Mirror neurons in the tree of life: mosaic evolution, plasticity and exaptation of sensorimotor matching responses

Abstract

Considering the properties of mirror neurons (MNs) in terms of development and phylogeny, we offer a novel, unifying, and testable account of their evolution according to the available data and try to unify apparently discordant research, including the plasticity of MNs during development, their adaptive value and their phylogenetic relationships and continuity. We hypothesize that the MN system reflects a set of interrelated traits, each with an independent natural history due to unique selective pressures, and propose that there are at least three evolutionarily significant trends that gave raise to three subtypes: hand visuomotor, mouth visuomotor, and audio-vocal. Specifically, we put forward a mosaic evolution hypothesis, which posits that different types of MNs may have evolved at different rates within and among species. This evolutionary hypothesis represents an alternative to both adaptationist and associative models. Finally, the review offers a strong heuristic potential in predicting the circumstances under which specific variations and properties of MNs are expected. Such predictive value is critical to test new hypotheses about MN activity and its plastic changes, depending on the species, the neuroanatomical substrates, and the ecological niche.

Keywords: exaptation; homology; manual gestures; mirror neurons; mosaic evolution; neonatal imitation; plasticity; sensorimotor learning; tool-use.

Fig. 1

Hand mirror neurons (MNs) in humans, chimpanzees and macaques. Specific anatomical regions of human (A), chimpanzee (B) and macaque (C) brains activated during observation of transitive grasping that overlap with regions activated during the execution of the same grasping action. Premotor, parietal and sensory areas are shown in green, red and yellow, respectively. In humans, the premotor cortex (PM) and inferior frontal gyrus (IFG) approximate Brodman areas BA44 (known as Broca’s area), BA45 and BA6, while the rostral part of the inferior parietal lobule (IPL) is thought to correspond to cytoarchitectonic areas PF and PFG. In the parietal area, mirror activity has also been found in the anterior sector of the intraparietal sulcus (AIP). In chimpanzees, the IFG occupies the anatomical region FCBm, following Bailey et al. (1950); while in the IPL, mirror responses occupy areas PF and PFG. In macaques, the PM contains area F5, where mostly hand MNs are located, and in the IPL MNs have been found in areas PF, PFG and in the AIP region. In humans and macaques, hand mirror responses have also been found in the motor cortex (M1) and supplementary motor area (SMA). The superior temporal sulcus (STS) is a sensorial region which lacks motor properties but is considered part of the hand mirror neuron system, because it is thought to be the main source of visual information.

Fig. 2

Actual and potential vocal learners and audio–vocal mirror neurons (MNs). Songbirds and humans, both vocal learners, have a similar pattern in the way premotor/motor areas are connected to the brainstem in the establishment of learned vocalizations. In songbirds (A) the robust nucleus of the arcopallium (RA) connects with motoneurons of the syrinx (XII) for the production of song, while a preparatory and modulatory role is played by the premotor nucleus high vocal centre (HVC), which is also implicated in the circuit underlying learning vocalizations together with the striatal nucleus (Area X) and the dorsolateral nucleus of the medial thalamus (DLM). Similarly, in humans (B), the face area of the primary motor cortex (the region involved in laryngeal motor control; LMC) connects to the nucleus ambiguous (Am), while Broca’s area, a premotor region, is thought to play a more anticipatory and planning role, and is though to be implicated in vocal learning together with the anterior striatum (Ast) and the anterior thalamus (At) (considered homologous to Area X and the DLM, respectively). Audio-vocal MNs have been found in the HVC and Broca’s area, respectively, in songbirds and humans. In non-singing birds or female songbirds (C), which lack vocal learning, the circuit controlling innate calls involves connections between the periacqueductal grey area (PAG) [the avian homolog to the dorsal medial nucleus of the midbrain (DM)] and motoneurons of the syrinx and larynx. A similar circuit has been described in non-human primates, such as the macaque (D), where connections between the PAG and the motoneurons of the Am are necessary, together with the anterior cingulate cortex (ACC), to produce emotionally spontaneous vocalizations. By contrast, the lateral part of the cerebral cortex of non-human primates, the premotor cortex, has a minor role in voluntary control, despite its most lateral part having motor representations of the mouth and larynx. Nevertheless, both female songbirds and macaques can be trained to vocalize voluntarily and consequently to establish connections between the premotor cortex and motoneurons, and therefore have the capacity to develop audio–vocal MNs in the premotor regions. The red arrow shows the direct motor projection from vocal motor regions (RA or LMC) to the brainstem (XII or Am). The white arrow represents the cortico-striatal loop for learning vocalizations, while the black arrow shows the conserved pathway for innate calls. The dashed black arrow shows possible connection between the cortico-striatal loop for learning and the direct motor projection involved in the production of vocalizations. LMAN, lateral magnocellular nucleus of the anterior nidopallium; VL, ventrolateral thalamus [considered homologous to the AT]; PM, premotor cortex. Modified from Pfenning et al. (2014).

Fig. 3

Mirror neurons (MNs) in the tree of life. MNs can be categorized in relation to the effector with which executed actions are performed (hand, mouth or vocal tract) and the modality of the sensorial input in which the same actions are perceived (vision or hearing). Hand MNs (blue lines) are observed in humans, apes, macaques and marmosets. The violet line indicates the later exaptation of circuits devoted to hand MNs for tool use and the existence (after a sensorimotor training with tools; in the laboratory; Ferrari et al., 2005) of tool-responding MNs in macaques, humans and likely in apes. Mouth MNs (red lines) have been reported in humans and macaques, and are likely present also in chimpanzees and marmosets. We further propose that mouth and hand MNs may be present in prosimians (red or blue lines with question marks). Audio–vocal MNs (yellow line) have been found in humans and songbirds.