AU5698286A - Determination of identity between two organisms - Google Patents

Description

DETERMINATION OF IDENTITY BETWEEN TWO ORGANISMS

TECHNICAL FIELD

The present invention relates to determining to an ascertainable degree of probability whether a first and second organism, the identity of one of which is known, ^* are identical. It is especially applicable to three main categories of organism: (1) those which reproduce principally by asexual means, in particular microorganisms such as bacteria, fungi and viruses, (2) those where it is important to identify a unique individual and (3) inbred populations which are reproducing sexually, such as some plant varieties.

BACKGROUND ART

As an example of uniquely identifying asexually reproducing organisms, one may consider the problem of identifying genetically engineered microorganisms. For a microorganism which has been engineered by the introduction of a plasmid the task of determining whether a particular microorganism is the same is relatively simple. The plasmid itself can be unambiguously identified at the molecular level. However, considerably greater problems of identification are posed in the case of microorganisms which have not been engineered in this way. The difficulties particularly apply to naturally occurring microorganisms and improved strains of such microorganisms which have been developed through mutation in the laboratory or by genetic engineering procedures which involve the organism's chromosome.

It is possible to mark a microorganism by introducing specific mutations or plas ids that are functionally unrelated to the properties that make the organism useful. Plasmids, however, usually reduce the fitness of microorganisms. Consequently, such plasmids are often lost during mass culture procedures. The easiest kind of mutational markers to introduce are mutations to drug resistance. Howewever, these mutations are often themselves disadvantageous. Further, both types of markers can be altered with relative ease.

As an example of identifying a unique organism, we may consider the problem of absolutely identifying a highly valued racing horse, or a stud bull. Classically such animals have been identified by both natural markings such as colour patterns and by deliberate markings such as heat-brands or tattoos. Natural colour patterns are not sufficiently unique for absolute identification, and both brands and tattoos can (and have) been deliberately altered.

DISCLOSURE OF THE INVENTION

A new approach has now been developed to the problem of identifying organisms, utilising unique properties of the organism that cannot be altered. The present invention takes advantage of the natural genetic variation that occurs in a species.

Accordingly, the present invention provides a method of determining to an ascertainable probability whether a first and a second organism, the identity of one of which is known, are identical, which method comprises:

(i) digesting genomic DNA of the first organism with one or more restriction endonuclease; (ii) separating by electrophoresis the DNA fragments thus-obtained; (iii) determining the positions of the fragments thus-separated which hybridise with one or more labelled probes ("hybridisation patterns", "band patterns" or "maps"), the or each probe comprising a fragment of DNA which has been derived randomly from the genomic DNA of an organism of the same species as the first or second organism; and (iv) comparing the positions of the fragments thus-determined with the positions of DNA fragments which bind to the or each said probe, which have been produced from genomic DNA of the second organism by digestion of genomic DNA of the second organism with the or each said restriction endonuclease and which have been subjected to electrophoresis in an identical manner to the DNA fragments obtained from genomic DNA of the first organism;

steps (i.) to (iv) being effected using an amount of probe DNA and one or more restriction endonucleases such that sufficient bands are revealed by the hybridisation in step (iii) to achieve a sufficiently low probability (X) that, when the comparison in step (iv) reveals that the two organisms appear identical, the two organisms will have failed to have been distinguished as genuinely different and unrelated as determined by:-

X = F^q (1)

wherein F is a fraction representative of the proportion of DNA fragments which are identical between restriction endonuclease digests of genomic DNA of pairs of independently-obtained organisms of the same species as the first and second organisms and q is the number of positions revealed by the probing in step (iii) . In other words, q is the total number of common positions of DNA digest fragments revealed by the pairwise comparison of the first and second organisms in step (iv) , when the two organisms have identical maps.

Preferably, steps (i) to (iv) are effected by

(i') digesting separately genomic DNA of the first organism and genomic DNA of the second organism with the same restriction endonuclease and, optionally, dividing each digest into portions; (ii') subjecting to electrophoresis side-by-side on a gel the digest, or a portion of the digest, for each organism;

(iii' )probing the gel using a said labelled probe and comparing the hybridisation patterns for the two organisms thus-revealed; and (iv') optionally repeating steps (ii') and (iii') for one or more further portion of the digest for each organism but using a said labelled probe comprising a different said fragment of DNA each time.

The procedure of steps (i¹) to (iv') can be effected two or more times using a different restriction endonuclease each time.

The present invention depends upon the degree to which the failure to detect a difference between hybridisation patterns for two organisms can be taken as evidence of identity. Where two genomes are digested by a restriction endonuclease, it is possible to estimate the fraction of base pair differences (P) between the two genomes from the fraction of conserved fragments (F) detected (W.B. Upholt, 1977, Nucleic Acid Res. 4_, 1257-1265; M. Nei and W.-H. Li, 1979, Proc. Nat. Acad. Sci. USA. 79, 5269-5273):

wherein n is the number of base pairs in a restriction endonuclease recognition site. This equation can be rearranged to permit estimation of F given an assumed value of P:

F = (l-P)²ⁿ/(2-(l-P)ⁿ) (3)

F is the probability that a fragment will be conserved. Consequently F^q is the probability that, if a digest produces q fragments, all the fragments will be conserved. This may then be used to determine how many different restriction endonucleases and different DNA probes need be employed in the present invention to determine whether two organisms are identical beyond reasonable doubt. Fragments produced when genomic DNA is digested by a restriction endonuclease can be detected by a labelled DNA probe after separation of the genomic DNA fragments by electrophoresis and hybridisation. The greater the amount of probe DNA and the more restriction endonuclease digests examined, the more DNA fragments will be revealed. In other words q increases. As F is a value less than 1, F^q becomes smaller as q increases. Thus the probability becomes less that two organisms, which show identical hybridisation patterns but which are in fact genuinely different, will fail to be distinguished.

The confidence (C) with which two strains can be assumed to be derived from the same clone is given by: C = 1 - F^q (4 )

where F and q are as defined above. In order to estimate C and X, the value of F has to be determined for a particular species of organism. In practice F is estimated experimentally for a number of independently-obtained organisms ("standard organisms") known to be of the same species. Thus using one or more labelled probe, hydridization patterns can be obtained for an appropriate number of non-clonally derived organisms of the same species, for example as determined by conventional taxonomic criteria. Sufficient standard organisms are used to obtain a statistically significant F value for the species, i.e. a F value representative for the species. Preferably 10 or more, for example about 12, such organisms are employed to achieve this. By a pairwise comparison in turn of all the hydridization patterns obtained, F values may be determined from the definition of F, the fraction of conserved fragments:

F = 2N/(a + b) (5)

where N is the total number of common bands between a pair of organisms under comparison and a and b are the respective numbers of fragments detected from all the restriction endonuclease digests and probes of genomic DNA of the two organisms of the pair. The experimental determination of F for a given species of organism enables users of this invention to predetermine the number of genomic DNA digests needed to be undertaken for a given amount of probe DNA in order to obtain a required level of confidence (C) or probability (X) that identity will not be wrongly claimed. Two organisms, therefore, can be analysed to determine the probability that they genuinely differ or the level of confidence that they are identical. If the band patterns are different then the two organisms cannot be identical and cannot have been derived clonally or by asexual reproduction or by sexual reproduction of appropriate highly in-bred organisms. If the band patterns appear to be identical a quantitative probability can be assigned that identity will be wrongfully claimed. Very accurate results can be achieved (see Example 3) .

The present method involves digesting genomic - DNA of the two organisms being analysed with restriction enzymes and separating the DNA fragments in the resulting digests by electrophoresis. The mobilities of the DNA fragments which are detected by a probe can be measured for each organism. Since the same restriction endonuclease has been used and electrophoresis has been conducted in the same way, if the hybridised fragment mobilities (or band patterns) are the same the organisms may be, but not necessarily are, identical. However, where the fragment mobilities are different, the two organisms must be different.

Where the band patterns are the same, it is possible to determine whether two organisms are identical with more certainty by employing more than one restriction endonuclease and/or more than one DNA probe. The greater the number of restriction endonucleases used to digest the genomic DNA and the larger the amount of DNA employed as probes, the lower is the probability that two different organisms will appear to be identical when they are not identical. Once the present method has been carried out a sufficient number of times with different restriction endonucleases and different probes and the band patterns for the two organisms are the same always, then it may be taken that the two organisms are identical beyond reasonable doubt. As the identity of one of the organisms is known, for example by conventional taxonomic criteria, then the other must be the same. Type II restriction endonucleases cut DNA at specific sites that depend upon the DNA sequence at that site. This high specificity means that a particular enzyme cuts (digests) the DNA from a particular organism into a specific set of DNA fragments. A different endonuclease with a different recognition sequence will produce a different set of fragments. The average length of these fragments depends upon the length of DNA required for recognition by a particular restriction endonuclease. For example, restriction endonucleases recognising 6-base sequences generate fragments of an average length of 4096 base pairs (bp) whilst those recognising 4-base sequences generate fragments of an average length of 256 bp. However, the distribution of recognition sites results in a large actual array of fragment lengths. On electrophoresis, the large amount of DNA in a cell combined with the variation in fragment length produces a smear of fragment sizes with little or no recognizable pattern.

However, under certain conditions DNA will bind specifically to DNA with the same sequence to produce a set of patterns unique to a particular organism. If a random fragment of DNA from an organism is cloned into a suitable vector and subsequently labelled, the cloned fragment acts as a probe which is capable of hybridising with DNA that has the same sequence. The fact that the probe is labelled enables its position to be detected. The comparison of two organisms according to steps (i) to (iv) of the method of the invention can be illustrated as follows. A random fragment of DNA from a first strain of an organism or a specific individual is used to make a labelled probe. The total genomic DNA from the first strain (or individual) is digested with a restriction endonuclease. The DNA fragments are separated by electrophoresis on a gel. If visualised, the digested DNA would be seen only as a smear. The digested DNA is hybridised with the probe. The probe will bind only to those fragments that show a certain degree of homology with the sequences in the probe. As a result, instead of a smear several bands can be detected by means of the label on the probe at the sites of hybridisation of homologous DNA sequences. The positions of these bands are determined by the fragment sizes. The total genomic DNA from a second strain or individual is digested by the restriction endonuclease and subjected to electrophoresis in an identical manner, and hybridised with the same probe. If any of the restriction sites in the hybridised DNA of the second strain or individual are different, caused by past mutational events, from those in the DNA of the first strain or individual the positions of the bands will be different, since digestion of the genomic DNA will produce fragments of different sizes due to the position of different restriction sites. If the positions are the same, the probe will have hybridised with identical DNA fragments and appear in identical positions in the gel. The strains or two individuals may therefore be the same. Further tests employing different restriction endonucleases and/or different probes will establish beyond reasonable doubt whether there is identity between the two strains or individuals.

The genomic DNA in step (i) may be digested separately with one or more restriction endonucleases. Respective portions of the DNA sample can be treated with different restriction endonucleases. Typically, the sample of genomic DNA may be divided into five portions each of which is treated with a different restriction endonuclease. Any suitable restriction endonuclease may be employed in the present invention. Restriction 5 endonucleases recognising 4-, 5- or 6-base sequences on the genomic DNA can be employed. Suitable endonucleases are Sau3A, PstI, Ba HI, XhoII, Hindlll, EcoRl, Neil, Bgll, Accl, Sail, Avail, Asp700 and Clal.

The DNA fragments obtained on digestion with a lOrestriction endonuclease are separated on a gel by electrophoresis in step (ii) . The DNA fragments derived from the organisms being compared with one another are subjected to electrophoresis in an identical manner, for example at the same time under the same conditions. This 5 is preferably achieved by subjecting the fragments to electrophoresis side-by-side on the same gel. Any suitable method of electrophoresis may be adopted. For example, the DNA fragments may be run on an agarose gel. The probe employed in step (iii) comprises a 0 randomly derived fragment of DNA. The fact that the probe DNA is derived randomly is important. The DNA for the probe is not specifically selected from functional genes which are under strong selection for conservation and consequently which will tolerate relatively few base 5 substitutions as mutations. It is mutations which alter the base sequence of a recognition site for a restriction endonuclease that provide the variability which is detected by this invention. Genes which are under pressure for selection because of their genetic functions 0 are likely to contain fewer variations. Consequently, such genes may be conserved between two different strains of an organism, particularly between two closely related organisms, and are not representative of the natural genetic variation between the two strains. On the other hand, by randomly choosing DNA for use as a probe in the present invention, the probe is likely to contain genes which are not under a strong pressure for selection. Consequently, these genes can tolerate many more mutations, are considerably less likely to be retained from strain to strain of the species in question and are representative of the natural genetic variation of individual strains.

The DNA for a probe can be produced by digesting, preferably partially digesting, the genomic DNA of a strain to be identified or a strain from the same species with a restriction endonuclease. The DNA fragments thus obtained can then be introduced into suitable vector plasmids to generate a bank of potential probes. Preferably the plasmids are screened and a suitable number, for example five, that contain different sequences of the original^' genomic DNA are chosen. Typically, the length of DNA fragment in the plasmid" is about 5 Kb or more, preferably 10 Kb or more. The length of the DNA fragment may vary depending upon, for example, the length of DNA that can comfortably be inserted in a particular plasmid. Further, the larger the amount of DNA in a probe, and the greater the number of probes employed, the greater the probability that two different strains will be distinguished. Where several probes are employed, the total amount of DNA may typically be about 30 Kb or more, for example about 50 Kb (for example five fragments of about 10 Kb each) . Any vector system can be used in this invention, another suitable example being cosmids.

The organism from the genomic DNA of which the DNA fragment incorporated in the probe is randomly derived is of the same species as an organism to which the_^ present invention is being applied. Consequently, where a first organism is being compared with a second organism, then the DNA fragment in the probe must have been derived from the genomic DNA an organism of the same species as the first or second organism and preferably has been derived from genomic DNA of the first or second organism or of one of the standard organisms employed in an experimental determination of an F value.

Each probe is labelled. For example, radioactively labelled probes may be prepared by nick translating with [α-³²P]dATP or [ -³²P]dCTP.

Alternatively, permanent biotinylated probes^'can be prepared by nick translating with biotinylated dUTP. The advantage of biotinylated probes is that they can be stored for future use whereas radioactive probes, generally, must be used within a short time of preparation.

After electrophoresis, the positions ^'of the DNA fragments which bind to a probe can be determined in any suitable manner. For example, the DNA fragments on each gel following electrophoresis may be transferred by the capillary method of Southern (E. Southern, 1975, J. Mol. Biol. S>8_, 503-517) to nitrocellulose membranes or to diazotised aminobenzyloxy ethyl (DBM) paper, or nylon membrane (e.g. Amersham pic's Hybond N) . Alternatively, DNA fragments may be more rapidly transferred by transverse electrophoresis to charge-modified nylon membranes such as Zeta Probe of Biorad Laboratories Ltd.

The DNA fragments are then contacted with a labelled probe. The position of any DNA fragments which hybridise with the probe is then detected by suitable means. A pattern of these positions can then be produced and compared with a pattern produced for a second organism or-, indeed, further organisms. Several such band patterns can be plotted for each organism using different restriction endonucleases and different probes. Where any of the patterns for two organisms are different, it follows that the two organisms are themselves different. Where a sufficient number of patterns have been compared and all are the same, then it can reasonably be assumed that the two organisms are identical. Indeed, the probability that two organisms will fail to be distinguished can be ascertained (equation (1)) . By using a sufficient amount of probe DNA and a sufficient number of restriction endonucleases (which may be 1) , a probability that two organisms will fail to be

-12 -15 distinguished of 10 or less, for example 10 or less,

10^" or less or even 10^" or less, may be achieved. A F value for a species of an organism may be determined experimentally by:

(a) digesting separately using the same restriction endonuclease genomic DNA of a number of independently-obtained organisms of the same species as the said first and second organisms sufficient to obtain a F value representative of the species and, optionally, dividing each digest into portions;

(b) subjecting to electrophoresis side-by-side on a gel the digest, or a portion of the digest, for each of the independently-obtained organisms;

(c) probing the gel using a labelled probe comprising a fragment of DNA derived randomly from the genomic DNA of an organism of the said species;

(d) comparing the hybridisation patterns on the gel thus revealed for pairwise combinations, preferably each and every pairwise combination, of the independently-obtained organisms; and

(e) optionally repeating steps (b) to (d) for one or more further portion of the digest for each of the independently-obtained organisms but using a said labelled probe comprising a different said fragment of DNA each time.

The procedure of steps (a) to (e) can be effected two or more times using a different restriction endonuclease each time.

Preferably the same set of restriction endonucleases as used in steps (i) and (iv) will be used in determining the species F value. However, this need not be the case and different sets of restriction endonucleases may be used in step (a) .

The organisms used in step (a) , the standard organisms, must be different strains of the same species. They must be independently-obtained, i.e. non-clonally derived. They must belong to the same species as the two organisms which it is wished to analyse for identity. Indeed, one of those two organisms preferably is included in the set of standard organisms. A sufficient number of standard organisms should be employed so that the F value that is obtained is statistically significant and can be considered as representative for the species in question.

Restriction endonuclease digests are prepared from samples of the genomic DNA of the standard organisms. The restriction endonucleases mentioned previously may be employed. The genomic DNA of each organism is digested with the same restriction endonuclease(s) . A genomic DNA sample of each organism is treated with a single restriction endonuclease. A panel of samples each digested by different restriction endonuclease can thus be built up for an organism.

The digests are subjected to electrophoresis to separate the DNA fragments in them. This may be achieved as described above. Digests for the same restriction endonuclease(s) are subjected to electrophoresis under identical conditions. Preferably, this is achieved by running them side-by-side on the same gel.

The or each labelled probe used in step (c) preferably comprises a fragment(s) of genomic DNA from one of the standard organisms or one of the two organisms under test. The probes may be constructed as described above. The same criteria apply. If more than one probe is used, the probes must contain different DNA fragments. The probes hybridise with DNA fragments amongst those which have been subjected to electrophoresis which possess DNA sequences homologous to those of the probe. For a given restriction endonuclease digest, a hybridisation pattern is therefore revealed for each of the standard organisms. This permits a pairwise comparison of the patterns to be undertaken. For a given restriction endonuclease digest, the hybridisation patterns of each and every possible pairwise combination of the standard organisms are examined to determine the fraction of the hybridised DNA fragments which are revealed by the probing which are identical for each pair. This fraction is the F value for that pair: see equation (5) above. The F values for all the pairs may be used to deduce the weighted average F value for that particular restriction endonuclease digest. A mean value of F for all the restriction endonuclease digests can be ontained from a weighted average of the F values of individual restriction endonuclease digests.

Once a F value has been obtained which may be considered representative of a particular species, that value may be used to ascertain the probability that two different organisms of the same species, which appear identical when analysed according to steps (i) to (iv) of the present method, in fact have failed to have been distinguished as different strains. Suppose a representative average F value (F ) has been obtained for the species to which the two organisms belong using a particular restriction endonuclease and a particular set of one or more probes. Using the same restriction endonuclease and the same probe set, if comparison of the two organisms according to steps (i) to (iv) reveals that they appear identical then the probability that they will have failed to have been distinguished is:

_1 (number of identical bands revealed by the probe set) ^Fav

i.e. X = F^q (equation (1))

As Fav is a number less than 1, the more bands revealed the lower will be the probability that the two organisms, if different, will have failed to have been distinguished. The number of bands examined can be increased by increasing the amount of probe DNA. It can also be increased by repeating steps (i) to (iv) for further restriction endonucleases. A mean value of F for all the restriction endonucleases can be obtained from a weighted average of the F values of the individual restriction enzyme digests and the total number of bands in the hybridisation patterns obtained for the restriction endonucleases summed (or pooled) so as to obtain a yet more accurate prediction of whether the two organisms will have failed to have been distinguished. That is, the greater the value of q the less will the probability be that erroneous identity is claimed.

It will of course be appreciated that when a F value is determined according to steps (a) to (e) above then: - the or each restriction endonuclease which is employed in step (i) need not be the same as the or each restriction endonuclease employed in the determination of the F value; however, preferably the same restriction endonuclease(s) will be used in both cases.

- the or each probe employed in steps (iii) and (iv) need not be the same as the or each probe used in the determination of the F value; however, preferably probe(s) will be used in both cases. - where average F values for particular restriction endonucleases are themselves averaged then the same standard organisms must have been employed in the determination of the average F value for each restriction endonuclease.

The invention can be applied to any organism. It will be appreciated that the term "organism" applies to a population of individuals which are clonally derived, e.g. microorganisms such as bacteria, fungi and viruses. The conventional taxonomic identity of one of the first and second organisms being compared must be known. As a minimum, this means that the species of one of the two organisms must be known so that an appropriate F value can be chosen/determined.

The invention is particularly applicable to organisms that reproduce by asexual means, for example microorganisms such as bacteria, fungi and DNA viruses. The invention can be applied to such microorganisms which are naturally occurring or are mutants which have been derived in the laboratory or, indeed, have been engineered _in_ vitro by the insertion of a plasmid or other genetic manipulations to genomic DNA. In the case of RNA viruses, cDNA obtained by use of a reverse transcriptase can be employed as the genomic DNA. The invention is also applicable to inbred populations which are reproducing sexually, such as some plant varieties, and to organisms where it is important to identify a unique individual. In order to effect absolute identification of a unique individual organism, a tissue sample may be obtained from the individual either shortly after birth, or at any subsequent time. Genomic DNA is prepared from that tissue sample and stored until such time as a question concerning identity arose. At such a time DNA is prepared from the individual in question, and that DNA is compared with the original DNA sample. In this case, the two DNA samples are compared by digestion with a series of restriction enzymes and hybridization to a set of probes containing random DNA fragments from an organism of the same species. Identity of the patterns of the two samples can be employed to establish the probability that the individual in question is the same as the individual from which the original DNA sample was taken. In this way the invention may be applied^* to humans and to animals, for example horses, especially race horses, and cattle, particularly stud bulls. This invention may also be employed by plant breeders.

The present invention therefore has wide application. If a strain of organism is thought to be identical to a particular strain this can quickly and unequivocally be established employing the present invention. A pattern for each organism can be built up employing different restriction endonucleases and different probes in the present invention. Where the patterns are the same, the probability that the two organisms will have failed to have been distinguished can be determined.

The present invention can also be employed to permit clinical identification of an unknown pathogen. This may be achieved by preparing a series of probes incorporating randomly derived fragments of the genomic DNA of the pathogen. The genomic DNA of the pathogen and the genomic DNA of various microorganisms to which it is thought to be identical can be digested with restriction endonucleases and subjected to electrophoresis under identical conditions in accordance with the invention. Using the already prepared probes, band patterns of the pathogen and of the organisms against which it is being tested can then be built up and compared for identity. Subsequently after determining the F value the confidence of correctly identifying the pathogen may be stated.

The present method may be augmented by subjecting cellular enzymes of the first and second organisms to electrophoresis to compare the electrophoretic mobilities of the enzymes. Differences in electrophoretic mobility indicate that the two organisms are different. However, if the electrophoretic mobility of enzymes of the two organisms is the same then the two organisms may be identical. The more enzymes that are tested the greater the chance of detecting whether two organisms are indeed identical. However, this means of establishing identity between two organisms is not as powerful as the restriction endonuclease analysis of the genomic DNA of the two organisms.

The electrophoretic mobility of enzymes of two organisms may be compared quite simply. A crude cell extract of each organism is subjected to an electric field in a gel, for example a starch or polyacrylamide gel.

After electrophoresis, each gel is immersed in a solution containing a substrate for the enzyme in question. Degradation of the substrate results in the appearance of a band at the position of the enzyme in the gel. If the enzymes from two organisms differ by one or more charged amino acids they are likely to migrate to different positions in the gel and a different band pattern results. A further technique for supplementing the present method of identifying an organism is to compare the respective protein patterns in gels of two organisms. Proteins can be separated by charge by isoelectric focussing. Proteins can also be separated by size, for example on SDS gels. These two techniques can be combined to provide a two dimensional separation of proteins on a single gel. The proteins appear on the gel as a pattern of spots. Where identical patterns of spots are produced for two organisms, the organisms may be identical. However, if the patterns are different then the two organisms are different.

BRIEF DESCRIPTION OF DRAWINGS

The three figures show band patterns produced in

Example 3. In each case, one of the lanes (labelled λ/p BR322) contained as a standard lamda DNA (which had been digested with restriction enzymes EcoRI and Hindlll) and pBR322 DNA (which had been digested with restriction enzyme Sau 3A) .

Figure 1 is a band pattern revealed for genomic DNA of 12 Lactobacillus plantarum strains, and lamda DNA, digested with restriction enzyme Accl, bound to nylon filters (Southern blot) and probed with nick-translated

[ 32P]-labelled plasmid pBTL30. The filter was autoradiographed at -70°C for 1 day to produce the X-ray film shown. The bands correspond to positions of genomic fragment and probe DNA hybridisation.

Figure 2 is a band pattern produced as for Figure 1 except that the restriction enzyme used was Asp700 and the probe was pBTL29. Figure 3 is an X-ray autorad iograph o f a g el o f type C (Section 8 of Example 3) , that is DNA d igested with r estr iction enzyme Asp700 , bound to nylon f i l ters

(Southern blot) and probed with nick-translated [ 32P]-labelled bacteriophage lamda DNA. A second control lane containing lamda DNA digested with Asp 700 was provided as a reference standard. This gel was used to assess whether or not the DNA digestion had been completed. MODES, INCLUDING THE BEST MODE, FOR CARRYING OUT THE INVENTION

EXAMPLE 1 : General Procedure (a)

1. Prepare genomic DNA from a strain A. Do partial digests of the genomic DNA with Sau3A restriction enzyme to give fragments of an average size of about 10 kbp. Ligate these fragments into the BamHI site of the plasmid pBR322. Transform a suitable host, selecting Ap transformants, and isolate 10 to 20 Tc^s transformants.

2. Identify five reco binant plasmids with different inserts of 10 kbp or greater.

3. Prepare larger amounts of these plasmids and make permanent biotinylated probes by nick translating with biotinylated dUTP. These probes can be stored for future use.

4. Digest ten portions of genomic DNA from strain A separately with a respective restriction endonuclease, five endonucleases having 6-base recognition sequences and five having 4-base recognition sites. The choice of enzymes should be dictated by the degree to which the enzyme is known to have randomly distributed recognition sites. Similarly, digest genomic DNA from a strain B which is to be compared with strain A. For each of the ten digests, carry out electrophoresis by running - the digest for strain A in a lane next to the digest for strain B on an agarose gel. All twenty digests 5 can be run on a single gel. Run five such identical gels.

5. For each gel do a Southern blot transfer of the DNA to a nitrocellulose filter or to DBM paper.

106. Probe each of the blots with a different one of the five probes. Compare the pattern of the hybridisations for strain A with that of the hybridisations for strain B. If the patterns are identical, the probability that strains A and B will

15 have failed to have been distinguished can be determined from a statistically significant F value for the species to which strains A and B belong and the number of bands revealed in the hybridisation patterns.

20 EXAMPLE 2: General Procedure (b)

General procedure (a) of Example 1 may be varied as follows:

1. Prepare genomic DNA from strain A. Do partial digest of the genomic DNA with restriction enzyme Pst I to

25 give fragments of an average size of about 10 kb. The fragments are ligated with the Pst I site of plasmid pBR322. The ligated plasmid now containing a piece of strain A DNA is transformed into a suitable host, selecting for Tc^r transformants and isolate

30 10-20 independent Ap^s transformants.

2. Identify the required number of recombinant plasmids with different strain A DNA inserts and size these inserts. 3. Prepare large amounts of these plasmids and either make permanent biotinylated probes or [ 32P]-labelled probes by nick translation. Biotinylated DNA probes can be stored for future use. [ 32P]-labelled DNA probes have to be used within days of preparation.

EXAMPLE 3

The following Example illustrates the practical operation of the invention by use of the organism Lactobacillus plantaru .

l. The microorganisms used

Twelve independently isolated strains of Lactobacillus plantarum were obtained as different representatives of the total L. plantarum population which were not clonally derived. Eleven of the strains were obtained from the National Collection of Industrial Bacteria (NCIB) and the twelfth from BioTechnica Limite 's collection. The latter had been independently analysed by NCIB and confirmed as a member of the species by conventional taxonomic criteria. The NCIB strains were 5914, 6105, 6376, 6461, 7220, 8016, 8026, 8102, 8299, 8531 and 11974. The twelfth strain, referred to here as BTLS1, was deposited on 25th September 1985 at NCIB, Torry Research Station, P.O. Box 31, 135 Abbey Road, Aberdeen AB9 8DG, GB. under accession number NCIB 12156. In this Example, the strains were used:

- to demonstrate that because of random mutations becoming fixed within the population (at different times and in different places) the various strains had different hybridization patterns as a result of restriction enzyme fragment polymorphisms; - to estimate the F value - the fraction of restriction endonuclease fragments which are conserved (i.e. are identical between the probed digests of two genomic DNA samples taken, separately, from two organisms) on a pairwise basis; and

- to estimate the probability that two different isolates would fail to be distinguished (F^q) .

For this analysis two other microorganisms were required for appropriate steps in the procedure.

These were E. coli strain SJ84R/pBR322 (as a source of plasmid pBR322 for constructing the probes) and E. coli RL57 (as the organism which could be transformed by the ligated probes and be analysed) .

Growth of microorganisms

A method was required to grow adequate quantities of the organisms to obtain DNA. The exact method does not matter so long as adequate quantities of DNA may be obtained. In this case the L. plantarum strains were separately grown on one quarter strength MRS broth (Oxoid) at a volume of 200 ml overnight at 30°C without any agitation. Exceptions were strain NCIB 8299, which was grown in 400 ml of one eighth strength MRS broth, and strain BTLSl, which was grown in 100 ml normal strength MRS broth. For strain E. coli SJ84R/pBR322 a picillin at 100 jig/ml was included in the broth to ensure

« maintenance of the plasmid. 3. Isolation of chromosomal DNA

The genomic DNA was isolated as follows. Organisms were harvested (7000 rpm, 10 min) , washed once in cold TES buffer (30 mM-Tris HCl, pH8.0; 5mM EDTA and 50mM-NaCl) and the pellets resuspended in 1.5 ml 0.1M-EDTA, 0.15M-NaCl with 100 μl of a solution of 70 mg/ml pronase E (Sigma, Registered Trade Mark) and incubated at 37°C for lh. 40 mg dry lyzozyme (Sigma) and 3.5 ml butanol buffer (13mM-Na-,HP0₄ ; 30mM-KH₂PO₄ ; ImM-EDTA; 65 (v/v) butanol) was added and incubated at 37°C for 20 minutes. Then 1 ml 20% (w/w) SDS was added and the lysate was frozen by placing on a dry ice/ethanol mixture and thawed at 65°C. The process was repeated three times.

1 ml 5M-NaC10. (freshly made) was added to the lysate and mixed. The mixture was extracted with 6 ml CHC1₃ - isoamyl alcohol (24:1, v/v) and placed on ice for 30 minutes before clearing the phases by centrifugation. The aqueous phase was removed with a wide bore pipette and the DNA spooled off with 2 vol of -20°C ethanol. The DNA was redissolved in 1 ml TE buffer (lOmM-Tris HCl, pH 8.0 and ImM-EDTA) with 200 units TI RNase (Boeringer) at 37°C for 1 hour. lOul 25 mg/ml protinase K (Boeringer) was added and incubated at 37°C for one hour.

The DNA was extracted twice with CHCl^, on ice, the aqueous phase was removed and 0.1 volume of 3M-sodium acetate, pH 5.2, added, followed by O.βmM isopropanol (room temperature). Precipitated DNA was spooled off. The spooled DNA was dipped in cold 70, 80, 90 and 100% ethanol sequentially (to remove the acetate) and dissolved in 500 μl TE buffer and subsequently dialysed against TE buffer. The purified DNA was quantified and ready for use in probe construction or digestion as appropriate. This method was based on that described by Marmur (1961, J. Mol. Biol. 3_, 208-218).

4. Isolation of plasmid DNA

For small scale preparation (so-called minipreps) the boiling method of Maniatis e_t al_^ (1982) (Molecular Cloning: a Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, U.S.A.) was used, followed by two isopropanol precipitations at room temperature. For large scale preparations the method of Hansen and Olsen (1978, J. Bact. 135, 227-238) was used.

5. Restriction enzymes The restriction enzymes used for constructing the probes or digesting genomic DNA were obtained from Bethesda Research Laboratories (BRL) or Boeringher Mannheim (BM) or Pharmacia. All were used in accordance with the manufacturers' instructions for use.

6. DNA gel electrophoresis

Horizontal agarose gel electrophoresis for separating DNA fragments was carried out in BRL apparatus in Tris-borate EDTA buffer as described by Maniatis et al_^ (1982, above) . Ethidium bromide (5ug/ml) was included in the agarose gel and electrophoresis buffer. DNA bands were visualized under medium wavelength U.V. and photographed as described by Maniatis e_t al. (1982, above) . 7. Construction of probes

Genomic DNA from Lactobacillus plantarum strain BTLSl was partially digested with restriction enzyme Pstl to give a majority of fragments in the size range 5-15 Kb DNA. The Pstl fragments were ligated with Pstl digested plasmid pBR322 DNA. Plasmid DNA was treated with 5 units of calf intestinal alkaline phosphatase for 60 min at 37°C in 0.OlM-Tris-HCl, pH 9.0 buffer containing 1.0mM-ZnCl₂ and l.OmM-MgCl,. This preliminary treatment prevented self ligation and contributed to a successful ligation between plasmid pBR322 DNA and the genomic DNA of Lactobacillus plantarum strain BTLSl. The ligation step was carried out using T4 ligase from BRL according to the specified instructions for use. The ligation mixture, following ligation of genomic and plasmid DNA, was used to transform E. coli strain RL57.

The transformation step followed the method of Maniatis et al (1982, above) with the exception that lOOmM CaCl, was used. Transformed E. coli strain RL57 cells were selected by plating on solidified L-broth containing 20 μg/ml tetracycline (Tc) . This ensured that only those cells now containing introduced pBR322 DNA grew. The independently obtained Tc^r cells were checked for ampicillin resistance (Ap^r) and sensitivity (Ap ) by plating on solidified L-broth with ampicillin at 50 μg/ml. Isolates which showed Ap were retained since these strains would have L. plantarum DNA ligated into the Pstl site which occurs within the ampicillin resistance gene of plasmid pBR322.

Mini-preps of plasmid DNA from Tc^r Ap^s strains of transformed E. coli RL57 were made and digested with restriction enzyme Pstl. The digested DNA was subjected to electrophoresis in 0.7% (w/v) agarose gels to estimate the size of the Lactobacillus plantarum DNA which had been ligated into the plasmid pBR322 DNA. From this survey of transformed E. coli RL57 isolates, four plasmids carrying inserted L. plantarum DNA were identified. They were designated pBTL8, pBTL23, pBTL29 and pBTL30. The sizes of the DNA inserted were estimated by comparison with a standard DNA digest with fragments of known size, namely bacteriophage lambda DNA digested with Hindlll and EcoRI restriction endonucleases. The following sizes were obtained: for pBTL8, 8.8 Kb of L. plantarum DNA had been inserted; for pBTL23, 8.0 Kb DNA; for ρBTL29, 5.7 Kb DNA; and for pBTL30, 10.4 Kb DNA. These four plasmids constituted the probes to be used for examining the genomic DNA of all the Lactobacillus plantarum strains.

In order to use the probes for genomic DNA investigations the probes needed to be labelled either with biotin-11-dUTP (for "biotinylated" probes) or with deoxycytidine 5'-[o- 32P]-triphosphate (herein referred to as [ 32P]-dCTP (for "radioactively labelled" probes) ) . In this Example [ 32P]-dCTP was used to produce radioactive probes using the nick translation procedure and kit supplied by Amersham International pic. The labelled probes were separated from unincorporated [ 32P]-dCTP by passage down a Sephadex (Registered Trade Mark) G-50 column previously equilibrated with 50mM-Tris HCl, ImM-EDTA and 0.1% (w/v) SDS. The leading radioactive peak (detected by a Gieger Muller monitor) was collected, and contained the [³²P]-labelled probes. The four plasmid probes contain pBR322 DNA as well as the cloned L. plantarum DNA. It was shown that nick translated [³²P]-labelled pBR322 DNA did not hybridise significantly with digested L. plantarum DNA under the conditions described here. Thus any hybridisation observed are due to genomic L. plantarum DNA hybridisation with probe L. plantarum DNA. 8. Examination of L. plantarum DNA Samples of genomic DNA (12.5 μg) containing dam-lambda bacteriophage DNA (150 μg) (obtained from BM) were digested with 30-250 units of restriction endonuclease, the amount depending on the enzyme used. The DNA samples were digested with a range of restriction enzymes separately. For this example, the following restriction enzymes were used: Neil, Bgll, AccI, Sail, Avail, Asp700 and Clal. After digestion with an enzyme each DNA sample was divided into three aliquots and electrophoresed on three 0.9% (w/v) agarose gels designated A, B and C. 5 μg of digested DNA was loaded per lane on gels A and B whilst 2.5 ug was loaded per lane on gel C. On gel C, in addition to the lanes loaded with sample DNA, one track was loaded with lambda DNA alone which had been digested with the same restriction enzyme as was used for the genomic DNA. For all three gels, one of the lanes contained lamda DNA (which had been digested with restriction enzymes EcoRI and Hindlll) and pBR322 DNA (which had been digested with restriction enzyme Sau3A) . The DNA fragments in this lane were used as molecular weight markers. All the gels were photographed immediately after electrophoresis and staining to check for the degree of DNA digestion and to provide a future calibration for the sizes of the fragments produced. After the agarose gels had been electrophoresed using standard procedures. the procedure was modified to produce a double filter blot for each of gels A and B (using the arrangement shown on page 385 of Maniatis e_t al_^ (1982, above) and a single filter blot for gel C. This procedure produced five separate blots, each blot containing the equivalent of approximately 2.5 μg digested DNA per lane bound to the filter. The five replicate filter blots were separately hybridised with the four pBTL probes and a probe of nick translated lambda DNA according to the procedure supplied by Amersham pic. The hybridisation of the nick translated probes with the digested DNA on the filters was established using standard procedures. Filters were hybridised and washed (stringently) in 6xSSC buffer (SSC buffer = 0.15M-NaCl and 0.015M-Na₃ Citrate, pH 7.0) and lxSSC buffer respectively at 65°C. Probes were denatured by boiling just prior to being incubated with the filters.

9. Autoradiography

The probed and hybridized filters, once dried, were wrapped in Saran Wrap. Filters were then autoradiographed for between 4 hours and 2 days at -70°C using fast tungstate intensifying screens.

10. Band Patterns

The band patterns produced are exemplified by Figures 1 to 3 of the accompanying drawings. The bacteriophage lambda probe was used to monitor the completeness of the genomic DNA restriction enzyme digests. From a complete lambda DNA digest a known number of restriction fragments will be generated when hybridised with nick translated [ 32P]-labelled lambda DNA. The appearance of additional restriction fragments after autoradiography indicates an incomplete or partial digestion of lambda DNA and hence an incomplete digestion of L. plantarum genomic DNA. That is gel C is used as a control to assess that the samples have been completely digested. Those lanes which showed incomplete digestion were ignored.

The X-ray films (or autoradiographs) produced from the filters hybridised with [32P]-labelled DNA probes showed a series of bands corresponding to fragments of DNA from the DNA samples which hybridised with the probes. Hybridising fragments were distinguished on the basis of their relative mobilities in the agarose gel. Thus, the largest hybridising fragments gave rise to bands on the autoradiograph which were closest to the gel's origin (as marked on the corresponding autoradiograph) .

Conversely, the smallest hybridising fragments had the greatest mobility and, therefore, gave rise to bands on the autoradiograph which were furthest away from the gel's origin. The positions of the bands relative to the gel's origin were recorded for each autoradiograph (each autoradiograph being obtained from 12 L. plantarum DNA samples digested with the restriction endonuclease, electrophoresed on one gel, bound to a Filter-Southern blotting - and hybridised with one probe). Therefore, the presence of an identical hybridising DNA fragment in more than one DNA sample was inferred by the presence of bands which moved the same distance from the gel's origin on the corresponding autoradiograph. Tables 2.1-2.4 and 3.1-3.4 below show the bands present for genomic DNA digestions with restriction enzymes Bgll and Aval respectively. In each probe the total number of bands visualized, as a result of hydridisation of the labelled probe with digested DNA, were recorded. The total number of bands for each set of probes for the Bgll and the Avail digests are shown in Table 1. For each plasmid probe and each restriction endonuclease digest, pairwise comparisons of the band patterns determined the number of common fragments (2N) between each pair of digests with a total of a and b fragments respectively. Tables 2.5-2.8 and 3.5-3.8 below show the pairwise comparisons for genomic DNA digestion with Bgll and Avail respectively. From the individual pairwise comparisons (i.e. Tables 2.5-2.8 and 3.5-3.8) a set of six tables (only two - Tables 2.9 and 3.9 - shown here) showing the pairwise comparison for the total probed DNA were constructed. Tables 2.9 and 3.9 show the summed pairwise comparison for genomic DNA digested with Bgll and Avail, respectively. Finally, Table 4 shows the complete pairwise comparisons between all the strains for the complete set of four probes and six restriction endonuclease digests.

The fraction of^* conserved fragments (F) for each pairwise comparison can be calculated (according to equation (5)) using the data in Tables 2.9 and 3.9 (for a pairwise determination of F for the complete set of four probes using only one genomic DNA digestion) and in Table 4 (for a complete pairwise determination of F for the total set of four probes and six genomic DNA digests) . These values of F are shown in Tables 5, 6 and 7 respectively. As expected since the twelve strains were independently isolated the values for F varied, showing that the degree of relatedness varied. In some instances, at the lower level of genomic DNA analysis (i.e. single restriction endonuclease digests) some pairwise comparisons showed apparently no common DNA fragments (for example strains 6461 v BTLSl; 8102 v BTLSl) when the four probes were used against genomic DNA digestion with restriction endonuclease Avail (Table 6) . A wider examination achieved by using six restriction endonucleases showed a few common bands (Table 4) and gave a very low F value (Table 7). Conversely, some strains were closely related: for example strains 7220 and 11974 had an F value of 0.963 when determined for four probes and one genomic digest (Table 5). Overall, this same pair of L. plantarum strains had an F value of 0.804 (Table 7). For none of the pairwise comparisons was an F value of 1.0 obtained, showing that none of the strains was identical.

11. Theoretical analysis for a thirteenth strain of L. plantarum

However, let us assume that a thirteenth strain, denoted the "test strain X" is compared with the standard set of twelve organisms and shown to produce an identical band pattern with one of the strains previously analyzed. In order to assert that the two strains are indeed identical and therefore clonally derived it is necessary to state either the confidence (C) with which the two strains can be assumed to be derived from the same clone and are therefore identical (equation (4)) or the probability (X) that the two strains will be claimed to be identical when in fact they are different (equation (1) ) . The range of confidences with the present data may best be considered by examining the F values calculated for the maximum, minimum and average F values calculated from the complete pairwise analysis, taking the number of bands for test strain X to be the maximum, minimum and average values determined for the twelve strains analysed (Table 8). This table shows that the closer F for the population approaches 1 then the larger the value of F and hence the lower the confidence level that identity between two observed patterns will be interpreted as. showing clonal identity between the two sources of DNA (strain X and the strain with which it is compared). That is, if strain X was probed with the 4 probes used in this study and the genomic DNA's only analysed with restriction enzyme Bgll, then if the highest F value was used, the confidence level C would be 0.29. That is there would be a 71% chance that genuine clonal identity would be wrongly claimed. Using the average F value and the same conditions then C=0.999658 and only an 0.04% chance that clonal identity would be wrongly claimed. Table 8 also shows that the probability of error rapidly diminishes if a greater number of genomic digests are undertaken. For example, for similar F values of F =0.412 (same 4 probes; 1 restriction enzyme) and F =0.410 (same 4 probes; 6 restriction enzymes) the probability of error decreases from 3.42 x 10~⁴ to 4.67 x lθ^"27. In practice, therefore, once the F value (and hence P value) for a population of independently isolated individuals of the same species has been experimentally determined, then the required degree of certainty may be met by appropriate choice of the number of genomic DNA digests undertaken and the amount of DNA used to probe the test and standard organisms.

12. Actual analysis of a thirteenth strain of L. plantarum

A sample of strain BTLSl was taken as an "unknown" thirteenth strain for comparison with a "known" sample of BTLSl. The "unknown" strain was designated strain Y. Strain Y had been reisolated from a natural source which had been inoculated with BTLSl. Genomic DNA from strain Y and from BTLSl was digested with the restriction enzyme Bgll and hybridised with probes pBTL8, pBTL23, pBTL29 and pBTL30 as above. Identical band patterns were obtained.

The four probes revealed a total of 24 identically-positioned bands for the Bgll digests of each of strain Y and BTLSl (Table 1) . On the basis of the F_ma_.„x, Fmm. and Fav values deduced for the standard set of twelve strains of L. plantarum (Table 5) , it could be predicted from equations (1) and (4) that the probabilities that strain Y and BTLSl were in fact different were:

for F max : 0. 963²⁴ = 0 . 40 for F av : 0. 412²⁴ _— . 5. 72 x 10^{"1 0} fo r F m m . : 0. 056²⁴ = 9 . 05 x 10^{" 31}

Genomic DNA from strain Y and from BTLSl was digested with restriction endonuclease Avail and hybridised with probes pBTL8, pBTL23, pBTL29 and pBTL30 as above to try to achieve a lower probability (greater confidence) that the two strains are different. Identical band patterns were revealed with 35 common bands per strain (Table 1) . On the basis of the

F-m.ax_., Fm. and Fa„v. values deduced for the standardset of twelve strains of L. plantarum (Table 6) , it could be predicted that the probabilities that strain Y and BTLSl were in fact different were:

To achieve a yet greater level of confidence that strain Y and BTLSl were in fact identical, genomic DNA from the strain was digested with Neil, Acil, Asp700 and Clal and hybridised with the four probes as above. Identical band patterns were revealed. Taking into account the results for the Bgll and

Avail digests, 159 bands were revealed (Table 1) . On the basis of the Fma„„, F_m_. and Fa„v_. values deduced for the standard set of twelve strains of L. plantarum (Table 7) for all six digests, the probabilities that strain y and BTLSl were in fact different were:

for F _ _ 1 max 0.898 5^D9 _ 3.72 x 10^"8 for F — av 0.410¹⁵⁹ — 2.71 x 10~⁶² for F_,._ — 159 - mm < 1 x 10^""

All the values in Sections 10 to 12 can be subjected to conventional statistical controls. Reference Example : Amount of probe DNA

Restriction endonucleases that recognise 4-base sequences produce fragments of about 0.25 kbp average length and those that recognise 6-base sequences produce fragments of about 4 kbp average length. Using DNA sequences of total length L kbp in probes, genomic digests with 4-base cutters will produce q=4L fragments and 6-base cutters will produce q=0.25L fragments. Assuming a random distribution of restriction sites and as F, and F₂ may be derived by assuming values for P in equation (3), the probability (X) that two different organisms will fail to be distinguished is:

(i) using y 4-base cutters to digest the genomic DNA:

X_χ = F_χ ^L (6)

(ii) using z 6-base cutters to digest the genomic DNA:

X₂ = F₂°-^25LZ (7)

(iii) using y 4-base cutters and z 6-base cutters to digest the genomic DNA:

X₃ = X₁-X₂ (8)

The above analysis illustrates that if two organisms* hybridisation patterns are identical, then knowledge of the fraction of base pair differences (P) between the two organisms together with the total length of DNA (L) used in the probes to examine the genomic DNA of the two organisms at which a known number of restriction enzymes recognising a given number of base sequences (y 4-base cutters; z 6-base cutters), enables an estimate of the probability that two genuinely different organisms are wrongly considered to be identical. It follows that the probability of wrongly claiming identity can be enhanced either by increasing the amount of DNA used to probe the two organisms' genomic DNA or by increasing the numbers of digest of genomic DNA undertaken using difference restriction enzymes. One such analysis is illustrated in Table 9. This Table shows the amount of probe DNA required to give a given error probability. It also follows that the smaller the values of P (i.e. the greater the conservation of bases at equivalent positions in the DNA of the two organisms) then the greater the number of probes and individual genomic digests required to produce the same degree of confidence that identity has not been wrongly claimed.

TABLE 1

Enzyme TOTAL NUMBER OF BRANDS VISUALISED FOR STRAINS: used for digestion BTLSl 8531 6105 8299 5914 8026 8016 6376 11974 6461 8102 7720

Ball 24 14 14 11 12 16 13 11 14 9 11 13 Avail 35 45 24 43 26 30 20 38 25 13 14 21

The total number of bands for the complete set of the four probes and the six restriction enzyme digests were:

BTLSl 8531 6105 8299 5914 8026 8016 6376 11974 6461 8102 7220

159 115 91 125 74 137 99 98 122 68 86 83

Average number of bands = 104.75 -*_ 27.28

Table 2.1 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Bgl I and probed with pBTL8. The bands are numbered in sequence from the origin of the agarose gel.

Band DNA Sample

Number

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

1 2 3 + + 4 + + + 5 + 6 + 7 8 9 + + + + 10

Total no. of 2 4 3 3 bands

Table 2.2 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Bgl I and probed with pBTL23. The bands are numbered in sequence from the origin of the agarose gel

Band DNA Sample Number

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

1 2

3 4 5 Table 2.2 cont

Band DNA Sample Number

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

6 +

7

8

9 +

1 0 +

1 1 + + + +

1 2 + + + +

1 3 + + + +

1 4 , + +

Total no . of 5 3 2 4 2 bands

Table 2.3 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Bgl I and probed with pBTL29. The bands are numbered in sequence from the origin of the agarose gel,

Band DNA Sample Number

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

1 + + + + + + 2 + + + + +

3 +

4

5 + 6 +

7 +

8 + Table 2.3 cont

Band DNA Sample Number

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

9 + 10 + 11 + 12 +

Total 10 1 1 1 1 1 no. of bands

Table 2.4 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Bgl I and probed with pBTL30. The bands are numbered in sequence from the origin of the agarose gel,

Band DNA Sample Number

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

1 + + +

2 + + + + + + + + + +

3 + + + + +

4 + + + + + + +

5 + + + + + + + + +

6 + + + + +

7 + + + + + + + + + + + +

8 + +

9 + +

1 0 + + + + + + + + +

1 1 +

Total no. of 5 7 5 5 5 5 8 5 5 bands TABLE 2.5: Pairwise comparison of band pattern data contained in Table 2.1. For each unique pair of DNA digests, with a and b fragments having N bands in common, the values are reported as 2N/(a+b)

TABLE 2.6: Pairwise comparison of band pattern data in

Table 2.2, determined as indicated for Table 2 .

TABLE 2.7: Pairwise comparison of band pattern data in Table 2.3, determined as indicated for Table 2.5

TABLE 2.8: Pairwise comparison of band pattern data in _^ Table 2.4, determined as indicated for Table -..._

TABLE 2.9: This Table shove the sum of 2N/(a+b) ratios for the four BTL probes used in the analysis of genomic DNA digested with restriction endonuclease Bgl I.

Table 3.1 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Ava II and probed with pBTLδ. The bands are numbered in sequence from the origin of the agarose gel.

Band DNA Sample Number

8531 8299 8026 6376 6461 7220

BTLS1 6105 5914 8016 11974 8102

1 +

2 +

3 + +

4 + +

5 +

6 + + + + + +

7 + + +

8 + + +

9 + +

10 + + + + +

11 +

12 + + + +

13 + + + + + + + + +

14 + + + + + + + + +

15 +

16 + + +

17 + + + + + +

18 + + +

19 + + + + + 0 + + + + 1 + + 2 + 3 - + 4 + + + 5 + + + + + 6 + + +

Total no. of 8 12 6 11 9 7 6 7 7 4 4 6 bands Table 3.2 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Ava II and probed with pBTL23. The bands are numbered in sequence from the origin of the agarose gel,

Band DNA Sample wuiuϋc

8531 8299 8026 6376 6461 7220 BTLS1 6105 5914 8016 11974 8102

1 + +

2 +

3 + +

4 + + + +

5 . + +

6 + +

7 + + + + +

8 + + +

9 ^' + +

10 + + *

11 + +

12 + + + +

13 + +

14 + + + + + + + + +

15 + +

16 + + + + + +

17 +

18 + + + + +

19 + +

20 + + + + +

21 + + +

22 + + + + +

23 + + +

24 + -f + + + + + + +

25 + + +

26 + + +

27 + + + + + + + + + +

28 + + + + + + +

29 + + + + 30 + + +

Total no, bands ' 9 16 9 19 5 13 5 18 8 2 4 5 Table 3.3 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Ava II and probed with pBTL29. The bands are numbered in sequence from the origin of the agarose gel .

Band DNA Sample Number

BTLSl 8531 6105 Θ299 5914 8026 8016 6376 11974 6461 8102 7220

1 + +

2 + +

3 + + + + + +

4 + +

5 +

6 +

7 +

8 + +

9 + +

10 +

11 +

12 +

13 +

14 +

Total no. of lO bands

Table 3.4 DNA from 12 Lactobacillus plantarum strains digested with restriction enzyme Ava II and probed with pBTL3θ. The bands are numbered in sequence from the origin of the agarose gel.

Band DNA Sample

Number -

BTLSl 8531 6105 8299 5914 8026 8016 6376 11974 6461 8102 7220

1

2

3

4

5

6 +

7 + +

8 + +

S + +

10

11

12 +

13

14 + +

15

16

17 +

18

19 +

20

21

22 +

23

24 +

25 +

26 +

27 +

28 +

29 +

30 +

31 +

32 +

33 +

34 + +

35 +

Total no. of 8 14 8 - 11 9 8 6 12 bands T_ABLE 3.5: Pairwise comparison of band pattern data in Table 3.1, determined as indicated for Table 2.5.

TABLE 3.6: Pairwise comparison of band pattern data in Table 3.2, determined as indicated for Table 2.5

TABLE 3.7: Pairwise comparison of banα pattern data in Table 3.3, determined as indicated for Table 2.5

TABLE 3.8: Pairwise comparison of band pattern data in Table 3.4, determined as indicated for Table 2.5

TABLE 3.9: Sum of pairwise comparison of the band patterns for data contained in Tables 3•5-3.8

This Table shows the sum of 2N/(a+b) ratios for the four pBTL probes used in the analysis of genomic DNA digested with restriction enzyme Ava II

TABLE 4 : Sum of pairwise comparison from four probes and six restrictive enzymes obtained by summing the data as shown in Tables 2.9 and 3.9 (as well as equivalent tables for the remaining four restriction enzymes - data not shown⁾

TA3LE 5 The fraction of conserved fragments F determined from the data in Table 2.9. F _raaχ = 0.963 for the pairwise comparison of strains 7220 and 11974. F _min = 0.056 for strains 6376 and BTLSl

F _av= 0.412.

TABLE 6 The fraction of conserved fragments F determined from the data in Table 3.9. F _maχ = 0.870 for the pairwise comparison of strains 7220 and 11974. F _mi_n = 0.033 for strains 5914 and BTLSl F _av= 0.354.

TABLE 7 The fraction of conserved fragment F determined from the data in Table 4. F max - 0.898 for the pairwise comparison of strains 8026 and 8531. F_mi_n = 0.057 for strains 8102 and BTLSl. F ^• = 0.410 (mean of 66 determinations of F^*) av

TABLE 8, Determination of F^ values

Fraction of conserved F^ values for q equal to: fragments, F, for:

9 13.50 24

(Data from Table 5: 4 probes and Bgll digestion of genomic DNA only)

F max = 0.963 0.71 0.60 0.40

F in = 0.056 5.42 x l-^"12 1.26 x 10^'17 9.05 x 10^"31

Fav - 0.412 3.42 x 10^"4 6.33 x 10^'6 5.72 x 10^"10

(Data from Table 6: F^ values for q equal to: 4 probes and Ava II digestion of genomic 13 27.83 45 DNA only)

F max = 0.870 0.16 0.02 1.90 x 10^"3

F min = 0.033 5.50 x 10^{' 20} 5.89 x 10^"42 2.15 x 10^"67

Fav = 0.354 1.37 x 10~⁶ 2.81 x 10^"13 5.07 x 10^"21

(Data from Table 7: -Λ values for q equal to: 4 probes and all six digests of genomic DNA) 68 104.75 159

F max = 0.898 6.65 x 10~⁴ 1.28 x 10~⁵ 3.72 x 10^"8

F min = 0.057 2.51 x 10~⁸⁵ < 1 x 10^"" £ 1 x 10^""

Fav = 0.410 4.67 x 10^*27 2.75 x 10^'41 2.71 x 10^"62

TABLE 9

Basepair Fraction of conserved fragments Amount of probe DNA required (Kb) differences (F) to f-ive an error probability of: (P) 6 base 5 base 4 base Cutter Cutter Cutter 1 x 10⁶ 1 x 10⁹ 1 x 10¹² 1 x 10¹⁵

0.0300 0.595 0.646 0.674 4.9 7.4 9.8 12.3 0.0150 0.768 0.801 0.819 9.7 14.5 19.4 24.2 0.0075 0.875 0.894 0.904 19.2 28.8 38.4 48.0 to 0.0037 0.935 0.946 0.951 38.3 57.4 76.5 95.7 0.0019 0.967 0.972 0.975 76.4 114.6 152.8 191.0 0.0009 0.983 0.986 0.987 152.6 228.9 305.2 381.5 0.0005 0.991 0.993 0.994 305.0 457.5 610.0 762.5 0.0002 0.996 0.996 0.967 609.9 914.8 1219.7 1524.6

Claims (21)

1. A method of determining to an ascertainable probability whether a first and a second organism, the identity of one of which is known, are identical, which method comprises:

(i) digesting genomic DNA of the first organism with one or more restriction endonuclease; (ii) separating by electrophoresis the DNA fragments thus-obtained; (iii) determining the positions of the fragments thus-separated which hybridise with one or more labelled probes, the or each probe comprising a fragment of DNA which has been derived randomly from the genomic DNA of an organism of the same species as the first or second organism; and (iv) comparing the positions of the fragments thus-determined with the positions of DNA fragments which bind to the or each said probe, which have been produced from genomic DNA of the second organism by digestion of genomic DNA of the second organism with the or each said endonuclease and which have been subjected to electrophoresis in an identical manner to the DNA fragments obtained from genomic DNA of the first organism;

steps (i) to (iv) being effected using an amount of probe DNA and one or more restriction endonucleases such that sufficient bands are revealed by the hybridisation in step (iii) to achieve a sufficiently low probability (X) that, when the comparison in step (iv) reveals that the two organisms appear identical, the two organisms will have failed to have been distinguished as genuinely different and unrelated as determined by:- X = F<3 (i )

wherein F is a fraction representative of the proportion of DNA fragments which are identical between restriction endonuclease digests of genomic DNA of pairs of independently-obtained organisms of the same species as the first and second organisms and q is the number of positions revealed by the probing in step (iii) .

2. A method according to claim 1, wherein steps (i) to (iv) are effected by:

(i") digesting separately genomic DNA of the first organism and genomic DNA of the second organism with the same restriction endonuclease and, optionally, dividing each digest into portions; (ii') subjecting to electrophoresis side-by-side on a gel the digest, or a portion of the digest, for each organism;

(iii' )probing the gel using a said labelled probe and comparing the hybridisation patterns for the two organisms thus-revealed; and (iv') optionally repeating steps (ii') and (iii') for one or more further portion of the digest for each organism but using a said labelled probe comprising a different said fragment of DNA each time.

3. A method according to claim 2, wherein the procedure of steps (i¹) to (iv¹) is effected two or more times using a different restriction endonuclease each time.

4. A method according to claim 1, wherein the or each probe comprises a fragment of DNA derived randomly from the genomic DNA of the first or second organism.

5. A method according to claim 1, wherein the

F value has been determined experimentally by: ( a) digesting separately using the same restr iction endonuclease genomic DNA of a number of independently-obtained organisms of the same spec ies as the said f irst and second organisms suff icient to obtain a F value representative of the species and, optionally, dividing each digest into portions;

(b) subjecting to electrophoresis side-by-side on a gel the digest, or a portion of the digest, for each of the independently-obtained organisms; (c) probing the gel using a labelled probe comprising a fragment of DNA derived randomly from the genomic DNA of an organism of the said species;

(d) comparing the hybridisation patterns on the gel thus revealed for pairwise combinations of the independently-obtained organisms; and

(e) optionally repeating steps (b) to (d) for one or more further portion of the digest for each of the independently-obtained organisms but using a said labelled probe comprising a different said fragment of DNA each time.

6. A method according to claim 5, wherein the procedure of steps (a) to (e) is effected two or more times using a different restriction endonuclease each time.

7. A method according to claim 5, wherein the hybridisation patterns for each and every possible pairwise combination of the independently-obtained organisms is compared in step (d) .

8. A method according to claim 5, wherein the or each probe employed in the determination of the F value comprises a fragment of DNA derived randomly from the genomic DNA of one of the said independently-obtained organisms or from genomic DNA of the first or second organism.

9. method according to claim 5, wherein one of the said independently-obtained organisms is the said first organism or the said second organism.

10. A method according to claim 5, wherein the or each restriction endonuclease employed in step (i) is the same as the or each restriction endonuclease employed in the determination of the F value.

11. A method according to claim 5, wherein the or each probe employed in steps (iii) and (iv) is the same as the or each probe used in the determination of the F value.

12. A method according to claim 1, which further comprises (v) calculating the degree of probability (X) that the two organisms will have failed to have been distinguished when the comparison in step (iv) reveals that the two organisms appear identical.

13. A method according to claim 1, wherein the first and second organisms are organisms which reproduce principally by asexual means.

14. A method according to claim 13, wherein the organisms are bacteria.

15. A method according to claim 13, wherein the organisms are fungi.

16. A method according to claim 13, wherein the organisms are viruses.

17. A method according to claim 1, wherein the organisms are plants.

18. A method according to claim 1, wherein steps (i) to (iv) are effected using an amount of probe DNA and a number of restriction endonucleases such that X is 10 or less.

19. A method according to claim 18, wherein X is 10 -25 o >rr lleessss..

20. A method according to claim 18, wherein X is 10 or less.

21. A method of determining to an ascertainable probability whether a first and a second organism, the identity of one of which is known, are identical, which method comprises:

(A) determining a fraction (F) representative of the proportion of DNA fragments which are identical between restriction endonuclease digests of genomic DNA of pairs of independently-obtained organisms of the same species as the first and second organisms by:

(a') digesting separately using the same restriction endonuclease genomic DNA of a number of independently-obtained organisms of the said species sufficient to obtain a F value representative of the species and, optionally, dividing each digest into portions; (b' ) subjecting to electrophoresis side-by-side on a gel the digest, or a portion of the digest, for each of the independently-obtained organisms; (c¹) probing the gel using a labelled first probe comprising a fragment of DNA derived randomly from the genomic of DNA of an organism of the said species;

(d') comparing the hybridisation patterns on the gel thus revealed for pairwise combinations of the independently- obtained organisms; and (e') optionally repeating steps (b¹ ) to (d') for one or more further portion of the digest for each of the independently-obtained organisms but using a said labelled first probe comprising a different said fragment of DNA each time; (B) digesting genomic DNA of the first organism with one or more restriction endonuclease;

(C) separating by electrophoresis the DNA fragments thus-obtained;

(D) determining the positions of the fragments thus-separated which hybridise with one or more labelled second probes, the or each probe comprising a fragment of DNA which has been derived randomly from the genomic DNA of an organism of the same _ . species as the first or second organism; (E) comparing the positions of the fragments thus-determined with the positions of DNA fragments which bind to the or each said probe, which have been produced from genomic DNA of the second organism by digestion of genomic DNA of the second organism with the or each said endonuclease and which have been subjected to electrophoresis in an identical manner to the DNA fragments obtained from genomic DNA of the first organism; and (F) when the comparison in step (E) reveals that the two organisms appear identical, calculating the probability (X) that the two organisms will have failed to have been distinguished as genuinely different and unrelated by:

(1)

wherein F is as defined above and q is the number of positions revealed by the probing in step (D) .