Multiple MONOPTEROS -dependent pathways are involved in leaf initiation

Initiation of leaves at the flanks of the shoot apical meristem occurs at sites of auxin accumulation and pronounced expression of auxin-inducible PIN genes, suggesting a feedback loop to progressively focus auxin in concrete spots. Since PIN expression is regulated by Auxin Response Factor (ARF) activity, including MONOPTEROS (MP) , it appeared possible that MP affects leaf formation as a positive regulator of PIN genes and auxin transport. Here we analyze a novel, completely leafless phenotype arising from simultaneous interference with both auxin signaling and auxin transport. We show that mp pin1 double mutants, as well as mp mutants treated with auxin-efflux inhibitors, display synergistic abnormalities, not seen in wild type regardless of how strongly auxin transport was reduced. The synergism of abnormalities indicates that the role of MP in shoot meristem organization is not limited to auxin transport regulation. In mp mutant background, auxin transport inhibition completely abolishes leaf formation. Instead of forming leaves, the abnormal shoot meristems dramatically increase in size harboring correspondingly enlarged expression domains of CLAVATA3 and SHOOTMERISTEMLESS , molecular markers for the central stem cell zone and the complete meristem respectively. The observed synergism under conditions of auxin efflux inhibition was further supported by an unrestricted PIN1 expression in mp meristems, as compared to a partial restriction in wildtype meristems. Auxin transport-inhibited mp meristems also lacked detectable auxin maxima. We conclude that MP promotes the focusing of auxin and leaf initiation in part through pathways not affected by auxin efflux inhibitors.


INTRODUCTION
Plants continuously produce lateral organs, primarily leaves and flowers, at the flanks of shoot apical meristems (SAMs). Considerable advances have been made over the past ten years on the understanding of the genetic basis of meristem maintenance, proliferation and lateral organ formation (see recent reviews by (Williams and Fletcher, 2005;Carraro et al., 2006;Shani et al., 2006;Tucker and Laux, 2007). At the center of the meristem, a central zone (CZ) of generally less frequently dividing cells provides cells for the more frequently dividing cells in the surrounding peripheral zone (PZ) and underlying rib zone (RZ) (Reddy et al., 2004). Together, the CZ, PZ and RZ make up the shoot apical meristem. The meristem provides cells for lateral organ, i.e. leaf and flower formation and underlying pith formation. The size of the central zone is regulated by a feedback loop involving the CLAVATA genes and the WUSCHEL (WUS) gene (Fletcher et al., 1999;Schoof et al., 2000;Clark, 2001). The meristem is specified and maintained by the involved in lateral organ formation (Aida et al., 1997;Aida et al., 1999;Hibara et al., 2003;Vroemen et al., 2003;Koyama et al., 2007).
It has long been known that the formation of lateral organs can be influenced by the plant hormone auxin (Reinhardt et al., 2000 and references therein). Application of auxin as well as auxin efflux inhibitors results in a range of phenotypes from altered numbers and positions of flowers and leaves to a complete block of flower formation from reproductive SAMs (Wardlaw, 1949;Meicenheimer, 1981;Okada et al., 1991;Mattsson et al., 1999). Recent advances suggest that auxin accumulation is required for lateral organ initiation and that auxin is transported to these sites by membrane bound efflux transport proteins which polarly localize to apical or basal ends of cells (Benkova et al., 2003;Reinhardt et al., 2003;Friml et al., 2004;Heisler et al., 2005;Petrasek et al., 2006). A key component in this process is PIN-FORMED1 (PIN1), a member of the PIN family of membrane bound auxin efflux proteins (Okada et al., 1991;Galweiler et al., 1998). Loss of function mutations in the PIN1 gene result in reduced auxin transport, and defective cotyledon and flower formation (Okada et al., 1991). Petrasek et al (2006) have recently shown that PIN auxin efflux proteins are sufficient to facilitate auxin efflux in yeast cells, suggesting that directionality of auxin flow can be regulated by the subcellular localization of PIN proteins. The PINOID (PID) gene, encoding a proteinserine/threonine kinase, acts as a positive regulator of polar auxin transport by regulating the sub-cellular localization of PIN1 (Bennett, 1995;Benjamins et al., 2001;Friml et al., 2004;Lee and Cho, 2006). Loss of function pid mutants display defects in lateral organ formation similar to pin1 mutants consistent with its role in regulating PIN1 mediated auxin efflux.
Auxin transport is promoted by the activity of the MONOPTEROS (MP) gene (Wenzel et al., 2007), which belongs to the Auxin Response Factor (ARF) family of transcription factors (Guilfoyle et al., 1998;Hardtke and Berleth, 1998). Members of this family are post-translationally activated in response to auxin via auxin-mediated degradation of members of the AUX/IAA family of nuclear repressor proteins that bind to ARFs and inhibit ARF dimerization and subsequent target gene transcription (Kim et al., 1997;Ulmasov et al., 1997;Leyser and Berleth, 1999;Ulmasov et al., 1999;Dharmasiri and Estelle, 2002;Liscum and Reed, 2002). Not only mutations in PIN1 and PID but also in the MP gene interfere with lateral organ formation on inflorescence meristems (Przemeck et al., 1996). Local auxin application can restore flower formation on the flanks of pin1 and pid, but not mp mutant inflorescences (Reinhardt et al., 2000;Reinhardt et al., 2003), suggesting that in mp mutants not the local supply of auxin, but auxin sensitivity is diminished. Similarly, cotyledon response assays show that mp mutants are more resistant to the effects of exogenous auxin treatments than the strong auxin resistant mutant allele axr1-12 demonstrating that mp mutants are severely defective in auxin signaling (Mattsson et al., 2003).
Recent reports show that ARFs, including MP, may regulate the expression of PIN genes (Sauer et al., 2006;Wenzel et al., 2007). To test if MP exerts its effect on lateral organ formation exclusively as a regulator of PIN genes and auxin transport, we created mp pin1 double mutants and also grew mp mutants on media supplemented with auxin efflux inhibitors. Here we show that mp pin1 double mutants, as well as mp mutants treated with auxin efflux inhibitors, display strong synergistic abnormalities.
These mutants fail to develop any lateral organs and the SAM develops into a leafless dome. The appearance of a synergistic defect indicates that the role of MP in shoot meristem organization is not limited to the regulation of auxin transport and the novel meristem phenotype implicates auxin transport and signaling in the regulation of meristem size.

mp pin1 double mutants fail to form leaves
The shoot meristems of both pin1 and mp single mutants produce a functional rosette of leaves from the vegetative SAM but are highly defective in the analogous process of flower formation from the reproductive SAM (Okada et al., 1991;Przemeck et al., 1996) ( Fig. 1A,B). To assess whether MP function in shoot organization acts exclusively through the regulation of auxin transport, we generated mp pin1 double mutants. Analysis of progeny from a cross between heterozygous mp and pin1 plants, resulted in the identification of a fraction of mp-like plants which had formed a leafless dome from the 6 SAM ( Fig. 1C,D). The segregation ratio of this novel phenotype was not significantly different from an expected theoretical value based on chi-square analysis (p = 0.75; Table   S1), supporting the notion that the individuals were double mutants. The domes had a smooth surface and lacked differentiated epidermal, trichome and stomata cells (Fig. 1D).
After 2-3 weeks of culture in short day conditions, the majority of the putative doublemutants had developed additional leafless dome structures arising from the base of the initial dome (Fig. 1E). Such domes were never observed in single mp or pin1 mutant populations. The appearance of a novel phenotype in the absence of both gene activities leads us to conclude that MP and PIN1 act, at least in part, in separate pathways (see discussion).
Phenotypes of mp pin1 double mutant plants ranged from highly fasciated domes ( Fig. 1F) to single or multiple dome formation and in the vast majority of all plants, leaf formation was absent. After 3-4 weeks of culture, many of the domes had formed one or more filament like projections from its surface. A large number of these projections were formed after prolonged culture (Fig. 1G,H). We interpreted these as inflorescences as they sometimes produced pistil-like or petal-like structures at their apices (Fig. 1I, data not shown). We found further evidence that MP acts on another pathway distinct from the regulation of PIN1 by the evaluation of mp pid and pin1 pid double mutants. The PID gene is known to be required for subcellular localization of PIN1 in plant cells transporting auxin (Friml et al., 2004) and may thus be thought to act in the same pathway as PIN1. Consistent with this interpretation, the mp pid double mutants produced phenotypes that were indistinguishable from the mp pin1 phenotype ( Fig. 1J; Table S1).
Further, as previously reported (Furutani et al., 2004), the pin1 pid double mutants were characterized by a variable degree of wide or fused leaves, but did not produce the leafless dome phenotype observed in mp pin1 or mp pid double mutants. (Fig. 1K). The fact that the pin pid double mutant displays defects that are qualitatively similar to those of both single mutants is consistent with PIN1 and PID acting in the same pathway, in line with molecular evidence (Friml et al., 2004). In summary, mp pin1 and mp pid double mutants produced an identical, novel synergistic phenotype, suggesting that MP function in shoot meristem organization goes beyond the regulation of auxin transport processes (see discussion).

Reduction of auxin transport does not abolish later organ formation
The phenotypes from the mp pin1 and mp pid double mutants suggests that in mp mutant background, leaf initiation becomes extremely sensitive to reduction of auxin transport.
To assess this possibility, we grew mp seedlings on medium supplemented with the polar auxin efflux inhibitor naphthylphtalamic acid (NPA). The observed defects very much resembled the phenotype of mp pin1 and mp pid double mutants (Fig. 1M). In addition a large part of the heterogeneity observed in double mutants was lost at NPA concentrations at or above 10 μM NPA, suggesting that the heterogeneity was due to a comparatively weaker reduction in auxin transport in pin1 or pid mutants. Similar phenotypes were obtained with other, chemically distinct auxin efflux inhibitors, i.e. 9hydroxyfluorene-9-carboxylic acid (HFCA) and 2,3,5-triiodobenzoic acid (TIBA) (Fig.   1N,O) when applied to mp mutants.
Since auxin transport is reduced in mp mutants (Przemeck et al., 1996), we next asked whether the leafless dome phenotype could simply be a consequence of particularly weak auxin transport. To this end, we grew wild-type (WT) plants and mp mutants in the presence of increasing NPA concentrations. As shown in Figure 2A, leaf formation in wild type, but also in pin1 and pid shoots could not be abolished by any concentration of NPA, not even at 100 μM NPA an eventually lethal concentration.
Upon exposure to NPA WT, pin1 and pid3 mutants developed leaf fusions or tubular leaves but never formed leafless domes ( Fig. 1L; Fig. 2A). In WT plants, 0.1 μM and 1 μM NPA had no significant effect on the numbers of leaves produced by 21 days after germination (DAG; Fig. 2B). In contrast, in mp mutants NPA concentrations as low as 0.1 μM resulted in a dramatic decrease in leaf initiation (Fig. 2B) and at concentrations of 1 μM NPA and higher, the majority of mp mutants developed leafless domes. The novel leafless domes continued to grow, demonstrating that their inability to produce leaves was not the expression of a general growth defect. We conclude that MP, in addition to promoting auxin transport, must stimulate another activity that leads to the actual formation and growth of leaf primordia (see discussion).

Plants with ectopic expression of MP display similar NPA hypersensitivity
The above results suggest that a loss of MP function is required for the formation of the

The shoot apical meristem enlarges during leafless dome formation
The strict requirement of defects in MP activity for the formation of leafless domes lead us to have a more careful look at the mp meristem and its ability to form leaves in the absence of PAT inhibition. We found various defects in phyllotaxy and growth of mp  5C,F). In summary, the leafless domes appear to have the organization of an enlarged shoot apex, comprising an apical meristem, and a basal radially organized stem region, but the central zone as well as the entire meristem region are enlarged and the basal region shows limited internal and external cellular differentiation.

Leafless domes fail to focus PIN1 expression and auxin
Previous studies have reported that PIN1 expression is upregulated at sites of flower primordia formation in the reproductive SAM (Heisler et al., 2005). We used a Pro PIN1 :PIN1:GFP marker to visualize PIN1 expression in vegetative SAMs defective in mp and/or auxin transport functions. Our analysis showed that PIN1 expression was most pronounced in discrete epidermal spots on the surface of vegetative WT SAM's and internal procambial midveins of young primordia (Fig. 6A), in agreement with previous findings from the reproductive SAM. In mp meristems, PIN1 expression domains were more diffuse, occurred in defective phyllotactic patterns, and expression appeared spuriously in cells that are normally not involved in primordia formation (Fig. 6B). PIN1 expression in NPA-grown WT seedlings was very weak or absent in the central zone area of the meristem thereby forming a ring of high expression in the peripheral zone possibly predicting the future formation of a tubular leaf (Fig. 6C). Remarkably, in NPA-grown mp plants, PIN1 expression was not even restricted to the peripheral zone and instead expression was evenly distributed throughout the entire surface of young domes, including the central zone and more basal parts of the leafless dome (Fig. 6D). To assess if the lack of PIN1 focus formation in NPA-grown mp plants is accompanied by a lack of auxin maxima formation, we analyzed the expression of the auxin responsive Pro DR5 :GUS marker. In WT seedlings, Pro DR5 :GUS is expressed initially at the apices of emerging leaf primordia, and also internally in leaf primordia in conjunction with the formation of procambial tissues but Pro DR5 :GUS expression is not found in the central and peripheral zones of the SAM (Fig. 6E) (Mattsson et al., 2003). In mp seedlings, the Pro DR5 :GUS expression in leaf primordia apices was always more diffuse than in WT seedlings (Fig. 6F). WT plants responded to NPA with a considerable delay in leaf primordia formation and when leaf primordia emerged, the Pro DR5 :GUS expression was found at the margins of the circular or close to circular leaf primordia (Fig. 6G). At no point did we observe localized Pro DR5 :GUS expression at the flanks of NPA-grown mp meristems (Fig. 6H). In summary, the leafless dome meristems of NPA-grown mp mutants show defects in the focusing of PIN1 expression and do not form local auxinresponse maxima as judged by Pro DR5 :GUS.

Leaf founder cell markers are expressed in leafless dome meristems
The synergistic phenotype in mp pin1 double mutants suggests that MP acts not only through regulation of polar auxin transport in the process of leaf formation, but may separately promote the growth of leaf primordia. Potential target genes could be involved in leaf founder cell fate specification or associated with subsequent organ outgrowth. The AINTEGUMENTA (ANT) and ASYMMETRIC LEAVES 1 (AS1) genes are expressed in the leaf founder cell population and subsequently during outgrowth of leaf primordia (Elliott et al., 1996;Long and Barton, 1998;Byrne et al., 2000). We used the expression of these genes to assess if leaf founder cell populations are established at the flanks of the meristem in leafless domes. In WT plants, we found that the expression of these markers preceded the formation of leaf primordia and that they were expressed in outgrowing primordia (Fig 6 I,M), in agreement with published results. The expression of ANT and AS1 in mp mutants appears identical to WT expression patterns (Fig. 6L,P) except for the defects in phyllotaxy already described (Fig. 3B,C). In response to NPA, WT plants expressed ANT and AS1 in a circular domain (Fig. 6K,O) consistent with the subsequent formation of a tubular leaf. We observed a similar ring-shaped expression of ANT and AS1 near the apex of leafless domes in NPA-grown mp plants (Fig. 6L,P). Thus, leaf founder cell populations appear to be specified in the peripheral zones of WT and mp plants treated with NPA but this specification is not sufficient for leaf formation in the later. The failure to form leaves in leafless domes appears to be due to a defect in outgrowth of leaf primordia. In WT plants, early leaf initiation can be detected by a switch from anticlinal to periclinal cell divisions in the L2 layer (Medford et al., 1992).
We screened longitudinal medial sections of more than 15 leafless domes without finding any indications of periclinal divisions in the L2 layer. Instead we observed smooth surfaces of the peripheral zone, and a pattern of cell walls in the L2 layer that indicated strict anticlinal cell division planes (Fig. 6Q,R).
In summary, we conclude that the defect in leaf primordia formation in NPAgrown mp plants does not involve a block in the formation of leaf founder cells, but appears to involve a block of subsequent periclinal divisions in the process of leaf outgrowth, which appears to depend on MP activity. we observed that mp mutants of various allele strengths are hypersensitive to NPA treatment and display synergistic defects in double mutants with pin1. These findings provide strong evidence for an involvement of MP in a process beyond the control of auxin transport. Importantly, the synergistic defects cannot be mimicked by applying increased concentration of NPA to WT or pin1 plants, further supporting that MP regulates further, hitherto unexplored processes to promote leaf initiation. As one of those processes, we propose that MP has a role in promoting the actual outgrowth of leaves and flowers. Notably, it has also been suggested that activating ARFs, including MP, could bind to the promoters of auxin-regulated leaf specification genes thereby promoting leaf formation in the peripheral zone of the meristem while interaction with other ARF's limit this action in the central zone of the meristem (Leyser, 2006). Given this scenario, ARF's like MP would therefore be implicated in also having functions in conferring differential properties to zones in the SAM. MP is another likely component of the postulated mechanism since mp mutants also fail to form flowers from the inflorescence meristem and have reduced auxin transport capacity (Przemeck et al., 1996) . Further, MP encodes an auxin response factor (Ulmasov et al., 1997;Hardtke and Berleth, 1998), which might be involved in the auxindependent regulation of PIN expression (Sauer et al., 2006;Wenzel et al., 2007). No flowers can be induced by local auxin application on the flanks of mp inflorescence meristems (Reinhardt et al., 2003), suggesting that it is not only auxin transport and auxin accumulation that is defective in mp mutants, but also a failure to trigger lateral organ outgrowth even when auxin is locally provided (Reinhardt et al., 2003). Thus published auxin application experiments already hint to a role of MP in controlling auxin responses in lateral organ outgrowth.

DISCUSSION
The inhibition of auxin transport in mp mutant backgrounds generates an unprecedented type of abnormal SAM development, which not only completely obstructs the formation of lateral organs but also vastly expands the shoot apex. Marker gene expression indicates that the enlarged apical dome is composed of expanded STM and CLV3 expressing domains surrounded by a wide circular peripheral zone, marked by ANT and AS1. Although no leaf primordia are formed under these conditions, there seems to be some dispersed growth as the ANT and AS1 expression domains are extremely wide.
Under conditions of normal auxin transport, ARFs acting redundantly to mp appear to be sufficient for triggering organ formation from the vegetative, yet not from the reproductive SAM, as mp mutants produce leaves. Conversely, inhibition of auxin transport seems to allow for sufficient auxin focusing in the epidermis to trigger vegetative leaf initiation as long as MP is functional. However, poorly defined leaf initiation points seem to be insufficient to trigger organ outgrowth through redundantly acting ARFs when MP is not functional. While failed leaf initiation may thus be explainable as the superimposition of defects in two interdependent steps, the reasons for the enlargement of the central zone seem to reflect other, unknown levels of control. It has been proposed that the restriction of leaf-initiating auxin focusing to the peripheral zone reflects auxin sensitivity zones due to the specific expression domains of competing ARFs (Leyser, 2006). In this interpretation it is plausible that the removal of an important ARF may destabilize the zoning sufficiently to promote cell proliferation also in the central zone. In this context it is remarkable that we observed equally strong PIN1-GFP expression in the peripheral and central zones uniquely in NPA exposed mp mutants. Formally, it is also possible that the expansion of the central zone could be a necessary consequence of defective lateral organ formation. Several levels of mutually antagonistic gene activities have been implicated in the control of stem cell pool size of the shoot meristem (reviewed in (Clark, 2001;Williams and Fletcher, 2005;Carraro et al., 2006;Tucker and Laux, 2007) in which some negative regulators originate from the peripheral zone. As there are no other leafless genotypes available, we cannot genetically separate leaflessness from SAM expansion. However, it should be noted that in the inflorescences of pin1 mutants devoid of lateral flowers, the size of the meristem and its constituent zones have been described as normal (Vernoux et al., 2000), arguing against a mechanism where signals negatively regulating shoot meristem size are derived from concrete flower or leaf primordia.
The sizes of SAMs vary considerably across the plant kingdom (Steeves and Sussex, 1989) and the influences of new regulators on SAM size are continuously being revealed (Chaudhury et al., 1993;Clark et al., 1993Clark et al., , , 1995Running et al., 2004;Green et al., 2005;Chiu et al., 2007). The discovery of highly abnormally sized SAMs as a consequence of simultaneous interference with auxin transport and ARF function may provide an entry point in the genetic analysis of auxin's role in this process.

Plant material and growth
The mp G12, G33, Tu399 , pid3 and pin1-1 mutant alleles used for double mutant and single mutant analysis have been described previously in (Berleth, 1993;Hardtke andBerleth, 1998) (Christensen et al., 2000;Benjamins et al., 2001) (Okada et al., 1991. All MP alleles used in this study are characterized as strong alleles and no differences were observed between different alleles and subsequent treatments or double mutant generation. The 35S::MP line was generated as described in Hardtke et al (2004)  grown on ATS medium (Lincoln et al., 1990) and exposed to NPA as described (Mattsson et al., 1999). For quantification of CLV3 and STM expression domains in meristems, images were taken and subsequently analyzed using ImageJ v1.37 software (NIH). mp seedlings germinate approximately 1 day after WT most likely due to lack of hypocotyl and root. Comparable developmental stages were chosen for each data set defined by the WT, for example; The 3 DAG stage is defined as 3 DAG WT plants and 4 DAG mp plants, with similar sizes of leaf primordia.

In situ hybridization, Histology and GUS assays
All gene fragments were amplified from cDNA generated from total RNA extracted from 14 day old WT seedlings using Trizol reagent (Invitrogen) and subsequently reverse transcribed using RevertAid™ M-MuLV Reverse Transcriptase (Fermentas) and cloned into pBluescript II Sk(-) (Stratagene). The ANT and AS1 fragments were generated as described (Long and Barton, 1998);(Byrne et al., 2000). Whole mount in situ hybridization procedure was as described (Zachgo et al., 2000) with some modifications including overnight fixation and agitation in a fresh solution containing 0.1 M triethanolamine (pH 8) and 0.5% (v/v) acetic anhydride for 15 min, followed by two washes in 1x PBT solution prior to hybridization for two days at 60°C. For histological analysis, plant material was fixed and sectioned as described in (Ruzin, 1999).
Localization of β−glucuronidase activity was carried out as described in (Mattsson et al., 2003).

Microscopy
A Zeiss LSM 410 was used to image Pro PIN1 :PIN1:GFP and Pro CLV3 :GFP:ER using a 488 nm excitation filter and 500-530 nm emission filter combination. Background red autofluorescence was detected using a 568 nm excitation filter and an LP 580 emission filter set. DIC images were taken on a Nikon Eclipse 600 microscope using a Canon D30 digital camera and tissue clearing and preparation were performed as described in (Mattsson et al., 1999      Bars represent average of measured areas from 6-13 meristems, error bars are standard deviations. ** illustrates significant difference between NPA-grown mp mutants compared to all other genotypes and treatments, as determined by student's t test analysis, p < 05. Representative images of measured areas are shown in Fig. S2.

Figure 6. Marker analysis of leafless dome meristems
Material grown on medium supplemented with 10μM NPA indicated as "+ NPA".