Divalent Cations in Tears, and Their Influence on Tear Film Stability in Humans and Rabbits

Abstract

Purpose.: Reduced tear film stability is reported to contribute to dry eye. Rabbits are known to have a more stable tear film than humans. Thus, we sought to examine the tears of rabbits and humans for metal cations, and to test how they influence tear film stability.

Methods.: Tears were collected from 10 healthy humans and 6 rabbits. Tear osmolality was measured by vapor pressure osmometer, and metals analyzed using inductively coupled plasma (ICP) mass spectrometry or ICP atomic emission spectroscopy. The influence of divalent cations on tears was analyzed by measuring surface tension using the Langmuir trough in vitro, using different concentrations of cations in the subphase, and grading the tear break-up in rabbits in vivo after instillation of chelating agents.

Results.: Rabbit tears had a higher osmolality compared to humans. Major metals did not differ between species; however, rabbits had higher levels of Mg²⁺ (1.13 vs. 0.39 mM) and Ca²⁺ (0.75 vs. 0.36 mM). In rabbit tears in vitro, diminishing divalent cations resulted in a decrease in the maximum surface pressure from 37 to 30 mN/m. In vivo, an increase in the amount of tear film that was broken-up was found. In contrast, when changing divalent cation concentrations in human tears, the maximum surface pressure remained at 26 mN/m.

Conclusions.: The normal osmolality of rabbit tears is significantly higher than that in humans. While divalent cations had little influence on human tears, they appear to have an important role in maintaining tear film stability in rabbits.

Introduction

Dry eye syndrome affects approximately 15% of the community. ¹ It seems to be associated strongly with decreased stability of the tear film ² and increased hyperosmolality. ³ The causes of tear film instability currently are unknown. Based on interblink intervals as a guide to tear film stability, ^4,5 humans have a much less stable tear film (interblink period ∼5 seconds) than other mammals, such as the rabbit (interblink period ∼10 minutes). ⁶ Comparative investigations of animal tears with human tears could lead to a better understanding of what leads to a more stable tear film. 

Experimentally, there have been attempts to correlate tear film stability with levels of different tear proteins and lipids or their interactions, and to relate tear stability with physicochemical properties, such as viscosity and surface tension. Gouveia and Tiffany reported that tear lipocalin interacts with other proteins and tear lipids to result in a significantly changed viscosity of human tears from non-Newtonian to Newtonian. ⁷ Differences in the protein profiles between human and rabbit tears have been reported, such as an abundance of lysozyme, lactoferrin, and lipocalin in the former, ⁸ and an abundance of lipophilins in the latter. ⁹ These differences could lead to different physicochemical properties of the tears and, hence, differences in tear film stability. Nagyova et al. found that the surface tension of human tears is caused by an interaction between tear lipocalin and polar lipids. ¹⁰ This finding has been investigated further by Tragoulias et al. using the Langmuir trough method. ¹¹ They showed that the pressure-area (Π-A) profiles of tears resembled closely the Π-A profiles of individual tear proteins rather than the meibomian lipids or mucins, particularly in terms of the hysteresis behavior. 

Hyperosmolality of tears is another feature that has been investigated and is one of the strongest predictors of dry eye syndrome. ³ Mudgil and Millar hypothesized that the link between hyperosmolality of tears and dry eye might be due to an increase in surface tension (or a decrease in surface pressure) as a result of a reorganization of the lipid layer. ¹² This reorganization may be influenced by the levels of divalent cations (Ca²⁺, Mg²⁺) in the aqueous layer, which in greater amounts, theoretically, would increase cross-linking of anionic surfactant lipids at the lipid layer and the aqueous layer interface to increase surface pressure. The anionic lipids are a class of lipids with negative charges. In tears, phospholipids are one of the major classes of anionic lipids that are found in rabbit and human tears. ^13,14 They possess their anionic charge due to their phosphate head groups. The divalent cations may interact with these phospholipids, helping to stabilize the lipid and inducing an increase of surface pressure. ¹⁵ However, in vitro experiments show that even tripling the ion concentrations generally, or divalent cations specifically, had little effect on the surface pressure of human meibomian lipid films. ¹² However, hyperosmolality in tears also may affect the proteins in the aqueous layer. An increase in osmolality leading to an increase in divalent cations could alter folding of proteins, ^16,17 and interactions between proteins that then could affect the surface tension and the stability of the tear film. 

The reason for the significant difference in tear film stability between humans and rabbits is not well understood. In the literature, the tear film break-up of rabbits still remains controversial. Studies using fluorescein have reported a tear break-up time ranging between 15 and 25 seconds. ^18–20 Another study, using non-invasive interferometry methods, however, reported no sign of break after 3 minutes. ²¹ Human tear break-up time has been reported to vary between 8 and 48 seconds. ^22–25 Identifying the differences, however, may be key to improving tear film stability in humans. In this study, we used human and rabbit tears to test the influence of divalent cation concentrations on surface tension in vitro using a Langmuir trough, and the influence of divalent cations in vivo by measuring non-invasive tear break-up time (NITBUT). NITBUT was used as an indicator of surface tension, as a negative correlation has been found previously between surface tension and NITBUT in normal and dry eye patients. ^26,27 In our study the normal ionic components of rabbit and human tears also were analyzed and compared. 

Materials and Methods

Chemicals

Water purified by ion exchange (ω = 18.2 MΩ; Millipore, Billerica, MA) was used. Ethylenediaminetetraacetate disodium salt (Na₂-EDTA) and ethylenediaminetetraacetate dipotassium magnesium salt (K₂Mg-EDTA) were obtained from Sigma-Aldrich, St. Louis, MO. All other chemicals used were purchased from AMRESCO Chemicals Inc., Solon, OH. 

Collection of Tear Samples

Human and animal ethics were approved by the Ethics Review Panel of the University of New South Wales and Brien Holden Vision Institute, respectively. The tenets of the Declaration of Helsinki were adhered to for human studies, and informed consent was obtained before enrolling human subjects into the study. Five male and five female human subjects between 25 and 38 years old were recruited for tear collection. All subjects were healthy non-contact lens wearers and were not taking medications. Subjects with ocular surface disease, current or past dry eye, or previous ocular surgery were excluded from the study. 

All animal experimental procedures conformed to the Association for Research and Vision in Ophthalmology (ARVO) statement for the Use of Animals in Ophthalmic and Vision Research. Six female New Zealand White rabbits weighing 2.8 to 6.4 kg were selected for this study. 

The collection time was up to 10 minutes for human subjects and 30 minutes for rabbits. Human participants were allowed to blink during the procedure, with the capillary being withdrawn each time. The collection was stopped if any reflex tearing was detected and resumed after 20 minutes. Basal tear samples from humans and rabbits (5–10 μL) were collected by placing a calibrated and fire polished glass 10 μL microcapillary tube with 0.5 mm internal diameter (Brand, Wertheim, Germany) into the lower tear meniscus, minimizing contact of the tip with the ocular surface to avoid inducing reflex tears. No sedation or anesthesia was used. Cells and any other particulate matter that might have been collected were removed from the samples by centrifugation (5 minutes at 5000 g). The supernatants were pooled at an equal contribution by volume from all subjects and stored at −80°C for all tests except for osmolality, for which individual samples were collected (5–10 μL), stored on ice, and then analyzed within one hour of collection. 

Tear Osmolality Measurement

Osmolality of human and animal tears was measured with a vapor pressure osmometer (Wescor, Logan, UT). An initial calibration of the osmometer was performed by applying 10 μL standards (290, 1000, and 100 mmol/kg; Wescor) to 3.2 mm diameter filter paper (Whatman, Maidstone, UK) on the sample plate. Since the sample size to be used was 2 μL, a further calibration was made by taking the average of three consecutive readings of 290 mmol/kg using a sample volume of 2 μL. Tear samples (2 μL) then were measured. The osmolality results were analyzed statistically. Bonferroni corrected Student's t-tests were used to determine the confidence limits for population mean with acceptance set at P = 0.02. 

Detection of the Major Metal Cations in Tears by Spectrometry

Quantitative analysis of tear metal cations was done by inductively coupled plasma mass spectrometer (ICP-MS) Perkin Elmer Elan DRC II (Waltham, MA) and inductively coupled plasma atomic emission spectrometer (ICP-AES) Varian Vista Pro (Palo Alto, CA) at the National Measurement Institute, Sydney, Australia. Tear samples (25 μL) were diluted 100-fold in 2% nitric acid solution and loaded for each assay. To test the repeatability of the technique, only human tears were used. The same batch of a human tear sample was tested on three separate occasions, and the mean value ± SD of the major metal cations was calculated, and coefficients of repeatability (1.96 × SD of mean difference between days) calculated. ²⁸ Since the daily variance or repeat testing was within acceptable limits and a relatively small volume of pooled rabbit tears was available, rabbit tears were tested only once.  

The Influence of Divalent Cations on Tear Film Stability—In Vitro Surface Pressure Experiment

The surface pressure of rabbit and human tears was measured on a double-barrier Langmuir trough (Nima Technology, Coventry, UK) with a maximum surface area of 80 cm². The trough was filled (∼80 mL) with an artificial tear (AT) buffer (Table 1). ¹² The effect of divalent cations was tested by varying their concentrations in the AT buffer (Table 1). A microsyringe (Hamilton Co., Bonaduz, Switzerland) was used to apply tear samples (10 μL) onto the surface of the AT buffer between the barriers, which were in the open position. Surface pressure was monitored using a Wilhelmy plate (Chr 1 filter paper; Whatman, Maidstone, UK). The surface was compressed continuously, and expanded between surface areas of 80 and 15 cm² at a rate of 15 cm²/minutes, and Π-A isocycles of tears were obtained with use of the software provided by Nima (Coventry, UK).  

Table 1.  

 

AT Subphase Buffers (pH 7.4) Used for Langmuir Trough

	Treatment of Tears	AT Subphase Buffer (mM)
		NaCl	KCl	NaHCO3	Na2HPO4	MOPS	MgCl2	CaCl2
Human	Normal	113.51	23.03	16.38	0.83	19.97	0.39	0.36
	High divalent cations	113.51	23.03	16.38	0.83	19.97	1.13	0.75
Rabbit	Normal	113.51	23.03	16.38	0.83	19.97	1.13	0.75
	With Na2-EDTA	113.51	23.03	16.38	0.83	19.97	0	0

All experiments were performed at physiologic (eye) temperature (35°C). ²⁹ Temperature of the AT subphase buffer in the Langmuir trough was maintained by heat exchange through a water jacket and regulated by a circulatory thermostat controlled bath (Ecoline RE106; Lauda, Lauda-Königshofen, Germany). The temperature of the AT subphase buffer was measured using a temperature probe (MadgeTech Inc., Warner, NH) immersed into the AT subphase buffer outside the barriers and the temperature was maintained within ± 1°C. 

The Influence of Divalent Cations on Tear Film Stability—In Vivo Experiment

Six rabbits were used and ocular health was determined by observation using a slit-lamp biomicroscope (Nikon FS-3 Zoom Photo Slit Lamp). Solutions of a chelating agent (Na₂-EDTA) and its saturated form (K₂Mg-EDTA) as a control were prepared in ion exchange purified water. Samples (10 μL) of the Na₂-EDTA (0.95 mM), K₂Mg-EDTA (0.95 mM), and MilliQ water were instilled separately into the inferior palpebral fold using an Eppendorf pipette by a masked investigator on different days allowing at least 24 hours between tests. The regimen meant that each eye of each rabbit received each of the drops twice in random order. On each occasion the rabbits were anesthetized with Midazolam 2 mg/kg (Roche, Dee Why, Australia) before drop instillation. Tear film was monitored using a slit-lamp biomicroscope immediately following drop instillation and graded after five minutes using an in-house grading scale (Table 2). 

Table 2.  

 

A Scale for Evaluating Extent of Tear Break-Up of Rabbit Tear Film 5 Minutes after Drop Instillation

	Grade
	0	1	2	3
Category	None	Mild	Moderate	Severe
Description	No tear break, smooth surface	<5% of tear film is broken	6–15% of the tear film is broken	>15% of the tear film is broken
Example photos

Results

Osmolality of Rabbit and Human Tears

The osmolality of collected tears was tested. Rabbit tears had a significantly (P < 0.02) higher osmolality (375.83 ± 18.17 mmol/kg) compared to that of human tears (282.50 ± 24.32 mmol/kg). 

Major Metal Cations in Tears

The human tear sample used for validating the reproducibility of the technique gave means ± SDs of Na⁺, K⁺, Mg²⁺, and Ca²⁺ of 104.5 ± 4.27, 18.1 ± 1.04, 0.49 ± 0.02, and 0.33 ± 0.03 mM, respectively, and coefficients of repeatability of 7.35, 2.04, 0.05, and 0.06. Having established the validity of the method, fresh samples of rabbit and human tears were analyzed within the same experiment to minimize technique variability. Pooled human and pooled rabbit tears had similar concentrations of Na⁺ and K⁺, but rabbit tears contained twice the concentration of Ca²⁺, and three times the concentration of Mg²⁺ (Table 3). Importantly, in terms of consistency, the fresh sample of tears gave values very similar to the stored sample used for test validation. 

Table 3.  

 

The Concentration of Na⁺, K⁺, Ca²⁺, and Mg²⁺ in Rabbit and Human Tears

	Major Metal Cations (mM)
	Na+	K+	Mg2+	Ca2+
Human tears	100.00	21.00	0.39	0.36
Rabbit tears	103.48	21.09	1.13	0.75

The Influence of Divalent Cations on Tear Film Stability—In Vitro Surface Pressure Experiment

Human and rabbit tears exhibited substantial surface activity with maximum surface pressures above 25 mN/m. The human tear Π-A profile was more complex than that for the rabbit with inflexions at about 55 and 30 cm², and greater hysteresis, but less maximum surface pressure (∼26 vs. ∼37 mN/m). The Π-A profile of normal rabbit tears, and the effect of the addition of Na₂-EDTA (i.e., diminishing the concentration of divalent cations) is shown in FigureA. Adding Na₂-EDTA to rabbit tears decreased the maximum surface pressure from 37 to 30 mN/m at 35°C (Fig.A). Neither increasing the concentration of divalent cations in human tear samples to the level in normal rabbit tears nor diminishing (addition of Na₂-EDTA) the concentration of divalent cations affected the Π-A isocycles of human tears. The isocycles overlapped and the maximum surface pressure remained at 26 mN/m (Figs.B, C). 

Figure. 

 

Π-A isocycles of tear samples on Langmuir trough at 35°C. (A) Rabbit tears with (solid line) and without (dotted line) mix of Na₂-EDTA. (B) Human tears with (solid line) and without (dotted line) increase concentration of divalent cations. (C) Human tears with (solid line) and without (dotted line) mix of Na₂-EDTA.

The Influence of Divalent Cations on Tear Film Stability—In Vivo Experiment

To investigate further the role of divalent cations on tear film stability, the effect of EDTA solutions on the stability of the rabbit tear film was studied. MilliQ water was compared to a chelating agent Na₂-EDTA and a saturated form K₂Mg-EDTA. 

After MilliQ water instillation, no change was observed in the tear film for the observation period (5 minutes). In contrast, application of Na₂-EDTA resulted in an increase in tear film break-up, which occurred immediately, and still was visible after five minutes (Table 4). After instillation of the saturated chelator K₂Mg-EDTA, there was no effect on tear film break-up with the grading remaining in the range 0 to 1 (Table 4). 

Table 4.  

 

The Extent of Rabbit Tear Break-Up after Instillation of MilliQ, Na₂-EDTA, and K₂Mg-EDTA according to the Evaluating Scale in Table 2.

Rabbit	MilliQ	Na2-EDTA	K2Mg-EDTA
A	0	3	1
B	0	3	1
C	0	2	0
D	0	3	1
E	0	2	0
F	0	2	1

 Each drop was performed on two separate occasions.

Discussion

In our study, the osmolality and ion concentrations measured in the human tear samples are consistent with those reported previously. In normal human tears, a range of 272 to 317 mmol/kg has been reported. ^30–37 To our knowledge, the osmolality of rabbit tears has not been reported previously and, surprisingly, was much higher than for human tears. The proportion of divalent cations also was markedly greater when taking into account the higher osmolality. In addition, ion concentration analysis showed that the large difference in osmolality between human and rabbit tears must be accounted for by something other than ions. Preliminary data indicate that proteins are the most likely candidates. ³⁸ The evaporation rate for rabbit tears has been reported to range from 7.8 to 41.6 × 10⁻⁷gcm⁻²sec⁻¹ and for humans to range from 4.1 to 32.4 × 10⁻⁷gcm⁻²sec⁻¹. ³⁹ Thus, with the longer interblink time of the rabbit it is possible that there is more evaporation of tears between blinks, which may be one reason for the increased osmolality of rabbit tears. These findings all are important considerations when using the rabbit as a model for testing eye drop formulations, particularly when the effects of osmolality are being investigated. One such example is when the effect of hypotonic electrolyte-balanced lubricant eye drop used to reduce tear film osmolality in dry eye patients is investigated in the rabbit model. ^40–42 When the results from these experiments are translated to support their effect on humans, the higher osmolality in rabbit tears must be taken into account. 

Given that rabbits have a more stable tear film than humans, ^6,43 the finding that rabbit tears are hyperosmolar compared to human tears is enigmatic given that hyperosmolality has been associated with dry eye and decreased tear break-up time. Our results show that in rabbit tears, the high concentration of divalent cations helps maintain a low surface tension (high surface pressure) while in human tears, changing this concentration has little influence. Divalent cations have been found to lead to the formation of complexes with proteins and alter their physical properties. ¹⁶ This indicates that the different effect of divalent cations on rabbit tears compared to human tears could be due to different proteins in tears between species leading to different complexes, which then affect the physical properties of tears. It is, perhaps, less likely that the divalent cation effects are due to differences in mucins between rabbits and humans, as rabbit mucin tends to have less negatively charged sialic acid associated with the mucin glycans than human mucin. ⁴⁴ Differences in the predominant negatively charged lipids, phospholipids, in rabbit and human tears also could have an effect on the physical properties of their tears. However, further experiments must be done to test this speculation. 

In human tears, the interaction of proteins with the divalent cations does not impact on the tear film in terms of physical properties, possibly due to the greater tolerance of the human tear system. This finding indicates that for human tears, the dry eye hyperosmolality association reported by Liu et al. ⁴⁵ may not be based on the change of physical properties of the tear film due to the interaction of proteins with divalent cations, but rather only on the sensation. 

In summary, rabbit tears have higher osmolality as well as higher levels of divalent cations compared to those of human tears. In humans, increasing the concentration of divalent cations does not impact on tear film stability. Unlike humans, in vitro and in vivo experiments have shown that higher divalent cation concentrations in rabbit tears contribute to their more stable tear film. This indicates that the high divalent cations in rabbit tears might be the key to explaining their very stable tear film. An understanding of the influence of divalent cations on rabbit tear film stability might allow for the development of new methods to prevent or treat dry eye in humans. 

Acknowledgments

Denise Lawler assisted with the animal works of this study, and Nerida Cole and Judith Flanagan critically reviewed the paper. 

Footnotes

 Supported by the Australian Federal Government through the Australian Postgraduate Award, and by a scholarship from the Brien Holden Vision Institute. Part of the work reported in this study was conducted while the first author was a recipient of the Cornea and Contact Lens Society of Australia research award.

Footnotes

 Disclosure: X.E. Wei, None; M. Markoulli, None; T.J. Millar, None; M.D.P. Willcox, None; Z. Zhao, None

Footnotes

 PMC Deposit Required: No