Vitelline Membrane Outer Layer 1 Homolog Interacts With Lysozyme C and Promotes the Stabilization of Tear Film

Abstract

Purpose.: The aim of this study was to explore the possible interactions between vitelline membrane outer layer 1 homolog (VMO1) and other tear proteins and to determine the function of VMO1 in tear fluid.

Methods.: Interactions between recombinant human VMO1 and several abundant tear proteins were determined by dot blot, His pull-down, immunoprecipitation, and Western blot assays, as well as by computer-assisted prediction and modeling of molecular interactions. Kirby-Bauer antibiotic testing was performed to determine whether VMO1 possesses antimicrobial activity. Tear samples were collected from dry eye patients and from healthy controls. The role of VMO1 in maintaining the stability of tear film was investigated by measurement of contact angles on Teflon, tear break-up time (TBUT) and the time-dependent reduction in tear film integrity in mice.

Results.: Vitelline membrane outer layer 1 homolog showed an interaction with lysozyme C (LYSC) in the dot-blot, His pull-down, and immunoprecipitation assays. Vitelline membrane outer layer 1 homolog revealed no zones of growth inhibition of standard strains of Staphylococcus aureus and Escherichia coli. Tears presented smaller contact angles on Teflon surfaces after the addition of VMO1 (P < 0.05). Vitelline membrane outer layer 1 homolog–treated mice showed longer TBUTs (P < 0.05). Tear films from VMO1-treated mice maintained their integrity for longer periods of time than tear films from the control group, and this effect was dose-dependent.

Conclusions.: Vitelline membrane outer layer 1 homolog interacts with LYSC and has positive effects on the stabilization of tear film.

Introduction

Tear fluid forms a thin film that has important functions in maintaining the normal physiology of the ocular surface.¹ Tears consist primarily of lipids, proteins, inorganic salts, and mucous components.² Tear proteins play important roles in defending the ocular surface from external harm, and they also modulate ocular wound healing and contribute to the maintenance of a stable tear film.^3,4 Alteration of tear proteins is associated with ocular surface diseases, such as dry eye.^5,6 

We previously demonstrated the presence of large amounts of vitelline membrane outer layer protein 1 homolog (VMO1) in camel tears.^7,8 Later, the presence of trace amounts of VMO1, as well as 1543 other proteins, in human tears was reported by Zhou et al.⁹ Camels, which live in a harsh and dry environment characterized by strong winds and high levels of ultraviolet radiation, have evolved many physiological adaptations to these conditions. In addition to ocular morphological adaptions, such as dense eyelashes, mobile lower lids, and highly developed lacrimal glands, tear proteins, especially those that are present in high concentration in camel tear film (e.g., VMO1), may contribute significantly to maintaining a healthy tear film.¹⁰ 

Vitelline membrane outer layer 1 homolog was first characterized in the outer layer of the vitelline membrane of hen's eggs, where it is present together with lysozyme, VMO2, and ovomucin.¹¹ Determination of the crystal structure of VMO1 revealed that VMO1 has a unique structure and that it may interact with glycosylated proteins.¹² We assumed that VMO1 may interact with other tear proteins and contribute to the establishment of a stable tear film. Full-length human recombinant VMO1 fused with a polyhistidine tag at the C-terminus was used in vitro and in vivo to validate our hypothesis. 

Materials and Methods

Dot-Blot Assays

To assess the interactions between VMO1 and three abundant tear proteins (>1 μg/μL), lactoferrin (TRFL), lysozyme C (LYSC), and lipocalin 1 (LCN1), dot-blot assays were performed as previously described.¹³ Briefly, solutions containing LYSC, TRFL (both from Abcam, Ltd., Cambridge, UK) and LCN1 (R&D Systems, Inc., Minneapolis, MN, USA) at increasing concentrations were spotted onto nitrocellulose membranes. After air drying, the membranes were blocked with BSA (10% in 20 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20, pH 7.4) for 1 hour and incubated with or without 4 μg of his-VMO1 (Abcam, Ltd.) in 500 μL binding buffer (20 mM sodium phosphate, 150 mM NaCl, pH 7.4) at 4°C overnight. Finally, the membrane was washed, and bound his-VMO1 was detected using an antibody against VMO1 (1:5000; GeneTex, Inc., Irvine, CA, USA). To exclude the possibility that his-tag (HHHHHH, MISC-014; Chinese Peptide Company, Hangzhou, China) may interact with LYSC, the his-tag, instead of his-VMO1, was used to perform dot-blot assay against LYSC. The method was similar to our above description. 

Predicting Interactions of VMO1

The structures of VMO1 (Q7Z5L0, UPI000004BA7A), LYSC (P61626, UPI000012EA62), TRFL (P02788, UPI000067C9DC), and LCN1 (P31025, UPI0000040631) were built using the homology method in Swissmodel (http://swissmodel.expasy.org/, in the public domain).^14,15 The Hex docking software (http://hex.loria.fr/, in the public domain) was used to predict interactions of VMO1 with the three other proteins.¹⁶ Following the Hex docking process, energy conformation scores were obtained and used to create a molecular model of the protein-protein interactions using Chimera software (http://www.cgl.ucsf.edu/chimera, in the public domain).¹⁷ 

His Pull-Down Assays

To characterize the interactions between his-VMO1 and LYSC, pull-down assays were performed as described by Morenilla-Palao et al.,¹⁸ with modifications. One microgram of his-VMO1 was mixed with 10 μL Dynabeads (Life Technologies, Ltd., Oslo, Norway) in 100 μL binding buffer for 20 minutes at 22°C. Then, 1 μg LYSC was added, and the mixture was incubated overnight at 4°C. The bead–protein complex was washed three times with binding buffer containing various concentrations of imidazole, and the beads were separated from the supernatant using a DynaMag-Spin apparatus (Life Technologies, Ltd.). The protein complexes were eluted from the beads with 40 μL of elution buffer (300 mM imidazole, 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, and 0.01% Tween-20) and analyzed by Western blotting.  

Immunoprecipitation Assays

For immunoprecipitation, 1 μg his-VMO1 and 1 μg LYSC were incubated with rabbit anti-LYSC antibody (5 μg/mL; Abcam, Ltd.) or rabbit control IgG (5 μg/mL; Cell Signaling Tech., Inc., Danvers, MA, USA) in 500 μL binding buffer for 16 hours at 4°C. Then, 40 μL protein A/G agarose suspension (Calbiochem, Darmstadt, Germany) was added, and the mixture was incubated for 2 hours at 4°C. Immunoprecipitated complexes bound to beads were boiled in SDS-PAGE sample buffer before Western blotting.¹⁹ 

Western Blotting Analysis

Western blotting was performed as previously described.⁸ Briefly, protein complexes were separated by 12% acrylamide SDS-PAGE and electroblotted onto polyvinylidene fluoride (PVDF) membranes (Merck Millipore Corp., Billerica, MA, USA). Proteins were detected using the Phototope-HRP Western Blot Detection System (Cell Signaling) with VMO1 or LYSC antibody (1:5000; Abcam, Ltd.). The visualized protein bands were recorded on a Kodak Image Station 4000MM (Eastman Kodak, Rochester, NY, USA). 

Antimicrobial Activity Testing

Kirby-Bauer antibiotic testing was performed to determine whether VMO1 inhibits the growth of bacteria.²⁰ The testing was conducted in accordance with the standard performance guidelines for antimicrobial susceptibility testing of the Clinical and Laboratory Standards Institute (http://www.clsi.org/, in the public domain) using two standard bacterial strains, Staphylococcus aureus (ATCC 25923) and Escherichia coli (ATCC 25922). Disks of filter paper 6 mm in diameter were impregnated with ampicillin, LYSC, his-VMO1, or sterile water and applied to Mueller-Hinton agar plates that had been inoculated with S. aureus or E. coli. After incubation of the plates at 35°C for 18 hours, zones of inhibition were photographed. 

Tear Sample Collection

Tear samples were collected from healthy controls and dry eye patients, who had completed the Ocular Surface Disease Index (OSDI) questionnaire before tear collection.²¹ Tear film breakup time (TBUT), scores of corneal staining with 1% sodium fluorescein, and Schirmer I test values were recorded. Glass capillary tubes (Drummond Scientific Co., Broomall, PA, USA) were used to collect tears after elicitation of the yawning reflex²² in healthy individuals (three males, mean age 26 ± 1 years) and dry eye patients (five males, mean age 27 ± 1 years, level 3 to 4 according to the Dry Eye Workshop²³). After centrifugation, the tear samples were pooled separately and stored at −80°C until use. All work involving human subjects was conducted in accordance with the guidelines of the Declaration of Helsinki and the Ethical Committee of the Zhongshan Ophthalmic Center, Sun Yat-sen University. Written informed consent was obtained from each participant. 

Contact Angle Detection

For contact angle measurements, 15-μL droplets were applied to a Teflon surface using a pipette. The samples were divided into five groups: tears, tears+H₂O, tears+his-VMO1, tears+ITRA, and water. After equilibration for 10 minutes at 80% ± 5% relative humidity and 20°C, the droplets were photographed using a camera (Canon EOS 5D Mark II; Canon, Inc., Tokyo, Japan) equipped with a +8 close-up lens. After aspiration, 10-μL droplets were applied within the same areas and the contact angles were photographed again.²⁴ The contact angles were analyzed using an ellipse fitting algorithm in ImageJ (http://imagej.nih.gov/ij/; provided in the public domain by the National Institutes of Health, Bethesda, MD, USA).^25–27 

Observation of Tear Film In Vivo

Twenty 10-week-old C57BL/6 mice (obtained from the Ophthalmic Animal Laboratory, Sun Yat-sen University, Guangzhou, China) were used for the in vivo study. All procedures involving animals were performed in accordance with the standards of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and approved by the Animal Ethics Committee of Zhongshan Ophthalmic Center. The mice were randomly assigned to groups treated with his-VMO1 or 0.9% NaCl solution (saline for injection; Tianjin Pharmaceutical Co., Ltd., Tianjin, China).²⁸ Recombinant human his-VMO1 was dissolved with 0.9% NaCl solution at a final concentration of 7.5 or 3.0 μg/μL, mixed with 0.25% sodium fluorescein, and administered under single-blinded conditions for the observer. In addition, the sequence homology between human VMO1 and mouse VMO1 was analyzed by amino acid sequence alignments (http://www.uniprot.org/, in the public domain). Mouse VMO1 (Q5SXG7) have 71% sequence identity with human VMO1 (Q7Z5L0), indicating they are highly homologous. 

Animals were anesthetized by intraperitoneal injection of 10% chloral hydrate (300 mg/kg) and observed under a slit lamp microscope (Sun Kingdom Medical Instruments, Chongqing, China), followed by the application of 1 μL 0.9% NaCl solution containing his-VMO1 (7.5 or 3.0 μg/μL) or 0.9% NaCl solution onto the ocular surface. The TBUT was recorded with cobalt blue light using a stopwatch. Three measurements were averaged for each eye.²⁸ In addition, changes in tear film after treatments were photographed using a Canon EOS 5D Mark II camera connected to a slit lamp in continuous shooting mode (three pictures per second) from 0 to 12 seconds. Using ImageJ software, the integrated density of green color from the same region of each cornea was calculated to assess the retention of the tears remaining on the cornea.²⁹ 

Statistical Analysis

All experiments were repeated at least three times unless otherwise indicated. One-way ANOVA was performed to analyze the results of the dot-blot, His pull-down, immunoprecipitation, and contact angle measurement assays, as well as the integrated density of the tear film, followed by least significant difference (LSD) post hoc analysis. Significant differences in TBUT were evaluated using independent t-tests. Differences in P values of less than 0.05 were considered statistically significant. 

Results

Interactions Between VMO1 and LYSC In Vitro

In the dot-blot assays, his-VMO1 was present in the areas with various amounts of LYSC, whereas no his-VMO1 was detected in areas with TRFL or LCN1 (Figs. 1A, 1B). The binding of his-VMO1 to LYSC was dose-dependent. When his-VMO1 was substituted with his-tag, LYSC dots could not be detected in the areas of his-tag (Fig. 1C). 

Figure 1

 

Dot-blot showing the binding of his-VMO1 to tear proteins. (A) Two microliters of LYSC, TRFL, and LCN1 containing 0 to 1000 ng of the respective proteins were spotted onto nitrocellulose membranes, and the membranes were incubated overnight in the presence or absence of his-VMO1. Bound his-VMO1 was detected using an anti-VMO1 antibody. (B) The relative integrated densities (IntDen) of the LYSC dots, shown as the mean ± SD (n = 3). (C) Two microliters of his-VMO1 or his-tag (containing 0, 0.25, 0.5, 5, 50 pmol/L, respectively, for both) were spotted onto nitrocellulose membranes, followed by incubation with or without LYSC. Bound LYSC was detected by anti-LYSC antibody.

The interaction between VMO1 and LYSC and the lack of interaction of VMO1 with TRFL and LCN1 was supported by computer-assisted molecular modeling. We found that the final binding energy conformation score for VMO1 and LYSC was −869.85, whereas the binding energy conformation scores for VMO1 and TRFL and VMO1 and LCN1 were −328 and −371, respectively. In the Chimera model, certain amino acids in VMO1 (i.e., Q153, E110, and Q70) can form hydrogen bonds with amino acids R59, R59/S100, and D85 of LYSC (Fig. 2). Besides the amino acids, N141, R143, L152, G154, S158, and W159 of VMO1 directly contacted amino acids K19, A101, Q104, D105, and N106 of LYSC. The model also showed that the binding area of LYSC was not in the LYSC active center that was previously reported.³⁰ 

Figure 2

 

Interaction model of VMO1 and LYSC analyzed by computer-assisted programs. (A) Interaction models of VMO1 (red) and LYSC (blue). (B) Enlarged yellow rectangular area in (A). In the model, Q153, E110, and Q70 of VOM1 bind to R59, R59/S100, and D85 of LYSC through hydrogen bonds.

To confirm the interaction between VMO1 and LYSC, His pull-down and immunoprecipitation assays were conducted. In the His pull-down assays, we found that with increasing concentrations of imidazole, which competitively binds to the Dynabeads in the wash buffer, the amount of bound LYSC decreased in parallel with the decrease in bound his-VMO1 (Fig. 3A). In the immunoprecipitation assays, both his-VMO1 and LYSC were detected in the anti-LYSC complexes by Western blotting with anti-VMO1 and anti-LYSC (Fig. 3B). 

Figure 3

 

Western blot assays demonstrating the association of VMO1 with LYSC after pull-down and immunoprecipitation. (A) One microgram of his-VMO1 and 1 μg LYSC were incubated together overnight, followed by washing the beads in increasing concentrations of imidazole. The pull-down complex was detected by Western blotting. The band intensities of both VMO1 and LYSC were reduced when the concentration of imidazole in the wash buffer was increased. (B) One microgram of his-VMO1 and 1 μg LYSC were incubated together at 4°C overnight. Rabbit anti-lysozyme antibody, rabbit control IgG, or beads only were then added to the protein mixtures, and immune complexes were detected by Western blotting. Each experiment was performed in triplicate.

VMO1 Shows No Antibacterial Activity

Disk diffusion antibiotic sensitivity tests were performed to assess the antibacterial activity of VMO1 (Fig. 4). After incubation on Mueller-Hinton agar plates for 18 hours, discs soaked with ampicillin or LYSC showed zones of inhibition against both S. aureus and E. coli. However, no zones of inhibition of bacterial growth were found around discs soaked with his-VMO1 or sterile water in plates containing either S. aureus or E. coli. 

Figure 4

 

Antimicrobial activity of his-VMO1. Discs containing ampicillin (1, 10 μg), his-VMO1 (2, 30 μg), lysozyme (3, 30 μg), and sterile water (4) were placed on the surface of an agar plate. Ampicillin and LYSC show bacteriostatic rings for S. aureus and E. coli. His-VMO1 is inactive, as is the water control.

VMO1 Decreases the Contact Angle of Tears on Teflon

Experiments on Teflon revealed an influence of his-VMO1 on the wettability of hydrophobic surfaces (Table; Fig. 5). Droplets, 15 μL in volume, of tears, tears+H₂O, tears+his-VMO1, tears+ITRA, and water were placed on a clean Teflon surface (Fig. 5A). Droplets of tears to which his-VMO1 had been added showed reduced initial contact angles (107.78°) compared with those of tears alone (109.51°), tears with ITRA (113.15°), tears with water (112.36°), and water alone (116.43°), indicating a reduction in surface tension. After aspiration, all of the corresponding contact angles were significantly reduced (Fig. 5B); the greatest reduction (31.23°) was observed in the tears mixed with his-VMO1. 

Figure 5

 

Measurements of the contact angles formed by droplets of solutions on a Teflon surface. (A) Each solution (15 μL) was applied to Teflon. From left to right: (a) water; (b) tears; (c) tears+H₂O; (d) tears+his-VMO1; and (e) tears+ITRA. The final concentrations of his-VMO1 and ITRA were both 7.5 μg/μL. (B) Photographs of each droplet were obtained 10 minutes after aspiration and reapplication of 10 μL of each solution in situ.

Table

 

Contact Angle Measurements of Tear Compounds (n = 3)

Compound	Direct Application	Aspiration With Reapplication of Solution
Water	116.43 ± 0.69	118.18 ± 0.33
Tear	109.51 ± 0.87	82.79 ± 1.20
Tear+H2O	112.36 ± 0.49	89.33 ± 2.29
Tear+his-VMO1	107.78 ± 0.48*	75.13 ± 1.15*
Tear+ITRA	113.15 ± 0.46	87.37 ± 0.74

*  P < 0.01, compared with other groups. Multiple comparison between the groups was performed using one-way ANOVA test followed by LSD post hoc test.

Vitelline Membrane Outer Layer 1 Homolog Increases the Stability of Tear Film

To determine whether VMO1 is advantageous in building a stable tear film, droplets of his-VMO1 solution (7.5 μg/μL) were applied to the ocular surfaces of mice. His-VMO1–treated mice showed significantly longer TBUTs than the group treated with 0.9% NaCl solution (P < 0.05; Fig. 6A). When the fluorescein retention of mouse cornea was evaluated by the image analysis, the results displayed a significant increase in the stability of mouse tear film in his-VMO1–treated eyes compared with that in 0.9% NaCl solution–treated animals (P < 0.05; Fig. 6B). Corneal fluorescein retention was time-dependent in all three groups of mice tested. Compared with the NaCl controls, the application of VMO1 at 7.5 μg/μL resulted in a significant increase in tear film retention from 3 to 12 seconds (P < 0.05). The application of eye drops containing 3.0 μg/μL VMO1 also increased tear retention, but the effect was less than that obtained with 7.5 μg/μL VMO1. 

Figure 6

 

Effects of his-VMO1 on mouse tear films detected by fluorescein staining. (A) His-VMO1 increases TBUT in mice. Droplets of 0.9% NaCl solution containing 7.5 μg/μL his-VMO1 were applied to the ocular surfaces of C57 BL/6 mice. 0.9% NaCl solution was used as a control. The data are presented as the mean ± SD (n = 20). *P < 0.05, his-VMO1-treated versus 0.9% NaCl-treated. (B) Relative fluorescein retention of tears measured by integrating the density of the fluorescein signal. The data are presented as the mean ± SD (n = 4). *P < 0.05, 7.5 μg/μL his-VMO1 versus 0.9% NaCl; +P < 0.05, 7.5 μg/μL his-VMO1 versus 3.0 μg/μL his-VMO1.

Discussion

The composition and physical properties of tear film have been intensively studied in recent years.³¹ It has been reported that interactions between lipids and proteins (LCN, LYSC, TRFL, and mucins) can reduce the surface tension of the tear film.^32–35 We observed a dose-dependent reduction in the binding of his-VMO1 to LYSC in His pull-down assays, dot-blot, and immunoprecipitation assays. Up to now, few data are available in the STRING database (http://string-db.org/, in the public domain) regarding possible specific binding partners of VMO1.³⁶ In the present study, of the three abundant tear proteins studied, LYSC was shown to bind specifically to his-VMO1. However, we cannot exclude the possibility that other tear proteins, especially mucins, may interact with VMO1 in vivo. 

In previous studies, VMO1 was reported to be abundant in many exocrine glands and secretions, including breast, pancreas, cerebrospinal fluid, urine, and respiratory secretions, and has been found in small quantities in human tears and a high abundance in camel tears.^7,9,37,38 In the present study, we found that the addition of his-VMO1 to tear fluid produced smaller contact angles, and longer TBUTs in mice. These observations imply that his-VMO1 may have positive effects on tear film stability in humans when it is used as eye drops. 

Tear film stability has been associated with low surface tension, and tear proteins contribute greatly to lowering the surface tension of tear fluid.³⁹ However, the tear fluid of dry eye patients is thought to have a higher surface tension, resulting in greater instability of the tear film.⁴⁰ We showed that the addition of his-VMO1 to tears reduced the contact angle of tears on Teflon and yielded a significant decrease in the contact angle after aspiration and reapplication, indicating a decrease in the surface tension of the tear fluid. In our experiments, the protein ITRA (isoionic point: 4.86, molecular weight: 24.0 kDa), which is not found in tears but has a similar isoionic point and molecular weight as VMO1 (isoionic point: 4.65, molecular weight: 21.5 kDa) (http://www.expasy.org/, in the public domain), was used as a control. It did not reduce the initial contact angle of tears placed on a Teflon surface. It was reported that interactions between lysozyme and lipids led the reduction of surface tension.⁴¹ Our results demonstrated that VMO1 interacted with LYSC, the latter might further associate with lipids and other tear proteins, leading to the reduction of the contact angle. However, the exact mechanism needs to be clarified in further study. 

The results of the in vivo experiments described in this work further supported our conclusions. The negative correlation between the surface tension of tear film and TBUT is widely acknowledged, and measurement of TBUT is regarded as a simple and repeatable method for determining tear film stability.⁴² We found that the TBUT of his-VMO1–treated mice is longer than that of 0.9% NaCl solution–treated mice. By photographing mouse tear films stained with fluorescein, we were able to show that his-VMO1–treated animals had a longer fluorescein retention time; this effect was more obvious in animals treated with 7.5 μg/μL his-VMO1 than in animals treated with 3.0 μg/μL his-VMO1. Thus, our findings revealed that his-VMO1 has the ability to stabilize tear films both in vitro and in vivo. 

In summary, the results of the current study demonstrate an interaction between VMO1 and LYSC. Vitelline membrane outer layer 1 homolog does not display any bacteriostatic activities. Furthermore, the results presented here show that VMO1 reduces the surface tension of tears, and produces a more stable tear film in vivo. These results suggest that VMO1 may be useful as a novel therapy for dry eye syndrome. 

Acknowledgments

Supported in part by a grant from the National Natural Science Foundation of China (No. 81170827), and by the Fundamental Research Funds of State Key Laboratory, China. 

Disclosure: Z. Wang, None; Z. Chen, None; Q. Yang, None; Y. Jiang, None; L. Lin, None; X. Liu, None; K. Wu, None