Functions of the anterior insula in taste, autonomic, and related functions
Introduction
The anterior insular cortex has been described as a region that represents interoception and is associated with all subjective feelings, and also with attention, cognitive choices and intentions, music, time perception and, unmistakably, awareness of sensations and movements, of visual and auditory percepts, of the visual image of the self, of the reliability of sensory images and subjective expectations, and of the trustworthiness of other individuals (Craig, 2002, Craig, 2009, Craig, 2011). It has also been suggested that the insula plays a key role in saliency, switching, attention, and control (Menon & Uddin, 2010). It has also been suggested that the mid-insula is the location of the human taste cortex (Craig, 2011, Small, 2010).
In this paper, evidence on some of the functions of the anterior insular cortex is considered. Part of the focus is on the taste and related viscero-autonomic functions of the anterior insula, for there is considerable evidence on this at the neuronal level in primates, and at the fMRI level in humans. Indeed, the approach taken here is to consider together, side-by-side, the primate neuronal recording and the human functional magnetic resonance imaging (fMRI) evidence, to build a clear foundation for understanding the processing in this region and the underlying principles in primates including humans. There are a number of differences between rodents and primates including humans in taste and related processing in the insula as follows, which lead to a focus in this paper on processing in primates including humans, with the aim of understanding the principles of operation of the insula in primates including humans.
First, there are major anatomical differences in the neural processing of taste in rodents and primates (Rolls, 2014, Rolls, 2015c, Rolls and Scott, 2003, Scott and Small, 2009, Small and Scott, 2009). In primates the rostral part of the nucleus of the solitary tract (NTS, the first central taste relay) projects to the taste thalamus and thus to the cortex (Fig. 1, Fig. 2); whereas in rodents the majority of NTS taste neurons project to the pontine parabrachial nucleus (PbN), referred to as the rodent ‘pontine taste area’ (Cho et al., 2002, Small and Scott, 2009) (Fig. 2). From the PbN the rodent gustatory pathway bifurcates into two pathways: (1) a ventral ‘affective’ projection to the hypothalamus, central gray, ventral striatum, bed nucleus of the stria terminalis and amygdala; and (2) a dorsal ‘sensory’ pathway, which first synapses in the thalamus and then the agranular and dysgranular insular gustatory cortex (Norgren, 1974, Norgren, 1976, Norgren, 1990, Norgren and Leonard, 1971) (Fig. 2). In primates (including humans) there is strong evidence to indicate that the PbN gustatory relay is absent (Small & Scott, 2009).
Second, a functional difference of rodent taste processing from that of primates is that physical and chemical signals of satiety have been shown to reduce the taste responsiveness of neurons in the nucleus of the solitary tract, and the pontine taste area, of the rat, with decreases in the order of 30% (Giza et al., 1993, Giza and Scott, 1983, Giza and Scott, 1987, Glenn and Erickson, 1976, Hajnal et al., 1999, Rolls and Scott, 2003, Scott and Small, 2009) (Fig. 2). (Given this evidence, as expected, neuronal responses in many areas of the rat brain including the insula and amygdala are decreased by satiety (de Araujo et al., 2006).) The implication of this whole body of evidence is that in rodents, sensory (perceptual) and reward (hedonic) processing are not independent. In contrast, in primates, the reward value of tastants is represented in the orbitofrontal cortex in that the responses of orbitofrontal cortex taste neurons are modulated by hunger in just the same way as is the reward value or palatability of a taste, and this is not found in the taste insula (Rolls, 2015c). In particular, it has been shown that orbitofrontal cortex taste neurons stop responding to the taste of a food with which a monkey is fed to satiety, and that this parallels the decline in the acceptability of the food (Critchley and Rolls, 1996a, Rolls et al., 1989). In contrast, the representation of taste in the primary taste cortex of non-human primates (Scott et al., 1986a, Yaxley et al., 1990) is not modulated by hunger (Rolls et al., 1988, Yaxley et al., 1988). Thus in the primary taste cortex of non-human primates (and at earlier stages of taste processing including the nucleus of the solitary tract (Yaxley, Rolls, Sienkiewicz, & Scott, 1985)), the reward value of taste is not represented, and instead the identity and intensity of the taste are represented (Rolls, 2014). A perceptual correlate of this is that when humans feed to satiety, the intensity of the flavor changes very little, whereas the pleasantness of the flavor decreases to zero (Rolls, Rolls, & Rowe, 1983), showing that in humans perceptual representations of taste and olfaction are kept separate from hedonic representations. This is adaptive, in that we do not go blind to the sight, taste, and smell of food after eating it to satiety, and can therefore still learn about where food is located in the environment even when we are not hungry (Rolls, 2014). Moreover, and consistently, activations in the human insular primary taste cortex are related to the intensity and not to the pleasantness of taste (Grabenhorst and Rolls, 2008, Grabenhorst et al., 2008) (see Fig. 10, Fig. 9).
The importance of cortical processing of taste in primates, first for identity and intensity in the primary taste cortex, and then for reward value in the orbitofrontal cortex, is that both types of representation need to be interfaced to visual and other processing that requires cortical computation. For example, it may have adaptive value to be able to represent exactly what taste is present, and to link it by learning to the sight and location of the source of the taste, even when hunger is not present and reward is not being produced, so that the source of that taste can be found in future, when it may have reward value. In line with cortical processing to dominate the processing of taste in primates, there is no modulation of taste responsiveness at or before the primary taste cortex, and the pathways for taste are directly from the nucleus of the solitary tract in the brainstem to the taste thalamus and then to the taste cortex (Fig. 1, Fig. 2) (Rolls, 2014).
Third, the architectonic division of the rat insula that is involved in taste processing appears to be different from that in primates (Evrard, Logothetis, & Craig, 2014).
Fourth, the prefrontal cortex (and for that matter the temporal lobe visual cortical areas) have also undergone great development in primates, and one part of the prefrontal cortex, the orbitofrontal cortex, is very little developed in rodents, yet is one of the major brain areas involved in taste and olfactory processing, and emotion and motivation, in primates including humans. With this great development of the orbitofrontal cortex in primates, there may be division of functionality, with the primate taste insula not performing taste-related hedonic functions (Rolls, 2015c). Indeed, it has been argued (on the basis of cytoarchitecture, connections, and functions) that the granular prefrontal cortex is a primate innovation, and the implication of the argument is that any areas that might be termed orbitofrontal cortex in rats (Schoenbaum, Roesch, Stalnaker, & Takahashi, 2009) are homologous only to the agranular parts of the primate orbitofrontal cortex, that is to areas 13a, 14c, and the agranular insular areas Ia (Passingham & Wise, 2012). It follows from that argument that for most areas of the orbitofrontal and medial prefrontal cortex in humans and macaques, special consideration must be given to research in macaques and humans. Indeed, there may be no cortical area in rodents that is homologous to most of the primate including human orbitofrontal cortex (Passingham and Wise, 2012, Preuss, 1995, Rolls, 2014, Rolls, 2015c, Wise, 2008).
Section snippets
Pathways
A diagram of the taste and related olfactory, somatosensory, and visual pathways in primates is shown in Fig. 1. The multimodal convergence that enables single neurons to respond to different combinations of taste, olfactory, texture, temperature, and visual inputs to represent different flavors produced often by new combinations of sensory input is a theme that has been addressed recently, as have the functions of different parts of this system in reward value processing, modulatory effects
Neuronal responses to taste
Rolls, Scott, and colleagues have shown that the primary taste cortex in the primate anterior insula and adjoining frontal operculum contains not only taste neurons tuned to sweet, salt, bitter, sour (Plata-Salaman et al., 1992, Plata-Salaman et al., 1993, Plata-Salaman et al., 1995, Plata-Salaman et al., 1996, Rolls and Scott, 2003, Scott et al., 1998, Scott et al., 1999, Scott and Plata-Salaman, 1999, Scott et al., 1991, Scott et al., 1994, Scott et al., 1986a, Smith-Swintosky et al., 1991,
An autonomic/visceral representation in the anterior/ventral insula
Parts of the insula receive visceral afferent inputs, and in turn are connected to efferent systems involved in the elicitation of autonomic responses (Al Omran and Aziz, 2014, Allen et al., 1991, Craig, 2002, Craig, 2009). In humans, a part of the anterior insula has activations related to visceral signals (Al Omran and Aziz, 2014, Critchley, 2005, Critchley and Harrison, 2013, Critchley et al., 2004). For example, Critchley et al. (2004) observed enhanced activation in the anterior insular
Saliency
It has been suggested that the anterior insula is a key part of a “saliency network” implemented by the insula and the anterior cingulate cortex (Menon & Uddin, 2010). The disparate functions ascribed to the insula are conceptualized by a few basic mechanisms: (1) bottom-up detection of salient events, (2) switching between other large-scale networks to facilitate access to attention and working memory resources when a salient event is detected, (3) interaction of the anterior and posterior
Somatosensory representations in the mid/posterior insular cortex
The mid and posterior insular cortex contain somatosensory areas (Kaas, 2012, Mufson and Mesulam, 1984). Effects found in these areas are briefly described for comparison with the effects found in the anterior insula.
In a mid/posterior insular area at Y = −2 to Y = −22 we found activations to touch on the forearm or hand (McCabe, Rolls, Bilderbeck, & McGlone, 2008). Interestingly, similar activations were not produced in these areas to the sight of touch, though they were in many other
Acknowledgments
This research was supported by the Medical Research Council. The participation of many colleagues in the studies cited is sincerely acknowledged. They include Ivan de Araujo, Gordon Baylis, Leslie Baylis, Wei Cheng, Hugo Critchley, Wanlu Deng, Paul Gabbott, Jianfeng Feng, Tian Ge, Fabian Grabenhorst, Mikiko Kadohisa, Morten Kringelbach, Christian Margot, Ciara McCabe, Francis McGlone, Tom Nichols, John O’Doherty, Barbara Rolls, Juliet Rolls, Thomas Scott, Zenon Sienkiewicz, Simon Thorpe, Maria
References (139)
- et al.
Responses of neurons in the primate taste cortex to glutamate
Physiology and Behavior
(1991) - et al.
Afferent connections of the orbitofrontal cortex taste area of the primate
Neuroscience
(1995) - et al.
Visceral influences on brain and behavior
Neuron
(2013) - et al.
Neural ensemble coding of satiety states
Neuron
(2006) - et al.
Componential Granger causality, and its application to identifying the source and mechanisms of the top-down biased activation that controls attention to affective vs sensory processing
Neuroimage
(2012) - et al.
Gastric modulation of gustatory afferent activity
Physiology and Behavior
(1976) - et al.
Value, pleasure, and choice in the ventral prefrontal cortex
Trends in Cognitive Sciences
(2011) - et al.
Human cortical representation of oral temperature
Physiology and Behavior
(2007) - et al.
Social and motivational functioning is not critically dependent on feedback of autonomic responses: Neuropsychological evidence from patients with pure autonomic failure
Neuropsychologia
(2004) Somatosensory system