Autoimmunity against carbonic anhydrase II affects retinal cell functions in autoimmune retinopathy

Autoimmunity has been suggested as a possible etiologic factor in retinal degeneration. Autoantibodies against various retinal proteins, including recoverin, enolase, and carbonic anhydrase II (CAII) have been found in patients with cancer-associated retinopathy (CAR) and in autoimmune retinopathy (AR) patients without diagnosed cancer (AR) at the time of initial examination { Adamus, 2004 #1111}. The presence of anti-CAII autoantibodies have also been reported in patients with retinitis pigmentosa (RP) with cystoid macular edema but rarely in the unaffected population (6% prevalence), suggesting their role in pathogenicity of disease { Heckenlively, 1999 #592}. The mechanism of CAII antibody generation in autoimmune diseases remains to be elucidated. We propose that autoantibodies against CAII that are detected in patients with retinopathy may be a consequence of the immune response against CAII found in other tissues. Serum autoantibodies to CAII could be a result of the cross-reactivity against other antigens that mimic CAII or were elucidated in response to cancer antigens { Chegwidden, 2000 #1722}. For autoantibodies to be effective and induce retinal disease they have to persist at high titers, reach the target autoantigen in the organ, and induce pathogenic effects on the tissue. A body of evidence from previous in vitro and in vivo studies on the role of autoantibodies in retinal degeneration indicates that anti-recoverin autoantibodies found in patients with retinopathy are cytotoxic to retinal cells and induce apoptotic death of retinal photoreceptor cells, which leads to the degeneration of the photoreceptor cell layer in our animal model of CAR { Adamus, 1998 #366; Cao, 2000 #486; Shiraga, 2002 #962; Ren, 2004 #1055; Ohguro, 1999 #467}. Previously, we showed that anti-recoverin antibody causes a significant increase in intracellular calcium level by blocking calcium binding capacity of recoverin, a calcium binding protein, leading to retinal cell mitochondrial-mediated apoptosis { Adamus, 2006 #1175; Shiraga, 2002 #962}. In addition, anti-enolase antibody led to apoptosis by three different pathways: by blocking the catalytic activity of enolase, which results in depletion of glycolytic ATP and disturbance of retinal cell glycolysis; by causing an increase in the intracellular calcium levels, which activates mitochondrial-mediated apoptosis; and by causing the translocation of Bax protein into the mitochondria, which will result in the release of cytochrome c into the cytoplasm, initiating apoptosis { Magrys, 2007 #1343}. CAII has been accepted as a one candidate target autoantigen recognized during the autoimmune response in retinopathy. CAII belongs to a family of physiologically important enzymes - carbonic anhydrases (CAs) that catalyze a reversible conversion of carbon dioxide to bicarbonate, participate in ion transport and pH control. CA isoenzymes differ in their tissue distribution, subcellular localization and their susceptibility to CA inhibitors. The sequence alignment suggested that the active site structure is conserved in two human cytoplasmic enzymes CAI and CAII. However, the amino acid sequence identity of human CA XIV relative to the other membrane-associated isozymes (CAIV, CAIX, and CAXII) is only 34–46% { Whittingtons, 2004 #1352}. Multiple CA isozymes are expressed in the retina, including membrane-associated CAIV and CAXIV, which have been implicated in facilitating the removal of CO 2 from the neuroretina and in photoreceptor function { Yang, 2005 #1210; Nagelhus, 2005 #1363}. Although systemic CA inhibition using acetazolamide increases retinal and cerebral blood flow, oxygen concentration in the optic nerve and blood CO 2 levels but these inhibitors can also impair photoreceptor function { Ilies, 2003 #1738}. The goal of the studies was to examine whether anti-CAII autoantibodies isolated from patients with AR induce pathogenic effects on the catalytic function of CAII, which may have metabolic consequences on cell survival. The in vitro studies were done using E1A-NR3 retinal cell line that represents immortalized undifferentiated retinal cells, expressing abundant retinal and neuronal markers of photoreceptor, Müller, and ganglion cell phenotypes { Seigel, 1999 #1198}. The cells possess the functional capacity to respond to specific neurotransmitors and apoptotic insults { Seigel, 2004 #1197;Adamus, 1997#168}. Our studies demonstrated that autoantibodies against CAII showed their pathogenic potential through the inhibition of the enzymatic activity, which in effect led to decrease the intracellular pH and increase the intracellular Ca 2+ in retinal cells. These cellular alternations had detrimental effects on retinal cells viability leading to cell death.

Sera were obtained from patients with progressive visual loss identified through the Ocular Immunology Laboratory - Oregon Health & Science University (OHSU). The study has been approved by the OHSU Institutional Review Board. Patients’ sera were affinity-purified on a Protein G column (Pierce) and then dialyzed, concentrated, and checked for purity by SDS-gel electrophoresis, as described previously { Adamus, 1998 #359}. Human CAII was purchased from Chemicon International and then purified on Ultralink Protein-G (Pierce) before use to remove impurities. Anti-human CAII II (Erythrocytes) polyclonal antibodies were also purchased from Chemicon International and used as control. E1A-NR3 retinal cell line was maintained in Dulbecco’s modified Eagle medium (DMEM), supplemented with 10% fetal bovine serum (FBS), 1% MEM nonessential amino acids, 1% MEM vitamins, and 50 mg/mL gentamycin { Adamus, 1993 #31}. Initial screening of patients’ sera was performed using human retinal proteins as described previously { Adamus, 2004 #1111} to select the sera that reacted with 30-kDa on the blot. Then, the reactivity of sera with 30-kDa retinal protein was confirmed on blots containing 1 μg pure CAII. Human retina sections from a donor eye were fixed in fresh 4% paraformaldehyde in PBS, pH 7.4, for 10 minutes at room temperature followed by washing and incubation with 0.3% H 2 O 2 in PBS for 10 min. Sections were then blocked in 10% normal goat serum, 1% BSA, 0.2% Tween 20 in PBS for 1 hour. Then, patients’ purified IgG at 50 μg/ml in 1% BSA, 0.2% Tween 20 in PBS were added for overnight incubation at 4°C; negative controls contain no antibodies. Sections were then washed as before, incubated in with biotinylated anti-human secondary antibody (Invitrogen) for 1 hour at room temperature, and washed again. Immunoreactivity was revealed by incubating the tissue sections in streptaviden-HRP conjugate (Invitrogen) for 30 minutes, washed, and incubated for 5 minutes with peroxidase DAB substrate (Pierce). Cells were allowed to attach to 96-well, flat-bottomed microtiter plates for four hours at a density of 2×10 4 cells/well in DMEM medium at a final volume of 100 μl per well. Doses of 75 μg, 150 μg, and 300 μg purified IgG were added to the culture for 72 hours. Three hours before the end of the antibody incubation, 25 μL of a 0.5% MTT (Thialzolyl blue, Sigma) solution was added to each well and then 100 μL of cell lysing buffer (10% SDS in 50% DMF) following by overnight incubation at 37°C. The color reaction was read in a BioRad Microplate Reader at 570 nm. The data are presented as a percentage of cell survival, calculated as follows: % cell survival={1−((A total −A sample )/A total )} × 100, where A total = OD of untreated cells and A sample = OD of the cells treated with antibodies. Microtiter plates were coated with 0.5 μg/well of CAII in 0.1 M Tris-HCl buffer, pH 9.0, overnight at room temperature. After blocking with 1% BSA in PBS for 1 hour and washing, 0.1 mg/ml of double serially diluted autoantibodies were added to each well and allowed to incubate for 1 hour, washed, and followed by 1 h incubation with anti-human IgG conjugated to peroxidase (Invitrogen.) Color reaction was developed for 30 minutes by incubation with peroxidase substrate (2–2′-azino-bis-(3-ethylbenz-thiazoline-6-sulfonic acid) in 0.1 M citrate-phosphate buffer, pH 4.5, containing 3% H 2 O 2 and measured at 405 nm using a Bio-Rad Microplate Reader. Microtiter plates coated with 0.5 μg/well of CAII in 0.1 M Tris-HCL buffer, pH 9.0, were blocked with 1% BSA, 5% normal goat serum (NGS) in PBS for 1 hour at room temperature. Patient serum (1:1000) was added and incubated for 1 hour followed by incubation with biotinylated Ig representing IgG1, IgG2, IgG3, IgG4, IgM, IgA, IgG (H+L) (Invitrogen) at 1:2000. Then, streptavidin conjugated to peroxidase (1:5000) was added for 30 min. Color reaction was developed for 30 minutes by incubation with peroxidase substrate (2–2′-azino-bis-(3-ethylbenz-thiazoline-6-sulfonic acid) in 0.1 M citrate-phosphate buffer, pH 4.5, containing 3% H 2 O 2 and measured at 405 nm using a Bio-Rad Microplate Reader. Changes in concentrations of free intracellular Ca 2+ were measured using the Ca 2+ -sensitive fluorescent dye Fluo-4 NW (Molecular Probes). Cells were allowed to attach to 96-well, flat-bottomed, half area, black walled microtiter plates overnight at a density of 5×10 3 cells/well in DMEM medium at a final volume of 100 μl per well. The medium was then removed and the cells were washed once with PBS containing 2.5 mM probenecid (Molecular Probes). Fluo-4 NW dye in 2.5 mM of probenecid acid was then added and the plate was incubated in the dark at 37°C for 30 minutes. The dye was then removed and the cells were washed once in Hank’s Balanced Salt Solution (HBSS) containing 2.5 mM of probenecid acid followed by 30 minutes of incubation at 37°C in the dark in the same solution. Then, a 5 minute baseline measurement of each well was read using an FLx800 Microplate Fluorescence Reader (Bio-Tek Instruments) with excitation 485/20, emission 528/20, and sensitivity 80 settings. Following baseline measurements purified patients’ purified autoantibodies and positive controls were added at a final concentration of 1 mg/ml and immediately read using the same settings as the baseline at 15 minute intervals. The results are presented as a fold of increase in intracellular [Ca 2+ ] from baseline. Changes in intracellular pH ([pH] i ) were measured using the pH-sensitive fluorescent dye BCECE-AM (Molecular Probes.) Cells were allowed to attach to 96-well, flat-bottomed, black walled microtiter plates overnight at a density of 2×10 4 cells/well in DMEM medium at a final volume of 100 μl per well. Purified IgG were added at a final concentration of 1 mg/ml for 8 hr incubation. After the incubation, the medium was removed and the cells were washed with HBSS and 100 μL of HBSS containing 2.5μL/well of 200 μM BCECE-AM was added to all wells. The plate was incubated for 20 min at 37°C protected from light in a tissue culture incubator and shaken periodically. Then, the dye was then removed from the wells and 100 μL of HBSS for 30 min incubation at 37°C and then read using an FLx800 Microplate Fluorescence Reader (Bio-Tek Instruments) with excitation 485/20, emission 530/20, and sensitivity 80 settings. Standard curve solutions were made from mixtures of 130 mM KH 2 PO 4 or 110 mM K 2 HPO 4 containing 20 mM NaCl in ddH 2 O. In wells designated for pH standards, 100 μL of standard pH solutions were added. The pH values were then calculated using the regression equation from the pH standard curve. Carbonic anhydrase activity was assayed by the colorimetric method using bromothymol blue { Rickli, 1964 #1387}. The reaction mixture contained 100 μL of 25 mM Tris-HCl, pH 8.2, with 1 mg/100 ml of bromothymol blue (Sigma) and was kept on ice. CA II was added to the reaction mixture to a final concentration of 100 ng. Various molar ratios of CAII II to anti-CA II antibodies were pre-incubated together for 1 hr at 4°C with rotational mixing. After incubation, 100 μL of 25 mM Tris-HCl, pH 8.2, bubbled 30 minutes with CO 2 on ice was added to the reaction mixture and the colorimetric change was timed from the moment of injection until the endpoint and measured at 430 nm using a spectrophotometer (Beckman). Activity units (U) are calculated according to the formula: U= 10*((T b /T c ) −1/mg protein) where T b is the time of uncatalyzed reaction, T c is the time of enzyme-catalyzed reaction endpoint, and mg protein is the amount of carbonic anhydrase II present. Significant differences between groups were assessed using analysis of variance (ANOVA), the nonparametric Mann-Whitney U test and Chi-square test. A p value <0.05 was considered significant.

Based on sample availability in quantities needed to perform experiments, we selected six retinopathy sera designated P1–P6 that reacted with a protein of molecular weight 30-kDa on the blot containing retinal proteins. As shown in Figure 1A , these sera reacted with purified human CAII on the blot. Thus, the antigen was confirmed to be carbonic anhydrase II. Immunostaining of human retina with patients’ purified autoantibodies revealed the cytoplasmic labeling of the cells in the ganglion cell layer, possible feet of the Muller cells, some cell bodies in the inner nuclear layer and the cytoplasm of inner and outer segments of photoreceptor cells ( Fig. 1B–E ). Fig. 2A shows titration curves for patients’ affinity purified IgG against pure human CAII determined by ELISA. These autoantibodies show high activity against CAII at titers up to 1:10,000. We also determined subclass distribution for immunoglobulins specific for CAII. As is shown in Fig. 2B , IgG subclass was IgG4 and IgG1 dominant. CAII is present in the cytoplasm of the cells and is responsible for controlling intracellular pH. To examine whether autoantibodies affect the catalytic function of CAII enzyme the purified anti-CAII autoantibodies P2 were incubated with CAII at increasing ratios of IgG and then the enzymatic activity was measured. Fig. 3A shows that P2 inhibited the catalytic activity in a dose dependent manner. The activity dropped from 90% with 10-fold excess of P2 IgG to 48% with 40-fold excess of IgG of the initial CAII activity. Autoantibodies without specificity to CAII (normal subject) had no effect on the enzymatic activity but other anti- CAII autoantibodies (P1, P4, P5) inhibited the CAII activity at the level similar to that of the control - specific goat anti-human CAII antibodies ( Fig. 3B ). These results demonstrated that binding anti-CA autoantibodies to CAII was associated with decreased enzymatic activity, suggesting that the site of the immunologic reaction at least partially involved the active catalytic site of the enzyme. Because anti-CAII autoantibodies impact the CAII catalytic function thus we hypothesized that inhibition of the CAII catalytic function will affect ([pH] i ) pH changes in the cell. To test this hypothesis we determined [pH] i of retinal cells grown in the presence of purified patients’ autoantibodies for 8 hr. After incubation with autoantibody the cells were loaded with the pH-sensitive fluorescent dye BCECE-AM followed by the [pH] i measurement. The [pH] i of untreated retinal cell was estimated to be 7.6. Within 1 hr after the addition of IgG the pH in antibody-treated cells stared to decrease (not shown). Fig. 4A shows that P1 – P6 autoantibodies caused a drop in [pH] i from 0.5 to 0.8 pH units after 8 hr-treatment, depending on anti-CAII antibody titers ( Fig. 2 ). The lowering of [pH] i to 6.8 corresponds to the threshold for activation of acid endonuclease and caspase 3 and correlates with the induction of apoptosis{ Gottlieb, 1996 #1195; Reber, 2002 #1733}. Normal IgG did not alter the [pH] i in retinal cells with the same amount of IgG added and time of incubation with cells. These findings suggest that the drop in [pH] i for a longer period of time will lead to intracellular cellular acidification, which may affect intracellular calcium levels{ Barnes, 1991 #1737; DeSantiago, 2007 #1751; Masuda, 2007 #1752}. To examine whether anti-CAII autoantibodies have effect on free intracellular calcium [Ca 2+ ] i , we investigated changes in [Ca2+] i in retinal cells grown in the presence of purified autoantibodies for a period of 120 min and measured at 15 min intervals. Figure 4A shows a time response curve for P3 IgG. The maximum Ca 2+ release was observed at 15 min and was significantly higher from normal control antibodies (p<0.05). Figure 4B shows that, at 15 min, other anti-CAII autoantibodies triggered a small but significant increase in [Ca 2+ ] i from about 3 fold to 8 fold compared to untreated cells. An ANOVA test revealed statistically significant differences between groups for different autoantibodies (all p =0.02). In summary, these results imply that changes in intracellular calcium may be linked to the inhibition of catalytic activity and acidification, which possibly causes a rise in [Ca 2+ ] i . To determine whether anti-CAII autoantibodies affect on retinal cell viability, cells were grown at different concentrations of purified autoantibodies (75 μg, 150 μg and 300 μg/ml) and viability was examined by a MTT assay. Fig. 5 shows decreased cell survival after 72-hr antibody treatment in a dose dependent manner. At the highest antibody dose, the cell survival dropped and ranged from 30% (P1) to 70% (P5) compared to untreated control. In conclusion, anti-CAII autoantibodies, depending on specific their titers, were cytotoxic to retinal cells and had a devastating effect on retinal cell survival.

Our studies were designed to understand the pathogenic effects of anti-CAII autoantibodies on retinal cells to determine their role in retinal degeneration. The major conclusions from our studies revealed that anti-CAII autoantibodies have the capacity to induce cellular damage by impairing its cellular function through (1) inhibiting the catalytic activity of CAII, (2) decreasing intracellular pH, (3) increasing intracellular calcium, (4) which in effect influence cell survival. The destabilized catalytic function of CAII and alterations in cytosolic pH were found very early after autoantibodies were added to the living cells and before apoptosis occurs; usually after 24 hr { Shiraga, 2002 #962}, further suggesting that autoantibodies are inducers of apoptosis. The decrease of [pH] i related to the function of the enzyme in the cell could cause acidification by changes in CO 2 /H 2 CO 3 metabolism and in proton pump activity, the Na+/H+ exchanger { Kniep, 2006 #1362; Reber, 2002 #1733}. Some anti-CAII autoantibodies caused a decrease of pH below pH 6.8, which could lead to caspase activation in those retinal cells. The efficiency of activation of caspase-3 by cytochrome c has been found to be pH sensitive, with an optimum pH of 6.6 to 6.8 { Gottlieb, 1996 #1195; Reber, 2002 #1733}. The earlier studies showed that acidification in Jurkat cells caused by anti-Fas antibody also preceded apoptosis, suggesting acidification is an early event in programmed cell death and may be essential for DNA destruction { Gottlieb, 1996 #1195}. Hence, one can envision that antibody entry into the cell and binding to CAII will disrupt metabolic functions, which, in effect, increase the propensity toward apoptosis ( Fig. 6 ). We believe that autoantibodies not only block the CAII catalytic function but also can block the CAII binding to the Na + /H + exchanger { Li, 2002 #1732}. Blocking of both CAII functions may lead to acidification, which are consistent with increases in intracellular calcium { Nicotera, 1998 #1063}. Our previous studies showed that the increase in [Ca 2+ ] induced by autoantibodies specific to recoverin or enolase led to apoptosis and cell death { Magrys, 2007 #1343; Adamus, 2006 #1175}. Even though these autoantibodies found in patients with retinopathy have different antigenic specificities, which affect different metabolic pathways, including phototransduction for recoverin, glycolysis for enolase, and pH control for CAII, they all induce an increase in the [Ca 2+ ] i . Thus, we propose that the marked elevations of [Ca 2+ ] i in retinal cells triggered by autoantibodies ultimately result in cell death. Such calcium-induced cytotoxicity may play a major role in autoimmune retinopathy and several other diseases { Mattson, 2007 #1703}. Anti-CAII autoantibodies are also found in autoimmune diseases such as systemic lupus erythematosus, Sjögren’s syndrome, autoimmune cholangitis, chronic pancreatitis, primary biliary cirrhosis, endometriosis, systemic sclerosis, and type I diabetes { Taniguchi, 2003 #1371; Bataller, 2004 #1348; Frulloni, 2000 #1359; Itoh, 1992 #1360; Alessandri, 2003 #1372; D’Cruz, 1996 #1374; Invernizzi, 1998 #380}. In all of these conditions, antibody levels of anti-CA II were higher in patients with disease than in the control group as reported from 3.8 – 10% in normal subjects. Such a broad occurrence of anti-CAII led to questions concerning the role of CA II in the pathogenesis of autoimmune retinopathy. Autoimmune response against CAII in CAR was present months before diagnosis of cancer in some retinopathy patients {Adamus #1871}. The dominant IgG subclass against CAII is IgG4. Interestingly, IgG4 is predominantly expressed under conditions of chronic antigen exposure { Aalberse, 2002 #1736}, which may be a case during cancer growth and turnover. It is possible that serum autoantibodies to CAII detected in patients with CAR/AR are a consequence of the cross-reactivity of autoantibody against another antigen that mimics CAII or were elucidated in response to cancer antigens { Chegwidden, 2000 #1722}. In addition, during oxidative stress, including the continuous destruction of oxidized erythrocytes appears to induce the formation of autoantibodies against certain erythrocyte components together with CAII { Iuchi, 2007 #1369; Kurien, 2006 #1749}. An insult by environmental factors, such as viruses, affects organs and triggers autoimmunity against CAII. However, it is unlikely that these autoantibodies are simply the result of organ degeneration because autoantibodies were not detected in patients with alcoholic or gall stone-related pancreatitis, during which degeneration of exocrine pancreatic tissues occurs { Nishimori, 2004 #1361}. There is evidence that CAII is a potent autoantigen. Cellular immune responses to CA II measured by PBMC proliferation assay were detected in idiopathic chronic pancreatitis and in Sjögren’s syndrome { Nishimori, 2004 #1361}. More interestingly, active immunization of susceptible mice with purified CA II resulted in inflammation of submandibular glands { Nishimori, 1995 #1356} and bile ducts and cholangiocytes in autoimmune cholangitis mouse model { Ueno, 1998 #1358}, suggesting a possible pathogenic role of anti-CA II in immune-mediated diseases. The target organs were inflamed in those animals but the eyes were not evaluated. The authors speculated that the disease may start with anti-CA II antibody formation in predispose mice (or humans), which bind to and impair the function of CA II, resulting in secondary effects that lead to inflammation at organ site where CAII function is critical. Some anti-CA II antibodies are able to clear the CAII antigen without mounting an additional immune response. Both effects may be mediated by antibody internalized but not degrade by the affected cells. In summary, our study showed that anti-CAII autoantibodies induce pathogenic effects on retinal cells by decreasing cell survival through distressing CAII cellular function. It is therefore possible that once such autoantibodies reach the retina they will provoke pathological changes in retinal physiology and cell death leading to retinal degenerating ( Figure 6 ). The current experiments may facilitate better understanding the role of the immune system in retinal degeneration and help to develop better strategies for therapy of autoimmune retinopathy.