In mouse models of Alzheimer’s, tickets for retrograde transport to neuronal cell bodies may become scarce, leaving BACE1 stranded at synapses. So hints a paper in the March 8 Journal of Neuroscience. Researchers led by Qian Cai of Rutgers University in Piscataway, New Jersey, also reported that restoring retrograde endosomal trafficking cleared excess BACE1, and Aβ, out of synapses, prevented their loss, and restored learning and memory. The researchers proposed the endosomal gridlock plays a pivotal role in synaptic destruction and neurodegeneration in AD.

Swollen axons, aka dystrophic neurites, are a hallmark of AD. Researchers have reported that the bulging processes, frequently found in the vicinity of amyloid plaques, are chock-full of BACE1 and one of its products, Aβ, as well as endosomal vesicles (see Kandalepas et al., 2013Oct 2016 conference news). Previous studies have described the slowing of retrograde transport of BACE1-laden endosomes out of axons, hinting that malfunctions in the pathway could explain the synaptic overcrowding (see Buggia-Prévot et al., 2014Sadleir et al., 2016). Cai’s previous studies detailed how synaptic endosomes hitch a ride to the soma by latching onto snapin, an adaptor that hooks up to dynein motors (see Cai et al., 2010). More recently, Cai and colleagues reported that this snapin-dynein transport malfunctioned in neurons from AD mice in culture, leading to a buildup of BACE1 and Aβ in distal axons (see Ye and Cai, 2014). 

For the current study, first author Xuan Ye and colleagues looked into the synaptic, cellular, and cognitive consequences of this transport hitch. Dovetailing with what Cai and other researchers had observed previously, the researchers spotted BACE1 in clusters congregating in dystrophic neurites in the hippocampus of hAPP-J20 mice, and determined that the enzyme co-localized with late endosomes crammed into presynaptic terminals. They observed elevated synaptic BACE1 in postmortem samples from AD patients as well. 

Crowding Clusters.

BACE1 (green) evenly distributes in the hippocampi of wild-type mice (left panels), but clumps into clusters that crowd presynaptic terminals (red) in hAPP mice (right panels). [Image courtesy of Ye et al., Journal of Neuroscience, 2017.]

To investigate the role of snapin-dynein retrograde transport of BACE1 in normal animals, the researchers next generated conditional knockout mice lacking snapin expression in neurons of the frontal cortex and hippocampus. Similar to hAPP mice, BACE1 co-localized with late endosomal and synaptic markers in these cKOs, and appeared to accumulate in presynaptic terminals of mossy fibers in the hippocampus. Electron microscopy revealed that BACE1 density shot up by sevenfold in the fibers of snapin-deficient mice compared to wild-type animals, and that nearly 90 percent of BACE1 resided in presynapses. Perhaps due to BACE1’s activity within these acidic organelles, mouse Aβ40 concentrations topped those in normal mice by more than 100 percent. The findings suggested that snapin-mediated retrograde transport of BACE1-laden endosomes played a key role in ridding synapses of the enzyme, along with its potentially damaging products.

The researchers hypothesized that snapin-mediated transport was defective in AD neurons, and that boosting snapin expression would restore it. Live imaging of hAPP neurons revealed that indeed, the flow of late endosomal BACE1 along axons moved predominantly in the anterograde direction: from the soma to the axon. In wild-type neurons, the traffic moved both ways. Overexpressing snapin in hAPP neurons partially restored retrograde transport, but did not affect anterograde. This boost of traffic out of the axon reduced the accumulation of endosomal BACE1 in presynaptic terminals.

Would snapin overexpression restore retrograde BACE1 trafficking in hAPP mice as well? To find out, the researchers injected adeno-associated viruses expressing snapin directly into the hippocampi of two- to three-month-old hAPP mice. Compared with animals injected with a control virus, these mice had more BACE1 in the soma of hippocampal neurons, and less in the mossy fiber axons, suggesting that transport from the axons was improved. Strikingly, snapin overexpression staved off synapse loss as well, restoring to wild-type levels the density of presynaptic terminals adorning the mossy fibers.

Snapin overexpression also reduced Aβ40 by more than 60 percent in preparations of hippocampal synaptosomes. Staining hippocampal slices with the A11 antibody, which recognizes soluble Aβ oligomers, revealed that snapin overexpression cut intracellular oligomers by more than half. Snapin’s benefits extended into the extracellular space as well, halving plaque area in the hippocampus, as measured by immunostaining with the 6E10 antibody. 

Snap! Aβ Gone.

Soluble intracellular Aβ (green) accumulates in hippocampal neurons of hAPP mice (left), in both soma (outside the circle) and mossy fiber axons (inside). Snapin overexpression reduces Aβ significantly (right). [Image courtesy of Ye et al., Journal of Neuroscience, 2017.]

Finally, the researchers asked whether snapin overexpression would stave off cognitive deficits in hAPP mice. By injecting AAV with the dynein adaptor gene they restored hAPP animals’ preference for novel, as opposed to familiar, objects, indicating that snapin corrected nonspatial learning and memory deficits. Judging by their ability to learn the location of a hidden platform, and their freezing behavior when placed into a cage where they had previously received a foot shock, the treated mice also performed at or near wild-type levels on spatial and contextual memory tests. Notably, snapin overexpression in wild-type mice had no effects on the animals’ cognition.

Overall, Cai proposes that the buildup of Aβ in BACE1-loaded synapses contributes to their subsequent demise in hAPP mice, and also to learning and memory problems and neurodegeneration. What kick-starts this destructive cascade remains unclear, although Cai speculated that cytoplasmic, soluble Aβ oligomers are to blame because she had previously found that they interfered with the coupling of snapin adaptors to endosomes (see Tammineni et al., 2017). She believes that Aβ oligomers could arise inside the cell, or from outside. The subsequent trafficking defects would then lead to a buildup of endosomes in axons and eventually to dystrophic neurites, where BACE1 accumulates to produce more Aβ and drive the whole process in a destructive feed-forward loop. Given that overexpressing snapin reduced plaque load, Cai also proposed that the synaptic pool of Aβ contributed to plaques.

Claudia Almeida of NOVA University in Lisbon, Portugal, raised the possibility of a slightly different scenario, commenting that snapin’s role in preventing Aβ accumulation could be more closely tied to transport of APP, rather than BACE1, to lysosomes for degradation.

Jochen Herms of the German Center for Neurodegenerative Diseases in Munich said Cai’s findings were impressive but he prefers the idea that Aβ fibrils, rather than soluble oligomers, harm neurites from the outside first, and that this damage ultimately triggers transport problems within. The subsequent accumulation of endosomes in synapses would ultimately lead to the release of intracellular Aβ, which could further contribute to plaques, he said. Herms’s outside-in concept stems from his recent data suggesting that dystrophic neurites only appear after the detection of fibrillar Aβ deposits (see Blazquez-Llorca et al., 2017). 

Cai pointed out that a failure of late endosomes to return to the soma, where they fuse with lysosomes, could also affect lysosomal degradation of proteins more broadly. In her other recent study, which revealed that Aβ oligomers interfere with dynein transport, Cai also reported that other digestive vesicles—autophagosomes—get stuck in synapses, leading to autophagic stress and an accumulation of damaged mitochondria in neurons from AD mice. She proposed that therapies aimed at restoring normal axonal trafficking, rather than solely blocking BACE1, have the potential to correct several cellular malfunctions in AD, and in other neurodegenerative diseases.—Jessica Shugart


  1. The authors report very exciting findings on snapin overexpression being able to revert Aβ accumulation at synaptic terminals and concomitantly improve cognitive performance. I am particularly convinced by the impact of snapin overexpression on amyloid oligomerization in the AD mouse brain. However, I am less convinced about it being mediated by a rescue of BACE1 transport specifically mediated by snapin. I can envision an alternative and also compatible scenario, where APP would be co-transported in endosomes with BACE1 and the lack of retrograde transport could enhance APP processing distally. Or there may even be a more general rescue of axonal impairments previously described in AD (e.g., work from the Goldstein lab and others).

    Caution is also necessary when using CI-MPR as a marker of late endosomes, since MPR cycles between the trans-Golgi network and endosomes, both early and late, to a proportion variable depending on the cell type (Lin et al.,  2004). In neurons, the endosomal localization of MPR has not been characterized. It is also important to note that the early endosomal marker EEA1 is specific to dendrites (Wilson et al., 2000) and thus should not be used to characterize axonal endosomes. 


    . Endocytosed cation-independent mannose 6-phosphate receptor traffics via the endocytic recycling compartment en route to the trans-Golgi network and a subpopulation of late endosomes. Mol Biol Cell. 2004 Feb;15(2):721-33. Epub 2003 Oct 31 PubMed.

    . EEA1, a tethering protein of the early sorting endosome, shows a polarized distribution in hippocampal neurons, epithelial cells, and fibroblasts. Mol Biol Cell. 2000 Aug;11(8):2657-71. PubMed.

  2. This is very interesting work. Ye et al. provide strong evidence that approaches that aim to restore axonal trafficking and lysosomal degradation of BACE1 may represent a potential therapeutic target for AD. It has been known for many years that BACE1 accumulates in AD brains. Several mechanisms have been proposed for such accumulation but evidence is accumulating in support of post-translational mechanisms. Over a decade ago we reported that BACE1 is degraded in lysosomes. We proposed that impaired lysosomal degradation of BACE1 is a leading candidate mechanism underlying BACE1 elevation in AD. Our hypothesis was based on our findings showing that the depletion of the clathrin adaptor GGA3 results in BACE1 accumulation owing to impaired trafficking of BACE1 to lysosomes. In neurons, lysosomal degradation of axonal protein depends on their retrograde transport to the soma, where mature lysosomes reside. Ye et al. clearly showed that impairment of retrograde axonal transport of BACE1 leads to presynaptic accumulation of BACE1 and, as a consequence, increased Aβ production. More importantly, they showed that the overexpression of snapin was able to restore retrograde trafficking of BACE1, reduce synaptic levels of Aβ, and improve cognitive deficits in hAPP Tg mice. Most likely other trafficking molecules in addition to snapin play keys role in BACE1 axonal trafficking and degradation. Thus, more studies are needed on this important topic.

  3. We thank our colleagues for taking the time to read our paper, and also appreciate their positive comments.

    Dr. Almeida’s points are well taken. As we discussed in the manuscript, the Roy lab elegantly showed that APP and BACE1 convergence and APP cleavage occur at presynaptic terminals, triggering amyloidogenesis (Das et al., 2013; Das et al., 2016). They demonstrated that APP and BACE1 are co-transported in axons and interact during this transit. Consistent with these findings, we showed the co-existence of APP and BACE1 in late endosomes that were purified from mouse brains (Ye and Cai, 2014). Our current study provides new lines of evidence that late endosome-loaded APP and BACE1 accumulate within presynaptic terminals of AD neurons. Thus, while we mainly focused on investigating retrograde transport of BACE1, it is likely that these late endosomes are also loaded with APP. Deficiency in snapin-mediated retrograde transport leads to retention of both APP and BACE1 within late endosomes in distal axons and at presynaptic terminals, enhancing β cleavage of APP. Remarkably, we detected the same phenotypes in mutant hAPP mice and snapin-KO mice: more severe accumulation of BACE1 within late endosomes and at presynaptic terminals. Our findings support the notion that defective retrograde transport causes the accumulation of both APP and BACE1 within late endosomes, thereby augmenting synaptic BACE1 processing of APP in AD neurons.

    With regard to Dr. Almeida’s comments on CI-MPR, it was originally shown to be a membrane protein preferentially located in late endosomes (Griffiths et al., 1988). We previously demonstrated that anti-CI-MPR-immunogold specifically labels the luminal vesicles of late endosomes in cortical neurons (Cai et al., 2010). Deleting snapin results in late endosomes clustered in the soma and processes of cortical neurons. In addition to CI-MPR, we alternatively examined the other marker of late endosomes—Rab7. Consistently, we found that BACE1 is retained within Rab7-associated late endosomes at presynaptic terminals of mutant hAPP Tg mice and in AD patient brains. Moreover, a significant portion of BACE1 co-localized and co-migrated with Rab7-labeled late endosomes along the same axon, with a biased long-distance retrograde transport toward the soma of neurons. Retrograde transport of BACE1-loaded late endosomes is impaired in AD axons. Our results are also consistent with previous studies from the Gouras lab showing that late endosomes/MVBs enriched with APP and Aβ42 accumulate in distal axons and synaptic compartments of vulnerable AD neurons (Takahashi et al., 2002Takahashi et al., 2004Gouras et al., 2005), suggesting that defects in retrograde transport results in enhanced BACE1 cleavage of APP within late endosomes.

    In our study, we did not use EEA1 as a marker to label axonal endosomes. As Dr. Almeida pointed out, early endosomal marker EEA1 is specific to dendrites. We found that EEA1 is absent from presynaptic terminals, which excludes the possibility of BACE1 retention within EEA1-associated early endosomes in the axon of AD neurons.


    . Snapin-regulated late endosomal transport is critical for efficient autophagy-lysosomal function in neurons. Neuron. 2010 Oct 6;68(1):73-86. PubMed.

    . Activity-Induced Convergence of APP and BACE-1 in Acidic Microdomains via an Endocytosis-Dependent Pathway. Neuron. 2013 Aug 7;79(3):447-60. PubMed.

    . Visualizing APP and BACE-1 approximation in neurons yields insight into the amyloidogenic pathway. Nat Neurosci. 2016 Jan;19(1):55-64. Epub 2015 Dec 7 PubMed.

    . Intraneuronal Abeta accumulation and origin of plaques in Alzheimer's disease. Neurobiol Aging. 2005 Oct;26(9):1235-44. PubMed.

    . The mannose 6-phosphate receptor and the biogenesis of lysosomes. Cell. 1988 Feb 12;52(3):329-41. PubMed.

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

    . Snapin-mediated BACE1 retrograde transport is essential for its degradation in lysosomes and regulation of APP processing in neurons. Cell Rep. 2014 Jan 16;6(1):24-31. Epub 2013 Dec 27 PubMed.

  4. This study provides interesting new insights into the possible disease relevance of the accumulation of endolysosomal organelles within the dystropic axons that surround amyloid plaques. In a parallel effort, my group has recently identified a distinct JIP3-dependent mechanism that is critical for the axonal transport and maturation of lysosomes and that perturbation of this pathway also results in enhanced APP processing and dramatically worsened amyloid plaque pathology (Gowrishankar et al, 2017).

    Thus, while it has long been known that autophagosomes, lysosomes and hybrids thereof robustly accumulate within the swollen axons around amyloid plaques (Terry et al, 1964; Nixon et al, 2005) and furthermore that such organelles contain enzymes that mediate APP processing (Kandalepas et al, 2013; Yu et al, 2004), new tools to selectively manipulate the axonal transport of such organelles are finally providing evidence that the abnormal axonal accumulation of endolysosomes is of Alzheimer's disease disease relevance.

    Gowrishankar S, Wu Y, Ferguson SM
    Impaired JIP3-dependent axonal lysosome transport promotes amyloid plaque pathology.
    J Cell Biol. 2017 Aug 7; PubMed: 28784610

    Kandalepas, P.C., K.R. Sadleir, W.A. Eimer, J. Zhao, D.A. Nicholson, and R. Vassar. 2013. The Alzheimer's beta-secretase BACE1 localizes to normal presynaptic terminals and to dystrophic presynaptic terminals surrounding amyloid plaques. Acta Neuropathol. 126:329-352.

    Nixon, R.A., J. Wegiel, A. Kumar, W.H. Yu, C. Peterhoff, A. Cataldo, and A.M. Cuervo. 2005. Extensive involvement of autophagy in Alzheimer disease: an immuno-electron microscopy study. Journal of neuropathology and experimental neurology. 64:113-122.

    Terry, R.D., N.K. Gonatas, and M. Weiss. 1964. Ultrastructural Studies in Alzheimer's Presenile Dementia. Am J Pathol. 44:269-297.

    Yu WH, Kumar A, Peterhoff C, Shapiro Kulnane L, Uchiyama Y, Lamb BT, Cuervo AM, Nixon RA. Autophagic vacuoles are enriched in amyloid precursor protein-secretase activities: implications for beta-amyloid peptide over-production and localization in Alzheimer's disease. The international journal of biochemistry & cell biology. 2004;36(12):2531-40. doi: 10.1016/j.biocel.2004.05.010. PMID: 15325590.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Does BACE Drive Neurites into Dystrophy, Shorting Circuits?

Research Models Citations

  1. J20 (PDGF-APPSw,Ind)

Antibody Citations

  1. Soluble Amyloid Oligomers
  2. APP or β-Amyloid (6E10)

Paper Citations

  1. . The Alzheimer's β-secretase BACE1 localizes to normal presynaptic terminals and to dystrophic presynaptic terminals surrounding amyloid plaques. Acta Neuropathol. 2013 Sep;126(3):329-52. PubMed.
  2. . Axonal BACE1 dynamics and targeting in hippocampal neurons: a role for Rab11 GTPase. Mol Neurodegener. 2014 Jan 4;9:1. PubMed.
  3. . Presynaptic dystrophic neurites surrounding amyloid plaques are sites of microtubule disruption, BACE1 elevation, and increased Aβ generation in Alzheimer's disease. Acta Neuropathol. 2016 Aug;132(2):235-56. Epub 2016 Mar 18 PubMed.
  4. . Snapin-regulated late endosomal transport is critical for efficient autophagy-lysosomal function in neurons. Neuron. 2010 Oct 6;68(1):73-86. PubMed.
  5. . Snapin-mediated BACE1 retrograde transport is essential for its degradation in lysosomes and regulation of APP processing in neurons. Cell Rep. 2014 Jan 16;6(1):24-31. Epub 2013 Dec 27 PubMed.
  6. . Impaired retrograde transport of axonal autophagosomes contributes to autophagic stress in Alzheimer's disease neurons. Elife. 2017 Jan 13;6 PubMed.
  7. . High plasticity of axonal pathology in Alzheimer's disease mouse models. Acta Neuropathol Commun. 2017 Feb 7;5(1):14. PubMed.

Further Reading


  1. . Significance of transcytosis in Alzheimer's disease: BACE1 takes the scenic route to axons. Bioessays. 2015 Aug;37(8):888-98. Epub 2015 Jun 30 PubMed.

Primary Papers

  1. . Regulation of Synaptic Amyloid-β Generation through BACE1 Retrograde Transport in a Mouse Model of Alzheimer's Disease. J Neurosci. 2017 Mar 8;37(10):2639-2655. Epub 2017 Feb 3 PubMed.