29 April 2002. Last week, 80 researchers from Caribbean nations, the U.S., and countries in Europe, the Middle East, and Asia met in balmy Montego Bay, Jamaica, for a 5-day conference that explored common molecular mechanisms in neurodegenerative diseases. Organized by Amanda McRae of the University of the West Indies, and Mark Smith of Case Western Reserve University in Cleveland, Ohio, the conference featured talks on schizophrenia, Parkinson's and Alzheimer's diseases, and age-related changes generally. For more information, see conference website. Below are selected highlights:
Hormones Gone Haywire-What Fires Up the Cell Cycle in Alzheimer's?
Several meeting presentations converged on an emerging theme in AD research, namely the question whether the observed death of neurons early on in AD is a consequence of a misguided attempt by these terminally differentiated cells to divide. In recent years, studies by Thomas Arendt (Arendt et al, 2000), Inez Vincent (Vincent et al. 2001), Karl Herrup (Yang et al 2001, Mark Smith (Raina et al. 2001), and other labs have detected an activation of cell-cycle-related proteins in degenerating neurons, including Ki67, cyclin A, cyclin D, cyclin E, cdk4, cdk5, and cdk7. Indeed, Arendt, of the University of Leipzig in Germany published on the induction of proliferation markers in AD neurons back in 1995 (Gartner et al.). As Arendt pointed out in his meeting presentation, all of these markers function in early phases of the cell cycle, suggesting that some as-yet unknown trigger pushes neurons back into the cell cycle but that the neurons die once they reach the S (or DNA replication) phase. This phenomenon occurs predominantly in brain areas with the highest degree of plasticity throughout life, and it also correlates with areas particularly vulnerable to tangle formation, Arendt added. Moreover, neurons upregulate cell-cycle inhibitors, such as p16, probably in a protective response, he said. Also activated in degenerating neurons are signal transduction pathways that transmit mitogenic signals and can lead either to cell division or to inflammation and cell death; these pathways were the subjects of talks by Arendt, Smith, and Donna Bozycko-Coyne of the biotechnology company Cephalon Inc. in West Chester, Pennsylvania, who focused on the activation in AD models of the JNK pathway.
Perhaps the most ambitious new hypothesis attempting to unite cell-cycle data with other pathological observations in Alzheimer's was presented jointly by Craig Atwood of Case Western Reserve University, and Richard Bowen of the biotechnology company Voyager Pharmaceuticals in Raleigh, North Carolina. With Smith, these investigators propose that it is a menopause- and andropause-related increase in gonadotropin hormones that reawakens the neuronal cell cycle and, in doing so, creates tau and amyloid pathology besides driving the neurons to death. (For example, a dividing neuron must hyperphosphorylate tau in order to take apart existing microtubules.) The role of gonadotropins in AD has not been studied extensively.
Since age is the leading risk factor for AD, these researchers argue, its root cause must also be age-related. The decline of estrogen and testosterone production lifts the negative feedback control that sex steroid exert on the production of their inducing hormones via the hypothalamic-pituitary axis. Consequently, the concentration throughout the body of the gonadotropins follicle-stimulating hormone (FSH) and luteinizing hormone (LH) rises. These gonadotropins stimulate not only sex steroid synthesis, but also cell proliferation in the gonads where, incidentally, AβPP and PS-1 are highly expressed.
In support of this idea, Bowen quoted data by others suggesting that LH receptors are present on neurons, preferentially in the hippocampus. He presented his own results of a study of serum FSH and LH levels in nursing home residents, showing people with AD have elevated levels compared to controls (Short et al., 2001.)
Atwood then presented unpublished data suggesting that cultured neurons increase incorporation of the cell proliferation marker BrdU when given physiologic doses of LH. He also detected low levels of intraneuronal LH in normal aging brain and increased levels in AD brain. These investigators presented their hypothesis with a sense of urgency because a LH-lowering drug with an established safety record exists in clinical practice. Leuprolide is used as a prostate-cancer treatment and could be evaluated for its potential use in AD, the investigators say. (Bowen has founded a biotech firm to do that.) In initial tests of leuprolide in cultured cells, Atwood found that it is non-toxic at physiological levels and able to inhibit growth after 3 days in culture. These experiments were done in neuroblastoma cells and need to be repeated with neurons more directly relevant to age-related neurodegeneration, in slice cultures of aging brain, or other model systems. Perhaps most intriguingly, Atwood also presented preliminary in-vivo data suggesting that treating C57/Bl6 wild-type mice for 4 and 8 weeks with leuprolide decreased Aβ levels.
This is a fascinating hypothesis that should be tested further, Arendt and others commented. Yet it its also clear that the literature contains numerous reports of other mitogens and their receptors that are upregulated in early stages of AD brain, including nerve growth factor, insulin-like growth factor, fibroblast growth factor, platelet-derived growth factor, and hepatocyte growth factor. Finally, Arendt cautioned, nobody has so far disproved the possibility that the reactivated cell cycle markers might serve a physiological function, as has been shown for other genes first thought to be expressed only during development.
ApoE - The Elephant at the Party?
The biggest genetic risk factor in AD is the E4 allele of the cholesterol transport protein ApoE. Yet surprisingly few scientists work on this molecule, perhaps because its genetic complexity combined with its "slippery" lipid-binding properties make it a daunting subject. Two talks at the meeting dealt with ApoE.
Daniel Michaelson, of Tel-Aviv University in Israel, extended his presentation last November at the Neuroscience meeting in San Diego with new data. His approach aims to understand the role of different ApoE alleles in synaptic plasticity and learning in a normal, positive environment and in response to age-related stressors. His group uses C57/Bl6 ApoE knockout mice, as well as ApoE knockouts that express either a human E3 or a human E4 gene with the human regulatory sequences, such that the human ApoE is expressed in different tissues.
Michaelson et al. placed the mice in an enriched environment--known to stimulate synaptogenesis, neurogenesis, and learning--for eight weeks and then tested learning speed and working memory with variations of the T-maze test. Mice carrying the E3 allele learned faster and seemed to improve their working memory in response to the enriched environment, but the E4 mice did not.
To work toward the molecular underpinning of this effect, Michaelson measured concentrations of the synaptic marker synaptophysin and of nerve growth factor in the mice's brain. Levels for both proteins increased after enrichment in the hippocampus of controls and E3 mice but not E4 mice. Surprisingly, E4 mice did show increased synaptophysin and nerve growth factor levels in their cortex, suggesting that E4 mice respond to a rich environment but have a specific problem in the hippocampus. These initial results are based on immunoblots of whole cortex homogenates, however. Histochemistry of cortical subregions, including the entorhinal and transentorhinal cortex, might lead towards an understanding of region-specific differences in ApoE genotype and neuronal vulnerability, Michaelson added.
He also described experiments simulating head injury, a known risk factor for dementia. He reported data suggesting that E4 mice had a higher mortality than E3 or control (i.e. ApoE knockout) mice and that among the survivors, E3 mice showed better neurological recovery (reflexes, etc.) than the E4 mice. Histologically, their lesions were also smaller.
Again to begin testing molecular substrates for this effect, Michaelson et al measured AβPP levels in the head-injured mice. Head injury leads to increased AβPP and the α-secretase cleavage product of AβPP, thought to be neurotrophic. His data suggest that E3, but not E4, favors a-secretase processing of AβPP, Michaelson said.
Finally, Michaelson described an experiment simulating brain inflammation using LPS as a stimulus, and found that astrocyte activation, but not microglial activation, appear to be dependent on ApoE genotype.
David Holtzman of Washington University, St. Louis, described work in his lab, in collaboration with Kelly Bales and Steve Paul at Eli Lilly, aimed at defining the role ApoE might play in converting Aβ into forms with a high β-sheet content and the resulting neuritic toxicity of the fibrillar deposits. He compared PDAPP transgenic mice, which deposit Aβ as both diffuse plaques (non-fibrillar Aβ) and as amyloid (fibrillar Aβ), with PDAPP-ApoE knockout mice, which build up Aβ deposits at the same time but more slowly, and with no fibrillar amyloid plaques until very late in life. To examine individual human ApoE genotypes on an ApoE-free mouse background, his group then crossed these transgene-knockouts with strains transgenic for either human ApoE2, E3, or E4 expressed in glial cells. He found that the hippocampus of human E2- and, to a lesser degree, E3-expressing mice resembled the hippocampus of ApoE knockout mice. There was little to no amyloid or neuritic plaques, even at old ages. Yet with human E4, fibrillar Aβ deposits again built up more quickly. This suggests that the association of ApoE4, much more than E2, facilitates the formation of the β-sheet conformation. Parts of his talk were published this month (Fagan et al. 2002.)
More Food for Thought on Homocysteine and AD
Gladys Maestre of University of Zulia in Maracaibo, Venezuela, reported unpublished results from the Maracaibo Aging Study, a community-based observational study of all people older than 55 in the Santa Lucia neighborhood of Maracaibo. Maestre and her colleagues found that the prevalence of dementia in this cohort was high, with 9.5 percent having AD, and 8.4 percent having other forms of dementia. This is consistent with Richard Mayeux's studies of Alzheimer's among Caribbean Hispanics, she said.
At this meeting, Maestre reported results of an analysis of 970 AD patients showing that plasma homocysteine levels were significantly higher in people with AD or other dementias than in the cognitively normal. Folate concentrations were lower in those with vascular dementia, but not in AD patients, when compared to non-demented study subjects. Maestre said this data supports the view that vascular factors such as homocysteine contribute to AD pathology.-Gabrielle Strobel.