Neurobiology of attachment to an abusive caregiver: short-term benefits and long-term costs

Abstract

Childhood maltreatment is associated with adverse brain development and later life psychiatric disorders, with maltreatment from the caregiver inducing a particular vulnerability to later life psychopathologies. Here we review two complementary rodent models of early life abuse, which are used to examine the infant response to trauma within attachment and the developmental trajectories that lead to later life neurobehavioral deficits. These rodent models include being reared with an abusive mother, and a more controlled attachment-learning paradigm using odor-shock conditioning to produce a new maternal odor. In both of these rodent models, pups learn a strong attachment and preference to the maternal odor. However, both models produce similar enduring neurobehavioral deficits, which emerge with maturation. Importantly, cues associated with our models of abuse serve as paradoxical safety signals, by normalizing enduring neurobehavioral deficits following abuse. Here we review these models and explore implications for human interventions for early life maltreatment.

Keywords: abuse; amygdala; fear conditioning; later life outcome; maternal odor; olfaction; predator odor; pup; rat; safety signal; stress; threat.

FIGURE 1

Schematic of short- and long-term effects following infant abuse. Infant abuse during the sensitive period (until PN10), using our odor-shock conditioning or abusive mother paradigms of abuse, produces converging neurobehavioral deficits across development. Despite abuse, infants form a strong attachment to their caregiver, which is mediated by the olfactory bulb (OB), anterior piriform cortex (PCTX), and locus coeruleus (LC). However, in infants, corticosterone (CORT) uncovers the effects of early life abuse. Heightened CORT in infancy prematurely engages the amygdala, and produces disrupted interactions with the mother in abused pups. By adulthood, the long-term effects of infant abuse on amygdala-dependent behaviors include decreased social behavior, increased depressive-like behaviors, as well as increased approach to predator odor, and poor maternal protection in a threatening situation.

FIGURE 2

All animals, regardless of rearing conditions, display fear to predator odor. However, as illustrated by these data, infant abuse alters the adult response to threat by decreasing passive defensive behaviors and increasing active defensive behaviors in adults. (A) This figure illustrates the Threat Response Selection Test (TRST). This test permits the assay of both active and passive defensive responses. In this test, individuals are placed in a center “start” chamber and given the option to (i) freeze, (ii) approach a predator odor (fox urine), or (iii) hide in a provided shelter. The rat is allowed free roam of the apparatus for 5 min, and active and passive fear responses are scored. Each animal is habituated to the apparatus without predator odor on the day before testing, so they are familiar with the apparatus and the location of the shelter. In the TRST, early life abused animals (B) spend significantly less time freezing, relative to normally reared controls [unpaired t-test, t(12) = 2.383, p < .05, n = 7], (C) spend less time hiding in the shelter [unpaired t-test, t(12) = 2.660, p < .05, n = 7], (D) spend significantly more time rearing (standing on hindlimbs) [unpaired t-test, t(12) = 3.417, p < .01, n = 7], (E) spend more time approaching the section of the apparatus that contains predator odor [unpaired t-test, t(12) = 2.339, p < .05, n = 7], and (F) spend significantly more time in direct contact with the predator odor [unpaired t-test, t(12) = 2.225, p < .05, n = 7]. Thus, these data illustrate that early life abused animals spend significantly less time engaged in passive fear behaviors than typically reared control animals (as indexed by time spent freezing and hiding in the provided shelter), and significantly more time engaged in active fear behaviors than typically reared control animals (as indexed by time spent rearing, approaching the predator odor, and contacting the predator odor). *Significantly different from all other groups, p < .05, unpaired t-test. Error bars represent SEM.

FIGURE 3

All mothers, regardless of rearing condition, display fear to predator odor. However, as illustrated in the TRST, mothers abused in infancy show a heightened threat response and display poor protection of their pups in a threatening situation relative to controls. Mothers that were abused in infancy: (A) exhibit significantly less time freezing [unpaired t-test, t(16) = 2.959, p < .01, n = 9], (B) spend less time hiding in the shelter [unpaired t-test, t(16) = 2.333, p < .05, n = 9], (C) spend significantly more time in the section of the arena that contains predator odor [unpaired t-test, t(16) = 2.342, p < .05, n = 9], and (D) fail to protect their pups in response to predator odor [unpaired t-test, t(16) = 2.546, p <.05, n = 9]. Poor pup protection is determined by total time spent scattering their pups, transporting their pups, and placing their pups close to the predator odor. *Significantly different from all other groups, p < .05. Error bars represent SEM.