Oligosaccharide substrate interactions with b-1,4-Galact

糖偶联物的寡糖部分在细胞的几个生物学过程中发挥着重要作用，包括糖蛋白的折叠和跨细胞间隔的运输。对于这些复杂低聚糖的生物合成，细胞中存在着一种复杂的机制。糖链合成缺陷会造成严重的病理后果，并导致多种人类疾病。低聚糖以高度的特异性结合到细胞蛋白质上，并调节糖蛋白的同源和异构化。由于低聚糖的构象灵活性，二糖单元的扭转角，特别是围绕A1-6-键的扭转角，以一种方式调整，使寡糖的侧基以促进与蛋白质结合残基的有利相互作用的方式自我定位。分枝低聚糖使蛋白质交联，并产生无限的蛋白质-碳水化合物复合体网络，从而调节各种细胞反应。 通过将低聚糖对接到结合位点和对复合物的分子动力学模拟来确定Gal-T1的寡糖结合部位：我们继续使用分子模拟方法来研究低聚糖与蛋白质的结合，尤其是各种寡糖底物与Gal-T1的结合，最近我们实验室通过测定其三维结构，无论是在与UDP-半乳糖和Mn2+的复合物中，还是在与α-乳蛋白和N-乙酰氨基葡萄糖的复合物中(见项目#Z01 BC 009305-06 LECB)。从Gal-T1 LA GlcNAc晶体结构中检测Gal-T1中的GlcNAc结合部位，发现在GlcNAc结合部位后面有一个“开放的沟渠状”延伸的糖结合部位，平均宽度和长度为10？和16岁？这个位点由三个区域的残基组成：残基280到289，残基319到325，残基359到368。在Gal-T1-LA络合物的晶体结构中，LA与这一区域结合；因此，它有望与含有壳二糖等低聚糖的GlcNAc竞争。作为Gal-T1底物的优选低聚糖数量有限。这些研究表明，在不同的含有二糖的GlcNAc中，只有b连接的二糖如GlcNAcb1，4-GlcNAc或GlcNAcb1，2-Man比a连接的二糖更受欢迎。事实上，与GlcNAc本身相比，a-甲基-GlcNAc不那么受欢迎。此外，与(GlcNAc)4相比，低聚糖如N-葡聚糖是更好的受体底物。连接到这种N-糖链上的蛋白质或肽的存在不会影响低聚糖的结合。为了探索低聚糖结合部位的大小和性质，对不同的二糖和N-葡聚糖配体在结合部位的对接进行了模拟研究。利用InsightII中的CVF力场和Discover模块计算了各配体的构象、Gal-T1与糖的分子间相互作用能。能量计算只考虑了来自任何配体原子的9A以内的蛋白质残基。通过系统地扫描配体的整个立体化学允许区域，配体的构象在结合位置上发生变化。总能量，包括分子间蛋白质-配体相互作用能和分子间配体能量，被用作确定配体在结合位置上可能的允许构象的指导。这些模拟研究表明，具有a-取代的GlcNAc如a-苄基-GlcNAc不能与Gal-T1结合，因为与高度保守的Tyr286残基有严重的空间接触，而具有b-取代的GlcNAc如b-苄基-GlcNAc可以在没有任何空间接触的情况下结合。二天线N-糖链与扩展糖结合位点上的GlcNAcs的对接表明，Gal-T1中的受体结合位点可以容纳从GlcNAc部分到天冬氨酸连接的GlcNAc的线性五糖。结合位点还可以容纳N-糖链的a-1-3臂(GlcNAc3b1-2Man3a1-3Manmb1-4GlcNAcb1-4GlcNAc-N)或a-1-6臂(GlcNAc6b1-2Man6a1-6Manmb1-4GlcNAcb1-4GlcNAc-N)，而不存在任何空间位阻。水中的MD模拟(75？Cube)对Gal-T1单独或与对接的双触角低聚糖结合部位的作用表明，在MANA1-3Manm和GlcNAc6b1-2Man6连接处的y构象变化显著地发生在结合的双触角低聚糖的1-3臂上。在双触角低聚糖结合的1-6臂中，Manmb1-4GlcNAc连接处的y和Man6a1-6Manm连接处的f的构象变化显著。MD模拟结果表明，当1-3-臂结合在结合部位时，低聚糖与Gal-T1之间存在疏水作用，从而有利于半乳糖转移到双触角低聚糖1-3-臂上的GlcNAc上。这一结论与已发表的生化数据一致。 在人类中，Gal-T1家族成员负责合成不同寡糖中的Gal部分，这表明虽然所有这些酶都将Gal转移到GlcNAc，但每个成员都识别GlcNAc所连接的剩余寡糖部分。人类Gal-T家族成员的序列比较表明，这些家族成员之间的GlcNAc结合位点只有很小的变化或没有变化，而扩展的寡糖结合区则显示出显著的差异，这表明这些酶可能偏好不同的含GlcNAc的寡糖作为它们的首选糖受体。为了确定结合部位低聚糖的确切结合方式，作为第一次尝试，我们成功地将Gal-T1突变体与二糖和五糖共结晶。目前正在确定这些络合物的晶体结构。

The oligosaccharide moieties of glycoconjugates play important roles in several biological processes of a cell, including the folding and transport of glycoproteins across cellular compartments. For the biosynthesis of these complex oligosaccharides, an intricate machinery exists in a cell. Defective glycan synthesis has serious pathological consequences and results in several human diseases. The oligosaccharide moieties bind to cellular proteins with high specificity and modulate the homo- and hetrodimerization of glycoproteins. Due to the conformational flexibility of oligosaccharides, the torsional angles of a disaccharide unit, especially around the a1-6-linkage, adjust in such a way that the side groups of the oligosaccharides orient themselves in a manner that promotes favorable interactions with the binding residues of the protein. Branched oligosaccharides cross-link proteins and generate infinite networks of protein-carbohydrate complexes, resulting in the modulation of various cell responses. Defining the oligosaccharide binding site of Gal-T1 by docking oligosaccharides into the binding site and MD simulation of the complexes: We have continued to use molecular modeling methods to study the binding of oligosaccharides to proteins, in particular the binding of various oligosaccharide substrates to Gal-T1, the three dimensional structure of which has recently been determined in our laboratory, either in complex with UDP-galactose and Mn2+ion, or in complex with a-lactalbumin and N-acetylglucosamine (see Project # Z01 BC 009305-06 LECB). Examination of the GlcNAc binding site in Gal-T1 from the Gal-T1 LA GlcNAc crystal structure reveals an "open canal shaped" extended sugar binding site that lies behind the GlcNAc binding site, with an average width and length of 10 ? and 16 ?, respectively. This site is formed by the residues from three regions; residues 280 to 289, residues 319 to 325, and residues 359 to 368. LA binds to this region in the crystal structure of Gal-T1-LA complex; therefore, it is expected to compete with the GlcNAc containing oligosaccharides such as chitobiose. A limited number of preferred oligosaccharides as substrates for Gal-T1 are known. These studies have shown that among the different GlcNAc containing disaccharides, only a b-linked disaccharide such as GlcNAcb1,4-GlcNAc or GlcNAcb1,2-Man is preferred over a-linked disaccharides. In fact a-methyl-GlcNAc is less preferred compared to GlcNAc by itself. Also, oligosaccharides such as N-glycans are more preferred acceptor substrates than a (GlcNAc)4. The presence of the protein or peptide attached to such an N-glycan does not influence the binding of an oligosaccharide. In order to probe the size and nature of the oligosaccharide binding site, a modeling study of the docking of various disaccharides and N-glycan ligands in the binding site were carried out. Each ligand conformation, inter-molecular interaction energy between Gal-T1 and the saccharide was calculated using CVFF force field and Discover module of InsightII. Only protein residues within 9 A from any of the ligand atoms were considered for energy calculations. The conformation of the ligand was varied in the binding site by systematically scanning the entire stereochemically allowed region of the ligand. The total energy, comprising the intermolecular protein-ligand interaction energy and intermolecular ligand energy, was used as a guide in determining the possible allowed conformations of the ligand in the binding site. These modeling studies show that GlcNAc with an a-linked substitution such as a-benzyl-GlcNAc can not bind to Gal-T1 because of severe steric contacts with the highly conserved Tyr286 residue, whereas GlcNAc with a b-linked substitution such as b-benzyl-GlcNAc can bind without any steric contacts. Docking of a biantennary N-glycan with GlcNAcs at its reducing ends in the extended sugar binding site reveals that the acceptor binding site in Gal-T1 can accommodate a linear pentasaccharide all the way from the GlcNAc moiety to the aspargine-linked GlcNAc. The binding site can also accommodate either the a-1-3 arm (GlcNAc3b1-2Man3a1-3Manmb1-4GlcNAcb1-4GlcNAc-N) or a-1-6 arm (GlcNAc6b1-2Man6a1-6Manmb1-4GlcNAcb 1-4GlcNAc-N) of the N-glycan without any steric hindrance. MD simulations in water (75 ? cube) for 750 ps of Gal-T1 alone or with the docked bi-antennary oligosaccharide in the binding site show that the conformational changes in the y at the Mana1-3Manm and GlcNAc6b1-2Man6 linkage are pronounced in the bound 1-3-arm of the bi-antennary oligosaccharide. Conformational changes in y at the Manmb1-4GlcNAc linkage, and f at the Man6a1-6Manm linkage are pronounced in the bound 1-6-arm of the bi-antennary oligosaccharide. The MD simulation results suggest hydrophobic interactions between the oligosaccharide and Gal-T1 when 1-3-arm is bound in the binding site, thus favoring the transfer of galactose to the GlcNAc on the 1-3-arm of the bi-antennary oligosaccharide. This conclusion concurs with the published biochemical data. In humans Gal-T1 family members are responsible for the synthesis of Gal moiety in different oligosaccharides, indicating that although all these enzymes transfer Gal to GlcNAc, each recognizes the remaining oligosaccharide moieties to which GlcNAc is attached differently. The sequence comparison of the human Gal-T family members reveals only a little or no variation in the GlcNAc binding site among the family members where as the extended oligosaccharide binding region shows significant variations, indicating that these enzymes may prefer different GlcNAc-containing oligosaccharides as their preferred sugar acceptors. To determine the exact mode of binding of the oligosaccharide in the binding site, as a first attempt we have been successful in co-crystallizing Gal-T1 mutants with di and a penta-saccharide. The crystal structure of these complexes is currently being determined.