New Approach in Protein Identification Accelerates Drug Development

The provided image illustrates the binding between different areas of the PRC2 protein (depicted on the right side) and survivin. To achieve this, researchers divided the PRC2 sequence into 15-amino acid units and arranged them on a grid, where each pixel represents 1 unit. The color of each pixel indicates the strength of binding to survivin, with blue representing no or weak binding, and bright pink representing strong binding. Image credit: Atsarina Larasati Anindya.

In the realm of living cells, proteins play diverse roles depending on the specific cell type. Scientists have now devised a novel method to identify proteins without relying on their structural properties. This approach offers advantages in terms of speed, simplicity, and reliability compared to previous methods.

Traditionally, it has been widely accepted that the function of a protein is primarily governed by its structure. The arrangement of atoms within proteins, known as atomic sequences, defines their structure and shape. However, many proteins lack a well-defined structure.

Dr. Gergely Katona from the University of Gothenburg has pioneered a new technique where proteins are analyzed based on the number of amino acids or different atoms they contain, rather than relying on structural characteristics, in order to identify them and determine their function. By employing this scanning method, the researchers achieved a reasonably accurate prediction of the specific combination of amino acids required for binding to the survivin protein. The method demonstrated an 80 percent reliability rate, surpassing the identification accuracy achievable using protein primary structures. The study results have been published in the scientific journal iScience.

Dr. Gergely Katona, a Biochemistry Professor at the University of Gothenburg, has highlighted the reduced significance of protein structure in this context. During their experiments with the new scanning method, the researchers tested several thousand peptides, each consisting of 15 amino acids. They observed that the content of amino acids within the peptides had a significant impact on their binding to survivin, whereas the structure of the peptides played a minimal role.

“Simple quantitative approaches have often proven successful in science. In this case, we counted the number of amino acids and achieved surprisingly accurate predictions of protein function,” explains Dr. Gergely Katona.

The researchers perceive multiple advantages to this protein scanning method. Furthermore, machine learning (AI) expedites the process of linking the number and types of amino acids to specific functions. Consequently, the development of novel biological drugs can be accelerated.

In addition to advancing the scanning technique, the researchers made an intriguing discovery during their experiments. They found a previously unknown function of the survivin protein, which prominently exists in embryonic cells and prevents programmed cell death. In cancerous tumors, survivin becomes deregulated, thereby facilitating cancer development.

The researchers also observed a direct influence of survivin on another protein called PRC2, responsible for modulating various DNA functions within the cell, akin to programming. Dysfunctional PRC2 has been associated with different forms of cancer. Although existing cancer drugs currently target both survivin and PRC2, the newfound link between survivin and PRC2 implies that drug design might need to be altered to prevent severe side effects.

“We observed that by reducing the level of survivin, PRC2 activity increased. The ultimate goal for pharmaceutical companies is to identify the precise targets within the atomic sequences to achieve a balanced modulation of the two proteins,” states Dr. Gergely Katona.

Reference: “Survivin prevents the polycomb repressor complex 2 from methylating histone 3 lysine 27” by Maja Jensen, Venkataragavan Chandrasekaran, María-José García-Bonete, Shuxiang Li, Atsarina Larasati Anindya, Karin Andersson, Malin C. Erlandsson, Nina Y. Oparina, Björn M. Burmann, Ulrika Brath, Anna R. Panchenko, Maria Bokarewa I., and Gergely Katona, published in iScience on May 29, 2023.
DOI: 10.1016/j.isci.2023.106976

Frequently Asked Questions (FAQs) about Protein identification

More about Protein identification

• University of Gothenburg: Official Website
• iScience Journal: Article
• Survivin protein: Wikipedia
• PRC2 protein: Wikipedia
• Cancer research: National Cancer Institute
• Drug development: Pharmaceutical Research and Manufacturers of America