(Meeting report by Malcolm Leissring, Harvard Medical School.) With the exception of Down's syndrome and rare, familial forms of Alzheimer’s, there is little evidence that AD is attributable to overproduction of the Aβ peptide. Therefore, deficits in the proteolytic degradation or clearance of Aβ may be the driving force behind the cerebral accumulation in many-possibly even most-cases of AD (for a recent review, see Selkoe, 2001). In a clear sign of the growing interest in this area, several talks at the 8th International Conference on Alzheimer’s Disease and Related Disorders in Stockholm focused on the biology of two Aβ-degrading enzymes, insulin-degrading enzyme (IDE) and neprilysin (NEP).
Dennis Selkoe (Abstract 552) presented the first evidence that defects in IDE can lead to accumulation of the Aβ peptide in vivo. The work, led by Wes Farris and colleagues, focused on a novel animal model of type 2 diabetes which harbors naturally occurring mutations in IDE. Known as the GK rat model, these animals were developed through a breeding strategy that selected for animals that performed poorly in glucose-tolerance tests. Subsequent genetic analysis by researchers at the Karolinska Institute in Sweden revealed that the diabetic phenotype was associated with two missense mutations in IDE (H18R and A890V).
Selkoe reported that GK rats showed a significant ~15-30 percent defect in the degradation of exogenously introduced Aβ in soluble brain fractions as well as in NaCO3-washed membrane fractions and intact primary fibroblasts. Significantly, primary neuronal cultures from these animals accumulated approximately 55 percent more endogenous Aβ1-40 and ~100 percent more Aβ1-42 in the conditioned medium than in control cultures. The finding that a small (~20-30 percent) decrement in IDE-mediated degradation can lead to such a large (~50-100 percent) increase in Aβ accumulation implies that IDE plays an important endogenous role in the regulation of brain Aβ.
Consistent with this, Selkoe reported preliminary data from a collaboration with Suzanne Guénette and Rudy Tanzi showing that brain Aβ levels are elevated in IDE knockout mice. Elsewhere in the Stockholm conference, evidence of genetic linkage between IDE and late-onset AD was reported by several sources, including Rudy Tanzi (Abstract 1206) and Anthony Brookes (Abstract 1557), suggesting that we will be hearing more about IDE in the years to come.
The next speaker, Takaomi Saido (Abstract 553), reported on his continuing studies of neprilysin (NEP), another Aβ -degrading protease. In the brains of APP23 transgenic mice crossed with NEP heterozygous (+/-) knockout mice, Saido observed a 50 percent increase in Aβ levels. Surprisingly, insoluble Aβ levels in these mice were unchanged, and the increase in overall Aβ was attributable to an approximately two-fold increase in soluble Aβ levels. This finding suggests that different Aβ -degrading proteases may act preferentially on different pools of Aβ . NEP mRNA levels were also reported to decrease with age, and this decrement was associated with particular brain regions (e.g., CA3, terminal zones of perforant path and entorhinal cortex), suggesting that NEP deficiency may play a role in the age-associated increase in the risk of AD.
In degradation assays comparing wild-type Aβ to several intra-Aβ mutants, Saido’s group found that each of the mutants was degraded significantly more slowly by NEP, suggesting that decreased degradation may play a role in certain familial AD cases. Saido also reported significant genetic linkage between AD and a SNP located 159 nucleotides past the stop codon of the NEP gene on chromosome 3. Finally, Saido described a model on which the NEP transript might be regulated by a ligand-receptor system. Using an activity-staining approach, Saido identified somatostatin as a ligand that upregulates NEP levels and proposed ligand supplementation therapy as a novel therapeutic approach to AD.
Roger Nitsch (Abstract 554) rounded out the trio of talks focused on Aβ -degradation. Nitsch reported that NEP mRNA and protein levels were significantly elevated in Aβ PP transgenic mice for as long as 30 weeks following a single intracranial injection of Aβ 1-42. The rise in NEP levels was associated with the prevention of plaque formation and reduced astrogliosis. This surprising result contrasts curiously with his previously reported finding that intracranial Aβ injection causes increased hyperphosphorylation of tau--in the former case Aβ seems therapeutic, while in the latter it appears pathogenic. Perhaps mice doubly transgenic for Aβ PP and tau will be capable of settling the issue.
It should be noted that IDE and NEP are by no means the only proteases implicated in the degradation of Aβ. Genetic and biochemical evidence continues to suggest a role for other proteases such as endothelin-converting enzyme (Abstract 669) and plasmin and its proteolytic activators. In my opinion, the abundance of Aβ -degrading proteases-rather than representing an Achilles’ heel-provides a wide-ranging and nuanced palette of drug targets that may one day allow us to modulate specific pools of Aβ. Judging from the range of data presented at the Stockholm conference, the future looks bright for Aβ-degradation research.