Bovine TB is a zoonotic infection of cattle caused by Mycobacterium bovis} Bovine TB was first diagnosed in free-ranging white-tailed deer (Odocoileus virginianus) in Michigan in November 1975.2 Since surveillance for bovine TB began in 1994, the extent and characteristics of outbreaks of the disease in deer and details of local and national management efforts have been extensively described.3–5 Bovine TB is of concern because of its ability to infect a wide variety of species,6 including humans,7,8 and resulting costs of infection for the livestock industry because of herd condemnations and loss of markets.9 After numerous years of surveillance and research, white-tailed deer remain the only proven reservoir for TB infection of US cattle (other than other infected cattle).10 Wildlife reservoirs have also been identified outside the United States, including badgers in the United Kingdom,9,10 brush-tailed opossums in New Zealand,10 and elk (Cervus elaphus) in Manitoba, Canada.10 Even with intense efforts to eradicate bovine TB in the United States, this disease is detected in 8 US cattle herds annually.11
Cattle with bovine TB pose serious risks to free-ranging wildlife if poor biosecurity practices allow for inadequate separation at the wildlife-livestock interface. Free-ranging deer (Odocoileus spp) and elk are of most concern because they often are attracted to shared food resources of cattle and disease transmission can occur directly from infected animals or indirectly through fomites.
The complex ecology and continuing reemergence of M bovis necessitate rapid, thorough national and international surveillance and a better understanding of transmission dynamics among an increasing number of wildlife reservoirs and hosts in a variety of ecosystems. Reliance on the in vivo tuberculin skin test coupled with an assay to measure in vitro release of interferon-γ for the identification of infected free-ranging deer is untenable. Current methods require trapping and handling of animals, which are difficult tasks for farm or ranch deer and are not feasible or cost-effective for free-ranging populations. Thus, an option for broad-based surveillance would be an easy-to-perform diagnostic test that yields unambiguous results and that could be performed on serum samples obtained from hunter-harvested deer or as an antemortem test for farmed deer. Novel serum- or urine-based molecular markers are needed for identifying and monitoring the progression of M bovis infection in free-ranging animal populations to effectively estimate prevalence of bovine TB so that cost-effective control measures can be implemented to prevent disease transmission to domesticated animals and humans.
Identification of novel molecular markers involves detection of circulating mycobacterial peptides, lipids, or metabolites in serum or plasma of infected animals. By use of this approach, 16 M bovis proteins, including MB2515c (transcriptional regulator [LuxR family]), MB1895c (cell wall biosynthesis), and MB 1554c or pks5, were detected in bovine TB-infected and -exposed cattle.12 Biomarkers were first identified by use of a gel-free multidimensional isobaric tag for relative and absolute quantitation proteomics and subsequently validated by use of a well-characterized cattle serum repository.12,13 An indirect ELISA that involved use of monoclonal antibodies synthesized against these peptides was developed to detect such biomarkers in serum and has been validated in samples obtained from host cattle12 and primates with experimentally induced TB (unpublished data).
Current diagnostic tests have a 1-size-fits-all method that does not take into consideration the disease prevalence for a given region. The biggest risk for spread of bovine TB in a region with a low disease prevalence (eg, the United States) is not wildlife or their natural movements; rather, the biggest risk is from human movement of infected cattle.14 If disease is detected in deer, efforts to reduce the free-ranging population and limit the risk of disease transmission can be costly as well as logistically and politically challenging.14,15
The purpose of the study reported here was to use novel Mycobacterium-specific peptides to develop a diagnostic test that could be used on deer sera for diagnosis of M bovis infection. Use of tests to detect TB-specific biomarkers would have great potential to aid in the early detection and monitoring of bovine TB in wildlife through the use of hunter-harvested samples.
Materials and Methods
Sample
Serum samples were obtained from white-tailed deer in a laboratory environment and from Minnesota DNR hunter-harvested free-ranging white-tailed deer.
Experimentally infected deer—Eight yearling white-tailed deer were maintained in a biosafety level 3 laboratory at the National Animal Disease Center. The deer were inoculated with 2 × 108 CFUs of M bovis strain 1315 via intratonsillar instillation (day 0). Sera were obtained on days 0, 19, 48, and 60. A group of 3 age-matched control white-tailed deer were injected with saline (0.9% NaCl) solution on day 0, and serum samples were obtained at the same time points as for the inoculated deer. The age-matched control deer were housed separately from the M bovis-inoculated deer.13,16 Approvals by a USDA institutional animal care and use committee were obtained for use of the animals in this study.
Minnesota DNR hunter-harvested free-ranging deer—Sera were obtained from 393 Minnesota DNR hunter-harvested white-tailed deer from 2007 through 2010 during an outbreak of bovine TB in Minnesota.17,18 The DNR obtained these samples as a part of a TB eradication program in collaboration with the Minnesota Board of Animal Health. Trained staff of the Minnesota DNR performed necropsies and collected blood from the chest cavity of hunter-harvested deer. Nine deer had gross lesions compatible with bovine TB and histologically confirmed granulomas containing acid-fast bacilli, and M bovis was isolated from samples that were submitted to the National Veterinary Diagnostic Laboratory. Disease-negative animals were defined as deer that had negative culture results for M bovis. These 384 disease-negative deer were considered an exposed uninfected population for the analysis.
Biomarkers
Sera from calves experimentally infected with M bovis were analyzed by use of multidimensional proteomics, whereby 32 host and 16 pathogen-specific peptides were identified that specifically increased in serum of M bovis-infected calves.13 In addition, 16 M bovis-specific peptides were identified in the same proteomics data set.12 The 3 most abundant pathogen-specific peptides (MB1895c [a hypothetical protein with a molybdenum sulfurase domain], MB2515c [a transmembrane family protein as determined by analysis of amino acid sequences], and pks5 [a polyketide synthase potentially involved in lipid metabolism]) were identified by use of a well-characterized cattle serum repository and an antigen capture ELISA12 with monoclonal antibodies against the peptides. These 3 pathogen peptide biomarkers were further validated for specificity in deer by use of sera from experimentally infected deer.12,14 For each sample tested, contemporaneous control samples were also tested.
Indirect ELISA
Serum from each of the 8 experimentally infected white-tailed deer, 3 age-matched control deer, 9 TB-positive Minnesota DNR hunter-harvested deer, and 384 exposed uninfected deer were diluted (1:50 dilution) in 0.05M carbonate-bicarbonate buffera (pH, 9.6). Fifty microliters of each diluted serum sample was transferred to separate wells in polystyrene flat-bottom ELISA plates.b Each sample was assayed in duplicate, and positive and negative control samples were included on each plate. Sera were allowed to absorb overnight at 4°C. Plates were washed 3 times with PBS solution (200 μL/well). Plates were blocked by incubation with 5% bovine serum albuminc in TBS solution (200 μL/well) for 2 hours at 37°C, washed 3 times (300 μL/well) with 1X PBS solutiond containing 0.05% Tween 20,e and incubated (100 μL/well) with primary monoclonal antibodies against MB2515c (1:5,000 dilution), MB1895c (1:5,000 dilution), or pks5 (1.2 mg/mL) or with a 1% solution of bovine serum albumin in TBS solution with 0.05% Tween 20 for 2 hours at room temperature (22°C). Plates were then washed as described previously and incubated after addition (100 μL/well) of goat anti-mouse horseradish peroxidase-conjugated IgGf diluted 1:10,000 in 1% bovine serum albumin in TBS solution with 0.05% Tween 20 for 2 hours at room temperature. Plates then were washed as described previously and were developed by incubation with tetramethylbenzidineg (100 μL/well) in the dark for 30 minutes at room temperature. The tetramethylbenzidine reaction was stopped by the addition of 2M sulfuric acid (50 μL/well), and the OD at 450 nm was recorded by use of a microplate reader.h
Data analysis
The OD data were uploaded into a spreadsheet,i and S/N values were calculated for each biomarker; S/N was defined as the ratio of OD for the 9 TB-positive Minnesota DNR hunter-harvested deer against the mean OD of all the exposed uninfected deer. Samples for the 9 TB-positive Minnesota DNR hunter-harvested deer were used as the positive control samples for comparison against samples from the experimentally infected deer and the remainder of the Minnesota DNR hunter-harvested deer. These ratios were uploaded to a graphing program,j and plots of time-course modulation of biomarkers in experimentally infected or Minnesota DNR hunter-harvested deer were generated. Box-and-whisker plots were generated for each biomarker by use of the same 9 TB-positive Minnesota DNR hunter-harvested deer.
Quantities of each biomarker for the 9 TB-positive Minnesota DNR hunter-harvested deer were simulated over time as a function of the total number of animals with positive test results against the total number of samples tested in a given time period. Prevalence of bovine TB as determined on the basis of biomarker presence was used to establish the extent and degree of disease burden since eradication of TB after the 2007–2010 outbreak was completed.
Results
Experimentally infected deer
The S/N for pks5 and MB2515c increased gradually over the infection cycle and reached a peak at 60 days after inoculation (Figure 1). In contrast, the S/N for MB1895c increased early in the infection cycle and decreased after day 48 after inoculation.
The S/N values for biomarkers pks5 (A), MB2515c (B), and MB1895c (C) in serum samples obtained from 8 yearling white-tailed deer (Odocoileus virginianus) experimentally infected with Mycobacterium bovis. Deer were inoculated with 2 × 108 CFUs of virulent M bovis strain 1315 (day 0), and sera were obtained on days 0, 19, 48, and 60. Notice that the scale on the y-axis differs among panels.
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
The S/N values for biomarkers pks5 (A), MB2515c (B), and MB1895c (C) in serum samples obtained from 8 yearling white-tailed deer (Odocoileus virginianus) experimentally infected with Mycobacterium bovis. Deer were inoculated with 2 × 108 CFUs of virulent M bovis strain 1315 (day 0), and sera were obtained on days 0, 19, 48, and 60. Notice that the scale on the y-axis differs among panels.
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
The S/N values for biomarkers pks5 (A), MB2515c (B), and MB1895c (C) in serum samples obtained from 8 yearling white-tailed deer (Odocoileus virginianus) experimentally infected with Mycobacterium bovis. Deer were inoculated with 2 × 108 CFUs of virulent M bovis strain 1315 (day 0), and sera were obtained on days 0, 19, 48, and 60. Notice that the scale on the y-axis differs among panels.
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
TB-specific biomarkers in TB-positive Minnesota DNR hunter-harvested deer
Sera from the 9 TB-positive Minnesota DNR hunter-harvested deer were assayed. Range of the OD values (measured at 450 nm) was 0.198 to 0.407 for pks5, 0.211 to 0.510 for MB2525c, and 0.153 to 0.78 for MB1895c. Range of the S/N values was 3.394 to 6.977 for pks5, 4.234 to 10.234 for MB2515c, and 1.463 to 4.571 for MB1895c. Thus, S/N values for all 3 biomarkers calculated by use of sera from the 9 TB-positive Minnesota DNR hunter-harvested deer were between 1.463 and 10.234. Range of the S/N values obtained for each biomarker was used to identify cutoff values to determine infection status. By use of these ranges, an S/N value ≥ 2.5 for any of the 3 biomarkers was considered a positive result (infected), and an S/N value < 2.5 was considered a negative result (uninfected).
Evaluation of box-and-whisker plots of the S/N values for the 3 pathogen-specific biomarkers revealed that pks5 and MB2515c were the most reliable biomarkers for use in determining infection status for the 9 TB-positive Minnesota DNR hunter-harvested deer (Figure 2). Interquartile ranges for pks5 and MB2515c overlapped; thus, these 2 biomarkers were considered to be the most reliable. The S/N values for MB1895c differed slightly from those of the other 2 biomarkers; hence, it was not considered to be a reliable biomarker.
Box-and-whisker plots of S/N values for 3 biomarkers in sera of 9 M bovis-infected Minnesota DNR hunter-harvested white-tailed deer. The S/N values for pks5 and MB2515c indicated that they would be reliable biomarkers for detection of bovine TB infection, whereas MB1895c would not be a reliable marker. Each box represents the interquartile range (25th to 75th percentiles), the horizontal bar in each box is the median, and the whiskers are the range of S/N values.
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
Box-and-whisker plots of S/N values for 3 biomarkers in sera of 9 M bovis-infected Minnesota DNR hunter-harvested white-tailed deer. The S/N values for pks5 and MB2515c indicated that they would be reliable biomarkers for detection of bovine TB infection, whereas MB1895c would not be a reliable marker. Each box represents the interquartile range (25th to 75th percentiles), the horizontal bar in each box is the median, and the whiskers are the range of S/N values.
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
Box-and-whisker plots of S/N values for 3 biomarkers in sera of 9 M bovis-infected Minnesota DNR hunter-harvested white-tailed deer. The S/N values for pks5 and MB2515c indicated that they would be reliable biomarkers for detection of bovine TB infection, whereas MB1895c would not be a reliable marker. Each box represents the interquartile range (25th to 75th percentiles), the horizontal bar in each box is the median, and the whiskers are the range of S/N values.
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
Prevalence of bovine TB in Minnesota deer
Infection status of 384 Minnesota DNR hunter-harvested white-tailed deer in northwestern Minnesota was evaluated by use of serum samples obtained from 2007 through 2010. Prevalence of bovine TB was estimated by use of 3 S/N cutoff values (2.5, 3.0, and 3.5) to provide information on the prevalence over time (Figure 3). Prevalence estimates indicated that bovine TB in wild Minnesota deer peaked in approximately 2009 (all 3 cutoff values) and decreased thereafter to almost undetectable levels (the most conservative cutoff value of 3.5). However, use of a cutoff value of 2.5 or 3.0 indicated that bovine TB appeared to be present, albeit at an extremely low level, in Minnesota deer.
Overall TB prevalence in sera of Minnesota DNR hunter-harvested deer by year for each biomarker determined by use of 3 S/N cutoff values (2.5 [circles], 3.0 [squares], and 3.5 [diamonds]).
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
Overall TB prevalence in sera of Minnesota DNR hunter-harvested deer by year for each biomarker determined by use of 3 S/N cutoff values (2.5 [circles], 3.0 [squares], and 3.5 [diamonds]).
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
Overall TB prevalence in sera of Minnesota DNR hunter-harvested deer by year for each biomarker determined by use of 3 S/N cutoff values (2.5 [circles], 3.0 [squares], and 3.5 [diamonds]).
Citation: American Journal of Veterinary Research 78, 6; 10.2460/ajvr.78.6.729
Discussion
Bovine TB is a zoonotic disease associated with devastating consequences for agriculture and public health. Mycobacterium bovis threatens domestic animals, humans, and the economy via a number of pathways. First, wildlife reservoirs of bovine TB have directly caused an increase in the incidence of TB in cattle in recent years, which resulted in loss of animals as well as trade dollars. Second, TB is a major zoonotic concern in rural areas in developing countries in which the animal-human interface is intensifying as patterns of land use change.
Existing strategies for the detection of TB in imported animals and local wildlife (in vivo tuberculin skin test coupled with assay of in vitro interferon-γ release) are inadequate to tackle a problem of this magnitude. Proposed strategies are for detection of host immune responses or identification of intact bacteria in lesions.
Existing strategies are rarely practical for implementation with free-ranging wildlife populations or specimens. Tests have poor performance metrics (eg, low sensitivity) and are labor-intensive, invasive, expensive, and unavailable at critical locations, specifically remote farming areas in the United States and developing countries where bovine TB is widespread. In addition, test results provide no information about the stage of disease.
In the present study, 3 pathogen-specific peptides were evaluated by the use of sera from experimentally infected deer and free-ranging deer. This appeared to be a practical sampling strategy for monitoring disease in free-ranging wildlife and would be sufficient to rapidly identify and allow a response to emerging outbreaks of TB in wildlife. Results indicated that pks5, MB2515c, and MB1895c can be used to detect infection in samples from TB-positive and exposed deer and that they can be used to distinguish between uninfected and infected animals. The most reliable pathogen biomarkers for bovine TB in the present study were pks5 and MB2515c. In another study,12 investigators found that pks5 was the most reliable biomarker for bovine TB. It was not surprising that the S/N values for experimentally infected deer were lower than those for free-ranging deer because the time frame in which the samples from the experimentally infected deer were obtained was extremely short (60 days), but samples of hunter-harvested deer were obtained from adult deer in which the response to infection potentially had a longer time to develop.
In the present study, the combination of pks5 and MB2515c was an optimal indicator of infection (or its absence). Simulation of the bovine TB infection status by use of a range of S/N values indicated that although bovine TB in Minnesota deer decreased after 2009, it may have persisted at extremely low levels as subclinical disease.
These findings confirmed that the calculated prevalence of disease differed depending on the S/N cutoff value. The cutoff value would differ on the basis of the areas in which disease surveillance is performed (eg, regions with endemic infection vs low-prevalence regions). Areas with endemic infection require tests with high sensitivity or low cutoff values to maximize accurate detection and minimize potential spillover to other domestic animals and wildlife. Thus, cutoff values would need to be optimized on the basis of the population prevalence. In areas with a low prevalence of bovine TB (such as Minnesota), a test with high specificity or a high cutoff value would be optimal to ensure that disease-free animals are not misclassified as test-positive animals. A region with a low disease prevalence (such as the United States) could greatly benefit from a diagnostic test that is highly specific, which should prevent misclassification resulting from animal exposure to other mycobacteria such as Mycobacterium avium subsp paratuberculosis. On the other hand, regions in Africa and Asia with a high disease prevalence need diagnostic tests with high sensitivity to identify all infected animals and minimize potential transmission to humans and other animals.
Existing diagnostic tests for bovine TB are inadequate for the purpose of disease surveillance and monitoring of free-ranging wildlife. Current methods for diagnosis of bovine TB rely on direct detection of the pathogen (through microscopy or culture) or on DNA amplification.19 Infection is determined by the presence of sustained T-cell reactivity with antigens of the Mycobacterium tuberculosis complex (tuberculin skin tests or assays of interferon-γ release in blood). These tests, although useful, do not provide an optimal foundation for control of bovine TB or a clear path to the development of improved strategies.19 Bacterial culture (the criterion-referenced standard for bovine TB testing) is time-consuming, which results in delays for obtaining results.
Critical challenges in diagnostic testing to detect disease caused by organisms of the M tuberculosis complex include a lack of effective tools to provide timely and accurate results. Research in TB diagnostic testing has yielded promising results. However, there is still a scarcity of tools for the diagnosis of TB in animals. We speculate that the direct detection of lesions attributable to infection with M tuberculosis or M bovis or their products (eg, peptides and DNA) in blood and urine by use of high-resolution de novo methods has better sensitivity than does bacterial culture of specimens, but the usefulness is still well below that required for a diagnostic test.15,20 Lipoarabinomannan has been extensively studied, and a commercially available test to detect lipoarabinomannan in urine is available, although its low sensitivity has limited its use.19 Detection of antibodies in serum and urine against several TB proteins including antigen 85 complex has shown promising results.19,20 Mycobacterium tuberculosis complex-specific antibody responses have been evaluated, but further optimization is needed because of the heterogeneity of the antibody response against the M tuberculosis complex.19
A highly sensitive and specific diagnostic test designed to evaluate disease progression and response to treatment would provide the ultimate diagnostic test for TB. Investigators of a recent study19 indicated target product profiles a biomarker should meet to provide the greatest impact as an ideal TB diagnostic test. Although the emphasis of that study19 was on TB in humans, the criteria remain the same for bovine TB, a zoonotic disease. An ideal biomarker-based assay should have high (> 98%) sensitivity and specificity, be noninvasive or minimally invasive, provide results rapidly, and be affordable. Investigators of 1 study21 indicated that a possible combination of biomarkers will confer diagnostic value for different settings (eg, one set of markers for differentiating between active and latent TB infections in humans or subclinical infections in other animals, and another set of markers to differentiate TB from other diseases). However, this would require collaborative efforts on the part of researchers on TB in humans and other animals (especially cattle).
Current methods for detection of bovine TB in animals are inadequate, and control programs cannot rely solely on test-and-slaughter of reactor animals. Methods need to be supplemented with results of comprehensive epidemiological investigations of outbreaks of bovine TB and surveillance at abattoirs, implementation of movement restrictions of animals, and use of biomarker technology.16,17,22,23 There is a need for a highly sensitive and specific diagnostic test for TB that can rapidly detect latent or subclinical disease in animals for various disease prevalences at a reasonable cost on an easily obtainable sample such as blood, urine, or exhaled breath.
In the present study, a biomarker detection assay was used to detect subclinical as well clinical infection in white-tailed deer and to differentiate between infected and noninfected experimentally inoculated deer. Information about new biomarkers is commonly published, but refinement, validation, and independent confirmation of such biomarkers often is not accomplished.19 Three biomarkers (pks5, MB2515c, and MB1895c) were validated for use in detection of TB infection of different host species (deer and cattle), with promising results. Application of TB-specific biomarkers in diagnostic tests would greatly enhance the ability to screen for this mycobacterial disease that traditionally has been difficult to monitor. Biomarkers for the M tuberculosis complex appeared to be valid indicators of TB infection, and we propose that they be incorporated into a point-of-care device for rapid screening for TB. Evaluation of highly specific serum biomarkers composed of mycobacterial peptides and proteins may provide insights into disease progression and interspecies transmission of M bovis infections between humans, cattle, and deer.
Acknowledgments
Supported by a Morris Animal Foundation grant (D14ZO-086) and a University of Minnesota Grand Challenges grant awarded to Dr. Sreevatsan. Dr. Wanzala was an MNDrive fellow.
ABBREVIATIONS
| DNR | Department of Natural Resources |
| OD | Optical density |
| pks5 | Polyketide synthase 5 |
| S/N | Signal-to-noise ratio |
| TB | Tuberculosis |
| TBS | Tris-buffered saline |
Footnotes
Sigma-Aldrich, St Louis, Mo.
Nunc-MaxiSorp, Thermo Fisher Scientific Inc, Waltham, Mass.
Blocker Blotto, Pierce, Rockford, Ill.
PBS Tween, Thermo Fisher Scientific Inc, Carlsbad, Calif.
G-BioSciences, St Louis, Mo.
Santa Cruz Biotechnology Inc, Dallas, Tex.
1-Step Ultra tetramethylbenzidine, Pierce, Rockford, Ill.
SpectraMax M2, Molecular Devices LLC, Sunnyvale, Calif.
Excel, Microsoft Corp, Redmond, Wash.
GraphPad Prism 6, GraphPad Software Inc, La Jolla, Calif.
References
1. Hingley-Wilson SM, Sambandamurthy VK, Jacobs WR Jr. Survival perspectives from the world's most successful pathogen, Mycobacterium tuberculosis. Nat Immunol 2003; 4: 949–955.
2. Schmitt SM, Fitzgerald SD, Cooley TM, et al. Bovine tuberculosis in free-ranging white-tailed deer from Michigan. J Wildl Dis 1997; 33: 749–758.
3. O'Brien DJ, Schmitt SM, Fierke JS, et al. Epidemiology of Mycobacterium bovis in free-ranging white-tailed deer, Michigan, USA, 1995–2000. Prev Vet Med 2002; 54: 47–63.
4. O'Brien DJ, Schmitt SM, Fitzgerald SD, et al. Managing the wildlife reservoir of Mycobacterium bovis: the Michigan, USA, experience. Vet Microbiol 2006; 112: 313–323.
5. Schmitt SM, O'Brien DJ, Bruning-Fann CS, et al. Bovine tuberculosis in Michigan wildlife and livestock. Ann N Y Acad Sci 2002; 969: 262–268.
6. O'Reilly LM, Daborn CJ. The epidemiology of Mycobacterium bovis infections in animals and man: a review. Tuber Lung Dis 1995; 76(suppl 1):1–46.
7. Wilkins MJ, Bartlett PC, Frawley B, et al. Mycobacterium bovis (bovine TB) exposure as a recreational risk for hunters: results of a Michigan hunter survey, 2001. Int J Tuberc Lung Dis 2003; 7: 1001–1009.
8. Wilkins MJ, Meyerson J, Bartlett PC, et al. Human Mycobacterium bovis infection and bovine tuberculosis outbreak, Michigan, 1994–2007. Emerg Infect Dis 2008; 14: 657–660.
9. Morris RS, Pfeiffer DU, Jackson R. The epidemiology of Mycobacterium bovis infections. Vet Microbiol 1994; 40: 153–177.
10. Corner LA. The role of wild animal populations in the epidemiology of tuberculosis in domestic animals: how to assess the risk. Vet Microbiol 2006; 112: 303–312.
11. USDA APHIS. Bovine tuberculosis disease information. Available at: www.aphis.usda.gov/aphis/ourfocus/animalhealth/animal-disease-information/cattle-disease-information/sa_tuberculosis/ct_bovine_tuberculosis_disease_information. Accessed Feb 22, 2016.
12. Lamont EA, Janagama HK, Ribeiro-Lima J, et al. Circulating Mycobacterium bovis peptides and host response proteins as biomarkers for unambiguous detection of subclinical infection. J Clin Microbiol 2014; 52: 536–543.
13. Seth M, Lamont EA, Janagama HK, et al. Biomarker discovery in subclinical mycobacterial infections of cattle. PLoS ONE 2009; 4:e5478.
14. Carstensen M, O'Brien DJ, Schmitt SM. Public acceptance as a determinant of management strategies for bovine tuberculosis in free-ranging US wildlife. Vet Microbiol 2011; 151: 200–204.
15. O'Brien DJ, Schmitt SM, Berry DE, et al. Estimating the true prevalence of Mycobacterium bovis in free-ranging elk in Michigan. J Wildl Dis 2008; 44: 802–810.
16. Palmer MV, Whipple DL, Waters WR. Experimental deer-to-deer transmission of Mycobacterium bovis. Am J Vet Res 2001; 62: 692–696.
17. Carstensen M, DonCarlos MW. Preventing the establishment of a wildlife disease reservoir: a case study of bovine tuberculosis in wild deer in Minnesota, USA. Vet Med Int 2011; 2011: 413240.
18. Glaser L, Carstensen M, Shaw S, et al. Descriptive epidemiology and whole genome sequencing analysis for an outbreak of bovine tuberculosis in beef cattle and white-tailed deer in northwestern Minnesota. PLoS ONE 2016; 11: e0145735.
19. Gardiner JL, Karp CL. Transformative tools for tackling tuberculosis. J Exp Med 2015; 212: 1759–1769.
20. Young BL, Mlamla Z, Gqamana PP, et al. The identification of tuberculosis biomarkers in human urine samples. Eur Respir J 2014; 43: 1719–1729.
21. Haas CT, Roe JK, Pollara G, et al. Diagnostic ‘omics’ for active tuberculosis. BMC Med 2016; 14: 37.
22. Vordermeier HM, de Val BP, Buddle BM, et al. Vaccination of domestic animals against tuberculosis: review of progress and contributions to the field of the TBSTEP project. Res Vet Sci 2014; 97: S53–S60.
23. Vordermeier HM, Jones GJ, Buddle BM, et al. Bovine tuberculosis in cattle: vaccines, DIVA tests, and host biomarker discovery. Annu Rev Anim Biosci 2016; 4: 87–109.
