Development of a Duplex Real-Time PCR Assay for the Differentiation of Blastomyces dermatitidis and Blastomyces gilchristii and a Retrospective Analysis of Culture and Primary Specimens from Blastomycosis Cases from New York (2005 to 2019)

Blastomycosis due to Blastomyces dermatitidis and Blastomyces gilchristii is a significant cause of respiratory mycoses in North America with occasional reported outbreaks. We developed a highly sensitive, specific, and reproducible TaqMan duplex real-time PCR assay for the differentiation of B. dermatitidis and B. gilchristii.

ABSTRACT

Blastomycosis due to Blastomyces dermatitidis and Blastomyces gilchristii is a significant cause of respiratory mycoses in North America with occasional reported outbreaks. We developed a highly sensitive, specific, and reproducible TaqMan duplex real-time PCR assay for the differentiation of B. dermatitidis and B. gilchristii. The new assay permitted retrospective analysis of Blastomyces cultures (2005 to 2019) and primary clinical specimens from blastomycosis cases (2013 to 2019) from New York patients. We identified B. dermatitidis as the predominant pathogen in 38 cases of blastomycosis, while B. gilchristii was a minor pathogen involved in five cases; these findings expand understanding of blastomycosis in New York. The duplex real-time PCR assay could be implemented in reference and public health laboratories to further understand the ecology and epidemiology of blastomycosis due to B. dermatitidis and B. gilchristii.

INTRODUCTION

Until recently, Blastomyces dermatitidis was considered a single species within the genus Blastomyces and the sole etiologic agent of pulmonary and disseminated blastomycosis (1). This knowledge base was behind the development of several diagnostic tests (2–5). With the recent advances in fungal biology, several distinct genetic populations were identified as new species within the genus Blastomyces. Among the new species are Blastomyces gilchristii (6, 7), Blastomyces percursus, and Blastomyces silverae (8, 9). Blastomyces helicus was proposed as a new species responsible for rare blastomycosis cases in the western part of North America. Finally, B. percursus was proposed as a new species for blastomycosis cases from Africa (10, 11). In addition, taxonomic revisions assigned Emmonsia parva as Blastomyces parvus and Emmonsia helica as Blastomyces helicus within an expanded Blastomyces genus (9). Blastomyces dermatitidis is the predominant species in North America. Blastomyces gilchristii might be the leading cause of blastomycosis in northwestern Ontario and northern Wisconsin areas where the disease is hyperendemic (10). The ecological niche(s) of B. dermatitidis and B. gilchristii remains undefined. Similarly, the role of remaining rare Blastomyces species in the overall incidence and outcome of blastomycosis cases requires further investigations.

All Blastomyces spp. are thermally dimorphic fungi, existing as a mold form at ambient temperature in the environment. Blastomyces spp. convert to the pathogenic yeast form in vitro at 37°C or when a susceptible mammalian host inhales the conidia of the mold form. The yeast forms of these pathogens have distinguishable microscopic features: B. dermatitidis and B. gilchristii produce abundant, large (8- to 20-μm) broad-based budding yeasts, B. percursus produces large yeast-like cells from fragmented swollen hyphal cells, B. helicus produces variably shaped yeast cells (4 to 5 μm) in short chains, and B. parva and B. silverae produce thin-walled giant cells and a few broad-based budding yeast-like cells (12). A close examination of the yeast morphology might help in the species differentiation in the primary specimens. However, microscopic diagnosis requires expertise in mycology, which is not available in all laboratories. Culture and histopathology will help diagnose blastomycosis, but molecular methods are needed for accurate identification of newly described Blastomyces species. There are currently no molecular or serologic diagnostic tests for the rapid and accurate identification of species within the genus Blastomyces. An available commercial test (Gen-Probe, Inc., San Diego, CA) is somewhat limited, as it can be used only with pure cultures of B. dermatitidis. Also, culture-based identification is not feasible in all laboratories because of the occupational hazard associated with the fungus mold form. Earlier, we developed a rapid and specific real-time PCR assay to identify and detect the then-known four haplotypes of B. dermatitidis (5).

Outdoor exposure and proximity to water have been associated with blastomycosis (13). Blastomycosis is reportable in only five states in the United States, namely, Arkansas, Louisiana, Michigan, Minnesota, and Wisconsin (13). In New York State, blastomycosis is considered an emerging disease (http://www.nycasm.org/Alerts/Notification_101674.pdf). Several recent case series and case reports describe blastomycosis from New York. However, areas in New York where blastomycosis is endemic remain undefined, and it is not included in the list of reportable diseases (14–17). Likewise, in the absence of a skin test, it is difficult to assess the geographic location of likely exposure to the pathogen. Limited publications on canine blastomycosis suggest that the disease is prevalent in the Saint Lawrence River valleys on the New York-Canada border (14). Pneumonia is the most common manifestation of blastomycosis; approximately half of all cases are asymptomatic (18). However, Blastomyces infection can lead to a severe and fatal disease, often because of respiratory failure (19). Disseminated infection can involve any organ, often including cutaneous abscesses and osteomyelitis, and is frequently accompanied by fever, weight loss, and night sweats (2).

BAD-1 is an important conserved adhesion-promoting protein and virulence factor of B. dermatitidis (20). Two large insertions were noted in the B. gilchristii BAD1 gene (21), while this gene was absent in African strains of Blastomyces spp. (22). Several conventional PCR assays were designed using the BAD1 gene (23) for the detection of B. dermatitidis DNA from clinical and soil samples (4, 24). Sidamonidze et al. (5) developed a real-time PCR assay targeting the BAD1 gene to identify B. dermatitidis in culture and primary specimens. In the present investigation, we focused on the sequence variations observed within BAD1 to develop a duplex real-time PCR assay for the differentiation of B. dermatitidis and B. gilchristii. We present data on the sensitivity, specificity, and reproducibility of the assay. The new assay was used for a retrospective analysis of culture and primary specimens from blastomycosis cases from 2005 to 2019 in New York. Our results highlight the idea that B. dermatitidis is the major pathogen in blastomycosis cases in New York, with B. gilchristii being involved in few cases.

MATERIALS AND METHODS

Strains, primary specimens, and DNA.

Seventy-nine isolates of Blastomyces spp. were evaluated in this study. Forty-eight clinical isolates came from the sporadic cases of blastomycosis between 2005 and 2019 from New York. Isolates from canine (6 isolates), feline (1 isolate), sea lion (2 isolates), bat (1 isolate), polar bear (1 isolate), soil (4 isolates), and human (2 isolates) sources were obtained from Gene M. Scalarone (Department of Biological Sciences, Idaho State University). American Type Culture Collection (ATCC) strains used were MYA2586 (soil), MYA2585 (dog), ATCC 56214 (human, Africa), and ATCC 56216 (human, Africa). Five well-characterized strains each of B. dermatitidis and B. gilchristii were obtained from the Public Health Laboratory, Public Health Ontario, Toronto, Ontario, Canada. The species-level identifications of Blastomyces cultures from Canada were confirmed by sequencing of the ITS2 and drk1 genes described in an earlier publication (7). Thirty-three primary clinical specimens from New York patients, including 12 paraffin-embedded tissues, six bronchial wash specimens, five lung tissue samples, three cerebrospinal fluid samples, two skin lesion tissue samples, two bone marrow samples, and one sample each of blood, brain lesion tissue, and sputum submitted for B. dermatitidis identification, were also part of this investigation. Additionally, DNA from closely and distantly related fungal pathogens was included as part of a specificity panel (5).

DNA extraction.

DNA samples from Blastomyces spp. were extracted using a Qiagen DNA minikit on the Qiacube automated extractor. In brief, fungal growth (approximately 5 by 5 mm) was removed from the agar surface using a sterile loop and added to a lysing reagent in a biosafety level 3 (BSL3) laboratory. The fungal suspension was incubated at 90°C for 1 h and then brought to a BSL2 laboratory for further processing. The heat-killed fungal suspension was homogenized three times in a Precellys homogenizer at 6,500 rpm for 15 s each time (program number 5; 6500-3x60-015). The homogenized suspension was transferred to a 2-ml screw-cap tube, leaving behind the beads. DNA from samples was extracted using the Qiagen DNA minikit in the QiaCube semiautomated DNA extractor, resulting in 50 µl of eluted DNA. All extracted DNAs were stored at −80°C. DNA of fungal species (yeasts and molds) other than Blastomyces spp. were procured from the Wadsworth Center Mycology Laboratory (WCML) DNA Collection Repository.

The Wadsworth Center Histopathology Core first sectioned DNA from paraffin-fixed tissues, and then the paraffin was dissolved with 1 ml of xylene, followed by two washes using 1 ml of 100% ethanol. The sample was then dried and extracted using the Qiagen DNA minikit as described above with an incubation temperature of 70°C instead of 90°C. For nonfixed tissues, an approximately 5- by 5-mm section of tissue was cut with a sterile blade, and extraction was done using the Qiagen DNA minikit as described for the isolates, with an incubation temperature of 70°C instead of 90°C.

Primers and probes.

Primers and probes for the real-time PCR assay were designed from the promoter region of the BAD1 gene using Geneious R9 9.1.6 software (Biomatters, Inc., San Diego, CA). The choice of BAD1 promoter was based upon our earlier successful utilization of this region for the singleplex real-time PCR assay for B. dermatitidis identification/detection (5). Multiple alignment of BAD1 promoter sequences revealed unique sequences for B. dermatitidis and B. gilchristii, which were used for primers and probes design (see Fig. S1 in the supplemental material). Oligonucleotide sequences of the primers and probes for B. gilchristii and B. dermatitidis are as follows: V2556 (B. gilchristii; forward), 5′-ATGGGTGCAAAATCCGCCTA-3′; V2558 (B. gilchristii; reverse), 5′-AATCTAGAAGCTGGAGCGCC-3′; V2557 (B. gilchristii; probe), 5′-FAM (6-carboxyfluorescein)-CCGTACTCC/ZEN/CTCCCCGGTACTCC-3′IABkFQ (Iowa Black fluorescent quencher); V2559 (B. dermatitidis; forward), 5′-GCAAAATCCGCCTACTACTA-3′; V2561 (B. dermatitidis; reverse), 5′-AGCTGAACCTGGAAGTATTG-3′; V2560 (B. dermatitidis; probe), 5′-Cy5-TCCCTACCC/TAO/CTGGCTACTTTTCT-3′IAbRQSp (Iowa Black red quencher spacer). The internal ZEN and TAO quencher was incorporated between bases 9 and 10 from the 5' end of the probes V2557 and V2560, respectively. This design decreased the distance between the dye and the quencher and was expected to reduce the background signal and achieve an improved dynamic range. The primers and probes were purchased from Integrated DNA Technologies, Inc. (Coralville, IA).

Real-time PCR assay.

Each DNA sample (isolate or primary specimen) was tested in duplicate in 20-μl reaction volumes using an optical 96-well plate. Each reaction mixture contained 1× PerfeCTa multiplex qPCR ToughMix (Quanta Biosciences), a 1,000 nM concentration of each B. gilchristii (V2556 and V2558) and B. dermatitidis (V2559 and V2561) primer, a 250 nM concentration of each B. gilchristii (V2557) and B. dermatitidis (V2560) probe, and 2 μl of DNA (approximately 1 to 10 ng) extracted from Blastomyces spp. or 5 μl of DNA extracted from primary clinical specimens. Each PCR run also included 2 μl (1 ng) of positive extraction control (M808; B. dermatitidis), 2 μl (1 ng) of positive amplification controls (M808 [B. dermatitidis] and MB97 [B. gilchristii]), and 2 μl of negative extraction (extraction reagents only) and negative amplification (sterilized nuclease-free water) controls. The unidirectional workflow kept the reagent preparation, specimen preparation, and amplification and detection areas separate to avoid cross-contamination. Cycling conditions on the ABI 7500 FAST system (Applied Biosystems, Thermo Fisher Scientific Inc., Waltham, MA) were initial denaturation at 95°C for 20 s, followed by 45 cycles of 95°C for 3 s and 60°C for 30 s. Based on limit of detection (LOD), a cycle threshold (C_T) value of ≤38 was reported as positive and >38 was reported as negative. Specimens were reported as inconclusive if PCR inhibition was observed for the primary specimens.

Data availability.

GenBank accession numbers for BAD1 are MT822768 to MT822773; GenBank accession numbers for the internal transcribed spacer (ITS) genes are MT822762 to MT822767.

RESULTS

Assay sensitivity, specificity, and reproducibility.

The duplex real-time PCR assay was highly sensitive, with linearity over 5 logs, a correlation coefficient of >0.99, and amplification efficiency of >92%. The limit of detection (LOD) of the assay was 1 pg of genomic DNA (gDNA) per PCR within the linear range of the standard curve for B. dermatitidis and B. gilchristii (Fig. 1 and Table 1). The assay was highly specific, as it did not cross-react with other closely and distantly related fungal pathogens (see Table S2). The duplex real-time PCR assay correctly identified B. dermatitidis and B. gilchristii from 10 genomic-DNA samples submitted in a blind fashion from Ontario, Canada, confirming the assay validity (Table 2). Additionally, DNA from B. dermatitidis and B. gilchristii strains run on three different days and within the same day yielded consistent C_T values with a cofficient of variance of <5.0, confirming assay reproducibility (Table S3A and B).

FIG 1.

Sensitivity of duplex real-time PCR assay. Serial 10-fold dilution series of gDNA of B. dermatitidis and B. gilchristii were prepared, and PCR was run in duplicate. The slope (s), correlation coefficient (R²), and amplification efficiency (E) are shown.

TABLE 1.

Sensitivity study of B. dermatitidis and B. gilchristii^a

DNA concn (ng/µl)	B. dermatitidis			B. gilchristii
	CT 1	CT 2	Mean CT	CT 1	CT 2	Mean CT
10	20.73	20.70	20.72	19.85	19.76	19.80
1	23.88	23.87	23.88	23.10	23.08	23.09
0.1	27.25	27.51	27.38	26.54	26.44	26.49
0.01	32.96	32.60	32.78	30.59	30.15	30.37
0.001	35.96	35.27	35.62	33.68	33.64	33.66
0.0001	36.91	37.48	37.20	36.22	37.74	36.98
0.00001	Undet	Undet	Undet	Undet	Undet	Undet

^aUndet, undetermined.

TABLE 2.

Blind panel of DNA of B. dermatitidis and B. gilchristii^a

Sample no.	Mean CT in real-time PCR assayb			Final ID
	Singleplex	Duplex
		B. dermatitidis (Cy5)	B. gilchristii (FAM)
1	23.02	23.48	Undet	B. dermatitidis
2	22.4	Undet	21.74	B. gilchristii
3	18.52	20.04	Undet	B. dermatitidis
4	22.31	22.64	Undet	B. dermatitidis
5	19.86	Undet	19.35	B. gilchristii
6	22.12	Undet	21.54	B. gilchristii
7	23.28	Undet	22.46	B. gilchristii
8	19.75	20.15	Undet	B. dermatitidis
9	22.23	22.39	Undet	B. dermatitidis
10	21.45	Undet	20.39	B. gilchristii

^aDNA was supplied by J.V.K. in a blind fashion.

^bUndet, undetermined.

Retrospective analysis of Blastomyces isolates and primary specimens.

The retrospective analysis of 79 isolates of Blastomyces spp. revealed 62 to be B. dermatitidis and 15 to be B. gilchristii (Table 3). One isolate (ATCC 56214), which was identified by singleplex real-time PCR assay as Blastomyces spp., was neither B. dermatitidis nor B. gilchristii by the duplex real-time PCR assay. It was later confirmed to be B. percursus by sequencing of the ribosomal genes and BLAST search. One strain of Blastomyces emzantsi (ATCC 21516), identified by sequencing, was not identified by the singleplex real-time PCR or the duplex real-time PCR assay, further confirming the specificity of the current duplex real-time PCR assay. Of 15 B. gilchristii isolates identified, 5 were well-characterized strains of B. gilchristii from Canada, 4 were from the Eagle River outbreak (Wisconsin) involving human, dog, and soil samples, and 6 isolates were recovered from five New York patients. All six New York isolates of B. gilchristii were also confirmed by sequencing of the BAD1 and ribosomal genes. Of 62 B. dermatitidis isolates identified, 5 were well-characterized strains of B. dermatitidis from Canada, 15 were from Illinois, Kentucky, Minnesota, Tennessee, and Wisconsin involving animals or soil and 42 isolates were from 38 patients from New York collected from 2005 to 2019. Of 33 primary specimens analyzed from 2013 to 2019, five specimens (one skin lesion, one bronchial wash, two tissue blocks of nasal masses, and one tissue block of unknown origin) from five patients were positive for B. dermatitidis DNA while none were positive for B. gilchristii DNA (Table 3).

TABLE 3.

Validity of duplex real-time PCR assay for culture identification of Blastomyces spp. and direct detection from primary specimens

Assay type and singleplex real-time PCR result (n)	No. of samples with duplex real-time PCR result
	B. dermatitidis		B. gilchristii
	Positive	Negative	Positive	Negative
Culture identification
Positive (78)	62	16	15	63
Negative (60)	0	60	0	60
Direct detection
Positive (5)	5	0	0	5
Negative (28)	0	28	0	28

Blastomycosis in New York.

Analysis of 43 cases of blastomycosis from New York (Table S1) revealed one to four cases per year from 2005 until 2016, but that number increased to 7 in 2017, and cases remained high, with 9 identified in 2018 and 8 identified in 2019 (Fig. 2). Men were more frequently infected than women with B. dermatitidis, while the number of cases was too small to derive this conclusion for B. gilchristii. Older people (over age 40) were found to be more prone to symptomatic infection irrespective of whether the agent was B. dermatitidis or B. gilchristii. B. dermatitidis was most commonly isolated from respiratory specimens, followed by skin/wounds/subcutaneous tissue and bone, while all six isolates of B. gilchristii were isolated from respiratory specimens (Table 4). Geographic distribution of blastomycosis cases revealed that the majority of patients presented with symptoms and lived in Mohawk Valley, Capital District, and Finger Lakes. Few cases were also reported from other regions in New York (Fig. S2).

FIG 2.

Blastomycosis cases in New York from 2005 to 2019. Blastomycosis cases confirmed earlier by singleplex real-time PCR assay for isolates and primary specimens were analyzed retrospectively with the newly developed duplex real-time PCR assay. Of 43 cases, 38 were confirmed as being caused by B. dermatitidis and 5 as B. gilchristii. There was a marked increase in blastomycosis cases from 2017 onward.

TABLE 4.

Characteristics of Mycology Laboratory (Wadsworth Center)-confirmed blastomycosis cases in New York from 2005 to 2019

Characteristic	Value for group
	Blastomycosis	B. dermatitidis	B. gilchristii
No. of patients	43	38	5
Sex
M	29	26	3
F	14	12	2
Age (yr)
20–29	4	4	0
30–39	4	4	0
40–49	8	7	1
50–59	11	10	1
60–69	13	10	3
>69	3	3	0
No. of specimens (n = 48)
Bronchial wash	28	22	6
Skin, wound, subcutaneous tissue	15	15
Bone	4	4
Brain	1	1
No. of specimens positive for DNA (n = 5)
Bronchial wash	1	1
Tissue block—nasal mass	2	2
Tissue block—source not provided	1	1
Skin wound fluid	1	1

DISCUSSION

In this study, we developed a duplex real-time PCR assay to differentiate B. dermatitidis from B. gilchristii. These two closely related species possibly overlap in their endemicity in North America. The duplex real-time PCR assay targeting the putative promoter region of the BAD1 gene was highly sensitive, specific, and reproducible. BAD1 and its promoter have been extensively used for the differentiation of Blastomyces species using restriction fragment length polymorphism (RFLP), PCR, and real-time PCR assays (5, 21, 24). The polymorphism in BAD1, its markedly different sizes (363 bp in B. dermatitidis and 663 bp in B. gilchristii), and the absence of this gene in African isolates of Blastomyces (5, 21, 22) confirmed the choice of this target as a highly specific one for the identification of B. dermatitidis and B. gilchristii in the present investigation.

The retrospective analysis of Blastomyces cultures and primary specimens confirmed 43 cases of blastomycosis in New York from 2005 to 2019 based on the samples received in our facility. We noted an increase in number of cases in the recent years. The reason behind the observed increase in blastomycosis cases in New York are not apparent; our results are in agreement with other investigations describing the increase in case counts in New York (15, 17, 25, 26). Of interest was the identification of five of 43 blastomycosis cases due to B. gilchristii. Interestingly, patients infected with B. gilchristii resided in different parts of New York. Additional studies are needed to determine if these cases are due to patients traveling between regions or a focal niche of B. gilchristii in New York. We found that most blastomycosis cases were caused by B. dermatitidis, with concentrations of cases in Mohawk Valley and the Capital District, followed by the Finger Lakes. The regional aspect of geographic risk for blastomycosis is not well understood. It is challenging to track the progression from exposure to disease onset in patients. A skin test or another reliable marker of prior exposure is not available, and B. dermatitidis is rarely recovered from the environment (10). Furthermore, the recognition of clinical cases in New York is sporadic. Therefore, evidence regarding areas of endemicity in New York is equivocal. Our results are in agreement with other reports indicating that B. dermatitidis has a broader geographic distribution throughout North America while B. gilchristii has a limited distribution in the northern United States and certain Canadian provinces (27).

We know little about the geographic and phenotypic differences between B. dermatitidis and B. gilchristii. However, preliminary studies suggest that there may be a difference in the clinical manifestation of the diseases between the two species (28). There is also a report of an acute respiratory distress syndrome-related fatal case due to B. gilchristii (29). BAD1 has been implicated as the major virulence factor of B. dermatitidis (20). The significant difference in size of the BAD1 promoter due to two large insertions between B. dermatitidis and B. gilchristii noted above and the issue of whether these insertions have any influence on BAD1 expression and virulence need further investigation (21). To the best of our knowledge, no studies are available to demonstrate the comparative virulence of B. dermatitidis and B. gilchristii in vertebrate or invertebrate model systems. The highly sensitive and specific assay developed in the present study would allow more comprehensive surveillance of blastomycosis to monitor disease incidence. Systematic disease reporting and surveillance efforts will help diagnose new blastomycosis cases. Prompt diagnosis ensures prompt initiation of treatment to decrease illness and death. Finally, public awareness campaigns such as health advisories might be needed for blastomycosis in New York. In summary, the newly developed duplex real-time PCR assay would ensure accurate laboratory diagnosis of B. dermatitidis and B. gilchristii and help expand the understanding of the ecology and epidemiology of blastomycosis.