Optimization of transgene expression in the trabecular meshwork (TM) by comparing capsid mutant AAV2 vectors

Abstract

Purpose : We have previously reported successful targeting of GFP expression to the rodent TM with capsid-mutant AAV2. The purpose of our study was to optimize targeting of the canine TM with AAV2 by altering capsid mutations, pupil size, and route of intraocular vector administration.

Methods : 18 eyes of 11 adult, wild-type dogs were treated by intraocular injections of AAV2-GFP with a small chicken beta actin promoter. TM fluorescence was evaluated by gonioscopy and immunohistochemistry (IHC). First, we compared 3 capsid mutants at 2 doses (10⁹ and 10¹¹ vg) following intracameral (IC) injection: single (Y444F), triple (Y444, 500, 730F), and quadruple (Y444, 500, 730F + T491V) capsid amino acid changes. Once we selected AAV2-Y444F as the benchmark vector, we compared doses over a range of 3 log units (10⁹-10¹² vg), and evaluated the effect of pupil size at the time of IC injection (latanoprost-induced miosis vs. atropine-induced mydriasis) on TM transgene expression. Finally, we compared IC vs. intravitreal (IVit) vector administration. The dogs were monitored for up to 4.6 months after treatment. GFP expression within the ICA was measured by IHC using ImageJ (% GFP-stained area within iridocorneal angle in a 4x low-power field).

Results : Of the 3 capsid mutant vectors only AAV2(Y444F) resulted in detectable TM GFP expression and was used as the benchmark vector in subsequent experiments. While at the 10⁹-vg dose GFP fluorescence was barely detectable (0.01% area), the area of transgene expression was much larger (0.35-0.4%) with 10¹⁰ and 10¹¹ vg, respectively. GFP fluorescence was most robust with 10¹² vg (4.1% area, including both TM and uveoscleral outflow pathway) with clinically detectable fluorescence by gonioscopy. The highest vector dose also resulted in uveitis at 5 weeks post injection, a likely adverse reaction to the expressed GFP. There were no differences in transgene expression between the 4 quadrants. The extent of GFP expression within the TM was only minimally affected by the pupil size at the time of IC vector injection, with the mydriatic pupil resulting in lower fluorescence (2.8 vs. 4.8%). Both IC and IVit vector administration resulted in comparable TM GFP expression; levels were slightly higher with IC injection (4.1 vs. 2.8%).

Conclusions : IC injection of AAV2(Y444F) at a dose of 10¹² vg resulted in the most robust transduction of the canine TM.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.