SEARCH

SEARCH BY CITATION

Keywords:

  • AtCullin1;
  • SCF;
  • COI1;
  • jasmonate

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The SKP1-Cullin/Cdc53-F-box protein ubiquitin ligases (SCF) target many important regulatory proteins for degradation and play vital roles in diverse cellular processes. In Arabidopsis there are 11 Cullin members (AtCUL). AtCUL1 was demonstrated to assemble into SCF complexes containing COI1, an F-box protein required for response to jasmonates (JA) that regulate plant fertility and defense responses. It is not clear whether other Cullins also associate with COI1 to form SCF complexes, thus, it is unknown whether AtCUL1, or another Cullin that assembles into SCFCOI1 (even perhaps two or more functionally redundant Cullins), plays a major role in JA signaling. We present genetic and physiological data to directly demonstrate that AtCUL1 is necessary for normal JA responses. The homozygous AtCUL1 mutants axr6-1 and axr6-2, the heterozygous mutants axr6/AXR6, and transgenic plants expressing mutant AtCUL1 proteins containing a single amino acid substitution from phenylalanine-111 to valine, all exhibit reduced responses to JA. We also demonstrate that ax6 enhances the effect of coi1 on JA responses, implying a genetic interaction between COI1 and AtCUL1 in JA signaling. Furthermore, we show that the point mutations in AtCUL1 affect the assembly of COI1 into SCF, thus attenuating SCFCOI1 formation.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Ubiquitin-mediated degradation of regulatory proteins plays important roles in the control of numerous processes including cell-cycle progression, signal transduction and transcriptional regulation (Deshaies, 1999; Hershko and Ciechanover, 1998). In this proteolysis system, the regulatory proteins are targeted for degradation by covalent ligation to ubiquitin through sequential actions of a ubiquitin-activating enzyme, a ubiquitin-conjugating enzyme, and a ubiquitin-protein ligase that is a key component essential for both the specificity and timing of protein degradation. Cullin is a central subunit of many ubiquitin-protein ligases. In humans, the Cullin1 protein (CUL1) functions as a scaffold protein in a conserved ubiquitin-protein ligase known as SCF that consists of Skp1, CUL1, an F-box protein, and Rbx1/Hrt/Roc1 (Zheng et al., 2002); the CUL2 protein functions in a ubiquitin-ligase complex containing the VHL tumor suppressor protein elongin-B, RBX1/ROC1, and elongin-C (an Skp1 functional homolog) (Ivan and Kaelin, 2001; Tyers and Jorgensen, 2000). The ubiquitin-ligase complexes containing other Cullins are poorly characterized, although at least six Cullin genes were identified in humans (Shen et al., 2002).

In Arabidopsis, there are 11 Cullin-related genes. Of these 11 Cullin members, five including At4g02570 (AtCUL1), At1g43140, At1g02980, At1g59800, and At1g59790 are most closely related to each other and form a single group, known as the AtCUL1 group, by phylogenetic analysis (Shen et al., 2002). AtCUL1 assembles into the SCF-type ubiquitin-ligase complexes containing the F-box protein TIR1 essential for auxin responses (Gray et al., 1999) or COI1 that mediates jasmonate (JA) signaling (Xu et al., 2002). Hellmann et al. (2003) demonstrated that AtCUL1 plays an important role in auxin signaling: point mutations of AtCUL1 in the axr6 mutants affected assembly of mutant AtCUL1 into a stable SCFTIR1 complex, and reduced degradation of the substrate AXR2/IAA7, causing defects in auxin responses. No genetic or physiological evidence to date demonstrates a biological function for AtCUL1 in the COI1-mediated JA responses. It is therefore not clear whether AtCUL1 alone or together with other Cullins (particularly the remaining members of the AtCUL1 group) is essential for JA signaling.

Jasmonates, which include jasmonic acid and its cyclopentane precursors as well as cyclopentenones (Reymond and Farmer, 1998), regulate a variety of plant developmental processes including pollen development, and also modulate responses to stress, wounding and defense against insects and pathogens (Cheong and Choi, 2003; Creelman and Mullet, 1997; Farmer, 2001; McConn et al., 1997; Staswick, 1992; Wasternack and Parthier, 1997). JA effects in Arabidopsis were defined mainly through analysis of JA biosynthetic mutants including fad (McConn and Browse, 1996), opr3/dde1 (Sanders et al., 2000; Stintzi and Browse, 2000; Stintzi et al., 2001) and aos (Park et al., 2002), and JA-insensitive mutants such as jar1 (Staswick, 1992; Staswick et al., 1998), coi1 (Feys et al., 1994), and jin1 and jin4 (Berger et al., 1996). Among these JA-insensitive mutants, coi1 is defective in all JA responses, and therefore defines a key regulator in the JA signal transduction pathway (Feys et al., 1994). The identification of COI1 as an F-box protein pointed to the involvement of the ubiquitylation pathway in JA signaling (Xie et al., 1998). The F-box protein COI1 assembles SCF complexes in planta with AtCUL1, and ASK1 and ASK2 (Xu et al., 2002), two of 21 Arabidopsis Skp1-like proteins (ASK) (Gagne et al., 2002).

The ASK1 protein, which assembles SCF complexes with TIR1 and COI1 in planta (Gray et al., 1999; Xu et al., 2002), was previously found to be essential for auxin signaling as the ask1 mutant seedlings (Yang et al., 1999) were resistant to auxin (Gray et al., 1999). However, we found that the ask1 mutant seedlings were sensitive to JA inhibition of root growth like wild-type plants, thus ASK1 alone does not play a necessary role in JA signaling (Xu et al., 2002). By analogy, the observation of a vital role for AtCUL1 in auxin signaling (Hellmann et al., 2003) does not necessarily indicate that AtCUL1 also plays an important role in JA signaling though AtCUL1 is a core subunit of both SCFCOI1 and SCFTIR1. As 11 Cullin members are present in the Arabidopsis genome, and it is not known whether other Cullin proteins (particularly the four members from the AtCUL1 group) also function in the SCF complexes containing COI1, it is not clear that only Cullin1 or two (even more) functionally redundant Cullins play a major role in JA signaling.

The JA signal activates SCFCOI1-mediated ubiquitylation and subsequent degradation of repressors of JA response (Xie et al., 1998; Xu et al., 2002). This model further suggests that in mutants with functional mutations within the crucial subunits of SCFCOI1, including coi1, ask1, ask2, and atcul1, putative repressors would not be degraded normally in response to JA, accumulate to high levels, and inhibit JA responses. Currently, the identity of these putative repressors is still a mystery though extensive efforts have been made; and mutations in the core subunits ASK1 or ASK2 do not cause obvious reduced sensitivity to JA inhibition of root growth (Xu et al., 2002). These results open the possibilities that either JA signaling may not depend mainly on SCFCOI1-mediated ubiquitylation, or ASK1-based SCFCOI1 is functionally redundant with ASK2-based SCFCOI1 and even more SCF-containing uncharacterized ASK. Thus, verification of the biological function of AtCUL1 in JA responses would provide new insights into the molecular basis of JA action.

We present genetic, molecular and physiological data to directly demonstrate that AtCUL1 is necessary for normal JA responses. The axr6-1 and axr6-2 mutants (Hobbie et al., 2000) with point mutations of F111V or F111I in AtCUL1 (Hellmann et al., 2003), and transgenic plants expressing mutant AtCUL1, all show reduced responses to JA. In addition, axr6 enhances the effect of coi1 on JA signaling. Furthermore, we demonstrate that the mutations of AtCUL1 reduce the assembly of the F-box protein COI1 into SCFCOI1 complexes. These results indicate that the SCF ubiquitin-protein ligase containing AtCUL1 plays an important role in JA signaling.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The axr6 mutants display reduction in JA sensitivity

To investigate JA responses mediated by the SCFCOI1 component AtCUL1, we screened for and identified a T-DNA insertion line in AtCUL1 (atcul1-5) at +3604 bp relative to the start codon ATG (data not shown). The atcul1-5 mutant is arrested at an early embryonic stage (data not shown), consistent with the observation of the null mutants atcul1-1 and atcul1-2 (Shen et al., 2002). The atcul1-5 mutant is therefore unavailable for analysis of JA response.

Two mutant alleles axr6-1 and axr6-2, which were identified through a genetic screen for auxin-resistant mutants (Hobbie et al., 2000), define a gene identical to AtCUL1 (Hellmann et al., 2003). Sequence analysis showed that a single nucleotide change leads to the substitution of the 111th amino acid in AtCUL1 from phenylalanine to valine (CUL1F111V) in axr6-1, and to isoleucine (CUL1F111I) in axr6-2 (Hellmann et al., 2003). The homozygous mutants axr6-1 and arx6-2 are able to germinate and produce one or two cotyledons (Hobbie et al., 2000), which makes it possible to investigate JA response at the transcriptional level.

The axr6-1, axr6-2 homozygous mutants and wild-type seedlings at a similar developmental stage were treated with methyl jasmonate (MeJA) and assayed for the expression of JA-induced-specific genes. As shown in Figure 1(a), the accumulation of the JA-induced-specific genes AtVSP, COR1, and LOX2 was clearly reduced in the axr6-1 and axr6-2 homozygous mutants compared with that in the wild type upon JA treatment, indicating that the point mutations of AtCUL1 lead to reduction in JA sensitivity.

image

Figure 1. axr6 mutants exhibit reduced jasmonate responses. (a) RNA gel blot analysis of AtVSP, COR1, and LOX2 in axr6-1, axr6-2 homozygous mutants, and Col-0 wild type (WT). 18S rDNA was used as a loading control. (b) Northern analysis of AtVSP, COR1, and LOX2 in heterozygous mutants axr6-1/AXR6 (axr6-1/+), axr6-2/AXR6 (axr6-2/+) and wild type. Total RNA was stained with ethidium bromide as a loading control. (c) Phenotype of 12-day-old seedlings grown on MS medium with or without 25 μm methyl jasmonate (MeJA). (d) MeJA dose–response curve of root growth. Root length of 9-day-old seedlings grown on MS medium containing 1, 5, and 25 μm MeJA was expressed as a percentage of root length on MS medium. Error bars represent SE (n > 30).

Download figure to PowerPoint

We next examined the heterozygous axr6/AXR6 mutants for alterations in JA response. As the heterozygous axr6/AXR6 mutants can develop roots, we carried out the root growth assay on different concentrations of MeJA. As shown in Figure 1(c,d), the axr6-1/AXR6 and axr6-2/AXR6 seedlings displayed weak but clear resistance to JA inhibition of root growth compared with wild type. When assayed for JA-induced-specific gene expression, axr6-1/AXR6 and axr6-2/AXR6 also exhibited reduced JA sensitivity. The expression level of these genes upon JA induction was mildly but notably reduced in axr6-1/AXR6 and axr6-2/AXR6 compared with that in wild type (Figure 1b). These results suggest an important role for AtCUL1 in JA signaling.

Transgenic expression of the mutant AtCUL1 results in partial resistance to JA

To verify whether the mutant AtCUL1 proteins are responsible for the reduction of JA responses, we generated transgenic plants expressing the mutant AtCUL1F111V, which was either 5′Myc-tagged under the control of CaMV 35S constitutive promoter (referred to as mAtCUL1F111V), or 3′Flag-tagged under the control of the endogenous promoter (referred to as fAtCUL1F111V). The wild-type AtCUL1 driven by its endogenous promoter was also 3′Flag-tagged as a control (referred to as fAtCUL1) (Figure 2a) (see Experimental procedures).

image

Figure 2. Molecular characterization and phenotype of transgenic plants. (a) Schematic drawing of constructs of Myc-tagged AtCUL1F111V (mAtCUL1F111V) and Flag(F)-tagged AtCUL1F111V (fAtCUL1F111V) or Flag-tagged AtCUL1 (fAtCUL1) (not to scale) (see Experimental procedures). (b) Northern analysis of the transgenic lines expressing Myc-tagged AtCUL1F111V (mFV1, mFV2, and mFV3 line), Flag-tagged AtCUL1F111V (fFV1 line) and Flag-tagged AtCUL1 (fCUL1 line). RNA blots were hybridized with the indicated probes including AtCUL1, Myc-, or Flag-containing fragments (Myc, Flag). (c) Western blot analysis of the plants described in (b) with the indicated antibodies including anti-AtCUL1, anti-ASK1, and anti-COI1. ASK1 or COI1 protein was used as a loading control. Approximately100 μg of total protein was loaded for each sample. (d) The morphology of wild type, axr6-1/AXR6 (axr6-1/+), and transgenic lines mFV1, fFV1, fCUL1, and vector (plants transformed with vector control). Top panel, soil-grown seedlings; middle panel, adult plants; bottom panel, enlarged siliques and main bolts. (e) 2,4-dichlorophenoxyacetic acid (2,4-D) dose–response column of root growth. Root length of 8-day-old seedlings grown on MS medium containing 0.1 μm 2,4-D was expressed as a percentage of root length on MS medium. Error bars represent SE (n > 15).

Download figure to PowerPoint

As shown in Figure 2(b,c), we identified three independent lines that highly express mAtCUL1F111V at the transcription level and at the translation level (referred to as mFV1, mFV2, and mFV3), a transgenic line with expression of fAtCUL1F111V (referred to as fFV1) and its control line similarly expressing transgenic fAtCUL1 (referred to as fCUL1). Compared with the control lines (fCUL1, wild type, and wild-type transgenic for empty vector), the transgenic lines expressing the mutant AtCUL1 protein (mFV1, mFV2, mFV3, and fFV1) all exhibited alterations similar to the axr6-1/AXR6 plants (Figure 2d, and data not shown): their rosette leaves were short and wrinkled, the plant height was severely reduced, their siliques lay at a more acute angle relative to the stem, and they also mimicked other phenotypes of axr6-1/AXR6, including insensitivity to auxin (Figure 2e, and data not shown).

We next investigated the possible effects of the mutant AtCUL1 expression on JA responses. Compared with the control lines (fCUL1, wild type, and the empty vector line), the transgenic lines expressing the mutant AtCUL1 protein (mFV1, mFV2, mFV3, and fFV1) all exhibited a clear reduction in the expression of AtVSP, COR1, and LOX2 upon various MeJA treatments (Figure 3a–c, and data not shown). Quantitative analysis of Figure 3(c) using a scanning densitometer showed that the abundance of AtVSP in the mFV3 transgenic line treated with MeJA for 6 or 12 h was approximately 55% of that in its corresponding wild-type control. The transgenic lines including mFV1, mFV2, mFV3, and fFV1 also displayed obvious resistance to JA inhibition of root growth (Figure 3d, and data not shown). These results demonstrate that transgenic expression of the mutant AtCUL1 protein results in reduced sensitivities to JA, and directly verify that the F111V mutation in AtCUL1 is responsible for the reduced JA sensitivity and other axr6 mutant phenotypes.

image

Figure 3. Expression of mutant AtCUL1F111V results in reduction in jasmonate response. (a) Northern analysis of AtVSP, and COR1 in wild type (WT), vector, mFV1 and fFV1 treated with methyl jasmonate (MeJA) (+) or water (−). Ethidium bromide staining of rRNAs shown at bottom indicates the loading amount of total RNA on the gel. (b) Transcript levels of AtVSP and LOX2 in WT and the mFV2 line upon treatment with MeJA for 0, 4, 8, and 12 h. (c) Transcript levels of AtVSP in WT and the mFV3 line treated with methyl jasmonate (MeJA) (+) or water (−) for 6 and 12 h. (d) MeJA dose–response curve of root growth. Root length of 9-day-old seedlings grown on MS medium containing 1, 5, and 25 μm MeJA was expressed as a percentage of root length on MS medium. Error bars represent SE (n > 30).

Download figure to PowerPoint

COI1 and AtCUL1 exhibit a genetic interaction in JA responses

To determine whether COI1 and AtCUL1 exhibit a genetic interaction in JA responses, we generated the double homozygous mutant coi1-2/coi1-2 axr6-1/axr6-1 (referred to as coi1 axr6) through crossing of axr6-1/AXR6 heterozygous plant with coi1-2, a leaky mutant allele of COI1 (Xu et al., 2002). As the coi1 axr6 seedlings exhibited severe morphological alterations similar to axr6-1 (Hobbie et al., 2000) (Figure 4a) and cannot be used for root growth analysis, we investigated JA response in coi1 axr6 only at the transcription level. As expected, the expression of JA-inducible genes was highly elevated in wild type upon JA treatment, but severely reduced in coi1-2 (Xu et al., 2002) (Figure 4b) and obviously reduced in arx6-1 (Figures 1a and 4b). However, in the coi1 axr6 double mutant seedlings treated with MeJA for various periods, the expression of JA-induced-specific genes such as AtVSP was not detectable (Figure 4b, and data not shown), and even was much lower than the background level of the untreated wild type as revealed by overexposure of X-ray film during Northern hybridization (data not shown). These data suggest that axr6 enhances the effect of coi1 on JA response at the transcriptional level.

image

Figure 4. Morphological phenotype and jasmonate-inducible gene expression in the coi1-2 axr6-1 double mutant. The coi1-2 axr6-1 (coi1axr6) double mutant, coi1-2, axr6-1 and wild type (a) were treated with methyl jasmonate (MeJA) (+) or water (−) for 6, 8 or 12 h. Total RNA from these plant materials was probed with AtVSP and 18S rDNA probes (b). 18S rDNA was used as a loading control.

Download figure to PowerPoint

To test whether axr6 enhances the effect of coi1 on JA inhibition of root growth, we generated the coi1-2/coi1-2 axr6-1/AXR6 double mutant (referred to as coi1 axr6/+) (homozygous for coi1-2 but heterozygous for axr6-1), which exhibits the axr6-1/AXR6 morphological phenotypes (Hobbie et al., 2000) except for the partial fertility like coi1-2 (Xu et al., 2002) (Figure 5a). As shown in Figure 5(b,c), the coi1 axr6/+ seedlings, when treated with various concentrations of MeJA, produced mildly but obviously longer roots compared with the single mutant coi1-2 or axr6-1/AXR6, indicating an enhanced resistance to JA inhibition of root growth in the coi1 axr6/+ double mutant. The coi1 axr6/+ double mutant exhibited no enhanced resistance to auxin compared with the single mutant axr6-1/AXR6 (Figure 5d). These data demonstrate that axr6 enhances the effect of coi1 on JA signaling, suggesting a genetic interaction between COI1 and AtCUL1.

image

Figure 5. Analysis of the coi1-2/coi1-2 axr6-1/AXR6 double mutant. (a) The morphology of the wild-type (WT), axr6-1/AXR6 (axr6/+), coi1-2, and coi1-2/coi1-2 axr6-1/AXR6 (coi1 axr6/+) seedlings (top panel) and flowering plants (bottom panel). (b) Phenotype of 12-day-old seedlings grown on MS medium or MS containing 25 μm methyl jasmonate (MeJA). (c) MeJA dose–response curve of root growth. Root length of 9-day-old seedlings grown on MS medium or MS containing 1, 5, and 25 μm MeJA was measured. Error bars represent SE (n > 30). (d) 2,4-dichlorophenoxyacetic acid (2,4-D) dose–response column of root growth. Root length of 8-day-old seedlings grown on MS medium containing 0.1 μm 2,4-D was expressed as a percentage of root length on MS medium. Error bars represent SE (n > 15).

Download figure to PowerPoint

Point mutations in AtCUL1 attenuate SCFCOI1 complex formation

We previously demonstrated that point mutations in COI1 disrupt SCFCOI1 formation leading to loss of JA responses (Xu et al., 2002). Here we tested whether the point mutations in AtCUL1 affect SCFCOI1 formation. We used the anti-Flag monoclonal antibody to immunoprecipitate protein complexes from the transgenic line fFV1 expressing the Flag-tagged mutant AtCUL1 (fAtCUL1F111V) and from the fCUL1 transgenic line expressing the Flag-tagged wild-type AtCUL1 (fAtCUL1), respectively.

Subsequent Western blotting analysis of the resulting immunoprecipitates showed that the COI1 protein was able to co-immunoprecipitate with both fAtCUL1F111V and fAtCUL1. However, the abundance of COI1 protein was significantly reduced in the fAtCUL1F111V-immunoprecipitates (Figure 6b), and quantitative analysis indicated that the abundance of COI1 in fAtCUL1F111V-immunoprecipitates was approximately one third of that in fAtCUL1-immunoprecipitates (Figure 6c). The ASK1 protein that is present in the fAtCUL1F111V-immunoprecipitates was also reduced to about one third of that in fAtCUL1-immunoprecipitates (Figure 6b,c), consistent with the observation that the interaction of ASK1 with the mutant AtCUL1 was reduced in the axr6-1 and axr6-2 mutants (Hellmann et al., 2003). We also found that the interaction of the mutant AtCUL1 with ASK2 was similarly reduced (data not shown), whereas the association of the mutant AtCUL1 with RBX1 appeared to be normal (Figure 6b,c).

image

Figure 6. Point mutation in AtCUL1 attenuates SCFCOI1 complex formation. Total protein extracts (a) from wild type, transgenic lines expressing Flag-tagged AtCUL1 (fCUL1) and Flag-tagged AtCUL1F111V (fFV1) were immunoprecipitated (IP) with anti-Flag antibody (b), then detected with the indicated antibodies. (c) Quantitative analysis of abundance of the immunoprecipitated proteins described in (b). The abundance of protein bands immunoprecipitated from the fFV1 line is calculated relative to each corresponding protein bands of the fCUL1 line that is set to 100. The experiment was repeated five times. Arrow shown in (b) indicates immunoglobulin light chain (LC).

Download figure to PowerPoint

In summary, these results suggest that the point mutation in AtCUL1 affects AtCUL1 protein interaction with ASK1 and ASK2, attenuates SCFCOI1 formation, and results in a reduction in JA responses.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

For plants, SCF-dependent ubiquitylation appears to be particularly important as the genome of Arabidopsis encodes almost 700 F-box proteins, which is more than that in similarly complex organisms from other kingdoms (Gagne et al., 2002). Although the function of only a limited number of F-box proteins is known, it is clear that they are important in many aspects of plant biology (Vierstra, 2003). In particular, plant hormone signaling seems to be subjected to SCF-dependent regulation (Frugis and Chua, 2002): the signaling pathways of auxin, JA, gibberellin and ethylene are controlled by SCF-like complexes involving the F-box proteins TIR1 (Ruegger et al., 1998), COI1 (Xie et al., 1998), SLEEPY1 (McGinnis et al., 2003)/GID2 (Sasaki et al., 2003), and EBF1/EBF2 (Guo and Ecker, 2003; Potuschak et al., 2003), respectively. Among these F-box proteins, TIR1 (Gray et al., 1999) and COI1 (Xu et al., 2002) form SCF-type complexes with AtCUL1 in planta.

Shen et al. (2002) showed that two different T-DNA-tagged knockout mutants (atcul1-1 and atcul1-2) are arrested during early embryogenesis. Consistent with this result, we also found that a null mutation of AtCUL1 (atcul1-5) leads to lethality in early embryogenesis. These dramatic defects in plant development caused by the loss of functional AtCUL1 suggest that AtCUL1 is the central component in many important SCF complexes. More useful for many purposes than the null mutants are the axr6-1 and axr6-2 mutants which both carry a base substitution in AtCUL1 (Hellmann et al., 2003). The axr6 mutants displayed abnormalities in embryo development and lethality in seedling growth whereas the heterozygous axr6/AXR6 mutant plants exhibited morphological abnormalities and defects in auxin signaling (Hellmann et al., 2003; Hobbie et al., 2000). Hellmann et al. (2003) demonstrated that axr6 mutations affect the ability of mutant AtCUL1 to assemble into a stable SCFTIR1 complex, resulting in reduced degradation of the substrate AXR2/IAA7. In this study, we specifically investigated a biological function for AtCUL1 in the JA signaling pathway.

The axr6 mutants exhibit reduced JA responses

Several experimental observations presented here show that functional AtCUL1 protein is required for normal JA response. (1) The axr6-1 and axr6-2 homozygous mutants display an obvious reduction in the induction of various genes that are regulated by JA. (2) Although to a lesser extent than the homozygous mutants, the axr6-1/AXR6 and axr6-2/AXR6 heterozygous mutants also exhibited alterations in JA responses, including a reduced induction of the JA-responsive genes and a partial JA insensitivity of root growth. (3) Furthermore, we demonstrated that axr6 has an enhanced effect on coi1, suggesting a genetic interaction between COI1 and AtCUL1 in JA signaling. (4) Finally, transgenic plants expressing the mutant AtCUL1 protein all mimicked the partial JA-insensitive phenotypes like the axr6 mutants, indicating that the mutant AtCUL1 confers gain-of-function of JA-insensitive phenotypes. The observation that JA responses were altered in all transgenic plants, which express various forms of mutant AtCUL1 fused with Myc or Flag epitope either at N-terminal or C-terminal under the control of either the 35S promoter or the endogenous promoter, also excludes the possibilities that the Myc or Flag tag may affect the overall folding of AtCUL1, which in turn may affect JA responses, and that alterations of JA responses may result from misexpression of AtCUL1 driven by the exogenous promoter. All these data clearly demonstrate an important biological role of AtCUL1 in JA signaling.

By analogy with the auxin signal transduction pathway, we proposed a model for JA action in which JA signal activates SCFCOI1-mediated ubiquitylation of repressors of JA response (Xie et al., 1998; Xu et al., 2002). However, genetic mutants for SCFCOI1 substrates have not been identified despite extensive efforts. The wild-type sensitivity to JA inhibition of root growth observed in the ask1 and ask2 null mutants might further raise the possibility that JA signaling may not depend mainly on the SCFCOI1-mediated ubiquitylation. The present results demonstrate an important role for AtCUL1 in JA responses, implicating the SCF-type ubiquitin ligase containing AtCUL1 plays a vital role in JA signaling. We showed that ask1 ask2 double mutants exhibit defects in embryogenesis and lethality in seedling growth whereas these severe phenotypes are not detectable in the single mutants (Liu et al., 2004). This finding indicates that the lack of defects of JA responses in the ask1 and ask2 null mutants is more likely caused by functional redundancy between ASK1- and ASK2- based SCFCOI1. Genetic redundancy among putative JA repressors may also make isolation of corresponding repressor mutants difficult by standard genetic mutagenesis screens. We identified a coi1 suppressor that represents a riboflavin pathway, indicating the possibility that the putative substrates of SCFCOI1 require the riboflavin pathway to exert their suppression action on JA responses (Xiao et al., 2004).

The axr6 mutations impair the formation of the SCFCOI1 complex

axr6 mutations affect the formation of SCFTIR1 (Hellmann et al., 2003). We investigated whether the mutations in AtCUL1 would reduce the formation of SCFCOI1in planta. We performed co-immunoprecipitation experiments of the SCFCOI1 complex from the transgenic lines expressing the mutant (the fFV1 line) and wild-type AtCUL1 (the fCUL1 line). Indeed we observed a strong reduction in COI1 and ASK1 in the immunoprecipitates from the fFV1 line, compared with the fCUL1 line. However, comparable levels of the RBX1 protein were immunoprecipitated from the fFV1 and fCUL1 lines. This result is consistent with the location of the axr6-1 mutation in the AtCUL1 N-terminal domain that interacts with Skp1, whereas RBX1 interacts with the C-terminus of the Cullin (Zheng et al., 2002). We speculate that the fFV1 line contains a low abundance of SCFCOI1 compared with fCUL1. Therefore, mutant AtCUL1 proteins confer the JA-insensitive phenotypes in the fFV1, mFV, and axr6/AXR6 lines probably via attenuation of SCFCOI1 formation.

We previously demonstrated that abolition of SCFCOI1 formation by the E22A single amino acid replacement in COI1 disrupts all the JA responses (Xu et al., 2002). The results described here show that the AtCUL1 point mutations in the axr6 mutants reduce the level of SCFCOI1 and thus cause a reduction in JA responses, suggesting that an adequate level of functional SCFCOI1 is important for JA responses. It is worthwhile to note that different JA responses may have different sensitivities to SCFCOI1 levels. As the axr6/AXR6 plants are fertile, it seems that the JA-regulated fertility may have less sensitivity to the SCFCOI1 level. Accordingly, Feng et al. (2003) showed that a slightly lower expression of COI1 has no effect on pollen fertility.

The axr6 mutants provide a powerful tool to study SCF in plants. However, great care needs to be taken when interpreting alterations of signal transduction pathways and development due to the large number of AtCUL1-based SCF complexes that may be downregulated in these mutants. This may particularly be true for the interpretation of hormonal effects, as many plant hormones act in a complex web of overlapping signaling pathways which function sometimes in synergistic or antagonistic ways (Gazzarrini and McCourt, 2003; Leon and Sheen, 2003).

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Plant materials

The AtCUL1 knockout mutant atcul1-5 was isolated from a T-DNA-tagged pool produced in the Wisconsin knockout facility (http://www.biotech.wisc.edu/Arabidopsis). The AtCUL1 mutant alleles axr6-1 and axr6-2 were previously described (Hobbie et al., 2000). Seeds were surfaced-sterilized, chilled at 4°C for 3 days, and then germinated and grown on plant growth medium (Murashige and Skoog supplemented with 1% sucrose) under a 16 h light (22–24°C)/8 h dark (16–19°C) photoperiod. Similar conditions were followed for soil-grown plants, except that the plants shown in Figure 5(a) were grown at 18–20°C.

Plasmid constructs and Arabidopsis transformation

A 4121-bp genomic fragment of the AtCUL1F111V gene was amplified by Pfu DNA polymerase (Stratagene, La Jolla, CA, USA) from the axr6-1 homozygote using the forward primer P1 (5′-ATGGAGCGCAAGACTATTGAC-3′) located at the AtCUL1 start codon and the reverse primer 4121P (5′-CCAGAGCTCCTAAGCCAAGTACCTAAACAT-3′) at the stop codon-supplemented SacI site. The fragment was then fused to the Myc epitope at SmaI/SacI sites in pMyc2 (Xu et al., 2002), resulting in mAtCUL1F111V (Figure 2a).

A 5221-bp genomic fragment containing the AtCUL1 promoter region and the coding sequence without the stop codon was amplified by Pfu DNA polymerase from wild type or axr6-1 homozygote, respectively, using the forward primer P1103 (5′-ACCGAGCTCCCGATTTCTATCT-3′) at 1103 bp upstream of the ATG start codon and the reverse primer 4118P (5′-TTTTCCCGGGAGCCAAGTACCTAAACATGT-3′) that was engineered with an SmaI site to replace the AtCUL1 stop codon and allow an in-frame fusion with Flag in the vector pFlag at SacI/SmaI sites. The pFlag was constructed in our laboratory by cloning the Flag epitope sequence (GATTACAAGGATGATGATGATAAGTGA) at the SmaI and BstEII sites into plasmid pCambia1305.2 vector with the gus reporter gene replaced. The AtCUL1 and AtCUL1F111V fragments were cloned into the pFlag vector at the SacI/SmaI sites, resulting in the fAtCUL1 and fAtCUL1F111V constructs (Figure 2a).

These constructs were verified by sequencing and introduced into Arabidopsis plants (Col-0) by the ‘floral-dip’ method of in plantaAgrobacterium tumefaciens-mediated transformation (Clough and Bent, 1998).

RNA gel blot analysis

AtVSP and 18S rDNA probes were as previously described (Xu et al., 2001, 2002); the AtCUL1 fragment (5′-TCTTGGAGCAAGGATGGGACTA-3′, 5′-CTGCTTGAGTGTAGGTAGTG-3′), the Myc-containing fragment (5′-TGACGTAAGGGATGACGCACA-3′, 5′-TCAATAGTCTTGCGCTCCAT-3′) and the Flag-containing fragment (5′-TCCTTAAGGAGCCAAACACC-3′, 5′-ATGTATAATTGCGGGACTCTAA-3′) were amplified with their specific primers. Special primers were also designed based on their DNA sequences to PCR-amplify probes COR1 (Benedetti et al., 1998) and LOX2 (Freire et al., 2000). Probe labeling and Northern hybridization methods were as previously described (Xu et al., 2001). Approximately 30 μg of total RNA for each sample was loaded in Northern blot analysis.

For JA treatment in Northern analysis, seedlings were grown on MS medium with 1% sucrose for 2–3 weeks except for 1 week in Figures 1(a) and 4(b), and then drenched in solution containing 100 μm MeJA (Aldrich, Milwaukee, WI, USA) or water for 6 h in daytime or for various periods (indicated in Figures).

Root length measurement

To test JA-inhibitory root growth, seedlings were grown on MS medium supplemented with various concentrations of MeJA for 9 days before measurement; all experiments were repeated three to five times. For axr6/AXR6 and coi1 axr6/+ seedlings, we first measured the root length of each F2 seedling, and then transferred each seedling onto MS medium containing 0.1 μm 2,4-dichlorophenoxyacetic acid (2,4-D) (Sigma, Saint Louis, MO, USA) for confirmation of the axr6/AXR6 and coi1 axr6/+ that are resistant to 2,4-D.

To test auxin-inhibitory root growth, seedlings were grown on MS medium supplemented with 0.1 μm 2,4-D for 8 days before measurement. The experiments were repeated three times.

Coimmunoprecipitation assay and protein gel blot analysis

The coimmunoprecipitation assay and Western analysis were as previously described (Xu et al., 2002).

Antibodies

Anti-COI1, anti-ASK1, anti-AtCUL1, anti-RBX1, secondary antibodies (goat anti-rabbit IgG-horseradish peroxidase, and goat anti-mouse IgG-HRP), and Flag affinity matrix were previously used in our laboratory (Xu et al., 2002).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank the Wisconsin knockout facility for T-DNA insertion pool. This work was supported by the Singapore Agency of Science, Technology and Research to D.X., the US National Science Foundation grant (IBN 998926) to L.H., and the French Ministry of Research's Action Concertee Incitative ‘Jeune Cherchurs’ grant to P.G.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References