A Systems Approach for the Fragment-Based Development of Selective Chemical Probes of Bromodomain Function

调节储存在DNA中的遗传信息表达的一个关键的细胞机制是通过蛋白质‘因子’，即基因转录。一组这样的蛋白质通过读取表观遗传学标记来影响基因表达水平，即安装在与DNA相关的其他蛋白质上的可逆化学修饰，以形成称为染色质的高度紧凑的结构。一种广泛出现的修饰是赖氨酸氨基酸的乙酰化，这是由含有一个到六个称为溴域的‘读者’结构域的蛋白质特异性识别的。人类基因组编码42个含溴结构域的蛋白质，给出了组成溴结构域蛋白家族的57个独特序列。越来越多的证据表明，溴结构域蛋白与包括癌症在内的各种疾病有关，然而许多溴结构域蛋白的具体功能尚不清楚。以剂量依赖的方式干扰生物靶标功能的强大的、细胞渗透的小分子是一种有效的方法，可以‘探测’靶标在特定生物过程中的作用，以及它与疾病的联系，从而实现其治疗潜力。与涉及基因敲除或RNAi的更传统方法相比，小分子有几个优势，包括允许对细胞内的影响进行空间和时间控制。然而，探针化合物的鉴定可能很费力，而且通常涉及筛选大型化合物文库。要开发出既对目标蛋白质有足够的效力，又具有高度选择性的“工具化合物”，使它们不与其他类似的蛋白质结合，可能是一项具有挑战性的工作。这往往阻碍了化学探针的成功应用，以确定分子靶标与调节靶标的生物学后果之间的关系。开发在这些领域取得进展的新方法和工具将对化学生物学领域和药物发现中的靶标确认产生直接影响。近年来，人们建立了一种新的、强大的方法来识别针对蛋白质的高质量结合蛋白。这包括筛选分子文库，即所谓的“片段”，它们比通常在“高通量筛选”中测试的分子要小得多。利用蛋白质结构技术表征从片段筛选中识别的HITS的结合模式，从而详细确定它们与蛋白质的相互作用。一旦确定了几个片段命中，关于它们相互作用的组合信息、结合位点的性质和它们的化学知识可以为将这些结构“阐述”成更有效的化学探针提供基础。在目前的提案中，我们将把基于片段的方法与蛋白质工程结合起来，蛋白质工程是一种通过将氨基酸从一种类型改变为另一种类型来在蛋白质上产生特定突变的技术。首先，我们将通过1)“生长”它们以挑选与结合位点的额外相互作用来阐述溴结构域靶向片段；2)“合并”结合在乙酰赖氨酸结合口袋的重叠部位的片段。这将为溴结构域产生紧密的结合配体。其次，我们将详细阐述这些分子，以适应在化学上补充结合位点中引入的突变的功能基团，例如，填充通过设计口袋而产生的空间，和/或将配体共价‘点击’到半胱氨酸上。这种修饰的化学探针应该对突变体具有高度的选择性，以对抗野生型或任何其他溴域。由于突变可以迅速引入任何溴域蛋白和细胞中，本方案将开发的方法和工具将使在系统一级对溴域蛋白的生物功能进行化学询问的一般战略成为可能。然后，这种方法也可以扩展到研究其他读者域系统。

A key cellular mechanism for regulating expression of the genetic information stored in DNA is by mean of protein 'factors' that gene transcription. One group of such proteins affects gene expression levels by 'reading' epigenetics marks, i.e. reversible chemical modifications that are installed on other proteins that associate with DNA to form the highly compacted structure known as chromatin. A widely occurring modification is acetylation of lysine amino acids, which is specifically recognized by proteins that contain between one and six 'reader' domains called bromodomains. The human genome encodes 42 bromodomain containing proteins, giving a total number of 57 unique sequences that make up the bromodomain protein family. There is increasing evidence that link bromodomain proteins in various diseases, including cancer, however specific functions of many bromodomain proteins are yet unknown. Potent, cell-permeable small molecules that perturb the function of a biological target in a dose-dependent fashion are a powerful way to 'probe' the role of the target in a particular biological process as well as its association to disease and thus its therapeutic potential. Small molecules have several advantages over more traditional approaches involving gene knock outs or RNAi, including allowing spatial and temporal controls on the effect within a cell. However, identification of probe compounds can be laborious and often involves screening of large compound libraries. It can be challenging to develop 'tool compounds' that are not only sufficiently potent against a target protein but also highly selective so they do not bind to other similar proteins. This often hampers the successful application of chemical probes to establish a relationship between a molecular target and the biological consequences of modulating the target. Developing new approaches and tools to make advances in these areas would have an immediate impact in the field of chemical biology and for target validation in drug discovery. Recent years have seen the establishment of a novel, powerful approach to identify high quality binders against proteins. This involves screening libraries of molecules, so-called 'fragments', that are much smaller than those usually tested e.g. in 'high-throughput screening'. The binding modes of 'hits' identified from a fragment screen are characterized using protein structural techniques so their interactions with the protein are determined in details. Once several fragment hits are identified, the combined information on their interactions, on the nature of the binding site and knowledge of their chemistry can provide a basis for 'elaborating' these structures into more potent chemical probes. In the current proposal, we will combine fragment-based approaches with protein engineering, a technique to generate specific mutations on a protein by changing amino acids from one type to another. First we will elaborate bromodomain-targeting fragments by 1) 'growing' them to pick additional interactions with the binding site; 2) 'merging' fragments bound at overlapping sites at the acetyl-lysine binding pocket. This will generate tight binding ligands for bromodomains. Second we will elaborate these molecules to accommodate functional groups that chemically complement the mutation introduced in the binding site, e.g. filling space created by engineering a pocket, and/or 'clicking' the ligand covalently onto a cysteine. Such modified chemical probes should be highly selective for the mutant against wild-type or indeed any other bromodomain. Since the mutation can be rapidly introduced into any bromodomain protein and in a cell, the methods and tools that will be developed in this programme would allow a general strategy to chemically interrogate the biological function of bromodomain proteins at the system level. This approach could then be extended to study other reader domain systems as well.