UPR activation specifically modulates glutamate neurotransmission in the cerebellum of a mouse model of autism.

An increasing number of rare mutations linked to autism spectrum disorders have been reported in genes encoding for proteins involved in synapse formation and maintenance, such as the post-synaptic cell adhesion proteins neuroligins. Most of the autism-linked mutations in the neuroligin genes map on the extracellular protein domain. The autism-linked substitution R451C in Neuroligin3 (NLGN3) induces a local misfolding of the extracellular domain, causing defective trafficking and retention of the mutant protein in the endoplasmic reticulum (ER). The activation of the unfolded protein response (UPR), due to misfolded proteins accumulating in the ER, has been implicated in pathological and physiological conditions of the nervous system. It was previously shown that the over-expression of R451C NLGN3 in a cellular system leads to the activation of the UPR. Here, we have investigated whether this protective cellular response is detectable in the knock-in mouse model of autism endogenously expressing R451C NLGN3. Our data showed up-regulation of UPR markers uniquely in the cerebellum of the R451C mice compared to WT littermates, at both embryonic and adult stages, but not in other brain regions. Miniature excitatory currents in the Purkinje cells of the R451C mice showed higher frequency than in the WT, which was rescued inhibiting the PERK branch of UPR. Taken together, our data indicate that the R451C mutation in neuroligin3 elicits UPR in vivo, which appears to trigger alterations of synaptic function in the cerebellum of a mouse model expressing the R451C autism-linked mutation.

Nitric oxide produced by non-motoneuron cells enhances rat embryonic motoneuron sensitivity to excitotoxins: comparison in mixed neuron/glia or purified cultures.

The present study compares the sensitivity to chronic exposure to glutamate agonists of SMI-32-positive rat-derived embryonic motoneurons under both mixed neuron/glia and purified cultures. We found that in spite of a trophic role of glia on cultured motoneurons, SMI-32-positive cells are more sensitive to excitotoxicity in the presence of glia than in purified culture, very likely through nitric oxide released by non-neuronal cells. The rank order of potency for inducing toxicity after 48 h incubation was AMPA>kainate>NMDA, with EC(50): 0.43, 4.9 and 49 microM, respectively, in mixed neuron/glia culture and 14, 32 and 135 microM in purified cultures. The effect of NMDA was dose-dependently potentiated by glycine, with similar potency in the two culture conditions. The effect of agonists was completely antagonized by the specific antagonists CNQX, BNQX and MK801 in both culture conditions. Motoneurons were similarly immunoreactive to NR1 and GluR2 antibodies under both mixed neuron/glia and purified cultures, thus confirming the presence of the calcium-impermeant AMPA receptor subtypes and of the obligatory subunit for NMDA receptors. The effect of kainate in mixed neuron/glia culture was reduced by the addition of 40 microM N-nitro-L-arginine or L-NAME, which shifted the EC(50) to 9 microM. By contrast, L-NAME did not modify the effect of kainic acid in purified cultures. These results suggest that the release of nitric oxide by non-neuronal cells in culture enhances glutamate excitotoxicity in SMI-32-positive cells, and that direct activation of ionotropic glutamate receptors is not enough to explain the mechanism of chronic motoneuron degeneration occurring in vivo in amyotrophic lateral sclerosis (ALS).

Biochemical and pharmacological evidence of a functional role of AMPA receptors in motor neuron dysfunction in mnd mice.

We studied ionotropic glutamate receptor subtypes and the effect of chronic treatment with NBQX [6-nitro-7-sulphamoyl-benzo(F)quinoxaline-2,3-dione], a selective (rs)-alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor antagonist, in the spinal cord of mnd mice. NBQX (8 mg/kg daily i.p. for 3 weeks starting from 24 weeks old) significantly improved the behavioural scores (hind leg extension reflex, cage rung grasping and gait) in mnd mice, measured after the last drug injection, and increased the number of mice with 'normal' gait (from 50% to 90%, P < 0.05). Receptor binding autoradiography of the competitive N-methyl-D-aspartate (NMDA) antagonist, [3H]CGP 39653, of [3H]AMPA and [3H]kainic acid in spinal cord sections, measured after 1 week of drug washout, were not significantly different in control and mnd mice, and were not modified by NBQX. GluR2/3 immunoreactivity, assessed using Western blotting, was significantly enhanced (by 59%, P < 0.01) in the spinal cord but not in the brain of 28-week-old mnd mice compared to age-matched control mice. NBQX treatment increased GluR2/3 immunoreactivity in the spinal cord of control mice and mnd mice by 327 +/- 74% (P < 0.01) and 212 +/- 52% (P < 0.01), respectively. The changes in GluR2/3 subunits may involve adaptive mechanisms of the receptor and play some role in the protective effect of NBQX. These findings suggest that selective antagonism of ionotropic non-NMDA receptors may be of value in the treatment of motor neuron disease.

Tumor necrosis factor is increased in the spinal cord of an animal model of motor neuron degeneration.

Autoimmunity and oxidative/excitotoxic damage are considered as possible pathogenetic mechanisms in amyotrophic lateral sclerosis (ALS). As tumor necrosis factor (TNF) is implicated in autoimmune diseases, including experimental autoimmune encephalomyelitis, and can be neurotoxic, we studied TNF production in a proposed animal model of ALS, the mnd mouse. These mice develop symptoms (progressive weakness of the limbs) as late as at 7 months of age. We measured TNF in serum, brain and spinal cord of mnd mice at 3 and 7 months of age. TNF was detectable in the brain and spinal cord (but not in the serum) at 7 months, while no TNF was detected in mnd mice at 3 months (asymptomatic) or in control mice of the same genetic background and the same age. Immunohistochemistry confirmed localization of TNF-alpha in motor neurons situated in the ventral horn of the spinal cord of 7-month old mnd mice. These results suggest the possibility of testing inhibitors of TNF production in this disease.

Spinal cord GLT-1 glutamate transporter and blood glutamic acid alterations in motor neuron degeneration (Mnd) mice.

This study characterizes for the first time neurochemical mechanisms in Mnd mice, initially described as a model of motor neuron disease and more recently proposed as a model for neuronal ceroid lipofuscinosis. A selective decrease (-30%) of [3H]glutamate uptake was found in spinal cord but not cortical synaptosomes of Mnd mice aged 28 weeks, when they show histopathological alterations, complete blindness and moderate neurological deficits. In spite of the widespread presence of stored material in neurons in many brain regions and spinal cord, the active transport of [3H]serotonin, [3H]dopamine and depolarization-induced [3H]serotonin release were not affected. Spinal EAAC1 glutamate transporter protein was significantly decreased in some but not all aged mice by 36% on average, possibly due to the loss of motor neurons. GLT-1 immunoreactivity was reduced by 34% in 28-week-old Mnd mice, while GLAST immunoreactivity was not affected. In Mnd mice aged 14 weeks, when there was no apparent alteration of motor function, the defect in the glial transporter protein GLT-1 was similar to that in 28-week-old mice (25%). Blood glutamic acid concentration was increased in Mnd mice aged 14-22 weeks. We suggest that the early decrease of GLT-1 protein might raise the extrasynaptic glutamic acid concentration, and contribute to the loss of motor neurons in affected mice, resulting in low [3H]glutamate uptake, low EAAC1 immunoreactivity and neurological deficits.