INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Immediately following cell division in higher plants, the two daughter cells are attached to each other as a result of the formation of the cell plate. These cells then maintain cell-to-cell attachment or separate from each other. The phenomenon in which separated cells epigenetically re-adhere, as seen in animal systems, is not typically observed in plants, but does occur in certain processes such as carpel fusion during gynoecium development (Walker, 1975; Siegel and Verbeke, 1989; van der Schoot et al., 1995), tissue union during grafting (Kollmann and Glockmann, 1985; Richardson et al., 1996; Wang and Kollmann, 1996), and cell repair in cut tissues. In Arabidopsis, mutants with ectopic fusion or adhesion of aerial tissues have been identified, including fiddlehead (Lolle et al., 1992, 1997; Lolle and Cheung, 1993; Lolle and Pruitt, 1999) and wax-1 (Jenks et al., 1996).

Studies on graft union and repair in cut tissues have focused on the differentiation of vascular elements in the tissue reunion process because the formation of the vascular bundle is a useful model system for studying cell differentiation and organization in higher plants (Stoddard and McCully, 1980; Moore and Walker, 1981; Kollmann and Glockmann, 1985; Monzer and Kollmann, 1986; Roberts, 1988; Tiedemann, 1989; Sachs, 2000). Although the molecular mechanisms controlling vascular differentiation are not yet fully understood, the involvement of phytohormones such as auxin and cytokinin in xylem and phloem differentiation has been suggested (Roberts, 1988; Mattsson et al., 1999; Sachs, 2000). However, the process of reunion in the cortex of cut tissues has not been analyzed.

In Japan, cucumber (Cucumis sativus) is often grafted onto squash (Cucurbita maxima Duchesne × C. moshata Duchesne) stock to prevent damage from soil-borne diseases during cultivation (Satoh, 1996). In this grafting procedure, the hypocotyls of the cucumber scion and squash stock are cut to two-thirds of their diameters in an upward or downward direction, respectively, and the cut ends are brought into contact and fixed with a clip. In this procedure, the apical tip and first leaf of the squash stock are removed, but the cotyledons of the scion and stock are preferentially left on the hypocotyl to improve grafting efficiency. Although the exact role of the cotyledon in the formation of the graft union is not understood, it is possible that the cotyledon produces compounds required for the formation of the graft union.

In this study, we performed morphological and histochemical analyses of the tissue reunion process in the cortex of cut cucumber and tomato (Lycopersicon esculentum) hypocotyls using light and transmission electron microscopy. We show that active pectin biosynthesis occurs in this process. Our results also suggest that gibberellin (GA), likely produced in the cotyledon, is required for the cell division during tissue reunion of the cortex in cucumber and tomato hypocotyls.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Morphological Analysis of Tissue Reunion by Light Microscopy

Cucumber hypocotyls were cut transversely to a depth of approximately one-half of their thickness (Fig. 1), and the following tissue reunion process in the cortex was observed by light microscopy. Mucus-like substances were stained with toluidine blue O in the cut surface immediately after the cutting (Fig. 2A, black arrows). One day after cutting, a wall-like structure was evident by toluidine blue O staining in the cut surface (Fig. 2B, white arrows). Three days after cutting, cortex cells near the cut surface initiated transverse cell division and longitudinal cell elongation toward the cut surface, and a layer of cell wall formed where the adjoining cortex cells attached to each other (Fig. 2C, white arrowheads). Randomly directed cell division and intrusive cell elongation subsequently occurred (Fig. 2D, asterisks). The cortex cells near the cut surface were interspersed (Fig. 2E, black stars). Further visible changes were not observed more than 7 d after cutting (data not shown). The morphological changes in the cortex are schematically summarized in Figure 2, A' to E'.

View larger version (33K): [in this window] [in a new window]   Figure 1.   Schematic illustration of how the hypocotyls were cut. The hypocotyls of 7-d-old plants were cut to one-half of their diameter, transversely, 3 cm from the base, using a razor blade (0.1 mm thick). The plants were then grown for an additional 10 d.

View larger version (76K): [in this window] [in a new window]   Figure 2.   Light micrographs (A-E) and schematic illustrations (A'-E') of the tissue reunion process in the cut hypocotyls of cucumber. A and A', Immediately following cutting of the hypocotyl. Black arrows indicate mucus-like substances stained with toluidine blue O. B and B', One day after cutting. White arrows indicate a wall-like structure intensely stained with toluidine blue O. C and C', 3 d after cutting. White arrowheads indicate a layer of cell wall where the confronted cortex cells attach to each other. D and D', 5 d after cutting. Asterisks indicate randomly directed cell division and intrusive cell elongation. E and E', 7 d after cutting. Black and white stars indicate cells in the tissue reunion region and cells in the non-reunion region, respectively. All sections were stained with toluidine blue O. Arrowheads indicate the location of the cut. co, Cortex; vb, vascular bundle. Scale bars indicate 100 µm.

4',6-Diamidino-2-Phenylindole (DAPI) Staining and 5-Bromo-2'-Deoxyuridine (BrdU) Labeling

To verify cell division in the tissue reunion process, the localization of nuclei was first visualized by DAPI staining as blue-white fluorescence, and then DNA synthesis in the nuclei was analyzed as green-yellow fluorescence by feeding the plant BrdU.

In the tissue reunion region, cells with DAPI-stained nuclei increased from 3 to 7 d after cutting the hypocotyl (Fig. 3, A-C). At 5 d, some of these nuclei were strongly labeled with BrdU (Fig. 3, G and H, white arrows). These results directly indicate that DNA synthesis occurred in the tissue reunion region. On the other hand, in the non-reunion region, few DAPI-stained nuclei were detected from 3 to 7 d after cutting the hypocotyl (Fig. 3, D, E, and J) and BrdU labeling was not detected in the nucleus (5 d; Fig. 3, I and J, white arrows).

View larger version (166K): [in this window] [in a new window]   Figure 3.   DAPI staining (A-G and I) and BrdU labeling followed by indirect immunofluorescence microscopy with anti-BrdU antibody (H and J). Sections were prepared from cells from the tissue reunion (A-C, G, and H) and non-reunion (D-F, I, and J) regions. A and D, 3 d after cutting the hypocotyl; B, E, G, and I, 5 d after cutting the hypocotyl; C and F, 7 d after cutting the hypocotyl. G and H, Magnified image of the area marked in B. White arrows indicate BrdU-labeled nuclei. I and J, Magnified image of the area marked in E. White arrowheads indicate the position of the cut. White arrowheads indicate the layer of cell wall where the confronted cortex cells attach to each other. Organelle DNA is seen as small dots. Scale bars indicate 50 µm.

The cell wall layer that formed where adjoining cortex cells attached to each other (Fig. 2C, white arrowheads) showed intense fluorescence under UV and blue light excitation at 3 d (Fig. 3A, white arrowheads). This was derived from autofluorescence of the cell wall, and did not reflect the localization of the nucleus and DNA. This autofluorescence weakened at 5 d (Fig. 3B, white arrowheads), and was barely detected at 7 d.

Transmission Electron Microscopy

To analyze the pectic substances, which are involved in intercellular attachment in higher plants, histochemical analysis in the tissue reunion region of the cortex was carried out using ruthenium red (RR; Iwai et al., 1999). RR is a cationic dye with six positive charges that form electrostatic bonds to the acidic groups of sugars, for example, carboxyl and sulfate groups (Lillie and Fullmer, 1976). Moreover, because RR also acts as a catalyst, saccharides are eventually oxidized, with a simultaneous reduction of OsO₄. Insoluble products with high electron densities are generated. The methyl groups of pectin can be removed and carboxyl groups can be restored by treatment with alkaline agents such as 1 N NaOH; increased labeling with antibodies that recognize demethylesterified pectin occurs after alkali treatments (Schindler et al., 1995). Thus, RR staining should reflect levels of nonmethylesterified pectic substances, and comparison of RR staining with and without alkali treatments should allow localization of methylesterified pectic substances (Iwai et al., 1999).

Sections prepared from cortex cells of regions undergoing tissue reunion (Fig. 4, A-C; indicated as black stars in Fig. 2E) and cells from non-reunion regions (Fig. 4, D-F; indicated as white stars in Fig. 2E) at 7 d after cutting were stained with OsO₄ (Fig. 4, A and D), double-stained with OsO₄ and RR (Fig. 4, B and E), or double-stained with OsO₄ and RR after treatment with 1 N NaOH (Fig. 4, C and F). The cell wall in the reunion region contained striped structures in which the membrane-like materials overlapped, and was much thicker than the cell wall in the non-reunion region (Fig. 4, A and D, CW). The striped structures were strongly stained by RR (Fig. 4B, CW, black arrows), and the entire cell wall structure in this region was intensely stained by RR after 1 N NaOH treatment (Fig. 4C, CW, black arrows). However, cell walls in the non-reunion region were stained by RR only in the middle lamella, even after treatment with 1 N NaOH (Fig. 4, E and F, CW, black arrows).

View larger version (109K): [in this window] [in a new window]   Figure 4.   Transmission electron micrographs of cell walls. Sections prepared from cells in the tissue reunion region (A-C; shown as black stars in Fig. 2E) and cells in the non-reunion region (D-F; shown as an white star in Fig. 2E) were stained with OsO4 (A and D), double-stained with OsO4 and RR (B and E), or double-stained with OsO4 and RR after 1 N NaOH treatment (C and F). All micrographs were taken 7 d after the hypocotyl was cut. CP, Cytoplasm; M, mitochondria; CW, cell wall. Arrows indicate cell wall stained with RR. Scale bars indicate 500 nm.

Effects of Organ Removal and Phytohormones

Involvement of organs and phytohormones in the tissue reunion process was further examined. Cell division in the cortex undergoing tissue reunion was strongly inhibited in the hypocotyl of a plant whose cotyledon had been removed (Fig. 5A, black arrow). In contrast, normal tissue reunion occurred in a plant whose shoot apex had been removed (data not shown). In the cotyledon-excised plant, cell division was restored by the application of gibberellic acid (GA₃) to the shoot apex and the cut surface of the cotyledon (Fig. 5B, black arrow). Cell division was not restored by indole-3-acetic acid (IAA) or by distilled water (D.W.; Fig. 5, C and D, black arrow).

View larger version (114K): [in this window] [in a new window]   Figure 5.   Effects of phytohormones on cell division during the tissue reunion process in cotyledon-removed plants. The cotyledon was excised from a 7-d-old plant (A), and a lanolin paste containing GA3 (B), IAA (C), or D.W. (D) was applied to cover the shoot apex and cut surface of the cotyledon. The final concentration of the phytohormones was 104 M. Black arrows indicate cells in the cortex of the cut hypocotyl. Black arrowheads indicate the position of the cut. All sections were stained with toluidine blue O. co, Cortex; vb, vascular bundle. Micrographs were taken 7 d after cutting. Scale bars indicate 100 µm.

The effects of phytohormone inhibitors on cell division in the cortex during the process of tissue reunion were further investigated. Cell division in the cortex was strongly inhibited when uniconazole-P, an inhibitor of GA biosynthesis (Izumi et al., 1985), was sprayed onto the cotyledon (Fig. 6A, black arrow), but was not inhibited when Tween 20 was sprayed (Fig. 6C). Simultaneous application of GA₃ and uniconazole-P restored cell division inhibited by uniconazole-P (Fig. 6B, black arrow). Normal tissue reunion occurred in the cortex of the hypocotyl when 10⁴ M 2,3,5-triiodobenzoic acid (TIBA), an inhibitor of polar auxin transport, or D.W. was applied to the surface of the shoot above the cut on the hypocotyl (Fig. 6, D and E).

View larger version (100K): [in this window] [in a new window]   Figure 6.   Effects of phytohormone inhibitors on cell division during tissue reunion. Uniconazole-P (104 M) plus 0.2% (v/v) Tween 20 was sprayed onto the abaxial surface of the cotyledon of 5-d-old plants once a day (A and B). After 2 d of treatment, uniconazole-P alone (A) or with 2 × 104 M GA3 (B) or 0.2% (v/v) Tween 20 (C) was sprayed once a day. Lanolin paste containing 104 M TIBA (D) or D.W. (E) was applied above the above the cut on shoots of 7-d-old plants. Black arrows indicate cells in the cortex of the cut hypocotyl. Black arrowheads indicate the position of the cut. All sections were stained with toluidine blue O. co, Cortex; vb, vascular bundle. Micrographs were taken 7 d after the hypocotyl was cut. Scale bars indicate 100 µm.

Tissue Reunion in the Tomato GA-Deficient Mutant (gib-1)

The involvement of endogenous GA in cell division in the cortex during the process of tissue reunion was investigated using a tomato GA-deficient mutant (gib-1; Groot et al., 1987; Rebers et al., 1999). In wild-type tomato seedlings, cell division in cortex undergoing tissue reunion (Fig. 7A) was strongly inhibited by removing the cotyledon (Fig. 7B), and the application of GA₄ restored cell division (Fig. 7C). Because similar results were obtained with tomato and cucumber seedlings, we further analyzed the process of cortex tissue reunion in a tomato GA-deficient mutant (gib-1). In this mutant, cell division did not occur in the process of cortex tissue reunion (Fig. 7D), and cell division was restored by microdrop application of GA₄ to the shoot apex and base of the cotyledon (Fig. 7E).

View larger version (93K): [in this window] [in a new window]   Figure 7.   Tissue reunion in wild-type (A-C) and GA-deficient mutant (gib-1; D and E) tomato. The hypocotyl of each plant was cut to one-half using the same methods as for cucumber hypocotyls and was observed 7 d after cutting the hypocotyl. A, Wild-type plant. B, Cotyledon-removed plant. C, Cotyledon-removed plant with lanolin paste containing 104 M GA4 applied to cover the shoot apex and cut surface of the cotyledon. D, gib-1. E, gib-1 with microdrop application of 104 M GA4 to the shoot apex and base of the cotyledon. Black arrowheads indicate the position of the cut. All sections were stained with toluidine blue O. Scale bars indicate 100 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Light microscopy of the cortex in the cut hypocotyl revealed that cell division and elongation began 3 d after cutting, and that the cortex was nearly completely united within 7 d (Fig. 2). DAPI staining and BrdU labeling experiments indicate that nDNA syntheses also occurred in this process (Fig. 3).

Spatially separated cells re-adhere in the reunion region, and the morphology of the cell wall in the reunion region (Fig. 2E, black stars) differs from that in the non-reunion region (Fig. 2E, white star), indicating that specific events must occur during intercellular attachment in the cell walls of the tissue reunion region.

Histochemical analysis in the tissue reunion region of the cortex was carried out to analyze pectic substances, which are involved in intercellular attachment in higher plants. In the tissue reunion region, RR staining of the cell walls of attached cells indicated an abundance of a nonmethylesterified pectic substance in the attached region (Fig. 4B, black arrows). Moreover, the strong RR staining of entire cell walls after alkali treatment (Fig. 4C, black arrows) indicates an abundance of methylesterified pectic substances. The pectic polysaccharides in the walls of young or actively growing cells are highly methylesterified, whereas the walls of mature cells contain strongly acidic pectins (Yamaoka and Chiba, 1983; Asamizu et al., 1984; Goldberg et al., 1986). These results suggest that active biosynthesis and accumulation of pectic substances occurs in the cell wall of attached cells in the reunion region of the cortex.

The results of experiments involving the excision of organs and the application of phytohormones and their inhibitors (Figs. 5 and 6) clearly suggest that GA is involved in cell division during tissue reunion in the cortex, and that the cotyledon is important in supplying GA. Because uniconazole-P inhibited the oxidation of ent-kaurene, and GA biosynthesis was blocked immediately after ent-kaurene, contents of biologically active GAs and their precursors after ent-kaurene were remarkably reduced in the cotyledon after treatment with uniconazole-P (Izumi et al., 1985). Moreover, the experiments with seedlings of the GA-deficient tomato mutant (Fig. 7) support this hypothesis. Also, in cotyledons of 7-d-old tomato seedlings, biologically active GAs (GA₁ and GA₄) and their precursors (GA₂₄, GA₁₉, GA₂₀, GA₄₄, GA₁₂, and GA₅₃) were identified by gas chromatography-mass spectrometry (data not shown).

Auxin and cytokinin are known to be critical in directing the formation of new vascular strands when the vasculature has been separated by wounding or by grafting (Mattsson et al., 1999; Sachs, 2000). GA controls various aspects of plant development such as germination, stem elongation, flowering, and fruit set and development by promoting cell division or cell elongation (Stuart et al., 1977; Groot et al., 1987; Cosgrove and Sovonick-Dunford, 1989; Toyomasu et al., 1998; Rebers et al., 1999; van den Heuvel et al., 1999; Yamaguchi and Kamiya, 2000); to our knowledge, this is the first report of the involvement of GA in cell division in the tissue reunion process.

Plants show various responses to mechanical damage. It is well known that various compounds such as ethylene, abscisic acid, jasmonic acid, oligopeptide systemin, and oligosaccharides, and other physical factors are involved in the wound-signal transduction pathway, and cross talk between some signaling pathways results in a different pattern of responses (Leon et al., 2001). How wound-signal transduction is involved in the tissue reunion process, especially in the production, transportation, and action of GAs, remains to be examined in the future.

In this study, we indicated that GA production in the cotyledon is involved in cell division during tissue reunion in the cortex of cucumber and tomato cut hypocotyls. The physiological and molecular mechanisms of the tissue reunion process, including the mechanisms of GA production and transportation, require further study.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Materials and Growth Conditions

Seeds of cucumber (Cucumis sativus cv Shimoshirazu-jibai) were obtained from Sakata Seed Co. (Kanagawa, Japan). The seeds were germinated and grown in artificial soil (Kurehakagaku, Tokyo) under white fluorescent light (32 µmol m² s¹) with 16-h days at 28°C. After 7 d of growth, the hypocotyl was cut to one-half of its diameter transversely 3 cm from the base, using a razor blade (0.1-mm thickness), and the plant was then grown as above for an additional 10 d.

Tomato (Lycopersicon esculentum) seeds of wild-type cv Moneymaker and gib-1 mutant (deficient in copalyl diphosphate synthase activity, and in the same genetic background as the wild type; Groot et al., 1987; Rebers et al., 1999) were obtained from Dr. Maarten Koornneef (Department of Genetics, University of Wageningen, Wageningen, The Netherlands). Because the endogenous level of GA was remarkably reduced (Groot et al., 1987), seeds of gib-1 were germinated in artificial soil supplied with 10 µM GA₄₊₇ for 3 to 5 d, and the seedlings were planted in artificial soil after germination and grown as described above. After 10 d of growth, the hypocotyl was then cut with the same methods as the cucumber.

Light Microscopy of Cut Cucumber Hypocotyl

Cut hypocotyls at various stages (0, 1, 3, 5, 7, and 10 d after cutting) were trimmed to 5-mm segments surrounding the cut surface. Samples were fixed in 2.5% (v/v) glutaraldehyde and 2% (w/v) paraformaldehyde in 0.1 M cacodylate buffer, pH 7.4, for 2 h at 4°C. After fixation, samples were washed in the same buffer, dehydrated in a graded ethanol series, and embedded in Technovit 7100 resin (Kulzer and Co., Werheim, Germany). The sections were prepared using an ultramicrotome glass knife (Reichert EM-ULTRACUT; Leica, Wetzlar, Germany). Sections were stained with 0.1% (w/v) toluidine blue O. Observations were made with a light microscope (DMRB; Leica).

DAPI Staining and BrdU Labeling

Cucumber plants at various stages were carefully removed from the soil, and the roots were put into 5-mL microtubes (Abbott Laboratories, North Chicago) and cultured with 50 µM BrdU in the presence of 1 µM 5-fluorodeoxyuridine overnight at 28°C. The labeled hypocotyl was fixed and embedded as described above. After sections were prepared, DAPI staining and BrdU labeling were carried out as described by Suzuki et al. (1995), with the following modifications. Sections on glass slides were soaked in 2 N HCl for 20 min to partially denature the DNA on the surface of the sections. After denaturing, samples were incubated with a mouse anti-BrdU antibody (Roche Molecular Biochemicals, Indianapolis) and Alexa 488-conjugated goat anti-mouse secondary antibody (Molecular Probes, Eugene, OR). The sections were finally stained with 1 µg mL¹ DAPI. The double-stained sections were observed under a fluorescence microscope (DMRB; Leica). DAPI-staining nuclei or BrdU-labeling DNA was detected as blue-white fluorescence under UV excitation or green-yellow fluorescence under blue light excitation, respectively.

Transmission Electron Microscopy

Seven days after cutting, the hypocotyls of cucumber were trimmed to 2-mm segments around the cut surface. Samples were fixed in 2.5% (v/v) glutaraldehyde and 2% (w/v) paraformaldehyde in 0.1 M cacodylate buffer, pH 7.4, for 2 h at 4°C. Samples were washed in the same buffer and post-fixed in 2% (w/v) OsO₄ in the same buffer overnight at 4°C. After fixation, samples were washed in the same buffer, dehydrated in a graded ethanol series, and embedded in Spurr's resin (TAAB, Aldermaston, Berkshire, UK). Ultrathin sections were prepared using a diamond knife on an ultramicrotome, and they were placed on copper grids. Sections on grids were double-stained with saturated uranyl acetate and Reynolds' lead citrate (Reynolds, 1963). Observations were made using a transmission electron microscope (JEM 100 CX-II; JEOL, Tokyo).

Staining of Pectic Substances with RR

Staining with RR and treatment with alkali were carried out as described previously (Iwai et al., 1999), with the following modifications. Samples were fixed in 2.5% (v/v) glutaraldehyde and 2% (w/v) paraformaldehyde plus 500 mg L¹ RR for 2 h at 4°C. Samples were washed in the same buffer and post-fixed in 2% (w/v) OsO₄ plus 500 mg L¹ RR overnight at 4°C. For alkali treatments, after prefixation with glutaraldehyde, paraformaldehyde, and RR, samples were washed in the same buffer and were then incubated in 1 N NaOH for 2 h at room temperature. After washing in the same buffer, samples were post-fixed with OsO₄ and RR overnight at 4°C.

Removal of Organs and Treatment with Phytohormones

The cotyledon or shoot apex, including the first leaf, was removed from 7-d-old cucumber plants using a razor blade. The hypocotyl was then cut in one-half as described above, and the plants were grown as above.

Lanolin paste containing GA₃, IAA, or D.W. was applied to the apical tip of cotyledon-removed plants to cover the shoot apex and cut surface of the cotyledon. The lanolin pastes were prepared by adding anhydrous lanolin to solutions of GA₃, IAA, or D.W. (3:1, v/v), and the final concentration of GA₃ and IAA was adjusted to 10³ M or 10⁴ M. The hypocotyl was then cut, and observations were made after 7 d of culture as described above.

Treatment with Phytohormone Inhibitors

Lanolin paste containing 10⁴ M TIBA, an inhibitor of polar auxin transport, was applied above the cut on shoots of 7-d-old cucumber plants. The hypocotyl was cut, and observations were made after 7 d, as described above. Five days after cutting, a solution of Tween 20 (0.2%, v/v) containing 10⁴ M uniconazole-P, an inhibitor of GA biosynthesis, with or without 2 × 10⁴ M GA₃, was applied to the abaxial surface of the cotyledon once a day. After 2 d of treatment, the hypocotyl was cut, and observations were made after 7 d of culture as described above.

Analysis of Tissue Reunion in Cut Tomato Hypocotyl

The cotyledon was removed from 10-d-old wild-type tomato plants using a razor blade, and the plants were grown as described above. Lanolin paste containing 10⁴ M GA₄ was applied to the tip of cotyledon-removed plants to cover the shoot apex and cut surface of the cotyledon. Observations were made after 7 d, as described above.

The hypocotyl of the 10-d-old gib-1 plant was cut, and GA₄ solution (10⁴ M) was applied to the shoot apex and the base of cotyledon once a day (10 µL d¹). Observations were made after 7 d of culture, as described above.