WO1996019585A1 - Typing of microorganisms - Google Patents

Abstract

The invention relates to a method of detecting, identifying and quantitating microorganisms, and to oligonucleotide probes for use in this method. In particular, the method relates to the typing of specific isolates of microorganisms, and discrimination between strains and allelic subtypes. The method uses amplification of the 16S-23S rRNA spacer region using a highly conserved region from the 3' end of the 16S-23S rRNA spacer region, and/or a highly conserved region from the 5' end of said region. The method enables epidemiological tracing of specific microorganism subtypes. Preferred primers are disclosed and claimed.

Description

TYPING OF MICROORGANISMS

This invention relates to a method of detecting, identifying and quantitating microorganisms, and to oligonucleotide probes for use in this method. In

particular, the method relates to the typing of specific isolates of microorganisms, and discrimination between strains and allelic subtypes.

Background and Prior Art

Correct identification of microorganisms is a vital part of microbiological practice, and is a

prerequisite for selecting the most suitable form of treatment of disease, for prevention of contamination of foods, and for prevention of cross-infection. In many cases it is important not only to identify the species of organism, but also to determine the strain and serotype, or even an allelic subtype. Such identification at subspecies level is particularly important in epidemiological tracing, for example in establishing the origin of

hospital-acquired (nosocomial) infection.

Disease caused by Staphylococcus aureus is most often the result of hospital-acquired infections. Strains of S. aureus that are resistant to the penicillinase¬resistant antibiotic methicillin are now common, the first major nosocomial epidemic of a methicillin-resistant strain of S. aureus (MRSA) having been described by Stewart and

Holt (1963). The determination of whether or not isolates of S. aureus represent a single strain is of considerable epidemiological value in a hospital setting.

Conventional methods of microbiological identification require culturing of microorganisms in a suitable growth medium, and this entails a delay of at least 24 hours, and often much longer. The methods

utilised are completely manual, and rely very heavily on the experience and skill of the microbiologist. Such methods do not lend themselves to automation. The only alternative presently widely used is immunological identification, which usually requires the use of monoclonal antibodies. A prerequisite for such immunological identification is that the species of the organism in question be known, or at least strongly

suspected. Some antibodies of broad specificity are available for use in preliminary screening. While

immunological methods can be automated, they are time consuming and expensive.

Other typing methods, which can be used for certain species only, include toxin detection, isolation of plasmids, bacteriophage, bacteriophage/bacteriocin typing systems, antibiotic susceptibility testing, protein typing by SDS-polyacrylamide gene electrophoresis, pulsed-field gel electrophoresis, immunoblotting, and restriction endonuclease analysis.

The availability of molecular biological methods, including oligonucleotide probing and polymerase chain reaction (PCR), offers a means of more accurate, and more rapid and precise identification. It also permits the identification of previously unknown organisms. Techniques based on analysis of ONA are more discriminating than traditional methods, and overcome the variability inherent in discriminating between strains by assays which rely upon the phenotype of the target organism.

The rRNA operon, rrn, is present in varying copy number in all bacteria, with some regions highly conserved and others highly variable (Neefs et al, 1990).

Consequently, when genomic DNA digested with a restriction enzyme is hybridized to rRNA operons, several bands are detected (Gamier et al, 1991). The Southern hybridization of rRNA operons (ribotyping) to detect restriction fragment length polymorphisms (RFLPs) between strains has been reported in many bacterial species, including Salmonella typhi strains (Altwegg et al, 1989), E. coli strains

(LiPuma et al, 1989), Xanthomonas maltophilia (Bingen et al, 1991), Legionella pneumophila strains (Harrison et al, 1992), and Stapbylococcus species and subspecies, including MRSA strains (Blumberg et al, 1992; Monzon et al, 1991; Prβheim et al, 1991; DeBuyser et al, 1992). However, Southern hybridization is slow and labour intensive.

Although the rRNA operon has a very high genetic stability and the length of the 16S rRNA gene is constant in all eubacteria (Neefs et al, 1990), the number of rRNA operons has been completely analyzed by Southern

hybridization or PCR of the rDNA in only a few eubacteria, including Escbericbia coli, demonstrating 7 operons (Morgan et al, 1977), Bacillus subtilis, 10 operons (Loughney et al, 1982), Clostridium perfrinςrens, 10 operons (Garnier et al, 1991), C. difficile, 10 operons (Gürtler, 1993) and Mycobacterium species (Bercovier et al, 1986) and

Mycoplasma species (Amikam et al, 1984) 1 or 2 operons respectively. The reports describing sequence data for the 16S-23S spacer region all include only a part of the total number of rRNA operons per genome, including E. coli demonstrating 5 spacers (Harvey et al, 1988), B. subtilis, 2 spacers (Loughney et al, 1982), Pleisomonas

sbiσelloides, 3 spacers (East et al, 1992), Aeromonas bydropbila, 3 spacers (East & Collins, 1993), Caulobacter crescentus (Feingold et al, 1985), Acboleplasma laidlawii (Nakagawa et al, 1992) and Enterococcus birae, (Sechi & Daneo-Moore, 1993) 2 spacers respectively.

International Patent Publication No. WO 91/16454 by N.B. Innogenetics S.A. describes use of hybridization probes consisting of at least 15 nucleotides from the spacer region between rRNA genes of non-viral organisms for detection of non-viral microorganisms, particularly

bacteria. The probes are species-specific, and are

preferably 15 to 100 nucleotides of the spacer region between the 16S and 23S rRNA genes. A separate

oligonucleotide probe is required for each microorganism species.

U.S. Patent No. 5,288,611 by Kohne describes methods and probes for identification and quantification of any organism or group of organisms containing rRNA, including previously unknown organisms. Probes specific for individual species and for groups of related species, and Probes hybridizing to rRNA or to tRNA are described.

U.S. Patent No. 5,292,874 by Milliman discloses hybridization probes specific for Stapbylococcus aureus probes, which detect a unique rRNA sequence in the 23S rRNA gene.

Japanese Patent Publication No. 6090793 by Takara Shuzo Co. Ltd. describes methods for detection of bacteria of the genus Lactobacillus, by detection of a sequence in the spacer region between the gene encoding 16S rRNA and the gene encoding 23S rRNA.

The disclosure of each of these patent specifications is incorporated herein by reference. In each case sequences within the spacer region were selected on the basis of specificity for the organism from which they were isolated, and used in hybridization assays or polymerase chain reaction (PCR) for specific identification of an organism. Each of U.S. Patent No. 5,288,611, U.S. Patent No. 5,292,874 and Japanese Patent Publication

No. 6090793 requires a separate oligonucleotide for each species of organism. None of these specifications mentions the existence or number of rrn alleles, or describes a method permitting differentiation of strains within a species, or of allelic variations.

International Patent Publication No. WO 93/11264 by E.I. Du Pont De Nemours & Company, the contents of which are incorporated herein by reference, discloses a method for identification of microorganisms by amplification of hypervariable spacer regions between highly conserved sequences encoding rRNA. These spacer region sequences are amplified, using primer sequences; the same pair of primers is used for all species of microorganisms; the sequences of these primers are highly conserved among prokaryotic organisms. The products of amplification are

characteristic of a given species. A further amplification step using arbitrarily primed polymerase chain reaction (AP-PCR; also known as randomly amplified polymorphic DNA; RAPD) is described as being able to differentiate serotype or strains within a species. However, it is evident that this method did not always enable differentiation between strains, and even if this differentiation was achieved, the patterns were not always clear. The conserved regions used as primers in WO 93/11264 are designated E and A herein, as described below. These findings have also been published elsewhere (Jensen et al, 1993; Jensen & Straus, 1993). The contents of these publications are also incorporated herein by reference.

PCR analysis of the 16S rRNA gene has been used to demonstrate species-specific differences (Gürtler et al, 1991) and strain differences (Vaneechoutte et al, 1992) in various bacterial species. Allelic species-specific differences within the 16S rRNA gene have been demonstrated in clostridia (Gürtler et al, 1991). The rRNA alleles of E. coli (Brosius et al, 1981) and B. subtilis (Loughney et al, 1982) have been shown to have variable length

16S-23S rRNA spacer regions.

We have now surprisingly found that the presence or absence of specific variable length rDNA spacer regions varies between strains within a given microorganism

species. By using different specific conserved regions of the rRNA operon, designated C and D herein (see Figure 1 below), as primers for polymerase chain reaction, and by using modified PCR conditions, we can achieve amplification of all alleles present in a microorganism sample, and thus we can differentiate between strains in any bacterial species, without the need for any further steps. The patterns obtained were stable within strains on repeated testing, using passaging either in vitro or in vivo, permitting discrimination within and between species, and designation of specific types within strains. Brief Description of the Drawings

Figure 1 shows the position of the

oligonucleotide primers (A, B, C, D, E) used in the prior art and in the present invention. The abbreviations ile and ala refer to the respective genes-encoding tRNA for isoleucine and alanine. Primers A and E are as disclosed in WO 93/11264.

Figure 2 illustrates the approaches used for the detection of rRNA alleles in C. difficile by Southern hybridization and PCR. The hatched bars (A, B, C and D) show positions of the respective PCR products (Table 3), the shaded bar denotes the 16S rRNA gene, the solid bar denotes the 23S rRNA gene, and the line joining the 16S and 23S gene depicts the spacer regions. The HindiII site is at position 975 of the 16S rRNA gene (Gürtler et al, 1991).

Figure 3 shows hybridization of PCR product B to Group II bands in genomic DNA isolated from C. difficile and C. bifermentans strains.

Lanes 1-10, C. difficile strains H13, H15, H16, H17, H18, H19, H20, H23, 9689 and 9689, respectively;

Lanes 11-13, C. bifermentans strains AM312,

AM360, and AM818, respectively;

Lane 14, pBR328 DNA digested with BgII and HinfI, labelled with photodigoxigenin;

Y indicates the position of an extra band visible in C. bifermentans products.

Figure 4 illustrates the hybridization of PCR product A to Group I and II bands in genomic DNA isolated from C. difficile strains.

Lane 1, pBR328 DNA digested with BgII and HinfI, labelled with photodigoxigenin;

Lanes 2-11, H24, H25, H26, H27, H28, H29, H30, H31, H32 and H33, respectively.

Figure 5 shows the detection of rRNA alleles in C. difficile strains by Southern hybridization. The symbols, box shadings and the position of the HindIII site are described in the legend to Figure 2. Bands depicted as Group I (Figures 3 and 4) correspond to fragments 5' of the HindiII site and Group II bands (Figures 4 and 5)

correspond to fragments 3' of the Hindi11 site, ★ refers to bands which are not present in all strains.

Figure 6 shows the constant and variable length regions within PCR product C amplified from C. difficile strains, as demonstrated by agarose gel electrophoresis of undigested (lanes 2-7) and HindIII-digested (lanes 10-15) PCR product C.

Lane 1, pBR328 DNA digest with HinfI and BglI ;

Lanes 2-7, H15, H24, H28, H30, H31 and H33, respectively;

Lane 8, no DNA control;

Lane 9, pBR328 DNA digested with HinfI and BglI ; Lanes 10-15, h13, H14, H17, H19, H23 and 630, respectively.

The standards are 2176, 1766, 1230, 1033, 653, 517, 453, 394, 298, 234 and 220 bp respectively.

C_d, constant HindIII-digested;

V_u, variable undigested; and

V_d, variable HindIII-digested.

Figure 7 shows denaturing PAGE of PCR product C amplified from strains of C. difficile.

Lanes 1-8, H17,H15, H14, H13,, 6989, 630 H23 and H19, respectively;

Lane 9, no DNA PCR control

Lanes 10-19, H33, H32, H31, H30, H29, H28, H27, H26, H25 and H24, respectively.

The sizes of the respective alleles are shown on the left [mean ± SEM (number of determinations)]. The molecular mass markers used (not shown) were λ DNA digested with HindIII and EcoRI (947 and 831 bp bands only) and SPPI DNA digested with EcoRI (1150 and 1000 bp bands only).

Figure 8 is a dendrogram showing the

relationships of C. difficile ribotypes. Using maximum parsimony, 50 equally parsimonious trees were found, one of which is shown. The same ribotypes were found in each of the circled branches (a, b, c) for all 50 trees. The root of the tree (C. bifermentans) had no bands in common with any of the ribotypes.

Figure 9 shows the distribution of rRNA genes and restriction sites in the region of interest in

Staphylococcus aureus. The solid line joining these genes can vary in length in the same strain or in different strains (Figure 12b). At the bottom of the diagram, the dashed lines show positions of the PCR products C, I & J, which were obtained using the primers R1392F and LR488, SP1F and SP2R and SP1F and LR20F respectively (Table 8). The dotted lines show the origins of HpaII fragments (E, F, G and H) obtained from PCR product C. The locations of other primer binding regions that were used to sequence HpaII fragment E are also shown.

Figure 10 illustrates denaturing PAGE of PCR products amplified from strains of S. aureus.

(a) Methicillin resistant S. aureus (MRSA) ribotype A, (lanes 1-3), and ribotype B (lanes 4-8). The sizes of the respective alleles are shown on the right

[mean ± SEM (number of determinations)].

(b) All MRSA ribotypes A-I (excluding C) and penicillin-sensitive ribotype Pa (ATCC 9144).

(c) Penicillin and methicillin sensitive

S. aureus strains.

Lanes 1-6: Ribotypes Pi, Pj, PF A(strain H11);

Lanes 9-16: B, Mi, Mh, Mi, Pi, Mh and Mj.

The sizes of the respective alleles are shown on the right [mean ± SEM (number of determinations)]. The molecular mass markers (lanes 7 & 8) used were λ DNA digested with HindIII and EcoRI (1375 and 947 bp bands only) and SPP1 DNA digested with EcoRI (1150 and 1000 bp respectively).

Figure 11 shows the alignments of 16S-23S spacer sequences from S. aureus. PCR product C from S. aureus strains (Table 9) was cloned into M12mp18RF and sequenced with the primers listed in Table 8 and Figure 9. The sequences were derived from the clones and isolates listed in Table 7. The sequences SA16S and SA223S were taken from Ludwig et al (1992). The alignment of rrn alleles with SA16S (a), rrnC, E, F, G, H, J, K & L (b) and rrn alleles SA23S (c) is shown. The symbols refer to an identical base (.), and absent base (-), † = (rrnA, E, J, L, F),

" = (clones 4, V17, V32) and § = (clones 4, V4, V8, V32, V43).

Figure 12 is a dendrogram showing the

relationships between the 16S-23S alleles using the data in Figure 13b or Table 7. The same tree was obtained using either sets of data. One parsimonious tree was obtained with the program DNA PENNY from the PHYLIP package. Then, using MACCLADE, of the 16 possible rerootings, the tree shown was selected because it was drawn using the longest allele (rrnA) as the root. The numerals indicate the numbers of changes/branch. The dotted lines show the clades which contain alleles without tRNA genes, and the solid lines show the clades which contain alleles with tRNA genes.

Summary of the Invention

The method of the invention avoids cumbersome steps required by previously available methods, and is suitable for testing large numbers of samples; it is also amenable to automation.

The method of the invention is particularly suitable for epidemiological studies, for example

identifying the source of hospital outbreaks of antibioticresistant microorganisms, or tracing the source of

microorganisms causing contamination of foodstuffs.

According to a first aspect, the invention

provides a method of identification of microorganisms, comprising the steps of extracting and purifying DNA from a sample suspected to contain bacteria, and subjecting the 16S-23S rRNA spacer region of said DNA to amplification, using a first primer comprising a sequence from the 5' end of the 16S rRNA gene, and a second primer comprising a sequence from the 3' end of the 23S rRNA gene, thereby producing fragments having detectable differences in size and number, and separating the amplified fragments.

Because the amplified fragments produced in the method of the invention are of variable length, they can be analysed directly, for example by electrophoresis; no other experimental step, such as hybridzation, is necesary in order to demonstrate differences between strains, although in some situations a hybridization step could be

advantageous.

The amplified products may be separated by any method which provides sufficient resolution. We have found that small conventional polyacrylamide gels have somewhat poor resolution, and our studies have used long denaturing polyacrylamide gels. However, the person skilled in the art will be aware that other separation methods, such as capillary electrophoresis or high performance liquid chromatography, may be used.

Similarly, the studies described herein have employed amplification of DNA sequences by polymerase chain reaction. However, the skilled person will be aware that other amplification methods are known, and may also be used. For example, ligase chain reaction, 3SR

amplification, strand displacement amplification,

Qβ replicase reaction, or branched DNA signal amplification are available (see Wolcott, 1992 for review). The skilled person will readily be able to test these alternative methods for suitability for use in the invention.

Optionally additional probes may be used, for example comprising the sequence encoding tRNA^ile and/or the sequence encoding tRNA^ala.

The sample will usually be a clinical sample such as blood, tissue, urine, faeces, sputum etc., a food sample, or an environmental sample such as a water sample or a soil sample. Other types of samples may be used, depending on the circumstances. The DNA may be extracted by any suitable method, but preferably the method is a rapid one. Extraction with guanidine hydrochloride or by boiling water followed by column purification are both suitable. In some cases, particularly where clinical specimens are used, it may be advantageous to effect a preliminary purification of the sample following DNA extraction. If the nature of the bacteria sought to be tested is known, this may be carried out using monoclonal antibody methods, such as those using monoclonal antibody conjugated to magnetic beads . Some broad spectrum antibodies are also available for this purpose.

While the method of the invention is specifically described with reference to Clostridium difficile and to Stapbylococcus aureus, it will be clearly understood that the invention is not limited to these organisms, and is applicable to any microorganism for which the sequences of the 16S rRNA gene and the 23S rRNA gene are known. These sequences enable suitable probes to be designed.

Preferably the primers used correspond to a highly conserved region from the 3' end of the spacer region, and to a highly conserved region from the 5' end of the spacer region respectively.

According to a second aspect, the invention provides amplification primers for use in the method of the invention. As described above, these primers correspond to highly conserved regions from the 3' end of the

16S-23S rRNA sacer region and the 5' end of the

16S-23S rRNA spacer region respectively. Preferably they correspond to regions from the 5' end of the 16S rRNA gene and to a region from the 3' end of the 23S rRNA gene respectively. Preferably the primers are 15 to 20

nucleotides long, since 15 nucleotides is generally

considered to be a minimum length of primer for PCR, but conservation is generally lost at greater than 20

nucleotides. In a preferred embodiment the primers are R1391F and LR488 or LR194F as herein defined. Most preferably LR488 is 15 to 19 nucleotides long, and R1391F is 15 to 18 nucleotides long. Primer C (LR488) is particularly preferred, because it is more highly conserved than primer A.

We have found that in fact there are three regions in the first 520 base pairs of the 23S rRNA gene which are highly conserved through a wide variety of species of bacteria and fungi, and which are useful as amplification probes in the present invention. These are summarised in Table 1.

Detailed Description of the Invention

The invention will now be described by way of reference only to the following non-limiting examples, and to the drawings referred to above Bacterial strains and their Cultivation

Bacterial strains used herein are listed in

Tables 2 and 3.

The identity of all strains of Clostridium difficile was determined by biochemical tests (Cato et al, 1986) and confirmed by gas-liquid chromatography (Sutter et al, 1985). Purified stocks were stored in cooked-meat broth at room temperature or in glycerol broth at -20°C. All strains were grown in brain heart infusion broth (BHI, Gibco). The stability of ribotype patterns was tested by passaging single colonies from horse blood agar plates every 2-3 days over a 5 week period. Toxin B production by C. difficile strains was detected by the method of

Boondeekhun et al (1993).

The identity of all S. aureus strains was determined by biochemical tests (Kloos & Schleifer, 1986), and antibiotic sensitivity tests were assessed by the agar dilution method (break points were determined according to NCCLS guidelines, Vol 13 No. 25, 1993). Purified stocks were stored in glycerol broth at -70°C. All strains were grown in trypticase soya broth (TSB,Oxoid). The stability of ribotype patterns was tested by passaging single

colonies from sheep blood agar plates five times a week for six weeks.

The numbers of strains isolated from various locations at various times are shown. The strains from Ireland are listed in Townsend et al (1987) as WG1761-3 (plus 6 other strains) while those from RFH are listed in Townsend et al (1984) as WG2710, 2715, 2720 and 2724 and in Townsend et al (1987) as WG2716 (plus 3 other strains) NCTC National Collection of Type Cultures, UK

ATCC American Type Culture Collection HRH Heidelberg Repatriation Hospital

RMH Royal Melbourne Hospital

RCH Royal Children's Hospital, Melbourne

RFC Royal Free Hospital, London

* including H11, H12, H14

§ including D46

† including H21

DNA isolation and amplification

Genomic DNA and plasmid DNA was isolated by the protocol of Gürtler et al (1991), except that the cell walls of S. aureus were disrupted by incubating the strain with 200 g lysostaphin ml^-1 at 37°C for 5-10 min. DNA regions (Figures 1 and 9) were amplified by the protocol of Gurtler (1991), except that the reaction volume and amount of DNA were halved, and 1.25 units Taq polymerase

(Boehringer) were used. The primers used are shown in Tables 4 and 7 below. For the amplification of M13 clone inserts in S. aureus, 50-100ng of single stranded Ml3 clone DNA was added to PCR mixtures using primers M13F and R. Restriction enzyme analysis

For C. difficile, purified PCR products R907-LR507 and R1391-LR507 were digested singly or doubly with 10-15 units HindiII and CfoI , as instructed by the manufacturer (Boehringer). Genomic DNA was digested with 30 units HindIII. The digested and undigested PCR products were resolved on 2% (w/v) low-gelling-temperature plus 2% (w/v) 'AR' agarose gels. The HindIII-digested genomic DNA was resolved on 1% (w/v) 'AR' agaorse gels.

For S. aureus, PCR product M13F-M13R was digested with 10-15 units Dral or HinfI , as instructed by the manufacturer (Boehringer). Genomic DNA was digested with 20 units HpaII. The digested PCR products and genomic DNA were resolved on 2% w/v low-gelling-temperature plus

2% (w/v) 'AR' agarose gels. Southern hybridization

The protocol of Gürtler et al (1991) was followed. PCR products were labelled with digoxigenin.

Denaturing PAGE

The amplification protocol described above was followed with some modifications. The reaction volume was decreased by a factor of two and 2 μCi [α-³²P]dATP (DuPont or Amersham) was added. The reduction of the unlabelled dNTPs by a factor of four increased the yield of labelled product. Radiolabelled DNA fragments were separated on a 0.4 mm thick, 38 cm wide and 50 cm high (Bio-Rad),

3.5% (w/v) polyacrylamide gel containing 7 M-urea (Sambrook et al, 1989). Gels were dried in a vacuum slab gel drier (Bio-Rad) for 2 h at 80°C. Autoradiographic exposure was 18-96 h.

DNA sequencing

Sequencing was performed by the dideoxynucleotide method of Sanger et al (1977) using the Bst DNA sequencing kit (BioRad). 7-Deaza-2'-deoxyguanosine triphosphate was used to minimize band compression due to GC-rich regions.

The primers used for sequencing are listed in Table 8 below and in Figure 9.

Data analysis

DNA sequences were processed and analysed by the following methods. The DNASIS program (version 6;

Pharmacia) was used to orient, join and edit DNA sequences. The orientation of inserts was deduced by alignment with the 16S or 23S rDNA seqences from B. subtilis (Green et al, 1985) or S. aureus (Ludwig et al, 1992). The 17 sequences (fragment E, Figure 1) were aligned using CLUSTAL V

(Higgins et al, 1992) and were aligned to the 16S and 23S rDNA sequences from S. aureus (Ludwig et al, 1992).

Further modifications to the alignment were done using MACCLADE software (Maddison & Maddison, 1992). Phylogenetic analysis was done with the software package PHYLIP using DNAPENNY (Felsenstein, 1993). The resulting treefile was then imported into MACCLADE for further analysis and presentation.

The presence or absence of PCR product C bands was analysed as follows:

(1) The presence or absence of bands (corresponding to region R1391-LR507) on autoradiograms was analysed by using the program Biolmage (Millipore). The average sizes of the 16 alleles (rrnA-P) were calculated from five separate gels ranging from 1-51 determinations for the respective alleles. Using these sizes as internal standards, molecular masses were assigned to respective bands from the different strains. Twenty-four strains from four gels were then compared at once. Presence or absence of bands was scored by a 1 or 0, respectively.

(2) The resulting data matrix prepared from four gels was analysed by maximum parsimony (Swofford, 1985).

Intense bands were reported as positive; when the faint bands were also reported as positive the results did not change. The resulting data matrix was analysed using MIX and DOLLOP in the program PHYLIP.

Example 1 Ribotyping of Strains of Clostridium

difficile

The DNA typing approaches used are shown in

Figure 2 and Table 4. Products A and B were hybridized to HindIII-digested genomic DNA isolated from C. difficile and C. bifermentans strains. Differences in HindIII sites on both flanking sides of the 16S rRNA gene were sought within and between strains. Products C and D were amplified from C. difficile strains in an attempt to find differences in the length of the 16S-23S spacer region within and between strains.

The products amplified cover the regions shown where S=16S rRNA gene and L=23S rRNA gene. The sense of the primers used is shown by R=reverse, F=forward and *=identical region, with R1391 being the complemnt of R1391F. The positions of all the primers are in regions which are highly conserved in eubacteria (Neefs et al, 1990; Guttell & Fox, 1988). The nucleotide numbering system is that of E. coli operon (Brosius et al, 1978). The positions of each product are schematically represented in Figure 2.

The bands detected by Southern hybridization (ribotyping) have been divided into Groups I and II, showing numerous Group II differences between strains and fewer Group I differences. Ribotyping of 21 isolates of C. difficile from the Heidelberg Repatriation Hospital and one from St Vincent's Hospital, Melbourne, Australia, produced 14 restriction fragment length polymorphism (RFLP) types, 10 of which are shown in Figure 4. There were 10 group I bands, demonstrating 10 rRNA alleles in

C. difficile.

Products A and B consist only of parts of the 16S rRNA gene (Table 4). We have shown previously that the 16S rRNA gene is of constant length between alleles and strains of C. difficile (Gürtler et al, 1991). When PCR product B was hybridized to C. difficile genomic DNA, Group II bands hybridized predominantly (Figure 3), the Group I bands hybridized faintly, because product B included 62 bp 5' of the HindIII site (1/10 of product B). When product B was hybridized to C. bifermentans genomic DNA digested with HindIII (Figure 3), no Group I bands hybridized and an extra band appeared (Y) due to an extra HindIII site at position 675 of the 16S rRNA gene (Gürtler et al, 1991). When PCR product A was used as a probe (Figure 4), the Group I and II bands hybridized with equal intensity. The orientation of the Group I and II bands is as shown in Figure 5 because product B hybridized predominantly to Group II bands and because the HindIII site lies 62 bp downstream from the 5' end of product B (Gürtler et al, 1991).

Example 2 Identification of a Variable Length Region

Between the 16S and 23S rRNA Gene of

Clostridium difficile

From Figures 3, 4 and 5 it can be seen that the Group II bands consist of the spacer region and part of the 23S rRNA gene. These Group II bands were of variable length, which could be explained by the presence of either a variable HindIII site or of an insertion within the spacer or the beginning of the 23S rRNA gene. To determine which possibility was correct, we amplified PCR products C and D, both of which include the spacer regions (Table 4). When the product C primer combinations of Table 4 were used, several bands (V_u) of varying molecular masses were obtained from each C. difficile strain, as shown in

Figure 6. The presence of bands varied from strain to strain. When V_u bands were digested with HindIII, a band appeared at 430 bp (C_d); this was of higher intensity than the digested variable length bands, V_d. Band C_d appeared in all the strains listed in Table 2 (results not shown). When the product D primer combination was used, the same HindIII band (C_d) was present, demonstrating that this band contains the 23S rRNA gene from position 80-507 (Table 5)5. The demonstration of band C_d shows that this region is of constant length between alleles. Taken with the Southern hybridization data of Figures 3 and 4, these results show that the variable length regions lie between the 16S and 23S rRNA genes. The exact base pair location of the spacer regions can only be determined when the separate alleles have been sequenced.

Example 3 Improving Resolution of Bands

The resolution of the variable length bands was low, as can be seen in Figure 6, and so it was decided to increase the resolution with long denaturing polyacrylamide gels. When this was done, the same amplification products, designated V_u in Figure 6 separated into between 5 and 9 bands per strain, with the presence of bands variable between strains. These results are shown in Figure 7.

Each band was assigned as an allele, resulting in a total of 16 alleles (A-P) of variable length. The constant length regions within the 16S and 23S genes were partially characterized, and the results are summarized in Table 5. The variability in length was due to variable length

16S-23S spacer regions between alleles.

PCR product C was amplified from various C.

difficile strains and separated by denaturing PAGE

(Figure 7). The presence of variable length alleles (rrnA- P) is shown. The size of each allele is shown in Figure 7. The outer limits of the constant regions are depicted by restriction enzyme-cut sites (see Figures 5 and 6). The strain numbers corresponding to the ribotype are listed in Table 2. The number of isolates in each ribotype is listed below each letter. Th constant length regions were

collated from results obtained in Figures 3, 4 and 6 and Gϋrtler et al (1991). The variable length regions wree collated using Biolmage software from Figure 7 and three other denaturing polyacrylamide gels.

Example 4 Relationship Between Ribotypes of

Clostridium difficile

When all of the C. difficile strains listed in

Table 2 were analysed, 24 strains were divided into 14 ribotypes, which are also shown in Table 5. The dendrogram depicted in Figure 8 shows that 3 clusters (a, b, c) are found in all trees analysed. Within ribotype G, 2 isolates were cultured from one patient at different times; within ribotype E, 3 isolates were cultured from one patient at different times. All other isolates which had identical patterns (ribotypes D, E, F, G and H) were from different patients.

Example 5 Stability of Band Patterns in ciostridinm difficile

The stability of band V_o sizes and patterns was investigated in detail by passaging five strains over a 5 week period. The alleles were scored as positive or negative by appearing visually identical and by having similar calculated molecular masses. The results,

illustrated in Figure 7 and Table 5, show that both the band sizes and patterns were highly reproducible in five C. difficile strains. The band sizes and patterns of strains H23 and 630 were reproducibly identical.

Product C was amplified from various C. difficile strains and separated by denaturing PAGE. Accumulated values (mean±SEM) taken from five separate electrophoresis runs are shown. The data include runs (PCR, DNA

preparations and electrophoresis) done over a 9 month period, as well as a stability testing experiment with the number of single colony passages per strain shown.

The main finding of the present study was that the presence or absence of specific variable length rDNA spacer regions varied between C. difficile strains. The patterns obtained were stable within strains upon repeated testing after passaging in vitro and in vivo, allowed the designation of strains to specific types, enabled

discrimination within and between species, and allowed for the easy testing of large numbers of strains. Thus the novel molecular typing method of the invention may be applied to epidemiological studies of C. difficile. Since 500 bp of the 5' end of the 23S gene was amplified, it was possible that the observed heterogeneity of PCR products was due to an insertion within the first 500 bp of the 23S gene. This possibility is supported by several findings. At least one extra cleavage site has been reported in the large rRNA subunit of Leptospira interrogans (Hsu et al, 1990) and Salmonella species (Hsu et al, 1992), producing several fragments smaller than 23S and a 90 bp intervening sequence has been shown to be excised during large subunit rRNA maturation (Burgin et al, 1990). The results presented in this specification show that in C. difficile, 430 bp 3' from position 507 of the 23S rRNA gene was of constant length and the 16S-23S spacer DNA was of variable length between alleles.

The 16S-23S spacer regions of B. subtilis (Void,

1985) and E. coli (Fournier & Ozeki, 1985) contain tRNA genes which vary in length from 75-90 bp. Of the 7 rrn operons in E. coli, all contain from 1-3 tRNA genes

(Brosius et al, 1981), while in B. subtilis operons, two out of the three analysed sets of the 10 rrn have been shown to contain tRNA genes (Loughney et al, 1982). It is possible that the 16S-23S spacer regions in C. difficile characterized herein may contain tRNA genes.

Example 6 Ribotyping of strains of Stapbylococcus

aureus

Genomic DNA was isolated from S. aureus as described above, and amplified using the primers described in Table 6 and Figure 9.

Numbering according to;

*, the published 23S rRNA sequence from S. aureus (Ludwig et al, 1992);

†, the aligned spacer sequence for S. aureus (Fig. 3); ¶, the published 16S rRNA sequence from S. aureus (Ludwig et al, 1992);

§, the published sequence for bacteriophage M13;

I, sequence obtained by the present inventor;

‡, addition of Hpa II sites (underlined) at 5' end of

R1391F & LR488.

Letters in brackets indicate the designations of primers as given in Figure.1. As stated above, primers A and E are as disclosed in WO 93/11264.

Using the DNA typing approach described above and the primers LR1391F, LR1392F and LR488 as shown in Figure 9, PCR product C was amplified from the S. aureus strains listed in Table 3. These included 281 MRSA from four geographically distinct clinical sources and 48 penicillin or methicillin sensitive strains from a single Melbourne source; several methicillin-resistant or sensitive strains from type culture collections were also used. These strains yielded various amplified products, of which only the most intense bands were considered to be alleles. In total, 15 alleles, designated rrnA to rrnO, were

recognised, with 14 varying in length from 935 to 1223 bp, as shown in Figure 10. rrnO was 906 bp in length (results not shown).

From Figure 10A, it can be seen that among the strains, two ribotypes, A and B, that were highly

reproducible in individual isolates were obtained, with 104 and 174 strains respectively (including 5 type strains). An additional 7 ribotypes were found among the remaining 9 MRSA strains; Figure 10b shows 8 MRSA ribotypes. Ribotype A was the major ribotype found between 1960 and 1989 in Melbourne (19/22 strains), Singapore (7/9 strains), Ireland (9/9 strains). New York (1/1 strain) and UK (12/12 strains). After 1989, ribotype B was the major ribotype found at the Heidelberg Repatriation Hospital; 176 were ribotype B and 57 were ribotype A.

In contrast to the MRSA strains, the sensitive strains showed considerably more variation in the presence or absence of bands, yielding an additional 26 ribotypes from the 48 strains studied. Figure 10c shows some of these strains. The MRSA ribotypes A, B and I included some of the penicillin or methicillin sensitive strains. The occurrence of the alleles in the various ribotype classes is summarized in Table 8.

PCR product C was amplified from various S.

aureus strains and separated by denaturing PAGE (Figure 10). The presence of variable length alleles (rrnA-O) is shown. The size of each allele is shown in Figure 10. The data were collated (using Biolmage Software) from Figure 10 and four other denaturing polyacrylamide gels.

† including H11; one isolate was penicillin-sensitive and another was methicillin-sensitive;

§ including D46 and H12; one isolate was methicillinsensitive;

¶ five isolates were methicillin-sensitive;

* including H14;

‡ including H21.

Example 7 Stability of Ribotypes of Stapbylococcus aureus

The stability of ribotypes A, B, C, D and Pa was investigated by 30 serial passages of strains 9144, H11, H12, H14 and H21 over a six-week period. The ribotype was assessed after every fifth passage by visual comparison with reference patterns, and was found to be stable except for strains H12 and H21. Strain H12 was identified as ribotype B at all passages except the fifth, where rrnL appeared, making it ribotype A. Strain H21 was originally found to be ribotype D; however during the stability experiment it was found to be ribotype A at all passages subsequently investigated (the colonies from which the DNA prepared were used completely). After plating out colonies from the original frozen stock, genomic DNA was prepared from 10 separate colonies of strain H21: in all cases the ribotype was found to be A. The instability of strains H12 and H21 could be explained by a contaminant or by the rearrangement, duplication or deletion of an rrn operon to yield rrnL or rrnM respectively. This instability did not affect the typing of S. aureus strains significantly, since it was an infrequent event.

The instability of rrn operons has been reported in B. subtilis (Widom et al, 1988) and E. coli (Hill & Harnish, 1982). These reports show evidence for the loss of an rrn operon, and there is also evidence for

recombination leading to chromosomal rearrangement of rrn operons (Hill & Harnish, 1982). The instability of

ribotype D (strain H21) could be explained by the loss of rrnH giving rise to the stable ribotype A. The relative instability of ribotype B in the present study was due to the loss of rrnL. The frequency of such events is low

[~10^-4; (Hill & Harnish, 1982)] and thus will have little practical effect on this method.

Example 8 Sequencing of Variable Length 16S-23S rRNA

Alleles

In order to isolate and compare the variable length sequences of the 16S-23S rRNA alleles, PCR product C from strains D46 (ribotype B), H11 and ATCC33952 (both ribotype A) was cleaved with HpaII and the resulting fragments cloned into M13 vectors. PCR product C

(Figure 1) was amplified as described above from S. aureus genomic DNA from strains D46, Hll, and 33952 using primers R1391F and LR488, R1391FH and LR488H or R1391F and LR194F

(Table 7) . For each strain, ten equivalent reactions were pooled, precipitated with 26% w/v polyethylene glycol in 20mM MgCl₂ (Paithankar & Prasad, 1991) and digested with Hpall. For D46, 1 g of the HpaII digested products were end repaired with 4 units of T4 DNA polymerase

(Boehringer), 200μM dNTPs, 33mM Tris-acetate (pH 8.0), 66mM potassium acetate, 10mM magnesium acetate, 0.5 mM

dithiothreitol, 0.1mg bovine serum albumin ml^-1 and

incubated at 11°C for 30 minutes. For 33952 and H11 the HpaII digested products (1-25ng) were ligated directly into AccI digested M13mp19RF (50ng: in a total of 10μl) and the end repaired HpaII digested products (1-25ng) from strain D46 were ligated into Smal digested M13mp19RF (50ng: in a total of 10 μl) with 1 unit T4 DNA ligase (Boehringer), 66mM Tris-HCl (pH 7.5), 5mM MgCl₂, ImM dithiothreitol,

1mM ATP and incubated at room temperature overnight. The competent JM109 E. coli cells (50μl; Promega) were

transformed with 2-3μl of the ligation mixtures according to the protocol described in Sambrook et al (1989). After plates were incubated overnight at 37°C, bacteriophage M13 plaques were either picked off and grown in Luria Broth (LB) or colony hybridizations (Sambrook et al, 1989) to HpaII digested PCR product C labelled with digoxigenin were performed. Positive plaques were then picked off and grown in LB (Sambrook et al, 1989). Single stranded

bacteriophage M13 DNA was then prepared from all the positive clones (Sambrook et al, 1989). To determine the presence and size of the inserts, the single stranded DNA from the M13 clones was used as a template in the PCR using M13F and R primers which flank the SmaI and Accl

restriction sites (Yanisch-Perron et al, 1985). Four clones were isolated from D46, one containing insert E, one containing insert F, one containing insert G and one containing insert H. Sixteen clones, designated V2-V17 were isolated from H11; seven contained insert E, four contained insert H, four contained insert G and one

contained insert F. With the sequence information of insert (H), primer LR194F was designed so as to contain a HpaII site. For strain 33952, primers LR194F and R1391F were used to obtain a mixture of PCR products, which were digested with Hpall to yield predominantly product E and cloned into Ml3mpl9; clones V18-V47 were isolated, and of these 9 contained insert E. These results are summarized in Table 9.

It was confirmed that the inserts were of variable length when amplified by PCR and then digested with Dral or HinfI (results not shown; restriction enzyme cleavage sites are indicated in Figure 1). It was

determined from the first insert sequenced (clone 4) that HinfI digested at the and of the tRNA^ile gene (Figures 1, 11b). All the PCR products which were digested with HinfI contained a tRNA^ile gene, and this was subsequently

confirmed by DNA sequencing, as shown in Figure lib.

The 16S-23S rDNA spacer sequences of 9 rrn operons were determined from 3 methicillin resistant

S. aureus strains. The variation in 16S-23S spacer length (303 bp to 551 bp) was accounted for by the type (tRNA^ala or tRNA^ile) and number (one, both or none) of tRNA genes, and by the presence or absence of other sequences of unknown function.

Example 9 Designation of Alleles

The designation of alleles set out in Table 9 was made by direct correlation with fragment C molecular weights (Figure 10A). The length of the spacer varied from 303 bp to 551 bp. The fragment E insert sequences were aligned to the 16S, 23S and 16S-23S spacer rDNA sequences, as shown in Figures 11a, lie and lib respectively. There were only 4 base pair differences in the 16S rDNA

sequences, in contrast to 71 base pair differences in the

23S rDNA sequences. In the 16S-23S spacer rDNA (Figure lib and Table 9) there were no differences in CS1 and CS2;

however, there were striking gaps between alleles in regions VS2, tRNA^ala, VS3, VS4, VS5, VS6, VS7, and VS8.

The tRNA^ile gene was present in rrnJ,G,F,C and A, while the tRNA^ala gene was only present in rrnA and C. The number of base pair differences between clones judged to be the same allele was 5 for rrnJ (from strains H11, D46 and 33952, isolated in 1982, 1992 and 1981 respectively), 2 for rrnH (from strains 33952 and H11), 4 for rmF (strain 33952) and 5 for rraC (33952). using phylogenetic analysis, the rrn alleles were divided into 3 distinct groups, which are shown in

Figure 12; "a" contains tRNA^ile and tRNA^ala, "b" contains tRNA^ile only and "c" contains no tRNA genes.

To confirm the presence of variable length

16S-23S spacer regions in genomic DNA, PCR products I and J were hybridized to genomic DNA isolated from S. aureus isolates. Between 4 and 7 bands ranging from -600 bp to ~850 bp were obtained for all strains, with variation between strains. The results are in close agreement with the results obtained with PCR-ribotyping (Figure 10) and DNA sequencing (Figure 11 and Table 9).

The type of tRNA gene found in the 16S-23S spacer varies in number (0, 1 or 2) and combination between operon and between species; A. bydropbila (East & Collins, 1993) and E. coli (Morgan et al, 1977) have tRNA^ala, tRNA^ile and tRNA^glu; B. subtilis (Loughney et al, 1982) and

C. crescentus (Feingold et al, 1985) have tRNA^ile and tRNA^ala; P. sbigelloides (East et al, 1992) has tRNA^fllu;

Metbanococcus vanielli (Jarsch & Bδck, 1983) and

Enterococcus birae (Sechi & Daneo-Moore, 1993) have tRNA^ala; and Mycobacterium bovis has no tRNA genes (Suzuki et al, 1987). The length of the spacer varies from 156 bp for M. vanielli (Jarsch & Bδck, 1983) to 551 bp (present specification). We have now found that there is

intraspacer and interspacer variation of other sequences besides tRNA genes.

Our results show that variable length 16S-23S spacer regions occur in genomic DNA whose size range is similar to the results obtained with PCR-ribotyping

(Figure 10) and DNA sequencing (Table 9). The majority of strains presented in this study were from the Heidelberg Repatriation Hospital (274 strains of a total of 322).

Among the MRSA strains, two ribotypes (A and B) that were highly reproducible in individual isolates were obtained (Figure 10a), with 101 and 180 strains respectively

(including 5 type strains). An additional 7 ribotypes were found among the remaining 9 MRSA strains (Figure 10b shows 8 MRSA ribotypes). Ribotype A was the major ribotype found (ribotype A/total no of strains in location) between 1960 and 1989 in Melbourne (19/22), Ireland (9/9), New York (1/1) and UK (12/12). After 1989, ribotype B was the major ribotype found at the HRH (176 were ribotype B and 57 were ribotype A). In contrast to the MRSA strains, the

sensitive strains showed considerably more variation in the presence or absence of bands, yielding an additional 26 ribotypes from the 48 strains studied (Figure 10c shows some of these strains). The MRSA ribotypes A, B and I included some of the penicillin or methicillin sensitive strains. The occurrence of the alleles in the various ribotype classes is summarized in Table 7.

Thus we have shown that the presence or absence of specific variable length rDNA spacer regions varies between S. aureus strains. The patterns obtained were mostly stable within strains upon repeated testing, allowed the designation of strains to specific types, discriminated within the species S. aureus, and allowed for the easy testing of large numbers of strains. With these criteria met, the molecular typing method described here is useful for epidemiological studies of S. aureus.

The variation in length and sequence of the

16S-23S spacer makes it an ideal candidate for typing of strains and species identification which can potentially be applied to any species of the bacterial kingdom. Our method permits reliable, rapid identification and typing on this basis.

The sequence information presented in Figure 11 is tabulated to show from which S. aureus strains the original PCR product was isolated, which clones were characterized, the size of the 16S-23S spacer (*), the size of the Hpall insert (fragment E: ‡), the presence or absence of sequences, with positions and lengths shown according to numbering in Figure 13), the base pair

differences between clones for the 13 regions (CS=conserved sequence, VS=variable sequence) and the GenBank accession numbers §. The size (base pairs, bp) of fragment C (31) can be obtained by adding 331bp (HpaII² to LR520) and 47bp (R1392 to HpaII¹) to fragment E (‡).

Whereas the bacteriophage typing system was formerly the standard method for S. aureus (Williams et al, 1953), many current MRSA strains are not typable by the International Set of Phages (Richardson et al, 1988), requiring the addition of further experimental phages.

RFLP analysis by pulsed-field-gel electrophoresis has been shown to be more discriminating than phage typing

(Schlichting et al, 1993). However, RFLP analysis relies on the stability of restriction enzyme recognition sites such that a point mutation within a site will result in a different RFLP. The sequence conservation of the rrn operons (Woese, 1987) argues for the use of the 16S-23S spacer region as a more stable and direct indicator of the evolutionary divergence of S. aureus strains, and is a valuable addition to the large number of typing methods available.

It will be apparent to the person skilled in the art that while the invention has been described in some detail for the purposes of clarity and understanding, various modifications and alterations to the embodiments and methods described herein may be made without departing from the scope of the inventive concept disclosed in this specification.

References cited herein are listed on the following pages, and are incorporated herein by this reference. REFERENCES

Altwegg, M., Hickmann-Brenner, F.W. & Farmer, J.J. Ill Journal of Infectious Diseases, 1989 160 145-149

Amikam, D., Glaser, G. & Razin, S.

J. Bacteriol., 1984 158 376-378

Bercovier, H., Kafri, O., & Sela, S.

Biochem. Biophys. Res. Commun., 1986 136 1136-41

Bingen, E.H., Denamur, E. Lambert-Zechvosky, N.Y.,

Bourdois, A., Mariani-Kurkdjian, P., Cezard, J-P.,

Navarro, J. & Elion, J.

J. Clin. Microbiol., 1991 29 1348-1350

Boondeekhun, H.S., Gürtler, V., Odd, M.L. Wilson, V.A. &

Mayall, B.C.

J. Med. Microbiol., 1993 38 384-387 Blumberg, H.M., Rimland, D., Kiehlbauch, J.A., Terry, P.M., & Wachsmuth, I.K.

J. Clin. Microbiol., 1992.30 362-9

Brosius, J., Palmer, M.L., Kennedy, P.J. & Noller, H.F. Proc. Natl. Acad. Sci. USA., 1978 72 4801-4805 Brosius, J., Dull, T.J., Sleeter, D.D. & Noller, H.F.

J. Molecular Biology, 1981 148 107-127

Burgin, A.B., Parodos, K., Lane. D.J. & Pace, N.R.

Cell, 1990 60 405-414

Cato, E.P., George, W.L. & Finegold, S.M

Bergey's manual of Systematic Bacteriology, 1986 2 1141-1200, Edited by P.H.A. Sneth, N.S. Mair, M.E. Sharpe & J.G. Holt. Baltimore: Williams & Wilkins De Buyser, M-L., Morvan, A., Aubert, S., Dilasser, F. & el, S.N.

J. Gen. Microbiol., 1992 138 889-99

East, A.K., Allaway, D. & Collins, M.D.

FEMS Microbiol. Lett., 1992 74 57-62

East, A.K. & Collins, M.D.

FEMS Microbiol. Lett., 1993 106 129-33

Feingold, J., Bellofatto, V., Shapiro, L., & Amemiya, K. J. Bacteriol., 1985 163 155-66 Felsenstein, J.

PHYLIP (Phylogeny Inference Package) Version 3.5p.

university of Washington, 1993

Fournier, M.J. & Ozeki, H.

Microbiological Reviews, 1985 49 379-397 Gamier, T., Canard, B., & Cole, S.T.

J. Bacteriol., 1991173 5431-8

Green, C.J., Stewart, G.C., Hollis, M.A., Void, B.S., &

Bott, K.F.

Gene, 1985 37 261-6 Gürtler, V.

J. Gen. Microbiol., 1993 139 3089-97 Gürtler, V., Wilson, V.A., & Mayall, B.C.

J. Gen. Microbiol., 1991 137 2673-9

Guttell, R.R. & Fox, G.E.

Nucleic Acids Research, 1968 16 r175-r269 Harrison, T.G., Saunders, N.A., Haththoutuwa, A., Doshi, N.

& Taylor, N.G.

J. Med. Microbiol., 1992 37 155-161

Harvey, S . , Hill, C.W., Squires, C. & Squires, C.L.

J. Bacteriol., 1988 170 1235-1238

Higgins, D.G., Bleasby, A.J., & Fuchs, R.

Comput. Appl. Biosci., 1992 8 189-191

Hill, C.W. & Harnish, B.W.

J. Bacteriol., 1982 149 449-457 Hsu, D., Pan, M-J., Zee, Y.C. & Le Febvre, R.B.

J. Bacteriology 1990 172 3478-3480

Hsu, D., Zee, Y.C, Ingraham, J. & Shh, L-M.

J. Gen. Microbiol., 1992 138 199-203

Jarsch, M. & Böck, A.

Nucleic Acids Res., 1983 11 7537-44

Jensen, M.A. & Straus, N.

PCR Methods and Applications, 1993 3 186-194

Jensen, M.A., Webster, J.A. & Straus, N.

Appl. Environ. Microbiol., 1993 59 945-52 Kloos, W.E. & Schleifer, K.H.

Bergey's Manual of Systematic Bacteriolosry, 1986 2

1013-1035 Ed. N.S. Mair P.H.A. Sneath M.E. Sharpe and J.G. Holt. Baltimore, Williams and Wilkins, 1986

LiPuma, J.J., Stull, T.L., Dasen, S.E., Pidcock, K.A.,

Kaye, D. & Korzeniowski, O.M.

J. Infect. Disease, 1989 159 526-532 Loughney, K., Lund, E. & Dahlberg, J.E.

Nucleic Acids Res., 1982 10 1607-24

Loughney, K., Lund, E. & Dahlberg, J.E.

J. Bacteriol., 1983 154 529-532 Ludwig, W., Kirchof, G., Klugbauer, N., Weizenegger, N., Betzl, D., Ehrmann, M., Hertel, C, Jilg, S., Tatzel, R., Zitzelsberger, H., Liebl, S., Hochberger, M., Shah, J., Lane, D. & Wallnoef, P.R.

Syst. Appl. Microbiol., 1992 15 487-501 Maddison, W.P. & Maddison, D.R.

MacClade, Analysis of phylogeny and character evolution. Version 3.0. Sunderland, Massachusetts: Sinauer Associates, 1992

Monzon, M.C., Aubert, S., Morvan, A. & Solh, N.E.

J. Med. Microbiol., 1991 35 80-8

Morgan, E.A., Ikemura, T. & Nomura, M.

Proc. Natl. Acad. Sci. USA., 1977742710-4

Nakagawa, T., Uemori, T., Asada, K., Kato, I. &

Harasawa, R.

J. Bacteriol., 1992 174 8163-8165

Neefs, J-M., Van de Peer, Y., Hendriks, L. & De Wachter, R. Nucleic Acids Research, 1990 18 r2237-2317

Paithankar, K.R. & Prasad, K.S.N.

Nuc. Acids Res., 1991 19 1346 Preheim, L., Pitcher, D., Owen, R. & Cookson, B.

Eur. J. Clin. Microbiol. Infect. Dis., 199110 428-36 Richardson, J.F., Chittasobhon, N. & Marples, R.R.

J. Med. Microbiol., 1988 25 67-74

Sambrook, J., Fritsch, E.F. & Maniatis, T.

Molecular Cloning, a Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989

Sanger, F., Nicklen, S. & Coulsen, A.R.

Proc. Natl. Acad. Sci. USA, 1977 74 5463-5467

Schlichting, C, Branger, C, Fournier, J.M., Witte, W., Boutonnier, A., Wolz, C, Goullet, P. & Doring, G.

J. Clin. Microbiol., 1993 31227-32

Sechi, L.A. & Daneo-Moore, L.

J. Bacteriol., 1993 175 3213-9

Stewart, G.T. & Holt, R.J.

Br. Med. J., 1963 1 308-311

Sutter, V.L., Citron, D.M., Edelstein, M.A.C. &

Finegold, S.M.

Wadsworth Anaerobic Bacteriology manual, 4th Ed., Belmont, California: Star Publishing Company, 1985 Suzuki, Y., Yoshinaga, K., Ono, Y., Nagata, A. & Yamada, T. J. Bacteriol., 1987169 839-43

Swofford, D.L.

(PAUP) version 2.4, University of Illinois, 1985

Townsend, D.E., Ashdown, N., Bolton, S., Bradley, J.,

Duckworth, G., Moorhouse, E.C. & Grubb, W.B.

J. Hosp. Infect., 1987 9 60-71 Townsend, D.E., Ashdown, N., Bradley, J.M., Pearman, J.w.

& Grubb, W.B.

Med. J. Aust., 1984 141 339-340

Vaneechoutte, M., Roussau, R., De Vos, P., Gillis, M., Janssens, D., Paepe, N., De Rouck, A., Fiers, T.,

Clayes, G. & Kersters, K.

FEMS Microbiology Letters, 1992 93 227-234

Void, B.S.

Microbiological Reviews, 1985 49 71-80 Widom, R.L., Jarvis, E.D., LaFauci, G. & Rudner, R.

J. Bacteriol., 1988 170 605-610

Williams, R.E.O., Rippon, J.E. & Dowsett, L.M.

Lancet, 1953 i 510-514

Woese, C.R.

Microbiol. Rev., 1987 51 221-271

Wolcott, M.J.

Clin. Microbiol. Rev., 1992 5 270-386

Yanisch-Perron, C, Vieira, J. & Messing, J.

Gene, 1985 33 103-119

Claims

CLAIMS :

1. A method of identification of microorganisms, comprising the steps of extracting and purifying DNA from a sample suspected to contain bacteria, and subjecting the 16S-23S rRNA spacer region of said DNA to amplification, comprising a highly conserved region from the 3' end of the 16S-23S rRNA spacer region, and/or a highly conserved region from the 5' end of said region, thereby producing fragments having detectable differences in size and number, and separating the amplified fragments.

2. A method according to Claim 1, wherein the primers used correspond to a highly conserved region from the 3' end of the spacer region, and to a highly conserved region from the 5' end of the spacer region respectively.

3. A method according to Claim 1 or Claims 2, wherein the primers comprise a sequence corresponding to a region from the 5' end of the 16S rRNA gene and/or to a region from the 3' end of the 23S rRNA gene.

4. A method according to any one of Claims 1 to 3, using a first primer comprising a sequence from the 5' end of the 16S rRNA gene, and a second primer comprising a sequence from the 3' end of the 23S rRNA gene.

5. A method according to any one of Claims 1 to 4, wherein the primers are 15 to 20 nucleotides long.

6. A method according to Claim 5, wherein the primers are respectively R1391F and selected from the group consisting of LR488 and LR194F, said primers being as herein defined.

7. A method according to Claim 6, wherein LR488 is 15 to 19 nucleotides long, and R1391F is 15 to 18

nucleotides long.

8. A method according to any one of Claims 1 to 7, wherein one or more additional probes are used.

9. A method according to Claim 8, wherein the

additional probe is the sequence encoding tRNA^ile and/or the sequence encoding tRNA^ala.

10. A method according to any one of Claims 1 to 9, wherein the amplification is performed by a method selected from the group consisting of polymerase chain reaction, ligase chain reaction, 3SR amplification, strand

displacement amplification, Qβ replicase reaction, and branched DNA signal amplification.

11. A method according to any one of Claims 1 to 9, wherein the amplified fragments are separated by a method selected from the group consisting of denaturing

polyacrylamide gel electrophoresis, capillary

electrophoresis, and high performance liquid

chromatography.

12. A method according to any one of Claims 1 to 11, wherein the sample is a clinical or environmental sample.

13. A method according to any one of Claims 1 to 11, wherein the microorganism is a Clostridium or a

Staphylococcus.

14. An amplification primer reagent for use in a method according to any one of Claims 1 to 13, comprising a highly conserved region from the 3' end of the 16S-23S rRNA spacer region, and/or a highly conserved region from the 5' end of said region.

15. An amplification primer reagent according to Claim 14, comprising a sequence corresponding to a region from the 5' end of the 16S rRNA gene and/or to a region from the 3' end of the 23S rRNA gene.

16. An amplification primer reagent according to Claim 15, comprising a region from the 5' end of the 16S rRNA gene and a region from the 3' end of the 23S rRNA gene.

17. An amplification primer reagent according to any one of Claims 14 to 17, wherein the primers are 15 to 20 nucleotides long.

18. An amplification primer reagent according to Claim 16, wherein the primers are respectively R1391F and selcted from the group consisting of LR488 and LR194F, said primers being as herein defined.

19. An amplification primer reagent according to Claim 18, wherein LR488 is 15 to 19 nucleotides long, and R1391F is 15 to 18 nucleotides long.