SEARCH

SEARCH BY CITATION

Keywords:

  • axillary meristem;
  • CUC;
  • miR164;
  • Arabidopsis;
  • LAS

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Aerial architecture in higher plants is established post-embryonically by the inception of new meristems in the axils of leaves. These axillary meristems develop into side shoots or flowers. In Arabidopsis, the NAC domain transcription factors CUP SHAPED COTYLEDON1 (CUC1), CUC2 and CUC3 function redundantly in initiating the shoot apical meristem and establishing organ boundaries. Transcripts of CUC1 and CUC2 are targeted for degradation by miR164. In this study, we show that cuc3-2 mutants are impaired in axillary meristem initiation. Overexpression of miR164 in the cuc3-2 mutant caused an almost complete block of axillary meristem formation. Conversely, mir164 mutants and plants harbouring miR164-resistant alleles of CUC1 or CUC2 developed accessory buds in leaf axils. Collectively, these experiments reveal that, in addition to CUC3, redundant functions of CUC1 and CUC2 as well as miR164 regulation are required for the establishment of axillary meristems. Studies on LAS transcript accumulation in mir164 triple mutants and cuc3-2 plants overexpressing miR164 suggest that regulation of axillary meristem formation by miR164 is mediated through CUC1 and CUC2, which in turn regulate LAS.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Shoot branching in plants is initiated by the formation of new meristems in the axils of leaves, which develop into secondary axes of growth. Axillary meristems recapitulate the function of the shoot apical meristem (SAM) by initiating several leaf primordia, resulting in the formation of axillary buds, which either grow out or remain dormant depending on their position along the shoot axis, the developmental phase of the plant and environmental factors (Mcsteen and Leyser, 2005).

The LATERAL SUPPRESSOR genes in Arabidopsis (LAS; Greb et al., 2003) and tomato (Solanum lycopersicum; Schumacher et al., 1999) encode putative transcription factors belonging to the GRAS family, and specifically regulate the initiation of axillary meristems during the vegetative phase of development. Mutations in MONOCULM1, the rice orthologue of LAS, cause defective tiller formation, altered rachis branches, and modified spikelets (Li et al., 2003). The tomato blind mutant displays a strong reduction in axillary meristem formation during vegetative and reproductive development (Schmitz et al., 2002). The Blind gene encodes a transcription factor of the R2R3 MYB class. In Arabidopsis, the Blind-homologous RAX genes control the formation of axillary meristems in overlapping zones along the shoot axis (Keller et al., 2006; Mueller et al., 2006). Transcripts of LAS, RAX1 and RAX3 accumulate in similar domains in the developing leaf axil where axillary meristems will form (Greb et al., 2003; Keller et al., 2006; Mueller et al., 2006), suggesting that the axillary region has a special identity which may be a prerequisite for the formation of new meristems.

CUC1, CUC2 and CUC3 encode NAC domain transcription factors (Ernst et al., 2004; Olsen et al., 2004). CUC1 and CUC2 are redundantly involved in the initiation of the SAM through the regulation of STM expression and in the establishment of cotyledon and floral organ boundaries (Aida et al., 1999; Long et al., 1996). The discovery of CUC3, encoding a protein with high functional similarity to CUC1 and CUC2, uncovered an additional level of redundancy in the function of these genes (Vroemen et al., 2003). When combined with cuc1 or cuc2, mutations in CUC3 lead to the formation of cup-shaped cotyledons, organ fusions and defects in shoot branching (Hibara et al., 2006; Vroemen et al., 2003).

A subset of NAC-domain transcription factors comprising CUC1, CUC2, NAC1, At5g07680, At5g61430 and At5g39610 is post-transcriptionally regulated by microRNA164 (miR164; Rhoades et al., 2002; Kasschau et al., 2003; Laufs et al., 2004; Mallory et al., 2004; Baker et al., 2005; Schwab et al., 2005). Plant microRNAs are endogenous, single-stranded, non-translated RNA molecules which are highly complementary to their target mRNAs and induce post-transcriptional gene silencing by catalysing cleavage of their targets (Bartel and Bartel, 2004). CUC1 and CUC2 mRNAs are cleaved within their miR164-binding site (Kasschau et al., 2003), and plants containing miR164-resistant versions of CUC1 or CUC2 show severe alterations in embryonic, vegetative and floral development due to an enlargement of various boundary domains (Baker et al., 2005; Laufs et al., 2004; Mallory et al., 2004; Nikovics et al., 2006). Constitutive overexpression of mirR164 was shown to phenocopy the cuc1 cuc2 double mutant by downregulating CUC1 and CUC2 transcript accumulation (Laufs et al., 2004; Mallory et al., 2004).

miR164 is encoded by three genes: MIR164A, MIR164B and MIR164C (Bonnet et al., 2004; Jones-Rhoades and Bartel, 2004; Reinhart et al., 2002; Wang et al., 2004). mir164a and mir164b mutants show increased lateral root formation (Guo et al., 2005), and a loss of function mutation in MIR164C leads to the formation of extra petals in early arising flowers (Baker et al., 2005). miR164a and miR164b mutations enhance the floral defects of mir164c plants substantially, showing that these miRNAs control flower development in a redundant manner and revealing a role of miR164 in the architecture of the inflorescence stem (Peaucelle et al., 2007; Sieber et al., 2007). MIR164A was also shown to regulate the development of leaf margins (Nikovics et al., 2006).

In this study, we have analysed the role of the Arabidopsis CUC genes and miR164 in the process of shoot branching. Characterization of plants containing a cuc3-2 knockout allele in combination with or without miR164 overexpression uncovered partially redundant functions of CUC1, CUC2 and CUC3 in axillary meristem formation. Analysis of mir164 knockout mutants and of transgenic plants carrying miR164-resistant CUC1 and CUC2 alleles demonstrated that miR164 regulation of CUC1/CUC2 transcript accumulation is required to achieve a wild-type shoot branching pattern. Expression studies suggest that CUC1 and CUC2 control axillary meristem development through regulation of LAS, whereas CUC3 may function in an LAS-independent manner.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

cuc3-2 is strongly compromised in axillary meristem formation during vegetative development

CUC3 encodes a putative NAC-domain transcription factor and its role in the establishment of boundaries between the cotyledons as well as between the SAM and lateral organs has been described previously (Vroemen et al., 2003). Because axillary meristems develop in the boundary zone between the SAM and leaf primordia, we have analysed the pattern of axillary bud formation in plants homozygous for the null allele cuc3-2 (Vroemen et al., 2003). In contrast to the Wassilewskija (Ws) wild type, cuc3-2 plants grown under short-day (SD) conditions show a strong reduction in the number of axillary buds originating from the axils of rosette leaves (Figure 1h). Most of the rosette leaves formed in the early and middle phase of vegetative development did not support the formation of axillary buds. However, axillary buds developed from a high proportion of the rosette leaves formed late in vegetative development (Figure 1h). Closer inspection of the empty leaf axils using a stereomicroscope and scanning electron microscopy (SEM) uncovered barren leaf axils without any indication of morphologically distinguishable axillary structures (Figure 1c,e).

image

Figure 1. CUC3 regulates axillary bud formation during vegetative development. (a) Growth habit of a Wassilewskija (Ws) wild type (left) and a cuc3-2 (right) plant grown for 35 days in short photoperiods and subsequently induced to flower in long days. (b, c) Close up of rosette leaf axils in Ws (b) and in cuc3-2 (c) showing presence (b, arrow) and absence (c, arrow) of axillary bud, respectively. (d, e) Scanning electron microscopy (SEM) micrographs of a Ws rosette leaf axil showing an axillary bud (d) and a barren cuc3-2 rosette leaf axil (e, arrow). (f, g) Comparison of STM mRNA accumulation in cross-sections through vegetative shoot apices of Ws (f) and cuc3-2 (g) plants. In addition to expression in the meristem and the interprimordial regions, focused STM expression was seen in the adaxial centre of older leaf primordia in the wild type (f, arrow) which is absent in cuc3-2. Bar = 400 μm in (f) and (g). (h, i) Schematic representation of axillary bud formation in leaf axils of cuc3-2 plants in comparison with Ws plants grown either under short-day conditions (h, = 18) or in long photoperiods (i, = 18). Leaf axils of plants were examined under a binocular microscope. Each column in (h) and (i) represents a single plant, with each square within a column representing an individual leaf axil. The horizontal line represents the border between the youngest rosette leaf and the oldest cauline leaf, with positions of progressively older rosette leaves below the line, and positions of progressively younger cauline leaves above it. The green colour denotes the presence of an axillary bud and yellow the absence of an axillary bud in any particular leaf axil.

Download figure to PowerPoint

Additionally, we studied the expression of the shoot meristem marker STM in cuc3-2 vegetative apices. Similar to its expression pattern in Ws wild-type apices (Figure 1f), STM transcript accumulation was also observed in the SAM and the interprimordial regions of cuc3-2 apices (Figure 1g). In accordance with previously reported studies (e.g. Long and Barton, 2000), STM expression was excluded from incipient and existing leaf primordia. However, the focused STM expression found in the adaxial boundary of older leaf primordia in the Ws wild type (Figure 1f, arrowhead; Greb et al., 2003) was conspicuously absent in cuc3-2 plants. These focused STM expression domains in wild-type plants are indicative of the organization of a new meristem in the axillary region, and thus their absence suggests that cuc3-2 plants are incapable of initiating these axillary meristems. From these histological and in situ hybridization experiments we concluded that the reduction in side-shoot formation in cuc3-2 was due to a failure in axillary meristem initiation rather than due to a defect in axillary bud outgrowth. During reproductive development, axillary bud formation in cuc3-2 did not deviate from the Ws wild type (Figure 1h). A strong reduction in the number of axillary buds in the axils of early rosette leaves was also observed in cuc3-2 plants grown under long photoperiods, but meristem development remained unaffected in the axils of cauline leaves (Figure 1i). Our results extend observations by Hibara et al. (2006), who reported a reduction in tertiary shoot formation in cuc3-101 and cuc3-105 plants. Taken together, these results demonstrate that the CUC3 gene plays a critical role in the genetic control of the formation of axillary meristems.

CUC1 and CUC2 transcripts accumulate in the axils of leaf primordia

CUC1 and CUC2, a pair of closely related genes, belong to a different clade of the NAC family of transcription factors than CUC3 (Vroemen et al., 2003; Zimmermann and Werr, 2005). However, they have been shown to function redundantly with CUC3 in the formation of the SAM and specification of cotyledon boundaries (Vroemen et al., 2003). Here, we monitored the distribution of CUC1 and CUC2 mRNA in vegetative shoot apices by RNA in situ hybridization experiments on tissue sections of Columbia (Col) wild-type plants that were grown under SD conditions and fixed 28 days after sowing. CUC1 transcripts were detected in the axils of young leaf primordia from P0 to P6/P7 (Figure 2a). The CUC1 expression domain was about three to five cell layers deep, including the L1–L3 layers of the SAM, and extended one or two cell layers into the adaxial–abaxial dimension. Transverse sections revealed that CUC1 transcripts accumulated in a band-shaped domain along the adaxial boundary of the leaf primordium (Figure 2b). Expression of CUC1 was not detectable in the axils of P7/P8 to P20/P21 primordia, but was again found from P21/P22 onwards, indicating the onset of axillary meristem activity in these leaf axils (data not shown).

image

Figure 2.  Patterns of CUC1 and CUC2 transcript accumulation in the vegetative shoot apex. Longitudinal (a, c) and transverse (b, d) sections through shoot apices of 28-day-old Columbia (Col) plants grown under short-day conditions were hybridized with a CUC1 (a, b) or a CUC2 (c, d) antisense probe. The arrows point to the CUC1 and CUC2 expression domains. Bars = 200 μm in (a) for (a) and (b), and in (c) for (c) and (d).

Download figure to PowerPoint

CUC2 mRNA was found to accumulate in a domain similar to the CUC1 domain (Figure 2c). Different from CUC1, CUC2 transcripts were detected in the axils of leaf primordia from P0 to P17. Transverse sections revealed accumulation of CUC2 transcript along the adaxial border of leaf primordia (Figure 2d). Vroemen et al. (2003) reported that during vegetative development CUC3 transcripts also accumulate in a narrow domain in the axils of leaf primordia. Taken together, these data suggest that during vegetative growth CUC1, CUC2 and CUC3 mRNAs accumulate in the leaf axil zone from which axillary meristems develop.

Since cuc3-2 mutants displayed a clear defect in axillary shoot development, we also analysed axillary bud formation in plants homozygous for null alleles of CUC1 and CUC2. Under long-day as well as SD growing conditions the shoot branching patterns of cuc1-1 and cuc2-1 mutants showed no aberration compared with their corresponding wild-type Landsberg erecta (Ler). The early growth arrest in cuc1 cuc2 double mutants (Aida et al., 1997) rendered it impossible to examine the pattern of shoot branching in these double mutants.

Overexpression of MIR164A or MIR164B in a cuc3-2 background enhances the cuc3-2 branching defect

CUC1 and CUC2 were predicted to be post-transcriptionally regulated by miR164 (Schwab et al., 2005). Overexpression of this miRNA was shown to reduce CUC1 and CUC2 RNA levels and phenocopy the cuc1 cuc2 double mutant phenotype (Laufs et al., 2004; Mallory et al., 2004). To uncover any function of CUC1 and/or CUC2 in the process of axillary meristem formation that might be masked by CUC3 activity, we characterized the shoot branching pattern of transgenic lines harbouring a 2x35S::MIR164A or 2x35S::MIR164B construct (Laufs et al., 2004), thus overexpressing miR164, in a Ws or a cuc3-2 background. One 2x35S::MIR164A and two 2x35S::MIR164B transgenic lines in the Ws background were analysed. All three transgenic lines showed fusions of cotyledons, sepals and stamens, as described previously (Laufs et al., 2004; Mallory et al., 2004). However, in both short and long photoperiods, overexpression of MIR164A or MIR164B in Ws plants did not lead to any deviation from the wild-type branching pattern, neither in the vegetative phase nor in the reproductive phase of development (Figure 3a).

image

Figure 3.  Axillary bud formation in 2x35s::MIR164B plants. Plants were grown for 35 days in short photoperiods and subsequently induced to flower in long photoperiods. (a, b) Growth habits of a transgenic plant harbouring the 2x35s::MIR164A construct in a Wassilewskija (Ws) wild-type background (a) and in a cuc3-2 mutant background (b). (c) Schematic representation of axillary bud formation in leaf axils of cuc3-2 2x35s::MIR164A in comparison with Ws wild-type plants. As explained in the legend to Figure 1, green and yellow boxes denote the presence or absence of an axillary bud, respectively, in leaf axils along the main shoot. (d) Ten-day-old cuc3-2 2x35s::MIR164A plant grown showing cup-shaped cotyledons. (e) Twenty-day-old cuc3-2 2x35s::MIR164A plant showing formation of a new shoot apex (arrow) at the hypocotyl. (f) Rosette leaves fused to the main shoot and displaying empty leaf axils (arrow) and distorted phyllotaxis in a mature cuc3-2 2x35s::MIR164A plant. (g) Empty leaf axils (arrow) and curled cauline leaves in a mature cuc3-2; 2x35s::MIR164A plant.

Download figure to PowerPoint

The branching patterns of two populations of plants homozygous for cuc3-2 and segregating for 2x35S::MIR164A (population A) and 2x35S::MIR164B (population B), respectively, were analysed. These plants were grown in short photoperiods for 30 days and then shifted to long days to induce flowering. From these populations we selected plants that formed cup-shaped cotyledons indicating a strong downregulation of CUC1 and CUC2 activity (Figure 3d). The PCR analysis demonstrated that these cup-shaped seedlings contained the 2x35S::MIR164 T-DNA. The majority of these plants initiated a new SAM in the hypocotyl region (Figure 3e), which produced rosette leaves and developed into a flowering shoot (Figure 3b). During vegetative (Figure 3f) as well as reproductive (Figure 3g) development, these plants failed to develop axillary buds in most of their leaf axils. Very rarely, in some plants a single axillary shoot was observed in a cauline leaf axil (Figure 3c). In addition, these plants were characterized by a reduction in the number of leaves, fusion of leaf petioles to the main axis, distorted phyllotaxis, curled cauline leaves, immature siliques and a dark green colour (Figure 3b,c,f,g). The enhancement of the cuc3-2 branching defect by overexpression of MIR164A or MIR164B suggests a redundant role for CUC1 and/or CUC2 in the control of axillary meristem formation.

Plants expressing miR164-resistant transcripts of CUC1 and CUC2 develop accessory side shoots

To further analyse the function of CUC1 and CUC2 in axillary bud formation, we investigated miR164-resistant CUC1 and CUC2 lines, 5mCUC1 and CUC2g-m4, respectively, for their pattern of axillary shoot development. 5mCUC1 and CUC2g-m4 are transgenic lines containing the miR164-resistant CUC1 and CUC2 genes driven by their endogenous promoters in a wild-type background (Mallory et al., 2004; Nikovics et al., 2006). These miR164-resistant alleles have silent mutations in their miR164-binding sites, which insulate them from miR164 regulation, leading to the accumulation of high levels of CUC1 and CUC2 mRNA.

During the reproductive phase, CUC2g-m4 plants frequently developed clusters of cauline leaves with highly reduced internodes (Figure 4a,d, arrows; Peaucelle et al., 2007). Notably, CUC2g-m4 plants developed accessory side shoots in the axils of cauline leaves (Figure 4a,b, arrows) and rosette leaves (Figure 4c, arrow). These additional side shoots developed in the zone between the primary side shoot and the leaf and appeared later than the primary side shoot. During vegetative development in short photoperiods, accessory buds were formed predominantly in the axils of late and middle rosette leaves (Figure 4f). In the reproductive phase, the formation of accessory buds was found to be more enhanced in early cauline leaf axils than in late cauline leaf axils (Figure 4f). An increase in accessory bud formation was observed irrespective of the length of the photoperiods that these plants experienced (data not shown). Accessory side shoots were also found at a very low frequency in cauline leaf axils of the Ws wild-type control (Figure 4f).

image

Figure 4.  Analysis of plants expressing miR164-resistant alleles of CUC1 and CUC2. (a) Close-up of cauline leaf axils in a CUC2g-m4 plant showing accessory bud formation and distorted phyllotaxis. (b) A SEM micrograph of a CUC2g-m4 cauline leaf axil displaying an accessory bud (arrow). Bar = 125 μm. (c) A SEM micrograph of a CUC2g-m4 rosette leaf axil showing an accessory bud (arrow). Bar = 250 μm. (d, e) Habitus of CUC2g-m4 (d) and 5mCUC1 (e) plants grown for 30 days in short photoperiods and subsequently in long days. (f, g) Schematic representation of accessory bud formation in leaf axils of a population of CUC2g-m4 (f) and 5mCUC1 (g) plants in comparison to Wassilewskija (Ws) plants. The pink colour denotes the presence of an accessory bud in a particular leaf axil, and green denotes the absence of accessory bud. Plants were grown for 30 days in short photoperiods and subsequently for 35 days under long-day conditions.

Download figure to PowerPoint

Formation of accessory buds was also observed in four independent 5mCUC1 lines. Similar to CUC2g-m4 plants, accessory bud formation was found mostly in early cauline leaf axils and decreased during late reproductive development (Figure 4g). However, unlike in CUC2g-m4 plants, accessory bud development was observed in 5mCUC1 plants, particularly during the middle phase of vegetative development; very few early and late rosette leaf axils harboured accessory buds (Figure 4g). Furthermore, the overall increase in accessory bud formation during vegetative as well as reproductive development was lower in CUC1-5m plants when compared with CUC2g-m4 plants. For example, almost 90% of CUC2g-m4 plants displayed accessory buds in early cauline leaf axils, whereas in 5mCUC1 plants fewer than 70% of early cauline leaf axils supported accessory bud formation (Figure 4f,g).

MIR164A and MIR164C are expressed in the boundary region between leaf primordia and the SAM

In Arabidopsis the MIR164 gene family comprises three members: MIR164A, MIR164B and MIR164C (Bonnet et al., 2004; Jones-Rhoades and Bartel, 2004; Reinhart et al., 2002; Wang et al., 2004). Previous studies using northern hybridization, RT-PCR, GFP-based transcriptional reporters and mRNA in situ hybridization have shown that these three miRNAs accumulate in leaves, floral organs and roots, and that miR164b is the most abundant of the three (Baker et al., 2005; Guo et al., 2005; Mallory et al., 2004; Nikovics et al., 2006; Sieber et al., 2007). In this study, we examined the expression patterns of the three MIR164 genes in vegetative shoot apices of Col wild-type plants grown for 30 days in short photoperiods, using β-glucuronidase (GUS)-based reporters. Expression from the MIR164A promoter was detected in the axillary regions of the shoot apex and within leaf primordia (Figure 5a,b). In the shoot apex, GUS staining was restricted to a few cells in the L1 layer of the boundary region between leaf primordia and the shoot meristem (Figure 5a,b). In leaf primordia, the MIR164A promoter directed GUS expression to the epidermal layer of both, the abaxial and the adaxial side, and to vascular bundles (Figure 5a,b). On the other hand, GUS staining was not observed in the SAM of plants transformed with the MIR164B promoter construct (Figure 5c,d). However, GUS staining was detected in the epidermal layer and the vascular bundles of leaf primordia (Figure 5c,d). In the epidermis, staining was more prominent on the abaxial side than on the adaxial side, whereas in vascular tissues GUS activity seems to be restricted to the phloem (Figure 5c,d). In contrast to the MIR164A and MIR164B expression patterns, MIR164C-directed GUS staining was specifically confined to the boundary region between the meristem and leaf primordia as well as to older leaf axils (Figure 5e,f). In young leaf axils, the expression was restricted to the L1 layer, whereas it comprised two or three cell layers in older leaf axils (Figure 5e,f). Notably, GUS staining in the MIR164C transgenic lines was not detected within leaf primordia, neither in the epidermis nor in the vasculature (Figure 5e,f).

image

Figure 5.  Patterns of MIR164A-, MIR164B- and MIR164C- promoter driven GUS expression in the vegetative shoot apex. Longitudinal (a, c, e) and transverse (b, d, f,) sections through shoot apices of MIR164A::GUS (a, b), MIR164B::GUS (c, d) and MIR164C::GUS (e, f) plants. Apices were harvested from plants grown for 30 days under short-day conditions and were incubated in X-Gluc staining buffer. Bars = 100 μm in (a) for (a), (c) and (e), and 400 μm in (b) for (b), (d) and (f).

Download figure to PowerPoint

miR164 regulates the number of axillary buds in leaf axils

Since interference in the miR164 regulation of CUC1 and CUC2 led to the formation of accessory side shoots, we further investigated the role of miR164 in axillary meristem development by characterizing the shoot branching patterns of plants homozygous for the loss of function alleles mir164a-4 (Nikovics et al., 2006), mir164b-1 (Mallory et al., 2004), mir164c-1 (eep1, Baker et al., 2005; Sieber et al., 2007), the three double mutants and the triple mutant.

In short photoperiods, mir164a-4, mir164b-1 and mir164c-1 single mutants developed accessory buds in their leaf axils (Figure 6c). In mir164a-4 and mir164b-1 mutants, accessory bud formation was observed only in the cauline leaf axils, with mir164a-4 displaying a much stronger phenotype than mir164b-1 (Figure 6d). This phenotype was accentuated in mir164a-4 mir164b-1 double mutants (Figure 6d). On the other hand, accessory bud formation was restricted to vegetative development in mir164c-1 (Figure 6d). Axils of rosette leaves formed during the middle phase of vegetative development displayed a greater tendency to harbour an accessory bud than early or late rosette leaf axils. mir164c-1 enhanced the mutant phenotype of mir164b-1, with the mir164b-1 mir164c-1 double mutant displaying a low number of accessory buds in late rosette leaf axils and an increase in the tendency to form accessory buds during reproductive development (Figure 6d). Surprisingly, mir164a-4 mir164c-1 plants developed accessory buds exclusively in cauline leaf axils (Figure 6d). However, these double mutants formed a small rosette within each rosette leaf axil, making it difficult to analyse accessory bud formation during the vegetative phase of development. Accessory bud formation was most prominent in mir164a-4 mir164b-1 mir164c-1 (mir164abc) triple mutants, both during the vegetative and the reproductive phase of development (Figure 6c,d). At a very low frequency, two accessory buds per cauline leaf axil were observed in the mir164abc mutant as well as in the mir164ab mutant. mir164abc frequently displayed extremely short internodes during reproductive development (Figure 6c) and all cauline leaf axils displayed an increased tendency to harbour an accessory bud (Figure 6d). During vegetative development, this tendency was stronger in the axils of late than of early rosette leaves (Figure 6d). Similar, though less pronounced, patterns of accessory bud formation were also seen in all mir164 mutant combinations grown in long photoperiods (data not shown). These results show that MIR164A, MIR164B and MIR164C redundantly regulate the number of side shoots, with a greater contribution of MIR164A during reproductive development and MIR164C during vegetative development.

image

Figure 6. mir164 mutants develop accessory buds. (a, b) Habitus of a mir164c (a) and a mir164b mir164c plant (b) grown for 30 days in short photoperiods and subsequently in long days. (c) Close up of cauline leaf axils of a mir164abc plant showing accessory bud formation (arrow) and distorted phyllotaxis. (d) Schematic representation of accessory bud formation in leaf axils of mir164a (a), mir164b (b), mir164c (c), mir164a mir164b (ab), mir164b mir164c (bc), mir164a mir164c (ac) and mir164a mir164b mir164c (abc) mutants in comparison to wild-type plants. The pink colour denotes the presence of an accessory bud in a particular leaf axil, and green denotes the absence of an accessory bud. mir164a and mir164b are in the Columbia (Col) background, and mir164c in Landsberg erecta (Ler). The wild type (Wt) depicted here is from the F2 of a cross between Col and Ler, and did not differ from Col or Ler with respect to accessory bud formation. A representative population of 10 plants out of 18 plants of each genotype, grown for 30 days in short photoperiods and subsequently for 30 days under long-day conditions, is shown here.

Download figure to PowerPoint

miR164 fine tunes axillary meristem initiation by regulating LAS through CUC1 and CUC2

Our studies on GUS expression from the promoters of the MIR164 genes revealed that the pattern of miR164 accumulation in the shoot apex partially overlapped with the expression domains of its targets, CUC1 and CUC2. Therefore we also examined the expression patterns of CUC1 and CUC2 in vegetative shoot apices of the mir164abc mutant grown for 30 days in short photoperiods. In mir164abc plants, CUC1 transcripts accumulated in a similar pattern as in the corresponding wild type, but the expression levels seemed to be upregulated in the axillary regions of the triple mutant (Figure 7a,b). On the other hand, we observed not only strongly elevated levels of CUC2 mRNA in mir164abc apices, when compared to wild-type apices, but also a broadening of its expression domain encompassing cells of the L1, L2 and L3 layers throughout the shoot apex (Figure 7c,d).

image

Figure 7.  Interactions between miR164, the CUC genes and LAS in the control of axillary meristem formation. Longitudinal sections through shoot apices of wild-type (a, c, e), mir164abc (b, d, h), cuc3-2 (f) and cuc3-2 2x35S::MIR164A (g) plants grown for 30 days under short-day conditions, hybridized with CUC1 (a, b), CUC2 (c, d) and LAS (e–h) anti-sense probes. mir164abc and wild-type plants had a mixed Columbia (Col)l/Landsberg erecta (Ler) background, whereas cuc3-2 and cuc3-2 2x35S::MIR164A were in the Wassilewskija (Ws) background. LAS expression in Col/Ler wild-type apices was comparable to that in Ws wild-type apices. Bar = 200 μm in (a) for (a) to (h).

Download figure to PowerPoint

The remarkable similarities between the expression patterns of miR164, the CUC genes and LAS, a well-known regulator of axillary meristem initiation in plants (Greb et al., 2003), together with the corresponding mutant phenotypes, suggested a hierarchical interaction between these genes in controlling axillary meristem development. To test this hypothesis, we examined LAS expression in vegetative apices of the cuc mutants. LAS transcript accumulation was found to be unaltered in apices of cuc1-1 and cuc2-1 mutants (Figure S1). Furthermore, we detected no change in LAS mRNA distribution in apices of the cuc3-2 mutant, which displays a defect in axillary meristem initiation, when compared with the corresponding wild type (Figure 7e,f). Apices of cuc3-2 35S::MIR164A plants, on the other hand, displayed a reduction in LAS mRNA accumulation (Figure 7g). Conversely, a substantial upregulation of the LAS transcript was observed in mir164abc mutant apices (Figure 7h). From these lines of evidence we conclude that LAS expression is positively regulated by concerted CUC activities. Taken together, these data suggest that the initiation of axillary meristems is confined by miRNA164 through the restriction of CUC1 and CUC2 mRNA accumulation, which in turn regulate LAS expression.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

CUC genes are redundantly required to initiate axillary meristem formation

The CUC genes are expressed in a variety of boundary regions between organs as well as between the SAM and developing organs (Keller et al., 2006; Takada et al., 2001; Vroemen et al., 2003). Because axillary meristems develop from the boundary region between a leaf primordium and the inner part of the shoot apex, we tested for a possible role of these genes in axillary meristem formation. Phenotypic analysis demonstrated that cuc3-2 plants do not develop axillary buds in rosette leaf axils in the early and middle phase of vegetative development. Axillary bud formation was much less compromised towards the top of the rosette and in the reproductive phase. The SEM analysis demonstrated that empty leaf axils did not contain any morphological structures that could be traced back to the activity of an axillary meristem. The absence of axillary meristem initiation in cuc3-2 was confirmed by the lack of a focused STM expression in the axils of older leaf primordia in these plants. Hibara et al. (2006) observed the lack of axillary shoots at a low frequency in cuc3-105 mutants. Our detailed analysis further establishes and refines the role of CUC3 as a key regulator of axillary meristem formation during the early and middle phases of vegetative development.

The block in axillary bud formation in cuc3-2 was not absolute, because axillary buds developed during the late vegetative phase and the reproductive phase. On the other hand, cuc1-1 and cuc2-1 mutants as well as wild-type plants overexpressing miR164 did not show a defect in axillary bud formation. To test for a possible redundant involvement of CUC1 and CUC2 in the process of axillary meristem initiation, we studied the consequences of misexpression of MIR164A and MIR164B under the control of the CaMV 35S promoter in the cuc3-2 background. Those plants that were characterized by a strong reduction in CUC1 and/or CUC2 activity, as indicated by the formation of cup-shaped cotyledons, showed an almost complete block in axillary bud formation in both rosette and cauline leaf axils. Taken together, these results suggest that, in addition to CUC3, CUC1 and/or CUC2 also play a role in the regulation of axillary meristem initiation.

The function of CUC1 and CUC2 in lateral meristem formation was further studied by analysing the shoot branching patterns of transgenic lines containing miR164-resistant variants of CUC1 and CUC2, namely 5mCUC1 and CUC2g-m4. These plants developed accessory buds in rosette and cauline leaf axils, demonstrating that increased CUC1 or CUC2 activities lead to additional axillary meristems. However, the increase in accessory bud formation was considerably stronger in CUC2g-m4 than in 5mCUC1, indicating that CUC2 has a more prominent role in axillary meristem initiation than CUC1. Altogether, these data suggest that CUC activity is not only required for axillary meristem formation but is also rate limiting during normal development and that additional axillary meristems form in response to an increase in CUC1 or CUC2 activity. Along the same lines, the phenotype of the cupuliformis mutant, which develops malformed leaf axils without lateral buds, put forward a role for the Antirrhinum CUC1/CUC2 homologue in axillary meristem formation (Weir et al., 2004).

miR164 controls the number of meristems per leaf axil

Schwab et al. (2005) have shown that six NAC-domain transcription factors contain binding sites for miR164, which is encoded by the genes MIR164A, MIR164B and MIR164C. From this subgroup of NAC-domain genes only CUC1 and CUC2 are expressed in the shoot apex (http://www.genevestigator.ethz.ch/). CUC3, the third redundant CUC gene that is expressed in the shoot tip, does not carry a recognition site for miR164. Phenotypic analysis revealed that mir164a-4 and mir164b-1 loss of function mutants developed accessory buds in the axils of cauline leaves, whereas in mir164c-1 accessory buds were formed in the axils of rosette leaves. This phenotype was enhanced in mir164a-4 mir164b-1 double mutants. On the other hand, accessory bud formation during the vegetative phase was reduced in mir164b-1 mir164c-1 and absent in mir164a-4 mir164c-1 mutants. This may be explained by the fact that the mir164c-1 mutant has a different genetic background (Ler) than the mir164a-4 and mir164b-1 mutants (Col) indicating the influence of accession-specific modifiers on this phenotype. However, the triple mutant developed accessory buds in most of its leaf axils. These results suggest that the three genes encoding miR164 differentially regulate CUC1 and CUC2 transcripts in overlapping zones along the shoot axis.

In situ hybridization analysis of miR164 accumulation in Nicotiana benthamiana revealed specific temporal and spatial expression patterns in developing flowers, ovules, ovaries, pollen sacs and anthers, and high levels of miR164 in meristems, procambial strands and vascular bundles (Valoczi et al., 2006). In Arabidopsis, mir164a-1 and mir164b-4 were shown to enhance the floral defects of mir164c-1, and the three genes, MIR164A, MIR164B and MIR164C, were reported to have partially overlapping and distinct expression patterns in floral apices, indicating partially redundant yet specific roles for these genes in regulating floral development (Sieber et al., 2007). Our results further support this concept of partial redundancy and specialization in the functions of the three MIR164 genes. Through GUS reporter studies, we showed that MIR164A and MIR164C are active in domains overlapping with the boundary between the SAM and leaf primordia, whereas MIR164B is not transcribed in these axillary regions. These observations correlate well with the phenotypes of the mir164 single mutants. Additionally, the mir164abc triple mutant displayed changes in CUC1 and CUC2 mRNA accumulation. Loss of miR164 activity leads to a considerably broader domain of CUC2 mRNA accumulation and an upregulation of CUC1 transcription, resulting in the formation of more than one axillary meristem. In addition to spatial regulation, miR164 may be needed to restrict the developmental time window of CUC1 and CUC2 mRNA accumulation.

Interplay between miR164, CUP SHAPED COTYLEDON genes and LATERAL SUPPRESSOR in the control of axillary meristem initiation

The cuc3-2 branching pattern is similar to the patterns of shoot branching observed in the Arabidopsis las-4 and rax1-3 mutants. In cuc3-2 and rax1-3 mutants, axillary meristem formation is impaired in the early phase of vegetative development and restored when the plant matures (Mueller et al., 2006). las-4 mutants do not usually develop axillary buds in the vegetative phase, but depending on growth conditions axillary shoots are formed in the axils of the topmost rosette leaves at a low frequency (Greb et al., 2003). Transcripts of the CUC1, CUC2, CUC3, RAX1 and LAS genes accumulate in overlapping domains in the axils of developing leaf primordia (Greb et al., 2003; Keller et al., 2006; Mueller et al., 2006; Vroemen et al., 2003). Taken together, the similarities in the mutant phenotypes and in the patterns of transcript accumulation raise the possibility that CUC1, CUC2, CUC3, RAX1 and LAS interact to establish and maintain the competence for axillary meristem formation during the phase of vegetative development. The RNA in situ hybridization analysis of CUC2 transcript accumulation in rax1-2 mutants suggested that RAX1 regulation of axillary meristem initiation is mediated through CUC2 (Keller et al., 2006; Figure 8). Hibara et al. (2006) showed that LAS and CUC3 have overlapping roles in various aspects of boundary formation, including the establishment of axillary meristems. We investigated a possible genetic interaction between LAS and CUC3 in axillary meristem development by studying the accumulation of LAS mRNA in the cuc3-2 shoot apex. Although failing to initiate axillary meristems, the cuc3-2 mutant accumulates LAS mRNA in a wild-type-like pattern, suggesting that CUC3 regulates axillary meristem development in a LAS-independent manner. However, because of the limitations of the RNA in situ hybridization technique, a slight downregulation in the level of LAS transcript would not have been detectable. Furthermore, LAS expression in cuc1-1 and cuc2-1 single mutants does not deviate from the wild-type pattern. On the other hand, overexpression of MIR164 in a cuc3-2 background results in a downregulation of LAS transcript accumulation. Along the same line, the LAS mRNA level is strongly upregulated in shoot apices of the mir164 triple mutant. These two lines of evidence suggest that miR164 indirectly regulates LAS function, through CUC1 and CUC2 (Figure 8). LAS expression was also shown to be downregulated in embryos of the cuc1 cuc2 double mutant (Hibara et al., 2006), supporting the conclusion that CUC1 and CUC2 are regulators of LAS activity. Future experiments will have to show whether CUC3 is redundantly involved in the same pathway or acts independently.

image

Figure 8.  Regulation of axillary meristem initiation. miR164 restricts expression of CUC1 and CUC2, but not CUC3, in the axillary regions. CUC1 and CUC2 promote axillary meristem formation through the regulation of LAS. CUC3 promotes the formation of axillary meristems, either through regulation of LAS or via an LAS-independent pathway. RAX1 has been previously shown to regulate the development of axillary meristems through CUC2 (Keller et al., 2006).

Download figure to PowerPoint

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant material

Arabidopsis thaliana ecotypes Ws, Col and Ler were used as wild types in this study. cuc1-1, cuc2-1 and cuc3-2 were obtained from the Nottingham Arabidopsis Stock Center (NASC). mir164b-1 and mir164a-4 were obtained from the Salk Institute Genomic Analysis Laboratory collection (SALK N636105) and the GABI insertion collection (GABI 867E03), respectively. mir164c-1, mir164ac, mir164bc and mir164abc were kindly provided by P. Sieber (University of Zurich, Switzerland). 2x35S::MIR164A and 2x35S::MIR164B constructs and the 5mCUC1 and CUC2g-m4 transgenic line have been described elsewhere (Laufs et al., 2004; Mallory et al., 2004; Nikovics et al., 2006). The 2x35S::MIR164A and 2x35S::MIR164B transgenic lines were crossed to cuc3-2 homozygotes. Seventeen F2 plants showing the sepal fusion phenotype of plants overexpressing miR164 (Laufs et al., 2004) and additionally displaying fusions of rosette leaves and fusions between leaves and the inflorescence stem, were selected from these two crosses. The progeny of these F2 plants segregated plants that were indistinguishable from the cuc3-2 homozygous parent and plants with leaf fusions, indicating that the F2 was homozygous for cuc3-2 and hemizygous for the 2x35S::MIR164A or B construct. Plants homozygous for both cuc3-2 and the 2x35S::MIR164A or 2x35S::MIR164B construct could not be identified, suggesting that they may be embryo lethal or arrested very early in development. The MIR164A::GUS line has been described before (ProMIR164A2.1:GUS; Nikovics et al., 2006).

Growth conditions

Plants were grown to maturity either in long days in greenhouses or in short days (8 h light 23°C and 16 h darkness 18°C) in a controlled environment. Wherever specifically mentioned in the text, SD plants were induced to flower after 30–35 days by shifting to long photoperiods.

Analysis of side shoot formation

Axillary bud formation in the axils of rosette and cauline leaves was examined using a stereomicroscope. The analyses were done by sequentially checking the oldest to the youngest leaf axils for initiation of a bud. The older leaves were successively removed to make the younger leaf axils available for inspection. At least 18 plants of each genotype were analysed in every experiment. Each experiment was repeated at least once.

DNA isolation and PCR

Plant DNA preparation was carried out using Qiagen DNeasy®96 Plant Kit (http://www.qiagen.com/) as per the manufacturer’s instructions. For the MIR164B::GUS reporter, 2043 bp of promoter sequence (from positions −2042 to −3 relative to the first nucleotide of mature miR164) were amplified from Col using miR164B-6 (TTG CTC ATC ACA CAC CTT CAT) and mir164B-18 attb1 (AAA AAA GCA GGC TTA ACT TGA CAT GAT ATA CAC CAC T) primers and used to drive the GUS reporter in the pBIB101.3 vector. A 2.3-kb promoter was shown by Guo et al. (2005) to complement the root phenotype of the mir164b-1 mutant. For the MIR164C::GUS reporter, 804 bp of promoter sequence (from positions –808 to –5 relative to the first nucleotide of mature miR164) was amplified from Ws using mir164c-13-attb2 (AAG AAA GCT GGG TTC AAG TGT TAC TCA CCC ATT ACT) and mir164c-14-attb1 (AAA AAA GCA GGC TGG ACC CAA ACT CAT CAC CTA TCT) primers and used to drive the expression of a GUS reporter in the Gateway pBI101-R1R2-GUS plasmid (Divol et al., 2007). The expression pattern driven by this promoter in the shoot apex is similar to that by a 1.8-kb promoter (A. Peaucelle and and P. Laufs, unpublished results) which contains a region sufficient for the function of MIR164C (Baker et al., 2005). The PCR detection of 2x35S::MIR164 T-DNA in segregating cuc3-2, 2x35S::MIR164/+ populations was performed using the primers MIR164A532F (TGG AGA AGC AGG GCA CGT GCA) and t35sR (CCT TAT CGG GAA ACT ACT CAC ACA T).

RNA in situ hybridization

Sample preparations and in situ hybridizations including probe hydrolysis were performed as described previously (Greb et al., 2003). The CUC1, CUC2 and LAS probes contained the nucleotide sequence 501–961, 496–1128 and 2–1348, respectively, relative to the ATG.

GUS assay

The GUS staining of apices harvested from plants grown for 30 days in short photoperiods was carried out as described in Sessions et al. (1999). Stained apices were embedded in Paraplast (Paraplast Plus, Kendall; http://www.kendallhq.com) and sectioned. Eight-micrometre sections were viewed by differential interference contrast microscopy after deparaffinization.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank the Nottingham Arabidopsis Stock Centre and the Salk Institute Genomic Analysis Laboratory for seeds stocks. The authors thank A. Mallory for the 5mCUC1 construct and P. Sieber for mir164ac, mir164bc and mir164abc seeds. We are grateful to R.-D. Hirtz for help with scanning electron microscopy and U. Pfordt and A. Kalda for excellent technical assistance. We thank G. Schmitz and M. Koornneef for critical reading of the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft through SFB 572 of the University of Cologne (Germany), and partly by the Genosome Project (Génoplante TRIL-046).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Figure  S1. Accumulation of LAS mRNA in cuc1 and cuc2. Longitudinal sections through shoot apices of Col (a), cuc1-1 (b) and cuc2-1 (c) plants grown for 28 days in short photoperiods and then for 2 days under long-day conditions were hybridized with an LAS antisense probe. Bar = 200 &mgr;m in (a) for (a), (b) and (c).

Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
TPJ_3483_sm_figs.pdf3867KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.