Do microglia promote the development of Alzheimer disease (AD) or hinder it? These immune cells are known to massively infiltrate amyloid plaques, lured by the presence of Aβ. However, in the process, microglia become activated, and there is continuing debate about whether activated microglia serve a neuroprotective function in the diseased brain. Yesterday in Neuron, the laboratories of Jean-Pierre Julien and Serge Rivest at Laval University in Quebec, Canada, entered this debate with a study that attempts to distinguish between resident and migratory, blood-derived microglia. First author Alain Simard and his colleagues report that bone marrow-derived microglia can reduce the number and size of amyloid deposits in an APPSwe/PS1 transgenic mouse model of AD. They show that this set of microglia actually infiltrate the core of amyloid plaques and phagocytose β-amyloid, and suggest these cells may be key in curbing the progression of AD.

The researchers transplanted green fluorescent protein (GFP)-expressing bone marrow cells into the bloodstream of irradiated APPSwe/PS1 transgenic mice. Then they immunohistochemically visualized an age-dependent increase in the number of plaque-associated GFP-positive microglia that peaked at 6 months of age, then slightly decreased around 9 months. Meanwhile, immunohistochemical staining of brain sections from treated animals revealed a consistent age-associated increase in the number and size of plaques. Based on these data, the authors argue that blood-derived microglia infiltrate the brain and migrate toward amyloid plaques after they have already formed and reached a particular size.

But how do these microglial cells get recruited and activated? To address this question, Simard and colleagues transplanted GFP-expressing bone marrow cells into irradiated wild-type mice and injected different isoforms of Aβ into their hippocampi. Upon examination of brain sections, they found that both Aβ40 and Aβ42 were able to stimulate the infiltration of GFP+ cells, suggesting that both proteins are behaving as chemoattractants for the bone marrow-derived microglia. Neither the control peptides Aβ31 and Aβ57, nor saline injections elicited this effect. Further investigation of the activation of microglia in response to Aβ42 in both wild-type and transgenic mice demonstrated the induction of an innate immune response without the inflammatory molecule TNF-α. The authors take this to mean that both exogenous and endogenous Aβ42 are capable of triggering similar, yet highly specific and atypical immune responses.

The dramatic infiltration of these blood-derived microglia around the age of 6 months and slight decrease at 9 months, when the number and size of plaques are still increasing, begs the question of whether these cells hinder or exacerbate senile plaque formation. To take a closer look, the researchers crossed the APPSwe/PS1 transgenic mice with ones that express mutant thymidine kinase (TK) and treated these animals with the antiviral drug ganciclovir. This drug is known to suppress bone marrow production of white blood cells, and thus, the treatment inhibited the recruitment of newly differentiated bone marrow-derived microglia, but not activation of resident microglia. Tissue analysis showed that ganciclovir treatment resulted in an increase in the number and size of plaques, as compared to matched saline-injected animals, indicating that the bone marrow-derived microglia restricted plaque size and number. Furthermore, ganciclovir treatment appeared to attenuate the immune response associated with the presence of amyloid plaques, a result the authors attribute to the prevention of infiltration of blood-derived microglia.

Finally, in analyzing stained tissues used in these studies, Simard and colleagues observed Aβ42 staining inside subcellular compartments within the GFP-positive, bone marrow-derived microglia, which co-localized with the lysosomal marker LAMP-2. From these data, the researchers assume that the blood-derived microglia were attempting to clear the amyloid deposits via phagocytosis in vivo. To confirm this observation, they treated cultured BV2 microglial cells with fluorescent red (Cy3)- conjugated Aβ42, and saw that the Aβ42 localized within the microglia, in line with their in vivo data.

In the end, the authors emphasize that previous studies have not distinguished between blood-derived and resident microglia, but have tended to fault microglia generally for contributing to plaque formation. Here the authors instead propose that resident microglia are present early on and perhaps play a role in plaque formation, while blood-derived ones appear later in an attempt to restrict the number and growth of plaques and/or clear them via phagocytosis. “The fact that newly recruited microglia are more efficient immune cells compared to their resident counterparts is clearly a beneficial mechanism in restricting disease progression,” they write. It is worth noting that some researchers interested in this topic raise a technical caveat about studies using irradiation, because that procedure is thought to temporarily weaken the blood-brain barrier. This would allow blood-borne cells to enter the brain in numbers that may not reflect the situation in human AD brain.—Erene Mina


  1. The origin, function, and sometimes even the existence of macrophages in the brain have been vigorously debated over the past century (see, for example, the historical overview by Peters, Palay, and Webster in The Fine Structure of the Nervous System, Oxford, 1991). Many issues have been resolved in recent years, but the cells remain surprisingly refractory to scientific interrogation. In Alzheimer disease, reactive microglia are a prominent cellular component of senile plaques, and hence, they have attracted the attention of researchers who wish to establish whether they are harmful or beneficial. The microglia themselves furnish evidence to support both views: As macrophages, they are equipped to rid the brain of unwanted material, yet this capability also gives them the potential to do collateral damage in the process.

    This intriguing paper by Simard, Rivest, and colleagues provides evidence for a beneficial role of microglia in removing excess β-amyloid in vivo. Their data indicate, in plaque-producing transgenic mice (including a model that also expresses thymidine kinase), that many macrophages associated with relatively mature senile plaques originate from the periphery, and that it is these cells, rather than the resident microglia, that phagocytose Aβ and thereby lower plaque load. The bone marrow-derived macrophages are summoned to dispose only of senile plaques that have evolved to the point where, presumably, they present a distinct threat to the integrity of the brain. This finding supports the view that activated microglia probably do not mediate the early deposition of Aβ in APP-transgenic mice (Stalder et al., 1999) or in aged humans (Vogelgesang et al., 2002). It also suggests a tentative resolution to the question of what microglia are doing in plaques: It depends in part on the origin of the cells. On balance, these interesting data in mice imply that promoting the phagocytosis of brain amyloid by bone marrow-derived microglia could reduce plaque load in AD.

    That said, provoking a cellular attack on Aβ is a potentially high-reward but also risky strategy, as has been made clear by the Aβ immunization trials for AD. With this in mind, there are some additional issues that might be considered before calling on peripheral macrophages to join the fray. First, the relevance of Aβ clearance in transgenic mice to the situation in human AD remains uncertain, as there is considerable evidence for critical differences between murine and human senile plaques/Aβ deposits, and possibly in the way that the two species respond immunologically to Aβ. Second, it is important to repeat this study in larger numbers of mice, and preferably in male and female mice, which can differ significantly in their tendency to deposit Aβ. Studies in biologically intermediate species might help to bridge this gap. In addition, it would be beneficial to know if, for example, oligomeric Aβ, tau, and neuronal integrity are affected by blood-borne microglia in animal models. Ultimately, behavioral improvement will be the arbiter of the value of any potential therapy.

    It would be ideal to impede the Aβ-cascade before the toxic effects of Aβ become manifest in the brain. If, however, recruitment of peripheral macrophages can be demonstrated to be selective, safe, and effective in humans, this approach could indicate a path to treating AD at a point in the pathogenesis of the disorder when other types of therapy might be too late.


    . Association of microglia with amyloid plaques in brains of APP23 transgenic mice. Am J Pathol. 1999 Jun;154(6):1673-84. PubMed.

    . Activated microglia do not mediate the early deposition of Abeta in carriers of the apolipoprotein Eepsilon4 allele. Clin Neuropathol. 2002 May-Jun;21(3):99-106. PubMed.

    View all comments by Lary Walker
  2. This is an interesting manuscript. It confirms an earlier study by Wisniewski et al., 1991, which reported that invading macrophages after brain injury phagocytose amyloid while resident microglia appear not to do so. However, the question of whether a CNS lesion, that is, disruption of the blood-brain barrier (BBB), is necessary for a “phagocytotic” activity of invading macrophages still remains unanswered in this new study. That is because Simard and collaborators inserted a catheter into the ventricle, which obviously affected the integrity of the BBB.

    The authors also suggest amyloid phagocytosis of the invading macrophages based on co-staining of a lysosomal marker with Aβ. Unfortunately this co-staining was not done for resident microglia. Colocalization of Aβ/amyloid at the level of confocal microscopy does not unequivocally prove amyloid phagocystosis (see, e.g., Wisniewski et al. above; Stalder et al., 2001). Because the role of resident microglia was not studied, further work is needed to elucidate the function and impact of resident microglia versus invading macrophages.


    . Phagocytosis of beta/A4 amyloid fibrils of the neuritic neocortical plaques. Acta Neuropathol. 1991;81(5):588-90. PubMed.

    . 3D-Reconstruction of microglia and amyloid in APP23 transgenic mice: no evidence of intracellular amyloid. Neurobiol Aging. 2001 May-Jun;22(3):427-34. PubMed.

    View all comments by Mathias Jucker
  3. I would like to thank Erene Mina and Drs. Walker and Jucker. They provide insightful comments regarding specific aspects of the study. I'd like to address a few points here.

    The first one regards irradiation and its effects on the blood-brain barrier (BBB). There is not very strong evidence that irradiation alters the BBB, and brain infiltration of bone marrow-derived cells has been reported with other techniques as well. Messengale and colleagues have validated this concept using both lethal irradiation and parabiosis techniques in mice (Massengale et al., 2005). Although most (if not all) GFP cells found in the brains of chimeric mice have a microglial phenotype, the overall contributions of such cells to the brain-resident microglial populations of normal mice remain quite low (e.g., 0.5-11.5 percent of resident microglia). This is what we generally observe in our mice (Simard and Rivest, 2004). In APP mice, however, there is a robust microglial recruitment toward the plaques, and those that derive from the bone marrow are attracted at a specific time of the disease. Other groups have observed a similar pattern in irradiated APP mice (Malm et al., 2005; Stalder et al., 2005), and one can appreciate the robust microglia infiltration in the plaques of non-irradiated mice (Fig. 1 and supp. movie 1). Therefore, I do not think that infiltration is caused by alteration of the BBB in irradiated mice, but is a natural process that is especially dynamic while the plaques progress. The mechanisms explaining why bone marrow-derived microglia are no longer recruited toward the plaques at a specific time point in the disease have yet to be unraveled with future experiments.

    Another point raised is that we did not look at the colocalization of Aβ in the lysosomes of the resident microglia. We actually did a meticulous analysis of such processes in the chimeric APP, and while GFP cells were almost always associated with lysosomal Aβ, the resident cells were not. This is the reason that we did not show these results, but we have discussed them.

    We observed phagocytosis by bone marrow-derived microglia during a very specific time. This takes place around 6 months of age in the APP/PS1 mice, and we no longer see these cells at 9 months. Therefore, cell recruitment (of bone marrow origin) and phagocytosis are dynamic and transient phenomena, which may explain why other groups have not detected it. This also explains why inhibition of cell recruitment (APP/TK mice) from 5 to 6 months has such profound consequences on plaque growth. We are now working on new genetic strategies to enhance and improve the recruitment of these cells for a longer period of time to see if we can prevent the amyloid cascade and cognitive deficit.

    Finally, multiple staining and 3D reconstructions using confocal laser-scanning microscopy are powerful tools to determine cellular compartmentalization, such as Aβ within the lysosomal GFP cells.


    . Bone-marrow-derived cells contribute to the recruitment of microglial cells in response to beta-amyloid deposition in APP/PS1 double transgenic Alzheimer mice. Neurobiol Dis. 2005 Feb;18(1):134-42. PubMed.

    . Hematopoietic cells maintain hematopoietic fates upon entering the brain. J Exp Med. 2005 May 16;201(10):1579-89. PubMed.

    . Bone marrow stem cells have the ability to populate the entire central nervous system into fully differentiated parenchymal microglia. FASEB J. 2004 Jun;18(9):998-1000. PubMed.

    . Bone marrow-derived microglia play a critical role in restricting senile plaque formation in Alzheimer's disease. Neuron. 2006 Feb 16;49(4):489-502. PubMed.

    . Invasion of hematopoietic cells into the brain of amyloid precursor protein transgenic mice. J Neurosci. 2005 Nov 30;25(48):11125-32. PubMed.

Make a Comment

To make a comment you must login or register.


No Available References

Further Reading


  1. . Invasion of hematopoietic cells into the brain of amyloid precursor protein transgenic mice. J Neurosci. 2005 Nov 30;25(48):11125-32. PubMed.
  2. . Experimental autoimmune encephalomyelitis repressed by microglial paralysis. Nat Med. 2005 Feb;11(2):146-52. PubMed.
  3. . Mononuclear phagocytes in the pathogenesis of neurodegenerative diseases. Neurotox Res. 2005 Oct;8(1-2):25-50. PubMed.
  4. . Bone-marrow-derived cells contribute to the recruitment of microglial cells in response to beta-amyloid deposition in APP/PS1 double transgenic Alzheimer mice. Neurobiol Dis. 2005 Feb;18(1):134-42. PubMed.
  5. . Dynamic complexity of the microglial activation response in transgenic models of amyloid deposition: implications for Alzheimer therapeutics. J Neuropathol Exp Neurol. 2005 Sep;64(9):743-53. PubMed.
  6. . Microglial activation facilitates Abeta plaque removal following intracranial anti-Abeta antibody administration. Neurobiol Dis. 2004 Feb;15(1):11-20. PubMed.

Primary Papers

  1. . Bone marrow-derived microglia play a critical role in restricting senile plaque formation in Alzheimer's disease. Neuron. 2006 Feb 16;49(4):489-502. PubMed.