, Ponnambalam S, Chelly J, Monaco AP (1998) A Golgi localization signal identified in the Menkes recombinant protein. Hum Mol Genet 7: 1245–1252 [PubMed].
Friml J, Yang X, Michniewicz M, Weijers D, Quint A, Tietz O, Benjamins R, Ouwerkerk PBF, Ljung K, Sandberg G, et al (2004) A PINOID-dependent binary switch in apical-basal PIN polar targeting directs auxin efflux. Science 306: 862–865 [PubMed].
Guzman P, Ecker JR (1990) Exploiting the triple response of Arabidopsis to identify ethylene-related mutants. Plant Cell 2: 513–523 [PubMed].
Hall AE, Bleecker AB (2003) Analysis of combinatorial loss-of-function mutants in the Arabidopsis ethylene receptors reveals that the ers1 etr1 double mutant has severe developmental defects that are EIN2 dependent. Plant Cell 15: 2032–2041 [PubMed].
Hawes C, Satiat-Jeunemaitre B (2005) The plant Golgi apparatus—going with the flow. Biochim Biophys Acta 1744: 93–107 [PubMed].
Hirayama T, Kieber JJ, Hirayama N, Kogan M, Guzman P, Nourizadeh S, Alonso JM, Dailey WP, Dancis A, Ecker JR (1999) RESPONSIVE-TO-ANTAGONIST1, a Menkes/Wilson disease-related copper transporter, is required for ethylene signaling in Arabidopsis. Cell 97: 383–393 [PubMed].
Hua J, Meyerowitz EM (1998) Ethylene responses are negatively regulated by a receptor gene family in Arabidopsis thaliana. Cell 94: 261–271 [PubMed].
Jurgens G (2004) Membrane trafficking in plants. Annu Rev Cell Dev Biol 20: 481–504 [PubMed].
Ma B, Cui ML, Sun HJ, Takada K, Mori H, Kamada H, Ezura H (2006) Subcellular localization and membrane topology of the melon ethylene receptor CmERS1. Plant Physiol 141: 587–597 [PubMed].
Nebenfuhr A, Ritzenthaler C, Robinson DG (2002) Brefeldin A: deciphering an enigmatic inhibitor of secretion. Plant Physiol 130: 1102–1108 [PubMed].
Qu X, Schaller GE (2004) Requirement of the histidine kinase domain for signal transduction by the ethylene receptor ETR1. Plant Physiol 136: 2961–2970 [PubMed].
Resnick JS, Wen CK, Shockey JA, Chang C (2006) From the cover: REVERSION-TO-ETHYLENE SENSITIVITY1, a conserved gene that regulates ethylene receptor function in Arabidopsis. Proc Natl Acad Sci USA 103: 7917–7922 [PubMed].
Ritzenthaler C, Nebenfuhr A, Movafeghi A, Stussi-Garaud C, Behnia L, Pimpl P, Staehelin LA, Robinson DG (2002) Reevaluation of the effects of Brefeldin A on plant cells using tobacco bright yellow 2 cells expressing golgi-targeted green fluorescent protein and COPI antisera. Plant Cell 14: 237–261 [PubMed].
Saint-Jore-Dupas C, Gomord V, Paris N (2004) Protein localization in the plant Golgi apparatus and the trans-Golgi network. Cell Mol Life Sci 61: 159–171 [PubMed].
Schaller GE, Bleecker AB (1995) Ethylene-binding sites generated in yeast expressing the Arabidopsis ETR1 gene. Science 270: 1809–1811 [PubMed].
Scheres B, Xu J (2006) Polar auxin transport and patterning: grow with the flow. Genes Dev 20: 922–926 [PubMed].
Surpin M, Raikhel N (2004) Traffic jams affect plant development and signal transduction. Nat Rev Mol Cell Biol 5: 100–109 [PubMed].
Wang W, Esch JJ, Shiu SH, Agula H, Binder BM, Chang C, Patterson SE, Bleecker AB (2006) Identification of important regions for ethylene binding and signaling in the transmembrane domain of the ETR1 ethylene receptor of Arabidopsis. Plant Cell 18: 3429–3442 [PubMed].
Wang W, Hall AE, O'Malley R, Bleecker AB (2003) Canonical histidine kinase activity of the transmitter domain of the ETR1 ethylene receptor from Arabidopsis is not required for signal transmission. Proc Natl Acad Sci USA 100: 352–357 [PubMed].
Wen CK, Smith R, Banks JA (1999) ANI1: a sex pheromone-induced gene in ceratopteris gametophytes and its possible role in sex determination. Plant Cell 11: 1307–1318 [PubMed].
Woeste KE, Kieber JJ (2000) A strong loss-of-function mutation in RAN1 results in constitutive activation of the ethylene response pathway as well as a rosette-lethal phenotype. Plant Cell 12: 443–455 [PubMed].
Xie F, Liu Q, Wen CK (2006) Receptor signal output mediated by the ETR1 N terminus is primarily subfamily I receptor dependent. Plant Physiol 142: 492–508 [PubMed].
Xiong L, David L, Stevenson B, Zhu JK (1999) High throughput screening of signal transduction mutants with luciferase imaging. Plant Mol Biol Rep 17: 159–170.
Xu J, Scheres B (2005) Dissection of Arabidopsis ADP-RIBOSYLATION FACTOR 1 function in epidermal cell polarity. Plant Cell 17: 525–536 [PubMed].
Yuan Z, Teasdale RD (2002) Prediction of Golgi type II membrane proteins based on their transmembrane domains. Bioinformatics 18: 1109–1115 [PubMed].

Articles from Plant Physiology are provided here courtesy of
American Society of Plant Biologists