. The APP Intracellular Domain Is Required for Normal Synaptic Morphology, Synaptic Plasticity, and Hippocampus-Dependent Behavior. J Neurosci. 2015 Dec 9;35(49):16018-33. PubMed.


Please login to recommend the paper.


  1. This paper by Klevanski and colleagues is interesting and shows that the C-terminal domain of βAPP is important for synaptic plasticity as well as for learning- and memory-linked behaviors in mice.

    Several preliminary papers have indicated that APP was involved in several physiological cerebral functions and that the deletion of its C-terminal end bearing a YENPTY domain (ΔCT15) could impair APP trafficking, and thereby its function. This was proposed to be due to a defect in the binding of this domain with various cytosolic signaling proteins. It has also been well described that the presence of endogenous members of the APP family (APLPs) could likely display complementary/redundant functions since single-APP knockout animals were healthy, while mice exhibiting double APP/APLP deficiency showed early lethality. Therefore, the delineation of the respective contribution of membrane-bound APP or APP-derived catabolites was rendered difficult, particularly because APP and APLP both undergo breakdowns by all canonical secretases, which yield similar degradation fragments, except Aβ peptides, which are only derived from APP.

    The authors had already previously described APLP2 knockout mice and APPΔCT15 knock-in mice. The originality of this new article stands in the design of a novel mouse model raised by crossing the APLP2 KO and APPΔCT15 knock-in mice to obtain APPΔCT15-KI/APLP2-KO double-mutant mice (referred to as APPΔCT15-DM). The aim of such design was to prevent any APLP2-mediated contribution to the phenotype triggered by APP-C-terminal deletion. The authors propose that the lack of the APP-CT15 domain yields learning and memory deficits, altered synaptic currents and Aβ generation.

    The main goal of the study, which was to delineate whether APP could function as a cell surface signaling entity or whether some of its various proteolytic fragments could account for its function, remains elusive. The findings do not establish whether CT15 deletion triggers its effect by abolishing the function of βAPP as a membrane-embedded entity or whether it impairs the function of a C-terminal catabolite. This could have been possibly resolved by preventing β- or γ-secretase cleavages in order to examine the putative function of full-length APPΔCT15.

    Concerning the possible contribution of CT15-bearing catabolites, it is extremely puzzling to note that the title of the article states that: “The APP intracellular domain is required for normal synaptic morphology …”! But it is widely understood that the APP intracellular domain corresponds to AICD, the ε-derived C-terminal fragment of APP thought to be the “analog” of the NICD fragment derived from Notch. AICD is regulated by the CT15 domain and displays transcriptional functions that could well account for some of the phenotype reported by Klevanski et al. But AICD is never discussed, even never cited in the article, while it is readily produced from the β-CTF fragment C99 and released in the cytosol. At the end of the day, while the authors report on a novel model in which the observation of CT15 deletion-associated phenotypes cannot be hindered by endogenous APLP2, the study falls short in delineating the respective weights of genuinely membrane-bound APP or its catabolites in the observed phenotypes.

  2. This very interesting study from Klevanski and co-workers follows a large set of analyses of APP mutant mice carrying different deletions. Importantly, it was shown previously that single APP-KO mice have only minor defects, whereas APP/APLP2 DKO mice exhibit severe deficits at the neuromuscular junction (NMJ) early in development. These were not, or were only partially, rescued by expression of sAPP or expression of full-length APP at the neuronal or muscular site, suggesting that APP trans-dimerization, reported by our group and others, might contribute to APP function at the synapse. Further different defects of APP family mutant mice in the central nervous system, including LTP, could be clearly attributed to sAPPα.

    In this study the authors used an APP knock-in mouse (APP∆CT15) lacking the last 15 amino acids at the C-terminus, including the binding site (NPTY motif) for numerous interactors, such as FE65, that may mediate APP signaling. Analysis of APP∆CT15 mice crossed back on an APLP2-deficient background revealed severe defects in NMJ morphology and function and deficits in synaptic plasticity (LTP). Together this study supports the hypothesis that APP family function in CNS and PNS is mediated by both secreted APP fragments, encompassing different sAPP species, and APP full-length protein, and it raises the question of which NPTY motif-interacting proteins might contribute to APP function at the peripheral and central synapse. 

  3. This new study from Ulrike Müller’s lab extends our understanding of the in vivo function of APP and its related family member, APLP2, in the mammalian nervous system. The Müller lab has used traditional, conditional, and knock-in mouse strategies over the last 15 years to produce several mouse strains aimed at understanding the physiological function of APP domains and of the APP protein family. This latest installment shows that a knock-in allele of APP bearing a C-terminal truncation (APP∆CT15) on an APLP2 KO background, APP∆CT15-DM (double mutant), displays attenuated versions of the phenotypes observed in APP/APLP2 DKO mice. The APP∆CT15-DM mice display less perinatal lethality than APP/APLP2 DKO mice, allowing for the characterization of phenotypes in the adult nervous system. The phenotypes observed include neuromuscular junction and associated motor function deficits, as well as impairments in cognitive function that may be due to underlying LTP deficits. Although similar observations were obtained for other DM mice, such as the APPsα-DM mice expressing knock-in alleles of sAPPα (Weyer et al., 2011) or those bearing APPY682G mutant alleles on an APLP2 KO background (APPYG/YG/APLP2—/—, Barbagallo et al., 2011), a closer look at the similarities and differences in phenotypes between these mouse models provides novel insights on the function of the last 15 residues of APP.

    All three double mutant lines, APP∆CT15-DM, APPsa-DM, and APPYG/YG/APLP2—/— mice, display neuromuscular junction deficits and impaired neuromuscular function. In the APP∆CT15-DM study, it is argued that the morphological and functional deficits in muscle function are responsible for the observed perinatal lethality. This is partly because these mice display deficits in sustained muscle contraction, but also because the reduced severity of NMJ deficits in these mice correlates with increased survival when compared to the APP/APLP2 DKO mice. As stated by the authors, this suggests that proper NMJ function is required for survival and that the YENPTY sorting motif of APP bestows an APP function at the neuromuscular junction necessary for survival. 

    One might ask which of the YENPTY domain bearing APP polypeptides, full-length APP, the APP CTFs, or AICD, is essential for normal nervous-system development and function. In the APP∆CT15-DM mice, APP-CTFα levels are presumably unchanged since sAPPα levels are similar to those of APLP2 KO controls in the brain. In contrast APP-CTFβ levels are dramatically reduced, suggesting β-CTFs or AICD may be the essential fragments. The former appears unlikely because of the several BACE1 KO mice previously generated, only one strain displayed some perinatal lethality, and that was proposed to be due to hyperactivity (Dominguez et al., 2005).

    A reduction in AICD could hinder nervous-system development by abrogation of Go complex signaling. However, a C-terminal fragment of APP missing a 19-amino-acid segment at the terminus still binds the Go complex (Nishimoto et al., 1993). What about AICD? Other AICD interactions might require the C-terminus and the YENPTY motif, however.

    Based on the above arguments, dysfunctional full-length APP, or perhaps dysfunctional AICD in the nucleus, may explain the phenotypes of the APP∆CT15-DM mice. However, given that APLP1/APLP2 DKO mice also display NMJ deficits and perinatal lethality (Klevanski et al., 2014) and that it is reasonable to assume a similar molecular mechanism for these phenotypes in all APP/APLP protein family mutant mice, the observation that the γ-secretase-generated APLP1 intracellular domain does not travel to the nucleus or participate directly in nuclear signaling (Gersbacher et al., 2013) suggests that nuclear AICD is not involved in the NMJ deficits or the perinatal lethality observed in APP/APLP protein family mutant mice. Collectively, then, these data support the notion that membrane-associated full-length APP, possibly involving APP family member trans-synaptic dimerization, is implicated in NMJ formation, maintenance and function. 

    One might also ask whether loss of interaction of APP YENPTY-binding proteins is implicated in the observed phenotypes. It is known that X11 and RanBP9 play important roles in the central nervous system and both compete with FE65 for binding to the YENPTY APP sequence. To date, only FE65 protein family KO mice are reported to show deficits in motor behavior (Suh et al., 2015). Further experimental data examining NMJ morphology and ascertaining whether CNS or PNS function is responsible for the motor deficits in FE65 protein family KO mice are needed. Such data would lend support to the notion that FE65 proteins regulate APP function in the nervous system.

    Lastly, it is interesting to note that APPYG/YG/APLP2—/— mice have elevated sAPPα levels in the brain, while APP∆CT15-DM mice have WT sAPPα levels. If one assumes that the spatial memory deficits in both mice arise from the same mechanism, then one would have to conclude that the spatial memory deficits of the APPYG/YG/APLP2—/— mice are not due to elevated sAPPα levels. However, there is accumulating evidence that sAPPα is involved in synaptic plasticity (see discussion by Klevanski et al.) and that sAPPα secretion is regulated at CNS synapses. Thus, regulation of sAPPα secretion may be lost in APP∆CT15-DM mice, while the response to sAPPα may be impaired due to the continual presence of sAPPα in the APPYG/YG/APLP2—/— mice. The cerebral sAPPα levels observed in these mice may be due to the effects of specific complements of APP-bound proteins on APP processing. In the case of APP∆CT15, loss of all YENPTY-interacting proteins is expected, while the APPY682G protein may still bind a subset of the YENPTY interacting proteins. 


    . APP and APLP2 are essential at PNS and CNS synapses for transmission, spatial learning and LTP. EMBO J. 2011 Jun 1;30(11):2266-80. PubMed.

    . The intracellular threonine of amyloid precursor protein that is essential for docking of Pin1 is dispensable for developmental function. PLoS One. 2011;6(3):e18006. PubMed.

    . Phenotypic and biochemical analyses of BACE1- and BACE2-deficient mice. J Biol Chem. 2005 Sep 2;280(35):30797-806. Epub 2005 Jun 29 PubMed.

    . Alzheimer amyloid protein precursor complexes with brain GTP-binding protein G(o). Nature. 1993 Mar 4;362(6415):75-9. PubMed.

    . Differential role of APP and APLPs for neuromuscular synaptic morphology and function. Mol Cell Neurosci. 2014 Jul;61:201-10. Epub 2014 Jul 4 PubMed.

    . Turnover of amyloid precursor protein family members determines their nuclear signaling capability. PLoS One. 2013;8(7):e69363. PubMed.

    . FE65 and FE65L1 amyloid precursor protein-binding protein compound null mice display adult-onset cataract and muscle weakness. FASEB J. 2015 Jun;29(6):2628-39. Epub 2015 Mar 10 PubMed.

Make a Comment

To make a comment you must login or register.