GLT1 overexpression in SOD1G93A mouse cervical spinal cord does not preserve diaphragm function or extend disease

• AAV8–Gfa2–GLT1was injected into cervical spinal cord ventral horn of SOD1G93A mice.
• AAV8–Gfa2–GLT1 increased intraspinal GLT1 expression several-fold in SOD1G93A mice.
• AAV8-based GLT1 overexpression did not protect respiratory phrenic motor neurons.
• GLT1 overexpression did not preserve diaphragm innervation or function.
• GLT1 overexpression did not delay disease onset, extend disease duration or prolong survival.

Amyotrophic lateral sclerosis (ALS) is characterized by relatively rapid degeneration of both upper and lower motor neurons, with death normally occurring 2–5 years following diagnosis primarily due to respiratory paralysis resulting from phrenic motor neuron (PhMN) loss and consequent diaphragm denervation. In ALS, cellular abnormalities are not limited to MNs. For example, decreased levels and aberrant functioning of the major central nervous system (CNS) glutamate transporter, GLT1, occur in spinal cord and motor cortex astrocytes of both humans with ALS and in SOD1G93A rodents, a widely studied ALS animal model. This results in dysregulation of extracellular glutamate homeostasis and consequent glutamate excitotoxicity, a primary mechanism responsible for MN loss in ALS animal models and in the human disease. Given these observations of GLT1 dysfunction in areas of MN loss, as well as the importance of testing therapeutic strategies for preserving PhMNs in ALS, we evaluated intraspinal delivery of an adeno-associated virus type 8 (AAV8)–Gfa2 vector to the cervical spinal cord ventral horn of SOD1G93A ALS mice for focally restoring intraspinal GLT1 expression. AAV8 was specifically injected into the ventral horn bilaterally throughout the cervical enlargement at 110 days of age, a clinically-relevant time point coinciding with phenotypic/symptomatic disease onset. Intraspinal delivery of AAV8–Gfa2–GLT1 resulted in robust transduction primarily of GFAP+ astrocytes that persisted until disease endstage, as well as a 2–3-fold increase in total intraspinal GLT1 protein expression in the ventral horn. Despite this robust level of astrocyte transduction and GLT1 elevation, GLT1 overexpression did not protect PhMNs, preserve histological PhMN innervation of the diaphragm NMJ, or prevent decline in diaphragmatic respiratory function as assessed by phrenic nerve–diaphragm compound muscle action potential (CMAP) recordings compared to control AAV8–Gfa2–eGFP injected mice. In addition, AAV–Gfa2–GLT1 did not delay forelimb disease onset, extend disease duration (i.e. time from either forelimb or hindlimb disease onsets to endstage) or prolong overall animal survival. These findings suggest that focal restoration of GLT1 expression in astrocytes of the cervical spinal cord using AAV delivery is not an effective therapy for ALS.