Intraparenchymal convection enhanced delivery of AAV in sheep to treat Mucopolysaccharidosis IIIC

Abstract

Background: Mucopolysaccharidosis IIIC (MPSIIIC) is one of four Sanfilippo diseases sharing clinical symptoms of severe cognitive decline and shortened lifespan. The missing enzyme, heparan sulfate acetyl-CoA: α-glucosaminide-N-acetyltransferase (HGSNAT), is bound to the lysosomal membrane, therefore cannot cross the blood-brain barrier or diffuse between cells. We previously demonstrated disease correction in MPSIIIC mice using an Adeno-Associated Vector (AAV) delivering HGSNAT via intraparenchymal brain injections using an AAV2 derived AAV-truetype (AAV-TT) serotype with improved distribution over AAV9.

Methods: Here, intraparenchymal AAV was delivered in sheep using catheters or Hamilton syringes, placed using Brainlab cranial navigation for convection enhanced delivery, to reduce proximal vector expression and improve spread.

Results: Hamilton syringes gave improved AAV-GFP distribution, despite lower vector doses and titres. AAV-TT-GFP displayed moderately better transduction compared to AAV9-GFP but both serotypes almost exclusively transduced neurons. Functional HGSNAT enzyme was detected in 24-37% of a 140g gyrencephalic sheep brain using AAV9-HGSNAT with three injections in one hemisphere.

Conclusions: Despite variabilities in volume and titre, catheter design may be critical for efficient brain delivery. These data help inform a clinical trial for MPSIIIC.

Keywords: AAV gene therapy; Convection Enhanced Delivery; HGSNAT; Large animal; Mucopolysaccharidosis.

Fig. 1.

Experimental protocol for AAV distribution studies. A Overview of treatment protocol and tissue analysis. B Different flow rates in 0.4% agarose gel phantoms. Experimental groups used in study 1 C and study 2 D. E Schematic of injection sites for Corpus Striatum (CS), White Matter (WM), Subcortical zone (SC) and Brainstem (BS) from a lateral and superior view. F Syringe driver used for AAV delivery. G Example surgical screenshot from Brainlab showing target accuracy. H Schematic of mediolateral sagittal brain levels 1-5 used for immunohistochemical analysis

Fig. 2

Comparative distribution of GFP expression in sheep brains in study 1. Images showing GFP expression in sheep brains 3 weeks post-injection. Sheep received AAV9-GFP (n=2/group) either into A four locations: the corpus striatum (CS), white matter (WM), subcortical zone (SC) and brainstem (BS), B the lateral ventricles (ICV), (C, D) CS and (E, F) WM with the latter two groups separated into two infusion flow rates (0.5 µL/min and 1 μL/min). G Average GFP expression across each sagittal brain level. H Average GFP expression across the entire hemisphere. Data are presented as mean ± s.e.m

Fig. 3

GFP expression from different injection sites in sheep brains in study 1. Images revealing the pattern of GFP expression 3 weeks post AAV9-GFP injection at 10× magnification. After delivery into the lateral ventricles transduced cells can be visualised on the A dorsal and B ventral surface of the ventricle. GFP transduced cells can be visualised in the C striatum, D medial geniculate nucleus , E motor cortex, F brainstem and G thalamus after direct injection into the CS. For WM injections, GFP positivity was detected in the H WM and the I neurons adjacent to the injection site. Scale bar = 100 µm

Fig. 4

Comparative distribution of GFP expression in sheep brains in study 2. Images showing GFP expression in sheep brains 3 weeks post-injection. Two groups received either A AAV9-GFP (n=3) or B AAV-TT GFP (n=2) into the corpus striatum (CS) only. C An additional group of sheep (n=2) received AAV-TT-GFP into the CS, white matter (WM) and brainstem (BS). D Average GFP expression across each sagittal brain level. E Average GFP expression across the entire hemisphere. Data are presented as mean ± s.e.m

Fig. 5

GFP expression from serotypes in sheep brains in study 2. Images revealing the pattern of GFP expression 3 weeks post AAV-GFP injection at 10× magnification. AAV9-GFP injection into the corpus striatum (CS) displays GFP transduced cells in the A striatum and B motor cortex. AAV-TT-GFP transduced cells can be visualised in the C striatum, D motor cortex, E medial geniculate nucleus and F thalamus. Scale bar = 100 µm

Fig. 6

AAV vectors target neurons in the corpus striatum of the sheep brain. Double staining of GFP+ (green) with NeuN+ neurons (magenta) in A AAV-TT-GFP and B AAV9-GFP treated sheep at 10× magnification. Arrows in the merged image indicate examples of co-localization of GFP+ and NeuN+ cells. Scale bar = 100 µm

Fig. 7

Immune reactions pre- and post-injection. Total IgG antibody responses against AAV9 and AAV-TT capsid proteins measured by ELISA on day 0 (the day of surgery) and 3 weeks post-surgery (at necropsy) in study 1 (A, B) and study 2 (C–J). A number of sheep showed pre-existing antibodies to the AAV-capsids that were still evident at necropsy. D One sheep developed an immune response to the AAV capsid. J One sheep (highlighted in red) was sacrificed on the day of surgery due to an adverse event but had no IgG antibodies against the capsid

Fig. 8

Comparison of HGSNAT expression in AAV9-HGSNAT-treated sheep 3 weeks post-surgery. Vector distribution is represented by three-dimensional schematic drawings of the brain. Each brain was divided into 1cm³ cubes. Quantitative comparison of HGSNAT activity above endogenous background level was assessed (A, C). HGSNAT values were determined as fold increase compared to the mean of 2 un-injected sheep for that relative brain section and VCN in two different sheep O95 and Y242 (C, D)