如果把你在《生物化学》里背过的糖酵解、三羧酸循环、激酶级联磷酸化统统倒进一只敞口烧杯里，生命并不会发生——分子会迅速稀释、信号会彼此串扰、质子梯度会在一瞬间消散。MITx 7.06.1x 的第一堂课上，Rebecca Lamason 教授一走上讲台就抛出了整个细胞生物学的出发点：生化和遗传学告诉我们分子能做什么，而细胞生物学真正迷人的地方，在于“语境（context）”——当这些生化反应被放进一个真实的活细胞空间里，细胞究竟如何划定边界、建立秩序、精准调控每一次反应？

要造出一个活细胞，自然界首先面对一个看似矛盾的工程学难题：细胞内外都是水相环境，它必须筑起一道严密的疏水高墙（barrier），把水溶性的离子、糖类和蛋白质牢牢挡住，从而区分“胞内与胞外”“细胞质基质与细胞器腔”；但与此同时，这道墙又绝不能是一堵死气沉沉的砖墙（brick wall）——细胞要爬行变形、要分裂增殖、要让散布在膜表面的受体迅速聚拢放大信号，这道墙本身又必须是一片能够自由流动的二维海洋（fluid）。

为什么由两性脂质自发拼成的 5 nm 薄膜，能同时做到“既是一道墙，又是一片海”？细胞如何调配甘油磷脂、鞘脂和胆固醇的比例来缓冲温度剧变？膜蛋白又是如何镶嵌进疏水核心、如何保持严格的拓扑不对称性、又是如何在实验室里被我们用荧光漂白恢复（FRAP）和表位标签（Epitope Tagging）看清动态与朝向的？让我们跟随 Lamason 教授的黑板粉笔与课堂实验，结合 MBoC 第 7 版的分子物理机制与翟中和《细胞生物学》第 5 版的考研核心口径，推开细胞质膜的大门。

---

1.1膜的屏障使命、自组装热力学与结构模型五步演进

一、从“油水沙拉酱瓶”看生物膜的边界本质

在 MIT 7.06 的第一堂课上，Rebecca Lamason 教授举起了一只由 Iain Cheeseman 教授帮她准备的透明演示瓶——瓶底是染成蓝色的水，上层浮着一层金黄色的油。她笑着问台下的本科生：“我拎着这只瓶子一路穿过整个 MIT 校园走过来，为什么油和水依然没有混在一起？”

台下学生回答：“因为它们是不混溶的（immiscible）——油是疏水的（hydrophobic），而水是亲水的（hydrophilic）。”

紧接着，Lamason 教授当堂用力摇晃瓶子。油层瞬间被打散成无数细小的油滴悬浮在蓝水中，但静置片刻后，小油滴为了尽可能把疏水尾部藏起来、减少与水分子的接触面积，纷纷彼此靠拢、融合（fuse），很快又重新分层。你在家里摇晃油醋汁沙拉酱（salad dressing）时看到的就是完全相同的热力学过程。

把这个日常物理现象放到真核细胞里，你会发现细胞面临着双重边界需求：第一重边界在细胞最外围，用来区分“细胞内（inside）”与“细胞外（outside）”；第二重边界在细胞内部，用来区分“细胞质基质（cytosol）”与“细胞器腔（lumen of an organelle）”。在翟中和《细胞生物学》第 5 版中，这两重边界对应着三个必须严格区分的基础概念：

1. 细胞质膜（plasma membrane，曾称细胞膜）：细胞质膜（plasma membrane）：围绕在细胞最外层，由脂质、蛋白质和糖类组成的生物膜，作为细胞界膜维持内环境稳定，并在物质运输、能量转换和信息传递中起关键作用。由于质膜厚度仅约 5～10 nm（脂双层核心约 5 nm），光学显微镜下看不到质膜结构，直到 20 世纪 50 年代电子显微镜问世后才首次直接观察到质膜。
2. 细胞内膜（intracellular membranes）：真核细胞内部由膜围成的各种细胞器膜。其中既有单层膜（内质网膜、高尔基体膜、溶酶体膜、过氧化物酶体膜等），也有双层膜（线粒体膜、叶绿体膜、核膜）。
3. 生物膜（biomembrane / membrane structure，又称膜相结构）：生物膜（biomembrane）：细胞质膜与细胞内膜（细胞器膜）的统称，主要由脂质、蛋白质和糖类组成，以连续的脂双层为共同基本结构 2009·名解 ⚑华农考过。

二、磷脂双分子层的自组装热力学：为什么形成双层囊泡而非单层胶束？

既然油滴能靠疏水作用把水挡在外面，细胞能不能直接用普通油脂（如三酰甘油）或者单层油膜来做细胞膜呢？Lamason 教授在黑板上画了一张示意图，引导大家思考：在刚才的油水瓶里，水面之上是空气，所以简单油分子可以把亲水头朝向水、疏水尾朝向空气排成单层；但在活细胞里，质膜外侧是水相（细胞外液），内侧也是水相（细胞质基质）！ 如果只排一层，疏水尾部必然有一侧会直接暴露在水分子面前，这在热力学上是绝对行不通的。

因此，细胞选择了具有双重性格的两亲性分子（amphipathic molecules，国内教材常称两性分子）——磷脂（phospholipids）。每个磷脂分子都拥有一个亲水的极性头部（polar head group）和两条疏水的非极性脂肪酸尾部（nonpolar fatty acid tails）。当大量磷脂进入水相时，它们无需任何酶催化或 ATP 供能，就能自发组装成磷脂双分子层（phospholipid bilayer）：两个单分子层的疏水尾部面对面聚拢在膜内部形成疏水核心（hydrophobic core），而极性头部则分别朝向两侧的水相。构成脂双层的每一层磷脂分子，在细胞生物学中被称为一个小叶（leaflet）——朝向胞外的一层叫外小叶（outer leaflet），朝向胞质的一层叫内小叶（inner leaflet）。

在考研中，脂质体（liposome）：根据磷脂分子在水相中可自发形成稳定脂双层膜而制备的人工球形双层膜小囊 2022·名解 是极其重要的高频考点。当单层脂分子铺展在水面时，极性端插入水相、非极性尾朝向空气；经剧烈搅动或超声处理后形成乳浊液，即可得到极性端向外的单层微团（micelle）或双层球形的脂质体（liposome）。切勿混淆微团与脂质体：微团是单层实心脂分子团（内部无水腔），而脂质体是双层封闭空心小囊（内部有水相内腔）。

由于脂双层内部是厚约 3～4 nm 的非极性碳氢核心，Lamason 教授在黑板上写下了一条决定全书后续章节的铁律：磷脂双分子层是一道卓越的屏障，能有效阻止亲水性物质（hydrophilic molecules）的自由扩散，包括蛋白质（proteins）、糖类（sugars）和带电离子（ions）。细胞要想摄取营养、排出废物或传递信号，就必须在脂双层中镶嵌专门的蛋白质机器——而在人类认识“脂质与蛋白质究竟如何拼成生物膜”的道路上，足足走过了近一个世纪的五步科学接力。

三、生物膜结构模型的五步演进史：每个模型都是被一个关键实验逼出来的

小柏学长在串讲专题 3 时反复叮嘱：生物膜结构发展的 5 个模型是简答与论述题的基石，绝不能只背名字，必须顺着“前一个模型解释了什么 → 哪一个新实验现象打脸了它 → 从而催生出下一个模型”的证据链来记！

1. 双层脂分子模型（Lipid Bilayer Model，1925 年）

1925 年，两位荷兰科学家 Gorter 和 Grendel 设计了一个极其精巧的定量实验：他们选定没有细胞核和细胞器膜干扰的人成熟红细胞，先统计红细胞数量并计算其总表面积，随后用丙酮等有机溶剂将红细胞质膜的膜脂全部抽提出来，滴加在空气-水界面上使其铺展成紧密排列的单分子层（monolayer）。测量结果令人震惊：抽提物在水面铺展成的单分子层面积，恰好是这些红细胞表面积的 2 倍！由此他们首次提出：细胞质膜是由双层脂分子构成的，极性头朝向内外水相，疏水尾相对排列，为后续所有生物膜模型奠定了脂双层核心基础。

3D流动镶嵌模型点击打开 3d.lkhan.ai 交互式三维结构模型↗

2. 三明治模型 / 夹板模型（Sandwich Model，1935 年）

双层脂分子模型诞生后不久，Danielli 和 Davson 在测量细胞质膜的物理性质时发现了一个反常现象：天然生物膜的表面张力远低于纯油-水界面的表面张力！如果人为在油滴表面吸附一层蛋白质，油滴的表面张力就会显著下降到与生物膜相近的水平。基于这一实验证据，他们于 1935 年提出了影响学界长达 20 余年的“三明治模型（蛋白质-脂质-蛋白质）”：认为质膜中间是连续的脂双层，内外两侧表面各平铺着一层呈伸展 β 折叠状态的球形蛋白质单层，并认为膜上存在由蛋白质围成的亲水微孔供水溶性物质通过。这一模型的重大贡献是首次把蛋白质引入了膜结构。

3D流动镶嵌模型点击打开 3d.lkhan.ai 交互式三维结构模型↗

3. 单位膜模型（Unit Membrane Model，1959 年）

20 世纪 50 年代透射电子显微镜（TEM）问世。J. D. Robertson（罗伯特森） 用四氧化锇（锇酸，OsO₄）固定细胞超薄切片后在电镜下观察，发现无论是细胞质膜还是内质网、线粒体等细胞内膜，切面全都呈现清晰的“暗-亮-暗（dark-light-dark）”三条带结构，总厚度约 7.5 nm（两侧暗带各约 2.0 nm，中间亮带约 3.5 nm）。Robertson 继承并发展了三明治模型，推测高电子密度的两侧暗带是蛋白质和磷脂极性头，中间低电子密度的亮带是脂双层的疏水脂肪酸链，并提出：细胞所有的生物膜都由这种统一的“蛋白质-脂质-蛋白质”单位膜（unit membrane）构成（该结论当时还获得了 X 射线衍射分析的支持）。

3D流动镶嵌模型点击打开 3d.lkhan.ai 交互式三维结构模型↗

然而，到了 20 世纪 60 年代末，两个致命的实验反证彻底动摇了三明治模型和单位膜模型的根基：

• 反证一（冰冻蚀刻电镜）：把快速冷冻的生物膜从脂双层中间疏水核心劈开（冷冻蚀刻技术，freeze-fracture），电镜下清晰看到脂双层内部镶嵌着大量直径 5～8 nm 的膜蛋白颗粒，说明蛋白质根本不是只平铺在膜表面，而是深深插入甚至贯穿了脂双层！
• 反证二（细胞融合与免疫荧光）：1970 年 Frye 和 Edidin 的人-鼠细胞融合实验证明，膜蛋白并不是静态固定在膜上的，而是可以在膜平面内自由侧向扩散的！

4. 流动镶嵌模型（Fluid Mosaic Model，1972 年）

在综合上述实验铁证的基础上，S. J. Singer 和 G. L. Nicolson 于 1972 年在 Science 上发表了里程碑式的流动镶嵌模型（fluid mosaic model）：以液态的磷脂双分子层为生物膜的基本连续主体，球形膜蛋白以不对称的方式或结合于膜表面、或嵌入/横跨脂双层，且膜脂和膜蛋白均可在膜平面内进行侧向运动。

Lamason 教授在课堂最后用一句话精准拆解了它的命名逻辑：“It's fluid because it has to move; it's mosaic because it's heterogeneous（说它流动，是因为分子必须运动；说它镶嵌/马赛克，是因为膜是由多种脂质和蛋白质异质性拼嵌而成的）”。在这个模型下，生物膜可被视为“蛋白质在脂双层分子中的二维溶液（two-dimensional fluid）”（华农判断题高频送分原话，判 √）。

5. 脂筏模型（Lipid Raft Model，1988 年）

随着研究的深入，科学家发现把生物膜单纯看作“膜蛋白在均一脂海中毫无拘束地自由漂浮”过于简化了——膜脂本身并非完全均一，膜蛋白与膜脂、膜蛋白与膜蛋白、膜蛋白与膜下细胞骨架之间存在着广泛的相互作用。1988 年，Simons 和 van Meer 等人在研究上皮细胞质膜分选时提出了脂筏模型（lipid raft model）：在以甘油磷脂为主体的生物膜上，胆固醇和鞘磷脂等富集区域形成相对有序的脂相微区，如同漂浮在脂双层上的“脂筏”一样，载着执行特定生物学功能的膜蛋白 2018·论述。

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 2 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 1 · Membranes 01 · Barriers⏱ 07:48

00:00This course is-- you're going to touch on a lot of different topics in cell biology. And we expect that you have some base knowledge with 7.05 (Biochemistry) and 7.03 (Genetics). And what I think the power of cell biology is, is it takes the things you learned in those courses and puts it in context.

00:17The signaling or the biochemical reactions you've learned about don't really mean anything if you don't think about what's actually happening in the cell at the time of the action. And the cell does a lot of things to really control, and enhance, and regulate these types of reactions.

00:32And understanding how they happen in the cell and the ways these are regulated, I find to be incredibly interesting. And so as a trainee, I've gone through a lot of different fields, looking at zebrafish pigmentation, immunology, and host-pathogen interactions.

00:51And honestly, the thing that has helped me through all of these very different fields is this foundational understanding of how cells do what they do. And so hopefully, no matter what path you guys go down, you will find this class to be very helpful to think about the context and the things that we should be considering for these processes.

01:10And so for the first section that I'm covering here, we're going to focus mostly on this idea of, what is a cell? How do you make the cell? And how does it delineate itself from its environment, setting up this really nice barrier?

01:24And then how does this barrier actually get traversed? The cell cannot have an impenetrable wall. It has to be able to receive information from outside. It has to be able to move things from outside to inside. And it also has to exchange things.

01:38And there's a ton of different ways, very elaborate ways, that eukaryotic cells really do this. And that's-- the majority of things we're going to talk about are focused on eukaryotic cells. So let's get started. For today, we're going to think about this barrier and how this barrier is set up.

01:53And then in the subsequent courses, we'll talk about how this thing is actually bypassed. All right. So we're going to cover three main topics today. We're going to first think about the components. What are the actual constituents of this barrier?

02:11And we'll introduce this idea of fluidity. And then we're going to think about this concept of heterogeneity and how heterogeneity within this-- these membranes, actually, is really important for function. And then lastly, we'll also talk about asymmetry, which seems somewhat similar.

02:43But this is really a cool phenomenon that happens in cells, that topology matters. And you'll see what that means in a few minutes. Cells have barriers. And they need barriers. And this allows them to know the difference between, say, the inside and the outside.

03:13And there's also ways in which this allows the cells to also know the difference, say, between-- it's in the context of an organelle-- between the cytosol and the lumen of that organelle. And so one of the main-- the big barrier that we often think about here is the plasma membrane.

03:43Whereas, here we have organelle membranes. And many of the things we're going to talk about regarding plasma membranes can also be thought about in the context of organelle membranes. But there's some interesting differences that we may or may not touch on throughout the class.

04:02So in order to think about this, it's important to consider how a cell could have evolved to have an-- the correct barrier, considering that it's sitting in an aqueous environment. Professor Cheeseman hooked me up with a really nice demo to simply explain how this works.

04:18So does anybody know what's actually in here? Can they tell what's going on? What's-- I think I heard something. Oil and water. Exactly. So we've got blue water here and oil on top. Why-- even though I walked all the way across campus with this, why is this not mixed?

04:39STUDENT: [INAUDIBLE] What was that? STUDENT: Because the oil or water are different consistencies. Scientifically, they're described as immiscible. Perfect. Any other explanations in addition to that? What about oil and water does not allow them to mix?

04:55STUDENT: Because oil is hydrophobic. Water is hydrophilic. Yeah, is that what you were going to say too? So exactly. So if we think about oil in a simple way, we actually have two parts. We've got the hydrophilic part. This is the part that's OK with water.

05:19And then we've got the hydrophobic part. And this is a very simple oil. And so what we're seeing right now is if we've got water, the oil is actually organizing itself with the hydrophilic heads there and all the hydrophobic tails sticking up.

05:41For those that can't see that, I can move that down. All right. What's going to happen when I shake this up? Is it going to mix? You can see it tries. And then what's going to happen? Separate. Yeah, this ability to self-assemble and self-organize is what is the basic property of how our membranes can actually separate themselves from an aqueous environment.

06:08Because what's happening here, you'll see some droplets that are forming if you're close. Or you can do this at home with salad dressing, like oil and water. But what you're actually seeing here is these oil droplets are forming by protecting those hydrophobic tails from the surrounding water.

06:31And eventually, once more and more of them form, and they come closer and closer together, they'll actually fuse together and create, eventually, maybe in a few minutes, a nice layer between these two spaces. All right.

06:44So one of the components that's part of our plasma membrane that creates this fantastic barrier is the phospholipids. And it has-- we'll talk about this in a second. It has many of the same properties of oil, but allow it to create this barrier between an aqueous environment and these hydrophobic regions.

07:11So that's one. There's also-- we're going to talk about cholesterol as well, which is another type of lipid. There's also glycolipids. And of course, there's proteins. And each of these has a very distinct and important role for how the cell behaves.

07:38And I'll touch on some of them today. But many of the things I'm going to teach you today, you can think about them and how they might regulate the cellular processes in the upcoming lectures as well.

🎬 Mod 1 · Membranes 02 · Phospholipids⏱ 05:19

00:00All right, so let's talk a little bit more about phospholipids and how they themselves act like an oil and actually create this barrier. And so for that, I'm going to draw the one and only structure that I will draw in class, and this is the structure of the actual phospholipid.

00:24All right. So at the top of the lipid, we're going to keep it simple right now. It has an alcohol group, and we'll talk about what this little black box is in a few minutes. This is linked to a phosphate. I always have to double check that I don't mess this up.

00:43And then this is connected to a glycerol backbone. And so we've got the phosphate. We've got glycerol. And then this is further connected to fatty acids. And these fatty acids can vary in length. We see something between a C 12 to C 24.

01:26All right, so this is the basic structure. And what is special about this is that it has this polar head group, which is hydrophilic, whereas the fatty acid tails confer the hydrophobic side. So just like the oil I drew up here, it has two parts that allow it to have this differential interaction with aqueous environments.

02:08And so the phospholipids that we see in cellular membranes are actually assembled in a very specific way to create this environment. Because you think about it, the external environment is aqueous, and so is the internal environment.

02:20So this kind of a situation would not work, right? Because you'd still have water on the other side with these hydrophobic tails sticking out. And so what we often see is what's called a phospholipid bilayer. And what this looks like is you actually see the phospholipids self-assemble with their polar head groups sticking out, and their hydrophobic core buried inside, protecting it.

03:00And we often call these two layers leaflets. We have an inner leaflet and an outer leaflet, depending on the orientation you're thinking about. And so there's two of them. OK now, if you were to shake up a cell or break it up, you could also get things like this, which are often called the droplets, or micelles.

03:21M I-- let me write that down. So same situation. Just like with a simple oil, it needs a way to bury those hydrophobic tails and protect it. And these hydrophobic regions are really important for some of the other things we're going to talk about, including how proteins themselves and other things are able to move through this.

03:47OK, so this phospholipid bilayer is fantastic, and it does a really good job of preventing diffusion of hydrophilic things. And that's going to include things like proteins, sugars, and ions. So obviously the cell needs to take these things up, but yet, there is this barrier in place.

04:20And so, even though it's here, whether we're going to learn in ways in which this barrier can be blocked. But just simply thinking about this phospholipid bilayer, it's important to consider what things can and cannot actually cross this, and what properties this confers.

04:34But this again is not a solid structure. This is not something that the cell builds, like a brick wall, and then is not able to mold and change. And so we also know that it has to be fluid. And so there's important reasons for this.

04:50You can imagine as cells are crawling around, they have to be able to move the membrane in response. Temperature can also have a big impact. And as you're going to learn a little bit today, signaling. The ways in which cells communicate with their environment requires that receptors that are found on the surface of the cell can actually come together and amplify the biochemical reaction.

05:13And so you can't have things stuck in this plasma membrane. It has to move.

1.2膜脂异质性、三大膜脂化学本质与膜流动性的调控法则

一、从“全课唯一化学结构式”看甘油磷脂与膜脂异质性

在第二讲与第五讲视频中，Lamason 教授拿起粉笔在黑板上画下了她在整门课中亲手画的唯一一个化学结构式——甘油磷脂（glycerophospholipid）。她一边画一边提醒大家：“你不需要死记硬背每个原子的化学式，但必须看懂它的四层积木结构如何决定膜的物理行为！”

从上往下看，一个标准的甘油磷脂分子由四个模块串联而成：

1. 极性醇基团（Alcohol group）：最顶端的“小黑盒”，如胆碱（choline）、乙醇胺（ethanolamine）、丝氨酸（serine）或肌醇（inositol）；
2. 磷酸基团（Phosphate group）：自带负电荷，与顶端的醇共同构成亲水的极性头部（polar head group）；
3. 甘油骨架（Glycerol backbone）：三碳骨架，前两个碳连脂肪酸，第三个碳连磷酸；
4. 两条脂肪酸尾部（Fatty acid tails）：赋予疏水性。Lamason 教授指出天然膜脂肪酸链长在 C₁₂～C₂₄ 之间；而在国内考研中，务必按翟中和教材背准：甘油磷脂的脂肪酸链碳原子数皆为偶数，大多数由 16 或 18 个碳原子组成（少量为 14 或 20 个碳原子），且常常含有一条饱和脂肪酸链和一条含 1～2 个顺式双键（cis-double bond）的不饱和脂肪酸链。

紧接着，Lamason 教授向全班提出了一个核心问题：“既然只要有极性头和疏水尾就能拼成脂双层，为什么细胞不干脆只用同一种磷脂来造膜，而要费尽心思维持膜脂的异质性（heterogeneity）？”

在教授与学生的精彩问答中，黑板上归纳出了膜脂必须保持异质性的四大生物学原因：

1. 构建信号激活微区（Activation domains / Signaling）：细胞不能用均质的膜做信号转导，必须靠特定脂质（如 PI、胆固醇+鞘磷脂）聚集成局部“生物化学枢纽（biochemical hubs）”，招募并浓缩下游激酶；
2. 赋予局部电荷（Charge）：不同醇头部的电荷截然不同（如磷脂酰丝氨酸 PS 带净负电），直接决定膜的表面电势与带正电蛋白的静电招募；
3. 塑造膜的曲率与几何形状（Shape and Curvature）：细胞不是呆板的方盒子，内吞、出芽、分裂都需要膜发生剧烈弯曲——圆柱形的 PC（卵磷脂）头尾等宽，适合紧密平铺成平坦长膜；而极性头极小的 PE（脑磷脂）呈圆锥形，并排聚集时会自发迫使膜发生弯曲（curvature），且堆积较为松散！
4. 调控与缓冲膜的流动性（Fluidity）：不同头部与尾部（饱和 vs 不饱和、胆固醇含量）直接改写膜的紧密堆积程度与熔点。

更妙的是，这种异质性不仅存在于同一张膜内，还体现在不同细胞器膜之间的脂质梯度上！Lamason 教授举例说：在细胞质膜中，小头圆锥形的 PE 仅占膜脂的 7% 左右；而在管网密布、频繁出芽囊泡的内质网（ER）膜中，PE 的比例高达 17%！翟中和教材同样给出了一个极具代表性的对比数字：高尔基体膜中鞘磷脂（SM）的含量是内质网膜的 6 倍左右。

二、生物膜三大类膜脂全解：甘油磷脂、鞘脂与固醇

在华农考研填空题与选择题中，膜脂的分类、代表分子与合成场所是必考送分题。生物膜的膜脂主要包括甘油磷脂（glycerophospholipids）、鞘脂（sphingolipids）和固醇（sterols，动物主要为胆固醇 cholesterol）三大类，它们全都是两亲性（两性）分子 2007/2016·填空。

1. 甘油磷脂（Glycerophospholipids）——含量之王，内质网合成

甘油磷脂是以 3-磷酸甘油（glycerol 3-phosphate） 为骨架的衍生物，占整个膜脂总量的 50% 以上，是生物膜中含量最丰富的膜脂，其主要合成场所是内质网（ER） 2016·填空 2022·选择。根据极性头部所连醇基的不同，主要分为四类（以及特殊的心磷脂）：

• 磷脂酰胆碱（Phosphatidylcholine, PC，俗称卵磷脂）：甘油磷脂中含量最高的一种，分子呈标准圆柱形，多分布于质膜外小叶。
• 磷脂酰乙醇胺（Phosphatidylethanolamine, PE，俗称脑磷脂）：极性头小，呈圆锥形，赋予膜曲率，多分布于质膜内小叶。切记俗名对应：卵磷脂＝PC（胆碱），脑磷脂＝PE（乙醇胺），考场千万别张冠李戴！
• 磷脂酰丝氨酸（Phosphatidylserine, PS）：四种主要膜磷脂中唯一带净负电荷的一种（PC、PE、SM 在生理 pH 下净电荷均为 0），正常情况下被严格限制在质膜内小叶（胞质侧），参与维持膜内负电势及细胞凋亡信号识别。
• 磷脂酰肌醇（Phosphatidylinositol, PI）：含量较少，集中于质膜内小叶（胞质侧）。PI 在细胞信号转导中具有极其重要的作用 2017·选择：其肌醇环被脂激酶磷酸化生成 PIP₂（如 PI(4,5)P₂），一方面可直接作为胞质信号蛋白的招募停靠位点，另一方面在受体的激活下被磷脂酶 C（PLC）切割生成两个第二信使——留守膜上的 DAG（二酰甘油，需结合内侧 PS 才能激活蛋白激酶 C 即 PKC） 和释放入胞质的 IP₃（肌醇三磷酸，动员内质网钙库释放 Ca²⁺）。
• 心磷脂（Cardiolipin，二磷脂酰甘油）：唯一的特殊例外——绝大多数甘油磷脂只有 2 条非极性脂肪酸尾，而心磷脂是两个磷脂酸通过共用一个甘油头部相连而成的“双倍磷脂”，拥有 4 条非极性脂肪酸尾部，富含于线粒体内膜和细菌质膜中（注意：是 4 条非极性脂肪酸尾，绝不是“4 条极性尾”）。

2. 鞘脂（Sphingolipids）——以鞘氨醇为骨架，高尔基体合成

鞘脂不是甘油衍生物，而是以鞘氨醇（sphingosine，一种含一条长疏水烃链的氨基醇）为骨架，其氨基再通过酰胺键共价结合一条长链脂肪酸，从而同样拥有两条非极性疏水尾。国内考研标准口径下，鞘脂的合成场所为高尔基体（Golgi apparatus）。

鞘脂根据极性头部的不同分为两支：

• 鞘磷脂（Sphingomyelin, SM）：极性头部为磷酸胆碱或磷酸乙醇胺，结构与甘油磷脂相似，其中含量最高的是神经鞘磷脂。鞘磷脂与甘油磷脂共同构成生物膜主体；由于其脂肪酸链较长且饱和度高，形成的脂双层更厚、流动性更低，与胆固醇共同构成脂筏的核心骨架。
• 鞘糖脂（Glycosphingolipids，简称糖脂 glycolipids）：极性头部没有磷酸，而是直接共价结合在鞘氨醇羟基上的一个或多个糖残基（单糖或寡糖链）。糖脂含量一般不足膜脂总量的 5%，但在神经细胞质膜上含量较高（占 5%～10%）。两类必背代表分子：

1. 脑苷脂（Cerebroside）：仅含一个单糖残基（如半乳糖脑苷脂），是动物细胞膜结构中最简单的糖脂（中性糖脂）；
2. 神经节苷脂（Ganglioside）：寡糖链较复杂，且含有一个或多个唾液酸（sialic acid，即 N-乙酰神经氨酸）残基而带净负电荷，在神经元质膜中最丰富（还是霍乱毒素的天然受体）；此外，决定人类 ABO 血型抗原的特异性寡糖链也存在于红细胞膜表面的鞘糖脂（及糖蛋白）上。

3. 固醇 / 胆固醇（Sterols / Cholesterol）——刚性四环两性分子

在动物细胞质膜（及极少数如支原体等无壁原核细胞）中，第三类关键膜脂是胆固醇（cholesterol）（植物细胞含豆固醇、谷固醇，真菌含麦角固醇）。胆固醇的化学结构极其独特：

• 极性头部：仅由一个微小的羟基（-OH）构成，紧靠在磷脂极性头部旁边；
• 刚性平面核心：由 4 个稠合闭环组成的刚性平面甾环结构（planar steroid ring）；
• 非极性尾部：一条简短的碳氢支链。

3D膜的流动性点击打开 3d.lkhan.ai 交互式三维结构模型↗

由于胆固醇的分子刚性极强、极性头太小，胆固醇是刚性很强的两性分子，它自身不能独立形成脂双层，只能插入磷脂分子之间；在动物细胞中，它由细胞质基质和内质网共同合成（体内大部分来自食物摄取），其含量一般不超过膜脂总量的 1/3（注：MBoC 提到某些真核质膜中胆固醇与磷脂摩尔比可达 1:1，考研答题一律按翟中和写“不超过膜脂的 1/3”）。胆固醇不仅能调节膜流动性、增加膜机械稳定性、降低水溶性小分子的通透性，还是合成固醇类激素（性激素、肾上腺皮质激素）、维生素 D 和胆汁酸的前体。

三、膜脂的四种运动方式与翻转酶（Flippase）

为什么说生物膜是一片“二维液体”？因为在生理温度下，膜脂分子一刻也不停地进行着四种形式的热运动 2011·简答：

1. 沿膜平面的侧向运动（Lateral diffusion）：侧向运动是膜脂分子最基本的运动方式，我们平时所说的“膜脂流动性”主要就是指侧向运动。它的速度有多快？MBoC 给出了惊人的测算：在同一单分子层内，一个磷脂分子每秒钟与邻近分子交换位置约 10⁷ 次，侧向扩散系数 D ≈ 10^-8 cm²/s，这意味着一个磷脂分子只需 1 秒钟 就能跑完一整个大肠杆菌（约 2 μm）的长度！
2. 绕轴心的自旋运动（Rotation）：脂分子围绕自身垂直于膜面的长轴快速旋转。
3. 脂肪酸尾部的摆动（Flexion）：疏水烃链像海草一样左右摆动，且靠近极性头部处受约束大、摆动较小，越靠近疏水尾端摆动幅度越大。
4. 跨双层的翻转运动（Flip-flop，又称横跨膜运动）：脂分子从脂双层的一侧小叶（leaflet）跨越疏水核心翻转到另一侧小叶。由于翻转时亲水的极性头部必须强行穿越非极性的疏水核心，热力学能垒极高，因此自发翻转运动极少发生（平均每个磷脂分子数小时乃至数天才能自发翻转一次）（注：胆固醇是唯一例外，因为它的极性头只有一个小小的 -OH，可以快速跨双层翻转）。

然而，在内质网膜（ER membrane）中，翻转运动却发生得极其频繁！为什么？因为内质网合成新磷脂的催化酶活性位点全部朝向胞质侧（PS 面）——所有新合成的磷脂一开始都被塞进胞质侧单分子层。如果不能及时翻过去一半，内质网膜就会因为单侧膨胀而崩溃！为此，细胞进化出了专门的膜蛋白——翻转酶（flippase / phospholipid translocator）：一类能够促进单个脂分子在脂双层两小叶之间进行跨膜翻转、从而维持膜双层生长平衡并造就膜脂不对称分布的特殊膜蛋白。MBoC 将这类酶细分为两类：内质网膜上的 Scramblase（混杂酶） 不耗能、无选择性地随机双向翻转磷脂以平衡两层数量；而质膜上的 Flippase（内翻酶，属于 P 型 ATP 酶） 则消耗 ATP，专门把外小叶的 PS 和 PE 定向抓回内小叶，维持质膜的高度不对称性。

四、膜的相变温度与决定膜脂流动性的五大法则

当环境温度发生改变时，脂双层的物理状态会发生戏剧性的转换，这引出了华农 2013 年考过的核心名词解释——细胞膜相变（membrane phase transition）：

细胞膜相变（membrane phase transition）：在不同温度下膜脂物理状态发生改变的过程。生理条件下膜脂多呈有序而又可流动的液晶态（liquid-crystalline state）；当温度下降至某一临界点时转变为刚性冻结的晶态（crystalline / gel state），而升温至一定温度时晶态又可熔解恢复为液晶态，这一发生相变的临界温度称为相变温度（transition temperature, T_m） 2013·名解。

那么，究竟有哪些因素决定了膜脂的相变温度 T_m 和流动性高低？小柏学长总结了一个无敌的底层推导口诀：“所有因素统统看磷脂尾部排得齐不齐、紧不紧——排得越整齐紧密，范德华引力越强，越容易冻结成晶态，相变温度 T_m 越高，膜流动性就越小！” 顺着这条物理逻辑，五大影响因素不背自通 2026·选择：

1. 脂肪酸链的不饱和度（Unsaturation）：饱和脂肪酸链没有双键，像一根笔直的木棍，彼此可以紧密贴合排列；而不饱和脂肪酸含有 1～2 个顺式双键（cis-double bond）。Lamason 教授在黑板上画了一条长出“小歪腿（kink）”的脂肪酸链：顺式双键使脂肪酸烃链产生约 30^° 的刚性折弯（kink），打断了邻近烃链之间的紧密堆积，削弱了范德华相互作用，因此不饱和程度越高，相变温度 T_m 越低，膜脂流动性越大 2026·选择！生活在寒冷环境中的细菌、酵母以及冬眠动物，正是靠上调脂肪酸去饱和酶、合成更多不饱和脂肪酸来降低相变温度，防止细胞膜在低温下“冻僵”。

3D膜的流动性点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. 脂肪酸链的长度（Chain length）：烃链越长，相邻尾部之间接触的表面积越大，疏水与范德华相互作用越强，越难移动。因此，脂肪酸链越短，尾部相互作用越弱，相变温度越低，膜脂流动性就越大。
2. 温度与相变温度（Temperature vs. T_m）：审题时务必分清“同一张膜变温”与“不同膜比相变温度”：对于同一张生物膜，环境温度越高，分子热运动越剧烈，流动性越大；而在相同温度下比较两种不同的生物膜，哪种膜的相变温度（T_m，即凝固点）越低，说明它越不容易冻结，其流动性就越大！
3. 卵磷脂/鞘磷脂比值（膜脂成分比例）：鞘磷脂（SM）的脂肪酸链长且高度饱和，相变温度一般高于甘油磷脂。因此，由鞘脂或卵磷脂（PC）组成的脂双层排列紧密、流动性较小；而由含小头或不饱和链较多的 PE、PI、PS 组成的膜流动性较大 2026·选择。
4. 胆固醇对流动性的双重调节（Cholesterol buffering）：这是 MITx 第六讲最精彩的一幕——为了让学生彻底搞懂胆固醇为什么被称为“膜流动性的缓冲剂（fluidity buffer）”，Lamason 教授当堂邀请了两位助教 Jimmy 和 Arish 走上讲台扮演两分子磷脂，随后又请第三位女助教走上台扮演一分子胆固醇，把双臂插在 Jimmy 和 Arish 之间挽住他们！

• 当室温下降（低于相变温度 T_m，低温条件）：没有胆固醇时，Jimmy 和 Arish 越走越慢，最后紧紧贴在一起冻成了晶态（流动性骤降）；但有了扮演胆固醇的助教插在两人中间后，当他们试图靠拢结冰时，学生们齐声喊道：“Cholesterol won't let 'em! Basically gonna pull them apart!”——在低温（相变温度以下）时，胆固醇楔入磷脂分子之间，像隔板一样阻止脂肪酸烃链彼此靠拢结晶，从而提高膜的流动性，防止低温下膜流动性突然降低。
• 当室温升高（高于相变温度 T_m，高温/生理条件）：没有胆固醇时，Jimmy 和 Arish 在台上到处乱跑，间距变得极大；而当胆固醇助教用双臂挽住他们时，两人想跑也跑不远了！——在高温（相变温度以上）时，胆固醇庞大的刚性平面甾环与磷脂靠近头部的烃链段相互作用，限制了脂肪酸链的自由摆动与侧向乱跑，从而增强膜的有序性、降低膜的流动性。

3D膜的流动性点击打开 3d.lkhan.ai 交互式三维结构模型↗

五、脂筏（Lipid Raft）：液态海洋上的“有序信号平台”

既然鞘磷脂拥有长而直的饱和疏水尾，而胆固醇拥有刚性平面的四环甾体，当这两种脂质在甘油磷脂的海洋里相遇时，会发生什么？它们会通过紧密的疏水键和范德华引力自发抱团，在液态无序（liquid-disordered）的膜上分离出一块块液态有序（liquid-ordered）的微结构域——这就是华农 2014 年名解、2018 年论述大题直接命中的脂筏（lipid raft）！

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 2 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 1 · Membranes 05 · Heterogeneity⏱ 08:51

00:00So fluidity is not the only thing we care about in terms of this membrane. And I mentioned that in addition to the components and the fluidity, we're going to talk about two other key features. And that first key feature is a heterogeneity.

00:30OK. So we're going to cover three basic places where we can see heterogeneity in this context. And the first one is in just thinking about the types of membranes and lipids is the actual alcohols. So I wanted to get back to that concept now can you think of a reason why we would need things to be heterogeneous?

00:52I think signaling I gave you. So activation domains is one reason. STUDENT: [INAUDIBLE] Yep. So that would sort of be still in the-- I'm thinking about it in terms of the activation domains. But in responding to different signals, you might see that in any second.

01:11Any other things that you can think of? I keep drawing squares cells. That's probably not-- or circle cells. They have different shapes. Right? You were going to say something too. STUDENT: I was going to say something [INAUDIBLE]..

01:29You're already minutes ahead of me. Perfect. Yes. Yeah, we're going to talk about that. This idea of charge. So charge is one reason why this heterogeneity is important. Activation domains, like I said, you can't have a homogeneous lipid bilayer.

01:53Shape is another thing. And of course, it all comes back to fluidity. The things that I'm going to mention here and in the next few minutes also impact the fluidity, and the types of experimental results you might get in this assay.

02:09OK. So the first one is the alcohol groups. And I can show you a couple of different slides. So we're getting back to this structure of the phospholipids where you see the fatty acid tails, the glycerol backbone, the phosphate.

02:27And then a variety of different alcohols that are available to them. And you don't have to memorize the structures. But you'll see the features that I'm going to talk about that make sense. And one of the features you can see is size.

02:39There's very different sizes of these head groups. And so for example, the one on your far right is phosphatidylcholine. Also goes by PC. And this has a very special feature. It doesn't confer it here in this sort of stick diagram.

02:57But when we think about PC, it looks a lot like all the phospholipid pictures I've been showing you where it looks kind of like a cylinder. Right? So the head and the tails are roughly the same size. And because of this, this lipid can do a very good job-- whoops, already drawing it backwards-- of tightly packing together.

03:26And this tight packing can certainly confer, and we'll talk about a little bit, changes in shape and fluidity. Those would certainly be important. So you might get these really long lipid bilayers. This is in contrast to phosphatidylethanolamine, or PE, which is on your far left now.

03:44This guy has this little tiny head, but still has those long tails. And as a consequence looks more like a cone shape. If I put a whole row of these together, what do you think will happen? PC looks like this. What's PE going to look like?

04:02Sort of. What was that? STUDENT: It's more spaced out. More spaced out. Well, the problem is if we do that-- let's go through that exercise. What's the problem with that guys? Space. The aqueous environment can contact the hydrophobic tails.

04:17And the whole job of this thing is to protect those tails from the aqueous environment. And with these teeny tiny heads, it's not going to work. Yeah, exactly. So you'll start to see that these little heads come together.

04:33And if my drawings were better, it'd be more elaborate. But you can actually see that they create this curvature. And so cells actually mix the different types of phospholipids to contribute to different shapes and sizes here or curvature.

04:51So that's a couple of different combinations that deal with fluidity. One more thing. Because this is so good at being nice and straight, it can pack really tightly like I said. Not only do these curve, they tend to be a little bit looser because again it's harder to smush them together.

05:09Now getting back to your response about the lipid that gets flipped out for apoptosis, that one is phosphatidylserine, which is the one in the middle. And it is notable because it has a negative charge associated with the head group that faces this cytosolic environment.

05:26It's similar to PE where it has sort of a conical or yeah, a cone shape. But it has a negative charge, which makes this special. All right. So one of the other special things is that because of this negative charge, it participates in a couple of different things like signaling.

05:45And it's often kept in the inner leaflet, which refers to the space that cytosolic. And the cell does a lot of work to keep this particular lipid in. And we'll talk about this when we get to the cell death portion about why this is important, and why this actually becomes a signal to neighboring cells that the cell is in trouble because it can't control the incorporation of phosphatidylserine on the inner leaflet.

06:15All right. So one of the other things to keep in mind too is that this heterogeneity doesn't just exist within the single membrane you're thinking about. It also exists within the different components of the cell. So for example, phosphatidylethanolamine.

06:31We see maybe 7% of the lipids in the plasma membrane have phosphatidylethanolamine. Whereas something like the ER has something like 17%. And you can see this throughout different types of lipids. You see different incorporations.

06:50So there's sort of a gradient sometimes depending on the type of lipid you're thinking about. All right. So there's one example of how the alcohol group is a really major contributor to the different ways that membrane can act.

07:05Another way we can think about this is the actual tails themselves. The fatty acid tails. And they come in two basic flavors. They can be saturated or unsaturated. Just as a reminder, saturated means that this tale looks like this.

07:30It doesn't have any double bonds. And so it makes a nice straight stick almost. Whereas unsaturated kind of looks like this. Assuming I drew this right. But we have often this cis double bond here. And because of this, you actually get lipids that have a slightly harder time packing again.

08:03So versus that where you get nice packing, nice straight lipids, you might get something where it has a little kink-- oops, draw this the other way-- where they can't pack as tightly because there's like a little leg sticking out almost in some ways.

08:24So these tend to be looser by the incorporation of unsaturated fatty acids. And we see them a lot in the fatty acid tails that are in our plasma membranes. And they sometimes are thinner because of the fact that it's not a nice stick straight plasma membrane.

08:41And so this is really important too because you don't want to have your cells freeze. You want to be able to control that and buffer against different types of temperatures.

🎬 Mod 1 · Membranes 06 · Other Lipids⏱ 06:38

00:00So, another set that we're gonna talk about before we take a break, are the other lipids. And one of the big ones that I think is really important to consider when we think about fluidity and activation domains, is cholesterol.

00:19So cholesterol looks like this. But what's notable is that it has a polar head group and a hydrophobic section, which means just like phospholipids, it's gonna orient itself in the membrane with its polar head here, and it's hydrophobic tail inside.

00:40All right. So one of the things that we can do is ask, how would the insertion of something that looks like this actually impact the fluidity of our membrane. And so we're gonna do an in-class experiment using TAs to sort of get through this idea.

00:57Because I find the idea of how cholesterol contributes to the fluidity of the membrane can be something that can stick for people. And so I need two TAs first that will volunteer to be my phospholipids. And we're all gonna do an experiment with them.

01:11All right guys. OK. So first, the experimental details are going to be that Jimmy and Arish are our phospholipids. And we're gonna do something where we actually change the temperature in the room so that we can see what happens to them in the plasma membrane.

01:32And we're gonna be measuring fluidity just as a concept. All right. Let's say room temperature, maybe it's the melting temperature. So there's something happening. What would you guys be doing? Moving around. You guys. Yeah.

01:48Nice, very good. Very good. All right. Guys, what's gonna happen if we drop the temperature? They're gonna just stop moving? Slowing down. What else, what else? Yes. So what's that gonna do to the fluidity? Is it gonna go down or up?

02:08Down, awesome. So as you get colder and colder and colder. All right. That's too much. All right. So now, that's what happens when it's cold, right. What happens when it gets really warm? So it's getting warmer, right. What do you guys say?

02:29Start running around. OK. Yeah, oh. Right. Exactly. Fluidity increases and you see space, right. This is really important to consider. All right. So I need another TA who's gonna do their best impression of cholesterol.

02:45So this is phospholipids only. All right. Cholesterol, one of your key features is that you do a very good job of interacting with your phospholipids. So you might want to stick your arms through them. OK. Given this, let's say we start here again with the addition of cholesterol, what's gonna happen when these guys try to get-- when we drop the temperature and they try to pack together.

03:15What's that? STUDENT: Cholesterol won't let 'em. Yeah. STUDENT: Basically gonna pull them apart. Yeah. So we'll see. She's maybe not standing straight. She might be like this. So you'll see a difference in the ability to pack, and therefore, you won't see as fluid of a membrane.

03:30So at low temperatures, you'll see something like that. Maybe even a little bit more conservative. Right? Does that make sense so far? So they can't pack as tightly and so cholesterol is really good at buffering against that.

03:41What's gonna happen when they try-- when we increase the temperature? Is she gonna let them go? No, let's see. See, their arms are down. There's less movement. Yeah, perfect. Exactly. So less fluidity. It depends on the temperature.

04:01You guys are good. Thank you very much. Good job, guys. So in the context of cholesterol, when we add cholesterol to this system, cholesterol is considered to buffer the fluidity of the membrane. And it totally depends on the temperature.

04:18Does that make sense? The last lipid that we're not gonna talk about in great detail, but it's good to know, is a glycolipid. And so unlike our phospholipids, which have this glycerol backbone and a phosphate core, glycolipids actually have a sugar group that's on the surface.

04:51And it turns out that our cells preferentially distribute these glycolipids on the outside. And they serve really important functions for, say cell communication, like in the immune system. Or in tissue development so the cells can connect to each other.

05:09And this brings up a really important point is, how do you get something only on the outside. And so something we're gonna talk about here just very briefly, and then we'll talk about more, is this idea of topology. And it turns out, when we make the glycolipids, the organelles here contribute to this idea of topology such that we put the glycolipid on the luminal side.

05:36And then when the vesicle is delivered from the ER to the surface, it then fuses and flips it out. And so this introduces this idea of asymmetry in topology that we're gonna talk about after a five-minute break. Any questions though, before we break?

05:59Sorry, say that again. Body temperature. Ooh. I would imagine-- because so what this point refers to is the melting temperature. I would imagine we're somewhere here. But that's a good question. I don't actually know where in my artificial graph we are in terms of body temperature.

06:24And I think it probably also depends on the thing you're looking at. So like the melting temperature for, say one phospholipid or other kind of thing could be different depending on the context of the lipid, phospholipid, and the cholesterol context.

1.3膜蛋白的分类、分离策略与跨膜移动的实验铁证（FRAP）

一、膜蛋白：生物膜功能的真正执行者与 T 细胞受体聚集实例

如果说脂双层搭起了生物膜的基本骨架，那么生物膜特定的生物学功能则主要由膜蛋白（membrane proteins）决定。在第三讲课堂上，Lamason 教授先抛出了一个生动的对比提问：“在普通的细胞质膜上，大约 50% 的质量是蛋白质。那么线粒体呢？线粒体是干什么的？”

台下学生齐声答道：“Powerhouse of the cell（细胞的动力工厂）！”

教授追问：“它靠氧化磷酸化（oxidative phosphorylation）制造 ATP，那你们猜线粒体膜上的蛋白质会更多还是更少？”学生喊道：“More！”——没错，普通细胞质膜的蛋白质与脂质质量比约为 1:1（约 50% 为蛋白质），而在执行高强度氧化磷酸化的线粒体内膜中，蛋白质含量高达 75%（髓鞘则主要起绝缘作用，蛋白质仅占约 25%）！此外，据基因组疏水性图谱预测，生物体全部基因编码的蛋白质中约有 30% 是跨膜蛋白，而目前临床上 50% 以上的小分子药物都是以膜蛋白（如 GPCR、离子通道、受体激酶）作为作用靶标（受体）的。

为什么这些嵌在膜上的蛋白质也必须像膜脂一样保持流动性（fluidity）？Lamason 教授在课堂上分享了她自己读研究生时的课题——T 细胞与靶细胞的免疫识别（Cell-cell signaling），并在大屏幕上播放了一段活细胞荧光显微录像：

• 假设一个被病毒感染的靶细胞表面展示着抗原信号，而 T 细胞表面携带着能够识别它的 T 细胞受体（TCR）。
• 你在做生化实验时一定深有体会：如果把极其稀薄（dilute）的酶和底物放进试管里，反应根本跑不起来，因为生化反应高度依赖局部浓度！在细胞表面也是如此——如果只有孤零零的一个 T 细胞受体撞上了靶细胞的配体，T 细胞非常聪明，它认为这点微弱信号可能是随机背景噪音，根本不足以启动免疫杀伤。
• 要想真正吹响免疫激活的号角，散落在 T 细胞表面各处的受体必须利用膜的流动性，在接触界面迅速向中心侧向滑行、聚拢（coalesce / cluster），形成高密度的受体簇（即免疫突触），从而成百倍地放大下游激酶级联信号！如果你人为破坏膜的流动性，阻止这些受体聚拢，T 细胞就彻底丧失了完全激活的能力。

二、膜蛋白的三大类型、结合方式与脂锚定机制

根据膜蛋白与脂双层结合的紧密程度、结合方式以及分离难易度，翟中和第 5 版与 MBoC 将膜蛋白严格划分为三大类型：

1. 外在膜蛋白（Peripheral Membrane Proteins，又称周边膜蛋白）

外在膜蛋白（peripheral membrane protein）：水溶性蛋白，不插入脂双层疏水核心，完全位于脂双层外侧，主要通过离子键、氢键等弱非共价键与膜表面的极性脂头部或内在膜蛋白的亲水结构域结合。

• 核心特征与分离：由于它暴露在水相中，其表面氨基酸组成与普通水溶性球蛋白无异；它与膜结合疏松，只需用温和的方法（如改变溶液离子强度/高盐或低盐洗脱、调节极端 pH、加入螯合剂去除 Ca²⁺/Mg²⁺ 或适当提高温度）即可将其从膜上分离下来，且膜本身的脂双层结构完全不被破坏。
• 典型实例：红细胞膜下的血影蛋白（spectrin）、肌动蛋白（actin）、细胞色素 c，以及结合在膜表面磷脂头部以催化降解衰老损伤膜的磷脂酶（phospholipase，也是多种蛇毒的活性溶血成分）。

2. 内在膜蛋白（Integral Membrane Proteins，又称整合膜蛋白）

内在膜蛋白（integral membrane protein）：部分或全部嵌入脂双层疏水核心、多为横跨脂双层的跨膜蛋白（transmembrane proteins），占整个膜蛋白总量的 70%～80%，与膜结合极其紧密，只有用去垢剂（detergent）使脂双层崩解后才能将其分离出来。

• 跨膜蛋白的三段式结构：由胞质外结构域（亲水）、跨膜结构域（疏水）和胞质内结构域（亲水）三部分组成，因此内在膜蛋白本身也是典型的两亲性分子。
• 内在膜蛋白与膜脂结合的三种分子作用力（由主到次）：

1. 疏水相互作用（最主要、最基本的方式）：跨膜结构域的非极性氨基酸残基侧链与脂双层内部的疏水脂肪酸尾部紧密互作；
2. 离子键（静电加固）：跨膜段两端紧贴膜表面的带正电荷氨基酸残基（如精氨酸 Arg、赖氨酸 Lys）与磷脂带负电的极性头部形成离子键，或者带负电的氨基酸残基通过 Ca²⁺、Mg²⁺ 等二价阳离子桥连与磷脂头部结合；
3. 共价结合脂肪酸链（额外锚定）：某些跨膜蛋白位于胞质侧的半胱氨酸（Cys）残基还会共价连接一条脂肪酸链（如棕榈酸），插入脂双层内小叶以进一步加固结合。

3. 脂锚定膜蛋白（Lipid-Anchored Membrane Proteins）

这是华农 2021 年直接考出的全英文名词解释真题 2021·名解 ⚑华农考过！脂锚定膜蛋白（lipid-anchored membrane protein）：通过共价相连的脂分子（脂肪酸链、异戊二烯烃链或 GPI 糖脂）插入脂双层疏水核心而锚定在膜上的膜蛋白，其水溶性的蛋白质本体部分完全位于脂双层之外。

3D膜蛋白的三种类型点击打开 3d.lkhan.ai 交互式三维结构模型↗

根据共价连接的脂锚化学性质及所在膜面（胞质侧 vs 胞外侧），脂锚定蛋白精确分为三种方式（务必背熟前两种在胞质侧、第三种在胞外侧！）：

1. 脂肪酸（肉豆蔻酸）结合于 N 端甘氨酸（胞质侧 PS 面）：一条 14 碳饱和脂肪酸（肉豆蔻酸，myristic acid）通过酰胺键共价结合到蛋白质 N 端的甘氨酸（Gly） 残基上，插入质膜内小叶。典型代表：肿瘤相关的酪氨酸激酶突变体 v-Src（以及 Src 家族激酶），定位于质膜胞质侧。
2. 异戊二烯烃链结合于 C 端半胱氨酸（胞质侧 PS 面）：一条 15 碳（法尼基，farnesyl）或 20 碳（香叶基香叶基，geranylgeranyl）的疏水烃链通过硫醚键共价连接在蛋白质 C 端附近的半胱氨酸（Cys） 残基上（部分蛋白还会通过可逆的棕榈酰化加上第二条饱和脂肪酸链，实现双重牢靠锚定与动态定位调控）。典型代表：参与细胞增殖信号转导的小 G 蛋白 Ras、以及介导膜泡定向融合的 Rab 蛋白，均定位于质膜或细胞器膜的胞质侧。
3. GPI 糖脂锚定（质膜外表面 ES 面）：蛋白质在内质网腔中完成修饰，其 C 端通过一段寡糖链共价连接到含有磷脂酰肌醇（PI）的糖脂分子上，简称 GPI 锚定蛋白（glycosylphosphatidylinositol-anchored protein）。与前两种位于胞质侧的脂锚截然相反，GPI 锚定蛋白 100% 分布于细胞质膜的外表面（ES 面），且偏好富集在脂筏微区中！可用特异性磷脂酶（如磷脂酰肌醇特异性磷脂酶 C，PI-PLC）切开脂锚将其释放到胞外溶液中。

三、去垢剂（Detergent）与临界胶束浓度（CMC）：拆解生物膜的分子手术刀

为什么洗盐水洗不下内在膜蛋白，非得用去垢剂（detergent，又称表面活性剂）不可？在第十讲和 MBoC 第 10 章中，去垢剂的作用机制被剖析得淋漓尽致：

去垢剂（detergent）：一端亲水（极性头）、一端疏水（单条非极性烃尾）的两亲性小分子，是分离与研究膜蛋白的常用试剂。与拥有两条疏水尾、呈圆柱形的磷脂不同，去垢剂通常只有一条疏水尾，整体呈圆锥形。当去垢剂浓度极低时，它以单分子形式溶于水；而当水溶液中去垢剂浓度达到某一特定阈值时，其疏水尾部便自发聚拢朝内、极性头朝外，形成球状的微团（胶束，micelles）——这一特征浓度就叫做微团临界浓度（critical micelle concentration, CMC，又称临界胶束浓度）：去垢剂在水中由单分子状态开始聚集形成微团时的临界浓度。

3D去垢剂点击打开 3d.lkhan.ai 交互式三维结构模型↗

当把去垢剂加入生物膜悬液时，去垢剂单体的疏水端会像楔子一样插入脂双层，一方面与膜脂混合形成“脂-去垢剂混合微团”，使脂双层彻底崩解；另一方面，去垢剂的疏水端会取代原本围绕在跨膜蛋白周围的磷脂分子，像救生圈一样包裹住膜蛋白的疏水跨膜结构域，而去垢剂的亲水端则朝向外部水相，从而把原本不溶于水的跨膜蛋白变成可溶性的“膜蛋白-去垢剂复合物”带入水溶液中！

在实验室与考场上，你必须根据实验目的在两类去垢剂之间做出正确选择：

1. 离子型去垢剂（Ionic detergents）——代表：SDS（十二烷基硫酸钠，sodium dodecyl sulfate）

• 结构与性质：极性头部带有强负电荷（硫酸根），作用极其剧烈。
• 对膜与蛋白的影响：不仅能彻底崩解生物膜、溶出最疏水的膜蛋白，而且高浓度 SDS 会钻入蛋白质内部疏水核心，破坏维持蛋白质空间构象的离子键、氢键和疏水相互作用，使蛋白质完全变性失活（并使多肽链按质量比结合大量负电荷）。
• 核心用途：专用于 SDS-聚丙烯酰胺凝胶电泳（SDS-PAGE），按分子量大小分离并分析膜蛋白的组成成分。

1. 非离子型去垢剂（Nonionic detergents）——代表：Triton X-100（曲拉通 X-100，或 β-辛基葡糖苷）

• 结构与性质：极性头部不带电荷，作用温和。
• 对膜与蛋白的影响：能有效瓦解脂双层并包裹跨膜蛋白的疏水跨膜段，但不会破坏蛋白质内部的非共价键与三级结构，因此不会使膜蛋白变性失活！
• 四大核心用途：① 分离纯化具有天然生物活性的膜蛋白（如纯化有活性的 7 次跨膜受体 GPCR 或 Na⁺/K⁺ 泵用于脂质体重建）；② 去除细胞的膜系统以研究细胞骨架与膜骨架（如用 Triton X-100 溶去红细胞膜脂与带 3 蛋白，保留完整的血影蛋白膜骨架）；③ 低温下从质膜分离脂筏；④ 免疫荧光实验中低浓度处理细胞，在质膜上打孔（通透化 permeabilization）以便抗体进入胞内（见 1.5 节 Lamason 教授经典实验）。

四、膜蛋白流动性的热力学本质与两大经典证明实验（人-鼠融合 & FRAP）

既然膜蛋白嵌在液态的脂双层中，它的运动需不需要消耗细胞的 ATP 能量呢？实验给出了干脆利落的否定回答 2022·选择 ⚑华农考过：

• 证据一：向细胞中加入呼吸链抑制剂或解偶联剂（抑制 ATP 合成与能量转换）以及蛋白质合成抑制剂，结果膜蛋白的侧向扩散速率完全不受影响 ⟶ 证明膜蛋白的运动不需要消耗代谢能量（ATP）；
• 证据二：一旦降低环境温度，膜蛋白的扩散速率便显著降低 ⟶ 证明膜蛋白在膜平面内的侧向扩散和旋转是依赖膜脂流动性的自发热运动（布朗运动）！

不过，天然质膜中膜蛋白的实际扩散速率，往往比把同一种蛋白放进纯净人工脂质体双层中慢得多（甚至有一部分蛋白根本不动）。为什么？因为活细胞中存在着四大限制膜蛋白流动的因素：① 膜下细胞骨架（皮层骨架）的锚定与“围栏”限制（若用微丝解聚药物“细胞松弛素 B（cytochalasin B）”处理细胞阻断微丝网络，膜蛋白的侧向流动性会大幅增加！）；② 膜蛋白与特定膜脂（如被困在致密的脂筏中）的相互作用；③ 膜蛋白彼此聚集成巨大的多蛋白复合体；④ 细胞外基质（ECM）或相邻细胞表面粘附分子的栓系与细胞连接（如上皮细胞的紧密连接阻止顶端与基底侧膜蛋白混溶）。

那么，在实验室里，“我们究竟是怎么知道膜是流动的？又该如何定量测出某一个膜蛋白的流动速率和受限程度？”（Lamason 教授原话："So how do we actually know something is fluid or how do we actually test if something is fluid?"）——这就引出了华农 2009、2015、2018、2020 年反复考查的两大传奇实验：人-鼠细胞融合实验与 FRAP（荧光漂白恢复技术）！

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 2 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 1 · Membranes 03 · Protein Properties⏱ 07:00

00:00And the next thing we should talk about then are membrane proteins. OK, draw this lipid. So you have a couple different classes of proteins that interact with this membrane that you'll see different layers of this throughout the course.

00:28And one of those are proteins that can actually break through this barrier. And they can either do this as a full pass, totally bypassing the layer, or sort of partially going through. And these kinds of proteins are called integral proteins.

00:58And that's because these proteins actually insert into the plasma membrane. And we'll talk about how exactly they can do this. Even though I've just set this whole thing up where nothing can go through, there are certain proteins with special features that allows them to actually get into this phospholipid bilayer.

01:14But these aren't the only proteins that interact with the membrane. There's also proteins that need to be dynamically recruited. And so we have proteins that can come, they are called peripheral. And we see that they often either interact with integral membrane protein-- a good example might be a receptor on the surface and it recruits a kinase that can then activate the downstream sets of a signaling pathway.

01:42They can also interact with the lipids themselves. And there's very special lipids that act to recruit proteins to further enhance the signaling efforts. OK? So this is a really nice way to think about it again in this idea of context.

01:57You guys have learned about different biochemical reactions and how those happen. But now, you have to think about, how do they happen in the context of a cell and at a membrane? And there's different properties that actually regulate this through this context here.

02:11All right, so it's important also to think about the different spaces within the cell. And so at the plasma membrane, we often see maybe 50% of the membrane is proteins. What about the mitochondria? What does the mitochondria do?

02:32STUDENT: Powerhouse of the cell. Powerhouse of the cell? Yeah, it does this through this oxidative phosphorylation reaction you guys learned about. So do you think it would have more or less proteins? STUDENT: More. I heard some mores.

02:45Good answer. Yeah, it's actually about 75%. And so there's a lot of big differences, not only if we think about the plasma membrane, but the organelles. And then the context, the things that they actually are supposed to be doing to create a functional cell, are going to change how much of proteins we see or in the context of other things we think about like cholesterol, lipids, that kind of stuff.

03:11And it all depends on the function and how these particular features enhance that function. Now, like I said, this idea of fluidity is also really important. All right, so fluidity-- these proteins can't stay static. They have to be able to move around.

03:32And so we see that both the lipids and the proteins have this ability to rapidly diffuse. And this just makes for a much more responsive cell to signaling to the different environments and engaging with that. And so one of the ways that I like to think about this is in the context of something I studied when I was a graduate student, which is cell-cell signaling.

04:08And in particular, I was studying how T cells, cells that are important for recognizing different infections, can sense that infection and actually respond to this. And so let's say we have a target cell that has a specific receptor.

04:28And we're going to keep it super generic, but we'll talk about this when we get to the biology of the immune system lecture. This is just a target cell receptor that's saying, I'm infected. OK? Here is your T cell and it has its own receptor that allows it to sense that there's an activation.

04:52Now, you guys probably remember or labs that you've done, if you put a really dilute amount of protein-- an enzyme, let's say-- into a buffer and you try to see a reaction, is it going to work? Sometimes these things are concentration dependent, right?

05:06You need to put the right amount into the reaction. Well, this is like a reaction in a tube but you're in a cell. And so if this first engagement happens, you might get some activation. But the T cell's smarter than that.

05:19It doesn't sense this as something that they should care about. This is not enough activation, a single receptor. So there needs to be a way to engage many receptors to fully enhance this. The problem is, these receptors are often throughout the cell.

05:36And so what these cells often do is they bring receptors together. So this activation leads to the recruitment of additional receptors that can then engage with more target cell receptors. And then we see this massive boost in activation.

05:57And this is a really great way for our immune cells to make sure that they don't activate unnecessarily. And so in really nice way to see that is in this slide here. And what they're doing is they're taking a live cell movie-- I'll play it in a second.

06:12This is looking at just the T cell in this sense, and they've labeled the receptors so they can watch them move. And you can see what that looks like here, and they're actually looking at this surface right here. So you can see them all sort of coalescing together.

06:31And this is really critical. If you disrupt this process and you don't let those receptors come together, you don't get full activation of this cell. And so the way that this happens, again, is you need to have things that are fluidly moving around.

06:44And because of this fluidity, we often call this phospholipid bilayer or these plasma membranes as following something called a fluid mosaic model. It's fluid because it has to move, its mosaic because it's heterogeneous.

🎬 Mod 1 · Membranes 04 · FRAP⏱ 06:24

00:00So how do we actually know something is fluid or how do we actually test if something is fluid? And so one of the techniques I'm going to teach you about today is a way to do that. And it's called a FRAP. Switch over here.

00:21All right. So FRAP stands for fluorescence recovery after photobleaching. OK. So in this context, we want to ask, let's say, we have a protein of interest. We want to know how fluid it is. The same experiment could technically be done with lipids if you have a specific way to label your lipids.

01:02But we're going to think about proteins today. OK. So we've got our-- we're going to keep it simple and just draw my lipid bilayer like this. We have this protein here. What's one way that we can actually do this live? What kind of a molecule would we need to see this or what technique would we need?

01:20GFP. Perfect. That's the example I was going to use. You can actually express a DNA construct that has your protein of interest fused to a fluorescent molecule. And then the cells will actually express this fusion in this context.

01:35And assuming this tag doesn't interfere with its function and localization, it'll go to the surface. Now if you do this, and especially if you do this for something that's way more abundant than that picture, you might see something like this.

01:47This is your cell. And the entire surface of the cell is green. You can't track a protein this way, right? There's not enough resolution. You guys talked about resolution. And so typically we're doing this on spinning disk on focals or some other kind of not high-high resolution system.

02:07And so the ways that we deal with this is we can do something called bleaching the spot. And the reason this can happen is because normally when we want to see a GFP molecule, you hit it with a laser. And you excite that molecule and then it emits photons, which is just light, essentially.

02:31It's energy but it's light. And so what we can do is we can-- and this is usually done with a short pulse of a laser because you don't want to destroy your protein. In this experiment, we want to destroy the GFP. And that's what this bleaching refers to.

02:48In this context, you're actually going to take the laser and shine it on a region of the cell. And you're going to hit it with so much light that you basically completely get rid of all the photons. The protein is still there.

03:12The GFP is still technically there. But it's now gone dark because we've burned it out. Does that make sense? OK. And we do this in a subregion because what's nice about this is, once we bleached, we get this spot where there's no signal.

03:30OK. And then we can take different images over time to see what happens in that region. And so you might start to see it start to fill in. And that's because it's fluid. There is an exchange that happens between the fluorescent molecules that are still here and the nonfluorescent ones that have been bleached.

03:51There's sort of this back and forth happening. And so over time, you'll start to see this fill in more and more and more. And then we can actually measure this by seeing what the rate of recovery is to get a sense of what this fluidity looks like.

04:06And so an example of this is you often will see these graphs that are made where they'll look at time. And then here they'll look at fluorescence, say 0% and 100%. And if we're just looking at this spot, that's the spot we care about, we'll see at the beginning a high 100% level of fluorescence.

04:35And then once we bleach it, it drops. Right. And then we watch this exchange rate happening. And depending on the protein, the context, things that we're looking at, you start to see a recovery at a specific rate. And often the diffusion rate is then calculated as the t half or the time it takes to get to 50% recovery.

05:05Any questions there? All right. So here's another way to also see this. And it brings up another additional component of this graph that's really important to pay attention to. So in this particular example, I showed you a perfectly fluid protein.

05:20It has recovered completely to 100%. But in this example, you can see it doesn't get to 100%. And that's because sometimes the proteins themselves are not 100% fluid. There is some reason. And we'll revisit this in a few minutes.

05:36There's often an immobile fraction that you will see. And there's different explanations for why something can be immobile. So it's good to keep that in mind. But you'll notice that the half height of recovery is not from the point at which it started it's the point in which the recovery happens so that recovery might be not 100% And so you have to figure out what 50% of that is.

05:57All right. So this was one of the really key experiments to understand fluidity, and to see that things are moving around, and then to test this. And you can do this in the context of the plasma membrane. You could do this in other different organelles.

06:11And you'll get different results depending on the type of thing you're looking at. And a lot of people want to know this because they want to see how stable is the protein that they're looking at in the complex that they're interested in.

06:21So it has a lot of power.

1.4膜的不对称性、拓扑学法则与红细胞膜骨架系统

一、从“四个膜面（ES / PS / EF / PF）”看生物膜的全面不对称性

在第七讲和第八讲中，Lamason 教授把视角从“同一膜平面内的横向异质性（lateral heterogeneity）”转向了“内外两个小叶之间的纵向差异”——这就是生物膜与流动性并列的第二大基本特征：膜的不对称性（membrane asymmetry） 2012/2015·简答。

膜的不对称性（membrane asymmetry）：生物膜脂双层的内外两个小叶中，膜脂、膜蛋白及糖缀合物（糖脂和糖蛋白）呈不均匀、具有严格方向性分布的特征。为了在超微结构水平精确描述膜的哪一面朝哪儿，细胞生物学家结合冰冻蚀刻技术（freeze-fracture electron microscopy）定义了生物膜的四个标准膜面（必须背熟缩写与全称！）：

3D膜的不对称性点击打开 3d.lkhan.ai 交互式三维结构模型↗3D冷冻蚀刻技术研究方法页的 3D 演示↗

1. 外表面（ES，Extracellular Surface，又称非胞质面）：直接与细胞外环境接触的质膜表面，其脂分子和膜蛋白部分构成外小叶（outer leaflet）；对于细胞内的内质网、高尔基体、分泌囊泡等细胞器膜而言，朝向细胞器腔（lumen）的一面即为 ES 面！
2. 原生质表面（PS，Protoplasmic Surface，又称胞质面）：直接与细胞质基质（cytosol）接触的膜表面，构成内小叶（inner leaflet）。
3. 外小叶断裂面（EF，Extracellular Fracture face） 与 原生质小叶断裂面（PF，Protoplasmic Fracture face）：当把快速冷冻的细胞用冷刀劈开时，最脆弱的地方正是脂双层中间仅靠疏水引力维系的疏水尾部交界处！膜从脂双层疏水核心中央裂开后，暴露出来的两个内部断裂面分别称为外小叶断裂面（EF 面，朝向胞质侧观察外小叶内表面）和原生质小叶断裂面（PF 面，朝向胞外侧观察内小叶内表面）。由于绝大多数跨膜蛋白与膜下细胞骨架相连或胞质侧结构域结合更紧，劈开时膜蛋白颗粒大多留在内小叶上，因此电镜下总是看到 PF 面上的膜蛋白颗粒远多于 EF 面，直观呈现了膜蛋白的不对称分布！

二、膜脂的不对称分布、成因与“PS 外翻吃我信号”

如果用不能穿透质膜的磷脂酶从细胞外部处理完整的人红细胞，再对比破裂血影膜的处理结果，你会发现同一种磷脂在内外两个小叶中的分布极不均匀 2012/2015·简答：

3D膜的不对称性点击打开 3d.lkhan.ai 交互式三维结构模型↗

• 外小叶（ES 面）富集：含胆碱极性头的磷脂酰胆碱（PC，卵磷脂）和鞘磷脂（SM）绝大多数分布在质膜外小叶；而所有的糖脂（glycolipids）则 100% 仅存在于质膜外小叶（及细胞内膜的 ES 面/腔面）！
• 内小叶（PS 面）富集：含末端伯氨基或肌醇环的磷脂酰乙醇胺（PE，脑磷脂）、磷脂酰丝氨酸（PS）和磷脂酰肌醇（PI）则主要分布在质膜内小叶（胞质侧）。
• 两层大致均匀：胆固醇在质膜内外两个小叶中的分布相对较为均匀（因为它的极性羟基头极小，可在两小叶间快速自发翻转平衡）。

为什么细胞宁可耗费宝贵的 ATP 也要把 PS、PE、PI 锁在内小叶，把糖脂和 PC 放在外小叶？Lamason 教授在第七讲黑板上画出了膜脂不对称性的四大生物学意义（与翟中和第 5 版完美契合）：

1. 维持特定功能与调控膜曲度的结构基础：例如糖脂的寡糖链必须朝向胞外才能执行细胞间识别；又如在内吞（endocytosis）和膜泡出芽时，细胞需要在弯曲部位混合小头圆锥形的 PE 与大头磷脂（如大头的磷酸肌醇），靠内外小叶脂分子的几何形状差异创造膜的内曲度（curvature）。
2. 维持胞内负电势并构建胞质侧信号转导枢纽（Biochemical hub）：Lamason 教授指出，活细胞质膜具有内负外正的膜电位（membrane potential），除了离子分布外，四种主要磷脂中唯一带净负电荷的 PS 集中在内小叶，直接造成内外两小叶显著的电荷差；同时，所有磷酸化的 PI（如 PIP₂、PIP₃） 极性头全部朝向细胞质基质一侧，这是 GPCR 和 RTK 介导胞内信号转导以及蛋白激酶 C（PKC 需结合内侧 PS 才能活化）发挥功能的必要前提！
3. 参与凝血生理过程：当血管受损、血小板（platelets）受到刺激活化时，钙离子激活混杂酶（Scramblase），使大量带负电的 PS 迅速由内小叶翻转到外小叶表面，为凝血酶原复合物提供带负电的装配平台，激活凝血级联反应。
4. 指示细胞生死状态——细胞凋亡的“Eat me（吃我）”信号：Lamason 教授用极其生动的拟人化语言描述了细胞凋亡（apoptosis）时的情景：“Something is wrong. I've now flipped all my lipids. I can't control myself. I'm dying. And come eat me!” 正常活细胞靠消耗 ATP 的 Flippase 拼命把 PS 锁在内小叶；一旦细胞进入凋亡程序，Flippase 失活而 Scramblase 激活，导致外小叶 PS 含量显著增加并暴露于细胞表面，成为招募巨噬细胞识别并吞噬凋亡细胞的“吃我（eat me）”信号，因此外小叶 PS 暴露（实验室常用荧光标记的 Annexin V 蛋白特异结合外翻的 PS）是检测早期细胞凋亡的经典指标！

三、膜蛋白与糖缀合物的不对称性及“膜拓扑学等价法则”

如果说膜脂的不对称是“相对含量不均”（除了糖脂绝对在外），那么膜蛋白与糖缀合物的不对称性则是 100% 绝对的！每一种膜蛋白分子在生物膜上都有唯一、明确的方向性（功能取向固定），这种拓扑结构在蛋白质于内质网合成插入膜时即已确定，在整个生命周期直至降解始终保持不变，膜蛋白绝对不能像膜脂那样发生跨双层翻转（flip-flop）（因为膜蛋白庞大的亲水结构域根本无法穿越疏水核心）！

在第八讲中，Lamason 教授画出了单次跨膜蛋白（single-pass transmembrane proteins）的两种基本取向分类，并传授了她的独家记忆秘诀：

• I 型跨膜蛋白（Type I transmembrane protein）：N 端（氨基端）朝向细胞外（或细胞器腔 ES 面），C 端（羧基端）朝向细胞质基质（PS 面）。教授的秒记口诀：“C's in the Cytosol（C 端在 Cytosol 胞质里，就是 Type I）！”
• II 型跨膜蛋白（Type II transmembrane protein）：方向正好相反，N 端朝向细胞质基质（PS 面），C 端朝向细胞外（ES 面）。
• 多次跨膜蛋白（Multipass transmembrane proteins）：如 7 次跨膜的 GPCR 或成孔通道蛋白，多肽链往返穿越脂双层多次，N 端和 C 端可在同侧或异侧。

为什么方向绝对不能搞反？教授在黑板上画了一个单次跨膜受体提问学生：“朝向细胞外的一端通常是什么功能域？朝向细胞内的一端又是什么功能域？”学生回答：“外面是受体配体结合域（ligand-binding domain），里面是催化/激酶结构域（catalytic / kinase domain）或接头结构域！”如果细胞粗心大意把这根蛋白插反了，配体结合域朝向胞质而激酶域露在胞外，这个受体就彻底报废了！

四、红细胞膜骨架（Membrane Skeleton）与血影（Ghost）经典实验

人的成熟红细胞直径约 7～ 8 μm，而人体最细的毛细血管直径只有 2～ 3 μm。在红细胞平均 120 天的寿命中，它要在动静脉之间往返数百万次、累计行程约 480 公里，成千上万次地剧烈挤压变形穿过比自己窄得多的毛细血管，却能瞬间弹回双凹圆盘状而绝不破裂！这惊人的柔韧性与机械强度从何而来？答案就在于紧贴在质膜胞质面下方的弹性纤维网架——膜骨架（membrane skeleton） 2022·简答！

膜骨架（membrane skeleton）：细胞质膜下与膜蛋白相连、由纤维蛋白组成的网架结构，从力学上参与维持质膜的形状与机械强度，并限制膜蛋白运动、协助质膜完成多种生理功能。

为什么科学家研究质膜与膜骨架时，最钟爱选用哺乳动物成熟红细胞作为模式材料？四个原因必背：

1. 无细胞核、无任何细胞内膜系统（无细胞器）：质膜是它唯一的生物膜，抽提时绝对没有核膜、线粒体膜或内质网膜的交叉污染；
2. 具有极高的弹性与机械强度，且这种力学特性完全由膜骨架赋予；
3. 极易纯化制备“血影（erythrocyte ghost）”：将红细胞放入低渗溶液（hypotonic solution）中，水分子渗入使红细胞膨胀破裂（溶血，hemolysis），释放出胞内的血红蛋白（hemoglobin）等可溶性胞质蛋白后，残留下来依然保持原有双凹圆盘形状的空壳，就叫做血影（erythrocyte ghost）：哺乳动物成熟红细胞经低渗处理破裂、释放血红蛋白等胞内可溶蛋白后，残留的细胞质膜及膜下膜骨架结构。

3D膜骨架（红细胞）点击打开 3d.lkhan.ai 交互式三维结构模型↗

对纯化的红细胞血影进行 SDS-PAGE 电泳，按分子量从大到小分离出了多条编号蛋白带，由此彻底查清了红细胞膜骨架的四大核心骨架蛋白（全属周边/外在膜蛋白）与两大跨膜连接靶标（属整合膜蛋白，非骨架蛋白！） 2022·简答：

1. 四大膜骨架蛋白（填空若给 4 个空，写这四个：血影蛋白、锚蛋白、带 4.1 蛋白、肌动蛋白）：

• ① 血影蛋白（Spectrin，又称红膜肽）：血影蛋白（spectrin）：构成红细胞膜骨架网络最主要的纤维蛋白成分（属周边膜蛋白），由 α 链和 β 链相互缠绕组成异二聚体，两个异二聚体再“头头相连”（端对端）形成长约 100 nm 的柔性四聚体。（若遗传缺陷导致血影蛋白缺乏或异常，红细胞会变成易碎的球形，引发遗传性球形红细胞增多症与溶血性贫血！）
• ② 肌动蛋白（Actin）：形成含约 13 个单体的短肌动蛋白纤维（short actin filaments），作为网架的交叉节点，每条短肌动蛋白丝约有 5 个与血影蛋白四聚体游离端（尾端）结合的位点。
• ③ 带 4.1 蛋白（Band 4.1 protein，及内收蛋白 Adducin）：位于连接复合体节点处，一方面显著加强血影蛋白四聚体游离端与短肌动蛋白纤维的结合，交织成二维弹性渔网；另一方面，带 4.1 蛋白直接与跨膜的“血型糖蛋白 A（glycophorin A）”的胞质侧 C 端（或带 3 蛋白）结合，形成将膜骨架锚定到质膜上的第一个锚定点！
• ④ 锚蛋白（Ankyrin）：锚蛋白（ankyrin）：红细胞膜骨架中的关键桥梁连接蛋白，其一个结构域结合血影蛋白四聚体的 β 链，另一个结构域结合跨膜整合蛋白“带 3 蛋白（Band 3）”的胞质面（N 端）位点，从而形成将膜骨架网牢牢固定在质膜上的主锚定点（带 4.2 蛋白在此处起辅助加固作用）。

1. 两大跨膜整合蛋白（千万注意命题陷阱！）：

• 带 3 蛋白（Band 3 protein，阴离子交换体）：带 3 蛋白（band 3 protein）：红细胞质膜上的跨膜整合蛋白（阴离子载体），以二聚体形式存在，每条多肽链以 α 螺旋往返穿越质膜 14 次，介导 Cl^-/HCO₃^- 跨膜交换以运送 CO₂，其伸向胞质侧的 N 端通过锚蛋白为膜骨架提供锚定位点。
• 血型糖蛋白 A（Glycophorin A）：富含唾液酸的单次跨膜 α 螺旋糖蛋白（常以跨膜段交叉的同源二聚体存在），胞外 N 端携有 MN 血型抗原，胞内 C 端通过带 4.1 蛋白连接膜骨架。
• 华农选择题与填空题最阴险的设错点：带 3 蛋白和血型糖蛋白虽然与膜骨架相连，但它们是“膜整合（跨膜）蛋白”，绝对不属于“膜骨架蛋白”，而且它们对维持红细胞的双凹圆盘形态不起决定作用！

我们是怎么用实验证明“血影蛋白和肌动蛋白是维持形状的周边蛋白，而带 3 蛋白和血型糖蛋白是不维持形状的整合蛋白”的？靠的就是以下两个经典的血影拆解实验（小柏学长高频强调）：

五、细胞质膜的七大基本功能与“结构功能观”论述法

在结束本节前，我们用小柏学长的“画圈记忆法”把翟中和教材总结的细胞质膜七大基本功能一网打尽（画一个圈代表质膜：① 画圈即隔绝 → ② 圈上有门做运输 → ③④ 圈上两个位点：外侧识别位点、内侧酶结合位点 → ⑤ 圈与圈相连 → ⑥ 圈自身凹凸特化 → ⑦ 圈坏了生病做药靶）：

1. 为细胞的生命活动提供相对稳定的内环境（界膜与区室化屏障）；
2. 选择性的物质运输（包括代谢底物的输入与代谢产物的排出，常伴随能量传递）；
3. 提供细胞识别位点，并完成细胞内外信息的跨膜传导；
4. 为多种酶提供结合位点，使酶促反应高效而有序地进行；
5. 介导细胞与细胞、细胞与细胞外基质（ECM）之间的连接；
6. 参与形成具有不同功能的细胞表面特化结构（如小肠上皮微绒毛、纤毛、鞭毛、胞饮小窝等）；
7. 与人类疾病及药物研发密切相关（膜蛋白异常与遗传病、恶性肿瘤、自身免疫病及神经退行性疾病相关，50% 以上小分子药物以膜蛋白为受体）。

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 2 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 1 · Membranes 07 · LIpid Asymmetry⏱ 05:19

00:00So we talked a lot about heterogeneity, which I sort of conceptually think about-- I'll take that, Annie-- which I sort of conceptually think about in the context of sort of a lateral distribution of things. But I think it's also important to think about this double, this two leaflet situation and how we can actually have heterogeneity in the context of an asymmetric distribution of things.

00:29So if we draw a cell, we can think about this first in the context of lipid asymmetry. So here's your cell. Maybe the cell is starting to do something like endocytosis. This is your outside, and this is your inside. What we find is that the types of lipids we see on the outside are distinct, often, from the ones we see on the inside.

01:18And so one of those examples we already sort of mentioned, which was phosphatidylserine we often see here because of that negative charge. We also have another example, which I didn't show in detail, phosphatidylinositol, which is also often kept on the inside.

01:40And it's really important in signaling. So this orientation matters. And again, you guys are going to talk about this when Professor Cheeseman comes back and does his lectures. You're going to talk a lot about signaling, and so it's important to think about this context here.

02:06And the way that I like to think about it is lipids, like phosphatidylserine and phosphatidylinositol, create almost like this biochemical hub that allows these reactions to occur in a much more efficient way. Because you can sort of concentrate them by creating little domains, within which certain types of PI modifications or phosphatidylserine are distributed.

02:40We're also going to talk about this in the context of charge, of how this contributes to something called a membrane potential. And so we often will see that our cells are much more negatively charged inside and much more positively charged outside.

03:05And some of this is due to the-- or a lot of it's due to the distribution of ions, but also phosphatidylserine. And the cell then uses this distribution to say to things, like during apoptosis and cell death, "Something is wrong.

03:19I've now flipped all my lipids. I can't control myself. I'm dying. And come eat me." That's pretty much the process. All right, so that's one way this is really important. Another way that the asymmetry is really important is for shape and curvature.

03:40And so for example, maybe in this space we might see an example of maybe more phosphatidylcholine or more even types of lipids that have a more cylindrical shape. So they can create a nice, long, flat, lipid bilayer. But in places like this, you might see-- draw this over here-- a mixture of small-headed and large-headed phospholipids that can create that inner curvature.

04:22Now it's also important to keep in mind is this asymmetry is not set up just at the plasma membrane. It's actually really important in other places of the cell. So for example, if we think about the ER, and the Golgi, and the secretory vesicles that then run through that system, to then send things to the surface of the cell, like the glycolipids or other proteins, it turns out that their orientation almost looks like the opposite.

04:55And it turns out that the lumen of an organelle or a vesicle is actually equal to the outside. And that's because when this thing comes in and it fuses, it needs to have the right asymmetric distribution and the right orientation.

05:12And so we think about this for lipids, or we think about this for proteins. We often refer to this as membrane topology.

🎬 Mod 1 · Membranes 08 · Protein Asymmetry⏱ 05:06

00:00So we don't just have asymmetry in the lipids, but we also have asymmetry that leads to a functional consequence in proteins as well. And so with this kind of a concept, we can start talking about the different types of integral proteins that I introduced.

00:32I sort of drew them very simply with just this blob that goes through, but they actually come in a variety of different flavors. And so if this is our outside, and this is our inside, we have examples where we have what's called a single pass.

00:50This is your phospholipid bilayer, the hydrophobic core. And this represents one of these membrane domains called a transmembrane domain. And in the context of this one, this one has its N-terminal segment facing outward and maybe its C-terminal segment facing inward.

01:20And we often refer to this as a type 1 protein. The way that I remember this one is C's in the cytosol, OK? Should help a little bit. And of course, there's the opposite orientation. We can also get ones that look very similar, but instead, their N-terminus is on the inside, and their C-terminus is on the outside.

01:41And these are called type 2. And these represent single pass-- singular transmembrane domain that we can see here. But of course, there's other examples, too. We can have proteins that pass through the membrane multiple times.

01:57G-protein-coupled receptors are a really good example of this. They have multiple transmembrane domains. That's really important for function. Pore-forming proteins that actually need to create a pore in the membrane often have multiple protein transmembrane domains.

02:11And they-- I've just sort of drawn this with their N and C in this orientation, but they can come in any orientation. There can be N and C on the outside, in the inside, in opposite orientations. It doesn't matter. So this is important because we need to not only think about this position, but we need to think about why this orientation matters.

02:30One more set of proteins, though, that I'll just mention is there is also ways, which we're not going to cover in this class, that proteins-- or at least in this particular lecture-- in which proteins can be anchored to the membrane.

02:41And this, again, confers some kind of function. All right, so just like with lipids, signaling is going to dictate why this orientation matters. So we think about this, again, is a simple example. Here's your cell. Here's a single-pass transmembrane protein.

03:00Here's the outside and the inside. What kind of functional domain do you often see facing the outside of a protein or in this context? STUDENT: Like, a receptor. Yeah. So some kind of receptor, something that can receive a signal, that you can think of this as, like, a ligand-binding domain.

03:31And then what about the internal port? STUDENT: Catalytic. Catalytic, is that what you said? Yeah, catalytic-- kinase domain, something like that, or an adapter domain, some way to interact with the intracellular environment.

03:46And so if you accidentally-- if you're messing this up as a cell, if you put this in the wrong orientation, this protein doesn't function. And so we're going to learn about next week where there's actual sequences within these proteins that help it distribute this asymmetric proteins in the right orientation and how it does this for type 1 and type 2, and why it matters, essentially, because we need to have them in the right orientation to really see what their function is.

04:13And then, of course, we'll touch on this idea of transport. So in order to get this protein in this orientation on the surface, it needs to first be built and expressed. And it actually gets inserted into the ER and the Golgi secretory pathway so it can eventually make its way here.

04:37Now, based on the asymmetric distribution I showed you for the lipids, where is the ligand-binding domain in this context? Is it inside the lumen, or inside the cytosol? STUDENT: Inside the lumen? Exactly. Whereas the kinase domain is still cytosolically exposed.

04:56So this is great, because this means when this thing is put in, it almost looks like it's in the opposite orientation. And when it fuses, it flips it out, all right?

1.5走进 MIT 实验室：未知膜蛋白的生物信息学预测与拓扑学实验鉴定

一、从氨基酸序列预测跨膜蛋白：跨膜 α 螺旋数学题与疏水性图谱

在第九讲一开场，Lamason 教授跟全班同学做了一个非常真实的科研场景互动：“你们当中不少人已经在实验室做本科生科研（UROP）了。假设导师今天交给你一个前人从未研究过的全新蛋白质 X（a novel protein X），在你甚至还没拿到移液器（pipettes）坐上实验台之前，你做的第一件事是什么？”

台下一位勇敢的同学喊道：“BLAST it!（查序列比对！）”

教授大加赞赏：没错！在动手做实验之前，我们先看它的氨基酸序列（primary sequence），就能直接预测它是不是一个跨膜蛋白、跨膜几次、甚至预测它的 N 端和 C 端朝向哪一侧！怎么预测？我们先看整合膜蛋白的跨膜结构域（transmembrane domain, TMD）在三维结构上有哪些构建方式。

翟中和第 5 版与 MBoC 指出，整合膜蛋白的跨膜结构域主要有三种构建形式（注意：如果判断题说“所有整合膜蛋白的跨膜结构域均形成 α 螺旋”，一定要判错（×），因为还有 β 折叠桶！）：

1. 跨膜 α 螺旋（Transmembrane α helix）——最普遍的跨膜方式

• 为什么跨膜段最爱形成 α 螺旋？ 脂双层疏水核心内部完全没有水分子，而多肽链主链上的肽键（C=O 和 N-H）都是强极性基团；如果直接暴露在疏水脂尾中在能量上极其不利，因此多肽链必须卷曲成规则的 α 螺旋，让主链上每一个肽键都在螺旋内部彼此形成饱和的分子内氢键，把极性主链完美藏在中央，同时把疏水氨基酸侧链（如 Leu、Ile、Val、Phe）全部伸向外侧与脂双层疏水核心接触！
• Lamason 教授的黑板数学题：为什么跨膜 α 螺旋恰好需要约 20 个疏水氨基酸？
  教授在黑板上一步步列出了算式：
  脂双层疏水核心厚度 ≈ 3～ 4 nm
  α 螺旋每转一圈（1 turn） = 3.6 个氨基酸残基 ≈ 0.54～ 0.56 nm
  跨越 3～ 4 nm 所需圈数 = (4 nm)/(0.56 nm/turn) ≈ 7～ 8 圈（turns）
  所需连续疏水氨基酸残基数 = (7～ 8)× 3.6 ≈ 20～ 25 个氨基酸残基！
  国内考研答题标准数字：一个跨膜 α 螺旋结构域通常由约 20 个（MBoC 写 20～30 个，MITx 写 20～25 个）连续的疏水氨基酸残基组成，长度约 3 nm，可垂直或倾斜穿过膜面（单次跨膜代表如血型糖蛋白 A，7 次跨膜代表如 G 蛋白偶联受体 GPCR）。
• 如何用疏水性图谱（Hydropathy plot）预测跨膜次数？
  给每个氨基酸赋予一个疏水性指数（hydropathy index，疏水残基为正值，亲水残基为负值），以约 20 个氨基酸为滑动窗口沿序列从 N 端到 C 端计算平均疏水性。只要在疏水性图谱上看到一个跨度约 20 个氨基酸、自由能显著突破正值阈值的尖峰，就对应着 1 个跨膜 α 螺旋；有几个这样的疏水峰，通常就预测该蛋白跨膜几次！

3D膜蛋白的三种类型点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. 跨膜 β 折叠片 / β 桶（β barrel）——孔蛋白（Porin）的专属结构

• 分布与实例：主要见于革兰氏阴性菌（如大肠杆菌）外膜、线粒体外膜和叶绿体外膜上的孔蛋白（porins）（翟中和教材举例：大肠杆菌孔蛋白为三聚体，每个亚基由 16 条反向平行 β 折叠片围成桶状大通道；MBoC 另展示了 8 股至 22 股不等的 β 桶）。
• 为什么疏水性图谱测不出 β 桶跨膜蛋白？ 因为 β 折叠股处于高度伸展状态，一般只需 10～12 个氨基酸残基即可穿过脂双层；而且在 β 桶壁上，氨基酸侧链是“一个朝桶外（接触脂尾，疏水）、紧接着下一个朝桶内（围成充水孔道，亲水）”交替排列的，没有连续 20 个疏水残基，故常规疏水性图谱无法检出 β 桶跨膜段！

1. 特殊两亲性 α 螺旋通道（外非极性、内极性）

• 当跨膜蛋白需要在膜中开出一条只允许特定亲水离子通过的狭窄通道时（如人红细胞膜的带 3 蛋白，14 次跨膜介导 Cl^-/HCO₃^- 运输），它由多个兼具极性与非极性侧链的两亲性 α 螺旋（amphipathic α helices）共同围成桶状：每个螺旋朝向脂双层外侧的一面全是非极性疏水残基（用以固定在膜脂中），而朝向通道内腔的一面则是极性亲水残基（用以引导极性离子跨膜）。

4. 如何从序列预测 N 端和 C 端谁在胞内、谁在胞外？——“正电荷在内法则（Positive-inside rule）”

疏水性图谱只能告诉你蛋白 X 在第几个氨基酸处有一个跨膜 α 螺旋，却没法直接告诉你它是 Type I（N 外 C 内）还是 Type II（N 内 C 外）。Lamason 教授在第九讲末尾给出了第二个生物信息学预测法宝：

• 还记得我们在 1.4 节讲过的“质膜内小叶富含带净负电荷的磷脂酰丝氨酸（PS），导致膜内侧带负电”吗？
• 为了通过静电吸引与内小叶带负电的磷脂头部结合并顺应内负外正的膜电势，跨膜螺旋紧邻的“胞质侧（cytosolic exposed）”尾段往往富含带正电荷的碱性氨基酸残基（赖氨酸 Lys、精氨酸 Arg）！这就是著名的正电荷在内法则（Positive-inside rule）。

二、MIT 课堂压轴实验：表位标签（Epitope Tagging）+ 选择性膜通透测定跨膜蛋白拓扑取向

现在，你通过 BLAST、疏水性图谱和正电荷法则，对未知蛋白 X 提出了一个清晰的科学假说（Hypothesis）：蛋白 X 是一个单次跨膜的 I 型（Type I）跨膜蛋白——它的 N 端暴露在细胞外（Outside），C 端留在细胞质基质内（Inside）。

你的导师看着你的预测图说：“猜得很好，但科学只认实验证据——去实验室用免疫荧光（Immunofluorescence）证明给我看！”在第十讲视频中，Lamason 教授在黑板上完整推演了这套在 MIT 考试和现代细胞生物学论文中出镜率最高的拓扑学鉴定实验（Epitope Tagging + Selective Permeabilization Assay）：

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 2 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 1 · Membranes 09 · Predicting Proteins⏱ 03:42

00:00So why do we care about this orientation? Because it's really useful to think about how a protein can function. And so how many of you guys are working in a lab right now? I figured the answer was pretty high. Have any of you been asked to study a novel protein that nobody knows anything about?

00:19Oh, the brave one in the class, nice. Yeah, you will someday don't worry. So what is-- OK since you did this, what was the first thing you did when you were told about this random protein? STUDENT: We tried to [INAUDIBLE] Oh, wow, you guys went really far.

00:39Well, OK, let's say you don't have a pipettes, before you even technically can sit in the lab and start playing with your protein [INAUDIBLE] what would be the first thing you might do, anyone? Or you. STUDENT: BLAST it?

00:52BLAST it. Yeah, do you guys know what that is? So look at the sequence, right? The sequence can be really informative for thinking about what your protein is doing. And so we can actually look at our sequences and find out if-- not ours but this protein sequence-- and figure out if it's actually a transmembrane protein and maybe its orientation.

01:11All right. So let's predict our favorite protein of interest. How do we do this? One of the first things we do is we need to figure out if it has a transmembrane spanning region. All right. Which groups of amino acids would you look for in this protein sequence to be part of this transmembrane domain?

01:42Hydrophobic residues, because that's the space that it needs to traverse it needs to go through this hydrophobic core. So you often see a higher preponderance of these hydrophobic leucine, isoleucine, and such. All right.

01:55So we've got that down. But we also need to consider that it needs to go through a specific size of this phospholipid bilayer. And so it turns out that this space is about four nanometers in space. And so we need enough hydrophobic residues to make it through this, and this is actually variable.

02:12It really depends on, again, context, lipids, that kind of thing. So we can start to make guesses. In this case, we're going to talk about alpha helix. I forgot to mention that transmembrane domains come in two flavors, alpha helix or beta strand.

02:28For this type example, we're going to think about an alpha helix. And so structurally we know that an alpha helix has one turn. It's about 3.6 amino acids. And this equals about 0.56 nanometers. So if you do the math, to get to four nanometers you need about seven to eight turns and that equals about 20 to 25 amino acids that need to be in this space.

02:59OK. So that's the second prediction, hydrophobic and it has to have a certain length to fit in the right spot. Now that just tells you it has a transmembrane domain. It doesn't tell you which direction it's in. So remember, I told you that we often have a positive charge and a lot more negative charges on the inside.

03:16And so another key feature that can give you a hint at orientation is that we often find positively charged amino acids in the tail segment that's cytosolic exposed. That way it can resist that charge. OK. So you've gotten your protein of interest.

03:32Your professor has asked you to now-- you have a hypothesis. I have protein X. I think it's in this direction. We need some way to actually test it.

🎬 Mod 1 · Membranes 10 · Predicting Proteins Epitopes⏱ 05:14

00:00Some way where we can actually test these using immunofluorescence. And so we need a tool kit. And let's say this is our-- we've predicted our transmembrane domain. We think we know the orientation that's going in. Here's N and C.

00:21So you need an antibody to this. And oftentimes, you could just say, oh, I'll just make an antibody to my protein of interest. That's a pain in the butt. You don't want to do that all the time. So what we often do is we add sequences, small sequences, to our protein of interest.

00:33And these come in the form of epitopes. And these are approximately eight to 13 amino acids. They come in different names, different sequences known as HA, flag, myc. And the reason is, because we've characterized these, we know that they don't typically have a detrimental function on our protein of interest.

00:57Maybe a larger thing like GFP. And we also have in our toolkit, antibodies that are specific to epitopes. All right, so we're going to take this protein, and just like with the GFP, we're going to fuse tags to it in the sequence.

01:28And so we'll do a flag sequence here at the N terminus, and we will add an HA sequence to the C terminus. And our prediction was-- you didn't know this, but our prediction was this-- the N terminus was on the outside, and the C terminus is on the inside.

01:49All right, what would happen if I stuck an anti-HA-- we take our cells and we fix them, so they're not moving around. Everything's stable. What would happen if I dump an anti-HA antibody onto the cell to see-- to try to detect it.

02:02Would it see it? No, why? STUDENT: [INAUDIBLE] Louder. STUDENT: [INAUDIBLE] Good, exactly. The barrier. We talked a lot about this barrier. So the first step you can do is, you can only see the things that are on the outside.

02:15And so what we often do is, we do a stepwise immunofluorescence where we first see the things on the outside, then we treat the cells, and then we look for the things on the inside. So the first step is, we would then add an antiflag antibody, which I'll abbreviate AB.

02:34And this guy we'll stick here. And you can either have this directly conjugated to a dye, or you can use secondary antibodies that recognize this, and then further amplify the signal. All right, so you need some way to break through this barrier.

02:51Because the antibody is a protein, and proteins do not easily-- like, antibodies don't easily pass through the membrane. So you need some way to essentially punch holes in this membrane. And so one of the ways that we can do this is through the addition of detergents, which you can sort of see here.

03:08And detergents, like lipids and like oils, have two parts. A hydrophobic section and a hydrophilic section. And you can see when we mix these into our system, they'll actually punch-- they'll actually intercalate into this space.

03:25And so, in this context, if you were doing a biochemical experiment and you wanted to obliterate your cell and get all of the proteins out of the membrane, you would use a high enough concentration so that that would happen.

03:35But we just use a small amount so that we just make small holes in this space. And one of the detergents we often use in this context is Triton X-100. All right, so now you have space. You've washed your samples, you created the holes.

03:54And the next thing you can add is your anti-HA antibody. And now what you're going to see essentially-- let's say we're looking in the microscope, and we're looking sort of down at the surface of the cell. You'll see a red spot, and then a blue spot right on top of that.

04:19You can't actually see top and bottom in this context, right? Because the resolution is a problem. Now, what if we were wrong? What if we added this, and our protein was in the opposite orientation. What color would we see instead?

04:40Who votes red? Who votes blue? Exactly. So let's draw that real quick. Blue on the outside, red on the inside. If we do it in this order, we try to add the flag, we don't get the red spot because the antibody can't get through.

04:58But when we add the detergent and add the other one, it can technically get through, but it really only cares about the externally exposed ones. And so in this context, we would have a cell that's not red, and just has a blue spot.

05:12Does that make sense?

★本章小结：核心逻辑链

1. 从生化反应到细胞边界：细胞内外及细胞器腔均为水相环境，必须依靠两亲性（两性）脂质分子通过疏水效应自组装成厚约 5 nm 的磷脂双分子层（lipid bilayer），将水溶性大分子与离子阻隔开来，形成细胞质膜与细胞内膜（二者统称生物膜 / 膜相结构）。
2. 分子几何形状决定自组装形态：单尾圆锥形两性分子（如去垢剂）在水中超过临界胶束浓度（CMC）时自组装成单层实心的微团（micelle）；而双尾近圆柱形的磷脂分子则自组装成双层片状结构，并为了消除边缘疏水尾暴露而自发闭合成含水内腔的双层球形囊泡——脂质体（liposome）。
3. 五大结构模型的实验演进链：红细胞单分子层铺展面积 2 倍 ⟶ 双层脂分子模型（1925）；质膜表面张力远低于油水界面 ⟶ 三明治模型（1935）；锇酸固定电镜下呈“暗-亮-暗”三层带 ⟶ 单位膜模型（1959）；冰冻蚀刻见膜内嵌蛋白颗粒＋人-鼠细胞融合证明膜蛋白侧向流动 ⟶ 流动镶嵌模型（1972，主流模型）；胆固醇与鞘磷脂相分离及抗低温非离子去垢剂提取 ⟶ 脂筏模型（1988，对流动性的补充而非否定）。
4. 三大膜脂与流动性调控法则：膜脂包括甘油磷脂（内质网合成，占 50% 以上，含 PC、PE、带负电的 PS、参与信号转导的 PI 及 4 条脂肪酸尾的心磷脂）、鞘脂（高尔基体合成，含神经鞘磷脂与仅分布于外小叶的鞘糖脂：脑苷脂最简单、神经节苷脂含唾液酸带负电）和胆固醇（胞质+内质网合成，刚性四环不超过 1/3）。膜的生理状态为液晶态；链越短、顺式不饱和双键越多、相变温度 T_m 越低、SM/PC 越少，膜流动性越大；胆固醇在高温下限制烃链运动降低流动性，低温下阻止烃链结晶提高流动性。
5. 三类膜蛋白与去垢剂拆解：外在（周边）膜蛋白靠离子键等弱键结合于膜表面，改变离子强度即可温和洗脱；脂锚定膜蛋白靠 N 端肉豆蔻酰化（v-Src，内侧）、C 端异戊二烯化（Ras/Rab，内侧）或 GPI 糖脂锚定（外侧）插入双层，可用磷脂酶切下；内在（整合）膜蛋白占 70%～80%，以跨膜 α 螺旋（约 20 个疏水残基）或 β 桶（10～12 个残基，孔蛋白）横跨疏水核心，必须用去垢剂（保持活性选非离子型 Triton X-100，电泳变性分析选离子型 SDS）崩解膜才能分离。
6. 膜蛋白流动性与 FRAP 双参数：膜蛋白的侧向扩散是不耗能的自发热运动（受温度和膜下细胞骨架限制，细胞松弛素 B 解聚微丝可增加流动性）。通过人-鼠细胞融合（40 min 混匀）与 FRAP（FPR） 可证明并定量：FRAP 曲线的恢复斜率（t_1/2）反映扩散速率 D，平台高度反映可移动比例（mobile fraction）与被骨架/微区固定的不可移动比例（immobile fraction）。
7. 膜的不对称性、拓扑等价与红细胞膜骨架：外小叶（ES 面）富集 PC、SM 和 100% 的糖脂与糖蛋白糖链，内小叶（PS 面）富集 PE、PI 和带负电的 PS（凋亡时 PS 外翻成为“吃我”信号）。细胞器/分泌泡腔面在拓扑上恒等于质膜外表面（ES 面）。红细胞膜骨架由血影蛋白（αβ 二聚体头头相连成四聚体）、肌动蛋白、带 4.1 蛋白、锚蛋白构成，经锚蛋白连带 3 蛋白、经带 4.1 连血型糖蛋白 A 固定于膜上，赋予红细胞双凹圆盘形态与抗变形韧性。利用 FLAG/HA 表位标签 + Triton X-100 选择性打孔免疫荧光 可精准测定未知跨膜蛋白的 N/C 端朝向。

---

★英文名词解释速背卡（2023–2026 全英文名解高频）

---

★易混概念与中英教材口径辨析

---

★华农历年真题章末综合自测

一、名词解释（全英文命题实战）

1. Biomembrane (Membrane structure) 2009·名解
2. Lipid raft 2014·名解
3. Liposome 2022·名解
4. Membrane phase transition 2013·名解
5. Lipid-anchored membrane protein 2021·名解
6. Fluorescence photobleaching recovery (FPR / FRAP) 2015·名解

【参考答案指引】 以上 6 题满分标准答案均已整理于上方 《表 1.3 第一章核心英文名词解释满分背诵卡》 中，请严格按“定义＋结构/原理＋功能/实例”三段式默写核对。

二、客观题精选（选择 / 判断 / 填空）

1. 【2007/2016·填空改编】 生物膜中包含的 3 种主要膜脂类型是＿＿＿＿、＿＿＿＿和＿＿＿＿，它们在化学结构上都是具有极性头部和非极性尾部的＿＿＿＿分子；其中含量最丰富的是＿＿＿＿，它在细胞的＿＿＿＿（填细胞器）中合成，而鞘脂则在＿＿＿＿（填细胞器）中合成。
2. 【2022/2026·选择综合】 下列关于生物膜流动性的叙述，错误的是（　　）
   A. 膜脂分子的运动包括侧向运动、自旋、尾部摆动和翻转运动，其中侧向运动是最基本的运动方式
   B. 膜脂中顺式不饱和脂肪酸含量越高、脂肪酸链越短，则相变温度越低，膜流动性越大
   C. 用呼吸链抑制剂抑制细胞内 ATP 合成，或用细胞松弛素 B 解聚微丝，均会导致膜蛋白侧向流动性显著降低
   D. 鞘磷脂（SM）和卵磷脂（PC）组成的脂双层流动性较小，而胆固醇在低温下可阻止磷脂烃链结晶从而防止流动性骤降
3. 【常考判断辨析】 判断下列说法的正误（正确的打 √，错误的打 × 并说明理由）：

• （1）生物膜可视为蛋白质在脂双层分子中的二维溶液，生理状态下细胞膜属于液态。（　　）
• （2）所有整合膜蛋白的跨膜结构域均由约 20 个疏水氨基酸残基形成 α 螺旋。（　　）
• （3）高尔基体分泌囊泡膜上的糖蛋白，其寡糖链朝向细胞质基质（PS 面），以便与质膜融合后暴露于细胞表面。（　　）
• （4）红细胞膜上的带 3 蛋白（Band 3）是穿越质膜 14 次的膜骨架蛋白，用非离子去垢剂 Triton X-100 处理血影后，带 3 蛋白溶出导致血影失去双凹圆盘形状。（　　）

【客观题答案与解析】

1. 甘油磷脂；鞘脂；固醇；两性（或两亲性）；甘油磷脂；内质网；高尔基体。
2. 选 C。膜蛋白的侧向扩散是不耗能的自发热运动，抑制 ATP 合成对其流动性无影响；而用细胞松弛素 B 阻断微丝网络会解除细胞骨架对膜蛋白的锚定与围栏限制，使膜蛋白侧向流动性大幅增加而非降低！
3. 判断解析：

• （1）前半句 √，后半句 ×（整题判 ×）：“生物膜可视为蛋白质在脂双层中的二维溶液”是教材原话（正确），但生理状态下细胞膜处于液晶态（liquid-crystalline state）而非普通液态。
• （2）×：除跨膜 α 螺旋外，细菌、线粒体和叶绿体外膜的孔蛋白（porins）是以多条 β 折叠片卷成 β 桶（β barrel） 跨越脂双层的（每股仅需 10～12 个氨基酸残基）。
• （3）×：根据膜拓扑学等价法则，分泌囊泡的内腔面（luminal side）才是 ES 面！糖基化发生在内质网和高尔基体腔内，故分泌囊泡上的寡糖链朝向囊泡腔内，当囊泡与质膜融合外翻后，腔面即变成质膜外表面（ES 面）。
• （4）×：带 3 蛋白是膜整合（内在）蛋白而非膜骨架蛋白；用 Triton X-100 处理血影后，带 3 蛋白和血型糖蛋白溶出消失，但由血影蛋白、肌动蛋白等构成的膜骨架依然维持原有的双凹圆盘形状，说明带 3 蛋白对维持红细胞形态不起决定作用。

三、主观大题综合实战（论述与实验设计）

1. 【2018·论述真题】何谓细胞膜结构的脂筏模型？请论述细胞膜脂筏结构特征和在细胞生命活动过程中的作用。 2018·论述

• 【满分答案】

1. 脂筏模型的定义：脂筏模型是 Simons 等人于 1988 年提出的生物膜结构模型，是对流动镶嵌模型中膜流动性的补充与新解读（而非否定）。该模型认为：在以甘油磷脂为主体的生物膜上，胆固醇和鞘磷脂等富集区域形成相对有序的脂相微区，如同漂浮在脂双层海洋上的“脂筏（lipid raft）”一样，载着执行特定生物学功能的膜蛋白。脂筏最初在高尔基体上形成，最终转移到细胞质膜上，部分脂筏还与膜下细胞骨架蛋白交联。
2. 脂筏的结构特征：

• 化学组成与物理性质：富含鞘磷脂（sphingomyelin）和胆固醇（cholesterol）；由于鞘磷脂较长且直的饱和非极性尾部彼此紧密作用，并与胆固醇近平面刚性疏水甾环形成疏水键，使脂筏表现为一种异质性、高度动态、分子排列较紧密、厚度较大、流动性较低的液态有序膜脂微区结构。
• 尺寸与蛋白容量：直径一般为 10～ 200 nm；具有抗低温非离子型去垢剂（如 4^°C Triton X-100）溶解的特性，可用非离子去垢剂从质膜分离；蛋白质组学显示其中富集多达 250 种蛋白质（如 GPI 锚定蛋白、N 端肉豆蔻酰化/棕榈酰化蛋白及特定跨膜受体），推测一个直径 100 nm 的脂筏可装载约 600 个蛋白质分子。

1. 在细胞生命活动过程中的作用：

• ① 细胞信号转导的时空平台：G 蛋白偶联受体（GPCR）、表皮生长因子受体（EGFR）、胰岛素受体、小 G 蛋白 Ras、Src 家族酪氨酸激酶等关键信号分子在脂筏中高度富集并组装成信号复合体，为高效、特异的跨膜信号传递提供时空保障；
• ② 参与物质跨膜运输与蛋白质分选：在非网格蛋白依赖的胞饮作用（如窖蛋白 caveolin 介导的内吞）以及高尔基体向质膜特定区域（如上皮细胞顶面）的蛋白质定向分选与膜泡运输中起重要作用；
• ③ 与人类疾病发生密切相关：多种病毒（如流感病毒、HIV）和细菌毒素利用脂筏作为入侵宿主细胞或出芽释放的门户；此外，阿尔茨海默病（Aβ 肽生成）及部分恶性肿瘤的发生发展均与脂筏结构或功能异常密切相关。

1. 【2009·实验设计真题 + MITx 课堂综合延伸】细胞膜的流动性如何用实验证明？请举例两种方法证明细胞膜的流动性并说明原理；若你进一步克隆了一个全新的单次跨膜受体蛋白 X，如何设计免疫荧光实验鉴定其 N 端和 C 端究竟哪一端朝向细胞外、哪一端朝向细胞质基质？ 2009·实验

• 【满分答案】
• 第一部分：证明细胞膜流动性的两种经典实验方法与原理

1. 荧光抗体免疫标记法（人-鼠细胞融合实验）：

• 原理与步骤：利用不同颜色的荧光抗体分别特异标记不同种属细胞表面的膜蛋白，通过细胞融合观察两种膜蛋白是否随时间在融合膜上发生侧向扩散混匀。具体将抗人膜蛋白抗体标记红色荧光（罗丹明）、抗小鼠膜蛋白抗体标记绿色荧光（荧光素），分别孵育人细胞和小鼠细胞后，用灭活的仙台病毒（或 PEG）介导两种细胞融合形成异核体。
• 结果与结论：融合初期异核体表面红、绿荧光各占一个半球呈分区分布；在 37^°C 下孵育 40 min 后，红、绿两色荧光在融合细胞表面变为完全均匀分布，直接证明膜蛋白可在膜平面内进行侧向移动，生物膜具有流动性。

1. 荧光漂白恢复技术（FPR / FRAP）：

• 原理与步骤：将亲脂或亲水荧光分子（如荧光素或 GFP）与膜脂或膜蛋白偶联标记活细胞；在激光共聚焦显微镜下，用高强度激光脉冲瞬间照射细胞膜某一微小区域，使该区域内的荧光分子发生不可逆淬灭，形成暗色的“光漂白区”；随后用低强度激发光连续追踪观察。
• 结果与结论：周围非漂白区内未淬灭的荧光标记分子因侧向热运动不断迁入漂白区，使漂白区的荧光强度随时间逐渐恢复。这不仅证明了膜具有流动性，还能根据荧光恢复速率（t_1/2）定量计算出膜脂或膜蛋白的侧向扩散系数 D 及可移动比例（mobile fraction）。
• 第二部分：利用双表位标签（Epitope Tagging）与去垢剂选择性通透鉴定单次跨膜蛋白 X 的 N/C 端拓扑取向（MITx 7.06 标准方案）

1. 实验原理：抗体是大分子蛋白质，无法穿过完整的磷脂双分子层，因此在未通透的完整细胞中，胞外加入的抗体只能识别暴露于质膜外表面（ES 面）的表位；而在用低浓度非离子去垢剂（如 Triton X-100）温和打孔（通透化）后，抗体即可进入胞质识别位于质膜内表面（PS 面）的表位。
2. 实验步骤：

• 构建双标签融合基因：利用基因工程在蛋白 X 基因的 N 端编码序列融合 FLAG 标签（8～13 aa 短肽），在 C 端编码序列融合 HA 标签，构建 FLAG-X-HA 真核表达载体并转染体外培养的宿主细胞；
• 平行分组与选择性通透：待融合蛋白表达并定位于质膜后，用 4% 多聚甲醛温和固定细胞，均分为两组：A 组（非通透组）仅加 PBS 缓冲液，不加去垢剂，保持质膜完整；B 组（通透组）加入低浓度非离子型去垢剂 Triton X-100（如 0.1%）短时处理，在质膜上打出小孔；
• 双色免疫荧光染色：向 A、B 两组细胞中同时加入红色荧光标记的 Anti-FLAG 单克隆抗体和蓝色荧光标记的 Anti-HA 单克隆抗体，孵育后充分洗涤，于荧光显微镜下分通道观察。

1. 预期结果与结论：

• 若 A 组（非通透）仅呈现红色荧光（FLAG 阳性、HA 阴性），而 B 组（通透）同时呈现红色与蓝色荧光 ⟶ 证明蛋白 X 的 N 端朝向细胞外（ES 面）、C 端朝向细胞质基质（PS 面），属于 I 型（Type I）跨膜蛋白；
• 若 A 组（非通透）仅呈现蓝色荧光（HA 阳性、FLAG 阴性），而 B 组（通透）同时呈现红色与蓝色荧光 ⟶ 证明蛋白 X 的 C 端朝向细胞外（ES 面）、N 端朝向细胞质基质（PS 面），属于 II 型（Type II）跨膜蛋白。

想象你正俯瞰一座每秒钟都在收发上千万件包裹的超级大都市。在真核细胞里，每一秒都有成千上万条新合成的多肽链从核糖体的出口隧道里钻出来。它们当中有留在细胞质基质里催化的糖酵解酶，有要去细胞核里调控基因转录的转录因子，有要穿过双层膜去线粒体基质发电的呼吸链亚基，还有要被装进囊泡、跨越质膜投送到血液里的胰岛素和抗体。一个最根本的问题摆在我们面前：既然所有核编码蛋白质都起始于同一个地方——细胞质基质中的游离核糖体，它们究竟凭什么不会迷路？

在 MITx 7.06.1x 第二模块的课堂上，Rebecca Lamason 教授在黑板上画下了一张纵横交错的“蛋白质邮政路线图（Protein Roadmap）”。她带着学生一步步追问：细胞如何给每个蛋白质印上专属的“邮政编码（postal code）”？我们怎么通过遗传删除与嵌合移接实验证明一段氨基酸序列既是靶向所必需（necessary）的、又是充分（sufficient）的？当多肽链穿过内质网膜上的狭窄通道时，核糖体共翻译的“推力”与腔内伴侣分子 BiP 的“布朗棘轮拉力”各自扮演什么角色？George Palade 又是如何用短短 3 分钟的放射性亮氨酸脉冲，在电子显微镜下拍出蛋白质从内质网到高尔基体再到分泌泡的动态电影的？

而在国内考研（尤其是华中农业大学 338/849 细胞生物学）的战场上，本章对应的翟中和《细胞生物学》第 5 版专题 5（细胞质基质与内膜系统）与专题 6（蛋白质分选与膜泡运输），更是当之无愧的“全书分值霸主”——从每年必考的全英文名词解释（Microsome、Endosome、Autophagosome、Protein Sorting、Protein Glycosylation、ER Stress、COPII/COPI、SNARE），到动辄 10～15 分的压轴论述与实验设计题（共翻译转运七步、线粒体输入三去向、三种包被膜泡与 Rab/SNARE 定点投送、高尔基体极性维持、内质网应激 UPR 三通路），几乎年年占据卷面四分之一以上的江山。让我们跟着 Lamason 教授的黑板粉笔，把 MIT 的硬核实验逻辑与翟中和教材的满分答题规范融为一炉。

---

2.1细胞区室化、内膜系统与蛋白质分选的“邮政路线图”

课堂一开始，Rebecca Lamason 教授没有直接抛出一堆复杂的细胞器名词，而是向台下抛出了一个看似再基础不过的问题：

"As a point of review, can anybody tell me where you actually do translation? Where do we start making our proteins?"

学生异口同声："Cytoplasm!"

"Cytoplasm, exactly! Transcription is in the nucleus — mRNA comes out. And then that's where we start to do protein synthesis, which is in the cytoplasm... This is super important to remember."

为什么要把“所有蛋白质合成都起始于细胞质基质”当作不可动摇的第一公理？因为只有站在这个起点上，你才能理解为什么真核细胞必须进化出一套精密绝伦的物流分拣系统。

一、真核细胞内部的三类结构与细胞质基质

如果用电子显微镜和生物化学分步离心审视一个真核细胞，细胞内部被膜系统大致分隔为三类结构：细胞质基质、细胞内膜系统、其他由膜包被的细胞器 2009/2017·选择。注意这里的断句是“细胞内部—被—膜系统分隔”，千万不要误读成“被膜系统”！

3D细胞质基质点击打开 3d.lkhan.ai 交互式三维结构模型↗

在这三类结构中，其他由膜包被的细胞器包括 线粒体、叶绿体、过氧化物酶体和细胞核（其中 双层膜包被的细胞器有且仅有三个：线粒体、叶绿体、细胞核，填空题第三空极易漏填细胞核；而核糖体、中心体等则属于非膜相结构 2009/2017·选择）。原核细胞没有任何内膜系统，通常只有核糖体一种细胞器。那么，占据细胞半壁江山的“大本营”——细胞质基质（cytosol）究竟是什么？

细胞质基质（cytosol）：指真核细胞细胞质中，除去可分辨的细胞器以外的胶状物质，约占细胞总体积的 50% 以上。它的主要成分是约占总体积 70% 的水和无机离子，以及以可溶性蛋白质为主的生物大分子。这里有一个极易踩中的判断题陷阱：细胞质基质约占“细胞总体积”的 50% 以上，绝不能写成“细胞质总体积”（部分盗版资料印错，考场判断必考）！细胞质基质绝非一袋混沌的水溶液，而是具有高度有序性的胶状体系：多数水分子以水合物形式结合在大分子表面；细胞质骨架纤维贯穿其中，充当空间结构的组织者，多数蛋白质直接或间接与骨架或生物膜结合，甚至连不同 mRNA 在卵母细胞基质中都呈严格的区域性分布以指导早期胚胎分化。

在考研简答题中，细胞质基质的六大核心功能可浓缩为“代谢、合成、骨架、修饰、寿命、折叠”十二个字：

1. 中间代谢的主要场所：糖酵解、磷酸戊糖途径、糖原的合成与分解都在此完成。
2. 蛋白质与脂肪酸合成：所有蛋白质合成的起始都在细胞质基质的游离核糖体上进行，脂肪酸的生物合成也在细胞质基质中完成（注意与光面内质网合成“膜脂/磷脂”严格区分！）。
3. 细胞骨架组织空间结构：维持基质的高度有序性。
4. 蛋白质修饰调节活性 2020·简答：基质中存在 五种关键蛋白质修饰：① 辅酶或辅基与酶结合（参与氧化还原，能量以质子形式在酶间传递）；② 磷酸化与去磷酸化（蛋白激酶与磷酸酶在 Tyr/Ser/Thr 上催化，快速调控信号级联与转录）；③ 糖基化（哺乳动物基质中主要是将单个 N-乙酰葡糖胺 GlcNAc 加到丝氨酸 Ser 的羟基上）；④ 甲基化（骨架蛋白 N 端甲基化防水解，组蛋白甲基化参与表观遗传）；⑤ 酰基化（如 Src 蛋白 N 端豆蔻酰化助其锚定质膜，微管蛋白乙酰化） 2020·简答。
5. 控制蛋白质的寿命（N 端规则与泛素-蛋白酶体降解）：多肽链翻译起始的甲硫氨酸（Met）被切除后，暴露出的 N 端第一个氨基酸残基是决定该蛋白质寿命的信号（N 端规则）：当 N 端为 Met、Ser、Thr、Ala、Val、Cys、Gly、Pro 这 8 种残基时蛋白质往往稳定，其余残基则往往不稳定，易被识别并送去降解。
6. 帮助变性或错误折叠的蛋白质重新折叠：依靠分子伴侣（尤其是热激蛋白家族）防止多肽聚集并协助其恢复天然构象。

二、细胞内膜系统：单层膜构筑的动态流水线

除了留在细胞质基质工作的蛋白质外，超过一半的新合成蛋白质必须前往各个膜包被的区室。其中最庞大、联系最紧密的一个家族，被统称为细胞内膜系统（endomembrane system）。

细胞内膜系统（endomembrane system）：指在结构、功能乃至发生上相互关联、由单层生物膜包被的细胞器或细胞结构组成的高度动态体系，主要包括内质网、高尔基体、溶酶体、胞内体（内体）和分泌泡等 2013/2019·论述 2021·简答。

为什么华农在 2013、2019、2021 年三次用大题追问“如何理解细胞内膜系统是一个结构与功能密切关联的动态统一体”？我们从 MBoC 的演化与生物物理视角一推就通：

1. 发生的同源性与拓扑连通：在演化上，内质网膜等被认为是由原始真核生物的质膜内陷演化而来（外核膜与糙面内质网直接连续，核周隙与内质网腔相通）。
2. 结构的动态转换（膜流）：内膜系统各成员之间并非孤立静止的孤岛，而是通过运输小泡（transport vesicles）的不断出芽（budding）与融合（fusion）实现膜脂双层与腔内物质的动态转换，并在微管、微丝等细胞骨架的牵引下精准定位 2009·多选。
3. 功能的流水线递进：从内质网（蛋白质与脂质合成、初加工、折叠与质控）→ 高尔基体（糖基化等深加工、分类包装与分选枢纽）→ 溶酶体/内体（细胞内消化、防御与受体回收）或分泌泡（胞吐分泌到胞外或更新质膜），环环相扣、层层递进 2013/2019·论述。

这里必须敲响一个一票否决级的警钟：过氧化物酶体（微体）虽然也是单层膜包被的细胞器，但在国内翟中和教材与考研标准答案中，过氧化物酶体和线粒体、叶绿体一样，均不属于细胞内膜系统！（尽管 MBoC 7e 因其前体泡可从 ER 出芽而将其视为分泌/内吞膜系统的特化分支，但在国内考场上遇到判断题“内膜系统包括过氧化物酶体”，一律判 ×！）

三、蛋白质分选的“两条基本途径”与“四种转运机制”

现在回到 Lamason 教授的黑板。当一个核基因编码的蛋白质在细胞质基质的游离核糖体上启动翻译后，它如何抵达上面这些细胞器？这就是蛋白质分选（protein sorting），又称蛋白质靶向转运（protein targeting）。

蛋白质分选（protein sorting / protein targeting）：又称蛋白质靶向转运，指真核细胞中核基因编码的蛋白质在合成后，依赖自身携带的分选信号（如信号肽、导肽等）及信号识别颗粒（SRP）、受体等因子的介导，定向转运到特定细胞部位以组装结构并发挥功能的过程，主要分为共翻译转运和翻译后转运两大基本途径 2010/2021/2025·名解。

3D共翻译转运：分泌蛋白进 ER 腔点击打开 3d.lkhan.ai 交互式三维结构模型↗

如果你仔细听 Lamason 教授在课堂上的梳理，你会发现她从两个正交的维度来拆解这张路线图：第一个维度是“转运与翻译在时间上的关系”（共翻译 vs 翻译后）；第二个维度是“蛋白质跨过或避开生物膜屏障的物理方式”（门控 vs 跨膜 vs 膜泡）。

1. 按翻译与转运的时机划分：两条基本途径

• 共翻译转运（co-translational translocation）：蛋白质合成起始于细胞质基质的游离核糖体，当新生肽 N 端信号肽暴露后经 SRP 引导转移至糙面内质网（rER）膜上，边合成边跨膜转运入内质网腔或插入内质网膜，随后经膜泡运输至高尔基体加工分选，最终运往溶酶体、质膜或分泌到细胞外的途径 2009·简答 2015·论述 2026·选择。
• 翻译后转运（post-translational translocation）：蛋白质在细胞质基质的游离核糖体上完全合成后释放，再凭自身携带的靶向序列（导肽、核定位信号等）在分子伴侣协助下转运至线粒体、叶绿体、过氧化物酶体、细胞核，或直接留在细胞质基质成为驻留蛋白和细胞骨架蛋白的途径 2009/2012·简答。

游离核糖体与内质网附着核糖体在结构、rRNA 和蛋白质组成上 100% 完全相同，二者属于同一个动态循环库；一个核糖体此刻是游离还是附着，唯一取决于它正在翻译的 mRNA 编码的蛋白质是否携带内质网靶向信号（信号肽）！如果没有任何分选信号，翻译出的蛋白质就默认留在细胞质基质中，成为永久的胞质居民。

2. 按跨越膜屏障的物理机制划分：三种（或四种）转运类型

在 MITx 课堂和 MBoC 第 12 章中，Lamason 教授将蛋白质在区室间的转移归纳为三大物理机制（而翟中和教材在此基础上多列了一条“细胞质基质中的蛋白质转运”，合称四种蛋白质分选基本类型 2009·简答）：

3D线粒体蛋白输入点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. 选择性门控转运（Gated transport）：发生在细胞质基质与细胞核之间。细胞核与细胞质基质在拓扑学上本来就是相通的（核膜孔道连接二者），门控转运（gated transport）：指在细胞质基质中合成的亲核蛋白等大分子，通过核孔复合体（nuclear pore complex, NPC）这一具有高度选择性的双向闸门，在细胞质与细胞核之间主动、双向地完成核输入或核输出的转运方式。折叠好的蛋白质甚至核糖体亚基无需解折叠即可穿过核孔（详见第四章）。
2. 跨膜转运（Transmembrane transport / Protein translocation）：发生在细胞质基质与内质网、线粒体、叶绿体、过氧化物酶体之间。Lamason 教授特别强调："This means the protein actually has to cross the plasma membrane or an organelle barrier in order to reach these different spaces." 跨膜转运（transmembrane transport）：指蛋白质在特定分选信号引导下，通过生物膜上的蛋白质移位装置（如 ER 膜上的 Sec61 移位子、线粒体 TOM/TIM 复合体），真正穿越脂双层屏障进入细胞器腔内或插入膜中的过程。通常要求多肽链处于未折叠的线性状态（依靠共翻译或分子伴侣维持未折叠），但过氧化物酶体除外（可跨膜输入完全折叠甚至寡聚化的蛋白）。
3. 膜泡运输（Vesicular transport）：发生在内质网、高尔基体、溶酶体、胞内体、分泌泡与质膜之间。Lamason 教授一针见血地指出它与前两种机制的根本差异："We don't actually cross the membrane in this case! We've packaged cargo in a membrane-bound compartment, and it just transitions to each of these things back and forth." 在膜泡运输中，可溶性蛋白质和跨膜蛋白质只在糙面内质网合成时跨膜一次，此后由供体膜出芽形成包被小泡并融合到靶膜，蛋白质从头到尾都待在膜泡腔内或膜泡膜上，再也不需要穿过任何脂双层！
4. 细胞质基质中的蛋白质转运（Protein transport in the cytosol）（翟中和第 4 类）：指蛋白质在细胞质基质内部的定向移动，上述三类分选过程均涉及该步骤，且与微管、微丝等细胞骨架系统及分子马达密切相关 2009·简答。

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 1 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 01 · Protein Roadmap⏱ 04:08

00:00As a point of review, can anybody tell me where you actually do translation? Where do we start making our proteins? STUDENT: Cytoplasm. Cytoplasm, exactly. So we have a ribosome, and a little mRNA, maybe a little protein coming out.

00:25That's because transcription is in the nucleus-- mRNA comes out. And then that's where we start to do protein synthesis, which is in the cytoplasm. Now, for some of the proteins that are made, they need to stay in the cytoplasm.

00:43So this is no big deal. But there's a lot of different places that proteins need to act. And so there's many different ways in which the cell has executed this function. How do things get in and out of the nucleus? We also need to deliver proteins to specific organelles, like the mitochondria, so that those things can actually get into the membrane.

01:09So we have to move to this space. And as you can already tell, we need to get things into the ER, as well, so there's also a path, or a route. And what's special is that each of these do slightly different things. They're going to use a special mechanism that relies on gated pores.

01:37And this refers to the nuclear pore complex that we'll talk about towards the end. And then some of these actually use a transmembrane type of transport. This means the protein actually has to cross the plasma membrane or an organelle barrier in order to reach these different spaces.

02:06So we have examples here for the mitochondria. We also are going to talk about this and how it works for the ER. OK? Now, a lot of things that get into the ER don't stay in the ER. They have to then traffic to other places in the cell.

02:20And there's the specific route that we'll talk about of how they go from the ER to the Golgi. And then often they'll go into additional vesicles, smaller vesicles like endosomes or lysosomes. And then some of these are also destined for the plasma membrane.

02:54And the destination, you'll see, is very much dictated by the type of protein and the type of information that's stored in that protein to tell it whether or not it should stop here or keep going. And it turns out it doesn't just go in this direction.

03:08It's also reversible, depending, again, on the cargo. Now, the way that this happened is very distinct from methods 1 and 2. We don't actually cross the membrane in this case. We do something that involves vesicle trafficking, which is our third form of movement.

03:35And notably, again, this is where we've packaged cargo in a membrane-bound compartment. And it just transitions to each of these things back and forth. Now, I've mentioned that protein synthesis starts in the cytoplasm.

03:51This is super important to remember. And then depending on the type of which one of these different routes it's going, some of these movements can continue to go on post-translationally, or they can be cotranslational. And so you'll see, through the lecture, which of these categories fit within that set.

2.2信号假说与分选信号：必要性与充分性的经典实验推演

早在 20 世纪五六十年代，现代细胞生物学之父 George Palade 就通过电镜观察发现了一个耐人寻味的现象：非分泌性的胞质蛋白主要在游离核糖体上合成，而分泌蛋白则无一例外在附着于糙面内质网（rER）的核糖体上合成。但为什么偏偏是合成分泌蛋白的核糖体跑去贴在内质网膜上？是这群核糖体天生跟别的核糖体长得不一样吗？实验很快否定了这一点——把附着核糖体从内质网上洗脱下来，它们在体外跟游离核糖体没有任何结构和功能差异。

1971 至 1975 年间，Palade 的学生 Günter Blobel 与 David Sabatini（以及剑桥大学的 César Milstein 团队）大胆提出了一个在当时饱受质疑、最终却为 Blobel 赢得 1999 年诺贝尔生理学或医学奖的构想——信号假说（signal hypothesis）。

一、科学思维的黄金准则：如何证明信号序列的“必要性”与“充分性”？

在 MITx 课堂上，Lamason 教授在黑板上画了两个细胞，用一场极具启发性的师生问答，展示了细胞生物学家证明一个元件功能的终极逻辑——必要性（Necessity）与充分性（Sufficiency）：

假设我们在细胞里有两个蛋白质：蛋白 A 是正常定位于内质网（ER）的蛋白，它的 N 端带有一段疑似的信号序列（signal sequence）；蛋白 B 是正常定位于细胞质基质（cytosol）的蛋白，没有这段序列。

1. 实验一（基因敲除/缺失实验，Loss-of-function）：

• 教授提问："What do you think will happen if we take the signal sequence and delete it off of A? Where will A go?"
• 学生回答："Cytoplasm!"
• 实验结论：当把 N 端信号序列从 ER 蛋白 A 上人为切除后，突变体蛋白 A 无法再进入内质网，而是滞留在细胞质基质中。删除该信号序列导致蛋白失去靶向内质网的能力，这在逻辑上证明了该信号序列对于蛋白进入内质网是“必要的（Necessary）”！

1. 实验二（嵌合蛋白嫁接实验，Gain-of-function）：

• 教授提问："Now, what would happen if we moved the signal sequence to B? Where is it going to go?"
• 学生回答："ER! It doesn't matter the identity of the protein itself, it matters what the signal sequence is that determines where it goes — Sufficiency!"
• 实验结论：利用重组 DNA 技术，把蛋白 A 的这段信号序列嫁接到原本纯属胞质蛋白 B 的 N 端，表达出的嵌合蛋白（Signal-B）竟然直接被导向并转运进了内质网腔！仅凭这一段短肽序列就能将原本无关的胞质蛋白重定向到目标细胞器，证明该信号序列对于指导蛋白质进入内质网是“充分的（Sufficient）”！

"We love it as scientists when we can show something is necessary and sufficient. It means you've got the minimal component that's driving this!"

不仅如此，Lamason 教授提醒大家，这一原理也是现代细胞生物学荧光成像的核心利器：如果你在实验室里想在活细胞中点亮某个特定细胞器（比如内质网或线粒体），只需要把绿色荧光蛋白（GFP 或 mNeonGreen）的基因与该细胞器的信号序列融合表达，GFP 就会像贴了邮政编码的信件一样，乖乖跑去把那个细胞器照得通亮！

二、Blobel 信号假说与微粒体（Microsome）体外重构实验

那么，Blobel 和 Dobberstein 在 1975 年究竟是用什么实验铁证让所有怀疑者心服口服的？他们利用了生物化学家手中的一件神器——微粒体（microsome）。

在华农考研真题中，微粒体（microsome） 是 2013、2016、2019 年连续三次考到的超高频名词解释！微粒体（microsome）：指细胞匀浆经超速离心处理后，由破碎的内质网形成的近似球形的囊泡结构；其核心组分为内质网膜与核糖体；虽为形态上的人工产物（细胞内并无微粒体这一实体结构），但仍具备蛋白质合成、糖基化及脂类合成等内质网的基本生化功能，故在体外无细胞实验研究中常等同于内质网 2013/2016/2019·名解。这里千万要注意一字之差的陷阱：微粒体（microsome，破碎的内质网囊泡）≠ 微体（microbody，即过氧化物酶体 peroxisome），二者毫无关系，切勿混淆！

3D共翻译转运：分泌蛋白进 ER 腔点击打开 3d.lkhan.ai 交互式三维结构模型↗

三、分选信号大家族全景图：信号肽、导肽与三维信号斑

既然每个细胞器都有自己专属的“邮政编码”，这些分选信号在化学结构上到底长什么样？在考研中，蛋白质分选信号（protein sorting signal） 2012·名解 分为两大形态类别：线性信号序列（signal sequence，包括信号肽、导肽、NLS 等） 与 三维空间构象信号斑（signal patch）。

3D共翻译转运：分泌蛋白进 ER 腔点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. 引导进入糙面内质网的“通行证”：信号肽（Signal Peptide）

信号肽（signal peptide）：位于分泌蛋白等新生多肽链 N 端、引导核糖体-新生肽复合物转移到内质网膜上的线性分选信号序列；一般由 16～26 个氨基酸残基组成，包括带正电荷的 N 端区、疏水核心区和含信号肽酶切割位点的极性 C 端区三部分 2007·填空 2012·名解 2017·选择。

仔细看上图的信号肽三段式结构：

• N 端区（N-region）：含 1～2 个带正电荷的碱性氨基酸（如 Lys、Arg），帮助信号肽在穿膜时把 N 端留在细胞质基质侧；
• 疏水核心区（H-region）：信号肽最核心的灵魂！无论不同分泌蛋白的信号肽氨基酸序列差异多大，其中心都必须包含一段由 8 个以上疏水（非极性）氨基酸残基组成的疏水核心区 2017·选择。正因为 Lamason 教授在课上反复强调 "The sequence itself turns out to be hydrophobic — that's all that really matters"，如果在选择题里看到“信号肽所含氨基酸均为亲水氨基酸”，直接秒选该项为错误选项 2017·选择！
• C 端区（C-region）：含数个极性氨基酸，具有被内质网腔面信号肽酶识别并专一性切割的位点。

2. 引导进入线粒体、叶绿体与过氧化物酶体的“通行证”：导肽（Leader Peptide）

在翟中和教材中，凡是指导游离核糖体合成的蛋白质在翻译后进入线粒体、叶绿体、过氧化物酶体等细胞器的靶向序列，广义上统称为导肽（leader peptide）（狭义上常专指线粒体前导肽）：

• 导肽（leader peptide / presequence）：引导游离核糖体合成的前体蛋白在翻译后跨膜进入线粒体（广义亦含叶绿体、过氧化物酶体）的靶向信号序列；线粒体导肽位于多肽 N 端，一般长 20～50 个氨基酸，富含疏水、碱性和羟基氨基酸，几乎不含带负电的酸性氨基酸，易形成一侧疏水、一侧带正电的“两亲性 α 螺旋（amphipathic α-helix）”，入基质后被基质蛋白酶切除 2012·名解 2012·简答。
• 转运肽（transit peptide）：位于核编码叶绿体前体蛋白 N 端，长约 40～50 个氨基酸，引导蛋白进入叶绿体基质后被基质蛋白酶切除，从而暴露出紧随其后的类囊体靶向序列。
• 过氧化物酶体靶向信号（PTS）：多数位于多肽链 C 端（PTS1，典型序列为 Ser-Lys-Leu [SKL] 三肽，由胞质受体 Pex5 识别；少数 PTS2 位于近 N 端，由 Pex7 识别），进入过氧化物酶体后不切除。

3. 不被切除的“永久工作证”：核定位信号（NLS）与信号斑（Signal Patch）

并不是所有的分选信号都是“一次性门票”（用完就被蛋白酶切掉）。为什么有些信号必须切掉，而有些信号绝不能切？我们从细胞周期的物理逻辑一推便知：

• 核定位信号（Nuclear localization signal, NLS）：富含赖氨酸（Lys）和精氨酸（Arg）等带正电碱性氨基酸的短序列，可位于亲核蛋白肽链的任何部位，且指导蛋白入核后“绝不被切除” 2012·简答！为什么不能切？因为有丝分裂期核膜会崩解，核内蛋白全洒回细胞质中；分裂末期核膜重建时，这些蛋白还指望着靠身上的 NLS 再次入核呢！
• 信号斑（Signal patch）：信号斑（signal patch）：指在一级线性序列上相距较远、但当多肽链折叠成天然三维构象后在蛋白质表面相互靠近并聚集成簇的特定氨基酸残基所构成的三维空间分选信号。最经典的例子就是溶酶体水解酶：当溶酶体酶前体在糙面内质网折叠后进入高尔基体顺面膜囊时，正是其表面的三维信号斑被高尔基体中的 N-乙酰葡糖胺磷酸转移酶特异识别，从而给它的甘露糖贴上 M6P（6-磷酸甘露糖）标签！

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 1 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 02 · Signal Hypothesis⏱ 04:34

00:00When this first came out, this was originally hypothesized that proteins had an inherent sequence by Gunter Blobel and David Sabatini, Sr., not Jr. They actually proposed that the proteins would have a special signal within this protein.

00:15And for a long time, nobody believed them. And so Blobel actually worked on this doggedly, despite all of this criticism, and eventually showed that this was the case. And so they had proposed, essentially, this idea of the signal hypothesis.

00:33And I'll tell you little bit more about what the signal is, and then we'll talk about why it is so special and what it was that made it so distinct, and why it was able to function so well. Essentially, in order for a protein to know where to go, it needs some type of a signal sequence.

00:54And this comes in a variety of different flavors. But typically, they're between 15 and 60 amino acids long. And if you have a sequence, in a signal sequence, you need some way to recognize it. And so all of these have corresponding sorting receptors.

01:21And something we'll cover in this lecture and the later lectures is, how do we think about different signals and different sorting receptors, and keeping track of those for the different compartments. And what's special about them is that they are destination-specific.

01:36So there has to be a match between, say, an ER or a Golgi specific sorting receptor and a signal sequence in order for the cargo to know where to go. And how did we actually think about this? And essentially, a lot of times, people refer to this as what's like, a postal code.

01:52So that things know where to go. And so there's two key features that we can think about when we think about how a signal sequence confers its activity. And so if we do this in the context of an actual experiment, this makes a little bit more sense.

02:08So here's a cell. We'll actually take two. So we'll do the ER. And let's say we have an ER-specific protein that we'll call A and a cytosolic-specific protein that we'll call B. And I just told you, in order for a protein to go into the ER, it has to have a signal sequence.

02:36So in this case, we'll give it an end terminal signal sequence. What do you think will happen if we take the signal sequence and delete it off of A? Where will A go? STUDENT: Cytoplasm. Cytoplasm. Exactly. It will end up in the cytoplasm in the cell.

02:56What this does, is it shows that the signal sequence is necessary. So that's one key thing. Now, what would happen if we moved the signal sequence to B? Where is it going to go? STUDENT: [INAUDIBLE] Say that again. Louder.

03:16ER. Very good. And so what does that show? Go ahead, right there. STUDENT: It doesn't matter the identity of the protein itself it matters what the signal sequence is that determines where it goes. Exactly. Is there another word that you guys think about, when you think about that idea?

03:35Very good explanation. Go ahead. STUDENT: Sufficiency. Sufficiency. Exactly. We love it as scientists, when we can show something is necessary and sufficient. It means you've got like, the minimal component that's driving this.

03:52In our case, as long as there's a signal sequence, it doesn't-- like you said, it doesn't matter what's here. This protein will go to that compartment, assuming of course, there is a sorting receptor available to see it.

04:03So this was a huge thing. And this seems very simple, but it's surprising how much pushback Blobel got for this. And then to reward him, he got the Nobel Prize to show that this was such a key feature. So this is one important thing.

04:16And we're going to talk about a lot of the different signals that confer this. And we as scientists often use these signals to mark things in the cell, and to specifically label if we want to see compartments. And so it's not only useful for the cell to use this, but it's useful as a cell biological technique as well.

2.3共翻译转运：SRP 翻译暂停、移位子开孔与跨膜蛋白拓扑学

现在，让我们把镜头推近到纳米尺度，跟着 Rebecca Lamason 教授一步步看清：一个携带 N 端信号肽的可溶性蛋白（或跨膜蛋白），究竟是如何在游离核糖体上露头、被押送到内质网膜上，并穿过脂双层屏障的。

一、共翻译转运的“四大金刚”：SRP、SRP 受体、移位子与信号肽酶

在共翻译转运的舞台上，除了前面讲过的信号肽外，还有四个缺一不可的核心分子机器。华农真题曾多次以名解、填空或论述题形式考查它们的定位、结构与功能 2007·填空 2009·名解 2015/2018·论述：

1. 信号识别颗粒（Signal-Recognition Particle, SRP） 2009·名解：

• 信号识别颗粒（SRP）：存在于细胞质基质中的核糖核蛋白复合体，由 6 种不同多肽亚基和一条长约 300 个核苷酸的 7S RNA 分子组成，包含信号肽结合结构域、核糖体结合结构域及结合 GTP 的 p54 亚基；能特异识别并结合暴露的新生肽 N 端信号肽与核糖体，使肽链延伸暂停，并引导核糖体-新生肽复合物与内质网膜上的 SRP 受体结合 2009·名解。
• MBoC 结构揭秘：为什么同一个 SRP 能够识别成百上千种氨基酸序列各不相同的 ER 信号肽？因为 SRP 的 p54 亚基信号结合口袋内衬有大量富含甲硫氨酸（Met）的 α 螺旋，甲硫氨酸无分支的疏水侧链极具柔性，像一块可塑的“疏水橡皮泥”，能轻松贴合各种长度和形状的疏水信号肽核心区！

1. SRP 受体（SRP Receptor，又称停泊蛋白 Docking Protein, DP）：

• SRP 受体（SRP receptor / docking protein, DP）：镶嵌于糙面内质网膜上的整合膜蛋白异二聚体，由胞质侧的 α 亚基和跨膜的 β 亚基组成（均含 GTP 结合域）；可特异识别并结合胞质中的 SRP，将核糖体-新生肽-SRP 三元复合物锚定到内质网膜上。

1. 移位子（Translocon / Protein Translocator，即 Sec61 复合体）：

• 为什么Lamason教授说这个通道叫 Sec61？因为它是Randy Schekman等人在酵母遗传筛选中分离出的第61号分泌缺陷突变体（sec61 mutant）！
• 移位子（translocon / Sec61 complex）：位于糙面内质网膜上供新生多肽链跨膜穿行的水性蛋白质通道复合体；核心由 Sec61α/β/γ 三个亚基组成（Sec61α 的 10 条跨膜 α 螺旋围成中央孔道），静息时由短螺旋构成的“塞子（plug）”封闭孔道以防 ER 腔内 Ca²⁺ 外漏，当环化信号肽楔入其侧门（lateral gate）时塞子移开、孔道开启。

1. 信号肽酶（Signal Peptidase）：

• 位于内质网膜腔面（紧靠 Sec61 移位子旁）的蛋白酶复合体，负责在多肽链跨膜延伸途中专一性切除 N 端信号肽。

二、共翻译转运七步标准流程：为什么 SRP 必须先“叫停”翻译？

在课堂上，Lamason 教授一边在黑板上画出核糖体和 SRP，一边特别强调了一个精妙的动力学设计：当 N 端信号序列刚从核糖体出口隧道露出时，SRP 一把抓住它，同时立刻叫停了核糖体的翻译延伸（halts translation / elongation arrest）！

为什么细胞要先踩一脚刹车？我们用反证法一想就明白：

1. 防止多肽链在胞质中提前折叠成球：Sec61 移位子的中央水性孔道非常狭窄（仅容未折叠的线性肽链或单根 α 螺旋通过）。如果核糖体在游离状态下以每秒十几个氨基酸的速度一路狂飙，等它慢悠悠飘到内质网膜时，整条多肽链早就翻译完并在胞质里折成了一个大球，再也塞不进 Sec61 通道了！
2. 保护细胞质免受危险酶类的破坏：许多经内质网途径合成的蛋白是溶酶体酸性水解酶或蛋白酶，如果任由它们在细胞质基质中合成释放并折叠出活性，就会在细胞质里乱切一气！

让我们把这七步标准过程逐条拆解清楚（考场论述题直接按此分点书写）：

1. 结合 SRP 与翻译暂停（第 1 步）：翻译起始于细胞质基质的游离核糖体。当新生多肽链延伸至约 80 个氨基酸残基时，N 端信号肽露出核糖体大亚基隧道，胞质中的 SRP 立即结合信号肽与核糖体（占据延伸因子结合位点），使肽链延伸暂时停止 2015·论述，从而防止新生肽 N 端受损或成熟前提前折叠（注意中英口径：MBoC 写翻译“减慢”，国内考场按翟中和统一写肽链延伸“暂停”）。
2. 锚定内质网膜（第 2 步）：SRP 携带核糖体-新生肽复合物移动并结合到糙面内质网膜上的 SRP 受体（停泊蛋白 DP）上；此时有 2 分子 GTP 分别结合到 SRP 的 p54 亚基和 SRP 受体的 α 亚基上 2007·填空 2015·论述，强化各复合物之间的相互作用。
3. 移交移位子与恢复延伸（第 3 步）：在 SRP 与 SRP 受体的协同下，核糖体-新生肽复合物与内质网膜上空闲的移位子（Sec61 复合体）结合；随后 SRP 与 SRP 受体共同水解结合的 GTP，SRP 脱离信号肽和核糖体，返回细胞质基质中重复使用，核糖体的肽链延伸随之恢复 2009·名解。
4. 信号肽开孔与袢环入腔（第 4 步）：呈环化（发夹）构象的信号肽与移位子组分结合并楔入侧门，使移位子塞子移开、孔道打开（该过程消耗能量），引导紧随其后的新生多肽链以“袢环（loop）”形式穿入内质网腔 2015/2018·论述。注意 Lamason 教授强调的细节：信号肽本身因为疏水，并不会进入腔内，而是卡在移位子侧门与脂双层之间！
5. 切除信号肽（第 5 步）：位于 内质网腔面的信号肽酶（signal peptidase）将信号肽切下 2017·判断，切下的疏水信号肽从移位子侧门滑入脂双层并被膜蛋白酶迅速降解。
6. 合成完成与腔内折叠（第 6 步）：核糖体紧贴在 Sec61 通道口上方，利用肽链翻译延伸的机械推力（driven by translation）将不断长出的多肽链源源不断地推入内质网腔，直至整条多肽链合成完毕，并在腔内分子伴侣（如 BiP）协助下折叠及发生糖基化等修饰 2026·选择。
7. 核糖体释放与通道关闭（第 7 步）：翻译终止后，核糖体大小亚基从内质网膜上解离释放回细胞质基质，移位子塞子复位、通道关闭，共翻译转运过程圆满终止。

三、内质网跨膜蛋白的整合：序列组合与“正电荷在内规则”决定跨膜拓扑学

前面讲的是像胰岛素、抗体这样的可溶性分泌蛋白——它们全身上下只有 N 端一段疏水信号肽，切掉之后整条链就彻底掉进了内质网腔里。但如果细胞要制造的是一个需要镶嵌在质膜或细胞器膜上的跨膜蛋白（如受体、离子通道、GPCR），它怎么停在膜上而不掉进腔里去呢？

在 2015 年华农论述真题中，题目专门要求“以细胞膜蛋白为例，说明蛋白质如何在内质网上合成并最终到达靶膜” 2015·论述 ⚑华农考过。所有的质膜、内质网、高尔基体和溶酶体跨膜蛋白，它们的跨膜拓扑结构（transmembrane topology，即跨膜几次、N 端和 C 端分别朝向胞质侧还是腔面/胞外侧）早在糙面内质网膜上共翻译插入的那一刻就已经被永久锁定，在后续的膜泡运输中再也不会翻转！

决定跨膜拓扑的密码，是新生多肽链中三类疏水序列的排列组合，外加一条极好用的电荷物理法则：

3DER 膜蛋白的拓扑点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. 三类决定跨膜的疏水序列

• 起始转移序列（Start-transfer sequence）：负责启动多肽链穿过 Sec61 移位子孔道。前面讲的可切除 N 端信号肽其实就是起始转移序列的一种；此外在多次跨膜蛋白内部也有不被切除的内在起始转移序列。
• 停止转移序列（stop-transfer anchor sequence, STA，又称内在停止转移锚定序列）：多肽链内部一段由 20～25 个疏水氨基酸残基组成的 α 螺旋序列；当其进入 Sec61 移位子通道时发出“停止穿膜”信号，因与脂双层疏水核心亲和力极强，从移位子侧门横向滑入磷脂双分子层并永久锚定，使后续合成的 C 端肽段留在细胞质基质侧 2015·论述 ⚑华农考过。
• 信号锚定序列（signal-anchor sequence, SA，又称内在信号锚定序列）：位于多肽链内部、兼具“信号肽靶向功能（被 SRP 识别并打开移位子）”与“跨膜锚定功能（从侧门滑入脂双层永久锚定且不被信号肽酶切除）”的疏水跨膜序列。

2. 单次与多次跨膜蛋白的 4 种拓扑类型（Ⅰ～Ⅳ 型）与“正电荷在内规则”

当一段内部的信号锚定序列（SA）进入 Sec61 通道时，它既没有可切除的 N 端信号肽打头，自己又要永久留在膜里，那它到底把自己的 N 端朝向胞质、还是把 C 端朝向胞质？

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 1 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 03 · Cotranslational Transport⏱ 04:27

00:00How a soluble protein gets into the ER. And in this case, we're going to talk about this in the context, first, of cotranslational. And then we'll talk about this for post-translational, because there's two basic ways this happens.

00:23All right, so you've got your ribosome, mRNA. So as this is happening in the cytosol, the mRNA is starting to be read. And a protein is going to be coming and emerging from the ribosome. And our signal sequence to get into the ER is actually located on the N-terminus of this protein.

00:40And so that's the first thing that's going to emerge here. The sequence itself turns out to be hydrophobic. That's all that really matters, as long as it's a short hydrophobic sequence. As this thing starts to emerge, the sorting receptor is actually sitting in the cytosol, as well, and is waiting for it to come out.

00:57And when it does that, a specific protein binds to this space here. And this is called SRP, which stands for signal recognition particle. Its job is to recognize the signal sequence. This protein binds to this space and actually interacts with the ribosome and halts translation, kind of holds it in this spot.

01:19And what it does is it then moves over to the ER where it can then deliver the translating protein into the ER. Here's the cytosol and the lumen. And the SRP is actually recognized by a receptor on the surface of the ER called the SRP receptor.

01:36And so it comes over here and sits. And since it's bound currently to a signal sequence that's sticking out, that's attached to a ribosome and a little bit mRNA, it's now bringing it to the ER just by simply interacting with this hydrophobic segment and this SRP.

01:52Nearby, there's actually a channel situated in the ER membrane. So remember, proteins are not going to easily want to cross. We need some way to transport it. So here we have our first sort of transporter, or channel-- kind of goes through a couple of different names.

02:07It's called Sec61. It's just a secretion mutant that was identified, number 61. And what's special about this is that, as it sits here, the SRP will actually then deliver this complex over to this channel. And it will actually sit on top of it.

02:23Now the ribosome actually sits on top. We still have mRNA associated with it. And that hydrophobic segment of the single sequence starts to go into this space. Now, because it's hydrophobic it's actually not going to completely go into the lumenal space.

02:39It'll actually get stuck in here. And what's going to happen is that, because the SRP is gone, it's no longer blocking translation, it is now free to continue to translate. And so what happens is more and more of the protein sequence starts to emerge and get pushed through the channel.

02:57This is very specific to this type of movement, that cotranslational transport, across the ER membrane, the force to push a protein is driven by translation. This ribosome sits in this space, blocks the protein from maybe getting out, and drives it through.

03:14Now, since it's a soluble protein, it needs to make it into the lumenal space. And it doesn't want to carry along this signal sequence here, because it's hydrophobic, and that's going to want to stay in the membrane. And so one more protein comes along.

03:27This is a signal peptidase. And its job is to basically come in and cleave this here. The signal sequence actually gets pushed out into the membrane space and left behind so that the soluble protein can continue to be pushed all the way through, delivered, and then folded in the lumenal space.

03:46So this is how you go through taking a soluble protein that's destined for the lumenal space of the ER at the very beginning and allow it to actually cross through, through the action of a sorting receptor-- basically the combination of SRP and the SRP receptor-- and delivering that to this very special set of proteins known as part of this Sec61 channel that allows the ribosome to sit on top and push proteins through through the force of translation.

04:14OK. And we've got chaperones in the lumenal space that allows these proteins to then be folded so they can be fully functional and deliver to wherever they need to be and to do whatever they need to do.

2.4翻译后转运：BiP 布朗棘轮模型与线粒体/叶绿体/过氧化物酶体输入

在上一节里，我们看到共翻译转运之所以能顺畅地把多肽链塞进内质网腔，靠的是核糖体直接坐在 Sec61 通道顶端，用翻译延伸的机械力把肽链往里“推（push）”。然而，Rebecca Lamason 教授紧接着在课堂上抛出了一个反例：

"Sometimes there's proteins that are completely translated in the cytosol, and yet they still need to get into the ER. And that means they're not going to have a ribosome that's going to push the protein through! So how do we actually get this to work?"

不仅在酵母等真核生物中有相当一部分可溶性 ER 蛋白走的是完全合成后再进入内质网的路线，在所有真核细胞中，核编码的线粒体蛋白、叶绿体蛋白和过氧化物酶体蛋白更是近乎 100% 在细胞质基质的游离核糖体上完全翻译完毕后，才进行翻译后跨膜转运（post-translational translocation） 2010/2015/2020·简答！没有了核糖体的推力，细胞靠什么物理机制把一条软塌塌的多肽链穿过狭窄的膜通道“拉（pull）”进细胞器内部？

一、内质网翻译后转运：分子伴侣 BiP 与“布朗棘轮模型（Brownian Ratchet）”

让我们先跟着 Lamason 教授看一条在胞质中完全合成的多肽链是如何翻译后进入内质网腔的：

3D线粒体蛋白输入点击打开 3d.lkhan.ai 交互式三维结构模型↗

首先，这条完全合成的多肽链要想钻进狭窄的 Sec61 通道，绝不能在细胞质里折叠成团。因此，在细胞质基质中必须有胞质分子伴侣（如 Hsp70）结合在多肽链上，消耗 ATP 使其保持未折叠的线性延伸状态（linear, unfolded fashion）。接着，它 N 端的信号序列同样被识别并送入内质网膜上的 Sec61 通道（伴随 Sec62/Sec63 等附加复合体组装），信号肽酶切掉信号序列。

现在，最严峻的物理难题来了：没有核糖体封堵通道口并往下推，通道里的多肽链受热运动（thermal fluctuations，即布朗运动）影响，既可以随机往腔内滑，也可以随机往胞质外退，甚至滑脱出去！是谁在内质网腔里把它拉进来的？

Lamason 教授在黑板上写下了一个名字——BiP（Binding Protein，结合蛋白），并向全班同学发起头脑风暴："How can a chaperone actually pull on a protein? Can anyone think about why this interaction might drive movement in an inward fashion?"

二、核编码蛋白质输入线粒体：翻译后跨膜转运的巅峰之作

如果说酵母内质网的翻译后转运在哺乳动物中还不算主角，那么在所有真核生物中，线粒体蛋白质的跨膜输入则是翻译后转运绝对的头号考场！

为什么线粒体表面几乎看不到附着核糖体？因为线粒体虽然拥有自己的环状 DNA，但只能编码不到 1%（人类仅 13 种）的线粒体蛋白；线粒体 99% 以上的蛋白质均由细胞核基因编码，在细胞质基质的游离核糖体上完全合成后，以翻译后跨膜转运方式输入线粒体的四个区室：外膜、膜间隙、内膜和基质 2010/2015/2020·简答。

1. 核心必背：前体蛋白输入线粒体基质的“六步标准流程”

穿过线粒体的双层膜进入最里面的线粒体基质（matrix），需要外膜上的 TOM 复合体（Translocase of the Outer Membrane，含受体 Tom20/Tom22 和通道孔蛋白 Tom40）与内膜上的 TIM 复合体（Translocase of the Inner Membrane，主要为 Tim23/Tim17 通道）在内外膜接触点（contact sites）协同配合 2010/2015/2020·简答：

1. 第①步（胞质伴侣维持未折叠）：游离核糖体合成的前体蛋白 N 端带有基质靶向序列（即线粒体导肽，长 20～50 aa，形成带正电的两亲 α 螺旋），胞质中的分子伴侣 Hsp70 与前体蛋白结合，消耗 ATP 使其保持未折叠（或部分折叠）的线性状态，防止其在胞质中提前折叠或聚集而无法穿孔 2012·简答。
2. 第②步（识别结合外膜输入受体）：前体蛋白通过 N 端基质靶向序列，被线粒体内外膜接触点附近的外膜输入受体 Tom20/Tom22 特异识别并结合。
3. 第③步（穿越内外膜双层通道）：胞质 Hsp70 解离，前体蛋白以未折叠形式先后穿过外膜输入孔道 Tom40 和内膜转运通道 Tim23/Tim17（在接触点处短暂同时跨越内外两层膜；此步导肽的正电荷在内膜“外正内负”的跨内膜电化学质子梯度/膜电位 ΔΨ 吸引下被电泳式拉入基质）。
4. 第④步（基质 Hsp70 耗能拉入）：露出内膜基质侧的多肽链与线粒体基质中的分子伴侣（基质 Hsp70 / mtHsp70）结合，基质 Hsp70 水解 ATP 产生机械力（布朗棘轮/动力冲程），驱动整条多肽链穿膜进入基质 2010/2015·简答。
5. 第⑤步（切除基质靶向序列）：多肽链进入基质后，基质蛋白酶（MPP）将 N 端的基质靶向序列（导肽）切除，同时基质 Hsp70 从新输入的蛋白质上释放。
6. 第⑥步（折叠为活性形式）：在基质分子伴侣（如 Hsp60 伴侣蛋白桶）的协助下，成熟多肽链正确折叠组装成具有生物活性的线粒体基质蛋白。

2. 进阶满分：输入线粒体内膜（a/b/c 三途径）与膜间隙（a/b 两途径）

如果 2020 年那样的简答题再次出现（“细胞核编码的蛋白是如何进入线粒体的？”），除了答完上面“一、输入基质六步”，还必须补齐“二、输入内膜三条途径”和“三、输入膜间隙两条途径” 2020·简答：

三、叶绿体与过氧化物酶体的翻译后输入机制

为了把翻译后转运拼图彻底补全，我们再用最精炼的语言拿下叶绿体和过氧化物酶体的蛋白质输入：

1. 叶绿体蛋白质输入（TOC/TIC 与类囊体双途径）：

• 叶绿体不仅有内外双层膜，基质里面还有第三层膜——类囊体膜（thylakoid membrane）及类囊体腔！
• 进入叶绿体基质：核编码前体蛋白凭 N 端 40～50 aa 的转运肽（transit peptide），以未折叠形式穿过外膜 TOC 和内膜 TIC 通道进入基质，随后转运肽被基质蛋白酶切除。为什么叶绿体基质输入不像线粒体那样利用跨内膜 H⁺ 梯度，而几乎完全依赖 ATP（和 GTP）水解？因为叶绿体把质子泵（光合电子传递链）全装在了内部的“类囊体膜”上，其内外包膜之间根本没有跨膜电化学质子梯度！
• 进一步进入类囊体（二级分选）：当基质蛋白酶切掉第一段 N 端转运肽后，紧随其后的第二段类囊体靶向序列暴露出来，通过两条主要途径进入类囊体：① 叶绿体 SRP 途径（如质体蓝素，在基质中保持未折叠，经无 RNA 的叶绿体 SRP、受体及 SecY 通道入腔后切去信号）；② pH 依赖途径（即 TAT 途径，双精氨酸途径）（如金属结合蛋白前体，信号含两个精氨酸 Arg-Arg，先在基质中完全折叠并结合金属辅因子，再依赖跨类囊体膜的 pH 梯度/H⁺ 梯度直接跨膜进入类囊体腔）。
• 2025 年华农单选题曾出过一个极具迷惑性的错项：“所有叶绿体蛋白在合成时都具有转运肽”——错！一则叶绿体自身基因组编码合成的蛋白没有转运肽；二则多数叶绿体外膜蛋白由内在序列引导直接插入外膜，也没有可切除的 N 端转运肽 2025·选择！

1. 过氧化物酶体（Peroxisome / 微体）的发生、输入与溶酶体对比 2012/2015·简答：

• 过氧化物酶体（peroxisome）：又称微体（microbody），是由单层膜围绕、内含一种或几种氧化酶类（如黄素氧化酶和标志酶过氧化氢酶，常含尿酸氧化酶晶格核心）的异质性细胞器 2015·简答。
• 三大功能必背 2012·简答：① 氧化解毒（如肝肾细胞中将饮入乙醇的约 25% 氧化为乙醛）；② 消除 H_2O2 毒害（黄素氧化酶利用 O₂ 氧化底物产生有毒的 H_2O2，紧接着标志酶——过氧化氢酶 [catalase] 立即将 H_2O2 分解为 H_2O 和 O₂，构成最简单的呼吸链）；③ 分解极长链脂肪酸等高能分子直接产热（注意：过氧化物酶体 β-氧化分解脂肪酸只直接释放热能，不偶联合成 ATP！要产 ATP 必须进线粒体）；④ 植物特化功能：叶肉细胞中参与光呼吸，油料种子萌发时作为乙醛酸循环体（glyoxysome）将脂肪酸经乙酰 CoA、琥珀酸转化为葡萄糖。
• 发生与蛋白输入特点：过氧化物酶体无自身 DNA，膜蛋白和基质酶均由核基因编码在胞质游离核糖体合成。它可以由内质网出芽的 Pex3/Pex16 前体小泡招募 Pex19 从头组装，或由已有过氧化物酶体经 Pex11 介导分裂增殖。最神奇的是：携带 C 端 PTS1（Ser-Lys-Leu）信号的酶在胞质中与受体 Pex5 结合（PTS2 与 Pex7 结合），经膜上 Pex14 易位体跨膜进入腔内时，“不需要解折叠”，能够直接转运完全折叠甚至寡聚化的蛋白质！若 PEX 基因突变导致蛋白无法输入，细胞内只剩空的过氧化物酶体膜泡，引发致死性的脑肝肾综合征（Zellweger 综合征）。

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 1 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 04 · Posttranslational Transport⏱ 06:53

00:00We also have to think about, sometimes there's proteins that are completely translated in the cytosol. And yet they still need to get into the ER. And that means they're not going to have a ribosome that's going to push the protein through.

00:10And so there has to be some other way for this to work. And so there's a slight variation. Some of the themes are very similar. But then there's some really interesting differences. And again, this is still-- or mostly thinking about, in this case, sort of a soluble protein.

00:40So in the cytosol, like I said, we have a protein that's been fully synthesized. It turns out it also has a signal sequence. But it also is bound by additional proteins, like chaperones. And they often keep these proteins in a linear fashion so that they're not fully folded.

01:05As you can see here, this is a fairly small, narrow channel. And so sticking a linear peptide or protein through would be much more efficient than a folded one. But because it has the signal sequence, it will go through very much the same process, where it's going to be recognized by an SRP and deliver to the SRP receptor.

01:26And we have a problem here. Remember, there is no force from ribosome translation to push it through. So how do we actually get this to work? So if we start here, where we have the original, we've already at least delivered the protein through the channel.

01:43And we've already deleted the signal sequence. But we need some way to get that protein all the way through. And it turns out there's another chaperone inside the lumen that goes by the name of BiP, which just, honestly, stands for binding protein, which is not very creative.

02:03[CHUCKLES] And BiP is special because it's also, like I said, a chaperone. And it has this ability to interact with the Sec61 channel. And the Sec61 channel actually activates the BiP activity. And BiP itself has an ATPase function, which I'll show you in a second.

02:27All right, so how can a chaperone actually pull on a protein? And so what we actually see is that, as it's emerging, BiP is able to bind. Can anyone think about why this interaction might drive movement in an inward fashion?

02:51STUDENT: Maybe BiP spins, so it kind of tugs the protein down. A spinning protein, OK. That's not a bad idea. [CHUCKLES] Any other ideas? Let's say you know you got rid of BiP in a genetic screen. You've identified that all your proteins are stuck in the cytosol.

03:07You've got to come up with some reason why interacting with this protein on the lumenal side could be important. Do you think this-- oh, go ahead. STUDENT: Could it do something to prevent the protein from leaving? Yep, that's one of the models.

03:22So she said, can it prevent the protein from moving back out? So remember, there's some thermal fluctuations happening. And you've been delivered by the SRP to this channel through the SRP and SRP receptor. But you might-- and especially once you've cleaved off that signal sequence, there might be some thermal fluctuation.

03:40And you might lose the protein out. And so one of the models is called a ratchet model, where you actually bind through a chain of BiPs. And as more and more protein emerges through thermal fluctuation, there's a new binding site for BiP to grab onto.

03:59And by locking onto it, it prevents it from going back out the channel. So that's sort of a Brownian ratchet model. And this binding as ATP-dependent. So we see that in the ADP state, BiP is able to bind. When it restores its ATP site, it actually then loses it.

04:27So there's a cycle of ATP-ADP binding that regulates this process. Now, there is somewhat-- maybe not a spinning protein-- [LAUGHS] but there is another way that they have also considered this, which is called a power stroke, which I think you are thinking about in the sense of like a singular protein doing something.

04:47And that model is essentially that it sits here. It's anchored. And through binding and unbinding, by regulating that ATPase cycle, you'll just keep pulling it through as a single protein. So this one is a chain. And this is sort of a single protein.

05:05And this is a challenging thing to figure out in cell biology. So there's still these two models that are not very well understood. It seems like most people favor this Brownian ratchet model. And you could probably start to think about how one could test this if you wanted to see, does the amount of BiP actually matter?

05:21Or, could you visualize the amount of BiP that's localizing to a protein as it's coming through? So based on this, they've sort of put forward this ratchet model as their favorite. So again, you have your signal sequence on this particular protein, which allows it to get delivered.

05:40You have the signal peptidase that needs to cleave this off. And now you have this free protein that somehow has to make its way in. And this is BiP. I mentioned that Sec61 actually activates BiP's activity. It actually specifically activates ATP hydrolysis, shown here.

05:58And you can see how there's this Brownian ratchet model, where a new protein, as it's sort of thermally fluctuating through as new protein binding sites emerge, this BiP is able to grab on. All right. So this would be a problem if-- what if we took got rid of BiP's signal sequence that needs to be in the ER?

06:15What would happen to BiP and the proteins that need to get in? Would that be-- would this particular process be specifically affected more than maybe a ribosome process? I see you nodding yes, so I'm going to just give it to you.

06:29Yeah. So this is very critical for a post-translational import. Yeah, and I don't-- it's not really sequence-specific at this point. It's not really something where there's a specific BiP-binding site, and only certain proteins have this.

06:43I think it's more of a chaperones just tend to sort of non-specifically bind. And so the idea would just be that it can just grab on to an emerging portion there.

2.5拓扑等价性、内质网加工质控与内质网应激（ERS & UPR）

当蛋白质千辛万苦通过共翻译或翻译后转运穿过 Sec61 通道进入内质网腔（或插入内质网膜）后，Rebecca Lamason 教授停下脚步，向大家提出了一个直击灵魂的设问：

"Why did we go to so much trouble to put something into the ER? Proteins don't really want to cross! And this is where we can rethink about this idea of topology... Remember, the inside lumenal space of our organelles, like the ER and the Golgi, is the equivalent of our outside space!"

一、拓扑等价性（Topological Equivalence）的物理与演化本质

什么叫拓扑等价（topologically equivalent）？观察下图：在真核细胞演化过程中，内质网、高尔基体、内体和溶酶体的膜系统是由原始质膜内陷形成的。因此，内质网腔、高尔基体腔、溶酶体腔、分泌泡腔与“细胞外空间（extracellular space）”在拓扑学上完全等价（均属于非胞质侧/腔面侧）；任何进入内质网腔的可溶性蛋白，无需再穿过任何一层生物膜，仅靠膜泡出芽与质膜融合即可直接排到细胞外；而跨膜蛋白暴露在内质网腔面的一侧（及其糖链），在融合到质膜后必定朝向细胞外侧！

3D内质网点击打开 3d.lkhan.ai 交互式三维结构模型↗

这一原理瞬间串联起了我们在第一章和本章学到的四个重要推论：

1. 为什么二硫键（-S-S-）只形成于内质网腔，而几乎不在细胞质基质中形成？ 因为细胞质基质是强还原性环境（富含还原型谷胱甘肽 GSH），半胱氨酸巯基（-SH）无法氧化成二硫键；而内质网腔模拟的是细胞外的强氧化性环境，在腔面蛋白二硫键异构酶（PDI）催化下专门为分泌蛋白（如胰岛素、抗体）和质膜外周结构域锻造稳定的二硫键！
2. 为什么内膜系统上的糖链永远朝向细胞器腔内，质膜上的糖萼永远朝向细胞外？ 因为催化 N-连接和 O-连接糖基化的所有糖基转移酶，其催化活性位点全部朝向内质网腔或高尔基体腔（拓扑学上的胞外侧）！（遇到判断题“内膜上的糖链朝向细胞质基质一侧”，秒判 ×！）

二、内质网的版图：糙面内质网（rER）与光面内质网（sER）全解

内质网（endoplasmic reticulum, ER） 是真核细胞中最大、最普遍、适应性最强的膜性细胞器，内质网膜约占细胞全部膜系统的一半（50%），其腔体体积占细胞总体积的 10% 以上。根据膜表面是否有核糖体附着，内质网分为呈扁囊状排列整齐的糙面内质网（rER）与呈分支管状的光面内质网（sER，滑面内质网） 2010·简答。注意：细胞中几乎不含纯粹孤立的光面内质网，rER 与 sER 是相互连通的整体网络。

对于光面内质网（sER），华农真题考得极为细致，以下四个功能点必须滚瓜烂熟：

1. 膜脂（磷脂与胆固醇）合成及跨膜转运三机制：

• 细胞所需的几乎全部膜脂（最主要是磷脂酰胆碱/卵磷脂）都在 sER 上合成 2009/2017·选择。合成磷脂的三种酶活性位点全部位于内质网膜的“细胞质基质侧”，利用胞质底物合成的磷脂最初只插入胞质侧单层（第一步即增大膜面积），随后在几分钟内经内质网膜上的“磷脂转位酶（爬行酶 scramblase）”翻转到腔面单层，实现双层膜对称生长。
• 内质网合成的磷脂如何运往其他膜？（填空题必背三机制）：① 通过出芽形成转运膜泡运往高尔基体、溶酶体和质膜；② 由水溶性的磷脂交换蛋白（PEP，脂转移蛋白）在膜间穿梭转运（线粒体、叶绿体和过氧化物酶体主要靠此途径获脂）；③ 供体膜与受体膜通过膜嵌入蛋白介导直接接触转运。
• 就在第③条里，藏着 2018 年华农考过的超纲名解——内质网-质膜接触点（ER-PM contact sites）：指内质网膜与质膜之间在纳米尺度（10～30 nm）上由系链蛋白固定而紧密靠近但“不发生膜融合”的特殊微区，是细胞器间进行非膜泡脂质交换、Ca²⁺ 信号传递和动态调控的重要平台 2018·名解 ⚑华农考过。

1. 肝细胞解毒作用（细胞色素 P450） 2014·简答 2025·选择：

• 肝细胞 sER 膜上的细胞色素 P450 单加氧酶系通过催化“羟基化（hydroxylation）”（以及氧化、还原、水解）反应，将脂溶性药物或毒物转化为水溶性物质，使其随尿液或胆汁排出体外 2025·选择。服用苯巴比妥等药物可诱导肝细胞 P450 解毒酶系大量合成，导致光面内质网面积成倍增加。

1. 肝细胞糖原代谢 2014·简答：

• 附着在肝细胞 sER 胞质面的糖原颗粒先被糖原磷酸化酶降解为 1-磷酸葡萄糖（G-1-P），在基质中由变位酶转为 6-磷酸葡萄糖（G-6-P）；因为带磷酸基团的 G-6-P 无法穿过质膜入血，必须转入 sER 腔内，由光面内质网腔面特有的“葡萄糖-6-磷酸酶（G-6-Pase）”催化去磷酸化生成游离葡萄糖，再释放入血以补充血糖 2014·简答。

1. 细胞内主要钙库与肌质网（Sarcoplasmic Reticulum） 2013·简答 2014·名解：

• 内质网腔内 Ca²⁺ 浓度高达 3 mmol/L，含有包括 BiP 在内的 4 种以上高容量 Ca²⁺ 结合蛋白（每分子可结合约 30 个 Ca²⁺）（顺便记住：高浓度 Ca²⁺ 可抑制 ER 出芽形成转运膜泡，且 Ca²⁺ 能把带负电的 BiP 桥连在带负电的磷脂头部上）。
• 肌质网（sarcoplasmic reticulum）：心肌和骨骼肌细胞中高度特化的光面内质网，是肌肉细胞的专用 Ca²⁺ 储库；膜上富含 Ca²⁺-ATP 酶（钙泵），利用 ATP 将胞质 Ca²⁺ 泵入腔内储存，当神经冲动传来时快速释放 Ca²⁺ 入胞质触发肌原纤维收缩 2013·简答 2014·名解。

三、糙面内质网腔内的加工、折叠与质量监控（Quality Control）

当新生多肽链进入糙面内质网腔后，它会立即经历四项紧锣密鼓的加工修饰 2018·论述：

1. N-连接糖基化（N-linked glycosylation）（最具代表性的共翻译修饰，详述见下）：

• 在内质网膜上，先以脂质分子磷酸多萜醇（dolichol phosphate）为载体，组装出一个标准的含 14 个糖残基的寡糖前体（2 个 N-乙酰葡糖胺 GlcNAc ＋ 9 个甘露糖 Man ＋ 3 个葡萄糖 Glc）；
• 当新生肽链进入 ER 腔露出 特征序列 Asn-X-Ser/Thr（其中 X 为除脯氨酸 Pro 以外的任意氨基酸） 时，腔面的寡糖基转移酶（OST）将这 14 个糖基组成的寡糖链一步整体转移到天冬酰胺（Asn）侧链的酰胺氮原子上 2013/2025·名解！随后在 ER 腔内被糖苷酶切掉 3 个葡萄糖和 1 个甘露糖，形成高甘露糖型核心糖链后运往高尔基体。

3DER 中的加工与质量监控点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. 二硫键形成与重排（PDI）：

• 蛋白二硫键异构酶（protein disulfide isomerase, PDI）：附着于内质网腔面的氧化还原酶（C 端含 KDEL 序列），能催化多肽链半胱氨酸残基间形成二硫键，并切断错误配对的异常二硫键，帮助蛋白质重新形成自由能最低的天然正确构象。

1. 分子伴侣协助折叠与亚基组装（BiP、钙连蛋白 Calnexin、钙网蛋白 Calreticulin）：

• 除了前面讲过的 BiP，内质网膜上的钙连蛋白（calnexin）与腔内的钙网蛋白（calreticulin）是专门识别糖蛋白上“单葡萄糖 N-糖链”的凝集素伴侣。

1. 特异性蛋白水解切割、酰基化与羟基化：如胶原蛋白前体脯氨酸与赖氨酸的羟基化、跨膜蛋白胞质侧 Cys 的软脂酰化等。

内质网的“开放监狱”法则：质量监控（ER Quality Control）与 ERAD

虽然合成一条几百个氨基酸的多肽链只需要几十秒到几分钟，但新合成的蛋白质往往要在内质网里停留几十分钟！为什么？因为内质网执行着铁面无私的质量监控（ER Quality Control） 2021·实验：

内质网中的 N-连接寡糖链充当着蛋白质折叠的“糖码计时器”：未折叠好的糖蛋白被 UDP-葡萄糖糖基转移酶（UGGT）反复加上 1 个葡萄糖，从而被钙连蛋白/钙网蛋白扣留在内质网中继续折叠；如果在内质网停留过久仍未折对，甘露糖苷酶就会剪掉其特定甘露糖残基，将其判定为“不可挽救的残次品”！

这些错误折叠或未能组装成寡聚体的残次品绝不允许进入高尔基体，而是启动内质网相关降解（ERAD）：错误折叠的多肽链经内质网膜上的 Sec61p 复合体（逆转位通道）被逆向转运回细胞质基质，随后经泛素化标记，被细胞质基质中的 26S 蛋白酶体彻底降解！若基因突变导致蛋白折叠异常（如囊性纤维化跨膜电导调节因子 CFTR ΔF508 突变，或 α₁-抗胰蛋白酶突变 2021·实验），突变蛋白即便仍有潜在催化活性，也会被 ER 质量监控系统拦截降解或滞留聚集，引发严重的“功能丢失”或“功能获得（如淀粉样斑块）”疾病。

四、⚑ 华农五年五考王牌：内质网应激（ERS）与未折叠蛋白应答（UPR）

如果遇到缺氧、缺糖、钙耗竭、病毒感染或突变蛋白大量表达，内质网里折叠错误的蛋白质来不及清除，越积越多，超过了内质网的负荷极限，细胞该怎么办？

内质网应激（endoplasmic reticulum stress, ERS）：指当内质网生理功能发生紊乱、钙稳态失衡，或未折叠/错误折叠蛋白质在内质网腔内过量积累时，细胞激活特定跨膜信号转导通路以应对危机的应激反应；主要包括未折叠蛋白应答（UPR）、内质网超负荷反应（EOR）、固醇调节级联反应以及启动细胞凋亡程序四个方面，是保护并恢复内质网稳态、监控蛋白质合成质量及决定细胞生死抉择的关键机制 2018/2024·简答 2022/2025·名解 2026·论述。

1. 内质网应激（ERS）的四种反应类型（2024 简答 / 2026 论述骨架）

1. 未折叠蛋白应答（Unfolded Protein Response, UPR）：未折叠或错误折叠蛋白在 ER 腔内聚集，触发信号通路使内质网分子伴侣（如 BiP/GRP78）和折叠酶（如 PDI）基因表达上调，同时暂停全局蛋白质翻译，以促进正确折叠、防止蛋白聚集，提高细胞存活能力 2018/2024·简答。
2. 内质网超负荷反应（ER Overload Response, EOR）：当正确折叠的蛋白质（尤其是膜蛋白）在内质网膜上过度蓄积时，激活转录因子 NF-κB，诱导产生促炎性细胞因子（如干扰素、白介素），进而激活细胞存活、凋亡、炎症及分化相关通路（注意：千万别像某讲义口误写成“线粒体超负荷反应”！）。
3. 固醇调节级联反应（Sterol Regulatory Cascade）：当内质网表面合成的胆固醇耗竭时，定位于 ER 膜的固醇调节元件结合蛋白（SREBP）在护送蛋白 SCAP 带领下转运至高尔基体，被蛋白酶（S1P/S2P）切割激活后入核，调控含固醇调节元件（SRE）的脂质合成相关基因转录；当胆固醇充足时，SREBP-SCAP 则滞留于 ER 膜抑制合成。
4. 启动细胞凋亡程序（ERS-induced Apoptosis）：若有害胁迫持续存在、内质网功能无法恢复稳态，内质网便释放储存的 Ca²⁺ 入胞质，激活钙依赖性蛋白酶 calpain（钙蛋白酶）；calpain 一方面裂解黏着斑蛋白破坏细胞骨架、切割抗凋亡蛋白 Bcl-xL 使其转为促凋亡分子，另一方面切割激活特异定位于内质网胞质侧的 caspase-12，最终触发细胞凋亡 2024·简答 2026·论述。

2. UPR 三条信号转导通路的“BiP 总开关”机制（2018 简答 / 2026 论述核心）

为什么内质网腔里未折叠蛋白一多，细胞核和核糖体就能立刻收到警报？奥秘全在于内质网膜上的三个跨膜感应蛋白——IRE1、PERK 和 ATF6，以及它们的共同看守者 BiP（GRP78）！

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 1 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 05 · Topology⏱ 01:20

00:00All right. So we've got our protein in either a soluble protein, or going through these two different types of transport. And now we can start moving out of the ER and thinking about why we went to so much trouble to put something into the ER.

00:17Proteins don't really want to cross. And this is where we can rethink about this idea of topology, which I introduced a few lectures ago. So if this is our cell, and we just brought in a protein into the ER, it's going to traffic to the Golgi.

00:49And this position-- remember, the inside lumenal space of our organelles, like the ER and the Golgi, is the equivalent of our outside space. And so what's so fantastic about this is that you bring these proteins in, you fold them, you modify them.

01:04You essentially make them fully functional before they even make it outside of the cell, because once you're out, you're in a hugely dilute space. You will not have chaperones and other modifying enzymes to facilitate a functional process.

01:17You need a fully functional protein here.

2.6分泌途径的发现与高尔基体枢纽：Palade 脉冲追踪实验、极性与溶酶体/自噬分选

当蛋白质在内质网里折叠妥当并通过了严苛的质量监控后，它就要正式踏上跨区室的旅程了。Rebecca Lamason 教授在黑板上画出了那条经典的向外箭头：ER → Golgi → Secretory vesicles → Plasma membrane（Exocytosis，胞吐/分泌作用），以及反方向从质膜向内的箭头：Plasma membrane → Endosomes → Lysosomes（Endocytosis，胞吞/内吞作用） 2017·选择。

教授紧接着向全班发问："We sort of take for granted that this road map exists. Turning back the clock, we didn't actually know this for a long time — how did we figure out the exact order in which proteins traffic through the cell?"

一、George Palade 的传世之作：脉冲追踪实验（Pulse-Chase）配电镜放射自显影

在电镜下，你看到的永远是一张张被定格的静态照片（static snapshots）：内质网里有蛋白，高尔基体里也有蛋白，分泌泡里也有蛋白。你怎么知道谁是第一站、谁是第二站、谁是终点站？

20 世纪 60 年代，George Palade 和同事以分泌功能极其旺盛的豚鼠胰腺外分泌腺泡细胞为模型，设计了细胞生物学史上最经典的动态追踪实验——脉冲追踪实验（Pulse-Chase Experiment）结合电镜放射自显影（EM Autoradiography）（这一成果直接助力 Palade 荣获 1974 年诺贝尔奖，也是 MITx 7.06 课堂必考的看家实验！）。

3D3D 演示：分泌蛋白从 ER 到质膜的运输观察 rER → 高尔基体 → 分泌泡全路线↗

3D共翻译转运：分泌蛋白进 ER 腔点击打开 3d.lkhan.ai 交互式三维结构模型↗

二、大分子运输的交通枢纽：高尔基体的结构区室、极性与维持机制

在分泌途径的十字路口，坐落着被誉为“细胞内大分子运输交通枢纽”和“糖类合成工厂”的细胞器——高尔基体（Golgi apparatus） 2025·选择。

高尔基体（Golgi apparatus / Golgi complex）：真核细胞中普遍存在的、由 4～8 个排列整齐的扁平膜囊（cisternae）堆叠成弓形主体、周围伴随大小不等囊泡所组成的高度动态且具有明显极性的单层膜细胞器；主要负责对内质网合成的蛋白质和脂质进行糖基化、硫酸化、水解切割等加工、分类与包装，是细胞内大分子转运的枢纽 2025·选择。（注：MBoC 写每个膜囊堆一般含 4～6 层扁囊；国内考场按翟中和标准答 4～8 个扁平膜囊！）

1. 高尔基体的五大结构区室与四种标志细胞化学反应

从接收内质网来货的入口（顺面，cis face，常朝向细胞核一侧凸面）到打包发货的出口（反面，trans face，常面向质膜一侧凹面；但注意判断题陷阱：高尔基体顺面并非总是位于凸面，在某些细胞或特定发育阶段顺面也可位于凹面！），高尔基体由外向内、由顺到反精细划分为 5 个功能区室：顺面高尔基网络（CGN）→ 顺面膜囊 → 中间膜囊 → 反面膜囊 → 反面高尔基网络（TGN） 2011/2014·名解。

其中最两端的两个网状结构是华农直接考过的名解原题：

• 顺面高尔基网络（cis Golgi network, CGN，顺面网状结构）：位于高尔基体顺面外侧、与顺面膜囊紧密相连的连续分支多孔网状结构，膜厚约 6 nm（比其他部位略薄，接近内质网膜），是高尔基体的物质入口区；负责接收来自内质网的新合成物质并进行初步分类，大部分转入中间膜囊继续加工，少部分逃逸的内质网驻留蛋白和脂质则从此经 COPI 膜泡遣返回内质网（此外还完成溶酶体酶寡糖磷酸化及参与冠状病毒装配） 2011·名解 2017·选择。
• 反面高尔基网络（trans Golgi network, TGN，反面网状结构）：位于高尔基体反面外侧、与反面膜囊相连并伸入反面细胞质的管网状结构，腔内 pH 低于其他部位，是高尔基体的物质出口区与蛋白质分选枢纽；既是网格蛋白/AP 包被膜泡和分泌泡的主要发源地，又承担半乳糖唾液酸化、酪氨酸硫酸化和肽激素水解活化等晚期修饰，并起“瓣膜（valve）”作用保证物质单向转运 2014·名解。

3D高尔基体点击打开 3d.lkhan.ai 交互式三维结构模型↗

2. 高频简答（2014 / 2026 两考）：如何理解高尔基体是极性细胞器？其极性是如何维持的？

华农在 2014 年和 2026 年两次用简答题正面考查高尔基体的极性 2014/2026·简答。考场作答必须分为“极性的三个体现方面”和“极性的四大维持机制”两段式：

• （一）如何从三个方面理解高尔基体的极性（2014 & 2026 前半问）：

1. 形态结构与位置极性：在细胞中往往具有恒定的空间位置与朝向（无极性动物细胞中常位于核附近、中心体/微管负端周围）；主体由扁平膜囊堆叠而成，近核一侧（常为凸面）为形成面/顺面（CGN），面向质膜一侧（常为凹面）为成熟面/反面（TGN），膜厚度从顺面（约 6 nm）向反面（约 8 nm）递增。
2. 化学组成（生化）极性：从顺面到反面各层膜囊具有完全不同的驻留酶系与特异性细胞化学标志反应——嗜锇反应特异染顺面膜囊，NADP 酶反应染中间膜囊，TPP 酶反应染反面 1～2 层膜囊，CMP 酶（及酸性磷酸酶）反应染反面管网结构（TGN）。
3. 功能与物质转运极性：物质严格遵循“顺面输入、反面输出”的单向流动——CGN 接收内质网来货并初筛（少部分遣返 ER，大部分进入中间膜囊）；中间膜囊进行核心糖基化修饰与多糖合成；反面与 TGN 进行晚期唾液酸化、硫酸化、水解加工，并作为分选枢纽包装出网格蛋白/AP 包被膜泡或分泌泡运往不同终点。

• （二）高尔基体的极性是如何维持的？（2026 简答追问 ⚑）：

1. 物质的单向膜泡流动与“瓣膜”机制：顺向运输（COPⅡ）与出口分选具有严格方向性，反面网状结构（TGN）在转运中起类似“单向瓣膜”的作用，保证物质由顺面向反面单向流动而不倒流；
2. COPI 介导的逆向回收纠错机制：无论是“膜泡运输模型（膜囊稳定）”还是“膜囊成熟模型（膜囊从顺面向反面渐次迁移成熟）”，不当向反面迁移的高尔基体各区室特异修饰酶或内质网逃逸蛋白，均通过反向运输的 COPI 包被膜泡（及 KDEL/KKXX 信号）按区室精准回收，防止区室成分混同；
3. 驻留酶系的跨膜固定与多酶复合物滞留：高尔基体各区室特征性的糖基转移酶与糖苷酶均为整合膜蛋白（活性位点在腔面），常通过跨膜区匹配不同膜厚或聚集成巨大多酶复合物而被固定在各自的膜囊间隔中，不被包入顺向运输小泡；
4. 细胞骨架与基质网架的锚定：高尔基体基质蛋白（如 GRASP、golgin）、血影蛋白（spectrin）网架以及微管和负端定向马达蛋白（胞质动力蛋白 dynein）共同锚定并维持高尔基体膜囊堆叠的动态空间极性 2026·简答。

三、N-连接糖基化 vs O-连接糖基化：全方位对比（2013 / 2024 / 2025 真题）

蛋白质糖基化（Protein glycosylation）是内质网和高尔基体联手完成的最核心修饰。2025 年华农同时考了 Protein Glycosylation 的英文名解和一道简答题“内质网和高尔基体中蛋白质糖基化的类型及其主要区别和联系” 2025·名解 2025·简答！

蛋白质糖基化（protein glycosylation）：指在蛋白质合成的同时或合成后，在一系列糖基转移酶和糖苷酶的催化下，将寡糖链共价连接到多肽链特定糖基化氨基酸残基位点上的过程；主要发生在内质网和高尔基体腔面，分为 N-连接糖基化和 O-连接糖基化两类，具有促进蛋白质正确折叠、增强抗蛋白酶稳定性、提供分选识别标志（如 M6P）、调节水溶性与表面负电荷及介导细胞间识别等重要功能 2013/2025·名解 2025·简答。

四、TGN 的三条分选去向与溶酶体酶 M6P 分选五步（2026 真题）

当加工成熟的蛋白质抵达高尔基体的出口站——反面高尔基网络（TGN）时，它们迎来了分道扬镳的时刻。高尔基体将蛋白质分选向三个终极去向：① 分泌到细胞外；② 整合到质膜上；③ 进入溶酶体等细胞器成为驻留蛋白。这对应着离开 TGN 的三条分选途径：

1. 组成型分泌途径（Constitutive secretory pathway，默认途径）：存在于所有真核细胞中。蛋白质无需特殊分选信号（分选信息仅存于蛋白自身），被连续不断地装入运输泡运往质膜并即刻融合胞吐（提供新合成的质膜脂质、跨膜蛋白及胞外基质蛋白）。在极性上皮细胞中，还会根据基因编码的内在信号进一步分选至顶面或基底外侧质膜。
2. 可调节性分泌途径（Regulated secretory pathway）：仅存在于特化的内/外分泌细胞（如胰岛 β 细胞分泌胰岛素、神经内分泌细胞分泌抗利尿激素 AVP 2024·论述、肥大细胞释放组胺）。可溶性分泌蛋白在 TGN 特定的微酸性与离子环境（约 pH 6.5、1 mmol/L Ca²⁺）中发生选择性聚集浓缩，包入分泌泡（分泌颗粒）中储存；在向质膜迁移过程中（或发生在 TGN 与分泌泡内），许多多肽类激素原（如胰岛素原、脑啡肽前体）经特异性蛋白水解切割才变成有活性的成熟激素 2025·选择，平时待命不释放，直到细胞受到胞外信号（如血糖升高或神经冲动）刺激引起胞质 Ca²⁺ 飙升时，才瞬间与质膜融合大量胞吐！
3. 溶酶体酶的 M6P（6-磷酸甘露糖）分选途径（五步必背！）：

五、内吞与消化终点站：胞内体（五考名解王）、溶酶体与细胞自噬（Autophagy）

既然讲到了溶酶体的发生，我们顺着这条流水线，把内膜系统里另外两个考频爆表的核心成员——胞内体（内体，endosome）、溶酶体（lysosome）与细胞自噬（autophagy）一举拿下！

1. 全书考频第一的名词解释：胞内体（Endosome，五考！）

在华农历年真题中，胞内体（endosome，又称内体；2015 年考“晚胞内体”） 在 2008、2010、2011、2015、2019 年足足考了五次名词解释，位列全书单词条考频榜首！

• 胞内体（endosome，内体）：动物细胞内由单层膜包裹的酸性囊泡状细胞器（分早期内体和晚期内体），主要功能是分选并传输由胞吞作用（内吞作用）新摄入的物质到溶酶体中进行降解；其膜上嵌有 ATP 驱动的 V 型质子泵（H⁺-ATPase），将 H⁺ 泵入腔内使 pH 降至 5～6，这种酸性环境能促使内吞的配体与受体（如 LDL 与 LDL 受体）解离，含空受体的小囊泡从内体出芽返回质膜重复使用，而配体则随晚期内体（溶酶体成熟的关键中间结构/前溶酶体）运往溶酶体消化 2008/2010/2011/2015/2019·名解。

2. 溶酶体的膜结构三特点、生老病死三阶段与鉴定实验

• 溶酶体（lysosome）：由单层膜围绕、内含约 40 种酸性水解酶类（最适 pH 约 5.0，标志酶为酸性磷酸酶）、形态大小高度异质的囊泡状细胞器，主要行使细胞内消化作用 2015·简答 2017·判断。（植物细胞中无典型溶酶体，由圆球体、糊粉粒和中央大液泡行使类似功能）。
• 为什么溶酶体里装满了凶猛的蛋白酶、核酸酶和脂酶，却不会把自己这层溶酶体膜给消化穿孔？ 这源于华农 2009、2017、2022 年三次考到的溶酶体膜三大结构特点：

1. 嵌有质子泵（V 型 H⁺-ATP 酶）：利用水解 ATP 的能量将细胞质中的 H⁺ 逆浓度梯度泵入溶酶体腔，形成并维持腔内约 pH 5.0 的酸性内环境；
2. 具有多种特异性载体蛋白：将消化降解产生的小分子（氨基酸、单糖、核苷酸）向外转运至细胞质基质供细胞重新利用；
3. 膜蛋白高度糖基化（全细胞所有生物膜中糖基化程度最高！）：朝向溶酶体腔面的高密度寡糖链形成致密的保护屏障，防止溶酶体自身膜蛋白被腔内酸性蛋白酶水解消化 2009/2017/2022·选择！（此外，万一少量水解酶泄漏到 pH 约 7.2 的中性细胞质基质中，其活性也会骤降而失活，构成双重安全锁；但若吸入硅尘刺破肺巨噬细胞大量溶酶体膜导致酶倾巢而出，则会引发矽肺/硅肺或类风湿性关节炎）。

• 溶酶体的“生老病死”三阶段：
• 初级溶酶体（Primary lysosome）：刚从高尔基体 TGN/前溶酶体形成，直径 0.2～0.5 μm，内容物均一、无明显颗粒，仅含处于潜伏未工作状态的酸性水解酶；
• 次级溶酶体（Secondary lysosome）：初级溶酶体与待消化底物囊泡融合而成，既含激活的水解酶，又含正在被消化的底物和消化产物 2009·多选（与吞噬泡/胞饮泡融合叫异噬溶酶体，与自噬泡融合叫自噬溶酶体 2007·名解）；
• 残质体（Residual body，又称后溶酶体）：消化完成后，无法被水解的残留物质（如衰老肝细胞和神经元溶酶体中堆积的未分解脂质氧化物——脂褐质 lipofuscin，即老年斑/脂褐体）留在泡内形成的结构 2026·选择 ⚑华农考过。
• 2010 年华农实验设计真题：“如何用实验证明，一个分离出的特定小泡实际为溶酶体？” 2010·实验
  答法紧扣“酶活 ＋ pH ＋ 超微形态”三步鉴定法：① 酶活性与细胞化学检测：加入特定底物检测标志酶——酸性磷酸酶（acid phosphatase）（如 Gomori 铅沉淀法电镜显色）及蛋白酶、糖苷酶等酸性水解酶活性；② 腔内 pH 测定：用 pH 敏感荧光探针（如 LysoTracker）或微电极测得小泡内腔 pH 约为 5.0；③ 电镜形态与膜特征观察：电镜下为直径 0.2～0.5 μm 的单层膜球形小泡（无尿酸氧化酶结晶），且免疫印迹显示其膜蛋白（如 LAMP1）高度糖基化并含 V 型质子泵 2010·实验。

3. 细胞自噬（Autophagy）：微自噬、巨自噬与分子伴侣介导自噬（2018 / 2024 / 2025 / 2026）

溶酶体消化的底物来自三条途径：内吞作用、吞噬作用、自噬作用 2017·判断。其中大隅良典（2016 诺奖得主）解析的细胞自噬（Autophagy）是华农近年最炙手可热的考点之一！

• 细胞自噬（autophagy）：在自噬相关基因（Atg）严密调控下，细胞利用双层膜自噬小泡包裹自身受损的细胞器、错误折叠蛋白聚集体或长寿蛋白，运送至溶酶体（动物）或液泡（酵母与植物）中融合降解并循环再利用其组成小分子的生理过程 2024·名解 2026·简答。
• 自噬体（autophagosome）：电镜下观察到的、在自噬信号刺激下由内质网、线粒体或高尔基体等膜延伸形成分离膜（隔离膜）并将胞浆待降解成分包裹封闭而成的“双层膜结构”细胞器，随后其外膜与溶酶体融合形成单层膜的自噬溶酶体（autolysosome） 2007/2012/2025·名解。
• 自噬的三种类型（2018 论述 / 填空必背） 2018·论述：

1. 微自噬（Microautophagy）：溶酶体或液泡的膜直接向内凹陷，将细胞质基质中的待降解物质包裹吞入腔内并降解；
2. 巨自噬（Macroautophagy，即通常所说的自噬）：利用内质网、线粒体等膜形成杯状分离膜，延伸包裹胞质底物闭合形成直径 400～900 nm 的“双层膜自噬体（autophagosome）”，随后自噬体外膜与溶酶体膜通过 SNARE 融合为自噬溶酶体，释放内膜与底物被酸性水解酶降解；
3. 分子伴侣介导的自噬（Chaperone-mediated autophagy, CMA）：胞质中的分子伴侣（如 Hsp70）特异识别靶蛋白表面暴露的五肽序列“KFERQ”，将其引导至溶酶体膜受体（LAMP-2A）并直接跨膜转运入溶酶体腔内降解（在持续饥饿供能及氧化损伤保护中起重要作用）。

• 自噬维持内环境稳定与抗逆境胁迫的生理意义 2018·论述 2026·简答：
  ① 常态稳态更新：正常生理下维持低基础水平，专职清除短寿蛋白酶体啃不动的长寿蛋白、蛋白质大聚集体和衰老受损细胞器（如线粒体自噬）；② 饥饿与逆境供能：在营养缺乏（氨基酸/血清饥饿，mTORC1 抑制解除、激活 ATG1 复合物）、缺氧等逆境胁迫下自噬迅速增强，降解自身大分子为细胞提供氨基酸等小分子原料和能量以维持生存；③ 免疫防御與疾病：清除入侵胞内的病原微生物，Atg 基因突变或自噬失调与帕金森病、阿尔茨海默病及肿瘤发生密切相关（注意判断题：自噬绝非单纯病理过程，而是重要的正常生理稳态机制！）。

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 2 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 07 · Vesicle Trafficking⏱ 03:29

00:00So we've talked a lot about how different types of proteins can get into the ER membrane. And we've talked about how soluble proteins can get into the ER lumen. But now we got to talk about how do they move around in the cell.

00:12So we need to start thinking about how that's done, and that's using a process called vesicle trafficking. And you'll see some very similar themes in terms of signal sequences and sorting receptors that really drive this direction in specificity.

00:37OK. ER. So in this particular lecture, we're going to speak about one specific direction, which is the arrows that I showed you that road map in the beginning, which is going from the ER to the Golgi to the plasma membrane.

01:03And this process is often called exocytosis. It just kind of means secretion. So you can not only send cargo out, you can also bring cargo in. And this is using a process called endocytosis. You guys see a common theme happening here?

01:36Vesicles, right? The protein doesn't actually have to cross in this case. The cargo stays inside a vesicle the whole time. And it can do this for both-- I'm showing you soluble proteins, but it can do soluble and transmembrane proteins.

02:00What's an example of a transmembrane protein that might want to come out? Receptor. Exactly. And then we also have examples where soluble proteins need to go out-- maybe like insulin and any cytokines, if you think about immunology a lot.

02:15So there's many, many examples of need things that need to come out. And we also need to take things in. And this is-- what's notable about this is this is also, like I said, it's reversible. It's fast. And there's this repeated cycles of movement.

02:33And it turns out, in order to go from the ER to the Golgi, we need things like small vesicles that bud off with our cargo of interest and then bind to the next destination so that we can go from ER to Golgi. And then same thing's going to happen here.

02:49There'll be a vesicle with a little cargo in it. And so there needs to be a way-- if you imagine, you're budding off a vesicle. You have some cargo. There's a massive amount of space in the cell. There has to be a way for this vesicle to then know where to dock and fuse at the correct spot.

03:07Now, that's something we're going to cover and a little bit, this idea of how different types of signals and receptors set up that specificity. But what's really critical is that we sort of take for granted that this road map exists.

03:21And I mentioned that we didn't actually know this for a long time, how exactly proteins trafficked through the cells.

🎬 Mod 2 · Transport 08 · Intro Pulse Chase⏱ 04:06

00:00So the first thing that was done to really get at this was actually started by a scientist named George Palade, who was actually a really well-trained electron microscopist. So he could see, we could see the organelles.

00:13And he needed some way to actually find proteins and track them through their movement. And so we ended up using a technique called pulse chase. We need some way to track newly-synthesized proteins. So step one is to label newly-synthesized proteins.

00:48And so let's say at time 0.0, we have our cell. What people did is they would actually add a radioactive amino acid, in this case, leucine. And they would maybe add this on for a few minutes. And this was considered the pulse, because just for a few minutes, you're providing a radioactive leucine, which will get incorporated into all newly-synthesized proteins.

01:17So now you've marked them all. And the question is, where do they go? And so let's say at time 0.3 minutes, we then flood the system with non-radioactive leucine. And now all the newly-synthesized proteins that start after time point 3 are no longer labeled.

01:35So for three minutes, we've labeled a subset of the proteins we want to track. And so this is why it's called pulse, and this is called chase. And this technique can actually be used in imaging capacity, a biochemical capacity by fractionating your cells.

01:54But at the time, George Palade did this using electron microscopy, because that was basically what he was trained in. So what he did is, after each of these time points he would fix the cells at different times and ask, where is the label?

02:13And then this is because he was so good-- he could look at these EM pictures, even though I can't look at these and see what I'm seeing all the time. But you can see different organelles and different black spots, which refers to the radiolabeled protein that he's tracking.

02:26And they do this at different time points and ask, where are all the black spots at time point x, y, and z? So what he found, if we actually do this experiment, we can see time of chase-- so we do 0 minutes to 120 minutes-- and then we can track what radioactivity are we seeing in different compartments.

02:59And at the time, we didn't know the order. And so what he was able to see was at early time points, he saw it in this early compartment. The majority of the signal was found in the ER. And then where did it go next? Golgi.

03:16So maybe around 15 minutes, we saw a peak for the Golgi. And then at some point, those then had to move into secretory vesicles. This was huge. We had a sense that there had to be some movement, but we didn't know what the order was.

03:43And by doing something that seems kind of simple, but using just really good electron microscopy-- and this was sort of a technique that had been coming around at that moment-- it seemed like a great way to actually visualize the movement of a new set of proteins.

03:57Now, this is just all newly-synthesized proteins. It's not going to be a specific one. But you could imagine the ways that you could potentially do this if you wanted to, for a specific protein of interest.

2.7“开放的监狱”与逆向回收：KDEL 受体、三种包被膜泡与小 G 蛋白开关

现在，让我们回到 MITx 7.06.1x 的课堂。当 George Palade 的脉冲追踪实验揭示了从内质网（ER）到高尔基体（Golgi）再到细胞外的高速顺向膜泡流后，Rebecca Lamason 教授敏锐地指出了一个致命的逻辑漏洞：

"One of the challenges is that vesicle trafficking is fast and reversible... How do proteins that are meant to be in the ER, that are soluble, don't just get shuttled all the way out and secreted — for example, BiP, right? BiP needs to stay in the lumen. It's not attached to the membrane, and yet there is a chance that it could get delivered and secreted from the cell!"

为什么细胞不干脆给 BiP 和 PDI 装一段跨膜螺旋，把它们死死钉在内质网膜上？教授笑着反问："You could say, well, why don't we just tether them into this cell? They have to move around! They have to be able to access different proteins as they're coming."

一、“开放的监狱”：KDEL 与 KKXX 逆向回收机制及 pH 亲和力开关

在内质网这个繁忙的车间里，可溶性驻留蛋白（如 BiP、PDI）的浓度高达毫摩尔级。为了防止它们流失，细胞设置了两道防线（翟中和教材形象地称之为内质网“开放的监狱”）：

1. 第一道防线（主动滞留/排斥在外）：ER 驻留蛋白之间通过弱相互作用加上高浓度 Ca²⁺ 桥连，聚集成庞大的多蛋白网状复合物，因体积过大而很难被挤进出芽的 COPⅡ 转运小泡中；
2. 第二道防线（逃逸抓回/逆向回收）：即便如此，总有一些可溶性的 BiP 或 PDI 分子会顺着液相顺流（bulk flow）混进 COPⅡ 小泡，“越狱”逃逸到高尔基体顺面网状结构（CGN）。怎么把这些逃逸蛋白抓回内质网？

3DER 中的加工与质量监控点击打开 3d.lkhan.ai 交互式三维结构模型↗

Lamason 教授在黑板上写下了那串大名鼎鼎的四肽邮政编码——KDEL（Lys-Asp-Glu-Leu，赖-天冬-谷-亮氨酸）：

• 可溶性内质网驻留蛋白的回收信号：内质网腔内的可溶性驻留蛋白（如 BiP、PDI）在其 C 端都携带有四肽驻留/回收信号 KDEL（或酵母中的 HDEL）。一旦它们逃逸到高尔基体顺面（CGN），就会被驻守在那里的跨膜 KDEL 受体（KDEL receptor） 一把抓住，打包装进逆向运输的 COPI 包被膜泡，押送回内质网腔！
• 内质网膜驻留蛋白的回收信号：如果是镶嵌在内质网膜上的跨膜驻留蛋白（如 SRP 受体）不小心溜到了高尔基体，它不需要 KDEL 受体中转，而是靠暴露在胞质侧 C 端的“KKXX”（两个赖氨酸加任意两个氨基酸）信号序列，直接与 COPI 包被的 α、β 亚基结合，被直接装进 COPI 膜泡运回内质网！

二、三种包被膜泡（COPII、COPI、Clathrin）：塑形出芽与分子开关

解决了货物识别问题，接下来看囊泡本身是如何从平坦的供体膜上鼓出来（budding）的。Lamason 教授在黑板上画了一张平坦的生物膜：磷脂双分子层本身倾向于保持低曲率的平坦状态，要把局部膜强行拗弯成一个直径仅 50～100 nm 的高曲率球形小泡，必须靠招募到膜胞质侧表面的包被蛋白（coat proteins）组装成笼状晶格（basket / latticework）来提供机械弯曲力，并同时把特定货物受体浓缩进芽体中！

在真核细胞（及翟中和教材标准体系）中，参与蛋白质膜泡运输的包被小泡严格分为三大类型：COPⅡ 包被膜泡、COPⅠ 包被膜泡、网格蛋白/接头蛋白（Clathrin/AP）包被膜泡 2014/2017/2022/2024·简答 2025/2026·选择。

3DCOPⅡ 膜泡：ER → 高尔基体点击打开 3d.lkhan.ai 交互式三维结构模型↗

三、小 G 蛋白分子开关（Sar1 与 ARF）如何精准操控包被的“穿上”与“脱下”？

在课堂上，Lamason 教授一再强调："You can't leave the coats on the whole time, or else the process won't continue!" 为什么囊泡一旦从供体膜上掐断出芽，就必须立刻把外面那层厚厚的包被蛋白“装甲”给脱掉（uncoating）？因为如果包被蛋白一直裹在囊泡外面，囊泡膜上的融合蛋白（v-SNARE）就被严严实实挡在里面，根本无法与靶膜接触融合（就像穿着厚棉袄没法泡澡一样）！

那么，细胞是用什么机关做到“需要出芽时在指定膜上快速穿上包被，出芽完成后又立刻脱下包被”的？答案是：单体小 G 蛋白（Sar1 和 ARF）的 GTP-GDP 分子开关循环！

3DCOPⅠ 膜泡与 ER 蛋白的回收点击打开 3d.lkhan.ai 交互式三维结构模型↗

1. COPII 包被膜泡的装配与去装配四步标准过程（2013 名解 / 2022 论述必背）

COPⅡ包被膜泡（COPII-coated vesicle）：从内质网（无核糖体的过渡型 ER 出口位点）出芽、介导新合成蛋白质与脂质由内质网顺向运输至高尔基体顺面（CGN）的包被膜泡，由小 G 蛋白 Sar1、内层 Sec23/Sec24 复合体、外层 Sec13/Sec31 复合体及辅助蛋白 Sec16 共同装配形成 2013·名解 2026·选择。

其四步装配与脱包被逻辑链条严丝合缝：

1. 第①步（Sec12 激活 Sar1-GTP 插膜）：游离在细胞质基质中的 Sar1-GDP（失活态，疏水尾藏在内部）遇到内质网膜上的整合膜蛋白 Sec12（作为鸟苷酸交换因子 GEF）；Sec12 催化 Sar1 释放 GDP 并结合 GTP 转为活化态 Sar1-GTP，构象改变暴露出 N 端两亲性疏水 α 螺旋，直接插入内质网膜的胞质侧单层（这不仅把 Sar1 锚定在膜上，还先撬动了膜曲率！）。
2. 第②步（招募内层包被 Sec23/Sec24 选货）：结合在 ER 膜上的 Sar1-GTP 提供结合位点，招募胞质中的 Sec23/Sec24 异二聚体，形成 Sar1-GTP/Sec23/Sec24 三重复合物（内层包被）；其中 Sec24 亚基负责识别并结合跨膜货物蛋白或货物受体胞质尾部的二酸分选信号（如 Asp-X-Glu），将待运货物选择性富集到出芽区。
3. 第③步（招募外层包被 Sec13/Sec31 成笼塑形）：内层复合物进一步招募 Sec13/Sec31 复合体（外层包被），后者像网格蛋白一样自组装成对称的多面体笼状骨架使膜弯曲成芽；同时内质网膜胞质面的纤维蛋白 Sec16 与已装配复合物相互作用，显著提高包被蛋白的聚合效率。
4. 第④步（Sec23 促 GTP 水解与脱包被）：包被组装与出芽完成后，内层包被中的 Sec23 亚基（作为 GTP 酶激活蛋白 GAP）激活 Sar1 的 GTP 酶活性，促使 Sar1 将 GTP 水解为 GDP；Sar1-GDP 收回疏水螺旋从膜上解离释放，进而引发整个 COPⅡ 包被解聚（去装配），脱下的包被亚基回到胞质循环利用，裸露的膜泡则启程前往高尔基体 2013·名解。

2. COPI 包被膜泡与非水解 GTP 类似物（GTPγS）实验技巧

COPⅠ包被膜泡（COPI-coated vesicle）：由小分子 GTP 结合蛋白 ARF 调控 7 种包被亚基（coatomer）装配形成、介导高尔基体反面向顺面膜囊以及顺面高尔基网络（CGN）向内质网逆向运输的包被膜泡；负责通过识别 KDEL 受体和 KKXX 信号将逃逸的内质网驻留蛋白及膜脂回收至内质网，是补充 COPⅡ 顺向流的纠错机制 2018·名解。

这里有一个极重要的实验考点：因为 ARF-GTP 水解为 ARF-GDP 是触发 COPI（以及 TGN 网格蛋白）脱包被的开关，科学家如果在体外无细胞膜泡出芽体系中加入不能被水解的 GTP 类似物（如 GTPγS 或 GMP-PNP），会发生什么？

加入不能水解的 GTPγS 后，ARF 永远锁定在激活的 GTPγS 结合态，包被可以正常组装并驱动膜泡出芽，但永远无法脱包被！于是细胞匀浆中会堆积大量穿着完整 COPI 包被的囊泡（coated vesicles），无法与靶膜融合，科学家便能用蔗糖密度梯度离心把这些带包被的囊泡轻松分离出来鉴定成分！

3. 网格蛋白/接头蛋白（Clathrin/AP）双层包被膜泡：三腿蛋白、接头四功能与发动蛋白（Dynamin）缢断

当膜泡从高尔基体 TGN 出芽运往溶酶体/内体，或者从质膜内陷进行受体介导的胞吞（如摄取 LDL）时，唱主角的是网格蛋白/接头蛋白包被膜泡（2022 年华农单选题专门考了它在 TGN 出芽时由外到内的层次排列：最外层网格蛋白 → 中间层接头蛋白 → 跨膜受体 → 最内腔面配体/货物 2022·选择）：

• 外层骨架：网格蛋白（Clathrin）
  网格蛋白（clathrin）：构成网格蛋白包被膜泡外层骨架的蛋白质，每个基本结构单位由 3 条重链和 3 条轻链组成特征性的“三腿结构（triskelion，三脚蛋白复合体）”，能在供体膜胞质面自组装成由五边形和六边形构成的多角形（足球状）笼状晶格，拉动生物膜弯曲成芽。
• 中间桥梁与选货专家：接头蛋白（Adaptor Protein, AP1/AP2/AP3 及 GGA）
  为什么拼成笼子的网格蛋白自己抓不住货物？因为它根本不直接接触脂双层和跨膜受体！它必须靠内层的接头蛋白（adaptor protein）牵线搭桥（经典 AP1、AP2、AP3 均为含 4 种亚基的异四聚体，GGA 为单链蛋白；其中 AP1 主要在 TGN，AP2 主要在质膜胞吞）。
  接头蛋白的四大核心功能（必背！）：① 连接外层网格蛋白与生物膜；② 识别并富集跨膜受体蛋白（选货）；③ 稳定包被的组装；④ 决定膜泡分选的特异性（分选特异性最终由被转运蛋白自身的分选信号决定）。
• 颈部切割刀：发动蛋白（Dynamin）
  当网格蛋白把膜拉成一个连着细长脖子的深窝（coated pit）时，它自己剪不断这根膜脖子！这时，发动蛋白（dynamin）：一种具有 GTP 酶活性的胞质蛋白，被招募到网格蛋白包被膜泡的狭窄颈部并聚合成螺旋环，通过水解 GTP 发生构象收缩（并招募脂质修饰酶），像勒紧绳套一样将膜泡颈部缢断，使独立包被膜泡从供体膜脱离。
  经典突变体证据：如果表达不能结合或不能水解 GTP 的发动蛋白突变体（如果蝇 shibire 温度敏感突变体），电镜下会看到质膜上挂满了带有细长脖子的网格蛋白包被膜泡芽，却一个都掐不下来！注意：这种依赖发动蛋白（dynamin）水解 GTP 进行颈部缢断的机制是网格蛋白包被膜泡特有的，在 COPⅠ 和 COPⅡ 膜泡形成中并未发现！
• 脱包被（Uncoating）：膜泡断裂后，除了 ARF 转为结合 GDP 外，还需要细胞质基质中的分子伴侣 Hsp70（在辅助蛋白 auxilin 激活下）水解 ATP 提供能量，迅速剥下外层三腿网格蛋白循环利用，从而暴露膜泡表面的 v-SNARE，为接下来的靶膜融合扫清障碍。

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 3 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 09 · KDEL ER⏱ 04:02

00:00OK, so we've got the order. But we still needed to know, how do these vesicles know where to go? And I've told you a little bit about how the proteins know how to get into the compartments, but how do the compartments themselves actually know where to go?

00:12And how do even the proteins in these spaces potentially dictate that? So one of the challenges that you're going to see again is that vesicle trafficking is fast and reversible. And then I'm going to tell you one example in which the cargo can dictate the return, how it's going to solve this.

00:43And I'm going to tell you another example of how the vesicles themselves can also direct this. So the solution is we need specific signals, again, and receptors, of course. And a really nice example of this is how proteins that are meant to be in the ER, that are soluble, don't just get shuttled all the way out and secreted-- for example, BiP, right?

01:14BiP needs to stay in the lumen. It's not attached to the membrane, and yet there is a chance that it could get delivered and secreted from the cell. And there has to be way to keep it back in the ER. We have ER. We have-- all right.

01:33So what we often see is that proteins that are destined to be secreted are getting rapidly packaged and delivered to the Golgi. But there's not really a clear way to prevent other soluble proteins from getting packaged and moved out, because this process is so fast.

01:52So this is a secreted protein. There's other proteins that are ER-resident proteins that need to stay. And they have to float around. So you could say, well, why don't we just tether them into this cell? They have to move around.

02:12They have to be able to access different proteins as they're coming. And so there has to be some way to make sure that if something happens and they get packaged and delivered to the Golgi, that they get returned, OK? So what we see is that they, ER-resident proteins, actually have a special sequence on their C-terminus that allows them to return back to this space.

02:35And this space, or this sequence, is called a KDEL. It's literally these amino acids at the C-terminus. And so they're distinct from secreted, because they have the sequence, whereas the secreted ones don't. And so what happens is, if this shows up in the Golgi-- zoom in a little bit here-- the Golgi actually has a special receptor that can recognize this KDEL.

03:04And it's called the KDEL receptor. And its job is to grab onto ER-resident proteins and deliver them back into the ER. And so this is actually a really neat trick for us as cell biologists, because it turns out in my lab, one of my students is doing this so she can label the ER.

03:29She's taking an mNeonGreen, or a GFP protein, and fusing it to a KDEL. And now we can actually see the inside lumenal space of the ER. And then same thing, if you have another type of signal sequence, you could mark different places of the cell.

03:44All right, so we know too, with this particular case and in this example, that's telling us, again, necessary and sufficient. If you get rid of the KDEL, this particular protein would get secreted from the cell. And if you move it to another protein, it will then give it this-- confer this specificity for this location.

🎬 Mod 2 · Transport 10 · Budding Fusion⏱ 04:34

00:00So we've got cargo that can dictate this. But like I said, there's also ways in which vesicles can regulate the trafficking to the specific locations. So let's talk a little bit more about how that's actually done, because this is sort of thinking about it on the cytosolically exposed surface, as opposed to something that's lumenal.

00:24And so I said vesicle trafficking is basically the repeated cycles of budding and fusion through each of the different compartments. And so we have this issue where we've got this membrane space here. We need to form a specifically sized vacuole that then is going to reconnect and fuse with this space.

00:55So we start with a flat membrane, and then we need some way to deform it. And one of the key proteins, or sets of proteins, that regulate this are things called coat proteins. Their job is really to help deform this membrane.

01:18And you'll see them-- let me clarify this here. This is the cytosol, and, say, the lumen. They'll get recruited and attach to the membrane space and actually cause it to deform. And this is actually what they look like.

01:37They come in a variety of different flavors, so we've got clathrin, we've got COPII and COPI. And each of them actually form little baskets or latticeworks that help them, through their interactions with each other, sculpt the membrane into this particular bud here.

01:53All right, so now you've got this coat protein, and you've created the vesicle. But now it's floating around in the cell and it needs to make sure it binds to the right location. And you might say, well, they're pretty close to each other.

02:09This isn't a big deal. And sometimes that's OK. But in cases of going from, say, the Golgi to the plasma membrane, that space might be more significant. And so you want to make sure that you specifically deliver the cargo to the right place.

02:21And so the next step of actually delivering this, you need to have some kind of target specificity. And that's through the action of proteins on the surface of the vesicle and here and the combination of surface proteins and fusion proteins.

02:50The Rab proteins we're going to talk about in a minute are our surface proteins, and then SNAREs confer the fusion event. In this case here, where we have different coats, they're actually specific for different locations in the cell.

03:13For example, COPI only works when you're going from the Golgi to the ER. COPII is in the other direction, ER to Golgi. And for clathrin, it's often going from the Golgi to the plasma membrane to endosomes and back. And when we talk about endocytosis, it also brings stuff from the plasma membrane back to endosomes.

03:45All right, so that's one layer of the specificity. Another layer is, each of these proteins come in a huge list of varieties, and they're specific to their targets and their cargo. So Rabs, we have about 60 kinds. You will not be required to memorize all 60, so don't worry.

04:04And there's about 35 kinds of SNAREs. So now we're in a really complex space. How do you actually regulate all of these, the right cargo with the right Rabs and SNAREs, to make the bud in the right location? There has to be some way in which the cell can regulate this and to do this only in specific locations at the right time.

04:26And so it turns out that the way that these are regulated-- and we're going to get a little bit deeper in the details-- is through the action of G-proteins.

🎬 Mod 2 · Transport 11 · Coat Proteins⏱ 01:47

00:00First we'll talk about how to make the coat. And this like I said, is regulated by a set of G-proteins-- let's say this is our membrane that we want to start building something on. In this case, we can have a membrane associated GEF.

00:14This is active GEF that will then activate our G-protein. So we start in the GDP bound state. It's now GTP bound. It's local. It's activating it on the membrane surface. And it can interact with this membrane surface, as well.

00:28And like I said, once the G-protein is active, it can then interact with other things, which then further amplifies this cascade of activity. So you'll often see additional proteins that assemble on the surface. And then those are sort of act like adaptors.

00:47And then those can then further recruit our coat proteins. OK? Now each of our coat proteins has a special G-protein buddy. So we've got COPI vesicles. They seem to prefer Arf. COPII prefers Sar1. And clathrin actually also prefers Arf and another protein called dynamin.

01:25All right. So, again, molecular switch turns this on, allows for the assembly of the coat. And then, we actually can see a GAP come along, turn this off, and that's how you get disassembly of the coat, is through activating that GTPase activity.

01:40So you need to have the cycle happen. You can't leave the coats on the whole time, or else the process won't continue.

2.8膜泡的精准寻靶与融合：Rab 效应器锚定与 SNARE“面包袋扎丝”四螺旋束

现在，脱掉包被的裸露运输囊泡（naked vesicle）正由驱动蛋白（kinesin）或胞质动力蛋白（dynein）沿着微管轨道往前运送。但细胞质里同时有成百上千种不同的囊泡在穿梭，从内质网出来的囊泡怎么知道自己该停靠在高尔基体顺面，而绝不会一头撞上溶酶体或质膜？当它停靠到位后，两层同样带负电、表面裹着水化层的磷脂双分子层又凭什么能克服巨大的静电排斥力融为一体？

在本章最后一讲中，Rebecca Lamason 教授揭开了膜泡运输最后、也最精彩的两道关卡——Rab 蛋白介导的远距离寻靶锚定（Tethering & Docking） 与 SNARE 蛋白介导的近距离物理融合（Fusion） 2014/2022/2025！

"For every Rab, there is a corresponding Rab effector that gives it that specificity. And what this does is it allows the vesicle to actually dock onto the target membrane... Now, this alone is not sufficient to bring the two lipid bilayers together. You need some physical force to actually squish those membranes together! And that's where SNAREs come in... Kind of like the twisty tie on the bread wrappers that you guys get, these proteins actually twist together!"

一、第一重特异性：Rab 小 G 蛋白与 Rab 效应器介导的寻靶锚定

在哺乳动物细胞中，Rab 蛋白（Rab GTPases）：属于单体小 G 蛋白超家族的最大分支（哺乳动物中超过 60 种），作为不同细胞器和运输膜泡表面的分子身份标签，通过 GTP/GDP 循环调控膜泡与靶膜的远距离识别与停靠锚定 2022·论述 2025·选择。每一种细胞器或特定运输囊泡表面都有其专属的 Rab 蛋白作为“身份识别码”（例如 Lamason 教授提到的：Rab1 特异定位于内质网与顺面高尔基体，Rab5 定位于质膜与早期内体，Rab7 定位于晚期内体与溶酶体，Rab8 定位于早期内体与反面高尔基体分泌泡）。这里务必警惕 2025 年华农真题选择题陷阱：Rab 蛋白属于介导囊泡寻靶锚定的单体小 G 蛋白，绝不属于包被蛋白（三类包被蛋白仅为 COPI、COPII 和 Clathrin） 2025·选择 ⚑华农考过！

3D膜泡的运输、锚定与融合点击打开 3d.lkhan.ai 交互式三维结构模型↗

Rab 蛋白驱动膜泡锚定的分子开关循环（论述题标准采分点）2022·论述：

1. 失活态驻留胞质：在细胞质基质中，处于 GDP 结合态的 Rab-GDP 与 GDP 解离抑制因子（GDI）结合，掩盖其疏水基团，保持可溶的失活状态。
2. GEF 激活与异戊二烯基插膜：当囊泡形成时，特定膜上的鸟苷酸交换因子（Rab-GEF）催化 Rab-GDP 换上 GTP 转为活化的 Rab-GTP；Rab-GTP 发生构象改变，暴露出共价连接在其 C 端的疏水“类异戊二烯基团（isoprenoid group，脂锚）”，借此插入转运膜泡的磷脂双分子层中。
3. 结合靶膜 Rab 效应器完成锚定：随着膜泡运抵目的地，膜泡表面的活性 Rab-GTP 被靶膜上特异配对的“Rab 效应器（Rab effector，如丝状系链蛋白或多亚基系链复合体）”识别并结合，将转运膜泡精准捕获并锚定（tethering/docking）在靶膜表面 2022·论述！
4. GTP 水解回收：在膜融合完成后，靶膜上的 Rab-GAP 促使 Rab 将 GTP 水解为 GDP，失活的 Rab-GDP 从膜上解离释放回细胞质基质，循环用于下一轮转运。

二、第二重特异性与融合引擎：v-SNARE 与 t-SNARE 四螺旋束及 NSF 解离

当 Rab 与 Rab 效应器把囊泡拉近到靶膜表面后，囊泡膜与靶膜之间仍然隔着一层水分子。由于生物膜磷脂亲水头部带有负电荷且紧密结合着水化层，两个独立的脂双层要想发生物理融合，必须克服极高的静电能垒，将两膜间距强行拉近到 1.5 nm 以内并排开界面水分子！承担这一“大力士”任务的，正是动物细胞中至少有 35 种以上的 SNARE 蛋白家族 2014·名解。

3D膜泡的运输、锚定与融合点击打开 3d.lkhan.ai 交互式三维结构模型↗

SNARE 蛋白（SNARE proteins, v-SNARE and t-SNARE）：介导转运膜泡与靶膜特异性识别并驱动脂双层物理融合的跨膜蛋白家族，分为位于转运膜泡（vesicle）上的单链 v-SNARE 和位于靶膜（target membrane）上的多链（通常提供 3 条螺旋）t-SNARE 两类；当膜泡锚定于靶膜后，v-SNARE 与配对的 t-SNARE 通过各自的螺旋卷曲结构域（coiled-coil domains）由远及近紧密缠绕，形成极其稳定的反式（trans）SNARE 四螺旋束复合体（four-helix bundle），利用缠绕释放的自由能将两膜强行拉近、挤出界面水分子并引发脂双层融合 2014·名解 2022·论述。

但故事还没有结束！当囊泡膜与靶膜融合成一张膜后，v-SNARE 和 t-SNARE 依然像拧成麻花的铁丝一样死死缠绕在同一张膜上（此时称为失活的 顺式 cis-SNARE 复合体）。正是因为包被蛋白会遮挡膜泡表面的 v-SNARE，所以囊泡在与靶膜进行 SNARE 配对融合之前，必须先完成“脱包被（uncoating）”以暴露出 v-SNARE 2022·论述！而且由于四螺旋束靠大量非共价键维系，稳定得像石头一样，融合后它们绝对不会自发松开（破伤风毒素和肉毒杆菌毒素正是通过专一性切割神经元突触囊泡的 SNARE 蛋白，阻断神经递质胞吐而导致肌肉麻痹的！）。

谁来拆开拧紧的 SNARE 四螺旋束？

• 膜融合完成后，必须由细胞质基质中的六聚体 AAA-ATP 酶“NSF（N-乙基马来酰亚胺敏感融合蛋白）”在衔接蛋白“α-SNAP（可溶性 NSF 附着蛋白）”的介导下结合到 SNARE 复合体上，通过催化“ATP 水解”提供机械能，像液压钳一样将缠绕的四螺旋束强行解离拆开 2014·名解 2022·论述！解离后的 v-SNARE 随回收泡返回供体膜，t-SNARE 留在靶膜原地恢复待命状态。注意能量节点区分：v-SNARE 与 t-SNARE 缠绕引发膜融合的过程本身由蛋白质构象折叠放能驱动，不直接消耗 ATP；真正消耗 ATP（由 NSF 水解）的是融合完成后的“SNARE 复合体解离回收”步骤！

---

🎧 MITx 7.06.1x 课堂原声英文字幕对照（本节共 1 讲 · 带时间轴）点击展开 / 收起 ▾

🎬 Mod 2 · Transport 12 · Rabs SNARES⏱ 02:46

00:00Once you've formed this vesicle, how do you actually then target and specifically fuse? And that's through the action, like I said, of Rabs and SNAREs. Now, Rabs themselves are also G-proteins, so they're going to go through the same cycle of on-off.

00:24And since there's 60 different kinds, each of these Rabs have specific destination and location specificity. So I think Rab1 is specific for the ER, whereas Rab8 is often found in early endosomes. And that's how you can get one layer of specificity.

00:41And so what we see, if we zoom in here, we've got a cell-- or, sorry, a vesicle with an active Rab, so it's in the Rab GTP state. It then needs to have a corresponding effector on the target membrane. And for every Rab, there is a corresponding Rab effector that gives it that specificity.

01:07And what this does is it allows the vesicle to actually dock on the-- or onto the target membrane. So this is your target here. Now, this alone is not sufficient to bring the two lipid bilayers together. You need some physical force to actually squish those membranes together.

01:24And that's where SNAREs come in. So as this thing sits here, we have two important sets of SNAREs. One sits on the vesicle, and it's called a v-SNARE or a vesicle SNARE. And then one sits on this space, called a t-SNARE.

01:51And these are also specific. I said there's 35 different kinds. There's different kinds for different compartments. And what happens is each of these proteins actually have coiled-coil domains that allow them to stick together and interact.

02:04And kind of like the twisty tie on the bread wrappers that you guys get, think about it that way. You can actually-- these proteins actually twist together. And essentially, this twisting action and this forcing the two membranes together is thought to drive fusion, so that you can actually get this vesicle to dock and fuse with this particular space.

02:26What we do know is that the SNAREs are essential for this to happen. But exactly how the lipids mix is still not very well understood, because it's really hard to see the different layers, the different leaflets, mixing together.

02:38But what's most important is just to think about a t-SNARE corresponding to the correct v-SNARE is what's going to drive this particular process.

★本章小结：核心逻辑链

1. 一个共同起点与两大分选分水岭：真核细胞内所有核基因编码的蛋白质，其翻译合成均起始于细胞质基质（cytosol）的游离核糖体。根据新生肽 N 端是否暴露出疏水的内质网信号肽（signal peptide，16～26 aa），蛋白质分选立即分道扬镳为两条基本途径：有信号肽者被 SRP 识别叫停翻译并押送至糙面内质网，走共翻译转运途径（靶向 ER、高尔基体、溶酶体、质膜及分泌到胞外）；无 ER 信号肽者在游离核糖体上完全合成后释放，凭导肽（线粒体前导肽、叶绿体转运肽、过氧化物酶体 PTS）或核定位信号（NLS）走翻译后转运途径（或无任何信号直接留在胞质）。
2. 跨越或绕开膜屏障的物理法则：细胞核与胞质拓扑相通，亲核蛋白凭不被切除的 NLS 经核孔复合体进行折叠态双向选择性门控转运；进入内质网、线粒体（经 Tom20/22-Tom40 与 Tim23/17）、叶绿体（TOC/TIC）的蛋白质必须保持未折叠线性状态进行跨膜转运（共翻译靠核糖体翻译延伸推力；翻译后靠腔内 BiP 或基质 mtHsp70 水解 ATP 驱动布朗棘轮单向拉入；唯独过氧化物酶体 Pex 系统可跨膜输入完全折叠蛋白）；而一旦进入内质网腔或插入内质网膜（拓扑方向由起始/停止转移序列与正电荷在内规则永久锁定），由于内膜系统各腔室与胞外空间拓扑等价，后续全程通过膜泡运输（出芽与融合）转移，再不跨膜！
3. 内质网与高尔基体的精密加工与质控流水线：糙面内质网利用磷酸多萜醇载体将 14 糖前体整体转移至 Asn-X-Ser/Thr 启动 N-连接糖基化，由 PDI 催化二硫键形成与重排、BiP 与钙连蛋白协助折叠；折错的残次品被质量监控拦截并经 Sec61 逆转位至胞质由 泛素-26S 蛋白酶体（五步） 降解（ERAD）；若未折叠蛋白过量堆积抢走 BiP，则激活 IRE1、PERK、ATF6 启动内质网应激（ERS/UPR）。合格蛋白由 Palade 脉冲追踪实验确立的路径进入极性的高尔基体（CGN → 顺面 → 中间 → 反面 → TGN），完成 N-糖链修饰、O-连接糖基化（Ser/Thr 逐个加糖）、硫酸化与蛋白原水解活化，并在 TGN 分流至组成型分泌、可调节分泌或凭 M6P 标志（五步） 运往溶酶体（与胞吞途径的胞内体及双层膜自噬体汇合完成细胞内消化）。
4. 膜泡运输的双向平衡与双重寻靶保险：Sar1-GTP 驱动 COPⅡ（Sec23/24 + Sec13/31）介导 ER → 高尔基体顺向运输；ARF-GTP 驱动 COPⅠ 介导逆向运输，配合 KDEL 受体（高尔基体微酸性结合、ER 中性释放） 与 KKXX 信号抓回逃逸的 ER 驻留蛋白；网格蛋白（三腿结构）＋ 接头蛋白（AP） 在 发动蛋白（dynamin） 缢断下介导 TGN 分泌泡与质膜内吞泡形成。脱包被后，囊泡凭 Rab-GTP 与靶膜 Rab 效应器实现远距离精准锚定，再由 v-SNARE 与 t-SNARE 缠绕成四螺旋束排开水分子完成膜融合，最终靠 NSF/SNAP 水解 ATP 解离回收。

---

★英文名词解释速背卡（2023–2026 全英文名解高频）

---

★易混概念与中英教材口径辨析

---

★华农历年真题章末综合自测