Mycobacterium species cause various diseases in domestic animals, such as cattle, pigs, sheep, horses, goats, and poultry (Malone and Gordon, Reference Malone and Gordon2017; McGill et al., Reference McGill, Saunders, Eastwood, Menache, Dalzell, Hill, Irving, Knight and Jones2018). Mycobacterial disease is generally characterized by a long incubation period and recurrence of disease depending on the immunological status of the host (Drain et al., Reference Drain, Bajema, Dowdy, Dheda, Naidoo, Schumacher, Ma, Meermeier, Lewinsohn and Sherman2018). Diagnosis of mycobacterial infection may be difficult, dependent on the stage of disease and the interpretation of the test (Britton et al., Reference Britton, Cassidy, O'Donovan, Gordon and Markey2016; Correia-Neves et al., Reference Correia-Neves, Fröberg, Korshun, Viegas, Vaz, Ramanlal, Bruchfeld, Hamasur, Brennan and Källenius2019; van den Esker and Koets, Reference van den Esker and Koets2019). For example, diagnosis can be difficult early in infection, and the tests may detect presence or exposure but not actually accurately determine disease outcome, as many animals may remain infected but exhibit no clinical signs for their lifetime (Stabel, Reference Stabel1998). Diagnosis is easier once the infection advances into a clinical stage. The detection of mycobacterial pathogens through bacterial culture, polymerase chain reaction (PCR), and real-time PCR in suspected animals can be used for the diagnosis of mycobacterial disease (Mehta et al., Reference Mehta, Raj, Singh and Khuller2014; Tucci et al., Reference Tucci, González-Sapienza and Marin2014; Leung et al., Reference Leung, Siu, Tam, Ho, Wong, Leung, Yu, Ma and Yam2018; Michelet et al., Reference Michelet, de Cruz, Karoui, Tambosco, Moyen, Hénault and Boschiroli2018). Additionally, diagnostic methods based on host responses, including interferon gamma (IFN-γ) assay, intradermal skin test, and serum enzyme-linked immunoassay (ELISA), can be used for the diagnosis of mycobacterial disease (Sockett et al., Reference Sockett, Conrad, Thomas and Collins1992; Britton et al., Reference Britton, Cassidy, O'Donovan, Gordon and Markey2016; Bernitz et al., Reference Bernitz, Clarke, Roos, Goosen, Cooper, van Helden, Parsons and Miller2018). Although the current diagnostic tests can detect mycobacterial infection, they had low sensitivity and specificity to subclinical or latent infection. Therefore, specific biomarkers, such as the expression of genes, proteins, metabolites, and microRNAs, have been characterized for understanding the pathogenesis and detection of mycobacterial disease in the early stage of infection (Kabeer et al., Reference Kabeer, Raja, Raman, Thangaraj, Leportier, Ippolito, Girardi, Lagrange and Goletti2011; Anderson et al., Reference Anderson, Kaforou, Brent, Wright, Banwell, Chagaluka, Crampin, Dockrell, French, Hamilton, Hibberd, Kern, Langford, Ling, Mlotha, Ottenhoff, Pienaar, Pillay, Scott, Twahir, Wilkinson, Coin, Heyderman, Levin and Eley2014; Iannaccone et al., Reference Iannaccone, Cosenza, Pauciullo, Garofalo, Proroga, Capuano and Capparelli2018).
Biomarkers are measurable biological indicators, such as nucleic acids (DNA, mRNA, microRNA, and small noncoding RNA), metabolites, and proteins, within biological fluids that reflect the physiological status and changes in healthy and/or diseased individuals (Strimbu and Tavel, Reference Strimbu and Tavel2010). Biomarkers can be derived from hosts or pathogens in infectious diseases (Tucci et al., Reference Tucci, González-Sapienza and Marin2014; Walzl et al., Reference Walzl, McNerney, du Plessis, Bates, McHugh, Chegou and Zumla2018). Because biomarker presence differs with the occurrence and progression of a specific disease, its abundance in blood or body fluids can be used as an indicator that enables measurement of physiological changes during disease progression (Mayeux, Reference Mayeux2004). Therefore, biomarkers can be used for various purposes, such as in the early diagnosis of disease, evaluation of prognosis, and prediction of response to therapy (Mayeux, Reference Mayeux2004). In detail, biomarkers are considered a useful noninvasive diagnostic tool for various diseases, such as cancers, autoimmune diseases, and infectious diseases, in both humans and animals (Schmidt et al., Reference Schmidt, Plate, Angele, Pfister, Wick, Koedel and Rupprecht2011; Shirmohammadi et al., Reference Shirmohammadi, Sohrabi, Jafarzadeh Samani, Effatpanah, Yadegarazari and Saidijam2018). In human medicine, CXCL13 was reported to increase in the cerebrospinal fluid (CSF) of patients with acute Lyme neuroborreliosis, and diagnosis based on the level of CXCL13 in CSF has been shown to have high sensitivity and specificity for the early diagnosis of atypical neuroborreliosis (Schmidt et al., Reference Schmidt, Plate, Angele, Pfister, Wick, Koedel and Rupprecht2011). Additionally, Hutchinson and colleagues suggested that the ratio and absolute number of heparin-binding hemagglutinin-specific CD4T cells was significantly higher in active tuberculosis (TB) patients than in latent TB patients (Hutchinson et al., Reference Hutchinson, Barkham, Tang, Kemeny, Chee and Wang2015). Furthermore, a rapid urine lipoarabinomannan (LAM) test based on an Mycobacterium tuberculosis (MTB)-specific pathogen-derived biomarker has been used for the diagnosis of HIV-associated TB patients (Lawn, Reference Lawn2012; World Health Organization, 2015).
In veterinary medicine, many studies have been performed to identify biomarkers that reflect physiological changes during disease progression (Foley et al., Reference Foley, Chapwanya, Callanan, Whiston, Miranda-CasoLuengo, Lu, Meijer, Lynn, O' Farrelly and Meade2015; Adnane et al., Reference Adnane, Chapwanya, Kaidi, Meade and O'Farrelly2017, Reference Adnane, Kelly, Chapwanya, Meade and O'Farrelly2018). Several biomarkers have been proposed for the prediction of clinical postpartum endometritis in cattle (Foley et al., Reference Foley, Chapwanya, Callanan, Whiston, Miranda-CasoLuengo, Lu, Meijer, Lynn, O' Farrelly and Meade2015; Adnane et al., Reference Adnane, Chapwanya, Kaidi, Meade and O'Farrelly2017, Reference Adnane, Kelly, Chapwanya, Meade and O'Farrelly2018). For example, the levels of interleukin (IL)-1β, IL-6, IL-8, serum amyloid A, and haptoglobin in cervicovaginal mucus were higher in clinical endometritis-affected cattle than in healthy cattle (Adnane et al., Reference Adnane, Chapwanya, Kaidi, Meade and O'Farrelly2017). Additionally, α1-acid glycoprotein was increased in the cervicovaginal mucus of clinical endometritis-affected cattle compared to healthy cattle (Adnane et al., Reference Adnane, Kelly, Chapwanya, Meade and O'Farrelly2018). In addition, inflammatory responses characterized by the production of IL-1 and IL-17 were activated only in the early postpartum stage and then returned to normal levels in healthy cattle, while the cytokines in endometritis-affected cattle persisted (Foley et al., Reference Foley, Chapwanya, Callanan, Whiston, Miranda-CasoLuengo, Lu, Meijer, Lynn, O' Farrelly and Meade2015). Upregulation of four cytokines (IL-1, IL-6, IFN-γ, and tumor necrosis factor (TNF)-α) was observed in subclinical infection of Mycobacterium avium subsp. paratuberculosis (MAP). Additionally, upregulation of three cytokines (transforming growth factor (TGF)-β, IL-10, and Traf-1) and downregulation of eight cytokines (IL-18, IL-17, IL-1 α/β, TNF-α, granulysin, RANTES, MCP-1, and IFN-γ) were observed in clinical infection of MAP (Wright et al., Reference Wright, Plain, Purdie, Saunders and de Silva2019). Furthermore, upregulation of four cytokines (IFN-γ, IL-17A, iNOS, and TNF-α) and downregulation of IL-10 were confirmed in latent bovine tuberculosis (bTB) (Wright et al., Reference Wright, Plain, Purdie, Saunders and de Silva2019). In contrast, TGF-β and IL-10 were increased in active bTB-affected animals (Wright et al., Reference Wright, Plain, Purdie, Saunders and de Silva2019). Recent studies have shown that a high IFN-γ response in young age can be used as a biomarker for the resilience of ovine paratuberculosis (de Silva et al., Reference de Silva, Begg, Plain, Purdie, Kawaji, Dhand and Whittington2013, Reference de Silva, Plain, Purdie, Begg and Whittington2018).
With the development of next-generation sequencing and bioinformatics, biomarkers have become widely used in the diagnosis of various diseases. To date, biomarkers have been used not only in human medicine but also in veterinary medicine, although to a lesser extent. In this review, we describe the current progress of research investigating diagnostic biomarkers of mycobacterial diseases in cattle.
Biomarkers of mycobacterial infections in cattle
Paratuberculosis
Paratuberculosis is a chronic enteric disease of ruminants that is caused by MAP. MAP infection causes granulomatous intestinal lesions and the formation of corrugation, which induce persistent diarrhea in the terminal stage of the disease (Sweeney, Reference Sweeney2011). MAP can endure under harsh conditions, such as low pH, high organic acidity, and low temperature (Larsen et al., Reference Larsen, Merkal and Vardaman1956; Jorgensen, Reference Jorgensen1977; Cook et al., Reference Cook, Flis and Ballard2013). Therefore, MAP can survive in such environments as water, soil, and grass for periods up to several years (Whittington et al., Reference Whittington, Marshall, Nicholls, Marsh and Reddacliff2004; Samba-Louaka et al., Reference Samba-Louaka, Robino, Cochard, Branger, Delafont, Aucher, Wambeke, Bannantine, Biet and Héchard2018). This property of MAP enables the circulation of the disease in an endemically infected herd after the first occurrence (Smith et al., Reference Smith, Schukken, Pradhan, Smith, Whitlock, Van Kessel, Wolfgang and Grohn2011; Fecteau et al., Reference Fecteau, Hovingh, Whitlock and Sweeney2013). The transmission of paratuberculosis mainly depends on the ingestion of contaminated feed, water, colostrum, and milk (Whitlock and Buergelt, Reference Whitlock and Buergelt1996). In general, MAP-infected cattle will undergo a subclinical stage that does not present symptoms for a long period, and during that subclinical stage, infected individuals discharge MAP through feces, resulting in continuous circulation of the bacteria in the environment and spread of the disease in the herd (Stabel, Reference Stabel1998). MAP has also been proven to be transmissible in utero and via semen (van Kooten et al., Reference van Kooten, Mackintosh and Koets2006). The isolation of MAP from clinical samples, such as feces, tissue, milk, and blood, has been considered the gold standard for the diagnosis of paratuberculosis (Collins, Reference Collins2011). Additionally, PCR assays have been developed to detect MAP in feces, and ELISA has been used to detect antibodies against MAP in the host animal (Arango-Sabogal et al., Reference Arango-Sabogal, Labrecque, Paré, Fairbrother, Roy, Wellemans and Fecteau2016; Gilardoni et al., Reference Gilardoni, Fernández, Morsella, Mendez, Jar, Paolicchi and Mundo2016; de Kruijf et al., Reference de Kruijf, Govender, Yearsley, Coffey and O'Mahony2017; Mathevon et al., Reference Mathevon, Foucras, Falguières and Corbiere2017). However, those methods have low sensitivity and specificity for the detection of MAP in subclinical infection (Sockett et al., Reference Sockett, Conrad, Thomas and Collins1992; Cousins et al., Reference Cousins, Whittington, Marsh, Masters, Evans and Kluver1999; Englund et al., Reference Englund, Bölske and Johansson2002; Bögli-Stuber et al., Reference Bögli-Stuber, Kohler, Seitert, Glanemann, Antognoli, Salman, Wittenbrink, Wittwer, Wassenaar, Jemmi and Bissig-Choisat2005; Nielsen and Toft, Reference Nielsen and Toft2008; Collins, Reference Collins2011).
In recent years, studies on host- or pathogen-related biomarker-based diagnostic methods have been conducted to overcome such problems as low sensitivity and specificity. Various diagnostic pathogen-derived biomarkers have been identified for the development of diagnostic methods for the detection of early stages of infection (Eda et al., Reference Eda, Bannantine, Waters, Mori, Whitlock, Scott and Speer2006; Shin et al., Reference Shin, Cho and Collins2008; Nagata et al., Reference Nagata, Kawaji and Mori2013; Li et al., Reference Li, Wagner, Freer, Schilling, Bannantine, Campo, Katani, Grohn, Radzio-Basu and Kapur2017, Reference Li, Bannantine, Campo, Randall, Grohn, Schilling, Katani, Radzio-Basu, Easterling and Kapur2019). Eda and colleagues developed ELISA based on the surface antigen of MAP for the diagnosis of paratuberculosis. ELISA based on surface antigen shows 97.4% specificity and 100% sensitivity with 38 JD-free and 51 fecal culture-positive serum samples (Eda et al., Reference Eda, Bannantine, Waters, Mori, Whitlock, Scott and Speer2006). Additionally, 96.6% of the low-level fecal shedders and 100% of the mid- and high-level shedders were detected by ELISA (Eda et al., Reference Eda, Bannantine, Waters, Mori, Whitlock, Scott and Speer2006). Shin et al. (Reference Shin, Cho and Collins2008) developed ELISA based on culture filtrate antigens and compared diagnostic sensitivity and specificity with five commercial ELISA kits (Shin et al., Reference Shin, Cho and Collins2008). ELISA had 56.3% sensitivity and 99% specificity, while all commercial ELISA kits had low sensitivity below 45% evaluated with serum samples of 444 paratuberculosis cases and 412 controls (Shin et al., Reference Shin, Cho and Collins2008). Additionally, Facciuolo and colleagues discovered novel secreted antigens of MAP as diagnostic biomarkers (Facciuolo et al., Reference Facciuolo, Kelton and Mutharia2013). These researchers identified four novel MAP antigens that reacted with 35 MAP-infected cattle serum samples, and these four novel antigens were reacted with six low MAP shedders and three fecal culture-positive and ELISA-negative serum samples (Facciuolo et al., Reference Facciuolo, Kelton and Mutharia2013). Recently, several researchers have focused on the development of diagnostic methods based on specific antigens. Nagata and colleagues developed ELISA based on enoyl coenzyme A hydratase (echA) of MAP (Nagata et al., Reference Nagata, Kawaji and Mori2013). MAP-echA ELISA had high specificity with immunized sera against five species of Mycobacterium species, while the Pourquier ELISA kit had low specificity, and the Map-echA ELISA had 96.7% sensitivity and 96.7% specificity with 30 MAP-positive serum samples and 30 MAP-negative serum samples (Nagata et al., Reference Nagata, Kawaji and Mori2013). Furthermore, the Map-echA ELISA had responses 2–7 months earlier than the Pourquier ELISA kit in all experimentally infected cattle (Nagata et al., Reference Nagata, Kawaji and Mori2013). Li and colleagues developed a multiplex immunoassay based on color-coded fluorescent beads coupled to six MAP antigens and evaluated diagnostic performance with 180 serum and 90 milk samples from cows at different stages of MAP infection. Multiplex immunoassays had higher sensitivity and specificity than commercial ELISA kits when using a combination of four antigens in both serum and milk samples (Li et al., Reference Li, Wagner, Freer, Schilling, Bannantine, Campo, Katani, Grohn, Radzio-Basu and Kapur2017). Additionally, three antigens (MAP1569, MAP2942c, and MAP2609) significantly discriminate fecal PCR-positive and ELISA-negative animals from uninfected animals, indicating high diagnostic value for detecting the early stage of MAP infection (Li et al., Reference Li, Wagner, Freer, Schilling, Bannantine, Campo, Katani, Grohn, Radzio-Basu and Kapur2017). Furthermore, these researchers constructed a MAP protein array with 868 purified recombinant MAP proteins and tested 180 serum samples from cows classified into four groups based on previous serum ELISA and fecal PCR tests (Li et al., Reference Li, Bannantine, Campo, Randall, Grohn, Schilling, Katani, Radzio-Basu, Easterling and Kapur2019). A total of 49 candidate antigens were identified in the different stages of infection groups compared with the uninfected group. These researchers evaluated the diagnostic potential of MAP antigens with combinations of the top reactive antigens in each group, and the sensitivities were 60.0, 73.3, and 81.7% in the negative high-exposure group, fecal-positive and ELISA-negative group, and fecal-positive and ELISA-positive group, respectively, with 90% specificity indicating the possibility of detection for the early stage of MAP infection (Li et al., Reference Li, Bannantine, Campo, Randall, Grohn, Schilling, Katani, Radzio-Basu, Easterling and Kapur2019).
Volatile organic compounds (VOCs) have been used as diagnostic biomarkers of various diseases, including cancer, inflammatory bowel disease, and infectious diseases (Kasbohm et al., Reference Kasbohm, Fischer, Küntzel, Oertel, Bergmann, Trefz, Miekisch, Schubert, Reinhold, Ziller, Fröhlich, Liebscher and Köhler2017; Lewis et al., Reference Lewis, Savage, Beeching, Beadsworth, Feasey and Covington2017; Monasta et al., Reference Monasta, Pierobon, Princivalle, Martelossi, Marcuzzi, Pasini and Perbellini2017; Tong et al., Reference Tong, Wang, Li, Liu, Chi, Liu, Guo, Li and Wang2017; Traxler et al., Reference Traxler, Bischoff, Saß, Trefz, Gierschner, Brock, Schwaiger, Karte, Blohm, Schröder, Miekisch and Schubert2018). Several researchers have focused on VOCs produced by MAP itself during in vitro growth and VOCs produced by host animals when infected with MAP (Trefz et al., Reference Trefz, Koehler, Klepik, Moebius, Reinhold, Schubert and Miekisch2013; Küntzel et al., Reference Küntzel, Fischer, Bergmann, Oertel, Steffens, Trefz, Miekisch, Schubert, Reinhold and Köhler2016; Kasbohm et al., Reference Kasbohm, Fischer, Küntzel, Oertel, Bergmann, Trefz, Miekisch, Schubert, Reinhold, Ziller, Fröhlich, Liebscher and Köhler2017). Various volatile markers have been discovered that are released by MAP in bacterial cultures (Trefz et al., Reference Trefz, Koehler, Klepik, Moebius, Reinhold, Schubert and Miekisch2013; Küntzel et al., Reference Küntzel, Fischer, Bergmann, Oertel, Steffens, Trefz, Miekisch, Schubert, Reinhold and Köhler2016). Trefz et al. (Reference Trefz, Koehler, Klepik, Moebius, Reinhold, Schubert and Miekisch2013) found that 34 VOCs were related to MAP growth and that the profile of VOCs between reference strain and field strains was significantly distinct. Additionally, the most abundant VOCs during the growth of MAP were furans, which are related to bacterial signaling in response to low-density growth (Trefz et al., Reference Trefz, Koehler, Klepik, Moebius, Reinhold, Schubert and Miekisch2013). Furthermore, Küntzel et al. (Reference Küntzel, Fischer, Bergmann, Oertel, Steffens, Trefz, Miekisch, Schubert, Reinhold and Köhler2016) described 43 VOCs related to MAP growth and revealed different volatile profiles that depended on bacterial density. In contrast to the previous study, the bacterial strain is not a crucial factor in VOC profiles, while the type of culture media or incubation period is crucial for changing VOC profiles (Küntzel et al., Reference Küntzel, Fischer, Bergmann, Oertel, Steffens, Trefz, Miekisch, Schubert, Reinhold and Köhler2016). Kasbohm et al. (Reference Kasbohm, Fischer, Küntzel, Oertel, Bergmann, Trefz, Miekisch, Schubert, Reinhold, Ziller, Fröhlich, Liebscher and Köhler2017) proposed a workflow based on VOC profiles for the diagnosis of MAP infection and compared diagnostic sensitivity and specificity with current diagnostic methods, such as bacterial culture, ELISA, and IFN-γ assay. A diagnostic method based on the combination of multiple VOCs with random forest classification showed the highest sensitivity in both breath and feces but showed low specificity compared to current diagnostic methods (Kasbohm et al., Reference Kasbohm, Fischer, Küntzel, Oertel, Bergmann, Trefz, Miekisch, Schubert, Reinhold, Ziller, Fröhlich, Liebscher and Köhler2017). Most recently, Maurer and colleagues compared microbial VOC profiles from various pathogenic mycobacterial cultures, including M. tuberculosis, Mycobacterium bovis (Mb), and MAP (Maurer et al., Reference Maurer, Ellis, Thacker, Rice, Koziel, Nol and VerCauteren2019). Although no exclusive VOC biomarkers were identified for MAP culture, differentiation of MAP culture from other mycobacterial species was possible when comparing VOC profiles of all mycobacterial cultures (Maurer et al., Reference Maurer, Ellis, Thacker, Rice, Koziel, Nol and VerCauteren2019). A summary of the identified pathogen-derived biomarkers of bovine paratuberculosis is presented in Table 1.
Identified pathogen biomarkers for bovine paratuberculosis
Current host biomarker-based diagnostic methods can be classified into three types – specific protein assay, transcriptomic analysis, and miRNA analysis – depending on the type of target indicator (Britton et al., Reference Britton, Cassidy, O'Donovan, Gordon and Markey2016). Several studies have been conducted to identify specific proteins that can discriminate MAP-infected animals from uninfected animals (Seth et al., Reference Seth, Lamont, Janagama, Widdel, Vulchanova, Stabel, Waters, Palmer and Sreevatsan2009; Stabel and Robbe-Austerman, Reference Stabel and Robbe-Austerman2011; You et al., Reference You, Verschoor, Pant, Macri, Kirby and Karrow2012; De Buck et al., Reference De Buck, Shaykhutdinov, Barkema and Vogel2014). When sera from infected and uninfected individuals were analyzed through protein fingerprinting, sera from infected cattle showed greater than 2-fold increases in the levels of six proteins and 2-fold decreases in two proteins (You et al., Reference You, Verschoor, Pant, Macri, Kirby and Karrow2012). The authors also observed that the proteins with significantly altered expression in infected cattle were involved in iron regulation, leukocyte/lymphocyte control, apoptosis, coagulation, and complement activation (You et al., Reference You, Verschoor, Pant, Macri, Kirby and Karrow2012). Proteomic analysis of sera from bTB- or paratuberculosis-affected cattle was performed to elucidate the responses of specific proteins to mycobacterial infection (Seth et al., Reference Seth, Lamont, Janagama, Widdel, Vulchanova, Stabel, Waters, Palmer and Sreevatsan2009). The results of this study showed that vitamin D binding protein (VDBP) was increased in both Mb- and MAP-infected cattle, while cathelicidin was increased significantly only in MAP-infected cattle (Seth et al., Reference Seth, Lamont, Janagama, Widdel, Vulchanova, Stabel, Waters, Palmer and Sreevatsan2009). These changes are believed to be related to the activation of the innate immune system, which is thought to be similar to reactions that occur in humans when infected with M. tuberculosis. The increase in transthyretin in MAP-infected cattle, which would increase the transport of vitamin A, resulted in increased differentiation of monocytes and activation of mucosal immunity to inhibit MAP proliferation (Seth et al., Reference Seth, Lamont, Janagama, Widdel, Vulchanova, Stabel, Waters, Palmer and Sreevatsan2009). Infection with MAP provokes a specific host response in the early stage of infection. The detection of early immune markers of MAP infection in an experimental infection model was reported for all infected groups maintaining a strong IFN-γ response during the study (Stabel and Robbe-Austerman, Reference Stabel and Robbe-Austerman2011). Moreover, T cell activation markers, such as CD25, CD26, CD45RO, and CD5, were significantly upregulated in MAP-infected calves compared to those in uninfected controls (Stabel and Robbe-Austerman, Reference Stabel and Robbe-Austerman2011). Comparative analysis of the metabolites in the sera of infected and uninfected cattle showed that concentrations of acetone, citrate, glycerol, and iso-butyrate were significantly altered in infected cattle and were indicative of increased lipid metabolism in infected cattle (De Buck et al., Reference De Buck, Shaykhutdinov, Barkema and Vogel2014). In addition, significant changes in amino acid concentrations were observed in the infected cattle, indicating that protein turnover or deficiencies occur in infected cattle (De Buck et al., Reference De Buck, Shaykhutdinov, Barkema and Vogel2014).
Transcriptomic analysis of host gene expression can be used to diagnose specific diseases (Holcomb et al., Reference Holcomb, Tsalik, Woods and McClain2017). Recently, several studies have been performed to discover biomarkers for the early diagnosis of paratuberculosis based on analyses of host transcriptome profiles (Cha et al., Reference Cha, Yoo, Park, Sung, Shin and Yoo2013; David et al., Reference David, Barkema, Guan le and De Buck2014a, Reference David, Barkema, Mortier, Ghosh, Guan le and De Buckb; Shin et al., Reference Shin, Park, Shin, Jung, Lee, Kim and Yoo2015a, Reference Shin, Park, Shin, Jung, Im, Park, Cho and Yoob; Park et al., Reference Park, Shin, Park, Jung, Cho and Yoo2016; Park et al., Reference Park, Park, Jung and Yoo2017a, Reference Park, Park, Cho, Kim, Jung, Shin, Lee, Kim and Yoob; Park et al., Reference Park, Park, Jung and Yoo2018). An in vitro study showed transcriptomic changes in mouse macrophages during MAP infection and revealed that different gene types were upregulated at 6 h postinfection, such as TNF, IRG1, and TFRC, while some genes, such as COX6A1 and YPEL3, were downregulated at 6 h postinfection (Cha et al., Reference Cha, Yoo, Park, Sung, Shin and Yoo2013). Based on these results, the authors suggested candidate genes that could be used as biomarkers during infection. Other researchers have attempted to identify host biomarkers through the investigation of animal models. Another study reported that gene expression profiles that changed at 3 weeks and 6 weeks after MAP infection in C57BL mice were related to metabolic, intracellular, cell communication, and immune system processes (Shin et al., Reference Shin, Park, Shin, Jung, Lee, Kim and Yoo2015a). Analysis of the gene expression profiles of MAP-infected cattle showed that the production and metabolism of reactive oxygen species are decreased, whereas IL-10 signaling, LXR/RXR signaling, and the complement system are activated during the subclinical infection stage (Shin et al., Reference Shin, Park, Shin, Jung, Im, Park, Cho and Yoo2015b). These results demonstrate the presence of a balanced response that provides an immune-limiting mechanism during the host immune response to MAP infection (Shin et al., Reference Shin, Park, Shin, Jung, Im, Park, Cho and Yoo2015b). Another study showed increased expression of CD46, ICOS, and CEP350 but decreased expression of CTLA4, YARS, and PARVB in whole blood of MAP-infected calves (David et al., Reference David, Barkema, Guan le and De Buck2014a). In addition, a comparison of seropositive and seronegative-infected calves confirmed that IL6ST, GP130, and CD22 have an important role in inducing the production of antibodies to MAP (David et al., Reference David, Barkema, Guan le and De Buck2014a). Additionally, gene expression analysis at 6 months after MAP infection confirmed downregulation of neutrophil beta-defensin-9 peptide (BNBD9-like), S100 calcium binding protein A9 (S100A9), G protein coupling receptor 77 (GPR77)/C5a anaphylatoxin chemotactic receptor (C5a2) and BOLA/MHC-I and upregulation of CD46 at 3, 6, 9, 12, and 15 months after experimental infection of MAP (David et al., Reference David, Barkema, Mortier, Ghosh, Guan le and De Buck2014b). Furthermore, studies to develop a diagnostic method for paratuberculosis based on the specific gene expression profiles in the host are being conducted. The expression of eight genes (TIMP1, HP, SERPINE1, TFRC, MMP9, DEFB1, DEFB10, and S100A8) was significantly increased in whole blood of MAP-infected cattle, and the diagnostic value of these biomarkers was evaluated (Park et al., Reference Park, Park, Jung and Yoo2017a). As a result, four genes (TIMP1, S100A8, DEFB1, and DEFB10) were shown to have the highest diagnostic accuracy in the subclinical MAP-infected group (Park et al., Reference Park, Park, Jung and Yoo2017a). Another study revealed the diagnostic potential of biomarkers for the diagnosis of early-stage MAP infection in cows and offspring that exhibit tissue that is MAP PCR-positive but feces that are MAP PCR-negative (Park et al., Reference Park, Park, Cho, Kim, Jung, Shin, Lee, Kim and Yoo2017b). Park et al. (Reference Park, Park, Jung and Yoo2018) showed that the host response occurring during the subclinical stage enhances MAP survival. Specifically, there was downregulation of Th17 cytokine genes and upregulation of PIP5K1C observed in whole blood of subclinical fecal shedders, which was subsequently related to the loss of granuloma integrity and upregulation of IRF5 and IRF7 and indicated the exhaustion of tryptophan metabolism, which induces inhibition of T cell proliferation (Park et al., Reference Park, Park, Jung and Yoo2018). Most recently, Alonso-Hearn and colleagues identified gene expression profiles from peripheral blood and ileocecal valves of MAP-infected cattle with RNA-seq (Alonso-Hearn et al., Reference Alonso-Hearn, Canive, Blanco-Vazquez, Torremocha, Balseiro, Amado, Varela-Martinez, Ramos, Jugo and Casais2019). Two genes (oxidized low-density lipoprotein receptor and tweety family member 2) were upregulated in MAP-infected animals with focal and diffuse lesions, while three genes, such as glycogen phosphorylase, insulin growth factor 2, and C–C motif chemokine ligand 14, were downregulated (Alonso-Hearn et al., Reference Alonso-Hearn, Canive, Blanco-Vazquez, Torremocha, Balseiro, Amado, Varela-Martinez, Ramos, Jugo and Casais2019).
Recently, studies involving microRNA-based diagnostic methods have been performed to investigate the diagnosis of early-stage paratuberculosis. Seven miRNAs (bta-mir-19b, bta-mir-19-2, bta-mir-1271, bta-mir-100, bta-mir-301a, bta-mir-32, and Novel:14_7917) were downregulated in MAP-infected animals compared to the levels in unexposed animals, whereas the expression levels of five and three miRNAs were decreased and increased, respectively, in the exposed group compared to the unexposed group. Six of the differentially expressed miRNAs were associated with the immune response, and two were novel miRNAs. These results suggest that miRNA expression is influenced by MAP infection and that such expression has a key role in regulating the host response to MAP infection (Malvisi et al., Reference Malvisi, Palazzo, Morandi, Lazzari, Williams, Pagnacco and Minozzi2016). In addition, Farrell et al. (Reference Farrell, Shaughnessy, Britton, MacHugh, Markey and Gordon2015) identified numerous novel miRNAs in bovine sera and showed the utility of RNA sequencing approaches when exploring the potential of miRNAs as novel diagnostic biomarkers for paratuberculosis in cattle (Farrell et al., Reference Farrell, Shaughnessy, Britton, MacHugh, Markey and Gordon2015). A summary of the identified host biomarkers of bovine paratuberculosis is presented in Table 2.
Identified host biomarkers for bovine paratuberculosis
Bovine tuberculosis
bTB, which is predominantly caused by Mb infection, occurs worldwide and causes considerable economic losses in the dairy industry (Caminiti et al., Reference Caminiti, Pelone, LaTorre, De Giusti, Saulle, Mannocci, Sala, Della Marta and Scaramozzino2016). The main route for Mb infection is inhalation of droplets from coughing Mb-infected animals, but infection can also occur by ingesting incompletely pasteurized milk from infected cattle (Michel et al., Reference Michel, Müller and van Helden2010). Since the progression of bTB may take months or years to develop clinical signs, the infected animal can spread bTB to many other animals before clinical signs begin to appear (Ramos et al., Reference Ramos, Silva and Dellagostin2015). Therefore, early detection of an infected animal is important in the control and eradication of the disease from a herd.
The diagnosis of bTB is mainly divided into two categories: first, to detect the pathogen in clinical samples and second, to identify the host immune response by the pathogen (Bernitz et al., Reference Bernitz, Clarke, Roos, Goosen, Cooper, van Helden, Parsons and Miller2018; Fontana et al., Reference Fontana, Pacciarini, Boifava, Pellesi, Casto, Gastaldelli, Koehler, Pozzato, Casalinuovo and Boniotti2018; Michelet et al., Reference Michelet, de Cruz, Karoui, Tambosco, Moyen, Hénault and Boschiroli2018). Detection of the pathogen is mainly performed by Mb-specific PCR and bacterial culture of Mb from clinical samples (Ereqat et al., Reference Ereqat, Nasereddin, Levine, Azmi, Al-Jawabreh, Greenblatt, Abdeen and Bar-Gal2013; Franco et al., Reference Franco, Paes, Ribeiro, de Figueiredo Pantoja, Santos, Miyata, Leite, Motta and Listoni2013). Additionally, even though isolation of Mb is the ‘gold standard’ for diagnosis of bTB, more than 12 weeks are required to achieve isolation of bacteria, and this method is affected by the type of culture media, decontamination method, and incubation time (Ramos et al., Reference Ramos, Silva and Dellagostin2015). Detection of the host immune response mainly depends on the intradermal test and IFN-γ assay (de Lisle et al., Reference de Lisle, Green and Buddle2017; Bernitz et al., Reference Bernitz, Clarke, Roos, Goosen, Cooper, van Helden, Parsons and Miller2018).
The intradermal test is based on delayed hypersensitivity in response to purified protein derivative (PPD) in the skin of the caudal fold (CFT) or neck (CIT). Intradermal tests have been used for primary screening tests of bTB infection due to low costs and high availability. Improved skin tests based on the mycobacterial antigens ESAT-6, CFP-10, and Rv3615c enabled differentiation between infected and uninfected vaccinated animals (Jones et al., Reference Jones, Whelan, Clifford, Coad and Vordermeier2012). Additionally, of all 41 uninfected and vaccinated animals tested, only one animal showed a detectable skin test response to ESAT-6/CFP-10/Rv3615c/Rv3020c, while 25 animals were positive for the single intradermal tuberculin (SIT) test (Jones et al., Reference Jones, Whelan, Clifford, Coad and Vordermeier2012). Recently, Srinivasan and colleagues developed and evaluated a novel peptide-based defined antigen skin test for the diagnosis of bTB and differentiation between infected and vaccinated cattle (Srinivasan et al., Reference Srinivasan, Jones, Veerasami, Steinbach, Holder, Zewude, Fromsa, Ameni, Easterling, Bakker, Juleff, Gifford, Hewinson, Vordermeier and Kapur2019). Peptide-based defined antigen skin tests detected 9 of 25 infected animals that were not detected by the traditional single intradermal comparative cervical tuberculin test (Srinivasan et al., Reference Srinivasan, Jones, Veerasami, Steinbach, Holder, Zewude, Fromsa, Ameni, Easterling, Bakker, Juleff, Gifford, Hewinson, Vordermeier and Kapur2019). However, intradermal testing has many limitations, such as labor-intensive work for injection and interpretation of results, need for a secondary visit to farm, and difficulty in standardization (de la Rua-Domenech et al., Reference de la Rua-Domenech, Goodchild, Vordermeier, Hewinson, Christiansen and Clifton-Hadley2006). The IFN-γ assay is based on the cell-mediated immune response produced by mycobacterial antigens. T lymphocytes produce IFN-γ in response to mycobacterial-specific antigens, such as early secretory antigen target 6 (ESAT-6) and culture filtrate protein 10 (CFP-10). The IFN-γ assay enables rapid and early diagnosis compared to bacterial culture or serological testing. However, the IFN-γ assay has several limitations, such as low specificity due to environmental mycobacteria and innate immune response in young cattle (Schiller et al., Reference Schiller, Oesch, Vordermeier, Palmer, Harris, Orloski, Buddle, Thacker, Lyashchenko and Waters2010; Jenkins et al., Reference Jenkins, Gormley, Gcebe, Fosgate, Conan, Aagaard, Michel and Rutten2018). Therefore, studies on the development of diagnostic methods based on host or pathogen-derived biomarkers of bTB, such as proteins, mRNA, and antigen-specific cytokines, have been conducted (Schiller et al., Reference Schiller, Oesch, Vordermeier, Palmer, Harris, Orloski, Buddle, Thacker, Lyashchenko and Waters2010).
Several studies have been conducted to identify pathogen biomarkers for the diagnosis of bTB. Lyashchenko and colleagues identified novel antigens with a multiantigen print immunoassay and serum samples of experimentally or naturally infected animals Lyashchenko et al. (Reference Lyashchenko, Grandison, Keskinen, Sikar-Gang, Lambotte, Esfandiari, Ireton, Vallur, Reed, Jones, Vordermeier, Stabel, Thacker, Palmer and Waters2017). Of the screened antigens, MPT70 and MPT83 had the highest antibody reactivity with infected sera, indicating that MPT70 and MPT83 had potential diagnostic value for Mb infection (Lyashchenko et al., Reference Lyashchenko, Grandison, Keskinen, Sikar-Gang, Lambotte, Esfandiari, Ireton, Vallur, Reed, Jones, Vordermeier, Stabel, Thacker, Palmer and Waters2017). In another study, ELISA was developed using ESAT-6/MPB70/MPB83 chimera recombinant protein for the diagnosis of bTB (Souza et al., Reference Souza, Melo, Ramos, Farias, Osório, Jorge, Vidal, Silva, Silva, Pellegrin and Araújo2012). ELISA based on ESAT-6/MPB70/MPB83 chimera protein had 83.2% sensitivity and 86.5% specificity when evaluated with 107 positive serum and 126 negative serum samples (Souza et al., Reference Souza, Melo, Ramos, Farias, Osório, Jorge, Vidal, Silva, Silva, Pellegrin and Araújo2012). Additionally, when these researchers screened 92 serum samples from cows classified into three groups based on the comparative intradermal tuberculin test (CITT), the sensitivity and specificity of ELISA were 79.5 and 75.5%, respectively (Souza et al., Reference Souza, Rodrigues, Gonçalves Jorge, Silva, Lilenbaum, Vidal, Etges, Kostovic and Araújo2019). In another study, the level of mannosylated LAM in serum was measured for use in the diagnosis of subclinical bTB infections. The results showed that the level of mannosylated LAM exhibited precise diagnostic performance with 100% sensitivity and 91.7% specificity for differentiating bTB-infected and bTB-exposed animals and 100% sensitivity and 93.9% specificity for differentiating bTB-exposed from bTB-negative animals. This finding indicates that serum mannosylated LAM levels can be a potential biomarker of subclinical bTB (Lamont et al., Reference Lamont, Ribeiro-Lima, Waters, Thacker and Sreevatsan2014a). Additionally, Mb-specific peptides, such as MB2515c, MB1895c, and MB1554c, were identified in the serum of experimentally infected cattle, and ELISA based on Mb-specific peptides was developed for the diagnosis of bTB (Lamont et al., Reference Lamont, Janagama, Ribeiro-Lima, Vulchanova, Seth, Yang, Kurmi, Waters, Thacker and Sreevatsan2014b). MB1554c had 99% sensitivity and 90% specificity when applied to naturally infected cattle (Lamont et al., Reference Lamont, Janagama, Ribeiro-Lima, Vulchanova, Seth, Yang, Kurmi, Waters, Thacker and Sreevatsan2014b). Marassi et al. (Reference Marassi, Medeiros, McNair and Lilenbaum2011) evaluated two indirect ELISA tests based on the Mb-specific proteins MPB70 and MPB83 for the diagnosis of bTB in naturally infected cattle (Marassi et al., Reference Marassi, Medeiros, McNair and Lilenbaum2011). However, diagnostic sensitivity was significantly reduced from 53.8 to 21.8% between Day 7 and Day 21 after the intradermal test (Marassi et al., Reference Marassi, Medeiros, McNair and Lilenbaum2011). Recently, Lyashchenko and colleagues developed an antigen capture immunoassay to detect immune complexes for the diagnosis of bTB (Lyashchenko et al., Reference Lyashchenko, Grandison, Keskinen, Sikar-Gang, Lambotte, Esfandiari, Ireton, Vallur, Reed, Jones, Vordermeier, Stabel, Thacker, Palmer and Waters2017). The immune complex consisting of antigen and immunoglobulin M (IgM) was formed at the early stage of bTB before antibody production (Lyashchenko et al., Reference Lyashchenko, Grandison, Keskinen, Sikar-Gang, Lambotte, Esfandiari, Ireton, Vallur, Reed, Jones, Vordermeier, Stabel, Thacker, Palmer and Waters2017). In a more recent study, Hadi et al. (Reference Hadi, Waters, Palmer, Lyashchenko and Sreevatsan2018) identified pathogen-derived biomarkers specific to Mb in the serum of calves experimentally infected with Mb using LC-MS/MS analysis (Hadi et al., Reference Hadi, Waters, Palmer, Lyashchenko and Sreevatsan2018). Mb-specific peptides, such as polyketide synthase, acyltransferase, and esterase, were identified in two animals and at 14 weeks and 36 weeks postinfection (Hadi et al., Reference Hadi, Waters, Palmer, Lyashchenko and Sreevatsan2018). A summary of the identified pathogen biomarkers of bTB is presented in Table 3.
Identified pathogen biomarkers for bTB
Several studies have been conducted to identify host-derived biomarkers for the diagnosis of bTB. Parsons et al. (Reference Parsons, McGill, Doyle, Goosen, van Helden and Gormley2016) suggested the use of IP-10 as a biomarker for the diagnosis of bTB because an increase in IP-10 was observed in whole blood after stimulation with Mb PPD but not after M. avium PPD and Phosphate buffered saline (PBS) stimulation. In addition, a statistically significant correlation was detected between IP-10 concentration and IFN-γ release in whole blood samples (Parsons et al., Reference Parsons, McGill, Doyle, Goosen, van Helden and Gormley2016). A similar study revealed the diagnostic potential of IP-10 for the diagnosis of bTB in African buffalo (Goosen et al., Reference Goosen, Cooper, Miller, van Helden and Parsons2015). Measuring the concentration of IP-10 in whole blood after antigen stimulation revealed that the IP-10 concentration was significantly higher in infected animals than in uninfected animals, and up to 93% of the infected animals were diagnosed with bTB (Goosen et al., Reference Goosen, Cooper, Miller, van Helden and Parsons2015). Additionally, Mb antigen-specific IP-10 production can differentiate TB reactor animals from noninfected animals (Coad et al., Reference Coad, Doyle, Steinbach, Gormley, Vordermeier and Jones2019). In detail, IP-10 cannot replace IFN-γ as a diagnostic biomarker of bTB. However, the combination of IFN-γ and IP-10 can increase the relative sensitivity of diagnosis from 68 to 73% in 124 naturally infected cattle (Coad et al., Reference Coad, Doyle, Steinbach, Gormley, Vordermeier and Jones2019).
However, contradictory results have been obtained in recent studies. To assess the diagnostic value of IL-17A and IP-10 in bTB, Xin et al. (Reference Xin, Gao, Yang, Li, Liang, Hou, Sui, Guo, Yuan, Zhu, Ding and Jia2018) compared the concentrations of IL-17A and IP-10 in whole blood in uninfected cattle that were PCR-positive or PCR-negative according to nasal swabs. The results showed a significant difference between uninfected and infected animals but no significant difference between PCR-positive and PCR-negative cattle (Xin et al., Reference Xin, Gao, Yang, Li, Liang, Hou, Sui, Guo, Yuan, Zhu, Ding and Jia2018). Lamont et al. (Reference Lamont, Janagama, Ribeiro-Lima, Vulchanova, Seth, Yang, Kurmi, Waters, Thacker and Sreevatsan2014b) proposed several host proteins, such as VDBP, fetuin, and α antitrypsin, as diagnostic biomarkers for bTB. Among the putative biomarkers, VDBP showed the highest sensitivity and specificity for the diagnosis of bTB, suggesting the possibility of obtaining a bTB diagnosis through the examination of host biomarkers (Lamont et al., Reference Lamont, Janagama, Ribeiro-Lima, Vulchanova, Seth, Yang, Kurmi, Waters, Thacker and Sreevatsan2014b). Additionally, upregulation of VDBP indicates increased transportation of vitamin D and activation of macrophages, which induce anti-mycobacterial activity (Lamont et al., Reference Lamont, Janagama, Ribeiro-Lima, Vulchanova, Seth, Yang, Kurmi, Waters, Thacker and Sreevatsan2014b). Elnaggar et al. (Reference Elnaggar, Abdellrazeq, Elsisy, Mahmoud, Shyboub, Sester, Khaliel, Singh, Torky and Davis2017) showed that the frequency of IFN-γ-producing CD4+ T cells and the plasma IL-1β level were higher in Mb-infected animals than in animals exposed to nontuberculosis Mycobacterium and uninfected animals. As a result, IL-1β can serve as a diagnostic biomarker to detect Mb-infected cows with high sensitivity and specificity (Elnaggar et al., Reference Elnaggar, Abdellrazeq, Elsisy, Mahmoud, Shyboub, Sester, Khaliel, Singh, Torky and Davis2017). Recently, Steinbach and colleagues established a dual IFN-γ/IL-2 fluorescence detection assay for quantification of the T cell immune response facilitated by bTB (Steinbach et al., Reference Steinbach, Vordermeier and Jones2019). The proportion of T cells that produce IL-2 alone was significantly higher in infected animals with visible lesions compared to the nonvisible lesion group (Steinbach et al., Reference Steinbach, Vordermeier and Jones2019). Gao et al. found that the level of IL-8 can distinguish between Mb-infected and noninfected cattle Gao et al. (Reference Gao, Guo, Li, Jia, Lin, Fang, Jiang, Zhu, Zhang, Ding and Xin2019). Additionally, the level of pentaxin can distinguish between Mb-infected cattle with PCR-positive and PCR-negative groups (Gao et al., Reference Gao, Guo, Li, Jia, Lin, Fang, Jiang, Zhu, Zhang, Ding and Xin2019). Furthermore, the proteomic profiles of infected PCR-positive and PCR-negative cattle were significantly differentiated, suggesting that a unique immune response to pathogens occurs between different disease stages in the progression of bTB (Gao et al., Reference Gao, Guo, Li, Jia, Lin, Fang, Jiang, Zhu, Zhang, Ding and Xin2019).
In addition, transcriptomic approaches to the detection of biomarkers of bTB infection have been conducted. With the development of novel analytic technology, including DNA microarray, RNA sequencing, and whole genome sequencing, the pangenomic scale of the gene expression profile was identified in bTB infection (MacHugh et al., Reference MacHugh, Gormley, Park, Browne, Taraktsoglou, O'Farrelly and Meade2009). Aranday-Cortes et al. (Reference Aranday-Cortes, Hogarth, Kaveh, Whelan, Villarreal-Ramos, Lalvani and Vordermeier2012) analyzed gene expression in spleen and lung cells of Mb-infected BALB/C mice and observed that the expression levels of granzyme A, granzyme B, CXCL9, IL-22, and CCR6 were markedly upregulated. Additionally, after stimulation with PPD, the expression levels of CXCL9, CXCL10, granzyme A, and IL-22 were significantly increased in peripheral blood mononuclear cells (PBMCs) of infected cattle compared to their expression levels in uninfected cattle (Aranday-Cortes et al., Reference Aranday-Cortes, Hogarth, Kaveh, Whelan, Villarreal-Ramos, Lalvani and Vordermeier2012). These results suggest that the potential biomarker candidates that were identified in the mouse model could be used as an alternative diagnostic method for bTB (Aranday-Cortes et al., Reference Aranday-Cortes, Hogarth, Kaveh, Whelan, Villarreal-Ramos, Lalvani and Vordermeier2012). In another study, the identification of biomarker candidates was conducted through the evaluation of immunological parameters by applying three different classification methods: linear discriminant analysis (LDA), quadratic discriminant analysis, and the k-nearest neighbor classifier (Blanco et al., Reference Blanco, Bigi and Soria2014). The LDA results showed the best diagnostic value for the IL-17 and IL-2 genes and were useful for confirming disease progression in cattle that were positive for tuberculin; however, there was little diagnostic value when used in combination with IL-10 gene analysis (Blanco et al., Reference Blanco, Bigi and Soria2014). Waters et al. (Reference Waters, Maggioli, Palmer, Thacker, McGill, Vordermeier, Berney-Meyer, Jacobs and Larsen2015) reported that cytokine gene expression associated with the Th17 response increased by 9-fold in PBMCs of infected cattle, suggesting that cytokines associated with the Th17 response might be a diagnostic biomarker of bTB (Waters et al., Reference Waters, Maggioli, Palmer, Thacker, McGill, Vordermeier, Berney-Meyer, Jacobs and Larsen2015). Recently, another study showed differential expression of 10 genes (CXCL9, THBS1, MMP9, IL-22, CXCL10, IFN-γ, IL-17, FYVE, CD14, and IL-1R) in PBMCs of bTB-infected animals (Klepp et al., Reference Klepp, Eirin, Garbaccio, Soria, Bigi and Blanco2019). Specifically, CXCL9 was upregulated in the PBMCs of Mb-infected animals, while THBS1 and MMP9 were decreased (Klepp et al., Reference Klepp, Eirin, Garbaccio, Soria, Bigi and Blanco2019). Palmer and colleagues evaluated the whole blood culture system based on the mycobacterial antigen ESAT-6/CFP-10/Rv3615c fusion protein (Palmer et al., Reference Palmer, Thacker, Rabideau, Jones, Kanipe, Vordermeier and Ray Waters2020). Several candidate biomarkers (CXCL9, CXCL10, IL-13, IL-21, and IL-22) were upregulated at both the gene and protein levels after stimulation with Mb-specific antigens (Palmer et al., Reference Palmer, Thacker, Rabideau, Jones, Kanipe, Vordermeier and Ray Waters2020). MicroRNAs have also been analyzed as diagnostic biomarkers of bTB (Golby et al., Reference Golby, Villarreal-Ramos, Dean, Jones and Vordermeier2014; Iannaccone et al., Reference Iannaccone, Cosenza, Pauciullo, Garofalo, Proroga, Capuano and Capparelli2018). Significant upregulation of miR-155 in PBMCs at 2 weeks and 11 weeks postinfection was observed in unvaccinated cattle (Golby et al., Reference Golby, Villarreal-Ramos, Dean, Jones and Vordermeier2014). Additionally, no differential expression of miR-155 was detected in preinfection samples from both unvaccinated and vaccinated cattle (Golby et al., Reference Golby, Villarreal-Ramos, Dean, Jones and Vordermeier2014). More recently, Iannaccone and colleagues found differential expression of mir-146a in milk samples from bTB-positive cattle (Iannaccone et al., Reference Iannaccone, Cosenza, Pauciullo, Garofalo, Proroga, Capuano and Capparelli2018). In receiver operating characteristic curve analysis, mir-146a showed 86 and 80% sensitivity and specificity, respectively, for the diagnosis of bTB (Iannaccone et al., Reference Iannaccone, Cosenza, Pauciullo, Garofalo, Proroga, Capuano and Capparelli2018).
Colostrum has an important role in early defense against infection by delivering antibodies, cytokines, and nutrients to newborn animals (Hurley and Theil, Reference Hurley and Theil2011). Sánchez-Soto et al. (Reference Sánchez-Soto, Ponce-Ramos, Hernández-Gutiérrez, Gutiérrez-Ortega, Álvarez, Martínez-Velázquez, Absalón, Ortiz-Lazareno, Limón-Flores, Estrada-Chávez and Herrera-Rodríguez2017) analyzed the expression of cytokine genes in colostrum-derived mononuclear cells of tuberculin-positive and tuberculin-negative cattle on farms where bTB was present. The results obtained by these researchers showed that TNF-α expression was higher in tuberculin-negative animals than in tuberculin-positive animals, whereas IL-6 was higher in tuberculin-positive cattle than in tuberculin-negative animals. In addition, IL-1α expression was higher in IFN-γ assay-positive cattle than in assay-negative cattle. These results suggest that the expression levels of IL-1α and TNF-α in colostrum-derived mononuclear cells could be used for the diagnosis of bTB (Sánchez-Soto et al., Reference Sánchez-Soto, Ponce-Ramos, Hernández-Gutiérrez, Gutiérrez-Ortega, Álvarez, Martínez-Velázquez, Absalón, Ortiz-Lazareno, Limón-Flores, Estrada-Chávez and Herrera-Rodríguez2017).
Several researchers have studied VOCs produced by mycobacterial pathogens during growth and VOCs produced by host animals when infected with Mb (Cho et al., Reference Cho, Jung and Oh2015; Ellis et al., Reference Ellis, Rice, Maurer, Stahl, Waters, Palmer, Nol, Rhyan, VerCauteren and Koziel2017; Küntzel et al., Reference Küntzel, Oertel, Fischer, Bergmann, Trefz, Schubert, Miekisch, Reinhold and Köhler2018). An electric nose has been used for the detection of VOCs in vapor and liquid samples, such as breath, serum, and urine (Fitzgerald et al., Reference Fitzgerald, Bui, Simon and Fenniri2017). Differences in VOCs in sera of bTB-infected and bTB-free cattle were analyzed for their potential use in the discrimination of bTB-infected cattle using the metal oxide sensor assay (Cho et al., Reference Cho, Jung and Oh2015). A metal oxide sensor assay showed the potential for detection of bTB infection using serum samples. However, the specificity and sensitivity of metal oxide sensors should be evaluated with field serum samples (Cho et al., Reference Cho, Jung and Oh2015). Küntzel et al. (Reference Küntzel, Oertel, Fischer, Bergmann, Trefz, Schubert, Miekisch, Reinhold and Köhler2018) identified the profiles of VOCs of bacterial cultures for discrimination and detection of diverse mycobacterial pathogens. The concentrations of 2-heptanol, 3-octanol, 2,3,5-trimethylfuran, acetone, and 3-pentanone were significantly increased in Mb culture, while five VOCs, including 2-ethylfuran, 2-propylfuran, 2-n-butylfuran, dibromochloromethane, and 2-heptanone, were significantly decreased (Küntzel et al., Reference Küntzel, Oertel, Fischer, Bergmann, Trefz, Schubert, Miekisch, Reinhold and Köhler2018). Ellis et al. (Reference Ellis, Rice, Maurer, Stahl, Waters, Palmer, Nol, Rhyan, VerCauteren and Koziel2017) suggested several VOC biomarkers in fecal samples, such as dimethyl sulfide, 3-methylthiophene, benzaldehyde, and 2-dodecanone, for the discrimination of BCG-vaccinated cattle after Mb infection (Ellis et al., Reference Ellis, Rice, Maurer, Stahl, Waters, Palmer, Nol, Rhyan, VerCauteren and Koziel2017). The available information about the identified host biomarkers for the diagnosis of bTB in cattle is presented in Table 4.
Identified host biomarkers for bTB
Bovigam, blood-based in vitro laboratory test for bovine tuberculosis; IGRA, interferon gamma release assay; IST, intradermal skin test; mQFT, modified QuantiFERON TB assay; PC-EC, peptide cocktail ESAT-6/CFP10; PC-HP, peptide cocktail ESAT-6, CFP10, Rv3615, and three other mycobacterium antigens; SCT, single cervical tuberculin; SICTT, single intradermal comparative tuberculin test; TST, tuberculin skin test; VOCs, volatile organic compounds.
Limitation of identified biomarkers
We have described the biomarkers of bTB and bovine paratuberculosis identified to date. Host biomarkers identified in previous studies showed differential expression in infected animals, including the subclinical stage of infection. However, the identified host biomarkers may not be specific to bTB and bovine paratuberculosis. Several host gene expression profiles or proteins can be nonspecific markers of inflammation or infection (Mayeux, Reference Mayeux2004). Therefore, pathogen-specific immune responses through stimulation with specific antigens of MAP or Mb should be applied to establish diagnostic methods based on host biomarkers. Most of the identified host biomarkers were not validated with well-characterized clinical samples. In that regard, host biomarkers should be validated through the test of sensitivity, specificity, reproducibility, and limit of detection. Pathogen-derived molecules, such as DNA, protein, and VOC, described in this review can be specific markers for bTB and bovine paratuberculosis. However, most of the developed diagnostic methods based on pathogen-derived biomarkers are limited to ELISA, which requires laborious procedures and a long time for specimen transport. Thus, there is still a requirement for desirable new diagnostic methods that are directly applicable to the field.
Conclusion and future perspectives
This review describes the current research trend of diagnostic biomarkers in mycobacterial infection of cattle. Diagnostic techniques using biomarkers have the advantage of diagnosing such diseases during the early subclinical stage. Moreover, these methods are noninvasive and enable the processing of large numbers of samples. To date, many studies have been conducted to identify diagnostic biomarkers in mycobacterial infection of animals. However, the biomarkers identified to date rarely have been applied in the clinical field of veterinary medicine. Validation and standardization of the identified potential biomarkers through field study is required for successful development of diagnostic methods based on biomarkers for clinical application. Further biomarker research should focus on identifying diagnostic methods that enable the diagnosis of disease at an early stage. Additionally, intelligent diagnostic methods are needed to discriminate between vaccinated and infected animals. Furthermore, the next generation of biomarker-based diagnostic methods should be on-site diagnostics that can be directly applied in the field.
Acknowledgments
This study was supported by the Strategic Initiative for Microbiomes in Agriculture and Food, Ministry of Agriculture, Food and Rural Affairs (Project No. 918020-4), BK21 PLUS and the Research Institute for Veterinary Sciences, Seoul National University, Republic of Korea. We are grateful to all members of the Laboratory of Veterinary Infectious Disease in Seoul National University for the discussion on this article.
Conflict of interest
The authors declare that they have no competing interests.
