Localization of Male-Specifically Expressed MROS Genes of Silene latifolia by PCR on Flow-Sorted Sex Chromosomes and Autosomes

Eduard Kejnovsk

, Pemysl Souek

, Jií irok

, Jaroslav Doleel

Corresponding author: Boris Vyskot, Laboratory of Plant Developmental Genetics, Institute of Biophysics, Academy of Sciences of the Czech Republic, Královopolská str. 135, CZ-612 65 Brno, Czech Republic., vyskot{at}ibp.cz (E-mail)

Communicating editor: D. CHARLESWORTH

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The dioecious white campion Silene latifolia (syn. Melandrium album) has heteromorphic sex chromosomes, XX in females and XY in males, that are larger than the autosomes and enable their separation by flow sorting. The group of MROS genes, the first male-specifically expressed genes in dioecious plants, was recently identified in S. latifolia. To localize the MROS genes, we used the flow-sorted X chromosomes and autosomes as a template for PCR with internal primers. Our results indicate that the MROS3 gene is located in at least two copies tandemly arranged on the X chromosome with additional copy(ies) on the autosome(s), while MROS1, MROS2, and MROS4 are exclusively autosomal. The specificity of PCR products was checked by digestion with a restriction enzyme or reamplification using nested primers. Homology search of databases has shown the presence of five MROS3 homologues in A. thaliana, four of them arranged in two tandems, each consisting of two copies. We conclude that MROS3 is a low-copy gene family, connected with the proper pollen development, which is present not only in dioecious but also in other dicot plant species.

THE majority of flowering plants are hermaphrodites, forming flowers with both female and male reproductive organs. Approximately 11% of plant species have unisexual flowers and 4% are dioecious with separate female and male individuals (for review, see GRANT et al. 1994 ). Heteromorphic sex chromosomes, which in plants are usually larger than autosomes, have evolved in only a few species like Silene latifolia or Rumex acetosa. Little is currently known about a molecular mechanism of sex determination in dioecious plants.

S. latifolia has recently become a popular model to study dioecy and evolution of plant sex chromosomes (for review, see CHARLESWORTH and GUTTMAN 1999 ; MONEGER et al. 2000 ). It possesses a pair of heteromorphic sex chromosomes; females are homogametic (2n = 24, XX) and males heterogametic (2n = 24, XY). The pair of sex chromosomes, X and Y, represents 16% of the total size of the male diploid genome (MATSUNAGA et al. 1994 ). The Y chromosome is 1.4 times larger than the X, which in turn is 1.6 times larger than the longest pair of autosomes and 1.9 times bigger than the average size of autosomes (calculated according to CIUPERCESCU et al. 1990 ). The intrinsic differences between autosomes and X and Y chromosomes make it possible to discriminate and sort individual chromosome types by flow cytometry (VEUSKENS et al. 1992 ). It is assumed that S. latifolia sex chromosomes evolved recently in contrast with the ancient origin of mammalian or insect sex chromosomes, thus offering the opportunity to study the early stages of sex chromosome evolution.

A number of research groups have recently isolated sex-specifically expressed or sex chromosome-linked genes and other DNA sequences in an attempt to find sex-determining genes in S. latifolia. They used (i) cDNA or genomic library subtraction methods (DONNISON et al. 1996 ; MATSUNAGA et al. 1996 ; BARBACAR et al. 1997 ; ROBERTSON et al. 1997 ; SCUTT et al. 1997 ; SCUTT and GILMARTIN 1998 ); (ii) microdissected and degenerate oligonucleotide-primed (DOP)-PCR amplified sex chromosomes as probes for screening of the cDNA library (DELICHERE et al. 1999 ) or for FISH (BUZEK et al. 1997 ; SCUTT et al. 1997 ); and (iii) PCR-based methods such as randomly amplified polymorphic DNA (MULCAHY et al. 1992 ). Although these attempts often resulted in isolation of repetitive sequences (DONNISON et al. 1996 ; BUZEK et al. 1997 ; SCUTT et al. 1997 ), a group of four male-specifically expressed genes, MROS genes, was discovered (MATSUNAGA et al. 1996 , MATSUNAGA et al. 1997 ). In addition, although not involved in sex determination, an active gene located on the Y chromosome, SlY1, with a functional homologue on the X, has been isolated (DELICHERE et al. 1999 ).

MROS (male reproductive organ-specific) genes are specifically expressed in male reproductive organs (MATSUNAGA et al. 1996 , MATSUNAGA et al. 1997 ): pollen grains (MROS1), late stages of anther maturation (MROS2), mature anther tapetum (MROS3), and early male flower buds (MROS4). The MROS genes were isolated by differential screening of the cDNA library prepared from male flower buds with cDNA from male and female flower buds separately. The MROS genes are present both in female and male genomes, so they cannot be located exclusively on the Y chromosome (MATSUNAGA et al. 1996 , MATSUNAGA et al. 1997 ). Other research groups isolated these genes or their alleles independently; Men-1 (SCUTT et al. 1997 ) is homologous to MROS1, Men-9 (ROBERTSON et al. 1997 ) and CCLS-4 (HINNISDAELS et al. 1997 ) are homologous to MROS3. Genetic analysis using a single-strand conformation polymorphism (SSCP) technique has indicated that only the MROS3 gene is located on the X chromosome, with a degenerate homologue (pseudogene) on the Y chromosome, whereas MROS1 and MROS2 are autosomal (GUTTMAN and CHARLESWORTH 1998 ). Moreover, two genomic clones of MROS3 were isolated recently and a single pollen PCR-based typing analysis suggests their localization either on the autosomes or in the homologous region of the X and Y chromosomes (MATSUNAGA et al. 1999 ).

Here we describe physical localization of MROS1 to MROS4 genes of S. latifolia by PCR on the flow-sorted X chromosomes and autosomes. We show that all these MROS genes are localized on the autosomes. In addition, at least two copies of MROS3 are X-linked, organized in the head-to-tail tandem array. Our data from PCR experiments and database homology searching indicate that MROS3 gene homologues are present also in other Silene species as well as in A. thaliana.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant material:
S. latifolia Poiret, S. diclinis (Lag.) Lainz, and S. vulgaris (Moench) Garcke plant material comes from the seed collection of the Institute of Biophysics, Brno, Czech Republic. As a routine source of S. latifolia mitotic chromosomes for flow sorting, hairy root cultures were established from a tetraploid female line after infection with Agrobacterium rhizogenes, strain A4RS (SIROKY et al. 1999 ). To synchronize the cell cycle in root tip meristems, aphidicolin (30 µM, Sigma, St. Louis) was added for 12 hr. Mitoses were then accumulated with 15 µM oryzalin (4 hr, Elanco).

Chromosome isolation and sorting:
Synchronized root tips were cut 1 cm from the root tip, rinsed in distilled water, and fixed for 20 min at 5° in 2% (v/v) formaldehyde made in Tris buffer (10 mM Tris, 10 mM Na₂EDTA, 100 mM NaCl, pH 7.5) supplemented with 0.1% Triton X-100 (DOLEZEL et al. 1992 ). After three 5-min washes in Tris buffer, meristem tips (1 mm) of 200 roots were cut and transferred to a 5-ml polystyrene tube containing 1 ml LB01 lysis buffer (DOLEZEL et al. 1989 ). The chromosomes were released by mechanical homogenization with a Polytron PT1200 homogenizer (Kinematica AG, Littau, Switzerland) at 15,000 rpm for 10 sec. The suspension was passed through a 50-µm pore size nylon mesh, stained with 4'6-diamidino-2-phenylindole (DAPI) at a final concentration of 2 µg/ml and analyzed at rates of 200–400 particles per sec using a FACSVantage flow cytometer (Becton Dickinson, San Jose, CA). The cytometer was equipped with an argon-ion laser tuned to multiline UV and run with a 300-mW output power. The system threshold was set on the fluorescence pulse height (FL1-H) and the gate window was set on a dot plot of FL1-H vs. forward light scatter to eliminate debris with extremely high or low fluorescence intensity. To achieve the highest purity in sorted fractions, two-step sorting was employed (LUCRETTI et al. 1993 ). Sorting gates were set on a dot plot of fluorescence pulse area vs. fluorescence pulse width and at least 25,000 chromosomes were sorted at rates of 5–10 per sec into a polystyrene tube containing 400 µl of 1.5x LB01. Before the second sorting run, DAPI was added to the suspension, enriched for given chromosome type, and chromosomes were sorted at rates of 20 per sec. For the analysis of purity in sorted fractions using fluorescence in situ hybridization (FISH), 1000 chromosomes were sorted into a 15-µl drop of buffer containing 5% sucrose on microscope slides (KUBALAKOVA et al. 2000 ). The slides were air dried after sorting and maintained at room temperature until use. For PCR, chromosomes were sorted into 33 µl of sterile deionized water in 0.5-ml PCR reaction tubes and stored at -70°.

Fluorescence in situ hybridization:
As a FISH probe for 25S rDNA, an internal biotinylated 2.5-kb EcoRI fragment of 25S rRNA gene was used. The hybridization mix (20 µl per slide) consisted of 200 ng of the labeled probe, 6 µg autoclaved salmon sperm DNA (Serva), 4 µl of 50% solution of dextran sulfate (Sigma), 10 µl formamide (Sigma), and 2 µl of 20x SSC. Denatured probe was added to the denatured slides with sorted chromosomes and hybridized for 12 hr. After a stringent washing, biotin was detected by FITC-conjugated avidin (Vector, Burlingame, CA). FISH signals were observed under Olympus AX 70 fluorescent microscope.

PCR on sorted chromosomes:
Before PCR, chromosomes were spun down and then 16 µl of master mix containing PCR buffer (Promega, Madison, WI), dNTP, MgCl₂, and primers were added. The final concentrations of the reagents were 0.2 mM dNTP, 0.2 µM primers, 1.5 mM MgCl₂, 50 mM KCl, and 10 mM Tris-HCl, pH 8.0. After initial denaturation at 94° for 10 min, temperature was decreased to 85° and 1 µl (5 units) of Taq polymerase (Promega) was added. DNA was amplified using 40 cycles (94°/40 sec–1 min, 50–60°/40 sec–1.5 min, 72°/1–2 min) with the final primer extension step at 72°/ 5 min. Temperatures and incubation times in PCR profile were varied depending on primers used. In all PCR experiments a PTC-200 thermal cycler (MJ Research) was used. Identity of PCR products was verified using restrictase digestion performed according to manufacturer's instructions. For reamplification with nested primers, 0.5 µl of original PCR product was used as a template for 20 cycles of PCR at conditions described above.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Flow sorting of chromosomes and purity of fractions:
Cultures from a tetraploid female (Fig 1A) were used to obtain a higher number of metaphase X chromosomes and autosomes per cell without a risk of contamination with Y chromosomes. Flow cytometric analysis of chromosome suspensions clearly discriminated a dominant peak corresponding to the population of autosomes and a smaller peak corresponding to the X chromosomes (Fig 1B). While the fractions of sorted autosomes obviously did not contain the (larger) X chromosomes or their separated chromatids (Fig 1C), sorted fractions of X chromosomes were contaminated to some extent with doublets of shorter autosomes (Fig 1D). To improve the purity of the X chromosome sorting, DNA content and chromosome length were analyzed simultaneously (see MATERIALS AND METHODS). Furthermore, a two-step sorting procedure was always used to ensure maximum purity. Chromosomal fractions were sorted in parallel into Eppendorf tubes for PCR analysis and onto microscopic slides to check their purity after DAPI staining and FISH.

View larger version (61K): In this window In a new window Download PPT slide   Figure 1. Sorting of chromosomes and their purity. (a) Fluorescence in situ hybridization on tetraploid metaphase plate (4n = 48, 4X) used for the sorting. "X" indicates the X chromosomes; bright spots at the end of 10 pairs of autosomes show 25S rDNA locations. (b) Flow karyotype obtained after the analysis of DAPI-stained chromosome suspension. Fluorescence in situ hybridization on flow-sorted autosomes (c) and the X chromosomes (d) with biotin-labeled 25S rDNA probe (bright spots represent the hybridization signals). Inserts in d show three possible variants of contaminating autosomal doublets found—both autosomes without any 25S rDNA label (left), only one autosome with the 25S rDNA cluster (middle), and two autosomes with the 25S rDNA signals (right).

Because of the high sensitivity of PCR, contamination of a chromosomal fraction lacking the gene in question with a few different chromosomes harboring this DNA sequence may result in a weak amplification fragment. To check the extent of cross-contamination by mis-sorted chromosomes, we applied both chromosomal fractions (X chromosomes and autosomes) onto microscopic slides in parallel with flow sorting of chromosomes into Eppendorf tubes for PCR. The chromosomal fractions on slides were then hybridized with labeled 25S rDNA. While the 25S rDNA clusters are distally located on 5 of 11 autosome pairs (45%) they are absent from the sex chromosomes (MATSUNAGA et al. 1994 ). The presence of 10 25S rDNA clusters on the tetraploid material used is shown using the FISH technique (Fig 1A). Microscopic observations of 500 chromosomes of each fraction after the fluorescence in situ hybridization with 25S rDNA probe did not indicate a detectable "statistical" contamination of autosomes by X chromosomes because 45% of chromosomes in the autosomal fraction were 25S rDNA positive (Fig 1C). However, in the X chromosome fraction, all autosomes were present in the form of doublets (Fig 1D). We observed 12% contamination of the X chromosomes by autosomes (44 X chromosomes and 6 autosomes in Fig 1D), i.e., 1% contamination by each type of 11 different autosomes.

A hairy root culture prepared from a male plant was used as a source of the Y chromosomes. Unfortunately, microscopic analysis of slides with sorted chromosomes showed an abundant presence of chromosomal clumps (consisting mainly of two to four autosomes) in the Y interest sorting zone, which prevented the sorting of the Y chromosomes to a purity acceptable for PCR (not shown).

The MROS1, MROS2, and MROS4 genes are autosomal:
Flow-sorted autosomes and X chromosomes were used as templates for PCR with primers specific for the individual MROS genes (Table 1). Primers used for amplification of MROS1, MROS2, and MROS4 were designed on the basis of the known cDNA nucleotide sequences (MATSUNAGA et al. 1996 , MATSUNAGA et al. 1997 ); in all cases the primer sites were located in coding regions of the MROS genes. The only exception was the MROS1-F5 primer, located in the MROS1 promoter region. While the MROS1 and MROS4 genes possess introns, there are no introns in MROS2. In all PCR experiments, the autosomal fractions contained 11 times more chromosomes (1100) than the fraction of the X chromosomes (100) to ensure the same numbers of each specific chromosome (since n = 11A + X). Using the MROS1 gene-specific primers MROS1-F5 and MROS1-R1, a single band of the expected size (1375 bp) was obtained with the sorted autosomes as a template, but not with the X chromosomes (Fig 2A). PCR with the MROS2 gene-specific primers MROS2-F1 and MROS2-R1 also resulted in a specific band of 785 bp only in the autosomal fraction (Fig 2B), while PCR on the X chromosomes did not yield any product. Similarly, the primers designed for the MROS4 gene led to amplification of a specific band of 600 bp only in the autosomes (Fig 2C). Female and male genomic DNA representing positive controls as well as negative controls without any DNA template were always included. For all three pairs of primers, female and male templates yielded the same products as in the autosomal fractions, while no product was detected in blank controls.

View larger version (47K): In this window In a new window Download PPT slide   Figure 2. Agarose gel electrophoresis of PCR products obtained with (a) MROS1 gene-specific primers MROS1-F5 and MROS1-R1, (b) MROS2 gene-specific primers MROS2-F1 and MROS2-R1, (c) MROS4 gene-specific primers MROS4-F1 and MROS4-R1, and (d) MROS3 gene-specific primers INF2 and R3X2. No template (0), 1100 autosomes (A), 100 X chromosomes (X), male genomic DNA (m), and female genomic DNA (f). M, DNA length markers: 1-kb ladder (a and d) and pBR322/AluI (b and c).

Due to the cross-contamination of the X chromosomal fraction with autosomes (as demonstrated in Fig 1D), it was necessary to perform control PCR experiments with increasing numbers of sorted chromosomes as a template. PCR with primers MROS4-F1 and MROS4-R1, specific for the autosomal MROS4 gene, resulted in a weak band of the expected size when only 55 autosomes (statistically representing 5 of each from 11 different autosomes) were used, while giving a strong band starting at 110 autosomes (Fig 3A). On the other hand, PCR with the same primer pair using an increasing number of X chromosomes resulted in a very weak band only when 500 X chromosomes were applied (Fig 3B). We believe that these data reflect contamination of 500 X chromosomes by <55 autosomes, thus representing 11% contamination at maximum.

View larger version (95K): In this window In a new window Download PPT slide   Figure 3. Agarose gel electrophoresis of PCR products obtained using the MROS4 gene-specific primers MROS4-F1 and MROS4-R1. Numbers of the flow-sorted autosomes (a) or X chromosomes (b) are indicated. No DNA template (0), male genomic DNA (m), and female genomic DNA (f) were used as controls. M, DNA length marker (pBR322/AluI).

The MROS3 gene is both autosomal and X-linked with tandemly arranged copies on the X chromosome:
MROS3 genes were amplified using the primers described previously by GUTTMAN and CHARLESWORTH 1998 . The MROS3 gene lacks introns. PCR with MROS3 primers INF2 and R3X2 yielded a strong band of 433 bp in the autosomes. In the X chromosomes, a specific 433-bp band plus several additional bands corresponding to fragments >1000 bp were present (Fig 2D). This result indicates that the MROS3 gene is present on the X chromosome in multiple copies, probably tandemly arranged. To confirm the tandem arrangement of MROS3 copies, we designed a pair of "inverse" primers INF1 and INR1 (Table 1) in the terminal regions of the MROS3 gene and directed outward as shown in Fig 4A. PCR with these primers resulted in amplification of a fragment of 1700 bp from X chromosomes (Fig 4B) but not from autosomes (not shown). This PCR product represents the intergenic region (spacer) between two copies of MROS3 genes flanked by the terminal parts of MROS3 genes. The specificity of this PCR product was confirmed by reamplification using two pairs of nested primers (Fig 4A). One pair of nested primers (INF1 + R3X2) reamplified a specific 73-bp-long region from the 3' end of the first (left) MROS3 gene; the second pair of nested primers (BF1 + INR1) reamplified the promoter region of the second (right) MROS3 gene (Fig 4B). The lengths of both nested PCR products were 73 and 400 bp, respectively, as expected from the known nucleotide sequence. In addition, we performed PCR using only one primer, INF1 or R3X2, to reveal potential tandemly arranged MROS3 genes in inverted orientation (not shown). The absence of any PCR product suggested that there are no tandemly arranged MROS3 genes in inverted orientation that are close enough to be amplified by PCR.

View larger version (50K): In this window In a new window Download PPT slide   Figure 4. A diagram of the tandem arrangement of two copies of the MROS3 gene separated with an intergenic spacer, PCR product, and two reamplification products expected on the basis of this hypothetical scheme (a). Agarose gel electrophoresis of PCR products (b) obtained with inverse and nested primers, respectively, as shown in a. As a template for PCR, 1100 sorted X chromosomes were used. M, DNA length markers: 1-kb ladder (left) and pBR322/AluI (right).

Evidence for specificity of PCR products:
On the basis of the knowledge of nucleotide sequences of MROS genes (MATSUNAGA et al. 1996 , MATSUNAGA et al. 1997 ), the specificity of PCR-amplified products was checked by restriction enzyme digestion into fragments of the expected sizes or by reamplification of PCR products using nested primers (Fig 5). The PCR fragment corresponding to the MROS1 gene (1375 bp) was reamplified with one original primer MROS1-R1 in combination with the nested primer MROS1-F1 located 247 bp from the 5' end, resulting in the PCR reamplification product of 1128 bp. Digestion of the MROS2 PCR product (785 bp) with MspI resulted in two fragments of 597 and 188 bp, as expected according to the one interprimer MspI restriction site. Similarly, the fragment obtained by PCR with MROS3 gene-specific primers (433 bp) was digested with MspI and yielded, as expected, two fragments of 257 and 176 bp in length. The PCR fragment corresponding to the MROS4 gene was reamplified using nested primers MROS4-F2 and MROS4-R2 located 60 and 71 bp apart from the 5' and 3' ends, respectively. As expected, the reamplification resulted in a PCR product 131 bp shorter than the original PCR fragment.

View larger version (48K): In this window In a new window Download PPT slide   Figure 5. Verification of the MROS PCR products by agarose gel electrophoresis of restriction fragments or reamplified products driven by nested primers. MROS1 identification, the basic PCR product (lane 1), and its product derived from the nested primers (lane 2). The MROS2 PCR product (lane 3) and the same after MspI digestion (lane 4; the smaller fragment is hardly visible). The PCR product with MROS3 primers (lane 5) and cut with MspI (lane 6). The PCR product with MROS4 primers (lane 7) and the same after reamplification with nested primers (lane 8). DNA length markers: 1-kb ladder (left) and pBR322/AluI (right).

The MROS3 gene is also present in other plant species:
To check whether the MROS3 gene is also present in other closely related plant species, we used MROS3 gene-specific primers INF1 and R3X2 for PCR with genomic DNA template prepared from other Silene species, the dioecious S. diclinis and the gynodioecious species S. vulgaris. In both species, we obtained a MROS3 band of the same size as in S. latifolia (not shown). We also used two copies of the S. latifolia MROS3 gene, MROS3a and MROS3b, published recently by MATSUNAGA et al. 1999 to search the entire Arabidopsis thaliana genome. This database search revealed that A. thaliana has five genes homologous to MROS3. They were named AtMROS3a (accession no. AAC19282), AtMROS3b (AAC19269), AtMROS3c (AAF02885), AtMROS3d (AAF02163), and AtMROS3e (AAF02153). None of the AtMROS3 genes have an intron. A database search showed that AtMROS3a and AtMROS3b as well as AtMROS3d and AtMROS3e are arranged as tandem repeats on chromosomes IV and III, respectively. Fig 6A shows a multiple alignment of the five A. thaliana and two S. latifolia MROS3 homologues and an evolutionary tree derived from this alignment (Fig 6B).

View larger version (53K): In this window In a new window Download PPT slide   Figure 6. (a) Alignment of deduced amino acid sequences of five A. thaliana AtMROS3 genes including AtMROS3a (accession no. AAC19282), AtMROS3b (AAC19269), AtMROS3c (AAF02885), AtMROS3d (AAF02163), and AtMROS3e (AAF02153) and two S. latifolia MROS3 genes including MROS3A (AB029398) and MROS3B (AB029399). The numbering on the right denotes the position of amino acid residues from the putative translational initiation. Dashes indicate gaps introduced to maximize the extent of homology among sequences. Solid boxes indicate conserved amino acids residues. Asterisks indicate complete consensus sequences. (b) Phylogenic analysis of deduced amino acid sequences of AtMROS3 genes and MROS3 genes. The clustering was performed using the program CLUSTAL X (THOMPSON et al. 1997 ). The branch lengths are proportional to the genetic distances by the neighbor-joining method (SAITOU and NEI 1987 ).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Very little is known about molecular mechanisms of sex determination in dioecious plants and the structure and evolution of plant sex chromosomes. S. latifolia has heteromorphic sex chromosomes that are much bigger than autosomes and it has been suggested that they represent an early stage in the evolution of sex chromosomes. Until now not many genes have been isolated in S. latifolia (MATSUNAGA et al. 1996 , MATSUNAGA et al. 1997 ; HINNISDAELS et al. 1997 ; ROBERTSON et al. 1997 ; SCUTT et al. 1997 ; DELICHERE et al. 1999 ) and only a few of them—MROS3 and SlX/SlY genes—have been localized on the sex chromosomes (GUTTMAN and CHARLESWORTH 1998 ; DELICHERE et al. 1999 ). In these studies genetic mapping methods were used, which cannot distinguish between localization of genes on autosomes and in the pseudoautosomal region of sex chromosomes. For example, the X-linkage of the MROS3 gene was derived from SSCP analysis, but the same pattern could also occur with some probability for an autosomal gene (GUTTMAN and CHARLESWORTH 1998 ). To supplement genetic analyses, direct physical techniques to localize genes on individual chromosomes should be used. These include fluorescence in situ hybridization and PCR on microdissected (MACAS et al. 1993A ) or flow-sorted chromosomes (MACAS et al. 1993B ). In situ hybridization of plant chromosomes is not yet a reliable technique for mapping short (<10 kb) single copy DNA sequences, while microdissection is a rather laborious method yielding only a small number of chromosomes of interest. PCR on flow-sorted chromosomes is a more straightforward and reliable approach allowing localization of single or low-copy genes or other DNA sequences.

Here we used this technique to localize the MROS genes on S. latifolia chromosomes. Chromosome sorting by flow cytometry requires chromosome suspensions with sufficient concentrations of intact chromosomes. The first protocol for chromosome isolation in S. latifolia involved the use of hairy root cultures (VEUSKENS et al. 1992 ). The procedure achieved a high degree of mitotic synchrony and clonal maintenance of large numbers of roots of either sex. Here we have used the same protocol to grow and synchronize hairy roots. However, unlike the earlier procedure, intact chromosomes were released from root meristems mechanically after mild formaldehyde fixation according to DOLEZEL et al. 1992 . The fixation made chromosomes resistant to mechanical shearing forces, and thus two-step sorting could be employed. Furthermore, fixed chromosomes were suitable for FISH after sorting onto microscope slides (cf. DOLEZEL et al. 1999 , DOLEZEL et al. 2001 ). While the purity of the X chromosome fraction was very good, we had problems sorting the Y chromosomes in sufficient purity, due to the presence of autosomal clumps. Similarly, VEUSKENS et al. 1992 , VEUSKENS et al. 1995 reported chromosome clusters to be the sole contamination in the Y fractions. Nevertheless, the authors reported 60–80% (VEUSKENS et al. 1992 ) and even 90% (VEUSKENS et al. 1995 ) purity in the Y chromosome fractions. Although we have used a different procedure for the chromosome release, the resolution of flow karyotypes was comparable and thus the reason for lower purity in our Y fractions is not clear. KUBALAKOVA et al. 2000 noted that the chromosome morphology may be changed during chromosome isolation, sorting, and drying on a flat surface and recommended identification of sorted chromosomes after specific labeling. As VEUSKENS et al. 1992 , VEUSKENS et al. 1995 evaluated the purity in the sorted Y chromosome fractions on the basis of chromosome morphology alone, one may speculate that the fractions were contaminated by the X chromosomes and/or autosomes, which were not detected.

PCR on flow-sorted plant chromosomes has proved to be a powerful and direct method for physical localization of genes and other DNA sequences. Here, for the first time, physical localization of genes using PCR and sequence-specific primers on the flow-sorted plant sex chromosomes is demonstrated. The sensitivity of PCR on sorted chromosomes is a critical factor because only a few copies of a gene of interest are sufficient to serve as a template. Moreover, the efficiency of PCR on chromosomes consisting of condensed DNA complexes with proteins could be lower in comparison to PCR using pure DNA. In our experiments, a relatively very high sensitivity was achieved and we were able to amplify the single copy MROS4 gene using 55 autosomes, representing an average of only 5 of each chromosome. This sensitivity is comparable to the data described by other authors (MACAS et al. 1993A , MACAS et al. 1993B ).

We showed that all four MROS genes are located on autosomes with at least two additional copies of MROS3 on the X chromosome. Our results support the previously published data indicating the X-linkage of MROS3 and autosomal localization of MROS1 and MROS2 (GUTTMAN and CHARLESWORTH 1998 ), as well as results describing two autosomal copies of the MROS3 gene (MATSUNAGA et al. 1999 ). In addition, GUTTMAN and CHARLESWORTH 1998 isolated an MROS3 pseudogene located on the Y chromosome. We can conclude that the MROS3 gene is not a single copy gene but rather a low-copy gene forming a gene family with a few members spread on the autosomes as well as the X and Y chromosomes.

MROS3 genes are also present in other Silene species. In addition, MROS3 homologues also were found in a nonrelated A. thaliana genome, suggesting an ancient origin of this gene. We showed that at least two MROS3 genes are arranged in tandem on the X chromosome of S. latifolia. Surprisingly, a database search revealed that in A. thaliana the MROS3 homologues also are tandemly arranged. However, on the basis of our results we cannot conclude whether this duplication event(s) took place in an ancestral genome or independently in S. latifolia and A. thaliana, giving rise to orthologous or paralogous genes, respectively.

The five A. thaliana MROS3 homologues have the same characteristics as the S. latifolia MROS3 genes: (i) The N-terminal regions in their encoded proteins are hydrophobic putative signal peptides (Fig 6A), (ii) the consensus sequence (P-G–PKGV) is found in homologous regions of the proteins (Fig 6A), and (iii) they lack introns. The proportion of proteins belonging to families of more than five members, such as AtMROS3, is relatively higher in the A. thaliana genome than in other eukaryotic genomes (THE ARABIDOPSIS GENOME INITIATIVE 2000). Moreover, most gene families are organized in tandem arrays. The tandem pairs AtMROS3a and AtMROS3b on chromosome IV and AtMROS3e and AtMROS3d on chromosome III are examples of this. AtMROS3c is localized on chromosome I. AtMROS3 genes fall into two groups in accordance with their chromosomal localization (Fig 6B). This suggests that the tandem arrays were generated by duplication of an ancestral duplicate gene. The duplication is conserved on the X chromosome of S. latifolia (Fig 4). It is possible that the X chromosome has some regions homologous to chromosome III or IV of A. thaliana and a study of chromosomal synteny may be interesting.

	ACKNOWLEDGMENTS

The authors are grateful to Drs. Jií Macas and Ioan Negrutiu for fruitful discussions. We are indebted to Drs. J. íhalíková and M. Kubaláková for help with the preparation of chromosome suspensions. We thank Dr. M. Lysák and K. Rychtarová for help with chromosome sorting and J. Weiserová, BSc. for excellent technical assistance. This work was supported by the Grant Agency of the Czech Republic (521/96/K117 and 521/99/0696) and National Science Foundation-Ministry of Education grant No. 380 (2000).

Manuscript received February 1, 2001; Accepted for publication April 6, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BARBACAR, N., S. HINNISDAELS, I. FARBOS, F. MONÉGER, and A. LARDON et al., 1997  Isolation of early genes expressed in reproductive organs of the dioecious white campion (Silene latifolia) by subtraction cloning using an asexual mutant. Plant J. 12:805-817[Medline].

BUEK, J., H. KOUTNÍKOVÁ, A. HOUBEN, K. ÍHA, and B. JANOUEK et al., 1997  Isolation and characterization of X chromosome-derived DNA sequences from a dioecious plant Melandrium album. Chromosome Res. 5:57-65[Medline].

CHARLESWORTH, D., and D. S. GUTTMAN, 1999 The evolution of dioecy and plant sex chromosome systems, pp. 25–49 in Sex Determination in Plants, edited by C. C. AINSWORTH. Bios Scientific Publishers, Oxford.

CIUPERCESCU, D. D., J. VEUSKENS, A. MOURAS, D. YE, and M. BRIQUET et al., 1990  Karyotyping Melandrium album, a dioecious plant with heteromorphic sex chromosomes. Genome 33:556-562.

DELICHÉRE, C., J. VEUSKENS, M. HERNOULD, N. BARBACAR, and A. MOURAS et al., 1999  SlY1, the first active gene cloned from a plant Y chromosome, encodes a WD-repeat protein. EMBO J. 18:4169-4179[Medline].

DOLEEL, J., P. BINAROVÁ, and S. LUCRETTI, 1989  Analysis of nuclear DNA content in plant cells by flow cytometry. Biol. Plant 31:113-120.

DOLEEL, J., J. ÍHALÍKOVÁ, and S. LUCRETTI, 1992  A high-yield procedure for isolation of metaphase chromosomes from root tips of Vicia faba. L. Planta 188:93-98.

DOLEEL, J., J. MACAS and S. LUCRETTI, 1999 Flow analysis and sorting of plant chromosomes, pp. 5.3.1.–5.3.33. in Current Protocols in Cytometry, edited by J. P. ROBINSON, Z. DARZYNKIEWICZ, P. N. DEAN, L. G. DRESSLER, A. ORFAO et al. John Wiley & Sons, New York.

DOLEEL, J., M. A. LYSÁK, M. KUBALÁKOVÁ, H. IMKOVÁ, J. MACAS et al., 2001 Sorting of plant chromosomes, pp. 3–31 in Methods in Cell Biology, Vol. 64, Ed. 3, part B, edited by Z. DARZYNKIEWICZ, H. A. CRISSMAN and J. P. ROBINSON. Academic Press, San Diego.

DONNISON, I. S., J. IROK, B. VYSKOT, H. SAEDLER, and S. GRANT, 1996  Isolation of Y chromosome-specific sequences from Silene latifolia and mapping of male sex determining genes using representational difference analysis. Genetics 149:1891-1899.

GRANT, S., A. HOUBEN, B. VYSKOT, J. IROK, and W. H. PAN et al., 1994  Genetics of sex determination in flowering plants. Dev. Genet. 15:214-230.

GUTTMAN, D. S. and D. CHARLESWORTH, 1998  An X-linked gene with a degenerate Y-linked homologue in a dioecious plant. Nature 393:263-266[Medline].

HINNISDAELS, S., A. LARDON, N. BARBACAR, and I. NEGRUTIU, 1997  A floral third whorl-specific marker gene in the dioecious species white campion is differentially expressed in mutants defective in stamen development. Plant Mol. Biol. 35:1009-1014[Medline].

KUBALÁKOVÁ, M., M. A. LYSÁK, J. VRÁNA, H. IMKOVÁ, and J. ÍHALÍKOVÁ et al., 2000  Rapid identification and determination of purity of flow-sorted plant chromosomes using C-PRINS. Cytometry 41:102-108[Medline].

LUCRETTI, S., J. DOLEEL, J. I. SCHUBERT, and J. FUCHS, 1993  Flow karyotyping and sorting of Vicia faba chromosomes. Theor. Appl. Genet. 85:665-672.

MACAS, J., W. WESCHKE, H. BAUMLEIN, U. PICH, and A. HOUBEN et al., 1993a  Localization of vicilin genes via polymerase chain reaction on microisolated field bean chromosomes. Plant J. 3:883-886[Medline].

MACAS, J., J. DOLEEL, S. LUCRETTI, U. PICH, and A. MEISTER et al., 1993b  Localization of seed protein genes on flow-sorted bean chromosomes. Chromosome Res. 1:107-115[Medline].

MATSUNAGA, S., M. HIZUME, S. KAWANO, and T. KUROIWA, 1994  Cytological analyses in Melandrium album: genome size, chromosome size and fluorescence in situ hybridization. Cytologia 59:135-141.

MATSUNAGA, S., S. KAWANO, H. TAKANO, H. UCHIDA, and A. SAKAI et al., 1996  Isolation and developmental expression of male reproductive organ-specific genes in a dioecious campion, Melandrium album (Silene latifolia). Plant J. 10:679-689[Medline].

MATSUNAGA, S., S. KAWANO, and T. KUROIWA, 1997  MROS1, a male stamen-specific gene in the dioecious campion Silene latifolia is expressed in mature pollen. Plant Cell Physiol. 38:499-502[Abstract/Free Full Text].

MATSUNAGA, S., K. SCHUTZE, I. S. DONNISON, S. R. GRANT, and T. KUROIWA et al., 1999  Single pollen typing combined with laser-mediated manipulation. Plant J. 20:1-8[Medline].

MONÉGER, F., N. BARBACAR, and I. NEGRUTIU, 2000  Dioecious Silene at the X-road: the reason Y. Sex. Plant Reprod. 12:245-249.

MULCAHY, D. L., N. F. WEEDEN, R. KESSELI, and S. B. CARROLL, 1992  DNA probes for the Y-chromosome of Silene latifolia, a dioecious angiosperm. Sex. Plant Reprod. 5:86-88.

ROBERTSON, S. E., Y. LI, C. P. SCUTT, M. E. WILLIS, and P. M. GILMARTIN, 1997  Spatial expression dynamics of Men-9 delineate the third floral whorl in male and female flowers of dioecious Silene latifolia. Plant J. 12:155-168[Medline].

SAITOU, N. and M. NEI, 1987  The neighbor-joining method: a new method for reconstructing phylogenetics trees. Mol. Biol. Evol. 4:406-425[Abstract].

SCUTT, C. P. and P. M. GILMARTIN, 1998  The Men-10 cDNA encodes a novel form of proline-rich protein expressed in the tapetum of dioecious Silene latifolia. Sex. Plant Reprod. 11:236-240.

SCUTT, C. P., Y. LI, S. E. ROBERTSON, M. E. WILLIS, and P. M. GILMARTIN, 1997  Sex determination in dioecious Silene latifolia—effects of the Y chromosome and the parasitic smut fungus (Ustilago violacea) on gene expression during flower development. Plant Physiol. 114:969-979[Abstract].

IROK, J., J. HODURKOVÁ, I. NEGRUTIU, and B. VYSKOT, 1999  Functional and structural chromosome analyses in autotetraploid Silene latifolia. Ann. Bot. 84:633-638[Abstract/Free Full Text].

Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. (2000) Nature 408:796-815[Medline].

THOMPSON, J. D., T. J. GIBSON, F. PLEWNIAK, F. JEANMOUGIN, and D. G. HIGGINS, 1997  The CLUSTAL X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882[Abstract/Free Full Text].

VEUSKENS, J., D. MARIE, S. HINNISDAELS and S. C. BROWN, 1992 Flow cytometry and sorting of plant chromosomes, pp. 177–188 in Flow Cytometry and Cell Sorting, edited by A. RADBRUCH. Springer Verlag, Berlin.

VEUSKENS, J., D. MARIE, S. C. BROWN, M. JACOBS, and I. NEGRUTIU, 1995  Flow sorting of the Y sex chromosome in the dioecious plant Melandrium album. Cytometry 21:363-373[Medline].

This article has been cited by other articles:

				T. R. Meagher Linking the Evolution of Gender Variation to Floral Development Ann. Bot., August 1, 2007; 100(2): 165 - 176. [Abstract] [Full Text] [PDF]

				R. Bergero, A. Forrest, E. Kamau, and D. Charlesworth Evolutionary Strata on the X Chromosomes of the Dioecious Plant Silene latifolia: Evidence From New Sex-Linked Genes Genetics, April 1, 2007; 175(4): 1945 - 1954. [Abstract] [Full Text] [PDF]

				M. Kubalakova, P. Kovarova, P. Suchankova, J. Cihalikova, J. Bartos, S. Lucretti, N. Watanabe, S. F. Kianian, and J. Dolezel Chromosome Sorting in Tetraploid Wheat and Its Potential for Genome Analysis Genetics, June 1, 2005; 170(2): 823 - 829. [Abstract] [Full Text] [PDF]

				D. A. Filatov Evolutionary History of Silene latifolia Sex Chromosomes Revealed by Genetic Mapping of Four Genes Genetics, June 1, 2005; 170(2): 975 - 979. [Abstract] [Full Text] [PDF]

				D. A. Filatov Substitution Rates in a New Silene latifolia Sex-Linked Gene, SlssX/Y Mol. Biol. Evol., March 1, 2005; 22(3): 402 - 408. [Abstract] [Full Text] [PDF]

				S. Matsunaga, E. Isono, E. Kejnovsky, B. Vyskot, J. Dolezel, S. Kawano, and D. Charlesworth Duplicative Transfer of a MADS Box Gene to a Plant Y Chromosome Mol. Biol. Evol., July 1, 2003; 20(7): 1062 - 1069. [Abstract] [Full Text] [PDF]

				K. Ishizaki, Y. Shimizu-Ueda, S. Okada, M. Yamamoto, M. Fujisawa, K. T. Yamato, H. Fukuzawa, and K. Ohyama Multicopy genes uniquely amplified in the Y chromosome-specific repeats of the liverwort Marchantia polymorpha Nucleic Acids Res., November 1, 2002; 30(21): 4675 - 4681. [Abstract] [Full Text] [PDF]

				I. Negrutiu, B. Vyskot, N. Barbacar, S. Georgiev, and F. Moneger Dioecious Plants. A Key to the Early Events of Sex Chromosome Evolution Plant Physiology, December 1, 2001; 127(4): 1418 - 1424. [Full Text] [PDF]