Radioiodinated Capsids Facilitate In Vivo Non-Invasive Tracking of Adeno-Associated Gene Transfer Vectors

Abstract

Viral vector mediated gene therapy has become commonplace in clinical trials for a wide range of inherited disorders. Successful gene transfer depends on a number of factors, of which tissue tropism is among the most important. To date, definitive mapping of the spatial and temporal distribution of viral vectors in vivo has generally required postmortem examination of tissue. Here we present two methods for radiolabeling adeno-associated virus (AAV), one of the most commonly used viral vectors for gene therapy trials, and demonstrate their potential usefulness in the development of surrogate markers for vector delivery during the first week after administration. Specifically, we labeled adeno-associated virus serotype 10 expressing the coding sequences for the CLN2 gene implicated in late infantile neuronal ceroid lipofuscinosis with iodine-124. Using direct (Iodogen) and indirect (modified Bolton-Hunter) methods, we observed the vector in the murine brain for up to one week using positron emission tomography. Capsid radioiodination of viral vectors enables non-invasive, whole body, in vivo evaluation of spatial and temporal vector distribution that should inform methods for efficacious gene therapy over a broad range of applications.

Figure 1. Modified Bolton-Hunter reagent labeling.

The reaction of diaryliodonium trifluoromethane-sulfonate salt with [¹²⁴I] NaI in acetonitrile/toluene solvent matrix produced an I-124 labeled modified Bolton-Hunter intermediate.

Figure 2. Gel electrophoresis assay to ascertain the identity of the labeled virus particle.

Samples were heated to 80 °C for 3 minutes causing capsid disintegration, resulting in three bands corresponding to VP1, VP2, and VP3 capsid protein constituents. The purification process removed unreacted free iodide as well as aggregates formed during radiolabeling.

Figure 3. Activity of I-124AAVrh.10CLN2 vector in vitro.

293ORF6 cells in a 12-well plate (4 × 10⁵) were infected with I-124AAVrh.10CLN2 (10,000 genome copies/cell). At 72 hours post infection cells were harvested and a cell lysate was prepared (200 μl). Kinetics of TPP-1 activity were assessed with 20 μl every 10 min for 60 min using a Cytofluor 4000TC plate reader. The plate was read from the bottom using 360/20 nm excitation and 460/25 emission filters. The gain was set at 70. Uninfected cell lysate (Mock) served as the control. TPP-1 activity is expressed as mean fluorescence unit/min/mg protein ± standard error.

Figure 4. Immunohistochemical assessment of human TPP-1 expressed by I-124AAVrh.10CLN2 using the Iodogen method.

Images are sequential coronal sections 5 μm apart and obtained on sacrifice, 4.5 weeks after vector administration. TPP-1 staining is observed in the cortex and striatum with halo-like patterns of cell bodies (left) as compared to background staining of the neighboring section using an irrelevant IgG (right).

Figure 5. PET/CT data acquired using I-124AAVrh.10CLN2 prepared using the Iodogen method.

The five images in the left column (Mouse 1) demonstrate binding of I-124 to the vector for several days after administration, while mice injected with free iodine exhibited rapid clearance from the brain and accumulation in the thyroid (Mouse 2, upper right). In the lower right panel (Mouse 3) vector has begun to spread beyond the injection site by Day 3.

Figure 6. Whole brain time activity curves for I-124 using the Iodogen (N = 12, circles) and modified Bolton-Hunter (N = 6, triangles) methods, compared with free I-124 (N = 6, squares). Different colors represent individual mice.