Multilineage transduction of resident lung cells in vivo by AAV2/8 for α1-antitrypsin gene therapy

Abstract

In vivo gene delivery has long represented an appealing potential treatment approach for monogenic diseases such as α1-antitrypsin deficiency (AATD) but has proven challenging to achieve in practice. Alternate pseudotyping of recombinant adeno-associated virus (AAV) vectors is producing vectors with increasingly heterogeneous tropic specificity, giving researchers the ability to target numerous end-organs affected by disease. Herein, we describe sustained pulmonary transgene expression for at least 52 weeks after a single intratracheal instillation of AAV2/8 and characterize the multiple cell types transduced within the lung utilizing this approach. We demonstrate that lung-directed AAV2/8 is able to achieve therapeutic α-1 antitrypsin (AAT) protein levels within the lung epithelial lining fluid and that AAT gene delivery ameliorates the severity of experimental emphysema in mice. We find that AAV2/8 efficiently transduces hepatocytes in vivo after intratracheal administration, a finding that may have significance for AAV-based human gene therapy studies. These results support direct transgene delivery to the lung as a potential alternative approach to achieve the goal of developing a gene therapy for AATD.

Figure 1

AAV2/8 mediates sustained transgene expression in the mouse abdomen and thorax. (a) A schematic representation of AAV8-CASI-luciferase expression vector backbone is shown. The CASI promoter, firefly luciferase transgene, WPRE, and SV40 late-polyadenylation signal are flanked by AAV2 inverted terminal repeats (ITR). (b) One week after IT administration, bioluminescence was present in both the abdominal and thoracic regions of AAV-luciferase recipients (n = 4, 1 × 10¹¹gc; AAV-GFP-treated mouse shown as negative control, n = 1), with photon flux (mean ± SD) in the abdominal region exceeding that in the thorax. (c) After 52 weeks, abdominal bioluminescence subsided allowing distinct visualization of persistent thoracic bioluminescence. (d,e) The bioluminescence kinetic is quantified as total photon flux (photon/s/cm²) and isolated according to anatomic region (lung or abdomen) as shown over a period of 52 weeks postadministration of AAV8.

Figure 2

Phenotyping of transduced cells following delivery of AAV 2/8 to the mouse lung. (a) A schematic represention of the AAV8-CASI-GFP vector, utilized in these experiments, is depicted. eGFP expression is under the control of the CASI promoter. (b,c) Four weeks after IT AAV8-CASI-GFP administration, frozen left lung sections counterstained with DAPI demonstrate multiple GFP+ cells in the airspaces (b) and distal airways (c). Immunofluorescent staining for CC10 (d), Fox J1 (d), T1α (e), prosurfactant protein C (e) and GFP exhibits colocalization of GFP with CC10, FoxJ1, and prosurfactant protein C. Area of inset images in panels d and e are designated by arrowheads. (f) Frozen lung section with DAPI counterstaining demonstrates a GFP+ resident alveolar macrophage within an alveolus. (g) Representative flow cytometric analysis of cells from bronchoalveolar lavage performed 4 weeks after IT AAV8-CASI-GFP (gated to isolate live alveolar macrophages); aggregate data (not shown) yielded 32% GFP+ cells (n = 4). (h) Representative flow cytometric analysis of whole lung digests 4 weeks after IT AAV8-CASI-GFP (n = 4) demonstrate GFP+ cells among CD45+, CD11c+, EpCAM+, and CD31+ cell populations. (i) Aggregate data from panel H experiment reveals percentages of each cell type found to be transduced (GFP+) by AAV8-CASI-GFP. (j) A frozen liver section at low magnification, 4 weeks after IT AAV8-CASI-GFP, illustrates efficient transduction of hepatocytes.

Figure 3

AAV2/8-based expression of human AAT protein in serum and epithelial lining fluid. (a) Mice administered 1 × 10¹¹gc of AAV8-CASI-hAAT via the IM (n = 4) or IT (n = 4) route underwent serial measurements of serum hAAT concentration (mean ± SD) by ELISA for 28 or 32 weeks, respectively. (b) Mice in each group were harvested at that timepoint for BAL. ELF hAAT concentrations were then quantified by ELISA.

Figure 4

Effects of AAV8-hAAT on lung function and morphometry in the setting of elastase-induced lung injury. (a,b) hAAT concentration within the serum and lung epithelial lining fluid 7 weeks after IT AAV8-CASI-hAAT or AAV8-CASI-GFP, and 3 weeks after porcine pancreatic elastase (PPE) or control vehicle. Dashed line represents theoretical therapeutic threshold hAAT concentration (^***P < 0.001, *P < 0.05). (c) Representative axial sections of paraffin-embedded, H&E-stained lung tissue according to vector and PPE treatment group. Additional images are shown in Supplementary Figure S6. (d–f) Airspace size is illustrated as equivalent alveolar diameter (Deq), maximum Deq (Max Deq), and area-weighted mean alveolar diameter (D₂). (g) Lung compliance across a range of PEEP settings was quantified in each treatment group (^**P < 0.01, see two-way analysis of variance, Supplementary Table S1). (h) Demonstration of the relationship between D2 and lung compliance.