Ethylene promotes root hair growth throughcoordinated EIN3/EIL1 and RHD6/RSL1activity in ArabidopsisYing Fenga,b, Ping Xub, Bosheng Lia, Pengpeng Lib, Xing Wena, Fengying Anb, Yan Gongb, Yi Xinb, Ziqiang Zhub,c,Yichuan Wanga,1, and Hongwei Guoa,d,1

aInstitute of Plant and Food Science, Department of Biology, Southern University of Science and Technology, Shenzhen, Guangdong 518055, China; bTheState Key Laboratory of Protein and Plant Gene Research, College of Life Sciences, Peking University, Beijing 100871, China; cCollege of Life Sciences,Nanjing Normal University, Nanjing 210023, China; and dPeking-Tsinghua Center of Life Sciences, Beijing 100871, China

Edited by José M. Alonso, North Carolina State University, Raleigh, NC, and accepted by Editorial Board Member Joseph R. Ecker November 14, 2017 (receivedfor review July 7, 2017)

Root hairs are an extensive structure of root epidermal cells andare critical for nutrient acquisition, soil anchorage, and environ-mental interactions in sessile plants. The phytohormone ethylene(ET) promotes root hair growth and also mediates the effects ofdifferent signals that stimulate hair cell development. However,the molecular basis of ET-induced root hair growth remains poorlyunderstood. Here, we show that ET-activated transcription factorETHYLENE-INSENSITIVE 3 (EIN3) physically interacts with ROOTHAIR DEFECTIVE 6 (RHD6), a well-documented positive regulatorof hair cells, and that the two factors directly coactivate the hairlength-determining gene RHD6-LIKE 4 (RSL4) to promote root hairelongation. Transcriptome analysis further revealed the parallelroles of the regulator pairs EIN3/EIL1 (EIN3-LIKE 1) and RHD6/RSL1 (RHD6-LIKE 1). EIN3/EIL1 and RHD6/RSL1 coordinately en-hance root hair initiation by selectively regulating a subset of coreroot hair genes. Thus, our work reveals a key transcriptional com-plex consisting of EIN3/EIL1 and RHD6/RSL1 in the control of roothair initiation and elongation, and provides a molecular frame-work for the integration of environmental signals and intrinsicregulators in modulating plant organ development.

ethylene | root hair | EIN3 | RHD6 | RSL4

Plants, unlike animals, are sessile organisms constantly exposedto various environmental changes. Organ morphogenesis in

response to different challenges is critical for plant survival. Roothairs are unicellular extensions of root epidermal cells that helpincrease water and nutrient uptake, and improve soil anchorage.As the outermost plant surface in the soil, root hairs rapidly andeffectively respond to environmental stimuli. Their flexible naturemakes them an exceptional model system for studying the orga-nization and coordination of gene regulatory networks modulatedby endogenous and exogenous signals (1, 2).In the angiosperm model plant Arabidopsis thaliana, the two root

epidermal cell fates are determined by the relative positioning ofcells above the inner layer of cortical cells. Hair (H) cells overliethe junction between two cortical cells, whereas nonhair (N) cellsoverlie only one cortical cell (3). Although numerous genes influ-ence root epidermal cell morphology after cell fate determination,the core regulatory network is driven by a transcriptional cascade(2, 4). GLABRA 2 (GL2), a homeodomain-leucine zipper tran-scription factor, plays a vital role in N cell differentiation (5, 6).GL2 maintains N cell fate mainly through direct repression of agroup of basic helix–loop–helix (bHLH) family transcription factorsthat positively regulate root hair initiation and elongation in H cells(7). Among them, ROOT HAIR DEFECTIVE 6 (RHD6) plays amajor role in promoting H cell development (8–10). RHD6 and itsclosest homolog, RHD6-LIKE 1 (RSL1), form class I of the bHLHgroup VIII subfamily. RHD6/RSL1 positively regulate four class IImembers of group VIII, RSL2–5, which also positively regulate roothair growth (11). However, only RSL4 has been shown to be directlyregulated by RHD6 (12). RSL4 transcripts rapidly accumulate

before hair cells enter the elongation stage, with the amount ofaccumulated RSL4 mRNA proportional to the final root hairlength (13). Recently, phytohormone application, such as auxinand cytokinin treatment, and nutrient starvation were found tostimulate root hair growth in an RSL4-dependent manner (12,14, 15), revealing that RSL4 is an essential node that integratesmultiple root hair elongation signals.The gaseous phytohormone ethylene (ET) is a well-known stress

hormone that helps plants to survive under various biotic andabiotic stresses (16, 17). Upon ET binding, ET receptors deactivatethe Ser/Thr kinase CONSTITUTIVE TRIPLE RESPONSE 1(CTR1), leading to the hypophosphorylation and proteolytic acti-vation of ETHYLENE INSENSITIVE 2 (EIN2) (18). Down-stream of EIN2, EIN3 and its functionally redundant homologEIN3-LIKE 1 (EIL1) are two master transcription factors thatactivate numerous signaling cascades and ET responses (16, 17).Activated EIN2 protects EIN3/EIL1 from proteasomal degrada-tion, at least in part, by repressing the E3 ligases EIN3-BINDINGF-BOX 1 (EBF1) and EBF2 (19, 20). Thus, EIN3/EIL1 proteinsaccumulate in the nuclei upon ET treatment (21, 22). One notablerole of ET is to stimulate root hair growth. Exogenous applicationof the ET biosynthesis precursor 1-aminocyclopropane-1-carboxylic

Significance

Root hairs are unicellular extensions of root epidermal cells thathelp plants increase water and nutrient uptake and improve soilanchorage, both of which are crucial for the globally recognizedgoal of yield improvement with reduced fertilizer use. Previousstudies have implicated numerous genes and phytohormones inthe control of root hair development. This work uncovers themolecular mechanism of ethylene (ET)-promoted root hair growthand identifies a transcriptional complex consisting of EIN3/EIL1 and RHD6/RSL1 as the key regulator of root hair initiationand elongation. As ET mediates the effects of various root hairstimuli, this work also elucidates a convergent signaling networkthat integrates diverse environmental cues and intrinsic signals tomodulate plant organ development.

Author contributions: Y.F., Z.Z., and H.G. designed research; Y.F., P.X., P.L., X.W., Y.G.,Y.X., and Y.W. performed research; X.W., F.A., and Y.W. contributed new reagents/ana-lytic tools; Y.F. and B.L. analyzed data; and Y.F. and H.G. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission. J.M.A. is a guest editor invited by theEditorial Board.

Published under the PNAS license.

Data deposition: The sequencing data reported in this paper have been deposited in theGene Expression Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accessionno. GSE107699).1To whom correspondence may be addressed. Email: wangyc@sustc.edu.cn or guohw@sustc.edu.cn.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1711723115/-/DCSupplemental.

13834–13839 | PNAS | December 26, 2017 | vol. 114 | no. 52 www.pnas.org/cgi/doi/10.1073/pnas.1711723115

Dow

nloa

ded

by g

uest

on

Oct

ober

27,

202

0

acid (ACC) or a loss-of-function mutation in CTR1 leads to muchlonger root hairs (23, 24). In contrast, the completely abolished ETresponse in the ein2 mutant leads to a short-hair phenotype andreduced sensitivity to other hair growth stimuli, such as auxin,strigolactones, and low boron (23, 25, 26).Although the roles of ET in promoting root hair growth and

mediating other hair growth signals are well established, the un-derlying molecular mechanism remains unknown. Our findingsshow that EIN3/EIL1 are necessary and sufficient for ET-inducedroot hair elongation. EIN3 and RHD6 interact and cooperativelypromote the hair elongation factor RSL4. Moreover, the functionsof EIN3/EIL1 and RHD6/RSL1 are coordinated during root hairinitiation. These findings elucidate the molecular mechanism of ETaction during root hair initiation and elongation, and provide in-sight into the coordination of environmental and developmentalsignals during plant organ development.

ResultsEIN3/EIL1 Are Critical for ET-Promoted Root Hair Elongation. Theein3 eil1 mutants were previously found to have reduced root hairdensity and compromised hair growth induction by jasmonic acid,suggesting that EIN3/EIL1 positively regulate root hair develop-ment (27). Phenotypic analysis indicated that EIN3 and EIL1appeared to function partially redundantly in regulating root hairlength (Fig. S1). To investigate the role of EIN3/EIL1 in mediatingET-induced root hair growth, we examined root hair length inyoung Arabidopsis wild-type Col-0 and ein3 eil1 seedlings upon ACCtreatment. In Col-0, a low concentration of ACC (20 nM) stimu-lated significant root hair elongation, with further elongationobserved under a higher concentration (100 nM). The ein3 eil1root hairs were shorter than those of Col-0 and were completelyunresponsive to ACC induction, suggesting that EIN3/EIL1 arenecessary for ET-induced root hair elongation (Fig. 1 A and B).We also tested the sufficiency of EIN3 protein in promotingArabidopsis root hair elongation. In the ein3 eil1 mutant back-ground, root hair length significantly increased with accumu-lated EIN3-FLAG fusion protein induced by 5 nM and 10 nMβ-estrogen application (Fig. 1 C and D). These results show thatEIN3/EIL1 are essential positive regulators mediating ET-promotedroot hair elongation.

ET-Promoted Root Hair Growth Requires RSL4/RSL2. Arabidopsis roothair formation is regulated by a series of transcription factors (2,4). We therefore assessed the interplay between EIN3/EIL1 andthese transcription factors during root hair development. RHD6/

RSL1 and RSL4/RSL2 positively regulate root hair differentiationdownstream of GL2 (7, 12, 23). In fact, there are no visible roothairs in the rhd6 rsl1 and rsl4 rsl2 mutants (9, 12). To determinewhether ET promotes hair growth through these factors, we assessedthe response of these mutants to ACC treatment. In rhd6 rsl1, root

Fig. 1. EIN3/EIL1 mediate ET-induced root hair elon-gation. (A) Representative root hairs from 6-d-oldCol-0 and ein3 eil1 on ACC media. (B) Quantifica-tion of root hair length in A. (C) Representativeroot hairs from 6-d-old iEIN3 ein3 eil1. (D) Quan-tification of root hair length in C. (Scale bars: A andC, 200 μm.) Data are means ± SD (n = 10 roots).One-way ANOVA with a post hoc Tukey honestsignificant difference (HSD) test (**P < 0.01) wasused. NS, not significantly different.

Fig. 2. ET-promoted root hair elongation requires RSL4/RSL2. (A) Representativeroot hairs from 6-d-old Col-0, ein3 eil1, rhd6 rsl1, and rsl4 rsl2 on ACCmedia. (Scalebar: 200 μm.) (B) Quantification of root hair length inA. Data are means ± SD (n =10 roots). One-way ANOVAwith a post hoc Tukey HSD test (**P < 0.01) was used.ND, not detectable; NS, not significantly different.

Feng et al. PNAS | December 26, 2017 | vol. 114 | no. 52 | 13835

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Oct

ober

27,

202

0

hairs grew to about 400 μm in length with ACC treatment (Fig. 2),suggesting rescue of the root hair formation defect. The expressionpattern of RHD6-GFP fusion protein driven by the native RHD6promoter remained largely unchanged during ACC incubation (Fig.S2A), thereby excluding the regulation of RHD6 expression by ET.Furthermore, RHD6 overexpression under the constitutive 35Spromoter in the ein3 eil1 mutant background resulted in longerhair growth, suggesting that RHD6 may function downstream ofor in parallel with EIN3/EIL1 (Fig. S2 B and C).In rsl4 rsl2, the roots remained hairless even with a high ACC

concentration, indicating that RSL4/RSL2 are required for ET-induced hair growth (Fig. 2). RHD6/RSL1 promotes root hair dif-ferentiation by positively regulating the RSL class II genes RSL2–5(11). Given the requirement of RSL4/RSL2 in ET-induced roothair growth, we hypothesized that EIN3/EIL1 might also positivelyregulate RSL class II genes in parallel with RHD6/RSL1.

RSL4 Is a Direct Target of EIN3. To determine whether EIN3/EIL1 regulate RSL genes, we measured RSL transcript levels inein3 eil1 and in transgenic plants with inducible EIN3 over-expression. Exogenous ET treatment increased RSL4 and RSL5transcript levels, but only the increase of RSL4 was completelydependent on EIN3/EIL1 (Fig. 3A). In plants with inducible EIN3overexpression, RSL4 mRNA levels increased as the inductiontime increased (Fig. 3B). RSL5 levels were very low in untreatedplants but dramatically increased after ET treatment and inductionof EIN3 overexpression (Fig. 3 A and B). RSL4 was previouslyreported to positively regulate RSL5 (11); thus, the effect of ETand EIN3 induction on RSL4 transcription likely promoted RSL5expression. In contrast, RSL2 and RSL3 transcript levels were eitherslightly reduced or largely unchanged upon exogenous ET applica-tion or EIN3 induction (Fig. 3 A and B).The changes in RSL4 mRNA abundance indicated that

RSL4 may be a direct target of EIN3. Analysis of the pro-moter region of RSL4 uncovered a putative EIN3-binding site(EBS, 5′-ATGTAT-3′, starting at -853 upstream of the RSL4 gene).In vitro electrophoretic mobility shift assays (EMSAs) confirmed the

specific binding of Escherichia coli-purified EIN3 protein (DNA-binding region, 141–352 aa) to the RSL4 EBS but not to the mu-tated EBS (5′-GGAGCC-3′) (Fig. 3C). Accordingly, EIN3 inductionof RSL4 was significantly impaired in planta when the promoter EBSmotif was mutated (Fig. S3A). We also conducted an in vivo chro-matin immunoprecipitation (ChIP) assay and verified the binding ofEIN3-FLAG fusion protein to the RSL4 EBS-containing region inArabidopsis roots using anti-FLAG antibody to precipitate chromatinfragments (Fig. 3D). Consistently, RSL4 was identified as an ET-responsive EIN3 target gene in a ChIP-sequencing assay usingendogenous anti-EIN3 antibody (28). Furthermore, RSL4 over-expression rescued the short root hair defect of ein3 eil1 (Fig. 3 Eand F). We therefore concluded that RSL4 is a direct target ofEIN3 and that ET promotes root hair elongation through tran-scriptional activation of RSL4 by EIN3/EIL1.

EIN3 and RHD6 Exhibit Protein–Protein Interaction. RHD6 was repor-ted to directly activate RSL4 transcription (12), although evidenceof RHD6 binding to the RSL4 promoter was lacking. We in-vestigated whether EIN3 and RHD6 coordinately regulate RSL4transcription. A yeast two-hybrid assay showed that EIN3 andRHD6 interacted with each other (Fig. S3B). Consistently, anin vitro pull-down assay demonstrated a direct interaction betweenHA-tagged EIN3 protein and His-tagged RHD6 protein (Fig. 4A).A luciferase complementation imaging (LCI) assay was also per-formed in tobacco leaves and Arabidopsis protoplasts, and verifiedthe EIN3–RHD6 interaction in planta (Fig. S3 C–E). EIN3-RFPprotein driven by EIN3 promoter colocalized with RHD6-GFPdriven by RHD6 promoter in H cell files of Arabidopsis roots (Fig.4B). A coimmunoprecipitation assay provided further in vivo evi-dence of the association between EIN3 and RHD6 (Fig. 4C).To determine whether EIN3 and RHD6 act together to regulate

RSL4, the quadruple mutant ein3 eil1 rhd6 rsl1 was generated bycrossing ein3 eil1 and rhd6 rsl1. Notably, due to the absence of twodistinct positive regulators, RSL4 was barely expressed in ein3eil1 rhd6 rsl1 and did not respond to ET treatment (Fig. 4D). Weconducted a dual-luciferase reporter experiment (29) in Arabidopsis

Fig. 3. RSL4 is a direct target of EIN3. (A) Relative(Rel.) expression levels of RSL class II genes in Col-0 and ein3 eil1 (ee) in response to ET. Six-day-oldseedlings were treated with air or ET at 10 ppm for4 h. (B) Rel. expression levels of RSL class II geneswith EIN3 overexpression induction. Six-day-oldiEIN3 ein3 eil1 ebf1 ebf2 (iEqm) seedlings weretransferred to Murashige and Skoog medium con-taining 2 μM β-estrogen for the amount of time in-dicated on the x axis (hours). (C) In vitro EMSAshowing EIN3 directly binding to the RSL4 promoter.(Upper) EIN3-binding site (ATGTAT, boldfaced) andthe mutated form (GAAGCC) are shown. The GST-fused EIN3 N terminus (GST-EIN3141-352) was purifiedfrom E. coli. Unlabeled cold probe (200-fold) wasadded for competition. Probe containing the mu-tated binding site (m) was used to assess bindingspecificity. (D) ChIP-qPCR shows that EIN3 binds tothe RSL4 promoter in vivo. The roots of 6-d-old iEqmwere collected after incubation in 10 nM β-estrogenfor 4 h to induce EIN3-FLAG overexpression. Frag-mented chromatin was precipitated with anti-FLAGantibody. (E and F) RSL4 overexpression restoredhair elongation in ein3 eil1. (E) Representative roothairs from Col-0, ein3 eil1, RSL4ox, and RSL4oxein3 eil1. (Scale bar: 200 μm.) (F) Quantification ofroot hair length in E. Data are means ± SD (n =10 roots). R4ox, RSL4ox. One-way ANOVA with a posthoc Tukey HSD test (**P < 0.01, *P < 0.05) was used.

13836 | www.pnas.org/cgi/doi/10.1073/pnas.1711723115 Feng et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

27,

202

0

root cell protoplasts and found that transient expression ofEIN3 together with RHD6 activated RSL4 transcription more ef-fectively than expression of either EIN3 or RHD6 alone (Fig. 4E).Based on these results, we conclude that EIN3 and RHD6 associatewith each other and coactivate RSL4 transcription.

ET Promotes Root Hair Initiation Through EIN3/EIL1 and RHD6/RSL1.Besides promoting root hair elongation, RHD6/RSL1 are knownregulators of root hair initiation (8, 9). ET also promotes root hairinitiation under hairless conditions (10, 30, 31). We carefully ob-served the surface changes of rhd6 rsl1 and rsl4 rsl2 in response toACC and found that ACC led to bulge formation in both mutants,a marker event of successful hair initiation (Fig. 5 A and B). Whenhigher ACC concentrations were used to treat rhd6 rsl1, thenumber of root hairs and hair length both increased (Fig. 2).Nevertheless, large portions of the root epidermal regions ofrhd6 rsl1 remained hairless (Fig. 5 A and B), suggesting that theeffect of ET on root hair initiation partially depends on the pres-ence of RHD6/RSL1. In rsl4 rsl2, all H positions formed bulgesupon ACC treatment (Fig. 5 A and B). However, the lack of RSL4/RSL2 led to tip-growth failure of the hair bulges, and no length-measurable hair was observed (Figs. 2 and 5). We further examinedhair initiation in ein3 eil1 rhd6 rsl1 and ein3 eil1 rsl4 rsl2. Neitherbulge formation nor hair growth was found even in the presence ofACC, illustrating the importance of EIN3/EIL1 and RHD6/RSL1in ET-induced root hair initiation.

Identification of Genes Coregulated by EIN3/EIL1 and RHD6/RSL1 inRoot Hair Initiation. For genome-wide analysis of EIN3/EIL1 andRHD6/RSL1 interaction, transcriptome profiles of Col-0, ein3 eil1,

rhd6 rsl1, and ein3 eil1 rhd6 rsl1 roots were obtained by RNA se-quencing. A total of 956 differentially expressed genes (DEGs) wereidentified in ein3 eil1 rhd6 rsl1 vs. Col-0. Biological processes in-volved in root hair development were statistically overrepresentedamong the 956 DEGs, including cell wall organization, cell tipgrowth, response to stimulus, and root epidermis differentiation(Fig. S4A). Moreover, the majority of the 956 genes had a greaterfold change in ein3 eil1 rhd6 rsl1 vs. Col-0 than in either doublemutant versus Col-0 (Fig. S4B), suggesting that gene expressioncoregulation by EIN3/EIL1 andRHD6/RSL1 occurs at loci throughoutthe genome and not only at RSL4.Next, we identified 187 genes induced by ET in an EIN3/EIL1-

dependent manner in rhd6 rsl1 (ET-promoted genes). Comparedwith other published genes related to root epidermis morphol-ogy, strikingly, 43 of 154 core H genes (10) were included (Fig.5C), but none of the 54 N genes was found, strongly indicatingthat EIN3/EIL1 and RHD6/RSL1 positively regulate hair for-mation. In rhd6 rsl1, the expression level of 43 H genes inducedby ET was still lower than that in Col-0 without ET. Such anexpression trend was consistent with the finding that ET onlypartially restored the hair growth defect in rhd6 rsl1 (Figs. 2 and5 A and B), suggesting that the coordinated activity of EIN3/EIL1 and RHD6/RSL1 is needed to fully activate root hair initia-tion. In light of the coactivation of the hair elongation factor RSL4by EIN3/EIL1 and RHD6/RSL1, we found 25 of the 43 H genes tobe RSL4-regulated genes (12, 32) (Fig. 5C). The remaining 18 genesnot influenced by RSL4 (Fig. 5C) were considered to be potentialinitiation stage factors. Furthermore, coexpression analysis of the

Fig. 4. EIN3 physically interacts with RHD6. (A) Pull-down analysis shows directinteraction between EIN3 and RHD6 in vitro. EIN3-HA and His-TF-RHD6 (TF-RHD6) were incubated for 4 h before precipitation by nickel-nitrilotriacetic acidagarose. TF, trigger factor. (B) Colocalization of EIN3 and RHD6 in H cells (whitearrows). Six-day-old F1 progeny of pRHD6::GFP:RHD6 × pEIN3::EIN3:RFP weretreated with ET for 2 h. (C) Coimmunoprecipitation of EIN3 and RHD6. Whole-protein extraction from root tips of pRHD6::GFP:RHD6was immunoprecipitatedby GFP-trap agarose, and endogenous anti-EIN3 antibody was used for blotting.IP, immunoprecipitation. (D) Relative (Rel.) expression level of RSL4 in ein3 eil1rhd6 rsl1 (ee61). Seedlings were treated with air or ET for 4 h. (E) Dual-luciferasereporter assay shows the coactivation of RSL4 by EIN3 and RHD6. The indicatedvectors were transformed into Arabidopsis root protoplasts. The activity ratio offirefly luciferase relative to Renilla luciferase was calculated as a metric of tran-scriptional activity. Error bars indicate ±SD of three biological replicates. One-wayANOVA with a post hoc Tukey HSD test (**P < 0.01, *P < 0.05) was used.

Fig. 5. ET promotes root hair initiation through EIN3/EIL1 and RHD6/RSL1.(A) Representative roots show hair bulge formation induced by 2 μM ACC in6-d-old hairless mutants. (Scale bar: 200 μm.) (B) Root hair/bulge percentage inH cells. The percentage was calculated based on bulge and hair numbers perH cell. Two hundred cells per sample were counted. ee, ein3 eil1; ee42,ein3 eil1 rsl4 rsl2; ee61, ein3 eil1 rhd6 rsl1; 42, rsl4 rsl2; 61, rhd6 rsl1; ND, notdetectable; NS, not significantly different. Error bars indicate ±SD of three bi-ological replicates. Statistical significance was determined by a Student’s t test(***P < 0.001). (C) Summary diagram for the identification of root hair initia-tion genes. RNA sequencing and differential expression analysis were per-formed using roots of Col-0, ee, 61, and ee61 after air or ET treatment. Amongthe 43 H genes, 25 were identified as RSL4-regulated genes (12, 32). Nine genes,not characterized in previous publications, are closely related to the known roothair genes based on coexpression analysis (Fig. S5). These nine genes plus18 non–RSL4-regulated H genes comprise the group of 27 hair initiation can-didate genes. EPGs, ET-promoted genes.

Feng et al. PNAS | December 26, 2017 | vol. 114 | no. 52 | 13837

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Oct

ober

27,

202

0

187 genes revealed compact clustering of previously publishedroot hair genes, including core root epidermal genes (10), RSL4-dependent genes (12, 32), and H cell-enriched genes (33) (Fig.S5). Although nine genes within the compact cluster were notfound in previous studies, they may nevertheless participate in roothair development considering the similarity of their expressionpatterns to those of other root hair genes. These nine genes andthe 18 non–RSL4-regulated H genes comprised a candidate poolof 27 putative downstream target genes of RHD6/RSL1 and EIN3/EIL1 involved in root hair initiation (Table S1). More analysisdetails are provided in SI Identification of Genes Coregulated byEIN3/EIL1 and RHD6/RSL1 in Root Hair Initiation.

EIN3/EIL1 and RHD6/RSL1 Mediate Diverse Root Hair Stimuli. Tran-scriptome profiling revealed a more general role of EIN3/EIL1 andRHD6/RSL1 as gene expression coregulators during root hairformation. RSL4, the target gene coactivated by EIN3 and RHD6,is required for several root hair-inducing signals, including nutrientdeficiency and hormone application (12, 14, 15). We thereforeassessed whether EIN3/EIL1 and RHD6/RSL1 mediate the effectsof these stimuli. Four root hair stimuli, application of auxin orcytokinin and depletion of phosphorus or nitrogen, were used toassess the response in rhd6 rsl1, ein3 eil1, and ein3 eil1 rhd6 rsl1.The rhd6 rsl1 was responsive to both hormone treatments but notto nutrient deficiency (Fig. 6 A–D and Fig. S7), suggesting thatRHD6/RSL1 are essential for mediating the effects of nutrientdepletion but not hormone application. The ein3 eil1 was re-sponsive to all four stimuli, although root hair length in the mutantwas relatively shorter than that of wild type under all conditions(Fig. 6 A–D). The ein3 eil1 rhd6 rsl1 quadruple mutant was in-sensitive to all four stimuli and had no visible root hairs (Fig. 6 A–Dand Fig. S7), highlighting the importance of EIN3/EIL1 andRHD6/RSL1 coordination in response to these treatments.

DiscussionResearch over the past few decades has clearly established theimportance of ET in regulating plant root hair development.However, the underlying molecular mechanisms were poorly un-derstood. In this study, we found that ET promotes root hair for-mation at both the initiation and elongation stages. Moreover, themaster regulators EIN3/EIL1 mediate the effects of ET at bothstages. EIN3 directly binds the promoter region of the hair length-determining gene RSL4 and activates its transcription to promoteroot hair elongation. EIN3 also physically interacts with RHD6,another essential regulator of root hair development upstream ofRSL4. EIN3/EIL1 and RHD6/RSL1 function in parallel and syn-ergistically as positive regulators of root hair initiation and elon-gation. Based on these findings, we propose the following model forET-induced root hair growth. In wild-type roots where EIN3/EIL1 levels are low, RHD6/RSL1 are mainly responsible for theinduction of RSL4 and root hair initiation genes to maintain normalroot hair growth. Upon ET treatment, EIN3/EIL1 accumulate andcomplex with RHD6/RSL1 to synergistically activate the expressionof hair initiation genes as well as RSL4 to potently increase hairgrowth. In rhd6 rsl1, ET-induced EIN3/EIL1 act independently topartially rescue the hair initiation and elongation defects (Fig. 6E).ET, a widely documented stress hormone, helps plants adapt

to various environmental challenges. In the proposed model, theET signal is integrated with the internal root hair developmentpathway, with EIN3/EIL1 conferring stress responsiveness to theEIN3–RHD6 transcription complex. Meanwhile, due to its strictexpression pattern in H cells, RHD6 confers spatiotemporal speci-ficity to the complex. In this way, different internal and externalsignals converge at key nodes through associated transcription fac-tors that provide flexibility and adaptability to ever-changing envi-ronments. Consistent with this hypothesis, the simultaneous lossof both EIN3/EIL1 and RHD6/RSL1 led to virtual insensitivity tovarious root hair-inducing signals, underscoring the central role ofthe EIN3/EIL1–RHD6/RSL1 transcription complex in signaling in-tegration. Notably, ET signaling is known to act upstream of auxinbiosynthesis in the root tip on cell elongation and polar root hair

initiation (34–38). The ein3 eil1 rhd6 rsl1 remains hairless in thepresence of auxin applications (Fig. 6A and Fig. S7 A and B), re-vealing that ET does not promote root hair initiation and elongationsimply through increasing auxin biosynthesis. Other effects on pro-moting hair formation by ET–auxin interplay, including trans-portation and signaling, still need further investigation. Similar toEIN3, ARF5 is also a direct regulator of RSL4 in mediating auxin-promoted polar growth (14). Combining findings in this study, RSL4promoter is suggested to be a direct, central point of convergence forauxin signaling via ARF5 and for ET signaling via EIN3, and that itis regulated by the major hair cell differentiation regulator RHD6.In addition to RSL4, EIN3/EIL1 and RHD6/RSL1 coregulate a

subset of genes that likely contribute to root hair initiation andelongation. Of the 27 candidate root hair initiation genes identified

Fig. 6. Relative root hair length of Col-0, ein3 eil1(ee), rhd6 rsl1(61), andein3 eil1 rhd6 rsl1(ee61) under the following stimuli: auxin [indole-3-acetic acid(IAA)] (A), cytokinin [6-benzylaminopurine (6-BA)] (B), nitrogen (N) deficiency(C), and phosphorus (P) deficiency (D). The length of Col-0 root hair was cali-brated as 1.0. Data are means ± SD (n = 10 roots). The values were comparedbetween treated and untreated samples within the same genotype. Statisticalsignificance was determined by a Student’s t test (***P < 0.001). ND, not de-tectable. (E) Working model. (Top Right) In the presence of ET in Col-0, highlevels of ET-induced EIN3/EIL1 protein accumulate (E3; red ovals). EIN3/EIL1 interact with RHD6/RSL1 (R6; yellow hexagons) in H cells to directly activatethe transcription of RSL4, an essential positive regulator of root hair tip growth,as well as a set of genes involved in hair initiation. (Top Left) In the absence ofET, root hair initiation and elongation are mainly regulated by R6, withE3 playing a minor role. (Bottom Right) In the rhd6 rsl1mutant, ET-induced andstabilized E3 independently promotes hair initiation and elongation to partiallyrestore hair growth of the hairless mutant. (Bottom Left) In the absence of bothET-induced E3 and functional R6, the hairless phenotype is observed.

13838 | www.pnas.org/cgi/doi/10.1073/pnas.1711723115 Feng et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

27,

202

0

in this study, some have been reported to participate in root hairformation. For example, overexpression of ROOT HAIR SPECIFIC3 (RSH3) leads to spiral, bent, and branched hair morphologies(39). LEUCINE-RICH REPEAT/EXTENSIN 1 (LRX1) encodesa chimeric leucine-rich repeat/extensin protein, and the lrx1 mutantexhibits aberrant root hair formation, including aborted, swollen,and branched hairs (40). Lotus japonicus (Lj)RHL1-LIKE 3(LRL3), encoding a bHLH subfamily XI protein, rescues thehairless defect of Ljrhl1 mutants (41).As master regulators, EIN3/EIL1 associate with a group of tran-

scriptional regulators in hormone cross-talk. All EIN3-associatedproteins reported thus far repress its biological function (27, 42, 43).Intriguingly, this study uncovered a class of EIN3-associated tran-scription regulators that enhance its activity, although how this en-hancement is achieved is unclear. Several possibilities can beconsidered. First, RHD6 may act as a positive regulator thatenhances EIN3 transcription activity. Although RHD6 directlyregulates RSL4 expression, no DNA-binding ability has beendemonstrated. RHD6 may be recruited by other DNA-bindingfactors, such as EIN3, to the RSL4 promoter. Second, RHD6 maydirectly bind to a specific DNA sequence in the RSL4 promoter,while EIN3 binds to the EBS motif. In turn, the association betweenthe two classes of transcription factors could mutually and greatly

enhance their respective DNA-binding ability. Third, RHD6 mayadopt a de-repression mechanism by competing with EIN3-associ-ating repressors in H cells, and RHD6 interaction could releaseEIN3 from an otherwise repressed state. Further investigation ofthese alternative mechanisms is needed to fully understand the ac-tion and importance of the EIN3–RHD6 complex.

Materials and MethodsPlant Materials. The A. thaliana mutant ein3 eil1 and transgenic plantiEIN3 ein3 eil1 ebf1 ebf2 and iEIN3 ein3 eil1 were described previously (43, 44).The rhd6 rsl1 and rsl4 rsl2 double mutants and transgenic pRHD6::GFP:RHD6and RSL4ox plants were gifts from Liam Dolan, University of Oxford, Oxford,United Kingdom. Details about plant growth conditions and treatments aredescribed in SI Materials and Methods. Plant transformation, root hair lengthmeasurement, gene expression, ChIP-qPCR, EMSA, yeast two-hybrid assay, pull-down analysis, LCI, dual-luciferase reporter assay, and sequencing analysis werecarried out according to protocols described in SI Materials and Methods.

ACKNOWLEDGMENTS. We thank Liam Dolan (University of Oxford) forsharing several plant materials. This work was supported by the NationalNatural Science Foundation of China (Grant 91017010), the Peking-TsinghuaCenter for Life Sciences, and start-up funding from the Southern University ofScience and Technology (to H.G.).

1. Giehl RF, von Wirén N (2014) Root nutrient foraging. Plant Physiol 166:509–517.2. Grierson C, Nielsen E, Ketelaarc T, Schiefelbein J (2014) Root hairs.Arabidopsis Book 12:e0172.3. Dolan L, et al. (1993) Cellular organisation of the Arabidopsis thaliana root.

Development 119:71–84.4. Schiefelbein J, Huang L, Zheng X (2014) Regulation of epidermal cell fate in Arabi-

dopsis roots: The importance of multiple feedback loops. Front Plant Sci 5:47.5. Di Cristina M, et al. (1996) The Arabidopsis Athb-10 (GLABRA2) is an HD-Zip protein

required for regulation of root hair development. Plant J 10:393–402.6. Masucci JD, et al. (1996) The homeobox gene GLABRA2 is required for position-

dependent cell differentiation in the root epidermis of Arabidopsis thaliana.

Development 122:1253–1260.7. Lin Q, et al. (2015) GLABRA2 directly suppresses basic helix-loop-helix transcription factor

genes with diverse functions in root hair development. Plant Cell 27:2894–2906.8. Masucci JD, Schiefelbein JW (1994) The rhd6mutation ofArabidopsis thaliana alters root-hair

initiation through an auxin- and ethylene-associated process. Plant Physiol 106:1335–1346.9. Menand B, et al. (2007) An ancient mechanism controls the development of cells with

a rooting function in land plants. Science 316:1477–1480.10. Bruex A, et al. (2012) A gene regulatory network for root epidermis cell differenti-

ation in Arabidopsis. PLoS Genet 8:e1002446.11. Pires ND, et al. (2013) Recruitment and remodeling of an ancient gene regulatory

network during land plant evolution. Proc Natl Acad Sci USA 110:9571–9576.12. Yi K, Menand B, Bell E, Dolan L (2010) A basic helix-loop-helix transcription factor

controls cell growth and size in root hairs. Nat Genet 42:264–267.13. Datta S, Prescott H, Dolan L (2015) Intensity of a pulse of RSL4 transcription factor

synthesis determines Arabidopsis root hair cell size. Nat Plants 1:15138.14. Mangano S, et al. (2017) Molecular link between auxin and ROS-mediated polar

growth. Proc Natl Acad Sci USA 114:5289–5294.15. Zhang S, et al. (2016) Multiple phytohormones promote root hair elongation by regu-

lating a similar set of genes in the root epidermis in Arabidopsis. J Exp Bot 67:6363–6372.16. Johnson PR, Ecker JR (1998) The ethylene gas signal transduction pathway: A mo-

lecular perspective. Annu Rev Genet 32:227–254.17. Ju C, Chang C (2015) Mechanistic insights in ethylene perception and signal trans-

duction. Plant Physiol 169:85–95.18. Ju C, et al. (2012) CTR1 phosphorylates the central regulator EIN2 to control ethylene

hormone signaling from the ER membrane to the nucleus in Arabidopsis. Proc Natl

Acad Sci USA 109:19486–19491.19. Li W, et al. (2015) EIN2-directed translational regulation of ethylene signaling in

Arabidopsis. Cell 163:670–683.20. Merchante C, et al. (2015) Gene-specific translation regulation mediated by the

hormone-signaling molecule EIN2. Cell 163:684–697.21. Potuschak T, et al. (2003) EIN3-dependent regulation of plant ethylene hormone

signaling by two Arabidopsis F box proteins: EBF1 and EBF2. Cell 115:679–689.22. Guo H, Ecker JR (2003) Plant responses to ethylene gas are mediated by SCF(EBF1/

EBF2)-dependent proteolysis of EIN3 transcription factor. Cell 115:667–677.23. Masucci JD, Schiefelbein JW (1996) Hormones act downstream of TTG and GL2 to

promote root hair outgrowth during epidermis development in the Arabidopsis root.

Plant Cell 8:1505–1517.24. Pitts RJ, Cernac A, Estelle M (1998) Auxin and ethylene promote root hair elongation

in Arabidopsis. Plant J 16:553–560.25. Kapulnik Y, et al. (2011) Strigolactones interact with ethylene and auxin in regulating

root-hair elongation in Arabidopsis. J Exp Bot 62:2915–2924.26. Martín-Rejano EM, et al. (2011) Auxin and ethylene are involved in the responses of root

system architecture to low boron supply inArabidopsis seedlings. Physiol Plant 142:170–178.

27. Zhu Z, et al. (2011) Derepression of ethylene-stabilized transcription factors (EIN3/EIL1) mediates jasmonate and ethylene signaling synergy in Arabidopsis. Proc NatlAcad Sci USA 108:12539–12544.

28. Chang KN, et al. (2013) Temporal transcriptional response to ethylene gas drivesgrowth hormone cross-regulation in Arabidopsis. Elife 2:e00675.

29. Hellens RP, et al. (2005) Transient expression vectors for functional genomics, quan-tification of promoter activity and RNA silencing in plants. Plant Methods 1:13.

30. Cao XF, Linstead P, Berger F, Kieber J, Dolan L (1999) Differential ethylene sensitivityof epidermal cells is involved in the establishment of cell pattern in the Arabidopsisroot. Physiol Plant 106:311–317.

31. Cho HT, Cosgrove DJ (2002) Regulation of root hair initiation and expansin geneexpression in Arabidopsis. Plant Cell 14:3237–3253.

32. Vijayakumar P, Datta S, Dolan L (2016) ROOT HAIR DEFECTIVE SIX-LIKE4 (RSL4) pro-motes root hair elongation by transcriptionally regulating the expression of genesrequired for cell growth. New Phytol 212:944–953.

33. Brady SM, et al. (2007) A high-resolution root spatiotemporal map reveals dominantexpression patterns. Science 318:801–806.

34. Stepanova AN, Hoyt JM, Hamilton AA, Alonso JM (2005) A link between ethylene andauxin uncovered by the characterization of two root-specific ethylene-insensitivemutants in Arabidopsis. Plant Cell 17:2230–2242.

35. R�uzicka K, et al. (2007) Ethylene regulates root growth through effects on auxinbiosynthesis and transport-dependent auxin distribution. Plant Cell 19:2197–2212.

36. Swarup R, et al. (2007) Ethylene upregulates auxin biosynthesis in Arabidopsis seed-lings to enhance inhibition of root cell elongation. Plant Cell 19:2186–2196.

37. Stepanova AN, et al. (2008) TAA1-mediated auxin biosynthesis is essential for hor-mone crosstalk and plant development. Cell 133:177–191.

38. Ikeda Y, et al. (2009) Local auxin biosynthesis modulates gradient-directed planarpolarity in Arabidopsis. Nat Cell Biol 11:731–738.

39. Won SK, et al. (2009) Cis-element- and transcriptome-based screening of root hair-specificgenes and their functional characterization in Arabidopsis. Plant Physiol 150:1459–1473.

40. Baumberger N, Ringli C, Keller B (2001) The chimeric leucine-rich repeat/extensin cellwall protein LRX1 is required for root hair morphogenesis in Arabidopsis thaliana.Genes Dev 15:1128–1139.

41. Karas B, et al. (2009) Conservation of lotus and Arabidopsis basic helix-loop-helixproteins reveals new players in root hair development. Plant Physiol 151:1175–1185.

42. Zhang X, et al. (2014) Jasmonate-activatedMYC2 represses ETHYLENE INSENSITIVE3 activity toantagonize ethylene-promoted apical hook formation inArabidopsis. Plant Cell 26:1105–1117.

43. An F, et al. (2012) Coordinated regulation of apical hook development by gibberellinsand ethylene in etiolated Arabidopsis seedlings. Cell Res 22:915–927.

44. An F, et al. (2010) Ethylene-induced stabilization of ETHYLENE INSENSITIVE3 andEIN3-LIKE1 is mediated by proteasomal degradation of EIN3 binding F-box 1 and2 that requires EIN2 in Arabidopsis. Plant Cell 22:2384–2401.

45. Gendrel AV, Lippman Z, Martienssen R, Colot V (2005) Profiling histone modificationpatterns in plants using genomic tiling microarrays. Nat Methods 2:213–218.

46. Bolger AM, Lohse M, Usadel B (2014) Trimmomatic: A flexible trimmer for Illuminasequence data. Bioinformatics 30:2114–2120.

47. Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheelertransform. Bioinformatics 25:1754–1760.

48. Robinson MD, McCarthy DJ, Smyth GK (2010) edgeR: A Bioconductor package for dif-ferential expression analysis of digital gene expression data. Bioinformatics 26:139–140.

49. Maere S, Heymans K, Kuiper M (2005) BiNGO: A Cytoscape plugin to assess over-representation of gene ontology categories in biological networks. Bioinformatics 21:3448–3449.

50. Ma C, Wang X (2012) Application of the Gini correlation coefficient to infer regula-tory relationships in transcriptome analysis. Plant Physiol 160:192–203.

Feng et al. PNAS | December 26, 2017 | vol. 114 | no. 52 | 13839

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Oct

ober

27,

202

0