Combining their individual fortés, the laboratories of Dave Holtzman at Washington University, St. Louis, and Brian Bacskai and Brad Hyman at Massachusetts General Hospital, Charlestown, treat us to the sight of neurons being treated before our eyes. In a study published online January 20, 2005, in the Journal of Clinical Investigation, fluorescently labeled dystrophic neurites in the cortex of PDAPP mice appear to shrink or disappear just days after Aβ antibodies are applied to the cortical surface.

For several years, Bacskai and Hyman have been wowing people with their "windows on the brain," openings in the skulls of mice through which they apply drugs to the cortex and then observe the results with two-photon microscopy. They have already visualized how plaques grow in AD transgenic mice, and have shown that amyloid plaques are cleared days after the application of Aβ antibodies to the cortical surface (Bacskai et al., 2001; Bacskai et al., 2002). Bacskai and Hyman have also assessed neuritic pathology in PDAPP mice, finding postmortem evidence that the Aβ antibodies can rapidly reduce neuritic pathology (Lombardo et al., 2003), though the researchers were unable to visualize the neurites through the windows.

Enter the glowing yellow mouse from Holtzman's lab. Holtzman and colleagues have crossed PDAPP mice with transgenic mice that express yellow fluorescent protein (YFP) throughout the neuronal cytosol, reaching into distal neurite tips. With these mice, the researchers have found evidence that neuritic dystrophy is much more extensive than was previously thought (Brendza et al., 2003).

Neurite stakeout detects disappearing pathology
The image on the left shows yellow fluorescent protein (YFP)-labeled neurites—imaged through a cranial window with two-photon microscopy—in a living PDAPP;thy-1:YFP double-transgenic mouse. Enlarged, bulbous, dystrophic neurites surrounding an amyloid plaque (not visualized in this image) can be seen. On the right, the same plaque three days after the cortical surface was treated with 10D5 anti-Aβ antibody. There is a reduction in YFP-labeled dystrophic neurites, and the arrows show two dystrophic areas that are present at day 0 and absent at day 3. (Scale bar = 10μm) [images courtesy of Bob Brendza, Washington University]

Bob Brendza in Holtzman's lab led the current collaboration, which also included Bill Klunk and Chet Mathis at the University of Pittsburgh, and Steve Paul and Kelly Bales at Eli Lilly. First, the researchers just watched, and noted that the numbers of dystrophic neurites surrounding individual plaques did not change over a period of three days. Similarly, individual dystrophic neurites under scrutiny were stable during this period. However, when they applied 10D5 Aβ antibodies to the cortical surface, the researchers observed significant morphological changes within only three days. They witnessed a substantial reduction, or disappearance altogether, of smaller dystrophic neurites, as well as the reduction or elimination of dystrophic regions on otherwise normal-looking dendrites or axons. Some of these neurites were followed for up to a week, with no return of the pathology. At the same time, however, larger dystrophic neurites appeared impervious to the antibody treatment, leading the authors to suggest that additional doses, or time, might be needed to get a more widespread benefit.

This method did not allow the researchers to quantify changes in amyloid burden. In a separate series of experiments, they found that the passive immunization reduced total Aβ and thioflavin S-positive plaques but had no effect on levels of PBS-soluble Aβ. This, Brendza and colleagues suggest, would lead to the conclusion that fibrillar Aβ is what the antibodies are targeting.

"Based on these data, it appears that axonal and dendritic structural damage associated with amyloid deposits is not permanent and is, at least in part, reversible over a relatively short time frame. Further, the specificity of the antibody shows that Aβ itself causes these reversible structural changes," write the authors. —Hakon Heimer

Q&A with Bob Brendza.

Q: How do your results extend those of your collaborators at MGH, who looked at effects on neurites following cortical administration of Aβ antibodies about a year and a half ago?
A: The Lombardo et al. paper was a postmortem study. It reported the very important finding that mice receiving antibody treatment had a significant reduction in abnormal neurite geometry that correlated with Aβ clearance. They didn't see much of a change in neuritic dystrophy when assessing it with APP immunoreactivity. YFP expression is a very sensitive marker for assessing neuritic dystrophy and is actually more sensitive than silver or APP staining in detecting dystrophic areas (especially smaller dystrophic swellings). Our 2003 J. Comp. Neurology paper actually compared the different methods. So, the three major differences between our JCI paper and the Lombardo et al. paper was that our study was in vivo, we tracked neuritic dystrophy in individual neuritic plaques, and we used a more sensitive method to monitor neuritic dystrophy and neurite morphology.

The other advantages of using YFP are that YFP requires no exogenous cofactors or substrates to fluoresce and diffuses freely throughout the cytoplasm; thus, it fluorescently labels all neuronal processes including axons, dendrites, and dendritic spines. These features of YFP allow neurites to be easily visualized in live tissue without any potentially disruptive manipulations. Since cytoplasmic YFP labels all neuronal processes indiscriminately, the morphology and dynamics of normal and dystrophic neurites of labeled neurons in PDAPP; YFP double Tg mice can be observed without concerns regarding tissue penetration or alteration of target molecules by the AD pathogenic environment, which are potential problems associated with the use of traditional histological probes.


  1. "Until the Last Dog(ma) Dies": Some Neuritic Dystrophy Is Reversible by Passive Immunization of PDAPP Mice
    A multidisciplinary group has demonstrated that at least some neuritic dystrophy in PDAPP mice is reversible. Holtzman from Wash U, Paul from Lilly, Mathis and Klunk from Pitt, and Bacskai and Hyman from MGH contributed their considerable talent to a new paper in the current issue of The Journal of Clinical Investigation. Using the open skull method and Congo red derivative methoxy-X04 devised by the MGH and Pittsburgh groups, respectively, the team followed with serial imaging the morphology of swollen (dystrophic) neurites surrounding cortical amyloid deposits in the PDAPP mouse. Conventional wisdom would have predicted that these swellings might be permanent, but the new paper describes how passive immunization with anti-Aβ antibodies had a significant effect on partially normalizing the shapes of the processes.

    The new paper builds on earlier work by the MGH group (Lombardo et al., 2003): The advance of the JCI paper is to study the same plaques serially during life, while the Lombardo paper relied on postmortem analysis. A key novelty is directly demonstrating a significant trend toward normalization of existing dystrophic neurites, i.e., watching the same, flagrantly abnormal neurites partially recover their normal morphology, thereby unequivocally documenting the reversibility of neuritic dystrophy.

    This dovetails well with the recent report in Neuron by Oddo and colleagues (see ARF related news story) who showed that amyloid deposits and neuritic antigens were attenuated by active immunization. Janus and colleagues (Janus et al., 2000) demonstrated Aβ-dependent behavioral deficits that were reversible with active immunization, and the Holtzman paper suggests that the same may be true for passive treatment. Together, the papers portend well for the principle that Alzheimer's pathology may be treatable, even after significant dystrophy has developed.

    This is important because the most sensitive means of detecting Alzheimer's today remains neuropsychological testing, implying, by definition, the existence of pathology and deficits that one would like to reverse, if possible. Interestingly, a recent PNAS paper (Lu et al., 2005) indicates that iconic memory deficits may predict Alzheimer's even before any important deficit is present or even otherwise detectable. All the usual mice-aren't-men caveats apply, of course, but these papers are very exciting because frontiers have been pushed back both in terms of early diagnosis as well as providing optimism for reversibility of pathology.


    . Amyloid-beta antibody treatment leads to rapid normalization of plaque-induced neuritic alterations. J Neurosci. 2003 Nov 26;23(34):10879-83. PubMed.

    . A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer's disease. Nature. 2000 Dec 21-28;408(6815):979-82. PubMed.

    . Fast decay of iconic memory in observers with mild cognitive impairments. Proc Natl Acad Sci U S A. 2005 Feb 1;102(5):1797-802. PubMed.

  2. The authors used multiphoton microscopy, an elegant technique, to monitor the dynamics of neuritic plaques in living mice. They used the PDAPP;Thy-1:YFP transgenic mouse model, which develops plaque pathology and expresses yellow fluorescent protein in a subset of neurons. Through cranial windows, Aβ deposits were analyzed with injected methoxy-X04, and dystrophic neurites with YFP-induced fluorescence. Over a period of 72 hours, the amyloid-associated neurites remained stable. However, after application of the anti-Aβ antibody 10D5 to the cortical surface, the number and total cross-sectional area of dystrophic neuritis decreased significantly. This clearly demonstrates again the value of passive immunization to reduce extracellular plaque load and the associated neuritic pathology.

    Although these results are very promising, novel transgenic mouse models teach us that extracellular amyloid plaques are not a major trigger for the dramatic neuron loss and brain atrophy. On the contrary, amyloid plaques do not correlate with the hippocampal neuron loss in the transgenic models (Schmitz et al., 2004; Casas et al., 2004). In both models, the increased amount of intraneuronal Aβ42 correlates best with the loss of neurons and brain tissue. These models support the idea that neuron loss and atrophy in AD is triggered by intracellular accumulation of Aβ42, and not by extracellular amyloid pathology.

    It is very likely that intraneuronal Aβ accumulation leads not only to neuron loss, but also disrupts many neuronal functions, for example, axonal transport. Therefore, it will be very important to learn more about the influence of passive immunization on the potential to reduce intraneuronal Aβ levels as has been shown recently (Oddo et al., 2004). Using a triple-transgenic model (3xTg-AD) that develops plaques and tau lesions, the authors showed that Aβ immunotherapy reduces not only extracellular Aβ plaques, but also intracellular Aβ accumulation, and leads to the clearance of tau hyperphosphorylation.

    Convincing evidence that passive immunization is able to rescue neuronal dysfunction and neuron loss is, however, still lacking.


    . Time sequence of maturation of dystrophic neurites associated with Abeta deposits in APP/PS1 transgenic mice. Exp Neurol. 2003 Nov;184(1):247-63. PubMed.

    . Hippocampal neuron loss exceeds amyloid plaque load in a transgenic mouse model of Alzheimer's disease. Am J Pathol. 2004 Apr;164(4):1495-502. PubMed.

    . Massive CA1/2 neuronal loss with intraneuronal and N-terminal truncated Abeta42 accumulation in a novel Alzheimer transgenic model. Am J Pathol. 2004 Oct;165(4):1289-300. PubMed.

    . Abeta immunotherapy leads to clearance of early, but not late, hyperphosphorylated tau aggregates via the proteasome. Neuron. 2004 Aug 5;43(3):321-32. PubMed.

    View all comments by Thomas Bayer
  3. This paper by the Holtzman group supports the original work of the Bacskai/Hyman group that also showed that reduction of Aβ in the brain can induce a rapid structural recovery of existing amyloid-associated neuritic dystrophy. The originality of this new paper is that the analysis was performed in a living mouse brain as opposed to postmortem in the original study published in 2003 (Lombardo et al.). The data presented support the AD-amyloid hypothesis and, indeed, suggest that reducing Aβ (via an immunotherapy, at least) will be effective.

    I would have liked to see if the same effect could be observed if the antibody is injected into the blood. Would this also reduce neurite dystrophy over the same time period? This is worthy of demonstration since, realistically, the current AD immunotherapy in development will require the injection of humanized Aβ antibodies into the bloodstream.

    Moreover, although the study clearly demonstrates histological improvements (reduction of neurite dystrophy) after the antibody treatment, it is important in the future to demonstrate if this also leads to a recovery of cognitive functions. It could also be very interesting to repeat the experiment with two kinds of antibody: one that recognizes only the monomeric form of Aβ and one that recognizes only its fibrillar form. This would answer this burning question: Shall companies develop an antibody that recognizes the fibrillar form of Aβ or its monomeric form (sink hypothesis)?

    Finally, the fact that the reduction of neuritic dystrophy can only be seen on the side ipsilateral to 10D5 application (when applied to 17-18-month-old PDAPP mice) suggests that the effect won't be observed unless there is a massive amount of antibody present.

    Altogether, this is an important paper because it shows clearly that ongoing axonal or dendritic damage by Aβ seems to be, in part, reversible.


    . Amyloid-beta antibody treatment leads to rapid normalization of plaque-induced neuritic alterations. J Neurosci. 2003 Nov 26;23(34):10879-83. PubMed.

    View all comments by Frank Bernier
  4. This paper by Brendza et al. elegantly confirms and extends previous studies by this group and others. It uses living PDAPP:Thy-1:YFP transgenic mice and multiphoton microscopy to show that passive immunization with anti-Aβ antibodies not only is able to reduce or eliminate cortical deposits of Aβ, but also the associated dystrophic neurites. These are novel and important studies, as they imply that the therapeutic effects of passive immunization may extend beyond fibrillar or aggregated Aβ deposits themselves to pathological dystrophic processes that are a prominent feature of Alzheimer disease (AD) brain pathology.

    This study is significant, as dystrophic processes are well-recognized but poorly understood components of AD brain pathology despite having been described nearly 20 years ago as sites of tau accumulation (e.g., Ihara, 1988). In addition, many studies have shown that dystrophic neurites also contain other elements including fragments of APP flanking the Aβ domain and neurofilament proteins (e.g., Arai et al., 1990). Although dystrophic neurites are the locus of more than 95 percent of immunohistochemically ascertainable, pathological tau amyloid as measured morphometrically (Mitchell et al., 2000), they are lesions with a complex composition and uncertain etiology. It is likely that they contribute to cognitive impairments in AD, although how they do this is incompletely understood.

    Thus, Brendza et al. point the way to render these enigmatic lesions more tractable to experimental investigation. It will be interesting to see if they or other investigators are able to take further steps toward elucidating the nature and pathological significance of dystrophic neurites with the elegant methods used in their current study. A better understanding of these issues will help clarify the extent to which dystrophic neurites should become a deliberate focus of therapeutic intervention in AD.

    Several of the investigators of the present paper also have shown recently how passive immunotherapy with anti-Aβ antibodies exacerbates cerebral amyloid angiopathy-associated microhemorrhage in amyloid precursor protein transgenic mice (Racke, 2005). It would be important to know if the methods used by Brendza et al. could be exploited to gain insight into mechanisms underlying this complication of Aβ immunotherapy.


    . Massive somatodendritic sprouting of cortical neurons in Alzheimer's disease. Brain Res. 1988 Aug 30;459(1):138-44. PubMed.

    . Defined neurofilament, tau, and beta-amyloid precursor protein epitopes distinguish Alzheimer from non-Alzheimer senile plaques. Proc Natl Acad Sci U S A. 1990 Mar;87(6):2249-53. PubMed.

    . Novel method to quantify neuropil threads in brains from elders with or without cognitive impairment. J Histochem Cytochem. 2000 Dec;48(12):1627-38. PubMed.

    . Exacerbation of cerebral amyloid angiopathy-associated microhemorrhage in amyloid precursor protein transgenic mice by immunotherapy is dependent on antibody recognition of deposited forms of amyloid beta. J Neurosci. 2005 Jan 19;25(3):629-36. PubMed.

    View all comments by John Trojanowski
  5. Thank you for offering such a variety of papers by people who are spending their lives looking for answers.

    PS: Footnotes for lay persons would help.

  6. Brendza and colleagues demonstrate that administration of anti-Aβ antibodies into APP-transgenic mouse brain results in recovery of dystrophic neuritis surrounding Aβ plaques. The authors employed multiphoton microscopy to detect dystrophic neuritis labeled by transgene-derived YFP. This detection method allows in-vivo observation of Aβ plaques and dystrophic neuritis in living mice, although it is not fully non-invasive, as it requires cranial surgery to make a small window on a cranial bone.

    The effect of the antibody administration is not very large, but is statistically significant. The authors’ observation indicates that Aβ removal leads to recovery of neuritic dystrophy and that the processes involved in dystrophic neuritis are reversible until they reach a certain point. The results shown by Brendza and colleagues thus provide additional support for therapeutic strategies targeting Aβ.

    There remains, however, a primary question whether the formation of dystrophic neurites is a main pathway causing cognitive dysfunction in the pathological cascade of Alzheimer disease development, since it is generally accepted that tauopathy rather than Aβ amyloidosis is more closely associated with the symptoms.

    One technical point to note is that the authors administered the anti-Aβ antibodies directly into the cranial windows. This protocol probably does not mimic the processes evoked by Aβ vaccination, active or passive, because intravenously administered anti-Aβ antibodies do not seem to penetrate into brain parenchyma.

    In any case, removal of Aβ deposition in the early stages of the disease development before tauopathy and neurodegeneration proceeds to an irreversible extent will certainly stop this tragic disease from rising. Thus, the key words for combating Alzheimer disease in a pragmatic manner are “presymptomatic diagnosis” and “preventive medication.” For this purpose, we need to identify the primary cause of Aβ deposition in sporadic Alzheimer disease development.

    View all comments by Takaomi Saido

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Imaging of amyloid-beta deposits in brains of living mice permits direct observation of clearance of plaques with immunotherapy. Nat Med. 2001 Mar;7(3):369-72. PubMed.
  2. . Non-Fc-mediated mechanisms are involved in clearance of amyloid-beta in vivo by immunotherapy. J Neurosci. 2002 Sep 15;22(18):7873-8. PubMed.
  3. . Amyloid-beta antibody treatment leads to rapid normalization of plaque-induced neuritic alterations. J Neurosci. 2003 Nov 26;23(34):10879-83. PubMed.
  4. . PDAPP; YFP double transgenic mice: a tool to study amyloid-beta associated changes in axonal, dendritic, and synaptic structures. J Comp Neurol. 2003 Feb 17;456(4):375-83. PubMed.

Further Reading

Primary Papers

  1. . Anti-Abeta antibody treatment promotes the rapid recovery of amyloid-associated neuritic dystrophy in PDAPP transgenic mice. J Clin Invest. 2005 Feb;115(2):428-33. PubMed.