The medial prefrontal cortex (mPFC) has been associated with diverse functions including attentional processes, visceromotor activity, decision making, goal directed behavior, and working memory. Using retrograde tracing techniques, we examined, compared, and contrasted afferent projections to the four divisions of the mPFC in the rat: the medial (frontal) agranular (AGm), anterior cingulate (AC), prelimbic (PL), and infralimbic (IL) cortices. Each division of the mPFC receives a unique set of afferent projections. There is a shift dorsoventrally along the mPFC from predominantly sensorimotor input to the dorsal mPFC (AGm and dorsal AC) to primarily 'limbic' input to the ventral mPFC (PL and IL). The AGm and dorsal AC receive afferent projections from widespread areas of the cortex (and associated thalamic nuclei) representing all sensory modalities. This information is presumably integrated at, and utilized by, the dorsal mPFC in goal directed actions. In contrast with the dorsal mPFC, the ventral mPFC receives significantly less cortical input overall and afferents from limbic as opposed to sensorimotor regions of cortex. The main sources of afferent projections to PL/IL are from the orbitomedial prefrontal, agranular insular, perirhinal and entorhinal cortices, the hippocampus, the claustrum, the medial basal forebrain, the basal nuclei of amygdala, the midline thalamus and monoaminergic nuclei of the brainstem. With a few exceptions, there are few projections from the hypothalamus to the dorsal or ventral mPFC. Accordingly, subcortical limbic information mainly reaches the mPFC via the midline thalamus and basal nuclei of amygdala. As discussed herein, based on patterns of afferent (as well as efferent) projections, PL is positioned to serve a direct role in cognitive functions homologous to dorsolateral PFC of primates, whereas IL appears to represent a visceromotor center homologous to the orbitomedial PFC of primates.

The nucleus reuniens (RE) is the largest of the midline nuclei of the thalamus and exerts strong excitatory actions on the hippocampus and medial prefrontal cortex. Although RE projections to the hippocampus have been well documented, no study using modern tracers has examined the totality of RE projections. With the anterograde anatomical tracer Phaseolus vulgaris leuccoagglutinin, we examined the efferent projections of RE as well as those of the rhomboid nucleus (RH) located dorsal to RE. Control injections were made in the central medial nucleus (CEM) of the thalamus. We showed that the output of RE is almost entirely directed to the hippocampus and "limbic" cortical structures. Specifically, RE projects strongly to the medial frontal polar, anterior piriform, medial and ventral orbital, anterior cingulate, prelimbic, infralimbic, insular, perirhinal, and entorhinal cortices as well as to CA1, dorsal and ventral subiculum, and parasubiculum of the hippocampus. RH distributes more widely than RE, that is, to several RE targets but also significantly to regions of motor, somatosensory, posterior parietal, retrosplenial, temporal, and occipital cortices; to nucleus accumbens; and to the basolateral nucleus of amygdala. The ventral midline thalamus is positioned to exert significant control over fairly widespread regions of the cortex (limbic, sensory, motor), hippocampus, dorsal and ventral striatum, and basal nuclei of the amygdala, possibly to coordinate limbic and sensorimotor functions. We suggest that RE/RH may represent an important conduit in the exchange of information between subcortical-cortical and cortical-cortical limbic structures potentially involved in the selection of appropriate responses to specific and changing sets of environmental conditions.

The paraventricular (PV) and paratenial (PT) nuclei are prominent cell groups of the midline thalamus. To our knowledge, only a single early report has examined PV projections and no previous study has comprehensively analyzed PT projections. By using the anterograde anatomical tracer, Phaseolus vulgaris leucoagglutinin, and the retrograde tracer, FluoroGold, we examined the efferent projections of PV and PT. We showed that the output of PV is virtually directed to a discrete set of limbic forebrain structures, including 'limbic' regions of the cortex. These include the infralimbic, prelimbic, dorsal agranular insular, and entorhinal cortices, the ventral subiculum of the hippocampus, dorsal tenia tecta, claustrum, lateral septum, dorsal striatum, nucleus accumbens (core and shell), olfactory tubercle, bed nucleus of stria terminalis (BST), medial, central, cortical, and basal nuclei of amygdala, and the suprachiasmatic, arcuate, and dorsomedial nuclei of the hypothalamus. The posterior PV distributes more heavily than the anterior PV to the dorsal striatum and to the central and basal nuclei of amygdala. PT projections significantly overlap with those of PV, with some important differences. PT distributes less heavily than PV to BST and to the amygdala, but much more densely to the medial prefrontal and entorhinal cortices and to the ventral subiculum of hippocampus. As described herein, PV/PT receive a vast array of afferents from the brainstem, hypothalamus, and limbic forebrain, related to arousal and attentive states of the animal, and would appear to channel that information to structures of the limbic forebrain in the selection of appropriate responses to changing environmental conditions. Depending on the specific complement of emotionally associated information reaching PV/PT at any one time, PV/PT would appear positioned, by actions on the limbic forebrain, to direct behavior toward a particular outcome over a range of outcomes.

The thalamus was subdivided into three major groups: sensorimotor nuclei (or principal/relay nuclei), limbic nuclei and nuclei bridging these two domains. Limbic nuclei of thalamus (or ‘limbic thalamus’) consist of the anterior nuclei, midline nuclei, medial division of the mediodorsal nucleus (MDm) and central medial nucleus (CM) of the intralaminar complex. The midline nuclei include the paraventricular (PV) and paratenial (PT) nuclei, dorsally, and the reuniens (RE) and rhomboid (RH) nuclei, ventrally. The ‘limbic’ thalamic nuclei predominantly connect with limbic-related structures and serve a direct role in limbic–associated functions. Regarding the midline nuclei, RE/RH mainly target limbic cortical structures, particularly the hippocampus and the medial prefrontal cortex. Accordingly, RE/RH participate in functions involving interactions of the HF and mPFC. By contrast, PV/PT mainly project to limbic subcortical structures, particularly the amygdala and nucleus accumbens, and hence are critically involved in affective behaviors such as stress/anxiety, feeding behavior, and drug seeking activities. The anatomical/functional characteristics of MDm and CM are very similar to those of the midline nuclei and hence the collection of nuclei extending dorsoventrally along the midline/paramidline of the thalamus constitute the core of the ‘limbic thalamus’.

The medial prefrontal cortex and the hippocampus serve well recognized roles in memory processing. The hippocampus projects densely to, and exerts strong excitatory actions on, the medial prefrontal cortex. Interestingly, the medial prefrontal cortex, in rats and other species, has no direct return projections to the hippocampus, and few projections to parahippocampal structures including the entorhinal cortex. It is well established that the nucleus reuniens of the midline thalamus is the major source of thalamic afferents to the hippocampus. Since the medial prefrontal cortex also distributes to nucleus reuniens, we examined medial prefrontal connections with populations of nucleus reuniens neurons projecting to hippocampus. We used a combined anterograde and retrograde tracing procedure at the light and electron microscopic levels. Specifically, we made Phaseolus vulgaris-leuccoagglutinin (PHA-L) injections into the medial prefrontal cortex and Fluorogold injections into the hippocampus (CA1/subiculum) and examined termination patterns of anterogradely PHA-L labeled fibers on retrogradely FG labeled cells of nucleus reuniens. At the light microscopic level, we showed that fibers from the medial prefrontal cortex form multiple putative synaptic contacts with dendrites of hippocampally projecting neurons throughout the extent of nucleus reuniens. At ultrastructural level, we showed that medial prefrontal cortical fibers form asymmetric contacts predominantly with dendritic shafts of hippocampally projecting reuniens cells. These findings indicate that nucleus reuniens represents a critical link between the medial prefrontal cortex and the hippocampus. We discuss the possibility that nucleus reuniens gates the flow of information between the medial prefrontal cortex and hippocampus dependent upon attentive/arousal states of the organism.

The medial orbital (MO) and ventral orbital (VO) cortices are prominent divisions of the orbitomedial prefrontal cortex. To our knowledge, no previous report in the rat has comprehensively described the projections of MO and VO. By using the anterograde tracer Phaseolus vulgaris leucoagglutinin and the retrograde tracer Fluoro-Gold, we examined the efferent projections of MO and VO in the rat. Although MO and VO projections overlap, MO distributes more widely throughout the brain, particularly to limbic structures, than does VO. The main cortical targets of MO were the orbital, ventral medial prefrontal (mPFC), agranular insular, piriform, retrosplenial, and parahippocampal cortices. The main subcortical targets of MO were the medial striatum, olfactory tubercle, claustrum, nucleus accumbens, septum, substantia innominata, lateral preoptic area, and diagonal band nuclei of the basal forebrain; central, medial, cortical, and basal nuclei of amygdala; paratenial, mediodorsal, and reuniens nuclei of the thalamus; posterior, supramammillary, and lateral nuclei of the hypothalamus; and periaqueductal gray, ventral tegmental area, substantia nigra, dorsal and median raphe, laterodorsal tegmental, and incertus nuclei of the brainstem. By comparison, VO distributes to some of these same sites, notably to the striatum, but lacks projections to parts of limbic cortex, to nucleus accumbens, and to the amygdala. VO distributes much more strongly, however, than MO to the medial (frontal) agranular, anterior cingulate, sensorimotor, posterior parietal, lateral agranular retrosplenial, and temporal association cortices. The patterns of MO projections are similar to those of the mPFC, whereas the projections of VO overlap with those of the ventrolateral orbital cortex (VLO). This suggests that MO serves functions comparable to those of the mPFC, such as goal-directed behavior, and VO performs functions similar to VLO such as directed attention. MO/VO may also serve as a link between lateral orbital and medial prefrontal cortices.

The nucleus reuniens (RE) of the midline thalamus has been shown to strongly innervate structures of the limbic forebrain, prominently including the hippocampus (HF) and the medial prefrontal cortex (mPFC) and to exert pronounced excitatory effects on HF and mPFC. It was unknown, however, whether RE projections to, and hence actions on, the HF and mPFC originate from a common or largely separate groups of RE neurons. Using fluorescent retrograde tracing techniques, we examined the patterns of distribution of RE cells projecting to HF, to the mPFC or to both sites via axon collaterals. Specifically, injections of the retrograde tracers Fluorogold (FG) or Fluororuby (FR) were made in the mPFC and in various subfields of HF and patterns of single (FG or FR) or double labeled (FG + FR) cells in RE were determined. Pronounced numbers of (single) labeled neurons were present throughout RE with FG or FR injections, and although intermingled in RE, cells projecting to the mPFC were preferentially distributed along the midline or in the perireuniens nucleus (pRE), whereas those projecting to HF occupied a wide mediolateral cross sectional area of RE lying between cells projecting to the mPFC. Approximately, tenfold more labeled cells were present in RE with ventral compared to dorsal CA1 injections. Like single labeled neurons, double labeled cells were found throughout RE, but were most densely concentrated in areas of greatest overlap of FG+ and FR+ neurons or mainly in the lateral one-third of RE, medial to pRE. Depending on specific combinations of injections, double labeled cells ranged from approximately 3-9% of the labeled neurons. The nucleus reuniens has been shown to be a vital link in limbic subcortical-cortical communication and recent evidence indicates a direct RE involvement in hippocampal and medial prefrontal cortical-dependent behaviors. The present findings indicate that RE is critically positioned to influence the HF and mPFC, and their associated behaviors, via separate or collateral projections to these sites.

The theta rhythm is the largest extracellular synchronous signal that can be recorded from the mammalian brain and has been strongly implicated in mnemonic processes of the hippocampus. We describe (a) ascending brain stem-forebrain systems involved in controlling theta and nontheta (desynchronization) states of the hippocampal electroencephalogram; (b) theta rhythmically discharging cells in several structures of Papez's circuit and their possible functional significance, specifically with respect to head direction cells in this same circuit; and (c) the role of nucleus reuniens of the thalamus as a major interface between the medial prefrontal cortex and hippocampus and as a prominent source of afferent limbic information to the hippocampus. We suggest that the hippocampus receives two main types of input: theta rhythm from ascending brain stem- diencephaloseptal systems and information bearing mainly from thalamocortical/cortical systems. The temporal convergence of activity of these two systems results in the encoding of information in the hippocampus, primarily reaching it from the entorhinal cortex and nucleus reuniens.