Search

Synaptic plasticity is an important cellular mechanism that underlies memory formation. In brain areas involved in memory such as the hippocampus, long-term synaptic plasticity is bidirectional. Major forms of bidirectional plasticity encompass long-term potentiation (LTP), LTP reversal (depotentiation) and long-term depression (LTD). Protein kinases and protein phosphatases are important players in the induction of both LTP and LTD, and the serine/threonine protein phosphatase-1 (PP1), in particular, has emerged as a key phosphatase in these processes. The goal of the present study was to assess the contribution of PP1 to bidirectional plasticity and examine the impact of a partial inhibition of PP1 on LTP, LTD and depotentiation in the mouse hippocampus. For this, we used transgenic mice expressing an active PP1 inhibitor (I-1*) inducibly in forebrain neurons. We show that partial inhibition of PP1 by I-1* expression alters the properties of bidirectional plasticity by inducing a shift of synaptic responses towards potentiation. At low-frequency stimulation, PP1 inhibition decreases LTD in a frequency-dependent fashion. It favours potentiation over depression at intermediate frequencies and increases LTP at high frequency. Consistently, it also impairs depotentiation. These results indicate that the requirement of bidirectional plasticity for PP1 is frequency-dependent and that a broad range of plasticity is negatively constrained by PP1, a feature that may correlate with the property of PP1 to constrain learning efficacy and promote forgetting.

The brain-specific Ras/Rap GTPase-activating protein synGAP is a major component of the postsynaptic density at glutamatergic synapses. It is a target for phosphorylation by Ca(2+)/calmodulin-dependent protein kinase II, which up-regulates its GTPase-activating activity. Thus, SynGAP may play an important role in coupling N-methyl-D-aspartate-type glutamate receptor activation to signaling pathways downstream of Ras or Rap. Homozygous deletion of synGAP is lethal within the first few days after birth. Therefore, to study the functions of synGAP, we used the cre/loxP recombination system to produce conditional mice mutants in which gradual loss of synGAP begins at approximately 1 week, and usually becomes maximal by 3 weeks, after birth. The resulting phenotypes fall into two groups. In a small group, the level of synGAP protein is reduced to 20-25% of wild type, and they die at 2-3 weeks of age. In a larger group, the levels remain higher than approximately 40% of wild type, and they survive and remain healthy. In all mutants, however, an abnormally high number of neurons in the hippocampus and cortex undergo apoptosis, as detected by caspase-3 activation. The effect is cell autonomous, occurring only in neuronal types in which the synGAP gene is eliminated. The level of caspase-3 activation in neurons correlates inversely with the level of synGAP protein measured at 2 and 8 weeks after birth, indicating that neuronal apoptosis is enhanced by reduction of synGAP. These data show that synGAP plays a role in regulation of the onset of apoptotic neuronal death.

Two types of synaptic depression have been described in the hippocampus, long-term depression and depotentiation of long-term potentiation known to recruit the serine/threonine protein phosphatases PP1, PP2A and PP2B (calcineurin). The contribution of each of these protein phosphatases is controversial. To examine the role of the Ca2+/calmodulin-dependent protein phosphatase calcineurin in long-term depression and depotentiation, we analysed the effect of genetically inhibiting calcineurin reversibly in the hippocampus, using the doxycycline-dependent rtTA system in transgenic mice. We show that reducing calcineurin activity has no effect on long-term depression but reversibly affects depotentiation. Consistently, the calcineurin inhibitor FK-506 reproduces the depotentiation impairment observed in the mutant mice but does not affect long-term depression in control animals. In contrast, the PP1/PP2A inhibitor okadaic acid fully blocks both long-term depression and depotentiation. These data demonstrate that the nature of signalling cascades induced by synaptic activity depends on the initial synaptic state. While depression of potentiated synaptic responses requires activation of PP1/PP2A and/or calcineurin, depression of basal synaptic responses depends only on PP1/PP2A activation.