Living with uninvited guests - comparing plant and animal responses to endocytic invasions

Lead Research Organisation: Royal Holloway, University of London
Department Name: Biological Sciences


Salmonella are Gram-negative bacterial pathogens capable of infecting a wide range of hosts, including humans, pigs, cows, chicken and even plants. Salmonella typhimurium is the causative agent of various human and animal diseases, reaching from enteritis to typhoid fever. According to the World Health Organisation, Salmonellosis is the most frequent food-borne disease with around 1,5 billion infections world-wide yearly. Although hygiene conditions have improved considerably, the number of Salmonella infections has increased over the last decade due to antimicrobial resistance and the ability of Salmonella to hide inside host cells. Novel approaches are needed to address this global health problem. Salmonella replicates within host cells in a membrane-bound compartment and is dependent on tolerance and resources of the host cell. To ensure survival and propagation, Salmonella therefore secretes proteins into the host cytoplasm using a type III secretion system. Some of the roles of these proteins are beginning to be revealed, in particular in modulating key host signal transduction pathways. However, to fully grasp the mechanisms of host-pathogen response, we need to take a system-wide view and determining the whole network of interactions between Salmonella proteins and the host proteins. Such deep insight will yield new approaches to target the pathogens. The identification of global networks of protein-protein interactions has been accelerated in recent years by the development of high-throughput technologies such as transcriptomics and proteomics. Here, we propose to identify protein-protein interactions and pathways as a means to understand the crosstalk between plant, animal or human hosts and Salmonella. By studying diverse hosts, we will address the following questions: How do different cells respond to bacterial invasion? To which subset of bacterial gene products are host cells exposed? Are there host 'weak points' that Salmonella exploits in animals and plant cells alike? We propose the idea that addressing these basic biological questions for divergent hosts such as plants and animals can help us elucidate the way the interaction between the host and the pathogen works at the mechanistic level. Analyzing the responses of different hosts to invasion, and integrating these results using a systems biology approach will expose the weaknesses and strengths in host responses. This will advance the field through the development of tools for the integration of data analysis, modelling and experimentation. In practical terms, this information can be exploited for drug development, diagnosis, disease forecasting, prevention and control. Our working hypothesis is that plants and animals respond to pathogen invasion in fundamentally similar ways, and that species-specificity is conferred by nuances on general themes. This means, that in principle, similar approaches should be applicable to design strategies to detect and fight Salmonella infections irrespective of host. We will test this hypothesis through an integrative cycle of computational and experimental approaches. The outcome will be either a unified model for the general host response, or separate models, one for each host. Thus, both, proving or disproving the hypothesis will be equally valuable.

Technical Summary

How do cells respond to bacteria, which are much larger than viruses and usually encode hundreds to thousands of genes? To which subset of the corresponding bacterial gene products are host cells exposed? This proposal will address these questions for host responses to Salmonella. The spectacular and unique tolerance of Salmonella to extreme divergence of host species, makes this organism an ideal case study to investigate fundamental biological mechanisms of host-pathogen communication. Studying the responses of very diverse hosts to the same pathogen, can reveal how pathogens evade detection and elimination from hosts. This has practical significance because Salmonella bacteria present in raw meats, fruits and vegetables causes salmonellosis, the world's most widely spread food poisoning. Due to antimicrobial resistance and the ability of Salmonella to hide inside host cells, the number of Salmonella infections increased over the last decade. Novel approaches are needed to address this growing global health problem. At the beginning of the 'food chain' that leads to infection of humans, are the infections of farm animals and plants. This proposal will focus on identification of protein-protein interactions and pathways as a means to understand the crosstalk between host and Salmonella, host being a plant, human or other animal. However, these networks and pathways are not static, but rather are regulated over time and space, and involve numerous regulatory mechanisms including post-translational modifications such as phosphorylation. Modulation of signal transduction pathways via phosphorylation during Salmonella invasion is clearly established, and individual Salmonella proteins can have different function depending on their location. The interactions between Salmonella and host are time-dependent, with different processes being affected at different time points. The work will therefore aim at creating dynamic models of Salmonella-host pathogen interactions.

Planned Impact

This research will increase our understanding and knowledge in the fields of to those working in human and animal health, food safety, applied medicine and public health. The systems biology component will be of interest to those working in proteomics and metabolomics. The theoretical development will be of interest to those working in computer science, bioinformatics, theoretical biology, and integrative biology. We anticipate that the research has the potential to lead to products that can be commercialized (either in the form of biomarkers, drug targets or software) and there is thus a positive impact on the commercial sector. The outputs that take the form of improved diagnostics will be of benefit for policy makers and public health officials in policy formulation and decision making. Through this it will be of benefit through the wider public through reducing the improved diagnostics and control of Salmonella and a reduction in the incidence of food poisoning. We anticipate that two types of outcomes will result from the proposed work. (a) direct results (b) methodology. Direct results (a) would include, for instance, the identification of biomarkers or targets for pharmaceuticals. For (b) we anticipate software solutions and statistical methodology to result from the proposed work. These results will have impact in the first place within the research community, but ultimately this could have wider applications which could have be used in improving public and animal health which could be commercially exploited. Biomarkers will aid and assist the detection of strains of Salmonella. Drug targets will benefit the wider population through increased health. We will disseminate our results through publication in the scientific literature, and if suitable, more widely accessed media, as well as through presentations and webpublication. The research activities of this project might turn into patentable and/or commercially exploitable results. We have close links with relevant industries: two of our partners are companies who will lead the development of both types of results into commercial products. Through the combined publication and dissemination of results, as well as pursuing the development of applications where suitable, we will maiximse the impact of our research on increasing knowledge and understanding as well as in public health and food safety.


10 25 50
Description This project was part of a European consortium in which we have identified similarities and differences in the interactions of Salmonella with animal and plant hosts. As model systems, we chose human and Arabidopsis, to represent extremely diverged species in the comparison.

We studied protein-protein ineractions and gene expression data. We carried out mass spectrometric analysis of samples on interactions of bait proteins with cell extracts from plants and animals. This is the major data collected on defining interactions experimentally and has yielded many exciting candidates.

We developed a method to predict potential protein-protein interactions on the basis of collected data by project partners. For this we developed an original approach, based on recently invented conformal predictors for reliable machine learning. Its advantage is that the size of the list of the most suspected interactions can be chosen according to a pre-selected confidence level, with guaranteed validity properties. The methodology was initially trained on the HIV-Human interaction problem and formulated. Then it was applied to the interaction between Salmonella proteins and human host, and later to Kinase-Substrate relations (collected by our project partners). The conformal predictor approach for Salmonella-Human interactions was with the results of an interolog PPI prediction approach as well as experimental evidence for human proteins that may be involved in the interplay between Salmonella and human. By this procedure we were able to predict highly confident list of potential Salmonella-human PPIs and state on their possible role in Salmonella pathogenicity. The biological conclusions are the following: 1. Salmonella may interact with human proteins involved in membrane trafficking and apoptosis regulation. 2. Salmonella may interact with integrins. 3. Salmonella triggers cytoskeletal rearrangement.

We have also developed a method to analyse time series data to compare data from different host organisms, that takes into account the variability between different replicas. The novelty of our approach lies in that it can compare time resolved expression data from different hosts with different time points by fitting. the time series to a priori defined profiles, and extract qualitative information from the fit. Our approach allows us to take the uncertainty, that is inherent to this data, into account during further analysis. The data generated by our project partners on Arabidopsis and Human gene expression data is difficult to compare, because the Arabidopsis experiments are done with whole plants as hosts, while for the Human experiments were done in cell cultures. We fitted the gene expression data to a number of different models and grouped genes by their function using the gene ontology database. We could then classify each group of genes (function) in various ways. We studied which functions were up/down regulated in the two experiments, and in this way could decide if a group of genes, corresponding to a biological function, was up or down regulated. Finally, we compared these results between the different species to see whether there are differences at functional level in their responses. Comparing Arabidopsis and Human gene expression according to gene function showed that in many ways Arabidopsis and Human cells react the same to Salmonella invasion. In both cases the immune system is activated while many metabolic functions are down regulated. The host cells are clearly under pressure and many stress related functions are activated, such as response to unfolded proteins. Interestingly, our analysis also showed that in both cases microtubuli are down regulated. This answers one of the key questions this project was set out to achieve, which was to determine in how far the Salmonella used a similar mechanism of invasion for two very different hosts.
Exploitation Route It will allow better prediction of protein-protein interaction and assist in the analysis of time resolved data by bioinformaticians
Sectors Pharmaceuticals and Medical Biotechnology,Other