When it comes to the biological activity of Aβ, a question often pondered is: How much is enough? In the March 31 Journal of Neuroscience, Barry Festoff and colleagues suggest that the answer may be: Not that much. Festoff, from the Veterans Affairs Medical Center, Kansas, together with coworkers there and elsewhere, suggest a mechanism whereby Aβ may cause neurotoxicity in the very early stages of Alzheimer’s disease. They report that levels of soluble Aβ that have no direct effect on microglia can nevertheless potentiate the effect of other molecules, namely, G protein-coupled receptor (GPCR) agonists.

This type of subthreshold response to Aβ has been observed previously in vascular cells (for example, see Wang et al., 2000). To test if microglia behaved similarly, first author Zhiming Suo and colleagues exposed primary mouse brain microglial cultures to GPCR activators, such as thrombin, after treating the cells with levels of Aβ that do (about 2.5 micromolar) or do not (around 0.01 to 0.1 micromolar) evoke a response when given alone. Suo found that in the latter case, the subthreshold level as much as tripled thrombin-induced expression of TNF-α.

So how could Aβ affect thrombin signaling? Because thrombin activates GPCRs, which must then be deactivated, Suo wondered if Aβ may prevent inactivation of these receptors. To test this idea he examined the effect of Aβ on G protein receptor kinases (GRKs), which are known to desensitize some GPCRs. The authors found that binding of GRK2 or GRK5 to two GPCRs—PAR1 and PAR4—is significantly reduced in the presence of subthreshold levels of Aβ. This, the authors discovered, may be the result of a compartmentalization effect. In the presence of Aβ1-42, GRK5 was rapidly moved from the membrane, where it is needed to switch off the GPCRs, to the cytosol. Though Aβ1-40 had a weaker effect on GRK5, and Aβ40-1 had no effect on the location of either kinase, GRK2 was unaffected by any version of the peptide.

These results are noteworthy for two reasons. First, they demonstrate that thresholds depend on what one is measuring. Second, they propose a potential mechanism for Aβ toxicity. The authors found that in the cortex of transgenic CRND8 mice, which exhibit early onset AD symptoms, GRK2 (and GRK5) appeared to be redistributed from the membrane fraction to the soluble fraction as the animals aged. This shift became “statistically significant at 10 weeks of age, which was one week before cognitive decline or ‘disease onset’ in these animals,” write the authors. Oddly enough, some molecular modeling predictions suggest that presenilin may be a GPCR (see ARF related news story), perhaps linking Aβ back to its maker via a GRK.—Tom Fagan


  1. This is an interesting study. Suo et al. propose a concept of "threshold" of Aβ in AD brain. There have been extensive studies on Aβ loads and correlation of cognitive decline. It is now clear that when so-called "subthreshold" levels of Aβ accumulate in the brain, they definitely do something, as the authors reported, i.e., sensitize microglia. More importantly, the authors claim that these subthreshold levels of Aβ may occur as oligomers. Glabe et al. (see ARF related news story) and Walsh et al. (see ARF related news story) have reported that oligomers of Aβ may be critical in causing neurodegeneration. Thus, neuron death probably has already occurred in AD brains at a very early stage, when clinical symptoms cannot yet be observed.

    Finding that low doses of Aβ cannot stimulate TNFα release from mouse microglia is interesting. We previously reported that Aβ can directly stimulate release of TNFα and other cytokines from microglia derived from rapidly autopsied brains with AD (Lue et al., 2001). It suggests that properties of murine glial cells might be different from those of human, especially AD microglia, and it is important to pursue this issue further. The result of the GRK abnormality occurring prior to cognitive decline in the Tg-CRND8 animal model is interesting. It will be important to study this phenomenon in patients with AD, as it might provide an alternative therapeutic opportunity for AD treatment.


    . Inflammatory repertoire of Alzheimer's disease and nondemented elderly microglia in vitro. Glia. 2001 Jul;35(1):72-9. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Erene Mina Reports on Presenilin’s Loops through the Membrane

Paper Citations

  1. . Toxicity of Dutch (E22Q) and Flemish (A21G) mutant amyloid beta proteins to human cerebral microvessel and aortic smooth muscle cells. Stroke. 2000 Feb;31(2):534-8.

Other Citations

  1. CRND8

Further Reading

No Available Further Reading

Primary Papers

  1. . Abnormality of G-protein-coupled receptor kinases at prodromal and early stages of Alzheimer's disease: an association with early beta-amyloid accumulation. J Neurosci. 2004 Mar 31;24(13):3444-52. PubMed.