Zhang C, Wu B, Beglopoulos V, Wines-Samuelson M, Zhang D, Dragatsis I, Südhof TC, Shen J.
Presenilins are essential for regulating neurotransmitter release.
Nature. 2009 Jul 30;460(7255):632-6.
Please login to recommend the paper.
To make a comment you must login or register.
A number of previous studies indicated that FAD mutations and genetic deletions of presenilins are linked with abnormal ER Ca2+ homeostasis (reviewed in 1,2). These two new papers provide a first look at changes in synaptic function resulting from such ER Ca2+ alterations. Chakrobotry at al. demonstrate that PS1-M146V mutation results in dramatically (12-fold) enhanced Ca2+ responses mediated by ryanodine receptors (RyanR) in the spines of hippocampal neurons from 3xTg mice. These effects are much more dramatic than the approximately twofold increase in RyanR-mediated responses in the soma of 3xTg neurons. The potential explanation of these results may be related to increased expression of RyanR2 in 3xTg mice (as suggested by the authors) or to overfilled ER Ca2+ stores due to loss of ER Ca2+ leak function resulting from PS1-M146V mutation (3).
Perhaps the most interesting findings in the paper resulted from studies of hippocampal synaptic plasticity in 3xTg mice. The short-term form of presynaptic plasticity (paired-pulsed facilitation, PPF) was normal in 3xTg mice when compared to wild-type mice. Long-term potentiation (LTP) was also normal. At the first glance, these results seem to indicate that changes in ER Ca2+ homeostasis induced by PS1-M146V mutation have no effect on short-term and long-term plasticity. However, when the RyanR was blocked by application of dantrolene, very significant differences emerged. Dantrolene had no effect on PPF and LTP in wild-type mice, but in 3xTg mice application of dantrolene increased PPF and reduced LTP. These observations suggest that the RyanR-mediated ER Ca2+ release is playing a much more important role in presynaptic and postsynaptic spines in 3xTg mice than in WT mice.
In the paper by Zhang at al., sophisticated genetic methods were used to dissect the contribution of presenilins to ER Ca2+ signaling and synaptic function on presynaptic and postsynaptic sides of the hippocampal synapse. Using CA1-specific (postsynaptic) and CA3-specific (presynaptic) Cre-driver lines crossed with PS1floxed/floxed; PS2-/- mice, these investigators discovered that hippocampal LTP, PPF, and high-frequency facilitation are impaired when presenilins are deleted presynaptically (CA3-Cre deletion), but not postsynaptically (CA1-Cre deletion). It should be noted here that PPF and high-frequency facilitation are presynaptic forms of short-term plasticity, so it is not surprising that the phenotype in these experiments has been observed in “presynaptic” but not in “postsynaptic” presenilin knockouts. Deficiency in high-frequency facilitation could be rescued by an increase in extracellular Ca2+ concentration and occluded in the presence of ER SERCA Ca2+ pump inhibitor thapsigargin or RyanR inhibitors ryanodine or dantrolene, but not in the presence InsP3R inhibitor Xestospongin C. These results indicate that defects in high-frequency facilitation in these mice are related to changes in RyanR-mediated Ca2+ release from the presynaptic ER.
In summary, both papers indicate that PS1-M146V FAD point mutation in presenlin-1 or genetic knockout of presenlin-1 in presynaptic neurons results in imbalance in ER Ca2+ handling and changes in RyanR-mediated Ca2+ release. Resulting differences in ER Ca2+ signaling are manifested as changes in short-term (PPF) and long-term (LTP) form of hippocampal synaptic plasticity. The mechanistic basis for these changes is not well understood, but may be linked to a loss of ER Ca2+ leak function of presenilins (3) resulting from PS1-M146V mutation (Chakrobotry et al.) or genetic deletion of PS1 (Zhang et al.) and resulting ER Ca2+ overload during stimulation.
Another example that loss of function of APP is likely the basic mechanism of neurodegeneration.
In a new study from Jie Shen’s lab, C. Zhang et al. provide compelling evidence that presenilins (PS) located at presynaptic terminals regulate neurotransmitter release during synaptic transmission. By using new engineered mice with specific deletion of both presenilin genes in presynaptic (CA3) or postsynaptic (CA1) hippocampal neurons, the authors conclusively demonstrate that PS regulate normal neurotransmitter release by controlling activity-induced ryanodine receptor-evoked Ca2+ release. Although the exact mechanism(s) by which presenilin regulates ryanodine receptor function need further investigation, this is the first report showing that regulation of Ca2+ signaling by PS is crucial for glutamate release and synaptic facilitation. These findings are important because they may explain, at least in part, the deficits on LTP and memory previously observed in mutant mice lacking both PS in excitatory neurons of the forebrain (Saura et al., 2004). Previous studies have demonstrated that FAD-linked mutations in presenilin-1 increase calcium levels through different mechanisms and cellular sources (Chan et al., 2000; Stutzmann et al., 2006; Tu et al., 2006; Cheung et al., 2008). In agreement with this idea, Beth Stutzmann’s group report elevated ryanodine receptor-evoked calcium release in CA1 hippocampal neurons of 3xTg-AD mice, a phenotype previously attributed to presenilin-1 mutation (Stutzmann et al., 2006). Notably, Chakroborty et al. found that aberrant calcium release from ER results in enhanced presynaptic (PPF) and reduced postsynaptic (LTP) plasticity in presymptomatic, not yet memory-impaired, 3xTg-AD mice. Together, these new findings suggest that PS dysfunction results in synaptic plasticity and memory deficits by altering calcium signaling.
One of the interesting points of J. Shen´s paper is the previously unappreciated or underestimated role of presenilin on presynaptic function. Indeed, the presynaptic localization of presenilins and the precursors of Aβ APP CTFs, in glutamatergic synapses (Saura et al., 2005; Zhang et al., 2009), strongly supports a role of PS/γ-secretase on presynaptic function. Inoue et al. demonstrated recently that PS localized at synapses are essential for synapse formation by regulating the γ-secretase cleavage of EphA4 (see Inoue et al., 2009 and ARF related news story). It is possible that besides regulating neurotransmitter release, presenilins may regulate the formation, maturation, and function of excitatory synapses by regulating the processing and/or function of EphA4 as well as other unidentified synaptic proteins. Since presenilins are also present in postsynaptic compartments, studies on postsynaptic presenilin/γ-secretase function need to be carefully evaluated in the future.
An important question related to AD pathology that has not been addressed in Shen’s present study is whether the hippocampal presynaptic and postsynaptic plasticity deficits caused by loss of presynaptic PS function result in neurodegeneration, such as that observed by loss of PS in the forebrain (Feng et al., 2004; Saura et al., 2004). While inherited mutations in the presenilin genes cause synaptic dysfunction and neurodegeneration in familial cases of AD, in general, PS transgenic mice characterized by presynaptic and/or postsynaptic plasticity deficits do not show signs of neurodegeneration, which suggests that plasticity deficits may not be directly linked to neuron death in PS mutant mice. This raises the possibility for alternative molecular mechanisms regulated by presenilins that impact neuron survival. Although the biological result of loss of PS function may not be strictly the same as FAD mutations, especially for APP processing and Aβ generation, the emerging new roles of PS on synaptic and memory function and neuronal survival may explain why these pathogenic mutations are so aggresive during the disease process.
Dr. Saura discussed that FAD presenilin mutation showed neurodegeneration in humans, but not in mice. Why this difference? We believe that homocysteic acid (HA) may be a pathogenic compound for Alzheimer disease. In our hands, HA induced neurodegeneration. We found recently that there is a big difference in urinary HA excretion between humans and mice. Human HA excretion into urine is 1,000 times higher than in mice. From our observation, humans show more severe HA-induced neurodegeneration than do mice, and HA production is related to presenilin activation (submitted for publication).