Genetic Interactions Between Leaf Polarity ... - Oxford Academic

Plant Cell Physiol. 48(5): 724–735 (2007) doi:10.1093/pcp/pcm040, available online at www.pcp.oxfordjournals.org ß The Author 2007. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: [email protected]

Genetic Interactions Between Leaf Polarity-Controlling Genes and ASYMMETRIC LEAVES1 and 2 in Arabidopsis Leaf Patterning Yanlei Fu, Lin Xu, Ben Xu, Li Yang, Qihua Ling, Hua Wang and Hai Huang * National Laboratory of Plant Molecular Genetics, Shanghai Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 300 Fenglin Road, Shanghai 200032, PR China

Introduction

During leaf development, establishment of adaxial– abaxial polarity is essential for normal leaf morphogenesis. This process is known to be strictly regulated by several putative transcription factors, microRNA165/166 (miR165/ 166), a trans-acting short-interfering RNA (tasiR-ARF), as well as proteins involved in RNA silencing. Among the putative transcription factor genes, ASYMMETRIC LEAVES1 and 2 (AS1 and 2) facilitate the specification of leaf adaxial identity; however, the mechanism by which AS1 and AS2 cooperate with other leaf polarity components remains largely undetermined. In the current study, we characterized the phenotype of mutants by combining as1 and as2 with mutations of several key transcription factors. Our data showed that double mutant plants carrying as1/as2 and rev, phb or phv enhanced as1/as2 defects by producing more severely abaxialized leaves. In contrast, triple mutants, obtained by combining as1/as2 with double mutant filamentous flower yabby3 (fil yab3) or kanadi1 kanadi2 (kan1 kan2), exhibited additive phenotypes. Additionally, while leaves of rev as2 contained high levels of FIL transcripts, only slightly elevated miR165/166 levels were noted, indicating that FIL and miR165/166 act in parallel in leaf patterning. Moreover, 35S::MIR165a/rev as2 transgenic plants resulted in a more severe abaxialized leaf phenotype than the rev and as2 single mutant plants transformed with the same 35S::MIR165a fusion. Genetic interactions between the key regulators during leaf patterning are discussed.

Leaves of higher plants derive from the peripheral zone of the shoot apical meristem (SAM). Once leaf primordia are initiated, polarity is conveyed along three axes: the proximal–distal, the medial–lateral and the adaxial–abaxial (ad/abaxial) axes (Waites et al. 1998, Hudson 2000, Byrne et al. 2001). Of these, the establishment of the ad/abaxial axis, required for subsequent lamina growth and asymmetric development, might be of primary importance (for a review, see Bowman et al. 2002). Several genes that play critical roles in leaf ad/abaxial polarity formation have been identified and extensively characterized in Arabidopsis. Studies have shown that class III HD-ZIP genes REVOLUTA (REV), PHABULOSA (PHB) and PHAVOLUTA (PHV) are expressed in the adaxial leaf domain and are required for specifying adaxial identity (McConnell and Barton 1998, McConnell et al. 2001, Emery et al. 2003, Zhong and Ye 2004). Several other genes, including FILAMENTOUS FLOWER (FIL), YABBY3 (YAB3), KANADI1 and 2 (KAN1 and KAN2), and AUXIN RESPONSE FACTOR4 (ARF4), are expressed in the abaxial leaf domain and promote abaxial identity (Chen et al. 1999, Sawa et al. 1999, Siegfried et al. 1999, Eshed et al. 2001, Kerstetter et al. 2001, Pekker et al. 2005, Fahlgren et al. 2006, Garcia et al. 2006). Recent studies also suggested that AS1 and AS2 genes mediate the leaf adaxial identity (Xu et al. 2003); AS1 encodes a protein that contains an R2-R3 MYB-domain (Byrne et al. 2000) while AS2 encodes a plant-specific leucine-zipper protein (Iwakawa et al. 2002) that associates with AS1 (Xu et al. 2003). Given that AS1 and AS2 might regulate the same genetic pathways during leaf morphogenesis (Xu et al. 2003), mutations in either the AS1 or AS2 gene could result in overall similar plant phenotypes (Ori et al. 2000, Sun et al. 2000, Semiarti et al. 2001). More recent data revealed that genes functioning in the RNA-induced gene silencing and the protein degradation pathways were also required for leaf adaxial identity. These genes, including RNA DEPENDENT RNA POLYMERASE6 (RDR6), SUPPRESSOR OF GENE

Keywords: Adaxial–abaxial polarity — Arabidopsis — Asymmetric leaves1(2) — Leaf development — MicroRNA165/166. Abbreviations: AGO7/ZIP, ARGONAUTE7/ZIPPY; ARF4, AUXIN RESPONSE FACTOR4; AS1/2, ASYMMETRIC LEAVES1 and 2; DCL4, DICER-LIKE4; FIL, FILAMENTOUS FLOWER; KAN, KANADI; Ler, Landsberg erecta; MIR165/166, MICRORNA165/166; PHB, PHABULOSA; PHV, PHAVOLUTA; RDR6, RNA DEPENDENT RNA POLYMERASE6; REV, REVOLUTA; RT–PCR, reverse transcription–PCR; SAM, shoot apical meristem; SEM, scanning electron microscopy; SGS3, SUPPRESSOR OF GENE SILENCING3; YAB, YABBY.

*Corresponding author: Email, [email protected]; Fax, þ86-21-54924015. 724

Genetic interactions in leaf polarity formation

SILENCING3 (SGS3), ARGONAUTE7/ZIPPY (AGO7/ ZIP), DICER-LIKE4 (DCL4), ARGONAUTE1 (AGO1), and those that encode 26S proteasome subunits, act synergistically with AS1 and AS2 for leaf ad/abaxial polarity formation (Lynn et al. 1999, Kidner and Martienssen 2004, Li et al. 2005, Garcia et al. 2006, Xu et al. 2006, Yang et al. 2006, Huang et al. 2006). Research has also demonstrated that RDR6/SGS3/ZIP and AS1/AS2 down-regulated miRNA165/166 and FIL in leaf development (Li et al. 2005, Garcia et al. 2006, Xu et al. 2006). To understand better the regulatory mechanisms underlying the establishment of leaf polarity, elucidation of genetic interactions among the regulatory components seems important. However, the genetic interactions between AS1–AS2 and other leaf polarity genes have not, to date, been fully elucidated. In the present study, we demonstrate that loss-of-function as mutants contained altered levels of REV, PHB, PHV and FIL mRNAs. In addition, we report the phenotypic analyses of mutants by combining as1 and as2 with other leaf polarity-defective mutants. Based on the phenotypic analyses of 35S::MIR165a transgenic plants, we discuss possible genetic models by which the leaf ad/abaxial polarity is normally established.

725

A 4 3.21 3 2.30 2

1

1

1 0.57 0.28 P < 0.01 P < 0.01 P < 0.01

0

r

Le

1 as

1 -10

01 S1 2-1 S::A s a 35

P < 0.05

0 ol-

C

-2d

iso

B 4 2.91

3 1.95 2

1

1

0.69

1

0.78

Results AS1 and AS2 positively regulate REV, PHB and PHV Previous studies showed that mRNA levels of the class III HD-ZIP gene PHB were elevated in 35S::AS2 transgenic plants (Lin et al. 2003), suggesting that AS2 might positively regulate PHB. To elucidate further the mechanism by which AS1 and AS2 genes may affect the accumulation of class III HD-ZIP gene transcripts, we utilized real-time reverse transcription–PCR (RT–PCR) to analyze REV, PHB and PHV expression in as1 and as2 mutants. Our data demonstrated that transcript levels of these three genes were all reduced in as1-101 and as2-101 rosette leaves as compared with those in wild-type plants (Fig. 1A–C), but were much lower than those in both 35S::AS1 transgenic plants and a gain-of-function as2 mutant, iso-2d (Nakazawa et al. 2003). These results indicated that normal AS1 and AS2 functions are required for the proper REV, PHB and PHV expression in leaves. Characterization of rev-6 as2-101, phb-6 as2-101 and phv-5 as2-101 phenotypes To investigate further the regulatory network in leaf ad/abaxial polarity formation, we constructed double mutants by combining as1-101 or as2-101 with rev-6, phb-6 and phv-5, and analyzed their phenotypes (see Supplementary Table S1). Compared with wild-type leaves (Fig. 2A), as2-101 (Fig. 2B) produced some rosette

0

P < 0.05 P < 0.05 P < 0.05

r 01 AS1 01 Le :: 2-1 1-1 as 35S as

P < 0.01

0 ol-

C

-2d

iso

C 4 3.33 3

2

1

0

1.51 1

0.65

1

0.73

P < 0.05 P < 0.05 P < 0.05

r 01 01 AS1 Le :: 2-1 1-1 as 35S as

P < 0.01

C

0 ol-

-2d

iso

Fig. 1 Altered REV, PHB and PHV expression. REV (A), PHB (B) and PHV (C) transcript levels from the first two pairs of rosette blades were analyzed using real-time RT–PCR. Leaves were collected from the 28-day-old plants of wild-type Ler, as1-101 and as2-101 single mutants, a 35S::AS1 transgenic line, and a gainof-function AS2 mutant iso-2d. Two other independent 35S::AS1 transgenic lines have also been analyzed, and results from these two lines were consistent with those of the line shown in this figure (data not shown). Bars indicate the SD.

726


A

Table 1 Frequencies of lotus- and needle-like leaves in mutant plantsa

B

C

D

Mutants

No. of plantsb

Lotus leaves (%)

Needle-like leaves (%)

as2-101 phv-5 as2-101 phb-6 as2-101 rev-6 as2-101

260 224 132 328

12.1 1.6 16.5 1.9 20.1 1.6 23.5 1.7

0.2 0.2 0.7 0.5 4.5 1.4 14.6 1.0

a

Plants were grown at 228C. Only the first two pairs of the rosette leaves of the 28-day-old plants were scored. Values are mean SD. b Total number of plants that were analyzed.

E

F

G

H

I

J

K

L

M

Fig. 2 Phenotypes of the rev-6 as2-101 double mutant. (A–D) Seedlings of wild-type Ler (A), as2-101 with lotus leaves (arrow) (B), rev-6 (C) and rev-6 as2-101 (D). (E–G) The altered rev-6 as2-101 phenotypes include a high frequency of needle-like leaves (D, arrowhead; E) and lotus leaves (D, arrows; F), and the rippled adaxial surface of the expanded leaves (G). (H, I) The wild-type adaxial (H) and abaxial (I) epidermal cells show distinct cellular patterns. (J) The adaxial surface of the expanded rev-6 as2-101 leaves has some long and narrow cells (arrowhead) similar to those on the abaxial surface of wild-type leaves. (K–M) The rev-6 as2-101 mutant plants contained petal adaxial epidermal cells (M) different from those on the adaxial (K) but similar to those on the abaxial surfaces of the wild-type petal (L). Bars ¼ 1 cm in (A–D), 200 mm in (E–G), 50 mm in (H–J) and 5 mm in (K–M).

leaves with a lotus leaf structure (arrow), with petioles noted underneath leaf blades, as well as a needle-like structure, both of which reflected the loss of leaf adaxial identity (Xu et al. 2003, Qi et al. 2004). The frequency of lotus leaf

and needle-like leaves in as2-101 was relatively low (Table 1), and rev-6 leaves (Fig. 2C) did not show such abnormalities. Interestingly, rev-6 as2-101 leaf surfaces became very scabrous (Fig. 2D, G) and the frequency of needle (Fig. 2D, E) and lotus structures (Fig. 2D, F) increased markedly (Table 1). Similar abnormal leaf phenotypes were also observed in the rev-6 as1-101 double mutant plants (see Supplementary Fig. S1B). While needle-like structures typically appeared in the first pair of rosette leaves in the as1-101 and as2-101 single mutant (Xu et al. 2002, Xu et al. 2003, Qi et al. 2004), they were also observed in the second pair of rosette leaves in rev-6 as2-101 (Fig. 2D) and rev-6 as1-101 (data not shown) plants. Scanning electron microscopy (SEM) analysis of leaf and petal epidermal cells was utilized to investigate the rev-6 as2-101 defects further. The adaxial epidermis of wild-type Ler leaves was characterized by an undulating surface composed of uniformly sized cells (Fig. 2H), while smaller and less uniformly sized cells mixed with some long and narrow cells were noted in the abaxial epidermis (Fig. 2I, arrowhead) (McConnell and Barton 1998). Although abaxial epidermal cells appeared normal (data not shown), the adaxial surface of rev-6 as2-101 leaves was mosaic, with both adaxial and abaxial epidermal patches (Fig. 2J, arrowhead). This aberrant phenotype appeared on leaves of the Antirrhinum mutant phantastica (Waites and Hudson 1995) and the Arabidopsis double mutant rdr6-3 as2-101 (Li et al. 2005), but was not observed on the as2 adaxial leaf surface. Petals of rev-6 as2-101 plants also displayed severe defects in ad/abaxial identity. Wild-type adaxial petal epidermis showed cone-shaped cells with straight lines (Fig. 2K), while the abaxial petal epidermis showed cobblestone-shaped cells with wavy lines (Fig. 2L) (McConnell and Barton 1998). Although the abaxial petal surface of rev-6 as2-101 was normal (data not shown), wavy lines were noted on all adaxial epidermal cells (Fig. 2M), similar to the petal abaxial epidermal cells of wild-type plants (Fig. 2L). These results indicate that petals in rev-6


as2-101 mutant plants are also abaxialized. We also constructed double mutants by combining as2-101 with another rev allele, rev-9, and found that the leaf phenotypes of rev-9 as2-101 and rev-6 as2-101 were very similar (see Supplementary Fig. S2). We next analyzed phb-6 as2-101 and phv-5 as2-101 double mutant leaves, and found that phb-6 as2-101 and phv-5 as2-101 also enhanced as2-101 leaf phenotypes, while they were less severe compared with those in rev-6 as2-101. Surfaces of expanded leaves of phb-6 as2-101 and phv-5 as2-101 were similar to those in wild-type plants (data not shown). However, the frequency of lotus- and needle-like leaves in double mutants was enhanced compared with those in the as2-101 single mutant (Table 1). The as1 double mutant exhibited overall similar phenotypes to the as2 double mutant, indicating that REV, PHB and PHV functions are required for a pathway which is controlled by both AS1 and AS2 genes, and REV might play a more critical role among the three in the pathway. Similarity of rev-6 as2-101 and rdr6-3 as2-101 double mutant leaf phenotypes In addition to the rippled leaf surface, rev-6 as2-101 leaves often produced outgrowths on the adaxial leaf side (Fig. 3A), some of which grew to form leaflets (see the boxed region in Fig. 3A, and Fig. 3B). Additionally, small leaflets generated additional outgrowths on their adaxial side (Fig. 3B, arrowhead). Although organ-like structures were not observed on cauline leaves, sepals and petals, they were rippled on their adaxial surfaces (Fig. 3C–E). Leaf phenotypes of rev-6 as2-101 resembled those of the rdr6-3 as2-101 double mutant (Li et al. 2005), which produced a high frequency of needle-like (Fig. 3F) and lotus-like leaves (Fig. 3G), and had highly rippled leaf surfaces (Fig. 3H). Moreover, rdr6-3 as2-101 leaves also produced outgrowths on the adaxial leaf side, some of which developed into leaflets (Li et al. 2005). We reported previously that levels of mature miR165/ 166 and FIL mRNA were elevated in rdr6-3 as2-101 double mutant leaves (Li et al. 2005). We further analyzed miR165/ 166 and FIL levels in order to determine whether the similar leaf phenotypes observed in rev-6 as2-101 resulted from molecular mechanisms similar to those in rdr6-3 as2-101. Surprisingly, while miR165/166 levels were much lower than those found in rdr6-3 as2-101 leaves (Fig. 3I), FIL levels were dramatically elevated in rev-6 as2-101 leaves, resembling those in rdr6-3 as2-101 leaves (Fig. 3J). In situ hybridization was utilized to investigate whether the altered transcript levels of FIL were also accompanied by changes in its expression pattern. In wild-type Ler (Fig. 3N), rev-6 (Fig. 3O) and as2-101 (Fig. 3P) plants, FIL mRNAs accumulated in the abaxial side of the leaf primordium,

727

whereas they were either extended towards the adaxial side of the expanded leaves (Fig. 3Q, arrowheads) or were present throughout the entire needle-like leaf primordium of rev-6 as2-101 (Fig. 3R). To examine further the role of FIL in the abnormal rev-6 as2-101 leaf morphology, we constructed fil-1 rev-6 as2-101 and analyzed the triple mutant leaves. Compared with the rev-6 as2-101 double mutant, fil-1 rev-6 as2-101 displayed an additional inflorescence phenotype (see Supplementary Fig. S3), similar to that of the fil rev double mutant (Chen et al. 1999). In addition, most fil-1 rev-6 as2-101 leaves did not show visible outgrowths (Fig. 3K, M), though the adaxial leaf surface was slightly pleated. While a small proportion of the fil-1 rev-6 as2-101 leaves still had ectopic outgrowths (Fig. 3L, M), the severity of this phenotype was greatly suppressed (Fig. 3L, for comparison see Fig. 3A). These results suggest that the ectopic expression and overexpression of the FIL gene may predominantly contribute to the abnormal rev-6 as2-101 leaf phenotypes. MIR165a overexpression in rev as2, phb as2 and phv as2 double mutants It is generally thought that miR165 and miR166 are important to leaf ad/abaxial polarity formation by repressing REV, PHB and PHV (Emery et al. 2003, Mallory et al. 2004, Zhong and Ye 2004). However, our previous results showed that overexpression of MIR165a in wild-type plants did not result in leaf defects in the ad/abaxial polarity (Li et al. 2005). To understand better miR165/166 action in leaf polarity formation, we introduced a 35S::MIR165a fusion into rev-6 as2-101, phb-6 as2-101 and phv-5 as2-101 double mutants, and subsequently analyzed the phenotypes of the transgenic plants. Wild-type Ler and each corresponding single mutant carrying the 35S::MIR165a fusion were used as controls. Our data demonstrated that 35S::MIR165a/Ler plants exhibited slightly down-curled leaves (Fig. 4A), while 35S::MIR165a/as2-101 plants had an increased frequency of lotus- and needle-like leaves (data not shown) with rippled leaf surfaces (Fig. 4B). Although 35S::MIR165a/rev-6 leaves (Fig. 4C) had leaf phenotypes similar to those of 35S::MIR165a/Ler plants, the morphology of 35S::MIR165a/rev-6 as2-101 transgenic plants was severely affected. A total of 61 independent transgenic lines were grouped into two categories. One category consisted of 17 lines, which showed milder plant phenotypes, with early appearing unexpanded leaves (Fig. 4D, arrowheads) and late appearing very rippled leaves (Fig. 4D, arrow). The second category consisted of 44 lines, which had normal-looking cotyledons, but with all other leaves needle-like in appearance (Fig. 4E). These plants were arrested at the seedling stages, and the needle-like leaves were covered


H

E

I rdr-6-3 as2-101

G

D

Wild-type

F

C

rev-6 as2-101

B

rev-6

A

as2-101

728

5s rRNA miR165

21nt

J

K

L

M

8.0 8.1

8 6 4 2

1.9 1

1.5

0

O

80 60 40 21.7 20 n = 100 n = 60

0

pe -101 v-6 -101 101 -ty re s2 s2ild as2 a a W -6 -3 rev rdr6

N

95.0

100 Frequency of leaves with outgrowths (%)

10

1

0 2-1

-6 rev

P

Q

as

01

2-1

as

e

1r

fil-

v-6

R

Fig. 3 Phenotypic comparison between rev-6 as2-101 and rdr6-3 as2-101. (A) The adaxial surface of some rev-6 as2-101 leaves produced ectopic outgrowths, some of which grew further to form leaflets with the rippled adaxial surface. (B) A rev-6 as2-101 leaflet that is the close-up of the boxed region in (A), which had additional outgrowths (arrowhead). (C–E) The adaxial surfaces of cauline leaves (C), sepals (D) and petals (E) in rev-6 as2-101 were also rippled. (F–H) The leaf phenotypes of rev-6 as2-101 are similar to those in rdr6-3 as2-101, which also produces needle-like (F) and lotus leaves (G), and the rippled adaxial surface of the expanded leaves (H). (I, J) The miR165/166 levels were distinct between rev-6 as2-101 and rdr6-3 as2-101 leaves (I), while the elevated FIL mRNA levels were consistent in both rev-6 as2-101 and rdr6-3 as2-101 leaves ( J). (K–M) The rev-6 as2-101 leaf phenotypes were suppressed in the fil-1 background. Most fil-1 rev-6 as2-101 leaves had no visible ectopic outgrowths (K, M). Although there were some outgrowth-containing leaves in fil-1 rev-6 as2-101 (L), the phenotypes appeared much weaker as compared with those in rev-6 as2-101. n, numbers of leaves analyzed. The first two pairs of rosette leaves from 28-day-old wild-type and mutant plants were used for RT–PCR and miR165/166 level analyses. (N–R) In situ hybridization using an antisense FIL probe on transverse sections of primordia in wild-type Ler (N), rev-6 (O), as2-101 (P), rev-6 as2-101 flat (Q) and needle-like (R) leaf primordia. Note that some flat primordia showed expanded FIL expression towards the adaxial side (Q) (arrowheads). Bars ¼ 1 mm in (A), 3 mm in (C), 500 mm in (B, D), 200 mm in (E, F–H, K, L), and 10 mm in (N–R).


A

D

B

C

729

C

E

F

C C

Fig. 4 Phenotypes of plants carrying a 35S::MIR165a construct. (A) A 35S::MIR165a/ Ler plant. (B) A 35S::MIR165a/as2-101 plant, showing slightly rippled leaves. (C) A 35S::MIR165a/rev-6 plant. (D) A 35S::MIR165a/rev-6 as2-101 plant, showing that rosette leaves were either very narrow and rippled (arrow) or unexpanded (arrowhead). (E) A more severe 35S::MIR165a/rev-6 as2-101 plant, showing all needle-like rosette leaves. (F) A needle-like leaf of the severe 35S::MIR165a/rev-6 as2-101 plant, showing an insufficiently differentiated epidermal pattern. F is a close-up of the boxed region in (E). c, cotyledon. Bars ¼ 2 mm in (A–E) and 50 mm in (F).

with long and narrow epidermal cells (Fig. 4F), resembling those on leaf petioles (Byrne et al. 2000, Sun et al. 2002). We also investigated 35S::MIR165a/phb-6 as2-101 and 35S::MIR165a/phv-5 as2-101 transgenic plants; however, the phenotypes of these plants were indistinguishable from those of the 35S::MIR165a/as2-101 transgenic plants. These results further supported the idea that among the three functionally redundant class III HD-ZIP members, REV was of premier importance for adaxial leaf identity. Phenotypes of plants carrying multiple mutations To elucidate further possible genetic interactions between AS1/2 and YAB or KAN genes, we constructed fil-1 yab3-2 as1-101, fil-1 yab3-2 as2-101, kan1-2 kan2-1 as1101 and kan1-2 kan2-1 as2-101 triple mutants. Compared with fil-1 yab3-2 (Fig. 5A) (Kumaran et al. 2002), the fil-1 yab3-2 as2-101 plants produced lotus leaves (Fig. 5B, arrowhead), resembling as2-101 single mutant plants. Other leaves of triple mutants (Fig. 5E) appeared to have

both fil-1 yab3-2 and as2-101 characteristics, with the leaf blade width between those of as2-101 (Fig. 5C) and fil-1 yab3-2 (Fig. 5D) plants (see Supplementary Table S2). fil-1 yab3-2 as1-101 showed similar plant phenotypes to those of fil-1 yab3-2 as2-101 (see Supplementary Fig. S1D, G). Additionally, kan1-2 kan2-1 double mutant plants showed narrow leaves with ectopic outgrowths emerging on the abaxial side (Fig. 5F, arrowhead, G) (Eshed et al. 2001, Eshed et al. 2004). The early appearing rosette leaves of kan1-2 kan2-1 as2-101 triple mutant plants also contained lotus leaves (Fig. 5H, arrow, I). However, the outer side of the lotus leaf, which is equivalent to the abaxial side of a wild-type leaf, produced ectopic outgrowths (Fig. 5H, arrowhead, I). The similar kan1-2 kan2-1 as2-101 phenotypes were also observed in the kan1-2 kan2-1 as1-101 triple mutant (see Supplementary Fig. S1I). The additive fil-1 yab3-2 as1-101, fil-1 yab3-2 as2-101, kan1-2 kan2-1 as2-101 and kan1-2 kan2-1 as2-101 triple mutant phenotypes suggested that the AS1/AS2 genes might not interact genetically with YAB and KAN genes.

730


A

F

J

B

C

G

D

H

E

I

K

L

M

ad

ab

Fig. 5 Phenotypes of fil-1 yab3-2 as2-101, kan1-2 kan2-1 as2-101 and kan1-2 kan2-1 rev-6 as2-101. (A, B) Seedlings of the double mutant fil-1 yab3-2 (A) and triple mutant fil-1 yab3-2 as2-101 (B). An arrowhead indicates a lotus leaf which was not observed in fil-1 yab3-2. (C–E) One of the first pair of rosette leaves from as2-101 (C), fil-1 yab3-2 (D) and fil-1 yab3-2 as2-101 (E). Note that the leaf of the fil-1 yab3-2 as2-101 triple mutant is narrower than that of the as2-101 single mutant but broader than that of the fil-1 yab3-2 double mutant. (F, G) A seedling (F) and a rosette leaf (G) of the kan1-2 kan2-1 double mutant, showing that outgrowths (arrowhead) had emerged on the abaxial leaf side. (H, I) A seedling (H) and a rosette leaf (I) of the kan1-2 kan2-1 as2-101 triple mutant, with outgrowths (arrowhead) appearing on the outside surface of a lotus leaf (arrow). (J–M) A seedling (J) and rosette leaves (K–M) of the kan1-2 kan2-1 rev-6 as2-101 quadruple mutant, with an adaxial view (K, L) and abaxial view (K, M) of rosette leaves, showing that ectopic outgrowths appeared on both adaxial and abaxial sides. ad, adaxial leaf side; ab, abaxial leaf side. Bars ¼ 0.5 cm in (A, B, F, H, J), 1 mm in (C–E, G, I) and 500 mm in (K–M).

The severe abaxialized leaves in rev-6 as2-101 or adaxialized leaves in kan1-2 kan2-1 resulted in outgrowths on the adaxial or abaxial leaf surfaces, respectively. To test whether ectopic outgrowth production was the result of genetic interaction between AS2–REV and KAN1–KAN2 controlled pathways, we constructed kan1-2 kan2-1 rev-6

as2-101 quadruple mutants. The plant stature of the quadruple mutant was very small (Fig. 5J), and ectopic outgrowths were found on both adaxial (Fig. 5K, L) and abaxial (Fig. 5K, M) sides of the mutant leaves. These results suggest that the AS2–REV and KAN1–KAN2 pathways function additively in leaf polarity formation.


Discussion AS1 and AS2 have been demonstrated to play critical roles in conferring leaf adaxial identity (Xu et al. 2003, Garcia et al. 2006). AS1 and AS2 can form protein complexes (Xu et al. 2003), which may regulate different downstream effectors or interact with other components that mediate formation of normal ad/abaxial leaf polarity. In the present study, we show more detailed analyses of as1/2 double, triple and quadruple mutants. We provide more genetic and molecular data to elucidate further the genetic interactions by which AS1 and AS2 cooperate with other leaf polarity components in the regulatory network for leaf polarity. Our data suggest that AS1 and AS2 positively regulate the class III HD-ZIP genes, probably via repression of negative regulators of these genes, such as miR165/166, or AS1/AS2 independently regulate both MIR165/166 and REV/PHB/PHV expression. Although previous studies suggested that AS2 negatively regulated some abaxial-promoting genes in YAB and KAN families (Lin et al. 2003), our data indicate that AS1/AS2 do not interact genetically with the YAB and KAN genes. Dosage effects of REV, PHB and PHV expression on leaf identity Previous data showed that plants with overexpression of AS2 resulted in an increased PHB level (Lin et al. 2003). We show in this study that AS1/AS2 positively regulate REV, PHB and PHV, evidenced by both gain-of-function and loss-of-function analyses (Fig. 1). These data led us to hypothesize that the AS1/AS2 regulation may be via repression of miR165/166. This hypothesis was supported by two lines of evidence: (i) miR165/166 levels were elevated in as2 backgrounds (Li et al. 2005, Xu et al. 2006); and (ii) dicer-like1 as2 and hyponastic leaves1 as2 double mutants, in which miRNA biogenesis is defective, can rescue leaf polarity defects of as2 (Xu et al. 2006). It has been previously proposed that REV, PHB and PHV promote leaf adaxial fate in a dosage-dependent manner, with REV as the major contributor (Prigge et al. 2005). Our data agreed with the dosage-effective HD-ZIP III gene action model. Since PHB and PHV levels were reduced in rev-6 as2-101 leaves (see Supplementary Fig. S4), the rev-6 as1101 or rev-6 as2-101 double mutant should be equivalent to plants with loss of function in REV and with reduction of PHB and PHV transcript levels. The loss of function in AS1/AS2 resulted in an elevation of miR165/166 levels, which in turn partially repressed PHB and PHV. The same effect may have occurred with phb as1, phb as2, phv as1 and phv as2, in which these double mutants displayed enhanced as1 or as2 abnormalities in leaf polarity. Given that REV plays a major role in leaf adaxial fate (Prigge et al. 2005),

731

the rev as1/2 double mutant exhibited more severe loss of adaxial identity than phb as1/2 and phv as1/2. Consequently, although the single mutants rev, phb and phv did not show leaf ad/abaxial defects, each of these mutations with as1 or as2 resulted in plants with more enhanced abnormalities in leaf adaxial identity. Other possible mechanisms for the leaf polarity control Plants with as2 mutation showed a small but detectable increase in the miR165 level (Li et al. 2005, Xu et al. 2006, and this work). However, leaves from wild-type plants carrying 35S::MIR165a contained a markedly elevated level of miR165 but did not show obvious ad/abaxial leaf abnormalities (Li et al. 2005, and this work). It is possible that the miR165 increase in as2 is developmentally more relevant to the leaf ad/abaxial polarity than that of 35S::MIR165a transgenic plants, as the increase in the as2 mutant may occur earlier in leaf development. This hypothesis is supported by a recent observation that PHB::MIR165b/ Ler plants could produce the severely abaxialized leaves (Alvarez et al. 2006). Although miR165 content was increased in 35S::MIR165a/Ler transgenic plants, in terms of the spatial- and temporal-based dosages, the miR165 level might not be high enough to repress their target gene transcripts. Further reduction in the effective dosage of REV, PHB and PHV transcripts by introducing 35S::MIR165a into the rev as2 background (with reduced PHB and PHV levels in the rev mutant) would result in plants with more severely abaxialized leaves. Although 35S::MIR165a/rev transgenic plants also reflect reduced PHB and PHV levels in the rev background, the reduction of PHB/PHV transcripts by 35S::MIR165a may not be as efficient as that by as2 mutation. Since AS1/AS2 and REV/PHB/PHV all encode putative transcription factors, our data also support an alternative model where AS1/AS2 independently regulate both MIR165/166 and REV/PHB/PHV expression. Although leaves of the 35S::MIR165a/Ler plants showed a dramatic reduction in transcript levels of class III HD-ZIP genes (Li et al. 2005), expression of these genes in leaf primordia of the transgenic plant is not known. If the 35S::MIR165a construct only produces very low miR165 dosages in the leaf primordium stages while AS1/AS2 genes in the transgenic plants can normally up-regulate the REV/PHB/PHV expression, leaf polarity in the transgenic plants should be less affected. Our data also suggest that REV/PHB/PHV may also be required for MIR165/166 repression. In the rev single mutant, miR165 levels were slightly elevated, indicating that REV may negatively regulate MIR165. Although the increase of miR165/166 in rev was small, we found that this result was highly reproducible by three independent experiments (data not shown).

732


If the REV has a function to repress miR165 in planta, it may act slightly later than AS1/AS2 during leaf development, so that as1/as2 mutations show leaf polarity defects but rev does not. In the future, it should be of great interest to test these models further with additional experimental results. The molecular mechanisms responsible for the similarity of rev as2 and rdr6 as2 leaf phenotypes Our results revealed that phenotypes of rev-6 as2-101 and the previously reported rdr6-3 as2-101 were very similar. Both produced a high frequency of lotus- and needle-like leaves. Additionally, the adaxial surface of the expanded leaves was rippled and exhibited outgrowths on the adaxial surface, with some outgrowths continuing to form leaflets (Li et al. 2005). Interestingly, miR165/166 levels in rev-6 as2-101 and rdr6-3 as2-101 leaves were markedly different. In rev-6 as2-101, miR165/166 levels were only slightly increased in leaves, similar to those in as2-101 single mutant leaves, whereas in rdr6-3 as2-101, miR165/166 levels were dramatically elevated. Leaf phenotypes in rev-6 as2-101 and rdr6-3 as2-101 may both result from the same substantial reduction of the class III HD-ZIP transcripts. In rdr6 as2, the dramatically elevated endogenous miR165/166 levels strongly repressed all REV, PHB and PHV genes. It was proposed that REV/PHB/PHV and FIL act antagonistically, and the reduced REV/PHB/PHV activity in the rdr6 as2 double mutant may result in the increased FIL expression. In contrast, the rev as2 leaf phenotypes were caused by loss of function in the major adaxial-promoting factor REV, together with a reduction of the PHB and PHV dosages by as2, as mutation in as2 resulted in an increase in miR165/166 levels. This also promoted FIL mRNA levels, which mimicked the molecular environments of the rdr6 as2 double mutant. Leaves producing outgrowths have been previously reported in the kan1-2 kan2-1 double mutant leaves (Eshed et al. 2001, Eshed et al. 2004). Unlike rev as2 and rdr6 as2 double mutants, from which outgrowths appeared on the adaxial leaf side, kan1-2 kan2-1 produced outgrowths on the abaxial leaf surfaces. It was previously proposed that the juxtaposition of adaxial and abaxial identity promotes outgrowths (Waites and Hudson 1995), and our results are in agreement with this hypothesis. In the abaxial leaf surfaces of rev as2 and rdr6 as2 or in the adaxial surfaces of kan1 kan2, leaf cell identities were normal. Therefore, no outgrowth was formed on the abaxial leaf side of rev-6 as2-101 and rdr6-3 as2-101 or on the adaxial leaf side of kan1-2 kan2-1. We further examined this hypothesis by longitudinal sectioning of the rev-6 as2-101 outgrowths, and found that all five outgrowths analyzed were sandwiched between the adaxial and abaxial cell patches (see Supplementary Fig. S5).

Genetic interaction analyses between AS1/AS2 and YAB and KAN genes Our results revealed that fil yab3 as1/2 and kan1 kan2 as1/2 triple mutants had the additive leaf phenotypes of fil yab3 and as1/2 or kan1 kan2 and as1/2 mutant plants, indicating that AS1/2 did not interact genetically with FIL-YAB3 and KAN1-KAN2. However, our other data showed that the severe rev-6 as2-101 phenotypes could be partially suppressed by the fil mutation. PHB/PHV/REV and FIL/YAB3 are known to be expressed in the mutually exclusive domains (Eshed et al. 2001) and are adaxial- and abaxial-promoting genes, respectively (McConnell et al. 2001, Eshed et al. 2001, Emery et al. 2004). In the as2 single mutant, PHB/PHV/REV transcripts may be reduced to a certain level, at which the activity balance between PHB/ PHV/REV and FIL/YAB3 is not dramatically affected. Thus the plants with fil, yab3 and as1/2 mutations only showed additive phenotypes. The severe leaf polarity phenotypes of rev as2 are caused by loss of function in a major player REV together with transcript reduction of PHB and PHV due to the as2 mutation. In this genetic background, the original balance between PHB/PHV/REV and FIL/YAB3 may be severely disrupted by suppression of PHB/PHV/REV, leading to a dramatic increase of FIL/YAB3 levels. Therefore, in the fil rev as2 triple mutant, fil mutation resulted in a partial rescue of the rev as2-caused phenotypes. Our results, and those from others (Lin et al. 2003), also demonstrated by RT–PCR analyses that AS1/2 acted upstream of genes in the YAB and KAN families, although these results may reflect a change in tissue identity in lossor gain-of-function AS2 lines. The AS1–AS2 protein complex appeared to be a critical repressor, as it is known to repress several important regulatory genes including BREVIPEDICELLUS, KNAT2, KNAT6, MIR165/166, FIL and KAN1 (Byrne et al. 2000, Semiarti et al. 2001, Lin et al. 2003, Li et al. 2005). A recent study showed that AS1 protein was able to bind HIRA, a homolog of the histone chaperone, that may function in chromatin remodeling for heterochromatic and euchromatic gene silencing (Phelps-Durr et al. 2005). In the future, biochemical data would be important in further elucidation of the AS1–AS2 molecular mechanisms in the network of leaf ad/abaxial polarity establishment.

Materials and Methods Plant materials and plant growth The Arabidopsis as1-101 and as2-101 mutants in the Landsberg erecta (Ler) genetic background were generated as described previously (Sun et al. 2000, Sun et al. 2002, Xu et al. 2002). Seeds of rev-6 (Ler) were provided by S.E. Clark (University of Michigan, Ann Arbor, MI, USA), seeds of yab3-2, phb-6/phb-6 phv-5/phv-5 rev-9/þ and kan1-2/kan1-2 kan2-1/þ (all in the

Genetic interactions in leaf polarity formation Ler background) were obtained from J.L. Bowman (University of California, Davis, CA, USA), seeds of iso-2d (Columbia genetic background) were from M. Matsui (RIKEN Yokohama Institute, Kanagawa, Japan) and seeds of fil-1 (Ler) were obtained from the Arabidopsis Biological Resource Center (Ohio State University, Columbus, OH, USA). Wild-type and mutant plants were grown on soil according to our previous conditions (Chen et al. 2000). Construction of double and multiple mutants Homozygous as2-101 plants were crossed to homozygous rev-6, phb-6 and phv-5, respectively. F1 progeny of those crosses were all phenotypically normal. For the rev-6 and as2-101 cross, double mutants were recognized among the F2 plants as those with the novel phenotypes of leaves, with the F2 segregation of 256 wildtype, 92 as2-101, 95 rev-6 and 30 with novel phenotypes. To define these phenotypically novel plants, we crossed one of them to wild-type Ler. Among a total of 381 F2 plants analyzed, 214 are wild-type plants, 71 and 73 showed as2-101 and rev-6 phenotypes, respectively, and 23 displayed similar novel phenotypes to that used in the cross. These results suggested that these plants are rev-6 as2-101 double mutant. To verify that we had obtained rev-6 as2-101, we crossed five putative rev-6 as2-101 plants to rev-6 and as2-101 single mutants, respectively, and the rev-6 as2-101 rev-6 and rev-6 as2-101 as2-101 crosses yielded only rev-6 and as2-101 plants in the F1 progeny, respectively. In addition, we also crossed an additional rev allele, rev-9, to as2-101, and the F2 progeny showed a segregation of wild type : rev-9 : as2-101 : rev-9 as2-101 like novel plants in a 9 : 3 : 3 : 1 ratio, indicating that the phenotypically novel plants are indeed the rev as2 double mutant (see Supplementary Fig. S2). Homozygous phb-6 and phv-5 had no obvious phenotypic changes (Emery et al. 2003). Since both phb-6 and phv-5 mutations were caused by transposome Ds insertions, the mutant plants could be genotyped by PCR. In the F2 population of crosses phb-6 as2-101 and phv-5 as2-101, we PCR-analyzed individual as2-like plants, and identified phb-6 as2-101 using primers 50 -GGTTCCCGTCCGATTTCGACT-30 (for Ds border), 50 -CTCTTCGCTTCTCCTTCCTCTCC-30 and 50 -TATTCAACCAACTAGATACG-30 , and phv-5 as2-101 using 50 -ACGGTCGGGAAACTAGCTCTAC-30 (for Ds border), 50 -GTTCCTTGTCCTTTCTCTCTCAG-30 and 50 -TGTCTGAA GACGAGCTGACTC-30 (Emery et al. 2003). The identified phb-6 as2-101 and phv-5 as2-101 double mutants were further PCR-verified in the F3 generation, in which plants showed no further segregation of as2 phenotypes from the Ds insertion. For the fil-1 yab3-2 as2-101 triple mutant, we first obtained the fil-1 yab3-2 double mutant, which had the previously reported phenotypes (Kumaran et al. 2002). We then pollinated homozygous as2-101 pollen to homozygous fil-1 yab3-2 flowers. Among a total of 782 F2 plants analyzed, 12 showed the consistent novel phenotypes. These phenotypically novel plants, which were close to 1/64 of the total F2 plants, were considered to be the putative fil-1 yab3-2 as2-101 triple mutants. For kan1-2 kan2-1 as2-101, the homozygous as2-101 pollen was used to pollinate kan1-2/kan1-2 kan2-1/þ flowers. Seeds from individual F1 plants were collected, and F2 plants were grown by lines. One F2 segregating population (a total of 343 plants) contained five plants with novel phenotypes, close to a 63 : 1 ratio. These five plants were considered to be the putative kan1-2 kan2-1 as2-101 triple mutant. For the fil-1 rev-6 as2-101 triple mutant, the homozygous rev-6 as2-101 pollen was used to pollinate fil-1 plants. Seeds from six individual F2 rev as2-like plants were collected, and F3 plants were grown by lines. Two independent F3 populations contained 15 and 32 plants with similar novel leaf phenotypes together with 48 and 98 rev

733

as2-like plants, respectively. The segregation ratio for these two F3 populations was close to 3 : 1, and therefore these phenotypically novel plants, which displayed less rippled rev as2 leaves and similar fil rev flowers (see Supplementary Fig. S3), were considered to be putative fil-1 rev-6 as2-101 triple mutants. For construction of kan1-2 kan2-1 rev-6 as2-101, the homozygous rev-6 as2-101 pollen was used to pollinate kan1-2/kan1-2 kan2-1/þ plants. Three separate F2 populations containing one, two and 19 plants with similar novel plant phenotypes were identified from a total of 358, 572 and 4,914 plants, respectively. The segregation is consistent with a 255 : 1 ratio and, therefore, these 22 plants were considered to be the putative kan1-2 kan2-1 rev-6 as2-101 quadruple mutants. Chi-square analyses were carried out for the double, triple and quadruple mutants as shown in Supplementary Table S1. 35S::MIR165a transgenic plants The 35S::MIR165a fusion was constructed and transformed into plants as described previously (Li et al. 2005). A total of 139, 82, 41, 10, 10 and 61 transgenic lines were obtained from transforming wild-type, as2-101, rev-6, phb-6 as2-101, phv-5 as2-101 and rev-6 as2-101 plants, respectively. The transgenic plants with the most consistent phenotypes from each transformation experiment were verified by PCR using a 35S-specific primer (50 -GCTCCTACAAATGCCATCA-30 ) and a primer matching the MIR165a precursor sequence (50 -AGAGGCAATAACA TGTTGG-30 ). Real-time RT–PCR Transgenic plants carrying the 35S::AS1 fusion were generated as described previously (Sun et al. 2002). Total RNA was extracted as described previously (Huang et al. 1995), and reverse transcription was performed with 1 mg of total RNA using a kit (Fermentas, Vilnius, Lithuania). Real-time PCR was performed with Taqman probes (Sangon, Shanghai, China) for REV and FIL or with the double-stranded DNA-specific dye SYBR green (Shenyou, Shanghai, China) for PHB and PHV following the manufacturer’s instructions. Amplification was monitored in real time by a fluorometric thermal cycle machine, Rotor-Gene 3000 (Corbett Research, Sydney, Australia). Taqman probes and other primers for real-time PCR were as follows: 50 -ACACCATCACCAGAATCCAGCACAATACCG-30 (Taqman probe), and 50 -GACCTTTAACTCTCCCGCTATGT-30 and 50 -CCTCGTAGATTGGCACAGTGT-30 for ACTIN; 50 -TCG AGCTGCCACTCCATTGCATCTT-30 (Taqman probe), 50 -AGA TGCCTGGGATGAAGCCT-30 and 50 -GGTTCTAAGCTAAC AAGACCACAG-30 (exon4/exon5) for REV; 50 -CCAGCTCGG TCCTCACTCTTACTTCAATCC-30 (Taqman probe), 50 -TGTCC TCCCAGCTTCTAACCA-30 and 50 -TCTCAGCTCCTCCAGAA TATCCT-30 (exon2/exon3) for FIL; 50 -TGTGGAGAATGGAAC CAC-30 and 50 -CTAGCAGAGTTCCTTTCC-30 (exon4/exon7) for PHB; and 50 -TTGCGGAGGAGACCTTGGCG-30 and 50 -GATAGTACCACCATTTCCAGTG-30 (exon4/exon6) for PHV (Williams et al. 2005). Quantifications of each cDNA sample were made in triplicate, and the consistent results from three separately prepared RNA samples were used. For each quantification, conditions were 1 E 0.80 and r2 0.980, where E is the PCR efficiency and r2 corresponds to the correlation coefficient obtained with the standard curve. For each quantification, a melt curve was realized at the end of the amplification experiment by steps of 18C from 60 to 998C. Results were normalized to that of ACTIN.

734


MicroRNA filter hybridization, in situ hybridization and microscopy The microRNA hybridization was performed according to our previous conditions (Li et al. 2005). In situ hybridization was performed with the antisense probe for the FIL gene (Li et al. 2005), according to a method described by Long and Barton (1998). Fresh leaves of plants were examined using a SZH10 dissecting microscope (Olympus, Tokyo, Japan), and photos were taken using a Nikon E995 digital camera (Nikon, Tokyo, Japan). Preparations of SEM were described previously (Chen et al. 2000). Supplementary material Supplementary material mentioned in the article is available to online subscribers at the journal website www.pcp.oxfordjournals.org.

Acknowledgments We thank M. Matsui (RIKEN Yokohama Institute, Kanagawa, Japan), S.E. Clark (University of Michigan, Ann Arbor, MI, USA) and J.L. Bowman (University of California, Davis, CA, USA) for seeds of mutants iso-2d, rev-6 and phb-6/ phb-6 phv-5/phv-5 rev-9/þ, yab3-2 and kan1-2/kan1-2 kan2-1/þ, respectively, X.Gao for SEM, and Z. Yuan for technical assistance. This research was supported by grants from the Chinese National Scientific Foundation (30630041 and 90208009), Chinese Academy of Sciences (KSCX2-YW-N-016), and the Shanghai Scientific Committee to H.H., and a grant from SIBS for the Plant Reproductive Development to H.M.

References Alvarez, J.P., Pekker, I., Goldshmidt, A., Blum, E., Amsellem, Z. and Eshed, Y. (2006) Endogenous and synthetic microRNAs stimulate simultaneous, efficient, and localized regulation of multiple targets in diverse species. Plant Cell 18: 1134–1151. Bowman, J.L., Eshed, Y. and Baum, S.F. (2002) Establishment of polarity in angiosperm lateral organs. Trends Genet 18: 134–141. Byrne, M.E., Barley, R., Curtis, M., Arroyo, J.M., Dunham, M., Hudson, A. and Martienssen, R.A. (2000) Asymmetric leaves1 mediates leaf patterning and stem cell function in Arabidopsis. Nature 408: 967–971. Byrne, M., Timmermans, M., Kidner, C. and Martienssen, R. (2001) Development of leaf shape. Curr. Opin. Plant Biol. 4: 38–43. Chen, C., Wang, S. and Huang, H. (2000) LEUNIG has multiple functions in gynoecium development in Arabidopsis. Genesis 26: 42–54. Chen, Q., Atkinson, A., Otsuga, D., Christensen, T., Reynolds, L. and Drews, G.N. (1999) The Arabidopsis FILAMENTOUS FLOWER gene is required for flower formation. Development 126: 2715–2726. Emery, J.F., Floyd, S.K., Alvarez, J., Eshed, Y., Hawker, N.P., Izhaki, A., Baum, S.F. and Bowman, J.L. (2003) Radial patterning of Arabidopsis shoots by class III HD-ZIP and KANADI genes. Curr. Biol. 13: 1768–1774. Eshed, Y., Baum, S.F., Perea, J.V. and Bowman, J.L. (2001) Establishment of polarity in lateral organs of plants. Curr. Biol. 11: 1251–1260. Eshed, Y., Izhaki, A., Baum, S.F., Floyd, S.K. and Bowman, J.L. (2004) Asymmetric leaf development and blade expansion in Arabidopsis are mediated by KANADI and YABBY activities. Development 131: 2997–3006. Fahlgren, N., Montgomery, T.A., Howell, M.D., Allen, E., Dvorak, S.K., Alexander, A.L. and Carrington, J.C. (2006) Regulation of AUXIN RESPONSE FACTOR3 by TAS3 ta-siRNA affects developmental timing and patterning in Arabidopsis. Curr. Biol. 16: 939–944. Garcia, D., Collier, S.A., Byrne, M.E. and Martienssen, R.A. (2006) Specification of leaf polarity in Arabidopsis via the trans-acting siRNA pathway. Curr. Biol. 16: 933–938.

Huang, H., Tudor, M., Weiss, C.A., Hu, Y. and Ma, H. (1995) The Arabidopsis MADS-box gene AGL3 is widely expressed and encodes a sequence-specific DNA-binding protein. Plant Mol. Biol. 28: 549–567. Huang, W., Pi, L., Liang, W., Xu, B., Wang, H., Cai, R. and Huang, H. (2006) The proteolytic function of the Arabidopsis 26S proteasome is required for specifying leaf adaxial identity. Plant Cell 18: 2479–2492. Hudson, A. (2000) Development of symmetry in plants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 51: 349–370. Iwakawa, H., Ueno, Y., Semiart, E., Onouchi, H., Kojima, S., Tsukaya, H., Hasebe, M., Soma, T., Ikezaki, M., Machida, C. and Machida, Y. (2002) The ASYMMETRIC LEAVES2 gene of Arabidopsis thaliana, required for formation of a symmetric flat leaf lamina, encodes a member of a novel family of proteins characterized by cysteine repeats and a leucine zipper. Plant Cell Physiol. 43: 467–478. Kerstetter, R.A., Bollman, K., Taylor, R.A., Bomblies, K. and Poethig, R.S. (2001) KANADI regulates organ polarity in Arabidopsis. Nature 411: 706–709. Kidner, C.A. and Martienssen, R.A. (2004) Spatially restricted microRNA directs leaf polarity through ARGONAUTE1. Nature 428: 81–84. Kumaran, M.K., Bowman, J.L. and Sundaresan, V. (2002) YABBY polarity genes mediate the repression of KNOX homeobox genes in Arabidopsis. Plant Cell 14: 2761–2770. Li, H., Xu, L., Wang, H., Yuan, Z., Cao, X., Yang, Z., Zhang, D., Xu, Y. and Huang, H. (2005) The putative RNA-dependent RNA polymerase RDR6 acts synergistically with ASYMMETRIC LEAVES1 and 2 to repress BREVIPEDICELLUS and microRNA165/166 in Arabidopsis leaf development. Plant Cell 17: 2157–2171. Lin, W.C., Shuai, B. and Springer, P.S. (2003) The Arabidopsis LATERAL ORGAN BOUNDARIES-domain gene ASYMMETRIC LEAVES2 functions in the repression of KNOX gene expression and in adaxial– abaxial patterning. Plant Cell 15: 2241–2252. Long, J.A. and Barton, M.K. (1998) The development of apical embryonic pattern in Arabidopsis. Development 125: 3027–3035. Lynn, K., Fernandez, A., Aida, M., Sedbrook, J., Tasaka, M., Masson, P. and Barton, M.K. (1999) The PINHEAD/ZWILLE gene acts pleiotropically in Arabidopsis development and has overlapping functions with the ARGONAUTE1 gene. Development 126: 469–481. Mallory, A.C., Reinhart, B.J., Jones-Rhoades, M.W., Tang, G., Zamore, P.D., Barton, M.K. and Bartel, D.P. (2004) MicroRNA control of PHABULOSA in leaf development: importance of pairing to the microRNA 50 region. EMBO J. 23: 3356–3364. McConnell, J.R. and Barton, M.K. (1998) Leaf polarity and meristem formation in Arabidopsis. Development 125: 2935–2942. McConnell, J.R., Emery, J., Eshed, Y., Bao, N., Bowman, J. and Barton, M.K. (2001) Role of PHABULOSA and PHAVOLUTA in determining radial patterning in shoots. Nature 411: 709–713. Nakazawa, M., Ichikawa, T., Ishikawa, A., Kobayashi, H., Tsuhara, Y., Kawashima, M., Suzuki, K., Muto, S. and Matsui, M. (2003) Activation tagging, a novel tool to dissect the functions of a gene family. Plant J. 34: 741–750. Ori, N., Eshed, Y., Chuck, G., Bowman, J.L. and Hake, S. (2000) Mechanisms that control knox gene expression in the Arabidopsis shoot. Development 127: 5523–5532. Pekker, I., Alvarez, J.P. and Eshed, Y. (2005) Auxin response factors mediate Arabidopsis organ asymmetry via modulation of KANADI activity. Plant Cell 17: 2899–2910. Phelps-Durr, T.L., Thomas, J., Vahab, P. and Timmermans, M.C. (2005) Maize rough sheath2 and its Arabidopsis orthologue ASYMMETRIC LEAVES1 interact with HIRA, a predicted histone chaperone, to maintain knox gene silencing and determinacy during organogenesis. Plant Cell 17: 2886–98. Prigge, M.J., Otsuga, D., Alonso, J.M., Ecker, J.R., Drews, G.N. and Clark, S.E. (2005) Class III homeodomain-leucine zipper gene family members have overlapping, antagonistic, and distinct roles in Arabidopsis development. Plant Cell 17: 61–76. Qi, Y., Sun, Y., Xu, L., Xu, Y. and Huang, H. (2004) ERECTA is required for protection against heat-stress in the AS1/AS2 pathway to regulate adaxial–abaxial leaf polarity in Arabidopsis. Planta 219: 270–276. Sawa, S., Watanabe, K., Goto, K., Liu, Y.G., Shibata, D., Kanaya, E., Morita, E.H. and Okada, K. (1999) FILAMENTOUS FLOWER,

Genetic interactions in leaf polarity formation a meristem and organ identity gene of Arabidopsis, encodes a protein with a zinc finger and HMG-related domains. Genes Dev. 13: 1079–1088. Semiarti, E., Ueno, Y., Tsukaya, H., Iwakawa, H., Machida, C. and Machida, Y. (2001) The ASYMMETRIC LEAVES2 gene of Arabidopsis thaliana regulates formation of a symmetric lamina, establishment of venation and repression of meristem-related homeobox genes in leaves. Development 128: 1771–1783. Siegfried, K.R., Eshed, Y., Baum, S.F., Otsuga, D., Drews, G.N. and Bowman, J.L. (1999) Members of the YABBY gene family specify abaxial cell fate in Arabidopsis. Development 126: 4117–4128. Sun, Y., Zhou, Q., Zhang, W., Fu, Y. and Huang, H. (2002) ASYMMETRIC LEAVES1, an Arabidopsis gene that is involved in the control of cell differentiation in leaves. Planta 214: 694–702. Sun, Y., Zhang, W., Li, F.L., Guo, Y.L., Liu, T.L. and Huang, H. (2000) Identification and genetic mapping of four novel genes that regulate leaf development in Arabidopsis. Cell Res. 10: 325–335. Waites, R. and Hudson, A. (1995) phantastica: a gene required for dorsoventrality of leaves in Antirrhinum majus. Development 121: 2143–2154. Waites, R., Selvadurai, H.R., Oliver, I.R. and Hudson, A. (1998) The PHANTASTICA gene encodes a MYB transcription factor involved in growth and dorsoventrality of lateral organs in Antirrhinum. Cell 93: 779–789.

735

Williams, L., Grigg, S.P., Xie, M., Christensen, S. and Fletcher, J.C. (2005) Regulation of Arabidopsis shoot apical meristerm and lateral organ formation by microRNA miR166g and its AtHD-ZIP target genes. Development 132: 3657–3668. Xu, L., Xu, Y., Dong, A., Sun, Y., Pi, L. and Huang, H. (2003) Novel as1 and as2 defects in leaf adaxial–abaxial polarity reveal the requirement for ASYMMETRIC LEAVES1 and 2 and ERECTA functions in specifying leaf adaxial identity. Development 130: 4097–4107. Xu, L., Yang, L., Pi, L., Liu, Q., Ling, Q., Wang, H., Poethig, R.S. and Huang, H. (2006) Genetic interaction between the AS1–AS2 and RDR6– SGS3–AGO7 pathways for leaf morphogenesis. Plant Cell Physiol. 47: 853–863. Xu, Y., Sun, Y., Liang, W. and Huang, H. (2002) The Arabidopsis AS2 gene encoding a predicted leucine-zipper protein is required for the leaf polarity formation. Acta Bot. Sin. 44: 1194–1202. Yang, L., Huang, W., Wang, H., Cai, R., Xu, Y. and Huang, H. (2006) Characterizations of a hypomorphic argonaute1 mutant reveal novel AGO1 functions in Arabidopsis lateral organ development. Plant Mol. Biol. 61: 63–78. Zhong, R. and Ye, Z.H. (2004) Amphivasal vascular bundle 1, a gain-offunction mutation of the IFL1/REV gene, is associated with alterations in the polarity of leaves, stems and carpels. Plant Cell Physiol. 45: 369–385.

(Received January 20, 2007; Accepted March 26, 2007)