Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
•The CUP-SHAPED COTYLEDON (CUC)/NO APICAL MERISTEM (NAM) family of genes control boundary formation and lateral organ separation, which is critical for proper leaf and flower patterning. However, most downstream targets of CUC/NAM genes remain unclear.
•In a forward screen of the tobacco retrotransposon1 (Tnt1) insertion population in Medicago truncatula, we isolated a weak allele of the no-apical-meristem mutant mtnam-2. Meanwhile, we regenerated a mature plant from the null allele mtnam-1. These materials allowed us to extensively characterize the function of MtNAM and its downstream genes.
•MtNAM is highly expressed in vegetative shoot buds and inflorescence apices, specifically at boundaries between the shoot apical meristem and leaf/flower primordia. Mature plants of the regenerated null allele and the weak allele display remarkable floral phenotypes: floral whorls and organ numbers are reduced and the floral organ identity is compromised. Microarray and quantitative RT-PCR analyses revealed that all classes of floral homeotic genes are down-regulated in mtnam mutants. Mutations in MtNAM also lead to fused cotyledons and leaflets of the compound leaf as well as a defective shoot apical meristem.
•Our results revealed that MtNAM shares the role of CUC/NAM family genes in lateral organ separation and compound leaf development, and is also required for floral organ identity and development.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Boundaries are composed of distinctive sets of cells that are formed between the shoot apical meristem (SAM) and lateral organs or between lateral organs during embryogenesis and post-embryonic development. Boundaries separate lateral organ primordia from the SAM and adjacent organ primordia, and promote SAM and organ primordium formation (Aida & Tasaka, 2006a,b; Rast & Simon, 2008). Boundary cells have specific morphological and cytological characteristics. Cells at the meristem and organ boundaries display a saddle-shaped surface and are elongated along the boundaries (Kwiatkowska, 2004, 2006; Reddy et al., 2004). Cell proliferation analysis reveals that boundaries between inflorescence and floral meristems, and between floral organs, consist of nondividing cells.
CUP-SHAPED COTYLEDON (CUC)/NO APICAL MERISTEM (NAM), a small group of plant-specific NAC transcription factors, plays important roles in the regulation of boundary development (Aida & Tasaka, 2006a; Rast & Simon, 2008). The CUC/NAM gene family includes CUP-SHAPED COTYLEDON1, 2 and 3 (CUC1, 2 and 3) in Arabidopsis thaliana, NO APICAL MERISTEM (NAM) in Petunia, GOBLET (GOB) in tomato (Solanum lycopersicum) and CUPULIFORMIS (CUP) in Antirrhinum majus (Souer et al., 1996; Aida et al., 1997; Takada et al., 2001; Vroemen et al., 2003; Weir et al., 2004; Berger et al., 2009). CUC/NAM genes are expressed in all boundaries between organ primordia and meristems from early embryogenesis to floral developmental stages. Mutations in CUC1/CUC2 of A. thaliana and in NAM of Petunia lead to fusion of cotyledons and some floral organs, as well as severe defects of the primary apical meristem (Souer et al., 1996; Aida et al., 1997; Takada et al., 2001). Double mutants cuc3/cuc2 or cuc3/cuc1 in A. thaliana and the co-suppression plants of NAM and its homolog genes NAM HOMOLOG-1 (NH-1) and NAM HOMOLOG-3 (NH-3) in Petunia exhibit lateral organ fusion during vegetative development (Souer et al., 1998; Hibara et al., 2006). Mutations in the CUP gene in snapdragon (Antirrhinum majus), however, strongly affect all lateral organ boundaries during embryonic, vegetative and reproductive development (Weir et al., 2004). These studies not only demonstrate that CUC/NAM genes play a pivotal role in organ separation and primary apical meristem formation but also show that different degrees of functional redundancy exist amongst the CUC/NAM family members.
Plant leaf primordia initiate from the flanks of the SAM and go through primary and secondary morphogenesis to form various leaf patterns, such as simple or compound leaves. Of particular importance for leaf patterning is the region at the leaf margins, the marginal blastozone or leaf marginal meristems, which maintains a morphogenetic activity and is responsible for the initiation of secondary structures, such as leaflets (Blein et al., 2010; Efroni et al., 2010). In compound-leafed species, both class I homeodomain KNOTTED1-like genes (KNOXI), which were initially identified for their roles in the maintenance of shoot meristem identity (Long & Barton, 1998), and the floral meristem identity gene LEAFY (LFY) and its orthologs, UNIFOLIATA (UNI ) in pea (Pisum sativum) and SINGLE LEAFLET1 (SGL1) in Medicago truncatula, are species-specific positive regulators in compound leaf development (Hofer et al., 1997; Champagne et al., 2007; Wang et al., 2008; Blein et al., 2010). Recent studies revealed that CUC/NAM genes also share a conserved function in compound leaf development and leaf margin formation (Nikovics et al., 2006; Blein et al., 2008, 2010; Bilsborough et al., 2011; Hasson et al., 2011). Reduced expression of NAM/CUC leads to the suppression of marginal outgrowth and thus formation of reduced and/or fused leaflets during compound leaf development in a diverse compound-leafed species (Blein et al., 2008). GOB, a NAM ortholog in tomato, is essential for proper specification of lateral organ boundaries at the apical meristem and proper specification of leaflet boundaries in developing compound leaves (Berger et al., 2009). Furthermore, it has been reported that the ectopic expression of CUC1 in the margins of developing leaves is sufficient to change their architecture from simple to compound in A. thaliana (Hasson et al., 2011). It is thus proposed that NAM/CUC genes have a common role in promoting leaflet formation and separation.
Previous studies suggest that CUC/NAM genes prevent organ fusion through repression of boundary cell growth and that the possible cytological function of these genes is to regulate cell division or orientation as well as cell expansion (Aida & Tasaka, 2006a,b; Rast & Simon, 2008). CUP in snapdragon directly interacts with a TCP-domain transcription factor, which has previously been shown to regulate organ outgrowth (Weir et al., 2004). Over the last decade, several regulators of CUC genes have been identified in A. thaliana, including SHOOTMERISTEMLESS (STM), PINFORMED1 (PIN1), MONOPTEROS (MP), and microRNA164 (Aida et al., 1999, 2002; Mallory et al., 2004; Aida & Tasaka, 2006a; Larue et al., 2009). STM plays a major role in SAM initiation and is also implicated in cotyledon separation. Activation of STM expression in the embryo apical end requires CUC1 and CUC2, whereas at later stages of A. thaliana embryogenesis STM is required for proper expression patterns of CUC1 and CUC2 (Aida et al., 1999; Hibara et al., 2003). Further studies showed that STM directly binds to the promoter of CUC1 and thus up-regulates CUC1 expression (Spinelli et al., 2011). PIN1 and MP repress CUC1 expression in cotyledons and promote CUC2 expression in cotyledon boundaries (Aida et al., 2002; Furutani et al., 2004). During leaf margin development, CUC2 promotes the generation of PIN1-dependent auxin accumulation while auxin represses CUC2 expression. This feedback loop regulates the activity of the conserved auxin efflux module in leaf margins to form stable serration patterns (Bilsborough et al., 2011). Both CUC1/2 and GOB are post-transcriptionally regulated by microRNA164 for fine-tuning of organ boundaries both temporally and spatially, especially in leaf margin development (Laufs et al., 2004; Mallory et al., 2004; Nikovics et al., 2006; Larue et al., 2009). These genes are negative or reciprocal regulators of CUC/NAM. More recently, it was reported that CUC1 directly activates the expression of LIGHT-DEPENDENT SHORT HYPOCOTYLS (LSH) (LSH3 and LSH4) in shoot organ boundaries (Takeda et al., 2011). Despite these observations, how CUC/NAM regulates downstream gene(s) to affect lateral organ development remains elusive.
In this study, two insertion mutant alleles of MtNAM, one null allele with a retrotransposon Tnt1 insertion, and one weak allele with a native Medicago Endogenous Retrotansposon 1 (MERE1) insertion, were characterized in detail in M. truncatula. The mtnam mutants display unique simplified floral organ phenotypes in addition to the common fused-cotyledon and leaflet phenotypes shared with other cuc/nam mutants. Microarray and real-time quantitative PCR analyses revealed that mutations in MtNAM down-regulate the expression of floral homeotic genes. MtNAM is expressed at boundaries between lateral organs/organ primordia and meristems. These results indicate that MtNAM plays an essential role in controlling floral organ formation and lateral organ separation in M. truncatula.
Materials and Methods
Seed treatment and plant growth
Seeds of wild-type Medicago truncatula Gaertn. R108 and mutant lines were scarified with concentrated sulfuric acid for 8 min and thoroughly rinsed with water, which was followed by sterilization in 30% bleach for 10 min, extensive rinsing with ddH2O, cold treatment at 4°C for 7 d on MS medium, and germination in a growth chamber with a regime of a 18 h light (25°C) : 6 h dark (22°C) photoperiod. After 2 wk, seedlings were transferred into Metro-Mix 350 (Scotts, Marysville, OH, USA) composite soil and grown in a glasshouse until maturation.
Tnt1 insertion mutant screening and molecular confirmation
The M. truncatula Tnt1 insertional mutant population was generated as described previously (Tadege et al., 2008). In a forward genetic screen for mutants with fused cotyledons or fused leaflets, we identified mutant lines NF1937 (mtnam-1) and NF1757 (mtnam-2). Recovery of flanking sequence tags (FSTs) in the Tnt1 insertion line NF1937 was carried out by thermal asymmetric interlaced (TAIL)-PCR as described previously (Cheng et al., 2011). Each individual FST was analyzed by tBLASTx against the M.truncatula genome sequence at the National Center for Biotechnology Information (http://blast.ncbi.nlm.nih.gov/). Candidate FSTs were selected for co-segregation analysis.
Segregated R1 seeds from NF1937 were treated and grown as described in the previous section. Homozygous plants were identified by their fused cotyledons, and heterozygous plants were identified by examining dissected maturing pods, which in heterozygous plants produced seeds with both fused cotyledons and normal cotyledons. Genomic DNA from both homozygous and heterozygous plants was extracted. Forward primer Tnt1-F and reverse primers from selected candidate FSTs were used for genotyping the NF1937 progenies.
PCR-based reverse screening for MtNAM mutant lines and molecular cloning of MtNAM
One c. 500-bp FST was found co-segregating with the fused cotyledon phenotype in NF1937 progenies. Part of this FST hits NAM/CUC family genes in Petunia hybrid and Arabidopsis thaliana but no M. truncatula genomic sequence matched this FST. To recover the full genomic sequence of the gene, two sets of primers, MtNAM-F, MtNAM-F1, MtNAM-R and MtNAM-R1 (sequences are listed in Table 1), were designed on both ends of the FST sequence. TAIL-PCR was carried out using wild-type R108 DNA as the template (Liu et al., 1995). After two rounds of PCR amplification, the full-length genomic DNA sequence of MtNAM was obtained. To obtain more mutant alleles of mtnam, PCR-based reverse genetic screening was performed as described previously using the pooled DNA from the Tnt1 mutant population (Tadege et al., 2008; Cheng et al., 2011).
Table 1. Primer sequences of genes used in the experiments
MtNAM, Medicago truncatula NO APICAL MERISTEM; SGL, SINGLE LEAFLET.
Regeneration of mtnam-1 plants
Immature seeds with fused cotyledons at the late cotyledon stage were collected from MtNAM/mtnam-1 heterozygous plants, sterilized for 15 min with 30% bleach + 0.01% Tween-20, and rinsed three times with autoclaved ddH2O. Immature embryos with fused cotyledons (homozygous mtnam-1) were cut into two halves and placed on the callus-induction medium. Callus subculture, somatic embryogenesis and plantlet regeneration were carried out following previous protocols (Trinh et al., 1998). Regenerated plants were transferred into soil and grown to maturation in the glasshouse.
In situ hybridization
Young seeds and shoot apices, at both vegetative and reproductive stages, were fixed and in situ hybridization was performed as described by Zhou et al. (2010). A 638-bp cDNA fragment from the nonconserved region of MtNAM was used as the probe for hybridization.
Flowers and seeds dissected from young pods 1–2 d after pollination were cleared in Hoyer’s solution (7.5 g of gum arabic, 100 g of chloral hydrate, 5 ml of glycerol and 60 ml of ddH2O). Ovules and embryos were further dissected and observed under a dissecting microscope.
Scanning electron microscopy
Inflorescence shoot apices were dissected and fixed in 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS; pH 7.2) for 12 h at 4°C. After rinsing with PBS for 3 h, samples were dehydrated in an ethanol series and critical-point dried. The samples were then mounted on metal stubs, sputter-coated with gold and observed under a Zeiss DSM-960A SEM (Carl Zeiss Inc.) at an accelerating voltage of 5 kV.
RNA extraction, RT-PCR and qRT-PCR
Roots, stems, leaves, vegetative shoots, inflorescence shoots, flower buds, young pods and young seeds were collected from wild-type R108 plants. Vegetative shoots and inflorescence shoots were also collected from both heterozygous and homozygous plants of mtnam-1 and mtnam-2. Total RNA was extracted using Tri-Reagent (Gibco-BRL Life Technologies, Grand Island, NY, USA) and treated with Turbo DNase I (Ambion). For RT-PCR and qRT-PCR, 3 μg of total RNA was used for reverse transcription using SuperScript III Reverse Transcriptase (Invitrogen) with the oligo (dT)20 primer. Two microliters of 1 : 20 diluted cDNA was used as the template. Gene-specific primers for RT-PCR and qRT-PCR are listed in Table 1. All qRT-PCR reactions were carried out using a 7900HT Fast Real-Time PCR System (Applied Biosystems), and the data were analyzed using sds 2.2.1 software (Applied Biosystems, Life Technologies, Grand Island, NY, USA). PCR efficiency (E) was estimated using the LinRegPCR program (Ramakers et al., 2003) and the transcript levels were determined by relative quantification using the M. truncatula actin gene (tentative consensus no. 107326) as the reference (Benedito et al., 2008).
For microarray analysis, total RNA was extracted from three biological replicates of the inflorescence shoots from both wild-type-looking plants (including wild-type and heterozygous plants from the same NF1937 segregating progeny) and regenerated homozygous mtnam-1 plants using the RNeasy Plant Mini Kit (Qiagen, Valencia, CA). Purified RNA was treated with Turbo DNase I (Ambion). 10 μg of total RNA from each sample was used for probe labeling. Hybridization against Affymetrix Medicago Genechip and scanning for microarray analysis were conducted according to the manufacturer’s instructions (Affymetrix, Santa Clara, CA, USA). Differentially expressed genes between wild-type-looking plants and homozygous mtnam-1 were selected using associative analysis as described previously (Benedito et al., 2008). Raw data from the experiment were deposited in the ArrayExpress database under the ArrayExpress accession number E-MEXP-3480.
The MtNAM gene sequence was deposited in GenBank under the accession number JF929904.
The fused cotyledon phenotype is caused by mutation of MtNAM in Medicago truncatula
Medicago truncatula mutant line NF1937 with fused cotyledons (Fig. 1a.2) was identified from a forward screen of the Tnt1 insertion population generated at the Samuel Roberts Noble Foundation (Tadege et al., 2008). Using the TAIL-PCR approach (Liu et al., 1995), 15 FSTs of Tnt1 insertions were recovered from this mutant. After BLAST searches against the NCBI GenBank, five candidate FSTs were chosen for genotyping the segregating F2 population from the self-pollinated heterozygous plants. One FST, which had no significant hits in the M. truncatula genome sequences in GenBank, was found co-segregating with the fused cotyledon phenotype. Using two sets of primers at the 3′ and 5′ ends of this FST sequence, a full-length genomic sequence was recovered by TAIL-PCR. The same two sets of primers were also used in combination with Tnt1-specific primers to screen the pooled DNA samples of the Tnt1 insertion population (Cheng et al., 2011) and two additional insertion alleles were identified. The two alleles, NF2689 and NF4725, exhibit similar fused cotyledon phenotypes as observed in NF1937. Sequence analysis revealed that the causative gene for the fused cotyledon phenotype in NF1937, NF2689 and NF4725 encodes an NAC-domain transcription factor, designated as MtNAM. MtNAM shares > 75% identity in amino acid sequences with other NAM/CUC family members. Phylogenetic analysis indicated that MtNAM belongs to the NAM clade and is closest to PsNAM1 and PsNAM2 in pea (Fig. 1b). NF1937, which was designated as mtnam-1, harbors a Tnt1 insertion at 1325 bp from ATG. NF4725, named mtnam-3, has a Tnt1 insertion at 363 bp, and NF2689, named mtnam-4, has a Tnt1 insertion at 1295 bp from ATG (Fig. 1c).
In addition, another mutant line, NF1757, with partially fused cotyledons and fused leaflets, was obtained from a forward screen of the Tnt1 insertion population. We failed to recover any causative Tnt1 insertions in NF1757 by TAIL-PCR using Tnt1-specific primers. Direct PCR using MtNAM-specific primers revealed an insertion of MERE1, a native retrotransponson of M. truncatula (Rakocevic et al., 2009), at 1811 bp from ATG of MtNAM, 67 bp upstream of the stop codon TAA (Fig. 1c). Although the full-length transcript of MtNAM was not amplified using MtNAM-specific forward and reverse primers (Fig. 1d; lanes 5 and 6), a transcript that is slightly bigger in size was amplified using the MtNAM forward primer and the MERE1 reverse primer in the mutant NF1757 (Fig. 1d; lanes 7 and 8). Sequence analysis of this RT-PCR product suggested that the deduced chimera protein has 375 amino acids, of which 362 aa are from MtNAM and 13 aa from MERE1 (Supporting Information Notes S1). Therefore, the mutant NF1757, designated as mtnam-2, is a weak allele of MtNAM.
MtNAM is required for cotyledon and leaflet separation and primary apical meristem formation
Mutant mtnam-1 seedlings display fused cotyledons and aborted primary apical meristem (Fig. 1a.2), while mtnam-2 seedlings exhibit a fused leaflet phenotype (Fig. 1a.3). We examined the development of these mutants from early embryogenesis to seedling stages in comparison with wild-type plants. Differences in the embryo morphology between mtnam-1 and wild type become apparent at the early heart stage. Unlike the heart-shaped embryos in wild type (Fig. 2a), embryos of mtnam-1 appear cylindrical (Fig. 2b). However, no significant differences were observed at this stage between mtnam-2 and wild-type embryos (data not shown).
At the cotyledon stage, the two cotyledons are separated and the apical meristem between the two cotyledons is visible in the wild type (Fig. 2c,f). In the weak allele mtnam-2, the two cotyledons are partially fused at the basal region and the embryonic meristem appears to develop normally (Fig. 2d,g). The juvenile leaf, which is typically elliptical in wild-type plants, is clearly deformed to be shield-shaped in the mtnam-2 plant (Fig. 1a.3). In contrast to wild-type and mtnam-2 plants, the two cotyledons in mtnam-1 are fused along the edges and no apical meristem was observed (Fig. 2e,h). After germination, wild-type seeds develop into young seedlings with a single juvenile leaf and three-leaflet compound leaves (Fig. 2i,k,n). Because in mtnam-2 the lower parts of the two cotyledons are fused, the mtnam-2 mutant grows into a seedling by breaking the side of the partially fused cotyledons with the developing SAM (Fig. 2i; arrow indicates the breaking point). The juvenile leaf in mtnam-2 is shield-shaped (Fig. 1a) and the three leaflets of compound leaves are fused into a simple leaf (Fig. 2n). Development of the mtnam-1 mutant arrests at the fused cotyledon stage. No primary apical meristem is developed (Fig. 2j,l). Furthermore, no escaped shoot meristem was observed with or without decapitating the hypocotyls (data not shown).
We were interested in understanding how loss-of-function of MtNAM affects development in mature plants. Using fused cotyledons of mtnam-1 as explants, mature mtnam-1 plants were regenerated through tissue culture. The regenerated mtnam-1 plants show nearly normal vegetative growth and no apparent defects in the SAM (Fig. 2n) even though no MtNAM expression was detectable (Fig. 1d). Like plants carrying the weak allele mtnam-2, the regenerated mtnam-1 null mutant also develops compound leaves with fused leaflets, and the fusion of leaflets is more severe than that in mtnam-2. The three major veins in the fused leaflets are distinctive, indicating the fused leaves are developed from three leaflets (Fig. 2n).
Mutation of MtNAM results in a reduced number of floral whorls and floral organs
Both weak allele mtnam-2 plants and regenerated null allele mtnam-1 plants develop into maturation with flowers but no seed pods. In order to uncover the phenotypes in flowers, we compared the flower morphology among mtnam-1, mtnam-2 and wild-type plants. Each flower in the wild type is organized in four whorls: sepals, petals, stamens and the carpel. The five sepals in the first whorl are fused at the base. The second whorl consists of one standard petal, two lateral petals or wings, and two fused short petals. The third whorl contains nine fused filaments and one free-standing filament. The innermost whorl is the centered carpel consisting of stigmas, style and ovary covered with glandular trichomes (Fig. 3a,b). In plants carrying the weak allele mtnam-2, flowers are similar in appearance to the wild-type flowers. Dissection of the mtnam-2 flowers revealed that the flower is also arranged in four whorls; the first and second whorls of sepals and petals are similar to those of the wild type in terms of organ numbers and shapes (Fig. 3c). The stamens in the third whorl and the carpel at the center show some differences from those of wild-type flowers. Some stamens are similar to those of the wild type, with nine fused filaments and one separated filament (Fig. 3c). Some stamens, however, have a reduced number of filaments (usually reduced to five to seven); and the free-standing filament is fused with the petals or carpel (Fig. 3d). Most anthers of mtnam-2 do not release pollen, although a few pollen grains are occasionally released to the outside of the anthers. The released pollen grains are able to germinate in vitro (data not shown). Unlike wild-type carpels which are covered only by glandular trichomes (Fig. 3l), the carpel in mtnam-2 is covered with long hairy trichomes (Fig. 3d,e,m), although glandular trichomes are still visible (Fig. 3e). Along the margin, carpel edges are not fully fused in the upper part of the ovary and some ovules are exposed (Fig. 3e). Occasionally germinating pollen grains are observed on stigmas (Fig. 3p). However, seeds are never developed in mtnam-2 plants. In the regenerated mtnam-1 plants, the flower size is reduced. Dissection of flowers revealed severe defects in whorl number, organ number and organ appearance compared with mtnam-2 and wild-type flowers (Fig. 3f–j). In mtnam-1 flowers, the first whorl (sepals) shows normal development, although sometimes the number of sepals is reduced to three to four (Fig. 3g). The second whorl (petals) and the third whorl (stamens) are usually simplified into one whorl, forming either sepal-like or petal-like structures, and the organ number is greatly reduced to one or two (Fig. 3g–j). Sometimes, anther-like tissues are formed on a petal-like structure, but no pollen is observed (Fig. 3o). At the center of the flower, the carpel is always observed. However, the carpel is small and its edges are not fused, resulting in exposed ovules (Fig. 3k). Some carpels have more severe defects, with no style and stigmas developed (Fig. 3g). Similar to mtnam-2, the carpel of mtnam-1 is also covered with dense long trichomes on the surface (Fig. 3h–j,n).
Inside the wild-type carpel, eight to ten ovules are arranged along the fused margin, and each mature ovule consists of funiculus, chalaza and nucellus containing a large embryo sac (Fig. 3q). In mtnam-2, eight to 10 ovules are also developed along the incompletely fused carpel margin. Both normal ovules with an embryo sac and abnormal ovules without an embryo sac are observed in the same ovary (Fig. 3r,s). Inside the carpel of mtnam-1, only three to six small ovules are observed along the unclosed carpel margin. The development of ovules is severely defective and no embryo sac is observed in the nucellus (Fig. 3t). In extreme cases, some ovules develop into carpel-like structures (Fig. 3u). Both male and female organs are sterile in mtnam-1 flowers.
We further examined the early development of floral organs in mtnam mutants and the wild type by scanning electron microscopy (SEM). Compared with the common radial arrangement of sepals, petals, stamens, and carpels in most angiosperm plant species, the initiation of floral organs in all whorls is unidirectional from the abaxial to the adaxial position in some legume species, including M. truncatula and pea. The most unique feature of flower development in these legume species is the formation of four common primordia, which are further differentiated into petals and stamens (Tucker, 1989, 2003; Ferrandiz et al., 1999; Benloch et al., 2003). Fig. 4 (a–c,f) shows the developmental stages of floral organs in wild-type flowers. Three distinctive whorls of floral organ primordia are initiated, including the sepal primordia in the outer whorl, common primordia in the middle whorl and the carpel primordium in the inner whorl (Fig. 4b). Later, petal and stamen primordia are differentiated from the common primordia in the middle whorl (Fig. 4c), and then individual floral organs are clearly differentiated (Fig. 4f). In mtnam-2, the sepal primordia are normally developed in the first whorl; common primordia and the carpel primordium are observed inside the first whorl. However, the arrangement of common primordia and the carpel primordium is disorganized (Fig. 4d,e). In mtnam-1, three whorls of organ primordia are initiated and separated (Fig. 4g–i). Although the sepal primordia appear normal, only one common primordium develops and no clear petal and stamen primordia are observed at later stages (Fig. 4h). At the center, the carpel primordium is differentiated (Fig. 4g–i). These observations indicate that the defects of mtnam flowers occur at both floral organ primordium initiation and later development stages.
All classes of floral homeotic genes are down-regulated in mtnam mutants
To gain insights into potential genes downstream of NAM/CUC, we compared the gene expression profiles in shoot apices at the reproductive stage between the regenerated mtnam-1 mutant plants and bulked wild-type-like plants (including heterozygous mtnam-1 and wild-type plants) by Affymetrix microarray. Of the 300 down-regulated genes (Table S1), nine MADS-box family genes and one APETALA2 (AP2)-like gene were identified, which represent homeotic genes related to floral organ development. Sequence analysis indicated that the 10 down-regulated floral homeotic genes in mtnam-1 fall into the A, B, C and E classes, including two genes in the A class, three in the B class, two in the C class and three in the E class (Table 2). The transcript levels of two B class genes, PI and AP3-like, are reduced > 20-fold. We further validated the microarray results for the 10 homeotic genes in individual mtnam-1 and mtnam-2 plants by real-time quantitative PCR. Results revealed that all 10 genes in the knockout mutant mtnam-1 and seven genes in the weak allele mutant mtnam-2 are down-regulated (Fig. 5). The down-regulation of the floral homeotic genes is in agreement with the severe flower organ defects observed in mtnam mutants.
Table 2. Down-regulation of Medicago truncatula floral homeotic genes in microarray analysis
Relative transcript level
Ten homeotic floral identity genes, which belong to classes A, B, C and E, are significantly down-regulated in the mtnam-1 mutant. The relative transcript level was compared with that of the control (wild-type-looking plants).
PISTILLATA (MtPI), B-class
APETALA3-like 2, B-class
SEPALLATA3-like protein, E-class
AGAMOUS-like protein 1, C-class
AGAMOUS-like protein 2, C-class
SEPALLATA1-like protein, E-class
SEPALLATA 3-like protein, E-class
AP2-L, AP2 domain transcription, B-class
APETALA1/FUL-like protein, A-class
APETALA3-like 2, B-class
MtNAM is expressed at boundaries between organs/organ primordia and meristem
To understand the function of MtNAM, we carried out tissue-specific expression analysis. Semi-quantitative PCR and real-time quantitative PCR analyses showed that MtNAM is highly expressed in shoot apices at both vegetative and reproductive stages; a low expression level was detected in mature flowers and young pods, whereas no expression was detected in roots, stems and mature leaves (Fig. 6a).
The elaborate expression pattern of MtNAM was further explored at various developmental stages by RNA in situ hybridization using the nonconserved region of MtNAM as the probe. During embryogenesis, MtNAM is detectable at the early globular embryo stage and the signal is detected at sites of the cotyledon primordial initiation (Fig. 6b.1). At the heart stage, MtNAM expression is detected between the embryonic meristem and the cotyledon primordia (Fig. 6b.2). The MtNAM signal is maintained at boundaries of cotyledons and the primary meristem at the cotyledon stage (Fig. 6b.3). During the vegetative developmental stage, MtNAM expression is found at boundaries between lateral organ primordia and the shoot apical meristem (Fig. 6b.4). The signal is also detectable at sites where leaflet primordia arise and at boundaries between leaflets (Fig. 6b.5, b.6). At the floral developmental stage, MtNAM is first detected in the floral meristem before floral organ primordia arise. Later, the signal is detected at boundaries between floral organ primordia (Fig. 6b.7, b.8). At the gynoecium stage, the MtNAM signal is detected at the margin of the carpel, where the ovule primordia are initiated. At later stages, signals are detected at boundaries between ovules and the placenta (Fig. 6b.9, b.10). Weak signals are also detectable between integuments in ovules (data not shown). Overall, MtNAM expression was observed at all boundaries, including those between organs of the same type, between different floral organ whorls and between developing organs and adjacent meristems. In addition, the MtNAM signal was detectable before the initiation of primordia.
MtNAM modulates compound leaf development
As described in the previous section, MtNAM is expressed at boundaries between leaflets, and mutations in MtNAM lead to the fusion of three leaflets, indicating that MtNAM is required for leaflet separation. It has been reported that SINGLE LEAFLET1 (SGL1), the FLOCAULA/LEAFY ortholog in M. truncatula, plays a key role in the initiation of leaflet primordia during compound leaf development; the loss-of-function sgl1 mutant exhibits a complete conversion of compound leaves into simple leaves (Wang et al., 2008). The difference between the simple leaves of sgl1 and the fused leaves of mtnam is that the three major veins from the three leaflets are distinguishable in the fused leaves of mtnam (Fig. 2n). To better understand compound leaf development, cross-pollination between sgl1 and mtnam-2 mutants was carried out. As both sgl1 and mtnam mutants are sterile, heterozygous sgl1 and mtnam-2 plants were used for cross-pollination. The resultant sgl1/mtnam-2 double mutant was confirmed by PCR-based genotyping and gene expression analysis (Fig. 7a,b). The double mutant contains a Tnt1 insertion in SGL1 and an MERE1 insertion in MtNAM (Fig. 7a), and has no detectable expression of SGL1 and MtNAM (Fig. 7b). Compared with the phenotype of the single leaves in sgl1 (Fig. 7c.2) and the phenotypes of the partially fused cotyledons and fused compound leaves in mtnam-2 (Fig. 7c.3), the sgl1/mtnam-2 double mutant shows additive phenotypes: partially fused cotyledons at the embryonic stage and shoot growing through the fused cotyledons during seed germination, and single leaves at the post-embryonic stage (Fig. 7c.4). Furthermore, the regulation of MtNAM and SGL1 expression was analyzed in mtnam and sgl1 mutants by real-time quantitative PCR. Results revealed that the expression of MtNAM is reduced in the sgl1 mutant, whereas the expression of SGL1 in both mtnam-1 and mtnam-2 mutants is similar to that of the wild type (Fig. 7d).
MtNAM plays a key role in separating adjacent lateral organs and in the establishment of the primary shoot meristem
In this paper, we report a detailed characterization of M. truncatula mutants, mtnam-1 and mtnam-2, which exhibit defects in leaf and flower development. At vegetative stages, mtnam mutants display fused cotyledons, a defective primary apical meristem, and fused leaflets of compound leaves. The first two phenotypes are shared among all known cuc/nam mutants in other plant species, including cuc1/cuc2, cuc1/cuc3, cup and nam (Souer et al., 1996; Aida et al., 1997; Takada et al., 2001; Vroemen et al., 2003; Weir et al., 2004), whereas the fused leaflet phenotype is only observed in nam/cuc mutants or NAM/CUC down-regulated plants of compound-leafed species (Blein et al., 2008; Berger et al., 2009), indicating that CUC/NAM genes have conserved functions in cotyledon separation, primary apical meristem establishment, and compound leaf development. Although the expression of MtNAM is detected in all lateral organ boundaries at post-embryonic stages, except for the fusion of leaflets in compound leaves, mtnam mutants do not exhibit other lateral organ fusion. This suggests that one or more other NAC-domain genes may have a redundant function in lateral organ separation, although no other NAM-like gene, except for MtNAM, has been identified in the M. truncatula genome.
Unlike cup, nam and gob mutants, in which escaped shoot meristems occasionally form and develop into mature plants, no escaped meristems were observed in the null mutant mtnam-1. Mature plants of mtnam-1 could only be obtained by tissue culture and subsequent plant regeneration, indicating that MtNAM is essential for the embryonic induction of the primary apical meristem, but is not necessary for the post-embryonic development of the apical meristem. Members of the NAM/CUC subfamily contain a conserved NAC domain at the N-terminus and three conserved motifs, V, L and W, in the C-terminal region. The NAC domain of CUC1 and CUC2 in A. thaliana is capable of binding to DNA and is important for promoting adventitious shoot formation, whereas the W motif plays an important role in transactivation (Taoka et al., 2004). The MERE1 insertion in mtnam-2 results in a chimeric transcript. The deduced protein from the chimeric transcript contains the NAC domain, the V and L motifs, and the miRNA164-binding site of MtNAM, but loses the conserved W motif at the C-terminus. The mutant plant shows no apparent defects in SAM development, indicating that the function of the deduced protein is sufficient for the induction of the primary meristem. The result is in agreement with the previous report that the specificity of CUC1 in promoting the formation of adventitious shoots resides in the conserved NAC domain (Taoka et al., 2004). However, the NAC domain alone is not sufficient for SAM development, for a transcript is also detected in the null mtnam-1 mutant, which encodes a deduced 198 amino acid truncated protein with an NAC domain (Notes S1). The partially fused cotyledons and fully fused leaflets in mtnam-2 further suggest that a fully functional MtNAM is required to repress cell proliferation locally in boundaries to separate lateral organs, and the W motif is crucial for the full function of MtNAM.
MtNAM is required for both floral organ separation and primordium development
During flower development, the fusion of sepals, petals or stamens and the occurrence of extra petals or petal whorls are commonly reported in cuc/nam mutants and in mutants with defective expression of CUC1 or miRNA164 (Souer et al., 1996; Ishida et al., 2000; Mallory et al., 2004; Weir et al., 2004; Baker et al., 2005), while a reduced petal number has been reported in A. thaliana plants overexpressing miRNA164 (Laufs et al., 2004). Similarly, fusion of the filaments with petals or carpels and a reduced filament number are observed in the weak allele mutant mtnam-2, indicating the conserved function of MtNAM for adjacent floral organ separation. However, unlike the formation of an extra petal whorl in cuc/nam mutants, the null mutant mtnam-1 displays a reduction of floral organs: one simple chimera whorl with greatly reduced organ numbers at the sites of petal and stamen whorls. This phenomenon implies a special floral organ development process in M. truncatula. In contrast to the centripetal and sequential flower ontogeny in other model species (Bossinger & Smyth, 1996), the floral organ primordia in M. truncatula initiate in the abaxial to adaxial unidirection without synchronization of organs in each whorl. Four common primordia occur sequentially in the second whorl and later differentiate into petals and stamens in the second and third whorls, while the carpel primordium is initiated early at the center (Benloch et al., 2003). The phenotypes of reduced common primordium and floral organ numbers indicate that the mutation of MtNAM represses outgrowth from common primordia. Taken together with the MtNAM expression pattern in floral organ primordia, our results suggest that MtNAM is required to promote the initiation and further development of the secondary primordia.
Plant aerial organs such as cotyledons, leaves, and floral organs have long been regarded as homologous structures (Esau, 1977; Pelaz et al., 2001). The difference between leaves and floral organs could be attributed to the differential expression of floral homeotic genes in flowers but not in leaves (Honma & Goto, 2001). Given the concept that a floral organ is a modified leaf, the unique developmental processes from common primordia to petals and stamens resemble those from the leaf marginal blastozone to secondary leaflet formation during compound leaf development. Down-regulation or mutation of CUC/NAM leads to the suppression of marginal outgrowth and fewer or fused leaflets in diverse compound-leafed species (Blein et al., 2008; Berger et al., 2009). The simplification of floral organ whorls and numbers in mtnam-1 is similar to the reduction of compound leaf complexity in gob mutants in tomato and the CUC/NAM down-regulated plants in other species (Blein et al., 2008; Berger et al., 2009). It is suggested that MtNAM shares an essential role in promoting secondary outgrowth with other CUC/NAM members.
MtNAM promotes meristem and primordium identity
Lateral organ primordia are initiated at the periphery of the apical meristem and are defined at early developmental stages. Each primordium acquires an identity by expressing an organ identity gene(s), which enables it to develop into the appropriate type of lateral organs. Members of the MADS-box gene family are the primary determinants of floral organ identity. The ABC model of floral organ determination has been proposed to explain how the unique combination of ABC types of genes determines the floral organ identity in each whorl of floral organs (Coen & Meyerowitz, 1991). An expanded ABC model suggests that class E activity is also required for the specification of each organ type (Gutierrez-Cortines & Davies, 2000). In mtnam mutants, the development of the second whorl is indeterminate and a chimera of petals and stamens or a sepal-like structure is observed, indicating the indeterminacy of floral organs in the whorl. In the center of the flower, the carpel is covered by long dense trichomes, which is a character of the sepal surface, and ovules inside the carpel are severely retarded or transformed into sepal-like structures in extreme cases. These observations indicate that floral organ identity in the inner whorls, except for the first sepal whorl, is impaired in mtnam mutants. The down-regulation of A, B, C and E classes of floral organ identity genes in the mtnam mutants is consistent with the displayed floral organ phenotypes, suggesting that MtNAM is not only required for organ separation, but is also necessary for the expression of organ identity genes to promote the development of floral organ primordia.
In A. thaliana, CUC2/CUC1 is expressed in the early globular embryos and activates the expression of the meristem identity gene STM in the meristem. Mutations in CUC1/2 result in abolished expression of STM and thus abortion of the primary apical meristem. Mis-expression of CUC1 leads to the ectopic formation of adventitious buds and to a change of leaf architecture from simple to compound and induces the expression of SAM-related and compound leaf genes (Aida et al., 1999; Hibara et al., 2003; Hasson et al., 2011). These studies suggest that CUC1/2 activates the expression of SAM identity and compound leaf determination genes and promotes shoot meristem formation. Similarly, MtNAM is also expressed in the early globular embryo in M. truncatula. The defect in the primary apical meristem in the mtnam-1 mutant is probably caused by the impaired expression of a shoot meristem identity gene(s).
Different roles of MtNAM and SGL1 in compound leaf development in M. truncatula
SGL1 plays a key role in compound leaf development. The loss-of-function mutant sgl1 exhibits complete conversion of compound leaves into simple leaves by repressing the initiation of leaflet primordia (Wang et al., 2008). MtNAM functions in separating initiated leaflets to form the compound leaf, and the loss of function leads to fused leaflets. The double mutant of SGL1 and MtNAM-2 shows single leaves resembling the sgl1 single mutant, indicating that the function of SGL1 in leaflet primordium initiation is epistatic to MtNAM on leaflet separation during compound leaf development, although the expression of SGL1 is reduced in mtnam mutants.
In addition to the common role in leaflet separation, NAM/CUC is also required for leaf margin outgrowth in leaf development, which is demonstrated in NAM/CUC-silenced or mutated plants with a greatly reduced number of secondary leaflets and a smooth leaf margin (Blein et al., 2008; Bilsborough et al., 2011). Mutants of MtNAM, however, show three fused leaflets and a wild-type-like serrated leaf margin, indicating that MtNAM does not play a prominent role in leaf margin development, or an unidentified MtNAM-like gene has the redundant function in leaf development in M. truncatula.
This work is supported by the National Science Foundation (NSF-0703285) and the Samuel Roberts Noble Foundation. The authors would like to thank Drs Elison Blancaflor and Ping Xu at the Noble Foundation for their critical reading of the manuscript. We also thank Kuihua Zhang for plant care and seed curation, Shulan Zhang for assistance with flanking sequence recovery, and Hee-Kyung Lee and Janie Gallaway for generating the Tnt1 lines and organizing forward screening.