Multiscale simulation of photoactive biomolecules

 

Our group's major efforts are focused on developing and applying new multiscale simulation methods to study the photoactive biomolecules that are essential in photopharmacology, biomedical imaging and photocatalysis, with the eventual goal of providing design principles to enhance their efficiency. The unique aspect of our multiscale simulation approach is that we bridge the accurate quantum mechanical treatment of photochemical reactions and the efficient molecular mechanics modeling of biomolecular motion.

​

Photoactive biological systems have promising applications in biomedical and energy sciences. Understanding photochemical reactions in biomolecules with molecular-level detail is critical for improving the design of new photoactive proteins, and molecular simulation is now ready to achieve this goal with ever-increasing capability. However, the multiscale nature of the photocontrolled biological systems poses significant challenges for simulation. Specifically, the photochemical reactions occur on short timescale and lengthscale, and typical force-field based simulations cannot describe their quantum mechanical nature. Meanwhile, the biomolecular motions initiated by the photochemical reactions occur on long timescale and lengthscale, and are beyond the reach of most quantum mechanical simulations. To address these challenges, we combine ab initio non-adiabatic dynamics simulation, force-field based molecular mechanics modeling, first principle electronic structure calculation, and enhanced sampling techniques to understand the multiscale nature of photoactive biomolecules.

​

• Light control of protein activity by molecular photoswitches in photopharmacology

​Using light to control protein activity is crucial for next-generation biomedical science because of its high spatiotemporal precision. One popular approach to achieve light control is to link proteins with molecular photoswitches such as azobenzene-derived compounds. Despite the success of this approach in recent years, several important questions remain to be answered in order to improve the photocontrol efficiency by molecular photoswitches. For example, how can we achieve high photoisomerization quantum yield and long absorption wavelength in protein environment? How can we maximize the difference in the protein’s function as a result of the photoisomerization reaction?

​

In order to investigate these questions, we combine ab initio non-adiabatic dynamics simulation and molecular mechanics modeling to understand the molecular mechanism underlying the photocontrol of proteins by molecular photoswitches. Such insights will eventually lead to better design principles for molecular photoswitches to be used as light-activatable drugs in precision medicine.

​

• Aggregation-induced emission phenomenon in biomedical imaging

​Fluorescence imaging is an important tool for disease diagnosis. In the past two decades, many organic fluorescent molecules were designed to possess aggregation-induced-emission (AIE) characteristics. The AIE-based fluorophore has enhanced fluorescence in the aggregate state compared to in solution, because the intramolecular motion leading to non-radiative decay is restricted when the fluorophore is in the aggregate. Compared to the traditional fluorophores, AIE-based fluorophores can detect biomolecules with high sensitivity and selectivity. Despite these advantages, there remain many challenges to improve their performance. Specifically, development of AIE-based fluorophores with increased fluorescence quantum yield, fluorescence wavelength and target selectivity are still necessary for next-generation bio-imaging technologies.

 

To achieve this goal, we develop multiscale simulation framework to study the fluorescence mechanism of the AIE-based fluorophores in biomolecules. Combining docking simulations, molecular dynamics simulations and free energy calculations, we aim to predict the fluorophore's selectivity of the target biomolecules. Using the ab initio excited-state molecular dynamics simulation with the quantum mechanics/molecular mechanics (QM/MM) scheme, we aim to quantify the fluorescence wavelength, the excited state lifetime and fluorescence quantum yield. Such information will be valuable for the design of highly sensitive and selective AIE-based fluorophores.

​

• ​Photoinduced electron transfer and electron bifurcation in proteins

​Electron transfer (ET) is fundamental for signal conduction and energy conversion in living organisms. Ultrafast photoinduced ET occurs in cryptochromes, which is a key component of the circadian clock and associated with diseases such as seasonal affective disorder, bipolar disorder, and depression. As another example, electron bifurcation separates two electrons from one donor to one branch of high-reduction-potential cofactors and the second branch of low-potentials cofactors, generating strong reductants for downstream processes. A well-known electron bifurcating enzyme is Complex III in the electron transport chain, where it extracts two electrons from the fully-reduced ubiquinol and delivers one to cytochrome c­1 and the other to cytochrome b in the Q-cycle. Electron bifurcation has been recognized as a general and fundamental mechanism of energy conservation in life.

​

Despite decades of study, many fundamental questions about ET in these biomolecules remain elusive. For example, how does nature prevent electron backflow and short-circuits to ensure high-efficiency charge separation? How do the electrostatics and dynamics of the protein affect the ET rate and electronic coherence between cofactors? Simulations quantifying the thermodynamics and dynamics of ET are necessary for answering these questions. However, such simulations are challenging due to the multiscale nature of biological ET: changes in the electronic state often couple with large and slow structural changes of biomolecules. Well-established ET models such as Marcus theory are undoubtedly very successful at describing biological ET, but they also have limitations. For example, the assumption of equilibrium statistical mechanics is often questionable in the regime of ultrafast ET in biomolecules such as cryptochromes, where nonergodic effects are prominent. Moreover, these models lack atomic-level details of real-time ET dynamics in biomolecules. In this regard, ab initio non-adiabatic dynamics simulations are indispensable to complement traditional ET models, because they directly propagate the coupled motions of nuclei and electrons without introducing assumptions such as ergodicity and (non)adiabaticity. However, existing methods in this category are not capable of accurately and efficiently describing the electronic structure and quantum dynamics of biological ET. To overcome these challenges, we develop new ab initio non-adiabatic dynamics methods to provide an unbiased description of the ET dynamics at the all-atom and all-electron resolution. Our project generates a powerful tool for understanding biological electron transfer in general. These key methodological advantages  enable us to comprehensively characterize the non-adiabatic reactivity in complex biomolecular systems with unprecedented accuracy. Furthermore, we integrate it with existing simulation tools and ET models in a multiscale simulation framework to understand ET in cryptochromes and electron-bifurcating enzymes.

​​

• Water splitting reaction catalyzed by photosystem II for renewable energy generation

Photosystem II (PSII) is essential for photosynthesis. It catalyzes the water photodissociation reaction and initiates the light-dependent reactions of photosynthesis. Understanding the mechanism of the proton-coupled electron transfer (PCET) reactions catalyzed by the oxygen-evolving complex (OEC) in PSII is necessary for the design of artificial water-splitting catalysts. Previous experimental and computational studies on PSII mainly focused on the static structures and the energetics of the intermediates of OEC. However, many important questions remain to be answered regarding the dynamical aspects of the PCET reactions connecting these intermediates, e.g., how are the proton and electron transfer events dynamically coupled to each other? How is the PCET mechanism influenced by the protein environment? Ab initio non-adiabatic dynamics simulations are necessary to answer such questions. However, one major challenge for such simulations is to accurately and efficiently describe the electronic structure of the OEC.

​

To overcome this challenge, we combine ab initio non-adiabatic dynamics method and GPU-accelerated advanced electronic structure methods to study the non-adiabatic dynamics of the PCET reactions in PSII. The simulations will provide new insight into the design principles of artificial water-splitting catalysts. 

​​

​​

Understanding the catalytic selectivity and activity of unspecific peroxygenases

​

The unspecific peroxygenases (UPOs), originally discovered in fungi, are capable of the selective oxyfunctionalisation of hydrocarbons. Because UPOs use peroxides as both the final electron acceptor and exclusive oxygen donor, they avoid the oxygen dilemma suffered by cytochrome P450 monooxygenases (P450s) and have higher stability and turnover numbers than the latter. These advantages make them promising biocatalysts for functionalizing unactivated C-H bonds under mild conditions without the use of toxic and expensive catalysts. However, the applications of UPOs are currently limited by their suboptimal chemical selectivity, reduced activity in the organic solvents (OS) of the substrates, and oxidative damage from reactive oxygen species. To improve UPOs’ selectivity, activity, and tolerance to OS and oxidative damage, it is critical to understand the molecular basis of their catalysis, which remains elusive to date. To this end, computational characterization of the catalytic cycle of UPOs is much needed, but significant challenges remain for modeling the constituent steps occurring at different timescales, such as substrate binding and C-H bond dissociation.

 

To fill this gap, we employ multiscale simulations, combining methods from both quantum mechanics and molecular mechanics regimes, to understand how substrate binding and the active site’s structural and electronic properties regulate the reaction selectivity and substrate specificity of the wild-type (WT) UPO from Agrocybe aegerita (AaeUPO). Moreover, we investigate how the catalytic activity of WT AaeUPO is influenced by the presence of OS and how the enzyme can potentially protect itself against oxidative damage during catalytic turnover via the hole-hopping mechanism. Furthermore, we characterize the molecular basis of how mutations fine-tune UPOs' catalytic selectivity, activity, and tolerance to OS. Our computational characterization of the catalytic mechanism will be systematically benchmarked against experiments. We also make computational predictions of UPO mutants with enhanced stability and activity in OS, which will then be verified experimentally. Overall, our research aims to provide new insights into the structure-function relationship of UPOs’ catalysis and assist in the engineering of next-generation biocatalysts with improved selectivity, activity, and tolerance to OS and oxidative damage. In the long run, our research will significantly impact the sustainable chemical transformation of feedstocks into value-added products.

​

Computational characterization of the functional mechanism of melibiose transporter

​

Melibiose permease (MelB) is a key model of cation-coupled symporter in the Major Facilitator Superfamily (MFS) that is crucial for various physiological and pathological processes. Despite extensive research, its symport mechanism remains unresolved, like most other MFS symporters. Building upon new structures of MelB in different conformations, we computationally characterized the all-atom free energy landscapes of the entire melibiose translocation process in MelB and its uniport mutant. The results align with experimental data and reveal a strong coupling between melibiose translocation and global protein conformational changes. Additionally, Na+ binding reduces the free energy barrier of melibiose translocation through allosteric coupling. Last but not least, our computational results elucidated the energetic contributions of different residues at the cation-binding pocket to the binding affinity of H+ and Na+, and affirmed Asp59 to be the sole protonation site of MelB. Taken together, these new insights are essential for understanding the cation-binding mechanism and unraveling the molecular origin of cation-coupled transport mechanisms in MFS symporters.