Yes / Background: Tracking recent transmission is a vital part of controlling widespread pathogens such as Mycobacterium tuberculosis. Multiple methods with specific performance characteristics exist for detecting recent transmission chains, usually by clustering strains based on genotype similarities. With such a large variety of methods available, informed selection of an appropriate approach for determining transmissions within a given setting/time period is difficult. Methods: This study combines whole genome sequence (WGS) data derived from 324 isolates collected 2005–2010 in Kinshasa, Democratic Republic of Congo (DRC), a high endemic setting, with phylodynamics to unveil the timing of transmission events posited by a variety of standard genotyping methods. Clustering data based on Spoligotyping, 24-loci MIRU-VNTR typing, WGS based SNP (Single Nucleotide Polymorphism) and core genome multi locus sequence typing (cgMLST) typing were evaluated. Findings: Our results suggest that clusters based on Spoligotyping could encompass transmission events that occurred almost 200 years prior to sampling while 24-loci-MIRU-VNTR often represented three decades of transmission. Instead, WGS based genotyping applying low SNP or cgMLST allele thresholds allows for determination of recent transmission events, e.g. in timespans of up to 10 years for a 5 SNP/allele cut-off. Interpretation: With the rapid uptake of WGS methods in surveillance and outbreak tracking, the findings obtained in this study can guide the selection of appropriate clustering methods for uncovering relevant transmission chains within a given time-period. For high resolution cluster analyses, WGS-SNP and cgMLST based analyses have similar clustering/timing characteristics even for data obtained from a high incidence setting. / ERC grant [INTERRUPTB; no. 311725] to BdJ, FG and CJM; an ERC grant to TS [PhyPD; no. 335529]; an FWO PhD fellowship to PM [grant number 1141217N]; the Leibniz Science Campus EvolLUNG for MM and SN; the German Centre for Infection Research (DZIF) for TAK, MM, CU, PB and SN; a SNF SystemsX grant (TBX) to JP and TS and a Marie Heim-Vögtlin fellowship granted to DK by the Swiss National Science Foundation. The computational resources and services used in this work were provided by the VSC (Flemish Supercomputer Center), funded by the Research Foundation - Flanders (FWO) and the Flemish Government – department EWI.

Campylobacter spp., principalmente as espécies C. coli e C. jejuni, são a causa mais comum de doença bacteriana veiculada por alimentos na Europa, Estados Unidos e alguns outros locais do mundo. No Brasil, há uma escassez de estudos de C. coli, o que dificulta avaliar a dimensão do envolvimento dessa bactéria como causadora de doença nos seres humanos e em animais, bem como, determinar o impacto de sua presença em alimentos e no meio ambiente. O objetivo desse trabalho foi caracterizar molecularmente linhagens de C. coli isoladas de origens diversas no Brasil pela pesquisa da presença de genes relacionados à virulência por PCR, perfil de sensibilidade a antimicrobianos e pela análise da similaridade genotípica por métodos de tipagem molecular. Adicionalmente, o Índice de Discriminação (D) de tais metodologias foi verificado. Foram estudadas 63 linhagens de C. coli, isoladas de humanos (12), animais (21), alimentos (10) e ambiente (20), entre os anos de 1995 e 2011, nos Estados do Rio de Janeiro, São Paulo e Minas Gerais. Todas as linhagens apresentaram os genes flaA, cadF e sodB. O gene cdtB foi detectado em 20 (31,7%) linhagens, o gene flhA foi detectado em 11 (17,5%) linhagens, o gene dnaJ foi encontrado em 10 (15,9%) linhagens, o gene pldA foi detectado em sete (11,1%) linhagens, o gene iamA foi detectado em três (4,8%) linhagens, os genes cdtC e docA foram encontrados em duas (3,2%) linhagens, os genes cdtA e crsA foram encontrados em uma (1,6%) linhagem e os genes ciaB, wlaN, virB11 e racR não foram detectados. Dentre as 63 linhagens estudadas, 42 foram susceptíveis a todos os antimicrobianos testados. Das 21 linhagens resistentes, 10 (15,9%) foram resistentes a tetraciclina e doxaciclina, seis (9,5%) foram resistentes a ciprofloxacina e uma (1,6 %) foi resistente a eritromicina. Somente quatro (6,3%) linhagens foram resistentes a pelo menos duas diferentes classes de antibióticos testados simultaneamente. O dendrograma de similaridade genética de Pulsed field gel electrophoresis (PFGE) agrupou as 63 linhagens estudadas em dois grupos principais denominados PFGE-A e PFGE-B com similaridade genômica de 44,9% entre eles. Entretanto, algumas linhagens isoladas de humanos, animais, ambiente e alimentos apresentaram uma alta similaridade genotípica acima de 80% entre elas e, foram agrupadas em sete subgrupos denominados PFGE-A1 a PFGE-A7. O dendrograma de similaridade genômica das sequências da SVR do gene flaA agrupou as linhagens ii estudadas em dois grupos principais designados SVR-A e SVR-B, com similaridade acima de 83,1 % entre eles. Ademais, o depósito das sequências da SVR do gene flaA no banco de dados online demonstrou que os alelos 30 e o 1647 foram os mais frequentemente encontrados e permitiu a comparação das linhagens estudadas com os alelos descritos no banco de dados. Sete alelos, dentre os 22 encontrados, não haviam sido previamente descritos. A análise do locus CRISPR por HRMA dividiu as linhagens de C. coli em quatro diferentes perfis de melting. O Multilocus sequence typing (MLST) foi utilizado para tipar 20 linhagens de C. coli e foram obtidos 18 STs diferentes dos quais apenas dois já haviam sido previamente descritos. O D das metodologias de PFGE, sequenciamento da SVR do gene flaA, análise do locus CRISPR por HRMA e MLST foi de 0,986, 0,916, 0,550 e 0,989, respectivamente. Pode- se concluir que o potencial patogênico das linhagens de C. coli não foi evidenciado o que pode estar relacionado ao fato da maioria dos estudos envolvendo patogênese terem sido realizados para a espécie C. jejuni. Algumas linhagens apresentaram-se resistentes aos antimicrobianos testados, o que é preocupante uma vez que tais linhagens podem disseminar genes de resistência a outras isoladas de diversas fontes. Os resultados gerados pelos métodos de tipagem molecular por PFGE e sequenciamento da pequena região variável (SVR) do gene flaA demonstraram uma alta similaridade genotípica entre algumas linhagens de C. coli, sugerindo que uma possível contaminação tenha ocorrido entre linhagens isoladas de fontes clínicas e não clínicas ao longo de 16 anos no Brasil. Ademais, a análise dos alelos da SVR do gene flaA nos permitiu concluir que os alelos prevalentes nas linhagens estudadas diferem daqueles encontrados nos países Europeus. Os dados obtidos por MLST sugerem que as linhagens estudadas possuem uma grande diversidade genética entre si e em comparação com as linhagens isoladas em diferentes locais do mundo. Finalmente, as técnicas de MLST e PFGE foram as mais eficientes e adequadas na genotipagem das linhagens de C. coli estudadas. / Campylobacter spp., mainly the C. coli and C. jejuni species, are the most common cause of bacterial disease conveyed by food in Europe, United States, and other places worldwide. In Brazil, there is a paucity of studies on C. coli, which makes it difficult to evaluate the involvement of this bacterium as a cause of diseases in humans and animals, as well as to determine the impact of its presence in food and the environment. The aim of this study was to molecularly characterize C. coli strains isolated from diverse origins in Brazil by searching for the presence of virulence-related genes by PCR, antimicrobial sensitivity profile, and analysis of the genotypic similarity by molecular typing methods. Addicionaly, the Discriminatory Index (D) of those methodologies was acessed. Sixty-three C. coli strains isolated from humans (12), animals (21), food (10), and the environment (20) between 1995 and 2011, in the States of Rio de Janeiro, São Paulo, and Minas Gerais were studied. All strains presented the flaA, cadF and sodB genes. The cdtB gene was detected in 20 (31.7%) strains; the flhA gene was detected in 11 (17.5%) strains; the dnaJ gene was detected in 10 (15.9%) strains; the pldA gene was detected in 7 (11.1%) strains ; the iamA gene was detected in three (4.8%) strains; the cdtC and docA genes were found in two (3.2%) strains; the cdtA and crsA were found in one (1.6%) strain and the ciaB, wlaN, virB11 and racR genes were not detected. Among the 63 strains studied, 42 were susceptible to all antimicrobials tested. Of the 21 resistant strains, 10 (15.9%) were resistante to tetracycline and doxaciclyne, six (9.5%) showed resistance to ciprofloxacin, and one (1.6%) was resistant to erythromycin. Only four (6.3%) strains were simultaneously resistant to at least two different classes of the antibiotics tested. The dendrogram of genetic similarity of Pulsed field gel electrophoresis (PFGE) grouped the 63 strains studied into two groups namely PFGE-A and PFGE-B with a genomic similarity of 44.9% among them. However, some strains isolated from humans, animals, the environment and food presented a high genotypic similarity above 80% and were subdivided into seven groups designated as PFGE-A1 to PFGE-A7. The dendrogram of genetic similarity of the SRV-flaA gene sequences grouped the strains studied into two groups namely SVR-A and SVR-B, with similarity above 83.1% among them. Besides, the deposit of the SVR sequences of the flaA gene in the online database showed that the alleles 30 and 1647 were the iv most frequently found and allowed the comparison between the strains studied with the alleles described in the database. Seven alleles, among the 22 found have never been described before. The CRISPR locus analysis divided the C. coli strains into four different melting profiles. The Multilocus sequence typing (MLST) was used to type 20 C. coli strains and revealed 18 different STs among which just two had been previously described. The D of PFGE, SVR- flaA sequence, HRMA of CRISPR locus analysis and MLST was 0.986, 0.916, 0.550 and 0.989, respectively. In conclusion, the pathogenic potential of the C. coli strains was not highlighted, which could be related to the fact that the majority of the pathogenicity studies were performed with C. jejuni species. Some strains showed resistance to the antibiotics tested what is a concern once those strains may spread the resistance genes to other strains isolated from different sources. The results obtained by PFGE and SVR-flaA sequence showed a high genomic similarity among some C. coli strains which may suggest that a possible contamination may have occurred among clinical and non-clinical sources during 16 years in Brazil. Furthermore, the analysis of SVR- flaA alleles allowed the conclusion that the prevalent alleles in the strains studied were different from those found in European countries. The data obtained by MLST suggests that the strains studied had a high genomic diversity among them and in comparison with strains isolated from different places worldwide. Finally, the MLST and PFGE technicques were the most efficient and adequate in genotyping the C. coli strains studied.

O objetivo deste trabalho foi investigar a presença de bactérias produtoras de β-lactamases adquiridas na microbiota Gram-negativa comensal de humanos e animais domésticos em uma comunidade remota na região da Floresta Amazônica. De março a julho de 2013 foram coletadas amostras de fezes de indivíduos atendidos e funcionários de um centro assistencial em saúde restrito a comunidades indígenas e de swab retal de animais de companhia da comunidade. Nas amostras de humanos foram detectados isolados de Escherichia coli, Klebsiella pneumoniae, Enterobacter aerogenes, Enterobacter kobei e Morganella morganii, carregando os genes blaTEM-1, blaCTX-M-15, blaCTX-M-14, blaCTX-M-8 e blaGES-5. Nas amostras de animais foram detectados apenas isolados de E. coli carregando os genes blaTEM-1, blaCTX-M-14, blaCTX-M-2 e blaCTX-M-8. Foi observada a relação clonal entre isolados de E. coli de origem humana e de origem animal. Estes resultados demonstram a disseminação de um problema endêmico em áreas urbanas para uma comunidade, em teoria, com baixa exposição a antibacterianos. / The aim of this study was to investigate the presence of acquired β-lactamase in the commensal Gram-negative microbiota of humans and domestic animals of a remote community in the Amazon Forest region. From March to July 2013 stool samples were collected from individuals attended in a health care center restricted to indigenous communities and from the local staff, and rectal swab samples were collected from companion animals in the community. In the human samples were detected Escherichia coli, Klebsiella pneumoniae, Enterobacter aerogenes, Enterobacter kobei and Morganella morganii isolates harboring blaTEM-1, blaCTX-M-15, blaCTX-M-14, blaCTX-M-8 e blaGES-5 genes. In the animal samples only E. coli strains harboring blaTEM-1, blaCTX-M-14, blaCTX-M-2 and blaCTX-M-8 were detected. The clonal relatedness between E. coli strains from human and animal samples was observed. These results demonstrate the dissemination of an urban endemic problem to a community, in theory, with low antimicrobial exposure.

Campylobacter jejuni é a espécie bacteriana mais comumente relacionada como causa de gastroenterite em humanos em vários países. Porém, o isolamento e o estudo de C. jejuni não são muito frequentes no Brasil, o que dificulta avaliar a dimensão dessa bactéria como causadora de doença em humanos e animais, bem como, determinar o impacto de sua presença em alimentos e no meio-ambiente. O objetivo desse trabalho foi avaliar a diversidade genética por cinco diferentes técnicas de tipagem molecular, o potencial patogênico pela pesquisa de 16 genes de virulência por PCR e o perfil de resistência pela concentração inibitória mínima por Etest® frente a quatro antimicrobianos e pela análise in silico de genes de resistência e pontos de mutação de linhagens de C. jejuni isoladas no Brasil. Foram estudadas 121 linhagens de C. jejuni isoladas de humanos (51), animais (35), alimentos (33) e ambiente (02) nos estados de Minas Gerais, São Paulo, Rio de Janeiro e Rio Grande do Sul, no período de 1996 a 2016. Todas as linhagens apresentaram os genes flaA, flhA, iamA, docA, ciaB, cdtA, cdtB, cdtC, racR, dnaJ, pldA, cadF, sodB e csrA. O gene wlaN foi detectado em 15 linhagens, e uma linhagem apresentou o gene virB11. Dentre as 121 linhagens estudadas, 68 linhagens foram resistentes a pelo menos um dos antimicrobianos testados. A resistência à ciprofloxacina, doxiciclina, tetraciclina e eritromicina foi observada em 43,8%, 34,7%, 34,7% e 4,9% das linhagens, respectivamente. O dendrograma de similaridade genética de Pulsed field gel electrophoresis (PFGE) agrupou as 121 linhagens estudadas em três grupos com similaridade genômica de 46,9% entre eles. Apesar da alta diversidade genômica entre as linhagens estudadas, algumas linhagens isoladas de diferentes fontes, locais e anos, apresentaram uma similaridade genotípica acima de 80% entre elas e, foram agrupadas em 21 subgrupos. Pelas sequências da SVR do gene flaA as linhagens estudadas foram agrupadas em dois grupos com linhagens isoladas de fontes clínicas e não clínicas e de humanos e animais com similaridade acima de 80,9 % entre elas e tipadas em 40 SVR-flaA alelos, sendo os alelos 57, 49 e 45 os mais frequentemente detectados. A análise do locus CRISPR por HRMA tipou as linhagens de C. jejuni em 23 diferentes variantes sendo que algumas variantes continham linhagens de origem clínica e não clínica e de humanos e animais. A árvore de SNPs gerada a partir dos dados do sequenciamento do genoma completo alocou as 116 linhagens sequenciadas em dois principais grupos. O grupo SNP-A agrupou 97 linhagens e o grupo SNP-B agrupou 19 linhagens, com linhagens de fontes clínicas e não clínicas e de humanos e animais, respectivamente. A técnica de Multilocus sequence typing (MLST) tipou as 116 linhagens de C. jejuni em 46 STs, e não foi observada a predominância de um ST. O índice de discriminação das metodologias de análise de SNPs no genoma completo, PFGE, MLST, sequenciamento das SVR do gene flaA e análise do locus CRISPR por HRMA foi 1,0, 0,982, 0,941, 0,939 e 0,874, respectivamente. Na análise in silico de genes de resistência e pontos de mutação, 95 linhagens apresentaram ao menos um gene de resistência ou ponto de mutação conhecido, sendo que a porcentagem de correlação entre os resultados de resistência fenotípicos e genotípicos foi maior que 66,7%; 94,6% e 96,8% para eritromicina, tetraciclina e ciprofloxacina, respectivamente. Conclui-se que a alta frequência da maioria dos genes de virulência pesquisados evidenciou o potencial patogênico das linhagens de C. jejuni estudadas. A resistência a antimicrobianos de primeira escolha utilizados para o tratamento da campylobacteriose encontrada nas linhagens estudadas é preocupante, podendo levar à falha terapêutica quando o tratamento é necessário. Os resultados obtidos pelas metodologias de tipagem molecular realizadas sugerem que uma possível contaminação possa ter ocorrido entre fontes clínicas e não clínicas e entre humanos e animais, ao longo de 20 anos no Brasil. Pelo índice de discriminação, foi observado que as metodologias de análise de SNPs no genoma completo e PFGE, em comparação com as outras técnicas de tipagem, foram as mais eficientes em discriminar as linhagens de C. jejuni do presente estudo. / Campylobacter jejuni is the most commonly bacterial species related as a cause of gastroenteritis in humans in several countries. However, the isolation and the study of C. jejuni have not been very frequently in Brazil, which makes it difficult to evaluate the involvement of this bacterium as a cause of diseases in humans and animals, as well as to determine the impact of its presence in food and the environment. The aim of this study was to evaluate the genetic diversity by five different molecular typing techniques, the pathogenic potential by searching for the presence of 16 virulence genes by PCR and the resistance profile by the minimum inhibitory concentration by Etest® against four antibiotics and by the in silico analyses of resistance genes and mutation points of C. jejuni strains isolated in Brazil. A total of 121 C. jejuni strains isolated from humans (51), animals (35), food (33) and the environment (02) in the States of Minas Gerais, Sao Paulo, Rio de Janeiro and Rio Grande do Sul, between 1996 to 2016 were studied. All strains presented the genes flaA, flhA, iamA, docA, ciaB, cdtA, cdtB, cdtC, racR, dnaJ, pldA, cadF, sodB and csrA. The wlaN gene was detected in 15 strains, and one strain presented the virB11 gene. Among the 121 strains studied, 68 strains were resistant to at least one of the antibiotics tested. Resistance to ciprofloxacin, doxycycline, tetracycline and erythromycin was observed in 43.8%, 34.7%, 34.7% and 4.9% of the strains, respectively. The Pulsed field gel electrophoresis (PFGE) dendrogram of genetic similarity clustered the 121 strains studied in three groups with a genomic similarity of 46.9% among them. Despite the high genomic diversity among the strains studied, some strains isolated from different sources, places and years, presented a genotypic similarity above 80% among them and were grouped into 21 subgroups. By flaA-SVR sequencing the strains studied were clustered into two groups with strains isolated from clinical and non-clinical sources and from humans and animals with a similarity above 80.9% among them and typed in 40 flaA-SVR alleles, being the alleles 57, 49 and 45 the most frequently detected. The analysis of the CRISPR locus by HRMA typed the C. jejuni strains in 23 different variants, with some variants containing strains from clinical and non-clinical origin and from humans and animals. The SNP tree generated from the whole genome sequencing data grouped the 116 strains sequenced into two major groups. SNP-A grouped 97 strains and SNP-B grouped 19 strains, with strains from clinical and non-clinical sources and from humans and animals, respectively. Multilocus sequence typing (MLST) technique typed the 116 C. jejuni strains in 46 STs, and it was not observed a predominant ST. The discrimination index of the analysis of SNPs in the whole genome, PFGE, MLST, flaA-SVR sequencing and analysis of the CRISPR locus by HRMA was 1.0, 0.982, 0.941, 0.939 and 0.874, respectively. In the in silico analyses of resistance genes and mutation points, 95 strains showed at least one resistance gene or known mutation point, and the percentage of correlation between phenotypic and genotypic resistance results was greater than 66.7%; 94.6% and 96.8% for erythromycin, tetracycline and ciprofloxacin, respectively. In conclusion, the high frequency of the majority of the virulence genes studied highlighted the pathogenic potential of the C. jejuni strains studied. Resistance to antimicrobials of first choice used for the treatment of campylobacteriosis found in the strains studied is worrying and may lead to therapeutic failure when treatment is required. The results obtained by the molecular typing methodologies performed suggest that a possible contamination may have occurred between clinical and non-clinical sources and between humans and animals over 20 years in Brazil. By the discrimination index, it was observed that the methodologies of analysis of SNPs in the whole genome and PFGE, in comparison to the other typing techniques, were the most efficients in discriminating the C. jejuni strains of the present study.

Salmonella spp. é reconhecida como uma das bactérias que mais causam doenças de origem alimentar no mundo. Dentre as diversas sorovariedades de Salmonella, a Typhimurium é uma das sorovariedades de maior ocorrência no mundo. Várias metodologias de tipagem fenotípicas e genotípicas foram desenvolvidas com o intuito de se delinear a epidemiologia e diversidade genotípica de Salmonella Typhimurium. Entretanto, a tipagem fenotípica é muitas vezes limitada por sua baixa capacidade de diferenciação de subtipos pertencentes a uma mesma sorovariedade de Salmonella, um problema minimizado pelos métodos genotípicos. No Brasil, foram realizados poucos estudos que genotiparam linhagens de S. Typhimurium. Os objetivos deste estudo foram caracterizar linhagens de S. Typhimurium isoladas de humanos, alimentos, animais e ambiente do animal no Brasil quanto ao seu potencial patogênico, perfil de resistência a antimicrobianos e diversidade genotípica. Foram estudadas 119 linhagens de S. Typhimurium, isoladas de material clínico de humanos (43), alimentos diversos (49), material clínico de suínos (22) e do ambiente de suínos (5), entre 1983 e 2013, provenientes de várias Estados do Brasil. A presença de 12 genes de virulência foi pesquisada por PCR. O perfil de resistência a 13 antimicrobianos foi realizado pelo método de discodifusão. A tipagem molecular foi realizada por Pulsed-field gel electrophoresis (PFGE), Enterobacterial repetitive intergenic consensus PCR (ERIC-PCR), Multiple-locus variablenumber tandem-repeats analysis (MLVA), Clustered regularly interspaced short palindromic repeats - Multi-virulence locus sequence typing (CRISPR-MVLST), Multilocus sequence typing (MLST) e sequenciamento do genoma completo para 92 linhagens de S. Typhimurium isoladas de humanos (43) e alimentos (46). As metodologias PFGE, ERIC-PCR e MLVA foram realizadas para 70 linhagens de S. Typhimurium isoladas de humanos (43), animais (22) e ambiente do animal (5). Todas as 119 linhagens apresentaram os genes sipA, flgK, flgL e invA. O gene sipD e o gene sopE2 foram encontrados em 118 (99,2%) linhagens. O gene fljB foi encontrado em 117 (98,3%) linhagens. O gene sopD foi presente em 114 (95,8%) linhagens, o gene sopB em 111 (93,3%) linhagens, o gene ssaR em 102 (85,7%) linhagens, o gene sifA em 86 (72,3%) linhagens e 45 (37,8%) linhagens apresentaram o gene plasmidial spvB. De um total de 119 linhagens, 64 (62,2%) linhagens foram resistentes a pelo menos um dos 13 antimicrobianos testados, sendo que 36 (30,3%) linhagens foram multi-droga resistentes (MDR). Na comparação dos isolados de humanos e alimentos, as linhagens isoladas de humanos antes de meados 1990, ficaram alocadas nos grupos PFGE-A, PFGE-B1, PFGE-B2, ERIC-A, ERIC-B, MLVA-A, MLVA-B1, MLVA-B2 e G1, G2, H para CRISPRMVLST. As linhagens isoladas de humanos após esse período ficaram alocadas nos grupos PFGE-B1, ERIC-A, MLVA-B1, MLVA-B2 e G2. As linhagens isoladas de alimentos ficaram alocadas nos grupos PFGE-A, PFGE-B1, ERIC-A, ERIC-B, MLVA-A, MLVA-B1, MLVAB2, G1 e G2. Por MLST, do total de 92 linhagens isoladas de humanos e alimentos, 77 linhagens foram tipadas como ST19. Pelo sequenciamento do genoma completo, as linhagens isoladas de alimentos e humanos ficaram alocadas no grupos I e J independente das datas de isolamento. Na comparação dos isolados de humanos e animais, as linhagens das duas origens ficaram alocadas nos grupos PFGE-D1, PFGE-D2, ERIC-C1, MLVA-C1 e MLVA-D. ii Conclui-se que a grande prevalência de genes de virulência nas linhagens de S. Typhimurium estudadas reforça o potencial das mesmas causarem doenças em humanos, bem como, os riscos de sua presença em alimentos, animais para consumo humano e ambiente. A ocorrência de S. Typhimurium multi-droga resistentes isoladas de alimentos diversos e de suínos para consumo é um alerta para o possível risco de humanos ingerirem alimentos contaminados por tais linhagens. Em conjunto os resultados de PFGE, ERIC-PCR, MLVA, CRISPR-MVLST sugerem que as linhagens de S. Typhimurium isoladas de humanos eram geneticamente mais diversificadas antes de meados de 1990, o que pode sugerir a seleção de um subtipo de S. Typhimurium mais adaptado, depois que Salmonella Enteritidis tornou-se a sorovariedade de maior ocorrência no Brasil após esse período. Com relação às linhagens isoladas de alimentos, os resultados de PFGE, ERIC-PCR, MLVA e CRISPR-MVLST sugerem que durante o período estudado houve a circulação de mais de um subtipo no país. Os resultados de MLST sugerem que tais linhagens tenham uma origem filogenética comum. Os resultados do sequenciamento do genoma completo sugerem que houve a circulação de mais de um subtipo de S. Typhimurium no país, com relação às linhagens de humanos e alimentos. Também alerta para o possível risco de linhagens MDR isoladas de alimentos contaminarem humanos e/ou disseminarem genes de resistência a antibióticos para linhagens de origem clínica e não clínica. Na comparação dos isolados de humanos e animais, os resultados de PFGE, ERICPCR e MLVA sugerem que algumas linhagens isoladas de suínos e humanos podem descender de um subtipo comum. Ademais, as linhagens MDR isoladas de suínos e do ambiente de suínos alertam para o possível risco de porcos usados para consumo contaminarem humanos, o ambiente e outros porcos. / Salmonella spp. is recognized as one of the most involved bacteria that cause food-borne diseases in the world. Among the various serovars of Salmonella, Typhimurium is one of the most frequent serovars worldwide. Several phenotypic and genotypic typing methods have been developed in order to delineate the epidemiology and genotypic diversity of Salmonella Typhimurium. However, phenotypic typing is often limited by its low capacity to differentiate subtypes belonging to the same serovar of Salmonella, a problem minimized by genotypic methods. In Brazil, few studies have been conducted that genotyped S. Typhimurium strains. The aims of this study were to characterize S. Typhimurium strains isolated from humans, food, animals and animal\'s environment in Brazil regarding its pathogenic potential, antimicrobial resistance and genotypic diversity. We studied 119 S. Typhimurium strains isolated from human clinical material (43), different foods (49), clinical material from pigs (22) and pigs environment (5), between 1983 and 2013 from various States of Brazil. The presence of 12 virulence genes was investigated by PCR. The resistance profile against 13 antimicrobial was performed by the disk diffusion method. Molecular typing was performed by Pulsed-field gel electrophoresis (PFGE), Enterobacterial repetitive intergenic consensus PCR (ERIC-PCR), Multiple-locus variable-number tandem-repeats analysis (MLVA), Clustered regularly interspaced short palindromic repeats - Multi-virulence locus sequence typing (CRISPR-MVLST), Multilocus sequence typing (MLST) and whole genome sequencing for 92 S. Typhimurium strains isolated from humans (43) and food (46). PFGE, ERIC-PCR and MLVA methods were performed for 70 S. Typhimurium strains isolated from humans (43), animals (22) and the animal\'s environment (5). All 119 strains showed the sipA, flgK, flgL and invA genes. The sipD and sopE2 genes were found in 118 (99.2%) strains. The fljB gene was found in 117 (98.3%) strains. The sopD gene was present in 114 (95.8%) strains, the gene sopB in 111 (93.3%) strains, the ssaR gene in 102 (85.7%) strains, the gene sifA in 86 (72.3%) strains and 45 (37.8%) strains showed the plasmid gene spvB. From a total of 119 strains, 64 (62.2%) strains were resistant to at least one of the 13 antimicrobials tested, and 36 (30.3%) strains were multi-drug resistant (MDR). In the comparison of isolates from humans and food, the strains isolated from humans before mid-1990s were allocated in PFGE-A, PFGE-B1, PFGE-B2, ERIC-A, ERIC-B, MLVA-A, MLVA-B1, MLVA-B2 and G1, G2, H for CRISPR-MVLST. The strains isolated from humans after this period were allocated in PFGE-B1, ERIC-A, MLVA-B1, MLVA-B2 and G2 clusters. The strains isolated from food were allocated in PFGE-A, PFGE-B1, ERIC-A, ERIC-B, MLVA-A, MLVA-B1, MLVA-B2, G1 and G2 clusters. By MLST, of the total of 92 strains isolated from humans and food, 77 strains were typed as ST19. By whole genome sequencing, the strains isolated from food and humans were allocated in I and J clusters independently of its isolation date. In the comparison of isolates from humans and animals, strains of the two origins were allocated in PFGE-D1, PFGE-D2, ERIC-C1, MLVA-C1 and MLVA-D clusters. In conclusion, the high frequency of virulence genes in the S. Typhimurium strains studied reinforces their potential hazard to cause disease in humans, as well as the risk of its presence in food, animals for human consumption and the environment. The occurrence of S. Typhimurium multi-drug iv resistant isolated from various food and pigs for consumption is an alert of the possible risk for humans to ingest contaminated food with those strains. Together the results of PFGE, ERIC-PCR, MLVA e CRISPR-MVLST suggest that S. Typhimurium strains isolated from humans were genetically more diverse before mid-1990s, which might indicate the selection of a more adapted S. Typhimurium subtype after Salmonella Enteritidis became the most prevalent serovar in Brazil. Regarding the strains isolated from food, the results of PFGE, ERIC-PCR, MLVA and CRISPR-MVLST suggest that during the studied period there was circulation of more than one subtype in the country. The MLST results suggest that these strains have a common phylogenetic origin. The results of the whole genome sequencing suggest that there may be more than one subtype circulating in the country, with respect to the strains of human and food origins. Also, alerts for the possible risk of MDR strains isolated from food to contaminate humans and/or disseminate antibiotic resistance genes for strains of clinical and non-clinical origin. In the comparison of isolates from humans and animals, the results of PFGE, ERIC-PCR and MLVA suggest that some strains isolated from pigs and humans may descend from a common subtype. In addition, the MDR strains isolated from pigs and pig environment warn for the possible risk of pigs used for human consumption to contaminate humans, the environment and other pigs.

Campylobacter spp., principalmente as espécies C. coli e C. jejuni, são a causa mais comum de doença bacteriana veiculada por alimentos na Europa, Estados Unidos e alguns outros locais do mundo. No Brasil, há uma escassez de estudos de C. coli, o que dificulta avaliar a dimensão do envolvimento dessa bactéria como causadora de doença nos seres humanos e em animais, bem como, determinar o impacto de sua presença em alimentos e no meio ambiente. O objetivo desse trabalho foi caracterizar molecularmente linhagens de C. coli isoladas de origens diversas no Brasil pela pesquisa da presença de genes relacionados à virulência por PCR, perfil de sensibilidade a antimicrobianos e pela análise da similaridade genotípica por métodos de tipagem molecular. Adicionalmente, o Índice de Discriminação (D) de tais metodologias foi verificado. Foram estudadas 63 linhagens de C. coli, isoladas de humanos (12), animais (21), alimentos (10) e ambiente (20), entre os anos de 1995 e 2011, nos Estados do Rio de Janeiro, São Paulo e Minas Gerais. Todas as linhagens apresentaram os genes flaA, cadF e sodB. O gene cdtB foi detectado em 20 (31,7%) linhagens, o gene flhA foi detectado em 11 (17,5%) linhagens, o gene dnaJ foi encontrado em 10 (15,9%) linhagens, o gene pldA foi detectado em sete (11,1%) linhagens, o gene iamA foi detectado em três (4,8%) linhagens, os genes cdtC e docA foram encontrados em duas (3,2%) linhagens, os genes cdtA e crsA foram encontrados em uma (1,6%) linhagem e os genes ciaB, wlaN, virB11 e racR não foram detectados. Dentre as 63 linhagens estudadas, 42 foram susceptíveis a todos os antimicrobianos testados. Das 21 linhagens resistentes, 10 (15,9%) foram resistentes a tetraciclina e doxaciclina, seis (9,5%) foram resistentes a ciprofloxacina e uma (1,6 %) foi resistente a eritromicina. Somente quatro (6,3%) linhagens foram resistentes a pelo menos duas diferentes classes de antibióticos testados simultaneamente. O dendrograma de similaridade genética de Pulsed field gel electrophoresis (PFGE) agrupou as 63 linhagens estudadas em dois grupos principais denominados PFGE-A e PFGE-B com similaridade genômica de 44,9% entre eles. Entretanto, algumas linhagens isoladas de humanos, animais, ambiente e alimentos apresentaram uma alta similaridade genotípica acima de 80% entre elas e, foram agrupadas em sete subgrupos denominados PFGE-A1 a PFGE-A7. O dendrograma de similaridade genômica das sequências da SVR do gene flaA agrupou as linhagens ii estudadas em dois grupos principais designados SVR-A e SVR-B, com similaridade acima de 83,1 % entre eles. Ademais, o depósito das sequências da SVR do gene flaA no banco de dados online demonstrou que os alelos 30 e o 1647 foram os mais frequentemente encontrados e permitiu a comparação das linhagens estudadas com os alelos descritos no banco de dados. Sete alelos, dentre os 22 encontrados, não haviam sido previamente descritos. A análise do locus CRISPR por HRMA dividiu as linhagens de C. coli em quatro diferentes perfis de melting. O Multilocus sequence typing (MLST) foi utilizado para tipar 20 linhagens de C. coli e foram obtidos 18 STs diferentes dos quais apenas dois já haviam sido previamente descritos. O D das metodologias de PFGE, sequenciamento da SVR do gene flaA, análise do locus CRISPR por HRMA e MLST foi de 0,986, 0,916, 0,550 e 0,989, respectivamente. Pode- se concluir que o potencial patogênico das linhagens de C. coli não foi evidenciado o que pode estar relacionado ao fato da maioria dos estudos envolvendo patogênese terem sido realizados para a espécie C. jejuni. Algumas linhagens apresentaram-se resistentes aos antimicrobianos testados, o que é preocupante uma vez que tais linhagens podem disseminar genes de resistência a outras isoladas de diversas fontes. Os resultados gerados pelos métodos de tipagem molecular por PFGE e sequenciamento da pequena região variável (SVR) do gene flaA demonstraram uma alta similaridade genotípica entre algumas linhagens de C. coli, sugerindo que uma possível contaminação tenha ocorrido entre linhagens isoladas de fontes clínicas e não clínicas ao longo de 16 anos no Brasil. Ademais, a análise dos alelos da SVR do gene flaA nos permitiu concluir que os alelos prevalentes nas linhagens estudadas diferem daqueles encontrados nos países Europeus. Os dados obtidos por MLST sugerem que as linhagens estudadas possuem uma grande diversidade genética entre si e em comparação com as linhagens isoladas em diferentes locais do mundo. Finalmente, as técnicas de MLST e PFGE foram as mais eficientes e adequadas na genotipagem das linhagens de C. coli estudadas. / Campylobacter spp., mainly the C. coli and C. jejuni species, are the most common cause of bacterial disease conveyed by food in Europe, United States, and other places worldwide. In Brazil, there is a paucity of studies on C. coli, which makes it difficult to evaluate the involvement of this bacterium as a cause of diseases in humans and animals, as well as to determine the impact of its presence in food and the environment. The aim of this study was to molecularly characterize C. coli strains isolated from diverse origins in Brazil by searching for the presence of virulence-related genes by PCR, antimicrobial sensitivity profile, and analysis of the genotypic similarity by molecular typing methods. Addicionaly, the Discriminatory Index (D) of those methodologies was acessed. Sixty-three C. coli strains isolated from humans (12), animals (21), food (10), and the environment (20) between 1995 and 2011, in the States of Rio de Janeiro, São Paulo, and Minas Gerais were studied. All strains presented the flaA, cadF and sodB genes. The cdtB gene was detected in 20 (31.7%) strains; the flhA gene was detected in 11 (17.5%) strains; the dnaJ gene was detected in 10 (15.9%) strains; the pldA gene was detected in 7 (11.1%) strains ; the iamA gene was detected in three (4.8%) strains; the cdtC and docA genes were found in two (3.2%) strains; the cdtA and crsA were found in one (1.6%) strain and the ciaB, wlaN, virB11 and racR genes were not detected. Among the 63 strains studied, 42 were susceptible to all antimicrobials tested. Of the 21 resistant strains, 10 (15.9%) were resistante to tetracycline and doxaciclyne, six (9.5%) showed resistance to ciprofloxacin, and one (1.6%) was resistant to erythromycin. Only four (6.3%) strains were simultaneously resistant to at least two different classes of the antibiotics tested. The dendrogram of genetic similarity of Pulsed field gel electrophoresis (PFGE) grouped the 63 strains studied into two groups namely PFGE-A and PFGE-B with a genomic similarity of 44.9% among them. However, some strains isolated from humans, animals, the environment and food presented a high genotypic similarity above 80% and were subdivided into seven groups designated as PFGE-A1 to PFGE-A7. The dendrogram of genetic similarity of the SRV-flaA gene sequences grouped the strains studied into two groups namely SVR-A and SVR-B, with similarity above 83.1% among them. Besides, the deposit of the SVR sequences of the flaA gene in the online database showed that the alleles 30 and 1647 were the iv most frequently found and allowed the comparison between the strains studied with the alleles described in the database. Seven alleles, among the 22 found have never been described before. The CRISPR locus analysis divided the C. coli strains into four different melting profiles. The Multilocus sequence typing (MLST) was used to type 20 C. coli strains and revealed 18 different STs among which just two had been previously described. The D of PFGE, SVR- flaA sequence, HRMA of CRISPR locus analysis and MLST was 0.986, 0.916, 0.550 and 0.989, respectively. In conclusion, the pathogenic potential of the C. coli strains was not highlighted, which could be related to the fact that the majority of the pathogenicity studies were performed with C. jejuni species. Some strains showed resistance to the antibiotics tested what is a concern once those strains may spread the resistance genes to other strains isolated from different sources. The results obtained by PFGE and SVR-flaA sequence showed a high genomic similarity among some C. coli strains which may suggest that a possible contamination may have occurred among clinical and non-clinical sources during 16 years in Brazil. Furthermore, the analysis of SVR- flaA alleles allowed the conclusion that the prevalent alleles in the strains studied were different from those found in European countries. The data obtained by MLST suggests that the strains studied had a high genomic diversity among them and in comparison with strains isolated from different places worldwide. Finally, the MLST and PFGE technicques were the most efficient and adequate in genotyping the C. coli strains studied.

La leptospirose est considérée comme la zoonose la plus répandue au monde mais les incidences sont les plus élevées dans les régions tropicales et en particulier sur les îles. Les îles du Sud-Ouest de l'Océan Indien ne dérogent pas à la règle puisque la maladie y représente un problème de santé humaine majeur sur plusieurs îles, notamment aux Seychelles qui enregistrent un des plus fort taux d'incidence humaine au monde. Sur la base des données disponibles, l'épidémiologie humaine apparait contrastée à l'échelle de la région : les cas cliniques sur Mayotte résultent d'infections par quatre espèces de leptospires distinctes alors qu'à La Réunion ou aux Seychelles, une seule espèce est à l'origine de la grande majorité des cas cliniques. L'objectif général de cette thèse est d'identifier certains des déterminants de cette épidémiologie singulière. Nous avons dans un premier temps complété les informations humaines en investiguant la leptospirose en Union des Comores, pays n'ayant jamais rapporté de transmission autochtone. Nos résultats indiquent que les populations humaines y sont exposées à des antigènes de leptospires comparables à ceux retrouvés sur l'île voisine de Mayotte. Ce résultat suggére que l'absence de leptospirose sur certaines îles est le résultat d'un déficit de surveillance. Nous avons ensuite caractérisé la diversité génétique des leptospires au sein de la faune de certaines îles, caractérisées par des niveaux d'endémisme élevés à même d'être en partie à l'origine de cette épidémiologie contrastée. Nous décrivons d'une part une importante diversité des leptospires pathogènes chez les chauves-souris (Chiroptères) malgaches. Nous montrons d'autre part que cette diversité de leptospires n'est pas structurée géographiquement mais présente au contraire une importante spécificité d'hôte, résultant de différents processus évolutifs incluant co-spéciation et host-switch. Nous avons exploité cette spécificité d'hôte pour éclairer l'épidémiologie de la leptospirose à Mayotte, où nous montrons que l'importante diversité bactérienne impliquée dans les cas cliniques résulte de la présence de nombreux réservoirs, dont certains originaires de Madagascar. Ainsi, il apparaît que l'épidémiologie humaine de la leptospirose dans le SOOI est le reflet d'assemblages distincts de leptospires cosmopolites et autochtones/endémiques maintenus et excrétés par des réservoirs animaux particuliers. / Leptospirosis is considered as the most widespread zoonosis worldwide but the incidence levels are higher in tropical regions and particularly on islands. The South-Western Indian Ocean islands are no exception and the disease is of major medical concern in several islands notably in Seychelles, displaying some of the highest human incidence ever reported. Based on available data, the human epidemiology appears contrasted in the region: on Mayotte, human cases result from the infection with four distinct Leptospira species whereas on Reunion Island or Seychelles, a single species causes the majority of clinical cases. The main objective of this thesis is to identify some of the drivers of this singular epidemiology. We first completed the information available on this human disease in the region by investigating the leptospirosis situation in the Union of the Comoros, a country where no autochthonous transmission has been reported so far. Our results indicate that Comorian populations are exposed to Leptospira, which are antigenically comparable to those detected in the neighbouring island of Mayotte. This finding suggests that the apparent absence of leptospirosis in some islands rather reflects a lack of surveillance. We then investigated the genetic diversity of Leptospira on distinct islands home to distinct endemic animal species that may shed distinct Leptospira lineages and thus at least in part explain the contrasted epidemiology of leptospirosis in the region. Specifically, we describe a high diversity of pathogenic Leptospira within Malagasy bats (Chiroptera) and further show that Leptospira diversity is not structured by geography. Instead, we show that these Leptospira display a strong specificity towards their hosts, which may result from different evolutionary processes including co-speciation and host switching. Using this tight host specificity, we investigated the leptospirosis epidemiology on Mayotte, where we show that the important bacterial diversity reported in clinical cases is due of the presence of several distinct animal species acting as reservoirs, some of which introduced from the neighbouring Madagascar. Altogether, results presented herein combined to data produced by our lab suggest that the epidemiology of leptospirosis in the SWOI results from distinct assemblages of cosmopolitan and autochthonous/endemic Leptospira. Keywords : Leptospirosis, Leptospira, Chiroptera, small mammals, hosts-parasites associations, MAT, MLST, co-phylogeny, Madagascar, Mayotte, Union of Comoros, South-Western Indian Ocean.

Antibiotic resistance challenges the practice and development of modern medicine. The aim of this thesis was to test the hypothesis that antibiotic resistance is reversible once the selection pressure of an antibiotic is removed. A decisive reduction (85%) in trimethoprim and trimethoprim-sulfamethoxazole over 24 months in Kronoberg County, Sweden, is described. The resistance baseline prior to the intervention and the effects of the intervention on resistance levels, trimethoprim resistance genes (dfr-genes) and population structure in Escherichia coli were studied. The effects of different algorithms for excluding patient duplicate isolates were small but systematic. An identical algorithm was used throughout. The drastic decrease in the use of trimethoprim containing drugs did not result in a corresponding decrease in trimethoprim resistance. This was true both for total trimethoprim resistance and for trimethoprim mono-resistance. The distributions of E. coli phenotypes, dfr-genes and E. coli sequence types were stable. The marginal effect on resistance rates was explained by a low fitness cost of trimethoprim resistance observed in vitro and the high levels of associated resistance in trimethoprim resistant isolates. Trimethoprim resistance was, although widespread in the E. coli population, more common in certain E. coli sequence types. The distributions of dfr-genes were different in E. coli and K. pneumoniae and between different E. coli sequence types. These results indicate mechanisms related to the genetic back-bone of E coli to be important for the acquisition and persistence of antibiotic resistance. The findings of this thesis indicates that, at least for some classes of antibiotics, we may have overestimated the usefulness of a strategy for reversing antimicrobial resistance based on the fitness cost of resistance. We have equally underestimated the conserving effects of associated resistance. The stability of the dfr-genes and E. coli sequence types underlines the importance of associated resistance and successful lineages in the spread and maintenance of antibiotic resistance in E. coli.

Neisseria meningitidis is a gram negative, and sometimes encapsulated, diplococcus that causes devastating disease worldwide. For the worldwide genetic surveillance of N. meningitidis, the gold standard for profiling the bacterium uses genetic loci found around the genome. Unfortunately, the software for analyzing the data for these profiles is difficult to use for a variety of reasons. This thesis shows my suite of tools called the Meningococcus Genome Informatics Platform for the analysis of these profiling data. To better understand N. meningitidis, the CDC Meningitis Laboratory and other world class laboratories have adopted a whole genome approach. To facilitate this approach, I have developed a computational genomics assembly and annotation pipeline called the CG-Pipeline. It assembles a genome, predicts locations of various features, and then annotates those features. Next, I developed a comparative genomics browser and database called NBase. Using CG-Pipeline and NBase, I addressed two open questions in N. meningitidis research. First, there are N. meningitidis isolates that cause disease but many that do not cause disease. What is the genomic basis of disease associated versus asymptomatically carried isolates of N. meningitidis? Second, some isolates' capsule type cannot be easily determined. Since isolates are grouped into one of many serogroups based on this capsule, which aids in epidemiological studies and public health response to N. meningitidis, often an isolate cannot be grouped. Thus the question is what is the genomic basis of nongroupability? This thesis addresses both of these questions on a whole genome level.

The possibility of identifying single nucleotide polymorphisms (SNPs) that would be useful for rapid bacterial typing was investigated. Neisseria meningitidis was the organism chosen for modelling the approach since informative SNPs could be found amongst the sequence data available for multi-locus sequence typing (MLST) at http://www.mlst.net. The hypothesis tested was that a small number of SNPs located within the seven gene fragments sequenced for MLST provide information equivalent to MLST. Preliminary investigations revealed that a small number of SNPs could be utilised to highly discriminate sequence types (STs) of clinical interest. Laboratory procedures demonstrated that SNP fingerprinting of N. meningitidis isolates is achievable. Further tests showed that laboratory identification of a defining SNP in the genome of isolates was to be a practical method of obtaining relevant typing information. Identification of the most discriminating SNPs amongst the ever-increasing amount of MLST sequence data summoned the need for computer-based assistance. Two methods of SNP selection devised by the author of this thesis were translated into computer-based algorithms by contributing team members. Software for two computer programs was produced. The algorithms facilitate the optimal selection of SNPs useful for (1) distinguishing specific STs and (2) differentiating non-specific STs. Current input information can be obtained from the MLST database and consequently the programs can be applied to any bacterial species for which MLST data have been entered. The two algorithms for the selection of SNPs were designed to serve contrasting purposes. The first of these was to determine the ST identity of isolates from an outbreak of disease. In this case, isolates would be tested for their membership to any of the STs known to be associated with disease. It was shown that one SNP per ST could distinguish each of four hyperinvasive STs of N. meningitidis from between 92.5% and 97.5% of all other STs. With two SNPs per ST, between 96.7% and 99.0% discrimination is achieved. The SNPs were selected from MLST loci with the assistance of the first algorithm which scores SNPs according to the number of base mismatches in a sequence alignment between an allele of an ST of interest and alleles belonging to all other STs at a specified locus. The second purpose was to determine whether or not isolates from different sources belong to the same ST, regardless of their actual ST identity. It was shown that with seven SNPs, four sample STs of N. meningitidis could, on average, be discriminated from 97.1% of all other STs. The SNPs were selected with the aid of the second algorithm which scores SNPs at MLST loci for the relative frequency of each nucleotide base in a sequence alignment as a measure of the extent of their polymorphism. A third algorithm for selecting SNPs has been discussed. By altering the method of scoring SNPs, it is possible to overcome the limitations inherent in the two algorithms that were utilised for finding SNPs. In addition, the third approach caters for finding SNPs that distinguish members of a complex from non-members.