Structural insights into alternative splicing-mediated desensitization of jasmonate signaling Feng Zhanga,b,c,d,e, Jiyuan Kec,d, Li Zhangb,f, Rongzhi Chenb, Koichi Sugimotob, Gregg A. Howeb,g,h, H. Eric Xuc,i, Mingguo Zhoua,1, Sheng Yang Heb,f,h,j,1, and Karsten Melcherd,1 a College of Plant Protection, Nanjing Agricultural University, 210095, Nanjing, Jiangsu Province, China; bDepartment of Energy Plant Research Laboratory, Michigan State University, East Lansing, MI 48824; cLaboratory of Structural Sciences, Van Andel Research Institute, Grand Rapids, MI 49503; dLaboratory of Structural Biology and Biochemistry, Van Andel Research Institute, Grand Rapids, MI 49503; eState Key Laboratory of Hybrid Rice, College of Life Sciences, Wuhan University, Wuhan 430072, China; fDepartment of Plant Biology, Michigan State University, East Lansing, MI 48824; gDepartment of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI 48824; hPlant Resilience Institute, Michigan State University, East Lansing, MI 48824; iKey Laboratory of Receptor Research, VARI-SIMM Center, Center for Structure and Function of Drug Targets, Shanghai Institute of Materia Medica, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China; and jHoward Hughes Medical Institute, Michigan State University, East Lansing, MI 48824
Contributed by Sheng Yang He, December 28, 2016 (sent for review October 14, 2016; reviewed by Roberto Solano and Ning Zheng)
Jasmonate ZIM-domain (JAZ) transcriptional repressors play a key role in regulating jasmonate (JA) signaling in plants. Below a threshold concentration of jasmonoyl isoleucine (JA-Ile), the active form of JA, the C-terminal Jas motif of JAZ proteins binds MYC transcription factors to repress JA signaling. With increasing JA-Ile concentration, the Jas motif binds to JA-Ile and the COI1 subunit of the SCFCOI1 E3 ligase, which mediates ubiquitination and proteasomal degradation of JAZ repressors, resulting in derepression of MYC transcription factors. JA signaling subsequently becomes desensitized, in part by feedback induction of JAZ splice variants that lack the C-terminal Jas motif but include an N-terminal cryptic MYC-interaction domain (CMID). The CMID sequence is dissimilar to the Jas motif and is incapable of recruiting SCFCOI1, allowing CMID-containing JAZ splice variants to accumulate in the presence of JA and to re-repress MYC transcription factors as an integral part of reestablishing signal homeostasis. The mechanism by which the CMID represses MYC transcription factors remains elusive. Here we describe the crystal structure of the MYC3–CMIDJAZ10 complex. In contrast to the Jas motif, which forms a single continuous helix when bound to MYC3, the CMID adopts a loop–helix–loop–helix architecture with modular interactions with both the Jas-binding groove and the backside of the Jasinteraction domain of MYC3. This clamp-like interaction allows the CMID to bind MYC3 tightly and block access of MED25 (a subunit of the Mediator coactivator complex) to the MYC3 transcriptional activation domain, shedding light on the enigmatic mechanism by which JAZ splice variants desensitize JA signaling.
|
signaling plant hormone plant insect
binding of TPL to ethylene-response factor-associated amphiphilic repression (EAR) motifs found at the N terminus of a subset of JAZ proteins or indirectly through the EAR motif-containing NINJA adaptor protein that binds to the central ZIM domain of JAZ proteins (Fig. 1A) (12–16). TPL proteins in turn recruit repressive histone deacetylases and chromatin remodelers (17, 18) to inhibit MYC target gene expression epigenetically. In response to stress or developmental cues, plants synthesize JA-Ile, which promotes the formation of a coreceptor complex consisting of JAZ (the Jas motif), JA-Ile, and coronatine insensitive 1 (COI1), an F-box protein that is a subunit of a Skp1–Cul1–F-box protein (SCF) ubiquitin E3 ligase (4, 19, 20). The JAZ–COI1 interaction leads to ubiquitination and proteasome-dependent degradation of JAZ repressors and thereby to the release of MYC proteins from transcriptional repression (21, 22). In addition to the highly conserved C-terminal Jas motif, JAZ10 also contains an N-terminal cryptic MYC-interaction domain (CMID) that can bind to the N terminus of MYC proteins but does not form complexes with JA-Ile and COI1 (23–25). Moreover, alternative splicing of JAZ10 mRNA generates three splice variants that differ in their C termini: JAZ10.1 (full-length JAZ10, amino acids 1–197), JAZ10.3 (amino acids 1–185, lacking the C Significance Jasmonate (JA) is a plant hormone involved in regulating defense response, growth, and development. A mechanistic understanding of JA signaling has great importance in agriculture, especially for enhancing plant resilience against biotic stresses and optimizing defense and growth in crop fields. This study reports the crystal structure of a key protein complex (MYC3– CMIDJAZ10) involved in the desensitization of JA signaling. Desensitization of JA signaling after an effective JA response is necessary to re-establish signal homeostasis and prevent JA signaling from running out of control, which could lead to severe growth and fitness penalties. The MYC3–CMIDJAZ10 desensitization complex structure reported here closes a major gap in our understanding of one of the most important hormone signaling pathways in plants.
| plant defense | plant pathogen |
T
he plant hormone jasmonate (JA) is a critical mediator of plant defense responses and an important regulator of plant growth and development (1–4). The bioactive form jasmonoyl isoleucine (JA-Ile) is produced by conjugation of jasmonic acid, a fatty acid-derived oxylipin, with isoleucine in the plant cytosol (5– 7). Under stress-free conditions, when JA-Ile levels are low, JA responses are restrained by a group of nuclear transcriptional repressors called “jasmonate ZIM-domain” (JAZ) proteins (2, 4, 8). In Arabidopsis, 13 genes (JAZ1–JAZ13) encode JAZ proteins with a highly conserved C-terminal Jas motif (9, 10). JAZ proteins physically bind and inhibit basic helix–loop–helix transcription factors, including MYC2, MYC3, and MYC4 (2, 9), through a dual mechanism: First, a recent study of the crystal structure of the MYC3 transcription factor in complex with JAZ1 and JAZ9 showed that the conserved Jas motif of JAZ proteins directly binds the transcriptional activation domain (TAD) of MYC proteins and thereby blocks access of the TAD to the MED25 subunit of the mediator complex and possibly to other coactivators whose interaction with the TAD is required for MYC target gene transcription (11) (Fig. 1A). Second, JAZ proteins recruit members of the TOPLESS (TPL) class of corepressors either through direct
1720–1725 | PNAS | February 14, 2017 | vol. 114 | no. 7
Author contributions: H.E.X., M.Z., S.Y.H., and K.M. designed research; F.Z., J.K., L.Z., R.C., K.S., and G.A.H. performed research; F.Z., J.K., L.Z., R.C., K.S., G.A.H., H.E.X., M.Z., S.Y.H., and K.M. analyzed data; and F.Z., S.Y.H., and K.M. wrote the paper. Reviewers: R.S., Centro Nacional de Biotecnologia (Consejo Superior de Investigaciones Científicas); and N.Z., University of Washington. The authors declare no conflict of interest. Data deposition: Crystallography, atomic coordinates, and structure factors reported in this paper have been deposited in the Protein Data Bank database (PDB ID codes 5T0Q and 5T0F). 1
To whom correspondence may be addressed. Email: [email protected], [email protected], or [email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1616938114/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1616938114
terminus of the Jas motif), and JAZ10.4, which lacks the entire Jas motif (amino acids 1–147, followed by 20 unrelated amino acids) (Fig. 1A). As a consequence, all three splice forms retain the ability to bind MYC and repress MYC target gene expression, but JAZ10.3 is partially resistant and JAZ10.4 is fully resistant to JA-induced degradation (8, 23, 26). Because JAZ10 expression itself is induced rapidly by JA, and JAZ10.4 protein accumulates following an increase in JA levels, JAZ10.4 has been suggested to be an important feedback regulator to mediate JA signal desensitization (25). This model has been supported further by genetic data. Although there is significant functional redundancy among JAZ proteins in Arabidopsis, jaz10 loss-of-function mutations result in JA hypersensitivity (27). Conversely, although plants overexpressing JAZ10.1 remain fully responsive to JA, overexpression of JAZ10.3 and JAZ10.4 causes moderate and severe JA insensitivity, respectively (8, 28, 29). JAZ1 also contains a functional N-terminal CMID, which has weak similarity to the JAZ10 CMID (24). However, because the CMID has no sequence similarity to the well-characterized Jas motif, how the CMID interacts with the MYC N-terminal domain and blocks MYC transcriptional activity is unknown. Here we report the crystal structure of CMIDJAZ10 in complex with the MYC3 N-terminal domain. For direct comparison, we also determined the crystal structure of the MYC3–JasJAZ10 complex. We found that CMIDJAZ10 and JasJAZ10 adopt strikingly Zhang et al.
different conformations, despite their overlapping binding sites in MYC3, leading in the case of the MYC3–CMIDJAZ10 complex to a modular MYC interaction. The MYC3–CMIDJAZ10 complex structure assisted us in the discovery of functional CMIDs in JAZ5, JAZ6, and possibly JAZ2, revealing a potentially widespread mechanism of JAZ splice variant-mediated desensitization of JA signal transduction. Results The MYC3 N-Terminal Domain and CMIDJAZ10.4 Are Sufficient to Mediate the MYC–JAZ10.4 Interaction. We used yeast two-hybrid
(Y2H) assays to determine the MYC3-binding region within JAZ10.4. A region in the N terminus of MYC3, MYC3(44–238), which is conserved among MYC2, MYC3, and MYC4 (30–32), contains the JAZ-interaction domain (JID)/TAD and is sufficient to interact with JAZ10.4 and JAZ1ΔJas. Conversely, the 43-aa CMIDJAZ10 [JAZ10(16–58)] is sufficient to interact with MYC3 (Fig. S1). We used a fusion protein strategy that we developed previously (11) to express and purify a stable MYC3/CMID complex consisting of MYC3(44–242) N-terminally fused to JAZ10(16–58) through a structurally flexible 12-aa linker (MYC3– CMIDJAZ10). This fusion protein formed high-quality crystals and allowed us to determine the structure of the complex at a resolution of 2.4 Å (Fig. 1B and Table S1). To be able to compare the MYC3–CMIDJAZ10 complex with the MYC3–JasJAZ10 complex, PNAS | February 14, 2017 | vol. 114 | no. 7 | 1721
PLANT BIOLOGY
Fig. 1. Structures of the MYC3(44–242)–CMIDJAZ10 and MYC3(44–242)–JasJAZ10 complexes. (A) Schematic structure of JAZ9, JAZ10.1, and JAZ10.4 proteins. Numbers indicate amino acid residues. JAZ10.4-N: amino acids 1–58; JAZ10.4-C: amino acids 59–167; CMID: amino acids 16–58; Core CMID: amino acids 35–58. (B) Structure of MYC3(44–242) in complex with CMIDJAZ10. (C) Structure of MYC3(44–242) in complex with JasJAZ10. (D) Overlay of the MYC3(44–242)– CMIDJAZ10 complex with the MYC3(44–242)–JasJAZ10 complex. Note that “JID” and “TAD” are only formal designations. Although they were originally mapped as two distinct regions, both regions are part of the same structural fold and are required for JAZ binding. (E) The CMIDJAZ10 forms extensive, modular interactions with the JID and TAD in MYC3. The MYC3(44–242)–CMIDJAZ10 structure with important interacting residues shown in stick presentation, and bonds are shown as dashed lines. CMIDJAZ10 is shown in magenta, JIDMYC3 is shown in cyan, and TADMYC3 is shown in green.
Fig. 2. CMIDJAZ10 and JasJAZ10 peptides competitively inhibit the MYC3– MED25 interaction. AlphaScreen competition assay of the interaction between His6Sumo–MED25(407–680) (50 nM) and biotinylated MYC3(5–242) (50 nM) with increasing concentrations of CMIDJAZ10 or JasJAZ10 peptide. n = 3 replicates; error bars indicate SD.
we also expressed and purified an MYC3(44–242)–14-aa linker– JAZ10(166–192) fusion protein, MYC–JasJAZ10, and determined its structure at a resolution of 2.2 Å (Fig. 1C and Table S1). The CMID Makes a Modular Interaction with MYC3 and Competes with the MYC3–MED25 Interaction. The structure of the MYC3–
JasJAZ10 complex is essentially superimposable with those of the previously determined MYC3–JasJAZ9 and MYC3–JasJAZ1 complexes (Fig. S2) (11), reflecting the high conservation of Jas motifs. In all three structures, the Jas motifs form a single continuous α-helix when bound to MYC3. In contrast, the CMID in the MYC3– CMIDJAZ10 complex consists of two short helices and two loops that make more extensive interactions with MYC3 than with the Jas helix (Fig. 1 B–D). Although residues 26–57 of the CMIDJAZ10 are resolved in the complex structure, only the C-terminal part of the CMID (residues 34–57, the core CMID) makes direct contacts with MYC3. The position of the N-terminal CMID helix (α1) overlaps with the position of the longer Jas helix in the MYC3–JasJAZ10 complex; α1 thus mimics the ability of the Jas motif to block the access of the TAD to transcription coactivators. In agreement with the position of α1, the CMID competes with the interaction between MYC3 and the activator-interacting domain (ACID) of MED25 (Fig. 2, orange competition curve). In addition (and absent in the MYC3–Jas interaction), the C-terminal helix and I46 of the α1–α2 connecting loop of CMIDJAZ10 wrap around the JID helix and make additional, predominantly hydrophobic, contacts with the back side of the binding groove (Fig. 1 B and E and Table S2).
generated JAZ10.4 double mutants in which we combined a mutation in the CMID α2 helix (L54A) with a mutation at the N-loop/α1 border (F36A), the α1 helix (I43A), or the α1–α2 loop (I46A). All three double mutants lost the ability to interact with MYC3 in the Y2H assay (Fig. 3B), even though these mutant proteins were expressed in yeast cells at levels equal to or higher than that of wild-type JAZ10.4 (Fig. S3A, Right), suggesting that at least two of these four structural elements must be mutated to abolish the MYC3–JAZ10.4 interaction. We also combined the L54A mutation with mutations in two hydrophilic residues in the α1 helix, S44 and R37, which form hydrogen bonds with MYC3. However, only mutants that contain the more drastic R37D charge reversal mutation, but neither the L54A/S44A and L54A/ R37A double mutants nor an L54A/S44A/R37A triple mutant, could disrupt the interaction with MYC3 (Fig. 3B), indicating that the interaction between the solvent-exposed MYC JID/ TAD and JAZ CMID is driven predominantly by Van der Waals interactions. Next, we introduced mutations into CMID-interacting residues of MYC3. Similar to the corresponding mutations in JAZ10.4, specific sets of double mutations in MYC3 were required to abolish or reduce the MYC3–CMID interaction. For example, the Y2H interaction between MYC3 and JAZ10.4, JAZ10.4-N, or the CMID was strongly reduced or disrupted only when we combined L125A (JID α3 helix) with either Y97A (JID β2 strand) or M155A (TAD α4 helix) or combined M155A with the charge-reversal mutation of E148. Interestingly, a chargereversal mutation of the E148-corresponding residue in MYC2, E165K, has been reported to cause JA hypersensitivity in Arabidopsis seedlings (33). Although expression of the L125A/ M155A mutant protein is reduced, the other mutant proteins are expressed at levels relative similar to or higher than wild type (Fig. S3B).
Disruption of the MYC–CMID Interactions Requires Mutations in at Least Two of the Modular Interfaces. To analyze the contribution
of the individual interfaces to the strength of the MYC3–CMID interaction, we mutated key residues in each interface (Table S2). First, we generated and expressed a set of single amino acid mutant forms of LexA–JAZ10.4 that collectively cover the CMID key interface residues identified in the structure and determined their interaction with the MYC3−activation domain (AD) fusion protein by Y2H assays. Similarly, we generated single amino acid mutations in the MYC3−AD protein. In contrast to the interaction between MYC3 and the Jas motif of JAZ9 (10), none of these single mutations could disrupt the MYC3–CMID interaction (Figs. 3A and 4A), suggesting that the MYC3–CMID interaction is more complex and likely requires the disruption of more than one interface. We therefore 1722 | www.pnas.org/cgi/doi/10.1073/pnas.1616938114
Fig. 3. Y2H analysis of the interaction between JAZ10.4 mutant proteins and wild-type MYC3. The experiment was repeated three times with same results. Blue yeast colonies indicate a positive interaction between two proteins. (A) Interactions between JAZ10.4 mutants with a single amino acid replacement and wild-type MYC3 proteins. (B) Interactions between JAZ10.4 mutants with double or triple amino acids replacements and wild-type MYC3 proteins. The same control colonies, LexA DB–JAZ10.4 WT(aa1-167) + WT MYC3−AD or vector, were used for A and B.
Zhang et al.
To identify more quantitative effects of mutations of hydrophilic residues in the MYC3 binding groove, we purified recombinant biotinylated MYC3(44–238) fusion protein and determined its binding to His6-tagged CMIDJAZ10 fusion protein using an AlphaScreen luminescence proximity assay. As shown in Fig. 4B, all mutations in the Jas/CMID-binding groove of MYC3 reduced the strength of the interaction; in particular, the Y97A/T156A and E148R/M155A double mutations strongly reduced (Y97A/T156A) or disrupted (E148R/M155A) the MYC3–CMIDJAZ10 interaction. Consistent with the interface being more extensive in the MYC3–CMID structure than in the MYC3–Jas structure, the CMIDJAZ10 was at least fourfold more efficient (IC50 = 58 nM) than JasJAZ10 (IC50 = 237 nM) in competing with the interaction of MYC3 with the ACID domain of MED25 (Fig. 2).
Discussion In this paper, we present the crystal structure of CMIDJAZ10 in complex with the MYC3 N-terminal domain and the biochemical characterizations we conducted to verify interface residues that are critical for the formation of the MYC3–CMIDJAZ10 complex. We also determined the crystal structure of the MYC3–JasJAZ10 complex. Direct comparison of these structures shows that, in contrast to the single Jas helix reported previously (11), the CMIDJAZ10 forms a more complex loop–helix–loop–helix structure that makes extensive, modular contacts with MYC3. The α1 PLANT BIOLOGY
Fig. 4. Y2H and AlphaScreen interaction analyses of the interaction between MYC3 mutant proteins and JAZ10.4 proteins. (A) Y2H assay. Blue yeast colonies indicate a positive interaction between two proteins. Light blue indicates a reduced interaction. The experiment was repeated three times with same results. (B) AlphaScreen interactions between CMIDJAZ10 and MYC3. AlphaScreen interaction assays between purified recombinant maltose binding protein (MBP)–CMIDJAZ10–His6 wild-type (10 nM) and biotinylated MYC3(44–238) mutant proteins (10 nM). n = 3; error bars indicate SD. SDS PAGE of biotin–MYC3(44–238) mutant proteins is shown below the graph.
CMIDJAZ10 structures; Fig. 1 B–D) but also for JAZ1. Because both CMID and Jas can bind MYC3, we further tested the ability of JasJAZ10 to compete with the MYC3–CMIDJAZ10 interaction and of CMIDJAZ10 to compete with the MYC3–JasJAZ10 interaction. As shown in Fig. 5, >30-fold more JasJAZ10 is required to compete with the MYC3–CMIDJAZ10 interaction than CMIDJAZ10mediated competition of the MYC3–JasJAZ10 interaction, suggesting that the CMID is the predominant MYC- interaction domain in fulllength JAZ10. Among the MYC3-interacting residues within the CMID of JAZ10, the best conserved ones are located in α1 and at the α1 border (S35, F36, I39, and G41) (Fig. S6). These residues also are conserved in the N termini of JAZ1, JAZ2, JAZ5, and JAZ6 (Fig. S4), raising the intriguing possibility that a functional CMID exists in JAZ2, JAZ5, and JAZ6 in addition to JAZ1. To test this hypothesis, we purified the corresponding regions as recombinant fusion proteins and tested their interaction with the MYC3 N-terminal domain by AlphaScreen assay. As shown in Fig. 6A, the fragment from JAZ2 interacts weakly (130,000 photon counts) with MYC3, indicating that at least JAZ5 and JAZ6 contain functional CMIDs. Quantitative homologous competition assays under conditions in which IC50 values approximate Kd values (SI Materials and Methods) indicate that CMIDJAZ5 and CMIDJAZ6 interact with MYC3 with submicromolar affinities (Fig. 6B).
Conservation of MYC3-Interacting Residues in the N-Terminal Half of CMIDJAZ10 in a Subset of JAZ Proteins. Like JAZ10, JAZ1 has an
N-terminal CMID (within amino acids 1–66) in addition to the C-terminal Jas motif (24). However, in contrast to the highly conserved Jas motif, the CMIDJAZ1 has only limited similarity to the CMIDJAZ10 (Fig. S4), and the residues important for the interaction of CMIDJAZ1 with MYC remain undefined. We mutated residues in CMIDJAZ1 that are conserved in CMIDJAZ10 (Fig. S5A) and residues in MYC3 that are important for interaction with CMIDJAZ10 to determine whether they play a role in the MYC–JAZ1 interaction (Fig. S5B). We found that the double mutations E148R/M155A and L125A/M155A compromised binding to full-length JAZ10 and JAZ1 proteins, which can interact with MYC through both their CMID and Jas domains. These results suggest that the corresponding amino acids play important roles in both MYC3–CMID and MYC3–Jas interactions, not only for JAZ10 (as observed in the JasJAZ10 and Zhang et al.
Fig. 5. Heterologous AlphaScreen competition assays in which untagged CMIDJAZ10 competes with the interaction between biotinylated–MYC3(44– 238) (50 nM) and MBP–JasJAZ10–His6 (50 nM) (orange curve and cartoon on the left), and untagged JasJAZ10 competes with the interaction between biotinylated MYC3(44–238) (50 nM) and MBP–CMIDJAZ10–His6 (50 nM) (blue curve and cartoon on the right). n = 3; error bars indicate SD.
PNAS | February 14, 2017 | vol. 114 | no. 7 | 1723
Fig. 6. The N termini of JAZ2, -5, and -6 interact directly with MYC3. (A) AlphaScreen interaction assays between purified recombinant MBP–NterminalJAZ–His6 (50 nM) and biotinylated MYC3(44–238) (50 nM). n = 3; error bars indicate SD. (B) AlphaScreen homologous competition of the interactions between biotinylated MYC3(44–238) (50 nM) and MBP–CMIDJAZ5–His6 (50 nM), and MYC3(44–238) (50 nM) and MBP–CMIDJAZ6–His6 (50 nM). The competitor was untagged MYC3(44–238) protein. n = 3, error bars indicate SD.
helix of the CMID, whose amino acid sequence is unrelated to that of Jas, mimics the Jas helix bound in the MYC Jas-binding groove and blocks the access of MED25 (and possibly other coactivators) to the TAD in MYC3; the loops and α2 helix of the CMID wrap around the JID helix from two opposite sites. This multipronged interaction firmly anchors α1 and the α1–α2 loop in the binding groove and masks the TAD; we have validated this masking biochemically by showing the ability of the CMID to compete with the interaction between MYC3 and the MED25 coactivator. The MYC3–CMIDJAZ10 interaction is highly resistant against single amino acid replacement mutations, providing support that the different interfaces observed in the structure are modular, i.e., they function independently. Indeed, only double mutations that disrupt at least two of the CMID interfaces were able to abolish the MYC–CMIDJAZ10 interaction. This redundancy may be biologically important to 1724 | www.pnas.org/cgi/doi/10.1073/pnas.1616938114
make the MYC–CMIDJAZ10 interaction resilient against random single mutations, because stable desensitization of JA signaling is likely important for avoiding uncontrolled pathway activation that has negative consequences on plant growth (34). A previous study showed that the N terminus of CMIDJAZ10 contains a 12-aa region with similarity to Jas JAZ10 (K23PKFQKFLDRRR34) (25). Although single or double mutations of the three arginine residues at the end of this motif did not affect the Y2H interaction between JAZ10.4 and MYC2, a triple R→A mutation abolished MYC2 binding (25). Surprisingly, in the MYC3–CMIDJAZ10 complex structure, this sequence sticks out of the binding groove, does not engage in side-chain interactions with MYC3, and is only moderately conserved in JAZ10 orthologs (Fig. S6). Instead, we show that the 23-aa sequence following the triple R constitutes the core CMID and makes all direct contacts with MYC3. Importantly, this core CMID is conserved in JAZ10 orthologous proteins across diverse plant species (Fig. S6), thus supporting its importance beyond Arabidopsis and suggesting that JA signal desensitization through JAZ10 splice variants may be a general mechanism. We speculate that mutating the three highly polar arginine residues at the core CMID border to hydrophobic alanines may interfere with the MYC–CMID interaction indirectly by allowing the N-terminal N-loop to fold back onto the core CMID. The structural insights gained in this study may have broader implications beyond CMIDJAZ10. We have shown that JAZ5 and JAZ6, and possibly JAZ2, have functional CMIDs at similar positions to those of JAZ10 and JAZ1 that share similarity with the TAD-binding α1 helix of CMIDJAZ10. Conversely, several MYC3 binding-groove residues that bind the α1 helix of CMIDJAZ10 (Fig. 1B) as well as the JasJAZ10 (Fig. 1C) and JasJAZ1 (11) helices are also important for CMIDJAZ1 binding. We therefore think it is likely that the CMIDs of JAZ1, JAZ5, and JAZ6 have equivalent helices that occupy the same binding groove in MYC and inhibit MYC activity by TAD masking, likely as part of a conserved mechanism to desensitize jasmonate signaling. It has been shown that JAZ3ΔJas lacks the CMID and is unable to bind MYCs but is still able to repress JA signaling. Two hypotheses have been proposed in the literature (35, 36): (i) JAZ3ΔJas may interact with and “poison” (i.e., inactivate) COI1 in a hormoneindependent fashion or (ii) JAZ3ΔJas may form heterodimers with MYC-interacting JAZs that are not efficiently degraded by COI1. Although neither of these models invokes alternative splicing, it is conceivable that heterodimerization of JAZ3ΔJas with the CMIDcontaining JAZs described in our study could contribute to the mechanism by which JAZ3ΔJas represses JA responses. In addition, because the CMIDs coexist with the Jas motif in some full-length JAZ proteins, our results raise the possibility that a subset of JAZ proteins may be able to form ternary MYC–JAZ–COI1 complexes in which the CMID mediates the interaction between JAZ and MYC, and Jas mediates the interaction between JAZ and COI1. This possibility could be tested in the future. Elucidation of the structure of the MYC–CMIDJAZ10 complex advances our knowledge of an essential but poorly understood step in JA hormone signaling—alternative splicing-mediated transcriptional desensitization (Fig. S7). Together with the previously reported crystal structures of COI1–JAZ (the coreceptor complex) and MYC–JAZ (the transcriptional repressor complex) (11, 21), the MYC3–CMIDJAZ10 (transcriptional desensitization complex) structure reported here closes a major gap in our understanding of the transcriptional repression, hormone-dependent activation, and splice variant-dependent transcriptional desensitization steps of a major important plant hormone signaling pathway in plants. Materials and Methods All proteins were expressed in and purified from Escherichia coli BL21 (DE3). MYC3(44–242)–CMIDJAZ10 was crystallized in 0.1 M Tris (pH 6.0), 20% (wt/vol)
Zhang et al.
ACKNOWLEDGMENTS. We thank Michelle Martin for administrative support, Jian Yao for reading the manuscript and critical comments, and staff members of the Life Science Collaborative Access Team of the Advanced Photon Source (APS) for assistance in data collection at the beam lines of sector 21, which is funded in part by the Michigan Economic Development Corporation and Michigan Technology Tri-Corridor Grant 085P1000817. Use of the APS was supported by the Office of Science of the US Department of
Energy (DOE) under Contract DE-AC02-06CH11357. This research was supported by China Scholarship Council Grant 201206850026 (to F.Z.); Gordon and Betty Moore Foundation Grant GBMF3037 (to S.Y.H.); NIH Grants R01AI060761 (to S.Y.H.), R01GM102545 and GM104212 (to K.M.), and DK071662 (to H.E.X.); DOE Chemical Sciences, Geosciences, and Biosciences Division, Office of Basic Energy Sciences, Office of Science Grant DE-FG0291ER20021 (to S.Y.H. and G.A.H.) (for infrastructural support); the Van Andel Research Institute (H.E.X. and K.M.); National Natural Science Foundation of China Grant 31300245 (to H.E.X.); Chinese Ministry of Science and Technology Grants 2012ZX09301001, 2012CB910403, 2013CB910600, XDB08020303, and 2013ZX09507001 (to H.E.X.); Amway-China (H.E.X.); The Japan Society for Promotion of Science Research Fellowship for Young Scientists 24-824 (to K.S.); Michigan AgBioResearch Grant MICL02278 (to G.A.H.); and the Discretionary Funding Initiative from Michigan State University (G.A.H.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
1. Campos ML, Kang J-H, Howe GA (2014) Jasmonate-triggered plant immunity. J Chem Ecol 40(7):657–675. 2. Chini A, et al. (2007) The JAZ family of repressors is the missing link in jasmonate signalling. Nature 448(7154):666–671. 3. Farmer EE, Alméras E, Krishnamurthy V (2003) Jasmonates and related oxylipins in plant responses to pathogenesis and herbivory. Curr Opin Plant Biol 6(4):372–378. 4. Thines B, et al. (2007) JAZ repressor proteins are targets of the SCF(COI1) complex during jasmonate signalling. Nature 448(7154):661–665. 5. Fonseca S, et al. (2009) (+)-7-iso-Jasmonoyl-L-isoleucine is the endogenous bioactive jasmonate. Nat Chem Biol 5(5):344–350. 6. Staswick PE, Tiryaki I (2004) The oxylipin signal jasmonic acid is activated by an enzyme that conjugates it to isoleucine in Arabidopsis. Plant Cell 16(8):2117–2127. 7. Wasternack C, Kombrink E (2010) Jasmonates: Structural requirements for lipidderived signals active in plant stress responses and development. ACS Chem Biol 5(1): 63–77. 8. Yan Y, et al. (2007) A downstream mediator in the growth repression limb of the jasmonate pathway. Plant Cell 19(8):2470–2483. 9. Melotto M, et al. (2008) A critical role of two positively charged amino acids in the Jas motif of Arabidopsis JAZ proteins in mediating coronatine- and jasmonoyl isoleucinedependent interactions with the COI1 F-box protein. Plant J 55(6):979–988. 10. Thireault C, et al. (2015) Repression of jasmonate signaling by a non-TIFY JAZ protein in Arabidopsis. Plant J 82(4):669–679. 11. Zhang F, et al. (2015) Structural basis of JAZ repression of MYC transcription factors in jasmonate signalling. Nature 525(7568):269–273. 12. Acosta IF, et al. (2013) Role of NINJA in root jasmonate signaling. Proc Natl Acad Sci USA 110(38):15473–15478. 13. Kagale S, Links MG, Rozwadowski K (2010) Genome-wide analysis of ethyleneresponsive element binding factor-associated amphiphilic repression motif-containing transcriptional regulators in Arabidopsis. Plant Physiol 152(3):1109–1134. 14. Kazan K (2006) Negative regulation of defence and stress genes by EAR-motif-containing repressors. Trends Plant Sci 11(3):109–112. 15. Pauwels L, et al. (2010) NINJA connects the co-repressor TOPLESS to jasmonate signalling. Nature 464(7289):788–791. 16. Ke J, et al. (2015) Structural basis for recognition of diverse transcriptional repressors by the TOPLESS family of corepressors. Sci Adv 1(6):e1500107. 17. Long JA, Ohno C, Smith ZR, Meyerowitz EM (2006) TOPLESS regulates apical embryonic fate in Arabidopsis. Science 312(5779):1520–1523. 18. Zhu Z, et al. (2011) Derepression of ethylene-stabilized transcription factors (EIN3/ EIL1) mediates jasmonate and ethylene signaling synergy in Arabidopsis. Proc Natl Acad Sci USA 108(30):12539–12544. 19. Katsir L, Schilmiller AL, Staswick PE, He SY, Howe GA (2008) COI1 is a critical component of a receptor for jasmonate and the bacterial virulence factor coronatine. Proc Natl Acad Sci USA 105(19):7100–7105. 20. Xie DX, Feys BF, James S, Nieto-Rostro M, Turner JG (1998) COI1: An Arabidopsis gene required for jasmonate-regulated defense and fertility. Science 280(5366):1091–1094. 21. Sheard LB, et al. (2010) Jasmonate perception by inositol-phosphate-potentiated COI1-JAZ co-receptor. Nature 468(7322):400–405. 22. Yan J, et al. (2009) The Arabidopsis CORONATINE INSENSITIVE1 protein is a jasmonate receptor. Plant Cell 21(8):2220–2236.
23. Chung HS, et al. (2010) Alternative splicing expands the repertoire of dominant JAZ repressors of jasmonate signaling. Plant J 63(4):613–622. 24. Goossens J, Swinnen G, Vanden Bossche R, Pauwels L, Goossens A (2015) Change of a conserved amino acid in the MYC2 and MYC3 transcription factors leads to release of JAZ repression and increased activity. New Phytol 206(4):1229–1237. 25. Moreno JE, et al. (2013) Negative feedback control of jasmonate signaling by an alternative splice variant of JAZ10. Plant Physiol 162(2):1006–1017. 26. Chung HS, Howe GA (2009) A critical role for the TIFY motif in repression of jasmonate signaling by a stabilized splice variant of the JASMONATE ZIM-domain protein JAZ10 in Arabidopsis. Plant Cell 21(1):131–145. 27. Chini A, Gimenez-Ibanez S, Goossens A, Solano R (2016) Redundancy and specificity in jasmonate signalling. Curr Opin Plant Biol 33:147–156. 28. Chung HS, et al. (2008) Regulation and function of Arabidopsis JASMONATE ZIMdomain genes in response to wounding and herbivory. Plant Physiol 146(3):952–964. 29. Demianski AJ, Chung KM, Kunkel BN (2012) Analysis of Arabidopsis JAZ gene expression during Pseudomonas syringae pathogenesis. Mol Plant Pathol 13(1):46–57. 30. Çevik V, et al. (2012) MEDIATOR25 acts as an integrative hub for the regulation of jasmonate-responsive gene expression in Arabidopsis. Plant Physiol 160(1):541–555. 31. Fernández-Calvo P, et al. (2011) The Arabidopsis bHLH transcription factors MYC3 and MYC4 are targets of JAZ repressors and act additively with MYC2 in the activation of jasmonate responses. Plant Cell 23(2):701–715. 32. Chen R, et al. (2012) The Arabidopsis mediator subunit MED25 differentially regulates jasmonate and abscisic acid signaling through interacting with the MYC2 and ABI5 transcription factors. Plant Cell 24(7):2898–2916. 33. Gasperini D, et al. (2015) Multilayered organization of jasmonate signalling in the regulation of root growth. PLoS Genet 11(6):e1005300. 34. Yang DL, et al. (2012) Plant hormone jasmonate prioritizes defense over growth by interfering with gibberellin signaling cascade. Proc Natl Acad Sci USA 109(19): E1192–E1200. 35. Browse J (2009) Jasmonate passes muster: A receptor and targets for the defense hormone. Annu Rev Plant Biol 60:183–205. 36. Chini A, Fonseca S, Chico JM, Fernández-Calvo P, Solano R (2009) The ZIM domain mediates homo- and heteromeric interactions between Arabidopsis JAZ proteins. Plant J 59(1):77–87. 37. Fairhead M, Howarth M (2015) Site-specific biotinylation of purified proteins using BirA. Methods Mol Biol 1266:171–184. 38. Kabsch W (2010) Xds. Acta Crystallogr D Biol Crystallogr 66(Pt 2):125–132. 39. Collaborative Computational Project, Number 4 (1994) The CCP4 suite: Programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50(Pt 5):760–763. 40. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Cryst 40(Pt 4):658–674. 41. Emsley P, Cowtan K (2004) Coot: Model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60(Pt 12 Pt 1):2126–2132. 42. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53(Pt 3): 240–255. 43. Withers J, et al. (2012) Transcription factor-dependent nuclear localization of a transcriptional repressor in jasmonate hormone signaling. Proc Natl Acad Sci USA 109(49):20148–20153.
Zhang et al.
PNAS | February 14, 2017 | vol. 114 | no. 7 | 1725
PLANT BIOLOGY
polyethylene glycol monomethyl ether 2000, and MYC3(44–242)–JasJAZ10 was crystallized in 20% (wt/vol) polyethylene glycol 3350 and 0.2 M magnesium formate dihydrate (pH 7.0). The MYC3(44–242)–CMIDJAZ10 and MYC3(44–242)–JasJAZ10 structures were determined by molecular replacement using the MYC3(44–238) structure (PDB ID code: 4RQW) as a search model. All biochemical experiments were performed three or more times with similar results. Detailed experimental protocols and information can be found in SI Materials and Methods.
Contributed by Sheng Yang He, December 28, 2016 (sent for review October 14, 2016; reviewed by Roberto Solano and Ning Zheng)
Jasmonate ZIM-domain (JAZ) transcriptional repressors play a key role in regulating jasmonate (JA) signaling in plants. Below a threshold concentration of jasmonoyl isoleucine (JA-Ile), the active form of JA, the C-terminal Jas motif of JAZ proteins binds MYC transcription factors to repress JA signaling. With increasing JA-Ile concentration, the Jas motif binds to JA-Ile and the COI1 subunit of the SCFCOI1 E3 ligase, which mediates ubiquitination and proteasomal degradation of JAZ repressors, resulting in derepression of MYC transcription factors. JA signaling subsequently becomes desensitized, in part by feedback induction of JAZ splice variants that lack the C-terminal Jas motif but include an N-terminal cryptic MYC-interaction domain (CMID). The CMID sequence is dissimilar to the Jas motif and is incapable of recruiting SCFCOI1, allowing CMID-containing JAZ splice variants to accumulate in the presence of JA and to re-repress MYC transcription factors as an integral part of reestablishing signal homeostasis. The mechanism by which the CMID represses MYC transcription factors remains elusive. Here we describe the crystal structure of the MYC3–CMIDJAZ10 complex. In contrast to the Jas motif, which forms a single continuous helix when bound to MYC3, the CMID adopts a loop–helix–loop–helix architecture with modular interactions with both the Jas-binding groove and the backside of the Jasinteraction domain of MYC3. This clamp-like interaction allows the CMID to bind MYC3 tightly and block access of MED25 (a subunit of the Mediator coactivator complex) to the MYC3 transcriptional activation domain, shedding light on the enigmatic mechanism by which JAZ splice variants desensitize JA signaling.
|
signaling plant hormone plant insect
binding of TPL to ethylene-response factor-associated amphiphilic repression (EAR) motifs found at the N terminus of a subset of JAZ proteins or indirectly through the EAR motif-containing NINJA adaptor protein that binds to the central ZIM domain of JAZ proteins (Fig. 1A) (12–16). TPL proteins in turn recruit repressive histone deacetylases and chromatin remodelers (17, 18) to inhibit MYC target gene expression epigenetically. In response to stress or developmental cues, plants synthesize JA-Ile, which promotes the formation of a coreceptor complex consisting of JAZ (the Jas motif), JA-Ile, and coronatine insensitive 1 (COI1), an F-box protein that is a subunit of a Skp1–Cul1–F-box protein (SCF) ubiquitin E3 ligase (4, 19, 20). The JAZ–COI1 interaction leads to ubiquitination and proteasome-dependent degradation of JAZ repressors and thereby to the release of MYC proteins from transcriptional repression (21, 22). In addition to the highly conserved C-terminal Jas motif, JAZ10 also contains an N-terminal cryptic MYC-interaction domain (CMID) that can bind to the N terminus of MYC proteins but does not form complexes with JA-Ile and COI1 (23–25). Moreover, alternative splicing of JAZ10 mRNA generates three splice variants that differ in their C termini: JAZ10.1 (full-length JAZ10, amino acids 1–197), JAZ10.3 (amino acids 1–185, lacking the C Significance Jasmonate (JA) is a plant hormone involved in regulating defense response, growth, and development. A mechanistic understanding of JA signaling has great importance in agriculture, especially for enhancing plant resilience against biotic stresses and optimizing defense and growth in crop fields. This study reports the crystal structure of a key protein complex (MYC3– CMIDJAZ10) involved in the desensitization of JA signaling. Desensitization of JA signaling after an effective JA response is necessary to re-establish signal homeostasis and prevent JA signaling from running out of control, which could lead to severe growth and fitness penalties. The MYC3–CMIDJAZ10 desensitization complex structure reported here closes a major gap in our understanding of one of the most important hormone signaling pathways in plants.
| plant defense | plant pathogen |
T
he plant hormone jasmonate (JA) is a critical mediator of plant defense responses and an important regulator of plant growth and development (1–4). The bioactive form jasmonoyl isoleucine (JA-Ile) is produced by conjugation of jasmonic acid, a fatty acid-derived oxylipin, with isoleucine in the plant cytosol (5– 7). Under stress-free conditions, when JA-Ile levels are low, JA responses are restrained by a group of nuclear transcriptional repressors called “jasmonate ZIM-domain” (JAZ) proteins (2, 4, 8). In Arabidopsis, 13 genes (JAZ1–JAZ13) encode JAZ proteins with a highly conserved C-terminal Jas motif (9, 10). JAZ proteins physically bind and inhibit basic helix–loop–helix transcription factors, including MYC2, MYC3, and MYC4 (2, 9), through a dual mechanism: First, a recent study of the crystal structure of the MYC3 transcription factor in complex with JAZ1 and JAZ9 showed that the conserved Jas motif of JAZ proteins directly binds the transcriptional activation domain (TAD) of MYC proteins and thereby blocks access of the TAD to the MED25 subunit of the mediator complex and possibly to other coactivators whose interaction with the TAD is required for MYC target gene transcription (11) (Fig. 1A). Second, JAZ proteins recruit members of the TOPLESS (TPL) class of corepressors either through direct
1720–1725 | PNAS | February 14, 2017 | vol. 114 | no. 7
Author contributions: H.E.X., M.Z., S.Y.H., and K.M. designed research; F.Z., J.K., L.Z., R.C., K.S., and G.A.H. performed research; F.Z., J.K., L.Z., R.C., K.S., G.A.H., H.E.X., M.Z., S.Y.H., and K.M. analyzed data; and F.Z., S.Y.H., and K.M. wrote the paper. Reviewers: R.S., Centro Nacional de Biotecnologia (Consejo Superior de Investigaciones Científicas); and N.Z., University of Washington. The authors declare no conflict of interest. Data deposition: Crystallography, atomic coordinates, and structure factors reported in this paper have been deposited in the Protein Data Bank database (PDB ID codes 5T0Q and 5T0F). 1
To whom correspondence may be addressed. Email: [email protected], [email protected], or [email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1616938114/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1616938114
terminus of the Jas motif), and JAZ10.4, which lacks the entire Jas motif (amino acids 1–147, followed by 20 unrelated amino acids) (Fig. 1A). As a consequence, all three splice forms retain the ability to bind MYC and repress MYC target gene expression, but JAZ10.3 is partially resistant and JAZ10.4 is fully resistant to JA-induced degradation (8, 23, 26). Because JAZ10 expression itself is induced rapidly by JA, and JAZ10.4 protein accumulates following an increase in JA levels, JAZ10.4 has been suggested to be an important feedback regulator to mediate JA signal desensitization (25). This model has been supported further by genetic data. Although there is significant functional redundancy among JAZ proteins in Arabidopsis, jaz10 loss-of-function mutations result in JA hypersensitivity (27). Conversely, although plants overexpressing JAZ10.1 remain fully responsive to JA, overexpression of JAZ10.3 and JAZ10.4 causes moderate and severe JA insensitivity, respectively (8, 28, 29). JAZ1 also contains a functional N-terminal CMID, which has weak similarity to the JAZ10 CMID (24). However, because the CMID has no sequence similarity to the well-characterized Jas motif, how the CMID interacts with the MYC N-terminal domain and blocks MYC transcriptional activity is unknown. Here we report the crystal structure of CMIDJAZ10 in complex with the MYC3 N-terminal domain. For direct comparison, we also determined the crystal structure of the MYC3–JasJAZ10 complex. We found that CMIDJAZ10 and JasJAZ10 adopt strikingly Zhang et al.
different conformations, despite their overlapping binding sites in MYC3, leading in the case of the MYC3–CMIDJAZ10 complex to a modular MYC interaction. The MYC3–CMIDJAZ10 complex structure assisted us in the discovery of functional CMIDs in JAZ5, JAZ6, and possibly JAZ2, revealing a potentially widespread mechanism of JAZ splice variant-mediated desensitization of JA signal transduction. Results The MYC3 N-Terminal Domain and CMIDJAZ10.4 Are Sufficient to Mediate the MYC–JAZ10.4 Interaction. We used yeast two-hybrid
(Y2H) assays to determine the MYC3-binding region within JAZ10.4. A region in the N terminus of MYC3, MYC3(44–238), which is conserved among MYC2, MYC3, and MYC4 (30–32), contains the JAZ-interaction domain (JID)/TAD and is sufficient to interact with JAZ10.4 and JAZ1ΔJas. Conversely, the 43-aa CMIDJAZ10 [JAZ10(16–58)] is sufficient to interact with MYC3 (Fig. S1). We used a fusion protein strategy that we developed previously (11) to express and purify a stable MYC3/CMID complex consisting of MYC3(44–242) N-terminally fused to JAZ10(16–58) through a structurally flexible 12-aa linker (MYC3– CMIDJAZ10). This fusion protein formed high-quality crystals and allowed us to determine the structure of the complex at a resolution of 2.4 Å (Fig. 1B and Table S1). To be able to compare the MYC3–CMIDJAZ10 complex with the MYC3–JasJAZ10 complex, PNAS | February 14, 2017 | vol. 114 | no. 7 | 1721
PLANT BIOLOGY
Fig. 1. Structures of the MYC3(44–242)–CMIDJAZ10 and MYC3(44–242)–JasJAZ10 complexes. (A) Schematic structure of JAZ9, JAZ10.1, and JAZ10.4 proteins. Numbers indicate amino acid residues. JAZ10.4-N: amino acids 1–58; JAZ10.4-C: amino acids 59–167; CMID: amino acids 16–58; Core CMID: amino acids 35–58. (B) Structure of MYC3(44–242) in complex with CMIDJAZ10. (C) Structure of MYC3(44–242) in complex with JasJAZ10. (D) Overlay of the MYC3(44–242)– CMIDJAZ10 complex with the MYC3(44–242)–JasJAZ10 complex. Note that “JID” and “TAD” are only formal designations. Although they were originally mapped as two distinct regions, both regions are part of the same structural fold and are required for JAZ binding. (E) The CMIDJAZ10 forms extensive, modular interactions with the JID and TAD in MYC3. The MYC3(44–242)–CMIDJAZ10 structure with important interacting residues shown in stick presentation, and bonds are shown as dashed lines. CMIDJAZ10 is shown in magenta, JIDMYC3 is shown in cyan, and TADMYC3 is shown in green.
Fig. 2. CMIDJAZ10 and JasJAZ10 peptides competitively inhibit the MYC3– MED25 interaction. AlphaScreen competition assay of the interaction between His6Sumo–MED25(407–680) (50 nM) and biotinylated MYC3(5–242) (50 nM) with increasing concentrations of CMIDJAZ10 or JasJAZ10 peptide. n = 3 replicates; error bars indicate SD.
we also expressed and purified an MYC3(44–242)–14-aa linker– JAZ10(166–192) fusion protein, MYC–JasJAZ10, and determined its structure at a resolution of 2.2 Å (Fig. 1C and Table S1). The CMID Makes a Modular Interaction with MYC3 and Competes with the MYC3–MED25 Interaction. The structure of the MYC3–
JasJAZ10 complex is essentially superimposable with those of the previously determined MYC3–JasJAZ9 and MYC3–JasJAZ1 complexes (Fig. S2) (11), reflecting the high conservation of Jas motifs. In all three structures, the Jas motifs form a single continuous α-helix when bound to MYC3. In contrast, the CMID in the MYC3– CMIDJAZ10 complex consists of two short helices and two loops that make more extensive interactions with MYC3 than with the Jas helix (Fig. 1 B–D). Although residues 26–57 of the CMIDJAZ10 are resolved in the complex structure, only the C-terminal part of the CMID (residues 34–57, the core CMID) makes direct contacts with MYC3. The position of the N-terminal CMID helix (α1) overlaps with the position of the longer Jas helix in the MYC3–JasJAZ10 complex; α1 thus mimics the ability of the Jas motif to block the access of the TAD to transcription coactivators. In agreement with the position of α1, the CMID competes with the interaction between MYC3 and the activator-interacting domain (ACID) of MED25 (Fig. 2, orange competition curve). In addition (and absent in the MYC3–Jas interaction), the C-terminal helix and I46 of the α1–α2 connecting loop of CMIDJAZ10 wrap around the JID helix and make additional, predominantly hydrophobic, contacts with the back side of the binding groove (Fig. 1 B and E and Table S2).
generated JAZ10.4 double mutants in which we combined a mutation in the CMID α2 helix (L54A) with a mutation at the N-loop/α1 border (F36A), the α1 helix (I43A), or the α1–α2 loop (I46A). All three double mutants lost the ability to interact with MYC3 in the Y2H assay (Fig. 3B), even though these mutant proteins were expressed in yeast cells at levels equal to or higher than that of wild-type JAZ10.4 (Fig. S3A, Right), suggesting that at least two of these four structural elements must be mutated to abolish the MYC3–JAZ10.4 interaction. We also combined the L54A mutation with mutations in two hydrophilic residues in the α1 helix, S44 and R37, which form hydrogen bonds with MYC3. However, only mutants that contain the more drastic R37D charge reversal mutation, but neither the L54A/S44A and L54A/ R37A double mutants nor an L54A/S44A/R37A triple mutant, could disrupt the interaction with MYC3 (Fig. 3B), indicating that the interaction between the solvent-exposed MYC JID/ TAD and JAZ CMID is driven predominantly by Van der Waals interactions. Next, we introduced mutations into CMID-interacting residues of MYC3. Similar to the corresponding mutations in JAZ10.4, specific sets of double mutations in MYC3 were required to abolish or reduce the MYC3–CMID interaction. For example, the Y2H interaction between MYC3 and JAZ10.4, JAZ10.4-N, or the CMID was strongly reduced or disrupted only when we combined L125A (JID α3 helix) with either Y97A (JID β2 strand) or M155A (TAD α4 helix) or combined M155A with the charge-reversal mutation of E148. Interestingly, a chargereversal mutation of the E148-corresponding residue in MYC2, E165K, has been reported to cause JA hypersensitivity in Arabidopsis seedlings (33). Although expression of the L125A/ M155A mutant protein is reduced, the other mutant proteins are expressed at levels relative similar to or higher than wild type (Fig. S3B).
Disruption of the MYC–CMID Interactions Requires Mutations in at Least Two of the Modular Interfaces. To analyze the contribution
of the individual interfaces to the strength of the MYC3–CMID interaction, we mutated key residues in each interface (Table S2). First, we generated and expressed a set of single amino acid mutant forms of LexA–JAZ10.4 that collectively cover the CMID key interface residues identified in the structure and determined their interaction with the MYC3−activation domain (AD) fusion protein by Y2H assays. Similarly, we generated single amino acid mutations in the MYC3−AD protein. In contrast to the interaction between MYC3 and the Jas motif of JAZ9 (10), none of these single mutations could disrupt the MYC3–CMID interaction (Figs. 3A and 4A), suggesting that the MYC3–CMID interaction is more complex and likely requires the disruption of more than one interface. We therefore 1722 | www.pnas.org/cgi/doi/10.1073/pnas.1616938114
Fig. 3. Y2H analysis of the interaction between JAZ10.4 mutant proteins and wild-type MYC3. The experiment was repeated three times with same results. Blue yeast colonies indicate a positive interaction between two proteins. (A) Interactions between JAZ10.4 mutants with a single amino acid replacement and wild-type MYC3 proteins. (B) Interactions between JAZ10.4 mutants with double or triple amino acids replacements and wild-type MYC3 proteins. The same control colonies, LexA DB–JAZ10.4 WT(aa1-167) + WT MYC3−AD or vector, were used for A and B.
Zhang et al.
To identify more quantitative effects of mutations of hydrophilic residues in the MYC3 binding groove, we purified recombinant biotinylated MYC3(44–238) fusion protein and determined its binding to His6-tagged CMIDJAZ10 fusion protein using an AlphaScreen luminescence proximity assay. As shown in Fig. 4B, all mutations in the Jas/CMID-binding groove of MYC3 reduced the strength of the interaction; in particular, the Y97A/T156A and E148R/M155A double mutations strongly reduced (Y97A/T156A) or disrupted (E148R/M155A) the MYC3–CMIDJAZ10 interaction. Consistent with the interface being more extensive in the MYC3–CMID structure than in the MYC3–Jas structure, the CMIDJAZ10 was at least fourfold more efficient (IC50 = 58 nM) than JasJAZ10 (IC50 = 237 nM) in competing with the interaction of MYC3 with the ACID domain of MED25 (Fig. 2).
Discussion In this paper, we present the crystal structure of CMIDJAZ10 in complex with the MYC3 N-terminal domain and the biochemical characterizations we conducted to verify interface residues that are critical for the formation of the MYC3–CMIDJAZ10 complex. We also determined the crystal structure of the MYC3–JasJAZ10 complex. Direct comparison of these structures shows that, in contrast to the single Jas helix reported previously (11), the CMIDJAZ10 forms a more complex loop–helix–loop–helix structure that makes extensive, modular contacts with MYC3. The α1 PLANT BIOLOGY
Fig. 4. Y2H and AlphaScreen interaction analyses of the interaction between MYC3 mutant proteins and JAZ10.4 proteins. (A) Y2H assay. Blue yeast colonies indicate a positive interaction between two proteins. Light blue indicates a reduced interaction. The experiment was repeated three times with same results. (B) AlphaScreen interactions between CMIDJAZ10 and MYC3. AlphaScreen interaction assays between purified recombinant maltose binding protein (MBP)–CMIDJAZ10–His6 wild-type (10 nM) and biotinylated MYC3(44–238) mutant proteins (10 nM). n = 3; error bars indicate SD. SDS PAGE of biotin–MYC3(44–238) mutant proteins is shown below the graph.
CMIDJAZ10 structures; Fig. 1 B–D) but also for JAZ1. Because both CMID and Jas can bind MYC3, we further tested the ability of JasJAZ10 to compete with the MYC3–CMIDJAZ10 interaction and of CMIDJAZ10 to compete with the MYC3–JasJAZ10 interaction. As shown in Fig. 5, >30-fold more JasJAZ10 is required to compete with the MYC3–CMIDJAZ10 interaction than CMIDJAZ10mediated competition of the MYC3–JasJAZ10 interaction, suggesting that the CMID is the predominant MYC- interaction domain in fulllength JAZ10. Among the MYC3-interacting residues within the CMID of JAZ10, the best conserved ones are located in α1 and at the α1 border (S35, F36, I39, and G41) (Fig. S6). These residues also are conserved in the N termini of JAZ1, JAZ2, JAZ5, and JAZ6 (Fig. S4), raising the intriguing possibility that a functional CMID exists in JAZ2, JAZ5, and JAZ6 in addition to JAZ1. To test this hypothesis, we purified the corresponding regions as recombinant fusion proteins and tested their interaction with the MYC3 N-terminal domain by AlphaScreen assay. As shown in Fig. 6A, the fragment from JAZ2 interacts weakly (130,000 photon counts) with MYC3, indicating that at least JAZ5 and JAZ6 contain functional CMIDs. Quantitative homologous competition assays under conditions in which IC50 values approximate Kd values (SI Materials and Methods) indicate that CMIDJAZ5 and CMIDJAZ6 interact with MYC3 with submicromolar affinities (Fig. 6B).
Conservation of MYC3-Interacting Residues in the N-Terminal Half of CMIDJAZ10 in a Subset of JAZ Proteins. Like JAZ10, JAZ1 has an
N-terminal CMID (within amino acids 1–66) in addition to the C-terminal Jas motif (24). However, in contrast to the highly conserved Jas motif, the CMIDJAZ1 has only limited similarity to the CMIDJAZ10 (Fig. S4), and the residues important for the interaction of CMIDJAZ1 with MYC remain undefined. We mutated residues in CMIDJAZ1 that are conserved in CMIDJAZ10 (Fig. S5A) and residues in MYC3 that are important for interaction with CMIDJAZ10 to determine whether they play a role in the MYC–JAZ1 interaction (Fig. S5B). We found that the double mutations E148R/M155A and L125A/M155A compromised binding to full-length JAZ10 and JAZ1 proteins, which can interact with MYC through both their CMID and Jas domains. These results suggest that the corresponding amino acids play important roles in both MYC3–CMID and MYC3–Jas interactions, not only for JAZ10 (as observed in the JasJAZ10 and Zhang et al.
Fig. 5. Heterologous AlphaScreen competition assays in which untagged CMIDJAZ10 competes with the interaction between biotinylated–MYC3(44– 238) (50 nM) and MBP–JasJAZ10–His6 (50 nM) (orange curve and cartoon on the left), and untagged JasJAZ10 competes with the interaction between biotinylated MYC3(44–238) (50 nM) and MBP–CMIDJAZ10–His6 (50 nM) (blue curve and cartoon on the right). n = 3; error bars indicate SD.
PNAS | February 14, 2017 | vol. 114 | no. 7 | 1723
Fig. 6. The N termini of JAZ2, -5, and -6 interact directly with MYC3. (A) AlphaScreen interaction assays between purified recombinant MBP–NterminalJAZ–His6 (50 nM) and biotinylated MYC3(44–238) (50 nM). n = 3; error bars indicate SD. (B) AlphaScreen homologous competition of the interactions between biotinylated MYC3(44–238) (50 nM) and MBP–CMIDJAZ5–His6 (50 nM), and MYC3(44–238) (50 nM) and MBP–CMIDJAZ6–His6 (50 nM). The competitor was untagged MYC3(44–238) protein. n = 3, error bars indicate SD.
helix of the CMID, whose amino acid sequence is unrelated to that of Jas, mimics the Jas helix bound in the MYC Jas-binding groove and blocks the access of MED25 (and possibly other coactivators) to the TAD in MYC3; the loops and α2 helix of the CMID wrap around the JID helix from two opposite sites. This multipronged interaction firmly anchors α1 and the α1–α2 loop in the binding groove and masks the TAD; we have validated this masking biochemically by showing the ability of the CMID to compete with the interaction between MYC3 and the MED25 coactivator. The MYC3–CMIDJAZ10 interaction is highly resistant against single amino acid replacement mutations, providing support that the different interfaces observed in the structure are modular, i.e., they function independently. Indeed, only double mutations that disrupt at least two of the CMID interfaces were able to abolish the MYC–CMIDJAZ10 interaction. This redundancy may be biologically important to 1724 | www.pnas.org/cgi/doi/10.1073/pnas.1616938114
make the MYC–CMIDJAZ10 interaction resilient against random single mutations, because stable desensitization of JA signaling is likely important for avoiding uncontrolled pathway activation that has negative consequences on plant growth (34). A previous study showed that the N terminus of CMIDJAZ10 contains a 12-aa region with similarity to Jas JAZ10 (K23PKFQKFLDRRR34) (25). Although single or double mutations of the three arginine residues at the end of this motif did not affect the Y2H interaction between JAZ10.4 and MYC2, a triple R→A mutation abolished MYC2 binding (25). Surprisingly, in the MYC3–CMIDJAZ10 complex structure, this sequence sticks out of the binding groove, does not engage in side-chain interactions with MYC3, and is only moderately conserved in JAZ10 orthologs (Fig. S6). Instead, we show that the 23-aa sequence following the triple R constitutes the core CMID and makes all direct contacts with MYC3. Importantly, this core CMID is conserved in JAZ10 orthologous proteins across diverse plant species (Fig. S6), thus supporting its importance beyond Arabidopsis and suggesting that JA signal desensitization through JAZ10 splice variants may be a general mechanism. We speculate that mutating the three highly polar arginine residues at the core CMID border to hydrophobic alanines may interfere with the MYC–CMID interaction indirectly by allowing the N-terminal N-loop to fold back onto the core CMID. The structural insights gained in this study may have broader implications beyond CMIDJAZ10. We have shown that JAZ5 and JAZ6, and possibly JAZ2, have functional CMIDs at similar positions to those of JAZ10 and JAZ1 that share similarity with the TAD-binding α1 helix of CMIDJAZ10. Conversely, several MYC3 binding-groove residues that bind the α1 helix of CMIDJAZ10 (Fig. 1B) as well as the JasJAZ10 (Fig. 1C) and JasJAZ1 (11) helices are also important for CMIDJAZ1 binding. We therefore think it is likely that the CMIDs of JAZ1, JAZ5, and JAZ6 have equivalent helices that occupy the same binding groove in MYC and inhibit MYC activity by TAD masking, likely as part of a conserved mechanism to desensitize jasmonate signaling. It has been shown that JAZ3ΔJas lacks the CMID and is unable to bind MYCs but is still able to repress JA signaling. Two hypotheses have been proposed in the literature (35, 36): (i) JAZ3ΔJas may interact with and “poison” (i.e., inactivate) COI1 in a hormoneindependent fashion or (ii) JAZ3ΔJas may form heterodimers with MYC-interacting JAZs that are not efficiently degraded by COI1. Although neither of these models invokes alternative splicing, it is conceivable that heterodimerization of JAZ3ΔJas with the CMIDcontaining JAZs described in our study could contribute to the mechanism by which JAZ3ΔJas represses JA responses. In addition, because the CMIDs coexist with the Jas motif in some full-length JAZ proteins, our results raise the possibility that a subset of JAZ proteins may be able to form ternary MYC–JAZ–COI1 complexes in which the CMID mediates the interaction between JAZ and MYC, and Jas mediates the interaction between JAZ and COI1. This possibility could be tested in the future. Elucidation of the structure of the MYC–CMIDJAZ10 complex advances our knowledge of an essential but poorly understood step in JA hormone signaling—alternative splicing-mediated transcriptional desensitization (Fig. S7). Together with the previously reported crystal structures of COI1–JAZ (the coreceptor complex) and MYC–JAZ (the transcriptional repressor complex) (11, 21), the MYC3–CMIDJAZ10 (transcriptional desensitization complex) structure reported here closes a major gap in our understanding of the transcriptional repression, hormone-dependent activation, and splice variant-dependent transcriptional desensitization steps of a major important plant hormone signaling pathway in plants. Materials and Methods All proteins were expressed in and purified from Escherichia coli BL21 (DE3). MYC3(44–242)–CMIDJAZ10 was crystallized in 0.1 M Tris (pH 6.0), 20% (wt/vol)
Zhang et al.
ACKNOWLEDGMENTS. We thank Michelle Martin for administrative support, Jian Yao for reading the manuscript and critical comments, and staff members of the Life Science Collaborative Access Team of the Advanced Photon Source (APS) for assistance in data collection at the beam lines of sector 21, which is funded in part by the Michigan Economic Development Corporation and Michigan Technology Tri-Corridor Grant 085P1000817. Use of the APS was supported by the Office of Science of the US Department of
Energy (DOE) under Contract DE-AC02-06CH11357. This research was supported by China Scholarship Council Grant 201206850026 (to F.Z.); Gordon and Betty Moore Foundation Grant GBMF3037 (to S.Y.H.); NIH Grants R01AI060761 (to S.Y.H.), R01GM102545 and GM104212 (to K.M.), and DK071662 (to H.E.X.); DOE Chemical Sciences, Geosciences, and Biosciences Division, Office of Basic Energy Sciences, Office of Science Grant DE-FG0291ER20021 (to S.Y.H. and G.A.H.) (for infrastructural support); the Van Andel Research Institute (H.E.X. and K.M.); National Natural Science Foundation of China Grant 31300245 (to H.E.X.); Chinese Ministry of Science and Technology Grants 2012ZX09301001, 2012CB910403, 2013CB910600, XDB08020303, and 2013ZX09507001 (to H.E.X.); Amway-China (H.E.X.); The Japan Society for Promotion of Science Research Fellowship for Young Scientists 24-824 (to K.S.); Michigan AgBioResearch Grant MICL02278 (to G.A.H.); and the Discretionary Funding Initiative from Michigan State University (G.A.H.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
1. Campos ML, Kang J-H, Howe GA (2014) Jasmonate-triggered plant immunity. J Chem Ecol 40(7):657–675. 2. Chini A, et al. (2007) The JAZ family of repressors is the missing link in jasmonate signalling. Nature 448(7154):666–671. 3. Farmer EE, Alméras E, Krishnamurthy V (2003) Jasmonates and related oxylipins in plant responses to pathogenesis and herbivory. Curr Opin Plant Biol 6(4):372–378. 4. Thines B, et al. (2007) JAZ repressor proteins are targets of the SCF(COI1) complex during jasmonate signalling. Nature 448(7154):661–665. 5. Fonseca S, et al. (2009) (+)-7-iso-Jasmonoyl-L-isoleucine is the endogenous bioactive jasmonate. Nat Chem Biol 5(5):344–350. 6. Staswick PE, Tiryaki I (2004) The oxylipin signal jasmonic acid is activated by an enzyme that conjugates it to isoleucine in Arabidopsis. Plant Cell 16(8):2117–2127. 7. Wasternack C, Kombrink E (2010) Jasmonates: Structural requirements for lipidderived signals active in plant stress responses and development. ACS Chem Biol 5(1): 63–77. 8. Yan Y, et al. (2007) A downstream mediator in the growth repression limb of the jasmonate pathway. Plant Cell 19(8):2470–2483. 9. Melotto M, et al. (2008) A critical role of two positively charged amino acids in the Jas motif of Arabidopsis JAZ proteins in mediating coronatine- and jasmonoyl isoleucinedependent interactions with the COI1 F-box protein. Plant J 55(6):979–988. 10. Thireault C, et al. (2015) Repression of jasmonate signaling by a non-TIFY JAZ protein in Arabidopsis. Plant J 82(4):669–679. 11. Zhang F, et al. (2015) Structural basis of JAZ repression of MYC transcription factors in jasmonate signalling. Nature 525(7568):269–273. 12. Acosta IF, et al. (2013) Role of NINJA in root jasmonate signaling. Proc Natl Acad Sci USA 110(38):15473–15478. 13. Kagale S, Links MG, Rozwadowski K (2010) Genome-wide analysis of ethyleneresponsive element binding factor-associated amphiphilic repression motif-containing transcriptional regulators in Arabidopsis. Plant Physiol 152(3):1109–1134. 14. Kazan K (2006) Negative regulation of defence and stress genes by EAR-motif-containing repressors. Trends Plant Sci 11(3):109–112. 15. Pauwels L, et al. (2010) NINJA connects the co-repressor TOPLESS to jasmonate signalling. Nature 464(7289):788–791. 16. Ke J, et al. (2015) Structural basis for recognition of diverse transcriptional repressors by the TOPLESS family of corepressors. Sci Adv 1(6):e1500107. 17. Long JA, Ohno C, Smith ZR, Meyerowitz EM (2006) TOPLESS regulates apical embryonic fate in Arabidopsis. Science 312(5779):1520–1523. 18. Zhu Z, et al. (2011) Derepression of ethylene-stabilized transcription factors (EIN3/ EIL1) mediates jasmonate and ethylene signaling synergy in Arabidopsis. Proc Natl Acad Sci USA 108(30):12539–12544. 19. Katsir L, Schilmiller AL, Staswick PE, He SY, Howe GA (2008) COI1 is a critical component of a receptor for jasmonate and the bacterial virulence factor coronatine. Proc Natl Acad Sci USA 105(19):7100–7105. 20. Xie DX, Feys BF, James S, Nieto-Rostro M, Turner JG (1998) COI1: An Arabidopsis gene required for jasmonate-regulated defense and fertility. Science 280(5366):1091–1094. 21. Sheard LB, et al. (2010) Jasmonate perception by inositol-phosphate-potentiated COI1-JAZ co-receptor. Nature 468(7322):400–405. 22. Yan J, et al. (2009) The Arabidopsis CORONATINE INSENSITIVE1 protein is a jasmonate receptor. Plant Cell 21(8):2220–2236.
23. Chung HS, et al. (2010) Alternative splicing expands the repertoire of dominant JAZ repressors of jasmonate signaling. Plant J 63(4):613–622. 24. Goossens J, Swinnen G, Vanden Bossche R, Pauwels L, Goossens A (2015) Change of a conserved amino acid in the MYC2 and MYC3 transcription factors leads to release of JAZ repression and increased activity. New Phytol 206(4):1229–1237. 25. Moreno JE, et al. (2013) Negative feedback control of jasmonate signaling by an alternative splice variant of JAZ10. Plant Physiol 162(2):1006–1017. 26. Chung HS, Howe GA (2009) A critical role for the TIFY motif in repression of jasmonate signaling by a stabilized splice variant of the JASMONATE ZIM-domain protein JAZ10 in Arabidopsis. Plant Cell 21(1):131–145. 27. Chini A, Gimenez-Ibanez S, Goossens A, Solano R (2016) Redundancy and specificity in jasmonate signalling. Curr Opin Plant Biol 33:147–156. 28. Chung HS, et al. (2008) Regulation and function of Arabidopsis JASMONATE ZIMdomain genes in response to wounding and herbivory. Plant Physiol 146(3):952–964. 29. Demianski AJ, Chung KM, Kunkel BN (2012) Analysis of Arabidopsis JAZ gene expression during Pseudomonas syringae pathogenesis. Mol Plant Pathol 13(1):46–57. 30. Çevik V, et al. (2012) MEDIATOR25 acts as an integrative hub for the regulation of jasmonate-responsive gene expression in Arabidopsis. Plant Physiol 160(1):541–555. 31. Fernández-Calvo P, et al. (2011) The Arabidopsis bHLH transcription factors MYC3 and MYC4 are targets of JAZ repressors and act additively with MYC2 in the activation of jasmonate responses. Plant Cell 23(2):701–715. 32. Chen R, et al. (2012) The Arabidopsis mediator subunit MED25 differentially regulates jasmonate and abscisic acid signaling through interacting with the MYC2 and ABI5 transcription factors. Plant Cell 24(7):2898–2916. 33. Gasperini D, et al. (2015) Multilayered organization of jasmonate signalling in the regulation of root growth. PLoS Genet 11(6):e1005300. 34. Yang DL, et al. (2012) Plant hormone jasmonate prioritizes defense over growth by interfering with gibberellin signaling cascade. Proc Natl Acad Sci USA 109(19): E1192–E1200. 35. Browse J (2009) Jasmonate passes muster: A receptor and targets for the defense hormone. Annu Rev Plant Biol 60:183–205. 36. Chini A, Fonseca S, Chico JM, Fernández-Calvo P, Solano R (2009) The ZIM domain mediates homo- and heteromeric interactions between Arabidopsis JAZ proteins. Plant J 59(1):77–87. 37. Fairhead M, Howarth M (2015) Site-specific biotinylation of purified proteins using BirA. Methods Mol Biol 1266:171–184. 38. Kabsch W (2010) Xds. Acta Crystallogr D Biol Crystallogr 66(Pt 2):125–132. 39. Collaborative Computational Project, Number 4 (1994) The CCP4 suite: Programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50(Pt 5):760–763. 40. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Cryst 40(Pt 4):658–674. 41. Emsley P, Cowtan K (2004) Coot: Model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60(Pt 12 Pt 1):2126–2132. 42. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53(Pt 3): 240–255. 43. Withers J, et al. (2012) Transcription factor-dependent nuclear localization of a transcriptional repressor in jasmonate hormone signaling. Proc Natl Acad Sci USA 109(49):20148–20153.
Zhang et al.
PNAS | February 14, 2017 | vol. 114 | no. 7 | 1725
PLANT BIOLOGY
polyethylene glycol monomethyl ether 2000, and MYC3(44–242)–JasJAZ10 was crystallized in 20% (wt/vol) polyethylene glycol 3350 and 0.2 M magnesium formate dihydrate (pH 7.0). The MYC3(44–242)–CMIDJAZ10 and MYC3(44–242)–JasJAZ10 structures were determined by molecular replacement using the MYC3(44–238) structure (PDB ID code: 4RQW) as a search model. All biochemical experiments were performed three or more times with similar results. Detailed experimental protocols and information can be found in SI Materials and Methods.
