Materials and Methods

Samples—Aspergillus fumigatus (ATCC 1022) and C albicans (ATCC 10231) were acquired from a commercial company^b and used to develop the qPCR assay. Twelve samples obtained from a clinical diagnostic laboratory were used to evaluate the qPCR assay. Finally, 29 samples obtained from 17 mares suspected of having fungal endometritis were evaluated with the qPCR assay.

Development of a 28S ribosomal qPCR assay to detect fungal organisms—Serial logarithmic dilutions of A fumigatus and C albicans in sterile (PCR-grade) PBS solution were used as control samples. Calcium alginate–tipped swabs were dipped in the logarithmic aliquots to mimic clinical samples collected from the uterus. These swab specimens were placed in 1.0 mL of sterile PBS solution for 1 hour and then vortexed for 1 minute. Each sample was poured into a 1.5-mL microcentrifuge tube and centrifuged at 10,000 × g for 3 minutes. Supernatant was removed, and the pellet was resuspended in 200 μL of sterile PBS solution. Fungi were lysed by incubation at 95°C for 10 minutes. Following cell lysis, 180 μL of tissue lysis buffer^c and 20 μL of proteinase K were added, and the mixture was briefly centrifuged and then incubated at 54°C for 1 hour. Extraction of DNA was performed with a commercial tissue extraction kit.^d Quantitation of fungal DNA was performed by use of a real-time PCR reaction that was adapted from an established protocol.⁵ Briefly, a volume of 20 μL that contained 10 μL of commercial fluorescent PCR dye,^e 0.4 μL of NL1 primer (5′-GCATAT-CAATAAGCGGAGGAAA AG-3′),⁸ 0.4 μL of 260R (5′-TTAGCTTTAGA TGRARTTTACCACC-3′),⁵ 5 μL of extracted template DNA, and 4.2 μL of sterile water was used in the real-time PCR assay. All reactions were performed in a thermocycler^f; conditions were 5 minutes at 95°C, which was followed by 40 cycles (each cycle consisted of a denaturation step of 10 seconds at 95°C followed by an annealing step of 35 seconds at 60°C). The DNA from amplicons with positive results was submitted to a commercial laboratory^g for sequence analysis. Chromatograms were analyzed,^h and results were compared with published sequences by use of a sequence-similarity toolⁱ at default settings.

Assay sensitivity for common fungal organisms was estimated from serial logarithmic dilutions of A fumigatus and C albicans in sterile PCR-grade PBS solution. The sensitivity experiment was repeated with sterile saline (0.9% NaCl) solution recovered after uterine lavage of mares with negative results for bacterial and fungal cultures to evaluate for the presence of PCR inhibitors in the uterus of mares. The DNA extracted from cultured samples of Klebsiella pneumoniae, Pseudomonas aeruginosa, and Streptococcus equi subsp zooepidemicus obtained from the Veterinary Diagnostic Laboratory at Colorado State University was used to confirm a lack of cross-reactivity of primers with common bacterial uterine pathogens. Interassay CV was determined by repeating the qPCR assay (15 replicates) for 1 extraction of both C albicans and A fumigatus.

Evaluation of extraction techniques to isolate fungal DNA—The efficiency of DNA extraction was evaluated after exposure of swab specimens to various concentrations (10⁻¹ to 10⁻⁶ g of DNA) of A fumigatus and C albicans DNA. Ten swab specimens from each concentration were soaked in sterile PBS solution for 2 hours, vortexed, and centrifuged; DNA was then extracted as described previously. A second set of 10 swab specimens was placed in sterile PBS solution and immediately vortexed and centrifuged; DNA was then extracted as described previously. The Ct at which the minimum threshold was exceeded was compared between extraction groups to semiquantitatively determine whether the total amount of DNA extracted differed between the 2 extraction techniques.

Comparison of traditional microbial culture and the qPCR assay to detect fungal organisms—The qPCR assay was subsequently evaluated for its use in detecting fungal pathogens from agar plates obtained from the Veterinary Diagnostic Laboratory at Colorado State University. Twelve samples were submitted for traditional microbial culture and analysis via qPCR assay. The interassay CV for the qPCR fungal assay was determined by performance of the assay on all 12 samples 5 times. The purposes of this experiment were to determine whether the relative detectable amount of DNA in the sample was similar between qPCR assays, whether DNA sequencing could be used to identify the same organism in each assay, and whether the same organism would be identified via traditional fungal culture and the qPCR assay.

Use of the qPCR assay to detect fungal organisms in mares suspected of having fungal endometritis—A total of 29 samples (endometrial swab specimens, biopsy specimens, or low-volume lavage samples) obtained from 17 mares suspected of having fungal endometritis were submitted by veterinarians in private practice or at referral microbiological laboratories. A diagnosis of fungal endometritis was based on results of microbial culture or cytologic evaluation (or both) of endometrial samples. Additional samples were collected from 46 mares considered to be reproductively normal on the basis of a lack of ultrasonographically detectable intrauterine fluid during a 4-week period, no growth of pathogens on microbial culture of endometrial samples, and no inflammatory cells or microbial pathogens observed during cytologic examination of endometrial samples.

Statistical analysis—The Ct in which the minimum threshold was exceeded was compared between groups by use of a Student t test. Comparisons in frequency of identification of fungal organisms between microbial culture and the qPCR assay were performed with the Fisher exact test. Values were considered significantly different at P < 0.05.

Results

Development of 28S ribosomal qPCR assay to detect fungal organisms—The broad-spectrum primers were able to amplify fungal DNA from A fumigatus and C albicans as determined by a positive signal detected by the thermocycler (data not shown). Logarithmic dilution of fungal organisms in uterine lavage fluid revealed that the assay was able to detect a minimum of 2 × 10⁻¹⁴ g of fungal DNA. The typical C albicans organism contains approximately 37 fg of DNA. Theoretically, the assay has a detection limit of 2 organisms. There was no difference in the minimum detectable limits of fungal DNA in PBS solution or uterine lavage fluid (data not shown). The interassay CV of the logarithmic dilutions of A fumigatus DNA from 15 replicates of a single extraction was 6%. The primers used in the qPCR assay did not amplify DNA of bacterial organisms commonly isolated from horses with endometritis.

Evaluation of fungal DNA extraction techniques—The Ct for the 2 fungal DNA extraction techniques did not differ significantly (data not shown).

Comparison of traditional microbial culture and the qPCR assay to detect fungal organisms—The qPCR assay was able to amplify fungal DNA from all 12 samples (Table 1). The interassay CV, as determined by performing the qPCR assay multiple times on all 12 samples (5 replicates/sample), was < 4%. The qPCR assay and subsequent genetic sequencing could be used to identify the genus and species of fungi in all 12 samples, compared with identification of only 8 of 12 samples via fungal culture and biochemical methods. Identification via biochemical methods and DNA sequencing was similar for 6 of 8 samples. Results of the qPCR assay followed by genetic sequencing were obtained within 3 days after sample acquisition. In contrast, microbial culture required a mean of 3 days (range, 2 to 5 days) to enable detection of a fungal organism and an additional 18 days (range, 11 to 42 days) for determination of genus and species.

Use of the qPCR assay to detect fungal organisms in mares suspected of having fungal endometritis—The qPCR assay detected fungal DNA in samples from 12 of the 17 mares. Sequence analysis of the amplicon provided identification of the genus and species for all 12 samples (Table 2). Yeast-type fungal pathogens were identified in 10 of the 12 mares, whereas hyphate-type fungal organisms were identified in the other 2 mares. Fungal culture and qPCR assay for fungal DNA both yielded negative results for all uterine samples collected from the 46 reproductively normal mares.

Discussion

The use of molecular techniques to diagnose bacterial, fungal, and viral infections in clinical laboratories has exponentially increased during the past decade.¹⁰ Real-time PCR amplification combined with sequencing of PCR products is a more rapid technique than is traditional microbial culture. In the present study, 8 hours typically was required for the detection of fungal DNA with the qPCR assay, and approximately 72 hours was required to determine identification of the organisms via DNA sequencing. The interval from sample acquisition to pathogen identification was largely a result of the time required for overnight shipment of the amplicon to a diagnostic laboratory for DNA sequencing.

It has been suggested that amplification inhibitors may adversely affect PCR diagnostic testing, which can cause false-negative results.¹¹ The qPCR assay detected similar concentrations of fungal DNA diluted in sterile PBS solution or uterine lavage fluid, which suggested that amplification inhibitors are not important in the uterus. The sensitivity of the assay was similar for both diluents, which suggested that important PCR inhibitors were not present in uterine lavage fluid.

The qPCR assay detected fungal DNA in samples obtained from 12 of 17 clinically affected mares in which fungal endometritis was diagnosed via cytologic examination or microbial culture of endometrial samples. The inability of the assay to detect fungal DNA in samples from the other 5 mares could have been attributable to a fungal DNA load that was below the sensitivity of the assay or the presence of amplification inhibitors, or it is possible that the fungal cultures yielded false-positive results.¹¹

It should be mentioned that fungal endometritis was originally diagnosed in 5 of the 17 mares via cytologic examination of endometrial samples, whereas microbial culture of uterine samples obtained from these 5 mares yielded negative results for fungal growth. These 5 mares had positive results for fungal DNA via the qPCR assay, and an organism was identified.

The qPCR assay appeared to be more sensitive than was routine microbial culture. One reason suggested for this is the ability of qPCR assays to amplify DNA of nonviable organisms and free fungal DNA.¹² The uterus is considered to be a privileged immunologic site from which no fungal organisms should ever be recovered. Consequently, detection of any fungal DNA from a uterine sample would be clinically relevant.¹³ Another issue is possible contamination of the sample with fungal DNA from equine skin or fecal material during sample collection. Care must be taken during sample collection to minimize contamination of samples submitted for microbial culture or qPCR assay.

Unfortunately, the qPCR assay cannot be used to determine antimicrobial susceptibility for fungal organisms. Culture of fungal organisms currently remains the standard for determination of antimicrobial susceptibility. However, fungal organisms detected in equine uterine samples generally have clearly defined susceptibilities to antifungal drugs, and identification of a fungal organism via qPCR assay will provide valuable information for development of a treatment plan. Thus, the rapid return of qPCR assay results can allow for empirical administration of drugs likely to be effective on the basis of genus and species identification of the organisms. In the future, a genetics-based approach to antifungal susceptibility may be possible.

In the study reported here, a rapid, sensitive, and repeatable qPCR assay was developed for detection of fungal DNA in equine uterine samples. The qPCR assay may prove to be clinically useful as an adjunct to microbial culture and cytologic examination of uterine samples for identification of fungal organisms in a timely manner and allow for early development and implementation of a treatment plan.