Predicting the durability and resistance risk of crop protection measures through experimental evolution of plant pathogens

进化有多可预测？如果进化史被重复，每次的结果是一样的，还是截然不同？这个问题几十年来一直吸引着进化生物学家，但当进化的特征是对药物、农药或其他治疗的抗性时，这个问题就有了迫切的实际意义。粮食安全不断受到植物病虫害的威胁，因此需要进行作物保护，以保障收成，避免因病虫害而浪费土地和其他投入物。农药目前是植物病害控制的主要组成部分，但正如抗生素的广泛使用导致耐药细菌的进化一样，农业杀菌剂、杀虫剂和除草剂的广泛使用也导致了耐药疾病、害虫和杂草的进化。任何有效的控制措施都会选择克服该控制措施的能力，无论该控制措施是治疗感染的药物，还是作物植物中的抗病基因。在农业方面，农药的替代品正在逐步转向，但这些替代品仍然存在抗药性的风险，对任何控制措施的抗药性都涉及相同的基本进化原理。第一例杀菌剂耐药性是50多年前在一种植物病原体中报道的。该项目将研究植物病原体抗性进化过程中涉及的一般进化原理，因此，几十年来对化学杀菌剂的抗性进化的经验教训可以应用于作物保护的新方法，并且可以从一开始就以进化智能的方式进行管理，在抗性发展成为问题之前减缓其发展。为了主动管理耐药性，我们需要能够预测它将如何演变。特定控制措施的抗性风险是高还是低？这些突变会导致低水平还是高水平的耐药性？它们会导致对一种特定产品或广泛产品的耐药性吗？我们能否预测准确的突变，并开发DNA测试，在突变首次出现时就检测到它们？本项目将采用实验进化、选择植物真菌病原菌对杀菌剂产生抗性等防治措施。我将使用杀菌剂选择，以便将结果与已经发生的现实世界抗性进化进行比较。我将测试对不同种类的杀菌剂的抗性进化的可重复性：抗性是否每次都是由相同的突变引起的，或者相同的实验是否每次都有不同的结果。我还将设置竞争实验，看看是否一些突变比其他突变具有更大的优势，以及这是否取决于环境条件，如温度和营养水平。这将告诉我们，当几种不同的突变都产生相似水平的抗性时，或者当抗性和竞争性之间存在权衡时，或者当不同的突变在不同的条件下得到青睐时，进化是否更难以预测。然后将这些方法应用于两种潜在的替代控制措施：生物控制和RNAi。我将测试引起植物疾病的真菌是否能够进化出对抑制其生长的细菌菌株的抗性，以及这种抗性是否通过相同的突变反复进化，或者是否出现了各种不同的机制。我还将测试一种引起植物疾病的真菌是否能够进化出对RNAi的抗性，RNAi是一种通过沉默特定基因表达而起作用的控制方法，这对设计RNAi来降低抗性风险意味着什么。本文所开发的方法也将适用于未来进一步的新的作物保护方法。

How predictable is evolution? If evolutionary history were repeated, would the result be the same every time, or dramatically different? This question has fascinated evolutionary biologists for decades, but when the trait evolving is resistance against a drug, pesticide or other treatment, the question takes on an urgent practical relevance. Food security is under constant threat from plant diseases and pests, so crop protection is needed to safeguard harvests and to avoid wasting land and other inputs on crops lost to pests and diseases. Pesticides are currently a major component of plant disease control, but just as the widespread use of antibiotics has led to the evolution of drug-resistant bacteria, the widespread use of agriculture fungicides, insecticides and herbicides has resulted in the evolution of pesticide-resistant diseases, pests and weeds. Any effective control measure will select for the ability to overcome that control measure, whether that control measure is a drug to treat an infection, or a disease-resistance gene in a crop plant. In agriculture, a gradual shift is occurring towards alternatives to pesticides, but these alternatives still have a risk of resistance, and the same fundamental evolutionary principles are involved for resistance to any control measure. The first case of fungicide resistance was reported in a plant pathogen over 50 years ago. This project will look at the general evolutionary principles involved in the evolution of resistance in plant pathogens, so lessons from decades of resistance evolution against chemical fungicides can be applied to new methods of crop protection and they can be managed in an evolution-smart way from the start, slowing the development of resistance before it becomes a problem. In order to manage resistance proactively, we need to be able to predict how it will evolve. Is the resistance risk for a particular control measure high or low? Will the mutations cause low or high levels of resistance? Will they cause resistance to one specific product or a wide range? Can we predict the exact mutations and develop DNA tests to detect those mutations as soon as they first emerge? This project will use experimental evolution, selecting a fungal plant pathogen for resistance against fungicides and other control measures. I will use fungicide selection so results can be compared to real-world resistance evolution that has already occurred. I will test how repeatable the evolution of resistance is for different classes of fungicides: whether resistance is caused by the same mutation every time, or whether the same experiment has different results each time. I will also set up competition experiments, to see whether some mutations have a bigger advantage than others, and whether this depends on environmental conditions such as temperature and nutrient levels. This will tell us whether evolution is less predictable when several different mutations all give a similar level of resistance, or when there are trade-offs between resistance and competitiveness or when different mutations are favoured under different conditions. These methods will then be applied to two potential alternative control measures: biological control, and RNAi. I will test whether a plant-disease-causing fungus is able to evolve resistance against a bacterial strain that inhibits its growth, and whether that resistance repeatedly evolved through the same mutation or whether various different mechanisms emerge. I will also test whether a plant-disease-causing fungus can evolve resistance against RNAi, a control method that works by silencing the expression of a specific gene, and what this means for designing RNAi to reduce the resistance risk. The methods developed here will also be applicable to further new crop protection methods in future.