Tumor-associated calcium signal transducer 2 is required for the proper subcellular localization of claudin 1 and 7: implications in the pathogenesis of gelatinous drop-like corneal dystrophy

Abstract

Gelatinous drop-like dystrophy (GDLD) is a rare autosomal recessive form of corneal dystrophy characterized by subepithelial amyloid depositions on the cornea. Previous clinical and laboratory observations have strongly suggested that epithelial barrier function is significantly decreased in GDLD. Despite the decade-old identification of the tumor-associated calcium signal transducer 2 (TACSTD2) gene as a causative gene for GDLD, the mechanism by which the loss of function of this causative gene leads to the pathological consequence of this disease remains unknown. In this study, we investigated the functional relationship between the TACSTD2 gene and epithelial barrier function. Through the use of immunoprecipitation and a proximity ligation assay, we obtained evidence that the TACSTD2 protein directly binds to claudin 1 and 7 proteins. In addition, the loss of function of the TACSTD2 gene leads to decreased expression and change in the subcellular localization of tight junction-related proteins, including claudin 1, 4, 7, and ZO1 and occludin, both in diseased cornea and cultured corneal epithelial cells. These results indicate that loss of function of the TACSTD2 gene impairs epithelial barrier function through decreased expression and altered subcellular localization of tight junction-related proteins in GDLD corneas.

Figure 1

Expression of the TACSTD2 protein in cornea (A), conjunctiva (B), skin (C), pharynx (D), esophagus (E), stomach (F), small intestine (G), colon (H), bladder (I), uterine cortex (J), and vagina (K). For negative control, normal cornea was immunostained with normal mouse IgG1 or IgG2a. L: Green indicates the signal for the TACSTD2 protein and red indicates the signal for propidium iodide. Original magnification was ×20 (C–K) or ×40 (A, B, and L).

Figure 2

Relationship between the TACSTD2 protein and TJ-related proteins. A: HCE-T-15 cells were lysed and immunoprecipitated using an anti-TACSTD2 antibody that was raised in a mouse (M) or goat (G). The immunoprecipitants were electrophoresed and immunoblotted using antibodies against the indicated proteins. B: PLA analysis was performed on normal corneal tissue using pairs of antibodies against the TACSTD2 and the CLDN1, 4, 7, TJP1, and OCLN proteins. For a negative control, PLA analysis was performed using normal mouse IgG(IgG₁ or IgG₂a) and normal goat IgG. Original magnification was ×40. C: PLA analysis was performed using pairs of antibodies against the TACSTD2 and the CLDN4 or 7 proteins and further immunostained with antibody against desmoplakin or golgin. Red indicates PLA signal, green indicates the signal of desmoplakin or golgin, and blue indicates nucleus stained with 4′,6′-diamidino-2-phenylindole. Original magnification was ×252.

Figure 3

Knockdown of the TACSTD2 gene leads to decreased epithelial barrier function accompanied with change in expression level and subcellular localization of the TJ-related proteins in HCE-T-15 cells. A: Effect of the lentiviral introduction of shRNA vectors against the TACSTD2 gene on the expression of the TACSTD2 gene. The vertical bar indicates the percentile inhibition of the TACSTD2 mRNA where expression of vehicle was set to 100%. Compared with the controls (shRNA-nontarget, mock, vehicle), two shRNA vectors against the TACSTD2 gene significantly down-regulated TACSTD2 mRNA at PID 4. B: Effect of the lentiviral introduction of shRNA vectors against the TACSTD2 gene on the expression of the TACSTD2 protein at PID 1, 4, and 7 in the HCE-T-15 cells. Compared with the vehicle at PID 7, the expression level of the TACSTD2 protein was gradually decreased day by day in the shRNA-introduced HCE-T-15 cells. Original magnification was ×40. C: Effect of the lentiviral introduction of shRNA vectors against the TACSTD2 gene on the epithelial barrier function in HCE-T-15 cells. The horizontal bar indicates the number of days postinfection (PID) and the vertical bar indicates TER in Ω · cm². D: Effect of the lentiviral introduction of shRNA vector against the TACSTD2 gene on the expression of the CLDN1, CLDN4, CLDN7, OCLN, and TJP1 proteins. Compared with the vehicle, the shRNA-introduced HCE-T-15 cells exhibited the decreased expression in the CLDN1 and seven proteins and altered subcellular localization from plasma membrane to cytoplasm or nucleus in all of those TJ-related proteins. Original magnification was ×40. E: Magnified images of D to indicate the subcellular localization of the TJ-related proteins. F: Results of Western blot analysis to examine the effect of shRNA introduction on the expression of the CLDN1, CLDN4, and CLDN7 proteins. The expression levels of the CLDN1 and seven proteins were decreased in the HCE-T-15 cells introduced with shRNA against the TACSTD2 gene (lane 1, shRNA-TACSTD2-1; lane 2, shRNA-TACSTD2-3) compared with the HCE-T-15 cells introduced with shRNA-nontarget (lane 3) or mock vector (lane 4). “shRNA-nontarget” means that the lentiviral vector expressing shRNA, which was designed to have no interference with any human genes. “Mock” means the lentiviral vector without an shRNA cassette. “Vehicle” means absence of lentiviral vector infection but presence of the treatment with the same concentration of polybrene for the same period. “Nontreated” means absence of lentiviral infection as well as polybrene treatment.

Figure 4

Knockdown of either of the CLDN1, 4, and 7 proteins significantly decreased the epithelial barrier function in the HCE-T-15 cells. HCE-T-15 cells were infected with lentiviral vectors expressing shRNA against the CLDN1 (A and B), 4 (C and D), and 7 (E and F) proteins. TER (A, C, and E) was measured up to seven days after the infection. The vertical bars indicate TER in Ω · cm². The horizontal bars indicate PID. Bar graphs (B, D, and F) demonstrate the effect of the lentiviral introduction of shRNA vectors on the expression of those genes. Vertical bar indicates the percentile inhibition of those mRNA where expression of the vehicle was set to 100%. Compared with the controls (shRNA-nontarget, mock, vehicle), for each of those genes, three shRNA vectors significantly down-regulated mRNA of their target at PID4.

Figure 5

Effect of the forcedly expressed TACSTD2 gene in HeLa cells on the protein expression of the CLDN1, 4, and 7 proteins. A: TER of HeLa cells was significantly low for more than two weeks after they reached confluence. B: RNA expression of TJ-related proteins in HeLa cells. Compared with HCE-T-15 cells, the expression level of the TACSTD2, CLDN4, and CLDN7 proteins was significantly low in HeLa cells. C: Results of immunostaining analysis against the HeLa cells transfected with the CLDN1, 4, or 7 genes together, with or without the TACSTD2 gene, using antibodies against the TACSTD2, CLDN1, CLDN4, or CLDN7 proteins. Note that the expression level of the CLDN1 and 7 proteins in the presence of the TACSTD2 gene is significantly higher than those in the absence of the TACSTD2 gene. Original magnification was ×40. D: Effect of the MG-132 treatment on the protein expression of the CLDN1, 4, and 7 proteins. HeLa cells transfected with the CLDN1, 4, or 7 genes were treated with proteasome inhibitor MG-132 for 24 hours. Compared with the negative control, the MG-132 treatment significantly increased the expression level of the CLDN1 and 7 proteins while that of the CLDN4 protein was virtually unchanged by the MG-132 treatment. Original magnification was ×40. E: Effect of the MG-132 treatment on the expression of the CLDN1, 4, and 7 proteins forcedly expressed in the HeLa cells. The MG-132 treatment significantly increased the expression level of the CLDN1 and 7 proteins in a dose-dependent manner but had virtually no effect on the expression level of the CLDN4 protein.

Figure 6

Expression of TJ-related proteins in the GDLD cornea. A: Clinical appearance of the cornea of a GDLD patient bearing a p.118Q>X nonsense mutation; grayish, protruding amyloid depositions are observed. B: Schematic representation of the structure of the TACSTD2 protein with mutations so far reported for the TACSTD2 gene indicated with amino acid numbers and types of mutation. The arrow indicates the p.118Q>X nonsense mutation. “SS” means signal sequence, “TY” means thyroglobulin repeat, “TM” means transmembrane domain, and “NH2” and “COOH” mean amino and carboxy termini, respectively. C: Expression of the TACSTD2, CLDN1, 4, and 7 proteins in normal and GDLD corneas. Original magnification was ×40. D: Results of Western blot analysis on the expression of CLDN1, 4, and 7 proteins in normal (NC) and GDLD corneas (GC). E: Results of qPCR analysis against the CLDN1, 4, and 7 and TACSTD2 mRNA in normal and GDLD corneas. Note that the expression level is almost identical between normal and GDLD corneas in the CLDN1, 4, and 7 mRNA.

Figure 7

Sequence comparison between the TACSTD2 and the EpCAM proteins and among the CLDN1, 4, and 7 proteins in their transmembrane domains. A: Both the TACSTD2 and the EpCAM proteins have an AxxxG motif (enclosed by a square) in their transmembrane (TM) domain. B: CLDN1 and 7 proteins have 2 AxxxG motifs in their third and forth transmembrane domains, while the CLDN4 protein does not have an AxxxG motif in any of its 4 transmembrane domains. C: Schematic representation of a molecular phylogenetic tree among 22 members of the CLDN gene family depicted in an unrooted N-J tree image. Note that the CLDN1 and 7 genes are quite similar over their entire coding regions while the CLDN4 gene is quite different from these two genes.