Arabidopsis ribosomal proteins control developmentalprograms through translational regulation of auxinresponse factorsAbel Rosado1,2, Ruixi Li1, Wilhelmina van de Ven, Emily Hsu, and Natasha V. Raikhel3

Center for Plant Cell Biology and Department of Botany and Plant Sciences, University of California, Riverside, CA 92521

This contribution is part of the special series of Inaugural Articles by members of the National Academy of Sciences elected in 2012.

Contributed by Natasha V. Raikhel, October 9, 2012 (sent for review August 14, 2012)

Upstream ORFs are elements found in the 5′-leader sequences ofspecific mRNAs that modulate the translation of downstream ORFsencodingmajor geneproducts. InArabidopsis, the translational con-trol of auxin response factors (ARFs) by upstream ORFs has beenproposed as a regulatory mechanism required to respond properlyto complex auxin-signaling inputs. In this study, we identify andcharacterize the aberrant auxin responses in specific ribosomal pro-tein mutants in which multiple ARF transcription factors are simul-taneously repressed at the translational level. This characteristiclends itself to the use of these mutants as genetic tools to bypassthe genetic redundancy among members of the ARF family in Ara-bidopsis. Using this approach, we were able to assign unique func-tions for ARF2, ARF3, and ARF6 in plant development.

ribosomal mutants | uORFs | auxin regulation

The phytohormone auxin plays key roles in the initiation andspecification of postembryonic organs emerging from the

apical meristems, the establishment of the apical-basal axis, andthe control of cell identity throughout plant development (1–3).In Arabidopsis, the response to auxin stimulus is controlled bya signal transduction pathway that includes at least 23 auxinresponse factors (ARFs) and 29 Aux/IAA interacting partnersthat form the complex network that activates or suppresses thetranscription of auxin response genes containing auxin responseelements (4–8). These regulatory elements are responsible fortranslating the local auxin concentration into specific gene ex-pression outputs and developmental responses (9–11).Being central elements for auxin signal transduction, ARF

protein amounts are tightly controlled in a process involvingARF-Aux/IAA protein–protein interactions and an SCF E3ubiquitin ligase-dependent degradation mechanism (5–8, 12–14).Although this regulatory mechanism provides a robust frame-work to understand auxin regulation mediated by the activation/inactivation of ARF proteins, it is limited in terms of explainingthe emerging role of specific components of the translationalmachinery that are required for plant development and properauxin responses (15–18). Translational control has been de-scribed as a regulatory mechanism for multiple plant physio-logical processes, such as polyamine homeostasis (19–21),sucrose sensing (22–24), and anthocyanin biosynthesis (25), butthe role of this regulatory mechanism in development is still notwell established.Translational regulation relies on several control signals usually

located in the 5′ untranslated regions (5′-leader sequences) up-stream of the main ORF (mORF), which can act independently(25) or in a coordinated manner (26–28). One important signalfound in both prokaryotes and eukaryotes is the upstream ORFs(uORFs), which are single or multiple protein-coding elementsoften found in long 5′-leader sequences (>100 nt) that can repressthemORF through two broadmechanisms: ribosomal stalling andinefficient reinitiation (12, 29–31). Different bioinformatic stud-ies have estimated that ∼20–30% of theArabidopsis genes containputative uORFs in their 5′-leader sequences (32–35), and tran-scription factors appear to be the most suitable candidates to be

regulated through this translational mechanism (36). However,information regarding the actual number of functional uORFsand the role of this regulatory mechanism in specific signalingpathways is scarce.Recent studies of a series of ribosome mutants, as well as

mutants with defects in translation reinitiation, have pointedtoward the auxin-signaling pathway as a candidate for trans-lational regulation through a uORF-dependent mechanism (15,16, 37). Specifically, it has been shown using in vivo assays inheterologous cells that multiple ARFs can be translationallyregulated through multiple uORFs located on their 5′-leadersequences (16) and that the ribosomal protein RPL24B and theelongation factor eIF3h are important regulatory elements fortranslational control of the auxin responses (15, 16). Still, suchaspects as whether multiple or single ribosomal proteins are in-volved in the translational regulation of the auxin responses, theanalysis of putative ARF-signaling components regulated throughuORFs in their original developmental context, and the potentialuse of translational mutants for the genetic analysis of the auxinresponses remain largely uncharted territory.In a previous report, we identified the ribosomal protein

RPL4A as an important element for the sorting of vacuolar car-goes in a process regulated by auxins (17). In this work, we ana-lyzed howRPL4A, as well as other ribosomal components, such asRPL4D and RPL5A, modulate auxin responses through thetranslational regulation of multiple ARF-containing 5′-leadersequences in Arabidopsis. We provide in vivo genetic evidencedemonstrating the requirement for the translational regulation ofARFs involving uORFs and for the proper regulation of auxin-mediated developmental programs. We also used ribosomalmutants as genetic tools to bypass ARF redundancy to revealunique functions for ARFs in Arabidopsis development.

ResultsMultiple Ribosomal Proteins Share Common Vacuolar Trafficking andDevelopmental Defects. In a previous genetic screen aimed atidentifying genes involved in protein trafficking toward the vac-uole, we isolated the ribosomal mutant rpl4a-1 from a T-DNA–mutagenized population in the Vac2 background (Vac2 T-DNA). The Vac2 line contains a CLAVATA3 (CLV3) proteinfused with the barley (Hordeum vulgare) lectin C-terminal vac-uolar sorting signal (CLV3:CTPPBL) in the clv3-2 mutant

Author contributions: A.R., R.L., and N.V.R. designed research; A.R., R.L., W.v.d.V., and E.H.performed research; N.V.R. contributed new reagents/analytic tools; A.R., R.L., W.v.d.V., E.H.,and N.V.R. analyzed data; and A.R., R.L., and N.V.R. wrote the paper.

The authors declare no conflict of interest.

Freely available online through the PNAS open access option.1A.R. and R.L. contributed equally to this work.2Present address: Departamento de Biología Molecular y Bioquímica, Universidad de Má-laga, Campus de Teatinos, E-29071 Malaga, Spain.

3To whom correspondence should be addressed. E-mail: natasha.raikhel@ucr.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214774109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1214774109 PNAS | November 27, 2012 | vol. 109 | no. 48 | 19537–19544

PLANTBIOLO

GY

INAUGURA

LART

ICLE

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

background. In this line, the CLV3:CTPPBL fusion protein istargeted to the vacuole and does not complement the clv3-2phenotype. Plants mutated in C-terminal signal (i.e., rpl4a-1)shunt CLV3:T7:CTPPBL to the default secretion pathway,thereby complementing the clv3-2 phenotype (17, 38). In additionto complementation of the clv3-2 phenotype, rpl4a-1 is charac-terized by its partial secretion of different vacuolar-targeted car-goes and aberrant auxin-related developmental responses (17,18). In this study, we tested whether the vacuolar traffickingdefects and aberrant auxin responses were specific for rpl4a-1 orwere a common feature of different components of the ribosome.For that purpose, we performed crosses between the Vac2transgenic line (38) and two ribosomal mutants previously iden-tified by their aberrant auxin responses (39–41). As shown in Fig. 1A and B, the introduction of the rpl4d (SALK_029203) mutant,which is the closest homolog to RPL4A in Arabidopsis (20), andthe rpl5amutant (SALK_089798) in the Vac2 background causedthe reversion of the clv3-2 phenotype; the early termination offloral meristems; and, in some instances (39% for rpl4d and 46%for rpl5a), the formation of pin-like structures (Fig. 1A, Insets).This result suggested that similar to rpl4a, the rpl4d and rpl5amutations induced the secretion of the vacuolar-targeted CLV3:T7:CTPPBL cargo (20).Once we established that rpl4d and rpl5a behaved similarly in

terms of vacuolar cargo secretion, we performed a more detailedphenotypic comparison between them. Early in their develop-ment, bothmutants displayed narrow, pointed first leaves (Fig. 1C,arrowheads) and retarded growth. The mutants also showed de-fective establishment of lateral organ boundaries, which causesmild fusions of the cauline leaves to the stem (Fig. 1D), and dis-played defects in vascular development that deviated from thereticulate pattern of the controls (Fig. 1E).Based on the extensive phenotypic similarities between rpl4d

and rpl5a, as well as with multiple ribosomal mutants indepen-dently reported in the literature (37, 41) (Table S1), we pro-posed that a common regulatory mechanism mediated by thewhole ribosome, rather than by ribosomal proteins acting in-dependently, regulated both vacuolar trafficking and differentdevelopmental programs. Further support for this hypothesis wasprovided by the nonallelic noncomplementation of the two re-cessive rpl4d and rpl5a mutants used in this study. In a previousreport (17), we have shown that the F1 progeny from an rpl4a ×rpl4d cross displays phenotypes similar to the recessive rpl4a andrpl4d parental lines and suggested that RPL4 haploinsufficiencymight be responsible for the observed phenotypes. Interestingly,a similar result was observed when the recessive rpl4d and rpl5amutants were crossed. As shown in Fig. 1F, although the rpl4d and

rpl5a backcrosses with their WT parents generated F1 progenieswith WT features, the F1 progenies from the rpl4d × rpl5a crossbehaved as if they were alleles of the same locus. Thus, the doubleheterozygote rpl4d/RPL4D;RPL5A/rpl5a displayed phenotypessimilar to those of the single homozygous mutants rpl4d/rpl4d orrpl5a/rpl5a. This result reflected a functional connection betweenthe two independent gene products and indicated that the ob-served developmental defects were likely due to changes in theprotein stoichiometry within the ribosome.Finally, our phenotypic analysis showed that although the rpl4d

and rpl5a mutant phenotypes in the Columbia (Col) backgroundwere relativelymild, the introgression of both ribosomalmutationsto a Landsberg erecta (Ler) background greatly increased the se-verity of the observed phenotypes (Fig. 1G). This result suggesteda strong influence of the genetic background on the ribosomalregulatory mechanism controlling those developmental programs.

Ribosomal Mutants Are Altered in Auxin Response Distribution andSensitivity.Because several rpl4d and rpl5a developmental defectshave been described as auxin-related [e.g., pointed leaf, fusedstem, aberrant vascular pattern (39–41)], we next explored theirpotential association with changes in auxin distribution andsensitivity. For that purpose, we analyzed the effect of exogenousauxin on the rpl4d and rpl5a mutants, focusing on well-knownauxin-dependent developmental processes, such as primary rootelongation, lateral root initiation, and shoot architecture.Defects in root elongation and lateral root initiation are typ-

ical characteristics of many mutants involved in auxin homeo-stasis or signaling (42). These features were also observed in therpl4d and rpl5a mutants grown in control conditions (Fig. 2A,Upper). To determine whether exogenously applied auxin couldrestore the defects in root elongation and lateral root de-velopment, rpl4d, rpl5a, and WT plants were germinated andgrown on Murashige and Skoog (MS) medium supplemented with100 nM 1-naphthaleneacetic acid (NAA) for 7 d. As shown in Fig.2A (Lower), the NAA-containing medium promoted lateral rootformation in WT plants; however, such treatment failed to res-cue the root elongation and lateral root development in the rpl4dand rpl5a mutants. Moreover, the NAA treatment inhibited WTroot elongation to a greater extent than it did in rpl4d and rpl5a,indicating that the mutant roots failed to respond properly toexogenously applied NAA.Next, we evaluated these results using histochemical staining

of the auxin-inducible DR5::GUS marker (43). In shoots, WTplants display an auxin maximum at the tip of the first leaf andDR5::GUS expression becomes diffuse around the margins, in-ducing a rounded leaf shape. In contrast, at the tip of the rpl4d

Fig. 1. Ribosomal mutants are defective in vacuo-lar trafficking and display common developmentaldefects. (A) Mutations in rpl4d and rpl5a suppressthe enlarged shoot apical meristem (SAM) pheno-type of the Vac2 (CLV3:CTPP in the clv3-2 back-ground) transgenic line. Adult plant phenotypes,and occasional PIN-like shoot meristems (Insets) areobserved in the Vac2/rpl4d and Vac2/rpl5a back-grounds. (B) Inflorescence phenotypes. (C) Seedlingphenotype displays the pointed first leaves in rpl4dand rpl5a (white arrows). (D) Adult rpl4d and rpl5aplants display mild fusions of the cauline leaves withthe stem (white arrows). (E) rpl4d and rpl5a plantsdisplay aberrant venation patterns. (F) Nonallelicnoncomplementation of the rpl4d and rpl5a mu-tants. (G) rpl4d and rpl5a plants display more severephenotypes in Col-Ler hybrid lines, such as a filamen-tous leaf, fused stem, and pin-formed shoot (whitearrows). (Scale bars: 1 cm.)

19538 | www.pnas.org/cgi/doi/10.1073/pnas.1214774109 Rosado et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

and rpl5a first leaf, the local auxin maximum was weaker andauxin gradients along the leaf margins were absent (Fig. 2B).This apparent altered auxin distribution partially explained theaberrant leaf geometry and vascular development defects ob-served in the mutant backgrounds (44). In the root of untreatedWT seedlings, GUS signal was expressed in the columella, lateralroot cap cells, and lateral root initiation sites, whereas the rpl4dand rpl5a mutants displayed slightly decreased GUS signal inthose tissues (Fig. 2C). After treatment with 100 nM NAA for60 min, the DR5::GUS signal in WT plants was stronger and wasobserved along the stele compared with untreated plants. Insimilar assays, the GUS signal was significantly weaker at theroot apex, stele, and lateral root initiation sites in both rpl4d andrpl5a mutants compared with WT plants, again suggesting a re-duced sensitivity to exogenously applied auxins (Fig. 2C).To analyze the effects of the ribosomal mutations on auxin

responses at the cellular level further, we used confocal laserscanning microscopy and the DR5rev::GFP marker (2). As shownin Fig. S1A, 5-d-old vertically grown WT seedlings displayedhighly localized GFP signal in the quiescent center (QC) andmature columella cells. In contrast, the rpl4a (Ler) and the rpl5a(Col) mutations induced the asymmetrical expansion of DR5-GFP signal in the QC, columella, and lateral root cap cells. Thisresult might be partially explained by the aberrant organizationand development of the columella cells, as indicated by stainingroot structure with styryl dye FM4-64 (<3 min) and the differ-ential starch staining using Lugol (Fig. S1B). Additionally, severephenotypes observed in the rpl4a (Ler) background, such as leafpolarity, geometry defects, and altered gravitropic responses,positively correlated with strong disturbances in the DR5::GUSand DR5rev::GFP signals (Figs. S1 A and B and S2 A–C).These results led us to propose that alterations of auxin fluxes

in the ribosomal mutant roots might be responsible for the ac-cumulation of auxins in the epidermal cell layers and the aber-rant differentiation of the columella cells. To test this hypothesis,we analyzed the polar distribution of the PIN-FORMED auxinefflux carriers PIN1 and PIN2 (45) in the rpl4d mutant back-ground. As shown in Fig. 2 D and E, the polar distribution of thePIN1::PIN1-GFP (46) and PIN2::PIN2-GFP (47) markers wassimilar in rpl4d and the WT controls; however, both PIN1-GFPand PIN2-GFP fluorescent signals (Fig. 2 D and E), as well as the

endogenous PIN1 protein amounts (Fig. S1C), were significantlyreduced in the rpl4d mutant background. In the rpl4a (Ler)background, the aberrant gravitropic responses previously de-scribed for this mutant (17) were also correlated with a strongdecrease of the PIN2-GFP signals, as well as with a reduction ofthe endogenous PIN2 protein content (Fig. S2 D and E). How-ever, AUX1, another auxin transporter (48), displayed similarexpression patterns in rpl4d and rpl5amutants compared with thecontrol (Fig. S2F). These results suggested that the observedroot phenotypes in rpl4d and rpl4a might be explained, at least inpart, by the reduced auxin transport as a consequence of a re-duction in PIN1 and PIN2 proteins but not by a general effect onall the auxin transporters.

Ribosome Regulates ARF5 and ARF7 Protein Quantity Translationally.A simple model to explain the specificity of the auxin-relatedphenotypes in ribosomal mutants is that they decrease rates oftranslation of auxin-response gene transcripts containing uORFs(15, 16) (Fig. 3A). In a previous report, Zhou et al. (16) per-formed a broad bioinformatics search of the Arabidopsis In-formation Resource (TAIR) release 10 database for genes withuORFs in their 5′-leader sequences that might be targets fortranslational regulation. They found that among auxin-relatedgenes, most ARFs harbor multiple uORFs, whereas uORFs wereuncommon among the ARF-interacting partners AUX/IAA,YUCCA biosynthetic genes, TIR1 auxin receptor homologs, andthe PIN family of auxin transporters (16). The enrichment ofuORFs in the ARF 5′-leader sequences, which might eitherencode bioactive peptides with the ability to reduce the trans-lation of the ARFs (Table S2) or regulate their translation ina sequence-independent manner, suggested that these mRNAsmight be specific targets for regulation in rpl4d and rpl5amutants. Therefore, we focused our analysis on different mem-bers of the ARF family of transcription factors.For genes with uORFs in the 5′-leader sequences, translating

the uORFs may cause ribosome stalling, which subsequentlyaffects the efficiency of translation initiation in downstreamgenes (49). Thus, we first tested whether ribosomal stalling wasresponsible for the observed phenotypes in the mutant back-grounds. As shown in Fig. 3B and Fig. S3A, the rpl4d and rpl5amutations only caused minor differences in the polysomal

Fig. 2. rpl mutants have reduced auxin sensitivity and response. (A) Root phenotypes of Col, rpl4d, and rpl5a after 7 d on Murashige and Skoog (MS) mediumwith mock or 100 nM 1-naphthaleneacetic acid (NAA). (B) DR5::GUS staining patterns in Col, rpl4d, and rpl5a first leaf in 3-wk-old plants. (Upper) Detail of the GUSstaining in the apical region of the leaf. (Lower) General view of the leaf shows the changes in leaf morphology and the presence of multiple GUS maxima in themutant backgrounds (arrowheads). (C) rplmutants show reduction of the DR5::GUS signal in both root tips and lateral root initiation sites. PIN1::PIN1GFP (D) andPIN2::PIN2-GFP (E) fluorescent signals are reduced in the rpl4d background. Signal Intensities are quantified by ImageJ and displayed in arbitrary units. Error barsrepresent SDs from 50 seedlings for each genotype, and results were analyzed using the Student t test (***P < 0.001). (Scale bars: A and B, 1 cm; C–E, 100 μm.)

Rosado et al. PNAS | November 27, 2012 | vol. 109 | no. 48 | 19539

PLANTBIOLO

GY

INAUGURA

LART

ICLE

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

profiles compared with the WT control. Thus, the rpl4d and rpl5aprofiles displayed 80S subunit intensity peaks similar to the WT,and no great depletion or accumulation of polysomes from 2 to 5units in size was observed. There was a minor reduction in theaccumulation of a larger number of polysomes in rpl4d, probablydue to cell type-specific effects. This result indicated that ribo-somal stalling was not the main regulatory mechanism affected inrpl4d and rpl5a mutants. Next, we measured the transcriptionallevels of different ARFs and PIN transporters in rpl4d, rpl5a, andWT backgrounds to discriminate between direct targets of theribosome and secondary effects of auxin deregulation. As shown inFig. 3C and Fig. S3B, no significant differences in the transcrip-tional levels of ARFs and PINs were observed between rpl4d andCol, except for PIN2, which showed an approximately 2.8-foldrepression in the mutant background. Interestingly, transcriptionalrepression was also observed for TMO3 and TMO5, two tran-scription factors acting downstream of ARF5 in the auxin-signalingpathway (50) (Fig. S3B). Next, we determined the transcriptionallevels of the same genes in WT and rpl4d isolated polysomalfractions. As shown in Fig. S3B, all analyzed genes followed sim-ilar trends in both total mRNA and polysomal isolated fractions.Thus, the total amounts of PIN2, TMO3, and TMO5 mRNAs, aswell as the mRNAs associated with rpl4d translating ribosomes,were reduced to approximately the same extent compared withWT. Based on these results, we concluded the following: (i) Therpl4d and rpl5a mutations did not cause a general defect in themRNA polysome loading, and (ii) PIN2, TMO3, and TMO5 werenot likely to be direct targets for the translational regulation me-diated by the ribosome owing to their decreased mRNA levels inboth total and polysomal-associated mRNA fractions.Next, we tested whether the translational regulation of two

uORF-containing transcripts (ARF5 and ARF7) that showed nodifferences at the transcriptional level was affected by the ribosome

mutations. For that purpose, we crossed the rpl4d and rpl5amutantswith a transgenic line harboring a promoter ARF5 construct, in-cluding the putative ARF5 uORFs fused to GFP (ARF5::3xGFP)(50) and a line harboring a promoter ARF7 construct including theputative ARF7 uORFs fused with GUS (ARF7::GUS) (51). Asshown in Fig. 3 D and E, translational repression (indicated by thedecrease in theGFP andGUS signals) was induced in the rpl4d andrpl5a mutants compared with their respective controls. This resultwas further validated by Western blot analysis, which showeda decreased protein level of ARF5 and ARF7 in both rpl4d andrpl5a mutants compared with the control (Fig. 3F). However,ARF8, which contains no uORFs, displayed a similar signal in rpl4dand rpl5a mutants when detected by Western blot (Fig. 3F). Theseresults confirmed the presence of a regulatory mechanism at thetranslational level mediated by the ribosome in both ARF5 andARF7 transcripts.

ARF3 uORFs Are Partially Responsible for Several Auxin-RelatedDefects in rpl4d and rpl5a. Having shown that a regulatory mech-anism mediated by ribosomal proteins translationally controlledARF protein content in vivo, we studied the role of uORFs in thisprocess. For that purpose, we transformed the rpl4d and rpl5amutants harboring the DR5::GUS marker (Fig. 2 B–E) witha previously reported ARF3 genomic fragment containing eitherfunctional or nonfunctional uORFs (15) (Fig. 4A). As shown inFig. 4 B–H, transgenic lines harboring the ARF3 genomic frag-ment with nonfunctional uORFs displayed increased DR5::GUSsignal after NAA treatment compared with either rpl mutantalone. On the other hand, transgenic lines harboring the ARF3genomic fragment with functional uORFs showed no changes inDR5::GUS signal after NAA treatment compared with either rplmutant alone. At the phenotypic level, ribosomal mutants har-boring the ARF3 genomic fragment with nonfunctional uORFs

Fig. 3. RPL proteins modulate auxin responsesthrough translational regulation of ARFs. (A) Sche-matic model depicts the three main regulatory stepsperformed by the ribosome during the translationof uORF containing ARFs. (B) Polysomal profiling ofCol and rpl4d obtained by sucrose gradient. (C)Quantitative RT-PCR result shows that the ARFtranscriptional level in total mRNA has no signifi-cant changes in rpl mutants. (D) ARF5::3xGFP signalis reduced in rpl4d and rpl5amutants. (E) ARF7::GUSsignal is reduced in rpl4d and rpl5a mutant back-grounds. (F) Western blot shows that ARF5 andARF7 protein level are reduced in rpl4d and rpl5amutants, whereas ARF8 remains the same. The im-age represents at least three biological replicates.(Scale bars: D, 100 μm; F, 1 cm.)

19540 | www.pnas.org/cgi/doi/10.1073/pnas.1214774109 Rosado et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

displayed significantly longer roots and an enhanced lateral rootnumber compared with either of the rplmutants (Fig. 4 B and C),whereas ribosomal mutants harboring the ARF3 genomic frag-ment with functional uORFs increased the main root length andthe lateral root number to a lesser extent than that observed whenthe ARF3 genomic fragment with nonfunctional uORFs was used(Fig. 4 E–H). These results suggested that a uORF-dependentmechanism causing ARF3 translational defects was partially re-sponsible for some of the auxin-related phenotypes observed inrpl4d and rpl5a. However, similar to the rplmutants, all transgeniclines (with or without functional ARF3 uORFs) showed a pointedleaf phenotype (Fig. 4B andC, Insets) suggesting that other ARFsmight be responsible for this phenotype or that several ARFs acttogether in determining leaf morphology.

Ribosomal Mutants Uncover Unique Functions for ARFs. Our modelfor uORF-mediated translational regulation of ARFs (Fig. 5)predicted that ribosomal mutants would be able to decrease si-multaneously the amount of several ARF proteins causingmultiple auxin-related defects. In accordance with this model, wehypothesized that redundancy among different ARFs might becompromised in the ribosomal mutant backgrounds, and thatribosomal mutants could therefore be used to identify otherwisehidden ARF functions. To investigate this, we performed geneticcrosses between the rpl4d and rpl5a mutants and differentmutants in auxin-responsive factors, such as arf2, arf3, and arf6,and then analyzed the double mutants for phenotypes not ob-served in their single-mutant lines. We found unique phenotypesin the double mutant arf2 rpl5a, such as developmental defects inseedlings (Fig. 6A). The arf2 rpl5a mutants also display pheno-typic variability in the rosette leaves, ranging from slightly en-larged leaves with mildly distorted shapes to plants that showedseverely distorted and round shapes reminiscent of asymmetricalleaf mutants (52) (Fig. 6B). These results implied a uniquefunction for ARF2 in seedling development and leaf polarity

establishment. The arf2 rpl4d double mutant showed mild phe-notypes with a higher rate of distorted and spiral rosette leaves(Fig. 6C). Both arf2 rpl4d and arf2 rpl5a double mutants showedseverely distorted stems (Fig. S4 A–C), highlighting the pre-viously described role of ARF2 in stem development (51) (Fig. S4A–C). The double mutants arf6 rpl4d and arf6 rpl5a showeddistorted and spiral rosette leaves, similar to arf2 rpl4d (Fig. S4 Dand F). Most of the arf6 rpl5a double mutants were smaller thanparental lines (Fig. S4F), and about 10% of them were severelydwarfed (Fig. 6D), indicating that ARF6 function was essentialfor proper development. The double mutant arf3 rpl4a in theLandsberg background showed terminated inflorescences (75%),which highlighted a general role for ARF3 in the regulation ofthe shoot apical meristem development (Fig. 6E). Interestingly,arf3 rpl4a also displayed prominent organ fusions (85%), whichrevealed a unique role for ARF3 in organ boundary establish-ment (Fig. 6F). In those plants, the combination of shoot apicalmeristem deregulation and organ fusions caused enhancedbranching (Fig. 6G) not observed in the individual arf3 (CS8556)or rpl4a single-mutant backgrounds.Together, our results highlighted the importance of the trans-

lational regulation of ARFs through uORFs for proper auxinsignaling and uncovered the ribosomal mutants as useful genetictools to assign unique functions for otherwise redundant ARFs.

DiscussionTranslational Regulation of Auxin Signaling Is Controlled by theRibosome. Throughout the Arabidopsis life cycle, specific geneexpression responses to auxin are mostly determined by inter-actions between signaling components that include the following:six TIR1/AFBs recruiting Aux/IAAs for proteasomal degrada-tion, 29 Aux/IAA repressor proteins, and 23 ARF transcriptionfactors. The different combinatorial interactions, expressionpatterns, and affinities among specific members of the TIR1/AFB-Aux/IAA-ARFs tripartite-signaling assembly provide the

Fig. 4. Transformation with modified ARF3 lacking uORF partially rescues auxin response and root defects in rplmutants. (A) Schematic map displays the twodifferent constructs used for transformation. The constructs harbor the ∼7.2-kb genomic fragment from ARF3 locus with (ARF3 genomic) or without (ARF3mutant) uORFs. The ARF3 mutant version was generated by substitution of the two uORF start codons ATG to TTG. Morphology of 7-d-old rpl4d (B) or rpl5a(C) seedlings transformed with either the ARF3 genomic construct or ARF3 mutant construct (pointed leaves in Insets). (D) DR5::GUS signal in rpl mutant rootstransformed with either the ARF3 genomic construct or ARF3 mutant construct after a 1 h of 100 nM NAA treatment. Root length measurement of 7-d-oldrpl4d (E) or rpl5a (G) seedlings transformed with either the ARF3 genomic construct or ARF3 mutant construct. Quantification of the lateral root number in7-d-old rpl4d (F) or rpl5a (G) plant transformed with either the ARF3 genomic construct or ARF3 mutant construct. (E–H) Forty seedlings for each genotypewere used, and results were analyzed using the Student t test (*P < 0.1; **P < 0.05; ***P < 0.001). (Scale bars: B and C, 1 cm; D, 100 μm.)

Rosado et al. PNAS | November 27, 2012 | vol. 109 | no. 48 | 19541

PLANTBIOLO

GY

INAUGURA

LART

ICLE

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

plants with the plasticity to translate complex auxin signals intoseparated developmental processes (8, 53, 54). In this context,control of the spatiotemporal expression and stability of theregulatory assembly components are essential to trigger adequateexpression profiles from the downstream gene sets.Multiple studies have shown a strong link between auxin

responses and the function of the ribosome. Several ribosomalproteins have been associated with auxin-related developmentalprocesses, such as cell proliferation, cell expansion, and polarityestablishment in leaves (39–41). Auxins also trigger the coordinateexpression of ribosomal proteinmRNAs (55, 56), increase de novoribosomal protein synthesis (57), and change ribosomal proteinphosphorylation patterns (58, 59). In our study, mutations inmultiple ribosomal genes caused specific auxin-related defectsrather than a nonspecific defect in translation. This suggests thatribosomal components are required as a translational regulatorymechanism to amplify auxin signals. To explain the specific role ofthe ribosomal proteins in this regulatory network, we proposeda simplified scheme in which specific pools of auxin-signalingcomponents are required for the regulation of different de-velopmental traits, such as root elongation, lateral root emer-gence, leaf polarity, and boundaries establishment. In this scheme,the auxin stimulus induced the coordinated expression of multipleauxin-signaling and ribosomal components. Next, the ribosome

translationally controls the amounts of each auxin-responsive el-ement using regulatory signals situated in the 5′-leader sequence.This translational mechanism would act as an additional regula-tory step in auxin signaling to decode complex auxin inputsproperly into specific developmental programs.In rpl mutants, alterations in dissociation of the 5′-leader se-

quence regulatory elements from the ribosome or during thetranslational reinitiation of the mORF would likely cause a shift inthe available auxin-signaling elements triggering the observeddevelopmental defects. Therefore, the translational regulation ofauxin-signaling components would be the primary cause for thespecific phenotypes observed in ribosomal mutant backgrounds.Based on results from our study and a previous report (17), wepropose a feedback regulation between the ribosome and auxin-signaling elements, where auxins regulate the expression of ribo-somal components and the ribosome modulates the translation ofauxin-responsive elements. This feedback mechanism might alsobe responsible for the posttranscriptional differences in the auxinresponses reported for different Arabidopsis accessions (60).It is also worthwhile to mention that rpl mutants are defective

in vacuolar sorting, as has been shown previously (17). Therefore,it would be of interest to investigate whether ribosomes are in-volved in nontranscriptional auxin-regulated sorting. Recently,many publications have been devoted to auxin-mediated endo-membrane trafficking independent of the TIR1/AFB nuclear-signaling pathway during the developmental process or in itsresponse to environmental stimuli (61–64). However, the linkbetween the two auxin pathways is still unclear. The fact that rplmutants are defective in both auxin signal transduction andendomembrane trafficking raises the hypothesis that ribosomesmight integrate auxin regulation in both pathways. Becauseribosomes are a molecular protein synthesis factory, it is plausibleto speculate that auxin might efficiently use ribosomes to fulfillglobal and specialized regulation during plant development.

Regulation of ARF Protein Levels Is Mediated Through uORFs. Toelucidate the molecular mechanism underlying the auxin-signalingdefects observed in ribosomal mutants, we focused on specificfeatures among TIR1/AFB-Aux/IAA-ARF mRNAs susceptibleto posttranscriptional regulation by the ribosome. There is a grow-ing body of information in the literature on the posttranscrip-tional regulation of the ARF transcript abundance by micro-RNA(miRNA) and transacting (ta)-siRNA (6, 65–69). These mecha-nisms are used by the plant for rapid removal of ARFs from cellsimmediately following signaling (70). Although it is reasonable tospeculate that ribosomal mutations may alter miRNA- and ta-siRNA–mediated translational repression of ARFs, our resultsshowed similar amounts of both total mRNAand polysomal-loadedARF mRNAs in WT and rpl4d backgrounds, indicating that this isnot the main cause of the observed phenotypes. Thus, we excludedmiRNA- and ta-siRNA–mediated translational repression con-trolled by our ribosomal mutants as a regulatory mechanism.A second model of ARF posttranscriptional regulation is based

on the presence of uORFs that are cis-regulatory elements in the

Fig. 5. Working model in which ribosomal mutants generate ARF-deprivedplants. (A) WT plant senses the auxin stimulus and transcriptionally activatesmultiple ribosomal proteins through an unknown mechanism and uORF-containing ARFs. The ARF mRNAs are translationally regulated through theiruORFs, and the ARF protein amounts required to activate specific de-velopmental programs (colored circles) are obtained. (B) Ribosomal mutantssense the auxin stimulus and transcriptionally activate multiple uORF-con-taining ARFs. Those ARFs are properly associated with polysomes but aretranslationally repressed due to mutations in the ribosome. Whether thosemutations alter the dissociation of the ribosome from the uORF or thesubsequent translational reinitiation of the mORF is still unclear. As a resultof the translational repression of multiple ARFs, the amount of ARF proteinsis reduced (colored circles) and auxin-related phenotypes are observed. Inthis ARF-deprived background, the plants are sensitized and can no longertolerate the loss of a particular ARF. As a result, arf rpl double mutants canbe analyzed to uncover specific ARF protein functions.

Fig. 6. rpl/arf double-mutant combinations un-mask unique functions for ARFs in development. (A–F) (Top) Phenotypes of the single arfmutants used asparental lines. (Middle) Phenotypes of the single rplmutant used as parental lines. (Bottom) Selectedphenotypes observed in the double-mutant combi-nations that were more severe than in the single-mutant parental lines and identify unique roles forknown ARFs in development. For all double-mutantcombinations, a higher degree of phenotypic vari-ability was present compared with their single-mu-tant counterparts, indicating the multigenic natureof the observed developmental outputs. (G) En-hanced branching in the rpl4a arf3 double-mutantcombination in Ler background. (Scale bars: 1 cm.)

19542 | www.pnas.org/cgi/doi/10.1073/pnas.1214774109 Rosado et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

5′-leader sequences of multiple transcription factors (includingARFs) (33), These uORFs can affect their specific interactionswith ribosomal components, and, as such, the translation of themORF (71). Supporting this model, Nishimura et al. (30) showedthat uORFs present in the ARF3 and ARF5 5′-leader sequencesimpede translation of the mORF in transfected protoplasts and inan in vitro translation system. A more recent study (16) confirmedthese results and showed that uORFs are uncommon amongTIR1/AFB and Aux/IAA family members but are a common feature formost of the ARF family members (with striking examples, such asthe seven ormore putative upstreamAUGs identified in theARF5,ARF6, and ARF11 5′-leader sequences). Moreover, evidence forthe direct role of a ribosomal component (RPL24B) in the trans-lational regulation of ARFs by virtue of their specific interactionswith mRNAs harboring uORFs was previously reported (15), andit has been proposed that the coordination between eukaryotictranslation initiation factors and the large ribosomal subunit helpsto fine-tune translation of the ARFs (16).In this study, we showed that multiple rpl mutants reduced the

ARF5 and ARF7 protein amounts without changing theirmRNA levels. This alteration of protein level is uORF-de-pendent because ARF8, which has no uORF, showed a similarprotein level in rpl mutants. Furthermore, the fact that elimi-nation of the ARF3 uORFs partially rescued some of the auxin-related phenotypes in rpl4d and rpl5a backgrounds supports thatmalfunction of the uORF-mediated regulation of ARF3 is atleast partially responsible for the phenotypic defects observed inrplmutants. The weak rescue phenotypes observed in roots couldbe explained by the low expression of ARF3 in this tissue and thesimultaneous disturbance of multiple ARFs in the rpl mutantseven though ARF3 restored normal function. Thus, we proposethat multiple components of the ribosome, such as rpl4a, rpl4d,rpl5a, rpl24b, and other ribosomal mutants displaying aberrantauxin responses, translationally control ARFs through theiruORFs in vivo. In fact, the translational defects of ARF5 andARF7 observed in rpl4d and rpl5d could explain several of theirdevelopmental phenotypes. Thus, the translational defects ofARF5 could be associated with defects in the vascular tissueformation and the root growth (72–74), as well as with thetranscriptional repression of the ARF5 targets TMO3 andTMO5 together with translational repression of the PIN1-GFPmarker. This might increase the tendency for pin-formed floralmeristems and cause the observed reduced auxin accumulationin root meristems (1, 75, 76). On the other hand, the trans-lational defects of ARF7 might be associated with defects in thegravitropic responses observed in the rpl4d and rpl5a ribosomalmutants (74, 77).

Use of Ribosomal Mutants to Assign Unique Functions to ARFs.Multiple studies have provided evidence for functional specif-icities, as well as redundancies, within the ARF family of pro-teins (51, 78–83), which, in our model, would depend onchanges in the composition of ARF–AUX/IAA complexescaused by the arf mutations. In most cases, the single arf mutantphenotype was enhanced in the double mutants. In some cases,

unique phenotypes, such as the delayed formation of lateralroots in the arf7 arf19 double mutant (82) or decreased jasmonicacid concentrations of arf6 arf8 double-mutant combinations(83), were observed. These genetic analyses have establishedpartially redundant functions for related ARFs but also sug-gested that different ARFs are functionally specialized. In thiscomplex scenario, where redundancy, target specificity, andpreferred interactions with Aux/IAA proteins are involved, wepropose an alternative genetic strategy based on the use ofdifferent ribosomal mutants for ARF function determination.In this strategy, ribosomal mutants would enhance defects in

auxin signaling by decreasing the translation of multiple uORF-containing ARFs. Under the condition of rpl mutation, arfmutants are likely to be translationally repressed by upstreamuORFs. Thus, although ARFs are not the only targets of RPLproteins, the information gained from the double mutants couldbe used to uncover hidden ARF functions by increasing the im-portance of an individual arf mutation in the ARF-deprivedbackground. We provide proof of principle that ribosomal mu-tants uncover unique ARF functions, such as the role of ARF2 inseedling development and leaf morphology (rpl5a mutant) andthe role ofARF3 in boundaries establishment (rpl4amutant). Thisstrategy demonstrated that different ribosomal mutants couldinduce different auxin outputs when crossed with the same arfmutant; for example, the double mutant arf3 rpl4a displayed organfusions and enhanced branching, whereas the double mutant arf3rpl4d was characterized by its terminated floral meristems.Thus, in a manner similar to that in which RNAi lines are

widely used to silence gene families at the transcriptional level,we propose that the ribosomal mutant collections could be usedto silence multiple elements of the ARF family at the trans-lational level and to uncover unique functions for ARFs that areusually masked by their redundancy.

Materials and MethodsPlants are grown under standard conditions. Details of plant materials andgrowth conditions, GUS staining, microscopy, RNA isolation, and RT-PCR arefound in the supporting information. Polysome profiling was performedusing freshly ground tissue from 7-d-old seedlings utilizing the differentialcentrifugation protocol described by Mustroph et al. (84).

ACKNOWLEDGMENTS. We thank Dr. Albrecht von Arnim (University ofTennessee) for his expert advice and for sharing the use of his uORF da-tabase, Dr. Dolf Weijers (Wageningen University) for kindly providing theARF5::3xGFP transgenic line, and Dr. Kiyotaka Okada (Kyoto University) forkindly providing the ARF3 genomic and ARF3 mutant constructs. We alsothank Dr. Tom J. Guilfoyle for sharing the ARF5, ARF7, and ARF8 antibodies.We thank Dr. Miguel Ángel Botella (Málaga University), Dr. Zhenbiao Yang,Dr. Julia Bailey-Serres, Dr. Glenn Hicks, and Dr. Mariano Perales (University ofCalifornia, Riverside) for helpful discussions. We also thank J. Brimo (Univer-sity of California, Riverside) for administrative support and L. Johnson fortechnical support. A.R. has received funding from the People Programme(Marie Curie Actions) of the European Union Seventh Framework ProgrammeFP7/2007-2013 under Research Executive Agreement (REA) IIF-2011-301419.This work was funded by the Division of Chemical Sciences, Geosciences, andBiosciences, Office of Basic Energy Sciences of the US Department of Energythrough Grant DE-FG02-02ER15295 to N.V.R.

1. Aida M, Vernoux T, Furutani M, Traas J, Tasaka M (2002) Roles of PIN-FORMED1 andMONOPTEROS in pattern formation of the apical region of the Arabidopsis embryo.Development 129(17):3965–3974.

2. Friml J, et al. (2003) Efflux-dependent auxin gradients establish the apical-basal axis ofArabidopsis. Nature 426(6963):147–153.

3. Shani E, Yanai O, Ori N (2006) The role of hormones in shoot apical meristem func-tion. Curr Opin Plant Biol 9(5):484–489.

4. Teale WD, Paponov IA, Palme K (2006) Auxin in action: Signalling, transport and thecontrol of plant growth and development. Nat Rev Mol Cell Biol 7(11):847–859.

5. Leyser O (2006) Dynamic integration of auxin transport and signalling. Curr Biol 16(11):R424–R433.

6. Guilfoyle TJ, Hagen G (2007) Auxin response factors. Curr Opin Plant Biol 10(5):453–460.

7. Chapman EJ, Estelle M (2009) Mechanism of auxin-regulated gene expression inplants. Annu Rev Genet 43:265–285.

8. Vernoux T, et al. (2011) The auxin signalling network translates dynamic input intorobust patterning at the shoot apex. Mol Syst Biol 7:508.

9. Abel S, Theologis A (1996) Early genes and auxin action. Plant Physiol 111(1):9–17.10. Hagen G, Guilfoyle T (2002) Auxin-responsive gene expression: Genes, promoters and

regulatory factors. Plant Mol Biol 49(3–4):373–385.11. Mockaitis K, Estelle M (2008) Auxin receptors and plant development: A new sig-

naling paradigm. Annu Rev Cell Dev Biol 24:55–80.12. Del Bianco M, Kepinski S (2011) Context, specificity, and self-organization in auxin

response. Cold Spring Harb Perspect Biol 3(1):a001578.13. Calderón Villalobos LI, et al. (2012) A combinatorial TIR1/AFB-Aux/IAA co-receptor

system for differential sensing of auxin. Nat Chem Biol 8(5):477–485.14. Vanneste S, Friml J (2012) Plant signaling: Deconstructing auxin sensing. Nat Chem

Biol 8(5):415–416.15. Nishimura T, Wada T, Okada K (2004) A key factor of translation reinitiation, ribo-

somal protein L24, is involved in gynoecium development in Arabidopsis. Biochem SocTrans 32(Pt 4):611–613.

16. Zhou F, Roy B, von Arnim AG (2010) Translation reinitiation and development arecompromised in similar ways by mutations in translation initiation factor eIF3h andthe ribosomal protein RPL24. BMC Plant Biol 10:193.

Rosado et al. PNAS | November 27, 2012 | vol. 109 | no. 48 | 19543

PLANTBIOLO

GY

INAUGURA

LART

ICLE

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021

17. Rosado A, et al. (2010) Auxin-mediated ribosomal biogenesis regulates vacuolartrafficking in Arabidopsis. Plant Cell 22(1):143–158.

18. Rosado A, Raikhel NV (2010) Application of the gene dosage balance hypothesis toauxin-related ribosomal mutants in Arabidopsis. Plant Signal Behav 5(4):450–452.

19. Hanfrey C, et al. (2005) A dual upstream open reading frame-based autor-egulatory circuit controlling polyamine-responsive translation. J Biol Chem 280(47):39229–39237.

20. Hu WW, Gong H, Pua EC (2005) The pivotal roles of the plant S-adenosylmethioninedecarboxylase 5′ untranslated leader sequence in regulation of gene expression atthe transcriptional and posttranscriptional levels. Plant Physiol 138(1):276–286.

21. Ivanov IP, Atkins JF, Michael AJ (2010) A profusion of upstream open reading framemechanisms in polyamine-responsive translational regulation. Nucleic Acids Res 38(2):353–359.

22. Wiese A, Elzinga N, Wobbes B, Smeekens S (2004) A conserved upstream openreading frame mediates sucrose-induced repression of translation. Plant Cell 16(7):1717–1729.

23. Rahmani F, et al. (2009) Sucrose control of translation mediated by an upstream openreading frame-encoded peptide. Plant Physiol 150(3):1356–1367.

24. Hummel M, Rahmani F, Smeekens S, Hanson J (2009) Sucrose-mediated translationalcontrol. Ann Bot (Lond) 104(1):1–7.

25. Wang L, Wessler SR (2001) Role of mRNA secondary structure in translational re-pression of the maize transcriptional activator Lc(1,2). Plant Physiol 125(3):1380–1387.

26. Franceschetti M, et al. (2001) Characterization of monocot and dicot plant S-adenosyl-l-methionine decarboxylase gene families including identification in the mRNA ofa highly conserved pair of upstream overlapping open reading frames. Biochem J 353(Pt 2):403–409.

27. Jin X, Turcott E, Englehardt S, Mize GJ, Morris DR (2003) The two upstream openreading frames of oncogene mdm2 have different translational regulatory proper-ties. J Biol Chem 278(28):25716–25721.

28. Yaman I, et al. (2003) The zipper model of translational control: A small upstream ORFis the switch that controls structural remodeling of an mRNA leader. Cell 113(4):519–531.

29. Gaba A, Wang Z, Krishnamoorthy T, Hinnebusch AG, Sachs MS (2001) Physical evi-dence for distinct mechanisms of translational control by upstream open readingframes. EMBO J 20(22):6453–6463.

30. Nishimura T, Wada T, Yamamoto KT, Okada K (2005) The Arabidopsis STV1 protein,responsible for translation reinitiation, is required for auxin-mediated gynoeciumpatterning. Plant Cell 17(11):2940–2953.

31. Roy B, et al. (2010) The h subunit of eIF3 promotes reinitiation competence duringtranslation of mRNAs harboring upstream open reading frames. RNA 16(4):748–761.

32. Munzarová V, et al. (2011) Translation reinitiation relies on the interaction betweeneIF3a/TIF32 and progressively folded cis-acting mRNA elements preceding shortuORFs. PLoS Genet 7(7):e1002137.

33. Tran MK, Schultz CJ, Baumann U (2008) Conserved upstream open reading frames inhigher plants. BMC Genomics 9:361.

34. Vaughn JN, Ellingson SR, Mignone F, Arnim Av (2012) Known and novel post-tran-scriptional regulatory sequences are conserved across plant families. RNA 18(3):368–384.

35. Takahashi H, Takahashi A, Naito S, Onouchi H (2012) BAIUCAS: A novel BLAST-basedalgorithm for the identification of upstream open reading frames with conservedamino acid sequences and its application to the Arabidopsis thaliana genome. Bio-informatics 28(17):2231–2241, 10.1093/bioinformatics/bts303.

36. Hayden CA, Jorgensen RA (2007) Identification of novel conserved peptide uORFhomology groups in Arabidopsis and rice reveals ancient eukaryotic origin of selectgroups and preferential association with transcription factor-encoding genes. BMCBiol 5:32.

37. Horiguchi G, Van Lijsebettens M, Candela H, Micol JL, Tsukaya H (2012) Ribosomesand translation in plant developmental control. Plant Sci 191–192:24–34.

38. Sohn EJ, et al. (2007) The shoot meristem identity gene TFL1 is involved in flowerdevelopment and trafficking to the protein storage vacuole. Proc Natl Acad Sci USA104(47):18801–18806.

39. Pinon V, et al. (2008) Three PIGGYBACK genes that specifically influence leaf pat-terning encode ribosomal proteins. Development 135(7):1315–1324.

40. Fujikura U, Horiguchi G, Ponce MR, Micol JL, Tsukaya H (2009) Coordination of cellproliferation and cell expansion mediated by ribosome-related processes in the leavesof Arabidopsis thaliana. Plant J 59(3):499–508.

41. Byrne ME (2009) A role for the ribosome in development. Trends Plant Sci 14(9):512–519.

42. Péret B, et al. (2009) Arabidopsis lateral root development: An emerging story. TrendsPlant Sci 14(7):399–408.

43. Ulmasov T, Murfett J, Hagen G, Guilfoyle TJ (1997) Aux/IAA proteins repress expres-sion of reporter genes containing natural and highly active synthetic auxin responseelements. Plant Cell 9(11):1963–1971.

44. Scarpella E, Marcos D, Friml J, Berleth T (2006) Control of leaf vascular patterning bypolar auxin transport. Genes Dev 20(8):1015–1027.

45. Petrásek J, et al. (2006) PIN proteins perform a rate-limiting function in cellular auxinefflux. Science 312(5775):914–918.

46. Benková E, et al. (2003) Local, efflux-dependent auxin gradients as a commonmodulefor plant organ formation. Cell 115(5):591–602.

47. Xu J, Scheres B (2005) Dissection of Arabidopsis ADP-RIBOSYLATION FACTOR 1function in epidermal cell polarity. Plant Cell 17(2):525–536.

48. Swarup R, et al. (2005) Root gravitropism requires lateral root cap and epidermal cellsfor transport and response to a mobile auxin signal. Nat Cell Biol 7(11):1057–1065.

49. Fernandez J, et al. (2005) Ribosome stalling regulates IRES-mediated translation ineukaryotes, a parallel to prokaryotic attenuation. Mol Cell 17(3):405–416.

50. Schlereth A, et al. (2010) MONOPTEROS controls embryonic root initiation by regu-lating a mobile transcription factor. Nature 464(7290):913–916.

51. Okushima Y, Mitina I, Quach HL, Theologis A (2005) AUXIN RESPONSE FACTOR 2(ARF2): A pleiotropic developmental regulator. Plant J 43(1):29–46.

52. Byrne ME, et al. (2000) Asymmetric leaves1 mediates leaf patterning and stem cellfunction in Arabidopsis. Nature 408(6815):967–971.

53. Lokerse AS, Weijers D (2009) Auxin enters the matrix—Assembly of response ma-chineries for specific outputs. Curr Opin Plant Biol 12(5):520–526.

54. Hayashi K (2012) The interaction and integration of auxin signaling components.Plant Cell Physiol 53(6):965–975.

55. Gantt JS, Key JL (1985) Coordinate expression of ribosomal protein mRNAs followingauxin treatment of soybean hypocotyls. J Biol Chem 260(10):6175–6181.

56. Key JL, et al. (1986) Auxin-regulated gene expression. Philos Trans R Soc Lond B BiolSci 314(1166):427–440.

57. Beltrán-Peña E, Aguilar R, Ortíz-López A, Dinkova TD, De Jiménez ES (2002) Auxinstimulates S6 ribosomal protein phosphorylation in maize thereby affecting proteinsynthesis regulation. Physiol Plant 115(2):291–297.

58. Pérez L, Aguilar R, Méndez AP, de Jiménez ES (1990) Phosphorylation of ribosomalproteins induced by auxins in maize embryonic tissues. Plant Physiol 94(3):1270–1275.

59. Turck F, Zilbermann F, Kozma SC, Thomas G, Nagy F (2004) Phytohormones partici-pate in an S6 kinase signal transduction pathway in Arabidopsis. Plant Physiol 134(4):1527–1535.

60. Delker C, et al. (2010) Natural variation of transcriptional auxin response networks inArabidopsis thaliana. Plant Cell 22(7):2184–2200.

61. Robert S, et al. (2010) ABP1 mediates auxin inhibition of clathrin-dependent endo-cytosis in Arabidopsis. Cell 143(1):111–121.

62. Xu T, et al. (2010) Cell surface- and rho GTPase-based auxin signaling controls cellularinterdigitation in Arabidopsis. Cell 143(1):99–110.

63. Paciorek T, et al. (2005) Auxin inhibits endocytosis and promotes its own efflux fromcells. Nature 435(7046):1251–1256.

64. Abas L, et al. (2006) Intracellular trafficking and proteolysis of the Arabidopsis auxin-efflux facilitator PIN2 are involved in root gravitropism. Nat Cell Biol 8(3):249–256,and correction (2006) 8(4):424.

65. Williams L, Carles CC, Osmont KS, Fletcher JC (2005) A database analysis methodidentifies an endogenous trans-acting short-interfering RNA that targets the Arabi-dopsis ARF2, ARF3, and ARF4 genes. Proc Natl Acad Sci USA 102(27):9703–9708.

66. Wu MF, Tian Q, Reed JW (2006) Arabidopsis microRNA167 controls patterns of ARF6and ARF8 expression, and regulates both female and male reproduction. De-velopment 133(21):4211–4218.

67. Garcia D, Collier SA, Byrne ME, Martienssen RA (2006) Specification of leaf polarity inArabidopsis via the trans-acting siRNA pathway. Curr Biol 16(9):933–938.

68. Hunter C, et al. (2006) Trans-acting siRNA-mediated repression of ETTIN and ARF4regulates heteroblasty in Arabidopsis. Development 133(15):2973–2981.

69. Xing H, et al. (2011) Genome-wide identification and expression profiling of auxinresponse factor (ARF) gene family in maize. BMC Genomics 12:178.

70. Axtell MJ, Bartel DP (2005) Antiquity of microRNAs and their targets in land plants.Plant Cell 17(6):1658–1673.

71. Xue S, Barna M (2012) Specialized ribosomes: A new frontier in gene regulation andorganismal biology. Nat Rev Mol Cell Biol 13(6):355–369.

72. Hardtke CS, Berleth T (1998) The Arabidopsis gene MONOPTEROS encodes a tran-scription factor mediating embryo axis formation and vascular development. EMBO J17(5):1405–1411.

73. Mattsson J, Ckurshumova W, Berleth T (2003) Auxin signaling in Arabidopsis leafvascular development. Plant Physiol 131(3):1327–1339.

74. Hardtke CS, et al. (2004) Overlapping and non-redundant functions of the Arabi-dopsis auxin response factors MONOPTEROS and NONPHOTOTROPIC HYPOCOTYL 4.Development 131(5):1089–1100.

75. Wisniewska J, et al. (2006) Polar PIN localization directs auxin flow in plants. Science312(5775):883.

76. Cole M, et al. (2009) DORNROSCHEN is a direct target of the auxin response factorMONOPTEROS in the Arabidopsis embryo. Development 136(10):1643–1651.

77. Harper RM, et al. (2000) The NPH4 locus encodes the auxin response factor ARF7,a conditional regulator of differential growth in aerial Arabidopsis tissue. Plant Cell12(5):757–770.

78. Sessions A, et al. (1997) ETTIN patterns the Arabidopsis floral meristem and re-productive organs. Development 124(22):4481–4491.

79. Stowe-Evans EL, Harper RM, Motchoulski AV, Liscum E (1998) NPH4, a conditionalmodulator of auxin-dependent differential growth responses in Arabidopsis. PlantPhysiol 118(4):1265–1275.

80. Goetz M, Vivian-Smith A, Johnson SD, Koltunow AM (2006) AUXIN RESPONSE FAC-TOR8 is a negative regulator of fruit initiation in Arabidopsis. Plant Cell 18(8):1873–1886.

81. Li J, Dai X, Zhao Y (2006) A role for auxin response factor 19 in auxin and ethylenesignaling in Arabidopsis. Plant Physiol 140(3):899–908.

82. Okushima Y, et al. (2005) Functional genomic analysis of the AUXIN RESPONSEFACTOR gene family members in Arabidopsis thaliana: Unique and overlappingfunctions of ARF7 and ARF19. Plant Cell 17(2):444–463.

83. Nagpal P, et al. (2005) Auxin response factors ARF6 and ARF8 promote jasmonic acidproduction and flower maturation. Development 132(18):4107–4118.

84. Mustroph A, Juntawong P, Bailey-Serres J (2009) Isolation of plant polysomal mRNAby differential centrifugation and ribosome immunopurification methods. MethodsMol Biol 553:109–126.

19544 | www.pnas.org/cgi/doi/10.1073/pnas.1214774109 Rosado et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

6, 2

021