Multimodal imaging and functional analysis of the chick NMDA retinal damage model

Abstract

Objectives: The chick is rapidly becoming a standardized preclinical model in vision research to study mechanisms of ocular disease. We seek to comprehensively evaluate the N-methyl-D-aspartate (NMDA) model of excitotoxic retinal damage using multimodal imaging, functional, and histologic approaches in NMDA-damaged, vehicle-treated, and undamaged chicks.

Methods: Chicks were either left undamaged in both eyes or were injected with NMDA in the left eye and saline (vehicle) in the right eye. TUNEL assay was performed on chicks to assess levels of retinal cell death one day post-injection of NMDA or saline and on age-matched untreated chicks. Spectral domain optical coherence tomography (SD-OCT) was performed weekly on chicks and age-matched controls day 1 (D1) up to D28 post-injection. Light adapted electroretinograms (ERG) were performed alongside SD-OCT measurements on post-injection chicks along with age-matched untreated controls.

Results: Untreated and vehicle-treated eyes had no TUNEL positive cells while NMDA-treated eyes accumulated large numbers of TUNEL positive cells in the Inner Nuclear Layer (INL), but not other layers, at D1 post injection. Significant inner retina swelling or edema was found on SD-OCT imaging at D1 post-injection which resolved at subsequent timepoints. Both the INL and the inner plexiform layer significantly thinned by one-week post-injection and did not recover for the duration of the measurements. On ERG, NMDA-treated eyes had significantly reduced amplitudes of all parameters at D1 with all metrics improving over time. The b-wave, oscillatory potentials, and ON/OFF bipolar responses were the most affected with at least 70% reduction immediately after damage compared to the fellow eye control.

Conclusion: This study establishes a normative baseline on the retinal health and gross functional ability as well as intraocular pressures of undamaged, vehicle-treated, and NMDA-damaged chicks to provide a standard for comparing therapeutic treatment studies in this important animal model.

Fig 1. Representative fundus imaging.

Choroidal vessels are visible in healthy chick eyes (A). While in NMDA-damaged chick eyes they can no longer be observed, instead we see the retina assume a homogenous white color potentially due to a lack of perfusion (B). The sharp and clear presence of the pecten rules out media opacity in NMDA-damaged eyes.

Fig 2. IOP measurements.

Chick IOP measurements on undamaged chicks and NMDA-damaged chicks (OD: Saline, OS: NMDA) taken weekly starting at post-hatch day 7 or post-injection day 1 till post-hatch day 35 or post-injection day 28.

Fig 3. NMDA induced TUNEL response in chick retina.

Untreated, vehicle (sterile saline), and NMDA injected chicks at D1 post injection. Both untreated and vehicle-injected chicks displayed no TUNEL-positive cells. NMDA-treated chicks saw a large number of TUNEL-positive cells in the INL, but not other retinal layers. The scale bar denotes 50μm. Abbreviation: GCL—ganglion cell layer, INL—inner nuclear layer, ONL—outer nuclear layer.

Fig 4. SD-OCT analysis of NMDA-treated and corresponding control eyes.

Representative b-scans of D1 vehicle (saline, left) and D1 NMDA (right) showing edema in IPL (A), D7 vehicle (saline, left) and D7 NMDA (right) (B), and D14 vehicle (saline, left) and D14 NMDA (right) (C). (D) Analysis of retinal thicknesses of the IPL, RNFL-IPL combined, INL, and total thickness between NMDA and saline (vehicle) treated eyes at P7/D1 to P21/D14 post injection and their respective age-matched untreated controls. Black brackets show IPL thickness, white brackets show INL thickness, asterisk denotes hyperreflective layer. Error bars are shown as standard deviation. Abbreviations: RNFL—retinal nerve fiber layer, IPL—inner plexiform layer, INL—inner nuclear layer, ONL—outer nuclear layer, RPE—retinal pigment epithelium.

Fig 5. ERG results of NMDA-damaged, vehicle-treated, and age-matched undamaged controls.

Undamaged chicks (n = 10 P7 to P21, n = 5 P28 to P35), NMDA-damaged chick eyes (OS), and saline vehicle-injected fellow chick eyes (OD) (n = 10, P7/D1 to P14/D7, n = 8, P21/D14, n = 6, P28/D21, n = 5, P35/D28) were examined with the Celeris ERG system. The data is organized by time (columns) and parameter (rows). From left to right, the columns represent one day post-injection (D1) and 7 days post-hatch (P7), D7/P14, D14/P21, D21/P28, and D28/P35. From top to bottom, the rows represent a-wave amplitudes, b-wave amplitudes, the amplitudes of the sum of oscillatory potentials (OPSum), flicker amplitudes at 20Hz, and ON/OFF bipolar cell response amplitudes. * indicates significance with vehicle (saline) controls, † indicates significance with undamaged controls.

Fig 6. Quantitative characteristic model of SD-OCT and ERG behavior during chick NMDA damage.

SD-OCT uses average layer thicknesses of NMDA-damaged chick eyes from D1 to D28 showing key features such as the edema present at D1 in the IPL and the hyperreflectivity present from D7 onwards in the INL (A). The ERG summary shows the average amplitudes as a percent of the fellow eye control at the 5 cd.sec/m² flash intensity from D1 to D28 (B).