RESEARCH

MOLECULAR INFORMATION TRANSFER IN THE CELL

We are physicists, chemists, biologists, and computer scientists, who address pressing questions in molecular biology of the cell signaling in health and disease. We develop computational methods and workflows by combining frameworks of molecular dynamics, quantum chemistry, and molecular evolution with multi-resolution experimental data. Current emphasis is on functional lipids in membrane-associated phenomena, protein dynamics in complex environments, and plasma membrane cell signalling.


Research Highlights (by area)

A majority of findings were achieved and/or verified in collaboration with experimental colleagues, including M.Gruebele, C.Rienstra, M.Burke, K.Hristova, J.Morrissey, and K.Zhang, among others.

Transmembrane signaling proteins: receptor tyrosine kinases

  • Funding: PI: NIH-R01 GM141298 2021/25, co-Is: Hristova & Zhang
  • Candidate structures of previously invisible RTK TM dimer structures (in progress)
  • Mutations play in EphA4 oligomerization and signaling (J. Biol. Chem, 2021)
  • Key side chains in binding of RTK TrkA juxtamembrane domain to membrane (JPCB, 2019

Functional lipid signatures & membrane-active agents

  • Funding: PI: NCSA CDDR (2021/23), co-I Kindratenko; co-I: NIH-tR01 GM123455 (2016/23),  PIs: Rienstra, Morrissey, Tajkhoshid
  • Cholesterol coupled dynamics in membrane (JACS, 2023)
  • Dynamic structure of anionic PS lipids shaped by Ca2+ in signaling (Biochem., 2018)
  • Contributed to capturing structure of the functional amphotericin B (AmB) – ergosterol sponge and development of safe AmB variant (Nature, published, 2023
  • Structure of the functional AmB sponge (Nat. Str. Mol. Biol., 2021)

In silico cell: dynamics in the crowd

  • Funding: key contributor: NSF 2205665 (2022/26), NIH-R01GM093318 (2017/21), PI: Gruebele
  • Perspective review article (JPCB, 2023)
  • Hinge-bending landscape of PGK in human cytoplasm (JPCL, 2024)
  • In-cell dynamics of ATP (JPCL, 2022)
  • Folding dynamics of diverse topologies – protein B (Prot. Sci., 2023) & WW domain (JPCB, 2020)
  • Protein-protein interactions are transient – a sign of quinary structure (JPCL, 2019)

Mechanisms of fast protein folding

  • Funding: key contributor: NSF 2205665 (2022/26), NIH-R01GM093318 (2017/21), PI: Gruebele
  • Small ubiquitous osmolyte TMAO protects proteins (Biophys. J., 2023)
  • Formation of a dry molten globule by a fast-folding protein (PNAS, 2019)
  • Differences in local and global probes of folding (JPCL, 2016)
  • Trap state explored by a fast-folding protein (PNAS, 2015)

Signaling Through the Cell Membranes: Receptor Tyrosine Kinases

Transmembrane signaling is vital to the cell life and ~50% of all transmembrane proteins are using single-pass TM domains to relay the molecular signals into the cytoplasm. The receptor tyrosine kinase (RTK) family contains ~60 members that control cell differentiation, proliferation, and migration. In RTKs, the extracellular and the intracellular kinase domains are connected to the TM domain though rather short but flexible juxtamembrane domains (JMD). Though active experimental and computational research on RTK signaling is ongoing, many challenges to address the role of JMD-membrane interactions persist.  We use  all-atom molecular dynamics simulations to capture spontaneous JMD binding and formation of the JMD-membrane complexes. The tens of µs timescales of modeling of wild type and designed JMD mutant interactions with membranes mimicking the cellular composition revealed the roles of the residue-lipid interactions. These computational findings are confirmed by our in vitro and in cell experiments. We find that an interplay between electrostatic sensing and hydrophobic membrane insertion dynamically adjusts the flexibly of JMD, which in turn controls positioning of the kinase domains in the cell. Intriguingly, a single point hydrophobic to basic mutation can lead to dramatically different JMD-membrane binding. Additionally, we find that highly charged anionic lipids can effectively restraint basic residues clearing the way for hydrophobic residues to latch to the membrane. This suggests the role molecular evolution plays even on the level of short <30 residue-long signaling peptides. (Wang, …, Diao, Zhang, Pogorelov, ChemRxiv, 2019, J. Phys. Chem. B, 2019).
Collaborators: Kalina Hristova (Johns Hopkins), Kai Zhang (UIUC), Jiajie Diao (U Cincinnati), Chad Rienstra (U Wisconsin-Madison).

Cell Membrane: Functional Lipids & Membrane-Active Agents

Cellular membranes play vital roles in numerous processes including protein activation and cell signaling. The plasma membrane is a complex and tightly-controlled heterogeneous environment that is dynamically shaped by proteins, lipids, and small molecules.
Sculpturing of plasma membrane by ions and anionic lipids. We employ a pivotal combination of the extensive HMMM all-atom MD simulations, comprehensive QM calculations on hundreds of lipid-ion complexes and detailed solid-state NMR spectroscopy on selectively-labeled diluted samples. Agreement between theory and experiment spans multiple spatial scales: 1) MD-based QM-calculated chemical shifts (hence chemical environments) of lipid carbon atoms in two identified long-lived lipid conformations agree with NMR measurements, 2) characteristic distances and coordination numbers of lipid-ion nanoclusters observed in MD simulations are in exact agreement with SSNMR REDOR measurements. Importantly and uniquely, use of diluted samples allows us to distinguish inter- and intra-lipid distances that otherwise are not separable. We find that the fundamental structural unit that activates the membrane is an anionic lipid tetramer, built from two types of lipid conformations, coordinated by Ca2+ (Hallock, …, Pogorelov, Biochemistry, 2018).
Collaborators: Martin Burke (UIUC), Chad Rienstra (U Wisconsin-Madison), James Morrissey (U Michigan, Ann Arbor), Emad Tajkhorshid (UIUC).

The ABC hypothesis. We are a part of a large multi-lab collaborative effort that investigates how membrane structure and dynamics influences functioning of proteins. Out collaborators employ solid-state NMR (Chad Rienstra), biochemical methodologies (James Morrissey), and molecular dynamics simulations (Emad Tajkhorshid). We were able to validate a novel hypothesis (Navoosi et al., JBC, 2011) that accounts for synergy between phosphatidylserine and phosphatidyl­etha­nol­amine in activation of the blood clotting cascade (termed the ABC, or Anything But Choline, hypothesis). We also showed that phospholipids with any headgroup other than choline strongly synergize with PS to enhance activation of coagulation factors. We propose that phosphatidylcholine and sphingomyelin (the major external phospholipids of healthy cells) are anticoagulant primarily because their bulky choline headgroups sterically hinder access to their phosphates.


Complex Cellular Environments: Signaling and Protein Dynamics in the Crowd

Protein-protein dynamics in E.coli sytoplasm. Proteins in vivo are immersed in a crowded environment of water, ions, metabolites, and macromolecules. In-cell experiments highlight how transient weak protein-protein interactions promote (via functional “quinary structure”) or hinder (via competitive binding or “sticking”) complex formation. Computational models of the cytoplasm are expensive. We tackle this challenge with an all-atom model of a small volume of the E. coli cytoplasm to simulate protein-protein contacts up to the 5 microsecond timescale on the special-purpose supercomputer Anton2. We use three CHARMM-derived force fields: C22*, C36m, and C36mCU (with CUFIX corrections). We find that both C36m and C36mCU form smaller contact surfaces than C22*. Although CUFIX was developed to reduce protein-protein sticking, larger contacts are observed with C36mCU than C36m. We show that the lifetimes of non-functional protein-protein contacts obeys a power-law distribution between 0.03-3 µs, with ~90% of all contacts lasting <1 µs — similar to the timescale for downhill folding (Rickard, Zhang, Gruebele, Pogorelov, JPCL, 2019).
Collaborators: Martin Gruebele (UIUC).

Protein folding characterized with multiprobes. We use computational and experimental techniques (in collaboration with the Gruebele Lab) to describe how the fast folding of proteins proceeds. We computationally designed and used multiple-site probes to study fast folding five-helix bundle lambda repressor protein variants and discovered that as these proteins fold, they can visit dehydrated compact misfolded states (Prigozhin, …, Gruebele, Pogorelov, PNAS, 2019) as well as previously unknown trap state that is populated from both the native and the denatured states and may slow down folding (Prigozhin, …, Pogorelov, Gruebele, PNAS, 2015). We also have shown (Sukenik, Pogorelov, Gruebele, JPCL, 2016) that in folding of the lambda repressor not all molecular probes report protein folding in similar fashion. We coupled experiments to computations to reveal that local probes reports differ from global ones, e.g. lower melting temperature, faster kinetics, and varying reaction to the environment. Interestingly, we also observed that some local probes can have global character.