Revisiting the hippocampal–amygdala pathway in primates: Association with immature-appearing neurons

Abstract

Elucidation of the ‘fear circuit’ has opened exciting avenues for understanding and treating human anxiety disorders. However, the translation of rodent to human studies, and vice versa, depends on understanding the homology in relevant circuits across species. Although abundant evidence indicates that the hippocampal–amygdala circuit mediates contextual fear learning, previous studies indicate that this pathway is more restricted in primates than in rodents. Moreover, cellular components of the amygdala differ across species. The paralaminar nucleus (PL) of the amygdala, a structure that is closely associated with the basal nucleus, is one example, having no clear homologue in rodents. In both human and nonhuman primates, the PL contains a subpopulation of immature-appearing neurons, which merge into the corticoamygdaloid transition area (CTA). To understand whether immature-appearing neurons are positioned to participate in fear circuitry, we first mapped the hippocampal–amygdala projection in the monkey. We then determined whether immature-appearing neurons were targets of this path. Retrograde results show that the hippocampal inputs to the amygdala originate in uncal region (CA1′) and the rostral prosubiculum, consistent with earlier studies. The amygdalohippocampal area, ventral basal nucleus, the medial paralaminar nucleus, and its confluence with the CTA are the main targets of this projection. Immature neurons are prominent in the PL and CTA, and are overlapped by anterogradely labeled fibers from CA1′, particularly in the medial PL and CTA. Hippocampal inputs to the amygdala are more focused in higher primates compared to rodents, supporting previous anatomic studies and recent data from human functional imaging studies of contextual fear. At the cellular level, a hippocampal interaction with immature neurons in the amygdala suggests a novel substrate for cellular plasticity, with implications for mechanisms underlying contextual learning and emotional memory processes.

Introduction

Across species, the hippocampus and amygdala are closely connected structures that influence emotional perception and experience. The path from the hippocampus to the amygdala has received much attention due to its role in contextual fear learning and memory, based on a wealth of rodent behavioral studies (Kim and Fanselow, 1992, Phillips and LeDoux, 1992, Muller et al., 1997, Rosen et al., 1998, Anagnostaras et al., 1999, Goosens and Maren, 2001, Hall et al., 2001, Anglada-Figueroa and Quirk, 2005, Calandreau et al., 2005, Hernandez-Rabaza et al., 2008, Herry et al., 2008). Co-activation of the hippocampus and amygdala in functional imaging studies support the idea that this circuit is also relevant for understanding contextual fear conditioning in humans (LaBar et al., 1998, Grillon et al., 2006, Alvarez et al., 2008, Alvarez et al., 2011, Marschner et al., 2008). More broadly, the recall of past emotional experiences involves both amygdala and hippocampal activity (Hamann et al., 1999), suggesting that these memories are an internal trigger of emotional states. Delineating the contextual fear circuit in humans is especially important in understanding the pathophysiology of mood and anxiety disorders. In these disorders, environments previously linked to aversive events can subsequently evoke exaggerated or inappropriately persistent fear responses (Phelps, 2004, Indovina et al., 2011). Similarly, in humans, implicit and explicit memories of past events are powerful contributors to emotional and mood states (Lane et al., 1997, Mayberg et al., 1999).

The hippocampus and amygdala in primates are larger, and have a greater cellular complexity, compared to lower species (reviews by Price et al., 1987, DeOlmos, 1990, Rosene and Van Hoesen, 1987). The CA1 region of the hippocampus in primates is greatly expanded at the rostral pole, and is cytoarchitecturally more diverse than in rodents. In particular, the anterior-most portion of CA1 (CA1′) forms a significant portion of the ‘uncus’, a region not found in species lower than the primates (Stephan, 1983). With respect to the amygdala, the most phylogenetically expanded subregion is the basolateral nuclear group (Stephan and Andy, 1977). Compared to the rodent, the primate basolateral nuclei have dramatic cellular gradients (Johnston, 1923, Crosby and Humphrey, 1941, Lauer, 1945, Jimenez-Castellanos, 1949). Another example of discontinuity between species is the paralaminar nucleus (PL), which is the last of the primate amygdala nuclei to form in both human and monkey, arising late in gestational development (Kordower et al., 1992, Ulfig et al., 2003). In the postnatal brain, the PL resides in the remnant of the former lateral ganglionic eminence, merging medially with the corticoamygdaloid transition area (CTA). The PL reaches its largest extent in the human, and has no clear counterpart in the rodent (review by deCampo and Fudge (2012)). Importantly, in both human and nonhuman primates, the region of the PL contains a subpopulation of immature-appearing neurons (Yachnis et al., 2000, Bernier et al., 2002, Fudge, 2004, deCampo and Fudge, 2012), whose function is yet to be defined. These cells contain proteins associated with neural immaturity, including doublecortin (DCX) and polysialylated neural cell adhesion marker (PSA-NCAM) (Francis et al., 1999, Angata et al., 2007).

Earlier anatomical studies in monkeys indicate that the hippocampal–amygdala pathway differs in primates and rodents (Rosene and Van Hoesen, 1977, Ottersen, 1982, Aggleton, 1986, Amaral, 1986, Saunders et al., 1988, Canteras and Swanson, 1992) (reviews by Pitkanen et al., 2000, McDonald, 1998), being far more restricted in higher primates. Given the important role of the hippocampal–amygdala path in contextual fear learning and memory, and its relevance to human anxiety and mood disorders, a closer examination of this projection in the primate is warranted. We sought first to examine the specificity of the hippocampal–amygdala pathway in the monkey using relatively small injections of retrograde tracers in specific subregions of the amygdala. We then analyzed the relationship of the projection relative to the distribution of immature neurons using anterograde tracer injections into the hippocampus. Our hypothesis was that hippocampal inputs would be positioned to influence the immature-appearing neurons in the amygdala, suggesting novel mechanisms for the acquisition and consolidation of contextual fear associations in primates. Since immature neurons in other brain regions respond in unique ways to excitatory signaling (Wang et al., 2000, Schmidt-Hieber et al., 2004, Ge et al., 2007), we were interested in exploring afferent influences on immature neurons in the amygdala.