like Homeobox gene, is required for lateral ... - Wiley Online Library

The Plant Journal (2015) 81, 480–492

doi: 10.1111/tpj.12743

LOOSE FLOWER, a WUSCHEL-like Homeobox gene, is required for lateral fusion of floral organs in Medicago truncatula Lifang Niu1,2,†, Hao Lin1,2,†, Fei Zhang1, Tezera W. Watira1, Guifen Li3, Yuhong Tang3, Jiangqi Wen3, Pascal Ratet4, Kirankumar S. Mysore3 and Million Tadege1,* 1 Department of Plant and Soil Sciences, Institute for Agricultural Biosciences, Oklahoma State University, 3210 Sam Noble Parkway, Ardmore, OK 73401, USA, 2 Biotechnology Research Institute, Chinese Academy of Agricultural Sciences, Beijing 100081, China, 3 Plant Biology Division, The Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway, Ardmore, OK 73401, USA, and 4 Institut des Sciences du Ve ge tal, Centre National de la Recherche Scientifique, Saclay Plant Sciences, 91198 Gif sur Yvette Cedex, France Received 30 July 2014; accepted 3 December 2014; published online 9 December 2014. *For correspondence (e-mail [email protected].) † These authors contributed equally to this study.

SUMMARY The Medicago truncatula WOX gene, STENOFOLIA (STF), and its orthologs in Petunia, pea, and Nicotiana sylvestris are required for leaf blade outgrowth and floral organ development as demonstrated by severe phenotypes in single mutants. But the Arabidopsis wox1 mutant displays a narrow leaf phenotype only when combined with the prs/wox3 mutant. In maize and rice, WOX3 homologs are major regulators of leaf blade development. Here we investigated the role of WOX3 in M. truncatula development by isolating the lfl/wox3 loss-of-function mutant and performing genetic crosses with the stf mutant. Lack of WOX3 function in M. truncatula leads to a loose-flower (lfl) phenotype, where defects are observed in sepal and petal development, but leaf blades are apparently normal. The stf lfl double mutant analysis revealed that STF and LFL act mainly independently with minor redundant functions in flower development, but LFL has no obvious role in leaf blade outgrowth in M. truncatula on its own or in combination with STF. Interestingly, LFL acts as a transcriptional repressor by recruiting TOPLESS in the same manner as STF does, and can substitute for STF function in leaf blade and flower development if expressed under the STF promoter. STF also complements the lfl mutant phenotype in the flower if expressed under the LFL promoter. Our data suggest that the STF/WOX1 and LFL/WOX3 genes of M. truncatula employ a similar mechanism of action in organizing cell proliferation for lateral outgrowth but may have evolved different cis elements to acquire distinct functions. Keywords: LFL, STF, WOX, floral organ fusion, leaf blade development, Medicago truncatula, homeobox, transcriptional repression, TPL.

INTRODUCTION Legumes comprise one of the largest monophyletic families with approximately 700 genera and 18 000 species (Dong et al., 2005) and are second only to grasses in economic and nutritional value (Graham and Vance, 2003; Benlloch et al., 2006). In addition to being used as an important source of dietary proteins for humans, legumes play a crucial role in soil fertility and maintaining the nitrogen cycle through fixing atmospheric nitrogen in symbiotic association with Rhizobium bacteria. The subfamily Papilionoideae is the largest of three subfamilies that make up 480

the family Leguminosae with approximately 455 genera and 12 000 species (Tucker, 2003; Dong et al., 2005), and includes economically important species such as pea, soybean, and the model species Lotus japonicus and Medicago truncatula. Members of Papilionoideae are known for their peculiar arrangement of zygomorphic flowers with pentamerous whorls of sepals and petals, two whorls of stamens, and a single carpel (Tucker, 2003). As a member of this subfamily, M. truncatula possesses 21 floral organs: five sepals, five petals, 10 stamens in two whorls, and a © 2014 The Authors The Plant Journal © 2014 John Wiley & Sons Ltd

LFL functions in floral organ fusion 481 single carpel. The five sepals are at the outermost and are fused at the base but separate on the top half. The five petals are of three types. The upper/adaxial petal, called the standard or vexillum, is the largest and envelops the rest of the petals at the early stages of flower development. The two bottom petals are fused together at the apex, forming a boat-like structure called the keel at the abaxial position. The two lateral petals, called the alae or wings, surround the two keel petals, which form a butterflyshaped corolla. The nine stamens are fused surrounding the central carpel forming the staminal tube. The tenth stamen, called vexillary stamen, is separate but loosely attached adaxially to the staminal tube at the base for most of the length of the filament and hangs free at the top (Tucker, 2003). The molecular mechanism by which this peculiar organ arrangement and fusion is regulated is unclear. It has been shown that WOX transcription factors play key roles in lateral organ development and organ fusion in different plant species. In Petunia x hybrida, MAEWEST (MAW), a member of the WOX1 subfamily, is required for petal, sepal, and carpel fusion, as well as for lateral outgrowth of the leaf blade (Vandenbussche et al., 2009). The lateral expansion of petals and sepals are strongly reduced in maw mutants resulting in a failure of proper fusion of the petals and sepals. The maw mutant flowers also display carpel fusion defects, with a split stigma and style phenotypes, resulting in reduced fertility (Vandenbussche et al., 2009). In M. truncatula and woodland tobacco (Nicotiana sylvestris), the WOX1 homologs STENOFOLIA (STF) and LAMINALESS1 (LAM1), respectively, are required for blade outgrowth and floral organ expansion (McHale, 1992; Tadege et al., 2011a,b). The lack of lateral expansion is evident in stf flowers in which the outer petal (vexillum) is narrow, and fails to cover the rest of the flower early in development. The staminal tube and ovary are partially open in which protruding ectopic ovules become visible, resulting in female sterility (Tadege et al., 2011a). The lam1 mutant phenotype is stronger than both the stf and maw mutant phenotypes, in which leaf blades are drastically reduced laterally and transition to the floral phase is prevented altogether (McHale, 1992; Tadege et al., 2011a). Rarely, lam1 plants flower spontaneously when grown at higher temperatures or when sprayed with GAs, but flowers are female sterile and display phenotypes similar to the stf mutant (McHale, 1992; Tadege et al., 2011a). Similarly, in pea, the WOX1 homolog, LATHYROIDES (LATH), is involved in controlling leaf blade lateral outgrowth, and zygomorphic flower development (Zhuang et al., 2012). These mutant phenotypes indicate that the WOX1 homologs play a conserved role in leaf blade lateral outgrowth and floral organ fusion in both the Leguminosae and Solanaceae families. However, in Arabidopsis, although overexpression of WOX1 leads to defects in meristem and leaf

development (Zhang et al., 2011), the wox1 mutant shows a normal leaf and flower phenotype (Vandenbussche et al., 2009; Nakata et al., 2012). The reason for this anomaly appears to have come from WOX1 functional redundancy with another related WOX gene, PRS also called WOX3 in Arabidopsis. In the prs/wox3 mutant, growth of the lateral sepals is repressed, and although the size and shape of the abaxial and adaxial sepals are normal, the cell files at the lateral margins are missing (Matsumoto and Okada, 2001). It was also found that stipules are deleted from the leaves and lateral stamens are deleted from the flowers of prs mutant (Nardmann et al., 2004). In wox1 prs double mutant, flowers have narrower petals and sepals, and the lateral petals are often missing or reduced as can be observed in single prs mutant (Vandenbussche et al., 2009). Interestingly, the wox1 prs double mutant displays a leaf blade defect comparable with Petunia maw, but fusion of the carpels and fertility are unaffected, indicating that WOX1 and PRS are functionally redundant in regulating leaf blade development in Arabidopsis. In monocots, the WOX1 gene is absent (Vandenbussche et al., 2009; Tadege et al., 2011a,b; Nardmann and Werr, 2013) and leaf blade outgrowth is regulated by WOX3 homologs. In maize, two WOX3 homologs, NARROW SHEATH 1 and 2 (NS1 and NS2) redundantly control leaf blade outgrowth in which the ns1 ns2 double mutant but not the single mutants display a leaf margin deletion phenotype (Scanlon et al., 1996; Nardmann et al., 2004). Similarly in rice, duplicate NS orthologs, NARROW LEAF2 and 3 (NAL2 and NAL3) together control leaf blade outgrowth (Cho et al., 2013; Ishiwata et al., 2013). It is intriguing to note that the strong leaf blade phenotypes of PRS homologs ns1 ns2 double mutant in maize are recapitulated by single WOX1 homologs stf and lam1 mutants in M. truncatula and N. sylvestris, respectively, while such a phenotype required the wox1 prs double mutant in Arabidopsis. Whether NS/WOX3 is the major regulator of leaf development in monocots and fulfills this role redundantly with WOX1 in dicots is not clear because the function of WOX3 has not been studied in dicots other than Arabidopsis. The leaf and flower phenotypes of wox1 mutants in other dicot species suggest that the functional redundancy between WOX1 and WOX3 may not be widespread. Nevertheless, WUS clade WOX genes have been proposed to have a common mechanism of action (Lin et al., 2013a,b) as demonstrated by complementation of the wox5 and prs mutant phenotypes by WUS, as well as by the partial complementation of prs phenotypes with WOX4 (Sarkar et al., 2007; Shimizu et al., 2009; Ji et al., 2010). We have recently shown that all modern/WUS clade WOX genes including WUS, WOX1-WOX7 that contain a WUSbox can substitute for STF/LAM1 function in regulating leaf blade outgrowth if expressed under the control of the STF promoter, while the intermediate/WOX9 and ancient/

© 2014 The Authors The Plant Journal © 2014 John Wiley & Sons Ltd, The Plant Journal, (2015), 81, 480–492

482 Lifang Niu et al. WOX13 clade members cannot, due to lack of the repressor WUS-box motif (Lin et al., 2013a,b). STF promotes cell proliferation at the adaxial-abaxial junction mainly acting as a transcriptional repressor and the evolutionarily conserved WUS-box, which is specific to the WUS clade WOX genes, partially contributes to this repressive activity (Lin et al., 2013a,b). The other partial repressive activity is conferred by another conserved motif at the C-terminus, STF-box, which is specific to STF/WOX1 homologs. The STF-box and WUS-box of STF cooperatively recruit the co-repressor TOPLESS (TPL) to repress AS2 in the leaf margin, and regulate leaf blade outgrowth and floral organ development in M. truncatula (Zhang et al., 2014). These observations suggest that this activity may be at least partially shared by other WUS clade WOX genes including WOX3 and may underlie the reason for functional redundancy or mutant complementation. To investigate the molecular basis of floral organ fusion in Papilionoideae, we characterized a loose-flower mutant, lfl, which is defective in floral organ fusion, by screening Tnt1 retrotransposon-tagged lines of M. truncatula. LFL encodes a WOX domain transcription factor, homologous to At-PRS. The lfl mutant offered us the opportunity to dissect the molecular function of WOX3 in M. truncatula leaf and flower development and enabled us to evaluate mechanistic similarities and any redundancy with STF analogous to the WOX1 and PRS redundancy in Arabidopsis. LFL/MtWOX3 mainly acts as a transcriptional repressor in regulating flower development and physically interacts with the co-repressor TPL similar to that of STF. However, genetic analysis indicated that LFL and STF act mainly independently with minor redundant functions in flower development, but LFL has no obvious role in leaf blade outgrowth in M. truncatula on its own or in combination with STF. Our data suggest that the STF/WOX1 and LFL/WOX3 genes of M. truncatula employ a similar repressive mechanism of action in organizing cell proliferation but may have evolved different cis elements to acquire distinct functions in regulating growth of separate lateral organs. RESULTS Identification of M. truncatula lfl mutants Two identical loose-flower mutants named lfl-1 and lfl-2 were identified in forward genetic screening of Tnt1 retrotransposon-tagged lines of M. truncatula genotype R108 (Tadege et al., 2008, 2009; Yarce et al., 2013). In the lfl mutant, the sepals failed to cover the corolla properly and the vexillum failed to enclose the rest of the petals, and the keel failed to enclose the stigma and anthers, resulting in the exposure of the stigma and anthers at early flower developmental stage (Figure 1a–f). A close examination of the sepals and petals showed fusion defects in lfl sepals (Figures 1g, h and S1) and petals (Figure 1i–l) compared

with the wild-type R108. In wild-type M. truncatula, the keel petals are joined and two wing petals are normally fused to the keel petals (Figure 1j), while in the lfl mutant, the degree of fusion of the keel petals is largely decreased and the wing petals are separated from the keel petals (Figure 1l). In addition, the carpel is enclosed by the staminal tube in wildtype, so that the pollen can fall on the stigma at the time of pollination (Figure 1m, o). In the lfl mutant, the carpel is not enclosed by the staminal tube and protrudes away from the staminal tube, reducing the efficiency of pollen deposition on the stigma leading to reduced fertility, although some flowers can still develop seed pods (Figure 1n, p). But, no obvious leaf defects were observed under our standard growth conditions. These phenotypes indicate that LFL function is highly relevant to the flower development in relation to lateral floral organ fusion. Molecular cloning and expression pattern of the LFL gene The LFL gene was cloned by PCR-based genotyping of flanking sequence tags (FST) in segregating populations (Tadege et al., 2008). Flanking sequence analysis of Tnt1 retrotransposon in the lfl-1 mutant revealed that one of 15 flanking sequences (FST 7) segregated with the mutant phenotype by PCR-based genotyping. The lfl mutant phenotype segregates as single Mendelian recessive mutation and all mutant plants genotyped homozygous for FST 7. We also confirmed by PCR analysis that, in the second allele (lfl-2), there was a Tnt1 insertion in FST 7 that segregates as homozygous with the mutant phenotype. The fulllength gene sequence corresponding to this particular FST was recovered and designated to represent the LFL gene. In both lfl-1 and lfl-2 alleles, the Tnt1 retrotransposon was inserted in the first exon of LFL with only one nucleotide difference (Figure 2a). RT-PCR analysis showed that the transcription of full-length LFL was abolished in both the lfl-1 and lfl-2 mutants (Figure 2b). To further confirm that the mutant phenotype was caused by disruption of this gene, a construct including a 2.6-kb promoter region immediately upstream the translational start and the LFL CDS was transformed into homozygous lfl-1 mutant plants and transgenic plants expressing the LFL transgene at various levels were obtained (Figure S2). Phenotypic observation showed that the lfl mutant phenotypes were fully complemented comparable with the wild-type (Figure 2c and Table S1). Sequence alignment and phylogenetic analysis showed that LFL encodes a putative homeodomain protein of the WOX family that is homologous to Arabidopsis PRS/WOX3. LFL and PRS share 42% amino acid identity using full-length sequences with strong conservation at the homeodomain and WUS-box regions (Figure S3). RT-PCR analyses in different tissues revealed that the expression level of LFL transcript is high in young and mature flower, medium in young leaf, pod, stem and cotyledon, low in mature leaf and root (Figure 2d).


LFL functions in floral organ fusion 483

(a)

(c)

(e)

(b)

(g)

(h)

(i)

(j)

(k)

(l)

(d)

(f)

(m)

(n)

(o)

(p)

Figure 1. The lfl mutant of M. truncatula shows defects in floral organ development. (a–f) Flower phenotype in the wild-type (a, c, e) and lfl mutant (b, d, f) at different stages of development. Bars in (a, b) = 200 lm, Bars in (c–f) = 1 mm. (g, h) Dissected sepals in the wild-type (g) and lfl mutant (h). Bars = 1 mm. (i, j) Dissected petals of the wild-type. The top view of vexillum (i), the bottom view (left) and top view (right) of fused alae and keel petals (j). Red arrows point to the bottom of fused alae and keel. White arrow points to the top of fused keel petals. Bars in (i, j) = 1 mm. (k, l) Dissected petals of lfl mutant. The top view of vexillum (k), separated alaes and keel (l). White arrow points to the top of fused keel. Bars in (k, l) = 1 mm. (m, o) Show the carpel is enclosed by the staminal tube in the wild-type. Bar in (m) = 1 mm, bar in (o) = 2.5 mm. (n, p) Show the staminal tube fails to enclose the carpel in the lfl mutant. White arrows point to stigmas. Bar in (n) = 1 mm, bar in (p) = 2.5 mm.

To further investigate the tissue-specific expression pattern of LFL, we performed promoter–GUS fusion and RNA in situ hybridization analysis. Fusing the 2.6-kb promoter to a b-glucuronidase (GUS) reporter gene uidA showed development and tissue-specific GUS expression pattern in flowers. In very young flower buds, GUS staining was detected at the sepal margin in the distal half and joint part, including the sepal tip (Figure 3a). As the flower develops further, the GUS signals appeared to be restricted to

the sepal fusion site and sepal lobes (Figure 3b). In the mature flower, GUS staining was detected at the joint region between the wings and keel petals, as well as at the tip of the staminal tube with strong staining of the anthers (Figure 3c, d). We further analyzed the tissue-specific expression patterns of LFL in vegetative shoot apex and developing flowers using RNA in situ hybridization. In the 4-week-old wild-type R108 shoot apex, LFL was mainly expressed at the primordium initiation site and incipient


484 Lifang Niu et al.

(a)

(b)

(c)

(d)

Figure 2. Molecular cloning and confirmation of the LFL gene. (a) Schematic representation of the gene structure of LFL showing the Tnt1 insertion sites in lfl-1 and lfl-2 mutants. (b) RT-PCR showing the level of LFL transcript in young flowers of wild-type (WT) and mutant lines. (c) Floral organ phenotypes of lfl-1 mutant complemented with pLFL::LFL. Bars = 2.5 mm. Upper panel shows flower phenotype and lower panel shows the staminal tube phenotype. (d) RT-PCR analysis of LFL expression in different plant organs. Actin was used as the loading control.

leaf primordium but not in the central region of the shoot apical meristem and leaves (Figure 3e). A very weak signal was detected in young leaf primordia (P2 and P3 stages). In the reproductive tissue, LFL was expressed in the emerging sepals and petals but not detected in the floral meristem (Figure 3f). A cross-section through the flower also detected a faint signal in the carpel fusion site (Figure 3g). These specific expression patterns are consistent with LFL functioning in regulating lateral fusion during the morphogenesis and maturation of floral organs. LFL mainly acts as a transcriptional repressor in regulating flower development To investigate the subcellular localization of LFL, we transiently expressed full-length LFL cDNA with

C-terminal fusion to green fluorescent protein reporter GFP driven by CaMV 35S promoter in Nicotiana benthamiana epidermal cells. The fluorescence of LFL–GFP was mainly localized to the nucleus (Figure S4), confirming that LFL functions as a nuclear localized transcriptional regulator. To test whether LFL functions as a transcriptional repressor or activator in floral organ development, we first examined its transcriptional activity in Arabidopsis protoplasts by using luciferase transient expression assays. The effector plasmid was constructed by fusing the LFL CDS to the GAL4 DNA binding domain, while the reporter plasmid was constructed by fusing the luciferase (LUC) gene to a 59 GAL4 binding sites. Both constructs were driven by the 35S promoter (Figure 4a). Bioluminescence measurements showed that luciferase activity was reduced by more than two-fold in the presence of the LFL protein (Figure 4b), indicating strong repressive activity. The WUS-box of WOX proteins is shown to be a repressor domain (Ikeda et al., 2009; Lin et al., 2013a,b). Mutation of two amino acids (leucine-to-alanine exchange) in the WUS-box of LFL (LFLm) or deletion of the C-terminal including the complete WUS-box of LFL (LFLt) greatly relieved this repression (Figure 4b), suggesting that LFL functions as a transcriptional repressor, and its WUS-box is mainly responsible for mediating this repressive activity. This is consistent with the proposed mechanism of WUS clade WOX genes primarily acting as transcriptional repressors (Lin et al., 2013a,b). To confirm the repressive function of LFL in floral organ development, complementation analysis was performed using the LFL mutant fusion constructs in lfl mutant plants. In these experiments, the mutant LFLm was used alone or fused to either the exogenous EAR motif repression domain SRDX or the activation domain of the herpes simplex virus VP16 protein and introduced into lfl-1 mutant driven by the LFL promoter. The pLFL::LFLm construct failed to complement the lfl-1 mutant flower phenotypes (Figure 4c, f and Table S1). pLFL::SRDX-LFLm transgenic plants, on the other hand, showed full complementation of lfl-1 mutant that was indistinguishable from wild-type (Figure 4d, g), while pLFL::LFLm-VP16 transgenic plants showed no complementation (Figure 4e, h), indicating that repressor but not activator activity is required for LFL function. The transcript abundance of LFL in transgenic lfl plants is shown in Figure S2. Taken together, these results indicate that LFL mainly acts as a transcriptional repressor in regulating floral organ development. LFL physically interacts with the M. truncatula TOPLESS protein In M. truncatula, STF physically interacts with the corepressor TOPLESS (Mt-TPL) to mediate leaf blade outgrowth. The STF-box and WUS-box of STF are



(a)

(e)

(b)

(c)

(f)

(d)

(g)

Figure 3. Expression pattern analysis of LFL by GUS staining and RNA in situ hybridization. (a–d) Histochemical GUS activity analysis in the flowers of wild-type plants transformed with pLFL::GUS construct. GUS staining in whole flower at different stages (a–c). GUS staining in the staminal tube (d). Red arrowheads indicate GUS staining in sepals and the fusion sites between alaes and keel. Bars = 1 mm. (e–g) RNA in situ hybridization analysis of LFL mRNA in vegetative and reproductive apices of the wild-type. Longitudinal sections of 4-week-old vegetative shoot apex (e) and reproductive apex (f). Cross-section of young flower (g). Arrowheads point to signals. Bars = 50 lm.

additively required for interaction with Mt-TPL (Zhang et al., 2014). As LFL mainly acts as a transcriptional repressor through its WUS-box in regulating floral organ development, we wondered whether LFL can interact with the co-repressor TPL. We performed yeast two-hybrid assay to examine interaction between LFL and Mt-TPL. We found that LFL can interact with Mt-TPL under stringent conditions (Figure 5a). Besides, LFL also interacted in Y2H with two of the four TPL-related (TPR) proteins tested (Figure S5). The interaction between LFL and Mt-TPL family proteins can be abolished by mutation of the WUS-box (Figures 5b and S5). Further bimolecular fluorescence complementation (BiFC) analysis using split YFP (Lu et al., 2010) confirmed that LFL can interact with Mt-TPL in the nucleus reconstituting the yellow fluorescence when both proteins are transiently expressed in N. benthamiana leaf epidermal cells (Figure 5c). Consistent with the Y2H results, mutation of the WUS-box of LFL (LFLm) abolished BiFC interaction with Mt-TPL (Figure 5d). These results indicate that LFL interacts with Mt-TPL through its

WUS-box and suggests that this interaction is required for LFL repressive function in floral organ development. Genetic analysis of LFL and STF in M. truncatula flower and leaf development In Arabidopsis, PRS is required for the development of lateral sepals and stamens in the flower and stipules at the leaf base (Matsumoto and Okada, 2001; Nardmann et al., 2004), and redundantly with WOX1 regulates leaf blade outgrowth, leaf margin development, and sepals and petals expansion (Vandenbussche et al., 2009; Nakata et al., 2012). We have previously reported that STF, the M. truncatula homolog of WOX1, is required for leaf blade outgrowth and leaf vascular patterning (Tadege et al., 2011a). To investigate the possible role of LFL in M. truncatula leaf development, the heterozygous lfl+/ was crossed with stf+/ and the resulting double mutant was analyzed for morphological phenotypes. Compared with wild-type and the phenotypes of single mutants, the lfl stf double mutant displayed a leaf phenotype similar to stf, but



(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

showed an additive effect on flower development (Figure 6). The lfl mutant leaf is indistinguishable from wild-type under our growth conditions (Figure 6a, b), and the stf mutant narrow leaf blade is indistinguishable from the lfl stf double mutant leaf blade (Figure 6c, d), suggesting that LFL is not directly involved in regulating leaf blade development individually or together with STF. In the lfl flower, the keel sepals are poorly fused and the wing petals are separated from the keel petals, while in the stf mutant flower, the petals are narrow and cannot enclose the anthers and stigma (Figure 6f, g), but in the lfl stf double mutant flowers, these phenotypes become additive combining both phenotypes (Figure 6h). The staminal tube is open in both the lfl and stf mutants but the carpel grows away from the staminal tube in lfl, while in stf the carpel remains attached to the staminal tube and often ectopic ovules are visible (Figure 6j, k, n, o). The lfl stf double mutant flower, conversely, combines the flower phenotypes of both lfl and stf mutants, and occasionally a second carpel arises from the first displaying a split carpel with two stigmas, which is similar to single stf mutant

Figure 4. LFL mainly acts as a transcriptional repressor in flower development. (a) Schematic representation of reporter and effector constructs used in transient expression assay. (b) Relative luciferase activities measured as bioluminescence in Arabidopsis protoplasts. Error bars indicates standard deviation (SD) (n = 3). **P < 0.01 (t-test). (c–h) Floral organ phenotypes of lfl-1 mutant plants complemented with mutated LFL construct pLFL::LFLm (c, f), exogenous repressor domain fused to the mutant pLFL::SRDX-LFLm1 (d, g), and exogenous activator domain fused to the mutant pLFL::LFLm1-VP16 (e, h). Bars = 1 mm.

(Figures 6l, p and S6). The opening of the ovary is wider and more frequent in the lfl stf double than in the stf single mutant, suggesting some redundant functions. These results clearly indicate that STF and LFL have distinct functions in floral organ and leaf blade development in M. truncatula with minimal redundancy in regulating petal and carpel development. The lack of obvious role for LFL in leaf blade development or its apparent lack of redundancy with STF explains why the stf single mutant displays a strong leaf blade outgrowth defect in the presence of wild-type LFL in M. truncatula. LFL and STF can function interchangeably if expressed under each other’s promoters The fact that both LFL and STF act as transcriptional repressors by interacting with TPL, and their WUS-box serves as a repressor domain required for the function of both proteins, prompted us to ask whether these two genes have similar mechanisms of action, and are functionally equivalent in regulating flower and leaf blade development. To test this hypothesis, we reciprocally expressed LFL in stf mutant



(a)

(b)

(c)

(d)

Figure 5. LFL physically interacts with the co-repressor Mt-TPL. (a) Interaction between LFL, and Mt-TPL in the Y2H assay. Interaction was examined by the growth of yeast cells on double dropout medium (DDO) and quadruple dropout medium (QDO)/X-Gal plates. Empty BD vector was used as a negative control. (b) Y2H interaction between the mutant form LFLm and Mt-TPL. (c, d) BiFC interaction in tobacco epidermal cells using split YFP between the WT LFL and Mt-TPL (c) and the mutant form LFLm and Mt-TPL (d). Bars = 50 lm.

driven by the STF promoter, and STF in lfl mutant driven by the LFL promoter. LFL complemented the stf mutant leaf blade outgrowth and flower phenotypes. Although a range of good, medium and poor complementation phenotypes were observed (Table S1), in the best complemented scenario, the leaves and flowers of stf appeared comparable with wild-type with reduced fertility (Figure 7a–d). The same construct also complemented both the leaf and flower phenotypes of the N. sylvestris lam1 mutant to a large extent (Figure S7). These observations indicate that LFL can replace the function of LAM1 and STF in regulating leaf blade development provided that it is expressed in the right

domains. This result is consistent with our previous finding that all WUS clade member of Arabidopsis WOX genes with WUS-box can substitute for STF in complementing the lam1 mutant leaf blade (Lin et al., 2013a,b). Conversely, STF can also complement the flower phenotypes of lfl when expressed under LFL promoter (Figure 7e–h and Table S1). Taken together, our data highlight a key observation that LFL/WOX3 and STF/WOX1 have mostly independent functions in flower development, and unlike the Arabidopsis WOX1 and PRS redundancy; LFL has no obvious redundant role with STF in leaf blade development in M. truncatula. Consistent with this, LFL and STF showed mostly non-overlapping expression patterns (Figure S8). We compared transcript abundance of these two genes in the same or comparable tissues by semi quantitative RT-PCR and RNA in situ hybridization. While both LFL and STF were abundantly expressed in juvenile flower, LFL was weakly detectable in juvenile leaf where STF was strongly expressed (Figure S8a). Comparing expression pattern by in situ in the vegetative shoot apex also revealed that LFL was expressed in the primordium initiation site but absent from the leaf margin while STF was expressed in the leaf margin after primordium formation (Figure S8b, c). These findings suggest that despite mechanistic conservation, LFL and STF may have evolved functional specialization through acquisition of different cis elements to independently regulate the development of different lateral organs. We have recently shown that STF binds to multiple regions of the Mt-AS2 promoter and represses its expression in the leaf margin (Zhang et al., 2014). To evaluate the extent of mechanistic conservation between LFL and STF action, we determined whether LFL can also repress the activity of Mt-AS2. Real-time PCR analysis of AS2 expression in the leaf tissue revealed that the level of AS2 in the lfl mutant was similar to that of wild-type R108 compared with approximately two-fold increase in the stf mutant, indicating that LFL normally does not affect the STF target gene, AS2, in the leaf. However, in the pSTF::LFL complemented stf mutant plants, AS2 expression was reduced back to R108 levels (Figure 8a), suggesting that LFL can repress AS2, if expressed in the STF domain. To test with a different approach that LFL can indeed repress AS2 promoter activity, we performed luciferase assay in Arabidopsis protoplast. The reporter construct was made by fusing the luciferase (LUC) gene to mini 35S driven by 3 kb Mt-AS2 promoter, and the GUS control and LFL effector constructs were driven by the 35S promoter (Figure 8b). Transformation of the LFL effector construct into the protoplast, reduced luciferase activity by five-fold compared with the GUS control (Figure 8c), confirming that LFL can repress the AS2 promoter activity. These observations are consistent with our hypothesis that LFL and STF may have evolved different cis elements to acquire distinct functional specialization in M. truncatula.



(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

DISCUSSION Repressive activity is critical for LFL function We have shown that the Arabidopsis PRS and maize NS homolog LFL has no obvious function in regulating leaf blade development in M. truncatula by characterizing two Tnt1 retrotransposon-tagged mutants lfl-1 and lfl-2. In the trifoliate legume model species M. truncatula, the function of LFL/WOX3 is restricted to the flower in which lfl-loss-offunction mutation leads to aberrant fusion of sepals, petals, and staminal tube resulting in a loose-flower phenotype and greatly reduced fertility (Figures 1 and S1). LFL primarily acts as a transcriptional repressor to orchestrate its functions in flower development (Figure 4) and physically interacts with the transcriptional co-repressor Mt-TPL (Figure 5). Both repressor activity and interaction with TPL are mediated by the WUS-box of LFL in which deletion or mutation of the WUS-box abolishes function and

Figure 6. Genetic analysis of LFL and STF in regulating lateral organ development. (a–d) Leaf phenotypes of wild-type (a), lfl (b), stf (c), lfl stf (d). Bars = 2 mm. (e–h) Flower phenotypes of wild-type (e), lfl (f), stf (g), lfl stf (h). Bars = 5 mm. (i–p) staminal tube and the carpel phenotypes in wild-type (i, m), lfl (j, n), stf (k, o), lfl stf (l, p). Bars = 2.5 mm. White arrows point to stigmas and red arrows point to ectopic ovules.

interaction with TPL (Figures 4 and 5), indicating that repression is central to LFL function, and TPL is a key component of this repressive activity. This is consistent with our previous proposal that all WUS clade WOX genes including WUS and WOX1-7 primarily function by a repressive mechanism (Lin et al., 2013a,b) and our recent report that the STF-TPL complex directly represses AS2 at the leaf margin to mediate blade outgrowth in M. truncatula (Zhang et al., 2014). Conservation and diversification of STF/WOX1 and LFL/ WOX3 functions In maize, duplicate PRS/WOX3 orthologs, NS1 and NS2, redundantly function in leaf blade lateral outgrowth (Nardmann et al., 2004). This is essentially similar in rice where two duplicated genes orthologous to NS, NAL2 and NAL3 together regulate leaf blade width (Cho et al., 2013; Ishiwata et al., 2013). The nal2 nal3 double mutant produces not only narrow leaves but also fewer lateral roots, opened



(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Figure 7. LFL and STF are interchangeable in regulating flower and leaf development. (a–d) Leaf and flower phenotypes of stf mutant plants complemented with pSTF::LFL construct. Bars = 1 mm. (e–h) Floral organ phenotypes of lfl mutant plants complemented with pLFL::STF construct. Bars = 1 mm.

spikelets, and thin grains (Cho et al., 2013). In both maize and rice, the double mutants lost most of the leaf blade area in the lateral compartment indicating that NS and its orthologs are major regulators of blade outgrowth in the medial–lateral axis. This is conceivable because the other

blade width regulator WOX gene, WOX1/MAW/STF/LAM1, appears to be missing in monocots (Vandenbussche et al., 2009; Tadege et al., 2011a,b; Nardmann and Werr, 2013; Zhang et al., 2014). In dicots, the WOX1 orthologs MAW in Petunia, STF in M. truncatula, LAM1 in N. sylvestris, and LATH in pea regulate leaf blade outgrowth and floral organ fusion as demonstrated by single mutant phenotypes (Vandenbussche et al., 2009; Tadege et al., 2011a; Zhuang et al., 2012). But, in Arabidopsis, WOX1 functions redundantly with PRS in regulating blade outgrowth (Vandenbussche et al., 2009; Nakata et al., 2012). WOX1 and PRS have overlapping expression pattern in leaf primordia (Nakata et al., 2012), which may explain their redundant function in Arabidopsis leaf blade outgrowth, whereas STF and LFL have mostly non-overlapping expression domains in M. truncatula (Figures 2, 3 and S8) accounting to their functional specialization. Though the WOX1 and PRS functional redundancy in Arabidopsis is intriguing, it may shed light on how the repressive WUS clade WOX genes function with a conserved mechanism (Lin et al., 2013a,b). Nevertheless, in addition to its redundancy with WOX1, Arabidopsis PRS also shows some functional differences from LFL. In the lfl mutant, the number of lateral organs is intact but fusion of sepals and petals, as well as the positioning of the carpel is affected, whereas in the prs mutant, stipules, lateral sepals and lateral stamens are missing (Matsumoto and Okada, 2001; Nardmann et al., 2004), but carpel development is intact. WOX1 function, on the other hand, is highly conserved albeit some variation in mutant phenotype exists. All the mutants described so far, maw, stf, lam1, and lath display dramatic phenotypes in leaf blade, petal and carpel development with no apparent defect in stamens, resulting in female sterile plants in stf, lam1, lath, and reduced fertility in maw (Vandenbussche et al., 2009; Tadege et al., 2011a; Zhuang et al., 2012). lfl/wox3 mutants have not been described in dicots other than Arabidopsis prs and it is unclear whether WOX3 in other dicots functions similar to LFL, PRS or the monocot counter parts. The presence of similar leaf blade phenotypes of wox1 mutants mentioned above in the simple and compound leaf Leguminosae and Solanaceae families prompted us to suggest that the WOX1 subclade may be the major regulator of lateral growth in plant lateral organs in dicots, whereas this function is orchestrated by the WOX3 subclade in monocots. In this scenario, the WOX1 and PRS redundancy may represent an exception restricted to Arabidopsis or a few dicots due to species-specific changes. Analysis of WOX3 function in other dicot species will be required to confirm this hypothesis. However, functional redundancy among WOX genes is not uncommon. For example, the appearance of a single narrow cotyledon without shoot apical meristem in wox1 2 3 5 quadruple mutant in Arabidopsis suggests that PRS and WOX1 have redundant functions with WOX2 and WOX5 in shoot pat-



(a)

(b)

(c)

Figure 8. LFL represses MtAS2 when ectopically expressed under STF promoter. (a) Relative expression level of MtAS2 by qRT-PCR in R108, lfl mutant, stf and stf complemented with pSTF::LFL. The expression level of MtAS2 in the R108 was set to 1.0. Error bars indicate standard deviation (SD) (n = 3). **P < 0.01 (t-test). (b) Schematic representation of reporter and effector constructs used in luciferase assay. The 3-kb MtAS2 promoter region was fused to a mini 35S promoter to drive the expression of luciferase reporter gene. (c) Relative luciferase activity in Arabidopsis protoplasts using LFL as effector compared with GUS control. Error bars indicate SD (n = 3). **P < 0.01 (t-test).

terning during embryo development (Breuninger et al., 2008). Thus, it is likely that STF and LFL in M. truncatula may also have other functions in different developmental stages redundant with other WOX genes. In this respect, LFL expression was detected at high levels in the lateral region of the meristem at the primordial initiation site and in the anthers but the lfl mutant showed no leaf initiation or obvious anther phenotypes, suggesting that it may be redundant in these tissues with other WOX genes. Our data provide a mechanistic insight into the function of LFL/ WOX3 in dicots that was not apparent from analysis of the prs single or prs wox1 double mutants in Arabidopsis. Evolutionary significance of the STF and LFL dichotomy in leaf blade development Because duplicate NS genes are known to regulate leaf blade development in maize, and PRS and WOX1 are redundant in blade outgrowth function in Arabidopsis, it might have been assumed by some that PRS and WOX1 are orthologs of the maize NS genes. Our data show that the PRS homolog, LFL, and the WOX1 homolog, STF, have distinct functions in M. truncatula. Unlike the NS or NAL genes of maize and rice, the STF/WOX1 and LFL/WOX3 genes are not duplicates. There is significant difference in both the gene structure and coding sequence that these two genes are phylogenetically distinct (Vandenbussche et al., 2009; Tadege et al., 2011a,b; Nardmann and Werr, 2013). It is likely that STF and LFL may not have even the same origin or evolved from each other (Nardmann and Werr, 2013). STF/WOX1 first appeared in the basal angiosperm A. trichopoda and maintained in the dicot but lost in the monocot lineage (Vandenbussche et al., 2009; Tadege et al., 2011a; Nardmann and Werr, 2013; Zhang et al., 2014), while LFL/ WOX3 is believed to have orthologs in the last common ancestor (LCA) of seed plants (Nardmann and Werr, 2013).

LFL/WOX3 is preserved in gymnosperms, basal angiosperms, monocots and dicots, and appeared to have undergone gene amplification in monocots. As monocots and dicots have different leaf structures, especially in the mediolateral axis and venation pattern, it is tempting to speculate that the STF/WOX1 and NS/WOX3 homologs might have contributed to the dicot/monocot evolutionary dichotomy by playing key roles in the design of leaf blade and floral organ architecture. In such a scenario, the LFL/WOX3 lineage in dicots would have limited function in the fusion of floral organs with occasional participation in leaf blade outgrowth as seen in Arabidopsis. It would be interesting to investigate if the activity of LFL/WOX3 homologs account for the difference in floral morphologies between archichlamydeae (dicots with unfused petals) and sympetalae (dicots with fused petals). Identifying the evolutionary origins of LFL and STF may help understand their contribution to the developmental evolution of monocots and dicots. EXPERIMENTAL PROCEDURES Mutant screening and cloning of LFL Insertional mutagenesis in M. truncatula genotype R108 using Tnt1 retrotransposon and screening conditions in the greenhouse have been previously described (Tadege et al., 2008; Yarce et al., 2013). Forward genetics screening of Tnt1-tagged lines under standard conditions (16 h/8 h and 24°C/20°C day/night cycles) in the greenhouse for flower mutants have identified two mutants with identical phenotypes of loose-flower designated lfl-1 and lfl-2. To identify the gene linked to the lfl phenotype, Tnt1 flanking sequences of lfl-1 mutant were recovered using TAIL-PCR (Tadege et al., 2008) and genotyped by PCR using FST-specific primers (Table S2). One of the 15 FSTs that segregated with the phenotype as homozygous insert was analyzed by BLAST search against the M. truncatula genome at the National Center for Biotechnology Information (NCBI), http://www.ncbi.nlm.nih.gov/ to obtain the full-length sequence. PCR and RT-PCR was carried out to amplify


LFL functions in floral organ fusion 491 the LFL genomic and CDS sequences, respectively. For quantitative RT-PCR, RNA extraction, reverse transcription and real-time PCR analysis were performed as previously described (Zhang et al., 2014). The FST corresponding to LFL in lfl-2 mutant was PCR amplified using a combination of Tnt1 specific primer LTR6 and LFL gene-specific primer. All primers used in this experiment are given in Table S2. The PCR products were sequenced and the sequences were aligned with the LFL genomic DNA sequence to determine the Tnt1 insertion sites.

Transgene construct and plant transformation To make complementation constructs, the LFL promoter containing a 2645-bp region immediately upstream of the translation start was cloned into the pMDC32 Gateway vector substituting 2 9 35S promoter to generate the pMDC32-pLFL destination vector by using pLFL-F and pLFL-R primers (Table S2). The LFL CDS sequences were amplified from M. truncatula R108 cDNA by using appropriate primers. The mutations in LFL (LFLm and LFLt) were introduced by PCR mediated mutagenesis using LFLm-R and LFLtR primers, respectively. SRDX-LFLm and LFLm-VP16 were constructed using synthetic sense and antisense primers with SRDX and VP16 sequences. All regions corresponding to the transgene were cloned into the pMDC32-pLFL destination vector by using the Gateway system (Invitrogen, http://www.lifetechnologies.com/). The 2645-bp LFL promoter was cloned into the pMDC162 vector by using pLFL-attB1 and pLFL-attB2 to generate the pLFL::GUS construct. Constructs were introduced into Agrobacterium tumefaciens by electroporation. A. tumefaciens strain AGL1 was used for M. truncatula transformation and strain GV2260 was used for N. sylvestris transformation as described (Tadege et al., 2011a).

Subcellular localization analysis The coding sequence of LFL was cloned into the pMDC83 Gateway vector to generate the pMDC83-LFL-GFP destination vector. The construct was introduced into A. tumefaciens by electroporation and the Agrobacterium was infiltrated into 4-week-old N. benthamiana leaves. P19 was used to inhibit transgenic silencing. GFP signal was observed 48–60 h after infiltration by a Leica TCS SP2 AOBS confocal laser scanning microscope (Leica Microsystems, http://www.leica-microsystems.com/).

In situ hybridization A 508-bp fragment of LFL CDS was amplified by PCR. The PCR product was labeled with digoxigenin (digoxigenin-11-UTP, Roche Diagnosis, http://www.roche.com/). RNA in situ hybridization was performed on shoot apices 4-week-old R108 plants or inflorescence as previously described (Zhang et al., 2014).

Histochemical GUS staining GUS staining assay was performed as described (Zhao et al., 2001) and images of GUS staining patterns of tissues were collected with a digital camera mounted on an Olympus SZX-16 Stereoscope (Hitschfel Instruments, http://www.hitschfel.com/).

Transient luciferase expression assay Construction of the reporter GAL4-LUC plasmid was described previously (Ikeda et al., 2009). For effectors plasmids, the coding sequences of LFL, LFLm, and LFLt were first cloned into pGBKT7. Then the coding regions of BD fusion were amplified using specific primers and cloned into p2GW7 using the Gateway system (Invitrogen) to yield effector plasmids. The coding region of LFL was ampli-

fied and cloned into p2GW7. The GW7-GUS effector and the proMtAS2-mini 35S reporter were described previously (Zhang et al., 2014).Transient expression assay was performed with Arabidopsis protoplasts according to Asai et al. (Asai et al., 2002). For each transformation 5 lg of reporter plasmid and 4 lg of effector plasmid were used. For normalization of the activity of the reporter gene, 0.5 lg of plasmid pRLC (Wang et al., 2010) was used as internal control.

Yeast two-hybrid and BiFC assay The bait and prey clones used in yeast two-hybrid (Y2H) assays were cloned into the Gateway (Invitrogen) version of pGBKT7-GW (bait) and pGADT7-GW (prey) vectors using specific primers (Table S2). Sets of constructs were co-transformed into Y2H gold yeast strain (Clontech, http://www.clontech.com/). Yeast transformants were selected on synthetic minimal dropout medium (DDO) deficient in tryptophan and leucine. Protein interaction tests were assessed on quadruple dropout (QDO) medium deficient in histidine, tryptophan, leucine and adenine in the presence of a color indicator, X-a-gal. BiFC assays were conducted as described (Lu et al., 2010). Briefly, LFL and LFLm were cloned to pEARLEYGATE201-YN while Mt-TPL was cloned to pEARLEYGATE202-YC by the LR reaction. LFL-YN, LFLm-YN, and Mt-TPL-YC were introduced into A. tumefaciens strain GV2260. Sets of combination were co-infiltrated to 4-week-old N. benthamiana leaves. P19 was used to inhibit transgenic silencing. YFP signal was observed 48–60 h after infiltration by a Leica TCS SP2 AOBS confocal laser scanning microscope (Leica Microsystems, http://www.leica-microsystems.com/).

Sequence alignment and phylogenetic analysis Amino acid sequences of WOX family proteins were aligned using ClustalW, and a neighbor-joining phylogenetic tree was constructed using MEGA 4 software. The most parsimonious tree with bootstrap values from 1000 trials was shown.

Accession numbers Sequence data used in this study can be found in the GenBank database under the following accession numbers: LFL (XM_003623010), Mt-TPL (KC525957), STF (JF276252), MtTPR1 (Medtr2g104140.1), MtTPR2 (Medtr3g043840.1), MtTPR3 (Medtr1g083700.1), MtTPR4 (Medtr7g112460.1), WUS (At2g17950), WOX1 (At3g18010), WOX2 (At5g59340), WOX3/PRS (At2g28610), WOX4 (At1g46480), WOX5 (At3g11260), WOX6/PFS2 (At2g01500), WOX7 (At5g05770), NS1 (NM_001111690), NS2 (NM_001111772), NAL2/3 (AB218893), LATH (JQ291249), NAO1 (JQ291250) and MAW (EU359004).

ACKNOWLEDGEMENTS We thank Drs. Elison Blancaflor and Jin Nakashima for their help with confocal microscopy and Yuhai Cui for providing Gatewaycompatible Y2H and BiFC vectors. We also thank the National Science Foundation equipment grant (DBI 0400580) for the confocal microscope. This work was supported by the National Science Foundation Grant IOS-1354422 (to M.T.) and the Chinese Academy of Agricultural Sciences Elite Youth Program (to L.N. and H.L.).

COMPETING INTERESTS The authors declare no competing financial interests. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article.


492 Lifang Niu et al. Table S1. Phenotype analysis of transgenic plants. Table S2. Primers used in this study. Figure S1. Comparison of fusion of sepals in wild-type R108 and lfl mutant. Figure S2. Transcript abundance of LFL in transgenic lfl plants. Figure S3. Phylogenetic analysis and sequence alignment of LFL and WUS clade WOX proteins in diverse species. Figure S4. The LFL protein is mainly localized in the nucleus. Figure S5. LFL Interacts with Mt-TPL family proteins. Figure S6. Carpel phenotypes of the stf mutant. Figure S7. Complementation of the lam1 mutant with Medicago LFL gene. Figure S8. Comparison of LFL and STF expression patterns.

REFERENCES Asai, T., Tena, G., Plotnikova, J., Willmann, M.R., Chiu, W.L., Gomez-Gomez, L., Boller, T., Ausubel, F.M. and Sheen, J. (2002) MAP kinase signalling cascade in Arabidopsis innate immunity. Nature, 415, 977–983. ~ as, Benlloch, R., d’Erfurth, I., Ferrandiz, C., Cosson, V., Beltran, J.P., Can ~ o, F. and Ratet, P. (2006) Isolation of mtpim L.A., Kondorosi, A., Maduen proves Tnt1 a useful reverse genetics tool in Medicago truncatula and uncovers new aspects of AP1-like functions in legumes. Plant Physiol. 142, 972–983. Breuninger, H., Rikirsch, E., Hermann, M., Ueda, M. and Laux, T. (2008) Differential expression of WOX genes mediates apical-basal axis formation in the Arabidopsis embryo. Dev. Cell, 14, 867–876. Cho, S.-H., Yoo, S.-C., Zhang, H., Pandeya, D., Koh, H.-J., Hwang, J.-Y., Kim, G.-T. and Paek, N.-C. (2013) The rice narrow leaf2 and narrow leaf3 loci encode WUSCHEL-related homeobox 3A (OsWOX3A) and function in leaf, spikelet, tiller and lateral root development. New Phytol. 198, 1071–1084. Dong, Z.C., Zhao, Z., Liu, C.W., Luo, J.H., Yang, J., Huang, W.H., Hu, X.H., Wang, T.L. and Luo, D. (2005) Floral patterning in Lotus japonicus. Plant Physiol. 137, 1272–1282. Graham, P. and Vance, C.P. (2003) Legumes: importance and constraints to greater use. Plant Physiol. 131, 872–877. Ikeda, M., Mitsuda, N. and Ohme-Takagi, M. (2009) Arabidopsis WUSCHEL is a bifunctional transcription factor that acts as a repressor in stem cell regulation and as an activator in floral patterning. Plant Cell, 21, 3493–3505. Ishiwata, A., Ozawa, M., Nagasaki, H. et al. (2013) Two WUSCHEL-related homeobox genes, narrow leaf2 and narrow leaf3, control leaf width in rice. Plant Cell Physiol. 54, 779–792. Ji, J., Strable, J., Shimizu, R., Koenig, D., Sinha, N. and Scanlon, M.J. (2010) WOX4 promotes procambial development. Plant Physiol. 152, 1346– 1356. Lin, H., Niu, L., McHale, N.A., Ohme-Takagi, M., Mysore, K.S. and Tadege, M. (2013a) Evolutionarily conserved repressive activity of WOX proteins mediates leaf blade outgrowth and floral organ development in plants. Proc. Natl Acad. Sci. USA 110, 366–371. Lin, H., Niu, L. and Tadege, M. (2013b). STENOFOLIA acts as a repressor in regulating leaf blade outgrowth. Plant Signal Behav. 8, pii: e24464. Lu, Q., Tang, X., Tian, G. et al. (2010) Arabidopsis homolog of the yeast TREX-2 mRNA export complex: components and anchoring nucleoporin. Plant J. 61, 259–270. Matsumoto, N. and Okada, K. (2001) A homeobox gene, PRESSED FLOWER, regulates lateral axis-dependent development of Arabidopsis flowers. Genes Dev. 15, 3355–3364.

McHale, N.A. (1992) A nuclear mutation blocking initiation of the lamina in leaves of Nicotiana sylvestris. Planta, 186, 355–360. Nakata, M., Matsumoto, N., Tsugeki, R., Rikirsch, E., Laux, T. and Okada, K. (2012) Roles of the middle domain-specific WUSCHEL-RELATED HOMEOBOX genes in early development of leaves in Arabidopsis. Plant Cell, 24, 519–535. Nardmann, J. and Werr, W. (2013) Symplesiomorphies in the WUSCHEL clade suggest that the last common ancestor of seed plants contained at least four independent stem cell niches. New Phytol. 199, 1081–1092. Nardmann, J., Ji, J., Werr, W. and Scanlon, M.J. (2004) The maize duplicate genes narrow sheath1 and narrow sheath2 encode a conserved homeobox gene function in a lateral domain of shoot apical meristems. Development, 131, 2827–2839. Sarkar, A.K., Luijten, M., Miyashima, S., Lenhard, M., Hashimoto, T., Nakajima, K., Scheres, B., Heidstra, R. and Laux, T. (2007) Conserved factors regulate signalling in Arabidopsis thaliana shoot and root stem cell organizers. Nature, 446, 811–814. Scanlon, M.J., Schneeberger, R.G. and Freeling, M. (1996) The maize mutant narrow sheath fails to establish leaf margin identity in a meristematic domain. Development, 122, 1683–1691. Shimizu, R., Ji, J., Kelsey, E., Ohtsu, K., Schnable, P.S. and Scanlon, M.J. (2009) Tissue specificity and evolution of meristematic WOX3 function. Plant Physiol. 149, 841–850. Tadege, M., Wen, J., He, J. et al. (2008) Large-scale insertional mutagenesis using the Tnt1 retrotransposon in the model legume Medicago truncatula. Plant J. 54, 335–347. Tadege, M., Wang, T.L., Wen, J., Ratet, P. and Mysore, K.S. (2009) Mutagenesis and beyond! Tools for understanding legume biology. Plant Physiol. 151, 978–984. Tadege, M., Lin, H., Bedair, M. et al. (2011a) STENOFOLIA regulates blade outgrowth and leaf vascular patterning in Medicago truncatula and Nicotiana sylvestris. Plant Cell, 23, 2125–2142. Tadege, M., Lin, H., Niu, L. and Mysore, K.S. (2011b) Control of dicot leaf blade expansion by a WOX gene, STF. Plant Signal Behav. 6, 1861–1864. Tucker, S. (2003) Floral development in legumes. Plant Physiol. 131, 911– 926. Vandenbussche, M., Horstman, A., Zethof, J., Koes, R., Rijpkema, A.S. and Gerats, T. (2009) Differential recruitment of WOX transcription factors for lateral development and organ fusion in Petunia and Arabidopsis. Plant Cell, 21, 2269–2283. Wang, H., Avci, U., Nakashima, J., Hahn, M.G., Chen, F. and Dixon, R.A. (2010) Mutation of WRKY transcription factors initiates pith secondary wall formation and increases stem biomass in dicotyledonous plants. Proc. Natl Acad. Sci. USA 107, 22338–22343. Yarce, J.C., Lee, H.K., Tadege, M., Ratet, P. and Mysore, K.S. (2013) Forward genetics screening of Medicago truncatula Tnt1 insertion lines. Methods Mol. Biol. 1069, 93–100. Zhang, Y., Wu, R., Qin, G., Chen, Z., Gu, H. and Qu, L.J. (2011) Over-expression of WOX1 leads to defects in meristem development and polyamine homeostasis in Arabidopsis. J. Integr. Plant Biol. 53, 493–506. Zhang, F., Wang, Y., Li, G., Tang, Y., Kramer, E.M. and Tadege, M. (2014) STENOFOLIA recruits TOPLESS to repress ASYMMETRIC LEAVES2 at the leaf margin and promote leaf blade outgrowth in Medicago truncatula. Plant Cell, 26, 650–664. Zhao, Y., Christensen, S.K., Fankhauser, C., Cashman, J.R., Cohen, J.D., Weigel, D. and Chory, J. (2001) A role for flavin monooxygenase-like enzymes in auxin biosynthesis. Science, 291, 306–309. Zhuang, L.-L., Ambrose, M., Rameau, C., Weng, L., Yang, J., Hu, X.-H., Luo, D. and Li, X. (2012) LATHYROIDES, encoding a WUSCHEL-Related Homeobox1 transcription factor, controls organ lateral growth, and regulates tendril and dorsal petal identities in garden pea (Pisum sativum L.). Mol. Plant, 5, 1333–1345.
