Protein Fluctuations and Cavity Changes Relationship

Protein cavities and tunnels are
critical for function. Ligand
recognition and binding, transport, and enzyme catalysis require cavities
rearrangements. Therefore, the flexibility of cavities should be guaranteed
by protein vibrational dynamics. Molecular dynamics simulations provide
a framework to explore conformational plasticity of protein cavities.
Herein, we present a novel procedure to characterize the dynamics
of protein cavities in terms of their volume gradient vector. For
this purpose, we make use of algorithms for calculation of the cavity
volume that result robust for numerical differentiations. Volume gradient
vector is expressed in terms of principal component analysis obtained
from equilibrated molecular dynamics simulations. We analyze contributions
of principal component modes to the volume gradient vector according
to their frequency and degree of delocalization. In all our test cases,
we find that low frequency modes play a critical role together with
minor contributions of high frequency modes. These modes involve concerted
motions of significant fractions of the total residues lining the
cavities. We make use of variations of the potential energy of a protein
in the direction of the volume gradient vector as a measure of flexibility
of the cavity. We show that proteins whose collective low frequency
fluctuations contribute the most to changes of cavity volume exhibit
more flexible cavities.