Identification of Histoplasma capsulatum, Blastomyces dermatitidis, an
http://www.100md.com
《微生物临床杂志》
Associated Regional and University Pathologists, Inc., Institute for Clinical and Experimental Pathology, Salt Lake City, Utah
ARUP Microbiology Laboratory, Salt Lake City, Utah
Department of Pathology and Laboratory Services, University of Arkansas for Medical Sciences, Little Rock, Arkansas
ABSTRACT
The performance of repetitive-sequence-based PCR (rep-PCR) using the DiversiLab system for identification of Coccidioides species, Blastomyces dermatitidis, and Histoplasma capsulatum was assessed by comparing data obtained to colony morphology and microscopic characteristics and to nucleic acid probe results. DNA from cultures of 23 Coccidioides, 24 B. dermatitidis, 24 H. capsulatum, 3 Arthrographis, and 2 Malbranchea isolates was extracted using a microbial DNA isolation kit as recommended by Bacterial Barcodes, Inc. Rep-PCR and probe results agreed for 97.2% of the dimorphic fungi when 85% similarity was used as the criterion for identification. Two H. capsulatum isolates were not identified, but no isolates were misidentified. From 43 of those cultures (15 Coccidioides, 14 B. dermatitidis, 14 H. capsulatum, 3 Arthrographis, and 2 Malbranchea), DNA also was extracted using an IDI lysis kit, a simpler method. Rep-PCR and probe results agreed for 97.7% of the dimorphic fungi when a criterion of 90% similarity was used for identification. One H. capsulatum isolate could not be identified; no isolates were misidentified. Using 85% similarity for identification resulted in one misidentification. These data suggest that the DiversiLab system can be used to identify Coccidioides and B. dermatitidis and, possibly, H. capsulatum isolates.
TEXT
The dimorphic fungi Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides species are important human pathogens found in certain areas of the United States where the species are endemic. Infections caused by these fungi are similar. Clinical manifestations differ, depending on the patient's immune status and fungal inoculum. In the immune-competent host, infection following administration of a low-level inoculum is often asymptomatic, whereas immunocompromised patients are at risk for severe, disseminated disease. Serologic assays for detection of fungal antigens (H. capsulatum) or antibodies (Coccidioides species) may be useful, but neither assay approach is 100% sensitive or specific (16, 17). Therefore, diagnosis of disease caused by these fungi frequently requires culture and identification of isolated colonies. The identification method used most commonly in clinical mycology laboratories is nucleic acid hybridization with commercial DNA probes (Accuprobe; Gen-Probe, Inc., San Diego, CA). Accuprobes provide rapid results and have high sensitivity and specificity (12), with the following exceptions. The probe for B. dermatitidis gives a false-positive result with Paracoccidioides brasiliensis (8), which is geographically restricted to Latin America, and with Gymnascella hyalinospora (6) and Emmonsia parva (Accuprobe product insert; GenProbe, Inc.), which are rare human pathogens. A case in which the H. capsulatum probe gave a false-positive result with Chrysosporium has been reported previously (1). Additionally, Accuprobe does not identify Coccidioides isolates to the species level; C. posadasii and C. immitis are both called C. immitis (Accuprobe product insert; GenProbe, Inc.). Currently, it does not appear to be clinically important to distinguish the Coccidioides species; however, this may change as additional information is acquired.
The primary goal of this study was to evaluate automated repetitive-sequence-based PCR (rep-PCR) using the DiversiLab system (Bacterial Barcodes, Inc., Athens, GA) for identification of H. capsulatum, B. dermatitidis, and Coccidioides species. This method has been shown to identify commonly encountered Aspergillus species (5), Fusarium species (4), and dermatophytes (10). Briefly, after DNA extraction, the rep-PCR method uses primers that target and bind to multiple noncoding, repetitive sequences (generally 30 to 500 bp) interspersed throughout the fungal genome. The outwardly facing primers generally amplify between repetitive elements, in contrast to inwardly facing primers, which amplify the repetitive element itself (as in variable-number tandem-repeat analysis) (15). Multiple DNA amplicons of different sizes and various quantities (intensities) are generated during PCR. To accomplish this result, we tested isolates previously identified by use of Accuprobes as well as isolates with colony and microscopic characteristics similar to those of the three pathogens under study, including Arthrographis and Malbranchea isolates, which can resemble Coccidioides isolates. Secondary goals were to investigate the use of an alternative method of extracting DNA from fungal isolates for use with the DiversiLab system and to examine the impact of the geographic location of the patient on the fingerprint pattern.
Study design. This study, which was approved by the Institutional Review Board of the University of Utah, was conducted in two phases. First, the feasibility of using automated rep-PCR to identify H. capsulatum, B. dermatitidis, and Coccidioides was investigated. In this phase, testing was performed and optimized at Bacterial Barcodes, Inc. The dimorphic fungal rep-PCR patterns were added to the existing mold database. In phase 2, conducted at ARUP, the accuracy of rep-PCR was evaluated, the database for dimorphic fungi was expanded, an alternative DNA extraction method was assessed, and the impact of the geographic location of the facility submitting the specimen on fingerprint pattern was investigated.
Isolates. All patient information associated with the clinical isolates in this study, which were randomly selected from the ARUP culture collection, was removed. The dimorphic fungi tested in both phases of the study were previously identified at ARUP by using Accuprobes; Arthrographis and Malbranchea isolates were identified by conventional methods plus negative Accuprobe results. In phase 1, 10 H. capsulatum, 10 B. dermatitidis, and 8 Coccidioides isolates were tested by rep-PCR. DNA was extracted using an UltraClean microbial DNA isolation kit (Mo Bio) (Mo Bio Laboratories, Solana Beach, CA) as previously described (10), labeled with the species name and a number, and sent to Bacterial Barcodes, Inc. Phase 2 included 14 H. capsulatum, 14 B. dermatitidis, 15 Coccidioides, 3 Arthrographis, and 2 Malbranchea isolates. For all isolates in phase 2, two subcultures were prepared, and DNA from each was extracted using a different method (see below). For the dimorphic fungi, the identity of the geographic location of the patient where the specimen was collected was retrieved after rep-PCR was completed.
Identification. The first step in the identification process involved assessment of colony morphology (7, 14). For suspected H. capsulatum, B. dermatitidis, or Coccidioides isolates, nucleic acid hybridization was performed using Accuprobes according to the manufacturer's directions. For isolates not suspected to be dimorphic fungi, slide cultures were prepared in a biosafety level 2 cabinet and sealed with parafilm. The sealed plate containing the slide culture was observed microscopically. After identification, cultures other than Coccidioides were allowed to sporulate, and the spores were stored in sterile water at room temperature. Coccidioides cultures were stored in a locked –80°C freezer in brain heart infusion broth-10% glycerol. Prior to testing, fungal isolates were grown on potato dextrose agar slants at 30°C for 5 to 14 days.
DNA extraction. All work was performed in a biosafety level 2 biological safety cabinet until cells were lysed. In both phases of the study, DNA was extracted from a spore and mycelial mass by using a Mo Bio kit, as recommended by Bacterial Barcodes, in accord with the manufacturer's instructions with the following modifications. The vortex time was increased to 30 min and the volume of MD3 to 900 μl. In phase 2, extraction was also performed using an IDI lysis kit (GeneOhm Sciences, San Diego, CA). Briefly, approximately 1 cm2 of mycelia and/or spores was transferred to 1 ml of sterile, nuclease-free water by scraping with a sterile wooden stick. The material was pelleted by centrifugation for 1 min at 6,000 x g. The supernatant was removed, and the fungal material was resuspended in 200 μl of IDI sample buffer. The suspension was transferred to the IDI lysis tube, which contained glass beads. Tubes were vortexed at high speed for 5 min, placed in a boiling water bath for 15 min, and then centrifuged for 5 min at 16,000 x g to pellet the glass beads and cellular debris. The supernatant was removed and stored at –20°C until analyzed by rep-PCR.
Rep-PCR. The fungal DNA was amplified using a DiversiLab mold kit (Bacterial Barcodes, Inc.) for DNA fingerprinting in accord with the manufacturer's instructions. Briefly, 2 to 4 μl of genomic DNA (50 to 100 ng) was added to the rep-PCR master mix with fungus-specific primers, 2.5 U AmpliTaq, and 10x PCR buffer (Applied Biosystems, Inc., Foster City, CA) for a 25 μl total reaction mixture. Thermal cycling parameters were as follows: initial denaturation at 94°C for 30 s; 35 cycles of denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and extension at 70°C for 90 s; and a final extension at 70°C for 3 min. Thermal cycling was performed on a GeneAmp PCR 2700 system (Applied Biosystems, Inc.) at ARUP and on a GeneAmp PCR 9700 system (Applied Biosystems, Inc.) at Bacterial Barcodes. The amplicons were stored at –20°C until tested.
Detection and analysis of rep-PCR products were implemented using the DiversiLab system, in which the amplified fragments of various sizes and fluorescent intensities were separated and detected using a microfluidics chip with the Agilent 2100 Bioanalyzer. Rep-PCR was repeated when the fluorescence was below 70 fluorescence units, the sample graph had a drop in fluorescence below baseline, or a bubble in the well prevented determination of rep-PCR pattern. Further analysis was performed with web-based DiversiLab software, version 2.1.66—which uses the Pearson correlation coefficient and unweighted pair group method with arithmetic mean to automatically compare the rep-PCR-based DNA fingerprints of unknown isolates. All bands in the gel-like image are considered in the analysis, although bands of greater intensity are weighted more heavily using the Pearson correlation coefficient.
Species identification of the fungal cultures was based on the percentage similarity and clustering profile obtained from the dendrograms as well as on a visual comparison of the virtual-gel images. The fingerprint pattern of each culture in this study was compared to all fingerprints contained in the mold database, which includes fingerprints of Trichophyton rubrum, T. tonsurans, T. mentagrophytes, Microsporum canis, M. nanum, M. gypseum, Epidermophyton floccosum, several species of Aspergillus, zygomycetes, Fusarium species, and Penicillium species. After analysis of the data in this study, the thresholds for identification of H. capsulatum, B. dermatitidis, and Coccidioides species were determined using guidelines provided by Bacterial Barcodes as follows. The similarity cutoff selected for species classification is the lowest level of similarity seen between samples of the same species identified by another method (e.g., the DNA probe method).
Results and discussion. Results of all cultures for which DNA was extracted with the Mo Bio kit are shown in Fig. 1 (phase 1 is designated by a letter and phase 2 by a number). None of the fingerprint patterns matched those of other fungi in the fungal database. When 85% similarity was used as the criterion for identification, rep-PCR and probe results agreed for 69 (97.2%) of the 71 dimorphic fungi tested in both phases of the study, and no isolates were identified incorrectly. All B. dermatitidis isolates grouped together with 90% similarity. There were two distinct clusters of Coccidioides isolates, one with >85% similarity and the other with >90% similarity. In contrast, there was more variability among the H. capsulatum isolates, and two isolates (Histoplasma H and I) were not identified. These two H. capsulatum isolates were retested by rep-PCR and probe; repeat and initial results were the same for both test methods. Of the five cultures resembling Coccidioides species, both Malbranchea isolates grouped together, but none of the three Arthrographis isolates were identified.
Fingerprint patterns of cultures extracted using the IDI lysis kit are shown in Fig. 2. All Coccidioides and B. dermatitidis, both Malbranchea, 13 of 14 H. capsulatum (all except Histoplasma 12), and 2 of 3 Arthrographis isolates were correctly identified using the criterion of 85% similarity. However, Arthrographis 1 was incorrectly called H. capsulatum. When 90% similarity was used as the criterion for identification, 42 (97.7%) of the 43 dimorphic fungi (all B. dermatitidis isolates, all Coccidioides isolates, and all but 1 H. capsulatum isolate [Histoplasma 12]) and both Malbranchea isolates were correctly identified. None of the Arthrographis isolates were identified, but no isolate was identified incorrectly. Fingerprint patterns of most cultures extracted with Mo Bio and IDI had 90% similarity; exceptions were four Coccidioides isolates (Coccidioides 12 to 15), six H. capsulatum isolates (Histoplasma 1 to 3, 5, 6, and 8), one Malbranchea isolate (Malbranchea 2), and one Arthrographis isolate (Arthrographis 2).
Figure 3 illustrates the relationship between geographic location and fingerprint pattern for the Coccidioides isolates tested in phase 2 by use of Mo Bio for DNA extraction. The 15 cultures grouped into two unrelated clusters. The four cultures from patients in California were in one cluster; the other cluster included cultures from patients in states other than California. When IDI lysis tubes were used for DNA extraction, one culture from California (Coccidioides 13; Fig. 2) did not group with the other samples submitted from California. There was no relationship between geographic location and fingerprint pattern for B. dermatitidis or H. capsulatum isolates (data not shown).
Dimorphic fungi are among the most important fungi encountered in a clinical mycology laboratory. Currently, use of nucleic acid probes is the method of choice for identifying these organisms. Our results showed that automated rep-PCR using the DiversiLab system can also be used to identify B. dermatitidis, Coccidioides, and, possibly, H. capsulatum isolates. Performance was excellent for identification of B. dermatitidis isolates, all of which had very similar fingerprint patterns, and Coccidioides isolates, which grouped into two unrelated clusters. When Mo Bio was used for DNA extraction, as recommended by Bacterial Barcodes, one cluster of Coccidioides isolates was composed only of isolates from California and the other included Coccidioides isolates only from states other than California. When IDI was used for extraction, one of the four California isolates grouped with the isolates from other states. These separate clusters could represent C. immitis, which in the United States has been detected only in California, in an area centered on the San Joaquin Valley, and C. posadasii, which has been detected almost exclusively in other states, although a few cases of C. posadasii infection have been found in California (3). Confirmation of this hypothesis requires additional testing, which is beyond the scope of this study. In contrast to B. dermatitidis and Coccidioides isolate results, the fingerprint patterns of H. capsulatum isolates were much more diverse, and a few isolates could not be identified. The addition of 14 isolates of H. capsulatum to the database in phase 2 allowed identification of Histoplasma J, which was not identified in phase 1, when there were only results for 10 H. capsulatum isolates in the database. This suggests that identification of a more variable species, such as H. capsulatum, may be improved as more isolates are added to the database.
Although the Mo Bio DNA extraction method is recommended by Bacterial Barcodes, it is quite labor intensive. For this reason, we examined an alternative method of extracting fungal DNA, one that uses the IDI lysis kit. We selected the IDI kit because it has the advantages of decreased time (30 min for IDI lysis tubes and 1.5 h for Mo Bio) and less manipulation of the sample (three tubes for IDI versus eight tubes for Mo Bio). A drawback to the IDI kit, however, is cost; list prices per extraction are $1.76 for Mo Bio and $5.25 for IDI. Identification results with the IDI kit were similar to those with Mo Bio; however, we had to apply a more stringent criterion for identification (i.e., 90% agreement rather than the 85% agreement criterion used with Mo Bio) to avoid misidentification of isolates. This suggests that DNA extraction methods other than Mo Bio, as recommended by Bacterial Barcodes, could be used with the DiversiLab system, but a thorough evaluation and validation prior to implementation is essential.
Purchasing the DiversiLab system solely for the purpose of identifying dimorphic fungi may not be practical for the following reasons. First, the number of dimorphic fungi encountered in most clinical mycology laboratories is small; second, for the automated rep-PCR method to be fully cost effective, a full chip of 13 wells should be used in a given run. However, for laboratories using the DiversiLab system and for those considering acquiring it, identification of these three dimorphic fungi is yet another use, in addition to bacterial strain typing (9, 11), identification of the commonly encountered Aspergillus species (5), Candida species (2), dermatophytes (10), and Fusarium species (4), and, potentially, identification of some mycobacteria (G. Hecox and G. Woods, Abstr. ASM Gen. Meet., abstr. C-026, 2005). For these laboratories, use of the DiversiLab system would be both effective and efficient with respect to cost and time compared to commercial DNA probe methods. The cost (list price) for fungal identification using the mold kit with the DiversiLab system is $27.98 per sample, assuming that a full chip of 13 wells is analyzed. The cost (list price) of the Accuprobe assay, assuming one patient sample and positive and negative controls are tested, is $97.50. The time to a result is slightly longer for rep-PCR: approximately 3.5 h postextraction, including analysis of data, for rep-PCR versus 2.5 h for the DNA probe method.
In summary, our data showed that the DiversiLab system can be used for identification of Coccidioides and B. dermatitidis isolates and that it has the potential to identify H. capsulatum isolates. As with many identification systems based on pattern comparison, a larger database should improve the performance criteria in such categories as typability and discriminatory power (13). In our opinion, expanding the database of rep-PCR patterns should improve the ability of the DiversiLab system to identify more genetically variable species such as H. capsulatum. Moreover, testing additional isolates that morphologically resemble the dimorphic fungi, including not only Arthrographis and Malbranchea isolates, which can resemble Coccidioides isolates, but also Chrysosporium and Emmonsia isolates, which may resemble B. dermatitidis or H. capsulatum isolates, will allow identification of the pathogenic dimorphic fungi examined in this study with greater confidence.
ACKNOWLEDGMENTS
We thank Katie Goates Ludwig and Claudia Barton for technical assistance. We acknowledge the staff at Bacterial Barcodes, Inc., for development of the rep-PCR for fungi and testing samples for the feasibility phase of this study. We thank Mimi Healy, Stacie Frye, and Maricel Lising for their input and technical expertise for all phases of this study.
Support for this project was provided in part by the Associated Regional and University Pathologists Institute for Clinical and Experimental Pathology.
FOOTNOTES
Corresponding author. Mailing address: ARUP Laboratories, 500 Chipeta Way, Salt Lake City, UT 84108. Phone: (801) 583-2787. Fax: (801) 584-5109. E-mail: june.pounder@aruplab.com.
REFERENCES
Brandt, M. E., D. Gaunt, N. Iqbal, S. McClinton, S. Hambleton, and L. Sigler. 2005. False-positive Histoplasma capsulatum Gen-Probe chemiluminescent test result caused by a Chrysosporium species. J. Clin. Microbiol. 43:1456-1458.
Chau, A. S., C. A. Mendrick, F. J. Sabatelli, D. Loebenberg, and P. M. McNicholas. 2004. Application of real-time quantitative PCR to molecular analysis of Candida albicans strains exhibiting reduced susceptibility to azoles. Antimicrob. Agents Chemother. 48:2124-2131.
Fisher, M. G., G. L. Koenig, T. J. White, and J. M. Taylor. 2002. Molecular and phenotypic description of Coccidioides posadasii sp. Nov., previously recognized as the non-California population of Coccidioides immitis. Mycologia 94:73-84.
Healy, M., K. Reece, D. Walton, J. Huong, S. Frye, I. I. Raad, and D. P. Kontoyiannis. 2005. Use of the DiversiLab system for species and strain differentiation of Fusarium species isolates. J. Clin. Microbiol. 43:5278-5280.
Healy, M., K. Reece, D. Walton, J. Huong, K. Shah, and D. P. Konyoyiannis. 2004. Identification to the species level and differentiation between strains of Aspergillus clinical isolates by automated repetitive-sequence-based PCR. J. Clin. Microbiol. 42:4016-4024.
Iwen, P. C., L. Sigler, S. Tarantolo, D. A. Sutton, M. G. Rinaldi, R. P. Lackner, D. I. McCarthy, and S. H. Hinrichs. 2000. Pulmonary infection caused by Gymnascella hyalinospora in a patient with acute myelogenous leukemia. J. Clin. Microbiol. 38:375-381.
Larone, D. H. 2002. Medically important fungi: a guide to identification, 4th ed. American Society for Microbiology, Washington, D.C.
Padhye, A. A., G. Smith, P. G. Standard, D. McLaughlin, and L. Kaufman. 1994. Comparative evaluation of chemiluminescent DNA probe assays and exoantigen tests for rapid identification of Blastomyces dermatitidis and Coccidioides immitis. J. Clin. Microbiol. 32:867-870.
Pounder, J. I., C. K. Shutt, B. Schaecher, and G. L. Woods. 2006. Clinical evaluation of repetitive sequence-based polymerase chain reaction using the DiversiLab system for strain typing of vancomycin-resistant enterococci. Diagn. Microbiol. Infect. Dis. 54:183-187.
Pounder, J. I., S. Williams, D. Hansen, M. Healy, K. Reece, and G. L. Woods. 2005. Repetitive-sequence-PCR-based DNA fingerprinting using the DiversiLab system for identification of commonly encountered dermatophytes. J. Clin. Microbiol. 43:2141-2147.
Shutt, C. K., J. I. Pounder, S. R. Page, B. J. Schaecher, and G. L. Woods. 2005. Clinical evaluation of the DiversiLab Microbial typing system using rep-PCR for strain characterization of Staphylococcus aureus. J. Clin. Microbiol. 43:1187-1192.
Stockman, L., K. A. Clark, J. M. Hunt, and G. D. Roberts. 1993. Evaluation of commercially available acridinium ester-labeled chemiluminescent DNA probes for culture identification of Blastomyces dermatitidis, Coccidioides immitis, Cryptococcus neoformans, and Histoplasma capsulatum. J. Clin. Microbiol. 31:845-850.
Streulens, M. J., and the Members of the European Study Group on Epidemiological Markers of the European Society for Clinical Microbiology and Infectious Disease. 1996. Consensus guidelines for appropriate use and evaluation of microbial epidemiological typing systems. Clin. Microbiol. Infect. 2:2-11.
Sutton, D. A., A. W. Fothergill, and M. G. Rinaldi. 1998. Guide to clinically significant fungi. Williams and Wilkins, Baltimore, MD.
Versalovic, J., T. Koeuth, and J. R. Lupski. 1991. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res. 19:6823-6831.
Wheat, J. L. 2003. Current diagnosis of histoplasmosis. Trends Microbiol. 11:488-494.
Yang, M. C., D. M. Magee, and R. A. Cox. 1997. Mapping of a Coccidioides immitis-specific epitope that reacts with complement-fixing antibody. Infect. Immun. 65:4068-4074.(June I. Pounder, Dewey Hansen, and Gail )
ARUP Microbiology Laboratory, Salt Lake City, Utah
Department of Pathology and Laboratory Services, University of Arkansas for Medical Sciences, Little Rock, Arkansas
ABSTRACT
The performance of repetitive-sequence-based PCR (rep-PCR) using the DiversiLab system for identification of Coccidioides species, Blastomyces dermatitidis, and Histoplasma capsulatum was assessed by comparing data obtained to colony morphology and microscopic characteristics and to nucleic acid probe results. DNA from cultures of 23 Coccidioides, 24 B. dermatitidis, 24 H. capsulatum, 3 Arthrographis, and 2 Malbranchea isolates was extracted using a microbial DNA isolation kit as recommended by Bacterial Barcodes, Inc. Rep-PCR and probe results agreed for 97.2% of the dimorphic fungi when 85% similarity was used as the criterion for identification. Two H. capsulatum isolates were not identified, but no isolates were misidentified. From 43 of those cultures (15 Coccidioides, 14 B. dermatitidis, 14 H. capsulatum, 3 Arthrographis, and 2 Malbranchea), DNA also was extracted using an IDI lysis kit, a simpler method. Rep-PCR and probe results agreed for 97.7% of the dimorphic fungi when a criterion of 90% similarity was used for identification. One H. capsulatum isolate could not be identified; no isolates were misidentified. Using 85% similarity for identification resulted in one misidentification. These data suggest that the DiversiLab system can be used to identify Coccidioides and B. dermatitidis and, possibly, H. capsulatum isolates.
TEXT
The dimorphic fungi Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides species are important human pathogens found in certain areas of the United States where the species are endemic. Infections caused by these fungi are similar. Clinical manifestations differ, depending on the patient's immune status and fungal inoculum. In the immune-competent host, infection following administration of a low-level inoculum is often asymptomatic, whereas immunocompromised patients are at risk for severe, disseminated disease. Serologic assays for detection of fungal antigens (H. capsulatum) or antibodies (Coccidioides species) may be useful, but neither assay approach is 100% sensitive or specific (16, 17). Therefore, diagnosis of disease caused by these fungi frequently requires culture and identification of isolated colonies. The identification method used most commonly in clinical mycology laboratories is nucleic acid hybridization with commercial DNA probes (Accuprobe; Gen-Probe, Inc., San Diego, CA). Accuprobes provide rapid results and have high sensitivity and specificity (12), with the following exceptions. The probe for B. dermatitidis gives a false-positive result with Paracoccidioides brasiliensis (8), which is geographically restricted to Latin America, and with Gymnascella hyalinospora (6) and Emmonsia parva (Accuprobe product insert; GenProbe, Inc.), which are rare human pathogens. A case in which the H. capsulatum probe gave a false-positive result with Chrysosporium has been reported previously (1). Additionally, Accuprobe does not identify Coccidioides isolates to the species level; C. posadasii and C. immitis are both called C. immitis (Accuprobe product insert; GenProbe, Inc.). Currently, it does not appear to be clinically important to distinguish the Coccidioides species; however, this may change as additional information is acquired.
The primary goal of this study was to evaluate automated repetitive-sequence-based PCR (rep-PCR) using the DiversiLab system (Bacterial Barcodes, Inc., Athens, GA) for identification of H. capsulatum, B. dermatitidis, and Coccidioides species. This method has been shown to identify commonly encountered Aspergillus species (5), Fusarium species (4), and dermatophytes (10). Briefly, after DNA extraction, the rep-PCR method uses primers that target and bind to multiple noncoding, repetitive sequences (generally 30 to 500 bp) interspersed throughout the fungal genome. The outwardly facing primers generally amplify between repetitive elements, in contrast to inwardly facing primers, which amplify the repetitive element itself (as in variable-number tandem-repeat analysis) (15). Multiple DNA amplicons of different sizes and various quantities (intensities) are generated during PCR. To accomplish this result, we tested isolates previously identified by use of Accuprobes as well as isolates with colony and microscopic characteristics similar to those of the three pathogens under study, including Arthrographis and Malbranchea isolates, which can resemble Coccidioides isolates. Secondary goals were to investigate the use of an alternative method of extracting DNA from fungal isolates for use with the DiversiLab system and to examine the impact of the geographic location of the patient on the fingerprint pattern.
Study design. This study, which was approved by the Institutional Review Board of the University of Utah, was conducted in two phases. First, the feasibility of using automated rep-PCR to identify H. capsulatum, B. dermatitidis, and Coccidioides was investigated. In this phase, testing was performed and optimized at Bacterial Barcodes, Inc. The dimorphic fungal rep-PCR patterns were added to the existing mold database. In phase 2, conducted at ARUP, the accuracy of rep-PCR was evaluated, the database for dimorphic fungi was expanded, an alternative DNA extraction method was assessed, and the impact of the geographic location of the facility submitting the specimen on fingerprint pattern was investigated.
Isolates. All patient information associated with the clinical isolates in this study, which were randomly selected from the ARUP culture collection, was removed. The dimorphic fungi tested in both phases of the study were previously identified at ARUP by using Accuprobes; Arthrographis and Malbranchea isolates were identified by conventional methods plus negative Accuprobe results. In phase 1, 10 H. capsulatum, 10 B. dermatitidis, and 8 Coccidioides isolates were tested by rep-PCR. DNA was extracted using an UltraClean microbial DNA isolation kit (Mo Bio) (Mo Bio Laboratories, Solana Beach, CA) as previously described (10), labeled with the species name and a number, and sent to Bacterial Barcodes, Inc. Phase 2 included 14 H. capsulatum, 14 B. dermatitidis, 15 Coccidioides, 3 Arthrographis, and 2 Malbranchea isolates. For all isolates in phase 2, two subcultures were prepared, and DNA from each was extracted using a different method (see below). For the dimorphic fungi, the identity of the geographic location of the patient where the specimen was collected was retrieved after rep-PCR was completed.
Identification. The first step in the identification process involved assessment of colony morphology (7, 14). For suspected H. capsulatum, B. dermatitidis, or Coccidioides isolates, nucleic acid hybridization was performed using Accuprobes according to the manufacturer's directions. For isolates not suspected to be dimorphic fungi, slide cultures were prepared in a biosafety level 2 cabinet and sealed with parafilm. The sealed plate containing the slide culture was observed microscopically. After identification, cultures other than Coccidioides were allowed to sporulate, and the spores were stored in sterile water at room temperature. Coccidioides cultures were stored in a locked –80°C freezer in brain heart infusion broth-10% glycerol. Prior to testing, fungal isolates were grown on potato dextrose agar slants at 30°C for 5 to 14 days.
DNA extraction. All work was performed in a biosafety level 2 biological safety cabinet until cells were lysed. In both phases of the study, DNA was extracted from a spore and mycelial mass by using a Mo Bio kit, as recommended by Bacterial Barcodes, in accord with the manufacturer's instructions with the following modifications. The vortex time was increased to 30 min and the volume of MD3 to 900 μl. In phase 2, extraction was also performed using an IDI lysis kit (GeneOhm Sciences, San Diego, CA). Briefly, approximately 1 cm2 of mycelia and/or spores was transferred to 1 ml of sterile, nuclease-free water by scraping with a sterile wooden stick. The material was pelleted by centrifugation for 1 min at 6,000 x g. The supernatant was removed, and the fungal material was resuspended in 200 μl of IDI sample buffer. The suspension was transferred to the IDI lysis tube, which contained glass beads. Tubes were vortexed at high speed for 5 min, placed in a boiling water bath for 15 min, and then centrifuged for 5 min at 16,000 x g to pellet the glass beads and cellular debris. The supernatant was removed and stored at –20°C until analyzed by rep-PCR.
Rep-PCR. The fungal DNA was amplified using a DiversiLab mold kit (Bacterial Barcodes, Inc.) for DNA fingerprinting in accord with the manufacturer's instructions. Briefly, 2 to 4 μl of genomic DNA (50 to 100 ng) was added to the rep-PCR master mix with fungus-specific primers, 2.5 U AmpliTaq, and 10x PCR buffer (Applied Biosystems, Inc., Foster City, CA) for a 25 μl total reaction mixture. Thermal cycling parameters were as follows: initial denaturation at 94°C for 30 s; 35 cycles of denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and extension at 70°C for 90 s; and a final extension at 70°C for 3 min. Thermal cycling was performed on a GeneAmp PCR 2700 system (Applied Biosystems, Inc.) at ARUP and on a GeneAmp PCR 9700 system (Applied Biosystems, Inc.) at Bacterial Barcodes. The amplicons were stored at –20°C until tested.
Detection and analysis of rep-PCR products were implemented using the DiversiLab system, in which the amplified fragments of various sizes and fluorescent intensities were separated and detected using a microfluidics chip with the Agilent 2100 Bioanalyzer. Rep-PCR was repeated when the fluorescence was below 70 fluorescence units, the sample graph had a drop in fluorescence below baseline, or a bubble in the well prevented determination of rep-PCR pattern. Further analysis was performed with web-based DiversiLab software, version 2.1.66—which uses the Pearson correlation coefficient and unweighted pair group method with arithmetic mean to automatically compare the rep-PCR-based DNA fingerprints of unknown isolates. All bands in the gel-like image are considered in the analysis, although bands of greater intensity are weighted more heavily using the Pearson correlation coefficient.
Species identification of the fungal cultures was based on the percentage similarity and clustering profile obtained from the dendrograms as well as on a visual comparison of the virtual-gel images. The fingerprint pattern of each culture in this study was compared to all fingerprints contained in the mold database, which includes fingerprints of Trichophyton rubrum, T. tonsurans, T. mentagrophytes, Microsporum canis, M. nanum, M. gypseum, Epidermophyton floccosum, several species of Aspergillus, zygomycetes, Fusarium species, and Penicillium species. After analysis of the data in this study, the thresholds for identification of H. capsulatum, B. dermatitidis, and Coccidioides species were determined using guidelines provided by Bacterial Barcodes as follows. The similarity cutoff selected for species classification is the lowest level of similarity seen between samples of the same species identified by another method (e.g., the DNA probe method).
Results and discussion. Results of all cultures for which DNA was extracted with the Mo Bio kit are shown in Fig. 1 (phase 1 is designated by a letter and phase 2 by a number). None of the fingerprint patterns matched those of other fungi in the fungal database. When 85% similarity was used as the criterion for identification, rep-PCR and probe results agreed for 69 (97.2%) of the 71 dimorphic fungi tested in both phases of the study, and no isolates were identified incorrectly. All B. dermatitidis isolates grouped together with 90% similarity. There were two distinct clusters of Coccidioides isolates, one with >85% similarity and the other with >90% similarity. In contrast, there was more variability among the H. capsulatum isolates, and two isolates (Histoplasma H and I) were not identified. These two H. capsulatum isolates were retested by rep-PCR and probe; repeat and initial results were the same for both test methods. Of the five cultures resembling Coccidioides species, both Malbranchea isolates grouped together, but none of the three Arthrographis isolates were identified.
Fingerprint patterns of cultures extracted using the IDI lysis kit are shown in Fig. 2. All Coccidioides and B. dermatitidis, both Malbranchea, 13 of 14 H. capsulatum (all except Histoplasma 12), and 2 of 3 Arthrographis isolates were correctly identified using the criterion of 85% similarity. However, Arthrographis 1 was incorrectly called H. capsulatum. When 90% similarity was used as the criterion for identification, 42 (97.7%) of the 43 dimorphic fungi (all B. dermatitidis isolates, all Coccidioides isolates, and all but 1 H. capsulatum isolate [Histoplasma 12]) and both Malbranchea isolates were correctly identified. None of the Arthrographis isolates were identified, but no isolate was identified incorrectly. Fingerprint patterns of most cultures extracted with Mo Bio and IDI had 90% similarity; exceptions were four Coccidioides isolates (Coccidioides 12 to 15), six H. capsulatum isolates (Histoplasma 1 to 3, 5, 6, and 8), one Malbranchea isolate (Malbranchea 2), and one Arthrographis isolate (Arthrographis 2).
Figure 3 illustrates the relationship between geographic location and fingerprint pattern for the Coccidioides isolates tested in phase 2 by use of Mo Bio for DNA extraction. The 15 cultures grouped into two unrelated clusters. The four cultures from patients in California were in one cluster; the other cluster included cultures from patients in states other than California. When IDI lysis tubes were used for DNA extraction, one culture from California (Coccidioides 13; Fig. 2) did not group with the other samples submitted from California. There was no relationship between geographic location and fingerprint pattern for B. dermatitidis or H. capsulatum isolates (data not shown).
Dimorphic fungi are among the most important fungi encountered in a clinical mycology laboratory. Currently, use of nucleic acid probes is the method of choice for identifying these organisms. Our results showed that automated rep-PCR using the DiversiLab system can also be used to identify B. dermatitidis, Coccidioides, and, possibly, H. capsulatum isolates. Performance was excellent for identification of B. dermatitidis isolates, all of which had very similar fingerprint patterns, and Coccidioides isolates, which grouped into two unrelated clusters. When Mo Bio was used for DNA extraction, as recommended by Bacterial Barcodes, one cluster of Coccidioides isolates was composed only of isolates from California and the other included Coccidioides isolates only from states other than California. When IDI was used for extraction, one of the four California isolates grouped with the isolates from other states. These separate clusters could represent C. immitis, which in the United States has been detected only in California, in an area centered on the San Joaquin Valley, and C. posadasii, which has been detected almost exclusively in other states, although a few cases of C. posadasii infection have been found in California (3). Confirmation of this hypothesis requires additional testing, which is beyond the scope of this study. In contrast to B. dermatitidis and Coccidioides isolate results, the fingerprint patterns of H. capsulatum isolates were much more diverse, and a few isolates could not be identified. The addition of 14 isolates of H. capsulatum to the database in phase 2 allowed identification of Histoplasma J, which was not identified in phase 1, when there were only results for 10 H. capsulatum isolates in the database. This suggests that identification of a more variable species, such as H. capsulatum, may be improved as more isolates are added to the database.
Although the Mo Bio DNA extraction method is recommended by Bacterial Barcodes, it is quite labor intensive. For this reason, we examined an alternative method of extracting fungal DNA, one that uses the IDI lysis kit. We selected the IDI kit because it has the advantages of decreased time (30 min for IDI lysis tubes and 1.5 h for Mo Bio) and less manipulation of the sample (three tubes for IDI versus eight tubes for Mo Bio). A drawback to the IDI kit, however, is cost; list prices per extraction are $1.76 for Mo Bio and $5.25 for IDI. Identification results with the IDI kit were similar to those with Mo Bio; however, we had to apply a more stringent criterion for identification (i.e., 90% agreement rather than the 85% agreement criterion used with Mo Bio) to avoid misidentification of isolates. This suggests that DNA extraction methods other than Mo Bio, as recommended by Bacterial Barcodes, could be used with the DiversiLab system, but a thorough evaluation and validation prior to implementation is essential.
Purchasing the DiversiLab system solely for the purpose of identifying dimorphic fungi may not be practical for the following reasons. First, the number of dimorphic fungi encountered in most clinical mycology laboratories is small; second, for the automated rep-PCR method to be fully cost effective, a full chip of 13 wells should be used in a given run. However, for laboratories using the DiversiLab system and for those considering acquiring it, identification of these three dimorphic fungi is yet another use, in addition to bacterial strain typing (9, 11), identification of the commonly encountered Aspergillus species (5), Candida species (2), dermatophytes (10), and Fusarium species (4), and, potentially, identification of some mycobacteria (G. Hecox and G. Woods, Abstr. ASM Gen. Meet., abstr. C-026, 2005). For these laboratories, use of the DiversiLab system would be both effective and efficient with respect to cost and time compared to commercial DNA probe methods. The cost (list price) for fungal identification using the mold kit with the DiversiLab system is $27.98 per sample, assuming that a full chip of 13 wells is analyzed. The cost (list price) of the Accuprobe assay, assuming one patient sample and positive and negative controls are tested, is $97.50. The time to a result is slightly longer for rep-PCR: approximately 3.5 h postextraction, including analysis of data, for rep-PCR versus 2.5 h for the DNA probe method.
In summary, our data showed that the DiversiLab system can be used for identification of Coccidioides and B. dermatitidis isolates and that it has the potential to identify H. capsulatum isolates. As with many identification systems based on pattern comparison, a larger database should improve the performance criteria in such categories as typability and discriminatory power (13). In our opinion, expanding the database of rep-PCR patterns should improve the ability of the DiversiLab system to identify more genetically variable species such as H. capsulatum. Moreover, testing additional isolates that morphologically resemble the dimorphic fungi, including not only Arthrographis and Malbranchea isolates, which can resemble Coccidioides isolates, but also Chrysosporium and Emmonsia isolates, which may resemble B. dermatitidis or H. capsulatum isolates, will allow identification of the pathogenic dimorphic fungi examined in this study with greater confidence.
ACKNOWLEDGMENTS
We thank Katie Goates Ludwig and Claudia Barton for technical assistance. We acknowledge the staff at Bacterial Barcodes, Inc., for development of the rep-PCR for fungi and testing samples for the feasibility phase of this study. We thank Mimi Healy, Stacie Frye, and Maricel Lising for their input and technical expertise for all phases of this study.
Support for this project was provided in part by the Associated Regional and University Pathologists Institute for Clinical and Experimental Pathology.
FOOTNOTES
Corresponding author. Mailing address: ARUP Laboratories, 500 Chipeta Way, Salt Lake City, UT 84108. Phone: (801) 583-2787. Fax: (801) 584-5109. E-mail: june.pounder@aruplab.com.
REFERENCES
Brandt, M. E., D. Gaunt, N. Iqbal, S. McClinton, S. Hambleton, and L. Sigler. 2005. False-positive Histoplasma capsulatum Gen-Probe chemiluminescent test result caused by a Chrysosporium species. J. Clin. Microbiol. 43:1456-1458.
Chau, A. S., C. A. Mendrick, F. J. Sabatelli, D. Loebenberg, and P. M. McNicholas. 2004. Application of real-time quantitative PCR to molecular analysis of Candida albicans strains exhibiting reduced susceptibility to azoles. Antimicrob. Agents Chemother. 48:2124-2131.
Fisher, M. G., G. L. Koenig, T. J. White, and J. M. Taylor. 2002. Molecular and phenotypic description of Coccidioides posadasii sp. Nov., previously recognized as the non-California population of Coccidioides immitis. Mycologia 94:73-84.
Healy, M., K. Reece, D. Walton, J. Huong, S. Frye, I. I. Raad, and D. P. Kontoyiannis. 2005. Use of the DiversiLab system for species and strain differentiation of Fusarium species isolates. J. Clin. Microbiol. 43:5278-5280.
Healy, M., K. Reece, D. Walton, J. Huong, K. Shah, and D. P. Konyoyiannis. 2004. Identification to the species level and differentiation between strains of Aspergillus clinical isolates by automated repetitive-sequence-based PCR. J. Clin. Microbiol. 42:4016-4024.
Iwen, P. C., L. Sigler, S. Tarantolo, D. A. Sutton, M. G. Rinaldi, R. P. Lackner, D. I. McCarthy, and S. H. Hinrichs. 2000. Pulmonary infection caused by Gymnascella hyalinospora in a patient with acute myelogenous leukemia. J. Clin. Microbiol. 38:375-381.
Larone, D. H. 2002. Medically important fungi: a guide to identification, 4th ed. American Society for Microbiology, Washington, D.C.
Padhye, A. A., G. Smith, P. G. Standard, D. McLaughlin, and L. Kaufman. 1994. Comparative evaluation of chemiluminescent DNA probe assays and exoantigen tests for rapid identification of Blastomyces dermatitidis and Coccidioides immitis. J. Clin. Microbiol. 32:867-870.
Pounder, J. I., C. K. Shutt, B. Schaecher, and G. L. Woods. 2006. Clinical evaluation of repetitive sequence-based polymerase chain reaction using the DiversiLab system for strain typing of vancomycin-resistant enterococci. Diagn. Microbiol. Infect. Dis. 54:183-187.
Pounder, J. I., S. Williams, D. Hansen, M. Healy, K. Reece, and G. L. Woods. 2005. Repetitive-sequence-PCR-based DNA fingerprinting using the DiversiLab system for identification of commonly encountered dermatophytes. J. Clin. Microbiol. 43:2141-2147.
Shutt, C. K., J. I. Pounder, S. R. Page, B. J. Schaecher, and G. L. Woods. 2005. Clinical evaluation of the DiversiLab Microbial typing system using rep-PCR for strain characterization of Staphylococcus aureus. J. Clin. Microbiol. 43:1187-1192.
Stockman, L., K. A. Clark, J. M. Hunt, and G. D. Roberts. 1993. Evaluation of commercially available acridinium ester-labeled chemiluminescent DNA probes for culture identification of Blastomyces dermatitidis, Coccidioides immitis, Cryptococcus neoformans, and Histoplasma capsulatum. J. Clin. Microbiol. 31:845-850.
Streulens, M. J., and the Members of the European Study Group on Epidemiological Markers of the European Society for Clinical Microbiology and Infectious Disease. 1996. Consensus guidelines for appropriate use and evaluation of microbial epidemiological typing systems. Clin. Microbiol. Infect. 2:2-11.
Sutton, D. A., A. W. Fothergill, and M. G. Rinaldi. 1998. Guide to clinically significant fungi. Williams and Wilkins, Baltimore, MD.
Versalovic, J., T. Koeuth, and J. R. Lupski. 1991. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res. 19:6823-6831.
Wheat, J. L. 2003. Current diagnosis of histoplasmosis. Trends Microbiol. 11:488-494.
Yang, M. C., D. M. Magee, and R. A. Cox. 1997. Mapping of a Coccidioides immitis-specific epitope that reacts with complement-fixing antibody. Infect. Immun. 65:4068-4074.(June I. Pounder, Dewey Hansen, and Gail )