Discussion on simulations to probe the molecular factors contributing to the observations in Zotter et al, JBC, 2017Zotter et al, JBC, 2017 published by Gideon Schreiber and colleagues

Present: RW, HXZ, NB, ANA, SKS, PF

Questions to address with simulations:

  1. Is the rate of diffusion (trans, rot) of the substrate slower in a crowded environment (like the cell) due to non-specific sticky (hydrophobic) interactions? Is the diffusion rate dependent on crowder type? on crowder conc? on substrate? on other conditions (IS, pH etc). Is slowing of the diffusion dependent on the 'sticky' fluorophores and not observed for the natural beta-lactam substrate?
  2. Can SDAMM-type simulations be compared directly to expts that could be done for in vitro systems by Gideon or by Hofmann at Weizmann?
  3. Is the bimolecular rate constant for diffusional association of the substrate to beta-lactamase slower in a crowded environment? Why?

Aim: Tackle the first set of questions and aim to have some progress that we can discuss with Gideon at BDBDB4.

To do:

  1. Generate 3D structures and force field parameters of CFF2, beta-lactam without D and A fluorophores, and FDG and RDG. Explore conformational space at a rather simple level (in MOE or Schrödinger)- generate e.g. for CFF2 an extended structure and a U-shaped structure that could interact with the b-lac active site. Generate parameters for SDA (charges and radii at pH 7 (7.4)) using Gaurav's scripts. Examine whether reasonable models have been derived. (Ariane will start this off with help from Gaurav etc)
  2. Build SDA boxes of molecules with several substrate molecules (ca. 20?, negatively charged) and HEWL (+ve charge) or Myoglobin (small net charge) crowders at different (initially 2) protein concentrations (low and high (near cellular)). Simulate each system for several microsec and compute Drot and Dtrans of substrates. Examine substrate-protein contacts (and check for substrate-substrate contacts).
  3. If we see trends in slower diffusion dependent on substrate and protein crowder type, do these relate to interaction strengths? Test this by FMAP calculations for the excess chemical potential of putting substrate molecule in system of crowders. Test this by doing calculations for a system that can be tested in vitro (discuss at BDBDB4)

other (just to bear in mind for now) : Systems with enzyme and substrate and no crowder: -Do NAM kon calculations for b-lac and different substrates and b-lac mutants: would give reference for uncrowded system -Do periodic box kon/MFPT calculations for b-lac + substrate + varying crowder concs. Post analysis may miss satisfaction of reaction criteria. Would need to introduce correct monitoring of rxn criteria during simulations into sda(mm).