【作者】 李莉；

【导师】 杨玉良；

【作者基本信息】 复旦大学 ， 高分子化学与物理， 2005， 博士

【摘要】 在生命体系中，脂质双分子层膜无处不在，它不仅表现出形状变化上的丰富多样性，而且还与细胞的各种生理功能紧密相关，如物质的定向输运、胞排、胞饮及信号传递等，而出芽是这些重要生命活动的首要基本过程。为了认识和发掘这些生物过程的内在机制，由多组分脂质分子构成的模型体系成为近年来膜生物学家和膜生物物理学家进行理论与实验研究的中心。生物学家已经发现，在富含胆固醇的模型体系，如DOPC／DPPC／Chol中，从高温的均相区域淬冷到低温的分相区域，膜表面会发生侧向相分离，形成微畴，而微畴诱导出芽的过程类似于细胞膜的出芽。借助于荧光显微镜，就可以通过研究微米尺度上的聚集态的形貌、形变、运动来了解纳米水平上的分子机理。另一方面，Helfrich的生物膜弯曲弹性理论在解释均匀膜的形变现象时取得了很大的成功，但对于多组分的囊泡体系，形变与分相的耦合使描述体系的自由能泛函变得复杂，在有关凸芽的生成与形状、凸芽间的融合、出芽过程的动力学规律等各项理论研究中，理论模型的合理性、计算或模拟结果与实验现象的对应性都有待实验的检验。因此，制备能发生出芽现象的多组分囊泡、观察微畴的形成与长大、凸芽的生成与融合等动态过程将有助于了解象神经细胞凸出、棘皮细胞的衰亡及上述各种细胞功能发生的内在原因，也有助于检验各项理论工作的正确性。为了研究多组分囊泡的侧向相分离与出芽过程的动力学，我们分别设计搭建和筹建了两套系统，以实现显微镜下的原位制样、实时观察、操作、记录样品的成膜、水合、溶涨、分相、出芽与形变等等过程；用弹性理论分析实验结果，并与现有的理论模型及模拟结果进行比较，主要的研究内容和结果可分为6个部分：1．提出了两种原位制备多组分出芽囊泡的制样方法：直接水合法与电场法；搭建了一套原位制样观测装置，将原有的偏光显微镜改装成荧光显微镜，配置了1394接口的CCD摄像头及特别适合动力学研究的时温图像采集软件。设计制作了由信号发生器与示波器控制的样品池Cell，实现原位电场法制样。将热台与预水合装置相连，形成了一套具有原位制样、预水合、水合功能，并可实时观察记录时温图像的工作系统。通过对模型体系各组分的研究，提出了最佳样品制备条件。2．研究了单个凸芽的生长过程，发现凸芽长大到极限尺寸后，并不立刻颈缩成球形，而是以管形形状稳定存在着。经过理论分析，证明从极限尺寸的半球形到管形凸芽的过程是一种能量有利的方式。对凸芽与凸芽及凸芽与微畴间相互融合的各种模式也进行了研究，发现凸芽之间是以某个夹角从底部开始融合，而不是以平行的方式相互融合。凸芽与微畴在膜中是以整体方式迁移扩散，融合的方式由模的弯曲刚性及脂质的整体扩散系数决定。这些实验结果与理论研究的模型不同。3．研究了多组分单层囊泡出芽的动态过程，发现最小面积一体积比的球形囊泡与较大面积一体积比的管形囊泡具有不同的形变出芽行为，可支配的超额膜面积是体系形成大量凸芽与形变的先决条件。由管形囊泡的出芽、融合行为及形变的研究，我们得到了凸芽数与生长时间的动力学曲线，拟合得到N_bud～t^-2／3的标度关系，与Kumar的Monte Carlo模拟及Laradji的DPD模拟结果一致。另外，发现球面上微畴与凸芽的排列具有球面结晶学特征，自组装中存在的缺陷排列成大角晶界，也称疤痕，这些疤痕是球面上平衡晶体的基本结构元素。4．在DPPC／DOPC／Chol与BSM／DOPC／Chol模型体系中均发现了紧密盘绕的单螺旋与双螺旋形囊管结构。通过对体系中流体施加剪切力，发现螺旋结构的膜之间不存在粘连关系，单螺旋结构还能象弹簧那样被均匀地拉伸并自动回复。盘绕中的错缺—孤子可以沿着螺管移动，并且是螺旋结构最终崩溃的原因。比较DPPC、BSM、DOPC单组分体系和DPPC／DOPC、DPPC／Chol、DOPC／Chol双组分体系，均未发现螺旋管结构出现，推测胆固醇引起的饱和组分与不饱和组分的分布不均匀是诱发螺旋结构形成的可能原因。5．研究了多组分多层膜囊泡的分相与形变，发现分相开始的初期，各层的分相是相互独立的，无空间上的相关性；随着时间的推移和相分离的继续进行，各层的相区会自行调整，形成相同相区的径向整齐排列；明场下发现这种排列与囊泡的形变区域相互对应。相邻层膜间的间隙将影响相区在膜内的迁移与囊泡形变的协调性。分析这种协调现象发生的原因，可能是不同相区膜的弯曲模量的差异与闭合囊泡存在体积限制两种因素共同作用的结果。6．用离散空间变分法研究多组分单层与双层膜囊泡的分相与形变，通过理论模拟，研究了界面张力、模量、压力差及自发曲率差等因素对单层囊泡出芽与形变的影响，还研究了双层膜囊泡的膜间隙对各层膜中相区分布、形变的影响。更多还原

【Abstract】 Lipid bilayer membranes are ubiquitous in biological systems, which exhibit rich shape transition behavior, and are closely related with cell functions, such as lipid transport, exocytosis, endocytosis, signaling and so on. Budding is always the first and important step in all these processes. To understand the mechanism of these biological processes, investigations of the budding on the multicomponent lipid vesicles have become a hot topic in membrane biology and biophysics. In this thesis, we focus on the study of the lateral phase separation and budding dynamics on the multicomponent lipid vesicles. The main results are summarized in the following:1. We have designed and setup two systems, which enable to prepare the samples in situ, and to study in real time the phase separation, budding, and its coalescence under the microscope.2. The growth of a single bud has been studied in real time under the microscope. It is found when the bud develops to a flat-bottom half sphere and reaches to its maximum size, it does not form a complete sphere on the membrane as predicted theoretically by Lipowsky et al., but continue to grow up to a tubular-shape bud with the fixed limiting radius. This new result can be interpreted by a model calculation based on a simple thermodynamic argument. The coalescence among buds and domains are also investigated in detail. The results show that the domains and buds diffuse in membrane via the aggregation of molecules, and the coalescence depends on both the bending rigidity and the bud （domain） diffusion coefficient. In addition, it is also observed that two tubular buds coalesce each other with an angle from their roots.3. For the budding dynamics of multicomponent unilamellar vesicles, it is found that the shape deformations are different for the spherical vesicles with smaller area-to-volume ratio and the tubular vesicles with larger area-to-volume ratio. The non-sphere shape vesicles are necessary for the formation of large number of buds, the number of buds on tubular vesicle as a function of growth time with a scaling law of N_bud～t^-2/3 has been determined, which agrees quite well with the theoretical works by Kumar et al and Laradji et al. Another interesting observation is that the packing of domains and buds on a large spherical surface obeys the crystallography on a sphere.4. Tight coiling single and double helical vesicles are observed in both DPPC/DOPC/Chol and BSM/DOPC/Chol lipid mixtures. Under some controlled perturbations, the single helix could be stretched smoothly and recovers automatically like a spring. In addition, the soliton-like coiling defects are observed to move along the helical structure, which seems to serve as the trigger of instability of helical vesicles.5. The phase separation and shape deformation of multilamellar vesicles in DPPC/DOPC/Chol mixture are investigated for the first time. It is found that the formation of domains on each lamellar is independent to each other at the initial stage, then domains start to move on the membrane spontaneously and line up along the radial direction. It is found that such a phase rearrangement is corresponds to the shape deformation.6. Phase separation and shape deformation of unilamellar and double lamellar vesicles are studies using the discrete-spatial variation method. It is found and discussed in details that the line tension, pressure difference, bending rigidity and spontaneous curvature are all important for the budding process, while the space between two lamellas in double lamellar vesicles are critical for the phase rearrangement.更多还原