Peripheral administration of a serine protease inhibitor blocks kindling

Evidence for disrupted NMDA receptor function in tissue plasminogen activator knockout mice

Tissue plasminogen activator (tPA), a serine protease immediate-early gene product expressed in brain areas important in learning and memory, has been shown to cleave the NR1 subunit of the NMDA receptor leading to a potentiated Ca(2+) influx. Mice lacking tPA (tPA-/- mice) have disrupted late phase-LTP in the hippocampus, possibly as a consequence of reduced Ca(2+) flux through NMDA receptors. In the present experiments, we investigated whether the NMDA antagonist dizocilpine might alter performance in tPA-/- mice in behavioural tasks shown to be sensitive to hippocampal lesions. tPA-/- mice and wild-type controls (WT) showed similar rates of acquisition and performance of a spatial working memory task (eight-arm radial maze). Dizocilpine (0.03-0.3 mg/kg, i.p.), given acutely, disrupted performance by increasing the number of errors equally across both genotypes. At asymptotic performance of a differential reinforcement of low response rate operant task (DRL), acute dizocilpine (0.03-0.3 mg/kg) impaired performance, but no differences between genotypes were observed. However, dizocilpine (0.1 mg/kg), given repeatedly during acquisition of a signalled-DRL15" task, retarded acquisition in tPA-/- but not WT mice. This treatment regime had no effect on locomotor activity in either genotype. tPA-/- mice showed no spatial learning deficits, but were more sensitive to dizocilpine during acquisition (though not expression) of a DRL task. This supports a role for tPA in modification of the NMDA receptor, although absence of tPA does not have consequences for all forms of NMDA-dependent mediated learning.

Integrins, synaptic plasticity and epileptogenesis

A number of processes are thought to contribute to the development of epilepsy including enduring increases in excitatory synaptic transmission, changes in GABAergic inhibition, neuronal cell death and the development of aberrant innervation patterns in part arising from reactive axonal growth. Recent findings indicate that adhesion chemistries and, most particularly, activities of integrin class adhesion receptors play roles in each of these processes and thereby are likely to contribute significantly to the cell biology underlying epileptogenesis. As reviewed in this chapter, studies of long-term potentiation have shown that integrins are important for stabilizing activity-induced increases in synaptic strength and excitability. Other work has demonstrated that seizures, and in some instances subseizure neuronal activity, modulate the expression of integrins and their matrix ligands and the activities of proteases which regulate them both. These same adhesion proteins and proteases play critical roles in axonal growth and synaptogenesis including processes induced by seizure in adult brain. Together, these findings indicate that seizures activate integrin signaling and induce a turnover in adhesive contacts and that both processes contribute to lasting changes in circuit and synaptic function underlying epileptogenesis.

Evidence that integrins contribute to multiple stages in the consolidation of long term potentiation in rat hippocampus

Three structurally distinct groups of antagonists were used to test the hypothesis that integrin adhesion receptors play an essential role in consolidating (stabilizing) long term potentiation of the Schaffer collaterals in rat hippocampus. Comparisons were made of percent potentiation at antagonist-treated versus control sites within CA1 stratum radiatum of the same hippocampal slice. Function blocking antibodies against the alpha5 subunit of the fibronectin receptor had no effect on baseline responses or initial potentiation but resulted in a >30% reduction, relative to within-slice control long term potentiation, 45 min later. Larger reductions were recorded in separate experiments continued for 4 h after the induction of potentiation. Alpha(v) and alpha2 subunit antibodies did not reliably affect the stabilization of potentiation. An antagonist peptide with preference for beta1 integrins produced a slowly developing decline of the type seen with alpha5 antibodies. A cyclic peptide antagonist reduced potentiation within 10 min of induction and caused an almost 40% decrease over 45 min. Two disintegrins (snake toxins that potently block integrins) were very effective in preventing the consolidation of long term potentiation: echistatin reduced potentiation by >70%, while triflavin caused approximately 50% decrease. The suppressing effects of echistatin were concentration-dependent, obtained with treatment after induction, and much more rapid than the effects of antibodies. Rapid declines in potentiation were particularly evident when the two disintegrins were applied together. These results indicate that hippocampal fibronectin receptors (alpha5/beta1 integrin) contribute importantly to a slowly developing phase of long term potentiation consolidation. They also suggest that other integrins are critical to aspects of consolidation occurring in the first few minutes after induction.

DRL performance in mice with deletion of tPA, uPA or PAI-1 genes

Plasminogen activators are serine proteases induced in the brain by electrical activity leading to synaptic remodelling. They are classified into two distinct subtypes, tissue plasminogen activating factor and urokinase plasminogen activating factor (tPA and uPA, respectively), which are both expressed in brain areas thought to be important in learning and memory. Plasminogen activator inhibitor-1 (PAI-1) is the primary inhibitor of tPA and uPA activity, and is expressed in corresponding brain areas. Mice lacking tPA show a deficit in the acquisition of a 15 s differential reinforcement of low rate of responding (DRL15") task relative to their wild types (WTs) under certain conditions. The current set of experiments were designed to investigate further the role of tPA and to extend our knowledge to uPA and PAI-1, using mice with the respective genes deleted (uPA -/- and PAI-1 -/- mice) in the DRL15" task. uPA -/- mice showed no disruption of DRL acquisition, but PAI-1 -/- mice showed a deficit similar to that seen in tPA -/- mice. In an attempt to compensate for this deficit, experiments using a fixed number of reinforcers or a signalled-DRL15" schedule, similar to that used in rat lesion studies of DRL, were performed. tPA -/- mice were able to complete the signalled-DRL task as well as their WTs, and, similarly, PAI-1 -/- mice were able to learn the fixed-number-of-reinforcers-DRL15" schedule and the signalled-DRL schedule. These data indicate that uPA deletion does not affect performance of a standard DRL15" task, whereas deletion of PAI-1 has the same behavioural consequences in these tasks as deletion of tPA. Deficits of both genotypes can be attenuated by providing either external information on completion of the delay or by equalizing the number of reinforcers obtained.

Delayed and isoform-specific effect of NMDA exposure on neural cell adhesion molecules in hippocampus

Brief stimulation of N-methyl-D-aspartate (NMDA) receptors has been shown to generate proteolytic fragments from the extracellular domain of neural cell adhesion molecules (NCAMs). In the present study, hippocampal slice cultures were used to demonstrate that such brief stimulation is followed by a delayed increase in the 180-kDa isoform NCAM-180. The slices were exposed to NMDA for 30 s followed by rapid quenching with the antagonist AP5. Immunoassays of the experimental samples indicated that concentrations of NCAM-180 were elevated above matched controls 2-3 h after the NMDA exposure, but not at earlier or later time points. This effect was isoform-specific as concentrations of the 140-kDa NCAM species were not found to increase. Interestingly, similar selectivity was evident with prolonged infusions of NMDA where, in contrast to the effect of brief stimulation, NCAM-180 content was reduced to 50% while levels of NCAM-140 were unchanged. Together with previous findings, the data indicate that the synaptic chemistries activated by NMDA differentially regulate NCAM-180 at the translation level and by localized activation of proteases.