This is Part 1 of a 5-part series. See also Part 2, Part 3, Part 4, and Part 5.
24 December 2007. Last August, Alzheimer disease researchers met with colleagues from other scientific fields and with foundation and NIH representatives in Bar Harbor, Maine, for the seventh annual workshop on Enabling Technologies for Alzheimer Disease Research. The participants’ goal was to identify current knowledge gaps that block progress toward a deeper understanding of AD and therapy development, and to identify opportunities for bridging these gaps. The workshop focused on three areas. An update on recent research advances set the stage for an intensive focus on the AD risk gene ApoE, and for a discussion of the emerging role of lipids. The second session examined new thinking in protein folding, and the third session introduced technical advances in imaging of mouse and human tissue, as well as in-vivo imaging. A separate discussion grew out of recent advances toward a better understanding of the clinical observation that some people with AD have epilepsy-like seizures. The report below begins with a brief summary of major advances in AD research over the past 2 years. It then summarizes 2 days of presentations and discussion around the three main topics, and concludes with a list of research priorities culled from the prior proceedings. Readers will find a broad description of how the big questions in AD research are changing, as well as ideas to update their own studies. As always, comments are welcome.
First, an overview of advances in the past year and new research questions they raised.
Progranulin and TDP-43: The discovery of mutations in the gene progranulin gave a genetic identity to a large fraction of FTDL-U. This is a common form of frontotemporal dementia that is not caused by tau mutations and is sometimes confused with AD in the clinic. The progranulin discovery adds haploinsufficiency of a growth factor involved in regulatory signaling of the cell cycle, motility, and injury response to the range of underlying causes for age-related dementia. Soon after, TDP-43 proved to be the major constituent of the pathogenic inclusions seen in progranulin-related FTDL-U. TDP-43 is a nuclear protein of poorly understood function. It also occurs in nuclear and cytoplasmic inclusions of other neurodegenerative diseases, including some cases of AD and all cases of sporadic ALS. Together, these twin discoveries are driving a realignment of the clinico-pathological-genetic delineations of neurodegenerative protein deposition diseases. Both genes offer new areas for mechanistic study. For details, see AD/PD meeting progranulin report; AD/PD meeting TDP-43 report.
AD genetics: Duplication of gene loci on chromosome 21, encompassing the APP gene, were found to cause autosomal-dominant familial AD with cerebral amyloid angiopathy in six different families (Rovelet-Lecrux et al., 2006; 2007). As have earlier reports of α-synuclein triplication causing early-onset Parkinson’s (Singleton et al., 2003), this finding supports the hypothesis that changes leading to elevated expression of a pathogenic, aggregation-prone protein increase its concentration past its point of solubility and in this way drive down the age of disease onset. A separate discovery in LOAD genetics pointed to a cell biological way of increasing Aβ levels over time (Rogaeva et al., 2007). AD-associated variants of the sortilin-related receptor (SORLA) appear less able to perform the receptor’s proposed function of trafficking APP towards recycling endosomes and away from amyloidogenic processing in BACE-containing late endosomes. This finding resulted from a large collaborative effort involving sharing sample sets of 6,000 genotyped patients. The search for pathogenic polymorphisms has since expanded to analysis of samples from some 10,000 patients, and the gene is estimated to be a risk factor in up to 10 percent of LOAD patients. SORLA has a weaker population effect than ApoE4 but confers risk independently of ApoE4 (SORLA Alzgene page).
A physiological substrate for a preferred AD drug target, the β-secretase BACE, was discovered (Willem et al., 2006). The enzyme appears to cleave the protein type 3 neuregulin-1 during peripheral myelination early in life. It is unclear whether this cleavage participates in myelination in the adult CNS, but drug developers pursuing BACE inhibition will watch for potential demyelinating effects.
The search for structural information on γ-secretase made initial progress with EM images of purified complex that visualize a globular, hollow chamber to 120A resolution (Lazarov et al., 2006). Atomic-scale X-ray crystallography structures of this 19 transmembrane-domain complex remain a future goal.
A new method in human CSF measurement has advanced mechanistic research into Aβ accumulation during the years prior to AD. By continuously monitoring the fractional concentration of isotope-labeled forms of Aβ in research volunteers over a 36-hour period, researchers have found a way to determine in real time how much Aβ is generated and cleared per hour. In a first study in healthy people, about 7.6 percent of total CSF Aβ turned out to have been newly made every hour, and 8.3 percent degraded (Bateman et al., 2006). This makes it possible to ask whether FAD mutation carriers indeed overproduce Aβ, whether individuals with late-onset AD have a problem with Aβ production or clearance, how relative concentrations of Aβ40 and 42 change prior to clinical disease, and how CSF Aβ in humans responds to anti-amyloid therapies.
The question of how Aβ causes toxicity has generated intense interest in synaptic biology. One study found that local overproduction of Aβ in hippocampal slices decreases the number of GluR2 subunits of AMPA-type glutamine receptors in the postsynaptic membrane, and that the internalization of these receptors leads to a reduction in dendritic spine number (Hsieh et al., 2006). Another study similarly found that Aβ oligomers added to slices decrease both the response of glutamate receptors to glutamate and the density of dendritic spines (Shankar et al., 2007). Studies largely concur that Aβ oligomers induce a state similar to long-term depression at the post-synapse.
Analysis of how the nervous system loses function in mouse models of amyloid buildup is broadening the importance of the tau protein. A recent study showed that halving the amount of tau protein in APP-overexpressing lines protected the mice from known learning and memory deficits. Tau reduction changed neither the amount of Aβ nor neuritic dystrophy around amyloid plaques, but it did prevent important functional deficits. This implies that tau somehow mediates the soluble amyloid toxicity that impairs brain function; neuronal dysfunction induced by excitotoxins also required the presence of tau (Roberson et al., 2007). An emerging notion from this work is that tau might disturb cortical and hippocampal networks in AD. A follow-up study showed that APP- overexpressing mice have non-convulsant seizures, as do some AD patients, and that removing tau eliminates these seizures (Palop et al., 2007; see ARF news, Q&A, commentary). Mechanisms remain to be explored. Possibilities include a new function of tau, or effects secondary to its role in axonal transport. Network dysfunction in AD, as opposed to neuronal loss, is a budding research area that implies new treatment options.
Amyloid imaging is able to visualize the buildup of amyloid in the brain of still-healthy carriers of FAD mutations years before AD symptoms are expected to set in (Klunk et al., 2007). Ongoing PIB imaging studies in community-based aging cohorts and in people with MCI are geared toward documenting and quantifying amyloid more broadly throughout the prodromal phase of AD, and exploring its usefulness for differential diagnosis. Amyloid imaging in mouse models has recently become possible, as well. (For news and commentary, see Maeda et al., 2007; Pike et al., 2007; Villemagne, 2007; ARF related news story; Johnson et al., 2007.)
Finally, some experimental treatments have reached phase 3 trials. This includes Alzhemed (first phase 3 trial has failed), Flurizan (phase 3 completed enrollment), AAB-001 antibody, and γ-secretase inhibitor LY450139 (phase 3 trials in planning). An ADCS phase 3 trial of IvIg, an off-the-shelf pooled antibody preparation used for various immunological conditions, is also in planning.
No company-sponsored novel tau-based compounds are known to have entered clinical trials. A recent academic animal study has focused attention on a widely used immunosuppressant drug, FK506, as a possible candidate for suppressing early pathogenesis and inflammation in tauopathies including Pick’s, FTDP-17, but also perhaps AD. FK506 prevented microglial activation and subsequent tau pathology in a tauopathy mouse model (Yoshiyama et al., 2007). Plans are underway for small-scale preventive testing in carriers of tauopathy mutations. Ensuing discussion of trial failures at the workshop reinforced the point that experimental drugs often are given too late in the disease to be able to have a detectable effect.—Gabrielle Strobel.
This is Part 1 of a 5-part series. See also Part 2, Part 3, Part 4, and Part 5.