New Computer Model Shows How Proteins are Controlled
Enzymes are large proteins that are involved in virtually every biological process, facilitating a multitude of biochemical reactions in our cells. Because of this, one of the biggest efforts in drug design today aims to control enzymes without interfering with their so-called active sites. This “at a distance” approach is called “allosteric regulation”, and predicting allosteric pathways for enzymes and other proteins has gathered considerable interest. Scientists from EPFL, with colleagues in the US and Brazil, have now developed a new mathematical tool that allows more efficient allosteric predictions.
Allosteric regulation is a fundamental molecular mechanism that modulates numerous cell processes, fine-tuning them and making them more efficient. Most proteins contain parts in their structure away from their active site that can be targeted to influence their behavior “from a distance”. When an allosteric modulator molecule binds such a site, it changes the 3D structure of the protein, thereby affecting its function.
The main reason allosteric sites are of such interest to drug design is that they can be used to inhibit or improve the activity of a protein, eg. the binding strength of an enzyme or a receptor. For example, diazepam (Valium) acts on an allosteric site of the GABAA receptor in the brain, and increases its binding ability. Its antidote, flumazenil (Lanexat), acts on the same site, but instead inhibits the receptor.
Generally speaking, an allosteric drug would also be used at a comparatively lower dose than a drug acting directly on the protein’s active site, thus providing more effective treatments with fewer side effects.
Developing an allosteric model
Despite the importance of allosteric processes, we still do not fully understand how a molecule binding on a distant and seemingly unimportant part of a large protein can change its function so dramatically. The key lies in the overall architecture of the protein, which determines what kinds of 3D changes an allosteric effect will have.
The lab of Matthieu Wyart at EPFL sought to address several questions regarding our current understanding of allosteric architectures. Scientists classify these into two types: hinges, which cause scissor-like 3D changes, and shear, which involve two planes moving side-by-side. Despite being clear mechanically, the two models do not capture all cases of allosteric effects, where certain proteins cannot be classified as having either hinge or shear architectures.
The researchers explored alternative allosteric architectures. Specifically, they looked at the structure of proteins as randomly packed spheres that can evolve to accomplish a given function. When one sphere moves a certain way, this model can help scientists track its structural impact on the whole protein.
Using this approach, the scientists addressed several questions that conventional models do not answer satisfactorily. Which types of 3D “architecture” are susceptible to allosteric effects? How many functional proteins with a similar architecture are they? How can these be modeled and evolved in a computer to offer predictions for drug design?
Using theory and computer power, the team developed a new model that can predict the number of solutions, their 3D architectures and how the two relate to each other. Each solution can even be printed in a 3D printer to create a physical model.
The model proposes a new hypothesis for allosteric architectures, introducing the concept that certain regions in the protein can act as levers. These levers amplify the response induced by binding a ligand and allow for action at a distance. This architecture is an alternative to the hinge and shear designs recognized in the past. The computational approach can also be used to study the relationship between co-evolution, mechanics, and function, while being open to many extensions in the future.
This article has been republished from materials provided by EPFL. Note: material may have been edited for length and content. For further information, please contact the cited source.
CRISPR Reveals New Targets for Promising Cancer DrugsNews
Novel screening method identifies new drug targets that could potentially enhance the effectiveness of PD-1 checkpoint inhibitors, a promising new class of cancer immunotherapy.READ MORE
New Algorithms Help Extract 3-D Biological Structure from Limited DataNews
CAMERA researchers capitalize on their Multi-Tiered Iterative Phasing approach to determine molecular structure of proteins and viruses from X-ray free electron laser data.READ MORE
Big-Data Analysis Points Toward New Drug Discovery MethodNews
A research team has developed a computational method to systematically probe massive amounts of open-access data to discover new ways to use drugs, including some that have already been approved for other uses.
Comments | 0 ADD COMMENT
EMBO Workshop: Integrating Systems Biology: From Networks to Mechanisms to Models
Apr 15 - Apr 17, 2018
EMBL Course: Introduction to Next Generation Sequencing
Apr 09 - Apr 12, 2018
EMBL Course: Introduction to Metabolomics Analysis
Mar 20 - Mar 23, 2018
EMBL Course: RNA Sequencing Library Preparation - How Low Can You Go?
Mar 19 - Mar 23, 2018