Roman strains as a psychogenetic model for the study of working memory: behavioral and biochemical data

Differences in memory performance: The effects of sex and reproductive experience on object recognition memory in high- and low-yawning Sprague‒Dawley rats

The novel object recognition (NOR) test is an efficient way to measure nonspatial memory in rodents. The NOR performance of female and male rats is sexually dimorphic because memory performance is better in the former than in the latter. In females, maternal experience enhances spatial memory. We used the NOR test to evaluate short- and long-term recognition memory in both sexes in the high- and low-yawning sublines of rats (HY and LY, respectively), which were generated via a strict inbreeding process from the Sprague‒Dawley (SD) strain for more than ninety generations. Additionally, we evaluated the effect of maternal experience using nulliparous, primiparous, biparous, and multiparous HY, LY and SD dams. Our results revealed that LY rats presented less thigmotaxis, with lower central square crosses and more vertical exploration in the open-field arena, suggesting that they experienced anxiety. Additionally, LY males performed significantly better than LY females in short- and long-term NOR memory, and LY males performed significantly better than SD rats did. Among females, two maternal experiences negatively affected short-term memory in the LY and HY sublines with respect to primiparous dams, and HY dams had better memory performance in the NOR test than did SD dams. Our findings suggest that the yawning sublines are suitable for studying the neurobiological basis of different memory processes under different endocrine conditions in highly inbred groups of rats.

Neurobehavioral Profiles of Six Genetically-based Rat Models of Schizophrenia- related Symptoms

Schizophrenia is a chronic and severe mental disorder with high heterogeneity in its symptoms clusters. The effectiveness of drug treatments for the disorder is far from satisfactory. It is widely accepted that research with valid animal models is essential if we aim at understanding its genetic/ neurobiological mechanisms and finding more effective treatments. The present article presents an overview of six genetically-based (selectively-bred) rat models/strains, which exhibit neurobehavioral schizophrenia-relevant features, i.e., the Apomorphine-susceptible (APO-SUS) rats, the Low-prepulse inhibition rats, the Brattleboro (BRAT) rats, the Spontaneously Hypertensive rats (SHR), the Wisket rats and the Roman High-Avoidance (RHA) rats. Strikingly, all the strains display impairments in prepulse inhibition of the startle response (PPI), which remarkably, in most cases are associated with novelty-induced hyperlocomotion, deficits of social behavior, impairment of latent inhibition and cognitive flexibility, or signs of impaired prefrontal cortex (PFC) function. However, only three of the strains share PPI deficits and dopaminergic (DAergic) psychostimulant-induced hyperlocomotion (together with prefrontal cortex dysfunction in two models, the APO-SUS and RHA), which points out that alterations of the mesolimbic DAergic circuit are a schizophrenia-linked trait that not all models reproduce, but it characterizes some strains that can be valid models of schizophrenia-relevant features and drug-addiction vulnerability (and thus, dual diagnosis). We conclude by putting the research based on these genetically-selected rat models in the context of the Research Domain Criteria (RDoC) framework, suggesting that RDoC-oriented research programs using selectively-bred strains might help to accelerate progress in the various aspects of the schizophrenia-related research agenda.

Neurobehavioral and neurodevelopmental profiles of a heuristic genetic model of differential schizophrenia- and addiction-relevant features: The RHA vs. RLA rats

The Roman High- (RHA) and Low-(RLA) avoidance rat lines/strains were generated through bidirectional selective breeding for rapid (RHA) vs. extremely poor (RLA) two-way active avoidance acquisition. Compared with RLAs and other rat strains/stocks, RHAs are characterized by increased impulsivity, deficits in social behavior, novelty-induced hyper-locomotion, impaired attentional/cognitive abilities, vulnerability to psychostimulant sensitization and drug addiction. RHA rats also exhibit decreased function of the prefrontal cortex (PFC) and hippocampus, increased functional activity of the mesolimbic dopamine system and a dramatic deficit of central metabotropic glutamate-2 (mGlu2) receptors (due to a stop codon mutation at cysteine 407 in Grm2 -cys407*-), along with increased density of 5-HT2A receptors in the PFC, alterations of several synaptic markers and increased density of pyramidal "thin" (immature) dendrític spines in the PFC. These characteristics suggest an immature brain of RHA rats, and are reminiscent of schizophrenia features like hypofrontality and disruption of the excitation/inhibition cortical balance. RHA rats represent a promising heuristic model of neurodevelopmental schizophrenia-relevant features and comorbidity with drug addiction vulnerability.

A Genetic Model of Impulsivity, Vulnerability to Drug Abuse and Schizophrenia-Relevant Symptoms With Translational Potential: The Roman High- vs. Low-Avoidance Rats

The bidirectional selective breeding of Roman high- (RHA) and low-avoidance (RLA) rats for respectively rapid vs. poor acquisition of active avoidant behavior has generated two lines/strains that differ markedly in terms of emotional reactivity, with RHA rats being less fearful than their RLA counterparts. Many other behavioral traits have been segregated along the selection procedure; thus, compared with their RLA counterparts, RHA rats behave as proactive copers in the face of aversive conditions, display a robust sensation/novelty seeking (SNS) profile, and show high impulsivity and an innate preference for natural and drug rewards. Impulsivity is a multifaceted behavioral trait and is generally defined as a tendency to express actions that are poorly conceived, premature, highly risky or inappropriate to the situation, that frequently lead to unpleasant consequences. High levels of impulsivity are associated with several neuropsychiatric conditions including attention-deficit hyperactivity disorder, obsessive-compulsive disorder, schizophrenia, and drug addiction. Herein, we review the behavioral and neurochemical differences between RHA and RLA rats and survey evidence that RHA rats represent a valid genetic model, with face, construct, and predictive validity, to investigate the neural underpinnings of behavioral disinhibition, novelty seeking, impulsivity, vulnerability to drug addiction as well as deficits in attentional processes, cognitive impairments and other schizophrenia-relevant traits.

Dopamine, Noradrenaline and Differences in Sexual Behavior between Roman High and Low Avoidance Male Rats: A Microdialysis Study in the Medial Prefrontal Cortex

Roman High- (RHA) and Low-Avoidance (RLA) outbred rats, which differ for a respectively rapid vs. poor acquisition of the active avoidance response in the shuttle-box, display differences in sexual activity when put in the presence of a sexually receptive female rat. Indeed RHA rats show higher levels of sexual motivation and copulatory performance than RLA rats, which persist also after repeated sexual activity. These differences have been correlated to a higher tone of the mesolimbic dopaminergic system of RHA rats vs. RLA rats, revealed by the higher increase of dopamine found in the dialysate obtained from the nucleus accumbens of RHA than RLA rats during sexual activity. This work shows that extracellular dopamine and noradrenaline (NA) also, increase in the dialysate from the medial prefrontal cortex (mPFC) of male RHA and RLA rats put in the presence of an inaccessible female rat and more markedly during direct sexual interaction. Such increases in dopamine (and its main metabolite 3,4-dihydroxyphenylacetic acid, DOPAC) and NA were found in both sexually naïve and experienced animals, but they were higher: (i) in RHA than in RLA rats; and (ii) in sexually experienced RHA and RLA rats than in their naïve counterparts. Finally, the differences in dopamine and NA in the mPFC occurred concomitantly to those in sexual activity, as RHA rats displayed higher levels of sexual motivation and copulatory performance than RLA rats in both the sexually naïve and experienced conditions. These results suggest that a higher dopaminergic tone also occurs in the mPFC, together with an increased noradrenergic tone, which may be involved in the different copulatory patterns found in RHA and RLA rats, as suggested for the mesolimbic dopaminergic system.

Spatial learning in the genetically heterogeneous NIH-HS rat stock and RLA-I/RHA-I rats: revisiting the relationship with unconditioned and conditioned anxiety

To characterize learning/memory profiles for the first time in the genetically heterogeneous NIH-HS rat stock, and to examine whether these are associated with anxiety, we evaluated NIH-HS rats for spatial learning/memory in the Morris water maze (MWM) and in the following anxiety/fear tests: the elevated zero-maze (ZM; unconditioned anxiety), a context-conditioned fear test and the acquisition of two-way active avoidance (conditioned anxiety). NIH-HS rats were compared with the Roman High- (RHA-I) and Low-Avoidance (RLA-I) rat strains, given the well-known differences between the Roman strains/lines in anxiety-related behavior and in spatial learning/memory. The results show that: (i) As expected, RLA-I rats were more anxious in the ZM test, displayed more frequent context-conditioned freezing episodes and fewer avoidances than RHA-I rats. (ii) Scores of NIH-HS rats in these tests/tasks mostly fell in between those of the Roman rat strains, and were usually closer to the values of the RLA-I strain. (iii) Pigmented NIH-HS (only a small part of NIH-HS rats were albino) rats were the best spatial learners and displayed better spatial memory than the other three (RHA-I, RLA-I and NIH-HS albino) groups. (iv) Albino NIH-HS and RLA-I rats also showed better learning/memory than the RHA-I strain. (v) Within the NIH-HS stock, the most anxious rats in the ZM test presented the best learning and/or memory efficiency (regardless of pigmentation). In summary, NIH-HS rats display a high performance in spatial learning/memory tasks and a passive coping strategy when facing conditioned conflict situations. In addition, unconditioned anxiety in NIH-HS rats predicts better spatial learning/memory.

The effect of partial reinforcement on instrumental successive negative contrast in inbred Roman High- (RHA-I) and Low- (RLA-I) Avoidance rats

Frustration is an emotional response that can be induced by the sudden devaluation of a reinforcer in the presence of greater reinforcement expectancies (e.g. instrumental successive negative contrast, iSNC). This emotional response seems to be similar to anxiety and can be attenuated by previous experiences of reward loss (e.g. partial reinforcement, PR, as opposed to continuous reinforcement, CR). In this study we used iSNC and PR procedures in order to compare the performance of two strains of rats psychogenetically selected on the basis of their emotional reactivity: the inbred Roman High- (RHA-I, low anxiety) and Low- (RLA-I, high anxiety) Avoidance rats. Animals were exposed to a straight alley, where they were changed from 12 pellets in the preshift phase (presented in 100% of trials-CR vs. 50% of trials-PR) to 2 pellets in the postshift phase, or exposed to 2 pellets throughout the training. The results indicated that the iSNC only appeared in RLA-I rats exposed to CR, as opposed to RLA-I animals exposed to PR and to RHA-I rats exposed to PR or CR. These data seem to support the implication of emotional responses in both iSNC and PR situations, and indicate that the behavioral reactivity to reward loss experiences is modulated by genetic variables.

Genetic animal models of anxiety

The focus of this review is on progress achieved in identifying specific genes conferring risk for anxiety disorders through the use of genetic animal models. We discuss gene-finding studies as well as those manipulating a candidate gene. Both human and animal studies thus far support the genetic complexity of anxiety. Clinical manifestations of these diseases are likely related to multiple genes. While different anxiety disorders and anxiety-related traits all appear to be genetically influenced, it has been difficult to ascertain genetic influences in common. Mouse studies have provisionally mapped several loci harboring genes that affect anxiety-related behavior. The growing array of mutant mice is providing valuable information about how genes and environment interact to affect anxious behavior via multiple neuropharmacological pathways. Classical genetic methods such as artificial selection of rodents for high or low anxiety are being employed. Expression array technologies have as yet not been employed, but can be expected to implicate novel candidates and neurobiological pathways.

Genetic basis of anxiety-like behaviour: a critical review

The way genetic and/or environmental factors influence psychiatric disorders is an enduring question in the field of human psychiatric diseases. Anxiety-related disorders provide a relevant example of how such an interaction is involved in the aetiology of a psychiatric disease. In this paper we review the literature on that subject, reporting data derived from human and rodent studies. We present in a critical way the animal models used in the studies aimed at investigating the genetic basis of anxiety, including inbred mice, selected lines, multiple marker strains, or knockout mice and review data reporting environmental components influencing anxiety-related behaviours. We conclude that anxiety is a complex behaviour, underlined not only by genetic or environmental factors but also by multiple interactions between these two factors.

Differential effects of CRH infusion into the central nucleus of the amygdala in the Roman high-avoidance and low-avoidance rats

Roman-high (RHA/Verh) and low (RLA/Verh) avoidance rats are selected and bred for rapid learning versus non-acquisition of two-way, active avoidance behavior in a shuttle box. RHA/Verh rats generally show a more active coping style than do their RLA/Verh counterparts when exposed to various environmental challenges. The central nucleus of the amygdala (CeA) is known to be involved in the regulation of autonomic, neuroendocrine and behavioural responses to stress and stress-free conditions, and it is considered in relation to coping strategies. Corticotropin-releasing hormone (CRH) seems to be a key factor in the control of the CeA output. Neuroanatomical studies have revealed that the majority of CRH fibers from the CeA have direct connections with autonomic regulatory nuclei in the brainstem, e.g. lateral parabrachial nucleus (lPB), ventrolateral periaquaductal gray (vlPAG). The modulating effects of CRH (30 ng) on CeA activity were studied by infusion of CRH into the CeA in freely moving male RHA/Verh and RLA/Verh rats under stress-free conditions. Heart-rate and behavioural activities were repeatedly measured before, during and after local administration of CRH or vehicle, after which early gene product FOS immunocytochemistry and CRH-mRNA in situ hybridisation were carried out in selected brain areas. CRH infusion into the CeA caused a long lasting increase in heart-rate and behavioural activation in the RHA/Verh rats, leaving the RLA/Verh rats unaffected. As a result of CRH infusion, the number of FOS positive cells in the CeA and lPB of RLA/Verh rats was increased whereas an opposite response was found in the RHA/Verh rats. However, CRH into the CeA of the Roman rat lines induced no pronounced effects on FOS staining in the vlPAG and CRH mRNA levels in the CeA. These results indicate that the CRH system of the CeA, connected with the output brainstem areas, is differentially involved in cardiovascular and behavioural responses.

Metyrapone reduces rat brain damage and seizures after hypoxia-ischemia: an effect independent of modulation of plasma corticosterone levels?

Hypoxia-ischemia is accompanied by abundant corticosterone secretion that could exacerbate brain damage after the insult. The authors demonstrate that the steroid synthesis inhibitor metyrapone (150 mg/kg subcutaneously) suppresses the hypoxia-ischemia-induced rise of plasma corticosterone levels (17.3 +/- 3.6 micrograms/dL) when compared with corticosterone-treated animals (72.2 +/- 4.8 micrograms/dL) immediately after hypoxia-ischemia. In parallel, metyrapone reduced brain damage (P < 0.05). Moreover, none of the metyrapone-treated animals displayed seizures, whereas seven of eight corticosterone-treated animals had seizures after hypoxia-ischemia. Although corticosterone administration in metyrapone-treated animals elevated plasma corticosterone levels (39.0 +/- 5.3 micrograms/dL), this did not result in a subsequent increase in brain damage and seizures when compared with metyrapone-treated animals. The authors conclude that metyrapone reduces brain damage and the incidence of seizures after hypoxia-ischemia but that this effect might partially be independent from its effect on modulating plasma corticosterone levels.

Stress and emotionality: a multidimensional and genetic approach

The use of behavioural tests aiming to assess the psychological components of stress in animals has led to divergent and sometimes arbitrary interpretations of animal behaviour. This paper presents a critical evaluation of behavioural methods currently used to investigate stress and emotionality. One of its main goals is to demonstrate, through experimental evidence, that emotionality may no longer be seen as a unidimensional construct. Accordingly, following a discussion about concepts, we propose a multiple-testing approach, paralleled by factor analyses, as a tool to dissociate and study the different dimensions of emotionality. Within this multidimensional context, genetic studies (illustrated here by different rat models) are shown to be particularly useful to investigate the neurobiology of stress/emotionality. A genetic approach can be used (i) to broaden and dissect the variability of responses within and between populations and (ii) to search for the molecular bases (i.e. genes and gene products) which underlie such a variability.

Cerebral tryptophan hydroxylase activity, and 5-HT1A receptor, 5-HT2A receptor, and 5-HT transporter binding in grouped and isolated Roman RHA and RLA rats: relationships with behaviours in two models of anxiety

Male Roman low-(RLA) and high-avoidance (RHA) rats differ when tested in the elevated plus-maze and the black/white box, but not when (isolated and) tested for their social interaction. Herein, we have analysed the impact of prior isolation on male Roman rats tested in the first two models of anxiety; moreover, because central serotonin (5-HT) systems in Roman rats have been scarcely studied, we have also analysed several anxiety-related indices of central serotonergic activity in grouped/isolated Roman rats. Group-housed RLA rats tested in the elevated plus-maze and the black/white box were less anxious than their RHA counterparts, thereby confirming our previous study. Isolation had anxiogenic (and hypolocomotor) effects, these being significant in RLA rats only. Tryptophan hydroxylase activity in midbrain (but not in cortex, hippocampus or hypothalamus) was lower in group-housed (but not in isolated) RLA rats than in RHA rats, a difference independent from changes in the regulatory properties of the enzyme. Neither midbrain and hippocampal [3H]8-hydroxy-2-(di-n-propylamino)-tetrlin binding at 5-HT1A receptors, nor midbrain [3H] citalopram binding at the 5-HT transporter was different between grouped/isolated RHA/RLA rats. Alternatively, a trend toward a lower hypothalamic [3H]citalopram binding in (group-housed) RLA rats than in RHA rats could be noted, whereas cortical [3H]ketanserin binding at 5-HT2A receptors was lower in RLA rats than in RHA rats, a difference prevented by prior isolation. This study opens the possibility that inter-line differences in 5-HT2A receptors partly (or totally) underlie the respective behaviours of RHA and RLA rats in the elevated plus-maze and the black/white box.

Environmental enrichment and postnatal handling prevent spatial learning deficits in aged hypoemotional (Roman high-avoidance) and hyperemotional (Roman low-avoidance) rats

The present study investigated the enduring effects of postnatal handling (administered during the first 21 days of life), and environmental enrichment (for a period of 6 months starting 3 weeks after weaning) on spatial learning in 24-month-old hypoemotional (Roman high-avoidance, RHA/Verh) and hyperemotional (Roman low-avoidance, RLA/Verh) rats. Two groups of 5-month-old rats from both lines were also included in the experiment as young controls. The Roman lines performed differently in the Morris water maze: Path lengths of RLA/Verh rats were shorter and they swam at lower speed than RHA/Verh rats, showing quicker and more efficient learning overall. Postnatal handling improved learning mainly in RHA/Verh rats, whereas environmental enrichment was able to prevent the deficits shown by aged controls of both lines. Young, enriched, and handled plus enriched animals exhibited better performance than impaired aged controls, to the point that aged enriched and handled plus enriched animals did not differ from young controls. Thus, besides indicating that RLA/Verh rats are better learners than RHA/Verh rats in the Morris water maze, this study demonstrates that environmental enrichment prevents the cognitive loss associated with aging, over the long term. Finally, the positive effects obtained with postnatal handling were dependent on the rat line.

Paradoxical differences in animal models of anxiety among the Roman rat lines

The Roman high-avoidance (RHA) and low-avoidance (RLA) rat lines have been selected upon their divergent active avoidance behaviours. On the basis of open-field behavioural analyses, it has been suggested that RLA rats display more emotionality than RHA rats. Herein, we have analysed the behaviours of male RHA and RLA rats in three tests putatively related to anxiety, namely the elevated plus-maze, the black/white box, and the social interaction test. In the elevated plus-maze, neither the number of total arm entries nor the percent number of open arm entries were different between the Roman lines. Alternatively, RLA rats spent more time on the open arms, compared to RHA rats. In the black/white box, both the latency to enter the black compartment, the number of shuttles, the time spent in the white compartment, and the general activity in the white compartment were higher in RLA rats, compared to their RHA counterparts. Lastly, socially isolated RHA and RLA behaved similarly when exposed to a social interaction test. It is suggested that under particular experimental conditions male RLA rats display less anxiety than male RHA rats, and that the open-field test may provide indices of activity rather than indices of anxiety.

GABAergic and dopaminergic transmission in the brain of Roman high-avoidance and Roman low-avoidance rats

The GABAergic and dopaminergic pathways in the central nervous system (CNS) play a pivotal role in the control of emotions and in the adaptive responses to stressful stimuli. The present study was aimed at characterizing a range of biochemical markers of GABA- and dopamine-mediated neurotransmission in the CNS of Roman high-avoidance (RHA/Verh) and Roman low-avoidance (RLA/Verh) rats, two psychogenetically selected lines that differ in their level of emotionality. The stimulatory effect of GABA on 36Cl- uptake was less pronounced in the cerebral cortex of RLA/Verh rats as compared to RHA/Verh rats, whereas no line-related changes were detected in [3H]GABA and [3H]flunitrazepam binding. On the other hand, the density of D1 dopamine receptors labeled with [3H]SCH 23390 was lower in the nucleus accumbens of RLA/Verh rats as compared to their RHA/Verh counterparts, whilst no line-dependent changes were observed in the binding parameters of D1 dopamine receptors in the striatum, amygdala, and prefrontal cortex. These biochemical differences may contribute to the distinct emotionality and responsiveness to the effects of psychoactive drugs of RHA/Verh and RLA/Verh rats.

The Maudsley rat strains as a probe to investigate noradrenergic-cholinergic interaction in cognitive function

Central noradrenergic function in relation to cognitive performance was studied in the Maudsley rat strains. Neurochemical studies revealed a higher response to acute stress in the locus coeruleus (LC) in the Maudsley reactives (MR) than in the Maudsley non-reactives (MNRA). Autoradiographic studies showed that MNRAs had greater 125I clonidine binding to alpha 2 receptors in LC, which was accompanied by a higher behavioral sensitivity to clonidine. MRs had a deficit in working memory, but were superior to MNRAs in two reference memory tasks. MRs displayed a stronger preference for novel objects, with no strain differences in general exploratory activity. The behavioral profile of the MRs is similar to rats treated with drugs which enhance noradrenergic function. Furthermore, MNRA rats had greater availability of muscarinic receptors, which correlated with behavioral performance in the spatial working memory task. The differences in noradrenergic and cholinergic systems and their relationship to the behavioral profile make the Maudsley strains a useful tool to probe the interaction between two neurotransmitter systems in cognitive function.

SESAME: a least-squares approach to the evaluation of protein structures computed from NMR data

A method is proposed for defining a probability distribution on an ensemble of protein conformations from a 2D NOE spectrum, while at the same time back-calculating the experimental spectrum from the ensemble. This enables one to assess the relative quality and significance of the conformations, and to test the consistency of the ensemble as a whole with the experimental spectrum. The method eliminates the need to integrate the cross-peak intensities and is surprisingly insensitive to random noise in the spectrum. In this communication, these advantages are demonstrated by applying the method to simulated data, for which the correct result is already known.

Development of monoamine systems after neonatal anoxia in rats

Neurochemical and morphological effects of neonatal anoxia on monoamine systems were studied after 100% N2 exposure for 25 min at 30 h postnatally (postnatal day 2-P2). At 20 min after anoxia, reductions of tissue levels of cerebellar noradrenaline (NA) and striatal dopamine (DA) and metabolites were seen, while 5-hydroxyindoleacetic acid (5-HIAA) was increased in cortex and cerebellum. At P7, NA increased in cerebellum, while serotonin (5-HT) and 5-HIAA decreased in cortex and cerebellum. At P21, increased hippocampal NA and striatal homovanillic acid (HVA) were found, while striatal 5-HT decreased and 5-HIAA increased in striatum and hippocampus. At P60, striatal 3,4-dihydroxyphenylacetic acid (DOPAC) and 5-HIAA levels were found to be enhanced. No effects were seen on 5-HT, tyrosine hydroxylase, or DARPP-32 immunostaining in cortex, hippocampus, and striatum. Thus, the neonatal anoxia induced both acute and persistent neurochemical abnormalities in monoamine systems that were not accompanied by morphological changes detectable with the methods used. The monoamine alterations found could be critically connected to the behavioral disturbances observed in rats after neonatal anoxia. The findings may also be of relevance to dysfunctions seen in humans after perinatal oxygen deficiency, e.g., the attention deficit hyperactivity disorder syndrome.

The Roman strains of rats as a psychogenetic tool for pharmacological investigation of working memory: example with RU 41656

This study examined the effects of RU 41656, a dopaminergic D2 agonist, on the differential working memory performances and on the differential activities of the neurochemical systems of the Roman high (RHA) and Roman low (RLA) avoidance strains of rats. Compared with RLA, RHA performed worse in three tests of working memory (spontaneous alternation, radial maze and object recognition) and had higher levels of exploratory locomotor activity. Hippocampal and frontal cortex choline acetyltransferase (ChAT) activities were lower in RHA. Frontal cortex DA and DOPAC levels, hippocampal and striatal 5-HT and NA levels were higher in RHA. RU 41656 induced a significant improvement in working memory performance of RHA, whereas in RLA it had no effect. It decreased exploratory locomotor activity in both strains. ChAT activity in hippocampus was not affected by RU 41656 in either strain, whereas in frontal cortex it was increased in RHA but not in RLA. Hippocampal NA levels were decreased by RU 41656 in RHA but not in RLA. These results confirm previous data concerning the promnesic effect of RU 41656 and extend the finding that the Roman strains are a psychogenetic model for the behavioural, neurochemical and psychopharmacological study of the working memory in rats.

Contribution of the Roman strains of rats to the elaboration of animal models of memory

Performances of male rats of the Roman High (RHA)- and Roman Low (RLA)-Avoidance strains were compared along four essential dimensions: working memory, reference memory, spontaneous locomotion and avoidance conditioning. Performances of RLA and RHA rats were significantly different for each dimension. As constantly reported, RHA rats were by far superior to RLA in avoidance conditioning. They had also higher levels of locomotor activity. On the opposite, RLA performed better than RHA in an appetitive working memory task, the delayed reinforced alternation, and were also superior in an appetitive reference memory task, the 5-unit linear maze. These results confirm the fact that RLA rats may acquire positively reinforced learning more rapidly than RHA rats and that the differences in active avoidance behavior between the two strains depend more on differential freezing behavior than on learning or memory capacities. Beyond the problem of the characterization of the Roman strains, these data might give indications on the relationships between behavioral tests widely used in rats, and on their use as memory models.