Effects of freezing on ability to detect Mycobacterium avium subsp. paratuberculosis from bovine tissues following culture

Abstract

Mycobacterium avium subsp. paratuberculosis (MAP) is the bacterium that causes Johne’s disease in cattle. Although infected cattle can be identified by examining fecal, blood, or milk samples, the gold standard is identification of MAP in tissue samples postmortem. Although tissue samples are commonly frozen, the ability to detect MAP in frozen–thawed tissue samples has apparently not been reported. We therefore determined the ability to detect MAP in tissue samples following freezing. Tissue samples were collected from calves that were either inoculated (IN) 3 mo prior, or contact-exposed (CE) for 3 mo. Following autopsy, tissues were immediately processed for culture, followed by DNA extraction and detection by qPCR. Samples were categorized as positive or negative based on the cycle threshold (Ct) value. The remaining unprocessed tissue samples were frozen at −80°C. After 18 mo, 50 tissue samples designated MAP-positive were thawed and processed for detection of MAP. Four (8%) samples were qPCR-negative, and Ct values of the remaining 46 samples were higher after freezing. Given the small numerical change in Ct values for MAP-positive samples after 18 mo of frozen storage, freezing and thawing may have had some deleterious effects on MAP detection in tissues. Although the decrease in ability to detect MAP-positive samples was minor for IN calves, there may be a greater effect for CE calves that should be considered when freezing tissue samples.

Johne’s disease is a production-limiting, costly disease in dairy cattle caused by Mycobacterium avium subsp. paratuberculosis (MAP).^7,14 Although identification of MAP-infected cattle is an important factor in control of the disease, diagnosis can be difficult given the prolonged incubation period, variable infection progression and immune responses, and unreliable detection tests as a result of poor test sensitivity.^1,11 Detection tests include individual fecal sampling and culture, interferon-γ test, antibody ELISA, and culture of tissue samples postmortem.^3,20 However, it is common to keep samples frozen prior to laboratory analysis.

Although tissue samples are often frozen for convenience or preservation, the freeze–thaw process may impact the resulting categorization of a sample as MAP-positive or -negative.⁸ However, research has focused primarily on the effects of freezing on fecal samples prior to culture. Initial studies indicated that the quantity of MAP recovered from fecal samples decreased after freezing at −80°C,¹⁷ although results from a 2011 study indicated that this temperature had little or no effect on viability.^9,16 However, despite tissue samples being commonly stored at −80°C before culture, there are apparently no reports regarding the effects of detecting positive samples in frozen–thawed tissues.

Deleterious effects of freezing on bacteria were attributed to formation of intracellular ice,¹² although a 2006 study reported that damage may be the result of intra- and extra-cellular osmotic imbalances during warming.⁵ Additionally, viability of MAP may vary depending on the sample matrix from which the bacteria are isolated (tissue vs. fecal; natural infection vs. artificially spiked); it was reported that fewer MAP were lost from the feces of naturally infected cattle after freezing compared to artificially spiked specimens.¹⁷ Culture followed by PCR is considered optimal for detection of MAP, because it allows for low numbers of bacteria to grow to a point at which genetic markers can be detected via PCR or quantitative PCR (qPCR).¹³ Although direct DNA extraction is documented, there is a risk that positive samples may be missed given the low numbers of bacteria in the tissue, especially among young stock soon after infection.¹⁹ Tissue culture is considered the gold standard for the detection of MAP infection.² However, following freezing, positive samples may be incorrectly identified as negative because of loss of viability and decreased growth. Our objective was to determine the effects of freezing tissues at −80°C for 18 mo on culture-based detection of MAP.

Tissue samples were collected during an experimental infection study.⁴ All protocols and the experimental design were approved by the University of Calgary Veterinary Sciences Animal Care Committee (Protocol AC14-0168). Briefly, 32 newborn Holstein–Friesian bull calves were collected from 13 dairy farms in Alberta (Canada) that had tested negative for MAP for at least 4 y. Calves were group-housed in 7 experimental pens consisting of 2 inoculated (IN) calves and 2 contact-exposed (CE) calves for 3 mo. Inoculum was prepared from a strain obtained from a clinical case (cow 69) in Alberta, and 2 calves in each experimental pen were inoculated with 2.5 × 10⁸ CFU on 2 consecutive days at 2 wk of age. Following 3 mo of group housing, IN calves were euthanized for tissue sampling, and CE calves were individually housed for an additional 3 mo before euthanasia. Full details regarding tissue collection, culture, and results of this conventional culture have been published.⁴ At the time of the initial study, only 2.5 g of the total sample of collected tissue was processed. The remaining tissue sample was stored frozen and then opportunistically used as a convenience sample for our study.

Of the tissue samples categorized in the initial study⁴ as culture-positive following fresh processing, 50 were systematically selected to be processed 18 mo after the initial autopsy and processing dates. The number of samples selected was based on a power calculation to detect a large or medium effect for categorization of samples or cycle threshold (Ct) values, respectively. Based on the power calculation, 32 samples were required to detect an effect at a p value of 0.05; however, given the lack of literature regarding the size of a potential freezing effect, 50 samples were selected. Given the variability of MAP-positive tissue cultures (intestinal tissue vs. lymph node),¹⁵ samples for processing after freezing were selected to include at least 1 intestinal tissue and 1 associated lymph node (LN). For IN calves, samples from the ileum were selected first. However, if no positive ileal or ileal LN sample was available, the next closest jejunal samples were selected. All 4 MAP-positive spleen samples were selected for additional processing. Additionally, 6 of the 50 samples were selected from CE calves, which represented 86% of all positive CE tissues. These tissues were thawed overnight in a 4°C refrigerator before processing. All sample-processing procedures, culture, DNA extraction, and detection were identical to those used in our initial study.⁴ In brief, LNs were cut into 2-cm cubes, mucosa was scraped from small intestinal samples, and samples were placed into pre-labeled Whirl-Pak bags (Nasco, Fort Atkinson, WI). Thereafter, a 2.5-g sample of tissue was removed from bags, weighed, and dissociated (gentleMACS dissociator, M tubes, Miltenyi Biotech, Auburn, CA) before undergoing a 24-h disinfection procedure specialized for retention of MAP.⁴ Briefly, the disinfection process involved incubation at 37°C for 3 h with 0.75% hexadecylpyridinium chloride (HPC) in half-strength brain heart infusion (BHI), followed by centrifugation (4,700 × g, 15 min) and re-suspension in a mixture of antibiotic solution (AS; para-JEM, TREK Diagnostic Systems, Cleveland, OH), water, and full-strength BHI and then incubated at 37°C overnight.⁴ Then, 1 mL of solution from the disinfected sample was added to culture bottles (TREK para-JEM) and incubated at 37°C for 49 d. Following culture, DNA was extracted from the culture bottles.⁶ A duplex qPCR assay targeting the MAP-specific F57 region and an internal amplification control was performed with primers and probes identical to those described.¹⁸ Samples were categorized as positive if the Ct value was <40.

All statistical analyses were performed using STATA 11.2 (StataCorp LP, College Station, TX). For all analyses, p ≥ 0.05 was considered significant. A Wilcoxon signed rank test was used to compare Ct values obtained before and after freezing of tissue samples, and a McNemar chi-squared test to compare number of positive samples before and after freezing.

We collected 91 culture-positive tissues from calves in the initial study,⁴ with 50 samples selected for processing after being frozen for 18 mo (Table 1). Of the 50 positive tissue samples that were processed and cultured after freezing, 4 samples were culture-negative based on qPCR. These samples included 2 spleen samples (1 IN, 1 CE), and 2 LN samples (both CE). LNs, and to some extent the spleen, have been shown to be fairly reliable tissues for identifying infected animals.¹⁵ Of the 4 spleen samples that were positive at initial culture, only 2 needed to be re-classified after a freeze–thaw cycle. Additionally, 3 of 4 negative samples following freezing originated from CE calves. Given the nature of transmission in the trial design, CE animals were not directly inoculated; rather, they became infected through direct and indirect contact with IN calves, which may have resulted in a lower infectious dose. A lower challenge dose leads to fewer tissue-positive samples and fewer MAP bacteria being present in infected tissues.^15,19 Although culture allows for a few bacteria to multiply to a detectable number, freezing may have reduced the viability of MAP present, causing a decrease in the ability to detect true positive samples. Fewer MAP, in combination with a decrease in viability caused by freezing, may explain why more CE samples changed status after freezing compared to IN samples. This apparent difference between IN and CE tissue samples to remain positive following freezing should be considered when culturing frozen samples from naturally infected animals, because the number of bacteria in tissue may differ. A negative culture result following freezing of a sample that would have been identified as positive if cultured before storing could have large implications for testing and surveillance purposes. Future studies would benefit from culturing field- and abattoir-derived tissue samples to determine effects of freezing collected from naturally infected animals.

Table 1.

Results of testing of tissue samples processed for Mycobacterium avium subsp. paratuberculosis following storage at −80°C for 18 mo. All samples selected for processing were initially positive by qPCR (Ct < 40) detection following culture of the fresh sample.⁴

Calf status/Location*	No. of culture-positive fresh samples	No. samples selected for processing after freezing†
Inoculated
Ileum (14)	9	9 (9)
Jejunum (42)	25	13 (13)
Ileum LN (14)	13	13 (13)
Jejunum LN (42)	34	6 (6)
Spleen (14)	3	3 (2)
Contact-exposed
Ileum (14)	0	0 (0)
Jejunum (42)	1	1 (1)
Ileum LN (14)	0	0
Jejunum LN (42)	5	4 (2)
Spleen (14)	1	1 (0)

LN = lymph node.

^*Numbers in parentheses are the total number of samples collected in initial study.

^†Numbers in parentheses are the number of samples testing positive after freeze–thaw cycle.

Positive tissue samples that were processed fresh had a mean Ct value of 24.8 (SD: 3.9, range: 20.4–37.1), whereas post-freezing samples had a mean Ct value of 26.6 (SD: 2.5, range: 24.2–37.0) following removal of negative samples (n = 4). Samples that were processed fresh had lower Ct values than those processed after freezing (p = 0.005). Although samples were frozen for an extended interval, which may have affected MAP viability, it is the number of freeze–thaw cycles that may negatively impact bacteria viability rather than freeze duration.^16,17 Quantification of MAP is possible based on DNA and resulting Ct values from qPCR,^10,18 but there is no way to differentiate between living and dead bacteria. Therefore, comparison of viable bacteria between fresh and frozen samples was not possible in our study. The lower Ct values would normally indicate a greater quantity of MAP in the samples; however, other deleterious effects from freezing may have caused slower growth, which would increase Ct values, and it cannot be determined how many MAP bacteria were killed in the freeze–thaw process. Despite an inability to quantify viable bacteria, it was noteworthy that the Ct values increased after a freeze and thaw cycle, and this should be considered when quantifying or detecting MAP from tissues in the future.

Although there was no difference between samples categorized as positive or negative (p = 0.13) given the change in Ct values, it is important to note that 4 positive samples were re-categorized as negative after freezing and thawing. Therefore, it is important to not only take multiple samples from a single animal, but if possible, process samples immediately. Additionally, naturally exposed animals may be more difficult to detect; therefore, freezing effects may be more important in this population. It should be taken into consideration that tissue samples were only cultured once before freezing; perhaps the change in Ct values following freezing may be the result of variability in sample homogeneity of MAP. However, samples processed after freezing had consistently higher CT values, consistent with an effect of freezing, rather than with within-sample variation. Additionally, only positive tissue samples were selected for re-processing following freezing; therefore, no conclusions regarding specificity were made, because no negative samples were re-processed after freezing.

Further research investigating MAP viability after a freeze–thaw cycle of tissues would provide further insights and would be of great importance for studies that culture MAP from tissues. Impacts on MAP viability could be investigated by plating on solid media or time-to-positive culturing techniques, rather than exclusively DNA-based measures of detection. Additionally, a study of intervals between freezing and thawing and various thaw methods may provide insight regarding optimal storage and thawing procedures for tissues suspected of being infected with MAP. Ultimately, further understanding regarding the loss of MAP in tissue samples following freezing will lead to more credible research findings regarding MAP infection, and identification of infected animals.