. Effects of synaptic modulation on beta-amyloid, synaptophysin, and memory performance in Alzheimer's disease transgenic mice. J Neurosci. 2010 Oct 27;30(43):14299-304. PubMed.


Please login to recommend the paper.


  1. Ten years ago, Gunnar Gouras began what has become a persistent call for attention to the potential importance of intraneuronal Aβ (Gouras et al., 2000). Since it is virtually impossible to be certain about which APP fragments are actually being observed in a histological setting, in an abundance of caution, herein we refer to this material as Aβ-like immunoreactivity (Aβ-LIR). In the intervening decade, Gouras has continued to find intraneuronal Aβ under every bed, and a number of other investigators (e.g., Oddo et al., 2003; LaFerla et al., 2007) have joined in with a chorus of endorsements of the “intraneuronal Aβ hypothesis.” Still, several fundamental questions have gone unaddressed, and these questions accrue new importance with the reports from several labs (including Gouras’s lab) that modulation of neurotransmission or autophagy stimulates clearance of intraneuronal Aβ, and that neurons disgorged of their Aβ-LIR material are happier and better functioning (Almeida et al., 2005; Almeida et al., 2006; Tampellini et al., 2007; Tampellini et al., 2009; Tampellini et al., 2010; Himeno et al., 2010; Caccamo et al., 2010; Spilman et al., 2010).

    Certainly, many labs, including our own (Petanceska et al., 2000; Gandy et al 2010), have observed striking intraneuronal pathology in transgenic mice. This is somewhat surprising since intraneuronal Aβ is difficult to recover from cultured cells, with the exception of systems employing Swedish APP (Haass et al., 1993) and perhaps with the important exception of cultured neurons (Turner et al., 1996). Compare this with the cultured media from these cells: There, newly generated Aβ is readily apparent (Haass et al., 1992). Why should we focus on the tiny amount of intraneuronal Aβ when clearly most Aβ is secreted? Further, when turning to human postmortem material, intraneuronal Aβ is observed occasionally at best. So, to recap, here are two strikes against intraneuronal Aβ: first, the tiny stoichiometry of retained versus secreted material, and second, the poor validation of intraneuronal Aβ-LIR from human AD neuropathology. To be fair, one could argue that the intraneuronal Aβ-LIR was first generated as secreted Aβ that has subsequently been endocytosed. Still, the poor validation in human material is worrisome.

    Despite these concerns, there are reasons to find intraneuronal Aβ-LIR to be an attractive target. In our recently described “oligomer only” Dutch APP mice (Gandy et al., 2010), the only site of Aβ-LIR accumulation is intraneuronal. This being the case, we are left to ponder the possibility that the mouse model might be exaggerating an authentic step in human AD pathogenesis, but which, in humans, occurs rapidly and/or so early in the development of the disease, and/or at such low levels, that we have been unable to capture it in postmortem human brain. Provisionally, we consider that the Aβ oligomers we are measuring are residing amongst the intraneuronal Aβ-LIR. The idea that intraneuronal Aβ oligomerization is a key step in pathogenesis brings to mind one formulation of how plaques are formed and why cerebral amyloidosis is miliary, i.e., that tiny chunks of indigestible Aβ-LIR material are extruded from neurons and initiate plaque formation. Although this next, unfortunately apt comment will induce snickers, this process has occasionally been compared to defecation.

    The link between diabetes and AD also appears to dovetail with this story. Insulin signaling is a dramatic stimulator of Aβ release (Gasparini et al., 2001; Liao and Xu, 2009), and metformin potentiates insulin-sensitive Aβ release (Chen et al., 2009) and may be associated with attenuated neuropathology (Beeri et al., 2008). It is difficult to avoid the formulation that insulin-stimulated secretion of Aβ appears to be a good thing. Tampellini et al (2009 and 2010) fits this model as well.

    So, while under-represented in the conventional narrative of the APP/Aβ lifecycle and not obvious in human neuropathology, there appears to be good reason to consider the reduction of intraneuronal Aβ-LIR as a valid target for AD therapy, even if we can’t quite figure out exactly what subcellular events transpired to create the intraneuronal Aβ-LIR material in the first place.


    . Beta-amyloid accumulation in APP mutant neurons reduces PSD-95 and GluR1 in synapses. Neurobiol Dis. 2005 Nov;20(2):187-98. PubMed.

    . Beta-amyloid accumulation impairs multivesicular body sorting by inhibiting the ubiquitin-proteasome system. J Neurosci. 2006 Apr 19;26(16):4277-88. PubMed.

    . Insulin in combination with other diabetes medication is associated with less Alzheimer neuropathology. Neurology. 2008 Sep 2;71(10):750-7. PubMed.

    . Molecular interplay between mammalian target of rapamycin (mTOR), amyloid-beta, and Tau: effects on cognitive impairments. J Biol Chem. 2010 Apr 23;285(17):13107-20. PubMed.

    . Antidiabetic drug metformin (GlucophageR) increases biogenesis of Alzheimer's amyloid peptides via up-regulating BACE1 transcription. Proc Natl Acad Sci U S A. 2009 Mar 10;106(10):3907-12. PubMed.

    . Days to criterion as an indicator of toxicity associated with human Alzheimer amyloid-beta oligomers. Ann Neurol. 2010 Aug;68(2):220-30. PubMed.

    . Stimulation of beta-amyloid precursor protein trafficking by insulin reduces intraneuronal beta-amyloid and requires mitogen-activated protein kinase signaling. J Neurosci. 2001 Apr 15;21(8):2561-70. PubMed.

    . Intraneuronal Abeta42 accumulation in human brain. Am J Pathol. 2000 Jan;156(1):15-20. PubMed.

    . Amyloid beta-peptide is produced by cultured cells during normal metabolism. Nature. 1992 Sep 24;359(6393):322-5. PubMed.

    . Cellular processing of beta-amyloid precursor protein and the genesis of amyloid beta-peptide. Cell. 1993 Dec 17;75(6):1039-42. PubMed.

    . Apomorphine treatment in Alzheimer mice promoting amyloid-β degradation. Ann Neurol. 2011 Feb;69(2):248-56. PubMed.

    . Intracellular amyloid-beta in Alzheimer's disease. Nat Rev Neurosci. 2007 Jul;8(7):499-509. PubMed.

    . Insulin signaling in sporadic Alzheimer's disease. Sci Signal. 2009;2(74):pe36. PubMed.

    . Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron. 2003 Jul 31;39(3):409-21. PubMed.

    . Mutant presenilin 1 increases the levels of Alzheimer amyloid beta-peptide Abeta42 in late compartments of the constitutive secretory pathway. J Neurochem. 2000 May;74(5):1878-84. PubMed.

    . Inhibition of mTOR by rapamycin abolishes cognitive deficits and reduces amyloid-beta levels in a mouse model of Alzheimer's disease. PLoS One. 2010;5(4):e9979. PubMed.

    . Internalized antibodies to the Abeta domain of APP reduce neuronal Abeta and protect against synaptic alterations. J Biol Chem. 2007 Jun 29;282(26):18895-906. PubMed.

    . Synaptic activity reduces intraneuronal Abeta, promotes APP transport to synapses, and protects against Abeta-related synaptic alterations. J Neurosci. 2009 Aug 5;29(31):9704-13. PubMed.

    . Effects of synaptic modulation on beta-amyloid, synaptophysin, and memory performance in Alzheimer's disease transgenic mice. J Neurosci. 2010 Oct 27;30(43):14299-304. PubMed.

    . Amyloids beta40 and beta42 are generated intracellularly in cultured human neurons and their secretion increases with maturation. J Biol Chem. 1996 Apr 12;271(15):8966-70. PubMed.

  2. Reply to comment by Sam Gandy and John Steele
    I appreciate the interest in intraneuronal Aβ by Gandy and Steele, who have been contributing to this topic. I also welcome the comment, because interaction and questioning is so important. Yes, it has been a long road, and I will underscore that many other Alzheimer’s investigators have joined this pursuit; see our recent review (Gouras et al., 2010). I have also noticed an increasing “grassroots” interest in this topic, at the poster level at conferences and in ever more published papers, although it still has not quite reached many review articles at top-tier journals. The aim of this work is not to start a fringe topic that is irrelevant, but to move AD forward, and the resistance to change has been remarkable. To address the major issues raised by Gandy and Steele:

    1. Lack of clear evidence in human AD brain: For me, intraneuronal Aβ actually all began with human brain. Looking through a microscope at sections from postmortem Down’s syndrome brains immuno-labeled with Aβ40 and 42 antibodies, I was struck by the marked intraneuronal labeling of particularly Aβ42 in Alzheimer’s prone neurons of already quite young Down’s syndrome brains; neurons of layer 2 in the entorhinal cortex and CA1 in the hippocampus were most prominently labeled. Thus, I would say that intraneuronal Aβ can be quite obvious in human brains. Many groups have now reported intraneuronal Aβ accumulation in human AD and Down’s syndrome brains. So why the concern about human brain by Gandy and Steele? I believe it is mainly because when you look at typical postmortem human AD brain (where plaque pathology tends to be advanced), intraneuronal Aβ42 is not as obvious. In fact, the same is seen with advanced plaque pathology in AD transgenic mice. We still do not have the full answer to this, but part of the answer—which we first realized with immuno-EM—is that most of the intraneuronal Aβ42 is present in neurites and synapses, which cannot be seen by light microscopy. In addition, conformational antibody specificities impact the amount of Aβ that one can see. For example, we have learned that standard Aβ42 antibodies are poor at detecting anything above dimers and trimers (they are best for monomers), limiting the amount of Aβ that can be detected with one antibody.

    2. Lack of biochemical evidence: It is true that this was a major issue. In unpublished work, we struggled with this. For example, we could add a lot of synthetic Aβ1-42 to cell lysates, but subsequently could retrieve only a small fraction of this added Aβ. Given its hydrophobic nature, we may have to study Aβ42 more like a lipid. Since one cannot run a Western for lipids, the lipid field has increasingly turned to using fluorescence labeling—an area that has been revolutionized in recent years by better microscopes, cameras, and imaging software. When using immunofluorescence for intraneuronal Aβ, one needs to keep using the critical controls (see point 3, below). But to get back to biochemistry, we were excited by recent work from a biochemistry lab at the Karolinska Institute (Hashimoto et al., 2010); using laser capture microdissection microscopy in conjunction with a sensitive ELISA, they showed that AD-vulnerable CA1 neurons had a much higher ratio of Aβ42 to 40 in sporadic AD compared to control brains. In addition, while this higher ratio was maintained in cerebellar Purkinje neurons, the absolute levels of Aβ were much lower. In addition, Gylys, Cole, and colleagues have been showing nice evidence consistent with Aβ accumulation within synapses using synaptosomal preparations of AD brains (Gylys et al., 2004).

    3. Secreted Aβ: I do not believe that it is as clear that secreted Aβ42 is so much more abundant than intraneuronal Aβ42. If we examine conditioned media of neurons grown at low density, we cannot even detect Aβ40, let alone Aβ42, while intraneuronal Aβ42 labeling is prominent. Again, the use of controls is critical: APP knockout neurons, in conjunction with biochemical antibody analysis, and also comparing Aβ labeling with antibodies directed at different domains of APP (including ones you believe the Aβ42 antibody may be labeling!).

    Further, our immuno-EM actually argues against oligomerized Aβ42 being actively extruded from cells, as Gandy and Steele speculate. The subcellular morphology appears too badly damaged to allow for such active microtubule-based extrusion (Takahashi et al., 2004). The EM supports that this intraneuronal Aβ becomes extracellular following synaptic/neuritic degeneration.

    Lastly, I will plug our recent paper that had received little attention, since I believe it is quite important. The main message is that we provide experimental evidence that plaque load does not correlate with Aβ-related destruction of synapses. The second message is that this paper counterbalances recent studies that have stressed the detrimental aspects of synaptic activity for AD. We show that reduced synaptic/cerebral activity destroys synapses in AD transgenic (but not wild-type mice), despite a decrease in plaques (whereas there is an increase in intraneuronal Aβ42). This might be important for clinical trials. Why? Because while amyloid imaging should be very helpful in diagnosing AD, we must be cautious using it as a surrogate marker for disease progression. We also hypothesize that intraneuronal Aβ is key to understanding why the aborted active vaccine trial showed continued cognitive decline despite plaque removal. Intraneuronal Aβ might also clarify reduced CSF Aβ42 in AD, which could be due to reduced release of the peptide from cells rather than accumulation in plaques. Our latest findings (Tampellini et al., 2010) support that synaptic activity has benefits in warding off AD. This doesn’t argue against the hypothesis that chronic sites of elevated synaptic/brain activity are prone to develop AD, or against detrimental effects of hyperexcitability, or sleep deprivation. We reconcile these various studies as pointing to Aβ altering the cellular machinery at synapses that no longer can effectively do what it could when young—efficiently secrete Aβ while also reducing intraneuronal Aβ42 with activity.

    To conclude, after looking at enough electron micrographs of various AD transgenic mice, as well as advanced human AD brains, we see that intraneuronal accumulating and aggregating Aβ42 clearly appears to be a major problem, which is why we focus on it! The realization that this problem initiates within synapses (Takahashi et al., 2002) is what prompted us to turn to how accumulating Aβ alters synapses. At the same time, Aβ and synapses are, of course, not acting in isolation, and important contributions to AD pathogenesis come from ApoE, tau, inflammatory cells, the vasculature, etc., in addition to other factors involved in brain aging.


    . Intraneuronal beta-amyloid accumulation and synapse pathology in Alzheimer's disease. Acta Neuropathol. 2010 May;119(5):523-41. PubMed.

    . Analysis of microdissected human neurons by a sensitive ELISA reveals a correlation between elevated intracellular concentrations of Abeta42 and Alzheimer's disease neuropathology. Acta Neuropathol. 2010 May;119(5):543-54. PubMed.

    . Synaptic changes in Alzheimer's disease: increased amyloid-beta and gliosis in surviving terminals is accompanied by decreased PSD-95 fluorescence. Am J Pathol. 2004 Nov;165(5):1809-17. PubMed.

    . Oligomerization of Alzheimer's beta-amyloid within processes and synapses of cultured neurons and brain. J Neurosci. 2004 Apr 7;24(14):3592-9. PubMed.

    . Beta-amyloid accumulation impairs multivesicular body sorting by inhibiting the ubiquitin-proteasome system. J Neurosci. 2006 Apr 19;26(16):4277-88. PubMed.

    . Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am J Pathol. 2002 Nov;161(5):1869-79. PubMed.

  3. We commend Gunnar Gouras for persevering with his studies on intraneuronal amyloid-β (Aβ). Gouras and his collaborators pioneered the idea that the Alzheimer’s disease (AD)-related Aβ pathology begins inside the neuron, and precedes both neurofibrillary tangles and Aβ plaque deposition (1). Over more than a decade, Gouras, and a few others, continued to gather evidence that the initial observations made in the AD brain hold true in mouse models of AD and cultured neurons, where they can be studied. The problem of intracellular Aβ will certainly preoccupy more and more investigators in the years to come. It also preoccupies us (2). Although views may vary, this Aβ accumulates within neurons early during AD and appears to correlate well with the incipient phases of the disease. While not definitively proven, it is likely that this intraneuronal Aβ contributes to the synapse pathology in AD, as Gouras correctly argues (3); it may also constitute the seed that nucleates AD plaques (see, e.g., 4).

    In the CAD neuronal cell line—one of our systems of study—intracellular Aβ (monomeric and oligomeric) is readily detectable in immunoblots (2), while secreted Aβ is more difficult to detect—in spite of the abundance of secreted sAPPs. This situation is similar to what Gouras noticed in his work with neuronal cultures (see his comment to this paper). This result suggests that Aβ accumulates and oligomerizes within neurons prior to its accumulation in the extracellular space. Intuitively, this should be so. Although Aβ can be cleaved off the recycled CTFs at the cell surface, and thus released without being truly secreted, a large body of evidence points to the fact that most Aβ is generated within the cell. Under normal conditions, most of it is probably degraded within the cell, and some is retrieved from the compartments where it was produced and released into the extracellular space. In AD, a larger fraction may be retained within the cell. Most importantly—as studies by Randy Nixon and collaborators consistently show—a larger fraction may not be properly degraded due to a failure of the endosome/lysosome/autolysosome system (5-10). In the end, the intracellularly accumulated Aβ becomes extracellular by a variety of mechanisms, including decomposition of the dying neuron.

    Animal models of disease, cultured neurons, and in vitro reconstitutions are the only systems available to researchers for dissecting the mechanisms of AD by experimentation. Obviously, if the data are not consistent with what is known about the human disease, they are discarded. This is certainly not the case with intracellular Aβ, which is present in the AD brain (1) in the soma and particularly in dystrophic neurites, mostly in endosomal/lysosomal and autophagic compartments (7). The study of intraneuronal Aβ and, in particular, of the mechanism of oligomerization and accumulation of Aβ within neurons—with animal models and cultured cells—will certainly contribute to the elucidation of the pathogenic mechanisms in AD.


    . Intraneuronal Abeta42 accumulation in human brain. Am J Pathol. 2000 Jan;156(1):15-20. PubMed.

    . Neuritic deposits of amyloid-beta peptide in a subpopulation of central nervous system-derived neuronal cells. Mol Cell Biol. 2006 Jul;26(13):4982-97. PubMed.

    . Intraneuronal beta-amyloid accumulation and synapse pathology in Alzheimer's disease. Acta Neuropathol. 2010 May;119(5):523-41. PubMed.

    . Seeding neuritic plaques from the distance: a possible role for brainstem neurons in the development of Alzheimer's disease pathology. Neurodegener Dis. 2008;5(3-4):250-3. Epub 2008 Mar 6 PubMed.

    . Macroautophagy--a novel Beta-amyloid peptide-generating pathway activated in Alzheimer's disease. J Cell Biol. 2005 Oct 10;171(1):87-98. PubMed.

    . Extensive involvement of autophagy in Alzheimer disease: an immuno-electron microscopy study. J Neuropathol Exp Neurol. 2005 Feb;64(2):113-22. PubMed.

    . Abeta localization in abnormal endosomes: association with earliest Abeta elevations in AD and Down syndrome. Neurobiol Aging. 2004 Nov-Dec;25(10):1263-72. PubMed.

    . Regional selectivity of rab5 and rab7 protein upregulation in mild cognitive impairment and Alzheimer's disease. J Alzheimers Dis. 2010;22(2):631-9. PubMed.

    . Microarray analysis of hippocampal CA1 neurons implicates early endosomal dysfunction during Alzheimer's disease progression. Biol Psychiatry. 2010 Nov 15;68(10):885-93. Epub 2010 Jul 23 PubMed.

    . Lysosomal proteolysis and autophagy require presenilin 1 and are disrupted by Alzheimer-related PS1 mutations. Cell. 2010 Jun 25;141(7):1146-58. PubMed.

Make a Comment

To make a comment you must login or register.