You are here

3D Information from just the Protein Sequence

Proteins are all around us in the biological world, like the hidden cogs that turn the hands of a watch; it is primarily proteins that form the structural and functional components of life as we know it. Just like beads on a string, proteins are made up of smaller units called amino acids that are joined together in a specific sequence. This specific sequence is unique to a protein, and gives rise to both the shape of the protein and its associated function.  While technology has allowed us to easily identify a protein’s sequence i.e. the sequence of amino acids it contains, what biologists are often concerned with is the three dimensional structure of a protein.

However, predicting the structure of a protein from its sequence is easier said than done. Prof. Raghavan Varadarajan and his lab at the Molecular Biophysics Unit, Indian Institute of Science have developed a method to accurately predict which amino acids in a protein that are in close proximity with each other, thereby simplifying the task of predicting the 3D structure of a protein.

“Sequences (of proteins) are identified at a higher rate than structures. Structures are typically determined by X-ray crystallography or NMR, the process can take a long time, and the protein needs to be amenable to purification and crystallization”, explains Prof. Varadarajan. A large number of proteins are still not understood completely because they are either difficult to purify at high concentrations or do not form crystals of the high quality needed for X-ray based structure determination.

 Prof. Varadarajan’s group have developed a method called ‘saturation suppressor mutagenesis’ to figure out which amino acids closely interact with each other in a protein while it is functioning inside the cell. Most often, amino acids that do not play any role in the actual chemistry of a protein’s function act as structural scaffolds, like the walls and pillars of a house. It is known that mutations in one amino acid can sometimes be compensated for by corresponding mutations (called suppressor mutations) in the amino acids spatially close by, thereby preserving protein function. For example, mutation from a small amino acid to a large amino acid can be compensated by a large to small suppressor mutation of an amino acid in close proximity. Thus balance is restored and the protein is functional again. This is like the carpenter cementing a larger brick next to a smaller brick to make sure that the wall is not filled with gaps because some of the bricks were broken.

While the concept is not new, traditional studies on suppressor mutations have isolated a few such mutations for a given protein . For the first time, Prof. Varadarajan and his team have devised a procedure to comprehensively search for suppressors, covering the entire sequence of amino acids. They did this by creating a single inactivating mutation in a protein, followed by sequential and methodical mutation of every other amino acid in the entire protein in combination with the inactivating mutation. 

Using the wall analogy, Prof. Varadarajan’s group started with a wall containing a single empty space where a large brick should have been. They know there is an empty space, but because the bricks were not cemented in the order they were delivered, they did not know which brick is next to the empty space. They went on to replace every brick with a larger brick one at a time, checking each time to see if the wall became stronger. In this way, they were able to find which bricks were in close proximity to the empty space. Using mutations in several places, the researchers were able to piece together proximity information for multiple pairs of amino acids and were thereby able to infer the 3D structure of the protein, with no prior knowledge of the actual structure.

Going further, the group was able to correct a previously solved structure that was ambiguous. Because the entire process is done inside the cell and not in a test-tube, Prof. Varadarajan’s methodology can help resolve ambiguities in existing structures and help predict functional structures where none are available. “Many drugs are designed based on the 3D structures of their protein targets, hence accurately predicting structures, especially for complexes of several proteins where crystallization may be difficult, becomes important”, he says. With bioinformatics giving us more and more sequence information, Prof. Raghavan Varadarajan’s research will go a long way in converting more of those sequences into meaningful 3D protein structures, impacting both basic research and drug discovery.

About the author: Prof. Raghavan Varadarajan is the Chairman, Molecular Biophysics Unit, Indian Institute of Science, Bangalore.

About the paper: The paper, “Residue proximity information and protein model discrimination using saturation-suppressor mutagenesis” appeared in the journal eLife.