Monosynaptic Retrograde Tracing From Prelimbic Neuron Subpopulations Projecting to Either Nucleus Accumbens Core or Rostromedial Tegmental Nucleus

Abstract

The prelimbic (PL) region of the medial prefrontal cortex (mPFC) has been implicated in both driving and suppressing motivated behaviors, including cocaine-seeking in rats. These seemingly opposing functions may be mediated by different efferent targets of PL projections, such as the nucleus accumbens (NAc) core and rostromedial tegmental nucleus (RMTg), which have contrasting roles in reward-seeking behaviors. We sought to characterize the anatomical connectivity differences between PL neurons projecting to NAc core and RMTg. We used conventional retrograde tracers to reveal distinct subpopulations of PL neurons projecting to NAc core vs. RMTg in rats, with very little overlap. To examine potential differences in input specificity for these two PL subpopulations, we then used Cre-dependent rabies virus (EnvA-RV-EGFP) as a monosynaptic retrograde tracer and targeted specific PL neurons via injections of retrograde CAV2-Cre in either NAc core or RMTg. We observed a similar catalog of cortical, thalamic, and limbic afferents for both NAc- and RMTg-projecting populations, with the primary source of afferent information arising from neighboring prefrontal neurons in ipsilateral PL and infralimbic cortex (IL). However, when the two subpopulations were directly compared, we found that RMTg-projecting PL neurons received a greater proportion of input from ipsilateral PL and IL, whereas NAc-projecting PL neurons received a greater proportion of input from most other cortical areas, mediodorsal thalamic nucleus, and several other subcortical areas. NAc-projecting PL neurons also received a greater proportion of contralateral cortical input. Our findings reveal that PL subpopulations differ not only in their efferent target but also in the input specificity from afferent structures. These differences in connectivity are likely to be critical to functional differences of PL subpopulations.

Keywords: RMTg; addiction; medial prefrontal cortex (mPFC); rabies; reinstatement.

Figure 1

Retrograde tracing of prelimbic (PL) neurons projecting to nucleus accumbens (NAc) core and rostromedial tegmental nucleus (RMTg). (A) Representative photo of retrogradely-traced neurons in PL following injections of Fluoro-Gold (FG) in NAc core (blue) and CTB in RMTg (magenta). Photo location is shown in the inset (magenta box). Scale bar = 100 μm. (B) Location and spread for retrograde tracer injections in NAc core (FG, red) and RMTg (CTB, blue). Each rat is represented by one translucent outline, and outlines are overlapped. Anterior-posterior (AP) levels represent mm from bregma. Scale bar = 1 mm. (C) Percentage of PL neurons that project to both NAc core and RMTg (i.e., double-labeled for both CTB and FG), as a proportion of the total number of PL neurons projecting to NAc core (all FG-labeled cells) or RMTg (all CTB-labeled cells). Averages (± SEM) are shown, as well as individual data points for each rat (n = 11, *p < 0.05).

Figure 2

Virus injection strategy for labeling monosynaptic inputs to NAc core-projecting vs. RMTg-projecting PL neurons. (A) CAV2-Cre was injected into either NAc core (upper schematic) or RMTg (lower schematic), and rabies helper viruses (AAVs) were injected into PL during the initial surgery. EnvA-RV-EGFP was injected into PL 21 days later. Rats were sacrificed 7 days later. Brains were analyzed for starter cells (mCherry + EGFP) and input cells (EGFP). (B) Representative photo of CAV2-Cre injection site in NAc core with biotin dextran marking the center of the injection (arrow). Scale bar = 200 μm. (C) Location for CAV2-Cre injections in NAc core (red) or RMTg (blue). Each rat is represented by one translucent dot marking the center of the injection site. Scale bar = 1 mm. (D) Representative low-magnification fluorescent photo of EnvA-RV-EGFP labeling in PL at the injection site. Photo location is shown in the inset (magenta box). White dashed-outline box indicates the area shown at higher magnification in panels (E–G). Scale bar = 500 μm. (E) mCherry expression (pseudocolored as magenta, from AAV helper viruses). (F) EGFP expression (green, from EnvA-RV-EGFP). (G) Merged image showing both mCherry and EGFP (with dual fluorescence shown in white). Neurons expressing both mCherry and EGFP are starter cells for retrograde labeling, whereas neurons expressing only EGFP are monosynaptic afferents to the starter cells. Scale bar = 100 μm.

Figure 3

Quantification of monosynaptic inputs to NAc core-projecting vs. RMTg-projecting PL neurons. (A) Location of starter cells projecting to NAc core (red) or RMTg (blue). Each dot is one starter cell, and starter cells are shown together for all rats (n = 4 NAc core, 4 RMTg). (B) Relationship between number of starter cells and number of input cells (r = 0.91; **p < 0.01). (C) Contralateral and ipsilateral inputs to PL neurons projecting to NAc core vs. RMTg. Percentages reflect the number of cells quantified for each brain region divided by the total number of cells quantified for the whole brain per rat. Averages (±SEM) are shown only for brain regions with inputs ≥0.2% in ≥2 rats are shown. Statistically significant post hoc differences are shown (*p < 0.05; ****p < 0.0001). Abbreviations for thalamic nuclei: AM, anteromedial; AVDM, dorsomedial part of anteroventral; AVVL, ventrolateral part of anteroventral; CL, centrolateral; CM, central medial; IAM, interanteromedial; IMD, intermediodorsal; LDDM, dorsomedial part of lateroventral; LPMR, mediorostral part of lateral posterior; MD, mediodorsal; MDC, central part of mediodorsal; MDL, lateral part of mediodorsal; MDM, medial part of mediodorsal; PC, paracentral; PF, parafascicular; PT, paratenial; PVA, anterior part of paraventricular; PVP, posterior part of paraventricular; Re, reuniens; Rh, rhomboid; VA, ventral anterior; VL, ventrolateral; VM, ventromedial; VPM, ventral posteromedial; VRe, ventral reuniens.

Figure 4

Representative brightfield photos showing monosynaptic afferents to PL neurons projecting to NAc core (top row) or RMTg (bottom row), with EGFP expression visualized via immunohistochemistry with DAB labeling (brown) and counterstaining via Methyl Green. (A,B) Ipsilateral and contralateral inputs in PFC; scale bar = 1 mm. High-magnification photo (A’) shows neuronal labeling in the area indicated by the white dashed-outline box; scale bar = 100 μm. (C,D) Inputs from thalamic nuclei; scale bar = 500 μm. High-magnification photo (C’) shows neuronal labeling in the area indicated by the white dashed-outline box; scale bar = 100 μm. (E–H) Inputs from claustrum, dorsal endopiriform nucleus, and BLA; scale bar = 200 μm. (I,J) Inputs from ventral hippocampus CA1; scale bar = 1 mm. High-magnification photo (I’) shows neuronal labeling in the area indicated by the black dashed-outline box; scale bar = 100 μm. AD, anterodorsal thalamic nucleus; AI, agranular insular cortex; AM, anteromedial thalamic nucleus; AVDM, dorsomedial part of anteroventral thalamic nucleus; AVVL, ventrolateral part of anteroventral thalamic nucleus; BLA, basolateral amygdala; Cg, cingulate cortex part 1; Cl, claustrum; CM, central medial thalamic nucleus; DEn, dorsal endopiriform nucleus; DP, dorsal peduncular cortex; DTT, dorsal tenia tecta; IAM, interanteromedial thalamic nucleus; IL, infralimbic cortex; LA, lateral amygdala; LO, lateral orbitofrontal cortex; MD, mediodorsal thalamic nucleus; MHb, medial habenular nucleus; mt, mammillothalamic tract; PC, paracentral thalamic nucleus; PL, prelimbic cortex; PT, paratenial thalamic nucleus; PVA, anterior part of paraventricular thalamic nucleus; Re, reuniens thalamic nucleus; Rh, rhomboid thalamic nucleus; Rt, reticular thalamic nucleus; sm, stria medullaris; Sub, submedius thalamic nucleus; VA, ventral anterior thalamic nucleus; VL, ventrolateral thalamic nucleus; VM, ventromedial thalamic nucleus; VRe, ventral reuniens thalamic nucleus.