While amyloid-β (Aβ) seems to have the majority vote for the protein culprit in Alzheimer disease (AD), the election is not quite as clear regarding which Aβ assembly state produces AD’s cognitively crippling characteristics. An increasing number of reports show that globular, soluble Aβ oligomers rather than Aβ plaques containing insoluble Aβ fibrils are to blame for the cognitive deficits (see ARF related news story and ARF news story). Using transgenic mouse models to manipulate Aβ assembly states, Lennart Mucke at the Gladstone Institutes/University of California and coauthors demonstrate that speeding up Aβ fibrillization into plaques diminishes Aβ oligomers and protects against cognitive deficits in AD mouse models. The report appears in the August 17 Journal of Biological Chemistry.

Taking advantage of the E22G Arctic mutation that accelerates Aβ aggregation into protofibrils and fibrils, the researchers studied the balance between fibrillar and nonfibrillar (or oligomeric) Aβ assembly states. First author Irene Cheng and coauthors show—for what they believe is the first time—three-dimensional morphological and quantitative characteristics of Aβ oligomers isolated directly from brain tissue, using atomic force microscopy (see ARF related news story). By analyzing synthetic Aβ peptides, the researchers confirmed that the Arctic mutation accelerates fibril formation and demonstrated that this effect actually lowers the relative abundance of Aβ oligomers.

Then the researchers evaluated how the Arctic mutation’s promotion of fibrillization altered Aβ assembly and cognition in three different mouse models, each engineered to have elevated levels of Aβ42 but with varying speeds of turning Aβ monomers into mature plaques. At 3-4 months, J20 mice had no plaques, ARC48 had lots of plaques, and ARC6 mice had few plaques. Oligomer levels were highest in J20 mice, a little lower in ARC48 mice, and undetectable in ARC6 mice.

Behavioral analyses using cued and spatial versions of the Morris water maze task showed a striking dissociation between cognitive deficits and Aβ plaques. “We find that cognitive deficits correlate best with oligomeric Aβ,” said Mucke of a pattern of results that is consistent with what coauthor Karen Ashe described last year (see ARF related news story) in which Aβ oligomers infused directly into the brain temporarily impaired memory. Now Mucke and colleagues extend those findings by demonstrating oligomer dependence of memory deficits across different mouse lines and normal memory in the ARC6 line, which had plaques but no oligomers.

The cognitive impairment associated with increased Aβ oligomers, regardless of plaques, indicates that shifting around Aβ assembly states can alter different aspects of the disease. The ARC6 mice showed quick transition of Aβ from monomers to plaques with decreased time spent in the oligomer state. “Even though they have more plaques, they have better memory,” Mucke said. “What matters here is the critical pool of oligomeric species.”

The results not only call into question whether plaque-detecting diagnostics (see ARF related news story) will correlate well with cognitive impairments, but also caution against pharmaceuticals that might inadvertently increase Aβ oligomers. “This paper may warn people in drug development that preventing fibril formation could be dangerous if it is associated with increasing oligomers,” said Mucke. “We’re highlighting how shifting Aβ quickly into fibrils could diminish the pool of pathogenic oligomers.”—Molly McElroy

Molly McElroy is a freelance writer based in Melbourne, Florida.

Comments

  1. Alzheimer scientists have long known that amyloid plaques are not always associated with cognitive dysfunction. Indeed, some cognitively normal individuals have impressive numbers of insoluble amyloid plaques. Recent studies have shifted the focus to soluble amyloid oligomers as a primary pathological effector of AD. These observations suggest the possibility that shifting the aggregation state of Aβ from soluble oligomers to insoluble fibrils may actually be beneficial and decrease cognitive dysfunction. This hypothesis has been tested in a recent paper by Lennart Mucke and coworkers in the recent issue of the Journal of Biological Chemistry. These studies used expression of mutant forms of Aβ in transgenic mice as a means of manipulating the aggregation state of Aβ. The authors report that the Arctic mutation actually increases the rate of fibril formation, in contrast to earlier reports that indicated that the mutation enhances oligomerization rather than fibril formation. Expression of the “Arctic” (E22G) mutant leads to accelerated amyloid fibril deposition in plaques and reduced levels of Aβ soluble oligomers, including Aβ*56. Compared to mice expressing wild-type Aβ (J20) the mice expressing the Arctic mutation (ARC6) had more plaques, but no cognitive deficits, suggesting that the accumulation of Aβ in plaque deposits actually may be beneficial. This study highlights concerns that therapeutic strategies targeted at reducing amyloid fibril accumulation may actually be detrimental if they lead to accumulation of toxic amyloid oligomers instead. The results also suggest that drugs that promote fibril formation may be therapeutically useful. A recent report indicated that several drugs that specifically inhibit oligomer at substoichiometric concentrations actually promote amyloid fibril nucleation (Necula et al., 2007), so these drugs may provide an alternative means of testing the hypothesis that fibril formation at the expense of oligomer formation may be beneficial.

    References:

    . Small molecule inhibitors of aggregation indicate that amyloid beta oligomerization and fibrillization pathways are independent and distinct. J Biol Chem. 2007 Apr 6;282(14):10311-24. PubMed.

  2. Cheng et.al. (1) reported interesting results showing that acceleration of amyloid plaque formation by introducing the “Arctic” mutation, which causes familial Alzheimer disease (FAD), within the Aβ sequence rescues memory deficits in their transgenic mice expressing human amyloid precursor proteins. This rescue effect is correlated with the reduction in the level of Aβ*56, which was previously shown to impair memory in rodent models of AD (2). Although this conclusion is well consistent with recent reports suggesting soluble Aβ oligomers but not amyloid fibrils are causative agents for memory defects, their results also raise an intriguing question: Why does the Arctic mutation cause FAD, if Aβ-Arctic is less toxic than wild-type Aβ (Aβ-Wt)?

    As the authors mention, one explanation would be due to a lack of Aβ-Wt in their ARC mice. Authors discuss the possibility that Aβ-Wt affects the aggregation kinetics of Aβ-Arctic, resulting in the production of more toxic intermediate forms. Another possibility could be that, although Aβ-Arctic itself is prone to fibrilize and produce less Aβ*56, it may facilitate misfolding of Aβ-Wt as a “toxic seed,” and increase Aβ-Wt assemblies including Aβ*56. This possibility may be addressed by crossing their J20 and ARC6 mice and observing whether existence of Aβ-Arctic increases the level of Aβ*56 formed by Aβ-Wt and enhances memory defects.

    Although Aβ-Arctic is less toxic than Aβ-Wt in the extracellular environment, it may be more toxic within neurons. It has been shown that the Arctic mutation facilitates intraneuronal Aβ aggregation in transgenic mouse model (3). This may be because the Arctic mutation decreases the accessibility of α-secretase and increases the production of Aβ within neurons (4). Knobloch et al. (5) reported that the formation of intracellular Aβ deposits coincides with memory defects in their Aβ-Arctic transgenic mice. In our Drosophila models, where Aβ peptide is expressed in the secretory pathways of neurons, Aβ42-Arctic is much more toxic than Aβ42-Wt and causes more severe memory defects and neurodegeneration accompanied by accelerated intraneuronal Aβ deposits. It would be interesting to examine whether intracellular Aβ deposits are observed in the Cheng et al. mice models.

    References:

    . Accelerating amyloid-beta fibrillization reduces oligomer levels and functional deficits in Alzheimer disease mouse models. J Biol Chem. 2007 Aug 17;282(33):23818-28. PubMed.

    . A specific amyloid-beta protein assembly in the brain impairs memory. Nature. 2006 Mar 16;440(7082):352-7. PubMed. RETRACTED

    . The Arctic Alzheimer mutation facilitates early intraneuronal Abeta aggregation and senile plaque formation in transgenic mice. Neurobiol Aging. 2006 Jan;27(1):67-77. PubMed.

    . The Arctic Alzheimer mutation favors intracellular amyloid-beta production by making amyloid precursor protein less available to alpha-secretase. J Neurochem. 2007 May;101(3):854-62. PubMed.

    . Intracellular Abeta and cognitive deficits precede beta-amyloid deposition in transgenic arcAbeta mice. Neurobiol Aging. 2007 Sep;28(9):1297-306. Epub 2006 Jul 31 PubMed.

  3. Amyloid Spin Doctors
    It certainly would seem that the Alzheimer disease (AD) research community has completed the 180-degree turnaround on their view of the toxic amyloid-β entity, i.e., from fibrils to oligomers. The days of plaque busters are presumably gone and the once toxic fibrils are now viewed as friend, not foe. While our group is probably the last that will go on record as defending amyloid-β in any guise (Perry et al., 2000; Joseph et al., 2001; Rottkamp et al., 2002; Smith et al., 2002a, b; Smith et al., 2002c; Lee et al., 2004a; Lee et al., 2004b; Lee et al., 2005; Lee et al., 2006b, 2006a; Lee et al., 2007), this about face reveals much about the scientific method and those that rigidly ignore its principles. Simply, the old analogue methods of in vitro amyloid toxicity are being replaced by the new digital methods of behavior in transgenic animals. In the past, using cell culture paradigms, fibrillar was the enemy and soluble the friend. Nowadays, using transgenic models, the reverse is true. However, the big question that often gets overlooked is what happens in humans with AD (i.e., the condition that such cell culture exponents and transgenic models purport to represent). Fibrillar amyloid is a poor predictor of disease (Castellani et al., 2006). However, whether oligomeric amyloid is more predictive or informative is unclear. If oligomeric amyloid-β is as poorly selective and specific for AD as was the case for fibrillar amyloid-β (Castellani et al., 2006), then this is all spin and no step.

    Finally, a question of major importance is where this spin leaves the ongoing vaccine. It has been shown that the vaccine reduces fibrils but increases soluble amyloid-β (Lee et al., 2006a; Patton et al., 2006). If fibrils are good and soluble amyloid is bad, surely the trial should be canceled. While we doubt that this will happen, when, and if (not if, but when), the trial fails (Perry et al., 2000; Smith et al., 2002b), the Amyloid Spin Doctors will have the answer in hand!

    References:

    . Neuropathology of Alzheimer disease: pathognomonic but not pathogenic. Acta Neuropathol. 2006 Jun;111(6):503-9. PubMed.

    . Copernicus revisited: amyloid beta in Alzheimer's disease. Neurobiol Aging. 2001 Jan-Feb;22(1):131-46. PubMed.

    . Perspectives on the amyloid-beta cascade hypothesis. J Alzheimers Dis. 2004 Apr;6(2):137-45. PubMed.

    . Challenging the amyloid cascade hypothesis: senile plaques and amyloid-beta as protective adaptations to Alzheimer disease. Ann N Y Acad Sci. 2004 Jun;1019:1-4. PubMed.

    . Amyloid-beta in Alzheimer's disease: the horse or the cart? Pathogenic or protective?. Int J Exp Pathol. 2005 Jun;86(3):133-8. PubMed.

    . Amyloid-beta in Alzheimer disease: the null versus the alternate hypotheses. J Pharmacol Exp Ther. 2007 Jun;321(3):823-9. PubMed.

    . Amyloid-beta vaccination: testing the amyloid hypothesis?: heads we win, tails you lose!. Am J Pathol. 2006 Sep;169(3):738-9. PubMed.

    . Amyloid beta: the alternate hypothesis. Curr Alzheimer Res. 2006 Feb;3(1):75-80. 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.

    . Amyloid-beta junkies. Lancet. 2000 Feb 26;355(9205):757. PubMed.

    . The state versus amyloid-beta: the trial of the most wanted criminal in Alzheimer disease. Peptides. 2002 Jul;23(7):1333-41. PubMed.

    . Ill-fated amyloid-beta vaccine. J Neurosci Res. 2002 Aug 1;69(3):285. PubMed.

    . Predicting the failure of amyloid-beta vaccine. Lancet. 2002 May 25;359(9320):1864-5. PubMed.

    . Dangers of the amyloid-beta vaccination. Acta Neuropathol. 2002 Jul;104(1):110. PubMed.

    View all comments by Xiongwei Zhu
  4. I appreciate Drs. Iijima and Iijima-Ando’s comments on our paper and would like to address a couple of points they raised.

    1. “Why does the Arctic mutation cause FAD, if Aβ-Arctic is less toxic than wild-type Aβ (Aβ-Wt)?” Our study demonstrates that high levels of oligomers containing either Aβ-Arctic (line ARC48) or Aβ-Wt (line J20) are associated with AD-like behavioral and neuronal impairments. Thus, Aβ-Arctic and Aβ-Wt would be expected to have comparable toxicities under conditions that promote the accumulation of pathogenic oligomers. Additional factors were discussed in our paper as kindly acknowledged by the commentators.

    2. “It would be interesting to examine whether intracellular Aβ deposits are observed in the Cheng et al. mice models.” Although we have yet to use immuno-electron microscopy to address this issue, we have used a variety of standard immunohistochemical approaches to look for intracellular Aβ in our models. Using antibodies that detect Aβ, but not longer C-terminal APP fragments, we have found no intraneuronal accumulation of Aβ-specific immunoreactivities in lines J20 or ARC48, as compared with nontransgenic controls. Cell culture studies suggest that Aβ can be difficult to detect in intracellular compartments because of its rapid secretion (see [1] and studies cited therein). The papers cited by the commentators (2,3) primarily used antibodies for the detection of intracellular Aβ that cross-react with C-terminal APP fragments. Although they also used Aβ40- and Aβ42-specific antibodies, no negative controls were shown to exclude nonspecific cross-reactivities of these antibodies with the tissues analyzed. Therefore, I am not yet fully convinced that the intracellular immunoreactivity demonstrated in the cited papers actually represents Aβ.

    References:

    . Intracellularly generated amyloid-beta peptide counteracts the antiapoptotic function of its precursor protein and primes proapoptotic pathways for activation by other insults in neuroblastoma cells. J Neurochem. 2004 Dec;91(6):1260-74. PubMed.

    . Intracellular Abeta and cognitive deficits precede beta-amyloid deposition in transgenic arcAbeta mice. Neurobiol Aging. 2007 Sep;28(9):1297-306. Epub 2006 Jul 31 PubMed.

    . The Arctic Alzheimer mutation facilitates early intraneuronal Abeta aggregation and senile plaque formation in transgenic mice. Neurobiol Aging. 2006 Jan;27(1):67-77. PubMed.

  5. Searching for the Culprit
    Numerous scientific studies provide evidence that amyloid- β protein (Aβ) plays a prominent role at the early stages of Alzheimer disease (AD). The pathway and mechanisms by which Aβ mediates toxicity are less clear. Despite substantial evidence that early oligomeric aggregates are key toxic species, many therapeutic strategies are still targeted against fibrillar aggregates. This thorough and elegant study by Cheng et al. conducted by Lennart Mucke’s group shows that such strategies need to be seriously re-evaluated.

    Using three transgenic mouse lines which express different relative amounts of oligomers versus fibrils, Cheng et al. demonstrated that the amount of soluble oligomers but not insoluble fibrils deposited in amyloid plaques correlated with learning and memory impairments of these animals. As a tool to vary the amounts of oligomers and amyloid burden in the three transgenic lines, the Arctic mutation (E22G), which in humans leads to familial AD (FAD) and in vitro enhances protofibril and fibril formation, was taken advantage of. In addition, the Swedish and Indiana FAD mutations were introduced into all three lines to boost the production of the most pathogenic Aβ42 species.

    Of the three lines (J20 with Aβ42-WT and ARC6 and ARC48 both with Aβ42-Arctic), J20 had the lowest and ARC48 the highest amyloid burden with large differences between them. The amount of the non-fibrillar Aβ*56 aggregates, which were first shown by Lesne et al. (1) to impair memory in Tg2576 mice, was measured in all three lines at 3-4 months of age. Here, J20 and ARC48 had comparable levels of Aβ*56, while in ARC6 the level of Aβ*56 was significantly lower. Most of the measured functional deficits were observed in J20 and ARC48 but not in ARC6 lines, demonstrating that the level of Aβ*56 but not the plaque burden and thus the amount of fibrillar deposits was correlated with toxicity.

    Interestingly, even premature mortality of these animals was detected in J20 and ARC48 but not in ARC6. The fascinating results of the present work enable one to examine in future whether the premature mortality in the three Tg lines correlates with the neuronal loss, if any, and test different therapeutic approaches aimed at reducing Aβ*56 levels to hopefully avoid premature mortality in these animals.

    In the discussion of their paper, Cheng et al. brought up important issues that are worth commenting on from a mechanistic perspective of the computer simulation expertise. First, ex-situ AFM image showed a globular ellipsoidal shape of Aβ*56 dodecamers with a volume estimate of 125-175 nm3. In discussion, Cheng et al. derived the volume of a dodecamer based on an assumption of a two-stranded β-sheet monomer. We found a radius of ~2.5nm for computer-simulated globular Aβ40 and Aβ42 pentamers (2), a result that was in agreement with earlier small-angle neutron scattering (3) and AFM (4) findings. Assuming a spherical Aβ pentamer, this amounts to a volume of 65 nm3. For a globular dodecamer of the same average density, the volume is proportional to the number of molecules, and thus the estimated volume of a dodecamer would be 65 x 12/5 nm3 = 156nm3, a value that is in agreement with the ex-situ AFM-estimated volume by Cheng et al. Thus, no assumption regarding the size of a monomer is required to theoretically estimate the size of a globular dodecamer.

    Also, we observed in these simulations ellipsoidal forms of oligomers made of 10-13 Aβ42 molecules, as a result of merging of two smaller size oligomers (made of 5-7 molecules) into one. This agrees with the scenario proposed by Bitan et al. (6), by which the paranuclei (pentamer/hexamer) formation was followed by an assembly of two paranuclei into larger oligomers.

    Second, Cheng et al. discuss the influence of the Arctic mutation on Aβ toxicity. This has been studied by several groups and yielded seemingly contradictory results, with some studies showing that Aβ-Arctic is more toxic and the other studies showing no significant effect. Here one should consider the fact that full-length peptides Aβ40 and Aβ42 display different behaviors already at the stage of folding (5) and oligomer formation (6). Computer simulations using an efficient discrete molecular dynamics approach and intermediate-resolution protein model (2,7) yielded results in agreement with the above experimental findings of Lazo et al. (5) and Bitan et al. (6).

    Consequently, any change in the protein sequence, including the Arctic mutation, might affect folding and oligomer formation of Aβ40 and Aβ42 in distinct ways. In particular, a recent study by Yun et al. (7) demonstrated that Aβ40 oligomerization was dominated by intermolecular interactions between pairs of K16-E22 regions, while in Aβ42 the regions closer to the C-terminus (L34-A42) were the most involved in assembly into oligomers. These computer simulation results suggest that the Arctic mutation (E22G) should affect more strongly oligomerization and neurotoxic properties of Aβ40 than of Aβ42. On the other hand, because fibril formation and the structure of Aβ40 and Aβ42 fibrils appear similar, the effect of the Arctic mutation on fibril formation might be similar in both alloforms.

    References:

    . A specific amyloid-beta protein assembly in the brain impairs memory. Nature. 2006 Mar 16;440(7082):352-7. PubMed. RETRACTED

    . In silico study of amyloid beta-protein folding and oligomerization. Proc Natl Acad Sci U S A. 2004 Dec 14;101(50):17345-50. PubMed.

    . Structure determination of micelle-like intermediates in amyloid beta -protein fibril assembly by using small angle neutron scattering. Proc Natl Acad Sci U S A. 2002 Jan 8;99(1):150-4. PubMed.

    . Self-assembly of Abeta(1-42) into globular neurotoxins. Biochemistry. 2003 Nov 11;42(44):12749-60. PubMed.

    . On the nucleation of amyloid beta-protein monomer folding. Protein Sci. 2005 Jun;14(6):1581-96. PubMed.

    . Amyloid beta -protein (Abeta) assembly: Abeta 40 and Abeta 42 oligomerize through distinct pathways. Proc Natl Acad Sci U S A. 2003 Jan 7;100(1):330-5. PubMed.

    . Elucidation of amyloid beta-protein oligomerization mechanisms: discrete molecular dynamics study. J Am Chem Soc. 2010 Mar 31;132(12):4266-80. PubMed.

    View all comments by Brigita Urbanc

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Orlando: Adding to ADDLs: Where Are They; What Might They Be Doing?
  2. SfN: Amyloid Oligomers—Not So Elusive, After All? Part 1
  3. Atomic Force Microscope Shows Antibodies Can Prevent Fibrils
  4. Aβ Star is Born? Memory Loss in APP Mice Blamed on Oligomer
  5. Philadelphia: Diagnosis Light—Toward Infrared Plaques

Further Reading

Primary Papers

  1. . Accelerating amyloid-beta fibrillization reduces oligomer levels and functional deficits in Alzheimer disease mouse models. J Biol Chem. 2007 Aug 17;282(33):23818-28. PubMed.