WO2005056821A1

WO2005056821A1 - Detection of cryptococcus in a sample by detecting a mini-intein encoding region of the prp8 gene

Info

Publication number: WO2005056821A1
Application number: PCT/NZ2004/000320
Authority: WO
Inventors: Russell Tony Poulter; Margaret Isobel Butler
Original assignee: Otago Innovation Limited
Priority date: 2003-12-11
Filing date: 2004-12-13
Publication date: 2005-06-23

Abstract

A method for the detection and/or assay of pathogenic Cryptococcus in a sample is provided. It involves detection or measurement of a polynucleotide in the sample encoding the intein of the PRP8 gene of a Cryptococcus pathogen.

Description

Detection of Cryptococcus in a sample by detecting a mini-intein encoding region of the PRP8 gene.

TECHNICAL FIELD The invention relates to a method for the detection and/or assay of the fungal pathogen from the genus Cryptococcus.

BACKGROUND ART Cryptococcus neoformans is an encapsulated fungal pathogen causing fatal meningitis in humans. Infection, initiated by inhalation into the lungs, occurs mainly in immunocompromised people, for example HIV/ AIDS patients, but can also occur in healthy people. Infection of the brain and meninges is the most common clinical manifestation resulting in dementia.

The Cryptococcus species have been diverging over the past 40 million years into 3 distinct varieties and four serotypes: opportunistic pathogens C. neoformans var. neoformans (serotype D) and C. neoformans var grubii (serotype A) and the primary pathogen C neoformans var gattii (aka C. bacillisporus)(seτotypes B and C).

Diagnosis of infectious disease is moving towards molecular methods (for example, detecting specific stretches of DNA present in the particular infecting organism).

This is partly because these molecular methods are faster and less subjective; compared to culture methods and/or morphological methods they require less experienced personnel to interpret the results. Molecular methods can also be more specific (ie only the relevant species is detected).

The polymerase chain reaction (PCR) is the present method of choice for many molecularly- based diagnostic systems. In this technique, a small region of DNA is preferentially amplified in vitro from a preparation of DNA extracted from the whole infecting organism or clinical sample.

For diagnosis especially of fungi, there is often difficulty in finding an appropriate, species- specific DNA target to amplify; most PCR applications so far rely on conserved sequences that are similar in many species and this can lead to false positive results. Inteins are genetic elements present within protein-coding sequences (intervening proteins). Inteins perform protein splicing. This is a self-catalyzed process by which the intein removes itself from the protein creating the functioning enzyme. These elements are rare in living organisms but examples are found in all three kingdoms. Only two inteins have been described in the nuclear genetic material of higher organisms. The applicant has discovered one of these within the PRP8 gene of Cryptococcus neoformans (see published PCT application WO 095036), the contents of which are fully incorporated herein by reference

Inteins from bacteria and higher organisms share a similar structure and mechanism of operation. Of the bacteria which are associated with humans, the only genus that carries inteins is Mycobacterium. This includes the causative agent of tuberculosis.

The PRP8 gene product is one of the most highly conserved proteins known (Luo et al RNA, 5, 893-908 (1991), comprising the core of the spliceosome. The PRP8 gene product is an indispensable component of the spliceosome and therefore essential for cell viability. Loss of PRP8 function would result in an inability to process introns from all mRNA transcripts. PRP8 is needed in very large amounts during rapid growth. Every intron in every message has to be removed before the mRNA can be active.

It is an object of the present invention to provide an assay for the detection and/or quantitation of pathogenic Cryptococcus species.

DISCLOSURE OF THE INVENTION

In one aspect the invention provides a method for the detection and/or assay of pathogenic Cryptococcus in a sample comprising detecting or measurement of a polynucleotide in the sample encoding the intein of the PRP8 gene of a Cryptococcus pathogen.

In a preferred embodiment of the invention, DNA of the sample is contacted with a polynucleotide that binds to the DNA encoding the intein and/or the flanking host gene (PRP8) and the binding is detected and/or measured. The polynucleotides may alternatively or additionally bind to DNA complementary to the coding sequence for the intein and/or flanking host gene.

In a further preferred embodiment the DNA encoding the intein in the sample is amplified. Preferably the DNA to be detected and/or amplified comprises at least a part of the DNA encoding the intein of the Cryptococcus PRP8 intein and a portion of the adjacent PRP8 sequence.

Preferably the DNA detected is the whole intein PRP8 sequence plus a portion of the immediately adjacent flanking sequences of PRP8.

In a different preferred embodiment of the invention a hybridisation probe binds to a sequence including nucleotides 340 and 341 of the H99_serA sequence shown in Figure 1 or to the corresponding nucleotides of another Cryptococcus serotype.

Currently the most preferred method for use in the invention is PCR.

PCR methods are well known by those skilled in the art (Mullis et al., Eds 1994 The Polymerase Chain Reaction, Birkhauser). The template for amplification may be selected from genomic DNA, mRNA or first strand cDNA derived from a sample obtained from the sample under test (Sambrook et al., 1989 Molecular Cloning - A laboratory manual, Cold Spring Harbour Laboratory, Cold Spring Harbour NY). The LightCycler (Roche Molecular Biochemicals, Mannheim, Germany) technology uses PCR and is capable of detecting and quantifying specific fungal DNA simultaneously.

The amplified DNA may be detected by any one of a number of techniques for example by gel electrophoresis. In real-time PCR protocols, such as the LightCycler, the PCR process is monitored either by fluorescence quantification of the DNA-binding dye S YBR Green I for the general detection of double-stranded DNA or by hybridization probes. LightCycler hybridization probes consist of a pair of oligonucleotides annealing next to each other on a given nucleic acid sequence and labeled with two different fluorescent dyes on their 3' and 5' ends, respectively. In the presence of the specific target sequence, the probes hybridize head-to-tail, bringing the two fluorescent dyes into close physical proximity.

When brought into proximity the fluorescent dyes can interact. The fluorescence intensity is directly proportional to the amount of specific target sequence present in the amplification mixture and measured during each PCR cycle. Real time PCR is capable of providing quantification of the infection by measuring the number of cycles needed to produce the amplified product. A low level infection will result in a low concentration of template in the clinical sample and this will require more cycles of amplification to achieve a given level of duplex product.

Real time PCR is also capable of providing sequence confirmation for the amplified product by a simple 'melting curve analysis'. A melting curve analysis relies on the fact that a double- stranded DNA (such as a PCR product and the associated, labelled fluorescent probes) will denature or disassociate at a specific sharply defined temperature. This temperature is determined by the length of the duplex, the base composition of the duplex and the presence of any mismatched bases. In a melting curve analysis, after the PCR process is completed, the temperature in the thermal chamber is slowly increased. When one of the probes melts off and the two fluorescent dyes are no longer in close proximity, fluorescence decreases and a characteristic melting point is observed for the given target sequence. Similar, single hybridisation probes are used in other 'real-time' PCR protocols, such as the TaqMan system of Applied Biosystems.

Real time PCR is capable of discriminating varieties of Cryptococcus by analysing the amplified product by a simple 'melting curve analysis'. Every mismatch present in the duplex formed by the PCR product and the hybridization probe will lower the observed melting point. It is possible to design hybridisation probes which will give distinct melting curves for the major varieties of

Cryptococcus. These probes can be designed on the basis of the DNA sequences which we have described for the varieties of Cryptococcus.

In another aspect the invention comprises a method of diagnosing a Cryptococcus infection using a method as described above.

It would be useful to be able to detect the presence of Cryptococcus infection at an early stage when it is present only in the lung, rather than later when it progresses to the nervous system. These methods also can be used to detect the three major clinical types. Most importantly assays of the invention include assays able to detect all strains of pathogenic Cryptococcus but without yielding false positives due to the presence of other microorganisms. All isolates of pathogenic Cryptococcus that have been analysed contain the mini-intein in their PRP8. No other organisms have been found with a mini-intein in their PRP8 gene. Most importantly this includes other, closely related, non-pathogenic yeasts. The mini-intein is therefore a defining characteristic of the pathogenic Cryptococcus. For efficient detection, one needs:

1. A PCR target in a gene that is present in all the strains likely to be clinically relevant.

2. The site for PCR primers must be highly conserved. The Cryptococcus mini-intein is present in a conserved part of the PRP8 protein and the terminal sites of the intein are also highly conserved. By designing primers which overlap the PRP8/intein boundary and the intein/PRP8 boundary, we will get a highly conserved PCR site which is also highly discriminatory.

3. The mini-intein target is also of an appropriate size for PCR analysis, in that if the flanking conserved regions are too far apart, it becomes more difficult to amplify the large PCR product. 4. In addition, to be able to specifically discriminate infecting Cryptococcus, one needs a target specific to the infecting organism. The PRP8 mini-intein is not present in any other organism. 5. In addition there are sufficient varietal differences within the internal sequence of the mini-intein to allow variety resolution.

The term "conserved" in relation to an intein polynucleotide sequence indicates that the conserved region has at least 90% identity, preferably 95%, more preferably 100% over a length of 14 or more nucleotides with all the corresponding region in the sequence of every strain of Figure 1.

One pair primers for detecting Cryptococcus neoformans/bacillosporus includes Fw-H99 5' CCATTCATCTGGACTAACC 3' (SEQ ID NO:l). This primer recognises a conserved region of the PRP8 gene ~160bp upstream from the intein sequence insertion point.

The second primer is

Rev-Intl804 5' GTCCTGCTCGTTCCATCG 3' (SEQ ID NO:2)

This primer recognises a conserved region (CGATGGAACGAGCAGGAC, SEQ ID NO:3) of the intein ~50bp in from the beginning of the intein. The PRP8 mini-intein is thus ideal in that it has conserved flanking sequences for primer design, these flanking regions are not too far apart for routine PCR amplification and the presence of the mini-intein is specific to the pathogenic Cryptococcus (and not present in other organisms). Our sequence analyses have also indicated that it is possible to use the intein sequence to discriminate between the three varieties of Cryptococcus neoformans. These varieties have specific, differing features relevant to epidemiology and treatment, and it is useful to be able to easily detect which is the infecting organism. The 3 varieties can be discriminated by restriction analysis of conventional PCR products, by sequencing these PCR products or by use of the Light Cycler and specifically designed primers. In an alignment of the inteins from diverse Cryptococcus, the three main clinical types can easily be distinguished.

The applicant has described a mini-intein of -500 base pairs within the PRP8 gene of Cryptococcus. Although the PRP8 gene is highly conserved (similar in sequence) across many species the applicant has shown that only in the pathogenic Cryptococcus is this gene interrupted by a mini-intein. We have analysed a large number of related non-pathogenic yeasts and none have the intein. We have analysed a large number of pathogenic Cryptococcus strains and all have the intein. The intein sequence is similar in diverse pathogenic Cryptococcus. The different serotypes/varieties can be distinguished on the basis of their intein sequence.

The conserved sequence of the PRP8 and intein means that it is possible to design a PCR assay using specific primers (the flanking sequences from which the amplification process is initiated). The assay will not be subject to false negatives or false positive results (because the presence of the intein is indicative of a pathogenic Cryptococcus).

Isolates of C neoformans var gattii caused an outbreak on Vancouver Island in Canada that involved >50 infections (some fatal) in otherwise healthy people. These isolates (both clinical and environmental) were unusual in that they were all fertile in mating tests, whereas environmental isolates from Australia are invariably sterile (published results of other workers). In addition, many PCR reaction primers which work in a wide variety of C. neoformans var. gattii strains do not work with the Vancouver strains. In contrast, the applicant's standard 'diagnostic' primers, designed to amplify the PRP8 intein do work in the Vancouver isolates. This reinforces our confidence in the effectiveness of these primers. It is clinically useful to be able to quantify an infection at diagnosis or during treatment. This can be done with 'real time PCR'. It would also be useful to be able to distinguish the various Cryptococcus serotypes as these have a different clinical performance. This can also be done by 'melting curve' analyses performed at the end of the real time PCR.

Preferred primer pairs have one member which binds to the junction of the PRP8 gene and the start of the intein and one member which binds to the end of the intein and the adjacent PRP8 sequence. That is to say, preferred primer pairs have one member which binds to the PRP8 gene upstream from the intein and to the start of the intein and a second member which binds to the end of the intein and the PRP8 gene downstream of the intein.

Preferred primers have a length of 15-25 Nucleotides, preferably 20 nucleotides.

Preferred primers have a similar composition, preferably approximately 50% guanine and cytosine.

Preferred primers may be equimolar mixtures of several very similar molecules (redundancy) so as to provide perfect complementarity to slightly divergent Cryptococcus sequences. Preferred primer pairs include FWdiag 5'ctgttctgggagaa[ag]gc[ct]tg 3' (SEQ ID NOs': 3-7) and diagRV 5 ' tcctcaaatcctga[ag]tt[ag]tg 3 ' (SEQ ID NOs' : 8- 11 ) where [ct] indicates redundancy, equimolar amounts of two oligonucleotides one with c and the other with t at that position. [ ag] redundancy indicates either an a or a g at that position.

It is also possible to use primers binding to two different conserved sites within the intein sequence.

A useful primer pair may be selected from the group comprising (a) cryp reverse 5' aaccggaccactt[c,t]gta 3' (SEQ ID NOs': 12 and 13) cryp int forw 5' tttgtccg[c,t]cttcctca 3' (SEQ ID NOs': 14 and 15) or (b) a pair of primers comprising at least 14 contiguous nucleotides of each member of the primer pair; or (c) a pair of primers complementary to the primer pair of (a) or (b). These primers are useful for a broad range of Cryptococcus strains.

A useful probe of the invention is a polynucleotide probe having a sequence selected from (a) 5'gaaaagctgatacttctgttgctctt 3' (SEQ ID NO: 16), (b) 5'gaaaagcttatacttctgttgctctt 3' (SEQ ID NO: 17), (c) 5'gaaaagccgatacttctgttgctctt 3' (SEQ ID NO: 18), (d) a partial sequence comprising at least 14 nucleotides of (a), (b) or (c) provided that the sequence comprises variable nucleotide residues 8 and 9 from the 5 'end, or (e) a sequence comprising at least 14 nucleotides complementary to any one of (a), (b), (c) and (d).

In a further aspect the invention provides a diagnostic kit which can be used for diagnosis of a Cryptococcus infection comprising the step of detecting the PRP8 intein. One kit includes a set of primers of the invention and reagents useful for amplifying DNA. Such a kit is useful for detecting the presence of Cryptococcus in a sample from an infected patient.

In a further aspect the invention provides a kit which can be used for quantification of a Cryptococcus infection in a real time PCR system. One kit includes a set of primers of the invention and reagents useful for amplifying DNA. Such a kit is useful for assessing the effectiveness of an antibiotic treatment against a Cryptococcus infection in a patient.

The invention will now be defined by specific examples which are illustrative only and are not intended to limit the invention in any way.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows intein DNA sequences from a number of Cryptococcus strains. The underlined portions are sites related to the primers of Example 2. The bold portion is complementary to the probe of Example 2.

Figure 2 shows a phylogenetic tree based on alignment of the DNA sequences of the Cryptococcus PRP8 inteins.

Figure 3 shows a phylogenetic tree based on an alignment of the DNA sequences of the Cryptococcus bacillisporus PRP8 inteins. Figure 4 shows an alignment of the splicing domain motifs of the PRP8 mini-inteins with the full-length intein from Cryptococcus laurentii.

Figure 5 shows Amplification Curves for 10-fold dilutions of Cryptococcus DNA amplified with intein-specific primers.

Figure 6 shows Melting peaks.

Figure 7 shows variation of melting T_m of 3 Cryptococcus varieties.

EXAMPLE 1 Materials and Methods

TABLE 1

Strains and species of Cryptococcus used in this study.

Source Species/variety Strain

Marianna Viviani, C. neoformans var. neoformans IUM, Milan C. neoformans var. grubii

Elizabeth Johnson, C. neoformans var. grubii PHLS, Bristol, UK

Wieland Meyer, C neoformans var. neoformans Westmead Hospital, C neoformans var. grubii Sydney C bacillisporus

Centraalbureau voor Schimmelcultures

Utrecht, The Netherlands C. bacillisporus (10 strains) C. laurentii CBS139 C amylolentus CBS6039 C curvatus CBS570 C. albidus CBS 142

C. heveanensis CBS569 C. skinneri CBS5029 C. dimennae CBS5770 C. humicola CBS577

Environmental Science and Research, C. uniguttulatus ATCC 66033

New Zealand C. neoformans var. neoformans CBS 132

Strains and media. The strains and species of Cryptococcus used in this analysis were obtained from the CBS, Utrecht, The Netherlands, or were gifts from members of the Cryptococcus research community (Table 1). Strains from the PHLS, Bristol (all serotype A) have a four digit designation; for example 8104. Strains supplied from IUM, Milan have a six digit designation; for example 93_3231. Strains prefixed with CBS were obtained from the Centraalbureau voor Schimmelcultures. Strains prefixed with LA or WM and M27055 were from Wieland Meyer at Westmead Hospital, Sydney. Strains prefixed by 'RB' and also 113A-5, 152A-1 and MAC 9 were isolated on Vancouver Island and were obtained from James Kronstad (University of British Columbia) via Wieland Meyer (Westmead Hospital, Sydney). B3501 is the strain sequenced at Stanford University. DNA from five strains isolated in Botswana (Bt) was kindly provided by Joe Heitman at Duke University. Strains were grown at 27°C in YPD medium (1% Difco Yeast extract, 2% peptone, 2% glucose; solidified with 1.5% agar when necessary). DNA isolation, amplification, restriction analysis and sequencing. Genomic DNA was isolated from 50ml overnight cultures essentially using the method of Philippsen et al. Methods Enzymol 194:169-182 (1991). Amplification of the intein sequence and flanking regions was accomplished with the Expand High Fidelity PCR system (Roche, Mannheim, Germany) as outlined in Butler et al. Yeast 18:1365-70 (2001). Primers were synthesised by Proligo, Singapore; the primer sequences were: FCnln 5' gcgaattcccacatggtgaatcgacg 3' (seq id no: 19), CnlnR 5' gctctagatcatctggactaaccagc 3' (SEQ ID NO:20), H99 5' gcgaattcccattcatctggactaacc 3' (SEQ ID NO:21), FW_diag 5' ctgttctgggagaa[ag]gc[ct]tg 3'(SEQ ID NOs':4-7) and diag_RV 5' tcctcaaatcctga[ag]tt[ag]tg 3' (SEQ ID NOs': 8-11); primers used for sequencing PCR products from C. laurentii are available from the inventors. The resulting PCR products were purified with Qiagen columns (Hilden, Germany) prior to restriction digest or sequencing. PCR products were digested with 1 U of the appropriate restriction enzyme and buffer for an hour; this was followed by separation of the fragments on al .5% agarose gel. PCR products were sequenced using an ABI 377 DNA Sequencer by the Centre for Gene Research at Otago University (http://microbes.otago.ac.nz/cgr/home.htm) or a capillary ABI 3730 Genetic Analyser at the Allan Wilson Centre Genome Service at Massey University

(http://awcmee.massey.ac.nz/genome-service.htm).

Sequence analysis. General sequence analyses were done using the Wisconsin GCG package (Genetics Computer Group, 575 Science Drive, Madison, Wis.). Sequence similarity searches were performed using the National Center for Biotechnology Information BLAST server (http://www.ncbi.nlm.nih.gov/BLAST - Multiple sequence alignments were constructed using CLUSTAL X at the European Bioinformatics Institute server (http://www.ebi.ac.uk/clustalw/) edited with Seaview (Galtier et al Comput. Appl. Biosci. 12 548, 1996) and shaded with MacBoxshade (http://www.isrec.isb-sib.ch/sib-isrec/boxshade/macBoxshade). Phylogenetic trees were constructed using PAUP4M0 (Swofford et al PAUP* Phylogenetic analysis using parsimony (*and other methods). Version 4 - Sinauer, Sunderland, Mass., USA. Intein sequences in the Cryptococcus PRP8 genes and flanking sequences of each variety have been assigned GenBank accession numbers AY422973, Cne PRP8 IUM93-3231 (C. neoformans var neoformans, serotype D); AY422974, Cne PRP8 8104 (C. neoformans var grubii, serotype A); AY422975, Cba PRP8 WM02.98 (C. bacillisporus). Mini-intein sequences of further strains, representative of the molecular types described in this study have been assigned GenBank accession numbers AY836243- AY836253. The full-length intein in C. laurentii, ClaPRP8, has been assigned to Gen Bank accession AY836254. Descriptions of the Cryptococcus PRP8 inteins have been added to the InBase records (http://www.neb.com/neb/inteins.html). Cryptococcus genome sequencing projects. Cryptococcus neoformans: the genome of serotype D strain B-3501A is being sequenced at the Stanford Genome Technology Center (SGTC; http://www-sequence.stanford.edu/group/Cneoformans/index.html). Results described in this paper refer to release cneoformans031220 (December 22, 2003). Serotype D strain JEC21 is being sequenced at The Institute for Genomic Research (TIGR; http://www.tigr.org/tdb/e2kl/cnal/). Serotype A strain H99 is being sequenced at the Whitehead Institute/MIT Center for Genome Research/Cryptococcus neoformans Sequencing Project (http://www-genome.wi.mit.edu) and at the Duke Center for Genome Technology (http://cneo.genetics.duke.edu/) and the Vancouver Genome Sequence Centre, BC Cancer Research Centre (http://www.bcgsc.bc.ca/). In addition, large numbers of cDNA sequences for strains B-3501 and H99 have been obtained by the Advanced Center for Genome Technology at the University of Oklahoma (http://www.genome.ou.edu/cneo.html).

RESULTS Occurrence of the PRP8 intein in Cryptococcus neoformans and C. bacillisporus. To examine the distribution of the Cne PRP8 intein, we used PCR to test for the presence of intein sequences in the PRP8 genes of multiple strains of Cryptococcus neoformans and the related yeast Cryptococcus bacillisporus. We used oligonucleotides designed to conserved sequences in the regions of the PRP8 gene flanking the intein insertion site. Amplification products were then sequenced directly to confirm that they encoded the intein and the sequences aligned to previously obtained intein and/or PRP8 gene sequences. Analysis of PCR products indicates that the intein sequence is present in the PRP8 gene of all 55 isolates of Cryptococcus neoformans (both variety grubii and variety neoformans and examples of probable hybrid AD serotypes) and 25 isolates of Cryptococcus bacillisporus. We have sequenced the PCR products from these strains of Cryptococcus and all contain uninterrupted open reading frames, as expected (see SEQ ID NOs' 22-95). Representative examples of these sequences have been submitted to GenBank (see Materials and Methods), all sequences are available as supplementary material. All serotype D intein sequences are 516bp long, serotype A intein sequences are 513bp and C. bacillisporus intein sequences are 510bp in length. These differences are due to single codon insertions/deletions.

Variation in the C. neoformans inteins. Serotype A DNA sequences are 94-95% identical to those of serotype D strains. All serotype A intein sequences contain a HinάJJl restriction site (at base pair 340) which is not found in the inteins of serotype D strains (or C bacillisporus). This restriction fragment length polymorphism is a simple method to distinguish the PCR products of serotype A strains. Sequences from serotype A strains form a phylogenetic group that is distinct from sequences found in serotype D strains (Fig 2). The 35 serotype A strains have six polymorphic sites (and four idiosyncratic base changes), pair-wise differences in this group range from 0-1.37%. Meyer et al. Electrophoresis 20, 1790-9 (1999) suggested, on the basis of PCR fingerprinting, that serotype A strains could be divided into two groups, VNI and VNII. However the intein sequences of representatives of these two groups are identical. As can be seen from Fig.2, the intein sequences from four of the Botswana serotype A strains (Bt63, Bt65, Bt85 and Btl31) are quite distinct from the other serotype A sequences. The intein sequences from Bt65 and Btl31 are identical (Fig. 2), this pair of strains falls into AFLP subgroup 1A (Litvintseva et al Eukaryotic Cell 2, 1162-68, 2003). The intein sequences from Bt63 and Bt85 are identical (Fig. 2), this pair of strains falls into AFLP subgroup IB. Sequences from serotype A strains form a phylogenetic group that is distinct from sequences found in serotype D strains (Fig 2). There is little variation in sequence within the 14 examples of serotype D; there are two polymorphic sites and three other base changes each found in only one strain. Pair-wise differences between the sequences range from 0-0.581.

Intein sequences from AD hybrid strains. Although it is accepted that most C. neoformans strains are haploid it is believed that some strains (serotype AD) are diploids formed by mating between serotype A and serotype D strains. Such hybrid AD strains occur naturally. It is possible that this genetic constitution is not completely stable giving rise to aneuploid (or even haploid) derivatives under natural conditions or during laboratory culture. This would explain the observation that some serotype AD strains are subsequently reported to have only a simple serotype. The intein PCR system was used to examine some putative hybrids. The strain from which the CnePRP8 intein was originally described is the type culture for C. neoformans, CBS 132. It has variously been described as a serotype D (Boekhout et al. Microbiology 147, 891-907, and a serotype AD strain (Katsu et al, FEMS Yeast Res. 4:377-388 (2004). The sequence in GenBank accession AF349436 was derived from direct sequencing of a PCR product generated and sequenced with primers FCnln and, CnlnR. The sequence is identical to that from serotype D strains. In order to determine if an alternative serotype A intein allele was present (but undetected by us) in this putative diploid, we cloned the FCnln /CnlnR PCR product and sequenced a number of clones. The cloned sequences are identical and belong to the serotype D group (data not shown). Subsequently, we designed serotype-specific oligonucleotide primers to sequence the mixed intein PCR products from AD serotype strains. PCR product was generated from CBS 132 using a pair of primers (FW diag and diag_RV) that amplify all C. neoformans. The PCR product could be sequenced with the FW diag, diag RV or D-specific primers but gave no sequence using the A-specific primers. In all cases the sequence was serotype D. We conclude that the CBS 132 clone we analysed carries only the serotype D intein sequence.

Further AD serotype strains were analysed by PCR and sequencing. These hybrid strains LAI 80, LA 183 and WM628 (gifts from Wieland Meyer, University of Sydney) are grouped as VNIII in the PCR fingerprinting method (Meyer pers.com.). These three strains provided PCR products using a pair of primers (FW diag and diag RV) that amplify all C. neoformans intein sequences. When sequenced directly with either the FW_diag or diag RV primers, these PCR products gave long sequences with regularly spaced peaks on the chromatograms. However, numerous sites had ambiguous base calls. These ambiguous sites corresponded to those that differ between serotype A and serotype D sequences. All three strains gave the appropriate, unambiguous sequences when amplified and/or sequenced with either the A-specific or D-specific primers, indicating that both the serotype A allele and the serotype D allele were present in LAI 80, LAI 83 and WM628. The two intein alleles in WM628 were completely sequenced and found to be identical to either serotype A or D. WM628_A allele sequence is identical to that from a large group including strain 8104; WM628_D allele sequence is identical to the group of strains which includes 88-3921.

Two more strains of putative hybrid origin, CBS 464 and CBS 950 (Boekhout et al), were also investigated. When PCR products were sequenced directly using the non-serotype specific primers, only serotype D sequences were present. In attempts to amplify the sequences with serotype-specific primers, only the sequence reactions with the D-specific primers were successful. These results indicate that only the serotype D intein allele is present in these strains, as was the case for CBS 132. These strains (CBS 132, CBS 464 and CBS 950) may be aneuploid, perhaps as a result of prolonged laboratory culture. Variation in the C. bacillisporus inteins. The strains of C. bacillisporus analysed were selected so as to facilitate the integration of the analysis with other molecular studies of this species. We were able to amplify the intein sequences from strains of C. bacillisporus using the primer pair FW diag and diag RV. C. bacillisporus intein DNA sequences are 85-87% identical to those of C. neoformans. Sequences from C. bacillisporus strains are more closely similar to those from serotype D strains (pair- wise differences range between 11.96%- 13.14%) than they are to serotype A strains (pair-wise differences range between 13.33%-15.11%). The C. bacillisporus strains have intein sequences that form a group distinct from those of the two varieties of C. neoformans (Fig.2). These strains also have much more variation in their intein sequences than the C. neoformans varieties, with 28 polymorphic sites (and two idiosyncratic sites). Pair-wise sequence differences between strains of C. bacillisporus range from 0- 3.53%. The sequences from C. bacillisporus fall into groups that correlate well with the molecular types as described by Ellis et al. (Med. Mycol. 38 Suppl 1, 173-82, 2000), Boekhout et al. (2001) Sugita et al. (Microbiol. Immunol. 45, 757-768, 2001) and Katsu et al. (2004)(Fig. 2). We recognise four major clades in the intein sequences from C. bacillisporus (termed ib-I, ib-II, ib-III and ib-IV). These groups correspond (almost exactly) with the PCR fingerprint groups (VGI, VGII, VGIII and VGIV respectively) proposed by Ellis et al (2000). The correlation of these clades with the groups recognised by others is shown in Fig.3.

Strains described as ib-I correspond to VGI (Ellis et al (2000). These strains form two groups (exemplified by CBS8273 and CBS7749) which correspond to ALFP groups 4A and 4B

(Boekhout et al., 2001) and with ITS groups 3 and 7 (Katsu et al., 2004). Taken together, the ib-I group contains sequences that are different from any other C. bacillisporus sequence at seven sites.

The strains from Vancouver Island have intein sequences that form a group with intein sequences of four strains from the AFLP 6 type of Boekhout et al., (2001). These four AFLP6 strains (CBS6956, 8684, 7750 and 1930) were of diverse geographical origin. This group of strains, which we term ib-II, all share an identical intein sequence (except for one difference in CBS6956) which is distinguished from any other C. bacillisporus at 11 sites. The strains recently isolated on Vancouver Island have been described as VGII (S. Kidd, 3^r Cryptococcus Genomics Conference, Vancouver, August 2003; see also 2) using the PCR fingerprinting typing method (Ellis et al 2000). There is no evidence these strains represent a hybrid group, as suggested by Boekhout et al (2001); for example there were no ambiguous sites in any of the sequences as occur in the intein sequences of AD strains. All of the ib-II intein sequences contain a BamRl restriction site at base pair 234. No other C. bacillisporus or C. neoformans inteins used in this study contain this sequence. This RFLP site therefore represents a simple way of recognising this group.

A third group, which we term ib-III, contains strains described as VG III (Ellis et al., 2000 and Meyer, pers. comm.) and strains described as AFLP 5B and 5C (Boekhout et al., 2001). This group all share three differences from all other C. bacillisporus strains but also includes one strain previously described as VG II (Meyer, pers. comm.). This represents the only discordance between the intein typing system and the PCR fingerprinting method.

Lastly, group ib-IV contains two strains previously typed as VGIV by Ellis et al (2000). The two ib-IV strains carry seven polymorphic intein sites. Inteins in other Cryptococcus species. We have used degenerate primers designed to anneal to various regions of the highly conserved PRP8 gene to screen by PCR other, closely related, basidiomycete species for the presence of intein coding sequences in their PRP8 genes. Some of these species were drawn from the Tremellales group of basidiomycetes which includes Cryptococcus neoformans and C. bacillisporus (16). Other Cryptococcus species examined were more distantly related to C. neoformans; the genus is polyphyletic and has members in all four of the major clades of the class Hymenomycetes. The C. neoformans inteins (Cne PRP8) and the C. bacillisporus inteins (Cba PRP8) are present only as mini-inteins; the insertion into the PRP8 gene is too short to encode an active endonuclease (HEG) and there are no detectable conserved motifs of a HEG domain. There are regions of high protein similarity at each end of the intein sequences, corresponding to the splicing domains described for inteins in general and previously for the Cryptococcus PRP8 intein in particular (Fig.4). During this screening we looked for PCR products corresponding in length (and sequence) to mini-inteins like that of C. neoformans but we also looked for alleles without an intein or those with a full-length intein. Species that tested negative for the presence of the PRP8 intein gave smaller PCR products of the expected size for an intein-less PRP8 site. These intein negative species included C. amylolentus (also known as Filobasidiella amylolentus), the species most closely related to C. neoformans. These short PCR products were sequenced to confirm that the PRP8 sequence did not contain an intein. There is no evidence of an intein present in the PRP8 genes of the basidiomycete yeasts C. heveanensis, C. skinneri, C. dimennae, C humicola, C. amylolentus, C. curvatus, C albidus, and C. uniguttulatus.

Unexpectedly, we detected a PCR product from C. laurentii (CBS 139) which was considerably larger than that found for C. neoformans or C. bacillisporus. We sequenced this PCR product directly with several specifically designed primers. The conceptual translation of these aligned sequences yielded an open reading frame of 1566bp encoding not only sequences similar to the splicing domains of CnePRP8 and CbaPRPδ but also internal sequences corresponding to a homing endonuclease of the LAGLIDADG type (Fig. 4). The putative homing endonuclease contains all of the conserved residues involved in HEG function. This intein has been named ClaPRP8 and the sequence data have been assigned the GenBank accession AY836254. The PRP8 sequences flanking the intein are highly similar to those flanking the mini-inteins CnePRP8 and CbaPRP8; the full-length C. laurentii intein is inserted at the same (allelic) position in the precursor protein. The intron that is downstream of the intein insertion site in the CnePRP8 (Butler et al, 2001) and CbaPRP8 inteins, however, is not present in the C. laurentii sequence. To confirm that the sequence was derived from C. laurentii, we amplified and sequenced 537bp from the ITS1, 5.8S ribosomal RNA gene, and ITS2 region of strain CBS 139. Our sequence has a 100% match to that derived by Scorzetti et al., (FEMS Yeast Res. 2, 495-517, 2002) from the same strain.

Inteins in the PRP8 gene of other fungi. We then looked for the presence of the intein in other fungi using the available genome sequence data. There is no intein in the PRP8 genes of the basidiomycetes Phanerochaete chrysosporium, Ustilago maydis or Coprinopsis cinerea; these three are the only basidiomycetes other than Cryptococcus neoformans for which large amounts of sequence data are available.

Our search of the available eukaryote sequence databases detected intein coding sequences in the PRP8 genes of four ascomycete fungi: Aspergillus nidulans, Aspergillus fumigatus, Histoplasma capsulatum and Paracoccidioides brasiliensis. A single EST entry from P. brasiliensis

(15)(Genbank accession CN242988) contains sequence encoding the C-terminal region of a PRP8 intein and the downstream PRP8 protein. All of these inteins occur at the same site within the PRP8 gene as Cne PRP8 and Cba PRP8. The complete ascomycete inteins have regions at each end that show similarity to the splicing domains of other inteins, especially to the Cryptococcus inteins in PRP8. They also have internal sequences corresponding to a homing endonuclease of the LAGLIDADG type with all of the conserved residues involved in HEG function.

The presence of a PRP8 intein in Aspergillus and Histoplasma is not expected to cause false positives in the methods of the present invention. These organisms are very genetically different from Cryptococcus. The preferred intein primers designed to be complementary to Cryptococcus such as (FWdiag 5'ctgttctgggagaa[ag]gc[ct]tg 3' (SEQ ID NOs' 4-7), diagRV 5' tcctcaaatcctga[ag]tt[ag]tg 3 'SEQ ID NOs':7-l l) do not amplify the intein from Aspergillus nidulans. This is due to the substantial number of mismatches between the diagRV primer and the diverged Aspergillus PRP8 and intein sequence (6 mismatches out of 20). In addition, the

Aspergillus intein is substantially longer (1815bp). This makes it less likely to amplify by PCR.

The PCR product would be easily resolved by gel electrophoresis or the real time PCR internal probes. The PRP8/intein sequences of Aspergillus fumigatus and Histoplasma capsulatum are similar to that of Aspergillus nidulans and these organisms will therefore not give false positive results with the Cryptococcus diagnostic PCR.

An intein is not present in the PRP8 gene of any species from either of the other ascomycete classes, hemiascomycetes (for example, Candida albicans, Saccharomyces cerevisiae) or archiascomycetes {Schizosaccharomycespom.be), for which PRP8 sequence data are available. A PRP8 intein is not present in humans.

The absence of a mini-intein in the PRP8 gene of species closely related to Cryptococcus neoformans and C. bacillisporus, which is particularly advantageous, may be the result of horizontal gene transfer from unrelated organisms possibly related to Aspergillus! Histoplasma.

EXAMPLE 2

This work used two assays on the LightCycler 2.0 instrument The first assay used primers designed to the intein sequence to generate a 156bp amplified product. SYBR Green detection (LC FastStart DNA Master PLUS SYBR Green kit, Roche Applied Science) was used as the detection technology. With this assay, only the primers and template were required in addition to the supplied master mix - no Mg titration was required. The primer sequences were as follows:

cryp reverse 5 ' aaccggaccactt[c,t]gta 3 ' (SEQ ID NOs' : 12 and 13) cryp int forw 5' tttgtccg[c,t]cttcctca 3' (SEQ ID NOs':14 and 15) this primer was labelled with FITC for use with HybProbe detection or unlabelled for use in SYBR Green detection

The second assay used a probe approach in order to differentiate the three varieties. Due to sequence variation, the 'primer-probe' approach was used. The same primers were used with one being internally labelled with the acceptor dye. The detection probe was adjacent to this primer and was 5 'labelled with FITC. This probe covered two single nucleotide polymorphisms (SNPs), each of which varied for one of the varieties.

This primer and probe was used with the LC FastStart DNA Master PLUS Hybridisation Probes kit (Roche Applied Science). Once again, the assay required no Mg optimisation.

The probe was 5' gaaaagctgatacttctgttgctctt 3' (SEQ ID NO: 16)

i. the probe has a perfect match to the sequence in Serotype D strains (C. neoformans var neoformans) ii. the probe has one mismatch to sequence in Serotype A strains (C. neoformans var grubii) 5 ' gaaaagct||atacttctgttgctctt 3 ' (SEQ ID NO : 17) iii. the probe has a different, single mismatch to sequence in C. neoformans var gattii strains (aka C. bacillisporus) 5' gaaaagcigatacttctgttgctctt 3' (SEQ ID NO: 18)

These mismatches affect the stability of the binding of the probe to this region of the Cryptococcus intein DNA. Thus, if the temperature is raised (for instance, during a melting curve analysis on the LightCycler), the probe will 'melt' off the intein sequence at a temperature which is specific for a particular type of PCR product (Tm). In our case, empirically, the mean Tm of C. neoformans var neoformans = 63.70°C. Mean Tm of C. neoformans var gattii = 61.29°C and the mean Tm of C neoformans var grubii = 58.25°C.

Thus the melting curve analysis will reliably separate/distinguish sequences derived from different serotypes. Our initial LightCycler analyses have indicated that it is possible to use primers designed to the intein sequence to detect the DNA of infecting Cryptococcus at relatively low levels (Figure 5) quantify the levels of the infecting Cryptococcus (Figures 6,7) detect the DNA of infecting Cryptococcus despite a high background of non-target DNA (Figure

5) discriminate between the three varieties of Cryptococcus neoformans using a fluorescent probe in a melting curve analysis( Figure 7)

Figure 5 shows (a) SYBR Green detection of Cryptococcus neoformans DNA by LightCycler2. Analytical sensitivity of the SYBR Green PCR assay, using Cryptococcus neoformans specific primers. Template was serially diluted in water. For each dilution, the corresponding amount of Cryptococcus DNA present in the reaction is given next to the amplification curve. These amplifications were done in the presence of lOx Candida albicans DNA. (b) Quantitative analysis of SYBR Green PCR assay. Crossing points determined by the LightCycler2 software were plotted against the logarithm of the target DNA concentration in the reaction.

Figure 6 shows SYBR Green detection of Cryptococcus neoformans. Non-target amplification products appearing in the lowest target DNA concentration reactions and in the negative control were easily resolvable from the target amplifications by melting curve analysis.

Figure 7 shows HybProbe detection of Cryptococcus neoformans DNA by LightCycler2.0 Discrimination between the three varieties of C. neoformans by melting point (Tm). Mean Tm of C. neoformans var grubii = 58.25°C. Mean Tm of C. neoformans var neoformans = 63.70°C. Mean Tm of C. neoformans var gattii = 61.29°C. Template amounts per reaction ranged from 0.2-2ng.

It will be appreciated that it is not intended to limit the invention to the above examples, only. Many variations, which may readily occur to a person skilled in the art, are possible without departing from the scope thereof.

Claims

1. A method for the detection and/or assay of pathogenic Cryptococcus in a sample comprising detection or measurement of a polynucleotide in the sample encoding the intein of the PRP8 gene of a Cryptococcus pathogen.

2. A method as claimed in claim 1 wherein DNA of the sample is contacted with a polynucleotide that binds to the DNA encoding a conserved region of the intein and a flanking portion of the host gene (PRP8) and the binding is detected and/or measured.

3. A method as claimed in claim 1 wherein the polynucleotide binds to DNA complementary to the coding sequence for the intein and the flanking portion of the host gene.

4. A method a claimed in any one of claims 1 -3 wherein the DNA encoding at least part of the intein in the sample is amplified.

5. A method as claimed in any one of claims 1-4 the DNA to be detected and/or amplified comprises a part of the DNA encoding the intein of the Cryptococcus PRP8 intein and a portion of the adjacent PRP8 sequence.

6. A method as claimed in any one of claims 1-5 wherein the method is a PCR method.

7. A method as claimed in claim 6 wherein a first primer recognises a conserved region of the PRP8 gene 160 bp upstream from the intein sequence insertion point and the second primer recognises a conserved region of the intein, 50 bp in from the beginning of the intein.

8. A method as claimed in claim 6 wherein the amplified DNA is detected by fluorescence quantitation of a DNA-binding dye.

9. A method for the detection and or assay of a pathogenic Cryptococcus in a sample comprising detecting or measurement of a polynucleotide binding in the sample wherein a probe binds to a sequence in the PRP8 intein which differs between strains having different serotypes.

10. A method as claimed in claim 9 wherein the probe binds to a sequence in the intein which differs in strains of C. neoformans var neoformans, C. neoformans var grubii and C. neoformans var gattii (also known as C. bacillisporus).

11. A method as claimed in claim 10 wherein the probe binds to nucleotides 340 and 341 of the H99_serA sequence shown in Figure 1 or to the corresponding nucleotides of another Cryptococcus sequence.

12. A method as claimed in claim 10 wherein the sequence of the probe comprises a sequence selected from (a) 5'gaaaagctgatacttctgttgctctt 3', (b) 5'gaaaagcttatacttctgttgctctt 3', (c) 5'gaaaagccgatacttctgttgctctt 3', (d) a partial sequence comprising at least 14 nucleotides of (a), (b) or (c) provided that the sequence comprises variable nucleotide residues 8 and 9 from the 5 'end, or (e) a sequence comprising at least 14 nucleotides complementary to any one of (a), (b), (c) and (d).

13. A method for the detection and/or assay of a pathogenic Cryptococcus in a sample comprising amplification of a polynucleotide sequence of the PRP8 intein which differs between strains of different serotypes and detection of the sequence of the serotypes.

14. A method as claimed in claim 12 wherein the amplified polynucleotide comprises a sequence selected from strains of C. neoformans var neoformans, C neoformans var grubii and C. neoformans var gattii.

15. A method as claimed in claim 13 wherein the sequence of the probe has a sequence selected from (a) 5'gaaaagctgatacttctgttgctctt 3', (b) 5'gaaaagcttatacttctgttgctctt 3', (c) 5'gaaaagccgatacttctgttgctctt 3', (d) a partial sequence comprising at least 14 nucleotides of (a), (b) or (c) provided that the sequence comprises variable nucleotide residues 8 and 9 from the 5 'end, or (e) a sequence comprising at least 14 nucleotides complementary to any one of (a), (b), (c) and (d).

16. A method as claimed in claim 1 wherein DNA of the sample is contacted with a polynucleotide that binds to DNA encoding a conserved region of the intein and the binding is detected and/or measured.

17. A method as claimed in claim 1 wherein DNA of the sample is contacted with a polynucleotide that binds to DNA complementary to a conserved region of the intein and the binding is detected and/or measured.

18. A method as claimed in claim 16 or 17 wherein DNA encoding the conserved region of the intein is amplified.

19. A method as claimed in any one of claims 16-18 wherein the method is a PCR method.

20. A method as claimed in any one of claims 16-19 wherein the sample is contacted with at least one polynucleotide having a sequence of 15-25 nucleotides.

21. A method of any one of claims 1 and 16-20 wherein the DNA of the sample is contacted with the primer pair selected from the group comprising (a) cryp reverse 5' aaccggaccactt[c,t]gta 3' (SEQ ID NOs': 12 and 13) cryp int forw 5' tttgtccg[c,t]cttcctca 3' (SEQ ID NOs': 14 and 15) or (d) a pair of primers comprising at least 14 contiguous nucleotides of each member of the primer pair; or (e) a pair of primers complementary to the primer pair of (a) or (b).

22. A polynucleotide probe having a sequence selected from (a) 5'gaaaagctgatacttctgttgctctt 3' (SEQ ID NO: 16), (b) 5'gaaaagcttatacttctgttgctctt 3' (SEQ ID NO: 17), (c) 5'gaaaagccgatacttctgttgctctt 3' (SEQ ID NO:18), (d) a partial sequence comprising at least 14 nucleotides of (a), (b) or (c) provided that the sequence comprises variable nucleotide residues 8 and 9 from the 5 'end, or (e) a sequence comprising at least 14 nucleotides complementary to any one of (a), (b), (c) and (d).

23. A polynucleotide primer having a sequence selected from (a) 5'aaccggaccacttcgta 3' (SEQ ID NO: 12)

(b) 5 'aaccggaccactttgta 3 ' (SEQ ID NO: 13)

(c) 5 'tttgtccgccttcctcctca 3 ' (SEQ ID NO: 14) (d) 5'tttgtccgtcttcctcctca 3' (SEQ ID NO:15)

(e) a sequence of at least 14 contiguous nucleotides selected from (a), (b), (c) and (d), and

(f) a sequence complementary to any of (a), (b), (c), (d) or (e).

24. A diagnostic kit for diagnosis of a Cryptococcus infection comprising a hybridisation probe as claimed in claim 22.

25. A diagnostic kit for diagnosis of a Cryptococcus infection comprising one or more primers as claimed in claim 23.