紫苏（Perilla frutescens (L.) Britton），属于唇形科，是一种药食同源植物，其环境适应性强，产量大，主要分布于中国、日本和韩国地区。紫苏富含多种生物活性物质，如紫苏油、多酚、α或γ生育酚和类胡萝卜素等。半乳糖甘油酯是一种重要的膜脂，其具有抗炎、抑菌和抗氧化等多种生物活性。紫苏含有丰富的单半乳糖甘油二酯（monogalactosyldiacylglycerol，MGDG）和双半乳糖甘油二酯（digalactosyldiacylglycerol，DGDG），植物叶片中的半乳糖甘油酯含量比茎和根中的丰富，同时关于紫苏中半乳糖甘油酯的研究较少，因而选择紫苏叶作为提取半乳糖甘油酯的原料。因分离纯化困难以及纯度低等缺陷，半乳糖甘油酯不能大规模工业化生产，同时关于其抗炎活性构效关系及作用位点的研究很少，因此研究了半乳糖甘油酯的分离纯化方法、抗炎活性构效关系及作用机制。

本研究建立了一套方便快捷的分离纯化紫苏中半乳糖甘油酯的方法，评价半乳糖甘油酯的抗炎活性，从而揭示其抗炎活性构效关系，并对半乳糖甘油酯的抗炎结合位点进行了探索。主要研究结果如下：

1、使用75%的乙醇提取紫苏粉，用水和乙酸乙酯萃取得到总脂，得率为12.4%。采用一步硅胶柱层析和制备型HPLC-ELSD纯化总脂中的半乳糖甘油酯并通过TLC鉴定，结果表明硅胶柱层析法使MGDG和DGDG型甘油糖脂完全分离，再经过制备型HPLC-ELSD纯化，MGDG和DGDG混合物共被分离成10个单峰。通过TLC再次鉴定，仍被鉴定为甘油糖脂类物质。采用GC-MS和LC-MS鉴定半乳糖甘油酯，共得到2种MGDG和3种DGDG，经分析型HPLC-ELSD检测，纯度大于95%，其中1种MGDG和1种DGDG在实验室前期工作中未被表征。通过NMR、脂肪酸连接位置以及半乳糖构型鉴定，MGDG被鉴定为1-O-9Z, 12Z, 15Z-十八碳三烯酰基-2-O-7Z, 10Z, 13Z-十六碳三烯酰基-3-O-(β-D-吡喃半乳糖)-sn-甘油，DGDG被鉴定为1-O-9Z,12Z-十八碳二烯酰基-2-O-9Z,12Z,15Z-十八碳三烯酰基-3-O-(α-D-吡喃半乳糖-(1'→6'')-β-D-吡喃半乳糖)-sn-甘油。

2、基于LPS诱导的小鼠巨噬细胞RAW264.7对5种半乳糖甘油酯进行了抗炎活性评价，从而揭示抗炎活性构效关系。通过MTT实验筛选5种半乳糖甘油酯和LPS的安全浓度，并筛选LPS诱导炎症产生的最佳浓度，采用ELISA法和RT-qPCR法测定半乳糖甘油酯对NO、IL-6、IL-1β和TNF-α含量以及相关炎症基因如iNOS、COX-2、IL-6、IL-1β和TNF-α表达量的影响。结果表明，5种半乳糖甘油酯的安全浓度为3.125、6.25和12.5μg/mL，LPS使得炎症产生的最佳浓度为500ng/mL。5种半乳糖甘油酯均能显著性抑制炎症因子的释放及炎症基因的表达（p<0.05），并呈剂量依赖性。通过对比不同结构的半乳糖甘油酯之间的抗炎活性差异，综合来看，发现MGDG 2比MGDG 1抗炎活性强，DGDG 1和DGDG 2比DGDG 3活性强，表明脂肪酸的链长影响半乳糖甘油酯的抗炎活性，随着链长的增加，活性逐渐增强。同时MGDG 2比DGDG 1的活性强，表明脂肪酸的不饱和度也影响抗炎活性，脂肪酸不饱和程度越高，抗炎活性越强。

3、通过分子对接软件DS模拟5种半乳糖甘油酯与TLR4/MD-2蛋白的结合能力，预测其抗炎结合位点。接下来评价TLR4/MD-2阻断剂和MGDG 2对相关炎症因子NO、IL-6、IL-1β和TNF-α以及iNOS、COX-2、IL-6、IL-1β和TNF-α基因的影响，验证半乳糖甘油酯的抗炎结合位点，从而探究抗炎机制。实验结果表明，5种化合物均能结合于MD-2蛋白的疏水空腔内。通过对比氢键数量及强弱、疏水键数量、范德华力及结合能大小，发现与其抗炎活性大小一致。综合来看，不同浓度的TLR4/MD-2阻断剂＋LPS组中炎症因子及基因表达水平随着浓度的增加逐渐下降，各组均显著低于LPS模型组（p<0.05），表明随着阻断剂浓度的增加，其与MD-2空腔结合得越来越多，导致LPS结合变少，从而抑制炎症因子分泌和基因表达。当采用阻断剂处理后，再使用MGDG 2处理，发现阻断剂＋MGDG 2＋LPS组中炎症因子及基因表达水平与同浓度的阻断剂＋LPS组相比显著性降低（p<0.05），同时随着浓度的增加，阻断剂＋MGDG 2＋LPS组的量逐渐接近对照组，表明随着阻断剂浓度的增加，MD-2的位点被封闭的越多，当MGDG 2与剩余位点继续结合，从而导致LPS结合的越来越少，甚至不能结合，表明半乳糖甘油酯的结合位点为TLR4/MD-2蛋白，其与LPS竞争MD-2结合位点，从而发挥抗炎作用。

Perilla frutescens (L.) Britton, belonging to the family Labiatae, is an edible-medicinal plant that is highly adaptable to environment and produces large yields, mainly in China, Japan and Southeast Korea, and rich in many bioactive substances, such as perilla oil, polyphenols, α or γ tocopherols and carotenoids. Galactoglycerolipids are important membrane lipids with a variety of biological activities such as anti-inflammatory, antibacterial and antioxidant. Perilla is rich in monogalactosyldiacylglycerol (MGDG) and digalactosyldiacylglycerol (DGDG), and the plant leaves are richer in galactoglycerolipids than the stems and roots. Studies on galactoglycerolipids in perilla are fewer, so perilla leaves were chosen as the raw material for extraction. Due to the difficulties in isolation and purification and low purity, galactoglycerolipids could not be produced industrially on a large scale, and there were few studies on their structure-anti-inflammatory activity relationship and mechanism of action. Therefore, the separation and purification method, structure-anti-inflammatory activity relationship and mechanism of action of galactoglycerolipids were studied.

In this study, a convenient and quick method was developed for isolation and purification of galactoglycerolipids from perilla. The anti-inflammatory activities of them were evaluated to reveal the structure-anti-inflammatory activity relationship, and the anti-inflammatory binding site was explored. The main findings were as follows:

1. The perilla powder was extracted with 75% ethanol. Total lipids were extracted with water and ethyl acetate. The yield was 12.4%. Galactoglycerolipids in total lipids were purified by one-step silica gel column chromatography and preparative HPLC-ELSD, and identified by TLC, showing that MGDGs and DGDGs were completely separated by silica gel column chromatography and purified by preparative HPLC-ELSD to gain 10 single peaks. The substances in these peaks were still glycerolipids re-identified by TLC. Two MGDGs and three DGDGs with more than 95% purity by analytical HPLC-ELSD were identified by GC-MS and LC-MS, of which one MGDG and one DGDG were not characterized in the preliminary laboratory work. By NMR, fatty acid linkage positions, and galactose configuration identification, MGDG was identified as 1-O-9Z,12Z,15Z-octadecatrienoyl-2-O-7Z,10Z,13Z-hexadecatrienoyl-3-O-(β-D-galactopyranosyl)-sn-glycerol, and DGDG was identified as 1-O-9Z,12Z-octadecadienoyl-2-O-9Z,12Z,15Z-octadecatrienoyl-3-O-(α-D-galactopyranosyl-(1'→6'')-β-D-galactopyranosyl)-sn-glycerol.

2. The anti-inflammatory activities of the five galactoglycerolipids were evaluated based on LPS-induced mouse macrophage RAW264.7, revealing the structure-activity relationship. The safe concentrations of the five galactoglycerolipids and LPS were screened by MTT assay, and the optimal concentration for LPS-induced inflammation production was screened. The effects of galactoglycerolipids on the levels of NO, IL-6, IL-1β and TNF-α and the expression of related inflammatory genes such as iNOS, COX-2, IL-6, IL-1β and TNF-α were determined by ELISA and RT-qPCR. The results showed that the safe concentrations of the five galactoglycerolipids were 3.125, 6.25 and 12.5μg/mL, and the optimal concentration of LPS-induced inflammation production was 500ng/mL. All five galactoglycerolipids significantly inhibited the release of inflammatory factors and gene expression (p<0.05) in a dose-dependent manner. By comparing the differences of anti-inflammatory activities of different structures of galactoglycerolipids, MGDG 2 had stronger anti-inflammatory activity than MGDG 1. DGDG 1 and DGDG 2 were more active than DGDG 3, indicating that the chain length of fatty acids affected the anti-inflammatory activities, and the activities were gradually enhanced with the increase of chain length. Also, MGDG 2 was more active than DGDG 1, indicating that the degree of unsaturation of fatty acids also influenced the anti-inflammatory activities. The higher the degree of fatty acid unsaturation, the stronger the anti-inflammatory activities.

3. The ability of the five galactoglycerolipids bind to TLR4/MD-2 protein was simulated by the molecular docking software DS, predicting their anti-inflammatory binding site. Next, the effects of TLR4/MD-2 blocking agent and MGDG 2 on related inflammatory factors such as NO, IL-6, IL-1β and TNF-α as well as iNOS, COX-2, IL-6, IL-1β and TNF-α genes were evaluated to validate the anti-inflammatory binding site of galactoglycerolipids, exploring the anti-inflammatory mechanism. The results showed that all five compounds were able to bind within the hydrophobic cavity of MD-2 protein. By comparing the number and strength of hydrogen bonds, the number of hydrophobic bonds, van der Waals forces and binding energy, they were consistent with their anti-inflammatory activities. Overall, the levels of inflammatory factors and gene expression in TLR4/MD-2 blocking agent + LPS group decreased gradually with increasing concentrations, and were significantly lower than those in LPS model group (p<0.05), indicating that as the concentration of the blocking agent increased, they bound more and more to the MD-2 cavity, resulting in less LPS binding and thus inhibiting inflammatory factors secretion and gene expression. When treated with blocking agent and then treated with MGDG 2, the levels of inflammatory factors and gene expression in blocking agent + MGDG 2 + LPS group were significantly lower than those in blocking agent + LPS group at the same concentration (p<0.05). As the concentration increased, the levels of blocking agent + MGDG 2 + LPS group gradually approached control group, indicating as the concentration of blocking agent increased, more MD-2 sites were closed. MGDG 2 continued to bind to the remaining sites, resulting in less and less LPS binding or even failing to bind. The binding site for galactoglycerolipids was shown to be the TLR4/MD-2 protein, which competed with LPS for the MD-2 binding site and thus exerted anti-inflammatory effects.