CAPRICE positively regulates stomatal formation in the Arabidopsis hypocotyl

Abstract

In the Arabidopsis hypocotyl, stomata develop only from a set of epidermal cell files. Previous studies have identified several negative regulators of stomata formation. Such regulators also trigger non-hair cell fate in the root. Here, it is shown that TOO MANY MOUTHS (TMM) positively regulates CAPRICE (CPC) expression in differentiating stomaless-forming cell files, and that the CPC protein might move to the nucleus of neighbouring stoma-forming cells, where it promotes stomata formation in a redundant manner with TRIPTYCHON (TRY). Unexpectedly, the CPC protein was also localized in the nucleus and peripheral cytoplasm of hypocotyl fully differentiated epidermal cells, suggesting that CPC plays an additional role to those related to stomata formation. These results identify CPC and TRY as positive regulators of stomata formation in the embryonic stem, which increases the similarity between the genetic control of root hair and stoma cell fate determination.

Introduction

The growth and development of multicellular organisms depend on the generation of cells with distinct cell fates. The plant epidermis has been used as an excellent model system to study cell fate specification. In Arabidopsis thaliana, epidermal cells in the hypocotyl are organized in files that run parallel to the long axis of the seedling. Files consisting of non-protruding cells are placed outside two cortical cell files, while that consisting of protruding cells overly a single cortical cell file.^1,2 Stomata only develop in epidermal files located outside two cortical cell files.^2,3 They are not present at germination and begin to form after 4 days, reaching a maximal density after 8 days postgermination.²

Some of the genes controlling stomata formation in the hypocotyl have been defined. GLABRA2 (GL2) encodes a homeodomain-leucine zipper protein,^4,5 and it is a repressor of stomata formation in cell files located outside a single cortical cell file,^2,3 where it is preferentially expressed.³ TRANSPARENT TESTA GLABRA (TTG) also negatively regulates the formation of ectopic stomata.^2,3 TTG encodes a small protein with WD40 repeats,⁶ and its expression pattern remains unknown. The WEREWOLF (WER) gene encodes a MYB protein and it is also preferentially expressed in epidermal cells files placed outside a single cortical cell file, where it negatively regulates stomatal cell fate.⁷ In addition, TTG and WER are positive regulators of the GL2 gene.^3,7

The strong reduction in the number of stomata in the hypocotyl of Arabidopsis plants overexpressing a basic helix-loop-helix (bHLH) transcriptional activator from maize, named R, suggests that an R-like activity negatively regulates stomatal formation in the embryonic stem.² In accordance with such a role, 35S:R plants exhibit GL2 expression in all cell files of the hypocotyl epidermis.³ Interestingly, mutations in both GLABRA3 (GL3) and ENHANCER OF GLABRA3 (EGL3), two Arabidopsis bHLHs closely related to R,^8,9 trigger the development of stomata in files overlying a single cortical cell file,¹⁰ revealing that both GL3 and EGL3 are the bHLH repressors of stomata formation in the embryonic stem. The yeast two-hybrid assay has shown that these bHLH proteins physically associate with both TTG and WER.^8,9,11

Current molecular, genetic and biochemical data support the idea that a complex consisting of WER, TTG, GL3 and/or EGL3 represses stomata formation by promoting GL2 expression in epidermal cell files overlying a single cortical cell file.^12,13 A similar mechanism guides non-hair specification in the root.^12–15 Here, I contribute to the model by exploring the possible role of a positive regulator of root hair formation, the single-repeat MYB protein CAPRICE (CPC), during stomata formation. Comparison of the CPC expression pattern, the CPC protein localization, and the cpc phenotype supports the proposal that CPC, whose expression is positively regulated by TOO MANY MOUTHS (TMM), represses stomaless cell fate perhaps through a lateral inhibition mechanism. The try and try cpc analysis also shows that the TRY and CPC genes, which are very similar in sequence,¹⁶ act in a redundant fashion to guide stomatal cell fate in the hypocotyl. In addition, these studies suggest that both CPC and TRY might play additional roles to those related to stoma/hair pattern formation.

Results

Stomatal pattern in the hypocotyl epidermis of wild type, cpc, try and try cpc mutants.

Examination of epidermal imprints with Nomarsky optics of wild type plants (Ler and Ws ecotypes) showed that non-protruding cell files are separated by 1 or 2 protruding cell files (Fig. 1A),¹ although 0 or 3 intervening protruding cell files (or vice versa) were also rarely observed (Fig. 1B). Occasionally, a single cell file continued with two parallel ones (Fig. 1C). The three files were identical (protruding or non-protruding) or one of the parallel cell files exhibited a different morphology to the other two cell files, which were identical. Also rarely, some files consisted of a column of protruding cells in contact with another of non-protruding ones (Fig. 1D).

Figure 1.

Hypocotyl epidermis of wild type, cpc, try and try cpc mutants. (A) In the wild type (Ws ecotype), cpc and try, protruding cell files are separated by 1 or 2 non-protruding ones. Stomata, whose number in cpc and try is reduced of that the corresponding wild type, are preferentially placed in non-protruding cell files. The try cpc double mutant has no stoma, and all epidermal files consist of protruding cells. (B–D) Rare events in wild type, cpc and try mutants. (B) Non-protruding cell files separated by more than 2 intervening protruding ones (asterisks). (C) A single file (asterisk) that continues in two parallel ones. (D) Files consisting of a column of protruding cells in contact with another of non-protruding ones (asterisks). All images represent agar imprints from 8-day-old seedlings. The cpc mutant is in the Ws ecotype; try is in Ler; pictures from Ler ecotype have been omitted, but they are similar to those from Ws. Bar, 100 µm.

As previously stated,^1,2 it was also confirmed that non-protruding cell files are located over two cortical cell files and can differentiate stomata in the upper 2/3 of the hypocotyl (Table 1; Fig. 1A), while protruding cell files are placed over a single cortical cell file and do not form stomata (Fig. 1A; Table 1). Although most of the stomata were located in cell files overlying two cortical cell files (composed of non-protruding cells) (Table 2), not all the cell files overlying two cortical cell files and being composed of non-protruding cells developed stomata. In fact, around half of the non-protruding epidermal cell files developed paired guard cells.

Table 1.

Phenotypic characteristics in the epidermis of the hypocotyl of wild type, cpc and try mutants

	Genetic background
Plant variable	Ws	Ler	cpc	try
Non-protruding cell files overlying two cortical cell files (%)	90	90	80	80
Protruding cell files overlying a single cortical cell file (%)	80	90	80	90

One randomly selected file per seedling (N = 10) was analysed. The cpc mutant is in Ws ecotype; try is in Ler.

Table 2.

Number and position of stomata in the hypocotyl of the wild type and several mutant backgrounds

	Genetic bacground
Plant variable	Ws	Ler	cpc	try	try cpc	CPC:CPC:GFP in cpc
Number of stomata per hypocotyla	17.7 ± 4.3	15.3 ± 2.6	10.5 ± 2.5	9.3 ± 2.5	0 ± 0	16.2 ± 2.7
Non-ectopic stomataa (%)	(230)b	(82)b	(145)b	(64)b	-	(95)b
	87.4 ± 11.6	90.5 ± 10.3	92.4 ± 12.6	74.2 ± 11.4		85.7 ± 12.1

^aMeans ± s.d.

^bNumber of stomata scored. N = 15. The cpc mutant is in Ws ecotype; try is in Ler. Non-ectopic stomata are those placed in epidermal files above two cortical cell files. Randomly selected stomata were scored from every seedling to calculate the percentage of non-ectopic stomata.

The cellular dimorphism in the hypocotyl epidermis of wild type plants was also apparent in the cpc (in Ws ecotype) and try (in Ler ecotype) mutants: files composed of non-protruding cells that differentiate stomata alternated with those consisting of protruding cells in which stomata do not develop (Fig. 1A). The rare events described in wild type plants (see first paragraph) were also observed in the cpc and try mutants (Fig. 1B–D). Both mutants displayed a reduced number of stomata per hypocotyl when compared with wild type plants (Table 2; Student's t-tests, p < 0.0001). In cpc, the position of these stomata did not differ from those developed in wild type plants: most of them were placed in cell files overlying two cortical cell files, which were composed of non-protruding cells (Table 2; χ2-test, p > 0.05). The try mutant showed a reduced number of non-ectopic stomata relative to wild type plants (Table 2; χ2-test, p < 0.0001). As in the wild type, stomata in cpc and try were located at the upper 2/3 of the hypocotyl. In addition, only some non-protruding epidermal cell files in the cpc and try mutants exhibited stomata. The try cpc double mutant exhibited no stoma, and all files from the apical to the basal part of the hypocotyl consisted of protruding cells (Table 2; Fig. 1A).

Expression pattern of CPC during hypocotyl epidermis development.

Histochemical staining of CPC:GUS plants showed vertical columns of strongly stained cells separated by columns of weakly stained cells in the hypocotyl epidermis at all developmental stages (3-, 5- and 8-day-old seedlings) (Fig. 2B). The intensity of GUS expression was similar and preferentially restricted to the apical portion of the hypocotyl, also in all developmental stages (Fig. 2A).

Figure 2.

Expression pattern of CPC:GUS during hypocotyl epidermis development. (A) GUS-expressing cells are preferentially located at the upper part of the hypocotyl in all developmental stages examined. (B) Cells in the strongly GUS-expressing files are longer than cells in the weakly expressing ones. It is known that files that do not develop stomata (and so consisting of protruding cells) differentiate longer cells.³ Bar, 5 mm in (A), and 125 µm in (B).

The detailed examination of these hypocotyls showed that the GUS gene is preferentially expressed in files characterized by the absence of stomata. By focusing alternatively protodermal and cortical cell files, it was determined that the strongly GUS-expressing cell files are located overlying a single cortical cell file. A lower but substantial signal of the GUS activity was also observed in epidermal cell files overlying two cortical cell files (Fig. 2B). An identical CPC promoter induction in stomaless-forming cell files of developing hypocotyls has been previously reported.¹⁷ The GUS activity in the epidermal cell files placed above two cortical cell files might reflect a low-level of the CPC promoter activity, or diffusion of the GUS product from the neighbouring and strongly stained stomaless-forming cell files. Strong GUS activity was also observed in paired guard cells.

Localization of the CPC protein during hypocotyl epidermis development.

In the hypocotyl epidermis, CPC is preferentially expressed in differentiating stomaless-forming cell files overlying a single cortical cell file (see previous section). To determine whether this CPC expression pattern overlaps with the localization of the CPC protein, we made use of a functional CPC:CPC:GFP construct (in Col-0 background), which was able to rescue cpc mutant (Table 2; Student's t-tests, p > 0.05). Seedlings from non-transgenic plants (Col-0 ecotype) were also analysed and used as negative controls for these experiments.

Confocal sections in the hypocotyl epidermal tissue of transgenic plants showed GFP fluorescence in all cell files and at every developmental stage (3-, 5- and 8-day-old seedlings) (Fig. 3A). GFP fluorescence was visualized in both the nucleus and surface of all epidermal cells (Fig. 3B). All epidermal nuclei were highly fluorescent but the intensity of the GFP signal in the cell surface was the strongest at the basal part of the hypocotyl, close to the root pole, and became diminished in the apical portion, where stomatal development takes place. This GFP-cell surface signal, which is due to the CPC cytoplasmic localization,¹⁸ moved towards the apical portion of the hypocotyl as seedlings developed (Fig. 3A).

Figure 3.

The CPC protein localization of CPC:CPC:GFP plants during hypocotyl epidermis development. (A) Epidermal confocal sections showing the strongest GFP signal in the lower portion of the hypocotyl at all developmental stages. Non-transgenic plant is 8-day-old; autofluorescence in all developmental stage was similar. (B) Overlaid epidermal confocal sections showing that the GFP protein localizes in the nucleus and cell surface of all epidermal cells. Owing to the weak GFP signal at the cell surface in the upper part of the hypocotyl, the GFP signal was amplified in such as part, so comparison of GFP signal cannot be made with these pictures. The red background in all the images is due to autofluorescence. Bar, 200 µm in (A), and 100 µm in (B).

Effect of both tmm-1 on CPC-promoter activity and try cpc on TMM-promoter activity.

Mutations in tmm-1, which encodes a LRR receptor-like protein without a cytoplasmic effector domain,¹⁹ also prevent stomata formation in the embryonic stem.²⁰ To determine whereas the CPC promoter induction is regulated by TMM, the CPC:GUS reporter gene fusion was introduced into the tmm-1 mutant background by genetic crosses. The hypocotyls of these tmm-1 CPC:GUS seedlings exhibited no GUS activity, showing that the tmm-1 mutation represses CPC promoter induction (Fig. 4A). GUS expression was detected in other organs as cotyledons or leaf primordium (Fig. 4B and C).

Figure 4.

Regulation of the CPC and TMM expression patterns. (A) The hypocotyls of tmm-1 CPC:GUS seedlings exhibit no GUS activity. GUS expression from tmm-1 CPC:GUS seedlings was detected in (B) the cotyledon and in (C) the leaf primordium. (D) GUS expression under the control of TMM promoter was observed in the stomatal precursor cells and immature stomata. (E) Seedlings containing both try cpc mutations and the TMM:GUS:GFP construct display no GUS activity in the hypocotyl epidermis. (F) In the cotyledon, the try cpc double mutants display TMM promoter activity in stomata precursor cells and immature stomata, similar to wild type plants. Bar, 200 µm in (A and E), 1,5 mm in (B and C), and 20 µm in (D and F).

The TMM promoter is induced in the stomatal lineage cells (precursors of meristemoid mother cells, meristemoids mother cells, meristemoids, guard mother cells and immature stomata (Fig. 4D).^19,21 Because hypocotyl epidermal cells of try cpc fail to enter into the stomatal pathway, it might be expected that TMM promoter is not induced in the hypocotyl epidermis of such a genetic background. Seedlings containing both try cpc mutations and the TMM:GUS:GFP construct displayed, as expected, no GUS activity in the hypocotyl epidermis (Fig. 4E). In the cotyledon, the try cpc double mutant displayed TMM promoter activity in meristemoids mother cells, meristemoids, guard mother cells and immature stomata, similar to wild type plants (Fig. 4F). This suggests that these single MYB proteins do not control TMM expression, but they affect indirectly the TMM promoter induction in the hypocotyl by repressing the entering into the stomatal pathway.

Discussion

CPC is a positive regulator of stomatal formation in the hypocotyl.

Stomatal development progresses from the upper to the basal part of the hypocotyl with no stomata formed in the basal third of the embryonic stem.² In addition, stomata only develop in epidermal cell files placed above two cortical cell files.^2,3 The cpc mutant exhibits a reduced number of stomata (about half of that the wild type), which are placed in the correct positions. GUS expression from CPC:GUS plants overlaps with the upper region in which stomatal formation occurs covering at least from the initiation of stomatal development (3-day-old plants) to the final stage in which all stomata are formed (8-day-old plants), but it is preferentially detected in cell files that do not form stomata. However, the CPC protein extends to the nucleus of all epidermal files, including those in which it is not strongly expressed such as those that form stomata. Together, this suggests that CPC protein (or mRNA) moves from epidermal cell files overlying a single cortical cell file to the nucleus of neighbouring cells placed above two cortical cell files, where it induces stoma formation.

The reduced number of stomata in the try mutant (more than half of that the corresponding wild type) and the absolute absence of such as cells in the try cpc double mutant indicate that these genes redundantly promote stomatal cell fate determination. In addition, the try mutant exhibits a reduced but substantial number of ectopic stomata, suggesting that TRY might also participate in stomaless-cell fate specification. These results on try stomatal phenotype differ from those previously reported, in which another try allele exhibited a stomatal phenotype similar to those of the corresponding wild type,² which indicates that the portion(s) of the protein affected in the mutant used in this work, and not those affected in the try mutant previously characterized, is essential during stomatal formation in the hypocotyl.

The presence of the CPC protein in the peripheral cytoplasm and the nucleus of fully differentiated epidermal cells at the basal part of the hypocotyl, where stomata development does not occur, led to speculation that CPC plays an additional role to those related to stomatal pattern formation. However, cells in the basal portion of the hypocotyl in the cpc mutant failed to show any obvious phenotype. Epidermal features such as the alternation of protruding cell files with non-protruding ones or the position of theses files relative to the cortical cell files (non-protruding cell files overlying two cortical cell files and protruding ones overlying a single one) are not affected by the cpc mutation. Different explanations can be advanced for this lack of an altered phenotype: (1) CPC might have a relatively minor role and/or might be only required under specific environmental conditions, such that mutation would cause a phenotype difficult to detect with the technology and/or under the growth conditions used in this work and/or, (2) CPC might be redundant, so that if mutated other gene(s) perform(s) the same function. Interestingly, although in either try and cpc protruding cell files alternate with non-protruding ones, in the try cpc double mutant all epidermal files consist of protruding cells (Fig. 1A), indicating that CPC and TRY together regulate such a phenotype. As the protruding versus non-protruding character is determined at the early stage of differentiation, it is likely that the CPC protein in the basal part of the hypocotyl participates in the maintenance of the non-protruding character. The localization of the CPC protein in the basal part of the hypocotyl also suggests that CPC also moves in a basal direction among cells of a given file. The fact that CPC and TRY together control al least two aspect of the epidermis differentiation (stomata formation and protruding character) suggests that these genes might regulate not only such as characters but also the identity of stomata forming-cell files.

CPC and epidermal cell differentiation in the Arabidopsis seedling.

Similar to stomata, root hairs develop in files placed outside two cortical cell files.^22–24 In addition, several lines of experimental evidence support the idea that a multimeric complex among WER, TTG, GL3 and/or EGL3 positively regulates GL2 expression in root and hypocotyl epidermal cells placed above a single cortical cell file, preventing both hair and stomatal cell fate determination.^12–15 Supporting that the same regulatory complex is responsible for regulating GL2 expression in the root and the hypocotyl, the same 500-bp fragment of the GL2 promoter is required for a proper GL2 expression pattern in the two organs.³ Moreover, the transcriptional regulation of GL2 by TTG and R depends on this fragment.³

But, how far does this similarity extend? CPC, TRY and ENHACER OF TRY AND CPC1 (ETC1), which encode single-repeat MYB proteins very similar in sequence, redundantly promote root hair formation.^16,17,25,26 The CPC expression pattern is preferentially restricted to the meristematic and elongation region of hairless cell files, but the CPC protein moves to the nucleus of hair-forming cells where it promotes root hair formation by repressing GL2 expression.^26,27 Although the cellular localization of TRY and ETC1 proteins is unknown, ETC1 (and perhaps TRY) is preferentially expressed in the non-hair cell files.^16,17 The repression of GL2 transcription seems to depend on the sequestration of GL3 and EGL3 by the single-repeat MYB proteins, which prevents the formation of the active complex (WER/GL3/EGL3/TTG) by inducing the formation of an inactive one (single-repeat MYB/GL3/EGL3/TTG) in hair-forming cells.^11–15

In the hypocotyl, CPC is expressed in epidermal cell files overlying a single cortical cell file, and this work suggests that the protein moves to the nucleus of those overlying two cortical cell files where it induces, in a redundant manner with TRY, guard cell formation. Thus, stomatal cell fate might also depend on the formation of an inactive complex among CPC (and perhaps TRY), TTG and EGL3 and/or EGL3 in cells from files overlying two cortical cell files. Interestingly, TMM positively regulates stomata formation by inducing CPC expression in the hypocotyl epidermal cells. Its possible role on root hair formation has not been investigated. Although, it is known that a receptor-like kinase, named SCRAMBLED, controls both CPC expression and root hair formation.²⁸

In spite of these parallels, the cell pattern resulting in the two embryonic organs is not similar. While all files overlying two cortical cell files differentiate hairs in the root,^22–24 only some of them produce stomata in the hypocotyl. In addition, while in a hair cell file all cells differentiate as root hairs,^22–24 in the hypocotyl only a few cells enter into the stomatal pathway. These differences might be explained by the existence of other negative regulators of stomatal cell fate that play no role in root hair differentiation.

Materials and Methods

Plant materials and growth conditions.

The cpc, try, try cpc, tmm-1, CPC:GUS and CPC:CPC:GFP strains of Arabidopsis thaliana have been previously described.^16,25,26,29 In stable transformation of Arabidopsis, previous studies have shown that this GFP stays in the cytosol.²⁶ The wild type strains used in this work were Landsberg erecta (Ler), Wassilewskija (Ws) and Columbia (Col-0).

Seeds were vernalized at 4°C for several days and surface-sterilized in 5% sodium hypochlorite. They were plated on Petri dishes containing Murashige and Skoog salts (Sigma) supplemented with 1% sucrose and solidified with 1% agar. Seedlings were germinated and grown on horizontally oriented dishes at 22°C in light (16-hours-light/8-hours-dark cycle).

Epidermal imprints.

Impressions of the epidermal surface of the hypocotyl were taken with 6% agarose as previously described.³⁰

Characterization of epidermal features.

To determine the number of stomata per hypocotyl, 8-day-old seedlings were mounted in slides, examined and digitised under Nomarsky optics with a Leica DC 300F camera attached to a Leica DMIRB inverted microscope. A total of 15 seedlings were used for each genetic background. Epidermal imprints of these seedlings, also under Nomarsky optics, were used to determine the type of epidermal file, protruding or non-protruding in which a given stoma was placed.

The localization of stomatal cells relative to the underlying cortical cells was determined by focusing alternatively on both epidermal and cortical focal planes. Stomata placed in epidermal files making contact with a single cortical cell file were defined as ectopics and those located in files that contacted two cortical cell files as nonectopics.^2,3 Data were obtained from 15 seedlings for each genetic background.

Results were evaluated by using χ2-test (95% confidence level) and Student's t-test (95% confidence level).

GUS staining.

For histochemical analysis of GUS expression, 3-, 5- and 8-day-old seedling harbouring the transgene were assayed for GUS activity as previously described.³¹ After histochemical GUS assays, seedling were cleared in a graded ethanol series and then examined by light microscopy with a Leica DC 300F camera attached to a Leica MZ6 stereomicroscope and to a Leica DMIRB inverted microscope.

GFP imaging of protein sub-cellular localization.

GFP fluorescence (and autofluorescence) from 3-, 5- and 8-day-old seedlings was monitorized with a DMIRB inverted Leica TCS SP2 confocal microscope using a 488 nm laser line of an Ar laser. For visualization of GFP, the emission window was set at 500–525 nm. For visualization of autofluorescence, the emission window was set at 580–690 nm. GFP fluorescence and autofluorescence acquisition was performed simultaneously using either the 10x/NA 0.30 PL fluotar or 40x/NA 1.25–0.75 PL APO (oil immersion) objectives. Optical sectioning along all cells was performed to capture the entire fluorescent signal. Sections were recorded with picture size of 1024 x 1024 pixels. Images were arranged and labelled using Adobe Photoshop 6.0.

Abbreviations

bHLH

basic helix-loop-helix

Col

columbia

CPC

caprice

EGL3

enhancer of glabra3

ETC1

enhacer of try and CPC1

GL2

glabra2

GL3

glabra3

Ler

landsberg erecta

TMM

too many mouths

TRY

triptychon

TTG

transparent testa glabra

WER

werewolf

Ws

wassilewskija