The Cerebellar Dopaminergic System

In the central nervous system (CNS), dopamine (DA) is involved in motor and cognitive functions. Although the cerebellum is not been considered an elective dopaminergic region, studies attributed to it a critical role in dopamine deficit-related neurological and psychiatric disorders [e.g., Parkinson's disease (PD) and schizophrenia (SCZ)]. Data on the cerebellar dopaminergic neuronal system are still lacking. Nevertheless, biochemical studies detected in the mammalians cerebellum high dopamine levels, while chemical neuroanatomy studies revealed the presence of midbrain dopaminergic afferents to the cerebellum as well as wide distribution of the dopaminergic receptor subtypes (DRD1-DRD5). The present review summarizes the data on the cerebellar dopaminergic system including its involvement in associative and projective circuits. Furthermore, this study also briefly discusses the role of the cerebellar dopaminergic system in some neurologic and psychiatric disorders and suggests its potential involvement as a target in pharmacologic and non-pharmacologic treatments.

Studies reported in the developmental and adult mammalian cerebellum the presence of extrinsic monoaminergic pathways. Studies have been mainly focused on the cerebellar functional role of serotonin (5-HT) and noradrenaline (NA); as a result, until now, the functional role of dopamine (DA) in the cerebellum has been widely disregarded.
In the human developmental cerebellum, at 16-18 and 26-28 weeks, a transient expression of noradrenergic neuronal cell bodies and processes occurs in the cerebellar cortex and in the deep cerebellar nuclei has been demonstrated (Yew et al., 1995). In addition, in the cerebellum of mammals, extensive distribution of the β 2 adrenergic subtype receptor (Pompeiano et al., 1989;Voogd et al., 1996) and, to a lesser extent, of β 1 , α 1 , and α 2 adrenergic subtype receptors have been demonstrated (Pompeiano et al., 1989;McCune et al., 1993;Rosin et al., 1996;Talley et al., 1996;Voogd et al., 1996). In the development, it has been found that the cerebellar noradrenergic system influences mainly the GABAergic synaptogenesis (Sievers et al., 1981;Sievers and Klemm, 1982;O'Leary and Leslie, 2003;Happe et al., 2004;Hirono et al., 2014). In the adult cerebellum, NA plays a pivotal role in the modulation of the glutamatergic and GABAergic synaptic signaling (Moises et al., 1983;Woodward et al., 1991;Hirono and Obata, 2006;Hirono et al., 2014;Lippiello et al., 2015). Noradrenaline exerts on the Purkinje neurons two types of influence. An increase of the intracellular levels of cAMP protein kinase-dependent by means on the betaadrenergic receptor (Kano et al., 1992;Cheun and Yeh, 1996); the levels of cAMP can, in turn, enhance a form of neuronal plasticity called rebound potentiation (RP; Kano et al., 1992;Cheun and Yeh, 1996;Kawaguchi and Hirano, 2002). Moreover, NA influences in the Purkinje neurons the expression of the immediate-early genes, c-fos and Jun-B (Pompeiano, 1998). The induction of immediate-early genes in the Purkinje neurons appears to play a role in the long-term biochemical changes involved in the maintenance of cerebellar long-term plasticity such as LTD (Pompeiano, 1998).
On the other hand, currently, the presence and the distribution of a dopaminergic system in the cerebellum and its functional role is controversial or neglected (Oertel, 1993;Ottersen, 1993;Kwong et al., 2000). However, several studies demonstrated the involvement of the cerebellum in DA related neurological and psychiatric disorders, such as Parkinson's disease (PD), schizophrenia (SCZ), autism spectrum disorders (ASD), and drug addiction (Glaser et al., 2006;Andreasen and Pierson, 2008;Mittleman et al., 2008;O'Hallaran et al., 2012;Lewis et al., 2013;Wu and Hallett, 2013;Parker et al., 2014, Carta et al., 2019Gil-Miravet et al., 2019;Miquel et al., 2020). Therefore, the goal of the present review is to provide a comprehensive overview of the presence, distribution, and functional role of the cerebellar dopaminergic system, also discussing its potential pathophysiological and clinical implications in some neurological and psychiatric DArelated disorders.
During the development of the mouse cerebellar cortex, a transient expression of TH in Purkinje neurons in different ages from postnatal day 3 (P3) to 11 months (M11) has been observed (Fujii et al., 1994). The TH expression appears in the Purkinje neurons at P8 in the cerebellar vermis, increases at P13-P15, reduces at P19, and then increases again after 1 month of age, reaching a maximum expression at 11 months (Fujii et al., 1994).
In the adult mouse cerebellum, the TH immunoreactive fibers are in the vermal lobules V and VI, whereas the lowest numbers are located in lobule X, and in each deep cerebellar nuclei, a dense plexus of TH immunoreactive varicose fibers has been mainly detected ( Table 1; Nelson et al., 1997). Whereas, TH immunoreactive cell bodies of Purkinje neurons have been found in the flocculus, paraflocculus, vermal lobules VI-X, and in the hemispheric lobules IX-X ( Table 1; Nelson et al., 1997). In pharmacological studies, in the mouse cerebellum DA specific binding sites of [ 3 H]DA and [ 3 H]spiperone has been detected (Panagopoulos and Matsokis, 1994).
Moreover, in the adult mouse cerebellum, specific binding of the DA uptake inhibitor 3 [H]GBR12935 in the paraflocculus, lobules IV, VI, IX, X, and lobule simplex Crus I and II has been detected (Delis et al., 2008). In the cerebellar cortex, the specific binding of 3 [H]GBR12935 was mainly distributed in the molecular layer and in the granular layer, while DAT immunoreactivity has been mainly detected in the cell bodies of the Purkinje neurons and in some neuron types of the deep cerebellar nuclei ( Table 1; Delis et al., 2008). Furthermore, in the mouse cerebellum, DARPP-32 immunoreactive Purkinje neuron cell bodies in the laminae of all lobules have been observed ( Table 1; Alder and Barbas, 1995).
In the rat cerebellum, the DA immunoreactivity presents a uniform distribution pattern in all lobules, and in the layers of the cerebellar cortex the DA immunoreactivity was mainly detected in the molecular layer in climbing fiber-like forms, while a small number of DA immunoreactive fibers within the Purkinje neuron layer and in the granular layer were found ( Table 1; Panagopoulos et al., 1991).
Furthermore, in the rat cerebellum, the TH immunoreactive fibers in the paraflocculus and crus I and II ansiform lobules have been mainly detected ( Table 1; Ikai et al., 1992); whereas, a high number of VMAT 2 immunoreactive 'puncta' (attributable to axon terminals or short sections of dendrites) has been observed in the lobule IX of the posterior cerebellum ( Table 1; Kim et al., 2009).
In the rat cerebellar cortex, a low number TH immunoreactive fibers variously oriented in the Purkinje neuron layer and in the granular layer has been detected; instead, in the molecular layer, a high number of climbing-like oriented TH immunoreactive fibers has been observed ( Table 1; Takada et al., 1993), and VMAT 2 small immunoreactive "puncta" were observed between the Purkinje neuron cell bodies and in the molecular layer in close relationship with the dendritic arborizations of the Purkinje neurons ( Table 1; Kim et al., 2009).
Biochemical analysis revealed significant levels of DA in the deep cerebellar nuclei of rat, with the highest DA levels being localized in the fastigial and dentate nuclei (Glaser et al., 2006), which is in line with a morphological study that revealed the presence of DAT immunoreactive fibers in all deep cerebellar nuclei (Delis et al., 2008). Moreover, a wide distribution of DAT immunoreactive fibers in the three layers of the cerebellar cortex and in the deep cerebellar nuclei has been revealed (Delis et al., 2008). In addition, DARPP-32 immunoreactive dendritic arborization of the Purkinje neurons in the molecular layer of all cerebellar lobules has been observed ( Table 1; Alder and Barbas, 1995).
In the rat cerebellar cortex, the presence of TH immunoreactive cell bodies of Purkinje neurons has been demonstrated in the lobules I and X of the vermis, in the paraflocculus, and in crus I and II ansiform lobules (Table 1; Takada et al., 1993) Instead, Kim et al. (2009) evidenced the presence of TH immunoreactive cell bodies of Purkinje neurons predominantly in the lobules VIII-X and a discontinuous presence in the lobules VI and VII, whereas a high number of DAT immunoreactive cell bodies of Purkinje neurons has been detected in the lobule IX of the posterior cerebellum (Table 1; Kim et al., 2009).
In the opossum cerebellum, most of TH immunoreactive fibers have been found in the lobules III-VIII of the vermis and, to a lesser extent, in lobules I and X ( Table 1; Nelson et al., 1997). Moreover, in the opossum cerebellar cortex, the TH immunoreactive fibers were mainly localized in the Purkinje neurons layer; they surround the cell bodies of the Purkinje neurons or run parallel to the plane of the Purkinje neuron layer, whereas, in the molecular layer, only a small amount was detected ( Table 1; Nelson et al., 1997), and in the granular layer, the TH immunoreactive fibers featured a random distribution (Nelson et al., 1997).
In addition, a moderate number of randomly distributed TH immunoreactive fibers has been detected in the deep cerebellar nuclei (Nelson et al., 1997).
In the cat cerebellum, the highest density of TH immunoreactive fibers were distributed in the vermal lobules V and VI and in the hemispheric lobules VI and crus I and II; meanwhile, the lowest density of fibers has been observed in lobules I-III and VIII-X (Table 1; Nelson et al., 1997). In the cat cerebellar cortex, the high density of TH immunoreactive fibers have been observed in the granular layer, where they present a random orientation, and in the Purkinje neurons layer, where they surround the cell bodies of the Purkinje neurons have been observed. Instead, in the molecular layer, only a few densities of TH immunoreactive fibers with a perpendicular orientation that often extend radially to the surface of the pial surface of the cortex have been detected (Table 1; Nelson et al., 1997). Finally, a moderate density of TH immunoreactive varicose fibers in the deep cerebellar nuclei have been also found ( Table 1; Nelson et al., 1997).
In the monkey cerebellum, we found a low density of TH immunoreactive fibers to be distributed in the lobules of the vermis and of both cerebellar hemispheres, whereas the DAT immunoreactive fibers were only observed in the vermis of the following lobules II, III, IV, VIIIA, VIIIB, IX, and X (Table 1; Melchitzky and Lewis, 2000).
In the monkey cerebellar cortex, a low density of TH immunoreactive fibers in the granular layer and in the molecular layer, has been detected while a higher density of TH plexuses and axonal terminals just beneath the Purkinje neuron cell bodies has been found. Conversely, DAT immunoreactive fibers to be randomly distributed in the granular layer; however, we also found forming plexuses around the deep pole of the cell bodies of the immunonegative Purkinje neurons has been observed. In contrast, in the molecular layer, no DAT immunoreactive fibers have been detected (Table 1; Melchitzky and Lewis, 2000).
Currently, in the monkey cerebellum studies, there is no evidence that proves the existence of dopaminergic neurons. Despite this, a biochemical study demonstrated significant levels of DA in all deep cerebellar nuclei, and the highest levels have been detected in the interpositus and dentate nuclei, (Glaser et al., 2006). Furthermore, in all cerebellar lobules of the monkey cerebellar cortex, a wide presence of DARPP-32 immunoreactive cell bodies and dendritic arborizations of Purkinje neurons has been detected ( Table 1; Alder and Barbas, 1995).
In the human cerebellum, immunohistochemical experiments revealed the presence of DAT immunoreactive fibers and neuronal cell bodies in lobules VII and IX (crus I and II, ansiform lobules, and tonsilla) and in the dentate nucleus (Table 1; Figures 1, 2; Flace, 2017Flace, , 2019bFlace, , 2020Flace et al., 2018aFlace et al., , 2019bFlace et al., , 2020. There is a significant presence of DAT immunoreactive dendritic arborization of the Purkinje neurons in the molecular layer of the human cerebellar cortex (Table 1; Figure 1). Moreover, the DAT immunoreactivity has been detected in form of clusters in the neuropil among the space FIGURE 2 | Dopamine receptor type 2 (DRD 2 ) immunoreactivity is observable in neuronal bodies and processes in the layers of the cerebellar cortex. In the ML:DRD 2 immunoreactive basket and stellate neuron cell bodies (arrows), DRD 2 immunoreactive primary, secondary, and apical dendrites of Purkinje neurons, DRD 2 immunoreactive Purkinje neuron cell bodies (single asterisk). In the granular layer, DRD 2 immunoreactivity in space of Held; DRD 2 immunoreactive Golgi neuron cell body (arrow) (Scale bar: 20 µm).
FIGURE 3 | Dopamine transporter (DAT) immunoreactivity in the dentate nucleus. The dopamine transporter (DAT) immunoreactivity is detectable in the dentate nucleus gray substance and in the neighboring white substance. DAT immunoreactive small neuron cell bodies (arrowheads); DAT immunonegative small neuron cell body (single asterisks) DAT immunoreactive cell body of projective neuron type, central neuron (arrow); fine clusters of DAT immunoreactivity in the neuropil of the nucleus and in the neighboring withe substance (Scale bar: 15 µm).
In the three layers of the cerebellar cortex, the dopaminergic receptor subtypes present a different distribution pattern. In the molecular layer, immunoreactivity to DRD 2 , DRD 3 , and DRD 5 receptors in the cell bodies and processes of stellate neurons, basket neurons, and in the dendritic arborizations of the Purkinje neurons has been detected (Table 2; Figure 2). Moreover, in the neuropil of the molecular layer, fine clusters of DRD 2 immunoreactivity were detected (Table 2; Figure 2; Camps et al., 1990;Ricci et al., 1995bRicci et al., , 1996Khan et al., 1998;Levant, 1998;Barili et al., 2000;Flace et al., 2018aFlace et al., , 2019aFlace et al., ,b, 2020.
In chemical neuroanatomy studies carried out on the cerebellum by means of antisera directed against the TH, the rate-limiting enzyme of DA biosynthesis and the presence of numerous TH immunoreactive fibers in the various lobules and laminae of the cerebellar cortex as well as in the deep cerebellar nuclei have been demonstrated (Austin et al., 1992;Ikai et al., 1992;Takada et al., 1993, Nelson et al., 1997. In addition, regarding the TH immunoreactivity, it should be indicated that it is related to the presence of NA or DA, or both, since by carrying out a selective depletion of NA, most of the immunoreactivity is abolished (Fuxe, 1965;Hökfelt and Fuxe, 1969;Bloom et al., 1971). Moreover, using biochemical techniques, low levels of DA were found in the cerebellum (Carlsson, 1959;Glowinski and Iversen, 1966;Landis and Bloom, 1975). In addition, using antisera against DBH, the enzyme responsible for the biosynthesis of NA, highlights the presence of fibers in the cerebellum, which presented only partial similarity to those observed in studies using directed antisera against DA (Verney et al., 1988;Panagopoulos et al., 1991). Furthermore, more recent studies in the cerebellum of various mammals species and in other regions of the central nervous system (CNS), demonstrated which distribution patterns of TH immuroreactivity were mainly correlated to catecholaminergic and not electively to the dopaminergic neurotransmission but did not exclude it; (Fallon and Moore, 1978;Hökfelt et al., 1984;Asan, 1993;Takada et al., 1993;Nelson et al., 1997). On the contrary, studies that electively used antisera against DAT evaluate specifically the distribution patterns related to dopaminergic neurotransmission (Melchitzky and Lewis, 2000;Delis et al., 2008;Flace et al., 2018aFlace et al., , 2019bFlace et al., , 2020. In addition, the DAT immunohistochemial studies evidenced the presence of a specific subpopulation of dopaminergic neuronal cell bodies and processes in the cerebellum (Melchitzky and Lewis, 2000;Delis et al., 2008;Flace et al., 2018aFlace et al., , 2019bFlace et al., , 2020, which is in FIGURE 4 | Dopamine receptor type 2 (DRD 2 ) immunoreactivity in the dentate nucleus. The (DRD 2 ) immunoreactivity is detectable in the dentate nucleus gray substance and in the neighboring withe substance; DRD 2 immunoreactive small neuron cell bodies (arrowheads); DRD 2 immunoreactive cell body of projective neuron type, central neuron (arrow); diffuse DAT immunoreactivity in the neuropil of the nucleus (Scale bar: 15 µm).
In fact, the relationship of these data suggest the existence in the cerebellum of detailed dopaminergic neurotransmitter mechanisms. For example, in terms of the distribution pattern of the DAT immunoreactivity (Melchitzky and Lewis, 2000;Delis et al., 2008;Flace et al., 2018aFlace et al., , 2019bFlace et al., , 2020 and of the dopaminergic D1-like and D2-like subtype receptors immunoreactivity (Camps et al., 1990;Bouthenet et al., 1991;Ricci et al., 1995a,b;Khan et al., 1998;Lazarov et al., 1998;Barili et al., 2000;Kim et al., 2009;Flace et al., 2018aFlace et al., , 2019aFlace et al., ,b, 2020, both were expressed in the Purkinje neurons cell bodies, dendritic arborizations, and axons, and this suggests the existence of a detailed cerebellar modulation by means of dopaminergic neurotransmission mechanisms in intrinsic and extrinsic cerebellar circuits.

PHYSIOLOGICAL ASPECTS ON THE DOPAMINERGIC CEREBELLAR SYSTEM
In animal model studies, it has been indirectly demonstrated an active role of DA in the cerebellum; indeed, the administration of lacosamide and morphine in a hypoglycemic animal model decreased the cerebellar level of dopamine significantly (Guzman et al., 2014). In the cerebellum of albino rats, the long administration of morphine sulfate determines a decrease in the levels of DA and histopathological changes (Bekheet et al., 2010). Hypoxic conditions induced in the cerebellum of neonatal rats a decrease in the DA levels and a reduced expression of the dopaminergic subtype receptors DRD 1 and DRD 2 ; these decreases are in part reversed by the supplementation of glucose, oxygen, and adrenaline (Joseph et al., 2010). Moreover, in the cerebellum of rodents, high levels of DA have been involved in neuronal synaptic mechanisms characterized by DA release and uptake (Efthimiopoulos et al., 1991;Dethy et al., 1997). In addition, in mouse cerebellar slices, the presence of a high-affinity Na + -dependent DA uptake system has been demonstrated, and this has been characterized by a K + -induced, Ca +2 -dependent dopamine release mechanism (Efthimiopoulos et al., 1991). Moreover, in several studies, it has been demonstrated in striatal medium spiny neurons a direct influence of DA in the mechanism of structural plasticity of dendritic spines (Yagishita et al., 2014). In the rat cerebellum, DA may influence in the Purkinje neurons the induction of RP a form of longlasting synaptic plasticity at inhibitory synapses by means of the cAMP-regulated protein DARPP-32 highly expressed in Purkinje neurons and involved in dopaminergic neuronal synaptic signaling (Alder and Barbas, 1995;Kawaguchi and Hirano, 2002).
Moreover, in rat Purkinje neurons dendrites, a release of DA from vesicular extrasynaptic and postsynaptic sites resulted in dopaminergic receptors paracrine and autocrine activation (volume transmission), which produced a Depolarization-Induced Slow Current (DISC; Kim et al., 2009). Moreover, in pharmacological experiments a close functional relationship in dopaminergic Purkinje neurons between DA signaling and DISC has been demonstrated; in fact, it was blocked by dopaminergic receptor antagonist (e.g., clozapine, haloperidol, and eticlopride), VMAT 2 inhibitors (reserpine and tetrabenazine), and dopamine reuptake inhibitors (e.g., rimcazole; Kim et al., 2009). Furthermore, it has been suggested in recent studies which TH immunoreactive Purkinje neurons and DRD 1 immunoreactive large projective neuron types of the dentate nucleus may be involved in the modulation of cerebellar cognitive functions (Locke et al., 2018(Locke et al., , 2020. The selective chemogenetic inhibition of the DRD 1 immunoreactive neuron type of the dentate nucleus could be involved in the impairment of cognitive functions such as spatial navigation memory, working memory, and pre-pulse inhibition of the acoustic startle reflex (Locke et al., 2018). In mice, a selective reduction of TH immunoreactive cerebellar Purkinje neurons has been correlated to a specific impairment of cognitive functions, such as behavioral flexibility, response inhibition, social recognition memory (Locke et al., 2020).
From the analysis of these experimental physiological and pharmacological studies, a potential role of the neuronal dopaminergic system at the cerebellar level emerges, especially in the synaptic and extrasynaptic neurotransmission and neuromodulation mechanisms (Efthimiopoulos et al., 1991;Dethy et al., 1997;Kawaguchi and Hirano, 2002;Kim et al., 2009) and, in cognitive functions related to the cerebellar activity (Locke et al., 2018(Locke et al., , 2020. Overall, they deserve further evaluation in order to better understand the relevance of the morphofunctional role played by the dopaminergic innervation in the cerebellum and their role in the behavioral functions of the cerebellum.
In addition, in cat and in rat, a direct cerebellar influence on the midbrain dopaminergic nuclei (A 8 -A 10 ) has been demonstrated. Fibers from the vermian cerebellar cortex and from the fastigial nucleus reach the ipsilateral VTA, whilst fibers from the interpositus and dentate nuclei reach the contralateral dorsal VTA and the medial and dorsal SNpc; moreover, 20% of the fibers had bilateral interconnections (Figure 9; Snider and Maiti, 1976).
In the rat cerebellum, using horseradish peroxidase (HRP) anterograde and retrograde transport methods, the efferents of the dentate and interpositus nuclei to the contralateral midbrain dopaminergic cell groups A 8 -A 10 have been demonstrated (Figure 9; Perciavalle et al., 2013).
Electrical stimulation of cat cerebellar dentate nucleus influenced the dopaminergic activity of the ipsilateral SNpc, which in turn increased the release of [ 3 H]-DA in the contralateral caudate nucleus and decreasing such release in the ipsilateral caudate nucleus. Moreover, the electrical stimulation of the fastigial nucleus increased only the release of [ 3 H]-DA in the ipsilateral caudate nucleus (Nieoullon et al., 1978), and the electrical stimulation of the posterior interpositus nucleus increased the release of [ 3 H]-DA in ipsilateral SNpc and in the contralateral caudate nucleus while decreasing the release [ 3 H]-DA in the ipsilateral caudate nucleus (Nieoullon and Dusticier, 1980). In addition, the electrical stimulation of the mouse cerebellar dentate nucleus was elicited mainly in the contralateral nucleus accumbens (NAc), determining an asymmetrical and lateralized DA release (Figure 9; Holloway et al., 2019).
Moreover, in the last decades, the developments in neuroscience research of non-invasive and in vivo diffusion Magnetic Resonance Imaging and tractography have been increasingly used for the neuroanatomical reconstruction of putative white substance tracts or links of the human brain (Jeurissen et al., 2014;Cacciola et al., 2016aCacciola et al., ,b, 2017aCacciola et al., ,b, 2019. Although relatively few studies focused on the connectivity of midbrain nuclei, some of these reported structural connectivities between dopaminergic midbrain regions and the cerebellum (Bareš et al., 2015;Milardi et al., 2016;Cacciola et al., 2017a;Flace et al., 2017Flace et al., , 2018aFlace et al., ,b, 2019aFlace et al., , 2020). An early work based on diffusion tensor imaging (DTI) and deterministic tractography aimed at the reconstruction of the median forebrain bundle (MFB), which represents the main white substance pathway connecting VTA and SNpc to the prefrontal cortex (PFC), found also a descending branch reaching to the cerebellum and in particular to the dentate nucleus through the superior cerebellar FIGURE 5 | Structural connectivity of SNpc and VTA, including putative midbrain-cerebellar connectivity. Data were obtained from the 100-unrelated-subjects sample of the HCP repository (see Van Essen et al., 2013) Diffusion datasets were processed using a multi-shell, multi-tissue constrained spherical deconvolution (MSMT-CSD) algorithm (see Jeurissen et al., 2014). A number of 10,000 streamlines passing through the left SNpc (dark red) and VTA (light red) regions of interest (see Pauli et al., 2018) (A) was generated. Streamlines were mapped to structural scans, transformed to MNI152 standard space, binarized, and summed up to obtain tract maximum probability maps (MPMs). A threshold of 50% was applied to show only tracts overlapping in at least half of the sample (B). Tractography was run on 30 high-quality 3T structural and diffusion data from the Human Connectome Project (HCP). Data were downloaded in a minimally pre-processed form and elaborated using the signal processing technique known as Constrained Spherical Deconvolution (CSD). Regions of interest (ROI) were delineated by means of multi-atlas automated segmentation: Substantia nigra (SN) and Ventro Tegmental Area (VTA) were resliced into subject space from Adcock's probabilistic atlas; dentate nucleus (both dorsal and ventral part) using the deep cerebellar nuclei atlas featured in SPM Anatomy Tract colors are attributed according to the spatial orientation of streamlines: superior-inferior (blue), anterior-posterior (green), and latero-lateral (red).
FIGURE 6 | Dentate-nigral interconnections. Coronal view shows the interconnections between the right dentate nucleus and the ipsilateral SN, and the left dentate nucleus and the ipsilateral SN. The fibers exited the cerebellum via the right and left superior cerebellar peduncles. peduncle (SCP; Coenen et al., 2012). These findings have been replicated in a study by using more advanced signal modeling algorithms and different tracking strategies (Coenen et al., 2018). Nevertheless, results coming from diffusion imaging should be interpreted with care due to the well-known limitations of the tractographic approach, such as the inability to detect axons or synapses and, then, to rule out the precise termination of putative white substance tracts at a cellular level as well as to distinguish between direct or indirect connectivity patterns and passing-by fibers (Jbabdi and Johansen-Berg, 2011). In particular, the inherently low spatial resolution of diffusion-weighted MRI makes it difficult to distinguish between SNpc, SNpr, and VTA, as their precise boundaries are not readily identifiable on conventional MRI scans (Chowdhury et al., 2013;Trutti et al., 2019). In addition, these results may be affected by passing-by fibers from the dento-rubro-thalamic tract (DRTT), which lies in close proximity to midbrain dopaminergic structures, despite FIGURE 7 | Dentate-VTA interconnections. Coronal view shows the interconnections between the right dentate nucleus and the ipsilateral and contralateral VTA, and the left dentate nucleus and the ipsilateral and contralateral VTA. The fibers exited the cerebellum via the right and left superior cerebellar peduncles. a recent study having suggested the potential dissociability of the cerebellar branch of MFB from DRTT (Hosp et al., 2019). In addition, in a human brain structural connectivity tractographic reconstruction of SNpc and VTA, we evidenced the existence of wide interconnections of the cerebellum with the SNpc and also with the VTA (Figure 5).
In addition, recently, by means of Constrained Spherical Deconvolution tractography (CSDt), Milardi et al. (2016) carried out a detailed analysis of direct links between the ventral and dorsal dentate nucleus and the ipsilateral SNpc (Figures 6, 8; Milardi et al., 2016). Subsequently, by means of CSDt, the existence of direct interconnections between the ventral and dorsal dentate nucleus and ipsilateral and contralateral VTA, predominantly characterized by an ipsilateral dentate-VTA links, has also been demonstrated (Figures 7, 8; Bareš et al., 2015;Milardi et al., 2016;Cacciola et al., 2017a;Flace et al., 2017Flace et al., , 2018aFlace et al., ,b, 2019bFlace et al., , 2020. FIGURE 8 | Dentate-SN and dentate-VTA interconnections. Coronal view shows the interconnections between the right dentate nucleus and left dentate nucleus to the ipsilateral SN, between the right dentate nucleus and left dentate nucleus to the ipsilateral and contralateral VTA. The fibers exited the cerebellum via the right and left superior cerebellar peduncles. The interconnection studies conducted with invasive methods in non-human mammals, and the analyses carried out in humans by means of tractographic neuroimaging methods highlight the presence of relevant interconnections of the cerebellum with the traditional dopaminergic areas of the brain. Moreover, this may likely suggest double direct functional DA interactions between the cerebellar dopaminergic system described in this review and the traditional DA cell groups system of the CNS (Figure 9; Björklund and Dunnett, 2007). In addition, these cerebellarmidbrain dopaminergic interconnections could represent part of the cerebellar projective circuits which allow the cerebellum to contribute to motor and cognitive functions (Koziol et al., 2014;Caligiore et al., 2017).

THE ROLE OF THE DOPAMINERGIC CEREBELLAR SYSTEM IN NEUROLOGIC AND PSYCHIATRIC DISORDERS
Though several studies suggested an involvement of the cerebellum in dopaminergic related neurologic and psychiatric disorders as PD (Jellinger, 1999(Jellinger, , 2017Lewis et al., 2013;Wu and Hallett, 2013), SCZ (Andreasen and Pierson, 2008;O'Hallaran et al., 2012;Parker et al., 2014), and ASD (O'Hallaran et al., 2012;Hampson and Blatt, 2015;Phillips et al., 2015), the precise role of the cerebellar dopaminergic system has not been fully characterized yet.
In this review, briefly, we analyzed some considerable experimental and clinical aspects of the cerebellum related to the dopaminergic system and its disorders.
Currently, only in few detailed studies has the direct involvement of a dopaminergic system at the cerebellar level in PD been analyzed. In a 6-hydroxydopamine (6-OHDA) animal model, increases in the DA level and its metabolites in the anterior cerebellum and as well as a decrease in the caudateputamen have been detected (Kolasiewicz et al., 2012). In the cerebellum of PD patients, a reduced mRNA expression of TH and of some dopaminergic receptor subtypes (DRD 1 -DRD 3 ) has been found (Hurley et al., 2003). In a human PD postmortem brain study it was shown that in the Purkinje neurons, a high expression of the calpain II (calpastatin), a calcium-dependent protease, resulted in overexpression in the dopaminergic neurons of SNpc (Mouatt-Prigent et al., 2000). PTEN-induced putative kinase 1 (PINK1) mutations related to the recessive genetic forms of parkinsonism, in the cerebellum of PD patients in Purkinje neuron and in several neuron types of dentate nuclei have been detected (Blackinton et al., 2007;Dodson and Guon, 2007).
The deposition of cerebellar α-synuclein (α-S) during PD remains unclear (Takahashi and Wakabayashi, 2001;Kingsbury et al., 2004). Indeed, some studies evidenced the presence of decreased or unchanged levels of α-S in the cerebellum (Tan et al., 2005;Westerlund et al., 2008), while others demonstrated a high mRNA expression of the α-S gene (SNCA) in the human cerebellum (Fuchs et al., 2008). Moreover, in the cerebellum of PD patients and of [A30P] transgenic mouse as well as in α-S in the molecular layer, the Bergmann glia Piao et al., 2003), in the Purkinje neurons, in the space of Held of the granular layer, in the neuropil and in cell bodies and processes of different neuron types of the dentate nucleus has been found (Kahle et al., 2000;Mori et al., 2003). Furthermore, an α-S neuroprotective activity in cerebellar granules against neurotoxicity of 6-OHDA has been also demonstrated (Monti et al., 2007).
In rat cerebellum, high mRNA expression of clusterin/apolipoprotein J, a glycoprotein involved in the regulation of α-S deposition (Sasaki et al., 2002;Emamzadeh, 2017) in the Purkinje neurons as well as in the neurons of the fastigial and interpositus nuclei, has been detected (Pasinetti et al., 1994).
Currently, no studies are available on the direct involvement of the cerebellar dopaminergic system in SCZ and in autism ASD. However, in several studies, the presence of cerebellar abnormalities in SCZ and ASD patients has been demonstrated. In SCZ, patterns of atrophy in the cerebellar cortex of the vermis have been demonstrated (Weinberger et al., 1980;Reyes and Gordon, 1981;Heath et al., 1982;Snider, 1982;Martin and Albers, 1995). In addition, reduced cerebellar cortical volumes (Laidi et al., 2015), a decreased cerebellar gray substance of Crus I and II ansiform lobules (Kühn et al., 2012), and a reduction in the gyrification index in the cerebellar vermis have also been observed (Schmitt et al., 2011). Moreover, in the cerebellum of SCZ patients, in a microscopical analysis, a loss or a reduced cell size of the Purkinje neurons has been revealed (Stevens, 1982;Tran et al., 1998). Furthermore, a reduced cerebellar expression of the Sp transcription factors and DRD 2 , both related to negative symptoms of SCZ, has been found (Pinacho et al., 2013).
In ASD morphological studies, in the cerebellar hemispheres a reduction of the number of the Purkinje neurons related to a reduction of the Nissl staining has been demonstrated (Bauman and Kemper, 1985;Kemper and Bauman, 1993). Furthermore, in the cerebellum of ASD patients, we also found a reduced Purkinje neuron density (Whitney et al., 2008;Skefos et al., 2014) together with the decreased cell body size of the Purkinje neuron (Fatemi et al., 2002).
Furthermore, electrical stimulations of the Purkinje neuron layer and of the dentate nucleus evokes a long-lasting increase of DA efflux in the PFC, and this suggests a possible disconnection between the Purkinje neurons and neuronal population of the dentate nucleus, which in turn can lead to aberrant DA signaling in the PFC and to abnormal behavior related to symptoms of SCZ and ASD (Mittleman et al., 2008;Rogers et al., 2013). Therefore, the cerebellum and its dopaminergic innervation and their interconnections to the other midbrain dopaminergic areas suggested a direct cerebellar involvement in the PD pathophysiological mechanisms (Lewis et al., 2013;Wu and Hallett, 2013;Yoo et al., 2019). Furthermore, the relevant role of the cerebellum is also strongly indicated in psychiatric disorders such as SCZ and ASD characterized by a significant dysregulation of the dopaminergic system (Andreasen et al., 1998;Strick et al., 2009;Mosconi et al., 2015).

THE ROLE OF THE DOPAMINERGIC CEREBELLAR SYSTEM IN THE TREATMENT OF NEUROLOGIC AND PSYCHIATRIC DOPAMINE-RELATED DISORDERS
Taken together, the data evidenced in the present review, suggested the existence of a cerebellar dopaminergic neuronal system, which can be the target for pharmacological, nonpharmacological, or combined therapeutic treatments (Miterko et al., 2019); here, we will briefly review some of the therapeutic aspects on the cerebellar dopaminergic system in PD, SCZ, and ASD.
In PD, neuroimaging studies have demonstrated L-DOPA administration resulted involved in asymmetrical effects in motor brain regions, highlighting differences in cerebellar activity (Martinu et al., 2014). In PD patients, an increased putamencerebellar activity after abstention of L-DOPA administration has been proven, suggesting a role for the cerebellum in compensatory mechanisms (Simioni et al., 2015).
In SCZ antipsychotic treatments, the cerebellum may also represent part of the pharmacologic target. In rat cerebellum, the atypical antipsychotic blonaserin and the anxiolytic buspirone engage extensively in dopamine receptor DRD 3 (Baba et al., 2015;Di Ciano et al., 2017); indeed, in the cerebellum an extensive distribution of the dopamine receptor DRD 3 has been demonstrated ( Barili et al., 2000;Kim et al., 2009). Furthermore, in genomic DNA isolated from the cerebellum, the atypical antipsychotic agent olanzapine increased methylation of genes related to the dopaminergic system, such as DRD 5 , DOPA decarboxylase (DDC8), and VMAT 2 (SCL18A2/VMAT2; Melka et al., 2013).
The cerebellum is extensively interconnected to the other brain regions involved in motor, cognitive, and affective functions (Milardi et al., 2016;Cacciola et al., 2017aCacciola et al., , 2019Caligiore et al., 2017;Bostan and Strick, 2018;Flace et al., 2018b). Although, these cerebellar interconnections have not yet been fully characterized, in studies, it has been demonstrated that the cerebellum may represent the ideal target of noninvasive brain stimulation therapies such as electrical or magnetic stimulations applied in therapies for neurological and psychiatric disorders (van Dun et al., 2017;Miterko et al., 2019;Quartarone et al., 2020). In PD patients, bilateral cerebellar repetitive Transcranial Magnetic Stimulation (rTMS) induced persistent clinical beneficial effects, reducing peak-dose L-DOPA-induced dyskinesia (Koch, 2010).
In healthy subjects, cerebellar vermal theta burst stimulation (TBS) produced downstream changes in neuronal activity in the frontal cortex (Schutter et al., 2003), and pharmacological treatment-resistant SCZ patients can improved cognitive functions (Demirtas-Tatlidede et al., 2010). The rTMS In ASD has been used to study excitatory/inhibitory imbalance (Uzunova et al., 2016) and can represent an innovative therapeutic approach for reducing some of the core and associated ASD symptoms (Oberman et al., 2016).
Finally, we suggest that the cerebellar dopaminergic system and its interconnections may represent an ideal candidate for innovative non-invasive treatments such as electrical or magnetic stimulations in neurological and psychiatric disorders (Demirtas-Tatlidede et al., 2010;Koch, 2010;Oberman et al., 2016;Miterko et al., 2019;Quartarone et al., 2020). These innovative therapeutic objectives constitute relevant elements of study and we hope that they can be achieved in a relatively short time.

AUTHOR CONTRIBUTIONS
PF designed the study, performed the experiments and the analysis of the experimental data, and participated in the writing of the manuscript. PL and DG shared the study project and participated in the writing of the manuscript. GB, ABi, SBe, JB, SBi, and ABr participated in the writing of the manuscript. GG, MG, and GA performed the analysis of the experimental data, participated in the writing of the manuscript. All authors contributed to the article and approved the submitted version.

FUNDING
We are particularly grateful to Maria Antonia Sironi and Graziano Benfenati, a person affected by Parkinson's disease; as advocates of free and independent scientific research, who have funded and permitted the publication of this review.