Two papers from the past week take aim at the β-sheet structure of proteins prone to pathological aggregation. A team led by Luis Serrano of the European Molecular Biology Laboratory in Heidelberg offers up a computer algorithm to help predict which proteins will aggregate and under what conditions. Vernon Ingram and colleagues of the Massachusetts Institute of Technology, in Cambridge, describe a small molecule with a big ability to inhibit the formation of Aβ fibrils in vitro.
It Takes Assumptions to TANGO
In an article published online September 12 in Nature Biotechnology, Serrano's group and their collaborators at the Flemish Institute for Biotechnology in Brussels describe how their program, dubbed TANGO, crunches recently derived information on the propensity of certain protein residues to promote β-sheet folding—in combination with data on the experimental environment such as pH, concentration, ionic strength, etc.—to predict the likelihood that a protein will form aggregates, as well as the portion of the protein sequence that likely underlies the aggregation. According to the project website (which you can visit if you're in a mood to TANGO), the algorithm is "based on simple physico-chemical principles of secondary structure formation extended by the assumption that the core regions of an aggregate are fully buried." (For another recent test drive of the TANGO algorithm, see Linding et al., 2004, in which Serrano and colleagues probe the relationship between protein structure and β-aggregation in globular and intrinsically disordered proteins.)
First author Ana-Maria Fernandez-Escamilla and colleagues report that when tested against experimental data, TANGO correctly predicted the aggregation propensity of 155 of 179 peptides, with 21 false positives and three false negatives. The researchers also tested a separate set of 71 disease-associated peptides and found that the program was correct on 65 of these.
Fernandez-Escamilla and colleagues paid special attention to Aβ, and their algorithm identified two regions on both Aβ40 and 42 that were likely to initiate aggregation. One of these turns out to be residues 17-21, home of familial AD mutations, while the second region was residues 31-36 of Aβ40 and 30-42 of Aβ42. TANGO accurately predicted a higher aggregation propensity of Aβ42 over Aβ40. "[T]he additional isoleucine and alanine induced a much higher aggregation propensity by recruiting a bigger part of the C-terminus into aggregation," write the authors. When tested against experimental data on aggregation of different familial AD mutants, the algorithm correctly predicted the propensity of the Dutch (E22Q), Arctic (E22G), and Flemish (A21G) variants, but not the Italian (E22K) variant.
Serrano's team mentions several provisos: TANGO is much better with peptides 30-40 amino acids or shorter, though it was pretty good with full-length proteins, too. And they close by noting that their program does not predict amyloid formation. "Although we found a good correlation between the aggregation tendency predicted by TANGO and the increase or decrease of amyloid formation in mutants of amyloidogenic peptides (Alzheimer peptides) and proteins (lysozyme and transthyretin), this does not mean that TANGO captures the specific sequence contribution for amyloid formation. Rather, it is probable that those peptides and proteins have amyloidogenic sequences and that β-sheet aggregation favors amyloidosis by promoting multimeric aggregated precursors," the authors write.
Small Is Good
In an article published this week in the online PNAS, Vernon Ingram’s research group at MIT update us on their quest for agents that decrease the propensity of Aβ to form β-sheets. First author Barbara Blanchard and colleagues report that their current favorite is 4,5-dianilinophthalimide (DAPH), known mainly as a protein-tyrosine kinase inhibitor. DAPH was identified in a screen of more than 3,000 small molecules. At low concentrations, when incubated with Aβ1-42, DAPH prevented the formation of fibrils, and it even manages to bust up already formed Aβ fibrils. This was apparently achieved by reducing the β-sheet content of the pools of Aβ.
Ingram and colleagues favor a theory of Aβ toxicity whereby soluble β-sheet-containing protofibrils or fibrils induce Ca2+ influx through AMPA glutamate receptors. "We hypothesize that those neurons that possess these particular receptors will be affected, perhaps explaining the very particular distribution of cell-type specificity in early AD," the authors write. In the current paper, they demonstrated that DAPH was able to significantly inhibit Aβ-induced Ca2+ influx into neuronal (CATH.a) cells.
Of obvious concern, DAPH is a protein-tyrosine kinase inhibitor with great specificity for the epidermal growth factor receptor kinase, but the authors are confident that with existing knowledge of the molecule's functional areas, they can eliminate the kinase inhibiting activity so that it might be safe for testing in AD or other protein misfolding disorders.—Hakon Heimer
- Linding R, Schymkowitz J, Rousseau F, Diella F, Serrano L. A comparative study of the relationship between protein structure and beta-aggregation in globular and intrinsically disordered proteins. J Mol Biol. 2004 Sep 3;342(1):345-53. PubMed.
- Fernandez-Escamilla AM, Rousseau F, Schymkowitz J, Serrano L. Prediction of sequence-dependent and mutational effects on the aggregation of peptides and proteins. Nat Biotechnol. 2004 Oct;22(10):1302-6. PubMed.
- Blanchard BJ, Chen A, Rozeboom LM, Stafford KA, Weigele P, Ingram VM. Efficient reversal of Alzheimer's disease fibril formation and elimination of neurotoxicity by a small molecule. Proc Natl Acad Sci U S A. 2004 Oct 5;101(40):14326-32. PubMed.