In this work, we show the existence of VENs-like cells in Wistar rats (Rattus norvegicus). These neuronal elements are found in layers III and V of the ACC and in layers III and V of the granular and agranular regions of the FI cortex. These cells have a fusiform shape with the same morpho-histological characteristics as those described by Betz (1881), Ramón y Cajal (1899) and later by von Economo and Koskinas (1925) and that have been reported in other species, including humans, great apes, cetaceans, elephants, manatees, walruses, horses, zebras, pygmy hippopotamuses, black rhinos, cows, pigs, macaques, deer, sheep, rock hyraxes and Indian parrots (Nimchinsky et al. 1999; Hof and Van Der Gucht 2007; Hakeem et al. 2009; Butti et al. 2009; Allman et al. 2010; Butti and Hof 2010; Evrard et al. 2012; Raghanti et al. 2015; Shubha and Sudhi 2017; Banovac et al. 2019; Correa-Júnior et al. 2020).
In this study, the identification was made by NADPH-d histochemistry, immunohistochemistry for nNOS, eNOS, NeuN, dopamine D3R, GABRQ and Nissl staining.
NADPH-d activity in the brain is mainly represented by the expression of nNOS (Dawson et al. 1991; Hope et al. 1991) and less reliably by the expression of eNOS (Dinerman et al. 1994) and inducible NOS (iNOS) (Tracey et al. 1993), which implies visualizing all NOS enzyme types (neural, endothelial, and inducible). To determine which type of specific NOS a cell contains, it is necessary to use antibodies for each enzyme. Therefore, we used anti-nNOS and anti-eNOS antibodies. We showed that some spindle-shaped and fork neurons were immunoreactive to nNOS, but none were immunoreactive with the eNOS antibody. Some authors have specifically reported the absence of VENs in rodents (Nimchinsky et al. 1999; Hakeem et al. 2009), while our results show the contrary. The explanation of the discrepancy between these works and ours may be due to the use of NADPH-d histochemistry, which allowed us to see the morphology of these neurons in the cortex of the rat more clearly. Another possibility may be that their nNOS antibody did not have the same sensitivity as ours.
The use of NADPH-d was key to revealing the presence of VENs-like cells in the rat. This staining also revealed the close anatomical interaction of these neuronal elements with capillary blood vessels. Moreover, the endplates of these neuronal extensions seem to hug the capillary, suggesting that these neurons could regulate vasomotion for regional metabolic demand, which has been proposed by other authors (Iadecola et al. 1993; Estrada and DeFelipe 1998; Larriva-Sahd et al. 2019). Although the role of astrocytes in the mediation of cerebral vascular flow has been well documented (Zonta et al. 2003; Takano et al. 2006), our findings may extend this regulatory ability to VENs.
Another important finding in this study is the presence of fork cells in the neighbourhood of the VENs-like cells, as described in other species. These cells were also positive for NADPH-d (Fig. 1a5 and Fig. 3b).
Evrard et al. (2012) reported the morphology of a VEN by retrograde labelling with Alexa 594, in which dendritic ramifications emerging from the neuronal body were observed. Similarly, we identified such a morphology in images obtained with NADPH-d staining in the ACC and FI of the rat. This same observation was described by von Economo (1926) in his original communication. This suggests a new way of visualizing the decimononic forms presented by von Economo and Koskinas (1925) for defining VENs and broadens the concept of connectivity, particularly with vascular elements (Fig. 1a1, a2, h, i and Fig. 2b). A description of a continuum shape of VENs in humans classified into three categories has been recently reported (Correa-Júnior et al. 2020): “VEN 1, which has main dendritic shafts but few branches and sparse spines; VEN 2, which shows an intermediate aspect; and VEN 3, which displays the most profuse dendritic ramification and more spines with varied shapes from proximal to distal branches”. In our case, all the VENs-like cells corresponded to the morphology between VENs 1–2 of the Correa-Junior et al., (2020) classification, as shown in Fig. 2.
We demonstrated the colocalization of nNOS and NeuN in some VENs-like cells and fork nerve cells found in layers III and V with double immunofluorescence, which indicated that these cells were neurons (Fig. 4). However, there were more NeuN-positive than nNOS-positive spindle-shaped and fork neurons, and more VENs-like cells demonstrated Nissl staining than lacked NADPH-d activity. In addition, more VENs-like cells were NADPH-d–positive than nNOS immunoreactive. Given the above, several assumptions can be made: a) not all VENs-like cells produce nitric oxide; b) not all nitrergic spindle-shaped neurons are immunoreactive for nNOS; and c) VENs-like cells are not immunoreactive for eNOS, which suggests that some fusiform neurons could express iNOS and produce NO on demand. To demonstrate this, more research is needed with a specific marker for iNOS and specific tasks to stimulate the synthesis of iNOS in these neurons.
We also found that the distribution of NADPH-d and nNOS-IR neurons in adult rats was similar to that reported in other studies (Bredt et al. 1991; Cha et al. 1998). Chung, Kim and Lee (2004) also describe the presence of nNOS-IR in layers II, III and V. As mentioned above, it has been reported that nitrergic neurons in the neocortex are only interneurons involved in local circuits; however, it is known that nitrergic neurons have heterogeneous morphological patterns and may have different biochemical markers and neurophysiological functions in all rat brains. Since the presence of nitrergic projection neurons in the cortex and hippocampus has been described (Vaid et al. 1996; Gerashchenko et al. 2008; Tricoire and Vitalis 2012; Funk and Kwan 2014; Stern et al. 2021), it is possible that the neurons we are describing are projection neurons that participate in extracortical circuits such as VENs. This last point will require further investigation.
Another remarkable finding is the presence of D3R and GABRQ immunoreactivity in a similar spindle-shaped neuron of the ACC layer V, as has also been reported in humans and primates, which have been defined as a part of receptors that are present in von Economo neurons (Allman et al. 2005, 2010; Evrard et al. 2012; Dijkstra et al. 2018; Gami-Patel et al. 2022).
Neurons related to functions of high cognitive-emotional order, language and other functions have been associated with NO production (Funk and Kwan, 2014). The emission of social vocalizations is an evolutionary activity in vertebrates, in which ethotransmission is species-specific and communicates the emotional state of the individual (Brudzynski 2013, 2019). Rats have an analogous mechanism of speech and vocalizations, identified as ultrasonic vocalizations, that express negative or positive emotional states in at least two different ranges of frequencies (Brudzynski 2013). As has been reported, speech and language are closely related to neurons that express NOS in mammals. In this sense, the presence of VENs-like cells that are also positive for NOS in rats supports our predictions.
Paradoxically, the presence of VENs-like cells in the rat had not been previously reported, despite the rat having been used as an animal model in learning and memory studies, even with cognitive challenges (Arakawa and Iguchi 2018; Arakawa 2019); they also display complex maternal and empathic social behaviours (Kaffman and Meaney 2007; Mohammadi et al. 2020) similar to those displayed by species known to have VENs, so it would be surprising not to find them in the rat or related rodents like mice. In this sense, it has been reported that mouse frontotemporal dementia models are effective in replicating the behaviours and morphological findings reported in humans (Roberson 2012; Vernay et al. 2015). Moreover, Hodge et al. (2020) suggested that in these rodent models, there could be cells homologous to VENs with a different morphology and with similar circuit function. However, the morphology of the spindle-shaped/fusiform cells reported in this work are similar to those described as VENs in humans and in other species. As reported by Raghanti et al. (2015), the presence of VENs in diverse species indicates that this type of neuron represents a phylogenetically ancient neuron rather than an emergent specialized neuron. This suggests that not all the characteristics present in human VENs could be found in VENs like cells in phylogenetically more distant species.
The presence of VENs in this mammal (Rattus norvegicus) is a crucial neuronal element that implies the emergence of self-awareness and social cognition in rodents.