Engineering and Evaluation of Human L-Methionase for Cancer Therapy

描述（由申请人提供）：神经母细胞瘤是儿童期最常见的颅外实体瘤，晚期疾病儿童的治愈率高达30%。显然需要新的化学治疗剂，因为目前的药物在高剂量下仅略微有效，导致毒性急性和严重的长期副作用。本提案的总体目标是采用蛋白质工程的现代技术来开发新一代非免疫原性和酶优化的酶，用于通过L-甲硫氨酸（L-Met）耗尽来对神经母细胞瘤和其他中枢神经系统（CNS）癌症进行化疗。L-Met不仅是蛋白质合成所必需的，也是甲基化还原和多胺生物合成的前体。肿瘤对L-Met的需求比正常组织大得多，并且当其可用性受到限制时会发生凋亡。静脉内施用细菌（假单胞菌）甲硫氨酸-g-裂解酶能够介导血清中L-Met的几乎完全耗尽，并且已经显示出显著抑制小鼠异种移植物中神经母细胞瘤、胶质母细胞瘤和前列腺癌的肿瘤生长。此外，已经报道了与微管解聚剂的强协同效应。不幸的是，在临床试验中，细菌酶显示出具有非常差的药理学性质（血清中的t1/2仅2小时），并且发现在灵长类动物中具有高度免疫原性，引起导致过敏性休克和死亡的严重不良反应。虽然人类基因组不编码任何甲硫氨酸裂解酶，但在初步研究中，我们部署了蛋白质工程策略以产生人类酶胱硫醚-g-裂解酶的潜在非免疫原性变体，其：（a）在体外和体内表现出高的L-Met降解活性，（B）对神经母细胞瘤细胞系表现出比它们的细菌对应物低的IC 50，和（c）在小鼠中稳定约10倍。在这里，我们将采用结构引导诱变和定向进化策略：1。工程师催化优化的“人L-甲硫氨酸酶”，即具有非常高的L-Met降解活性、在血清中更好的稳定性和高选择性的胱硫醚-g-裂合酶。2.通过位点特异性聚乙二醇化或通过融合至长内在无序多肽序列（XTEN），开发“人L-甲硫氨酸酶”的优化形式，以延长其在循环中的持久性，并确定其药代动力学和药效学特性。3.评价这些酶在使用诊断或复发时建立的临床细胞系形成的人神经母细胞瘤肿瘤小鼠异种移植模型中的疗效。将研究这些酶作为单药治疗和与长春新碱联合治疗的效用。 公共卫生相关性：神经母细胞瘤是儿童期最常见的颅外实体瘤，晚期患儿的治愈率高达30%。显然需要新的化学治疗剂，因为目前的药物在高剂量下仅略微有效，导致毒性急性和严重的长期副作用。我们正在开发一种由人类酶改造的药物，该药物可以通过攻击神经母细胞瘤细胞的关键代谢点来杀死肿瘤，并且当与低无毒剂量的当前化疗药物结合使用时，显示出梦幻般的前景。

DESCRIPTION (provided by applicant): Neuroblastoma is the most common extra-cranial solid tumor of childhood with an appalling 30% cure rate in children with advanced disease. There is a clear need for new chemotherapeutics, as current drugs are only marginally effective at the high doses that result in toxic acute and grave long term side effects. The overall objective of this proposal is to employ modern techniques of protein engineering to develop a new generation of non-immunogenic and pharmacologically optimized enzymes for chemotherapy of neuroblastomas and other central nervous system (CNS) cancers through L-Methionine (L-Met) depletion. L-Met is required not only for protein synthesis but also as the precursor for methylation reductions and for the biosynthesis of polyamines. Tumors have a much greater requirement for L-Met than normal tissues and become apoptotic when its availability is restricted. i.v. administration of bacterial (Pseudomonas) methionine-g-lyase is able to mediated near complete depletion of L-Met in serum and has been shown to drastically inhibit tumor growth of neuroblastomas, glioblastomas and prostate carcinomas in mouse xeongrafts. Furthermore strong synergistic effects with microtubule depolymerization agents have been reported. Unfortunately, in clinical trials the bacterial enzyme was shown to have very poor pharmacological properties (t 1/2 in serum only 2 hrs) and was found to be highly immunogenic in primates eliciting severe adverse responses that resulted in anaphylactic shock and death. While the human genome does not encode any methionine lyase enzymes, in preliminary studies we deployed protein engineering strategies to generate potentially non-immunogenic variants of the human enzyme cystathionine-g-lyase that: (a) exhibit high L-Met degradation activity in vitro and in vivo, (b) display a lower IC50 for neuroblastoma cell lines than their bacterial counterparts and (c) are about 10-fold more stable in mice. Here we will employ structure guided mutagenesis and directed evolution strategies to: 1. Engineer catalytically optimized "human L-methioninases" i.e. cystathionine-g-lyase enzymes with very high activity for L-Met degradation, even better stability in serum and high selectivity. 2. Develop optimized formats of the "human L-methioninases" for prolonged persistence in circulation by either site-specific PEGylation or by fusion to long intrinsically disordered polypeptide sequences (XTEN) and determine their pharmacokinetic and pharmacodynamic properties. 3. Evaluate the efficacy of these enzymes in the mouse xenograft model of human neuroblastoma tumors formed using clinical cell lines established either in diagnosis or relapse. The utility of these enzymes will be investigated both as monotherapy and in combination therapy with vincristine. PUBLIC HEALTH RELEVANCE: Neuroblastoma is the most common extra-cranial solid tumor of childhood with an appalling 30% cure rate in children with advanced disease. There is a clear need for new chemotherapeutics, as current drugs are only marginally effective at the high doses that result in toxic acute and grave long term side effects. We are developing a drug engineered from a human enzyme that can kill tumors as a single agent by attacking a key metabolic point of neuroblastoma cells and is showing fantastic promise when used in combination with low non-toxic doses of current chemotherapeutics.