The amyloid precursor protein was first cloned in 1987. Twenty years and 4,000+ papers later, the function of APP is still shrouded in mystery, despite the overwhelming weight of evidence linking the precursor protein to Alzheimer disease. Is the field getting any closer to pinning down a physiological role for APP? Is it involved in membrane trafficking, cell signaling, calcium homeostasis? Theories abound, but which hold most water?

The purpose of the live discussion was to review some of the latest research on APP and try to build a consensus on its physiological function. It also provided a sneak preview of a similarly focused symposium at this year’s Society for Neuroscience meeting in San Diego, California. View details on this symposium. For more on this live discussion, see background text below.

Sanjay W. Pimplikar and Suzanne Guenette led this live discussion on 11 October 2007. Readers are invited to submit additional comments by using our Comments form at the bottom of the page.


Background Text
By Sanjay W. Pimplikar

The cloning of amyloid precursor protein (APP) was reported in 1987, and since then this protein has been a focus of intense study due to its role in generating Aβ. APP is expressed ubiquitously. Early functional information on its role was obtained from in vitro studies on neuronal and non-neuronal tissue culture and cell models. These reports were followed by in vivo studies using animal models that were genetically manipulated to delete or to overexpress APP. Although most in vivo studies utilized the mouse as a model, a number of observations were made in worms, flies, and other rodent systems. As a result, a massive amount of data on the possible function of APP are available, which are reviewed in many excellent recent articles (1-3). Although no definitive function has yet been assigned to APP, recent studies from multiple groups implicate APP in a) larval development and pharyngeal pumping in Caenorhabditis elegans; b) behavioral alterations, axonal transport, dendritic arborization, and neuronal viability in Drosophila; and c) neuromuscular junction formation and neuronal positioning in the developing cortex, and axon outgrowth in mice. In addition, APP has been implicated in cell signaling, gene expression, calcium homeostasis, and membrane trafficking, and this list is likely to be incomplete since APP has been associated with many cellular processes. An ancillary symposium related to this subject matter and entitled "Function of APP Gene Family Members and Clues to AD Pathogenesis: Studies from Worms to Mammals" will be held at the Annual Meeting of the Society for Neuroscience, San Diego on Sunday, 4 November 2007 at 6:30-9:30 p.m.

The primary aim of this live discussion is to invite experts (and interested researchers) in the field to discuss the rather vast amount of data and see if a consensus can be reached about APP's normal physiological function or functions. Although we know how APP deletion or overexpression affects a given experimental parameter, it has proved difficult to translate this knowledge into a definitive framework that brings us closer to its normal physiological function in humans. With this central goal in mind, we plan to discuss and deliberate the available data in three parts:

1. What are the different functional roles ascribed to APP in vitro and in animal systems? As pointed out above, APP has been implicated in a number of functions. Are the data consistent or are there anomalies? Does APP perform a different function in non-neuronal cells? Does AICD really regulate the transcription? If so, how?

2. Is there a common cellular or molecular mechanism that underlies the observed phenomena? This issue has not been addressed systematically. At first glance, APP seems like the “Jack of All Trades,” which performs a wide array of cellular activities. Is there a common unifying mechanism that could connect such diverse functions as formation of neuromuscular junctions, membrane trafficking, behavioral alterations, and postnatal lethality?

3. Finally, if we recognize a unifying cellular function of APP, will it give us a better insight into the pathogenesis of Alzheimer disease? The present attempts to understand the etiology of the disease are focused on the Aβ peptide as a potential proximal agent of pathogenesis. Since APP processing is invariably coupled to APP function, can elucidation of the normal function of APP (and hence the biological necessity of APP processing) lead to a more precise mechanism of AD pathogenesis?

1. Zheng H, Koo EH. The amyloid precursor protein: beyond amyloid. Mol Neurodegener. 2006 Jul 3;1:5. Abstract

2. Senechal Y, Larmet Y, Dev KK. Unraveling in vivo functions of amyloid precursor protein: insights from knockout and knockdown studies. Neurodegener Dis. 2006;3(3):134-47. Review. Abstract

3. Wolfe MS, Guenette SY. APP at a glance. J Cell Sci. 2007 Sep 15;120(Pt 18):3157-61. No abstract available. Abstract


  1. The function of APP is an important question with many possible answers, since, like so many other trans-membrane proteins of similar molecular design, APP is likely to influence many membrane-related functions that depend on the context in which it is studied. Is it also possible that it has its own unique functions which have not yet been revealed? The many "activities" attributed to APP that are described above may well be influenced by it in some way, but these activities may have little to do with the pathological consequences attributed to its proteolytic products, much the same way that angiotensinogen "functions" in the renin/angiotensin cascade.

    However, from the point of view of the AD problem I am intrigued by the following statement: "Since APP processing is invariably coupled to APP function, can elucidation of the normal function of APP (and hence the biological necessity of APP processing) lead to a more precise mechanism of AD pathogenesis?"

    APP processing is certainly coupled to Aβ peptide production, but why is processing necessarily also coupled to other activities of the intact APP molecule? If there is evidence that it is, it would be of great interest to add this to the discussion.

  2. Response to comment by Vincent Marchesi
    Dr. Marchesi correctly captures the essence of the forthcoming Live Discussion by posing the question “…Is it possible that APP has its own unique functions which have not yet been revealed?” and by raising the possibility that the known activities of APP may have little to do with its pathological consequences. Both the Live Discussion and the Ancillary Symposium titled “APP Function and AD Pathogenesis” at the SfN Annual Meeting next month are organized precisely to find answers to such questions. It has been felt that there are lots of data on APP “activities” but little information on its “function.” One reason behind this disconnect could be that the AD field has focused intensely, and for the most part rightly, on “A-betalogy.” I believe we will advance the field substantially if we pause, step back, and ask, Why is APP cleaved in the first place? We hope that these two events will stimulate more researchers to focus on APP function and seek a different angle to study disease pathogenesis. It may well turn out that the knowledge of APP function will not make us any wiser about AD, but this exercise will certainly lead to a better understanding of the biological need of APP processing.

    The intrigue about the statement that “…APP processing is invariably coupled to APP function….” is valid. I am not aware of any study that has directly addressed this issue. This has proved experimentally challenging because there are no mutations in APP that render it completely resistant to cleavage by any of the secretases. Nonetheless, based on the fact that APP undergoes well-orchestrated cleavage events, and that these are conserved throughout evolution, it is safe to believe that APP processing and APP function are tightly linked. The above statement may be unacceptable in a formal scientific paper, but I hope it fits at the informal setting of Alzforum, which strives to stimulate meaningful discussion on important issues.

  3. Yes, this is a great question: "Does APP have a function of its own (i.e., as a full-length protein), or does it function only through the polypeptides that are released through the complex processing of APP?" I am inclined to think that most of the APP functions are associated with the cleavage products, especially since they are distributed so differently within the cell. However, this question is difficult to answer. Immunocytochemical detection in neurons of the endogenous APP, done with various antibodies, surprisingly shows that the epitopes corresponding to different domains of APP are segregated, and mostly localize to distinct regions of the cell. Also, within the neurites, the epitopes from the amino-, carboxy-, and the middle (A-beta) regions of APP are localized to distinct transport carriers, throughout the processes. Of course, these results are difficult to interpret, because one never knows if certain epitopes are indeed not present at a certain location, or if they are blocked from interacting with the antibody because of interactions with other cellular proteins.

    Similarly, exogenous expression is plagued by two problems: first, labeling APP at only one terminus of the protein (e.g., with GFP) leads to the uncertainty of whether what you see is in cells is still full-length APP or cleaved polypeptides. Second, overexpressing APP certainly alters its normal processing, and the results may not be relevant to what is actually happening under physiological levels of expression. From this point of view, even expressing an APP that is tagged at both ends may not be able to tell if the molecule would be intact at endogenous levels of expression (unless the expression level of the double-tagged APP is very low). Using biochemistry to quantify how much APP is intact and how much is present as cleaved polypeptides is not easy. I still need to see a blot that identifies all APP cleavage products present in a brain sample or in cultured cells. Anyway, let's see what the future will bring with regard to the APP functions.

Make a Comment

To make a comment you must login or register.


Webinar Citations

  1. APP Family: Clues to Function in Health and AD

Paper Citations

  1. . The amyloid precursor protein: beyond amyloid. Mol Neurodegener. 2006;1:5. PubMed.
  2. . Unraveling in vivo functions of amyloid precursor protein: insights from knockout and knockdown studies. Neurodegener Dis. 2006;3(3):134-47. PubMed.
  3. . APP at a glance. J Cell Sci. 2007 Sep 15;120(Pt 18):3157-61. PubMed.

External Citations

  1. Society for Neuroscience meeting

Further Reading


  1. . Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science. 1999 Oct 22;286(5440):735-41. PubMed.