Ben A. Bahr

Amyloid β protein is internalized selectively by hippocampal field CA1 and causes neurons to accumulate amyloidogenic carboxyterminal fragments of the amyloid precursor protein

A critical issue concerning Alzheimers disease is its selectivity, which leads to cellular degeneration in certain brain areas but not in others, and whether this pathogenic selectivity involves products of the amyloid precursor protein (APP). Here, we show that the amyloid β protein Aβ1–42 is accumulated gradually and is retained intact by field CA1, but not by other subdivisions, of organotypic hippocampal slice cultures. In contrast, the slightly shorter Aβ1–40 peptide was not sequestered selectively. Sequestration of Aβ1–42 was followed by the build‐up of carboxyterminal fragments of the endogenous precursor protein that were identified by immunoprecipitation. Unlike the peptide uptake, this induction appeared to be stochastic at the cellular level. In addition, the APP fragments were distributed more broadly within the CA1 pyramidal neurons than the sequestered Aβ1–42, and they appeared to be localized to synaptic terminals in the molecular layer of the dentate gyrus and in the stratum lacunosum‐moleculare of the subfield CA3. Concentrations of synaptophysin, a presynaptic marker, decreased as the number of neurons producing amyloidogenic species increased. These results indicate that exogenous Aβ1–42 sets into motion a sequence that involves 1) selective uptake of the peptide by vulnerable cells at risk in Alzheimers disease, 2) markedly enhanced production of amyloidogenic precursor material, and 3) slow deterioration of central synapses. J. Comp. Neurol. 397:139–147, 1998.

Induction of β-Amyloid-Containing Polypeptides in Hippocampus: Evidence for a Concomitant Loss of Synaptic Proteins and Interactions with an Excitotoxin

Long-term cultures of brain slices were used to test if the lysosomotropic agent chloroquine induces beta-amyloid-related peptides in hippocampus and if such effects are accompanied by other manifestations of brain aging. Chloroquine administration resulted in the appearance of a carboxyl-terminal fragment of the beta-amyloid precursor protein (APP); the 27-kDa antigen was detectable after 24 h, increased rapidly for 6-10 days, and was eliminated upon drug washout. Immunocytochemical analyses showed that beta-amyloid immunoreactivity accumulated in the perikarya of pyramidal neurons, primarily in the form of punctate bodies. These effects were accompanied by a correlated loss (and recovery) of the presynaptic marker synaptophysin and by a delayed reduction of postsynaptic glutamate receptors, while cytoskeletal proteins were unchanged. Acute administration of chloroquine had no evident effects on synaptic responses but prolonged applications caused a decrease in the maximum amplitude of field potentials. Finally, a brief pretreatment with the excitotoxin kainic acid had little effect with regard to APP fragments or synaptophysin, but altered the events following from a subsequent infusion of chloroquine. Buildup of the 27-kDa APP fragment and loss of synaptophysin were more rapid and, more importantly, did not reverse upon washout of chloroquine. These findings indicate that lysosomal dysfunction in hippocampus results in the accumulation of a particular APP fragment and suggest that this event, or a variable correlated with it, is linked to the loss of synaptic proteins. They also raise the possibility that certain aspects of brain aging reflect a synergism between lysosomal disturbances and excitotoxicity.

Stimulation of NMDA receptors activates calpain in cultured hippocampal slices

The hypothesis that intense stimulation of NMDA receptors activates calpain was tested in long-term cultures of hippocampus. Slices prepared from 10-day-old rats were maintained for periods of up to 6 weeks and then assayed for a stable breakdown product that results from the proteolysis of spectrin by calpain. The breakdown product increased dramatically during the first 24 h after tissue preparation and then decreased to a low level that remained unchanged for weeks. NMDA caused a 2- to 3-fold increase in breakdown product that rose linearly with time (5-30 min) and was blocked by the receptor antagonist MK-801. The effect of NMDA was the same throughout the culture period and was dependent upon the concentration of extracellular calcium with no effect at 2 mM and maximal effect at 4 mM calcium. These results indicate that rapid activation of calpain occurs in undamaged hippocampal neurons following stimulation of NMDA receptors.

Translational suppression of calpain I reduces NMDA-induced spectrin proteolysis and pathophysiology in cultured hippocampal slices.

Transfection of cultured hippocampal slices for five days with antisense oligonucleotides directed against mRNA encoding calpain I resulted in an approximately 60% decrease in the amount of caseinolytic activity stimulated by 10 microM calcium. Increases in a single proteolytic fragment of spectrin produced by 10-20 min of NMDA receptor stimulation were substantially (approximately 50%) reduced in antisense treated slices; this effect was not obtained in slices exposed to NMDA for 45 min. Attenuation of NMDA receptor-induced spectrin proteolysis by the antisense oligonucleotides was confirmed in immunoassays using antibodies that recognize multiple spectrin breakdown products and in immunocytochemical experiments with an antibody that detects an individual calpain I-mediated fragment. Translational suppression of calpain I did not detectably affect evoked synaptic responses but markedly improved their recovery from a 15 min infusion of NMDA. These results indicate that spectrin breakdown products provide a useful index of in situ calpain I activity and support the hypothesis that the protease plays a significant role in excitotoxicity.

Mouse telencephalon exhibits an age-related decrease in glutamate (AMPA) receptors but no change in nerve terminal markers

The central excitatory amino acid receptor selective for alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) was examined in brain tissue from mice at 3 and 25 months after birth. Antibodies against the rat GluR-A glutamate receptor subunit (selective for kainate and AMPA) labeled a mouse brain component of about M(r) 100,000. Telencephalic tissue from the older group of mice exhibited 31% less immunoreactivity towards this component as compared with that from the young group. Binding of [3H]AMPA also decreased with age in the telencephalon to an extent which was similar to the loss of receptor immunoreactivity. Scatchard analysis revealed that this reduction is due to a decrease in receptor density and not to a change in binding affinity. In contrast, there were only small age-related changes in AMPA receptor immunoreactivity and binding levels in the brain stem and cerebellum. Binding to dopamine, serotonin, or GABA receptors was not significantly reduced in the older mice. Since the nerve terminal markers synaptophysin and the SV2 glycoprotein were not detectably different in the two groups of mice, the age-related reduction in AMPA receptors is not likely to be due to a general decrease in synaptic density. These data suggest that glutamatergic neurotransmission mediated by AMPA-type receptors is selectively impaired with aging in the telencephalon.

Spectrin breakdown products increase with age in telencephalon of mouse brain

Calcium activated proteolysis of brain spectrin produces characteristic breakdown products (BDPs), the concentrations of which increase markedly in many instances of brain pathology. Results reported here indicate that levels of the BDPs rise with age (3-30 months) in the telencephalon but not in the hindbrain of Balb/c mice. These observations suggest that spectrin breakdown is a pathologic biochemical marker which increases with age in some but not all brain regions.

AMPA Receptor Development in Rat Telencephalon: [3H]AMPA Binding and Western Blot Studies

Abstract: Telencephalic membranes from rats of different embryonic (E16, E19) and postnatal (P2, P7, P14, adult) ages were assessed for α‐[3H]amino‐3‐hydroxy‐5‐methylisoxazole‐4‐propionic acid ([3H]AMPA) binding and for immunoreactivity levels of AMPA receptor subunits (GluR1, GluR2/3, and GluR4). In addition, the synaptic markers synaptophysin and NCAM140 (a neural cell adhesion molecule isoform) were examined by immunoblot. The density of [3H]AMPA binding sites increased steadily with advancing age. This increase was due mainly to the development of the large low‐affinity component (KD = 400 nM) that dominates the [3H]AMPA binding profile of adult rat brain membranes. As resolved by two‐site regression analysis, the high‐affinity component (KD = 15 nM) of the [3H]AMPA binding increased by approximately twofold from E16 to adult, whereas the low‐affinity component increased by 25‐fold. Staining for GluR1 and GluR2/3 increased steadily with increasing age at all time points examined; synaptophysin and NCAM140 exhibited similar ontogenic immunostaining profiles. GluR4 immunoreactivity was first evident at P14 and increased by adulthood. These results indicate that AMPA receptor density increases steadily during development and that this increase is coincident with the ontogenic expression of other synaptic components. Furthermore, there is a shift toward a preponderance of low‐affinity [3H]AMPA binding, which occurs during the period when AMPA receptors are being sorted into postsynaptic regions, suggesting that some element of the postsynaptic membrane environment modulates AMPA receptor properties.

Distinct distributions of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptor subunits and a related 53,000 Mr antigen (GR53) in brain tissue

Polyclonal antibodies against specific carboxy-terminal sequences of known alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptor subunits (GluR-4) were used to screen regional homogenates and subcellular fractions from rat brain. Affinity purified anti-GluR1 (against amino acids 877-899), anti-GluR2/3 (850-862), and anti-GluR4a and anti-GluR4b (868-881) labeled distinct subunits with the expected molecular weight of approximately 105,000. These antigens were shown to have distinct distributions in the brain. While GluR2/3 epitopes had a distribution profile similar to that of the presynaptic marker synaptophysin, GluR1 was notable for its abundance in the hippocampus and its relatively low density in neocortical areas, and GluR4 was highly enriched in cerebellar tissue. An additional antigen (glutamate receptor-related, GR53) of lower molecular weight (50,000-59,000) was recognized in rat, human, frog, chick and goldfish brain samples by anti-GluR4a as well as by anti-GluR1 at, an antibody that specifically recognizes the extracellular aminoterminal domain of GluR1 (amino acids 163-188). Both antibodies also labeled antigens of approximately 105,000 mol. wt in brain tissue from all species tested. The approximately 53,000 mol. wt antigen was concentrated 10-20-fold in synaptic membranes vs homogenates across rat brain regions. Both the 105,000 and the 53,000 mol. wt proteins were also concentrated in postsynaptic densities, and neither of the two antigens were evident in seven non-brain tissue samples. These data indicate that AMPA receptors have regionally different subunit combinations and that some AMPA receptor composites include proteins other than the conventional 105,000 mol. wt GluR subunits.

Variants of the receptor/channel clustering molecule gephyrin in brain: distinct distribution patterns, developmental profiles, and proteolytic cleavage by calpain.

The postsynaptic molecule gephyrin is involved in clustering neurotransmitter receptors. To test for protein variants that correspond to alternatively spliced gephyrin mRNAs, antibodies were made against 1) an amino‐terminal domain of gephyrin (GNN) and 2) its invariant carboxy‐terminus (GNC). Both antibodies recognized an antigen with the expected molecular weight of 93–95 kDa in rat and human brain tissue, as well as five additional proteins between 90 and 108 kDa. Most of these variants were found distributed throughout the brain, and their developmental profiles paralleled those of synaptic markers. Interestingly, the pattern of antigens immunostained across brain regions by anti‐GNN was markedly distinct from that labeled by anti‐GNC, a difference consistent with carboxy‐terminal modification. In control experiments in which hippocampal membranes were treated to activate endogenous proteases, there was no evidence that certain gephyrin variants originate from proteolysis. A subset of the antigens was, however, rapidly degraded during the treatment. A corresponding production of stable, carboxy‐terminal gephyrin fragments of 48–50 kDa occurred within 1 min of proteolytic activation and was blocked by the selective calpain inhibitor CX295. These findings suggest that multiple gephyrin proteins are active in the brain and that some of their roles may require functional modulation by limited proteolysis. J. Neurosci. Res. 49:381–388, 1997.

Age-Related Phosphorylation and Fragmentation Events Influence the Distribution Profiles of Distinct Tau Isoforms in Mouse Brain

Native tau isoforms were analyzed in adult mouse brain to determine whether they are differentially distributed and to identify molecular alterations that modify individual isoforms in an age-dependent manner. In general, the distribution profiles of 42–50 kDa tau were distinct from those of larger, hyperphosphorylated species of 55–69 kDa. The hippocampus and neocortex had concentrated levels of 55 kDa tau, and moderate amounts of 62–69 kDa isoforms. The latter species were similarly expressed in thalamic and hindbrain tissue; however, the noncortical regions were uniquely enriched in high molecular weight tau (97–110 kDa). When assessing hippocampal tau across age, increasing levels of 69 kDa tau were found to correlate with a gradual reduction in 42–50 kDa isoforms. Endogenous phosphatase activity induced an opposite correlation, thus supporting the idea that certain isoform conversions that occur with age stem from hyperphosphorylation. Age-related increases in 69 and 97 kDa tau also corresponded to enhanced levels of tau29, a putative tau fragment that exhibited an atypical localization (concentrated in olfactory bulb and hindbrain samples). These findings indicate that phosphorylation and fragmentation events influence tau distribution patterns, and that the former modification may promote the latter. They also raise the possibility that brain regions targeted by Alzheimer disease are distinguished by distinct tau profiles.