Developmental and Adult Striatal Patterning of Nociceptin Ligand Marks Striosomal Population With Direct Dopamine Projections Authors: Emily Hueske, Carrie Stine, Tomoko Yoshida, Jill R Crittenden, Akshay Gupta, Joseph C Johnson, Ananya S Achanta, Smitha Bhagavatula, Johnny Loftus, Ara Mahar, Dan Hu, Jesus Azocar, Ryan J Gray, Michael R Bruchas, Ann M Graybiel Published: 2024 ABSTRACT ABSTRACTCircuit influences on the midbrain dopamine system are crucial to adaptive behavior and cognition. Recent developments in the study of neuropeptide systems have enabled high‐resolution investigations of the intersection of neuromodulatory signals with basal ganglia circuitry, identifying the nociceptin/orphanin FQ (N/OFQ) endogenous opioid peptide system as a prospective regulator of striatal dopamine signaling. Using a prepronociceptin‐Cre reporter mouse line, we characterized highly selective striosomal patterning of Pnoc mRNA expression in mouse dorsal striatum, reflecting the early developmental expression of Pnoc. In the ventral striatum, Pnoc expression in the nucleus accumbens core was grouped in clusters akin to the distribution found in striosomes. We found that PnoctdTomato reporter cells largely comprise a population of dopamine receptor D1 (Drd1) expressing medium spiny projection neurons localized in dorsal striosomes, known to be unique among striatal projection neurons for their direct innervation of midbrain dopamine neurons. These findings provide a new understanding of the intersection of the N/OFQ system among basal ganglia circuits with particular implications for developmental regulation or wiring of striato‐nigral circuits.Keywords: dendron bouquet, dopamine, nociceptin, opioid, orphanin F/Q, striatum, striosome Introduction The midbrain neurons of the substantia nigra influence motor and cognitive behaviors related to reinforcement learning, motivation, behavioral choice, and movement initiation (da Silva et al. 2018; Hikosaka et al. 2014; Howe and Dombeck 2016; Schultz 2016). The circuits that innervate these dopamine neurons can bias, enhance, or diminish these modulatory functions. A labyrinthine striosomal compartment of the striatum that winds through the surrounding matrix contains neurons unique among striatal projection neurons (SPNs) in their direct innervation of midbrain dopamine‐containing neurons, forming a striosome‐dendron bouquet capable of shutting down dopamine neuron activity (Crittenden et al. 2016; Matsushima and Graybiel 2020; Watabe‐Uchida et al. 2012; Evans et al. 2020). Here, we report that among these bouquet‐innervating striosomal neurons there is a population selectively defined by the developmental expression of prepronociceptin (Pnoc), a gene encoding the opioid‐like ligand nociceptin/orphanin FQ (N/OFQ) associated with pain, stress, and reinforcement‐related processing by way of its dampening effect on dopamine (Flau et al. 2002; Norton et al. 2002; Parker et al. 2019; Zheng, Grandy, and Johnson 2002). N/OFQ is an opioid neuropeptide originally named for its induction of hyperalgesia (Meunier et al. 1995; Reinscheid et al. 1995). Its functions have since been shown to extend beyond pain modulation to influence motivation, stress, hormone regulation, learning and memory, mood regulation, feeding, and substance use disorder (D'Oliveira da Silva, Zaveri, and Moulédous 2023; Jenck et al. 1997; Kash et al. 2015; Koob 2008; Mogil et al. 1996; Parker et al. 2019). In the central nervous system (CNS), N/OFQ acts as a ligand for the N/OFQ opioid peptide (NOP) receptor, a G‐protein–coupled receptor that is widely distributed throughout the brain (Mollereau and Mouledous 2000). In contrast to its receptor's broad expression, N/OFQ ligand expression is more restricted and primarily localized in regions related to stress, feeding, and dopamine regulation (Neal et al. 1999; Norton et al. 2002; Ubaldi et al. 2021), including the central amygdala (CeA), hypothalamus, and bed nucleus of stria terminalis (Hardaway et al. 2019; Hernandez et al. 2021; Kash et al. 2015; M. A. Smith et al. 2020), as well as in a population of paranigral ventral tegmental area (VTA) neurons that locally modulate VTA dopamine neurons (Parker et al. 2019). There is a strong precedent for the importance of opioid signaling in the striatum, with extensive literature pointing to numerous striatal functions of other endogenous opioid systems, most notably mu and kappa, in the coordination of key reward and feeding behaviors (Al‐Hasani et al. 2015; Bodnar et al. 1995; Castro et al. 2021; Castro and Berridge 2014; Castro and Bruchas 2019; Ragnauth et al. 2000). In addition, dysregulated striatal opioid signaling can contribute to a wide range of neurological disorders, including mood disorders, substance use disorder, Huntington's disease, Parkinson's disease, and schizophrenia, among others (Banghart et al. 2015; Canales and Graybiel 2000; Clark and Abi‐Dargham 2019; Cui et al. 2014; Kennedy et al. 2006; Margolis et al. 2023; Morigaki et al. 2020; Sgroi and Tonini 2018). This is due in part to the distinct patterns of expression exhibited by different opioid receptors and peptides throughout the striatum, including differential expression in striosome and matrix compartments (Brimblecombe and Cragg 2017; Crittenden and Graybiel 2011; Graybiel 1990; Tajima and Fukuda 2013). Notably, N/OFQ has also recently been identified as a potential treatment candidate for major depressive disorder (Der‐Avakian et al. 2017; D'Oliveira da Silva, Zaveri, and Moulédous 2023; Gavioli and Calo’ 2013; Post et al. 2016; Ubaldi et al. 2021; Witkin, Wallace, and Martin 2019). Although Pnoc‐expressing neurons can be found throughout the striatum, their organization within striatal neuronal types and their anatomical intersection with dopamine circuitry across striatal subregions have remained largely unknown, especially in comparison to the other striatal opioid systems. Despite this, several studies have implied functional significance for the N/OFQ system in both the dorsal striatum and nucleus accumbens. Studies of dorsal striatal dopamine release following activation or blockade of NOP receptors in SNc highlighted bidirectional modulation of dopamine (Marti et al. 2004; Murphy and Maidment 1999; Olianas et al. 2008). In the nucleus accumbens, N/OFQ infusion has been shown to reduce dopamine release (Koizumi et al. 2004; Murphy and Maidment 1999) and inhibit dopamine synthesis at DA nerve terminals in the NAc by reducing tyrosine hydroxylase phosphorylation (Olianas et al. 2008). With regard to behavioral function, NOP receptor activation in the nucleus accumbens has been implicated in the modulation of reward‐related behaviors and the reinforcing properties of drug use in several studies (Bebawy et al. 2010; Caputi et al. 2014; Lutfy, Do, and Maidment 2001; Lutfy et al. 2002; Marquez et al. 2008; Murphy and Maidment 1999; Romualdi et al. 2007; Sakoori and Murphy 2004; Vazquez‐DeRose et al. 2013). Although there is evidence of N/OFQ's modulatory effects on striatal dopamine, little is known about the organization of striatal circuits subserving N/OFQ signaling and how they intersect with the midbrain dopamine system. This organization may be critical to N/OFQ's influence on approach‐avoidance and reward behaviors since several studies have demonstrated diversity in the function of opioid peptide/receptor systems specific to their location within different striatal subregions (Al‐Hasani et al. 2015; Castro et al. 2021; Castro and Berridge 2014; Cui et al. 2014; Resendez et al. 2013; Tejeda et al. 2017). In the current study, we investigated the spatial and molecular expression pattern of the N/OFQ ligand Pnoc across the dorsal and ventral striatum. To visualize Pnoc + neurons, we crossed mice with Pnoc‐driven expression of Cre recombinase (Pnoc‐Cre) with the Cre‐dependent Ai14 (tdTomato) reporter line to induce a lasting tdTomato fluorescent label in Pnoc+ neurons (PnoctdTomato). Using this mouse line, we visualized the spatial distribution of Pnoc + neurons across the dorsal striatum and the nucleus accumbens and co‐stained for various SPN and interneuron molecular markers to identify the neuronal subtypes labeled by PnoctdTomato reporter expression. We also evaluated differences in neonatal and adult striatal Pnoc expression, which suggests that N/OFQ function in the striatum may vary from development into adulthood. Results PnoctdTomato Cells in the Striatum Exhibit a Patterned, Clustered Organization that Corresponds With Immuno‐Identified Striosomes in the Dorsal Striatum To visualize Pnoc‐expressing neurons, we bred mice harboring Pnoc‐driven expression of Cre recombinase (Pnoc‐Cre) with a Cre‐dependent “reporter line,” loxP‐Stop‐loxP‐tdTomato (Ai14; LSL‐tdTomato), such that offspring enduringly “report” red fluorescent protein, tdTomato, in cells that have shown expression of Pnoc (PnoctdTomato), whether current or previously. Examination of these PnoctdTomato cells in the striatum demonstrated distinct spatial distribution patterns across both dorsal and ventral regions (Figure 1). In the dorsal striatum, PnoctdTomato cells were largely confined to striosomes, as identified by overlap with the known striosomal marker mu‐opioid receptor, MOR1 (Figure 1a–d). Discrete patterning was also evident in the ventral striatum, with PnoctdTomato neurons found predominantly in the dorsomedial portion of the medial shell, in high‐density clusters in the nucleus accumbens core, and scattered sparsely and more homogeneously throughout the lateral shell (Figure 1a,e–g). Across the entire striatum, PnoctdTomato neurons were distributed in similar proportions between the dorsal striatum, NAc core, and medial NAc shell, with a much smaller proportion found in the lateral NAc shell (Figure 1h). PnoctdTomato neurons were consistently found in clusters that localized with striosomes along the dorsal striatum's anterior to posterior axis at a ∼20‐fold higher density than found in the surrounding matrix (0.0007 ± 0.0001 cells/µm in striosomes; 3.8 × 10−5 ± 1.2 × 10−5 cells/µm in matrix; Figures 1i and S1). These findings demonstrate that Pnoc‐expressing neurons in the dorsal striatum are highly selective to the striosomal compartment and show clustered patterning in the ventral striatum. PnoctdTomato reporter expression shows compartmental distribution across the striatum and corresponds with striosomes in the dorsal striatum. (a) Composite image showing PnoctdTomato (magenta) and striosomal marker MOR1 (cyan) in the striatum. (b–d) Higher magnification images of the top boxed region from panel (a) with PnoctdTomato (b, magenta) and MOR1 (c, cyan) shown separately. Merged image (d) shows strong overlap between PnoctdTomato neurons and MOR1 expression. (e–g) same as (b–d) but for the lower boxed region in the ventral striatum from panel (a). (h) Quantification of PnoctdTomato spatial distribution across the dorsal and ventral (nucleus accumbens core, medial shell, and lateral shell) striatum (bar graph, one‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, N = 12 mice, N = 19 sections, 1–2 sections/mouse. Data represented as mean ± SEM across mice). (i) Quantification of PnoctdTomato spatial localization in striosomes vs. the matrix across the anterior to posterior axis in the dorsal striatum (two‐way ANOVA shows main effect of striatal compartment ***p < 0.001, no significant effect of anterior‐posterior axis p = 0.0851 bar graph, two‐way ANOVA with Bonferroni's multiple comparisons test, ***p < 0.001, ****p < 0.0001, N = 8 mice, 3 sections/mouse). Data represented as mean ± SEM across mice. ANT = anterior; DS = dorsal striatum; lNAcSh = lateral accumbens shell; MID = medial; mNAcSh = medial accumbens shell; NAc = nucleus accumbens; NAcC = nucleus accumbens core; POS = posterior. Striatal PnoctdTomato Neurons Predominantly Overlap With D1 SPNs, With Greater Heterogeneity Among SPN Subtypes in the Ventral Striatum To further molecularly classify striatal Pnoc + neurons, we evaluated PnoctdTomato co‐expression with markers for different SPN subtypes. We crossed our Pnoc‐Cre;LSL‐tdTomato reporter line with Drd1‐GFP (D1GFP) or Drd2‐GFP (D2GFP) reporter mice to simultaneously visualize PnoctdTomato neurons colocalized with either D1+ or D2+ neurons labeled by GFP (Figure 2). We found a significant colocalization of PnoctdTomato with D1GFP‐type SPNs in both dorsal (87.9% ± 0.7%) and ventral (63.7% ± 10.8%) striatum (Figure 2a–i), to the near exclusion of D2GFP SPNs (dorsal 1.3% ± 0.2%; ventral 4.3% ± 1.1%, Figure 2j–r). Due to a lack of full genetic penetrance that is characteristic of GFP reporter lines, we also quantified the SPN cell types by in situ hybridization (ISH) of Drd1 and Drd2, which were consistent with our initial findings (Figure 2s–u). The distribution of PnoctdTomato neurons that co‐localized with Drd1, Drd2, both, or neither, remained consistent across the dorsolateral, ventrolateral, and medial dorsal striatum (Figure S2‐1). By contrast, PnoctdTomato SPN subpopulations in the ventral striatum appeared more heterogeneous, with just under half of PnoctdTomato neurons in the lateral NAc shell (48.0% ± 12.5%) and almost a third in the medial shell (29.8% ± 10.5%) not co‐localized with either Drd1 or Drd2 (Figure S2‐2). These results indicate that PnoctdTomato labeling predominantly overlaps with Drd1 expression across the dorsal and ventral striatum, suggesting that striatal Pnoc expression largely occurs within D1‐type SPNs. PnoctdTomato expressing striatal neurons cluster with Drd1 SPNs to near exclusion of Drd2. (a) Sagittal image of PnoctdTomato cells (magenta) colocalizing with D1‐GFP‐expressing SPNs (green) in dorsal (b–e) and ventral (f–i) striatum. (b) Higher magnification image of top boxed region from panel (a) in the dorsal striatum. (c–e) Higher magnification images from inset in (b) of dorsal striatum. Arrowheads point to cells expressing markers for both PnoctdTomato and D1‐GFP. (f) Higher magnification image of the bottom boxed region from panel (a) in the ventral striatum (nucleus accumbens). (g–i) Higher magnification images from inset in (f) of the ventral striatum. (j) Sagittal image of minimal overlap between PnoctdTomato cells (magenta) and D2‐GFP expression (green) in dorsal (k–n) and ventral (o–r) striatum (same as [b–e] and [f–i] but for D2‐GFP expression). (s) Summary of IHC‐stained PnoctdTomato reporter cell colocalization with in situ hybridization‐labeled Drd1 and Drd2 SPNs in dorsal striatum. (t) Quantification of PnoctdTomato cells on Drd1 and Drd2 labeled SPNs across the anterior to posterior axis in the dorsal striatum (main effect of neuron type, two‐way ANOVA, ****p < 0.0001; bar graph, two‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, N = 6 mice, 3 brain sections/mouse). (u) Same as (s) for the nucleus accumbens. (v) Quantification in nucleus accumbens of PnoctdTomato labeling in Drd1 and Drd2 SPNs (bar graph, one‐way ANOVA with Bonferroni's multiple comparisons test, ***p < 0.001, *p < 0.05, N = 6 mice, N = 8 sections, 1–2 sections/mouse). Data represented as mean ± SEM across mice. ANT = anterior; DS = dorsal striatum; MID = medial; NAc = nucleus accumbens; POS = posterior. A Small, Interneuronal Population of PnoctdTomato Cells Are Marked by SST/NPY Expression We next asked whether PnoctdTomato cells are found within any interneuronal populations in the striatum by staining for the interneuron markers neuropeptide Y (NPY), somatostatin (SST), choline acetyltransferase (ChAT), and parvalbumin (PV) in combination with PnoctdTomato labeling (Figure 3). The majority of striatal PnoctdTomato neurons did not overlap with any interneuronal markers (dorsal striatum: 92.3% ± 0.5% PnoctdTomato only; ventral striatum: 81.0% ± 5.3% PnoctdTomato only), a small dorsal striatal population of PnoctdTomato neurons overlapped with NPY+/SST+ interneurons (5.2% ± 0.5%), often seen near the borders of striosomal PnoctdTomato cell clusters, and a marginally larger subset in the ventral striatum overlapped with NPY+/SST− interneurons (16.9% ± 5.1%) (Figures 3a–e,k,l and S3). In both dorsal and ventral striatum, Pnoc reporter expression showed little to no overlap with PV+ or ChAT+ striatal interneurons (Figure 3f–j). These PnoctdTomato interneuronal expression patterns did not detectably vary across the anterior to posterior striatal axis (Figure 3k). These multiple findings indicate that striatal PnoctdTomato cells primarily overlap with D1‐type SPNs and not with striatal interneurons. PnoctdTomato reporter expression marks a population of NPY and SST co‐expressing interneurons. (a) Coronal image of PnoctdTomato cells (magenta), somatostatin (white, SST), and neuropeptide Y (cyan, NPY) expression in the striatum. Higher magnification images of inset from panel (a) for (b) merged and (c–e) individual channels showing PnoctdTomato (c) colocalization with a subset of striatal interneurons co‐expressing SST (d) and NPY (e). (f–j) Same organization as (a–e) for additional interneuron markers; PnoctdTomato cells (magenta) show little colocalization with choline acetyltransferase (green, ChAT) or parvalbumin (blue, PV) expressing striatal interneurons. (k) Top: Summary of quantified PnoctdTomato cells colocalized with different interneuron markers in the dorsal striatum. Bottom: Quantification of PnoctdTomato cells colocalized with different interneuron markers across the anterior to posterior axis in the dorsal striatum (main effect of neuron type, two‐way ANOVA, ****p < 0.0001; bar graph, two‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, N = 6 mice, 3 sections/mouse). (l) Same as (k) but for nucleus accumbens (bar graph, one‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, **p < 0.01, *p < 0.05, N = 6 mice, N = 11 sections, 1–2 sections/mouse). Data represented as mean ± SEM across mice. ANT = anterior; lNAcSh = lateral accumbens shell; MID = medial; mNAcSh = medial accumbens shell; NAcC = nucleus accumbens core; POS = posterior. Expression of Striatal Pnoc mRNA is Subthreshold to Levels Required for Cre‐Based Recombination of PnoctdTomato Reporter in Adult Mice In contrast to the abundant Pnoc+ labeling found throughout the striatum using the tdTomato reporter line, when we instead probed for Pnoc mRNA expression using in situ hybridization, we detected surprisingly little Pnoc mRNA in adult striatal tissue with no appearance of striosomal patterning (Figure 4a–p). Indeed, although Pnoc levels in reporter‐positive (tdT+) neurons were elevated relative to a negative control (NC) labeled tissue, they were not different from Pnoc levels in reporter‐negative (tdT−) neurons (Figure 4q,r). This puzzling finding suggests that while not nonexistent, Pnoc mRNA expression in adult animals is not sufficient to drive Pnoc promoter‐driven Cre expression at levels required for Cre‐dependent recombination of the LSL‐tdTomato reporter. Supporting this, we observed a lack of Cre‐dependent recombination in adult tissue following injection of a viral vector encoding a Cre‐dependent (DIO‐) membrane GFP construct into the dorsal striatum of adult Pnoc‐Cre mice (Figure S4‐1). Intriguingly, average Pnoc levels in both tdT+ and tdT− cells were statistically higher than NC‐labeled tdT+ cells assayed in adjacent brain sections labeled with an NC ISH probe in place of the Pnoc probe (Figure 4q,r). This indicates that there may be low, basal level expression throughout the striatum of the Pnoc gene that is subthreshold to levels required for Cre‐dependent recombination. Functional studies will be required to evaluate if there is any physiological relevance of this low‐level expression. Adult expression of striatal Pnoc mRNA is sparse relative to reporter‐labeled PnoctdTomato cells. (a–d) Coronal image showing composite (a) of DAPI (blue, b), Pnoc mRNA expression evaluated by ISH (green, c), and reporter‐labeled PnoctdTomato cells (magenta, d) in the adult striatum. Magnified insets from the dorsolateral striatum (e–g), dorsomedial striatum (h–j), nucleus accumbens core (k–m), and nucleus accumbens shell (n–p) show PnoctdTomato reporter‐labeled cells that do not co‐label with adult Pnoc mRNA expression. (q) Quantification of mean Pnoc mRNA intensity in PnoctdTomato cells (tdT+, green), nearby non‐reporter labeled cells (tdT−, yellow), and in PnoctdTomato cells from adjacent control sections labeled with a negative control probe (NC, gray) in adult dorsal striatum. (Two‐way ANOVA with Bonferroni's multiple comparisons test, ***p < 0.001, **p < 0.01, ns = non‐significant, N = 4 mice, N = 7 sections, 1–2 sections/mouse). (r) Same as (q) but for adult ventral striatum (two‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, ***p < 0.001, ns = non‐significant, N = 4 mice, N = 7 sections, 1–2 sections/mouse). (s) Proportion of PnoctdTomato reporter cells identified as Pnoc mRNA positive (Pnoc+/tdT+, green) or negative (Pnoc−/tdT+, magenta) in adult dorsal striatum. (t) Same as (s) but for adult ventral striatum. All data represented as mean ± SEM across mice. DLS = dorsolateral striatum; DMS = dorsomedial striatum; lNAcSh = lateral accumbens shell; mNAcSh = medial accumbens shell; NAcC = nucleus accumbens core. Based on these findings, we also quantified the proportion of striatal PnoctdTomato reporter cells that were positive for Pnoc mRNA, classifying cells as positive based on Pnoc signal levels that outstrip the homogenous “background” labeling seen in tdT− cells (Figure 4s,t). Only between ∼1% and 5% of tdT+ cells were found to be Pnoc+ across all striatal subregions (Figure 4s,t). These results indicate that active Pnoc mRNA expression in adult striatal tissue is both markedly low and sparse, occurring at levels insufficient to drive the Cre recombinase expression that induces tdTomato‐reporter labeling. Taken together, our findings suggest that the majority of reporter‐labeled PnoctdTomato neurons in the striatum are representative of transient Pnoc gene expression that occurred at an earlier time point. Pnoc mRNA Is Developmentally Expressed in Striosomes In Pnoc‐Cre;LSL‐tdTomato reporter animals, PnoctdTomato labeling of a given cell is triggered by the earliest sufficient expression of Cre‐recombinase in that cell and will persist throughout the lifetime of the mouse. Given this, the discrepancy between the robust PnoctdTomato reporter labeling and the dearth of adult Pnoc mRNA expression suggested that the Pnoc expression in the dorsal striatum could be developmentally regulated, with higher expression levels during embryonic and/or postnatal stages followed by a downregulation of expression as animals reach maturity. To investigate this possibility, we used ISH to detect Pnoc mRNA expression in the striatum of postnatal Day 1 (P1) Pnoc‐Cre;LSL‐tdTomato reporter mice (Figure 5). In contrast to our findings in adult tissue, Pnoc mRNA expression in neonatal tissue did recapitulate the reporter‐labeled patterning of PnoctdTomato cells (Figure 5a–p). Average Pnoc intensity in tdT+ cells across the dorsolateral striatum and all ventral striatal subregions was significantly elevated compared to both nearby tdT− cells and NC‐labeled PnoctdTomato cells from adjacent sections (Figure 5q,r). In the dorsomedial striatum, average Pnoc intensity in tdT+ cells relative to NC controls neared but did not reach statistical significance after averaging across mice (tdT+ vs. NC p = 0.0581, Figure 5q). We found that the vast majority (∼70%–90%) of neonatal PnoctdTomato cells were positive for Pnoc mRNA expression across all striatal subregions (Figure 5s,t). Neonatal expression of striatal Pnoc mRNA recapitulates reporter‐labeled patterning in PnoctdTomato cells. (a–d) Coronal image showing composite (a) of DAPI (blue, b), Pnoc mRNA expression evaluated by ISH (green, c), and reporter‐labeled PnoctdTomato cells (magenta, d) in postnatal Day 1 (P1) brain sections. Magnified insets from the dorsolateral striatum (e–g), dorsomedial striatum (h–j), nucleus accumbens core (k–m), and nucleus accumbens shell (n–p) show PnoctdTomato reporter‐labeled cells co‐labeled with neonatal Pnoc mRNA expression. (q) Quantification of mean Pnoc mRNA intensity in PnoctdTomato cells (Pnoc−/tdT+, green), nearby non‐reporter labeled cells (Pnoc−/tdT−, yellow), and in PnoctdTomato cells from adjacent control sections labeled with a negative control probe (NC, gray) in neonatal dorsal striatum. (Two‐way ANOVA with Bonferroni's multiple comparisons test, **p < 0.01, ns = non‐significant, N = 4 mice, N = 8 sections, 2 sections/mouse). (r) Same as (q) but for neonatal ventral striatum (two‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, ***p < 0.001, ns = non‐significant, N = 4 mice, N = 7 sections, 1–2 sections/mouse). (s) Proportion of PnoctdTomato reporter cells identified as Pnoc mRNA positive (Pnoc+/tdT+, green) or negative (Pnoc−/tdT+, magenta) in neonatal dorsal striatum. (t) Same as (s) but for neonatal ventral striatum. (u) Proportion of PnoctdTomato reporter cells identified as Pnoc mRNA positive in neonate (light purple) vs. adult (dark purple) tissue (two‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, N = 4 mice/group, N = 7 sections, 1–2 sections/mouse). (v) Fold‐change in mean Pnoc mRNA intensity in tdT+ cells relative to background (tdT− cells) in neonate (light purple) vs. adult (dark purple) tissue (two‐way ANOVA with Bonferroni's multiple comparisons test, ****p < 0.0001, ***p < 0.001, **p < 0.01, ns = non‐significant, N = 4 mice/group, N = 7 sections, 1–2 sections/mouse). All data represented as mean ± SEM across mice. DLS = dorsolateral striatum; DMS = dorsomedial striatum; NAcC = nucleus accumbens core; NAcSh = nucleus accumbens shell. When we compared the proportion of PnoctdTomato cells expressing Pnoc mRNA in adult vs. neonatal tissue, we found a significantly higher proportion of Pnoc + tdTomato‐labeled cells in neonates relative to adults across the striatum (neonatal vs. adult percent tdT+ cells positive for Pnoc p < 0.0001 for all striatal subregions, Figure 5u). This result is congruent with our hypothesis that striatal Pnoc mRNA expression is developmentally enriched. In addition, we compared the intensity of Pnoc mRNA expression in tdT+ cells relative to background levels (tdT−) in adult vs. neonatal tissue, finding that Pnoc mRNA intensity in neonatal tdT+ cells is elevated relative to adult levels across the dorsolateral and ventral striatum (p < 0.01 for all striatal subregions except DMS, Figure 5v). In the dorsomedial striatum (DMS), the average, normalized Pnoc mRNA intensity did not reach significance in neonatal tissue relative to adults (p = 0.4174, Figure 5v). However, there is still a higher proportion of Pnoc + cells among tdT+ cells (Pnoc +/tdT+) in neonatal DMS, indicating that Pnoc mRNA expression is more abundant in neonatal DMS than in adult DMS. Together, these findings demonstrate that Pnoc gene expression in the striatum is generally elevated and more abundant during earlier developmental stages, becoming subsequently downregulated by adulthood. PnoctdTomato Neurons in Striosomes Directly Project to Dopamine Neurons Striosomal Drd1 (D1) neurons are unique among striatal SPNs, or nearly so, in their direct innervation of midbrain dopamine neurons, in particular the ventral tier dopamine neurons of the SNc. However, we recently reported that PnoctdTomato cells represent ∼40% of striosomal Drd1 + neurons (Lazaridis et al. 2024), and it is not known whether all striosomal Drd1 + neurons forge such a projection. Our identification of PnoctdTomato neurons as predominantly overlapping with Drd1 SPNs suggests that dorsal striatal PnoctdTomato cells might correspond to this projection population. Examining midbrain sections of the PnoctdTomato reporter mice, we observed specific innervation of the nigral striosome‐dendron bouquets by the labeled axons of PnoctdTomato cells (Figure 6a) both in the adult (Figure 6b–d) and in early P5 mouse tissue (Figure 6e–g). We then established that axons projecting specifically from striosomes account for these bundled fibers entwining themselves with dopamine‐containing dendrons, rather than from another source, by injecting a Flp‐dependent membrane GFP‐expressing AAV vector into the dorsal striatum of Pnoc‐Cre;loxP‐Stop‐loxP‐FlpO (PnocFlpO) mice (Figure 6h–m). Again the striosome‐dendron bouquets were innervated by the population of striosomal neurons defined by early Pnoc expression. These findings demonstrate that Pnoc+ striatal neurons comprise a population of dorsal striatal dSPNs highly selective to the striosomal compartment and that these Pnoc+ neurons forge the specialized striatonigral circuit targeting striosome‐dendron bouquets early in development. Our results open the door for further investigation of the functional significance of this population and its ventral striatal counterpart, also showing clustered patterning of Pnoc+ expression. Striosomal fibers from PnoctdTomato targeted neurons project to the striosome‐dendron bouquet in SNc in both adults and neonates. In PnoctdTomato reporter mice (a), tdTomato‐labeled fibers (magenta) project to TH‐labeled dopamine neurons (green) in adults (b–d) and neonates (e–g). (h) Schematic of mouse genetic strategy by viral labeling of striosomal cells in Pnoc‐Cre;loxP‐Stop‐loxP‐FlpO mice with a Flp‐dependent AAV expressing membrane‐GFP (mGFP). Virally targeted mGFP‐expressing PnocFlp neurons (i–j) exhibit projections to dopamine dendrons in SNc (k–m). Discussion The N/OFQ system has a demonstrated ability to disrupt the landscape of dopamine signaling with functional consequences relevant to stress and mood regulation, substance use disorder, and Parkinson's disease, among others. The striatal circuits and cell types that N/OFQ intersects with to drive these effects have remained poorly defined. Here, we lay the groundwork for investigating N/OFQ‐mediated striatal contributions to these dopamine‐related functions. Our findings reveal an anatomical and molecular organization of Pnoc expression primarily in striatal D1+ neurons, indicative of the structure that underlies N/OFQ's functional interactions with striatal‐nigral circuitry. We also report dynamic expression in the striatum throughout the development of this opioid neuropeptide ligand known for its modulatory effects on behavior and dopamine function. We found that PnoctdTomato neurons exhibit highly selective striosomal patterning in the dorsal striatum, and within the ventral striatum, we found distinctly concentrated PnoctdTomato cell clusters in the core and medial shell, but not in the lateral shell. In particular, we have demonstrated that striosomal PnoctdTomato cells in the dorsal striatum largely co‐express Drd1 but not Drd2, thereby comprising part of the unique striosomal direct pathway neurons that make direct connections with dopamine‐containing neurons via striosome‐dendron bouquet structures in the ventral tier SNc (Crittenden et al. 2016; Evans et al. 2020; Matsushima and Graybiel 2020; Watabe‐Uchida et al. 2012). In the ventral striatum, we highlight a newly identified population of molecularly heterogeneous Pnoc neurons with distinct spatial distribution patterns across the nucleus accumbens shell and core and adjoining ventral territories. Drd1 was similarly the leading marker found co‐expressed with PnoctdTomato neurons in the ventral striatum; however, a significant proportion of PnoctdTomato neurons did not overlap with either Drd1 or Drd2 expression in the lateral NAc shell. To determine whether these ventral striatal direct pathway neurons share a circuit motif of dopamine projection with their dorsal striosomal neighbors will require careful mapping of individual ventral striatal populations. Given the known importance of the ventral striatum for many of the functions for which N/OFQ has been implicated and in which dopamine is involved, we suggest that the afferent and efferent projections of these distinct ventral populations are an important next step in pursuit of evaluating ventral striatal N/OFQ function. We found PnoctdTomato to be a superlative marker of striosomes across development and into maturity, but that this reporter‐based patterning is reflective of developmentally transient expression of the Pnoc gene that becomes downregulated by adulthood (Figures 4, 5, and S4‐1). Such transient developmental expression is reminiscent of a number of striatal genes that exhibit striosomal patterning early in development, including FoxP2 (Enard 2011; Fong et al. 2018), Asc1, and Dlx1 (Kelly et al. 2018). Striosomal neurons are born during embryonic days ∼10.5–13.5 and lead the migration out into the striatal primordium ahead of the migration of neurons destined for the matrix compartment (Graybiel and Hickey 1982; Kelly et al. 2018; Lança et al. 1986; Matsushima and Graybiel 2020). In the course of development, striosomes and matrix exhibit different patterns of neurochemical and mRNA expression (Eblen and Graybiel 1995; Gokce et al. 2016; Graybiel and Hickey 1982; Kelly et al. 2018; Lança et al. 1986; Märtin et al. 2019; Matsushima and Graybiel 2020; Saunders et al. 2018) and develop distinct efferent and afferent patterns of connectivity. Striosomal neurons are more connected with limbic regions than are neurons of the matrix; they receive enriched inputs from limbic regions of the cortex, the bed nucleus of the stria terminalis, and amygdala as well as providing the main source of striatal projections to midbrain dopamine neurons (Crittenden et al. 2016; Crittenden and Graybiel 2011; Graybiel and Matsushima 2023; Lee et al. 2024; Ragsdale and Graybiel 1988; J. B. Smith et al. 2016; Watabe‐Uchida et al. 2012). Due to concerns surrounding the selectivity of opioid antibodies and the widely known challenges associated in particular with labeling neuropeptides in fixed tissue (Liu, Kwok, and Fernstrom 1991), we instead utilized detection methods that probe for Pnoc gene expression. It is important to note however that mRNA transcript expression does not necessarily equate to the presence of the fully processed N/OFQ peptide. While the proteolytic processing of the precursor peptide pronociceptin is not entirely known, the enzyme prohormone convertase 2 (PC2), which is central to this process for many neuropeptides, including the other opioid peptides, has been shown to metabolize pronociceptin into N/OFQ (Allen et al. 2001). PC2 is abundantly present across the dorsal striatum and nucleus accumbens in both D1 and D2 SPNs (Montalban et al. 2022), so it is certainly plausible that functional N/OFQ peptide can be generated throughout the striatum. Still, this is a limitation of the current study and an important consideration for the interpretation of our findings. Striatonigral N/OFQ signaling could play a role in the development of the striosome‐dendron bouquet circuit. Studies following the discovery of N/OFQ showed that high levels of Pnoc transcript in undifferentiated neural cells are accompanied by neurite outgrowth (Saito et al. 1996, 1995, 1997; Zaveri, Waleh, and Toll 2006). N/OFQ signaling also activates the master transcriptional regulator, NFκB, important in the regulation of genes involved in synaptic plasticity, memory formation, cognition, development, and neuronal survival (Donica et al. 2011; Levenson, Pizzi, and Sweatt 2006). In addition, N/OFQ modulates immune factors such as TNFα and IL‐6 (Fu et al. 2007), crucial to the formation and maintenance of neural circuit development and synaptic plasticity (Boulanger 2009). Our finding of robust Pnoc expression selectively in striosomal direct pathway neurons slated for dopaminergic interconnection opens the door for evaluation of how these molecular genetic and immune interactions unfold in a circuit of wide functional and pathological consequences. Still to be determined is the impact on the striatonigral circuitry of developmental perturbations of Pnoc expression. The potential for the development of circuit vulnerabilities and dysfunction through the mis‐wiring of striosomal neurons with the dopamine system is evident from the nature of those neurons with which striosomes form interconnections. Here, we show that the striosomal neurons that developmentally expressed Pnoc directly innervate vSNc dopamine neurons (Figure 6). The dopamine neurons of the vSNc, which are the recipient neurons of these striosomal‐dendron bouquet axons, carry particular susceptibility to degeneration in Parkinson's disease (Damier et al. 1999; Giguère et al. 2018; Kordower et al. 2013; Mendez et al. 2005; Pereira Luppi et al. 2021; Yamada et al. 1990). More widely, the cortico‐striosome‐dopamine circuit is implicated in mood regulation and dysregulation as striosomes are differentially implicated from surrounding matrix in reinforcement‐related updating paradigms (Bloem et al. 2022, 2017; Yoshizawa, Ito, and Doya 2018), and, in particular, in influencing cost‐benefit conflict decision‐making in both rodents (Friedman et al. 2017, 2015, 2020) as well as non‐human primates (K.‐I. Amemori et al. 2018; K.‐I. Amemori and Graybiel 2012; S. Amemori et al. 2020; S. Amemori, Graybiel, and Amemori 2021), findings that have been extended to human patient populations (Ironside et al. 2020; Pedersen et al. 2021; Rolle et al. 2022). These studies suggest circuits to study for a mechanistic understanding of how the cortico‐striato‐nigral circuit and the nigro‐striato‐nigral loop operate in dopamine regulation in the context of such disorders. As a result of its modulation of dopamine, the N/OFQ system has attracted interest as a therapeutic target for modulating dopamine in both Parkinson's disease, substance use disorder, and mood regulation (Gavioli and Calo’ 2013; Lambert 2008; Post et al. 2016; Witkin, Wallace, and Martin 2019; Zaveri 2011), including in alleviation of depressive symptoms resulting from developmental inflammatory insults arising from maternal immune activation (Medeiros et al. 2015). Our findings point to the value of further work on this developmental impact. Materials and Methods Mouse Lines and Husbandry All animal procedures were approved by the Committee on Animal Care at the Massachusetts Institute of Technology, were performed in accordance with the U.S. National Research Council Guide for the Care and Use of Laboratory Animals, and were maintained under conditions of constant temperature (25°C) and humidity (50%) and a 12:12 h light/dark cycle. The following mouse strains were used for behavioral, neural recording, and histological experiments: Pnoc‐Cre (Parker et al. 2019), Ai14 (Madisen et al. 2010), and LSL‐Flpo (He et al. 2016). We crossed Pnoc‐Cre mice on a C57BL/6J background (Jackson Laboratory) to the Cre‐dependent Ai14 (LSL‐tdTomato reporter line) or Cre‐dependent LSL‐Flpo line on an FVB background (Taconic Biosciences) corrected for behavioral optimization. Consequently, F1 hybrid mice are produced to suit behavioral characterization (Menalled and Brunner 2014), avoiding the early hearing loss of the C57BL/6J mouse strain and poor vision of the FVB strain. Surgical Preparations of Stereotactic Striatal Injections All surgeries were performed on adult mice over 8 weeks of age. Animals were anesthetized briefly with 5% isoflurane and maintained at 1%–2% throughout the surgery. For intrastriatal delivery of viral vectors, Pnoc‐Cre or Pnoc‐Cre;LSL‐Flp mice were injected unilaterally with AAVs packaging Cre‐ or Flp‐dependent membrane GFP and synaptophysin‐mRuby constructs (AAV2/8 vector packaging hSyn‐DIO‐mGFP‐T2A‐Synaptophysin‐mRuby and hSyn‐FDIO‐mGFP‐T2A‐Synaptophysin‐mRuby; Biohippo, Gaithersburg, MD) targeting the anterior dorsal striatum. All viruses were injected at a rate of 100 nL/min. Injection coordinates empirically optimized for F1 hybrid C57BL/6J × FVB mice were based on Friedman et al. (2020) for ages < 6 months: AP: +1.1 mm; ML: ±1.7 mm; DV: 2.0 mm from the brain surface. Histological Labeling of Brain Tissue Animals were transcardially perfused with a phosphate‐buffered saline (PBS) solution followed by 4% paraformaldehyde in PBS. Extracted brains were soaked in the paraformaldehyde solution for 24 h and then maintained in 30% sucrose in PBS at 4°C for cryoprotection. After 24 h, brains were frozen in dry ice and stored at −80°C. A sliding microtome was used to cut 30‐µm‐thick sagittal or coronal sections from the brains of the adults and 54‐µm‐thick coronal sections from the brains of P1 to P20 mice. Sections were stored at 4°C in 0.02% sodium azide in 0.1 M PB ahead of preparation for immunofluorescence (IF) staining. For fluorescent in situ hybridization (FISH), 10‐µm‐thick coronal sections were cut from adult mouse brains, and 18‐µm‐thick coronal sections were cut from the P1 to P20 mouse brains. Sections were mounted on slides and stored at −80°C until use. For IF staining, sections were rinsed three times for 10 min in PBS with 0.2% Triton X‐100 (Tx), then blocked with tyramide signal amplification (TSA) blocking solution (Akoya Bioscience FP1012) for 20 min at room temperature and then incubated with primary antibodies (Table 1) diluted in TSA blocking solution for 48–72 h at 4°C. Sections were again rinsed with PBS‐Tx for 3 × 10 min and were incubated with fluorescent conjugated secondary antibodies (Table 1) for 2 h at room temperature. Sections were rinsed with PB for 10 min, then counterstained with DAPI solution diluted in PB for 2 min. After rinsing in PB for three times, sections were mounted with ProLong Gold and stored at 4°C until imaging. For FISH, the RNAscope Multiplex Fluorescent Reagent Kit (Cat. 320850; Advanced Cell Diagnostics (ACD), Hayward, CA) and manufacturer's protocols were used. For neonatal and adult tissue, the Multiplex Fluorescent Detection Reagents kit V2 (Cat. 323110; Advanced Cell Diagnostics (ACD), Hayward, CA) was used, and the manufacturer's protocols were followed. The probes used in this study are shown in Table 2 with the combination of Channels 1–3. For IF‐FISH co‐staining, RNase‐free blocking reagent and co‐detection Target Retrieval Reagents from the RNA‐Protein Co‐detection Ancillary Kit (Cat No. 323180, ACD) were used. TABLE 1Antibodies used in this study.AntibodiesSourceCatalog numberRRIDPolyclonal chicken anti‐GFPAbcamab13970AB_300798Monoclonal rat anti‐mCherryInvitrogenM11217AB_2536611Monoclonal rabbit anti‐MOR1Abcamab134054AB_3122135Polyclonal rabbit anti‐THAbcamab112AB_297840Monoclonal rat anti‐somatostatinMilliporeMAB354AB_2255365Polyclonal rabbit anti‐neuropeptide YAbcamab221145AB_2894872Polyclonal guinea pig anti‐PVSynaptic Systems195004AB_2156476Polyclonal goat anti‐ChaTMilliporeAB144PAB_2079751Goat anti‐chicken IgY (H + L), Alexa Fluor 488InvitrogenA‐11039AB_2534096Goat anti‐rat IgG (H + L), Alexa Fluor 546InvitrogenA‐11081AB_2534125Goat anti‐rabbit IgG (H + L), Alexa Fluor 647InvitrogenA‐21245AB_2535813Donkey anti‐guinea pig IgG (H + L), DyLight 405Jackson ImmunoResearch706‐475‐148AB_2340470Donkey anti‐goat IgG (H + L), Alexa Fluor 488InvitrogenA‐11055AB_2534102Donkey anti‐rat IgG (H + L), Alexa Fluor 647InvitrogenA78947AB_2910635Donkey anti‐rabbit IgG (H + L), BiotinInvitrogenA16039AB_2534713Streptavidin, Alexa Fluor 750InvitrogenS21384N/A Antibodies used in this study. TABLE 2In situ hybridization probes used in this study.Probe nameSourceCatalog number of Channel 1Mm‐Drd1aAdvanced cell diagnostics406491Mm‐Drd2Advanced cell diagnostics406501Mm‐PnocAdvanced cell diagnostics437881tdTomatoAdvanced cell diagnostics317041Mm‐Oprm1Advanced cell diagnostics315841Mm‐Lypd1Advanced cell diagnostics318361Negative control probeAdvanced cell diagnostics320871 In situ hybridization probes used in this study. Image Analysis Images were collected on a confocal TissueFAXs SL slide scanner (TissueGnostics) and were exported in tiff format. For the IF‐FISH co‐staining experiment, images were collected on the Epifluorescent TissueFAXs slide scanner (TissueGnostics) after IF staining (image set 1), and then images were collected again following FISH (image set 2). These two image sets were exported in TIFF format and then registered by the ImageJ plugin, StackRegJ (Stowers Institute for Medical Research). Registered images were then evaluated using ImageJ (FIJI), QuPath (Bankhead et al. 2017), or HALO image analysis software (IndicaLabs). For the detection and classification of cells and colocalization with cell type markers, QuPath software was used to annotate striatal subregions, and the cell detection module was used to detect cells based on tdTomato signal or nuclear signals NeuN or DAPI. Detections were visually inspected and corrected for missed or false‐positive detections, and measurements were collected for mean intensity values and classification for each detection in each annotated subregion in each brain section. For classification of cell type, the QuPath classifier module was used to establish thresholds based on the average intensity values of each marker for classifying cells for one or multiple probe and/or antibody labels. For adult and neonatal quantification, Pnoc ISH labeling was evaluated in Pnoc‐tdTomato‐positive and tdTomato‐negative cells, and in the NC section tdTomato positive cells were treated identically and simultaneously but labeled with an ISH probe against soil in place of the Pnoc probe. Neonatal and adult tissue require different histological processing due to differences in tissue density, and mean probe intensity values cannot be directly compared between adult and neonatal sections. NC labeling was done to draw comparisons between the tissues by allowing for normalization and thresholding purposes. Statistical Analyses Statistical analyses were performed as indicated using GraphPad Prism version 10.0.0 for Windows, GraphPad Software, Boston, MA, USA, www.graphpad.com. All data are expressed as mean ± SEM. Author Contributions Emily Hueske: conceptualization (lead), data curation (co‐lead), formal analysis (co‐lead), investigation, funding acquisition, project administration (lead), resources, supervision (co‐lead), validation, visualization, writing–original draft preparation (co‐lead), writing–reviewing and editing (co‐lead). Carrie Stine: conceptualization, data curation (co‐lead), formal analysis (co‐lead), funding acquisition, project administration, supervision (co‐lead), validation, visualization (lead), writing–original draft preparation (co‐lead), writing–reviewing and editing (co‐lead). Tomoko Yoshida: conceptualization, data curation (co‐lead), formal analysis, investigation (lead), methodology (lead), project administration, resources (lead), supervision (co‐lead), validation, writing–reviewing and editing. Jill R. Crittenden: conceptualization, data curation, formal analysis, investigation, project administration, validation. Akshay Gupta: data curation, formal analysis. Joseph C. Johnson: data curation, formal analysis. Ananya S. Achanta: data curation, formal analysis. Smitha Bhagavatula: data curation, formal analysis. Johnny Loftus: data curation, formal analysis, investigation. Ara Mahar: data curation, investigation, resources. Dan Hu: investigation. Jesus Azocar: investigation, resources. Ryan J. Gray: data curation, formal analysis. Michael R. Bruchas: conceptualization, project administration, supervision, funding acquisition (co‐lead), writing–reviewing and editing. Ann M. Graybiel: conceptualization, project administration, supervision, funding acquisition (co‐lead), writing–reviewing and editing. Peer Review The peer review history for this article is available at https://publons.com/publon/10.1002/cne.70003. Supporting information Supporting Information Acknowledgments We thank Sophia Robertson for assistance with histological cell counting. We also thank Diego Pizzagalli for comments on the manuscript. This research was supported by NIH/NIMH P50 MH119467 (Project 3 to AMG; Project 4 to MRB), NIH/NIMH R01 MH060379 (AMG), NIH/NIDA F31 5F31DA059438‐02 (CS), Saks‐Kavanaugh Foundation (AMG), Hock E. Tan and K. Lisa Yang Center for Autism Research (AMG), Mallinckrodt Foundation (MRB), Simons Foundation grant to the Simons Center for the Social Brain at MIT (AMG). Emily Hueske, Carrie Stine, and Tomoko Yoshida contributed equally to this work. Funding: This research was supported by NIH/NIMH P50 MH119467 (Project 3 to AMG; Project 4 to MRB), NIH/NIMH R01 MH060379 (AMG), NIH/NIDA F31 5F31DA059438‐02 (CS), Saks‐Kavanaugh Foundation (AMG), Hock E. Tan and K. Lisa Yang Center for Autism Research (AMG), Mallinckrodt Foundation (MRB), Simons Foundation grant to the Simons Center for the Social Brain at MIT (AMG). Contributor Information Michael R. Bruchas, Email: mbruchas@uw.edu. Ann M. Graybiel, Email: graybiel@mit.edu. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Supporting Information Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.