Abstract
Rapid detection and differentiation of Aspergillus and Mucorales species in fungal rhinosinusitis diagnosis are desirable, since the clinical management and prognosis associated with the two taxa are fundamentally different. We describe an assay based on a combination of broad-range PCR amplification and reverse line blot hybridization (PCR/RLB) to detect and differentiate the pathogens causing fungal rhinosinusitis, which include five Aspergillus species (A. fumigatus, A. flavus, A. niger, A. terreus, and A. nidulans) and seven Mucorales species (Mucor heimalis, Mucor racemosus, Mucor cercinelloidea, Rhizopus arrhizus, Rhizopus microsporus, Rhizomucor pusillus, and Absidia corymbifera). The assay was validated with 98 well-characterized clinical isolates and 41 clinical tissue specimens. PCR/RLB showed high sensitivity and specificity, with 100% correct identifications of 98 clinical isolates and no cross-hybridization between the species-specific probes. Results for five control isolates, Candida albicans, Fusarium solani, Scedosporium apiospermum, Penicillium marneffei, and Exophiala verrucosa, were negative as judged by PCR/RLB. The analytical sensitivity of PCR/RLB was found to be 1.8 × 10−3 ng/μl by 10-fold serial dilution of Aspergillus genomic DNA. The assay identified 35 of 41 (85.4%) clinical specimens, exhibiting a higher sensitivity than fungal culture (22 of 41; 53.7%) and direct sequencing (18 of 41; 43.9%). PCR/RLB similarly showed high specificity, with correct identification 16 of 18 specimens detected by internal transcribed spacer (ITS) sequencing and 16 of 22 detected by fungal culture, but it also has the additional advantage of being able to detect mixed infection in a single clinical specimen. The PCR/RLB assay thus provides a rapid and reliable option for laboratory diagnosis of fungal rhinosinusitis.
INTRODUCTION
Rhinosinusitis is a common disease with multiple manifestations and can even cause death in acute invasive cases. Fungi are considered to be major etiological agents, of which members of the order Mucorales and genus Aspergillus (order Eurotiales) are the most pathogenic (5, 10, 13). The course, therapeutic treatment, and prognosis of fungal rhinosinusitis caused by Aspergillus and Mucorales species are radically different; therefore, early diagnosis and accurate identification of pathogenic fungal species are crucial for effective treatment and clinical decision-making (21). Currently, diagnosis of fungal sinusitis still depends on histopathological examination and culture from nasal biopsy, but conventional culture-based phenotypic identification techniques often include significant delays and can fail to yield growth in clinical samples (17). In a significant number of cases, fungal culture is negative and only formalin-fixed paraffin-embedded (PE) tissue specimens are available for diagnosis of fungal infection. However, rapid diagnosis of surgical tissues is urgently needed in acute invasive infection cases. In addition, histopathological observations of fungal shape and arrangement may not be sufficient for the accurate identification of fungal species if only a limited quantity of anamorphic fungal hyphae is present. Therefore, to improve the outcome for fungal rhinosinusitis patients, the rapid and accurate detection and identification of pathogenic fungal species are needed to allow early initiation of targeted therapy.
In this study, we developed an assay combining broad-range PCR amplification and reverse line blot hybridization (PCR/RLB) to detect and differentiate Aspergillus and Mucorales pathogens in tissue specimens. The fungal pathogens tested included five Aspergillus species (Aspergillus fumigatus, A. flavus, A. niger, A. terreus, and A. nidulans) and seven species from the order Mucorales (Mucor heimalis, Mucor racemosus, Mucor cercinelloidea, Rhizopus arrhizus, Rhizopus microsporus, Rhizomucor pusillus, and Absidia corymbifera). The PCR/RLB assay was validated using a panel of 98 well-characterized isolates and 41 clinical specimens, including 28 PE and 13 surgical tissues.
MATERIALS AND METHODS
Fungal strains used in this study.
All of the fungal strains were obtained from the Research Center for Medical Mycology, Beijing University (Table 1). A total of 98 isolates belonging to 12 fungal species were studied, and all species were represented by species-specific probes in the RLB assay (Table 1). Isolates comprised five Aspergillus species (A. fumigatus, A. flavus, A. niger, A. terreus, and A. nidulans; 57 strains) and seven Mucorales species (M. heimalis, M. racemosus, M. cercinelloidea, Rhizopus arrhizus, Rhizopus microsporus, Rhizomucor pusillus, and Absidia corymbifera; 41 strains). Five isolates, belonging to Candida albicans, Fusarium solani, Scedosporium apiospermum, Penicillium marneffei, and Exophiala verrucosa, were used as negative controls.
Table 1.
Ninety-eight well-characterized fungal clinical isolates identified by the RLB assay
| No. of isolates testeda | Conventional identificationb | Identification by: |
|
|---|---|---|---|
| ITS sequence analysisc | RLB assay | ||
| 15 | A. fumigatus | A. fumigatus | A. fumigatus |
| 12 | A. flavus | A. flavus | A. flavus |
| 11 | A. niger | A. niger | A. niger |
| 15 | A. terreus | A. terreus | A. terreus |
| 4 | A. nidulans | A. nidulans | A. nidulans |
| 11 | M. heimalis | M. heimalis | M. heimalis |
| 4 | M. racemosus | M. racemosus | M. racemosus |
| 5 | M. cercinelloidea | M. cercinelloidea | M. cercinelloidea |
| 5 | Rhizopus arrhizus | Rhizopus arrhizus | Rhizopus arrhizus |
| 10 | Rhizopus microsporus | Rhizopus microsporus | Rhizopus microsporus |
| 2 | Rhizomucor pusillus | Rhizomucor pusillus | Rhizomucor pusillus |
| 4 | Absidia corymbifera | Absidia corymbifera | Absidia corymbifera |
| 1 | Candida albicans | Candida albicans | Noned |
| 1 | Fusarium solani | Fusarium solani | None |
| 1 | Scedosporium apiospermum | Scedosporium apiospermum | None |
| 1 | Penicillium marneffei | Penicillium marneffei | None |
| 1 | Exophiala verrucosa | Exophiala verrucosa | None |
These strains were provided by the Research Centre for Medical Mycology and Mycoses of Peking University.
These strains were identified by standard morphological methods.
Species identification by sequence analysis of the ITS region with universal primers ITS1 and ITS4.
None, five negative control strains belonging to Candida albicans, Fusarium sp., Scedosporium apiospermum, Penicillium sp., and Exophiala sp. were not detected by the RLB assay.
Clinical specimens.
PE tissues (n = 28) and surgical tissues (n = 13) of fungal rhinosinusitis specimens and 19 control samples without fungal infections were obtained from the Department of Pathology of Tongren Hospital, Capital Medical University, Beijing, China. The control samples were taken from patients suffering from chronic sinusitis (CS) (n = 14), inverted papilloma of maxillary sinus (IPMS) (n = 3), carcinoma of maxillary sinus (CMS) (n = 1), and rhinoscleroma (RS) (n = 1). A single pathologist (H. Liu) reviewed all of the hematoxylin and eosin, periodic acid-Schiff-stained slides to confirm the diagnosis and ensure the presence of representative tissue. In addition, the authors (R. Li) reviewed the results of fungal culture for all patients. The cases included fungal ball (n = 39) and invasive fungal rhinosinusitis (n = 2). All diagnoses in this research were established using current standard criteria and were confirmed by the observation of fungal hyphae in the histopathology examination (13). The surgical tissue was sampled by the excision of a standard, representative region of the tissue with a size of 5 mm by 5 mm by 5 mm. Surgical specimens for PCR analysis were immediately snap-frozen in liquid nitrogen and stored at −80°C until use. The surgical specimens for fungal culture were cut into pieces with a diameter of 3 to 5 mm and cultured on Saboraud dextrose agar plates at 28°C immediately after the biopsy.
DNA extraction.
For all fungal isolates, DNA extraction was performed by the benzyl chloride extraction method as previously described (25). DNA extraction was performed in a class II laminar flow cabinet. Briefly, 0.3 g of mycelium was harvested into a centrifuge tube. To each sample, 600 μl extraction buffer (100 mM Tris-HCl, pH 9.0, 40 mM EDTA), 50 μl 10% SDS, and 300 μl benzyl chloride were added. The tube was vortexed and incubated at 55°C for 1 h with shaking or repeated vortexing at 10-min intervals to keep the two phases thoroughly mixed. Then, 50 μl of 3 M sodium acetate (NaOAc), pH 5.2, was added, followed by incubation on ice for 20 min. After centrifugation at 12,000 rpm for 10 min, the supernatant was collected, mixed with an equal volume of phenol-chloroform, and centrifuged again at 12,000 rpm for 10 min. Then, the supernatant was collected and the DNA was precipitated with isopropanol. The pellet was dissolved in Tris-EDTA (TE) buffer and stored at −20°C until use in the experiments. For the clinical specimens, the QIAamp DNA FFPE tissue kit (Qiagen, Germany) and the QIAamp DNA minikit (Qiagen, Germany) were used to extract DNA from the PE and surgical tissues, respectively. Control of contamination was performed as described before, e.g., specimen manipulations and DNA extractions were performed in a class II laminar flow cabinet (16, 23).
Oligonucleotide design.
Relevant fungal DNA sequences spanning the fungal ribosomal DNA gene complex (ITS1, 5.8S rRNA, and ITS2) were accessed from GenBank and compared using the DNAMAN software program (Lynnon Biosoft). Three pairs of 5′-end biotin-labeled primers (Sangon, China), based on the universal fungal primer pairs ITS1 and ITS4, ITS1 and ITS2, and ITS3 and ITS4 (7, 16, 23), were used for the amplification of the ITS, ITS1, and/or ITS2 region (Table 2). To allow optimal hybridization under the same conditions, probes were also designed to have similar physical characteristics: lengths between 20 and 28 bp and melting temperatures between 54.48°C and 63.68°C (Table 2). Oligonucleotide probes for Aspergillus species were used as described previously or modified from these previously described probes (16, 23). Mucorales species-specific probes were designed by the authors, targeting ITS1 or ITS2 sequences (Table 2). All oligonucleotide probes included a 5′-amine group to facilitate covalent linkage to the nylon membrane and allow membranes to be stripped and reused repeatedly (Sangon, China).
Table 2.
Oligonucleotide primers and probes used in this study
| Primer or probea | Target | Length (bp) | Tm (°C)b | GenBank accession no. | Sequence (5′ to 3′) |
|---|---|---|---|---|---|
| ITS1bc | 18S rRNA universal fungal 5′ primer | 19 | 61.88 | AF455531 | 19TCCGTAGGTGAACCTGCGG37 |
| ITS2bc | 5.8S rRNA universal fungal 3′ primer | 20 | 57.80 | AF455531 | 23GCTRCGYTCTTCATCGATGC207 |
| ITS3bc | 5.8S rRNA universal fungal 5′ primer | 20 | 55.75 | AF455531 | GCATCGATGAAGARCGYAGC |
| ITS4bc | 28S rRNA universal fungal 3′ primer | 20 | 55.75 | L28817 | TCCTCCGCTTATTGATATGC |
| AFUMpc | ITS2 region of A. fumigatus | 23 | 56.60 | AY660923 | AGCCGACACCCAACTTTATTTTT |
| AFLUpc | ITS2 region of A. flavus | 22 | 54.48 | AF454111 | ACGCAAATCAATCTTTTTCCAG |
| ANIGpc | ITS2 region of A. niger | 20 | 57.80 | AF454120 | GCCGACGTTTTCCAACCATT |
| ATERpc | ITS2 region of A. terreus | 27 | 55.93 | AF454134 | GCATTTATTTGCAACTTGTTTTTTTCC |
| ANIDpc | ITS2 region of A. nidulans | 23 | 58.39 | AY660924 | GGC GTC TCC AAC CTT ATT TTT CT |
| MHEIp | ITS2 region of M. heimalis | 28 | 62.01 | EU326196 | TTAGGCCTGAACTATTGTTCTTTCTGCC |
| MRACp | ITS2 region of M. racemosus | 28 | 62.01 | AY213662 | CCTCTCGATCTGTATAGATCTTGAAACC |
| MCIRp | ITS2 region of M. cercinelloidea | 26 | 57.26 | EU484243 | ATACTGAGAGTCTCTTGATCTATTCT |
| RARRp | ITS1 region of R. arrhizus | 23 | 61.95 | EU337012 | CGCCTTACCTTAGGGTTTCCTCT |
| RMICp | ITS1 region of R. microsporus | 24 | 63.68 | FJ810505 | CTCTGGCGATGAAGGTCGTAACTG |
| RPUSp | ITS1 region of R. pusillus | 24 | 63.68 | EF136361 | CTCTATTGGTGAGCCGCGATTCTC |
| ACORp | ITS1 region of A. corymbifera | 24 | 58.55 | DQ118984 | TTGGCCTAACTTAAGGTCTTCTCT |
Suffixes: “b” indicates 5′ biotin-labeled primers; “p” indicates 5′ hexylamine-labeled probes.
The Tm values for oligonucleotide primers or probes were provided by the company that synthesized the oligonucleotides (Sangon).
PCR of fungal isolates.
PCR using the primer pair ITS1 and ITS4 was performed to amplify the ITS1, 5.8S rRNA, and ITS2 regions. PCR assays were performed with 25-μl reaction mixtures containing 0.5 U TaKaRa Taq polymerase, 2.5 μl PCR buffer (TaKaRa, Japan), 125 μM (each) dATP, dCTP, dGTP, and dTTP (TaKaRa, Japan), 1 μM (each) forward and reverse primers, and 2 μl of DNA template. Amplification was performed on a Mastercycler gradient thermocycler (Eppendorf, Germany). The thermal cycling conditions were 95°C for 5 min, followed by 30 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 60 s, with a final extension step at 72°C for 10 min. For the clinical specimens, the primer pair ITS1 and ITS2 was used to amplify the ITS1 region, and the pair ITS3 and ITS4 was used to amplify the ITS2 region. These reactions were carried out with the same PCR mixture described above, but the thermal cycling conditions were modified as follows: 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 30 s, with a final extension step at 72°C for 10 min.
RLB hybridization assay.
The RLB assay was performed as previously described (6, 24). We labeled a pair of probes on the membrane for a single species; therefore, the positions of positive samples on the grid will appear against the background as two dark signals (6). The same membrane could be reused on at least nine occasions following the stripping of bound PCR products without a loss of signal. The melting temperatures of the probes ranged between 60 and 80°C (Table 2), and the optimal hybridization temperature was determined to be 60°C. In brief, the membrane was washed at 60°C with prewarmed (to 60°C) 2× SSPE (1× SSPE is 0.18 M NaCl, 10 mM NaH2PO4, and 1 mM EDTA [pH 7.7])–0.5% SDS for 5 min. The amplified PCR products were hybridized with nylon membrane-bound probes at 60°C for 1 h (lanes 1 and 2 and 44 to 45 were loaded with 2× SSPE–0.1% SDS without PCR products, as negative controls in each reaction). The membrane was washed twice (10 min each time) at 60°C with prewarmed (to 60°C) 2× SSPE–0.5% SDS and incubated in 15 ml of streptavidin-peroxidase conjugate (Roche Diagnostics) diluted 1:4,000 in 2× SSPE–0.5% SDS for 60 min at 42°C. The membrane was further washed with 2× SSPE–0.5% SDS at 42°C and then at 25°C. If present, bound PCR products were detected by chemiluminescence using electrochemiluminescence detection liquid (Roche) and visualized by exposure for 7 min to X-ray film (Hyperfilm; Amersham). The membrane was washed twice with 250 ml 1% SDS for 30 min at 80°C, with rocking, to remove the PCR products bound on the membrane and then with 20 mM EDTA for 15 min at room temperature. The membrane was sealed in a plastic bag with approximately 10 ml of 20 mM EDTA to avoid dehydration and was stored at 4°C until reuse.
Sequencing of amplification products.
The amplification products from pure fungal cultures and from tissue specimens that yielded a single amplification product were purified with the QIAquick kit according to the manufacturer's protocol (Qiagen, Germany) and sequenced directly with a cycle sequence protocol (Sangon, China) using primer ITS4 (Table 1). Sequence similarities were assessed with a search for homology to GenBank sequences using the BLAST search program. Sequence similarities greater than 98% over a range of at least 75% of the ITS region were required for species-level identification.
Sensitivity and specificity of RLB.
To evaluate the sensitivity and specificity of the RLB assay, the results were compared with those obtained using culture-based identification and ITS sequencing. To determine the analytical sensitivity of RLB, we used 10-fold serial dilutions to obtain a gradient of concentrations of A. fumigatus genomic DNA that was subsequently used to perform the PCR/RLB assay. The starting concentration of 180 ng/μl was determined on the spectrophotometer.
RESULTS
Detection and identification of fungal strains by RLB assay.
A total of 98 well-characterized fungal strains, including 12 common Aspergillus or Mucorales species linked to fungal rhinosinusitis, were tested by the RLB assay (Table 1). These resulted in positive dots, as expected, indicating hybridization with species-specific probes but without cross-reactions between different probes (Fig. 1). To assess the specificity of the assay, we compared the results of PCR/RLB for all 98 studied isolates with conventional morphology-based identification and ITS sequence analysis. The RLB assay correctly identified 98 isolates and showed 100% correlation with the other two methods (Table 1). No signal was detected by RLB assay for the five control strains, Candida albicans, Fusarium solani, Scedosporium apiospermum, Penicillium marneffei, and Exophiala verrucosa.
Fig. 1.
Results of multiplex PCR-based reverse line blot hybridization assay (mPCR/RLB) hybridization for 12 strains. Lanes 1 to 12 represent the 12 strains, in the following order: A. fumigatus, A. flavus, A. niger, A. terreus, A. nidulans, M. heimalis, M. racemosus, M. cercinelloidea, R. arrhizus, R. microsporus, R. pusillus, and A. corymbifera; the 12 labeled probes are in the same order as in Table 2. Duplicate rows for the same probe represent different probe concentrations (0.2 and 0.4 pM). PCR products for the 12 strains hybridized with the species-specific probes but did not hybridize with any of the other probes.
Detection of fungal DNA in PE tissue and surgical tissue specimens by RLB assay.
The feasibility of the PCR/RLB assay was studied with PE tissue (n = 28) and surgical tissue specimens (n = 13), including 39 cases of clinical diagnosis as fungal ball (FB), one case of chronic invasive fungal rhinosinusitis (CIFRS), and one case of acute invasive fungal rhinosinusitis (AIFRS) from patients with fungal infection confirmed by the presence of fungal hyphae in pathological samples (Fig. 2). The amplification products were sequenced to identify the detected species. Table 3 summarizes the data for all patients.
Fig. 2.
RLB results from clinical samples. The labeled probes are in the same order as in Fig. 1. Lanes 1 and 2 and 44 and 45 are blank (loaded with 2× SSPE–0.1% SDS without PCR products, as negative controls). Lanes 3 to 43 are described in Table 3 (see footnote a).
Table 3.
Detection of fungal DNA in tissue specimensa
| Patient no. | Specimen type | Clinical diagnosis | Identification by: |
|||
|---|---|---|---|---|---|---|
| Culture | Histopathology | ITS1/ITS2 sequencing | RLB assayb | |||
| 1 | PE | FB | A. niger | Fungal hyphae | Nonec | A. niger |
| 2 | PE | FB | A. flavus | Fungal hyphae | A. flavus | A. flavus |
| 3 | PE | FB | Scedosporium apiospermum | Fungal hyphae | Scedosporium apiospermum | None |
| 4 | PE | FB | A. flavus | Fungal hyphae | A. flavus | A. flavus |
| 5 | PE | FB | A. nidulans | Fungal hyphae | None | A. nidulans |
| 6 | PE | CIFRS | A. flavus | Fungal hyphae | None | A. flavus + A. nigerd |
| 7 | PE | FB | Scedosporium apiospermum | Fungal hyphae | None | None |
| 8 | PE | FB | A. fumigatus | Fungal hyphae | A. fumigatus | A. fumigatus |
| 9 | PE | FB | Penicillium spp. | Fungal hyphae | None | None |
| 10 | PE | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 11 | PE | FB | None | Fungal hyphae | None | A. niger |
| 12 | PE | FB | Fusarium spp. | Fungal hyphae | None | None |
| 13 | PE | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 14 | PE | AIFRS | Rhizopus spp. | Fungal hyphae | None | Rhizopus arrhizus |
| 15 | PE | FB | Aspergillus spp. | Fungal hyphae | None | A. fumigatus + A. flavus |
| 16 | PE | FB | None | Fungal hyphae | None | A. fumigatus |
| 17 | PE | FB | None | Fungal hyphae | None | A. fumigatus |
| 18 | PE | FB | A. flavus | Fungal hyphae | A. flavus | A. flavus |
| 19 | PE | FB | Exophiala spp. | Fungal hyphae | None | None |
| 20 | PE | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 21 | PE | FB | None | Fungal hyphae | A. niger | A. flavus + A. niger |
| 22 | PE | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 23 | PE | FB | None | Fungal hyphae | None | A. fumigatus + A. flavus |
| 24 | PE | FB | None | Fungal hyphae | None | A. fumigatus |
| 25 | PE | FB | None | Fungal hyphae | None | A. flavus |
| 26 | PE | FB | None | Fungal hyphae | A. fumigatus | A. fumigatus |
| 27 | PE | FB | None | Fungal hyphae | None | A. fumigatus |
| 28 | PE | FB | None | Fungal hyphae | None | A. fumigatus |
| 29 | ST | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 30 | ST | FB | A. flavus | Fungal hyphae | A. flavus | A. flavus |
| 31 | ST | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 32 | ST | FB | None | Fungal hyphae | None | A. flavus |
| 33 | ST | FB | A. fumigatus | Fungal hyphae | A. fumigatus | A. fumigatus |
| 34 | ST | FB | Aspergillus spp. | Fungal hyphae | None | A. fumigatus |
| 35 | ST | FB | A. flavus | Fungal hyphae | A. flavus | A. flavus |
| 36 | ST | FB | None | Fungal hyphae | None | A. flavus |
| 37 | ST | FB | Aspergillus spp. | Fungal hyphae | A. flavus | A. flavus |
| 38 | ST | FB | Fusarium spp. | Fungal hyphae | None | None |
| 39 | ST | FB | None | Fungal hyphae | A. flavus | A. flavus |
| 40 | ST | FB | A. fumigatus | Fungal hyphae | None | A. fumigatus |
| 41 | ST | FB | A. fumigatus + A. flavus | Fungal hyphae | None | A. fumigatus |
| 42 | PE | CS | None | None | None | None |
| 43 | PE | CS | None | None | None | None |
| 44 | PE | CS | None | None | None | None |
| 45 | PE | CS | None | None | None | None |
| 46 | PE | IPMS | None | None | None | None |
| 47 | PE | IPMS | None | None | None | None |
| 48 | PE | IPMS | None | None | None | None |
| 49 | PE | CMS | None | None | None | None |
| 50 | PE | RS | None | None | None | None |
| 51 | ST | CS | Streptococcus aureus | None | None | None |
| 52 | ST | CS | Streptococcus aureus | None | None | None |
| 53 | ST | CS | Streptococcus aureus | None | None | None |
| 54 | PE | CS | Streptococcus aureus | None | None | None |
| 55 | PE | CS | Streptococcus aureus | None | None | None |
| 56 | ST | CS | Streptococcus pneumoniae | None | None | None |
| 57 | PE | CS | Streptococcus pneumoniae | None | None | None |
| 58 | ST | CS | Haemophilus influenzae | None | None | None |
| 59 | ST | CS | Streptococcus aureus + Streptococcus pneumoniae | None | None | None |
| 60 | ST | CS | Streptococcus aureus + Haemophilus influenzae | None | None | None |
Abbreviations: PE, paraffin-embedded tissue; FB, fungal ball; CIFRS, chronic invasive fungal rhinosinusitis; AIFRS, acute invasive fungal rhinosinusitis; ST, surgical tissue; CS, chronic sinusitis; IPMS, inverted papilloma of maxillary sinus; CMS, carcinoma of maxillary sinus; RS, rhinoscleroma.
The results of the RLB assay (no. 1 to no. 41) are shown in Fig. 2 (lane 3 to lane 43).
None, this sample could not be detected by tissue culture, ITS1/ITS2 sequencing, or RLB assay.
These samples included mixed infections, with more than one species detected.
Of the 41 specimens, 22 were culture positive. Eighteen of these were identified as A. flavus (n = 6; Table 3, specimens from patients no. 2, 4, 6, 18, 30, and 35), A. fumigatus (n = 3; Table 3, no. 8, 33, and 40), A. nidulans (n = 1; Table 3, no. 5), A. niger (n = 1; Table 3, no. 1), Rhizopus spp. (n = 1; Table 3, no. 14), Scedosporium apiospermum (n = 2; Table 3, no. 3 and 7), Fusarium spp. (n = 2; Table 3, no. 12 and 38), Penicillium spp. (n = 1; Table 3, no. 9), and Exophiala spp. (n = 1; Table 3, no. 19). No. 41 was a mixed infection, with both A. flavus and A. fumigatus (n = 1; Table 3), but was identified as A. fumigatus by RLB assay. No. 15 and 34 were morphologically identified as Aspergillus spp., but the RLB assay yielded more detail. No. 15 was identified as an A. fumigatus plus A. flavus mixed infection, and no. 34 was identified as A. flavus by RLB assay, but direct sequencing failed for these two samples.
Direct sequencing was possible for only 18 (43.9%) of 41 clinical specimens. Of these, 13 were identified as A. flavus (Table 3, no. 2, 4, 10, 13, 18, 20, 22, 29, 30, 31, 35, 37, and 39), three as A. fumigatus (Table 3, no. 8, 26, and 33), one as Scedosporium apiospermum (Table 3, no. 3), and one as A. niger (Table 3, no. 21). The results for 8 of the 18 samples (Table 3, No. 2, 3, 4, 8, 18, 30, 33, and 35) that could be sequenced were in agreement with the results of cultivation. Results for 16 of 18 samples were in agreement with the RLB identification, except no. 3, which was identified as Scedosporium apiospermum but was not detectable by RLB, and no. 21, which was identified as A. niger but was identified as A. flavus plus A. niger by RLB. Twenty-three of 41 specimens could not be directly sequenced, including 19 showing smudge amplification products by gel electrophoresis that, in spite of their lack of sequencing results, could still be detected by the RLB assay (Fig. 2). The remaining 4 of these 23 (17.4%) specimens yielded more than one amplification product, which was further proved to represent mixed infections; specifically, A. fumigatus plus A. flavus (Table 3, no. 15 and 23) and A. flavus plus A. niger (Table 3, no. 6 and 21), which were detected by the RLB assay. The mixed infections were verified by cloning prior to sequencing (data not shown).
The RLB assay showed a higher sensitivity than the fungal culture and direct sequencing methods, with organisms in 35 of 41 (85.4%) specimens detected and identified by the assay. For six of the 41 (14.6%) clinical specimens with a diagnosis of fungal rhinosinusitis for which organisms failed to be detected by RLB, subsequent identifications were Scedosporium apiospermum (no. 3 and 7), Penicillium spp. (no. 9), Fusarium spp. (no. 12 and 38), and Exophiala spp. (no. 19). The results obtained with these six specimens served to confirm the high specificity of the probes. In future investigations, probes corresponding to these four genera should be added to the system. For 10 specimens with a diagnosis of bacterial rhinosinusitis in which organisms failed to be detected by RLB, identifications were Streptococcus aureus (no. 51, 52, 53, 54, and 55), Streptococcus pneumoniae (no. 56 and 57), Haemophilus influenzae (no. 58), Streptococcus aureus plus Streptococcus pneumoniae (no. 59), and Streptococcus aureus plus Haemophilus influenzae (no. 60). For nine specimens with a diagnosis of chronic sinusitis (CS) (no. 42, 43, 44, and 45), inverted papilloma of the maxillary sinus (IPMS) (no. 46, 47, and 48), carcinoma of the maxillary sinus (CMS) (no. 49), and rhinoscleroma (RS) (no. 50) without any infection, organisms also failed to be detected by RLB.
Sensitivity and specificity of RLB.
The RLB assay was capable of detecting 101 to 102 conidia/ml in normal saline. This is similar to the sensitivity of RLB observed in other studies (16, 23). Further investigations using 10-fold serial dilutions of genomic DNA of A. fumigatus indicated that the minimum concentration that can be detected by the PCR/RLB assay is 1.8 × 10−3 ng/μl (data not shown).
DISCUSSION
Over the last decade, the morbidity and mortality of fungal rhinosinusitis have increased dramatically, with an average of over 140 cases per year requiring surgical intervention in our hospital between 2005 and 2009. Therefore, rapid detection and accurate identification of Aspergillus and Mucorales species is needed for early diagnosis and subsequently effective antifungal treatment.
To date, PCR-based molecular identification methods, including PCR (1, 4), seminested PCR (19), quantitative real-time PCR (3), multiplex PCR (12), PCR-restriction fragment length polymorphism (RFLP) (9), randomly amplified polymorphic DNA (RAPD) (20), and ITS sequence analysis (2, 22), have been used to identify Aspergillus and Mucorales species. However, these assays are not readily adaptable for use in the clinical lab due to disadvantages, including the following: (i) PCR, seminested PCR, and real-time PCR can identify only one species at a time, (ii) it is difficult to draw clear conclusions from the results of RFLP and RAPD, (iii) the number of primers for the Multiplex PCR system is limited, and (iv) sequence analysis is time-consuming and often fails to identify samples with mixed infections and samples of poor quality or that are in small quantities. In the present study, detection by direct sequencing was successful for only 43.9% of the clinical specimens.
By comparison, RLB technology overcomes these shortcomings by allowing the simultaneous detection and identification of clinically important pathogenic microorganisms, including bacteria, viruses, parasites, and fungi (6, 16, 23). Playford and Zeng et al. successfully utilized RLB to detect and identify Candida, Aspergillus, and Cryptococcus species, but unfortunately, Mucorales, an emerging, fatal pathogen, particularly in immunocompromised patients, was not included in their studies (11, 14, 16, 18, 23). The frequency of Aspergillus and Mucorales infections is increasing because of the increase in the life span of patients with neoplasms and the widespread use of immunosuppressive therapy. Invasive aspergillosis and mucormycosis are fungal diseases that affect multiple organs, including, among others, the sinuses, brain, and lungs. These diseases result in 50% or greater mortality despite first-line therapy.
The present study describes a panfungal PCR/RLB method for the identification of common pathogens involved in fungal rhinosinusitis, including five Aspergillus species and seven Mucorales species. The PCR/RLB assay demonstrates a high level of sensitivity and specificity. The validity of the PCR/RLB assay was confirmed by analysis using 98 well-characterized isolates, including comparison with conventional morphological identification and ITS sequence analysis. For all 98 isolates tested, identical results were obtained by the three methods, demonstrating the reliability of the RLB assay in accurately identifying culture isolates. No cross-hybridization was found among the 98 strains tested. In addition, five control strains, six clinical specimens infected by other fungal species, and three clinical specimens infected by bacterial species were not detected in this study, supporting the specificity of the probes.
The diagnostic applicability of broad-range PCR and RLB has been confirmed by the identification of bacterial, viral, and fungal pathogens in clinical specimens (6, 16, 23). In the present study, the feasibility for clinical application of the assay was studied by testing 28 PE tissues and 13 biopsy tissues from 41 patients, of which 22 were positive by tissue culture. The RLB assay showed the highest sensitivity, with accurate identification of 35 of 41 (85.4%) clinical specimens, compared with 22 of 41 (53.7%) specimens for fungal culture and 18 of 41 (43.9%) specimens for direct sequencing (Table 3). PCR/RLB showed similar high specificities, with the correct identification 16 of 18 specimens detected by ITS sequencing and 16 of 22 specimens detected by fungal culture (Table 3). Eighteen specimens that could not be identified by ITS sequencing and 19 that resisted cultivation were identified by the RLB assay. Interestingly, 10 tissue samples (no. 11, 16, 17, 23, 24, 25, 27, 28, 32, and 36) that were identified as Aspergillus spp. by RLB but not by cultivation or sequencing were confirmed to be Aspergillus spp. by immunochemical staining with MUC5B antibody (15) (data not show). Ten tissue samples with bacterial infections and 9 tissue samples without infection were not identified by PCR/RLB, which confirmed the high specificity of the assay.
The platform also shows excellent potential for the development of additional fungal probes targeting other organisms, such as Scedosporium apiospermum, Penicillium spp., Fusarium spp., and Exophiala spp., which were not detected in the specimens by RLB assay in our study, since the number of probes in the system can be further increased to include all 43 species. An additional advantage is that mixed infections were correctly identified by RLB assay but were not detected by direct PCR and subsequent ITS sequencing in four specimens in our study.
Besides the detection and identification of fungi in rhinosinusitis patients, the RLB assay, with its ability to simultaneously detect and identify Aspergillus and Mucorales species in clinical specimens from different sources, provided results in ∼7 h, versus 3 days for direct sequencing and 1 to 4 weeks for culture-based identification. Therefore, it is suitable for use in most reference laboratories to perform simple, quick, and high-throughput assays for epidemiological and diagnostic purposes.
However, RLB technology is still an open system, which can yield amplicon contamination if stringent laboratory measures (e.g., DNA extraction and preparation of reagents in a biological safety cabinet, physically separate areas for DNA extraction, preparation of reagents and storage, PCR, post-PCR, and RLB) are not followed (6, 8, 16, 23).
In conclusion, these data revealed that the application of a fungal PCR/RLB assay has the advantage of accurate and simultaneous detection and identification of species of Aspergillus and Mucorales directly from surgical and PE tissue specimens from fungal rhinosinusitis patients. The assay represents a promising diagnostic option, with its higher rate of identification possibly leading to a more timely initiation of targeted therapy and better clinical outcomes.
ACKNOWLEDGMENTS
This work was supported in part by the youth fund of First Hospital, Beijing University, Community Projects of The Ministry of Health (200802026), and the National Science & Technology Major Project, funded by the Ministry of Science & Technology of China (2008zx 10004-002).
We thank Bioedit for English editing.
Zuotao Zhao and Lili Li made similar contributions to the work and should be regarded as co-first authors.
Conflict-of-interest disclosure: we have no competing financial interests to declare.
Footnotes
Published ahead of print on 16 February 2011.
REFERENCES
- 1. Bialek R., et al. 2005. PCR based identification and discrimination of agents of mucormycosis and aspergillosis in paraffin wax embedded tissue. J. Clin. Pathol. 58:1180–1184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Eyigor H., et al. 2008. Characterization of fungi in chronic rhinosinusitis using polymerase chain reaction and sequencing. Eur. Arch. Otorhinolaryngol. 265:651–655 [DOI] [PubMed] [Google Scholar]
- 3. Francesconi A., et al. 2008. Automated and manual methods of DNA extraction for Aspergillus fumigatus and Rhizopus oryzae analyzed by quantitative real-time PCR. J. Clin. Microbiol. 46:1978–1984 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hofman V., et al. 2010. Usefulness of molecular biology performed with formaldehyde-fixed paraffin embedded tissue for the diagnosis of combined pulmonary invasive mucormycosis and aspergillosis in an immunocompromised patient. Diagn. Pathol. 5:1–7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. International Rhinosinusitis Advisory Board 1997. Infectious rhinosinusitis in adults: classification, etiology and management. Ear Nose Throat J. 76:5–22 [PubMed] [Google Scholar]
- 6. Kong F., Gilbert G. L. 2006. Multiplex PCR-based reverse line blot hybridization assay (mPCR/RLB)—a practical epidemiological and diagnostic tool. Nat. Protoc. 1:2668–2680 [DOI] [PubMed] [Google Scholar]
- 7. Kumar M., Shukla P. K. 2005. Use of PCR targeting of internal transcribed spacer regions and single-stranded conformation polymorphism analysis of sequence variation in different regions of rRNA genes in fungi for rapid diagnosis of mycotic keratitis. J. Clin. Microbiol. 43:662–668 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Lau A., et al. 2007. Development and clinical application of a panfungal PCR assay to detect and identify fungal DNA in tissue specimens. J. Clin. Microbiol. 45:380–385 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Machouart M., et al. 2006. Genetic identification of the main opportunistic Mucorales by PCR-restriction fragment length polymorphism. J. Clin. Microbiol. 44:805–810 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Meltzer E. O., et al. 2004. Rhinosinusitis: establishing definitions for clinical research and patient care. J. Allergy Clin. Immunol. 114:S155–S212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mohindra S., Mohindra S., Gupta R., Bakshi J., Gupta S. K. 2007. Rhinocerebral mucormycosis: the disease spectrum in 27 patients. Mycoses 50:290–296 [DOI] [PubMed] [Google Scholar]
- 12. Nagao K., et al. 2005. Genetic identification and detection of human pathogenic Rhizopus species, a major mucormycosis agent, by multiplex PCR based on internal transcribed spacer region of rRNA gene. J. Dermatol. Sci. 39:23–31 [DOI] [PubMed] [Google Scholar]
- 13. Nichols L. 1997. Fungal sinusitis. N. Engl. J. Med. 337:1633–1634 [PubMed] [Google Scholar]
- 14. Nosari A., et al. 2000. Mucormycosis in hematologic malignancies: an emerging fungal infection. Haematologica 85:1068–1071 [PubMed] [Google Scholar]
- 15. Piao Y. S., Zhang Y., Yang X., He C. Y., Liu H. G. 2008. The use of MUC5B antibody in identifying the fungal type of fungal sinusitis. Hum. Pathol. 39:650–656 [DOI] [PubMed] [Google Scholar]
- 16. Playford E. G., et al. 2006. Simultaneous detection and identification of Candida, Aspergillus, and Cryptococcus species by reverse line blot hybridization. J. Clin. Microbiol. 44:876–880 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Reiss E., Morrison C. J. 1993. Nonculture methods for diagnosis of disseminated candidiasis. Clin. Microbiol. Rev. 6:311–323 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ribes J. A., Vanover-Sams C. L., Baker D. J. 2000. Zygomycetes in human disease. Clin. Microbiol. Rev. 13:236–301 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Rickerts V., et al. 2006. Diagnosis of invasive aspergillosis and mucormycosis in immunocompromised patients by seminested PCR assay of tissue samples. Eur. J. Clin. Microbiol. Infect. Dis. 25:8–13 [DOI] [PubMed] [Google Scholar]
- 20. Vastag M., Papp T., Kasza Z., Vágvölgyi C. 2000. Intraspecific variation in two species of Rhizomucor assessed by random amplified polymorphic DNA analysis. J. Basic Microbiol. 40:269–277 [DOI] [PubMed] [Google Scholar]
- 21. Vennewald I., Henker M., Klemm E., Seebacher C. 1999. Fungal colonization of the paranasal sinuses. Mycoses 42:33–36 [PubMed] [Google Scholar]
- 22. Woo P. C., et al. 2010. Internal transcribed spacer region sequence heterogeneity in Rhizopus microsporus: implications for molecular diagnosis in clinical microbiology laboratories. J. Clin. Microbiol. 48:208–214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Zeng X., et al. 2007. Reverse line blot hybridization assay for identification of medically important fungi from culture and clinical specimens. J. Clin. Microbiol. 45:2872–2880 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Zhao Z., Kong F., Gilbert G. L. 2006. Reverse line blot assay for direct identification of seven Streptococcus agalactiae major surface protein antigen genes. Clin. Vaccine Immunol. 13:145–149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Zhu H., Qu F., Zhu L. H. 1993. Isolation of genomic DNAs from plants, fungi and bacteria using benzyl chloride. Nucleic Acids Res. 21:5279–5280 [DOI] [PMC free article] [PubMed] [Google Scholar]