Nobel Prize 2023 in Physiology or Medicine : Drew Weissman and Katalin Karikó Pioneering mRNA Vaccine Breakthrough Against COVID-19

Nobel Prize 2023 in Physiology or Medicine : Drew Weissman and Katalin Karikó Pioneering mRNA Vaccine Breakthrough Against COVID-19

In 2023, the Nobel Prize in Physiology or Medicine was awarded to Drew Weissman and Katalin Karikó for their pioneering work in the development of mRNA vaccine technology, a breakthrough that proved to be pivotal in combating the COVID-19 pandemic. Their discoveries, focusing on nucleoside base modifications, enabled the creation of mRNA vaccines that have become a cornerstone of global public health. This achievement has not only transformed vaccine development but also offers promising avenues for the treatment of other diseases. To understand the depth of their contribution, it is essential to explore the scientific context, their individual and collaborative research, the challenges they overcame, and the far-reaching impact of their work.

The Background of mRNA and Its Promise

The concept of using mRNA as a therapeutic tool is not new. Messenger RNA (mRNA) is a molecule that carries genetic instructions from DNA to ribosomes, where it directs the synthesis of proteins. In theory, mRNA could be used to instruct cells to produce proteins that could stimulate an immune response, providing an innovative approach for developing vaccines. The idea of creating vaccines based on mRNA was a logical one, but in practice, several challenges had to be overcome before it could be realized as a viable therapeutic option.

For decades, researchers had been interested in using mRNA to produce proteins inside the body. However, one of the major challenges was the instability of unmodified mRNA when introduced into human cells. Furthermore, the immune system recognized foreign RNA as a threat, leading to rapid degradation and inflammation, which hindered its potential use in vaccines.

Katalin Karikó's Early Career and Vision

Katalin Karikó, a Hungarian-born scientist, had long been fascinated by the potential of mRNA. She first joined the University of Szeged in Hungary, where her focus was on RNA biology. She later moved to the United States, where her career faced significant obstacles. Karikó was a dedicated researcher, but much of her work, particularly her focus on mRNA as a therapeutic tool, was considered unconventional. In the 1990s, her ideas about mRNA were dismissed by many in the scientific community. The consensus was that mRNA would not be a viable tool for therapeutic applications.

Despite these challenges, Karikó was persistent. Her work on mRNA centered on developing a method to stabilize RNA, making it suitable for use in vaccines and therapies. This would be a long and difficult road, but Karikó’s determination never wavered. She spent years researching how to modify mRNA so that it could enter cells without triggering an immune response that would lead to its degradation.

Her breakthrough came when she collaborated with Drew Weissman, an immunologist at the University of Pennsylvania, who had expertise in understanding how the immune system reacts to RNA. Their collaboration would prove to be the turning point in mRNA vaccine development.

Drew Weissman’s Expertise and the Breakthrough

Drew Weissman had spent much of his career studying how the immune system reacts to foreign molecules. His focus was on understanding how the body identifies and responds to pathogens, including viruses. Weissman’s expertise in immunology made him an ideal partner for Karikó’s work. When the two met in the early 2000s, they began exploring the idea of using mRNA as a therapeutic tool, a novel concept at the time.

In particular, Weissman’s knowledge of how the immune system interacts with RNA was key. The immune system is designed to recognize and attack foreign RNA, as it often signals the presence of viruses. This immune reaction was one of the main reasons why mRNA had not been successfully used in vaccines before. The immune system would essentially destroy the mRNA before it could do any work.

Weissman and Karikó’s breakthrough came when they discovered that modifying the nucleosides—the building blocks of RNA—could reduce the immune response triggered by mRNA. They developed a method of altering one of the bases in the mRNA molecule, a modification known as pseudouridine, that allowed the mRNA to enter cells without being immediately degraded or provoking an immune response. This modification, which is a relatively small change in the chemical structure of RNA, proved to be the key to stabilizing mRNA and making it effective for vaccine development.

The Role of Nucleoside Modifications

The core of their discovery was the concept of nucleoside base modifications. Nucleosides are the building blocks of RNA, and their sequence determines the genetic information carried by the molecule. The immune system recognizes certain patterns in RNA, particularly the presence of unmethylated RNA, which triggers a response to destroy the foreign molecule.

Karikó and Weissman’s work focused on the modification of nucleosides to reduce the immune response. Specifically, they used pseudouridine, a modified nucleoside, to replace uridine in the mRNA. This subtle change prevented the immune system from recognizing the mRNA as a foreign invader, allowing the mRNA to persist longer in the body. With this modification, the mRNA could deliver its instructions to cells, enabling them to produce proteins and stimulate an immune response without triggering the body’s defense mechanisms.

This discovery had enormous implications. It not only made mRNA more stable and effective for use in vaccines, but it also paved the way for the development of mRNA-based therapies for a variety of diseases, including cancer, genetic disorders, and more. The ability to modify nucleosides in RNA opened up an entirely new field of medicine.

The Development of mRNA Vaccines for COVID-19

In late 2019, the world was faced with an unprecedented challenge: the emergence of the COVID-19 pandemic. The rapid spread of the SARS-CoV-2 virus demanded an equally swift scientific response. Traditional vaccine development, which often takes years, was not an option in this case. Researchers around the world needed a solution that could be developed quickly and at scale.

The work done by Karikó and Weissman on mRNA provided the foundation for the rapid development of COVID-19 vaccines. The mRNA vaccine platforms developed by companies like Pfizer-BioNTech and Moderna were based directly on the discoveries made by Karikó and Weissman. By using the mRNA technology, these vaccines were able to instruct cells to produce the spike protein of the SARS-CoV-2 virus, which then triggered an immune response and protected the body from infection.

The mRNA vaccines were developed in record time, with the first doses being administered in December 2020, less than a year after the virus was first identified. This was a groundbreaking achievement, and the success of the mRNA vaccines in combating COVID-19 highlighted the power of the technology that Karikó and Weissman had spent years developing.

Challenges and Breakthroughs Along the Way

The journey to this breakthrough was not without its challenges. Karikó and Weissman faced skepticism from their peers and struggled to secure funding for their research. In particular, Karikó’s focus on mRNA as a therapeutic tool was initially seen as unfeasible. The technology was considered too risky, and many doubted whether it would ever be effective in humans. However, Karikó’s persistence and vision, combined with Weissman’s expertise in immunology, allowed them to overcome these obstacles.

Their work was not an overnight success. It required years of trial and error, refining the technology and overcoming numerous scientific hurdles. But ultimately, their breakthrough changed the landscape of vaccine development forever. The mRNA vaccines for COVID-19 are not only a triumph in the fight against the pandemic, but they also represent a new era in medicine, offering hope for the treatment of many other diseases.

The Impact of Their Work

The impact of Karikó and Weissman’s work extends far beyond the COVID-19 pandemic. Their research has opened the door to a new class of vaccines and therapies, which could potentially be used to treat a wide variety of diseases, from cancer to genetic disorders. The ability to rapidly design mRNA-based vaccines and therapies allows for more flexible and responsive treatments, which is particularly important in the face of emerging diseases.

In addition to the development of vaccines, their work has implications for cancer treatment. Cancer cells often produce abnormal proteins that can be targeted by the immune system. Using mRNA, scientists can potentially instruct the body’s cells to produce these abnormal proteins, prompting an immune response that targets and destroys the cancer cells.

Conclusion

In recognition of their groundbreaking work, Drew Weissman and Katalin Karikó were awarded the 2023 Nobel Prize in Physiology or Medicine. Their discoveries concerning nucleoside base modifications in mRNA were pivotal in the development of the mRNA vaccines that have saved countless lives during the COVID-19 pandemic. Their research has not only revolutionized vaccine technology but has also opened the door to new treatments for a variety of diseases. The Nobel Prize is a fitting acknowledgment of their perseverance, vision, and transformative contributions to science and medicine. Their work is a testament to the power of collaboration, innovation, and scientific determination in solving some of the world’s most pressing health challenges.

Sources from Nobelprize.org

AlphaFold vs Traditional Methods of Protein Modeling

AlphaFold vs Traditional Methods of Protein Modeling

The field of protein modeling has evolved dramatically over the past few decades, with significant advances that have improved the accuracy and speed of predicting protein structures. Among the most notable developments is the emergence of AlphaFold, an artificial intelligence (AI)-driven tool developed by DeepMind, which has revolutionized the way researchers predict protein structures. To understand its significance, it's essential to compare AlphaFold with traditional methods of protein modeling.

 

Traditional Methods of Protein Modeling

Traditional methods of protein structure prediction rely heavily on experimental data and computational techniques that have been refined over the years. These methods can be broadly categorized into homology modeling, threading (fold recognition), and ab initio (de novo) methods.

Homology Modeling

Homology modeling is based on the principle that proteins with similar sequences are likely to fold into similar three-dimensional structures. This method requires a known protein structure (template) that is homologous (i.e., shares sequence similarity) to the protein of interest. The steps in homology modeling typically include:

• Identifying homologous sequences from a protein database (such as PDB).
• Aligning the target sequence with the template structure.
• Constructing the model by replacing the residues of the template with those of the target protein, while keeping the backbone structure intact.

While this method is useful for predicting structures of proteins with known homologs, its accuracy drops significantly when there are no close homologs available.

Threading (Fold Recognition)

Threading is used when no close homologs are available. In this method, the target sequence is "threaded" onto a library of known protein structures to identify the most likely fold. The challenge here is that the protein sequence might fold into a novel structure that is not represented in the template library, making this method computationally expensive and less reliable for novel folds.

Ab Initio (De Novo) Methods

Ab initio methods attempt to predict protein structures from scratch without relying on known templates. These methods simulate the physical interactions between atoms, such as Van der Waals forces, hydrogen bonds, and electrostatic interactions, to explore the possible three-dimensional conformations of a protein. The complexity of the problem lies in the vast number of possible conformations a protein can adopt, making this approach computationally demanding and time-consuming. Popular ab initio methods include:

• Rosetta: A widely used tool for protein structure prediction and design.
• FOLDIT: A crowdsourced protein-folding game that helps predict protein structures.

Although ab initio methods have shown improvements over the years, they often struggle with large proteins and are highly sensitive to computational resources.

AlphaFold

AlphaFold represents a paradigm shift in protein structure prediction, using deep learning techniques to make accurate predictions based on a protein's amino acid sequence. Developed by DeepMind, AlphaFold is an AI-driven tool that leverages large-scale data and advanced neural networks to predict protein structures with remarkable accuracy.

Key Features of AlphaFold

• Deep Learning Approach: AlphaFold utilizes neural networks to predict protein structures. It is trained on a vast dataset of known protein structures and sequences. By learning patterns and relationships in this data, AlphaFold is able to generate highly accurate models for a wide range of proteins.
• Protein Language Modeling: One of the key breakthroughs in AlphaFold is its ability to model the "language" of proteins. AlphaFold analyzes sequences of amino acids and predicts how they fold into three-dimensional structures. It takes into account evolutionary information, co-evolutionary signals, and other contextual information to guide the folding process.
• Accuracy: In the CASP13 and CASP14 (Critical Assessment of Techniques for Protein Structure Prediction) competitions, AlphaFold significantly outperformed other methods, achieving near-experimental accuracy in many cases. It demonstrated its ability to predict the structures of proteins that had previously been unsolved, including proteins with no close homologs.
• Integration with Structural Biology: While AlphaFold doesn't replace experimental methods, it complements them. For example, when experimental data (such as cryo-EM or X-ray crystallography data) is available, AlphaFold can help refine and interpret these data. Conversely, for proteins where experimental structures are not available, AlphaFold provides a valuable alternative.

Comparison: AlphaFold vs Traditional Methods

(i) Speed

• Traditional Methods: Techniques like ab initio modeling and threading can be computationally expensive and time-consuming, often requiring days or weeks to generate a single model.
• AlphaFold: AlphaFold is significantly faster than traditional methods, thanks to the power of deep learning. It can predict protein structures in a matter of hours or even minutes, depending on the size of the protein.

(ii) Accuracy

• Traditional Methods: The accuracy of traditional methods depends on the availability of homologous proteins (for homology modeling) or known folds (for threading). Ab initio methods, while more general, often struggle with complex or novel proteins and can produce low-accuracy models.
• AlphaFold: AlphaFold has demonstrated near-experimental accuracy in many cases. Its predictions rival those of experimental techniques in terms of structural accuracy, especially for smaller proteins or those with no homologous structures.

(iii) Dependence on Experimental Data

• Traditional Methods: Homology modeling relies heavily on the availability of experimental data for homologous proteins. Ab initio and threading methods also rely on existing structural knowledge to guide the modeling process.
• AlphaFold: AlphaFold can predict structures without relying on experimental data for similar proteins. It uses evolutionary and co-evolutionary information to make predictions, which allows it to generate accurate models even for proteins with no known homologs.

(iv) Versatility

• Traditional Methods: Homology modeling is limited by the availability of known structures, and ab initio methods struggle with large proteins or novel folds. Threading methods have a similar limitation in terms of known folds.
• AlphaFold: AlphaFold is versatile and can generate accurate models for a wide range of proteins, including those with no known homologs. It can also handle both small and large proteins, making it applicable to a broad spectrum of protein types.

(v) Cost

• Traditional Methods: Experimental methods (such as X-ray crystallography, cryo-EM, or NMR) can be costly and time-consuming. Computational methods like ab initio modeling may require significant computational resources.
• AlphaFold: While AlphaFold itself requires computational resources for training, the cost of using AlphaFold to predict protein structures is relatively low compared to traditional experimental methods, making it more accessible to researchers.

(vi) Complementary Use with Experimental Techniques

• Traditional Methods: Experimental methods are often required to obtain the most accurate and reliable protein structures.
• AlphaFold: AlphaFold can complement experimental methods, helping to refine models or predict structures in the absence of experimental data. For example, when cryo-EM data is available, AlphaFold can improve the interpretation of electron density maps.

Limitations of AlphaFold

Despite its impressive performance, AlphaFold has a few limitations:

• Post-translational Modifications: AlphaFold does not always accurately predict post-translational modifications (such as phosphorylation or glycosylation), which can be crucial for understanding protein function.
• Protein-Protein Interactions: AlphaFold excels at predicting the structure of individual proteins, but it may not be as effective at predicting protein-protein interactions or the structures of multi-subunit complexes.
• Structural Dynamics: AlphaFold provides a static model of protein structure and does not take into account the dynamic nature of proteins, which can change conformation based on environmental conditions or interactions with other molecules.

Conclusion

AlphaFold has revolutionized the field of protein modeling by providing a fast, accurate, and versatile tool for predicting protein structures, outperforming traditional methods like homology modeling, threading, and ab initio approaches in many cases. It has opened new possibilities in drug discovery, disease research, and synthetic biology. However, traditional methods, especially experimental techniques, remain indispensable for validating and refining AlphaFold predictions, particularly in cases where dynamic information or post-translational modifications are critical. The combination of AI-driven tools like AlphaFold with traditional experimental techniques holds great promise for advancing our understanding of proteins and their functions.

Photo from iStock