How many bacterial pathogens are there?
Posted on December 9, 2022 by Dr Michiel Vos
A recent publication, 'A comprehensive list of bacterial pathogens infecting humans' in Microbiology provides a comprehensive list of every bacterial pathogen that infects humans. One of the authors, Dr Michiel Vos, takes us behind the paper and how he hopes the list can assist other microbiologists.
Can you introduce yourself, your role and your research area?
I am Michiel Vos, a Senior Lecturer in Evolutionary Microbiology at the University of Exeter, UK. I am part of the European Centre for Environment and Human Health in the Medical School and based at the Environment and Sustainability Institute on the Cornwall Campus.
I trained as a marine biologist but got drawn into microbiology for my PhD and have since been working in microbial ecology, including the study of natural populations of bacteria and horizontal gene transfer. As I am now in a medical school, part of my research focuses on free-living opportunistic pathogens, especially how the environment can select for antimicrobial resistance and virulence.
How did this research come about?
I was approached by first author Abigail Bartlett, who then was doing her undergraduate degree in Medical Sciences at Exeter, about a possible lab internship over summer. As we still had COVID-19 restrictions in place, I proposed to do a desk-based study instead. The research question was quite a simple one: how many bacterial pathogens of humans are there? When I asked this to colleagues, I always got wildly different answers. Of course, the answer depends on how you define a pathogen and how hard you look. I will not go into the details of our search strategy, but Abi searched intensively (although there are undoubtedly some species we missed).
What to call a pathogen is a bit contentious. There is no doubt that Vibrio cholerae (which causes cholera) or Yersinia pestis (the cause of plague) are pathogens, but what about Cupriavidus metallidurans for instance? This is an environmental chemolithoautotroph that only very rarely causes infection, and then only in immunocompromised patients.
We decided to call a species an ‘established pathogen’ when it was described as being pathogenic in three or more independent publications. This is quite a pragmatic definition, taking into account the view of the wider research community. We called pathogens that were reported fewer than three times ‘putative’. Lumping both categories we found a total of 1,513 species.
Did anything in the findings surprise you? If so, why?
I was a bit surprised by the large number of human pathogens, around 7% of all described bacterial species. Of course, only a tiny fraction of bacteria have been formally described, and many of those that are, are relevant to humans and so more likely to cause disease.
New pathogens are described at a rapid rate, with around 13% of all reported pathogens described in the last decade alone. It is possible that we are now exposed to a greater variety of pathogens but my guess is that this is predominantly due to our improved ability to tell species apart through whole-genome sequencing. For instance, co-author Richard Bendall and I, along with colleagues at the University of Exeter and the Royal Cornwall Hospital, described a new pathogen ourselves a few years back. Our colleague John Lee, who I call ‘the Staph whisperer’, noticed a Staphylococcus that looked a bit different on blood agar. After whole genome sequencing, we found it was indeed different and we described it as Staphylococcus cornubiensis (named after Cornwall).
What do you hope the future implications of this research will be?
Our paper is relatively straightforward, basically a long list of pathogens along with basic information on the taxonomic distribution of species. However, we hope it can be a useful resource to clinical microbiologists, as well as biologists studying the evolution of pathogenicity. Co-author Daniel Padfield has done an amazing job of archiving not only the pathogen list but also all descriptions of workflows and scripts to recreate figures on GitHub and Zenodo.
Ideally, we would like to update this list on a yearly basis and potentially also add relevant traits to it. We are still looking for funding or for volunteers to do this, so if any readers want to come forward, please contact me!
