INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Improvements in the management and treatment of debilitated medical and surgical patients have led to an unwelcome increase in the number of life-threatening infections due to true pathogenic and opportunistic fungi (2, 17, 30). Candidosis and aspergillosis are the most common mycoses in immunocompromised patients (6). The advent of new antifungal drugs has improved the prospects for management of these infections; however, diagnosis remains difficult, and early initiation of antifungal therapy is critical in reducing the high mortality rate in immunocompromised patients (8, 11).

Conventional methods for diagnosis of fungal infection rely primarily on the identification of species- or genus-specific morphological characteristics that often require histological examination or in vitro culture (6, 19). Recently several authors have described PCR assays targeting different regions of the fungal genome for detection of Candida (9, 38) and Aspergillus (10, 18, 27, 32) species. DNA-based assays for the detection and identification of fungal species provide a potentially more specific and sensitive alternative to conventional culture and detection methods.

We describe a reverse hybridization line probe assay (LiPA) which when combined with PCR amplification allows the specific detection and identification of fungal species. The previously described fungus-specific PCR primers (40) were used to amplify the internal transcribed-spacer (ITS) region of the ribosomal RNA complex from all fungi. The ITS region has sufficient sequence variation to allow for the design of species-specific probes to discriminate between different species (33, 41). In this study, species-specific oligonucleotide probes were designed from within the ITS region for the following species: Candida albicans, Candida parapsilosis, Candida tropicalis, Candida krusei, Candida glabrata, Candida dubliniensis, Cryptococcus neoformans, Aspergillus fumigatus, Aspergillus nidulans, Aspergillus flavus, and Aspergillus versicolor. These probes were incorporated into a LiPA, combined with PCR amplification of the ITS region, and evaluated on a panel of typed fungi and clinical isolates.

Several researchers have described DNA-based tests for Candida infections with detection limits as low as 2 cells/ml of blood sample (14-16, 39). In this report we describe the application of the fungal PCR-LiPA technology to the detection of Candida spp. inoculated into blood samples. As part of the study we evaluated a number of sample preparation methods and concluded that pretreatment of inoculated blood samples (10⁵ to 10¹ Candida cells) with a minibead-beating step followed by DNA extraction using a previously described method (4) or by using the QIAmp tissue extraction kit from Qiagen enabled detection limits of 2 to 10 cells/ml of blood tested. The fungal PCR-LiPA assay can easily be integrated into clinical laboratories for the identification of fungal species following isolation, with the potential to identify the pathogen directly from clinical samples within a single working day.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Microorganisms. All typed fungal isolates (Table 1) used in this study were obtained from the National Collection of Pathogenic Fungi, Mycology Reference Laboratory, PHLS Central Public Laboratory, London, and from John Banks, Central Sciences Laboratory, Sand Hutton, York, United Kingdom. The typed culture of C. dubliniensis CD36 was supplied by Derek Sullivan, Moyne Institute of Preventative Medicine, Trinity College, Dublin, Ireland. Clinical isolates were obtained on agar plates from the Department of Medical Microbiology, University College Hospital Galway, Galway, Ireland. For the purpose of generating serial dilutions of Candida cells for sensitivity studies in inoculated blood samples, 1 ml of a 10-ml overnight culture of Candida cells was centrifuged at 13,000 rpm (14,926 × g; Heraeus Sepatech Biofuge A) for 5 min, the cell pellet was resuspended in 1 ml of dH₂O, and the blastoconidia were microscopically enumerated on a hemocytometer.

Culture conditions. All fungi were grown from lyophilized pellets on Sabouraud glucose agar (Oxoid Ltd.), and a single yeast colony or a 50-µl aliquot of a spore suspension in 0.2% Tween 20 was transferred to 25 ml of Sabouraud glucose (Oxoid Ltd.) broth for DNA extraction. All yeast isolates were grown at 37°C for 48 h, with shaking at 160 rpm for liquid cultures, with the exception of Cryptococcus albidus, which was grown at 30°C for 72 h. All Aspergillus isolates were grown at 30 to 37°C for 48 to 72 h, while all other filamentous isolates were grown at 22 to 30°C for 5 through 10 days. Clinical isolates were subcultured onto Sabouraud agar and grown at 37°C for 72 h. C. albicans clinical isolates were verified using either the germ tube test in serum or the Murex C. albicans CA50 kit (Murex Diagnostics Ltd., Bereluk, B.V. Straatweg, Dreubelen).

DNA extraction. (i) Preparation of template DNA from filamentous fungi. The method used for extraction of DNA from filamentous fungi was a modification of methods previously described (3, 10). Briefly, filamentous fungal mycelia grown in 25 ml of Sabouraud broth as described above were harvested by filtration and washed once with sterile distilled water. The harvested mycelia were transferred to a microcentrifuge tube containing 0.5-mm-diameter zirconium glass beads (Biospec Products, Bartlesville, Okla.) stored in 0.2% sodium dodecyl sulfate (SDS) (BDH). Cell destruction was achieved by shaking the microcentrifuge tube at maximum speed for 190 s in a Mini Beadbeater (Biospec Products, Bartlesville, Okla.). Nucleic acids were purified with 900 µl of L6 buffer (10 M guanidium thiocyanate, 0.1 M Tris-HCl [pH 6.4], 0.2 M EDTA, 2.6% [vol/vol] Triton X-100) (Sigma-Aldrich Ltd.) and 40 µl of silica dioxide suspension (Sigma-Aldrich Ltd.) at room temperature for 10 min. The sample was then centrifuged at 13,000 rpm (14,926 × g) for 1 min, and the silica pellet was washed twice with 1 ml of L2 buffer (10 M guanidium thiocyanate, 0.1 M Tris-HCl [pH 6.4]), followed by two washes in 1 ml of 70% ethanol (BDH; Merck Ltd.) and one wash in 1 ml of 100% acetone (BDH; Merck Ltd.). The pellet was air dried, and the DNA was eluted in 100 µl of sterile 0.1× TE buffer (10 mM Tris-HCl, 1 mM EDTA [pH 8.0]) (Sigma-Aldrich Ltd.).

(ii) Preparation of template DNA from yeast isolates. DNA was extracted from yeast isolates using a modification of a previously described method (21). Briefly, 25 ml of yeast culture grown as described above was harvested by centrifugation at 6,000 rpm (5,000 × g; Beckman J2-21 centrifuge) for 15 min and washed once in distilled water. The cell pellet was resuspended in 7 ml of lysis buffer (10 mM Tris-HCl [pH 8.0], 250 mM EDTA [pH 8.0], 0.5% Triton X-100 [vol/vol]) supplemented with 3 mg of lyticase enzyme (Sigma-Aldrich Ltd.) per ml and incubated at 37°C overnight. Spheroplasts were subsequently lysed by incubating the samples with 3 mg of proteinase K per ml (Roche-Boehringer Mannheim Diagnostics, Mannheim, Germany) at 55°C for 2 h. The proteinase K was inactivated at 95°C for 10 min. An equal volume of phenol-chloroform-isoamyl alcohol (25:24:1; Sigma-Aldrich, Steinheim, Germany) was added, the tube was mixed by inversion and centrifuged at 12,000 rpm (19,800 × g; Beckman J2-21 centrifuge) for 1 h. The aqueous layer was transferred to a fresh tube, and an equal volume of ice-cold isopropanol was added. The DNA was precipitated by centrifugation at 12,000 rpm (19,800 × g; Beckman J2-21 centrifuge) for 15 min. The pellet was washed twice with 70% ethanol. Air-dried pellets were resuspended in 200 µl of 0.1× TE buffer (Sigma-Aldrich Ltd.) and stored at 20°C. DNA concentrations were estimated against known concentrations of lambda DNA (New England Biolabs) using a densitometer with Bio-ID V.96 software (Vilber Lourmat, Marne La Vallée, France) and by spectrophotometric analysis of absorbances at 260 and 280 nm using a Heios  spectrophotometer. Candida and Cryptococcus DNA was diluted to a working concentration of 10 ng/µl. Aspergillus DNA was diluted to a 1-ng/µl working stock.

(iii) Rapid preparation of template DNA from yeast cells. To facilitate PCR analysis of a large number of clinical isolates, a method was developed for the rapid preparation of template DNA from yeast isolates. A single colony was removed from the plate using a micropipette tip and resuspended by grinding in 100 µl of lysis buffer (0.1 M EDTA, 0.1 M NaOH). The sample was vortexed, and a 5-µl aliquot was used per 100-µl PCR mixture.

(iv) Preparation of C. albicans DNA template from inoculated blood samples. Blood samples (200 µl) inoculated with yeast cells (1 × 10⁵ to 1 × 10¹ cells) were lysed in 800 µl of lysis buffer (10 mM Tris-HCl [pH 7.5], 10 mM EDTA, 50 mM NaCl) at room temperature for 10 min and centrifuged at 13,000 rpm (14,926 × g; Heraeus Sepatech Biofuge A) for 5 min, the supernatant was discarded, and the pellet was resuspended in 100 µl of sterile H₂O. Blood samples (1 ml and 5 ml) inoculated with yeast cells (10⁵ to 10¹ cells) were lysed in 3 ml of lysis buffer and centrifuged at 2,000 rpm (500 × g; Beckman J2-21 Centrifuge) for 10 min, the supernatant was discarded, and the pellet was resuspended in 100 µl of sterile H₂O. Glass beads (0.5-mm-diameter zirconium glass beads stored in 0.2% SDS) were added to the resuspended pellet, the sample was vortexed in a minibead beater at top speed for 190 s, and the DNA was extracted as previously described (4). Briefly, the suspension was removed, following bead beating, to a fresh microcentrifuge tube containing 900 µl of L7 buffer (10 M GuSCN, 100 mM Tris-HCl [pH 6.4], 200 mM EDTA, 2.6% [wt/vol] Triton X-100, and alpha-casein [Sigma-Aldrich Ltd.]) (added to a final concentration of 1 mg/ml), and 40 µl of silica suspension for 200-µl and 1-ml samples and was scaled up to 80 µl for 5-ml samples. The sample was vortexed at maximum speed for 30 s followed by incubation at room temperature for 10 min. The sample was vortexed again for 30 s and spun at 12,000 rpm (12,700 × g; Heraeus Sepatech Biofuge A) for 1 min. The supernatant was removed, and the pellet was washed twice in 1 ml of L2 buffer (10 M GuSCN, 100 mM Tris-HCl [pH 6.4]) which was vortexed for 30 s and spun at 12,000 rpm (12,700 × g; Heraeus Sepatech Biofuge A) for 1 min. The pellet was then washed twice in 1 ml of 70% ethanol followed by a final wash in 1 ml of 100% acetone. The pellet was dried in a heating block at 60°C for 10 min and resuspended in 100 µl of 0.1× TE buffer for 30 min. Similarly, blood samples (200 µl, 1 ml, and 5 ml) inoculated with yeast cells (10⁵ to 10¹ cells/ml) were lysed and minibead beaten, and DNA was extracted from the suspensions by using a QIAmp tissue kit (Qiagen, Los Angeles, Calif.).

Sequence information, oligonucleotide probes, and primer design. Oligonucleotide primers were obtained from Genosys Biotechnologies (Europe) Ltd., Cambridgeshire, United Kingdom. The oligonucleotide primer pairs used in this study were previously shown to amplify the 5.8S rDNA and the adjacent ITS regions (40). To facilitate the detection of ITS PCR products in the LiPA, PCR primers were modified at the 5' end with a biotin moiety. The oligonucleotide primers UP1 and RP1 (UP1, 5'-GCCTAATGTAATCCATGGCG-3'; RP1, 5'-CTCCATTGGATTATCCCAGCA-3') for amplification of a 306-bp internal standard control (ISC) were designed by Innogenetics (Ghent, Belgium) and were supplied modified with 5' biotin moieties for this study.

PCR. PCRs were performed in a final volume of 100 µl. PCR conditions for the amplification of the ITS region from yeast isolates were as follows; each reaction mixture contained 0.25 mM deoxynucleotide triphosphates (DU:dNTPs [2:1]), 1× reaction buffer (Promega), 3 mM MgCl₂, 1 U of uracil DNA glycosylase (25) (Roche-Boehringer Mannheim), 40 pmol each of the forward and reverse primers (ITS 5-ITS 4 pair for full ITS amplification, and ITS 5-ITS 2 pair for ITS-1 amplification), 2.5 U of Taq polymerase (Promega), and 5 µl of template DNA made to a final volume of 100 µl with nuclease-free water (Sigma-Aldrich, Ltd.). PCR amplification was performed in a Touchdown thermocycler (Hybaid, Middlesex, United Kingdom), with the following cycling conditions: 37°C for 10 min for one cycle followed by 94°C for 2 min for one cycle followed by 30 cycles of DNA denaturation at 94°C for 30 s, primer annealing at 55°C for 30 s, and DNA extension at 72°C for 2 min, with a final extension cycle at 72°C for 10 min.

LiPA. The INNO-LiPA fungal assay (Innogenetics) was performed essentially as described previously (35) and is based on the reverse-hybridization principle. Oligonucleotide probes for the LiPA were enzymatically provided with a polydeoxythreonine tail. Subsequently, probes were immobilized as parallel lines onto a nitrocellulose membrane, with the top line containing a positive-control biotinylated DNA. Briefly, 10 µl of biotinylated PCR product was denatured in a LiPA tray by adding an equal volume of denaturing solution (NaOH-EDTA) and incubating at room temperature for 5 min. A 1-ml aliquot of preheated (50°C) hybridization buffer (2× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate] and 0.1% SDS) together with a LiPA strip was added and incubated at 50°C for 30 min in a shaking water bath. The strips were stringently washed three times in 1 ml of hybridization buffertwice for 1 min at room temperature and once at 50°C for 15 min. These washes were followed by one wash in 1-ml of rinse solution for 1 min followed by incubation in 1 ml of an alkaline phosphatase-linked streptavidin conjugate at room temperature for 30 min. The strips were washed twice in 1 ml of rinse solution and once in 1 ml of substrate buffer for 1 min each at room temperature. Finally the strips were incubated in 1 ml of substrate solution (5-bromo-4-chloro-3-indolylphosphate and nitroblue tetrazolium diluted in substrate buffer) for 30 min at room temperature. The color reaction was stopped by adding distilled water to the strips. After drying, the strip results were interpreted by eye.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

PCR amplification of the ITS with universal fungal primers. Four PCR primers designed from the conserved nucleotide sequences of 18S, 5.8S, and 28S rRNA genes were used to amplify the ITS region. These primers were used in the following combinations: ITS 5-ITS 4, ITS 5-ITS 2, and ITS 3-ITS 4 to amplify the full ITS region, the ITS 1 region, and the ITS 2 regions, respectively. Using these primer combinations and the optimized PCR conditions described in Materials and Methods, PCR products were generated from all of the typed fungi listed in Table 1. Figure 1 illustrates the different amplicon sizes obtained for the full ITS region amplified from a selection of different fungus species. For consistent amplification of the ITS region from filamentous fungi, PCR conditions were modified to contain 15% glycerol and a hotstart at 95°C for 10 min. No PCR products were amplified with the ITS PCR primers for genomic DNA isolated from Clostridium perfringens, Mycobacterium bovis, Listeria monocytogenes, Escherichia coli, or human DNA.

View larger version (52K): [in this window] [in a new window]   FIG. 1.   Full-ITS PCR products amplified from different fungal species. Lanes 1 through 12 show full-ITS amplicons from C. albicans NCPF 3302, C. guillermondii NCPF 3896, C. glabrata NCPF 3700, C. tropicalis NCPF 3870, C. dubliniensis CD36, C. krusei NCPF 3847, C. krusei 3922, C. krusei NCPF 3845, C. kefyr NCPF 3898, A. fumigatus NCPF 2109, A. nidulans NCPF 7063, and A. niger NCPF 2828. Lane 13 represents a no-template PCR negative control. M, 100-bp molecular size marker.

Sequence analysis and probe design. Oligonucleotide probes were designed from ITS sequences available in the GenBank database. For C. dubliniensis, C. neoformans, and A. fumigatus strains for which insufficient sequence information was available in the GenBank database, the full ITS region was sequenced from two isolates of each species. One sequencing reaction was performed on each strand, and a consensus sequence was obtained using the UWGCG Fragment Assembly program.

Specificity of LiPA. The specificities of the individual probes in the LiPA were evaluated against the panel of fungi listed in Table 1. Biotinylated full-ITS amplicons generated from these species were reverse hybridized to the LiPA strips at 50°C. Figure 2 illustrates the specificity of a selection of these probes in the LiPA. All of the probes hybridized specifically with the ITS PCR products from the appropriate species with the following exceptions. The CA1 probe reacted with the full-ITS PCR products from C. dubliniensis. The CN1 probe reacted weakly with C. laurentii PCR products (data not shown). A. versicolor full-ITS amplicons generated from 11 individual isolates varied in size from approximately 650 to 700 bp and were found to cross-react with the A. nidulans AN1 probe and the A. flavus AFL1 probe (data not shown). Additionally, during this study, ITS sequences for A. parasiticus and A. nominus were submitted to the GenBank database (accession numbers AF027862, AF027860, AF027864, and AF027861). The AFL1 probe designed for the specific detection of A. flavus shared 100% homology with both of these ITS sequences, resulting in the design of a new A. flavus probe (AFL4) for the specific identification of A. flavus ITS PCR products. The AFL4 probe was found to hybridize with full-ITS and ITS1 PCR products from A. flavus, and although it showed weak cross-hybridization with full-ITS PCR products from some A. versicolor isolates, it did not show cross-reaction with the ITS1 PCR products from these isolates. The full-ITS regions from three A. versicolor isolates were sequenced to investigate whether the cross-reaction of some A. versicolor full-ITS PCR products with the AFL4 probe was a result of sequence homology within the full-ITS sequence of these A. versicolor isolates and the AFL4 probe sequence. The full-ITS sequences from these A. versicolor isolates showed 100% homology with the GenBank sequence entry (L76745) for A. versicolor and showed no significant homology with the AFL1 or AFL4 probe sequences in similar Blast searches. As a result of cross-reaction of A. versicolor full-ITS PCR products with the A. nidulans AN1 probe, a second DNA probe, AN2, was designed for the specific detection of A. nidulans which has been shown to have no cross-reaction with the full-ITS PCR products from A. versicolor and will replace AN1 for the detection of A. nidulans in future LiPA strips. The two DNA probes CD1 and CD2 for the detection of C. dubliniensis were designed from ITS 2 sequence information. DNA sequencing of the full-ITS region from C. dubliniensis allowed the design of an ITS 1 probe CD3, specific for the detection of C. dubliniensis.

View larger version (110K): [in this window] [in a new window]   FIG. 2.   LiPAs of full-ITS products amplified from typed fungal isolates. The positions of the control line and the 16 species-specific probes for fungal pathogens are shown on the left hand side of the photograph. LiPAs 1 through 12 represent full-ITS PCR amplicons from the following species: 1, C. albicans NCPF 3345; 2, C. parapsilosis NCPF 3872; 3, C. glabrata NCPF 3700; 4, C. tropicalis NCPF 3870; 5, C. krusei NCPF 3922; 6, C. krusei NCPF 3847; 7, C. dubliniensis CD36; 8, C. neoformans var. gatti NCPF 3756; 9, C. neoformans var. neoformans NCPF 3232; 10, A. fumigatus NCPF 2109; 11, A. nidulans NCPF 7063; 12, A. flavus NCPF 2199.

Analysis of clinical isolates. Further evaluation of the LiPA was carried out on a collection of clinical isolates and quality control samples from the Quality Assurance Laboratory, PHLS Central Public Laboratory, and isolates from diverse specimen types obtained from the Bacteriology Department, University College Hospital, Galway. PCR template DNA was prepared from the clinical isolates using DNA extraction methods A and C as described in Materials and Methods. Full-ITS PCR amplicons generated from the isolates were hybridized to the LiPA strips. A total of 127 clinical isolates were analyzed, and of these, 100 isolates were identified, including 77 dimorphic yeast species (66 C. albicans isolates, 4 C. glabrata isolates, 2 C. parapsilosis isolates, 2 C. krusei isolates, and 1 isolate each from C. tropicalis, C. dubliniensis, and C. neoformans) and 23 members of the Aspergillus genus (20 isolates of A. fumigatus, 2 isolates of A. flavus, and 1 isolate of A. nidulans). The identity of the C. albicans isolates was confirmed using the Murex C. albicans CA50 kit. The LiPA did not identify the 27 other fungal isolates for which DNA probes were not designed and included in the LiPA assay. These included 3 dimorphic yeast isolates (1 Saccharomyces sp., 1 C. laurentii isolate, and 1 C. guilliermondii isolate), 10 dermatophytes (isolates of Microsporum canis, Trichophyton rubrum, and Trichophyton verrucosum), and 14 filamentous fungal isolates which did not hybridize to DNA probes in the LiPA assay (3 A. niger isolates, 2 A. terreus isolates, 1 A. glaucus isolate, 4 Penicillium sp. isolates, 2 Fusarium sp. isolates, and 2 unknown isolates). With the exception of C. albicans, the number of isolates of each species tested was low, and while these isolates demonstrated the potential of the PCR-LiPA for species identification, ideally the study should be expanded to a larger number of isolates to confirm the reliability of the PCR-LiPA for species identification.

Sensitivity of the LiPA. The sensitivity of the PCR-LiPA was evaluated by performing PCR amplification of serial dilutions of purified genomic DNA (100 ng to 1 fg) isolated from C. albicans, C. neoformans, and A. fumigatus isolates using both the ITS 5-ITS 4 and the ITS 5-ITS 2 primer pairs. PCR products were hybridized to the LiPA strips, and detection limits of 50 pg for full-ITS and 50 fg for ITS 1 PCR amplicons generated from A. fumigatus genomic DNA were obtained by the LiPA with the A. fumigatus probes. Detection limits of 100 fg for full-ITS and ITS 1 amplicons generated from C. albicans DNA were obtained by the LiPA, with all three C. albicans probes, and 10 pg for full ITS and 1 pg for ITS 1 amplicons were obtained for the C. neoformans probes. Comparable detection limits were achieved for each species by Southern blot hybridization with the appropriate digoxigenin-labeled probes.

Application of the PCR-LiPA to blood samples. In this study, 200-µl, 1-ml, and 5-ml blood samples were inoculated with decreasing concentrations of C. albicans cells (10⁵ to 10¹ cells). The inoculated blood samples were pretreated to lyse and remove the erythrocytes, and the pellet of lymphocytes and yeast cells was resuspended, glass beads were added, and the suspension was minibead beaten for 190 s to mechanically disrupt the yeast cell walls in preparation for DNA extraction. DNA was extracted from the suspensions as previously described (4) and scaling up the silica suspension from 40 to 80 µl to facilitate DNA extraction from the 5-ml inoculated samples. A 20-µl aliquot of extracted DNA from the 200-µl and 1-ml blood samples and a 20-µl aliquot of a 1/10 dilution of DNA extracted from 5-ml blood samples containing the decreasing concentrations of yeast cells were PCR amplified with the ITS 5-ITS 4 primer pair, and 10 µl of each of the PCR products was analyzed by agarose gel electrophoresis with a second 10-µl aliquot hybridized to the LiPA strips. The detection limits achieved following agarose gel electrophoresis and hybridization in the LiPA assay were comparable at 50 cells/ml for 200-µl and 10 cells/ml for 1-ml blood samples, with a detection limit of 2 cells/ml for 5-ml blood samples (Fig. 3a and b). The 200-µl, 1-ml, and 5-ml blood samples inoculated with yeast cells (10⁵ to 10¹ cells/ml) were pretreated as previously described, and the DNA was extracted from the suspensions by using the QIAmp tissue kit from Qiagen by following the kit manual. PCR amplification of 20 µl of the extracted DNA with the ITS 5-ITS 4 primer pair followed by hybridization of 10 µl of the PCR mixtures to the LiPA strips yielded detection limits of 50 cells/ml for 200-µl blood samples, 10 cells/ml for 1-ml blood samples, and 2 cells/ml for the 5-ml blood samples, which were comparable to the detection limits achieved with the guanidium thiocyanate-silica DNA extraction method. An ISC in the form of an "artificial plasmid" was included in the PCR to monitor for PCR inhibition in the extracted DNA samples. This ISC was designed to be PCR amplified with a UP1-RP1 primer pair to yield a PCR product of 306 bp, which was smaller than the full-ITS PCR amplicons from Candida spp. (>600 bp). When the ISC was included at a concentration of 10⁵ copies of the PCR ISC plasmid/100 µl of mixtures of decreasing dilutions of C. albicans cells extracted from 1 ml of blood, a detection limit of 10 cells/ml was achieved (Fig. 3c and d).

View larger version (68K): [in this window] [in a new window]   FIG. 3.   PCR amplification and LiPA detection of ITS region from C. albicans inoculated into 1-ml blood samples and extracted as previously described (4). (A) PCR amplification of ITS region from C. albicans inoculated into 1-ml blood samples and extracted as previously described (4). Lane 1, 100-bp ladder; lane 2, 1× 105 cells; lane 3, 104 cells; lane 4, 103 cells; lane 5, 102 cells; lane 6, 101 cells; lane 7, PCR positive control; lane 8, PCR positive control; lane 9, negative PCR control; lane 10, 100-bp ladder. (B) LiPA hybridization of 10 µl of full-ITS PCR products from C. albicans inoculated into 1-ml blood samples and extracted as previously described (4). LiPA 1, 105 cells; LiPA 2, 104 cells; LiPA 3, 103 cells; LiPA 4, 102 cells; LiPA 5, 101 cells, LiPA 6, positive control; LiPA 7, positive control; LiPA 8, negative control. (C) PCR amplification of ITS region in the presence of an ISC from C. albicans inoculated into 1-ml blood samples and extracted as previously described (4). Lane 1, 100-bp ladder; lane 2, 104 cells; lane 3, 103 cells; lane 4, 102 cells; lane 5, 101 cells; lane 6, sample preparation negative control; lane 7, positive control; lane 8, positive control; lane 9, PCR-negative control; lane 10, PCR-negative control. (D) LiPA hybridization of 10 µl of full-ITS PCR products from C. albicans and ISC PCR products from C. albicans inoculated into 1-ml blood samples and extracted as previously described (4). LiPA 1, 104 cells; LiPA 2, 103 cells; LiPA 3, 102 cells; LiPA 4, 101 cells; LiPA 5, sample preparation negative control; LiPA 6, PCR-negative control; LiPA 7, PCR-positive control. Positive, positive control probe; CA1, C. albicans ITS probe; CA2, C. albicans ITS probe; CA3, C. albicans ITS probe; ISC, internal standard control probe.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Invasive mycotic infections are contributing to increased morbidity and mortality among immunocompromised patients. While the availability of new antifungal drugs has improved the management of infections, in many cases diagnosis of the fungal agent remains difficult. In this study we report on the development of a reverse-hybridization LiPA which, when combined with PCR amplification, detects and identifies clinically significant fungal pathogens including Candida, Aspergillus and Cryptococcus species.

Previous reports describe PCR assays for the detection of fungal pathogens that target highly conserved regions such as the 18S (8, 13, 27) and the 28S (27, 34) regions or single copy genes such as actin (20) or heat shock protein 90 (7). In this study the ITS region of the rRNA complex was chosen as a target for species-specific-probe design. This region is an ideal target for species identification because it has sufficient variation to allow for discrimination between species (1, 12), and this target is present in multiple copies in the fungal genome.

In this study we describe universal amplification of the ITS region from fungi combined with the hybridization of the ITS PCR products to species-specific oligonucleotide probes on the LiPA strip for the specific identification of fungal pathogens. The new species, C. dubliniensis, first described by Sullivan et al. (36), is typically identified as C. albicans by routine phenotypic methods although a PCR assay based on amplification of the pH-regulated PHR1 and PHR2 genes has been recently described for distinguishing these species (22). Sequence analysis of the ITS regions from both C. albicans and C. dubliniensis in this study revealed a high degree of homology between both species, and we observed cross-reaction between the ITS PCR products from these two species and their respective oligonucleotide probes. The CD2 probe, which was specific for C. dubliniensis, was designed from the ITS 2 region, and a second probe, CD3, designed from ITS 1 with a single-base-pair mismatch to C. albicans in the ITS1 probe region, was also specific for detection of C. dubliniensis. In this study we also observed cross-reaction of ITS PCR products from A. versicolor isolates with the A. flavus probe (AFL1) and the A. nidulans probe (AN1). In addition we observed different full-ITS amplicon sizes (650 to 700 bp) among 11 different isolates of A. versicolor that were PCR amplified as part of this study. Other authors have reported on genetic variation within the ITS region between isolates of the same species (28, 31, 32), in particular for Pneumocystis carinii (23, 26), which has been found to have up to 15 different ITS subtypes. However, cloning and sequencing the full-ITS region from three A. versicolor isolates showed 100% homology with A. versicolor sequences in the GenBank database.

The detection limit of the LiPA was 10-fold lower than that of agarose gel electrophoresis for DNA that was PCR amplified for C. albicans, A. fumigatus, and C. neoformans. Detection of C. albicans inoculated into blood was as low as 2 to 10 cells/ml and compared well to sensitivities reported by other researchers. A microtiter enzyme immunoassay (EIA) for the detection of Candida species in blood with a detection sensitivity of 2 cells/200 µl of sample and a real-time fluorescent PCR-based assay for the detection of three Candida species in a single reaction tube with a detection sensitivity of 100 cells/200 µl have previously been described (12, 15). A nested PCR assay for the detection of C. neoformans in cerebrospinal fluid is reported to have a detection limit of 10 cells (29). Other researchers determined a detection level of 5 pg for A. fumigatus rDNA 18S gene amplicons in a microtiter plate EIA (10). A nested specific PCR-EIA for A. fumigatus (13) based on sequences derived from the 18S rRNA gene had a detection limit of 1.7 ng/µl of genomic DNA, while a similar PCR-EIA for an Aspergillus mitochondrial gene had a sensitivity of 0.6 fg/ml (18). We report a detection limit for ITS 1 PCR amplicons from A. fumigatus of 50 fg or one fungal genome equivalent (5) in the PCR-LiPA assay.

During the course of this study we evaluated a number of different DNA extraction methods with the aim of developing a universal approach for the preparation of fungal DNA from Candida, Cryptococcus, and Aspergillus spp. in blood and/or respiratory specimens. Evaluation of different methods for the extraction of fungal DNA from blood has previously been reported in the literature (24). Most of the methods evaluated involved enzymatic approaches previously described (16, 24), with some modifications. In our laboratory these methods failed to consistently yield high-quality DNA, resulting in variation in the detection limits achieved between experiments (data not shown). We also noted that DNA extracted from A. fumigatus by these methods required a "hot start" PCR approach for sensitive PCR amplification of the extracted DNA. The modified guanidium thiocyanate-silica method recently described (4) proved to be an efficient DNA extraction method and enabled detection limits of 2 to 10 cells/ml to be consistently achieved from fungal DNA extracted from blood samples. The QIAmp tissue kit from Qiagen also proved to be an equally reliable and user-friendly approach for extracting DNA from blood samples.

We have developed a PCR-LiPA for detection and identification of clinically important species of Candida, Aspergillus, and Cryptococcus, and we have demonstrated the potential of the technology for species confirmation of a small number of clinical fungal isolates and the direct detection of C. albicans in inoculated blood samples. Application of the PCR-LiPA is being expanded to a range of clinical specimen types, including tissue specimens in many cases of which fungal morphology is not sufficiently distinct to be diagnostic, and a rapid nonculture-based identification method would have direct application.