Accumulation of HDL Apolipoproteins Accompanies Abnormal Cholesterol Accumulation in Schnyder's Corneal Dystrophy

Collection of Corneal Tissue and Aqueous Fluid

Corneas from four patients with Schnyder's dystrophy (Table 1⇓) were obtained at the time of penetrating keratoplasty. A 7- to 7.5-mm central region of cornea was removed during the keratoplasty procedure. This region included the central corneal opacity but did not include the peripheral arcus. A similar portion of normal corneal tissue was harvested from donors between 3 and 8 hours after death. Portions of the corneas were removed and fixed in 10% phosphate-buffered formalin (pH 7.0) for electron microscopy. The remaining corneal tissues were frozen on dry ice and stored at −70°C until analysis. Aqueous fluid obtained at the time of keratoplasty was also kept frozen. Some plasma lipid and lipoprotein data were reported previously for patients 1 to 3 listed in Table 1⇓.¹⁷ These and additional lipoprotein data are included here for reference (Table 2⇓).

Lipid Analysis of Corneas

Corneal tissues were thawed, blotted dry, and weighed before they were finely minced into pieces no larger than 1 mm³. The tissue pieces were collected into 1 mL of distilled H₂O and extracted with 21 mL of a mixture of 2:1 (vol/vol) chloroform-methanol according to Folch et al.²⁴ Aliquots of extracts were assayed in duplicate for their content of unesterified and esterified cholesterol according to the enzymatic-fluorometric method of Gamble et al.²⁵ Phospholipid content was determined (in duplicate) by the colorimetric method described by Bartlett.²⁶ Duplicate determinations did not vary more than 10%.

Immunostaining of Apolipoproteins

Frozen sections of unfixed corneal tissue were prepared from three normal individuals and the 47-year-old woman with Schnyder's corneal dystrophy (the second cornea removed from patient 2). Sections were adhered to gelatin-coated microscope slides and stored at −70°C until immunostaining was performed. For staining, tissue sections were brought to room temperature and fixed in 10% phosphate-buffered formalin or absolute acetone for 10 minutes. Treatment of tissue sections with 3% H₂O₂ for 5 minutes followed by two successive rinses in Dulbecco's phosphate-buffered saline (DPBS) quenched any endogenous peroxidase activity. Excess buffer was removed from around the sections that were then incubated with a blocking solution (1% bovine serum albumin and 0.1% Tween-20 in DPBS) for 15 minutes. After the blocking step, sections were incubated with the indicated concentration or dilution (in DPBS) of mouse monoclonal primary antibody. Mouse monoclonal antibodies were either whole ascites fluid or IgG fractions purified from ascites and included the following: purified antibodies with specificities against apoA-I (clone 3F10), apoA-II (clone 4A2), and apoB (clone 2B1), all from PerImmune, Inc, and whole ascites fluid (diluted 1:100) with specificity against apoE (MAB 1048) from Chemicon International, Inc. Additionally, a negative staining reaction was assessed by substituting the purified primary mouse monoclonal antibody with the same concentration of purified mouse IgG (catalog No. 55939, Cappel) or by substituting anti-apoE-containing ascites fluid with a similar dilution of control ascites fluid containing an anti-theophylline antibody (catalog No. MAB550-215/12, Chemicon).

After exposure to primary antibody for 30 minutes, sections were rinsed twice in DPBS. Primary antibody was detected with a kit (LSAB kit 683, Dako Corp) that used biotin-streptavidin methodology. Sections were incubated for 30 minutes with biotinylated goat anti-mouse immunoglobulin diluted in DPBS containing 1% bovine serum albumin and 0.1% Tween-20. After this incubation, sections were rinsed twice with DPBS, incubated 30 minutes with peroxidase-conjugated streptavidin in DPBS, and rinsed twice again with DPBS. Sections were then incubated 10 minutes with substrate solution consisting of 0.07% 3-amino-9-ethylcarbazole and 0.007% H₂O₂ in 0.1 mol/L acetate buffer (pH 4.8). After a final two rinses with DPBS, sections were mounted in glycerol gelatin.

Immunoblot Analysis of Corneal Apolipoproteins

A sample of normal cornea (from a 49-year-old man) and a sample of affected cornea (from a 49-year-old man, patient 4) were thawed and placed in buffer A (25 mmol/L HEPES [pH 7.4] containing 1 mmol/L Na₂-EDTA, 0.5 mg/mL 4-(2-aminoethyl)benzenesulfonyl fluoride HCl (Pefabloc SC), 1 μg/mL aprotinin, 1 μg/mL pepstatin A, and 1 μg/mL leupeptin). The samples were rinsed with two changes of buffer A. Then a quadrant of the central cornea with a radius of 2.5 mm (from the center of the cornea) was cut out from each sample. Excess buffer A was drained on a gauze pad, and wet weights of the cornea samples were determined. Each sample was transferred to a glass vial containing 0.5 mL of solubilization buffer (62.5 mmol/L Tris-HCl [pH 6.8], 10% glycerol, 5% sodium dodecyl sulfate [wt/vol], and 2.5% dithiothreitol [wt/vol]) and homogenized with a rotating blade (Tissue Tearor, Biospec Products). After protein was extracted from corneal tissue for 2 hours on ice, samples were transferred to 1.5-mL Eppendorf tubes and centrifuged at 10 000g to obtain a supernatant with the extracted protein.

Next, samples were delipidated and concentrated by precipitation with chloroform-methanol according to the procedure of Wessel and Flügge.²⁷ Precipitates were redissolved in solubilization buffer containing 0.01% bromphenol blue and electrophoresed according to Laemmli.²⁸ Electrophoresis was performed with 7.5×10-cm 10% acrylamide gels and a Mini Protean II cell (Bio-Rad). A Mini Trans-Blot electrophoretic transfer cell (Bio-Rad) was used to blot electrophoresed samples onto nylon membranes. Blotting was performed in 10 mmol/L 3-(cyclohexylamino)-propanesulfonic acid buffer (ie, CAPS) (pH 9.5) containing 20% methanol. Blots were probed overnight at 4°C with affinity-purified mouse IgG monoclonal antibodies (0.2 μg/mL) with specificities against apoE (clone 6C5, University of Ottawa Heart Institute) and apoA-I (clone 3F10, PerImmune). Purified apoA-I and apoE standards prepared from plasma were also obtained from PerImmune. Blotting and detection were performed with a chemiluminescence immunoblotting kit (catalog No. 77248, Schleicher and Schuell) with the use of affinity-purified, alkaline phosphatase-conjugated, goat anti-mouse IgG and Lumigen PPD (4-methoxy-4-[3-phosphatephenyl]-spiro[1,2-dioxetane-3,2′-adamantane]) as substrate (dilutions and conditions as specified by Schleicher and Schuell).

Quantification of Apolipoproteins in Aqueous Fluid

ApoA-I and apoE were quantified with the use of a sandwich-type enzyme-linked immunoabsorbent assay. Wells of microtiter plates (96-well Immulon 1, Dynatech Laboratories, Inc) were coated overnight at 4°C with a monoclonal capture antibody (5 μg/mL for apoA-I and 1 μg/mL for apoE) in 100 μL of 0.2 mol/L carbonate-bicarbonate coating buffer, pH 9.4 (catalog No. 28382, Pierce). The monoclonal capture antibodies for apoA-I and apoE were the same as those used for immunoblot analyses. After removal of the coating buffer, wells were incubated for 2 hours at room temperature with blocking buffer (PBS containing 3% bovine serum albumin, pH 7.4) and then rinsed four times with wash buffer (PBS containing 0.05% Tween-20). Next, 100 μL of samples or 100 μL of apolipoprotein standards (diluted in blocking buffer) was added to wells and incubated for 2 hours at 37°C. Then wells were rinsed four times with wash buffer and incubated at 37°C for 1 hour with 100 μL of polyclonal goat anti-human apoA-I (2 μg/mL) or apoE (3 μg/mL) antibodies (IgG fraction) (catalog Nos. 06500154 and 06501904, respectively, Biogenesis Inc) diluted in blocking buffer. Next, wells were rinsed four times with wash buffer and incubated at 37°C for 1 hour with 100 μL of affinity-purified polyclonal rabbit anti-goat IgG-peroxidase conjugate (catalog No. A4174, 30 U/mL, Sigma Chemical Co) diluted 1:10 000 with blocking buffer. Wells were rinsed six times with wash buffer and then incubated for 15 minutes at room temperature with 100 μL of 0.4 g/L TMB peroxidase substrate system (catalog No. 507600, Kirkegaard and Perry Laboratories, Inc). Lastly, 100 μL of 2 mol/L sulfonic acid was added to each well, and the absorbance at 450 nm was determined.

Ultrastructural Analysis of Cornea

Pieces of central cornea (1 mm³) initially fixed at the time of surgery in 10% phosphate-buffered formalin were later refixed in 4% glutaraldehyde in 0.15 mol/L phosphate buffer (pH 7.2) at room temperature for 1 hour. Tissues were rinsed with the phosphate buffer, postfixed for 2 hours in 1% osmium tetroxide in phosphate buffer, dehydrated in ascending concentrations of ethanol, and embedded in epoxy resin. Thin sections were prepared and stained with uranyl acetate and lead citrate.

Clinical Appearance of Corneas

All patients displayed a central disclike opacification, diffuse stromal haze, and peripheral arcus lipoides (Fig 1⇓). Three patients (patients 2, 3, and 4) (Table 1⇑) had central subepithelial crystal deposition. Relative lucency in the middle of the central corneal disc opacity or “bull's eye configuration” was noted in two patients (patients 1 and 4). The disciform central opacity affected the full thickness of stroma in all but one case. In this patient (patient 2), the disciform central opacity seemed to be affecting the anterior stroma. All patients had a marked decrease in visual acuity and corneal sensation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Central and peripheral opacities in a cornea from a patient with Schnyder's dystrophy. The eye of patient 2 (47-year-old woman) shows peripheral arcus (→ ←), central disclike opacification (crossed ↦), and central subepithelial crystal deposition (↦).

Ultrastructural Appearance of Corneas

Electron microscopic analysis revealed that vesicles and solid-appearing granules were dispersed within the extracellular matrix of the corneal stroma (Fig 2c⇓) and Bowman's layer (the acellular anterior layer of the cornea) of affected corneas (Fig 2b⇓). The vesicles and granules were spherical, elongated, or irregular in shape and ranged between 100 and 300 nm along their longest diameter. The vesicles were usually limited by two to five tightly apposed membranes and appeared similar to oligolamellar liposomes (Fig 2c⇓). Extracellular empty crystalline clefts were also observed in some corneal regions (data not shown). No vesicles, granules, or crystals were present in the normal central cornea (Fig 2a⇓). There was no evidence of any inflammatory component such as the presence of granulocytic or mononuclear cells accompanying the extracellular accumulation of lipid particles, although some keratocytes were necrotic.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Ultrastructure of abnormal lipid particles that accumulate in corneas of patients with Schnyder's dystrophy. Bowman's layer of normal (a) and affected (patient 1) (b) central cornea. Many multilamellar vesicles (arrows) and electron-dense granules (arrowhead) are present in the extracellular matrix of the stroma of the affected cornea (c). Bar=200 nm for c; bar=2000 nm for a and b.

Accumulation of Cholesterol and Phospholipid in Cornea

Chemical analysis showed a substantial accumulation of cholesterol (both unesterified and esterified) and phospholipid in affected corneas (Table 1⇑). There was a greater than 10-fold increase in the cholesterol content and a 5-fold increase in the phospholipid content of affected compared with normal corneas. The percentage of cholesterol that was unesterified in normal corneas averaged 50% and was slightly but significantly increased to 63% in affected corneas. The molar ratio of unesterified cholesterol to phospholipid in affected corneas (1.5:1) was increased 3-fold compared with the ratio (0.5:1) in normal corneas.

Crystals, presumably cholesterol, accumulate in corneas of patients with Schnyder's dystrophy.¹ However, it is not known whether these crystals are composed of the esterified or unesterified form of cholesterol. Therefore, we examined the lipid content of a region of an affected cornea that was shown by slit lamp examination to contain crystals. The cholesterol content of this region of cornea was 15 μmol/g tissue (wet wt), of which 84% was unesterified. In addition, the molar ratio of unesterified cholesterol to phospholipid was high, at 1.9:1. Both findings are consistent with the crystals being composed of predominantly unesterified cholesterol.

Apolipoproteins in Cornea and Aqueous Fluid

Deposition of HDL-associated apolipoproteins in the cornea accompanied accumulation of extracellular lipids. ApoA-I immunostaining was substantially increased in the center of the affected cornea compared with staining in the center of the normal corneas (Fig 3⇓). ApoA-I immunostaining was associated with both keratocytes (thin elongated cells) and the extracellular matrix of the affected cornea (Fig 4a⇓). This pattern of staining was similar to the distribution of apoA-I in the normal cornea.³⁰ However, apoA-I immunostaining associated with keratocytes and the extracellular matrix of the affected cornea was increased compared with the normal cornea (Fig 4a and 4b⇓⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

ApoA-I immunostaining in a cornea affected with Schnyder's dystrophy compared with a normal cornea. The affected cornea (a) shows intense apoA-I immunostaining in the central region, whereas the central region of the normal cornea (b) shows much less immunostaining. Only the periphery of the normal cornea shows a degree of apoA-I immunostaining similar to that seen in the central region of the affected cornea (the periphery of the affected cornea was not removed during keratoplasty). Bar=1 mm.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Immunostaining of apolipoproteins in normal and affected corneas. Frozen sections of the midportion of the central cornea (not including the peripheral arcus) from a patient with Schnyder's dystrophy (a second cornea removed from patient 2 at age 51 years) and an individual of similar age without Schnyder's dystrophy were fixed for 10 minutes with 10% phosphate-buffered formalin. Then sections were incubated for 30 minutes with 1 μg/mL of the purified mouse IgG monoclonal antibodies with specificity against apoA-I, apoA-II, and apoB and whole ascites fluid (diluted 1:100) containing monoclonal antibody against apoE. Control sections were incubated with similar concentrations of purified mouse IgG or whole ascites fluid containing monoclonal antibody against theophylline. Antibodies were detected with a biotin-streptavidin-peroxidase system as described in “Methods.” Incubations with purified mouse IgG or ascites fluid containing anti-theophylline control antibody showed no staining, similar to g and h. Arrows point to immunostained keratocytes, and rectangles enclose regions of immunostained extracellular matrix. In h, bar=50 μm.

Similar to apoA-I immunostaining, apoE and apoA-II immunostaining were much more intense in the affected compared with the normal corneas (Fig 4c through 4f⇑). Most of the increased apoA-II and apoE immunostaining occurred in the extracellular matrix. Slight apoE but almost no apoA-II immunostaining occurred in the extracellular matrix of normal corneas. In contrast to apoA-I, there was no obvious apoE or apoA-II immunostaining associated with keratocytes in normal corneas. However, some apoA-II immunostaining was associated with keratocytes in the affected cornea.

The central corneal endothelium that lines the posterior corneal surface showed immunostaining for all three apolipoproteins in affected corneas but showed only slight immunostaining for apoA-I in normal corneas (not shown). In the affected cornea, all three apolipoproteins showed focal immunostaining of central Bowman's layer. In the normal cornea this region did not show apolipoprotein staining (not shown). ApoB immunostaining was not present in the corneal tissue examined (Fig 4g and 4h⇑⇑). However, the corneal tissue did not include the peripheral cornea, where apoB staining was previously shown in normal corneas.^{30 31}

Fixation of corneas with 10% phosphate-buffered formalin for 10 minutes resulted in more intense immunostaining of apoA-I, apoA-II, and apoE compared with fixation of corneas with absolute acetone. Replacement of the specific mouse monoclonal antibodies with the same concentration of purified mouse IgG or control (ie, anti-theophylline) ascites fluid did not produce staining of normal or affected cornea. This was similar to the lack of staining observed with the monoclonal antibody directed against apoB (Fig 4g and 4h⇑⇑).

Protein was extracted from equivalent portions of normal and affected corneas. Equal aliquots of the extracted proteins were electrophoresed through 10% acrylamide gels under reducing conditions. The separated corneal apolipoproteins (apoA-I and apoE) were detected by immunoblot analysis and are shown in Fig 5⇓. Immunoblot analysis showed greater levels of apoA-I and apoE in the affected cornea compared with the normal cornea. In addition, the immunostained apolipoproteins exhibited the expected M_r of 28 000 for apoA-I and 36 000 for apoE, similar to the plasma standards for these apolipoproteins. IgG in the corneal extracts was detected through the cross-reactivity of the goat anti-mouse IgG with human IgG. Equal amounts of IgG (M_r=50 000) were present in the affected and normal corneal extracts.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Immunoblot analysis of corneal apolipoproteins. Protein was extracted from equivalent central regions (not including the peripheral arcus) of corneas from a patient with Schnyder's dystrophy (patient 4) and a man of similar age without Schnyder's dystrophy. Extraction was performed by homogenizing corneal tissue in a solubilization buffer containing 62.5 mmol/L Tris-HCl (pH 6.8), 10% glycerol, 5% sodium dodecyl sulfate (wt/vol), and 2.5% dithiothreitol (wt/vol). Samples were delipidated and concentrated by precipitation with chloroform-methanol before being electrophoresed in 10% acrylamide gels and electroblotted onto nylon membranes. Blots were incubated overnight at 4°C with affinity-purified mouse IgG monoclonal antibodies that detected apoA-I or apoE as indicated. The monoclonal antibodies were detected with chemiluminescence with alkaline phosphatase-conjugated, goat anti-mouse IgG and Lumigen PPD as substrate. Because the goat anti-mouse IgG cross-reacted with human IgG, it was possible to detect IgG in the corneal extracts. This provided a means to assess the amount of this protein in the cornea relative to the amount of probed apolipoprotein in the cornea. Shown are standards prepared from plasma (a), extract from an affected cornea (b), and extract from a normal cornea (c). Ten nanograms of apoA-I and 20 ng of apoE standards were electrophoresed. Ten microliters of corneal extract was electrophoresed for detection of apoE, and 20 μL of corneal extract was electrophoresed for detection of apoA-I. The IgG and apoA-I were detected in the same blot. A longer exposure of the blot to the chemiluminescent substrate was required to detect apoE compared with apoA-I.

Because immunostain and immunoblot analyses showed an increase in the amount of apoA-I and apoE in the affected cornea, we examined whether there was an increase in apoA-I and apoE in the aqueous fluid of one patient (patient 2). The concentrations of apoA-I and apoE were 2.2 μg/mL and 0.5 μg/mL, respectively. These levels were not significantly different from concentrations of apoA-I (2.7±0.4 [SE] μg/mL) and apoE (0.7±0.3 [SE] μg/mL) determined in the aqueous fluid of four patients who had corneal transplants for reasons other than Schnyder's corneal dystrophy.