Aβ, a sticky peptide associated with neurotoxicity and inflammation in Alzheimer’s disease, normally plays the villain. Now, in the August 1 Science Translational Medicine, researchers led by Lawrence Steinman at Stanford University, Palo Alto, California, cast the peptide as a protagonist, reporting that Aβ in the blood ameliorates multiple sclerosis-like symptoms. The researchers injected the peptide into mouse models of MS and found that the treatment delayed disease onset and improved motor symptoms in the rodents. They traced the effect to a modulation of the peripheral immune system that quieted inflammation. The data demonstrate that “what may be true in one context may be different in another,” Steinman said. “In AD, the context may be highly pathological, whereas in other conditions, that same molecule may provide benefit.”

Scientists contacted by Alzforum described the paper as fascinating and thought provoking, while noting that more experiments will be needed to figure out what these data mean in terms of Aβ’s role in inflammation and disease. “I was very excited by this paper,” said Cindy Lemere at Brigham and Women’s Hospital, Boston, praising the thoroughness of the study and the use of multiple mouse models to show an effect. Terrence Town at the University of California, Los Angeles, agreed: “This is a well-executed, elegant piece of immunology.” The Multiple Sclerosis Discovery Forum reported on the significance of this work for MS (see related report).

In multiple sclerosis, immune cells from the blood infiltrate the brain and attack white and gray matter, leading to lesions and loss of myelin (see related news on the MS Discovery Forum). The disease typically strikes young adults and causes weakness, fatigue, and difficulty moving. To model the disease in animals, researchers inject myelin or lipid proteins into mice to induce experimental autoimmune encephalomyelitis (EAE). This demyelinating disease does not recapitulate all the features of MS, but has been used to develop several approved therapies for the disease (see related news on the MS Discovery Forum).

Steinman and colleagues, who previously developed the MS drug natalizumab (marketed as Tysabri®), were intrigued by the fact that people with MS make autoantibodies to Aβ, which hangs around damaged brain regions. They wondered if Aβ was pumping up inflammation. To test this, first author Jacqueline Grant injected 300 μg of synthetic Aβ40 or Aβ42, in the form of monomers and oligomers, or solvent control into the peritoneal cavity of mice with EAE. To the authors’ surprise, the treatment improved disease symptoms. In one paradigm, Grant and colleagues injected the peptide three times a week, beginning before clinical disease onset. Symptoms emerged later and were milder in animals that received the peptide, showing that the treatment could delay the disease. In other experiments, the researchers injected Aβ into symptomatic, partially paralyzed mice. Aβ42 had the strongest effect, reversing paralysis within two days, but Aβ40 also lessened disease severity compared to control animals.

The authors saw similar results in three different models of progressive MS, including EAE induced with myelin oligodendrocyte glycoprotein (MOG). They also used a fourth model, which was induced with proteolipid protein and is considered to represent relapsing-remitting MS. This is the most common form of the disease, in which periods of worsening symptoms are followed by remissions in which a person recovers some function. The relapsing-remitting animals had less inflammation after treatment, but only modest clinical improvement compared to controls, which the authors suggest might reflect the remitting nature of this model; some control animals may have spontaneously improved, making it more difficult to see a clinical effect.

These experiments showed that more Aβ in the blood could be good for EAE. Would less Aβ be bad? To investigate this, the authors looked at amyloid precursor protein (APP) knockout mice, which make no Aβ. These animals developed more severe symptoms after EAE induction than control mice did, strengthening the idea that Aβ protects against the disease.

What might explain the clinical improvement? The authors measured levels of 26 cytokines and chemokines in the blood of treated mice, finding an overall dampening of proinflammatory signaling molecules. When they analyzed subsets of immune cells by flow cytometry, they saw changes in the proportion of various cell types. In particular, treated mice had more helper T cells that secrete interferon-γ (IFN-γ), which the authors note acts to suppress the immune system in EAE. Importantly, the researchers saw normal cell viability, indicating Aβ was not simply killing off subsets of immune cells. Steinman noted that these results contrasted with what they saw in in-vitro experiments, where Aβ42 was cytotoxic for immune cells. “There was a complete disconnect between what this material could do in a dish and what it was doing in an animal,” Steinman said, adding that this highlights how differently molecules can behave in distinct physiological contexts.

Another important question was whether Aβ was acting on the peripheral immune system or the brain. The authors took T cells from mice that had received either Aβ or solvent injections, stimulated the cells with MOG antigen, and injected them into wild-type mice. This paradigm would normally induce EAE in the recipients. Animals that received T cells from Aβ-treated mice developed milder symptoms than those getting T cells from controls. This shows that something about the peripheral immune system had changed in Aβ-treated mice, the authors suggest. In addition, they did not see any Aβ deposits in the cortex or hippocampus of treated mice, further implying that the peptide was not acting on the brain.

Is this dampening of immune function specific to Aβ, or could other proteins produce the same effect in EAE mice? Steinman told Alzforum that in ongoing work, he has found that scrambled Aβ peptides do not provide benefit. On the other hand, other amyloidogenic proteins can suppress inflammation. This implies that something about the shape of these proteins, rather than the peptide sequence, modulates the immune system. Steinman added that he has no plans to pursue Aβ as a therapy for MS due to concerns about the toxicity of the molecule in the brain. However, he previously showed that an amyloidogenic protein found in the lens of the eye, α-B crystallin, also reduced inflammation and reversed paralysis in EAE mice (see Ousman et al., 2007, and Medical Research News story). He has formed a company, Cardinal Therapeutics, Menlo Park, California, and is currently seeking funding with the hopes of taking α-B crystallin forward into clinical trials for MS.

Exactly how Aβ modulates the immune system in EAE is still unknown. Some commentators suggested that injection of the human peptide may create local inflammation, drawing immune cells to the peritoneum and away from the brain. Other commentators noted, however, that the authors saw an overall dampening of inflammation in the peritoneum, not a rampant infection, making this unlikely. Town favors the idea that the peptide instead “diverts” the attention of T cells, making the lymphocytes respond to Aβ and thereby abandon the auto-aggressive phenotype they normally assume in EAE. “If that’s what is going on here, it represents an important immunological finding,” he said. The strategy could make a promising MS therapy if a safe enough peptide or protein could be found to inject, Town said. In that regard, Pritam Das at the Mayo Clinic in Jacksonville, Florida, cautioned that amyloid deposits in peripheral organs can cause severe disease. This would be one of the main things to watch for in a trial of amyloidogenic material, he said.

Some other data support the idea that Aβ in the blood could be protective. A recent study found that people with higher plasma Aβ42 had less cognitive decline over nine years than those with lower levels (see ARF related news story on Yaffe et al., 2011). Intriguingly, the immune system is altered in people with Down’s syndrome (DS), who carry an extra copy of the APP gene, giving them lifelong higher levels of Aβ. Lemere told Alzforum that clinical observations of people with DS by Florence Lai at McLean Hospital, Belmont, Massachusetts, show that this population is highly susceptible to autoimmune disorders such as hypothyroidism. About a third of people with DS have that condition, increasing to half in those who also develop AD. This may be because they form amyloid deposits in the brain starting early in life, which could pull Aβ out of the periphery, Lemere suggested. This might take the brakes off the immune system. In support of this idea, some research has shown that Aβ can travel from blood to brain (see ARF related news story). It would be interesting to know if people with APP mutations are also at higher risk of autoimmune disorders with age, Lemere said.

Several antibodies that target Aβ are currently in trials for AD (see, e.g., ARF related news story and ARF news story). If these treatments lower the amount of Aβ in the blood, could that increase inflammation? Das suggested that clinicians might want to closely monitor peripheral immune responses in these populations. However, Lemere pointed out that plasma Aβ has actually been shown to be higher during passive immunotherapy, as antibodies pull the peptide out of the brain or hold it in the blood. In theory, this might increase the benefit from these treatments. “We just have to wait and see what the data show,” Steinman said.—Madolyn Bowman Rogers


  1. These results are highly unexpected and in need of replication, but they are also very interesting in the light of what we currently know about Aβ homeostasis in several diseases of severe neuroinflammation, for example, bacterial meningitis, HIV-associated dementia, and multiple sclerosis. In these conditions, CSF levels of Aβ38, 40, and 42; soluble APP fragments; and BACE1 are lower than in neurologically healthy controls and or in AD patients. Perhaps a general reduction in the expression of APP-derived proteins plays a role in neuroinflammation?

  2. Steinman suggests that their results could imply a protective or pathological role for Aβ, depending on the context (i.e., good in MS, bad in AD).

    However, together with evidence showing an apparent protective role for Aβ in the brain against a variety of stress stimuli, including microglia activation, ROS, and cholesterol dysregulation (Castellani et al., 2009; Castello and Soriano, 2012), their results would be consistent with an alternative scenario, in which Aβ has a bona fide protective function, both in blood and in brain.

    That protective role would be achieved within a tightly regulated concentration range of Aβ, and dysregulation leading to peptide concentrations outside of that optimal range would lead, or contribute, to pathology. That would be true whether there is too little or too much Aβ.

    The obvious implication is that the current efforts to reduce Aβ levels in healthy brains as an approach to preventing AD are likely to be neutral at best, and would accelerate neuronal dysfunction at worst.

    It would seem that we need a better understanding of the basic biology of Aβ (and of its precursor APP) if we want to successfully determine whether amyloid-centric approaches to AD are scientifically sound or ought to be discontinued.


    . Reexamining Alzheimer's disease: evidence for a protective role for amyloid-beta protein precursor and amyloid-beta. J Alzheimers Dis. 2009;18(2):447-52. PubMed.

    . Rational heterodoxy: Cholesterol reformation of the amyloid doctrine. Ageing Res Rev. 2012 Jul 6; PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. In the Blood: What Can Plasma Aβ Tell Us About Alzheimer’s?
  2. Peripheral Aβ Seeds CAA and Parenchymal Amyloidosis
  3. A Close Look at Passive Immunotherapy Newbie, Crenezumab
  4. No Pony in There: Bapi Fails Mild to Moderate ApoE4 Carriers

Paper Citations

  1. . Protective and therapeutic role for alphaB-crystallin in autoimmune demyelination. Nature. 2007 Jul 26;448(7152):474-9. PubMed.
  2. . Association of plasma beta-amyloid level and cognitive reserve with subsequent cognitive decline. JAMA. 2011 Jan 19;305(3):261-6. PubMed.

External Citations

  1. related report
  2. related news on the MS Discovery Forum
  3. related news on the MS Discovery Forum
  4. Medical Research News story

Further Reading

Primary Papers

  1. . β-Amyloid: Enemy or Remedy?. Sci Transl Med. 2012 Aug 1;4(145):145fs24. PubMed.
  2. . Reversal of paralysis and reduced inflammation from peripheral administration of β-amyloid in TH1 and TH17 versions of experimental autoimmune encephalomyelitis. Sci Transl Med. 2012 Aug 1;4(145):145ra105. PubMed.