© The Indian Express Pvt Ltd
Tags:
Journalism of Courage
Proteins are some of the most important life-sustaining molecules in any living organism. They perform a critical role in almost all biological processes. In fact, life itself would not be possible in absence of proteins.
This is why these large and complex molecules have been subjects of scientific studies for decades. This research, over the years, has led to significant breakthroughs when it comes to the understanding of the structure and functioning of these proteins, many of which have garnered Nobel Prizes.
This year’s Nobel Prize in Chemistry too was awarded for research on proteins. This award, however, is slightly different to previous ones. Scientists David Baker, Demis Hassabis and John Jumper have been honoured not for presenting any new insights into proteins themselves, but for developing tools that make it vastly easier and quicker to decipher their structures, and make entirely new proteins.
Hassabis and Jumper, who share one half of the prize, are co-creators of an artificial intelligence-based tool called AlphaFold that can predict the structure of a protein with outstanding accuracy. Baker, winner of the other half, used similar computational tools to create new proteins, which are not available in nature, but can perform many useful functions. Together, the trio managed to accomplish things that scientists have been striving to achieve for several decades.
Proteins are made of long chains of amino acids, which themselves are small organic molecules containing carbon, hydrogen, nitrogen and oxygen and sometimes sulphur. There are 20 different amino acids that serve as the building blocks of proteins. Different combinations of amino acids, arranged in a sequence and folded tightly into unique three-dimensional shapes, form the proteins that are vital to almost all biological processes.
Certain kinds of proteins, called enzymes, can speed up biochemical reactions within the body, while others can provide structural support to cells and tissues. Then there are some proteins that help in immune response, while others can store nutrients or energy.
It has long been known that the structure of proteins determines their functions. A few decades ago, it was discovered that it is the sequence of amino acids that decide how they will fold in three-dimensional shapes, thus determining the final structure of proteins. This discovery earned biochemist Christian Anfinsen the Chemistry Nobel in 1972, which he shared with Stanford Moore and William H Stein.
“So, the sequence of amino acids determines the structure of proteins, and the structure in turn determines the function of the proteins. But there are so many sequences of amino acids that are possible. It is like words in any language. They can be arranged in different ways to make a very large number of sentences. But not all combinations are valid,” P Balram, former director of Indian Institute of Science, told The Indian Express. “So just like in language, there has to be a grammar to the arrangement of amino acids in proteins, and this grammar has been very difficult to understand,” he said.
Although scientists have been trying to unravel the structure of proteins for several years for the insight this provides regarding biological processes in the body, progress in this regard has been slow.
“Traditional methods of decoding the structure of proteins, through x-ray crystallography, is a laborious and time-consuming process. It takes months, if not years, to figure this out,” Arun Shukla, of the Department of Biological Sciences and Bioengineering at IIT Kanpur, told The Indian Express.
Still, over the years, the structures of thousands of proteins have been catalogued. Simultaneously, a database containing all known sequences of amino acids has also been created.
Hassabis and Jumper built upon this work to create AlphaFold. The AI tool predicts the structures of proteins using known sequences of amino acids from the database. These predictions were then matched to catalogued protein structures in the other database. With training, the AI tool gained sufficient accuracy in predicting protein structures, given a particular sequence of amino acids.
Although developed just a few years ago, the tool has already undergone multiple upgrades. Today it is being used by a large number of researchers.
“This method is a spectacular departure from all previous efforts to decode the structure of proteins. Earlier efforts involved a lot of chemistry and physics. This one uses data and computation. This is extremely interesting and promising,” Balram said.
Baker relied on slightly different kinds of computations to design completely new, synthetic proteins that are not found in nature. “He has produced new sequences of amino acids that are valid proteins and perform useful and desirable functions. It is like constructing new sentences. This opens up huge new possibilities,” Balram said.
The new proteins can perform functions that naturally-synthesised proteins are not designed to. Theoretically, for example, a synthetic protein can be designed to degrade plastics which are otherwise not biodegradable.
The alternative way of producing synthetic proteins is a long process. “One can take the evolutionary approach, for example. Protein molecules can be allowed to undergo mutations until they acquire the desired characteristics. But this can involve hundreds of mutations and can take years. Baker’s process of producing synthetic proteins is much more efficient and effective,” biologist Alok Bhattacharya of the Ashoka University told The Indian Express.
The work of the three scientists has huge implications in drug discovery, and in overcoming stubborn diseases that occur due to protein disorders.
Interestingly, both the Physics and Chemistry Nobel prizes this year have been given to artificial intelligence related research. In fact, while acknowledging the foundational work in AI that led to the Nobel Prize in Physics this year, the Nobel Committee had listed AlphaFold as one of the examples of the kind of impact that was already evident. The Chemistry prize has now picked up the co-creators of AlphaFold for the honour.
“These AI-powered breakthroughs in predicting the structure of proteins have been creating a lot of buzz in the research community, and a Nobel recognition was already being talked about. So,this hasn’t come as a surprise,” Shukla said.
Balram said after the Physics Prize was announced on Monday, he had wondered whether AlphaFold would also be very quickly recognised. “It was done the very next day,” he said.
