|
Figure 1. CNGB3 disease-associated mutations impart a gain-of-function phenotype in heteromeric CNG channels. (A) Diagram showing CNGB3 subunit topology with approximate locations of disease associated mutations examined in this paper. (B) Representative current traces from inside-out patches excised from Xenopus oocytes expressing heteromeric human A3+B3 channels activated by a saturating concentration (1 mM) of cGMP in the presence (arrowhead) or absence of 25 μM L-cis-diltiazem. Current traces were elicited using voltage steps from a holding potential of 0 mV, to +80 mV, −80 mV, and then returning to 0 mV. (C) Box plots summarizing the sensitivity to diltiazem block of channels containing disease-associated mutations compared to wild type A3+B3 heteromeric (WT) and CNGA3-only homomeric channels (A3). Data are expressed as the ratio of the 1 mM cGMP current amplitude at +80 mV in the presence and absence of diltiazem (as in B). Boxes represent 25th–75th percentiles, lines show the median, and whiskers represent the 5th–95th percentiles. (D) Summary of current densities at 5 μM cGMP for CNGB3 disease-associated mutations. Current-density estimates are based on cGMP dose-response data and estimates of patch area using patch electrode resistance (Sakmann and Neher, 1995). Four of the five mutations examined had an increase in the 5 μM current density compared to wild type A3+B3 channels (p < 0.001, single-factor ANOVA, n = 6–23; Holm's t-test, **p < 0.01, ***p < 0.001).
|
|
Figure 2. Macular degeneration-associated Y469D and L595F mutations in CNGB3 each increase the sensitivity of heteromeric A3+B3 channels to physiological ligands. (A) Representative dose-response curves for activation of A3+B3 wild type (WT; filled circles), A3+B3Y469D (Y469D; open circles), A3+B3G558C (G558C; open triangles), and A3+B3L595F (L595F; open squares) channels by cGMP. Currents were normalized to the maximum cGMP current (IMAX = 1.0). Continuous curves represent fits of the dose-response relationship to the Hill equation as indicated in the Materials and Methods. Best-fit parameters are as follows: WT, K1/2 = 15 μM, nH = 2.4; Y469D, K1/2 = 10 μM, nH = 1.4; G558C, K1/2 = 13 μM, nH = 2.1; L595F, K1/2 = 5.7 μM, nH = 2.1. Shaded area represents approximate physiological cGMP concentration in photoreceptors in the dark (Pugh and Lamb, 1993). (B) Summary of apparent cGMP affinity (K1/2) for channels formed by CNGA3 with wild-type CNGB3 or CNGB3 having indicated disease-associated mutations. Data are based on best-fit Hill curves and expressed as mean K1/2 (± S.E.M.). The K1/2 cGMP was significantly reduced for the Y469D and L595F groups compared to wild type (p < 0.001, Kruskal-Wallis, n = 12–22; ***p < 0.001, Mann-Whitney U-test). (C) Summary of mean Hill coefficients (nH) for channels formed by CNGA3 with wild-type CNGB3 or CNGB3 having disease-associated mutations. The nH was significantly reduced for the Y469D, G558C, and L595F groups compared to wild type (p < 0.001, single-factor ANOVA, n = 5–9; ***p < 0.001, **p < 0.01 Holm's t-test). (D) Representative current traces of heteromeric human A3+B3 channels after activation by a sub-saturating concentration of cGMP (5 μM). Currents were scaled to the maximum current (IMAX) in a saturating concentration of cGMP (1 mM) as illustrated by the WT current traces. Relative sub-saturating current amplitudes (I5μM/IMAX) are as follows: WT = 0.05; Y469D = 0.25; G558C = 0.13; L595F = 0.43. (E) Representative current traces after activation by a saturating concentration of cGMP (1 mM, black line) or a saturating concentration of cAMP (10 mM, gray line). (F) Summary of relative agonist efficacy (IcAMP/IcGMP) for channels containing wild-type CNGB3 or disease-associated mutations, expressed as mean (± S.E.M.). The efficacy of cAMP was significantly enhanced for the Y469D and L595F channels compared to wild type (p < 0.001, Kruskal-Wallis, n = 5–12; ***p < 0.001, Holm's t-test).
|
|
Figure 3. Effects of CNGB3 disease-associated mutations on the absolute agonist efficacy of cGMP and the spontaneous open probability. (A, Left) Overlay of four current traces from A3+B3 channels, activated by a sub-saturating concentration of cGMP (25 μM) to produce an approximately half-maximal current amplitude. Shaded area indicates the region of interest (ROI) used for the calculation of mean current and mean isochrone variance for each concentration of cGMP. (A, Right) ROI overlays at various cGMP concentrations (μM cGMP indicated on the right). Current and variance calculations were performed as described in the Materials and Methods. The mean isochrone variance is minimal at extreme low and high open probabilities and maximizes at 50% open probability. For wild type cone CNG channels, ~50% open probability is achieved at cGMP concentrations near the K1/2. The mean isochrone variance at saturating cGMP (i.e., 1 mM) reflects the maximum open probability (PO, MAX), such that a decreased variance indicates an increased PO, MAX (provided that PO, MAX > 50%). (B) Current-variance plot for A3+B3 wild type (closed circles) and A3+B3L595F (open circles) after activation by cGMP at +80 mV. The L595F mean current and variance estimates were scaled to WT IMAX to facilitate comparison of the initial slope from zero current and the variance at saturating ligand concentrations. Continuous lines produced by fitting with a 2nd order polynomial, as described in the Materials and Methods, using the following best-fit parameters: WT (solid), i = 3.1 pA, N = 1550 channels; L595F (hashed), i = 3.2 pA, N = 1370 channels. The mean variance at saturating cGMP is reduced for L595F channels (gray arrowhead) compared to WT channels (black arrowhead), indicating an increased PO, MAX for L595F channels. (C) Summary of maximal open probabilities after activation with saturating cGMP (1 mM). Maximal open probabilities were calculated as described in the Materials and Methods. The PO, MAX was significantly increased for Y469D and L595F channels compared to wild type (p < 0.01, single-factor ANOVA, n = 4–9; Holm's t-test, *p < 0.05, **p < 0.01). (D) Summary of unitary conductance values based on best-fit polynomials described in (B). The CNGB3 disease-associated mutations did not significantly alter the unitary conductance (p = 0.28; single-factor ANOVA). (E) Relationship between the spontaneous isochrone variance (σ2SP) and estimated number of channels (N) in the membrane patch for WT (black circles), L595F (open circles) and A3 only (gray circles) containing channels. The σ2SP and N estimates were determined as described in the Materials and Methods. Data fit with linear models using the following best-fit slopes (with the units: A2/channel x10−26): A3+B3 WT = 1.53; A3+B3 L595F = 4.98; A3 WT = 0.58. The slope of the relationship between σ2SP and N (dσ2SP/dN) was elevated for L595F channels compared to WT, reflecting an increase in the spontaneous open probability (PO, SP) for L595F channels. (F) Spontaneous open probabilities were calculated from trend-line slopes (dσ2SP/dN) from E. PO, SP calculations were based on the relationship: dσ2/dN = i2Poq, as described in Materials and Methods. Error estimates of PO, SP were propagated from the standard error of slopes from (E). We observed that the spontaneous open probability is significantly lower for A3 homomeric channels (A3WT, PO, SP = 0.50 ×10−3) compared WT heteromeric channels (A3+B3 WT, PO, SP = 1.6 x10−3) (***p < 0.001, extra sum-of-squares F-test, n = 6, 7). These estimates are in reasonable agreement with previous characterization of mouse cone CNG channels (Gerstner et al., 2000). We also observed that the PO, SP is significantly increased for L595F-containing channels (A3+B3 L595F, PO, SP = 4.3 ×10−3) compared to WT heteromeric channels (**p < 0.001, extra sum-of-squares F-test, n = 7, 8), reflecting a change in the intrinsic gating properties of L595F-containing channels.
|
|
Figure 4. Effects of single nucleotide polymorphisms on ligand sensitivity of heteromeric A3+B3 channels. (A) CNGA3 and CNGB3 single nucleotide polymorphisms (SNPs) examined here form functional A3+B3 heteromeric channels. Box plots summarizing the sensitivity to block by 25 μM diltiazem in the presence of 1 mM cGMP for SNP-containing channels compared to wild type A3+B3 and A3-only channels. (B) Summary of K1/2 cGMP for wild type and SNP-containing A3 or A3+B3 channels. Data are based on best-fit Hill parameters and expressed as mean K1/2 (± S.E.M.). The K1/2 cGMP was significantly increased for the A3+B3W234C and A3T153M+B3 groups compared to wild type A3+B3 channels (p < 0.001, single-factor ANOVA, n = 4–7; Holm's t-test, **p < 0.01, ***p < 0.001). There was no significant difference between wild-type A3 and A3T153M homomeric channels (Holm's t-test, p > 0.05). (C) Summary of relative agonist efficacy for wild type and SNP-containing A3 or A3+B3 channels. The relative efficacy of cAMP was significantly decreased for A3+B3W234C and A3T153M+B3 channels compared to wild type A3+B3 channels (p < 0.001, Kruskal-Wallis, n = 4–7; Holm's t-test, *p < 0.05). There was no significant difference between wild-type A3 and A3T153M homomeric channels (Student's t-test, p > 0.05). (D) Summary of 5 μM cGMP current densities for select CNG channel polymorphisms. The estimated 5 μM current density for heteromeric A3T153M+B3 channels was decreased compared to wild type A3+B3 channels (p < 0.05, single-factor ANOVA, n = 4–8; Holm's t-test, *p < 0.05).
|
|
Figure 5. CNGA3-T153M polymorphism decreases the maximum open probability for heteromeric but not homomeric CNG channels. (A) Current-variance plot for A3T153M homomeric (open circles) and A3T153M+B3 (closed circles) channels following activation by various concentrations cGMP at +80 mV (as in Figure 3). Continuous lines were produced by fitting with a 2nd-order polynomial using the following best-fit parameters: A3T153M (solid), i = 2.5 pA, N = 370 channels; A3T153M+B3 (hashed), i = 2.5 pA, N = 370 channels. Arrows denote mean variance at saturating cGMP for A3T153M (solid) and A3T153M+B3 (hashed). (B) Maximum open probability was decreased for heteromeric A3T153M+B3, but not homomeric A3T153M channels compared to wild-type channels (***p < 0.001, Student's t-test, n = 4–8).
|
|
Figure 6. Gain-of-function/loss-of-function continuum for cone CNG channel disease-associated mutations and polymorphisms based on current densities at a physiological concentration of cGMP. Plot comparing 5 μM current density ratios for CNGB3 disease associated mutations and CNG channel polymorphisms with wild type A3+B3 channels (star). The plot includes CNGB3 disease-associated mutations described here (open circles) and several CNGB3 disease-associated mutations described previously (Peng et al., 2003b—open triangles; Bright et al., 2005—open squares). Channels with a current-density ratio greater than 1.0 are predicted to generate an elevated photoreceptor dark current compared to wild type A3+B3 channels and thus can be characterized as having a putative gain-of-function phenotype. Conversely, channels with a current-density ratio less than 1.0 are predicted to generate an attenuated photoreceptor dark current compared to wild type channels and can thus be characterized as having a putative loss-of-function phenotype.
|
