28 July 2011. Scientists know that some familial AD mutations cause a toxic flood of calcium to enter neurons, but they know much less about how that calcium does its dirty work. In the Jul 22 Proceedings of the National Academy of Sciences, researchers led by Kevin Foskett at the University of Pennsylvania, Philadelphia, shed some light on these downstream pathways. They found that calcium constitutively activates the transcription factor cAMP response element binding protein (CREB), well known for its role in promoting learning and memory. Foskett and colleagues link overactive CREB to increased cell death in vitro, suggesting that, while a little CREB activity may be good for your brain, perhaps too much is bad.
Kim Green at the University of California, Irvine, praised the study, noting that “for years we have been talking about how these presenilin mutations, and perhaps presenilin itself, modulates calcium, but we have never really looked at the consequences of those calcium changes.” He was not involved in the work.
Numerous previous studies have shown that familial AD presenilin-1 (PS1) mutations result in exaggerated calcium release from internal cellular stores, although researchers disagree about the exact mechanism (see, e.g., Stutzmann et al., 2006; Stutzmann et al., 2007; Bezprozvanny et al., 2008; ARF related news story on Green et al., 2008; ARF related news story on Green and LaFerla, 2008; see also ARF live discussion). Previous work by Foskett and colleagues linked the inositol 1,4,5-triphosphate receptor (InsP3R) to PS1-induced calcium release (see ARF related news story).
First author Marioly Müller wanted to look more closely at cellular events driven by calcium toxicity. She used cultured neural cell lines in which she expressed either wild-type or mutant PS1. Cells containing the familial M146L PS1 mutation, but not those with wild-type PS1, had large spontaneous calcium spikes, which the researchers abolished by blocking InsP3R either pharmacologically or with short interfering RNA. Looking downstream, Müller and colleagues found that cells with mutant presenilin also contained more activated CREB, which accumulated in the nucleus. The researchers could block CREB activation by either knocking down mutant PS1 or inhibiting InsP3R.
The authors went on to dissect the pathway by which excess calcium turned on CREB, finding that it involved activation of Ca2+/CaM kinase kinase β (CaMKKβ) and Ca2+/calmodulin-dependent protein kinase IV (CaMKIV). Müller and colleagues also linked the pathway to cell death in vitro, finding that cells with mutant PS1 died in fivefold greater numbers and were more sensitive to the toxic effects of Aβ than cells with wild-type PS1. Inhibiting InsP3R, CaMK, or CREB returned cell death to normal levels.
To extend these results in vivo, the authors used the triple-transgenic 3xTgAD mouse, which expresses the M146V PS1 mutation. At four to six weeks of age, before these mice show any amyloid or tau deposition, or any cognitive deficits, they had a three- to fivefold increase in activated CaMKIV and CREB, as well as in several genes turned on by CREB. A mouse that carried only the M146V mutation gave similar results.
Because inhibiting CREB reduced cell death to normal levels in vitro, the authors suggest that constitutively active CREB might play a role in cell death in familial AD. This idea runs a bit against the grain: CREB activation is usually considered beneficial, as many studies have touted its importance for learning and memory (see, e.g., ARF related conference story; ARF news story; and ARF related news story). CREB levels are typically down in AD brains (see, e.g., Saura and Valero, 2011 and ARF related news story on Caccamo et al., 2010). Increasing CREB improves memory and protects synapses in AD models (see ARF related news story on Gong et al., 2004).
What explains this discrepancy? One difference, the authors suggest, is that constitutive CREB activity could be an early feature that precedes Aβ pathology, while lowered CREB levels are a consequence of amyloid accumulation and occur much later. Another factor is the chronic nature of the activation. Other studies have found harmful effects, such as memory problems, cognitive deficits, and neurodegeneration, from continuous CREB activity (see Ramos et al., 2003; Lopez de Armentia et al., 2007; and Viosca et al., 2009). Some of the genes that CREB turns on, such as c-fos and nitric oxide synthase, also cause cell degeneration when expressed too highly, the authors note.
However, Green pointed out that the young PS1 mutant mice have no cognitive problems or neurodegeneration, despite their elevated CREB. People with mutant PS1 also show no difficulties for the first several decades of life, implying that if CREB is overactive in the human brain, it is not immediately harmful. Perhaps excess CREB activity has no effect on cognition, or even helps to offset the negative effects of too much calcium, Green suggested. Another factor to keep in mind, Green said, is that presenilin mutations are found in only a small fraction of AD cases. This study “does not tell us anything about sporadic AD,” he noted, particularly as wild-type presenilin did not have a pathological effect in these studies. One of the next questions to be answered, Green told ARF, is whether the effect of mutant presenilin on calcium and CREB is also a physiological role of presenilins in normal tissues, which would make these results more broadly applicable to AD.—Madolyn Bowman Rogers.
Müller M, Cárdenas C, Mei L, Cheung KH, Foskett JK. Constitutive cAMP response element binding protein (CREB) activation by Alzheimer’s disease presenilin-driven inositol trisphosphate receptor (InsP3R) Ca2+ signaling. PNAS Early Edition, 22 July, 2011. Abstract