Deficit in Attachment Behavior in Mice Lacking the µ-Opioid Receptor Gene

See allHide authors and affiliations

Science  25 Jun 2004:
Vol. 304, Issue 5679, pp. 1983-1986
DOI: 10.1126/science.1095943


Endogenous opioid binding to μ receptors is hypothesized to mediate natural rewards and has been proposed to be the basis of infant attachment behavior. Here, we report that μ-opioid receptor knockout mouse pups emit fewer ultrasonic vocalizations when removed from their mothers but not when exposed to cold or male mice odors. Moreover these knockout pups do not show a preference toward their mothers' cues and do not show ultrasonic calls potentiation after brief maternal exposure. Results from this study may indicate a molecular mechanism for diseases characterized by deficits in attachment behavior, such as autism or reactive attachment disorder.

The opioid system controls nociceptive and addictive behaviors, with a prominent role of μ-opioid receptors (Orpm) in these responses. Mice lacking the μ receptor gene (Orpm/) show a loss of morphine-induced analgesia, reward, and dependence (1); increased sensitivity to painful stimuli (2, 3); reduced reward to nonopioid drugs of abuse (4); and altered emotional responses (5). Endogenous opioid binding to μ receptors has been considered one of the neural mediators of infant attachment behavior (610), although the role of μ receptors in mediating the rewarding properties of mother-related stimuli has not been determined (11, 12).

Attachment behavior entails the display of affiliative behaviors and the establishment of a special bond with the animal's caregivers (13). It is characterized by the selective approach to, and interaction with, specific individuals and by the display of emotional distress during acute periods of separation from these individuals. This behavioral system has been suggested to represent the developmental roots of the affiliative system emerging during the adult life and to modulate behaviors such as social play and sexual and parental behavior (12).

Scott (14) was the first to use the distress vocalizations as a model to study attachment behavior in puppies. In mouse pups, the response to separation from the mother consists of a protest phase characterized by a high rate of ultrasonic vocalization (UV) and hyperactivity, which facilitates reunion with her. UVs have been used as an index of distress in several experimental paradigms (15) and are strongly reduced in pups given μ-opioid agonist agents (1618). However, the inconsistency in the results obtained with antagonist administration (19, 20) has questioned the role of μ-opioid receptors in the pup's vocal response to social separation.

Orpm–/– pups emitted fewer calls than their wild-type controls (Orpm+/+) in response to isolation at various developmental time points (21) (Fig. 1). Moreover, morphine administration significantly reduced UVs in 8-day-old wild-type pups, whereas this drug had no effect on Orpm–/– calls (21). The lower UV performance of Orpm–/– pups was not associated with evident differences in maternal responsiveness toward the offspring. In fact, mothers of the two genotypes did not differ in their behavior toward pups in response to either nest relocation (Table 1) or to a test already validated for maternal emotionality and motivation (Table 2) (21). The absence of detectable differences in maternal responsiveness toward pups suggests that the fewer distress calls emitted by Orpm/– pups are the result of a reduced sensitivity to isolation. We thus performed a second experiment to measure the UV response of isolated pups to stressful physical (cold) and social (clean and novel male bedding compared to nest odor bedding) stimuli (Fig. 2). Wild-type pups responded to isolation and exposure to clean bedding with a high frequency of ultrasonic calls; this effect was significantly reduced in the presence of familiar bedding. In contrast, isolated Orpm/ pups showed very low amounts of UVs when isolated and exposed to clean bedding and no further reduction in the amount of calls emitted in the presence of familiar bedding. In fact the clean bedding condition was not sufficient to elicit the distress behavioral response in Orpm/– pups. However, pups of both genotypes showed a similar UV response when exposed to stressful stimuli such as cold or strange male odors, ruling out the possibility that the mutant animals have a general impairment in their capacity to emit high levels of ultrasonic calls or in their sensitivity to olfactory and thermal stimuli.

Fig. 1.

Ultrasonic calls in response to maternal separation. Mean number of ultrasonic calls of Orpm/ (n = 17) and Orpm+/+ (n = 16) isolated pups tested during a 5-min session in an empty beaker at room temperature. Pups' ultrasounds were measured on postnatal days (PNDs) 4, 8, and 12. Four litters per genotype were tested, and litter size ranged between three and six in both lines. The two-way ANOVA for repeated measures indicated significant main effects [for genotype, F(1/62) = 5.08 and P < 0.05; for day, F(2/62) = 8.34 and P < 0.001] and no significant interaction effect.

Fig. 2.

Ultrasonic calls in response to maternal versus other cues. Mean number of ultrasonic calls of PND8 pups isolated and exposed for 5 min to either different social stimuli (clean, n = 10 Orpm/ and n = 19 Orpm+/+; mother/nest, n = 10 Orpm/ and n = 24 Orpm+/+; or male, n = 13 Orpm/ and n = 16 Orpm+/+) or cold, n = 12 Orpm/ and n = 16 Orpm+/+. The two-way ANOVA indicated significant effects of the treatment [F(3/112) = 17.88 and P < 0.001] and of the genotype × treatment. [F(3/112) = 5.22 and P < 0.01]. Asterisks indicate P < 0.05 between pups of the two genotypes in the same treatment; open circles, P < 0.05 versus pups of the same genotype exposed to different treatments (Tukey HSD).

Table 1.

Comparison of mothers' responsiveness toward their pups (PND4) in the two homozygotic lines, in standard condition, after nest relocation in a more complex environment (mean s ± SE). No statistically significant differences were found in the behavior of mothers of the two genotypes.

With pups Exploring the cage Exploring the T maze
Orpm+/+ (n = 6) 180±25 363±54 394±79
Orpm-/- (n = 5) 138±46 312±40 461±42
Table 2.

Comparison of mothers' responsiveness toward their pups in the two homozygotic lines in a behavioral test validated for mother's emotional and motivational response (29). Mother's behavior in the three-compartment cage measured on PND8. No statistically significant differences were found in the behavior of mothers of the two genotypes.

Latency to enter male's compartment (mean s ± SE) Time to reach pups (mean s ± SE) Transitions
Orpm+/+ (n = 8) 95.2±23.1 257.7±37.5 5.0±1.1
Orpm-/- (n = 8) 100.5±17.2 295.9±23.8 4.0±0.7

These results suggest that the absence of mother and nest odor does not elicit a distress vocal response in Orpm/– pups. Opioids have a facilitatory role in odor preference acquisition and nipple-milk conditioning (22). Orpm/– pups may thus lack the normal activation of the endogenous opioid system, mediating the rewarding values of maternal stimuli that facilitate infant behavior toward the mother. This hypothesis was further investigated by assessing the maternal potentiation of UV (23). Maternal potentiation of UVs is suggested to have its basis in a learning process whereby pups associate their calling with positive maternal-related reinforcements. The removal of the mother, after a brief exposure, causes a violation of reward expectancy and an increase in ultrasonic calls. Wild-type pups showed the expected increase in UV after a short exposure to the mother, whereas mutant pups did not show a maternal potentiation effect (Fig. 3). The absence of maternal potentiation of UV in Orpm–/– pups supports the role of μ receptors in mediation of the positive affective state evoked by maternal cues.

Fig. 3.

Maternal potentiation of ultrasonic calls. Effects of mother's presence and mother's removal on ultrasonic calls relative to pretest in 12-day-old Orpm+/+ (n = 8) and Orpm–/– (n = 14) pups. The two-way ANOVA for repeated measures indicated only a significant genotype per repeated measures effects [F(1/20) = 4.49 and P < 0,05]. Asterisk indicates P < 0.05 (Tukey HSD) versus Orpm+/+ animals.

Pup attachment to the mother, providing warmth, milk, and protection, has its basis primarily in olfactory cues. Therefore, we evaluated the olfactory acuity and preference of Orpm–/– and Orpm+/+ pups. We found no difference between the abilities of 8-day-old pups of the two genotypes to discriminate between their mother and nest and clean bedding. In contrast, when the pups had to choose between their own and an unfamiliar mother and nest bedding, only 36% of Orpm–/– compared with 100% of Orpm+/+ pups chose the familiar bedding. In addition, the percentage of time spent in the familiar bedding was lower in the mutant pups (Fig. 4).

Fig. 4.

Responses to familiar versus unfamiliar odors. Percentage of time spent by pups with different genotype in their own mother and nest versus the clean bedding compartment and versus the compartment covered with nest odor of an unknown mother with pups of similar age. Asterisk indicates P < 0.001 (Student t test, t24 = 4.46) versus Orpm+/+ animals. Numbers at the bottom of the bars represent the number of pups choosing that compartment over the total number of pups tested. The two lines of pups did not differ when choosing between own and clean bedding (Fisher exact probability test, P = 0.15) but differed when choosing between their own and an unknown dam's bedding (Fisher exact probability test, P = 0.001).

These results indicate that mice lacking μ-opioid receptors show deficits in two independent measures of attachment behaviors. Orpm–/– pups were not able to selectively approach their mothers, and maternal deprivation was not sufficient to elicit distress calls. Infant vocal behavior has been suggested to reflect anxiety in pups (24). However, lower fearfulness during maternal separation alone cannot account for the deficit in attachment behavior displayed by these knockout pups. Pups of both genotypes showed the same amount of UV when exposed to cold or male odor cues. Most interestingly, Orpm–/– pups seemed to be specifically less sensitive to the absence of maternal cues. This could be because of the absence of association between maternal stimuli and pleasurable states mediated by the μ-opioid system.

The mesocorticolimbic dopamine reward system has been implicated in social attachment whereby oxytocin and vasopressin peptides play a critical role in the association between the social stimuli and the brain reward circuits (25). Our data support the hypothesis that the μ-opioid system is a crucial component of the social reinforcement pathways modulating the positive affective states associated with mother stimuli. Panksepp and colleagues (7, 16, 26) elaborated a model in which the social contact provides comforting effects, in part opioid-mediated, and proposed that social separation determines a distress analogous to opiate withdrawal. Orpm/ pups would thus not experience either state because of the absence of the receptors mediating the response to social isolation or social comfort. Pharmacological activation of μ-opioid receptors also decreased UVs (1618), suggesting that exogenous opioids mimic the positive state associated with mother's presence, thus reducing the UV response to isolation. We hypothesize that the deficit in the μ-opioids system of our knockout mice annihilate the natural association between reward and maternal stimuli, making these animals less sensitive to maternal separation.

Alterations of the attachment system have been considered as etiological factors for several psychiatric syndromes. Our data show that μ receptors are critical players in attachment disorders, supporting the classical hypothesis of Panksepp and co-workers (27, 28) that a malfunctioning of the endogenous opioids system may be implicated in the social indifference displayed by autistic infants. The present data also highlight mice lacking μ-opioid receptors as a useful animal model to evaluate the consequences of deficits in the affiliative system during development and adulthood.

Supporting Online Material

Materials and Methods

SOM Text

References and Notes

Stay Connected to Science

Navigate This Article