Reduced Membrane Insertion of CLC-K by V33L Barttin Results in Loss of Hearing, but Leaves Kidney Function Intact

Abstract

In the mammalian ear, transduction of sound stimuli is initiated by K⁺ entry through mechano-sensitive channels into inner hair cells. K⁺ entry is driven by a positive endocochlear potential that is maintained by the marginal cell layer of the stria vascularis. This process requires basolateral K⁺ import by NKCC1 Na⁺-2Cl^--K⁺ co-transporters as well as Cl^- efflux through ClC-Ka/barttin or ClC-Kb/barttin channels. Multiple mutations in the gene encoding the obligatory CLC-K subunit barttin, BSND, have been identified in patients with Bartter syndrome type IV. These mutations reduce the endocochlear potential and cause deafness. As CLC-K/barttin channels are also expressed in the kidney, patients with Bartter syndrome IV typically also suffer from salt-wasting hyperuria and electrolyte imbalances. However, there was a single report on a BSND mutation that resulted only in deafness, but not kidney disease. We herein studied the functional consequences of another recently discovered BSND mutation that predicts exchange of valine at position 33 by leucine. We combined whole-cell patch clamp, confocal microscopy and protein biochemistry to analyze how V33L affects distinct functions of barttin. We found that V33L reduced membrane insertion of CLC-K/barttin complexes without altering unitary CLC-K channel function. Our findings support the hypothesis of a common pathophysiology for the selective loss of hearing due to an attenuation of the total chloride conductance in the stria vascularis while providing enough residual function to maintain normal kidney function.

Keywords: Bartter syndrome; CLC channel; barttin; hearing loss; patch clamp.

Figure 1

V33L barttin reduces macroscopic currents of CLC-K channels. (A) Transmembrane topology of barttin with the position of V33L indicated by an arrow. (B) Representative current recordings from HEK293T cells co-expressing ClC-Ka with WT or V33L barttin. (C) Voltage-dependence of steady-state currents in cells co-expressing ClC-Ka with WT (n = 16) or V33L barttin (n = 10). (D) Representative current recordings from HEK293T cells co-expressing ClC-Kb with WT or V33L barttin. (E) Voltage-dependence of steady-state currents in cells co-expressing ClC-Kb with WT (n = 50) or V33L barttin (n = 48). All error bars indicate s.e.m.

Figure 2

V33L barttin leaves single channel properties of ClC-Ka/barttin and ClC-Kb/barttin unaffected. (A) Representative noise analyses from cells co-expressing ClC-Ka with either WT (open circles) or V33L barttin (filled circles). The y-axis intercept provides the unitary conductance and the slope of the fitted line the number of protopores according to Equation 5. Inset, Mean values indicate similar unitary protopore conductances for ClC-Ka/barttin with WT or V33L barttin (V33L: n = 8, WT: n = 13, p = 0.3). (B) Voltage dependence of current variance by amplitude ratios for ClC-Kb/barttin with WT or V33L barttin according to Equation 6 (V33L: n = 34, WT: n = 41, p = 0.7 at −165 mV).

Figure 3

V33L does not affect association of barttin to ClC-Ka or ClC-Kb. (A,B) Representative fluorescence scans of SDS-PAGE gels demonstrate the co-purification of WT or V33L barttin-mCherry with an anti-GFP antibody targeting either eGFP-ClC-Ka (A) or eGFP-ClC-Kb (B). (C,D) Ratios of band intensities for co-purified barttin-mCherry to the intensity of the eGFP-ClC-Ka (C) or ClC-Kb (D) are similar for WT and V33L barttin (ClC-Ka: n = 4 each, p = 0.7; ClC-Kb: n = 4 each, p = 0.2).

Figure 4

V33L barttin reduces the ratio of whole-cell currents to eGFP-fluorescence intensities. (A) Plot of the steady-state current at −75 mV vs. whole-cell fluorescence for cells co-expressing ClC-Ka with WT (open circles) or V33L barttin (filled circles). (B) Mean current-fluorescence ratios are significantly lower in cells co-expressing ClC-Ka with V33L barttin (n = 14) than with WT barttin (n = 19, p = 0.003) indicating lower membrane insertion with V33L barttin. (C) Correlation between steady-state current at −175 mV and whole-cell fluorescence for cells co-expressing ClC-Kb with WT (open circles) or V33L barttin (filled circles). (D) Mean current-fluorescence ratios are significantly lower in cells co-expressing ClC-Kb with V33L barttin (n = 31) than with WT barttin (n = 40, p = 0.004).

Figure 5

Representative confocal images of living MDCK II cells confirm a lower membrane insertion of CLC-K channels when co-expressed with V33L barttin. (A,B) MDCK II cells expressing WT (A) or V33L barttin-mCherry (B) show a clearly visible membrane staining. (C,D) MDCK II cells co-expressing WT or V33L barttin-mCherry with ClC-Ka mVenus show a more clearly delineated membrane staining for ClC-Ka co-expressed with WT barttin (C) than with V33L barttin (D). (E,F) MDCK II cells co-expressing either barttin with ClC-Kb demonstrate a higher abundance of intracellular ClC-Kb channels when co-expressed with V33L barttin (F).

Figure 6

V33L barttin reduces the membrane insertion of ClC-Ka and ClC-Kb in MDCK II cells. (A,B) Representative fluorescence scan of a SDS-PAGE gel from the purified fraction after surface biotinylation (A, eluate) or the whole cell lysate (B, lysate). The upper panels depict eGFP-ClC-Ka or eGFP-ClC-Kb in different glycosylation states (“#” complex-glycosylated, “°” core-glycosylated, and ^“*” unglycosylated). The lower panels show WT or V33L barttin-mCherry. (C–E) Relative surface expression calculated as ratio of eluate to whole cell lysate intensity for barttin expresed alone (n = 6, p = 0.004), of ClC-Ka (D, n = 6) or ClC-Kb (E, n = 6) co-expressed with WT or V33L barttin.