Phloem-Mobile Aux/IAA Transcripts Target to the Root Tip and Modify Root Architecture


  • Michitaka Notaguchi,

    1. Department of Plant Biology, College of Biological Sciences, University of California, Davis, CA 95616, USA
    Search for more papers by this author
    • Present address: JST ERATO Higashiyama Live-Holonics Project, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, auAichi 464-8602, Japan

  • Shmuel Wolf,

    1. The Robert H. Smith Institute of Plant Sciences and Genetics in Agriculture and the Otto Warburg Minerva Center for Agricultural Biotechnology, The Hebrew University of Jerusalem, The Robert H. Smith Faculty of Agriculture, Food and Environment, Rehovot 76100, Israel
    Search for more papers by this author
  • William J. Lucas

    Corresponding author
    1. Department of Plant Biology, College of Biological Sciences, University of California, Davis, CA 95616, USA
    Search for more papers by this author

Tel: +1 530 752 1093; Fax: +1 530 752 5410; E-mail:


In plants, the phloem is the component of the vascular system that delivers nutrients and transmits signals from mature leaves to developing sink tissues. Recent studies have identified proteins, mRNA, and small RNA within the phloem sap of several plant species. It is now of considerable interest to elucidate the biological functions of these potential long-distance signal agents, to further our understanding of how plants coordinate their developmental programs at the whole-plant level. In this study, we developed a strategy for the functional analysis of phloem-mobile mRNA by focusing on IAA transcripts, whose mobility has previously been reported in melon (Cucumis melo cv. Hale's Best Jumbo). Indoleacetic acid (IAA) proteins are key transcriptional regulators of auxin signaling, and are involved in a broad range of developmental processes including root development. We used a combination of vasculature-enriched sampling and hetero-grafting techniques to identify IAA18 and IAA28 as phloem-mobile transcripts in the model plant Arabidopsis thaliana. Micro-grafting experiments were used to confirm that these IAA transcripts, which are generated in vascular tissues of mature leaves, are then transported into the root system where they negatively regulate lateral root formation. Based on these findings, we present a model in which auxin distribution, in combination with phloem-mobile Aux/IAA transcripts, can determine the sites of auxin action.


Recent phloem sap analyses, performed on a broad range of plant species, have revealed that large numbers of macromolecules, including proteins, mRNA, and small RNA, are contained within the phloem translocation system (Xoconostle-Cázares et  al. 1999; Lough and Lucas 2006; Le Hir et  al. 2008; Turgeon and Wolf 2009; Atkins et  al. 2011). These macromolecules could potentially serve as signaling agents to convey developmental or environmental condition information detected in mature source leaves to distantly-located sink organs, including the shoot and root apical meristems. Evidence in support of this notion has been obtained by studies aimed at identifying the florigenic signal(s) that transmits the photoperiodic condition through the phloem to mediate the vegetative-to-floral transition. Here, the florigenic agent was shown to be a protein, FT/Hd3a (Corbesier et  al. 2007; Tamaki et  al. 2007), which is generated in source companion cells and moves through the sieve tube system (Lin et  al. 2007) to upregulate the developmental pathway involved in flowering (Wigge 2011; Turnbull 2011).

The capacity of the phloem to deliver RNA species has also been established through studies on viral systemic infection (Lough and Lucas 2006), as well as studies on endogenous mRNA, such as GAIP (Haywood et  al. 2005), PFP-LeT6 (Kim et  al. 2001), and BEL5 (Banerjee et  al. 2006), and small RNA (Kehr and Buhtz 2008; Molnar et  al. 2010). These examples heighten the importance of the discovery that the sieve tube system contains a large and unique population of RNA species (Lough and Lucas 2006; Kehr and Buhtz 2008). By elucidating the roles of these various RNA species as candidate signaling agents in plant growth and development, pathogen defense, and environmental signaling, we stand to gain important insights into the evolution of the plant signaling system that operates to coordinate events at the organismal level.

Importantly, the phloem translocation stream is also known to carry phytohormones such as auxin, cytokinin, gibberellins, jasmonates, and methylsalicylates (Allen and Baker 1980; Hoad 1995; Kamboj et  al. 1998; Corbesier et  al. 2003; Rocher et  al. 2006; Chhun et  al. 2007; Jager et  al. 2007; Truman et  al. 2007) that mediate in regulating a wide array of developmental and defense programs (Chapman and Estelle 2009; Vanneste and Friml 2009; Vlot et  al. 2009; Kudo et  al. 2010). In this regard, the recent discovery that phloem sap contains transcripts for members of the Aux/IAA gene family that encode key transcriptional regulators in auxin signaling (Omid et  al. 2007) is highly intriguing. Because auxin controls the degradation of the encoded Aux/indoleacetic acid (IAA)proteins, this raises the question as to why plants would have evolved the capacity to carry both auxin and the Aux/IAA transcripts together through phloem.

To investigate this apparent conundrum, an experimental system was developed to identify the Arabidopsis thaliana phloem-mobile Aux/IAA transcripts. Application of micro-grafting techniques then allowed us to identify the target location in which these transcripts act within the root tip to contribute to the general control over root architecture.


Screening for phloem-mobile IAA transcripts in A. thaliana

Our initial experiments were aimed at confirming the earlier observation that the melon (Cucumis melo cv. Hale's Best Jumbo) phloem translocation stream carries IAA transcripts for contig F-308 (accession number EB715302) and contig F-571 (accession number EB714898) (Omid et  al. 2007). Reverse transcription-polymerase chain reaction (RT-PCR) experiments showed that elevated F-308 and F-571 mRNA levels were present in isolated leaf vasculature tissues (prepared according to Endo et  al. 2005), compared with mature melon leaves (Figure  1A and Figure S1A available online). Furthermore, both transcripts were highly enriched in the melon phloem sap collected from stems. Using a melon (stock):pumpkin (scion) hetero-grafting system, we showed that F-308 transcripts moved across the graft union, via the phloem, from the melon stock into the pumpkin (Cucurbita maxima cv. Big Max) scion (Figure  1B). In contrast to findings from Omid et  al. (2007), F-571 transcripts were not detected in phloem sap collected from these pumpkin scions (Figure S1B).

Figure 1.

Identification of Arabidopsis thaliana phloem-mobile IAA transcripts. (A) Detection of F-308 transcripts in leaves (Leaf), isolated leaf vasculature tissue (VAS) and phloem sap (PS) from melon plants. Four genes were used as the reference for Leaf and VAS assays (Figure S1A). CmePP16L was used as an internal control. (B) Ability of F-308 transcripts to move long-distance in the phloem. Hetero-grafting experiments between melon (Cme) and pumpkin (Cm) confirmed the movement of F-308 transcripts across the melon:pumpkin graft union. CmPP16 was used as an internal control (Xoconostle-Cázares et  al. 1999).(C) Phylogenic tree constructed using A. thaliana, cucumber (Huang et  al. 2009), and melon (Omid et  al. 2007) IAA gene sequences using the neighbor-joining method. Numbers above the branches indicate bootstrap support as a percentage. Normalized expression values for transcripts detected in the phloem sap from cucumber (Csa) and watermelon (Cla) (Guo et  al. 2012), are shown to the right (see details in Table S1). IAA transcripts having enriched expression in VAS, and those shown to be phloem-mobile, are highlighted in pink and blue, respectively.(D) Analysis of vasculature-enriched total RNA isolated from mature A. thaliana leaves. Six IAA transcripts (highlighted in pink in C), but not IAA14, were preferentially detected in the VAS compared to whole leaves. FT served as a control for a known VAS-enriched transcript (Corbesier et  al. 2007; Tamaki et  al. 2007; Turnbull 2011; Wigge 2011), and ACT2 was used as a control.(E) Hetero-grafted A. thaliana (At) stock and Nicotiana benthamiana (Nb) scion. Arrowhead indicates the site of the graft union. Scale bar =  5  cm.(F) Test for long-distance transport of IAA transcripts from the At stock into the Nb scion. Intact At and Nb, and Nb scions grafted to At stocks were analyzed. GAI and β-ACT were used as controls.

To perform a functional analysis of phloem-mobile IAA transcripts via a molecular genetics approach, studies were next performed on A. thaliana. A phylogenetic analysis of the IAA gene family identified IAA7 and IAA14/SOLITARY-ROOT (SLR) as the closest A. thaliana homologs to F-308 (Figure  1C). To test whether transcripts for these two genes also accumulate preferentially in leaf vascular tissue, mRNA was extracted from the tissue (Endo et  al. 2005) for analysis (Figure S2A, B). As neither IAA7 nor IAA14 transcripts were enriched in these vascular tissues, micro-grafting studies (Turnbull et  al. 2002) were performed to test whether the mRNA for a gain-of-function mutant of IAA14 (slr-1) (Fukaki et  al. 2002) moves a long distance through the phloem. Because these transcripts did not move across the graft union (Figure S3), we turned our attention to six IAA genes whose expression was enriched in the leaf vascular tissue (Figure  1D and Figure S2C). Interestingly, inspection of a cucumber, watermelon, and pumpkin phloem transcriptome database (Guo et  al. 2012) identified enriched transcripts that differed from those in A. thaliana (Figure  1C, right, and Table S1).

To obtain direct evidence of long-distance transport, a new stem-grafting system was developed with A. thaliana as the donor stock and Nicotiana benthamiana as the recipient scion (Figure  1E and Figure S4). Analysis of mRNA extracted from hetero-grafted N. benthamiana scions confirmed the long-distance delivery of A. thaliana GAI transcripts (Haywood et  al. 2005; Figure  1F). Importantly, strong IAA18 and IAA28 signals were also detected in the N. benthamiana scions (Figure  1F).

IAA18 and IAA28 are expressed in the vasculature tissue of mature leaves

To establish the tissue distribution of phloem-mobile IAA transcripts, RT-PCR analyses were performed on 2-week-old plants. At this developmental stage, phloem-based long-distance signaling has been shown to occur, as reported in flowering control studies (Kardailsky et  al. 1999; Kobayashi et  al. 1999; Notaguchi et  al. 2009). Cotyledons and fully-expanded leaves, representing source organs, together with apical and root tissues as sink organs were analyzed (Figure  2A). Consistent with previous reports (Rogg et  al. 2001; Uehara et  al. 2008; Ploense et  al. 2009; De Rybel et  al. 2010), both transcripts were detected in all organs, although IAA18 expression in roots was lower than that in aerial organs, whereas IAA28 expression reflected the opposite situation.

Figure 2.

Phloem-mobile IAA transcripts are expressed in the vasculature of Arabidopsis thaliana mature leaves. (A)IAA18 and IAA28 expression in cotyledons (CL), true leaves (TL), the shoot apex, and roots of 2-week-old plants.(B, C) GUS expression analysis in mature leaves of PIAA18::GFP-GUS (upper) and PIAA28::GFP-GUS (lower) transgenic plants.(B) Strong expression was observed in the vasculature tissues of the CL (1-week-old) and TL (2-week-old).(C) Cross sections of the mid veins in TL. Ph, Phloem; Xy, Xylem. Scale bars =  2  mm (B) and 10 μm (C).

Transgenic plants harboring β-glucuronidase (GUS) fused to the 3.2  kb and 3.7  kb upstream regions of IAA18 and IAA28, respectively, were analyzed to identify cellular expression domains for these genes. In both plant lines, a strong GUS signal was observed in leaf vascular tissues (Figure  2B). Transverse sections of the mid-vein indicated that IAA18 is expressed in both phloem and xylem cells, whereas for IAA28, a GUS signal was detected in the xylem (Figure  2C). Similar patterns were observed for the minor veins.

Graft-transmission of dominant mutant iaa18 and iaa28 activity reduced lateral root formation

To address the function of phloem-mobile IAA18 and IAA28 transcripts, their loss-of-function mutants iaa18–3 (GK-800H03-025387) and iaa28–2 (SALK_129988C) were analyzed in terms of root and shoot developmental phenotypes (Figure S5). No significant differences were observed in primary root cellular arrangement, cell division activity, lateral root formation, lateral root number, or general root architecture. These findings held true for both iaa18–3 and iaa28–2 mutants, as well as for the iaa18;iaa28 double mutant. In addition, there was no change in target gene expression in the root systems of these mutant lines compared with wild-type plants. These studies indicate the likely presence of gene redundancy for control over lateral root formation.

Earlier studies established that dominant IAA18 and IAA28 mutants, in which these proteins are not targeted for degradation by the 26S proteasome, exhibit stable repressor activities in the auxin signaling cascade (Chapman and Estelle 2009; Vanneste and Friml 2009; Figure  3A). The dominant IAA18 mutant crane-2 (Uehara et  al. 2008), the dominant IAA28 mutant iaa28–1 (Rogg et  al. 2001), and the closest F-308 homolog slr-1, all showed severe defects in lateral root formation, consistent with their important function in root development (Rogg et  al. 2001; Fukaki et  al. 2002; Uehara et  al. 2008; Péret et  al. 2009; Figures 3B, C and Figure S3B). Interestingly, wild-type scions grafted to crane-2 stocks failed to show an altered floral phenotype (Figure S6). The crane-2 and iaa28–1dominant mutants (hereafter referred to as diaa18 and diaa28, respectively) were next used to explore the role of their phloem-mobile transcripts in lateral root development.

Figure 3.

Long-distance action of diaa18 and diaa28 mutant scions reduces lateral root formation in wild-type rootstock. (A) Auxin signaling cascade in wild-type and diaa18 mutant plants in the presence (shaded) or absence (unshaded) of exogenous auxin.(B, C) Grafting studies performed with diaa18(B) and diaa28(C) mutants using wild-type (Col or Ws) rootstocks. Number of lateral roots in a 3  cm region back from the root tip, counted 17  d after grafting of the indicated combination of scions:rootstocks. Intact Col/Ws and diaa18/diaa28 mutant plants served as controls. For exogenous auxin experiments, an equivalent set of plants was transferred to medium containing 1 μM indoleacetic acid (IAA) 14  d after grafting, and analyzed 3  d later. Mean ± SE; asterisks indicate values that differ significantly (see Tables S2 and S3).

Micro-grafting experiments were performed to test whether the ability of the diaa18 to suppress lateral root formation is transmitted from the scion to the rootstock. A combination of homo- (Col-0:Col-0 and Ws-0:Ws-0) and hetero- (diaa18:Col-0 and diaa28:Ws-0) grafts were established using 4-day-old wild-type and 5-day-old mutant seedlings, with the graft junction located at the hypocotyl region. Based on earlier studies showing that functional graft unions are established by 14  d after grafting (Notaguchi et  al. 2009), the number of emerged lateral roots located within a 3  cm region back from the root tip was noted 17  d post-grafting. Wild-type rootstocks grafted to diaa18 scions had fewer lateral roots compared with wild-type homo-grafted plants (Figure  3B and Table S2). Grafting experiments using the iaa18;iaa28 double mutant as the recipient rootstock grafted to the diaa18 scion did not result in enhancement in the suppression of lateral root formation (Figure S7 and Table S3). Collectively, these results provided support for the hypothesis that the activity of diaa18 to suppress lateral root formation is transmitted through a phloem graft junction. Here, it is important to note that grafting, per se, did not cause any reduction in lateral root formation, as the number of lateral roots was the same in homo-grafted and intact wild-type plants (Figure  3B and Table S2).

Parallel experiments were also performed in which the grafted plants were treated with exogenous auxin. This hormone treatment promotes pericycle cell division, resulting in a significant increase in the number of lateral roots. This upregulation involves the degradation of IAA18 and other IAA proteins, resulting in the release of auxin response factor (ARF) transcription factors from their IAA repressors (Laskowski et  al. 1995; Uehara et  al. 2008; Figure  3A). Importantly, under the application of auxin, a significant reduction in the number of lateral roots was still observed in the diaa18:Col-0 hetero-grafting studies (Figure  3B and Table S2). This finding provided further support for the hypothesis that the action of diaa18 is graft-transmissible through phloem.

Similar grafting experiments were next conducted with diaa28. In contrast to the findings in diaa18, no effect on lateral root number was observed in the diaa28:Ws-0 hetero-grafting experiments (Figure  3C and Table S4). However, under the application of auxin, a small decrease in lateral root number was obtained compared with Ws-0:Ws-0 homo-grafted plants. These results may well reflect the expression pattern of IAA28/diaa28 in the xylem tissues of mature leaves.

IAA18 and IAA28 transcripts, but not their proteins, are delivered long-distance into the root tip

To ascertain the nature of the long-distance signal, we next tested whether the diaa18 and IAA28 transcripts can move from the scion, across the graft union, into the recipient rootstock. For these experiments, the iaa18;iaa28 double loss-of-function mutant was used as the recipient rootstock, and nested-PCR was used to amplify potentially imported diaa18 and IAA28 transcripts from the scions (Figure  4). Importantly, diaa18 transcripts were detected in all grafted rootstocks analyzed; however, IAA28 transcripts were not consistently detected in the iaa18;iaa28 rootstocks (Figure  4A), indicating a low level of transcript delivery into the roots.

Figure 4.

Long-distance transport of IAA18/diaa18 and IAA28 transcripts from donor scions into iaa18;iaa28 double mutant rootstocks.Shoot wild-type RNA (total) samples diluted (104- and 105-fold) with total RNA from iaa18;iaa28 double mutant shoots served as a positive control. For a negative control, RNA extracted from roots of homo-grafted or intact iaa18;iaa28 double mutant plants was used. Whole rootstocks (A) and 3–4  mm of primary and secondary lateral root tips (B) were analyzed.

It is well known that plants in nitrogen or phosphate-limited conditions either produce more lateral roots or upregulate lateral root growth (Nibau et  al. 2008; Figure S8A). A nutrient deprivation experiment was thus performed to test whether such responses to limited nutrients are triggered by a reduction of IAA18/IAA28 gene expression or attenuation of IAA18/IAA28 transcript delivery through the phloem. Nutrient withdrawal did not alter the expression patterns for IAA18 and IAA28 (Figure S8B, C). Interestingly, although lateral root growth was enhanced in these experiments, both diaa18 and IAA28 transcripts could still be detected in the rootstock (Figure S8D). Thus, the response to nutrient limitation seems to be an independent mechanism from that of the IAA18/IAA28 phloem delivery system.

Lateral roots are initiated from founder cells generated in the basal meristem of the root tip (Fukaki and Tasaka 2009; Péret et  al. 2009). In light of this fact, we analyzed both primary and lateral root tips (3–4  mm) taken from Col-0 (scion):iaa18;iaa28 (rootstock) hetero-grafted plants. Transcripts for both IAA18 and IAA28, derived from the grafted scions, were detected in these excised primary and lateral root tips (Figure  4B). These studies, taken together with our analyses performed on entire rootstocks, unambiguously established that IAA18/diaa18 transcripts are targeted to the root, including the meristematic region of both primary and lateral roots.

A heterologous system was next used to test whether IAA18 and/or IAA28 proteins can enter the phloem sieve tube system. Here, a Zucchini yellow mosaic virus (ZYMV) vector was used to generate diaa18–4×Myc and diaa28–4×Myc proteins in pumpkin leaves (Hsu et  al. 2004; Ma et  al. 2010). Protein was isolated from stem vascular bundles and phloem sap, and Western analysis was performed using a c-Myc monoclonal antibody. FT protein was used as an internal control for a phloem-mobile protein (Lin et  al. 2007), and RbcS served as a control against phloem sap contamination from surrounding cells. Although these ZYMV-expressed proteins were present in the vascular tissue, only FT was detected in the phloem translocation stream (Figure  5). These results indicate that, in pumpkin, neither IAA18 nor IAA28 protein has the capacity to enter the sieve tube system. By extrapolation, we believe that these two proteins are unlikely to function as long-distance signaling agents in A. thaliana. This notion is consistent with the unstable nature of these proteins (Rogg and Bartel 2001).

Figure 5.

Ability of IAA18-Myc and IAA28-Myc to enter the sieve tube system was tested using a viral vector system.Pumpkin plants were inoculated with Zucchini yellow mosaic virus (ZYMV) carrying diaa18–4xMyc, diaa28–4xMyc, or FT-4Myc constructs. Phloem sap and stem vasculature tissue collected from these ZYMV-infected plants was analyzed by Western blot using an anti-c-Myc monoclonal antibody (top). Ponceau-S staining was used as the loading control (bottom). Note that FT-Myc but not IAA18-Myc or IAA28-Myc was detected in the phloem sap.

Model for IAA18 regulation of lateral root formation

Analysis of 8-d-old PIAA18::GFP-GUS seedlings revealed that GUS staining was present along the root axis and within the phloem pole of transverse sections taken from the mature region of the primary root (Figure  6A). However, sections taken from the terminal 3–5  mm region of the primary root tip were devoid of a GUS signal (Figure  6A). It is well established that auxin is delivered from the shoot to the root via the phloem (Hoad 1995; Vanneste and Friml 2009), and that it begins to accumulate in the transition zone between the meristem and elongation zone of the root tip (Brunoud et  al. 2012). It has also been reported that the level of cytokinin, an auxin antagonist, is low in the basal meristem zone (Miyawaki et  al. 2004; Kieffer et  al. 2010). Our studies established that the IAA18 gene is not expressed in this basal meristem zone, but that IAA18 transcripts are delivered via the phloem into this same region of the root (Figure  4B). These findings thus support a model in which phloem delivery of IAA18 transcripts, auxin, and cytokinin to the phloem unloading domain (Imlau et  al. 1999; Stadler et  al. 2005) serves to coordinate the site(s) of auxin action that controls lateral root formation (Figure  6B).

Figure 6.

Shoot-derived IAA transcripts are transported into the root tip and contribute to the regulation of lateral root patterning. (A) Expression profile of IAA18 established using PIAA18::GFP-GUS transgenic plants. A GUS signal was observed in root primordia (arrowheads) and the mature region of the root. Note that a GUS signal was not detected in the root tip (bracket). Cross-sections show a GUS signal in the phloem tissues at site 1, but absence of a GUS signal at site 2. Scale bars =  1  cm and 20 μm (insets).(B) Model for the role of phloem-mobile IAA18 transcripts in root development. Phloem-based signaling to regulate lateral root formation in the root tip is illustrated. Together with auxin and its antagonist, cytokinin, IAA18 transcripts are transported down the root axis through the phloem (meta sieve element, mSE), and enter the root tip. In the basal meristem, oscillations in the PDR5::GUS auxin report reflect the regular establishment of lateral root initiation sites within xylem pole pericycle (XPP) cells (De Smet et  al. 2007; Moreno-Risueno et  al. 2010). This is accomplished by an interaction between IAA18 and cytokinin as negative repressors, and auxin as the positive regulator, to yield periodic initiation of lateral root primordia. Here, it should be noted that we cannot discount the involvement of additional phloem-mobile IAA signaling agents, including IAA28.

What cells in the root tip are targeted by IAA18 transcripts?

Phloem delivers IAA18 transcripts into the primary and lateral root unloading zone, and presumably from there, these transcripts need to move cell-to-cell through plasmodesmata (PD) to reach target cells. To investigate this pathway, transgenic PIAA18::GFP-GUS plants were used to probe the expression pattern along the plant axis. A strong GUS signal could be detected in leaf vascular tissues, with a relatively weaker signal detected within the root (Figures 2B and 6A). A 2-fluorescent protein reporter system was next developed, based on the GAL4-VP16 (GV) and GAL4-target (UAS) promoter unit, to identify cells within which the imported IAA18 transcripts were being translated. A GFPER reporter was used to monitor promoter activity, while an RFPER reporter was developed to visualize the spatial distribution of the expressed test transcript (Figure S9A). To examine movement in the root tip, a modified trigger was used, and RFPER was used as the reporter for promoter activity, while GFPER served to indicate transcript distribution (Figure S9B).

These reporter systems were tested by examining their activities in both trichome accessory cells of mature leaves (as these cells can be readily visualized by confocal scanning laser microscopy, CSLM; Figure S9A) and the root tip (Figure S9B). In both tissues examined, the fluorescent patterns for the promoter and the IAA18 transcripts were identical. These results indicate that either the protein machinery required for the delivery of these transcripts and their subsequent movement through PD may be absent, or that such trafficking is prevented in these cell types.


In this study, we developed an experimental system for the identification of phloem-mobile transcripts in A. thaliana (Figure  1 and Figures S2–S4). The utility of this hetero-grafting system is that it allows one to test for the mobility of transcripts without the need to generate transgenic plants. Application of this method provides the means to extrapolate from known mobile transcripts in melon in order to identify A. thaliana members of this Aux/IAA gene family that separately evolved to function as long-distance signaling agents in auxin-based signaling.

At the protein level, members of the Aux/IAA gene family are highly conserved (Paponov et  al. 2009; Figure S10). Therefore, it is interesting to consider the underlying evolutionary mechanism that allowed different IAA gene family members to acquire the capacity to move through the phloem translocation stream of melon and A. thaliana (Figure  1C and Table S1). It is most likely that specific cis-sequence/structural motifs were acquired to impart transcripts with the capacity to both enter and exit the sieve tube system. Identification of the putative cis-acting genetic elements that, in various plant species, function to mediate phloem-mobility of Aux/IAA transcripts, will offer important insights into the underlying evolutionary mechanism(s) that led to expression of IAA transcripts within the vascular tissues of source leaves, and these transcripts’ subsequent movement into the root system, where they cooperate to negatively regulate lateral root formation (Figures 3 and 5).

In this study, we identified phloem-mobile transcripts that were delivered to a region of the plant in which their expression was not detected (Figure  4 and 6A). In this regard, the IAA18 system differs from previously-characterized phloem-mobile transcripts in that earlier studies found the genes to be expressed in both mature source leaves and in the target tissues (Kim et  al. 2001; Haywood et  al. 2005; Banerjee et  al. 2006). In the situation where there is overlap between the target tissue and gene expression, imported transcripts may function either to give rise to a dosage-effect, or may participate in some level of epigenetic regulation.

For this study, the IAA18 transcripts imported into the primary and lateral root tips were derived from either the source leaves (Figure  4) or the phloem located along the translocation pathway (Figure  6A). The developmental process underlying lateral root initiation is known to occur within the xylem pole pericycle (XPP) (De Smet et  al. 2007; Moreno-Risueno et  al. 2010). Our research establishes that the import of IAA18 transcripts into the root tip affects the extent to which lateral roots develop (Figure  3). Thus, it seems logical to assume that the IAA18-based ribonucleoprotein (RNP) complex would need to move through the PD-interconnecting cells from the phloem unloading zone to the XPP (Figure  6B). Based on our cell targeting studies (Figure S9), it would appear that a cellular regulatory mechanism may operate – along the non-cell-autonomous pathway – to exert tight control over post-phloem trafficking of the IAA18 transcripts.

In conclusion, our research provides the experimental basis for a new concept in which auxin distribution, in combination with phloem-mobile Aux/IAA transcripts, can determine the site(s) of auxin action. The challenge ahead will be to identify the molecular components of the IAA18-RNP complex, and the mechanism underlying the post-phloem trafficking of this RNP complex to the target tissue(s) where IAA18 protein contributes to the auxin-based regulation of lateral root formation.

Materials and Methods

Plant materials and growth conditions

Pumpkin (Cucurbita maxima cv. Big Max) and melon (Cucumis melo cv. Hale's Best Jumbo) plants were grown on soil in a growth chamber under long-day conditions (16  h light at 28 °C/8  h dark at 20 °C). Arabidopsis thaliana plants were grown in a growth chamber under long-day conditions (16  h light/8  h dark) at 22 °C, with white fluorescent lights (100 μmol/m2  per s). Nicotiana benthamiana plants were grown under the same photoperiod as A. thaliana, but the growth temperature and light intensity were 27 °C and 150 μmol/m2  per s, respectively. Nutrients were supplied daily as described on the UC Davis Controlled Environment Facility website (UC Davis Controlled Environment Facility 2012). A. thaliana ecotype Columbia-0 (Col-0) and Wassilewskija-0 (Ws-0) were used as the wild-types. slr-1 (Fukaki et  al. 2002) and crane-2 (Uehara et  al. 2008) were obtained from H. Fukaki. iaa28–1 (Rogg et  al. 2001) was obtained from B. Bartel. T-DNA insertion alleles of IAA genes in Col-0 background (SALK_013158, SALK_122796, SALK_129988, SALK_138286) were received from the Arabidopsis Biological Resource Center (ABRC, USA), and one allele (GK-800H03) in Col-0 background was received from the Nottingham Arabidopsis Stock Center (NASC, UK). Ws-0 background alleles (FLAG_454G12, FLAG_121D4) were received from INRA-Versailles Genomic Resource Center (France) (Samson et  al. 2002) (Figure S5A). Each homozygous line was isolated by PCR genotyping, using T-DNA border markers. The GAL4 enhancer trap line was obtained from the NASC. For expression and phenotypic analyses in A. thaliana, seeds were surface-sterilized and germinated on 1/2 × Murashige and Skoog (MS) medium.

Plasmid construction and plant transformation

All plasmid constructs for plant transformation were generated using standard molecular biology procedures. For PIAA18::GFP-GUS and PIAA28::GFP-GUS, an upstream fragment of IAA18 (3  211  bp) or IAA28 (3  760  bp) was amplified from genomic DNA by PCR, and then cloned into pENTR D-TOPO (Invitrogen, Grand Island, NY, USA). Note that these promoter fragments are longer than those used in earlier GUS reporter studies in which expression was not observed in shoots (Rogg et  al. 2001; Uehara et  al. 2008). Transfer of the DNA fragment from the entry clone to pBGWFS7 (Karimi et  al. 2002) was performed by an LR reaction using LR Clonase (Invitrogen). Recombinant pBGWFS7s were selected on media containing 100 μg/mL spectinomycin.

The reporter systems for transcript visualization were constructed using GATEWAY binary vectors pGWBN1–7 originating from pGWB2 (Nakagawa et  al. 2007), or pTA7002 (Aoyama and Chua 1997). For pGWBN1, a fragment from attR1 to attR2 of pGWB2, including the chloramphenicol-resistance marker (Cmr) chloramphenicol acetyl transferase for selection in bacteria and the ccdB negative selection marker in bacteria, was amplified by PCR with Np1 primers and cloned into pCR-Blunt II-TOPO (Invitrogen). The fragment was digested by SbfI and PmeI and introduced into the same sites as pTA7002, resulting in the removal of the 35S promoter from pTA7002. For pGWBN2, a fragment of mCherryER was amplified from ER-rb (Nelson et  al. 2007) by PCR with Np2 primers, and cloned into pCR-Blunt II-TOPO. The fragment was digested by SpeI and introduced into the same site as pGWBN1. For pGWBN3, a fragment of PUAS::mGFP5ER was amplified from pZY375 (Rim et  al. 2009) by PCR with Np3 primers, and cloned into pCR-Blunt II-TOPO. The fragment was digested by HindIII and SpeI and introduced into the HindIII and XbaI sites of pGWB2, resulting in the removal of the 35S promoter from pGWB2. For pGWBN4, a fragment of PUAS::mCherryER was amplified from pGWBN2 by PCR with Np4 primers, and cloned into pCR-Blunt II-TOPO. The fragment was digested by HindIII and NheI and introduced into the HindIII and XbaI sites of pGWB2, resulting in the removal of the 35S promoter from pGWB2.

The final plasmids were obtained by LR reaction with the generated destination vectors. For PDR5::GVG;PUAS::RFPER, a fragment of eDR5 was amplified from cloning plasmid pATM-DR5_LUCplus (Covington and Harmer 2007) by PCR with Np5 primers, and cloned into pENTR D-TOPO. Transfer of the DNA fragment from the entry clone to pGWBN2 was performed by LR reaction. For PUAS::GFPER-IAA18 and PUAS::RFPER-IAA18, a fragment of IAA18 gene with untranslated regions (UTRs) was amplified from genomic DNA by PCR with Np6 primers, and cloned into pENTR D-TOPO. Transfer of the DNA fragment from the entry clone to pGWBN3 and pGWBN4 was performed by LR reaction. Other final plasmids shown in Figure S9 were similarly constructed using fragments of IAA28, GAI, F-308, and CmPP16. Recombinant pGWBN2s were selected on media containing 50 μg/mL kanamycin. Recombinant pGWBN3s and pGWBN4s were selected on media containing 50 μg/mL kanamycin and 50 μg/mL hygromycin. Primer information is provided in Table S5.

The constructs described above, in binary vectors, were introduced into Agrobacterium tumefaciens strain GV3101, and transformed into A. thaliana plants ecotype Col using the floral-dip procedure (Clough and Bent 1998).

Grafting procedures

Hetero-grafting between melon stock and pumpkin recipient scion was carried out as follows. Stocks were prepared by removing the shoot apical region of 2-w-old melon plants, leaving two to three leaves to allow for new lateral bud growth from the axial meristem. A centrally-located, 2–3  cm long slit was then made into the cut stem surface. Scions were prepared by making a wedge-shaped cut at the hypocotyl-root junction of 1-w-old pumpkin seedlings. After graft assembly by inserting the scion into the 2–3  cm long slit of the stock, the graft site was sealed with Parafilm. The scion and graft site were then covered with a clear plastic bag for the next 7  d. Phloem sap was harvested from the pumpkin scions 4  weeks after grafting.

Hetero-grafting between A. thaliana ecotype Col-0 stock and N. benthamiana scion was carried out as follows. Stocks were prepared by removing the growing axillary buds and apical region of the inflorescence from the shoots of 4-w-old bolting A. thaliana plants, leaving 7–10  cm of the primary stem. A 1–2  cm long slit was then made in a central position on the cut surface. Scions were prepared from an inflorescence stem of 3 to 4-w-old N. benthamiana plants by making a wedge-shaped cut at the bottom of the excised stem. After graft assembly by inserting the scion into the slit of the stock, the graft site was sealed with Parafilm and the scion and graft site were covered with a clear plastic bag for the next 7  d. The N. benthamiana scions were excised 4  w after grafting and used for RNA extraction. The same procedures were used for the homo-grafting experiments described in Figure S4C.

A. thaliana I-shaped (Figure S4D, upper panel) and Y-shaped (Figure S6) micro-grafting were performed as previously described (Turnbull et  al. 2002; Notaguchi et  al. 2009), with modifications. In brief, a wedge-shaped graft was assembled on the hypocotyl of 4-d-old seedlings grown on 1/2 × MS medium containing 0.8% agar at 22 °C. For dominant iaa mutants, 5-d-old seedlings were used in parallel grafting studies. These grafting procedures were carried out on nylon membranes and seedlings were grown for 5–6  d on 1/2 × MS medium containing 1.5% agar at 27 °C. Grafted seedlings were subsequently transferred to 1/2 × MS medium containing 0.8% agar and were grown at 22  °C until they were used for analysis. I-shaped homo-grafts of 6-d-old N. benthamiana seedlings, and 4-d-old A. thaliana scions hetero-grafted to 6-d-old N. benthamiana rootstocks were performed using the same procedures as described above (Figure S4D, middle and lower panels).

Analysis of lateral roots

The apical part of the primary root was excised and placed on an agar plate. Terminal 3  cm regions of the roots were analyzed, under a binocular microscope, for the presence of emerged lateral roots.

RT-PCR analyses

Total RNA was extracted using TRIzol reagent (Invitrogen) and then treated with RNase-free DNaseI (Invitrogen), according to the manufacturer's instructions. cDNA was synthesized in a 20 μL RT reaction mixture containing 0.5 μg RNA as a template, using Superscript III (Invitrogen). After RT, the mixture was diluted with 30 μL of glass-distilled water, and 1-μL aliquots were used for PCR assays. For nested PCR, 1-μL aliquots of the first PCR products were used for the second PCR. Primers and PCR conditions are provided in Table S3. PCR products were resolved by electrophoresis on agarose or polyacrylamide gels, and visualized by ethidium bromide staining. For the grafted samples tested in Figures 1B, F, Figure 4, and Figure S8D, the amplification of target sequences was confirmed by direct sequencing of the PCR products.

Melon vasculature tissues were hand stripped from mature leaves. Phloem sap was collected from incisions made in the stem located directly beneath the apex, and exudates were transferred to TRIzol LS reagent (Invitrogen). Vasculature-enriched samples of mature A. thaliana leaves were isolated essentially as described previously (Endo et  al. 2005). For the dilution series of wild-type shoot RNA samples in Figure  4, aliquots of wild-type shoot total RNA were diluted with varying amounts of iaa18;iaa28 shoot total RNA. Procedures followed to synthesize cDNA from these RNA samples were carried out as described above.

GUS expression analyses

For GUS staining, tissues were fixed with 90% acetone for 10  min on ice, rinsed three times with 50  mM sodium phosphate buffer, infiltrated for 15  min with staining solution (1.0  mg/mL X-Gluc, 50  mM sodium phosphate buffer, pH 7.0, 5  mM potassium ferrocyanide, 5  mM potassium ferricyanide, 0.2% Triton X-100) under vacuum, and subsequently incubated for 3–24  h at 37  °C in the dark. Leaf and root tissues were dehydrated through a graded ethanol series and embedded in Technovit 7100 (Heraeus Kulzer, Germany), and 8 μm sections were subsequently prepared using a microtome (HM340E; Microm, Germany).

Confocal scanning laser microscopy

Confocal scanning laser microscopy (CSLM) analyses of the GFP and RFP fluorescent signals generated by the trigger and reporters 1 and 2 shown in Figure S9A, and the RFP and GFP signals generated by the trigger and reporters 1 and 2 shown in Figure S9B, were carried out as follows. Plant lines used in these studies were generated by crossing a M0169 GAL4 driver line (NASC 2012) harboring a “trigger” and “reporter 1” with a series of PUAS::RFPER transgenic plants referred to “reporter 2” (Figure S9A, upper panel). 2-w-old seedlings were used in these experiments (Figure S9A, lower panel). For the second reporter system, plant lines were generated by crossing a PDR5::GVG;PUAS::RFPER transgenic plant harboring a “trigger” and “reporter 1” with a series of PUAS::GFPER transgenic plants referred to “reporter 2” (Figure S9B, upper panel). For root analyses, hydroponically grown 9-d-old seedlings were treated overnight with 1 μM dexamethasone (DEX), and roots were then analyzed (Figure S9B, lower panel).

EdU-staining (Kotogány et  al. 2010) was performed according to the manufacturer's protocol for the Click-iT Edu Alexa Fluor 488 HCS Assay (Invitrogen).

All fluorescent signals were analyzed by CSLM (model DM RXE 6 TCS-SP2 AOBS; Leica Microsystems, Germany) using an Ar/ArKr laser.

Zucchini yellow mosaic virus infection system

Polymerase chain reaction fragments encoding A. thaliana RbcS, FT, and the diaa18 and diaa28 mutants each harboring a single nucleotide substitution, were transferred into the SphI and KpnI site of the Zucchini yellow mosaic virus (ZYMV)-based viral vector (Hsu et  al. 2004; Ma et  al. 2010). All viral constructs were verified by sequencing. Microprojectile bombardment was performed as described previously (Ma et  al. 2010). The ZYMV-based viral vectors were precipitated onto 1-μm gold particles, which were then bombarded onto cotyledons of 8-d-old pumpkin seedlings. Phloem sap and stem vasculature tissues were harvested 2 w after bombardment. For Western blotting analysis, anti-c-Myc monoclonal antibody (1:5  000; Sigma, St. Louis, MO, USA) and anti-mouse IgG antibody conjugated with horseradish peroxidase (1:5  000; Sigma) were used as the first and secondary antibody reactions, respectively. Immunodetection was performed with chemiluminescence reagent ECL-plus (GE Healthcare, Pittsburgh, PA, USA) and X-ray film (Kodak Biomax MS; Kodak, Japan).

(Co-Editor: Chun-Ming Liu)


We would like to thank Drs Hidehiro Fukaki, Bonnie Bartel, Tsuyoshi Nakagawa, Jae-Yean Kim, and Stacy Harmer, as well as the ABRC, the NASC, and the INRA for materials. We are thankful to Dr. Yasunori Ichihashi for helpful discussion, Dr. Takashi Akagi for advice on phylogenic analysis, and Elizabeth Faulkner for plant care and technical assistance. This work was supported by a grant from the United States-Israel Binational Science Foundation (BSF 2007052; to W.J.L. and S.W.), and by a Postdoctoral Fellowship for Research Abroad from the Japanese Society for the Promotion of Science (awarded to Michitaka Notaguchi).

Author Contributions

W.J.L. and S.W. conceived of the project, M.N. and W.J.L. developed the experimental plan, M.N. performed the experiments, M.N. and W.J.L. wrote the manuscript, and S.W. was an advisor for the study.