All types of amyloidosis are structurally characterized by the cross beta-pleated sheet conformation of the fibrils, irrespective of their biochemical composition. The clinical observation that the anthracycline 4'-iodo-4'-deoxy-doxorubicin (IDOX) can induce amyloid resorption in patients with immunoglobulin light chain amyloidosis was the starting point for this investigation of its possible mechanism of action. IDOX binds strongly to all five types of natural amyloid fibrils tested: immunoglobulin light chains, amyloid A, transthyretin (methionine-30 variant), beta-protein (Alzheimer), and beta 2-microglobulin. Quantitative binding studies showed that IDOX, but not doxorubicin, binds strongly to amyloid fibrils. This binding is saturable and involves two apparently distinct binding sites with Kd values of 5.9 x 10(-11) M and 3.4 x 10(-9) M. IDOX inhibited in vitro insulin amyloid fibrillogenesis. In vivo studies using the experimental amyloid murine model confirmed the specific targeting of IDOX to amyloid deposits. Preincubation of amyloid enhancing factor with IDOX significantly reduced the formation of amyloid deposits. It is hypothesized that IDOX exerts its beneficial effects through the inhibition of fibril growth, thus increasing the solubility of existing amyloid deposits and facilitating their clearance. IDOX may represent the progenitor of a class of amyloid-binding agents that could have both diagnostic and therapeutic potential in all types of amyloidoses.

We identified a novel missense mutation in the apolipoprotein A-I gene, T2069C Leu(174) --> Ser, in a patient affected by familial systemic nonneuropathic amyloidosis. The amyloid deposits mostly affected the heart of the proband, who underwent transplantation for end-stage congestive heart failure. Amyloid fibrils of myocardial and periumbilical fat samples immunoreacted exclusively with anti-ApoA-I antibodies. Amyloid fibrils extracted from the heart were constituted, according to amino acid sequencing and mass spectrometry analysis, by an amino-terminal polypeptide ending at Val(93) of apolipoprotein A-I (apoA-I); no other significant fragments were detected. The mutation segregates with the disease; it was demonstrated in the proband and in an affected uncle and excluded in three healthy siblings. The plasma levels of high-density lipoprotein and apoA-I were significantly lower in the patient than in unaffected individuals. This represents the first case of familial apoA-I amyloidosis in which the mutation is outside the polypeptide fragment deposited as fibrils. Visualization of the mutation in the three-dimensional structure of lipid-free apoA-I, composed of four identical polypeptide chains, indicates that position 174 of one chain is located near position 93 of an adjacent chain and suggests that the amino acid replacement in position 174 is permissive for a proteolytic split at the C-terminal of Val(93).

The availability of a genetic model organism with which to study key molecular events underlying amyloidogenesis is crucial for elucidating the mechanism of the disease and the exploration of new therapeutic avenues. The natural human variant of β2-microglobulin (D76N β2-m) is associated with a fatal familial form of systemic amyloidosis. Hitherto, no animal model has been available for studying in vivo the pathogenicity of this protein. We have established a transgenic C. elegans line, expressing the human D76N β2-m variant. Using the INVertebrate Automated Phenotyping Platform (INVAPP) and the algorithm Paragon, we were able to detect growth and motility impairment in D76N β2-m expressing worms. We also demonstrated the specificity of the β2-m variant in determining the pathological phenotype by rescuing the wild type phenotype when β2-m expression was inhibited by RNA interference (RNAi). Using this model, we have confirmed the efficacy of doxycycline, an inhibitor of the aggregation of amyloidogenic proteins, in rescuing the phenotype. In future, this C. elegans model, in conjunction with the INVAPP/Paragon system, offers the prospect of high-throughput chemical screening in the search for new drug candidates.

Availability of living organisms to mimic key step of amyloidogenesis of human protein has become an indispensable tool for our translation approach aiming at filling the deep gap existing between the biophysical and biochemical data obtained in vitro and the pathological features observed in patients. Human β2-microglobulin (β2-m) causes systemic amyloidosis in haemodialysed patients. The structure, misfolding propensity, kinetics of fibrillogenesis and cytotoxicity of this protein, in vitro, have been studied more extensively than for any other globular protein. However, no suitable animal model for β2-m amyloidosis has been so far reported. We have now established and characterized three new transgenic C. elegans strains expressing wild type human β2-m and two highly amyloidogenic isoforms: P32G variant and the truncated form ΔN6 lacking of the 6 N-terminal residues. The expression of human β2-m affects the larval growth of C. elegans and the severity of the damage correlates with the intrinsic propensity to self-aggregate that has been reported in previous in vitro studies. We have no evidence of the formation of amyloid deposits in the body-wall muscles of worms. However, we discovered a strict correlation between the pathological phenotype and the presence of oligomeric species recognized by the A11 antibody. The strains expressing human β2-m exhibit a locomotory defect quantified with the body bends assay. Here we show that tetracyclines can correct this abnormality confirming that these compounds are able to protect a living organism from the proteotoxicity of human β2-m.

No abstract