Currently buprenorphine and methadone are approved for use in the treatment of opioid dependence, but only methadone is approved for use during pregnancy in the United States [1, 2]. Fetal exposure to methadone, however, results in a neonatal abstinence syndrome that can be prolonged and require pharmacological management [3, 4]. Withdrawal signs include irritability, high shrill crying, poor suckling, and inability to sleep [5]. Prenatal methadone exposure also results in long-term consequences in children, including fewer goal-directed behaviors, more speech and cognitive deficits, and poorer social skills [6, 7]. These children also show greater anxiety and aggression and poorer fine and gross motor coordination [8, 9]. However, human studies may be confounded by environmental factors such as socioeconomic status and maternal care, making it difficult to determine direct effects of prenatal opioid dependence [10].

While earlier studies failed to find a relationship between dose of methadone and neonatal withdrawal or health measures, a recent study from New Zealand found that higher doses of methadone during pregnancy were related to increased likelihood of preterm birth, smaller at birth (weight, length, and head circumference), a longer postnatal hospital stay and increased mortality [11]. Also of interest are findings indicating that the reason for methadone use during pregnancy can affect the outcome of the infant. The offspring of women who were receiving methadone for pain management had a significantly reduced neonatal abstinence syndrome compared to infants whose mothers received similar dosing for opioid addiction. They also had higher body weights at birth and larger head circumference. There was no difference in premature labor between the two groups [12]. These data suggest that the other factors beyond drug pharmacology can mediate the effects of developmental exposure. Because of the difficulty in separating environmental factors from pharmacological factors in clinical populations, animal models of prenatal drug exposure have been developed and have been useful in identifying pharmacologically-induced vulnerability factors.

Animal models have also demonstrated long-term dysfunction following prenatal opioid exposure in behavioral, endocrinological, and immunological domains. There is convergent evidence that certain behaviors, such as measures of emotionality, subsequent drug self-administration, operant responding, and place preference are sensitive to developmental opioid exposure [13–17]. One postulated mechanism for these effects is the direct pharmacological action of opioids on brain development, since prenatal opioid exposure affects morphological [18, 19] and neurochemical parameters, including those of the endogenous opioid, dopamine, adrenergic, and cholinergic systems [20–24].

One of the major systems affected by prenatal opioid exposure is the nociceptive system. As early as the 1970s it was noted that prenatal exposure to morphine led to an attenuated antinociceptive response when affected offspring were given morphine prior to hotplate testing at 3, 5, and 11 weeks of age [25]. The attenuated response was similar to that of rats that had been made tolerant to morphine prior to test and the authors speculated that in utero exposure led to the development of tolerance. These results have been replicated by other groups using different morphine exposures [26] as well as other different prenatal opioid exposure models (e.g., levorphanol [27]; methadone [28, 29]). There are some important caveats to note. In some paradigms an enhanced antinociceptive response has been noted. Kirby et al. [30] found that when 20 mg/kg/day morphine was divided across fours doses a day from gestation days 12–20 there was an enhanced response in comparison to saline-exposed offspring, while when the dose was divided into two doses a day they had similar tail-flick latencies as the saline-exposed offspring. The direction and the magnitude of the effects can also be dependent on the age when nociception is assessed. Prenatal methadone exposure enhanced analgesic action of morphine on postnatal day 4, but attenuated it on postnatal day 21 [28]. Prenatal morphine attenuated the antinociceptive response to morphine on postnatal day 14 [26].

Reward properties of drugs in adulthood are also affected. An early study by Glick et al. [31] found that prenatal morphine decreased the number of days needed to acquire morphine self-administration as an adult. Similarly prenatal methadone increased the amount of morphine consumed in exposed adult rats in the two bottle choice test [32]. Prenatal exposure to morphine increased rates of heroin and cocaine self-administration in adulthood, increasing the sensitivity to lower doses [33]. Because both cocaine and heroin self-administration were enhanced, it is unlikely due to solely to changes in opioid receptor density or affinity.

One of the most robust findings in the prenatal opioid literature is the effect on the cognition, specifically the acquisition and retention of new information. Spatial learning models have been extensively examined. We have found that exposure throughout gestation in rats to the long-acting opiate, l-α-acetylmethadol (LAAM) resulted in poor performance in acquisition of the radial arm maze [34]. Prenatal LAAM-exposed rats had more reference and working memory errors, but were able to acquire the task after 5 days of training and there was no difference from prenatal water controls in retention trials 24 h later. Similar effects were found in the Morris water maze for rats exposed to oxycodone in utero, where differences in the search strategy utilized by the prenatally exposed rats were noted [35]. Radial arm maze [36] and Morris water maze [37] deficits were also observed after exposure to prenatal heroin in rodents. Morphine administered to rats in embryonic days 11–18 increased latency in the radial arm maze [38]. Exposure to morphine on embryonic days 12–16 caused a deficit in long-, but not intermediate-term memory in the one-trial passive avoidance task paradigm in the chick [39].

Prenatal opiate exposure has also been known to result in cellular and molecular alterations related to learning and memory. For example, prenatal exposure to heroin caused pre and postsynaptic alterations in the septo-hippocampal cholinergic system [40]. These changes, which include choline transporter activity, G-protein levels, as well as basal and carbachol-stimulated PKC activity, are postulated to play a role in changeS in long-term potentiation (LTP), widely considered as a cellular mechanism for learning and memory [41]. Prenatal morphine caused impairment in Morris water maze performance that was concomitant with alterations in hippocampus LTP and LTD (long-term depression), as well as NMDA receptor-mediated plasticity and phosphorylation of CREBSerine-133 [42]. In an in vivo study, prenatal morphine attenuated the maintenance phase of LTP in the lateral perforant path in the hippocampus, while not altering induction or intermediate LTP in the lateral or medial perforant paths [43].

The consequences of opioid exposure are not limited to the nervous system. The immunosuppressive properties of opioids have been well documented and will not be detailed here [44, 45]. However, with respect to opioid exposure during development, Shavit et al. [46] found that prenatal morphine blunted natural killer cell activity and the fever response to the endotoxin lipopolysaccharide (LPS) in adult rats. We found a blunted fever response to LPS in hatchlings exposed in ovo to the l-alpha acteylmethadol (LAAM) metabolite NLAAM [47]. Similarly, we found a blunted fever response to LPS in adult rats that had been prenatally exposed to LAAM. There was no change in basal body temperature or in response to saline injection. There were indications that the neural-immune network was altered because levels of the mature form of the IL-1b protein were elevated in the hypothalamus of prenatally LAAM-treated rats. Interestingly, circulating levels of IL-1b were not affected, nor were protein levels in the spleen [48].

There are multiple mechanisms by which opioids exert their effects on the developing nervous system. Not surprisingly, opioid receptor signaling has been indicated, specifically mu opioid receptors are involved [49]. Darmani et al. [50] found that prenatal methadone altered mu receptor affinity in both gestation day 7 and on postnatal day 7. Scatchard analyses revealed that the receptor density was unchanged, but the affinity for the mu selective ligand DAMGO was decreased (Darmani et al., 1992).

Studies examining prenatal morphine exposure have demonstrated involvement of NMDA receptors. For example, Tao et al. [26] found that the NMDA antagonist dextromethorphan when co-administered with morphine throughout pregnancy and for the first postnatal month reduced neonatal mortality, normalized body weight, attenuated postnatal withdrawal, and prevented the down-regulation of hippocampal NMDA receptors that were associated with the developmental morphine exposure. Interestingly, the role of glutamatergic neurotransmission in the effects of opioid exposure may vary on the specific receptor subtype and the developmental stage. NMDA receptors antagonists have low efficacy in blocking withdrawal effects during the first postnatal week, but high efficacy in the second and third postnatal weeks. In contrast, antagonists to the AMPA and metabotropic glutamate receptors are equally effective throughout the postnatal—preweaning period [51, 52]. Alterations in the cholinergic system have been suggested to underlie some of the cognitive deficits associated with prenatal opioid exposure. Vatury et al. [53] found that choline transporter sites were distributed in a different pattern in the CA1, CA3, and dentate gyrus regions in the hippocampus of mice exposed prenatally to heroin than in vehicle controls.

Studies using various pharmacological manipulations in animals have suggested that neonatal opioid withdrawal may be involved in some of the acute and long-term effects of prenatal opioid exposure [54–56]. This may be a consequence of the neurochemical changes associated with opioid withdrawal and/or some of the metabolic and neuroendocrine manifestations. Neonatal opioid withdrawal is mediated by multiple neurochemical systems, with the serotonin₂ receptors and alpha₂ adrenergic receptors playing important roles [57–61]. Studies have also implicated a role for nitric oxide [51]. Other potential mediators are changes in metabolic demands and respiration, as well as activation of the hypothalamic–pituitary–adrenal (HPA) axis that can occur as the neonate undergoes withdrawal [62]. In the chick embryo, we found that precipitated withdrawal from chronic opioid exposure activated the HPA axis, increasing corticosterone concentrations [61]. In the absence of drug exposure itself, these physiological and metabolic changes associated with opioid withdrawal on their own could affect development of multiple organ systems and impact long-term health and behavior.

Animal models of prenatal opioid exposure have corroborated some of the findings of the human clinical literature, in particular, alterations in cognitive functioning and processing of nociceptive signals. Mechanisms for these effects include dysregulation of neurotransmitter systems and molecular signaling pathways. These findings suggest therapeutic targets via both pharmacological as well as environmental regimens.