WO1992005280A1

WO1992005280A1 - Identification of organisms

Info

Publication number: WO1992005280A1
Application number: PCT/GB1991/001618
Authority: WO
Inventors: Gregory Glenn Lennon; Hans Rudolph Lehrach
Original assignee: Imperial Cancer Research Technology Limited
Priority date: 1990-09-21
Filing date: 1991-09-20
Publication date: 1992-04-02
Also published as: GB9020621D0; EP0550531A1; JPH06501385A

Abstract

Nucleic acid (usually PCR-amplified) from organisms is screened with a library of short probes to determine whether each probe hybridises to the nucleic acid and, if so, how many times. A 'hybridisation signature' is thereby created with a score (0, 1, 2 etc) for each probe. The method can be used to provide an initial screening of bacteria gathered from the wild to weed out those of an undesired type and those studied previously, and also for typing organisms of medical or industrial interest.

Description

IDENTIFICATION OF ORGANISMS

The present invention relates to the identification of organisms, particularly microorganisms, and especially eubacteria.

There are many situations in which one wishes to identify whether a microorganism is present in a test sample and/or to identify the nature of the microorganism. For example, the involvement of bacteria in some di-seases is suspected but not yet proved and cell cultures can become contaminated with organisms which are difficult to culture and identify by conventional methods. See, for example, US Patent No 4 851 330, Gδbel et al (1987 J . Gen . Microbiol . 133, 1969-1974) and Chen et al (1989 FEMS Microbiol . Lett . 57, 19-24) which relate to the use of probes to hybridize to ribosomal RNA and thereby identify the presence or nature of a microorganism. These probes are roughly 17-30 nucleotides long and are designed to be specific to the rRNA of the organism sought. Only one or a few specific probes are used in any given assay.

In forensic DNA •^■fingerprinting" methods, relatively short and therefore frequently-binding probes have been used, but only as markers to identify size-sorted fragments generated by restriction enzymes. The use of the probes alone would not be informative. In contrast, we have now found benefits in using a relatively large number of short probes to generate a "hybridisation signature" for a given organism, the "hybridisation signature" consisting of the series of determinations of hybridisation positivity for each probe according to whether the probe binds to the sample nucleic acid (and, if it binds, whether it binds multiply) or not.

Thus, one aspect of the invention provides a method of identifying an organism, the method comprising (1) obtaining nucleic acid derived from the organism, and (2) determining, for each of at least 5 nucleotide probes, whether the probe hybridises to the said nucleic acid.

Preferably, the number of probes used is at least 10, 15, 20, 25 or more when one wishes to distinguish between large numbers of samples (for example hundreds) . 10-20 would typically be enough in such a situation to establish a meaningful signature, provided that each probe does not hybridise so seldom or so frequently to the sample nucleic acid as to give the same highly negative or highly positive hybridisation signature to all the samples which one is testing and between which one wishes to differentiate. However, if one wishes only to classify an organism into one of a few categories, for example two, then fewer probes will usually be needed. This is also the case when the probes bind to more than one site on the sample nucleic acid, since each probe is then capable of yielding more than a binary "yes/no" result and nucleic acid which binds the probe three times can be differentiated from nucleic acid which binds it only once or twice. The key feature of the methods of the present invention is that a plurality of short probes are used in concert to generate an informative "hybridisation signature".

Thus, a second aspect provides a method of assigning a "hybridisation signature" to a given organism's nucleic acid based on its pattern of hybridisations to more than one oligonucleotide probe.

Suitably, each probe is such that, under appropriate hybridisation conditions, it will bind to about 25-75% of all nucleic acid of the types being assayed, preferably 30-70%, more preferably 40-60% and most preferably 40-50%. Each probe, whilst being specific (subject to the hybridisation conditions) for a given sequence corresponding to the length of the probe, is non-specific in the sense that, on its own, it cannot be used with any precision to distinguish between different lengths of nucleic acid which are all significantly longer than the probe. The specificity of the assay method thus derives from the fact that a relatively high number of probes are used. The probes are typically 4-15 nucleotides long, preferably 5-10 and often 8 or 9 nucleotides long. A mixture of two or more similar probes (for example 5'- TGCATGGC-3 ' and 5 '-TGCATGGG-3 ' ) can be treated as a single probe (ie used in a single hybridisation step) to increase the frequency with which the composite probe binds to sample nucleic acid. The term "probe" is used in this specification to include such composite probes.

It has been found that probes for regions of the organisms^• genome which vary considerably from species to species are particularly useful in identifying similarities and differences between organisms of differing species. If such variable regions are flanked by conserved regions, then it is relatively easy to produce primers for DNA amplification by the PCR method, even when the relevant highly variable sequence of the organism is unknown. We have found that ribosomal RNA genes, particularly the 16S gene, are suitable according to both of these criteria. Specific PCR primers include

(A) 5^»-GAGAGTTTGATCCTGGCTCAG-3 •

(B) 5 '-TGCCTCCCGTAGGAGTCTGG-3 ' (C) 5^»-GGTAGTCCACGCCGTAAACG-3 • and (D) 5 ^•-ATCTCACGACACGAGCTGACG-3 •

and analogues and fragments thereof. The analogues and fragments are those which provide for specific amplification by the PCR of DNA adjoining that to which the said four specific primers hybridise.

Probes and primers can be made by known means, for example by simple chemical synthesis using apparatus supplied by Applied Biosystems Inc.

Unless the amount of nucleic acid in each sample can be arranged to be the same in each case, or can be measured by some other means, it is convenient to normalise the amount by determining the extent of binding of a probe which is known to be common to all of the samples under consideration, for example one in a highly conserved region. Such probes are hereinafter termed "conserved probes", as opposed to the previously discussed "variable probes" . The binding of the variable probes is then expressed as a ratio to the binding of the conserved probes. Conserved probes are not subject to the same size constraints as variable probes.

Probes may comprise all or part of the sequence of a PCR primer.

The identification of the organism which is achieved by methods in accordance with the invention may be more or less precise according to how many probes are used and may be more or less informative according to what reference data one has for the same and other organisms. Thus, the methods may be used to place the organism in any of the taxonomic groups (Kingdom, phylum, class, sub-class, order, sub-order, family, genus, species or sub-species), to identify similarities with organisms having desirable or undesirable characteristics or even to identify the particular strain of organism. The methods are particularly useful in screening microorganisms, especially bacteria, which are collected more or less randomly from the wild to determine whether the sample microorganism has been studied previously or whether it falls into a group of microorganisms which are believed to be of particular interest. Heretofore, much effort has been wasted in characterising bacteria which turn out to have been studied " previously or turn out to be members of a group which are known to be of no interest. The studying of bacteria and fungi to find those which secrete valuable antibiotics is one particular process in which the methods of the invention can be used as an initial screen. The methods of the invention, in using several small probes, differ from those which employ large specific probes to identify regions unique to an individual or specific to a desired gene. In the mass screening exercises described above, the number of organisms being investigated is typically large (several hundred at a time) and about 10-20 probes will usually be necessary to differentiate between them adequately.

Another use for the methods of the invention is to identify, at least to an extent, the nature of a pathogen. For example, in the screening of cervical smears, papillomaviruses which are found must be typed, especially to determine whether they are type 16 or type 18, the two types thought to be responsible for cervical cancers and other abnormalities. There are over 35 papillomavirus types which have so far been isolated in such smears. If probes are used which generate only binary hybridisation signatures (binding or not binding) then six probes may be necessary to distinguish between 35-60 types (as 2⁶ is 64) . However, fewer probes may be employed if the probes bind multiply or if one wishes only to know whether the virus is type 16 or 18 or another type. In such a situation, 2, 3, 4 or 5 probes may be used, with consequent savings of time and materials. It should be noted that the probes are not individually capable of specifically binding to a sequence motif unique to the virus type in question; it is still the hybridisation signature obtained by the use of a plurality of probes which yields the desired information.

Further uses include the typing of other pathogens, for example HIV, rhinoviruses and Streptococcus species, and the typing of desirable or undesirable bacterial and viral strains in cheese-making, for example the lysin-producing bacteriophages and the desirable bacteria Lactococcus lactis subsp. lactis and subs, cremoriε .

The methods of the invention are generally applicable to identifying or typing viruses, bacteria, mycoplasmas

(including acholeplasmas) , spirochaetes, fungi, protozoa, plants and animals. Plants (including trees) may be typed as an initial screening process to identify those most likely to be useful in some context (for example desirable bread-making strains of wheat, oil- or pharmaceutical- bearing plants or nutritionally valuable or disease- resistant strains of crop plants) and this may be particularly valuable as part of a plant breeding program where one may otherwise have to wait until the hybrid is mature before being able to identify its characteristics.

The nucleic acid which is probed may be DNA, RNA or cDNA.

The method may involve any suitable combination of nucleic acid and probe. This includes, but is not limited to, chromosomal nucleic acid isolated in toto (DNA) , ribonucleic acid, or specific nucleic acid regions amplified in order to increase the sensitivity of the method. In the case of specifically amplified nucleic acid, the important concept is that regions of nucleic acid variability will be located between highly conserved nucleic acid sequences, such that oligonucleotide primers can be designed that will complement the conserved regions and will allow the amplification of the variable regions located in between them. Examples of suitable loci include the ribosomal RNA genes, such as the bacterial 16S and 23S genes, and regions between them. Primers are likely to consist of oligonucleotides of length appropriate to their use in the polymerase chain reaction. Probes are likely to consist of oligonucleotide nucleic acids between six and ten bases in length, tagged either radioactively or non- radioactively. The nucleic acids to be probed may be positioned on a support such as a nylon filter membrane. Hybridisation of the probes to the nucleic acids must allow sufficient specific hybrid molecules to form and to be detected. Accurate quantitation of the amounts of probe bound to each nucleic acid is important, and may be achieved through the use of storage phosphor screen autoradiography or by other means.

Examples of the invention will now be described with reference to the accompanying drawings, in which:

Figure 1 shows partial DNA sequences from the 16S ribosomal RNA genes of four bacteria;

Figure 2 illustrates the generation of distinct hybridisation signatures for four bacteria, using only two probes;

Figure 3 illustrates diagrammaticc.lly the use of six probes to generate respective hybridisation signatures for six organisms;

Figure 4 shows partial DNA sequences from several human papillomaviruses; and

Figure 5 shows partial sequences (as they might appear following reverse transcription) from four human rhinoviruses. EXAMPLE 1

Figure 1 shows nucleotide sequence alignments of the beginning of several 16S ribosomal RNA genes from different bacterial organisms. The sequence of the E. coli 16S rRNA gene is the reference sequence (GenEMBL Accession No J01695) . The next three sequences are from distantly related bacteria Streptomyces coelicolor, GenEMBL Accession No Y00411; Pseudojϊionas aer ginosa , GenEMBL Accession No X06684; Pseudomonas testoεteroni, GenEMBL Accession No M11224) . A colon is used to indicate sequence conservation between the 16S rRNA sequence of these species and E . coli . The last two sequences (P. aeru . and P . test . ) are from two species of the same genus, and illustrate that variable regions remain variable even between two closely related species. Examples of the sequences of oligonucleotides used as primers for the polymerase chain reaction (PCR) are indicated by overlying arrows, and are termed Primer A and Primer B. Examples of the sequences of oligonucleotides used as probes in the hybridisation are underlined as Probe 1 (GACCCTCG, S . coel . ) and Probe 2 (CTGAGACAC, present in all species shown) .

Preparation of amplified 16S rRNA nucleic acid regions. Total DNA from the organisms used was kindly provided by Prof D A Hopwood. Target DNA (3 nanograms) was used in the PCR reaction with 500 nanograms of each amplification primer. The PCR was carried out using 2.5 units of Taq DNA polymerase (Perkin-Elmer Cetus) in a 50 μl reaction according to the manufacturer's conditions. The reactions consisted of 30 cycles of 1 minute at 94°C, 1 minute at 55°C, and 2 minutes at 73°C. Pairs of oligonucleotide primers enabling amplification of specific rRNA regions were used as described above for primer pair A and B. Primer pair C and D consisted of oligonucleotides of sequence 5 ^• -GGTAGTCCACGCCGTAAACG-3 ' and 5 ^• - ATCTCACGACACGAGCTGACG-3 ' .

Preparation of membranes. Equimolar amounts of 2X denaturant (3M NaCl, 1 N NaOH) were added directly to PCR products. Approximately 2 μl of each product was then dotted manually onto GeneScreen membranes (DuPont) that had been pre-wet in 0.5 N NaOH, 1.5 M NaCl. The filters were neutralised in 50 mM Na₂HP0₄, pH 6.5, then dried, baked at 80°C for one hour, and UV-irradiated for 2 minutes.

Preparation of radioactive oligonucleotide probes. Oligonucleotides were synthesised on an Applied Biosystems 380A machine and labelled with [α-³²P]ATP (3000 Ci/mmol^"1; Amersha International) and T4 polynucleotide kinase. The sequence of probes p3 and p4 were 5 '-CCACGTTG-3 ' and 5 '- TGCATGGC-3 ' , respectively.

Hybridisation of probes to filters. Filters were prehybridised for 30 minutes at room temperature in hybridisation solution (4X SSC, 7% sarkosyl) , then hybridised with probe at a concentration of 3nM for 3 hours at 80°C in a volume of 250 μl per 28 cm² filter. Filters were washed for one hour at 12°C in two changes of 100 ml of hybridisation solution, then exposed to either film or storage phosphor screens.

Analysis of hybridisation signals. Storage phosphor screens were quantitated (PhosphorImager, Molecular Dynamics Inc) and analysed using the Phosphorlmager • software. Ratios of signals were calculated as the amount of signal generated by the variable region probe (for example, probe 1 as indicated above) divided by the amount of signal generated by the conserved probe (probe 2, as indicated above) .

Figure 2 is a schematic illustration of the hybridisation information obtained using four oligonucleotide probes (two as indicated in Figure 1) against a nylon filter membrane onto which the PCR products were spotted. Two separate regions of the 16S rRNA gene were amplified independently using primer pair A and B (as indicated in Figure 1) , or primer pair C and D. Probes pi and p2 were designed to be used as probes for nucleic acid samples from the region amplified using primer pair A and B; similarly, probes p3 and p4 for DNA from primer pair C and D. Hybridisations were done in .duplicate, and averages were used. Hybridisation signals were quantified by storage phosphor screen autoradiography, and signals from variable region probes (pi and p3) normalised against signals from conserved region probes (p2 and p4) . The four organisms classified by these means are shown to have been assigned four different hybridisation signatures.

The use of a control organism of known seguence allows one to compare the ratios obtained with any probe for that organism to the ratios obtained with the same probe for every other organism. Ratios close to the control "positive" ratio are thus also classified as positive, whereas those close to a control "negative" ratio are classified as negative. Ratios double that of the control single copy sequence are classified as representing two copies of the probe sequence.

Examination of the schematic and ratios as shown in Figure 2 indicates that all four organisms, including the two (S . lividans and S . glaucescenε) whose 16S rRNA sequence is unknown at this date, acquired different hybridisation signatures, thus confirming experimentally that these organisms can be distinguished by the methods described herein. In application to larger numbers of (different) organisms, more probes are to be used as discussed earlier.

Figure 3 is a diagrammatic representation of the concept of assigning hybridisation signatures based on the accumulation of positive or negative assignments to each of six probes assayed against an array of nucleic acid samples. The number of different hybridisation signatures possible in any one set of experiments is, for experiments where only presence or absence is measured, 2^N where N is the number of probes used. Thus, the use of 20 probes in such an experiment yields 2²⁰, ie over one million different (in this case, binary) hybridisation signatures. Each type of nucleic acid sample will reproducibly be assigned one of these signatures. In experiments where the presence of the probe sequence in a nucleic acid sample can be refined further to indicate the number of times that sequence is independently present in the sample, there are even greater numbers of hybridisation signatures. As diagrammed here, the assay detects either 0, 1, or 2 copies of the probe in the nucleic acid of the organisms tested.

EXAMPLE 2

Figure 4 shows nucleotide sequence alignments of portions of the genomes of several different human papillomaviruses, including two associated with a high risk of developing cervical cancer (Pphl6 and Pphl8) . As in Figure 1, a colon indicates nucleotide conservation between the sequences shown. Sequences for Pphll, Pphl6, Pphlδ, and Pph33cg are from GenEMBL (Accession Numbers M14119, K02718, X05015, and M12732, respectively) . A pair of potential PCR primers are marked as primers E and F, and probes for both a non- conserved (probe 5) and a conserved (probe 6) region are indicated. The probes are prepared and used as in Example 1 above.

EXAMPLE 3

Figure 5 shows nucleotide sequence alignments of portions of four human rhinovirus genomes as they might appear after reverse transcription. Sequence from human rhinoviruses types 14, 89, IB, and 2 are shown (from GenEMBL sequences, Accession Numbers X01087, M16248, D00239, and X02316, respectively) . As before, potential primers are indicated (Primers G and H) and an example is shown of both a probe from a variable region (probe 7 and a conserved region (probe 8) . The probes are prepared and used as in Example 1 above.

Claims

1. A method of identifying an organism, the method comprising (1) obtaining nucleic acid derived from the organism, and (2) determining, for each of at least 5 nucleotide probes, whether the probe hybridises ~o the said nucleic acid.

2. A method of assigning a hybridisation signature to a given organism's nucleic acid based on its pattern of hybridisations to more than one oligonucleotide probe.

3. A method according to Claim 1 or 2 comprising the step of amplifying the amount of a selected region of nucleic acid before the hybridisation to it as said.

4. A method according to Claim 3 comprising selecting single-stranded DNA fragments suitable for use as primer pairs in a polymerase chain reaction, the pairs being adapted to hybridise to genomic regions highly conserved throughout evolution, and flanking less well conserved regions.

5. A method according to any one of the preceding claims wherein each probe will hybridise to the nucleic acids of multiple organisms.

6. A method according to any one of the preceding claims wherein the probe consists of 4-15 nucleotides.

7. A method according to any one of the preceding claims comprising normalising for the amount of nucleic acid on a given position on a given filter comprising an analysis of a hybridisation using a conserved probe common to all nucleic acids on said filter.

8. A PCR primer consisting^' of single-stranded DNA fragments adapted to hybridise to ribosomal RNA gene sequences.

9. A PCR primer according to Claim 7 selected from

(A) 5 '-GAGAGTTTGATCCTGGCTCAG-3 '

(B) 5 '-TGCCTCCCGTAGGAGTCTGG-3 '

(C) 5 '-GGTAGTCCACGCCGTAAACG-3 ' and

(D) 5 •-ATCTCACGACACGAGCTGACG-3 '

and analogues and fragments thereof.

10. A single-stranded DNA probe selected from

(A) 5 '-TGCATGGC-3 '

(B) 5--CCACGTTG-3'

(C) 5--GACCCTCG-3 '

(D) 5--CTGAGACAC-3'

11. Single-stranded DNA fragments according, to Claim 9 consisting of the fragments used as the primer pairs themselves as in Claim 4.

12. A kit for classifying organisms comprising a plurality of nucleic acid variable probes, each probe being 4-15 nucleotides long.

13. A kit according to Claim 12 wherein each probe is 5-10 nucleotides long.

14. A kit according to Claim 12 or 13 additionally comprising conserved probes which hybridise to conserved regions of nucleic acid in the organisms being classified.

15. A kit according to any of Claims 12-14 additionally comprising one or more pairs of PCR primers.