Developing a neurobehavioral animal model of infant attachment to an abusive caregiver

Abstract

Background: Both abused and well cared for infants show attachment to their caregivers, although the quality of that attachment differs. Moreover, the infant's attachment to the abusive caregiver is associated with compromised mental health, especially under stress. In an attempt to better understand how abuse by the caregiver can compromise mental health, we explore the neural basis of attachment in both typical and abusive environments using infant rats, which form attachments to the mother through learning her odor. Here, we hypothesize that the neural circuitry for infant attachment differs based on the quality of the attachment, which can be uncovered during stressful situations.

Methods: We used infant rats to compare infant attachment social behaviors and supporting neurobiology using natural maternal odor, as well as two odor-learning attachment paradigms: odor-stroke (mimics typical attachment) and odor-.5 mA shock conditioning (mimics abusive attachment). Next, to uncover differences in behavior and brain, these pups were injected with systemic corticosterone. Finally, pups were reared with an abusive mother to determine ecological relevance.

Results: Our results suggest that the natural and learned attachment odors indistinguishably control social behavior in infancy (approach to the odor and interactions with the mother). However, with corticosterone injection, pups with an abusive attachment show disrupted infant social behavior with the mother and engagement of the amygdala.

Conclusions: This animal model of attachment accommodates both abusive and typical attachment and suggests that pups' social behavior and underlying neural circuitry may provide clues to understanding attachment in children with various conditions of care.

Figure 1

During the sensitive period (PN8): (A) pups approach (Y-maze) the natural maternal odor or learned attachment odor produced by pairing a novel odor with either stroking or .5 mA shock. (B–D) The natural maternal odor, as well as the odor previously paired with either shock or stroking supported interactions with the mother. Specifically, if the natural maternal odor is removed, pups show little interactions with the mother (B), such as the age-specific behavior of probing (pushing nose into mother’s fur) and nipple attachment (C–D). However, an airstream of either maternal odor or the odor previously paired with stroke or shock (B) enhances interactions with the mother and also (C–D) reinstates nipple attachment. Unpaired and odor-only presentations do not support approach response, mother interactions, or nipple attachment, indicating the importance of learning in producing the maternal odor. *p < .05 between groups (n = 5–6 for all groups). PN, postnatal day.

Figure 2

During sensitive period (PN8), (A) natural maternal odor and odors previously paired with shockor stroke increased c-Fos expression in the glomerular layer of the olfactory bulb. Systemic CORT injection (3.0 mg/kg) 30 minutes before the odor presentation did not change c-Fos expression. Bars represent the number (mean ± SEM) of c-Fos-positive cells counted bilaterally in the glomerular layer of the olfactory bulb. *p < .05 between groups (n= 5–6 for all groups). (B) Schematic representation of the division (midlateral) of the olfactory bulb analyzed. CORT, corticosterone; PN, postnatal day.

Figure 3

During sensitive period (PN8), systemic CORT injection (3.0 mg/kg) 30 minutes before the odor presentation increased significantly c-Fos only in the paired odor-shock group. Bars represent the number (mean ± SEM) of c-Fos-positive cells counted bilaterally in each amygdala nuclei: (A) cortical, (B) medial, (C) basolateral complex, (D) central. *p< .05 between groups (n= 4–7 for all groups). (E) Schematic representation of amygdala nuclei analyzed. BLA, basolateral complex; ceA, central; coA, cortical; CORT, corticosterone; meA, medial; PN, postnatal day.

Figure 4

During the sensitive period (PN8), a systemic CORT injection (3.0 mg/kg) 30 minutes before Y-maze test (A) switches the normal odor approach (preference) to odor avoidance only in paired odor-shock pups. CORT injection also increased the time spent probing (B), increased the latency to nipple attach (C), and decreased the time spent nipple attached (D) in paired odor-shock pups. CORT injections had no effects on paired odor-stroke, maternal odor, and odor-only groups. *p < .05 between groups (n= 5 for all groups). CORT, corticosterone; PN, postnatal day.

Figure 5

Stress-mother reared pups (from PN3 to PN8) present a decreased approach to maternal odor in the Y-maze test (A), increased the time probing the mother’s ventrum during nipple searching (B), increased the latency to nipple attach (C), and decreased the time nipple attached (D) in the mother-pup interaction test. *p < .05 between groups (n= 6 for all groups). PN, postnatal day.

Figure 6

Stress-mother reared pups (from PN3 to PN8) do not change the expression of c-Fos in the olfactory bulb (A) after maternal odor presentation. c-Fos expression is significantly increased in amygdala nuclei analyzed in stress-mother reared pups. Bars represent the number (mean ± SEM) of c-Fos-positive cells counted bilaterally in each olfactory bulb and amygdala nuclei: (B) cortical, (C) medial, (D) basolateral complex, (E) central. *p < .05 between groups (n= 5 for all groups). PN, postnatal day.