Whole genome sequencing identifies bacterial factors affecting transmission of multidrug-resistant tuberculosis in a high-prevalence setting

Avika Dixit, Luca Freschi, Roger Vargas, Róger Calderón, James C. Sacchettini, Francis Drobniewski, Jerome T. Galea, Carmen Contreras, Rosa Yataco, Zibiao Zhang, Leonid Lecca, Sergios‐Orestis Kolokotronis, Barun Mathema, Maha FarhatScientific 2019 · DOI

TL;DR

This study uses whole-genome sequencing (WGS) to provide a high-resolution view of multidrug-resistant tuberculosis (MDR-TB) transmission in a high-burden setting in Peru. The findings reveal that traditional genotyping methods underestimate the age and complexity of TB transmission clusters, and highlight specific bacterial factors that may contribute to the success of MDR-TB strains.

What Problem Does This Paper Address?

Tuberculosis (TB) remains a significant global health threat, with multidrug-resistant TB (MDR-TB) posing a particularly challenging problem. Understanding how TB strains, especially drug-resistant ones, transmit within communities is crucial for effective control strategies. Traditional genotyping methods, like MIRU-VNTR, have been used to identify TB transmission clusters, but their resolution is limited, especially in high-burden settings where TB has been circulating for a long time. This can lead to an underestimation of the age of outbreaks and a less precise understanding of transmission dynamics. The paper aims to address this gap by using WGS to provide a more detailed picture of TB transmission, identify the evolutionary history of drug resistance, and pinpoint bacterial factors that might enhance the transmissibility and success of MDR-TB strains.

Methods

The researchers employed a multi-faceted approach:

  • Whole-Genome Sequencing (WGS): 61 isolates from a large MIRU-VNTR-defined TB cluster in Peru were sequenced. This provided a much higher resolution of genetic variation compared to traditional methods.
  • Phylogenetic Analysis: A time-calibrated phylogenetic tree was constructed using Bayesian methods (BEAST) to estimate the divergence times of the TB strains and the acquisition of drug resistance. This allowed them to infer the evolutionary history of the cluster.
  • Transmission Link Inference: They compared transmission links inferred from epidemiological data (household contacts) with those inferred from genomic data (SNP differences). This helped assess the utility of WGS for identifying transmission events.
  • Comparative Genomics: The genetic makeup of the study cluster was compared against a large database of over 12,000 public Mtb genomes to identify bacterial factors (single nucleotide polymorphisms or SNPs) that were unique to the cluster and potentially under positive selection in other TB lineages.
  • Drug Susceptibility Testing (DST): Phenotypic drug resistance was determined for the isolates, and this was correlated with genotypic findings to understand the emergence and acquisition of resistance.

Datasets

  • Study Dataset: 61 whole-genome sequences of Mycobacterium tuberculosis (Mtb) isolates from a household-based TB transmissibility study in northern Lima, Peru. These isolates belonged to a large MIRU-VNTR cluster of 148 strains.
  • Public Database: A dataset of 12,032 public Mtb genomes curated from the literature and the NCBI ReseqTB initiative was used for comparative genomic analysis to identify bacterial factors under positive selection.

Key Insights and Ideas

  • MIRU-VNTR Underestimates Cluster Age: The study demonstrates that clusters identified by MIRU-VNTR typing can be much older than previously thought, circulating for decades in high-burden settings. This challenges the assumption that such clusters represent recent outbreaks.
  • WGS Reveals Complex Evolutionary History: WGS provided a much finer-grained view of the genetic diversity within the MIRU-VNTR cluster, revealing complex evolutionary patterns and multiple independent acquisitions of drug resistance.
  • Drug Resistance Emergence is Multi-Event: Isoniazid resistance emerged once, but rifampin resistance emerged at least three times independently within the cluster. Other drug resistance types also emerged recently, highlighting the dynamic nature of resistance acquisition.
  • Bacterial Factors May Drive Transmission Success: The identification of five cluster-defining SNPs under positive selection, including those in genes like esxV (involved in host-pathogen interaction), suggests that specific bacterial adaptations may contribute to the transmissibility and success of these MDR-TB strains.
  • Genomic Data Enhances Transmission Inference: WGS, especially when including variations in regions like PE/PPE genes and indels, offers a significantly higher resolution for inferring transmission links compared to traditional methods and even epidemiological data alone.

Key Results

  • The core cluster of 57 Mtb isolates diverged approximately in 1968 (95% HPD: 1945–1985), indicating a long history of circulation.
  • Isoniazid resistance arose once, while rifampin resistance emerged at least three times independently.
  • Five cluster-defining SNPs were identified as potentially contributing to transmissibility and were under positive selection in other TB lineages.
  • The study found that household contact tracing alone was insufficient to identify all transmission links, with genomic data revealing many more connections within the community.
  • Including variations in PE/PPE regions and indels in phylogenetic analysis led to a decrease in the number of inferred transmission links by up to 25%, suggesting a more accurate representation of genetic distance.

Limitations and Future Work

  • Subset Sequencing: Only a subset of isolates (61 out of 148) were sequenced, which could potentially lead to an underestimation of diversity and missed transmission links.
  • Molecular Clock Rate Reliance: The dating estimates are dependent on the assumed molecular clock rate, which is based on previously reported data.
  • Contact Tracing Scope: Contact tracing was limited to household contacts, potentially missing community-level transmission events.
  • Future Work: The authors suggest that WGS data should be integrated with epidemiological investigations for more effective transmission cluster identification and contact network reconstruction. Further research into the functional roles of the identified cluster-defining SNPs and their contribution to bacterial fitness and transmissibility is warranted. The study also highlights the need for improved methods to incorporate variations in PE/PPE regions and indels into phylogenetic analyses for more accurate transmission inference.

Summary updated

Abstract

Whole genome sequencing (WGS) can elucidate Mycobacterium tuberculosis (Mtb) transmission patterns but more data is needed to guide its use in high-burden settings. In a household-based TB transmissibility study in Peru, we identified a large MIRU-VNTR Mtb cluster (148 isolates) with a range of resistance phenotypes, and studied host and bacterial factors contributing to its spread. WGS was performed on 61 of the 148 isolates. We compared transmission link inference using epidemiological or genomic data and estimated the dates of emergence of the cluster and antimicrobial drug resistance (DR) acquisition events by generating a time-calibrated phylogeny. Using a set of 12,032 public Mtb genomes, we determined bacterial factors characterizing this cluster and under positive selection in other Mtb lineages. Four of the 61 isolates were distantly related and the remaining 57 isolates diverged ca. 1968 (95%HPD: 1945-1985). Isoniazid resistance arose once and rifampin resistance emerged subsequently at least three times. Emergence of other DR types occurred as recently as within the last year of sampling. We identified five cluster-defining SNPs potentially contributing to transmissibility. In conclusion, clusters (as defined by MIRU-VNTR typing) may be circulating for decades in a high-burden setting. WGS allows for an enhanced understanding of transmission, drug resistance, and bacterial fitness factors.
BibTeX
@inproceedings{dixit2019whole,
  author = {Avika Dixit and Luca Freschi and Roger Vargas and Róger Calderón and James C. Sacchettini and Francis Drobniewski and Jerome T. Galea and Carmen Contreras and Rosa Yataco and Zibiao Zhang and Leonid Lecca and Sergios‐Orestis Kolokotronis and Barun Mathema and Maha Farhat},
  title = {{Whole genome sequencing identifies bacterial factors affecting transmission of multidrug-resistant tuberculosis in a high-prevalence setting}},
  booktitle = {Scientific},
  year = {2019},
  doi = {10.1038/s41598-019-41967-8},
}
Figure