I am a Royal Society University Research Fellow in the Department of Infectious Disease Epidemiology. I research the evolution and epidemiology of bacteria using a combination of laboratory work and theoretical approaches.
Working with Brian Spratt and his group we use MLST (Multi Locus Sequence Typing) in which a few genes are sequenced as representative of the rest of the genome, for the purposes of strain definition. The sequences at these genes can easily be compared to others over the internet, making this an ideal tool for global epidemiology.
Population Genetics of Bacteria
The impact of natural selection, or demographic process such as rapid population growth and bottlenecks, can be inferred from examining the distribution of genetic variation within a species. In bacteria, this is more complicated than might be expected because of recombination. This means that bacteria, rather than simply dividing into two daughter cells, each of which are genetically identical, on occasion acquire genetic material from other, more distantly related strains. In humans, genes are inherited from two parents. The process of recombination in bacteria can be likened to a child inheriting genes from three or more parents. To make matters worse, different species vary markedly in the frequency of such events. From a medical point of view, they are interesting because they can generate new combinations of genes associated with antibiotic resistance or virulence.
Working with Christophe Fraser, we have developed a multilocus model of evolution in bacteria, which allows varying amounts of recombination, and validated it using MLST and samples of three medically important species: N. meningitidis, S. pneumoniae and S. aureus. The model is described in this paper in PNAS. Unexpectedly, we found that the samples we studied showed no signs of natural selection, once we accounted for multiple samples of the same short transmission chain. From these ‘micro-epidemics’, which are a consequence of non-random contact patterns among humans, a pattern of neutral, random drift emerges.
The model can be used to estimate the relative rate at which genes are shuffled by recombination, and as a basis for more sophisticated simulation studies of bacterial evolution, which are ongoing.
Molecular Epidemiology of S. pneumoniae
Over the last few years I have been working closely with researchers based at the National Public Health Institute (KTL) in Finland. As part of the Finnish Otitis Media (FinOM) studies, scientists have collected large numbers of bacteria of the species S. pneumoniae from ear infection (otitis media) and asymptomatic carriage. We were able to use these to demonstrate the relationships between certain strains of S. pneumoniae and ear infection, finding that strains only differ slightly in their ability to cause ear infection, unlike in invasive disease. This has implications for the control of otitis media with new conjugate vaccines. Data from these studies were also used to validate a model of bacterial evolution (see above).
Following the introduction of widespread vaccination against S. pneumoniae in the US, I have been working with scientists based at Harvard and Boston University to track the consequences for the pneumococcal population. We have noted the emergence of serotype 19A in both asymptomatic carriage and invasive disease, and are currently working to predict the likely future trajectory of this serotype, which is thriving in the post vaccine era.
The species problem
What constitutes a species has long been a source of debate in biology, and particularly so for bacteria. It is now known that bacteria exchange genes over far greater evolutionary distances than is the case for multicellular organisms, and can acquire through recombination (see above) genetic information from organisms which would not normally be considered to be the same species. This, together with predominantly clonal inheritance, means that the biological species concept of Mayr can therefore not be applied to bacteria.
Recently, we have approached this question from a different angle: it is reasonable to assume that any species worthy of the name will form a cluster in sequence space (the theoretical space of all possible sequences). We tested this using publicly available MLST data from the Neisseria, which frequently undergo recombination both within and between currently recognised species. We found that clusters in sequence space are observed, but that these are not discrete, ideal platonic entities. Rather their borders are diffuse, giving rise to what we call ‘fuzzy’ species. Work is underway to apply this to other bacteria, and to investigate how such fuzzy species may emerge.
I am interested in the combination of factors that have made HIV-1 such a succesful virus (more than 40 million currently infected worldwide). Again with Christophe Fraser and others, I have been working to improve our understanding of the relationship between viral load, the number of viral particles in the blood, and infectiousness on one hand, and life expectancy on the other.
I contribute articles to the national press and online media, concerning infectious disease and evolution.
I teach on a number of undergraduate and postgraduate courses in life sciences and medicine, including the MSc in Modern Epidemiology and the Short Course in Infectious Disease Epidemiology organised by Peter White and Christophe Fraser.
Hanage WP, 2011, Charles Darwin in modern epidemiology and public health: the celebration continues, Journal of Epidemiology and Community Health, Vol:65, ISSN:0143-005X, Pages:6-7
et al., 2010, Re-emergence of the type 1 pilus among Streptococcus pneumoniae isolates in Massachusetts, USA, Vaccine, Vol:28, ISSN:0264-410X, Pages:4842-4846
et al., 2010, HIV-1 Transmitting Couples Have Similar Viral Load Set-Points in Rakai, Uganda, Plos Pathogens, Vol:6, ISSN:1553-7366
Hanage WP, 2010, The New Foundations of Evolution On the Tree of Life, Science, Vol:327, ISSN:0036-8075, Pages:645-646
et al., 2009, Identifying Currents in the Gene Pool for Bacterial Populations Using an Integrative Approach, Plos Computational Biology, Vol:5, ISSN:1553-734X