T.M. Edwards

New perspectives on the mechanisms through which nitric oxide may affect learning and memory processes

Nitric oxide (NO) has been well established as a molecule necessary for memory consolidation. Interestingly, the majority of research has focused on only a single mechanism through which NO acts, namely the up-regulation of guanylate cyclase (GC). However, since NO and NO-derived reactive nitrogen species are capable of interacting with a broad array of enzymes, ion channels and receptors, a singular focus on GC appears short-sighted. Although NO inhibits the action of a number of molecules there are four, in addition to GC, which are up-regulated by the direct presence of NO, or NO-derived radicals, and implicated in memory processing. They are: cyclic nucleotide-gated channels; large conductance calcium-activated potassium channels; ryanodine receptor calcium release (RyR) channels; and the enzyme mono(ADP-ribosyl) transferase. This review presents evidence that not only are these four molecules worthy of investigation as GC-independent mechanisms through which NO may act, but that behavioural evidence already exists suggesting a relationship between NO and the RyR channel.

The role of intracellular calcium stores in synaptic plasticity and memory consolidation.

Memory processing requires tightly controlled signalling cascades, many of which are dependent upon intracellular calcium (Ca(2+)). Despite this, most work investigating calcium signalling in memory formation has focused on plasma membrane channels and extracellular sources of Ca(2+). The intracellular Ca(2+) release channels, ryanodine receptors (RyRs) and inositol (1,4,5)-trisphosphate receptors (IP3Rs) have a significant capacity to regulate intracellular Ca(2+) signalling. Evidence at both cellular and behavioural levels implicates both RyRs and IP3Rs in synaptic plasticity and memory formation. Pharmacobehavioural experiments using young chicks trained on a single-trial discrimination avoidance task have been particularly useful by demonstrating that RyRs and IP3Rs have distinct roles in memory formation. RyR-dependent Ca(2+) release appears to aid the consolidation of labile memory into a persistent long-term memory trace. In contrast, IP3Rs are required during long-term memory. This review discusses various functions for RyRs and IP3Rs in memory processing, including neuro- and glio-transmitter release, dendritic spine remodelling, facilitating vasodilation, and the regulation of gene transcription and dendritic excitability. Altered Ca(2+) release from intracellular stores also has significant implications for neurodegenerative conditions.

Inhibition of guanylate cyclase and protein kinase G impairs retention for the passive avoidance task in the day-old chick.

Nitric oxide (NO) is a highly labile chemical messenger which has previously been implicated in memory processes in a variety of learning paradigms and species. However, there is only limited evidence to suggest which enzymes are acted upon by NO during the formation of memory. The present study investigates the role of guanylate cyclase (GC) and protein kinase G (PKG) in a form of passive avoidance learning known to be dependent on nitric oxide activity. It was determined that in vivo pharmacological inhibition of GC using either 1H-[1,2,4]oxadiazolo[4,3,-a]quinoxalin-1-one or 6-anilino-5,8-quinolinedione resulted in two transitory memory retention deficits centred around 40 and 120 min posttraining, respectively. In contrast, inhibition of PKG with N-[2-(methylamino)ehtyl]-5-isoquinoline-sulfornamide hydrochloride (H-8) resulted in a single temporary retention loss centered at 120 min posttraining. These temporary retention losses appear to be specific to memory since they were dose-dependent and could not be explained by nonspecific performance effects. Further, these results suggest that these agents inhibit memory retrieval rather than formation, since memory is subsequently available. The current findings indicate that guanylyl cyclase mediates two memory retrieval processes, the latter of which appears to be PKG-dependent. In contrast, since inhibition of NO results in a permanent retention loss, it is suggested that NO is required for memory formation through GC-independent processes.

Pharmaco-behavioural evidence indicating a complex role for ryanodine receptor calcium release channels in memory processing for a passive avoidance task.

Calcium signalling is an important process underlying neuronal function and consequently behaviour. The release of calcium from intracellular stores via the ryanodine receptor calcium release (RyR) channel has been implicated in both synaptic plasticity and to a limited extent in memory processing. While past investigations have suggested a role for RyR channels in long-term memory, the present study suggests their action is more complex. Using a single trial passive avoidance task developed for the day-old chick, it is proposed that RyR channels are necessary both prior to the expression of long-term memory and also in retrieval processes. Specifically, 5 mM dantrolene (a specific RyR channel blocker) resulted in a persistent retention loss from 40 min post-training while 10 nM dantrolene produced a transient retention loss centred at 40 min post-training. We speculate that in the context of memory formation, RyR channels may be activated by nitric oxide and in the context of memory retrieval may lead to the activation of large conductance calcium-activated potassium BK(Ca) channels which, when blocked by 50 nM iberiotoxin, also demonstrated a transient retention loss centred at 40 min post-training.

Zaprinast consolidates long-term memory when administered to neonate chicks trained using a weakly reinforced single trial passive avoidance task.

A weakly reinforced variant of the single trial passive avoidance task developed for the day-old chick typically fails to consolidate long-term memory. However, administration of zaprinast, a phosphodiesterase (PDE) type 5 inhibitor, (ic; 10 microl/side) immediately post-training resulted in a dose-dependent increase in retention at test 180 min post-training. Further, 100 microM zaprinast resulted in high levels of retention at test 180 min post-training when administered from 10 min before training to 10 min after training. Finally, 100 microM zaprinast, when administered immediately post-training, resulted in the consolidation of long-term memory at a number of times of test extending as late as 24 h post-training. Inhibition of PDE type 5 is known to increase cellular cGMP levels. Previous investigations using a strongly reinforced variant of this task have suggested a role for cGMP in memory retrieval, we now postulate that cGMP is also necessary for memory formation in chicks trained using passive avoidance.

Inhibition of mGluR1 and IP3Rs impairs long-term memory formation in young chicks

Calcium (Ca(2+)) is involved in a myriad of cellular functions in the brain including synaptic plasticity. However, the role of intracellular Ca(2+) stores in memory processing remains poorly defined. The current study explored a role for glutamate-dependent intracellular Ca(2+) release in memory processing via blockade of metabotropic glutamate receptor subtype 1 (mGluR1) and inositol (1,4,5)-trisphosphate receptors (IP(3)Rs). Using a single-trial discrimination avoidance task developed for the young chick, administration of the specific and potent mGluR1 antagonist JNJ16259685 (500nM, immediately post-training, ic), or the IP(3)R antagonist Xestospongin C (5microM, immediately post-training, ic), impaired retention from 90min post-training. These findings are consistent with mGluR1 activating IP(3)Rs to release intracellular Ca(2+) required for long-term memory formation and have been interpreted within an LTP2 model. The consequences of different patterns of retention loss following ryanodine receptor (RyR) and IP(3)R inhibition are discussed.

A ryanodine receptor agonist promotes the consolidation of long-term memory in young chicks.

Young chicks were trained on a weakly reinforced variant of a single-trial discrimination avoidance task which typically fails to consolidate the long-term memory stage. The ryanodine receptor (RyR) agonist 4-chloro-m-cresol (500 microM, i.c.) persistently promoted high retention until at least 24 h post-training when administered between the time of training and 20 min post-training. The consolidation of the long-term memory stage by RyR activation implicates intracellular calcium release in triggering long-term memory.

HIV-1 protein gp120 rapidly impairs memory in chicks by interrupting the glutamate-glutamine cycle

Learning and memory impairments are frequently observed in patients suffering from AIDS Dementia Complex (ADC). These effects have been linked to the presence of gp120, an HIV viral coat glycoprotein. The present study investigated the possibility that gp120 prevents the uptake of extracellular glutamate by astrocytes, leading to an interruption of the glutamate-glutamine cycle and a subsequent impairment of memory. Ten microliters of 10nM gp120 was bilaterally injected into the region of the intermediate medial mesopallium of day-old chicks at various times before, or after, training using a single-trial passive avoidance task. Gp120 was found to significantly impair memory retention when injected 10-40 min after training. Memory impairments were evident within 5 min of gp120 administration and remained evident 24h later. Further, the amnestic effect of gp120 could be overcome with glutamine or with precursors of glutamate synthesis, but only weakly by glutamate. These results support the conclusion that the amnestic effect of gp120 is due to an impaired uptake of glutamate by astrocytes and a subsequent interruption of glutamine supply to neurones. The data indicate that the glutamate-glutamine cycle may be a useful therapeutic target in the treatment of ADC.

Inhibition of endogenous carbon monoxide production induces transient retention losses in the day-old chick when trained using a single trial passive avoidance learning task

Carbon monoxide (CO) is most often thought of as an exogenous toxin rather than as a possible endogenous nootrope. However, a limited number of studies have suggested that CO is necessary in memory processing for at least some tasks. While nitric oxide (NO) and CO are known activators of guanylyl cyclase (GC), only the effect of NO on GC has been extensively investigated as a mechanism underlying memory processing. The aim of the present study was to determine if inhibition of CO production would have an effect on memory processing. Using chicks trained on a single trial passive avoidance task, inhibition of CO production using zinc (II) deuteroporphyrin IX 2,4-bis ethylene glycol (ZnBG; 5 microM) resulted in two transient retention losses occurring at around 40 and 130 min post-training. The timing of these transient retention losses was similar to those observed following inhibition of GC, using the same species and task in a previous study. This supports the notion that CO is necessary in memory processing for this task and may act through a GC-dependent mechanism. As ZnBG also directly inhibits GC or nitric oxide synthase (NOS) at high concentrations, a second experiment was carried-out to confirm the specificity of ZnBG for heme oxygenase (HO) at the concentration used. The action of ZnBG was challenged with the HO agonist hemin (100 microM) and the transient deficits were abolished. This confirmed that the action of ZnBG on memory was through a CO-related mechanism rather than directly on GC or NOS. In this way the specificity of ZnBG (5 microM) for HO could be confirmed. The results support a role for endogenous CO in memory processing, possibly through activation of GC. In addition, the transient retention losses observed following administration of ZnBG suggest that CO may be necessary for memory retrieval and not formation as previously thought.

D-Lactate inhibition of memory in a single trial discrimination avoidance task in the young chick.

L-Lactate is a metabolite possibly able to meet some neuronal energy demands. However, a clear role for L-lactate in behaviour remains elusive. Administration of the inactive isomer D-lactate (1.75 mM; ic), immediately post-training, resulted in a persistent retention loss from 40 min post-training when used in conjuction with a single trial discrimination avoidance task designed for the young chick. Furthermore, 1mM noradrenaline (ic) administered 20 min post-training overcame the retention loss induced by D-lactate. Although not directly demonstrated in the current study, it is plausible that D-lactate inhibited memory processing by competing with L-lactate for uptake into neurons. The time of onset of the retention loss induced by D-lactate is in accord with findings where the action of noradrenaline is inhibited. The successful challenge of D-lactate inhibition by a high concentration of noradrenaline may suggest a relationship by some unidentified mechanism.