Correlation between development of female flower buds and expression of the CS-ACS2 gene in cucumber plants

Transcription

1 Journal of Experimental Botany, Vol. 58, No. 11, pp , 2007 doi: /jxb/erm141 Advance Access publication 13 July, 2007 This paper is available online free of all access charges (see for further details) RESEARCH PAPER Correlation between development of female flower buds and expression of the CS-ACS2 gene in cucumber plants Sayoko Saito 1, Nobuharu Fujii 1, Yutaka Miyazawa 1, Seiji Yamasaki 2, Seiji Matsuura 3, Hidemasa Mizusawa 3, Yukio Fujita 3 and Hideyuki Takahashi 1, * 1 Graduate School of Life Sciences, Tohoku University, Katahira, Aoba-ku, Sendai , Japan 2 Fukuoka University of Education, 1-1 Akamabunkyomachi, Munakata, Fukuoka , Japan 3 Tohoku Seed Co., Ltd., 1625 Nishihara, Himuro-cho, Utsunomiya , Japan Received 9 January 2007; Revised 2 May 2007; Accepted 29 May 2007 Abstract Ethylene plays a key role in sex determination of cucumber flowers. Gynoecious cucumber shoots produce more ethylene than monoecious shoots. Because monoecious cucumbers produce both male and female flower buds in the shoot apex and because the relative proportions of male and female flowers vary due to growing conditions, the question arises as to whether the regulation of ethylene biosynthesis in each flower bud determines the sex of the flower. Therefore, the expression of a 1-aminocyclopropane-1- carboxylic acid synthase gene, CS-ACS2, was examined in cucumber flower buds at different stages of development. The results revealed that CS-ACS2 mrna began to accumulate just beneath the pistil primordia of flower buds at the bisexual stage, but was not detected prior to the formation of the pistil primordia. In buds determined to develop as female flowers, CS-ACS2 mrna continued to accumulate in the central region of the developing ovary where ovules and placenta form. In gynoecious cucumber plants that produce only female flowers, accumulation of CS-ACS2 mrna was detected in all flower buds at the bisexual stage and at later developmental stages. In monoecious cucumber, flower buds situated on some nodes accumulated CS-ACS2 mrna, but others did not. The proportion of male and female flowers in monoecious cucumbers varied depending on the growth conditions, but was correlated with changes in accumulation of CS-ACS2 mrna in flower buds. These results demonstrate that CS-ACS2-mediated biosynthesis of ethylene in individual flower buds is associated with the differentiation and development of female flowers. Key words: ACC synthase, 1-aminocyclopropane-1-carboxylic acid (ACC), CS-ACS2, cucumber, Cucumis sativus L., ethylene, gynoecious, monoecious, sex expression. Introduction Cucumber plants usually produce separate male and female flowers in the same individual, but some cucumber lines produce bisexual flowers. Cucumber flower buds differentiate as lateral organs on the leaf axils. In general, the initiation of flower bud differentiation is observed on nodes situated three or four nodes below the apical meristem in the cucumber shoot (Fujieda, 1966). In cucumber flower buds, primordia of the sepals, petals, stamens, and pistils are formed centripetally from the outside (Atsmon and Galun, 1960; Malepszy and Niemirowicz-Szczytt, 1991). At the early stages of development, flower buds contain primordia of both stamen and pistil, and sex determination occurs due to the selective arrest of development of either the staminate or pistillate primordia just after the bisexual stage (Atsmon and Galun, 1960; Malepszy and Niemirowicz-Szczytt, 1991). Continued stamen development and arrest of pistil development results in male flowers, whereas continued pistil development and arrest of stamen development results in female flowers. Sex determination occurs in flower buds situated at nodes below the apical meristem irrespective of the age of the plant (Fujieda, 1966). Depending on the ratio of male, female, and bisexual flowers produced by the plant, cucumber plants are * To whom correspondence should be addressed. hideyuki@ige.tohoku.ac.jp ª 2007 The Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

2 2898 Saito et al. classified into five genotypes: monoecious, gynoecious, andromonoecious, hermaphrodite, and androecious (Galun, 1961; Shifriss, 1961; Kubicki, 1969; Malepszy and Niemirowicz-Szczytt, 1991). The most common cucumbers, the monoecious line, produce both male and female flowers. Gynoecious plants produce only female flowers. The andromonoecious line produces male and bisexual flowers, whereas the hermaphrodite line produces only bisexual flowers. Two major genes, F and M, control these sex expression phenotypes in cucumber (Galun, 1961; Shifriss, 1961; Kubicki, 1969). That is, M-F-, M-ff, mmff, and mmf- confer gynoecious, monoecious, andromonoecious, and hermaphrodite phenotypes, respectively. The F gene is relatively dominant and promotes the production of female flowers. The M gene acts on the production of female flowers, and plants with recessive m alleles (mm) produce bisexual flowers. In addition to these genotypes, androecious cucumbers that produce only male flowers have been reported (Galun, 1961). The A gene acts downstream of the F gene, and plants become androecious when both A and F are recessive (aaff). The genetically controlled sex expression in cucumbers varies in response to environmental and hormonal cues, and ethylene is a major regulator responsible for the sex of cucumber flowers (reviewed in Frankel and Galun, 1977; Yin and Quinn, 1995; Perl-Treves, 1999; Yamasaki et al., 2005). In particular, plants or shoot apices of gynoecious lines produce more ethylene than monoecious lines (Rudich et al., 1972a, b, 1976; Fujita and Fujieda, 1981; Trebitsh et al., 1987), and monoecious and andromonoecious lines treated with ethylene or an ethylene-releasing reagent produce increased numbers of female and bisexual flowers, respectively (MacMurray and Miller, 1968; Iwahori et al., 1969; Shannon and De La Guardia, 1969). In addition, treatment of cucumber plants with compounds that inhibit ethylene biosynthesis or that block ethylene signalling resulted in a decrease in the number of female or bisexual flowers (Beyer, 1976; Atsmon and Tabbak, 1979; Takahashi and Suge, 1980; Takahashi and Jaffe, 1984). The F locus governs the ethylene biosynthesis that is correlated with the development of female flowers (Trebitsh et al., 1997; Mibus and Tatlioglu, 2004; Knopf and Trebitsh, 2006). Recently, it was proposed that ethylene signals influence the M gene product and thereby inhibit stamen development in flower buds (Yin and Quinn, 1995; Yamasaki et al., 2001). Because the ability to produce and respond to ethylene appears to account for the different sex expression genotypes in cucumber as described above, it was hypothesized that the site of ethylene biosynthesis for sex differentiation resides in the flower buds (Yamasaki et al., 2001, 2003b). In monoecious cucumbers in particular, ethylene biosynthesis is predicted to correspond with the appearance of female flowers. Because sex expression in monoecious cucumber dramatically varies depending on node position of the main stem and environmental cues such as temperature and photoperiod (Nitsch et al., 1952; Ito and Saito, 1960; reviewed in Frankel and Galun, 1977), it is a particular interest to know whether those factors affect ethylene production of each flower bud within the shoot apices. The hypothesis by Yamasaki et al. (2001) suggests that monoecious (M-ff) cucumbers develop two types of flower buds; one produces more ethylene and the other produces less ethylene. However, this hypothesis remains to be proven, and no evidence for the regulation of ethylene biosynthesis by flower buds is available at present. That is probably because measuring ethylene biosynthesis in individual flower buds during the early stages of development is technically difficult, and whole plants or shoot apices have been used for the measurement of ethylene biosynthesis in studies of the sex differentiation of cucumber flowers. To date, it has been proved that 1-aminocyclopropane-1-carboxylic acid (ACC) is an immediate precursor of ethylene (Adams and Yang, 1979) and that the expression of ACC synthase genes (ACS) generally correlates with ethylene biosynthesis in plants. Kamachi et al. (1997, 2000) identified CS-ACS2 and showed that mrna accumulation in shoot apices was correlated with the genotypes of sex expression in cucumber. That is, CS-ACS2 expression is more abundant in the shoots of gynoecious cucumbers than monoecious cucumbers and occurs mainly in female flower buds (Kamachi et al., 2000; Yamasaki et al., 2003b). However, it remains to be determined whether variations in the sex proportion of male and female flowers in shoot apices of monoecious cucumber are accompanied by changes in the expression of the CS-ACS2 gene in flower buds at the time of sex determination. The relationship of CS-ACS2 expression with sex differentiation therefore needs to be studied in individual flower buds at different developmental stages. To understand the role of ethylene biosynthesis by individual flower buds in sex differentiation and the fluctuations in the proportion of female flowers in monoecious cucumber plants, the relationship between the expression of CS-ACS2 in flower buds and the development of female flowers in monoecious (MMff) cucumber plants was examined, and these expression patterns compared with that of gynoecious (MMFF) plants. The results suggest that the level of CS-ACS2 expression in flower buds is correlated with the appearance of female flowers in monoecious cucumber, and that flower buds of gynoecious cucumber maintain a constant high level of CS-ACS2 expression at the time of sex determination. Materials and methods Plant growth and evaluation of sexuality Seeds of monoecious RS-M (MMff) and gynoecious RS-G (MMFF) cucumbers (Cucumis sativus L.) were provided by the Tohoku Seed

3 Sex differentiation of cucumber flowers 2899 Co., Ltd. (Utsunomiya, Japan). These two cucumbers are near isogenic lines with the exception of the F gene. Cucumber seeds were germinated on wet filter paper in a Petri dish at 28 C in the dark overnight. The resulting seedlings were transferred to plastic pots (12 cm diameter) containing the soil composite Kureha Engei Baido (0.4 g N, 1.9 g P, 0.6 g K kg 1 ; Kureha Chemical Co., Tokyo). When the leaf blade of the fifth leaf was approximately 2 cm long (defined as the 5-leaf stage), plants were transplanted into larger pots (27 cm diameter) filled with Kureha Engei Baido soil composite. Plants were adequately watered and supplied with fertilizer (0.002% (v/v) Hyponex (Hyponex-Japan, Osaka)) during the experiments. Plants were grown in a greenhouse until anthesis of the flowers on node 25 of the main stems, and the sex of flowers on each node was recorded. Cucumber plants were grown in two different seasons, mid-september to the end of November in 2003 and mid-august to the end of October in 2004, and they were designated as Experiment I and Experiment II, respectively. There were differences in temperature and day-length conditions in the two experiments (Table 1). Flower sex was determined until node 25 or an even higher node at 7 weeks after germination. Day-length became shorter from approximately 14 h to 12 h during the experiments, and it was approximately min longer in Experiment II than Experiment I. The average day temperature and the minimum day temperature in Experiment II were 2 9 C higher than in Experiment I. The differences were greater than 4 C during weeks 2 to 7 in particular, when cucumber plants differentiated flower sex until node 25 or higher on the main stems. Because short-day and low night temperature conditions are favourable for cucumber plants to produce female flowers (Nitsch et al., 1952; Ito and Saito, 1960; reviewed in Frankel and Galun, 1977), plants grown in Experiment I could produce a greater number of female flowers compared with those grown in Experiment II as discussed below. Expression analysis of CS-ACS2 by RT-PCR Total RNA was isolated from shoot apices, stems, leaves, and roots when plants were at the 5-leaf stage. Shoot apices of approximately 1 cm long included immature leaves. Stems and leaves were obtained from the third internodes and the fourth leaves, Table 1. Temperature and day-length conditions during Experiment I and Experiment II Temperature and day-length data recorded in Sendai were obtained from the Japan Meteorological Agency and the National Astronomical Observatory of Japan, respectively. Week after germination Average temperature ( C) a Minimum temperature ( C) b Day-length (h) c Exp I Exp II Exp I Exp II Exp I Exp II a Average value ( C) of temperatures measured every hour for 7 d. b Average value ( C) of minimum temperatures observed between h to h for 7 d. c Duration (h) between dawn and dusk of the first day of the week. respectively. These samples were excised with a razor blade and immediately frozen in liquid nitrogen and stored at 80 C prior to the extraction of nucleic acids. Total RNA was extracted from the cucumber organs with the TRI reagent (Sigma Chemical Co., St Louis). cdnas synthesis was performed with 0.5 lg of total RNA using ReverTra Ace a (Toyobo, Osaka) and random hexamers (Takara Bio Inc., Shiga, Japan) in a 20 ll volume. PCR was performed with TaKaRa Ex Taqä (Takara Bio Inc.) using 1 ll of the cdna solution and specific primer sets for CS-ACS2 and Cs-actin. The primers were as follows: CS-ACS2 forward primer, 5#-TTC AAG CGT TGA ACT TTC GCG-3#, and CS-ACS2 reverse primer, 5#-GAA TGT CTT CGA TTG TGG ACC G-3#, for CS-ACS2; Cs-actin forward primer, 5#-GAC ATT CAA TGT GCC TGC TAT-3#, and Cs-actin reverse primer, 5#-CAT ACC GAT GAG AGA TGG CTG-3#, for Cs-actin (Yamasaki et al., 2001). The PCR conditions were 94 C for 30 s, 55 C for 30 s, and 72 C for 1 min. After 24 cycles for CS-ACS2 and 20 cycles for Cs-actin, the PCR products were analysed on a 1.0% agarose gel by electrophoresis and visualized by ethidium bromide staining. Expression analysis of the CS-ACS2 gene by in situ hybridization The CS-ACS2 fragment was isolated as described previously (Yamasaki et al., 2001), labelled according to the instructions provided with the DIG (digoxigenin) RNA Labelling and Detection Kit (Roche Diagnostics, Basel), and used as a probe for in situ hybridization. Shoot apices, approximately 1 cm long, were harvested from cucumber plants when the leaf blade of the fourth or the sixth leaf was approximately 2 cm long (defined as the 4-leaf stage and the 6-leaf stage, respectively). All flower buds were collected from shoot apices of five plants in Experiment I and three plants in Experiment II using a stereomicroscope. Each excised flower bud was put into a vial and fixed with 0.05 M sodium phosphate buffer (ph 7.2) containing 4% paraformaldehyde and 0.25% glutaraldehyde. Samples were incubated twice with fresh fixative for 5 min each time, and subsequently underwent a secondary fixation for 90 min. After fixation, the samples were dehydrated with an ethanol series, which was then replaced by butanol, and finally embedded in Paraplast Plus (Oxford, Labware, St Louis). For in situ hybridization, a previously described method by Saito et al. (2004) was modified according to Sugiyama et al. (2006) as follows. Sections (9 lm thick) were placed on silicon-coated glass slides (Matsunami Glass Ind., Osaka) and incubated at 47 C overnight. The paraffin was removed from the slides by immersion in xylene. Sections were rehydrated and incubated for 10 min in 0.2 N HCl, 100 mm TRIS-HCl (ph 7.5), and 50 mm EDTA containing 7.5 lg ml 1 proteinase K (Roche Diagnostics) at 25 C, and for 20 min in 100 mm triethanolamine (ph 8.0) and 0.25% acetic anhydrate. After dehydration in an ethanol series, slides were air-dried before application of the hybridization solution. Hybridization solution (100 ll) containing 50% formamide, 53 SSC, 53 Denhardt s solution, 100 lg ml 1 Escherichia coli trna, and 500 lg ml 1 poly(a) was applied to each slide. After 2 h of prehybridization, 50 ll of the hybridization solution was discarded, and 100 ll hybridization solution containing 3 lg ml 1 of digoxigenin (DIG)-labelled RNA probe was added to each slide. These slides were incubated in a humid chamber at 65 C overnight, and unhybridized probe was removed by immersing the slides in 23 SSC for 2 h at 65 C. Slides were immersed in NT buffer (150 mm NaCl, 100 mm TRIS-HCl, ph 7.5), and incubated for 60 min with 100 ll of 1.5% (w/v) blocking reagent (Roche Diagnostics) at room temperature. Subsequently, 100 ll of anti- DIG-alkaline-phosphatase-conjugated antibody (Roche Diagnostics)

4 2900 Saito et al. diluted into 1:1000 in NT buffer was applied to each slide. The slides were washed with NT buffer and TNM buffer (100 mm TRIS-HCl, ph 9.5, 100 mm NaCl, and 50 mm MgCl 2 ). The sections were stained for 2 h in 200 ll of 175 lg ml 1 nitroblue tetrazolium (Roche Diagnostics) and 450 lg ml 1 5-bromo-4- chloro-3-indolyl-phosphate (Roche Diagnostics) in TNM buffer. The staining reaction was stopped by immersion in 10 mm TRIS- HCl (ph 7.5) and 100 mm EDTA. The slides were dehydrated with an ethanol series, which was then replaced by xylene, sealed with MP500 (Matsunami Glass Ind.), and covered with a glass cover slip. The sections were observed under a microscope (BX50, Olympus, Tokyo), and photographs were taken with a digital camera (Camedia E-10, Olympus). All sections obtained from three to five plants were observed, but only those of one plant are represented in each figure. Results Sex expression and its variation in cucumber plants Plants were grown until anthesis of flowers on node 25, and the sex of the flowers on each node of the main stem was recorded. Figure 1 shows the sex expression of RS-G and RS-M cucumber plants in Experiment I. In this experiment, the first flower appeared on nodes 3 4 in both cucumber lines, and the percentage of flowered nodes with female flowers was 100% in RS-G and 89.1% in RS-M cucumbers. RS-M plants produced male flowers on nodes 10 12, although female flowers developed on other nodes. Figure 2 shows the sex expression of RS-G and RS-M cucumber plants grown for Experiment II. The first flower appeared on node 5 in RS-G and on node 3 in RS-M cucumbers. The percentage of flowered nodes with female flowers in RS-G and RS-M cucumber plants was 100% and 21.9%, respectively, under the growth conditions of this experiment. RS-M cucumber plants developed female flowers on nodes 6 7, 12 13, 19 20, and 24 25, and male flowers were produced at the other nodes. Numbers of nodes with female flower on the main stems in RS-M cucumber in Experiment I and Experiment II were 20.7 and 5.6, respectively, and the difference was statistically significant by Student t test (P <0.01). Thus, sex expression of the gynoecious RS-G cucumber was relatively stable, but that of the monoecious RS-M cucumber varied dramatically in the two experiments. Expression of the CS-ACS2 gene in different organs of cucumber plants An ACC synthase gene, CS-ACS2, was isolated from cucumber plants and shown to be expressed in shoots and female flowers (Kamachi et al., 1997, 2000; Yamasaki et al., 2003b). However, no detailed analysis of the expression of the CS-ACS2 gene in other organs of cucumber plants has been conducted. Therefore, RT-PCR was used to analyse CS-ACS2 expression in different organs of the monoecious cucumber RS-M and the gynoecious cucumber RS-G (Fig. 3). CS-ACS2 expression Fig. 1. Sex of the flowers of gynoecious (RS-G) and monoecious (RS- M) cucumber plants grown in 2003 (Experiment I). The diagrammatic data show the sex of the flowers on each node of the main stems of seven individuals. Closed and open circles represent female and male flowers, respectively. Lower nodes without either circle represent the vegetative nodes. Both RS-G and RS-M plants produced their first flowers on nodes 3 or 4 on the main stems. Flowers produced by RS-G plants were all female, but RS-M plants produced male flowers at approximately nodes Fig. 2. Sex of the flowers of gynoecious (RS-G) and monoecious (RS-M) cucumber plants grown in 2004 (Experiment II). The diagrammatic data show the sex of the flowers on each node of the main stems of five individuals. Closed and open circles represent female and male flowers, respectively. Lower nodes without either circle represent the vegetative nodes. The first flower appeared on node 5 in RS-G and on node 3 in RS-M plants. All flowers produced by RS-G plants were female, but RS-M plants produced female flowers on node 6 7, 12 13, 19 20, and 24 25, which were separated by 4 5 nodes with male flowers. was specific to the shoot apex in both monoecious and gynoecious cucumber plants. Moreover, the spatial expression analysis used here revealed that accumulation of

5 CS-ACS2 mrna was not detected in leaves, stems, or roots. CS-ACS2 mrna accumulated more abundantly in the shoot apices of gynoecious cucumber than that of the monoecious line, as reported previously. Accumulation of CS-ACS2 mrna in roots was detected when the number of PCR cycles was increased to 30, but there was no difference in the signal intensity between monoecious and gynoecious cucumber plants (data not shown). To identify the specific site of CS-ACS2 expression, the shoot apices were analysed by in situ hybridization. In both RS-G and RS-M cucumber plants, CS-ACS2 mrna accumulated beneath the pistil primordia or in the region of the flower bud locules, but not in other organs such as the immature leaves and stems (Fig. 4A, B). When a CS-ACS2 sense probe was used, no signals were detected in any organs including the flower buds (data not shown). Expression of the CS-ACS2 gene in flower buds at different developmental stages in gynoecious and monoecious cucumber plants Since CS-ACS2 was expressed predominantly in the flower buds of cucumber shoots, its mrna accumulation was examined in each flower bud in the shoot apices of cucumber plants. Because the proportion of male and female flowers in monoecious cucumbers varies in response to environmental cues, shoot apices for in situ hybridization were obtained from cucumber plants grown in different seasons for Experiment I and Experiment II. In Experiment I, flower buds on nodes 3 14 were obtained from cucumber plants at the 4-leaf stage. In cucumber plants at this stage, flower buds had already initiated and developed by node 17 18, and the flower sex had been determined by node In the gynoecious RS-G cucumber, flower buds on nodes 13 and 14 were at the stage of primordia formation of the sepals and stamens (Fig. 5A, B). Flower buds on nodes 11 and 12 were at a bisexual stage (Fig. 5C, D), but the sex of flower buds on the lower nodes had been determined, and these flower buds possessed pistils that had begun to form a cavity for ovary development (Fig. 5E L). Accumulation of CS-ACS2 mrna was seen in flower buds obtained from node 5 through 12 (Fig. 5C J). No accumulation of CS-ACS2 was observed in flower buds on nodes 13 and 14, in which pistil primordia had not yet been formed (Fig. 5A, B). Flower buds on nodes 11 and 12, which were at the bisexual stage, as judged by the presence of primordia of both pistils and stamens, showed CS-ACS2 mrna accumulation only beneath the pistil primordial (Fig. 5C, D). In flower buds, on nodes 5 10, ovary development started after sex determination, and CS-ACS2 mrna accumulated in the locule of the ovary, the cavity where the ovule will develop (Fig. 5E J). Female flower buds on nodes 3 and 4 developed to approximately 4 mm long and did not show CS-ACS2 mrna accumulation (Fig. 5K, L). Sex differentiation of cucumber flowers 2901 Fig. 3. Accumulation of CS-ACS2 mrna in different organs of gynoecious (RS-G) and monoecious (RS-M) cucumber plants. Total RNA was extracted from roots, stems, leaves and shoot apices of plants at the 5-leaf stage and analysed by RT-PCR. Fig. 4. Accumulation of CS-ACS2 mrna in the shoot apices of gynoecious (A) and monoecious (B) cucumber plants. Shoot apices were harvested from 5-leaf stage plants and sectioned longitudinally for in situ hybridization. Am, apical meristem of the main shoot; Le, immature leaf; Sm, stem; Se, sepal; Pe, petal; St, stamen or its primodium; Pi, pistil or its primodium; Ov, ovule. Scale bar represents 1 mm. CS-ACS2 signal (arrowhead) is observed specifically in the pistillate tissue of flower buds. Flower buds from nodes 3 through 13 were examined in monoecious RS-M cucumber plants (Fig. 6A K). As in the RS-G gynoecious cucumber, flower buds on nodes were at the bisexual stage (Fig. 6B D). CS-ACS2 mrna accumulation was seen in flower buds on nodes 4 9 and node 12 (Fig. 6B, E J). The accumulation site of CS-ACS2 mrna in flower buds of monoecious cucumber was similar to that of gynoecious cucumber. That is, flower buds on node 12 belonged to the bisexual stage, and CS-ACS2 mrna accumulated just beneath the pistil primordial (Fig. 6B). Flower buds on nodes 4 9 had begun to form a cavity for locule development in the ovary, and CS-ACS2 mrna was detected surrounding the cavity (Fig. 6E J). Flower buds on nodes 3, 10, 11, and 13 did not show CS-ACS2 expression (Fig. 6A, C, D, K). In Experiment II, flower buds were obtained from nodes 5 through 18 of RS-G and from nodes 5 through 17 of RS-M cucumber plants at the 6-leaf stage (Figs 7A N, 8A M). In cucumber plants at this stage, flower buds had

6 2902 Saito et al. Fig. 5. In situ hybridization analysis of CS-ACS2 mrna accumulation in flower buds from successive nodes on the main stem of gynoecious (RS-G) cucumber plants (Experiment I). Flower buds at different stages of development were harvested from nodes 14 (A) to 3 (L) on the main stem of RS-G plants grown in Numbers in the parentheses indicate the node position on the main stem. Se, sepal; Pe, petal; St, stamen or its primodium; Pi, pistil or its primodium; Ov, ovary. Scale bars for nodes 14 (A) to 11 (C) and for nodes 10 (D) to 3 (K) represent 200 lm and 500 lm, respectively. Fig. 6. In situ hybridization analysis of CS-ACS2 mrna accumulation in flower buds from successive nodes on the main stem of monoecious (RS-M) cucumber plants (Experiment I). Flower buds at different stages of development were harvested from nodes 13 (A) to 3 (K) on the main stem of RS-M plants grown in Numbers in the parentheses indicate the node position on the main stem. Se, sepal; Pe, petal; St, stamen or its primodium; Pi, pistil or its primodium; Ov, ovary. Scale bars for nodes 13 (A) to 11 (C) and for nodes 10 (D) to 3 (K) represent 200 lm and 500 lm, respectively. developed up to node 21, and the flower sex had been determined by node In RS-G cucumber plants, flower buds on nodes 17 and 18 were forming stamen primordia (Fig. 7A, B). Sex determination had occurred in flower buds on nodes 5 through 13, which had begun to form a cavity beneath the pistil or to develop ovules (Fig. 7F-N). CS-ACS2 expression was strongly observed in flower buds on nodes 5 16 in RS-G cucumber plants (Fig. 7C N). Flower buds on nodes were at the bisexual stage and showed CS-ACS2 mrna accumulation

7 (Fig. 7C E). However, flower buds on nodes 17 and 18, which had not developed to the bisexual stage, showed no accumulation of CS-ACS2 mrna (Fig. 7A, B). In RS-M cucumber plants, flower buds on nodes 14 through 17 were still at the bisexual stage or before the stage of Sex differentiation of cucumber flowers 2903 formation of pistil primordia (Fig. 8A D), whereas flower buds on nodes 5 through 13 had already undergone sex differentiation (Fig. 8E M). The flower buds on nodes 7 and 13 appeared to be developing as female flowers, and CS-ACS2 mrna accumulation occurred only in the Fig. 7. In situ hybridization analysis of CS-ACS2 mrna accumulation in flower buds from successive nodes on the main stem of gynoecious (RS-G) cucumber plants (Experiment II). Flower buds at different stages of development were harvested from nodes 18 (A) to 5 (N) on the main stem of RS-G plants grown in Numbers in the parentheses indicate the node position on the main stem. Se, sepal; Pe, petal; St, stamen or its primodium; Pi, pistil or its primodium; Ov, ovary. Scale bars for nodes 18 (A) to 15 (D) and for nodes 14 (F) to 5 (N) represent 200 lm and 500 lm, respectively. Fig. 8. In situ hybridization analysis of CS-ACS2 mrna accumulation in flower buds from successive nodes on the main stem of monoecious (RS-M) cucumber plants (Experiment II). Flower buds at different stages of development were harvested from nodes 17 (A) to 5 (M) on the main stem of RS-M plants grown in Numbers in the parentheses indicate the node position on the main stem. Se, sepal; Pe, petal; St, stamen or its primodium; Pi, pistil or its primodium; Ov, ovary. Scale bars for nodes 17 (A) to 14 (D) and for nodes 13 (E) to 5 (M) represent 200 lm and 500 lm, respectively.

8 2904 Saito et al. flower buds on nodes 7 and 13 in the RS-M cucumber plants (Fig. 8E, K). Thus, the nodes that produced female flowers well coincided with those expressed CS-ACS2 gene in flower buds in both Experiment I and Experiment II. Location of the CS-ACS2 expression in developing ovary To clarify further the tissue that accumulates CS-ACS2 mrna, flower buds were obtained from node 9 of RS-G cucumber plants grown for Experiment I and crosssections were prepared for in situ hybridization. CS-ACS2 expression occurred specifically surrounding the tissues destined to develop as ovules and placenta in the female flowers of cucumber plants (Fig. 9A C). Discussion Ethylene has been shown to regulate the sex differentiation of cucumber flowers (reviewed in Yin and Quinn, 1995; Perl-Treves, 1999; Yamasaki et al., 2005). The regulatory mechanism of sex differentiation by ethylene still remains to be clarified, but the following model was proposed recently (Yamasaki et al., 2001). When the genotype is F-, all flower primordia produce enough ethylene to induce female flowers. Alternatively, when the genotype is ff, some flower primordia do not produce enough ethylene to induce the development of female flowers. In addition, the product of the M locus mediates the inhibition of stamen development by ethylene. When the genotype is mm, the ethylene signal is not transmitted, and stamen development is not inhibited. The sex expression of the gynoecious genotype (M-F-) is relatively more stable than that of the monoecious genotype (M-ff). Indeed, compared with gynoecious cucumbers, sex expression in monoecious cucumbers varies more readily in response to growth regulators such as gibberellins and ethylene, as well as environmental cues such as temperature and day-length (reviewed in Frankel and Galun, 1977; Perl-Treves, 1999; Yamasaki et al., 2005). According to this model, all flower buds of gynoecious cucumbers produce enough ethylene due to the presence of the F gene and develop as female flowers. By contrast, two types of flower buds are present in shoot apices of monoecious cucumbers (M-ff), and the ones destined to develop as female flowers produce more ethylene, while the ones destined to develop as male flowers produce less ethylene. However, no evidence for this has been reported because it is difficult to separate ethylene biosynthesis by the flower buds at the early developmental stages from that produced by vegetative tissues or other flower buds in the shoot apices. The ability of flower buds to produce ethylene was evaluated by monitoring CS-ACS2 expression in order to verify the model described above and to understand how monoecious cucumber plants regulate and alter the sex of their flowers under various circumstances. CS-ACS2 was cloned as a gene that encodes a 1-aminocyclopropane-1- carboxylic acid (ACC) synthase specifically associated with femaleness in cucumber flowers (Kamachi et al., 1997, 2000). Its specific expression in flower buds is confirmed here and its expression correlated with the development of female flowers (Figs 3, 4). Therefore, the accumulation of CS-ACS2 mrna in cucumber flower buds was examined at different developmental stages to evaluate their ability to produce ethylene and to determine the positional relationship between CS-ACS2 gene expression and the sex of the flower in the main shoots. In this study, RS-G gynoecious cucumber plants produced only female flowers in two experiments, but the proportion of female flowers in RS-M monoecious cucumber plants varied in the two experiments done under different growth conditions. In Experiment I, RS-M cucumber plants produced female flowers, with the exception of nodes 10 12, where male flowers appeared. In contrast, in Experiment II, RS-M mainly produced male flowers, with the exception of nodes 6 7, 12 13, 19 20, and 24 25, where female flowers developed. This difference was probably a result of the longer day-length and higher temperatures in Experiment II, compared with Experiment I, because the conditions of short day-length Fig. 9. Accumulation of CS-ACS2 mrna in the developing ovary of cucumber plants. Female flower buds were harvested from node 9 of gynoecious (RS-G) plants at the 5-leaf stage for in situ hybridization (A, B). (A, B) The location of CS-ACS2 mrna accumulation (arrowhead) in the cross-section of the developing ovary. (C) The cross-section of the ovary at anthesis, in which the arrowhead indicates the region of the placenta. Scale bars in (A), (B), and (C) represent 500 lm, 100 lm, and 1 mm, respectively.

9 and low temperature are favourable for femaleness in cucumber plants (reviewed in Frankel and Galun, 1977; Perl-Treves, 1999; Yamasaki et al., 2005). Indeed, the daylength in Experiment II was approximately 1 h longer than that in Experiment I (Table 1). Also, the temperature in Experiment II was much higher than that in Experiment I (Table 1). During weeks 2 to 7, during which flower sex on nodes observed in this study was determined, minimum temperatures observed at night were approximately 4 9 C lower in Experiment I than those in Experiment II. It is known that night temperatures influence sex expression in cucumber more effectively than day-length (Ito and Saito, 1960; reviewed in Frankel and Galun, 1977). Night temperatures of 17 C to 9 C are the most suitable conditions for the differentiation of female flowers. For example, a monoecious cultivar of cucumber, Sagami- Hanziro, grown under a 16-h photoperiodic condition produced female flowers on two nodes out of 25 on the main stem at a night temperature 25 C, but produced female flower on nine nodes at a night temperature 15 C (Ito and Saito, 1960; reviewed in Frankel and Galun, 1977). The numbers of female flower nodes in Sagami- Hanziro cucumber grown at 17 C under 8-h photoperiodic conditions and at 24 C under 16-h photoperiodic conditions were 27.2 and 4.6, respectively, out of the first 30 nodes on the main stems (Ito and Saito, 1960; reviewed in Frankel and Galun, 1977). In this study, the minimum temperatures recorded during the period of Experiment II was higher than 17 C for the first six weeks after germination, while the minimum temperature gradually decreased from 17 C to 9 C in Experiment I. The variation of sex expression of monoecious cucumber in our two experiments was similar to that due to night temperature in other monoecious cucumber. It is therefore concluded that temperature conditions mainly affected sex expression in the two experiments. Shoot apices harvested from plants at the 4-leaf stage (Experiment I) or the 6-leaf stage (Experiment II) included flower buds at different developmental stages. The accumulation of CS-ACS2 mrna corresponded well to flower buds that ultimately developed into female flowers. That is, the nodes with flower buds that expressed the CS-ACS2 gene (Figs 5 8) coincidently produced female flowers (Figs 1, 2). In gynoecious cucumbers, which produced only female flowers, flower buds accumulated CS-ACS2 mrna both at the bisexual stage and at later developmental stages. In monoecious cucumbers, by contrast, flower buds on nodes in Experiment I, which developed into male flowers, did not show any CS-ACS2 mrna accumulation. Other nodes, where flower buds developed into female flowers, showed CS- ACS2 mrna accumulation. In Experiment II, CS-ACS2 mrna accumulation was observed in nodes 7 and 13 where the flower buds developed into female flowers. The other nodes developed male flowers, and did not show Sex differentiation of cucumber flowers 2905 any CS-ACS2 mrna accumulation at the younger stages. Flower buds on nodes and of monoecious cucumbers in Experiment II developed as female, but they were either before the bisexual stage or as yet undifferentiated at the time of sampling for in situ hybridization. The results of this study suggest that expression of the CS- ACS2 gene is positively correlated with the differentiation of female flowers in cucumber plants and that fluctuations in ethylene biosynthesis in flower buds correspond to fluctuations in the proportion of female and male flowers in monoecious cucumbers. Thus, changes in the proportion of male and female flowers in response to environmental cues could be caused by a change in CS- ACS2-mediated ethylene biosynthesis in individual flower buds at the bisexual stage. The mechanism by which the expression of the CS-ACS2 gene is modified in response to temperature and day-length, is not known, but indeed ethylene evolution and CS-ACS2 expression are higher under the short-day conditions than the long-day conditions in the shoot apices of monoecious cucumber (Yamasaki et al., 2003a). When Trebitsh et al. (1997) identified the CS-ACS1 gene encoding ACC synthase, they found that monoecious cucumber plants possessed a single copy of this gene, whereas gynoecious plants possessed an additional copy, CS-ACS1G, that is strongly linked to the F locus. CS-ACS1G expression occurs prior to the CS-ACS2 expression that is ethylene-inducible and more abundant in gynoecious plants than monoecious ones (Kamachi et al., 2000; Yamasaki et al., 2001). In gynoecious cucumber plants, therefore, ethylene biosynthesis mediated by CS-ACS1G could trigger further biosynthesis of ethylene by CS-ACS2. The in situ hybridization analysis in this study demonstrated that the CS-ACS2 gene was up-regulated in flower buds destined to develop into female flowers. Accumulation of CS-ACS2 mrna was seen in all flower buds at the bisexual stage and thereafter in gynoecious cucumber flower buds. No CS-ACS2 signals were detected in flower buds earlier than the bisexual stage. This result suggests that ethylene biosynthesis due to CS-ACS2 gene expression in flower buds is a factor responsible for determining the sex of the cucumber flowers because the sex is determined by the arrest of either pistil or stamen development after the bisexual stage. CS-ACS2 mrna accumulated in the pistillate tissue of flower buds at the bisexual stage, but not in the stamen or its primordia. This suggests that an ACC synthase other than CS-ACS2 regulates ethylene biosynthesis in the stamen primordia or, alternatively, that the ethylene produced via CS-ACS2 in the pistillate tissue participates in the arrest of stamen development. Because pistil and stamen primordia are both enclosed in the flower buds and immature leaves, it is hypothesized that CS-ACS2-produced ethylene accumulates enough to influence stamen development. Accordingly, ethylene produced by CS-ACS2 could act both

10 2906 Saito et al. to develop the pistillate tissue and suppress staminate tissue. CS-ACS2 mrna continues to accumulate in the region of ovule formation during the early stages of locule/ovary development. Ethylene may play a role not only in maintaining femaleness but also in ovule or ovary development. Previously, the pistil and ovary were shown to produce ethylene, which was correlated with senescence of petals or leaves after pollination (Ten Have and Woltering, 1997; Clark et al., 1997). In tobacco, the ACC oxidase gene (ACO) is expressed in flower buds at the time of ovule development, and suppression of ACO gene expression resulted in incomplete development of ovules (De Martinis and Mariani, 1999). It is therefore probable that ethylene plays an important role in ovule/ovary development of cucumber flowers. This study demonstrates that shoot apices of monoecious cucumber differentiate two types of flower buds that can be distinguished on the basis of their levels of CS- ACS2 gene expression. It is concluded that fluctuation of ethylene biosynthesis in individual flower buds at the bisexual stage accounts for the fluctuation of the proportion of female flowers in monoecious cucumbers. Acknowledgements We thank Dr Ryuji Sugiyama of Toyama University for his kind advice on in situ hybridization. We also thank Dr Yuko Saito of our laboratory for her helpful advice and discussion. This work was supported by a Grant-in-Aid for Scientific Research (B) ( ) from the Japan Society for the Promotion of Science to HT. This study was also carried out as a part of the Ground-Based Research Announcement for Space Utilization promoted by the Japan Space Forum. References Adams DO, Yang SF Ethylene biosynthesis: identification of 1-aminocyclopropane-1-carboxylic acid as an intermediate in the conversion of methionine to ethylene. Proceedings of the National Academy of Sciences, USA 76, Atsmon D, Galun E A morphogenetic study of staminate, pistillate and hermaphrodite flowers in Cucumis sativus L. Phytomorphology 10, Atsmon D, Tabbak C Comparative effects of gibberellin, silver nitrate and aminoethoxyvinyl glycine on sexual tendency and ethylene evolution in the cucumber plant (Cucumis sativus L.). Plant and Cell Physiology 20, Beyer E Silver ion: a potent antiethylene agent in cucumber and tomato. HortScience 11, Clark DG, Richards C, Hilioti Z, Lind-Iversen S, Brown K Effect of pollination on accumulation of ACC synthase and ACC oxidase transcripts, ethylene production and flower petal abscission in geranium (Pelargonium3hortorum L.H. Bailey). Plant Molecular Biolology 34, De Martinis D, Mariani C Silencing gene expression of the ethylene-forming enzyme results in a reversible inhibition of ovule development in transgenic tobacco plants. The Plant Cell 11, Frankel R, Galun E Pollination mechanisms, reproduction, and plant breeding. Heidelberg: Springer-Verlag. Fujieda K A genecological study on the differentiation of sex expression in cucumber plants. Bulletin of Horticultural Research Station, Japan, Series D 4, Fujita Y, Fujieda K Relation between sex expression types and cotyledon etiolation of cucumber in vitro. I. On the role of ethylene evolved from seedlings. Plant and Cell Physiology 22, Galun E Study of the inheritance of sex expression in the cucumber: the interaction of major genes with modifying genetic and non-genetic factors. Genetica 32, Ito H, Saito T Factors responsible for the sex expression of the cucumber plant. XII. Physiological factors associated with the sex expression of flowers. Tohoku Journal of Agricultural Research 11, Iwahori S, Lyons JM, William LS Induced femaleness in cucumber by 2- chloroethanephosphonic acid. Nature 222, Kamachi S, Mizusawa H, Matsuura S, Sakai S Expression of two 1-aminocyclopropane-1-carboxylate synthase genes, CS-ACS1 and CS-ACS2, correlated with sex phenotypes in cucumber plants (Cucumis sativus L). Plant Biotechnology 17, Kamachi S, Sekimoto H, Kondo N, Sakai S Cloning of a cdna for a 1-aminocyclopropane-1-carboxylate synthase that is expressed during development of female flowers at the apices of Cucumis sativus L. Plant and Cell Physiology 38, Knopf RR, Trebitsh T The female-specific Cs-ACS1G gene of cucumber. A case of gene duplication and recombination between the non-sex specific 1-aminocyclopropane-1-carboxylate synthase gene and a branched-chain amino acid transaminase gene. Plant and Cell Physiology 47, Kubicki B Investigation of sex determination in cucumber (Cucumis sativus L.). Genetica Polonica 10, Malepszy S, Niemirowicz-Szczytt K Sex determination in cucumber (Cucumis sativus) as a model system for molecular biology. Plant Science 80, MacMurray AL, Miller CM Cucumber sex expression modified by 2-chloroethanephosphonic acid. Science 162, Mibus H, Tatlioglu T Molecular characterization and isolation of the F/f gene for femaleness in cucumber (Cucumis sativus L.). Theoretical and Applied Genetics 109, Nitsch J, Kurtz EB, Livermann JL, Went FW The development of sex expression in curcurbit flowers. American Journal of Botany 39, Perl-Treves R Male to female conversion along the cucumber shoot: approaches to studying sex genes and floral development in Cucumis sativus. In: Ainsworth CC, ed. Sex determination in plants. Oxford: BIOS Scientific Publishers Ltd, Rudich J, Baker LR, Scott JW, Sell HM Phenotypic stability and ethylene evolution in androecious cucumber. Journal of the American Society for Horticultural Science 101, Rudich J, Halevy AH, Kedar N. 1972a. Ethylene evolution from cucumber plants as related to sex expression. Plant Physiology 49, Rudich J, Halevy AH, Kedar N. 1972b. The level of phytohormones in monoecious and gynoecious cucumbers as affected by photoperiod and ethephon. Plant Physiology 50, Saito Y, Yamasaki S, Fujii N, Hagen G, Guilfoyle T, Takahashi H Isolation of cucumber CsARF cdnas and expression of the corresponding mrnas during gravity-regulated morphogenesis of cucumber seedlings. Journal of Experimental Botany 55,

11 Sex differentiation of cucumber flowers 2907 Shannon S, De La Guardia MD Sex expression and the production of ethylene induced by auxin in cucumber (C. sativus L.). Nature 223, 186. Shifriss O Sex control in cucumbers. Journal of Heredity 52, Sugiyama R, Oda H, Kurosaki F Expression of ASK1-like genes in arrested stamens of female Silene latifolia plants. Journal of Plant Research 119, Takahashi H, Jaffe MJ Further studies of auxin and ACC induced feminization in the cucumber plant using ethylene inhibitors. Phyton 44, Takahashi H, Suge H Sex expression in cucumber plants as affected by mechanical stress. Plant and Cell Physiology 21, Ten Have A, Woltering EJ Ethylene biosynthetic genes are differentially expressed during carnation (Dianthus caryophyllus L.) flower senescence. Plant Molecular Biology 34, Trebitsh T, Rudich J, Riov J Auxin, biosynthesis of ethylene and sex expression in cucumber (Cucumis sativus L.). Plant Growth Regulation 5, Trebitsh T, Staub JE, O Neill SD Identification of a 1-aminocyclopropane-1-carboxylic acid synthase gene linked to the female (F) locus that enhances female sex expression in cucumber. Plant Physiology 113, Yamasaki S, Fujii N, Matsuura S, Mizusawa H, Takahashi H The M locus and ethylene-controlled sex determination in andromonoecious cucumber plants. Plant and Cell Physiology 42, Yamasaki S, Fujii N, Takahashi H. 2003a. Photoperiodic regulation of CS-ACS2, CS-ACS4, and CS-ERS gene expression contributes to the femaleness of cucumber flowers through diurnal ethylene production under short-day conditions. Plant, Cell and Environment 26, Yamasaki S, Fujii N, Takahashi H. 2003b. Characterization of ethylene effects on sex determination in cucumber plants. Sexual Plant Reproduction 16, Yamasaki S, Fujii N, Takahashi H Hormonal regulation of sex expression in plants. Vitamins and Hormones 72, Yin T, Quinn JA Tests of a mechanistic model of one hormone regulating both sexes in Cucumis sativus (Cucurbitaceae). American Journal of Botany 82,