Surface talk: Cells' communication systems 16/12/04. By Giles Newton When two cells meet, they talk to each other through the proteins that stud their surfaces.

We live in a world full of sights, sounds, smells and sensations – a constant flow of information used by the brain to understand our environment and other people around us. Likewise, the cells in our body are part of a talkative community, gossiping and chit-chatting with each other, sending and receiving orders and directions, keeping the tissues and organs running smoothly.

Each cell is surrounded by a fatty membrane, an impermeable barrier to most biological molecules. Communication is therefore carried out by special proteins that cross the membrane: one end sticking out of the cell, ready to receive a signal; the other end sticking into the cell, ready to pass the signal on. It is these proteins that fascinate Gavin Wright, who moved to the Wellcome Trust Sanger Institute in 2003.

"My interest has always been in the interactions between proteins on the surface of cells," he says. "If we want to know how two cells talk to each other, we need to look at the cell surface, where communication between cells begins."

Cell-surface proteins are found in almost all organisms, but vertebrates seem to find them particularly useful. Indeed, about 3000 human genes – roughly 10 per cent of the total – produce such proteins. But which proteins interact with each other? Dr Wright is taking an ambitious, high-throughput approach – termed functional proteomics – to find out who talks to whom.

"Think of it as a telephone conversation, and you're eavesdropping on a person at one end of the conversation," he says. "You might know the identity of that person, and be able to listen to them talk, but you won't understand the full conversation unless you know who they're talking to and what they're talking about. Likewise, if you know that a cell-surface protein is expressed in cells in the brain, you need to find out which molecule that protein binds to, and where that molecule is expressed."

So far, relatively few interactions have been discovered. Dr Wright's plan is to look through an entire genome sequence – the zebrafish genome – to find all the genes encoding cell-surface proteins, then to produce the proteins and pair them to see which interact with each other. Rather than tackling all potential cell-surface proteins at once, they are being subdivided into families, starting with probably the largest family, the immunoglobulins.

"There are probably about 500 different zebrafish proteins with at least one immunoglobulin domain," says Dr Wright. "So what we're doing is a matrix-testing to see whether each protein binds to itself or to one or more of the other 499 proteins. As we test each protein 'home and away', that's 250 000 interactions to investigate."

Brief encounter

For many years, scientists have studied the interactions between proteins inside cells. But in larger, more complex organisms, the space outside cells has a different chemistry and studies of protein interactions are fraught with technical challenges. "It is hard, because these interactions are special," admits Dr Wright.

"And the interactions that we know a bit about, mainly from the mammalian immune system, show that they are incredibly fleeting – they last for less than a second."

This brevity causes headaches for researchers, but is crucial to the dynamic nature of the body's cells and tissues. When two cells meet, hundreds of proteins will come together at once, forming a kind of Velcro. Cells that have to move, such as migrating immune cells or the tips of growing nerves, have only moments to decide whether to carry on moving, to change direction, or to stop.

Dr Wright gets around this problem by bolting several protein molecules together, which strengthens the interactions with their partner. "After that it's merely a question of pushing proteins through the high-throughput system and finding the interactions."

Having found that two proteins interact, Dr Wright then wants to know why. "Just knowing that protein A binds protein B isn't particularly intellectually satisfying. We want to know what happens as a consequence of that interaction, and what this means for the organism. This is where the function part of our project comes in, and we've chosen to use the zebrafish as our model organism."

The zebrafish is an attractive organism for such studies. It is easy to breed and keep, and has transparent eggs that allow the development of the embryo to be followed. And by attaching coloured probes to the messenger RNA produced from a gene (a technique called in situ hybridisation), the cells, tissues or organs in which the gene is expressed are revealed in the translucent embryo.

"If we find that two proteins bind to each other, we can look to where the genes are expressed, and get a better idea of where the interactions are occurring in the animal," says Dr Wright. "If one is expressed in nerves and another in muscles, for example, that may be very interesting. We'll also use loss-of-function techniques to see what happens if the fish does not produce one or more of the proteins."

It is the union of high-throughput proteomics and detailed functional studies that, Dr Wright suggests, will bring many insights into the communication between cells. "If we put all the data together, we can be reasonably confident of the function of an interaction between two cell-surface proteins. Vertebrates, whether fish or humans, have lots of these proteins, so they look to be extremely important for complex animals. I think we will have lots of interactions to test."

Image credit: Nicoletta Baloyianni