Greg M. Cole

Vitamin E protects nerve cells from amyloid βprotein toxicity

The amyloid beta protein (ABP) is a 40 to 42 amino acid peptide which accumulates in Alzheimers disease plaques. It has been demonstrated that this peptide and a fragment derived from it are cytotoxic for cultured cortical nerve cells. It is shown here that ABP and an internal fragment encompassing residues 25 to 35 (beta 25-35) are cytotoxic to a clone of PC12 cells at concentrations above 1 x 10(-9)M and to several other cell lines at higher concentrations. Between 10(-9) and 10(-11) M beta 25-35 protects PC12 cells from glutamate toxicity. The antioxidant and free radical scavenger vitamin E inhibits ABP induced cell death. These results have implications regarding the prevention and treatment of Alzheimers disease.

The regulation of amyloid β protein precursor secretion and its modulatory role in cell adhesion

The regulation and function of two forms of the amyloid beta protein precursor (ABPP) that are released into the growth-conditioned medium of the PC12 nerve cell line were examined. Nerve growth factor increases the release of the form of ABPP without the protease-inhibitor domain relative to the protein containing the protease inhibitor and increases the overall rate of ABPP secretion 2-fold. In contrast, fibroblast growth factor increases the rate of ABPP secretion approximately 7-fold. Both forms of the secreted ABPP molecule are, in turn, able to stimulate adhesion of PC12 cells to substrata to which they are adsorbed about 10-fold more efficiently on a molar basis than Iaminin.

Amyloid beta protein precursor is a mitogen.

The form of the secreted amyloid beta-protein precursor which contains the protease inhibitor sequence is mitogenic for Swiss 3T3 cells, while the precursor molecule lacking the protease inhibitor domain is not. A ten-fold stimulation of DNA synthesis occurs at 8 x 10(-9) M protein.

Solvent effects on beta protein toxicity in vivo

Human beta (1-40) and rat beta (1-42) were dissolved in three different solvents and stereotaxically injected into rat hippocampus with the contralateral side injected with control reverse sequence peptide or vehicle alone. Results at 1 week showed gross toxicity of the 35% acetonitrile solvent which was markedly enhanced by 3 nmol of beta protein but not by reverse sequence peptide. Beta peptide in water also appeared more toxic than reverse sequence, but the results were less clear cut. In contrast, 3 nmol of beta peptide in a cyclodextrin/PBS solution produced no marked short-term toxic effects. Peripheral injection of substance P failed to prevent toxicity. We conclude that solvent effects play a major role in acute beta protein neurotoxicity.

Lack of long-term effects after β-amyloid protein injections in rat brain

Rat β(1–42) peptide (β/A4) or phosphate buffered saline (PBS) was bilaterally injected into the hippocampus (HIP) or the lateral ventricle (ICV) of 3-month-old Fischer-344 rats. Fifteen months later, the animals ability to learn a spatial memory task was tested using the Morris water maze. Acquisition of the task was impaired by the bilateral injection of either peptide or PBS into the hippocampus. Hippocampal-injected animals showed an increased average latency to find the platform by approximately 6 s (p < 0.05). However, injection of rat β-peptide into the hippocampus or lateral ventricles failed to induce behavioral impairment when compared to vehicle injected controls. Retention of this task was not significantly impaired in any group. The spatial acuity test, a trial without the platform, revealed that both groups of animals that received hippocampal injections were impaired, spending 23% less time in the target quadrant compared to ICV-injected animals (p < 0.005). Hippocampal ChAT activity was decreased in β/A4-injected animals but not significantly (p < 0.06). β/A4-immunoreactivity was detected at the bottom of the needle track and the adjacent parenchyma of β/A4 hippocampal injected animals after 16 months. However, long-term in vivo deposition of β/A4 in both regions did not result in an upregulation of hippocampal amyloid precursor protein (APP) expression and there was no qualitative neuronal loss in the hippocampus. These data suggest that even small CA1 lesions in aging rats caused by two injections using a blunt 26-gauge needle produce a significant spatial navigation deficit in the Morris water maze, but in vivo rat β/A4 (1–42) deposition into the hippocampus or the lateral ventricles failed to induce specific long-term behavioral, biochemical, or histological effects.

An Endosomal-Lysosomal Pathway for Degradation of Amyloid Precursor Protein a

We previously reported evidence for a lysosomal degradative pathway for APP and C-terminal fragments thereof, based on Western and immunocytochemical analysis of drug-treated cells. Here, we verify the existence of a lysosomal degradative pathway for APP using pulse chase immunoprecipitation analysis of drug-treated cells and fibroblasts with and without a known lysosomal hydrolase targeting defect. The results are consistent with the hypothesis that part or all of the beta-protein domain of APP is normally degraded by lysosomes. A mechanism for beta-protein deposition based on this data is hypothesized.

Beta Amyloid Protein Clearance and Microglial Activation

Progression of AD involves a slow accumulation of Aβ peptide deposited extracellularly in the neuropil and vasculature of the brain. Aβ peptides of MW from 4 to 5 kD (Aβ40, Aβ42 or Aβ43) are normally produced by many cells, but the mutations in early onset familial AD (fAD) cause increased production of the rapidly aggregating Aβ1–42 (Borchelt et al. (1997); Younkin (1995)). However, in approximately 95% of cases of AD, there is Aβ accumulation without genetically increased Aβ production. Thus, in the vast majority of AD cases, which arise out of interaction of aging with genetic risk factors, other aspects of Aβ metabolism appear to be important. For example, the increased risk, and earlier onset, of AD from the apolipoprotein (Apo) E4 allele, which is strongest between 65 and 80 years of age, is not associated with increased Aβ production, but rather with reduced Aβ clearance or enhanced amyloid formation. Other potential genetic risk factors for late-onset AD may also influence AD pathogenesis at levels beyond Aβ production, including alpha-2 macroglobulin (Liao et al. 1998), alpha-1 antichymotrypsin (Thome et al. (1995)), interleukin 1 (Nicoll et al. (2000)), and transforming growth factor beta (Luedecking et al. (2000)). Although factors regulating Aβ degradation and clearance have received little attention compared with factors regulating Aβ production associated with early-onset AD genes, there is now a large enough literature to consider the issues.

Diet, Abeta Oligomers and Defective Insulin and Neurotrophic Factor Signaling in Alzheimer’s Disease

Epidemiology suggests that the risk for developing Alzheimer’s disease (AD) is reduced by the dietary omega-3 fatty acid docosahexaenoic acid (DHA) and increased by type II diabetes, corresponding to diet-induced obesity (metabolic syndrome). The brains of AD patients show evidence of insulin resistance, including insulin and neurotrophic factor signaling defects. We investigated whether the insulin signaling defects play a causal role in the loss of dendritic spines and arbor, which is associated with memory loss in AD brain. We focused on evaluating increases in cytosolic hyperphosphorylated insulin receptor substrate 1 (IRS-1) and losses of total cytosolic IRS-1, which are signatures of insulin resistance. IRS is critical as an adaptor protein, coupling insulin/trophic receptors to survival signaling via Akt (a serine threonine protein kinase). We demonstrated in cultured neurons that hyperphosphorylation of IRS-1 accompanied the Aβ oligomer (AβO)-induced loss of dendritic spines and arbor and the increases in the levels of phosphorylated tau (pTau) and activated c-Jun N-terminal kinase (JNK), a known tau kinase; these effects were antagonized by JNK-specific inhibitors as well as by omega-3 fatty acid DHA, a non-specific inhibitor of JNK. We observed IRS-1 signaling defects in three trangenic models of AD. In a Tg model of severe neuron loss, the 5x FAD APP/PS1 mice (Robert Vassar, Northwestern University, Chicago, IL) showed a loss of nuclear and stronger granular perikaryal and dendritic inclusion pIRS-1 in neurons from the cortex and CA1 region of the hippocampus by six months, mirroring changes seen in the AD brain. In the cortex of 22-month-old APPsw Tg2576 mice, total IRS-1 levels were reduced, consistent with its degradation by Aβ and the loss of insulin-like signaling. In the 3xTg-AD model, mice were placed on a high fat (21%) omega-3-depleting “diabetogenic” diet for four months, and they showed elevated phospho-IRS-1 (Ser 616) and JNK-sensitive phospho-tau (Ser 422) in the hippocampus. Four months of dietary treatment with fish oil or curcumin corrected hyperphosphorlation of JNK, IRS-1 and tau, but the combination of fish oil and curcumin was most effective at restoring cognitive defects (in Y maze). Thus, treatment with fish oil/DHA, curcumin or a combination, which both limit amyloid accumulation and/or JNK activation, has the potential to improve insulin/trophic signaling and downstream synaptic and cognitive deficits that cause AD. Consistent with these observations, several small clinical trials suggest that omega-3 supplements can arrest the progression of AD at early stages.

Erratum: The regulation of amyloid β protein precursor secretion and its modulatory role in cell adhesion (Neuron 3, 689-694, 1989)

[This corrects the article on p. 224-228 in vol. 46.].