Early odor learning is important in securing mother-infant bonding in the postnatal environment (Hofer & Sullivan, 2001). Endogenous opioids have been shown to play a prominent role in the postnatal attachment process. Specifically in the rat neonate, opioids facilitate odor preference learning (Barr & Rossi, 1992; Kehoe & Blass, 1986a; Panksepp, Nelson, & Siviy, 1994; Randall, Kraemer, Dose, Carbary, & Bardo, 1992; Roth & Sullivan, 2001,2003; Shide & Blass, 1991), and nipple-milk conditioning (Petrov, Varlinskaya, Becker, & Smotherman, 1998, Petrov, Varlinskaya, & Smotherman, 2000; Robinson, Arnold, Spear, & Smotherman, 1993; Robinson & Smotherman, 1997). Suggestive of their rewarding value in neonates, opioids are sufficient to alleviate separation distress (Carden, Barr, & Hofer, 1991; Goodwin, Molina, & Spear, 1994; Kehoe & Blass, 1986b; Panksepp, Herman, Conner, Bishop, & Scott, 1978). Additionally, Moles, Kieffer, and D’Amota (2004) have recently demonstrated that mice neonates lacking μ-opioid receptors fail to show preferences toward maternal odor and do not show distress when separated from the mother, indicative that opioids play a prominent role in modulating the rewarding experience of mother-infant interactions.

We have previously demonstrated that opioids also play a critical role when neonates learn odor preferences despite pain. Specifically, in rat neonates up to PN9, presentations of odor and shock (0.5 mA) result in learned odor preferences (Camp & Rudy, 1988; Roth & Sullivan, 2001; Sullivan, Landers, Yeaman, & Wilson, 2000). Indicative of the aversive nature of this conditioning paradigm, a shock intensity of 0.5 mA is similar to that commonly used in adult fear conditioning experiments to effectively elicit avoidance (e.g., LaLumiere, Buen, & McGaugh, 2003; Paschall & Davis, 2002; Wilensky, Schafe, & LeDoux, 1999). Additionally, threshold to shock does not appear to change developmentally, and 0.5 mA shock elicits both pain vocalizations (White, Adox, Reddy, & Barfield, 1992) and escape responses (Emerich, Scalzo, Enters, Spear & Spear, 1985; Stehouwer & Campbell, 1978; Sullivan et al., 2000). Thus, neonates display unique learning characteristics about pain. Using this paradigm, we have shown that opioid receptor antagonism during neonatal odor-shock conditioning prevents acquisition of a conditioned odor preference (Roth & Sullivan, 2001). In sharp contrast, opioid receptor antagonism immediately following the odor-shock conditioning produces an odor aversion instead of the typical odor preference (Roth & Sullivan, 2001, 2003). Given these results from our previous studies, one might also expect opioid receptor antagonism post odor-stroke conditioning to elicit an odor aversion. However, results from the present study indicate this is not the case, and rather disruption of the endogenous opioid system following odor-stroke conditioning only blocks consolidation of an odor preference. We suggest that the discrepancy between these studies appears to reflect an important role of endogenous opioids in the neonate’s hedonic response to stimuli. Though our understanding of the neural circuitry mediating infant perception of rewarding and aversive stimuli remains premature, our present behavioral results and those from odor-shock conditioning suggest that neonates can perceive and assign hedonic value to the stimuli, as we are only able to induce a neonate odor aversion when using an aversive conditioning paradigm with opioid receptor manipulation.

Together, results from our previous behavioral studies and the present indicate that opioids play a pivotal role in ensuring learned odor preferences, especially in response to aversive stimuli. Indeed, neonates do not readily learn avoidances or aversions typically produced from passive avoidance, active avoidance, or inhibitory conditioning (reviewed in Roth, Wilson, & Sullivan, 2004). The limitations on aversive learning corresponds to a developmental period when the pups are confined to the nest (Bolles & Woods, 1965) and learned odor aversions would hinder proximity seeking of the mother, and thus the nutrition, warmth, and protection necessary for survival (Hofer & Sullivan, 2001). And the infant learning circuitry appears optimized to support readily learned odor preferences (Sullivan, 2003). Specifically, brain areas activated by odor-stroke, odor-milk, or odor-shock conditioning in rat neonate are the locus coeruleus, olfactory bulb, and piriform cortex (reviewed in Roth et al., 2004; Roth & Sullivan, 2005). Additionally, learning during odor-stroke or odor-shock conditioning in neonates does not appear to involve significant participation of the amygdala (Sullivan & Wilson, 1993; Sullivan et al., 2000; Roth & Sullivan, 2005). An understanding of the role of the endogenous opioid system in this neural circuitry and behavior should provide a better understanding of how the infant brain is optimized for attachment behavior and the mechanisms of mother-infant interactions on infant neural and emotional development.