ABSTRACT
A Blastomyces dermatitidis nested PCR assay targeting the gene encoding the Wisconsin 1 (WI-1) adhesin was developed and compared with a nested PCR targeting the 18S rRNA gene (rDNA) of members of the family Onygenaceae. We examined 73 paraffin-embedded tissue samples obtained from nine dogs which died of blastomycosis and nine dogs which succumbed to lymphosarcoma according to autopsy findings; amplifiable canine DNA was extracted from 25 and 33 specimens from the two groups, respectively. The B. dermatitidis PCR amplified DNA from 8 of 13 tissue samples in which yeast cells were detected by microscopy. Sequencing revealed that all PCR products were homologous to the B. dermatitidis WI-1 adhesin gene. No PCR product was amplified from 12 microscopically negative biopsy specimens from dogs with blastomycosis or from 33 biopsy specimens from dogs with lymphosarcoma. The 18S rDNA PCR amplified DNA from 10 and 9 tissue samples taken from dogs which died of blastomycosis and lymphosarcoma, respectively. Only six products were identified as being identical to B. dermatitidis 18S rDNA; they were exclusively obtained from specimens positive by the B. dermatitidis nested PCR. For specificity testing, 20 human biopsy specimens proven to have histoplasmosis were examined, and a specific H. capsulatum product was amplified by the 18S rDNA PCR from all specimens, whereas no product was obtained from any of the 20 samples by the B. dermatitidis PCR assay. In conclusion, the PCR targeting a gene encoding the unique WI-1 adhesin is as sensitive as but more specific than the PCR targeting the 18S rDNA for detection of B. dermatitidis in canine tissue.
Blastomycosis is a systemic fungal infection that occurs mainly in dogs and humans. The organism that causes the infection, Blastomyces dermatitidis, grows in a mycelial form in the soil, and infection occurs from inhalation of the spores. The infection is most common in North America, but it also occurs in Africa. The North American cases occur mainly along the Mississippi and Ohio River basins and around the Great Lakes. In some areas of the state of Wisconsin, the mean annual incidence of blastomycosis in dogs may reach 1,420 per 100,000 population (1). The geographical distribution of cases of blastomycosis in humans is similar to that in dogs, but the rates in dogs are about 10 times those in humans (9).
Blastomycosis in dogs is a systemic disease with infection of the lungs, lymph nodes, eyes, skin, bone, testes, and brain (16). The agar gel immunodiffusion test has a 95% specificity for the detection of blastomycosis, but up to 30% of dogs with early infection do not have antibodies against B. dermatitidis (15). A radioimmunoassay with the Wisconsin 1 (WI-1) antigen identified antibodies in 92% of infected dogs (13). This study attempts to identify a role for PCR in the early detection of blastomycosis in dogs.
PCR assays that amplify sequences of fungal genes have been introduced successfully for the diagnosis of fungal infections (11, 12, 17, 18, 21, 22). rRNA genes (rDNAs) are often targeted in order to achieve high sensitivity because several gene copies are usually present within a single genome. Accordingly, we developed a sensitive 18S rDNA-based nested PCR assay to monitor murine histoplasmosis. The assay also amplifies DNA from the closely related species B. dermatitidis and Paracoccidioides brasiliensis (4, 5). From previous experience with paracoccidioidomycosis and histoplasmosis (3, 7), a distinctive gene of B. dermatitidis was used to develop a diagnostic nested PCR with a high specificity.
In order to evaluate this novel assay with canine specimens, paraffin-embedded tissue samples were examined. The use of this kind of specimen permits repeatable examination. The organisms seen on histopathology are characteristic of B. dermatitidis, but identification by culture is impossible due to formalin fixation. However, if specific genes are targeted, PCR might be used for further identification. The quality and amount of extractable DNA in formalin-fixed tissue may vary depending on the amount of time that the tissue remained in formalin (2, 19). A sensitive PCR targeting a canine gene is therefore necessary as a control for DNA extraction. This assessment is essential in judging the diagnostic value of the PCR assays for the detection of B. dermatitidis DNA in formalin-fixed canine tissue.
MATERIALS AND METHODS
The dogs used in this study were patients at the Veterinary Medical Teaching Hospital, University of Tennessee, Knoxville, from 1992 to 2000, which died due to either blastomycosis or lymphosarcoma. Case selection was based on a review of necropsy findings and the availability of multiple tissue samples in paraffin blocks to identify dogs that had proven blastomycosis. Some tissues from each dog had histologic evidence of B. dermatitidis organisms, and some tissues did not. Hematoxylin-eosin-stained slides of tissue specimens from all dogs, including those diagnosed with lymphosarcoma, were then reviewed for the presence or absence of B. dermatitidis organisms. In addition, duplicate slides with tissue samples from most dogs with blastomycosis that stained positive with Gomori methenamine silver for fungal organisms were also reviewed. None of the samples from dogs with lymphosarcoma had histologic evidence of fungal organisms. Small tissue samples were then cut out of the corresponding paraffin blocks for each tissue reviewed and were labeled according to the presence or absence of B. dermatitidis organisms. Histologic samples were marked as positive for B. dermatitidis when typical 7- to 15-μm, spherical, thick-walled yeasts were identified within the tissue. A total of 35 samples were taken from nine dogs (median, 4 samples per animal; range, 2 to 9 samples per animal) which succumbed to blastomycosis. Another 38 specimens were obtained from nine dogs (median, 4 samples per animal; range, 3 to 5 samples per animal) which died of lymphosarcoma. The organ distributions were similar for both groups: 8 lungs from each group; 5 and 13 lymph nodes from dogs with blastomycosis and lymphosarcoma, respectively; 3 and 4 spleens from dogs with blastomycosis and lymphosarcoma, respectively; 9 and 6 kidneys from dogs with blastomycosis and lymphosarcoma, respectively; and 5 and 7 livers from dogs with blastomycosis and lymphosarcoma, respectively. In addition samples from the brain, prostate gland, testis, a thoracic mass, and nasal cavity were obtained from dogs with blastomycosis. The paraffin-embedded tissue samples were blinded and sent for further testing to the Institute for Tropical Medicine, University Hospital Tübingen, Tübingen, Germany.
In order to evaluate specificity, biopsy specimens from humans with proven histoplasmosis from a former study were examined (7).
DNA extraction.One thousand microliters of xylene was added to one Eppendorf tube containing two 5-μm sections of a biopsy specimen. It was incubated on a shaker for 5 min at room temperature and was subsequently centrifuged at 10,000 × g for 2 min. The supernatant was removed, and 1,000 μl of absolute ethanol was added, followed by centrifugation at 10,000 × g for 3 min. After the supernatant was removed and the ethanol and centrifugation steps were repeated, the supernatant was removed and the samples were air dried. As the next step, 180 μl of ATL buffer from the QIAamp tissue kit (Qiagen, Hilden, Germany) and proteinase K (final concentration, 2 mg/ml; Qiagen) were added. After incubation at 55°C overnight, the samples were boiled for 5 min and then exposed to three cycles of freezing in liquid nitrogen for 1 min and boiling for 2 min to disrupt the fungal cells. After the DNA was cooled to room temperature, it was extracted by using the QIAamp tissue kit (Qiagen) based on binding of the DNA to silica columns, in accordance with the instructions of the manufacturer.
Design of primers for B. dermatitidis adhesin PCR.Outer primers blasto I (5′-AAG TGG CTG GGT AGT TAT ACG CTA C-3′) and blasto II (5′-TAG GTT GCT GAT TCC ATA AGT CAG G-3′) are complementary to nucleotide positions 1552 to 1576 and 1914 to 1890, respectively, of the gene coding for the immunodominant WI-1 adhesin (GenBank accession number U37772 , from the database of the National Center for Biotechnology Information, Washington, D.C.) and amplify a 363-bp sequence of this unique antigen. Inner primers blasto III (5′-TGA ATC TGC TTG GCA AAT GCC GTT G-3′) and blasto IV (5′-AGG CGC AGG AGA GGT AAA ATT GGC A-3′) are complementary to positions 1595 to1619 and 1860 to 1836, respectively, and amplify a specific 266-nucleotide sequence.
Design of primers for H. capsulatum-B. dermatitidis 18S rDNA PCR.A previously described H. capsulatum 18S rDNA PCR was used (5). Outer primers fungus I (5′-GTT AAA AAG CTC GTA GTT G-3′) and fungus II (5′-TCC CTA GTC GGC ATA GTT TA-3′) are complementary to a highly conserved region of the small-subunit rRNA gene of Histoplasma capsulatum (GenBank accession number X58572 ) and B. dermatitidis (GenBank accession number AF320010 ) and amplify a 429-bp sequence of several fungi pathogenic for humans. Inner primers histo I (5′-GCC GGA CCT TTC CTC CTG GGG AGC-3′) and histo II (5′-CAA GAA TTT CAC CTC TGA CAG CCG A-3′) are complementary to positions 643 to 666 and 873 to 849 of the 18S rDNA, respectively, and amplify a specific 231-nucleotide sequence. In contrast to H. capsulatum, a single mismatch with the B. dermatitidis 18S rDNA sequence occurs at position 663, where guanine is replaced by adenine.
PCR assays.The reaction mixture for the primary PCRs consisted of 10 μl of DNA extract in a total volume of 50 μl with final concentrations of 10 mM Tris-HCl (pH 8.3), 50 mM KCl, and 2.5 mM MgCl2 (10× Perkin-Elmer buffer II plus MgCl2 solution [Roche Molecular Systems, Branchburg, N.J.]); each primer of the outer primer sets (Roth, Karlsruhe, Germany) at a concentration of 1 μM; 1.5 U of AmpliTaq DNA polymerase (Roche); and each deoxynucleoside triphosphate (Promega, Madison, Wis.) at a concentration of 100 μM. The reaction mixture for the nested PCRs was identical to that for the primary PCRs, except that 1 μl of the first reaction mixture, each deoxynucleoside triphosphate at a concentration of 50 μM, and each primer of the inner primer sets at a concentration of 1 μM were used. The reaction mixtures with the outer primer sets were thermally cycled once at 94°C for 5 min; 35 times at 94°C for 30 s, 50°C for 30 s, and 72°C for 1 min; and then once at 72°C for 5 min. For the nested PCR products, the reaction mixtures were thermally cycled once at 94°C for 5 min, 30 times at 94°C for 30 s and 72°C for 1 min, and then once at 72°C for 5 min. The high melting temperatures of the inner primer sets allowed two-step nested PCRs with high stringency.
GAPDH PCR.In order to prove the presence of amplifiable DNA, a nested PCR with a target sequence within the human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene (GenBank accession number J04038.1 ) was carried out as described before (7). Outer primers gap 1 (5′-GAC AAC AGC CTC AAG ATC ATC-3′) and gap 2 (5′-GAC GGC AGG TCA GGT CCA CCA-3′) amplify a 610-nucleotide sequence of genes (positions 3816 to 4425) and a 441-nucleotide sequence of pseudogenes, respectively. Inner primers gap 3 (5′-AAT GCC TCC TGC ACC ACC-3′) and gap 4 (5′-ATG CCA GTG AGC TTC CCG-3′) amplify 325-bp products (positions 3932 to 4372) and 248-bp products, respectively.
The reaction mixture was identical to that for the PCR assays described above, except that all primers were used at a concentration of 0.3 μM and 2 μl of the first-round reaction product was used for the nested PCR. The reaction mixtures with the outer primer sets were thermally cycled once at 94°C for 5 min; 35 times at 94°C for 30 s, 56°C for 30 s, and 72°C for 45 s; and then once at 72°C for 5 min. The reaction with the inner primer set was identical, except that 40 cycles were carried out.
All PCRs were run in a Primus PCR thermocycler Tc 9600 instrument (MWG Biotech, Ebersberg, Germany). The nested PCR products were analyzed by electrophoresis on 1.5% agarose gels, stained with ethidium bromide, and visualized on a UV transilluminator.
Cloning.The amplicon of the primary PCR with DNA extracted from a laboratory strain of B. dermatitidis (V-9636; M. G. Rinaldi, Fungus Testing Laboratory, University of Texas Health Science Center at San Antonio) was purified with Qiagen spin columns. It was inserted into the pCR 2.1-TOPO cloning vector by using the Original TA Cloning kit (Invitrogen, Groningen, The Netherlands) in accordance with the instructions of the manufacturer. After culturing and harvesting of the bacteria, the plasmid DNA was purified by using the Qiagen Plasmid Maxi kit, consisting of alkaline lysis of bacteria, separation, and binding of plasmid DNA to an anion-exchange resin; wash steps; and final elution. The DNA concentration was measured by determination of the absorption at 260 nm. Serial dilutions were used in order to determine the detection limit of the nested PCR assay. The amplified product was sequenced to prove homology to the sequence of the B. dermatitidis WI-1 adhesin gene in the GenBank database.
Controls.Ten microliters containing 100 fg (∼5,000 genome equivalents) of purified cloned plasmid DNA was used as a positive control in each PCR assay. In order to monitor crossover contamination, sterile water was included in the DNA extraction and was used as a negative control after every fifth sample in the nested PCR assay. Reaction mixtures without DNA were run in the first and nested PCRs to detect contamination.
Sequencing.The nested PCR products were purified by using the QIAquick PCR purification kit (Qiagen), based on DNA binding to a silica membrane. Automated sequencing was done with the BigDye terminator cycle sequencing kit and the primers used for the nested PCR, in accordance with the instructions of the manufacturer. PCR products were analyzed on an ABI 373 automated DNA sequencer (Applied Biosystems Division, Perkin-Elmer Biosystems, Foster City, Calif.). Sequences were generated from both strands, edited, and aligned with Sequence Navigator software (Applied Biosystems) and used for a BLAST search of the GenBank database.
After all PCRs had been done, the diagnosis and the results of microscopy were unblinded for further analysis.
RESULTS
In order to evaluate DNA extraction, DNA samples were amplified by a published PCR method targeting a canine housekeeping gene for detection of an acidic ribosomal phosphoprotein fragment (20). A PCR product was obtained from only 18 samples (data not shown), which was attributed to the limited sensitivity of this single-round PCR assay. There is strong homology between the mRNA sequences of canine and human GAPDH (GenBank accession numbers AB038240 and NM002046 , respectively), so we tried a previously described nested human GAPDH PCR assay. By using DNA extracted from the blood of healthy dogs, a single 249-bp product was obtained. Sequencing revealed that this product from canine blood is 95% homologous to the canine GAPDH mRNA sequence and 86 to 88% homologous to the human GAPDH mRNA sequence (data not shown). By application of this assay, one distinct 249-bp fragment was amplified from 58 of 73 canine samples (Table 1). A nested PCR product of a similar size is obtained from human GAPDH pseudogenes, of which several copies occur in the human genome (8). Failure to amplify the target sequence is an unequivocal sign of the absence of amplifiable canine DNA. This may be due to the failure to extract DNA or the destruction of DNA by formalin fixation.
Results of PCR assays and sequencing of 73 canine paraffin-embedded tissue samples examined
As shown in Table 1, the B. dermatitidis adhesin PCR amplified DNA from 8 of 13 samples which were positive by microscopy. The positive specimens consisted of three lymph node specimens, three lung tissue specimens, and one specimen each of kidney tissue, testis tissue, and prostate gland tissue from three different dogs. According to sequencing, the sequences of all PCR products except for those for two samples from one dog were identical to the WI-1 adhesin gene sequence deposited in GenBank. These PCR products showed an identical 5-nucleotide insert (3′-TGTAC-5′) at position 1658 of the GenBank sequence.
The B. dermatitidis adhesin PCR assay failed to amplify DNA from the remaining 5 microscopically positive and the 12 microscopically negative samples from dogs with blastomycosis. No PCR product was recovered from any sample from dogs with lymphosarcoma.
In contrast, the 18S rDNA PCR amplified DNA from 19 samples (Table 1). Sequencing identified only the sequences from six of these as being identical to the B. dermatitidis 18S rDNA sequence. All six PCR products were obtained from samples which were positive by the B. dermatitidis adhesin PCR assay as well. The remaining four sequences of PCR products from spleen, liver, kidney, and lymph node tissue specimens from three dogs with blastomycosis and nine sequences of PCR products from lymph node, liver, and spleen tissue specimens from five dogs with lymphosarcoma were homologous to the sequences of several species of the class Euascomycetes. Colonization or contamination with nonpathogenic ubiquitous fungi during autopsy is the most likely explanation, because all DNA extraction controls and reaction mixture controls were proven to be negative in all PCR assays carried out. The presence of contamination during the extraction procedure and crossover contaminants can therefore be excluded.
Specificity.The 18S rDNA PCR amplified DNA from all 20 human samples with histologically proven histoplasmosis. Sequencing identified all products as being identical to H. capsulatum 18S rDNA. In contrast, the B. dermatitidis adhesin PCR did not amplify DNA from any of these specimens from humans with histoplasmosis.
Detection limit.Serially diluted cloned plasmid DNA was repeatedly used as a template in the B. dermatitidis adhesin PCR assay. A product was detected by using a minimum amount of 0.1 fg of plasmid DNA. If it is assumed that a single copy of the WI-1 adhesin gene is present per genome, a detection limit of 5 genome equivalents can be calculated.
DISCUSSION
The two nested PCR assays are capable of identifying B. dermatitidis in formalin-fixed, paraffin-embedded canine tissue samples. The detection limits and sensitivities of the two assays turned out to be similar. No false-positive results were obtained by the PCR targeting the WI-1 adhesin gene. In contrast, the 18S rDNA nested PCR assay revealed products in 13 additional samples which required sequencing to identify them as nonspecific reaction products. These findings confirm earlier observations on diagnostic PCR assays for closely related fungi (5, 7). A PCR assay targeting rDNA might achieve a higher sensitivity for the detection of small numbers of a variety of fungal pathogens by use of universal primers. However, the ubiquitous occurrence of these conserved genes in clinical specimens is detrimental because nonpathogenic contaminating or colonizing fungi can cause considerable nonspecific amplification. Even the high stringency of our nested PCR with a high annealing temperature of 72°C cannot avoid nonspecific amplifications which may not even be recognized by sequencing (6).
The failure of a sensitive nested PCR targeting the multicopy GAPDH gene to amplify canine DNA from 21% of our samples reveals the well-known disastrous effect of formalin fixation on DNA (2, 19). A single-round PCR targeting a housekeeping gene is not sensitive enough to prove extraction of small amounts of amplifiable canine DNA. A nested PCR assay targeting the human GAPDH gene was also a sensitive indicator of the presence of canine DNA. Analogous to the situation for human cells, several gene copies can be assumed to exist in the canine genome. Failure to amplify this gene indicates that DNA was destroyed or that extraction failed.
For five microscopically positive samples, both fungal PCR assays failed to amplify specific DNA, most likely due to the presence of small amounts of fungal DNA in the total volume extracted. The presence of mutations in the primer binding region is another possible explanation, but the presence of significant variations in at least two regions resulting in two negative results for PCR assays that target two different conserved fungal genes seems highly unlikely.
Recently, an in situ hybridization protocol for the identification of yeast-like organisms in tissue sections that uses specific and panfungal oligonucleotides complementary to 18S and 28S rDNA was published (10). A sensitivity of 90% and a specificity of 97% were reported when 20 samples positive for B. dermatitidis, as confirmed by culture, were examined. Compared to PCR assays that require intact DNA strands as long as 450 bp, the binding of specific oligonucleotides is less sensitive to the destruction of DNA by formalin. Thus, in situ hybridization assays may be superior to PCR assays for the detection of B. dermatitidis DNA in formalin-fixed tissues. However, the diagnostic PCR assays are intended for use with various unfixed clinical specimens like bronchoalveolar lavage fluid, cerebrospinal fluid, lung tissue, and bone tissue specimens and vitreous aspirates. PCR might be especially suitable for examination of fine-needle biopsy specimens and vitreous aspirates because only small amounts are usually available. Lymph node enlargement is common in blastomycosis, and up to 50% of dogs with systemic disease have ocular involvement (14), making PCR assays of fine-needle aspirate material a practical diagnostic tool.
In conclusion, we have developed a specific nested PCR assay for detection and identification of B. dermatitidis DNA in canine tissue samples by targeting a gene coding for a unique fungal adhesin protein. In contrast, the products of the equally sensitive PCR assay targeting the 18S rDNA must be confirmed by sequencing in order to avoid a high rate of false-positive results.
ACKNOWLEDGMENTS
The excellent technical assistance of Sonja Jauch and Beate Salten is highly appreciated.
Anna Cascante Cirera received a grant from German Academic Exchange Service (DAAD), Bonn, Germany.
FOOTNOTES
- Received 27 March 2002.
- Returned for modification 27 September 2002.
- Accepted 25 October 2002.
- Copyright © 2003 American Society for Microbiology
