Distribution of Serotonergic Transporter Innervation in the Nucleus Accumbens and Ventral Pallidum Is Highly Conserved Among Primates

ABSTRACT

The nucleus accumbens (NAcc) and ventral pallidum (VP) are key nodes in the mesolimbic reward pathway that facilitate stimulus salience, including the regulation of social motivation and attachment. Primate species display variation in social behaviors, including different levels of impulsivity, bonding, and aggression. Previous research has implicated neuromodulation of the reward pathway in the differential expression of various social behaviors, suggesting that differences in neurotransmitter innervation may play a role in species‐specific patterns. To explore this, we examined serotonergic innervation in the NAcc and VP among primates. We used stereology to quantify serotonin transporter‐immunoreactive (SERT‐ir) axon length density in the NAcc and VP of 13 primate species, including humans, great apes, and cercopithecid and platyrrhine monkeys. Our data show that serotonergic innervation density within both the NAcc and VP is highly conserved among species. This finding contrasts with our previous findings of higher levels of SERT‐ir axons in the dorsal striatum of humans and great apes relative to monkeys, a human‐specific increase in dopaminergic innervation within the NAcc and VP, and a human‐specific increase of neuropeptide Y in the NAcc, highlighting the mosaic nature of innervation patterns among species.

We quantified serotonin transporter (SERT) axon length density in the reward pathway of 13 primate species. Serotonergic innervation density within both the nucleus accumbens and ventral pallidum is highly conserved among species, in contrast to our previous findings of human‐specific differences in dopaminergic and neuropeptide Y innervation.

1. Introduction

One of the many functions of the central serotonergic system is to guide social interactions (Harmer et al. 2004; Hysek et al. 2014; Knutson et al. 1998) through its actions with the mesolimbic reward pathway (Dölen et al. 2013; Tanaka et al. 2007; Virk et al. 2016). The key structures in this network include the anterior cingulate (ACC) and orbital prefrontal (OFC) cortices, the ventral striatum, the ventral pallidum (VP), and the midbrain dopamine neurons (Haber and Knutson 2010). Humans and nonhuman primates are highly social species that rely heavily on behavioral inhibition and emotional regulation to facilitate bonding and partnership (Brosnan and de Waal 2014; Crockett et al. 2013; Higley et al. 1996; Raleigh and McGuire 1991). Differential serotonergic innervation of regions of the striatum is implicated in impulse control (Miyazaki et al. 2014; Raleigh and McGuire 1991), personality style (Erritzoe et al. 2019; Kanen et al. 2021; Plieger et al. 2014), and aggression (Higley et al. 1996; Weld et al. 1998). Humans and apes have comparatively higher levels of serotonin than monkeys in the dorsal striatum (Raghanti et al. 2018), potentially contributing to their complex social cognition.

To build on this previous work, and because the mesolimbic reward pathway is central to species‐specific behavior, we chose to examine serotonergic innervation of the nucleus accumbens (NAcc) and VP, key nodes of the reward pathway. The NAcc is part of the ventral striatum and receives dopaminergic inputs from the ventral tegmental area (VTA) and substantia nigra (SN). The VP receives inputs from the NAcc and sends projections to several regions, including the mediodorsal thalamus, lateral habenula, ventral striatum, VTA, SN, and subthalamic nucleus (Haber 2011). The NAcc plays a role in motivational and emotional responses to environmental stimuli, and the VP is involved in encoding the value of reward as well as the motor actions to attain the reward (O'Doherty et al. 2004; Pagnoni et al. 2002). The reward pathway is central to mediating predictability and risk information, along with assessing likely outcomes from different behavioral choices. Higher‐order cortical and subcortical forebrain structures are engaged in making complex choices about fundamental needs and choices involving secondary rewards (Corlett et al. 2004; de Visser et al. 2011; Elliott et al. 2003; Haber 2011).

Although serotonergic innervation is widespread throughout the brain, variation and regional specificity are made possible through several distinct mechanisms. The serotonergic system signals through seven classes of receptors with several subtypes (Marin et al. 2020), all of which are G protein‐coupled receptors (GPCRs) except for the 5‐HT3 cation channel (Derkach et al. 1989; McCorvy and Roth 2015). Serotonin receptors are found at both pre‐ and postsynaptic locations and interact with various intercellular pathways, including adenylyl cyclase (Albert et al. 1999; Bouhelal et al. 1988), phospholipase C (Conn and Sanders‐Bush 1984), and Janus kinase‐2/signal transducers and activators of transcription (JAK2/STAT3) (Singh et al. 2010). Some individual serotonergic neurons send axon collaterals to more than one brain region to integrate specific circuits, as seen in the dorsal raphe projections to NAcc and prefrontal cortex (PFC) (Van Bockstaele et al. 1993) as well as to the caudate–putamen and nucleus reticularis gigantocellularis pars α in rat (Li et al. 2001). In mice, genetically defined subgroups of serotonergic neurons have local axon collaterals within the dorsal raphe that likely participate in autoregulatory feedback networks of distinct systems (Bang et al. 2012). Each of these molecular mechanisms may contribute to a variety of behaviors.

Dopamine in the NAcc is also known to promote prosocial behaviors (Heifets et al. 2019), and dopaminergic innervation is increased in the human dorsal and ventral striatum relative to that of nonhuman primates (Hirter et al. 2021; Raghanti et al. 2016). Serotonergic modulation of the dopaminergic system is implicated in natural reward and drug abuse (Bowers et al. 2000; Deurwaerdère et al. 2004; Navailles et al. 2008; O'Dell and Parsons 2004; Yan and Yan 2001). It is unknown if differences also exist among primates in serotonin levels in the NAcc and VP.

The VP facilitates behavioral responses to reward by coding hedonic reward and reward learning in rats (Smith and Berridge 2007; Tindell et al. 2004, 2006). Neuronal firing within the VP of rats correlates with the speed of upcoming approach movement, proximity to the target at cue onset, and efficiency of approach movement path (Lederman et al. 2021). Serotonin in the VP attenuates locomotor and rearing behaviors in rats through the 5‐HT‐2C receptor subtype (Graves et al. 2013), and the 5‐HT‐1A receptor can either enhance or suppress neuronal firing (Heidenreich and Napier 2000). The VP also contributes, through 5‐HT‐2A receptors, to the modulation of prepulse inhibition of the startle response (PPI), an important measure of sensorimotor gating (Sipes and Geyer 1997).

The central serotonergic system has undergone considerable modifications across taxa (E. C. Azmitia 2020). In rodents, large clusters of cell bodies act together to influence general circuits, but in primates, distinct raphe clusters send myelinated outputs to support multiple discrete cortical functions. Humans may have more precise distributions (Azmitia 1987; E. C. Azmitia and Segal 1978; E. Azmitia and Gannon 1983). The dorsal raphe serotonergic system innervates most of the forebrain, including the reward pathway. Ascending projections from the raphe nuclei display species‐specific differences (Figure 1). In mice, the projections from the dorsal component of the dorsal raphe (B7) pass through the medial longitudinal fasciculus (MLF) to the NAcc along with other parts of the basal ganglia and are mainly, but not exclusively, ipsilateral. Relatively sparse projections from the B9 supralemniscal raphe group pass through the medial forebrain bundle (MFB) to the NAcc bilaterally with an ipsilateral predominance (Muzerelle et al. 2016). The MFB is the major ascending serotonergic pathway in rats and has a similar absolute size in long‐tailed macaques. Within the MFB, macaques display a greater percentage (25%) of myelinated versus unmyelinated serotonin‐ir axons compared to rats (0.7%). Used sparsely in rats, the dorsal raphe cortical tract (DRCT) is much larger in long‐tailed macaques (E. Azmitia and Gannon 1983). In contrast to macaques, the ascending serotonergic fiber pathways in squirrel monkeys follow a similar path to that of humans. Thick fascicles mainly from the dorsal, and less so from the medial raphe, ascend through the superior cerebellar peduncle and VTA and pass through the MFB in the lateral hypothalamic area (Parent et al. 2011; Wallman et al. 2011). Species‐specific differences are apparent in the anatomy of serotonergic projection patterns, including laminar specificity differences within the lateral geniculate nucleus and cortical area 17 of cynomolgus versus squirrel monkeys (Blasdel and Lund 1983; Morrison and Foote 1986). It is unknown if species differences exist in innervation density levels of serotonin projections to the reward pathway.

FIGURE 1.

Diagram of relative proportions of three ascending serotonergic pathways in mouse, rat, long‐tailed macaque, and human with squirrel monkey. Diagram is based on separate studies in mouse (Muzerelle et al. 2016), rat and macaque (E. Azmitia and Gannon 1983), squirrel monkey and human (Parent et al. 2011), as well as human alone (Wallman et al. 2011). MLF, medial longitudinal fasciculus; DRCT, dorsal raphe cortical tract; MFB, medial forebrain bundle.

Axon density measures are a way to compare innervation levels among various species and may reflect functional differences. Previous studies have reported on the differences among species for serotonergic innervation within various regions of the brain. Stimpson et al. (2015) found that bonobos have more than twice the density of serotonergic axons within the amygdala than chimpanzees, with most pronounced differences in the basal and central nuclei. This difference is thought to affect reactivity to the social environment through modulation of major output nuclei that project to cortical and autonomic centers. Bonobos are more tolerant of conspecifics in competitive contexts, whereas chimpanzees more frequently use aggression during conflicts (Hare et al. 2007; Surbeck et al. 2017). Humans have a higher density of serotonergic axons than chimpanzees and bonobos in the central and accessory basal nuclei of the amygdala, which may suggest a human‐specific emphasis on the modulation of amygdaloid circuitry between cognitive and autonomic pathways by the serotonergic system (Lew et al. 2019). Humans and apes together have higher serotonergic axon density than monkeys in dorsal striatal regions, potentially contributing to social strategic skills necessary for bonding and partnership (Raghanti et al. 2018). Humans and chimpanzees also have a greater density of serotonergic axons in layers V and VI in prefrontal cortical areas 9 and 32 compared to macaque monkeys (Raghanti et al. 2008). These regions of the cortex are important for working memory and theory of mind (Gallagher and Frith 2003; Petrides 2000; Stuss et al. 2001), functions that support human and great ape cognitive specializations important for social behavior. Within the reward pathway, humans have higher levels of dopamine innervation in both the NAcc and VP than nonhuman primates, which is thought to promote behavioral flexibility and mediate incentive salience (Hirter et al. 2021). Humans also have elevated neuropeptide Y within the NAcc relative to nonhuman primates, potentially supporting sociality and a preference for dietary fat to support enlarged brains (Raghanti et al. 2023). While these studies show differences among a few species, we chose to examine 13 species to gain a clearer view of evolutionary relationships.

2. Materials and Methods

2.1. Specimens

Postmortem brain samples from 88 individuals representing 13 species were used for the present study, including platyrrhine monkeys (owl monkeys, cotton‐top tamarins, common marmosets, and tufted capuchins), cercopithecid monkeys (rhesus macaques, Japanese macaques, pig‐tailed macaques, moor macaques, and olive baboons), African great apes (western lowland gorillas, chimpanzees, and bonobos), and humans (Table 1). Four to seven individuals per species were used, with the sexes balanced as equally as possible. All individuals were adults, free of gross neuropathology, and samples were taken from the left hemisphere. Human brain samples were provided by the National Disease Research Interchange (NDRI), the Northwestern University Alzheimer's Disease Center Brain Bank, and the El Paso County Coroner in Colorado (as approved by the Colorado College Institutional Review Board, #011311‐1). Nonhuman primate brains were received from American Zoo and Aquarium‐accredited zoos or research institutions and maintained in accordance with each institution's animal care and use guidelines. All animals died from either natural causes or from humane euthanization for medical reasons. The acquisition and processing of postmortem human and nonhuman brain materials are exempted from the requirement of approval by institutional animal care and human subject use committees. All postmortem intervals were less than 17 h. The moor macaques were perfused transcardially with 4% paraformaldehyde according to methods previously described (Hof et al. 1996; Hof and Nimchinsky 1992) as part of unrelated experiments. All other brain samples were immersion‐fixed in 10% buffered formalin for a minimum of 7 days, then moved to a 0.1 M buffered saline solution with 0.1% sodium azide, and finally stored at 4°C until further processing.

TABLE 1.

List of individuals including sex and age.

Species	Common name	Sex	Age (years)
Callithrix jacchus	Common marmoset*	M	2.75
Callithrix jacchus	Common marmoset*	F	6.2
Callithrix jacchus	Common marmoset+	M	5.7
Callithrix jacchus	Common marmoset	F	5.1
Callithrix jacchus	Common marmoset	F	11.3
Callithrix jacchus	Common marmoset	M	6
Callithrix jacchus	Common marmoset	M	4.5
Callithrix jacchus	Common marmoset+	F	6.11
Callithrix jacchus	Common marmoset+	M	8.5
Saguinus oedipus	Cotton‐top tamarin+	F	9.9
Saguinus oedipus	Cotton‐top tamarin*	M	10.9
Saguinus oedipus	Cotton‐top tamarin+	M	8.4
Saguinus oedipus	Cotton‐top tamarin+	M	9.5
Saguinus oedipus	Cotton‐top tamarin	F	8
Saguinus oedipus	Cotton‐top tamarin	F	6
Saguinus oedipus	Cotton‐top tamarin	F	16
Aotus trivirgatus	Northern owl monkey	M	>18
Aotus spp.	Owl monkey	M	18
Aotus spp.	Owl monkey	F	3
Aotus vociferans	Spix's night monkey	F	5
Sapajus apella	Tufted capuchin	M	2.9
Sapajus apella	Tufted capuchin	M	16.6
Sapajus apella	Tufted capuchin	M	15.9
Sapajus apella	Tufted capuchin	F	12.6
Sapajus apella	Tufted capuchin	F	17.5
Sapajus apella	Tufted capuchin	F	18.3
Macaca mulatta	Rhesus macaque*	M	8
Macaca mulatta	Rhesus macaque	M	13
Macaca mulatta	Rhesus macaque	M	13
Macaca mulatta	Rhesus macaque	F	14
Macaca mulatta	Rhesus macaque	F	11
Macaca mulatta	Rhesus macaque	F	12.5
Macaca nemestrina	Pigtailed macaque	M	15.73
Macaca nemestrina	Pigtailed macaque	M	4.28
Macaca nemestrina	Pigtailed macaque	M	2.5
Macaca nemestrina	Pigtailed macaque	F	15.1
Macaca nemestrina	Pigtailed macaque	F	9
Macaca nemestrina	Pigtailed macaque	F	5.95
Macaca maura	Moor macaque*	M	8
Macaca maura	Moor macaque	F	5
Macaca maura	Moor macaque	F	7
Macaca maura	Moor macaque	F	7
Macaca maura	Moor macaque	F	8
Macaca maura	Moor macaque	M	10
Macaca fuscata	Japanese macaque	M	19
Macaca fuscata	Japanese macaque	F	19
Macaca fuscata	Japanese macaque	F	14
Macaca fuscata	Japanese macaque	F	10
Macaca fuscata	Japanese macaque	M	11
Macaca fuscata	Japanese macaque	M	9
Papio anubis	Olive baboon	F	9.5
Papio anubis	Olive baboon	M	7
Papio anubis	Olive baboon	M	13
Papio anubis	Olive baboon	F	8
Papio anubis	Olive baboon	M	5.96
Papio anubis	Olive baboon	F	12
Gorilla gorilla	Western gorilla*	M	40
Gorilla gorilla	Western gorilla*	M	21.5
Gorilla gorilla	Western gorilla*	F	40
Gorilla gorilla	Western gorilla*	M	49
Pan troglodytes	Chimpanzee*	M	28.7
Pan troglodytes	Chimpanzee	M	19.3
Pan troglodytes	Chimpanzee*	M	20.3
Pan troglodytes	Chimpanzee+	M	24
Pan troglodytes	Chimpanzee+	M	26.9
Pan troglodytes	Chimpanzee+	M	25.3
Pan troglodytes	Chimpanzee*	F	27
Pan troglodytes	Chimpanzee*	M	21
Pan troglodytes	Chimpanzee*	F	12
Pan troglodytes	Chimpanzee+	F	36.2
Pan troglodytes	Chimpanzee+	F	35.4
Pan paniscus	Bonobo*	M	34
Pan paniscus	Bonobo+	M	5
Pan paniscus	Bonobo	M	25
Pan paniscus	Bonobo+	M	34
Pan paniscus	Bonobo*	F	25
Pan paniscus	Bonobo	F	52
Homo sapiens	Human*	M	53
Homo sapiens	Human	M	56
Homo sapiens	Human	M	44
Homo sapiens	Human*	M	44
Homo sapiens	Human*	M	42
Homo sapiens	Human	F	25
Homo sapiens	Human+	F	23
Homo sapiens	Human+	F	49
Homo sapiens	Human+	F	33

^*Denotes NAcc only; +denotes VP only.

2.2. Sample Processing

All samples were cryoprotected in a graded series of sucrose solutions (10%, 20%, and 30%) prior to sectioning. Brains were frozen in dry ice and cut into 40‐µm‐thick sections using a Leica SM2000R freezing sliding microtome (Buffalo Grove, IL). Sections were placed into sequentially numbered, individual microcentrifuge tubes containing freezer storage solution (30% distilled water, 30% ethylene glycol, 30% glycerol, and 10% 0.244 M phosphate‐buffered saline [PBS]) and stored at −20°C until further processing. Every 10th section was Nissl‐stained with 0.5% cresyl violet to visualize boundaries of the regions of interest for immunohistochemical staining and stereological quantification of neuron cell densities (Nv).

Figure 2 shows the two areas of interest for the current study. The NAcc is defined as the portion of the ventromedial striatum ventral and medial to the inferior border of the internal capsule in sections rostral to the anterior commissure, and the VP is defined as the area immediately ventral to the anterior commissure (S.‐L. Ding et al. 2016; Haber 2011; Paxinos et al. 2000, 2012).

FIGURE 2.

Photomicrographs showing Nissl‐stained sections including the NAcc in human‐56 (a, e), baboon‐12 (b, f), pig‐tailed macaque‐15 (c, g), and capuchin‐16 (d, h). Photomicrographs including the VP in chimpanzee‐36 (i, m), rhesus macaque‐14 (j, n), Japanese macaque‐9 (k, o), and cotton‐top tamarin‐9 (l, p). Each sampling region is indicated by a dashed line. Age in years is listed after each speciman. a.c., anterior commissure; C, caudate nucleus; P, putamen. All sections are 40‐µm thick. Scale = 1 cm.

2.3. Immunohistochemistry

Sections including the NAcc and VP adjacent to Nissl‐stained sections were immunohistochemically processed for SERT using the avidin–biotin–peroxidase method (see Figure 3). Sections were pretreated for antigen retrieval by incubating in 0.05% citraconic acid (pH 7.4) at 85°C–90°C in a water bath for 30 min. After rinsing, endogenous peroxidase was quenched using a solution of 75% methanol, 2.5% hydrogen peroxide (30%), and 22.5% distilled water for 20 min at room temperature. Sections were pre‐blocked in a solution of 0.1 M PBS (pH 7.4), 0.6% Triton X‐100, 4% normal serum, and 5% bovine serum albumin. Sections were then placed in a monoclonal anti‐mouse SERT primary antibody (Millipore MAB5618, Billerica, MA; RRID: AB_2190560) at a dilution of 1:100,000 for 24 h at room temperature, followed by 24 h at 4°C (Raghanti et al. 2008). The molecular weight of the SERT antibody is reported to be 60–70 kDa, and its specificity was characterized previously (Henry et al. 2003; Ramsey and DeFelice 2002; Serafeim et al. 2002). Sections were then incubated in a biotinylated secondary antibody (1:200) in a solution of PBS and 2% normal serum, followed by the avidin–peroxidase complex (PK‐6100; Vector Laboratories, Burlingame, CA). Finally, a 3,3ʹ‐diaminobenzidine‐peroxidase (DAB) substrate with nickel enhancement was used as the chromogen (SK‐4100; Vector Laboratories).

FIGURE 3.

Photomicrographs of SERT‐stained tissue within the NAcc for human‐25 (a), gorilla‐40 (b), pig‐tailed macaque‐15 (c), rhesus macaque‐11 (d), and common marmoset‐11 (e) as well as within the VP for human‐23 (f), chimpanzee‐26 (g), moor macaque‐10 (h), olive baboon‐9 (i), Japanese macaque‐19 (j), and cotton‐top tamarin‐6 (k). Images demonstrate the intensity of the immunostaining. Staining was robust in all species and showed many SERT‐ir varicose axons. Age in years is listed after each speciman. Magnification = 100×. Scale = 20 µm.

2.4. Data Collection and Analysis

Quantitative analyses were performed using computer‐assisted stereology on an Olympus BX‐51 photomicroscope system equipped with a digital camera and StereoInvestigator software version 11 (MBF Bioscience, Williston, VT). Subsampling techniques were performed for each species to optimize appropriate sampling parameters that would yield consistent coefficients of variation less than 0.10 (Slomianka and West 2005).

Three equidistant sections were sampled per individual per brain region. SERT‐ir axon length was measured using the Space Balls probe at 100× (N.A. 1.4) under Koehler illumination using a hemisphere set at 7 µm with a 2% guard zone (Calhoun and Mouton 2000; Kreczmanski et al. 2005; Mouton et al. 2002). Axons were marked where they intersected the outline of the hemisphere, and section thickness was measured at every fifth sampling site, with an average mounted section thickness for immunostained sections being 24.62 ± 9.4 µm in the NAcc and 28.74 ± 10.3 µm in the VP. Sampling grids were set at 300 × 300 µm for human, chimpanzee, bonobo, gorilla, baboon, rhesus macaque, and capuchin, and 225 × 225 µm for pig‐tailed macaque, moor macaque, Japanese macaque, tamarin, marmoset, and owl monkey. The mean number of sampling sites in the NAcc per individual was 80.89 ± 8.9 with a range of 22–95. The mean number of sampling sites in the VP per individual was 78.96 ± 9.8 with a range of 49–90. ALv was calculated as the total fiber length divided by the planimetric measurement of the reference volume sampled, as previously described (Raghanti et al. 2008). Neuron density was used as a denominator to account for the fact that axons are innervating neurons regardless of the widely differing brain sizes seen among species in this study (Hirter et al. 2021).

Neuron density (Nv) was assessed in Nissl‐stained sections using the optical dissector combined with a fractionator sampling scheme. Nv is reported as the sum of neurons counted with the sum of optical dissectors divided by the product of total dissectors sampled and their volume. The NAcc and VP were outlined at 4× magnification, and neurons were manually counted using a 60× objective (N.A. 1.35) with a counting frame of 50 × 50 µm and a height of 7 µm. Neurons were counted when the nucleolus was in focus within the counting frame and displayed the presence of a distinct nucleolus and lightly stained proximal portions of dendritic processes (e.g., Hirter et al. 2021).

Statistical analyses were performed using SPSS (version 13.0), and the level of significance (α) was set at 0.05 for all statistical tests. We first analyzed the effects of sex by using t‐tests on all species in both regions. An independent‐samples Kruskal–Wallis test was run on neuron density (Nv) in each region. An independent‐samples Kruskal–Wallis test was used to evaluate differences in SERT‐ir axon densities normalized by neuron densities (ALv/Nv) among species in both brain regions. We did not find it necessary to perform a regression analysis of axon length density on brain volume, because our previous findings have been wholly consistent with our results from using neuron density as a denominator, even within a given subset of the current samples (Hirter et al. 2021).

3. Results

Table 2 provides summary data for the variables included in our analyses. No sex differences were found in any species except for capuchins. In the VP, males had significantly higher innervation levels than females (t(4) = 3.619, p = 0.011). However, there was a small sample for this species, and the lack of sex differences for the other species suggests that sex is not a confounding factor in these analyses.

TABLE 2.

Summary data including SERT ALv/Nv and Nv in both the NAcc and VP for each species.

Species	N	SERT ALv/Nv	Nv	N	SERT ALv/Nv	Nv
	NAcc	NAcc	NAcc	VP	VP	VP
Marmoset	6	130.52 ± 67.0	141,024 ± 83,363	7	381.6 ± 274	58,456 ± 17,387
Tamarin	4	123.00 ± 32.8	112,258 ± 15,704	6	389.5 ± 90.4	51,589 ± 8982
Owl monkey	4	107.81 ± 54.5	101,140 ± 51,251	4	234.3 ± 136	78,360 ±27,520
Capuchin	6	212.43 ± 136	61,559 ± 44,521	6	318.7 ± 51.7	33,030 ± 9108
Rhesus macaque	6	231.21 ± 354	67,525 ± 22,498	5	380.8 ± 187	29,968 ± 3486
Pigtailed macaque	6	227.15 ± 102	48,222 ± 21,872	6	839.3 ± 520	12,860 ± 3806
Japanese macaque	6	174.64 ± 184	72,330 ± 27,335	6	1055 ± 827	18,129 ± 8888
Moor macaque	6	234.79 ± 60.8	28,479 ± 8121	5	2054± 1287	5091.6 ± 855
Baboon	6	179.60 ± 206	73,094 ± 34,837	6	896.9 ± 690	14,579 ± 7541
Gorilla	4	119.90 ± 56.7	63,428 ± 17,740	0
Chimpanzee	6	372.04 ± 208	37,798 ± 12,530	6	980.4 ± 388	13,133 ± 6622
Bonobo	4	56.03 ± 27.44	100,051 ± 31,880	4	713.7 ± 665	19,541 ± 7938
Human	6	156.89 ± 119	28,110 ± 13,604	6	1134 ± 1067	11,464 ± 11,555

There was a significant difference among species for Nv within the NAcc (H(12) = 41.693, p < 0.001; Figure 4a). Pairwise comparisons with Bonferroni correction showed that humans have significantly lower neuron density than marmosets (p = 0.012) and tamarins (p = 0.019) and that moor macaques have significantly lower neuron density than marmosets (p = 0.006) and tamarins (p = 0.012).

FIGURE 4.

(a) Histograms showing Nv among species within the NAcc. Nv is lower for humans than for marmosets and tamarins, and Nv is lower for moor macaques than for marmosets and tamarins. (b) Box plots showing SERT ALv/Nv among species. SERT ALv/Nv in the NAcc is highly conserved.

No differences among species were detected for ALv/Nv in the NAcc (H(12) = 16.8, p = 0.114; Figure 4b).

In the VP, a species difference was detected for Nv (H(11) = 53.1, p < 0.001; Figure 5a). Pairwise comparisons with Bonferroni adjusted p‐values showed that moor macaques had lower neuron density than marmosets (p = 0.000), owl monkeys (p = 0.001), and tamarins (p = 0.001), and humans had lower neuron density than marmosets (p = 0.008), owl monkeys (p = 0.021), and tamarins (p = 0.029). Chimpanzees had lower neuron density than marmosets (p = 0.036).

FIGURE 5.

(a) Histograms showing Nv among species within the VP. Nv is lower for moor macaques than owl monkeys, tamarins, and marmosets. (b) Box plots showing SERT ALv/Nv among species in the VP. SERT innervation density is higher for moor macaques than capuchins, owl monkeys, and marmosets.

A species difference was detected for ALv/Nv in the VP (H(11) = 30.7, p = 0.001; Figure 5b). Pairwise comparisons with Bonferroni‐adjusted p‐values showed that moor macaques had higher innervation densities than marmosets (p = 0.031), owl monkeys (p = 0.033), and capuchins (p = 0.047).

4. Discussion

The current data show that SERT ALv/Nv is highly conserved in both the primate NAcc and VP, important nodes in the mesolimbic reward pathway. Despite vastly differing brain and body sizes and a variety of social organizations, these two areas have similar innervation densities among 13 primate species. These results differ from our previous work showing species differences in the dorsal striatum of primates (Raghanti et al. 2018). Also, in contrast to the present findings, the mesolimbic dopaminergic system has significant differences in innervation density among primate species (Hirter et al. 2021), notwithstanding the close interaction of the two neurotransmitters (Peters et al. 2021).

4.1. Interactions With Dopamine

At least three classes of serotonin receptors are known to modulate dopamine release within the NAcc of rodents (Alex and Pehek 2007). 5‐HT1B receptors modulate dopamine release at high serotonin levels or with pharmacological manipulation (Hållbus et al. 1997; Yan and Yan 2001). 5‐HT2C receptors tonically inhibit dopamine release in the NAcc (Deurwaerdère et al. 2004; Gobert et al. 2000). 5‐HT3 receptor agonists increase dopamine release (Campbell and McBride 1995; Chen et al. 1991; Z.‐M. Ding et al. 2015), while antagonists show no acute effect but chronically cause decreased dopamine levels (Invernizzi et al. 1995). Dopamine release in the NAcc is modulated indirectly by serotonin receptors in the VTA, another important region of the reward circuit (Navailles et al. 2008; O'Dell and Parsons 2004), and by 5‐HT1A autoreceptors in the raphe nuclei (Andrews et al. 2005).

Serotonin and dopamine interactions are implicated in the abuse of ethanol and cocaine. 5‐HT3 receptors mediate ethanol‐induced dopamine release in the NAcc (Campbell and McBride 1995) and modulate its reinforcing effects (Z.‐M. Ding et al. 2015). 5‐HT1A autoreceptors modulate cocaine‐induced elevation of serotonin and dopamine levels in the NAcc (Andrews et al. 2005). 5‐HT2B receptor inactivation in dopamine neurons lowers ventral striatal dopamine activity and cocaine self‐administration (Doly et al. 2017), and 5‐HT2C receptors modulate cocaine‐induced NAcc dopamine output (Navailles et al. 2008). Humans are uniquely vulnerable to addiction (Calvey 2019; Mavridis 2015), which may be because of our uniquely high levels of dopaminergic innervation within the NAcc and VP (Hirter et al. 2021). Selective 5‐HT2C receptor activation decreases cocaine and methamphetamine intake and drug‐seeking behavior in rhesus macaques (Berro et al. 2017), and serotonin release dampens stimulant effects of amphetamine‐type drugs (Baumann et al. 2011). It could be hypothesized that higher serotonergic levels in the NAcc would lower ethanal‐ or cocaine‐induced dopamine output or self‐administration, in contrast to our conserved levels of serotonin, which may contribute to this vulnerability.

4.2. Ventral Versus Dorsal Striatum

Reward behaviors are supported in different ways by the ventral and dorsal striatum. The NAcc, along with the orbitofrontal cortex (OFC) and VTA, is involved in Pavlovian processes, while the dorsal striatum, along with the ACC, is involved in instrumental behavior (Canese et al. 2011; Marche et al. 2017; Van Den Bos et al. 2014). Serotonin is associated with instrumental behavior, which in mice may be increased with a 5‐HT 2C receptor selective ligand (Bailey et al. 2016). Humans and apes have more serotonin than monkeys in the dorsal and medial caudate nucleus and dorsal putamen (Raghanti et al. 2018). This likely contributes to the high levels of instrumental learning demonstrated in humans and apes (Boesch et al. 2019; van Leeuwen et al. 2024).

The human ventral striatum responds to fairness considered beyond material value (Tabibnia et al. 2008) and shows a pattern of activity consistent with inequality‐averse social preferences (Tricomi et al. 2010). In contrast, the dorsal striatum is associated with retaliation (de Quervain et al. 2004; Krämer et al. 2007; Strobel et al. 2011). Dietary depletion of tryptophan reduces ventral striatal responses to fairness and increases dorsal striatal responses during rejection of unfair offers or costly punishment (Crockett et al. 2013). Humans share negative reactions to inequity with several cooperative species, while it is uncertain if with apes, we share the stronger component of fairness, equalizing outcomes (Bräuer and Hanus 2012; Brosnan and de Waal 2014). Given similar experimental conditions, chimpanzees, like human children and adults, choose a more equitable split of rewards in an ultimatum game when influenced by a partner, compared to more selfish choices when the partner has no recourse (Proctor et al. 2013). Though reports conflict, chimpanzees will sometimes display refusals when they receive a better reward than their partner (Bräuer and Hanus 2012; Brosnan et al. 2010), while capuchins punish a conspecific partner who gains unequal access to resources (Leimgruber et al. 2016). In contrast, marmosets, owl monkeys, and squirrel monkeys do not respond negatively to inequitable outcomes (Freeman et al. 2013). The sense of fairness in humans, along with the ability to constrain retaliation, at least in part, contributes to our sociality. Presumably, other facets of the serotonergic system besides innervation density, such as receptor subtype, contribute to these behavioral differences among species.

4.3. Species Differences

Interestingly, moor macaques have higher serotonergic axon density than some platyrrhine species in the VP due to low neuron density levels. Moor macaques are an isolated species on the island of Sulawesi. The divergence in neuron density within the VP of moor macaques may be explained in part by their isolated geographic location. They inhabit the southwest peninsula and are part of the silenus lineage, thought to be the earliest dispersal of macaques to colonize Asia (Okamoto et al. 2000; Thierry 2007). A socioecological model predicts that species considered to be resident‐nepotistic‐tolerant, including moor macaques, will have high levels of between‐group contest (Sterck et al. 1997). Interestingly, moor macaques have very low intergroup aggression. In a study of 85 encounters between a group and their neighbors, no intergroup interactions with body contact were observed, and females never directed aggression toward other groups (Okamoto and Matsumura 2002). Low neuron density is typically associated with less function (Kverková et al. 2022; Storks et al. 2024). In the VP, neuronal firing in rats correlates with vigor and proximity to a target (Lederman et al. 2021), while serotonin attenuates locomotion and rearing (Graves et al. 2013). Further research is needed to determine what effects low neuron density in the VP may have on motor actions to attain a reward and reactivity to outgroup conspecifics.

Unsurprisingly, humans and chimpanzees have lower neuron densities than some of the small‐bodied species of South America. Even though the human striatum is smaller than expected for body size compared to other primates (Barger et al. 2014; Raghanti et al. 2016; Yin et al. 2009), brain size scales allometrically with body size, basal metabolic rate, and life‐history traits across primates (Leigh 2004; Martin 1981). The decrease in neuron densities in tandem with increasing brain size seen in this study likely reflects allometric scaling patterns that have been reported previously from other brain regions (Hirter et al. 2021; Sherwood et al. 2020).

There are some limitations to this study. As mentioned above (Section 2.1), the moor macaques were transcardially perfused, while the other species were all immersion fixed. We normalized our axon length density measures by neuron counts in part to minimize any differences due to tissue preparation or shrinkage, but we cannot completely rule out a difference due to fixation techniques. Another limitation of the present study is the inability to perfectly age‐match individuals among each species. All individuals are considered adults, yet age ranges have some variation (Table 1). While brain size scales allometrically with body size and life‐history traits as mentioned above (Leigh 2004), and our age ranges roughly follow this trend, we cannot completely rule out an effect of age in our samples. Among our specimens, only two individuals were within the last 25% of adult potential lifespan: one female bonobo and one male gorilla (AnAge: The Animal Ageing and Longevity Database, n.d.). The effects of aging on innervation density are not well understood and cannot be completely ruled out as a contributing factor in the present study.

5. Conclusion

The NAcc and VP, as subcortical structures involved in limbic circuits that support behavior, are important targets of evolutionary study. The mesolimbic reward pathway is crucial for selecting species‐typical behaviors by activating an appetitive state that motivates animals to search for life‐supporting stimuli and avoid harm (Alcaro et al. 2007). Population genomic scans in baboons revealed the dopamine receptor‐mediated pathway as a system of high differentiation between two species that is the likely cause of adaptive behavioral differences (Bergey et al. 2016). Serotonin supports reward‐seeking through activation of the dopaminergic system (Peters et al. 2021).

The current study contributes to the understanding of the reward system by revealing highly conserved levels of serotonergic innervation in the NAcc and VP among humans, apes, cercopithecid monkeys, and platyrrhine monkeys. These levels contrast with elevated dopamine levels for humans and apes. Notably, fitness is enhanced by linking success with higher dopamine levels, including social behaviors (O'Connell and Hofmann 2011), and serotonin, as the major modulator of the mesolimbic dopaminergic system, does not have elevated levels. This unique neurochemical profile may contribute to our vulnerability to addiction.

Author Contributions

H.N.S. and M.A.R. contributed to the study design. H.N.S., D.N.J., and E.L.M. performed research. P.R.H. and C.C.S. contributed new reagents/analytic tools. H.N.S. and M.A.R. analyzed data and wrote the first draft of the paper. D.N.J., C.C.S., and P.R.H. provided comments. All authors read and approved the final manuscript.

Conflicts of Interest

Dr. Patrick Hof and Dr. Mary Ann Raghanti are the Editorial Board members of CNE Journal and the co‐authors of this article. To minimize bias, they were excluded from all editorial decision‐making related to the acceptance of this article for publication. The authors declare no other conflicts of interest.

Peer Review

The peer review history for this article is available at https://publons.com/publon/10.1002/cne.70083

Smith, H. N. , Jones D. N., Munger E. L., Hof P. R., Sherwood C. C., and Raghanti M. A.. 2025. “Distribution of Serotonergic Transporter Innervation in the Nucleus Accumbens and Ventral Pallidum Is Highly Conserved Among Primates.” Journal of Comparative Neurology 533, no. 8: 533, e70083. 10.1002/cne.70083

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.