Amyloid-β (Aβ) peptides in cerebrospinal fluid are among the few established molecular signs of impending Alzheimer disease. By contrast, tracking Aβ dynamics directly in the tiny spaces between cells of affected brain areas has proven a far greater challenge. Reporting in the August 29th issue of Science, researchers at Washington University in St. Louis, Missouri, and at the University of Milan and affiliated hospitals in Italy, have done just that in vivo in human brain trauma patients. Using microdialysis to collect brain interstitial fluid (ISF) samples from 18 patients undergoing invasive monitoring for acute brain injury, the scientists found that Aβ levels in brain ISF correlated strongly with neurological status. They dipped as patients declined and rose as they improved. Though somewhat paradoxical, these findings appear to fit with in vitro studies suggesting that neuronal activity boosts Aβ release. The new work also raises intriguing questions about the physiological role of soluble Aβ in the brain, and provides tools to examine how such functions may fade in AD and related conditions.

Earlier in vitro and animal studies (Kamenetz et al., 2003; Cirrito et al., 2005) have linked higher synaptic activity with increased Aβ production, and a recent study has suggested endocytosis as a key mechanism connecting the two (see ARF related news story). Collecting and analyzing hippocampal brain ISF samples from mice in vivo required a fair bit of technical know-how (Cirrito et al., 2003). Tracking ISF Aβ levels in brains of living patients presents additional challenges—not the least of which is the scarcity of individuals willing to have a microdialysis catheter stuck into their skull for study purposes. Acknowledging such concerns, researchers led by David Brody and David Holtzman of Washington University, and Sandra Magnoni of Ospedale Maggiore Policlinico, a major trauma center in Milan, turned to intensive care unit (ICU) patients who were already getting invasive brain procedures done as part of their clinical care. These 18 ICU patients—12 in Milan and six in St. Louis, none diagnosed with AD or dementia—had holes drilled into their heads for pressure monitoring after acute brain injury, and agreed to have a microdialysis catheter placed into subcortical white matter during the same procedure.

Analyzing the microdialysis samples using enzyme-linked immunosorbent assay (ELISA), the researchers measured Aβ levels every 1 to 2 hours in all patients, starting 12 to 48 hours post-injury at the time the catheter was placed. This required some procedural innovation. The standard microdialysis perfusion fluid allows virtually no Aβ recovery, perhaps because Aβ is so sticky, Brody told ARF. In earlier mouse ISF studies, John Cirrito in the Holtzman lab found a way around this by including bovine serum albumin in the perfusion fluid. BSA is a non-specific blocking agent that apparently keeps Aβ from adhering to the collection vials. In the new study, the researchers similarly improved Aβ recovery by spiking the perfusion fluid with sterile human albumin. Brody noted that this modification could impact future studies looking at different Aβ forms, such as oligomers, which might have different binding properties.

Analysis of a small number of trauma patients in an ICU setting allows at best for limited experimentation and control of conditions, and would be expected to produce variable data. Nevertheless, the researchers uncovered a number of surprising trends. First and foremost, the scientists saw an overall rise in brain ISF Aβ levels over the course of the first few days in most patients: median Aβ concentrations at 60 to 72 hours were 59 percent higher than during the first 12 hours of analysis. This trend was no measurement artifact because concentrations of urea, which control for the stability of the microdialysis catheter function, held steady in the ISF samples over the same time frame.

From a scientific perspective, the initially low Aβ in brain ISF was unexpected. Early-life traumatic brain injury has been reported to increase later risk of AD, presumably by setting in motion its pathophysiological manifestations—namely, buildup of toxic Aβ (see ARF Live Discussion). “With that in mind,” Brody said, “we expected a big spike of Aβ immediately after injury that then subsided during recovery.”

They saw the opposite—initially low Aβ levels that steadily rose as the patients recovered. The Aβ measurements seemed to jibe with other wellness parameters routinely used in ICU monitoring. For instance, low brain ISF Aβ was found to correlate with low glucose and a high lactate/pyruvate ratio, indicators of abnormal brain metabolism. Low brain ISF Aβ also correlated with high intracranial pressure and extreme brain temperatures—problems likely to disrupt neuronal function. “Thus, brain ISF Aβ increased as overall physiology normalized,” the authors wrote.

The observed brain Aβ dynamics seemed to make biological sense in light of previous work in mice suggesting that neuronal activity drives Aβ secretion (see ARF related news story). Changes in brain ISF Aβ from baseline appeared to track, and in some cases precede, changes in global neurological status as assessed by Glasgow Coma Scores in 13 patients from whom these serial measurements could be readily obtained. In other words, as patients regained brain function, their Aβ levels increased. Brody noted, however, that electrical activity in the brains of the patients was not directly measured in this study; hence, a connection between improved neurological status and increased synaptic activity can only be inferred indirectly.

The apparent rise in ISF Aβ during the patients’ recovery could reflect not only increased production of Aβ but also a reduction in its elimination, suggested Roy Weller, University of Southampton School of Medicine, United Kingdom, via email (see full comment below). He and colleagues reported recently that levels of soluble Aβ in the brain depend not only on its level of production but also on the efficiency by which it is cleared from ISF (Weller et al., 2008).

An existing research focus on the role for soluble Aβ in AD pathogenesis intensified when researchers led by Dennis Selkoe of Brigham and Women’s Hospital, Boston, published evidence for neurotoxicity of Aβ oligomers isolated from human AD brains (Shankar et al., 2008). Selkoe commented via email that he finds the data by Brody and colleagues “compelling and biologically sensible.”

The results prompt questions about whether Aβ assumes different roles in short-term versus longer-term scenarios, said Doug Smith, director of the Center for Brain Injury and Repair at the University of Pennsylvania School of Medicine in Philadelphia. “In the chronic setting, you might suspect that having a lot of Aβ floating around is not a good thing,” Smith said in a phone interview. “But in the acute setting…you have to consider that there may be something protective there. I’d be curious to know—if these patients go on releasing Aβ in the brain—whether at some point it affects cognition.”

Smith and Brody acknowledge that the new data provide no clear evidence for a pathophysiologic process linking acute brain injury with later development of AD. Along similar lines, Smith and colleagues reported recently that Aβ plaques induced days after traumatic brain injury (TBI) did not seem to result in AD, as long-term TBI survivors had no virtually no Aβ plaques (Chen et al., 2008). Regarding the current study, Brody noted that his team would not have detected an early, transient rise in brain ISF Aβ, as the ISF microdialysis catheters were not placed into the patients’ brains until at least 12 hours post-injury.

Brad Hyman of Massachusetts General Hospital, Charlestown, offered this comment via email: “While the implications for a ‘normal’ function of Aβ are intriguing, it is still not completely clear whether the data reflect an active role for Aβ or simply establish that it is a marker for neuronal activity.”

Sorting out what happens in AD is another key issue for follow-up work, Hyman noted (see full comment below). In brain imaging studies, lower CSF Aβ42 has been correlated with higher levels of brain amyloid, suggesting that amyloid plaques trap Aβ peptides normally destined for export through the CSF (see ARF related news story). Furthermore, in a prospective study by Holtzman and colleagues, cognitively normal people with low concentrations of CSF Aβ42 and Aβ40 (along with high levels of tau, another CSF biomarker) had increased risk of conversion to dementia during a 3-4 year follow-up period (see ARF related news story).

Interestingly, the current study demonstrated that brain ISF and CSF are not as closely linked as many in the field might have hoped. The Aβ dynamics traced in brain ISF—the rise as neurological status improved, the fall as it worsened—was not reflected in ventricular CSF samples taken in the same patients. “If you only looked at CSF, you would have missed those key dynamics,” Brody said. He cautioned, however, that it remains to be seen whether the ISF-CSF mismatch observed in patients with severe brain injuries holds for AD patients, as well. Likewise, because this study looked at ventricular CSF, i.e. sampled in the brain, it is hard to say whether the ISF/CSF differences cast doubt on the validity of lumbar CSF Aβ and tau measurements used as predictive biomarkers in AD, Brody said. (Ventricular CSF was conveniently sampled in the new study because many patients already had ventricular catheters placed for clinical purposes, whereas lumbar punctures can be dangerous in these patients, he said.) “Quite honestly, this study is just the beginning,” acknowledged Brody. “It raises just as many questions as it answers.”—Esther Landhuis


  1. This is a fantastic study using an extraordinarily powerful technique to study human physiology and pathophysiology. While the implications for a "normal" function of Aβ are intriguing, it is still not completely clear whether the data reflect an active role for Aβ or simply establish that it is a marker for neuronal activity. Sorting this out will be fascinating.

    The remaining other big question - hopefully soon to be determined - is what happens in Alzheimer disease? Do low CSF Aβ levels reflect low synthesis? Does the diminished rate of plaque accumulation as the disease progresses reflect low synthesis rates? Hopefully this technology can shed light on these long-standing paradoxes.

    View all comments by Bradley Hyman
  2. Amyloid-β Shows Another Facet
    In this article, Brody et al. showed that concentrations of amyloid-β (Aβ) in brain interstitial fluid (ISF), in vivo, increased following head injury and subarachnoid hemorrhage as neurological status improved. Conversely, concentrations of Aβ fell when neurological status declined. The authors conclude that neuronal activity regulates the concentration of extracellular Aβ, and that declining levels of Aβ reflect depressed neuronal function.

    To some extent, Brody et al. underestimate the potential significance of their findings. As emphasized by the authors, observations derived from in vivo studies in human patients are extremely valuable as they relate directly to the human condition and allow the generation of hypotheses that can be tested experimentally. In addition, their studies have produced data regarding the physiological functions of soluble Aβ that are relevant to the role of Aβ in Alzheimer disease (AD).

    A number of recent studies suggest that soluble Aβ in the brain may have a more significant role in the pathogenesis of AD than the plaques of insoluble Aβ. PET imaging has shown that, as a group, patients with mild cognitive impairment have a similar amount of insoluble Aβ in their brains as patients with AD [3]. Furthermore, removal of Aβ plaques from the brain does not prevent progressive neurodegeneration in AD [2]. Other studies indicate that high levels of soluble Aβ in the brain correlate more closely with cognitive decline in AD than does insoluble Aβ plaque load [5,6] and that removal of Aβ plaques from the brain by Aβ immunotherapy increases the level of soluble Aβ in the brain [8].

    The perceived role of soluble Aβ in the pathogenesis of cognitive impairment in AD makes it all the more urgent to take a fresh look at the physiological and pathological roles of Aβ. Little is known about the physiological role of soluble Aβ in the brain in vivo. The result’s of Brody et al. suggest that neurons produce Aβ in response to acute brain injury and that the higher levels of Aβ in the subcortical white matter correlate with improvement of neurological status. But, why should the levels of soluble Aβ rise in those patients showing improvement in neurological status?

    Levels of soluble Aβ in the brain depend not only on its level of production but also on the efficiency of elimination of Aβ from the interstitial fluid [11]. Several pathways for elimination of Aβ are recognized; they include degradation by neprilysin and absorption into the blood by various mechanisms [11]. Aβ also drains with brain ISF along basement membranes in the walls of cerebral capillaries and arteries where it may deposit as cerebral amyloid angiopathy (CAA) [11]. The rise in concentration of Aβ in the ISF during the recovery period following acute brain injury in the study of Brody et al. could reflect a reduction of elimination of Aβ as well as its increased production.

    Failure of elimination of Aβ along perivascular drainage routes and the development of CAA is observed with increasing age when arteries stiffen [11] and following denervation [1] when vessel tone may be affected. Diffuse vascular injury is well recognized in patients with head injuries [9], and vascular spasm is seen in both head injury and following subarachnoid hemorrhage [7]. Vascular factors may play a role in impeding the elimination of Aβ along artery walls and in the development of CAA following head injury [4].

    Brody et al. emphasize that their results correlate with in vitro studies showing that neuronal and synaptic activity dynamically regulates the concentration of extracellular soluble Aβ. However, it is not known whether the effects of the increased production of Aβ are purely local or whether Aβ has an effect on artery walls as it drains out of the brain along perivascular pathways.

    The role of Aβ in the pathogenesis of AD requires some re-evaluation. Aβ in its soluble form in the brain seems to be as important as insoluble plaques of Aβ, if not more important, in the pathogenesis of AD. Toxic forms of soluble Aβ appear to have a role in the pathogenesis of AD [10], and the techniques used by Brody et al. may provide the opportunity to study this further in vivo. However, we may ask whether toxicity is the sole role for Aβ in AD. The deposition of fibrillar Aβ in the perivascular drainage pathways in CAA may block the elimination not only of Aβ but also of other brain metabolites. This may lead to loss of homeostasis and neuronal malfunction as a factor in cognitive impairment in AD. Evaluation of the composition of brain ISF may help to answer questions about the quality of the neuronal environment in patients with AD.


    . Cholinergic deafferentation of the rabbit cortex: a new animal model of Abeta deposition. Neurosci Lett. 2000 Mar 31;283(1):9-12. PubMed.

    . PET amyloid ligand [11C]PIB uptake is increased in mild cognitive impairment. Neurology. 2007 May 8;68(19):1603-6. PubMed.

    . Cerebral amyloid angiopathy in traumatic brain injury: association with apolipoprotein E genotype. J Neurol Neurosurg Psychiatry. 2005 Feb;76(2):229-33. PubMed.

    . Soluble amyloid beta peptide concentration as a predictor of synaptic change in Alzheimer's disease. Am J Pathol. 1999 Sep;155(3):853-62. PubMed.

    . Soluble pool of Abeta amyloid as a determinant of severity of neurodegeneration in Alzheimer's disease. Ann Neurol. 1999 Dec;46(6):860-6. PubMed.

    . Posttraumatic vasospasm: the epidemiology, severity, and time course of an underestimated phenomenon: a prospective study performed in 299 patients. J Neurosurg. 2005 Nov;103(5):812-24. PubMed.

    . Amyloid-beta peptide remnants in AN-1792-immunized Alzheimer's disease patients: a biochemical analysis. Am J Pathol. 2006 Sep;169(3):1048-63. PubMed.

    . Diffuse vascular injury in fatal road traffic accident victims: its relationship to diffuse axonal injury. J Forensic Sci. 2003 May;48(3):626-30. PubMed.

    . A beta oligomers - a decade of discovery. J Neurochem. 2007 Jun;101(5):1172-84. PubMed.

    . Perivascular drainage of amyloid-beta peptides from the brain and its failure in cerebral amyloid angiopathy and Alzheimer's disease. Brain Pathol. 2008 Apr;18(2):253-66. PubMed.

  3. This fits very nicely with accumulating evidence on the involvement of Aβ peptides in maintenance and repair. The “toxic” hydrogen peroxide Aβ produces (Behl et al., 1994) is actually a signaling molecule used by growth factors to activate many of their downstream effectors (Rhee et al., 2003). Growth factors, of course, are involved in repair as well as growth.

    Aβ produces another molecule besides hydrogen peroxide that is generally considered toxic but might actually be involved in maintenance and repair. Aβ1-42 was recently found to activate sphingomyelinases, whose product ceramide kills cells (Malaplate-Armand et al., 2006) but can also activate autophagy (Scarlatti et al., 2004), an important component of maintenance and repair systems which malfunctions in Alzheimer’s (Nixon et al., 2005).

    In fact, “toxic” ceramide has an even more surprising function. It mediates the effects of NGF on outgrowth of cultured hippocampal neurons, according to Brann et al. (1999). It seems that low levels of ceramide promote growth, intermediate levels, autophagy, and high levels, apoptosis. Since Aβ produces ceramide (see above), this means that the concentration of Aβ might be critical, too. Low levels of Aβ peptides promote neurite outgrowth, and high levels inhibit it (Postuma et al., 2000). We know that high levels kill cells, so perhaps intermediate levels promote autophagy?


    . Hydrogen peroxide mediates amyloid beta protein toxicity. Cell. 1994 Jun 17;77(6):817-27. PubMed.

    . Ceramide signaling downstream of the p75 neurotrophin receptor mediates the effects of nerve growth factor on outgrowth of cultured hippocampal neurons. J Neurosci. 1999 Oct 1;19(19):8199-206. PubMed.

    . Soluble oligomers of amyloid-beta peptide induce neuronal apoptosis by activating a cPLA2-dependent sphingomyelinase-ceramide pathway. Neurobiol Dis. 2006 Jul;23(1):178-89. PubMed.

    . Extensive involvement of autophagy in Alzheimer disease: an immuno-electron microscopy study. J Neuropathol Exp Neurol. 2005 Feb;64(2):113-22. PubMed.

    . Substrate-bound beta-amyloid peptides inhibit cell adhesion and neurite outgrowth in primary neuronal cultures. J Neurochem. 2000 Mar;74(3):1122-30. PubMed.

    . Cellular regulation by hydrogen peroxide. J Am Soc Nephrol. 2003 Aug;14(8 Suppl 3):S211-5. PubMed.

    . Ceramide-mediated macroautophagy involves inhibition of protein kinase B and up-regulation of beclin 1. J Biol Chem. 2004 Apr 30;279(18):18384-91. PubMed.

    View all comments by Jane Karlsson

Make a Comment

To make a comment you must login or register.


News Citations

  1. Link Between Synaptic Activity, Aβ Processing Revealed
  2. Paper Alert: Synaptic Activity Increases Aβ Release
  3. Brain Imaging Speaks Volumes about AD and the Aβ Sink
  4. Biomarker Roundup: Collecting Clues from MRIs to RNAs

Webinar Citations

  1. Sports Concussions, Dementia, and APOE Genotyping: What Can Scientists Tell the Public? What’s Up for Research?

Paper Citations

  1. . APP processing and synaptic function. Neuron. 2003 Mar 27;37(6):925-37. PubMed.
  2. . Synaptic activity regulates interstitial fluid amyloid-beta levels in vivo. Neuron. 2005 Dec 22;48(6):913-22. PubMed.
  3. . In vivo assessment of brain interstitial fluid with microdialysis reveals plaque-associated changes in amyloid-beta metabolism and half-life. J Neurosci. 2003 Oct 1;23(26):8844-53. PubMed.
  4. . Perivascular drainage of amyloid-beta peptides from the brain and its failure in cerebral amyloid angiopathy and Alzheimer's disease. Brain Pathol. 2008 Apr;18(2):253-66. PubMed.
  5. . Amyloid-beta protein dimers isolated directly from Alzheimer's brains impair synaptic plasticity and memory. Nat Med. 2008 Aug;14(8):837-42. PubMed.
  6. . A lack of amyloid beta plaques despite persistent accumulation of amyloid beta in axons of long-term survivors of traumatic brain injury. Brain Pathol. 2009 Apr;19(2):214-23. PubMed.

Further Reading


  1. . Synaptic activity regulates interstitial fluid amyloid-beta levels in vivo. Neuron. 2005 Dec 22;48(6):913-22. PubMed.
  2. . Fluctuations of CSF amyloid-beta levels: implications for a diagnostic and therapeutic biomarker. Neurology. 2007 Feb 27;68(9):666-9. PubMed.
  3. . Acute stress increases interstitial fluid amyloid-beta via corticotropin-releasing factor and neuronal activity. Proc Natl Acad Sci U S A. 2007 Jun 19;104(25):10673-8. PubMed.
  4. . Endocytosis is required for synaptic activity-dependent release of amyloid-beta in vivo. Neuron. 2008 Apr 10;58(1):42-51. PubMed.

Primary Papers

  1. . Amyloid-beta dynamics correlate with neurological status in the injured human brain. Science. 2008 Aug 29;321(5893):1221-4. PubMed.