Hindered diffusion through an aqueous pore describes invariant dye selectivity of Cx43 junctions

Abstract

The permselectivity (permeance/conductance) of Cx43-comprised gap junctions is a variable parameter of junctional function. To ascertain whether this variability in junctional permselectivity is explained by heterogeneous charge or size selectivity of the comprising channels, the permeance of individual Cx43 gap junctions to combinations of two dyes differing in either size or charge was determined in four cell types: Rin43, NRKe, HeLa43, and cardiac myocytes. The results show that Cx43 junctions are size- but not charge-selective and that both selectivities are constant parameters of junctional function. The consistency of dye selectivities indicates that the large continuum of measured junctional permselectivities cannot be ascribed to an equivalent continuum of individual channel selectivities. Further, the relative dye permeance sequence of NBD-M-TMA approximately Alexa 350 > Lucifer yellow > Alexa 488 >> Alexa 594 (Stokes radii of 4.3 A, 4.4 A, 4.9 A, 5.8 A, and 7.4 A, respectively) and the conductance sequence of KCl > TEACl approximately Kglutamate are well described by hindered diffusion through an aqueous pore with radius approximately 10 A and length 160 A. The permselectivity and dye selectivity data suggest the variable presence in Cx43-comprised junctions of conductive channels that are either dye-impermeable or dye-permeable.

Figure 1

Cx43 conductance measurements show a lack of charge selectivity among ions. (A) Channel event conductances (40 mV transjunctional voltage) determined from fitting channel event frequency histograms with single Gaussian peaks for KCl (324 events, two preparations), TEACl (148 events, one preparation), and Kglutamate (138 events, one preparation). (B) Plot of channel event conductance versus solution conductivity for Cx43wt junctions measured using KCl (solid circle), TEACl (open diamond), and Kglutamate (solid square) as primary-current-carrying ions. The dashed line represents conductances relative to KCl predicted by solution conductivities.

Figure 2

Dual dye charge selectivity measurement method. (A) An image sequence from a typical experiment comparing NBD-M-TMA (left column) to Alexa 350 (right column) (false color) (NRKe cells). Yellow scale bar, 10 μm. (B) Experimental data from A plotted as a function of time either as fluorescence normalized to total intensity (donor + recipient) at t₀ (upper panel) or as the –ln((F_eq – F_(t)/F_total(t/t0))/(F_eq – F₍₀₎)) (Eq. 8) (lower panel). Data from both dyes are well fit to a single rate constant (1.0 min⁻¹) by Eq. 7 (upper panel) or by linear regression of data points converted using Eq. 8 (lower panel), as evidenced by the fit lines through the data points.

Figure 3

Cx43 junctions are not charge-selective for diffusion of dyes. (A) Rate constants for NBD-M-TMA (B_NBD-M-TMA) versus Alexa 350 (B_{Alexa 350}) for multiple Cx43 junctions (solid circle) (n = 4, HeLa43; n = 12, Rin43; n = 25, NRKe; and n = 1, cardiac ventricular myocytes), cytoplasmic bridges (open triangles, n = 8), and Cx40 junctions (open squares, n = 6, Rin40). The dashed line represents relative calculated aqueous diffusion constants for NBD-M-TMA/Alexa 350 (Table 1). (B) Average B_NBD-M-TMA/B_Alexa350 (charge selectivity) for junctions shown in A. The dashed line represents relative calculated aqueous diffusion constants for NBD-M-TMA/Alexa 350. (asterisk) Cx40 junctions were significantly different (p < 0.05) from both Cx43 junctions and cytoplasmic bridges. Selectivity of Cx43 junctions was not significantly different from that of bridges.

Figure 4

Dual dye size selectivity measurement method. (A) Results from a typical dye experiment showing intercellular dye diffusion over time for Alexa 350 (left) and Alexa 594 (right). Yellow scale bar, 10 μm. (B, upper graph) Total cell fluorescence intensity values for donor (upper curves) and recipient (lower curves) cells normalized to the total (donor + recipient) intensity of the first image. Curves represent the fits of the data by Eq. 7 to determine the rate constants (B_dye) of diffusion (B_Alexa350 = 2.8 min⁻¹ and B_Alexa594 = 0.08 min⁻¹). (B, inset) Loss in total fluorescence intensity (donor + recipient) over time. The lines are fits to an exponential decay function with rate constants of 0.02 min⁻¹ (Alexa 350) and 0.002 min⁻¹ (Alexa 594). (B, lower graph) Plot of the intensities from the upper graph converted using Eq. 8 (see Methods). Lines represent linear regressions, the slopes of which equal the diffusion rate constants (B_Alexa350 = 2.7 min⁻¹ and B_Alexa594 = 0.08 min⁻¹). (C) Results from a typical dye experiment showing dye transfer over time for the same cell pair for Alexa 350 (left) and Alexa 488 (right). Yellow scale bar, 10 μm. (D, upper graph) The resulting total cell fluorescence intensity values for donor (upper curves) and recipient (lower curves) cells normalized to the total (donor + recipient) intensity of the first image. Curves represent fits of the data by Eq. 7 (B_Alexa350 = 2.96 min⁻¹ and B_Alexa488 = 0.77 min⁻¹). (D, lower graph) Plot of the intensities from E converted using Eq. 8. Lines are linear regressions, the slopes of which represent the diffusion rate constants (B_{Alexa 350} = 2.81 min⁻¹ and B_{Alexa 594} = 0.76 min⁻¹).

Figure 5

Cx43 junctions show constant size selectivity. (A) Plot of permeance rate constants (B_dye) for Lucifer yellow (solid circles) (Rin43, n = 6; NRKe, n = 3; HeLa43, n = 3); Alexa 488 (open triangles) (NRKe, n = 14; Rin43, n = 9; HeLa43, n = 5); and Alexa 594 (open squares) (NRKe, n = 11; Rin43, n = 10; HeLa43, n = 2; cardiac ventricular myocytes, n = 2) versus Alexa 350 (B_{Alexa 350}) from individual cell pairs for Cx43wt junctions. Lines are linear regressions for each dye combination. All combinations were significantly correlated (p < 0.05). (Inset) The same plot as in A for cytoplasmic bridges, where the dotted lines represent the regression fits from junctions. (B) Average B_{Alexa dye}/B_{Alexa 350} of experiments shown in A. (asterisk) All dye combinations are significantly different (p < 0.05) from cytoplasmic bridges for the same dye combination.

Figure 6

Dye diffusion is independent in dual dye measurements. Plot of average rate constants for Alexa 350, Lucifer yellow, Alexa 488, and Alexa 594 from the dual dye size selectivity experiments (dual dye) from NRKe and Rin43 cells (Alexa 350, n = 53; Lucifer yellow, n = 9; Alexa 488, n = 23; Alexa 594, n = 21) compared to the average rate constant for each dye when injected alone (single dye) using the same cell types (Alexa 350, n = 23; Lucifer yellow, n = 7; Alexa 488, n = 10; Alexa 594, n = 9). None of the dyes showed significantly different rate constants when measured alone compared to when measured in concert with another dye.

Figure 7

Permeance measurements yield a 10-Å pore radius estimate. Logarithmic scale plot of relative junctional permeances (B_dye/B_{Alexa 350}) versus calculated Stokes-Einstein radii (see Methods) for each of the dyes used in the size-selectivity measurements: Alexa 350 (solid circles), Lucifer yellow (black inverted triangle), Alexa 488 (solid square), and Alexa 594 (solid diamond). Also shown (gray inverted triangle) is the relative permeance of Lucifer yellow plotted against its radius corrected for diffusion across cytoplasmic bridges. Dashed lines represent predicted relative permeances for simple aqueous pores of the given channel radii according to Eq. 13. (Inset) Logarithmic scale plot of relative permeance (B_dye/B_{Alexa 350}) across cytoplasmic bridges versus calculated Stokes-Einstein radii for each of the dyes used in the size-selectivity measurements (symbols as in main figure). Also plotted here is the radius of Lucifer yellow predicted by its relative permeance to Alexa 350 (gray inverted triangle). Dashed line is a simple line graph connecting the Alexa series.

Figure 8

Model for independent regulation of conductance (A) and permeance (B) for Cx43 gap junctions. (A) A model for regulation of conductance independent of permeance by changing the number of dye-impermeable conductive channels without changing the number of dye-permeable conductive channels. (B) A model for regulation of permeance independent of conductance by conversion of channels between dye-permeable and dye-impermeable conductive states. In both A and B, the lower diagrams indicate the resulting changes in each parameter of junctional function that would occur as a result of the gradual shift in channel populations shown above.