Group selection as a novel tool to screen and improve biological pesticides

In modern agriculture and horticulture there is more and importance placed on exploiting naturally occurring organisms, such as the pathogens of insects, for pest management. Microbes such as Bacillus thuringiensis (Bt) are currently the main active ingredients in organically licensed applications targeted at caterpillar, mosquito and beetle pests. In addition, the same bacterial species is an important source of proteins for genetically modified plants, as it produces many proteins that are highly toxic to a narrow range of insects but safe for humans and wildlife.

Nevertheless, identifying novel virulent bacteria, and identifying new proteins that are toxic to insects, is a difficult and challenging task. The foundation of this proposal is the idea that if we can understand how investment in virulence factors (toxic proteins produced by microbes) increases the reproduction and fitness of microbes, then we can apply selection-based methods in the laboratory to isolate mutants or variants with high virulence.

Prior to this proposal we made an important advance in understanding the selection pressure that maintains virulence. Pathogens such as Bt have to invest energy and resources in exporting virulence factors that help them infect insects. Investing in virulence is therefore costly for individual cells, but essential, as bacterial reproduction requires the death of the host. Individual level selection (within hosts) can favour mutants, known as 'cheaters', which rely on the proteins produced by their neighbours for successful infection. Conversely, selection at the group level (between infections) maintains virulence. Preliminary experiments have shown that group level selection (in combination with a supply of mutations) can lead to the emergence of mutants with increased levels of virulence, even if these mutations are costly to the individual.

The group selection methods we have developed offer a new tool for: exploring existing microbial diversity and trying to improve microbial biocontrol agents. Moreover, for Bt, the critical proteins involved in killing host are the Cry toxins, and there is enormous applied interest in generating new and better Cry toxins that can kill hard-to-target pests. Using genetic engineering methods we can artificially construct a collection of Bt strains that expresses a diverse collection of these proteins, all with a subtly different make-up. Here we propose to use our new selection methods to pick out new and interesting proteins from this diverse mix.

A major question for this proposal is to better understand when and how we might be able to apply these group selection methods. Evolutionary theory can help us refine our methods, for instance by predicting how the make up of groups affects the power of selection. Another consideration, that is important for understanding the applied potential of this method, is to understand whether we can only use selection to improve strains with low fitness, or whether we can also improve strains that are already quite effective at killing insect hosts.

Another important prediction is that selection experiments might help us better understand how pathogens kill insect hosts, and identify the proteins responsible in Bt. When bacteria have evolved increases in virulence, we can use the change in DNA sequence, or the proteins expressed, to identify which molecules are responsible for increased virulence. Given our lack of knowledge on the function of the of the majority of proteins produced by Bt, we expect to gain significant extra knowledge regarding the means by which these bacteria can kill hosts and overcome their immune systems.

Pesticides based on the bacterium B. thuringiensis (Bt), or its protein toxins, have proved to be an effective and environmentally benign means of controlling important pest insects. Their use is limited by the low susceptibility of some pests to known strains or toxins, and is threatened by the development of resistance to existing products. Thus there remains a need to develop alternatives, the search for which is hampered by the need now to screen increasingly large collections of strains in order to identify any novel, useful isolates. We have recently used evolutionary theory to develop a protocol that can evolve, and enrich for, strains with improved virulence, and which overcomes the normal problem of enriching for those able to cheat by reducing their own metabolic load whilst relying on virulence factors produced by others.

We will expand on this finding to further our understanding of social evolution and test several hypotheses in this area, primarily how population structure and genetic background influences virulence evolution. This will be achieved by applying our selection regime to pools of diverse bacterial strains mixed randomly or grouped by sequence similarity, and also to mixtures of strains with known variations in virulence. We will also apply our selection to variants of toxins whose ancestors have differing levels of efficacy. A suite of genetic and biochemical analyses including genomic sequencing and proteomic comparisons will characterize the basis of the increased virulence. As well as helping us understand the principles of social evolution the project has clear application to the discovery of novel or modified virulence factors that can be developed as commercial products, particularly against S. frugiperda, the Fall Army Worm, which is the focal insect of this study and a major worldwide pest of maize and other crops.

This research, whilst based on basic science theory, has clear commercial applications and thus potentially significant impact for a wide range of beneficiaries.

Agrochemical companies
As evidenced by the willingness of DowDuPont to support this project there are a number of outcomes likely to be of direct relevance to this sector. This project should confirm our previous findings that it is possible to evolve a Bt strain with improved virulence towards an insect pest, of commercial importance, that has acquired resistance to existing products. The same approach can be used to evolve a strain that only currently has weak activity against a target pest, this could be worthwhile either if the target pest is particularly difficult to kill, or because naturally-occurring pathogens are generalists, which is true of a number of insect parasites or pathogens.

Aside from improving the efficacy of a strain with known, but poor, activity this work aims to develop a protocol that will significantly streamline the cost and time involved in screening large strain collections. There are many companies that are developing Bt strains as live microbial products, including both large international ones as well as smaller enterprises chasing a niche market, which could benefit from this research. In addition, biotech companies (such as DowDuPont) are more interested in discovering novel protein toxins or other virulence factors that can be incorporated into GM crops. It is anticipated that the protocols being developed by this work will lead to the identification of such products either through modification or overexpression in the evolved strain, or through their presence in strains selected from a collection. Furthermore, the toxin mutagenesis approach could lead to improved variants of existing proteins. Understanding the mechanism of improved virulence could also lead to protocols by which existing products could be modified to improve efficacy and therefore economic competitiveness.

Academia
Although this research has clear applied outcomes of improving bacterial strains for pest control it does so by addressing some fundamental scientific questions in social evolution. As a result the work will be of interest to a diverse range of academics who can use the results not only to facilitate future work into insect pathogens but also to help understand the evolution of virulence and potentially apply the principles to other systems.

Skills and training.
This project embraces a range of different technical approaches and methodologies and makes use of the complementary skill sets of the PI and Co-Is. Due to the tight collaborative nature of the work the PDRAs employed on the project will not only benefit from the many professional development opportunities available at both Exeter and Sussex, but will have the added benefit of exposure to a much broader range of ideas and opportunities than would have been encountered at a single site. The industrial input into this project will also provide them with an insight into how industry functions and how it can productively interact with academia.

UK Economy
Although insect pest control in the UK is not as big a commercial problem as it is elsewhere in the world, there are still opportunities for the development of products based on the outcome of this work. For instance, there is considerable commercial interest in expanding and improving the set of pathogenic isolates that can control aphids, in order to replace neonicotinoids.

Public and other stakeholders
While there is still debate in the UK over the use of GM crops, it is accepted that pest control is required and products involving Bt strains are licenced for use - particularly in organic farming. This work has the potential to increase the efficacy of these products and facilitate the growth of organic farming to meet public demand.