| Picture: 2012 Accomplishment By A Senior Scientist Award
Gunnar von Heijne.
Photo by Max Brouwer
Perhaps it all began with the French lessons. As a a young PhD student in theoretical physics at the Royal Institute of Technology (KTH) in Stockholm, Gunnar von Heijne decided, on whim, to brush up on his rusty, schoolboy French. He took a few lessons and also subscribed to the French popular science magazine La Recherche.
Flicking through its pages, he came across a short article on protein secretion and the signal hypothesis, the mechanism that describes the way secretory proteins cross a membrane.
At the time, the late-1970s, very little was known about this process, but some ideas were beginning to emerge. For example, it was thought that a so-called signal peptide---a short chain of amino acids---at the end of the protein carried the signal that determined how the proteins are transported out of the cell.
The article confused him, however. It showed a diagram of a hydrophobic signal peptide squeezing through the similarly hydrophobic membrane. "That didn't make sense to me. The hydrophobic peptide ought to become anchored in the membrane,"he says.
The puzzle piqued his interest. He solved it by calculating the energetics of a polypeptide chain passing through lipid bilayer, which he published in 1979. This work by a theoretician created ripples in a field dominated by experimentalists.
And so began the career for which he now receives the Accomplishment by a Senior Scientist Award from the International Society for Computational Biology (ISCB). "Gunnar is one of the big stars of our field,"says Burkhard Rost, president of the ISCB. "He is one of the few who completely change the field using computational methods."Polypeptide energetics was only the start, however.
By the early 1980s, molecular biologists had begun to determine the sequence of amino acids in the signal peptides from different proteins. However, little had been done to study the properties of signal peptide sequences as a group.
von Heijne changed this. He began comparing the sequences, looking for recurring patterns that might help to identify them. "I looked at 20 to 30 signal peptides. Once you did that, some clear patterns emerged that had not been seen before,"he says.
He found that small, uncharged amino acids tended to occupy certain positions in signal peptide chains, the -3 and -1 positions. It is at this site that the signal peptide is later cleaved from the protein once it has passed through a biomembrane. This pattern has since become known as the (-3, -1)--rule.
"Nowadays you would say this was a very trivial bioinformatics study,"he says modestly. However, this was an important discovery and von Heijne's paper has since become one of the most highly cited in the field.
He then used the newly discovered patterns to make predictions about proteins. For example, it became possible to create an algorithm that would take a protein sequence and predict whether it had a signal peptide at the end.
Initially, that was not very useful. When molecular biologists sequenced a gene or messenger RNA, they generally knew what they were working on; whether it would have a signal peptide on the end or not.
But that changed when sequencing became faster and biologists started to sequence things they didn't know much about. "The algorithms have continually improved and are now extremely useful,"he says.
Secretory proteins have to move across a lipid bilayer through a molecular machine called a translocon. The signal peptide guides the ribosome that makes the protein, towards the translocon. This triggers the opening of this protein-conducting channel through the membrane.
But other types of protein only make the journey partway, becoming embedded half in and half out of the membrane. These so-called membrane proteins use the same translocon machinery as the secretory proteins. "So it was a natural step to start looking at these membrane proteins next,"says von Heijne.
The part of the protein that ends up in the membrane is very different to the parts outside exposed to water. This transmembrane section must be much more hydrophobic. So the trick to predicting which parts of a protein become embedded in the membrane is to look for the segments that are most hydrophobic.
Once you know the transmembrane segments, an interesting problem is to determine how the protein becomes woven into the membrane. For example, if it has four hydrophobic sections, there are two ways in which it can be arranged in the membrane: with the termini pointing either in or out. But which orientation should the protein take?
"We discovered a very simple principle that determines this,"he says. The regions that connect the transmembrane segments contain positively charged amino acids, which give them an electric potential. The simple principle is that the segments with the greatest number of positive charges end up inside the membrane, an idea that has since become known as the "positive inside rule".
"This is very important work and provides some of the best data on membrane proteins,"says Alfonso Valencia, chair of the ISCB awards committee.
In the late 1980s, von Heijne began to realise that he could gain significant insight into these and other problems by doing experiments rather than just theory work. So he set up his own lab. "I trained as a chemist so I wasn't a complete novice in a wet lab,"he says.
This first idea was to see whether it was possible to make proteins that inserted "upside-down"into the membrane. He could show that by changing the location of the positively charged amino acids in a protein, it is possible to make it take up the opposite orientation.
This link between his theoretical and practical work has been important for him. Bioinformatics studies often throw up patterns that may or may not have biological relevance. "The only way to determine whether they are important is to do the experiments,"he says.
"It's hard to overstate the significance of von Heijne's work. Membranes and transmembrane proteins are the gates and gatekeepers to our cells; they determine what gets in and what stays out,"explains Rost. "That's why around two-thirds of drugs target membrane proteins."Understanding the structure of transmembrane proteins provides crucial insight into how cells work and is also useful for future drug development. "That's why the methods developed by Gunnar are so important,"says Rost.
To continue his work, von Heijne set up the Stockholm Bioinformatics Centre at the beginning of the millennium. And today, von Heijne runs the Centre for Biomembrane Research in Stockholm, where he has brought together computational, modelling, and experimental groups. Few places can boast the same breadth of experience under one roof.
Throughout this time, von Heijne has maintained an impressive work--life balance as a scientist, a husband, and a father. He says that's been possible, at least in part, because he was working in a new field with few competitors. "I never felt stressed that we'd be scooped. I work hard but not crazily ."Others clearly admire his positive approach, which he combines with a relaxed attitude. "He also looks ten years younger than he has any right to!"says one envious colleague.
For a while in the 1980s, he spent half his time working as a science journalist for the Swedish National Radio. "You decide on Monday what you broadcast on Friday so there is immediate feedback, which has a good pulse to it,"he says.
But for von Heijne, doing science is more satisfying than reporting it. "Radio stories have a short half life; they're on air, then they're gone,"he says. "The rewards in science are greater and longer lasting."It's surprising how far schoolboy French can take you.
This article is excerpted from the May 2012 issue of PLoS Computational Biology. To link to the full journal article please visit www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1002535