Ethylene Regulates Differential Growth via BIGARF-GEF-Dependent Post-Golgi SecretoryTrafficking in ArabidopsisOPEN

Kristoffer Jonsson,a Yohann Boutté,b Rajesh Kumar Singh,a Delphine Gendre,a and Rishikesh P. Bhaleraoa,1

a Umeå Plant Science Centre, Department of Forest Genetics and Plant Physiology, Swedish University for Agricultural Sciences,SE-901 83 Umeå, SwedenbCNRS-University of Bordeaux, UMR 5200 Membrane Biogenesis Laboratory, INRA Bordeaux Aquitaine, 33140 Villenave d’Ornon,France

ORCID ID: 0000-0003-4384-7036 (R.P.B.)

During early seedling development, the shoot apical meristem is protected from damage as the seedling emerges from soil bythe formation of apical hook. Hook formation requires differential growth across the epidermis below the meristem in thehypocotyl. The plant hormones ethylene and auxin play key roles during apical hook development by controlling differentialgrowth. We provide genetic and cell biological evidence for the role of ADP-ribosylation factor 1 (ARF1)-GTPase and itseffector ARF-guanine-exchange factors (GEFs) of the Brefeldin A-inhibited GEF (BIG) family and GNOM in ethylene- andauxin-mediated control of hook development. We show that ARF-GEF GNOM acts early, whereas BIG ARF-GEFs act at a laterstage of apical hook development. We show that the localization of ARF1 and BIG4 at the trans-Golgi network (TGN) dependson ECHIDNA (ECH), a plant homolog of yeast Triacylglycerol lipase (TLG2/SYP4) interacting protein Tgl2-Vesicle Protein23 (TVP23). BIGs together with ECH and ARF1 mediate the secretion of AUX1 influx carrier to the plasma membrane from theTGN during hook development and defects in BIG or ARF1 result in insensitivity to ethylene. Thus, our data indicate a divisionof labor within the ARF-GEF family in mediating differential growth with GNOM acting during the formation phase whereasBIGs act during the hook maintenance phase downstream of plant hormone ethylene.

INTRODUCTION

The apical hook develops immediately after seed germination,protecting the apical meristem as the etiolated seedling pene-trates through soil. The curvature of the apical hook results fromasymmetric cell elongation rates at opposite sides of the bendinghypocotyl (Silk and Erickson, 1978; Raz and Koornneef, 2001),whicharecoordinatedby theplant hormonesethyleneand indole-3-acetic acid (IAA/auxin) (Li et al., 2004; Gallego-Bartolome et al.,2011). Ethylene promotes hook curvature largely by modulatingauxin biosynthesis anddistribution (Harper et al., 2000; Zádníkováet al., 2010). Bending during hook development requires asym-metric distribution of auxin across the hypocotyl. The asymmetricauxin distribution with auxin response maxima localized on theconcave side of the hook can be visualized using synthetic auxinresponse reporters such as DR5 (Vandenbussche et al., 2010;Zádníková et al., 2010). An increased auxin response on theconcave side causes a reduced cell elongation rate relative to theconvex side, resulting in hook formation. The establishment ofauxin response maxima on the concave side of the apical hookrelies on active, carrier-mediated polar auxin transport (PAT)(Vandenbussche et al., 2010; Zádníková et al., 2010). Members of

both the AUX/LIKE-AUX1(LAX) influx carrier and PIN-FORMED(PIN) efflux carrier families play key roles in PAT-mediated hookdevelopment, underlined by the hook defects observed in aux/laxand pinmutants, or by exposing seedlings to the carrier inhibitorsN-(1-naphtyl)phtalamic acid or 1-naphtoxyacetic acid that disruptPAT (Vandenbussche et al., 2010; Zádníková et al., 2010). Theabundance and localization of PAT-mediating carriers is modu-lated by membrane trafficking processes. Whereas studies of thePIN family of auxin carriers have provided insights into how en-docytosis facilitates their localization and abundance at theplasma membrane (PM) (Geldner et al., 2003; Dhonukshe et al.,2007), much less is known about the machinery involved in AUX/LAX protein trafficking, which also play an important role duringhook development as suggested by the mutant phenotypes(Vandenbussche et al., 2010).We have identified a key role for the evolutionarily conserved

protein ECHIDNA (ECH), a plant homolog of the yeast Tri-acylglycerol lipase (TLG2/SYP4) interacting protein Tgl2-VesicleProtein 23 (TVP23), in secretory trafficking at the post-Golgicompartment trans-Golgi network (TGN) of AUX1 in the apicalhook (Gendre et al., 2011; Boutté et al., 2013). The ech mutantdisplays severe attenuation of the secretory trafficking of AUX1 tothe PM, altered auxin distribution, insensitivity to the ethyleneprecursor aminocyclopropane-1-carboxylate (ACC), and hookdevelopmental defects. Nevertheless, how ECH mediates se-cretory trafficking of AUX1 remains poorly understood. BIGproteins, which act as a guanine-exchange factor (GEF) for ADP-ribosylation factor (ARF) GTPases, are required for de novo PIN1delivery to the PM (Richter et al., 2014). BIGs, like ECH, localize to

1 Address correspondence to rishi.bhalerao@slu.se.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Rishi Bhalerao (rishi.bhalerao@slu.se).OPENArticles can be viewed without a subscription.www.plantcell.org/cgi/doi/10.1105/tpc.16.00743

The Plant Cell, Vol. 29: 1039–1052, May 2017, www.plantcell.org ã 2017 ASPB.

the TGN, a post-Golgi compartment that serves as a hub forsecretory and endocytic traffic. ARF-GEFs primarily catalyzenucleotide exchange for ARF GTPases, which mediate vesicleformation (Peyroche et al., 1996). Interestingly, the number ofsecretory vesiclesat theTGNhasbeenshown tobe reduced in theech mutant, implicating ECH in a hitherto unknown aspect ofvesicle development (Boutté et al., 2013). The localization of ECHand BIGs at the TGN and the secretory defects observed for echand big mutants prompted us to investigate whether BIGs likeECH were involved in hook development and whether they areinvolved in AUX1 trafficking in concert with ECH. Intriguingly,previous studies have hinted at a role for an ARF-GEF that isresistant to the action of the fungal toxin Brefeldin A (BFA) in AUX1trafficking (Kleine-Vehn et al., 2006). BIG3, a member of the BIGfamily is resistant to BFA further supporting the investigationinto the role of BIGs in AUX1 trafficking via TGN. Therefore, weinitiated thecharacterizationofBIGsand their targetARF1 inAUX1trafficking.

Our data demonstrate that BIG4 colocalizes with ECH andARF1 at the TGN and like ECH, BIG function is essential forARF1 localization at the TGN. Disruption of ARF1 or BIG functionphenocopies the AUX1 trafficking and hook developmental de-fects observed in ech. Moreover, big mutants, like ech, exhibitresistance to exaggeration of the apical hook by ethylene pre-cursor ACC. Taken together, this work provides insight into themechanism of ARF-GEF-mediated, ECH-dependent secretorytrafficking at the TGN in hormonal control of differential growth.

RESULTS

The ARF-GEFs GNOM and BIG1-4 Act at Distinct Stages ofApical Hook Development

Following germination, wild-type seedlings progressively bendtheir hypocotyls to formanapical hook in thedark, aprocess that iscomplete 24 to 36hpostgermination (designated as the formationphase). The apical hook is maintained for ;48 h (maintenancephase), followed by a gradual opening over 72 to 96 h post-germination (opening phase), after which the hypocotyl becomescompletely straight.We employed a genetic andpharmacologicalapproach to investigate the role of ARF-GEFs in apical hookdevelopment, using the fungal toxin BFA, which inhibits the ac-tivity of ARF-GEFs (Peyroche et al., 1999). When wild-type seedswere germinated on media supplemented with 5 mM BFA, apicalhook development was completely abolished (Figure 1A), in-dicating a role for ARF-GEFs in this process. The Arabidopsisthaliana genome harbors eight ARF-GEFs, divided into twosubclasses, one comprising GNOM and GNL1-2 and the otherBIG1-5, with BIG1-4 regulating processes distinct from BIG5(Richter et al., 2014). Next, we took a genetic approach to obtaina better understanding of the contribution of distinct ARF-GEFs tohook development. We used Arabidopsis seedlings expressinga BFA-resistant variant of GNOM (GNOMR) (Geldner et al., 2003)and analyzed the effect of BFA on hook development in theseseedlings (Figure 1A). In contrast to the wild type, the introductionGNOMR into the wild-type background completely restored hookformation, even in thepresenceof 5mMBFA (Figure 1A).However,

GNOMR seedlings were unable to maintain the apical hook in thepresence of 5 mM BFA compared with seedlings grown withoutBFA (Figure 1A). We observed that GFP-tagged GNOM driven byits native promoterwas expressed uniformly at equal levels duringboth hook formation and maintenance (Supplemental Figure 1).Thus, although expressed ubiquitously during distinct stages ofhook development, GNOM acts primarily during the hook for-mation phase, whereas other ARF-GEFs may be essential duringthe subsequent stages of hook development.ARF-GEFs BIG1-4 were recently shown to act redundantly and

independently of GNOMduring lateral root initiation (Richter et al.,2014). Therefore, we analyzed whether BIGs play a role in hookdevelopment. While big1-big4 single mutants, as well as variousdouble and triple mutant combinations, displayed no discernible

Figure 1. BIG1-4 and GNOM Operate during Distinct Phases of ApicalHook Development, and BIG1-4 Are Required for Hook Maintenance.

(A)Upon treatment with 5 mMBFA, hook formation is disrupted in the wildtype, while introduction of GNOMR in the wild-type background restoreshook formation but not maintenance.(B) big1 big2 big3 and big1/+ big2 big3 big4 exhibit a reduced apical hookmaintenance phase compared with the wild type.(C) While under mock conditions, big3 displays hook development in-discernible from that of the wild type, and 0.5 mM BFA treatment disruptsapical hook development of thebig3mutant but not thewild type. For eachgenotype, n $ 16. Error bars represent SE of the mean.

1040 The Plant Cell

hook defects (Supplemental Figures 2A and 2B), big1 big2 big3and big1/+ big2 big3 big4 mutants exhibited an apical hook de-velopment with amarkedly shortenedmaintenance phase (Figure1B). This indicates that BIGARF-GEFs are not essential during theformation phase but act redundantly during the maintenancephase. A quadruple big mutant is not viable and could thereforenot be analyzed (Richter et al., 2014). Of the BIG ARF-GEFs, onlyBIG3 is resistant to BFA (Richter et al., 2014). Thus, treating thebig3 mutant with BFA allows simultaneous inhibition of BIG1-4function as well as GNOM. While hook formation of wild-typeplants remained unaffected in the presence of 0.5 mM BFA, big3mutant treated with 0.5 mM BFA displayed severely perturbedapical hook development, with hook angle peaking at ;150° in-stead of 180° as in the wild type and with a completely abolishedmaintenance phase (Figure 1C). The hook defect in the BFA-treated big3 plants could be rescued with the introduction of anengineered, BFA-resistant variant of BIG4-YFP (ProUBQ10:BIG4R-YFP) into the big3 background, but not with the in-troduction of a BFA-sensitive BIG4-YFP (ProUBQ10:BIG4-YFP)(Supplemental Figure 2C). These results further illustrate thefunctional redundancy among BIG1-4 family members duringhook development. It has been previously shown that GNOM cantake over the function of theBFA-resistantGNL1 in the absenceofGNL1 (Richter et al., 2007). Therefore, we tested whether GNOMcan similarly take over the function of BIGs in hook development.However, the introduction of GNOMR into big3 did not reverse thesevere hook defect of BFA-treated big3 (unlike BFA-resistantBIG4) (Supplemental Figure 2D), indicating that GNOM cannotsubstitute for BIG1-4. Taken together, our results show that ARF-GEFs BIGs and GNOM are required at distinct stages of hookdevelopment.

BIGs Act Redundantly in Ethylene-MediatedHook Development

Previous studies have demonstrated that the apical hook main-tenance phase is regulated by ethylene (Raz and Koornneef,2001), and exogenously added ethylene or ethylene precursorACC induces exaggeration of the apical hook (Kang et al., 1967;Bleecker et al., 1988). Therefore,we investigatedwhetherBIGsareinvolved in ethylene-mediated control of hook development.Hence, we examined whether the hook maintenance defect ofBFA-treated big3 could be suppressed by ACC. While wild-typeseedlings exhibited anexaggeratedhook curvature in response totreatmentwith10mMACCboth in thepresenceandabsenceof0.5mM BFA (Figure 2A), reaching an average angle of 270°, the big3mutants did not exaggerate their hook curvature in response to10 mMACC treatment when grown in the presence of 0.5 mMBFA(Figure 2B). This indicates that the BIG ARF-GEFs are required,andact independently ofGNOM,duringethylene-mediatedapicalhook maintenance.

BIGs Interact Genetically with ECH in Hook Development

The hook defects in big mutants, e.g., attenuated maintenanceand insensitivity of hook development to ethylene, are similar tothose described in the ech mutant (Boutté et al., 2013). Further-more, like ECH, BIG3 localizes to the TGN (Richter et al., 2014).

Therefore, we examined whether BIG ARF-GEFs and ECH mightact in acommonpathwayduringhookdevelopment.Weobservedthat hook development of the ech mutant displayed a stronghypersensitivity to BFA treatment, implying that ARF-GEF func-tion may be compromised in the ech mutant (Figure 3A). Fur-thermore, an ech big2 big3 triple mutant exhibited enhanceddefects in hook development when compared with ech and big2big3 double mutants alone (Figure 3B), suggesting that ECH andBIG1-4 act synergistically in hook development.

AUX1 Trafficking to the PM Requires BIG1-BIG4 Function

Wehavepreviously demonstrated that the level of AUX1at thePMis strongly reduced in the ech mutant (Boutté et al., 2013). Fur-thermore, theaux1mutantdisplays insensitivity to ethyleneduringhook development (Vandenbussche et al., 2010), as do the echand bigmutants. This led us to investigate if AUX1 delivery to thePM requires BIG1-4 function. In contrast to mock-treated wildtype and big3, or wild-type seedlings grown in the presence of0.5 mM BFA, AUX1-YFP levels at the PM decreased to 43% ofmock levels during the formation phase and 11% of mock levelsduring the maintenance phase in big3mutant grown onmediumsupplemented with 0.5 mM BFA (Figure 4). Compared with thestrong reduction of AUX1-YFP plasma membrane levels duringthe maintenance phase in big3 seedlings grown on 0.5 mMBFA,the transcript levels of AUX1-YFP were only mildly reduced inbig3 grown on 0.5 mM BFA during the maintenance phase(Supplemental Figure 3), indicating that the observed attenua-tion of PM localization of AUX1 in BFA-treated big3 is primarilyposttranscriptional.It has been suggested that AUX1 trafficking may rely on a BFA-

resistant ARF-GEF (Kleine-Vehn et al., 2006). In agreement withthis observation, no intracellular agglomerations of AUX1-YFPwere observed in wild-type epidermal cells in the hook that weremock-treated, subjected to a 3-h treatment with 50 mM BFA, oralternatively pretreated with 50 mM cycloheximide (CHX) for 1 hprior to a 3-h treatment with 50 mMCHX together with 50 mMBFA(Figures 5A to5C).While AUX1-YFP localizationwasunaffected inmock-treated big3 mutant (Figure 5D) like the wild type, big3mutants treated with 50 mM BFA for 3 h displayed strong in-tracellular agglomerations of AUX1-YFP (Figure 5E). Importantly,this formation of AUX1-YFP agglomerations in big3 upon BFAtreatment was blocked when the big3 seedlings were pretreatedwithCHXat50mMfor1hprior to a3-h treatmentwith50mMCHX+50 mM BFA (Figure 5F). Additionally, the intracellular AUX1-YFPagglomerations in BFA-treated big3 seedlings strongly over-lapped with the TGN-localized VHA-a1-RFP (Dettmer et al.,2006) (Pearson’s coefficient, 0.4296 0.083) (Figures 5G to 5L;Supplemental Figure 4A). By contrast, pretreatmentwith 50 mMCHX for 1 h followed by a 3-h treatment with 50 mMCHX + 50 mMBFA strongly agglomerated VHA-a1-RFP but not AUX1-YFP(Pearson’s coefficient 0.007 6 0.055) (Supplemental Figures 4Aand 4B). Importantly, the AUX1-YFP agglomeration in BFA-treated big3 was suppressed by expression of a BFA-resistantBIG4-RFP but not a BFA-sensitive BIG4-RFP (SupplementalFigures 5A o 5D) or expression of GNOMR (Supplemental Figures5E and 5F). Thus, these results suggest that GNOMdoes not playa role in the traffickingofAUX1-YFP to thePM, incontrast to its role

Ethylene-Mediated Differential Growth by ARF-GEFs 1041

in PIN1 trafficking. Taken together, these results indicate that, likeECH, BIG1-4 mediate de novo AUX1 trafficking to the plasmamembrane, via the TGN, independently of GNOM function.

BIG4 and ECH Colocalize at the TGN and Their LocalizationIs Interdependent

BIG1-4 were recently shown to localize to the TGN in Arabidopsisroots (Richter et al., 2014). At least two domains have beenidentified in the TGN, one of which is labeled by VHA-a1, SYP61,and ECH, whereas the other of which only partially overlaps withVHA-a1 but is marked with RABA2a (Chow et al., 2008; Gendreet al., 2011; Uemura et al., 2014). Since ech and big1-4 mutantsdisplay very similar phenotypes, we investigated whether BIGscolocalize with ECH at the TGN. We observed a strong colocal-ization between ECH-YFP and BIG4-RFP (61%) as well as be-tween ECH-YFP and the TGN-localized VHA-a1-RFP (64%) andBIG4-RFP and VHA-a1-GFP (65%) (Figures 6A to 6D), usinga quantitative morphology-based method for assessing coloc-alization percentage (Boutté et al., 2006). This observation alongwith the phenotypes of the ECH and BIG ARF-GEF mutants

strongly suggests that ECH and BIG ARF-GEFs operate at thesame TGNdomain. Therefore, we testedwhether ECH is requiredfor BIG ARF-GEF localization at the TGN. While BIG4R-YFP ex-hibited an almost exclusively punctate labeling in wild-typeseedlings (Figure7A),BIG4R-YFPwasstronglymislocalized inechmutants, exhibiting a diffuse rather than punctate pattern (Figure7B). Similarly, we tested whether ECH localization required BIGARF-GEF function to assess the localization interdependence ofECH and BIG ARF-GEFs. While ECH-YFP localized to punctatestructures in mock-treated wild-type and big3, ECH-YFP ag-glomerated in the wild-type after 1 h of 50 mM BFA treatment(Figures 7C to 7E). By contrast, 1 h of 50mMBFA treatment of big3led to ECH-YFP exhibiting a diffuse pattern (Figure 7F) differentfrom that observed in BFA-treated wild-type seedlings, indicatingthat ECH localization requires BIG ARF-GEF function and, hence,that the localization of ECH and BIG ARF-GEFs at the TGN isinterdependent.ECHhasbeenshown tobe important for the localizationofVHA-

a1 as well as other proteins such as YIPs and SYP61 at the TGN(Gendreetal., 2011,2013).Disruptionof theV-typeprotonATPasefunctionbyConcanamycinA (ConcA) results in themislocalization

Figure 2. BIG1-4 Are Required for Ethylene-Mediated Hook Maintenance.

(A) Wild-type seedlings respond to 10 mM ACC treatment with exaggerated hook curvature both in the presence or absence of 0.5 mM BFA.(B)big3mutant seedlingsareunable toexaggerate in response to10mMACCin thepresenceof0.5mMBFA.Foreachgenotype,n$16.Errorbars representSE of the mean.

1042 The Plant Cell

of several TGNproteins, suchasSYP61, that are alsomislocalizedin the ech mutant. This result prompted us to suggest that themislocalization of TGN proteins in the echmutant may be causedby the disruption of VHA-a1 function. Therefore, we testedwhether the mislocalization of BIG ARF-GEFs in ech was due todisruption of VHA-a1 function by targeting VHA-a1 activity usingConcA. Our results show that a 2-h 5 mMConcA treatment did notaffect the localization of BIG3-YFP or BIG4R-YFP (Figures 7G to7J). By contrast, the primary target of ConcA, VHA-a1-GFP,strongly agglomerated into large ConcA-induced bodies, in-dicating the effectiveness of theConcA treatment (Figures 7K and7L). Thus, ECH is required for the proper localization of BIG ARF-GEFs at the TGN independently of VHA-a1.

BIG ARF-GEFs and ECH Are Required for ARF1 Localizationat the TGN

BIG3 has been previously shown to act as an ARF-GEF for theARF1GTPase in vitro (Nielsen et al., 2006). SinceBIG4colocalizeswith ECHat theTGN,we investigatedwhetherARF1alsocolocalizeswith BIG ARF-GEFs at the TGN. We observed that ARF1-labeledstructures colocalized with both BIG4 and VHA-a1 (40 and 47%,respectively) (Figures 8A to 8C), indicating that ARF1 and BIG4

localize to the same TGN domain. The mislocalization of BIG ARF-GEFs in the ech mutant led us to investigate whether the TGN lo-calization of ARF1, the potential in vivo target of BIG3, could becompromised in the echmutant. Whereas ARF1-GFP labeling in thewild type was largely restricted to punctate structures (Figure 9A),ARF1-GFP in the ech mutant exhibited a diffuse pattern with fewpunctate structures, indicating themislocalization of ARF1 in the echmutant (Figure 9F). To examine the role of BIGs in ARF1 localization,we visualized ARF1-GFP in the big3 mutant upon BFA treatment.While ARF1-GFP exhibited punctate labeling undermock conditionsin both the wild type and the big3 mutant (Figures 9B and 9G), thepunctate labelingofARF1-GFPwas lostaftera15-min treatmentwith50 mM BFA in big3 (Figure 9H), before any discernable effect on itslocalization was observed in BFA-treated wild type (Figure 9C). Bycontrast, VHA-a1-GFP labeling was unaffected after a 15-min BFAtreatment in both the wild type and big3 (Figures 9D, 9E, 9I, and 9J).The rapid mislocalization of ARF1 in big3 upon BFA treatmentsuggests thatARF1localizationmaydependonTGN-localizedBIG1-4 function. Introduction of a BFA-resistant BIG4-RFP but not a BFA-sensitive BIG4-RFP reversed the ARF1-GFP mislocalization in big3upon a 30-min 50 mM BFA treatment, indicating that BIG ARF-GEFfunction like ECH is required for proper ARF1 localization at the TGN(Supplemental Figures 6A to 6D).

Figure 3. Loss of ECH Causes BFA Hypersensitivity, and ECH Interacts Genetically with BIG2 and BIG3.

(A) Hook development in ech is perturbed when exposed to 0.5 mM BFA, while wild-type seedlings remain unaffected.(B) ech big2 big3 triple mutant exhibits more severe apical hook development compared with ech and big2 big3mutants. For each genotype, n$ 16. Errorbars represent SE of the mean.

Ethylene-Mediated Differential Growth by ARF-GEFs 1043

ARF1 Members Are Essential for Ethylene-Mediated ApicalHook Development and AUX1 Trafficking

ARF1 localization at the TGN relies on BIGs and ECH. Therefore,we investigated whether ARF1 also plays a role in hook de-velopment, like BIG ARF-GEFs and ECH. However, the high re-dundancy among ARF1 subclass members (Xu and Scheres,2005) makes it difficult to address the ARF1 function. Therefore,we generated transgenic plants expressing a dominant-negativeGDP-locked mutant of the ARF1 member ARFA1c with low GTPaffinity fused to CFP (ARFA1cT31N-CFP) (Dascher and Balch,1994; Xu and Scheres, 2005), under control of a b-estradiol-inducible UBQ10 promoter to interfere with ARF1 function. Incontrast to plants that expressed a wild-type variant of ARFA1c(ARFA1cWT-GFP) or untransformed wild-type controls (Figures10A and 10B), induction of ARFA1cT31N-CFP mutant expressionled to severely perturbed hook development. Upon induction ofARFA1cT31N-CFP, thehook formationpeakedat145° (Figure10C)and the hook maintenance phase was completely abolished,closely resembling the ech and big mutant phenotypes. Fur-thermore, ACC treatment of the wild-type or ARFA1cWT-GFPseedlings resulted inanexaggeratedhookangleof;270° (Figures10A and 10B), while seedlings expressing ARFA1cT31N-CFP didnot exaggerate hook curvature upon ACC treatment (Figure 10C).The transcript level of ARFA1cT31N-CFP upon b-estradiol in-duction was similar to that of ARFA1cWT-GFP (SupplementalFigure 7), indicating that the observed effect of ARFA1cT31N-CFPon hook development was not due to a higher expression ofARFA1cT31N-CFP relative to that of ARFA1cWT-GFP.

ARF1 colocalizes with BIGs at the TGN, and the disruption ofARF1 function, as in the ech or big mutants, results in failure to

respond to ACC. Therefore, we examined whether ARF1 functionis also required, like ECH and BIG1-4, for AUX1-YFP traffickingto the PM. AUX1-YFP fluorescence levels at the PMdecreasedto ;25% of wild-type levels following ARFA1cT31N-CFP in-duction, whereas AUX1-YFP levels at the PM were unaffectedby ARFA1cWT-GFP induction (Figures 11A to 11F; SupplementalFigure 8). In the wild type, as well as in seedlings expressingARFA1cWT-GFP, AUX1-YFP was exclusively localized to theplasma membrane (Figures 11A, 11B, 11D, and 11E), with nointracellular signal detected. By contrast, a strong intracel-lular AUX1-YFP signal was observed in seedlings expressingARFA1cT31N-CFP (Figure 11F). Thus, AUX1 delivery to the PMrequires the function of the ARF1 subclass. These observa-tions strongly suggest that ECH, BIG1-4, and ARF1 operate in

Figure 4. AUX1-YFP Fluorescence at the Plasma Membrane Is Reducedin big3 upon BFA Treatment.

While under mock conditions, AUX1-YFP levels are equal in the wild typeand big3 during apical hook formation (24 h postgermination) and main-tenance (48 h postgermination); when grown on medium supplied with0.5 mMBFA, AUX1-YFP levels are unaffected in the wild type but stronglyreduced in big3 during both hook formation and maintenance phases. Allexperimentswereperformed in epidermal cells of the apical hook. For eachtreatment, five cells from each of 10 seedlings were analyzed. ***Student’st test P value < 0.0001.

Figure 5. BIG1-4 Are Required for de Novo Delivery of AUX1-YFP at theTGN.

(A) to (F)AUX1-YFPagglomerates inbig3uponBFAtreatment.Whileundermock conditions, AUX1-YFP remains exclusively at thePM inboth thewildtype (A) and big3 (D); upon 3 h of 50 mM BFA, AUX1-YFP agglomeratesintracellularly in big3 (E) (white arrowhead) but not wild-type (B). Ag-glomerationsareblocked inbig3whenseedlingsarepretreatedwith50mMCHX for 1 h followedby 3 h of 50mMCHX+50mMBFA treatment (F), whileAUX1-YFP remains unaffected in the wild type (C).(G) to (L) AUX1-YFP agglomerations in big3 overlap with VHA-a1-GFP. Inbig3 under mock conditions, AUX1-YFP resides at the PM (G), while VHA-a1-GFP exhibits a punctate labeling (H) (overlay in [I]). Upon 3 h of 50 mMBFA treatment in big3, AUX1-YFP agglomerations (white-bordered ar-rowhead) (J) overlap with VHA-a1-GFP (K) (overlay in [L]). All experimentswere performed in epidermal cells of the apical hook at 48 h post-germination. Bars = 20 mm.

1044 The Plant Cell

a common pathway during hook development and AUX1 traf-ficking to the PM.

DISCUSSION

Plants exhibit remarkable plasticity during development and areable to modify their pattern of growth to adapt to the environment.

They often rely on differential cell elongation to change theirgrowth pattern, as exemplified in the apical hook developmentthat follows seed germination. Here, we demonstrate that theGTPase ARF1 and its effectors ARF-GEFs of the BIG family playa role in apical hook development by mediating the secretory traf-ficking of the auxin influx carrier AUX1 from the TGN to the PM inconcert with the previously described TGN-localized protein ECH.

Figure 6. BIG4 Colocalizes with ECH at the TGN.

Colocalizationexperiment shows thatBIG4-RFPexhibits61%overlapwithECH-YFPand65%overlapwith theTGNmarkerVHA-a1-GFP ([A], [C], and [D]),whileECH-YFPexhibits64%overlapwithVHA-a1-RFP ([B]and [D]). Fiveepidermal cells ineachof$10apical hooksat48hpostgerminationwereanalyzedfor each treatment. Error bar represents SD of the mean. Bars = 10 mm.

Figure 7. BIG ARF-GEF and ECH Localization Is Interdependent, but BIG Localization Is Unaffected by Concanamycin A Treatment.

While BIG4R-YFP exhibits punctate labeling in the wild type (A), in ech, BIG4R-YFP labeling is diffuse (B). Under mock conditions, ECH-YFP exhibitspunctate labeling in both the wild type and big3 ([C] and [E]). Upon 1 h of 50 mM BFA treatment, ECH-YFP agglomerates into large intracellular bodies inwild-type (D), while in big3, ECH-YFP exhibits a diffuse labeling upon 1 h of 50 mMBFA treatment (F). Under mock conditions, BIG3-YFP, BIG4R-YFP, andVHA-a1-GFP exhibit punctate labeling ([G], [I], and [K]). Upon 2 h of 5mMConcA treatment, BIG3-YFP andBIG4R-YFP remain punctate ([H] and [J]), whileVHA-a1-GFP agglomerates (L). All experiments were performed in epidermal cells of the apical hook at 48 h postgermination. Bars = 20 mm.

Ethylene-Mediated Differential Growth by ARF-GEFs 1045

Stage-Specific Role of ARF-GEFs in ApicalHook Development

The specific defects in secretory vesicle (SV) morphology andnumber previously observed in ech hinted at ECH being essentialfor proper secretory vesicle genesis at the TGN (Boutté et al.,2013).Other knownplayers in vesicle formation areARF-GTPases(D’Souza-Schorey and Chavrier, 2006) that cycle between GTP-and GDP-bound states and ARF-GEFs that play a key role in ARFGTPase functionby facilitating theGDP-to-GTPexchangeonARFGTPases (Anders and Jürgens, 2008). Our results showed thathook development in the ech mutant is hypersensitive to treat-ment with the ARF-GEF-inhibitor BFA, suggesting that ARF-GEFfunction may be compromised in the echmutant (Figure 3A). Ourdissection of ARF-GEF involvement during hook developmentrevealed that both the BFA-sensitive ARF-GEF GNOM and ARF-GEFs of the BIG family, BIG1-4, play a role in hook development.While the BFA-sensitive ARF-GEF GNOM is essential mainlyduring the formation phase of apical hook development (Figure

1A), the maintenance phase relies on the BIG ARF-GEF subclassmembers BIG1-4 (Figures 1B and 1C). It has previously beenshown that ARF-GEFs can functionally replace each other. Forexample, GNOM can substitute for GNOM-LIKE1, and GNOM-LIKE2 can substitute for GNOM (Richter et al., 2007, 2011).However, this does not appear to be the case for GNOM andBIG1-4 in hook development, as the distinct ARF-GEFs havestage-specific roles, with BIG1-4 operating redundantly and in-dependently of GNOM to mediate apical hook maintenance(Figures 1B and 1C; Supplemental Figures 2A to 2D). Moreover,since GNOM is expressed at all stages of hook development(Supplemental Figure 1), the stage-specific roles of GNOM andBIGs are not due to their differential stage-specific expression.Thus, the results presentedhere suggest thatARF-GEF regulationof hook development is separated temporally, with distinctpathways operating at different developmental phases. It re-mains to be seen whether the apical hook defects observedfollowing the disruption of GNOM and BIG function are related

Figure 8. ARF1 Structures Overlap with BIG4 at the TGN.

ARF1-GFP-labeledstructuresexhibit 40and47%overlapwithBIG4-RFPandVHA-a1-RFP, respectively. Fiveepidermal cells ineachof$10apicalhooksat48 h postgermination were analyzed for each treatment. Error bars represent SD of the mean. Bars = 10 mm.

Figure 9. ECH and BIGs Are Required for Proper ARF1-GFP Localization.

ARF1-GFP localization is punctate in thewild type (A)while strongly disrupted in ech (F).While 15min of 50mMBFAhas noeffect onARF1-GFP localizationin the wild type (C) compared withmock (B), punctate ARF1-GFP in big3 under mock conditions (G) becomes strongly mislocalized in big3 upon 15min of50mMBFA treatment (H). By contrast, VHA-a1-GFP localizationundermock conditions in either thewild type (D)orbig3 (I) remains unaffected upon15minof 50mMBFA treatment in both thewild type (E) and big3 (J). All experiments were performed in epidermal cells of the apical hook at 48 h postgermination.Bars = 20 mm.

1046 The Plant Cell

to their distinct roles in trafficking, with GNOM participating inrecycling (Geldner et al., 2003) andBIG1-4 involved in secretionrather than recycling (Richter et al., 2014).

BIG1-4 and ECH Act in Concert during Ethylene-MediatedHook Development

The plant hormone ethylene is involved in hook maintenance, andethylene-insensitive mutants, such as ein2, have severe defects inthemaintenance phase (Guzmán and Ecker, 1990; Vandenbusscheet al., 2010). Our data showed that BIG1-4 are required for ethylene-mediated hook maintenance (Figures 2A and 2B). In this respect,disruption of BIG1-4 function phenocopies the echmutant, which isalso ethylene-insensitive (Boutté et al., 2013). These observationssuggested that ECH and BIG1-4 could operate in a common

pathway in hook development where ethylene is an upstreamregulator. In accordance with this hypothesis, we observed that anech big2 big3 triple mutant displayed a strongly enhanced apicalhook defect compared with the ech or big2 big3 double mutant(Figure3B).Thus,BIG2andBIG3act togetherwithECH inapathwaythat controls hook development downstream of ethylene.

BIG1-4 Mediate AUX1 Trafficking to the PM at an ECH-Positive TGN Domain

Previous studies have revealed that AUX1 is required for ethylene-induced exaggeration of hook curvature and that in response toethylene, AUX1 PM turnover increases preferentially on theconcave side of the hook (Roman et al., 1995; Vandenbusscheet al., 2010; Boutté et al., 2013). Delivery of de novo-synthesizedAUX1 to thePM is strongly attenuated in theechmutant, andhookdevelopment in the ech mutant is insensitive to ethylene (Bouttéet al., 2013). In agreement with the genetic interaction betweenECH and BIGs and the overall similarity of the big and echmutantphenotypes, our data now revealed that BIG1-4 are essential forthe delivery of newly synthesized AUX1 to the PM, independentlyofGNOM (Figures 4 and5A to 5F; Supplemental Figures 5A to 5F).Furthermore, our colocalization study indicates that BIG3 andECH reside on a similar TGN subdomain (Figures 6A to 6D). Thus,genetic andcell biological datasuggest thatbothBIG1-4andECHare required at the TGN for AUX1 trafficking to thePMduring hookdevelopment, highlighting the role of secretory trafficking inregulation of differential growth of the apical hook.

ECH and BIG1-4 Act in Concert Independently of VHA-a1 inAUX1 Trafficking from the TGN

TGN has a key role in secretion to the PM (Gendre et al., 2015).In recent years, the characterization of ECH and VHA-a1, a

Figure 10. ARF1 Is Required for Ethylene-Mediated Hook Development.

Kinematic analyses of apical hook development reveals that hook cur-vature is exaggerated in wild-type and plants expressing ARFA1cWT-GFPin response to 10mMACC treatment ([A] and [B]). Compared with the wildtype, plants expressing ARFA1cT31N-CFP exhibit a shortened mainte-nance phase and insensitivity to 10 mM ACC treatment (C). For eachgenotype and treatment, n $ 16. Error bars represent SE of the mean.

Figure 11. ARF1 Is Required for Proper AUX1-YFP Localization.

AUX1-YFPPM levelsareequal in thewild type (A), ARFA1cWT-GFP (B), andARFA1cT31N-CFP (C) under mock conditions. Upon 5 nM b-estradiol in-duction of expression, AUX1-YFP PM fluorescence remains unchanged inthe wild type (D) and ARFA1cWT-GFP (E), while AUX1-YFP is stronglyreduced at the PM and retained intracellularly in ARFA1cT31N-CFP (whitearrowhead) (F). All experiments were performed in epidermal cells of theapical hook at 48 h postgermination. Bars = 20 mm.

Ethylene-Mediated Differential Growth by ARF-GEFs 1047

component of the TGN-localized vacuolar H+ ATPase, and otherTGN-resident proteins have revealed certain details about themechanisms underlying TGN function and protein secretion fromthe TGN to the PM (Dettmer et al., 2006; Gendre et al., 2011;Uemura et al., 2012; Asaoka et al., 2013). Disruption of VHA-a1 orECH results in the mislocalization of TGN-resident proteins andalters TGN morphology (Dettmer et al., 2006; Viotti et al., 2010;Boutté et al., 2013). Disruption of TGN-localized vacuolar H+

ATPasealtersTGNpH levels, resulting indefective trafficking fromtheTGN (Luo et al., 2015). VHA-a1haspreviously beenconnectedto the secretion of BRI1 (Dettmer et al., 2006). However, despitethe colocalization betweenECHandVHA-a1 at the TGNaswell asthe mislocalization of VHA-a1 in the echmutant, ECH-dependentAUX1 secretory trafficking in the apical hook and ECH-dependentpectin secretion in seed coat cells occur independently of VHA-a1(Boutté et al., 2013; Gendre et al., 2013; McFarlane et al., 2013).Like ECH, BIGs (BIG3/4) localize to the TGN and colocalize withVHA-a1 (Richter et al., 2014) (Figures6Cand6D) andECH (Figures6A and 6D). Thus, a crucial question to addresswaswhether BIGsact together with VHA-a1 or independently of it in hook de-velopment. We observed that BIG localization was severely af-fected in the ech mutant, as BIG4R-YFP exhibited a cytosol-likeinstead of a punctate pattern (Figures 7A and 7B). Similarly, ECHlocalization was perturbed by the disruption of BIG ARF-GEFfunction (Figures 7C to 7F), indicating the mutual dependence for

TGN localization of ECH and BIGs. In contrast, when VHA-a1functionwas inhibitedwithConcA, BIGARF-GEF localizationwasunaffected (Figures 7G to 7L), indicating that the TGN localizationof BIGARF-GEFs analyzed here does not rely on VHA-a1 functiondespite the strong colocalization between them (Richter et al.,2014). Thus, while the secretion of AUX1 requires TGN function, itismediated byBIG1-4 andECH inaVHA-a1-independentmannerduring hook development. These observations reveal the com-plexity of TGN structure and function and indicate that evenwithina subdomain of the TGN, multiple distinct secretory pathwaysoperate in a highly regulated manner.

ARF1 GTPases Mediate Ethylene-Regulated HookDevelopment and Require ECH and BIG1-4 for TheirProper Localization

ARF-GEFs regulatevesicle formationbyactivatingGTPasesof theARF/Sar family (Peyroche et al., 1996; Anders and Jürgens, 2008).The ARF-GEF BIG3 was previously shown to catalyze GDP/GTPexchange specifically for members of the ARF1 subclass, local-ized mainly to the TGN and Golgi (Nielsen et al., 2006; Robinsonet al., 2011). We observed that the disruption of ARF1 function byexpressing a dominant-negative variant of ARFA1c (ARF1T31N)exhibited severely perturbed hook development (Figures 10A to10C), with complete abolishment of the maintenance phase,

Figure 12. Model of AUX1 Secretory Trafficking at the TGN.

(A) In the wild type, ECH and BIG ARF-GEFs localize to the TGN allowing ARF1 GTPases to be recruited, facilitating vesicle formation required for AUX1delivery to the PM, a prerequisite for the ethylene response during hook development.(B) In the absence of ECH, both BIG ARF-GEFs and ARF1 are lost from the TGN, leading to defects in AUX1 delivery to the PM and insensitivity to ethyleneduring hook development.(C)WhenBIGARF-GEF function is perturbed, ECHandARF1aremislocalized,which results inAUX1delivery to thePMbeinghamperedand insensitivity toethylene during hook development.

1048 The Plant Cell

phenocopying ech and big1/+ big2 big3 big4 or BFA-treated big3mutant. Moreover, defects in hook development caused by theexpression of the dominant negative ARFA1c could not be sup-pressed by ethylene treatment as shown earlier for ech or big3upon BFA treatment (Figure 10C). ARF1T31N, which has a lowaffinity forGTP, is thought tosequesterendogenousARF-GEFsbymimickingGDP-bound inactiveARF1, hampering theactivationofendogenous ARF1, therebymimicking the loss of ARF1 and ARF-GEF function (Dascher and Balch, 1994). Perturbing ARF1 func-tion also greatly affected the delivery of AUX1 to the PM, as AUX1wasstrongly retained in intracellular structuresandhighly reducedat thePM (Figures 11A to11F;Supplemental Figure 8). This placesthe ARF1 into the same genetic pathway as ECH and BIG1-4 inhook development. Our data also showed that ARF1 colocalizedstrongly with BIG4 at the TGN (Figures 8A to 8C), and ARF1 la-beling dramatically shifted from strictly punctate to highly diffusewhen either BIG1-4 or ECH function was perturbed, suggestingthat ARF1 is mislocalized and probably no longer associatedwithamembranecompartment in theabsenceofECHorBIGARF-GEFs (Figures 9A to 9C and 9F to 9H). These results suggest thatARF1 activation via ARF-GEF-mediated GDP/GTP exchangemay have been compromised in the absence of ECH and BIGARF-GEFs.

ARF1 has been associatedwith COPI-mediated pathways at theGolgi (Goldberg, 1999;Pimpl etal., 2003).ARF1also localizes to theTGN inplants (Robinsonet al., 2011) andhasbeendemonstrated tohave the capacity to mediate COPI-independent vesicle formationat the TGN in a cell-free system derived from PC12 cells (pheo-chromocytomaof the rat adrenalmedulla) (Barr andHuttner, 1996).Interestingly, inhibiting BIG1-4 function does not affect COPI lo-calization (Richter et al., 2014). Thus, BIG1-4 may regulate ARF1functionat theTGNinArabidopsis.Similarly,defects inecharemostprominent at the TGN, and importantly BIGs are mislocalized fromthe TGN in ech (Figures 7A and 7B). Thus, ECH may also mediateARF1-dependent processes preferentially at the TGN, potentiallyvia a BIG ARF-GEF-dependent pathway.

We previously showed that the number of SVs at the TGN isreduced in the ech mutant, while the number of clathrin-coatedvesicles (CCVs) remained unaffected (Boutté et al., 2013). UnlikeCCVs,SVsmaybeara thincoat or even lackacoat (Donohoeetal.,2007), forming via compositional modifications in membranelipids that induce curvature of the membrane (Bard andMalhotra,2006; Gendre et al., 2015). Notably, in vitro studies in rat pituitaryGH3 cells demonstrated that Phospholipidase D (PLD) stimulatesSV budding at the TGN (Chen et al., 1997). PLD mediates theconversion of the cylindrical phospholipid phosphatidylcholine tothe more conical phosphatidic acid (Pappan et al., 1998), causingthe membrane to arch. PLD activity is stimulated by ARF1, whichprovides a potential mechanistic framework for SV formation atthe TGN (Chen et al., 1997). Analogous mechanisms for the for-mation of SVs from TGN have not been demonstrated as yet inplants. Nevertheless, our data showing reduced SVs in ech andcommonalities in cell biological anddevelopmental phenotypesofech and bigmutants warrant the investigation of the role of ECH/BIGs/ARF1 and lipids in SV formation at the TGN in the future.

The TGN is a key hub for secretory and endocytic pathways. Atthe least two distinct domains, defined by ECH/VHAa1/SYP61and by RabA2a, appear to be involved in the trafficking of distinct

TGN-PM and endocytic pathways (Dettmer et al., 2006; Chowet al., 2008; Gendre et al., 2011, 2013). Nevertheless, our knowl-edge of protein sorting and separation of diverse pathways fromtheTGNremains incomplete.Wepreviously showed thatECHandVHA-a1, despite localizing to the same TGN domain, could havedistinct functions, and ECH but not VHA-a1 is required for thetrafficking of AUX1, a key component required for ethylene-mediated hook development (Boutté et al., 2013). Furthermore,ethylene application has been shown to increase the turnover ofAUX1 on the concave side of the hook. Our results presented herereveal additional components of the AUX1 trafficking pathwayessential to ethylene-mediated hook development (Figure 12).This pathway composed of BIG ARF-GEFs and their target ARF1acts in AUX1 trafficking independently of VHA-a1 as describedearlier for ECH (Boutté et al., 2013). As summarized in the model(Figure12),whenECH isnotpresentat theTGN,BIG localizationatthe TGN is affected and vice versa, which in turn results in the lossof ARF1 from the TGN and severe reduction of AUX1 levels at thePM and attenuation of the ethylene response in the hook (Figure12). Both ARF1 and BIG ARF-GEFs have broad roles in traffickingrelative to those uncovered so far for ECH. For example, ARF1regulates additional processes independent of ECH, such asCOPI-mediated retrograde trafficking from the Golgi (Stefanoet al., 2006). Similarly, BIG1-4 operate in pathways distinct fromECH function, such as trafficking of cargo to the vacuole (Richteret al., 2014). Therefore, it is tempting to speculate that ECHmightfunction as a component of the machinery that while requiringARF1 and BIGs, could facilitate compartment and/or pathwayspecificity for trafficking via TGN. Unraveling the mechanismsbehind trafficking pathway specificity may thus provide insightinto the intricate complexity of endomembrane trafficking andhow the multifaceted network of pathways controls myriad de-velopmental processes spanning from embryogenesis to se-nescence.

METHODS

Plant Material

The Arabidopsis thaliana ecotype Columbia-0 (Col-0) and the followingmutants were used in this study: ech (Gendre et al., 2011), big1, big2, big3,big4, big1 big3, big1 big4, big2 big3, big3 big4, big1 big2 big3, big1 big2big4,big1big3big4,big2big3big4, andbig1/+big2big3big4 (Richteretal.,2014). The following transgenic lines were used: GNOMM696L-myc(GNOMR) (Geldner et al., 2003), ProAUX1:AUX1-YFP (Swarup et al., 2004),ProVHA-a1:VHA-a1-GFP, ProVHA-a1:VHA-a1-RFP (Dettmer et al., 2006),ProECH:ECH-YFP (Gendre et al., 2011), ProBIG3:BIG3-YFP, ProUBQ10:BIG4-YFP, ProUBQ10:BIG4R-YFP, ProBIG3:BIG4-RFP, ProBIG3:BIG4R

-RFP (Richter et al., 2014), and ARF1-GFP (Xu and Scheres, 2005).

Growth Conditions

Plantsweregrownonplates suppliedwith0.53MS (2.2g/LMurashigeandSkoog nutrient mix [Duchefa]), 0,8% (w/v) plant agar (Duchefa), 0.5% (w/v)sucrose, and2.5mMMES(Sigma-Aldrich) bufferedatpH5.8withKOH.Forconfocal microscopy analysis, seeds were stratified for 2 d at 4°C, givena 6-h white light treatment, and subsequently grown in darkness on ver-tically oriented agar plates at 21°C for the appropriate time (indicated foreach experiment in the results section and/or figure legends). For time-lapse analysis of apical hook development, seedlings were grown vertically

Ethylene-Mediated Differential Growth by ARF-GEFs 1049

on plates in a dark room at 21°C illuminated only with an infrared LED lightsource (850 nm). Seedlings were photographed at 4-h intervals usinga Canon D50 camera without an infrared filter.

Inhibitor Treatments and Chemical Induction

For time-lapse analysis of apical hook development, seedlingswere grownon agar plates supplied with the respective inhibitor/induction agent orequivalent amounts of solvent for mock treatments. For inhibitor pre-treatments of dark-grown seedlings, seedlings were incubated in liquidmedium (0.53MS supplied with 0.5% [w/v] sucrose and 2.5 mMMES, pHadjusted to 5.8 with KOH) supplied with the indicated inhibitor or anequivalent amount of solvent during the time indicated. Stock solutionswere 50 mM BFA, 5 mM ConcA, 100 mM ACC, and 50 mM CHX (Sigma-Aldrich), dissolved in DMSO. For induction of ARFA1cWT-GFP and AR-FA1cT31N-CFP expression, seeds were germinated and grown on platessuppliedwith 5 nMb-estradiol (Sigma-Aldrich) or the equivalent amount ofsolvent from a 5 mM stock solution.

Genotyping of big T-DNA Mutant Lines

Genotyping was performed using a Phire Plant Direct PCR kit (ThermoFisher Scientific),with the followingprimers.Primersused forbig1 (GK-452B06)T-DNA mutant analysis are as follows: wild-type allele, 59-CTAATCCGT-GGCGATTGTTT-39 and 59-GACCTGGCGTTCAATTTTGT-39; T-DNA allele,59-CTAATCCGTGGCGATTGTTT-3 and 59-ATATTGACCATCATACTCATTGC-39. Primersused forbig2 (SALK_033446) T-DNAmutant analysis are as follows:wild-type allele, 59-GAAGGGTCCTGGGAAAAGTC-39 and 59-TCATCCACTT-CACCACCAGA-39; T-DNA allele, 59-TCATCCACTTCACCACCAGA-39 and 59-ATTTTGCCGATTTCGGAAC-39. Primers used for big3 (SALK_044617) T-DNAmutant analysis are as follows: wild-type allele, 59-TAAAACTGGCGGAGGAA-GAA-39 and 59-TCTGCTCTTGGGTCGAAACT-39; T-DNA allele, 59-TAAAA-CTGGCGGAGGAAGAA-39 and 59-ATTTTGCCGATTTCGGAAC-39. Primersused for big4 (SALK_069870) T-DNA mutant analysis are as follows: wild-typeallele, 59-TCCGTGGTGATTGCTTGTTA-39 and 59-TTTGTAAGCCCTTCGTTG-CT-39; T-DNA allele, 59-TTTGTAAGCCCTTCGTTGCT-39 and 59-ATTTTGCC-GATTTCGGAAC-39.

For big1 heterozygosity analysis of the big1/+ big2 big3 big4 mutant,individual seedlings were genotyped after completion of the time-lapseimaging experiment.

Confocal Laser-Scanning Microscopy and Quantitative Analyses ofFluorescence Intensity and Colocalization

All fluorescence detection by confocal laser-scanning microscopy wasperformed using a Carl Zeiss LSM780 equipped with a 403 lens (ZeissC-Apochromat 403/1.2 W Corr M27). For examination of AUX1-YFPfluorescence intensities at the plasma membrane, the plasma membranewas acquired by manually outlining a 0.5-mm segmented line along theplasma membrane using ImageJ software. For each experiment, fiveepidermal cells from each of $10 seedlings were measured (n $ 50). ForGNOM-GFP fluorescence intensity examination, the interior of individualcells was demarcated manually using polygonal selection tool in ImageJ,and fluorescence intensity was measured. For each time point, five epi-dermal cells from each of 10 seedlings were measured (n = 50). Coloc-alization of endomembrane compartments was performed using thesimultaneous line-scanning mode, with a four-line average. Individualchannels were segmented, subcellular objects were demarcated, andgeometrical centers (centroids) of objects were calculated using the 3Dobjects counter plug-in in ImageJ. Colocalization of two objects wasacceptedwhen thedistancebetween themwasbelow theXY resolution (R)limit according to R = 0.61l/NA. For each comparison, five epidermalcells fromeach of 10 seedlingsweremeasured (n=50). For AUX1-YFP andVHA-a1-RFP intracellular overlap analysis, the cell interior signal was

separated from theplasmamembrane signal bymanual demarcationusingthepolygonal selection tool in ImageJandselectedas the regionof interest.The ImageJcolocalizationplug-inwasused toobtainPearson’scorrelationcoefficient for regions of interest of individual cells. For each experiment,five epidermal cells fromeachof 10 seedlingswere analyzed (n=50). For allquantitative analyses of fluorescence intensity, confocal images wereacquired using identical acquisition parameters (laser power, photo-multiplier, offset, zoom factor, and resolution) between the examinedgenotypes. All experiments were performed in the apical hook of dark-grown seedlings.

Generation of Transgenic Plants

ARF1WT-GFP and ARF1T31N-CFP sequences were amplified from extractedgenomic DNA from plants carrying these constructs (Xu and Scheres, 2005)using the following primers: CACCATGGGGTTGTCATTCGGAAAGTTG andTTACTTGTACAGCTCGTCCATGCCG. Fragments were introduced into thepENTR/D-TOPO cloning vector (Invitrogen), and sequences were verified byDNA sequencing. Afterwards, LR reaction was performed to introduceARF1WT-GFP and ARF1T31N-CFP into the pMDC7.b vector. All constructswere transformed into Col-0 wild-type plants. Plants were selected on agarplates supplied with hygromycin (25 mg/mL) (Duchefa). Three independentlines were analyzed.

RNA Isolation and Quantitative Real-Time PCR Analysis

Total RNA was extracted from 3-d-old dark-grown seedlings usinga Spectrum Plant Total RNA isolation kit (Sigma-Aldrich) according to themanufacturer’s protocol. RNA (5 mg) was treated with RNase-free TURBODNase (Life Technologies, Ambion), and 1 mg was then used for cDNAsynthesis using an iScript cDNA synthesis kit (Bio-Rad). ELFawas used asthe referencegene for all experiments after screeningandvalidationof asetof four different referencegenes (UBQ, TIP, ELFa, andactin) usingGeNormsoftware. CFP/GFP/YFP primers were designed from conserved se-quences and the same primers were used for all experiments. Quantitativereal-time PCR analyses were performed with a Roche LightCycler 480 IIinstrument, and relative expression values were calculated using theD-Ct-method (DCt) to calculate relative expression values of genes ofinterest. Real-time PCR primers were as follows: CFP/GFP/YFP-Fwd,59-GCACGACTTCTTCAAGAGCGCCATGCC-39; CFP/GFP/YFP-Rev, 59-CCGTCCTCCTTGAAATCGATTCCCTT-39; ELFa Fwd, 59-TGGTGACG-CTGGTATGGTTA-39; and ELFaRev, 59-TCCTTCTTGTCCACGCTCTT-39.

Statistical Analyses

For statistical analyses, Microsoft Excel 2010 was used. For generation ofbox plot graphs, Minitab 16 was used. To evaluate statistical differencesin plasma membrane intensities, Student’s t test was used. For all tests,*P < 0.05 and ***P < 0.0001.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL datalibraries under the following accession numbers: ECH (At1g09330), BIG1(At4g38200), BIG2 (At3g60860), BIG3 (At1g01960), BIG4 (At4g35380),GNOM (At1g13980), and ARFA1c (At2g47170).

Supplemental Data

Supplemental Figure 1. GNOM is expressed ubiquitously throughoutapical hook development.

Supplemental Figure 2. BIG1-BIG4 act redundantly and indepen-dently of GNOM during apical hook development.

Supplemental Figure 3. qRT-PCR analysis of AUX1-YFP expression.

1050 The Plant Cell

Supplemental Figure 4. AUX1-YFP overlaps with TGN-marker VHA-a1-RFP in big3 upon BFA treatment.

Supplemental Figure 5. BFA-resistant BIG4 but not BFA-resistantGNOM reverses BFA-induced AUX1-YFP agglomerations in big3.

Supplemental Figure 6. BFA-resistant BIG4 reverses BFA-inducedARF1 mislocalization in big3.

Supplemental Figure 7. qRT-PCR analysis of inducible ARF1 ex-pression.

Supplemental Figure 8. Quantification of AUX1-YFP PM fluores-cence.

ACKNOWLEDGMENTS

WethankGerdJürgens for sharingmaterials andhelpful discussionson themanuscript. We also thank Karin Schumacher, Malcolm J. Bennett, andBen Scheres for sharing published materials. This work was supported bygrants from the Knut and Alice Wallenberg Foundation, Vetenskapsrådet,Vinnova, and Berzelii. The funders had no role in the study design, datacollection and interpretation, or the decision to submit the work forpublication.

AUTHOR CONTRIBUTIONS

K.J. conceived, designed, drafted, and revised the article, and acquired ,analyzed, and interpreted data. R.P.B. and Y.B. conceived, designed,drafted, and revised the article, and analyzed and interpreted data. D.G.analyzed and interpreted data and drafted the article. R.K.S. acquired,analyzed, and interpreted data.

Received September 26, 2016; revised March 3, 2017; accepted April 19,2017; published April 24, 2017.

REFERENCES

Anders, N., and Jürgens, G. (2008). Large ARF guanine nucleotideexchange factors in membrane trafficking. Cell. Mol. Life Sci. 65:3433–3445.

Asaoka, R., Uemura, T., Ito, J., Fujimoto, M., Ito, E., Ueda, T., andNakano, A. (2013). Arabidopsis RABA1 GTPases are involved intransport between the trans-Golgi network and the plasma mem-brane, and are required for salinity stress tolerance. Plant J. 73:240–249.

Bard, F., and Malhotra, V. (2006). The formation of TGN-to-plasma-membrane transport carriers. Annu. Rev. Cell Dev. Biol.22: 439–455.

Barr, F.A., and Huttner, W.B. (1996). A role for ADP-ribosylationfactor 1, but not COP I, in secretory vesicle biogenesis from thetrans-Golgi network. FEBS Lett. 384: 65–70.

Bleecker, A.B., Estelle, M.A., Somerville, C., and Kende, H. (1988).Insensitivity to ethylene conferred by a dominant mutation in Ara-bidopsis thaliana. Science 241: 1086–1089.

Boutté, Y., Crosnier, M.T., Carraro, N., Traas, J., and Satiat-Jeunemaitre, B. (2006). The plasma membrane recycling pathwayand cell polarity in plants: studies on PIN proteins. J. Cell Sci. 119:1255–1265.

Boutté, Y., Jonsson, K., McFarlane, H.E., Johnson, E., Gendre, D.,Swarup, R., Friml, J., Samuels, L., Robert, S., and Bhalerao, R.P.

(2013). ECHIDNA-mediated post-Golgi trafficking of auxin carriersfor differential cell elongation. Proc. Natl. Acad. Sci. USA 110:16259–16264.

Chen, Y.G., Siddhanta, A., Austin, C.D., Hammond, S.M., Sung,T.C., Frohman, M.A., Morris, A.J., and Shields, D. (1997). Phos-pholipase D stimulates release of nascent secretory vesicles fromthe trans-Golgi network. J. Cell Biol. 138: 495–504.

Chow, C.M., Neto, H., Foucart, C., and Moore, I. (2008). Rab-A2 andRab-A3 GTPases define a trans-golgi endosomal membrane do-main in Arabidopsis that contributes substantially to the cell plate.Plant Cell 20: 101–123.

Dascher, C., and Balch, W.E. (1994). Dominant inhibitory mutants ofARF1 block endoplasmic reticulum to Golgi transport and triggerdisassembly of the Golgi apparatus. J. Biol. Chem. 269: 1437–1448.

Dettmer, J., Hong-Hermesdorf, A., Stierhof, Y.D., and Schumacher,K. (2006). Vacuolar H+-ATPase activity is required for endocytic andsecretory trafficking in Arabidopsis. Plant Cell 18: 715–730.

Dhonukshe, P., Aniento, F., Hwang, I., Robinson, D.G., Mravec, J.,Stierhof, Y.D., and Friml, J. (2007). Clathrin-mediated constitutiveendocytosis of PIN auxin efflux carriers in Arabidopsis. Curr. Biol.17: 520–527.

Donohoe, B.S., Kang, B.H., and Staehelin, L.A. (2007). Identificationand characterization of COPIa- and COPIb-type vesicle classesassociated with plant and algal Golgi. Proc. Natl. Acad. Sci. USA104: 163–168.

D’Souza-Schorey, C., and Chavrier, P. (2006). ARF proteins: roles inmembrane traffic and beyond. Nat. Rev. Mol. Cell Biol. 7: 347–358.

Gallego-Bartolome, J., Arana, M.V., Vandenbussche, F.,Zadnikova, P., Minguet, E.G., Guardiola, V., Van Der Straeten,D., Benkova, E., Alabadi, D., and Blazquez, M.A. (2011). Hierarchyof hormone action controlling apical hook development in Arabi-dopsis. Plant J. 67: 622–634.

Geldner, N., Anders, N., Wolters, H., Keicher, J., Kornberger, W.,Muller, P., Delbarre, A., Ueda, T., Nakano, A., and Jürgens, G.(2003). The Arabidopsis GNOM ARF-GEF mediates endosomal re-cycling, auxin transport, and auxin-dependent plant growth. Cell 112:219–230.

Gendre, D., Jonsson, K., Boutté, Y., and Bhalerao, R.P. (2015).Journey to the cell surface–the central role of the trans-Golgi net-work in plants. Protoplasma 252: 385–398.

Gendre, D., McFarlane, H.E., Johnson, E., Mouille, G., Sjödin, A.,Oh, J., Levesque-Tremblay, G., Watanabe, Y., Samuels, L., andBhalerao, R.P. (2013). Trans-Golgi network localized ECHIDNA/Yptinteracting protein complex is required for the secretion of cell wallpolysaccharides in Arabidopsis. Plant Cell 25: 2633–2646.

Gendre, D., Oh, J., Boutté, Y., Best, J.G., Samuels, L., Nilsson, R.,Uemura, T., Marchant, A., Bennett, M.J., Grebe, M., andBhalerao, R.P. (2011). Conserved Arabidopsis ECHIDNA proteinmediates trans-Golgi-network trafficking and cell elongation. Proc.Natl. Acad. Sci. USA 108: 8048–8053.

Goldberg, J. (1999). Structural and functional analysis of the ARF1-ARFGAP complex reveals a role for coatomer in GTP hydrolysis.Cell 96: 893–902.

Guzmán, P., and Ecker, J.R. (1990). Exploiting the triple response ofArabidopsis to identify ethylene-related mutants. Plant Cell 2: 513–523.

Harper, R.M., Stowe-Evans, E.L., Luesse, D.R., Muto, H.,Tatematsu, K., Watahiki, M.K., Yamamoto, K., and Liscum, E.(2000). The NPH4 locus encodes the auxin response factor ARF7,a conditional regulator of differential growth in aerial Arabidopsistissue. Plant Cell 12: 757–770.

Kang, B.G., Yocum, C.S., Burg, S.P., and Ray, P.M. (1967). Ethyleneand carbon dioxide: mediation of hypocotyl hook-opening re-sponse. Science 156: 958–959.

Ethylene-Mediated Differential Growth by ARF-GEFs 1051

Kleine-Vehn, J., Dhonukshe, P., Swarup, R., Bennett, M., andFriml, J. (2006). Subcellular trafficking of the Arabidopsis auxininflux carrier AUX1 uses a novel pathway distinct from PIN1. PlantCell 18: 3171–3181.

Li, H., Johnson, P., Stepanova, A., Alonso, J.M., and Ecker, J.R.(2004). Convergence of signaling pathways in the control of differ-ential cell growth in Arabidopsis. Dev. Cell 7: 193–204.

Luo, Y., et al. (2015). V-ATPase activity in the TGN/EE is required forexocytosis and recycling in Arabidopsis. Nat. Plants 1: 15094.

McFarlane, H.E., Watanabe, Y., Gendre, D., Carruthers, K.,Levesque-Tremblay, G., Haughn, G.W., Bhalerao, R.P., andSamuels, L. (2013). Cell wall polysaccharides are mislocalized tothe vacuole in echidna mutants. Plant Cell Physiol. 54: 1867–1880.

Nielsen, M., Albrethsen, J., Larsen, F.H., and Skriver, K. (2006). TheArabidopsis ADP-ribosylation factor (ARF) and ARF-like (ARL) sys-tem and its regulation by BIG2, a large ARF-GEF. Plant Sci. 171:707–717.

Pappan, K., Austin-Brown, S., Chapman, K.D., and Wang, X. (1998).Substrate selectivities and lipid modulation of plant phospholipase Dalpha, -beta, and -gamma. Arch. Biochem. Biophys. 353: 131–140.

Peyroche, A., Paris, S., and Jackson, C.L. (1996). Nucleotide ex-change on ARF mediated by yeast Gea1 protein. Nature 384: 479–481.

Peyroche, A., Antonny, B., Robineau, S., Acker, J., Cherfils, J., andJackson, C.L. (1999). Brefeldin A acts to stabilize an abortive ARF-GDP-Sec7 domain protein complex: involvement of specific resi-dues of the Sec7 domain. Mol. Cell 3: 275–285.

Pimpl, P., Hanton, S.L., Taylor, J.P., Pinto-daSilva, L.L., andDenecke, J. (2003). The GTPase ARF1p controls the sequence-specific vacuolar sorting route to the lytic vacuole. Plant Cell 15:1242–1256.

Raz, V., and Koornneef, M. (2001). Cell division activity during apicalhook development. Plant Physiol. 125: 219–226.

Richter, S., Geldner, N., Schrader, J., Wolters, H., Stierhof, Y.D.,Rios, G., Koncz, C., Robinson, D.G., and Jürgens, G. (2007).Functional diversification of closely related ARF-GEFs in proteinsecretion and recycling. Nature 448: 488–492.

Richter, S., Müller, L.M., Stierhof, Y.D., Mayer, U., Takada, N.,Kost, B., Vieten, A., Geldner, N., Koncz, C., and Jürgens, G.(2011). Polarized cell growth in Arabidopsis requires endosomalrecycling mediated by GBF1-related ARF exchange factors. Nat.Cell Biol. 14: 80–86.

Richter, S., Kientz, M., Brumm, S., Nielsen, M.E., Park, M., Gavidia,R., Krause, C., Voss, U., Beckmann, H., Mayer, U., Stierhof, Y.D.,

and Jürgens, G. (2014). Delivery of endocytosed proteins to thecell-division plane requires change of pathway from recycling tosecretion. eLife 3: e02131.

Robinson, D.G., Scheuring, D., Naramoto, S., and Friml, J. (2011).ARF1 localizes to the golgi and the trans-golgi network. Plant Cell23: 846–849, reply 849–850.

Roman, G., Lubarsky, B., Kieber, J.J., Rothenberg, M., and Ecker,J.R. (1995). Genetic analysis of ethylene signal transduction inArabidopsis thaliana: five novel mutant loci integrated into a stressresponse pathway. Genetics 139: 1393–1409.

Silk, W.K., and Erickson, R.O. (1978). Kinematics of HypocotylCurvature. Am. J. Bot. 65: 310–319.

Stefano, G., Renna, L., Chatre, L., Hanton, S.L., Moreau, P.,Hawes, C., and Brandizzi, F. (2006). In tobacco leaf epidermalcells, the integrity of protein export from the endoplasmic reticulumand of ER export sites depends on active COPI machinery. Plant J.46: 95–110.

Swarup, R., et al. (2004). Structure-function analysis of the pre-sumptive Arabidopsis auxin permease AUX1. Plant Cell 16: 3069–3083.

Uemura, T., Suda, Y., Ueda, T., and Nakano, A. (2014). Dynamicbehavior of the trans-golgi network in root tissues of Arabidopsisrevealed by super-resolution live imaging. Plant Cell Physiol. 55:694–703.

Uemura, T., Kim, H., Saito, C., Ebine, K., Ueda, T., Schulze-Lefert,P., and Nakano, A. (2012). Qa-SNAREs localized to the trans-Golginetwork regulate multiple transport pathways and extracellulardisease resistance in plants. Proc. Natl. Acad. Sci. USA 109: 1784–1789.

Vandenbussche, F., Petrásek, J., Zádníková, P., Hoyerová, K.,Pesek, B., Raz, V., Swarup, R., Bennett, M., Zazímalová, E.,Benková, E., and Van Der Straeten, D. (2010). The auxin influxcarriers AUX1 and LAX3 are involved in auxin-ethylene interactionsduring apical hook development in Arabidopsis thaliana seedlings.Development 137: 597–606.

Viotti, C., et al. (2010). Endocytic and secretory traffic in Arabidopsismerge in the trans-Golgi network/early endosome, an independentand highly dynamic organelle. Plant Cell 22: 1344–1357.

Xu, J., and Scheres, B. (2005). Dissection of Arabidopsis ADP-RIBOSYLATION FACTOR 1 function in epidermal cell polarity.Plant Cell 17: 525–536.

Zádníková, P., et al. (2010). Role of PIN-mediated auxin efflux inapical hook development of Arabidopsis thaliana. Development 137:607–617.

1052 The Plant Cell

DOI 10.1105/tpc.16.00743; originally published online April 24, 2017; 2017;29;1039-1052Plant Cell

Kristoffer Jonsson, Yohann Boutté, Rajesh Kumar Singh, Delphine Gendre and Rishikesh P. BhaleraoTrafficking in Arabidopsis

Ethylene Regulates Differential Growth via BIG ARF-GEF-Dependent Post-Golgi Secretory

 This information is current as of October 21, 2020

 

Supplemental Data /content/suppl/2017/05/17/tpc.16.00743.DC2.html /content/suppl/2017/04/24/tpc.16.00743.DC1.html

References /content/29/5/1039.full.html#ref-list-1

This article cites 46 articles, 24 of which can be accessed free at:

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists