Intrapleural administration of AAV9 improves neural and cardiorespiratory function in Pompe disease

Abstract

Pompe disease is a neuromuscular disease resulting from deficiency in acid α-glucosidase (GAA), results in cardiac, skeletal muscle, and central nervous system (CNS) pathology. Enzyme replacement therapy (ERT) has been shown to partially correct cardiac and skeletal muscle dysfunction. However, ERT does not cross the blood-brain barrier and progressive CNS pathology ensues. We tested the hypothesis that intrapleural administration of recombinant adeno-associated virus (rAAV9)-GAA driven by a cytomegalovirus (CMV) or desmin (DES) promoter would improve cardiac and respiratory function in Gaa(-/-) mice through a direct effect and retrograde transport to motoneurons. Cardiac magnetic resonance imaging revealed significant improvement in ejection fraction in rAAV9-GAA-treated animals. Inspiratory phrenic and diaphragm activity was examined at baseline and during hypercapnic respiratory challenge. Mice treated with AAV9 had greater relative inspiratory burst amplitude during baseline conditions when compared with Gaa(-/-). In addition, efferent phrenic burst amplitude was significantly correlated with diaphragm activity in both AAV9-DES and AAV9-CMV groups but not in Gaa(-/-). This is the first study to indicate improvements in cardiac, skeletal muscle, and respiratory neural output following rAAV administration in Pompe disease. These results further implicate a role for the CNS in Pompe disease pathology and the critical need to target the neurologic aspects in developing therapeutic strategies.

Figure 1

Vector genome quantification and retrograde transport of rAAV9 following intrapleural injection. (a) Vector copy number was significantly higher in cardiac, diaphragm, and the spinal cord in rAAV9-CMV or -DES Gaa^−/− mice (P < 0.05). Data are presented as mean ± SEM. Gaa^−/− = sham-treated Gaa^−/−, AAV9-CMV = rAAV9-CMV-hGAA + Gaa^−/−, AAV9-DES = rAAV9-DES-hGAA + Gaa^−/−. Dotted line represents threshold of detection. (b) Immunohistochemical detection of GFP staining in (i.) cervical and (iv.) thoracic spinal cord. Inset represents positive motoneuron (ii.) GFP, and (iii.) GAA staining in the cervical spinal cord and (v.) GFP staining in the thoracic spinal cord. CMV, cytomegalovirus; DES, desmin; GAA, acid α-glucosidase; GFP, green fluorescent protein; rAAV, recombinant adeno-associated virus; vg, vector genome.

Figure 2

GAA activity and glycogen detection following intrapleural injection. GAA activity in the (a) myocardium, (b) diaphragm, and (c) liver 6 months after rAAV9 vector delivery in Gaa^−/− mice (P < 0.05). (d) Representative sections of (i.) Gaa^−/−, (ii.) AAV9-CMV, and (iii.) AAV9-DES stained for glycogen in the diaphragm using PAS staining. Although glycogen (intense black staining) is still visible in AAV9 vector-treated animals, it appears there is a severe reduction in the overall number and size of deposition. Data are presented as mean ± SEM. Gaa^−/− = sham-treated Gaa^−/−, AAV9-CMV = rAAV9-CMV-hGAA + Gaa^−/−, AAV9-DES = rAAV9-DES-hGAA + Gaa^−/−, WT = wild-type (129SVE). *P < 0.05 compared with untreated Gaa^−/−; ^P < 0.05 compared with AAV9-CMV; ^#P < 0.05 compared with AAV9-DES. CMV, cytomegalovirus; DES, desmin; GAA, acid α-glucosidase; PAS, periodic acid-Schiff; rAAV, recombinant adeno-associated virus.

Figure 3

Assessment of cardiac function at 4.7-T. (a) Cardiac magnetic resonance imaging detected improved ejection fraction and (b) stroke volume in rAAV9-treated Gaa^−/− mice (P < 0.05). Gaa^−/− = sham-treated Gaa^−/−, AAV9-CMV = rAAV9-CMV-hGAA + Gaa^−/−, AAV9-DES = rAAV9-DES-hGAA + Gaa^−/−. *P < 0.05 compared with untreated Gaa^−/−. CMV, cytomegalovirus; DES, desmin; GAA, acid α-glucosidase; rAAV, recombinant adeno-associated virus.

Figure 4

The impact of AAV-GAA therapy on respiratory motor output in anesthetized Gaa^−/− mice. (a) Breathing frequency was similar between control, untreated Gaa^−/− mice and those injected with AAV. (b) However, the normalized amplitude of the efferent phrenic inspiratory burst was significantly greater in both AAV9-CMV– and AAV9-DES–treated mice. (c) The efferent diaphragm burst was also normalized to peak output, but this analysis revealed no differences between the groups. Gaa^−/− = sham-treated Gaa^−/−, AAV9-CMV = rAAV9-CMV-hGAA + Gaa^−/−, AAV9-DES = rAAV9-DES-hGAA + Gaa^−/−, WT = wild-type (129SVE). *P < 0.05 compared with untreated Gaa^−/−. CMV, cytomegalovirus; DES, desmin; GAA, acid α-glucosidase; rAAV, recombinant adeno-associated virus.

Figure 5

The relationship between efferent phrenic burst amplitude and diaphragm burst amplitude. Anesthetized mice were studied during spontaneous breathing, and the raw burst amplitudes (i.e., arbitrary units (a.u.)) of the inspiratory phrenic neurograms and diaphragm EMG were analyzed. Untreated Gaa^−/− mice showed no relationship between these variables, but both groups of AAV9-treated mice had a linear relationship. The lines indicate the linear regression as follows: solid line: Gaa^−/−; dotted line: AAV9-DES; dashed line: AAV9-CMV. AAV, adeno-associated virus; CMV, cytomegalovirus; DES, desmin; EMG, electromyogram; GAA, acid α-glucosidase.