• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of embojLink to Publisher's site
EMBO J. Jun 17, 2002; 21(12): 3081–3095.
PMCID: PMC126045

Homolog interaction during meiotic prophase I in Arabidopsis requires the SOLO DANCERS gene encoding a novel cyclin-like protein


Interactions between homologs in meiotic prophase I, such as recombination and synapsis, are critical for proper homolog segregation and involve the coordination of several parallel events. However, few regulatory genes have been identified; in particular, it is not clear what roles the proteins similar to the mitotic cell cycle regulators might play during meiotic prophase I. We describe here the isolation and characterization of a new Arabidopsis mutant called solo dancers that exhibits a severe defect in homolog synapsis, recombination and bivalent formation in meiotic prophase I, subsequently resulting in seemingly random chromosome distribution and formation of abnormal meiotic products. We further demonstrate that the mutation affects a meiosis-specific gene encoding a novel protein of 578 amino acid residues with up to 31% amino acid sequence identity to known cyclins in the C-terminal portion. These results argue strongly that homolog interactions during meiotic prophase I require a novel meiosis-specific cyclin in Arabidopsis.

Keywords: Arabidopsis/cyclin/meiosis/recombination/synapsis


Meiosis is required for eukaryotic sexual reproduction and provides an important mechanism for generating genetic diversity among individuals of a species. Cytological studies indicate that the meiotic prophase I is a long and complex stage that differs from the mitotic prophase (see reviews by Ashley and Plug, 1998; Dawe, 1998; Zickler and Kleckner, 1999). During prophase I, chromosomes condense and homologous chromosomes (homologs) pair. The prophase I continuum is divided into five substages based on chromosomal characteristics: first, in the leptotene stage, chromosomes begin to condense and can be seen as thin thread-like structures. Homolog pairing initiates in leptotene and continues extensively in the zygotene stage. In the pachytene stage, fully synapsed homologs are observed as thick thread-like structures under a light microscope. The synaptonemal complex (SC) forms at this stage and can be observed in detail using electron microscopy. During the diplotene stage, homologs become desynapsed along much of their length, but remain attached through chiasmata. Finally, during the diakinesis stage, chromosomes contract lengthwise to produce highly condensed bivalents. Therefore, an important outcome of prophase I is the formation of bivalents. Although the mechanisms may differ, it is universal that homologs are attached until the onset of anaphase I.

Molecular genetic and biochemical studies in yeasts, Caenorhabditis elegans and Drosophila have provided many of the molecular insights into homolog interactions during prophase I (Roeder, 1997; Dernburg et al., 1998; Orr-Weaver, 1999; Zickler and Kleckner, 1999). Such interactions include pairing, synapsis and recombination, which are intimately associated with each other. Meiotic recombination is thought to begin in late leptotene or zygotene stages after homolog pairing has initiated, and continues in the pachytene stage while the homologs are synapsed. In budding yeast, double-stranded DNA breaks (DSBs) are generated by the Spo11p protein and are required for both recombination and SC formation. In C.elegans, DSBs also initiate homologous recombination, which is not required for SC formation (Dernburg et al., 1998). In budding yeast, the recA homologs Dmc1p and Rad51p are required for recombination and SC formation (Bishop et al., 1992; Rockmill et al., 1995; Shinohara et al., 1997).

In addition, sister chromatids are closely associated due to sister chromatid cohesion during prophase I (Klein et al., 1999; Hirano, 2000). In budding yeast, sister chromatid cohesion depends on the cohesin complex, which contains at least four proteins: Smc1p, Smc3p, Scc1p and Scd3p (Michaelis et al., 1997; Nasmyth, 1999). Rec8p, a meiosis-specific Scc1p homolog, from both fission and budding yeasts, is essential for normal meiosis (Molnar et al., 1995; Orr-Weaver, 1999; StoopMyer and Amon, 1999; Watanabe and Nurse, 1999). During meiotic prophase I and metaphase I, sister chromatid cohesion along the chromosomal arms is required to maintain bivalents. The separation of sister chromatids along the arms at the metaphase I–anaphase I transition allows the homologs to separate. Sister chromatid cohesion at the centromere serves to maintain sister association until anaphase II, when the sisters separate. Therefore, both the recombinational crossover and sister chromatid cohesion contribute to the formation of chiasmata. Clearly, regulation of both recombination and sister chromatid cohesion is critical for proper segregation of homologs, although very little is known about such regulation at the molecular level, including the identity of genes that play regulatory roles.

Numerous studies have been conducted to understand the molecular control of the mitotic cell cycle, particularly in animal and fungal systems (Futcher, 1991; Reed, 1991; Pines, 1993; Roberts et al., 1994). Among known mitotic cell cycle regulators, cyclins and cyclin-dependent protein kinases (CDKs) play central roles in controlling all major phases of the cell cycle, such as the G2/M and G1/S transitions (Murray, 1994; Andrews and Measday, 1998; Gitig and Koff, 2000). Cyclins were first identified as proteins showing a cyclical pattern of accumulation and destruction during early embryonic development in marine invertebrates (Evens et al., 1983; Swenson et al., 1986). In budding yeast, genetic analysis with various combinations of mutations in the B-type cyclins CLB1, CLB2, CBL3 and CLB4 indicates that CLB1 and CLB4 are required for meiosis II, but not for meiosis I (Grandin and Reed, 1993; Dahmann and Futcher, 1995). Furthermore, the budding yeast B-type cyclins CLB5 and CLB6 are required for pre-meiotic DNA replication, which seems indirectly to affect the initiation of recombination and SC formation (Dirick et al., 1998; Stuart and Wittenberg, 1998; Smith et al., 2001). Also, mouse cyclin A1 is required for the progression from pachytene to diplotene (Liu et al., 1998). However, whether there is a more direct requirement for cyclins in homolog synapsis and bivalent formation during prophase I is not known.

Plants have been one of the most important sources of our knowledge about meiosis at the cytological level (Dawe, 1998; Zickler and Kleckner, 1999). Mutants with defects in homolog interactions have been described in several plants (Kaul and Murphy, 1985; Dawe, 1998). In particular, a wheat asynaptic mutant seemed to have normal leptotene but unpaired/unattached homologs throughout prophase I (La Cour and Wells, 1970). In addition, the maize desynaptic (dy) mutant was shown to form an SC, but could not fully maintain the SC at late pachytene; bivalents separate into univalents at a variable frequency during diplotene and diakinesis (Nelson and Clary, 1952; Maguire, 1978). However, the molecular nature of these mutant defects is as yet unknown.

Recently, normal male meiosis in Arabidopsis has been described using fluorescence microscopy (Ross et al., 1996; Peirson et al., 1997), paving the way for molecular genetic analysis of Arabidopsis meiosis. For example, the Arabidopsis SYN1 gene (also called DIF1) encoding a protein similar to the yeast Rec8p cohesin subunit is essential for normal meiotic chromosome condensation and homolog interactions (Bai et al., 1999; Bhatt et al., 1999). In addition, disruption of an Arabidopsis SPO11 homolog (AtSPO11-1) causes defects in meiotic recombination, SC formation and bivalent formation (Grelon et al., 2001). Furthermore, recA homologs AtDMC1 and AtRAD51 are also required for meiosis in Arabidopsis (Klimyuk and Jones, 1997; Couteau et al., 1999), and the Rad51 protein has also been implicated in meiosis of maize and lily (Anderson et al., 1997; Franklin and Cande, 1998; Franklin et al., 1999).

We have now isolated a meiotic mutant (solo dancers, or sds) that is severely defective in interactions between homologs in prophase I. Furthermore, we show that the SDS gene is expressed specifically in male and female meiocytes and encodes a protein that is similar to known cyclins. Our results demonstrate for the first time that a meiosis-specific novel cyclin-like protein is critical for normal homolog synapsis and bivalent formation, suggesting that a CDK may play a key role in controlling the normal chromosomal events during meiotic prophase I in Arabidopsis.


Isolation and genetic characterization of a meiotic mutant

In higher plants, male meiosis occurs inside the anther portion of the male reproductive organ, the stamen. To identify genes important for male meiosis, we screened for Ds insertional mutants with reduced fertility (see Materials and methods). Wild-type Arabidopsis plants (Figure 1A) produce large seedpods with dozens of seeds. We found that the progeny of one of the Ds lines segregated for normal and nearly sterile plants that had small seedpods (Figure 1B); furthermore, the normal and mutant plants were in a ratio of ~3:1, suggesting that the parental plant was heterozygous for a recessive mutation. When the mutant plants also expressed the Ac transposase, they could produce fertile sectors due to the excision of the Ds element (Figure 1C). Whereas normal flowers produce many pollen grains (Figure 1D), the mutant flowers produced little or no normal pollen (Figure 1E). Nevertheless, the mutant was able to produce some seeds when pollinated with wild-type pollen, indicating that it was female fertile, although the seed set was much lower than normal. In addition, progeny of a cross from such a pollination were normal, confirming that the mutation was indeed recessive. We have named this new mutant solo dancers (sds) for its meiotic phenotype (see below).

figure cdf285f1
Fig. 1. Wild-type and sds mutant plant, flower and pollen development. (A) A wild-type plant with normal seedpods (arrowheads). (B) An sds mutant plant, with small seedpods (arrows). (C) An sds mutant plant with small seedpods ...

The sds mutant appeared to be normal in vegetative and flower development, producing normal floral organs (Figure 1E; other data not shown). Wild-type Arabidopsis anthers in an unopened floral bud contain normal pollen grains (Figure 1F). However, the sds mutant anthers had many abnormal pollen grains with variable sizes (Figure 1I). The defective pollen grains degenerated before the flower opened (data not shown). Analysis of immature anthers showed that the mutant microspores also had different sizes (Figure 1J), unlike normal microspores (Figure 1G). Further examination revealed that whereas a normal meiosis always produces four microspores of equal size in a tetrad (Figure 1H), the mutant male meiosis produced 2–8 microspores of variable sizes (Figure 1K shows a ‘hexad’ with six spores). From 565 sds meioses, we found six (1.1%) diads, 20 (3.5%) triads, 293 (51.9%) tetrads, 141 (25%) pentads, 80 (14.2%) hexads, 21 (3.7%) heptads and four (0.7%) octads. Therefore, the sds mutant is defective in male meiosis.

The sds mutant is defective in homolog synapsis and bivalent formation in meiotic prophase I

To determine meiotic defect(s) of the sds mutant, we compared wild-type and mutant male meiosis using a chromosome spread and 4′,6-diamino-2-phenylindole dihydrochloride (DAPI) staining procedure (Ross et al., 1996). During leptotene (Figure 2A) (Ross et al., 1996), chromosomes begin to condense and form visible thin lines. Chromosomes continue to condense and begin synapsis at zygotene (Figure 2B); at pachytene (Figure 2C), homologs have completed their synapsis, as indicated by the appearance of thick lines. In late diplotene (Figure 2D), homologs desynapse along much of the chromosome, leaving only limited association at the chiasmata; the chromosomes are also condensed further. At late diakinesis (Figure 2E), chromosomes are very highly condensed and the five bivalents can be easily recognized.

figure cdf285f2
Fig. 2. Male meiosis I in wild-type and the sds mutant. Shown are images of DAPI-stained chromosomes. (A–E) Wild-type prophase I at leptotene, zygotene, pachytene, diplotene and diakinesis, respectively; note that (E) shows five brightly ...

In sds mutant cells, chromosomal patterns similar to the wild-type leptotene through pachytene stages were observed (Figures (Figures2F–H).2F–H). Although it is difficult to determine whether diplotene in the sds mutant (Figure 2I) was normal, it was obvious that by diakinesis the homologs were not attached in sds cells, as indicated by the observation of 10 univalents (Figure 2J). In addition, we noticed that the sds cells with chromosome images similar to that in Figure 2H were very infrequent, suggesting a defect in synapsis. To determine the distribution of prophase I cells, we examined hundreds of wild-type and sds meiotic cells. Because male meiosis in Arabidopsis is slightly asynchronous, a population of male meiotic cells from a single flower can cover a few adjacent meiotic stages or substages of the long prophase I. When data from samples containing only or largely prophase I cells were summarized (Figure 3), we found that among the wild-type prophase I cells (three samples, 717 cells), nearly half were pachytene cells, and relatively few were leptotene cells, with the other stages having moderate numbers of cells. In particular, pachytene was also the most frequent class in each individual sample (data not shown). In contrast, the sds mutant showed a dramatically different distribution for prophase I substages (six samples, 1241 cells); the number of cells at the leptotene and zygotene stages was higher than normal, whereas cells at the pachytene stage were very rare. In this case, pachytene-like cells were the least frequent in each sample. The sds distribution of prophase I stages further supports the idea that synapsis is defective in the mutant.

figure cdf285f3
Fig. 3. Distribution of prophase I cells in wild type and the sds mutant. Chromosome spreads were examined as shown in Figure 2. The number of images at each stage is shown above the appropriate bar. The sds leptotene, zygotene, pachytene and ...

Therefore, the sds mutant had a clear defect in homolog synapsis and bivalent formation in male meiosis, suggesting that normal homolog synapsis could not be achieved in the absence of SDS gene function, and chromosomes form univalents at the end of prophase I. On the other hand, chromosome condensation seems to be unaffected by the sds mutation, unlike the syn1 mutant (Bai et al., 1999; Bhatt et al., 1999). We likened the wild-type chromosome interactions and condensation to a highly choreographed duet dance in which homologs ‘dance’ in pairs. Therefore, the mutant had a clear phenotype of homologs behaving as ‘solo dancers’.

Next, we compared wild-type and sds meiosis from metaphase I to telophase I for possible additional mutant defects. In wild-type cells at metaphase I (Figure 2K and L), the five bivalents align at the equatorial plane. At anaphase I, homologs separate (Figure 2M), and move towards the opposite pole of the spindle (Figure 2N). In contrast, the univalents in the sds mutant did not align completely; some of them were quite far from the equator (Figure 2O and P). The abnormal distribution seemed to persist (Figure 2Q and R). It seemed that usually more than two clusters of chromosomes were formed. We examined chromosome distribution at late meiosis I. For wild-type cells at anaphase I, 62 out of 66 cells (93.9%) showed a 5:5 even distribution of chromosomes; the other four cells had <10 chromosomal DAPI spots, most probably due to a superimposition of some chromosomes. On the other hand, sds male meiotic cells exhibited many abnormal distribution patterns. Among 71 sds cells at anaphase I, 46 cells (64.8%) had two groups of chromosomes with the following distributions: 5:5 (nine cells), 6:4 (16 cells), 7:3 (14 cells), 8:2 (six cells) and 9:1 (one cell). The remaining cells had one (14 cells; 19.7%), two (eight cells; 11.3%) or three (three cells; 4.2%) chromosomes at the equator, with various distributions of other chromosomes (data not shown) on either side of the equator.

Because the sds mutant also had reduced female fertility, we examined female meiosis in the wild-type and the sds mutant. As shown in Figure 4, wild-type prophase I at diakinesis displays five bivalents (Figure 4A), which become highly condensed at metaphase I (Figure 4B). The homologs separate at anaphase I, showing two groups with five chromosomes each (Figure 4C). In contrast, female meiosis in the sds mutant at diakinesis showed a variable number of univalents. Among 48 nuclei observed, 23 (47.9%) had only univalents (e.g. Figure 4D), similar to the sds male meiosis. Seven (14.6%) and eight (16.7%) nuclei had one (Figure 4G) or two bivalents (data not shown), respectively, with additional chromosomes forming univalents. Only a very small number of nuclei had four (2/48, or 4.2%) or five (2/48) bivalents (data not shown). These 8% of female meioses probably accounted for most of the female fertility when normal pollen was used. At metaphase I in the sds mutant, as expected, many female meioses had only univalents (Figure 4E) and others had some bivalents (Figure 4H). At anaphase I, one or more pairs of chromosomes were often observed, suggesting that they were homologs that had been associated in bivalents (Figure 4F and I). Other anaphase I images are consistent with an absence of bivalents (data not shown). These results indicate that in the sds mutant, female meiosis is defective in a way similar to male meiosis, but to a lesser extent.

figure cdf285f4
Fig. 4. Female meiosis I in wild type and the sds mutant. Shown here are DAPI-stained images of chromosome spreads from the wild type (A–C) or the sds mutant (D–I). (A) A diakinesis image showing five bivalents. (B) At ...

During meiosis II in the wild-type male meiocytes (Figure 5), the two groups of chromosomes are separated by a band of organelles (arrows in Figure 5A–D). Partially condensed chromosomes were observed during prophase II (Figure 5A); at metaphase II, they congressed at the two equators (Figure 5B). Sister chromatids then separated at anaphase II (Figure 5C), moved to opposite poles at telophase II (Figure 5D) and decondensed to form four nuclei (Figure 5E). In the sds mutant, chromosome distribution was abnormal, most probably due to a defective meiosis I, but chromosome condensation and sister chromatid separation seemed to be quite normal. Chromosomes condensed at prophase II (Figure 5F and K), but alignment was abnormal (Figure 5G and L). Then sister chromatids separated at anaphase II (Figure 5H and M), and moved apart at telophase II (Figure 5I and N). Distributions of meiotic cells from metaphase I to the end of meiosis II were not dramatically different between the wild type and the sds mutant (data not shown).

figure cdf285f5
Fig. 5. Male meiosis II in wild type and the sds mutant. (A–O) Images obtained from chromosome spreads. (A–D) Wild-type meiosis II at prophase, metaphase, anaphase and telophase, respectively. The arrows in these panels ...

Our results indicate that the sds mutant is defective in homolog synapsis and bivalent formation in prophase I. In addition, there was no obvious defect during later stages of meiosis I or during meiosis II other than those that are probable consequences of the failure to form bivalents. Furthermore, clear separation of sister chromatids at meiosis II indicates that meiotic DNA replication was not obviously affected in the mutant. Therefore, the sds mutant is specifically defective in homolog interaction during prophase I.

The sds mutant is defective in meiotic recombination

The phenotypes of the sds mutant were similar to those of the recombination mutants Atspo11 and Atdmc1, suggesting that sds might also be defective in recombination. To test such a defect in the sds mutant, we generated sds/sds plants that were heterozygous (L1L2/C1C2) for two molecular markers on each of chromosomes II and V (see Materials and methods). Although the sds mutant was nearly sterile, it was able to produce 50–100 seeds per plant (a normal plant can easily produce several thousand seeds). When the seeds from sds plants were planted, many of them developed to maturity and exhibited the same sds defects in fertility and meiosis (data not shown), indicating that the seeds from sds plants were not from cross-pollination. The progeny of sds/sds plants allowed the analysis of recombination frequency in the sds mutant. The sds/+ or +/+ siblings were also heterozygous for the same molecular markers and served as positive controls for recombination.

The progeny of both sds and normal plants (L1L2/C1C2) were analyzed for genotypes of these markers. As shown in Table I, for the markers nga1126 and AthBIO26 on chromosome II, the control plants had 17.0% recombinants, whereas the sds plants had no detectable recombinants. Similarly, for the markers nga76 and nga106 on chromosome V, the control plants showed 16.5% recombinants, in contrast to the 4.5% recombinants among sds plants. Therefore, meiotic recombination is greatly reduced in the sds mutant. Our results were obtained from viable seedlings, which resulted from viable pollen and embryos.

Table I.
Meiotic recombination in normal and sds mutant plants

The progeny of the sds mutant are often aneuploid

Because sds mutant meiosis appeared to have random chromosome segregation at meiosis I, we suspected that seeds from sds plants might carry an abnormal number of chromosomes, which could cause visible developmental defects. When the seeds from sds plants were germinated on plates and examined under the dissecting microscope, we found that a considerable fraction of seedlings (53 out of a total of 124) showed abnormal development (Figure 6). Unlike normal seedlings with two similarly sized cotyledons (Figure 6A), seedlings from the seeds of sds mutant plants had three cotyledons (two seedlings, see Figure 6B and C), two uneven cotyledons (six, Figure 6D), one cotyledon (four, Figure 6E), one cotyledon and no hypocotyl (one, Figure 6F) or had root defects (40, data not shown). The seedling defects were similar to those of the Atdmc1 mutant seedlings (Couteau et al., 1999). Because sds seedlings from sds/+ heterozygous plants never showed any developmental defects prior to meiosis, the abnormal development of progeny of sds plants was unlikely to be caused directly by the sds mutation.

figure cdf285f6
Fig. 6. Wild-type and sds seedlings. (A) A wild-type seedling with two evenly sized cotyledons. (BF) Abnormal seedlings from seeds of the sds mutant, with three cotyledons (B and C), two unequal cotyledons (D), one cotyledon (E), ...

To test directly the hypothesis that the abnormal seedlings were aneuploid, we then determined the number of chromosomes in wild-type seedlings and the progeny of sds plants. As expected, all 13 wild-type seedlings examined had 10 chromosomes. On the other hand, when 13 normal-appearing progeny of sds were examined, one had 11 chromosomes and another had nine chromosomes, whereas the others had 10. Even more strikingly, when 13 sds seedlings with abnormal development were tested, only one had 10 chromosomes; the remaining 12 seedlings had the following numbers of chromosomes: nine (four seedlings), 11 (two seedlings), 12 (one seedling), 13 (three seedlings) and 15 (two seedlings). Therefore, progeny of sds plants were often aneuploid, presumably due to abnormal chromosome segregation at meiosis in sds plants.

The SDS gene encodes a cyclin-like protein

The Ds element in sds carries an NPT (neomycin phosphotransferase) gene conferring kanamycin resistance (Sundaresan et al., 1995), which co-segregated with the sds mutant phenotype. If the Ds element was indeed inserted into the SDS gene, excision of the Ds element would produce normal-appearing revertant sectors. We screened among sds mutant plants carrying an Ac element and found that nine of 13 sds plants produced phenotypically normal revertant sectors. The flowers of one large sector (Figure 1C) produced normal pollen (data not shown); this ruled out the possibility that the seeds of the revertant sector were due to contaminating pollen. Seeds of three revertant sectors were planted, and each segregated for mutant plants; therefore, the fertile plant sectors were heterozygous, consistent with an excision of the Ds element from one of the two alleles. These results strongly support the idea that the sds mutation was caused by a Ds insertion.

The genomic sequences flanking the Ds element were obtained using thermal asymmetric inter-laced PCR (TAIL-PCR; Liu et al., 1995) and shared the same 8 bp immediately adjacent to the Ds sequence, consistent with the 8 bp duplication characteristic of Ds insertions (Figure 7A). Using primers matching genomic sequences flanking the Ds element, we amplified a PCR fragment of the expected size from wild-type and revertant genomic DNAs, but not from mutant DNAs (data not shown). We analyzed the sequences of the PCR fragment from seven revertants, and found that four had wild-type sequences, while the other three had 6 or 9 bp insertions (Ds excision footprints) which did not disrupt the open reading frame (Figure 7A). These results confirm that we have identified the genomic insertion that is responsible for the sds mutant phenotype, and that the insertion most probably disrupts a protein-coding region.

figure cdf285f7
Fig. 7. The SDS locus and gene structure. (A) The DNA and predicted animo acid sequences of the SDS locus near the Ds insertional site. The sds mutant sequence shows an 8 bp duplication (underlined). The revertants have 9 or 6 bp ...

Using the SDS genomic sequence as a probe, we isolated a 2.1 kb cDNA clone after screening ~1 × 106 plaques of a floral cDNA library (Fan et al., 1997). The SDS sequence also matches a sequenced bacterial artificial chromosome (BAC) clone (F10B6; accession No. AC006917) ~20 cM from the distal end of the left arm of chromosome 1. Comparison between the SDS genomic and cDNA sequences indicates that there are seven exons and six introns (Figure 7B). The exons are 1117, 168, 167, 129, 181, 45 and 294 bp, and the introns are 189, 249, 297, 88, 93 and 96 bp. It should be noted that the SDS cDNA sequence differs from the exon sequence of the predicted gene (gene ID At1g14750) from the genomic sequence in two places: the actual introns 2 and 6 were longer at the 3′ and 5′ ends by 51 and 57 nucleotides, respectively. The additional 17 and 19 amino acid residues, respectively, encoded by the additional predicted exon sequences would have interrupted the cyclin core of the SDS protein (see below).

The predicted SDS protein contains 578 amino acids (Figure 8A). The C-terminal one-third of SDS is similar to the cyclin core region of several known Arabidopsis cyclins (Figure (Figure8A–C).8A–C). Like other cyclins, its N-terminal region contains a putative ‘destruction box’ motif with a conserved element and several lysine residues (Figure 8A; Hunt, 1991). Most of the SDS N-terminal region consists of several putative PEST elements, rich in proline (P), glutamate (E) or aspartate (D), serine (S) and threonine (T) residues (Rogers et al., 1986) (Figure 8A). These PEST elements may play a role in targeting SDS for proteolytic degradation. In addition, the SDS protein contains several potential sites for phosphorylation by CDKs.

figure cdf285f8
Fig. 8. Sequence and comparison of the SDS protein. (A) The predicted SDS amino acid sequence (accession No. AJ457977). A putative destruction ...

Some cyclins, such as the yeast G1 cyclins Cln1, -2 and -3, have the cyclin core region near their N-termini (Nash et al., 1988; Hadwiger et al., 1989). In plants, previously known cyclins have been classified based on amino acid sequence similarity in the cyclin core into three types, A, B and D, which are grouped further into subtypes (Renaudin et al., 1996). Among these known plant cyclins, D-type cyclins also contain a cyclin core near their N-termini (Figure 9A). In contrast, in plant A- and B-type cyclins, the cyclin core is located near the C-terminus (Figure 9A). Therefore, the location of the cyclin core in the SDS protein is similar to those of plant A- and B-type cyclins, although the SDS N-terminal region is unusually large. Within the cyclin core, members of the same type share ≥49% amino acid sequence similarity and those of the same subtype can share 68–97% identity; however, levels of amino acid sequence identity between different types are only 20–39% (Figure 8B and C). Because the putative SDS cyclin core has 23–32% identity to known cyclins, we hypothesize that SDS is a new type of cyclin.

figure cdf285f9
Fig. 9. Analysis of the SDS protein. (A) Regions in SDS and several known Arabidopsis cyclins. The regions containing putative PEST elements and the cyclin core homologous regions are shaded as indicated. The numbers indicate the beginning and ...

As a first test for the hypothesis that SDS is a new type of cyclin, we investigated whether SDS can interact with known Arabidopsis CDKs Cdc2a and Cdc2b (Hirayama et al., 1991), using the yeast two-hybrid assay (Fields and Song, 1989). A portion of the SDS coding region containing the cyclin homology was fused with the coding region for the GAL4 DNA-binding domain to yield the bait construct (Figure 9A). The Cdc2a and Cdc2b coding regions were each fused with that for the GAL4 transcriptional activation domain to produce the prey constructs. The bait and prey constructs were introduced into α or a mating-type yeast strains, respectively, and were brought together by yeast mating. The yeast cells carry a HIS3 reporter gene that is transcriptionally controlled by GAL4 and that allows cells to grow on media lacking histidine. If a bait and a prey interact to a sufficient level, then the activation of the HIS3 reporter gene will support yeast growth without histidine. As shown in Figure 9B, SDS was able to allow yeast cells to grow without histidine in the presence of Cdc2a or Cdc2b, supporting the hypothesis that SDS can interact with these CDKs.

The SDS gene is expressed specifically in meiotic cells

To obtain additional clues about how SDS functions, we analyzed the expression of the SDS gene. The SDS cDNA clone was very rare; in addition, Northern hybridizations detected only very faint signals in floral bud RNA samples (data not shown). Because the sds mutant had only fertility defects, we first examined SDS expression in developing wild-type Arabidopsis flowers, particularly at stages near meiosis, by RNA in situ hybridization (Figure 10). SDS expression was not detectable in the inflorescence meristem or in very early floral primordia (data not shown). Similarly, no SDS expression was found in a stage 8 developing flower (Figure 10A, as defined previously by Smyth et al., 1990), when anther differentiation had begun but meiosis had not occurred. At about stage 9 (Figure 10B), or the time of male meiosis, we observed a strong SDS signal that was restricted to the male meiocytes (arrows) within the anther locules, but not in other tissues of the anther or in other floral organs. At stage 11 (Figure 10C), when pollen development was progressing, again there was no detectable SDS expression. A sense probe control produced only background levels of signals (data not shown). Therefore, our results strongly suggest that in the anther, SDS is expressed specifically in male meiotic cells.

figure cdf285f10
Fig. 10. The SDS gene expression pattern. (A–C) In situ RNA hybridization results from a radioactive probe. The left panel in each is a bright field image showing the tissues, the center panel is a dark field image, and the right panel is a composite ...

Because SDS was also needed for normal female meiosis, we also tested its expression in female meiocytes. Unlike male meiocytes which are in groups, individual female meiocytes are surrounded by somatic cells. Because RNA in situ hybridization using radiolabeled probes does not have high enough resolution to pinpoint accurately expression in a single cell, we switched to using non-radioactive probes for detecting SDS expression in the female meiocyte. As shown in Figure 10D, the female meiocyte (arrow) clearly expressed SDS, whereas control sense probe (Figure 10E) showed only background. The somatic cells in the ovule did not have SDS signal (Figure 10D); further, other tissues of the female reproductive organ also lacked SDS signals (data not shown). Therefore, in the female reproductive organ, SDS is expressed specifically in the female meiocyte.

To test whether SDS was expressed in other organs at low levels, we performed RT–PCR experiments with RNA samples isolated from roots, leaves, floral stems, young inflorescences, older floral buds, open flowers and young fruits (seedpods). Our results (Figure 10F, top panel) indicate that SDS was expressed preferentially in young floral buds, but not in older buds, mature flowers or vegetative organs. As a negative control, we found that SDS expression was not detectable in floral buds and flowers of the sds mutant (Figure 10F). Moreover, as a positive control for mRNA and RT–PCR, we tested for the presence of mRNA from the non-specific ASK1 gene (Yang et al., 1999) in the same RNA samples; our results confirmed that ASK1 mRNA was present in all organs tested (Figure 10F, bottom panel). Therefore, our results strongly suggest that SDS is expressed in a highly specific manner, both spatially and temporally.


The Arabidopsis SDS gene is required for homolog synapsis and bivalent formation

We have described here the isolation of solo dancers (sds), a new mutant in Arabidopsis that exhibits clear defects in male and female meioses, resulting in greatly reduced male fertility and partial female fertility, respectively. Although the sds mutant male meiocytes displayed normal chromosome condensation during prophase I, there were many more zygotene cells and very few pachytene cells. One possibility is that in the sds mutant normal synapsis of homologs cannot be achieved for most cells. The observed sds zygotene stage cells could include the cells that had partial or no synapsis; further analysis is needed to determine whether there is homolog pairing in sds cells. The observations of individual univalents in all sds cells at late prophase I and early metaphase I, and of separated arms of sister chromatids for at least some chromosomes at diakinesis (Figure 2J) indicate that sds was defective in homolog synapsis and had a severe limitation in the formation and/or maintenance of bivalents.

In female meiosis of the sds mutant, many univalents were present at late prophase I and metaphase I, although one or two bivalents were often observed and four or five bivalents occasionally were found. Therefore, female meiosis was partially functional, providing an explanation for a higher female fertility than male fertility. More mild female meiosis and fertility phenotypes than male phenotypes have also been observed in several maize meiotic mutants (I.Golubovskaya and W.Zacheus Cande, personal communication). It is possible that the SDS function is not essential for female meiosis because one or more other cyclins (Renaudin et al., 1996) might fulfill the SDS function partially. Another explanation for differences in female and male meiotic phenotypes is that female and male meioses may be regulated differently.

We observed that progeny of the sds mutant often exhibit abnormal seedling development. Because such abnormal seedlings were not observed among progeny of SDS/sds plants, the abnormal development was probably a consequence of abnormal chromosome number, resulting from the meiotic defect of the sds parent. Indeed, analysis of chromosome number in such abnormal seedlings revealed that their chromosome numbers ranged from nine to 15. It is interesting to note that abnormal embryo and seedling development was also observed when a dominant-negative form of the CDK Cdc2a was expressed in the embryo (Hemerly et al., 2000). The abnormal seedling patterning supports the idea that normal cell division is important for pattern formation during embryo development. Although the molecular basis for the abnormal development of the sds mutant progeny is not known, it is possible that an imbalance of gene function resulting from the aberrant chromosome number caused a disruption of normal cell division.

The SDS gene encodes a novel meiosis-specific cyclin-like protein

We have isolated the SDS gene and found that the predicted SDS protein contains a C-terminal domain that resembles known cyclin proteins from Arabidopsis and other plants. Previously identified plant cyclins can be divided into three types: A-, B- and D-type cyclins (Renaudin et al., 1996) (Figure 8). Each type of plant cyclin shares some characteristic sequence motifs with the corresponding type from animals, although the overall levels of amino acid sequence identity in the cyclin core region are not high between corresponding types of animal and plant cyclins (Renaudin et al., 1996). In addition, the levels of amino acid sequence identity are high between members of the same type, ranging from 49 to 84% (Renaudin et al., 1996) (Figure 8C). The SDS cyclin-like domain is most similar to the B2 subtype of Arabidopsis B cyclins (31% identity) and slightly less similar to the B1 subtype and A-type cyclins (24–30% identity). The degrees of amino acid identity and similarity between SDS and other plant cyclins is within the range of those between A-, B- and D-type cyclins (20–39%) (Renaudin et al., 1996) (Figure 8C). Therefore, SDS is distinct from all known cyclin types and may represent a novel type of cyclin. Furthermore, the finding that SDS can interact with the Arabidopsis CDKs Cdc2a and Cdc2b (Hirayama et al., 1991) in a yeast two-hybrid assay further supports the hypothesis that SDS is a cyclin. It will be interesting to determine whether one or both of the CDKs tested are involved in regulating meiosis.

The SDS protein is unusual in having an extra long N-terminal region. Many cyclins have a non-conserved region in addition to the cyclin core region. In plants, the A- and B-type cyclins have an N-terminal non-conserved region that make up about half of the protein, whereas D-type cyclins have a small C-terminal non-conserved region (Figure 9A). The SDS N-terminal region is rich in hydrophilic residues, particularly in putative PEST elements (Rogers et al., 1986). PEST elements are found in cyclins (Nash et al., 1988; Salama et al., 1994) (Figure 9A) and other unstable proteins (Chevaillier, 1993; Gorl et al., 2001). It was found that the truncated yeast Cln3 proteins with larger deletions in the PEST region had greater stability (Tyers et al., 1992), indicating that the size of the PEST region is correlated with the degree of instability. Therefore, the unusually large PEST region of SDS suggests that the SDS protein might be very unstable.

Our in situ RNA hybridization results indicate that SDS is expressed in male and female meiocytes, whereas its expression was not detected in other cells and at other stages. Furthermore, RT–PCR experiments also showed that the only detectable expression was found in flowers containing meiotic cells, and not in vegetative organs or other stages of reproductive structure. These results strongly suggest that SDS is a meiosis-specific gene. Molecular genetic studies in yeast indicate that yeast cyclins are all expressed during mitosis. Even the CLB proteins that play important roles in meiosis are not meiosis specific (Grandin and Reed, 1993; Dahmann and Futcher, 1995; Dirick et al., 1998; Stuart and Wittenberg, 1998; Smith et al., 2001). Therefore, our results, including mutant phenotype, protein sequence and expression pattern, argue strongly that SDS may be a novel cyclin that regulates aspects of meiotic chromosome behavior in prophase I.

Meiosis can be considered as a modified cell cycle (Wolgemuth et al., 1995; Zickler and Kleckner, 1999). Similarly to mitosis, meiosis is also preceded by DNA replication and involves chromosome condensation and sister chromatid segregation. Thus, it is reasonable to expect that some cyclin functions would be conserved between mitosis and meiosis. For example, the yeast B-type cyclins CLB5 and CLB6 are required for the meiotic S phase (Smith et al., 2001), as they are for mitotic S phase (Epstein and Cross, 1992; Schwob and Nasmyth, 1993). Similarly, the mouse cyclin A2 is highly expressed in pre-leptotene spermatocytes, suggesting that it may be involved in controlling meiotic DNA replication (Ravnik and Wolgemuth, 1999). In addition, it was shown that the yeast CLB1 and CLB4 proteins, which are redundant with CLB2 and CLB3 for mitosis, are required for meiosis II but not meiosis I (Grandin and Reed, 1993; Dahmann and Futcher, 1995).

However, meiosis differs from the mitotic cell cycle in that it has an additional extended prophase I during which homologs undergo pairing, synapsis and recombination; this is followed by metaphase I through telophase I to allow homologs to segregate in a reductional division. To accommodate two divisions, sister chromatid cohesion removal is biphasic, first along the arms to allow homologs to separate, then at the centromere to allow sister separation (Orr-Weaver, 1999). Therefore, coordination of events in meiosis I would be likely to involve regulatory proteins that are functionally distinct from mitotic cell cycle regulators. Studies with the mouse cyclin A1 protein have provided strong support for meiosis-specific cyclins (Liu et al., 1998, 2000). Mouse cyclin A1 protein is present just before or during meiosis I, but not in meiosis II (Ravnik and Wolgemuth, 1999). Furthermore, in mutant mouse with a disrupted gene for cyclin A1, male meiosis arrests before metaphase I (Liu et al., 1998). Our results indicate that SDS also functions as a meiosis-specific regulatory gene, as discussed further in the next section.

Possible roles of SDS in regulating synapsis, recombination and/or sister chromatid cohesion

The sds mutant is defective in homolog synapsis and bivalent formation. It is possible that during meiosis in the sds mutant, chromosomes can find their partners and begin pairing, but the synapsis could not be achieved. Such an incomplete or abnormal association might be unstable or indistinguishable from the zygotene stage at the light microscope level. Alternatively, chromosome pairing and SC formation might not occur in the sds mutant. Because the SDS protein is similar to cyclins, SDS might activate one or more CDK(s) during early to mid-prophase of meiosis I. The mammalian SCP1 protein, which is a probable component of the SC transverse filaments, has a potential target site for CDK that is thought to play a role in SC assembly and/or disassembly (Meuwissen et al., 1992). In addition, the mammalian Cdk2 protein was found to localize to newly synapsed axes and to disappear by mid-pachytene stage (Ashley et al., 2001). Therefore, it is possible that SDS-dependent CDK(s) might regulate the activity of pairing and/or synapsis proteins.

In animals and yeast, mutants defective in recombination exhibit homolog separation prematurely before metaphase I (Dernburg et al., 1998; Woods et al., 1999). In Saccharomyces cerevisiae, meiotic recombination is initiated by DNA DSBs, the formation of which requires several genes, including SPO11 (Smith and Nicolas, 1998; Paques and Haber, 1999). DSBs are also required for SC formation, so mutations affecting DSBs also cause defects in synapsis. Arabidopsis male meiosis undergoes recombination once or twice per chromosome, suggesting that recombination is also needed for normal meiosis in Arabidopsis (Copenhaver et al., 1998). In addition, mutations in Arabidopsis AtSPO11and AtDMC1 genes cause defects in meiotic recombination, and SC and bivalent formation (Couteau et al., 1999; Grelon et al., 2001), similar to the sds defects. Furthermore, we showed here that the sds mutant had greatly reduced frequency of meiotic recombination. Because meioses that lacked any recombinational crossover probably resulted in a greater number of non-viable pollen and embryos, our results were probably an overestimate of the true recombination frequency in the sds mutant. It is possible that the sds defect in meiotic recombination and that in synapsis may be caused by the same molecular defect. Alternatively, recombination and synapsis may each be regulated by SDS via distinct mechanisms.

The S.cerevisiae B-type cyclins CLB5 and CLB6 were found to control recombination initiation and SC formation (Smith et al., 2001). However, the fact that CLB5 and CLB6 are required for meiotic DNA replication suggests that the defects of Clb5 Clb6 cells in recombination and SC formation might be a consequence of impaired replication. In contrast, meiotic DNA replication in the sds mutant appears to be normal. In mouse, cyclin A1 may play a critical role in the activation of the M-phase-promoting factor (CDK) during mouse spermatogenesis prior to metaphase I (Liu et al., 2000). Unlike the sds mutant, in the mouse mutant defective in cyclin A1, male meiocytes can form synapsed chromosomes at pachytene (Liu et al., 1998). Therefore, the proposed SDS functions in regulating synapsis and recombination might be distinct from those of the mouse cyclin A1 and the yeast CLB5 and CLB6. The idea that cyclin–CDK may regulate meiotic recombination is also supported by the recent observation that the mammalian Cdk2 protein co-localizes with the mismatch repair protein MLH1 at sites of reciprocal recombination (Ashley et al., 2001).

The Atspo11 mutant can form some bivalents (Grelon et al., 2001), suggesting that the Atspo11 mutant is not totally defective in recombination due to gene redundancy (Grelon et al., 2001). The fact that sds male meiosis is more defective than the Atspo11 mutant in bivalent formation could be explained if the sds mutation could possibly affect another aspect of prophase I, such as sister chromatid cohesion, which is also required for bivalent formation. The observations that the meiotic recombination in the sds mutant was not zero and that there were a small number of pachytene-like images support the idea that homolog synapsis occurred at a very low frequency in sds. Therefore, during sds male meiosis, the failure of the infrequent homolog interactions observed at the pachytene stage to be maintained as bivalents at the diakinesis stage could be explained if sister chromatids separated pre-maturely along the arms, either at or before diakinesis; this idea is supported by the observation that some univalents at diakinesis had ‘X’ or ‘Y’ shapes. There are several lines of evidence for the idea that sister chromatid condensation and cohesion are regulated by protein phosphorylation (Wei et al., 1998; Suja et al., 1999; Wei et al., 1999; Kaszas and Cande, 2000). Therefore, if SDS is indeed a meiosis-specific cyclin, one possible mechanism for its function could be to regulate the phosphorylation of proteins involved in sister chromatid cohesion.

Recently, the Arabidopsis SWITCH gene was shown to be required for sister chromatid cohesion during meiosis (Mercier et al., 2001). In early male meiosis of the swi1.2 mutant, homologs fail to synapse and sister chromatids separate during prophase I; consequently, 20 individual chromatids, instead of five bivalents, were observed at metaphase I, followed by abnormal chromosome segregation. In swi1.2 female meiosis, chromosome segregation resembles that in mitosis, suggesting that sister chromatid cohesion is maintained until anaphase I. Therefore, SWI1 may play related but distinct roles in male and female meioses; alternatively, female defects in swi1.2 may be less severe than that in male meiosis. SWI1 encodes a novel protein, so it is not clear what its biochemical function might be (Mercier et al., 2001). Because chromosomes do condense in the swi1.2 mutant and cohesins also play a role in chromosome condensation (Bai et al., 1999; Bhatt et al., 1999), the SWI1 protein may specifically regulate sister cohesion but not condensation. If, as suggested here, SDS plays a role in controlling the timing of the separation of sister chromatids, SDS might regulate the activity of SWI1.

Our results presented here strongly support the idea that we have identified a meiosis-specific gene that is required for normal homolog synapsis and recombination in early to mid-prophase I of Arabidopsis meiosis. Furthermore, the remaining low frequency of recombination and pachytene-like images, combined with the lack of bivalents at late prophase I and metaphase I, support the idea that SDS may regulate the timing of sister chromatid separation. The SDS sequence and yeast two-hybrid results support the hypothesis that SDS is a new type of cyclin, suggesting that it may regulate recombination and sister chromatid cohesion via phosphorylation of proteins involved in these processes. Further analysis is necessary to test these possibilities and to identify the target proteins of SDS. SDS provides a new and important entry point into understanding the regulation of homolog synapsis, recombination and bivalent formation in plant meiosis using genetic, cytological, genomic and biochemical approaches.

Materials and methods

Plant materials, mutant isolation and genetic analyses

Both the wild-type and the sds mutant plants were of the Landsberg erecta ecotype and were grown under long day conditions at 22–25°C unless otherwise indicated. Ds insertional lines were generated following crosses between homozygous Ds-carrying plants and homozygous Ac-carrying plants (Sundaresan et al., 1995; Yang et al., 1999). One Ds line was found to segregate for mutant plants that showed a severe reduction in fertility and was chosen for further analysis. F2 progeny from crosses by pollinating sds pistils with normal pollen were grown to maturity and scored for fertility defects. The seeds of sds plants were germinated on MS plates and the developmental phenotypes of the seedlings were examined after 8 days. Some of these seedlings were grown to maturity to determine fertility.

To screen for revertant sectors, F2 progeny of the original Ds × Ac cross were grown to maturity and sds mutant plants were retained. A majority (~3/4) of these sds plants should also carry the Ac element, and Ds excision would produce a revertant sector. Because stable sds mutant plants never produced large seedpods with a full seed set, putative revertant sectors were identified as either a normal branch with multiple large seedpods or a single large seedpod. Seeds from these putative revertant sectors were planted and grown to maturity. DNAs were isolated from the phenotypically normal plants for further analysis. For one large sector of a branch, flowers were examined using a dissecting microscope and were found to have normal pollen.

Phenotypic characterization

Fresh plant samples were either photographed and the photographs then scanned to yield digital images or were photographed using a dissecting microscope and an Optronics digital camera. Anthers prior to dehiscence were stained (Alexander, 1969) to distinguish viable and dead pollen grains. Developing microspores and meiotic products from fresh floral buds were stained using an aqueous solution of toluidine blue (0.05%). Chromosome spreads were prepared as described by Ross et al. (1996) and stained with 5 µl of DAPI (1 µg/ml). Digital images were obtained by using the Optronics camera mounted on a Nikon fluorescence microscope. Female meiosis was analyzed essentially according to Armstrong and Jones (2001), using floral buds at stages 10–11 (Smyth et al., 1990). Karyotyping was performed on mitotic images essentially as described previously (Maluszynska and Heslop-Harrison, 1991), using root tips from 4-day-old seedlings.

Recombinational analysis

For recombinational studies, plants were generated that had the appropriate genotypes with respect to both the SDS gene and molecular markers for detecting recombination. First sds/+ (Ler) was crossed with the Columbia (Col) ecotype to introduce the sds mutation into a different ecotype so that we could use molecular markers to detect recombinational events. We identified the desired F1 plants from such crosses using kanamycin resistance and screened the F2 generation by PCR for sds/+ plants that were homologous for the Col allele at both of two microsatellite markers (Bell and Ecker, 1994) on either chromosome II or V. The chromosome II markers were nga1126 and AthBIO26, and the chromosome V markers were nga76 and nga106 (http://genome.salk.edu/SSLP_info/SSLPsordered.html). These F2 plants were then crossed with our original sds/+ (Ler) plants to yield both sds/sds mutant plants (experimental) and sds/+ or +/+ normal plants (control); all of these plants were heterozygous (L1L2/C1C2) for two molecular markers on either chromosome II or V. The experimental and control plants were grown to maturity. Their seeds were then germinated on MS plates and DNAs were isolated from the seedlings and used for genotyping by PCR using the primers for the two markers on the relevant chromosome.

DNA and RNA analyses

Genomic DNA was isolated as described previously (Huang et al., 1994). To isolate the SDS gene, we obtained ~0.8 and 0.6 kb genomic fragments (data not shown) from the 5′ and 3′ ends of the Ds element, respectively, using the TAIL-PCR procedure (Liu et al., 1995). TAIL-PCRs were performed using Ds-specific primers (Ds3-1, Ds3-2, Ds-3-3, D5-1, Ds5-2 and Ds5-3; Grossniklaus et al., 1998) and arbitrary degenerate primers (AD2 and AD4; Liu et al., 1995). The SDS-specific primers were designed based on the sequence information of the TAIL-PCR products: oMC257, 5′-CGAAATTTCGAATCACCAAT-3′; oMC258, 5′-TAGTTTCAAGCTTTCGTACG-3′; oMC273, 5′-CGCAGATGCATAAATTCAAACG-3′; and oMC274, 5′-TGGGAATCGTGGATCTACAG-3′. To isolate the SDS cDNA, a probe was generated by PCR using the oMC257 and oMC273 primers and Arabidopsis genomic DNA and used to hybridize with a floral cDNA library (Fan et al., 1997). Approximately 1 × 106 plaques were screened and one positive clone was obtained that contained a 2.1 kb insert; it was designated as pMC2311.

RT–PCR was performed using 10 µg of total RNAs from roots, stems, leaves, open flowers, mature buds (larger than ~1.5 mm), young buds (<1 mm), sds open flowers, sds mature buds and sds young buds. The RNAs were treated with RNase-free DNase I (Roche) followed by inactivation of the DNase I. The RNAs were then used for reverse transcription and the resulting cDNAs were used as template for PCRs. The SDS gene-specific primers used for the RT–PCRs were oMC299 (5′-GGAGCTTGAGATAGTCGGAT-3′) and oMC310 (5′-GTATGAGCGACACAAATGGTGGCAC-3′). As a control, the same cDNAs were used to amplify the ASK1 gene using the oMC221 and oMC222 primers (Yang et al., 1999). The PCR products were hybridized with SDS-specific or ASK1-specific probes, respectively.

In situ RNA hybridization experiments were performed as described previously (Drews et al., 1991) using probes generated as follows. A 619 bp EcoRI–XhoI fragment from the SDS cDNA in pMC2311 was subcloned into pGEM7Zf(+), and the resultant clone was named as pMC2317. pMC2317 was digested with XhoI or EcoRI and used for transcription with SP6 or T7 RNA polymerases, respectively, to generate sense or antisense RNA probes. For expression in the female reproductive organ, non-radioactive digioxiginin-labeled probes were synthesized from the same templates as for the radioactive probes, using a Riboprobe transcription system (Promega, Madison, WI) in vitro RNA synthesis kit.

Yeast two-hybrid assay

The yeast two-hybrid assay (Fields and Song, 1989) was performed to test for interaction between SDS and Arabidopsis CDKs Cdc2a and Cdc2b (Hirayama et al., 1991). A plasmid construct expressing a fusion of the SDS protein with the GAL4 DNA-binding domain (bait, Figure 9A) and fusions of Cdc2a or Cdc2b proteins with the GAL4 transcriptional activation domain (prey) were created by homologous recombination in yeast (Ma et al., 1987). Briefly, to create the expected construct, yeast cells were transformed with a linearized vector plasmid and a DNA fragment with the desired insert and ends that were homologous to the vector. The linearized plasmid was derived from the pAS1-CYH2 vector (Durfee et al., 1993) by SmaI and NcoI digestion, and the insert DNA fragments were obtained by PCR using cDNA templates and gene-specific primers. The SDS bait construct contains a portion of SDS with the cyclin core homologous region, using primers oMC544 (5′-TACGCTAGCTTGGGTGGTCATATGGCCATGGAGCTTGAGATAGTCGGATGC-3′) and oMC545 (5′-ATTAGCTTGGCTGCAGGTCGACGG ATCCCCTATTCTTTTTAAGAATATGAG-3′). For a positive control, the Arabidopsis cyclin cycB1.1 coding region (Hemerly et al., 1992; Renaudin et al., 1996) was amplified using primers oMC546 (5′-TACGCTAGCTTGGGTGGTCATATGGCCATGATGATGACTTCTCGTTCGATT-3′) and oMC547 (5′-ATTAGCTTGGCTGCAGGTCGACGGATCCCCCTAAGCAGATTCAGTTCCGGT-3′). The bait con struct for a fusion between GAL4 and the p53 protein was used as a negative control (Durfee et al., 1993).

Two prey constructs were created similarly using Cdc2a and Cdc2b PCR fragments and the pAD-GAL4 vector. For Cdc2a, the primers oMC323, 5′-GAAGATACCCCACCAAACCCAAAAAAAGAGATGGATCAGTACGAGAAAGTT-3′, and oMC324, 5′-TGCGGGGTTTTTCAGTATCTACGATTCATACTAAGGCATGCCTCCAAGATC-3′) were used. For Cdc2b, primers oMC548, 5′-CCACCAAACCCAAAAAAAGAGATCGAATTCATGGAGAAGTACGAGAAGCTAG-3′ and oMC549, 5′-CTCTGCAGTAATACGACTCACTATAGGGCTTCAGAACTGAGACTTGTCAAG-3′ were used. These PCR fragments were co-transformed with the pAD-GAL4 plasmid linearized by digestion with SmaI and EcoRI.

The bait constructs were introduced individually into the Y187 yeast strain (MATα; Durfee et al., 1993) by selecting for transformants on plates lacking tryptophan. Similarly, the prey constructs were transformed into the PJ69-4A strain (MATa; James et al., 1996) with selection for growth without leucine. The bait strains and prey strains were mated pair-wise to form diploids that contain one of the baits and one of the preys. Diploid cells were selected on plates lacking both tryptophan and leucine, and then tested for growth on plates without tryptophan, leucine and histidine.


We thank H.Choi, M.Kim, N.Mehta, Y.Sun and J.Wang for assistance with plant work, T.Mullagan and A.Omeis for plant growth and maintenance, and S.Teplin and D.Grove for sequence analysis. We thank P.Doerner for providing the Arabidopsis Cdc2a and Cyc1 cDNA clones, R.Dixit for providing the Arabidopsis Cdc2b cDNA clone, S.Elledge for the gift of p53 bait construct and yeast strains, and E.Craig for yeast strains. We are also grateful to A.Villeneuve, C.Makaroff, F.Solomon, W.Z.Cande and W.Ni for helpful discussions and critical reading of this manuscript. This work was supported by funds from Department of Biology and the Life Sciences Consortium at the Pennsylvania State University, and by grants from the US Department of Agriculture, the National Science Foundation and the National Institutes of Health. Y.A. is grateful for the support from the Promotion and Mutual Aid Corporation for Private Schools of Japan for his sabbatical visit.


  • Alexander M.P. (1969) Differential staining of aborted and nonaborted pollen. Stain Technol., 44, 117–122. [PubMed]
  • Anderson L.K., Offenberg,H.H., Verkuijlen,W.M.H.C. and Heyting,C. (1997) RecA-like proteins are components of early meiotic nodules in lily. Proc. Natl Acad. Sci. USA, 94, 6868–6873. [PMC free article] [PubMed]
  • Andrews B. and Measday,V. (1998) The cyclin family of budding yeast: abundant use of a good idea. Trends Genet., 14, 66–72. [PubMed]
  • Armstrong S.J. and Jones,G.H. (2001) Female meiosis in wild-type Arabidopsis thialiana and two meiotic mutants. Sex. Plant Reprod., 13, 177–183.
  • Ashley T. and Plug,A. (1998) Caught in the act: deducing meiotic function from protein immunolocalization. In Handel,M.A. (ed.), Meiosis and Gametogenesis. Academic Press, San Diego, Vol. 37, pp. 201–239. [PubMed]
  • Ashley T., Walpita,D. and de Rooij,D.G. (2001) Localization of two mammalian cyclin dependent kinases during mammalian meiosis. J. Cell Sci., 114, 685–693. [PubMed]
  • Bai X., Peirsion,B.N., Dong,F., Cai,X. and Makaroff,C.A. (1999) Isolation and characterization of SYN1, a RAD21-like gene essential for meiosis in Arabidopsis. Plant Cell, 11, 417–430. [PMC free article] [PubMed]
  • Bell C.J. and Ecker,J.R. (1994) Assignment of 30 microsatellite loci to the linkage map of Arabidopsis. Genomics, 19, 137–144. [PubMed]
  • Bhatt A.M., Lister,C., Page,T., Fransz,P., Findlay,K., Jones,G.H., Dickinson,H.G. and Dean,C. (1999) The DIF1 gene of Arabidopsis is required for meiotic chromosome segregation and belongs to the REC8/RAD21 cohesin gene family. Plant J., 19, 463–472. [PubMed]
  • Bishop D.K., Park,D., Xu,L. and Kleckner,N. (1992) DMC1: a meiosis-specific yeast homolog of Escherichia coli recA required for recombination, synaptonemal complex formation and cell cycle progression. Cell, 69, 439–456. [PubMed]
  • Chevaillier P. (1993) PEST sequences in nuclear proteins. Int. J. Biochem., 25, 479–482. [PubMed]
  • Copenhaver G.P., Browne,W.E. and Preuss,D. (1998) Assaying genome-wide recombination and centromere functions with Arabidopsis tetrads. Proc. Natl Acad. Sci. USA, 95, 247–252. [PMC free article] [PubMed]
  • Couteau F., Belzile,F., Horlow,C., Grandjean,O., Vezon,D. and Doutriaux,M.P. (1999) Random chromosome segregation without meiotic arrest in both male and female meiocytes of a dmc1 mutant of Arabidopsis. Plant Cell, 11, 1623–1634. [PMC free article] [PubMed]
  • Dahmann C. and Futcher,B. (1995) Specialization of B-type cyclins for mitosis or meiosis in S. cerevisiae. Genetics, 140, 957–963. [PMC free article] [PubMed]
  • Dawe R.K. (1998) Meiotic chromosome organization and segregation in plants. Annu. Rev. Plant Physiol. Plant Mol. Biol., 49, 371–395. [PubMed]
  • Dernburg A.F., McDonald,K., Moulder,G., Barstead,R., Dresser,M. and Villeneuve,A.M. (1998) Meiotic recombination in C.elegans initiates by a conserved mechanism and is dispensable for homologous chromosome synapsis. Cell, 94, 387–398. [PubMed]
  • Dirick L., Goetsch,L., Ammerer,G. and Byers,B. (1998) Regulation of meiotic S phase by Ime2 and a Clb5,6-associated kinase in Saccharomyces cerevisiae. Science, 281, 1854–1857. [PubMed]
  • Drews G.N., Bowman,J.L. and Meyerowitz,E.M. (1991) Negative regulation of the Arabidopsis homeotic gene AGAMOUS by the APETALA2 product. Cell, 65, 991–1002. [PubMed]
  • Durfee T., Becherer,K., Chen,P.-L., Yeh,S.-H., Yang,Y., Kilburn,A.E., Lee,W.-H. and Elledge,S.J. (1993) The retinoblastoma protein associates with the protein phosphatase type 1 catalytic subunit. Genes Dev., 7, 555–569. [PubMed]
  • Epstein C.B. and Cross,F.R. (1992) CLB5: a novel B cyclin from budding yeast with a role in S phase. Genes Dev., 6, 1695–1706. [PubMed]
  • Evens T., Rosenthal,E., Youngblom,J., Distel,D. and Hunt,T. (1983) Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell, 33, 389–396. [PubMed]
  • Fan H.-Y., Hu,Y., Tudor,M. and Ma,H. (1997) Specific interactions between the K domains of AG and AGLs, members of the MADS domain family of DNA binding proteins. Plant J., 12, 999–1010. [PubMed]
  • Fields S. and Song,O. (1989) A novel genetic system to detect protein–protein interactions. Nature, 340, 245–246. [PubMed]
  • Franklin A.E. and Cande,W.Z. (1998) Rad51 distribution is altered in desynaptic2, a maize meiotic mutant that has abnormal chromosome pairing. Mol. Biol. Cell, 9, 404A.
  • Franklin A.E., McElver,J., Sunjevaric,I., Rothstein,R., Bowen,B. and Cande,W.Z. (1999) Three-dimensional microscopy of the Rad51 recombination protein during meiotic prophase. Plant Cell, 11, 809–824. [PMC free article] [PubMed]
  • Futcher A.B. (1991) Saccharomyces cerevisiae cell cycle: cdc28 and the G1 cyclins. Semin. Cell Biol., 2, 205–212. [PubMed]
  • Gitig D.M. and Koff,A. (2000) Cdk pathway: cyclin-dependent kinases and cyclin-dependent kinase inhibitors. Methods Mol. Biol., 142, 109–123. [PubMed]
  • Gorl M., Merrow,M., Huttner,B., Johnson,J., Roenneberg,T. and Brunner,M. (2001) A PEST-like element in FREQUENCY determines the length of the circadian period in Neurospora crassa. EMBO J., 20, 7074–7084. [PMC free article] [PubMed]
  • Grandin N. and Reed,S.I. (1993) Differential function and expression of Saccharomyces cerevisiae B-type cyclins in mitosis and meiosis. Mol. Cell. Biol., 13, 2113–2125. [PMC free article] [PubMed]
  • Grelon M., Vezon,D., Gendrot,G. and Pelletier,G. (2001) AtSPO11-1 is necessary for efficient meiotic recombination in plants. EMBO J., 20, 589–600. [PMC free article] [PubMed]
  • Grossniklaus U., Vielle-Calzada,J.P., Hoeppner,M.A. and Gagliano,W.B. (1998) Maternal control of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis. Science, 280, 446–450. [PubMed]
  • Hadwiger J.A., Wittenberg,C., Richardson,H.E., de Barros Lopes,M. and Reed,S.I. (1989) A family of cyclin homologs that control the G1 phase in yeast. Proc. Natl Acad. Sci. USA, 86, 6255–6259. [PMC free article] [PubMed]
  • Hemerly A., Bergounioux,C., Van Montagu,M., Inze,D. and Ferreira,P. (1992) Genes regulating the plant cell cycle: isolation of a mitotic-like cyclin from Arabidopsis thaliana. Proc. Natl Acad. Sci. USA, 89, 3295–3299. [PMC free article] [PubMed]
  • Hemerly A.S., Ferreira,P.C., Van Montagu,M., Engler,G. and Inze,D. (2000) Cell division events are essential for embryo patterning and morphogenesis: studies on dominant-negative cdc2aAt mutants of Arabidopsis. Plant J., 23, 123–130. [PubMed]
  • Hirano T. (2000) Chromosome cohesion, condensation and separation. Annu. Rev. Biochem., 69, 115–144. [PubMed]
  • Hirayama T., Imajuku,Y., Anai,T., Matsui,M. and Oka,A. (1991) Identification of two cell-cycle-controlling cdc2 gene homologs in Arabidopsis thaliana. Gene, 105, 159–165. [PubMed]
  • Huang H., Weiss,C.A. and Ma,H. (1994) Regulated expression of the Arabidopsis Gα gene GPA1. Int. J. Plant Sci., 155, 3–14.
  • Hunt T. (1991) Destruction’s our delight. Nature, 349, 100–101. [PubMed]
  • James P., Halladay,J. and Craig,E.A. (1996) Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics, 144, 1425–1436. [PMC free article] [PubMed]
  • Kaszas E. and Cande,W.Z. (2000) Phosphorylation of histone H3 is correlated with changes in the maintenance of sister chromatid cohesion during meiosis in maize, rather than the condensation of the chromatin. J. Cell Sci., 113, 3217–3226. [PubMed]
  • Kaul M.L.H. and Murphy,T.G.K. (1985) Mutant genes affecting higher plant meiosis. Theor. Appl. Genet., 70, 449–466. [PubMed]
  • Klein F., Mahr,P., Galova,M., Buonomo,S.B., Michaelis,C., Nairz,K. and Nasmyth,K. (1999) A central role for cohesins in sister chromatid cohesion, formation of axial elements and recombination during yeast meiosis. Cell, 98, 91–103. [PubMed]
  • Klimyuk V.I. and Jones,J.D. (1997) AtDMC1, the Arabidopsis homologue of the yeast DMC1 gene: characterization, transposon-induced allelic variation and meiosis-associated expression. Plant J., 11, 1–14. [PubMed]
  • La Cour L.F. and Wells,B. (1970) Meiotic prophase in anthers of asynaptic wheat: a light and electron microscopical study. Chromosoma, 29, 419–427.
  • Liu D., Matzuk,M.M., Sung,W.K., Guo,Q., Wang,P. and Wolgemuth,D.J. (1998) Cyclin A1 is required for meiosis in the male mouse. Nature Genet., 20, 377–380. [PubMed]
  • Liu D., Liao,C. and Wolgemuth,D.J. (2000) A role for cyclin A1 in the activation of MPF and G2–M transition during meiosis of male germ cells in mice. Dev. Biol., 224, 388–400. [PubMed]
  • Liu Y., Mitsukawa,N., Oosumi,T. and Whitier,R.F. (1995) Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetric interlaced PCR. Plant J., 8, 457–463. [PubMed]
  • Ma H., Kunes,S., Schatz,P.J. and Botstein,D. (1987) Plasmid construction by homologous recombination in yeast. Gene, 58, 201–216. [PubMed]
  • Maguire M.P. (1978) Evidence for separate genetic control of crossing over and chiasma maintenance in maize. Chromosoma, 65, 173–183.
  • Maluszynska J. and Heslop-Harrison,J.S. (1991) Localization of tandemly repeated DNA sequences in Arabidopsis thaliana. Plant J., 1, 159–166.
  • Mercier R., Vezon,D., Bullier,E., Motamayor,J.C., Sellier,A., Lefevre,F., Pelletier,G. and Horlow,C. (2001) SWITCH1 (SWI1): a novel protein required for the establishment of sister chromatid cohesion and for bivalent formation at meiosis. Genes Dev., 15, 1859–1871. [PMC free article] [PubMed]
  • Meuwissen R.L., Offenberg,H.H., Dietrich,A.J., Riesewijk,A., van Iersel,M. and Heyting,C. (1992) A coiled-coil related protein specific for synapsed regions of meiotic prophase chromosomes. EMBO J., 11, 5091–5100. [PMC free article] [PubMed]
  • Michaelis C., Ciosk,R. and Nasmyth,K. (1997) Cohesins: chromosomal proteins that prevent premature separation of sister chromatids. Cell, 91, 35–45. [PubMed]
  • Molnar M., Bahler,J., Sipiczki,M. and Kohli,J. (1995) The rec8 gene of Schizosaccharomyces pombe is involved in linear element formation, chromosome pairing and sister-chromatid cohesion during meiosis. Genetics, 141, 61–73. [PMC free article] [PubMed]
  • Murray A.W. (1994) Cyclin-dependent kinases: regulators of the cell cycle and more. Chem. Biol., 1, 191–195. [PubMed]
  • Nash R., Tokiwa,G., Anand,S., Erickson,K. and Futcher,A.B. (1988) The WHI1+ gene of Saccharomyces cerevisiae tethers cell division to cell size and is a cyclin homolog. EMBO J., 7, 4335–4346. [PMC free article] [PubMed]
  • Nasmyth K. (1999) Separating sister chromatids. Trends Biochem. Sci., 24, 98–104. [PubMed]
  • Nelson O.E. and Clary,G.B. (1952) Genetic control of semi-sterility in maize. J. Hered., 43, 205–210.
  • Orr-Weaver T.L. (1999) The ties that bind: localization of the sister-chromatid cohesin complex on yeast chromosomes. Cell, 99, 1–4. [PubMed]
  • Paques F. and Haber,J.E. (1999) Multiple pathways of recombination induced by double strand breaks in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev., 63, 349–404. [PMC free article] [PubMed]
  • Peirson B.N., Bowling,S.E. and Makaroff,C.A. (1997) A defect in synapsis causes male sterility in a T-DNA-tagged Arabidopsis thaliana mutant. Plant J., 11, 659–669. [PubMed]
  • Pines J. (1993) Cyclins and cyclin-dependent kinases: take your partners. Trends Biochem. Sci., 18, 195–197. [PubMed]
  • Ravnik S.E. and Wolgemuth,D.J. (1999) Regulation of meiosis during mammalian spermatogenesis: the A-type cyclins and their associated cyclin-dependent kinases are differentially expressed in the germ-cell lineage. Dev. Biol., 207, 408–418. [PubMed]
  • Reed S.I. (1991) G1-specific cyclins: in search of an S-phase-promoting factor. Trends Genet., 7, 95–99. [PubMed]
  • Renaudin J.P. et al. (1996) Plant cyclins: a unified nomenclature for plant A-, B- and D-type cyclins based on sequence organization. Plant Mol. Biol., 32, 1003–1018. [PubMed]
  • Roberts J.M., Koff,A., Polyak,K., Firpo,E., Collins,S., Ohtsubo,M. and Massague,J. (1994) Cyclins, Cdks and cyclin kinase inhibitors. Cold Spring Harbor Symp. Quant. Biol., 59, 31–38. [PubMed]
  • Rockmill B., Sym,M., Scherthan,H. and Roeder,G.S. (1995) Roles for two RecA homologs in promoting meiotic chromosome synapsis. Genes Dev., 9, 2684–2695. [PubMed]
  • Roeder G.S. (1997) Meiotic chromosomes: it takes two to tango. Genes Dev., 11, 2600–2621. [PubMed]
  • Rogers S., Wells,R. and Rechsteiner,M. (1986) Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science, 234, 364–368. [PubMed]
  • Ross K.J., Fransz,P. and Jones,G.H. (1996) A light microscopic atlas of meiosis in Arabidopsis thaliana. Chromosome Res., 4, 507–516. [PubMed]
  • Salama S.R., Hendricks,K.B. and Thorner,J. (1994) G1 cyclin degradation: the PEST motif of yeast Cln2 is necessary, but not sufficient, for rapid protein turnover. Mol. Cell. Biol., 14, 7953–7966. [PMC free article] [PubMed]
  • Schwob E. and Nasmyth,K. (1993) CLB5 and CLB6, a new pair of B cyclins involved in DNA replication in Saccharomyces cerevisiae. Genes Dev., 7, 1160–1175. [PubMed]
  • Shinohara A., Gasior,S., Ogawa,T., Kleckner,N. and Bishop,D.K. (1997) Saccharomyces cerevisiae recA homologues RAD51 and DMC1 have both distinct and overlapping roles in meiotic recombination. Genes to Cells, 2, 615–629. [PubMed]
  • Smith K.N. and Nicolas,A. (1998) Recombination at work for meiosis. Curr. Opin. Genet. Dev., 8, 200–211. [PubMed]
  • Smith K.N., Penkner,A., Ohta,K., Klein,F. and Nicolas,A. (2001) B-type cyclins CLB5 and CLB6 control the initiation of recombination and synaptonemal complex formation in yeast meiosis. Curr. Biol., 11, 88–97. [PubMed]
  • Smyth D.R., Bowman,J.L. and Meyerowitz,E.M. (1990) Early flower development in Arabidopsis. Plant Cell, 2, 755–767. [PMC free article] [PubMed]
  • StoopMyer C. and Amon,N. (1999) Meiosis: Rec8 is the reason for cohesion. Nat. Cell Biol., 1, E125–E127. [PubMed]
  • Stuart D. and Wittenberg,C. (1998) CLB5 and CLB6 are required for premeiotic DNA replication and activation of the meiotic S/M checkpoint. Genes Dev., 12, 2698–2710. [PMC free article] [PubMed]
  • Suja J.A., Antonio,C., Debec,A. and Rufas,J.S. (1999) Phosphorylated proteins are involved in sister-chromatid arm cohesion during meiosis I. J. Cell Sci., 112, 2957–2969. [PubMed]
  • Sundaresan V., Springer,P., Volpe,T., Haward,S., Jones,J., Dean,C., Ma,H. and Martienssen,R. (1995) Patterns of gene action in plant development revealed by enhancer trap and gene trap transposable elements. Genes Dev., 9, 1797–1810. [PubMed]
  • Swenson K.I., Farrell,K.M. and Ruderman,J.V. (1986) The clam embryo protein cyclin A induces entry into M phase and the resumption of meiosis in Xenopus oocytes. Cell, 47, 861–870. [PubMed]
  • Tyers M., Tokiwa,G., Nash,R. and Futcher,B. (1992) The Cln3–Cdc28 kinase complex of S.cerevisiae is regulated by proteolysis and phosphorylation. EMBO J., 11, 1773–1784. [PMC free article] [PubMed]
  • Watanabe Y. and Nurse,P. (1999) Cohesin Rec8 is required for reductional chromosome segregation at meiosis. Nature, 400, 461–464. [PubMed]
  • Wei Y., Mizzen,C.A., Cook,R.G., Gorovsky,M.A. and Allis,C.D. (1998) Phosphorylation of histone H3 at serine 10 is correlated with chromosome condensation during mitosis and meiosis in Tetrahymena. Proc. Natl Acad. Sci. USA, 95, 7480–7484. [PMC free article] [PubMed]
  • Wei Y., Yu,L., Bowen,J., Gorovsky,M.A. and Allis,C.D. (1999) Phosphorylation of histone H3 is required for proper chromosome condensation and segregation. Cell, 97, 99–109. [PubMed]
  • Wolgemuth D.J., Rhee,K., Wu,S. and Ravnik,S.E. (1995) Genetic control of mitosis, meiosis and cellular differentiation during mammalian spermatogenesis. Reprod. Fertil. Dev., 7, 669–683. [PubMed]
  • Woods L.M., Hodges,C.A., Baart,E., Baker,S.M., Liskay,M. and Hunt,P.A. (1999) Chromosomal influence on meiotic spindle assembly: abnormal meiosis I in female Mlh1 mutant mice. J. Cell Biol., 145, 1395–1406. [PMC free article] [PubMed]
  • Yang M., Hu,Y., Lodhi,M., McCombie,R. and Ma,H. (1999) The Arabidopsis SKP1-LIKE1 gene is essential for male meiosis and may control homologue separation. Proc. Natl Acad. Sci. USA, 96, 11416–11421. [PMC free article] [PubMed]
  • Zickler D. and Kleckner,N. (1999) Meiotic chromosomes: integrating structure and function. Annu. Rev. Genet., 33, 603–754. [PubMed]

Articles from The EMBO Journal are provided here courtesy of The European Molecular Biology Organization
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...