EIN2 Mediates Direct Regulation of Histone Acetylation in the Ethylene Response

EIN2 mediates direct regulation of histone acetylation in the ethylene response

Fan Zhanga,b,1, Likai Wanga,b,1, Bin Qic, Bo Zhaoa,b, Eun Esther Koa,b, Nathaniel D. Riggana,b, Kevin China,b, and Hong Qiaoa,b,2

aInstitute for Cellular and Molecular Biology, The University of Texas at Austin, Austin, TX 78712; bDepartment of Molecular Biosciences, The University of Texas at Austin, Austin, TX 78712; and cDepartment of Molecular, Cellular, and Developmental Biology, University of Colorado Boulder, Boulder, CO 80309

Edited by Steven E. Jacobsen, University of California, Los Angeles, CA, and approved August 10, 2017 (received for review May 15, 2017) Ethylene gas is essential for developmental processes and stress of EBF1 and EBF2 (19, 20). In the nucleus, the EIN2-C transduces responses in plants. Although the membrane-bound protein EIN2 is signals to the transcription factors EIN3 and EIL1, which are key for critical for ethylene signaling, the mechanism by which the ethylene activation of expression of all ethylene-response genes (21, 22). We signal is transduced remains largely unknown. Here we show the recently discovered that acetylation at H3K23 and H3K14Ac is in- levels of H3K14Ac and H3K23Ac are correlated with the levels of volved in ethylene-regulated gene activation in a manner that de- EIN2 protein and demonstrate EIN2 C terminus (EIN2-C) is sufficient to pends on both EIN2 and EIN3 (23, 24). rescue the levels of H3K14/23Ac of ein2-5 at the target loci, using Here we show that the levels of H3K14/23Ac are positively CRISPR/dCas9-EIN2-C. Chromatin immunoprecipitation followed by correlated with the EIN2 protein levels and demonstrate that deep sequencing (ChIP-seq) and ChIP-reChIP-seq analyses revealed that EIN2-C is sufficient to rescue the levels of H3K14Ac and EIN2-C associates with histone partially through an interaction with H3K23Ac in ein2-5 at loci targeted using CRISPR/dCas9-EIN2-C. EIN2 nuclear-associated protein1 (ENAP1), which preferentially binds Chromatin immunoprecipitation followed by deep sequencing to the genome regions that are associated with actively expressed (ChIP-seq) and ChIP-reChIP-seq analyses reveal that EIN2-C genes both with and without ethylene treatments. Specifically, in the associates with histones partially through an interaction with presence of ethylene, ENAP1-binding regions are more accessible upon EIN2 nuclear-associated protein1 (ENAP1), which preferentially the interaction with EIN2, and more EIN3 proteins bind to the loci

where ENAP1 is enriched for a quick response. Together, these results binds to the genome regions that are associated with actively PLANT BIOLOGY reveal EIN2-C is the key factor regulating H3K14Ac and H3K23Ac in expressed genes in both the presence and absence of ethylene. response to ethylene and uncover a unique mechanism by which Specifically, in the presence of ethylene, ENAP1-binding regions ENAP1 interacts with chromatin, potentially preserving the open chro- are more accessible upon the interaction with EIN2, and more matin regions in the absence of ethylene; in the presence of ethylene, EIN3 proteins bind to the loci where ENAP1 is enriched for a EIN2 interacts with ENAP1, elevating the levels of H3K14Ac and quick response. Together, these results reveal EIN2-C is the key H3K23Ac, promoting more EIN3 binding to the targets shared with factor regulating H3K14Ac and H3K23Ac in response to ethylene ENAP1 and resulting in a rapid transcriptional regulation. and uncover a unique mechanism by which ENAP1 interacts with chromatin, potentially preserving the open chromatin regions in histone acetylation | ethylene | Arabidopsis | CRISPR/dCas9 | ChIP-reChIP-seq Significance

he plant hormone ethylene is essential for a myriad of Upon receipt of different cues, transcription factors bind to Tphysiological and developmental processes. It is important in specific DNA sequences to recruit the general transcriptional responses to stress, such as drought, cold, flooding, and pathogen machinery for gene expression. Chromatin modification plays a infection (1, 2), and modulates stem cell division (3). The com- central role in the regulation of gene expression by providing mon aquatic ancestor of plants possessed the ethylene signaling transcription factors and the transcription machinery with dy- pathway and the mechanism has been elucidated by analysis of namic access to an otherwise tightly packaged genome. We use Arabidopsis (4). Ethylene is perceived by a family of receptors Arabidopsis to study how chromatin perceives ethylene sig- bound to the membrane of the endoplasmic reticulum (ER) that naling, an important plant hormone in plant growth, devel- are similar in sequence and structure to bacterial two-component opment, and stress responses. We demonstrate that the histidine kinases (5–9). Each ethylene receptor has an N-terminal essential factor EIN2, which mediates ethylene signaling from the transmembrane domain, and the receptors form dimers that bind endoplasmic reticulum to the nucleus, directly regulates histone ethylene via a copper cofactor RAN1 (10, 11). Signaling from one acetylation through an interaction with a histone-binding protein. of the receptors, ETR1, induces its physical association with the This study reveals the novel mechanism of how chromatin per- ER-localized protein RTE1 (12). The ethylene receptors function ceives the hormone signals to integrate into gene regulation. redundantly to negatively regulate ethylene responses (5) via a downstream Raf-like protein kinase called CTR1 (13, 14). Author contributions: F.Z., B.Q., and H.Q. designed research; F.Z., L.W., B.Q., B.Z., E.E.K., N.D.R., and K.C. performed research; L.W. and H.Q. analyzed data; and F.Z. and H.Q. In the absence of ethylene, both the ethylene receptors and wrote the paper. CTR1 are active, and CTR1 is associated with the ER membrane The authors declare no conflict of interest. through a direct interaction with ETR1 (13). The CTR1 downstream This article is a PNAS Direct Submission. factor, EIN2, is localized to the ER membrane, where it interacts with ETR1 (15). The protein stability of EIN2 is regulated by two Freely available online through the PNAS open access option. F-boxproteins,ETP1andETP2,whichmediateditsdegradationby Data deposition: The data reported in this paper have been deposited in the National – Center for Biotechnology Information Sequence Read Archive (NCBI SRA) Gene Expres- the ubiquitin proteasome pathway (16). In the absence of ethylene, sion Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accession nos. the CTR1-mediated phosphorylation at the C-terminal end of EIN2 GSE97288, GSE81200, and GSE83214). (EIN2-C) leads to a repression of EIN2 activity (17). In the presence 1F.Z. and L.W. contributed equally to this work. of ethylene, the EIN2-C is dephosphorylated, cleaved from the rest of 2To whom correspondence should be addressed. Email: [email protected]. the protein, and translocated into the nucleus (18) or into the P-body This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. (19, 20). In the P-body, the EIN2-C mediates translational repression 1073/pnas.1707937114/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1707937114 PNAS Early Edition | 1of6 Downloaded by guest on October 2, 2021 the absence of ethylene; in the presence of ethylene, EIN2 interacts with ENAP1, elevating the levels of H3K14/23Ac, promoting more EIN3 proteins binding to the targets shared with ENAP1 and resulting in a rapid transcriptional regulation. Results EIN2 Is the Key Regulator of Histone Acetylation of H3K14 and H3K23 in Response to Ethylene. Previous work demonstrated that EIN2 is involved in the regulation of H3K14Ac and H3K23Ac levels in ethylene response (23). To explore the molecular mechanisms that underlie this regulation, we examined the levels of H3K14Ac and H3K23Ac in EIN2S645A transgenic plants, in which EIN2-C is con- stitutively localized to the nucleus, and these plants display a constitutive ethylene responsive phenotype (18). To our surprise, significant amounts of H3K14Ac and H3K23Ac were elevated in EIN2S645A in comparison to Col-0 plants treated with ethylene (Fig. 1A). Based on this result and our previous data (23), we speculated that EIN2-C itself is the key factor for histone acetylation in the ethylene response. We then examined the global levels of H3K14Ac and H3K23Ac in EIN2 gain-of-function plants (EIN2ox)andCol-0 and ein2-5 mutant plants treated with or without ethylene by Western blot. As previously reported, both H3K14Ac and H3K23Ac were elevated in Col-0 plants by ethylene treatment (Fig. 1B). The levels of H3K14Ac and H3K23Ac were higher in EIN2ox plants than in Col-0 plants both with and without ethylene treatments (Fig. 1B), and they were lower in ein2-5 mutant than in Col-0 plants in the absence of ethylene. No ethylene-induced elevations were detected in the ein2-5 mutant (Fig. 1B).ToconfirmtheWesternblotresultatthe molecular level, we revisited previously published ChIP-seq data of H3K14Ac and H3K23Ac from Col-0 and ein2-5 mutant plants (23). In the ein2-5 mutant, levels of both H3K14Ac and H3K23Ac were lower than in Col-0 plants, whereas no significant difference was detected for H3K9Ac (Fig. 1 C–E and Fig. S1). Taken together, these results suggest the EIN2-C has a key role in the regulation of Fig. 2. gRNA-EIN3-T/dCas9-EIN2-C partially restores the ein2-5 phenotype histone acetylation of H3K14 and H3K23 in the ethylene response. (A) Diagram showing the construction of binary vector containing both dCas9 and guide RNAs. Guide RNA expression is driven by the OsU3 promoter EIN2-C Is Able to Restore the Levels of Histone Acetylation at (39), and dCas9-EIN2-C is driven by the 35S promoter. (B) The seedling phe- ein2-5 notype of CRISPR/dCas9-EIN2-C transgenic plants is shown. Three-day-old Ethylene-Regulated Gene Loci in the Mutant. To further ex- μ amine the function of EIN2-C in histone regulation, we decided etiolated seedlings were grown on the medium with or without 10 M ACC before being photographed. (C and D) The measurements of hypocotyls to use the CRISPR-dCas9 system to test the function of EIN2-C (C) and roots (D) in 3-d-old etiolated seedlings of the plants are shown. in histone acetylation in specific loci. As shown in Fig. 2A,two Plotted are means ± SD of at least 30 seedlings. Different letters indicate previously described point mutations were introduced into Cas9 to significant differences (P < 0.05).

generate the deactivated dCas9 (25) and the EIN2-C coding sequence was fused with that of dCas9. We designed two guide RNAs for targeting EBF2 promoter regions (gEBF2a and gEBF2b), regions at which H3K23Ac is elevated by ethylene in an EIN2- dependent manner (23). We also designed a guide RNA for targeting to the loci containing the EIN3-binding motif (gEIN3-T) (Fig. 2A) (22). The constructs described in Fig. 2A were introduced into the ein2-5 mutant; dCasS9-EIN2C without guide RNAs was introduced into the ein2-5 mutant and served as a control. The ein2-5 mutants that expressed gRNA-EIN3-T/dCas9- EIN2-C (gRNA-EIN3-T/dCas9-EIN2-C/ein2-5) were slightly dwarfed compared with the ein2-5 mutant treated with or without ethylene, and the phenotype of gRNA-EBF2a,b/dCas9-EIN2- C/ein2-5 plants was similar to that of the ein2-5 mutant (Fig. 2B). The expression of dCas9-EIN2-C was detected in all of the plants examined (Fig. S2A). ChIP–quantitative PCR (ChIP-qPCR) Fig. 1. EIN2 is the key factor that regulates H3K14Ac and H3K23Ac in of EIN2-C in gRNA-EIN3-T/dCas9-EIN2-C/ein2-5 and gRNA- ethylene response. (A) Total histone proteins extracted from 3-d-old etiolated S645A EBF2a/b/dCas9-EIN2-C/ein2-5 plants confirmed the presence of EIN2 and Col-0 seedlings were analyzed by Western blot for H3K14Ac and EIN2-C at target loci (Fig. S2B). We next conducted ChIP-qPCR H3K23Ac. (B) Total histone proteins extracted from 3-d-old etiolated EIN2ox, to quantify H3K14Ac and H3K23Ac in the targeted loci in Col-0, Col-0, and ein2-5 seedlings treated with air or with ethylene were analyzed by Western blot for H3K14Ac and H3K23Ac. (C–E) Average normalized H3K9Ac ein2-5, gRNA-EIN3-T/dCas9-EIN2-C/ein2-5,andgRNA-EBF2a,b/ (C), H3K14Ac (D), and H3K23Ac (E) ChIP-seq signals in regions 2 kbp up- and dCas9-EIN2-C/ein2-5 plants treated with or without ethylene. The downstream of genes in ein2-5 and Col-0 plants are plotted. Transcription start levels of H3K14Ac and H3K23Ac in gRNA-EIN3-T/dCAS9-EIN2- sites (TSS) and transcription termination sites (TTS) were aligned. C/ein2-5 and gRNA-EBF2a,b/dCas9-EIN2-C/ein2-5 plants were

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1707937114 Zhang et al. Downloaded by guest on October 2, 2021 enriched significantly relative to those in Col-0 plants treated with EIN2 Associates with Histone Partially Through the Interaction with air, and the levels were comparable to those in Col-0 plants ENAP1. EIN2-C does not contain DNA or chromatin-binding do- treated with ethylene (Fig. 3 A and B). No significant enrichment mains, so it is not clear how EIN2-C associates with chromatin to of H3K14Ac or H3K23Ac was detected in the ein2-5 mutant with regulate histone acetylation. Given that EIN2-C interacts with or without ethylene treatments (Fig. 3 A and B). Levels of ENAP1, a histone-binding protein (23), we conducted ENAP1 ChIP sequencing and ENAP1-ChIP/EIN2-reChIP sequencing by H3K14Ac and H3K23Ac in gRNA-EIN3-T/dCas9-EIN2-C/ein2-5 using a long-arm cross-linker, ethylene glycol bis-(succinimidyl and gRNA-EBF2a,b/dCas9-EIN2-C/ein2-5 were not regulated by succinate) (EGS) in pENAP1-ENAP1-YFP plants with or without ethylene due to the constitutive expression of EIN2-C in the target ethylene treatment. We identified 7,484 and 6,460 ENAP1 target loci. In addition, no off-target activity for dCas9-EIN2-C was ob- genes in the plants treated with air and ethylene, respectively (Fig. served in gRNA-EIN3-T/dCas9-EIN2-C/ein2-5 and gRNA-EBF2a, 4A), and most peaks were located in the promoter regions (Fig. S4 b/dCas9-EIN2-C/ein2-5 plants (Fig. 3 A and B). A and B). By ENAP1-ChIP/EIN2-reChIP sequencing in pENAP1- Expression of the target genes in gRNA-EIN3-T/dCas9-EIN2- ENAP1-YFP plants treated with ethylene, 1,739 genes bound by C/ein2-5 and gRNA-EBF2a,b/dCas9-EIN2-C/ein2-5 plants was both ENAP1 and EIN2 were identified (Fig. 4B), and more than significantly elevated compared with levels in Col-0 plants without 50% of ENAP1-ChIP/EIN2-reChIP targets were overlapped with ethylene treatment (Fig. 3C), indicating a positive correlation be- ENAP1-binding targets in the presence of ethylene (Fig. 4B), and tween the levels of H3K23Ac and H3K14Ac and gene expression. the binding signals were significantly enriched in those targets with Additionally, levels of mRNA produced from the target loci were not thepresenceofethylene(Fig. S4D). The genome browser data and ENAP1-ChIP/EIN2-reChIP qPCR of selected loci further subjected to ethylene regulation due to the constitutive expression of confirmed these findings (Fig. 4 C–E). EIN2-C (Fig. 3C). Further, the dCas9-EIN2-C/ein2-5 control plants, We then examined the association of ENAP1-ChIP/EIN2- which do not contain guide RNAs, displayed a slightly dwarf phe- reChIP target genes with H3K23Ac. We found that more than notype, and the gene expression in the control plants was slightly 70% of the genes bound by ENAP1 and EIN2 were also marked elevated, but with no locus specificity compared with gRNA-EIN3- by H3K23Ac in the presence of ethylene (Fig. 4F). ChIP-qPCR T/dCas9-EIN2-C/ein2-5 and gRNA-EBF2a,b/dCas9-EIN2-C/ein2-5 of H3K23Ac further confirmed the association of EIN2-C with (Fig. S3 A–E). Taken together, these results demonstrate EIN2-C H3K23Ac (Fig. 4G and Fig. S4E). is sufficient to rescue the levels of H3K14Ac and H3K23Ac in the

ein2-5 mutant, and the elevation of H3K23Ac and H3K14Ac is EIN2-C Regulates ENAP1 Binding in Ethylene Signaling. To explore PLANT BIOLOGY correlated with the activation of gene expression. how EIN2-C functions through ENAP1 in response to ethylene, we first examined ENAP1 binding both with and without ethylene treatments by ChIP-seq, using formaldehyde as a cross-linker (Fig. S5A). Nearly 7,000 and 5,000 ENAP1-binding targets were identified from the plants treated with air and ethylene, respectively (Fig. 5A). Most of the ENAP1 binding was in the promoter regions under both conditions (Fig. S5 B and C). To our surprise, we detected down-regulation of ENAP1 binding upon ethylene treatment (Fig. 5B and Fig. S5D), even though the protein levels of ENAP1 did not differ in the presence and absence of ethylene (Fig. S5E). However, previously we did not detect ethylene-induced differential binding of ENAP1 by using a long-arm cross-linker both in ChIP-seq and in ChIP-qPCR (Figs. 4D and 5C and Fig. S4C). We speculate that the differential binding of ENAP1 detected by using formaldehyde is due to the type of interaction that oc- curs between ENAP1 and EIN2 upon ethylene treatment. To test this idea, we conducted ENAP1-ChIP-qPCR, using formaldehyde or the long-arm cross-linker in plants treated with ethylene for different periods of time. No significant differences in ENAP1 binding were detected over time, using the long-arm cross-linker (Fig. 5D and Fig. S5F). However, using the formaldehyde cross-linker, ENAP1 binding was decreased after 15 min of ethylene treatment (Fig. 5E). This is the time frame in which the nuclear translocation of EIN2-C is observed (18). To explore the connection between the down-regulation of ENAP1 binding and its interaction with EIN2, we conducted ENAP1-ChIP/EIN2-reChIP-qPCR in plants treated with ethylene fordifferentperiodsoftimeasshowninFig.4D and E. The levels of EIN2 binding detected were higher in plants treated with ethylene (Fig. 5F), which anticorrelated with ENAP1 binding. Further, the ChIP-qPCR assay showed that the levels of H3K23Ac were elevated with the same treatments (Fig. 5G), which showed a positive correlation with the levels of EIN2 binding. Taken together, these results suggest that in the presence of ethylene, Fig. 3. EIN2-C restores H3K14Ac and H3K23Ac levels in the ein2-5 mutant changes in ENAP1 binding are caused by an interaction with EIN2- background. (A and B) ChIP-qPCR to examine the levels of H3K14Ac (A) and H3K23Ac (B) from 3-d-old etiolated seedlings of Col-0, ein2-5, and the C, which results in an increase in H3K23Ac over the target loci. transgenic plants, treated with air or ethylene. (C) qRT-PCR to examine the target gene expression in 3-d-old etiolated seedlings of the plants shown. EIN2 Is Required for Alterations in ENAP1 Binding in the Ethylene Each ChIP-qPCR or qRT-PCR was repeated at least three times, and a repre- Response. To further explore how EIN2 regulates ENAP1 binding sentative result is presented. in response to ethylene, we examined ENAP1 binding by ChIP-seq

Zhang et al. PNAS Early Edition | 3of6 Downloaded by guest on October 2, 2021 etiolated Col-0 seedlings (Fig. S7A). We then examined average ENAP1 ChIP-seq signals in the five groups of genes. Near transcription start-site (TSS) regions, highly expressed genes showed significantly higher levels of average ENAP1 ChIP signals than in other groups of genes (Fig. 7 A and B). The majority of the ENAP1-bound genes were in the top two groups of most highly expressed genes (Fig. S7B), suggesting that ENAP1 prefers to bind actively transcribed gene regions. Histone acetylation is a well- known mark of gene activation, and we found that about 50% of H3K9Ac, 63% of H3K14Ac, and 53% of H3K23Ac marked genes were bound by ENAP1 (Fig. S7 C–E). To further confirm that ENAP1 prefers to bind actively transcribed gene regions, we examined DNA accessibility at the ENAP1-associated regions, using formaldehyde-assisted isolation of regulatory elements (FAIRE)-qPCR in ENAP1ox, amiR-ENAP1/2, and Col-0 plants untreated or treated with ethylene for different periods of time (Fig. 7C and Fig. S7F). Compared with Col-0 plants, the DNA accessibility in ENAP1-bound regions was slightly higher in ENAP1ox plants and significantly lower in amiRENAP1/enap2 plants, and differences became more significant with longer periods of ethylene treatments (Fig. 7C and Fig. S7F). This result further confirms that ENAP1 associates with actively transcribed gene regions. Fig. 4. EIN2 associates with histone partially through ENAP1. (A) Venn di- EIN3 is the key transcription factor in the ethylene signaling agram showing the ENAP1-bound targets with and without the ethylene pathway (23). Therefore, we examined the average EIN3 ChIP- treatments. ChIP sequencing using a long-arm cross-linker was conducted in 3-d- seq signals and ENAP1 ChIP-seq signals in TSS regions in the old etiolated seedlings of pENAP1-ENAP1-YFP seedlings treated with air or ethylene. (B) Comparison of ENAP1-bound targets with ENAP1-ChIP/EIN2-C- reChIP binding targets. The first ChIP was conducted by using anti-GFP antibody, and the second ChIP (reChIP) was conducted by using anti-EIN2-C antibody with a long-arm cross-linker. (C) Genome browser traces of ENAP1 ChIP-seq and the ENAP1-ChIP/EIN2-C-reChIP-seq data from representative genes. (D)Fold change in ENAP1-ChIP-qPCR signal relative to Col-0 in air obtained using a long- arm cross-linker in both air and ethylene. (E) ChIP-qPCR of selected loci to confirm the ChIP-reChIP-seq result of EIN2-C. (F) Venn diagram showing the overlap of ENAP1-bound targets with H3K23Ac-marked genes. (G) ChIP-qPCR of H3K23Ac showing the elevation of H3K23Ac in EIN2-C–associated genes.

in the ein2-5 mutant both with and without ethylene treatments (Fig. S6A). The ENAP1 binding in the ein2-5 mutant was enriched in the promoter regions in both conditions as it was in Col-0 plants (Fig. S6 B and C). However, the ethylene-induced change in ENAP1 binding was not detected in the ein2-5 mutant (Fig. 6A). These results were confirmed by ChIP-qPCR in ENAP1-YFP/ein2-5 and ENAP1-YFP/Col-0 plants (Fig. 6B). We also noticed that even though ENAP1 protein levels in Col- 0 plants were similar to those in the ein2-5 mutant (Fig. S6D), the average ENAP1-binding signals in the ein2-5 mutant were lower than in Col-0 plants even without ethylene treatment (Fig. 6B), and this result was further confirmed by comparing the ChIP-seq data from ENAP1 in Col-0 and ein2-5 mutant plants (Fig. 6C). We then examined the gene expression in ein2-5 and ENAP1OE/ein2-5 Fig. 5. ENAP1 binding is regulated by ethylene treatment. (A) Venn dia- plants with or without ethylene treatment. As in the ein2-5 mutant, gram showing overlap of ENAP1-bound targets in air and ethylene. The the ethylene-induced gene expression was largely impaired in ChIP-seq was done by using formaldehyde as a cross-linker in 3-d-old etio- ENAP1OE/ein2-5 lated pENAP1-ENAP1-YFP seedlings treated with air or ethylene. (B) Mean plants. More than 92% of genes regulated by enrichment profiles [Log2 reads per kilobase per million mapped reads ethylene in Col-0 plants were not differentially expressed in (RPKM) (ChIP/Input)] around 2 kb up- and downstream of TSSs in target ENAP1OE/ein2-5 plants in the presence and absence of ethylene genes with or without ethylene treatment using formaldehyde cross-linker. (Fig. 6D). These results provide additional molecular evidence that (C) Mean enrichment profiles around ±2 kb of TSSs in target genes with or the function of ENAP1 in the ethylene response is EIN2 dependent. without ethylene treatment using a long-arm cross-linker. (D) ChIP-qPCR to examine ENAP1 binding in pENAP1-ENAP1-YFP plants treated with ethylene ENAP1 Binds to Active Gene Regions and EIN3 Prefers to Bind ENAP- for different periods of time by using a long-arm cross-linker. (E)ChIP-qPCR Targeted Regions. The majority of ENAP1 binding was in the using a formaldehyde cross-linker to examine ENAP1 binding in pENAP1- promoter regions (Figs. S4 A and B and S5 B and C), suggesting ENAP1-YFP 3-d-old etiolated plants treated with ethylene for different periods the potential function of ENAP1 in gene regulation. To examine of time relative to no treatment. (F) ChIP-reChIP PCR to examine the binding of EIN2 through the interaction with ENAP1. The first ChIP was performed with whether there is a correlation between ENAP1 binding and gene anti-GFP antibody and reChIP with anti-EIN2-C antibody in pENAP1-ENAP1-YFP expression that would indicate a function in regulation of gene plants treated with ethylene for different periods of time by using a long-arm expression, we divided the ≈30,000 genes in the Arabidopsis ge- cross-linker. (G) ChIP-qPCR to examine H3K23Ac in pENAP1-ENAP1-YFP 3-d-old nome into five groups based on their expression levels in 3-d-old etiolated seedlings treated with ethylene for indicated times.

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1707937114 Zhang et al. Downloaded by guest on October 2, 2021 proximity of ENAP1 to DNA is altered by ethylene treatment. The elevation of H3K23Ac in the presence of ethylene (Fig. 5G) and the FAIRE-qPCR in Fig. 7C provide additional evidence to support our hypothesis that the chromatin is more accessible in ENAP1-bound regions in the presence of ethylene than in the absence of ethylene. Further, we showed that EIN2-C shares a subset of binding targets with ENAP1 by ChIP-reChIP-seq (Fig. 4); all these data suggest a potential mechanism that in the presence of ethylene, ENAP1 interacts with EIN2-C, which leads to the elevation of H3K14Ac and H3K23Ac, resulting in the chromatin regions being more open in the presence of ethylene. A previous study showed that the ethylene response can be detected within 30 min at the transcriptional level (22), and we observed that in the presence of ethylene, EIN3, the key transcription factor in ethylene signaling (21, 22, 38), prefers to bind ENAP1-targeted regions (Fig. 7 D and E). Furthermore, in the absence of ethylene, ENAP1 prefers to bind to the DNA regions Fig. 6. Ethylene-induced change of ENAP1 binding is EIN2 dependent. (A) Mean enrichment of ENAP1 ChIP-seq signals along target gene bodies in 3-d-old etiolated ein2-5 seedlings treated with air or ethylene. Genes were ranked according to relative mRNA expression levels and divided into five equal sets. (B) ENAP1-ChIP-qPCR in 3-d-old etiolated Col-0 and pENAP1- ENAP1-YFP and PENAP1-ENAP1-YFP/ein2-5 seedlings. (C) Mean enrichment of ENAP1-ChIP-seq signals along 2 kb up- and downstream of TSSs of target genes in Col-0 or ein2-5 mutant seedlings. (D) Overlap of genes regulated by ethylene in Col-0, ein2-5, and ENAP1OE/ein2-5 seedlings.

genes cobound by ENAP1 and EIN3 or in the genes bound only PLANT BIOLOGY by EIN3. The average EIN3-binding signals were significantly higher in cobound genes than in the genes bound only by EIN3 (Fig. 7D and Fig. S7G). No significant differences were detected for the average ENAP1 ChIP-seq signals near TSS regions of genes that were cobound with EIN3 or bound only by ENAP1 (Fig. 7E), showing that EIN3 prefers to bind to ENAP1 target regions, which are more accessible. Discussion Histone acetylation has been demonstrated to function in different plant hormones (23, 26–29), and several histone deacetylases are known to be involved in the regulation of gene expression (30–37). The data presented here showed that the C-terminal domain of EIN2 (EIN2-C) is the key factor that drives alterations in histone acetylation during the ethylene response. However, the exact bio- chemistry functions of EIN2-C remain unknown. We performed a careful motif search in the EIN2-C but no histone acetyltransferase (HAT) motif was identified. It is possible that in the presence of ethylene, EIN2-C is translocated into the nucleus, where it functions as an important scaffold protein to recruit other components required for regulating histone acetylation. As shown by Western blot, H3K14Ac and H3K23Ac were significantly elevated in EIN2S645A plants, and therefore it is possible that EIN2-C functions as a HAT. Even though it does not contain a known HAT motif, it is possible that, when properly folded in vivo, EIN2-C forms a HAT activity domain. A crystal structure of EIN2-C will provide insight into how EIN2-C functions. In addition, isolation of the EIN2-C–containing Fig. 7. ENAP1 prefers to bind actively transcribed genes. (A) Mean enrich- protein complex that forms in the nucleus in the presence of eth- ment of ENAP1 ChIP-seq signals along gene bodies in 3-d-old etiolated ylene will reveal whether EIN2-C associates with a known HAT. pENAP1-ENAP1-YFP seedlings. Genes were ranked according to relative In this study, we observed differences in ENAP1 binding when mRNA expression levels and divided into five equal sets. (B) Heat maps show different cross-linkers were used in ChIP-seq. The differences ENAP1 enrichment in all genes in Arabidopsis along with relative mRNA were not due to the protein levels, because ENAP1 protein levels values. Shown is a scatter plot for mRNAs ranked according to gene ex- are similar in the presence and absence of ethylene (Fig. S5D). pression levels in 3-d-old Col-0 seedlings under air treatment. Heat maps We then revisited our ChIP-seq data and found that similar total are ranked according to gene expression levels. (C) FAIRE-qPCR showing reads and mapping rates were obtained using ENAP1 ChIP-seq ENAP1 binding is associated with DNA accessibility in the plants shown. (D) Mean enrichment of EIN3 ChIP-seq signals (RPKM) 2 kb up- and down- both with and without ethylene treatments (Table S1), showing stream of TSSs of the genes cobound by ENAP1 and EIN3 (red) or of the that the ChIP-seq data are not a reason for ENAP1-binding genes bound only by EIN3 (blue) (22). (E) Mean enrichment of ENAP1 ChIP- change. Analyses of ENAP1 binding in response to ethylene seq signals around 2 kb up- and downstream of TSSs of the genes cobound performed using different cross-linkers strongly suggest that the by ENAP1 and EIN3 (red) or of the genes bound only by ENAP1 (blue).

Zhang et al. PNAS Early Edition | 5of6 Downloaded by guest on October 2, 2021 that are accessible (Fig. 7 A–C), indicating that ENAP1 may EIN2 and ENAP1 in the ein3-1 mutant will provide more insight serve as a place holder in open chromatin regions to allow EIN3 to into the molecular function of EIN3 in histone acetylation in the quickly bind to target loci for a rapid ethylene response. It should ethylene response. be noted that not all of the binding targets of EIN3 are overlapped with ENAP1-bound regions (Fig. 7D). Due to the weak ethylene- Materials and Methods responsive phenotype of amiR-ENAP1/enap2 plants, we speculate Details of plant growth and measurements of roots and hypocotyls, ChIP-seq that other factors in addition to ENAP1 are involved in EIN3- and RNA-seq processing and analysis, and ChIP-PCR and FAIRE-qPCR analysis mediated gene regulation. are described in SI Materials and Methods. All of the primers for ChIP-qPCR, Based on data shown here, we propose a unique function qPCR, and FAIRE qPCR are listed in Table S2. mechanism, by which ENAP1 interacts with histones in the absence of ethylene, potentially preserving the open chromatin ACKNOWLEDGMENTS. We thank S. Sung and his laboratory members for status. This enables a rapid response to ethylene stimulation. In comments, V. Iyer and S. Roux for proofreading, and Natalie Ahn and Nancy the presence of ethylene, EIN2 interacts with ENAP1 and other Vega for plants and laboratory maintenance. We thank the Salk Institute Genomic Analysis Laboratory for providing seeds. We thank the genomic unidentified factors, elevating H3K14Ac and H3K23Ac levels sequencing and analysis facility of the Institute of Cellular and Molecular Biology and promoting EIN3 binding to the targets shared with ENAP1, at The University of Texas at Austin for RNA-seq and ChIP-seq. We thank V. Iyer and resulting in a rapid transcriptional response to the presence for advice on ChIP-seq data analysis. This work was supported by grants from the of ethylene (Fig. S8). Further examining the binding activity of National Institutes of Health (to H.Q.) (NIH-1R01GM115879-01).

1. Bailey-Serres J, et al. (2012) Making sense of low oxygen sensing. Trends Plant Sci 17:129–138. 28. Liu X, et al. (2012) Histone acetyltransferases in rice (Oryza sativa L.): Phylogenetic 2. Achard P, et al. (2007) The plant stress hormone ethylene controls floral transition via analysis, subcellular localization and expression. BMC Plant Biol 12:145. DELLA-dependent regulation of floral meristem-identity genes. Proc Natl Acad Sci 29. Zhu YX (2010) The epigenetic involvement in plant hormone signaling. Chin Sci Bull USA 104:6484–6489. 55:2198–2203. 3. Ortega-Martínez O, Pernas M, Carol RJ, Dolan L (2007) Ethylene modulates stem cell 30. Zhou C, Zhang L, Duan J, Miki B, Wu K (2005) HISTONE DEACETYLASE19 is involved in division in the Arabidopsis thaliana root. Science 317:507–510. jasmonic acid and ethylene signaling of pathogen response in Arabidopsis. Plant Cell 4. Ju C, Chang C (2015) Mechanistic insights in ethylene perception and signal trans- 17:1196–1204. duction. Plant Physiol 169:85–95. 31. Wu K, Zhang L, Zhou C, Yu CW, Chaikam V (2008) HDA6 is required for jasmonate 5. Hua J, Meyerowitz EM (1998) Ethylene responses are negatively regulated by a re- response, senescence and flowering in Arabidopsis. J Exp Bot 59:225–234. ceptor gene family in Arabidopsis thaliana. Cell 94:261–271. 32. Zhu Z, et al. (2011) Derepression of ethylene-stabilized transcription factors (EIN3/ 6. Chang C, Kwok SF, Bleecker AB, Meyerowitz EM (1993) Arabidopsis ethylene-response EIL1) mediates jasmonate and ethylene signaling synergy in Arabidopsis. Proc Natl gene ETR1: Similarity of product to two-component regulators. Science 262:539–544. Acad Sci USA 108:12539–12544. 7. Hua J, et al. (1998) EIN4 and ERS2 are members of the putative ethylene receptor 33. Chen LT, Wu K (2010) Role of histone deacetylases HDA6 and HDA19 in ABA and gene family in Arabidopsis. Plant Cell 10:1321–1332. abiotic stress response. Plant Signal Behav 5:1318–1320. 8. Bleecker AB, Esch JJ, Hall AE, Rodríguez FI, Binder BM (1998) The ethylene-receptor family 34. Wang Z, et al. (2013) Arabidopsis paired amphipathic helix proteins SNL1 and from Arabidopsis: Structure and function. Philos Trans R Soc Lond B Biol Sci 353:1405–1412. SNL2 redundantly regulate primary seed dormancy via abscisic acid-ethylene antag- 9. Hirayama T, et al. (1999) RESPONSIVE-TO-ANTAGONIST1, a Menkes/Wilson disease-related onism mediated by histone deacetylation. Plant Cell 25:149–166. copper transporter, is required for ethylene signaling in Arabidopsis. Cell 97:383–393. 35. Perrella G, et al. (2013) Histone deacetylase complex1 expression level titrates plant 10. Schaller GE, Ladd AN, Lanahan MB, Spanbauer JM, Bleecker AB (1995) The ethylene growth and abscisic acid sensitivity in Arabidopsis. Plant Cell 25:3491–3505. response mediator ETR1 from Arabidopsis forms a disulfide-linked dimer. JBiolChem 36. Lee K, Park OS, Jung SJ, Seo PJ (2016) Histone deacetylation-mediated cellular de- 270:12526–12530. differentiation in Arabidopsis. J Plant Physiol 191:95–100. 11. Sakai H, et al. (1998) ETR2 is an ETR1-like gene involved in ethylene signaling in 37. Colville A, et al. (2011) Role of HD2 genes in seed germination and early seedling Arabidopsis. Proc Natl Acad Sci USA 95:5812–5817. growth in Arabidopsis. Plant Cell Rep 30:1969–1979. 12. Dong CH, et al. (2010) Molecular association of the Arabidopsis ETR1 ethylene re- 38. Guo H, Ecker JR (2003) Plant responses to ethylene gas are mediated by SCF(EBF1/ ceptor and a regulator of ethylene signaling, RTE1. J Biol Chem 285:40706–40713. EBF2)-dependent proteolysis of EIN3 transcription factor. Cell 115:667–677. 13. Huang Y, Li H, Hutchison CE, Laskey J, Kieber JJ (2003) Biochemical and functional 39. Xie K, Yang Y (2013) RNA-guided genome editing in plants using a CRISPR-Cas system. analysis of CTR1, a protein kinase that negatively regulates ethylene signaling in Mol Plant 6:1975–1983. Arabidopsis. Plant J 33:221–233. 40. Jinek M, et al. (2012) A programmable dual-RNA-guided DNA endonuclease in 14. Kieber JJ, Rothenberg M, Roman G, Feldmann KA, Ecker JR (1993) CTR1, a negative adaptive bacterial immunity. Science 337:816–821. regulator of the ethylene response pathway in Arabidopsis, encodes a member of the 41. Pedmale UV, et al. (2016) Cryptochromes interact directly with PIFs to control plant raf family of protein kinases. Cell 72:427–441. growth in limiting blue light. Cell 164:233–245. 15. Bisson MM, Groth G (2011) New paradigm in ethylene signaling: EIN2, the central 42. Ramirez F, Dundar F, Diehl S, Gruning BA, Manke T (2014) deepTools: A flexible regulator of the signaling pathway, interacts directly with the upstream receptors. platform for exploring deep-sequencing data. Nucleic Acids Res 42:W187–W191. Plant Signal Behav 6:164–166. 43. Xing D, Wang Y, Hamilton M, Ben-Hur A, Reddy AS (2015) Transcriptome-wide identi- 16. Qiao H, Chang KN, Yazaki J, Ecker JR (2009) Interplay between ethylene, ETP1/ETP2 fication of RNA targets of Arabidopsis SERINE/ARGININE-RICH45 uncovers the un- F-box proteins, and degradation of EIN2 triggers ethylene responses in Arabidopsis. expected roles of this RNA binding protein in RNA processing. Plant Cell 27:3294–3308. Genes Dev 23:512–521. 44. Kim D, et al. (2013) TopHat2: Accurate alignment of transcriptomes in the presence of 17. Ju C, et al. (2012) CTR1 phosphorylates the central regulator EIN2 to control ethylene insertions, deletions and gene fusions. Genome Biol 14:R36. hormone signaling from the ER membrane to the nucleus in Arabidopsis. Proc Natl 45. Trapnell C, et al. (2012) Differential gene and transcript expression analysis of RNA- Acad Sci USA 109:19486–19491. seq experiments with TopHat and Cufflinks. Nat Protoc 7:562–578. 18. Qiao H, et al. (2012) Processing and subcellular trafficking of ER-tethered EIN2 control 46. Bai MY, Fan M, Oh E, Wang ZY (2012) A triple helix-loop-helix/basic helix-loop-helix response to ethylene gas. Science 338:390–393. cascade controls cell elongation downstream of multiple hormonal and environ- 19. Merchante C, et al. (2015) Gene-specific translation regulation mediated by the mental signaling pathways in Arabidopsis. Plant Cell 24:4917–4929. hormone-signaling molecule EIN2. Cell 163:684–697. 47. Zhang F, et al. (2016) Phosphorylation of CBP20 links microRNA to root growth in the 20. Li W, et al. (2015) EIN2-directed translational regulation of ethylene signaling in ethylene response. PLoS Genet 12:e1006437. Arabidopsis. Cell 163:670–683. 48. Andrews S (2013) FastQC: A quality control tool for high throughput sequence data. 21. Chao Q, et al. (1997) Activation of the ethylene gas response pathway in Arabidopsis by Available at www.bioinformatics.babraham.ac.uk/projects/fastqc. Accessed August 3, 2016. the nuclear protein ETHYLENE-INSENSITIVE3 and related proteins. Cell 89:1133–1144. 49. Lamesch P, et al. (2012) The Arabidopsis Information Resource (TAIR): Improved gene 22. Chang KN, et al. (2013) Temporal transcriptional response to ethylene gas drives annotation and new tools. Nucleic Acids Res 40:D1202–D1210. growth hormone cross-regulation in Arabidopsis. eLife 2:e00675. 50. Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler 23. Zhang F, et al. (2016) EIN2-dependent regulation of acetylation of histone H3K14 and transform. Bioinformatics 25:1754–1760. non-canonical histone H3K23 in ethylene signalling. Nat Commun 7:13018. 51. Ramikie TS, et al. (2014) Multiple mechanistically distinct modes of endocannabinoid 24. Wang L, et al. (2017) Ethylene induces combinatorial effects of histone H3 acetylation mobilization at central amygdala glutamatergic synapses. Neuron 81:1111–1125. in gene expression in Arabidopsis. BMC Genomics 18:538. 52. Robinson JT, et al. (2011) Integrative genomics viewer. Nat Biotechnol 29:24–26. 25. Hilton IB, et al. (2015) Epigenome editing by a CRISPR-Cas9-based acetyltransferase 53. Zhang Y, et al. (2008) Model-based analysis of ChIP-Seq (MACS). Genome Biol 9:R137. activates genes from promoters and enhancers. Nat Biotechnol 33:510–517. 54. Shin H, Liu T, Manrai AK, Liu XS (2009) CEAS: Cis-regulatory element annotation 26. Yamamuro C, Zhu JK, Yang Z (2016) Epigenetic modifications and plant hormone system. Bioinformatics 25:2605–2606. action. Mol Plant 9:57–70. 55. Bardet AF, He Q, Zeitlinger J, Stark A (2011) A computational pipeline for compara- 27. Pandey R, et al. (2002) Analysis of histone acetyltransferase and histone deacetylase tive ChIP-seq analyses. Nat Protoc 7:45–61. families of Arabidopsis thaliana suggests functional diversification of chromatin 56. Wu M-F, et al. (2015) Auxin-regulated chromatin switch directs acquisition of flower modification among multicellular eukaryotes. Nucleic Acids Res 30:5036–5055. primordium founder fate. eLife 4:e09269.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1707937114 Zhang et al. Downloaded by guest on October 2, 2021