Antibiotic resistance determination using Enterococcus faecium whole-genome sequences: a diagnostic accuracy study using genotypic and phenotypic data

BACKGROUND

DNA sequencing could become an alternative to in vitro antibiotic susceptibility testing (AST) methods for determining antibiotic resistance by detecting genetic determinants associated with decreased antibiotic susceptibility. Here, we aimed to assess and improve the accuracy of antibiotic resistance determination from Enterococcus faecium genomes for diagnosis and surveillance purposes.

METHODS

In this retrospective diagnostic accuracy study, we first conducted a literature search in PubMed on Jan 14, 2021, to compile a catalogue of genes and mutations predictive of antibiotic resistance in E faecium. We then evaluated the diagnostic accuracy of this database to determine susceptibility to 12 different, clinically relevant antibiotics using a diverse population of 4382 E faecium isolates with available whole-genome sequences and in vitro culture-based AST phenotypes. Isolates were obtained from various sources in 11 countries worldwide between 2000 and 2018. We included isolates tested with broth microdilution, Vitek 2, and disc diffusion, and antibiotics with at least 50 susceptible and 50 resistant isolates. Phenotypic resistance was derived from raw minimum inhibitory concentrations and measured inhibition diameters, and harmonised primarily using the breakpoints set by the European Committee on Antimicrobial Susceptibility Testing. A bioinformatics pipeline was developed to process raw sequencing reads, identify antibiotic resistance genetic determinants, and report genotypic resistance. We used our curated database, as well as ResFinder, AMRFinderPlus, and LRE-Finder, to assess the accuracy of genotypic predictions against phenotypic resistance.

FINDINGS

We curated a catalogue of 228 genetic markers involved in resistance to 12 antibiotics in E faecium. Very accurate genotypic predictions were obtained for ampicillin (sensitivity 99·7% [95% CI 99·5-99·9] and specificity 97·9% [95·8-99·0]), ciprofloxacin (98·0% [96·4-98·9] and 98·8% [95·9-99·7]), vancomycin (98·8% [98·3-99·2] and 98·8% [98·0-99·3]), and linezolid resistance (after re-testing false negatives: 100·0% [90·8-100·0] and 98·3% [97·8-98·7]). High sensitivity was obtained for tetracycline (99·5% [99·1-99·7]), teicoplanin (98·9% [98·4-99·3]), and high-level resistance to aminoglycosides (97·7% [96·6-98·4] for streptomycin and 96·8% [95·8-97·5] for gentamicin), although at lower specificity (60-90%). Sensitivity was expectedly low for daptomycin (73·6% [65·1-80·6]) and tigecycline (38·3% [27·1-51·0]), for which the genetic basis of resistance is not fully characterised. Compared with other antibiotic resistance databases and bioinformatic tools, our curated database was similarly accurate at detecting resistance to ciprofloxacin and linezolid and high-level resistance to streptomycin and gentamicin, but had better sensitivity for detecting resistance to ampicillin, tigecycline, daptomycin, and quinupristin-dalfopristin, and better specificity for ampicillin, vancomycin, teicoplanin, and tetracycline resistance. In a validation dataset of 382 isolates, similar or improved diagnostic accuracies were also achieved.

INTERPRETATION

To our knowledge, this work represents the largest published evaluation to date of the accuracy of antibiotic susceptibility predictions from E faecium genomes. The results and resources will facilitate the adoption of whole-genome sequencing as a tool for the diagnosis and surveillance of antimicrobial resistance in E faecium. A complete characterisation of the genetic basis of resistance to last-line antibiotics, and the mechanisms mediating antibiotic resistance silencing, are needed to close the remaining sensitivity and specificity gaps in genotypic predictions.

FUNDING

Wellcome Trust, UK Department of Health, British Society for Antimicrobial Chemotherapy, Academy of Medical Sciences and the Health Foundation, Medical Research Council Newton Fund, Vietnamese Ministry of Science and Technology, and European Society of Clinical Microbiology and Infectious Disease.

Comparing genomic variant identification protocols for Candida auris

Genomic analyses are widely applied to epidemiological, population genetic and experimental studies of pathogenic fungi. A wide range of methods are employed to carry out these analyses, typically without including controls that gauge the accuracy of variant prediction. The importance of tracking outbreaks at a global scale has raised the urgency of establishing high-accuracy pipelines that generate consistent results between research groups. To evaluate currently employed methods for whole-genome variant detection and elaborate best practices for fungal pathogens, we compared how 14 independent variant calling pipelines performed across 35 Candida auris isolates from 4 distinct clades and evaluated the performance of variant calling, single-nucleotide polymorphism (SNP) counts and phylogenetic inference results. Although these pipelines used different variant callers and filtering criteria, we found high overall agreement of SNPs from each pipeline. This concordance correlated with site quality, as SNPs discovered by a few pipelines tended to show lower mapping quality scores and depth of coverage than those recovered by all pipelines. We observed that the major differences between pipelines were due to variation in read trimming strategies, SNP calling methods and parameters, and downstream filtration criteria. We calculated specificity and sensitivity for each pipeline by aligning three isolates with chromosomal level assemblies and found that the GATK-based pipelines were well balanced between these metrics. Selection of trimming methods had a greater impact on SAMtools-based pipelines than those using GATK. Phylogenetic trees inferred by each pipeline showed high consistency at the clade level, but there was more variability between isolates from a single outbreak, with pipelines that used more stringent cutoffs having lower resolution. This project generated two truth datasets useful for routine benchmarking of C. auris variant calling, a consensus VCF of genotypes discovered by 10 or more pipelines across these 35 diverse isolates and variants for 2 samples identified from whole-genome alignments. This study provides a foundation for evaluating SNP calling pipelines and developing best practices for future fungal genomic studies.

Rapid Genomic Characterization and Global Surveillance of Klebsiella Using Pathogenwatch

BACKGROUND

Klebsiella species, including the notable pathogen K. pneumoniae, are increasingly associated with antimicrobial resistance (AMR). Genome-based surveillance can inform interventions aimed at controlling AMR. However, its widespread implementation requires tools to streamline bioinformatic analyses and public health reporting.

METHODS

We developed the web application Pathogenwatch, which implements analytics tailored to Klebsiella species for integration and visualization of genomic and epidemiological data. We populated Pathogenwatch with 16 537 public Klebsiella genomes to enable contextualization of user genomes. We demonstrated its features with 1636 genomes from 4 low- and middle-income countries (LMICs) participating in the NIHR Global Health Research Unit (GHRU) on AMR.

RESULTS

Using Pathogenwatch, we found that GHRU genomes were dominated by a small number of epidemic drug-resistant clones of K. pneumoniae. However, differences in their distribution were observed (eg, ST258/512 dominated in Colombia, ST231 in India, ST307 in Nigeria, ST147 in the Philippines). Phylogenetic analyses including public genomes for contextualization enabled retrospective monitoring of their spread. In particular, we identified hospital outbreaks, detected introductions from abroad, and uncovered clonal expansions associated with resistance and virulence genes. Assessment of loci encoding O-antigens and capsule in K. pneumoniae, which represent possible vaccine candidates, showed that 3 O-types (O1-O3) represented 88.9% of all genomes, whereas capsule types were much more diverse.

CONCLUSIONS

Pathogenwatch provides a free, accessible platform for real-time analysis of Klebsiella genomes to aid surveillance at local, national, and global levels. We have improved representation of genomes from GHRU participant countries, further facilitating ongoing surveillance.

A community-driven resource for genomic epidemiology and antimicrobial resistance prediction of Neisseria gonorrhoeae at Pathogenwatch

BACKGROUND

Antimicrobial-resistant (AMR) Neisseria gonorrhoeae is an urgent threat to public health, as strains resistant to at least one of the two last-line antibiotics used in empiric therapy of gonorrhoea, ceftriaxone and azithromycin, have spread internationally. Whole genome sequencing (WGS) data can be used to identify new AMR clones and transmission networks and inform the development of point-of-care tests for antimicrobial susceptibility, novel antimicrobials and vaccines. Community-driven tools that provide an easy access to and analysis of genomic and epidemiological data is the way forward for public health surveillance.

METHODS

Here we present a public health-focussed scheme for genomic epidemiology of N. gonorrhoeae at Pathogenwatch ( https://pathogen.watch/ngonorrhoeae ). An international advisory group of experts in epidemiology, public health, genetics and genomics of N. gonorrhoeae was convened to inform on the utility of current and future analytics in the platform. We implement backwards compatibility with MLST, NG-MAST and NG-STAR typing schemes as well as an exhaustive library of genetic AMR determinants linked to a genotypic prediction of resistance to eight antibiotics. A collection of over 12,000 N. gonorrhoeae genome sequences from public archives has been quality-checked, assembled and made public together with available metadata for contextualization.

RESULTS

AMR prediction from genome data revealed specificity values over 99% for azithromycin, ciprofloxacin and ceftriaxone and sensitivity values around 99% for benzylpenicillin and tetracycline. A case study using the Pathogenwatch collection of N. gonorrhoeae public genomes showed the global expansion of an azithromycin-resistant lineage carrying a mosaic mtr over at least the last 10 years, emphasising the power of Pathogenwatch to explore and evaluate genomic epidemiology questions of public health concern.

CONCLUSIONS

The N. gonorrhoeae scheme in Pathogenwatch provides customised bioinformatic pipelines guided by expert opinion that can be adapted to public health agencies and departments with little expertise in bioinformatics and lower-resourced settings with internet connection but limited computational infrastructure. The advisory group will assess and identify ongoing public health needs in the field of gonorrhoea, particularly regarding gonococcal AMR, in order to further enhance utility with modified or new analytic methods.

Globetrotting strangles: the unbridled national and international transmission of Streptococcus equi between horses

The equine disease strangles, which is characterized by the formation of abscesses in the lymph nodes of the head and neck, is one of the most frequently diagnosed infectious diseases of horses around the world. The causal agent, Streptococcus equi subspecies equi, establishes a persistent infection in approximately 10 % of animals that recover from the acute disease. Such 'carrier' animals appear healthy and are rarely identified during routine veterinary examinations pre-purchase or transit, but can transmit S. equi to naïve animals initiating new episodes of disease. Here, we report the analysis and visualization of phylogenomic and epidemiological data for 670 isolates of S. equi recovered from 19 different countries using a new core-genome multilocus sequence typing (cgMLST) web bioresource. Genetic relationships among all 670 S. equi isolates were determined at high resolution, revealing national and international transmission events that drive this endemic disease in horse populations throughout the world. Our data argue for the recognition of the international importance of strangles by the Office International des Épizooties to highlight the health, welfare and economic cost of this disease. The Pathogenwatch cgMLST web bioresource described herein is available for tailored genomic analysis of populations of S. equi and its close relative S. equi subspecies zooepidemicus that are recovered from horses and other animals, including humans, throughout the world. This article contains data hosted by Microreact.

Public health surveillance of multidrug-resistant clones of Neisseria gonorrhoeae in Europe: a genomic survey

BACKGROUND

Traditional methods for molecular epidemiology of Neisseria gonorrhoeae are suboptimal. Whole-genome sequencing (WGS) offers ideal resolution to describe population dynamics and to predict and infer transmission of antimicrobial resistance, and can enhance infection control through linkage with epidemiological data. We used WGS, in conjunction with linked epidemiological and phenotypic data, to describe the gonococcal population in 20 European countries. We aimed to detail changes in phenotypic antimicrobial resistance levels (and the reasons for these changes) and strain distribution (with a focus on antimicrobial resistance strains in risk groups), and to predict antimicrobial resistance from WGS data.

METHODS

We carried out an observational study, in which we sequenced isolates taken from patients with gonorrhoea from the European Gonococcal Antimicrobial Surveillance Programme in 20 countries from September to November, 2013. We also developed a web platform that we used for automated antimicrobial resistance prediction, molecular typing (N gonorrhoeae multi-antigen sequence typing [NG-MAST] and multilocus sequence typing), and phylogenetic clustering in conjunction with epidemiological and phenotypic data.

FINDINGS

The multidrug-resistant NG-MAST genogroup G1407 was predominant and accounted for the most cephalosporin resistance, but the prevalence of this genogroup decreased from 248 (23%) of 1066 isolates in a previous study from 2009-10 to 174 (17%) of 1054 isolates in this survey in 2013. This genogroup previously showed an association with men who have sex with men, but changed to an association with heterosexual people (odds ratio=4·29). WGS provided substantially improved resolution and accuracy over NG-MAST and multilocus sequence typing, predicted antimicrobial resistance relatively well, and identified discrepant isolates, mixed infections or contaminants, and multidrug-resistant clades linked to risk groups.

INTERPRETATION

To our knowledge, we provide the first use of joint analysis of WGS and epidemiological data in an international programme for regional surveillance of sexually transmitted infections. WGS provided enhanced understanding of the distribution of antimicrobial resistance clones, including replacement with clones that were more susceptible to antimicrobials, in several risk groups nationally and regionally. We provide a framework for genomic surveillance of gonococci through standardised sampling, use of WGS, and a shared information architecture for interpretation and dissemination by use of open access software.

FUNDING

The European Centre for Disease Prevention and Control, The Centre for Genomic Pathogen Surveillance, Örebro University Hospital, and Wellcome.

Microreact: visualizing and sharing data for genomic epidemiology and phylogeography

Visualization is frequently used to aid our interpretation of complex datasets. Within microbial genomics, visualizing the relationships between multiple genomes as a tree provides a framework onto which associated data (geographical, temporal, phenotypic and epidemiological) are added to generate hypotheses and to explore the dynamics of the system under investigation. Selected static images are then used within publications to highlight the key findings to a wider audience. However, these images are a very inadequate way of exploring and interpreting the richness of the data. There is, therefore, a need for flexible, interactive software that presents the population genomic outputs and associated data in a user-friendly manner for a wide range of end users, from trained bioinformaticians to front-line epidemiologists and health workers. Here, we present Microreact, a web application for the easy visualization of datasets consisting of any combination of trees, geographical, temporal and associated metadata. Data files can be uploaded to Microreact directly via the web browser or by linking to their location (e.g. from Google Drive/Dropbox or via API), and an integrated visualization via trees, maps, timelines and tables provides interactive querying of the data. The visualization can be shared as a permanent web link among collaborators, or embedded within publications to enable readers to explore and download the data. Microreact can act as an end point for any tool or bioinformatic pipeline that ultimately generates a tree, and provides a simple, yet powerful, visualization method that will aid research and discovery and the open sharing of datasets.

Genome of the Host-Cell Transforming Parasite Theileria annulata Compared with T. parva

Theileria annulata and T. parva are closely related protozoan parasites that cause lymphoproliferative diseases of cattle. We sequenced the genome of T. annulata and compared it with that of T. parva to understand the mechanisms underlying transformation and tropism. Despite high conservation of gene sequences and synteny, the analysis reveals unequally expanded gene families and species-specific genes. We also identify divergent families of putative secreted polypeptides that may reduce immune recognition, candidate regulators of host-cell transformation, and a Theileria-specific protein domain [frequently associated in Theileria (FAINT)] present in a large number of secreted proteins.

Sequencing and analysis of the genome of the Whipple's disease bacterium Tropheryma whipplei

BACKGROUND

Whipple's disease is a rare multisystem chronic infection, involving the intestinal tract as well as various other organs. The causative agent, Tropheryma whipplei, is a Gram-positive bacterium about which little is known. Our aim was to investigate the biology of this organism by generating and analysing the complete DNA sequence of its genome.

METHODS

We isolated and propagated T whipplei strain TW08/27 from the cerebrospinal fluid of a patient diagnosed with Whipple's disease. We generated the complete sequence of the genome by the whole genome shotgun method, and analysed it with a combination of automatic and manual bioinformatic techniques.

FINDINGS

Sequencing revealed a condensed 925938 bp genome with a lack of key biosynthetic pathways and a reduced capacity for energy metabolism. A family of large surface proteins was identified, some associated with large amounts of non-coding repetitive DNA, and an unexpected degree of sequence variation.

INTERPRETATION

The genome reduction and lack of metabolic capabilities point to a host-restricted lifestyle for the organism. The sequence variation indicates both known and novel mechanisms for the elaboration and variation of surface structures, and suggests that immune evasion and host interaction play an important part in the lifestyle of this persistent bacterial pathogen.