Complementary memory storage sites in mice: their development as affected by the competitive NMDA receptor antagonist, CPP.

Bitemporal injections of puromycin consistently induce amnesia of aversive maze learning in mice when administered within 3 days of training. These bitemporal puromycin injections lose their amnestic effectiveness if the latency between training and injection is extended beyond 6 days. Consistent with other evidence, we conclude that in our experimental paradigm, complementary memory storage sites normally develop in additional cerebral areas within 6 days following training. Previous experiments have indicated that the central adrenergic and cholinergic systems are critically involved in this process. We now present evidence that administration of the NMDA receptor antagonist, CPP, blocks the development of these complementary memory storage sites. As suggested by studies of long-term potentiation, NMDA receptor-dependent postsynaptic calcium appears to be essential for the development of these storage sites and indeed to trigger their development.

Long-term suppression of the development of complementary memory storage sites in mice: functional interdependence of acetylcholine and dopamine.

Bitemporal injections of puromycin consistently induce amnesia of aversive maze learning in mice when given within 3 days after training. These injections consistently fail to induce amnesia when given 6 or more days after training. Consistent with the evidence from other laboratories, we interpret these results to indicate that the initial, temporal memory storage sites are supplemented 6 days after training by the development of complementary storage sites in other cerebral areas. Previous experiments have shown that this process is suppressed for 30-60 days by a single SC injection of scopolamine, a muscarinic antagonist. We now find that this suppressive action of scopolamine can be completely nullified by haloperidol, a dopaminergic antagonist. This finding supports the view that there may be a therapeutic role for dopamine antagonists in the treatment of cognitive dysfunction associated with cholinergic loss.

Long-term suppression in mice of the development of complementary memory storage sites: effect of a muscarinic antagonist.

Bitemporal injections of puromycin consistently induce amnesia of aversive maze-learning in mice when administered within 3 days of training. These bitemporal puromycin injections lose their amnestic effectiveness, if the latency between training and injection is extended beyond 6 days. Consistent with other evidence, we conclude that in our experimental paradigm, complementary memory storage sites normally develop in additional cerebral areas during the 6 days following training. Previous experiments have indicated that the central adrenergic system is critically involved in this process. We now present evidence that the central cholinergic system is also critically involved. This conclusion is based upon our results with the muscarinic receptor antagonists, scopolamine and methyl scopolamine.

Long-term suppression of the cerebral spread of a memory: effects of idazoxan and clonidine.

Bitemporal injections of puromycin consistently induce amnesia of aversive maze-learning in mice when administered within 3 days of training. These bitemporal puromycin injections lose their amnestic effectiveness if the latency between training and injection is extended beyond 6 days. Consistent with other evidence, we believe that memory (in our task) "spreads" during the 6 days following training. Since previous experiments have indicated that the central noradrenergic system is involved in this process of "memory spread," we have examined the effect of stimulation or blockade of the alpha 2-receptor. To this end, we administered a single dose of the alpha 2-adrenoceptor antagonist, idazoxan, or the alpha 2-agonist, clonidine. Idazoxan (1 mg/kg, SC) had no effect on engram spread. Clonidine (25 micrograms-125 ng/kg, SC), by contrast, suppressed engram spread for at least 30 days after treatment. When mice were tested at 60 and 90 days after treatment, spontaneous recovery (i.e., engram spread) was evident in only about 50% of the clonidine treated mice. Coadministration of idazoxan with clonidine blocked the effects of clonidine on "memory spread."

Blockade of beta 1- but not of beta 2-adrenergic receptors replicates propranolol's suppression of the cerebral spread of an engram in mice.

Bitemporal injections of puromycin that primarily affect the hippocampal-entorhinal area induce amnesia of aversive maze-learning in mice for 3 days after training but are ineffective 6 or more days after training. At these later times, additional puromycin sites covering widespread forebrain areas are necessary to induce amnesia, a result that we attribute to the cerebral spread of the engram during the 6-day period. We have reported that blockade of about 60% of cerebral beta-adrenergic receptors by a single, subcutaneous injection of (-)-propranolol, a nonselective beta-receptor antagonist, inhibited engram spread for 60-90 days, at which time engram spread spontaneously occurred. In the present experiments using single doses of antagonists that appeared to block 60% of beta 2- or beta 1-adrenergic receptors, it was found that the selective beta 2 antagonist ICI 118,551 was without effect on engram spread, whereas the selective beta 1 antagonist betaxolol inhibited the spread for at least 3 months. Propranolol's effect consequently appears to be accounted for by its blockade of beta 1 receptors.

Blockade of peripheral beta-adrenergic receptors fails to suppress the cerebral spread of an engram in mice.

Using bitemporal injections of puromycin, we have reported observations interpreted to indicate that a systemic injection of (-)-propranolol (50 micrograms/kg) drastically suppressed the spread of an engram in mice from the hippocampalentorhinal area to widespread cerebral areas. The present experiments were made with a non-selective, irreversible beta-adrenergic receptor antagonist that fails to cross the blood-brain barrier in order to test the possibility that the propranolol-induced blockade of peripheral beta-receptors might contribute to its effect on engram spread. Prolonged blockade of peripheral receptors by the irreversible antagonist had no effect on engram spread, suggesting that propranolol's effect was centrally mediated.

The effect of the beta-adrenergic receptor antagonist, propranolol, on the cerebral spread of a memory trace in mice.

Bi-temporal injections of puromycin that primarily affect the hippocampal-entorhinal areas consistently induce amnesia of aversive maze-learning in mice for 3 days after training but are consistently ineffective if given 6 or more days after training. At these later times, additional puromycin sites covering widespread areas of the forebrain are necessary to induce amnesia. Consistent with other evidence, these observations are interpreted to indicate that the locus of the memory trace becomes more widespread during the 6-day period. A single subcutaneous injection of (-)-propranolol (50 micrograms/kg) given either before or 2 days after training suppressed engram spread for 60-90 days, at which time engram spread spontaneously occurred. This effect of propranolol was stereospecific. Suppression of engram spread persisted for a prolonged period in spite of the rapid recovery (about 4 hr), following treatment, of the normal level of specific binding of 3H-dihydroalprenolol in membrane preparations of the cerebral hemispheres and of 125I-pindolol in selected areas of the forebrain, diencephalon and brainstem.

Studies on memory: the cerebral spread of an engram in mice as affected by inhibitors of dopamine beta-hydroxylase.

Bitemporal injections of puromycin that primarily affect the hippocampal-entorhinal areas consistently cause amnesia of maze-learning in mice for 3 days after training but become consistently ineffective if given 6 or more days after training. At these later times, additional puromycin injection sites covering widespread areas of the forebrain are necessary to induce amnesia. These observations are interpreted to indicate that the locus of the engram has become more widespread within the 6-day period. Treatment with inhibitors of dopamine beta-hydroxylase for 3 days following training, retarded the spread of memory from a matter of days to a period of weeks. Repeated treatment with the inhibitors restricted engram spread for about 3 months; again spread was evident about a month after the last treatment. These observations imply that the mechanisms responsible for engram spread are capable of surviving for extraordinarily long periods of time.

Time-dependency of neurohypophyseal peptide attenuation of puromycin amnesia in mice.

[Ile3, Arg8]vasopressin (arginine-vasotocin), as well as the C-terminal tripeptides of the neurohypophyseal hormones arginine and lysine vasopressin, Pro-Arg-Gly-NH2 and Pro-Lys-Gly-NH2, were protective against puromycin-induced amnesia in mice when administered 24h before training. The N-protected tripeptide derivative, Z-Pro-Lys-Gly-NH2, was effective when given 5 days before training. The effectiveness of all peptides to attenuate puromycin-induced amnesia decreased as the interval between training and peptide treatment increased, indicating that the peptides influence memory processes, rather than general arousal. Z-Pro-Lys-Gly-NH2 was active at 24h after training, when the other peptides were no longer effective. Although it seems clear that neurohypophyseal hormones per se can attenuate puromycin-induced amnesia, these results are in line with the possibility that some portion of hormone action may be mediated via formation of longer-lived hormone fragments in the CNS.

Distribution, survival and biological effects in mice of a behaviorally active, enzymatically stable peptide: pharmacokinetics of cyclo(Leu-Gly) and puromycin-induced amnesia.

Cyclo(Leu-Gly), the enzymatically resistant diketopiperazine formally derived from the C-terminal dipeptide sequence of oxytocin, exhibits activity in several behavioral systems. The distribution of cyclo(Leu-14C(U)Gly) in brain, and the time course of the disappearance of this labeled peptide from brain and plasma after subcutaneous injection into mice have been studied. The intact peptide was distributed equally in the five cerebral areas studied, for up to 96 hours after injection. Two exponential components were determined for peptide disappearance rates in plasma and brain; peptide half-lives in plasma up to 10 hr and from 24--96 hr after injection were, respectively, 0.8 and 33 hr; in brain, 1.0 and 42 hr. The peptide was found to accumulate in brain intracellular space to some degree. The time course of distribution of labeled cyclo(Leu-Gly) in subcellular fractions of mouse brain was also examined, and the concentration of peptide in the synaptosomal fraction was significantly correlated with the degree of protection against puromycin-induced amnesia of a maze-learning test. The results obtained not only confirm that cyclo(Leu-Gly) penetrates brain tissue intact and remains intact after peripheral administration in order to exert its behavioral effects, but, moreover, suggest an intriguing dynamic relationship between peptide concentration in the synaptosomal fraction and behavioral activity.

ADH and related peptides: effect of pre- or posttraining treatment on puromycin amnesia.

The peptide Z-Pro-Leu-Gly-NH2 attenuated puromycin-induced amnesia in mice when administered 5 days prior to training, while arginine vasopressin, lysine vasopressin and cyclo(Leu-Gly), were effective when given 24 hr before training. The activity of all peptides to inhibit puromycin-induced amnesia decreased as the interval after training and before peptide administration increased, suggesting that the peptides influence memory processes rather than generalized arousal mechanisms.

Dose-response relationships in attenuation of puromycin-induced amnesia by neurohypophyseal peptides.

Intracerebral injections of puromycin one day after training of mice in a Y-maze cause amnesia when the animals are tested 7 days later. This amnesia was shown to be attenuated by various neurohypophyseal hormones, analogs and fragments, administered subcutaneously immediately after training. Dose-response relationships have been obtained for the attenuation of puromycin-induced amnesia in mice by selected neurohypophyseal peptides. All of the compounds tested reduce the amnesia in a dose-related way, suggesting that these peptides may interact with specific receptors to induce their central effect. Among the peptides studied the two most potent--i.e., those that cause substantial retention of memory at the lowest doses--are the neurohypophyseal hormone arginine vasopressin and Z-prolyl-leucyl-glycinamide (Z-MIF).

Effect of two inhibitors of dopamine beta-hydroxylase on maturation of memory in mice.

Bitemporal injections of puromycin that primarily affect the hippocampal-enthorhinal cortical areas suppress memory of maze-learning in mice for 3 days after training but are ineffective 6 or more days after training. At these later times, injections affecting widespread areas of the brain in addition to the hippocampal-entorhinal area are necessary for amnesia. These observations are interpreted to indicate that the locus of the memory trace has enlarged at 6 days to include other parts of the central nervous system in addition to the hippocampal-entorhinal area. To produce an imbalance of neurotransmitters and so to test their importance in enlargement of the memory trace's locus, we treated mice for 7 days after training with inhibitors of dopamine beta-hydroxylase. These mice, unlike untreated controls, developed amnesia after bitemporal injections of puromycin. In view of additional control experiments, we interpret these results to suggest that an imbalance of transmitters suppresses the normal enlargement of the locus of the memory trace.

Neurohypophyseal hormones, analogs, and fragments: their effect on puromycin-induced amnesia.

Neurohypophyseal hormones and several of their analogs, as well as N-terminal and C-terminal fragments, have been studied for their ability to attenuate puromycin-induced amnesia in mice. [8-Lysine]vasopressin, [8-arginine]vasopressin, and the analogs des-9-glycinamide-[8-lysine]vasopressin, [1-beta-mercaptopropionic acid, 8-lysine]vasopressin, [1,6-aminosuberic acid, 8-lysine]vasopressin, [4-leucine, 8-lysine]vasopressin, glycyl-glycyl-glycyl-[8-lysine]vasopressin, [1-beta-mercaptopropionic acid, 8-D-arginine]vasopressin, and [1,6-aminosuberic acid, 8-arginine]vasopressin are active. [8-Arginine]oxytocin as well as oxytocin and all of its other analogs tested are inactive with the striking exception of glycyl-glycyl-glycyl-oxytocin. The structural aspects of the neurohypophyseal hormones which appear to be important for significant activity in memory consolidation include the combination of a cyclic moiety containing the Tyr and Phe residues along with a basic residue in position 8. Another series of active compounds comprises C-terminal neurohypophyseal peptides and analogs thereof, including the naturally occurring Pro-Leu-Gly-NH2 and, most surprisingly, Leu-Gly-NH2, as well as its derivatives D-Leu-Gly-NH2 and the diketopiperazine, cyclo(-Leu-Gly-).

The effect of acetoxycycloheximide on rate of accumulation of cerebral catecholamines from circulating tyrosine as related to its effect on memory.

The rate of accumulation of newly synthesized catecholamines and endogenous catecholamine levels in mice were determined after an amnestic intracranial dose of acetoxycycloheximide. Rates of accumulation were found to be severly decreased for a period of 12 hr after injection. Total catecholamine levels were increased from 1 to 4 hr, then decreased 17 hr following injection. These results are discussed in terms of previous behavioral observations. It is pointed out that they weaken the assumption that acetoxycycloheximide's amnestic effects are solely due to inhibition of protein synthesis.

Puromycin's suppression of memory in mice as affected by caffeine.

It has previously been shown that expression of maze-learning in mice is blocked for long periods of time by puromycin injected intracerebrally one or more days after the training experience. Treatment with caffeine after training has now been found to reduce greatly the amnestic effects of puromycin. With a high dose of caffeine (200 mg/kg) this reduction is evident 6 days after treatment with puromycin. With a lower dose of caffeine (25 mg/kg) the effect becomes evident only after a more extended period of time. In view of control experiments, we suggest that caffeine modifies factors necessary for the expression of memory and that this alteration makes puromycin relatively ineffective in blocking memory.

Effect of corticotropin and desglycinamide 9 -lysine vasopressin on suppression of memory by puromycin.

The puromycin-induced blockade of expression of maze learning in mice can be prevented by subcutaneous administration of "Purified Cortrophin Gel" up to 3 days prior to training. A similar protective effect was not found when highly purified corticotropin was tested, but was observed after administration of desglycinamide(9)-lysine vasopressin. It appears that pressorpeptides are more potent protective agents than corticotropin when administered subcutaneously, and that vasopressin contamination of the commercial preparation was probably responsible for the protective effects previously reported. These experiments suggest that vasopressin and its congeners modify memory consolidation in such a way that the "expression" of memory becomes insensitive to puromycin.

Pituitary peptides and the suppression of memory by puromycin.

It has previously been shown that the expression of memory of maze-learning in mice is blocked by intracerebrally-injected puromycin one or more days after the training experience, and that bilateral adrenalectomy before training protects memory against this effect of puromycin. Purified cortrophin gel, injected subcutaneously up to 3 days before training, has now been found to give results like those obtained with adrenalectomy and, additionally, to provide a high degree of protection when injected up to 16 hr after training. Several possible explanations for these effects have been considered and found to be inadequate.

Studies on memory: distribution of peptidyl-puromycin in subcellular fractions of mouse brain.

Previous experience has led us to suggest that sustained blockage of memory in mice upon administration of puromycin is caused by the persistence of peptidyl-puromycin at synapses. The block can be removed and memory restored by intracerebral injections of water and by subcutaneous or intraperitoneal injections of several psychotropic drugs. The present experiments provide evidence that peptidyl-puromycin persists for long periods in subcellular fractions of synaptosomes (nerve endings), and that the peptide is lost from synaptosomal, but not from mitochondrial, fractions when memory is restored. These findings strengthen our interpretation of the mode of action of puromycin on memory.