Periodic Reporting for period 1 - HyBOP (Hydrogen Bond Networks as Optical Probes)
Berichtszeitraum: 2022-11-01 bis 2025-04-30
Fluorescence takes place throughout the natural world. Most conventional chemical wisdom proposes that in organic entities, fluorescence occurs in conjugated systems, such as in the aromatics. However, in biological settings, the interaction of light with matter occurs in media built up of dense networks of hydrogen bonds. Recent experiments suggest that it is possible to observe fluorescence from these networks too. This could open the possibility of designing hydrogen-bond networks with enhanced fluorescence, offering enormous fundamental and practical potential.
The overarching goal of HyBOP is to decipher, using advanced computer simulations, the exotic optical properties of hydrogen-bond networks and to harness them as probes of water-mediated forces. To achieve this, HyBOP will tackle the following challenges:
1) Establish the ground rules for creating fluorescent hydrogen-bond networks in biological materials.
2) Understand how to drive the electrons and nuclei of water networks into regimes where they can fluoresce.
3) Use the optical behaviour of these networks to probe hydrophobic forces in nature.
To uncover the complex chemistry of hydrogen-bond network fluorescence, and guide the discovery of new fluorophores, we will deploy state of the art electronic excited-state molecular dynamics in combination with machine-learning techniques. This will provide HyBOP with ground-breaking knowledge which will lay a theoretical framework to motivate development of new experimental probes of hydrophobicity.
HyBOP seeks to bring hydrogen-bond networks to the forefront of chemistry in their use as optical probes; by laying the theoretical ground-work for designing non-invasive fluorophores in biophysics, opening up a new window into the origins of autofluorescence in medical diagnostics and finally, provoking frontier electron and nuclear spectroscopy, HyBOP will have a spill-over effect and build new synergies across several branches of the physical sciences.
The overarching goal of HyBOP is to decipher, using advanced computer simulations, the exotic optical properties of hydrogen-bond networks and to harness them as probes of water-mediated forces. To achieve this, HyBOP will tackle the following challenges:
1) Establish the ground rules for creating fluorescent hydrogen-bond networks in biological materials.
2) Understand how to drive the electrons and nuclei of water networks into regimes where they can fluoresce.
3) Use the optical behaviour of these networks to probe hydrophobic forces in nature.
To uncover the complex chemistry of hydrogen-bond network fluorescence, and guide the discovery of new fluorophores, we will deploy state of the art electronic excited-state molecular dynamics in combination with machine-learning techniques. This will provide HyBOP with ground-breaking knowledge which will lay a theoretical framework to motivate development of new experimental probes of hydrophobicity.
HyBOP seeks to bring hydrogen-bond networks to the forefront of chemistry in their use as optical probes; by laying the theoretical ground-work for designing non-invasive fluorophores in biophysics, opening up a new window into the origins of autofluorescence in medical diagnostics and finally, provoking frontier electron and nuclear spectroscopy, HyBOP will have a spill-over effect and build new synergies across several branches of the physical sciences.
The ERC Consolidator Team hosted at the ICTP has made several important achievements in the last two years. Some of these are highlighted below:
1) We have uncovered a generic mechanism which we dub as the 'carbonyl-lock-mechanism' which we believe provides a general framework for rationalizing the origins of non-aromatic fluorescence in many organic/peptide aggregates. We have deployed state-of-the-art excited-state simulations to understand this process. In essence, we find that by locking/constraining the carbonyl mode, we can limit accessibility to the conical intersection thereby enhancing the possibility of emission. This paper was published in Nature Communications in 2023 (https://www.nature.com/articles/s41467-023-42874-3(öffnet in neuem Fenster)).
2) We developed an unsupervised learning approach for detecting angular jumps in liquid water which we have found to involve a highly collective and cooperative process thereby challenging the idea of a jump mechanism which is localized in the water hydrogen-bond network. We are now extending this tool to investigate the mechanism in salt solutions. This paper was published in Nature Communications also in 2023 (https://www.nature.com/articles/s41467-023-37069-9(öffnet in neuem Fenster)). This protocol was also used to investigate the jump-mechanism in supercooled water to examine the putative second critical point where we discovered that the transition seems to be strongly coupled to these angular jumps. This paper was published recently in PNAS (https://www.pnas.org/doi/10.1073/pnas.2407295121(öffnet in neuem Fenster)).
3) We are now working on a framework to use unsupervised learning techniques to: 1) train and develop new potentials for salt solutions in close collaboration with Prof. Paul Cremer from UPENN who is conducting new experiments in this area and 2) developing techniques to discover the relevant modes in a system that take the excited state dynamics to the conical intersection - stay tuned for results soon!
1) We have uncovered a generic mechanism which we dub as the 'carbonyl-lock-mechanism' which we believe provides a general framework for rationalizing the origins of non-aromatic fluorescence in many organic/peptide aggregates. We have deployed state-of-the-art excited-state simulations to understand this process. In essence, we find that by locking/constraining the carbonyl mode, we can limit accessibility to the conical intersection thereby enhancing the possibility of emission. This paper was published in Nature Communications in 2023 (https://www.nature.com/articles/s41467-023-42874-3(öffnet in neuem Fenster)).
2) We developed an unsupervised learning approach for detecting angular jumps in liquid water which we have found to involve a highly collective and cooperative process thereby challenging the idea of a jump mechanism which is localized in the water hydrogen-bond network. We are now extending this tool to investigate the mechanism in salt solutions. This paper was published in Nature Communications also in 2023 (https://www.nature.com/articles/s41467-023-37069-9(öffnet in neuem Fenster)). This protocol was also used to investigate the jump-mechanism in supercooled water to examine the putative second critical point where we discovered that the transition seems to be strongly coupled to these angular jumps. This paper was published recently in PNAS (https://www.pnas.org/doi/10.1073/pnas.2407295121(öffnet in neuem Fenster)).
3) We are now working on a framework to use unsupervised learning techniques to: 1) train and develop new potentials for salt solutions in close collaboration with Prof. Paul Cremer from UPENN who is conducting new experiments in this area and 2) developing techniques to discover the relevant modes in a system that take the excited state dynamics to the conical intersection - stay tuned for results soon!
The results coming out from our ERC described earlier are opening up two key directions:
1) Our work has allowed us to discover design principles that are now inspiring the developing of using large-language models (LLMS) to essentially imagine/hallucinate new peptide sequences that may display enhanced optical activity which would enhance UV-vis absorption and possibly fluorescence. We are currently trying to pursue some new funding for applications in this direction.
2) Our results on collective angular jumps in water is opening up new directions on trying to measure and probe causality in aqueous solutions involving proteins solvated in water. This will allow us to understand the directionality in which information flows in a system where you have a protein and water coupled to each other. We believe that this will provide key insights into non-equilibrium processes in water as well as in understanding more deeply out-of-equilibrium hydrophobicity, a concept that we speculated about in the original ERC submission.
1) Our work has allowed us to discover design principles that are now inspiring the developing of using large-language models (LLMS) to essentially imagine/hallucinate new peptide sequences that may display enhanced optical activity which would enhance UV-vis absorption and possibly fluorescence. We are currently trying to pursue some new funding for applications in this direction.
2) Our results on collective angular jumps in water is opening up new directions on trying to measure and probe causality in aqueous solutions involving proteins solvated in water. This will allow us to understand the directionality in which information flows in a system where you have a protein and water coupled to each other. We believe that this will provide key insights into non-equilibrium processes in water as well as in understanding more deeply out-of-equilibrium hydrophobicity, a concept that we speculated about in the original ERC submission.