This is part 3 of our 3-part series. Also see part 1 and part 2, or download PDF.

BACE Biology: Who Are Its Handlers?
Investigators are probing intensely the question of which other proteins interact with BACE. Perhaps the interacting proteins control BACE activity, and in this way influence Aβ production? Among the list of proteins thought to bind BACE, the nogo family is emerging as an intriguing group. Long studied for its ability to repulse outgrowing neurites in the brain and spinal cord, nogo and its receptor are forming a point of convergence between the formerly separate problems of axonal regeneration and AD pathology.

In Madrid, Rabinder Prinjha of the GlaxoSmithKline research group in Harlow, Great Britain, asked how nogo modulates APP processing via its link to BACE. Prinjha first noted that nogo-A and the nogo receptor are members of a large family of proteins, called reticulons, the physiological function of which is barely known. Four human reticulon genes are known to date with reticulon-4A being better known as nogo-A.

Nogo-A (reticulon4-A) and BACE occur together in the endoplasmic reticulum of cultured cells, Prinjha said. To test if they might interact in vivo, his team looked in APP- and APP/PS1-transgenic mouse lines and saw nogo-A and BACE both being upregulated in cortical areas surrounding plaques. Then the researchers tested the effect of nogo-A and its relatives in the reticulon family on APP processing, and noticed that they all affected Aβ production in distinct ways. In short, both overexpression and RNAi knockdown studies were used to confirm that reticulon3-A1 functions to decrease Aβ production, whereas reticulon4-A (nogo-A) increases it. The precise location of the proteins within the cell drove the effect. Prinjha did not address the mechanism of the interaction, but said that he suspects it to change the endocytosis or subcellular localization of BACE. Notably, besides the large pool in the ER, a second, smaller pool of nogo-A occurs on the cell surface. Similarly, Wataru Araki, of Japan’s National Institute of Neuroscience in Tokyo, reported that reticulon3 and reticulon4-B and -C bind BACE1. (Reticulons4-B/C are smaller isoforms of reticulon4). He suggested that these reticulons appear to inhibit BACE1’s ability to cleave APP by some interaction with BACE1 outside of the enzyme’s active site. For more on this up-and-coming topic, see Park et al., 2006; Gil et al., 2006; Yan et al., 2006).

Different sorting proteins also appear able to control APP processing by directing BACE¹s journey through intracellular compartments, primarily between endosomes and the trans-Golgi network. They include GGA1 (see ARF 2006 Eibsee conference report; He et al., 2005) and sortilin. On the latter, Gina Finan, working with Tae-Wan Kim at Columbia University in New York, reported that postmortem brain samples of AD patients contain less sortilin than controls. Sortilin forms a complex with BACE1 and appears to reduce Aβ secretion through its role in trafficking BACE1, the researchers suggest.

BACE in the Aging Brain: What Goes Wrong?
A number of groups have established that BACE activity tends to go up with age, and more steeply in AD. What could cause this? One hypothesis came to the fore when Robert Vassar’s group at Northwestern University in Chicago picked up findings from brain imaging, which has shown mild hypometabolism in the aging brain and even more in the AD brain. Wondering if the BACE increase might follow diminished perfusion—that is, insufficient oxygen and glucose supplies to the brain—Rodney Velliquette began to model energy starvation. Last November, the scientists reported that a single injection of chemical inhibitors of ATP generation had a long-lasting effect on the brain in that BACE levels (and Aβ production) shot up, and stayed up for a week (Velliquette et al., 2005). Reflecting Citron’s comment about the importance of post-translational modifications, Vassar noted that this increase occurred at the level of BACE protein, not gene expression.

But one injection does not model a slow disease such as AD, and in Madrid, Vassar followed up with a second, chronic study. It mildly starved ATP production in Tg2576 mice for a 3-month period, beginning prior to amyloid deposition at 9 months of age until 12 months of age when plaques are forming. BACE, Aβ levels, and plaque load all went up in the treated mice, Vassar reported. This suggests that, perhaps, sporadic AD could have an upstream, stress-related beginning that would drive BACE. “It is established that cerebral blood flow decreases in aging and particularly in AD brain. We do not know if this is just a correlation or a pre-existing driving force. It is something to look into because aging is the major risk factor in AD. As we age, cardiovascular disease increases and could put the brain under chronic energy stress,” Vassar speculated.

To sort out the time course of these events, Vassar’s group needed to make a better monoclonal antibody. All antibodies they could get their hands on were “dirty,” showing non-specific binding on Western blots and in brain tissue. With Skip Binder of Northwestern, who is noted for his skill in generating antibodies, the Chicago scientists used BACE knockout mice as immunization hosts, because they have never seen this protein. Out came a cleaner anti-BACE monoclonal antibody that recognizes but a single band on blots, Vassar said. This antibody helped the scientists characterize the BACE increase in various materials, including the Tg2576 and the group’s aggressive 5x-transgenic strain (see ARF SfN meeting story). The BACE antibody co-stains with neuronal markers but not astrocytic ones, particularly in dystrophic neurons. What’s more, BACE staining correlates with plaque development, and visualizes BACE around plaque cores. The BACE increase appears to be associated with Aβ42 deposition. This raises a chicken-and-egg question about which comes first in the course of AD pathogenesis: Does BACE first go up and induce Aβ42 deposition, or do Aβ42 deposits induce a secondary BACE increase? Vassar suspects a stress-induced feedback loop at play here.

The new antibody, Vassar hopes, will help with the analysis of what upstream factors can trigger BACE. Vassar particularly wonders whether any of those upstream factors would make for a good drug target so that, ultimately, a drug would become available that prevents only the age- or AD-related BACE increase, not all BACE altogether. “We may need BACE around for other functions,” Vassar said.

BACE: The Newest Biomarker?
Last but not least, one ICAD presentation moved research on BACE into the bustling realm of biomarker research. Yong Shen at Sun Health Research Institute in Sun City, Arizona, reported results of a collaborative study that, tantalizingly, suggested BACE1 might make for a decent biomarker. Shen’s lab was among the first to notice the BACE1 increase in AD brain, (see Li et al., 2004), a finding others have since confirmed. In Madrid, Shen used the same ELISA assay used for that study to assess BACE1 levels in the CSF of 80 sporadic AD cases, 59 MCI cases, and 69 age-matched controls. His data suggest that BACE levels in these MCI cases were twice as high as in the healthy controls but—surprisingly—returned to control levels once a person has progressed to overt AD. If replicated, this data would suggest that BACE1 might eventually serve to predict AD. Much remains to be sorted out about this prospect. In Shen’s hands, a BACE activity assay tracked with the ELISA for BACE detection and with total Aβ levels in that all three measures correlated in the CSF. These are early days for BACE1 research in CSF, but one study published to date would tend to confirm that enzymatically active BACE1 can be detected in human CSF (Verheijen et al., 2006).

All told, AD researchers still consider BACE the ideal target to test the amyloid hypothesis of Alzheimer disease. To be sure, BACE1 gets more complex the deeper the field digs into its function and regulation. Still, many scientists agree that inhibiting BACE1 may be cleaner than inhibiting γ-secretase, because that enzyme complex has a startling array of functions other than snipping APP once BACE is through with it. If a BACE inhibitor were to get into the brain and remove Aβ as expected, yet failed to improve dementia, then the amyloid hypothesis would be in serious jeopardy. With such inhibitors now coming online, this day of reckoning may be drawing nearer.—Gabrielle Strobel.


  1. Previous studies by us have demonstrated increased BACE1 protein activity levels in both postmortem and antemortem CSF of subjects with AD compared to controls (Holsinger et al., 2004). Surprisingly, we also detected full-length BACE1 protein, indistinguishable from that observed in human brain (Holsinger et al., 2002), in CSF by Western blotting. This immunoreactive species was, however, different from that observed in another recent study (Verheijen et al., 2006). Therefore, the finding by Shen and colleagues is of great interest as the differences in the three studies may be attributable to the different species recognized by the antibodies.


    . Increased beta-Secretase activity in cerebrospinal fluid of Alzheimer's disease subjects. Ann Neurol. 2004 Jun;55(6):898-9. PubMed.

    . Increased expression of the amyloid precursor beta-secretase in Alzheimer's disease. Ann Neurol. 2002 Jun;51(6):783-6. PubMed.

    . Detection of a soluble form of BACE-1 in human cerebrospinal fluid by a sensitive activity assay. Clin Chem. 2006 Jun;52(6):1168-74. PubMed.

  2. I have been working on the regulation of BACE activity together with Rudy Tanzi, and presented our new data at the ICAD meeting. Our data show that GGA3 regulates BACE by modulating BACE degradation in the lysosome. We have found that caspase-3 degrades GGA3 during cerebral ischemia, leading to increased levels of BACE. This then leads to increased Aβ production associated with ischemia. Most importantly, we also found that GGA3 levels are decreased in the AD brain and inversely correlate with increases in BACE levels.

    We believe this is the first study elucidating a molecular mechanism regulating BACE activity. Regarding the physiological role of BACE, data from many labs, including ours, support the hypothesis that BACE is a stress-induced protease in the adult brain. Although during development, BACE upregulation may play a role in the processing of some substrates, BACE does not seem to be necessary to reach adulthood without an abnormal phenotype. The likely explanation is that the BACE substrates (including neuregulin-1) are also substrates for α-secretase.

  3. It’s of great interest that Guiseppina Tesco and Rudy Tanzi find that caspase-3 degrades GGA3, which leads to increased levels of BACE, and that GGA3 levels are decreased in the AD brain. I refer to the Greeve et al. study (1) in which they report that expression of seladin-1 reduces caspase-3 activation. I have proposed that reduced ACTH secretion as a response to chronic hypercortisolism may explain the reduced seladin-1 reported in AD (2). It would seem that the result may then be increased caspase-3-mediated degradation of GGA3.

    A couple of interesting studies report that RNA interference (RNAi) of GGA3 expression results in accumulation of the cation-independent mannose-6 phosphate receptor and internalized epidermal growth factor (EGF) within enlarged early endosomes, and PS-FAD expressed higher levels of the cation-independent mannose-6 phosphate receptor (3,4).

    See also:

    Comment by Mary Reid


    . The human DIMINUTO/DWARF1 homolog seladin-1 confers resistance to Alzheimer's disease-associated neurodegeneration and oxidative stress. J Neurosci. 2000 Oct 1;20(19):7345-52. PubMed.

    . Interactions of GGA3 with the ubiquitin sorting machinery. Nat Cell Biol. 2004 Mar;6(3):244-51. PubMed.

    . Presenilin mutations in familial Alzheimer disease and transgenic mouse models accelerate neuronal lysosomal pathology. J Neuropathol Exp Neurol. 2004 Aug;63(8):821-30. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Madrid: BACE News Roundup, Part 1
  2. Madrid: BACE News Roundup, Part 2
  3. Assembly, Traffic Escorts, Fats All Control APP Processing
  4. SfN: Where, How Does Intraneuronal Aβ Pack Its Punch? Part 1

Paper Citations

  1. . Alzheimer precursor protein interaction with the Nogo-66 receptor reduces amyloid-beta plaque deposition. J Neurosci. 2006 Feb 1;26(5):1386-95. PubMed.
  2. . Nogo-A expression in the human hippocampus in normal aging and in Alzheimer disease. J Neuropathol Exp Neurol. 2006 May;65(5):433-44. PubMed.
  3. . Reticulon proteins: emerging players in neurodegenerative diseases. Cell Mol Life Sci. 2006 Apr;63(7-8):877-89. PubMed.
  4. . GGA proteins mediate the recycling pathway of memapsin 2 (BACE). J Biol Chem. 2005 Mar 25;280(12):11696-703. PubMed.
  5. . Energy inhibition elevates beta-secretase levels and activity and is potentially amyloidogenic in APP transgenic mice: possible early events in Alzheimer's disease pathogenesis. J Neurosci. 2005 Nov 23;25(47):10874-83. PubMed.
  6. . Amyloid beta peptide load is correlated with increased beta-secretase activity in sporadic Alzheimer's disease patients. Proc Natl Acad Sci U S A. 2004 Mar 9;101(10):3632-7. PubMed.
  7. . Detection of a soluble form of BACE-1 in human cerebrospinal fluid by a sensitive activity assay. Clin Chem. 2006 Jun;52(6):1168-74. PubMed.

Other Citations

  1. PDF

Further Reading