Development of inhibitors targeting Plk1 polo-box domain

抗有丝分裂药物，如紫杉烷类和长春花生物碱，直接抑制微管（MT）的动态功能，已被证明在治疗癌症中有效。然而，这些常规MT靶向剂的使用受到其剂量限制性毒性的严重限制，这是由于MT在许多细胞类型（特别是肠上皮细胞和淋巴细胞）中介导有丝分裂和非有丝分裂细胞过程的广泛功能。因此，通过靶向对癌细胞增殖至关重要的有丝分裂特异性蛋白来开发一类新的抗癌治疗剂是重要的研究方向。Polo-like kinase 1（Plk 1）是目前最具吸引力的抗癌靶点之一。通过靶向Plk 1的催化活性来产生Plk 1特异性抑制剂的努力已被证明是困难的，这是由于与其他结构相关激酶的催化结构域的相似性。在这里，我们建议开发一类新的单特异性Plk 1抑制剂，采用一种新的方法来靶向非催化，但功能上必不可少的，PBD的Plk 1。为此，我们与马里兰州贝塞斯达的NCATS合作进行了高通量筛选。从这个筛选中，我们已经从初级筛选中鉴定了3，000种化合物，通过二级培养基通量和三级基于细胞的测定将其缩小到最终的两种化合物（1 S和B7）。通过这些母体化合物的系统分子建模/对接，我们获得了六个显着改进的化合物，其在体外ELISA中显示出Plk 1 PBD抑制活性，其水平与先前表征的PBD结合配体之一PLHSpT（Kd = 450 nM）相似。由于原始HTS先导化合物可能属于称为泛测定干扰化合物或PAINS的一类，因此我们将在命中-先导化合物优化期间进一步纳入许多药物样特征，以获得高影响力的化学探针，其最终在概念验证小鼠肿瘤模型中表现出对Plk 1 PBD的所需功效。作为药物发现的第一步，我们生成了Plk 1 PBD与其中一种先导化合物的共晶结构。该结构模型将用于进行构效关系研究和进一步优化化合物。作为这项工作的结果，我们获得了几个初始命中，进行了几轮的计算机衍生化和基于ELISA的Plk 1 PBD抑制分析。这些努力导致发现了七种具有不同化学型的化合物（NC 21至NC 27）。用纯化的化合物进行的体外FP和ELISA测定证实了它们的抗PBD特异性，Kd值在100-200 nM范围内。在过去的两年中，我们进行了命中到铅优化（与Ken Jacobson博士，NIDDK，NIH合作），并开发了20种抑制剂，这些抑制剂似乎在体外试验中有效抑制Plk 1 PBD抑制剂。然而，这些抑制剂的疗效和生物利用度低于预期。现在，努力的重点是提高效力和渗透性。这项工作已经产生了一篇论文（Alverez CN等人，J.Med.Chem.2020），其描述了几种对Plk 1 PBD具有特异性的新型三唑并喹唑啉衍生的小分子前药。为了跟进这一初步研究结果，我们产生了各种衍生物，以增加针对Plk 1 PBD的效力和特异性。这一努力导致了随后的两篇论文（Park，JE等人，ACS Pharm. & Transl.科学，2023和Park，JE等人，PNAS，2023，出版中）、美国专利申请（第18/028，463号; 2023年3月24日提交）和美国临时专利申请（第63/455，608号; 2023年3月30日提交），其涵盖了第一种变构Plk 1 PBD抑制剂，称为Allopole，具有与Plk 1 PBD 1结合的独特模式。Allopole与Plk 1 PBD的结合模式揭示了一个变构结合位点，该位点可以选择性地破坏PBD与其磷酸结合靶点之间的典型相互作用。考虑到Allopole结合口袋的结构独特性和封闭性质，在典型的磷酸表位结合位点上靶向该口袋可能是有利的，这需要能够介导多种水介导的相互作用的带负电荷的配体。我们正在利用Allopole结合位点的独特性质来开发Plk 1 PBD特异性抑制剂。

Anti-mitotic drugs, such as taxanes and vinca alkaloids that are directed at inhibiting the dynamic function of microtubules (MTs), have proved effective in treating cancer. However, the use of these conventional MT-targeting agents is severely constrained by their dose-limiting toxicities due to the widespread functions of MTs in mediating both mitotic and non-mitotic cellular processes in many cell types, notably intestinal epithelial cells, and lymphocytes. Therefore, developing a new class of anti-cancer therapeutics by targeting mitosis-specific proteins critical for cancer cell proliferation is an important line of investigation. Polo-like kinase 1 (Plk1) is one of the most attractive targets for anti-cancer therapy. Efforts to generate Plk1-specific inhibitors by targeting the catalytic activity of Plk1 have proven difficult due to similarities with the catalytic domains of other structurally related kinases. Here, we propose to develop a new class of mono-specific Plk1 inhibitors by employing a novel approach to target the non-catalytic, but functionally essential, PBD of Plk1. To this end, we have carried out a high throughput screen in collaboration with NCATS, Bethesda, MD. From this screen, we have identified 3,000 compounds from a primary screen, which were narrowed down to the final two compounds (1S and B7) through secondary medium throughput and tertiary cell-based assays. Through systematic molecular modeling/docking of these parent compounds, we obtained six significantly improved compounds, which showed Plk1 PBD inhibition activity at levels similar to that of one of the previously characterized PBD-binding ligands, PLHSpT (Kd = 450 nM), in an in vitro ELISA. Since the original HTS leads may belong to a class known as pan-assay interference compounds, or PAINS, we will further incorporate many drug-like characteristics during the hit-to-lead optimization in order to obtain a high-impact chemical probe that ultimately exhibits the desired efficacy against the Plk1 PBD in proof-of-concept mouse tumor models. As the first step of drug discovery, we generated the co-crystal structure of Plk1 PBD with one of the lead compounds. This structural model will be used to conduct structure-activity relationship studies and further optimize the compounds. As a result of this effort, we obtained several initial hits, subjected to several rounds of in silico derivatization and ELISA-based Plk1 PBD inhibition analysis. These efforts led to discovery of seven compounds (NC21 to NC27) with distinct chemotypes. In vitro FP and ELISA assays with purified compounds confirmed their anti-PBD specificity with Kd values in the 100-200 nM range. For the past two years, we have carried out hit-to-lead optimizations (in collaboration with Dr. Ken Jacobson, NIDDK, NIH) and developed 20 inhibitors that appear to potently inhibit Plk1 PBD inhibitors in vitro assays. However, the efficacy and bioavailability of these inhibitors exhibit less than expected. Now, the effort is focused on improving potency and permeability. This work has resulted in a paper (Alverez CN et al., J. Med. Chem. 2020) that describes several novel triazoloquinazoline-derived small-molecule prodrugs specific against Plk1 PBD. To follow up on this initial research outcome, we generated various derivatives to increase the potency and specificity against Plk1 PBD. This effort resulted in two subsequent papers (Park, JE et al., ACS Pharm. & Transl. Sci., 2023 and Park, JE et al., PNAS, 2023, in press), a US patent application (No. 18/028,463; filed on March 24, 2023), and a US provisional patent application (No. 63/455,608; filed on March 30, 2023) that covers the first allosteric Plk1 PBD inhibitor, called Allopole, with a unique mode of binding to Plk1 PBD1. The mode of Allopole binding to Plk1 PBD revealed an allosteric binding site that could selectively disrupt the canonical interactions between the PBD and its phospho-binding targets. Given the structural uniqueness and occluded nature of the Allopole-binding pocket, it could be advantageous to target this pocket over the canonical phosphoepitope-binding site, which requires a negatively charged ligand capable of mediating multiple water-mediated interactions. We are taking advantage of the unique properties of the Allopole-binding site to develop Plk1 PBD-specific inhibitors.