Adaptive landscapes of antibiotic resistance: population size and 'survival-of-the-flattest'.

改变基因序列的自发突变是进化的关键引擎。然而，如果突变率超过一个“临界突变率”，自然选择就会过于频繁地改变种群的基因构成。例如，我们可以把突变看作是在“适合度”最高的基因序列具有高适合度的情况下，后代从父母身边移开。如果有太多的突变(高于临界突变率)，选择可能能够使一个种群保持在一个宽广的适应度峰值上，但不能保持在一个狭窄的适应度峰值上，即所谓的“最平的生存”。这种临界突变率通常被认为比在典型生物有机体中看到的要高得多。然而，我们最近发现，在计算机模拟中，在非常小的群体(&lt；100个个体)中，关键突变率要低得多。这表明最扁平的生存可能发生在生物突变率的正常范围内。因此，这一提议的目的是以抗生素耐药性进化为测试案例，测试最扁平的生存是否真的存在于生物学中。这需要一个综合的、多学科的方法，进行决定性的湿实验室实验，将计算机模拟与生物学相结合，并开发一个将研究结果与数学理解联系起来的严谨和预测性的理论框架。我们的第一个目标将是刻画对不同抗生素耐药的细菌的适应状况，以确定在实践中是否存在狭义和宽泛的适应峰之间的适当区别。在这里，我们将重点放在含有抗生素的环境中，但将使用实验进化(在高和最小种群规模下)和完整的基因组测序来考虑整个基因组的突变。我们的第二个目标是开发与生物学相一致的计算机模拟和理论。具体地说，我们需要将我们在模拟中关于临界突变率的发现与现有理论联系起来。然后，我们需要将理论和模拟都推向更符合生物现实的假设。例如，我们将测试更现实的健康状况，特别是那些通过我们的第一个目标经验确定的健康状况。我们的最终目标是严格检验理论、模拟和生物实验之间的一致性。特别是，我们将使用小种群规模的实验进化来测试我们是否能像预测的那样检测临界突变率的影响，并进一步表征实验进化的菌株，以测试所涉及的生物机制的一致性。正确地取消这种新的种群规模效应在抗生素耐药性的进化中可能会产生广泛的影响。抗生素耐药性细菌不可避免地出现在种群规模较小的突变中，而对一种抗生素产生耐药性的细菌将被另一种抗生素击倒到较小的种群规模。我们在这里测试的假设将决定在如此小的人口规模下是否保持特定的抗生素耐药性，是否取决于不同抗生素以理论上可预测的方式施加的不同适应环境的细节--这可能是对抗抗菌素耐药性的关键信息。总之，这项工作将把实验方法与生物学、模拟和理论中的进化紧密联系起来，以确定最平坦的人的生存是否以及如何影响到抗生素耐药性进化这一日益关键的问题。

Spontaneous mutation altering genetic sequences is a key engine of evolution. However, if the rate of mutations exceeds a 'critical mutation rate', changes occur too frequently for natural selection to maintain the population's genetic makeup. For example, one can think of mutation as moving offspring away from their parents in a 'fitness landscape' where peaks are genetic sequences with high fitness. If there is too much mutation (above a critical mutation rate), selection may be able to keep a population on a broad fitness peak, but not on a narrow one, so-called 'survival-of-the-flattest'. Such critical mutation rates are generally believed to be much higher than those seen in typical biological organisms. However, we have recently discovered that, in computer simulations, critical mutation rates are much lower in very small populations (<100 individuals). This suggests that survival-of-the-flattest could be occurring within the normal range of biological mutation rates.This proposal therefore aims to test whether survival-of-the-flattest really does occur in biology, using antibiotic resistance evolution as a test case. This requires an integrated, multi-disciplinary approach, conducting decisive wet-lab experiments, aligning computer simulations with biology and developing a rigorous and predictive theoretical framework that relates findings to mathematical understanding. Our first objective will be to characterise the fitness landscapes of bacteria resistant to different antibiotics to determine whether appropriate distinctions exist in practice between narrow and broad fitness peaks. Here we will focus on environments containing antibiotics, but will use experimental evolution (at high and minimal population sizes) with complete genome sequencing to consider mutations across the genome. Our second Objective is to develop computer simulations and theory to align with the biology. Specifically we need to relate our findings about critical mutation rates in simulations to existing theory. We then need to move both theory and simulation towards more biologically realistic assumptions. For instance we shall test more realistic fitness landscapes, particularly those determined empirically through our first Objective. Our final Objective is rigorously to test the coherence of the theory, simulation and biological experiment with each other. In particular we shall use experimental evolution at small population sizes to test whether we can detect the effect of critical mutation rates as predicted and further characterise the experimentally evolved strains to test the consistency of biological mechanisms involved.Rigorously unpicking this novel population size effect in the evolution of antibiotic resistance could have broad impacts. Antibiotic resistant bacteria inevitably appear by mutation with a small population size and bacteria resistant to one antibiotic will be knocked down to small population sizes by another. The hypotheses we shall test here will determine whether the maintenance of particular antibiotic resistances at such small population sizes, could depend on the details of the different fitness landscapes imposed by different antibiotics in a theoretically predictable way - potentially crucial information in combatting antimicrobial resistance. In sum this work will closely link experimental approaches to evolution in biology, simulation and theory to determine if when and how survival-of-the-flattest impinges on the increasingly critical issue of antibiotic resistance evolution.