1. Lys-Trp-Lys: Structure and Solvation by Molecular Dynamics Simulations Ali Hassanali

2. Loop Motion, Rigid-body Helix-Coil transitions
Protein Dynamics
Atomic Fluctuations,
Side-Chain Motion
Ligand docking flexibility
10-15 – s
Loop Motion, Rigid-body
Motion of helices
10-9 – s
Active site conformation adaptation, Binding
Domain and Subunit
motion
10-6 – s
Hinge-bending, catalysis
Folding/unfolding,
Helix-Coil transitions
Hormone activation, catalysis
10-3 – 104 s
Water Dynamics – Protein Dynamics – Biological Functioning

3. Indole-Water/Indole-Protein Debate
Agreement on two time-scales existing – disagreement on what causes the long time-scales.
Fundamentally: Experimental techniques characterize degree of perturbed water differently
MRD: Magnetic Resonance Dispersion (Halle. et. al.)
Dielectric Relaxation (Steinhauser et.al. , Weingartner et.al. )
NMR
Fast water
Long time scale is dominated by fluctuation of side-chains.
QENS (quasi electron neutron scattering) (Head-Gordon et.al)
TDFSS (Zewail et.al., Zhong et.al., Bagchi et.al.)
Slow water
Long time scale is attributed to dynamic exchange between bulk and hydration shell
protein
biological?? water

4. Theoretical Model
La state Mulliken charges obtained from CASSCF calculations by Sobolewski and Domcke Chem. Phys. Lett. 315 (3-4): 293, 1999.
S0 state dipole ~ 2.41 D and La state dipole ~ 4.96 D
“Excitation” achieved by instantaneously changing ground-state charges to excited state charges
No QM integrated into this technique. (Callis. et al)
No adjustment to ground-state charges to avoid force-field clashing.

5. KWK Isomers KN N KC Ground State: KN < N < KC
Excited State: N < KN < KC
Isomeric structures driven by interactions and potentially hydrophobic interactions.

6. Free-Energy Calculations
Free energy difference between each isomer
and reference structure determined
Thermodynamic Integration techniques used
MD simulations yield free-energy differences
and free-energy derivatives
Ground State:
Excited State:
KN :71%, N: 20%, KC: 9%
KN :35.8%, N: 50.2%, KC: 13.9%
Ground-state free energy calculations with GROMOS96 yield same ordering

7. making any changes to the GROMOS96 forcefield.
Importance of these interactions in chemistry and biology: stabilizing protein
structure during/after protein folding, protein-ligand recognition. Dougherty.
Science, 271(5246): 163, 1996.
Chipot et. al. JACS 118 (12): , show that forcefields underestimate
binding energy between toluene and ammonium in gas phase by ~ 6 kcal mol-1
They also demonstrate that this binding energy is reduced in solvent between
the range kcal mol-1
Aliste et. al. Biochemistry, 42(30):8976,2003 conduct their analysis without
making any changes to the GROMOS96 forcefield.
LJ well-depths ~ 10 kcal mol-1
A realistic model will account for unequal polarization effects and one lysine should be farther away from the indole (Minoux and Chipot, J.Am. Chem. Soc., 121(44):10366, 1999)

8. Fluorescence Anisotropy
Karplus et al. Biochemistry, 22(12):2884, 1983
Data fit with a double exponential for both MD and experimental data
Short time component attributed to local wobbling motion of tryptophan
Long time component represents tumbling of peptide with pi-cation complex.
Pi-cation complex doesn’t mean lysine is rigid! (Qiu et. al. J.Phys. Chem.,B 109(35): 16901,2005)
Note first 100 fs of experimental data represents internal conversion between 1La and 1Lb states

9. Explicit time dependence on La excited state potential surface
Stokes Shifts
Double energy difference between absorbed and emitted photons
Depends on the electronic excitation energy. Gas phase excitation energy
of indole chromophore is ignored in our calculations due to cancel out
Explicit time dependence on La excited state potential surface
Excited t
No modeling of Lb La conversion
Ground

10. Competitive Solvation
Long equilibrium excited and ground state runs
Inverse correlation between indole-protein and indole-water solvation energies
Indole-protein/water fluctuate over a very large range but total solvation energy does not!
Competition independent of water model. SPC demonstrates same phenomena
Range of ground-state fluctuations is less but still exists.
Relationship to rugged landscape of proteins/peptides where gains/loss in enthalpy and entropy pave the free energy landscape?

11. Stokes Shifts Results
Stokes shifts data for three isomers shown with error-bars
Three important time scales
Inset shows inertial regime
Inertial regime dominated by indole-water solvation
Ground
Excited t

12. N-Complex Fits MD Experiment
Total Stokes shift show remarkable agreement
Experimental inertial decay is slower than MD:
experimental or MD (forcefield) artifact?
No isomerization detected in experiment
Small errors in barrier size can affect rates of
activated processes in MD
Were Stokes shifts observed long enough to
witness isomerization?

13. N-Complex Solvation Steady lowering of solvation energy
drives overall process
Initial drop over 1 ps is dominated by
indole-water contribution
0.79 ps and 54 ps solvation process
involves compensation between
protein and water to solvate indole
For KN and KC , isomerization
makes similar analysis difficult

14. N-Complex Peptide Readjustment
Clear motion of lysine NT ammonium toward indole by ~ 0.08 nm
C-terminus lysine side-chain moves by same distance but farther away
N-terminus lysine side-chain also shows minor adjustment

15. KN & KC Peptide Readjustment
KC shows clear response by moving toward indole and then relaxing
KN shows a less discernible process perhaps due to faster isomerization rates
Remote lysines for both isomers show clear response by moving away from
the indole ring
Other coordinates for KN might reveal similar behavior as KC

16. Peptide Dynamics, Isomerization
NH3 +
NH3 +
KN N
KC N
Chain-extension breaks pi-cation and perhaps hydrophobic interactions
Isomerization rates for KN is faster than KC
KC isomerization is driven by kinetics, even though it is at a higher free energy
than KN

17. Peptide-Motions

18. Acknowledgements Sherwin Singer Dongping Zhong Tanping Li
Members of Singer and Zhong groups
Thank you for listening!
Funding Sources:
LIGDP and NSEC