Birthing newborn neurons in the adult brain is a complex process with lots of moving parts. Papers in the July 30 eNeuro illuminate the need for two key proteins, both well known for their role in Alzheimer’s disease. Rudolph Tanzi and Se Hoon Choi at Massachusetts General Hospital, Boston, report that lack of BACE1 permits over-proliferation of neural progenitors, but prevents their differentiation into mature neurons. The results imply that BACE1 inhibition, if too powerful, could reduce the number of mature neurons in the hippocampus. Writing in the second paper, Steven Kernie and colleagues at Columbia University, New York, posit that ApoE4 stunts the maturation of newborn neurons in adult mice, thinning their dendritic arbors. Together, these two papers suggest that, at least in the case of adult neurogenesis, BACE inhibition might have a similar effect to ApoE4.

  • Proteins involved in neurodegeneration linked to adult neurogenesis.
  • BACE1 promotes maturation of progenitor cells in the mouse hippocampus.
  • ApoE3, but not E4, supports the growth and branching of dendrites in new neurons.

“These findings point toward the need to understand the neurodevelopmental roles of proteins involved in neurodegeneration while designing potential treatments for AD,” wrote Beate Winner, Friedrich-Alexander-University of Erlangen-Nürnberg, Germany, to Alzforum. Winner was not involved in either study.

Forever Young. Without BACE (right), more newborn neurons proliferate than in mice with BACE (left), but fewer develop into mature neurons. [Courtesy of Emma Brand/Massachusetts General Hospital.]

No BACE1, No New Neurons
BACE1 is best known for making the first enzymatic cut in the Ab precursor protein (APP) as it is processed to release Ab. More recently, the enzyme was recognized for regulating developmental neurogenesis (Hu et al., 2013). Young BACE1 knockout mice had fewer newborn neurons in their hippocampi. Tanzi, Choi, and colleagues wondered if knocking out the enzyme had a similar effect on adult neurogenesis as well.

To find out, co-first authors Zena Chatila, Eunhee Kim, and Clara Berlé used heterozygous BACE+/- mice from the lab of Mark Albers, Massachusetts General Hospital, and bred them to have either two, one, or no copies of the BACE1 enzyme. After aging them to two months, they gave each a single intraperitoneal injection of bromodeoxyuridine (BrdU), a stand-in for thymidine in newly formed DNA, to label newborn neurons. They euthanized half the mice one day after injection and sectioned their brains to see if BACE1 affected proliferation of progenitor cells. The remaining mice were aged for four weeks to give newborn neurons time to mature and integrate into the surrounding circuitry. The researchers then looked for BACE1 effects on survival and differentiation of nascent neurons.

After one day, the number of proliferating cells in the dentate gyri of BACE1 knockout mice almost doubled that in the BACE1 hetero- or homozygotes. However, four weeks later, relatively few of these young cells had matured. In the knockouts, up to 40 percent fewer of the BrdU+ cells expressed NeuN, a marker of neurons, as compared to cells in control mice. Most of these didn’t express doublecortin, olig2, or GFAP either; markers of immature neurons, oligodendrocytes, and astrocytes, respectively. Instead, this pool of neural progenitor cells appeared to stall in an undifferentiated state. There were no differences among BACE1 homo- and heterozygous mice, suggesting proliferation and maturation proceed normally with only partial knockout of the enzyme.

Based on these results, the authors propose that in people with AD, complete inhibition of BACE1 could lead to a pool of cells that get stuck in a stem-cell-like state. “While full BACE1 inhibition is not a goal for patients, 50 percent or slightly more appears fully acceptable,” wrote Stefan Lichtenthaler, German Center for Neurodegenerative Diseases, Munich (see full comment below). “Importantly, for prevention trials it is likely that even less than 50 percent BACE1 inhibition would be sufficient, which should prevent basically all potential side effects.” He suggested repeating the study with a BACE1 inhibitor to determine how much inhibition allows normal neurogenesis.

Carmen Birchmeier, Max Delbrück Center for Molecular Medicine, Berlin, pointed out that it is not yet clear how detrimental it would be if it adult neurogenesis was curtailed. BACE1 cleaves many substrates and BACE KOs have many other troubling phenotypes, including reduced muscle spindle formation, uncoordinated movement, and spontaneous seizures “Independently of this [study], the suggestion not to fully inhibit BACE1 but rather reduce its activity is good, just to be on the safe side.”

ApoE4 Stops Neurons Branching Out
Kernie and colleagues similarly wondered if ApoE influenced adult neurogenesis. They previously reported that the murine version of the protein puts the brakes on progenitor proliferation and prevents stores of neural stem cells from becoming depleted too early (Yang et al., 2011). Does it affect the ongoing development of these cells, too?

Stunted Growth.

Newborn neurons in wild-type mice and those with human ApoE3 have complex dendritic arbors and long dendrites (left panels). When ApoE is knocked out or with human ApoE4, dendrites appear shorter and less complex. [Courtesy of Chatila et al., 2018.]

First author Yacine Tensaouti and colleagues compared adult wild-type mice with ApoE knockouts. They also compared ApoE3 targeted replacement mice to their ApoE4 counterparts. In these animals, human isoforms replace the murine gene. By injecting the dentate gyri of six-week-old animals with a GFP-encoding retrovirus that only incorporates into dividing cells, they specifically labeled the newborn cells in the hippocampi. Four weeks later, they examined them.

Compared with the adult-born neurons in wild-type mice and in animals with human ApoE3, those in both the ApoE knockout and in ApoE4 targeted replacement mice had shorter, sparser dendrites with fewer spines and branch points (see image at right). In all mice, astrocytes that expressed ApoE wrapped around these new dendrites. The results hint that in astrocytes murine ApoE and human ApoE3 give dendrites a maturational boost, while ApoE4 does not. 

“We know that ongoing neurogenesis is important for keeping a healthy, dynamically functioning hippocampus,” Kernie told Alzforum. “What this study shows is that if mice have ApoE4, their ability to do that is not as robust as in mice that express ApoE3. That probably impairs the ability of the hippocampus to develop normally and to respond to injury.” However, it is still unclear how neurogenesis influences AD in people, he noted.

Guojun Bu, Mayo Clinic, Jacksonville, Florida, who was not involved in the study, thought it was well-executed. He still thinks the main reason for the increased AD risk brought by ApoE4 is that the protein drives earlier and more abundant amyloid pathology. However, there is increasing evidence that ApoE4 also has a loss of function compared to E3. “In a model such as this, where there’s no AD-related pathology, a deficiency of ApoE4 supports this notion.” He cautioned that the authors did not test whether the differences they observed in neurogenesis and morphology had any impact on function.—Gwyneth Dickey Zakaib


  1. This is an interesting paper adding to our understanding of physiological BACE1 functions. It demonstrates that BACE1 controls neuronal differentiation. This study is different from a previous paper by Riqiang Yan (Hu et al., 2013) in that it now analyzes adult hippocampal neurogenesis (AHN), which is linked to life-long learning and may be impaired in AD.

    Full BACE1 inhibition—as shown in this paper—would also impair AHN and should obviously be avoided by therapeutic BACE inhibition. However, the study also demonstrates that heterozygous BACE1 KO mice have no defect in AHN. Thus, while full BACE1 inhibition is not a goal for patients, 50 percent or slightly more appears fully acceptable. And that is exactly what is currently done in the clinical trials with BACE inhibitors. Importantly, for prevention trials it is likely that even less than 50 percent of BACE1 inhibition would be sufficient, which should prevent basically all potential side-effects.

    Going forward, it would be important to repeat this study with a pharmacological BACE1 inhibitor to determine to what extent BACE1 can be inhibited without side effects on AHN. In fact, out of the many phenotypes reported in BACE1-deficient mice, only very few have so far been recapitulated in inhibitor-treated adult mice, in particular muscle-spindle alterations (Cheret et al.,  2013) and spine density and LTP defects (Zhu et al., 2018). Additionally, it would be great to establish which BACE1 substrate contributes to the AHN phenotype. An obvious candidate is Jagged 1, based on the previous study from Riqiang Yan. 

    In summary, this beautiful new study enlarges the spectrum of physiological BACE1 functions. With regard to clinical BACE1 inhibitors, it reinforces the general view in the field that BACE1 inhibition of somewhat more than 50 percent may be acceptable and should not come too close to 100 percent. This is currently achieved in clinical trials and suggests that mechanism-based side effects resulting from BACE1 inhibition may be controlled in patients. 


    . BACE1 regulates hippocampal astrogenesis via the Jagged1-Notch pathway. Cell Rep. 2013 Jul 11;4(1):40-9. PubMed.

    . Bace1 and Neuregulin-1 cooperate to control formation and maintenance of muscle spindles. EMBO J. 2013 Jun 21; PubMed.

    . Beta-Site Amyloid Precursor Protein Cleaving Enzyme 1 Inhibition Impairs Synaptic Plasticity via Seizure Protein 6. Biol Psychiatry. 2016 Dec 26; PubMed.

  2. The findings of Chatila and colleagues in BACE1 knockout mice—namely increased proliferation, decreased neurogenesis, and accumulation of undifferentiated neural precursor cells (NPCs) in the subventricular zone of the adult hippocampal dentate gyrus (DG)—are interesting and potentially important considerations for chronic use of BACE inhibitors in Alzheimer’s disease patients. These results from adult mice are consistent with previous observations of decreased DG neurogenesis (Hu et al., 2013), accumulation of clusters of Doublecortin (Dcx)-positive immature neuronal cells in the sub-pial zone and delayed migration during development of the DG (Hou et al., 2017). Collectively, these results indicate that the abnormalities in neurogenesis in the BACE1 constitutive knockout model, even in the adult, are likely to have arisen from the lack of BACE1 throughout development.

    Whether these observations have a major bearing on possible consequences of inhibiting BACE1 in the mature or the aging brain depends not only on the level of inhibition (current BACE inhibitor trials are targeting ~60-75 percent inhibition) but also whether adult neurogenesis is occurring at an appreciable level in the mature and aging human brain and important for memory. Recent literature highlights the ongoing controversy surrounding this issue (Sorrells et al., 2018; Boldrini et al., 2018), however, if adult hippocampal neurogenesis does not decline dramatically with age, then there is clearly a strong imperative to identify the BACE1 substrate/s involved and develop substrate-sparing inhibitors.


    . BACE1 regulates hippocampal astrogenesis via the Jagged1-Notch pathway. Cell Rep. 2013 Jul 11;4(1):40-9. PubMed.

    . BACE1 Deficiency Causes Abnormal Neuronal Clustering in the Dentate Gyrus. Stem Cell Reports. 2017 Jul 11;9(1):217-230. Epub 2017 Jun 29 PubMed.

    . Human hippocampal neurogenesis drops sharply in children to undetectable levels in adults. Nature. 2018 Mar 15;555(7696):377-381. Epub 2018 Mar 7 PubMed.

    . Human Hippocampal Neurogenesis Persists throughout Aging. Cell Stem Cell. 2018 Apr 5;22(4):589-599.e5. PubMed.

  3. Chatila and colleagues have identified a role for BACE1 in the regulation of hippocampal neurogenesis in adult mice, raising a new concern regarding the use of BACE inhibitors for the treatment of Alzheimer’s disease. The increase in undifferentiated neural precursor cells (NPCs) seen in the dentate gyri of constitutive BACE1 KO mice adds to the physiological and behavioural changes identified in these mice, which include hypomyelination, axon guidance errors, and alterations in synaptic number and function (Munro et al., 2016). It will be important to identify which BACE1 substrate(s) contribute to this phenotype, including APPα/β as considered by the authors; BACE1 has a number of validated substrates expressed in the adult hippocampus such as seizure-related gene 6 (Sez6), Sez6-like (Pigoni et al., 2016) and CHL1. Additionally, it remains to be determined what the functional effect will be of this increase in undifferentiated cells in the hippocampus.

    It is now essential to determine whether the alterations in neurogenesis observed by Chatila et al. are seen in mature, wild-type mice treated with BACE inhibitors, or whether this phenomenon is at least partly the result of other (e.g. compensatory) changes resulting from constitutive BACE1 KO. This study also reinforces the importance of identifying the optimal level of BACE1 inhibition for the treatment of Alzheimer’s disease which will slow disease progression while minimizing potential side effects. Unlike BACE1 KO mice, BACE1 heterozygous mice show no changes in NPC proliferation and may not have a decrease in neuronal differentiation of NPCs (although, as the authors discuss, this point is not completely clear). Clinically, the most appropriate level of inhibition of BACE1 proteolytic activity remains to be determined, although complete inhibition is not the aim.

    Past animal studies have shown that BACE inhibition can result in changes to dendritic spine plasticity (Filser et al., 2015; Zhu et al., 2016) and cognitive deficits (Filser et al., 2015). It is important to ascertain any additional mechanism-based side effects of BACE inhibition that could counteract the potential benefits of this treatment strategy, particularly in light of the recent clinical trials of BACE1 inhibitor verubecestat, which had no effect on cognitive decline in patients. Here, Chatila et al. have identified a physiological function of BACE1 that could significantly impact on plasticity and memory function and should be given attention in future studies involving BACE inhibitors.


    . Functions of the Alzheimer's Disease Protease BACE1 at the Synapse in the Central Nervous System. J Mol Neurosci. 2016 Nov;60(3):305-315. Epub 2016 Jul 25 PubMed.

    . Seizure protein 6 and its homolog seizure 6-like protein are physiological substrates of BACE1 in neurons. Mol Neurodegener. 2016 Oct 5;11(1):67. PubMed.

    . Pharmacological inhibition of BACE1 impairs synaptic plasticity and cognitive functions. Biol Psychiatry. 2015 Apr 15;77(8):729-39. Epub 2014 Oct 29 PubMed.

    . Beta-Site Amyloid Precursor Protein Cleaving Enzyme 1 Inhibition Impairs Synaptic Plasticity via Seizure Protein 6. Biol Psychiatry. 2016 Dec 26; PubMed.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . BACE1 regulates hippocampal astrogenesis via the Jagged1-Notch pathway. Cell Rep. 2013 Jul 11;4(1):40-9. PubMed.
  2. . ApoE is required for maintenance of the dentate gyrus neural progenitor pool. Development. 2011 Oct;138(20):4351-62. PubMed.

Further Reading


  1. . Impact of the apolipoprotein E polymorphism, age and sex on neurogenesis in mice: Pathophysiological relevance for Alzheimer's disease?. Brain Res. 2013 Oct 16; PubMed.
  2. . Adult neurogenesis and neurodegenerative diseases: A systems biology perspective. Am J Med Genet B Neuropsychiatr Genet. 2017 Jan;174(1):93-112. Epub 2016 Feb 16 PubMed.

Primary Papers

  1. . BACE1 Regulates Proliferation and Neuronal Differentiation of Newborn Cells in the Adult Hippocampus in Mice. eNeuro. 2018 Jul-Aug;5(4) Epub 2018 Aug 3 PubMed.
  2. . ApoE Regulates the Development of Adult Newborn Hippocampal Neurons. eNeuro. 2018 Jul-Aug;5(4) Epub 2018 Aug 2 PubMed.