Corneas obtained from the Wisconsin Lions Eye Bank were from donors aged 23, 5(5, 58, 60, (51, 63, 64, 67,. 75, 77, 79, 81, and 85 years (average age 65 Â± 16.
Corneal Synthesis of a 1-Proteinase Inhibitor (al-Antitrypsin) Sally S. Twining,* Takeo Fukuchi,\ Beatrice Y.J.T. Patricia M. Wilson,* and Goran Boskovic*
Purpose. To determine if the cornea synthesizes cvl-proteinase inhibitor (a 1-ant.il rypsin). Methods. Human coineas were placed in organ culture lor 24 hours in the presence of :':>Smeihionine 10 radiolabel corneal proteins. Monoclonal antibodies were used to precipitate labeled cvl-proteinase inhibitor. The imnninologically isolated inhibitor was elect rophoresed on polyacrylamide gels and visualized by autoradiography or by staining for protein. Human corneas were also fixed wiih formalin and imbedded in para (Tin. Sections were probed with •''H-labeled complementary DNA probes lo the coding region of «l-proteinasc inhibitor. Results. Metabolically labeled cvl-proteinase inhibitor was recovered from organ-cultured corneas and the cornea-conditioned medium. Specific messenger RNA was observed in the cornea by in situ hybridization most prominently in corneal epithelial cells. Conclusions. cvl-Protcinase inhibitor is synthesized and released by human corneal epithelial cells. These results indicate that the cornea has the ability to locally control degradation through synthesis of this inhibitor without total dependence on a supply of the inhibitor from the vascular system. Invest Ophthalmol Vis Sci. 1994;35:458-462.
JYlaintenance of corneal transparency is critical for vision. To achieve this goal, the matrix of the stroma, which is composed of collagen, proteoglycans and other extracellular proteins,1 must be protected from prot.eolyt.ic degradation. Proteinase inhibitors play a major role in preventing and controlling proteolysis. The human cornea contains cvl-proteinase inhibitor (also called cvl-ant itrypsin),2 which is the major inhibitor of polymorphonuclear leukocyte elastase.3 This inhibitor is present in all three layers of the cornea at average levels of 29.5 ng for the epithelium, 54.3 ^g for the stroma, and 3.5 /ig for the endothelium per cornea.2 cvl -Proteinase inhibitor is localized inside the cells, in the extracellular matrix of the stroma and in Descemet's membrane.
From the *Departments of Biochemistry and Ophthalmology, Medic.nl College of Wisconsin, Milwaukee, and the iDepartment of Ophthalmology and Visual Sciences, University of Illinois at Chicago College of Medicine, Chicago, Illinois. This study was supported by Grants from the National Institutes of Health, ROIF.Y06663 (SSI), R0I-EY03890 (HYJTY). R0I-HY05628 (HYJTY) and P30EYQ1931 (Core Grant for Vision Research). Submitted for publication: June I, 1993; revised August 9. 1993; accepted August 17, 1993. Proprietary interest category: N. Reprint requests: Sally S. Twining, Department of Biochemistry, Medical College of Wisconsin, 8701 Watertown Plank Road. Milwaukee, Wl 53226.
Proteinase-inhibitor complexes as well as degradation products are observed in corneal extracts. cvl-Proteinase inhibitor is part of the serpin family of inhibitors that inhibit serine proteinases.' Other homologous members of this family include cvl-antichymotrypsin, cvl-antiplasmin, antithrombin III, protease nexin, and plasminogen activator inhibitor. Serpins and serine proteinases react with a .1:1 stoichiometry. These proteinases interact with the inhibitor within an exposed reactive loop. The a 1-proteinase inhibitor reactive loop is composed of 1(5 residues. This loop is the most variable region of the serpins. The major physiologic target enzyme for cvJ-proteinase inhibitor is neuirophil elastase with a 1\ of 10"H M"1 and a ka of 6.5 X 107 M's"1.'5 cvl -Proteinase inhibitor inhibits many serine proteinases with most having K; values of 10"4 to 10"8 M"1. It is probable that the major role of a 1 -proteinase inhibitor in the cornea is to protect the cornea from degradation by neuirophil elastase during inflammation. In addition, this inhibitor may also function as a backup inhibitor for other serine proteinases in the cornea. cvl-Proteinase inhibitor is present in most body fluids including blood aqueous humor and tears.2 Additionally it has been localized in many tissues, yet. the
Investigative Ophthalmology & Visual Science. February 1!)'.)'!. Vol. 'lf>. No. 2 Copyright @ Association for Research in Vision and Ophthalmology
Corneal Synthesis of al-Proteinase Inhibitor
synthesis of this inhibitor has been reported in only a few tissues.4'5 Because of the high levels of synthesis of this inhibitor by hepatocytes where messenger RNA (mRNA) levels are approximately 200 X higher than in other cells,5 it was initially believed to be expressed specifically in this tissue and distributed to other tissue via the circulatory system. Recently, blood monocytes, alveolar macrophages, and enterocytes4'5 are also found to synthesize the inhibitor. In this article we report the synthesis of al-proteinase inhibitor by the human cornea. METHODS Organ Culture Corneas obtained from the Wisconsin Lions Eye Bank were from donors aged 23, 5(5, 58, 60, (51, 63, 64, 67, 75, 77, 79, 81, and 85 years (average age 65 ± 16 years). Each cornea was cultured in 1 ml leucine-, lysine-, and methionine-deficient Minimum Essential Media (Gibco-BRL, Grand Island, NY) with Earle's salts supplemented with 52 jiig L-leucine, 72 jug L-lysine, 100 nC\ 35S-methionine (Amersham, Arlington Heights, IL, cell labeling grade) and 2.2 mg/ml NaHCO3. The organ culture was performed in the presence of 95% air/5% CO2 at 37°C for various times up to 24 hours. In some experiments, corneas were cultured in the aforementioned culture media for 2.25 hours and then switched to the same media without radiolabel but with 10% fetal calf serum (Gibco-BRL). At. the time of harvest, the medium was saved and the corneas were either used whole or dissected into epithelial, stromal, and Descemet's membrane-endothelium. Each cornea or corneal layer was frozen in liquid nitrogen, freeze-fractured and extracted in 10 mM Tris buffer, pH 7.2, containing 0.14 M NaCl, 10 mM phenylmethylsulfonyl fluoride, 10 mM ethylenediaminetetraacetic acid, 5 /xM E-64, 1 /xM pepsiatin, 1% non idet P-40, 1% sodium deoxycholate, and 0.1% sodium dodecyl sulfate (0.4 ml for endothelium and epithelium and 1.0 ml stroma or whole cornea). The proteinase inhibitors, phenylmethylsulfonyl fluoride, ethylenediaminetetraacetic acid, E-64, and pepstatin (Sigma Chemical Co, St. Louis, MO), were used to prevent degradation of al -proteinase inhibitor during the extraction and assay procedures. Immunobinding Assay Because of the nonspecific binding to antibody-bound resins, a method was developed based on the assay of Chao et. al,6 which lakes advantage of the bivalent nature of antibodies. al-Proteinase inhibitor was bound to Durapore membrane 96-well plates (Millipore, Bedford, MA) by incubating 10 /ig of the inhibitor (Athens Research Technologies, Athens, GA) in 100 n\
of 50 mM Tris buffer, pH 7.5, for 45 minutes at room temperature. Nonbound sites were blocked by incubation for 1.5 hours at room temperature with 200 n\ of 50 mM Tris buffer, pH 7.5, containing 25 mg/ml bovine serum albumin (Sigma), 150 mM NaCl, and 0.02% sodium azide. The wells were then washed once with 200 /xl of 50 mM Tris buffer, pH 7.5, containing 150 mM NaCl and then three times with the same buffer containing 0.05% Triton X-l 00. After washing, the wells were incubated with 200 n\ of 50 mM Tris buffer, pH 7.5, containing 51 itg/nil of mouse monoclonal antibody to human a\-proteinase inhibitor (Medix Biotech, Inc., Foster City, CA) for 1 hour at room temperature. The wells were washed as before. The wells were then incubated with 200 ix' of corneal extract or corneal conditioned media for 2 hours at room temperature. The wells were washed as before and then dried. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer (20 /xl/ well) was added and incubated at 90°C for 5 minutes. These samples (10 /xl/sample) were loaded on 10% polyacrylainide gels. After electrophoresis, one set of gels was silver stained (BioRad, Richmond, CA) and the second set was impregnated with the scintillant, 2,5-diphenyloxazole (PPO, Amersham) and exposed to x-ray film (Konica, Tokyo, Japan). In Situ Hybridization of Complementary DNA to al-Proteinase Inhibitor mRNA In situ hybridization of cvl -proteinase inhibitor mRNA to 3H-labeled complementary DNA was carried out using a procedure developed by Zhu et al for keratin mRNA in the cornea.7 Briefly, corneas were obtained from the Illinois Eye Bank. The donors (14, 17, 19, 41, 46, and 68 years of age; 39 ± 19 years) had no known ocular diseases. Corneas were fixed immediately in 4% formaldehyde in 0.08 M phosphate buffer, pH 7.0, for 24 hours, then processed and embedded in paraffin. Corneal sections (5 /xm) were cut, deparafTinized, rehydrated, and digested with proteinase K (Boehringer Mannheim, Indianapolis, IN). After digestion, the sections were washed, postfixedin 4% paraformaldehyde, dehydrated, and dried. Escherichia coli HB101, containing a PBR 322 plasmid inserted with 1.4 kB complementary DNA to a\-proteinase inhibitor (clone phAT85), was obtained from American Type Culture Collection (Rockville, MD). The complementary DNA insert contains the entire coding region for the inhibitor, exons 2 to 5, plus 48 bases on the 5' end and 130 bases past the stop codon including part of the poly A tail.8 3H-labeled DNA probes were prepared using random primers (Amersham) and 3H-dATP (ICN, Irvine, CA) according to the method of Haseba et al.9 The mRNA in the corneal sections was hybridized to labeled probe (2 X 106 cpm/ml, 0.1 ixg probe/ml) at
Investigative Ophthalmology & Visual Science, February 1994, Vol. 35, No. 2
45°C overnight. To test for specificity of binding, additional cornea] sections were incubated using the same amount of labeled probe plus 10 fig unlabeled probe/ml and the same incubation conditions as experimental sections. The corneal sections were washed as described previously7 then treated with 960 U/ml SI nuclease (Sigma). After rinsing, the slides were dehydrated, air dried, coated with 50% Kodak NT-2 nuclear emulsion (IBI, New Haven, CT) and exposed for 6 to 8 weeks. After developing, the tissue was counterstained in Mayer's hematoxylin (Sigma) and mounted in Permount (Fisher Scientific, Itasca, IL). In the authors' opinion, methods for securing human tissue were humane, included proper consent and approval, and complied with the Declaration of Helsinki.
RESULTS Human corneas, cultured in the presence of 35S-methionine in Minimal Essential Medium or in the same medium plus fetal calf serum for periods up to 24 hours, synthesized a protein that bound to monoclonal antibodies made to al-proteinase inhibitor. This protein comigrated with isolated cd-proteinase inhibitor from human plasma (Fig. 1). This 58 kD band was found in corneal extracts from total corneas and cornea conditioned medium (Fig. 1). It was also found in the epithelial, stromal, and endothelial layers of corneas, which were incubated intact, separated into layers and then extracted (data not shown).
When corneas were preincubated for 2.25 hours in Minimal Essential Medium plus 35S-methionine and then transferred to medium without the radiolabel but containing 10% fetal calf serum, the initial form of the inhibitor detected in the cornea was a 60 kD form (Fig. 1). By 4 hours after the preincubation period, a band at 58 kD was also observed, which increased in intensity by 6 hours with concomitant disappearance of the 60 kD band. A 56 kD band appeared in the medium at 2 hours after the preincubation period and increased with time, but not as dramatically as in the corneal extracts. By 6 hours, this band ran at approximately 58 kD. In the cornea at 4 hours, a unique band at 49 kD was observed. A 96 kD band was observed in both the cornea and the medium at 4 and 6 hours. Low amounts of additional high-molecular-weight bands were also observed. Similar bands were noted when fetal calf serum was not present, however, the rate of synthesis was decreased. Two bands at 43 and 76 kD were partially nonspecifically absorbed to the albumin-blocked Durapore membrane with or without incubation with antibodies (Fig. 1, control cornea). These results using mouse monoclonal antibodies were similar to those using rabbit polyclonal antibodies to the inhibitor. Because the corneas were cultured intact, the site of synthesis of al-proteinase inhibitor was not determined. To address this, corneal sections were probed with a 3H-complementary DNA probe to al-proteinase inhibitor mRNA using an in situ hybridization procedure. The probe reacted strongly to the epithelial
Hrs after Preincubation
Medium 0 2 4 6
Total Cornea 0 2 4 6 "S Methlonlne
Control Cornea 6 6
Cornea Medium a1PI 6 6 Silver Stain
FIGURE l. Synthesis of metabolically labeled al-proteinase inhibitor by the cornea with time in organ culture. Corneas were incubated in Minimal Essential Medium containing 35S-methionine for 2.25 hours and then transferred to the same medium without radiolabel but with 10% fetal calf serum for increasing amounts of time. Proteins isolated by the immunobinding assay were separated by sodium dodecyl sulfate electrophoresis and visualized by fluorography. Companion gels, including isolated plasma al-proteinase inhibitor, were silver stained. The four corneas used for this study were from 67- and 79-year-old donors. Control cornea extract A: no al-proteinase inhibitor bound to the Durapore membrane but the membrane was blocked with bovine serum albumin and the antibody was added to the well. Control cornea extract B: No al-proteinase inhibitor bound to the Durapore membrane or antibody added to the well but the well was blocked with bovine serum albumin.
Corneal Synthesis of al-Proteinase Inhibitor
cells (Fig. 2, A and C). Although the labeling was not as prominent as in the epithelial cells, mRNA labeling was seen associated with the stromal keratocytes and some of the endothelial cells (Fig. 2, C and D). Dilution with nonlabeled probe (Fig. 2B) reduced the hybridization to the background level. Similar results were obtained using corneas from donors aged 14 to 68 years indicating that the ability of the cells to synthesize mRNA for al-proteinase inhibitor was not dependent on age. DISCUSSION We have shown the presence of al-proteinase inhibitor mRNA and metabolically labeled protein in human corneas. In addition, we have demonstrated the ability of the cornea to release newly synthesized inhibitor
2. al-Proteinase inhibitor mRNA localization by in situ hybridization of normal human corneal tissue. Corneal sections from a 17-year-old donor were hybridized with a 3 H-labeled complementary DNA probe to al-proteinase inhibitor in the absence (A), (C), and (D) or the presence of 100X unlabeled probe (B). For (A) and (B) the magnification is X 106 and for (C) and (D) X 422. (C) is from the anterior portion and (D) the posterior portion of the same section as (A). (B) is a serial section to (A). Photographs are given as the dark field negative. Arrows designate areas of concentration of labeled probe associated with stromal keratocytes. Arrowheads indicate the labeling of an endothelial cell. FIGURE
into the culture medium. The first form observed was 56 kD at 2 and 4 hours after preincubation. The major form observed at 6 hours and at 24 hours is 58 kD. The presence of al-proteinase inhibitor in the medium could be attributable to cell death and release of cytoplasmic contents into the medium. However, this inhibitor, in the cornea, is probably targeted for secretion through posttranslation glycosylation in a similar manner as the hepatocyte and monocyte forms of the inhibitor. The glycosylation pattern of corneal al-proteinase inhibitor is not yet known. The plasma form contains 3 asparagine linked high mannose oligiosaccharides.10 That the inhibitor has been localized in the extracellular matrix of the cornea2 also suggests that the inhibitor is secreted from the corneal cells. The high amount of al-proteinase inhibitor mRNA detected in the epithelial layer and that al-proteinase inhibitor is released into the cornea conditioned media suggests the epithelial layer may be the source of at least part of the al-proteinase inhibitor found in the tears. The major form of the newly synthesized enzyme observed under reducing conditions both in the cornea and in the medium was 58 kD, which comigrated with samples of human al-proteinase inhibitor isolated from plasma. The initial forms observed were approximately 60 kD in the cornea and 56 kD in the medium suggesting the presence of several processing forms. These are most likely due to processing of the complex carbohydrates on the inhibitor. The 52 kD band could possibly be the nonglycosylated precursor of the inhibitor, similar to that noted for liver and monocytes.5 A 96 kD band observed in the cornea and the medium probably represents a complex between a proteinase and the inhibitor. Additional high-molecular-weight bands may represent other complexes. The smaller band at 43 kD is probably a degradation product. The molecular weights of al-proteinase inhibitor complexes and degradation products precipitable with antibodies to this inhibitor depend on the type and source of the cells.5 In our hands, the observed molecular weight of al-proteinase inhibitor depends on the sodium dodecyl sulfate-polyacrylamide gel electrophoresis system used. In this study using reducing sodium dodecyl sulfate gels, plasma al-proteinase inhibitor and one form of the immunoprecipitated corneal inhibitor was 58 kD. Previously, we observed the plasma form and the corneal form of this inhibitor at 53 kD on nonreducing gels.2 If al-proteinase inhibitor synthesized by the cornea is exactly the same as that produced by the liver is not known. Other tissues studied, macrophages, monocytes, and enterocytes produce an inhibitor identical to that of the liver one but the regulation of synthesis is different as are the mRNA transcripts.45 There are
Investigative Ophthalmology & Visual Science, February 1994, Vol. 35, No. 2
specific promoter regions on the gene used for liver and macrophage synthesis of the inhibitor. Three « 1 proteinase mRNA transcripts are produced constitutively by macrophages by alternate splicing that are distinct from the normal hepatocyte mRNA. One contains three additional exons, a second contains two additional exons plus part of the third and the other contains two extra exons. 4 These three can be produced by hepatocytes on stimulation by the cytokine IL6. The al-proteinase inhibitor synthesized by the three transcripts is identical to that of the liver. In the longer one, there are two additional short open reading frames. Regulation of the synthesis of this inhibitor in the liver and monocytes occurs both at the level of translation and transcription. 4 Experiments are in progress in our laboratories exploring the regulation of synthesis of this gene in the cornea. Synthesis of this inhibitor by the cornea, however, does not rule out the liver as an additional source of inhibitor in the cornea through diffusion from the limbal vessels or the fluids that bathe the cornea. Local corneal synthesis of al-proteinase inhibitor may be critical the cornea's protection during local inflammation. Key Words al-proteinase inhibitor, a 1-antitrypsin, cornea, protein synthesis, imrnunoassay References I. Walmian SR, Hart Jr WM. The cornea. In: Adler FH, ed. Adler's Physiology of the Eye. Si. Louis: Mosby;
2. Twining SS, Everse SJ, Wilson PM, Yue BYJT, Chan S-K. Localization and quantification of a-1-proteinase inhibitor in the human cornea. Curr Eye Res. 1989;8:389-393. 3. Heidtmann H, Travis J. Human a,-protcinase inhibitor. In: Barrett AJ, Salvcnson GS, ed. Proteinase Inhibitors. London: Elsvier; J986;441-456. 4. Hafeez W, Ciliberto G, Perlmulier DH. Constitutive and modulated expression of the human , antitrypsin gene. Different transcriptional initiation sites used in three different cell types./ Clin Invest. 1 992;89:12141222. 5. Perlmutter DH, Cole FS, Kilbridge P, Rossing TH, Colten HR. Expression of the a,-proteinase inhibitor gene in human monocytes and macrophages. ProcNatl Acad Sci U S A. 1985;82:795-799. 6. Chao S, Chao L, Chao J. Enhanced specificity in iminunoscreening of expression cDNA clones using radiolabeled antigen overlay. Biotechniques. 1989;7:6872. 7. Zhu G, Ishizaki M, Haseba T, Wu RL, Sun T-T, Kao VVVV-Y. Expression of K12 keratin in alkali-burned rabbit corneas. Curr Eye Res. 1992;] 1:875-887. 8. Long GL, Chandra T, Woo SIX, Davie EW, Kurachi K. Complete sequence of the cDNA for human alpha1-antitrypsin and the gene for the S variant. Biochemistry. 1984;23:4828-4837. 9. Haseba T, Nakazawa M, Kao CW-C, Murthy R, Kao VVW-Y. Isolation of wound-specific cDNA colones from a cDNA library prepared with mRNAs of alkaliburned rabbit corneas. Cornea. 1991; 10:322-329. 10. Mega T, Lujan E, Yoshida A. Studies on the oligosaccharide chains of human a^protease inhibitor. / Biol Chem. 1980:255:4057-4061.