IOVS Journal Watch
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Glasgow, B. J.
Right arrow Articles by Knobler, C. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Glasgow, B. J.
Right arrow Articles by Knobler, C. M.
(Investigative Ophthalmology and Visual Science. 1999;40:3100-3107.)
© 1999 by The Association for Research in Vision and Ophthalmology, Inc.

Tear Lipocalins

Potential Lipid Scavengers for the Corneal Surface

Ben J. Glasgow1,2, Gary Marshall3, Oktay K. Gasymov1, Adil R. Abduragimov1,2, Taleh N. Yusifov1 and Charles M. Knobler3

1 From the Departments of Pathology, 2 Ophthalmology, the Jules Stein Eye Institute, and the 3 Department of Chemistry and Biochemistry, University of California, Los Angeles, School of Medicine.


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
PURPOSE. To investigate the dynamic effect of tear lipocalins (TLs), the major lipid-binding protein in tears, at aqueous–cornea and lipid–aqueous interfaces, and their potential contribution to surface tension in the tear film.

METHODS. Human apo- and holo-TLs were applied to the aqueous subphase in a Langmuir trough, and changes in surface pressure were measured. Changes in the contact angle of tear components were observed on Teflon and ferric-stearate–treated surfaces. A nitroxide-labeled derivative of lauric acid and a fluorescence-labeled derivative of palmitic acid were used to monitor the dynamic interaction of lipid removed from a hydrophobic surface by the major tear components in solution.

RESULTS. TLs increase the surface pressure at the aqueous–air interface by penetrating, spreading, and rearranging on the surface. Apo-TLs show a longer diffusion-dependent induction time than holo-TLs due to more extensive oligomerization of the apoprotein. Kinetic analysis of relaxation time suggests that apo-TLs have more rapid surface penetration and rearrangement than holo-TLs, indicative of a more flexible structure in apo-TLs. TLs reduce the contact angle of solutions on lipid films, a property that is greater with TLs than other tear proteins. TLs, unlike lysozyme and lactoferrin, remove labeled lipids from hydrophobic surfaces and deliver them into solution.

CONCLUSIONS. TLs are potent lipid-binding proteins that increase the surface pressure of aqueous solutions while scavenging lipids from hydrophobic surfaces and delivering them to the aqueous phase of tears. These data suggest important functional roles for TLs in maintaining the integrity of the tear film.


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
The notions about the structure of the tear film are controversial and evolving. The traditional concept has been that the tear film, 7 µm in total thickness, is composed of three distinct layers1 : a 0.02- to 0.05-µm precorneal mucin layer, a 4- to 7-µm middle aqueous layer, and a 0.01-µm external lipid layer.2 3 Ultrastructural observation of the rat tear film suggests that mucins may exist as a network distributed in the aqueous body of the tear film, rather than an as a separate discrete layer.4 Nichols described a mucin layer that measures 2 to 7 µm above the corneal surface and is intimately associated with corneal microvilli and presumably is anchored to the glyocalyx of the conjunctiva.5 Laser confocal microscopy suggests that the human aqueous–mucin layer may be much thicker than first estimated, 41 to 46 µm.6 In the current model of the tear film, the aqueous–mucin layer is covered by two thin layers of lipid. Polar lipids such as phospholipids lie adjacent to the aqueous–mucin layer and nonpolar lipids, such as cholesterol, are present at the tear–air interface.7 8 9 The thickness of the lipid layer, estimated by observation of interference patterns, measures between 0.06 and 0.18 µm in the open human eye.10 11

Much information has been published regarding the hydrophobicity of the corneal surface before and after mucolytic or abrasive procedures were applied. Holly12 proposed that the corneal epithelium has a low surface energy and a critical surface tension of 28 dynes/cm (with no adsorbed mucins). Mucins presumably raise the critical surface tension of the cornea to be more wettable (38 dynes/cm).2 However, other studies provide scanning electron microscopic evidence to suggest that methods used in the previous studies cause severe damage to the corneal epithelium and may be flawed on a theoretical basis as well.13 More recent evidence suggests that the surface tension of the cornea is much higher (67.5–72 dynes/cm) than previously determined, with minimal change after treatment with mucolytic agents.14 15 16 Lipid contamination of the corneal surface, either because of loss of mucin or contamination of a mucinous surface, lowers the surface tension and renders the cornea unwettable.2 17 This situation is possible whenever the tear film thins, such as in dry eye disease.18 19 The mucin covering the cornea may be compromised in dry eye disease, as is suggested by the rose bengal staining of epithelial cells.20 A mechanism to remove meibomian lipids that inadvertently come to rest on the mucins or directly on the corneal epithelium is necessary to prevent drying of the ocular surface.21

Tear lipocalins (TLs), the major lipid-binding proteins in tears, bind a broad array of lipids including fatty acids, cholesterol, phospholipids and glycolipids.22 TLs increase the solubility of lipids in the tear film and may promote rapid equilibration to form a layer of lipid on the tear film surface. Lipids are released from TLs in the acidic local environment achieved by lateral proton conduction at the aqueous-lipid interfaces.23 The dissociation of the lipid–protein complex at the interface is associated with defined structural changes in TLs, including a reduction in overall structural rigidity, decreased ß structure, relaxation of aromatic–side-chain asymmetry and a transition through a molten globule state.23 The relative affinity of TLs to tear film lipids has been determined from displacement assays using fluorescent ligands.24 TLs have greatest affinity for the relatively insoluble long-chain fatty acids and phospholipids. Therefore, TLs are potential scavengers of lipids, particularly those that could perturb the wettability of the cornea. Because TLs are the only protein component in tears that demonstrates significant binding of lipid components,22 they comprise the most suitable protein candidate in tears to remove lipids from the cornea. TLs are promiscuous in binding a broad array of lipids and are ideally suited for scavenging the wide range of meibomian lipids that could spill onto the corneal surface.25 This study was designed to determine the role of TLs as potential scavengers and carriers of lipid from the corneal surface.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
Purification of Tear Proteins
TLs were purified from pooled human tear samples by size-exclusion and ion-exchange column chromatographies, as previously described.22 26 Delipidation was performed by chloroform-methanol extraction.22 Protein concentrations were determined by the Biuret method.27 Lysozyme and lactoferrin from human tears were purified by column chromatographies, as previously described.22 28 29 Purity of the final samples was assessed by sodium dodecyl sulfate (SDS) tricine polyacrylamide gel electrophoresis (PAGE).22 For experiments using tears depleted of TLs, 2 ml of tears was applied to the Sephadex G-100 (Sigma, St. Louis, MO) column as described previously,22 except the elution buffer was 0.01 M ammonium acetate (pH 7.4; Fig. 1A ). Eluted fractions, excluding 60 to 72 that contained TLs (Fig. 1A) , were combined, lyophilized and redissolved in 2 ml water. Gel electrophoresis confirmed that the resultant mixture was depleted of TLs (Fig. 1B) . Nanopure water was used in all experiments.



View larger version (27K):
[in this window]
[in a new window]
  		Figure 1. (A) Gel filtration of tears in 0.01 M ammonium acetate buffer. Fractions 60 through 72 contain TLs. All other fractions were pooled, lyophilized, and recombined in water to the original volume to create TL-depleted tears. (B) Coomassie-stained gel of fractions. Lane 1: 5 µg purified TLs; lane 2: 2 µl tears; lane 3: 2 µl TL-depleted tears

 
Determination of the Multimeric State of TLs
For determination of the multimeric state of purified TLs, gel filtration was performed on a Sephadex G-100 column (2.5 x 120 cm) equilibrated with 50 mM Tris-HCl and 100 mM NaCl (pH 8.4). Four milligrams holo- and apo-TLs was applied to the column. The column was calibrated with the markers bovine serum albumin (68,000 Da), carbonic anhydrase (29,000 Da), and egg white lysozyme (14,500 Da).

TL Adsorption Isotherms in a Langmuir Trough
For these experiments a Langmuir trough, constructed from Teflon, and measuring 125 x 68 x 4 mm was used with a motor-driven barrier to control the area of the liquid–air interface and a pressure transducer fitted with a filter paper Wilhelmy plate to measure surface pressure. The transducer was calibrated to baseline for a clean air–water interface. The temperature of the trough was maintained between 33°C and 34°C by water circulation in the base of the trough and four Peltier elements situated at the bottom of the trough for rapid heating or cooling. A thin sheet of indium-tin oxide–coated heated cover glass prevented evaporation and dust contamination during the experiments. Adsorption isotherms were measured from baseline by slowly injecting TLs into the subphase of the Langmuir trough (0.1 M sodium phosphate buffer, pH 7.0). Surface pressure was recorded on a strip chart recorder for the duration of the experiment. Adsorption isotherms were measured at various concentrations for native TLs and apo-TLs.

Contact Angle Measurements to Assess TL Binding to Lipid Substrates
Contact angle measurements were performed on Teflon and on glass slides coated with thick films of ferric stearate. Droplets of water and solutions deposited on the slides from a microliter syringe were allowed to equilibrate for 10 minutes in a humidified chamber and then imaged with a camera equipped with a close-up lens. The contact angles of the drops were obtained within a precision of ±5° from measurements of the limiting slopes on large-scale printed images. All experiments were performed at 20°C.

Scavenging Properties of TLs by Electron Paramagnetic Resonance and Fluorescence
A two- to threefold molar excess (compared with TLs) of a nitroxide derivative of lauric acid (C-12 spin label)22 was dissolved in ethanol and placed in clean polystyrene cuvettes. After evaporation of ethanol under nitrogen, either TLs (50 µM in 10 mM sodium phosphate, pH 7.3) or buffer alone was gently added to the cuvette containing the spin-labeled lipid and incubated. The entry of spin label into solution was monitored by removing aliquots from the cuvette at various time points. The samples (2 µl) were loaded in quartz capillaries sealed at one end. Electron paramagnetic resonance (EPR) spectra were recorded at X-band with a spectrometer (model E-109; Varian, Sunnyvale, CA) at 2 mW incident microwave power in a loop-gap resonator, as previously described.22 The concentration of spin label in solution was calculated by double integration of the EPR spectra and compared with a standard solution of 2,2,6,6-tetramethylpiperidine-N-oxyl (Tempol; Sigma, St. Louis, MO).22 The total concentration of spin label was plotted versus time. The concentration of bound and free spin label was estimated by double integration of the spectrally titrated EPR signals.

16-(9-Anthroyloxy)palmitic acid (0.04 micromoles; 16-AP; Molecular Probes, Eugene, OR) was dissolved in ethanol and placed in clean quartz cuvettes. After evaporation of ethanol under nitrogen, 0.01 micromoles (5 µM final concentration) of apo-TLs, lactoferrin, or lysozyme in 10 mM sodium phosphate (pH 7.3) were gently added to the cuvette containing the 16-AP. Entry of 16-AP bound to protein in solution was monitored at various time points. Fluorescence measurements were made at 25°C in a thermostated cuvette with a spectrofluorometer (SPEX Fluorolog-3; Jobin Yvon, Edison, NJ), bandwidth for excitation was 2 nm, {lambda} ex = 361 nm, and for emission was 4 nm, {lambda} em = 452 nm. Fluorescence at 467 nm was used to detect free 16-AP in the solution.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
Surface Pressure Measurements in the Langmuir Trough
Experiments in the Langmuir trough demonstrated that native holo-TLs generated a surface pressure with a slow induction phase that began within 30 minutes of injection. The surface pressure increased slowly and began to plateau at approximately 3 hours. In contrast apo-TLs had a longer induction phase of between 30 and 60 minutes, after which the surface pressure increased rapidly and plateaued slightly earlier than that of holo-TLs. The final surface pressure was greater at each concentration for the apo-TLs than for holo-TLs (Fig. 2) . Changes in the surface pressure provide a way of monitoring penetration of protein into the surface and molecular rearrangements of adsorbed protein molecules.30 Analysis of the shape of the adsorption curves provides information regarding the relaxation times of the protein. For holo-TLs, the curves suggested one exponential. The shapes of the adsorption curves for apo-TLs clearly did not have simple exponential forms. They were consistent with two relaxation times, a short one, roughly 40 minutes, and another, roughly 10 times longer.



View larger version (14K):
[in this window]
[in a new window]
  		Figure 2. Comparison of surface pressure–time data for apo- and holo-TLs at various concentrations. (long-dashed line) holo-TLs 2 x 10-4%; (bold solid line) holo-TLs 4 x10-4%; (short-dashed line) apo-TLs 2 x 10-4%; (thin solid line) apo-TLs 4 x 10-4%.

 
Multimeric State of TLs
The elution profiles of apo- and holo-TLs in gel filtration experiments are shown in Figure 3 . Holo-TLs eluted almost entirely in a single peak at a molecular weight of approximately 35 to 36 kDa (dimer). Apo-TLs eluted in three peaks: the major peak at 35 to 36 kDa was a dimer (65%), a second smaller peak at a molecular weight of 70 to 73 kDa was a tetramer, and the remainder eluted in the void volume.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 3. Gel filtration of purified apo- ({circ}) and holo-TLs ({triangleup}). Holo-TL emerged from the column in a single peak at a molecular weight of 35,000 to 36,000 Da, whereas apo-TL was distributed over three peaks at molecular weights of 35,000 to 36,000 Da (65% of protein mass) and 70,000 to 73000 Da (~17% of protein mass). The remainder eluted in the void volume (~18% of protein mass). OD, optical density.

 
Contact Angle Experiments
Multiple experiments on Teflon and thick films of ferric stearate demonstrated the influence of a solution of holo-TLs on the wettability of hydrophobic surfaces (Table 1) . A representative experiment is shown in Figure 4 . Droplets, 40 µl in volume, of buffer, tears, and reconstituted tears depleted of TLs were placed on a clean Teflon surface (Fig. 4A) . The droplets with protein showed reduced initial contact angles for tears (80°) and tears depleted of TLs (85°) indicating a reduction in surface tension compared with buffer (104°). After aspiration, droplets of 20 µl of buffer (a smaller volume was used to ensure that the drop remained within the boundaries of the prior 40-µl droplet) were replaced within the same area (Fig. 4B) . The corresponding contact angles, 54° and 57°, respectively, were reduced indicating, a general interaction typical of most proteins with the Teflon surface (Table 1) . The experiment was performed with purified TLs, lactoferrin, and lysozyme on Teflon (Figs. 4C , 4D ) with similar results. The same experiment was performed on a ferric stearate surface (Figs. 4E , 4F ). The initial contact angle of 146° for buffer is typical of hydrocarbon surfaces. When a 40-µl droplet of tears was applied, the contact angle was slightly lower (133°), indicating a reduction in surface tension compared with the contact angle of tears depleted of TLs (143°). After aspiration and reapplication of 20 µl of buffer, a much lower contact angle was observed (105°), indicating a marked interaction with ferric stearate (Fig. 4F) . When the major protein components of tears were tested, the contact angle was most reduced on the spot where TLs (110°) had been placed, indicating a marked interaction with the ferric stearate surface (Figs. 4G , 4H ). A much smaller reduction was observed for other components, 124° for lactoferrin and 131° for lysozyme.


View this table:
[in this window]
[in a new window]
  		Table 1. Contact Angle Measurements with Tear Components

 


View larger version (72K):
[in this window]
[in a new window]
  		Figure 4. Solution droplets on Teflon and ferric stearate multilayer surfaces from which contact angle measurements were taken. (A) Each solution (40 µl) was applied on Teflon, from left to right: buffer, tears, and TL-depleted tears. (B) Photograph taken 10 minutes after aspiration of the drop in (A) and reapplication of 20 µl buffer in the original spot. (C) Each solution (40 µl) on Teflon, from left to right: TL, lactoferrin, and lysozyme. (D) Photograph taken 10 minutes after aspiration of the drop in (C) and reapplication of 20 µl buffer in the original spot. (E) Each solution (40 µl) was applied on ferric stearate from left to right: buffer, tears and TL-depleted tears. (F) Photograph taken 10 minutes after aspiration of the drop in (E) and reapplication of 20 µl buffer in the original spot. (G) Each solution (40 µl) on ferric stearate, from left to right: TL, lactoferrin, and lysozyme. (H) Photograph taken 10 minutes after aspiration of the drop in (G) and reapplication of 20 µl buffer in the original spot.

 
Scavenging Properties of TLs by EPR and Fluorescence
The dynamic interaction of TLs and the C-12 spin-labeled compound was easily detected in the EPR experiments. Nitroxide-labeled compound, which is mobile (free in solution), had a spectrum characterized by three sharp resonance peaks. The spectrum broadened as the spin label was reduced in motion (bound to protein). All spectra were composites, consisting of bound and free components. The C-12 spin label dried to a polystyrene surface showed no significant diffusion in buffer, even after hours of incubation; no signal was detected in the overlying solution. However, when the solution containing TLs was incubated with the spin label, a signal became evident within 2 minutes. Inspection of the spectra show that the relative heights of the bound and free signals changed with time (Fig. 5) . The spectrum taken from the sample 20 minutes after TLs were added showed a greater amount of bound signal than free (Fig. 5A) . Comparison with the spectra at 150 minutes showed that the overall signal intensity increased with time, and the relative signal from the mobile component of the spin label also increased (Fig. 5B) . The total concentration of spin label in solution increased, and the relative amount of the free spin label increased as the C-12 compound dissociated from TLs (Figs. 5 , 6 ). Regression analysis shows that these data are well represented by a first-order kinetic rate equation. Spectral titration and double integration of the EPR signal (Fig. 5) revealed a concentration of 23.4 µM bound spin label compared with 5.5 µM free spin label at 150 minutes.



View larger version (20K):
[in this window]
[in a new window]
  		Figure 5. EPR spectra of spin-labeled C-12 derivative dried in a cuvette and overlaid with 50 µM TLs in 10 mM sodium phosphate (pH 7.3). At 20 minutes (A) there was a major broad component spectra with a smaller sharp resonance peak. At 150 minutes (B) there was a marked increase in the overall amplitude of the peaks with a greater increase in the amplitude of the sharp resonance peak, indicating a relatively larger free component. The magnetic field scan width is 128 Gauss.

 


View larger version (15K):
[in this window]
[in a new window]
  		Figure 6. Concentration of total C-12 spin label (C12SL) entering the solution with TLs plotted versus time. The concentration of C-12 spin label was determined by double integration of the corresponding EPR signal and standard solution of 2,2,6,6-tetramethylpiperidine-N-oxyl (Tempol; Sigma).

 
Similar results were evident for 16-AP (Fig. 7) . 16-AP demonstrated enhanced fluorescence and a blue shift of the emission peak when protein bound. Fluorescence at 452 nm was detected in cuvettes in which 16-AP was incubated with apo-TLs but not in the cuvettes with lactoferrin or lysozyme. No fluorescence was detected at 467 nm, indicating that 16-AP was not present, free in solution in the cuvettes containing lysozyme and lactoferrin.



View larger version (19K):
[in this window]
[in a new window]
  		Figure 7. Fluorescence intensity produced by 16-AP bound to protein. 16-AP was dried in cuvettes and overlaid with {circ} apo-TLs, {triangleup} lactoferrin or {square} lysozyme (5 µM final concentration) in 10 mM sodium phosphate (pH 7.3). No free 16-AP was detected in the cuvettes containing lactoferrin or lysozyme.

 

   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
It is evident from the TL adsorption isotherm experiments (Fig. 2) that TLs were surface active; apo-TLs were more surface active than holo-TLs. Analysis of the plots according to Graham and Phillips31 showed kinetic differences in the behavior of different proteins in phases. In the first phase, the induction phase, protein diffuses through the subphase to be adsorbed on the surface.31 This process is controlled by diffusion. The induction phase for lysozyme or casein occurs within 10 to 15 minutes, during which time the surface concentration increases rapidly.31 In the case of apo-TLs, there was a delay in the increase of the film pressure in this period compared with holo-TLs. The delay may have been due to a slower initial rate of diffusion of the apo-TLs. The molecular weight of the major species of TLs is 17,446 Da.32 However, from our data and that of others, holo-TL exists as a noncovalently associated dimer.33 Oligomerization of the apo-TLs could be responsible for the initial slow rate of diffusion of apo-TLs. In support of oligomerization of apo-TLs, gel filtration studies showed that holo-TLs elutes as a dimer and apo-TLs elutes as a heterogeneous mixture of multimeric forms including a dimer and tetramer.

Graham and Phillips31 describe a fast relaxation time, {tau}1, that occurs as the surface concentration increases with a continued rise in surface pressure. Adsorption and probably simultaneous unfolding of protein molecules occur at the interface. During the fast relaxation time, the film pressure of apo- TLs increased more rapidly and to a higher level than during the same period for holo- TLs. It has been demonstrated that the rate of change of surface pressure is a function of protein structure; flexible molecules such as rat serum high-density apolipoprotein or ß-casein cause more rapid decreases in surface tension than do globular proteins under similar conditions.31 34 35 Therefore, the more rapid increase in film pressure with apo-TLs than holo-TLs may be attributed to a more flexible structure. Graham and Phillips31 describe a second, longer relaxation time that is found in some proteins and is ascribed to the rearrangement of molecules in the surface layer. Thus, the appearance of two relaxation times for apo-TLs and not for holo-TLs suggests that apo-TLs may have a more relaxed structure capable of more facile unfolding and surface intermolecular rearrangement. These findings are in keeping with our previously published observations that apo-TLs have 18% less ß structure, an overall reduction in the optical activity, reduced stability against urea unfolding, and less conformational asymmetry involving aromatic residues compared with holo-TLs, all indicative of a less rigid structure in apo-TLs.23 36

At first glance, the time course for the surface adsorption isotherms, which was approximately 1 hour, may appear too slow to be clinically relevant. However, adsorption isotherms for other fatty acid–binding proteins such as albumin occur on a similar time scale. The rate of formation of a monolayer film at the air–water interface above a solution is determined by diffusion over a boundary layer approximately 1 mm thick.37 In the tear film, the distance for diffusion of the apo-TLs was much smaller, approximately 10 µm. In a diffusive process the time (t) is related to the distance (x) by x2 = Dt, where D is the diffusion constant. Thus, the time for establishment of a substantial fraction of a monolayer on the tear film is t ~ 3600 seconds (10/1000)2 = 0.36 seconds, consistent with times of 0.5 to 1 second at which a film is established in vivo. Of course, the in vivo conditions are much more complex than this simplification.

In the current tear film structural model the mucin–aqueous layer was covered by a superficial lipid layer. In this model, TLs, dissolved in aqueous and presumably confined to the mucin–aqueous solutions would have no access to an air–water interface. TLs could exert influence on the surface tension of the tear film only by indirectly affecting the film pressure of the superficial lipid layer. Clinical studies support an interaction of protein at the interface of the lipid layer with the mucin–aqueous layer of the tear film. In patients with dry eye, the administration of drugs that increase TLs and total tear proteins, results in the reduction of surface tension.38 From inspection of these data the increase in TLs was relatively greater than the increase in total proteins. Because other proteins were not individually measured and TLs were measured on the basis of the relative increase in peak area from a chromatogram, it was not possible to determine the contribution of TLs to this affect. Although the precise structure and lipid composition of the tear film remain to be elucidated, the adsorption isotherms demonstrated the influence of TLs on an aqueous surface and provided the first step toward clarifying the interactions of TLs at tear film interfaces.

The possible interaction of TLs at the interface between the cornea and mucin–aqueous layer is evident from contact angle experiments. Young’s equation30 relates the contact angle {theta} of a liquid drop on a solid surface to three surface tensions, that between the liquid and air ({gamma}LA), that between the liquid and the surface ({gamma}LS), and that between the surface and air ({gamma}SA):

The initially decreased contact angles for proteins applied to ferric stearate may therefore be attributed to decreases in {gamma}LA caused by the adsorption of the proteins at the air–liquid interfaces of the droplets or by decreases in the {gamma}LS associated with the changes in the surface, or a combination of both effects. It was clear from the contact angle measurements after aspiration and reapplication of buffer that {gamma}LS for spots with previous contact with buffer alone had not been changed; the surface had retained its hydrophobic character. In contrast, the lower contact angle for buffer applied to the spots that previously contained purified TLs or tears shows that after contact with TLs, the wettability of the surface was enhanced. The effect was greater for solutions with TLs than with lysozyme, lactoferrin, or tears depleted of TLs. TLs in solution bind to lipids that are adsorbed on a surface. This situation is analogous to that occurring when meibomian lipids have inadvertently spilled directly on the corneal surface. Such a contaminated surface would exhibit a low critical surface tension, depending on the types of lipids, and would be unwettable.2 Our data show that after a solution of TLs was incubated with a film of lipid on a solid support, there was an initial reduction of contact angle compared with buffer or solutions relatively depleted of TLs. The further reduction in the contact angle when buffer was replaced on surface from 143° for buffer to 110° for TLs indicate that the surface became more wettable. This phenomenon was the result of TLs interacting with the surface and probably binding and/or removing the lipid. Extrapolated to the clinical situation TLs have the potential to bind lipids contaminating the cornea and could either remove the lipids or replace the hydrophobic surface with a more wettable surface that could allow spreading of the mucin–aqueous layer. Ideally, lipid deposited directly on the cornea should be completely removed.

EPR experiments confirmed the claim that TLs can remove lipids from a solid hydrophobic surface and deliver them into solution. Spin-labeled C-12 did not diffuse readily into a purely aqueous solution; no signal was detected. When TLs were added to the solution, a signal identified within 2 minutes indicated that nitroxide was in the solution. The initial spectra show that a large proportion of the total signal was accounted for by nitroxide bound to TLs in solution. Spectra taken at later times indicated an increasing total concentration of detectable spin label in solution with progressively more unbound spin label. Because the concentration of bound spin label approached one half the total protein concentration at 150 minutes, the apparent dissociation constant for this reaction approached the concentration of free ligand (5.5 µM), assuming the ligand-to-protein-binding ratio was 1:1. This estimation is consistent with that previously calculated.22 Therefore, the delivery of free ligand in solution was the result of simple dissociation of the protein ligand complex. The data obtained from incubation of tear proteins with 16-AP demonstrate that the removal of lipid did not occur by some nonspecific detergent effect of protein. Incubation with neither lysozyme nor lactoferrin resulted in a fluorescent signal from 16-AP. The scavenging of 16-AP was produced only by TLs in tears, and it occurred through a specific binding mechanism.

Taken together, the data indicate that TLs bound to the spin-labeled lipid, removed lipid from the hydrophobic surface and delivered lipid into the solution that was both bound and unbound to protein. This mechanism has important ramifications for the cornea and dry eye disease. Previous investigators have demonstrated that mucin defects and thinning of the tear film may lead to disruption because meibomian lipids may deposit directly on the corneal surface or on the mucin layer converting it to a hydrophobic surface that is prone to desiccation.12 21 However, the mucin–aqueous phase of tears contains TLs that have the potential to bind, cover, and remove these lipids allowing reconstitution of the integrity of the tear film. Because TLs are the only proteins in tears that show a significant bound lipid component, TLs are the key candidates to act as scavengers.22 The lipid-binding role of other components such as mucin has not been elucidated. The promiscuous binding nature of TLs is in keeping with the role as scavengers. In our experiments removal of lipids by TLs began within minutes. Because TLs are found in all normal tear samples,26 the removal of contaminating lipids is probably continuously ongoing in the healthy state. However, in disease in which there is destruction of the lacrimal glands, this protective mechanism may be compromised. It has been previously shown in patients with dry eye that the concentration of TLs correlates with tear film stability.38 Our data suggest that TLs interact with both interfaces of the mucin–aqueous layer to promote lipid solubility and tear film stability.


   Acknowledgements
 
The authors thank Wayne Hubbell for providing the nitroxide derivative of lauric acid and access to the electron paramagnetic resonance instrument.


   Footnotes
 
Supported by U. S. Public Health Service, National Institutes of Health Grants EY11224, EY12080 and EY00331, an unrestricted grant from Research to Prevent Blindness, the Bundy Foundation, and National Science Foundation Grant CHE 94-08584.

Submitted for publication January 22, 1999; revised May 18 and July 12, 1999; accepted July 28, 1999.

Commercial relationships policy: N.

Corresponding author: Ben J. Glasgow, Department of Pathology, Jules Stein Eye Institute, UCLA, School of Medicine, C100 Stein Plaza, Los Angeles, CA, 90095. E-mail: bglasgow{at}mednet.ucla.edu


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 

  1. Holly, FJ, Lemp, MA (1977) Tear physiology and dry eyes Surv Ophthalmol 22,69-87[Medline][Order article via Infotrieve]
  2. Holly, FJ, Lemp, MA (1971) Wettability and wetting of corneal epithelium Exp Eye Res 11,239-250[Medline][Order article via Infotrieve]
  3. Brauninger, GE, Shaw, DO, Kauffman, HE (1972) Direct physical demonstration of oily layer on tear film surface Am J Ophthalmol 73,132-134[Medline][Order article via Infotrieve]
  4. Chen, H-B, Yamabayashi, S, Ou, B, Tanaka, Y, Ohno, S, Tsukahara, S (1997) Structure and composition of rat precorneal tear film Invest Ophthalmol Vis Sci. 38,381-387[Abstract/Free Full Text]
  5. Nichols, BA, Chiappino, ML, Dawson, CR (1985) Demonstration of the mucous layer of the tear film by electron microscopy Invest Ophthalmol Vis Sci 26,464-473[Abstract/Free Full Text]
  6. Prydal, JI, Artal, P, Woon, H, Campbell, FW (1992) Study of human precorneal tear film thickness and structure using laser interferometry Invest Ophthalmol Vis Sci 33,2006-2011[Abstract/Free Full Text]
  7. McCulley, JP, Shine, WE (1998) Meibomian secretions in chronic blepharitis Adv Exp Med Biol 438,319-326[Medline][Order article via Infotrieve]
  8. McCulley, JP, Shine, WE (1997) A compositional based model for tear film lipid layer Trans Am Ophthalmol Soc 95,79-88[Medline][Order article via Infotrieve]
  9. Greiner, JV, Glonek, T, Korb, DR, Leahy, CD (1996) Meibomian gland phospholipids Curr Eye Res 15,371-375[Medline][Order article via Infotrieve]
  10. McDonald, JE (1969) Surface phenomena of the tear film Am J Ophthalmol 67,56-64[Medline][Order article via Infotrieve]
  11. Korb, DR, Greiner, JV, Glonek, T, et al (1998) Human and rabbit lipid layer and interference pattern observations Adv Exp Med Biol 438,305-308[Medline][Order article via Infotrieve]
  12. Holly, FJ (1974) Surface chemistry of tear film component analogs J Colloid Interface Sci 49,221-231
  13. Cope, C, Dilly, PN, Kaura, R, Tiffany, JM (1986) Wettability of the corneal surface: a reappraisal Curr Eye Res 5,777-785[Medline][Order article via Infotrieve]
  14. Tiffany, JM (1990) Measurement of wettability of the corneal epithelium, I: particle attachment method Acta Ophthalmol 68,175-181
  15. Tiffany, JM (1990) Measurement of wettability of the corneal epithelium, II: contact angle method Acta Ophthalmol 68,182-187
  16. Tiffany, JM (1994) Composition and biophysical properties of the tear film: knowledge and uncertainty Sullivan, DA eds. Lacrimal Gland, Tear Film, and Dry Eye Syndromes ,231-238 Plenum Press New York.
  17. Sharma, A. (1993) Energetics of corneal epithelial cell-ocular mucus-tear film interactions: some surface-chemical pathways of corneal defense Biophys Chem 47,87-99[Medline][Order article via Infotrieve]
  18. Doane, MG (1994) Abnormalities of the structure of the superficial lipid layer on the in vivo dry-eye tear film Adv Exp Med Biol 350,293-298[Medline][Order article via Infotrieve]
  19. Golding, TR, Bruce, AS, Mainstone, JS (1997) Relationship between tear-meniscus parameter and tear-film breakup Cornea 16,649-691[Medline][Order article via Infotrieve]
  20. Tseng, SCG, Zhang, SH (1995) Interaction between rose bengal and different protein components Cornea 14,427-435[Medline][Order article via Infotrieve]
  21. Holly, FJ (1973) Formation and rupture of the tear film Exp Eye Res 15,515-525[Medline][Order article via Infotrieve]
  22. Glasgow, BJ, Abduragimov, AR, Farahbakhsh, Z, Faull, KF, Hubbell, WL (1995) Tear lipocalins bind a broad array of lipid ligands Curr Eye Res 14,363-372[Medline][Order article via Infotrieve]
  23. Gasymov, OK, Abduragimov, AR, Yusifov, TN, Glasgow, BJ (1998) Structural changes in tear lipocalins with ligand binding Biochim Biophys Acta 1386,145-156[Medline][Order article via Infotrieve]
  24. Gasymov, OK, Abduragimov, AR, Yusifov, TN, Glasgow, BJ (1999) Binding studies of tear lipocalin: the role of the conserved tryptophan in maintaining structure stability and ligand affinity Biochim Biophys Acta 1433,307-320[Medline][Order article via Infotrieve]
  25. Nicolaides, N, Kaitaranta, JK, Rawdah, TN, Macy, JT, Boswell, FM, III, Smith, RE (1981) Meibomian gland studies: comparison of steer and human lipids Invest Ophthalmol Vis Sci 20,522-536[Abstract/Free Full Text]
  26. Glasgow, BJ (1995) Tissue expression of lipocalins in human lacrimal and von Ebner’s glands: colocalization with lysozyme Graefes Arch Clin Exp Ophthalmol 233,513-522[Medline][Order article via Infotrieve]
  27. Bozimowski, D, Artiss, JD, Zak, B. (1985) The variable reagent blank: protein determination as a model J Clin Chem Clin Biochem 23,683-689[Medline][Order article via Infotrieve]
  28. Selsted, ME, Martinez, RJ (1978) Lyozyme: primary bactericidin in human plasma serum active against Bacillus subtilis Infect Immun 20,782-791[Abstract/Free Full Text]
  29. Querinjean, P, Masson, PL, Heremans, JF (1971) Molecular weight single-chain structure and amino acid composition of human lactoferrin Eur J Biochem 20,420-425[Medline][Order article via Infotrieve]
  30. Adamson, AW (1982) Physical Chemistry of Surfaces 4th ed. ,338 Wiley New York.
  31. Graham, DE, Phillips, MC (1979) Proteins at liquid interfaces J Colloid Interface Sci 70,403-414
  32. Glasgow, BJ, Abduragimov, AR, Yusifov, TN, et al (1998) A conserved disulfide motif in human tear lipocalins influences ligand binding Biochemistry 37,2215-2225[Medline][Order article via Infotrieve]
  33. Redl, B, Holzfeind, P, Lottspeich, F. (1992) cDNA cloning and sequencing reveals human tear prealbumin to be a member of the lipophilic-ligand carrier protein superfamily J Biol Chem 267,20282-20287[Abstract/Free Full Text]
  34. Mitchell, J, Irons, L, Palmer, GJ (1970) A study of the spread of and adsorbed films of milk proteins Biochim Biophys Acta 200,138-150[Medline][Order article via Infotrieve]
  35. Camejo, G, Colacicco, G, Rapport, MM (1968) Lipid monolayers: interactions with the apoprotein of high density plasma lipoprotein J Lipid Res 9,562-569[Abstract]
  36. Glasgow, BJ, Abduragimov, AR, Yusifov, TN, Gasymov, OK (1998) Studies of ligand binding and CD analysis with apo- and holo- tear lipocalins Adv Exp Med Biol 483,105-112
  37. Ter Minassian–Saraga, L. (1955) Étude de l’adsorption et de la désorption aux surfaces liquides J Chim Phys. 52,181-192
  38. Schoenwald, RD, Vidvauns, S, Wurster, DE, Barfknecht, CF (1997) Tear film stability of protein extracts from dry eye patients administered a sigma agonist J Ocul Pharmacol Ther 13,151-161[Medline][Order article via Infotrieve]



This article has been cited by other articles:


Home page
 IOVSHome page
T. J. Millar, P. Mudgil, I. A. Butovich, and C. K. Palaniappan
Adsorption of Human Tear Lipocalin to Human Meibomian Lipid Films
Invest. Ophthalmol. Vis. Sci., January 1, 2009; 50(1): 140 - 151.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Ophthalmol.Home page
P S Tsai, J E Evans, K M Green, R M Sullivan, D A Schaumberg, S M Richards, M R Dana, and D A Sullivan
Proteomic analysis of human meibomian gland secretions.
Br. J. Ophthalmol., March 1, 2006; 90(3): 372 - 377.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
O. K. Gasymov, A. R. Abduragimov, P. Prasher, T. N. Yusifov, and B. J. Glasgow
Tear Lipocalin: Evidence for a Scavenging Function to Remove Lipids from the Human Corneal Surface
Invest. Ophthalmol. Vis. Sci., October 1, 2005; 46(10): 3589 - 3596.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
H. Saaren-Seppala, M. Jauhiainen, T. M. T. Tervo, B. Redl, P. K. J. Kinnunen, and J. M. Holopainen
Interaction of Purified Tear Lipocalin with Lipid Membranes
Invest. Ophthalmol. Vis. Sci., October 1, 2005; 46(10): 3649 - 3656.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Ophthalmol.Home page
M Yamada, H Mochizuki, M Kawai, K Tsubota, and T J Bryce
Decreased tear lipocalin concentration in patients with meibomian gland dysfunction
Br. J. Ophthalmol., July 1, 2005; 89(7): 803 - 805.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
K. Peters, G. R. Dennis, P. J. Anderton, and T. J. Millar
The Effects of Novel Amphipathic Block Copolymers on Stabilization of the Rat Tear Film
Invest. Ophthalmol. Vis. Sci., December 1, 2003; 44(12): 5089 - 5094.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
O. K. Gasymov, A. R. Abduragimov, T. N. Yusifov, and B. J. Glasgow
Relaxation of {beta}-Structure in Tear Lipocalin and Enhancement of Retinoid Binding
Invest. Ophthalmol. Vis. Sci., October 1, 2002; 43(10): 3165 - 3173.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Glasgow, B. J.
Right arrow Articles by Knobler, C. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Glasgow, B. J.
Right arrow Articles by Knobler, C. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS