Anatomical Inputs From the Sensory and Value Structures to the Tail of the Rat Striatum

Abstract

The caudal region of the rodent striatum, called the tail of the striatum (TS), is a relatively small area but might have a distinct function from other striatal subregions. Recent primate studies showed that this part of the striatum has a unique function in encoding long-term value memory of visual objects for habitual behavior. This function might be due to its specific connectivity. We identified inputs to the rat TS and compared those with inputs to the dorsomedial striatum (DMS) in the same animals. The TS directly received anatomical inputs from both sensory structures and value-coding regions, but the DMS did not. First, inputs from the sensory cortex and sensory thalamus to the TS were found; visual, auditory, somatosensory and gustatory cortex and thalamus projected to the TS but not to the DMS. Second, two value systems innervated the TS; dopamine and serotonin neurons in the lateral part of the substantia nigra pars compacta (SNc) and dorsal raphe nucleus projected to the TS, respectively. The DMS received inputs from the separate group of dopamine neurons in the medial part of the SNc. In addition, learning-related regions of the limbic system innervated the TS; the temporal areas and the basolateral amygdala selectively innervated the TS, but not the DMS. Our data showed that both sensory and value-processing structures innervated the TS, suggesting its plausible role in value-guided sensory-motor association for habitual behavior.

Keywords: dorsomedial striatum; habitual behavior; rostral-caudal axis; sensory input; tail of striatum; value input.

Figure 1

Injection sites of retrograde tracers in the tail of the striatum (TS) and dorsomedial striatum (DMS). (A) Coronal slices of Nissl-stained striatum in the rostral-caudal axis. Striatum outlines are indicated by black lines filled with blue. TS, tail of striatum; Amy, amygdala; Thal, thalamus. (B) Target coordinates of the injection site in the striatum. The striatum of rat #4 was reconstructed in the dorsal and lateral views, and coordination of the TS was marked by a cross based on the rat brain atlas (Paxinos and Watson, 1997). (C) Coronal view of the TS and DMS showing the red and green retrobead injection sites respectively. Fluorescent signals of red and green retrobeads were overlaid on Nissl-stained brain slices. Red and black arrows indicate the injection sites.

Figure 2

Cortical, subcortical and brain stem projections to the TS and DMS. (A) Dorsal view of the striatum showing red and green retrobead injection sites in the TS and DMS of rat #4. (B) Retrogradely labeled neurons from the TS injection site in a coronal brain section. The fluorescent signals were mostly found in the soma of neurons (right). These labeled neurons are marked as red dots and overlaid on a Nissl-stained brain section (left). V2M, medial part of the secondaryvisual cortex; V1, primary visual cortex; V2L, lateral part of the secondary visual cortex; A2, secondary auditory cortex; A1, primary auditory cortex; TeA, temporal association cortex; PRh, perirhinal cortex; Ent, entorhinal cortex. (C) TS-projecting neurons on coronal slices across the rostral-caudal axis. (D) Distributions of TS- and DMS-projecting neurons at 400-μm intervals in the rostral-caudal axis (TS-projecting neurons, n = 6 hemispheres; DMS-projecting neurons, n = 4 hemispheres). (E–G) Proportions of TS-projecting neurons in the cortical, subcortical, and brain stem regions (n = 6 hemispheres). TS-projecting neurons were mainly found in the midbrain, at ventral to the cerebral aqueduct, and dorsal to the medial longitudinal fasciculus, which may be the parvicellular part of the oculomotor nucleus (putative 3PC). PBP, parabrachial pigmented nucleus of the VTA; MT, medial terminal nucleus of the accessory optic tract. ***p < 0.001.

Figure 3

The TS receives cortical and thalamic visual inputs: visual cortex and dorsolateral geniculate nucleus (DLG) projections. (A) TS-projecting neurons in the rostral, middle and caudal parts of the visual cortex. V1, primary visual cortex; V2L, lateral part of the secondary visual cortex; V2M, medial part of the secondary visual cortex. (B) DMS-projecting neurons were found in the visual cortex. Scale bars: 1 mm. (C) An example of Nissl-stained slice showing the location of TS-projecting neurons in the V2M layers (Rat #2, left side of the brain). Scale bar: 500 μm. (D) Rostral-caudal distribution of TS-projecting neurons at 400-μm intervals in the subregions of the visual cortex (n = 6 hemispheres). (E) Rostral-caudal distribution of TS-projecting neurons at 400-μm intervals in the DLG (n = 6 hemispheres). (F) TS-projecting neurons in the rostral, middle and caudal parts of the DLG. (G) No DMS-projecting neurons were found in the DLG. Scale bars: 1 mm. **p < 0.01,***p < 0.001.

Figure 4

SN and DRN inputs to the TS and DMS. (A) Retrogradely labeled neurons in and around the SN (black dotted line). TS-projecting neurons (red dots) were located in the lateral region of the SN, mostly in the substantia nigra pars lateralis (SNl). DMS-projecting neurons (green dots) were mostly found in the medial region of the SN. SNr, substantia nigra pars reticulata; SNc, Substantia nigra pars compacta. Scale bar: 1 mm. (B) Rostral-caudal distributions of TS- and DMS-projecting neurons at 400-μm intervals in the SN (n = 6 hemispheres and n = 4 hemispheres, respectively). **p < 0.01, ***p < 0.001. (C,D) TS- and DMS-projecting neurons in the SN were dopaminergic. Signals of retrogradely labeled neurons (left) colocalized with tyrosine hydroxylase (TH)-positive signals (middle). Merged images (right) show double-labeled neurons (arrow heads in example). Blue is pseudocolor for TH-positive cells. Scale bars: 25 μm. (E) TS- projecting neurons in the DRN. DMS-projecting neurons were not found in the slice. The green-labeled area indicates the DRN. DRN, dorsal raphe nucleus. Scale bar: 1 mm. (F) Rostral-caudal distribution of TS-projecting neurons at 400-μm intervals in the DRN (n = 6 hemispheres). **p < 0.01. (G) Serotonin-positive signals (right) were overlaid on the Nissl-stained brain stem (left). MRN, median raphe nucleus. Scale bars: 500 μm. (H) TS-projecting neurons in the DRN were serotonergic. Retrogradely labeled signals (left) were colocalized with serotonin-positive signals (middle). Merged images (right) show double-labeled neurons (arrow heads). Scale bar: 25 μm. **p < 0.01, ***p < 0.001.

Figure 5

The TS receives cortical and thalamic auditory inputs: auditory cortex and medial geniculate body (MGB) projections. (A) TS-projecting neurons in the rostral, middle, and caudal parts of the auditory cortex. A1, primary auditory cortex; A2, secondary auditory cortex. (B) DMS-projecting neurons were not found in the auditory cortex. (C) Rostral-caudal distribution of TS-projecting neurons at 400-μm intervals in the subregions of the auditory cortex (left) and proportions of projecting neurons in the cortex (right; TS-projecting neurons, n = 6 hemispheres; DMS-projecting neurons, n = 4 hemispheres). (D) Rostral-caudal distribution of TS-projecting neurons at 400-μm intervals in the MGB (left) and their proportions in the MGB subregions (right; TS-projecting neurons, n = 6 hemispheres). MGB, medial geniculate body; MGV, ventral part of the MGB; MGD, dorsal part of the MGB; MGM, medial part of the MGB. (E) TS-projecting neurons in the rostral, middle, and caudal parts of the MGB. (F) No DMS-projecting neurons were found in the MGB. Scale bars: 1 mm. **p < 0.01,***p < 0.001.

Figure 6

Somatosensory, gustatory and olfactory inputs to the TS and DMS from cortical and thalamic regions. (A) TS- and DMS-projecting neurons (red and green dots, respectively) in the somatosensory cortex. S1, primary somatosensory cortex; S2, secondary somatosensory cortex. (B) Rostral-caudal distributions of TS- (red line) and DMS-projecting (green line) neurons at 400-μm intervals in the somatosensory cortex (left) and proportions of projecting neurons in the cortex (right; TS-projecting neurons, n = 6 hemispheres; DMS-projecting neurons, n = 4 hemispheres). (C,D) The same format as in (A,B), showing projections from the endopiriform cortex (En). (E,F) The same format as in (A,B), showing inputs from the insular cortex. GI, granular insular cortex; DI, dysgranular insular cortex; AID, dorsal part of the agranular insular cortex; AIV, ventral part of the agranular insular cortex; AIP, posterior part of the agranular insular cortex. (G,H) The same format as in (A,B), showing projections from the thalamus. VL, ventrolateral thalamic nucleus; AV, anteroventral thalamic nucleus; LPLR, lateral part of the lateral posterior thalamic nucleus; PC, paracentral thalamic nucleus; Po, posterior thalamic nuclear group; VPM, ventral posteromedial thalamic nucleus (medial part of the VPN); VPL, ventral posterolateral thalamic nucleus (lateral part of the VPN). Scale bars: 1 mm. **p < 0.01, ***p < 0.001.

Figure 7

Limbic inputs to the TS and DMS. (A) TS- and DMS-projecting neurons (red and green dots, respectively) in the medial prefrontal cortex (mPFC; left) and proportions in the cortex (top-right). Rostral-caudal distributions of TS- (red line) and DMS-projecting (green line) neurons at 400-μm intervals in the mPFC cortex (bottom-right; TS-projecting neurons, n = 6 hemispheres; DMS-projecting neurons, n = 4 hemispheres). PrL, prelimbic cortex; IL, infralimbic cortex. (B) The same format as in (A) showing cingulate cortex (Cg) inputs. (C) The same format as in (A), showing perirhinal cortex (PRh) inputs. (D) The same format as in (A), showing entorhinal cortex (Ent) inputs. (E) The same format as in (A), showing temporal association cortex (TeA) inputs. (F) The same format as in (A), showing amygdala input proportions in the subcortex. BLA, anterior part of the basolateral amygdaloid nucleus; BLP, posterior part of the basolateral amygdaloid nucleus; LaV, ventral part of the lateral amygdaloid nucleus; LaD, dorsal part of the lateral amygdaloid nucleus. Scale bars: 1 mm. **p < 0.01, ***p < 0.001.

Figure 8

Inputs from the sensory associative structures, motor cortex and orbitofrontal cortex (OFC) to the TS and DMS. (A) TS- and DMS-projecting neurons (red and green dots, respectively) in the claustrum (Cl; left) and their proportions in the subcortex (top-right). Rostral-caudal distributions of TS- (red line) and DMS-projecting (green line) neurons at 400-μm intervals in the Cl (bottom-right; TS-projecting neurons, n = 6 hemispheres; DMS-projecting neurons, n = 4 hemispheres). (B) The same format as in (A), showing parietal association cortex (PtA) inputs. (C) The same format as in (A), showing motor cortex inputs. M1, primary motor cortex; M2, secondary motor cortex. (D) The same format as in (A), showing OFC inputs. Scale bars: 1 mm. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 9

A model of habit learning in the TS by sensory and value inputs. The rat TS directly receives sensory and value inputs from the cortex, thalamus and brain stem. This anatomical convergence of sensory and value inputs in the TS allows for learning of a sensory-response association for habitual behavior. Furthermore, the thalamic sensory inputs might produce a stimulus-induced quick response after long-term learning.