Molecular Vision 2010; 16:2055-2061 <http://www.molvis.org/molvis/v16/a221>
Received 24 February 2010 | Accepted 30 September 2010 | Published 13 October 2010

Proteomic analysis of the tear film in patients with keratoconus

Isabel Lema,1 David Brea,2 Raquel Rodríguez-González,2 Elío Díez-Feijoo,1 Tomás Sobrino2

The first two authors contributed equally to the work

1Instituto Galego de Oftalmoloxía, Servizo Galego de Saúde, University of Santiago de Compostela, Santiago de Compostela, Spain; 2Clinical Neuroscience Research Laboratory, Hospital Clínico Universitario, IDIS, University of Santiago de Compostela, Santiago de Compostela, Spain

Correspondence to: Isabel Lema, M.D., Ph.D., Instituto Galego de Oftalmoloxía, Hospital de Conxo, 15706 - Santiago de Compostela, A Coruña, Spain; Phone: +34 981531627; FAX: +34 981531627; email: mariaisabel.lema@usc.es

Abstract

Purpose: To identify proteins differentially expressed between the tear film of keratoconus (KC) patients and control subjects using two dimensional electrophoresis (2-DE) and mass spectrometry-based techniques.

Methods: Twenty two patients (44 eyes) diagnosed with bilateral KC and 22 control subjects (44 eyes) were studied in a prospective case-control study. Keratoconus screening programs and Orbscan II topographies were performed on all participants. Tear samples were collected by the Schirmer I method using filter paper. Proteins were extracted from the Schirmer strips and separated by 2-DE. Comparison of protein patterns was performed using PDQuest Software and protein differences were identified by mass spectrometry. Finally, results were validated by western-blot.

Results: Four spots were identified to be differentially expressed between KC patients and control subjects. Three of them were more expressed in healthy subjects and they were identified as zinc-α2-glycoprotein (ZAG), lactoferrin, and IGKC (immunoglobulin kappa chain). The other spot was more expressed in KC patients and it was identified as ZAG. Differences in ZAG seem controversial in two different spots because different posttranslational modifications, however, analysis of both spots revealed that globally, ZAG is overexpressed in healthy subjects. Founded differences in ZAG, lactoferrin, and IGKC expression were subsequently validated by western blot.

Conclusions: IGKC protein, ZAG, and lactoferrin are under-expressed in the tears of patients diagnosed with bilateral KC compared with healthy subjects. These differences could contribute to the knowledge of the pathophysiology of this disease.

Introduction

Keratoconus (KC) is a primary corneal ectasia, generally bilateral and progressive, with a conical shape as a result of the thinning of the stroma. This thinning induces irregular astigmatism, myopia, and protusion, leading to mild to marked impairment in the quality of vision [1]. KC has its onset at puberty, and it progresses until the third or fourth decade of life. The calculated incidence of KC is between 1/500 and 1/2,000 in the general population [1,2]. KC affects young persons in a very productive period of their lives. Yet the cause of this disorder, considered to be the most common indication for penetrating keratoplasty in developed countries [3], remains unknown.

KC may be isolated or may be associated with various clinical conditions, or even form part of a syndrome. Its etiology remains poorly understood, but it has been reported to arise as a consequence of biochemical alterations to corneal collagenase, which breaks down collagen leading to stromal thinning [4]. Other authors have suggested that it may arise as a consequence of keratocyte loss [5], which is on average about 19% lower in contact lens users [6]. In some atopic diseases, KC may be due to apoptosis of corneal cells, possibly mediated by interleukin-1 [7]. Rahí et al. [8], found atopic diseases in 35% of patients with KC, in contrast to 12% of control subjects. KC is frequently associated with ocular itchiness and rubbing, provoking chronic epithelial lesions [7]. The cornea is part of an integrated system - the ocular surface - which contains specific and non-specific immune molecules. Tissue degradation in thinning disorders, such as KC, involves the expression of inflammatory mediators, such as proinflammatory cytokines, cell adhesion molecules, and matrix metalloproteinases [9-12]. Moreover, previous studies from our group have demonstrated that the levels of proinflammatory cytokines (interleukin-6 [IL-6] and the tumor necrosis factor alpha [TNF-α]) and matrix metalloproteinase-9 (MMP-9) are significantly increased in the tears of patients with KC [13]. Despite these advances, there are currently no diagnostic markers of KC and the pathophysiology of this disease is largely unknown.

Proteomics is a discipline that study complex mixtures of proteins. Two-dimensional electrophoresis (2-DE) and mass spectrometry are very useful techniques to find out protein differences between different types of samples. Tear proteomics has been used to find out potential cancer markers [14], to study responses in patients with fungal keratitis [15] and for identification of tear proteases [16].

In this study we used proteomic tools, such as two dimensional electrophoresis (2-DE) and mass spectrometry, to find out protein differences in tear samples that allow the differentiation between eyes of KC patients and normal controls.

Methods

Subjects and clinical examination

We have designed a prospective, case-controlled study in which 22 bilateral KC patients (44 eyes) and 22 normal control subjects (44 eyes) were recruited from the Contact lens Unit, Instituto Galego de Oftalmoloxía, Complexo Hospitalario Universitario de Santiago de Compostela, Spain. KC patients (50.0% male; mean age, 31.3±7.8 years) and control subjects (54.5% male; mean age, 30.8±7.6 years) had not worn contact lenses for a week prior the study. All patients were expressly cited for the purposes of this study, and all examinations were performed by the same researcher between September 2007 and April 2008. Data collected included gender, age, patient’s ocular history (eye rubbing), medical history (atopy), and family history of KC. This research was performed in accordance with the Declaration of Helsinki of the World Medical Association (2000). Protocol was approved by the Hospital's Research Ethics Committee, and informed consent was given by all patients and control subjects.

The biomicroscopy exam was performed to detect signs of KC and the corneal topography study was performed to quantify topographic parameters. The modified Rabinowitz-McDonnell test was used to confirm the diagnosis of KC, and the severity of KC was classified according to the steepest simulated keratometry reading on the keratometric map (K2 <45 D, mild; K2 between 45 and 52 D, moderate; K2 >52 D, severe KC) [17].

Inclusion criteria were; 1) Bilateral KC patients, severe in both eyes or severe and moderate, with biomicroscopic signs in both eyes. 2) Subjects in the control group with normal corneal topography maps.

Exclusion criteria included; 1) Previous surgical intervention in the anterior segment, or childhood corneal trauma or disease. 2) Existence of active or systemic inflammation, or ocular disease, or current treatment with systemic or local anti-inflammatory drugs.

Instrumental

Basic examination instruments were a Topcon biomicroscope and Topcon refraction column (Topcon, Essebaan, Netherlands). As specific examination instruments we used a Keratron Scout (Opticon 2000; SpA, Rome, Italy) and an Orbscan II corneal topographer (Orbtek, Salt Lake City, UT).

Two-dimensional electrophoresis was performed in IPGphor (Bio-Rad, Hercules, CA) for the first dimension and ETTAN-DALTSIX for second dimension (GE, Uppsala, Sweden). Western blot was performed using Miniprotean 3 system and Trans blot SD semi-dry transfer cell (Bio-Rad). Gels and blots were digitized in FX-Proplus (Bio-Rad). Finally, MALDI-TOF mass spectrometer (MS) analyses were performed using a 4700 Proteomics Analyzer MALDI TOF/TOF MS (Applied Biosystems, Carlsbad, CA).

Tear analysis

Tear samples were obtained by placing a Schirmer strip over the lower lip, approximately 3 mm from the lateral canthus, without previous instillation of drugs or vital dyes. The subject was instructed to close the eyes for the 5 min test duration, the wet length was not recorded but was observed to be within normal ranges in all cases (>10 mm). We collected four Schirmer strip for subject; samples were frozen and stored at −20 °C immediately. Proteins were extracted from the Schirmer strips by incubating with 200 μl of 100 mM ammonium bicarbonate at room temperature for 1 h. Then 600 μl cold acetone were added and incubated for 2 h at −20 °C. Precipitation of proteins was performed by centrifugation at 18,000× g. Precipitated proteins were resuspended in solubilization buffer (9 M urea, 2% CHAPS). Proteins were quantified by modified Bradford assay using BSA as protein standard.

Two-dimensional electrophoresis

Two replicate gels were performed for 12 patients (24 eyes) and 12 control subjects (24 eyes). Protein (20 μg) was applied on 17 cm 3–10NL IPG strips and first dimension was performed until 45 kV h was achieved. Second dimension of gel electrophoresis was performed on 12.5% polyacrilamide gels on ETTAN-DALTSIX (GE) at 5 W per gel for 15 min, and then at 17 W per gel until the dye front reached the bottom of the gel. Analytical gels were stained with Flamingo Fluorescent Stain and visualized on Molecular Imager FX Pro-plus (Bio-Rad). Preparative gels were stained in 0.1% Coomassie Blue Brilliant R-250.

Gels were digitalized using a Bio-Rad Molecular Imager FX Pro-plus. Protein patterns were analyzed using PDQuest 7.0 from Bio-Rad. Gel spots of interest were manually excised from preparative gels and digested according to Shevchenko et al. [18].

MALDI-TOF MS analyses were performed. The obtained peptide mass fingerprints were used for protein identification. The non-redundant NCBI or SwissProt database was searched using MASCOT 1.9 through the Global Protein Server v3.5 from Applied Biosystems. All determinations were performed without knowledge of the corresponding clinical data (blind test).

Western blot

Tears collected from 10 patients (20 eyes) with KC and 10 control subjects (20 eyes) were used to subsequent validation of expression differences found using proteomics. Ten μg of total protein extracted from the Schirmer strips were loaded and size-separated in a 10% sodium dodecyl sulfate-polyacrilamide gel electrophoresis (95 V). The proteins were blotted in semi-dry conditions using polyvinyl difluoride membranes (Millipore, Billerica, MA) and incubated with primary antibodies raised in mouse against lactoferrin (1:2,000; Abcam, Cambrigde, UK), immunoglobulin kappa constant (IGKC) (1:2,000; Abcam) and zinc-α2-glycoprotein (1:700; Santa Cruz Biotechnology, Santa Cruz, CA) overnight at 4 °C. For detection, a goat anti-mouse Cy3-labeled secondary antibody (1:3,000; GE) was used and images were acquired using Molecular Imager FX Pro-plus (BioRad). QuantityOne software (BioRad) was used to analyze the expression of the different bands of protein. To verify equal protein loading, membranes were stripped at 50 °C for 30 min in stripping buffer containing 62.5 mM Tris, 2% SDS and 100 mM β-ME, subsequently re-blocked and probed with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (1:3,000; Abcam). Experiments were performed in duplicate.

Statistical analysis

Descriptive statistical analyses were performed with percentage for categorical variables. Discontinuous variables were expressed as median [quartiles]. Statistical significance for intergroup differences was assessed by the χ2 test for categorical variables. One-way ANOVA and the Mann–Whitney test were performed for comparison between groups. A value of p<0.05 was considered to be statistically significant. On the other hand, protein patterns were analyzed using PDQuest 7.0 from Bio-Rad. The reproducibility of the gels was calculated by Pearson's coefficient. Differences were considered when spot intensity was significantly different between patients and controls (p<0.01) and at least two times greater in one group respect to the other. The statistical analysis was conducted using SPSS 16.0 (IBM, Chicago IL) for Windows XP (Microsoft Corporation, Redmond, WA).

Results

Clinical Features

Table 1 shows clinical characteristics of patients with keratoconus and control subjects. No age-related or sex-related statistical differences were detected between KC patients and control subjects. A family history of KC was reported for 3 patients (13.6%) and 1 control (4.5%). Ten KC patients (45.5%) and five controls (22.7%) were diagnosed as having atopic disease (most commonly to pollen or acaroids). Twelve KC patients (54.5%) admitted frequent and vigorous eye rubbing and 3 control eye rubbing (13.6%).

All of the KC eyes showed either moderate or severe stage of progression: 18 eyes (40.9%) presented moderate KC (K2 between 45 D and 52 D), 26 eyes (59.1%) had severe KC (K2 >52 D).

Mean K2 was 54.8±7.1 D in the KC eyes, versus 43.7±1.6 D in the control subjects (p<0.0001). Mean of the thinnest point pachymetry was 386.2±71.6 µm in the KC eyes, versus 542±23.9 µm in the control subjects eyes (p<0.0001).

Proteomic analysis

An average of 532 spots was observed in both groups (Figure 1). Gels were highly reproducible (r2=0.95) between groups. No qualitative differences have been found, however three spots were found to be expressed at least twofold in healthy subjects than in KC patients and one spot was found twofold in KC patients than in healthy subjects (Figure 2). The expression levels are shown in Table 2. Furthermore these differences were statistically significant (p<0.001). Mass spectrometry analysis revealed that quantitative differences more expressed in healthy subjects were; IGKC (Immunoglobulin Kappa Chain) protein, zinc-α2-glycoprotein (ZAG; Figure 3), and lactoferrin and quantitative difference more expressed in KC patients was identified as ZAG.

Western blot

Western blot analysis for IGKC protein, ZAG, and lactoferrin showed a statistically significant decrease in protein expression (optic mean intensity) in KC patients compared to healthy controls (p<0.05; Figure 4).

Discussion

Despite extensive basic and clinical studies of keratoconus in recent years, the precise mechanisms underlying this pathology still remain largely unknown. In this study, we performed a detailed proteomic analysis of both eyes in patients with bilateral KC and control subjects. Our results show the usefulness of proteomics to find out protein differences between KC patients and healthy subjects in tear samples. By comparing the global protein expression in tears, we found three spots whose expression was greater in the healthy subjects (IGKC protein, ZAG, and lactoferrin) and one spot was found to be more expressed in KC patients (ZAG). It is not strange to detect two or even more spots that belong to the same protein in a 2-DE gel. This is an indication that the protein has different posttranslational modifications. In our case, we have found two different spots for ZAG, one of them is threefold less expressed in KC patients respect to the controls, while the second spot is fourfold more expressed in patients than in control subjects. This is not a contradiction; one posttranslational modification could be more expressed in one group while the other could be just in the opposite way. Proteomics papers have described differences in posttranslational modifications in tears, specially homopolymers formation, protease degradation and glycosilation. Since ZAG is a glycoprotein, we hypothesized that posttranslational modification should be a glycosilation [19]. Nevertheless, what it is important is that overall the protein is more expressed in controls than in patients, as reflected in the western-blot results, where each band includes all posttranslational modifications (spots in 2-DE gels) of a protein. Therefore, we conclude that ZAG is under-expressed in KC patients respect to healthy subjects.

Tear lactoferrin is an 82 kDa protein produced in the acinar cells of the lacrimal gland. Lactoferrin is present in normal tears of humans and was found to be invariant through age and independent of sex. It has been reported reduced levels in certain diseases such as Sjögren syndrome, idiopathic dry eye, vernal conjunctivitis, contact lens induced giant papillary conjunctivitis, trachoma, herpes simplex keratitis, chronic irritative conjunctivitis keratoconjunctivitis sicca, and clinical dry eye with a marked keratopathy [20]. In tear film lactoferrin is thought to be antimicrobial and antiinflammatory. In particular, it has been demonstrated that Lipopolysaccharide-induced production of IL-6, TNF-α, IL-1-β, and IL-8 is inhibited by lactoferrin, in various human monocytic cell lines [21-24]. Previous studies from our group have demonstrated that the levels of proinflammatory cytokines (interleukin-6 [IL-6] and the tumor necrosis factor alpha [TNF-α]) are significantly increased in the tears of patients with KC [13]. Likewise, it has been postulated that reactive oxygen and nitrogen species are raised in keratocunus that, in turn, induce collagenases and gelatinases responsible for the corneal thinning process in this disease [13,25]. Consequently, it is tempting to postulate that lactoferrin might be protective in keratoconus because of its role in decreasing the expression of proinflammatory cytokines and collagenases.

ZAG is a 41-kDa protein component of plasma, saliva, and tears. It stimulates lipid breakdown in adipocytes and is associated to the extreme weight loss that occurs in some cancers [26]. On the other hand, structural organization and fold is similar to MHC class I antigen-presenting molecule; hence, ZAG may have a role in the expression of the immune response [27]. We have demonstrated that the tears of patients with keratoconus contain a smaller amount of ZAG. Contradictory results were found in patients with Graves ophthalmopathy where the expression of ZAG is increased [28]. However, the specific role of ZAG in tears is not known.

Finally, we found a significant decrease in protein expression for IGKC in KC patients. IGKC and immunoglobulin kappa light chain variable region constitutes the light chains of an antibody and is related with immune response. In recent years, several clinical studies of KC support the idea that its pathogenesis involves an immune and inflammatory component [29,30]. Indeed, previous studies from our group, together with others, have shown that the levels of proinflammatory cytokines are significantly increased in the tears of patients with KC [10,13,31].

The limited sample size of the present study weakens the strength with which we may draw our conclusions. However, the robustness of the results gives support to the need to explore the role of these molecular markers as a new therapeutic or diagnostic tool in the KC. Although sample size is not so large, similar proteomic studies have used similar sample sizes [32]. Another limitation is that proteomics based on two-dimensional gels has a sensitivity in the order of magnitude of nanograms. Therefore, proteins whose expression is below the nanogram scale cannot be detected on the gels. Consequently, we do not rule out the possibility that other differences (not identified because of the sensitivity of staining methods) could exist between these two groups.

In conclusion, this study reveals that IGKC protein, zinc-α2-glycoprotein, and lactoferrin are under-expressed in the tears of patients diagnosed with bilateral KC compared with healthy subjects. However, the precise role of each of these molecular factors still needs to be defined, but the results suggest that the pathogenesis of KC might involve immune and inflammatory processes.

Acknowledgments

This project has been supported by a grant from the Xunta de Galicia, PGIDIT07PXIB918088PR.

References

  1. Krachmer JH, Feder RS, Belin MW. Keratoconus and related noninflammatory corneal thinning disorders. Surv Ophthalmol. 1984; 28:293-322. [PMID: 6230745]
  2. Rabinowitz YS. Keratoconus. Surv Ophthalmol. 1998; 42:297-319. [PMID: 9493273]
  3. Dobbins KR, Price FW, , Jr Whitson WE. Trends in the indications for penetrating keratoplasty in the midwestern United States. Cornea. 2000; 19:813-6. [PMID: 11095055]
  4. Tuft SJ, Moodaley LC, Gregory WM, Davison CR, Buckley RJ. Prognostic factors for the progression of keratoconus. Ophthalmology. 1994; 101:439-47. [PMID: 8127564]
  5. Kaldawy RM, Wagner J, Ching S, Seigel GM. Evidence of apoptotic cell death in keratoconus. Cornea. 2002; 21:206-9. [PMID: 11862097]
  6. Erie JC, Patel SV, McLaren JW, Nau CB, Hodge DO, Bourne WM. Keratocyte density in keratoconus. A confocal microscopy study(a). Am J Ophthalmol. 2002; 134:689-95. [PMID: 12429244]
  7. Wilson SE, He YG, Weng J, Li Q, McDowall AW, Vital M, Chwang EL. Epithelial injury induces keratocyte apoptosis: hypothesized role for the interleukin-1 system in the modulation of corneal tissue organization and wound healing. Exp Eye Res. 1996; 62:325-7. [PMID: 8795451]
  8. Rahi A, Davies P, Ruben M, Lobascher D, Menon J. Keratoconus and coexisting atopic disease. Br J Ophthalmol. 1977; 61:761-4. [PMID: 603783]
  9. Chandler JW. Immunology of the ocular surface. Int Ophthalmol Clin. 1985; 25:13-23. [PMID: 3891657]
  10. Bonini S, Lambiase A, Sgrulletta R, Bonini S. Allergic chronic inflammation of the ocular surface in vernal keratoconjunctivitis. Curr Opin Allergy Clin Immunol. 2003; 3:381-7. [PMID: 14501439]
  11. Zhou L, Sawaguchi S, Twining SS, Sugar J, Feder RS, Yue BY. Expression of degradative enzymes and protease inhibitors in corneas with keratoconus. Invest Ophthalmol Vis Sci. 1998; 39:1117-24. [PMID: 9620070]
  12. Collier SA, Madigan MC, Penfold PL. Expression of membrane-type 1 matrix metalloproteinase (MT1-MMP) and MMP-2 in normal and keratoconus corneas. Curr Eye Res. 2000; 21:662-8. [PMID: 11148603]
  13. Lema I, Duran JA. Inflammatory molecules in the tears of patients with keratoconus. Ophthalmology. 2005; 112:654-9. [PMID: 15808258]
  14. de Freitas Campos C, Cole N, Van Dyk D, Walsh BJ, Diakos P, Almeida D, Torrecilhas A, Laus JL, Willcox MD. Proteomic analysis of dog tears for potential cancer markers. Res Vet Sci. 2008; 85:349-52. [PMID: 18164356]
  15. Ananthi S, Chitra T, Bini R, Prajna NV, Lalitha P, Dharmalingam K. Comparative analysis of the tear protein profile in mycotic keratitis patients. Mol Vis. 2008; 14:500-7. [PMID: 18385783]
  16. de Souza GA, Godoy LM, Mann M. Identification of 491 proteins in the tear fluid proteome reveals a large number of proteases and protease inhibitors. Genome Biol. 2006; 7:R72 [PMID: 16901338]
  17. Zadnik K, Barr JT, Gordon MO, Edrington TB. Biomicroscopic signs and disease severity in keratoconus. Collaborative Longitudinal Evaluation of Keratoconus (CLEK) Study Group. Cornea. 1996; 15:139-46. [PMID: 8925661]
  18. Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing of proteins from silver-stained polyacrylamide gels. Anal Chem. 1996; 68:850-8. [PMID: 8779443]
  19. Green-Church KB, Nichols KK, Kleinholz NM, Zhang L, Nichols JJ. Investigation of the human tear film proteome using multiple proteomic approaches. Mol Vis. 2008; 14:456-70. [PMID: 18334958]
  20. Flanagan JL, Willcox MD. Role of lactoferrin in the tear film. Biochimie. 2009; 91:35-43. [PMID: 18718499]
  21. Puddu P, Valenti P, Gessani S. Immunomodulatory effects of lactoferrin on antigen presenting cells. Biochimie. 2009; 91:11-8. [PMID: 18539153]
  22. Choe YH, Lee SW. Effect of lactoferrin on the production of tumor necrosis factor-alpha and nitric oxide. J Cell Biochem. 1999; 76:30-6. [PMID: 10580998]
  23. Mattsby-Baltzer I, Roseanu A, Motas C, Elverfors J, Engberg I, Hanson LA. Lactoferrin or a fragment thereof inhibits the endotoxin-induced interleukin-6 response in human monocytic cells. Pediatr Res. 1996; 40:257-62. [PMID: 8827774]
  24. Håversen L, Ohlsson BG, Hahn-Zoric M, Hanson LA, Mattsby-Baltzer I. Lactoferrin down-regulates the LPS-induced cytokine production in monocytic cells via NF-kappa B. Cell Immunol. 2002; 220:83-95. [PMID: 12657243]
  25. Shoham A, Hadziahmetovic M, Dunaief JL, Mydlarski MB, Schipper HM. Oxidative stress in diseases of the human cornea. Free Radic Biol Med. 2008; 45:1047-55. [PMID: 18718524]
  26. Russell ST, Zimmerman TP, Domin BA, Tisdale MJ. Induction of lipolysis in vitro and loss of body fat in vivo by zinc-alpha-(2)-glycoprotein. Biochim Biophys Acta. 2004; 1636:59-68. [PMID: 14984739]
  27. Hassan MI, Waheed A, Yadav S, Singh TP, Ahmad F. Zinc alpha 2-glycoprotein: a multidisciplinary protein. Mol Cancer Res. 2008; 6:892-906. [PMID: 18567794]
  28. Baker GR, Morton M, Rajapaska RS, Bullock M, Gullu S, Mazzi B, Ludgate M. Altered tear composition in smokers and patients with Graves ophthalmopathy. Arch Ophthalmol. 2006; 124:1451-6. [PMID: 17030713]
  29. Chandler JW. Ocular surface immunology. In: Pepose JS, Holland GN, Wilhelmus KR, editors. Ocular infection and immunity. St Louis: Mosby; 1996. p. 104–11.
  30. Bonini S, Lambiase A, Juhas T, Rama P. Inflammatory immune-associated diseases of the cornea. In: Ben Ezra D, editor. Ocular inflammation: basic and clinical concepts. London: Martin Dunitz; 1999. p. 151–68.
  31. Lema I, Sobrino T, Durán JA, Brea D, Díez-Feijoo E. Subclinical keratoconus and inflammatory molecules from tears. Br J Ophthalmol. 2009; 93:820-4. [PMID: 19304583]
  32. Srivastava OP, Chandrasekaran D, Pfister RR. Molecular changes in selected epithelial proteins in human keratoconus corneas compared to normal corneas. Mol Vis. 2006; 12:1615-25. [PMID: 17200661]