Page 27 - CBAC Newsletter 2017

protein and temporal resolution required to simulate physiological function (pico-seconds to seconds.) Previously,
a customized super-computer called Anton was specifically built to perform all-atom calculations of Kv1.2 dynamics
(a voltage gated potassium ion channel.) Over the course of a month, a single trajectory of protein conformational
changes over a long-enough time scale (230 micro-seconds) was simulated and analyzed. This kind of computing
ability is not a readily available resource to scientists and since multiple protein conformational trajectories are
required to understand the overall mechanism of voltage gating, it is an impractical approach for advanced analysis.
Furthermore, many cardiac ion channels activate at a slower rate and would require calculations over a much longer
time scale (seconds).
The Rudy Lab has developed a framework to compute the aforementioned protein dynamics at the atomistic level
over the gating time scale of cardiac ion channels using serial computing. The CHPC has been extremely helpful in
providing the resources required to perform this massive undertaking. We chose the cardiac slowly-activating
delayed rectifier potassium ion channel (IKs; extremely large protein, ~ 4000 amino-acids) as a prototype to explore
this type of protein dynamics simulation. This ion channel is a complex of pore forming homo-tetramer KCNQ1
(alpha-subunit) modulated by KCNE1 (beta-subunit). Tertiary subunits in the system include PIP2 and Calmodulin
with calcium ions. Mutations in the IKs ion channel are responsible for the long QT syndrome that cause arrhythmias.
The IKs is essential to the repolarization phase of the ventricular action potential and plays a key role as part of the
‘repolarization reserve’. Additionally, increased IKs current results in reduced action potential durations; IKs is
therefore critical for rate adaptation and beta-adrenergic modulation of the cardiac cell. The simulations conducted
in the CHPC will provide the information required to simulate the IKs structural changes during an action potential
in a ventricular cell, thus providing valuable insight into arrhythmogenic mechanisms. Furthermore, this framework
can be utilized for any kind of protein dynamics simulation and is therefore of general interest and impactive. CHPC
based computations have allowed us to simulate normal behavior, ligand modulation, atomistic mechanism of
subunit interactions and much more. We further plan to use the CHPC to study subunit stoichiometry, arrhythmia-
causing mutations and drug effects.

CBAC Center Heartbeat | 22