Linhardt R J, Cohen D M, Rice K G
Division of Medicinal and Natural Products Chemistry, College of Pharmacy, University of Iowa, Iowa City 52242.
Biochemistry. 1989 Apr 4;28(7):2888-94. doi: 10.1021/bi00433a021.
Computer simulation studies were used to prepare an ensemble of heparin number chains. The polydispersity of these chains was simulated by introducing a specific "fraction of terminators", and it closely resembled the experimentally observed polydispersity of a porcine mucosal, glycosaminoglycan heparin. The same percentage of simulated chains contained antithrombin III (ATIII) binding site sequences as are typically found to contain ATIII binding sites using affinity chromatography. Heparin lyase action was then simulated by using Michaelis-Menten kinetics. In one model, heparin chains were constructed from the random assembly of monosaccharide units using the observed mole percentage of each. After simulated depolymerization, the final oligosaccharides formed were compared to the observed oligosaccharide products. The simulation which assumed a random distribution of monosaccharide units in heparin did not agree with experimental observations. In particular, no ATIII binding site sequences were found in the simulated number chains. The results of this simulation indicate that heparin is not simply a random assembly of monosaccharide units. These results are consistent with the known, ordered biosynthesis of heparin. In a second model, heparin chains were constructed from randomly assembled oligosaccharides at the mole percentage in which each is found in the final product mixture. The action of heparin lyase was then simulated, and the distribution of the oligosaccharide products was measured throughout the simulated time course of the depolymerization reaction. The simulated rate of formation and final concentration of a particular oligosaccharide which contains a portion of heparin's ATIII binding site were similar to those observed experimentally. These results are consistent with the random distribution of ATIII binding sites within glycosaminoglycan heparin.(ABSTRACT TRUNCATED AT 250 WORDS)
计算机模拟研究用于制备肝素数链集合。通过引入特定的“终止剂分数”来模拟这些链的多分散性，其与实验观察到的猪黏膜糖胺聚糖肝素的多分散性非常相似。与使用亲和色谱法通常发现含有抗凝血酶III（ATIII）结合位点的比例相同的模拟链含有ATIII结合位点序列。然后使用米氏动力学模拟肝素裂解酶的作用。在一个模型中，肝素链由单糖单元的随机组装构建而成，使用每种单糖单元观察到的摩尔百分比。模拟解聚后，将形成的最终寡糖与观察到的寡糖产物进行比较。假设肝素中单糖单元随机分布的模拟结果与实验观察结果不一致。特别是，在模拟数链中未发现ATIII结合位点序列。该模拟结果表明肝素并非简单地由单糖单元随机组装而成。这些结果与已知的肝素有序生物合成一致。在第二个模型中，肝素链由最终产物混合物中每种寡糖以其摩尔百分比随机组装而成。然后模拟肝素裂解酶的作用，并在解聚反应的模拟时间进程中测量寡糖产物的分布。含有肝素一部分ATIII结合位点的特定寡糖的模拟形成速率和最终浓度与实验观察结果相似。这些结果与糖胺聚糖肝素中ATIII结合位点的随机分布一致。（摘要截于250字）
