Role of Spindle Cells in the Social Cognition of Apes and Humans

4.30 Role of Spindle Cells in the Social Cognition of Apes and Humans K K Watson and J M Allman, California Institute of Technology, Pasadena, CA, USA ª 2007 Elsevier Inc. All rights reserved.

479 479 481 482 482 483 484

4.30.1 Morphology 4.30.1.1 Background and History 4.30.1.2 Sterology and Development 4.30.2 Immunohistochemistry and Functional Insights 4.30.2.1 Uncertainty, Dopamine, and Serotonin 4.30.2.2 Social Behavior and Vasopressin 4.30.3 The Social Cognition Hypothesis

Glossary von Economo cells

immunohistochemistry

Nissl stain

Golgi stain

dopamine

serotonin Large bipolar neurons described by the anatomists Constantin von Economo and Georg Koskinas, in 1925. They are restricted to anterior cingulate cortex and frontoinsula cortex, and present in great apes and humans, but not other primates. The use of antibodies to recognize and label specific molecules in a tissue specimen. A classic cell-body stain used for nervous tissue. Cresyl violet, a typical Nissl stain, stains the somas of neurons and the nuclei of glia purple. A metal impregnation technique used to label neurons, glia, or nerve fibers. It is used in some applications to randomly stain the cell bodies and processes of a small percentage of cells in nervous tissue, allowing the dendritic arborization of a few individual neurons to be distinguished from the surrounding cells. A neurotransmitter derived from tyrosine. Dopaminergic cells are located in discrete nuclei in the midbrain, which project heavily to the frontal cortex and the basal ganglia. Dopamine is implicated in motor processes and reward processing, and is also the precursor to epinephrine and norepinephrine.

A molecule derived from tryptophan. Serotonin is found widely throughout the central nervous system, where it acts as a neurotransmitter, and in the periphery. In the gastrointestinal system, serotonin causes the peristaltic action of the smooth muscles.

4.30.1 Morphology 4.30.1.1 Background and History

One way to divide neurons into distinct classes is to base the categories on shape, as is evident in the names of the pyramidal and stellate cells (see The Evolution of Neuron Types and Cortical Histology in Apes and Humans). Similarly, when Nimchinsky et al. (1999) identified a type of cell that was unique to great apes and humans, its identity was based on its distinctive morphology. At the time, they termed this population the spindle cells, but to avoid potential confusion with other uses of this name we now refer to them as von Economo (VE) cells. This name is chosen in honor of the neuroanatomist Constantin von Economo, who, with Georg Koskinas, first described this distinctive class of neurons in 1925 (von Economo and Koskinas, 1925). Upon inspection of his Golgi preparations of human cortex, he noted that these large cells were located in layer 5 and restricted to two regions of the human brain: the anterior cingulate cortex (ACC) and in posterior orbitofrontal cortex adjacent to the insula, a region that he termed frontoinsular cortex (FI, Figure 1). Both of these regions lack a granular layer 4; as in motor cortex, this agranularity may reflect a functional specialization.

480 Role of Spindle Cells in the Social Cognition of Apes and Humans

FI

(a)

ACC

(b) Figure 1 Cytoarchitechtonic divisions of the adult human brain as drawn by von Economo. Lateral view of the brain is shown in (a), with FI indicated in orange. In the medial view (b), ACC is visible and indicated in orange. Drawing modified from von Economo, C. and Koskinas, G. 1925. Die Cytoarchitectonik der Hirnrinde erwachsenen Menschen. Springer, by Atiya Hakeem.

In a cresyl violet-stained sample of human or great ape cortex, these cells may be easily distinguished from the neurons around them due to their symmetric, bipolar soma shape and their large size (Figure 2). Years later, it was discovered that these distinctive-looking cells were only present in the great apes and the humans, which implies that they evolved within the last 15 My (Figure 3). In the case of the VE cells in the FI, they are present only in the great African apes and not the orangutan. This

pushes the likely emergence of VE cells in that region to 9Mya, which in turn predates the rapid expansion of the hominid brain by 3–6My (Kumar and Hedges, 1998; Wood and Collard, 1999). Using the Golgi technique, we found that the apical dendrites of the VE neurons are quantitatively very similar to the apical dendrites of the neighboring pyramidal neurons; however, the branching pattern of the basal dendritic trees is much simpler on VE cells compared to the pyramids (Figure 4) (Watson et al.,

Role of Spindle Cells in the Social Cognition of Apes and Humans 481

2006). The somatic symmetry that is apparent in Nissl stains is retained for some distance along the dendritic tree for VE cells. The Golgi technique also revealed spines on the VE neurons of the frontoinsula, which implies that these large cells are excitatory.

In order to determine the projection targets of neurons, typical tract-tracing methods require invasive applications of the tracing material and the subsequent sacrifice of the animal. This technique would obviously be difficult or unethical in hominoids. However, there are several pieces of evidence to indicate that the VE cells are projection cells. First, they are located in layer V, a classical output later. Second, they are immunohistochemically labeled with SMI-32, a selective antibody for nonphosphorylated neurofilament (NPN) (Nimchinsky et al., 1995). Upon phosphorylation, these molecules are translocated from the soma to the axon, where they increase the diameter of the axon and, consequently, transmission speed. Third, the volume of the VE cell somas is correlated with encephalization, with the largest VE neurons occurring in the largest brains of the primates. This is true also of the Betz cells of the motor cortex, which are known to project long distances (Sherwood et al., 2003). 4.30.1.2 Sterology and Development

VE cells appear in the 35th week after conception in very small numbers. Unlike typical pyramidal cells, the VE cells are still very sparse at birth, suggesting that they differentiate from pyramidal cells or migrate in after birth (Allman et al., 2005). Although we do not at this time know the exact age of proliferation, the population size is very

Figure 2 Nissl-stained brain section from a 7-month-old human infant. The somas of two VE cells are stained purple in the center of the image, with the apical dendrite oriented upwards. Note the elongated shape and presence of a single large basal dendrite.

Primates Strepsirrhini

Haplorrhini

Lemuridae Ring tailed lemur Lesser bamboo lemur

Indridae

Platyrrhini

Tarsii

Tarsiidae

Tamarin

Ph. tarsier

Galagonidae Lesser galago

Titi monkey

Aotinae Owl monkey

Cercopithecinae

Human (10+)

Mandrill

Gorilla (2)

Guinea baboon

Howler monkey

Potto

Atelinae Spider monkey

Chimpanzee (4)

Macaque

Bonobo (2)

Colobinae

Alouattinae

Slender loris

Hominidae

Sooty mangaby

Ioridae Slow loris

Hominoidea

Cercopithecoidea

Callitrichinae

Callicebinae

Verrreaux’s sifaka

Catarrhini

Langur

Pongidae Orangutan (2)

Hhylobatidae White handed gibbon (3)

Cheirogaleidae Gray mouse lemur

Figure 3 Primate phylogenetic tree. Species that have VE neurons in frontoinsula are indicated in orange. The number of brains that have been inspected is indicated. Based on the Semendeferi, Welker, Yakovlev, and Allman Brain Collections; analyzed by Atiya Hakeem, Nicole Teteault and John Allman. Figure by Atiya Hakeem based on Rowe, D. 1996. The Pictorial Guide to the Living Primates. Pogonias Press.

482 Role of Spindle Cells in the Social Cognition of Apes and Humans 4.30.2.1 Uncertainty, Dopamine, and Serotonin

The VE cell regions appear to be strongly activated during periods of high uncertainty. In an fMRI study during which subjects were engaged in a simple gambling task, activation in both FI and ACC got increasingly stronger as the uncertainty in the task increased (Critchley et al., 2001). In a similar vein, both regions were activated during a reversal task, in which a subject attempts to maximize reward during a task that changes contingencies when the subject’s behavior stabilizes (O’Doherty et al., 2003). A series of incorrect answers will prompt the subject to switch strategies, at which point both ACC and FI show increased activity. Recordings from individual dopaminergic neurons in the macaque monkey ventral tegmentum reveals a similar pattern of activation. During trials with high uncertainty of reward, dopamine neurons exhibit a gradual increase in firing rate across the duration of the trial (Fiorillo et al., 2003). Dopaminergic neurons project heavily to frontal cortex and limbic regions, including the VE cell regions. Using D3-specific antibodies on human brain tissue, we found that the VE cells are labeled strongly for this high-affinity dopamine receptor on the somas and apical dendrites (Figure 5). VE cells are labeled in this manner more often than the neighboring pyramidal cells, with 85% of VE cells Figure 4 Neurolucida models of Golgi-stained neurons from the FI of a 23-year-old male human. Note the relative absence of dendritic complexity (i.e., sparse branching) on the VE neuron (left) in comparison to the FI pyramidal neuron (right). Neurolucia tracings by Tiffianie Jones.

near the adult level in the single 4-year-old human specimen that we have analyzed. Humans have far more VE neurons than the great ape species. All species, however, show a consistent asymmetry, with about 30% more VE neurons in the right hemisphere compared to the left.

4.30.2 Immunohistochemistry and Functional Insights What is the function of the VE cells? The fMRI literature reveals that two major types of paradigms activate the VE cell regions: decision-making in the context of high uncertainty and social paradigms. This in turn allows us to make educated guesses about what sort of molecules, particularly surface receptors, might be expressed by the VE cells, which we could then probe using immunohistochemistry on postmortem human specimens.

Figure 5 Dopamine D3 receptor labeling on two VE neurons (right) and a pyramidal neuron (lower left) in human ACC. Brown diaminobenzidine (DAB) deposits indicate the presence of the receptor on the soma and the apical dendrites.

Role of Spindle Cells in the Social Cognition of Apes and Humans 483

being labeled compared to only 50% of the layer V pyramidals. Taken together, this evidence implies that the VE cells take part in a circuit that is involved in processing uncertainty. Dopaminergic input is known to be associated with learning, signaling the extent to which a reward is unexpected. Serotonin has been proposed to mediate the aversive component of learning, signaling the presence or absence of punishment in a manner similar to relationship between dopamine and reward (Daw et al., 2002). Immunohistochemistry reveals that the VE neurons express at least two serotonin receptors, including 5HT-1b and 5HT-2b (Figure 6). 5HT-1b is associated with the inhibition of aggression; applying a 5HT-1b antagonist or knocking out the receptor in mice increases aggressive behavior, while applying a 5HT-1b agonist decreases aggression (Bouwknecht et al., 2001; de Almeida et al., 2001). The 5HT-2b receptor, while heavily prevalent in the human gut, is rare in the human central nervous system (Borman et al., 2002). We found that, in the VE cell regions, the 5HT2b receptor was specific to layer V and present on VE cell somas, pyramidal somas, and short segments of apical trunks in the absence of somatic

(a)

(b)

Figure 6 Immunohistochemical (DAB) labeling of the serotonin 1b (a) and 2b (b) receptors on VE neurons in the ACC. Both samples are Nissl counterstained to reveal unlabeled cells. Scale bar applies to both images.

labeling. In the human gastrointestinal system, this receptor causes peristalsis through the induction of smooth muscle contraction. Interestingly, insular cortex has recently been shown to be involved in interoception, that is, the representation and monitoring of one’s internal states (Craig, 2004). The presence of 5HT-2b on both the VE neurons and in the viscera may indicate a functional connection between the two, perhaps a collateral projection that allows the CNS to rapidly process information relevant to the body. The presence of such a connection would be highly relevant to Damasio’s (1994) theory of somatic states, which hypothesizes that many decisions are made on the basis of signals arising from regulatory processes that occur in the periphery. 4.30.2.2 Social Behavior and Vasopressin

Functional imaging paradigms associated with social behavior also reliably activate both VE cell regions. For example, both ACC and FI are active during the act of lying (telling untruths), and they are both active when a subject receives an unfair offer while playing the ultimatum game (Sanfey et al., 2003; Spence et al., 2004). Studies by Bartels and Zeki (2000, 2004) show both regions are active when subjects view the face of their love partner or child. Singer et al. (2004b) showed that both VE cell regions are active when a person feels empathy for pain, that is, when they know that their loved one, outside of the scanner, is being delivered an electric shock. Interestingly, the extent of activation an individual shows under these conditions is directly correlated to that individuals score on a trait measurement for empathy. Finally, in a separate study, Singer et al. (2004a) demonstrated that left FI is specifically active when subjects view faces of individuals who are reported to behave in a trustworthy fashion. Fortunately, there is an excellent molecular model that allows us to specifically implicate the VE neurons in these various social behaviors. A body of work by Insel and Young indicates that the oxytocin and vasopressin V1a receptors mediate social bonding (Insel et al., 1998; Young et al., 2001; Lim et al., 2004). Insel et al. (1998) also suggest that these molecules may interact with dopamine to impart the rewarding aspects of social bonding. Our immunohistochemical results show that the antibodies specific for the V1a receptor label a subpopulation of VE cells, as well as pyramidal neurons in layers 2/3 and 5 of ACC and FI.

484 Role of Spindle Cells in the Social Cognition of Apes and Humans

4.30.3 The Social Cognition Hypothesis In light of the above evidence, we hypothesize that the recently evolved VE neurons are a functional specialization of a circuit involved in making appropriate responses during quickly changing, ambiguous circumstances (Allman et al., 2005). Links between the VE cells and interoception – including, literally, gut feelings – could provide the basis for their role in fast decision-making in the absence of explicit reasoning. In apes and humans, complex social interactions between conspecifics provide a forum in which this cognitive capacity would prove to be particularly useful (see Human Cognitive Specializations). This is because participants must rapidly synthesize an enormous number of relevant but often ephemeral informational cues in order to act appropriately. We thus propose that VE cells mediate the rapid assessments and behavioral modifications required for the successful navigation of social interactions.

References Allman, J. M., Watson, K. K., Tetreault, N. A., and Hakeem, A. Y. 2005. Intuition and autism: A possible role for Von Economo neurons. Trends Cogn. Sci. 9, 367–373. Bartels, A. and Zeki, S. 2000. The neural basis of romantic love. Neuroreport 11, 3829–3834. Bartels, A. and Zeki, S. 2004. The neural correlates of maternal and romantic love. Neuroimage 21, 1155–1166. Borman, R. A., Tilford, N. S., Harmer, D. W., et al. 2002. 5-HT(2B) receptors play a key role in mediating the excitatory effects of 5HT in human colon in vitro. Br. J. Pharmacol. 135, 1144–1151. Bouwknecht, J. A., Hijzen, T. H., van der Gugten, J., Maes, R. A., Hen, R., and Olivier, B. 2001. Absence of 5-HT(1B) receptors is associated with impaired impulse control in male 5-HT(1B) knockout mice. Biol. Psychiatry 49, 557–568. Craig, A. D. 2004. Human feelings: Why are some more aware than others? Trends Cogn. Sci. 8, 239–241. Critchley, H. D., Mathias, C. J., and Dolan, R. J. 2001. Neural activity in the human brain relating to uncertainty and arousal during anticipation. Neuron 29, 537–545. Damasio, A. R. 1994. Descartes’ Error. G.P. Putnam. Daw, N. D., Kakade, S., and Dayan, P. 2002. Opponent interactions between serotonin and dopamine. Neural Netw. 15, 603–616. de Almeida, R. M., Nikulina, E. M., Faccidomo, S., Fish, E. W., and Miczek, K. A. 2001. Zolmitriptan – a 5-HT1B/D agonist, alcohol, and aggression in mice. Psychopharmacology (Berl.) 157, 131–141. Fiorillo, C. D., Tobler, P. N., and Schultz, W. 2003. Discrete coding of reward probability and uncertainty by dopamine neurons. Science 299, 1898–1902. Insel, T. R., Winslow, J. T., Wang, Z., and Young, L. J. 1998. Oxytocin, vasopressin, and the neuroendocrine basis of pair bond formation. Adv. Exp. Med. Biol. 449, 215–224.

Kumar, S. and Hedges, S. B. 1998. A molecular timescale for vertebrate evolution. Nature 392, 917–920. Lim, M. M., Wang, Z., Olazabal, D. E., Ren, X., Terwilliger, E. F., and Young, L. J. 2004. Enhanced partner preference in a promiscuous species by manipulating the expression of a single gene. Nature 429, 754–757. Nimchinsky, E. A., Vogt, B. A., Morrison, J. H., and Hof, P. R. 1995. Spindle neurons of the human anterior cingulate cortex. J. Comp. Neurol. 355, 27–37. Nimchinsky, E. A., Gilissen, E., Allman, J. M., Perl, D. P., Erwin, J. M., and Hof, P. R. 1999. A neuronal morphologic type unique to humans and great apes. Proc. Natl. Acad. Sci. USA 96, 5268–5273. O’Doherty, J., Critchley, H., Deichmann, R., and Dolan, R. J. 2003. Dissociating valence of outcome from behavioral control in human orbital and ventral prefrontal cortices. J. Neurosci. 23, 7931–7939. Rowe, D. 1996. The Pictorial Guide to the Living Primates. Pogonias Press. Sanfey, A. G., Rilling, J. K., Aronson, J. A., Nystrom, L. E., and Cohen, J. D. 2003. The neural basis of economic decision-making in the ultimatum game. Science 300, 1755–1758. Sherwood, C. C., Lee, P. W., Rivara, C. B., et al. 2003. Evolution of specialized pyramidal neurons in primate visual and motor cortex. Brain Behav. Evol. 61, 28–44. Singer, T., Kiebel, S. J., Winston, J. S., Dolan, R. J., and Frith, C. D. 2004a. Brain responses to the acquired moral status of faces. Neuron 41, 653–662. Singer, T., Seymour, B., O’Doherty, J., Kaube, H., Dolan, R. J., and Frith, C. D. 2004b. Empathy for pain involves the affective but not sensory components of pain. Science 303, 1157–1162. Spence, S. A., Hunter, M. D., Farrow, T. F., et al. 2004. A cognitive neurobiological account of deception: Evidence from functional neuroimaging. Philos. Trans. R. Soc. Lond. B: Biol. Sci. 359, 1755–1762. von Economo, C. and Koskinas, G. 1925. Die Cytoarchitectonik der Hirnrinde des erwachsenen Menschen. Springer. Watson, K. K., Jones, T. K., and Allman, J. M. 2006. Dendritic architecture of the von Economo neurons. Neuroscience (in press). Wood, B. and Collard, M. 1999. The human genus. Science 284, 65–71. Young, L. J., Lim, M. M., Gingrich, B., and Insel, T. R. 2001. Cellular mechanisms of social attachment. Horm. Behav. 40, 133–138.

Further Reading Allman, J. M., Watson, K. K., Tetreault, N. A., and Hakeem, A. Y. 2005. Intuition and autism: Role of Von Economo neurons. Trends Cogn. Sci. 9, 367–373. Craig, A. D. 2004. Human feelings: why are some more aware than others? Trends Cogn. Sci. 8, 239–241. Singer, T. B., Seymour, B., O’Doherty, J., et al. 2004. Empathy for pain involves the affective but not sensory components of pain. Science. 303, 1157–1162.