Fig 2: Characterization and quantification of ClC-3 in LP2 fraction A, BCharacterization of SV-enriched LP2 fraction. (A) Western blots for ClC-3, ClC-4, synaptophysin, and PSD95 (postsynaptic marker) of fractions obtained during synaptic vesicle-enriched fraction (LP2) preparation using differential centrifugation as described (Huttner et al, 1983; Takamori et al, 2000). (B) Immunoblots for different cellular markers comparing whole lysate (H) and synaptic vesicle-enriched fraction (LP2) (GM130, Golgi marker; PDI, ER marker; rab4, early endosomes and SV; cathepsin D, lysosomal marker; synaptophysin, SV marker).C–EQuantification of ClC-3 in LP2 fractions (C) Western blot of lysates from HEK cells transiently overexpressing either mVGLUT1 or hClC-3, which were both fused at their N terminus to GFP. Probing for the GFP tag allows to directly compare the relative amount of ClC-3 and VGLUT1 in the lysate for its use as calibration standards for anti-ClC-3 and anti-VGLUT1 antibodies in Western blots of tissue standards [as in (D) and (E)]. (n = 3 independent transfections, horizontal lines and error bars represent mean values SEM). The standard contains about 4.33 × more ClC-3 molecules than the standard for VGLUT1. (D) Example immunoblot for VGLUT1 in which the GFP-VGLUT1 standard [same lysate as in (A)] is compared to different amounts of LP2 fraction. Below, example plot in which WB signals are plotted against rising amounts of the LP2 fraction, quantified by protein contents. The standard for VGLUT1 roughly corresponds to 2.5 µg of LP2 fraction (averaged from n = 6 Western blots). (E) Same as shown in (D) but for ClC-3. The standard for ClC-3 roughly corresponds to 140 µg of the LP2 fraction (n = 7). From (A) and (B), we calculated a molecular ratio for ClC-3/VGLUT1 ˜ 2.5/(140*4.33) = 4.12 × 10-3 = 1/243 in our LP2 fraction. Considering that ClC-3 may form not only hetero- but also homodimers and that SVs are ˜ 2-fold enriched in the SV compared to the LP2 fraction (Huttner et al, 1983) and that there are ˜ 10 copies of VGLUT1 per vesicle (Takamori et al, 2006), our data yield an upper limit of one ClC-3-containing transporter on roughly every 25th SV. This upper limit assumes that in LP2 fractions ClC-3 is only present on SVs. Any contamination with other ClC-3-containing compartments like endosomes, as found in this work, will lower this estimate. The ˜ 20-fold lower value for the ClC-3/VGLUT1 ratio given by (Schenck et al, 2009) may have been caused by their calibration of the VGLUT1 antibody (Synaptic Systems) that recognizes an epitope at the C terminus of VGLUT1. It might have been shielded by the GFP, which these authors fused directly to the carboxy terminus of VGLUT1, thereby overestimating the number of VGLUT1 molecules in their preparation (S. Takamori, personal communication).

Fig 3: Subcellular localization of VenusClC-3 and ClC-4 in neurons AGFP antibodies immunolabel VenusClC-3 (green) in the somata of CA1 pyramidal neurons of Clcn3 ven/ven, but not Clcn3 +/+ brain sections. sp, stratum pyramidale (scale bar: 20 µm).BCo-immunolabeling of VenusClC-3 (green) with ClC-4 (red) in the CA3 region of Clcn3 ven/ven mice (scale bar: 20 µm). sl, stratum lucidum; sp, stratum pyramidale.CCo-immunostaining of VenusClC-3 (green) with EEA1 (red, left panel), Lamp-1 (red, middle panel), or synaptophysin (red, right panel) in either CA1 or CA3 region of Clcn3 ven/ven mice. Arrows indicate structures of overlap (scale bar: 5 µm).DGFP antibody immunolabels VenusClC-3 (green) in neurites of cultured neurons. EEA1 and VGLUT1 are co-stained (both in red; left and right panel, respectively). The dendritic marker MAP2 is stained in blue (right panel; scale bars: 10 µm). DNA stained with DAPI.EImmunoblot analysis of fractions collected during preparation of SVs from brain of Clcn3 ven/ven mice. VenusClC-3 (detected with GFP antibody) co-purified with SV marker protein synaptophysin.FFluorescence of immobilized Venus-tagged SVs imaged with TIRF microscopy. Scale bar, 2 µm.G, HCo-localization between the Venus-tagged vesicles with antibodies against VGLUT1 and synaptophysin was analyzed. (G) 11.4 ± 0.6 (SD) of VGLUT1-labeled SVs and 8.2 ± 2 (SD) of synaptophysin-labeled SVs showed Venus fluorescence (n = 6). (H) 65 ± 5.4 (SD) and 79.4 ± 3.8 (SD) of Venus-tagged SVs showed co-localization with VGLUT1 and synaptophysin, respectively (n = 6).ICo-immunolabeling of VenusClC-3 (green) with synaptophysin (red) of immobilized SVs imaged with TIRF microscopy. Vesicle in quadrant delimited by dashed line is shown in higher magnification in merge. Scale bar, 2 µm. Source data are available online for this figure.