J Physiol. 2006 Aug 15; 575(Pt 1): 5–10.
Published online 2006 Jun 29. doi: 10.1113/jphysiol.2006.111203
Hugh A Pearson and Chris Peers
Abstract
Alzheimer's disease is recognized post mortem by the presence of extracellular senile plaques, made primarily of aggregation of amyloid β peptide (Aβ). This peptide has consequently been regarded as the principal toxic factor in the neurodegeneration of Alzheimer's disease. As such, intense research effort has been directed at determining its source, activity and fate, primarily with a view to preventing its formation or its biological activity, or promoting its degradation. Clearly, much progress has been made concerning its formation by proteolytic processing of the amyloid precursor protein, and its degradation by enzymes such as neprilysin and insulin degrading enzyme. The activities of Aβ, however, are numerous and yet to be fully elucidated. What is currently emerging from such studies is a diffuse but steadily growing body of data that suggests Aβ has important physiological functions and, further, that it should only be regarded as toxic when its production and degradation are imbalanced. Here, we review these data and suggest that physiological levels of Aβ have important physiological roles, and may even be crucial for neuronal cell survival. Thus, the view of Aβ being a purely toxic peptide requires re-evaluation.
Aβ formation
It is now over two decades since amyloid β peptide (Aβ) was first sequenced and recognized as a potential marker of Alzheimer's disease (Glenner & Wong, 1984). Soon after, this 39–43 amino acid peptide was identified as the major component of the extracellular plaques that define this major form of dementia (Wong et al. 1985). Since then a wealth of academic and commercial research has been aimed at understanding where this peptide comes from, for a very simple reason; if we can stop its production, we might also prevent Alzheimer's disease. Alzheimer's disease is a major problem, and one likely to grow with an ageing population since it is primarily a disease of old age. The intensity and volume of dedicated research has resulted in a detailed current understanding of Aβ production which is summarized in Fig. 1. In essence, Aβ is a cleavage product of a large, transmembrane protein, termed APP (amyloid precursor protein). APP can undergo cleavage down one of at least two pathways. In the first pathway, cleavage by the enzyme α-secretase prevents Aβ formation, and instead produces the neuroprotective sAPPα fragment. However, if sequential cleavage by β- and then γ-secretases predominates, Aβ is formed. This peptide can then aggregate over time to produce senile plaques, during which period it also evokes numerous neurotoxic effects (some of which may require a degree of oligomerization), or it can be degraded by enzymes such as neprilysin, insulin degrading enzyme or endothelin converting enzyme (Turner et al. 2004). Clearly, the net balance of production versus degradation and clearance will determine levels present in cerebrospinal fluid (CSF), which are low but nevertheless measurable even in individuals showing no signs of dementia whatsoever (Selkoe & Schenk, 2003). The presence of Aβ in the CSF of non-demented individuals and in media from neuronal cell cultures (Tamaoka et al. 1997; Haass et al. 1992) indicates that, as well as having a potential pathological role in Alzheimer's disease, Aβ has a role in the normal physiology of the central nervous system.
Cartoon depicting the proteolytic processing of amyloid precursor protein (APP) via non-amyloidogenic (left) and amyloidogenic (right) cleavage
The physiology of APP processing
The production and degradation of Aβ has given us many insights into potential target processes for therapeutic intervention aimed at preventing Aβ formation or accelerating its degradation. Certainly, numerous compounds have been designed to interfere with either the β- or γ-secretases, but progress has been hampered by the fact that these enzymes also cleave other substrate proteins, so it is not only the production of Aβ that would be impaired (see reviews by Selkoe & Schenk, 2003; Vardy et al. 2005). Furthermore, although the potential for beneficial intervention remains great, it must be remembered that this proteolytic pathway is a physiological process – only when net Aβ levels become excessive can this process be regarded as pathological. One major, and often overlooked, aspect of APP processing is that numerous peptides can be generated. These – and what is known of their roles – are summarized in Table 1. Clearly, there is much more to learn about the roles and activities of these peptides, but these are beyond the scope of this review. Instead, we will focus on the potential physiological roles of Aβ. Such a discussion is complicated by the fact that many studies on the effects of Aβ focus on the toxic actions of the peptide. Concentrations of Aβ used in such studies are often far higher than levels found in CSF (1–10 nM). We have therefore directed our attention to studies where levels of applied Aβ are in the < 100 nM range, or where endogenous Aβ production/breakdown has been modified.
Cleavage products of amyloid precursor protein (APP) and their suggested physiological roles
Physiological control of synaptic activity
Several lines of evidence indicate that Aβ may have a role in controlling synaptic activity. Kamenetz et al. (2003) found that evoked activity of hippocampal neurones in brain slices increased the production of Aβ primarily by increasing trafficking of APP towards β secretase sites at the cell membrane. This would promote Aβ formation, but also increase production of other fragments (Fig. 1) such as AICD which may also modulate synaptic activity. At physiological expression levels of APP, this provided a negative feedback, since Aβ depresses synaptic activity. Without such depression, synaptic activity could become excessive, leading to excitotoxicity. Indeed, γ secretase inhibition led to increased EPSC frequency (Kamenetz et al. 2003), and kainate-induced seizures are potentiated in APP knockout mice (Steinbach et al. 1998). Further to these studies, a recent report indicated that specific stimulation of NMDA receptors up-regulated APP, inhibited α-secretase activity and promoted Aβ production (Lesne et al. 2005). Collectively, these studies argue strongly that APP processing, and the presence of Aβ itself, are closely associated with synaptic activity and may serve to provide physiological control of activity, guarding against excessive glutamate release.
What happens in the absence of Aβ?
In primary cultures of central neurones inhibition of endogenous Aβ production (by exposure to inhibitors either of β- or γ-secretases) or immunodepletion of Aβ caused neuronal cell death (Plant et al. 2003). Importantly, this appeared neurone-specific, since a variety of non-neuronal cells were unaffected by the same treatments. Perhaps most important, however, was the fact that neuronal cell death in response to secretase inhibition could be restored by addition of physiological (picomolar) levels of Aβ. The most common isoform, Aβ(1–40), was most effective in this regard, and the commonly used fragment Aβ(25–35), which retains many of the toxic properties of Aβ, was almost completely ineffective (Plant et al. 2003). These findings provided compelling evidence for a role for Aβ in neuronal survival. The underlying mechanism remains to be determined, but may involve altered expression of K+ channels. The activity of various K+channel types has been implicated in neuronal survival or death, in part because they govern excitability and hence the excitotoxicity of released glutamate, but also because intracellular [K+] is a key determinant of apoptosis (Yu, 2003). Changes in K+ channel expression may also provide an explanation of the findings of Kamenetz et al. (2003). We recently reported that inactivating K+ currents of central neurones are suppressed in amplitude by inhibition of the production of endogenous Aβ (Plant et al. 2005). As previously, this effect could be recovered by low levels of exogenous Aβ. This therefore probably reflects a physiological role for Aβ in controlling both excitability and cell survival.
We would predict from these findings that transgenic animals in which APP expression had been knocked out would show severe neurological deficits or lethality. This is not the case. APP-null mice show reduced branching of dendrites and fewer synaptic boutons but no reduction in neuronal number despite an absence of Aβ (Dawson et al. 1999). Furthermore, these reported neurological changes may not be a result of a deficiency in Aβ but of a deficiency in another product of APP processing or of APP itself. This is supported by the fact that an APP/APLP (amyloid precursor like protein) double knockout is lethal (von Koch et al. 1997). In addition, a transgenic knockout of β-secretase in which levels of Aβ are reduced to less than 10% of controls show no obvious deficits in either behaviour or neurology (Luo et al. 2001). On the other hand, one could argue that compensatory mechanisms in APP-null transgenic animals prevent toxic changes from occurring. One approach that would answer this would be to test the effects of an APP conditional knockout, but as far as we are aware these have yet to be made.
Hypoxia and APP processing
It has long been known that individuals who have suffered severe or chronic periods of hypoxia are more likely to develop Alzheimer's disease subsequently (Desmond et al. 2002). Indeed, hypoxic/ischaemic conditions up-regulate APP mRNA and protein (Hall et al. 1995; Kokmen et al. 1996; Jendroska et al. 1997; Shi et al. 2000), with consequent increases in Aβ. In vitro, hypoxia increases Aβ production (Taylor et al. 1999) with numerous consequences for cell function. Interestingly, many of these concern Ca2+ signalling (Smith et al. 2003, 2004), and disruption of Ca2+ signalling is a key event underlying neuronal death in Alzheimer's disease (LaFerla, 2002). One important aspect is the alteration of functional expression of ion channels: in cerebellar granule neurones, prolonged hypoxia leads to a selective up-regulation of L-type Ca2+channels. This effect requires Aβ formation, since it could be prevented by secretase inhibitors (Webster et al. 2006). This mechanism was explored further using a recombinant expression system: results (summarized as part of the cartoon of Fig. 2) indicated that hypoxia triggered increased Aβ production. This effect was dependent on an increase of reactive oxygen species (ROS) derived specifically from mitochondria (this in itself is seemingly paradoxical, but continues to gain increasing support; see Guzy et al. 2005). Once formed, the Aβ interacted directly with the recombinant L-type Ca2+ channel (α1C) subunit either to promote trafficking of channels towards, or inhibit retrieval of channels from, the plasma membrane. The net effect was increased Ca2+ channel protein at the cell membrane and hence increased Ca2+ conductance.
The question arises, however, of whether this altered channel trafficking represents a physiological or pathological response of cells to hypoxia. On one hand, it is convenient to propose the idea of altered ion channel trafficking as a mechanism accounting for the clinical observations that prolonged hypoxia promotes development of Alzheimer's disease (see earlier). Alternatively, this can be regarded as a physiological process. This work – like the vast majority of in vitro cellular studies – was conducted using a ‘normoxic’ level of 150–160 mmHg (i.e. room air-equilibrated conditions). This is, of course, hyperoxic for any cell. For central neurones, this is a particularly high ambient level of O2. Central nervous system O2 levels vary, but rarely exceed approximately 40 mmHg (Li et al. 2005). Indeed, commonly, they can be half this value, as measured in animals breathing normal air (Grote et al. 1996). A compelling recent study has indicated that if neurones are isolated and maintained in primary culture at physiological O2 levels, they appear more robust, and can tolerate hypoxia to much greater levels than cells cultured at 20% O2 (Li et al. 2005). The protection appeared to derive in part from the production of VEGF (vascular endothelial growth factor), which in turn was driven by the stabilization of the α1 subunit of the transcriptional regulator, hypoxia inducible factor (HIF-1α). Indeed, pharmacological activation of HIF-1α promoted neuronal survival, as did exogenous VEGF (Li et al. 2005). But do these effects involve Aβ in any way? At present we simply do not know, but available evidence points to a potentially exciting link. Soucek et al. (2003) examined the relationship between HIF-1, Aβ and glucose metabolism in central neurones particularly with a view to examining potentially protective shifts in metabolic activity (generating greater reducing equivalents via glycolysis and the hexose monophosphate shunt). Crucially, they found that low levels of Aβ could induce HIF-1 and thereby protect cells from the toxicity of exposure to high levels of Aβ. In addition, activation of HIF also mimicked the protective effects of low levels of Aβ. The authors themselves concluded that ‘an early function of Aβ in ageing is neuroprotection’ (Soucek et al. 2003).
Summary
This brief review has highlighted some of an accumulating but diffuse collection of work pointing to important physiological roles for Aβ. Given the presence of specific enzymatic pathways for the constitutive generation of Aβ, coupled with the fact that there exist selective uptake, breakdown and clearance pathways for its removal, it seems inconceivable that Aβ does not have a role to play in the normal function of the nervous system. Aβ should not, therefore, be regarded merely as a toxic factor that requires eradication to avoid dementia. Clearly, there is evidence to suggest essential modulation of synaptic activity and neuronal survival, but what the full physiological extent of Aβ activity (or indeed the activity of other products of APP proteolysis) remains to be seen.
Acknowledgments
Our work in this field has been/is supported by the MRC, the Wellcome Trust, The Alzheimer's Research Trust and the Alzheimer's Society.
References
Dawson GR, Seabrook GR, Zheng H, Smith DW, Graham S,
O’Dowd G et al. (1999). Age-related cognitive deficits,
impaired long-term potentiation and reduction in synaptic
marker density in mice lacking the beta-amyloid precursor
protein. Neuroscience 90, 1–13.
Desmond DW, Moroney JT, Sano M & Stern Y (2002).
Incidence of dementia after ischemic stroke: results of a
longitudinal study. Stroke 33, 2254–2260.
Dumont M, Lalonde R, Ghersi-Egea JF, Fukuchi K & Strazielle
C (2005). Regional acetylcholinesterase activity and its
correlation with behavioral performances in 15-month old
transgenic mice expressing the human C99 fragment of APP.
J Neural Transm; DOI: 10.1007/s00702-005-0373-6.
Furukawa K, Barger SW, Blalock EM & Mattson MP (1996).
Activation of K+ channels and suppression of neuronal
activity by secreted beta-amyloid-precursor protein. Nature
379, 74–78.
Glenner GG & Wong CW (1984). Alzheimer’s disease: initial
report of the purification and characterization of a novel
cerebrovascular amyloid protein. Biochem Biophys Res
Commun 120, 885–890.
Grote J, Laue O, Eiring P & Wehler M (1996). Evaluation of
brain tissue O2 supply based on results of PO2 measurements
with needle and surface microelectrodes. J Auton Nerv Syst
57, 168–172.
Guzy RD, Hoyos B, Robin E, Chen H, Liu L, Mansfield KD
et al. (2005). Mitochondrial complex III is required for
hypoxia-induced ROS production and cellular oxygen
sensing. Cell Metab 1, 401–408.
Haass C, Schlossmacher MG, Hung AY, Vigo-Pelfrey C, Mellon
A, Ostaszewski BL et al. (1992). Amyloid beta-peptide is
produced by cultured cells during normal metabolism.
Nature 359, 322–325.
Hall ED, Oostveen JA, Dunn E & Carter DB (1995). Increased
amyloid protein precursor and apolipoprotein E
immunoreactivity in the selectively vulnerable hippocampus
following transient forebrain ischemia in gerbils. Exp Neurol
135, 17–27.
Jendroska K, Hoffmann OM & Patt S (1997). Amyloid beta
peptide and precursor protein (APP) in mild and severe
brain ischemia. Ann NY Acad Sci 826, 401–405.
Kamenetz F, Tomita T, Hsieh H, Seabrook G, Borchelt D,
Iwatsubo T, SisodiaS&Malinow R (2003). APP processing
and synaptic function. Neuron 37, 925–937.
C 2006 The Authors. Journal compilation C 2006 The Physiological Society
10 H. A. Pearson and C. Peers J Physiol 575.1
Kokmen E, Whisnant JP, O’Fallon WM, Chu CP & Beard CM
(1996). Dementia after ischemic stroke: a population-based
study in Rochester, Minnesota (1960–84). Neurology 46,
154–159.
LaFerla FM (2002). Calcium dyshomeostasis and intracellular
signalling in Alzheimer’s disease. Nat Rev Neurosci 3,
862–872.
Leissring MA, Murphy MP, Mead TR, Akbari Y, Sugarman
MC, Jannatipour M et al. (2002). A physiologic signaling
role for the gamma-secretase-derived intracellular fragment
of APP. Proc Natl Acad Sci U S A 99, 4697–4702.
Lesne S, Ali C, Gabriel C, Croci N, MacKenzie ET, Glabe CG
et al. (2005). NMDA receptor activation inhibits
alpha-secretase and promotes neuronal amyloid-beta
production. J Neurosci 25, 9367–9377.
Li D, Marks JD, Schumacker PT, Young RM & Brorson JR
(2005). Physiological hypoxia promotes survival of cultured
cortical neurons. Eur J Neurosci 22, 1319–1326.
Luo Y, Bolon B, Kahn S, Bennett BD, Babu-Khan S, Denis P
et al. (2001). Mice deficient in BACE1, the Alzheimer’s
beta-secretase, have normal phenotype and abolished
beta-amyloid generation. Nat Neurosci 4, 231–232.
Pardossi-Piquard R, Petit A, Kawarai T, Sunyach C, Alves d C,
Vincent B et al. (2005). Presenilin-dependent transcriptional
control of the Abeta-degrading enzyme neprilysin by
intracellular domains of betaAPP and APLP. Neuron 46,
541–554.
Plant LD, Boyle JP, Smith IF, Peers C & Pearson HA (2003). The
production of amyloid beta peptide is a critical requirement
for the viability of central neurons. J Neurosci 23, 5531–5535.
Plant LD, Webster NJ, Boyle JP, Ramsden M, Freir DB, Peers C
& Pearson HA (2005). Amyloid beta peptide as a
physiological modulator of neuronal ‘A’-type K+ current.
Neurobiol Aging In Press
Ramsden M, Plant LD, Webster NJ, Vaughan PF, Henderson Z
& Pearson HA (2001). Differential effects of unaggregated
and aggregated amyloid beta protein (1–40) on K+ channel
currents in primary cultures of rat cerebellar granule and
cortical neurones. J Neurochem 79, 699–712.
Ramsden M, Henderson Z & Pearson HA (2002). Modulation
of Ca2+ channel currents in primary cultures of rat cortical
neurones by amyloid beta protein (1–40) is dependent on
solubility status. Brain Res 956, 254–261.
Rockenstein E, Mante M, Alford M, Adame A, Crews L,
Hashimoto M et al. (2005). High beta-secretase activity
elicits neurodegeneration in transgenic mice despite
reductions in amyloid-beta levels: implications for the
treatment of Alzheimer disease. J Biol Chem 280,
32957–32967.
Selkoe DJ & Schenk D (2003). Alzheimer’s disease: molecular
understanding predicts amyloid-based therapeutics. Annu
Rev Pharmacol Toxicol 43, 545–584.
Shi J, Yang SH, Stubley L, Day AL & Simpkins JW (2000).
Hypoperfusion induces overexpression of beta-amyloid
precursor protein mRNA in a focal ischemic rodent model.
Brain Res 853, 1–4.
Smith IF, Boyle JP, Green KN, Pearson HA & Peers C (2004).
Hypoxic remodeling of Ca2+ mobilization in type I cortical
astrocytes: involvement of ROS and pro-amyloidogenic APP
processing. J Neurochem 88, 869–877.
Smith IF, Plant LD, Boyle JP, Skinner RA, Pearson HA & Peers
C (2003). Chronic hypoxia potentiates capacitative Ca2+
entry in type I cortical astrocytes. J Neurochem 85,
1109–1116.
Smith RP, Higuchi DA & Broze GJ Jr (1990). Platelet
coagulation factor XIa-inhibitor, a form of Alzheimer
amyloid precursor protein. Science 248, 1126–1128.
Soucek T, Cumming R, Dargusch R, Maher P & Schubert D
(2003). The regulation of glucose metabolism by HIF-1
mediates a neuroprotective response to amyloid beta
peptide. Neuron 39, 43–56.
Steinbach JP, Muller U, Leist M, Li ZW, NicoteraP&Aguzzi A
(1998). Hypersensitivity to seizures in beta-amyloid
precursor protein deficient mice. Cell Death Differ 5,
858–866.
Tamaoka A, Sawamura N, Fukushima T, Shoji S, Matsubara E,
Shoji M et al. (1997). Amyloid beta protein 42 (43) in
cerebrospinal fluid of patients with Alzheimer’s disease.
J Neurol Sci 148, 41–45.
Taylor SC, Batten TF & Peers C (1999). Hypoxic enhancement
of quantal catecholamine secretion. Evidence for the
involvement of amyloid beta-peptides. J Biol Chem 274,
31217–31222.
Turner AJ, Fisk L & Nalivaeva NN (2004). Targeting
amyloid-degrading enzymes as therapeutic strategies in
neurodegeneration. Ann N Y Acad Sci 1035, 1–20.
Vardy ER, Catto AJ & Hooper NM (2005). Proteolytic
mechanisms in amyloid-beta metabolism: therapeutic
implications for Alzheimer’s disease. Trends Mol Med 11,
464–472.
von Koch CS, Zheng H, Chen H, Trumbauer M, Thinakaran G,
Van der Ploeg LH, Price DL & Sisodia SS (1997). Generation
of APLP2 KO mice and early postnatal lethality in
APLP2/APP double KO mice. Neurobiol Aging 18,
661–669.
Webster NJ, Ramsden M, Boyle JP, Pearson H & Peers C (2006).
Amyloid peptides mediate hypoxic increase of L-type Ca2+
channels in central neurones. Neurobiol Aging 27, 439–445.
Wong CW, Quaranta V & Glenner GG (1985). Neuritic plaques
and cerebrovascular amyloid in Alzheimer disease are
antigenically related. Proc Natl Acad Sci U S A 82,
8729–8732.
Yu SP (2003). Regulation and critical role of potassium
homeostasis in apoptosis. Prog Neurobiol 70, 363–386