Serum Autoantibody against Glutathione S-Transferase in Patients with Glaucoma

Abstract

purpose. To identify retinal proteins that are the targets of serum autoantibodies in patients with glaucoma.

methods. To identify retinal antigens that are recognized by the sera of patients with glaucoma, immunoreactive bands were separated, by using two-dimensional gel electrophoresis of the bovine retinal soluble fraction. A 29-kDa band was then selected for further analysis. Tryptic peptides of the 29-kDa band were analyzed using electrospray mass spectrometry to identify the protein. After protein identification, immunoreactivity against this newly identified protein was studied by Western blot analysis using sera from 65 patients with glaucoma (25 with primary open-angle glaucoma [POAG]; 40 with normal-pressure glaucoma [NPG]) and 25 age-matched healthy subjects. In addition, serum antibody titers were compared in these groups, by using a specific enzyme-linked immunosorbent assay (ELISA).

results. The 29-kDa band was identified as glutathione S-transferase (GST). Western blot analysis revealed that serum antibodies against GST antigen were recognized in 34 (52%) of 65 patients with glaucoma (22 of NPG and 12 of POAG) and 5 (20%) of 25 age-matched control subjects (χ² test, P < 0.05). By ELISA, it was also found that patients with glaucoma had higher titers of anti-GST antibody, compared with the control group (Mann–Whitney test; NPG versus control, P = 0.013; POAG versus control, P = 0.0006).

conclusions. These findings indicate that GST is one of the retinal antigens targeted by the serum antibodies detected in some patients with glaucoma.

Several observations have provided compelling reasons to propose that the immune system may have a role in the development and/or progression of glaucomatous optic neuropathy in some patients. Nagasubramanian et al. ¹ found elevated levels of serum IgG, IgM, and antinuclear antibodies in patients with glaucoma. Cartwright et al. ² demonstrated an epidemiologic association of immune-related disease in a cohort of patients with normal-pressure glaucoma (NPG). In addition, our observations revealed that in some patients with glaucoma, there is an increased prevalence of monoclonal gammopathy, ³ retinal immunoglobulin deposition, ⁴ and elevated serum antibody titers to retinal antigens. ^{5 6 7 8} We also demonstrated that heat shock proteins, including bacterial and human hsp60, ⁶ hsp27, andα B-crystallin, ⁷ as well as rhodopsin, ⁵ are among the retinal antigens targeted by serum antibodies detected in patients with glaucoma. In addition, our recent observations have shown that the mechanism of hsp27 antibody-mediated cytotoxicity to retinal ganglion cells is associated with the disruption of the integrity of cytoskeleton. ⁹  

Because our previous findings indicate that there may be additional target antigens in the retina of patients with glaucoma, ^{5 6 7 8} we sought to identify previously uncharacterized specific autoantibodies against retinal proteins in patients with glaucoma. We report here that glutathione S-transferase (GST) is one of the retinal antigens targeted by serum antibodies detected in some patients with glaucoma. 

Materials and Methods

Patient Selection

Blood samples were collected after detailed consents were obtained according to the tenets of the Declaration of Helsinki. Forty age-matched patients with NPG (69.5 ± 9 years), 25 patients with POAG (66.7 ± 11 years), and a control group of 25 healthy subjects (68.5 ± 9 years) were included. The inclusion and exclusion criteria for these groups have been described previously. ¹⁰ Briefly, the diagnostic criteria for NPG consisted of the presence of open iridocorneal angles, no evidence of intraocular pressure higher than 23 mm Hg, glaucomatous changes in visual fields and optic nerve cupping, and the absence of alternative causes of optic neuropathy. The diagnostic criteria for the POAG were similar to those for NPG, except that untreated intraocular pressure levels were higher than 23 mm Hg. Visual field loss of patients was evaluated with the Humphrey Field Analyzer, 30-2 program (Humphrey Instruments, San Leandro, CA). Our criteria for visual field abnormalities included a corrected pattern SD with P < 0.05 or a glaucoma hemifield test outside normal limits obtained with at least two reliable and reproducible visual field examinations. The subjects in the control group had no evidence of an immune-related systemic disease or an ocular disease based on their general health status, systemic disease history, current medications, and routine ophthalmologic examination, including visual acuity test and funduscopy. 

Western Blot Analysis

Bovine retinas were dissected from eyes obtained at a local abattoir within 3 hours of death and homogenized in ice-cold lysis buffer containing 2 mM HEPES, 2 mM EDTA (pH 7.4), and protease inhibitors (50 μM phenylmethylsulfonyl fluoride and 1 μg/ml each of aprotinin, antipain, bacitracin, bestanin, chymostatin, leupeptin, and pepstatin A) for 5 minutes at 4°C. After centrifugation at 1,000g for 10 minutes, the pellet (consisting of nuclei and unbroken cells) was discarded, and the membrane fraction was pelleted by centrifugation at 35,000g for 20 minutes. The supernatant or soluble fraction was saved and the membrane pellet washed twice with a solution containing 50 mM Tris, 154 mM NaCl (pH 7.4), and homogenized in lysis buffer. Fractions were stored at −80°C until use. The protein concentrations in the membrane and soluble fractions were determined using the bicinchroninic acid (BCA) method (Sigma, St. Louis, MO). 

Samples, including bovine retinal proteins and purified GST (Sigma) were separated by electrophoresis in 12% sodium dodecyl sulfate (SDS)-polyacrylamide gels at 160 V for 1 hour and electrophoretically transferred to polyvinylidene fluoride membranes (Millipore, Marlboro, MA) using a semidry transfer system (Bio-Rad, Hercules, CA). After transfer, membranes were blocked in a buffer (50 mM Tris-HCl, 154 mM NaCl, and 0.1% Tween-20 [pH 7.5]) containing 5% nonfat dry milk for 1 hour and then overnight in the same buffer containing a dilution of primary antibody and sodium azide. Primary antibodies were either patient sera (dilution, 1:1000) or antibody against GST (dilution, 1:2000; Sigma). After several washes and a second blocking for 20 minutes, the membranes were incubated with secondary antibodies (IgG or IgM) conjugated with horseradish peroxidase (1:2000; Fisher Scientific, Pittsburgh, PA) for 1 hour. Immunoreactive bands were visualized by enhanced chemiluminescence using commercially available reagents (Amersham Life Science, Arlington Heights, IL). 

Two-Dimensional Gel Electrophoresis

Two-dimensional gel electrophoresis of bovine retinal proteins was performed according to the manufacturer’s instructions (Bio-Rad). During the electrophoresis, the capillary tubes (Bio-Rad) were used in first dimension for isoelectric focusing. The tubes were filled with gel monomer solution, containing 9.2 M urea, 4.5% acrylamide/bis stock, 1% Bio-Lyte pH 5–7 ampholyte, 4% Bio-Lyte pH 3–10 ampholyte, 5% detergent solution (0.5 M 3-([3-cholamidopropyl] dimethylammonio)-2-hydroxy-1-propanesulfonate [CHAPS] and 10% Nonidet P-40), 0.1% N,N,N,′N′,-tetramethyl-ethylenediamine (TEMED), and 0.2% of 10% ammonium persulfate (APS). The samples were treated with a solution containing 0.15 mM dithiothreitol (DTT) and 0.35 mM SDS. Upper running buffer was 20 mM NaOH, and lower running buffer was 10 mM H₃PO₄. Electrophoresis was performed at 200 V for 2 hours, followed by 500 V for 2 hours, and 800 V overnight. The proteins were then further separated by electrophoresis in 12% SDS-polyacrylamide gels, transferred to membranes, and subjected to Western blot analysis as described. 

Electrospray Mass Spectrometry

Protein identification by electrospray mass spectrometry was performed at the Protein and Nucleic Acid Chemistry Laboratories of Washington University. Protein containing acrylamide gel pieces was first digested with sequencing-grade trypsin (Promega, Madison, WI). Tryptic digests were then separated on a high-pressure liquid chromatograph (model UMA; Michrom BioResources, Auburn, CA) using a reversed-phase trifluoroacetic acid (TFA)-lite solvent system on a C-18 column. The flow stream from the high-pressure liquid chromatograph, immediately after UV detection, was introduced into an ion-trap electrospray mass spectrometer (ThermoQuest model LCQ Classic; Finnigan, San Jose, CA) for mass spectrometry of the tryptic peptides. The mass spectrometry data for each peptide was then sequentially compared with the database of tryptic digests of all known proteins using commercial software (Sequest; Sequest Technologies, Lisle, IL) that matches the mass spectra against the electronically produced mass spectra of all possible tryptic peptides proteins in the database. Standard reagents and conditions were used. 

Enzyme-Linked Immunosorbent Assay

For the enzyme-linked immunosorbent assay (ELISA) 96-well microtiter plates (Packard Instruments, Meriden, CT) coated with GST (50 ng/ml in sodium carbonate buffer [pH 8.8]) were incubated overnight at 4°C. After the plates were washed, any remaining binding sites were blocked using 1% normal goat serum and 0.1% sodium azide at room temperature for 2 hours. Sera from patients or control subjects, diluted 1:500 in phosphate-buffered saline (PBS) containing Tween-20, plus 1% normal goat serum and sodium azide, were added to duplicate wells of antigen-coated plates and incubated overnight at 4°C. The serum was removed by washing with PBS, and secondary antibody (goat anti-human IgM conjugated with horseradish peroxidase; 1:2000; Fisher) was added. After 2 hours’ incubation at room temperature, the secondary antibody was washed with PBS, and color was developed using a kit (Biotrin, Dublin, Ireland). The plates were read at 410 nM with a plate reader (Packard Instruments). Negative control wells prepared without antigen or primary antibody and positive control wells in which increased concentrations of specific antibody to GST were used as primary antibody were simultaneously processed. Optical density values were recorded at three independent measurements, and the average value was calculated for each patient. Optical densities are presented as the mean ± SD calculated within each study group. The sensitivity ranges of this method were 0 to 2000 pg/ml, with intra-assay precision of less than 10%. 

Results

When we used sera from patients with glaucoma to probe Western blot analysis of retinal proteins, various immunoreactive bands were observed, several of which are present in many of the patients. ^{3 4 5 6 7 8} We sought to identify the 29-kDa band that was prominent in Western blot analysis using sera from several patients. We first performed two-dimensional acrylamide gel electrophoresis to separate proteins with identical molecular weight within the 29-kDa immunoreactive band. Figure 1 shows a representative Western blot using two-dimensional gel electrophoresis. Identical gels were stained with Coomassie blue, and the 29-kDa protein band at 5.6 isoelectric point (Fig. 1 , arrow) was cut and used for the identification of protein by electrospray mass spectrometry. 

Figure 2 shows the separation of tryptic peptides of the 29-kDa spot by high-pressure liquid chromatography. After searching the mass spectra data of individual peptides against the database of electronically produced mass spectra of all possible peptides that was obtained by the tryptic digests of the known proteins as described, the 29-kDa spot was identified as GST (class μ). We then performed Western blot analysis using patient sera against purified GST. Western blot analysis revealed that sera from 34 (52%) of 65 patients with glaucoma recognized 29-kDa purified GST protein. Figure 3 shows Western blot analysis of representative sera from patients with glaucoma and control subjects to probe either supernatant fraction of bovine retinal proteins or purified GST antigen. Sera from 22 (55%) of 40 patients with NPG and 12 (48%) of 25 patients with POAG exhibited prominent immunoreactivity against GST. However, among sera from 25 age-matched healthy control subjects, only 5 (20%) showed a fair immunoreactivity against GST (χ² test; NPG versus control, P = 0.005; POAG versus control, P = 0.04). No statistical difference was detected between POAG and NPG (P = 0.58). 

Last, we compared serum antibody titers against GST in glaucoma and control groups by using a specific ELISA. Serum ELISA titers (optical densities) of anti-GST antibody were 0.056 ± 0.01 and 0.051 ± 0.01 in patients with POAG and NPG, respectively, whereas the titer was 0.039 ± 0.01 in the control group (Mann–Whitney test, NPG versus control, P = 0.013; POAG versus control, P = 0.006). No difference was detected between NPG and POAG (P = 0.25). 

Discussion

When the sera from patients with glaucoma are used to probe Western blot analysis of retinal proteins, various immunoreactive bands are observed, several of which seem to be present in many of the patients. ^{3 4 5 6 7 8} Our previous studies have shown that retinal heat shock proteins and rhodopsin are among the most common retinal autoantigens targeted by serum antibodies in many patients with glaucoma. ^{5 6 7} In this study, we examined an immunoreactive band at 29 kDa on Western blot analysis of the supernatant fraction from bovine retina that was probed by sera from patients with glaucoma. The related retinal antigen was identified as GST class μ. Thus, GST is another retinal autoantigen recognized by serum antibodies detected in many patients with glaucoma. 

The reason for the selection of an immunoreactive band at 29 kDa for further characterization was based on the strength of the signal detected on Western blot analysis. In addition, the serum autoantibody recognizing the 29-kDa protein in retinal extracts was IgM. IgM has been identified to play a pathogenic role in producing autoimmune polyneuropathy, and the antigenic targets of IgM are more readily identified than the targets of IgG or IgA. ¹¹  

The GST supergene family, which encodes detoxification enzymes, is widely expressed in mammalian tissue cytosols and membranes. GSTs catalyze the conjugation of reduced glutathione with a wide variety of electrophiles that include known carcinogens and various compounds that are products of oxidative stress, including oxidized DNA and lipid. Indeed, several lines of evidence suggest that the level of expression of GST is a crucial factor in determining the sensitivity of cells to a broad spectrum of toxic chemicals and oxidative stress. ^{12 13}  

The cytosolic enzymes are encoded by at least five distantly related gene families designated GST classes α, μ, π, ς, andθ . ^{14 15 16} All the cytosolic subunits have a molecular weight of 45,000 and are dissociable into subunits of approximately 26,000 Da. The molecular weight of the immunoreactive band that we selected appeared higher than 26 kDa, because the prestained marker that we used migrated more slowly than nonstained markers. This was further confirmed by using 26-kDa purified GST, which also migrated as a 29-kDa band under these conditions. 

GST is present in glial and neuronal cells of the central nervous system and in the retina, as well as all throughout the human body, ^{17 18 19 20 21 22} and its subclasses are similar in various mammalian tissues. ²³ In addition, regional and cellular distribution of immunoreactivity to different subunits of GST was studied in rat and bovine retinas, and class μ immunoreactivity was detected throughout the retina, including the retinal ganglion cell layer. ^{21 22} High levels of GST in retinal glial cells suggest one native neuroprotective mechanism provided by the glial cells in which they conjugate toxicants with glutathione through the action of GST. ²⁴ This may be an important component of adaptive retinal responses in the glaucomatous retina, because oxidative stress and the generation of reactive oxygen species appear to be involved in neuronal cell death during glaucomatous neurodegeneration. ^{25 26} Increased titers of autoantibodies to GST in some patients with glaucoma may therefore represent a generalized response to tissue stress and/or damage as a consequence of the glaucomatous neurodegeneration process and thereby secondary production of serum antibodies to GST in the glaucomatous retina. Similarly, the members of the GST gene family have been identified as antigens in autoimmune hepatitis, ²⁷ in which GST has been identified to be an important component of hepatic detoxification mechanisms. 

Whether the circulating antibodies against GST are a consequence of tissue stress and damage in the retina or whether they have a pathogenic significance should be further evaluated. The data presented herein support previous observations ^{3 4 5 6 7 8} that activated immunity may be involved in the etiology and/or sustaining of glaucomatous optic neuropathy in some patients. Further studies will facilitate better understanding of the role of activated immunity in glaucomatous neurodegeneration. 

 Supported in part by Grant EY12314 from the National Eye Institute; the Glaucoma Research Foundation; The Glaucoma Foundation; the American Health Assistance Foundation, Washington, DC; and an unrestricted grant to Washington University School of Medicine, Department of Ophthalmology and Visual Sciences from Research to Prevent Blindness.

 Submitted for publication August 31, 2000; revised November 13, 2000; accepted January 8, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Identification of 29-kDa antigen by two-dimensional gel electrophoresis. Supernatant fraction of bovine retina (30 μg) was separated with two-dimensional gel electrophoresis and probed using serum from a patient with glaucoma (dilution, 1:1000). Arrow: immunoreactive spot subjected to sequence analysis.

Figure 2.

 

Identification of 29-kDa bovine retinal antigen. Electrospray mass spectrometry of the tryptic peptides revealed this protein to be GST class μ.

Figure 3.

 

Western blot analysis of serum from age-matched control subjects and patients with glaucoma with either POAG or NPG. Sera from some but not all patients with glaucoma (dilution, 1:1000) recognized purified GST (3 μg/lane), as well as 29-kDa protein in bovine retina supernatant fraction (BRS; 30 μg).