You are here

Article Text

Article menu
Clinical science
Original article
Association of tear proteins with Meibomian gland disease and dry eye symptoms
  1. Louis Tong1,2,3,
  2. Lei Zhou1,4,
  3. Roger W Beuerman1,2,4,
  4. Shao Zhen Zhao5,
  5. Xiao Rong Li5
  1. 1Singapore Eye Research Institute, Singapore, Singapore
  2. 2Singapore National Eye Centre, Singapore, Singapore
  3. 3Duke-NUS Graduate Medical School, Singapore, Singapore
  4. 4Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
  5. 5Tianjin Medical University Eye Center, Tianjin, China
  1. Correspondence to Professor Roger W Beuerman, Singapore Eye Research Institute, 11 Third Hospital Avenue, Singapore 168751; rwbeuer{at}mac.com

https://doi.org/10.1136/bjo.2010.185256

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Dry eye syndrome is defined as ocular surface dryness and discomfort associated with an abnormal tear film from either tear deficiency or excessive tear evaporation.1 Significant ocular surface damage and visual disability are also common, and it is a healthcare problem that affects millions of people with a prevalence estimated to be as high as 11–22% of the general population.2 3

    Although dry eye is a multi-factorial disease with contribution from lifestyle and occupational habits,1 studies have suggested that other ocular factors such as Meibomian gland disease (MGD) and eyelid inflammation may contribute to dry eye.4 Abnormalities of tear lipid properties in MGD may contribute to tear instability and excessive tear evaporation, and hence evaporative dry eye.1 MGD itself is an inflammatory condition,5 often associated with chronic inflammation associated with aberrant fibrosis and scarring around the Meibomian glands and the lid margin. Currently, it is not clear if the molecular mediators of inflammation related to MGD6 are fundamentally different from the mediators of T cell inflammation present in other types of dry eye, such as autoimmune dry eye and Sjogren syndrome. It is also not known if the pro-inflammatory molecules secreted by the Meibomian gland can spill over to the tear film to cause ocular surface dysfunction or irritation. If mediators of inflammation in MGD are different from those in other types of dry eye, they will increase our understanding of the contribution of MGD to dry eye. In addition, if these molecules also mediate ocular surface damage or loss of visual function, they would be potential targets of therapy in this subset of dry eye patients, and would increase the value of MGD assessment in dry eye patients.

    In dry eye the results of standard clinical tests do not correlate well with the level and type of patient complaints,7 8 suggesting that different components or subtypes of dry eye may be associated with different symptomatology. Most types of symptoms in dry eye are either irritative in nature (eg, pain), or related to reversibility of vision (eg, transient blurring).

    Tear fluid is a complex mixture of proteins, lipids, salts, mucin and other small organic molecules.1 The secretory acinar cells of the lacrimal gland are the major source of tear proteins and fluid, and the Meibomian glands are the major source of tear lipids.9 There is little doubt that Meibomian lipids were altered in MGD.10 Since the tear lipid layer is adjacent to the aqueous protein-containing layer of the tearfilm, MGD may at least in theory, also derange the protein content of the tears directly or indirectly. Analysis of peptides from Meibomian gland secretions of volunteers has shown the presence of peptides that may be related to inflammation, such as interferon regulatory factor 3 and tyrosine kinase C,6 but this type of investigation has not been done in tears of MGD patients. Our group has previously found that changes of tear protein profiles were correlated with conditions such as ocular surface wounding and ocular surface diseases.11–13 Our current hypothesis is that the ocular surface micro-environment, including specific derangement of tear inflammatory proteins, may be associated with the severity of MGD. A second hypothesis is that since unique disturbances in tear proteins are found in dry eye patients with MGD, the level of specific proteins may be correlated to specific symptoms of dry eye.

    In this study, we use a quantitative proteomics method, that is isobaric tagging for relative and absolute quantification (iTRAQ) technology with two dimensional nanoliquid chromatography (LC)-nano-electrospray ionisation (ESI)-mass spectrometry (MS)/MS, to examine for significant correlations between MGD severity and dry eye symptoms.

    Experimental procedures

    Chemicals

    Formic acid, trifluoroacetic acid, ammonium bicarbonate, ammonium acetate and phosphate-buffered saline (PBS) were purchased from Sigma (St Louis, Missouri, USA). High performance liquid chromatography (HPLC) grade acetonitrile, water, acetic acid and methanol were from Fisher Scientific (Pittsburgh, Pennsylvania, USA).

    Patients

    Clinical examinations included subjective symptoms, Schirmer's test (without anaesthesia), tear break-up time (TBUT), and other general ophthalmic examinations such as visual acuity and lid margin and Meibomian gland status. Dry eye was defined as having dry eye symptoms with at least one abnormal finding out of three objective clinical tests (Schirmer's test ≤10 mm at 5 min without anaesthetic, tear break-up time ≤10 s, or corneal fluorescein staining higher than 2 using Oxford scheme).14 MGD severity was graded into 0, 1, 2 and 3. MGD was defined by the presence of lid margin telangiectasia (grade 1) or Meibomian gland orifice plugging (grade 2) or both (grade 3) on slit lamp biomicroscopic examination. A grade of 0 was given when these signs were absent. Dry eye symptoms and some irritative ocular surface symptoms were evaluated (table 1) and graded into 0 (nil), 1 (occasional), 2 (some of the time), 3 (often) or 4 (always). None of the patients had Sjogren's syndrome. Patients had been receiving artificial tear treatment but not anti-inflammatory treatment.

    View this table:
    Table 1

    Grading of symptoms in the study

    Tear collection and tear protein elution

    Tear fluids for all patients were collected using Schirmer's strips. After collection, Schirmer's strips were immediately frozen at –80°C until analysis. The first 10 mm of the Schirmer's strip was cut into small pieces and soaked in 150 μl of phosphate-buffered saline (PBS) for 3 h to elute tear proteins. Total tear protein concentration of each sample was measured using a Micro BCA Protein Assay Kit (Pierce Biotechnology, Inc, Rockford, Illinois, USA).

    Study design and iTRAQ sample preparation

    The details of iTRAQ technology and the use of pooled control samples have been described in a previous report14 (refer to supplementary file 1 for more details of the experiment). Eighteen control tear samples were pooled to form a global control and labelled with iTRAQ reagent 114. iTRAQ 115, 116, and 117 were used to label three individual dry eye samples (total eight sets of iTRAQ experiments). Relative quantification of proteins using iTRAQ technology is based on the ratio of peak areas of m/z 114, 115, 116 and 117 from MS/MS spectra. The iTRAQ 114 referred to a pooled sample of normal controls. The iTRAQ 115, 116 and 117 were individual tear samples from different patients with dry eye. The levels of α-enolase, α-1-acid glycoprotein 1, S100A8 (calgranulin A), S100A9 (calgranulin B), S100A4 and S100A11 (calgizzarin), prolactin-inducible protein (PIP), lipocalin-1, lactoferrin and lysozyme were hence expressed as ratios.

    Statistical analysis

    The software SPSS for Windows was used to evaluate correlation coefficients between continuous variables. Tear was only collected from one eye, so the MGD severity of the same eye was used in correlation analysis. For the dry eye symptoms, the eye with the higher grading of MGD was used in the analysis. The strength and direction of the correlation was determined by the absolute value of the Pearson correlation coefficient r and the sign of the r respectively. The MGD severity was an ordinal and not scale variable, so when we used MGD severity as an ordinal independent variable, we also computed η in addition to r, which took into consideration possible non-linearity in the ordinal independent variable. Where there is absence of non-linearity, η will be equal to r. When both the independent and dependent variables were ordinal, for example, the MGD level and the redness experienced, Kendall's τ was calculated. Differences in the mean of continuous variables was evaluated by analysis of variance (ANOVA).

    Results

    Patient demographics

    A total of 24 patients were included; the mean age of patients was 35. 5±15.4 years. There was a slight female preponderance, with 15 women and nine men. Nine (37.5%), eight (33.3%), three (12.5%) and two (8.3%) of patients had MGD grading of 0, 1, 2 and 3, respectively (in the worse eye). Only two of the participants had unequal MGD grading in the right and left eyes, with a difference in grading of 1 step in each of these patients. Two patients did not have clinical data on MGD grading.

    The median Schirmer's test value was 5.0 (range 1.0–17.0) mm and median tear break-up time was 3.5 (range 1.0–10.0) s. Eighteen patients (75%) showed corneal fluorescein dye staining in the eye with tear collection.

    The mean and median scores for each of the symptoms evaluated in this study are shown in table 1. There was no significant difference (all p>0.05) in the symptom grading between those without MGD (grade 0) and those with MGD (grades 1–3) (data not shown). However, on ANOVA, there is a trend for transient blurring of vision to occur more frequently in higher MGD grades. Statistically, the trend was more pronounced comparing those with level 1 versus level 3 MGD, but this may not be significant with Bonferroni correction (p=0.046). Significant correlations were found between grittiness grades with MGD level (r=0.45, p=0.036 figure 1A), as well as between redness grades with MGD level (r=0.45, p=0.037, figure 1B); these, however, were found to be not statistically significant on non-parametric correlation (τ=0.303, p=0.116 and τ=0.286, p=0.128, respectively).

    Figure 1
    Figure 1

    Scatter diagrams showing the association between Meibomian gland severity and symptoms (A and B) and between Meibomian gland severity and tear protein levels (C and D). MGD, Meibomian gland disease.

    Association of tear proteins with MGD

    Table 2 shows the correlation coefficients of tear proteins with MGD level. In particular S100A8 (η=0.615) and S100A9 (η=0.590) were significantly correlated to MGD level (figure 1C,D respectively). Table 3 shows the mean protein ratios for S100A8 and S100A9 in each level of MGD severity. There was a trend for increasing S100A8 and S100A9 ratios with increasing levels of MGD severity (p=0.033 and 0.048, respectively).

    View this table:
    Table 2

    Correlation analysis between different tear protein levels and MGD levels

    View this table:
    Table 3

    Mean S100A8 and S100A9 protein ratios in different Meibomian gland severity levels (0–3)

    Association of tear proteins with dry eye symptoms

    Table 4 shows the correlations between tear protein ratios and the symptoms elicited, and their statistical significance. However, non-parametric analysis was also performed with the following results: the possible association of S100A8 protein with presence of grittiness (η=0.484, p=0.177), redness (η=0.598, p=0.043) and transient blurring (η=0.581, p=0.117); the possible association of S100A9 with symptoms of redness (η=0.594, p=0.045) and transient blurring (η=0.559, p=0.152); the possible association of lipocalin-1 with heaviness of the eyelids (η=0.522, p=0.118) and tearing (η=0.673, p=0.029); and the possible association of lactoferrin with presence of pain (η=0.536, p=0.193) and tearing (η=0.527, p=0.211). In summary, the correlations that were significant on both Pearson correlation and η were S100A8 and S100A9 levels with redness, and lipocalin-1 level with tearing.

    View this table:
    Table 4

    Correlation analysis between tear proteins and symptoms

    Discussion

    In this study on dry eye patients, we found that increasing tear levels of S100A8 and S100A9 proteins were correlated with MGD severity in dry eye patients. Second, we also found that the levels of these proteins were also associated with symptoms of redness and transient blurring. The severity of symptoms in dry eye patients did not differ significantly between MGD severity levels, but this may be related to the small sample size.

    MGD is associated with various anaerobic bacteria, alteration of lipases and formation of inflammatory abscesses.15 It is also a condition that responds to anti-inflammatory treatment.16 Therefore, it is not surprising that pro-inflammatory molecules are present in the tear of patients with MGD.

    We believe that this is the first report that has linked tear S100A8 and S100A9 proteins to MGD. S100A8 and S100A9 are members of calcium-regulated cornified envelope proteins, which are important for diverse cellular processes such as stress signalling, barrier function and innate immunity.17 Since S100A8 and S100A9 may promote or respond to epidermal hyperproliferation,17 it may also be associated with hyper-keratinisation of the Meibomian ductal epithelium found in MGD. Previously, S100A8 and S100A9 have been found to be elevated in solid tissues in inflammatory conditions,18 19 as well as in tears associated with pterygium and in the pterygium tissue.11 20 They have also been detected in normal conjunctiva tissue.20 The higher levels of S100A8 and S100A9 observed in tears may be related to keratinisation of the Meibomian gland ducts as part of the pathology of MGD. They could also be released from ocular surface epithelial cells after sustaining ocular surface damage due to evaporative dry eye. Last but not least, they could be released by macrophages or other immune cells that might be increased in MGD.

    A previous report has linked decreased tear lipocalin to MGD.21 Lipocalin-1 is known to be present in Meibomian gland secretions,6 and may be adsorbed onto the lipid layer in human tear film.22 In our study, the tear lipocalin-1 level was not significantly correlated to MGD severity; however, the correlation (r=−0.4) was almost significant (p=0.065) and may have been statistically significant had there been a larger sample size.

    Previously, there has not been any study, which has linked tear proteins or MGD to specific types of symptoms, nor was transient blurring of vision linked to MGD. It is known that lifestyle factors, such as the ‘office eye syndrome’ are associated with lipid abnormalities and foamy changes in the tear.23 However, the link with MGD and ocular symptoms has not been reported in patients. The ‘office eye syndrome’ is associated with use of computer video display units, which is associated with reduced blink rates and ocular symptoms. S100A8 and S100A9 may be produced as a response to oxidative changes in the tear, since they are also concerned with redox regulation.18 Oxidative changes in MGD may result in fast and uncontrolled epoxidation of Meibomian lipids, resulting in reduced tear stability.9 Transient blurring reflects tear instability, which can be overcome by repeated blinking to transiently restore the pre-ocular tearfilm. This may explain the association of S100A8 and S100A9 with transient blurring.

    It is interesting that lactoferrin was marginally associated with tearing and pain rather than transient blurring. Although lactoferrin is known to be associated with tear lipids22 and dry eye symptoms,24 it is largely related to lacrimal tear function rather than MGD.25 Therefore, it may be contributing to dry eye pathology via a non-MGD mechanism.

    The current technology used in this study allows for accurate determination of tear proteins in a minute amount of sample of tears.14 The levels of some of these proteins have also been validated by an independent ELISA. However, in the group of non-dry eye participants, we did not perform an evaluation of MGD and did not establish the normative values for the protein markers, which would be a step towards designing a diagnostic test. We did not employ more definitive and objective evaluation techniques for MGD such as Meibography. In addition, the sample size is relatively small and findings may or may not be generalisable to larger populations. Given the complexities of clinical definitions in dry eye and MGD and the likelihood of overlapping mechanisms, increasing the sample size by a moderate amount will still be insufficient to assess the independent effects of each variable. We believe that the current correlation analyses were valid and were sufficient to direct more studies in the future to establish the clinical significance or relevance of our findings.

    Current MS technology is not able to detect levels of cytokines at pg/ml. We are now performing cytokine analysis using an alternative technology, a bead-based indirect sandwich immunofluorescence assay (Luminex, Austin, Texas, USA), to determine if there was any correlation between inflammatory interleukin expression in the tears and S100A8 or S100A9. Such an approach may distinguish between MGD and other conditions such as autoimmune dry eye.

    This study throws insight into the nature of dry eye, and strongly suggests that MGD may have distinct contribution to dry eye disease via elevation of proinflammatory protein. Furthermore, subject to further evaluation, certain irritative symptoms may be used to detect the contribution of MGD in patients with dry eye, and hence directing therapy to the eyelids.

    Conclusion

    Distinct tear proteins are associated with MGD in dry eye patients, and some proteins were associated with distinct dry eye symptoms.

    References

    1. ↵
      1. Anonymous
      . The epidemiology of dry eye disease: report of the Epidemiology Subcommittee of the International Dry Eye WorkShop (2007). Ocul Surf 2007;5:93–107.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Schein OD,
      2. Munoz B,
      3. Tielsch JM,
      4. et al
      . Prevalence of dry eye among the elderly. Am J Ophthalmol 1997;124:723–8.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Sullivan DA
      . Tearful relationships? Sex, hormones, the lacrimal gland, and aqueous-deficient dry eye. Ocul Surf 2004;2:92–123.
      OpenUrlCrossRefPubMed
    4. ↵
      1. Bron A,
      2. Tiffany JM
      . Contribution of Meibomian gland disease to dry eye. Ocul Surf 2004;2:149–64.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Obata H
      . Anatomy and histopathology of human Meibomian gland. Cornea 2002;21:S70–4.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Tsai PS,
      2. Evans JE,
      3. Green KM,
      4. et al
      . Proteomic analysis of human Meibomian gland secretions. Br J Ophthalmol 2006;90:372–7.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Schein OD,
      2. Tielsch JM,
      3. Munoz B,
      4. et al
      . Relation between signs and symptoms of dry eye in the elderly. A population-based perspective. Ophthalmology 1997;104:1395–401.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Lemp MA
      . Report of the National Eye Institute/Industry workshop on Clinical Trials in Dry Eyes. Clao J 1995;21:221–32.
      OpenUrlPubMed
    9. ↵
      1. Butovich IA
      . The Meibomian puzzle: combining pieces together. Prog Retin Eye Res 2009;28:483–98.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Joffre C,
      2. Souchier M,
      3. Gregoire S,
      4. et al
      . Differences in Meibomian fatty acid composition in patients with meibomian gland dysfunction and aqueous-deficient dry eye. Br J Ophthalmol 2008;92:116–19.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Zhou L,
      2. Beuerman RW,
      3. Ang LP,
      4. et al
      . Elevation of human alpha-defensins and S100 calcium-binding proteins A8 and A9 in tear fluid of patients with pterygium. Invest Ophthalmol Vis Sci 2009;50:2077–86.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Zhou L,
      2. Beuerman RW,
      3. Foo Y,
      4. et al
      . Characterisation of human tear proteins using high-resolution mass spectrometry. Ann Acad Med Singapore 2006;35:400–7.
      OpenUrlPubMedWeb of Science
    13. ↵
      1. Zhou L,
      2. Beuerman RW,
      3. Huang L,
      4. et al
      . Proteomic analysis of rabbit tear fluid: defensin levels after an experimental corneal wound are correlated to wound closure. Proteomics 2007;7:3194–206.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Zhou L,
      2. Beuerman RW,
      3. Chan CM,
      4. et al
      . Identification of tear fluid biomarkers in dry eye syndrome using iTRAQ quantitative proteomics. J Proteome Res 2009;8:4889–905.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. McCulley JP,
      2. Shine WE
      . Meibomian gland function and the tear lipid layer. Ocul Surf 2003;1:97–106.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Matsumoto Y,
      2. Shigeno Y,
      3. Sato EA,
      4. et al
      . The evaluation of the treatment response in obstructive meibomian gland disease by in vivo laser confocal microscopy. Graefes Arch Clin Exp Ophthalmol 2009;247:821–9.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Eckert RL,
      2. Broome AM,
      3. Ruse M,
      4. et al
      . S100 proteins in the epidermis. J Invest Dermatol 2004;123:23–33.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Lim SY,
      2. Raftery MJ,
      3. Goyette J,
      4. et al
      . Oxidative modifications of S100 proteins: functional regulation by redox. J Leukoc Biol 2009;86:577–87.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Foell D,
      2. Wittkowski H,
      3. Ren Z,
      4. et al
      . Phagocyte-specific S100 proteins are released from affected mucosa and promote immune responses during inflammatory bowel disease. J Pathol 2008;216:183–92.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Riau AK,
      2. Wong TT,
      3. Beuerman RW,
      4. et al
      . Calcium-binding S100 protein expression in pterygium. Mol Vis 2009;15:335–42.
      OpenUrlPubMedWeb of Science
    21. ↵
      1. Yamada M,
      2. Mochizuki H,
      3. Kawai M,
      4. et al
      . Decreased tear lipocalin concentration in patients with meibomian gland dysfunction. Br J Ophthalmol 2005;89:803–5.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Miano F,
      2. Calcara M,
      3. Millar TJ,
      4. et al
      . Insertion of tear proteins into a meibomian lipids film. Colloids Surf B Biointerfaces 2005;44:49–55.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Franck C
      . Fatty layer of the precorneal film in the ‘office eye syndrome’. Acta Ophthalmol (Copenh) 1991;69:737–43.
      OpenUrlPubMed
    24. ↵
      1. Yolton DP,
      2. Mende S,
      3. Harper A,
      4. et al
      . Association of dry eye signs and symptoms with tear lactoferrin concentration. J Am Optom Assoc 1991;62:217–23.
      OpenUrlPubMed
    25. ↵
      1. Danjo Y,
      2. Lee M,
      3. Horimoto K,
      4. et al
      . Ocular surface damage and tear lactoferrin in dry eye syndrome. Acta Ophthalmol (Copenh) 1994;72:433–7.
      OpenUrlPubMed

    Supplementary materials

    Footnotes

    • Funding Grants NMRC/0808/2003, NMRC/CPG/002/2003, NMRC/0982/2005 and NMRC/1206/2009 from National Medical Research Council (NMRC), Singapore; an unrestricted grant from Allergan (Irvine, California, USA); and grant 07JCYBJ09800 from Tianjin Committee of Science and Technology Fundamental Research Project.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval This study was conducted with the approval of the Institutional Review Board of the Singapore Eye Research Institute and Tianjin Medical University Eye Center.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    Linked Articles

    • At a glance
      Highlights from this issue
      Harminder S Dua Arun D Singh
      British Journal of Ophthalmology 2011; 95 i-i Published Online First: 19 May 2011. doi: 10.1136/bjophthalmol-2011-300369