TMEM16B, a novel protein with calcium-dependent chloride channel activity, associates with a presynaptic protein complex in photoreceptor terminals

Abstract

Photoreceptor ribbon synapses release glutamate in response to graded changes in membrane potential evoked by vast, logarithmically scalable light intensities. Neurotransmitter release is modulated by intracellular calcium levels. Large Ca(2+)-dependent chloride currents are important regulators of synaptic transmission from photoreceptors to second-order neurons; the molecular basis underlying these currents is unclear. We cloned human and mouse TMEM16B, a member of the TMEM16 family of transmembrane proteins, and show that it is abundantly present in the photoreceptor synaptic terminals in mouse retina. TMEM16B colocalizes with adaptor proteins PSD95, VELI3, and MPP4 at the ribbon synapses and contains a consensus PDZ class I binding motif capable of interacting with PDZ domains of PSD95. Furthermore, TMEM16B is lost from photoreceptor membranes of MPP4-deficient mice. This suggests that TMEM16B is a novel component of a presynaptic protein complex recruited to specialized plasma membrane domains of photoreceptors. TMEM16B confers Ca(2+)-dependent chloride currents when overexpressed in mammalian cells as measured by halide sensitive fluorescent protein assays and whole-cell patch-clamp recordings. The compartmentalized localization and the electrophysiological properties suggest TMEM16B to be a strong candidate for the long sought-after Ca(2+)-dependent chloride channel in the photoreceptor synapse.

Figure 1.

Topological model of TMEM16B. TMEM16B contains eight putative transmembrane segments indicated as gray cylinders. The large N-terminal domain and the C terminus with the conserved PDZ class I binding motif are predicted to be present in the cytoplasmic space. Cysteine residues that are highly conserved among the TMEM16 protein family are encircled. Mutation of one of these cysteines (asterisk) in the TMEM16E paralog causes gnathodiaphyseal dysplasia 1. The positions of four N-linked glycosylation sites in the first and third extracellular loop are indicated; the encircled one is highly conserved among the TMEM16 paralogues.

Figure 2.

Expression analysis. A, RT-PCR assay to analyze TMEM16B expression in a panel of 20 human tissues. G3PDH served as a control. B, RT-PCR to analyze Tmem16b expression in a panel of 15 mouse tissues. Gusb served as a control.

Figure 3.

Biochemical properties of TMEM16B. A, Western blot analysis of MBP fusion proteins containing N- and C-terminal TMEM16B protein fragments and total protein lysates of 293-EBNA cells overexpressing TMEM16B with monoclonal antibody TMEM16B-7H10 and polyclonal antibody TMEM16B-303. B, C, Western blot analysis (B) of TMEM16B immunoprecipitated from 293-EBNA cell after digestion with EndoH and PNGase F glycosidases and (C) of TMEM16B in mouse retina and brain homogenates. D, Immunofluorescence detection of TMEM16B in 293-EBNA cells. E, Western blot analysis of TMEM16B solubilized with Triton X-100 or SDS. P, Insoluble pellet; S, supernatant. Data shown in B–E were generated with TMEM16B-7H10 antibodies.

Figure 4.

Localization of TMEM16B in mouse retina. A, Immunofluorescence microscopy on a cryosection of mouse retina labeled with TMEM16B-7H10. Differential interference contrast image on the left shows the different retinal layers. OS/IS, Outer and inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. B, Outer plexiform layer simultaneously stained with the TMEM16B-7H10 (red) and antibodies toward PSD95, VELI3, and synaptophysin. Yellow staining in the merged images (right) indicates colocalization. White arrows point to synaptic terminals with plasma membranes labeled by TMEM16B-7H10 (red) and synaptic vesicles stained by synaptophysin antibodies (green). Scale bars: 25 μm.

Figure 5.

Interaction of TMEM16B with PSD95 and MPP4. A, Bacterially expressed GST (control) or GST fusion protein containing the C terminus of TMEM16B (GST+C-term) were incubated with lysates from 293-EBNA cells transfected with PSD95, VELI3, and MPP4. B, GST-pulldown assay performed with a TMEM16B deletion mutant lacking the C-terminal three amino acids (GST+C-termΔ3). C, GST (control) or GST+C-term fusion proteins were incubated with 293-EBNA cell lysates transfected with β-PSD95, MPP4 or cotransfected with MPP4 and α- or β-PSD95. A–C, Bound proteins were identified by immunoblotting with antibodies as indicated below the respective images. Inputs represent 2% of total cell lysates in all experiments.

Figure 6.

Loss of TMEM16B from Mpp4 ^−/− photoreceptor synapses. A, B, Retinal sections from wild-type (A) and Mpp4 ^−/− mutant (B) mice were double-labeled with TMEM16B-7H10 (green) and MPP4 (red) antibodies. Yellow color in the merged wild-type image (A) indicates colocalization of TMEM16B and MPP4 in the OPL. Scale bars: 50 μm.

Figure 7.

Halide-sensitive YFP assay demonstrates Ca⁺²-dependent anion conductance in HEK293 cells expressing TMEM16B. A, Representative traces show fluorescence (arbitrary units) in HEK293 cells transfected with YFP-I152L (dashed line) or with YFP-I152L and TMEM16B (solid line). Fluorescence quenching was induced by isotonic application of 20 mm I⁻ (gray arrow) and 1 μm ionomycin (black arrow). B, Summary of I⁻ influx-mediated maximal fluorescence quenching (mean ± SEM). Gray boxes show maximal quenching after the addition of I⁻, black boxes demonstrate maximal quenching after ionomycin treatment. Addition of ionomycin induced significant fluorescence quenching in all YFP- I152L-transfected cells (*p < 0.0001; paired Student's t test) (n = 51). Ionomycin-mediated fluorescence quenching was significantly enhanced when YFP-I152L was cotransfected with TMEM16B (**p < 0.001; ANOVA) (n = 61). Pretreatment with DIDS had no effect on ionomycin-induced I⁻ influx in cells expressing TMEM16B (n = 31).

Figure 8.

Whole-cell patch-clamp analysis of HEK293 cells expressing TMEM16B. A, B, Representative voltage-clamp recordings from a HEK293 cell transfected with YFP (A) and YFP and TMEM16B (B) stimulated with a ramp protocol shown in the inset in A. Membrane potentials were clamped at a holding potential of −40 mV and the cells were depolarized to 50 mV in 10 mV steps in 50 ms followed by −10 mV steps to −130 mV in 50 ms. Intracellular Ca²⁺ concentration was 100 nm. C, Current–voltage relationships from experiments shown in A and B. The currents in the I–V plots are expressed as current densities which were calculated as described previously (Wesolowski et al., 2007). Each value represents the mean current densities (± SEM) measured from cells at each voltage pulse. D–G, Representative voltage-clamp recordings and current–voltage relationships from HEK293 cells transfected with YFP (D, E) and YFP and TMEM16B (F, G) obtained by the protocol shown in the inset in D. The membrane currents were continuously recorded at a holding potential of −40 mV for 125 s. During this time the cells were repeatedly stimulated every 2.5 s by −20 mV voltage steps (each 100 ms) to −140 mV followed by depolarization steps to 60 mm. Treatment with 1 μm ionomycin is indicated by black bars. The mean current densities (± SEM) of cells with and without ionomycin application (control) were plotted against the test potentials of the electrical stimulation. Insets in F show magnifications of current deflections obtained with HEK293 cells expressing TMEM16B before and after ionomycin application. H, I, Current–voltage relationships from HEK293 cells transfected with YFP and TMEM16B obtained with Ringer or low-Cl⁻-Ringer solution by the protocol shown in A. Current–voltage relationship in I was measured after ionomycin application. Replacing Ringer bath solution with symmetric Cl⁻ reduced current amplitudes and caused statistically significant shifts of the reversal potentials to positive values (p < 0.001) (J).