Fig 1.
Processing of Vitamin A in The Visual Cycle.
Enzymatic processing within the visual cycle begins with delivery of vitamin A (all-trans-retinol) from the blood circulation. Upon entry into the RPE, all-trans-retinol is converted to a retinyl ester through the activity of lecithin retinol acyl transferase (1). The resulting all-trans-retinyl ester pool represents a storage form of vitamin A upon which RPE65 acts to generate 11-cis-retinol (2); 11-cis-retinol is then oxidized by an 11-cis-specific retinol dehydrogenase to form the visual chromophore, 11-cis-retinal (3). The visual chromophore is delivered to rod and cone outer segments (4) where it combines with opsins to form visual pigments (e.g., rhodopsin). Light activation of rhodopsin initiates visual transduction processes and liberates all-trans-retinal as a photoproduct. Reduction of all-trans-retinal, via all-trans-retinal dehydrogenase, produces all-trans-retinol (5), which is transferred back to the RPE for recycling. The continued activity of RPE65 in the light state ensures sustained levels of rhodopsin, closure of ion channels through transducin activation, and reduced oxygen demand.
Fig 2.
Inhibition of Isomerase Activity and Visual Chromophore Regeneration.
The effect of emixustat on isomerase activity and visual chromophore regeneration was evaluated in in vitro and in vivo assays. Reaction conditions for in vitro assays are described in Methods. Retinoids in the sample extracts were measured by HPLC (RE, retinyl esters; ROL, retinols). Panel A shows representative HPLC chromatographic tracings from a control sample containing no emixustat (black trace) and a sample containing 1 μM emixustat (red trace). Note absence of the 11-cis-ROL peak in the sample containing emixustat. The dose-response curve for inhibition of 11-cis-ROL production by emixustat is provided in panel B (each data point represents mean ± SEM, n = 8). The mean 11-cis-ROL level in control samples (no emixustat) was taken as 0% inhibition; half maximal inhibitory concentration (IC50) for 11-cis-ROL production in the presence of emixustat is provided in the panel B inset. The effect of emixustat on visual chromophore regeneration in vivo was evaluated in wild-type mice as described in Methods. In these analyses, the amount of 11-cis-retinal production, at 2 hours in darkness following a photobleach (measured by HPLC as syn- and anti-11-cis-retinal oximes, s-11-cis-ROx and a-11-cis-ROx, respectively), in mice treated with varied doses of emixustat was examined. Sample chromatographic tracings of the retinoid profile from a vehicle-treated mouse (black trace) and a mouse treated with 1 mg/kg emixustat (red trace) are shown in panel C (syn- and anti-All-trans-retinal oximes, s-All-trans-ROx and a-All-trans-ROx, and all-trans-retinyl esters, REs, are also shown). Note reduced level of syn- and anti-11-cis-retinal oxime peaks and increased REs in the trace from the emixustat-treated mouse. The dose-response curve for inhibition of visual chromophore production by emixustat is provided in panel D (each data point represents mean ± SEM, n = 8). The mean 11-cis-retinal oxime level in control samples (no emixustat) was taken as 0% inhibition. The calculated half maximal effective dose (ED50) for inhibition of visual chromophore production by emixustat is provided in the panel D inset.
Fig 3.
Pharmacodynamic Effect of Emixustat on Rod Function.
The recovery of rod photoreceptor function following a photobleach was measured by ERG. Mice were treated with varied doses of emixustat (n = 4/dose group), or vehicle (n = 4/group), and ERG responses were recorded every 2 minutes for 50 minutes using a light stimulus 0.01 cd*s/m2. Panel A shows representative ERG waveforms in vehicle- and emixustat-treated mice. Arrows in panel A indicate the b-wave amplitude peaks which are plotted as a function of recovery time after photobleach in panel B. B-wave response amplitudes (μV), at each corresponding time point of recovery, are shown as mean values ± SEMfor each of the treatment groups (panel B). Emixustat treatment caused a dose-dependent suppression of b-wave response amplitudes. The dose required for half-maximal suppression of ERG b-wave recovery was determined to be 0.21 mg/kg.
Fig 4.
The ability of emixustat to provide protection from light damage was assessed as described in Methods. Mice received a single dose of emixustat, or vehicle, prior to light exposure (8,000 lux white light, 1 hour). Histological analyses and determination of ONL thickness was performed following a 2-week recovery period. Panel A shows tissue sections prepared from untreated, dark-adapted mice, untreated light-exposed mice, and mice pre-treated with either 0.3 or 1.0 mg/kg emixustat (ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer). Quantitative analysis of ONL layer thickness (panel B) was used to calculate the ED50 for preservation of the ONL in emixustat-treated mice (ED50 = 0.20 mg/kg). There was a statistically significant preservation of ONL thickness in mice treated with 1.0 or 3.0 mg/kg emixustat compared to the vehicle-treated, light control group (*, p<0.01).
Fig 5.
Reduced Lipofuscin Autofluorescence and A2E Accumulation.
The effect of emixustat on lipofuscin autofluorescence and A2E levels was examined in an animal model of autosomal recessive Stargardt disease (Abca4-/- mice). Histological analysis of lipofuscin autofluorescence in untreated, strain- and age-matched wild-type mice, and Abca4-/- mice treated with either vehicle or emixustat (0.3 or 3.0 mg/kg) is shown in panels A—D, respectively. All mice were 5 months of age. Lipofuscin fluorophores were extracted from RPE eyecups and analyzed by HPLC as described in Methods. Representative chromatograms from eyecup extracts of mice treated with emixustat (3 mg/kg for 3 months) and vehicle (red and black tracings, respectively) are shown in panel E. UV-vis spectra associated with the indicated peaks (numbered 1–5) are shown in the panel inset. Peak 1 in the chromatogram was determined to be A2E based upon spectral identity and co-elution with an authentic A2E standard. Quantitative analysis of A2E levels (based on area units of peak 1) in Abca4-/- mice is shown in panel F (peaks 2–4 were not quantified). Numbers of mice analyzed for A2E quantitation are as follows: Day 0: n = 17; Vehicle: n = 6; 0.03 mg/kg: n = 6; 0.1 mg/kg: n = 7; 0.3 mg/kg, n = 8 1.0 mg/kg, n = 8; 3 mg/kg, n = 8. A2E levels increased from ~5 to ~20 pmoles/eye over a 3-month period in vehicle-treated Abca4-/- mice. Abca4-/- mice treated with emixustat showed a dose-dependent reduction of A2E which was statistically significant at doses ≥ 0.30 mg/kg/day, relative to vehicle treated Abca4-/- mice (*, p<0.05). The ED50 for the effect of emixustat on reducing accumulation of A2E was 0.47 mg/kg/day.
Fig 6.
Reduced Retinal Neovascularization.
The effect of emixustat on retinal neovascularization was studied in the mouse OIR model. Seven day-old mouse pups were subjected to hyperoxia (75% oxygen) for 5 days. On P12, the mice were returned to room air and daily treatments with ruboxistaurin (10 mg/kg), emixustat (0.03–3.0 mg/kg), or appropriate vehicles were administered as described in Methods. Retinal flat mounts were prepared and areas of NV were quantified; these data were compared to data from control mice that were maintained in a normoxic environment (21% oxygen). Mice that were moved from a hyperoxic to normoxic environment, without treatment, showed a significant extent of retinal NV (~30% of the retinal area). Treatment with the ruboxistaurin (positive control) reduced the area of NV to ~20% of the total retinal area. In mice treated with emixustat, a dose-dependent reduction in retinal NV was observed. The reduction in NV at the highest emixustat dose (3.0 mg/kg/day) approached a level that was comparable to that obtained with ruboxistaurin (Fig 6A; *, t-test, p<0.05). The ED50 for reduction of retinal NV in emixustat-treated mice was 0.46 mg/kg/day. Representative retinal flat mounts from an untreated, normoxic control, an untreated OIR control, and in a 3 mg/kg emixustat-treated mouse are shown in panels B, C, and D, respectively. Areas of NV, outlined in red tracings, were identified and quantified using Adobe Photoshop software.