Human brain tissue from a neurotypical 58-year-old male, previously analyzed in multiple studies (García-Cabezas et al., 2007; Sancha-Velasco et al., 2023; Uceda-Heras et al., 2024), was obtained post-mortem and cut in brain slabs within the stereotaxic space of Talairach and Tournoux (García-Cabezas et al., 2023). Brain slabs (1 cm thick) were postfixed in 4% paraformaldehyde for 24–48 h and then cryoprotected in 30% phosphate-buffered sucrose. Then, small blocks containing the diencephalic and mesencephalic regions were separated from the slabs, cryoprotected in sucrose solutions, and coronally sectioned at 50 μm thickness using a freezing microtome. Consecutive sections were processed to examine the cytoarchitecture, myeloarchitecture, and chemoarchitecture of the brain tissue using cresyl violet staining, silver myelin staining (Gallyas, 1979), cytochrome oxidase histochemistry (Wong-Riley, 1979) and acetylcholinesterase histochemistry (Cavada et al., 1995), respectively. Tyrosine hydroxylase (TH) detection was performed using a mouse monoclonal anti-TH antibody (MAB318; 1:200–1:400; Chemicon, Temecula, CA) followed by a rabbit anti-mouse secondary antibody (AB240; 1:30; Chemicon) and incubation in mouse peroxidase-antiperoxidase (PAP) (PAP14; 1:600; Chemicon). The reaction was developed and intensified, with TH revealed using the sensitive glucose oxidase-diaminobenzidine (DAB)-nickel method (See details in Sánchez-González et al., 2005).

Pregnant and adult Swiss mice, as well as pregnant, adolescent, and adult Sprague–Dawley (SD) rats, were obtained from the animal facilities at the University of Murcia. All animals were weighed and housed under identical conditions in standard cages (50 cm × 35 cm × 35 cm) with a 2–3 cm layer of dry cork bedding. Pregnant animals were sacrificed to obtain E12.5 mouse embryos and E13.5 rat embryos. The housing rooms were maintained at a temperature of 22–25°C with a relative humidity of 45–60%. Adolescent and adult animals had ad libitum access to a standard chow diet (ENVIGO, diet 2014, United States) and filtered water.

Swiss mouse and Sprague–Dawley rat brains were obtained and processed according to established protocols (Ferran et al., 2015a; Ferran et al., 2015b). The brains were perfused with a saline solution, followed by fixation with 4% paraformaldehyde in 0.1 M phosphate buffer (PB; pH 7.4). After extraction, the brains were fixed in 4% paraformaldehyde at 4°C for 16 h. Some brains were subsequently washed in phosphate-buffered saline (PBS) and cryoprotected in 15 and 30% sucrose solutions in 0.1 M PBS (pH 7.4). These brains were sectioned using a sliding microtome (Micron HM430, Thermo Scientific, United States) into 50 μm sagittal, horizontal, and transverse sections. Sections were collected as parallel or consecutive series on SuperFrost Plus slides (Menzel-Gläser, Braunschweig, Germany) and processed for hybridization and/or immunohistochemistry. Other brains were washed in PBS and embedded in 4% agarose (low electroendosmosis-EEO agarose; catalog No. 8008; Pronadisa, Spain) to obtain 100 μm vibratome sections. These sections were processed as free-floating samples for immunohistochemistry (Ferran et al., 2015a; Ferran et al., 2015b).

A Lim1 cDNA fragment was obtained through RT-PCR and cloned into a TA vector for subsequent RNA probe synthesis. Fresh postnatal mouse brain tissues were homogenized using the Precellys Evolution system (Bertin Technologies, France) with a single 20-s cycle at 6500 RPM in 2 mL tubes (CK14). Total RNA was extracted using the NZY Total RNA Isolation Kit (Nzytech, MB13402, Portugal) and treated with DNase I (Invitrogen, Cat. 18,068-015, United States). cDNA synthesis was performed using Superscript III reverse transcriptase (Invitrogen, Cat. 18080-044, Spain) and oligo dT-anchored primers. The resulting cDNA was used as a template for PCR amplification with Taq polymerase (Promega, Cat. M8305, Spain) and specific primers:

The amplified PCR products were cloned into the pGEM-T Easy Vector (Promega, Cat. A1360, Spain) and sequenced by ACTI (University of Murcia, Spain), yielding a 439 bp fragment (NCBI Accession Number: NM_008498.3, position 893–1,331).

Brain sections for in situ hybridization, obtained using a vibratome, were collected on SuperFrost Plus slides and processed following previously published protocols (Ferran et al., 2015a; Ferran et al., 2015b). Linear cDNA templates for Gbx2, Lim1 and Pax6 were generated by PCR amplification of cloned fragments (see details for Gbx2 probe and Pax6 in Puelles et al., 2016) and (Puelles, 2019). Labeled sense and antisense RNA riboprobes were synthesized using digoxigenin-11-UTP (Roche, Lewes, United Kingdom) (Ferran et al., 2015a; Ferran et al., 2015b).

A detailed protocol for the immunohistochemical reaction has been described (Ferran et al., 2015a; Ferran et al., 2015b). Briefly, tissue sections were first treated with 0.3% hydrogen peroxide to inactivate endogenous peroxidases. Primary antibodies —mouse anti-NeuN (MAB377, Sigma-Aldrich, 1:4000), rabbit anti-Calbindin (CB38, Swant, 1:4000), and rabbit anti-TH (NB300-109, Novusbio, 1:200, Bio-Techne R&D Systems, Spain)—were incubated overnight at 4°C in rat and/or mouse sections. After washing, sections were incubated for 2 h with biotinylated secondary antibodies (goat anti-rabbit IgG (H + L) and goat anti-mouse IgG (H + L), Vector Laboratories, BA-1000-1.5 and BA-9200-1.5, 1:200). A streptavidin-peroxidase complex (Vectastatin-ABC kit, Vector Laboratories, United States; PK4000) was then applied for 1 h at room temperature. Finally, peroxidase activity was visualized using 0.03% 3,3′-diaminobenzidine (DAB, Sigma, St. Louis, MO, United States) with 0.003% hydrogen peroxide. Antibody specificity was confirmed by previous studies for TH (Bilbao et al., 2022), NeuN (Mullen et al., 1992) and CB (Caballero et al., 2014). Additional control experiments, omitting the primary antibody, showed no residual immunostaining.

Processed in situ and immunohistochemistry sections were digitalized with a ScanScope CS digital slide scanner (Aperio Technologies, Vista, CA, United States). Size, contrast, brightness, and focus in the images were adjusted by applying Adobe Photoshop CS3. Figures were produced using Adobe Illustrator CS2 (Adobe Systems Inc., San Jose, CA, United States).

According to modern interpretations, the regionalization of the forebrain results in larger proneuromeric regions along the anteroposterior axis. These regions, from rostral to caudal, can be identified as the secondary prosencephalon, the diencephalon proper, and the midbrain (Figure 1B) (Albuixech-Crespo et al., 2017; Ferran and Puelles, 2018; Puelles, 2018; Ferran et al., 2022). The rostral proneuromere, known as the secondary prosencephalon, gives rise to two neuromeres: the peduncular hypothalamic prosomere (hp1) and the terminal hypothalamo-telencephalic prosomere (hp2) (Figure 1C) (Puelles et al., 2012a; Ferran et al., 2015c; Puelles and Rubenstein, 2015). Adjacent to these and within the diencephalon proper, there are three diencephalic prosomeres (Figure 1C, dp1–dp3). These neuromeres are also known as “pretectum” (dp1), “thalamus” (dp2), and “prethalamus” (dp3) (Puelles and Rubenstein, 1993; Rubenstein et al., 1994; Puelles et al., 2012b). Finally, in the most caudal proneuromere of the forebrain (often identified as the midbrain) there are 2 midbrain prosomeres (or mesomeres) (Figure 1C, mp1 and mp2) (Puelles E. et al., 2012c; Puelles, 2013; Watson and Puelles, 2017; Puelles, 2018, 2019; Puelles and Hidalgo-Sánchez, 2023). In the rhombencephalic region (hindbrain), the developing brain is organized into 13 distinct rhombomeres (Figure 1C, r0–r11) (Puelles, 2013; Puelles, 2018; Puelles and Hidalgo-Sánchez, 2023). This segmentation provides a developmental framework for understanding the organization of the forebrain and midbrain, including the spatial distribution of molecular markers from which structures like the SN and VTA emerge.

To assess whether the SN and VTA follow a polyneuromeric distribution, we will first define the key molecular and anatomical features of the midbrain and neighboring neuromeres (Figures 2A–C). The midbrain alar plate comprises four rostrocaudal rather than just the two classic colliculi: the tectal gray, superior colliculus, inferior colliculus (all within mp1), and alar preisthmus within mp2, but the basal plate contains the oculomotor nucleus complex in mp1 (Figures 1B,C, 2A–C) (Puelles E. et al., 2012c; Puelles, 2013; Watson et al., 2017; Puelles, 2018, 2019; Puelles and Hidalgo-Sánchez, 2023). The rostral boundary of the midbrain, which borders the diencephalic pretectal region (dp1), is discernible from early developmental stages in all vertebrates by the caudal limit of Pax6 expression in the alar plate (Ferran et al., 2007; Ferran et al., 2008; Ferran et al., 2009; Merchan et al., 2011; Morona et al., 2011, 2017; Brożko et al., 2022; Ferran and Puelles, 2024). Throughout later stages, anatomical landmarks can be used to identify this boundary. The pretecto-midbrain or diencephalon-midbrain boundary is defined by a plane extending from the dorsal roof plate to the ventral floor plate. This plane passes behind the posterior commissure in the alar plate, but anterior to the oculomotor complex (3 cranial nucleus) and between the red parvocellular and magnocellular nuclei, in the basal plate (the parvocelullar nucleus is in the basal plate of dp1 and the magnocellular nucleus in the basal plate of mp1) (Figure 2C) (Ferran et al., 2007; Ferran et al., 2009; Puelles E. et al., 2012c; Puelles, 2013; Watson et al., 2017; Puelles, 2018, 2019; Puelles and Hidalgo-Sánchez, 2023). The caudal midbrain boundary adjoins the isthmic region which contains the pathetic nucleus (4 cranial nucleus) and its associated nerve fibers (Watson et al., 2017; Puelles, 2019). Additionally, the interpeduncular nucleus which spans the isthmic (r0) and r1 rhombomeres, and the decusation of the superior cerebellar peduncle observed in the isthmic rhombomere help to identify the midbrain-isthmic boundary (dscp) (Figure 2C) (Lorente-Canovas et al., 2012). Key diencephalic landmarks can also help to identify interneuromeric boundaries. The retroflex tract (or habenulo-interpeduncular tract) marks the dp1-dp2 border, separating thalamus (dp2) from pretectum (dp1) (Figure 2C). This tract extends dorsoventrally from the habenula to the basal plate, passing through the caudal part of dp2 just rostral to dp1, but then along the basal plate to the interpeduncular nucleus (Figure 2C) (Ferran and Puelles, 2024). On the rostral side of the dp2, the mammillothalamic tract (mth), spanning from the mamillary body to the anterior thalamic complex, has a portion that defines the dp2-dp3 boundary (Figure 2C). Finally, the fornix tract (fx), coursing through rostral hp1, marks the hp1-hp2 boundary (Bilbao et al., 2022).

At early embryonic stages (E12.5 in mice and E13.5 in rats), the main neuromeric boundaries can be identified molecularly, aiding in the detection of TH-positive neurons that will later contribute to the adult substantia nigra pars compacta (SNc) and ventral tegmental area (VTA) (Figure 3). Cells belonging to the SNc primordium (SNcp) and VTA primordium (VTAp) are observed in both mice and rats within the diencephalic prosomeres (dp1-dp3), midbrain prosomeres (mp1-mp2) and the rostral hindbrain rhombomere (r0) (Figures 3A–I). Several studies have established molecular markers at this developmental stage to define interneuromeric boundaries in the diencephalon, midbrain and rhombencephalon (see previous section for details). For instance, Gbx2 expression delineates the rostral p2/p3 boundary and the caudal p2/p1 boundary at E12.5 (Figure 3E). Also, Pax6 is expressed in the alar plate of dp1 (pretectum), with its caudal border making the diencephalic-midbrain border (dp1-mp1) at E12-5 (Figure 3F). Finally, Lim1 expression further aids in identifying both the diencephalic-midbrain boundary and the p2/p3 boundary across the alar and basal plates (Figure 3G).

During the adolescent and adult stages of rodents (mice and rats) TH-positive neurons within the SN and VTA persist in locations similar to those observed in early developmental stages. These neurons remain situated in the basal plate across all diencephalic prosomeres (dp1-dp3), midbrain prosomeres (mp1-mp2), and the rostral rhombomere (r0) (Figures 4A–C, 5A–E, 6A–C, 7A–E). Some anatomical landmarks, as described in the previous section, help delineate the interneuromeric boundaries. The mammilo-thalamic tract, situated in the rostral part of the diencephalic prosomere 2 (dp2), serves as a landmark for the dp2/dp3 boundary. This boundary demarcates distinct TH-positive neuronal populations belonging to the SNc and VTA and located in the basal plate of dp2 and dp3 neuromeres (Figures 5E, 7B). Just caudal to the retroflex tract (rf) lies the boundary between the dp1 and dp2 neuromeres. This tract extends from the dorsal alar plate to the ventral basal plate, dividing the SNc and VTA TH-positive neurons into a rostral dp2 group and a caudal dp1 group (Figures 4B, 5E, 6B, 7C). The border between diencephalon and midbrain (di-mb or dp1-mp1) lies caudal to the posterior commissure (alar plate; Figures 4C, 5B, 6A,B, 7D), and rostral to the oculomotor complex (3N) (basal plate; Figure 6B). In the basal plate, the border separating the parvocellular (RPM, dp1) and magnocellular (RMN, mp1) red nuclei (Figure 6B), delineates TH-positive neurons within the SN and VTA into dp1 and mp1 neuromeres (Figures 5E, 6B). The RMC caudal border indicates the location of mp2 (Figure 6B). The location of the decusation of the superior cerebellar peduncle (dscp) and the interpeduncular nucleus in r0 recognizes in their anterior parts the boundary between midbrain (mp2) and isthmic region (r0) (Figures 4B,C, 6A,B, 7B–D).

To determine the distribution of SN and VTA neuronal groups in the primate Macaca mulatta, we analyzed two Nissl-stained sagittal sections obtained from BrainMaps.org . The selected sections, which pass through SN or SN/VTA regions, reveal their distribution in the basal plate of distinct neuromeres, including diencephalic (dp1-dp3), midbrain (mp1-mp2), and rostral hindbrain (r0) segments (Figures 8A–C). The most rostral group is located in the basal plate of dp3, caudal to the subthalamic nucleus (Sth) (Figure 8A). Key anatomical landmarks help identifying interprosomeric boundaries, such as the mammillothalamic tract (mth) between dp2 and dp3, and the retroflex tract (rf), marking the boundary between dp1 and dp2 (Figure 8B). At the alar plate level of dp1 (pretectal region), the posterior commissure (pc) is visible, with its caudal border demarcating the dp1/mp1 (diencephalon-midbrain) boundary (Figure 8B). Additionally, the parvocellular and magnocellular red nuclei serve as markers for the dp1/mp1 border (Figure 8B). The sections also highlight the oculomotor complex (3 N) in the midbrain basal plate and its fibers (3n) extending toward the interpeduncular fossa (Figure 8B). Both 3 N and 3n indicate the dp1/mp1 border rostrally and the mp1/mp2 border caudally. The mp2/r0 border can be recognized by the rostral parts of the decusation of the superior cerebellar peduncle (dscp) and the IP nucleus (Figures 7A,B). This preliminary analysis suggests that a significant population of SN and VTA neurons is situated in the most rostral rhombomere (r0) (Figures 8A–C).

Next, we analyzed human brain sections stained with Nissl, cytochrome oxidase (CO), tyrosine hydroxylase (TH) and acetylcholinesterase (ACTH) histochemistry, focusing on regions encompassing through the subthalamic nucleus. These sections allowed us to examine the spatial relationships between de SN and VTA with diencephalic prosomeres (Figures 9A–E´). Additionally, we studied more caudal sections to assess their association with the midbrain prosomere 1 (mp1) (Figures 10A–E´). These coronal sections were cut perpendicular to a reference plane defined by two anatomical landmarks: the superior border of the anterior commissure (ac) and the inferior border of the posterior commissure (pc) (Figures 9E, E´, 10E,E´). Rostral sections traversing the Sth, reveal its location within the hypothalamo-telencephalic prosomere 1 (hp1), which is further characterized by cerebral peduncle (cp) fibers coursing along its surface (Figures 9A–D). Adjacent to hp1, dp3 is identified by the presence of the reticular nucleus in its alar plate (Figure 9D). The dp3 alar plate continues to its basal plate bordering the subthalamic nucleus and extending to the surface at the cp (Figures 9A–D). The extensive thalamic nuclear region (Th, dp2) and parvocellular red nucleus (RPN, dp1) are readily identifiable through distinct anatomical landmarks, confirming the presence of both neuromeres in these sections. Based on this anatomical analysis, we concluded that the SN primarily localizes to dp3 and dp2 prosomeres, while VTA and possibly a small population of SN neurons occupy dp1 (Figures 9A–D). In the analyzed caudal sections, the dp3 prosomere is clearly delineated by the presence of the reticular nucleus within its alar plate (Figure 10D). The alar plate of dp2 prosomere is identified by the presence of the extensive thalamic (Th) complex, the habenular region (Hb), and the lateral geniculate nucleus (LG) (Figures 10A–D). At the basal plate level, we observe an important distinction: while the parvocellular red nucleus marks dp1 territory, the magnocellular red nucleus identifies mp1 (midbrain prosomere 1). This organization reveals that: (1) a distinct SN population resides within dp1 and (2) both SNc and VTA neuronal groups are predominantly located in mp1 at this level of section. The midbrain mp1 prosomere is further defined by the path of oculomotor nerve fibers (3n) traversing this region (Figures 10A–D). These findings demonstrate that in adult humans, both the VTA and SN span multiple neuromeric domains, including diencephalic prosomeres (dp1-dp3), midbrain prosomeres (mp1-mp2) and the rostral hindbrain rhombomere (r0) (Figures 9E, 10E´).