The RAR1 Interactor SGT1, an Essential Component of R Gene-Triggered Disease Resistance

See allHide authors and affiliations

Science  15 Mar 2002:
Vol. 295, Issue 5562, pp. 2073-2076
DOI: 10.1126/science.1067554


Plant disease resistance (R) genes trigger innate immune responses upon pathogen attack. RAR1 is an early convergence point in a signaling pathway engaged by multiple R genes. Here, we show that RAR1 interacts with plant orthologs of the yeast protein SGT1, an essential regulator in the cell cycle. Silencing the barley geneSgt1 reveals its role in R gene–triggered,Rar1-dependent disease resistance. SGT1 associates with SKP1 and CUL1, subunits of the SCF (Skp1-Cullin–F-box) ubiquitin ligase complex. Furthermore, the RAR1-SGT1 complex also interacts with two COP9 signalosome components. The interactions among RAR1, SGT1, SCF, and signalosome subunits indicate a link between disease resistance and ubiquitination.

Plant disease resistance (R) genes are key components in pathogen perception; R genes activate a battery of defense reactions, collectively called the hypersensitive response (HR) (1). A number of R genes from various plant species have been isolated and characterized in detail (2). Although different R genes confer resistance to a variety of pests, including bacteria, viruses, fungi, nematodes, and insects, many R gene products share common structural modules such as a nucleotide-binding site and a leucine-rich repeat domain. The structural similarity among R proteins suggests the existence of common or related signal transduction mechanisms downstream of pathogen perception (3). Several genes required for R gene function have been identified, and some may represent such downstream signaling components. For example,EDS1 encodes a lipase-like protein, and NDR1encodes a putative membrane-associated protein (4,5). However, the biochemical function(s) of these proteins remains to be elucidated.

Barley Rar1 (HvRar1) was originally isolated as a gene required for resistance signaling triggered by multipleR genes, including Mla6 and Mla12, which confer resistance to the pathogenic powdery mildew (Blumeria graminis f. sp. hordei) (6). HvRar1 encodes a protein with two highly similar but distinct zinc-binding domains called CHORD-I and -II (cysteine- and histidine-rich domain). RAR1 homologs are present in plants, protozoa, metazoa, and fungi but are absent in yeast. The sequence and paired arrangement of CHORDs are highly conserved, a finding that suggests the preservation of a common biochemical function. Gene silencing of Rar1 homologs inCaenorhabditis elegans perturbed specific developmental processes (6). In humans, a RAR1 homolog (melusin) was isolated as a muscle-specific protein that interacts in vitro with the cytoplasmic domain of integrin (7). This finding suggests that the conserved RAR1s are key components in diverse biological processes in eukaryotes.

To investigate the molecular function(s) of RAR1 in plants, we searched for interacting proteins by using a yeast two-hybrid screen in whichArabidopsis RAR1 (AtRAR1) served as bait (8). We identified two interacting proteins sharing extensive amino acid similarity to each other (87%) and to yeast SGT1 (48 and 50%, respectively), which we designated AtSGT1a and AtSGT1b (Fig. 1, A and C). HvRAR1 also interacts with both AtSGT1a and AtSGT1b, indicating that the SGT1-binding function of plant RAR1 proteins has been conserved in monocots and dicots. We also cloned an Sgt1 homolog from barley (HvSgt1) by using reverse transcription–polymerase chain reaction (RT-PCR) with degenerate primers. Only a single sequence was identified; thus, barley probably contains only one copy of Sgt1. Both AtSGT1a and AtSGT1b are more similar to one another than to SGT1 proteins of monocots, suggesting that these two Arabidopsis homologs are the result of recent gene duplication. Alignments of amino acid sequences defined five domains in plant SGT1 proteins: a tetratricopeptide repeat domain (TPR) (9), two variable regions (VR1 and VR2), the CS motif (present in metazoan CHORD and SGT1 proteins), and the SGS motif (SGT1-specific motif) (Fig. 1B).

Figure 1

The deduced SGT1 protein sequence and physical interaction between SGT1 and RAR1. (A) Sequence alignment of SGT1 proteins. Sc, Saccharomyces cerevisiae; Hs, human; Hv, barley; Os, rice; At, Arabidopsis. The black (100%), dark gray (80%), and light gray (60%) boxes represent levels of amino acid identity or similarity. GenBank accession numbers: HvSGT1, AF439974; AtSGT1a, AF439975; AtSGT1b, AF439976 (22). (B) Graphical representation of the structures of HvRAR1, CeCHP (Ce,C. elegans), and HvSGT1 domains. The numbers on the right indicate the number of amino acids per protein. (C) Yeast two-hybrid analysis of the interaction between SGT1 and RAR1.HvRAR1, Hvrar1-1 (6),AtRAR1, AtSGT1a, AtSGT1b, and fragments ofAtRAR1 and AtSGT1a were amplified by PCR and cloned into the Eco RI and Xho I sites of the pLexA (binding domain fusion) or pB42AD (activator domain fusion) vectors, as indicated. Interactions were detected by induction (+) or no induction (–) of lacZ and LEU2 reporter genes under the control of the lexA gene in yeast EGY48 cells; nt, not tested. TPR, tetratricopeptide repeat domain; VR1 and VR2, variable regions; SGS, SGT1-specific motif.

Deletion analysis in the yeast two-hybrid system revealed that the CS motif of AtSGT1a interacts with CHORD-II but not with CHORD-I or Cys-Cys-Cys-His (CCCH) of AtRAR1 (Fig. 1C). A single amino acid substitution in CHORD-I (Hvrar1-1) of HvRAR1 did not abolish interaction with AtSGT1, corroborating the finding that CHORD-II is sufficient for the RAR1-SGT1 complex formation (Fig. 1C). Thus, although the sequences of CHORD-I and CHORD-II are highly related, these domains may serve distinct roles. Interestingly, the CS motif of SGT1 proteins is also found at the COOH terminus of the metazoan RAR1 homolog, CHP (6) (Fig. 1B). Therefore, the RAR1-SGT1 interaction represents an example of the Rosetta Stone principle (10), whereby the presence of two domains in a single polypeptide in one species (in this case, CHORD-II and CS in metazoan CHP) indicates that the two domains interact with one another when they are present in separate proteins in another species. These results suggest that RAR1 and SGT1 participate in a related biochemical mechanism(s) in eukaryotic organisms that contain both genes.

In yeast, Sgt1 (suppressor of the G2 allele ofskp1) is required for cell cycle progression at the G1/S-phase and G2/M-phase transitions (11). At nonpermissive temperatures, both AtSGT1aand AtSGT1b can complement G1 and G2arrest in temperature-sensitive sgt1 mutants (Fig. 2). No other SGT1 homologs have been found in theArabidopsis genome, indicating that these proteins are the functional orthologs of yeast Sgt1.

Figure 2

Arabidopsis SGT1a and SGT1bcomplement the sgt1 mutation in yeast. Yeast strains YKK57 (sgt1-5) and YKK65 (sgt1-3) were transformed with AtSGT1a, AtSGT1b (cloned into p426GAL), or vector p426GAL. The transformants were selected on Sc-Ura 2% glucose plates and spread on Sc-Ura 2% raffinose plates. The strains were spread on Sc-Ura 2% galactose plates and incubated for 4 days to test their ability to complement the temperature-sensitive sgt1-5 andsgt1-3 growth defects.

To determine whether HvSGT1 is essential in Rgene–triggered disease resistance, we used a single-cell transient double-stranded RNA inhibition (dsRNAi) system for silencing the corresponding endogenous genes (12). In this assay, a DNA construct encoding an intron-spliced dsRNA is delivered via gold-coated particles into epidermal cells of barley leaves. A plasmid-expressing red fluorescent protein (RFP) was delivered with the test constructs to mark the transformed cells. To monitor the silencing efficiency, we cobombarded a third plasmid-expressing green fluorescent protein (GFP) fused to the COOH terminus of either HvRAR1 or HvSGT1. Forty-eight hours after delivery, the green fluorescence of cells expressing HvRAR1-GFP or HvSGT1-GFP was markedly reduced, whereas the levels of red fluorescence remained unchanged (Fig. 3A). This finding indicates thatHvSgt1 and HvRar1 expression can be silenced in a sequence-dependent manner.

Figure 3

HvSgt1 is required for full resistance to powdery mildew triggered by Mla6. (A) dsRNAi was used to silence single-cell expression of HvRar1 andHvSgt1. Detached barley leaves were bombarded with the constructs, as described (23). After 48 hours, transformed epidermal cells were identified using a microscope with the GFP or RFP filter. Magnification, ×300 (B) Effect of silencing ofHvRAR1 and HvSGT1 on Mla6- andMla1-resistance pathways in single cells. After particle-mediated delivery of corresponding dsRNAi constructs and the GUS marker plasmid, leaves were incubated at 20°C for either 48 hours (white bar) or 96 hours (black bar) before inoculation with powdery mildew spores. After 48 hours, samples were stained for GUS activity and fungal structures (with Coomassie blue). The fraction of susceptible GUS-positive cells in bombardments lacking a correspondingR gene construct but containing an empty dsRNAi vector was defined as 100% susceptibility (the control fraction). The relative susceptibility was then calculated as the fraction of susceptible GUS-positive cells bombarded with a test dsRNAi construct and an R gene, divided by the control fraction, ×100. (mean ± SD of three experiments). The inset (magnification, ×300) shows a typical susceptible leaf epidermal cell expressing GUS in which powdery mildew successfully formed an intracellular haustorium and initiated growth of aerial hyphae.

Next, we tested the effect of silencing HvRar1 andHvSgt1 in Mla1- and Mla6-specific disease resistance by inoculating bombarded leaves with spores of powdery mildew strains expressing the corresponding avirulence genesAvrMla1 or AvrMla6 (Fig. 3B). In these experiments, we used β-glucuronidase (GUS) as a marker for transformed cells. Silencing of HvRar1 for 96 hours impaired the resistance triggered by Mla6, although only in 10% of the attacked epidermal cells, whereas Mla1-dependent resistance was not affected. This finding confirmed previous genetic and molecular analyses that indicated that HvRar1 is required for Mla6-triggered but not forMla1-triggered resistance to powdery mildew (13–15). A reduction inMla6-triggered resistance became apparent inHvSgt1-silenced cells 48 hours after bombardment and was more pronounced after 96 hours; at that time, about 63% of epidermal cells were rendered susceptible to attack by the A6 isolate containingAvrMla6. Mla1-triggered resistance inHvSgt1-silenced cells was largely unaffected (1.7%) even at 96 hours, indicating that HvSgt1 is required forMla6-specific but not for Mla1-specific resistance. When HvRar1 and HvSgt1 were cosilenced, Mla6-triggered resistance was highly disabled (83%). These results suggest cooperation between HvRAR1 and HvSGT1 inMla6-triggered resistance. The minor, if any, contribution of HvRAR1 and HvSGT1 to Mla1-triggered resistance is surprising because MLA1 and MLA6 share 95% sequence identity (14, 15). The minor contribution may indicate the existence of another signaling pathway for MLA1. Alternatively, a signal emitted from MLA1 may be stronger than the signal emitted from MLA6, and it may require only residual amounts of HvRAR1 and HvSGT1 to transmit signals to downstream components.

To examine whether HvSGT1 and HvRAR1 also interact in planta, we used antibodies to HvRAR1 and HvSGT1 for coimmunoprecipitation experiments with leaf extracts of barley plants grown in the absence of pathogens. Anti-HvRAR1 immune complexes from wild-type plants contained HvSGT1, which was not detected in rar1-2 plants (Fig. 4). In a reciprocal experiment, HvRAR1 was detected in HvSGT1 complexes; this finding confirms that in the absence of pathogens in planta, HvRAR1 associates with HvSGT1.

Figure 4

HvSGT1 interacts with HvRAR1, SCF, and COP9 siganalosome subunits in vivo. Extracts from wild-type Rar1or mutant rar1-2 barley plants were immunoprecipitated with antibodies to RAR1 or SGT1. Samples of eluted fractions (40 μl) were analyzed by immunoblotting with antibodies to RAR1, SGT1, SKP1 (24), CUL1 (25), CSN4, and CSN5 (Affiniti, Exeter, UK).

In yeast, SGT1 associates with the kinetochore complex and the SCF-type (Skp1-cullin–F-box) E3 ubiquitin ligase by interacting with SKP1 (11). Immunoblotting analysis confirmed that HvSGT1 complexes contained SKP1 homolog(s) (Fig. 4). CUL1 homolog(s), another subunit of the SCF ubiquitin ligase, was also detected in the HvSGT1 immunoprecipitates. Interestingly, the association between HvSGT1 and these subunits also occurs in rar1-2 mutant plants. Because neither SKP1 nor CUL1 homolog(s) was detected in the HvRAR1 immunoprecipitates, at least two distinct HvSGT1 complexes may exist, one containing the subunits of the SCF ubiquitin ligase and the other containing HvRAR1.

Recently, the COP9 signalosome was shown to directly interact with SCF E3 ubiquitin ligases (16, 17). This finding prompted us to investigate the possible interaction between the HvRAR1-HvSGT1 complex and COP9 subunits CSN4 and CSN5. Analyzing HvRAR1 and HvSGT1 immunoprecipitates with antibodies to CSN4 and CSN5 revealed the presence of barley CSN4 (HvCSN4) and CSN5 (HvCSN5) in the HvRAR1-HvSGT1 complex from wild-type Rar1plants (Fig. 4). HvCSN4 and HvCSN5 were also detected in HvSGT1 immunoprecipitates from rar1-2 mutant plants; this result indicates that HvSGT1 associates with HvCSN4 and HvCSN5 independently of HvRAR1.

By its interaction with SCF complexes, the yeast protein SGT1 exerts its essential activity in the degrading of SIC1 (a Clb5/cdc28 kinase inhibitor) and CLN1 (a G1 cyclin) (11). Recent studies have also shown that mouse SGT1 is associated with the SCFSkp2 complex (16). Thus, one possible role of plant SGT1 could be to target resistance-regulating proteins for degradation by the 26Sproteasome via specific SCF complexes. In this scenario, the target proteins would be negative regulators of disease resistance.

SCF E3 ubiquitin ligases can modify protein activity. For example, monoubiquitination regulates membrane transport, subcellular localization, transcription, and protein kinase activity (18). Thus, the SGT1 complex may participate in modification of protein activity or may have a dual role for activation and degradation of the target via ubiquitination. Such a dual role of ubiquitination has been reported in the regulation of transcriptional activation (19). In this context, the R gene product itself could be a target of the SGT1 complex for activation and degradation. Indeed, the R protein RPM1 rapidly degrades during the resistance response (20). This may lead to receptor desensitization that could restrict the HR to a small number of host cells at sites of attempted pathogen invasion.

What is the role of the RAR1-SGT1 interaction in disease resistance? Because yeast does not contain RAR1 homologs, RAR1 itself cannot be required for SCF E3 ligase function, per se. The presence of HvCSN4 and HvCSN5 in the HvRAR1-HvSGT1 complex suggests a link to the COP9 signalosome, which is closely related to the lid complex of the 26S proteasome (21). RAR1 could connect, via SGT1, the COP9 signalosome to SCF E3 ubiquitin ligases to promote ubiquitination of still unknown target(s) in Rgene–mediated signal transduction. Future systematic single-cell silencing experiments of SCF and COP9 signalosome components in combination with protein interaction studies should provide more detailed insights into the role of these complexes in Rgene–specified disease resistance.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: ken.shirasu{at}


View Abstract

Stay Connected to Science

Navigate This Article