Disrupting direct inputs from the dorsal subiculum to the granular retrosplenial cortex impairs flexible spatial memory in the rat.

In a changing environment, animals must process spatial signals in a flexible manner. The rat hippocampal formation projects directly upon the retrosplenial cortex, with most inputs arising from the dorsal subiculum and terminating in the granular retrosplenial cortex (area 29). The present study examined whether these same projections are required for spatial working memory and what happens when available spatial cues are altered. Consequently, injections of iDREADDs were made into the dorsal subiculum of rats. In a separate control group, GFP-expressing adeno-associated virus was injected into the dorsal subiculum. Both groups received intracerebral infusions within the retrosplenial cortex of clozapine, which in the iDREADDs rats should selectively disrupt the subiculum to retrosplenial projections. When tested on reinforced T-maze alternation, disruption of the subiculum to retrosplenial projections had no evident effect on the performance of those alternation trials when all spatial-cue types remained present and unchanged. However, the same iDREADDs manipulation impaired performance on all three alternation conditions when there was a conflict or selective removal of spatial cues. These findings reveal how the direct projections from the dorsal subiculum to the retrosplenial cortex support the flexible integration of different spatial cue types, helping the animal to adopt the spatial strategy that best meets current environmental demands.

Collateral rostral thalamic projections to prelimbic, infralimbic, anterior cingulate and retrosplenial cortices in the rat brain.

As the functional properties of a cortical area partly reflect its thalamic inputs, the present study compared collateral projections arising from various rostral thalamic nuclei that terminate across prefrontal (including anterior cingulate) and retrosplenial areas in the rat brain. Two retrograde tracers, fast blue and cholera toxin B, were injected in pairs to different combinations of cortical areas. The research focused on the individual anterior thalamic nuclei, including the interanteromedial nucleus, nucleus reuniens and the laterodorsal nucleus. Of the principal anterior thalamic nuclei, only the anteromedial nucleus contained neurons reaching both the anterior cingulate cortex and adjacent cortical areas (prefrontal or retrosplenial), though the numbers were modest. For these same cortical pairings (medial prefrontal/anterior cingulate and anterior cingulate/retrosplenial), the interanteromedial nucleus and nucleus reuniens contained slightly higher proportions of bifurcating neurons (up to 11% of labelled cells). A contrasting picture was seen for collaterals reaching different areas within retrosplenial cortex. Here, the anterodorsal nucleus, typically provided the greatest proportion of bifurcating neurons (up to 15% of labelled cells). While individual neurons that terminate in different retrosplenial areas were also found in the other thalamic nuclei, they were infrequent. Consequently, these thalamo-cortical projections predominantly arise from separate populations of neurons with discrete cortical termination zones, consistent with the transmission of segregated information and influence. Overall, two contrasting medial-lateral patterns of collateral projections emerged, with more midline nuclei, for example, nucleus reuniens and the interoanteromedial nucleus innervating prefrontal areas, while more dorsal and lateral anterior thalamic collaterals innervated retrosplenial cortex.

Time to retire the serial Papez circuit: Implications for space, memory, and attention.

After more than 80 years, Papez serial circuit remains a hugely influential concept, initially for emotion, but in more recent decades, for memory. Here, we show how this circuit is anatomically and mechanistically naïve as well as outdated. We argue that a new conceptualisation is necessitated by recent anatomical and functional findings that emphasize the more equal, working partnerships between the anterior thalamic nuclei and the hippocampal formation, along with their neocortical interactions in supporting, episodic memory. Furthermore, despite the importance of the anterior thalamic for mnemonic processing, there is growing evidence that these nuclei support multiple aspects of cognition, only some of which are directly associated with hippocampal function. By viewing the anterior thalamic nuclei as a multifunctional hub, a clearer picture emerges of extra-hippocampal regions supporting memory. The reformulation presented here underlines the need to retire Papez serially processing circuit.

A Direct Comparison of Afferents to the Rat Anterior Thalamic Nuclei and Nucleus Reuniens: Overlapping But Different.

Both nucleus reuniens and the anterior thalamic nuclei are densely interconnected with medial cortical and hippocampal areas, connections that reflect their respective contributions to learning and memory. To better appreciate their comparative roles, pairs of different retrograde tracers were placed in these two thalamic sites in adult rats. Both thalamic sites receive modest cortical inputs from layer V that contrasted with much denser projections from layer VI. Despite frequent overlap in layer VI, ventral prefrontal and anterior cingulate inputs to nucleus reuniens were concentrated in the deepest sublayer (VIb). Meanwhile, inputs to the anterior thalamic nuclei originated more evenly from both sublayers VIa and VIb, with the result that they were often located more superficially than the projections to nucleus reuniens. Again, while the many hippocampal (subiculum) neurons projecting to nucleus reuniens and the anterior thalamic nuclei were partially intermingled within the deep cellular parts of the subiculum, cells projecting to nucleus reuniens consistently tended to lie even deeper (i.e., immediately adjacent to the alveus). Variable numbers of double-labeled cells were present in those cortical and subicular portions where the two cell populations intermingled, though they remained in a minority. Our data also show how projections to these two thalamic sites are organized in opposing dorsal/ventral and rostral/caudal gradients across both the cortex and hippocampal formation. While the anterior thalamic nuclei are preferentially innervated by dorsal cortical sites, more ventral frontal sites preferentially reach nucleus reuniens. These anatomic differences may underpin the complementary cognitive functions of these two thalamic areas.

The anterior thalamic nuclei and cognition: A role beyond space?

The anterior thalamic nuclei are a vital node within hippocampal-diencephalic-cingulate circuits that support spatial learning and memory. Reflecting this interconnectivity, the overwhelming focus of research into the cognitive functions of the anterior thalamic nuclei has been spatial processing. However, there is increasing evidence that the functions of the anterior thalamic nuclei extend beyond the spatial realm. This work has highlighted how these nuclei are required for certain classes of temporal discrimination as well as their importance for processing other contextual information; revealing parallels with the non-spatial functions of the hippocampal formation. Yet further work has shown how the anterior thalamic nuclei may be important for other forms of non-spatial learning, including a critical role for these nuclei in attentional mechanisms. This evidence signals the need to reconsider the functions of the anterior thalamic within the framework of their wider connections with sites including the anterior cingulate cortex that subserve non-spatial functions.

Chemogenetics Reveal an Anterior Cingulate-Thalamic Pathway for Attending to Task-Relevant Information.

In a changing environment, organisms need to decide when to select items that resemble previously rewarded stimuli and when it is best to switch to other stimulus types. Here, we used chemogenetic techniques to provide causal evidence that activity in the rodent anterior cingulate cortex and its efferents to the anterior thalamic nuclei modulate the ability to attend to reliable predictors of important outcomes. Rats completed an attentional set-shifting paradigm that first measures the ability to master serial discriminations involving a constant stimulus dimension that reliably predicts reinforcement (intradimensional-shift), followed by the ability to shift attention to a previously irrelevant class of stimuli when reinforcement contingencies change (extradimensional-shift). Chemogenetic disruption of the anterior cingulate cortex (Experiment 1) as well as selective disruption of anterior cingulate efferents to the anterior thalamic nuclei (Experiment 2) impaired intradimensional learning but facilitated 2 sets of extradimensional-shifts. This pattern of results signals the loss of a corticothalamic system for cognitive control that preferentially processes stimuli resembling those previously associated with reward. Previous studies highlight a separate medial prefrontal system that promotes the converse pattern, that is, switching to hitherto inconsistent predictors of reward when contingencies change. Competition between these 2 systems regulates cognitive flexibility and choice.

Distributed interactive brain circuits for object-in-place memory: A place for time?

Rodents will spontaneously learn the location of an individual object, an ability captured by the object-in-place test. This review considers the network of structures supporting this behavioural test, as well as some potential confounds that may affect interpretation. A hierarchical approach is adopted, as we first consider those brain regions necessary for two simpler, 'precursor' tests (object recognition and object location). It is evident that performing the object-in-place test requires an array of areas additional to those required for object recognition or object location. These additional areas include the rodent medial prefrontal cortex and two thalamic nuclei (nucleus reuniens and the medial dorsal nucleus), both densely interconnected with prefrontal areas. Consequently, despite the need for object and location information to be integrated for the object-in-place test, for example, via the hippocampus, other contributions are necessary. These contributions stem from how object-in-place is a test of associative recognition, as none of the individual elements in the test phase are novel. Parallels between the structures required for object-in-place and for recency discriminations, along with a re-examination of the demands of the object-in-place test, signal the integration of temporal information within what is usually regarded as a spatial-object test.

Organisation of cingulum bundle fibres connecting the anterior thalamic nuclei with the rodent anterior cingulate and retrosplenial cortices.

Despite considerable interest in the properties of the cingulum bundle, descriptions of the composition of this major pathway in the rodent brain have not kept pace with advances in tract tracing. Using complementary approaches in rats and mice, this study examined the dense, reciprocal connections the anterior thalamic nuclei have with the cingulate and retrosplenial cortices, connections thought to be major contributors to the rodent cingulum bundle. The rat data came from a mixture of fluorescent and viral tracers, some injected directly into the bundle. The mouse data were collated from the Allen Mouse Brain Atlas. The projections from the three major anterior thalamic nuclei occupied much of the external medullary stratum of the cingulum bundle, where they were concentrated in its more medial portions. These anterior thalamic projections formed a rostral-reaching basket of efferents prior to joining the cingulum bundle, with anteromedial efferents taking the most rostral routes, often reaching the genu of the corpus callosum, while anterodorsal efferents took the least rostral route. In contrast, the return cortico-anterior thalamic projections frequently crossed directly through the bundle or briefly joined the internal stratum of the cingulum bundle, often entering the internal capsule before reaching the anterior thalamus. These analyses confirm that anterior thalamic connections comprise an important component of the rodent cingulum bundle, while also demonstrating the very different routes used by thalamo-cortical and cortico-thalamic projections. This information reveals how the composition of the cingulum bundle alters along its length.

Deconstructing the Direct Reciprocal Hippocampal-Anterior Thalamic Pathways for Spatial Learning.

The hippocampus is essential for normal memory but does not act in isolation. The anterior thalamic nuclei may represent one vital partner. Using DREADDs, the behavioral consequences of transiently disrupting anterior thalamic function were examined, followed by inactivation of the dorsal subiculum. Next, the anterograde transport of an adeno-associated virus expressing DREADDs was paired with localized intracerebral infusions of a ligand to target specific input pathways. In this way, the direct projections from the anterior thalamic nuclei to the dorsal hippocampal formation were inhibited, followed by separate inhibition of the dorsal subiculum projections to the anterior thalamic nuclei. To assay spatial working memory, all animals performed a reinforced T-maze alternation task, then a more challenging version that nullifies intramaze cues. Across all four experiments, deficits emerged on the spatial alternation task that precluded the use of intramaze cues. Inhibiting dorsal subiculum projections to the anterior thalamic nuclei produced the severest spatial working memory deficit. This deficit revealed the key contribution of dorsal subiculum projections to the anteromedial and anteroventral thalamic nuclei for the processing of allocentric information, projections not associated with head-direction information. The overall pattern of results provides consistent causal evidence of the two-way functional significance of direct hippocampal-anterior thalamic interactions for spatial processing. At the same time, these findings are consistent with hypotheses that these same, reciprocal interactions underlie the common core symptoms of temporal lobe and diencephalic anterograde amnesia.SIGNIFICANCE STATEMENT It has long been conjectured that the anterior thalamic nuclei might be key partners with the hippocampal formation and that, respectively, they are principally responsible for diencephalic and temporal lobe amnesia. However, direct causal evidence for this functional relationship is lacking. Here, we examined the behavioral consequences of transiently silencing the direct reciprocal interconnections between these two brain regions on tests of spatial learning. Disrupting information flow from the hippocampal formation to the anterior thalamic nuclei and vice versa impaired performance on tests of spatial learning. By revealing the conjoint importance of hippocampal-anterior thalamic pathways, these findings help explain why pathology in either the medial diencephalon or the medial temporal lobes can result in profound anterograde amnesic syndromes.

Stable Encoding of Visual Cues in the Mouse Retrosplenial Cortex.

The rodent retrosplenial cortex (RSC) functions as an integrative hub for sensory and motor signals, serving roles in both navigation and memory. While RSC is reciprocally connected with the sensory cortex, the form in which sensory information is represented in the RSC and how it interacts with motor feedback is unclear and likely to be critical to computations involved in navigation such as path integration. Here, we used 2-photon cellular imaging of neural activity of putative excitatory (CaMKII expressing) and inhibitory (parvalbumin expressing) neurons to measure visual and locomotion evoked activity in RSC and compare it to primary visual cortex (V1). We observed stimulus position and orientation tuning, and a retinotopic organization. Locomotion modulation of activity of single neurons, both in darkness and light, was more pronounced in RSC than V1, and while locomotion modulation was strongest in RSC parvalbumin-positive neurons, visual-locomotion integration was found to be more supralinear in CaMKII neurons. Longitudinal measurements showed that response properties were stably maintained over many weeks. These data provide evidence for stable representations of visual cues in RSC that are spatially selective. These may provide sensory data to contribute to the formation of memories of spatial information.

Separate cortical and hippocampal cell populations target the rat nucleus reuniens and mammillary bodies.

Nucleus reuniens receives dense projections from both the hippocampus and the frontal cortices. Reflecting these connections, this nucleus is thought to enable executive functions, including those involving spatial learning. The mammillary bodies, which also support spatial learning, again receive dense hippocampal inputs, as well as lighter projections from medial frontal areas. The present study, therefore, compared the sources of these inputs to nucleus reuniens and the mammillary bodies. Retrograde tracer injections in rats showed how these two diencephalic sites receive projections from separate cell populations, often from adjacent layers in the same cortical areas. In the subiculum, which projects strongly to both sites, the mammillary body inputs originate from a homogenous pyramidal cell population in more superficial levels, while the cells that target nucleus reuniens most often originate from cells positioned at a deeper level. In these deeper levels, a more morphologically diverse set of subiculum cells contributes to the thalamic projection, especially at septal levels. While both diencephalic sites also receive medial frontal inputs, those to nucleus reuniens are especially dense. The densest inputs to the mammillary bodies appear to arise from the dorsal peduncular cortex, where the cells are mostly separate from deeper neurons that project to nucleus reuniens. Again, in those other cortical regions that innervate both nucleus reuniens and the mammillary bodies, there was no evidence of collateral projections. The findings support the notion that these diencephalic nuclei represent components of distinct, but complementary, systems that support different aspects of cognition.

Do the rat anterior thalamic nuclei contribute to behavioural flexibility?

The rodent anterior thalamic nuclei (ATN) are vital for spatial memory. A consideration of their extensive frontal connections suggests that these nuclei may also subserve non-spatial functions. The current experiments explored the importance of the ATN for different aspects of behavioural flexibility, including their contribution to tasks typically associated with frontal cortex. In Experiment 1, rats with ATN lesions were tested on a series of response and visual discriminations in an operant box and, subsequently, in a water tank. The tasks included assessments of reversal learning as well switches between each discrimination dimension. Results revealed a mild and transient deficit on the operant task that was not specific to any stage of the procedure. In the water tank, the lesion animals were impaired on the reversal of a spatial discrimination but did not differ from controls on any other measure. Experiment 2 examined the impact of ATN damage on a rodent analogue of the 'Stroop', which assesses response choice during stimulus conflict. The lesion animals successfully acquired this task and were able to use contextual information to disambiguate conflicting cue information. However, responding during the initial presentation of conflicting cue information was affected by the lesion. Taken together, these results suggest that the ATN are not required for aspects of behavioural flexibility (discrimination learning, reversals or high-order switches) typically associated with the rat medial prefrontal cortex. The results from Experiment 2 suggest that the non-spatial functions of the ATN may be more aligned with those of the anterior cingulate cortex.

When is the rat retrosplenial cortex required for stimulus integration?

The rodent retrosplenial cortex is known to be vital for spatial cognition, but evidence has also pointed to a role in processing nonspatial information. It has been suggested that the retrosplenial cortex may serve as a site of integration of incoming sensory information. To examine this proposal, the current set of experiments assessed the impact of excitotoxic lesions in the retrosplenial cortex on two behavioral tasks that tax animals' ability to process multiple and overlapping environmental stimuli. In Experiment 1, rats with retrosplenial lesions acquired a negative patterning discrimination, a form of configural learning that can be solved only by learning the conjunction of cues. Subsequent transfer tests confirmed that both the lesion and control animals had solved the task by using configural representations. Furthermore, in Experiment 2, a 2nd cohort of retrosplenial lesion animals successfully acquired conditioned inhibition. Nevertheless, the same animals failed a subsequent summation test that assesses the ability to transfer what has been learned about one stimulus to another stimulus in the absence of reinforcement. Taken together, these results suggest that in the nonspatial domain, the retrosplenial cortex is not required for forming associations between multiple or overlapping environmental stimuli and, consequently, retrosplenial engagement in such processes is more selective than was previously envisaged. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Anterior thalamic nuclei, but not retrosplenial cortex, lesions abolish latent inhibition in rats.

The present study examined the effects of excitotoxic lesions in 2 closely related structures, the anterior thalamic nuclei and the retrosplenial cortex, on latent inhibition. Latent inhibition occurs when nonreinforced preexposure to a stimulus retards the subsequent acquisition of conditioned responding to that stimulus. Latent inhibition was assessed in a within-subject procedure with auditory stimuli and food reinforcement. As expected, sham-operated animals were slower to acquire conditioned responding to a stimulus that had previously been experienced without consequence, relative to a non-preexposed stimulus. This latent inhibition effect was absent in rats with excitotoxic lesions in the anterior thalamic nuclei, as these animals conditioned to both stimuli at equivalent rates. The retrosplenial lesions appeared to spare latent inhibition, as these animals displayed a robust stimulus preexposure effect. The demonstration here that anterior thalamic nuclei lesions abolish latent inhibition is consistent with emerging evidence of the importance of these thalamic nuclei for attentional control. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Retrosplenial cortex and its role in spatial cognition.

Retrosplenial cortex is a region within the posterior neocortical system, heavily interconnected with an array of brain networks, both cortical and subcortical, that is, engaged by a myriad of cognitive tasks. Although there is no consensus as to its precise function, evidence from both human and animal studies clearly points to a role in spatial cognition. However, the spatial processing impairments that follow retrosplenial cortex damage are not straightforward to characterise, leading to difficulties in defining the exact nature of its role. In this article, we review this literature and classify the types of ideas that have been put forward into three broad, somewhat overlapping classes: (1) learning of landmark location, stability and permanence; (2) integration between spatial reference frames; and (3) consolidation and retrieval of spatial knowledge (schemas). We evaluate these models and suggest ways to test them, before briefly discussing whether the spatial function may be a subset of a more general function in episodic memory.

Collateral Projections Innervate the Mammillary Bodies and Retrosplenial Cortex: A New Category of Hippocampal Cells.

To understand the hippocampus, it is necessary to understand the subiculum. Unlike other hippocampal subfields, the subiculum projects to almost all distal hippocampal targets, highlighting its critical importance for external networks. The present studies, in male rats and mice, reveal a new category of dorsal subiculum neurons that innervate both the mammillary bodies (MBs) and the retrosplenial cortex (RSP). These bifurcating neurons comprise almost half of the hippocampal cells that project to RSP. The termination of these numerous collateral projections was visualized within the medial mammillary nucleus and the granular RSP (area 29). These collateral projections included subiculum efferents that cross to the contralateral MBs. Within the granular RSP, the collateral projections form a particularly dense plexus in deep Layer II and Layer III. This retrosplenial termination site colocalized with markers for VGluT2 and neurotensin. While efferents from the hippocampal CA fields standardly collateralize, subiculum projections often have only one target site. Consequently, the many collateral projections involving the RSP and the MBs present a relatively unusual pattern for the subiculum, which presumably relates to how both targets have complementary roles in spatial processing. Furthermore, along with the anterior thalamic nuclei, the MBs and RSP are key members of a memory circuit, which is usually described as both starting and finishing in the hippocampus. The present findings reveal how the hippocampus simultaneously engages different parts of this circuit, so forcing an important revision of this network.

The rat retrosplenial cortex as a link for frontal functions: A lesion analysis.

Cohorts of rats with excitotoxic retrosplenial cortex lesions were tested on four behavioural tasks sensitive to dysfunctions in prelimbic cortex, anterior cingulate cortex, or both. In this way the study tested whether retrosplenial cortex has nonspatial functions that reflect its anatomical interactions with these frontal cortical areas. In Experiment 1, retrosplenial cortex lesions had no apparent effect on a set-shifting digging task that taxed intradimensional and extradimensional attention, as well as reversal learning. Likewise, retrosplenial cortex lesions did not impair a strategy shift task in an automated chamber, which involved switching from visual-based to response-based discriminations and, again, included a reversal (Experiment 2). Indeed, there was evidence that the retrosplenial lesions aided the initial switch to response-based selection. No lesion deficit was found on an automated cost-benefit task that pitted size of reward against effort to achieve that reward (Experiment 3). Finally, while retrosplenial cortex lesions affected matching-to-place task in a T-maze, the profile of deficits differed from that associated with prelimbic cortex damage (Experiment 4). When the task was switched to a nonmatching design, retrosplenial cortex lesions had no apparent effect on performance. The results from the four experiments show that many frontal tasks do not require the retrosplenial cortex, highlighting the specificity of their functional interactions. The results show how retrosplenial cortex lesions spare those learning tasks in which there is no mismatch between the internal and external representations used to guide behavioural choice. In addition, these experiments further highlight the importance of the retrosplenial cortex in solving tasks with a spatial component.

The retrosplenial cortex and object recency memory in the rat.

It has been proposed that the retrosplenial cortex forms part of a 'where/when' information network. The present study focussed on the related issue of whether retrosplenial cortex also contributes to 'what/when' information, by examining object recency memory. In Experiment 1, rats with retrosplenial lesions were found to be impaired at distinguishing the temporal order of objects presented in a continuous series ('Within-Block' condition). The same lesioned rats could, however, distinguish between objects that had been previously presented in one of two discrete blocks ('Between-Block' condition). Experiment 2 used intact rats to map the expression of the immediate-early gene c-fos in retrosplenial cortex following performance of a between-block, recency discrimination. Recency performance correlated positively with levels of c-fos expression in both granular and dysgranular retrosplenial cortex (areas 29 and 30). Expression of c-fos in the granular retrosplenial cortex also correlated with prelimbic cortex and ventral subiculum c-fos activity, the latter also correlating with recency memory performance. The combined findings from both experiments reveal an involvement of the retrosplenial cortex in temporal order memory, which includes both between-block and within-block problems. The current findings also suggest that the rat retrosplenial cortex comprises one of a group of closely interlinked regions that enable recency memory, including the hippocampal formation, medial diencephalon and medial frontal cortex. In view of the well-established importance of the retrosplenial cortex for spatial learning, the findings support the notion that, with its frontal and hippocampal connections, retrosplenial cortex has a key role for both what/when and where/when information.

The importance of mammillary body efferents for recency memory: towards a better understanding of diencephalic amnesia.

Despite being historically one of the first brain regions linked to memory loss, there remains controversy over the core features of diencephalic amnesia as well as the critical site for amnesia to occur. The mammillary bodies and thalamus appear to be the primary locus of pathology in the cases of diencephalic amnesia, but the picture is complicated by the lack of patients with circumscribed damage. Impaired temporal memory is a consistent neuropsychological finding in Korsakoff syndrome patients, but again, it is unclear whether this deficit is attributable to pathology within the diencephalon or concomitant frontal lobe dysfunction. To address these issues, we used an animal model of diencephalic amnesia and examined the effect of mammillothalamic tract lesions on tests of recency memory. The mammillothalamic tract lesions severely disrupted recency judgements involving multiple items but left intact both recency and familiarity judgements for single items. Subsequently, we used disconnection procedures to assess whether this deficit reflects the indirect involvement of the prefrontal cortex. Crossed-lesion rats, with unilateral lesions of the mammillothalamic tract and medial prefrontal cortex in contralateral hemispheres, were unimpaired on the same recency tests. These results provide the first evidence for the selective importance of mammillary body efferents for recency memory. Moreover, this contribution to recency memory is independent of the prefrontal cortex. More broadly, these findings identify how specific diencephalic structures are vital for key elements of event memory.