Mechanisms underlying the early risk to develop anxiety and depression: A translational approach

Abstract

Anxious temperament (AT) is an early life disposition that markedly increases the risk to develop stress related psychopathology such as anxiety and depressive disorders. Since anxiety and depression are common, and frequently have their onset early in life, a better understanding of the factors related to their childhood onset will facilitate the development of new more effective neurally informed interventions. A nonhuman primate (NHP) developmental model of childhood AT has been established, which has provided an understanding of the neural systems and molecular mechanisms mediating the development of AT. Multimodal neuroimaging studies reveal altered brain metabolism across prefrontal, limbic (e.g. central nucleus of the amygdala (Ce) and anterior hippocampus), and brainstem regions, as well as altered functional connectivity involving the Ce. Heritability studies demonstrate that individual variation in AT is heritable, and genetic correlational analyses demonstrate that metabolism in the posterior orbital frontal cortex, the bed nucleus of the stria terminalis, and the periaqueductal gray share a genetic substrate with AT. On a molecular level, the finding of reduced expression of Ce neuroplasticity genes provides the basis for a neurodevelopmental hypothesis focused on the Ce. Viral vector methods for altering gene expression in the Ce of young NHPs are currently being used as a prelude to conceptualizing novel molecularly targeted early life interventions.

Figure 1

(A) The three experimental conditions of the human intruder paradigm elicit distinct fear and anxiety-related behaviors in young rhesus monkeys. When alone and separated from their cagemate (Alone condition, left), young monkeys actively explore the test cage and emit “coo” calls, thought to reflect an attempt to attract help from their mothers or other conspecifics. In the next condition, a human intruder presents his or her profile, while avoiding direct eye contact with the monkey (No Eye Contact condition, center). In this situation, the monkeys typically orient their focus to the intruder, trying to evade discovery by remaining completely still (freezing) and reducing their coo volcalizations. In the third condition, the human intruder enters the room and stares at the animal (Stare condition, right). This direct threat condition elicits aggressive and submissive behaviors (e.g., barking, threatening gestures, lip smacking, and cage rattling). From Kalin (1993). Copyright 2002 by Scientific American, Inc. Reprinted by permission. (B) Anxious Temperament (AT) is calculated as the mean z-scores of NEC-induced freezing, coo vocalizations, and plasma cortisol levels. (C) AT is a relatively stable trait. In this example, taken from Fox et al. (2008), AT assessed at two time points with an approximate 4-month interval was significantly correlated. Reprinted with permission.

Figure 2

(A) During uncertain anticipation, relative to the certain anticipation of fear faces viewed inside the MRI scanner, children with anxiety disorders exhibited greater activation in the left amygdala (red cluster). The bar graph displays the percent signal change extracted from the cluster, and shows that the group difference is driven by a higher amygdala response to the uncertain cue in children with anxiety. Error bars represent SEM. Region of interest (face-responsive voxels within the amygdala) outlined in green. Modified from Williams et al. (2015), with permission. (B) Regions where brain metabolism was significantly associated with individual differences in rhesus monkey AT (p<.05, Šidák corrected for multiple comparisons across the whole-brain). Regions include ortibal proisocortex/anterior insula (OPro/AI; shown in [a]), subgenual anterior cingulate, temporal cortex, bed nucleus of the stria terminalis (BST; shown in [b]), central nucleus of the amygdala (Ce; shown in [c]), anterior hippocampus (aHip; shown in [d]) and brainstem regions including the periaqueductal gray (PAG; shown in [e]). Reprinted with permission from Fox et al. (2015). (C) The effects of central nucleus (Ce) lesions on components of AT. Monkeys with Ce lesions displayed less freezing (top), emitted more coo calls (middle), and released less adrenocorticotropic hormone (ACTH, bottom) during exposure to the human intruder paradigm. Adapted by permission from Kalin, Shelton, and Davidson (2004).

Figure 3

A tripartite prefrontal-limbic-midbrain circuit involved in the genetic transmission of AT. (A) Regions where brain metabolism demonstrated a significant genetic correlation with AT, thus sharing a genetic substrate, include: the ortibal proisocortex/anterior insula (OPro/AI, red), bed nucleus of the stria terminalis (BST, orange), and the periaqueductal gray, (PAG, peach). (B-D) Using high-resolution anatomical and chemoarchitectonic imaging in a separate group of monkeys, regions were precisely localized as: agranular OPro/AI (B); the BST region lying between the anterior commissure (ac) and the [18-F] fallypride identified dopamine receptor-rich ventral striatum that includes the nucleus accumbens (green; NAcc) (C); and the vlPAG-region in the gray matter surrounding the ventricle, superior to the [11-C] DASB identified serotonin transporter-rich dorsal raphe nucleus, DRN (D). Modified from Fox et al. (2015), and reproduced with permission.

Figure 4

Ce expression of mRNA for the neurotrophic receptor, NTRK3, negatively predicts AT. (A) Microarray data showed that individuals with higher levels of Ce NTRK3 mRNA expression exhibited lower AT. Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) confirmed the negative relationship between Ce NTRK3 mRNA expression levels and AT (r = - 0.49; P = 0.029). (B) Individuals showing higher levels of NTRK3 mRNA expression, indexed by qRT-PCR, show reduced amygdala metabolism in vivo (green) [FDR-corrected within the stable AT-related region (outlined in red)]. (C) Schematic of the neuroplasticity-associated NTRK3 (tropomyosin receptor kinase [Trk]) pathway. A similar pattern in relation to AT was found for IRS2, an intracellular kinase signaling molecule (orange) and RPS6KA3 (pink), two downstream mediators of NTRK3 activation. Other molecules in the NTRK3 pathway are also depicted in light gray. Adapted with permission from Fox et al. (2012) and from Fox & Kalin (2014).

Figure 5

In vivo estimation and postmortem verification of dorsal amygdala corticotropin-releasing factor (CRF) overexpression. (A) The gadolinium cloud in the dorsal amygdala, central nucleus (Ce) region, during and immediately following adeno-associated virus type 2 (AAV2)-CRF delivery provided an estimate of the location and extent of the infusions. (B) Camera lucida drawings of CRF expression from postmortem tissue reflected the extent of viral infusion as estimated from the intraoperative gadolinium signal. Gray regions represent neuropil staining and the black dots represent CRF overexpressing cells. (C) Acetylcholinesterase staining defined the boundaries of the amygdalar nuclei. (D) Adjacent sections were used for CRF immunohistochemistry demonstrating marked overexpression in the dorsal amygdala, Ce region. (E) Based on the intraoperative gadolinium images, we estimated the infusion extent in standard space to examine the overlap of the gadolinium injection clouds across the five experimental animals. The colors represent the number of animals with gadolinium signal at each voxel. Note the bilateral overlap across all experimental animals within the Ce region (yellow). (F) Compared with their matched control animals, the CRF overexpressing animals demonstrated increased postsurgical levels of AT (mean +/- SEM). Significance was determined using a paired-samples t-test comparing dorsal amygdala CRF animals and their cagemate control animals (CRF group [post-pre] – control group [post-pre]) (p<, .05, one-tailed; see inset and Kalin et al., (2016) for details). (G) Compared with their matched control animals, the CRF overexpressing animals demonstrated increased [post-pre] change in metabolism within the dorsal amygdala, OPro/AI, and hippocampus (yellow, p <, .01, two-tailed, uncorrected). ABmc, accessory basal nucleus, magnocellular subdivision; Astr, amygdalostriatal transition zone; Bmc, basal nucleus, magnocellular subdivision; CeLpc, central nucleus, lateral central subdivision; CeM, central nucleus, medial subdivision; L, lateral nucleus; R, right.