Mechanisms of Neuronal Death: An in vivo Study in the Lurcher Mutant Mice

Summary

Lurcher is a gain-of-function point mutation located in the gene encoding the δ ₂ subunit of glutamate receptors (GRID2). The Lurcher mutation is lethal when homozygous. Heterozygous mice are ataxic due to a massive neuronal loss in their cerebellum. Lurcher Purkinje cells expressing the mutated allele are depolarized and die from the second postnatal week onwards, suggesting an excitotoxic process. Target-related cell death affects more than 90% of granule cells and 60–75% of olivary neurons, the two Purkinje cell afferences.

Thus, the Lurcher heterozygous mouse is an ideal model to study in vivo the mechanism of two types of neuronal death: an excitotoxicity-like process and target-related neuronal death. The timing of Purkinje cell death onset in Lurcher mice, around P10, is concomitant with the beginning of synaptogenesis between parallel fiber and Purkinje cell, suggesting a potential role of granule cell-Purkinje cell interaction in the timing of Purkinje cell death. X-irradiation of Lurcher mice during granule cell genesis is a means to reduce granule cell number. In these mice, Purkinje cells degenerate with the same timing as in Lurcher controls, suggesting that granule cells do not influence this process although they differentiate surprisingly better than in non-irradiated mutants.

The molecular cascade leading to apoptosis, a particular type of cell death, has been well defined, especially in vitro. Two families of proteins have an essential role in the regulation of apoptosis: the Bcl-2 family and the caspases. TUNEL-labeling studies have suggested the involvement of apoptosis in both types of neuronal death affecting the Lurcher nervous system.

The caspases are structurally similar cystein proteases that cleave their substrates specifically after an aspartate residue. They are synthesized as a precursor that is activated by cleavage, resulting in the formation of a large and a small subunit. Two heterodimers then associate to form the final active protease. Three categories of caspases can be distinguished by the specificity of their substrate cleavage site: caspases generating mature proinflammatory cytokines (caspase-1, -4, -5), effector caspases (caspase-2, -3, -7 and CED3) and initiator caspases (caspase-6, -8, -9) of apoptosis. The analysis of caspase-3 expression has shown an up-regulation of pro-caspase-3 in 25% of Purkinje cells of Lurcher mice. Activation of this caspase was also detected in 1–3% of these cells, as was TUNEL labeling. Dying granule cells and olivary neurons also contained activated caspase-3. These results further suggest the involvement of an apoptotic process in the two types of neuronal death occurring in Lurcher mice.

The Bcl-2 family contains both pro- and anti-apoptotic members. Overexpression of the anti-apoptotic protein Bcl-2 in Lurcher mice does not rescue Purkinje cells but is able to delay this process, as Purkinje cells can still be found in two-month-old Lurcher mice overexpressing Bcl-2. We further analyzed the involvement of the Bcl-2 family by studying the role of the proapoptotic protein Bax. Bax up-regulation has been shown in both Purkinje cells and granule cells in Lurcher mice. To analyze the effect of Bax inactivation, we generated double-mutants, i.e. Bax knock-out Lurcher mice. One-month-old animals had a 40% increase in granule cell number. This increase was still observed in two-month-old animals, showing that Bax inactivation persistently inhibited the target-related death of granule cells. However, olivary neuron degeneration was not prevented in Bax knock-out Lurcher mice, showing that Bax involvement in target-related cell death depends on the neuronal population. In one-month-old animals, Purkinje cell number was the same in Bax knock-out Lurcher mice and in Lurcher controls. However, an increased number of Purkinje cells is detected in P15 Bax knock-out Lurcher mice. Thus, Bax inactivation is not sufficient to inhibit Purkinje cell death induced by the Lurcher mutation, but it is able to delay this process for a short period. In the Bax knock-out Lurcher mice, caspase-3 activation is inhibited in both Purkinje cells and granule cells, whereas pro-caspase-3 up-regulation in Purkinje cells is not influenced. Granule cell rescue in this model can be correlated to the inhibition of caspase-3 activation. Interestingly, the inhibition of caspase-3 activation is not sufficient to rescue Purkinje cells, suggesting that another pathway, for example another caspase or a caspase-independent mechanism, is able to mediate Lurcher Purkinje cell death.

The study of Lurcher mice highlights the point that different pathways underlie neuronal death depending on the death stimuli and also on the neuronal population. Different proteins in granule cells and olivary neurons mediate target-related neuronal death. Moreover, Lurcher Purkinje cells express different apoptosis-inducing molecules, activated caspase-3 and Bax, but are not rescued by their inhibition, suggesting that in one cell type several pathways of cell death can be induced by one stimulus.