Astrocyte-selective AAV gene therapy through the endogenous GFAP promoter results in robust transduction in the rat spinal cord following injury

Abstract

Adeno-associated viral (AAV) vectors are a promising system for transgene delivery into the central nervous system (CNS) based on their safety profile and long-term gene expression. Gene delivery to the CNS has largely been neuron centric but advances in AAV technology are facilitating the development of approaches to enable transduction of glial cells. Considering the role of astrocytes in the on-going secondary damage in spinal cord injury (SCI), an AAV vector that targets astrocytes could show benefit as a potential treatment. Transduction efficiency, transgene expression and cellular tropism were compared for the AAV serotypes AAV5, AAV9 and AAVRec2 whereby destabilised yellow fluorescent protein (dYFP) was controlled by the GFAP or the truncated GfaABC₁D promoter. The vectors were tested in primary spinal cord astrocyte cell culture, spinal cord slice culture and an in vivo model of SCI contusion. AAV5 resulted in greater transduction efficiency, transgene expression and astrocyte tropism compared with AAV9 and AAVRec2. In a rodent model of SCI, robust transgene expression by AAV5-GFAP/GfaABC₁D-dYFP was observed through 12 mm of spinal cord tissue and expression was largely restricted to astrocytes. Thus, AAV5-GFAP/GfaABC₁D carries the potential as a potential gene therapy vector, particularly for transducing astrocytes in the damaged spinal cord.

Fig. 1

The AAV5 serotype leads to the most efficient transduction and strongest transgene expression in primary spinal cord astrocytes. a Vector constructs containing a dYFP transgene under the control of a GFAP or GfaABC₁D promoter were packaged into AAV serotypes (5, 9 or Rec2) and transduced cultures; AAV5-GFAP-dYFP, AAV5-GfaABC₁D-dYFP, AAV9-GFAP-dYFP, AAVRec2-GFAP-dYFP (4 × 10⁹ vg/well). dYFP gene expression was visualised using fluorescent immunocytochemistry and captured at ×10 magnification using a Nikon Eclipse TE2000-U microscope. Each is a representative image. Scale bar = 200 µm. b The number of dYFP-positive cells per image were counted. Values represent the mean and standard error of the mean (n = 4; independent cultures). Two-way ANOVA followed by Tukey’s multiple comparisons were used to determine statistical significance (not reported on the graph). c dYFP gene expression was visualised using fluorescent immunocytochemistry and images captured at ×10 magnification using a Nikon Eclipse TE2000-U microscope using identical settings and the number of dYFP-positive cells were counted. The integrated densities were determined for each image using ImageJ and used to determine the fluorescent intensity per cell. Values represent the mean and standard error of the mean (n = 4 independent cultures; three technical repeats for each; three images per technical repeat). One-way ANOVA followed by Tukey’s post-hoc analysis *P < 0.05, **P < 0.001

Fig. 2

Organotypic slice cultures transduced with AAV vectors show robust gene expression and AAV5 results in the greatest astrocyte-tropic transductions. Spinal cord were excised from P10 to 15 rat pups, sliced and then cultured. Slices were transduced with AAV vectors: AAV5-GFAP-dYFP, AAV5-GfaABC₁D-dYFP, AAV9-GFAP-dYFP, and AAVRec2-GFAP-dYFP (1 × 10⁹ viral genomes each). Gene expression was developed for 7 days before fixing. Immunohistochemistry was used to visualised dYFP gene expression (green) and GFAP expression (red). a ×10 magnification images were stitched together to display the whole slice. b Images were captured on an EVOS FL Auto imaging station at ×20 magnification to analyse colocalisation of dYFP and GFAP. Images are representative images. c Data are represented as percentage of GFAP-positive dYFP-positive cell ± S.E.M. (n = 4 independent cultures; three technical repeats for each; three images per technical repeat) independent slice cultures. A one-way ANOVA with Tukey’s multiple comparison test was used to determine statistical differences between the vectors, *P < 0.05

Fig. 3

High transgene expression by GFAP-dYFP or AAV5-GfaABC₁D is observed through >12 mm of spinal cord tissue and does not worsen locomotor scores. Rats received 175 KDyne contusive injuries and then immediately received four injections of either AAV5-GFAP-dYFP or AAV5-GfaABC₁D (4 × 10⁹ vg at each). Four weeks later, the animals were euthanized and DAB immunohistochemistry was performed to visualise dYFP transgene expression in the spinal cord tissue. a Images were captured at ×2.5 magnification on a Leica DMR upright microscope. b Images were thresholded and ImageJ was used to determine the percentage of tissue that was immunoreactive for transgene expression. Each bar represents the mean ± SEM of transgene expression within 600 µm blocks from the injury centre (n = 9). c The BBB open field locomotor scale was used to investigate whether the injections had an impact on their locomotion. Data represent the mean ± SEM (sham n = 4; n = 9 for the other groups). A two-way repeated-measure ANOVA followed by a Bonferroni’s multiple comparison was used to determine if there was any statistical difference between contusion groups

Fig. 4

AAV5 vector dYFP expression driven by GFAP or GfaABC₁D promoters results in astrocyte-specific transgene expression. Constructs containing a dYFP reporter transgene driven by either a GFAP or GfaABC₁D promoter were packaged into AAV5 vectors and four infusions of 4 × 10⁹ viral genomes into the contused rodent spinal cord. Tissue was processed for immunohistochemistry to detect dYFP (green) colocalisation with the cell type markers: GFAP, Iba1, Olig2 and NeuN (red). Images were captured at ×60 magnification on an Olympus FV1000 confocal microscope (n = 9). Images presented are representative images. Scale bar = 50 µm