Skip to main content
PMC home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2006 Aug 18;103(36):13286–13293. doi: 10.1073/pnas.0605528103

ETHYLENE-INSENSITIVE5 encodes a 5′→3′ exoribonuclease required for regulation of the EIN3-targeting F-box proteins EBF1/2

Gabriela Olmedo *,†,‡, Hongwei Guo *,§,‡, Brian D Gregory *,‡, Saeid D Nourizadeh ¶,‡, Laura Aguilar-Henonin *,†, Hongjiang Li §, Fengying An §, Plinio Guzman *,†, Joseph R Ecker *,‖
PMCID: PMC1550774  PMID: 16920797

Abstract

Ethylene is a gaseous plant growth regulator that controls a multitude of developmental and stress responses. Recently, the levels of Arabidopsis EIN3 protein, a key transcription factor mediating ethylene-regulated gene expression, have been demonstrated to increase in response to the presence of ethylene gas. Furthermore, in the absence of ethylene, EIN3 is quickly degraded through a ubiquitin/proteasome pathway mediated by two F-box proteins, EBF1 and EBF2. Here we report the identification of ETHYLENE-INSENSITIVE5 as the 5′→3′ exoribonuclease XRN4. Specifically, we demonstrate that EIN5 is a component of the ethylene signal transduction cascade acting downstream of CTR1 that is required for ethylene-mediated gene expression changes. Furthermore, we find that the ethylene insensitivity of ein5 mutant plants is a consequence of the over-accumulation of EBF1 and EBF2 mRNAs resulting in the under-accumulation of EIN3 even in the presence of ethylene gas. Together, our results suggest that the role of EIN5 in ethylene perception is to antagonize the negative feedback regulation on EIN3 by promoting EBF1 and EBF2 mRNA decay, which consequently allows the accumulation of EIN3 protein to trigger the ethylene response.

Keywords: Arabidopsis, growth regulation, signal transduction

The plant hormone ethylene regulates a variety of developmental and stress responses in plants, including seed germination, cell elongation, cell fate, sex determination, fruit ripening, flower senescence, leaf abscission, defense against pathogens, and responses to mechanical trauma (1, 2). The most widely documented ethylene response is the so-called “triple response” of dark-grown (etiolated) seedlings. In the reference plant Arabidopsis thaliana the triple response is characterized by inhibition of hypocotyl and root elongation, a thickened hypocotyl, and an exaggerated apical hook (3). The triple-response phenotype has been used extensively to isolate components of the ethylene signal transduction pathway (4). Several mutants that display an aberrant triple response have been isolated in Arabidopsis. One class of mutants (i.e., etr1, etr2, ein2, ein3, ein4, ein5, ein6, and ein7) shows a reduction or complete loss of ethylene responsiveness in the presence of exogenous ethylene or its biochemical precursor, 1-aminocyclopropane-1-carboxylic acid (ACC). A second class of mutants (eto1, eto2, eto3, and ctr1) constitutively exhibits a triple response, either because of the overproduction of ethylene (Eto− mutants) or as a consequence of constant activation of the ethylene signaling pathway (Ctr−). The third class of mutants, which includes ebf1, ebf2, eer1, and rte1, exhibits an enhanced sensitivity to exogenous ethylene, suggesting they are negative regulators of the ethylene signaling pathway (5–10). A genetic and molecular analysis of these mutants has unraveled a largely linear ethylene signal transduction pathway leading from initial hormone perception to transcriptional regulation (4, 11). Briefly, ethylene gas is perceived by a family of five endoplasmic reticulum-localized, membrane-bound receptors (ETR1, ETR2, ERS1, ERS2, and EIN4) that have been demonstrated to be negative regulators of ethylene signaling (12, 13). To be functional, the ethylene receptors require a copper cofactor whose loading likely depends on RAN1, a homolog of the human Menkes Wilson P-type ATPase copper transporters (14, 15). Recent data suggest that RTE1, an evolutionarily conserved protein with unknown biochemical activity, plays an important role in the function of the ethylene receptor ETR1 (6). Interestingly, the tomato homolog of RTE1, GR, has been demonstrated to be required for ethylene responsiveness in fruit tissues (16). The first signaling component downstream of the ethylene receptors is CTR1, which is also a negative regulator of ethylene signaling (17). EIN2, an integral membrane protein of unknown function, is genetically downstream of CTR1 and is a major positive regulator of the pathway because loss-of-function mutations result in complete ethylene insensitivity (18). Functioning downstream of EIN2 is a small family of DNA-binding transcription factors that includes EIN3 and various EIN3-like proteins (5, 19). Loss-of-function mutations in EIN3 cause partial ethylene insensitivity, whereas loss-of-function mutations in EIL1 or EIL2 result in weak ethylene insensitivity (19, 20). Moreover, ectopic expression of EIN3 or EIL1 is sufficient to induce a complete ethylene response phenotype, suggesting that these proteins act as master regulators of the ethylene response (19, 21). Interestingly, recent studies have suggested that EIN3 protein levels rapidly increase in response to ethylene and that this response requires several ethylene signaling pathway components, including the ethylene receptors (ETR1 and EIN4), CTR1, EIN2, EIN5, and EIN6 (5). Furthermore, in the absence of or upon depletion of ethylene gas, EIN3 is quickly degraded through a ubiquitin/proteasome pathway mediated by two F-box proteins, EBF1 and EBF2 (5, 9, 10, 22).

In this study we describe the identification of ETHYLENE-INSENSITIVE5 as the previously described 5′→3′ exoribonuclease XRN4. We show that the ein5 mutation is able to partially suppress the constitutive ethylene response phenotype of ctr1 in both seedling and adult plants, thereby suggesting that EIN5 is a component of the ethylene signal transduction cascade acting downstream of CTR1. The involvement of EIN5 in the ethylene signaling cascade is further suggested by the finding that ein5 mutant plants are defective for ethylene-mediated gene expression. In addition, the mRNAs for the EIN3-regulating F-box proteins EBF1 and EBF2 accumulate in ein5 mutant plants in the presence and absence of ethylene gas, likely resulting in the under-accumulation of EIN3 and ultimately ethylene insensitivity. Mutation of either EBF1 or EBF2 in ein5 mutant plants suppresses the ethylene insensitivity phenotypes of these plants. Taken together, these results suggest that the role of EIN5 in ethylene perception is to antagonize the negative feedback regulation on EIN3 by promoting EBF1 and EBF2 mRNA decay, which consequently allows the accumulation of EIN3 protein to trigger the ethylene response.

Results

ein5 Mutants Specifically Affect the Ethylene Response.

The EIN5 locus is one of five novel ethylene-insensitive (Ein−) complementation groups identified as a result of a large-scale screen for ethylene response mutants in Arabidopsis (23). Mutations at the EIN5 locus confer insensitivity to high levels of exogenous and endogenous ethylene, which is characterized by the elongated hypocotyl and root compared with those of wild-type seedlings (Fig. 1A). Mutation of the CTR1 gene confers a constitutive triple-response phenotype on plants (17). CTR1 is a Raf-like protein kinase that negatively regulates downstream ethylene signaling events upon response to the ethylene receptors (17). Thus, the absence of CTR1 results in constitutive triggering of the ethylene response pathway. We wished to determine whether EIN5 function is epistatic to the CTR1 kinase. To do this, we made crosses between ein5-1 and ctr1-1 mutant plants. We found that in ein5-1 ctr1-1 double-mutant plants ein5 was able to partially suppress the constitutive ethylene phenotype of ctr1 in both seedling and adult stages (Fig. 1 B and C–F, respectively), thus demonstrating that EIN5 affects a step in the ethylene signaling pathway downstream of CTR1.

Fig. 1.

Fig. 1.

Open in a new tab

Ethylene insensitivity of ein5 and partial suppression of the constitutive phenotype of ctr1. (A) Comparison of 3-day-old etiolated wild-type, ein5, and ein2 seedlings grown in 10 ppm ethylene in hydrocarbon-free air. Wild type displays the ethylene-mediated triple-response phenotype, which is absent in ein2 and significantly reduced in ein5. (B–F) ein5 partially suppresses the constitutive ethylene phenotype of ctr1 in both seedling and adult stage. (B) Comparison of 3-day-old etiolated ctr1, ein5/ctr1, and wild-type seedlings grown in hydrocarbon-free air. ctr1-1 displays a constitutive triple response in the absence of exogenous ethylene. (C–F) Three- to 4-week-old adult wild-type (wt), ein5, ctr1, and ein5/ctr1 plants.

EIN5 Identified as the 5′→3′ Exoribonuclease XRN4.

We cloned the EIN5 gene using a map-based approach (Fig. 2A). ein5-1 and ein5-7 as well as new alleles were mapped to the lower half of chromosome 1, near marker nga128 (Fig. 2A). Using the 128-bp deletion-containing allele ein5-11 as a tool, we were able to precisely identify EIN5 as coding for the previously described 5′→3′ exoribonuclease XRN4 (Fig. 2B). Confirmation of EIN5's identity as XRN4 was provided by examination of multiple ein5 allele sequences. Seven of the 10 sequenced ein5 alleles caused frame-shift mutations that predicted a premature termination of the protein (Fig. 2C and Table 1). In addition, the mutation in ein5-4 caused a complex rearrangement of the gene, whereas ain1-2 and ain1-7 have G-to-A transitions that affect, respectively, the donor and acceptor sites of exon 3 (Fig. 2A). It has previously been reported that XRN4/EIN5 shares homology with two 5′→3′ exoribonucleases from Saccharomyces cerevisiae, Xrn1p and Rat1p (24). In addition, many other eukaryotic organisms also have homologs of XRN4/EIN5 (Fig. 2D; see also Fig. 7, which is published as supporting information on the PNAS web site). Interestingly, mutation of S. cerevisiae XRN1 results in pleiotropic defects including meiotic arrest, reduced spore viability, and slow growth (25). Correspondingly, we were interested to learn whether XRN4/EIN5 could complement the pleiotropic growth defects associated with an xrn1 mutant yeast strain. To do this, we introduced EIN5 into a S. cerevisiae xrn1 mutant strain using a yeast expression plasmid. We observed that a yeast expression plasmid carrying either EIN5 or the S. cerevisiae Xrn1 gene could complement an xrn1 mutant (Fig. 8B, which is published as supporting information on the PNAS web site). Similarly, as previously described, we observed that that introduction of XRN2 rescues the rat-1ts mutation but EIN5 does not (26) (Fig. 8A).

Fig. 2.

Fig. 2.

Open in a new tab

Map-based cloning of ETHYLENE-INSENSITIVE5. (A) Partial genetic map of Arabidopsis chromosome 1. EIN5 was mapped to SSLP marker nga128 and to restriction fragment-length polymorphisms identified within BACs T13C11, F3K22, and F20D21. The position of markers is indicated above the chromosome line, and under each one is displayed the number of recombinants between the marker and ein5 per number of chromosomes examined. A physical map of the region was constructed by using one yeast artificial chromosome, clone CIC11E10, and nine BAC clones. The genetic distance to EIN5 is shown below the chromosome line, and a broken arrow points to the location in the BAC containing the marker. At the bottom, the intron–exon structure of EIN5 and the location of mutations in ain1-2 and ain1-7 alleles are schematically shown. Both of these alleles have G→A transitions that affect the donor and acceptor sites of exon 3. (B) Identification of polymorphisms between the genomes of ein5-11 and wild-type plants. The allele ein5-11, which contains a 128-bp deletion, was used to locate EIN5. BAC F20D21 was used to identify polymorphisms between the genomes of ein5-11 and wild type. Southern blot analysis of restriction enzyme digests of Col-0 and ein5-11 DNA was performed with 32P-labeled DNA from BAC F20D21. The arrows point to the observed polymorphisms. (C) Schematic representation of mutations in ein5 alleles. The nucleotide positions for each mutation are as follows: ein5-1, 1-bp deletion at position 4292; ein5-2, 4-bp deletion at 3393; ein5-3, 5-bp deletion at 2417; ein5-5, 23-bp deletion at 4217; ein5-6, 22-bp deletion at 1724; ein5-7, 1-bp deletion at 1749; ein5-11, 128-bp deletion at 163. Position is given with reference to the ATG marking the start of the ORF in the unspliced sequence. (D) Phylogenetic analysis of XRN 5′→3′ exonucleases. The tree includes representative XRN genes from Drosophila (Dm, Drosophila melanogaster), humans (Hs, Homo sapiens), mice (Mm, Mus musculus), worms (Ce, Caenorhabditis elegans), rice (Os, Oriza sativa), yeast (Sc, S. cerevisiae), and Arabidopsis (At, A. thaliana). The phylogenetic tree was generated by using the neighbor-joining method with a bootstrap value of 1,000 replicates and analyzed by using NJplot.

Table 1.

Molecular nature of EIN5 alleles

Allele Mutagen Position of mutation† Ecotype Source
ein5-1 X-ray A 1-bp deletion at 4292; exon 15 Col-0 23
ein5-2 X-ray A 4-bp deletion at 3393; exon 12 Col-0 23
ein5-3 Fast neutron A 5-bp deletion at 2417–2421; exon 8 Col-0 This work
ein5-4 Fast neutron Rearrangement Col-0 This work
ein5-5 T-DNA A 23-bp deletion at 4217; exon 15 Col-0 J. M. Alonso, R. Solano, and J.R.E., unpublished observations
ein5-6 T-DNA A 22-bp deletion plus 5 bp at 1724–1745; exon 5 Col-0 J. M. Alonso, R. Solano, and J.R.E., unpublished observations
ein5-7 (ein7) X-ray A 1-bp deletion at 1749; exon 5 Col-0 23
ein5-11 T-DNA A 128-bp deletion at 163–291 Col-6 J. M. Alonso, R. Solano, and J.R.E., unpublished observations
ain1-2‡ EMS G→A at the donor site of the third exon/intron junction at 1074 C24 Gift from D. Van der Straeten (University of Ghent, Ghent, Belgium)
ain1-7‡ EMS G→A at the acceptor site of the third exon/intron junction at 1149 Col Gift from A. Bleecker (University of Winsconsin, Madison, WI)
Open in a new tab

†Relative to the start ATG of the unspliced gene.

‡ain, ACC-insensitive mutants.

Altered Ethylene-Mediated Gene Expression in ein5 Mutants.

The observed Ein− phenotype of ein5 mutant plants suggests that XRN4/EIN5 plays an important role in the ethylene signal transduction pathway. Therefore, mutation of XRN4/EIN5 might be expected to result in altered ethylene-mediated gene expression. To examine whether XRN4/EIN5 is required for proper ethylene-mediated gene expression, we performed Northern blot analysis using total RNA from 3-day-old etiolated seedlings of wild-type Col-0, ein5-1, ein5-6, ein5-11, and ein3-1 grown in air or 10 ppm ethylene. We found that wild-type Col-0 seedlings treated with ethylene (Fig. 3A, Col-0+) exhibited a dramatic increase in expression of the ethylene-responsive genes AtGST2, AtEBP (coding for the ethylene-responsive protein element), AtACO2, basic chitinase (BCH1), and defensin (PDF1.2) compared with Col-0 seedlings grown in hydrocarbon-free air (Fig. 3A, compare Col-0+ to Col-0−). Conversely, this same set of genes showed a reduced ethylene induction of expression in ein5 mutant seedlings (Fig. 3A) as compared with wild-type plants, although not as dramatic as the reduction observed in ein3 mutant seedlings (Fig. 3A). All three ein5 alleles tested (ein5-1, ein5-6, and ein5-11) exhibited a clearly reduced expression of defensin, basic-chitinase (BCHI), and GST (ATGST2) as compared with wild-type plants (Fig. 3A). Taken together, these findings provide further evidence for the involvement of EIN5 in the ethylene signaling pathway.

Fig. 3.

Fig. 3.

Open in a new tab

Defects in ethylene-mediated gene expression of ein5 mutant plants. (A) RNA gel blot analysis of total RNA from 3-day-old etiolated seedlings of wild-type Col-0, ein5-1, ein5-6, ein5-11, and ein3-1 grown in air or 10 ppm ethylene (C2H4). (Upper) Ten micrograms of total RNA from 3-day-old etiolated seedlings of wild-type Col-0, ein5-1, and ein3-1 grown in air (−) or 10 ppm ethylene (+) were sequentially hybridized with probes to GST (AtGST2 and At4g02520), ethylene-responsive binding protein (AtEBP and At3g16770), and ACC oxidase (AtACO2 and At1g62380). Finally, the blot was rehybridized with probes to ubiquitin and Tau3 as a loading control. (Lower) Twenty micrograms of total RNA from 3-day-old etiolated seedlings of wild-type Col-0, ein5-1, ein5-6, ein5-11, and ein3-1 grown in air (−) or 10 ppm ethylene (+) were sequentially probed with probes to basic chitinase (BCH1 and At3g12500), defensin (PDF1.2 and At5g44420), and AtGST2. Finally, the blot was rehybridized with a probe to 25S rRNA, and this blot is shown as a loading control. (B) Hierarchical cluster analysis of ATH1 microarray data obtained from etiolated seedlings exposed to ethylene (E) or hydrocarbon-free air (A). (Upper) Clustering of 50 genes up-regulated >3.0 times. (Lower) Clustering of 50 genes down-regulated >3.0 times. (Genes are listed in Tables 3 and 4.) Each cell corresponds to the level of expression of each gene in a given condition. Red indicates an increased level of expression, and blue indicates a decreased level of expression. Expression level is relative to the average of normalized expression of two GeneChips per condition. Genes are clustered together based on similarities of expression level. The bar is the scale for normalized expression.

To confirm and expand on these observations we used Affymetrix (Santa Clara, CA) Arabidopsis gene expression arrays, which interrogate the expression of >22,000 genes (27). We hybridized mRNA samples from 3-day-old etiolated seedlings of wild-type Col-0, ein5-1, and ein2-5 grown in air or treated for 4 h with 10 ppm ethylene gas to compare transcription profiles. We found that the expression of 244 genes was induced and 384 genes were repressed at least 2-fold after 4 h of ethylene exposure in wild-type etiolated seedlings (Table 2). Among these transcripts there are several previously known ethylene responsive genes such as ERF1 (21) and many genes involved in cell wall metabolism, confirming the success of the ethylene treatment for this experiment (data not shown). Fig. 3B shows graphically the 50 transcripts induced and repressed at the highest level (100 total) in wild-type Col-0 seedlings upon ethylene treatment and a comparison between expression levels of these same 100 genes in ein2-5 and ein5-1 seedlings upon hormone treatment (see also Tables 3 and 4, which are published as supporting information on the PNAS web site). We found that the loss of ethylene-mediated gene expression was more significant in ein2-5 mutant plants compared with ein5-1 mutant plants (Fig. 3B). In ein2-5 mutant plants there was almost no induction or repression of the 100 genes that are ethylene-responsive in wild-type Col-0 plants. Interestingly, in ein5-1 mutant plants these genes exhibited an intermediate level of expression and repression between wild-type Col-0 and ein2-5, as expected because of the observed intermediate ethylene insensitivity of ein5 mutant plants. Taken together, these results suggest that, rather than affecting a subset of the genes that respond to ethylene, ein5 affects the expression levels of most ethylene-responsive genes.

Table 2.

Genes exhibiting an altered expression in ein5-1 and ein2-5 with respect to Col-0

Allele No. of genes increased* No. of genes decreased* % altered with respect to Col-0 (air)†
ein5-1 127 123 1.56
ein2-5 193 119 1.95
Open in a new tab

*At least 2-fold difference.

†Sixteen thousand genes expressed.

EIN5 Regulates the EIN3-Targeting F-Box Proteins EBF1/2 mRNAs.

Next, we wanted to identify potential target RNAs regulated by XRN4/EIN5 that could possibly elucidate its function in the ethylene signaling pathway. To do this, we used a set of custom Affymetrix genome tiling arrays that span the entire Arabidopsis genome at high resolution (28). We prepared RNA samples from 10-day-old light-grown seedlings of wild type (Col-0) and ein5-1 to compare transcript profiles. From the expression profiling we found that EBF1 and EBF2 were highly overexpressed in ein5-1 mutant plants compared with wild-type Col-0 plants (Fig. 4A and B). This result was confirmed by examination of the data previously obtained using the gene expression arrays. In addition, using whole-genome tiling arrays allowed us to examine the expression of many other genes not present on the gene expression arrays (see also Fig. 9, which is published as supporting information on the PNAS web site, and data not shown).

Fig. 4.

Fig. 4.

Open in a new tab

EIN5 regulates the level of mRNA for EIN3-targeting F-box proteins EBF1 and EBF2. (A) Levels of EBF1 mRNA were determined by Affymetrix Arabidopsis tiling array expression analysis. The top track displays the level of EBF1 mRNA from ein5-1 mutant plants. The middle track displays the level of EBF1 mRNA from wild-type (Col-0) plants. The bottom track is the annotated gene model for the EBF1 locus. (B) Levels of EBF2 mRNA determined by Affymetrix Arabidopsis tiling array expression analysis. The top track displays the level of EBF2 mRNA from ein5-1 mutant plants. The middle track displays the level of EBF2 mRNA from wild-type (Col-0) plants. The bottom track is the annotated gene model for the EBF2 locus. (C) RNA gel blot analysis of 10 μg of total RNA from 3-day-old etiolated seedlings of wild-type Col-0 and ein5-6 grown in air or treated for various times with 10 ppm ethylene (C2H4) gas with a 3′ EBF1 probe. The times of ethylene treatment (or lack thereof) are indicated above the lanes. After stripping, the blot was rehybridized with a 25S rRNA probe, and this blot is shown as a loading control. Normalized values of EBF1 mRNA to the rRNA control (with Col-0 without ethylene treatment set at 1.0) are indicated at the bottom of the rRNA blot. (D) RNA gel blot analysis of 10 μg of total RNA from 3-day-old etiolated seedlings of wild-type Col-0 and ein5-6 grown in air or treated for varying times with 10 ppm ethylene (C2H4) gas with a 3′ EBF2 probe. The times of ethylene treatment (or lack thereof) are indicated above the lanes. After stripping, the blot was rehybridized with a 25S rRNA probe, and this blot is shown as a loading control. Normalized values of EBF2 mRNA to the rRNA control (with Col-0 without ethylene treatment set at 1.0) are indicated at the bottom of the rRNA blot. (E) RNA gel blot analysis of 10 μg of total RNA from rosette leaves of 3-week-old wild-type Col-0, ein5-1, ein5-6, and ein5-7 plants with a 3′ EBF1 probe. The rRNA is shown as a loading control. (F) RNA gel blot analysis of 10 μg of total RNA from rosette leaves of 3-week-old wild-type Col-0, ein5-1, ein5-6, and ein5-7 plants with a 3′ EBF2 probe. The rRNA is shown as a loading control.

To validate these results and to further characterize the regulation of EBF1 and EBF2 mRNA levels by EIN5, we performed Northern blot analysis using total RNA from 3-day-old etiolated seedlings of wild-type Col-0 and ein5-6 grown in air or 10 ppm ethylene for various amounts of time. We found that ein5-6 mutant seedlings accumulated significantly more EBF1 mRNA than did wild-type Col-0 seedlings both before and during ethylene treatment (Fig. 4C). Furthermore, the levels of EBF1 mRNA in the ein5-6 mutant seedlings accumulated to even higher levels compared with wild-type Col-0 plants after 4 and 12 h of growth in ethylene (Fig. 4C). Similar to the previous results, we also found that ein5-6 mutant seedlings accumulated significantly more EBF2 mRNA than did wild-type Col-0 seedlings both before and during ethylene treatment (Fig. 4D). Interestingly, we observed that, upon treatment with ethylene, wild-type Col-0 plants accumulated increased levels of EBF2 mRNA. However, the increase of EBF2 mRNA was still significantly higher in the ein5-6 mutant seedlings compared with wild-type Col-0 seedlings after treatment with ethylene (Fig. 4D).

Next we wanted to examine whether EBF1 and EBF2 mRNA levels were also significantly higher in light-grown adult ein5 mutant plants compared with wild-type Col-0 plants. We performed Northern blot analysis using total RNA from rosette leaves of 3-week-old wild-type Col-0, ein5-1, ein5-6, and ein5-7 plants. As was the case in 3-day-old etiolated seedlings, the ein5 mutant plants accumulated significantly higher levels of EBF1 and EBF2 mRNAs than did wild-type Col-0 plants (Fig. 4 E and F). Taken together, these results demonstrate that the XRN4/EIN5 exoribonuclease regulates EBF1 and EBF2 at the posttranscriptional level and suggest that regulation of these EIN3-targeting F-box proteins is the role of XRN4/EIN5 in the ethylene signal transduction pathway.

Mutation of EBF1 and EBF2 Can Suppress the Ein− Phenotype of ein5 Mutant Plants.

We wanted to determine whether regulation of EBF1 and EBF2 mRNA levels was relevant to the role of EIN5 in ethylene signal transduction. We generated double mutants between ein5 and ebf1 or ebf2. As shown in Fig. 5A and B, ebf2-1 completely suppressed the Ein− phenotype of ein5-1, whereas ebf1-1 partially suppressed the Ein− phenotype of ein5-1. Additionally, three independent attempts were made to obtain ein5-1 ebf1-1 ebf2-1 triple-mutant plants. We were never able to obtain a homozygous triple mutant plant, but we found that ein5-1 ebf1-1 ebf2-1+/− and ein5-1 ebf1-1+/− ebf2-1 were highly hypersensitive to ethylene (data not shown). Therefore, the inability to obtain triple-mutant plants was likely caused by embryo lethality of this combination of mutations. Taken together, these results suggest that the Ein− phenotype of ein5 mutant plants is caused by the over-accumulation of EBF1 and EBF2 mRNAs, likely resulting in a decrease in EIN3 levels.

Fig. 5.

Fig. 5.

Open in a new tab

ein5 defect is suppressed by ebf1 and ebf2 mutations. (A) Triple-response phenotype of 3-day-old etiolated seedlings grown on MS medium supplemented with (+ACC) or without (−ACC) 10 μM ACC. (B) Dosage response of wild-type Col-0 and various mutants. Etiolated seedlings were grown on MS medium supplemented with different concentrations of ACC for 3 days. The length of hypocotyls from 10 seedlings was measured, and the mean values and standard deviations were plotted. (C) Immunoblot assays of EIN3 protein in ein5 and ein5 ebf double mutants. Etiolated seedlings grown in air for 3 days were treated with or without ethylene (C2H4) gas (10 ppm) for 4 h before tissues were harvested for immunoblot assays. CRY1 (Arabidopsis cryptochrome 1) protein was used as a loading control.

To test directly whether EIN3 levels are depleted in ein5 mutant plants compared with wild-type Col-0, Western blot analysis with an EIN3-specific antibody was done by using total protein from 3-day-old etiolated seedlings of wild-type Col-0 ein5-1, ein5-1 ebf1-1, and ein5-1 ebf2-1 grown in air or 10 ppm ethylene. As expected, upon ethylene treatment EIN3 accumulated in wild-type Col-0 plants, and this ethylene-induced accumulation of EIN3 was ablated in ein5-1 mutant plants (Fig. 5C) (5). Interestingly, ethylene-induced accumulation of EIN3 was completely restored in ein5-1 ebf2-1 double-mutant plants but not in ein5-1 ebf1-1 double mutant plants, providing a likely explanation for the difference in ability of the ebf1 and ebf2 to suppress the Ein− phenotype of ein5-1 mutant plants (Fig. 5). Taken together, these results suggest that the over-accumulation of EBF1 and EBF2 mRNAs in ein5 mutant plants results in an inability to accumulate EIN3 in response to ethylene, thereby causing ethylene insensitivity.

Discussion

Here we show that EIN5 is a component of the ethylene signaling pathway that acts downstream of CTR1, which is necessary for proper ethylene-mediated gene expression. Interestingly, microarray analysis carried out on wild-type Col-0, ein2, and ein5 etiolated seedlings (exposed to air or to ethylene) (Fig. 3) revealed that a set of genes that are highly induced by ethylene treatment in the wild-type plant exhibits an intermediate expression level in ein5 as compared with ein2, where there is only marginal induction of these genes. It is interesting to notice that the profile of expression coincides with the phenotypic appearance of the plants with the wild type exhibiting the full triple response, and ein2 being the extreme, with none of the triple-response features evident (Fig. 1). In ein5 seedlings there is an intermediate Ein− phenotypic response to ethylene gas (Fig. 1). Upon cloning EIN5 we were surprised to find that it codes for the previously described 5′→3′ exoribonuclease XRN4. XRN4/EIN5 has previously been shown to degrade mRNA in a 5′→3′ direction (18, 29). Therefore, this enzyme is likely involved in the mRNA decay pathway in which deadenylation is followed by the hydrolysis of the 5′ cap structure and processive degradation of the mRNA body (30). Specific to the ethylene signaling pathway, we find that XRN4/EIN5 targets for decay the mRNAs of the EIN3-regulating F-box proteins EBF1 and EBF2 through a yet unknown mechanism (Fig. 6). We propose that the function of XRN4/EIN5 in the ethylene signal transduction pathway is likely to antagonize the negative feedback regulation on EIN3 by promoting EBF1 and EBF2 mRNA decay, which results in the accumulation of EIN3 protein, triggering the ethylene response (Fig. 6). Analysis of the ethylene-insensitive mutant ein5 has led us to uncover a previously unknown aspect of the regulation of the ethylene response, namely the participation of an enzyme with known functions in mRNA turnover and RNA interference, XRN4/EIN5 (24, 29, 31).

Fig. 6.

Fig. 6.

Open in a new tab

A model of the role of EIN5 in the ethylene signal transduction pathway. Ethylene (C2H4) is perceived by repressing the action of receptor complexes including ETR/ERS/EIN4 receptors, RTE1, and Raf-like protein kinase CTR1, which negatively regulates downstream signaling component EIN2. Upon ethylene treatment, EIN2 is derepressed and could thus transmit the signal into the nucleus to activate a number of transcription factors, including EIN3 and EIL1. EIN3 directly binds to the regulatory elements of target genes and induces the expression of yet other transcription factors (i.e., ERFs and EDFs) that would ultimately regulate a series of ethylene responses. In the absence of ethylene signal, a Skp1-Cullin1-F-box complex consisting of one of two F-box proteins, EBF1 and EBF2, targets EIN3 protein for degradation via an ubiquitin/proteasome pathway. Interestingly, EBF1/EBF2 gene expression is induced by ethylene in an EIN3-dependent manner, which forms a negative feedback regulation on the EIN3 function. EIN5, a 5′→3′ exoribonuclease, is involved in facilitating the turnover of EBF1/EBF2 mRNA through a yet unknown mechanism. Therefore, EIN5 is proposed to antagonize the negative feedback regulation on EIN3 by promoting EBF1 and EBF2 mRNA decay, which consequently allows the accumulation of EIN3 protein to trigger the ethylene response. Red arrows and blue bars represent positive and negative regulations, respectively. The dotted lines represent regulatory steps in which a direct physical link between upstream and downstream components has yet to be demonstrated.

Although XRN1 is required for many processes in yeast, it was observed that in yeast strains lacking this protein (xrn1Δ) only a small fraction of transcripts are over-accumulated, thus suggesting that 3′→5′ decay, not 5′→3′ decay, may be the major mRNA decay activity in yeast cells (32). In concordance, our analysis carried out with gene expression arrays suggests that 3′→5′ decay, not 5′→3′ decay, may be the major mRNA decay activity in plants. A comparison of the expression data from air-treated seedlings of wild-type Col-0 and ein5 mutant plants did not reveal a global effect on the abundance of transcripts expressed in plants (data not shown). Additionally, whole-genome tiling array expression profiling also identified a specific class of EIN5 substrates (Fig. 9 and data not shown), including EBF1 and EBF2 mRNAs (Fig. 5 A and B). In concordance with these microarray results, a previous analysis of ein5 mutant plants versus wild-type Col-0 plants carried out on 8,000 Arabidopsis genes identified only 14 genes that seemed to accumulate in ein5 mutant plants (29). Furthermore, ein5 mutant plants do not manifest gross developmental defects, but have only slightly serrated leaf edges (Fig. 1D). Taken together, these results suggest that XRN4/EIN5 is likely an exonuclease required for the decay of specific target mRNAs, whereas a 3′→5′ activity suffices for general mRNA decay in Arabidopsis.

XRN4/EIN5 is a 5′→3′ exoribonuclease with homologs in many eukaryotic organisms (Figs. 3D and 7). Interestingly, the closest homolog of XRN4/EIN5 in Caenorhabditis elegans, xrn-1, is required for proper embryonic development. Loss of this key protein results in embryos that fail to complete ventral enclosure, where the outer layer of cells normally closes over the mesoderm in a purse-string movement (33). Furthermore, S. cerevisiae strains containing mutation in XRN1 manifest pleiotropic defects including meiotic arrest, reduced spore viability, and slow growth (25). Recently, XRN1 has also been demonstrated to play a role in yeast iron homeostasis (34). In plants, XRN4/EIN5 has already been demonstrated to degrade selected mRNAs including the 3′ products of the endonucleolytic cleavage directed by selected microRNAs (29). Here we further demonstrate that the XRN4/EIN5 exonuclease is a required component of the ethylene signal transduction pathway. Taken together, these results suggest that within plants and other eukaryotic organisms XRN4/EIN5 and its homologs are likely active in many other pathways in which target mRNAs are required for specific function.

Materials and Methods

Plant Lines and Growth Conditions.

The Columbia ecotype of Arabidopsis was used in this study. Seeds were surface-sterilized as described previously (35), resuspended in a suitable volume of top agar (0.3% low-melt agarose), and spread onto Murashige and Skoog agar [Murashige and Skoog salts (Invitrogen, Carlsbad, CA) and 0.8% agar (pH 5.7)]. Seeds were cold-treated for 4 days (4°C) and then were covered with aluminum foil and moved to an incubator at 24°C. Adult plants were grown in potting soil (Metro Mix 250; Grace-Sierra, Boca Raton, FL) under continuous illumination at 23°C under a 16-h light/8-h dark cycle. The ein5-2 and ein5-7 (formerly ein7) mutants were previously identified (23). The ein5-1, ein5-2, and ein5-7 alleles were isolated from an x-ray-mutagenized population (ecotype Columbia) (17). ein5-3 and ein5-4 were screened out from a fast neutron mutagenized population. ein5-5, ein5-6, and ein5-11 were recovered from the T-DNA collection of The Salk Institute for Biological Studies (36). ain1-2 and ain1-7 come from an ethyl methanesulfonate (EMS)-mutagenized population. Only ain1-2 was derived from a C24 ecotype. Ethylene treatment of Arabidopsis seedlings grown on Petri plates was performed in containers by flowing through hydrocarbon-free air supplemented with 10 ppm ethylene or were treated with hydrocarbon-free air alone (17).

Genetic Analysis of Double Mutants.

Double mutants were generated by genetic crosses, and homozygous lines were identified by PCR-based genotyping. Arabidopsis seeds were surface-sterilized and plated on the surface of MS medium supplemented with different concentrations of ACC. After 4 days at 4°C the plates were wrapped in foil and kept in a 24°C incubator for 3 days before the phenotypes of seedlings were analyzed.

Northern Blot Analysis.

Total RNA was isolated by using the RNeasy Plant Mini Kit (Qiagen, Valencia, CA), fractionated by electrophoresis on 1.2% agarose/6% formamide gels, and transferred to Hybond nylon membranes (Amersham Biosciences, Piscataway, NJ). Each DNA probe was 32P-labeled by the multiprime labeling system (Amersham Biosciences), and hybridization, stripping, and reprobing were carried out as described (19).

Microarray Experiments.

Cold-treated seeds in MS plates were placed in chambers at 24°C in the dark with hydrocarbon-free airflow for 3 days, after which some of the chambers were connected to ethylene gas at 10 ppm while the others remained on air treatment. Four hours later the seedlings of each plate were quickly collected and frozen in liquid nitrogen. Total RNA was prepared from both air-treated and ethylene-treated etiolated seedlings of Arabidopsis Col-0, ein2-5, and ein5-1 by using the RNeasy Plant Mini Kit. Biotinylated target RNA was prepared from 16 μg of total RNA by using the procedure described by the manufacturer (Affymetrix, Santa Clara, CA). Briefly, a primer encoding a T7 RNA polymerase promoter fused to (dT)24 (Genset Oligos, La Jolla, CA) was used to prime double-stranded cDNA synthesis using the SuperScript Choice System (Invitrogen). The resulting cDNA was transcribed in vitro by using the BioArray HighYield RNA Transcript Labeling Kit (Enzo Biochem, New York, NY) in the presence of biotinylated UTP and CTP to produce biotinylated target cRNA. The labeled target cRNA was purified, fragmented, and hybridized to Arabidopsis microarrays (Affymetrix ATH1 gene expression arrays) according to protocols provided by the manufacturer in hybridization oven model 640 (Affymetrix). The arrays were washed and stained with streptavidin–phycoerythrin by using a GeneChip Fluidics Station model 400 and then scanned with a Gene Array Scanner (Hewlett-Packard, Palo Alto, CA). Scanned images were processed and quantified by using GeneChip Suite 3.2. Genespring software (Silicon Genetics, Redwood City, CA) was used to manage and filter the array data. Each measurement was divided by the 50.0th percentile of all measurements in that sample. The percentile was calculated with all normalized measurements above 10. For samples where the bottom 10th percentile was less than the negative of the 50.0th percentile, it was used as a background and subtracted from all of the other genes first. Experiments were carried out in duplicate and hybridized to two sets of expression chips. Expression values were analyzed by using GeneSpring software version 4.2.

For the oligonucleotide tiling array experiments 10-day-old light-grown Arabidopsis Col-0 and ein5-1 seedlings were transferred to a Petri plate containing incubation buffer (1 mM Pipes, 1 mM sodium citrate, 1 mM KCl, and 15 mM sucrose) and maintained in agitation at 75 rpm for 30 min covered with foil. At this point vacuum was applied for exactly 15 s and cordycepin (Sigma Chemical, St. Louis, MO) was added to a final concentration of 150 μg/ml. The seedlings were then incubated for two more hours in the dark at room temperature and kept in agitation at 75 rpm in Petri plates. The seedlings were then briefly blotted on filter paper to dry and kept in liquid nitrogen. Total RNA was extracted from the samples by using the RNeasy plant kit (Qiagen, Valencia, CA). Biotinylated target RNA was prepared from 120 μg of total RNA from each sample by using the GeneChip Expression Analysis system (Affymetrix). Each sample of the purified and fragmented labeled target cRNA was hybridized to one set of 12 custom-made genome tiling arrays (Affymetrix) that span the entire Arabidopsis genome at high resolution (28). Hybridization, washes, and staining were carried out as described above for the Affymetrix array ATH1. Cell files were obtained by using GCOS software (Affymetrix), and ChipViewer software was used to visualize and analyze the tiling chip data (28).

Immunoblot Assays.

Immunoblot assays were performed as described (5). In brief, 3-day-old dark-grown seedlings were treated with 10 ppm ethylene or hydrocarbon-free air for 4 h before protein samples were extracted for Western blot analysis. The protein extracts were fractionated by 4–12% gradient Tris-glycine Novex precast gels (Invitrogen) blotted onto a nitrocellulose filter. The blot was probed first with anti-EIN3 antibody and was subsequently stripped with 0.2 N glycine (pH 2.5) three times and reprobed with anti-CRY1 antibody (37).

Supplementary Material

Supporting Information
pnas_0605528103_index.html (4.6KB, html)

Acknowledgments

We thank Huaming Chen, Juan Caballero, and Enrique Ibarra for assistance with microarray analysis; Octavio Martínez-de la Vega and H. Chen for configuring the tiling array database; H. Chen and J. Yazaki for depositing gene expression and tiling array data in the Gene Expression Omnibus database; P. Shinn for depositing DNA sequences in the GenBank database; M. A. Urich for technical assistance; A. Johnson (University of Texas, Austin, TX) for providing yeast strains; and K. Emerson and J. Simon for assistance with preparation of digital artwork. G.O. and P.G. were supported by a CONACyT sabbatical grant from Mexico, and G.O. was supported by a Fulbright–García Robles Scholar Visitor fellowship. B.D.G. is a Damon Runyon Fellow supported by the Damon Runyon Cancer Research Foundation (DRG-1909-06). This work was supported by National Science Foundation Grant MCB-0516888 and Department of Energy Grant DE-FG02-04ER15517 (to J.R.E.).

Abbreviation

ACC

1-aminocyclopropane-1-carboxylic acid.

Footnotes

Conflict of interest statement: No conflicts declared.

Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession no. BT026022), and gene expression and tiling array data have been deposited in the Gene Expression Omnibus database, www.ncbi.nlm.nih.gov/geo (accession no. GSE5174).

References

  • 1.Johnson P. R., Ecker J. R. Annu. Rev. Genet. 1998;32:227–254. doi: 10.1146/annurev.genet.32.1.227. [DOI] [PubMed] [Google Scholar]
  • 2.Wang K. L., Li H, Ecker J. R. Plant Cell. 2002;14(Suppl.):S131–S151. doi: 10.1105/tpc.001768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Ecker J. R. Science. 1995;268:667–675. doi: 10.1126/science.7732375. [DOI] [PubMed] [Google Scholar]
  • 4.Chen Y. F., Etheridge N., Schaller G. E. Ann. Bot. (London) 2005;95:901–915. doi: 10.1093/aob/mci100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Guo H., Ecker J. R. Cell. 2003;115:667–677. doi: 10.1016/s0092-8674(03)00969-3. [DOI] [PubMed] [Google Scholar]
  • 6.Resnick J. S., Wen C. K., Shockey J. A., Chang C. Proc. Natl. Acad. Sci. USA. 2006;103:7917–7922. doi: 10.1073/pnas.0602239103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Larsen P. B., Chang C. Plant Physiol. 2001;125:1061–1073. doi: 10.1104/pp.125.2.1061. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Larsen P. B., Cancel J. D. Plant J. 2003;34:709–718. doi: 10.1046/j.1365-313x.2003.01762.x. [DOI] [PubMed] [Google Scholar]
  • 9.Potuschak T., Lechner E., Parmentier Y., Yanagisawa S., Grava S., Koncz C., Genschik P. Cell. 2003;115:679–689. doi: 10.1016/s0092-8674(03)00968-1. [DOI] [PubMed] [Google Scholar]
  • 10.Gagne J. M., Smalle J., Gingerich D. J., Walker J. M., Yoo S. D., Yanagisawa S., Vierstra R. D. Proc. Natl. Acad. Sci. USA. 2004;101:6803–6808. doi: 10.1073/pnas.0401698101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Nehring R., Ecker J. R. In: Plant Hormones: Biosynthesis, Signal Transduction, Action! Davies P. J., editor. Dordrecht: Kluwer; 2004. pp. 350–368. [Google Scholar]
  • 12.Chang C., Stadler R. Bioessays. 2001;23:619–627. doi: 10.1002/bies.1087. [DOI] [PubMed] [Google Scholar]
  • 13.Chang C., Bleecker A. B. Plant Physiol. 2004;136:2895–2899. doi: 10.1104/pp.104.900122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hirayama T., Kieber J. J., Hirayama N., Kogan M., Guzman P., Nourizadeh S., Alonso J. M., Dailey W. P., Dancis A., Ecker J. R. Cell. 1999;97:383–393. doi: 10.1016/s0092-8674(00)80747-3. [DOI] [PubMed] [Google Scholar]
  • 15.Woeste K. E., Kieber J. J. Plant Cell. 2000;12:443–455. doi: 10.1105/tpc.12.3.443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Barry C. S., Giovannoni J. J. Proc. Natl. Acad. Sci. USA. 2006;103:7923–7928. doi: 10.1073/pnas.0602319103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Kieber J. J., Rothenberg M., Roman G., Feldmann K. A., Ecker J. R. Cell. 1993;72:427–441. doi: 10.1016/0092-8674(93)90119-b. [DOI] [PubMed] [Google Scholar]
  • 18.Alonso J. M., Hirayama T., Roman G., Nourizadeh S., Ecker J. R. Science. 1999;284:2148–2152. doi: 10.1126/science.284.5423.2148. [DOI] [PubMed] [Google Scholar]
  • 19.Chao Q., Rothenberg M., Solano R., Roman G., Terzaghi W., Ecker J. R. Cell. 1997;89:1133–1144. doi: 10.1016/s0092-8674(00)80300-1. [DOI] [PubMed] [Google Scholar]
  • 20.Alonso J. M., Stepanova A. N., Solano R., Wisman E., Ferrari S., Ausubel F. M., Ecker J. R. Proc. Natl. Acad. Sci. USA. 2003;100:2992–2997. doi: 10.1073/pnas.0438070100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Solano R., Stepanova A., Chao Q., Ecker J. R. Genes Dev. 1998;12:3703–3714. doi: 10.1101/gad.12.23.3703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Yanagisawa S., Yoo S. D., Sheen J. Nature. 2003;425:521–525. doi: 10.1038/nature01984. [DOI] [PubMed] [Google Scholar]
  • 23.Roman G., Lubarsky B., Kieber J. J., Rothenberg M., Ecker J. R. Genetics. 1995;139:1393–1409. doi: 10.1093/genetics/139.3.1393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kastenmayer J. P., Green P. J. Proc. Natl. Acad. Sci. USA. 2000;97:13985–13990. doi: 10.1073/pnas.97.25.13985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Larimer F. W., Stevens A. Gene. 1990;95:85–90. doi: 10.1016/0378-1119(90)90417-p. [DOI] [PubMed] [Google Scholar]
  • 26.Kastenmayer J. P., Johnson M. A., Green P. J. Methods Enzymol. 2001;342:269–282. doi: 10.1016/s0076-6879(01)42551-1. [DOI] [PubMed] [Google Scholar]
  • 27.Redman J. C., Haas B. J., Tanimoto G., Town C. D. Plant J. 2004;38:545–561. doi: 10.1111/j.1365-313X.2004.02061.x. [DOI] [PubMed] [Google Scholar]
  • 28.Yamada K., Lim J., Dale J. M., Chen H., Shinn P., Palm C. J., Southwick A. M., Wu H. C., Kim C., Nguyen M., et al. Science. 2003;302:842–846. doi: 10.1126/science.1088305. [DOI] [PubMed] [Google Scholar]
  • 29.Souret F. F., Kastenmayer J. P., Green P. J. Mol. Cell. 2004;15:173–183. doi: 10.1016/j.molcel.2004.06.006. [DOI] [PubMed] [Google Scholar]
  • 30.Meyer S., Temme C., Wahle E. Crit. Rev. Biochem. Mol. Biol. 2004;39:197–216. doi: 10.1080/10409230490513991. [DOI] [PubMed] [Google Scholar]
  • 31.Gazzani S., Lawrenson T., Woodward C., Headon D., Sablowski R. Science. 2004;306:1046–1048. doi: 10.1126/science.1101092. [DOI] [PubMed] [Google Scholar]
  • 32.He F., Li X., Spatrick P., Casillo R., Dong S., Jacobson A. Mol. Cell. 2003;12:1439–1452. doi: 10.1016/s1097-2765(03)00446-5. [DOI] [PubMed] [Google Scholar]
  • 33.Newbury S., Woollard A. RNA. 2004;10:59–65. doi: 10.1261/rna.2195504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Lee A., Henras A. K., Chanfreau G. Mol. Cell. 2005;19:39–51. doi: 10.1016/j.molcel.2005.05.021. [DOI] [PubMed] [Google Scholar]
  • 35.Vogel J. P., Woeste K. E., Theologis A., Kieber J. J. Proc. Natl. Acad. Sci. USA. 1998;95:4766–4771. doi: 10.1073/pnas.95.8.4766. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Alonso J. M., Stepanova A. N., Leisse T. J., Kim C. J., Chen H., Shinn P., Stevenson D. K., Zimmerman J., Barajas P., Cheuk R., et al. Science. 2003;301:653–657. doi: 10.1126/science.1086391. [DOI] [PubMed] [Google Scholar]
  • 37.Lin C., Ahmad M., Gordon D., Cashmore A. R. Proc. Natl. Acad. Sci. USA. 1995;92:8423–8427. doi: 10.1073/pnas.92.18.8423. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
pnas_0605528103_index.html (4.6KB, html)
pnas_0605528103_05528Table3.xls (19.5KB, xls)
pnas_0605528103_05528Table4.xls (19KB, xls)
pnas_0605528103_1.pdf (207.7KB, pdf)
pnas_0605528103_2.pdf (285.6KB, pdf)
pnas_0605528103_05528Fig8.jpg (30.7KB, jpg)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES