Biodistribution of adeno‐associated virus type 2 carrying multi‐characteristic opsin in dogs following intravitreal injection

Abstract Gene therapy of retinal diseases using recombinant adeno‐associated virus (rAAV) vector‐based delivery has shown clinical success, and clinical trials based on rAAV‐based optogenetic therapies are currently in progress. Recently, we have developed multi‐characteristic opsin (MCO), which has been shown to effectively re‐photosensitize photoreceptor‐degenerated retina in mice leading to vision restoration at ambient light environment. Here, we report the biodistribution of the rAAV2 carried MCO (vMCO‐I) in live samples and post‐mortem organs following intraocular delivery in wild‐type dogs. Immunohistochemistry showed that the intravitreal injection of vMCO‐I resulted in gene transduction in the inner nuclear layer (INL) but did not induce detectable inflammatory or immune reaction in the dog retina. Vector DNA analysis of live body wastes and body fluids such as saliva and nasal secretions using quantitative polymerase chain reaction (qPCR) showed no correlative increase of vector copy in nasal secretions or saliva, minimal increase of vector copy in urine in the low‐dose group 13 weeks after injection and in the faeces of the high‐dose group at 3–13 weeks after injection suggesting clearance of the virus vector via urine and faeces. Further analysis of vector DNA extracted from faeces using PCR showed no transgene after 3 weeks post‐injection. Intravitreal injection of vMCO‐I resulted in few sporadic off‐target presences of the vector in the mesenteric lymph node, liver, spleen and testis. This study showed that intravitreal rAAV2‐based delivery of MCO‐I for retinal gene therapy is safe.

since there is not a FDA-approved therapy that can halt the degeneration. 8 Importantly, other than retinal prostheses, there are no clinically meaningful therapies that can restore vision once vision loss has ensued. 9 Retinal implants are available for late-stage RP patients under a Humanitarian Device Exemption (Argus II, Second Sight) and are indicated for use in one eye for patients with severe to profound RP, defined as bare light or no light perception in both eyes.
However, only partial restoration of vision is possible due to inherent limitations to the retinal implant device. Also, a complex, invasive surgical procedure is involved, can only be used in one eye, requires rigorous training with behaviourial experts and includes a number of warnings and precautions related to electromagnetic interference. The ability of optogenetics to address the unmet need in RP has received considerable attention 10 for its potential clinical utility, and unlike retinal implants, invasive surgery is not required. Optogenetics can be delivered to both eyes and offers the theoretical advantage of better resolution and one-time administration. Also, while direct electrical stimulation approaches require mechanical contact of electrodes to the retinal cells, indirect stimulation approaches such as optogenetic stimulation 11  which effectively re-photosensitizes photoreceptor-degenerated retina in mice leading to vision restoration at ambient light environment. 12 Compared to other opsins, MCO has the advantage of functioning in ambient light and broad visible spectrum. 12 Significant photocurrent is generated in MCO-sensitized cells at white light intensity levels close to ambient light conditions without compromising the fast kinetics required to form vision. Owing to ambient light sensitivity, no external device-based light stimulation is needed for MCO-activation, thus eliminating potential phototoxicity. Further, MCO is polychromatic opsin that has broad activation spectrum, and therefore, subjects with MCO-sensitized retina will have potential to regain vision at different colour environments.
The development of recombinant adeno-associated viral (rAAV) vectors provides excellent vehicles for efficacious delivery and long-term expression of gene therapy molecules, thus opening up new vistas for curing degenerative diseases. Several in vitro and in vivo systems are used for preclinical studies to produce information assuring safe administration of the investigational drug to humans. 13 The MCO-encoding gene, with distal CMV promoter and ON-bipolar cell-specific mGluR6-enhancer along with mCherry as a reporter and enhancer, is packaged in an adeno-associated virus type 2 (AAV2) vector for transducing reti- The relevance of the vMCO-I in vision restoration and study outcomes are fundamental for successful translation of optogeneticsbased technologies in human. 13 Determining the AAV vector biodistribution is often conducted prior to early-phase clinical trials for the safety. [13][14][15] To determine whether the distribution of the vector may pose a risk to the patient, the level of genomes should be quantified in major organs as well as other tissues or fluids pertinent to the disease, gene therapy vector, transgene and route of administration. 15 Though biodistribution studies in mice at different time points after intravitreal injection of different doses of vM-CO-I have been conducted, a study in a larger animal model that is more anatomically and physiologically relevant is required to justify first-in-human dosing. This will minimize uncertainty surrounding the scaling of vector dose from animal to human. The large size of the canine eye, the volume of the vitreous and its consistency and the anatomy of the retina that comprises a cone rich fovea-like central region in the canine macula is close to that of the human.

| Vector construct of ambient light activatable multi-characteristic opsin
Multi-characteristic opsin (MCO) gene under metabotropic glutamate receptor mGluR6 with reporter mCherry (mGluR6-MCO-I-mCherry) was designed, constructed using rDNA technology and cloned at the restriction sites (BamH I and SalI) of pAAV-MCS vector ( Figure S1). The cloned mGluR6-MCO-I-mCherry sequence was validated by sequencing and sequence alignment. After validating the cloned sequence, the plasmid construct was used for production of vMCO-I.

| Preparation of animals and intravitreal injection conditions
All animals were cared for and treated in accordance with the Nanoscope Technologies sponsored CRO's IACUC approved protocol (# NS-1702). The monocular intravitreal injections, similar to those that would be used in human eyes, were performed in three different groups of wild-type beagle dogs. Each group consisted of four dogs (two males and two females). Intravitreal injections of the right eyes were performed with the control vehicle (75 µl, 8.6 × 10 12 Viral Genomes/ml, ie VG/ml) administered to Group 1 dogs, the high-dose (75 µl, 8.6 × 10 12 VG/ml) vMCO-I administered to Group 2 dogs and the low-dose (75 µl, 1 × 10 12 VG/ml) vMCO-I article administered to the Group 3 dogs (Table S1A). Stagger 1 dogs were injected on 1 day, and then, the stagger 2 dogs were injected on a following day. The viral titre used in the intravitreal injection was determined by qPCR with standard curve generated by linearized

| Tissue extraction
At the pre-scheduled time point (13 weeks post-injection), the dogs were euthanized and different organs (heart, liver, spleen, kidney, mesenteric, mandibular, testis/ovary) from each dog in the three different groups were collected. The organ tissues were kept in the 1.8 ml cryovials and stored at −80°C. Each vial was properly labelled with study number, animal identification number, date of extraction and name of organ. Faeces, saliva and nasal secretions were collected at 1, 3, 13 weeks post-injection (Table S1B).

| Immunofluorescence microscopy
Dog tissue samples were fixed in 10% neutral-buffered formalin and embedded in paraffin wax. De-paraffinized tissue sections (6 mm thick) were used for immunohistochemistry staining. The immunohistochemistry reagents included a washing solution (0.5% Triton in 1× PBS), a blocking solution (4% goat serum in washing solution), primary antibodies solutions against mCherry, PKC-alpha, interferongamma and CD45 (Table S7), and secondary antibodies such as Dylight 488, Alexa Fluor 488 and Alexa Fluor 568 (Table S8). Table S9 shows the dilution of the respective antibodies. Tissue sections of both control and vMCO-I-injected dogs were initially reacted with primary antibody (Tables S7 and S9). Tissue sections were washed three times and treated with secondary antibodies (Tables S8 and   S9) to detect the expression of the transgene or the induction of an immune reaction. The staining was performed as described by Christopher Kerfoot et al. 16 Fluorescence imaging was carried out on the immunoassayed slices using 20X oil objective under Olympus confocal microscope (Fluoview FV1000).

| DNA extraction
Genomic DNA was extracted from tissue samples using the Phenol/ chloroform DNA extraction technique. 17 DNA from faeces, saliva, nasal secretion and urine was extracted using the Thermo Fisher Scientific GeneJET Genomic DNA Purification Kit (cat# K0722) according to the manufacturer's protocol.

| Quantitative PCR analysis
qtuantitative polymerase chain reaction was performed using Takara

| Statistical analysis
Data are expressed as the mean of vector copy number (Av.) ± standard deviation (SD). One-way ANOVA analysis was carried out. were analysed using the same method. However, to test inter-group differences, the regression analyses used generalized linear models with baseline values as covariate.

| RE SULTS
The main objective of this study was to determine the biodistribution of the vMCO-I in live samples, targeted retina and nontargeted organs after intravitreal injections in wild-type dogs. The schematic of the vector used ( Figure S1) shows that it contains two AAV2 inverted terminal repeats (ITR) that flank the cloning sites.
The study comprised of three groups of dogs with each group consisting of four dogs (two males and two females in each group) at the euthanization time point of 13 weeks (Table S1A,B). The primary end points in this study were the immunohistochemistry examination, transgene expression in the retina and vector genome detection in live body wastes (urine, faeces) and body fluids (saliva and nasal secretions) as well as in non-targeted tissues (lung, liver, kidney, mandibular/mesenteric lymph nodes, heart, spleen and testis/ovary).  Figure 1C. In Figure 1D

| Biodistribution of vMCO-I in live samples
In order to understand the biodistribution of vMCO-I and its clearance, we assayed for its presence in body wastes (urine and faeces) and body fluids (saliva and nasal secretions). Table S2 shows (Table S4).
Detection of vector DNA in dog urines 13 weeks after intravitreal injection in Groups 1, 2 and 3 was carried out. Table S5 shows positive (+ve) vector amplification after intravitreal injection in all three groups. In Figure 6, we show qPCR analysis of vector copy number in urine of dogs intravitreally injected with AAV2 vehicle (Group 1) or vMCO-I (low or high dose). The detected vector copy number in low-dose vMCO-I-injected group (Group 3) was found to be higher than that injected with high dose (Group 2). Further, the vector copy number in Group 2 was lower than the AAV vehicleinjected control group (Group 1).

| Biodistribution of intravitreally injected AAV2 in non-targeted tissues
We assayed for the biodistribution of vMCO-I using genomic DNA  (Table S6). However, as shown in Figure 7A, the vector copy / ng DNA was <2. For the dogs in Group 2 (high-dose vMCO-I), the qPCR signals were detected in the liver, kidney, mandibular lymph node and testis of dog-2001 (Table S6). In dog 2002, the mesenteric lymph node, heart and testis showed signals of the presence of the virus ITR. In dog-2501, lung, kidney, mesenteric lymph node, heart, spleen and ovary showed some positive results. No virus ITR signals were detected in any of the dog-2502 organs (Table S6). In Group 3 (lowdose vMCO-I), in male dogs, the presence of virus ITR was detected in the spleen, and testis; however, in the female group, the virus signal was detected additionally in heart apart from spleen and ovary as per the qPCR analysis (Table S6). Since there was no consistence of +ve amplification in any organ in all the animals of any group, average with standard deviation was plotted and subjected to oneway ANOVA test. As shown in Figure 7A, except mesenteric lymph node, no significant differences were observed between the groups. to be very efficient in gene delivery to vascular endothelial cells. 44,45 In the brain, most AAV serotypes show neuronal tropism, whereas AAV5 also transduces astrocytes. 46 we assessed the biodistribution of the vector in major organs of the intravitreally injected wild-type dogs, as well as body wastes and body fluids. Non-significant levels of the vMCO-I vector ITRs were observed in some dog tissues during qPCR analysis, but the gels from qPCR and PCR did not show any ITR product (~4 kb). This is probably due to the increased sensitivity of qPCR via use of SYBR Green, or formation of primer-dimers and higher (forty) cycle numbers. 57 Nevertheless, qPCR is a sensitive method for demonstrating gene transfer and evaluating biodistribution following AAV-mediated transduction in vivo. 58 In this study, we did not find any significant

CO N FLI C T O F I NTE R E S T
The author Samarendra Mohanty has equity interest in Nanoscope Technologies LLC and Nanoscope Therapeutics Inc. Kissaou Tchedre, Subrata Batabyal, Ananta Ayyagari and Sai Chavala has equity interest in Nanoscope Therapeutics Inc.