Early exposure to nicotine during critical periods of brain development: Mechanisms and consequences

Abstract

Tobacco use during pregnancy continues to be a major problem with more than 16% of pregnant women in the United States continuing to smoke during pregnancy. Tobacco smoke is known to contain more than 4,000 different chemicals, and while many of these compounds have the potential to interfere with proper neurodevelopment, there is direct evidence that nicotine, the major psychoactive substance present in tobacco, acts as a neuroteratogen. Nicotine activates, and subsequently desensitizes, neuronal nicotinic acetylcholine receptor subtypes (AChRs), which are expressed in the developing central nervous system (CNS) prior to the in-growth of cholinergic neurons. Nicotinic AChRs are present by the first trimester of development in both humans and rodents, and activation of these receptors by acetylcholine is thought to play a critical role in CNS development. The purpose of the current review is to provide an overview of the role that nicotinic AChRs play in the developing CNS and to describe the effects of nicotine exposure during early development on neuronal cell biology, nicotinic AChR expression and neurotransmitter system (e.g., dopamine, norepinephrine, serotonin) function. In particular, differences that occur as a result of the timing and duration of nicotine exposure will be discussed. Emphasis will be placed on preclinical studies examining particular periods of time which correspond to periods of prenatal development in humans (i.e., first, second and third trimesters). Finally, the effects of early nicotine exposure on neurobehavioral development as it pertains to specific disorders, i.e., attention deficit hyperactivity disorder (ADHD), depression and addiction, will be discussed.

1. Introduction

Despite the best efforts of anti-smoking campaigns and legislation, tobacco use continues to be a major problem (1). Worldwide, more than 1.2 billion individuals are current smokers, and approximately 6 million tobacco-related deaths occur each year (2,3), a number that is estimated to increase to more than 8 million by the year 2030 (4). While smoking has deleterious effects in all populations, there is particular concern in regards to pregnant women, since they expose both themselves and the fetuses they carry to the numerous environmental toxins present in tobacco smoke. According to the 2008 National Survey on Drug Use and Health, 16.4% of pregnant women between the ages of 15 and 44 smoke tobacco (5). In comparison, only 10.6% of pregnant women use alcohol (5). Furthermore, while the overall number of women smokers in the United States has declined, the percentage of pregnant smokers remained steady between 1995 and 2005 (6), indicating that attempts to curb smoking in pregnant women have been unsuccessful. Women who stop smoking during pregnancy have a high rate of relapse postpartum (6), which results in continued exposure to tobacco smoke constituents to infants, both through environmental smoke exposure and breast-feeding. According to the Centers for Disease Control, women who smoke during pregnancy have an increased likelihood of premature rupture of placental membranes, placental abruption, and placenta previa during pregnancy (7). Infants exposed to tobacco smoke during pregnancy are 30% more likely to be born premature or with low birth weights, and are up to 3 times more likely to die of Sudden Infant Death Syndrome (SIDS). Further, a recent study found that pregnant women exposed to environmental tobacco smoke (i.e."secondhand smoke”) also have substantial urine cotinine concentrations (8). Thus, in addition to direct inhalation of tobacco smoke, passive inhalation is sufficient to produce concentrations of nicotine that may be harmful to the fetus (9). Most importantly, in utero exposure to constituents in tobacco smoke results in long-term and profound alterations in brain development that last into adulthood (10).

Tobacco smoke is known to contain more than 4,000 different chemicals, including various carcinogens and environmental toxins (11,12). While many of these compounds have the potential to interfere with proper neurodevelopment, there is direct evidence that nicotine, the major psychoactive substance present in tobacco, is a “neuroteratogen” (13). Nicotine activates and subsequently desensitizes nicotinic acetylcholine receptors (AChRs) (14). Neuronal nicotinic AChRs are expressed in the developing central nervous system (CNS) prior to innervation by cholinergic neurons (15–18), and the cholinergic neurotransmitter system is thought to play a critical role in CNS development (19). Nicotinic AChRs are present during the first trimester in both humans and rodents (19) and have transient patterns of expression throughout the course of early brain development (19–22). Chronic nicotine treatment in utero produces alterations in neuronal cytoarchitecture, nicotinic AChR expression and the function of various neurotransmitter systems (23–27). Activation of nicotinic AChRs by nicotine may prematurely induce developmental events, such as cholinergic-mediated signaling that causes neurons to transition from replication to differentiation (28,29). Prenatal nicotine exposure has been found to induce apoptotic cell death and decrease cell size in numerous brain regions (29–33). Further, activation of nicotinic AChRs by nicotine interferes with proper development of neurotransmitter systems, including dopamine (DA), norepinephrine (NE) and serotonin (5–HT) (25,27,34–36). Nicotine-induced plasticity in these neurotransmitter systems has been implied by changes in neurobehavioral development. Locomotor hyperactivity (30,37–39), depression (40–42) and changes in sensitivity to nicotine (43,44) as well as other psychostimulants (45,46) have been reported following nicotine exposure in utero. Thus, nicotine exposure during early development alters proper development of neuroanatomy, nicotinic AChR expression and neurotransmitter system function and leads to long-term changes in the trajectory of neurobehavioral development.

Due to obvious ethical reasons, it is necessary to employ preclinical animal models in order to successfully address questions regarding human brain maturation, and as a result, cross-species extrapolation must be employed. All mammals pass through the same stages of brain development; however, the timing associated with these stages of development varies greatly between species with regard to parturition (47–49). Both rats and mice have short gestational periods (22.5 and 19.5 days, respectively) and are generally resistant to disease (49). As the general species of choice for biomedical research, the neuroanatomy and neurophysiology of rats are well mapped and have been correlated with behavioral assessments (49). Mouse neuroanatomy and neurophysiology also are well established and this species is generally considered to be better suited to genetic manipulation (49). As a result, rats and mice are the two animal models most often employed in studies investigating brain development, and as such, are the animal model focused upon in this review. In rats and mice, the period of time corresponding with the first and second trimesters of human brain maturation take place prenatally, while the third trimester equivalent occurs postnatally (47–50). When comparing rats to humans, the first 11 gestational days (GD) represent the “first trimester equivalent”, GD 12–22 represent the “second trimester equivalent” and the first 10 postnatal days (PND) comprise the “third trimester equivalent” (47–50). Generally, trimester equivalents in mice are similar to rats, but may be a day or two shorter, as a result of the shorter gestational period. In humans, the vast majority of brain development occurs prenatally. Consequently, the period of time known as the “brain growth spurt”, a period of rapid, transient brain growth, which begins in the third trimester in humans and ends by two years of age, occurs entirely postnatally in rats (47,48). By comparing these analogous stages of brain development, it becomes possible to draw reliable interpretations when attempting to extrapolate data from animal models to humans.

The purpose of the current review is to present an overview of the role of nicotinic AChRs in the developing CNS and to describe the effects of nicotine exposure during early development on neuroanatomy, nicotinic AChR expression and neurotransmitter system function. We will also describe how the effects of nicotine exposure during early development vary depending on the timing and duration of exposure. Emphasis will be placed on preclinical studies examining the period of time corresponding to prenatal development in humans, since numerous studies have demonstrated profound effects of nicotine and other drugs of abuse on neural development during this stage (47–48,51–53). Finally, alterations in neurobehavioral development that occur as a result exposure to nicotine during early development will be discussed.

2. Role of nicotinic AChRs in early CNS development

2.1. Nicotinic acetylcholine receptors

Nicotinic AChRs belong to the Cys-loop superfamily of ligand-gated ion channels that includes γ-aminobutyric acid (GABAA), glycine, and 5-HT3 receptors, and are comprised of five transmembrane subunits arranged around a central, water-filled pore (14,54,55). Activation of nicotinic AChRs by endogenous (e.g., acetylcholine; ACh) or exogenous (e.g., nicotine) agonists alters the orientation of the second transmembrane domain (M2) of all five subunits, and results in a widening of the water-filled pore to the point that cations (i.e., K⁺, Na⁺, Ca⁺⁺) pass through (56). Influx of Na⁺ and Ca⁺⁺ produces a rapid change in membrane potential and local increases in intracellular Ca⁺⁺ concentrations, which leads to membrane depolarization (56). In the brain, twelve distinct genes have been identified that encode ten different nicotinic AChR subunits, α2-α10 and β2-β4 (57,58). Nicotinic AChRs can be either homomeric or heteromeric in regards to subunit composition (14,54,59). α7-α9 form homomeric nicotinic AChRs, with five identical subunits incorporated into a receptor, while the remaining α subunits (α2–α6) combine with β subunits (β2–β4) to form heteromeric receptors (60). To date, the α8 subunit has only been identified in avian species (61); however, all the remaining subunits are expressed in mammals (58,60). Numerous aspects of nicotinic AChR function are dependent upon the specific subunit conformation (14,56). α7 nicotinic AChRs have the highest permeability to Ca⁺⁺, while heteromeric nicotinic AChRs (e.g., α4β2 AChRs) are less permeable to Ca⁺⁺ and consequently more permeable to Na⁺ (62–63). Similarly, α7 nicotinic AChRs desensitize much more rapidly than heteromeric receptors (64). Conversely, this affinity of α7 nicotinic AChRs for nicotine is relatively low (EC₅₀ = 200 µM) when compared with α4β2 nicotinic AChRs (EC₅₀ = 1.6 µM; 14). Incorporation of additional subunits further modifies the function of nicotinic AChRs. For instance, the presence of the α5 subunits in α4β2 nicotinic AChRs (i.e., α4α5β2) prevents the upregulation of protein expression that is commonly observed following chronic nicotine treatment (65).

Nicotinic AChR subtypes also have varying degrees of expression and distribution throughout the brain. In mammals, α4β2 nicotinic AChRs are by far the most prevalent, comprising more than 90% of all neuronal nicotinic AChRs (59,66). α7 nicotinic AChRs are the second most prevalent subtype, and are also widely distributed throughout the CNS (59). Receptors containing other subunits are far more limited in distribution, with expression typically being limited to specific brain regions. For example, α2-containing nicotinic AChRs are primarily found in hippocampus, cortex and amygdala (67,68), while α3β4 nicotinic AChRs are expressed primarily in autonomic nuclei in the brainstem (20,69). α6-containing nicotinic AChRs are highly expressed in striatum, nucleus accumbens, ventral tegmental area, substantia nigra, locus coeruleus and the retina, but are expressed only at extremely low levels in all other brain regions (70–72). The α10 subunit combines with α9 to form a heteromeric nicotinic AChR that is found only in the cochlea (73). Thus, in addition to the differences in function discussed above, nicotinic AChR subtypes differ greatly in terms of expression and distribution within the CNS.

2.2. Nicotinic AChR expression in the developing brain

Nicotinic AChR subtypes differ in regards to both temporal and regional patterns of expression in the developing brain (74) (Figure 1). In humans, α7 mRNA and protein are expressed in pons and midbrain in the early first trimester (at 5 weeks after conception) (75,76), followed by a transient increase in expression, and by 9 weeks, α7 can be detected in spinal cord, subcortical forebrain, and cortex (17,75,76). Similar to α7 nicotinic AChRs, α4β2 nicotinic AChR binding sites are detected very early in development (6 weeks) in pons and midbrain, followed by increased expression in cortex and cerebellum (8 weeks) and, later, in hippocampus and basal ganglia (25 weeks) (16,17,76,77). Both α7 and α4β2 nicotinic AChRs peak late in the third trimester before decreasing to levels seen in adults (17,76). Unfortunately, a clear picture of expression levels for other nicotinic AChR subtypes has yet to be fully elucidated in human brain at any stage of development.

Figure 1.

Temporal and regional expression of nicotinic AChR subunit mRNA in developing human and rat brain. Timeline for initial expression of mRNA for different nicotinic AChR subunits is illustrated. Given the diversity of nicotinic AChR subtypes as well as the differences in subunit composition and conformation (i.e. spatial arrangement of the different subunits within the receptor), mRNA expression is most often used to identify the potential for the presence of a specific nicotinic AChR subunit in different brain regions. However, it is important to note that expression of mRNA does not necessarily indicate the presence of functional receptors that incorporate that subunit. Abbreviations: GD, gestational day; PND, postnatal day.

Nicotinic AChR expression in the developing rat brain is strikingly similar to that observed in humans (78,79) (Figure 1). In rats, α7 mRNA and protein can be detected by GD 12–13, increases throughout development and peaks on PND 7 (corresponding to the third trimester equivalent in humans), before decreasing to much lower levels during adulthood (80–82). Regional expression patterns of α7 are also similar between rats and humans, with expression initially observed in cortex and thalamus, and later in hippocampus and spinal cord (80,81,83,84). In spinal cord, mRNA for α4 and β2 subunits is detected between GD 11–13, and both mRNA and protein levels increase throughout pre- and postnatal development (22,85). Interestingly, while mRNA for β2 nicotinic AChR subunits can be detected by GD 12–13, α4 mRNA is not robustly expressed until late in gestation (GD 17–19) (22), a finding that may account for the relatively late appearance (GD 20) of [³H]nicotine and [³H]epibatidine binding sites in the developing cortex (22,80,85). Postnatal levels of α4 and β2 mRNA vary depending upon the specific brain region. Cortical α4 mRNA peaks on PND 14 and remains high, whereas in hippocampus α4 expression displays an inverted-U shaped curve, with higher levels on PND 7 and 14 than on PND 1 or 28 (82). In contrast with human studies, prenatal expression of mRNA for nicotinic AChR subtypes besides α7 and α4β2 have been characterized in rats and mice. mRNA for α2, α3 and β4 nicotinic AChR subunits appear between GD 12–13, while α5 does not appear until GD 18 (20, 22, 80, 86, 87). Further, in rats, α6 and β3 nicotinic AChR mRNA does not appear until postnatal development (25, 88); indicating that nicotinic AChRs incorporating these subtypes do not appear until the third trimester equivalent. It is important to note that the majority of these studies examined temporal and regional differences in nicotinic AChR mRNA expression, and changes in gene expression do not always translate to changes in the number functional nicotinic AChR binding sites. Nevertheless, similar to humans, rat brains display temporal, regional and subtype-specific differences in expression of nicotinic AChR subunit mRNA in the developing brain.

2.3. Nicotinic AChRs modulate neurotransmitter system function in the developing brain

Nicotinic AChRs are involved critical processes of CNS development such as gene expression, neurite outgrowth, chemotactic signaling for neuronal projections, cell proliferation, differentiation, neurogenesis/gliogenesis, and apoptosis (19, 89). The temporal, regional and subtype-dependent heterogeneity of nicotinic AChRs allows cholinergic signaling the potential to have dramatic effects on brain maturation processes through a variety of mechanisms (78, 79). For instance, activation of α7 nicotinic AChRs by ACh induces neurite retraction, while inhibition of α7 nicotinic AChRs using the subtype-selective antagonist α-bungarotoxin (α-BTx) induces neurite extension (90–92). Further, application of the α7-selective antagonist α-BTx prevents normal apoptotic cell death during development (93–95). Activation of α4β2 nicotinic AChRs propagates spontaneous burst activity along the length of the spinal cord that is required for axonal pathfinding of motor neurons (96, 97). Nicotinic AChRs may also play a critical role in the switch in GABAergic signaling from excitatory, as is found in late gestational and early postnatal life, to inhibitory, as is found in adults (89, 98). During development, the excitatory GABA signal acts as a trophic factor and guides neuronal migration and neurite outgrowth (99–101), and activation of α7 nicotinic AChRs appears to modulate this effect (102, 103). Interestingly, a recent study found that α7 nicotinic AChRs co-localize with GABA_A receptors, indicating that cholinergic input may exert local postsynaptic effects at GABAergic synapses (102). Mice lacking the α7 nicotinic AChR subunit maintain an excitatory GABA profile for a longer period of time than is observed in wild-type mice (103), indicating that this subtype plays a critical role in determining the timing for the conversion of GABA signaling from excitatory to inhibitory.

Modulation of neurotransmitter release by presynaptic nicotinic AChRs is the most prevalent and best characterized role of nicotinic AChRs in the CNS (14, 54), and is believed to play a major role in proper development of catecholamine neurotransmitter systems. In particular, nicotinic AChRs are thought to play a critical role in the development of nigrostriatal and mesocorticolimbic DA systems (79). mRNA for D1 and D2 DA receptors are expressed early in development (104–106); however, functional binding sites do not appear until innervation from dopaminergic midbrain neurons occurs (104, 107). Activation of D1 receptors has been shown to have an antimitogenic effect, reducing entry of cells into the S phase of mitosis (108, 109), whereas activation of D2-like receptors has a pro-mitogenic effect, increasing DNA synthesis (109, 110). Interestingly, survival of DA neurons appears to be regulated by α4β2 nicotinic AChRs, which are highly expressed on DA terminals and are functional during development (25). Evidence for this comes from studies of α4 and β2 knockout mice, which have a higher incidence of unpruned terminal arbors and alterations in presynaptic neurotransmitter release than wild-type mice. Conversely, mice lacking the α6 nicotinic AChR subunit, which is almost exclusively expressed on DA nerve terminals in adult mice (72) do not exhibit alterations in the development of the mesostriatal DA system (111) This is due mostly likely to the fact that the α6 subunit is not incorporated into functional nicotinic AChRs until postnatal development (25, 88), after the majority of DA neurons are already present.

Nicotinic AChRs also modulate NE and 5-HT release in the developing brain. NE is important for guiding cortical differentiation, and for induction of astrogliosis and glial cell proliferation (112, 113). NE neurons in the locus coeruleus express high levels of nicotinic AChRs that regulate embryonic neurotransmitter release early in development (88, 114). Evidence from fetal rat locus coeruleus cells grown in culture suggests that these nicotinic AChRs regulate embryonic NE release (115). Moreover, the ability of nicotinic AChRs to modulate NE release in the neonatal rat hippocampus increases dramatically between birth and PND 10 (116). Similar to NE neurons, nicotinic AChRs are present on 5-HT neurons, and are expressed as early as GD 11 in rats, a time corresponding to the 5^th week of human development (112). 5-HT is thought to play a role in the development of cortical layers and differentiation of neuronal progenitors, and excessive 5-HT release prevents the normal development of the somatosensory cortex (117, 118). Both α7 and α4β2 nicotinic AChRs modulate 5-HT release, (119, 120), and by extension, cortical layer development and differentiation. Thus, by modulating neurotransmitter release, nicotinic AChRs indirectly influence multiple neurotransmitter systems, which in turn mediate normal brain maturational processes.

3. Effects of early nicotine exposure on neural development

3.1. Cell proliferation and cell death

Nicotine readily crosses the human placental barrier and studies have demonstrated that nicotine from inhaled tobacco smoke can easily distribute to the fetus (76, 121). Given the pervasive expression of nicotinic AChRs and the numerous processes that these receptors modulate during brain maturation, it is not surprising that exposure to nicotine during early of development results in deleterious effects, including alterations in both the expression and function of nicotinic AChRs within the CNS. As discussed above, cholinergic signaling plays an important role in mediating neuronal development (19, 89), and it has been hypothesized that nicotine acts as a neuroteratogen, in part, by prematurely evoking cholinergic signaling, and chronically desensitizing nicotinic AChRs, which precludes activation by endogenous ACh and leads to a “discoordination” of brain development (13, 28).

Early efforts by Slotkin and colleagues found that daily injections (subcutaneous; sc) of nicotine (6 mg/kg) to pregnant Sprague-Dawley rats from GD 4–20 decreased DNA content (~21%) in the brains of offspring, a result indicative of cell loss (122). Further, administration of nicotine during this stage of development, which corresponds to the first and second trimesters of human development, resulted in desynchronization of the ontogenetic patterns of DNA, RNA and protein in every brain region assessed. Prenatal nicotine administration also has been found to decrease cell numbers and/or cell size in numerous brain regions, including cortex, hippocampus, cerebellum, brainstem, somatosensory cortex, cuneate nucleus and hypothalamus (29–33, 39,123–126). Moreover, postnatal nicotine treatment during a stage corresponding to the third trimester of human prenatal development and the brain growth spurt also decreases cell numbers in the developing brain (48). Neonatal (PND 4–9) administration of nicotine (6 mg/kg/day) decreased the number of mitral cells and Purkinje cells in olfactory bulb and cerebellar vermis, respectively (127, 128), and increased expression of cell death markers within the CA3 subregion of hippocampus (129). Collectively, these results indicate that nicotine exposure during numerous stages of early brain development produces cell loss in brain.

In the developing brain, cell loss can result from either the direct toxic effects of a compound, which can activate necrotic or apoptotic signaling pathways, or from compound-induced inhibition of cell cycle progression, i.e., decreased proliferation rates. Hence, determining the precise mechanism of cell loss is particularly important when considering nicotine, since nicotinic AChRs have been shown to regulate both cell death signaling pathways and proliferation (33, 130–132). Prenatal nicotine exposure increases c-fos expression and terminal deoxynucleotidyl transferase (TUNEL) staining, two markers indicative of apoptosis (31, 126, 133, 134). Nicotine is also hypothesized to act as a proliferation enhancer in brain, an effect that results in increased cell numbers (13, 19, 132). In order to reconcile the idea of increased cell death and enhanced proliferation, one must consider the effects of nicotine on different types of cells within the brain. As mentioned above, early studies found that prenatal nicotine decreases DNA content within brain; however, this effect was transient in nature, dissipating prior to the start of adolescence (29, 122). Interestingly, the increase in DNA content occurred after completion of the neurogenesis stage, and during the ongoing process of gliogenesis (47), suggesting that increases in DNA content, and presumably the total number of cells, was the result of an increase in the number of glia, rather than new neurons. This is supported by studies demonstrating that prenatal nicotine increases glial fibrillary acidic protein (GFAP) expression, a marker of astrocyte activation, in cerebellum and hippocampus (30, 125). Thus, nicotine may induce cell death in neurons by activation of apoptotic pathways, while simultaneously enhancing the proliferation rates of glia. A recent study using nonhuman primates provides additional support for this hypothesis. Rhesus monkeys were exposed to environmental tobacco smoke either during gestation through the first 13 postnatal months or only postnatally from 6–13 months of age, and markers of cell damage were assessed in cerebrocortical brain regions (135). Both perinatal and prenatal nicotine exposure reduced total DNA content of neurons while simultaneously increasing the number of glia present, suggesting that environmental tobacco smoke produces effects similar to those found following nicotine treatment alone. In the latter study, damage to neuronal projections accompanied by reactive sprouting was also observed (135). Collectively, these results suggest that prenatal nicotine treatment induces cell death in neurons by activating apoptotic pathways, while simultaneously increasing the total number of glia by enhancing proliferation. It is possible that these effects of nicotine vary substantially across different brain regions and across different windows of drug exposure. Additional studies are needed to elucidate the specific nicotinic AChR subtypes and downstream signaling mechanisms involved in these effects of nicotine.

3.2. Nicotinic AChR expression

Alterations in proliferation rates and amount of cell death are extreme examples of the deleterious effects of developmental nicotine exposure. Nevertheless, neuroteratogens, such as nicotine, often can have negative consequences on the developing brain that are more subtle. In particular, nicotine-mediated changes in nicotinic AChR expression and function can have far reaching effects on normal brain development. Cultured cortical neurons derived from fetal human brain show an increase in [³H]epibatidine binding to non-α7-containing nicotinic AChRs and [³H]cytisine binding to β2-containing nicotinic AChRs, indicating an increase in nicotinic AChR number (136). Increased expression of α4 and α7 mRNA has also been observed in the brains of fetuses belonging to mothers who smoked tobacco during pregnancy (15). Results from animal models of brain development have contributed insight to the effects of prenatal nicotine treatment on nicotinic AChRs, particularly in regards to regional differences in expression. An early study by van de Kamp and Collins (137) found that nicotine (2 mg/kg/day via sc injection) from GD 11–21 increased [³H]nicotine binding in hippocampus, striatum, midbrain and cortex, indicating an increase in the high-affinity α4β2 nicotinic AChR subtype, and no change in α-[¹²⁵I]-BTx binding to the low affinity α7 nicotinic AChR (137). In contrast, an identical dose of nicotine delivered from GD 4–21 via osmotic minipump decreased [³H]epibatidine binding in the prefrontal cortex (PFC), nucleus accumbens, substantia nigra, and ventral tegmental area, indicating a decrease in non-α7 nicotinic AChRs in these regions (36). Thus, prenatal nicotine-mediated changes in nicotinic AChR expression differ in terms of both brain region and subtype.

Prenatal nicotine also modulates gene expression of nicotinic AChR subunits and again regional- and subtype-related differences are readily apparent. Daily injections of nicotine (1.5 mg/kg; sc) from GD 3–21 were found to increase α2, α4, α7 and β2 mRNA in whole rat brains, while no change in mRNA was observed for α3, α5, β3 and β4 subunits (24). Similarly, prenatal nicotine treatment (2 mg/kg/day) delivered via osmotic minipump from GD 7–21, a period of time roughly equivalent to the last half of the first trimester and the second trimester, increased α7, α4 and β2 mRNA in both cortex and hippocampus (138). In contrast, a more recent study using osmotic minipumps to deliver an identical dose of nicotine from GD 4 through PND 10, a period of time encompassing all three human trimester equivalents, reported decreased α4, α7 and β2 mRNA expression in hippocampus (139). Discrepancies among these different studies in regard to the direction of change in mRNA expression may be explained by differences in the length of nicotine exposure, i.e., first and second trimester only vs. all three trimesters. Furthermore, nicotine administration (2 mg/kg/day via osmotic minipump) from GD 4–21 was also found to decrease α3, α4, α5 and β4 mRNA expression in reward-relevant brain regions, such as PFC, nucleus accumbens, substantia nigra, and ventral tegmental area (36). Additional studies are needed to identify the specific temporal window of vulnerability for nicotine-mediated changes in gene and protein expression.

Similar to the findings with prenatal nicotine exposure, neonatal nicotine administration during a period of time corresponding to the third trimester of human development, also alters nicotinic AChR expression in brain. Neonatal rat pups injected daily with nicotine (3.4 mg/kg; sc) from PND 1-21 had increased [³H]epibatidine and [³H]nicotine binding in cortex, hippocampus, striatum, thalamus and brainstem, indicating an increase in non-α7 nicotinic AChRs, specifically the high affinity α4β2 subtype (140). Further, similar to findings from prenatal nicotine exposure studies (137), no change in α-[¹²⁵I]-BTx binding was observed, indicating that α7 nicotinic AChR expression is not altered (140). Additionally, oral administration of nicotine (6 mg/kg/day) from PND 1–8 increased expression of heteromeric nicotinic AChRs (i.e., non-α7) in cortex, hippocampus, thalamus and medial habenula, but not cerebellum (129, 141). Importantly, neonatal nicotine treatment is not associated with changes in α2, α3, α4, α7 or β2 mRNA expression (129, 140). Chronic nicotine treatment increases nicotinic AChR expression in adult brain, and this is usually independent of steady state changes in mRNA expression levels (142–144). Recently, it has been hypothesized that nicotine increases nicotinic AChR expression by acting as a pharmacological chaperone, increasing incorporation of subunits into functional nicotinic AChRs and shuttling them to the cell surface (142, 145). Thus, it can be hypothesized that in rats, prenatal nicotine exposure (i.e., first and second trimester) alters nicotinic AChR expression accompanied by either an increase or decrease in mRNA expression; whereas postnatal nicotine treatment (i.e. third trimester) alters nicotinic AChR expression without altering mRNA levels. Therefore, in the developing brain both translational and post-translational mechanisms contribute to prenatal nicotine-induced nicotinic AChR upregulation. Further, it is worth noting that some controversy exists regarding the functional status of nicotinic AChRs upregulated in response to agonist. Agonist-induced receptor desensitization is hypothesized to contribute to nicotinic AChR expression plasticity (145–147), and many studies have associated behavioral/physiological tolerance to nicotine as a result of these changes (147). However, other studies, particularly those using in vivo approaches, have reported increases in nicotinic AChR function following agonist-induced upregulation (59, 147, 148). These processes may depend on nicotinic AChR subtype, cellular localization, brain region and regional circuitry.

3.3. Neurotransmitter systems

In addition to changes in nicotinic AChR expression, several studies have also demonstrated that prenatal nicotine during early development alters neurotransmitter system function. For instance, daily injections of nicotine (1.5 mg/kg; sc) from GD 3–21 decreases expression of choline acetyltransferase (ChAT), the vesicular acetylcholine transporter (VChAT) and the high affinity choline transporter (CHT1) in both the forebrain and hindbrain (149). Further, neonatal rats (PND 4–9) that received nicotine (6 mg/kg/day) via oral intubation had decreased GABA concentrations within the developing olfactory bulb (127). The effects of nicotine on the development of monoamine neurotransmitter systems, specifically DA, NE and 5-HT have been characterized to a far greater extent. Prenatal nicotine (1.5 mg/kg/day) delivered via osmotic minipump throughout gestation decreased the total number of DA D2 receptor binding sites (B_max) in rat striatum (150) and ventral tegmental area (151). These results differ from an early clinical study demonstrating that fetuses from mothers who smoked during pregnancy exhibited increased DA binding in whole brain homogenates (152). Additionally, prenatal nicotine exposure appears to have region-specific effects on DA content and/or turnover. In cortex and hypothalamus, prenatal nicotine decreases DA turnover (153–156). In contrast, prenatal or postnatal nicotine exposure increases the synthesis of DA from tyrosine, enhances DA turnover and decreases DA content in striatum (151, 154, 157). These results are in agreement with another study demonstrating that prenatal nicotine (3 mg/kg/day from GD 4–20) decreases nicotine-evoked DA release in nucleus accumbens of adolescent rat pups (158), and suggests that prenatal nicotine exposure blunts the ability of nicotine to evoke DA release.

In comparison to DA, there is paucity of literature regarding the effects prenatal nicotine on the NE neurotransmitter system. However, the argument can be made that this system may be even more sensitive to nicotine than the DA system (114, 159). Support for this hypothesis comes from studies demonstrating that cultured NE neurons from the locus coeruleus have nicotinic AChRs with a much higher affinity for nicotine than those found on DA terminals (25, 114). Further, prenatal nicotine administration is associated with increased NE content and turnover in forebrain and hypothalamus (153, 160). Prenatal nicotine exposure also sensitizes nicotinic AChRs on NE terminals, as evidenced by an increase in nicotine-evoked [³H]NE release from slices of parietal cortex taken from pups administered nicotine prenatally (3 mg/kg/day from GD 4–21) compared to controls (114). These findings are in contrast with studies showing that the same prenatal nicotine exposure (3 mg/kg/day from GD 4–20) decreases nicotine-evoked DA release (158), and suggest that prenatal nicotine exposure may produce opposite effects on DA and NE neurotransmitter systems.

Studies have also investigated the effects of prenatal nicotine exposure on 5-HT neurotransmission. Similar to DA and NE, prenatal nicotine exposure alters expression of 5-HT receptors (157, 161). For instance, studies using rats and non-human primates have demonstrated that both prenatal and postnatal nicotine exposure increases 5-HT₂ receptors (157, 161, 162). Conversely, studies have demonstrated that 5-HT_1A receptors can be either upregulated or downregulated by prenatal nicotine exposure (3 mg/kg/day from GD 4–21), and this is dependent upon both brain region and sex (157, 163). Moreover, changes in 5-HT receptor expression also are dependent upon the specific stage of development at which nicotine is administered. Decreased 5-HT_1A expression is observed more commonly following prenatal nicotine exposure (6 mg/kg/day from GD 4–21) (163); whereas increased 5-HT_1A expression has been observed following postnatal nicotine administration (162). Prenatal nicotine administration (3 and 6 mg/kg daily via osmotic pump) from GD 4–21 decreased 5-HT turnover in both forebrain and cerebellum (155, 157), and increased [³H]paroxetine binding in forebrain, cortex, midbrain and brainstem, indicating an increase in 5-HT transporter (SERT) (27, 164). Changes such as these in the 5-HT system may play a role in the increased incidence of SIDS that is observed in children exposed to tobacco smoke in utero, and this has been recently reviewed (78, 79). In summary, prenatal nicotine has numerous effects on developing monoaminergic neurons, including changes in receptor expression, transporter function and neurotransmitter turnover, all of which play an important role in mediating neurobehavioral development in adolescence and adulthood.

4. Neurobehavioral changes caused by early nicotine exposure

4.1. Attention deficit hyperactivity disorder

While there are numerous behavioral disorders that have been associated with maternal tobacco smoking (165, 166), an increased of incidence of ADHD is by far the best documented (167–171). Clinical research has shown that the pathophysiology of ADHD is likely the result of dysfunction in catecholamine neurotransmitters systems that are prominently mediated by nicotinic AChRs (42, 172, 173). A recent hypothesis suggests that ADHD symptoms are mediated by extracellular concentrations of DA being “out of tune” in the PFC (174, 175). Extracellular concentrations of DA and NE are associated to the relevance of the stimuli occurring in the environment (174). Moderate extracellular concentrations of DA and NE allow for proper filtration of irrelevant information and increased attention on the specific task at hand, respectively. When extracellular NE concentrations are too high, excess DA is released, leading to over activation of D1 and D4 receptors, which dysregulates filtration and leads to attention being focused on items not germane to the task at hand (174, 175). Conversely, if insufficient concentrations of extracellular DA are present, too few D1 and D4 receptors are activated, which prevents proper filtration of information (174, 175). As was discussed above, prenatal nicotine has been shown to decrease DA release and increase NE release in different brain regions. Thus, prenatal nicotine-mediated changes in extracellular DA and NE concentrations may directly contribute to the ADHD symptomology observed in children of mothers that smoke during pregnancy. However, to date no studies have examined the effects of prenatal nicotine on DA and NE release and/or content in PFC. Further, Becker and colleagues (176) recently found that males with prenatal smoke exposure who were homozygous for the 10-repeat allele of the dopamine transporter 1 (DAT1 10r) demonstrated increased hyperactivity-impulsivity than males that either had the DAT1 10r allele and were not exposed to tobacco smoke, or did not have the DAT1 10r and were exposed to tobacco smoke in utero. Interestingly, in females, no gene×environment interaction was noted, indicating gender differences as well (176). Together, these results suggest that prenatal nicotine exposure may interact with unknown genetic factors to increase the incidence of ADHD.

Maternal smoking more than triples the likelihood of being diagnosed with ADHD (177) and studies estimate that maternal smoking has increased the number of children in the United States with ADHD by more than 270, 000 (178). Exposure to tobacco smoke in utero increases inattention, hyperactivity, externalization problems and aggression in both young children and adolescents (179–182). Further, the severity of ADHD symptomology in childhood is positively correlated with the number of cigarettes smoked per day during pregnancy (183). Moreover, exposure to environmental tobacco smoke (i.e. “second-hand smoke) results in inattention and hyperactivity in children, indicating that even passive exposure to tobacco smoke can produce ADHD-like symptoms (184). A recent study found that children of mother that smoked during pregnancy were more resistant to treatment of ADHD than peers from mothers that did not smoke during pregnancy (185).

In addition to nicotine, tobacco smoke contains more than 4,000 chemicals (11–12), several of which, such as lead (186), have been linked to ADHD. Thus, the argument could be made that nicotine is only be a contributing factor to ADHD, and that the cumulative effects of the various neuroteratogens present within tobacco smoke are necessary to produce ADHD. However, numerous preclinical studies have demonstrated that nicotine alone is capable of producing ADHD-like symptoms in animal models. Prenatal nicotine administration, both with or without subsequent postnatal exposure, increases locomotor activity in the open-field (187–190). In addition, hyperactivity in the open-field is far more prevalent in males than females (37), a finding that reflects the clinical literature (166, 169). Further, both prenatal exposure (GD 4–21) and prenatal and postnatal (GD1 to PND 10) to nicotine in animal models produces anxiety-like behavior and attentional deficits, two symptoms commonly observed in children with ADHD (139, 191). These studies support the hypothesis that nicotine exposure is a principle, if not the primary, cause of hyperactivity and attentional deficits that occur following prenatal exposure to tobacco smoke.

4.2. Depression

A strong correlation exists between tobacco smoking and mood disorders (192, 193). Individuals with clinical depression are more likely to use tobacco than non-depressed individuals, have greater difficulty quitting and have withdrawal symptoms (193–195). Smokers with a history of major depression are more likely to experience symptoms of depression when undergoing cessation (196). Similarly, mothers that smoke during pregnancy are more likely to experience postpartum depression (197). Thus, it is not surprising that several clinical studies have observed an increase in mood disorders in children that were exposed to tobacco in utero (40–42). For instance, children of women who smoked during pregnancy were found to have an increase in oppositional behavior and emotional instability at three years of age (198). Children of women who smoked 10 or more cigarettes per day during pregnancy were far more likely to demonstrate symptoms of depression, even at 18-months of age (199), indicating that maternal smoking is capable of altering the trajectory of neurobehavioral development even at a very young age. Additionally, children exposed to prenatal tobacco smoke with self-reported depression were far more likely to use tobacco during adolescence than their peers (200). At least one preclinical study has also reported that prenatal exposure to nicotine increases depression-like symptoms. Offspring of C57/Bl mice treated from GD 1 through PND 6 with nicotine (50 µg/ml oral nicotine solution to dams) showed an increased latency to escape in the learned helplessness behavioral paradigm (190). Further, as discussed above, prenatal nicotine exposure is well known to decrease 5-HT turnover in brain (155,157), and decreased 5-HT concentrations in brain are often seen in depression (10, 162). Likewise, there is accumulating preclinical evidence that the functional state of the 5-HT neurotransmitter system during prenatal and postnatal development plays a role in modulating the response to stress and depression in adulthood (190, 201–203). Future clinical and preclinical research is needed to further elucidate the effects of early nicotine exposure on affective disorders.

4.3. Addiction

As discussed in detail above, exposure to nicotine during early development has an immense impact on normal development of the monoamine neurotransmitter systems. Given the role that these systems play in addiction, several studies, both clinical and preclinical have investigated the effects of prenatal nicotine on addiction in adolescents and adults. Maternal smoking during pregnancy, particularly during the third trimester of gestation is strongly associated with an increased risk of the child's tobacco experimentation and subsequent nicotine dependence (200, 204). Children whose mothers reported smoking during pregnancy were more than twice as likely to have initiated cigarette smoking during adolescence and to develop nicotine dependence compared to peers whose mothers did not smoke (205, 206). Maternal smoking has also been associated with an increased likelihood of abuse of other illicit substances, such as cocaine and marijuana, in adolescence (207, 208).

Results from preclinical studies suggest that prenatal nicotine exposure may alter the rewarding effects of nicotine and other psychostimulants in the brain. For instance, nicotine (3 mg/kg/day) exposure from GD 4–21 via osmotic minipump increased nicotine- and cocaine-mediated locomotor sensitization, a measure of the rewarding properties of a drug, in adult rats (209, 210). This increase in locomotor sensitization is observed in rats as old as PND 450 (210), indicating that the effects of prenatal nicotine on the rewarding properties of a drug last throughout adulthood. Furthermore, studies using mice have found that cocaine-induced conditioned place preference, another measure of the rewarding properties of a drug, is enhanced in mice treated with nicotine from GD 1 through early postnatal development (~PND 6; 190). Prenatal nicotine exposure (3 or 6 mg/kg/day from GD 4-21) also has been shown to increase operant responding for nicotine and cocaine (46, 211), suggesting that in addition to enhancing the inherent rewarding properties of psychostimulants, prenatal nicotine also increases the reinforcing properties of a drug. Interestingly, prenatal nicotine also increased operant responding for sucrose, suggesting that responding for natural rewards is also augmented (46). These changes in the rewarding and reinforcing properties of nicotine may be mediated by prenatal nicotine-induced alterations in nicotine-evoked DA and NE release (116, 158). Collectively, these studies suggest that early exposure to nicotine may increase the probability of an individual to develop a substance abuse problem in adulthood (166, 200).

5. Conclusions

Findings from the clinical and preclinical literature discussed in the current review demonstrate that nicotine exposure during early CNS development produces alterations in neuronal cytoarchitecture, nicotinic AChR expression, neurotransmitter system function and ultimately, neurobehavioral development. Premature activation of cholinergic signaling pathways in the developing CNS by nicotine leads to “discoordination” of brain development (13, 28). Studies such as those discussed here provide evidence that prenatal nicotine exposure alters the neurobehavioral trajectory of an individual, resulting in long-term negative consequences. Furthermore, these studies suggest that nicotine replacement therapy may not be an ideal smoking cessation therapy in pregnant women (6). It has been previously hypothesized that nicotine replacement therapy is a safer alternative compared to continued maternal smoking during pregnancy (212, 213), and it is difficult to argue against this point. However, it should be reinforced that no study has ever demonstrated that nicotine replacement therapy is a harmless alternative to smoking, nor has nicotine replacement been found to be an efficacious cessation approach in pregnant women. Thus, while it is true that this may protect the developing fetus from other harmful environmental toxins present in tobacco smoke, one must be careful not to suggest that nicotine replacement is “safe” in pregnancy, particularly in light of the delayed, but persistent, effects discussed in this review.

Moreover, considerable gaps still remain in our knowledge regarding the specific effects of nicotine on the developing brain. The majority of studies that have examined nicotine-mediated changes in neuroanatomy, nicotinic AChR expression, neurotransmitter function and neurobehavioral development have looked at each of these parameters separately. Future studies will have the challenge of designing parametric analyses using the same dose of nicotine for the same length of time during gestation and measuring changes in cytoarchitecture and nicotinic AChR expression and function concurrently. In particular, studies are lacking describing the effects of prenatal and/or neonatal nicotine exposure on nicotine-evoked neurotransmitter release and catecholamine transporter function in adolescent and adult animals. Finally, given that environmental tobacco smoke has been found to produce changes in neurobehavioral development, there is a great need to compare the effects of direct and passive nicotine exposures on the developing CNS.