Index INTRODUCTION MATERIALS AND METHODS Plants Differential display and isolation of SlMF1 full-length cDNA Expression analyses Genomic distribution analyses RESULTS SlMF1 is specifically expressed in stamens and petals SlMF1 is a multicopy gene and is located on sex chromosomes DISCUSSION References

INTRODUCTION

Although sex chromosomes are found in most animals, they are limited to dioecious plants in which staminate and pistillate flowers are produced in separate individuals. Plant sex chromosomes have arisen independently in taxonomic groups from liverworts to angiosperms (Okada et al. 2001; Ishizaki et al. 2002; Charlesworth 2002). Heteromorphic sex chromosomes have been reported in several dioecious species of angiosperms (Vyskot and Hobza 2004). The Y chromosome of Silene latifolia in the family Caryophyllaceae is the most well-studied sex chromosome in plants (Matsunaga and Kawano 2001; Negrutiu et al. 2001). The Y chromosome is largely euchromatic with the exception of centromeric and subtelomeric regions (Buzek et al. 1997; Matsunaga et al. 1999a; Kazama et al. 2003). As compared with the mammalian Y chromosome that emerged approximately 320 million years ago, the Y chromosome in the genus Silene emerged quite recently, that is, during the last 10 million years. Analyses of the Y chromosome deletions induced by γ-rays or X-rays suggest that the Y chromosome has three functions: suppression of pistil development and initiation and completion of stamen development (Negrutiu et al. 2001). Recently, many Y-linked genetic markers were isolated from the male genome of S. latifolia (Donnison et al. 1996; Zhang et al. 1998; Nakao et al. 2001; Obara et al. 2002). In particular, analyses of male-specific AFLP markers confirmed the functions assigned to the regions of the Y chromosome (Lebel-Hardenack et al. 2002). Flow sorted X chromosomes derived from female cultured cells are useful for rapid confirmation of the linkage of the X chromosome (Kejnovsky et al. 2001). PCR analyses with flow sorted X chromosomes revealed that a male reproductive organ-specific gene MROS3 had a paralogue on the X chromosome (Matsunaga et al. 1999b; Kejnovsky et al. 2001).

After floral identity genes determine the developmental fate of each whorl, the first, second, third, and fourth whorl develop sepals, petals, stamens, and carpels, respectively (Coen and Meyerowitz 1991). In dioecious plants, after floral identity is determined, unisexual flowers develop through the suppression or promotion of cell division activity in each sex primordium (Matsunaga et al. 2004a). In male flowers, the fourth-whorl gynoecium primordium is suppressed and later becomes a rudimentary gynoecium in the form of a filamentous rod that lacks an ovary and pistils (Grant et al. 1994). In the third whorl of female flowers, the growth of stamen primordia is arrested and the tissues degenerate before the flower opens (Uchida et al. 2003, 2005). Flower bud-specific genes were expected to contribute to the regulation of sex-specific development of unisexual floral organs. Therefore, several researchers attempted to characterize reproductive organ-specific genes (Barbacar et al. 1997; Lebel-Hardenack et al. 1997; Matsunaga et al. 1996, 1997; Scutt et al. 2002; Sugiyama et al. 2003) or orthologs of Arabidopsis thaliana genes that regulate flower development (Hardenack et al. 1994; Ageez et al. 2003; Matsunaga et al. 2003, 2004b). We have isolated a reproductive organ-specific gene SlMF1 from male flower buds, and analyzed its expression pattern in flower buds and its sex chromosome linkage. Interetingly, SlMF1 was an X chromosome-linked gene that was specifically expressed in developing anthers and petals.

MATERIALS AND METHODS

Plants

We used 2 inbred S. latifolia lines–the Japanese laboratory strain K1 and the USA laboratory strain MR4X64 for the molecular experiments. The plants were grown in a temperature-controlled chamber or a green house at 22°C.

Differential display and isolation of SlMF1 full-length cDNA

Differential display was performed on mRNA isolated from young male and female flower buds. Total RNA and mRNA isolation were performed in accordance with the method of Moore et al. (2003). The oligonucleotides in an RNA image kit 2 (Genehunter, Nashville, TN, USA) were used for differential display. Amplified fragments were subcloned using a TOPO TA Cloning kit (Invitrogen). The nucleotide sequences were determined using an ABI 3100 genetic analyzer and a BigDye terminator cycle sequencing kit (Applied Biosystems). We obtained full-length cDNAs from the screening of our constructed Tripl Ex2 cDNA library (Clontech) of male flower buds (Matsunaga et al. 2003). The full length cDNA sequence was confirmed by RACE-PCR using a SMART RACE cDNA amplification kit (Clontech).

Expression analyses

Northern hybridization was performed using an AlkPhos Direct Labelling and Detection System with CDP-Star (Amersham) with 15 μg of total RNA from male and female flower buds and leaves. Quantitative RT-PCR was performed in accordance with the method of Matsunaga et al. (2003) using a Smart Cycler (Takara) with a QuantiTect SYBR green kit (Qiagen). The gene for the GTPase beta subunit (SlGb), which is expressed constitutively in all organs, was used as an internal standard to estimate the relative expression of mRNA. The relative expression of mRNA for a given tested gene was defined as its mean mRNA expression value divided by that for SlGb, with the same cDNA used as template. Relative expression values and corresponding standard deviations for the transcripts were calculated from three experimental replicates. The oligonucleotide primer sets used for quantitative RT-PCR were as follows: 5'-GACATGGTGACAGCCATAGCAACA-3' and 5'-TCACGAGAAGCAGAGACTATCTGT-3' for SlGb; SlMF1-F: 5'-GTGTCACCCAATTCTACTCA-3' and SlMF1-R: 5'-ACGACGCCACTGGAAGATAG-3' for SlMF1; MROS4-F: 5'-TAGTTGTGCAAATGGCTCCC-3' and MROS4-R: 5'-TCCGAAACACAATGGCCTTC-3' for MROS4. Preparation of biotin-labeled probes for SlSF1 was performed in accordance with the method of Matsunaga et al. (2004a). In situ hybridization was performed as described previously (Matsunaga et al. 2003), with an automatic ISH robot AIH-101B (Aloka) using tyramide amplification in the GenPoint system (Dako).

Genomic distribution analyses

Genomic DNA was isolated from young leaves using an automatic DNA isolation system PI50 (Kurabo). Twenty five micrograms of male and female genomic DNA were digested with HindIII, resolved in 1% agarose gel, blotted onto a membrane, and hybridized with the full-length cDNA of SlMF1 as a probe. Southern hybridization was performed using an AlkPhos Direct Labelling and Detection System with CDP-Star (Amersham).

Flow-sorted chromosomes were collected as described in Kejnovsky et al. (2001). The set of oligonucleotide primers for the 390-bp SlMF1 fragment consisted of SlMF1-F and SlMF1-R.

RESULTS

SlMF1 is specifically expressed in stamens and petals

We performed differential display with 48 combinations of 7 primers using very early young male and female flower buds and identified 38 male flower bud-specific fragments. Northern hybridization using total RNA from male and female flower buds revealed that two RNA fragments strongly hybridized to the total RNA of male flower buds. One fragment was termed SlMF1 (a Silene latifolia male flower-specific gene) and further analyzed. We isolated the full-length cDNA of SlMF1 from the cDNA library of male flower buds. The sequence of the full-length cDNA was confirmed using the 5' RACE method. The length of the cDNA excluding the polyA tail was 744 bp (DDBJ accession no. AB238692), and its longest ORF encoded a 50-amino acid protein that did not show significant homology with other reported proteins. Northern hybridization using total RNA from male and female flower buds and leaves showed that the SlMF1 transcript preferentially accumulated in male flower buds (Fig. 1a). The transcript was also detected to a small extent in female flower buds; however, accumulation of the transcript in male and female leaves was less than below the detection limit. Analysis of its expression in floral organs (male and female sepals and petals, male stamens, female styles and ovaries) was examined using quantitative RT-PCR (Fig. 1b). This analysis indicated that the SlMF1 transcript was preferentially expressed in male and female petals and male stamens, with strong expression in male stamens. In contrast, its expression in male and female sepals and female ovaries was very low. Quantitative RT-PCR with specific primers for a male reproductive organ-specific gene MROS4 (DDBJ accession no. AB238691) using the same RNA samples showed that the transcript preferentially accumulated in male stamens (Fig. 1c). This result is consistent with the previously reported profile of MROS4 expression using Northern hybridization (Matsunaga et al. 1996). This suggests that the results of our quantitative RT-PCR precisely reflect the accumulation level of transcripts in floral organs.

View Details

Fig. 1. Expression analyses of SlMF1. (a) Northern hybridization analysis of SlMF1. The upper panel shows a Northern hybridization image using an AlkPhos Detection System with CDP-Star (Amersham). The lower panel shows an electrophoresis image of rRNA bands stained with ethidium bromide. Lanes 1, 2, 3, and 4 contain 15 μg of total RNA from male flower buds, female flower buds, male leaves, and female leaves. (b and c) Quantitative RT-PCR analyses of the SlMF1 and MROS4 transcripts. The upper and lower panels show values representing the relative expression of the SlMF1 and MROS4 transcripts that were standardized based on the expression of the GTPase β subunit, which is expressed constitutively in all organs. Bars 1 to 7 represent the relative expression levels in male sepals, male petals, male stamens, female sepals, female petals, female styles, and female ovaries, respectively. Data are expressed as mean ± SEM (n = 3).

To analyze the expression profile in greater detail, we performed in situ hybridization using longitudinal sections of young and mature male flower buds at stages 3 and 9 described by Grant et al. (1994) (Fig. 2). The biotin-labeled probe was detected using tyramide amplification and the signal appeared brown. When the section of the young male flower bud at stage 3 was allowed to hybridize with the antisense probe, significant signals could be detected in the whole floral meristem (Fig. 2a). When the young female flower bud at the same stage was used for in situ hybridization, the same signal distribution was found in the whole floral meristem (data not shown). When a section of the mature male flower bud at stage 9 was allowed to hybridize with the antisense probe, the signals could be detected in anthers and petals (Fig. 2c). In particular, strong signals could be detected in pollen mother cells and tapetal cells. When the mature female flower bud at the same stage was used for in situ hybridization, the signals were detected in petals, but were not detected in vestigial stamens (data not shown). In contrast, the sense probe produced no hybridization signal above background in the sections of both young and mature male flower buds (Fig. 2b and d).

View Details

Fig. 2. In situ hybridization of the SlMF1 gene. Longitudinal sections of young male flower buds are allowed to hybridize with the biotin-labeled probe. Hybridization signals appear brown. Sections were allowed to hybridize with the antisense probe of SlMF1 (a and c) and the sense probe of SlMF1 (b and d). (a and b) Male flower buds at stage 3; (b and d) male flower buds at stage 9. Po, pollen mother cell; Pe, petal; Ta, tapetum. Bars in (b) and (d) represent 50 μm or 150 μm, respectively.

SlMF1 is a multicopy gene and is located on sex chromosomes

To confirm the genomic distribution of SlMF1, we performed Southern hybridization using male and female genomic DNA from a pair of male and female parents and three progeny of the USA laboratory strain (Fig. 3a), and from a pair of male and female parents and four progeny of the Japanese laboratory strain (Fig. 3b). Six fragments of 20.0, 5.8, 4.9, 4.7, 3.6, and 2.9 kb hybridized strongly and three fragments of 15.0, 8.0, and 7.3 kb hybridized weakly in the male genome. This result shows that SlMF1 is a multicopy gene. The 5.8-kb fragment was detected only in the male genomic DNAs of both strains. This indicates that the 5.8-kb fragment is linked to the Y chromosome. Regardless of whether the plants were male or female, the 4.9- and 4.7-kb fragments were detected in some plants, but were not detected in others. This suggests that these fragments are autosomal or pseudoautosomal.

View Details

Fig. 3. The chromosomal locations of SlMF1. (a) Genomic Southern hybridization analysis using genomic DNA digested with HindIII. Each lane indicates DNA from male (1) and female (5) parents or male (2–4) and female (6–8) offspring of the USA laboratory line. (b) Genomic Southern hybridization analysis using genomic DNA digested with HindIII. Each lane indicates DNA from male (1) and female (6) parents or male (2–5) and female (6–10) offspring of the Japanese laboratory line. (c) PCR analyses on flow-sorted chromosomes. The PCR products were amplified with the set of SlMF1-F and SlMF1-R primers for the 390-bp fragments of SlMF1. PCR was performed using female (♀) and male (♂) genomic DNA, flow-sorted autosomes (A), X chromosomes (X), or no chromosomes (0) as templates. M represents a 100-bp ladder that was used as a DNA length marker.

To determine the X chromosomal location of SlMF1, we performed PCR with flow-sorted X chromosomes and autosomes using the SlMF1-F and SlMF1-R primers for the 390-bp region of SlMF1 cDNA. This PCR method can clearly confirm X chromosome linkage without the effects of contaminated autosomes in without the effects of contaminated autosomes in the flow-sorted X chromosomes as described previously (Kejnovsky et al. 2001). When PCR was performed using male and female genomic DNA, a single 290-bp fragment was amplified (Fig. 3). This genomic sequence is completely consistent with the cDNA sequence, indicating that the region has no intron. The fragments could be amplified from both flow-sorted X chromosomes and autosomes. The amplified sequences from both X chromosomes and autosomes were completely matched. The amplified sequences were also consistent with those from both male and female genomes.

DISCUSSION

SlMF1 was expressed in all the whorls of young flower buds; however, its expression was limited to the petals and anthers of mature male flower buds and the petals of mature female flower buds. This expression profile suggests that SlMF1 functions not only in the formation of young flower buds but also in the development of anthers and petals during maturation of the flower buds. The first morphological difference between males and females was found in the size of the fourth whorl of flower buds at stage 5 (Grant et al. 1994). Before morphological sex differences are clearly visible, certain differences in gene expression patterns are observed between male and female flower buds at stage 3. In situ hybridization analysis with two orthologs of floral identity genes–a PISTILLATA homolog SLM2 and an APETALA3 homolog SLM3–revealed that these are expressed more closely toward the center of the fourth male whorl compared with the fourth female whorl at stage 3, suggesting that these two genes correlate with a reduction in the size of the fourth whorl (Hardenack et al. 1994). In contrast, SlMF1 was expressed in all the whorls of both male and female flower buds at stage 3. This result suggests that SlMF1 is not directly involved in sex determination. At stage 9 of the male flower bud, pollen mother cells and tapetal cells begin to develop in the anther and petal primordia start to elongate rapidly. Two cell division genes, SlH4 and SlCycA1, were detected in both the anthers and petals, indicating that cells in the anthers and petals divide actively at this stage (Matsunaga et al. 2004). In particular, dividing cells could be detected in the upper region of the petal primordia. Consequently, the SlMF1 transcript accumulated in cells in the upper region of the petal primordia. This suggests that SlMF1 functions in the development of anthers and petal primordia.

Until now, six functional Y-linked genes have been characterized in this species. DD44, SlY1, SlY3, SlY4, and Slss are housekeeping genes with X-linked paralogs that are ubiquitously expressed in both sex organs (Moore et al. 2003; Delichere et al. 1999; Atanassov et al. 2001; Nicolas et al. 2005; Filatov et al. 2004). Only SlAP3Y has no X-linked homolog; however, it has an autosomal paralog (Matsunaga et al. 2003). The expressed sequence tags (ESTs)–Men-153 and Men-470 of two male flower buds have Y-linked homologous sequences (Scutt et al. 2002). When Southern hybridization was performed with each EST as a probe, male-specific fragments corresponding to Y-linked sequences were detected. However, the full-length cDNAs of these ESTs have not been reported. Genomic Southern hybridization with SlMF1 shows a pattern similar to that observed with these ESTs. This indicated that SlMF1 was a multicopy gene with homologous sequences on autosomes and the X and Y chromosomes. With the exception of individual polymorphisms, the 390-bp nucleotide sequences that were amplified from RNA, genomic DNA, flow-sorted autosomes and X chromosomes, were completely matched with SlMF1 cDNA sequences. Although we cannot completely exclude the possibility that either the paralog on the autosomes or the X chromosome may become a pseudogene, it is highly possible that SlMF1 has at least two paralogs on autosomes and the X chromosome. If so, these paralogs must have duplicated recently because no difference was observed in the 290-bp sequences between autosomal and X-linked SlMF1. Quantitative RT-PCR showed that the SlMF1 transcript was detected not only in male anthers and petals but also in female petals. This result demonstrates that SlMF1 is not derived from only a Y-linked gene. Moreover, we could not determine whether or not the 5.8-kb Y-linked sequence seen in Southern hybridization analyses is expressed. Characterization of the X- and Y-linked genomic fragments in further studies will provide information regarding whether the sex chromosome-linked genes are functional. Moreover, evolutionary analyses through comparison between the X- and Y-linked sequences of SlMF1 will reveal the degree of degeneration of the Y chromosome.

This work was supported by Grants-in-Aid for Scientific Research to S. M. (No. 16770049) from the Ministry of Education, Science, Culture, Sports, Science and Technology, Japan, and by Grant Agency of the Czech Republic (No. 204/05/2097).