Rapid identification disease resistance genes from plant genomes by resistance gene enrichment sequencing (RenSeq) of EMS-derived susceptible mutants

Plant disease causes significant yield losses in agriculture. Wheat and potato are two of the most important crops worldwide, including India and the UK. Among the most damaging diseases of wheat are the rusts. Stripe rust occurs wherever the crop is grown causing average yearly yield losses of up to 10% in some regions. Stem rust was until the green revolution associated with regular crop failures and famine. The resistance introduced then has now been broken by new strains of the fungus, which started appearing in Africa 14 years ago. The potato late blight disease, the cause of the Great Irish Potato famine in the 1840s, is still a serious impediment to potato cultivation today. Pesticides can control these diseases but they are expensive, at odds with sustainable intensification of agriculture, and in developing countries and for subsistence farmers, they are simply unaffordable.

Wild relatives of domesticated crops contain many useful disease resistance (R) genes. Introducing this natural resistance is an elegant way of managing disease. However, traditional methods for introducing R genes typically involve long breeding trajectories to avoid linkage drag, i.e. the simultaneous introduction of deleterious traits. Furthermore, R genes tend to be overcome by the pathogen within a few seasons when deployed one at a time.

Our long-term strategy is to isolate, by molecular cloning, as many new R genes as possible, and introduce them in combinations using GM methods. Molecular cloning makes it possible, indeed straightforward, to put several new genes together in the same location in the genome, allowing breeders to work with them as a "single" gene and avoiding linkage drag. Moreover, from first principles, a pyramid of R genes with distinct specificities should be more durable.

Traditional map-based cloning of R genes, however, is still challenging. First, large tracts of plant genomes are inaccessible to map-based genetics due to lack of recombination. Second, most R genes belong to a structural class of genes called NB-LRRs, which tend to reside in complex clusters, and many hundreds of NB-LRRs populate a typical plant genome. The scientist therefore frequently delimits a map interval containing multiple NB-LRRs and must find out which confers the resistance of interest. An approach, which has been successfully used to narrow down the candidate list to a single NB-LRR, is mutagenesis and screening for susceptible mutants. This creates discrete variations whereby a simple comparison of mutant and wildtype can identify the R gene.

We propose a strategy that will significantly increase the rate of R gene identification. In a first step of our workflow, we will screen large numbers of mutagenized plants for susceptible mutants. In a second step, we will use a state-of-the art sequencing technique recently implemented in our lab to selectively capture and sequence all the NB-LRRs in a plant genome. This will allow us to rapidly and cheaply compare wildtype with mutants to identify and clone resistance genes.

The outputs of this research will be three-fold: (i) using known controls we will implement our generic strategy to isolate R genes from complex genomes, (ii) we will apply this strategy to identify novel R genes from potato and wheat (against late blight and wheat rusts respectively), and (iii) we will test our key wheat rust R genes in Indian and UK environments.

We envisage that not only will our strategy significantly accelerate R gene cloning, it could also be used to pursue R genes not amenable to standard genetics, e.g. in low- or non-recombinogenic regions of the genome including centromeres, alien introgressed segments, and translocations. In wheat, this would allow accessing a plethora of useful R genes currently unusable due to linkage to deleterious yield-depressing alleles.

The isolation of plant disease resistance (R) genes by traditional map-based cloning is typically an expensive and time consuming process, requiring development of thousands of molecular markers and screening of large segregating populations. Furthermore, many perfectly good R genes reside in genetically inaccessible regions of the genome.

We recently developed a NextGen resistance gene enrichment-sequencing (RenSeq) workflow that allows rapid scrutiny of the hundreds of NB-LRR R gene candidates within a plant genome.

In this proposal we will use RenSeq to quickly and cheaply compare the NB-LRRome of resistant parents and EMS-derived susceptible mutants to accelerate the identification of agriculturally relevant R genes.

In particular we will:

Improve RenSeq by adapting it to long read technologies (MiSeq and PacBio). Next, we will perform RenSeq on a recombinant inbred line population. This will allow us to test and optimise various bioinformatics parameters - the best parameters will be those that allow the integration of most NB-LRR markers into the genetic map with a low false discovery rate. As a proof of concept, we will use our optimised RenSeq workflow to re-clone R genes from previously published EMS-derived mutants.

Apply optimised RenSeq to agriculturally relevant R genes. We will generate EMS mutant populations in (i) a wild diploid potato relative with resistance to potato late blight, (ii) a wild diploid wheat relative with resistance to wheat stem rust, and (iii) various introgression lines of hexaploid bread wheat with resistance to wheat stripe rust and/or leaf rust. The choice of germplasm will allow us to test our RenSeq workflow on increasingly complex situations from a small diploid genome (1 Gb) to a large hexaploid genome (17 Gb).

Assay germplasm for resistance in Indian and UK environments. We will exchange our key wheat germplasm lines between the UK and India to test their resistance efficacy under local conditions.

A durable solution to wheat stem and stripe rust is a paramount problem in global food security. In recent years, new, more aggressive races of these two pathogens have appeared globally, increasing the potential for economic damage and food scarcity. Because the pathogens have shown the ability to overcome each new resistance gene that is deployed in agriculture, developing a capability to isolate large numbers of useful genes and deploy them in combinations is critical to successfully overcome these diseases. The proposed research aims to do exactly that: develop new methods for rapidly accessing novel resistance genes that can be deployed in wheat breeding programs.

Direct beneficiaries of this work include public wheat germplasm improvement programs (such as those at CIMMYT), the country-level breeding efforts that they support in the developing world (including Kenya, Ethiopia, Pakistan, India, and China), growers of wheat, and consumers of wheat in those countries. In the absence of a durable solution to wheat stem and wheat stripe rust, wheat agriculture, particularly in Africa and South Asia, will live under the constant threat of epidemics and resulting crop failure caused by this disease.

In the UK, new races of wheat stripe rust have appeared in recent years that infect previously resistant UK wheat cultivars. In addition, although stem rust has not been important in the UK or in most European wheat production for many years, climate change brings a new threat from this disease, which requires relatively warm conditions early in the growing season to have full impact. Although in high-value agriculture the pathogen can be controlled with chemical fungicide application, genetic resistance is vastly preferable from the standpoints of cost and environmental sustainability. Thus, UK and European wheat breeding companies, wheat growers, and consumers are also potential direct beneficiaries of this.

Identification of resistance genes that can be deployed for transgenic resistance to stem and stripe rust in wheat, in combinations to maximize durability of resistance are longterm (10-15) years, depending on acceptance of GM wheat.

Aegilops sharonensis, our choice as source for several of the novel resistance genes in the present proposal, is of interest to the global community of wheat researchers because its genome is closely related to the D and B genomes of hexaploid bread wheat. As part of another BBSRC-funded grant we have recently sequenced the genome of Ae. sharonensis and we are presently collaborating with the IWSC to perform genome-wide comparisons with the other wheat progenitors. The combined tools from these projects, including a deep sequence and vastly improved genetic map of this organism, will enable any interested researcher to more easily identify genes associated with useful agronomic traits in Ae sharonenesis. In addition to resistance genes for other diseases, these will include genes for tolerance to abiotic stresses such as heat and drought, and increased efficiency of nutrient utilization. These resources are already being made available to the research community.

The research staff working on this project will develop additional technical expertise in a diverse range of skills, including molecular genetics, bioinformatics, and plant biotechnology, specifically in wheat, an exceedingly important crop in UK and Indian agriculture. They will develop an increased understanding of the application of how basic science discoveries get translated into practical outcomes for agriculture by working with our downstream deployment partners. This expertise can be applied in future career positions in academia, research institutes, and in companies, both small agricultural biotech firms and global multi-national agriculture companies. These benefits will be realized immediately upon the next career progression steps taken by the individuals involved in the project.