Structural basis for ApoE4-induced Alzheimer’s disease

Our long-term goal is to understand the mechanism that makes polymorphism in the apolipoprotein E (ApoE) protein one of the strongest genetic risk factors for Alzheimer’s disease (AD).

AD is the 6th leading cause of death in the USA and there are no effective treatments. Moreover, the prevalence of this age-related neurodegenerative disease is likely to increase as our population ages. Therefore, there is a great need to understand AD and develop therapeutics. ApoE is an appealing target because this lipid transporter is one of the strongest genetic risk factors for AD. Carriers of ApoE4 are up to 15-fold more likely to develop AD than carriers of the more common ApoE3 isoform, while ApoE2 appears to be protective against AD. Subsequent experiments have confirmed that ApoE4 plays a causal role in AD. However, the mechanism coupling ApoE and AD remains unclear.

Neither of the substitutions that distinguish ApoE4 and ApoE2 from ApoE3 occurs in a functional site (C112R in ApoE4 and R158C in ApoE2), suggesting they indirectly impact AD risk by altering the protein’s conformational preferences. Understanding the structural differences between neurotoxic and non-toxic isoforms could enable the design of ‘structure correctors’ that combat AD by stabilizing non-toxic conformations. However, characterizing these differences remains challenging. Partial crystal structures of the different isoforms are essentially identical and the rest of the protein has largely defied structural characterization because of disordered regions and a propensity to form heterogeneous oligomers.

We are uncovering the structural determinants of ApoE-induced neurotoxicity by building and analyzing atomically-detailed Markov state models (MSMs) of neurotoxic and non-toxic variants. We are testing these models with single molecule Förster resonance energy transfer (smFRET) experiments in collaboration with Andrea Soranno and Carl Frieden.

Our models will provide insight into the connection between ApoE and AD and a foundation for precision medicine efforts, such as predicting the implications of rare ApoE isoforms for AD and designing therapeutics. Our methods will also be applicable to other proteins with disordered regions.

Allosteric impact of non-active site mutations on enzymatic function

Our objective is to uncover a mechanistic understanding of how non-active-site (NAS) mutations exploit allosteric coupling to impact enzymatic activities.

NAS mutations are often critical for optimizing new functions, appearing frequently in the evolution of drug resistance and directed evolution studies. We are initially focusing on understanding NAS mutations in multiple β-lactamases that confer these enzymes with activity against new substrates. These mutations play a crucial role in the evolution of antibiotic-resistant infections, which kill tens of thousands of Americans and cost our nation billions of dollars every year. Our research will uncover a mechanistic understanding of how drug resistance emerges that will enable future efforts to predict and combat the evolution of antibiotic-resistant infections.

A predictive understanding of NAS mutations remains elusive because of the great diversity of mechanisms that couple distant residues. For example, a dihedral from an NAS residue can undergo concerted structural fluctuations with an active site dihedral, suggesting that perturbing the NAS dihedral is likely to influence the structure of the active site. While there are many methods for identifying concerted structural fluctuations, the sampling efficiency of these methods is limited by the ruggedness of proteins’ energy landscapes. There is also mounting evidence that correlations between the amplitudes of fluctuations often dominate allostery. However, the field lacks a systematic method for integrating knowledge of the amplitudes of fluctuations into a holistic model of allostery.

We are integrating novel computational methods with in vitro and in vivo experiments to converge on a quantitative understanding of the full spectrum of correlated fluctuations responsible for allosteric coupling. For example, we are deploying the FAST algorithm we developed to facilitate comprehensive sampling of proteins’ energy landscapes by leveraging Markov State Models (MSMs) to efficiently sample conformations with pre-specified features. We are also developing new methods for identifying both concerted structural changes and correlations between the amplitudes of fluctuations.

Together, these methods are revealing the active site structures of β-lactamases that are responsible for different activities and NAS residues that can modulate the populations of these structural states through strong coupling to the active site. To test our insights, we are predicting mutations that modulate the populations of key functional states and, ultimately, alter β-lactamases’ ability to degrade specific antibiotics. Then we are experimentally testing 1) whether these mutations have the intended impact on β-lactamases’ activities and 2) whether the variants we design are capable of protecting bacteria from the target antibiotic.

Completion of this work will result in a general framework for understanding allosteric communication that will serve as a basis for future efforts to predict drug resistance, design new antibiotics, and manipulate allostery in other systems.

Design of allosteric inhibitors to restore the efficacy of existing antibiotics

We’ve developed a combination of computational and experimental approaches for identifying cryptic allosteric sites, which are druggable pockets that only open when a protein fluctuates away from its crystal structure. The ability to target these sites could create a wealth of new opportunities for drug design, so now we’re in the process of developing a tool chain for discovering small molecules that bind cryptic sites.

Our primary model system is TEM-1 beta-lactamase, which contributes to antibiotic resistance by degrading beta-lactam antibiotics, such as penicillin. This is an important application as bacteria are rapidly evolving resistance to all of our existing antibiotics and most pharmaceutical companies are no longer investing in new antibiotics. Together, these trends have raised concerns that we may be entering a post-antibiotic era, in which common bacterial infections are life-threatening events.

We’re using rational drug design tools developed to target crystal structures to target experimentally-validated structures of cryptic allosteric sites from our simulations. Then we’re using a battery of experimental tests to see if our best hits inhibit beta-lactamase by binding in their predicted poses. Successful compounds may be useful for restoring the efficacy of existing antibiotics.

Demonstrating the applicability of our approach in multiple settings

Having made significant progress on the problems outlined above, we are testing the generalizability of our approach in a variety of settings. Examples include:
1) Viral proteins that are often assumed to be undruggable, e.g. from Ebola and SARS-CoV-2.
2) Myosin motors, which are some of the most mutated proteins in human disease.
3) G proteins, which essential components of GPCR signaling and present complementary opportunities for drug design.

Adaptive sampling algorithms

Molecular simulations are an important component of all of our work. Therefore, we are also working on new methods for running these simulations as efficiently as possible. Rather than trying to run one long simulation, our main focus is on developing adaptive sampling methods. These algorithms run many independent simulations in batches. The starting points for each batch of simulations are chosen by analyzing all previous batches of simulations and determining where more simulation data would be most beneficial. Running simulations in this manner allows us to make effective use of commodity hardware and to ensure that we don’t keep simulating the same events over and over again, which is a common problem with long, conventional simulations. Deciding where to run new simulations and how to integrate the information from many independent simulations into a single, statistically rigorous model gets into lots of interesting physical chemistry and machine learning.

Extraction of valuable insight from large simulation datasets

Finally, we are drawing on information theory and deep learning to automate the extraction of valuable insights from large simulation datasets. A major theme is trying to minimize the assumptions that are built into our approaches, allowing us to be surprised by all the unexpected ways evolution has solved different problems.