Association between vascular endothelial growth factor-mediated blood-brain barrier dysfunction and stress-induced depression.

Several lines of evidence suggest that stress induces the neurovascular dysfunction associated with increased blood-brain barrier (BBB) permeability, which could be an important pathology linking stress and psychiatric disorders, including major depressive disorder (MDD). However, the detailed mechanism resulting in BBB dysfunction associated in the pathophysiology of MDD still remains unclear. Herein, we demonstrate the role of vascular endothelial growth factor (VEGF), a key mediator of vascular angiogenesis and BBB permeability, in stress-induced BBB dysfunction and depressive-like behavior development. We implemented an animal model of depression, chronic restraint stress (RS) in BALB/c mice, and found that the BBB permeability was significantly increased in chronically stressed mice. Immunohistochemical and electron microscopic observations revealed that increased BBB permeability was associated with both paracellular and transcellular barrier alterations in the brain endothelial cells. Pharmacological inhibition of VEGF receptor 2 (VEGFR2) using a specific monoclonal antibody (DC101) prevented chronic RS-induced BBB permeability and anhedonic behavior. Considered together, these results indicate that VEGF/VEGFR2 plays a crucial role in the pathogenesis of depression by increasing the BBB permeability, and suggest that VEGFR2 inhibition could be a potential therapeutic strategy for the MDD subtype associated with BBB dysfunction.

Dimethylarginine dimethylaminohydrolase 1 as a novel regulator of oligodendrocyte differentiation in the central nervous system remyelination.

Remyelination is a regenerative process that restores the lost neurological function and partially depends on oligodendrocyte differentiation. Differentiation of oligodendrocytes spontaneously occurs after demyelination, depending on the cell intrinsic mechanisms. By combining a loss-of-function genomic screen with a web-resource-based candidate gene identification approach, we identified that dimethylarginine dimethylaminohydrolase 1 (DDAH1) is a novel regulator of oligodendrocyte differentiation. Silencing DDAH1 in oligodendrocytes prevented the expression of myelin basic protein in mouse oligodendrocyte culture with the change in expression of genes annotated with oligodendrocyte development. DDAH1 inhibition attenuated spontaneous remyelination in a cuprizone-induced demyelinated mouse model. Conversely, increased DDAH1 expression enhanced remyelination capacity in experimental autoimmune encephalomyelitis. These results provide a novel therapeutic option for demyelinating diseases by modulating DDAH1 activity.

Computing Social Value Conversion in the Human Brain.

Social signals play powerful roles in shaping self-oriented reward valuation and decision making. These signals activate social and valuation/decision areas, but the core computation for their integration into the self-oriented decision machinery remains unclear. Here, we study how a fundamental social signal, social value (others' reward value), is converted into self-oriented decision making in the human brain. Using behavioral analysis, modeling, and neuroimaging, we show three-stage processing of social value conversion from the offer to the effective value and then to the final decision value. First, a value of others' bonus on offer, called offered value, was encoded uniquely in the right temporoparietal junction (rTPJ) and also in the left dorsolateral prefrontal cortex (ldlPFC), which is commonly activated by offered self-bonus value. The effective value, an intermediate value representing the effective influence of the offer on the decision, was represented in the right anterior insula (rAI), and the final decision value was encoded in the medial prefrontal cortex (mPFC). Second, using psychophysiological interaction and dynamic causal modeling analyses, we demonstrated three-stage feedforward processing from the rTPJ and ldPFC to the rAI and then from rAI to the mPFC. Further, we showed that these characteristics of social conversion underlie distinct sociobehavioral phenotypes. We demonstrate that the variability in the conversion underlies the difference between prosocial and selfish subjects, as seen from the differential strength of the rAI and ldlPFC coupling to the mPFC responses, respectively. Together, these findings identified fundamental neural computation processes for social value conversion underlying complex social decision making behaviors.SIGNIFICANCE STATEMENT In daily life, we make decisions based on self-interest, but also in consideration for others' status. These social influences modulate valuation and decision signals in the brain, suggesting a fundamental process called value conversion that translates social information into self-referenced decisions. However, little is known about the conversion process and its underlying brain mechanisms. We investigated value conversion using human fMRI with computational modeling and found three essential stages in a progressive brain circuit from social to empathic and decision areas. Interestingly, the brain mechanism of conversion differed between prosocial and individualistic subjects. These findings reveal how the brain processes and merges social information into the elemental flow of self-interested decision making.

GABA storage and release in the medial globus pallidus in L-DOPA-induced dyskinesia priming.

Levo-dihydroxyphenylalanine (L-DOPA) is the most effective treatment for Parkinson's disease; however, most patients develop uncontrollable abnormal involuntary movements known as L-DOPA-induced dyskinesia. L-DOPA-induced dyskinesia can be reduced by pallidotomy of the medial globus pallidus or pallidal deep brain stimulation, suggesting that the medial globus pallidus plays a significant role in the development of L-DOPA-induced dyskinesia. In the present study, the pathological changes of the medial globus pallidus in L-DOPA-induced dyskinesia were studied in rat models of Parkinson's disease (unilateral 6-hydroxydopamine lesioning) and L-DOPA-induced dyskinesia (L-DOPA injection in Parkinson's disease-model rats twice daily for 2 weeks, confirmed by display of dyskinesia-like abnormal involuntary movements). L-DOPA-induced dyskinesia-model rats displayed medial globus pallidus hypertrophy, enlarged axon terminals surrounding the dendrites of medial globus pallidus neurons, and increased density of synaptic vesicles in enlarged axon terminals on the lesioned side. Synaptic terminal enlargement reversed after discontinuation of L-DOPA. Histological studies revealed the enlarged synaptic terminals were those of GABAergic striatal (direct pathway) neurons. A single injection of L-DOPA enhanced GABA release in the medial globus pallidus on the lesioned side in L-DOPA-induced dyskinesia-model rats compared to Parkinson's disease-model rats. In addition, microinjection of muscimol, a GABAA receptor agonist, into the medial globus pallidus on the lesioned side of Parkinson's disease-model rats induced dyskinesia-like abnormal involuntary movements. Microinjection of bicuculline, a GABAA receptor antagonist, into the medial globus pallidus on the lesioned side alleviated L-DOPA-induced dyskinesia in Parkinson's disease-model rats that had received L-DOPA prior to the microinjection. These results indicate that priming for L-DOPA-induced dyskinesia comprises excessive GABA storage in axon terminals of the direct pathway and that expression of L-DOPA-induced dyskinesia is associated with enhanced GABA release into the medial globus pallidus after L-DOPA dosing and the resultant excessive stimulation of GABAA receptors.

Visualization by P-flow: gradient- and feature-based optical flow and vector fields extracted from image analysis.

We proposed a method for extracting the optical flow suitable for visualization, pseudo-flow (P-flow), from a natural movie [Exp. Brain Res.237, 3321 (2019)EXBRAP0014-481910.1007/s00221-019-05674-0]. The P-flow algorithm comprises two stages: (1) extraction of a local motion vector field from two successive frames and (2) tracking of vectors between two successive frame pairs. In this study, we show that while P-flow takes a feature (vector) tracking approach, it is also classified as a gradient-based approach that satisfies the brightness constancy constraint. We also incorporate interpolation and a corner detector to address the shortcomings associated with the two approaches.

Abnormal axon guidance signals and reduced interhemispheric connection via anterior commissure in neonates of marmoset ASD model.

In autism spectrum disorder (ASD), disrupted functional and structural connectivity in the social brain has been suggested as the core biological mechanism underlying the social recognition deficits of this neurodevelopmental disorder. In this study, we aimed to identify genetic and neurostructural abnormalities at birth in a non-human primate model of ASD, the common marmoset with maternal exposure to valproic acid (VPA), which has been reported to display social recognition deficit in adulthood. Using a comprehensive gene expression analysis, we found that 20 genes were significantly downregulated in VPA-exposed neonates. Of these, Frizzled3 (FZD3) and PIK3CA were identified in an axon guidance signaling pathway. FZD3 is essential for the normal development of the anterior commissure (AC) and corpus callosum (CC); hence, we performed diffusion tensor magnetic resonance imaging with a 7-Tesla scanner to measure the midsagittal sizes of these structures. We found that the AC size in VPA-exposed neonates was significantly smaller than that in age-matched controls, while the CC size did not differ. These results suggest that downregulation of the genes related to axon guidance and decreased AC size in neonatal primates may be linked to social brain dysfunctions that can happen later in life.

Vection induced by low-level motion extracted from complex animation films.

This study examined the contributions of low-, mid- and high-level visual motion information to vection. We compared the vection experiences induced by hand-drawn and computer-generated animation clips to those induced by versions of these movies that contained only their pure optic flow. While the original movies were found to induce longer and stronger vection experiences than the pure optic flow, vection onsets were not significantly altered by removing the mid- and high-level information. We conclude that low-level visual motion information appears to be important for vection induction, whereas mid- and higher-level display information appears to be important for sustaining and strengthening this vection after its initial induction.

Fast voltage-sensitive dye imaging of excitatory and inhibitory synaptic transmission in the rat granular retrosplenial cortex.

Rodent granular retrosplenial cortex (GRS) has dense connections between the anterior thalamic nuclei (ATN) and hippocampal formation. GRS superficial pyramidal neurons exhibit distinctive late spiking (LS) firing property and form patchy clusters with prominent apical dendritic bundles. The aim of this study was to investigate spatiotemporal dynamics of signal transduction in the GRS induced by ATN afferent stimulation by using fast voltage-sensitive dye imaging in rat brain slices. In coronal slices, layer 1a stimulation, which presumably activated thalamic fibers, evoked propagation of excitatory synaptic signals from layers 2-4 to layers 5-6 in a direction perpendicular to the layer axis, followed by transverse signal propagation within each layer. In the presence of ionotropic glutamate receptor antagonists, inhibitory responses were observed in superficial layers, induced by direct activation of inhibitory interneurons in layer 1. In horizontal slices, excitatory signals in deep layers propagated transversely mainly from posterior to anterior via superficial layers. Cortical inhibitory responses upon layer 1a stimulation in horizontal slices were weaker than those in the coronal slices. Observed differences between coronal and horizontal planes suggest anisotropy of the intracortical circuitry. In conclusion, ATN inputs are processed differently in coronal and horizontal planes of the GRS and then conveyed to other cortical areas. In both planes, GRS superficial layers play an important role in signal propagation, which suggests that superficial neuronal cascade is crucial in the integration of multiple information sources.NEW & NOTEWORTHY Superficial neurons in the rat granular retrosplenial cortex (GRS) show distinctive late-spiking (LS) firing property. However, little is known about spatiotemporal dynamics of signal transduction in the GRS. We demonstrated LS neuron network relaying thalamic inputs to deep layers and anisotropic distribution of inhibition between coronal and horizontal planes. Since deep layers of the GRS receive inputs from the subiculum, GRS circuits may work as an integrator of multiple sources such as sensory and memory information.

Developmental genetic profiles of glutamate receptor system, neuromodulator system, protector of normal tissue and mitochondria, and reelin in marmoset cortex: potential molecular mechanisms of pruning phase of spines in primate synaptic formation process during the end of infancy and prepuberty (II).

This is the second report of a series paper, which reports molecular mechanisms underlying the occurrence of pruning spine phase after rapid spinogenesis phase in neonates and young infant in the primate brain. We performed microarray analysis between the peak of spine numbers [postnatal 3 months (M)] and spine pruning (postnatal 6M) in prefrontal, inferior temporal, and primary visual cortices of the common marmoset (Callithrix jacchus). The pruning phase is not clearly defined in rodents but is in primates including the marmoset. The differentially expressed genes between 3M and 6M in all three cortical areas were selected by two-way analysis of variance. The list of selected genes was analyzed by canonical pathway analysis using "Ingenuity Pathway Analysis of complex omics data" (IPA; Ingenuity Systems, Qiagen, Hilden, Germany). In this report, we discuss these lists of genes for the glutamate receptor system, G-protein-coupled neuromodulator system, protector of normal tissue and mitochondria, and reelin. (1) Glutamate is a common neurotransmitter. Its receptors AMPA1, GRIK1, and their scaffold protein DLG4 decreased as spine numbers decreased. Instead, GRIN3 (NMDA receptor) increased, suggesting that strong NMDA excitatory currents may be required for a single neuron to receive sufficient net synaptic activity in order to compensate for the decrease in synapse. (2) Most of the G protein-coupled receptor genes (e.g., ADRA1D, HTR2A, HTR4, and DRD1) in the selected list were upregulated at 6M. The downstream gene ROCK2 in these receptor systems plays a role of decreasing synapses, and ROCK2 decreased at 6M. (3) Synaptic phagosytosis by microglia with complement and other cytokines could cause damage to normal tissue and mitochondria. SOD1, XIAP, CD46, and CD55, which play protective roles in normal tissue and mitochondria, showed higher expression at 6M than at 3M, suggesting that normal brain tissue is more protected at 6M. (4) Reelin has an important role in cortical layer formation. In addition, RELN and three different pathways of reelin were expressed at 6M, suggesting that new synapse formation decreased at that age. Moreover, if new synapses were formed, their positions were free and probably dependent on activity.

Developmental expression profiles of axon guidance signaling and the immune system in the marmoset cortex: potential molecular mechanisms of pruning of dendritic spines during primate synapse formation in late infancy and prepuberty (I).

The synapse number and the related dendritic spine number in the cerebral cortex of primates shows a rapid increase after birth. Depending on the brain region and species, the number of synapses reaches a peak before adulthood, and pruning takes place after this peak (overshoot-type synaptic formation). Human mental disorders, such as autism and schizophrenia, are hypothesized to be a result of either too weak or excessive pruning after the peak is reached. Thus, it is important to study the molecular mechanisms underlying overshoot-type synaptic formation, particularly the pruning phase. To examine the molecular mechanisms, we used common marmosets (Callithrix jacchus). Microarray analysis of the marmoset cortex was performed in the ventrolateral prefrontal, inferior temporal, and primary visual cortices, where changes in the number of dendritic spines have been observed. The spine number of all the brain regions above showed a peak at 3 months (3 M) after birth and gradually decreased (e.g., at 6 M and in adults). In this study, we focused on genes that showed differential expression between ages of 3 M and 6 M and on the differences whose fold change (FC) was greater than 1.2. The selected genes were subjected to canonical pathway analysis, and in this study, we describe axon guidance signaling, which had high plausibility. The results showed a large number of genes belonging to subsystems within the axon guidance signaling pathway, macrophages/immune system, glutamate system, and others. We divided the data and discussion of these results into 2 papers, and this is the first paper, which deals with the axon guidance signaling and macrophage/immune system. Other systems will be described in the next paper. Many components of subsystems within the axon guidance signaling underwent changes in gene expression from 3 M to 6 M so that the synapse/dendritic spine number would decrease at 6 M. Thus, axon guidance signaling probably contributes to the decrease in synapse/dendritic spine number at 6 M, the phenomenon that fits the overshoot-type synaptic formation in primates. Microglial activity (evaluated by quantifying AIF1 expression) and gene expression of molecules that modulate microglia, decreased at 6 M, just like the synapse/dendritic spine number. Thus, although microglial activity is believed to be related to phagocytosis of synapses/dendritic spines, microglial activity alone cannot explain how pruning was accelerated in the pruning phase. On the other hand, expression of molecules that tag synapses/dendritic spines as a target of phagocytosis by microglia (e.g., complement components) increased at 6 M, suggesting that these tagging proteins may be involved in the acceleration of pruning during the pruning phase.

Morphologic changes of dendritic spines of striatal neurons in the levodopa-induced dyskinesia model.

Maladaptive plasticity at corticostriatal synapses plays an important role in the development of levodopa-induced dyskinesia. Recently, it has been shown that synaptic plasticity is closely linked to morphologic changes of dendritic spines. To evaluate morphologic changes of dendritic spines of two types of striatal medium spiny neurons, which project to the internal segment of globus pallidus or the external segment of globus pallidus, in the levodopa-induced dyskinesia model, we used 6-hydroxydopamine-lesioned rats chronically treated with levodopa. Dendritic spines were decreased and became enlarged in the direct pathway neurons of the model of levodopa-induced dyskinesia. The same levodopa treatment to normal rats, in which no dyskinesia was observed, also induced enlargement of dendritic spines, but not a decrease in density of spines in the direct pathway neurons. These results suggest that a loss and enlargement of dendritic spines in the direct pathway neurons plays important roles in the development of levodopa-induced dyskinesia.

Marmoset monkeys evaluate third-party reciprocity.

Many non-human primates have been observed to reciprocate and to understand reciprocity in one-to-one social exchanges. A recent study demonstrated that capuchin monkeys are sensitive to both third-party reciprocity and violation of reciprocity; however, whether this sensitivity is a function of general intelligence, evidenced by their larger brain size relative to other primates, remains unclear. We hypothesized that highly pro-social primates, even with a relatively smaller brain, would be sensitive to others' reciprocity. Here, we show that common marmosets discriminated between human actors who reciprocated in social exchanges with others and those who did not. Monkeys accepted rewards less frequently from non-reciprocators than they did from reciprocators when the non-reciprocators had retained all food items, but they accepted rewards from both actors equally when they had observed reciprocal exchange between the actors. These results suggest that mechanisms to detect unfair reciprocity in third-party social exchanges do not require domain-general higher cognitive ability based on proportionally larger brains, but rather emerge from the cooperative and pro-social tendencies of species, and thereby suggest this ability evolved in multiple primate lineages.

Erroneous predictive coding across brain hierarchies in a non-human primate model of autism spectrum disorder.

In autism spectrum disorder (ASD), atypical sensory experiences are often associated with irregularities in predictive coding, which proposes that the brain creates hierarchical sensory models via a bidirectional process of predictions and prediction errors. However, it remains unclear how these irregularities manifest across different functional hierarchies in the brain. To address this, we study a marmoset model of ASD induced by valproic acid (VPA) treatment. We record high-density electrocorticography (ECoG) during an auditory task with two layers of temporal control, and applied a quantitative model to quantify the integrity of predictive coding across two distinct hierarchies. Our results demonstrate a persistent pattern of sensory hypersensitivity and unstable predictions across two brain hierarchies in VPA-treated animals, and reveal the associated spatio-spectro-temporal neural signatures. Despite the regular occurrence of imprecise predictions in VPA-treated animals, we observe diverse configurations of underestimation or overestimation of sensory regularities within the hierarchies. Our results demonstrate the coexistence of the two primary Bayesian accounts of ASD: overly-precise sensory observations and weak prior beliefs, and offer a potential multi-layered biomarker for ASD, which could enhance our understanding of its diverse symptoms.

Altered projection-specific synaptic remodeling and its modification by oxytocin in an idiopathic autism marmoset model.

Alterations in the experience-dependent and autonomous elaboration of neural circuits are assumed to underlie autism spectrum disorder (ASD), though it is unclear what synaptic traits are responsible. Here, utilizing a valproic acid-induced ASD marmoset model, which shares common molecular features with idiopathic ASD, we investigate changes in the structural dynamics of tuft dendrites of upper-layer pyramidal neurons and adjacent axons in the dorsomedial prefrontal cortex through two-photon microscopy. In model marmosets, dendritic spine turnover is upregulated, and spines are generated in clusters and survived more often than in control marmosets. Presynaptic boutons in local axons, but not in commissural long-range axons, demonstrate hyperdynamic turnover in model marmosets, suggesting alterations in projection-specific plasticity. Intriguingly, nasal oxytocin administration attenuates clustered spine emergence in model marmosets. Enhanced clustered spine generation, possibly unique to certain presynaptic partners, may be associated with ASD and be a potential therapeutic target.

Neuronal DSCAM regulates the peri-synaptic localization of GLAST in Bergmann glia for functional synapse formation.

In the central nervous system, astrocytes enable appropriate synapse function through glutamate clearance from the synaptic cleft; however, it remains unclear how astrocytic glutamate transporters function at peri-synaptic contact. Here, we report that Down syndrome cell adhesion molecule (DSCAM) in Purkinje cells controls synapse formation and function in the developing cerebellum. Dscam-mutant mice show defects in CF synapse translocation as is observed in loss of function mutations in the astrocytic glutamate transporter GLAST expressed in Bergmann glia. These mice show impaired glutamate clearance and the delocalization of GLAST away from the cleft of parallel fibre (PF) synapse. GLAST complexes with the extracellular domain of DSCAM. Riluzole, as an activator of GLAST-mediated uptake, rescues the proximal impairment in CF synapse formation in Purkinje cell-selective Dscam-deficient mice. DSCAM is required for motor learning, but not gross motor coordination. In conclusion, the intercellular association of synaptic and astrocyte proteins is important for synapse formation and function in neural transmission.

A reappraisal of the default mode and frontoparietal networks in the common marmoset brain.

In recent years the common marmoset homolog of the human default mode network (DMN) has been a hot topic of discussion in the marmoset research field. Previously, the posterior cingulate cortex regions (PGM, A19M) and posterior parietal cortex regions (LIP, MIP) were defined as the DMN, but some studies claim that these form the frontoparietal network (FPN). We restarted from a neuroanatomical point of view and identified two DMN candidates: Comp-A (which has been called both the DMN and FPN) and Comp-B. We performed GLM analysis on auditory task-fMRI and found Comp-B to be more appropriate as the DMN, and Comp-A as the FPN. Additionally, through fingerprint analysis, a DMN and FPN in the tasking human was closer to the resting common marmoset. The human DMN appears to have an advanced function that may be underdeveloped in the common marmoset brain.

In vitro myelination using explant culture of dorsal root ganglia: An efficient tool for analyzing peripheral nerve differentiation and disease modeling.

Peripheral nerves conducting motor and somatosensory signals in vertebrate consist of myelinated and unmyelinated axons. In vitro myelination culture, generated by co-culturing Schwann cells (SCs) and dorsal root ganglion (DRG) neurons, is an indispensable tool for modeling physiological and pathological conditions of the peripheral nervous system (PNS). This technique allows researchers to overexpress or downregulate molecules investigated in neurons or SCs to evaluate the effect of such molecules on myelination. In vitro myelination experiments are usually time-consuming and labor-intensive to perform. Here we report an optimized protocol for in vitro myelination using DRG explant culture. We found that our in vitro myelination using DRG explant (IVMDE) culture not only achieves myelination with higher efficiency than conventional in vitro myelination methods, but also can be used to observe Remak bundle and non-myelinating SCs, which were unrecognizable in conventional methods. Because of these characteristics, IVMDE may be useful in modeling PNS diseases, including Charcot Marie Tooth disease (CMT), in vitro. These results suggest that IVMDE may achieve a condition more similar to peripheral nerve myelination observed during physiological development.

Local and long-distance organization of prefrontal cortex circuits in the marmoset brain.

The prefrontal cortex (PFC) has dramatically expanded in primates, but its organization and interactions with other brain regions are only partially understood. We performed high-resolution connectomic mapping of the marmoset PFC and found two contrasting corticocortical and corticostriatal projection patterns: "patchy" projections that formed many columns of submillimeter scale in nearby and distant regions and "diffuse" projections that spread widely across the cortex and striatum. Parcellation-free analyses revealed representations of PFC gradients in these projections' local and global distribution patterns. We also demonstrated column-scale precision of reciprocal corticocortical connectivity, suggesting that PFC contains a mosaic of discrete columns. Diffuse projections showed considerable diversity in the laminar patterns of axonal spread. Altogether, these fine-grained analyses reveal important principles of local and long-distance PFC circuits in marmosets and provide insights into the functional organization of the primate brain.

Multi-modal brain magnetic resonance imaging database covering marmosets with a wide age range.

Magnetic resonance imaging (MRI) is a non-invasive neuroimaging technique that is useful for identifying normal developmental and aging processes and for data sharing. Marmosets have a relatively shorter life expectancy than other primates, including humans, because they grow and age faster. Therefore, the common marmoset model is effective in aging research. The current study investigated the aging process of the marmoset brain and provided an open MRI database of marmosets across a wide age range. The Brain/MINDS Marmoset Brain MRI Dataset contains brain MRI information from 216 marmosets ranging in age from 1 and 10 years. At the time of its release, it is the largest public dataset in the world. It also includes multi-contrast MRI images. In addition, 91 of 216 animals have corresponding high-resolution ex vivo MRI datasets. Our MRI database, available at the Brain/MINDS Data Portal, might help to understand the effects of various factors, such as age, sex, body size, and fixation, on the brain. It can also contribute to and accelerate brain science studies worldwide.

Reciprocal connectivity of identified color-processing modules in the monkey inferior temporal cortex.

The inferior temporal (IT) cortex is the last unimodal visual area in the ventral visual pathway and is essential for color discrimination. Recent imaging and electrophysiological studies have revealed the presence of several distinct patches of color-selective cells in the anterior IT cortex (AIT) and posterior IT cortex (PIT). To understand the neural machinery for color processing in the IT cortex, in the present study, we combined anatomical tracing methods with electrophysiological unit recordings to investigate the anatomical connections of identified clusters of color-selective cells in monkey IT cortex. We found that a color cluster in AIT received projections from a color cluster in PIT as well as from discrete clusters of cells in other occipitotemporal areas, in the superior temporal sulcus, and in prefrontal and parietal cortices. The distribution of the labeled cells in PIT closely corresponded with that of the physiologically identified color-selective cells in this region. Furthermore, retrograde tracer injections in the posterior color cluster resulted in labeled cells in the anterior cluster. Thus, temporal lobe color-processing modules form a reciprocally interconnected loop within a distributed network.