袁岚峰对话诺奖得主约阿希姆·弗兰克（一）神器冷冻电镜究竟是什么？诺奖发明者亲自解说 | 科技袁人

关注风云之声

提升思维层次

导读

最近，世界顶尖科学家协会与字节跳动帮我联系到一位诺贝尔奖得主：约阿希姆·弗兰克，他和另外两位科学家雅克·杜波谢和理查德·亨德森因为对冷冻电镜技术的贡献获得了2017年的诺贝尔化学奖。

视频链接：

西瓜视频：https://www.ixigua.com/6904546032164700683本视频发布于2020年12月11日，播放量已超三百万

我们的节目经常有我跟著名科学家的对话。最近，世界顶尖科学家协会与字节跳动帮我联系到一位诺贝尔奖得主：约阿希姆·弗兰克，他和另外两位科学家雅克·杜波谢和理查德·亨德森因为对冷冻电镜技术的贡献获得了2017年的诺贝尔化学奖。

说到冷冻电镜这个词，许多人可能是，虽然不知道原理，但是听说过，这是一种解析生物分子结构的神器。

例如中国科学家施一公和颜宁等人，就用这种技术取得了很多成果。那么冷冻电镜究竟是什么？它为什么这么神奇？弗兰克教授作为一位1940年生的长者，对中国的科学事业有什么建议？下面我就来和弗兰克教授连线。

【约阿希姆·弗兰克与袁岚峰对话】

袁岚峰：

So dear Professor Frank, Dear Prof. Frank, I heard that you just took a surgery several days ago, yet you still accept this interview. We are very grateful, and hope you recover soon. So first, how to pronounce your first name?

亲爱的弗兰克教授，我听说您前几天刚做了一个手术，然而您仍然接受了这个访谈。我们非常感谢，也祝您早日康复。首先，请问您的名字如何发音？是约阿希姆吗？

约阿希姆·弗兰克：

Joachim. Joachim.

是约阿希姆

袁岚峰：

Joachim Okay? I see, thank you.

约阿希姆，好的，我知道了。

The reason of your Nobel Prize was your contribution to developing cryo-electron microscopy, i.e., cryo-EM. I was not familiar with this technology before, but I heard this name long before, because some famous Chinese scientists, e.g, Shi Yigong and Yan Ning, employed cryo-EM to analyze the structures of many biomolecules.

您是因为研发冷冻电镜获得诺贝尔奖的。我以前对于这项技术并不了解，不过早就听说过这个名字，因为中国有一些著名的科学家如施一公和颜宁用这项技术解析了很多生物分子的结构。

Recently I read some reviews of cryo-EM, so my understanding to it has increased a lot. I got to know that there are many interesting stories, e.g., each one of you and the two scientists sharing the Nobel Prize with you, Prof. Dubochet and Prof. Henderson, solved some problems in this technology, respectively. I am curious, how would you introduce cryo-EM and your contributions to the public?

最近我看了一些对冷冻电镜的综述，对它增加了很多了解，同时发现这其中有很多有趣的故事，例如您和跟您同时得奖的杜波谢教授和亨德森教授分别解决了什么困难。我很感兴趣，您会如何向公众介绍冷冻电镜技术和你们的贡献？

约阿希姆·弗兰克：

Okay, one has to know that this all started in the 70s, 1970s and 80s. And at the time that I that I started, there was no cryo-EM. In fact, the press release of the Nobel foundation says that I got the award for work mainly between 1975 and 1986.

好吧，人们应该知道，这一切都始于70年代、1970年代到1980年代。当时还没有冷冻电镜。事实上，诺贝尔基金会的新闻稿说，我主要是由于1975到1986年间工作而获奖。

I did not use cryo-EM, it did not exist at the time. So my contributions really have something to do with the mathematics and computation of how to put images together in order to form a three -dimensional image.

我那时还没有用冷冻电镜，它当时还根本不存在。所以我的贡献主要与数学和计算有关，如何把图像组合起来，形成一个三维的图像。

And the main idea is really that the specimens, the samples from which the image come from, is not crystalline. The molecules are not put together and ordered the arrangement, but rather, the individual molecules are free to move in all different directions.

主要的想法是，这些用于图像的来源的样品并不是结晶的。这些分子不是以一种有序的方式排列在一起的，而是而是各个分子可以自由地向各个方向运动。

So the main idea is really to obtain structure from such an arrangement of molecules that are in completely random orientation, that's the main idea.

所以主要的想法就是从这样完全随机取向的分子中，得到结构。

And one has to know that electrons damage the molecules very strongly, and so it doesn't work just by looking at a single molecule because by the time you get a sharp image, the molecule has disappeared.

我们还必须知道，电子对分子的损伤非常强烈，所以仅仅观察看到单个的分子是行不通的，因为当你得到清晰的图像时分子早就消失了。

So all techniques using electron microscopy need to be concerned with keeping the radiation damage very low.

因此，所有使用电子显微镜的技术都需要注意保持极低的辐射损伤。

And that can only be done by distributing the radiation load or distributing the electrons set up being used over a very large number of molecules, so each molecule that is being image contributes just a tiny amount to the overall information.

这只能通过分配辐射负载即使用的电子分布在大量的分子上来实现，所以每个被成像的分子对整体信息的贡献只是很小的一部分。

So today we would call all these techniques deep data analysis, yeah at the time it did not exist, the past, you have to think about tens of thousands, hundreds of thousands of molecules, and each contributes a very, very faint image.

所以今天我们将所有这些技术称为深度数据分析。（袁：像现在人工智能中的深度学习。）是的，在当时它并不存在。过去你要考虑数万个，几十万个分子，每一个都只贡献一个非常非常微弱的图像。

And now you have to make sense of that information. And one of the biggest challenges is to find the orientation of the molecules to each other.

你现在必须让这些信息产生意义。其中最大的挑战之一是找到分子之间的取向。

That information doesn't come from anywhere. You have to find it; you have to find it computationally.

这些信息完全不是直截了当的，你必须通过计算找到它。

袁岚峰：

Yeah, it sounds to me that there is a famous story. That some blind people are touching an elephant, and they say this elephant is like a wall or like a of fun or something like that. So your technique sounds like a contribution to find a way to reconstruct the elephant from the blind people.

在我听起来，这就像是那个著名的盲人摸象的故事。一些盲人在摸大象，其中有的人说这只大象像一堵墙或是扇子等其他之类的东西。您对这项技术的贡献听起来就像是找到了一种从这些盲人的描述中重建大象图像的方法，对吗？

约阿希姆·弗兰克：

You could think of it this way, right. So the molecules that when I started developing the technology with, were prepared by what's called negative staining.

是的，你可以这样理解，对。所以当我开始开发这项技术的时候，这些分子是通过所谓的负染色来制备的。

So they are embedded in a heavy metal salt and air dried on the grid. But this is a very bad and imperfect way of preparing molecules, and only later, through the work of Jacques Dubochet, the cryo-EM techniques came into use, and that is a much better way of keeping molecules in their native environment.

它们被嵌入一种重金属盐中，并在网格上风干。但是，这是一种非常糟糕和不完美的分子制备方法，直到后来，通过雅克·杜波谢的工作，低温技术才开始使用，这是一种好得多的把分子保持在它的原生环境中的方法。

袁岚峰：

I see, so nowadays you don't use the heavy metal salt? It’s not needed for now?

我懂了。那么现在你们不用重金属盐了吗？现在不需要了吗？

约阿希姆·弗兰克：

It's only used sometimes when one gets started with a new structure, and when one wants to have essentially the first flavor of you know, what does it look like? This is much easier to use negative staining because you see the shapes very clearly immediately.

现在只是有时会用，当人们开始研究一个新结构的时候，当你想品尝它的初次味道时，意思是，它看起来像什么？这时使用负染色方法会更容易，因为你可以立刻看到非常清楚的形状。

You don't see it as faint images, so that's a typical start. People just do a very quick and dirty rule on molecules with negative staining. It's still popular, but it's only used right at the beginning.

它对你不显示为模糊的图像了，这是一种典型的开始。人们用负染色这种方法，只是为了对分子做一种快速而粗糙的研究。这种现在仍然很流行，但只是用于研究工作的开始阶段。

袁岚峰：

I see, so you said the word is negative staining or what is the word staining?

您说道的staining是什么意思呢？

约阿希姆·弗兰克：

Oh, staining, that's a word which is used to create contrast.

staining就是染色，就是去创造差别。

袁岚峰：

Oh, I see, S-T-A-I-N, stain. Okay.

我懂了，是S-T-A-I-N, stain。

约阿希姆·弗兰克：

And it's called negative staining because in the end, when person see the molecule one only sees the stain, and when see the molecule, the molecule is negative, it stands out by excluding stain. That’s how it works.

它被称为负染色是因为最终，当人们看到分子时，人们只看到染色的背景，分子的图像是负的，它通过排除而突显出来。这就是它的工作原理。

袁岚峰：

I see, I see. So what you really see is the staining, and the molecule is some voids in the staining.

我懂的，我懂的，你真正看到的是染色背景，而分子是染色背景中的一些空隙。

约阿希姆·弗兰克：

So then, when the other technique came along, you have to think of the fusion of two techniques.

所以，当另一种技术出现时，你必须考虑两种技术的融合。

One technique didn't use a cryo because it was not available, and then the other one was a technique where cryo was developed, but not with Single particle idea.

其中一种技术没有使用低温，因为当时还没有低温技术；在另一种技术中，发展了冷冻技术，但不是用于单粒子的想法，他们只把它用于晶体样品。

But they are only applied to a crystalline specimen because that's what people knew, they thought structure only makes sense if it is a structure of crystals, you know.

它们只适用于晶体标本，因为当时人们知道的只有这么多，他们认为结构只有在是晶体结构的情况下才有意义。

袁岚峰：

Yes, yes, this applies to XRD, the X-ray crystallography.

是的，是的，这适用于X射线衍射，X射线结晶学。

约阿希姆·弗兰克：

Yes, yes, that's really it is. And you mentioned Yigong Shi before and Ning Yan, they are both X-ray crystallographer, they learned pretty late in their career how to use cryo-EM.

是的，的确如此。你之前提到过施一公和颜宁，他们都是X射线晶体学家，他们在他们职业生涯相当靠后的时期学会了如何利用冷冻电镜。

袁岚峰：

Yes, they mainly use XRD before this, right?

是的，在这之前他们主要使用X射线衍射，对吧？

约阿希姆·弗兰克：

Yeah, only, exclusively, there was no cryo-EM, or it was there but it didn't reach a higher resolution.

是的，只用X射线衍射，当时还没有冷冻电镜，或者说，当时有冷冻电镜，但它还没有达到更高的分辨率。

袁岚峰：

So, thank you very much. So the third question is that I heard that based on your works, the Chinese scientist Cheng Yifan made one more important contribution to cryo-EM, i.e., introducing direct electronic detectors. By doing so, he for the first time got the atomic resolution structure of the TRPV1 ion channel, which facilitates our understanding to the question “why we feel spicy”. Could you please introduce this interesting story and more important discoveries made by cryo-EM? How is the situation of this field for now and future?

我听说华人科学家程亦凡在你们工作的基础上，对冷冻电镜又有一个重要的贡献，就是引入直接电子探测器。由此他第一次获得了原子分辨的TRPV1离子通道结构，这有助于我们理解“为什么会感到辣味”。您能不能介绍一下这个有趣的故事，以及其他的由冷冻电镜导致的重要的发现？现在和将来冷冻电镜的发展状况如何？

约阿希姆·弗兰克：

Right, right. Yeah, I also know Yifan Cheng very well. It's not correct that he introduced the cameras, he had no contribution to this. The cameras were developed by three academic labs. Two of them were in the United States and one was at the MRC in Cambridge, and they've worked respectively with three different commercial companies.

是的，我也很了解程亦凡。但是说他发明了相关摄像机是不对了，他对此没有贡献。这些相机是由三个学术实验室研制的。其中两个实验室在美国，另一个实验室在剑桥医学研究委员会，他们分别与三个不同的商业公司合作。

And then all these commercial companies came out with the first working commercial product in 2012, and Yifan Cheng, because of his association with one of the academic developers, was able to use this camera immediately when it came out.

所有这些商业公司在2012年推出了第一款实用的商业产品，而程亦凡因为和其中一位学术开发机构的关系，可以在这款相机面世的时候就能够立即使用它。

And then his group made very important contributions to the use of the camera, to the kind of data processing that needs be used. And these programs are still used today.

他的团队在相机的使用方法以及对需要使用的数据处理方法方面做出了非常重要的贡献，这些程序至今仍在使用。

So this was really he excelled in making use of the technology very, very fast. And so when in 2013, he published two Nature articles, there was a run on cryo-EM. There were many people never heard of cryo-EM and so forth, and then all of a sudden they said, oh my god, we can even look at channels.

他的杰出之处在于，非常非常快地利用这项技术。在2013年，他在《自然》杂志发表了两篇文章，引领了冷冻电镜的风潮。当时有很多人从来没有听说过冷冻电镜之类的技术，然后他们突然说，天哪，我们甚至可以看清通道。

Okay, it is difficult to use X-ray crystallography to image channels because it's very hard to crystallize them, but to prepare them for cryo-EM, it is possible to make the molecule feel like they are still in the membrane.

是的，人们很难用X射线晶体学来给通道成像，因为通道很难结晶。但是利用冷冻电镜，就可能让分子认为自己还在分子膜中。

So when there are techniques that to surround the molecule with a lipid layer, so the molecule doesn't even know that it's not in a membrane, and so it becomes a single particle and one obtains structure even though the molecule is not in its native environment, you know, it's like a fake environment that is brought in.

所以，当有技术用脂质层包围分子时，分子甚至不会发现它其实并不在膜中，所以它变成了一个单一的粒子。这样即使分子不在其自然环境中，我们也能获得结构，你知道，这就像是引入一个假环境。

So he made use of this new camera with new programs that he developed, and at the same time he solved the problem of embedding the molecules in this fake environment. And that makes the molecule feel at home okay. And he came out with atomic structure.

所以他有效利用了新相机和他开发的新程序，同时他解决了在这个假环境中嵌入分子的问题。这让分子感觉还在自己家里似的。然后他想到了原子结构。

So people just went wild everywhere and that's when, and the main company that profited from it, was FEI, which is now called Thermo Fisher. Thermo Fisher took over recently, but FEI profited from it because there was a run on these microscopes, you know, XXX bought eight of them and some were delivered to Saudi Arabia, where they were put in the desert, they are distributed everywhere.

然后各地的人们都激动起来了。从中获利的主要公司是FEI，它现在叫赛默飞。赛默飞最近接手了FEI公司，但FEI公司从中的确获利了，因为这些显微镜炙手可热，你知道吗，XXX买了8台，还有些被送到沙特阿拉伯的沙漠里，这些显微镜现在遍布全球。

扩展阅读：

作者简介：袁岚峰，中国科学技术大学化学博士，中国科学技术大学合肥微尺度物质科学国家研究中心副研究员，科技与战略风云学会会长，“科技袁人”节目主讲人，安徽省科学技术协会常务委员，中国青少年新媒体协会常务理事，入选“典赞·2018科普中国”十大科学传播人物，微博@中科大胡不归，知乎@袁岚峰（https://www.zhihu.com/people/yuan-lan-feng-8）。

  背景简介：视频发布于2020年12月11日《神器冷冻电镜究竟是什么？诺奖发明者亲自解说》（https://www.ixigua.com/6904546032164700683）。

责任编辑：孙远