(172d) Regulation of Amyloidogenesis and Proteolysis in the Beta-Amyloid/Cathepsin B/Cystatin C Network

Cystatin C (CysC) is an important protease inhibitor of the central nervous system. Hereditary Cystatin C Amyloid Angiopathy (HCCAA) is a rare and fatal condition caused by misfolding, aggregation, and deposition of CysC variant L68Q. This single mutation enhances the unfolding rate of the monomer to a partially-unfolded intermediate state, allowing the protein to dimerize by domain-swapping. Since L68Q is prone to both domain-swapping and to fibrillogenesis, it has been hypothesized that CysC forms amyloid via propagated domain-swapping. We tested this idea by characterizing the stability of a domain-swapping-resistant CysC mutant, V57N. Contrary to the propagated domain-swapping hypothesis, V57N readily formed fibrils despite its inability to form domain-swapped dimers. Furthermore, we discovered that, at physiological conditions, CysC can self-associate into small meta-stable oligomers that retain the secondary structure of monomeric CysC and are not domain-swapped. These non-swapped oligomers exhibited some pre-amyloid characteristics, yet were more resistant to fibril formation than CysC monomers. We propose a new hypothesis that CysC amyloid aggregation initiates from the monomeric state and does not require domain-swapping. The Alzheimer’s peptide, Aβ, is well known to associate into amyloid fibrils. There are a few reports suggesting that CysC interacts with Aβ and may inhibit Aβ fibril formation, although the published data are inconclusive. We used several biophysical tools to compare the efficacy of CysC monomers (mCys), domain-swapped dimers (dCys), and non-swapped oligomers (oCys) in inhibiting Aβ aggregation. Both mCys and dCys caused a modest reduction in Aβ fibrillation. Surprisingly, oCys was a much more potent inhibitor of Aβ fibrillation. We hypothesize that this strong interaction is a specific consequence of oCys pre-amyloid structure, suggesting that heterotypic interactions between pre-amyloid-like oligomers arising from different proteins may potently arrest further maturation into fibrils.

Because CysC can interact with Aβ and inhibit its aggregation, it has been speculated that CysC could be a neuroprotective agent in Alzheimer’s, a hypothesis supported in some but not all transgenic animal studies. However, one of the primary functions of CysC is to inhibit the cysteine protease cathepsin B (catB). CatB is known to proteolytically degrade monomeric Aβ into harmless fragments, and may also digest preformed oligomers and protofibrils. This raises the question: what is the net outcome of the interactions of these three proteins, when CysC binds to Aβ, CatB degrades Aβ, and CysC inhibits CatB. Does CysC inhibit Aβ fibrillogenesis, or does CysC inhibit CatB, preventing the degradation of Aβ? In this way, the putative neuroprotective functions of CysC and CatB are antagonized by the other’s presence. Further complicating the situation, monomeric CysC is metastable, and as described above, it can re-fold into domain-swapped dimers that no longer inhibit CatB, or it can associate into oligomers that retain CatB inhibitory activity but interact more strongly with Aβ.

We developed a mathematical model to simulate the complex interactions between Aβ, CatB and CysC. We measured CysC binding affinity with Aβ through a solution-based FRET assay. It has generally been believed that CatB activity is constrained to acidic cellular compartments and is drastically diminished upon secretion to neutral extracellular fluids. However, we monitored CatB proteolysis of Aβ by and discovered that even at neutral pH, this reaction was fast enough to be physiologically relevant. We quantified enzymatic degradation rates of Aβ₄₂, Aβ₄₀, and CysC by CatB to determine k_cat/K_M values. We validated the model with experimental data, then used the model to define a set of conditions under which CysC and CatB concentrations critically affect the amount of free Aβ and the rate of Aβ amyloidogenesis. Several published studies have explored the relationship between CysC and Aβ in animal models, with researchers reaching different conclusions. Our analysis can be used to interpret these in vivo studies mechanistically, and to propose novel strategies for controlling amyloidogenesis and interfering with disease processes.