Not a dry eye in the house

Dry eye syndrome, formally known as keratoconjunctivitis sicca, is not a life-threatening condition yet it causes discomfort to an estimated 11-22% of the population worldwide. It is more prevalent in Asia, with an incidence of about 20-24%. As we age, we are more likely to become sufferers, as are people who wear contact lenses or spend a lot of time in front of a computer screen.

The most common causes of dry eye syndrome are reduced tear production, excessive tear evaporation and dysfunctional production of mucus or lipids in the tear layer. Despite its high incidence, diagnosis is not a formality because there is no definitive test for the condition. Instead, subjective tests such as measuring the volume of tear fluid or the tear breakup time are applied, some of which might be brought about by other conditions.

This uncertainty, combined with the prevalence in Asia, prompted scientists in Asia to seek an alternative approach. Proteomics techniques have been used sparingly in tear research to date, partly due to the low volume of tears, but they have suggested correlations between proteins and some conditions such as inflammatory disease.

So, proteomics were selected to study dry eye syndrome by Roger Beuerman from the Singapore Eye Research Institute, with co-researchers from the National University of Singapore, the Singapore National Eye Centre, Singapore Immunology Network, Tianjin Medical University Eye Centre, China, and Allergan Pharmaceuticals, Irvine, CA.

Tears were collected from 56 patients with dry eye and 40 healthy controls by inserting porous Schirmer strips under the lower eyelids. The proteins were eluted from the strips with phosphate-buffered saline and digested with trypsin. The resulting peptides were derivatised with the iTRAQ reagents, which differ only in the degree of isotopic substitution. The control samples were reacted with one reagent and the dry eye samples with others.

The labelled samples were combined and subjected to two-dimensional HPLC with cation exchange and reversed-phase columns in series. The final eluent was analysed by electrospray-tandem mass spectrometry and the relative peptide abundances were determined from peak area ratios for the labelled ions at m/z 114, 115, 116 and 117. The proteins associated with the peptides were identified by database searching.

A total of 93 proteins were identified with high confidence and 10 of these were differentially expressed between the dry eye patients and the controls and were potential biomarkers for dry eye syndrome. These 10 proteins were associated with aqueous secretion deficiency by the lacrimal gland and the inflammatory response in the eye.

The six up-regulated proteins were alpha-enolase, alpha1-acid glycoprotein 1, S100 A8, S100 A9, S100 A4 and S100 A11. The four down-regulated proteins were prolactin-inducible protein, lipocalin-1, lactoferrin and lysozyme.

On a simple single-protein basis, the level of alpha-enolase correctly identified dry eye in 85% of cases. Prolactin-inducible protein was the best of the down-regulated proteins with an accuracy of 81%. However, in keeping with current thinking that one protein alone may not be sufficient to characterise a clinical condition, the researchers used a statistical model to identify a panel of four predictive proteins. alpha-Enolase, prolactin-inducible protein, lipocalin-1 and S100 A9 gave a diagnostic accuracy of 96%.

In a further finesse of the results, the levels of alpha1-acid glycoprotein 1, S100 A8, S100 A9 were correlated to the severity of the dry eye condition. They are all inflammatory proteins and were able to distinguish between mild, moderate or severe dry eye.

Increased levels of S100 A8 and S100 A9 also correlated with older age of the patients but there was no age association for the remaining candidate biomarkers. All of the observed changes in protein levels were similar in patients from Singapore and Tianjin, China.

The identification of two groups of proteins that can effectively diagnose the condition and predict its severity shows the value of this proteomics approach. It is suitable for the small sample volumes and has led to a diagnostic panel of proteins that could be incorporated into a simpler clinical test for dry eye syndrome.

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.