Our study reports the proteins identified in tears of different subgroups of DE patients by using a novel method of quantitative proteomics. Statistically significant differences in the protein ratios were detected between normal and DE groups. Our study showed a greater number of proteins down-regulated in MSDE than other DE groups (
Table 3). Unique proteins were associated with each subgroup of DE. iTRAQ technology is a relatively new protein quantification method, especially in tear film, that allows the analysis of multiple samples to be obtained simultaneously.
To date, there are very few studies in the literature on tears proteomics in humans using this novel quantitative proteomic method. Zhou et al.
13 and Tong et al.
14 have employed iTRAQ technique to study proteins in human tears. Zhou et al. identified tear biomarkers in DE patients in a recent study that demonstrated 4 down-regulated proteins (PIP, lipocalin-1, lactoferrin, and lysozyme) in DE patients.
13 These results are comparable to our study in which three (PIP, lipocalin-1, and lysozyme) of the four proteins were detected. However, the number of upregulated proteins is different in both studies.
A comparison of defensin (NP-1 and NP-2), levels in tear film in rabbits before and after corneal wounding (days 1–3) was conducted by Zhou et al. with iTRAQ experiments.
19 Similarly, iTRAQ also has been used to quantify the relative difference in tear protein N-glycosylation levels between controls and patients with climatic droplet keratopathy.
41 Tong et al. concluded that unique tear proteins (S100A8 and S100A9) are associated with meibomian gland dysfunction (MGD) in DE patients, and the level of certain proteins, such as S100A8, S100A9, and lipocalin-1, correlated significantly with grittiness, and symptoms of redness, transient blurring, heaviness of the eyelids, and tearing, respectively.
14 Our results need exploration to correlate individual protein levels with specific symptomatology of DE.
Mass spectrometry analysis by Versura P et al., conducted in tear samples of 60 evaporative DE patients, detected the following proteins in tears, including lactotransferrin, serum albumin, lipocalin-1, lipophilin-c, extracellular glycoprotein lacritin, proline-rich protein-4, lysozyme C, lipophilin-a, prolactin-inducible protein, proline-rich protein-1, Ig gamma-1 chain C region, serotransferrin, and zinc-α-2-glycoprotein.
20 A statistically significant decrease in lactoferrin, lipophilin, and lipocalin amount was found in patients versus controls; however, our study did not show a change in lysozyme and zinc-alpha-2-glycoprotein amounts. Tear proteomics on contact lens wearers,
42 Sjögren's syndrome patients,
13,16–18 and keratoconus patients
43,44 have shown upregulation and down-regulation of unique proteins.
Our study has shown two upregulated proteins, namely aldehyde dehydrogenase and haptoglobin, in the MDE group. These two upregulated proteins did not appear in the MSDE group. At this stage, it is unclear as to why there was an upregulation noted only in the MDE group. Interestingly, a recent report by Joseph et al. examined epithelial and stromal proteins from keratoconus and normal cornea using label-free Nano-ESI-LC MS (MS)2, and showed that there was an upregulation of aldehyde dehydrogenase and haptoglobin.
45
There are mixed views in the literature on the levels of certain proteins in different age groups. For upregulated tear proteins, Zhou et al. reported that higher levels of proteins, such as S100A8 and S100A9, were associated significantly with increased signs of dryness,
13 and this association still remained in all age groups. No significant correlation was found for the rest of the tear biomarker candidates that were assessed. A previous study reported that, due to the reduction in secretion function of the lacrimal glands in the elderly, there was a negative correlation between age and some tear protein levels, such as lysozyme and lactoferrin.
24 However, another recent study conducted by Zhou et al. showed that age correlation assessments did not demonstrate significant association between age, and the levels of lysozyme and lactoferrin.
13
The biologic functions of commonly reported down-regulated tear proteins in DE, such as lysozyme, lactotransferrin, and lipocalin, include antibacterial protection of the ocular surface
2 and general protection factor of epithelial cell surfaces, respectively.
46 PIP was found to be decreased in patients with MGD, which is involved in water transport function.
14 A recently conducted study in Sjögren's syndrome tears also showed a few similar proteins, which were detected in our current study, including Ig alpha-2 chain C, Ig alpha-1 chain C, zinc-alpha-2 glycoprotein, and polymeric-immunoglobulin.
47 The determination of biologic functions of other proteins highlighted in results in relation to DE needs further research.
In our study, several proteins were identified multiple times (greater than 2 times, but less than 6 times); however, these proteins were not reported here, which may be of relevance. Differential protein expression can be detected between normal and DE patients using iTRAQ.
The sample size of our study was small and warrants further work to show consistency in the number of proteins expressed, and the ratio of upregulated and down-regulated proteins in larger samples. Tear collection methods, patient grouping techniques, and tear processing techniques have a major role in tear proteomics. In our study, a full battery of DE diagnostic tests was not performed, thus common DE diagnostic elements, aqueous production, and symptoms, were used to classify patients. Our study was exploratory in nature, and future exploratory and confirmatory studies could use this as well as other grouping schemes, including additional tests, such as the evaluation of the status of the meibomian glands. Schirmer's test is particularly problematic and variable, and may be influenced by reflex tearing to some extent, so its value does not reflect merely basal aqueous production. In this classification, this fact is particularly relevant as some of the mixed category may have the measured Schirmer's partly as a result of reflex tearing.
Consistent approaches to tear collection and analysis additionally are required. There also is a need for the standardization of the number of replicates while performing tear proteomics. Research data obtained using new proteomics tools, such as iTRAQ, must be validated with comprehensive bioinformatic tools, such as PIR. The use of protein ID mapping supports the functional inference of down-regulated or upregulated proteins in relation to the severity of DE. Continued work to characterize involved pathways is needed. Future work to validate findings using Western blot and ELISA assays are planned.