C H A P T E R

29
The Emotional Brain
L. Pessoa
University of Maryland, College Park, MD, United States

INTRODUCTION REGIONS OF THE EMOTIONAL

BRAIN
In his seminal contribution, in 1878, Broca
described the great limbic lobe comprising In this section, several important regions of
the interior surface of the cortex and a set of the emotional brain are discussed. Although
related structures. These structures were later the exposition is centered on particular regions,
part of the first major circuit theory of emo- they do not work in isolation. Instead, they par-
tion, namely, the circuit of Papez. By the mid- ticipate in large, distributed neural networks
dle of the 20th century, ideas by Papez and whose constituent regions, collectively, carry
others were further refined and extended by out functions and behaviors of interest. There-
MacLean into the concept of the “limbic sys- fore, although the exposition emphasizes a
tem,” according to which emotion depends region’s participation in emotional functions,
on a well-defined and delimited set of brain to understand the emotional brain it is criti-
regions. The “emotional brain” was suggested cal to characterize large-scale circuits within
to be responsible for both response systems which regions are embedded (further discussed
and feeling states/processes, given that emo- in sections From Regions to Circuits and How
tion was characterized by a combination of Are Emotions Organized in the Brain?). In this
changes in physiological arousal, behav- sense, there is no one, separately identifiable
ioral expression, and, importantly, subjective “emotional brain,” but instead regions that play
experience. important roles in emotion (Fig. 29.1).
This chapter describes several components
of the emotional brain. As outlined in the expo-
sition in the following sections, the concept of Amygdala
a “limbic system” has given way to a modern Anatomically, the amygdala is a complex
understanding of emotion as highly integrated structure containing more than a dozen richly
with other brain systems. It is thus preferable interconnected nuclei. However, based on sev-
to avoid the term “limbic system” altogether, as eral types of information, including connectiv-
advocated for some time by many leading emo- ity and distribution of neurotransmitters, some
tion researchers. have questioned whether this collection of nuclei

Conn’s Translational Neuroscience

http://dx.doi.org/10.1016/B978-0-12-802381-5.00046-4 635 Copyright © 2017 Elsevier Inc. All rights reserved.
636 29.  THE EMOTIONAL BRAIN

FIGURE 29.1  Emotional brain. Several brain structures participate in emotional mechanisms in important ways. Some of
the regions discussed in the chapter are shown schematically here. PAG, periaqueductal gray.

should be considered a functional–anatomical medium spiny neurons. Furthermore, In terms

unit. Indeed, the title of a provocative paper of embryological origin, the amygdala consists
by leading researchers was “What is the amyg- of distinct groups of nuclei with distinct precur-
dala?” Nevertheless, it is useful to consider at sors linked to the lateral and central amygdala
least two subunits of this structure: the “lateral (among others).
amygdala” (at times called “basolateral amyg- The two amygdala subdivisions display con-
dala”) involving the lateral, basal, and accessory nectivity patterns that are quite distinct from
basal nuclei, and the “central amygdala” involv- each other (Fig. 29.3). The lateral amygdala
ing the central nucleus (Fig. 29.2). These nuclei receives substantial sensory information from
are typically further subdivided into subnuclei, cortex and is richly connected with temporal,
such as the medial and lateral divisions of the cingulate, insular, and prefrontal cortex; connec-
central amygdala. tions with parietal cortex also exist but are more
Structurally, the lateral amygdala exhibits a modest. In contrast, the central amygdala has
cytoarchitecture that is closer to that seen in cor- extensive interconnections with the hypothala-
tex, which is organized in a layered manner. The mus and brain stem nuclei (including sites in the
majority of lateral amygdala neurons are gluta- midbrain, pons, and medulla) that are involved
matergic (excitatory) spiny projection neurons in behavioral, autonomic, and neuroendocrine
and about 20% of its neurons are GABAergic responses; cortically, the central amygdala is
(inhibitory) interneurons. The central amyg- interconnected with the medial prefrontal cor-
dala exhibits a more simplified cytoarchitecture tex, orbitofrontal cortex, and insula.
with incipient lamination, and is a striatum- The amygdala is a key brain structure for
like structure that is composed of GABAergic aversive memory acquisition, a notion that is
FIGURE 29.2  Human amygdala. (A) Lateral view of an MRI scan with the approximate location of the amygdala out-
lined on the lateral surface; line represents location of coronal section in panel B. (B) Coronal section showing position of left
and right amygdala. (C) Coronal section with box around left amygdala. (D) Enlarged, Nissl-stained section of left amyg-
dala nuclei from panel C. Acc, accessory; PAC, periamygdaloid cortex; PC, posterior cortical. Reproduced with permission from
Schumann CM, Bauman MD, Amaral DG. Abnormal structure or function of the amygdala is a common component of neurodevelop-
mental disorders. Neuropsychologia 2011;49:745–759.

FIGURE 29.3  Amygdala connectivity. (A) Output connections from the lateral amygdala. Connections to parts of the
parietal cortex, including the intraparietal sulcus, are not shown. Numbers refer to approximate locations of Brodmann areas.
(B) Output connections of the central amygdala. Connections to cortex are not shown. (A) Reproduced with permission from
Nieuwenhuys R, Voogd J, van Huijzen C. The human central nervous system. 4th ed. New York: Springer-Verlag; 2008. (B) Adapted with
permission from Nieuwenhuys R, Voogd J, van Huijzen C. The human central nervous system. 4th ed. New York: Springer-Verlag; 2008.
638 29.  THE EMOTIONAL BRAIN

supported by studies using multiple experi- rodents, responses include freezing behavior
mental paradigms and measures of conditioned (technically, a nonresponse), alterations in auto-
responses (Box 29.1). Note that the term “aver- nomic nervous system activity, release of stress
sive conditioning” is used here instead of “fear hormones, analgesia, and facilitation of reflexes.
conditioning” because fear is a subjective men- Subsequently, conditioned responses can be sup-
tal state not accessible in nonhuman animals pressed if the CS is repeatedly presented alone,
(unfortunately, many researchers continue to a phenomenon called “extinction” (see section:
use the term “fear” when studying rodents or Interactions Between the Amygdala and Medial
other animals). Prefrontal Cortex).
Aversive conditioning employs techniques The lateral amygdala receives multiple sen-
of classical conditioning (also called Pavlov- sory signals, including auditory, visual, and
ian conditioning). In aversive conditioning, somatosensory stimuli that convey informa-
the subject is exposed to a neutral conditioned tion about both conditioned and unconditioned
stimulus (CS), such as a tone or a light, which stimuli. Such convergence positions the lat-
typically co-terminates with an aversive uncon- eral amygdala to be able to form associations
ditioned stimulus (US), such as a foot-shock. between the two. Evidence supporting the idea
With training, the CS (tone/light) acquires aver- that the lateral amygdala is critical for aversive
sive properties and, when subsequently pre- learning comes from multiple sources. First,
sented alone, elicits a conditioned response. In lesion or inactivation of the amygdala during

BOX 29.1

STUDYING “FEAR” AND “ANXIETY”

It is productive to operationalize “fear” as suggesting that, if these terms are used at all,
acute threat and “anxiety” as potential threat, clear operationalizations are needed. In nonhu-
such as in the Research Domain Criteria (abbre- man animals, terms such as “fear” and “anxiety”
viated RDoC) initiative of the National Insti- are preferably avoided given that they reflect
tute of Mental Health. Acute threat is linked to subjective mental states.
imminent and discrete threat, such as when a Aversive conditioning constitutes the central
CS signals the possibility of a shock. Potential experimental paradigm to study acute threat. In
threat is linked to contexts in which harm may this paradigm (see main text), an initially neutral
be experienced, such as when entering a room in stimulus evokes a negative state through asso-
which a shock may be delivered. More generally, ciation with an aversive event. Potential threat
acute and potential threat can be distinguished is studied, among others, via exposure to threat-
according to the degree of certainty regarding ening contexts (such as a brightly lit space for
the timing and form of the threat. Behaviorally, rodents, or a dark room for humans) and via con-
cues that indicate an immediate threat give rise text conditioning (where unpredictable aversive
to “fearful” defensive behaviors, whereas more events are paired with a certain context or place).
diffuse, distal, or unpredictable threat cues pro- Past research tended to emphasize the role of
duce “anxious” risk assessment behavior. In the particular brain areas in generating states linked
literature, there is widespread inconsistency in to acute or potential threat. However, mounting
how terms such as “fear,” “anxiety,” “stress,” evidence indicates that these states rely on neu-
“distress,” “worry,” and so on are employed, ronal circuits that involve multiple brain regions.
 Regions of the Emotional Brain 639
learning prevents associations from taking place particular US. This information can then be used
(but typically these lesions, though they might by the central amygdala and its hypothalamic,
include the lateral amygdala, are not limited to midbrain, and other brain stem targets, thus giv-
this sector). Second, responses elicited by the CS ing rise to “emotional” responses such as freez-
in the lateral amygdala are modified after pair- ing, fear-potentiated startle, and modulation of
ing with the US. Specifically, during aversive arousal and attention.
conditioning, cells in the lateral amygdala of rats This serial scheme of amygdala function in
show plasticity during exposure to a CS. Cells which the lateral amygdala passes on informa-
that are responsive to the CS prior to condition- tion to the central amygdala is a valuable heuris-
ing increase their firing after it. Third, recent tic and research tool. However, it captures only
studies employing modern techniques based on a specific mode of amygdala functioning. This
optogenetics indicate that activity-dependent is because the central amygdala receives direct
plasticity in the lateral amygdala is necessary for inputs from cortical and subcortical structures
aversive learning. For example, aversive condi- too (cortical input includes the medial prefrontal
tioning can be inactivated through optogenetic cortex and orbitofrontal cortex), which contrib-
depression of sensory inputs to the lateral amyg- ute to the acquisition of Pavlovian associations.
dala, and later reactivated by optogenetic activa-
tion of these inputs. The Bed Nucleus of the Stria Terminalis
The central amygdala is a major participant
and the Extended Amygdala
in the generation of conditioned responses—
at times referred to as “emotional responses.” The strial terminalis is a major fiber pathway
This role is supported by projections to multiple of the amygdala that courses upward and loops
autonomic and skeletomotor regions involved around medially in the direction of the hypo-
in aversive conditioned responses, including thalamus. Among others, it carries fibers from
the following: the periaqueductal gray (PAG), the amygdala to the thalamus, hypothalamus,
which participates in freezing behavior that is and septum; it also carries fibers projecting from
characterized by the absence of overt behavior; these areas back to the amygdala. Whereas the
the lateral hypothalamus, which participates in stria terminalis fiber bundle is relatively easy to
sympathetic activation; and caudal parts of the visualize in three dimensions, it is much harder
reticular formation in the pons, which partici- when viewed via standard cross-sectional ana-
pate in the potentiation of the startle reflex. The tomical planes. Along the stria terminalis, it is
central amygdala also projects to multiple brain possible to identify several small masses of cells,
stem nuclei that originate neurotransmitter sys- which together comprise the bed nucleus of the
tems with widespread outputs to the forebrain, stria terminalis (BNST; Fig. 29.4).
including the following: the dopaminergic Broadly speaking, the BNST can be subdi-
ventral tegmental area and adjacent substantia vided into lateral and medial parts. The lateral
nigra pars compacta; the noradrenergic locus BNST (BNSTlat) is closely associated with the
coeruleus; the serotonergic raphe nuclei; and the central amygdala, and together they have been
cholinergic basal forebrain nuclei. In all, several suggested to be part of a central “extended
conditioned responses depend on the central amygdala.” The BNSTlat projects to the hypo-
amygdala and its projections to these regions. thalamus and brain stem, as does the central
One way to summarize some of the func- nucleus of the amygdala (as reviewed previ-
tions of the lateral and central amygdala during ously). Indeed, the central amygdala and the
aversive conditioning is as follows. The lateral BNSTlat have similar anatomical connectiv-
amygdala establishes associations, thus allow- ity patterns. Furthermore, neurons in the cen-
ing the CS to access the motivational value of its tral amygdala and the BNSTlat share similar
640 29.  THE EMOTIONAL BRAIN

FIGURE 29.4  Bed nucleus of the stria terminalis (BNST). (A) Approximate location of the BNST in the human brain. (B)
Coronal myelin sections showing cell masses of the BNST. Abbreviations for the structures of main interest: BSTLD, laterodor-
sal part of the BNST; BSTLDcn, central subdivision of the BSTLD; BSTM, medial part of the BNST; CdL, lateral aspect of the
caudate; CdM, medial aspect of the caudate; CdV, ventral aspect of the caudate; ic, internal capsule; LH, lateral hypothalamus;
Pu, Putamen; st, stria terminalis. (B) Reproduced with permission from Yilmazer-Hanke DM. Amygdala. In: Mai JK, Paxinos G,
editors. The human nervous system. 3rd ed. Amsterdam: Elsevier; 2012. p. 759–834.

morphological and neurochemical character- threat (“fear”) and the BNST is more closely asso-
istics (they are mostly GABAergic as found in ciated with potential threat (“anxiety”). Recent
striatal brain regions). research, including with novel techniques such
Functionally, the BNSTlat is important for as optogenetics, is consistent with this general
maintaining prolonged threat preparedness that idea but also reveals important ways in which
is associated with the processing of potential amygdala and BNST circuits participate in mul-
threat (see Box 29.1). Consider the paradigm of tiple aspects of threat processing. For example,
light-enhanced startle. Rats produce a startle activation/deactivation of projection neurons
reflex when presented with brief noise bursts. from the lateral to the central amygdala can be
The amplitude of this acoustic startle can be mea- used to influence the time mice will spend on
sured during different contexts. In one study, open fields (a behavior that is typically actively
during the first phase, rats were tested in the avoided by mice), a form of potential threat.
dark. During the second phase, rats were tested Thus, the lateral and central amygdala are not
again in the dark or, alternatively, in the pres- only involved in the processing of acute threat
ence of bright light (which is highly aversive to (the latter is revealed by standard conditioning
rats). Exposure to bright light for 5–20 min led to paradigms).
a significant increase in acoustic startle amplitude
going from the dark to the light (light-enhanced
startle) compared with startle amplitude when Hypothalamus
the rats remained in the dark. Critically, inactiva- The hypothalamus contains a number of
tion of the BNST (but not the central amygdala) nuclei and, as its name implies, is located just
eliminates the light-enhanced startle. below the thalamus (and thus just above the
An influential framework suggests that the brain stem) (Fig. 29.1; see also Fig. 29.4). The
amygdala is more closely associated with acute importance of the hypothalamus in emotion
 Regions of the Emotional Brain 641
was established in the first decades of the 20th a view that is conveyed by its designation as
century, as highlighted by the work of Bard and the “head ganglion” of the autonomic nervous
Cannon, who showed that coordinated emo- system. Indeed, the hypothalamus has robust
tional expressions were compromised when the connections that synapse at multiple brain stem
hypothalamus was excised. The involvement of sites, including the PAG, midbrain reticular
the hypothalamus in emotion and other basic formation, dorsal raphe, parabrachial nucleus,
functions was fittingly summarized in 1929 by locus coeruleus, nucleus of the solitary tract,
Cushing (considered by many to be the father as well as sites in the medulla and spinal cord.
of modern neurosurgery) in the following way: Notably, several of these sites have “ascending”
“Here in this well-concealed spot, almost to be projections that reach the hypothalamus too.
covered with a thumb nail, lies the very main- The stress response is among the many func-
spring of primitive existence–vegetative, emo- tions in which the hypothalamus participates; in
tional, reproductive.” this case as part of the hypothalamic–pituitary–
Since the 1920s and 1930s, our knowledge adrenal (often abbreviated HPA) axis. Neurons
of hypothalamic function has been greatly in the paraventricular nucleus of the hypothala-
extended and refined, and current understand- mus, which synthesize corticotropin-releasing
ing concurs with the earlier notion that the hypo- hormone (CRH), integrate stress-relevant sig-
thalamus is involved in several important basic nals from multiple brain regions and launch the
operations. For example, it participates in many neuroendocrine response to stress. CRH stimu-
complex homeostatic mechanisms and contrib- lates the release of the adrenocorticotropic hor-
utes to neuroendocrine outputs. Indeed, hypo- mone from the pituitary, which in turn drives
thalamic mechanisms contribute to a wealth of the release of glucocorticoids from the adrenal
processes: circadian rhythms, wakefulness and glands (sitting atop the kidneys). These “stress
sleep, stress responses, temperature regulation, hormones,” such as corticosterone and cortisol,
food intake, thirst, sexual behaviors, and defen- have a wide range of effects, including increas-
sive behaviors. In all these cases, the hypothala- ing blood pressure and glucose levels, and sup-
mus works in concert with a multitude of other pressing inflammatory and immune responses.
sites, several of which are located in the brain The hypothalamus is also implicated in
stem and spinal cord. defensive behavior. In the cat, for example, the
In broad terms, the hypothalamus has been behavioral components of this reaction include
subdivided into several parts: rostrocaudally ear retraction, hissing, growling, and strik-
into the preoptic, anterior, medial (also called ing with the forepaw; autonomic responses
tuberal), and posterior regions, and mediolat- include pupil dilation, piloerection, accelerated
erally into periventricular, medial, and lateral heart rate, and elevated blood pressure. Classic
zones. Within those sectors, multiple “nuclei” work by Hess and many others in the 1920s–50s
or “areas” have been described. However, the showed that a defense response can be elicited
hypothalamus poses considerable challenges by electrical stimulation of brain areas in and
for systematic study because compact group- around the hypothalamus. In the literature,
ings of neuronal elements are not always seen specific defensive behaviors are at times linked
in some “areas.” In fact, even the overall borders to specific “centers.” For example, in the rat, a
of the hypothalamus are ill defined (eg, caudally hypothalamic “attack area” has been described,
the hypothalamus transitions gradually into the and includes parts of the lateral hypotha-
midbrain tegmentum). lamic area and the ventromedial hypothalamic
Historically, the hypothalamus has been con- nucleus. Whereas characterizations such as this
ceptualized in terms of “descending” systems, have conferred a certain degree of specificity to
642 29.  THE EMOTIONAL BRAIN

FIGURE 29.5  Hypothalamus and structure–function relationship. Hypothalamic regions (R1, …, R6) are involved in
many functions (or behaviors), and functions rely on many hypothalamic regions. The three-dimensional landscape repre-
sents levels of activity across space. The patterns of activity indicate that multiple regions of the hypothalamus are engaged
during a certain function/behavior. In this manner, the function is associated with a pattern of activity across the hypothala-
mus (and elsewhere in the brain), and not activation of a specific hypothalamic nucleus. Adapted with permission from Newman
SW. The medial extended amygdala in male reproductive behavior. A node in the mammalian social behavior network. Ann N Y Acad Sci
1999;877:242–257.

hypothalamic areas, it is important to empha- cortex. Subcortically, the PAG receives input
size that there is no one-to-one mapping from a variety of structures, including amyg-
between area and function. Instead, hypotha- dala, hypothalamus, dorsal raphe, parabrachial
lamic areas are involved in many functions (eg, nuclei, pontine and medullary reticular forma-
stress responses, temperature regulation, food tion, and spinal cord. Many of these regions
intake, and defensive behavior), and at the same receive reciprocal projections from the PAG.
time functions rely on many hypothalamic areas Cytoarchitectonic features and physiologi-
(Fig. 29.5). Most importantly, the functions in cal responses indicate that PAG subregions are
which the hypothalamus participates are deter- organized more or less like columns that are
mined by neuronal networks that extend well aligned in parallel to the long axis of the brain
beyond this structure. stem (vertically in humans; horizontally in
rodents) (Fig. 29.6). At least two columns help
organize defensive behaviors (although recent
Periaqueductal Gray
data also reveal local integration of inputs, that
The brain is bathed in a colorless fluid also is, within columns). Excitation of neurons in the
found in the spine called the cerebrospinal fluid “active” column generates behaviors such as
which, among other things, provides buoyancy facing and backing away, or a full-blown flight
and protection to the brain. The PAG (also called reaction, reactions that are very similar to natu-
central gray) surrounds the cerebral aqueduct ral ones produced by a rat or cat when threat-
containing cerebrospinal fluid (hence the name) ened or attacked. Excitation of neurons in the
in the rostral brain stem (Figs. 29.1 and 29.6). “passive” column generates an entirely different
Cortically, the PAG receives input from several reaction, namely, the cessation of ongoing activ-
areas along medial prefrontal cortex, including ity and profound hyporeactivity, with the animal
cingulate areas and dorsomedial sectors, as well neither orienting nor responding to its environ-
as orbitofrontal cortex, insula, and temporal ment. This type of freezing behavior is similar to
 Regions of the Emotional Brain 643
information and, like the parabrachial nucleus,
provides an interface between autonomic and
sensory systems. Activation and behavioral
data indicate that PAG columns are important
during responses to different classes of pain.
Interestingly, the PAG is involved in analgesia
too. Early studies in rats that used simple with-
drawal reflexes as a measure of pain showed
that stimulation of the PAG inhibited responses
to noxious stimulation. Subsequent experiments
demonstrated antinociceptive effects in humans
too. For example, electrical stimulation of the
PAG produced clinically significant pain relief.

Cingulate Cortex
The cingulate cortex surrounds the entire
extent of the corpus callosum. The anterior sec-
tor of the cingulate gyrus (Fig. 29.7A) is involved
in a broad range of processes, including willed
action, executive function, and emotion. A
remarkable property of this cortical tissue is that
its “descending” projections are probably more
extensive than that of any other cortical region
FIGURE 29.6  Periaqueductal gray (PAG). The PAG is and includes major connections to autonomic
organized as elongated columns of neurons. Figure idea based regulatory structures; notably to lateral hypo-
on Figure 1 of Bandler R, Shipley MT. Columnar organization in thalamus, PAG, parabrachial nucleus, and the
the midbrain periaqueductal gray: modules for emotional expres-
nucleus of the solitary tract. This connectivity is
sion? Trends Neurosci 1994;17:379–389.
consistent with studies that have documented
effects of cingulate electrical stimulation on vir-
that of an animal that has incurred an injury, or tually all autonomic processes, as well as many
after defeat in a social encounter. An important endocrine mechanisms.
feature of the behaviors elicited by PAG excita- Conversely, the cingulate cortex is the recipi-
tion is that they are coordinated actions; that is, ent of ascending signals from the brain stem.
they are not simply isolated reactions (such as a The most notable of these is perhaps the one
knee-jerk reflex), but full-blown behaviors. from the nucleus of the solitary tract, the major
The PAG plays an important role in nocicep- visceral sensory cell group in the brain. Sev-
tion (pain signaling). The PAG receives direct eral nociceptive circuits also reach anterior and
(and indirect) nociceptive input from the dorsal mid-cingulate areas indirectly via the thalamus.
horn of the spinal cord, which is sensitive to noci- Thus, the cingulate cortex participates in more
ceptive information from the viscera, skin, and than “motor” (ie, output) autonomic functions.
other organs. In addition, the PAG is reciprocally The anterior part of the cingulate cortex is one
interconnected with the parabrachial nucleus of the most intensely studied areas in human
and the nucleus of the solitary tract. The latter neuroimaging, and is activated by a wide range
fulfills a key role in the processing of visceral of experimental paradigms. Based on early
644 29.  THE EMOTIONAL BRAIN

FIGURE 29.7  Anterior sector of the cingulate cortex. (A) Medial surface of cortex with the anterior sector of the medial
prefrontal cortex outlined in black. General locations of dorsal (D) and ventral (V) parts of the medial prefrontal cortex are
indicated. Anatomically, the cingulate cortex is delimited dorsally by the cingulate sulcus, although cortex dorsal to the sul-
cus is often referred to as “cingulate cortex.” S, subgenual anterior cingulate cortex. (B) Cognition and emotion in the medial
frontal cortex. Foci of activation across studies of negative affect and cognitive control, demonstrating extensive overlap
between the two. (B) Kindly provided by Alex Shackman and adapted with permission from Shackman AJ, Salomons TV, Slagter HA,
Fox AS, Winter JJ, Davidson RJ. The integration of negative affect, pain, and cognitive control in the cingulate cortex. Nat Rev Neurosci
2011;12:154–167.

imaging studies, it was proposed that this ante- treatment-resistant depression patients. A cur-
rior region could be subdivided into a rostral rent hypothesis is that treatment efficacy is
“affective division” and a caudal “cognitive divi- mediated via cascading effects of the initial sub-
sion,” sectors putatively sensitive to emotional genual stimulation on a distributed network of
and cognitive manipulations, respectively. Recent brain regions, such as the amygdala, hypothala-
meta-analyses of neuroimaging studies that take mus, striatum, and orbitofrontal cortex.
into account hundreds of studies have revealed, In rodents, parts of the medial PFC, the so-
instead, considerable overlap of sites in medial called “prelimbic” and “infralimbic” cortices,
prefrontal cortex engaged during negative affect that have important roles in emotion are dis-
and cognitive control (Fig. 29.7B). Activation sites cussed in section “Interactions Between the
of negative affect studies encompass a broad Amygdala and Medial Prefrontal Cortex.”
expanse of medial prefrontal cortex, as do those
of cognitive control studies, although the lat-
Orbitofrontal Cortex
ter exhibit a discernible concentration of sites in
more dorsal regions. Indeed, studies of cognitive Orbitofrontal cortex is the portion of prefron-
control observed activation sites in dorsomedial tal cortex that sits just above the orbits of the
prefrontal cortex more often than they did in ven- eyes and extends posteriorly several centimeters
tromedial prefrontal cortex. Notably, however, to form the frontal base of the brain (Fig. 29.8).
in studies of negative affect, activation of dorsal Based on anatomical connectivity information, it
and ventral prefrontal cortex was equally likely. has been suggested that the orbitofrontal cortex
These findings thus undermine the notion of a can be subdivided into orbital and medial compo-
strict segregation between emotion and cognition nents, or networks. The orbital network receives
in medial prefrontal cortex. wide-ranging sensory information and appears
The participation of the anterior cingulate to integrate it, particularly in relation to the
cortex in emotion has important clinical impli- assessment of food and reward. The medial net-
cations. Chronic deep brain stimulation of the work is distinctively and heavily connected with
white matter in the vicinity of the subgenual regions of the medial wall of the brain, including
cingulate cortex (Fig. 29.7A) produces consid- those of cingulate cortex and surrounding areas.
erable remission of symptoms in at least some In contrast to the orbital network, the medial
 Regions of the Emotional Brain 645

FIGURE 29.9  Anterior insula. The insula is a territory of

cortex folded within the lateral sulcus between the frontal
and temporal lobes, and is not visible from the lateral surface
of the brain. The approximate location of the anterior insula
(AI) is shown.

FIGURE 29.8  Orbitofrontal cortex. Ventral surface of the

brain showing the approximate location of the orbitofrontal parts, including the viscera, signals reporting
cortex. temperature, pain, itches, muscular sensations,
touch, and other bodily sensations. Both sympa-
network receives few sensory inputs. Impor- thetic and parasympathetic bodily signals are at
tantly, the medial orbitofrontal cortex projects first conveyed to the posterior insula, and then
to the hypothalamus and other areas involved to the anterior insula via the mid-insula. These
in visceral processing. In fact, via the hypo- properties have led researchers to suggest that
thalamus, the descending medial orbitofrontal the anterior insula participates in the representa-
influence has the potential to extend as far as tion of “feelings” from the body.
autonomic centers in the spinal cord. In contrast, In all, whereas an output or motor role is
there are relatively few projections to the hypo- often emphasized for the cingulate cortex, an
thalamus from the orbital network. input or sensory role is frequently emphasized
for the insula. But as discussed earlier, the cin-
gulate cortex also receives visceral, ascending
Anterior Insula signals from important autonomic nuclei. In the
Insular cortex is the part of cortex folded case of the insula, descending connections to the
within the lateral sulcus (also called the “Syl- hypothalamus and brain stem (for instance, the
vian fissure”) between the frontal and temporal parabrachial nucleus) also exist. Thus, in both
lobes—it is thus not visible from the lateral sur- cingulate cortex and the insula, bidirectional
face of the brain (cortex covering it is referred to circuits are present, although they are likely
as the “operculum,” meaning “lid”) (Fig. 29.9). asymmetrical in their efficacies (stronger role
The anterior insula is strongly involved in the for output for the cingulate and stronger role for
processing of bodily signals as it contains repre- input for the insula).
sentations of the internal state of the body. This Interestingly, in human neuroimaging stud-
representation involves signals from many body ies, researchers have found that the anterior
646 29.  THE EMOTIONAL BRAIN

insula is engaged by a remarkable array of tasks extremely high connectivity (see section: Prin-
spanning perception, cognition, and emotion. ciple 1: Massive Anatomical Connectivity).
Indeed, the anterior insula has been suggested to The pattern of connectivity between the
be among the most functionally diverse regions amygdala and prefrontal cortex is of particu-
of the brain. lar interest given the latter’s role in cognitive
functions. In addition to substantial connec-
tions between the amygdala and both medial
FROM REGIONS TO CIRCUITS and orbital aspects of prefrontal cortex, bidirec-
tional connectivity is present along the lateral
Earlier we reviewed some of the regions that surface too, although these connections are rela-
participate in emotional processing. But more tively weak. More generally, to understand how
broadly, how should we understand the brain amygdala signals are potentially “broadcast” to
basis of complex mental functions such as emo- all sectors of prefrontal cortex, it is important
tion? Many accounts of mental function empha- to consider prefrontal connectivity itself. In one
size the importance of particular brain regions. study, although the amygdala was estimated to
In the case of emotion, the hypothalamus, for be directly connected to about 40% of prefron-
example, was considered an emotion “center” tal regions, about 90% of prefrontal cortex was
for several decades. Likewise, the amygdala has shown to be capable of receiving amygdala sig-
played a major role in models of the emotional nals after a single additional connection within
brain in the past few decades. However, it is prefrontal cortex. This “one-step” property
increasingly recognized that functions arise via shows that amygdala signals are not necessar-
the coordinated action of multiple brain regions ily confined to orbital and medial prefrontal
that participate in distributed circuits. In this territories, both of which are directly connected
section, particular examples of this type of “dis- with the amygdala. Furthermore, this architec-
tributed processing” are illustrated in the con- ture indicates how amygdala signals are able to
text of emotional processing. influence responses in lateral prefrontal cortex,
even though direct connections to this part of
cortex are relatively weak.
Large-Scale Amygdala Networks Let us consider the potential implications of
A remarkable property of the amygdala is the amygdala’s connectivity on its influence on
its massive interconnectivity; as many as 1000 signal processing in the brain. Even a simpli-
separate cortical and subcortical pathways may fied view of its anatomical connectivity shows
exist. The connectivity is all the more notable that, minimally, it belongs to three networks—in
given that it involves all cortical lobes (though addition to prefrontal networks—as discussed
connections to parietal cortex are sparser) as in the previous paragraph. The first is a “visual
well as subcortex (see Fig. 29.3). Combined, network,” as the amygdala receives fibers from
these properties indicate that the amygdala is anterior parts of temporal cortex. Indeed, cells in
an extensively interconnected hub, namely, a the amygdala respond to visual stimuli, includ-
region that exhibits a high degree of anatomical ing faces, with response latencies a little longer
connectivity (Fig. 29.10). Indeed, computational than cells in anterior visual cortex (a region that
analysis of databases that collate anatomical con- responds to abstract visual stimuli such as spe-
nectivity suggests that several amygdala nuclei cific shapes, as well as faces). The amygdala, by
(including the lateral nucleus and the accessory its turn, influences visual processing via a set of
basal nucleus) should be considered as part of projections that reach most of ventral occipito-
a “core” brain circuit, all of whose regions have temporal cortex (including primary visual
 From Regions to Circuits 647

FIGURE 29.10  Brain connectivity graph. The amygdala (Amyg) is highly connected brain region. Note that the central
position of the amygdala in the figure does not mean that it is the most connected (or “central”) region in the brain; other
highly connected regions, including the hippocampus (Hipp) and entorhinal cortex (ER), could have been displayed at the
center. Reproduced with permission from Pessoa L. On the relationship between emotion and cognition. Nat Rev Neurosci 2008;9:148–158
and adapted with permission from Young MP, Scannell JW, Burns GA, Blakemore C. Analysis of connectivity: Neural systems in the
cerebral cortex. Rev Neurosci 1994;5:227–249.

cortex). Thus, for instance, responses to emo- events, even those removed from actual reward
tional faces are stronger than those to neutral delivery, if they provide information about the
ones in occipito-temporal cortex. likelihood of a future reward.
The second is the well-known “autonomic In total, the amygdala affiliates with dif-
network,” as evidenced by connectivity with ferent sets of regions, and thus participates in
subcortical structures such as the hypothalamus several networks, in a highly flexible and con-
and PAG. Via this network, the amygdala par- text-dependent manner. Other notable aspects
ticipates in many complex autonomic mecha- of amygdala connectivity include interactions
nisms, such as enhancing bodily arousal. between the amygdala and the basal forebrain
The third is a “value network,” as evidenced that are important for attentional functions (see
by its connectivity with orbitofrontal cortex, section: Basal Forebrain), as well as how sub-
medial prefrontal cortex, and striatum. The stantial projections from the amygdala to visual
orbitofrontal cortex, for example, is important cortex influence competition mechanisms in
for determining the relative value of the current visual cortex, thus contributing to attention as
“state,” and orbitofrontal responses differentiate well.
648 29.  THE EMOTIONAL BRAIN

Interactions Between the Amygdala and the extinction and expression of behaviors asso-
Medial Prefrontal Cortex ciated with aversive stimuli. Finally, evidence
As described, the pairing of an otherwise for contributions of different parts of the medial
neutral CS with an aversive US leads to aver- prefrontal cortex in extinction and response
sive learning; subsequent presentation of the CS expression has also been garnered in humans.
alone generates several of the properties of the Like in rodents, more dorsal (and posterior in
US, including freezing behavior and other phys- humans) parts are involved in response expres-
iological responses that prepare the organism to sion (and appraisal/evaluation of emotional
handle the impending danger. However, upon stimuli) and ventral (and anterior in humans)
repeated presentations of the CS without accom- parts are involved in response extinction.
panying USs, the acquired responses dissipate.
For example, freezing behavior to a tone that
was previously paired with shock decreases.
Hypothalamic Circuits
This phenomenon, which is called extinction, However important the hypothalamus may
does not simply amount to forgetting, but instead be for descending influences, a significant recent
appears to involve new learning that inhibits or insight is that the mammalian cerebral cortex and
overrides past learning. In other words, extinc- the hypothalamus share massive bidirectional
tion is not a passive forgetting process, but an connections. In the rat, the best-studied mam-
active learning one. In the rat, a ventral por- mal species, there are four major routes from the
tion of the medial prefrontal cortex is critically hypothalamus to the cerebral cortex (Fig. 29.11).
involved in extinction. Animals with lesions to These include a robust direct projection to all
this so-called infralimbic region of medial pre- parts of the cortical mantle, and three indirect
frontal cortex show resistance to extinction, that routes by way of the thalamus, basal brain nuclei
is, they show resistance to unlearning the origi- (specifically, basal forebrain and amygdala), and
nal CS-to-US association. brain stem. In all, the hypothalamus appears to
The role of prefrontal cortex in extinction be the largest source of nonthalamic direct input
resonates with the long-held idea that this to the cortex. In the rat, some notable targets of
region helps regulate behavior, for instance, as hypothalamic input include regions in medial
espoused by the 19th century British neurologist prefrontal cortex (infralimbic, prelimbic, and
Hughlings Jackson who proposed that “higher” anterior cingulate cortex) and insular cortex.
centers exert inhibitory control over “lower” Interestingly, less prominent projections of the
brain regions. However, the picture of higher hypothalamus to lateral prefrontal cortex and
regions inhibiting lower ones is too impover- even to primary sensory areas have been found.
ished. Indeed, recent studies implicate a region The connections between prefrontal cortex and
of the medial prefrontal cortex called the pre- the hypothalamus have also been investigated
limbic cortex in “fear” excitation (this region is in nonhuman primates, where they were found
immediately dorsal to the infralimbic cortex). to closely resemble those observed in rats. Nota-
Importantly, the prelimbic cortex appears to be bly, the hypothalamus has widespread projec-
critical for fear expression. Prelimbic neurons tions to all sectors of the prefrontal cortex.
exhibit sustained increases in activity lasting tens In sum, whereas the hypothalamus is involved
of seconds that parallel the time course of aver- in a host of “basic” control functions, it is part
sive behaviors. In all, the prelimbic and infralim- of an extensive bidirectional connective system
bic cortex appear to integrate multiple types of with cortex and many other subcortical struc-
inputs with emotional significance to regulate tures in a manner that allows for integration of
 From Regions to Circuits 649

FIGURE 29.11  Hypothalamic ascending connectivity.

Summary of the four major pathways from the hypothala-
mus to cerebral cortex schematized on a flattened represen-
tation of the rat brain. The basal ganglia here refer to the
basal forebrain and the amygdala complex. Note that one of
the indirect connections first descends to the brain stem. BG,
basal ganglia; BS, brain stem; CTX, cortex; HY, hypothala-
mus; TH, thalamus. Reproduced with permission from Risold
PY, Thompson RH, Swanson LW. The structural organization of
connections between hypothalamus and cerebral cortex. Brain Res
Brain Res Rev 1997;24:197–254.

wide-ranging signals. Critically, the hypothala- FIGURE 29.12  Basal forebrain. The basal forebrain
comprises heterogeneous structures close to the medial
mus is linked to other structures that have them- and ventral surfaces of the brain. The basal forebrain origi-
selves broad connectivity, including the basal nates ascending cholinergic (and GABAergic) projections
forebrain and the amygdala, further expanding that innervate extensively throughout the entire cortex. The
its potential for influencing information pro- approximate location of the basal forebrain is shown.
cessing. These insights thus resonate with the
observation by Morgane that the hypothalamus parietal, cingulate, frontal (including lateral and
is “only a single link in a larger multicircuit com- orbital regions), and insular cortex (Mesulam,
plex rather than … an entity unto itself control- 2000); substantial connections have also been
ling any behavior as some form of command found to both the hippocampus and amygdala
center” (p. 7). This view stands in stark contrast (Mesulam et al., 1992).
to the traditional viewpoint that the hypothala- Acetylcholine released onto cortical neurons
mus functions as a “head ganglion” that func- facilitates their response. Studies suggest that
tions as an overall controller region. directly applying acetylcholine or stimulating
the basal forebrain can increase the signal-to-
noise ratio of neuronal responses and sharpen
Basal Forebrain
response tuning curves. More broadly, given
The basal forebrain comprises a heterogeneous its widespread connections, the basal forebrain
set of structures close to the medial and ventral is in a favorable position to influence multiple
surfaces of the cerebral hemispheres (Fig. 29.12). sites across the brain, including sensory regions
These include the magnocellular basal nucleus, which play a central role in responding to envi-
which contains the basal nucleus of Meynert and ronmental stimuli. Overall, these topographi-
other cell groups. The basal forebrain originates cally distributed effects result in increased
ascending cholinergic (and GABAergic) projec- vigilance, alertness, or attention.
tions that innervate extensively throughout the Emotional content can influence signal pro-
entire cortex. Major projections reach visual, cessing across the brain via the basal forebrain
650 29.  THE EMOTIONAL BRAIN

because of the inputs that this structure receives. Distributed Circuits and the Evolution of
Connections from cortex originate from sites in the Emotional Brain
orbitofrontal, temporal, insular, and cingulate
cortex, all of which are important for emotion. From an evolutionary perspective, it is some-
Major subcortical projections to the basal fore- times stated that cognition can be considered
brain originate from the amygdala (in particular, “the tip of the iceberg,” often associated with
the central amygdala), hypothalamus, and vari- “newer” parts of the cortex, while the “rest of the
ous brain stem cell groups, including midbrain iceberg” involves emotional processes that fre-
dopaminergic cells and the locus coeruleus. In quently rely on “old” subcortical brain regions.
all, given its overall connectivity pattern, the Furthermore, so the story goes, the “old” emo-
basal forebrain is in a position to influence pro- tional system functions in a largely autonomous
cessing across the brain based on multiple types manner.
of signals, including those central to emotion. The brain has what can be considered old sys-
tems. For example, anatomical and behavioral
data suggest that ray-finned fishes and land ver-
Lateral Prefrontal Cortex tebrates probably share an ancestor that already
The role of lateral prefrontal cortex in a range possessed an amygdala. And the amygdala of
of cognitive operations is well documented. the ancestral amniote (which includes reptiles,
Indeed, this region is perhaps the paradigmatic birds, and mammals) is believed to have pos-
“cognitive” brain region. Yet, the connectiv- sessed an amygdala with so-called pallial and
ity pattern of lateral prefrontal cortex is quite subpallial components, namely the ancestral
extensive, indicating that its functions should form of the basolateral and central amygdala
be conceptualized more broadly. For example, highlighted in this chapter. However, “old”
the dorsolateral prefrontal cortex is not only an regions do not exist within a “layered” architec-
important hub, but is also strongly connected ture, with newer structures simply added on top
with other regions that are topologically cen- of old ones, such that the ones on top control the
tral themselves, which considerably expands ones at the bottom (Fig. 29.13). The examples dis-
the reach of the dorsolateral prefrontal cortex cussed in the preceding sections, including the
in influencing, and being influenced by, other amygdala and the hypothalamus, illustrate that
regions. Being robustly bidirectionally con- the old/new conceptualization that is central to
nected to medial and orbital prefrontal areas, as many versions of the emotional brain, does not
well as to the insula, the lateral prefrontal cortex take into account the extensive anatomical con-
has access to afferent signals from the body and, nectivity of brain circuits.
at the same time, the region can influence signals Recent studies focusing on the amygdala
that reach the body. Furthermore, as previously further inform the understanding of the evolu-
discussed, the amygdala is not only directly con- tion of a region that is important for emotion.
nected to a substantial set of prefrontal regions, For example, the lateral, basal, and accessory
but nearly all of prefrontal cortex is capable of basal subregions of the amygdala are consider-
receiving amygdala signals after a single addi- ably more “developed” in monkeys than in rats
tional connection within prefrontal cortex itself (based on morphological characteristics, such as
(what was referred to as the “one-step” prop- cell counts and the volume of subregions). One
erty). Overall, the connectivity pattern of the suggestion is that the differences between rats
lateral prefrontal cortex allows it to play a sig- and monkeys are linked to the degree of con-
nificant role in the integration of “cognitive” nectivity of these nuclei with other brain struc-
and “emotional” signals. tures, in line with the framework of correlated
 How Are Emotions Organized in the Brain? 651

FIGURE 29.13  Brain evolution. (A) Brain evolution conceptualized as an additive process, with newer structures being
added on top of older ones, and exerting control over them. Although this view was espoused by some comparative neu-
roanatomists in the early 20th century, and popularized well into later parts of the century (most notably in the idea of the
“triune brain”), it does not reflect current knowledge of comparative neuroanatomy. Importantly, it does not take into account
the extensive anatomical connectivity of brain circuits in general, and between cortical and subcortical regions in particular,
as schematically shown in part (B).

evolution between components of functional brain” and, as some have pointed out, the term
systems. In this view, the lateral, basal, and “limbic system” substitutes naming for under-
accessory basal nuclei are more developed in standing. Thus, the designation “emotional
primates than in rodents, paralleling the greater brain” is preferable, and the term “limbic sys-
development of the cortical areas with which tem” should be abandoned.
these nuclei are interconnected in primates. In Nevertheless, two concepts related to early
other words, the relative development of these notions of the “limbic system” are worth con-
amygdala nuclei is influenced by their inter- sidering. One concept is the broad organizing
connections with other brain structures. Fur- principle that along a medial–lateral axis of the
thermore, such correlated evolution would be brain, a gradient of affective/motivational sen-
responsible for a higher convergence and inte- sitivity is observed. By and large, medial brain
gration of information in the primate amygdala. structures (along the inner edge or limbus of
the brain) are more sensitive to affective and
motivational variables, while more lateral brain
HOW ARE EMOTIONS ORGANIZED regions are less so. But the sensitivity is a matter
IN THE BRAIN? of degree because all brain regions are suscep-
tible to emotional and motivational influences,
In the late 1940s, MacLean proposed the idea to some degree. There are important exceptions
of a “limbic system” to explain how emotion to this broad principle. For example, the anterior
is organized in the brain. Although the term is insula, which participates in important emotion-
probably one of the most widely used in neu- related functions, lies more laterally (although it
roscience, the concept has proven too unwieldy is covered by the operculum of the frontal and
and unstable to be scientifically useful. Impor- parietal cortices).
tantly, agreement regarding the brain regions Another concept is the broad principle that
that belong to this system has never been parts of the brain that are more sensitive to emo-
attained. Because of this, the term is often used tional variables are less structured. Many brain
in a circular fashion to indicate the “emotional regions important for emotional processing are
652 29.  THE EMOTIONAL BRAIN

subcortical, and subcortex is cytoarchitectur- brain must be understood from the standpoint of
ally poorly structured. Cortex varies in degree this general architectural property. Furthermore,
of lamination, from three main layers to six or it is noteworthy that many regions implicated in
more well-demarcated layers. Agranular cortex emotion were proposed to be part of the brain’s
contains three layers (it lacks a conspicuous core circuit. Finally, the property of extensive
layer IV), dysgranular cortex has four layers, anatomical connectivity is further illustrated,
and granular cortex contains six layers. Because for example, by the discussion of the amygdala
cortex along Broca’s “great limbic lobe” (ie, and hypothalamus in section “From Regions to
the medial edge of the hemispheres) is mostly Circuits.”
not granular, it follows from its location and
the medial–lateral gradient noted previously
that less individuated cortex is relatively more Principle 2: Anatomical Connectivity to
involved in emotional processing. and From the Body
Next, three other principles of the emotional Because emotion is closely linked to bodily
brain are described. states, brain circuits that involve signals to and
from the body are important for emotion.
To illustrate the to-the-body part, consider the
Principle 1: Massive Anatomical
hypothalamus. Classical studies showed that
Connectivity “decortication” (ie, removal of cortex) alone did
Computational analysis of anatomical con- not abolish emotional expression, but that decor-
nectivity demonstrates that both cortical and tication plus ablation of the hypothalamus has
subcortical brain regions are massively inter- a much more substantial effect on this expres-
connected (Fig. 29.14). Massive connectivity is sion. Current understanding of hypothalamic
not limited to specific sectors of the brain but function indeed reveals that this structure is
instead includes all of them: occipital, tempo- involved in important body-related operations.
ral, parietal, and frontal lobes, as well as cingu- As described in section “Hypothalamus,” hypo-
late, insula, thalamus, and basal brain regions thalamic mechanisms contribute to a wealth of
(subcortical nuclei at the base of the forebrain, processes, including circadian rhythms, wake-
including the amygdala and basal ganglia). fulness and sleep, stress responses, temperature
Related analyses have described a tightly regulation, food intake, thirst, sexual behaviors,
integrated “core circuit” spanning all of the and defensive behaviors. To carry out these
brain sectors mentioned earlier. This core circuit functions, the hypothalamus works in concert
is topologically central, that is, strongly con- with a multitude of other sites, several of which
nected to all other regions of the core and the are located in the brain stem and spinal cord.
rest of the brain. The core circuit was proposed Cortical regions with substantial body-related
to have several important properties: (1) it is a signals include the medial orbitofrontal cortex,
subnetwork that is far more tightly integrated the cingulate gyrus, and the insula—illustrating
than the overall brain network; (2) information the from-the-body part. Consider, for example,
likely spreads more swiftly within the core than the insula: the physiological condition of the
through the overall brain network; and (3) the entire body is conveyed to the posterior insu-
overall brain network communicates with itself lar cortex, including individually mapped and
mainly through the core. distinct feeling-related signals from the body.
Although the property of massive anatomi- These signals are also present in the mid-insula
cal connectivity is a principle that applies to the and the anterior insula, which integrate body-
entire brain, the organization of emotion in the related signals with activity that is associated
 Principle 3: High Functional Connectivity With Most of the Brain 653

FIGURE 29.14  Computational analysis of whole-brain anatomical connectivity. The analysis considered existing data of
an extensive set of cortical and subcortical regions (indicated via colored rectangles) spanning major brain sectors. The basal
ganglia refer to nuclei at the base of the forebrain, including the amygdala. Reproduced with permission from Modha DS, Singh R.
Network architecture of the long-distance pathways in the macaque brain. Proc Natl Acad Sci USA 2010;107:13485–13490.

with emotionally salient stimuli (eg, via sensory PRINCIPLE 3: HIGH FUNCTIONAL
regions, the temporal pole, and the amygdala). CONNECTIVITY WITH MOST OF
Thus, the insula, especially its posterior part, THE BRAIN
can be considered interoceptive cortex, much
like parts of parietal cortex are somatosensory Understanding how regions contribute to
cortex, for example. brain function requires characterizing how they
The combination of principles 1 and 2 reveals are “functionally connected,” in addition, of
that brain regions that influence the body, and course, to how they are anatomically connected.
conversely brain regions that are influenced by Functional connectivity is a concept that was
the body, extend well beyond those with more devised to characterize how neurons interact
direct anatomical pathways to and from the body. and was defined as the “coherence” among the
654 29.  THE EMOTIONAL BRAIN

FIGURE 29.15  Functional connectivity. Maps show correlations between a seed region in the posterior cingulate cortex
(PCC) and all other voxels in the brain for a single subject during resting fixation (both positively and negatively correlated
regions are shown). Thus, all colored voxels are functionally connected with the PCC. The bottom plot shows the time course
for a single run for the seed region (PCC, yellow), a region positively correlated with this seed region in the medial prefrontal
cortex (MPF, orange), and a region negatively correlated with the seed region in the intraparietal sulcus (IPS, blue). Reproduced
with permission from Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van Essen DC, Raichle ME. The human brain is intrinsically orga-
nized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci USA 2005;102:9673–9678.

activity of different neurons. Functional connec- of functional connectivity (as measured during
tivity thus answers the following question: how rest with functional MRI). Thus, largely normal
coordinated is the activity of two brain regions coordinated activity can emerge in brains with
that may or may not be directly anatomically dramatically altered structural connectivity. In
connected (Fig. 29.15)? all, the functional organization of the brain is
Whereas it is natural to anticipate a func- driven by factors that go beyond direct struc-
tional association between brain regions that tural connectivity.
are directly connected, the relationship between Therefore, to understand the organization of
structural and functional connectivity is not emotion in the brain, in addition to character-
always a simple one. A striking example of izing the brain’s structural anatomy, it is neces-
structure–function dissociation is illustrated by sary to characterize the functional relationships
an unusual population of adults without the between brain regions. Importantly, anatomical
corpus callosum, a structure that provides the architectural features support the efficient com-
major communication pathway between the two munication of information even when strong
hemispheres. Although starkly different struc- direct structural connections are not present,
turally relative to controls, individuals without and support functional interactions that will
the callosum exhibited very similar patterns vary based on context. This is illustrated, for
 Further Reading 655
example, by the “one-step” property of amyg- circuits that involve nonlocal connections that
dala–prefrontal cortex connectivity discussed bring regions, which are anatomically far from
previously (amygdala signals reach nearly all one another, together into functional circuits.
prefrontal regions via a single connectivity step In all, to understand how emotions are orga-
within prefrontal cortex; section “From Regions nized requires understanding general princi-
to Circuits”). This anatomical property allows ples of information processing in the brain (as
the amygdala to engage in functional interac- a whole). Thus, to understand the emotional
tions with, for instance, lateral prefrontal regions brain, it is critical to understand the sensory,
that are not supported by strong direct anatomi- motor, and cognitive brain. In this way, it is also
cal connections. possible to characterize and understand the
full ramifications of emotional psychopatholo-
gies, such as the anxiety disorders and depres-
CONCLUSIONS sion, whose impact on mental function is not
restricted to emotion, but necessarily encom-
The search to understand the emotional brain passes additional dimensions, most notably
is as active today as it ever was. With time, the list cognition.
of areas that have been proposed to play impor-
tant roles in emotion has grown from having
just a few members (hypothalamus, amygdala) Further Reading
to a large list. Notably, this list is not “fixed,” but [1] Bains JS, Cusulin JIW, Inoue W. Stress-related synap-
some members are included and highlighted by tic plasticity in the hypothalamus. Nat Rev Neurosci
2015;16:377–88.
some authors but not others. This is not capri- [2] Bandler R, Shipley MT. Columnar organization in the
cious; instead, it is because brain regions do not midbrain periaqueductal gray: modules for emotional
implement a single function (say, the amygdala expression? Trends Neurosci 1994;17:379–89.
is responsible for just aversive learning), but [3] Craig AD. How do you feel—now? The anterior insula
participate in many functions (the amygdala and human awareness. Nat Rev Neurosci 2009;10:59–70.
[4] Davis M, Walker DL, Miles L, Grillon C. Phasic vs sus-
is involved in attention and decision making, tained fear in rats and humans: role of the extended
among other functions). Therefore, the choice is amygdala in fear vs anxiety. Neuropsychopharmacol
one of emphasis. For example, the hippocampus 2010;35:105–35.
is centrally important for memory processes. [5] Fields H. State-dependent opioid control of pain. Nat
However, although it was not discussed in this Rev Neurosci 2004;5:565–75.
[6] Grupe DW, Nitschke JB. Uncertainty and anticipation
chapter, it also participates in emotional pro- in anxiety: an integrated neurobiological and psycho-
cessing, and has been particularly investigated logical perspective. Nat Rev Neurosci 2013;14:488–501.
in its contributions to anxiety. [7] Heimer L, Van Hoesen GW, Trimble M, Zahm DS.
Another reason the list of emotional brain Anatomy of neuropsychiatry. Amsterdam: Academic
regions is not fixed is because the boundaries of Press; 2008.
[8] Johansen-Berg H, Gutman DA, Behrens TE, Matthews
the emotional brain are ill defined. As reviewed PM, Rushworth MF, Katz E, et al. Anatomical connec-
in the present chapter, the brain is characterized tivity of the subgenual cingulate region targeted with
by an architecture that exhibits massive intercon- deep brain stimulation for treatment-resistant depres-
nectivity. This connectivity implies that signals sion. Cereb Cort 2008;18:1374–83.
in a given structure can readily influence other [9] LeDoux JE. Emotion circuits in the brain. Annu Rev
Neurosci 2000;23:155–84.
parts of the brain in a few synapses—in many [10] Martínez-García F, Novejarque A, Lanuza E. Evolu-
cases in as few as one or two synapses. Impor- tion of the amygdala in vertebrates. In: Kaas JH, editor.
tantly, brain regions do not implement functions Evolution of nervous systems, vol. 2. Oxford: Academic
in isolation. Instead, they are part of distributed Press; 2007. p. 255–334.
656 29.  THE EMOTIONAL BRAIN

[11] Mesulam MM. Principles of behavioral and cognitive [16] Pessoa L. The cognitive-emotional brain: from interac-
neurology. Oxford University Press; 2000. tions to integration. Cambridge: MIT Press; 2013.
[12] Mesulam M, Hersh LB, Mash DC, Geula C. Differential [17] Pessoa L. Understanding brain networks and brain
cholinergic innervation within functional subdivisions organization. Phys Life Rev 2014;11:400–35.
of the human cerebral cortex: a choline acetyltransfer- [18] Sotres-Bayon F, Quirk GJ. Prefrontal control of fear:
ase study. J Comp Neurol 1992;318(3):316–28. more than just extinction. Curr Opin Neurobiol
[13] Millan MJ. Descending control of pain. Prog Neurobiol 2010;20:231–5.
2002;66:355–474. [19] Tovote P, Fadok JP, Lüthi A. Neuronal circuits for fear
[14] Nieuwenhuys R, Voogd J, van Huijzen C. The human and anxiety. Nat Rev Neurosci 2015;16:317–31.
central nervous system. 4th ed. New York: Springer- [20] Tovote P, Esposito MS, Botta P, Chaudun F, Fadok
Verlag; 2008. JP, Markovic M, et al. Midbrain circuits for defensive
[15] Morgane PJ. Historical and modern concepts of hypo- behaviour. Nature 2016;534:206–12.
thalamic organization and function. In: Morgane PJ, [21] Vogt BA, editor. Cingulate neurobiology and disease.
Panksepp J, editors. Handbook of the hypothalamus, Oxford: Oxford University Press; 2008.
vol. 1. New York: Marcel Dekker Inc.; 1979. p. 1–64.