INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Invasive aspergillosis is associated with a high rate of mortality among patients receiving bone marrow transplantations or those with leukemia, other cancers, or respiratory ailments. Medical advances that predispose patients to invasive aspergillosis include intravenous catheterization and treatment with immunosuppressive drugs, radiation, and high doses of corticosteroids, among others (2, 5, 24, 28). Early diagnosis and initiation of antifungal therapy are therefore essential to reduce the high rate of mortality. Traditional diagnostic methods used in the clinical laboratory include microscopy, culture, or antigen detection. However, these methods are time-consuming, have low degrees of sensitivity, are difficult to standardize, and/or are nonspecific (3, 6, 7, 13, 14). Furthermore, misidentification can usually occur because some fungi may be poorly characterized by classical techniques or by the inexperienced person. Delays in diagnosis may impede the selection of antifungal agents when identification of the pathogen is required (21, 23). PCR-based techniques that target DNA have avoided many of these problems and provide an alternative approach. PCR-based assays have recently been adapted to the detection and identification of pathogenic Aspergillus spp. The high degrees of sensitivity and specificity of these methods provide a method for the early diagnosis of invasive aspergillosis (6, 7, 10, 12). The goal of our study was to develop a practical, quick, cheap, and specific nested PCR method for the identification of A. fumigatus, the most common of the Aspergillus pathogens.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Organisms and culture. Twenty-four isolates of A. fumigatus and variants, eight isolates of Aspergillus nidulans, seven isolates of Aspergillus flavus and variants, eight isolates of Aspergillus terreus, nine isolates of Aspergillus niger, one isolate of each of Aspergillus parasiticus, Aspergillus penicilloides, Aspergillus versicolor, Aspergillus wangduanlii, Aspergillus qizutongii, Aspergillus beijingensis, and Exophiala dermatitidis, four isolates of Candida, four isolates of bacteria, and human DNA were used. All molds were initially identified by morphology and all yeasts were identified by fermentation and assimilation tests with the API 20C AUX system (Table 1). The bacterial type cultures were provided by The Quality Control Center for Clinical Laboratory, Ministry of Health (Beijing, People's Republic of China). All filamentous fungi were inoculated onto potato dextrose agar slants (Difco, Detroit, Mich.) at 27°C for 72 to 120 h and were stored at 4°C until needed. Yeasts were inoculated in yeast extract-peptone-glucose broth (1% yeast extract, 2% Bacto Peptone, 1% glucose; Difco) and were shaken at 27°C for 18 h. Bacteria were inoculated into Luria-Bertani medium (Bacto-Tryptone, 10 g/liter; Bacto Yeast Extract, 5 g/liter; NaCl, 10 g/liter [pH 7.0]) and were shaken for 18 h at 27°C.

DNA preparation. (i) Fungal DNA isolation methods were adopted as described previously (29). Briefly, mycelia were gently removed from cultures, while yeasts were pelleted in a 1.5-ml Eppendorf tube. Five hundred microliters of extraction buffer (100 mM Tris-HCl [pH 9.0], 40 mM EDTA), 60 µl of 20% sodium dodecyl sulfate, and 300 µl of benzyl chloride were added to each sample. The reaction mixture was vortexed and incubated in a 50°C water bath for 40 min and then shaken for 10 min so that the two phases were mixed thoroughly. Then, 60 µl of 3 M sodium acetate (pH 5.0) was added, and the tube was kept on ice for 20 min. After centrifugation at 3,500 × g at 4°C for 15 min, the supernatant was collected and DNA was precipitated with isopropanol (1:1). The DNA pellet was resuspended in 300 µl of TE buffer (10 mM Tris-HCl [pH 7.4]-1 mM EDTA), and 1.5 µl of RNase (10 mg/ml) was added. After 5 min, the samples were extracted with phenol-chloroform (1:1 [vol/vol]) and, following chloroform extraction, were precipitated with isopropanol. The DNA pellet was resuspended in 200 µl of TE buffer, and 3 to 5 µl was electrophoresed. All fungal DNA samples were purified with the DNA Rapid Purification kit (BioDev, Beijing, People's Republic of China) after DNA extraction. (ii) Extraction of DNA from bacteria and peripheral blood from healthy humans was done by previously described procedures (25), with minor modifications.

DNA sequencing. The fragments containing the internal transcribed spacer regions 1 and 2 (ITS1 and ITS2) of the ribosomal DNA (rDNA) complex were amplified with previously published (8, 20) panfungal primers, primers ITS1, ITS2, ITS3, and ITS4 (ITS1, 5'-TCC GTA GGT GAA CCT GCG G-3'; ITS2, 5'-GCT GCG TTC TTC ATC GAT GC-3'; ITS3, 5'-GCA TCG ATG AAG AAC GCA GC-3'; ITS4, 5'-TCC TCC GCT TAT TGA TAT GC-3') (amplification conditions are described below). The PCR product was purified with a commercial purification kit and sequenced in both directions by using an ABI 373 sequencing machine with Applied Biosystems Taq DyeDeoxy Terminator Cycle-Sequencing Ready Reaction kits according to the manufacturer's instructions. The isolates sequenced included two isolates each of A. flavus, A. nidulans, A. niger, and A. terrus and six isolates of A. fumigatus (Table 1). Another isolate of A. fumigatus (isolate 00582) was also sequenced to verify the specificity of the nested PCR with primers Asp1 and AFUM1 (Asp1, 5'-CGG CCC TTA AAT AGC CCG GTC-3'; AFUM1, 5'-TTA CGA TAA TCA ACT CAG ACT GCA TA-3').

PCR amplification. All PCR were the same for each template and set of primers used. Each of the reaction mixtures contained 2.5 µl of 10× PCR buffer (100 mM Tris-HCl [pH 9.0] at 25°C, 15 mM MgCl₂, 500 mM KCl, 1.0% Triton X-100), 0.5 U of Taq DNA polymerase (Promega), 1 µl of deoxynucleoside triphosphates (dATP, dCTP, dGTP, and dTTP [10 mM each]; Boehringer Mannheim GmbH, Mannheim, Germany), 20 pmol of each primer, and 1 µl of sample DNA. Ultrapure sterile water was added to a final volume of 25 µl. Primers ITS1, ITS2, ITS3, and ITS4 were used for our sequencing amplifications. The DNA samples used for the nested PCR are indicated in Table 1. When doing the nested PCR, the specific primers used were Asp5 and AFUM2 (Asp5, 5'-GAT AAC GAA CGA GAC CTC GG-3'; AFUM2, 5'-ACC TTA GAA AAA TAA AGT TGG GTG-3') in the first amplification. For the second step of the nested PCR, the products of the first step (1 µl) were used as a template, and primers Asp1 and AFUM1 were used as the specific primers. PCR was performed in a GeneAmp PCR system 9600 instrument (Perkin-Elmer Applied Biosystems, Foster City, Calif.) at 95°C for 5 min for denaturation, 95°C for 30 s for denaturation, 58°C for 30 s for annealing, and 72°C for 1 min for primer extension for 30 cycles, with 5 min of extension at 72°C used for the final cycle. A more rapid scheme may also be used for the identification protocol, in which the reaction conditions are the same as those described above, except that the denaturation and the annealing-extension temperatures are 96 and 70°C, respectively, and that the times for the first and second step PCR are 5 and 10 s, respectively.

Multiple sequence alignments and primer design. Multiple sequence alignments were performed with the Pileup and Pretty programs from the Multiple Sequence Analysis program group (provided in Web ANGIS, the 3rd version of ANGIS [Australian National Genomic Information Service]). Specific primers were designed according to the sequence alignment and reference data (see Fig. 1).

Agarose gel electrophoresis. PCR products were analyzed by electrophoresis on 2% (wt/vol) agarose gels stained with 0.5 µg of ethidium bromide per ml. Volumes of 10 µl of PCR product and 2 µl of Blue/Orange 6× loading dye were loaded in each lane. Electrophoretic conditions were 100 V for 45 min in 0.5× TAE buffer (Tris-acetate, EDTA electrophoresis buffer [50× concentrated stock solution which includes, per liter, 242 g of Tris base, 57.1 ml of glacial acetic acid, and 100 ml of 0.5 M EDTA; pH 8.0]). Markers are also run in parallel to approximate the sizes of the PCR products.

Specificities of the primers. The specificities of the primers were tested with the organisms indicated in Table 1. Sample DNA was amplified with the appropriate primers to ensure that they were free of PCR inhibitors. A fragment of about 996 bp can be amplified from the positive sample by a single PCR with the outer specific primers, and a fragment of 643 bp can be amplified by the nested PCR with the inner specific primers. A positive amplicon from A. fumigatus 00582 of the expected size (about 643 bp) from a gel was excised, purified, and sequenced as described above for verification.

Sensitivity of the nested PCR. The concentration of DNA from two A. fumigatus isolates (isolates 00308 and 00586) was determined with a UV spectrophotometer (UV 2100; UV-VIS Recording Spectrophotometer; Shimadzu Co., Kyoto, Japan) by using 10-fold serial dilutions in TE buffer. The sensitivity of the nested PCR was estimated at the maximal dilution titer by using the A. fumigatus-specific primers.

Isolate identification study. Eight isolates of various Aspergillus species previously identified by morphology methods were selected, inoculated onto Sabouraud agar (Oxoid, Hampshire, United Kingdom), and incubated at 35°C for 24 h. Four A. fumigatus isolates, three A. flavus isolates, and one A. niger isolate were used. The samples were coded and presented for processing. A 3-mm² section (approximately) of mycelium was used for DNA extraction and amplification. The PCR products were analyzed by agarose gel electrophoresis.

Precautions against contamination and controls. The universal precautions suggested by Kwok and Higuchi (15) were used to eliminate possible contamination of samples. Cross-contamination by aerosols of spores was reduced by using BSC (Bio-Clean Bench; Sanyo, Tokyo, Japan) or by physical separation of laboratory areas when we were culturing organisms, isolating sample DNA, preparing PCR mixtures, and analyzing PCR products and by using disposable pipettes. Other precautions included autoclaving of all materials following their use. The same amplification mixture without template was used as a negative control, and at the same time, positive controls were provided. The fungal DNAs were detected with appropriate primers to verify the identities of the isolates in the samples and to exclude the existence of PCR inhibitors.

Nucleotide sequence accession numbers. The sequence (ITSI-5.8S-ITS2 region) acession numbers for the isolates (in parentheses) are as follows: AF07890 (00593), AF07891 (00594), AF07892 (00595), AF109329 (00572), AF19330 (00574), AF109328 (00624), AF07894 (00599), AF07893 (00600), AF07899 (00356), AF07898 (00643), AF07897 (00632), AF07896 (00630), AF07895 (00613), and AF109327 (00612). The accession numbers for the other sequences obtained from GenBank for use with primers designed from small-subunit rRNA and ITS1-5.8S-ITS2 (in parentheses) are as follows: A. fumigatus, AB008401, M55626, and M60301 (AF07889); A. flavus, X78537 and D63696 (AB008414); A. nidulans X78539, AB008403, and U77377 (U93686); A. terreus, X78540 and AB008409 (AJ001333); A. niger, X78538 and D63697 (U93685); A. parasiticus, D63699 (AB008418 and AF027862); Candida albicans, AJ005123, M60302 and X53497 (L76774, X71088, and L47111).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Multiple sequence alignments and primers. Multiple sequence alignments were developed from the sequence data and information from GenBank (Fig. 1). From the multiple sequence alignment results, we noticed that (i) the small-subunit, 5.8S, and large-subunit regions from all fungi analyzed are conserved, but small differences are observed; (ii) both ITS1 and ITS2 regions are quite diverse; and (iii) greater distinctions between species than between isolates within a species are observed (alignment data not show). The ITS1 region displayed more interspecies variation than the ITS2 region; approximately four separate variable regions were seen. These results are in agreement with those from other detailed studies (11a, 19). We also determined that if universal or genus-specific primers are used, the conserved regions (18S, 5.8S, and 28S) are the best targets, but if the specific primers are used, the ITS region is the choice as the target. Therefore, our specific primers were designed from the sequence analysis of the ITS region. The locations of the primers used in the present study are indicated in Fig. 1. Upon sequence analysis of ITS1, primer AFUM1 was generated; its sequence is identical to those of all seven A. fumigatus reference isolates tested but different from those of the other isolates tested. Primer AFUM2 was generated from the sequence of ITS2; likewise, its sequence was also identical to those in the A. fumigatus isolates tested but different from those of the other isolates tested (data not show). Primer Asp5 (forward) combined with primer AFUM2 (reverse) as outer primers specific for A. fumigatus amplified a 996-bp amplicon (approximately); primer Asp1 (forward) combined with primer AFUM1 (reverse) as inner primers specifically amplified an A. fumigatus target gene which resulted in a 643-bp amplicon (Fig. 2 and 3).

View larger version (58K): [in this window] [in a new window]   FIG. 1.   Multiple sequence alignment results for the consensus sequences of the five regions of the rDNA complex. Relevant primers are shown. SSU, small-subunit rRNA gene, partially displayed; ITS1, internal transcribed spacer region 1, completely displayed; 5.8SU, the 5.8S subunit rRNA gene, nucleotides 79 to 148 are omitted; ITS2, internal transcribed spacer region 2, completely displayed; LSU, large-subunit rRNA gene, partially displayed; the dashes in the sequences for the individual species names indicate identity to the consensus nucleotide; dashes in the consensus sequence indicate no consensus nucleotide; dots indicate deletion of the nucleotide at that point; shading indicates the positions of the primers.

View larger version (73K): [in this window] [in a new window]   FIG. 2.   Results of PCR for specificity of amplification. Lane 1, 100-bp ladder; lanes 2, 4, 6, and 8, A. fumigatus 00308, 00309, 00581, and 00586, respectively, lane 3, A. flavus 00327; lane 5, A. niger 00336; lane 7, A. terreus 00321; lane 9, A. nidulans 00359; lane 10, single amplification of A. fumigatus 00308; lane 11, negative control; lane 12, positive control (isolate 00321). The arrowhead to the left of lane 1 indicates a 500-bp marker.

View larger version (63K): [in this window] [in a new window]   FIG. 3.   Results of nested PCR for identification of isolates in clinical specimens. Lane 1, 100-bp ladder; lanes 2, 4, 6, 8, A. fumigatus 01200, 01340, 01887, and 01910, respectively; lane 3, A. flavus 01276; lane 5, A. flavus 01606; lane 7, A. flavus 01909; lane 9, A. niger 01633; lane 10, known positive isolate (isolate 001124); lane 11, known negative isolate (isolate 00337); lane 12, negative control. The arrowhead to the left of lane 1 indicates a 500-bp marker.

Precautions against contamination and control. All samples were tested with the appropriate primers to rule out false-negative reactions. Actually, in this study we found several samples that gave false-negative reactions. However, purification of the template DNA by using the DNA Rapid Purification kit overcame this problem except for one isolate of A. niger (isolate 00335). That isolate was eliminated from our study, because amplification was not observed after purification. No contamination occurred when we used a negative control.

Specificity of the nested PCR with the designed primers. Seventy-three samples were used in order to evaluate the specificity of the nested PCR with the specific primers designed in the present study. The presence of amplicons of about 643 bp on agarose gels as detected under UV light was considered a positive result by nested PCR (Fig. 2). The results are summarized in Table 1. The results indicate that the specific primers designed for detection of A. fumigatus amplified 16 isolates of A. fumigatus, 6 isolates of Neosartorya fischeri variants, 1 isolate of A. fumigatus var. neoellipticus, and 1 isolate of Aspergillus brevipes. Therefore, by using the A. fumigatus-specific primers designed in the present study, all isolates of A. fumigatus and closely related isolates were successfully amplified. These results are in agreement with the sequence data in GenBank, as determined with the Basic Blast program with the specific primers. The amplicon from A. fumigatus 00582 has been sequenced and compared with A. fumigatus sequences deposited in GenBank. The results show that the amplification position and length are correct. Using the sequenced amplicon, we found that our sequence also aligned with the sequences of A. fumigatus isolates deposited in GenBank. At the same time, no false-positive reactions were found. Interestingly, A. beijingensis (17) can be amplified with the outer primers but not with the inner primers. These results indicate that this new isolate is distinct at the molecular level from A. fumigatus. Cross-reactions with other species of fungi have not been observed.

Sensitivity of the nested PCR. A total of 10 µl of each of the PCR products was loaded onto agarose gels for electrophoresis. The visible bands in the nested PCR were seen at DNA dilutions of up to 1:13 to 1:14. This corresponds to 10 to 100 ag of sample DNA. The nested PCR is therefore more sensitive than other methods, such as antigen detection or Southern blotting.

Identification of clinical isolates. To determine the utility of the nested PCR with primers designed for the accurate identification of A. fumigatus, a blind test was carried out. Eight clinical isolates morphologically confirmed to be A. fumigatus were tested. Following incubation of the culture plates for 24 h at 35°C and amplification of the isolates by using specific primers, the products were analyzed by agarose gel electrophoresis. By this method, all A. fumigatus cultures were identified correctly to the species level (Fig. 3), and all identifications were made in less than 12 h after receipt of the culture.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Fungi from patient samples need to be correctly identified to the species level. This information often influences the type, dosage, and duration of antifungal therapy and also provides epidemiological surveillance data (16). A variety of different PCR strategies have been tried in recent years. Most of the published data used 18S rDNA or 28S rDNA as target DNA but rarely used the ITS regions. However, the resolving power (specificity) is low because of the conservation of these sequences, so that Aspergillus can be detected only at the genus level. Further separation requires an additional, labor-intensive step that is also time-consuming (6, 7, 11-13, 20).

Turenne et al. (27) adopted differences in the length of the ITS region for the identification of medically important fungi. However, the amplicon lengths were not specific enough to distinguish species, especially when larger samples were used. Although sequencing is quite accurate, its labor-intensive nature and expense restrict its use in the clinical laboratory (11a, 27).

The data published thus far indicate that the ITS region is sufficiently heterogeneous to differentiate species. Therefore, the ITS region seems to be an excellent target for the determination of species of fungi (11a, 24). Because the primers determine the specificity and sensitivity of PCR, they were extensively considered in the present study. The position and relationship of the primers to the fungal isolates are shown in Fig. 1. By use of the specific primers that target the ITS region and nested PCR, an easy, cost-effective assay that has a high sensitivity is achieved. The whole procedure can be finished within 4 h if the rapid scheme is used for amplification.

Besides the target gene and primers, the DNA extraction method should also be considered carefully. Several papers have detailed progress in improving methods of DNA extraction and also the problems of PCR amplification of fungi (22). Bougnoux et al. (1) reported on an inhibitor of PCR that could not be inactivated. In our study, several false-negative reactions occurred, but most of them could be resolved after purification of sample DNA with glass milk. One sample of A. niger was eliminated from the analysis because of a negative result in the control test. The presence of a PCR inhibitor in the first step of the PCR in the present study may also have been one of the causes of the negative results. Therefore, a good method of extraction of DNA for PCR amplification is an important step in the identification scheme. Progress in this field may lead to breakthroughs in the diagnosis of infections caused by medically important fungi or the identification of the medically important fungi.

There are many criticisms of PCR for the detection of fungal DNA, especially when panfungal primers or nested PCR is used, since molds are so prevalent in the environment. In a collaborative study among five European centers, the frequency and risk of contamination due to airborne spore inoculation or carryover contamination in fungal PCRs were analyzed. The investigators concluded that the risk of contamination is no higher in any fungal PCR assays than in other diagnostic PCR-based assays if general precautions are taken (18). In the present study, the results obtained with the negative controls demonstrate that contamination did not occur. Therefore, the incidence of contamination can be avoided if stringent methods are adopted. For the application of PCR-based strategies to the diagnostic laboratory, we strongly recommend the procedures developed by Bretagne et al. (4).

Although we have identified A. fumigatus and related species, there are still some questions that need to be answered. For example, first, is it necessary to identify A. fumigatus to the species level? In this study we could not discriminate A. fumigatus from N. fischeri, a related telemorph species of A. fumigatus section fumigati and A. brevipes. They are taxonomically included in A. fumigatus section fumigati (9, 16, 28). Our results support the belief that they are probably derived from a common ancestor of A. fumigatus. At the same time, they also demonstrate the problems associated with the traditional morphology-based nomenclature system or whether the target region that we chose for identification is optimal (9, 26) for the subgrouping of A. fumigatus section fumigati. Second, most clinical observations identify infectious agents at the genus level and samples are studied on a smaller scale with primers or probes. If a larger number of isolates was used, the numbers of samples with false-negative or false-positive results may increase due to the great diversity of fungi. It has been reported that more than 80% of fungal isolates can be isolated only once when tested at the molecular level (5, 28). Some papers have reported that some cross-reactions exist when large number of samples are tested (6). Therefore, the great diversity of fungi may be the most significant drawback for the development of a PCR-based diagnostic method. It seems that the problem is not resolvable at present. Third, as we noticed from our data and data presented elsewhere, the differences in both rDNA and mitochondrion genes are sufficient for discrimination of fungi at the genus level but are not enough to discriminate strains within a species. This problem reflects the need for more sequence information or even whole-genome sequence analysis. In subsequent years, more work will be needed to correlate genotypes with the morphologies or physiologies of fungi. These results will provide important information for use in clinical diagnosis.

In conclusion, specific primers with a nested PCR successfully identified A. fumigatus section fumigati. The assay was quick, sensitive, and economical. Although there are still problems that need to be addressed before it can be used in clinical laboratories, we consider this method to be very useful for the identification of fungi.