Alzheimer‘s disease therapy: focus on synapses

There is a growing race to discover a therapy for Alzheimer‘s disease (AD) following years of progress in uncovering the pathological, molecular and cellular basis of this leading cause of dementia and ever more prevalent disease of aging. Reduction in amyloid-β (Aβ) peptide, the main constituent of the disease-linked cerebral amyloid plaques, has become the major target for emerging experimental therapies [1]. Other therapeutic directions include prevention of tau hyperphosphorylation and aggregation [2], cholesterol reduction [3] and antioxidant treatment [4,5]. Although the aberrant accumulation of Aβ peptides into plaques is the neuropathological hallmark of the disease, the molecular mechanism(s) whereby Aβ accumulates and is involved in neuronal dysfunction and degeneration remain unclear. In fact, Aβ peptides are normally generated by cells, especially neurons, and neuronal activity stimulates the secretion of Aβ peptides [6].

It has long been known that alterations in synapses, rather than the Aβ plaques or tau-containing neurofibrillary tangles, are the best pathological correlate of cognitive decline in the disease [7]. The larger and more abundant Aβ precursor protein (APP) is ubiquitously expressed and in neurons is transported rapidly to synapses [8], where cleavages to generate Aβ peptides are increasingly thought to occur. Transgenic mouse models of AD that accumulate Aβ abnormally have shown that both behavioral and synaptic deficits precede plaque and tangle pathology. Moreover, immunoelectron microscopy studies demonstrated that the aberrant accumulation of the most disease-linked Aβ42 peptides occurs in late endosomal organelles, especially in distal processes and synaptic compartments, and is directly associated with subcellular pathology even prior to plaque formation [9]. Elucidating the molecular steps by which Aβ peptides accumulate and lead to initial dysfunction at synapses has become a major effort in the field.

Synapses are sites of high metabolic activity in the brain, also reflected by their abundance of mitochondria. It is postulated that this might lead to especially elevated oxidative stress and free radical generation at synapses. Remarkably, it was demonstrated that the extracellular addition of Aβ peptides to cultured nerve cells leads to the binding of this Aβ to synapses of selective neurons [10]. The dynamic relationship between the intra- and extracellular pools of Aβ are increasingly being appreciated [11]. Secreted Aβ is initially generated within cells and addition of extracellular Aβ can markedly upregulate the intracellular pool of Aβ [12]. Furthermore, it was reported that addition of Aβ was toxic to wild-type neurons but not to cultured neurons derived from well-established APP knockout mice [13]. More recently, it was reported that point mutations in the NPXY motif in the C-terminus of APP was sufficient to block extracellular Aβ toxicity [14]. Therefore, increasing data support a relation between extra- and intracellular Aβ that requires APP.

This focus on synapses, in turn, could also be a warning of caution in the pursuit of therapies aimed at reduction of Aβ peptides. If Aβ generation and secretion are augmented by synaptic activity, could this Aβ be serving a normal function that might be perturbed by Aβ reduction? In addition, a growing paradox in the field is that activity-induced Aβ secretion, at least for the current amyloid-cascade hypothesis, does not appear easy to reconcile with the gradually progressive decrease in neuronal activity that occurs in the disease. Thus, based on this hypothesis, patients with AD might be expected to develop less rather than more deposition of Aβ. The alternative intraneuronal Aβ-cascade hypothesis postulates that it might rather be reduced synaptic activity fostering reduced Aβ secretion while augmenting intraneuronal Aβ accumulation that may be especially detrimental [6,11]. The intraneuronal Aβ hypothesis does not exclude that there is also an important role for extracellular Aβ. It posits that marked accumulation of Aβ within processes leads to initial synaptic dysfunction followed by degeneration that thereby creates high levels of extracellular Aβ, which that lead to a chain reaction of extracellular Aβ fostering further intracellular Aβ accumulation and pathology.

Currently, the major therapeutic directions for Aβ reduction are inhibition of either of the two proteases that generate Aβ from APP (β- and γ-secretase inhibitors) and immunotherapy for Aβ. Unfortunately, both β- and γ-secretases were found to have other physiological substrates, which could limit their effectiveness. At the same time, it might be possible that a modest reduction in one or both of these proteases could provide sufficient Aβ lowering while not leading to excessive inhibition that would be detrimental for normal cleavages. Aβ immunotherapy has been an exciting development in AD therapy [1]. Although an active Aβ vaccine clinical trial was halted because 6% of patients developed meningoencephalitis, follow-up studies on patients from this study reported remarkable clinical stabilization in some patients with elevated antibody titers. Currently, several new active and passive immunotherapy clinical trials for AD are in progress. Hypotheses for how antibodies against Aβ can reduce Aβ pathology and behavioral abnormalities in AD transgenic mice include Aβ efflux out of the brain, inflammatory cell recruitment and degradation of plaques, and neutralization of extracellular Aβ-induced synaptic damage. Recently, an additional mechanism was demonstrated, whereby Aβ antibodies reduce intraneuronal Aβ and protect synapses in cultured neurons from AD mouse models. Remarkably, this was shown to occur via Aβ-antibody binding to the Aβ domain within its precursor, APP, at the cell surface, followed by internalization into neurons and Aβ reduction [15]. The possibility that antibodies can act within cells has additional interesting implications in neurology for autoimmune and paraneoplastic diseases where the mechanisms of antibody-induced neuronal dysfunction remain unclear.

At a time where optimism for development of a disease-modifying medication for this debilitating disease of the aging brain is greater than ever; we need to remain vigilant about the complexity of this disease and continue our efforts to further uncover the basic biology of AD that combines and impairs some of the most complex processes we know: aging, synapses and cognition.

Financial disclosure

The authors have no relevant financial interests including employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties related to this manuscript.