Plant Physiol. (1998) 117: 73-84

A Common Position-Dependent Mechanism Controls Cell-Type Patterning and GLABRA2 Regulation in the Root and Hypocotyl Epidermis of Arabidopsis¹

Chen-Yi Hung, Yan Lin, Meng Zhang, Susan Pollock, M. David Marks, and John Schiefelbein^*

Department of Biology, University of Michigan, Ann Arbor, Michigan 48109-1048 (C.-Y.H., Y.L., M.Z., J.S.); and Department of Genetics and Cell Biology, University of Minnesota, St. Paul, Minnesota 55108-1095 (S.P., M.D.M.)

	ABSTRACT

Top Abstract Introduction Methods Results Discussion References

A position-dependent pattern of epidermal cell types is produced during root development in Arabidopsis thaliana. This pattern is reflected in the expression pattern of GLABRA2 (GL2), a homeobox gene that regulates cell differentiation in the root epidermis. GL2 promoter::GUS fusions were used to show that the TTG gene, a regulator of root epidermis development, is necessary for maximal GL2 activity but is not required for the pattern of GL2 expression. Furthermore, GL2-promoter activity is influenced by expression of the myc-like maize R gene (35S::R) in Arabidopsis but is not affected by gl2 mutations. A position-dependent pattern of cell differentiation and GL2-promoter activity was also discovered in the hypocotyl epidermis that was analogous to the pattern in the root. Non-GL2-expressing cell files in the hypocotyl epidermis located outside anticlinal cortical cell walls exhibit reduced cell length and form stomata. Like the root, the hypocotyl GL2 activity was shown to be influenced by ttg and 35S::R but not by gl2. The parallel pattern of cell differentiation in the root and hypocotyl indicates that TTG and GL2 participate in a common position-dependent mechanism to control cell-type patterning throughout the apical-basal axis of the Arabidopsis seedling.

	INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

A fundamental problem in developmental biology is understanding how diverse cell types are specified in multicellular organisms. The formation of the root epidermis provides a simple model for investigating this problem in plants. In many species, two distinct cell types are formed during root epidermis development, root-hair cells and hairless cells (Cormack, 1949; Bunning, 1951; Cutter, 1978). The specification of cell fate in the root epidermis (i.e. the process that is responsible for causing a newly generated epidermal cell to differentiate into a root-hair cell or a hairless cell) varies in different plant species. In many monocots, cell specification is associated with an asymmetric cell division; the smaller daughter cell differentiates into a root-hair cell and the larger one becomes a mature, hairless cell (Sinnot and Bloch, 1939; Avers, 1963). In crucifers (e.g. Arabidopsis), cell specification is associated with the relative position of the epidermal cells. Immature epidermal cells located outside an anticlinal cortical cell wall (i.e. in contact with two underlying cortical cells) differentiate into root-hair cells, whereas cells located outside a periclinal cortical cell wall (i.e. in contact with a single cortical cell) differentiate into mature, hairless cells (Bunning, 1951; Cutter, 1978).

In Arabidopsis the differentiating epidermal cells, like many other cells of the root, are organized into columns (or files), with the newly formed cells located near the apex (in the meristematic region) and the older cells located farther from the apex (Dolan et al., 1993; Schiefelbein et al., 1997). Therefore, the differentiating epidermal cell files resemble an "assembly line" of cells, with each cell more developmentally advanced than the one before it, enabling an accurate assessment of the developmental fate of each cell. Furthermore, the precursors to the root-hair and hairless cells can be accurately identified prior to cell maturity. The developing root-hair cells (trichoblasts) possess more densely staining cytoplasm and reduced vacuolation relative to the developing hairless cells (atrichoblasts; Dolan et al., 1994; Galway et al., 1994).

The correlation between cell position and cell type in the root epidermis of Arabidopsis implies that cell signaling plays an important role in cell specification. However, the molecular events involved in the presumed cellular signaling and the subsequent cell differentiation are unknown. To define the molecules controlling cell specification in the Arabidopsis root epidermis, genetic studies have been used to identify loci that affect the normal pattern of root-hair and hairless cells (Dolan et al., 1994; Galway et al., 1994; Masucci and Schiefelbein, 1994; DiCristina et al., 1996; Masucci et al., 1996; Schneider et al., 1997; Wada et al., 1997). The analysis of two of these loci, TRANSPARENT TESTA GLABRA (TTG) and GLABRA2 (GL2), suggest that they encode negative regulators of root-hair-cell differentiation or, alternatively, positive regulators of hairless cell differentiation (Galway et al., 1994; Masucci et al., 1996). TTG and GL2 are also required for the appropriate formation of trichomes and the production of seed-coat mucilage (Koornneef, 1981; Koornneef et al., 1982; Rerie et al., 1994).

In the Arabidopsis root the recessive ttg mutations cause nearly all root epidermal cells, regardless of their position, to differentiate as root-hair cells (Galway et al., 1994). Expression of the maize R (Lc) cDNA under control of the cauliflower mosaic virus 35S promoter in Arabidopsis suppresses the ttg defects and causes an excessive number of root epidermal cells to differentiate as hairless cells (Galway et al., 1994). Together with the ability of the 35S::R transgene to suppress all of the other ttg defects, these results suggest that an R homolog (a myc-like bHLH transcriptional activator; Ludwig et al., 1989) in Arabidopsis may act at the same point or downstream from the TTG product to control these various processes (Lloyd et al., 1992; Galway et al., 1994).

Like the ttg mutations, mutations in the gl2 gene cause root hairs to form on essentially every root epidermal cell (Masucci et al., 1996). However, the ectopic root-hair cells in the gl2 mutant differ from those in the ttg mutant because they retain the cellular characteristics of wild-type hairless cells during their formation, including differences in cell vacuolation, cytoplasmic density, and cell length (Masucci et al., 1996). The GL2 gene encodes a homeodomain protein of the HD-Zip class and is expressed preferentially in the differentiating hairless epidermal cells within the meristematic and elongation regions of the root, which implies that GL2 acts as a cell-position-dependent transcriptional regulator to repress root-hair formation (Rerie et al., 1994; DiCristina et al., 1996; Masucci et al., 1996). Because GL2 affects a subset of the processes controlled by TTG (Masucci et al., 1996), and because the steady-state level of GL2 RNA is reduced in the ttg mutant (DiCristina et al., 1996), TTG may be a positive regulator of GL2.

In the present study the regulation of the GL2 homeobox gene and its role in epidermis development were analyzed. One goal of this research was to define genes/proteins that influence the spatial and/or quantitative regulation of GL2 during root development. To accomplish this, a series of GL2 promoter::GUS reporter gene fusions were constructed and used as sensitive, in vivo reporters of GL2-promoter activity. These led to the identification of a promoter region necessary for the cell-position-dependent expression of GL2 during root development. In addition, we show that maximal GL2-promoter activity in the root is TTG dependent and is influenced by the 35S::R transgene but not by gl2 mutations. Furthermore, these studies led to the discovery of a cell-position-dependent pattern of GL2-promoter activity and cell differentiation within the developing hypocotyl of Arabidopsis. Our results indicate that TTG and GL2 participate in a common mechanism to specify cell-type patterning during the development of the root and hypocotyl epidermis.

	MATERIALS AND METHODS

Top Abstract Introduction Methods Results Discussion References

Genetic Stocks and Growth Conditions

Two transgenic lines (1439 and 1434) bearing the 35S::R construct and displaying similar phenotypic effects were used in these studies; their root defects were originally described by Galway et al. (1994).

The ttg gl2 double mutant was constructed by crossing single-mutant plants and examining F₂ progeny for the ttg phenotype. These plants were subsequently test crossed with the gl2 mutant to identify the ttg gl2 double mutant. The 35S::R gl2 line was generated by crossing 35S::R plants with the gl2 mutant, backcrossing the F₁ to the gl2 mutant, and testing the progeny to identify individuals bearing the 35S::R construct and homozygous for the gl2 mutation.

Unless otherwise noted, seedlings were grown on vertically oriented Petri dishes on agarose-solidified medium containing 1% Suc, 0.6% agarose, and mineral nutrients under continuous illumination (Estelle and Somerville, 1987) following a 2-d chilling period, as previously described (Schiefelbein and Somerville, 1990).

Microscopy

To determine the location of root-hair-bearing cells, 5-d-old seedlings were submerged in molten 3% agarose. After the agarose solidified, transverse root sections were hand-cut with a double-edged razor blade and stained with a solution of 0.002% toluidine blue dye in artificial pond water. The location of root-hair cells relative to the underlying cortical cells was determined by viewing sections from at least five roots from each line.

The agarose-embedding technique was also used to analyze GUS staining within the hypocotyl of 3-d-old seedlings. To observe both the histochemical GUS staining and the position of hypocotyl cells, the cell walls of these sections were subsequently stained with FB28, as described previously (Galway et al., 1994), and viewed simultaneously under fluorescent and visible light.

The cellular characteristics of the hypocotyl epidermis were determined by GUS staining, clearing with 95% ethanol, and staining with toluidine blue dye. The relative cell length and stomatal formation in GL2::GUS-expressing and -nonexpressing files were measured from 10 cells of each file type from each of four independent sets of five seedlings grown for 5 d on the agarose-solidified medium described above. Cells in the stomatal complexes were not included in the cell-length calculations.

Plastic transverse sections were obtained from roots embedded in resin (JB-4, Polysciences, Warrington, PA) stained with 0.05% toluidine blue O, as previously described (Masucci and Schiefelbein, 1996).

Gene Fusions and Transgenic Plants

GUS-Reporter Gene Assays

Quantitative analysis of GUS activity in seedlings was performed essentially as described previously (Gallagher, 1992). Groups of 10 to 15 seedling root apices (excised 0.5 cm from the tips of 3- to 5-d-old seedlings) or hypocotyl/cotyledon segments (excised at the root-hypocotyl junction from 3-d-old seedlings) were ground with a plastic pestle in a microfuge tube in 200 µL of extraction buffer composed of 50 mm NaPO₄, 1 mm EDTA, 0.1% sodium lauryl sarcosine, 0.1% Triton X-100, and 10 mm DTT. The homogenized sample was centrifuged for 10 min, and 100 µL of the supernatant was assayed for GUS activity in a 30-min reaction at 37°C with 1 mm 4-MUG and 20% methanol. The reaction was terminated by the addition of an equal volume of 0.2 m Na₂CO₃, and fluorescence from the product was measured with a fluorometer (model TKO 100, Hoefer, Piscataway, NJ).

To accurately assess the effect of mutations and constructs on the activity of a particular GL2::GUS transgene, groups of seedlings used in the quantitative GUS assays were selected, based on their phenotype, from an F₂ pool derived from a cross between the mutant or construct line with the GL2::GUS line. The relative effect of a mutation or construct on GUS activity was determined by comparing the GUS activity (expressed as millimoles of product per milligram of protein) in the mutant seedlings with that of the wild-type seedlings within the F₂ pool. To ensure that the GL2::GUS construct was equally distributed in each phenotypic class within an F₂ pool, seedlings were tested for the presence of the construct. At least three assays were conducted for each mutant/transgene combination, and two independent lines containing a particular GL2 promoter::GUS transgene were tested for each combination.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Analysis of Cell-Position-Specific Activity of the GL2 Promoter in Roots

View larger version (13K): [in this window] [in a new window]   Figure 1. Effect of GL2-promoter fragments on GUS-reporter expression in roots and hypocotyls of Arabidopsis seedlings. The GL2-promoter fragments fused to the GUS-reporter gene are shown on the left. The ability of these constructs to drive GUS expression in a cell-position-dependent manner in the seedling root and/or hypocotyl was determined by histochemical staining with the X-Gluc substrate. +, Typical GUS expression pattern detected; , abnormal/no GUS expression detected. The relative root/hypocotyl GUS-activity value was determined by comparing the GUS activity (calculated as millimoles of product per milligram of protein per minute) in root versus hypocotyl extracts from a common set of 3-d-old seedlings bearing the indicated transgene. Values are means ± sd. Xh, XhoI; H, HindIII; Hp, HpaI; RI, EcoRI; RV, EcoRV; X, XbaI; M, MscI; and , deletion.

The results of this analysis indicate that a 500-bp EcoRV/XbaI DNA fragment located between position 840 and 1340 in the GL2 promoter is necessary to direct the appropriate pattern of GUS expression in differentiating hairless epidermal cells (Fig. 1). In plants bearing each of the constructs containing this fragment, GUS activity in the root was limited to epidermal cells located outside a periclinal cortical cell wall (i.e. the normal GL2 root-expression pattern; Masucci et al., 1996). No GUS activity could be detected in the roots of plants bearing the MHp construct, which lacks this fragment. Likewise, plants containing the MR construct did not exhibit detectable staining of differentiating epidermal cells, although they did display a trace of GUS staining in emerging secondary roots.

Effect of ttg and 35S::R on GL2-Promoter Activity in Roots

Qualitative GUS Assay

View larger version (97K): [in this window] [in a new window]   Figure 2. Spatial expression pattern of GL2::GUS-reporter-gene fusion construct during root development in Arabidopsis seedlings. Four-day-old seedlings were assayed for GUS activity by histochemical staining with the X-Gluc substrate. A, Wild-type root containing the GL2::GUS transgene. Bar = 50 µm. B, Wild-type root containing the GL2::GUS transgene. GUS-expressing cells are preferentially located in specific epidermal cell files. Bar = 20 µm. C, ttg-1 mutant root containing the GL2::GUS transgene. Bar = 50 µm. D, ttg-1 mutant root containing the GL2::GUS transgene. GUS-expressing cells are preferentially located in specific epidermal cell files. Bar = 20 µm. E, 35S::R root containing the GL2::GUS transgene. Bar = 50 µm. F, 35S::R root apex containing the GL2::GUS transgene. GUS-expressing cells are not clearly located in specific epidermal cell files. Bar = 20 µm. G, Wild-type root containing the GL2::GUS transgene; transverse plastic section taken from the late meristematic region. GUS expression is limited to epidermal cells located outside periclinal cortical cell walls (i.e. in contact with a single cortical cell). At this developmental stage, a single layer of lateral root cap cells surrounds the epidermis. Bar = 20 µm. H, ttg-1 mutant root containing the GL2::GUS transgene; transverse plastic section taken from the late meristematic region. No GUS expression is observed. At this developmental stage, a single layer of lateral root cap cells surrounds the epidermis. Bar = 20 µm. I, 35S::R root containing the GL2::GUS transgene; transverse plastic section taken from the late meristematic region. GUS expression is observed throughout the epidermis, cortex, and lateral root cap. At this developmental stage, a single layer of lateral root cap cells surrounds the epidermis. Bar = 20 µm. J, ttg-1 mutant root containing the GL2::GUS transgene; thick transverse section from agarose-embedded root. GUS expression is observed in epidermal cells located outside periclinal cortical cell walls. Bar = 20 µm. K, Wild-type root apex containing the GL2::GUS transgene. Whole-mount root preparation showing GUS expression near the meristem initials but not within the lateral or columella root cap cells. The dense staining visible in the upper portion of this root is due to GUS-expressing epidermal cells above and below the plane of focus. Bar = 20 µm. L, 35S::R root apex containing the GL2::GUS transgene. Whole-mount root preparation showing GUS expression throughout region containing meristem initials and within the lateral root cap cells but not within the columella root cap cells. The dense staining visible in the upper root is due to GUS-expressing cells above and below the plane of focus. Bar = 20 µm. M, 35S::GUS root apex. Whole-mount root preparation showing preferential GUS expression throughout region containing meristem initials, the root cap, and the developing vascular tissue. Bar = 20 µm. These GL2::GUS seedlings all contain the full-length 4-kb GL2 promoter::GUS transgene. Note that root hairs are not visible in these images near the root apex; hairs form on epidermal cells at a later developmental age, just beyond the field of view shown in A, C, and E.

To further investigate the role of the TTG pathway on GL2 expression in roots, we tested the effect of the 35S::R construct, which is able to suppress ttg mutant root defects (Galway et al., 1994). Seedling roots containing both the 35S::R and the GL2::GUS constructs displayed ectopic GUS activity compared with the wild-type GL2::GUS pattern (Fig. 2, E and F). Transverse sections from the 35S::R GL2::GUS roots showed that GUS activity accumulated throughout the epidermis (indicating ectopic expression in the differentiating root-hair cells), in lateral root cap cells, and in cortical cells (Fig. 2I). Therefore, the expression of the myc-like R gene by the 35S promoter is sufficient to induce ectopic GL2-promoter activity during root development.

To further analyze GUS expression in the 35S::R GL2::GUS line, GUS activity in a 35S::GUS transgenic line was examined and compared with GL2::GUS and 35S::R GL2::GUS roots (Fig. 2, K-M). This analysis is useful because the 35S::GUS provides an indication of the likely location of R protein accumulation in the 35S::R line. The comparison showed that the strong 35S-promoter activity in the region of the root meristem initials and lateral root cap coincides with ectopic expression of GL2::GUS in these regions in the 35S::R GL2::GUS line. However, despite 35S-promoter activity in columella root cap cells and vascular tissue, the 35S::R GL2::GUS roots lack detectable GUS activity in these regions (Fig. 2, K-M). Therefore, ectopic GL2 expression occurs in some but not all of the cells that are likely to accumulate the R protein, which implies that additional factor(s) are necessary to activate the GL2 promoter in these root tissues.

Quantitative GUS Assay

The quantitative GUS analysis of seedlings possessing the 35S::R and GL2::GUS constructs showed that the 35S::R roots have a slight reduction in GUS activity relative to the wild type (Table I). This result was somewhat unexpected, because ectopic GUS expression is present in the 35S::R GL2::GUS roots (Fig. 2), which would be expected to lead to an increase in the overall root GUS activity. Apparently, the quantitative effect of 35S::R on root GUS expression is due to a reduction in GUS expression in the cells that normally express GL2 (differentiating hairless cells), as well as the induction of GUS expression in cells that normally do not express GL2.

To define the GL2-promoter region required for the ttg and 35S::R effects on GL2 expression, the GL2-promoter deletion::GUS lines (Fig. 1) were introduced into the ttg and 35S::R genetic backgrounds and each combination was tested for GUS activity. As shown in Table II, the ttg mutant was found to impart the same relative effect on each GL2::GUS construct; the amount of GUS activity and the intensity of GUS staining was reduced (but the cell-position expression pattern was not affected) in each line that exhibited GUS expression. This shows that the DNA fragment in the GL2 promoter responsible for the ttg effect could not be separated from the 500-bp EcoRV/XbaI DNA fragment responsible for the cell-position-dependent root expression (Fig. 1). The analysis of GL2::GUS lines bearing the 35S::R construct generated a similar outcome; the amount of GUS activity was slightly reduced (relative to the wild type) and ectopic GUS expression was observed (by X-Gluc staining) in the roots from each line (Table II). The correlation between the GL2-promoter region required for the cell-position-dependent expression (determined in Fig. 1) and the region required for the ttg and 35S::R effects suggests that the TTG and/or an R homolog influence GL2-promoter activity at or near the same site required for cell-position-dependent promoter activity.

Effect of the gl2 Mutant on GL2-Promoter Activity

Genetic Analysis of TTG, 35S::R, and GL2 during Root Development

In a second experiment, the ttg gl2 double mutant was constructed and analyzed. The root epidermis in ttg-1 gl2-1 roots was indistinguishable from either single mutant with respect to the production of root-hair and hairless cells (Table I), and no synergistic effect on root development was detected. This indicates that TTG and GL2 are not likely to possess redundant functions in root development and is consistent with the notion that TTG and GL2 act within the same pathway.

GL2-Promoter Activity during Hypocotyl Development

View larger version (102K): [in this window] [in a new window]   Figure 3. The spatial-expression pattern of the GL2-GUS-reporter-gene fusion construct during hypocotyl development. Seedlings harboring the 4-kb GL2 promoter::GUS transgene were stained for GUS activity using X-Gluc. A, Wild-type hypocotyl from 3-d-old seedling. Bar = 100 µm. B, Wild-type hypocotyl epidermis from 3-d-old seedling. GUS-expressing cells are located within specific epidermal cell files. Bar = 50 µm. C, Wild-type 3-d-old seedling sectioned through the hypocotyl. A ring of GUS-expressing hypocotyl epidermal cells is visible. Bar = 100 µm. D, Wild-type hypocotyl from 3-d-old seedling; transverse agarose section. GUS expression is present in epidermal cells located outside a periclinal cortical cell wall. E, Wild-type hypocotyl from 5-d-old seedling. Note that cells in the GL2::GUS-expressing files are longer than cells in the nonexpressing files. Bar = 40 µm. F, Wild-type hypocotyl from 5-d-old seedling. Stomatal development (arrowhead) occurs in non-GL2::GUS-expressing cell files. Bar = 40 µm. G, Wild-type hypocotyl and cotyledons from 3-d-old seedling. GUS-expressing cells are visible at the margin of cotyledons. Bar = 100 µm. H, Wild-type hypocotyl/cotyledon junction region from 3-d-old seedling. Bar = 50 µm. I, Wild-type cotyledon; transverse section taken near cotyledon apex (3-d-old). Arrowhead indicates a GUS-staining epidermal cell. Bar = 50 µm. J, Wild-type hypocotyl from 7-d-old seedling. GUS expression is visible in the developing leaf primordia and trichomes. Bar = 200 µm. K, ttg-1 mutant hypocotyl from 3-d-old seedling. Bar = 100 µm. L, ttg-1 mutant hypocotyl from 3-d-old seedling. GUS-expressing cells appear to be located within specific epidermal cell files. Bar = 50 µm. M, ttg-1 mutant hypocotyl from 3-d-old seedling; transverse agarose section. GUS expression is present in epidermal cells located outside a periclinal cortical cell wall (i.e. in contact with a single cortical cell). Bar = 50 µm. N, 35S::R hypocotyl from 3-d-old seedling; transverse agarose section. GUS expression is present in cells located throughout the epidermis. Bar = 50 µm. O, 35S::R hypocotyl from 3-d-old seedling. Bar = 100 µm. P, 35S::R hypocotyl from 3-d-old seedling. GUS-expressing cells are located throughout the epidermis. Bar = 50 µm.

Two additional features of the GL2::GUS expression pattern were observed. First, although no GUS expression was detected within the cotyledon proper, the GL2::GUS seedlings displayed GUS activity in a ring of cells around the circumference of the cotyledons (Fig. 3, G and H). These GUS-expressing cells at the cotyledon margin appeared to be epidermal cells located outside a periclinal cortical cell wall (Fig. 3I). Second, the analysis of GUS accumulation in hypocotyls of GL2::GUS seedlings at different developmental stages showed that the intensity of the hypocotyl GUS staining was greatest in 2- or 3-d-old seedlings and becomes diminished and largely localized to the upper portion of the hypocotyl at later stages of development (Fig. 3, A and J). This suggests that GL2-promoter activity in the hypocotyl may be correlated with the differentiation of the hypocotyl epidermal cells.

To define the GL2-promoter region(s) involved in directing the hypocotyl-epidermis expression, GUS accumulation was examined by X-Gluc staining in 3-d-old seedlings containing the GL2 promoter::GUS constructs illustrated in Figure 1. This analysis showed that the 500-bp EcoRV/XbaI fragment at 840 to 1340 is necessary for the cell-file-specific expression of GUS in hypocotyls (Fig. 1). Arabidopsis seedlings bearing GL2::GUS transgenes lacking this 500-bp fragment exhibited no detectable GUS activity in the hypocotyls. This result shows that the region of the GL2 promoter necessary for position-dependent expression in the root is also required for proper expression in the hypocotyl epidermis.

Although the same GL2-promoter region is required for the root- and hypocotyl-expression pattern, other promoter elements may exist that exert different quantitative effects on the root and hypocotyl expression. To test this possibility, the ratio of root to hypocotyl GUS activity was determined for each of the GL2::GUS lines by conducting quantitative GUS assays on extracts from the root and hypocotyl of a common set of 3-d-old seedlings. If a GL2-promoter element exists that differentially affects root versus hypocotyl expression, then a GL2::GUS transgene that lacks this element would be expected to display an altered ratio of root to hypocotyl GUS activity. The results of this experiment showed that the root to hypocotyl GUS activity ratio was similar in each of the GL2::GUS transgene lines (Fig. 1), which indicates that GL2-promoter activity is not differentially regulated in the root and hypocotyl in a quantitative manner by promoter elements located outside the 500-bp EcoRV/XbaI fragment.

Position-Dependent Cell Differentiation in the Hypocotyl Epidermis

Effect of Mutations on GL2-Promoter Activity in the Hypocotyl

Compared with hypocotyls bearing the GL2::GUS in a wild-type background, ttg-1 GL2::GUS hypocotyls exhibit a greatly diminished level of GUS activity, as assessed by X-Gluc staining (Fig. 3, K and L) and quantitative 4-MUG GUS assays (Table I). The small amount of hypocotyl GUS activity was in epidermal cells located outside a periclinal cortical cell wall, indicating that the spatial pattern of GL2::GUS expression was unaltered (Fig. 3M).

In the 35S::R GL2::GUS hypocotyls, GUS activity is detected in all epidermal cells and is particularly intense near the upper end of the hypocotyl (Fig. 3, N-P). In addition, ectopic GUS expression was observed extending beyond the margins of the cotyledons in these seedlings (Fig. 3O; data not shown). Nonetheless, the overall amount of hypocotyl GUS activity in the 35S::R GL2::GUS seedlings was not significantly different from the wild-type GL2::GUS (Table I).

The qualitative and quantitative analysis of GUS activity in hypocotyls of the gl2-1 GL2::GUS seedlings did not indicate a significant difference compared with the wild-type GL2::GUS activity (Table I; data not shown).

Taken together, analysis of the ttg, 35S::R, and gl2 showed that these factors exert a similar effect on the hypocotyl GL2-promoter activity as they do on the root GL2 activity, which suggests that the same regulatory mechanism is responsible for controlling root and hypocotyl GL2 expression.

Effect of Mutations on Cell Differentiation in the Hypocotyl Epidermis

In gl2-1 GL2::GUS hypocotyls, a significantly greater proportion of stomata were found in GUS-expressing cells (21% ectopic stomata in gl2 versus 5% in the wild type), although the length of the GUS-expressing cells relative to the non-GUS-expressing cells was similar to the wild type (Table III). Likewise, ttg-1 GL2::GUS hypocotyls possessed a greater proportion of ectopic stomata than the wild type but did not exhibit a statistically significant deviation in cell length (Table III). These alterations in stomatal patterning indicate that TTG and GL2 are required to restrict stomatal development to the epidermal cells located over the anticlinal cortical cell walls of the hypocotyl.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Control of GL2 in Root Development

Although the ttg-1 mutant exhibits 29% of the GL2-promoter activity present in wild-type roots, this level of GL2 expression is apparently inadequate to cause hairless cell differentiation, because root hairs are produced on essentially every root epidermal cell in the ttg-1 mutant (Table I; Galway et al., 1994). These results imply that the inhibition of root-hair formation in atrichoblasts is sensitive to the quantity of the GL2 homeodomain protein. This supposition is supported by the previously reported finding that the gl2 mutation displays a dosage effect in the roots; with gl2/+ seedlings possessing a significant proportion of ectopic root-hair cells (Masucci et al., 1996). Therefore, a major factor in appropriate epidermal cell specification in the root may be sufficient induction of GL2 expression in atrichoblasts (by TTG and probably other genes) to ensure that hair formation is inhibited.

The GUS-expression pattern in 35S::R GL2::GUS roots shows that the R-gene product is sufficient to induce expression of the GL2 promoter in cells that normally do not exhibit detectable GL2 activity. In particular, GL2 expression is greatly enhanced in epidermal cells located outside anticlinal cortical cell walls (Fig. 2I), which is correlated with a change in the developmental fate of these cells (Table I; Galway et al., 1994). This implies that, in at least some cells, a protein related to the myc-like R product (or a protein controlled by an R-like protein) may be the limiting factor controlling GL2 expression and root epidermal cell fate. In this regard, it is interesting that GUS activity accumulates in lateral root cap cells of the 35S::R GL2::GUS roots, because the cells of the epidermis and lateral root cap are derived from a common initial cell in the root meristem (Dolan et al., 1993; Scheres et al., 1994). Thus, regulation of the activity of an R-like protein may be the critical factor responsible for differential activation of GL2 in the daughter cells of the epidermis/lateral root cap initial and, more specifically, within a particular subset of epidermal cells. Additional support for the view that an R-like protein acts through GL2 comes from the phenotype of the 35S::R gl2 line (Table I), which shows that 35S::R requires a functional GL2 protein for its effects on root epidermis development.

It is interesting to note that a slight decrease in overall GUS activity is observed in the 35S::R GL2::GUS roots, despite some ectopic GUS accumulation. This suggests that the cells exhibiting high levels of GL2 expression in the wild type (differentiating hairless cells) are inhibited to some extent in their GL2-promoter activity in the 35S::R background. This may be due to the titration of a transcription factor partner(s) required for GL2-promoter activity by the high-level expression of the myc-like bHLH maize R protein.

The analysis of the GL2-promoter region shows that an approximately 500-bp region at 840 to 1340 is necessary for cell-position-dependent promoter activity in the root epidermis. In addition, the effects of the ttg and 35S::R factors on GL2-promoter activity are also dependent on the presence of this 500-bp fragment. This fragment contains a putative myb-binding site (TACTAACAGTATA), which opens the possibility that the TTG and/or an R-like protein may interact with (or control the activity of) a myb-like protein to regulate GL2 in this region. It is interesting to note that the same 500-bp region important for GL2-promoter activity in the root/hypocotyl epidermis has also been found to be important for trichome and leaf primordia expression of GL2 in the Arabidopsis shoot (D. Szymanski and M.D. Marks, unpublished data).

Control of Epidermis Development in the Arabidopsis Hypocotyl

The position-dependent pattern of GL2-promoter activity is correlated with a position-dependent pattern of hypocotyl epidermal cell differentiation. Our results show that hypocotyl epidermal cells located over an anticlinal cortical cell wall (non-GL2::GUS-expressing cells) differ from epidermal cells located over periclinal cortical cell walls (GL2::GUS-expressing cells) because they are of reduced length and preferentially develop stomata (Table III). In previous studies of cell differentiation in the light-grown Arabidopsis hypocotyl epidermis, two morphologically distinct cell types were identified, with protruding epidermal cell files located outside periclinal cortical cell walls and burrowed cell files located outside anticlinal cortical cell walls (Wei et al., 1994; Gendreau et al., 1997). Together, these results show that cells of the hypocotyl epidermis, like the root epidermis, undergo position-dependent cell differentiation, which generates a common pattern of epidermal cell types in the root and hypocotyl. The identification of a common patterning mechanism for cell types within the root and hypocotyl epidermis, together with the previous finding of common radial patterning of tissue types in the root and hypocotyl (Scheres et al., 1995), provides evidence for patterning events that control development throughout the apical-basal axis of the Arabidopsis seedling.

Although stomata have been observed in the Arabidopsis hypocotyl (Reed et al., 1993; Wei et al., 1994) and the patterning of stomata in the leaf epidermis has been described (Larkin et al., 1997), the results from our study provide evidence for a previously unreported (to our knowledge) cell-position-dependent pattern of stomatal development in the hypocotyl. These results show that stomata preferentially develop in cell files located outside an anticlinal cortical cell wall (Table III). One possible reason for this pattern is that it may enable more efficient gas exchange through the apoplast. Furthermore, we find that ttg and gl2 mutations significantly alter stomatal patterning in the hypocotyl, enabling a greater proportion of cells located outside periclinal cell walls to produce stomata (approximately 25% ectopic stomata in each mutant; Table III). Therefore, TTG and GL2 are required to ensure that cells located outside a periclinal cortical cell wall differentiate into hairless cells in the root epidermis and differentiate into nonstomatal cells in the hypocotyl epidermis.

The difference in cell length that we have identified between hypocotyl epidermal cells in adjacent files indicates that there is a position-dependent difference in the extent of cell division during hypocotyl development. Specifically, this implies that developing hypocotyl cells located outside anticlinal cortical cell walls undergo a greater number of cell divisions than developing cells located outside periclinal cortical cell walls. This is similar to the position-dependent difference in root epidermal cell division recently reported (Berger et al., 1998) and provides further support for a close relationship between the epidermal developmental mechanisms operating in the root and hypocotyl. Because the gl2 and ttg mutations do not appear to significantly alter the relative hypocotyl cell length present in the wild type (Table III), it is likely that other genes are involved in regulating this cell characteristic.

A Model for Epidermal Cell Differentiation in the Root and Hypocotyl

View larger version (12K): [in this window] [in a new window]   Figure 4. Proposed pathway for the regulation of cell differentiation in the root and hypocotyl of Arabidopsis. The TTG is proposed to activate an R-like bHLH protein that positively controls the transcription of GL2. The GL2 homeodomain protein is proposed to control root and hypocotyl epidermal cell differentiation. See text for additional discussion. Arrows indicate positive action; blunted lines indicate negative regulation. RHD6, Root hair defective 6; and RHL, root hairless.

In the future, this proposed gene pathway is likely to be refined by the analysis of additional loci that have recently been shown to influence root epidermis development, such as the CPC, ERH, and RHL loci (Schneider et al., 1997; Wada et al., 1997). In particular, genetic evidence indicates that the CPC gene product, which contains a myb-like DNA-binding domain, may act as a negative regulator of GL2 (Wada et al., 1997).

In addition to providing insight into the components and arrangement of the pathway that controls cell differentiation in the seedling epidermis, the results from this study provide clues regarding the regulation of this pathway. Because the GL2-promoter activity is present in a similar pattern throughout the apical-basal axis in the Arabidopsis seedling, including specific cells of the root, hypocotyl, and cotyledon, the proposed pathway may be initiated during embryo development. It is possible that a "prepattern" of TTG/R/GL2 expression may be established at an early stage of embryogenesis when the radial pattern of cells is organized (Scheres et al., 1994), such that epidermal cells located outside periclinal cortical cell walls throughout the apical-basal axis initiate expression of this pathway.

An interesting developmental issue that is amenable to analysis in this system is the manner in which common regulatory genes are used to direct the development of different cells or tissues in a multicellular organism. As described here, the developing root and hypocotyl use common genetic components to generate a similar pattern of cell types and gene activity. In addition, the TTG and GL2 genes are used to regulate trichome formation and seed coat mucilage, two other epidermal developmental processes (Koornneef, 1981; Koornneef et al., 1982; Marks, 1997). Despite these similarities, there is also evidence of variation in the control of the TTG/GL2 pathway in different tissues. For example, the hypocotyl GL2 expression is reduced approximately 3-fold compared with root GL2 expression in the ttg-1 mutant background (Table I). Therefore, despite some common components, the regulation of cell differentiation in different tissues is likely to depend on differences in the organization of the pathway, on redundancy for some of the factors in different tissues, and/or on differences in the degree of the influence of the common components. The continued analysis of the TTG/GL2 pathway in different Arabidopsis tissues is expected to lead to a better understanding of the complex regulation of cell specification in plants.

	FOOTNOTES

   Received November 14, 1997; accepted January 26, 1998.

	ABBREVIATIONS

Abbreviations: bHLH, basic helix-loop-helix. 4-MUG, 4methylumbelliferyl -d-glucuronic acid. X-Gluc, 5-bromo-4-chloro-3-indolyl--d-glucuronic acid.

	ACKNOWLEDGMENTS

We thank James Masucci, Haiyang Wang, Ben Scheres, Fred Berger, Katharina Schneider, Liam Dolan, and Takuji Wada for helpful discussions. We acknowledge the contribution of genetic stocks by the Arabidopsis Biological Resource Center (Ohio State University, Columbus).

	LITERATURE  CITED

Top Abstract Introduction Methods Results Discussion References

Avers CJ (1963) Fine structure of phleum root meristem cells. II. Mitotic asymmetry and cellular differentiation. Am J Bot 50: 140-148

Berger F, Hung C-Y, Dolan L, Schiefelbein J (1998) Control of epidermal cell division in the root meristem of Arabidopsis thaliana. Dev Biol (in press)

Bunning E (1951) Uber die Differenzierungsvorgange in der Cruciferenwurzel. Planta 39: 126-153 [CrossRef][ISI]

Cormack RGH (1949) The development of root hairs in angiosperms. Bot Rev 15: 583-612 [ISI]

Cutter EG (1978) Plant Anatomy. Clowes & Sons, London, pp 94-106

DiCristina MD, Sessa G, Dolan L, Linstead P, Baima S, Ruberti I, Morelli G (1996) The Arabidopsis Athb-10 (GLABRA2) is an HD-Zip protein required for regulation of root hair development. Plant J 10: 393-402 [CrossRef][ISI][Medline]

Dolan L, Duckett C, Grierson C, Linstead P, Schneider K, Lawson E, Dean C, Poethig S, Roberts K (1994) Clonal relations and patterning in the root epidermis of Arabidopsis. Development 120: 2465-2474 [Abstract/Free Full Text]

Dolan L, Janmaat K, Willemsen V, Linstead P, Poethig S, Roberts K, Scheres B (1993) Cellular organisation of the Arabidopsis thaliana root. Development 119: 71-84 [Abstract]

Estelle MA, Somerville C (1987) Auxin-resistant mutants of Arabidopsis thaliana with an altered morphology. Mol Gen Genet 206: 200-206 [CrossRef][ISI]

Gallagher SR (1992) GUS Protocols. Academic Press, London

Galway M, Masucci J, Lloyd A, Walbot V, Davis R, Schiefelbein J (1994) The TTG gene is required to specify epidermal cell fate and cell patterning in the Arabidopsis root. Dev Biol 166: 740-754 [CrossRef][ISI][Medline]

Gendreau E, Traas J, Desnos T, Grandjean O, Caboche M, Hofte H (1997) Cellular basis of hypocotyl growth in Arabidopsis thaliana. Plant Physiol 114: 295-305 [Abstract]

Koornneef M (1981) The complex syndrome of ttg mutants. Arabidopsis Inf Serv 18: 45-51

Koornneef M, Dellaert SWM, van der Veen JH (1982) EMS- and radiation-induced mutation frequencies at individual loci in Arabidopsis thaliana (L) Heynh. Mutat Res 93: 109-123 [ISI][Medline]

Larkin JC, Marks MD, Nadeau J, Sack F (1997) Epidermal cell fate and patterning in leaves. Plant Cell 9: 1109-1120 [CrossRef][ISI][Medline]

Lloyd AM, Walbot V, Davis RW (1992) Arabidopsis and Nicotiana anthocyanin production activated by maize regulators R and C1. Science 258: 1773-1775 [Abstract/Free Full Text]

Ludwig SR, Habera LF, Dellaporta SL, Wessler SR (1989) Lc, a member of the maize R gene family responsible for tissue-specific anthocyanin production, encodes a protein similar to transcriptional activators and contains the myc-homology region. Proc Natl Acad Sci USA 86: 7092-7096 [Abstract/Free Full Text]

Marks MD (1997) Molecular genetic analysis of trichome development in Arabidopsis. Annu Rev Plant Physiol Plant Mol Biol 48: 137-163 [CrossRef][ISI]

Masucci JD, Rerie WG, Foreman DR, Zhang M, Galway ME, Marks MD, Schiefelbein JW (1996) The homeobox gene GLABRA2 is required for position-dependent cell differentiation in the root epidermis of Arabidopsis thaliana. Development 122: 1253-1260 [Abstract]

Masucci JD, Schiefelbein JW (1994) The rhd6 mutation of Arabidopsis thaliana alters root-hair initiation through an auxin- and ethylene-associated process. Plant Physiol 106: 1335-1346 [Abstract]

Masucci JD, Schiefelbein JW (1996) Hormones act downstream of TTG and GL2 to promote root hair outgrowth during epidermis development in the Arabidopsis root. Plant Cell 8: 1505-1517 [Abstract]

Reed JW, Nagpal P, Poole DS, Furuya M, Chory J (1993) Mutations in the gene for the red/far-red light receptor phytochrome B alter cell elongation and physiological responses throughout Arabidopsis development. Plant Cell 5: 147-157 [Abstract]

Rerie WG, Feldmann KA, Marks MD (1994) Genes Dev 8: 1388-1399 [Abstract/Free Full Text]

Scheres B, Di Laurenzio L, Willemsen V, Hauser M-T, Janmaat K, Weisbeek P, Benfey PN (1995) Mutations affecting the radial organization of the Arabidopsis root display specific defects throughout the embryonic axis. Development 121: 53-62 [Abstract]

Scheres B, Wolkenfelt H, Willemsen V, Terlouw M, Lawson E, Dean C, Weisbeek P (1994) Embryonic origin of the Arabidopsis primary root and root meristem initials. Development 120: 2475-2487 [Abstract/Free Full Text]

Schiefelbein JW, Masucci JD, Wang H (1997) Building a root: the control of patterning and morphogenesis during root development. Plant Cell 9: 1089-1098 [CrossRef][ISI][Medline]

Schiefelbein JW, Shipley A, Rowse P (1992) Calcium influx at the tip of growing root-hair cells of Arabidopsis thaliana. Planta 187: 455-459

Schiefelbein JW, Somerville C (1990) Genetic control of root hair development in Arabidopsis thaliana. Plant Cell 2: 235-243 [Abstract/Free Full Text]

Schneider K, Wells B, Dolan L, Roberts K (1997) Structural and genetic analysis of epidermal cell differentiation in Arabidopsis primary roots. Development 124: 1789-1798 [Abstract]

Sinnot EW, Bloch R (1939) Cell polarity and the differentiation of root hairs. Proc Natl Acad Sci USA 25: 248-252 [Free Full Text]

Wada T, Tachibana T, Shimura Y, Okada K (1997) Epidermal cell differentiation in Arabidopsis determined by a Myb-homolog, CPC. Science 277: 1113-1116 [Abstract/Free Full Text]

Wei N, Kwok SF, von Arnim AG, Lee A, McNellis TW, Piekos B, Deng X-W (1994) Arabidopsis COP8, COP10, and COP11 genes are involved in repression of photomorphogenic development in darkness. Plant Cell 6: 629-643 [Abstract/Free Full Text]

Copyright Clearance Center:   0032-0889/98/117/0073/12
 
© 1998 American Society of Plant Physiologists