Human tear protein biomarkers are useful for detecting both ocular and systemic diseases. Unfortunately, existing tear film sampling methods (Schirmer strips (SS) and microcapillary tubes (MCT)) have significant drawbacks. Here we present an alternative tear protein sampling method using soft contact lenses (SCLs). First, we optimized SCL protein sampling in vitro, then performed in vivo studies in 6 human subjects. Using Etafilcon A SCLs and 4M guanidine-HCl for protein removal, we sampled an average of 60 +/- 31 ug of protein per eye. We also performed objective and subjective assessments of all sampling methods. Proteomic analysis by mass spectrometry (MS) revealed unique subsets of proteins for SS and SS/MCT methods. Additionally, SCLs had reduced levels of reflex tear protein zymogen granule protein 16 homolog B (ZG16B) versus SS and MCT. These experiments demonstrate SCLs as an accessible tear sampling method which may minimize reflex tearing and potentially aid biomarker discovery.

Soft contact lenses (SCLs) have recently been introduced as an alternative method for human tear protein sampling. However, SCLs are available in a variety of chemical compositions which affect protein binding specificity. Here we analyzed 8 different SCL materials to identify an optimal lens for tear protein sampling. Polymer contamination, mass spectrometry (MS) sample preparation method, total protein capture, individual protein specificity, and SCL cost were all assessed. Using a filter-aided sample prep (FASP) method with 4M guanidine for protein removal, only etafilcon A and verofilcon A did not have significant polymer contamination. Polymer was successfully removed using phosphate buffered saline (PBS) with S-Trap columns for all SCL materials, though yielding a slightly lower number of protein identifications per sample. Minor quantitative differences were observed between SCL materials. However, we also saw significant intersubject variation in protein abundance. Of all the assessed SCL materials, verofilcon A lenses yielded the most total protein while comfilcon A and senofilcon A had the least protein variability. As a newly released daily disposable modality (Precision 1, Alcon), verofilcon A has one of the longest predictable production schedules and one of the lowest costs per lens, making it beneficial for large-scale experiments and diagnostics. Furthermore, we demonstrate how protein binding bias with SCL tear sampling is useful for intra-experiment normalization. Overall, these experiments have led us to optimize our previous protocol for SCL tear protein sampling, highlighting important differences between SCL materials and identifying etafilcon A and verofilcon A as optimal materials for tear protein sampling.