They arise from the olfactory bulb itself, the anterior olfactory nucleus (located in the groove occupied by the olfactory tract) and the lateral olfactory stria (Y Soudry et al. 2011). Let us recall that at the macroscopic level, this stria reaches the prepiriform cortex at the limen (several fibres reaching the primary olfactory cortex located in the insular angle in front of the agranular cortex of the insula) and gives a dissectible large contingent (see Fig. 7.​1), detaching almost perpendicularly to reach the convexity which raises the cortical nucleus of the amygdala, at the level of the periamygdaloid anterior cortex. Research carried out by DL Rosene and CW Van Hoesen in 1977 and then by ST Carmichael et al. in 1994 that used, in macaques, a combination of anterograde and retrograde axonal tracers to study these inputs confirmed that the fibres of olfactory origin come to the cortical nucleus of the amygdala and the periamygdaloid cortex (semilunar gyrus and gyrus ambiens). Otherwise, the prepiriform cortex which also receives a macroscopic quota from lateral olfactory stria projects at the baso-lateral nucleus of the amygdala (TPS Powell et al. 1965). The olfactory bulb and prepiriform cortex also target the three parts (dorsal, ventral and ventromedial) of the entorhinal cortex (JE Krettek and JL Price 1977, online 2004) whose deep layers project to the amygdala (M Hoïstad and H Barbas 2008).

They have been studied by JP Aggleton et al. in 1980. They mainly come from the ventromedial hypothalamus and from other hypothalamic nuclei (including paraventricular and arcuate nuclei and the lateral hypothalamus) which are also involved in these inputs that go to the medial and central nuclei of the amygdala, leading visceral inputs. Note also that the anatomical dissection has already shown us bundles of nerve fibres connecting the lateral hypothalamus to the central nucleus of the amygdala (see Fig. 7.​13).

The thalamic inputs come from the dorsal and ventral regions of the thalamus. They were identified in 1979 by OP Ottersen and Y Ben-Ari in cats and in 1980 by WR Mehler in monkeys and also the same year by M Norita and K Kawamura (1980) and by JP Aggleton et al., and then confirmed in 1990 by JE LeDoux et al. (1990b).

The thalamic nuclei which receive inputs from the spinothalamic tract (JE LeDoux 1987) project to the baso-medial nucleus (JE LeDoux et al. 1990a).

Inputs coming from the parvocellular part of the VPM, the ventral postero-medial nucleus (nucleus that transmits the visceral taste information), go to the lateral nucleus of the amygdala.

Inputs coming from the medial division of the medial geniculate nucleus, the posterior intralaminar nucleus or the posterior medial complex also go to the lateral nucleus of the amygdala (JM Edeline and NM Weinberger 1992). Thus, during fear conditioned experiments, auditory stimuli (coming from the medial division of the medial geniculate body) go to the lateral nucleus of the amygdala, while contextual stimuli from the hippocampus go to the baso-lateral and baso-medial nuclei (NS Canteras and LN Swanson 1992; JE LeDoux 2000).

They have also been the subject of much research, particularly in 1980, in the rat (OP Ottersen) and in the monkey (M Norita et al.; WR Melher; JP Aggleton et al.) and in 2000 (K Semba et al.). Thus, Aggleton et al. who practised in macaques macaca mulatta, fluorescent markings with the enzyme horseradish peroxidase, could highlight inputs coming from the BST and the nucleus of the horizontal part of the diagonal band and ending at the medial and central nuclei of the amygdala. They also observed inputs from the innominate substance reaching the basal nuclei of the amygdala: These afferent fibres belong to large cholinergic neurons, which corresponds to the area in which is found the basal nucleus of Meynert. Other inputs from the ventral pallidum project to the lateral and baso-lateral nuclei of the amygdala (OP Ottersen 1980).

As demonstrated by the works of WR Mehler (1980), M Norita and K Kawamura (1980), JP Aggleton et al. (1980), OP Ottersen and Y Ben-Ari (1979), OP Ottersen (1981) and J Ciriello et al. (1994), all brainstem neurons send axons to the central nucleus of the amygdala which they approach through the ventral pathway. This is the case of neurons in the periaqueductal grey matter, the pars compacta of the substantia nigra, the ventral tegmental area, nuclei of the medulla oblongata (locus coeruleus, dorsal raphe nucleus, lateral parabrachial nucleus, nuclei of the solitary tract, interpeduncular nucleus) and the ventrolateral nucleus of the spinal cord. According to J Hanaway et al. (2001), all these fibres borrow in their ascending path, the medial forebrain bundle and dorsal longitudinal fasciculus (see Figs. 7.​21 and 8.1).

The lateral parabrachial nucleus provides the largest contingent of these afferent fibres. It is involved in the severe, disabling tinnitus which results from interactions between auditory neural sites (cochlear and vestibular) and non auditory neural sites (parabrachial nuclei). These nuclei project the aberrant auditory stimuli towards the central nucleus of the amygdala. This one transfers its emotional responses to the insula which transforms them into conscious emotional feelings. We can see on fMRI the activations of the amygdala and the insula during the severe tinnitus and the regression of these activations when tinnitus was treated with high-frequency acoustic treatment (ML Lenhardt et al. 2007).

This is also from the parabrachial area and directly from the spinal cord that the central nucleus of the amygdala can receive nociceptive inputs (R Burstein and S Potrebic 1993). Since as in the amygdala, the nucleus is the effector of the lateral and basal nuclei of the amygdala, the messages transmitted to it by these nuclei can be modified by nociceptive inputs it receives, which can generate nociceptive outputs to the target organs.

The fibres come from the dorsal raphe nucleus, and fibres of the locus coeruleus (which are noradrenergic fibres) require not only the ventral pathway but also the terminal stria. As for inputs coming from substantia nigra and the ventral tegmental area, they explain the abundant dopaminergic innervation of the central nucleus (Y Ben-Ari 1981).

The work of DL Rosene, GW Van Hoesen (1977) and RE Saunders et al. (1988) in rhesus monkeys has clarified that these inputs come from CA1 from the prosubiculum and essentially go to the baso-medial nucleus and the ventral cortical nucleus. The lateral and central nuclei also receive hippocampal inputs.

In 2000, the Finnish team of A Pitkänen et al., which has made anterograde and retrograde studies in rats, confirmed these data stating that the inputs to the amygdala come from the rostral half of the entorhinal cortex, from CA1, from the subiculum and from areas 35 and 36 of the perirhinal cortex. The nuclei that receive these inputs are the lateral, baso-lateral, baso-medial, central nuclei as well as the amygdalo-hippocampal area.

In 2008, M Höistad and H Barbas reviewed the temporal projections to the amygdala from the medial part of the temporal pole, the entorhinal and perirhinal areas, the granular and the dysgranular insula (the densest connections) and the agranular insula and the parahippocampal region (scattered connections (see for structure of the insula, Fig. 8.2)). They also confirmed that there are feedback connections between the entorhinal cortex, which projects from its deep layers to the amygdala and receives back projections in layers II–III, which ultimately target the hippocampus. This pathway could explain “how the amygdala can attach emotional value to environmental stimuli, participate in the sequence of information processing of emotions and modulate the formation of emotional memories”.

They are particularly important and come from different associative cortical areas. They project for most to the baso-lateral nuclear group of the amygdala, which plays a major role for the receipt of inputs. GW Van Hoesen in 1981, L Stefanacci, DG Amaral in 2002 and JE LeDoux et al. (1990b) showed that the auditory inputs come from the anterior half of the superior temporal gyrus, that is to say, the rostral part of area 22 (extension of the lower section of the posterior long insular gyri or secondary acoustic area AII).

MJ Webster et al., in 1991 and JE LeDoux et al., in 1990a, have shown that inputs from the associative visual cortex come from the lower part of the temporal lobe, that is to say, areas 20 and 21.

We must also consider the very important role of the inferior longitudinal fasciculus (Fig. 7.​21) which directly brings to amygdala the visual stimuli. RM Bauer, in 1982, reported the case of a patient whose cranial trauma had caused temporal haemorrhages responsible for the bilateral deterioration of inferior longitudinal fasciculus. This patient had lost any emotion aroused by a visual substratum and regretted having been brought to cancel a subscription of the magazine Playboy whose diligent reader he was before his accident and for which he had no more, from now on, any interest.

Note that we have seen in terms of macroscopic anatomy the bundles of fibres that connect the amygdala to the rostral part of each temporal gyrus, conducting resections of the successive gyri from top to bottom of the temporal lobe.

AG Herzog, GW Van Hoesen (1976) and then BH Turner et al. in 1980 and L Stefanacci and DG Amaral in 2000, showed showed that inputs from the multisensory association areas come from the rostral portion of the perirhinal cortex (areas 35 and 36), the cortex forming the dorsal bank of the superior temporal sulcus, the entorhinal cortex, the superior temporal gyrus, the parahippocampal cortex and the temporopolar area (area 38). Let us recall again that the macroscopic study showed us the obvious and important connections between the amygdala and the temporal pole (see Fig. 5.​2).

EJ Mufson et al. (1981) studied inputs involved in gustatory and visceral functions: They come from anterior insular areas agranular areas (Fig. 8.2) that themselves receive impulses from the ventral postero-superior and ventral postero-inferior thalamic nuclei and from the parvicellular division of the ventral postero-medial nucleus of the thalamus. The agranular and dysgranular insular areas project to the lateral nucleus of the amygdala (L Stefanacci and DG Amaral 2000).

MM Mesulam, EJ Mufson in 1985; L Stefanacci, DG Amaral in 2002, studied the posterior part of the insula (Fig. 8.2): it contains a somatosensory association area (granular area), postero-superior, which is connected with the adjacent second sensory area (SmII) (located on the upper lip of the posterior segment of the lateral sulcus) which is involved in the process of nociceptive information.

In a more recent study (2008), M Höistad and H Barbas showed that the medial temporal pole, the entorhinal and perirhinal areas and the agranular and dysgranular parts of the insula have the densest connections with the amygdala, while the lateral temporal pole, the granular insula and the parahippocampal region have sparser connections.

All these inputs coming from the temporal regions (anterior superior temporal, anterior middle temporal and inferior temporal gyri, medial and lateral parts of the temporal pole) and from the insula go mainly to the lateral nucleus of the amygdala and little or not all to other nuclei (L Stefanacci and DG Amaral 2000). The temporal pole is the only one to also target the baso-medial nucleus (AG Herzog and GW Van Hoesen 1976).

We also know now thanks to experiments on conditioned fear (JE LeDoux 2000) that the cortex is not always involved in the immediate analysis of stimuli and the thalamus (receiver of the stimulus to the “medial division of the medial geniculate body”) can directly transfer the stimulus (auditory case) to the lateral nucleus of the amygdala. This has been demonstrated by JS Morris et al. in 1999 by fMRI studies on humans: These authors have observed that during conditioning, amygdala activity changes are correlated with those of the thalamus as is not the case with the cortex which proves the importance of the direct thalamo-amygdaloid trajectory and the direct effector possibilities of the amygdala, in extreme emergency, without cortical modulation! But even in these cases, the stimulus is going to reach the brain analyser controller, some milliseconds later!

Inputs to the amygdala from the frontal lobe are of extreme importance. They are numerous and include the posterior orbitofrontal areas, areas 13 and 14; the medial prefrontal areas, areas 11 and 12 (HT Ghashghaei and H Barbas¹ 2002; H Barbas 2007); the anterior cingulate area, area 24 (DN Pandya et al. 1973); the prelimbic area, area 32; and the infra-limbic area, area 25. All these inputs from the frontal lobe target the baso-lateral and the baso-medial nuclei of the amygdala. Numerous authors have devoted their research to these inputs, identifying a number of details:

GR Leichnetz, J Astruc (1977) showed that the prefrontal cortex not only projects to the amygdala but also to the substantia innominata and the pallidum.

L Stefanacci and DG Amaral (2000) found found that the orbitofrontal cortex; the prefrontal medial cortex, areas 10, 11, 12, 13, 13a and 14; and the anterior cingulate cortex, areas 25, 24 and 32, mainly target the baso-lateral and baso-medial nuclei and the orbitofrontal cortex and prefrontal medial cortex (but not the anterior cingulated cortex) and also project to the cortical and medial nuclei as well as to the periamygdaloid cortex.

NL Rempel-Clower, in 2007, observed in rats, as H Barbas did the same year in the monkey, that the orbitofrontal cortex also projects to the intercalated masses of the amygdala that are GABAergic and thus exerts inhibitory influences within the amygdala. H Barbas was able to clarify that, in fact, the orbitofrontal posterior cortex projects to the amygdala, through 2 “pathways” that have opposite effects on the central autonomous structures: The first path is the one mentioned above, which leads from the cortex to the intercalated mass and whose activation can ultimately disinhibit central autonomic structures during emotional arousal. The second system innervates the central nucleus of the amygdala that has inhibitory projections to the hypothalamus and autonomous structures of the brainstem and can cause inhibition of central autonomous structures, resulting in an autonomous homeostasis¹. HT Ghashghaei and H Barbas in 2002 showed that the lateral prefrontal cortex “which has executive functions” specifically projects to the layer 5 of the orbitofrontal cortex. In 2007, the same authors found that the orbitofrontal neurons sending their axons to the amygdala are from this cortical layer 5. They so demonstrate that also the lateral prefrontal cortex has a controlling role over the inner workings of the amygdala (especially on the posterior half of the amygdala, the intermediate sector of the baso-lateral nucleus and the magnocellular portion of the baso-medial nucleus). It was also in 2007 that M Medalla et al. were able to demonstrate that at the level of the prefrontal cortex, there are two specialised classes of inhibitor neurons, belonging respectively to each of the two pathways mentioned above and expressing the duality and thus the complexity of inhibitory control: the area of inhibitory neurons 32, dedicated to emotional communication, and those in area 10 (see mapping Fig. 7.​8), dedicated to working memory functions.

We were also able to demonstrate by dissection the connection between the terminal stria and the medullary stria of the thalamus (or stria medullaris). We see perfectly a direct posterior branch detaching from the stria terminalis prior to its development to continue directly with the medullary stria that will rejoin the habenula (see Fig. 6.​5), the starting point of the habenulo-interpeduncular tract.

They are created by the baso-lateral, baso-medial and central nuclei of the amygdala. They immediately use the ventral pathway and then target the structures of various regions: the septal region (particularly to the diagonal band nucleus), the sublenticular region (notably to the basal nucleus of Meynert), the BST, the thalamus and the lateral hypothalamic area.