Edward Lemke opens doors for molecular biologists

Introducing a single reactive artificial amino acid into a protein opens up a whole new world for molecular biologists – we speak to one of the scientists who made this possible and find out more...

Edward Lemke is one of a group of scientists from the European Molecular Biology Laboratory (EMBL) in Germany who have developed a new method to label any protein of their choice. The method – based on click chemistry – enables them to label even rare proteins very precisely for optical imaging.

This month we caught up with Edward to find out more about artificial amino acids, labelling proteins and the ‘click’ the reaction.

What is the purpose of labelling proteins, why is it necessary?

Labelling proteins allows us to find and follow them inside cells. A single cell, typically 10 micrometers in diameter – 10 times smaller than a strand of hair – contains thousands of different types of proteins that can assemble into larger molecular machines which carry out vital tasks, keeping the cell alive. What’s more, the same protein can perform different tasks at different times or in different regions of the cell. So, a cell is not a homogenous bag of little molecules, but rather a highly regulated, interconnected system. In order to understand this network, it is not enough to just know what proteins a cell contains; we must understand what each protein is doing when and where.

Unfortunately, proteins are too small to easily tell them apart. So, the general strategy is to attach something to the protein, like a label, that allows them to be easily identified by their colour, e.g. using a microscope. The most popular approach is to install fluorescent markers on a protein of interest, which glow when you shine non-invasive light on your sample. Their fluorescence can be easily detected under a light microscope. This way, the invisible is made visible, and we can monitor when a certain protein is doing something just by looking for the fluorescent signal.

However, many fluorescent tags (like fluorescent proteins) are fairly large, which means they might interfere with the function of the protein they’re being used to study. Thus, part of the field of chemical biology is trying to develop surrogates that are smaller and ideally brighter than existing tags.

How did you label the proteins?

Our labelling strategy is based on a ‘click’ reaction, in which, as the name implies, two molecules ‘click with each other and permanently connect. One of the most powerful ‘click’ reactions is between two chemical groups called a strained alkyne and an azide. Although they are artificial, these chemical groups are not toxic. We developed a technology in which one of these partners, the strained alkyne, is genetically encoded in the host cell. This allows us to create a custom-made E.coli bacterium, in which the strained alkyne is part of an artificial amino acid which replaces one of the building blocks of any protein of our choice. As this modification is very small – the artificial amino acid is basically the size of the natural one it replaces – it typically does not interfere with protein function. The slightly modified protein is hardly distinguishable from the native form, and even detecting this modification typically requires advanced and very sensitive mass spectrometry techniques. Nevertheless, this small change is enough to make the protein prone to react with azides, even if only a few of them are present. As azides do not occur naturally, any azide-containing reagent that we can artificially feed to the cell or the organism will permanently ‘click’ to our protein of choice.

The process involves using an artificial amino acid - why does it need to be artificial and how did you make it?

All proteins present in living beings are made from the same 20 amino acids. These natural amino acids are combined in a specific sequence for each protein, and each amino acid can appear several times in a proteins as most of them are several hundred amino acids long. Consequently, natural amino acids cannot be used to install labels and mark only one specific protein. However, we can overcome this problem by using an artificial amino acid, which is produced synthetically by us in the lab. Our artificial amino acid looks very much like the natural amino acid lysine; the major difference is that we modified its side chain so that it contains a strained alkyne.

We introduced this modification by developing a protocol to synthesise this compound in high purity and large amounts. A major plus for using an artificial amino acid is that we can essentially fool the host translation machinery and piggyback on the natural system. The artificial amino acid is so similar to the natural amino acid that we can trick the cell’s protein-producing machinery into making the modified protein for us, without setting off any of the damage-control processes that normally kick into action if a faulty protein is produced. This is much easier and more efficient than us trying to incorporate the artificial amino acid into the protein ourselves. And in terms of labelling, which is our goal here, we are only adding a few more atoms that interfere much less with a protein than attaching a large fluorescent tag (like a fluorescent protein) to it, so this will hopefully enable us to follow proteins as they perform their natural tasks unhindered.

How do you hope to use this technique?

This approach is already opening up plenty of opportunities. Proteins can now be encoded with the ‘clickable’ functionality and then readily clicked to small fluorescent dyes. Those dyes are stable enough that even a single molecule can be visualised using state-of-the-art techniques and so we can observe individual protein dynamics and function directly. Thanks to this approach, we can now unambiguously identify proteins even when they’re in complex mixtures – within a cell or living organisms.

In other words, we can watch proteins in action in greater detail, with greater precision and certainty by making only a tiny change, and this will facilitate our understanding of how they work. For example, the machinery that safely stores our DNA in a condensed form so that it fits within a single cell (Chromatin) is made from a huge number of proteins. One can now aim to visualise where individual parts of this DNA-condensing machinery assemble and at what time, which will help us to understand how our genetic material can also be safely duplicated during reproduction.

What's next in this line of research?

We developed this system in the bacterium E. coli, so the most important step is to transfer the system to mammalian cells, like our own. Using this new strategy, we can then see where and when proteins are at work within a living organism. Furthermore, we can use our technology to ‘click’ tools to the protein that allow us to control the protein with light. Remotely controlling the cellular machinery can be used to trigger specific actions and thus simplify biomedical research on complex processes. Most pressing, however, is the need to move from ‘black and white TV’ to ‘color TV’ – finding a way to incorporate different artificial tags at different places, to finally watch movies of the molecular building blocks of a living cell at work.