Purpose:
To evaluate the potential tear matrix effect on the sensitivity and selectivity of quantitation of tear inflammatory mediators using antibody arrays including linearity of the signal derived from a 4- point tear titration as well as the spiked recovery using the standard criteria of ± 20%.
Methods:
Tears were collected from both eyes using microcapillary tubes and Schirmer strips from subjects categorized into Normal (N) and Dry Eye (DE) groups using the modified Delphi/DEWS dry eye severity grading scheme. Tears were also collected from normal contact lens-wearing subjects (CL). A RayBiotech Human Inflammation-3 Quantibody® array and a 10-analyte custom Quantibody array were analyzed. The arrays were scanned using the Axon GenePix and data was processed using RayBiotech Q-Analyzer® software. Cytokine standards were serial diluted to create a standard curve. 2µL, 5µL (in duplicate) and 10µL tears were diluted to 100µL with dilution buffer to test for linearity and reproducibility of the assay. In addition a 5µL tear sample was diluted with standards to calculate the spike signal by comparing to endogenous signal from the tear sample. 5µg, 10µg and 20µg of protein extracted from Schirmer strips were similarly assayed.
Results:
For the Inflammation-3 array, 9 proteins from tears and 13 from Schirmer extracts were linear and met the spiking recovery criteria, and 8 proteins from both Schirmer extracts and tears met the same criteria on the custom array (Table 1). More proteins were observed in the DE and CL groups compared to the N group while similar proteins were seen in both DE and CL groups.
Conclusions:
Stringent criteria need to be applied on array-based outcome measures to avoid false positives from clinical data. It is crucial to exercise careful consideration in the choice of analytes and optimize performance of the array prior to analyses of clinical study samples.
Keywords: cornea: tears/tear film/dry eye • cytokines/chemokines • contact lens