Sources of input to the rostromedial tegmental nucleus, ventral tegmental area, and lateral habenula compared: A study in rat

Abstract

Profound inhibitory control exerted on midbrain dopaminergic neurons by the lateral habenula (LHb), which has mainly excitatory outputs, is mediated by the GABAergic rostromedial tegmental nucleus (RMTg), which strongly innervates dopaminergic neurons in the ventral midbrain. Early reports indicated that the afferent connections of the RMTg, excepting its very strong LHb inputs, do not differ appreciably from those of the ventral tegmental area (VTA). Presumably, however, the RMTg contributes more to behavioral synthesis than to simply invert the valence of the excitatory signal coming from the LHb. Therefore, the present study was done to directly compare the inputs to the RMTg and VTA and, in deference to its substantial involvement with this circuitry, the LHb was also included in the comparison. Data indicated that, while the afferents of the RMTg, VTA, and LHb do originate within the same large pool of central nervous system (CNS) structures, each is also related to structures that project more strongly to it than to the others. The VTA gets robust input from ventral striatopallidum and extended amygdala, whereas RMTg biased inputs arise in structures with a more direct impact on motor function, such as deep layers of the contralateral superior colliculus, deep cerebellar and several brainstem nuclei, and, via a relay in the LHb, the entopeduncular nucleus. Input from the ventral pallidal-lateral preoptic-lateral hypothalamus continuum is strong in the RMTg and VTA and dominant in the LHb. Axon collateralization was also investigated, providing additional insights into the organization of the circuitry of this important triad of structures.

Keywords: inhibitory control; lateral habenula; rostromedial tegmental nucleus.

Figure 1

Photomicrographs illustrating representative cholera toxin β subunit (Ctβ) injection sites in the rostromedial tegmental nucleus (RMTg in A and B), ventral tegmental area (VTA in C) and lateral habenula (LHb in D). Outlines shown in the inset in B reflect the dense core and less dense periphery of the injection in relation to immunoreactivity against the μ-opioid receptor. See list for additional abbreviations. Scale bar: 1 mm.

Figure 2

Graphs depicting numbers of VTA-, RMTg-, and LHb-projecting neurons along the rostrocaudal axis of the brain. Shown are means± SEM of numbers of retrogradely labeled neurons per brain section on the ipsilateral (A) and contralateral (B) sides relative to the injection site in rats injected into the VTA (open boxes), RMTg (closed triangles), and LHb (open circles). Approximate rostral and caudal limits of a number of relevant brain structures are indicated by horizontal arrows at the top of panel B. Abbreviations (left to right): PrL: prelimbic cortex (cx); Acb: nucleus accumbens; VP: ventral pallidum; BST: bed nucleus stria terminals; LPO: lateral preoptic area; LHb: lateral habenula; LH: lateral hypothalamus; VTA: ventral tegmental area; SC: superior colliculus; PPTg: pedunculopontine tegmental nucleus; DCN: deep cerebellar nuclei. VTA vs. LHb: °p < 0.05, °°p < 0.01, °°°p < 0.001, °°°°p < 0.0001; VTA vs. RMTg: #p < 0.05, ##p < 0.01, ###p < 0.001, ####p < .0001; RMTg vs. LHb: *p < 0.05, **p < 0.01, *** p < 0.001, **** p < 0.0001. Results of the ANOVAs are given in Table 2.

Figure 3

Bar graph illustrating the means± SEM of numbers of retrogradely labeled neurons counted in structures throughout the brain following injections of cholera toxin β subunit (Ctβ) into the ventral tegmental area (VTA, green bars), rostromedial tegmental nucleus (RMTg, magenta bars) and lateral habenula (LHb, black bars). The structures under each subheading (A. - E.) are listed in approximately rostrocaudal order. * - IPAC is interstitial nucleus of the posterior limb of the anterior commissure. Vertical bars with asterisks indicate statistically significant differences with *, **, and *** indicating p < 0.05, p < 0.01 and p < 0.001, respectively (Tukey's posthoc tests). Results of ANOVAs are given in Table 3

Figure 3

Bar graph illustrating the means± SEM of numbers of retrogradely labeled neurons counted in structures throughout the brain following injections of cholera toxin β subunit (Ctβ) into the ventral tegmental area (VTA, green bars), rostromedial tegmental nucleus (RMTg, magenta bars) and lateral habenula (LHb, black bars). The structures under each subheading (A. - E.) are listed in approximately rostrocaudal order. * - IPAC is interstitial nucleus of the posterior limb of the anterior commissure. Vertical bars with asterisks indicate statistically significant differences with *, **, and *** indicating p < 0.05, p < 0.01 and p < 0.001, respectively (Tukey's posthoc tests). Results of ANOVAs are given in Table 3

Figure 4

Photomicrographs illustrating retrogradely labeled neurons in the accumbens (Acb in A-C), a heavily labeled structure lodged beneath the infralimbic cortex (outlined area to the right of IL in D-F) and the magnocelluar division of the hypothalamic paraventricular nucleus (PaHmc in G-I) following injections of cholera toxin β subunit (Ctβ) into the ventral tegmental area (VTA, panels A, D and G), rostromedial tegmental nucleus (RMTg, panels B, E and H) and lateral habenula (LHb, panels C, F and I). Note the density of retrogradely labeled neurons in the Acb (A), beneath the infralimbic cortex (D) and surrounding the PaHmc (G) is greatest after injection of tracer into the VTA (A, D and G), whereas it is densest in the juxta-paraventricular lateral hypothalamus (jPaLH in I) following tracer injections into the LHb (C, F and I). Insets in A, B and C are enlargements of the respective boxed areas in the photomicrographs. See list for additional abbreviations. Scale bar: 1 mm in A-C, 200 μm in D-F and 100 μm in G-I.

Figure 5

Photomicrographs illustrating retrograde labeling in the deep layers of the superior colliculus (SC) following injections of cholera toxin β subunit (Ctβ) into the rostromedial tegmental nucleus (RMTg in A), ventral tegmental area (VTA in B) and lateral habenula (LHb in C). Insets are enlargements of the boxed areas in the respective micrographs. Tracer was injected into the brain to the right of the midline. Note following injection of tracer into the RMTg that retrograde labeling is distributed almost exclusively contralateral to the RMTg injection (asterisk in A). Substantially less dense labeling was distributed bilaterally but with an ipsilateral bias following tracer injection into the VTA (double asterisks in B). Following LHb injections, labeling in the SC was almost entirely ipsilateral and, although there were moderate numbers of labeled neurons, they were very sparsely labeled with immunoperoxidase reaction product, such that higher magnification was necessary to visualize them (inset in C). See list for additional abbreviations. Scale bar: 1 mm.

Figure 6

Photomicrographs illustrating retrograde labeling in the transition area between the lateral preoptic area (LPO) and lateral hypothalamus (LH) following injections of Ctβ into the lateral habenula (LHb, A and B), rostromedial tegmental nucleus (RMTg, C and D) and ventral tegmental area (VTA, E and F). Insets in B, D and F show the respective injection sites. Note after tracer injections into the LHb (A), that an arciform, dorsmedial to ventolateral oriented slab of densely packed retrogradely labeled neurons parallels the curved lateral margin of the anterior hypothalamic area (AHA). The somatodendritic architecture of many of the densely labeled neurons in that aggregation exhibits a similar dorsomedial to ventrolateral tilt (B). Note also that robust labeling is present in the juxta-paraventricular lateral hypothalamus (jPaLH) and moderate numbers of labeled neurons occupy the AHA following LHb injections (but not statistically different). In contrast, only moderate numbers of neurons were present in the LPO-LH transition after injections of tracer into the RMTg (C and D) and VTA (E and F) and the general shape of the labeling pattern is quite different from that observed after LHb injections of tracer as shown in A. The apparent density of labeling in the LPO-LH transition is enhanced following VTA injections due to the presence of many anterogradely labeled axons (E and F). Few neurons were present in the AHA following RMTg or VTA injections. The jPaLH contained moderate numbers of labeled neurons following VTA injections, as did the hypothalamic paraventricular nucleus, which exhibited few labeled neurons after LHb or RMTg injections. See list for additional abbreviations. Scale bar: 1 mm.

Figure 7

Frequency histogram (presented as a line graph) showing how much of the retrograde labeling (binned on the abscissa as percent of total labeling) was present in how many of structures (ordinate, total is 70, i.e., all the structures listed in Fig. 3) following injections of tracer into the ventral tegmental area (VTA - open squares), rostromedial tegmental nucleus (RMTg - closed triangles) and lateral habenula (LHb - open circles). Note that many of the structures had minimal labeling (less than 0.625% of total) following all of the injections, whereas most of the other structures were contained in the bins representing sparse to moderate labeling (e.g., 0.625-5%). A remaining small minority of structures, which, of note, were different for each of the three injection sites (see Tables 5 and 6) exhibited a substantial amount of labeling (e.g., 5% or more).

Figure 8

Micrographs showing single-labeled neurons (green and magenta) and double-labeled neurons (white arrows) in the lateral preoptic region following injections of (A) cholera toxin β subunit (Ctβ) into the VTA and Fluorogold (FG) into the RMTg (map is shown in Figure 9) and (B) FG into the LHb and Ctβ into the RMTg (map is shown in Figure 10). Scale bar: 100 μm.

Figure 9

Map showing several sections through the basal forebrain ordered rostrocaudally from A to D and illustrating retrogradely labeled neurons following injection of the retrograde tracers cholera toxin β subunit (Ctβ) into the ventral tegmental area (circles) and Fluorogold (FG) into the rostromedial tegmental nucleus (crosses). Neurons exhibiting both tracers (double-labeled) are shown as red dots. Note that the double-labeled neurons are not located preferentially in any particular structure (see also Fig. 10). Quantitative comparisons should not be made given the disparity in robustness of retrograde labeling after FG versus Ctβ injections. See list for abbreviations.

Figure 10

Map showing several sections through the basal forebrain ordered rostrocaudally from A to D and illustrating retrogradely labeled neurons following injection of the retrograde tracers cholera toxin β subunit into the rostromedial tegmental nucleus (circles) and Fluorogold into the lateral habenula (crosses). Neurons exhibiting both tracers (double-labeled) are shown as red dots. Note that the double-labeled neurons are located preferentially in the lateral preoptic area, which probably reflects the greater concentration of retrogradely labeled neurons in the lateral preoptic area following lateral habenula injections of tracer (see also Fig. 11 and Table 6). See list for additional abbreviations.

Figure 11

Graphs illustrating the numbers of retrogradely double-labeled neurons observed in various brain structures in cases in which (A) cholera toxin β subunit (Ctβ) was injected into the rostromedial tegmental nucleus (RMTg) and Fluorogold (FG) was injected into the lateral habenula (LHb) or (B) Ctβ was injected into the ventral tegmental area and FG was injected into the RMTg. Note that only the RMTg-LHb injection pair (A) produced preferential double-labeling in a particular structure, the lateral preoptic area. * - p < 0.05.