Embryonic alteration of motoneuronal morphology induces hyperexcitability in the mouse model of amyotrophic lateral sclerosis

Highlights

• We use embryonic E17.5 SOD1^G93A mouse spinal cords.

• We show that motoneurons are hyperexcitable early in development.

• 3D reconstructions reveal a reduced dendritic arborization.

• Computer simulations indicate that hyperexcitability relies on morphological changes.

Abstract

Although amyotrophic lateral sclerosis (ALS) is an age-dependent fatal neurodegenerative disease in which upper and lower motoneurons (MNs) are targeted for death in adults, increasing lines of evidence indicate that MNs display physiological and morphological abnormalities during postnatal development, long before disease onset. Here, using transgenic mice overexpressing the G93A mutation of the human Cu/Zn superoxide dismutase gene (SOD1), we show that SOD1^G93A embryonic lumbar E17.5 MNs already expressed abnormal morphometric parameters, including a deep reduction of their terminal segments length. Whole-cell patch-clamp recordings from acute spinal cord preparations were made to characterize functional changes in neuronal activity. SOD1^G93A E17.5 MNs displayed hyperexcitability compared to wild-type MNs. Finally, we performed realistic simulations in order to correlate morphometric and electrophysiological changes observed in embryonic SOD1^G93A MNs. We found that the reduced dendritic elongation mainly accounted for the hyperexcitability observed in SOD1^G93A MNs. Altogether, our results emphasize the remarkable early onset of abnormal neural activity in the commonly used animal model for ALS, and suggest that embryonic morphological changes are the primary compensatory mechanisms, the physiological adjustments being only secondary to morphological alterations.

Introduction

Amyotrophic lateral sclerosis (ALS) also known as Lou Gehrig's disease is an age-dependent fatal paralytic disease that results from the degeneration of upper descending cortical neurons and lower motoneurons (MNs) located in the brainstem and spinal cord (Boillee et al., 2006). ALS is the third neurodegenerative cause of adult death after Alzheimer's disease and Parkinson's disease, with an incidence of 1–2 per 100,000 humans that peaks in the sixth decade of life, and a lifetime risk at about 1 in 1000 (Boillee et al., 2006, Pasinelli and Brown, 2006). Usually, respiratory paralysis leads to death 2 to 5 years after onset of the disease. To date, causes of ALS remain elusive. Many hypothesis have been explored to explain ALS pathogenesis including mitochondrial dysfunction (Higgins et al., 2003), axonal transport (De Vos et al., 2007), protein misfolding and endoplasmic reticulum stress (Atkin et al., 2006), and glutamate excitotoxicity (Van Den Bosch et al., 2006).

The majority of the ALS cases are sporadic with unknown etiology, but ~ 10% of ALS patients suffer from a familial form of ALS. If mutations in TAR DNA-binding protein (TDP-43, encoded by the TARDBP gene) (Kabashi et al., 2008, Sreedharan et al., 2008), fused in sarcoma (encoded by the FUS/TLS gene) (Kwiatkowski et al., 2009, Vance et al., 2009) and Cu/Zn superoxide dismutase gene (SOD1) that encodes an antioxidant enzyme (Deng et al., 1993, Rosen et al., 1993) account for approximately 30% of classical inherited ALS, SOD1 mutations represent the most common cases. More than 100 mutations of the SOD1 gene are known in ALS patients (Boillee et al., 2006).

Among these mutations, SOD1^G93A, in which glycine is substituted by alanine at residue 93, has been particularly studied, and a transgenic mouse line overexpressing this mutant human SOD1^G93A has been produced. Interestingly, SOD1^G93A mouse expresses phenotypic and pathological symptoms resembling ALS in humans (Cleveland and Rothstein, 2001, Gurney et al., 1994).

Because a phenotype becomes apparent at quite late stages in the SOD1^G93A mouse ALS model, i.e. abnormal gait at P50 (Wooley et al., 2005), most studies performed on this mouse line have focused at late post-natal asymptomatic stages (Kanning et al., 2010). However, in spite of an absence of clear phenotype, it is possible that early changes accompanied by compensatory mechanisms occur in SOD1 mice as suggested by studies performed during the first two post-natal weeks. Changes include alteration of MN morphology and excitability (Amendola and Durand, 2008, Bories et al., 2007, Pambo-Pambo et al., 2009) as well as transient delays in development of gross locomotor abilities (van Zundert et al., 2008). In addition, data collected from cultured SOD1^G93A embryonic (E13.5) MNs demonstrate defects in axonal transport (De Vos et al., 2007, Kieran et al., 2005), enhanced sensitivity of MNs to Fas- or NO-triggered cell death (Raoul et al., 2002) and impaired mitochondrial dynamics (Magrane et al., 2012), highlighting early developmental pathological features in the SOD1 rodent model (Kanning et al., 2010).

There is increasing evidence that the toxicity of mutant SOD1 is linked to its propensity to misfold and to aggregate (Liu et al., 2012). Based on the use of an antibody (C4F6) specifically recognizing a “toxic” form of the mutant SOD1 protein, a recent study (Brotherton et al., 2012) suggests that a subset of SOD1 protein (misfolded form) accumulates in MNs when they become sick. This study also indicates that C4F6 staining is observed throughout the disease course (presymptomatic through end stage), and even shows staining in certain MNs at early stages (P15). This indicates that toxic form of mutated SOD1 may be present at perinatal stages.

In the present study, by analyzing SOD1^G93A MNs at the late embryonic stage E17.5 when they undergo major developmental changes and when motor spinal networks become functional (Branchereau et al., 2000), we show that SOD1^G93A MNs are hyperexcitable and exhibit a reduced dendritic arborization. Interestingly, using computer simulations, we show that the reduced growth of embryonic SOD1^G93A MNs accounts, for the most part, for their hyperexcitability.