Abstract
Purpose :
While numerous novel therapeutic biologics are being developed for ocular diseases, there is little available information about the tissue expression levels of the target proteins for these agents. The variable density of ocular matrices and limited sample availability present unique challenges for ocular target protein quantitation. To reliably quantitate proteins in multiplex assays, we have developed and qualified a method that reduces interference from complex ocular matrices.
Methods :
Frozen cynomolgus and 6 pairs of donor human eyes were partially thawed, and the macula taken as a 10mm temporal punch. Aqueous, vitreous and extra-macula retina/choroid were collected after eye dissection removal and of the anterior segment. Aqueous/vitreous samples were diluted 1:1 with PBS + protease inhibitors. Vitreous was liquefied by passing through 23/25G needles, centrifuging at 12,500G (15 mins/4C), and passing the supernatant over a 0.45µm PVDF Durapore column. Retina/choroid were separated, homogenized in lysis buffer (T-PER vs TER-I) + protease inhibitors, supernatant collected, rested for 1hr (4C), and centrifuged at 12,500G (15 mins/4C). Total protein was measured using a BCA assay and EMD Millipore multiplex assays were used to quantitate 59 different markers of angiogenesis, complement and immunity. A range of sample dilutions were used to determine specificity, accuracy, precision, and linearity.
Results :
Tissue viscosity, lysis buffer and total protein all interfered with Luminex bead recovery. The viscosity of aqueous/vitreous could be overcome by tissue processing. Processing of all tissues yielded protein concentrations ≥1µg/µl. T-PER lysis buffer interfered with phycoerythrin conjugation yielding results below the limit of quantitation (BLQ) while TER-I produced quantifiable protein levels. Dilution was critical to overcome protein interference. A total protein/well of 15-30µg was established as the minimal required dilution (MDR) for acceptable standard recovery (±20%). Spike/QC recovery was performed to assess linearity and specificity, confirming ‘Goodness-of-Fit’ in all matrices.
Conclusions :
Preanalytical variables such as tissue processing, lysis buffer, and total protein loaded, must be optimized to allow accurate determination of protein analytes in ocular matrices.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.