Long-term molecular surveillance provides clues on a cattle origin for Mycobacterium bovis in Portugal

1 Vol.:(0123456789) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports Long‑term molecular surveillance provides clues on a cattle origin for Mycobacterium bovis in Portugal Ana C. Reis1,2, Rogério Tenreiro2, Teresa Albuquerque3, Ana Botelho3 & Mónica V. Cunha1,2* Animal tuberculosis (TB), caused by Mycobacterium bovis, is maintained in Portugal in a multi‑host system, with cattle, red deer and wild boar, playing a central role. However, the ecological processes driving transmission are not understood. The main aim of this study was thus to contribute to the reconstruction of the spatiotemporal history of animal TB and to refine knowledge on M. bovis population structure in order to inform novel intervention strategies. A collection of 948 M. bovis isolates obtained during long‑term surveillance (2002–2016, 15 years) of cattle (n = 384), red deer (n = 303) and wild boar (n = 261), from the main TB hotspot areas, was characterized by spoligotyping and 8 to 12‑loci MIRU‑VNTR. Spoligotyping identified 64 profiles and MIRU‑VNTR distinguished 2 to 36 subtypes within each spoligotype, enabling differentiation of mixed or clonal populations. Common genotypic profiles within and among livestock and wildlife in the same spatiotemporal context highlighted epidemiological links across hosts and regions, as for example the SB0119‑M205 genotype shared by cattle in Beja district or SB0121‑M34 shared by the three hosts in Castelo Branco and Beja districts. These genomic data, together with metadata, were integrated in a Bayesian inference framework, identifying five ancestral M. bovis populations. The phylogeographic segregation of M. bovis in specific areas of Portugal where the disease persists locally is postulated. Concurrently, robust statistics indicates an association of the most probable ancient population with cattle and Beja, providing a clue on the origin of animal TB epidemics. This relationship was further confirmed through a multinomial probability model that assessed the influence of host species on spatiotemporal clustering. Two significant clusters were identified, one that persisted between 2004 and 2010, in Beja district, with Barrancos county at the centre, overlapping the central TB core area of the Iberian Peninsula, and highlighting a significant higher risk associated to cattle. The second cluster was predominant in the 2012–2016 period, holding the county Rosmaninhal at the centre, in Castelo Branco district, for which wild boar contributed the most in relative risk. These results provide novel quantitative insights beyond empirical perceptions, that may inform adaptive TB control choices in different regions. The control and eradication of infectious diseases can be very difficult for pathogens that are able to thrive across multiple host species. This is the case of animal tuberculosis (TB), an infectious disease mainly caused by Mycobacterium bovis (M. bovis), which circulates within livestock and across several wildlife species worldwide, through direct and indirect routes1–4. Animal TB affects livestock productivity and trading, animal welfare, and also biodiversity, as it globally or locally threatens ecosystem services. The zoonotic potential of M. bovis bacteria is also liable as new infection human cases are reported every year5,6. Cattle (Bos taurus) is described as the main host species for M. bovis, while several wildlife hosts can support pathogen maintenance, depending on the ecosystem and epidemiological scenario. Over the years several molecular reports demonstrated inter-species transmission between cattle and wildlife, and spatial interaction studies associated wildlife as a risk factor for TB infection in cattle. Badger (Meles meles) in United Kingdom and Republic of Ireland; African buffalo (Syncerus caffer) in South Africa, brushtail possums (Trichosurus vulpecula) in New Zealand, white-tailed deer (Odocoileus virginianus) in EUA, and red deer (Cervus elaphus) and wild boar (Sus scrofa) in the Iberian Peninsula (Portugal and Spain) are the best characterized wildlife reservoirs5,7–10. In Iberian ecosystems, cattle and non-bovine livestock species managed in extensive husbandry, and red deer, wild OPEN 1Centre for Ecology, Evolution and Environmental Changes (cE3c), Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal. 2Biosystems & Integrative Sciences Institute (BioISI), Faculdade de Ciências da Universidade de Lisboa, Lisboa, Portugal. 3National Institute for Agrarian and Veterinary Research (INIAV IP), Oeiras, Portugal. *email: [email protected] 2 Vol:.(1234567890) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ boar, fallow deer and small carnivores compose the host community that can support M. bovis maintenance11. Some spatial clusters of infection in central and south-eastern regions of Portugal, in the periphery of the Iberian TB core area, have been detected, where average TB prevalence may reach as much as 50% in wild ungulate populations alone7,10,12. Expansion from the core Iberian area is believed to be driven by high host densities, artificial management and super-shedder wild populations13,14. In Portugal, a national eradication program for animal TB in cattle population commenced in 1987, being presently based on the detection and culling of reactors to the single intradermal comparative tuberculin (SICCT) test and the interferon-γ test of cattle aged over 6weeks, also including routine surveillance at slaughterhouses, mandatory sanitary classification of herds and regions, compulsory slaughter of positive animals, compensation to owners of slaughtered animals and pre-movement assessments15. Bacteriological culture and histopathology are the officially accepted diagnostics to confirm animal TB infection. Portugal has made significant progress in the control of bovine TB since the nationwide implementation of the eradication program approved by the European Union in 1992 (Council Decision 92/299/CE). Currently, the program covers mainland Portugal and the autonomous regions of Azores and Madeira16. During the last decade, TB herd prevalence remained low, varying between 0.11% in 2008 and 0.29% in 2017, with a peak in 2010 (0.91%)17. However, the national epidemiological indicators reflect different contexts throughout mainland Portugal, with TB herd prevalence ranging from 0.08% in the North up to 1.24% in Alentejo (south) in 201717. Algarve, in the extreme south of Portugal, is the single region that already achieved the officially TB-free status (Decision 2012/204/EU), where a specific surveillance program applies for the maintenance of this status (Fig.1). The eradication program has suffered changes along the years since its first implementation. While invivo surveillance and SICCT testing used to be homogenous across cattle herds, it has recently adapted towards the evolution of the epidemiological situation. There are now regional differences in the application of the surveillance scheme, according to herd type (beef, dairy, breeding), TB history and geographical location. The heterogeneity of TB herd prevalence across mainland Portugal indicates that some regional circumstances may favour intraspecific and interspecific M. bovis transmission, such as: (1) the management in extensive husbandry of livestock in the central and south regions that promote aggregation and enable the direct and indirect interaction of domestic with wild animals in communal pastures; (2) the wide host community of wildlife species potentially susceptible to TB that free-range in Portugal; and (3) the overabundance of widely distributed big game species that are artificially managed for hunting purposes, namely red deer and wild boar. Following several scientific publications reporting M. bovis occurrence in wild populations across two regional foci10,18–20 and the field knowledge reported by veterinary services of some municipalities, combined Figure1. Map of mainland Portugal highlighting the high TB prevalence area for cattle and the epidemiological risk area for wild ungulates (red deer and wild boar) and evidencing the geographic distribution of M. bovis isolates under analysis. The middle plots represent the evolution of the total number of M. bovis isolated from the different host species (red for cattle (C), green for red deer (RD) and blue for wild boar (WB)) sampled per geographic region, and the plot at the right compares the global number of M. bovis isolates per geographic region. 3 Vol.:(0123456789) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ with evidence for growing abundances of wild ungulates and the proliferation of big game hunting, in 2011 the national authorities defined an epidemiological risk area for big game species in south-central Portugal, along the border with Spain, where specific surveillance measures apply21. All big game specimens hunted in this risk area are mandatorily subjected to post-mortem examination in the field by a credentialed veterinarian. Samples from suspected TB lesions are collected and forwarded to the National Reference Laboratory (NRL) to confirm TB. Hunting by-products and animals with TB-compatible lesions (fully rejected) are disposed according to European Regulations (EC) no. 1069/2009 and 854/200421. The resilience of M. bovis in wildlife hosts remains a challenge for TB eradication in the cattle population and raises cattle and wildlife management issues in addition to public health concerns. The national context in Portugal is of special complexity, since M. bovis is maintained in this multi-host system with at least three maintenance hosts. Their single or joint contribution for M. bovis transmission and spatial patterns are not yet quantified. Although reports of M. bovis occurrence in other mammal species (e.g. caprine, ovine, carnivores) have remained occasional22,23, the role of each species in these transmission processes is unknown for now. Under these complex circumstances, appropriate epidemiological studies applied to high burden areas are required. So, in this work, we have focused on perceived hotspot areas where the livestock-wildlife interface is prominent, with widely distributed cattle herds of varying dimensions and abundant wild ungulate populations. We take advantage of a large collection of M. bovis isolates which has been gathered for almost two decades in the NRL and that can be exploited at the molecular level. A detailed molecular approach making use of spoligotyping (Spacer oligonucleotide typing) and MIRU-VNTR (Mycobacterial Interspersed Repetitive Units-Variable Number Tandem Repeats)22–27, combined with Bayesian clustering inference analyses, phylogenetic and network analyses, was envisaged with the aim to refine knowledge on the M. bovis population structure and to assess the spatiotemporal distribution of the most common genotypes, enabling the inference of the spatial risk, the most epidemiologically relevant hosts and the most probable transmission routes. Methods The general workflow followed in this study that involved genotypic characterization by two methods and statistical analyses is described in detail in Fig.2. Study area. The study area comprises the three regions regarded in Portugal as the main hotspot areas for animal TB at the livestock-cattle interface. They are enclosed in the districts (administrative level sample unit) of Castelo Branco, Portalegre and Beja, which locate on inner central and south Portugal, within the epidemiological risk area for TB in big game species that simultaneously holds chronically infected cattle herds that are managed in an extensive regimen (Fig.1). Castelo Branco was the most represented district, with 535 M. bovis isolates, followed by Portalegre, with 217, and Beja, with 196 isolates (Fig.1). The study area is simultaneously used for livestock production and big game hunting activities. Bacterial isolation and identification. Nine hundred and forty-eight M. bovis isolates were recovered (n = 948), in a 15-year period (2002–2016), from three host species, namely cattle (Bos taurus) (n = 384), red deer (Cervus elaphus) (n = 303) and wild boar (Sus scrofa) (n = 261) (Fig.1). Figure2. General workflow followed in this study (general template obtained in https ://tinyp pt.com/). 4 Vol:.(1234567890) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ This 15-year molecular analysis concerns three time periods: period 1 (2002–2008; 7years) in which herd prevalence steadily decreased and surveillance in wildlife was sporadic and uneven; period 2 (2009 and 2012; 4years) in which national herd prevalence raised more than fourfold and carcass examination of hunted big game species became mandatory in the epidemiological risk area from 2011 onwards; and, finally, period 3 (2013–2016; 4years) that was characterized by a decrease in herd prevalence after mandatory augmented surveillance by stakeholders, both at the livestock and wildlife flanks. Sample collection was performed within an official context according to specific national legislation or in the context of research projects held by the NRL. No animals were slaughtered or hunted for this study purpose. Cattle samples (lungs, lymph nodes, liver and/or spleen) were obtained after sanitary slaughter following previous positive intradermal testing or from evidence of macroscopic TB-like lesions at routine abattoir inspection. Wildlife samples (lungs, lymph nodes, liver and/or spleen) were collected from hunted animals presenting TBcompatible lesions that were detected during veterinarian examination in the field, which is obligatory for 19 counties (municipal level) enclosed in the hotspot areas. Animal tissue samples were pooled and processed in a biosecurity level 3 facility at INIAV, IP (National Institute for Agrarian and Veterinary Research), the NRL for animal TB diagnosis, following the protocol guidelines recommended in the OIE Manual for Terrestrial Animals (a detailed procedure is described in Cunha etal.7); and inoculated onto Stonebrink (Biogerm, Portugal) and Löwenstein-Jensen pyruvate (Biogerm, Portugal) solid media and liquid medium (BACTEC 9000MB system, Becton Dickinson Microbiology Systems). Cultures were incubated at 37°C and inspected weekly for growth for a minimum period of 12weeks. Pelleted cultures were heat-killed to extract the DNA and species identification was performed by restriction fragment length polymorphism analysis of PCR-amplified gyrB gene (PCR-REA) as described in Cunha etal.7 or using commercially available systems [GenoType Mycobacterium (Hain diagnostics, Germany)], according to manufacturer instructions. In all PCR batch assays, DNA from two positive controls (M. tuberculosis H37RV and M. bovis BCG) and a no-template control (purified water) were included. Spatiotemporal cluster analyses. A space–time analysis with a multinomial probability model28 was performed in SaTScan (version 9.6, https ://www.satsc an.org/) to assess the influence of host species on spatiotemporal risk clustering. In total, 249TB cases in cattle, grouped in 135 livestock herds, and 474TB cases in wildlife hosts, grouped in 91 officially-delimited hunting areas, were considered, meaning 76.4% of the global dataset, for which geographic coordinates of the location centroids of herds/hunting areas, were available. A database was created by geocoding each TB case to the geographic coordinates of the corresponding livestock herd or officially-delimited hunting area and no substantial change in geographic coordinates was assumed over the study period. The multinomial probability model applied here evaluates whether there are any clusters where the distribution of TB cases is different from the rest of the region under study. To operationalize the model, the study area is scanned using cylinders of varying sizes, with the base representing space and the height representing time, and the number of cases inside the cylinders is compared to the case expectations outside. Space–time clusters were identified at a maximum temporal and spatial window of 50.0% of the study period and case data, respectively. It was only allowed one set of geographical coordinates per location (herd/hunting area) and no overlapping clusters were permitted. Statistically significant clusters were evaluated at a p-value < 0.05, using Monte Carlo permutations with 999 replications of the dataset, under the null hypothesis of complete spatial randomness. The map was generated using QGIS (Quantum GIS development Team 2018, version 3.10, https ://www.qgis. org/en/site/) to visualize spatial‐temporal clusters and outbreak locations. Molecular characterization of M. bovis isolates. Spoligotyping was performed with in-house made membranes following the standard published protocol29, with the modifications described by Botelho etal.30. The presence or absence of the 43 spacers contained in the Direct Repeat (DR) locus was represented by a binary code. Spoligotyping patterns (prefix SB followed by four digits) were obtained from the Spoligotype Database website (http://www.mbovi s.org/index .php). DNA from M. bovis BCG and M. tuberculosis H37Rv isolates were used as positive controls and purified water as no-template control. The relative proportion of each one of the top six most common spoligotyping profiles, over the remaining in the different study years was obtained. A global and regional trend analysis was performed by applying a linear regression. A representative collection of isolates (n = 524) was further subjected to MIRU-VNTR analyses. Two selection criteria were applied: firstly, isolates were selected to cover all spoligotyping groups; and secondly, to cover in equal proportion different host species, geographic regions and year of isolation. Moreover, nine of the 29 orphan spoligotypes (i.e. with only one isolate) were also included. MIRU-VNTR typing was performed using a previously established set of 8-loci panel constituted by VNTR3232, ETR-A, ETR-B, ETR-C, QUB11a, QUB11b, MIRU4 and MIRU2620. These loci were amplified by PCR, in single reaction tubes, using the primers and cycling conditions previously described by Duarte etal.20. Positive control (DNA extracted from M. bovis BCG) and a no-template control (purified water) were included in all PCR assays. PCR products were separated in 2% agarose gels, prepared in Tris Buffer EDTA 1×, at 90V for 3h. The number of tandem repeats in each amplified locus was attributed by correlation with the amplicon size according to previously published tables31. Complete profiles were grouped by similarity in MIRU-VNTR types (M types). Profiles with non-amplifiable loci were excluded from this classification. 5 Vol.:(0123456789) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ A small subset of the MIRU-VNTR results (< 10%) reported in this study have been previously published by Duarte etal.20 (n = 23, between 2002 and 2007) and by Cunha etal.7 (n = 27, between 2007 and 2009), but were included to span the analyses for a wider period. A case was considered as polyclonal infection when two amplification products for one or more MIRU-VNTR loci were detected, as in Reis etal.32. In these cases, the presence of double alleles was confirmed by repeating the PCR and electrophoresis experiments. For the subgroup with double alleles (n = 37), four additional loci (MIRU16, MIRU40, ETR-E and QUB26) were added to the initial panel with the purpose of further discriminating the isolates, following the rational indicated by Reis etal.32. Similar to the 8-loci panel, these four additional loci were amplified by PCR, in single reaction tubes, using the primers described by Supply etal.33 for MIRU16, MIRU40 and ETR-E, and Skuce etal.34 for QUB26. The cycling and electrophoresis conditions applied were equal to the ones already described for the previous panel. The genotypic profiles were defined by the combination of spoligotyping profile and MIRU type. The calculation of the discriminatory power (D), that is the average probability that the typing method will assign a different type to two unrelated isolates sampled in the population, for both typing methods, and the individual allelic diversity (h) of the loci panel tested were obtained. These parameters were calculated based on Simpson’s index of diversity, as described by Hunter and Gaston35, using an online tool available on the website In silico simulation of molecular biology experiments (In silico, 2018)36. Only complete profiles with single alleles were considered for the calculation of the discriminatory power of MIRU-VNTR typing. Network analyses. To explore the relationships established within each host-geographic region pair, two networks involving spoligotyping (n = 948) or genotypic profiles (n = 487) (i.e. based on spoligotyping and MIRU) were visualized with Gephi (version 0.9.2, https ://gephi .org/). Each host-geographic region combination was considered a node, and the number of shared spoligotyping or genotypic profiles were used as connection lines between nodes. The connection lines were established based on absolute values; no statistical transformation was performed. Phylogenetic analysis. BioNumerics software (version 6.6, http://www.appli ed-maths .com/bionu meric s) (Applied Maths, Sint-Martens-Latem, Belgium) was used to visualize evolutionary relationships among the M. bovis isolates by generating Minimum Spanning Trees (MSTs), using a combined dataset composed by spoligotyping and 8-loci MIRU-VNTR (n = 487). Profiles with double alleles were excluded from these analyses. A composite character experiment was created with spoligotyping and MIRU-VNTR data, and an advanced cluster analysis was performed using the categorical option as a similarity coefficient. The single locus variant was used as a criterion for maximum distance between nodes. Bayesian population analysis. To further explore the population structure of M. bovis isolates, a Bayesian clustering algorithm, implemented in the software STRU CTU RE (version 2.3.4, https ://web.stanf ord.edu/ group /pritc hardl ab/struc ture.html)37 was applied to a dataset composed by MIRU-VNTR profiles (n = 487). Bayesian inference method and MST analysis were applied to the same M. bovis isolates. STRU CTU RE was used to infer the most likely number of putative ancestral populations (K), and to assign individuals to these populations. It was ran using the admixture model as an ancestry prior model. The admixture model considers that the data may have been originated from the admixture of K putative ancestral populations at some time in the past, so it was selected because it is a more flexible model and, thus, suitable for dealing with the complexities of a real data population. Five independent simulations were performed using Markov Chain Monte Carlo (MCMC) with an initial 10.000 ‘burn-in’ period and 100.000 iterations, following ‘burn-in’. For each simulation, K values were tested between 2 and 15, with 10 replicated runs for each value of K. To estimate the number of putative ancestral populations, delta K was calculated by the Evanno method38, using STRU CTU RE HARVESTER39. The estimate of delta K is based on the rate of change in the log probability of data between successive K values. The individual Q matrix corresponds to the estimated ancestry coefficients for each individual, that is the part of the genome that each individual inherited from ancestors. A threshold of Q ≥ 0.50 was applied to assign individuals to ancestral populations. New MST analyses were then performed using BioNumerics software (version 6.6), according to the criterion described in “Phylogenetic analysis” section, and identifying M. bovis by the classification provided by STRU CTU RE analysis. Allelic richness. For allelic richness analyses, M. bovis isolates were grouped according to geographic region and populations defined by STRU CTU RE. The mean allelic richness of each M. bovis group was evaluated using the statistical method of rarefaction implemented in the software HP-RARE (version 1.0, https ://www.monta na.edu/kalin owski /softw are/hp-rare.html)40, which standardizes samples of different sizes, compensating for sampling disparity, so that the allelic diversity across large and small sample datasets could be compared. Statistical analyses. The statistical significance between the associations of spoligotyping profiles proportion and geographic region or host species were tested using a chi-square test (α = 0.05). An unbiased estimate of the exact significance level was performed through Monte Carlo approximation with confidence interval of 99% and 10,000 repeated sampling. Chi-square test (α = 0.05) was also used to evaluate the statistical significance of the association between M. bovis ancestral populations defined by STRU CTU RE software and geographic region or host species. Contingency tables and tests were performed in IBM SPSS Statistics (version 24.0.0, https :// www.ibm.com/analy tics/spss-stati stics -softw are). Results were considered to be significant if p < 0.05. 6 Vol:.(1234567890) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ Results Demography analyses of M. bovis isolates. The M. bovis dataset considered in this study is well represented as far as time, space and animal hosts are concerned, however the strain set of the first period, especially from 2002 to 2006, is restricted (n = 85), corresponding to only 9% of the total dataset, which may be explained by reduced and uneven surveillance during that time period. The number of isolates has two peaks in 2007 and 2010 and values rise since 2013, coinciding with the increase in TB surveillance of the wildlife component. Until 2006, Beja was the most represented region (cumulative number of isolates = 46), but since then and until 2016, Castelo Branco (n = 511) outnumbered the number of M. bovis isolates recovered (n = 150 for Beja and n = 202 for Portalegre) (Fig.1). Cattle was the most represented host species until 2011 (n = 258), then wildlife species were sampled in higher numbers (n = 437) and with similar proportions each. M. bovis became regularly isolated from wildlife hosts after 2011, which is related with the settlement of the epidemiological risk area for TB in game species (Fig.1). The space–time clustering behaviour of TB cases was evaluated with a multinomial probability model, in a total of 723 cases, reported between 2002 and 2016, in cattle (n = 249), red deer (n = 247) and wild boar (n = 227). Two significant clusters were identified (p-value = 0.001) and were both composed by TB cases originated from the three host species (Fig.3). The 173.17km-radius spatial cluster in Barrancos county, Beja district, included a total of 108TB cases (n = 105 for cattle, n = 2 for red deer and n = 1 for wild boar) and persisted between 2004 and 2010, pointing out to a remarkable higher relative risk (RR) for cattle (RR = 4.16), when comparing with wild boar (RR = 0.025) and red deer (RR = 0.047) (Fig.3). The other cluster, with 31.49km-radius, holds the county Rosmaninhal at the centre, in Castelo Branco district, and embraces 270TB cases (n = 13 for cattle, n = 131 for red deer and n = 126 for wild boar) from 2012 to 2016. In this cluster, wild boar (RR = 2.08) presents a higher relative risk compared to red deer (RR = 1.90) and cattle (RR = 0.093) (Fig.3). Genotypic diversity of M. bovis. Spoligotyping was used as first-line genotyping tool for the selected 948 M. bovis isolates, distinguishing 64 patterns with varying prevalence (overall discriminatory power, D = 0.92) (Supplementary Fig.S1). Three spoligotypes, SB2529, SB2530 and SB2531, were recorded for the first time, and were introduced at the M.bovis.org database. Twenty-nine patterns with lower frequency (encompassing 2 to 42 Figure3. Multinomial space–time clusters (black-lined circles) of M. bovis calculated for the 2002–2016 period. Categories refer to host species. Relative Risk (RR) estimates indicate whether or not, greater than expected numbers occurred in a given category. RR > 1 represents a greater than expected number of individuals of a certain category inside the spatial-time cluster compared to outside. The radius of each space–time cluster is indicated. Map was generated with QGIS software (version 3.10, https ://www.qgis.org/en/site/). 7 Vol.:(0123456789) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ isolates) and 29 orphan patterns represented 32% and 3% of total isolates, respectively (Supplementary Fig.S1 and Supplementary TableS1). SB1174 (n = 158; 16.7%), SB0122 (n = 123; 13.0%), SB0121 (n = 111; 11.7%), SB1264 (n = 92; 9.7%), SB0265 (n = 75; 7.9%) and SB0119 (n = 57; 6.0%) were the top 6 profiles and together account for 65% of total isolates (Supplementary Fig.S1, TableS1). These six profiles were documented in the three host species and, with the exception of SB1174 and SB1264, they were distributed across the three geographic regions under analysis. The proportion of SB0119 and SB0121 varied across the study period, exhibiting a global negative trend, while SB0122 and SB1264 showed a positive tendency (Fig.4). For a subset of 524 M. bovis isolates (selected based on host species, location, year—see “Methods” section), a panel of 8-loci MIRU-VNTR was applied, resulting in 447 complete profiles with single alleles, grouped by similarity in 157 MIRU types and following the denomination of Cunha etal.7; 37 complete profiles with double alleles in one or more loci; and 40 profiles with one systematically un-amplified locus (Supplementary Fig.S2, TableS2). The majority of MIRU type groups (n = 95; 60.1%) were singletons, whereas a minority comprised 2 to 37 isolates (Supplementary Fig.S2). The six most common MIRU types (M34, M205, M219, M225, M251 and M261) account for 35% of sub-typified isolates (Supplementary TableS2). The contribution of singletons for global analysis was higher in the first time period, representing 26% of total MIRU types, whereas in the last considered time period singletons decreased to 21%. The largest MIRU type groups (M261, M34 and M225), encompassing over 30 isolates each, were found in the three host species under analysis; whereas M34 was recorded in the three regions, and M225 and M261 were restricted to Castelo Branco and Portalegre (Supplementary TableS2). The overall discriminatory power of MIRU-VNTR typing was higher than spoligotyping (D = 0.97) and the allelic diversity of individual locus ranged from 0.28 to 0.68, in decreasing order: ETR-A (h = 0.68), QUB11b (h = 0.58), ETR-B (h = 0.57), ETR-C (h = 0.53), VNTR 3232 (h = 0.42), QUB11a (h = 0.41), MIRU4 (h = 0.32) and MIRU26 (h = 0.28). All MIRU-VNTR loci were polymorphic, ranging between three (MIRU4) and 10 (VNTR3232) different alleles per locus. MIRU-VNTR typing enabled a finer discrimination of M. bovis isolates within the same spoligotype group, with the exception of two isolates with profile SB1273 that presented the same MIRU type. Therefore, the combined dataset of results from both typing methods enabled the definition of 208 genotypic profiles (Supplementary TableS3). There were genotypic profiles common to the three host species, but never to the three geographic regions under analysis; however, common profiles across bordering regions (Castelo Branco and Portalegre or Portalegre and Beja) were found. Polyclonal infection was suggested in 37 cases, through the identification of double alleles in one to four loci. Double alleles were registered across all loci, with the exception of QUB11a, and in the majority of profiles (n = 31; 83.8%) were restricted to one single locus (Supplementary TableS4). M. bovis genotypic variants were isolated from the three host species and geographic regions throughout the 2006–2015 period. In a small proportion of cases (n = 6), restricted to wild hosts from Castelo Branco, double alleles were recorded in two to four loci (Supplementary TableS4). M. bovis genotypic diversity across TB hotspot regions. The most represented region, Castelo Branco, was also the region that yielded the highest number of spoligotypes, with the identification of 48 profiles (D = 0.8974), followed by Beja, with 28 (D = 0.8966), and Portalegre, with 23 (D = 0.7762) (Fig.5). Analysis of spoligotyping diversity evolution through time, reveals an overall reduction in the three TB hotspots (Fig.5). From time period 1 to time period 3, D values varied between 0.88 to 0.85 in Castelo Branco, 0.92 to 0.67 in Portalegre and 0.81 to 0.76 in Beja. Eleven spoligotypes are common to the three geographic regions, accounting for more than half of M. bovis isolates. The panel of Castelo Branco-specific spoligotypes (56%) was higher when comparing with Beja (36%) and Portalegre (13%) (Fig.5, Supplementary TableS1). The prevalence rates varied across geographic regions, with SB0122 (21.6%) being most commonly reported in Castelo Branco, SB1174 (43.5%) in Portalegre and SB0265 and SB1190 (15.3%) equally prevalent in Beja (Supplementary TableS1). In Castelo Branco and Portalegre, the cumulative percentage of the six most common profiles increased by 20%, representing 78 and 81%, respectively, of the dataset in the time period 3. In Beja, profiles SB1174 and SB1264 were substituted by SB0295 and SB1190, and this cumulative evolution was subtle, from 61% in period 1 to 76% in period 3. Statistically significant associations were established within the spoligotyping profiles and the three locations. Beja was the district with more positive associations established (p-value < 0.01), wherein seven profiles were highlighted (SB0119, SB0120, SB0140, SB0265, SB0295, SB1172 and SB1190); Castelo Branco with four (SB0122, SB1195, SB1264 and SB 1266); and, finally, Portalegre with three (SB1167, SB1174 and SB1483) (p-value < 0.01) (Supplementary TableS5). MIRU-VNTR typing grouped the 254 isolates from Castelo Branco into over 100 distinctive MIRU types, while Portalegre and Beja isolates were split into 48 and 35 different MIRU types, respectively (Supplementary Fig.S3). M34(15.9%) was the most frequent in Beja; M225 (13.4%) in Castelo Branco; and M261 (21.4%) in Portalegre (Supplementary TableS2). The discriminatory power of MIRU-VNTR was higher than spoligotyping (D = 0.97) and the stratified analysis also disclosed similar values. ETR-A and MIRU26 were the loci with highest and lowest allelic diversity, respectively, both in Castelo Branco and Portalegre, while in Beja, QUB11a was the most discriminatory and QUB11b the least. 8 Vol:.(1234567890) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ Three MIRU types (M34, M109 and M199), encompassing 42 isolates, were common to the three geographic regions (Supplementary Fig.S3, TableS2). However, when combining spoligotyping and MIRU type, only adjacent geographic regions presented common genotypic profiles, in a total of 19 common to Castelo Branco and Portalegre and 2 common to Portalegre and Beja (Supplementary TableS3). Figure4. Variation in the proportion of M. bovis isolates included within the top 6 spoligotyping profiles. The solid line represents global predictions from a linear regression with 95% confidence interval, while punctuated lines represent the prediction for each geographic hotspot. Each graph is identified by the spoligotyping profile. 9 Vol.:(0123456789) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ M. bovis genotypic diversity across host species. Red deer yielded the highest number of spoligotyping profiles, with the 303 M. bovis isolates being grouped into 39 profiles, when comparing with the 384 cattle isolates that grouped into 38 profiles and 261 wild boar isolates that collapsed into 30 profiles, with 17 profiles common to the three host species (Fig.6). The discriminatory power presents a differential evolution through time periods, and the underlying overall discriminatory power ranged from 0.92 in cattle isolates to 0.88 in red deer (Fig.6). Frequencies of spoligotyping profiles vary across host species, SB0121 (17.7%) is the most frequent in cattle, SB0122 (21.8%) in red deer and SB1174 (21.8%) in wild boar. The spoligotypes SB0122, SB1174 and SB1264 are the top three most common in the wildlife hosts, only with variable frequencies. When considering the percentages of the six most common profiles, per host species, the increase of the cumulative effect, through time, only occurred in wildlife datasets. The contribution for cattle dataset is stable, around 55%; while for wild boar it ranges from 61 to 76% and for red deer from 52 to 85%, from period 1 to 3, respectively. Positive and significant relationships were established with all host species, being one profile (SB1264) associated to both wild boar and red deer (p-value < 0.01) (Supplementary TableS6). Red deer established positive associations with SB0122, SB1195 and SB1264; wild boar was associated with SB1174, SB1257, SB1264 and SB1265; and cattle with SB0119, SB0121, SB0140, SB0295, SB1090 and SB1172 (p-value < 0.01). Complete MIRU-VNTR profiles included 194 isolates from cattle grouped into 100 MIRU types, 138 from red deer clustered into 55 MIRU types and 115 from wild boar into 58 MIRU types (Supplementary Fig.S4). Seventeen MIRU types were common to three host species, representing 45% of the global dataset (Supplementary Fig.S4). The M34, M225 and M261 types were most commonly reported in specific host species datasets, M34 (11.8%) in cattle, M225 (13.9%) in wild boar and M225 and M261 (12.3%) in red deer. Figure5. Spoligotyping analysis per geographic region: (a) comparison between the total number of M. bovis isolates and the spoligotyping profiles registered; (b) common and specific-spoligotyping profiles; (c) evolution of the discriminatory power (D values) of spoligotyping for the global dataset (black line) and for each geographic region. Blue represents Castelo Branco (CB), orange Portalegre (PG) and purple Beja (BJ). Figure6. Spoligotyping analysis per host species: (a) comparison between the total number of M. bovis isolates and the spoligotyping profiles registered; (b) common and specific-spoligotyping profiles; (c) evolution of the discriminatory power (D values) of spoligotyping, for the global dataset (black line) and for each host species. Red represents cattle (C), green red deer (RD) and blue wild boar (WB). 16 Vol:.(1234567890) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ Received: 24 March 2020; Accepted: 17 November 2020 References 1. Gortázar, C. et al. Bovine tuberculosis in Doñana biosphere reserve: the role of wild ungulates as disease reservoirs in the last Iberian lynx strongholds. PLoS ONE 3, e2776 (2008). 2. Naranjo, V., Gortázar, C., Vicentea, J. & de la Fuente, J. Evidence of the role of European wild boar as a reservoir of Mycobacterium tuberculosis complex. Vet. Microbiol. 127, 1–9 (2008). 3. Barasona, J. A., Torres, M. J., Aznar, J., Gortázar, C. & Vicente, J. DNA detection reveals Mycobacterium tuberculosis complex shedding routes in its wildlife reservoir the Eurasian wild boar. Transbound. Emerg. Dis. 64, 906–915 (2017). 4. Cano-Terriza, D. et al. Management of hunting waste as control measure for tuberculosis in wild ungulates in south-central Spain. Transbound. Emerg. Dis. 65, 1190–1196 (2018). 5. Palmer, M. V., Thacker, T. C., Waters, W. R., Gortázar, C. & Corner, L. A. L. Mycobacterium bovis: a model pathogen at the interface of livestock, wildlife, and humans. Vet. Med. Int. 2012, 236205. https ://doi.org/10.1155/2012/23620 5 (2012). 6. Luciano, S. A. & Roess, A. Human zoonotic tuberculosis and livestock exposure in lowand middle-income countries: a systematic review identifying challenges in laboratory diagnosis. Zoonoses Public Health 67, 97–111 (2020). 7. Cunha, M. V. et al. Implications and challenges of tuberculosis in wildlife ungulates in Portugal: a molecular epidemiology perspective. Res. Vet. Sci. 92, 225–235 (2012). 8. Corner, L. A. L., Murphy, D. & Gormley, E. Mycobacterium bovis Infection in the Eurasian badger (Meles meles): the disease, pathogenesis, epidemiology and control. J. Comp. Pathol. 144, 1–24 (2011). 9. Fitzgerald, S. D. & Kaneene, J. B. Wildlife reservoirs of bovine tuberculosis worldwide: hosts, pathology, surveillance, and control. Vet. Pathol. 50, 488–499 (2012). 10. Vieira-Pinto, M. et al. Combined evaluation of bovine tuberculosis in wild boar (Sus scrofa) and red deer (Cervus elaphus) from central-east Portugal. Eur. J. Wildl. Res. 57, 1189–1201 (2011). 11. Cowie, C. E. et al. Interactions between four species in a complex wildlife: livestock disease community: implications for Mycobacterium bovis maintenance and transmission. Eur. J. Wildl. Res. 62, 51–64 (2016). 12. Santos, N. et al. Spatial analysis of wildlife tuberculosis based on a serologic survey using dried. Emerg. Infect. Dis. 24, 2169–2175 (2018). 13. Madeira, S. et al. Factors that influence Mycobacterium bovis infection in red deer and wild boar in an epidemiological risk area for tuberculosis of game species in Portugal. Transbound. Emerg. Dis. 64, 793–804 (2015). 14. Santos, N., Almeida, V., Gortázar, C. & Correia-Neves, M. Patterns of Mycobacterium tuberculosis-complex excretion and characterization of super-shedders in naturally-infected wild boar and red deer. Vet. Res. 46, 129–139 (2015). 15. DGAV. Programme for the eradication of bovine tuberculosis, bovine brucellosis or sheep and goat brucellosis (submitted for obtaining EU cofinancing-submission number 1483698737655-9828). https ://ec.europ a.eu/food/fundi ng/anima l-healt h/natio nal-veter inary -progr ammes _en Accessed Mar 2020 (2017). 16. DGAV. Programme for the eradication of bovine tuberculosis, bovine brucellosis or sheep and goat brucellosis (B. melitensis) (submitted for obtaining EU cofinancing). https ://ec.europ a.eu/food/fundi ng/anima l-healt h/natio nal-veter inary -progr ammes _enAccessed Mar 2020 (2019). 17. DGAV. Relatório técnico de sanidade animal—tuberculose Bovina. http://srvba mid.dgv.min-agric ultur a.pt/porta l/page/porta l/ DGV/gener icos?gener ico=24303 689&cboui =24303 689 Accessed Mar 2020 (2017). 18. Santos, N. et al. Epidemiology of Mycobacterium bovis infection in wild boar (Sus scrofa) from Portugal. J. Wildl. Dis. 45, 1048–1061 (2009). 19. Cunha, M. V. et al. Multihost tuberculosis: insights from the portuguese control program. Vet. Med. Int. 2011, 795165. htt ps ://doi. org/10.4061/2011/79516 5 (2011). 20. Duarte, E. L., Domingos, M., Amado, A., Cunha, M. V. & Botelho, A. MIRU-VNTR typing adds discriminatory value to groups of Mycobacterium bovis and Mycobacterium caprae strains defined by spoligotyping. Vet. Microbiol. 143, 299–306 (2010). 21. DGAV. Plano de Controlo e Erradicação de Tuberculose em Caça Maior (2011). 22. Matos, F. et al. Snapshot of Mycobacterium bovis and Mycobacterium caprae infections in livestock in an area with a low incidence of bovine tuberculosis. J. Clin. Microbiol. 48, 4337–4339 (2010). 23. Matos, A. C. et al. New Insights into Mycobacterium bovis prevalence in wild mammals in Portugal. Transbound. Emerg. Dis. 63, 313–322 (2016). 24. Hauer, A. et al. Genetic evolution of Mycobacterium bovis causing tuberculosis in livestock and wildlife in France since 1978. PLoS ONE 10, e0117103. https ://doi.org/10.1371/journ al.pone.01171 03 (2015). 25. Boniotti, M. B. et al. Molecular typing of Mycobacterium bovis strains isolated in Italy from 2000 to 2006 and evaluation of variablenumber tandem repeats for geographically optimized genotyping. J. Clin. Microbiol. 47, 636–644 (2009). 26. Allix, C. et al. Evaluation of the epidemiological relevance of variable-number tandem-repeat genotyping of Mycobacterium bovis and comparison of the method with IS6110 restriction fragment length polymorphism analysis and spoligotyping. J. Clin. Microbiol. 44, 1951–1962 (2006). 27. Navarro, Y. et al. Multiple sampling and discriminatory fingerprinting reveals clonally complex and compartmentalized infections by M. bovis in cattle. Vet. Microbiol. 175, 99–104 (2015). 28. Jung, I., Kulldorff, M. & Otukei John, R. A spatial scan statistic for multinomial data. Stat. Med. 29, 1910–1918 (2010). 29. Kamerbeek, J. et al. Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology. J. Clin. Microbiol. 35, 907–914 (1997). 30. Botelho, A., Canto, A., Leão, C. & Cunha, M. V. Clustered regularly interspaced short palindromic repeats (CRISPRs) analysis of members of the Mycobacterium tuberculosis complex. In Veterinary Infection Biology: Molecular Diagnostics and High-Throughput Strategies (eds Cunha, M. V. & Inácio, J.) 373–389 (Springer, Berlin, 2015). 31. Supply, P. et al. Proposal for standardization of optimized mycobacterial interspersed repetitive unit-variable-number tandem repeat typing of Mycobacterium tuberculosis. J. Clin. Microbiol. 44, 4498–4510 (2006). 32. Reis, A. C., Albuquerque, T., Botelho, A. & Cunha, M. V. Polyclonal infection as a new scenario in Mycobacterium caprae epidemiology. Vet. Microbiol. 240, 108533. https ://doi.org/10.1016/j.vetmi c.2019.10853 3 (2020). 33. Supply, P., Lesjean, S., Savine, E., Kremer, K. & Locht, C. Automated high-throughput genotyping for study of global epidemiology of Society. J. Clin. Microbiol. 39, 3563–3571 (2001). 34. Skuce, R. A. et al. Discrimination of Mycobacterium tuberculosis complex bacteria using novel VNTR-PCR targets. Microbiology 148, 519–528 (2002). 35. Hunter, P. R. & Gaston, M. A. Numerical index of the discriminatory ability of typing systems: an application of simpson’s index of diversity. J. Clin. Microbiol. 26, 2465–2466 (1988). 36. In silico simulation of molecular biology experiments. http://insil ico.ehu.es.Accessed Nov 2018. 37. Pritchard, J. K., Stephens, M. & Donnelly, P. Inference of population structure using multilocus genotype data. Genetics 155, 945–959 (2000). 17 Vol.:(0123456789) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ 38. Evanno, G., Regnaut, S. & Goudet, J. Detecting the number of clusters of individuals using the software STRU CTU RE: a simulation study. Mol. Ecol. 14, 2611–2620 (2005). 39. Earl, D. A. & von Holdt, B. M. Structure harvester: a website and program for visualizing Structure output and implementing the evanno method. Conserv. Genet. Resour. 4, 359–361 (2012). 40. Kalinowski, S. T. HP-RARE 1.0: a computer program for performing rarefaction on measures of allelic richness. Mol. Ecol. Notes 5, 187–189 (2005). 41. Merker, M. et al. Evolutionary history and global spread of the Mycobacterium tuberculosis Beijing lineage. Nat. Genet. 47, 242–249 (2015). 42. Duarte, E. L., Domingos, M., Amado, A. & Botelho, A. Spoligotype diversity of Mycobacterium bovis and Mycobacterium caprae animal isolates. Vet. Microbiol. 130, 415–421 (2008). 43. Rodríguez, S. et al. High spoligotype diversity within a Mycobacterium bovis population: clues to understanding the demography of the pathogen in Europe. Vet. Microbiol. 141, 89–95 (2010). 44. Rodriguez-Campos, S. et al. Splitting of a prevalent Mycobacterium bovis spoligotype by variable-number tandem-repeat typing reveals high heterogeneity in an evolving clonal group. J. Clin. Microbiol. 51, 3658–3665 (2013). 45. García-jiménez, W. L. et al. Spoligotype diversity and 5-year trends of bovine tuberculosis in Extremadura, southern Spain. Trop. Anim. Health Prod. 48, 1533–1540 (2016). 46. Romero, B. et al. Persistence and molecular evolution of Mycobacterium bovis population from cattle and wildlife in Doñana national park revealed by genotype variation. Vet. Microbiol. 132, 87–95 (2008). 47. Parra, A., Larrasa, J., García, A., Alonso, J. M. & de Mendoza, J. H. Molecular epidemiology of bovine tuberculosis in wild animals in Spain: a first approach to risk factor analysis. Vet. Microbiol. 110, 293–300 (2005). 48. de Mendoza, J. H. et al. Bovine tuberculosis in wild boar (Sus scrofa), red deer (Cervus elaphus) and cattle (Bos taurus) in a mediterranean ecosystem (1992–2004). Prev. Vet. Med. 74, 239–247 (2006). 49. Hauer, A. et al. Accurate phylogenetic relationships among Mycobacterium bovis strains circulating in France based on whole genome sequencing and single nucleotide polymorphism analysis. Front. Microbiol. 10, 955. https ://doi.org/10.3389/f micb .2019.00955 (2019). 50. Smith, N. H., Gordon, S. V., de la Rua-Domenech, R., Clifton-Hadley, R. S. & Hewinson, R. G. Bottlenecks and broomsticks: the molecular evolution of Mycobacterium bovis. Nat. Rev. Microbiol. 4, 670–681 (2006). 51. Corner, L. A. L. The role of wild animal populations in the epidemiology of tuberculosis in domestic animals: how to assess the risk. Vet. Microbiol. 112, 303–312 (2006). 52. Kukielka, E. et al. Spatial and temporal interactions between livestock and wildlife in south central Spain assessed by camera traps. Prev. Vet. Med. 112, 213–221 (2013). 53. Barasona, J. A. et al. Spatiotemporal interactions between wild boar and cattle: implications for cross-species disease transmission. Vet. Res. 45, 122. https ://doi .org/10.1186/s1356 7-014-0122-7 (2014). 54. Warren, R. M. et al. Microevolution of the direct repeat region of Mycobacterium tuberculosis: implications for interpretation of spoligotyping data. J. Clin. Microbiol. 40, 4457–4465 (2002). 55. Reyes, J. F., Chan, C. H. S. & Tanaka, M. M. Impact of homoplasy on variable numbers of tandem repeats and spoligotypes in Mycobacterium tuberculosis. Infect. Genet. Evol. 12, 811–818 (2012). 56. Skuce, R. A. et al. Mycobacterium bovis genotypes in Northern Ireland: herd-level surveillance (2003 to 2008). Vet. Res. 167, 684–689 (2010). 57. Balseiro, A. et al. Spatial relationships between Eurasian badgers (Meles meles) and cattle infected with Mycobacterium bovis in northern Spain. Vet. J. 197, 739–745 (2013). 58. Serraino, A. et al. Monitoring of transmission of tuberculosis between wild boars and cattle: genotypical analysis of strains by molecular epidemiology techniques. J. Clin. Microbiol. 37, 2766–2771 (1999). 59. Andrievskaia, O., Turcotte, C., Berlie-surujballi, G., Battaion, H. & Lloyd, D. Genotypes of Mycobacterium bovis strains isolated from domestic animals and wildlife in Canada in 1985–2015. Vet. Microbiol. 214, 44–50 (2018). 60. Glaser, L. 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 11, e0145735. https ://doi.org/10.1371/journ al.pone.01457 35 (2016). 61. Biek, R. et al. Whole genome sequencing reveals local transmission patterns of Mycobacterium bovis in sympatric cattle and badger populations. PLoS Pathog. 8, 1003008. https ://doi.org/10.1371/journ al.ppat.10030 08 (2012). 62. Barasona, J. A. et al. Environmental presence of Mycobacterium tuberculosis complex in aggregation points at the wildlife/ livestock interface. Transbound. Emerg. Dis. 64, 1148–1158 (2016). 63. Caley, P., Hickling, G. J., Cowan, P. E. & Pfeiffer, D. U. Effects of sustained control of brushtail possums on levels of Mycobacterium bovis infection in cattle and brushtail possum populations from Hohotaka, New Zealand Mycobacterium bovis infection in cattle and brushtail possum. N. Z. Vet. J. 47, 133–142 (1999). 64. Reilly, L. A. & Courtenay, O. Husbandry practices, badger sett density and habitat composition as risk factors for transient and persistent bovine tuberculosis on UK cattle farms. Prev. Vet. Med. 80, 129–142 (2007). 65. Cosgrove, M. K., Brien, D. J. O. & Ramsey, D. S. L. Baiting and feeding revisited: modeling factors influencing transmission of tuberculosis among deer and to cattle. Front. Vet. Sci. 5, 306. https ://doi.org/10.3389/fvets .2018.00306 (2018). 66. Birch, C. P. D., Goddard, A. & Tearne, O. A new bovine tuberculosis model for England and Wales (BoTMEW) to simulate epidemiology, surveillance and control. BMC Vet. Res. 14, 273. https ://doi.org/10.1186/s1291 7-018-1595-9 (2018). 67. Lahue, N. P., Banos, J. V., Acevedo, P., Gortázar, C. & Martínez-López, B. Spatially explicit modeling of animal tuberculosis at the wildlife-livestock interface in ciudad real province, Spain. Prev. Vet. Med. 128, 101–111 (2016). 68. Martínez-López, B. et al. Farm-level risk factors for the occurrence, new infection or persistence of tuberculosis in cattle herds from south-central Spain. Prev. Vet. Med. 116, 268–278 (2014). 69. Bouchez-zacria, M., Courcoul, A. & Durand, B. The distribution of bovine tuberculosis in cattle farms is linked to cattle trade and badger-mediated contact networks in south-western France, 2007–2015. Front. Vet. Sci. 5, 173. https ://doi.org/10.3389/fvets .2018.00173 (2018). 70. Green, D. M., Kiss, I. Z., Mitchell, A. P. & Kao, R. R. Estimates for local and movement-based transmission of bovine tuberculosis in British cattle. Proc. R. Soc. 275, 1001–1005 (2008). 71. Goodchild, A. V., Watkins, G. H., Sayers, A. R., Jones, J. R. & Clifton-Hadley, R. S. Geographical association between the genotype of bovine tuberculosis in found dead badgers and in cattle herds. Vet. Res. 170, 259. https ://doi.org/10.1136/vr.10019 3 (2012). 72. Gopal, R., Goodchild, A., Hewinson, G. & Domenech, R. D. R. Introduction of bovine tuberculosis to north-east England by bought-in cattle. Vet. Rec. 159, 265–271 (2006). 73. Grear, D. A., Kaneene, J. B., Averill, J. J. & Webb, C. T. Local cattle movements in response to ongoing bovine tuberculosis zonation and regulations in michigan, USA. Prev. Vet. Med. 114, 201–212 (2014). 74. Delahay, R., Langton, S., Smith, G., Clifton-Hadley, R. & Cheeseman, C. The spatio-temporal distribution of Mycobacterium bovis (bovine tuberculosis) infection in a high-density badger population. J. Anim. Ecol. 69, 428–441 (2000). 75. Navarro, Y. et al. Detailed chronological analysis of microevolution events in herds infected persistently by Mycobacterium bovis. Vet. Microbiol. 183, 97–102 (2016). 18 Vol:.(1234567890) Scientific Reports | (2020) 10:20856 | https://doi.org/10.1038/s41598-020-77713-8 www.nature.com/scientificreports/ 76. Gormley, E., Corner, L. A. L., Costello, E. & Rodriguez-Campos, S. Bacteriological diagnosis and molecular strain typing of Mycobacterium bovis and Mycobacterium caprae. Res. Vet. Sci. 97, 30–43 (2014). 77. Rodriguez-Campos, S. et al. European 2—a clonal complex of Mycobacterium bovis dominant in the Iberian Peninsula. Infect. Genet. Evol. 12, 866–872 (2012). 78. Berg, S. et al. African 2, a clonal complex of Mycobacterium bovis epidemiologically important in east Africa. J. Bacteriol. 193, 670–678 (2011). 79. Muller, B. et al. African 1, an epidemiologically important clonal complex of Mycobacterium bovis dominant in Mali, Nigeria, Cameroon, and Chad. J. Bacteriol. 191, 1951–1960 (2009). 80. Smith, N. H. et al. European 1: a globally important clonal complex of Mycobacterium bovis. Infect. Genet. Evol. 11, 1340–1351 (2011). 81. Branger, M. et al. The complete genome sequence of Mycobacterium bovis Mb3601, a SB0120 spoligotype strain representative of a new clonal group. Infect. Genet. Evol. 82, 104309. https ://doi.org/10.1016/j.meegi d.2020 (2020). 82. Smith, N. H. et al. Ecotypes of the Mycobacterium tuberculosis complex. J. Theor. Biol. 239, 220–225 (2006). Acknowledgements MVC thanks the national authority for nature conservation and forests, ICNF (Gonçalo Lopes), for providing georeferenced location of the hunting areas, and the national veterinary authority, DGAV (DSSPA, Ana Caria), for kindly providing georeferenced data from cattle herds. MVC also acknowledges DVM João Serejo for fruitful discussion on TB along the years and excellent field work, the generous collaboration of hunting associations and the laboratorial assistance by Celeste Matos, Alice Batalha and Ana Canto at the NRL for animal TB. This work was funded by Programa Operacional de Competitividade e Internacionalização (POCI) (FEDER component), Programa Operacional Regional de Lisboa, and Fundação para a Ciência e a Tecnologia (FCT/MEC), Portugal, in the scope of the project “Colossus: Control Of tubercuLOsiS at the wildlife/livestock interface uSing innovative natUre-based Solutions” (ref. POCI-01-0145FEDER-029783) awarded to MVC as PI. This work was also supported by strategic funding to cE3c and BioISI Research Units (UID/BIA/00329/2020 and UID/ Multi/04046/2020). ACR is the recipient of a PhD fellowship by FCT/MEC (PD/BD/128031/2016). Author contributions M.V.C. conceived the study and directed all experiments. A.B. and T.A. provided the bacterial isolates from the NRL. A.C.R. performed the experimental molecular work under the supervision of M.V.C. and R.T. Formal data analyses were performed by A.C.R. and M.V.C. A.C.R. and M.V.C. wrote and revised all drafts of the manuscript. R.T. gave critical feedback on final draft. All authors approved the final version for submission. Competing interests The authors declare no competing interests. Additional information Supplementary information is available for this paper at https ://doi.org/10.1038/s4159 8-020-77713 -8. Correspondence and requests for materials should be addressed to M.V.C. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/. © The Author(s) 2020