Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
A molecular cytogenetic map of Medicago truncatula (2n = 2x = 16) was constructed on the basis of a pachytene DAPI karyogram. Chromosomes at this meiotic prophase stage are 20 times longer than at mitotic metaphase, and display a well differentiated pattern of brightly fluorescing heterochromatin segments. We describe here a pachytene karyogram in which all chromosomes can be identified based on chromosome length, centromere position, heterochromatin patterns, and the positions of three repetitive sequences (5S rDNA, 45S rDNA and the MtR1 tandem repeat), visualized by fluorescence in situ hybridization (FISH). We determined the correlation between genetic linkage groups and chromosomes by FISH mapping of bacterial artificial chromosome (BAC) clones, with two to five BACs per linkage group. In the cytogenetic map, chromosomes were numbered according to their corresponding linkage groups. We determined the relative positions of the 20 BACs and three repetitive sequences on the pachytene chromosomes, and compared the genetic and cytological distances between markers. The mapping resolution was determined in a euchromatic part of chromosome 5 by comparing the cytological distances between FISH signals of clones of a BAC contig with their corresponding physical distance, and showed that resolution in this region is about 60 kb. The establishment of this FISH pachytene karyotype, with a far better mapping resolution and detection sensitivity compared to those in the highly condensed mitotic metaphase complements, has created the basis for the integration of molecular, genetic and cytogenetic maps in M. truncatula.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The Fabaceae (legumes) are the third largest family of flowering plants, and display a striking variety of plant types ranging from small annual herbs to massive tropical trees. Among the legumes, the subfamily Papilionoideae contains the majority of agronomic species, including the galegoid species such as Pisum sativum (pea) and Medicago sativa (alfalfa), and beans such as Phaseolus, Vigna and Glycine spp. In addition to their agronomic importance as producers of oil and protein for the diets of humans and livestock, legumes have the distinction of being one of the major sources of biologically available nitrogen in the biosphere, due to their unique ability to establish a symbiosis with several genera of bacteria that are collectively called rhizobia. This symbiotic interaction results in the formation of root nodules where the bacteria are capable of reducing atmospheric nitrogen. Due to their narrow host range, none of these microsymbionts can interact with model plants such as Arabidopsis and rice, and hence legume species have been proposed as a model system for unravelling the molecular mechanisms that control the development and functioning of this beneficial plant–microbe interaction. Medicago truncatula (barrel medic) has been selected for this purpose (Barker et al., 1990). This annual, autogamous diploid forage crop has 16 chromosomes and a relatively small genome (1.15 pg/2C, Blondon et al., 1994) of about the same size as that of rice and four times larger than that of Arabidopsis. In addition, molecular and genetic studies of M. truncatula became possible with recently developed efficient transformation and regeneration procedures, and the establishment of bacterial artificial chromosome (BAC) libraries and genetic maps (Cook, 1999).
In order to position BACs and other DNA sequences along the chromosomes, molecular cytogenetic maps are needed. Among the tools for such maps, fluorescence in situ hybridization (FISH) is undoubtedly the most versatile and accurate for ordering repetitive and single-copy DNA sequences on the chromosomes with respect to centromeres, telomeres and heterochromatic regions (De Jong et al., 1999; Jiang and Gill, 1994). Cytogenetic maps can be highly informative to support the construction of physical maps and map-based cloning projects, and to position genes in large heterochromatic regions where linkage distances are inaccurate due to the low levels of meiotic recombination (Lambie and Roeder, 1986; Roberts, 1965; Zhong et al., 1999).
Microscopic preparations with mitotic metaphase complements for FISH are mostly prepared from root tips. Chromosomes at this stage are highly condensed, which limits the optical resolution of adjacent FISH targets to Mbs rather than kbs. Where higher resolution is required, pachytene chromosomes are more appropriate because their complements measure 10–40 times longer and display a differentiated pattern of heterochromatic and euchromatic regions (Albini and Schwarzacher, 1992; Fransz et al., 1998; Fransz et al., 2000; Shen et al., 1987; Zhong et al., 1996). This heterochromatin banding, along with chromosome length and centromere position, is important for the identification of individual chromosomes. However, the production of high-quality pachytene spreads in comparison to metaphase chromosomes is technically more demanding, and results may differ between related species and even between genotypes of the same species (reviewed by De Jong et al., 1999).
The published karyotypes of M. truncatula were based on simple morphometric measurements of metaphase plates from accession V357 (Agarwal and Gupta, 1983) and the accessions R-108-1 and Jemalong J5 (Gerbah et al., 1999). However, the morphology of metaphase chromosomes was found to be too similar to allow reliable identification of all chromosomes on the basis of relative arm lengths and centromere positions. FISH of 5S and 45S rDNAs, which were applied as additional markers in the R-108-1 and Jemalong J5 accessions (Gerbah et al., 1999), showed that both genotypes have a single NOR (nucleolar organizing region), and that Jemalong J5 had three 5S rDNA loci whereas R-108-1 has only two.
Pachytene karyotypes of several Medicago species were described, including tetraploid and diploid Medicago sativa (Buss and Cleveland, 1968; Gillies, 1968; Gillies, 1970b), and species closely related to M. sativa such as M. falcata (Gillies, 1970a), M. glomerata (Gillies, 1971), M. coerulea, M. glandulosa, M. glutinosa and M. prostrata (Gillies, 1972). Chromosomes at the pachytene stage of microspoocytes were measured and characterized by average total chromosome length, arm length, arm length ratio and number, position and size of chromatic area.
This study is the first pachytene chromosome map of M. truncatula based on the Jemalong A17 genotype, which is the standard line in most current molecular genetic studies (Nam et al., 1999; Trieu et al., 2000; D.-J. Kim and D.R. Cook, unpublished results). The chromosomes were stained with DAPI and their fluorescence images were used for karyotype analysis on the basis of morphometric data and heterochromatin patterns. In addition, the pericentromeric repeat MtR1 and 5S rDNA were tested as probes in FISH experiments to facilitate chromosome identification. For assigning linkage groups to the cytogenetic map, we hybridized BAC clones positioned on the genetic map of M. truncatula (D.-J. Kim and D.R. Cook, unpublished results) to the pachytene complements, and mapped their positions with respect to centromeres and heterochromatin segments.
The pachytene karyotype
Pollen mother cells at late pachytene clearly displayed eight fully paired bivalents with lengths varying from 29 to 68 µm and a total complement length of 406 µm. Chromo somes were submetacentric or metacentric, with centromere indexes between 27 and 47% (Figure 1a; Table 1). They were numbered according to their corresponding linkage maps, as decided during the Second Medicago truncatula Workshop (Amsterdam, The Netherlands, July 22–23 1999), with their two arms denoted as S (short) and L (long), respectively. DAPI staining of pachytene chromosomes demonstrated striking differences in chromatin density. Brightly fluorescing heterochromatic blocks were detected in the pericentromeric regions of all chromosomes, although their lengths were different for each chromosome (Figure 1a; Table 1). The centromere itself appeared as a primary constriction in which the chromatin fluoresces far more weakly than the flanking heterochromatic regions (Figure 1a). Distal regions of the chromosome arms generally consist of weakly fluorescing euchromatin. In addition to the pericentromeric heterochromatin, smaller heterochromatic knobs were observed on the short arms of chromosomes 3, 4 and 7, and both arms of chromosome 6, but their number was variable and could be observed in few cells only. The total length of all heterochromatic areas is 59.3 µm, about 15% of the length of the complement.
Table 1. Absolute and relative lengths of individual chromosomes and chromosome regions, positions of 5S rDNA, NOR and MtR1 on chromosomes
Chromosome identification was based on their length, centromere position, size of pericentromeric heterochromatin, and the presence of diagnostic heterochromatic knobs. Eight cells were selected in which all eight bivalents could be discerned and fully traced along their length. Based on this morphometric characterization, the chromosomes could be distinguished as follows (Table 1).
Group of three longest chromosomes (1, 3, 4). These chromosomes measure 60–68 µm and have submedian centromere positions with centromere index (CI) values of 36, 27 and 30%, respectively. Based on centromere position, chromosome 1 can be distinguished from chromosomes 3 and 4. Chromosomes 3 and 4 have similar symmetrical heterochromatic regions, and differ only slightly in arm lengths and centromere positions. Distinction between them without additional diagnostic markers was therefore doubtful.
Group of three medium-sized chromosomes (2, 5, 7). These chromosomes measure approximately 50 µm. The former two have median centromere positions (centromere indexes of 47 and 46%, respectively), whereas chromosome 7 has a submedian centromere (CI = 30%). Furthermore, chromosome 5 has a characteristic pattern of four conspicuous pericentromeric heterochromatic knobs. In contrast, chromosomes 2 and 7 have only one knob on each arm. Furthermore, chromosome 5 contains the secondary constriction (nucleolar organizer region). This weakly fluorescing region is located on the long arm, close to the centromere, and is often clumped together with heterochromatic blocks of other chromosomes.
Group of two smallest chromosomes (6, 8). These chromosomes are 29 and 33 µm long, respectively, and can easily be distinguished by differences in chromatic patterns. Chromosome 6 has several heterochromatic chromomeres on both arms, whereas chromosome 8 has two larger heterochromatic blocks on either side of the centromere.
The total length of the pachytene chromosomes is 406 µm, which is about 20 times longer than that of the mitotic metaphase chromosomes (data not shown). The morphological features of the eight bivalents have been used to construct an ideogram (Figure 2a). We selected the 5S rDNA, 45S rDNA and the MtR1 tandem repeats as additional diagnostic markers to facilitate the identification of the individual chromosomes in the cell complement. FISH hybridization revealed that the 5S rDNA loci are located on chromosomes 2, 5 and 6 (Figure 1b,d). A major 5S rDNA region occurs on the distal part of the pericentromeric heterochromatin of chromosome 5, on the arm containing a single heterochromatic knob. A second, smaller 5S rDNA region is located on the long arm of chromosome 2, close to the border of the pericentromeric heterochromatin. A third 5S rDNA site is present on chromosome 6, 17% of the arm length distally from the centromere. FISH with the 45S rDNA probe demonstrated a bright spot on the secondary constriction of chromosome 5, between two proximal heterochromatic knobs (Figure 1b,d). The same number of 5S and 45S rDNA loci were observed by FISH studies on metaphase chromosomes (Gerbah et al., 1999). The MtR 1 tandem repeat has a 166 bp motif, was identified in two randomly isolated BAC clones, BAC75N01 and BAC53F10 (Table 2), and will be described elsewhere in more detail (O. Kulikova, T. Huguet, H. de Jong and T. Bisseling, unpublished results). This repeat is located in the pericentromeric regions of the chromosome arms 1-L, 2-L, 4-S, 7-S and 7-L, and 8-L (Figure 1c). The MtR1 signal on 8-L is weaker than that of the other MtR1 signals. Thus MtR1 is a good marker to distinguish chromosomes 3 and 4.
Table 2. Characteristics of BAC clones and DNA markers
BEST; homology to Arabidopsis hypothetical protein Z97335. BAC 54F15 survey sequencing reveals homology to multiple genesd
57.2 ± 1.1
BEST; similar to NBS-LRR disease resistance protein (AB019186) RPR1 of Oryza sativa
16.2 ± 0.1
BEST; no known homology for marker template. BAC 10F20 survey sequencing reveals homology to multiple genes
26.3 ± 1.0
BEST; no known homology for marker template. BAC 01P05 contains putative MYB family transcription factor
43.7 ± 1.2
BEST; BAC 64B21 is contiguous with BAC 42H09 from which survey sequencing reveals homology to multiple genes, including Arabidopsis ein3-like family
81.0 ± 1.0
PCR marker 3′′ of ENOD40 coding region. BAC 23A06 contains ENOD40 gene and putative receptor protein kinase
40.1 ± 1.6
BEST; no known homology for marker template. BAC 35012 survey sequencing reveals homology to multiple genes
64.2 ± 1.1
BEST; no known homology for marker template
BEST; no known homology for marker template. BAC45I09 survey sequencing reveals homology to Mt ESTs
BEST; similar to beta-transducin. BAC 19L20 contains Medicago truncatula cycloartenol synthase gene (Y15366.1)
100.0 ± 0.0
BEST; homology to tomato callus EST AW029689
83.1 ± 1.0
BEST; BAC contains homology to LBS-LRR-TIR family of resistance genes
89.3 ± 2.1
BEST; no known homology for marker template
57.3 ± 1.1
BEST; no known homology for marker template
44.8 ± 1.1
BEST; similar to peptide transporter. BAC contains homology to Arabidopsis genomic DNA by tblastX
37.3 ± 0.1
BEST; similar to Arabidopsis hypothetical protein. BAC 41H08 contains Medicago truncatula EST AW775698
53.1 ± 0.1
Together, as shown in Figure 2a and Table 1, the karyotype analysis of the pachytene chromosome morphology and FISH patterns of the 5S rDNA, 45S rDNA, and MtR1 repeats allow identification of all eight chromosomes.
Integration of cytogenetic map and linkage groups
The numbering convention for the eight genetically identified linkage groups of M. truncatula was adopted from M. sativa (Kalóet al., 2000) as determined by comparative map analysis (D.-J. Kim and D.R. Cook, unpublished results). For assigning individual linkage groups to the chromosomes, we selected BACs of each linkage group and used these as probes for FISH mapping to the pachytene chromosomes (Table 2). Concurrent hybridizations with the 5S rDNA and MtR1 repeats were used to assist chromosome identification. Only FISH signals that occurred in at least 90% of the cell complements were considered for quantitative analysis. Examples of representative FISH patterns are shown in Figure 2(b). We measured the distance of the FISH signal in a relative scale from centromere to telomere in seven to ten cells (Table 2). Their distance values were averaged for drawing their positions on the cytogenetic map (Figure 2a).
To provide a frame of reference to the genetic map of M. truncatula, five or more sequence-characterized genetic markers are indicated for each linkage group (Figure 2a). These markers were developed on the basis of BAC end-sequence information (see Experimental procedures). Two to five BAC clones were used for mapping on pachytene chromosomes. Figure 2a shows their genetic map position along with their corresponding position on the pachytene FISH map. Detailed information on individual markers is given in Table 2 and at http://chrysie.tamu.edu/perl-bin/mt_marker_query.2.pl.
As the distal parts of most chromosomes are euchromatic, it is expected that the genetic and FISH map distances would be directly correlated. Table 3 gives a comparison of genetic map distances between six marker pairs and their corresponding microscopic distances on the pachytene FISH map. The ratio of genetic and cytogenetic distance values ranges from 0.9 to 1.9 cm µm−1 for most marker pairs. However, this value is markedly higher (3.0–3.1 cM µm−1) for the pairs located on the small chromosomes 6 and 8.
Table 3. Comparison of genetic and cytogenetic distances between neighbouring BAC clone pairs
Genetic distance (cM)
Microscopic distance (µm)
Resolution of FISH mapping
The BAC clones 58F01 and 59K07 from a BAC contig of chromosome 5 (G. Gualtieri, R. Geurts and T. Bisseling, unpublished results) were selected for estimating the resolution of FISH mapping in a euchromatic region of a pachytene chromosome 5 (Figure 2a). The BACs were hybridized and detected with digoxigenin-FITC (green signal) and biotin-Texas Red (red signal), respectively (Figure 3a,b). A small band of yellow fluorescence in between the red and green spots represents the region of signal overlap. The mid-points of the sequences covered by these BACs are separated by 150 kb. This corresponds to a microscopic distance of approximately 0.5 µm between the centres of the red and green spots on the pachytene chromosome, implying a chromatin density of 300 kb µm−1 for that euchromatin region. With the spatial resolution limit of 0.2 µm for the fluorescence microscope, mapping resolution in this chromosome segment can be estimated at about 60 kb. This is confirmed by a comparable FISH experiment with the BAC clones 45I09 and 63C24 located in the same region of chromosome 5 (Figure 2a). These BACs are about the same size, approximately 40 kb, and are separated by about 55 kb (Figure 3c,d). Fluorescence microscopic observation revealed a prominent yellow spot flanked by small green and red regions, confirming that mapping resolution of adjacent targets in this region is about 60 kb.
We have shown here that DAPI-stained pachytene chromosomes of M. truncatula are suitable for construction of a detailed karyotype. The fully paired chromosomes at this meiotic prophase stage measure 406 µm (Table 1), 20 times longer than their counterparts at mitotic metaphase, and allows mapping with a high resolution (60 kb). In general, pachytene chromosomes clearly display the differentiation of large heterochromatic blocks around the centromere, whereas the distal parts of the arms are euchromatic. This pattern of well defined heterochromatic areas is reminiscent of the conspicuous heterochromatin blocks on Arabidopsis thaliana pachytene chromosomes (de Jong et al., 1999; Fransz et al., 1998; Ross et al., 1996), but differs strongly from other small genome species. For example, rice exhibits numerous smaller heterochromatic knobs distributed along all chromosome arms (Khan, 1975). With this simple organization of solid heterochromatin blocks flanking the centromeres and long stretches of euchromatin in the distal areas, M. truncatula now becomes an attractive model species for cytogenetic analyses.
The combination of chromosome length, centromere position and heterochromatin patterns of the DAPI-stained pachytene complements proved to be sufficient to identify all chromosomes except 3 and 4. However, with the MtR1 repeat it became possible to distinguish these two chromosomes as well. In general, chromosomes were hybridized with 5S rDNA and MtR1 to facilitate the recognition of the chromosomes, even when the chromosomes were clustered or partly overlapping.
Our second goal was to assign the eight chromosomes to the linkage groups of M. truncatula, which was recently established for this species (G.B. Kiss, T. Huguet, D.-J. Kim and D. Cook, unpublished results). We selected two to five markers for each linkage group, and mapped the corresponding BACs as probes in FISH experiments on pachytene cells. Markers belonging to a certain linkage group always hybridized to the same chromosome.
Heterochromatic regions are supposed to be rich in repeated sequences and to contain a low density of expressed genes (Dean and Schmidt, 1995). Euchromatin contains markedly fewer repeated sequences, and has a higher density of genes. The repetitive nature of heterochromatic regions is consistent with the location of MtR1 and the highly repeated ribosomal genes in heterochromatic pericentromeric regions. Furthermore, most of the linkage group-specific BACs gave distinct FISH signals in euchromatic parts of the chromosomes and no background labelling was observed, suggesting that the interstitial segments of M. truncatula chromosomes contain a relatively low number of repeated sequences. In contrast, BAC paintings on pachytene chromosomes of maize, in which transcribed genes are separated by areas of repetitive DNA sequences (Bennetzen et al., 1994; SanMiguel et al., 1996), blocking with Cot-100 fraction of genomic DNA was required to suppress signals generated by these repeated sequences (Sadder et al., 2000). Results from whole BAC clone sequencing have indicated a gene density of approximately one predicted gene per 6 kb in gene-rich regions of M. truncatula (D.J. Kim and D. Cook, unpublished results), which is consistent with the expectation that the arms of M. truncatula are rich in transcribed genes and that repeat sequences will be under-represented in the intergenic regions. This implies that positional cloning strategies for genes located in the euchromatic regions will not be hampered by a high density of repeated sequences, such as might be expected in heterochromatic parts of chromosomes.
The degree of chromatin condensation in a euchromatic part of chromosome 5 was shown to be approximately 300 kb µm−1, similar to that of Arabidopsis euchromatin which varies between 150 and 300 kb µm−1 (Fransz et al., 1998; Fransz et al., 2000). Assuming that the average degree of condensation in the euchromatic regions of M. truncatula is about 300 kb µm−1, the fraction of the M. truncatula genome contained within euchromatic and heterochromatic regions, respectively, can be estimated. As shown in Table 1, the total length of pachytene chromosomes is 406 µm, of which about 350 µm is euchromatic. Thus the total euchromatic fraction of M. truncatula DNA is estimated at 105 Mb (300 kb µm−1 × 350 µm). As the genome size is about 500 Mb, 395 Mb (almost 80% of the M. truncatula genome) is estimated to occupy heterochromatic regions. These values are the result of a rough calculation, as the degree of condensation within euchromatin has been determined in only one region. Nevertheless, it strongly indicates that the majority of the genome is located in the heterochromatic region. As this part of the genome is clustered in the pericentromeric regions which are predicted to contain few transcribed genes, it will be attractive and possible to focus a future M. truncatula genome-sequencing program on the euchromatic parts of the chromosome arms.
The detailed pachytene karyotype, along with 20 BACs which could be positioned in the euchromatic region of the long or short arms of the chromosomes (Figure 2), now provide the basis for a high-resolution cytogenetic map, and will be the first step in the integration of the physical, chromosomal and genetic maps. Once sufficient markers have been mapped to cover all chromosome regions, an informative cytogenetic map will provide an indispensable tool for map-based cloning studies where reliable estimation of the physical lengths between adjacent markers and their precise position with respect to centromere, telomere and heterochromatin areas is required. Furthermore, it can reveal interesting cytogenetic properties. For example, our studies indicate that the small chromosomes 6 and 8 display a higher cM/µm ratio than the other chromosomes. A more detailed comparison of the genetic and cytogenetic map is therefore essential to show whether this difference holds true for the other euchromatic regions of these small chromosomes, and such comparisons can provide the basis for studies of differential chromosomal behaviour.
Preparation of meiotic pachytene chromosomes
The method we developed for the preparation of Medicago pachytene spreads is adapted from the Arabidopsis procedure described by Ross et al. (1996). Immature flower buds of 1.5–1.8 mm in length were directly fixed in ethanol/acetic acid (3 : 1) for at least 3 h and can be stored in this fixative at −20°C for several months. For cell-wall digestion the buds were rinsed three times for 1 min in deionized water and transferred to a pectolytic enzyme mixture [0.3% (w/v) pectolyase Y23 (Sigma-Aldrich, St. Louis, MO, USA), 0.3% (w/v) cytohelicase (Sepracor, Jaures, France) and 0.3% (w/v) cellulase RS (Sigma) in citrate buffer (10 mm sodium citrate buffer pH 4.5)] at 37°C for 2 h. The vulnerable material was rinsed again with deionized water, and each flower bud was transferred to a droplet of water on a microscope slide. Anthers were dissected from flower buds with fine needles and transferred to a grease-free slide. The resulting cell suspension was spread on a clean glass slide with 30 µl 60% acetic acid at 45°C for 1 min. Finally, 1 ml ice-cold ethanol/acetic acid (3 : 1) was added in a circle around the suspension before leaving the slides to dry.
Probe preparation and labelling
Clone pCT4.2, which contains a 5S ribosomal DNA repeat unit (≈500 bp) of Arabidopsis thaliana in pBS (Campell et al., 1992), was PCR-labelled with biotin-16-dUTP (Boehringer Mannheim, Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer's instructions. Clone pTA71, which contains a 9.1 kb fragment of 18S-5.8S-26S rDNA of common wheat (Gerlach and Bedbrook, 1979), was labelled with digoxigenin-11-dUTP using the high-primed labelling kit (Boehringer Mannheim). The MtR1 repeat was identified in the randomly isolated BAC 75N01 (20 kb insert) and BAC 53F10 (30 kb) clones of M. truncatula(Table 2). The ends of these BAC clones are composed of a series of 166 bp long direct repeats, named MtR1 (Figure 1). Two oligonucleotide primers (5′-AAAAAT TCGAATGCACCAAAACTGG-3′ and 5′-TCAGGATCTCATGAA ACTGCTCTTTT-3′) were used to amplify a 307 bp fragment by PCR with DNA of BAC75N01. This fragment was subcloned in the pGEM-T (Promega Corporation, Madison, WI, USA). Clone pMtR1 containing a 307 bp fragment of the 166 bp MtR1 repeat motif was labelled by PCR with digoxigenin-11-dUTP (Boehringer Mannheim).
BAC clone isolation and manipulation
Bacterial artificial chromosome (BAC) clones of M. truncatula genotype Jemalong A17 were identified, either by means of hybridization to high-density filter arrays obtained from the Clemson University Genomics Institute (http://www.genome.clemson.edu), or by PCR screening of a multiplexed DNA copy of the M. truncatula BAC library as described by Nam et al. (1999). BAC DNA was isolated according to the alkaline lysate method and labelled with either biotin-dUTP or digoxigenin-dUTP using the nick translation mix (Boehringer Mannheim) for FISH. BAC end sequencing was performed on whole BAC clones using primers complementary to the pBeloBAC11 vector (‘left primer’: AACGCC AGGGTTTTCCCAGTCACGACG; ‘right primer’: ACACAGGAA ACAGCTATGACCATGATTACG). The low-pass survey sequence reported in Table 2 and at http://chrysie.tamu.edu/medicago was obtained by sequencing of fragmented sublibraries from individual BAC clones. Briefly, BAC DNA was sheared to a range of 1–3 kb and subcloned into the SmaI site of pUC18. The template DNA for sequencing was obtained either by PCR using universal primers, or by isolation of plasmid DNA. Plasmid DNA was sequenced using a universal primer directed against the pUC18 poly linker (CAGGAAACAGCTATGACCATGATTACGA).
Fluorescence in situ hybridization
FISH was performed as described in detail by Zhong et al. (1996) without pepsin treatment. For hybridizations with the BAC10F20 and BAC63C24 clones, the addition of competitor DNA was required. We added a 100-fold excess of fragmented genomic DNA to the hybridization mixture before denaturation at 80°C for 10 min and pre-annealing at 37°C for 1 h. After this treatment, the mixture was applied to slides (Jiang et al., 1995). Genomic DNA was isolated from seedlings of M. truncatula according to the CTAB DNA extraction method (Rogers and Bendish, 1988), and subsequently fragmented by autoclaving at 15 lb cm−2 for 5 min. Biotin-labelled probes were detected with Avidin–Texas Red and amplified with biotin-conjugated goat-anti-Avidin and Avidin–Texas Red. Digoxigenin-labelled probes were detected with sheep-antidigoxigenin-fluorescein (FITC) and amplified with rabbit-anti-sheep-FITC. Double fluorescence detection of probes was performed according to Fransz et al. (1996). Chromosomes were counterstained with DAPI (4′,6-diamidino-2-phenylindole) in Vectashield antifade solution (Vector Laboratories, Inc., Burlingame, CA, USA), 5 µg ml−1. Some chromosome preparations were re-used for FISH with a new set of probes according to the method of Heslop-Harrison et al. (1992). Chromosome preparations were studied and photographed under a Zeiss Axioskop fluorescence microscope equipped with separate excitation filter sets for DAPI (01), FITC (09) and Texas Red (14). FISH signals with different colours were recorded on a single photograph by double exposure. The colour negatives were scanned at 1000 dpi and their digital images were optimized for contrast and brightness using adobe photoshop 5.0.2. To separate individual bivalents, each bivalent was digitally excised and copied into a new image. Chromosomes were measured with MicroMeasure, a freeware software programme from Colorado State University (http://www.colostate.edu/Depts/Biology/MicroMeasure). Chromosome nomenclature was according to Levan et al. (1964).
Genetic markers were developed on the basis of BAC end-sequence information (a so-called BAC end sequence tag or BEST), obtained from BAC clones listed in Table 2. Briefly, oligonucleotide primers were designed based on the corresponding BAC end-sequence information and used to PCR amplify and sequence genomic DNA from the M. truncatula A17 and A20 genotypes (Penmtesa and Cook, 2000). Restriction enzyme polymorphisms were identified by comparing sequence differences between the parental genotypes against known restriction sites. The resulting CAPS (cleaved amplified polymorphic sequences) or length polymorphism markers were mapped on a population of 93 F2 progeny from a genotype A17 × A20 cross (D.-J. Kim and D.R. Cook, unpublished results). Polymorphic DNAs were resolved on a 1.5% agarose gel and visualized by ethidium bromide staining. Primers, PCR conditions and restriction enzyme information is given at http://chrysie.tamu.edu/perl-bin/mt_marker_query.2.pl. DNA was extracted from the mapping population by means of the Nucleon Phytopure kit (Amersham Pharmacia Biotech, Inc., Piscataway, NJ, USA), according to the manufacturer's instructions.
This work was financially supported by grants INTAS-96-1371, NWO 047-011-000, NSF IBN 9872664 and QLRT-2000-30676.