ADAPTIVE EVOLUTION OF NOVEL FUNCTIONAL PEPTIDES

这项建议旨在发展数学， 计算机、重组DNA、选择和筛选程序 实现全新多肽或融合的适应性进化 具有有用的催化、配体结合、结构或其他性质的蛋白质 功能。潜在用途从工业催化到 生产新药和疫苗。我们目前有 产生了大约1000万个克隆到lambda中的新基因- Gtll，并似乎已成功鉴定出新的融合 使庆大霉素对宿主细菌产生抗药性的多肽。 数学工作包括发展有关的理论 可能肽空间中“自适应行走”的性质， 通过点突变或重组，变成局部的多肽 或期望函数的全局最优解。我们将完成 目前的工作是表征不同的新型多肽，如 融合蛋白在大肠杆菌中对庆大霉素的抗药性 溶原菌。然后我们将测试点的相对效率 适应性爬山中突变体与重组的比较。在……里面 此外，我们将选择新的肽或融合蛋白与 一种理想的催化功能。值得注意的是，我们将选择测试版- 半乳糖苷酶功能和磷酸丙糖异构酶功能 适当的细菌缺失突变体。 我们努力的一个广泛目的是获得新的多肽 可以模仿几乎任意信号的生物效应 激素、生长因子等分子，甚至致病因子 抗原。中心思想很简单：为小说筛选或选择 与靶标的形状互补相结合的多肽 分子。形状补体可能是一种受体 生长因子，或抗生长因子抗体。作为一种 最初的例子是，我们将筛选表皮生长因子 (EGF)与抗体发生交叉反应的模拟多肽 EGF靶分子。此外，我们还将开发一种通用的 选择新的模拟任意多肽的程序 表位。这种模拟多肽可能作为药物或 疫苗。这项工作的基本重要性包括 功能在多肽空间中的分布分析 为应用分子技术开辟道路 进化论。

This proposal aims at the development of mathematical, computer, recombinant DNA, selection and screening procedures to attain adaptive evolution of entirely novel peptides or fusion proteins with useful catalytic, ligand binding, structural or other features. Potential uses range from industrial catalysis to production of new drugs and vaccines. We have at present generated about 10,000,000,000 novel genes cloned into lambda- gtll, and appear to have successfully identified novel fusion peptides conferring gentamycin resistance to host bacteria. Mathematical work includes development of theory about the character of "adaptive walks" in the space of possible peptides, via point mutations or recombination, to peptides which are local or global optima for a desired function. We shall complete current work characterizing different novel peptides which as fusion proteins confer resistance to gentamycin in E. Coli lysogens. We will then test the relative efficiency of point mutants versus recombination in adaptive hill climbing. In addition we shall select for novel peptides or fusion proteins with a desired catalytic function. Notably we will select for beta- galactosidase function, and triosephosphate isomerase function in appropriate bacterial deletion mutants. A broad purpose of our efforts is to obtain novel peptides which can mimic the biological effects of almost arbitrary signal molecules such as hormones, growth factors, even pathogenic antigens. The central idea is simple: screen or select for novel peptides which are bound by a shape complement of the target molecule. The shape complement might be a receptor for a growth factor, or an antibody against the growth factor. As an initial example, we shall screen for epidermal growth factor (EGF) mimetic peptides which cross react to antibodies against the EGF target molecule. In addition we shall develop a general procedure to select for novel peptides mimicing arbitrary epitopes. Such mimetic peptides may be useful as drugs or vaccines. The fundamental importance of this work includes analysis of the distribution of function in peptide space and opening the way towards a technology of applied molecular evolution.