Aintegumenta and Aintegumenta-Like6 regulate auxin-mediated flower development in Arabidopsis

Background Two related genes encoding AP2/ERF-type transcription factors, AINTEGUMENTA (ANT) and AINTEGUMENTA-LIKE6 (AIL6), are important regulators of floral growth and patterning in Arabidopsis. Evidence suggests that these genes promote several aspects of flower development in response to auxin. To investigate the interplay of ANT, AIL6 and auxin during floral development, I have examined the phenotypic consequences of disrupting polar auxin transport in ant, ail6 and ant ail6 mutants by either genetic or chemical means. Results Plants containing mutations in ANT or AIL6 alone or in both genes together exhibit increased sensitivity to disruptions in polar auxin transport. Both genes promote shoot growth, floral meristem initiation and floral meristem patterning in combination with auxin transport. However, differences in the responses of ant and ail6 single mutants to perturbations in auxin transport suggest that these two genes also have non-overlapping activities in each of these developmental processes. Conclusions The enhanced sensitivity of ant and ail6 mutants to alterations in polar auxin transport suggests that these mutants have defects in some aspect of auxin physiology. The inability of ant ail6 double mutants to initiate flowers in backgrounds disrupted for auxin transport confirm the proposed roles for these two genes in floral meristem initiation.


Background
The phytohormone auxin is a central regulator of many aspects of plant growth and development. Within the shoot apical meristem, auxin accumulates in a graded and dynamic manner with sites of auxin maxima correlating with the sites of lateral organ initiation [1][2][3]. The formation of auxin gradients within the meristem appears to result primarily from directional transport of the hormone and local biosynthesis (reviewed in [4]). Directional transport is mediated by PINFORMED (PIN) proteins, which act as auxin efflux carriers and exhibit polarized plasma membrane localization [5]. Mutations in PIN1 result in inflorescences that do not initiate floral meristems and instead grow as pin-like structures [6]. This phenotype can be rescued by application of auxin to the meristem, demonstrating the importance of auxin to floral meristem initiation and the pivotal role that PIN1 plays in the generation of auxin gradients within the meristem [7]. These conclusions are also supported by studies showing that wildtype plants grown in the presence of the auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) phenocopy pin1 mutants [6].
In addition to regulating floral meristem outgrowth, auxin also regulates floral organ development. Disruptions in polar auxin transport via transient NPA application result in flowers with reduced numbers of floral organs, narrower sepals and petals, malformed anthers and gynoecium with altered patterning [8]. These alterations in gynoecium patterning are very similar to those resulting from mutations in ETTIN (ETT), which encodes an auxin response factor (ARF) [8][9][10]. Mutations in PIN1 or PINOID (PID), which encodes a Ser-Thr kinase that controls PIN1 polarity, can produce several abnormal flowers prior to inflorescence termination in a pin-like structure [6,[11][12][13]. These flowers exhibit a range of defects that can include alterations in organ number (typically fewer sepals and stamens and more petals), fusion of floral organs, and valveless gynoecia [6,13]. Disruptions in flower development have also been observed in plants defective in auxin biosynthesis. Biosynthesis of the major auxin, indole-3-acetic acid (IAA), involves both tryptophan (Trp)-dependent and Trp-independent pathways (reviewed in [14]). Mutations in multiple members of the YUCCA (YUC) family of flavin monooxygenases, which act in the tryptamine (TAM) Trp-dependent pathway, display reductions in floral organ number, altered organ morphology, valveless gynoecia and sterility [15]. Similar phenotypes are observed in plants mutant for tryptophan aminotransferase TAA1 and the related TAR2 gene, which act in the indole-3-pyruvic acid (IPA) Trp-dependent pathway [16,17].
The molecular mechanisms by which auxin regulates floral meristem outgrowth and organogenesis within the flower are not well defined. Real time imaging of inflorescence meristems has demonstrated that PIN1 upregulation in floral anlagen is correlated with downregulation of the meristem regulator SHOOTMERISTEM-LESS (STM) as well as the boundary protein CUP-SHAPED COTYLEDON 2 (CUC2) [2]. Two genes that are likely to promote floral meristem initiation and specification downstream of auxin are the growth-promoting gene ANT and the floral meristem identity gene LEAFY (LFY). These two genes are early markers of a floral meristem fate and their expression is altered in pin1 mutants [18][19][20]. In addition, exposure of wild-type inflorescences to NPA results in a decrease in ANT mRNA accumulation in floral anlagen as early as 24 hours after treatment [21]. These data are consistent with a model in which auxin accumulation upregulates ANT and LFY expression to promote primordium outgrowth and floral identity, respectively [18]. However ant mutants do not display any defects in floral meristem initiation. Thus either ANT does not play a role in floral meristem initiation or it acts in parallel with other genes to promote primordium outgrowth.
Besides its proposed role in floral meristem initiation, ANT regulates floral meristem patterning, specification of floral organ identity, growth of floral organs and gynoecium patterning [19,[21][22][23][24][25]. In two of these processes (lateral organ growth and gynoecium patterning) ANT function been linked directly to auxin [25,26]. ANT promotes growth downstream of the auxin-inducible gene ARGOS (auxin-regulated gene involved in organ size), while in maturing organs ANT expression is repressed by ARF2, a repressor of organ growth [26,27]. In the gynoecium, ANT acts together with REVOLUTA and polar auxin transport to specify development of the carpel medial domain [25].
In several of its roles in flower development, ANT acts in a partially redundant manner with the related AINTEGUMENTA-LIKE6 (AIL6). While ail6 flowers have a wild-type appearance, ant ail6 flowers have more severe defects than ant. ant ail6 flowers lack petals, stamens and normal gynoecium, are dramatically reduced in size, and exhibit defects in floral organ positioning [21]. Altered expression of the auxin-responsive reporter AGH3-2:GUS in ant ail6 inflorescence meristems and flowers suggests that these floral defects may be a consequence of altered patterns of auxin accumulation and/ or responsiveness [21]. In addition to floral defects, ant ail6 plants exhibit decreased apical dominance, reduced stature, and altered vascular patterning, phenotypes similar to those found in plants disrupted in auxin physiology [21].
Like ANT, AIL6 expression is upregulated in incipient lateral organ primordia. This suggests that AIL6 might act redundantly with ANT in floral meristem initiation, but ant ail6 double mutants are still able to initiate floral meristems. ant ail6 inflorescence meristems do eventually stop initiating flowers, but this is due to a general growth arrest of the entire inflorescence apex [21] rather than specific termination of floral meristem initiation with continued growth of the inflorescence apex as is observed in pin1 mutants. Thus, it is still not known whether ANT and AIL6 regulate floral meristem initiation.
To further probe the roles of ANT and AIL6 in floral development processes potentially regulated by auxin, I have examined the consequences of losing ANT and AIL6 function singly or together in plants compromised for auxin transport via genetic or pharmacological means. These experiments demonstrate that both mutants are sensitized to defects in auxin transport, consistent with a role for these genes in regulation of some aspect of auxin physiology during flower development. The phenotypic consequences of loss of ANT function and loss of AIL6 function under the same conditions are somewhat different, suggesting that these two genes make distinct contributions to early stages of flower development. Floral meristem initiation is completely suppressed in plants compromised in polar auxin transport and containing mutations in both ANT and AIL6, demonstrating that these two genes do function in the initiation of floral meristems from the inflorescence meristem.

Findings
Mutations in ANT and AIL6 enhance the floral initiation and floral patterning defects of pid mutants Because pid mutants make several flowers prior to termination of the inflorescence meristem, they provide a sensitized background in which to study the role of ANT and AIL6 in auxin-mediated floral meristem initiation. I introduced ant-4, ail6-2 and ant-4 ail6-2 into the strong pid-1 allele and the intermediate pid-2 allele [13]. No flowers were produced in pid-1 ant-4 ail6-2 or pid-2 ant-4 ail6-2 triple mutants ( Figure 1A, B). While flowers were produced in pid-1 ant-4, pid-1 ail6-2, pid-2 ant-4 and pid-2 ail6-2 plants, they were reduced in number compared to pid-1 and pid-2 single mutants ( Figure 1A, B). These results indicate that ANT and AIL6 have partially overlapping roles in promoting floral meristem initiation.
The flowers produced by pid-1 ant-4 and pid-1 ail6-2 plants exhibited more severe defects than those observed in pid-1 (Figure 2A-F). pid-1 ant-4 flowers were extremely small with reduced numbers of petals and stamens ( Figure 2C, D). A similar phenotype was observed in pid-2 ant-4 flowers ( Figure 2G, H). The dramatic effect on petal number was surprising since ant-4 flowers exhibit only slight reductions in petal number in early-arising flowers while pid flowers have increased numbers of petals [13,21]. pid-1 ail6-2 flowers resembled pid-1 flowers except for an increased incidence of sepal fusion. Eighty percent of pid-1 ail6-2 flowers exhibited sepal fusion while 41.2% of pid-1 flowers exhibited sepal fusion ( Figure 2E, F). ail6-2 single mutants exhibited no sepal fusion or any other alterations in flower development compared with wild type [21].
Mutations in ANT and AIL6 enhance shoot defects in plants genetically disrupted for auxin transport The pid-1 ant-4 ail6-2 triple mutants also exhibited severe defects in leaf size and stem growth. Leaves of the triple mutant were thinner and shorter than in pid-1 ( Figure 1D). The naked pin-like inflorescences of these plants were extremely short in height and often not visible above the basal rosette of the plant ( Figure 1C, D). Leaf size and plant height were partially compromised in pid-1 ant-4 plants but not significantly altered in pid-1 ail6-2 plants ( Figure 1C, D). Similar results were obtained with the pid-2 allele.
ant mutants exhibit increased sensitivity to the effects of NPA on floral organ development I next examined the consequence of inhibiting polar auxin treatment via application of 10 μM NPA to ant-4 flowers. This concentration of NPA had some phenotypic consequences on wild-type Arabidopsis flower development, such as reductions in stamen number ( Figure 3G) but does not dramatically reduce floral meristem initiation, as is the case when wild-type inflorescences are treated with 100 μM NPA [8]. Thus it can be used to examine the relative sensitivity of ant-4 and Ler flowers to NPA treatment. Examination of mock and NPA-treated ant-4 plants revealed temporally distinct phenotypes. The first effects of NPA treatment were observed in ant-4 flowers opening approximately 10 days after treatment and continued through day 12 after treatment. A second phenotypic class was observed in ant-4 flowers opening 13-15 days post treatment.
In 10-12 day post NPA-treated ant-4 flowers, second whorl petals were often replaced with filaments or very thin petals ( Figure 3A-C). Organs were classified as filaments if they were radially symmetric and as thin petals if they had a flattened appearance. Alterations in petal development were not observed in 10-12 day post mock-treated ant-4 plants or NPA-treated Ler plants ( Figure 3A, C). To further characterize the second whorl organs that develop in these NPA-treated ant-4 flowers, SEM was used to examine their epidermal surface morphologies. Epidermal cells of the filaments lacked petal cell shape and cuticular thickenings suggesting that these organs did not possess any petal identity ( Figure 3D). Epidermal cells of thin petals did show cuticular thickenings characteristic of petals ( Figure 3E). However, these organs often lacked the conical cells normally present on the adaxial surface of petals, suggesting that these organs, like the filaments, had disruptions in organ polarity. No significant differences were observed in the numbers of sepals, stamens or carpels of 10-12 day post NPA-treated ant-4 or Ler plants ( Figure 3F). Occasionally, a valveless gynoecia phenotype was observed in the fourth whorl of 10-12 day post NPA-treated ant-4 flowers.
In 13-15 day post NPA-treated ant-4 plants, petals were often missing entirely (rather than being replaced with filaments) ( Figure 3G) and sepals were often fused to each other ( Figure 3H-J). The incidence of sepal fusion was much greater in NPA-treated ant-4 flowers (0.90 incidence/flower) compared with NPA-treated Ler flowers (0.20 incidence/flower) ( Figure 3H).

ail6 mutants exhibit increased sensitivity to the effects of NPA on floral meristem initiation
Treatment of Ler inflorescences with 10 μM NPA had no dramatic effects on floral meristem initiation ( Figure  4A, B). However, treatment of ail6-2 inflorescences with 10 μM NPA resulted in a temporary suspension of floral meristem initiation ( Figure 4C, D). Since primordia initiation was restarted at a later date, there was a gap in the developmental series of floral buds represented on an inflorescence ( Figure 4C, D). SEM analyses suggested that this time period corresponds to the initiation of at least six flowers in wild type ( Figure 4B, D). Floral organ number was similarly affected in NPA-treated Ler and ail6-2 flowers ( Figure 4E). While sepal number was similar in NPA-treated Ler and ail6-2 flowers, the incidence of sepal fusion was increased in NPA-treated ail6-2 flowers (0.46 incidence/flower) compared to NPA-treated Ler flowers (0.16 incidence/flower) ( Figure  4F, G). No obvious effect on floral meristem initiation or floral organ development was observed in ant-4 ail6-2 plants treated with NPA as compared with mocktreated plants. Any effect on floral meristem initiation would be difficult to observe since termination of the inflorescence meristem occurs in untreated ant-4 ail6-2 inflorescences prior to the expected effect of NPA treatment.

ANT and AIL6 promote floral meristem initiation
Although it has been proposed that ANT promotes primordium outgrowth in response to auxin accumulation, neither ant nor ant ail6 double mutants display defects in floral meristem initiation from the inflorescence meristem. The results presented here provide the first evidence that ANT and AIL6 are important factors regulating floral meristem initiation. Combining either ant or ail6 with pid reduces the number of flowers initiated by the inflorescence meristem, while the loss of both ANT and AIL6 function together prevents the formation of any flowers in pid mutants. These results suggest that ANT and AIL6 act in a redundant fashion to promote floral meristem initiation. Other pieces of evidence however suggest that the activities of these two genes are not completely overlapping and that AIL6 may play a more important role in this process as compared with ANT. Floral meristem initiation was temporarily terminated in ail6-2 inflorescences treated with 10 μM NPA but not in Ler or ant-4 inflorescences similarly treated. Thus floral meristem initiation is inhibited by lower levels of NPA in ail6-2 inflorescences. Furthermore, loss of AIL6 function in the pin1-3 background results in a nearly complete loss of lateral branches from inflorescences. ANT and AIL6 act with polar auxin transport in floral meristem patterning One defect in flower development observed in mutants with disruptions in polar auxin transport such as pid [13] or in wild-type plants treated with NPA was fusion of floral organs. This may be a consequence of alterations in the positioning or outgrowth of organ primordia from the floral meristem and/or the inability to establish boundaries between adjacent organs. It is known that polar auxin transport is critical to cotyledon boundary formation during embryogenesis [28]. In postembryonic development, separation between lateral organs and the meristems from which they arise is also likely to involve polar auxin transport (reviewed in [4,29]). Live imaging of inflorescence meristem shows that upregulation of PIN1 expression in lateral organ founder cells was correlated with reduced expression of meristem and boundary markers such as STM and CUC2 [2]. Correspondingly, boundary genes such as JAGGED LATERAL ORGAN (JLO) appear to repress PIN1 and promote STM expression in boundary cells [30]. These antagonistic relationships result in the separation of indeterminate cells within the meristem and determinate cells within the lateral organ primordia [29]. The results presented here showing floral organ fusion in some NPA treated wildtype flowers suggest that a common mechanisms is involved in organ separation in shoot apical meristems and floral meristems. Sepal fusion was enhanced in pid-1 ail6-2 flowers compared to pid-1 flowers and in NPA treated ant-4 and ail6-2 flowers as compared with NPA treated Ler flowers. Thus, ANT and AIL6 appear to function together with polar auxin transport in separation of sepals, while ANT also acts in stamen boundary specification. These data provide a link between auxin and the role of ANT and AIL6 in floral meristem patterning.
ANT promotes petal development in combination with auxin transport ant-4 mutants in which polar auxin transport was reduced also exhibited severe effects on petal development. Few petals were initiated in pid ant-4 double mutants or in the later-arising NPA-treated ant-4 flowers. The distinct temporal effects observed after NPA treatment on ant-4 flowers were likely a consequence of the floral development stage at which auxin transport was inhibited. Flowers maturing 10-12 days after NPA treatment presumably correspond to those in which sepal and petal primordia were already initiated while flowers maturing 13-15 days after NPA treatment most likely correspond to younger floral meristems in which sepal and petal primordia had not yet initiated. In Ler plants, the length of time between floral meristem initiation and flower opening is approximately 13.25 days [31]. In the earliest flowers affected by NPA treatment, second whorl organs were present but often developed as filaments or very thin petals. The lack of petal blade outgrowth and the absence of conical cells in thin petals indicate that petal polarity is disrupted. A role for auxin in mediating adaxial/abaxial patterning during organogenesis has been suggested by the phenotype of plants lacking two related ARFs, ETT and ARF4 [32]. ett arf4 double mutants have defects in lateral organ polarity that mimic loss of function mutations in KANADI genes that specify abaxial identity [32]. In later-arising NPA-treated ant-4 flowers, second whorl organs were usually missing suggesting that polar auxin transport is required for petal initiation. It is not clear why petal initiation is particularly sensitive to NPA treatment in the ant-4 background.

Conclusions
Here I show that combining mutations in ANT and AIL6 with disruptions in polar auxin transport results in severe effects on shoot and flower development. ANT and AIL6 appear to have both overlapping and distinct roles in the process of floral meristem initiation downstream of auxin, while ANT appears to play a more important role than AIL6 in later stages of auxinmediated floral meristem patterning and floral organogenesis. The results are consistent with a model in which ant and ail6 mutants are disrupted in some aspect of auxin physiology. Furthermore they support the idea that all of the different functions of these two genes are linked with auxin dynamics within shoot tissues.

SEM
Tissue for SEM was fixed, dried, dissected and coated as described previously [23]. SEM analyses were performed on a FEI Quanta 200 ESEM.