|
Figure 1. Quantification of SLC4A11 expression. (A) Confocal immunofluorescence of oocytes injected with the indicated RNA transcripts. HA-tagged proteins were detected using mouse anti-HA antibody and visualized using chicken anti-mouse IgG conjugated with Alexa Fluor 488 (green). (B) Oocytes, injected with cRNA encoding HA-epitope tagged SLC4A11, AQP1, NIP5;1 or with water alone were fractionated to produce a membrane-rich extract. Samples corresponding to 10 oocytes were probed on immunoblots with anti-HA antibody and caveolin antibody as a marker of the amount of membrane protein present, to enable measurement of the amount of HA-tagged protein/oocyte. (C) HEK293 cells were transfected with SLC4A11 cDNA or with vector alone were probed on immunoblots with a polyclonal antibody generated in rabbit against a synthetic peptide corresponding to amino acids 37–50 of human SLC4A11, in the putative cytoplasmic N-terminal domain of the protein. (D) AE1-Ct (40 ng protein) and 50 µg of protein from HEK293 cells transfected to express HA-epitope-tagged human AE1 (25) were immunoblotted and probed with mouse monoclonal anti-AE1 antibody, IVF12 (46). Densitometry of the immunoblot revealed the relative amounts of AE1 antigen present in the two preparations. (E) HEK293 cell lysate, containing 60 ng AE1, was electrophoresed along with a lysate of X. laevis oocyte membranes, corresponding to 10 oocytes. Densitometry revealed that each oocyte therefore contains 15 ng SLC4A11, corresponding to 9.1 × 109 SLC4A11/oocyte.
|
|
Figure 2. Water-flux through SLC4A11. (A) Xenopus oocytes were injected with H2O or H2O containing RNA transcripts encoding N-terminally HA-tagged SLC4A11, AQP1, NIP5;1 or untagged hCNT3. Osmotic swelling was induced by perfusing the oocytes with 44 mOsm/kg hypotonic solution at time zero, and oocyte volume was calculated from images captured digitally every 15 s. Top of panel diagrammatically shows a spherical oocyte swelling in response to hypo-osmotic medium. (B) Osmotic water permeabilities (Pf) were calculated (23) for oocytes and corrected for the value observed in H2O-injected oocytes (2.9 ± 0.09 × 10–3 cm s−1). Bars represent mean ± SE from three separate swelling assays with 10 oocytes per assay. N.S., not significant. (C) HEK293 cells were transiently co-transfected with eGFP cDNA and SLC4A11 (WT and R125H mutant), hCNT3, AQP1 cDNAs or empty vector. Cells were perfused alternately with isotonic (black bar) and hypotonic (white bar) medium. eGFP fluorescence in regions of interest was measured digitally from images captured every 250 ms, but here only one data point every 3 s is shown. Top of panel diagrammatically represents a eGFP-expressing cell swelling in response to hypotonic medium. White circle represents the region of interest monitored to measure the change of eGFP concentration. (D) Rate of fluorescence change, calculated by linear regression of the initial intensity change during the first 15 s of perfusion with hypotonic buffer, was corrected for activity of vector-transfected cells (0.19 ± 0.005 s−1). Data represent the mean ± SE of six independent swelling experiments with 30–40 cells measured per assay. *Significant difference (P < 0.01) compared with hCNT3.
|
|
Figure 3. Loss of water-flux upon mutation of an NPS motif. (A) Schematic topology models for SLC4A11 and for the AQP family of proteins, with the outside of the cell at the top. Positions of conserved Asparagine–Proline-X (NPX) motifs are indicated by arrows. (B) Amino acid sequence alignments of the region around the NPX motif found in SLC4A11 (boxed). Species are: Human (Homo sapiens (Hs)), Chimpanzee (Pan troglodytes (PT)), Mouse (Mus Musculus, Mm), Rat (Rattus norvegicus, Rn), Guinea pig (Cavia porcellus, Cp), Dog (Canis lupus familiaris, Cl), Pig (Sus Scrofa, Ss), Horse (Equss caballus, Ec), Cattle (Bos Taurus, Bt) and Chicken (Gallus gallus, Gg). The degree of conservation is indicated at the top by amino acid frequency, represented by the relative height of each single-letter coded amino acid (WebLogo software) (47). (C) Water permeability of HEK293 cells expressing WT or N639A SLC4A11 was assessed, by following the rate of change of eGFP fluorescence after cells were exposed to hypotonic medium. (D) Comparison of WT and N639A rates of cell swelling following hypo-osmotic challenge. The rate of fluorescence change was corrected for activity of vector-transfected cells (0.19 ± 0.02 s−1). Data represent mean ± SE of three independent swelling experiments with 30–40 cells measured per assay.
|
|
Figure 4. Water-flux is proportional to SLC4A11 expression level and is inhibited by stilbene disulfonates. (A and B) HEK293 cells were co-transfected with eGFP cDNA and indicated amount of cDNA encoding HA-tagged wild-type (WT) SLC4A11. (A) SLC4A11 expression was quantified on immunoblots of cell lysates probed with anti-HA antibody. (B) Cell swelling of HEK293 cells, expressing various amounts of SLC4A11, was assessed by the eGFP dilution assay (as in Fig. 2C). The rate of swelling observed for vector-alone transfected cells was subtracted from the values observed for cells expressing SLC4A11. (C) Chemical structures of SLC4A11 inhibitors used here. (D and E) HEK293 cells were co-transfected with eGFP and either SLC4A11, AQP1 or vector cDNA. Water-flux following shift to hypo-osmotic medium, assessed by measuring the rate of eGFP fluorescence change (as in Fig. 2C), was corrected for rate in vector-transfected cells. (D) Water-flux assays were performed in the presence of varied concentrations of DNDS, a non-covalently acting compound classically used to inhibit proteins of the SLC4 family (48). Half-maximal effective concentration determined from the curve is 24 ± 1 µM DNDS, n = 3. (E) eGFP dilution water-flux assays were performed in the absence or presence of 100 µM H2DIDS, a covalently acting stilbene disulfonate. Rates of fluorescence change were corrected for the value found for vector-transfected cells. *P < 0.05. N.S., not significant. n = 3.
|
|
Figure 5. Corneal oedema and progressive corneal dysfunction in slc4A-a11â/â mice. (A) Slit lamp examination of WT and slc4a11â/â mice. Corneas of 48-week WT and slc4a11â/â mice seen under retro-illumination (top panels) and examined using tangential slit illumination (bottom panels). Arrows indicate corneal haze observed in slc4a11â/â mice. (B) Corneal thickness measured by in vivo confocal microscopy (n = 5 per genotype). Central corneal thickness was significantly different between WT and slc4a11â/â mice at all time points (P < 0.01). The corneal thickness in heterozygous (slc4a11+/â) mice was not different compared with wild-type mice. (C) Scanning electron micrographs of corneal endothelia from WT and slc4a11â/âmice at 32 weeks. Scale bars are 20 µm. (D) Statistical analysis of corneal endothelial cell density.
|
|
Figure 6. SLC4A11 and AQP1 in human cornea. (A) Bright field and confocal immunofluorescence microscopy images of paraffin-embedded human cornea. Cornea sections were incubated with rabbit anti-SLC4A11 or rabbit anti-AQP1 serum, followed by goat anti-rabbit IgG conjugated with Alexa Fluor 594 (red). Nuclei were detected with DAPI (blue). Scale bars represent 10 µm. (B) The endothelial cell layer separates the corneal stroma from the aqueous humour. Fluid accumulates in the stroma, which is reabsorbed by the endothelial cell layer to prevent corneal oedema. Localization of AQP1 and SLC4A11 to opposite poles of endothelial cells suggests they may work cooperatively to provide a water conductive pathway through the endothelial layer. T.J. represents the tight junction that connects cells of the endothelium.
|
|
Figure 1. Quantification of SLC4A11 expression. (A) Confocal immunofluorescence of oocytes injected with the indicated RNA transcripts. HA-tagged proteins were detected using mouse anti-HA antibody and visualized using chicken anti-mouse IgG conjugated with Alexa Fluor 488 (green). (B) Oocytes, injected with cRNA encoding HA-epitope tagged SLC4A11, AQP1, NIP5;1 or with water alone were fractionated to produce a membrane-rich extract. Samples corresponding to 10 oocytes were probed on immunoblots with anti-HA antibody and caveolin antibody as a marker of the amount of membrane protein present, to enable measurement of the amount of HA-tagged protein/oocyte. (C) HEK293 cells were transfected with SLC4A11 cDNA or with vector alone were probed on immunoblots with a polyclonal antibody generated in rabbit against a synthetic peptide corresponding to amino acids 37–50 of human SLC4A11, in the putative cytoplasmic N-terminal domain of the protein. (D) AE1-Ct (40 ng protein) and 50 µg of protein from HEK293 cells transfected to express HA-epitope-tagged human AE1 (25) were immunoblotted and probed with mouse monoclonal anti-AE1 antibody, IVF12 (46). Densitometry of the immunoblot revealed the relative amounts of AE1 antigen present in the two preparations. (E) HEK293 cell lysate, containing 60 ng AE1, was electrophoresed along with a lysate of X. laevis oocyte membranes, corresponding to 10 oocytes. Densitometry revealed that each oocyte therefore contains 15 ng SLC4A11, corresponding to 9.1 × 109 SLC4A11/oocyte.
|
|
Figure 2. Water-flux through SLC4A11. (A) Xenopus oocytes were injected with H2O or H2O containing RNA transcripts encoding N-terminally HA-tagged SLC4A11, AQP1, NIP5;1 or untagged hCNT3. Osmotic swelling was induced by perfusing the oocytes with 44 mOsm/kg hypotonic solution at time zero, and oocyte volume was calculated from images captured digitally every 15 s. Top of panel diagrammatically shows a spherical oocyte swelling in response to hypo-osmotic medium. (B) Osmotic water permeabilities (Pf) were calculated (23) for oocytes and corrected for the value observed in H2O-injected oocytes (2.9 ± 0.09 × 10–3 cm s−1). Bars represent mean ± SE from three separate swelling assays with 10 oocytes per assay. N.S., not significant. (C) HEK293 cells were transiently co-transfected with eGFP cDNA and SLC4A11 (WT and R125H mutant), hCNT3, AQP1 cDNAs or empty vector. Cells were perfused alternately with isotonic (black bar) and hypotonic (white bar) medium. eGFP fluorescence in regions of interest was measured digitally from images captured every 250 ms, but here only one data point every 3 s is shown. Top of panel diagrammatically represents a eGFP-expressing cell swelling in response to hypotonic medium. White circle represents the region of interest monitored to measure the change of eGFP concentration. (D) Rate of fluorescence change, calculated by linear regression of the initial intensity change during the first 15 s of perfusion with hypotonic buffer, was corrected for activity of vector-transfected cells (0.19 ± 0.005 s−1). Data represent the mean ± SE of six independent swelling experiments with 30–40 cells measured per assay. *Significant difference (P < 0.01) compared with hCNT3.
|
|
Figure 5. Corneal oedema and progressive corneal dysfunction in slc4A-a11â/â mice. (A) Slit lamp examination of WT and slc4a11â/â mice. Corneas of 48-week WT and slc4a11â/â mice seen under retro-illumination (top panels) and examined using tangential slit illumination (bottom panels). Arrows indicate corneal haze observed in slc4a11â/â mice. (B) Corneal thickness measured by in vivo confocal microscopy (n = 5 per genotype). Central corneal thickness was significantly different between WT and slc4a11â/â mice at all time points (P < 0.01). The corneal thickness in heterozygous (slc4a11+/â) mice was not different compared with wild-type mice. (C) Scanning electron micrographs of corneal endothelia from WT and slc4a11â/âmice at 32 weeks. Scale bars are 20 µm. (D) Statistical analysis of corneal endothelial cell density.
|
|
Figure 6. SLC4A11 and AQP1 in human cornea. (A) Bright field and confocal immunofluorescence microscopy images of paraffin-embedded human cornea. Cornea sections were incubated with rabbit anti-SLC4A11 or rabbit anti-AQP1 serum, followed by goat anti-rabbit IgG conjugated with Alexa Fluor 594 (red). Nuclei were detected with DAPI (blue). Scale bars represent 10 µm. (B) The endothelial cell layer separates the corneal stroma from the aqueous humour. Fluid accumulates in the stroma, which is reabsorbed by the endothelial cell layer to prevent corneal oedema. Localization of AQP1 and SLC4A11 to opposite poles of endothelial cells suggests they may work cooperatively to provide a water conductive pathway through the endothelial layer. T.J. represents the tight junction that connects cells of the endothelium.
|
