Disentangling the identity of the zona incerta: a review of the known connections and latest implications

The zona incerta (ZI) is a subthalamic region composed by loosely packed neurochemically mixed neurons, juxtaposed to the main ascending and descending bundles. The extreme neurochemical diversity that characterizes this area, together with the diffuseness of its connections with the entire neuraxis and its hard-to-reach positioning in the brain caused the ZI to keep its halo of mystery for over a century. However, in the last decades, a rich albeit fragmentary body of knowledge regarding both the incertal anatomical connections and functional implications has been built mostly based on rodent studies and its lack of cohesion makes difficult to depict an integrated, exhaustive picture regarding the ZI and its roles. This review aims to provide a unified resource that summarizes the current knowledge regarding the anatomical profile of interactions of the ZI in rodents and non-human primates and the functional significance of its connections, highlighting the aspects still unbeknown to research. Abbreviations


Introduction
The zona incerta (ZI) is a subthalamic structure conserved among mammalians and in reptiles and avians putative counterparts of the ZI have been identified (Pritz, 2022).Since the last decades of the past century, research has focused on defining the function of this structure and has revealed the extreme multifaceted roles the ZI plays in several functional domains, including the cardiovascular and sexual cycle regulation, ingestion and locomotion control, arousal and attention (for a review, see Mitrofanis, 2005).However, the interest over this subthalamic area shows no signs of ceasing: new and appealing insights are emerging by recent studies which support a role of the ZI also in emotional states regulation (Venkataraman et al., 2019;Li et al., 2021), novelty seeking and investigation (Monosov et al., 2022;Ogasawara et al., 2022) and neuroplastic processes (Wang et al., 2020b).Nonetheless, clinical studies have provided promising perspectives over the ZI as a therapeutic target for treating the motor symptoms of the Parkinson's Disease, showing the Deep Brain Stimulation (DBS) applied on the ZI to be more effective than when applied on a classical therapeutic target as the subthalamic nucleus (Plaha et al., 2006); interestingly, a recent study has also raised the possibility that the DBS applied on the ZI may result effective on the symptoms of the Obsessive-Compulsive Disorder (Haber et al., 2023).
Beside the conspicuous richness of functions it is involved in, the ZI also shows an outstanding heterogeneity that spans among several levels, regarding its histology, neurochemistry and connections that cover approximately the entire brain (Mitrofanis and Mikuletic, 1999;Mitrofanis, 2005), allowing the ZI to influence the activity of a multitude of centers (Londei et al., in press) involved in as many processes.However, a clear and unified picture of both the functions and connections of this intriguing area is still lacking.
The present review aims at collecting and summarizing, in graphical form and with standardized nomenclature, the current knowledge on incertal connectivity offered by anatomical tracing studies on rodents and to merge the anatomical data with the latest insights on functional aspects to provide an accessible, complete, and updated tool for future studies on the ZI.Although the focus of the review is on the rodents, we will also consider a summary of the anatomy of the ZI in non-human primates.

Cytoarchitecture and chemoarchitecture in rodents
From a histological perspective, the ZI encompasses several cytoarchitectonic sectors classified on the basis of the morphology, size and staining of the local neurons which originally allowed the identification of six sectors (Kawana and Watanabe, 1981): pars rostro-polaris, pars dorsalis, pars ventralis, pars caudalis, pars magnocellularis and pars retro-polaris.However, a four-sector cytoarchitectonic classification has become predominant in the last years.This system encompasses the rostral (ZIr), dorsal (ZId), ventral (ZIv), and caudal (ZIc) sector (Kim et al., 1992;Nicolelis et al., 1992;Kolmac and Mitrofanis, 1999;Mitrofanis, 2005), with the latter including the pars caudalis, pars magnocellularis, and pars retro-polaris by Kawana and Watanabe (1981) (Nicolelis et al., 1995).Fig. 1 summarizes the main cytoarchitectonical and neurochemical features of the four sectors.Concerning the neuronal morphology and size, the ZIr is characterized by small-sized, fusiform-or round-shaped cells, with sparse medium-sized neurons (Kawana and Watanabe, 1981).Similarly, the ZId presents neurons ranging from small to medium in size, with round, fusiform but also multipolar morphology.A cell-free fibrous lamina separates the ZId from the ZIv, populated by medium-sized multipolar and fusiform cells (Kawana and Watanabe, 1981;Trageser et al., 2006).A higher cell density, appearing by the end of the third postnatal week, contributes to differentiate ZIv from ZId (Nicolelis et al., 1995).A rich diversity characterizes the cytoarchitecture of the ZIc, with fusiform, rounded, polygonal, and multipolar cells of different sizes, ranging from small to large (Kawana and Watanabe, 1981).However, two additional criteria are adopted to distinguish each sector: the neurochemical profile and Fig. 1.Chemoarchitecture and cytoarchitecture of the zona incerta.Green boxes indicate the reported size and morphology of cells and red boxes indicate the cell type based on immunohistochemistry, relatively to the incertal sector.To be noted, red boxes are presented with a color darkness gradient which reflects the relative richness of different cell types within the different incertal sectors (based on Kawana and Watanabe, 1981;Kim et al., 1992;Kolmac and Mitrofanis, 1999;Watson et al., 2014).
the cytochrome oxidase (CO) histochemistry.Concerning the first, some cell types are evenly distributed among sectors whereas some others clearly delineate the borders between sectors.Specifically, the higher concentration of somatostatin and tyrosine hydroxylase positive neurons defines the ZIr; similarly, the higher concentration of nicotinamide adenine dinucleotide phosphate (NADPH)-diaphorase and glutamate positive neurons characterizes the ZId (Kolmac and Mitrofanis, 1999), while the ventral sector is rich in GABA and GAD-immunostained cells, which are more numerous and more intensely labeled compared to the other sectors (Kim et al., 1992;Kolmac and Mitrofanis, 1999).The ZIv is also characterized by a more intense CO staining compared to the surrounding sectors and, interestingly, this feature characterizes the ZIv since birth (Kim et al., 1992;Nicolelis et al., 1992;Nicolelis et al., 1995).Albeit less sharply defined compared to the other sectors, some features contribute to identify the ZIc: it appears to be more richly populated by calbindin-positive cells and more extensively marked by AchE neuropilar staining compared to ZIr, ZIdl, and ZIv, and it results negative to SMI-32 immunostaining, whilst ZIv and ZId result positive (Watson et al., 2014).Finally, few serotonin-positive cells were equally distributed in the four incertal sectors (Kolmac and Mitrofanis, 1999).

Thalamus
The zona incerta (ZI) interacts with several diencephalic structures, as shown by Fig. 2 (and Table 1 for references), among which the thalamus.Higher order thalamic nuclei, that is nuclei whose activity is driven by cortical inputs (Guillery, 1995) and encompass the associative and intralaminar nuclei, are more densely connected with the ZI compared to the primary order nuclei, which receive their driving inputs from lower cerebral centers (Guillery, 1995;Power et al., 1999).Furthermore, the incertal cells originating projections to the associative and intralaminar thalamic centers seem to distribute discretely in respect to the borders of, respectively, the ZIv and ZId, while projections targeting the primary order nuclei arise from cells distributed among every sector with no major concentration in any of them (Power et al., 1999).
A detailed description of both anatomical features and possible functional roles is available regarding the relationship between the ZI and the higher order nuclei, particularly the posterior complex of the thalamus (PO).Incerto-thalamic fibers, arising mainly from the ventral sector of the nucleus (Lavallée et al., 2005;Trageser et al., 2006;Watson et al., 2015), target the shaft of the dendrites (rarely the soma), often in proximity to principal afferents, of the non-GABAergic thalamic cells forming large, mostly symmetric, synaptic contacts and form occasionally punctum adherens-like structure also (Barthó et al., 2002).Albeit the incertal projections to the PO are generally considered to be GABAergic (Barthó et al., 2002;Lavallée et al., 2005;Trageser et al., 2006;Watson et al., 2015), a study reports the existence of incerto-thalamic projections terminating with asymmetric synapses, structures characterized by a thick postsynaptic density and considered to be typical of excitatory contacts (Power and Mitrofanis, 2002;Romaus-Sanjurjo et al., 2021).From the functional perspective, several studies have investigated the role of the ZI in modulating the "permeability" of the PO to sensory subcortical inputs, collaborating to the regulation of the parameters of its responses to whisker stimulation (Trageser and Keller, 2004;Lavallée et al., 2005;Trageser et al., 2006).In fact, it has been shown that incertal lesions disrupting the tonic inhibition that the ZI exerts on the PO allow the latter to respond to whisker deflections with latencies, duration and magnitude coherent with a primary-order nature of this nucleus, whereas in absence of lesions to ZI, PO's responses kinetics were more similar to what expected for a high-order nucleus (Trageser and Keller, 2004).Subsequent studies have highlighted that incertal control over the PO ability to respond to sensory stimuli is modulated by the cholinergic innervation (Masri et al., 2006;Trageser et al., 2006).The stimulation of the brainstem activating system (lateral dorsal tegmental nucleus and the pedunculopontine nucleus, PPN), which mimics a state of wakefulness and alertness, causes the release of acetylcholine on the ZI, resulting in the suppression of its activity and eventually disinhibiting the PO.The disruption of the incertothalamic inhibition lets the ascending sensory inputs to drive the activity of the PO, allowing it to function as a primary order nucleus.When the brainstem activating system is not active, and the acetylcholine is not released on the ZI, mimicking a state of sleep or anesthesia, the incertal inhibition is exerted on the PO that acts as a higher order nucleus whose activity is driven not by sensory inputs coming from subcortical structures but by cortical inputs (Trageser et al., 2006).
Numerous studies have lately investigated the involvement of the ZI-PO pathway in the pain perception, especially considering pathological states.It has been shown that spontaneous pain following a spinal lesion, a condition often referred to as central pain syndrome, is associated with the reduced activation of incertal neurons (Masri et al., 2009;Moon and Park, 2017) and subsequently by an hyperactivation of the PO neurons (Masri et al., 2009).This circuit expresses the cannabinoid type-1 receptor (CB1-R), which makes it susceptible to the regulation operated by endogenous cannabinoids, whose modulation influences both the electrical activity of PO neurons and the pain thresholds (Wang et al., 2020a).Interestingly, pathological conditions of neuropathic pain are associated with a downregulation of those receptors (Wang et al., 2020a).
The ZI sends also dopaminergic and GABAergic projections to the nucleus reuniens (RE) (Barthó et al., 2002;McKenna and Vertes, 2004;Venkataraman et al., 2021;Lin et al., 2023), with terminals smaller than those entering other nuclei of the dorsal thalamus (Barthó et al., 2002).On the functional side, recent studies have shown the involvement of the ZI-RE projections in fear-related behaviors: the optogenetical stimulation of this ZI-RE GABAergic pathway reduces fear generalization and enhances the recall of the extinction learning after one day from training, while the stimulation of the dopaminergic pathway only has effect on the extinction recall and not on the fear generalization (Venkataraman et al., 2021).Additionally, the optogenetic manipulation of the incertal projections from somatostatin-positive GABAergic cells to the RE induce defensive-like behaviors in different experimental contexts: the activation of such projections induces freezing and bradycardia in an open field context, while their inactivation induces an increase in the latency to return to the shelter and a reduction in the time spent hiding when presented with an overhead looming stimulus (Lin et al., 2023), suggesting a role of this pathway in the modulation of fear-related behaviors.
Furthermore, a pathway from the ZI to the paraventricular thalamus (PVT) has been identified by several studies (Zhang and Van Den Pol, 2017;Zhao et al., 2019;Ye et al., 2022;Wu et al., 2023a).From a neurochemical and functional perspective, GABAergic neurons of the ZI target glutamatergic neurons of the PVT (Zhang and Van Den Pol, 2017), and, moreover, these incertal terminals can be regulated by serotonin release from raphe nucleus which inhibits the release of GABA (Ye and Zhang, 2021).These neurochemical mechanisms have been involved in the regulation of feeding behavior: the stimulation of incertal terminals targeting the PVT evoked food-foraging behavior and increased binge-eating-like food intake, feeding time and evoked a preference for high-fat food (Zhang and Van Den Pol, 2017).A recent study also shed light on the involvement of ZI-PVT GABAergic pathway in the regulation of nociceptive behavior, since the chemogenetic manipulation of this pathway leads to the modulation of the pain threshold and its activation reduces hypersensitivity in a neuropathic pain mouse model (Wu et al., 2023a).
The ZI also sends axons that target the thalamic reticular nucleus (RT).Quantitatively, axons departing from the ZIr constitute a denser projection compared to the one originated by axons arising from cells in the ZIc, while projection from the ZIv is denser than the one from the ZId but sparser than the one from the ZIr (Çavdar et al., 2006).between the ZI and the other area considered.Each external area is placed in larger boxes indicating its anatomical allocation on the basis of the classification reported by the Allen Mouse Brain Atlas.Asterisks refer to areas labeled with different nomenclatures in respect to the ones reported by the Allen Brain Atlas, either for brevity (for example, "RVM" includes multiple areas recognized by the Allen Brain Atlas) or for complications related to the use in literature of several nonoverlapping nomenclature systems ("Temporal cortex" and "VISs").For more details, please refer to "Cortical structures" and "Hindbrain" sections.
Labeled terminal-like structures were also observed in several other thalamic nuclei, among which the lateral posterior (LP), lateral dorsal (LD), ventral medial (VM), paracentral (PCN) central lateral (CL) and central medial (CM), after anterograde tracer injection in the ZI (Barthó et al., 2002).Injections of tracer in several thalamic nuclei revealed retrograde-labeled neurons in ZI, confirming the results obtained by anterograde tracing for CL, LP and LD relationships with the ZI (Power et al., 1999).Furthermore, anterograde tracing was also observed, suggesting that the injected thalamic nuclei (parafascicular, PF, and PO besides the aforementioned ones) also reciprocate the projections they receive from the ZI (Power et al., 1999).
Although much evidence has been collected on the roles and characteristics of incertofugal projections targeting the thalamus, and more specifically the PO, the RE and the PVT, afferent projection to the ZI from the thalamus have been poorly characterized.
An exception is the PF, which projects its axons to the ZI (Park et al., 2022) and also receives afferents from the incertal dorsal sector that form asymmetric synapses with the thalamic neurons (Power et al., 1999;Power and Mitrofanis, 2002).Interestingly, decreases in activity of the PF-ZI pathway have been correlated with the initiation of specific movements, such as turn left or right, executed in an open field paradigm.Additionally, different increases in activity of this pathway have been correlated with vertical or horizontal sniffing, thus encoding the dimensionality of the movement (Park et al., 2022).

Cortical structures
The relationship between the ZI and the cortical structures has been in the spotlight of recent research that attempted to qualify such interconnectivity also in terms of the functional implications, that, however, remain elusive up to date.From the anatomical perspective, in adult rats, the incertal input has been shown to reach the primary somatosensory cortex (SSp) as well as other cortical areas, such as the primary visual cortex, the motor cortex, and auditory/temporal cortex (Lin et al., 1990;Schroeder et al., 2023), with the relative density of each projection being somewhat debated (Lin et al., 1990;Power et al., 2001;Schroeder et al., 2023).
During the early postnatal phase, the incertocortical projection seems to be larger compared to the one seen in the adult phase, suggesting a crucial role for the ZI in the early stages of cortical development (Lin et al., 1990).In line with this hypothesis are the studies showing that, during the first postnatal weeks, incertal cells are already well developed, active and form a plexus of GABAergic fibers that branch in the layer 1 of the somatosensory and motor cortex, exerting an excitatory effect on the apical dendrites of the pyramidal cells that is involved in controlling plastic processes such as the synaptogenesis, spines development and dendrites branching (Dammerman et al., 2000;Chen and Kriegstein, 2015).
Concerning the allocortex, several studies employing retrograde tracers injected in the hippocampal formation have evidenced labeled cells in the ZI, suggesting the existence of an incertal projection to these allocortical regions which also resulted to be α-melanocyte-stimulatinghormone (αMSH)-positive (Riley and Moore, 1981;Köhler et al., 1984).Corticoincertal back projections have been observed in several studies in which neural tracers were injected in various cortical areas or in the ZI.As in these studies different cortical nomenclatures were adopted, for the sake of comparison and clarity in the present work we refer to the Allen Mouse Brain Atlas cortical subdivision.Based on anterograde tracer injections in various cortical areas, Mitrofanis and Mikuletic (1999) reported projections to the ZI from several areas, such as the frontal (Fr1), cingulate (Cg1), forelimb (FL), parietal (Par1) and occipital (Oc1) areas.The Fr1 seems to correspond to the primary motor cortex, Cg1 to the dorsal part of the anterior cingulate area (ACAd) by Swanson (1992Swanson ( , 1998)), FL and Par1, together with the hindlimb area Table 1 References for Fig. 2. Each index links each connection of Fig. 2 with the relative reference reported on the middle column; the right column provides the technique/s employed by each study to trace projections from/to the ZI.Since each technique bears its drawbacks and/or limitations, most of them connected to false negatives deriving from differential sensitivity of different tracers and false positives deriving from the dye uptake from passing fibers instead of terminals, only studies in which the authors explicitly refer to direct projection from/to the ZI, or to terminals and targets to indicate sites of synaptic contact, are included.However, for a complete evaluation of the validity of the tracing techniques, please refer to the mentioned studies.(HL) seems to constitute the primary somatosensory area (SmI) by Zilles and Wree (1995) which in turn seems to coincide to the SSp by Swanson (1992Swanson ( , 1998)), while Oc1 may correspond to the primary visual cortex (Paxinos, 2004).They showed that the strongest projection arises from the Cg1 and targets the dorsal and rostral ZI, while projections arising from the Fr1, Par1 and FL target the ZId and the ZIv, with terminals labeled after anterograde tracer injection in Fr1 and Par1 located, respectively, more medially and more laterally compared to the ones observed after injection in FL; lastly, projections from Oc1 are sparser compared to the other corticoincertal pathways (Mitrofanis and Mikuletic, 1999).Projections from the anterior cingulate area (ACA) and from the SSp were noted also by Roger and Cadusseau (1985) and by Shammah-Lagnado et al. (1985), and the latter study also reports a projection from the primary motor cortex (MOp), area 17 and area 18, which correspond to the primary visual (VISp) and the extrastriate cortex (arbitrarily referred to as "VISs" in the present work) and from the retrosplenial cortex (RSP) which is confirmed by the retrograde tracing experiment by Mitrofanis and Mikuletic (1999) that, after injection in dorsal and caudal ZI, observed labeled cells both in its granular and agranular subdivision.It has also been shown that the projection to the ZI originates from the fifth layer (Mitrofanis and Mikuletic, 1999;Urbain and Deschênes, 2007), and that the number of labeled cells in the primary (MOp) and secondary (MOs) motor cortex is quantitatively more prominent compared to that in the somatosensory area (SS) (Urbain and Deschênes, 2007).
On the functional side, recent evidence suggests an involvement of the pathway arising from the anterior cingulate cortex and targeting the ZI in the behavioral adaptation to safe or dangerous contexts, influencing the activity of the superior colliculus (SC) and determining a possible escape response.Specifically, a disynaptic inhibitory pathway from the cingulate cortex to the SC, mediated by the ZI, arises from a neuronal population displaying preferential activation in safe conditions, and its activation suppresses escape responses.On the other hand, an excitatory direct projection from the cingulate cortex to the SC arises from a neuronal population displaying danger-related activity, and its activation contributes to enhance escape responses (Wu et al., 2023b).
As introduced in the last paragraph, alterations of the activity of the ZI are associated with neuropathic pain (Masri et al., 2009;Moon and Park, 2017).Interestingly, the hyperalgesia that characterizes this pathological condition is reduced by the motor cortex stimulation, which has been proposed to enhance the activity of the ZI neurons, and thus restoring the proper inhibition to the PO, since the stimulation of the ZI produces similar therapeutic effects (Lucas et al., 2011).
Support to this hypothesis comes from the evidence that the stimulation of the motor cortex leads to the activation of putative POprojecting incertal neurons, that is, neurons in the ventrolateral sector of the ZI, and to the inhibition of spontaneous activity of the PO neurons.This thalamic inhibition would thus prevent the flow of nociceptive information to get to the cortex (Cha et al., 2013).However, parallel mechanisms for nociceptive regulation have been reported.Based on a dissection of the MOp-ZI circuit in a spinal nerve injury model it has been demonstrated that hindlimb MOp (layer 5) projects to ZId and ZIv and that the former seems to play a key role in pain relief mechanisms (Gan et al., 2022).Indeed, MOp projections almost exclusively target non-GABAergic cells of the ZId and, through collaterals, also the periaqueductal grey (PAG), which both project to the locus coeruleus (LC) and to the rostral ventromedial medulla (RVM), structures involved in the descending modulation of the nociceptive transmission (Gan et al., 2022).Another point discussed in the previous paragraph was the role of the ZI in regulating the flow of sensorial information through the PO to the cortex, which has been hypothesized to rely on the activation of the arousal-linked cholinergic nuclei.However, an alternative hypothesis has been proposed, which emphasizes the role of the descending corticoincertal projections in the regulation of PO ability to respond to peripheral stimuli.This hypothesis is supported by the observation that the stimulation of the motor cortex drives the suppression of the responses to whisker deflection of the vibrissae-sensitive incertal neurons, populating the ZIv, through an intra-incertal network of GABAergic axon collaterals (Urbain and Deschênes, 2007).Therefore, since the ZIv is the main source of projections to the PO, it is possible that during palpation, the object detection-oriented active whisking, the motor cortex input to the ZI causes a window of disinhibition in PO that allows the sensory information collected by vibrissae stimulation to flow and to get to the cortex (Lavallée et al., 2005;Urbain and Deschênes, 2007).
The discrepancy of the effects produced by the stimulation of the motor cortex in Cha et al. (2013) and Urbain and Deschênes (2007), which reported respectively activation and suppression of ZI cells, may be accounted for by the different duration of the cortical stimulation applied in the two studies, with a prolonged stimulation, applied by Cha et al. (2013), that may lead to the depletion of GABAergic reserves of incertal cells receiving a direct innervation from the motor cortex (Cha et al., 2013).
Projections from the somatosensory areas originate consistently from the SSp (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Nicolelis et al., 1992;Wang et al., 2019), and from the supplemental somatosensory cortex (SSs) (Zhou et al., 2018) and have been involved in flight behavior.Specifically, by processing multisensory stimuli, ZI mediates the facilitatory effect of somatosensory stimuli on a flight behavior triggered by an auditory stimulus, through its connections with the PO (Wang et al., 2019).
Furthermore, sporadic labeling was found in several cortical regions by a study injecting retrograde tracers into the ZI; these regions include the prelimbic (PL) cortex and the infralimbic (ILA) area (Shammah--Lagnado et al., 1985).A recent study has found a role of PL-ZI projections in the processing of motivational value.Such information on the motivation, when received and elaborated by the ZI, is used to drive the investigatory behavior with the contribution of the PAG, which receives direct projections from ZI cells recipient of prelimbic projections (Ahmadlou et al., 2021).Interestingly, it has been recently discovered an implication of the dmPFC-ZI pathway in fear generalization.Specifically, the modulation of the excitatory projections from the dmPFC to the ZI influences the activity of incertal dopaminergic cells and affects the ability to discriminate contexts and to display appropriate fear responses (Tong et al., 2023).Finally, in rodents several studies based on anterograde and retrograde tracers injections at the cortical level and in the ZI, respectively, have described projections from the temporal cortex to the ZI, specifically from the secondary auditory cortex (addressed as Te2 and Te3 in Mitrofanis, 2002b, andas Au2 in Zhao et al., 2019), which corresponds to the belt region located around the primary auditory area (AUDp), in which, instead, only few retrogradely labeled cells have been observed after retrograde tracer injections in ZI (area Te1 in Shammah-Lagnado et al., 1985).

Basal ganglia and related nuclei
The ZI exchanges extensive projections with several mesencephalic structures.One of the most intriguing interactions regards the substantia nigra, particularly its pars reticulata (SNr), rather than its pars compacta (SNc), which among the basal ganglia is a major source of inputs and target of outputs of the ZI (Shammah-Lagnado et al., 1985;Kolmac et al., 1998;Heise and Mitrofanis, 2004).Specifically, the study by Heise and Mitrofanis (2004) has shown labeling of both cells and terminals in the two parts of the SN after injection of a retrograde and anterograde tracer in the ZI.However, more lateral injections of neural tracer in the ZI have resulted in a scarcer labeling in the SNc compared to the labeling obtained in SNr, which resembled the one obtained by medial injections.Accordingly, while projections from and to the SNr may involve the entire ZI, connections with the SNc may involve mostly the medial ZI (Heise and Mitrofanis, 2004), which, interestingly, is the incertal region thought to take part in motor preparation and execution (Yang et al., 2022).These results are also in agreement with the study by Shammah-Lagnado et al. (1985) that found only scarce and occasional labeling of the SNc when the ZI was injected with a retrograde tracer.
When considering the interaction between the ZI and the SN, it is worth mentioning the clinical interest linked to the involvement of this subthalamic zone in Parkinson's disease (PD), a degenerative disorder marked by events of neuronal death in the SNc and the subsequent nigrostriatal denervation that, besides causing a cascading alteration of activity in several brain centers, leads to movement and coordination impairments.This condition is also associated with the hyperactivity of ZI neurons (Périer et al., 2000;Li et al., 2022) which can be explained either by an enhanced excitatory input or by a loss of inhibitory control over their activity.Notwithstanding no conclusive explanation is currently available, several, probably coexisting, hypotheses have been proposed.In particular, it has been hypothesized that the SC, whose activity is reduced in PD-like conditions, may provide a weaker inhibitory input to the ZI compared to control conditions and thus may favor the hyperactivity of incertal cells.On the other hand, the PPN, whose activity is enhanced after nigrostriatal denervation, may provide a stronger excitatory drive to incertal cells, enhancing their activity (Périer et al., 2000;Merello et al., 2006).
However, the study by Li et al. (2022) has outlined that the photo-stimulation of the SNr-projecting glutamatergic neurons at 30 Hz, a frequency that coincides with the mean firing rates of incertal neurons in PD patients, induces coordination impairments, akinesia and bradykinesia and thus suggests a causal relationship between ZI alterations and the development of classical parkinsonian symptoms.
On the clinical side, Plaha et al. (2006) have shown that the Deep Brain Stimulation applied on the ZI is more efficient in ameliorating the motor symptoms of the PD compared to its application on a classical therapeutic target as the subthalamic nucleus.Ossowska proposed that one of the reasons at the basis of the DBS' therapeutic efficacy may be a possible inhibitory effect on the ZI that would restore the normal flow of information through the basal ganglia circuits (Ossowska, 2020).Such hypothesis is coherent with recent experimental data showing that the inhibition of the glutamatergic neurons of the caudal ZI ameliorates the motor symptoms displayed by a PD mouse model (Li et al., 2022).Accordingly, since the stimulation of incertal GABAergic neurons leads to the amelioration of parkinsonian motor symptoms too, it has been hypothesized that such manipulation could restore the proper inhibitory control of local GABAergic neurons over glutamatergic ones (Chen et al., 2023).
Very few information is available regarding the relationship between the dorsal pallidal structures and the ZI: it shares sparse bidirectional projections with both the internal and external part of the globus pallidus (GPi and GPe) (Shammah-Lagnado et al., 1985;Heise and Mitrofanis, 2004), corresponding respectively to the entopeduncular nucleus and the globus pallidus (GP), with the efferent incertal projections being predominantly glutamatergic (Heise and Mitrofanis, 2004).
The PPN, specifically the pars dissipata, shares one the densest connectivity, among the basal ganglia-related nuclei, with the ZI (Heise and Mitrofanis, 2004).Such projections arise from all the sectors with a preference for the dorsal one and have a mixed glutamatergic and GABAergic nature, with the latter prevailing (Heise and Mitrofanis, 2004).Interestingly, the ZI-PPN GABAergic pathway has been involved in the control of the expression of signaled active avoidance, since its stimulation leads to the suppression of such responses (Hormigo et al., 2020).

Superior colliculus
The motor-related deep and intermediate layers of the superior colliculus (SCm), which are one of the main targets of the basal ganglia outputs, receive dense projections from a small field within the ZIv.These projections distribute following a topographical organization, in which the medial ZIv projects laterally in the SCm, the lateral ZIv projects medially (Kim et al., 1992;Kolmac et al., 1998;Comoli et al., 2012), the rostral ZIv projects rostrally and the caudal ZIv projects caudally (Kim et al., 1992).The reciprocal connection was described by Kolmac et al. (1998) as non-topographical from the deep layers of the SC to the ZId while Kim et al. (1992) and Watson et al. (2015) reported not only that such pathway is topographically arranged when considering the ZIv, with the medial SC projecting to the lateral part of the ventral ZI (Watson et al., 2015), but it also arises from the intermediate gray layer (Kim et al., 1992).Contrarily, very sparse or no projections have been identified between the ZI and the sensory-related superior colliculus (SCs), involving mainly the lateral field of the ZI (Roger and Cadusseau, 1985;Power et al., 2001).From a neurochemical perspective, the SCm provides the subthalamic area with a glutamatergic innervation and targets the GABAergic neurons of the ZI (Shang et al., 2019).However, a population of GABAergic neurons in the SC that receives afferents from the cortex and projects to the ZI has also been reported (Martín-Cortecero et al., 2023).Additionally, the ZI provides the SCm with a GABAergic and dopaminergic innervation (Kim et al., 1992;Bolton et al., 2015;Liu et al., 2022), exerting a depressive effect on the activity of this nucleus (Bolton et al., 2015).Given the highly integrative nature of the ZI that collects both interoceptive and exteroceptive information from multiple centers, the inhibition provided by the ZI to the SCm may be involved in the focalization of burst of activity of the collicular neurons, resulting in the selection of the right orientating action towards a specific spatial target according to the internal state of the animal (May and Basso, 2018).Furthermore, it has been suggested that the SCm-ZI pathway is involved in the delivery of sensory information that drives hunting (Shang et al., 2019).

Periaqueductal grey
Recently, great consideration has been posed on the relationship between the ZI and the PAG for its possible role in controlling internal states.Incertal efferent connections to the PAG arise from multiple subpopulations of the ZIr, including somatostatin-positive (SOM+) and calretinin-positive (CR+) GABAergic neurons and Vglut2-positive (Vglut2 +) glutamatergic neurons.Specifically, CR+ neurons project to dmPAG, Vglut+ neurons project to l/vlPAG, while SOM+ neurons project to all the three periaqueductal fields (Li et al., 2021).Each of these populations also has a specific functional role in controlling anxiety.For instance, the activation of SOM+ neurons reduces the exploration of the open arms of the elevated plus maze, which would suggest an increase in the anxiety levels; on the contrary, both the activation of CR+ neurons and the inhibition of Vglut+ neurons induce anxiolysis (Li et al., 2021).Additionally, Chou et al. (2018) have shown that the GABAergic projections from the ZIr to the excitatory neurons of the vl/dl PAG have a role in the modulation of the amplitude of both innate and conditioned flight response, since the stimulation of such fibers leads to suppressive control over the downstream PAG neurons whose activity correlates with the magnitude of the defensive behavior shown by the animal (Chou et al., 2018).Other studies have focused on the role of the ZI-PAG pathway in hunting.Zhao and colleagues found an inhibitory pathway from the ZI to the l/vl PAG involved in hunting.This projection seems to specifically control hunting rather than feeding motivation in general, since the activation of the incertal terminals in the PAG increased the hunting success rate, the attack frequency, and decreased the attack latency, but no influence was detected on the high-fat food intake (Zhao et al., 2019).However, a more recent report suggests that inputs to the PAG from different subcortical structures are involved in a specific phase of the hunting process, with the incertal inputs to PAG being preferentially involved in the prey detection and chase phase (Yu et al., 2021).
An interesting recent development links the ZI to the novelty-seeking and exploratory behavior.Specifically, Ahmadlou et al. (2021) have shown that the activation or inhibition of ZI projections to PAG leads respectively to an increase or decrease of the time and depth of the investigation.Moreover, they have identified an incertal inhibitory neuronal subpopulation, expressing tachykinin 1, that receives afferents from the PL and sends projections to the lPAG, which seems to play a key role in exploratory behavior since its silencing results in the suppression of the investigatory behavior of an object or a conspecific.

Other mesencephalic structures
The connections between the ventral tegmental area (VTA) and the ZI are bidirectional and not topographically organized (Shammah-Lagnado et al., 1985;Kolmac et al., 1998).The possible role of these pathways has been recently related to social behavior, stress and also to feeding.Specifically, in a murine model of depression exposed to an inescapable stress, structural plasticity in both the ZI-VTA and VTA-ZI pathways was associated to the social buffering effect, that is the partial recovery from the stressful event due to the exposition to conspecifics (Cai et al., 2022).Additionally, the ZI-VTA pathway plays a role in the regulation of feeding behavior since its activation and inactivation induce modulation in food intake by acting on motivation and meal initiation (de Git et al., 2021).Regarding connections between the ZI and the midbrain reticular nucleus (MRN), bidirectional projections have been seen between these structures (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Kolmac et al., 1998;Mitrofanis, 2002a;Zhou et al., 2018), with the reticuloincertal projection being loosely topographical (Shammah-Lagnado et al., 1985).
Similarly, topography seems to be loose in the anterior pretectal nucleus (APN)-ZI projection, which is only in a minor part GABAergic and preferentially targets the ZIv (Giber et al., 2008).Finally, there is a non-topographical bidirectional connection between the red nucleus (RN) and the ZI, connecting the parvocellular lamina of the RN and the medial part of each incertal sector (Shammah-Lagnado et al., 1985;Mitrofanis, 2002a).

Hypothalamus
The ZI is continuous to the hypothalamus both from an anatomical and functional perspective (Mitrofanis, 2005).Several studies have shown the existence of projections extending bidirectionally between several fields of the hypothalamus and the ZI.The ventromedial (VMH) and lateral (LHA) hypothalamic areas have been consistently cited as the hypothalamic structures most related to the ZI by several neuroanatomic studies (Barone et al., 1981;Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985).The existence of projections from the VMH to the ZI has also been confirmed by electrophysiology studies which revealed that the stimulation of the VMH resulted in the orthodromic activation of dopaminergic neurons of the medial ZI (Eaton and Moss, 1989).Interestingly, recent studies have focused on the connectivity and functions of a specific population of incertal neurons that express the marker Lhx6.Liu et al. (2017) discovered the involvement of such population in the promotion of sleep through the projections it receives from distributed centers of the brain that control sleep/wake transitions and through the inhibition they exert on wake-promoting hypocretin and GABAergic neurons of the LHA (Liu et al., 2017).Moreover, it has been shown that such population promotes selectively REM, and not NREM, sleep, although researchers hypothesize that this effect could be mediated by the incertal influence over the pons, that directly orchestrates REM sleep (Vidal-Ortiz et al., 2023).Oh et al. (2020) reported that Lhx6-positive neurons additionally send projections to the supramammillary nucleus (SUM), which hypothetically mediate the interface between the ZI and several other regions such as the limbic areas and the claustrum (Oh et al., 2020).It is worth noting that the medial preoptic area (MPO) is a source of fibers that pass through the ZI along their path to the PPN and give off some terminals to the ZI (Swanson et al., 1987).However, no further information is available, to our knowledge, regarding the neurochemical nature and the functional significance of this pathway.

Cerebellum
A dearth of information is nowadays available regarding the relationships between the cerebellum and the ZI.Specifically, two deep cerebellar nuclei, the interposed (IP) and the dentate (DN; often referred to as "nucleus lateralis", Paxinos, 2004), are strongly connected to the ZI (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Zhou et al., 2018), with the former providing one of the most extensive projections to the ZI (Mitrofanis and De Fonseka, 2001).IP projections are predominantly contralateral, and terminals are distributed in the medial part of the four incertal sectors, while projections from ZI to IP are scarcer, and terminate preferentially in the ipsilateral side (Mitrofanis and De Fonseka, 2001).However, to our knowledge no functional characterization has been ascribed to such incertal-cerebellar pathways.

Hindbrain
As outlined in the previous paragraphs, the ZI is considerably interconnected with somatosensory-related areas and a multitude of studies have also unraveled the existence of afferents from both the dorsal column nuclei (DCN) (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Nicolelis et al., 1992) and the trigeminal sensorial nuclei (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Nicolelis et al., 1992;Williams et al., 1994;Veinante and Deschênes, 1999;Veinante et al., 2000;De Chazeron et al., 2004;Lavallée et al., 2005;Urbain and Deschênes, 2007;Simpson et al., 2008).Among the trigeminal nuclei, the heaviest projection arises from the contralateral principal sensory nucleus (PSV) (Shammah-Lagnado et al., 1985;Simpson et al., 2008); moreover, based on anterograde and retrograde tracing studies, the pars oralis (SPVO) and pars interpolaris (SPVI) of the contralateral spinal trigeminal complex are considered to provide more extensive afferent projections to the ZI compared to the pars caudalis (SPVC) (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Veinante et al., 2000;De Chazeron et al., 2004).From the functional perspective, Urbain and Deschênes (2007) have shown that a lesion of the ascending projections arising from the SPVI, but not from the PSV, and terminating in the ZI prevents this subthalamic nucleus from displaying electrophysiological responses to the whisker stimulation but not to the stimulation of the perioral or nose area, raising the possibility that the PSV-ZI pathway may transmit information regarding the perioral area, and the SPVI-ZI pathway may transmit vibrissae-related somatosensory information (Urbain and Deschênes, 2007).Given the rich interconnections that the ZI establishes with somatosensory centers at different levels of the neuroaxis and the existence of a rough somatotopy, the ZI has been proposed to function as a distinct channel of somatosensorial information that operates in parallel to the other somatosensorial pathways (Nicolelis et al., 1992).
The ZI also receives afferents originating from the parabrachial nucleus (PB), and several studies have specifically identified the Kölliker-Fuse subnucleus (KF) as source of afferents (Shammah-Lagnado et al., 1985;Urbain and Deschênes, 2007).Parabrachial afferents have been functionally implicated in the formation of the conditioned taste aversion (CTA) (Sakai and Yamamoto, 1999), a specific and adaptive form of learning that, by associating a negative visceral state with a specific gustatory perception, allows the animal to avoid the food that most probably caused the malaise.Sakai and Yamamoto proposed that the pathway encompassing parabrachial projections to the ZI, incertal projections to the intralaminar thalamic complex (MITC) and projections from the latter to the basolateral amygdala may constitute an highly impacting route involved in the CTA formation together with other routes, such as the one from the PB to the central amygdala (CEA) and the one from the PB to the basolateral amygdala encompassing the cortex (Sakai and Yamamoto, 1999).
Scant information is available regarding the incertal anatomical relationship with other nuclei of the hindbrain.For example, several studies have identified reciprocal projections between the ZI and the G. Arena et al. pontine reticular nucleus (PRN) (Roger and Cadusseau, 1985;Shammah-Lagnado et al., 1985;Kolmac et al., 1998;Zhao et al., 2019;Wang et al., 2020aWang et al., , 2020b;;Ahmadlou et al., 2021) with the oral part of the PRN being more densely targeted by incertal projections compared to the caudal part (Shammah-Lagnado et al., 1987).However, no neurochemical profile has been provided for such projections.On the contrary, a more detailed chemical characterization has been conducted on the neurons of the LC projecting to the ZI, which have been shown to be positive to a wide range of peptides, among which substance P, bombesin, and galanin (Lechner et al., 1993).Functionally, Oh et al. (2023) have hypothesized a possible role for the incertal projections to the monoaminergic and cholinergic nuclei, among which the LC, in the regulation of paradoxical sleep, since they receive projections from a population of Lhx6 + neurons of the ventral ZI, which has been involved in paradoxical sleep regulation.However, only a small fraction (4%) of this population projected to the LC in this study, leaving the possibility for future studies to address the questions whether other incertal populations may target the LC and what kind of information is exchanged between these two structures.Lastly, the pain-modulator RVM (Chen and Heinricher, 2022), which includes the raphe magnus (RM), raphe pallidus (RPA), the gigantocellular and several parts of the paragigantocellular reticular nucleus (Hossaini et al., 2012), has been reported by several studies to be anatomically connected to the ZI (Shammah-Lagnado et al., 1985;Zhao et al., 2019), with some incertal cells projecting to the RM being tyrosine hydroxylase (TH) positive (Sim and Joseph, 1989).The role of the ZI-RVM pathway in the light of their joint implication in pain modulation is a perspective to be considered yet.

Other areas
Injections of anterograde tracer in the central, medial and basolateral amygdalar nuclei (CEA, MEA, BLA) produced robust labeling of anterograde-like elements in the ZIr, while injections in CEA also labeled a few fibers in the ZIv and injections in MEA yielded sparser terminal labeling in ZIc and ZId (Reardon and Mitrofanis, 2000), consistently to CEA-ZI projections showed by Roger and Cadusseau (1985) and Shammah-Lagnado et al. (1985).On the other hand, efferents from the ZI to the amygdalar nuclei have only been shown targeting the CEA (Wagner et al., 1995;Fu et al., 2020) and the basomedial (BMA) nucleus (Zhang et al., 2022).Only CEA-ZI and ZI-BMA pathways, among the circuits involving the ZI and the amygdala, have been investigated also in a functional perspective.Concerning the former, a recent study has implicated a particular subset of CEA-ZI GABAergic projections in the regulation of behavioral pain responses.Specifically, Singh et al. (2022) have shown that the optogenetic manipulation of this pathway influences the sensitivity to evoked pain, so that its inhibition reduces the hypersensitivity to thermal and mechanical stimulation in the injured paw of a mouse model of neuropathic pain, and its activation increases the sensitivity to the same stimuli in the uninjured paw.Moreover, it has been shown that the silencing of the inhibitory CEA inputs to incertal parvalbumin-positive neurons disrupts the acquisition and the retrieval of fear memory (Zhou et al., 2018).
Interestingly, the ZI-BMA pathway has also been implicated in the associative aversive learning: the activation of this pathway enhances freezing responses in a mouse model with aversive learning disabilities (Zhang et al., 2022).
The ZI also relates to the medial septal complex (composed of the medial septum, MS, and the diagonal band nucleus, NDB): injection of retrograde tracer in the MS/NDB labeled neurons in the ZIr which also resulted to be immunoreactive to the neuropeptide EI; a subsequent injection of anterograde tracer in the ZI confirmed such projection highlighting the existence of terminal-like structures in the MS (Bittencourt and Elias, 1998).Moreover, varicosities that indicate the existence of synapses have also been identified on the fibers arising from the rostral ZI and coursing through the bed nuclei of the stria terminalis (BST), the diagonal band nucleus (NDB) as well as through the lateral septum (LS) (Wagner et al., 1995).Lastly, descending fibers from the substantia innominata (SI) to the PPN course through the ZI, forming also terminal boutons particularly in the ZIc (Mogenson et al., 1985).

Organization and connections of the ZI in non-human primates
Also the macaques ZI has been subdivided into four regions: a rostral (ZIr), a caudal (ZIc) and a central region, which is further subdivided into a dorsal (ZId) and a ventral (ZIv) one (Watson et al., 2014).Each subdivision is characterized by cyto-, chemo-, and immunoarchitectonical features, like in rodents: ZIc is highly positive for acetylcholinesterase (Watson et al., 2014), tyrosine hydroxylase positive neurons are mainly located in ZIr, nitric oxide synthase positive cells in ZId, parvalbumin positive neurons in ZIv, whereas calbindin positive neurons are equally located in all sectors (Mitrofanis et al., 2004).It is known, since early connectional studies, that the macaque ZI is a target of subcortical projections from the deep layers of the SC, the pretectal region, and the nucleus of the optic tract (Benevento and Fallon, 1975;Benevento et al., 1977;Harting et al., 1980;Mustari et al., 1994;Büttner-Ennever et al., 1996).Other projections, targeting mainly the ZIv, have been described as originating from the principal sensory trigeminal and dorsal funicular nuclei (Smith, 1975) and, more sparsely, from the spinothalamic tract (Boivie, 1979;Apkarian and Hodge, 1989).Furthermore, the ZI receives also noradrenergic (Ginsberg et al., 1993) and dopaminergic (García-Cabezas et al., 2009) innervation.After anterograde neural tracer injections at the cortical level, several studies observed labeled terminals in the ZI.Specifically, projections originated from: i) frontal and parietal oculomotor-related areas, i.e. frontal eye field, supplementary eye field, and parietal area LIP (Huerta et al., 1986;Huerta and Kaas, 1990;Asanuma et al., 1985); ii) primary somatosensory and primary motor areas (Künzle, 1977;Coudé et al., 2018); iii) medial prefrontal, orbitofrontal, and frontal opercular areas ( Öngür et al., 1998;Simonyan and Jürgens, 2003); iv) mesial parietal, caudal cingulate, and preoccipital areas (Leichnetz, 1990(Leichnetz, , 2001;;Parvizi et al., 2006); v) antero-medial temporal regions (Ogasawara et al., 2022).Accordingly, also in macaques the ZI receives a large variety of cortical and subcortical inputs, but the topography especially of the cortical projections within the ZI has only been occasionally described.
On the other hand, anatomical studies so far indicate that the projections of the macaque ZI target a few exclusively subcortical structures.Based on a series of retroanterograde tracer injections, May and Basso (2018) have shown that the ZI projects to the pretectal region and the SC, where it targets the intermediate and deep layers, which are the source of collicular oculomotor output.These projections originate mostly from ZIv.The macaque ZI is also a source of projections to the dorsomedial pontine nuclei and the adjacent reticularis tegmenti pontis (Giolli et al., 2001), which are precerebellar structures involved in visuomotor control of saccades (see, e.g., May, 2006).The ZI projects also to the mediodorsal (MD) nucleus of the thalamus, especially to its anterodorsal part (Erickson et al., 2004), and, in the squirrel monkey, to the PAG (Dujardin and Jürgens, 2005), which is in turn a source of projections to the MD nucleus (Erickson et al., 2004).This thalamic projection suggests a role for the ZI in modulating the activity of the prefrontal cortex (Mitchell et al., 2014).Interestingly, the MD nucleus of the macaque is characterized by a richer dopaminergic innervation and local GABAergic interneuronal circuitry with respect to rodent dorsal thalamus, suggesting a possible differential impact of the external inhibitory signals, including those from ZI, on thalamic information processing in the two species (see, Perry et al., 2021).Finally, some incertal neurons project to the GP (DeVito et al., 1980).All these projections are described as almost exclusively ipsilateral.
A recent study described more in detail the connectivity of the macaque ZIr, based on anterograde tracer injections at the cortical level and a retroanterograde tracer injection in the ZIr (Haber et al., 2023).
Specifically, this part of the ZI is mainly a target of cortical projections from the dorsolateral prefrontal and anterior cingulate cortex, originating from layer V.The main subcortical inputs originate from intralaminar thalamic nuclei, dopaminergic structures of the VTA and substantia nigra, the medial hypothalamus, the reticular formation, and the PPN.Some cells were also observed in the amygdala, DR, and PAG.All these structures are also targets of the ZIr output, with richer projections reaching the thalamus, the reticular formation, and the PPN.Interestingly, one of the most prominent outputs of the ZIr is to the thalamic lateral habenular nucleus (LH).The connectivity pattern of the ZIr seems to be markedly different from that of the more caudal parts of the ZI, which receive projections mostly from motor-related areas and do not project to the LH.Accordingly, the ZIr could be a connectional hub for reinforcement learning processes.Interestingly, another recent study demonstrated that the intermediate part of the ZI (likely ZId and ZIv) is causally involved in novelty seeking behavior, by processing novelty prediction signals originating from the anterior medial temporal cortex and controlling gaze shift (Ogasawara et al., 2022).Altogether, these data suggest that the macaque ZI integrates different cortical and subcortical signals for modulating behavior and allowing behavioral flexibility (Haber et al., 2023).

Conclusions and future perspectives
As emerged by the present work, piecing what we know about the ZI together reveals a patchy knowledge of this area, notwithstanding the numerous and constructive attempts that have been made to try to grasp its real nature.Such gaps exist on multiple levels, spanning from the anatomical characterization to the neurochemical nature.However, on the functional level more than on the others such gaps reveal their weight, since several perspectives have emerged over time that have opened as many research paths, resulting in divergent and multiple frontiers for the study of the ZI that exacerbate the lack of cohesion that today exists.Therefore, a common line of research, oriented to unify rather than separate, is needed to puzzle out the real identity of this area.Moreover, many research questions await consideration.For instance, no functional characterization is available regarding the relationship between the ZI and the cerebellum, which would be interesting to consider in the light of the cerebellar involvement in both motor and cognitive functions (Koziol et al., 2014).Similarly, details of connections exchanged between the ZI and several structure, as the hippocampal areas, the basolateral amygdala, large territories of the isocortex and numerous thalamic nuclei remain to be disclosed in terms of neurochemical identity, anatomical ultrastructure and functional implications, and provide valid questions to be addressed by future studies.

Fig. 2 .
Fig. 2. Anatomical connections established by the ZI.Unidirectional/bidirectional arrows indicate that unidirectional/bidirectional projections have been shownbetween the ZI and the other area considered.Each external area is placed in larger boxes indicating its anatomical allocation on the basis of the classification reported by the Allen Mouse Brain Atlas.Asterisks refer to areas labeled with different nomenclatures in respect to the ones reported by the Allen Brain Atlas, either for brevity (for example, "RVM" includes multiple areas recognized by the Allen Brain Atlas) or for complications related to the use in literature of several nonoverlapping nomenclature systems ("Temporal cortex" and "VISs").For more details, please refer to "Cortical structures" and "Hindbrain" sections.