INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS AND DISCUSSION MATERIALS AND METHODS LITERATURE CITED

The model plant Arabidopsis has many advantages that make it suitable for use in classical anatomy and morphology courses. However, its small size may present a challenge for the study of some aspects of plant anatomy, particularly for the novice. Easily visible molecular markers that are expressed in a single cellular or subcellular compartment have been employed to characterize both normal patterns of development as well as for the analysis of mutant phenotypes (Benfey et al., 1993; Malamy and Benfey, 1997; Roe et al., 1997; Topping and Lindsey, 1997; Tsukaya and Uchimiya, 1997; Berger et al., 1998; Sabatini et al., 1999; Sessions et al., 1999). Such tools usually exploit sensitive reporter genes such as -glucuronidase (GUS) or Green Fluorescent Protein, which are generally easy to use and not prohibitively expensive. Expression patterns that mark cellular or anatomical features would be a useful teaching aid, and can readily be employed in a classroom setting to aid students in visualization of anatomical compartments.

Gene and enhancer traps serve as a rich source of molecular markers, and large collections of trap lines have been generated in Arabidopsis (Bechtold et al., 1993; Topping et al., 1994; Sundaresan et al., 1995; Campisi et al., 1999; Springer, 2000). These trap lines contain a reporter gene that has been inserted randomly into the genome. When insertion occurs within or adjacent to a gene or enhancer region, the reporter gene is expressed under the control of the native promoter or enhancer elements. Thus, patterns of reporter gene expression represent the expression of an endogenous chromosomal gene.

In the collections of enhancer and gene trap lines that have been generated, many different reporter gene expression patterns have been observed (Sundaresan et al., 1995; Campisi et al., 1999; Springer, 2000). Although some ubiquitous patterns that show expression in most or all plant cells and tissues have been observed, more commonly expression is detected in a limited number of cells or tissue systems. These more limited patterns are often interesting and informative regarding compartments of gene regulation. Although botanists have described anatomical and functional compartments in plants, enhancer trap transposants have revealed additional subdivisions that do not always correspond to those previously described.

Here, 30 Arabidopsis enhancer trap transposants are described. These lines were selected because they have GUS reporter gene expression patterns that are useful for teaching anatomy. These transposant lines will help students to identify cell types, tissues, tissue systems, and more complex patterns of gene regulation during development. Although molecular characterization of individual transposants will no doubt lead to the identification of important genes in plant development, at this stage they are useful as markers and provide a novel teaching resource.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS AND DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Leaf Markers

A number of transposant lines that exhibited cell type-specific reporter gene expression patterns were identified (Tables I and II). Arabidopsis trichomes, single-celled epidermal hairs, are quite easy to identify, and as such were one of the first cell types to be studied by mutant analysis (Herman and Marks, 1989; Hülskamp et al., 1994; Larkin et al., 1994). The precursor cells that will give rise to mature trichomes, however, are difficult to observe. Two enhancer trap lines were identified that will aide in visualization of trichome development. GUS activity in transposant UCR1 was observed in all mature trichomes but was not detected in earlier stages of trichome development (Fig. 1A). In contrast, GUS activity in transposant UCR2 was visible throughout all stages of trichome development, beginning with the unexpanded trichome precursor cells (Fig. 1B).

View larger version (124K): [in this window] [in a new window]   Figure 1.   Enhancer trap transposants with GUS activity in the shoot. A, UCR1, GUS activity in mature trichomes. B, UCR2, GUS activity in trichomes and trichome precursor cells (arrowhead). C, UCR3, GUS activity in cotyledon guard cells. D, UCR4, GUS activity in stipules. E, UCR5, GUS activity in cotyledon epidermis, image shows a transverse hand-section. F, UCR6, GUS activity in palisade mesophyll of cotyledon, image shows a transverse hand-section and a whole mount (inset). G, UCR7, GUS activity in cotyledon vein. H, UCR8, GUS activity in veins of cotyledon and hypocotyl. I, UCR9, GUS activity in developing vascular tissue of leaf blade and petiole (arrowhead). J, UCR10, Longitudinal hand section through shoot apex showing GUS activity at base of leaf primordia. K, UCR11, GUS activity in rings of cells at the base of leaf primordia. L, UCR12, GUS activity in trichomes, petiole margins, and hypocotyl vasculature. M, UCR13, GUS activity in cotyledon hydathodes, petiole, and upper hypocotyl. N, UCR14, GUS activity in veins of hypocotyl and cotyledon petiole. O, UCR15, GUS activity is excluded from leaf primordia. P, UCR16, GUS activity in cotyledon and leaf blade. Scale bars = 100 µm.

A second epidermal cell type, the guard cell, may also be difficult for the novice to identify. Transposant UCR3 showed GUS activity in mature guard cells throughout the seedling, including those on cotyledons, leaves, hypocotyls, and petioles (Fig. 1C). Weak GUS activity was also occasionally detected in the vasculature in this transposant line.

Stipules, structures that form at each leaf base, are quite reduced in Arabidopsis, consisting of 20 to 30 cells (Bowman, 1993). Because of their small size and proximity to the meristem, they can be difficult to identify and may be confused with initiating leaf primordia. GUS activity in transposant UCR4 specifically marked stipules (Fig. 1D) and was not detected in any other location in the vegetative shoot.

Each of the three plant tissue systems (dermal, ground, and vascular) was identified in the following group of enhancer trap lines. Transposant UCR5 showed GUS activity throughout the entire cotyledon epidermis, including guard and pavement cells (Fig. 1E). GUS activity was not detected in ground or vascular tissues (Fig. 1E) or in young developing leaves (data not shown). GUS expression was confined to the mesophyll in transposant UCR6, with the most intense expression detected in palisade cells in the cotyledon (Fig. 1F). In transposant UCR7, GUS activity was detected in the vasculature. Upon closer examination, staining appeared to be confined to a single axial cell type (Fig. 1G). Cells staining for GUS activity were elongate and adjacent to or within the xylem tissue but scattered throughout the shoot vasculature and were likely differentiating vessel elements. In contrast, transposant UCR8 showed GUS expression localized to the phloem in vasculature throughout the plant (Fig. 1H). This enhancer trap line is a useful tool for visualizing the different vascular patterns found in lateral organs, from the simple venation pattern in cotyledons to the increasingly branched, reticulate pattern in leaves. Vein continuity between the leaf, hypocotyl, and roots can also easily be seen.

GUS activity in transposants UCR9 and UCR10 marked the early stages of vascular development. High levels of GUS activity were visible in differentiating veins of the leaf blade and petiole and in the root tip (see below) in UCR9 (Fig. 1I). The mature veins of the hypocotyl and cotyledon were unstained (data not shown). After flowering, expression expanded to the floral buds and mature petals, in addition to developing veins (Table I). In transposant UCR10, GUS activity was detected in a triangular wedge of tissue at the base of leaf primordia that appeared to mark the forming midvein (Fig. 1J), although staining appeared before differentiation of vascular tissue. The pattern was transient, so that only weak expression was detectable in older primordia.

Root Markers

The size and regular organization of the Arabidopsis root make it useful for studies of root anatomy (Scheres and Wolkenfelt, 1998). Five enhancer trap lines have been chosen that individually identify each radial layer of the mature root (Fig. 2, A-E). The epidermis, including root hairs and atrichoblast cell files, was marked by GUS activity in transposant UCR17 (Fig. 2A). In contrast, GUS activity in transposant UCR18 marked the root cortex, which consists of a single layer of large, vacuolated cells (Fig. 2B). GUS activity in this line was not completely restricted to the cortex, however, and faint staining was occasionally also detected in the epidermis. GUS activity was detected exclusively in the root endodermal layer in transposant UCR19 (Fig. 2C). Strong GUS expression was also observed in the stigmatic papillae and placenta of the pistil and in pollen tubes in UCR 19 (Table II).

View larger version (122K): [in this window] [in a new window]   Figure 2.   Enhancer trap transposants with GUS activity in the root. A, UCR17, GUS activity in epidermis. B, UCR18, GUS activity in cortex. C, UCR19, GUS activity in endodermis. D, UCR8, GUS activity in phloem and pericycle. Arrow points to a lateral root (lr). E, UCR20, GUS activity in developing xylem. F, UCR21, GUS activity in outer layers of vascular cylinder and vascular initial cells in the RAM. G, UCR9, GUS activity in central layers of the vascular cylinder, originating in the vascular initials. H, UCR22, GUS activity in the zones of elongation and differentiation, localized to the epidermis and cortex. I and J, UCR23, GUS activity in trichoblast cell files in the elongation (I) and differentiation (J) zones. K, UCR24, GUS activity in the root cap. L through N, GUS activity in the bottom two tiers (L, UCR25), middle tier (M, UCR26), and initials (N, UCR27) of the columella root cap. O, UCR28, GUS activity in the lateral root cap. P and Q, Developmental series of lateral root initiation. P, UCR29, GUS activity in the first few cells of dividing pericycle (P1). GUS activity was visible throughout the lateral root primordia (P2-P5), was progressively restricted to the root tip (P6), and was not detected in mature roots (P7). Q, UCR30, GUS activity was not detected in early stage primordia (Q1) but was visible before the lateral root primordia emerged from the primary root (Q2). GUS activity was restricted to the organizing RAM (Q3-Q4) and disappeared at later stages (Q5-Q6). Images were captured using differential interference contrast microscopy of root whole mounts. Scale bars = 50 µm. Scale bar in Q1 refers to Q1 through Q6.

Many enhancer trap lines that mark different compartments within the vascular cylinder in the root were identified, and two were chosen for inclusion in this collection. GUS activity in transposant UCR8 was detected in all vascular tissues of the plant. In the root, staining was localized to the outer layers of the vascular cylinder, corresponding to the pericycle and phloem (Fig. 2D). GUS activity in transposant UCR20 marked xylem parenchyma and differentiating xylem and was detected in the center of the vascular cylinder in the mature root (Fig. 2E). GUS activity was not detected in the root tip in either transposant line (data not shown). In transposant UCR20, high levels of GUS expression were also observed in inflorescence stem nodes, perhaps corresponding to the formation of new vasculature during bud emergence (Table II). Transposant UCR21 displayed GUS activity in the outer layers of both developing and mature vascular tissue (Fig. 2F), whereas in transposant UCR9, GUS activity was detected in the inner layers of developing vasculature (Fig. 2G) but was not detected in mature tissues (data not shown). GUS expression in both lines was detectable in root vascular precursor cells, including the initial cells of the root apical meristem (RAM) and the cell files derived from them. They therefore serve to illustrate the origin of vasculature in the root. Low levels of GUS activity were also visible in the lateral root cap in transposant UCR9 (Fig. 2G).

In addition to radial patterning, roots show distinct organization along the apical-basal axis. Elements along this axis begin at the root tip with the root cap and the RAM. The elongation zone, which exhibits a high rate of cell division and differentially elongates during gravitropism (Mullen et al., 1998), is behind the tip. The zone of differentiation or specialization is further back from the tip, and is marked by the emergence of root hairs. A number of enhancer trap lines were identified that showed distinct patterns of GUS expression along this apical-basal axis. GUS activity in transposant UCR22 was detected in cortical and epidermal cells in a region of the elongation zone near the zone of differentiation (Fig. 2H). Transposant UCR23 also showed GUS activity in the epidermis of this region, however, only alternating cell files were stained (Fig. 2I). Examination of the epidermis also revealed weak staining in differentiating root hair cells (Fig. 2J), indicating that GUS activity in UCR23 marked trichoblast cells. This line nicely illustrates the concept of root hair patterning, which has been the subject of intense research (Masucci et al., 1996; Dolan and Costa, 2001; Lin and Schiefelbein, 2001).

The root cap was marked by GUS activity in transposant UCR24 (Fig. 2K). Three lines showed root cap staining that was further restricted to the bottom two tiers of cells (UCR25; Fig. 2L), a middle tier (UCR26; Fig. 2M), and the columella initials (UCR27; Fig. 2N). Finally, GUS activity was restricted to the lateral root cap in transposant UCR28 (Fig. 2O). This transposant series demonstrates the compartmentalization of the root cap, something that is not apparent from histological analyses but has been experimentally demonstrated. For example, the lateral root cap and columella have distinct embryonic origins and may also serve distinct functions (Esau, 1965; Bowman, 1993; Dolan et al., 1994; Baum and Rost, 1996). The cells in the center of the columella have been shown to house the receptor for gravity sensing (Tanaka et al., 2002). The cells at the periphery of the root cap are continuously sloughed off and replaced, and in some plants, these cells secrete mucilage that serves to protect the root from microbial pathogens (Hawes et al., 1998) and may help to lubricate the root tip as it pushes through the soil (Esau, 1965).

Enhancer trap lines were isolated with GUS activity that was temporally regulated during lateral root development. Lateral root primordia form in the pericycle, and weak GUS activity in transposant UCR29 was detected in the pericycle during the earliest stages of lateral root primordium formation (Fig. 2, P1). The intensity of GUS staining increased as the primordium developed, becoming very strong during later stages, marking the entire primordium (Fig. 2, P2-P5). In older lateral roots, GUS activity diminished but was retained in the root cap (Fig. 2, P6). In mature roots, GUS activity was no longer detectable (Fig. 2, P7). In a second transposant line, UCR30, GUS activity was first detectable at a later stage of lateral root development, beginning in the apical regions of pre-emergent lateral root primordia (Fig. 2, Q1-Q2). As the primordium emerged, GUS staining became restricted to the root tip and was most intense at the site of the developing RAM (Fig. 2, Q3-Q4). GUS staining faded quickly, disappearing in early lateral roots (Fig. 2, Q5-Q6). The transient nature of GUS expression in both of these enhancer trap lines demonstrates that lateral root formation is an extremely dynamic process.

Complex Patterns of Gene Regulation

In addition to providing visual anatomical markers, enhancer trap lines that demonstrate more complex patterns of gene expression were also chosen for inclusion in this collection. These patterns may serve to make connections between seemingly unrelated tissues or cell types and may not always coincide with known morphological features. For example, transposant UCR2 showed GUS activity in trichomes (Fig. 1B) and in root hairs (Table I), suggesting that trichomes and root hairs share a common feature. Both cell types undergo tip growth, and in addition, mutational analyses have demonstrated the use of common regulatory genes that function as cell fate determinants during the development of both trichomes and trichoblast cells (for review, see Dolan and Scheres, 1998).

A number of enhancer trap lines were identified that showed GUS expression in an intriguing pattern at the base of all lateral organs. Molecular characterization of one such line identified the LATERAL ORGAN BOUNDARIES gene, the founding member of a large family of novel genes (Shuai et al., 2002). An example of such an expression pattern is represented by transposant UCR11. GUS activity in UCR11 was detected in a single ring of epidermal cells at the base of each leaf primordium (Fig. 1K) and at the base of all lateral organs in the inflorescence (Table I). GUS activity was detected at the base of young primordia and persisted in mature leaves. In transposant UCR27, GUS activity, which marked the columella initials (Fig. 2N), was also observed in a ring of cells encircling leaf bases (Table II).

In transposant UCR12, GUS activity was detected in trichomes and in the marginal cells of leaf petioles (Fig. 1L). The marginal staining was strong at the leaf base, and disappeared completely near the blade. GUS activity in UCR12 was also present in the hypocotyl, stipules, anthers, and root vasculature (Fig. 1L; Table I). No staining was detected in the veins of leaves or cotyledons, however. The significance of this staining pattern is currently unknown, and it has been provided as a representative of the many complex expression patterns detected by enhancer traps.

Seedlings of transposant UCR13 displayed GUS activity in the cotyledon petioles and hydathodes, but staining was otherwise excluded from the cotyledon blade (Fig. 1M). GUS activity was also detected in the upper hypocotyl, but was not detected in leaf primordia. After flowering, GUS activity was detected in the inflorescence stem and the carpels of the flower but was excluded from other floral organs. In transposant UCR14, the veins in the hypocotyl and petiole were marked by GUS activity, but other veins were unstained (Fig. 1N). Veins in the inflorescence stem, flower pedicel, style, and silique valve were also marked by GUS activity (Table I).

In transposant UCR15, GUS activity marked the cotyledons, hypocotyl, and the majority of the root, including the root cap (Fig. 1O; Table I). GUS activity was excluded from leaf primordia (Fig. 1O), RAM, and elongation zone (data not shown). This expression pattern appeared to correspond to regions that did not contain rapidly dividing cells. After flowering however, strong GUS activity in UCR15 was observed in most tissues of the inflorescence, including young flower buds (Table I), indicating that GUS expression isn't simply excluded from rapidly dividing cells. Transposant UCR16 was selected for its prominent staining of leaf primordia. GUS activity was detected only in the blade regions of leaf primordia (Fig. 1P). Staining faded slightly in older expanding leaves, and remained detectable in mature leaves and cotyledons. GUS activity was also detected in developing, but not mature anthers, suggesting a transient role in anther development (Table I). Weak GUS activity was also detected in the cortical cells of the collet, the interface of root and hypocotyl (Table I).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS AND DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Transposant Lines

A collection of enhancer and gene trap lines was generated using previously described methods (Sundaresan et al., 1995). All transposants have been made publicly available through the Arabidopsis Biological Resource Center (Ohio State University, Columbus; http://www.Arabidopsis.org/abrc).

Plant Growth

For growth on sterile media, seed was surface sterilized by treating with 95% ethanol for 5 min, followed by treatment with a solution of 20% household bleach (1% sodium hypochlorite) and 0.1% Tween 20 for 5 min, followed by three rinses in sterile water. The sterilized seed was transferred to germination medium containing 0.43% Murashige and Skoog salts (Invitrogen, Carlsbad, CA), 1% Suc, and 0.8% agar, adjusted to pH 5.7 with 1.0 N KOH. Plants were grown to maturity in Sunshine Mix no. 2 (Sun Gro Horticulture, Alberta, Canada) supplemented with 14-14-14 Osmocote (Scotts, Marysville, OH) at a rate of 75 g per cubic foot and Marathon systemic insecticide (Olympic Horticulture, Mainland, PA) at a rate of 25 g per cubic foot. Imbibed seeds were cold-treated at 4°C for 4 d and then transferred to a growth chamber at 22°C with a light intensity of 200 microeinsteins m² s¹, and a 16-h light/8-h dark cycle.

Histochemical Localization of GUS Activity

Tissue was stained for GUS activity in staining solution containing 100 mM sodium phosphate, pH 7.0, 10 mM EDTA, 0.1% Triton X-100, 1 mg mL¹ 5-bromo-4-chloro-3-inolyl--D-GlcUA, cyclohexylammonium salt (X-Gluc, Biosynth International, Naperville, IL), 100 µg mL¹ chloramphenicol, 2 mM potassium ferricyanide, and 2 mM potassium ferrocyanide as previously described (Sundaresan et al., 1995). In brief, tissue was harvested directly into a volume of staining solution sufficient to cover the tissue and placed under house vacuum for 10 min. Staining was carried out at 37°C in the dark for 48 h. After staining, the tissue was incubated in 70% ethanol to remove the chlorophyll. The ethanol was changed several times until the tissue was clear. Stained tissues were examined using a stereomicroscope or mounted on a glass slide under a cover glass in 50% glycerol, and monitored using differential interference contrast optics on a Leica DMR compound microscope. Images were captured using a Spot digital camera (Diagnostic Instruments, Sterling Heights, MI). Tissues and cell types were identified based on comparison with published reports (Esau, 1965; Sachs, 1991; Bowman, 1993; Meyerowitz and Somerville, 1994).