Molecular and chromosomal characterization of repeated and single-copy DNA sequences in the genome of Dasypyrum villosum

Molecular characterization of repeated and single-copy DNA sequences in the genome of Dasypyrum Hereditas Restriction fragment length polymorphism of ribosomal DNA repeated unit and single-copy DNA fragments and chromosomal distribution of a highly repeated sequence, have been studied to assess molecular markers and the extent of their heterogeneity in Dasypyrurn villosum. Substantial variation has been found for the length of the intergenic spacer of ribosomal genes clustered in different alleles at Nor-VI locus of heterozygous individuals, but not within the cluster of rDNA of homozygous individuals. After Southern blots and hybridization to an intergenic spacer probe, each cluster of rDNA was detected as a single band with at least four variants differing for the number of 130 bp suhrepeats in the intergenic spacer. One recombinant plasrnid contained a 2270 bp DNA insert from the D. ui//osum genome that upon Sph 1 restriction endonuclease digestion was cleaved in three 380 bp repeat elements and one 1090 bp fragment. When Southern blots of Sph I digested D. villosum DNAs of different genotypes were hybridized to the '2P-labelled 380 bp repeat, a distinct ladder consisting of multiples of a basic repeat unit of about 380 bp in length was revealed on autoradiograms. The in situ hybridization of the 'H-labelled 380 bp repeat element showed that one chromosome pair (7V) was not labelled. In the other pairs, silver grains remained clustered at or near the telomeres. Dot-blot hybridization analysis of DNAs from a range of diploid, tetraploid, and hexaploid Triticeae species indicated that the 380 bp repeated element was a specific feature of the D. dosum genome. Other cloned DNA sequences of D. villosurn showed a large restriction length polymorphism and one was located on V chromosomes.

Restriction fragment length polymorphism of ribosomal DNA repeated unit and single-copy DNA fragments and chromosomal distribution of a highly repeated sequence, have been studied to assess molecular markers and the extent of their heterogeneity in Dasypyrurn villosum. Substantial variation has been found for the length of the intergenic spacer of ribosomal genes clustered in different alleles at Nor-VI locus of heterozygous individuals, but not within the cluster of rDNA of homozygous individuals. After Southern blots and hybridization to an intergenic spacer probe, each cluster of rDNA was detected as a single band with at least four variants differing for the number of 130 bp suhrepeats in the intergenic spacer.
One recombinant plasrnid contained a 2270 bp DNA insert from the D. ui//osum genome that upon Sph 1 restriction endonuclease digestion was cleaved in three 380 bp repeat elements and one 1090 bp fragment. When Southern blots of Sph I digested D. villosum DNAs of different genotypes were hybridized to the '2P-labelled 380 bp repeat, a distinct ladder consisting of multiples of a basic repeat unit of about 380 bp in length was revealed on autoradiograms. The in situ hybridization of the 'H-labelled 380 bp repeat element showed that one chromosome pair (7V) was not labelled. In the other pairs, silver grains remained clustered at or near the telomeres. Dot-blot hybridization analysis of DNAs from a range of diploid, tetraploid, and hexaploid Triticeae species indicated that the 380 bp repeated element was a specific feature of the D. d o s u m genome. Other cloned DNA sequences of D. villosurn showed a large restriction length polymorphism and one was located on V chromosomes. Pace, Dept. Agrobiology and Agrochemistry, University of Tuscia, 01 100 Viterbo, Ifaly Dasypyrurn oillosum (L.) CANDARGY is a widespread outcrossing diploid (2n = 14) annual species, which belongs to the Triticeae tribe (FREDERIKSEN 1991).

C. De
The genome of D. villosum, indicated V by SEARS (1953), is considered an important donor of genes to wheat for improving powdery mildew resistance ( DE PACE et al. 1988a), take-all (SCOTT 1981), and plant and seed storage protein content (DELLA GATTA et al. 1984;SHEWRY et al. 1987;DE PACE et al. 1988a). The hybridization of D. vi[losum with diploid, tetraploid and hexaploid species of Triticum has been performed (SANDO 1935;SEARS 1953;HYDE 1953), and other intergeneric hybridizations involving D. villosum were tempted ( NAKAJIMA 1962). Individual V chromosomes have been added to both durum (BLANCO et al. 1987) and bread wheat (SEARS 1953;HYDE 1953).
Morphological, isozyme and seed storage protein markers have been identified to tag each of the seven V chromosome pairs ( MONTEBOVE et al. 1987;DE PACE et al. 1988;BENEDETTELLI and HART 1988) and, on average, two markers for each V chromosome are known. However, this situation is not satisfactory to monitor gene transfer from D. uillusum to wheat through intergeneric hybridization. Moreover, information on the allelic polymorphism for gene markers in D. villosum is limited (DE PACE 1987;DE PACE et al. 1988b), although they are of paramount importance for a deliberate choice of D. villosum phenotypes that possess the desired alleles for genes that may improve given wheat characters. Another problem is the identification of V chromosomes following their introduction into wheat. The Giemsa banding may be helpful, but its potential is limited, since the banding pattern of the alien chromosomes may be non-reproducible when introduced into the wheat nucleus. D. villosum DNA markers obtained by restriction fragment length polymorphism (RFLP) or repeated sequence abundance polymorphism (RSAP) (DVORAK et al. 1988) may be a valid means for the identification of defined, polymorphic, DNA sequences to be used for V chromosome identification and assessment of linkage between those sequences and other useful genes.
In this paper, we report on (i) the polymorphism for restriction sites of repeated and single-copy DNA sequences in the V genome, (ii) the molecular and chromosomal characterization of a speciesspecific highly repeated DNA sequence, and (iii) the feasibility of using single-copy sequences as markers for the V chromosomes.

Plant material
The species studied and their origin are listed in Table 1. Seeds (caryopses) were germinated in Petri dishes on wet paper at room temperature in the dark. Plantlets to be used for DNA or RNA extraction were washed, dried, quickly frozen in liquid nitrogen and stored at -80°C. The roots of other plantlets, to be used for cytological preparations, were treated with a 0.2 % aqueous solution of colchicine at room temperature and fixed in ethanol-acetic acid 3: 1 (v/v).

DNA isolation
For DNA preparation, about 0.5 g of fresh tissues were pulverized with pestle and mortar in the presence of liquid nitrogen and homogenized in 5 ml of pH 8 buffer containing 0.1 M Tris, 0.05 M EDTA, 0.5 M NaCI, and 0.01 M mercaptoethanol.
DNA extraction was performed as described by DELLAPORTA et al. (1983)

DNA cloning
Purified DNA was cleaved with Bum HI plus Pst I restriction enzymes and ligated into the Bam HI/Pst I site of pUC8. Escherichia coli DH5 a competent cells were transformed with the ligated hybrid plasmids and plated on LB medium containing 100 ng/ml ampicillin and 2 YO of Bluo-Gal (BRL). White colonies were transferred to a similar medium and colony hybridization was performed according to GRUN-STElN and OGNESS (1975) using Hybond-C (Amersham) filters and D. uillosum genomic DNA which was 32P-labelled by nick-translation according to RIGBY et al. (1977). After autoradiography, the colonies were divided in two groups: group I comprised colonies giving very low hybridization signal; group I1 comprised those giving very strong signals. From 200 colonies of group I, 5 were chosen for further characterization and were indicated as pDVA7, pDVA9, pDVII, pDV17, and pDV18 (henceforth pA7, pA9, p l l , p17, and p18, respectively). From five colonies of group 11, 1 (~2270) was chosen for further characterization.
Probe preparation DNA probe. -Plasmids from the chosen colonies were isolated using the mini-preparation boiling method (SAMBROOK et al. 1989), and digested with Barn HI plus Pst 1. The DNA was fractionated in 0.8 % low melting point agarose gel and gel portions containing the D. uillosum inserts were excised from the gels, melted at 65°C. for five minutes, and the DNA concentration was adjusted at approximately 25 ng/pl. Random sequence hexanucleotides were used to obtain "P-labelled probes with a specific activity of about 1 x lo9 cpm/pg. The labelling was performed using the "Multiprime" Kit (Amersham). A plasmid containing a 2.7 kb portion of the nontranscribed intergenic spacer of Triticum aestiuum ribosomal DNA (pTA250.4) was kindly provided by R. Appels (C.S.I.R.O., Canberra, Australia).

Southern blot hybridization
Genomic DNAs were digested with restriction endonucleases, fractionated in 1 % agarose gel, and blotted onto Zeta-probe membrane( Biorad). The membranes were prehybridized in 5 x SSC, 6 x Denhardt, 0.05 M Tris, 0.4 % SDS, 10mM EDTA, and 0.4 mg/ml heat-denatured salmon sperm DNA, at 62°C for 4 h. Heat-denatured 32P labelled DNA was then added ( I x 1O8cpm per ml of hybridization mixture) and allowed to hybridize for 12 h at 62°C. The membranes were then washed in 2 x SSC, 0.1 % SDS (2 x 5', RT), 0.1 x SSC, 0.1 YO SDS ( 1 x 15', 62"C), and exposed to X-OMAT film for autoradiography. Probe stripping was performed soon after autoradiography by washing once in a large volume of 0.1 x SSC/O.5 YO SDS at 95°C for 20 min. Filters were reprobed at least five times without losing hybridization efficiency.

Dor blot hybridization
One pg of DNA was spotted on Zeta-probe (Biorad) filters using the Minifold apparatus (Schleicher and Schuell). Filters were then dried at 80°C for two hours in a vacuum oven and hybridized as described for Southern blot hybridization.

In situ hybridization
Tips of colchicine-treated and fixed roots were squashed after treatment with a 5 % aqueous solution of pectinase (Sigma) as already described (CIONINI et al. 1985). Cytological hybridization was performed according to MACGREGOR and MIZUNO (1976). The DNA of the nuclei and chromosomes was denatured in 0.07N NaOH for 3min at room temperature and a highly repeated DNA sequence, which had been 'H-labelled by nick-translation (RrCBY et al. 1977) to a specific activity of 3.5 x 106cpm/ pg, was used as hybridization at a concentration of 1 pg/ml. After incubation, unbound labelled DNA was removed by stringent washings (HEN-NEN et al. 1975) and the preparations were covered with Ilford L, emulsion. After exposure times ranging from one week to three months, the slides were developed and stained with Giemsa (Merck).

Ribosomal DNA
When cleaved with Eco RI, D. villosurn genomic DNA released two fragments of 6.5 kb and 2.5 kb, respectively, that hybridized with 32P-labelled 25s rRNA (Fig. IA). The same DNA produced three fragments of 9.0 kb, 5.1 kb, and 3.9 kb that hybridized to 25s rRNA probe upon cleavage with Barn H (Fig. 1B). After double digestion with Eco RI plus Barn HI, rDNA fragments of 5.0 kb, 3.9 kb, 2.5 kb, 1.5 kb, and 1.1 kb in size were found (Fig. 1C). The 1.5kb fragment was very faint in Fig. 1C because it is an 18s rRNA fragment that cross-hybridizes to the 25s rRNA probe due to common sequence they may share as observed in other higher plants by MIASSOD and CECCHINI (1976). The 1.5 kb fragment gave a strong hybridization signal when the 18s rRNA was used as probe (Figure not shown).
On the basis of these results, and of those obtained after hybridization of the same digests with 32P-18S RNA and 32P-labelled pT A250.4 (not shown), the restriction map of the rDNA repeat of D. oillosum was constructed (Fig. 2). It can be seen that each rDNA repeat has: (i) an average length of 9.0 kb; (ii) two Eco RI restriction sites, one located in the intergenic spacer (ips), the other in the 25s transcribed region, and (iii) one Burn HI restriction site in both 18s and 25s transcribed regions, which divide the rDNA repeat into two fragments (A and B, respectively). The appearance of a band matching the 9.0 kb rDNA fragment indicated that a Bum HI restriction site is methylated in about 50 % of the rDNA repeats. The double digestion with Eco RI plus Bum HI indicated that the methylated Barn HI site must be in the 25s rDNA region.
The restriction fragment pattern observed for D. villosurn rDNA repeats is in accordance with that observed in D. oillosum and other Triticeae species for the position of the Eco RI restriction sites and the length of the igs subrepeats (130 bp) (GILL and APPELS 1989).
However, the occurrence of the methylated Barn HI site was not reported previously.
Southern blots of the Bum HI digested genomic DNA of D. villosurn extracted from single plantlets belonging to different populations followed by hybridization with 32P-pTA 250.4 (Fig. 3), clearly showed that the length of the rDNA fragment A might differ between individuals, due to the variation in number of 130 bp subrepeats in the igs. Fragment A appeared as a single bond in certain individuals (phenotypes A, B, C, and D); or as two bands in other individuals (phenotypes G, H, I, and J). The double-banded phenotypes contained different combinations of fragment A in the singlebanded phenotypes (Fig. 4).
Progeny testing of individuals showing different Bum HI phenotypes (not shown) suggested that the variability for rDNA fragment A was governed by one locus (Nor-Vl), with four codominant alleles indicated as Nor-Vla, Nor-Vlb, Nor-Vlc, and Nor-Vld (Fig. 4). Each allele can be visualized as a cluster of highly homogeneous repeats, all having the same number of subrepeats in the igs. Eight out of ten phenotypes resulting from all the possible combinations of the four alleles have been actually observed (phenotypes A, B, C, D, G, H, I, and J of Fig. 3).  On the basis of these findings, the restriction map reported in Fig. 6 was constructed. About half of the 2270 bp DNA element has an evident repeated structure, since it is composed by three 380 bp units on its left side ending with the Bum HI restriction site; the other half is represented by the Sph I-Pst I 1090 bp fragment.

Chromosomal distribution of the 380 bp repeated unit
Southern blots of partially digested genomic DNAs with Sph I were hybridized, using 380 bp or 1090 bp DNA elements (Fig. 7). Different D. villosum individuals showed difference in the length and abundance of the stretches of subrepeated 380 bp units. The same finding was obtained through the analysis of the presence of the 380 bp element in each D. villosum chromosome, which was possible using CS + DV disomic addition lines (because the 380 bp unit is virtually absent from the wheat genome). The 380 bp unit was absent in the chromosome 7V. In the octoploid amphiploid CS x DV the number of bands hybridizing to the 380 bp unit is lower than those in D. villosum. Using the 1090 bp fragment as a probe, hybridization on the DNA in the lower portion of the Southern blot was similar, though lighter than that observed using the 380 bp unit. This result indicates that the two DNA elements share partial nucleotide sequence homology. Additional hybridization signals (that were absent when the 380 bp unit was used) were observed for the DNA fragments of 1.60 kb and 3.22 kb, and very high molecular weight genomic DNA in both D. villosum and CS. This observation indicates that nucleotide sequences The results obtained in Southern blots of the DNAs from CS + DV disomic addition lines (Fig.   7) were confirmed through cytological hybridization experiments. A metaphase plate after in situ hybridization with 'H-labelled 380 bp DNA sequences is shown in Fig. 8. All chromosome pairs but one are substantially labelled at or near the telomeres. Though the seven chromosome pairs of D. villosum are rather similar to each other in length and shape and the chromosome morphology is unavoidably affected by the hybridization procedure, a careful analysis of a number of metaphase plates allows to state that the nonlabelled chromosome pair was the shorter one of the two subtelocentric pairs. In labelled pairs, silver grains occurred at both the telomeres with the exception of the longer subtelocentric pair: in its longer arm, the telomere was not labelled, whereas silver grains occurred at an intercalary position. Other labelling sites were observed, always near the telomeres, in the three submetacentric pairs, in both arms of the metacentric pair and in the longer, not satellited arm of the satellited pair. In this pair, also the satellite was labelled. Clear labelling over the background was not observed at or near the centromere in any chromosome. Apart from the nonlabelled pair and the centromeric or pericentromeric regions, the distribution of silver grains was similar to that of heterocrhomatin as disclosed after C-banding (DONG FRIEBE et al. 1987). The same labelling pattern over the V chromosome was observed in the hexaploid amphiploid D. villosum x Triticum turgidum var. durum cv 'Modoc' (not shown). No clear localization of labelling was detectable, even after long times of exposure to photographic emulsion, when the 380 bp sequence was hybridized in situ to the chromosomes of rye and wheat.
32P-labelled 1090 bp and 380 bp sequences were used as probes in dot blot hybridizations to genomic DNA from different Triticeae species with genomes closely (T. boeoricum, T. urartu, T. monococcum, Ae. speltoides, T. searsii, Ae. squarrosa, T. araraticum, T. timopheevii, T. dicoccum, T. turgidum var. durum, and T. aestioum) or distantly ( D . villosum and Secale cereale) related to the A, B, and D genomes of cultivated wheats (Fig. 9). It was found that the 380 bp sequence is highly specific to the V genome, since it did not hybridize to the DNA of any species except D. villosum. On the other hand, the 1090 bp element gave hybridization signal on the DNA of all the species tested. The hybridization signal was very strong for D. villosum, as expected, and for T. boeoticum, T . monococcum, T. timopheevii, and T. dicoccum. Unlike the 380 bp subrepeat, the 1090 bp element seems to share nucleotide sequence homology with the wheat ribosomal igs in pTA 250.4.
It is worth noting the different hybridization signal from the DNA of different individuals of Ae. squarrosa or Ae. speltoides. This result suggests that the redundancy of the 1090 bp sequences may change in the genome of different individuals within a species.

Other DNA sequences
Five clones (pA7, pA9, p l l , p17, and p18), randomly selected from our genomic D. villosum DNA library, were tested as potential RFLP probes.
The D . villosum genomic DNA of individual plantlets belonging to different populations was digested with Eco RI, Hind 111, and Taq I restriction enzymes, and hybridized to the "P-labelled probes (Fig. 10).
A simple pattern with one or two bands per lane was observed when the genomic DNA was digested with Eco RI (Fig. 10A). These prominent bands showed polymorphism when the pl 1 probe was used.
The p18 probe revealed polymorphism for several minor bands; the presence of these bands indicates that this probe may have a core sequence related to moderately repeated elements.
The Hind 111 restriction enzyme left the majority of the genomic D. villosum DNA uncut (Fig. 10B). The fragments that hybridized to the pA9 and pl 1 probes were cut and a simple hybridization pattern with two prominent bands were revealed. The probe pl 1 revealed the same type of polymorphism detected in the genomic DNA digested with Eco RI, but the polymorphic fragments were shorter than those detected using Eco RI.  Table 1) from different D. vi//osurn populations.
Hybridization probes were 32P-labelled pA7, pA9, p l l , p17, and p18. Tuq I restriction enzyme (Fig. 1OC) was efficient in cleaving the D. villosum genomic DNA and produced fragments 250 to 1100 bp in length. A simple hybridization pattern with one to three prominent bands was observed. RFLP was revealed by pA7 and p18 probes, and in both cases the hybridizing genomic DNA fragments of an individual (120/2) were different in size from those of the other individuals indicated in Fig. 1OC.
Even though the number of probes and enzymes used in the RFLP analysis was not high, three probes out of five (60 %) detected polymorphism, and every restriction enzyme used produced RFLP that was detected by at least one probe. Four out of 15 (27 'YO) of the probe-restriction enzyme combinations studied revealed polymorphism. The RFLP analysis of the disomic addition lines did not show any additional DNA fragments or fragments which were variable in length when compared to CS (Fig. 11). This situation occurred even though in CS x DV amphiploid there is a release of DNA fragments not present in the parental species. This may be explained recalling that the D. villosum genome used for producing the CS x DV amphiploids was different from that used for producing disomic addition lines. Therefore, the probes used here hybridized with defined DNA fragments of CS, CS + DV discimic addition lines, and D. oillosum; the only difference being the intensity of the hybridization signal. The obscuring effect of DNA fragments in CS that are similar in size and hybridization ability to those of D . villosum, prevented the chromosomal location of the unique or moderately repeated D. villosum fragments inserted in our probes.
As shown in Fig. 12, a significant diversity among CS + DV disomic addition lines was evidenced for the intensity of the hybridization signal when the Hind 111-pA9 combination was used.
The disomic addition line CS + lV, CS + 2V, and CS + 6V showed an autoradiographic signal as strong as that in D. villosum and CS x DV amphiploid for the two hybridizing DNA fragments.
The other disomic addition lines (CS + 4V, CS+ SV, and CS+7V) gave a faint autoradiographic signal as that in CS. As already noted for the p18 probe ( Fig. 1 I), the probe pA9 most likely shares nucleotide sequence homology with repeated elements in DNA fragments that are particularly abundant on certain V chromosomes. This probe may be used to distinguish chromosome lV, 2V, and 6V from the others.  Table I. Molecular sizes are as in Fig. 7.

Conclusions
Our results increase the knowledge on molecular both Eco RI and Hind 111) and p18 (using Tag I), may be helpful in population genetic studies. On the contrary, these probes cannot be used as markers in genetic linkage studies through disomic addition lines.