Click here to close
Hello! We notice that you are using Internet Explorer, which is not supported by Xenbase and may cause the site to display incorrectly.
We suggest using a current version of Chrome,
FireFox, or Safari.
Two serial filters control P2X7 cation selectivity, Ser342 in the central pore and lateral acidic residues at the cytoplasmic interface.
Markwardt F
,
Schön EC
,
Raycheva M
,
Malisetty A
,
Hawro Yakoob S
,
Berthold M
,
Schmalzing G
.
???displayArticle.abstract???
The human P2X7 receptor (hP2X7R) is a homotrimeric cell surface receptor gated by extracellular ATP4- with two transmembrane helices per subunit, TM1 and TM2. A ring of three S342 residues, one from each pore-forming TM2 helix, located halfway across the membrane bilayer, functions to close and open the gate in the apo and ATP4--bound open states, respectively. The hP2X7R is selective for small inorganic cations, but can also conduct larger organic cations such as Tris+. Here, we show by voltage-clamp electrophysiology in Xenopus laevis oocytes that mutation of S342 residues to positively charged lysines decreases the selectivity for Na+ over Tris+, but maintains cation selectivity. Deep in the membrane, laterally below the S342 ring are nine acidic residues arranged as an isosceles triangle consisting of residues E14, D352, and D356 on each side, which do not move significantly during gating. When the E14K mutation is combined with lysine substitutions of D352 and/or D356, cation selectivity is lost and permeation of the small anion Cl- is allowed. Lysine substitutions of S342 together with D352 or E14 plus D356 in the acidic triangle convert the hP2X7R mutant to a fully Cl--selective ATP4--gated receptor. We conclude that the ion selectivity of wild-type hP2X7R is determined by two sequential filters in one single pathway: (i) a primary size filter, S342, in the membrane center and (ii) three cation filters lateral to the channel axis, one per subunit interface, consisting of a total of nine acidic residues at the cytoplasmic interface.
Fig. 1.Visualization of candidate residues involved in controlling the cation selectivity of the hP2X7R. A) Overall structure of the hP2X7Rwt homotrimer homology-modeled based on the cryo-EM structure of the apo-closed conformation of the rP2X7R (26), viewed perpendicular to the membrane normal. The hP2X7Rwt extends over ∼71 (N302B to D329C) and ∼50 Å (D356B to S584B) in the extracellular and intracellular directions, with a maximum width of ∼71 Å (G418B to S290A). Each of the three identical subunits (ABC) is shown in a different color (A) pale green, B) pale cyan, C) light pink), except for TM1 and TM2, which are shown in white and yellow, respectively, for easier identification. The three TM2 domains line the channel pore. Several residues are highlighted as spheres: C4 and C5, the first residues solved in the apo and open rP2X7R cryo-EM structures, respectively, and the last residue Y595 (corresponding to the C-terminal end) in their chain colors; S339 and S342 around the gate in aquamarine and cyan, respectively; in the ectodomain, the acidic residues in red and the residues that coordinate the ectodomain ATP4− binding by ionic interactions (K64, K66, R294, K311) and hydrogen bonding (T189, N292) in blue and marine, respectively. B), C) Enlarged lateral views of the transmembrane region of the closed and open states of hP2X7Rwt, including the channel pore as predicted by MOLEonline (33). Channels are shown in duplicate, with surrounding protein (purple) and, for a free view of the channel shape, in white, with only the selected residues (one or two per triplet) accessible to cysteine-reactive reagents from the extracellular space: I331 (yellow–orange), V335 (pale yellow), S339 (aquamarine), and S342 (cyan). The dashed red lines indicate the outer and cytoplasmic boundaries of the membrane as predicted by the OPM database (50). The open-channel model (C) lacks a sufficiently wide opening to the cytoplasm for ions to exit, as does the closed channel model (B). D), E) Cytoplasmic views on the channel pore and surrounding acidic residues (E14 in orange, D352 in red, D356 in firebrick) at the level of the membrane–cytoplasmic interface after masking of the cytoplasmic structure. Obviously, residues S339 and S342 move laterally to open the pore (compare D to E). In contrast, the acidic residues arranged in an isosceles triangle (yellow) do not move significantly during channel opening (compare D to E).
Fig. 2.Ramp currents mediated by hP2X7Rwt and mutants expressed in X. laevis oocytes. The ATP-induced ramp currents shown were calculated as the difference between the ramp currents before and during 0.1 mM ATP−4 application. Only the ramp currents in the voltage range of -70 to +40 mV are shown because capacitive currents are generated by voltage jumps from holding potentials of -40 to -80 mV. The name of the hP2X7R construct expressed is shown at the top of each of the three rows of panels, and the only salt (except ∼5 mM HEPES, diameter ∼10 Å) of the extracellular solution is shown at the left margin. The bold curved lines show the measured currents; the thin dashed straight lines show the linear fit of the current–voltage relationship near the reversal potential. From this approximation, the Vrev values (shown in each of the figures here and in Fig. 3) and the conductance (slope of the lines whose statistics are shown as Gat Vrev in Fig. 3) were calculated. I) Vrev was determined by extrapolation (see Materials and methods). The in-panel numbering (A–H) is used to uniquely identify each panel in the results text description.
Fig. 3.Permeation characteristics of hP2X7R constructs expressed in X. laevis oocytes. The bars represent the reversal potential (A) and the corresponding slope conductance at Vrev (B) of the indicated hP2X7R constructs during application of 0.1 mM ATP−4 in one of three extracellular solutions based on Na+Cl− (open columns), Tris+Cl− (red columns), and Tris+Glu− (purple columns). Data are mean ± SEM of 6–7 oocytes. A) The constructs are arranged in four correspondingly numbered groups that differ significantly in their reversal potential in Tris+Cl−. Group 1: hP2X7Rwt. Group 2: Substitution of single or multiple acidic residues by alanine or lysine as indicated; note that the Vrev shift due to substitution of Na+ by Tris+ (i.e. the difference between the white and red columns) of all mutants in this group is significantly smaller than that of hP2X7Rwt, indicating that they all have reduced selectivity for small cations. Group 3: Note that the S342K mutation further reduces the Vrev shift in Tris+Cl− versus NaCl compared to the mutants in group 2, indicating a further reduced selectivity for the small cation Na+; the additional mutation of E14 or D356 to lysine additionally produces a significant Vrev shift in Tris+Glu− versus Tris+Cl−, most likely reflecting an occurring anion permeability with reduced permeability to Glu− compared to Cl−. Group 4: Na+ substitution by Tris+ has no effect on Vrev, indicating a loss of cation permeability. The Vrev shift in Tris+Glu− versus Tris+Cl− was further strongly increased when the S342K mutation was combined either with D352K alone or with E14K and D356K together, an effect that could not be surpassed by mutating all three acidic residues to lysines or by additionally incorporating the S339A mutation.
Fig. 4.Cation selectivity filter at the cytoplasmic interface of hP2X7Rwt. Lateral (A) and bottom (B) views of the open hP2X7Rwt with labeled residues lining the open pore or located more peripherally (F350, I351, colored gray). The pore was modeled using MOLEonline. From the frequent detection of single lateral pores in the open hP2X7R channel by the MOLEonline software (see SI Appendix, Fig. S3) and consistent with the data shown here, we conclude that the fluid in the cavity below the tri-S342 ring (sky blue) and the lateral pores is continuum, with access controlled by the gate at S342 (selectivity filter 1). C) Location of filters 1 and 2 relative to phospholipid head groups. Our electrophysiological data indicate that cations exit through three exits (indicated by blue arrows), one between each subunit interface and each flanked by E14 of one subunit and D352/D356 of the adjacent subunit (size-selective filter 2). The membrane boundaries of the SWISS homology-modeled open hP2X7Rwt were determined using the “OPM” server (https://opm.phar.umich.edu/) with the conditions “Mammalian plasma membrane,” “Allow curvature yes,” “Topology N-term in” (50, 59). The red and blue pseudo-atoms were automatically added by the OPM server to mark the hydrophobic boundaries of the lipid bilayer. The downloaded PDF file was visualized with PyMOL. Note the proximity of the cation selectivity filter to the phospholipid headgroups and the apparent lack of coverage of the transmembrane channel by the main and side chains of flanking amino acid residues, both of which may be related to the effect of lipids on ion permeation (25) (channel portion in A) uncovered by amino acid main and side chains).
