Grégory Dubourg

Culture of previously uncultured members of the human gut microbiota by culturomics

Metagenomics revolutionized the understanding of the relations among the human microbiome, health and diseases, but generated a countless number of sequences that have not been assigned to a known microorganism1. The pure culture of prokaryotes, neglected in recent decades, remains essential to elucidating the role of these organisms2. We recently introduced microbial culturomics, a culturing approach that uses multiple culture conditions and matrix-assisted laser desorption/ionization–time of flight and 16S rRNA for identification2. Here, we have selected the best culture conditions to increase the number of studied samples and have applied new protocols (fresh-sample inoculation; detection of microcolonies and specific cultures of Proteobacteria and microaerophilic and halophilic prokaryotes) to address the weaknesses of the previous studies3–5. We identified 1,057 prokaryotic species, thereby adding 531 species to the human gut repertoire: 146 bacteria known in humans but not in the gut, 187 bacteria and 1 archaea not previously isolated in humans, and 197 potentially new species. Genome sequencing was performed on the new species. By comparing the results of the metagenomic and culturomic analyses, we show that the use of culturomics allows the culture of organisms corresponding to sequences previously not assigned. Altogether, culturomics doubles the number of species isolated at least once from the human gut.

High-level colonisation of the human gut by Verrucomicrobia following broad-spectrum antibiotic treatment

The gut microbiota is mainly composed of the phyla Firmicutes, Bacteroidetes, Actinobacteria and Proteobacteria; the Verrucomicrobia phylum is occasionally observed. Antibiotics can change the bacterial diversity of the gut, with limited changes in the proportions of phyla. In this study, the gut repertoire of two patients who received a broad-spectrum antibiotic regimen was studied. As part of a large gut microbiota study, two stool samples were analysed: one sample was collected after broad-spectrum antibiotic therapy in a patient with Coxiella burnetii vascular infection (Patient A); and the other sample was collected from a patient admitted to the Intensive Care Unit (Patient B). Samples were subjected to Gram staining, electron microscopy, 16S rRNA V6 amplicon pyrosequencing and fluorescence in situ hybridisation (FISH). In parallel, the antibiotic susceptibility of Akkermansia muciniphila Muc(T) strain was studied and this strain was observed by electron microscopy. Pyrosequencing revealed that a large proportion of the sequences were associated with Verrucomicrobia (proportions of 44.9% and 84.6% for Patients A and B, respectively). All of the phylotypes were represented by a single species (A. muciniphila), and neither patient presented significant gastrointestinal disorders. Electron microscopy and FISH with specific Verrucomicrobia probes confirmed the presence of the bacterium. The Muc(T) strain was susceptible to imipenem and doxycycline but resistant to vancomycin and metronidazole. Dramatic colonisation of the human gut microbiota by the Verrucomicrobia phylum following a broad-spectrum antibiotic regimen occurred without significant gastrointestinal manifestations, suggesting that influenced by external factors such as antibiotics, the gut repertoire remains partially unknown.

Clinical detection and characterization of bacterial pathogens in the genomics era.

The availability of genome sequences obtained using next-generation sequencing (NGS) has revolutionized the field of infectious diseases. Indeed, more than 38,000 bacterial and 5,000 viral genomes have been sequenced to date, including representatives of all significant human pathogens. These tremendous amounts of data have not only enabled advances in fundamental biology, helping to understand the pathogenesis of microorganisms and their genomic evolution, but have also had implications for clinical microbiology. Here, we first review the current achievements of genomics in the development of improved diagnostic tools, including those that are now available in the clinic, such as the design of PCR assays for the detection of microbial pathogens, virulence factors or antibiotic-resistance determinants, or the design of optimized culture media for ‘unculturable’ pathogens. We then review the applications of genomics to the investigation of outbreaks, either through the design of genotyping assays or the direct sequencing of the causative strains. Finally, we discuss how genomics might change clinical microbiology in the future.

Culturomics and pyrosequencing evidence of the reduction in gut microbiota diversity in patients with broad-spectrum antibiotics

The human gut flora is currently widely characterised using molecular techniques. Microbial culturomics (large-scale culture conditions with identification of colonies using MALDI-TOF or 16S rRNA) is part of the rebirth of bacterial culture that was initiated by environmental microbiologists for the design of axenic culture for intracellular bacteria in clinical microbiology. Culturomics was performed on four stool samples from patients treated with large-scale antibiotics to assess the diversity of their gut flora in comparison with other culture-dependent studies. Pyrosequencing of the V6 region was also performed and was compared with a control group. Gut richness was also estimated by bacterial counting after microscopic observation. In total, 77 culture conditions were tested and 32,000 different colonies were generated; 190 bacterial species were identified, with 9 species that had not been isolated from the human gut before this study, 7 newly described in humans and 8 completely new species. A dramatic reduction in diversity was observed for two of the four stool samples for which antibiotic treatment was prolonged and uninterrupted. The total number of bacteria was generally preserved, suggesting that the original population was replaced but was sustained in size. Discordances between culture and pyrosequencing biodiversity biomarkers highlight the depth of bias of molecular studies. Stool samples studied showed a dramatic reduction in bacterial diversity. Considering the variable antibiotic concentration in the gut, this reduction in the number of species is possibly linked to the production of bacteriocin in the upper digestive tract by specific bacteria, such as Lactobacillus spp.

Investigation of Acinetobacter baumannii resistance to carbapenems in Marseille hospitals, south of France: a transition from an epidemic to an endemic situation

Carbapenem‐resistant Acinetobacter baumannii infections are a worldwide endemic nosocomial threat. Between December 2010 and April 2011, an increase of carbapenem‐resistant A. baumannii infections occurred in several Marseille University Hospitals. The aim of this study was to investigate the increase of carbapenem‐resistant A. baumannii infections and to characterize the mechanisms of carbapenem resistance. The increase was detected by a homemade computer surveillance program, known as EPIMIC, that monitors antibiotic resistance profiles on a weekly basis. During this period, positive samples of carbapenem‐resistant A. baumannii were retrieved from patients hospitalized in different units. Genotyping of the isolates was performed using pulsed‐field gel electrophoresis (PFGE) and multi‐locus sequence typing (MLST), and carbapenemase gene analyses were performed to detect the presence of carbapenemases and to determine the relationships of the isolates. Carbapenem‐resistant A. baumannii were isolated in a total of 11 patients who were hospitalized in different hospitals units. We identified the presence of the blaOXA23‐like carbapenemase‐encoding gene in all of the isolates and found four major PFGE groups and different MLST groups. These results demonstrate a current evolution in the A. baumannii epidemiology in Marseille with a switch from an epidemic situation to an endemic situation and with several circulating clones.

MALDI-TOF identification of the human gut microbiome in people with and without diarrhea in Senegal

Background In Africa, there are several problems with the specific identification of bacteria. Recently, MALDI-TOF mass spectrometry has become a powerful tool for the routine microbial identification in many clinical laboratories. Methodology/Principal Findings This study was conducted using feces from 347 individuals (162 with diarrhea and 185 without diarrhea) sampled in health centers in Dakar, Senegal. Feces were transported from Dakar to Marseille, France, where they were cultured using different culture conditions. The isolated colonies were identified using MALDI-TOF. If a colony was unidentified, 16S rRNA sequencing was performed. Overall, 2,753 isolates were tested, allowing for the identification of 189 bacteria from 5 phyla, including 2 previously unknown species, 11 species not previously reported in the human gut, 10 species not previously reported in humans, and 3 fungi. 2,718 bacterial isolates (98.8%) out of 2,750 yielded an accurate identification using mass spectrometry, as did the 3 Candida albicans isolates. Thirty-two bacterial isolates not identified by MALDI-TOF (1.2%) were identified by sequencing, allowing for the identification of 2 new species. The number of bacterial species per fecal sample was significantly higher among patients without diarrhea (8.6±3) than in those with diarrhea (7.3±3.4; P = 0.0003). A modification of the gut microbiota was observed between the two groups. In individuals with diarrhea, major commensal bacterial species such as E. coli were significantly decreased (85% versus 64%), as were several Enterococcus spp. (E. faecium and E. casseliflavus) and anaerobes, such as Bacteroides spp. (B. uniformis and B. vulgatus) and Clostridium spp. (C. bifermentans, C. orbiscindens, C. perfringens, and C. symbosium). Conversely, several Bacillus spp. (B. licheniformis, B. mojavensis, and B. pumilus) were significantly more frequent among patients with diarrhea. Conclusions/Significance MALDI-TOF is a potentially powerful tool for routine bacterial identification in Africa, allowing for a quick identification of bacterial species.

Patterns of Kingella kingae Disease Outbreaks.

Background: Kingella kingae outbreaks occur sporadically in childcare centers but remain poorly understood and difficult to identify. Methods: To provide the basis of a better knowledge of K. kingae outbreaks patterns that may help to guide identification and management strategies, we collected epidemiological, clinical and laboratory data from all reported K. kingae outbreaks, and those from 2 new Israel outbreaks in 2014. Results: Nine outbreaks were identified in the USA, Israel and France from 2003 to 2014. Twenty-seven children with a median age of 14 ± 4.1 months were affected, male:female ratio of 1.4:1. Outbreaks demonstrated seasonal patterns from the 10th to the 45th weeks, a mean duration of 13.1 ± 8.4 days, a mean attack rate of 17.3 ± 5.1% and a case-fatality rate of 3.7% (1/27). Seventy-four percentage of children had fever (20/27), and the mean values of white blood cell count and C-reactive protein level were 14.6 ± 4.5 × 109/L and 23.8 ± 24.1 mg/L, respectively. Osteoarticular infections accounted for 88.9% of cases (24/27), bacteremia 7.4% (2/27), endocarditis 3.7% (1/27) and meningitis 3.7% (1/27). Specific real-time polymerase chain reaction demonstrated higher performance than culture methods in the diagnosis of case patients and investigations of oropharyngeal K. kingae carriage among close contacts, and multilocus sequence typing methods revealed that ST-6 and ST-25 invasive strains were responsible for multiple country-dependent outbreaks. Coviral infections were identified in the majority of K. kingae outbreaks, notably those causing oral ulcers. Conclusions: K. kingae outbreaks displayed severe K. kingae diseases that were poorly confirmed with culture methods. We argue for the use of genomic technologies to investigate further K. kingae outbreaks.

Scalp eschar and neck lymphadenopathy after tick bite: an emerging syndrome with multiple causes

The clinical and epidemiological features of 56 patients with scalp eschar associated with neck lymphadenopathy after a tick bite (SENLAT) syndrome were evaluated at the National French Rickettsial Center. Eschar swabs, crusts, and biopsies as well as ticks and blood samples were acquired for molecular and serological assays. SENLAT predominantly affects children (p < 0.05), followed by 40- to 70-year-olds, and it is found mostly in women (p < 0.05). The seasonal distribution has two peaks: one in the spring (55%) and one in the autumn (30%). The etiological agent was identified in 18 cases, which include Rickettsia slovaca in 13 cases with incidences of two co-infections with Rickettsia raoultii and one case caused by Rickettsia sibirica mongolitimonae. Other possible agents that were found in attached ticks were Candidatus R. rioja, Coxiella burnetii, and Borrelia burgdorferi. The tick vector was Dermacentor marginatus in almost all cases, with the exception of one case, in which Ixodes ricinus was identified as the vector. Our findings show that SENLAT is a clinical entity characterized as a local infection controlled by the immune system and is neither pathogen- nor vector-specific.

Clostridium difficile 027 emerging outbreak in Marseille, France.

Author(s): Jean-Christophe Lagier, MD, PhD; Gregory Dubourg, PharmD; Nadim Cassir, MD; Pierre-Edouard Fournier, MD, PhD; Philippe Colson, PharmD, PhD; Hervé Richet, MD, PhD; Philippe Brouqui, MD, PhD; Didier Raoult, MD, PhD Source: Infection Control and Hospital Epidemiology, Vol. 34, No. 12 (December 2013), pp. 1339-1341 Published by: The University of Chicago Press on behalf of The Society for Healthcare Epidemiology of America Stable URL: http://www.jstor.org/stable/10.1086/673995 . Accessed: 28/06/2014 14:14

Emerging methodologies for pathogen identification in positive blood culture testing

Bloodstream infections (BSIs) represent a major cause of death in developed countries and are associated with long-term loss of functions. Blood culture remains the gold standard for BSI diagnosis, as it is easy to perform and displays a good analytical sensitivity. However, its major drawback remains the long turnaround time, which can result in inappropriate therapy, fall of survival rate, emergence of antibiotic resistance and increase of medical costs. Over the last 10 years, molecular tools have been the alternative to blood cultures, allowing early identification of pathogens involved in sepsis, as well detection of critical antibiotic resistance genes. Besides, the advent of MALDI-TOF revolutionized practice in routine microbiology significantly reduced the time to result. Reviewed here are recent improvements in early BSI diagnosis and these authors’ view for the future is presented, including innovative high-throughput technologies.