SEARCH

SEARCH BY CITATION

Keywords:

  • branching;
  • evolution;
  • leaf venation;
  • meristem;
  • polar auxin transport (PAT);
  • Selaginella kraussiana ;
  • shoot development

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Summary

  • To provide a comparative framework to understand the evolution of auxin regulation in vascular plants, the effect of perturbed auxin homeostasis was examined in the lycophyte Selaginella kraussiana.
  • Polar auxin transport was measured by tracing tritiated IAA in excised shoots. Shoots were cultured in the presence of auxin efflux inhibitors and exogenous auxin, and developmental abnormalities were documented.
  • Auxin transport in Selaginella shoots is exclusively basipetal, as in angiosperms. Perturbed auxin transport results in the loss of meristem maintenance and abnormal shoot architecture. Dichotomous root branching in Selaginella appears to be regulated by an antagonistic relationship between auxin and cytokinin.
  • The results suggest that basipetal polar auxin transport occurred in the common ancestor of lycophytes and euphyllophytes. Although the mechanisms of auxin transport appear to be conserved across all vascular plants, distinct auxin responses govern shoot growth and development in lycophytes and euphyllophytes.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Vascular plants comprise the lycophytes and euphyllophytes, with the latter including monilophytes, gymnosperms and angiosperms. Evidence from fossils and phylogenetic analyses indicates that lycophyte and euphyllophyte lineages probably diverged over 400 million yr ago and that the common ancestor comprised leafless branching axes (Kenrick & Crane, 1997; Pryer et al., 2001). As such, indeterminate apical growth probably evolved before the lycophyte–euphyllophyte divergence, but complex shoot architecture evolved independently in each lineage. To date, comparative studies have suggested that, although some developmental pathways evolved in parallel in both lycophytes and euphyllophytes (Harrison et al., 2005; Sanders et al., 2010), others play distinct roles in each group (Floyd & Bowman, 2006).

In angiosperms, shoot architecture is regulated, in part, by the modulation of auxin transport and homeostasis. For example, vascular tissue is patterned along polar auxin transport (PAT) routes in stems and leaves (reviewed in Berleth et al., 2000), and branching is controlled by patterns of auxin flow (reviewed in Müller & Leyser, 2011). A simplified model of leaf initiation describes how auxin flux between developing primordia and the shoot apex creates local auxin maxima that establish leaf initiation sites and thus determine phyllotactic pattern (Reinhardt et al., 2000, 2003). In Arabidopsis and tomato, loss of PAT results in a loss of leaf production (Reinhardt et al., 2000) and, in maize, aberrant PAT disrupts leaf placement and hence phyllotactic pattern at the apex (Lee et al., 2009). Similar effects are seen in gymnosperms, in that disrupted PAT causes a loss of cotyledon boundaries and abnormal meristem development in spruce embryos (Larsson et al., 2008). In combination, these observations suggest that the role of auxin transport in shoot patterning is conserved across seed plants.

In angiosperms, auxin efflux is inhibited by 1-n-naphthylphthalamic acid (NPA) and 2,3,5-triiodobenzoic acid (TIBA) (Depta & Rubery, 1984). These compounds have therefore been used to examine the role of PAT in the regulation of plant growth (e.g. Mattsson et al., 1999; Rashotte et al., 2000; Reinhardt et al., 2000; DeMason & Chawla, 2004) and to infer the underlying mechanisms (Poli et al., 2003). In Arabidopsis, morphological perturbations induced by auxin efflux inhibitors are accompanied by perturbed function of PIN-FORMED (PIN1) and P-GLYCOPROTEIN (PGP) auxin efflux carriers (Geldner et al., 2001; Geisler et al., 2005). Disruption of PIN1 function leads to similar morphological perturbations, including the loss of leaf formation and disrupted phyllotaxy (Gälweiler et al., 1998; Reinhardt et al., 2003; Blilou et al., 2005; Lee et al., 2009). Although the disruption of PAT in the moss Physcomitrella patens results in morphological aberrations (Fujita et al., 2008), the role of auxin in the regulation of moss development is not well understood. In the lycophyte Selaginella, previous work has suggested that auxin transport is bidirectional (Wochok & Sussex, 1973), but the lack of controls in that study prevented conclusive interpretations of the data. To ascertain more precisely how auxin and PAT regulate developmental processes in lycophytes, we have characterized the morphological consequences of perturbed PAT in the lycophyte Selaginella kraussiana. The data obtained provide a comparative framework to infer how the role of auxin has changed during land plant evolution and, specifically, to compare the mechanisms that regulate shoot architecture in lycophytes and euphyllophytes.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Plant growth conditions

Selaginella kraussiana (Kunze) A. Braun plants were obtained from the Royal Botanic Gardens at Kew. Plants are routinely maintained on peat-based soil at 23°C under continuous light. Segments of these plants were cut for auxin transport experiments. Plants were similarly grown for inhibition and supplementation experiments, but the shoots were excised, sterilized in 3% bleach for 3 min and grown in sterile culture on C-fern medium (http://www.c-fern.org/) at pH 7.1 with 5% agar at 23°C. Preliminary work showed that explants could be successfully transferred to liquid culture medium and grown with no visible perturbations to plant morphology. Plants were added to 10 ml liquid C-fern medium pH 7.1 in 25-ml culture flasks either with no supplements or with one of the following: dimethyl sulphoxide (DMSO) only; 10 μM NPA; 40 μM NPA, 20 μM TIBA; 100 nM 1-naphthaleneacetic acid (NAA); 40 μM NPA + 100 nM NAA; 1 μM benzyl amino purine (BAP); 500 nM BAP + 100 nM NAA (all dissolved in DMSO). The auxin analogue NAA was used instead of IAA as IAA is not stable when exposed to light. Plants were grown for 2 months with solutions renewed every 2 wk.

Auxin transport measurements

Sections of young stem were cut and the 5-mm apical region was removed, leaving 20-mm sections with a minor branch approximately midway between each cut end. Auxin transport assays were modified from the protocol of Lewis & Muday (2009). Segments were placed in 1.5-ml tubes with either the apical or basal end in contact with 40 μl of working solution in the base of the tube (Fig. 1a). Working solutions comprised 100 nM [3-H]IAA (from a stock solution of 37 MBq ml−1 dissolved in ethanol; Perkin-Elmer, Waltham, MA, USA), 100 nM [3-H]IAA + 10 μM NPA or 100 nM [3-H]benzoic acid ([3-H]BA; from a stock solution of 37 MBq ml−1 dissolved in ethanol; American Radiochemicals, St Louis, MO, USA) as a diffusion control. All solutions were adjusted to give equal concentrations of DMSO and ethanol, and were made up to 1 ml with 2-morpholinoethanesulphonic acid (Mes) solution, pH 5.6 (Sigma). Experiments were left at 23°C overnight in a glass container with wet tissues in the bottom. On removal, stem segments were washed in water and blotted with tissue paper. Five-millimetre segments from the end of the stem that had not been in contact with the test solution were taken for analysis (Fig. 1a). In addition, 5 mm of the side branch were removed for analysis (Fig. 1a). Each 5-mm segment was added to 5 ml of Ultima Gold Scintillation Cocktail (Perkin-Elmer) in a scintillation vial and left to diffuse for at least 2 h. Counts per minute were measured three times for 3 min each in a Beckman) LS 6500 scintillation counter, and an average of the three readings was calculated. The efficiency of the scintillation counter was estimated at 60% from standards of 20 μl of test solution. Counts per minute were converted to attomoles as in Lewis & Muday (2009). Experiments were repeated three times to give a total of 33 replicates.

image

Figure 1. Polar auxin transport (PAT) occurs basipetally in shoots and is sensitive to 1-n-naphthylphthalamic acid (NPA). (a) Schematic diagram of experimental methodology. Shoots were immersed in auxin at the apical or basal end. Segments from the opposite end of the shoot and a branch were sampled for analysis. (b) Graph showing quantities of tritiated auxin detected in Selaginella kraussiana shoot segments after immersion in 100 nM tritiated auxin. Values are the mean of 33 biological replicates. Bars represent ± 2SE. Only basipetal transport of auxin resulted in quantities significantly larger than diffusion controls. BA, benzoic acid.

Download figure to PowerPoint

Measurements of plant morphology

Overall plant morphology was recorded and photographed. Plants were then fixed in 50% ethanol, 7% acetic acid, 43% water. Mature leaves were removed, cleared in a graded ethanol series and immersed in 1 : 1 ethanol : histoclear. Leaves were then mounted on slides in DPX mounting medium and viewed under UV light using a Leica DRMB microscope; Leica, Wetzlar, Germany). Leaf and vein measurements were made from photographs using ImageJ software (http://rsbweb.nih.gov/ij/index.html).

Scanning electron microscopy

Plants were dehydrated in a graded ethanol series, dried in an AutoSamdri 815 critical point drier (Tousimis, Rockville, MD, USA), mounted, gold/palladium sputter-coated (Polaron Equipment, Watford, Hertfordshire, UK) and viewed in a JEOL JSM 5510 SEM (JEOL, Welwyn Garden City, Hertfordshire, UK).

Statistics

The bar graph in Fig. 1(b) was produced using Microsoft Excel. Statistical analyses were conducted and boxplots were produced using the R software package (http://www.r-project.org/).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

PAT is basipetal and sensitive to NPA

Selaginella kraussiana shoots are dorsiventral and unequally dichotomizing, with branches formed after the production of six (minor branch) or eight (major branch) leaf pairs. To examine auxin flux in Selaginella shoots, auxin movement was quantified basipetally, acropetally and into a branch (Fig. 1a). Figure 1(b) demonstrates that the average amount of [3-H]IAA transported basipetally was c. 10-fold greater than the amount of [3-H]IAA transported acropetally or the amount of [3-H]BA diffusing through the tissue. The quantity of auxin transported basipetally was significantly different (< 0.001) from control treatments and from acropetal transport. Importantly, the quantities of auxin in the acropetal transport study and in the minor branch samples were not significantly different from diffusion controls. As such, the transport of auxin appears to be exclusively basipetal. A concentration of 10 μM NPA was sufficient to completely inhibit PAT (Fig. 1b), just as it is in Arabidopsis (Lewis & Muday, 2009). The complete loss of PAT on treatment with the auxin efflux inhibitor NPA (Fig. 1b) suggests that the molecular components mediating PAT in Selaginella are similar to those in seed plants.

Loss of PAT perturbs leaf venation

To investigate how PAT regulates vascular development in Selaginella, auxin transport inhibitors (NPA and TIBA) and the auxin analogue NAA were applied to growing shoots in liquid culture. In Selaginella leaves, a single vein extends into the leaf to c. 80% of the leaf length (Fig. 2a). Leaves of control plants treated with DMSO were identical to leaves of untreated plants (Fig. 2b), and those treated with NAA were identical to DMSO controls (Fig. 2c). In plants treated with NPA, however, the leaf vein appeared thicker than normal and tracheids were disorganized (Fig. 2d). In a few extreme cases, a branched vein was observed (not shown). Transverse sections of the veins revealed that untreated or DMSO-treated veins comprised fewer than 10 tracheids in diameter (Fig. 2g), whereas veins of NPA-treated leaves comprised c. 50 tracheids in diameter (Fig. 2h). A similar increase in size of the stem vasculature, as a result of an increase in tracheid diameter, was also observed in NPA-treated plants (Fig. 2i,j). Plants treated with NPA + NAA showed a similar increase in tracheid diameter (not shown). In plants treated with NPA or TIBA, the length of the vascular strand relative to the length of the leaf was reduced significantly (in all cases, < 0.02) when compared with untreated or DMSO-treated plants (Fig. 2k). No effect was seen on ligule morphology in any of the treatments. These data suggest that the development of leaf venation and stem vasculature in Selaginella is regulated by PAT.

image

Figure 2. Loss of polar auxin transport (PAT) perturbs leaf venation. (a–f) Cleared ventral leaves of Selaginella kraussiana under UV light showing veins of untreated plants (a), and of plants treated with dimethyl sulphoxide (DMSO) (b), 100 nM 1-naphthaleneacetic acid (NAA) (c), 40 μM 1-n-naphthylphthalamic acid (NPA) (d), 10 μM NPA (e) and 20 μM 2,3,5-triiodobenzoic acid (TIBA) (f). Arrows show point at which vein terminates. Bar, 500 μm. (g, h) Transverse sections of leaf vein treated with DMSO (g) and 40 μM NPA (h). Note increased number of tracheids in (h). Bar, 20 μm. (i, j) Transverse sections of stem vasculature treated with DMSO (i) and 40 μM NPA (j). Note larger tracheids in (j). Bar, 40 μm. (k) Boxplot showing length of vein relative to length of the leaf; the box defines the interquartile range with the whiskers defining the minimum and maximum values and the dots representing outliers. Plants treated with auxin efflux inhibitors have shorter veins.

Download figure to PowerPoint

Perturbed auxin homeostasis affects phyllotaxy

To examine how auxin influences Selaginella shoot development, excised shoots were grown in sterile liquid culture supplemented with NPA and/or NAA. Untreated shoots and DMSO-treated control shoots produced dorsal and ventral leaves in pairs (Fig. 3a,b), in the same way as untreated plants grown on soil. The shoot apices of these plants were bilaterally symmetrical along the dorsiventral axis (Fig. 3b). By contrast, shoots treated with NPA produced ventral-like leaves that encircled half of the stem in an alternate pattern (Fig. 3c). The morphology of the NPA-treated shoot apex was also altered, having a flattened appearance (Fig. 3d, arrow). Shoots showing altered structure were more frequent on treatment with 40 μM NPA than with 10 μM NPA (Fig. 3i). Thus, NPA treatment disrupts Selaginella shoot structure, in a dose-dependent manner, at concentrations similar to those that disrupt plant structure in seed plants (e.g. Reinhardt et al., 2000; DeMason & Chawla, 2004; Larsson et al., 2008). Interestingly, plants that were treated with a combination of 40 μM NPA and 100 nM NAA displayed normal leaf placement (Fig. 3e,f), suggesting that auxin supplementation can prevent NPA-mediated disruption of dorsiventrality. Treatment of shoots with 100 nM NAA appeared to have no effect on shoot structure (Fig. 3g,h). These data suggest that auxin transport regulates leaf patterning and placement in Selaginella, but does not promote leaf initiation.

image

Figure 3. Perturbed auxin homeostasis affects phyllotaxy. (a–h) Scanning electron microscopy (SEM) images of Selaginella kraussiana shoots and apices (arrows) treated with dimethyl sulphoxide (DMSO) (a, b), 40 μM 1-n-naphthylphthalamic acid (NPA) (c, d), 40 μM NPA + 100 nM 1-naphthaleneacetic acid (NAA) (e, f) and 100 nM NAA (g, h). Bars: (a, c, e, g) 500 μm; (b, d, f, h) 50 μm. (i) Table showing number of shoots with abnormal or normal phyllotaxy.

Download figure to PowerPoint

Loss of PAT disrupts meristem maintenance

Shoots treated with NPA for at least 2 months often developed a flattened shoot apex and growth terminated (Fig. 4a,b). This result implies that a reduction in auxin transport ultimately causes meristem identity in apical cells to be lost. Interestingly, although the addition of NAA together with NPA rectified leaf placement (Fig. 3e), apical growth still terminated (Fig. 3f). In shoots treated with NPA plus NAA, however, although the apical cells ceased to proliferate, the shoot apex appeared to be morphologically normal (Fig. 4c,d). These results suggest that PAT is necessary to maintain indeterminacy of the apical meristem and that a high auxin concentration at the apex is required for dorsiventral patterning.

image

Figure 4. Loss of polar auxin transport (PAT) disrupts apical meristem maintenance. (a) Selaginella kraussiana shoot apex treated with 40 μM 1-n-naphthylphthalamic acid (NPA) terminating in a leaf. Presumed remnants of the shoot apex at arrow. Bar, 500 μm. (b) Close-up of a branching apex of a plant treated with NPA. The meristem in the left branch has terminated (white arrow) following the production of a single leaf (arrowhead); the right branch still has a visible meristem (white arrow) with two leaves in abnormal positions (arrowheads). Bar, 100 μm; (c) Terminating shoot apex (white arrow) treated with 40 μM NPA + 100 nM 1-naphthaleneacetic acid (NAA). Bar, 100 μm. (d) Close up of an apex (white arrow) treated with 40 μM NPA + 100 nM NAA; the shoot has ceased to grow, but the structure of the shoot is intact and leaves have initiated in the correct positions (arrowheads). Bar, 100 μm.

Download figure to PowerPoint

Loss of PAT perturbs leaf shape

To ascertain whether auxin influences leaf shape, the length and width of mature leaves were measured. Leaves of plants treated with NPA were broader than those of control plants with a significantly smaller (< 0.001) length to width ratio (Fig. 5a). No significant difference (P = 0.66) was found between the length to width ratios of leaves treated with 40 or 10 μM NPA. Because treatment with NPA causes leaves to occupy a larger region of the stem (Fig. 3d), the change in shape may be a direct consequence of perturbed leaf initiation processes. However, an NPA-mediated reduction in leaf length was also observed (Fig. 5b). Interestingly, TIBA treatment also caused a significant difference in leaf length, but not a significantly different length to width ratio. This difference could be a result of differences in strength and/or binding properties of the two inhibitors (Depta & Rubery, 1984). Either way, these results demonstrate that normal PAT is required to maintain correct leaf shape.

image

Figure 5. Loss of polar auxin transport (PAT) perturbs leaf shape. (a) Boxplots showing the length to width ratio of Selaginella kraussiana ventral leaves under different treatments. The box defines the interquartile range with the whiskers defining the minimum and maximum values and the dots representing outliers. Plants treated with 1-n-naphthylphthalamic acid (NPA) showed a decreased length to width ratio. (b) Boxplots showing the lengths of ventral leaves under different treatments. Plants treated with NPA had shorter leaves. NAA, 1-naphthaleneacetic acid; TIBA, 2,3,5-triiodobenzoic acid.

Download figure to PowerPoint

Exogenously applied auxin perturbs root branching but not shoot branching

To ascertain whether auxin plays a role in the regulation of dichotomous branching, shoot and root structures were examined. In untreated and DMSO-treated control plants, roots under 50 mm in length were either unbranched or branched once (Fig. 6a,b). By contrast, roots of plants treated with 100 nM NAA were highly branched (Fig. 6c) and roots of plants treated with NPA developed callus-like tissue at the root tip (Fig. 6d). These observations suggest that high auxin concentrations promote dichotomous branching of the root. Because the ratio of auxin to cytokinin is important in the regulation of root development and root branching in angiosperms (Laplaze et al., 2007; reviewed in Moubayidin et al., 2009), the effect of addition of the cytokinin analogue, BAP, together with NAA was examined. Figure 6(e) shows that the addition of BAP restored normal morphology, suggesting that the auxin to cytokinin ratio regulates dichotomous branching in the root.

image

Figure 6. Exogenously applied auxin perturbs root, but not shoot, branching. (a–e) Typical Selaginella kraussiana root tips of untreated plants and plants treated with dimethyl sulphoxide (DMSO) (a,b), 100 nM 1-naphthaleneacetic acid (NAA) (c), 40 μM 1-n-naphthylphthalamic acid (NPA) (d) and 100 nM NAA + 500 nM benzyl amino purine (BAP) (e). Bar, 500 μm. (f) Untreated shoots with anisotropic branching after the production of six to eight leaf pairs. Bar, 3 mm. (g, h) Excised shoots treated with DMSO (g) or 40 μM NPA (h). Shoots extend for over 12 leaf pairs without branching (arrows). Bar, 3 mm. (i) Table showing percentage of plants cultured in different media exhibiting reduced branching frequency. For each group, = 17. No treatment ameliorated the effect of excision.

Download figure to PowerPoint

In Selaginella, nonphotosynthetic rhizophores are produced at shoot branch points and convert into a root after a few millimetres. Previous studies have shown that treatment of Selaginella plants with NPA results in the conversion of these aerial root-like structures into a shoot-like organ (Wochok & Sussex, 1976). Under our experimental conditions, such effects were observed in two of 20 plants examined (data not shown).

Selaginella shoots normally branch after the production of six to eight leaf pairs (Fig. 6f). However, when shoot tips are excised, branching is temporarily stalled and up to 20 leaf pairs can be formed without branching (Fig. 6g). To discover whether auxin plays a role in the regulation of branching, excised plants were placed into solutions containing NPA, NAA or a combination of both. Branching patterns were observed after 2 months, at which point plants treated with NPA showed a mixture of shoots that had terminated and shoots that were still growing. In all cases, growing shoots exhibited the same branching pattern as untreated controls (Fig. 6h,i). These data suggest that auxin may not play a role in the regulation of dichotomous shoot branching.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Auxin transport mechanisms are conserved in lycophytes and euphyllophytes. To provide a comparative framework for the assessment of the role of auxin in the development of early diverging vascular plants and seed plants, we investigated the mechanisms of auxin transport in the lycophyte Selaginella kraussiana and characterized the morphological consequences of the perturbation of transport processes. We first showed that, as in seed plants, PAT is exclusively basipetal and is sensitive to known PAT inhibitors (Fig. 1). The finding that PAT is exclusively basipetal in Selaginella contrasts with previous reports which concluded a 2 : 1 basipetal to acropetal ratio in Selaginella shoots (Wochok & Sussex, 1973). However, given that these earlier studies did not demonstrate the extent of diffusion using auxin efflux inhibitors or other organic acids (see Lewis & Muday, 2009 for a discussion of such methods), the results may not be reliable. Notably, the quantity of auxin taken up into Selaginella shoots was c. 100 times lower than that in angiosperms, such as maize or Arabidopsis (Fig. 1; e.g. Rashotte et al., 2000; Poli et al., 2003; Lewis & Muday, 2009). It is unclear whether the smaller quantities reflect differences in the experimental method, size of the tissue, amount of vascular tissue or reduced efficiency of auxin transport systems within Selaginella. However, the data suggest that the molecular mechanisms underlying PAT are likely to be conserved across all vascular plants. Consistent with this suggestion, eight PIN1-like sequences are present in the Selaginella genome (Banks et al., 2011).

The evolution of PAT mechanisms

The relevance of auxin transport mechanisms before the evolution of vascular plants is uncertain. Current evidence suggests that basipetal PAT occurs in moss seta, but studies reporting the presence or absence of acropetal transport are conflicting (Poli et al., 2003; Fujita et al., 2008). Notably, genes orthologous to PIN1 and to auxin response genes are found in the genome of Physcomitrella patens (Křeček et al., 2009; Paponov et al., 2009), indicating that at least some of the molecular components mediating auxin responses are present in nonvascular plants. Furthermore, NPA causes the disruption of moss embryo development, indicating that PAT in moss is sensitive to auxin efflux inhibitors (Fujita et al., 2008). Poli et al. (2003) suggested that both liverworts and mosses show partial sensitivity to auxin efflux inhibitors, but without diffusion controls their results are hard to interpret. Table 1 summarizes what is known of the morphological effects of reduced auxin transport. Although information is lacking or incomplete for groups other than angiosperms, evidence suggests that PAT plays a ubiquitous role in patterning sporophyte architecture and in regulating vascular tissue development. Intriguingly, a recent study has demonstrated that NPA-sensitive PAT occurs in single cells of Chara corralina at a rate higher than would be expected from diffusion alone (Boot et al., 2012). Putative PIN-like gene sequences have also been identified in streptophyte algae (De Smet et al., 2011). In combination with the data presented here, these findings suggest that basipetal auxin transport may have been co-opted from algal ancestors during the early evolution of the multicellular sporophyte, and that mechanisms were elaborated coincident with the evolution of complex branching systems.

Table 1. Comparison of the effects of polar auxin transport (PAT) inhibitors on plant structure in each of the major groups of land plants
 LiverwortsMossesHornwortsLycophytesMonilophytesGymnospermsAngiosperms
  1. Based on Wochok & Sussex (1976); Liu et al. (1993); Reed et al. (1998); Mattsson et al. (1999); Sieburth (1999); Reinhardt et al. (2000); DeMason & Chawla (2004); Hou et al. (2004); Gregorich & Fisher (2006); Scarpella et al. (2006); Barkoulas et al. (2008); Larsson et al. (2008); Müller & Leyser (2011).

  2. ?, information not known; –, not applicable, organ not present.

Shoot architectureGametophyte shoot??No effect on Physcomitrella patens gametophyte leafy shoot Leaves are produced, meristem arrest in Selaginella??Leafless shoots, increased branching, meristem continues growth in Arabidopsis
Leaf morphologyDecreased length to width ratio in Selaginella??Shorter petiole, decreased length to width ratio in Arabidopsis. Loss of leaflet initiation in pea and Cardamine hirsuta
Root architectureCallus formation on root tips of SelaginellaDecreased growth rate but no effect on root production in Ceratopteris richardii?Inhibition of lateral root formation in Arabidopsis
Vascular tissueDisrupted vascular development and leaf venation, larger tracheids in Selaginella??Disrupted vascular development and leaf venation, increased number of stem vascular strands in Arabidopsis
OtherEmbryo??Pertubations in diploid embryo of P. patensGametophyte, embryo??Rhizophores redetermined to shootsDisrupted gametophyte development – extra meristem formation in Ceratopteris richardiiAbnormal embryo development in Picea abiesAbnormal embryo development in Arabidopsis

Auxin mediates dichotomous branching in roots but not shoots of Selaginella

As with all lycophytes, Selaginella shoots and roots branch dichotomously, a trait probably inherited from ancestral vascular plants (Kenrick & Crane, 1997). By contrast, seed plant shoots branch through axillary buds at nodes, and roots branch endogenously from the pericycle. Auxin is a key regulator of branching in both roots and shoots of seed plants. In shoots, auxin is produced in leaves and is transported basipetally, repressing axillary bud outgrowth and maintaining apical dominance (reviewed in Müller & Leyser, 2011). Auxin transport thus provides a ‘top-down’ control of axillary branching in flowering plants. By contrast, the loss of regular branching patterns in excised shoots of Selaginella (Fig. 6) hints at a ‘bottom-up’ control of dichotomous branching in this species. Given that auxin transport is exclusively basipetal (Fig. 1b) and that exogenous auxin cannot ameliorate the excision-mediated disruption of shoot branching (Fig. 6a–c), it seems unlikely that auxin is a regulator of dichotomous shoot branching.

In Arabidopsis, lateral root development is promoted by auxin and inhibited by cytokinin (Laplaze et al., 2007). Similarly, exogenous auxin promotes and cytokinin inhibits the dichotomous branching of Selaginella roots (Fig. 6c,e). From a developmental perspective, the dichotomous branching of Selaginella root tips is fundamentally different from endogenous lateral root formation in angiosperms, yet these data suggest that antagonistic roles of auxin and cytokinin may underpin both mechanisms.

Conserved role of PAT in boundary formation at the shoot apex

Apices of Selaginella are bilaterally symmetrical with the long edge of the apex parallel to the dorsiventral axis. A pair of apical initials ultimately gives rise to pairs of dorsal and ventral leaves that are normally produced on either side of this axis (Harrison et al., 2007). In plants treated with NPA, the shoot apex loses this morphology and leaves are produced from broad regions of the apex, crossing the former dorsiventral axis. Each of these large leaves encompasses the zone normally occupied by one dorsal and one ventral leaf, plus the associated stem tissue. Boundaries defining dorsiventral patterning and leaf phyllotaxy, therefore, fail to be delimited when PAT is inhibited (Fig. 3c). Similar effects have been reported in seed plants, in which multicellular, layered meristems pattern phyllotaxy. For example, spruce embryos treated with NPA grow with fused cotyledons and perturbed meristem patterning (Larsson et al., 2008), and Brassica embryos treated in a similar way phenotypically resemble the embryos of Arabidopsis pin-1 mutants which have fused cotyledons (Liu et al., 1993). In combination, these observations suggest that the role of auxin in maintaining the boundaries between meristem and leaf zones in the apex is conserved between lycophytes and seed plants.

Distinct roles for auxin in the initiation of leaves in lycophytes and euphyllophytes

Arabidopsis inflorescences deficient in the auxin efflux carrier protein PIN1, and tomato shoots cultured on auxin efflux inhibitors, produce shoots lacking leaves (Reinhardt et al., 2000, 2003). Thus, PAT is essential for leaf initiation in angiosperms, but not for shoot meristem maintenance. By contrast, leaves are formed in NPA-treated Selaginella plants, but PAT inhibition ultimately terminates shoot growth. PAT is thus required for proliferation of the two apical initials in the Selaginella apex, but not for leaf initiation, suggesting that different mechanisms regulate shoot growth in lycophytes.

In angiosperms, the failure to initiate leaves when PAT is inhibited can be overcome by the addition of high concentrations of auxin together with NPA. In these circumstances, a greater number of cells than normal are recruited to form leaves, consistent with the idea that auxin maxima at the apex determine leaf positions (Reinhardt et al., 2000). Paradoxically, treatments with NPA and NAA in Selaginella reduce the number of cells recruited into leaves and hence restore normal dorsiventral patterning and leaf positioning (Fig. 3). As such, we suggest that the disruption of phyllotaxis in NPA-treated Selaginella results from perturbed boundary formation, as opposed to a failure to initiate leaves per se (Reinhardt et al., 2003; Lee et al., 2009). By extension, these results imply that auxin does not play a direct role in leaf initiation in Selaginella.

Although the addition of auxin negates the effects of inhibited PAT on dorsiventrality in the shoot of Selaginella, it does not prevent the termination of shoot growth. The morphology of the terminated shoot, however, differs in the presence and absence of auxin. In NPA alone, cells in the apex are all recruited into leaves, whereas, in NPA plus NAA, apices retain normal morphology, but the apical initials stop proliferating. To explain these results, we propose that meristematic activity is dependent on a differential gradient of auxin between the apical cells and their derivatives. This gradient can be perturbed when auxin is either too high (NPA and NAA) or too low (NPA) in the apical cells. As such, both treatments lead to meristematic arrest.

Conserved role for PAT in vascular differentiation, but not in venation patterns

As in angiosperms, the inhibition of PAT in Selaginella results in disrupted leaf venation (Fig. 2; Mattsson et al., 1999; Sieburth, 1999; Scarpella et al., 2006). In Arabidopsis and other angiosperms, the addition of PAT inhibitors results in thicker veins, isolated tracheids and leaf traces that do not connect with the stem vasculature (Mattsson et al., 1999; Sieburth, 1999). In a similar manner, PAT inhibition in Selaginella results in a thicker vein comprising an increased number of tracheids (Fig. 2, Table 1). However, although NPA-treated Arabidopsis leaves display an increase in venation at leaf margins and a reduction in venation more basally (Mattsson et al., 1999; Sieburth, 1999), the single vein in Selaginella leaves is thicker at the leaf base and does not reach the margin at the leaf tip. Mattsson et al. (1999) noted that the perturbed venation pattern in Arabidopsis suggests that a vascular inducing signal emanates from the apex of the leaf, a suggestion supported by the tracking of auxin responses in NPA-treated leaves (Scarpella et al., 2006). In this context, despite basipetal auxin transport through the stem, leaf venation in Selaginella may be patterned by transport from the stem towards the leaf apex. This scenario predicts that, with NPA treatment, auxin would become increasingly confined to the stem vasculature and would, in turn, promote cell expansion in the procambium. Consistent with this hypothesis, tracheid diameters are larger in NPA-treated stems and leaves than in untreated or DMSO-treated plants, and leaf veins do not extend to the leaf tip (Fig. 2). Taken together, these data suggest that vascular tissue development, but not venation patterning, are similarly regulated by PAT in lycophytes and euphyllophytes.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was funded by a European Research Council Advance Investigator Grant (EDIP) to J.A.L. We thank Hugh Dickinson for assistance with scanning electron microscopy, Nick Kruger and Chris Snowden for help and advice with auxin transport experiments, Julie Bull for plant propagation assistance and John Baker for photography. Guillaume Chomicki-Bayada assisted with preliminary data collection. Jill Harrison, Ottoline Leyser and Liam Dolan provided valuable discussions.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References