Skip to main content

Association of Aquaporin Transmembrane Channels and Influence of their Oligomerization State on the Transport Properties

Final Report Summary - AQPASSOC (Association of Aquaporin Transmembrane Channels and Influence of their Oligomerization State on the Transport Properties)

Final report summary

Contacts: matteo.ceccarelli@dsf.unica.it lorant.janosi@dsf.unica.it

http://www.dsf.unica.it/~matteo/unica/unica/Main.html

OBJECTIVES.

The main objective of this application was to investigate by computational means how the oligomerization state and the environment influence the transport characteristics of channel proteins. These proteins facilitate the transport of ions and molecules across the cell membrane. As protein channel model in our simulations, we used the human aquaporin 5 (HsAQP5), which is mostly localized in the membrane of cells adjacent to air-interacting surfaces. All aquaporins conduct water molecules, and sometimes small sugar alcohol molecules as well.

Improper functioning of aquaporins in humans, consisting in impaired water regulation, can lead to diseases and disorders, such as nephrogenic diabetes insipidus or congenital cataract. This can be caused by alterations in (i) the structure of individual pores, (ii) their oligomerization state, or (iii) the environment outside the cell (e.g. pH, salt concentration). Our proposal targeted the understanding of what is the impact of the oligomerization state on transport properties, and what is the impact of the environmental factors on the oligomerization state. Since all known aquaporins form tetramers, but it has been shown that individual channels also conduct water, answering these questions will shed light onto why and how these pores associate.

RESULTS. We began our investigation by determining the transport properties of HsAQP5 in its natural tetrameric oligomerization state, using fully atomistic molecular dynamics (MD) simulations, and then compare these results to the single channel ones. Since water flux through aquaporins is relatively high, we used unbiased MD simulations to investigate these properties. The results we obtained were unexpected at the time we wrote this proposal. Our study is the first to report that in human aquaporins in particular, and more generally in mammalian aquaporins, a gating water transport mechanism seems present for regulating the water conduction across the channel. While such mechanism was found experimentally in other kingdoms of life, like yeast and plant, it was not reported in mammalians aquaporins. Our results call for an experimental verification, however the primary sequence alignment of human aquaporins with those of yeast and plants supports the existence of a similar gating mechanism.

Specifically, we found that different monomeric conformations within the tetramer lead to a distribution of structures, which can be characterized as open or closed. The switch between the two states of a channel is a tap-like mechanism at the cytoplasmic end which regulates the water passage through the pore. The channel is closed by a translation of a specific histidine residue localized on the 'wall' of the pore to the inside. On the other end of the channel (selectivity filter), the water flow is regulated as well by narrowing down the entry to the channel. Furthermore, we quantified this by calculating water permeability for a channel when the gate is open and the selectivity filter is wide or narrow. The obtained osmotic permeability for the fully open channel was found in good agreement with the reported experimental value.

Larger fluctuations of the pore radius and the presence of only the open state in the single (unassociated) channels suggests that the tetrameric oligomerization state plays an important role in stabilizing both the water channel structure and the water regulation through it.

While water transport is extremely fast through the pores, the switching between open and closed states happens only once in tens or hundreds of nanoseconds, making these simulations computationally very expensive. Nonetheless, in order to study how dynamics and structural distribution change with lipid composition, we extended our simulations of the tetramer in another bilayer environment, and found that qualitatively the HsAQP5 behavior remained the same, suggesting that the gating mechanism is mainly intrinsic to the aquaporin, i.e. is sequence-specific.

Finally, we calculated the free energy profiles of water along the protein channel both from equilibrium simulations (for open channels) and from non-equilibrium steered molecular dynamics (SMD) simulations (for both open and closed channels). While the equilibrium simulations revealed the small barrier values inside the channel, consistent with previous results and with the fast water permeation, they could not solve the high barrier in the cytoplasmic end of a closed channel. Combining SMD with free-energy profile calculation methods, we were able to determine the barrier and that water permeation will occur on over millisecond timescale, meaning that for biological purposes, there is no water passage through the channel.

The project has also proposed the calculation of free energy profile of association of single channels using coarse-grained models due to the computationally prohibitive requirements for performing them for fully atomistic systems. However, the model we proposed was unable to maintain and reproduce the structural behavior of aquaporins. The main reason behind this issue was the formation of vacuum inside the pores. While one can restrain the backbone conformation, the sidechains of the aminoacids arrange themselves in incorrect conformations. Furthermore, imposing further restraints on the coarse-grained structure would eliminate any possibility for reproducing the behavior that revealed the gating mechanism described in the 'Results' section.

Therefore, in order to calculate accurate association free energy profiles, and how they are influenced by the environment, we believe it is necessary to remain at atomistic level of description of the aquaporins. Considering the immense computational requirements to fulfill such a task, we started to develop an improvement on parallelizing the MD simulations over large number of supercomputer cores. This algorithm will be used within the Monte Carlo simulations proposed, and previously shown, to compute free energy profiles of association in large biomolecular systems.

CONCLUSIONS. Our study reports for the first time a water-regulating gating mechanism in HsAQP5 that, while similar to other reported mechanisms, is specific to humans, and represents a novel finding in regards to mammalian aquaporins. Considering the great importance of this finding, due to the potential drug targeting of HsAQP5, we extended our research to characterizing structurally and energetically this gating mechanism. While water transport occurs in single channels as well, we found that tetrameric oligomerization state is essential in regulating the water passage through the channels. Given that the current coarse-grained models do not reproduce the atomistic behavior of the channels, we continued our investigation on the channel association process by developing enhanced molecular dynamics parallelization code on extensive number of cores. The aim of this approach is to perform the association study in fully atomistic model.

Related documents