Small gases, such as dioxygen (O2), nitric oxide (NO), carbon dioxide (CO2) and ammonia (NH3), participate in numerous biochemical processes in living cells. CO2 and NH3 are involved in acid-base homeostasis; NO is involved in signaling; and O2 is involved in metabolism. As a vast majority of these processes take place inside of cells, gas transport across biological membranes becomes one of the most fundamentally physiological processes. The movement of polar NH3 gases is facilitated through membrane channels, such as Rh proteins and aquaporins (AQPs). Nonpolar O2, NO and CO2 gases may simply diffuse through the lipid bilayer, but there are membranes that are impermeable to these small gases. AQPs are one of the most extensively studied membrane channels and are comprised of thirteen isoforms in human cells with unique selectivity features to gases. Not only are they known to function as water channels, they also faciliate the transport of glycerols and NH3, as well as CO2 and NO. Although gas permeation across biological membranes has been extensively studied in particular by Boron, a long-time collaborator, the existence of gas channels has remained controversial especially for CO2. Thus, fundamental aims to be addressed include: (1) elucidating the structural mechanism underlying solute selectivity in AQPs, and (2) the role of AQPs in transporting gases across membranes.


Boron was among the first to discover gas-impermeable membranes and aquaporin 1 (AQP1) as a gas channel in erythrocytes. The group also demonstrated the selectivity of gases in different AQP isoforms. However, the results are based on electrophysiology and X-ray crystallography experiments and can be varied depending on approaches. Moreover, membranes with low permeability to gases, which are of erythrocytes and lens fibers, present high contents of cholesterols and sphinomyelins. To provide microscopic and structural descriptions of how different membrane lipid compositions modulate the transport and selectivity of gases in AQPs, MD simulations will be performed using the Center-developed software NAMD (TRD1). Since the transport of gases involves interactions between gases and amino acids and/or lipids, it requires all-atom MD simulations. Free energy methods, such as umbrella sampling, will be employed to assess energetic cost of gas conduction and degrees of gas selectivity. The Center-developed software VMD (TRD2) will provide atomic images of gas diffusion and the interface for calculating partitioning profiles and permeability coecients of gases.

State of the work

Boron recently showed that one AQP isoform (AQP1) is permeable to both CO2 and NH3 whereas the other isoform (AQP5) is permeable only to CO2. However, it is also essential to first demonstrate the necessity of gas channels. We have carried out MD simulations of O2, CO2 and NO molecules in the presence of membranes with cholesterols and sphingomyelins. These simulations are different from our previous simulations which were conducted with glycerophospholipids as the only constituents of a lipid bilayer. The preliminary results have established a membrane model which shows high cholesterol and sphingomyelin contents in membranes significantly reduce membrane partitioning and permeability of these gases.

Research plan

The next steps will be structural and functional characterizations of AQPs in transporting gases across membranes by performing MD simulations using membrane models with and without cholesterols and sphingomyelin. MD simulations and free-energy calculations of CO2 and NH3 will be performed to characterize their selectivity in AQP1 and AQP5. The results obtained from computational studies will assist in designing effective experimental mutagenesis studies and provide molecular details for experimental observations.

Publications Database
  • Movement of NH3 through the human urea transporter B: a new gas channel. R. Ryan Geyer, Raif Musa-Aziz, Giray Enkavi, Paween Mahinthichaichan, Emad Tajkhorshid, and Walter F. Boron. American Journal of Physiology - Renal Physiology, 304:F1447-F1457, 2013.
  • Exploring transmembrane diffusion pathways with molecular dynamics. Yi Wang, Saher A. Shaikh, and Emad Tajkhorshid. Physiology, 25:142-154, 2010.
  • Nitric oxide conduction by the brain aquaporin AQP4. Yi Wang and Emad Tajkhorshid. PROTEINS: Structure, Function, and Bioinformatics, 78:661-670, 2010.
  • Exploring gas permeability of cellular membranes and membrane channels with molecular dynamics. Yi Wang, Jordi Cohen, Walter F. Boron, Klaus Schulten, and Emad Tajkhorshid. Journal of Structural Biology, 157:534-544, 2007.