In Vivo Screening of Mixture-Based Combinatorial Libraries

描述（由申请人提供）：我们的工作假设是，从基于混合物的组合文库中直接体内筛选和鉴定单个化合物将产生更“高级”的治疗候选物，同时减少药物发现过程中固有的时间和成本。目前的药物发现筛选策略几乎总是涉及基于靶点的体外生化或基于细胞的测定。这些用作鉴定活性化合物的主要手段，这些活性化合物接下来在动物研究中评估其活性。正是在这个后期阶段，大多数化合物由于毒性、缺乏功效和生物利用度差而失败。 在支持我们的工作假设，成功的初步研究，利用鼠甩尾疼痛模型已经证明，明确区分可以实现活性和非活性的混合物。例如，在甩尾试验中发现由50个不同的氨基酸在三个位置组成的125，000个四肽的混合物具有抗伤害感受活性，其中仅单独定义了其N-末端位置。该混合物（含有Dmt-DALDA;已知在体内有活性）的活性具有比吗啡长5至10倍的作用持续时间，而以每mg计，活性仅比吗啡低3-5倍。此外，发现在μ、δ或κ阿片样物质结合测定中没有活性的所选其他混合物具有明显的体内甩尾活性。这提出了一个令人兴奋的可能性，即新的受体位点或多个受体相互作用是所发现的活动的原因。 本提案中的两个目标将作为这种方法的一般概念证明，并将为以后使用一系列现有的基于杂环混合物的库进行研究奠定基础。第一个Aim将使用甩尾测定来完成迭代去卷积过程，以鉴定该活性混合物的三个位置处的最具活性的氨基酸。我们相信，这将能够鉴定相对于Dmt-DALDA具有增强活性的活性个体序列，或者至少将鉴定Dmt-DALDA（两者都是可接受的概念验证最终结果）。第二个目标涉及体内筛选构成该6，250，000个不同序列（50 × 503）的四肽文库的全部50种混合物。除了鉴定新的阿片特异性激动剂之外，该方法还可以鉴定对调节非阿片受体的“孤儿”疼痛具有活性的新的抗伤害感受化合物。如果成功的话，基于混合物的组合文库的直接体内测试不仅会促进疼痛调节，而且会促进生物医学研究和药物发现过程。

DESCRIPTION (provided by applicant): Our working hypothesis is that the direct in vivo screening and identification of individual compounds from mixture-based combinatorial libraries will yield more "advanced" therapeutic candidates, while decreasing the time and costs inherent in the drug discovery process. Current drug discovery screening strategies virtually always involve target based in vitro biochemical or cell-based assays. These serve as the primary means to identify active compounds that are next assessed for their activity in animals studies. It is at this later stage that the majority of compounds fail due to toxicity, lack of efficacy, and poor bioavailability. In support of our working hypothesis, successful preliminary studies utilizing the murine tail flick pain model have demonstrated that clear differentiation can be achieved between active and inactive mixtures. For example, a mixture of 125,000 tetrapeptides, made up of 50 different amino acids at three positions and with only its N-terminal position individually defined, was found to have antinociceptive activity in the tail flick assay. The activity of this mixture (that contained Dmt-DALDA; known to be active in vivo) had a 5 to 10 time longer duration of action than morphine, while being only 3-5 times less active than morphine on a per mg basis. Additionally, other mixtures chosen that had no activity in the mu, delta or kappa opioid binding assays were found that had clear in vivo tail flick activity. This raises the exciting possibility that novel receptor sites or multiple receptor interactions are responsible for the activities found. The two Aims in this proposal will serve as a general proof of concept for this approach and will lay the foundation for later studies with a range of existing heterocyclic mixture-based libraries. The first Aim will use the tail flick assay to complete an iterative deconvolution process to identify the most active amino acids at the three positions of this active mixture. We believe that this will enable the identification of active individual sequences that have enhanced activity relative to Dmt-DALDA or, as a minimum, will identify Dmt-DALDA (both are acceptable proof of concept end results). The second Aim involves the in vivo screening of the entire 50 mixtures making up this tetrapeptide library of 6,250,000 different sequences (50 x 503). In addition to identifying novel opioid specific agonists, this approach may enable the identification of novel antinociceptive compounds that have activity at "orphan" pain modulating non-opioid receptors. If successful, the direct in vivo testing of mixture-based combinatorial libraries will advance not only pain modulation, but biomedical research and the drug discovery process in general.