• Aquaporin
• Images
• References
• slideshow html

Aquaporin (PDB ID: 3ZOJ) from Pichia pastoris

Created by: Duncan Cannon

            Aquaporins are found in all domains of life and are highly conserved. They are a class of transmembrane channels that allow water to flow down its concentration gradient when open. Some are very selective for water, while others (usually called aquaglyceroporins) allow for small molecules like glycerol to pass through. They are important for maintaining water homeostasis in the cell, which can affect the osmolarity of solutes (1). In humans, aquaporins are very important in kidney function. There are at least 10 different aquaporins found in mammals, seven of which are found in specialized kidney cells (2). Mutations in these aquaporins have been related to various kidney disorders such as nephrogenic diabetes insipidus. A human aquaporin has also been implicated in hyponatremic acute brain edema (2). The P. pastoris aquaporin, referred to as Aqy1, has been crystalized in the closed conformation. One structure of Aqy1 (PDB ID: 3ZOJ) was resolved to 0.88 Å, which revealed information about the protein-water interactions as well as water-water interactions (3).

            Aquaporins are found as transmembrane proteins and, in vivo, are normally found as homotetramers. The structure of Aqy1 was crystalized as a single subunit, but an A. fulgidus aquaporin (PDB ID: 3NE2) was crystalized as a homotetramer. As integral membrane proteins, they each have a transmembrane domain consisting of 6 transmembrane helices and two half helices that form a pseudo-transmembrane helix. The pseudo-transmembrane helix is very important in selectivity. The channel has an hourglass shape that is seen to bottleneck around the NPA, which is the region where the two halves of the pseudo-helix meet, so that in the center of the channel the water molecules are only allowed to pass in a single file manner. This hourglass shape is due to the right-handed helical bundle structure (1). A chloride ion can be seen bound between two helices at the top of the channel. The molecular weight of the P. pastoris aquaporin is 29914 Da, and the pI is 6.41 (4). In humans and plants, these channels are regulated by phosphorylation or pH change (5). Aqy1 has been reported as possibly being regulated by phosphorylation and mechanosensitivity (5).

            One of the important features of the aquaporin is its ability to select against proton transport so that it does not interfere with the cell’s established pH. Aquaporins have multiple different structural features that contribute to this. As stated above, the protein has 6 transmembrane helices as well as a seventh pseudo transmembrane helix. The two half helices are aligned so that the N-termini of the helices both point to the center of the protein. This means that the positive dipole of the two helices points to this center region. This positive potential serves as a barrier to any positive charge (3). This also orients the molecules in the channel in a specific manner, which is important to preventing transport of protons. Protons can flow quickly through a solution via a Grotthuss mechanism (3), which describes the process by which water molecules transfer protons through hydrogen bond interactions (Figure 1). When water molecules are hydrogen bonded to one another, they are able to exchange the proton that is bound between them. The water molecule that recently acquired a proton can pass along another proton to another water that it is donating a hydrogen bond to. As will be seen in the discussion of the selectivity filter, the water molecules are oriented so that their hydrogen bond pattern eliminates the possibility of transport by a Grotthus mechanism.

            The NPA motif is a highly conserved region near the center of the channel. Mutations of this region have been shown to allow for the passage of sodium ions (6). There are two critical residues in the region, Asn-112 and Asn-224. One of these asparagine residues is located on one of the pseudo-transmembrane helices, and the other residue is located on the other helix. This means that the NPA motif occurs at the region of highest positive potential. These residues coordinate with water molecules passing through the channel, forcing them into a specific configuration. Water molecules on the intracellular side of the NPA region donate hydrogen bonds to the water further down the channel. The two water molecules located within the NPA region have equal probability to donate hydrogen bonds with one another. The B-factors, which measure the disorder of a molecule or atom in a crystal structure, for the two molecules at the top of the NPA are very high. This indicates that they are disordered and can be found bonded in multiple states with relatively high probability for each state, hindering transport by a Grotthuss mechanism (3).

            Another domain located just above the NPA motif, is the selectivity filter (SF). The selectivity filter varies among certain aquaporins and related channels, especially ones that allow urea and glycerol to pass. However, there are two important residues that are conserved among highly water selective aquaporins, His-212 and Arg-227 (3). Within the selectivity filter, there are four possible binding sites for water, all four of which are too close together to be simultaneously occupied. The suggested mechanism of transport is pairwise motion of water molecules, similar to the mechanism of potassium movement through potassium channels. Positions 1 and 2 are both hydrogen bonded to the Nη² of Arg-227 and the carbonyl oxygen of Gly-220. Positions 3 and 4 are both hydrogen bonded to Nε of Arg-227, the carbonyl oxygen of Ala-221, and Nε of His-212 (3).

            Two different techniques are useful for finding homologues, PSI-BLAST and Dali Server search. PSI-BLAST queries the amino acid sequence of a protein and searches for sequence homologies in other proteins in the database. Proteins with similar sequences have low E values, indicating that there is a low probability that the homologous regions are due to random chance based on query length and database size. PSI-BLAST results reveal that the closest human homologue is hAQP1 (PDB ID: 1FQY), which has an E value of 8e-29, indicating that this Aqy1 is likely a good model for understanding some of human aquaporin function (7). A Dali Server search compares the query structure with other known structures in the PDB. The algorithm that it uses calculates a Z-score. A higher Z-score indicates a more structurally similar the protein (a Z-score greater than 2 indicates a similar fold). An aquaporin from the archaeal organism M. marburgensis (PDB ID: 2EVU) yields a Z-score of 30.7, indicating a high level of similarity (8). Aqy1 is highly selective for water, however, the class of aquaporins known as aquaglyceroporins also permit the transport of glycerol and certain other small organics like urea. AqpM, from M. marburgensis, falls somewhere in-between these two categories (1). Aqy1 and AqpM are highly similar with respect to three-dimensional structure but differ in selectivity. This is attributed to a mutation in the SF of AqpM, which changes the histidine, His-212 in Aqy1, to isoleucine, widening the SF by ~0.68 Å (1). This allows for weaker selectivity for water, but not as weak as in typical aquaglyceroporins, which have a glycine at this position.

            The high resolution of the Aqy1 structure allows for an in depth understanding of the mechanism of water selectivity that improves upon previous selectivity models. It has been shown how the SF as well as the NPA work in tandem to prevent proton transfer through the channel. The selectivity for water can be seen by understanding similar structures of aquaporins that are not as selective as Aqy1. Additionally, a previousl paper implicates a model of gating based on the movement of the N terminal cytosolic domain.