DNA Extraction Method Affects the Detection of a Fungal Pathogen in Formalin-Fixed Specimens Using qPCR

Abstract

Museum collections provide indispensable repositories for obtaining information about the historical presence of disease in wildlife populations. The pathogenic amphibian chytrid fungus Batrachochytrium dendrobatidis (Bd) has played a significant role in global amphibian declines, and examining preserved specimens for Bd can improve our understanding of its emergence and spread. Quantitative PCR (qPCR) enables Bd detection with minimal disturbance to amphibian skin and is significantly more sensitive to detecting Bd than histology; therefore, developing effective qPCR methodologies for detecting Bd DNA in formalin-fixed specimens can provide an efficient and effective approach to examining historical Bd emergence and prevalence. Techniques for detecting Bd in museum specimens have not been evaluated for their effectiveness in control specimens that mimic the conditions of animals most likely to be encountered in museums, including those with low pathogen loads. We used American bullfrogs (Lithobates catesbeianus) of known infection status to evaluate the success of qPCR to detect Bd in formalin-fixed specimens after three years of ethanol storage. Our objectives were to compare the most commonly used DNA extraction method for Bd (PrepMan, PM) to Macherey-Nagel DNA FFPE (MN), test optimizations for Bd detection with PM, and provide recommendations for maximizing Bd detection. We found that successful detection is relatively high (80-90%) when Bd loads before formalin fixation are high, regardless of the extraction method used; however, at lower infection levels, detection probabilities were significantly reduced. The MN DNA extraction method increased Bd detection by as much as 50% at moderate infection levels. Our results indicate that, for animals characterized by lower pathogen loads (i.e., those most commonly encountered in museum collections), current methods may underestimate the proportion of Bd-infected amphibians. Those extracting DNA from archived museum specimens should ensure that the techniques they are using are known to provide high-quality throughput DNA for later analysis.

Fig 1. Comparison of overall Bd recovery success of Macherey-Nagel DNA FFPE extraction (MN) vs. PrepMan (PM).

MN-extracted swabs were 31% more effective than PM at detecting Bd from formalin-fixed specimens that had been previously identified as Bd-positive in the field.

Fig 2. Probability of Batrachochytrium dendrobatidis (Bd) detection after formalin fixation.

We used parameters from the best-fit generalized linear mixed models (Table 3) to predict Bd detection probabilities of the two different extraction methods, Macherey-Nagel DNA FFPE (MN), and PrepMan (PM). The differences in Bd detection probability between MN and PM are greatest at low and moderate Bd loads (as much as 50% at the median infection level of this study) and become more similar at higher Bd loads.

Fig 3. Nonzero zoospore equivalent values before and after formalin fixation for MN and PM-extracted swabs.

There was no significant relationship between pre-and post-preservation Bd loads on individual swabs for either the MN (R² = 0.0006; R² (adj.) = -0.026; p = 0.879) or the PM (R² = 0.003; R² (adj.) = -0.046; p = 0.798) extracted specimen swabs.

Fig 4. Post-fixation Bd detection success (top) and zoospore equivalents (bottom) across all swab events (n = 58).

(a) Bd detection success was significantly greater for swab event E as compared to all other swab events (p < 0.0001). (b) The two ZE outliers in Event E are from specimens that had pre-preservation Bd loads of less than 1 zoospore equivalent and were not detected, in any number of runs, by any of the PM-extracted swabs. Results are based on analysis of singlicate data.