Periodic Reporting for period 1 - MIMIC (Mapping and modulating integrin mediated interactions)
Reporting period: 2020-01-07 to 2022-01-06
All complex organisms are composed of millions of living cells that are in constant interaction with each other and with the environment. These interactions are brought about by proteins embedded in the cell membranes, called receptors, forming interactions with proteins and other molecules outside of the cell, called ligands. These extracellular proteins can be signals that change the cells’ behaviour (for example hormones), can be on the surface of other cells (forming cell-cell contacts, for example in tissue formation or during the activity of immune cells), or even pathogens such as viruses (to get inside the cell where they replicate).
One of the most important receptors on the surface of human cells are the integrins. This receptor family has 24 members and while all of them are similar in their characteristics, they all recognise slightly different ligands. Virtually all of our cells have various integrins on their surface that aid the cell in performing its function. For example, our platelets have a very high number of the integrin called alpha(IIb)beta(3), and these integrins can bind to large proteins in the blood in order to clump together and clog wounds to stop bleeding. However, when the regulation of the platelets goes wrong, this clumping together can lead to very serious conditions, such as thrombosis. In addition to thrombosis, the activity and misregulation of integrins have been connected to a very wide range of other diseases as well, such as Crohn’s disease, multiple sclerosis, or cancer.
In light of their importance, it is clear why integrins have been the focus of biological research for many years. Until now about 80,000 scientific papers have been published about how integrins fulfill their functions and how they are misregulated in diseases. However, there are currently only 6 drugs used in clinics that target integrins, albeit many more were in development over the years. The fundamental problem that hinders efficient drug development is that integrins are very large and complex receptors and we do not fully understand how they recognise their ligands and what these interactions do precisely.
The main objective of MIMIC is to understand which integrin recognises which ligands by combining computational modeling and experimental biology. Based on the interactions between integrins and various ligands already described, we build models that can capture the most important features of these ligands that determine whether they can bind to integrins. Once these models are accurate enough, we can find new integrin ligands in various organisms, going beyond human proteins and potentially understanding how certain viruses - such as SARS-CoV-2 - cause diseases. This knowledge can help to better understand how our cells work and can also provide targets for future pharmaceutical drug development to combat severe and life-threatening diseases.
One of the most important receptors on the surface of human cells are the integrins. This receptor family has 24 members and while all of them are similar in their characteristics, they all recognise slightly different ligands. Virtually all of our cells have various integrins on their surface that aid the cell in performing its function. For example, our platelets have a very high number of the integrin called alpha(IIb)beta(3), and these integrins can bind to large proteins in the blood in order to clump together and clog wounds to stop bleeding. However, when the regulation of the platelets goes wrong, this clumping together can lead to very serious conditions, such as thrombosis. In addition to thrombosis, the activity and misregulation of integrins have been connected to a very wide range of other diseases as well, such as Crohn’s disease, multiple sclerosis, or cancer.
In light of their importance, it is clear why integrins have been the focus of biological research for many years. Until now about 80,000 scientific papers have been published about how integrins fulfill their functions and how they are misregulated in diseases. However, there are currently only 6 drugs used in clinics that target integrins, albeit many more were in development over the years. The fundamental problem that hinders efficient drug development is that integrins are very large and complex receptors and we do not fully understand how they recognise their ligands and what these interactions do precisely.
The main objective of MIMIC is to understand which integrin recognises which ligands by combining computational modeling and experimental biology. Based on the interactions between integrins and various ligands already described, we build models that can capture the most important features of these ligands that determine whether they can bind to integrins. Once these models are accurate enough, we can find new integrin ligands in various organisms, going beyond human proteins and potentially understanding how certain viruses - such as SARS-CoV-2 - cause diseases. This knowledge can help to better understand how our cells work and can also provide targets for future pharmaceutical drug development to combat severe and life-threatening diseases.
In the initial phases of MIMIC we analysed the available literature to compile an extensive catalogue of proteins that reside outside of the cell and are known to be able to interact with various integrins. This part of the project is exceptionally labour intensive as each publication has a different structure, style, and terminology. From this body of literature, we encoded all of the relevant information in a standard format and made it publicly available for the scientific community to use. It is located on a freely accessible website at http://mimicdb.embl.de/. As the next step, we analysed the sequences and structures of these proteins to distil common features that are needed to interact with specific integrins to build models of what are the requirements of integrin-binding at the molecular level.
Having these models enabled the search for new integrin ligands that have never been measured yet. Amongst the newly identified integrin ligands, we found several that can be connected to various proteins of the SARS-CoV-2 virus. We noticed that based on its sequence and structure features, the Spike protein of the virus contains a likely binding site for integrins. Previously it was known that the human cell receptor for the virus is a protein called ACE2, but this finding indicated that in addition, SARS-CoV-2 uses integrins as well to infect human cells. Further analysing ACE2 and integrins, we found that these receptors are likely to be able to interact with other proteins inside the cell that are connected to various processes, most notably endocytosis and autophagy - the process of the cell engulfing material from the outside and digesting certain parts on the inside. These interactions were experimentally verified by our collaborators. Both processes are known to be linked to viruses in general and our findings offer ways to potentially counteract the viral infection with know drugs that were developed for other purposes (https://www.science.org/doi/10.1126/scisignal.abd0334).
In addition to Spike, we suspected that the viral protein called ORF7a is likely to bind to integrins, but upon experimental testing, this turned out to be false. However, by now we know that this protein also binds to some other receptor on the surface of immune cells and is likely to influence how the immune system works. We are currently devising experiments to determine which of the 5000 cell surface receptor ORF7a binds to, so we can later develop strategies to counteract this interaction thereby potentially boosting the immune response to viral infection.
MIMIC also served with many other new candidates for integrin binding. We do not have the capacity to test all of these within the framework of the project, and therefore we make all of this data public, including the results of the literature searches, the built models, and the found candidate ligands for other members of the community to use. These results were being communicated in peer-reviewed scientific articles, as well as conference talks, Twitter posts (@_BalintMeszaros), press releases (https://www.embl.org/news/science/protein-sequences-and-sars-cov-2/) and YouTube videos (https://www.youtube.com/watch?v=Hyt-IbJ2WJc).
Having these models enabled the search for new integrin ligands that have never been measured yet. Amongst the newly identified integrin ligands, we found several that can be connected to various proteins of the SARS-CoV-2 virus. We noticed that based on its sequence and structure features, the Spike protein of the virus contains a likely binding site for integrins. Previously it was known that the human cell receptor for the virus is a protein called ACE2, but this finding indicated that in addition, SARS-CoV-2 uses integrins as well to infect human cells. Further analysing ACE2 and integrins, we found that these receptors are likely to be able to interact with other proteins inside the cell that are connected to various processes, most notably endocytosis and autophagy - the process of the cell engulfing material from the outside and digesting certain parts on the inside. These interactions were experimentally verified by our collaborators. Both processes are known to be linked to viruses in general and our findings offer ways to potentially counteract the viral infection with know drugs that were developed for other purposes (https://www.science.org/doi/10.1126/scisignal.abd0334).
In addition to Spike, we suspected that the viral protein called ORF7a is likely to bind to integrins, but upon experimental testing, this turned out to be false. However, by now we know that this protein also binds to some other receptor on the surface of immune cells and is likely to influence how the immune system works. We are currently devising experiments to determine which of the 5000 cell surface receptor ORF7a binds to, so we can later develop strategies to counteract this interaction thereby potentially boosting the immune response to viral infection.
MIMIC also served with many other new candidates for integrin binding. We do not have the capacity to test all of these within the framework of the project, and therefore we make all of this data public, including the results of the literature searches, the built models, and the found candidate ligands for other members of the community to use. These results were being communicated in peer-reviewed scientific articles, as well as conference talks, Twitter posts (@_BalintMeszaros), press releases (https://www.embl.org/news/science/protein-sequences-and-sars-cov-2/) and YouTube videos (https://www.youtube.com/watch?v=Hyt-IbJ2WJc).
MIMIC established the first publicly available database that offers an exhaustive list of known integrin interacting proteins. It is also the first effort to build publicly available models to find previously unknown candidate integrin ligands from large protein sequence and structure databases. Identifying new ligands this way serves several purposes: on one hand, it expands our knowledge of how cells interact with their environments, which is important in basic research. On the other hand, understanding the molecular principles of these interactions offers us ways of efficiently modulating them by developing molecules that alter these binding events. Understanding the molecular mechanisms of diseases and having effective strategies to combat them is becoming ever more important. Integrins are connected to diseases that impact all modern societies, including cancer, autoimmune diseases, Alzheimer’s, cystic fibrosis, autism spectrum disorder and schizophrenia, and several viral infections such as HPV-16, HIV, or West Nile. The successful medication of these diseases will contribute to the goal of the European Innovation Partnership on active and healthy ageing. Hopefully, the results obtained in MIMIC will be able to contribute to these goals by providing novel interactions serving as the basis for drug development.