Abstract
Purpose: :
Fundus imaging was performed in wild type and knockout mice to prescreen for abnormalities and characterize the models prior to enrollment in research protocols. The failure to identify and exclude mice with preexisting retinal conditions has the potential to confound experimental outcomes and/or study conclusions.
Methods: :
Imaging was performed in mice approximately 2 months of age. The strains investigated include C57BL/6J, BALB/c, complement knockouts C3, C4, factor B, factor H, and SOD1. Knockout mice all originated from C57BL/6J background. Anesthesia was induced IP using 90mg/kg of sodium pentobarbital. Mydriasis was induced with 0.5% Tropicamide/Phenylephrine. Corneal hydration was maintained using Systane Ultra and balanced salt solution drops. Fundus imaging was performed using confocal scanning laser ophthalmoscopy (SLO) and spectral-domain optical coherence tomography (OCT). After imaging eyes received ophthalmic antibiotic ointment to prevent corneal dehydration and permit subsequent imaging at a later date. Abnormalities found were documented and tabulated as a percentage related to the total and type of mouse imaged.
Results: :
A routine was developed for obtaining baseline SLO and OCT images. Imaging data from over 700 mice was obtained. The use of dual imaging systems, with up to six separate imaging modalities available, permitted comprehensive screening for abnormalities in the central and peripheral fundus regions. Abnormalities were detected in every mouse strain to varying degree. Abnormalities in wild type mice were more numerous in BALB/cJ than in C57BL/6J line. BALB/cJ abnormalities are widespread and are associated with retinal infoldings. SOD1 and factor H knockout mice also showed a high frequency of abnormalities. The distribution of abnormalities found among SOD1 mice were 15, 27 and 17% for wild type (+/+), heterozygous (+/-) and knockout (-/-) animals, respectively. SOD1 lesions appear to be associated with and/or directly under retinal vessels. Factor H mice exhibited a population of two types of abnormalities with one specific type accounting for the majority (85%) detected and involving the peripheral retina. The percentage of abnormalities observed for each strain/knockout line was as follows: C3-/- (1.8%), BALB/cAnNTac (3.3%), C57BL/6J (3.4%), factor B-/- (8.3%), C4-/- (12.3%), factor H-/- (18.5%), SOD1+/+, +/-, -/- (20.6%), and BALB/cJ (100%).
Conclusions: :
Retinal abnormalities exist to varying degree in presumably normal mice, including wild type animals. This study underscores the importance of prescreening mice using one or more of the available fundus imaging modalities to avoid enrolling animals with preexisting fundus abnormalities into experimental protocols.
Keywords: imaging/image analysis: non-clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • retina