2~fl08 7107~1) MET~OD OF IDENTIFYING DNA SEQ~CES IN C~ROMOSOMES OF PLANTS
The present invantion relates to a method of identifying DNA
sequences in the nuclear genome of a plant and to a method of plant breeding using this method.
It is often desirable to incorporate DNA sequences into the genome of a plant from another source. Conventional plant breeding techniques have crossed different strains of the same species or even different species in order to introduce genes coding for desirable properties (e.g. disease resistance, protein quality) into the resultant plant variety. The traditional method relied on cultivating large numbers of random hybrids between donor and acceptor species or varieties ant selecting those hybrids which had the desired property for further propagation and inbreeding. This is a slow procedure. The transfer of genes relies on the transfer of whole chromosomes from one plant to another, followed by several generations of selection, often involving backcrossing. During that time, recombination between chromosomes will occur. In many varieties only small chromosome segments are transferred from the donor species into the target plant.
It is now often possible to determine which DNA sequences are responsible for a given characteristic in the plant having that sequence. Recent developments in molecular biology enable more ; direct insertion of DNA segments into new varieties. In all cases,it is helpful to have a technique to follow the incorporation of alien DNA into the genome, without relying only on field testing of plants.
The methods currently used to follow and identify incorporatsd alien DNA which may carry important genes in a target plant are expensive and difficult to interpret. Thus transfer of DNA sequences may be identified by the morphology of the plant, by the morphology of the chromosome itself, by chromosome banding, by determining ths iso-enzymes present, or by RFLP (restriction fragment length polymorphism). All these methods have disadvantages. Another method is to hybridize the chromosome3 in a hybrid plant cell in situ with cloned DNA sequences as disclosed by N Lapitan et al (J Heredity 77, 415-419, 1986). By using clones, the probe DNA
sequences will all be identical. The sequences cloned are selected so as to be characteristic of one of the sources of the DNA in the target plant. However, the clones are difflcult to isolate, and are unlikely to be generally applicable.
It is disclosed by Le et al in an abstract published in connection with the XVIth International Congress of Genetics, Toronto (1988), that total genomic DNA has been used to distinguish chromosomes or large chromosome segments of rye origin from the chromosomes originating from wheat in a rye-wheat hybrid and a wheat ~0 variety carrying a rye interchange. This method relies Gn producing a single stranded version of the chromosomes and bringing them into contact with labelled single stranded DNA isolated from the rye parent. The single stranded chromosomes of the hybrid which are derived from rye will link preferentially to the labelled complementary probe DNA from the rye source.
The metnod disclosed by Le et al (loc. cit.) will identify the origin of whole chromosomes or large chromosome segments in plants where donor and acceptor species differ greatly - for example, in hybrids between parents from different genera of plants. However, in plant breeding it is often desirable to introduce short sequencPs of DNA (typically a few tens of thousands of base pairs), from a related source, into the nuclear genome of a target plant. Thus resistance to eyespot disease in one commercial wheat variety is provided by a DNA sequence from Ae~ilops ventricosa, a wild diploid wheat species. In producing such modîfied plants, whether by sexual hybridization or by more direct laboratory manipulation, there is a need to determine whether a sequence of DNA from an alien source is present in a given plant. This enables a plant breeder to see if a DNA sequence has been been transferred and to follow that DNA
sequence through generations to see if it is changing in length or posi~ion on the chromosome~ and to ex~mine its pairiDg behaviour at meiosis. If the technique described by Le et al. (loc. cit.) is applied to the nuclear genome of a plant whose DNA is derived from two closely related sources, then the total labelled genomic single stranded probe DNA from one source will hybridize (combine to form double stranded DNA) not only with the single stranded nuclear DNA
from the same source but also with the sequences from the other source because of close homologies.
There is therefore a need to find a simple method of identifying DNA sequences from a given source in the nuclear genome of a target plant containing DNA sequences from more than one source which may be close-ly related. In addition, there is a need to find a method which can be applied to different situations, including hybrids and plants carrying alien DNA sequences involving a wide range of species.
According to the present invsntion a method for identifying the origin of one or more sequences in the nuclear genome of a target eukaryotic plant containing genetic material from at least two different sources comprising hybridi~ing DNA from the target plant with labelled total genomic DNA fragments selected to hybridize to DMA from one of the sources, while preventing hybridization of the probe to sequences common to more than one source with unlabelled total genomic DNA fragments selected to block such common sequences by hybridizing with them in the target plant DNA and/or in the labelled DNA, and detecting the sites of hybridization of the labelled probe to the DNA of the target plant.
There are various schemes for classifying living organisms.
The scheme used for the purpose of the present specification is that given on page 33 of "Elementary Microbiology" by 0 ~yss, 0 B Williams and E W ~ardener Jr., John Wiley & Sons Tnc 19S3. In X00~ )8 this classlfication the plant kingdom includes bacteria which are included in a phylum identified as the ProtoPhyta (primitive plants without intracellular membranes). Eukaryotic plants therefore are all tbe members of the plant kingdom except the Protophyta.
The plant is preferably a vascular plant, more preferably one of the Spermatophyta (angiosperms, gymnosperms). The plant is most preferably monocotyledonous, e.g. one of the Gramineae. The target plant may contain DNA from Hordeum, Secale, Triticum, or Ae~ilo~
species. Specific target plants to which the present inveniton may be applied are those containin~ a) DNA from a Hordeum and ~rom a Secale source, b) DNA from a Triticu or a AeRilops source and a Secale source.
-In this specification the "target plant" is the plant whose DNAsequences are under investigation. The tareet plant potentially contains DNA sequences originating from more than one distinct "source" introduced by sexual methods or genetic engineering (i.e.
using a viral vector). The "tarBet DNA sequencesl' are some of the DNA sequences which may be genomes, chromosomes or chromosome segements potentially included in the nuclear genome of the target plant which are to be detected and followed through successive generations.
The "source" of the DNA ion the target plant may ba the parents (e.g. in the first generation hybrid H. chilense x ~k_~E19~e9~) or be more distant ancestors of the target plant (e.g S. cereale and T. durum Desf. ln Triticale, or T. monococcum and others in T
The total genomic DNA used as labelled probe or blocking DNA
can be isolates from the sources of DNA in the target plant. The probe or blocking DNA may also be isolated from "remote sources"
which are taxa related to the sources (e.g. o~ the same family, or the same genus).
At least one, and preferably all sources of the DNA sequences in the target plant should be distinct. One source should preferably have some recognizable characteristic not present in the other sources which results from a difference in its genetic make-up ~)O~flOfil -~e.g. disease reslstance, morphology and also including gene expression factors).
The fragments o~ nuclear genome used to block DNA sequences in common between the labelled probe and more than one of the sources in the target plant will be selected according to the degree of difference between the sources in the target plant. Where the sources are closely related (e.~. H. vulgare and H. bulbosum) it will be preferably to use total genomic DNA from one of the sources as block.
~here the sources are less closely related for example Hordsum vul~are and Secale africanum the total genomic DNA used as a blocking agent may be from a remote source if it is less closely related to the source of labelled DNA than to the other source or sources of DNA in the target plant.
The term "hybridization" as used for the artificial reannealing of single stranded forms of the DNA from plants with single stranded labelled DNA is understood in the art (as shown by Le et al and Lapitan et al (loc. cit.), for review see Henderson, Int.Rev.Cytol.
76, 1-46, 1982). It is, of course, distinct from the hybridization which takes place as a consequence of sexual reproduction in nature. The single stranded DNA from the tarBet plant may be in the form of chromosomes prepared for microscopy. Thus the present invention may be applied to DNA in the target plant which DNA is present as a metaphase or interphase chromosome preparation from the target plant. The DNA may be extracted from the tar8et plant and immobiliæed on a membrane, optionally on restricted areas of the membrane. The DNA from the tar8et plant may be digested ~ith one or more reskriction enæymes and si7.e separated by electrophoresis before transfer to the membrane. Alternatively the DNA may be immobiliæed on the membrane by squashing target plant tissue and transferring the exudate on to the membrane. Methods of producing single stranded DNA and of hybridizing it with labelled DNA probes are generally known and are for example disclosed by Lapitan et al and Henderson (loc. cit.).
The process of the present invention makes use of labelled total genomic DNA as a probe to hybridize to the target DNA
sequences, bloc~sd with unlabelled total ~enomic DNA from other sources. The lengths of the labelled and unlabelled probe DNA
sequences are generally less than that of the total length of the DNA sequence in the chromosome, S0 to 1000 base paira typically.
The length reduction may be achievsd by, for e~ample, the DNA
extraction process itself, autoclaving the DNA, sonication or mechanical shearing. The DNA labelling procèdure may also produce fragments of DNA sequences - oligo-labelling giv~s fr~gments which 10 have a length of 80-120 base pairs.
Total genomic DNA from one of the sources present in the target plant, is labelled and referred to as "labelled DNA". Preferably total genomic DNA ~rom one or more of the sources present in the target plant, but not DNA from the source which supplied the labelled DNA, is used as "blockin~ DNA" and is unlabelled. The DNAs are m~de single-stranded before being placed under conditions which allow hybridiæation. Before and/or during the hybridization, the blocking DN~, which may be present in more than five times the concentration of the labelled DNA is brought in contact with (a) the target plant DNA, (b) the labelled DNA or (a) and (b) combined. In case (a), the labelled sequences can hybridize with those sequences of the target plant DNA which are ~till in single stranded form because they do not correspond to sequences in the blocking DNA. In case (b) many of those sequences of the labelled DNA which are not common to the blocking DNa remain single-stranded, and they are then the only labelled sequences remaining single-stranded and thus available to hybridize to the target plant DNA.
The blocking DNA hybridizes with the sequences of the target plant derived from the same souxce as the blocking DNA, and also sequences of the target plant DNA and/or labelled DNA which are in common with the blocking source DNA. In all cases, hybrid DNA
between unlabelled and unlabelled, between labelled and unlabelled and between labelled and labelled will for~. Because of the higher concentration of unlabelled DNA, the frequency of the latter will be lower, leaving labelled, generally low copy number sequences available to hybridize to corrssponding sequences of the DNA from the target plant.
After hybridization, the labelled DNA is detected in some way as being different from the unlabelled DNA sequences, for example due to the presence of radioactive atoms or other atoms or groups, e.g. biotin or mercury. ~ethods of labelling DNA sequences and detection of labelled DNA after hybridization are well known.
Radioactive lab lling may be used, which may be detected by autoradiography. A preferred method is labelling with biotin, as described by, for example, Lapitan et al (loc. cit.), using nick translation. The presence of biotin labelled material may be detected by fluorescence or colorimetric techniques. An alternative method is to label the DNA with an ~nzyme and to detect the label by a reaction catalysed by the enzyme.
The marking of the potentially present target DNA may be positive or negative. Positive marking would mark the target DNA by the presence of e.8. radioactive atoms or biotin groups after the hybridization. Alternatively, negative marking may distinguish the tarBet DNA by the absence of label.
The invention will now be described by reference to the following examples.
Example 1 and Comparative Test A
Plant DNAs: Total genomic DNA was isolated from leaves o~ ordeum vulgare, H. chilense Roem ~ Schult, Secale cereale, S afri anu Stapf and the hybrids H. vul~are x S. africanum and H. chilense x S.
africanum. Following a standard dot blotting protocol, the DNA was denatured and transferred to a hybridization transfer membrane (GeneScreenPlus membrane, E I du Pont de Nemours & Co Inc, Boston, USA), applying approximately 0.08, 0.2, 0.8 and 2.0 ~g DNA of each ~pecies to different dots of the dot blotter, before alkaline denaturation.
Labelled DNA: Total genomic DNA from S. africanum was labelled by oligo-labelling using biotin~l1-dUTP.
Blockin~ DNA: Autoclaved total genomic DNA from H. vul~are.
The transfer membrane was pre-hybridized with 200 ~g of denatured blocking DNA in 2 ml of a 3tandard hybridization solution (Sharp PJ, Kreis M, Shewry PR, Gale ND. Theor. Appl. Genet.
75:289-290, 1988~. For hybridization, 0.5 ~g denatured labelled DNA
and 1 mg denatured herring carrier DNA were added to the plastic bag and incubated overnight at 70C. Post-hybridization washes were performed with 0.16xSSC (20xSSC: 3N sodium chloride, 0.3M sodium citrate, pH7) at 60C. Hybridization under such conti~ions should allow sequences with 80X homology to form hybrids.
Hybridized biotinylated DNA was visualised (made visible) using ths streptavidin/alkaline phosphatasç colorimetric detection system by Bethesda Research Laboratories (Maryland, USA).
Hybridization to both S. africanum and S. cereale was strongly visible in the dots containing 2.0 and 0.8 ~g of DNA and only weakly to 0.2 ~g DNA. The hybridization to the DNA of the two Hordeum species was only detected in the 2.0 ~g dot. Differentiation between S. africanum and the Hordeum species and detection of the S.
africanum DNA in the hybrids was possible (Example 1), but impo~sible between S. afr anum and S. cereale (Test A).
Example 2 DNA from S. cere~ale, used as blocking DNA, in a hybridization expsriment with S. afri anum as labelled DNA (carrled out as in Example l), allo~ed discrimination between species in one g~nus, S
africanum and S. cereale.
Example 3 and Comparative Test B
A hybridization membrane was prepased and hybridized as described in Example 1. Total genomic DNA from H. vul~are (0.5 ~g) was labelled with biotin by nick-translation, and autoclaved total genomic DNA (lO0 ~g) from H. chilsnse was used as blocking DNA.
The dots containing H. vul~are and H. vul~_re x S. africanum DNA (Example 2) showed more hybridization signal than the dots containing H. chilense and H. chilense x S. africanum DNA (TPst B).
Example 4 and Comparative Tests C and D
Plant DNAs: Total genomic DNA from H. vulgare and H. bulbosum Nevski was digested with the restriction enzymes EcoR~ and DraI~ size separated by 2gaross gel electrophoresis and transferred to Hybond 20~
N~ (~mersham International plc, Amersham, U.K.) support membrane using alkaline transfer methodology.
Lab lled DNA: Tot~l genomic DNA from H. bulbosum.
Blockin~ DNA: Autoclaved total genomic DNA from H. vulgare.
For probe labelling, hybridization and detection of hybridization s~tes, the chemiluminescence method ECL (Amersham) was used following the manufacturer's protocol. Briefly, the transfer membrane was pre-incubated for 30 min in the hybridization solution and denatured blocking DNA ~6 ~g/ml). 12 ng/ml of denatured horse-radish peroxidase labelled DNA was added and hybridization carried out overnight at 42C. Post-hybridization washes were adjusted by sodium ion concentration to allow s~quences with an estimated 90% identity to remain annealed. Hybridization sites were detected by the emission of light directly recorded on film.
Strong signal resulting from hybridization of labelled probe to genomic DNA was observed on the tracks containing DNA rom H.
bulbosum, while little signal was detected on the tracks of ~.
vul~are DNA (Test C). In contrast, when no blocking DNA was used (Test D) substantial crosshybridization to H. vul~are was detected.
The method of hybridization with labelled total genomic DNA together with blocking enables the differentiation of related species within the same genus.
Exam~e 5_and Comparative Test E
Total genomic DNA from H. ch _ense, H. vul~are, S. africanum was digested with EcoRI and treated as in Example 4.
The "Southern" transfer was blocked with 8 ~g/ml total genomic DNA from S. africanum and hybridized with 17 ng/ml labelled total genomic DNA from H. chilense. Post-hybridization washes were carried out at 80X and 90% stringency.
At both stringencies, strong hybridization signal was observed on the DNA track of H. chilense. Cross hybridization to both DNA
from H. vul~are and S. africanum was reduced by the iddition of the unlabelled S. aPricanum DNA (as compared with a similar blot ~.
hybridized without blocking DNA). (Test ~.
i~0~ 8 Exam~le 6 DNA from H. vul~are, used as a blocking DNA~ ln a hybridization experlment (carried out a~ in Example 5) suppressed th~ cross hybridization of labelled DNA ~rom H. chilense to DNA from both S.
africanum and H. vul~are.
Examples 5 and 6 showed that the DNA from a remote source can be used to block the common sequences in hybridization experiments.
Example 7 and C~ g~ E~e~_~
DNAs: DraI restriction enzyme fragments of total genomic DNA from hexaploid triticale (x Triticosecale Wittmark) cv. Lasko, three different cultivars of wheat, (T. aestivum L. cv. Chinese Spring, cv. Beaver and cv. Glenson), and rye, (S. cereale), were treated as in Example 4.
Labelled DNA: 10 ngtml total genomic DNA from S. cereale was labelled as in Example 4.
Blockin~ DNA: 3 ~g/ml unlabelled autoclaved total genomic DNA from T. aestivum cv. Chinese Spring.
The DNA of _. aestivum cv. Chinese Spring (Test F) showed only weak hybridization. Stronger hybridization was detected to DNA of S. cereale, triticale, T. aestivum cv. Beaver and cv. Glennson.
Thus, using total genomic DNA Prom rye as a probe and appropriate blocking, rye DNA can be discriminated in triticale (a hybrid between wheat and rye) and in wheat varieties containing a rye chromosome segment (cv. Beaver and Glennson). Signal quantification showed that the hybridization was approximately proportional to the amount of rye material present.
Example 8 DNAs: Chromosome preparations were made from fixed root tips from H. chilense x S. africanum by placing them in a mixture of 4%
cellulase and 40% liguid pectinase in O.OlM citric acidtsodium citrate buffer (pH 4.o) for 1-2h at 37C. They were subsequently squashed in 45% acetic acid following standard cytological procedures.
Labelled DNA: Total genomic DNA from S. afr_ca~num was labelled by nick~translation using biotin-ll-dUTP.
Blocking DNA: Autoclaved total genomic DNA from H. chilense.
Hybridization was carried out using standard techniques (Schwarzacher T, Leitch AR, Bennett MD, Heslop-Harrison JS.
Ann. Bot. 64:315-324, 1989). The chromosomes wer~ denatured in deionized 70% formamide in 2xSSC for 2 minutes at 68-72C, dehydrated and air dried. Overnight hybridization was carrisd out with 0.1 ~g of denatured biotinylated DNA with 1 ~g of denatured blocking DNA in 2~ ~1 of a solution o~ 502 formamide, lOZ dextran sulphate, 0.1~ sodium dodecyl sulphate and 2xSSC overnight. After hybridization, the slide was washed in 50% formamide in 2xSSC at 42~C. Hybridization under these conditions should occur at 80-85%
Hybridized labelled probe was detected using fluoresceinated avidin, and amplified with biotinylated anti-avidin. Chromatin was counterstained with propidium iodide (1-2 ~g/ml in phosphate buffered saline).
The avidin in situ hybridization signal and propidium iodine were excited at 450-490 nm, the former fluoresces greenish yellow, the latter red. At metaphase the seven larger chromosomes originating from S. afr_canum were distinctly yellow. The seven smaller chromosomes ~rom H. chilense showed no detectable label.
Thus the tschnique clearly separated the two evolutionarily related chromosome sets.
Example 9 Prosphases, telophases and interphases which were -treated as described in ~xample 8 showed distinct domains belonging to either yellow labelled chromosomes from ~S. africanum or red unlabelled chromosomes from H. chilense.
Example 10 Chromosome preparations from root tips of H. vul~are x S.
africanum were made and treated as described in Example 8. In situ hybridization was carried out using 0.05 ~g biotinylated total genomic DNA from S. africanum as labelled DNA and 1.5 ~g of total genomic DNA from H. chilense as block.
The two sets of chromosomes were clearly differentiated as labelled and unlabelled. At metaphase, the seven larger chromosomes from S. aEric~num were labelled and fluoresced brightly yellow, while the ~even smaller chromosomes from H. v~l~are were orange-red showing no detectable label. At prosphase, telophase and interphase, distinct domains of either yellow or red chromosomes were distinguishable.
Example 11 In situ hybridi~ation experiments were performed as described in example 8 using chromosome preparations from the wheat variety Beaver carrying a lB/lR translocation, biotinylated total genomic DNA from S. cereale as a probe and unlabelled total genomic DNA from T. durum. The Texas Red conjugated avidin used here detected the hybridization sites by red fluorescence, while unlabelled chromatin remained invisible. The translocated rye segment oould be identified at both metaphase and interphase. Exact measurements of the breakpoint of the translocation were possible.