Integrated Metabolic Modeling, Culturing, and Transcriptomics Explain Enhanced Virulence of Vibrio cholerae during Coinfection with Enterotoxigenic Escherichia coli

Alyaa M. Abdel‐Haleem, Vaishnavi Ravikumar, Boyang Ji, Katsuhiko Mineta, Xin Gao, Jens Nielsen, Takashi Gojobori, Ivan Mijakovic̀mSystems 2020 · DOI

TL;DR

This study uses integrated metabolic modeling, culturing, and transcriptomics to show that Vibrio cholerae thrives in coinfections with E. coli due to metabolic cross-feeding, leading to enhanced virulence. This work identifies potential drug targets that are essential for V. cholerae in both single and coinfection scenarios, offering a broader spectrum of therapeutic strategies.

What Problem Does This Paper Address?

Most research on pathogen drug targets focuses on single-species infections. However, in reality, infections often involve multiple microbes (polymicrobial infections). This can significantly alter a pathogen's behavior, including its growth and essential gene requirements, making single-target strategies less effective. Specifically, Vibrio cholerae (causing cholera) and enterotoxigenic Escherichia coli (ETEC) frequently co-occur in diarrheal infections, and these dual infections are known to be more severe. The authors aim to understand how coinfection with ETEC impacts V. cholerae's metabolic capabilities and gene essentiality, and to identify drug targets that remain effective in these complex scenarios.

Methods

The researchers employed a multi-pronged approach:

  • Genome-Scale Metabolic Modeling (GEMs): They constructed a detailed metabolic model of V. cholerae (iAM-Vc960) and a coinfection model combining V. cholerae and ETEC. These models simulate metabolic fluxes and predict growth under various conditions.
  • In Vitro Culturing: They experimentally grew V. cholerae and ETEC together (cocultures) and separately (monocultures) to observe their growth dynamics.
  • Dual RNA Sequencing (RNAseq): They analyzed the gene expression profiles of both pathogens during coculture to understand how their transcriptomes change in response to coinfection.
  • In Silico Gene Essentiality Analysis: Using their models, they predicted which genes are essential for pathogen survival and growth, both in single infections and coinfections.

Datasets

  • Genomic Data: The genome of V. cholerae V52 was sequenced and annotated. Existing genomic data for ETEC strains were also utilized.
  • Metabolic Network Reconstructions: Previously published GEMs for ETEC and other enteric pathogens (like E. coli and Shigella) were used for comparison and to build the coinfection model.
  • Experimental Gene Essentiality Data: Data from three independent studies on experimentally validated essential genes for V. cholerae were used to validate the model's predictions.
  • Transcriptomic Data: Dual RNAseq data from cocultures and monocultures were generated and analyzed.

Key Insights and Ideas

  • ETEC as a Metabolic Enabler: ETEC doesn't just coexist; it actively provides V. cholerae with essential metabolites (cross-feeding) that V. cholerae cannot produce itself. This is a major driver of V. cholerae's enhanced growth in coinfection.
  • ETEC's Downside: While ETEC benefits V. cholerae, ETEC itself experiences suppressed growth in the coinfection. This suppression is likely due to non-metabolic factors, such as the type VI secretion system (T6SS) employed by V. cholerae.
  • Transcriptomic Shifts: Coinfection significantly alters the gene expression landscape of V. cholerae, particularly in metabolic pathways, indicating a distinct physiological state compared to single infections.
  • Druggable Targets in Polymicrobial Infections: The study highlights that essential genes can differ between single and coinfection scenarios. Identifying targets essential in both situations is crucial for developing broad-spectrum antimicrobials.

Key Results

  • Enhanced V. cholerae Growth: Model predictions and in vitro experiments confirmed that V. cholerae grows significantly better when coinfected with ETEC compared to growing alone.
  • ETEC Growth Suppression: Coculture experiments showed a decrease in ETEC abundance, likely mediated by V. cholerae's virulence factors.
  • Metabolic Differences: V. cholerae was found to have less versatile metabolic capabilities than E. coli and Shigella, relying more on cross-feeding in coinfection.
  • Gene Essentiality Discrepancies: In silico gene knockout simulations revealed that some genes essential for V. cholerae in single infections become non-essential in coinfection, and vice-versa.
  • Identification of Core Essential Genes: The study identified 47 metabolic genes critical for V. cholerae growth in both single and coinfection settings, many of which are involved in cofactor biosynthesis and metabolism.

Limitations and Future Work

  • Non-Metabolic Factors: The metabolic models could not fully explain the observed decrease in ETEC growth, suggesting that non-metabolic interactions (like T6SS activity) play a significant role and require further investigation.
  • Model Refinement: The authors acknowledge that their V. cholerae model is not exhaustive and can be further refined with more experimental data.
  • In Vitro vs. In Vivo: While the study integrates modeling and in vitro experiments, it notes that the human gut environment is more complex (hypoxic, varying pH, diverse commensals) and these factors are not fully captured.
  • Future Directions: The authors suggest expanding models to include virulence factors, exploring the impact of environmental conditions (pH, oxygen), and validating the identified drug targets experimentally. They also propose applying this integrative framework to other polymicrobial infections.

Summary updated

Abstract

Gene essentiality is altered during polymicrobial infections. Nevertheless, most studies rely on single-species infections to assess pathogen gene essentiality. Here, we use genome-scale metabolic models (GEMs) to explore the effect of coinfection of the diarrheagenic pathogen <i>Vibrio cholerae</i> with another enteric pathogen, enterotoxigenic <i>Escherichia coli</i> (ETEC). Model predictions showed that <i>V. cholerae</i> metabolic capabilities were increased due to ample cross-feeding opportunities enabled by ETEC. This is in line with increased severity of cholera symptoms known to occur in patients with dual infections by the two pathogens. <i>In vitro</i> coculture systems confirmed that <i>V. cholerae</i> growth is enhanced in cocultures relative to single cultures. Further, expression levels of several <i>V. cholerae</i> metabolic genes were significantly perturbed as shown by dual RNA sequencing (RNAseq) analysis of its cocultures with different ETEC strains. A decrease in ETEC growth was also observed, probably mediated by nonmetabolic factors. Single gene essentiality analysis predicted conditionally independent genes that are essential for the pathogen's growth in both single-infection and coinfection scenarios. Our results reveal growth differences that are of relevance to drug targeting and efficiency in polymicrobial infections.<b>IMPORTANCE</b> Most studies proposing new strategies to manage and treat infections have been largely focused on identifying druggable targets that can inhibit a pathogen's growth when it is the single cause of infection. <i>In vivo</i>, however, infections can be caused by multiple species. This is important to take into account when attempting to develop or use current antibacterials since their efficacy can change significantly between single infections and coinfections. In this study, we used genome-scale metabolic models (GEMs) to interrogate the growth capabilities of <i>Vibrio cholerae</i> in single infections and coinfections with enterotoxigenic <i>E. coli</i> (ETEC), which cooccur in a large fraction of diarrheagenic patients. Coinfection model predictions showed that <i>V. cholerae</i> growth capabilities are enhanced in the presence of ETEC relative to <i>V. cholerae</i> single infection, through cross-fed metabolites made available to <i>V. cholerae</i> by ETEC. <i>In vitro</i>, cocultures of the two enteric pathogens further confirmed model predictions showing an increased growth of <i>V. cholerae</i> in coculture relative to <i>V. cholerae</i> single cultures while ETEC growth was suppressed. Dual RNAseq analysis of the cocultures also confirmed that the transcriptome of <i>V. cholerae</i> was distinct during coinfection compared to single-infection scenarios where processes related to metabolism were significantly perturbed. Further, <i>in silico</i> gene-knockout simulations uncovered discrepancies in gene essentiality for <i>V. cholerae</i> growth between single infections and coinfections. Integrative model-guided analysis thus identified druggable targets that would be critical for <i>V. cholerae</i> growth in both single infections and coinfections; thus, designing inhibitors against those targets would provide a broader spectrum of coverage against cholera infections.
BibTeX
@inproceedings{abdelhaleem2020integrated,
  author = {Alyaa M. Abdel‐Haleem and Vaishnavi Ravikumar and Boyang Ji and Katsuhiko Mineta and Xin Gao and Jens Nielsen and Takashi Gojobori and Ivan Mijakovic̀},
  title = {{Integrated Metabolic Modeling, Culturing, and Transcriptomics Explain Enhanced Virulence of Vibrio cholerae during Coinfection with Enterotoxigenic Escherichia coli}},
  booktitle = {mSystems},
  year = {2020},
  doi = {10.1128/msystems.00491-20},
}
Figure