Reflexions, the University of Liège website that makes knowledge accessible


Revealing the 3D structure of proteins

10/14/13

At the present time, it is still difficult and costly to obtain a representation of the three-dimensional structure of proteins. But such a representation contains important indications about their function. That is why scientists have been trying to develop computer-based methods of making predictions about the properties of certain proteins that are faster than current methods. 
Julien Becker, Francis Maes and Louis Wehenkel are using their automatic predictive tool to predict the formation of disulfide bridges (connections that form between the sulfur atoms of two amino acids called cysteines) that are present within a certain protein. Their work has led to the creation of a predictive tool known as x3CysBridges, which allows them to calculate the probability that the cysteines of a single protein will form disulfide bridges. The international scientific community has already begun to use this predictive tool.

protein strucutreProteins are very large molecules that serve a variety of functions within a cell or an organism; proteins are essential to all life forms. They are made up of one or more chains of amino acids, which in virtue of their sequences and their interactions give to each protein a particular three-dimensional structure, which is closely related to the function of the protein.

From a simple sequence of amino acids to its final three-dimensional structure, the macromolecule which a protein is passes through four levels of structuring known as primary, secondary, tertiary, and quaternary. At the end of these stages the protein begins to fold up into a stable three-dimensional structure that will allow it to fulfil its proper function. At that point the protein may assume in one or the other of two general types of structure: a helix or a sheet.

When proteins have a similar sequence of amino acids, they tend to have similar three-dimensional structures. But it sometimes happens that two proteins that have identical sequences have a three-dimensional structure (and thus a function) that is different with regard to the solvent or the environment in which they are found. In fact, proteins interact with other elements in their environments, and these influence the final shape of their structure.

Bridges that stabilize proteins

In a general way, different types of interactions between different amino acids, and between these amino acids and their environment, permit stabilization of the structure of proteins. Among these interactions, we find covalent bonds, and disulfide bridges in particular. These connections form through oxidation, and they are formed between atoms of sulphur contained in two amino acids that are called cysteines. “We find disulfide bridges especially at the level of extracellular proteins, which are excreted, or at the level of membranes because that allows them to be stabilized, said Julien Becker, a doctoral candidate working under the supervision of Prof Louis Wehenkel in the Systems and Modeling research unit that is part of GIGA. “Inside the cell, the environment is fairly stable and there is not so much of a need to further stabilize proteins.” For example, disulfide bridges are found in the proteins of snake or scorpion venom, in antibodiesproduced by the immune system, in insulin - the hormone that regulates the concentration of sugar in the blood – and in many other proteins.

“Obtaining a 3-D representation of a protein is difficult and very costly, ” said Becker. “It is not actually possible for many proteins, and in the cases where it is possible, we have to use methods like X-ray crystallography or nuclearmagnetic resonance spectroscopy. This is why scientists try to develop computer-based methods in order to make faster predictions about three-dimensional protein structure, from which they obtain indications about the properties of the proteins that will form. This is the area in which Julien Becker and Louis Wehenkel work. “The Holy Grail in this area is the ability to predict the 3-D structure of proteins,” Becker said.

Toward a tool for predicting 3-D structure

Finding a method to predict the three-dimensional structure of proteins that is computer-based is a tremendous challenge. Researchers are obliged to begin working on the simplest aspects of the problem. “We have been using computers to predict the properties of proteins for 30 years. In the beginning we just looked at their makeup in terms of amino acids and tried to see if it was a matter of soluble or non-soluble residues,” Becker said. “Today we use automatic prediction more and more and we work with thousands of proteins which we try to characterize using mathematical techniques”. Automatic prediction is a technique that allows rules for the formation of protein structure to be produced automatically on the basis of previous data that contains examples, cases that have already been examined and validated. “If you take the example of a database that has pictures of cars, trucks and motorcycles in it, you see that these images are annotated so that you know which image belongs to which category, ” Becker explained. “Then we apply an algorithm to that information that will look for similar forms. Once the model has been learned, we can feed new images into it and it will automatically classify them into the correct category”.
Predictor 
With regard to proteins, scientists assign values to various properties that they want to predict such as the accessibility of residues, the secondary structure in the form of a helix or in the form of a sheet, etc. In general, researchers try to generate one predictor per type of property. “My idea was to combine all that into one predictor, and in order to do that I started working on problems a little more complicated than disulfide bridges,” he said. “Here the difficulty is not to predict a property for a single residue but for a pair of residues”.

Calculating the probability of the formation of disulfide bridges

Small protein predictionCysteines are the only types of residues that can form bridges themselves, from one cysteine to another. “The three-dimensional structure of the protein will eventually be significantly constrained by these bridges. So, when we look for the cysteines that are connected to each other, we can already get a good idea of the structure of the protein, Becker said. The main thing is to find a means of predicting the formation of disulfide bridges between different cysteines present in a protein. “In this kind of learning, we try to represent every object, in this case the pairs of cysteines, as a vector of numerical values between zero and one. My idea attempts to use a series of ways of representing a pair of residues and then to see which of these is the most successful at predicting disulfide bridges”. Becker took into account criteria such as the length of the protein (that is, the number of residues), the number of cysteines, their relative position, and so on. Then he developed an algorithm that could select the best descriptions, up to the point where adding a description would not improve the predictive tool’s performance. “Before applying this algorithm, we tested three major families of learning machines in order to choose the one that works best for the application of our algorithm”.

Disordered regions of proteins: a new challenge...

The work of Julien Becker led to the creation of a predictive tool called x3CysBridges, that allows researchers to calculate the probability that the cysteines of a single protein will form disulfide bridges. The principle is simple: researchers using the tool enter a sequence of amino acids that they want to test and the predictor carries out the necessary calculations. Following this, the researchers are informed about the predictor’s answer, that is, which cysteines will form bridges, via e-mail. The results of this work have been published in the Journal PLoS ONE (1). “We have already had several requests from biologists who wanted to use this service, ” Becker said.

Today Becker is working on a new challenge: “I am in the process of repurposing this method to do automatic learning with regard to disordered regions, ” he said. Within a protein, there are regions called “ordered” whose three-dimensional structure is stable; but there are also disordered regions in which there is a great deal of variability. These protein segments do not adopt a stable structure, and in general they are used to “attract” metabolites within or outside a cell. “This is the same kind of study that we went through for disulfide bridges; I try to learn a representation in the form of numerical values for each one of the residues of a protein in order to construct a predictive tool for disordered regions”. Little by little, the performance of predictive tools is improved, and so this young scientist pursues the goal of a single automatic predictive tool that will allow him to predict the three-dimensional structure of proteins.

(1) Becker J, Maes F, Wehenkel L. On the relevance of sophisticated structural annotations for disulfide connectivity pattern prediction. PLoS One. 2013;8(2):e56621. doi: 10.1371/journal.pone.0056621. Epub 2013 Feb 15.


© Universit� de Li�ge - https://www.reflexions.uliege.be/cms/c_352385/en/revealing-the-3d-structure-of-proteins?printView=true - March 29, 2024