Rhodococcus equi is a soil-dwelling facultative intracellular pathogen that can infect many mammals, including humans. R. equi is most well known for its ability to cause severe pyogranulomatous disease in foals, primarily involving the lungs although other body systems may also be affected. The disease is endemic on many horse-breeding farms worldwide and poses a severe threat to the horse breeding industry because there is no vaccine available. Current prophylaxis is based on systematic preventative treatments with macrolides combined with rifampicin, which are also used to treat clinical cases of the disease in foals. In this thesis I have used a combination of wet laboratory and bioinformatic approaches to identify the molecular basis of emerging combined resistance to macrolides and rifampicin in R. equi foal isolates from the USA. The genomes of a selection of resistant and susceptible strains from across the USA were sequenced and assembled. Resistance genes were systematically searched by reciprocal best-match BLASTP comparisons to known antibiotic resistance determinants. This led to the discovery of a novel erythromycin ribosomal methylase (erm) gene, erm(46), in all resistant strains. Complementation analysis in a susceptible R. equi strain showed that erm(46) was sufficient to confer resistance to all macrolides, lincosamides, and streptogramin B. The erm(46) gene is carried by an integrative conjugative element (ICE) which is transferable between R. equi strains. The ICE is formed by two distinct parts, a class I integron associated with an IS6100 sequence and the erm(46) determinant carried by a sub-element which contains putative actinobacterial conjugative translocase apparatus and a transposase/integrase. All resistant strains also carry the same non synonymous point mutation in rpoB conferring rifampicin resistance. Thus, these strains are carrying double resistance to the most commonly used antibiotics to treat R. equi worldwide. Phylogenetic analysis based on the core genome demonstrated that all resistant strains are clonal. This indicates that although conjugal acquisition of the erm(46) conjugative element may occur at a high frequency, the need for the concurrent presence of a second rpoB mutation for survival in the macrolide and rifampicin dominated farm environment has effectively selected for the spread of a single clone. In the second section of this work, we sequenced a further 20 R. equi genomes from difference sources (equine, porcine, bovine, human), including representatives of each of the seven major genogroups previously defined in our laboratory based on pulsed field gel electrophoresis. I have used the newly acquired genetic information to study the genome of R. equi and analyse its diversity within and outwith its species group. This enabled us to explore the pan genome and define that R. equi is a genetically well-defined bacterial species. Our results provide definitive evidence that resolves the current dispute over R. equi classification, specifically they do not support the recent proposal (based on classical polyphasic bacterial taxonomical methods) that R. equi should be transferred to a new genus. Our core-genome phylogenomic analyses unambiguously show that the genus Rhodocococcus is monophyletic and that R. equi forms a clade together with the most recently described related environmental species R. defluvii that radiates from within the genus. Together with other shared biological and genetic characteristics, namely the unique niche-adaptive mechanism based on evolutionarily related extrachromosomal replicons, R. equi should be conseidered a bona fide member of the genus Rhodococcus. We also confirm that Rhodococcus spp. and Nocardia spp. are sufficiently distinct to warrant them belonging to different genera. In conclusion, this work used whole genome sequencing to characterize the molecular basis underlying the emergence and clonal spread of multi-resistant R. equi in horse breeding farms in the USA. This work also highlights the limitations of classical taxonomical approaches in bacterial systematics, and illustrates the importance of incorporating modern phylogenomic approaches to understand the evolutionary relationships between bacterial strains and their accurate taxonomic position.

Rhodococcus equi is a common cause of severe pneumonia in foals, and is an opportunistic pathogen in immunocompromised humans. In combination, erythromycin and rifampin are the most commonly used antimicrobials in treating R. equi in foals. To provide reliable treatment, it is imperative to determine the mean inhibitory concentrations (MICs) of other antimicrobial agents in the event that certain strains of R. equi develop resistance to the current treatment. Several strains of R. equi have developed resistance to various antibiotics. In this study, R. equi strain 288 was completely resistant to rifampin with a MIC > 256ug/ml. The MICs of ethambutol, clarithromycin, azithromycin, isoniazide, ethionamide, rifampin, erythromycin, and linezolid of ninety-five R. equi isolates were also determined in this study. These isolates were obtained from the lungs and transtracheal washes of foals. In addition to these strains, three National Committee for Laboratory Clinical Standards (NCCLS) quality control strains were also tested: R. equi ATCC 6939, R. equi ATCC 33701, and S. pneumoniae 49619. Each drug was tested in triplicate and the MIC 50’s and MIC 90’s were determined for each drug. Ethambutol, isoniazide, and ethionamide were completely ineffective against R. equi. with MICs > 250ug/ml. Rhodococcus equi strains were more susceptible to clarithromycin (MIC 90 = 0.23 ug/ml) than to azithromycin (MIC 90 = 2.33 ug/ml), rifampicin (MIC 90 = 0.67ug/ml), erythromycin (MIC 90 = 1.2ug/ml), and linezolid (MIC 90 = 4ug/ml).

Rhodococcus equi is a common cause of severe pneumonia in foals, and is an opportunistic pathogen in immunocompromised humans. In combination, erythromycin and rifampin are the most commonly used antimicrobials in treating R. equi in foals. To provide reliable treatment, it is imperative to determine the mean inhibitory concentrations (MICs) of other antimicrobial agents in the event that certain strains of R. equi develop resistance to the current treatment. Several strains of R. equi have developed resistance to various antibiotics. In this study, R. equi strain 288 was completely resistant to rifampin with a MIC > 256ug/ml. The MICs of ethambutol, clarithromycin, azithromycin, isoniazide, ethionamide, rifampin, erythromycin, and linezolid of ninety-five R. equi isolates were also determined in this study. These isolates were obtained from the lungs and transtracheal washes of foals. In addition to these strains, three National Committee for Laboratory Clinical Standards (NCCLS) quality control strains were also tested: R. equi ATCC 6939, R. equi ATCC 33701, and S. pneumoniae 49619. Each drug was tested in triplicate and the MIC 50’s and MIC 90’s were determined for each drug. Ethambutol, isoniazide, and ethionamide were completely ineffective against R. equi. with MICs > 250ug/ml. Rhodococcus equi strains were more susceptible to clarithromycin (MIC 90 = 0.23 ug/ml) than to azithromycin (MIC 90 = 2.33 ug/ml), rifampicin (MIC 90 = 0.67ug/ml), erythromycin (MIC 90 = 1.2ug/ml), and linezolid (MIC 90 = 4ug/ml).

Orientador: Márcio Garcia Ribeiro / Resumo: Os patógenos entéricos são um grupo complexo de micro-organismos que causam doenças em animais e humanos. Os gatos domésticos podem ser considerados carreadores de patógenos entéricos com potencial zoonótico devido ao hábito de auto-limpeza, estabelecer amplo território, pelo instinto de caça e eliminar subclinicamente patógenos. O presente estudo investigou a ocorrência de Escherichia coli, Salmonella sp., Clostridium spp. e Rhodococcus equi nas fezes de 200 gatos domésticos sem sinais entéricos, dos quais 100 pertenciam ao ambiente urbano e 100 à áreas rurais. Marcadores de virulência de E. coli, Clostridium spp. e R. equi também foram investigados. Do total de 200 amostras fecais avaliadas foram identificadas 175 (88%) linhagens de E. coli das quais 93 (93%) de gatos de ambiente urbano e 82 (82%) de gatos de área rural. Houve diferença significante (p=0,03) entre os grupos para frequência mais alta de isolamento de E. coli em gatos de ambiente urbano. Dentre os marcadores de virulência de E. coli foram identificados os genes eae (13%), escN (13%), stx1 (1%), stx2 (0,6%), aatA (0,6%) e ipaH (0,6%). Foram isoladas 86 (43%) linhagens de C. perfringens A, das quais em 21 (24%) foi detectado o gene cpb2, que codifica a toxina beta 2 e em uma (1%) linhagem foi identificado o gene cpe, responsável pela codificação da enterotoxina. Também foram identificados cinco (2%) isolados de C. difficile, dos quais um (20%) apresentou os genes tcdA e tcdB, codificadores das toxinas A e B, re... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Enteric pathogens are a complex group of microorganisms associated to diseases in animals and humans. Domestic cats may be considered carriers of enteric pathogens with zoonotic potential, due to habit of self-cleaning, to establish wide territory of life, by hunting behavior, and subclinical elimination of pathogens through feces. The aim of present study was investigate the occurrence of Escherichia coli, Salmonella sp., Clostridium spp., and Rhodococcus equi in 200 fecal samples of cats without enteric signs, of which 100 cats belonging to urban area and 100 to farm environment. Virulence markers from E. coli, Clostridium spp. and R. equi were also tested. Among 200 fecal samples were identified 175 (88%) E. coli strains, from these, 93 (93%) were from feces of cats breeding in urban areas, whereas 82 (82%) were from animals from farms. Statistical difference was observed (p=0.03) to higher identification of E. coli from cats breeding in urban area. Genes eae (13%), escN (13%), stx1 (1%), stx2 (0,6%), aatA (0,6%) e ipaH (0,6%) were indentified from virulent markers in E. coli strains. Eighty-six (43%) C. perfringens A were isolated. From these, cpb2 gene (which encodes beta 2 toxin) was detected in 21 (24%) isolates, and cpe gene (which encodes enterotoxin) was identified in another (1%) isolate. Five (2%) strains of C. difficile were also identified. From these, only one (20%) presented tcdA e tcdB genes, which encodes A and B toxins, respectively. Were isolated also unco... (Complete abstract click electronic access below) / Mestre