Toxicology 333 (2015) 63–75

Amelioration strategies fail to prevent tobacco smoke effects onneurodifferentiation: Nicotinic receptor blockade, antioxidants, methyldonors

Theodore A. Slotkin a,*, Samantha Skavicus a, Jennifer Card a, Edward D. Levin b,Frederic J. Seidler a

aDepartment of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, NC 27710, USAbDepartment of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC 27710, USA

A R T I C L E I N F O

Article history:Received 12 February 2015Received in revised form 12 March 2015Accepted 14 April 2015Available online 17 April 2015

Keywords:AntioxidantsMethyl donorsNeurodifferentiationNicotinePC12 cellsTobacco smoke extract

A B S T R A C T

Tobacco smoke exposure is associated with neurodevelopmental disorders. We used neuronotypicPC12 cells to evaluate the mechanisms by which tobacco smoke extract (TSE) affects neuro-differentiation. In undifferentiated cells, TSE impaired DNA synthesis and cell numbers to a muchgreater extent than nicotine alone; TSE also impaired cell viability to a small extent. In differentiatingcells, TSE enhanced cell growth at the expense of cell numbers and promoted emergence of thedopaminergic phenotype. Nicotinic receptor blockade with mecamylamine was ineffective in preventingthe adverse effects of TSE and actually enhanced the effect of TSE on the dopamine phenotype. A mixtureof antioxidants (vitamin C, vitamin E, N-acetyl-L-cysteine) provided partial protection against cell loss butalso promoted loss of the cholinergic phenotype in response to TSE. Notably, the antioxidants themselvesaltered neurodifferentiation, reducing cell numbers and promoting the cholinergic phenotype at theexpense of the dopaminergic phenotype, an effect that was most prominent for N-acetyl-L-cysteine.Treatment with methyl donors (vitamin B12, folic acid, choline) had no protectant effect and actuallyenhanced the cell loss evoked by TSE; they did have a minor, synergistic interaction with antioxidantsprotecting against TSE effects on growth. Thus, components of tobacco smoke perturb neuro-differentiation through mechanisms that cannot be attributed to the individual effects of nicotine,oxidative stress or interference with one-carbon metabolism. Consequently, attempted ameliorationstrategies may be partially effective at best, or, as seen here, can actually aggravate injury by interferingwith normal developmental signals and/or by sensitizing cells to TSE effects on neurodifferentiation.

ã 2015 Elsevier Ireland Ltd. All rights reserved.

Contents lists available at ScienceDirect

Toxicology

journa l homepage : www.e l sev ier .com/ loca te / toxicol

1. Introduction

In addition to its major contribution to perinatal morbidity andmortality (Abbott and Winzer-Serhan, 2012; DiFranza and Lew,1995; Pauly and Slotkin, 2008), prenatal tobacco smoke exposuresubstantially increases the risk of neurodevelopmental disorders,including learning disabilities, attention deficit/hyperactivitydisorder and conduct disorders (Cornelius and Day, 2009; DiFranzaand Lew, 1995; Gaysina et al., 2013; Pauly and Slotkin, 2008;Wakschlag et al., 1997). These outcomes reflect, in large measure,

Abbreviations: ANOVA, analysis of variance; ChAT, choline acetyltransferase;MDA, malondialdehyde; NAC, N-acetyl-l-cysteine; TH, tyrosine hydroxylase; TSE,tobacco smoke extract.* Corresponding author at: Box 3813 DUMC, Duke Univ. Med. Ctr., Durham, NC

27710, USA. Tel.: +1 919 681 8015; fax: +1 919 684 8197.E-mail address: t.slotkin@duke.edu (T.A. Slotkin).

http://dx.doi.org/10.1016/j.tox.2015.04.0050300-483X/ã 2015 Elsevier Ireland Ltd. All rights reserved.

the adverse impact of nicotine on brain development (Pauly andSlotkin, 2008; Slikker et al., 2005; Slotkin, 2004, 2008). Neuro-transmitter signals provide key neurotrophic information for brainassembly (Dreyfus, 1998; Hohmann, 2003; Lauder, 1985), so thatthe inappropriate timing and intensity of cholinergic receptorstimulation by nicotine preempts normal developmental process-es, leading to defects in neuronal cell replication and differentia-tion, in axonogenesis and synaptogenesis, and in the formation ofneural circuits (Pauly and Slotkin, 2008). However, given theadvent of nicotine replacement products for smoking cessation, aswell as alternative nicotine delivery devices, it becomes importantto distinguish whether the thousands of other components oftobacco smoke also play a role in adverse neurodevelopmentaloutcomes. A number of animal studies have identified effects ofcigarette smoke that, in general, resemble those of nicotine at thebiochemical, structural and functional levels (Bruijnzeel et al.,2011; Fuller et al., 2012; Golub et al., 2007; Gospe et al., 2009; Lobo

64 T.A. Slotkin et al. / Toxicology 333 (2015) 63–75

Torres et al., 2012; Sekizawa et al., 2008; Slotkin et al., 2006a,b).Nevertheless, these exposure models simply reinforce the resem-blance between smoke exposure and the effects of nicotine, ratherthan distinguishing between them. Additionally, they add anuncontrolled variable, since repetitive, involuntary exposure in asmoke-filled chamber is likely to elicit stress.

To study the direct effects of tobacco smoke on neuro-differentiation without participation of these confounds, werecently compared nicotine to other tobacco smoke products inPC12 cells, a neuronotypic cell line widely used to study neuro-differentiation (Costa, 1998; Teng and Greene, 1994). This cell lineis poorly responsive to nicotine despite the presence of nicotinicacetylcholine receptors, generally requiring concentrations as highas 100–200 mM for a full effect (Abreu-Villaça et al., 2005; Avilaet al., 2003; Gueorguiev et al., 2000). Using tobacco smoke extract(TSE) at concentrations where nicotine by itself had little or noeffect, we found that TSE promotes the transition from cellreplication to neurodifferentiation (Slotkin et al., 2014), resultingin deficits in cell numbers. Additionally, TSE alters neurodiffer-entiation outcomes, promoting emergence of the dopaminergicphenotype over the cholinergic phenotype. In the current study,we used these basic findings to address two issues. First, what arethe mechanisms underlying the effect of TSE on neurodifferentia-tion, and second, can we use that information to ameliorate theeffects? We focused on three likely mechanisms: actions onnicotinic receptors (the target for nicotine), oxidative stress andinterference with one-carbon metabolic pathways. Smokingcauses fetal oxidative stress (Aycicek and Ipek, 2008; Fayolet al., 2005; Gitto et al., 2002; Orhon et al., 2009) and consequently,attempts have been made to ameliorate the adverse effectsthrough vitamin C supplementation. In vivo, vitamin C preventstobacco smoke-induced lung damage in the fetus and improvesneonatal respiratory outcome (McEvoy et al., 2014; Proskocil et al.,2005). While vitamin C also prevents oxidative damage to thedeveloping brain (Slotkin et al., 2011), it increases fetal nicotinelevels, enhancing damage attributable to inappropriate nicotinicreceptor stimulation (Slotkin et al., 2005). Here, using an in vitrosystem that eliminates pharmacokinetic factors, we explored theability of three antioxidants in combination or separately toameliorate the effects of TSE: vitamin C, vitamin E and N-acetyl-L-cysteine (NAC). Likewise, smokers are often advised to supplementtheir diets with methyl donors, including vitamin B12, folic acid,and choline (Boeke et al., 2013; Steegers-Theunissen et al., 2013).Accordingly we also evaluated amelioration strategies with theseagents.

2. Methods

TSE (Arista Laboratories, Richmond, VA) was prepared fromKentucky Reference cigarettes (KY3R4F) on a Rotary SmokeMachine under ISO smoke conditions. The smoke condensatewas collected on 92 mm filter pads, which were then extracted byshaking for 20 min with dimethyl sulfoxide, to obtain a solution ofapproximately 20 mg of condensate per ml. Condensate aliquotswere stored in amber vials at �80 �C until used. Two cigaretteswere smoked to produce each ml of extract and the final productcontained 0.8 mg/ml (5 mM) nicotine (determined by the manu-facturer).

2.1. Cell cultures and TSE treatments

Because of the clonal instability of the PC12 cell line (Fujitaet al., 1989), the experiments were performed on cells that hadundergone fewer than five passages. As described previously (Qiaoet al., 2003; Song et al., 1998), PC12 cells (American Type CultureCollection CRL-1721, obtained from the Duke Comprehensive

Cancer Center, Durham, NC) were seeded onto poly-D-lysine-coated plates in RPMI-1640 medium (Sigma Chemical Co., St. Louis,MO) supplemented with 10% horse serum (Sigma), 5% fetal bovineserum (Sigma), and 50 mg/ml penicillin streptomycin (Invitrogen,Carlsbad, CA). Incubations were carried out with 5% CO2 at 37 �C,standard conditions for PC12 cells. For studies in the undifferenti-ated state, the medium was changed 24 h after seeding to includetest agents. TSE was evaluated at low and high levels (“TSE low”

and “TSE high”), calculated to produce final concentrations of 1 mMor 10 mM nicotine in the culture medium, corresponding to a1/5000 and 1/500 dilution of the condensate. As a positive controlfor comparison with TSE, we used nicotine bitartrate (Sigma) at afinal concentration of 10 mM, a concentration previously verified inthe PC12 model (Abreu-Villaça et al., 2005; Qiao et al., 2003;Slotkin et al., 2014). To control for the dimethyl sulfoxide vehicle inthe TSE samples, all cultures contained final concentrations of 0.2%dimethyl sulfoxide (Sigma), which has no effect on PC12 cellgrowth or differentiation (Qiao et al., 2001; Song et al., 1998).Studies were then conducted after 24 h of exposure.

To initiate neurodifferentiation (Jameson et al., 2006b; Slotkinet al., 2007b; Teng and Greene, 1994), the medium was changed toinclude 50 ng/ml of 2.5 S murine nerve growth factor (PromegaCorporation, Madison, WI). TSE exposure commenced simulta-neously with the addition of nerve growth factor, so as to bepresent throughout neurodifferentiation. The medium waschanged every 48 h with the continued inclusion of nerve growthfactor and continued for either 4 days or 6 days, depending on themeasured endpoint. At the end of each study, the cultures wereexamined under a microscope to verify the outgrowth of neurites.

2.2. Amelioration treatments

Each amelioration treatment was begun simultaneously withthe exposure to TSE, whether in undifferentiated or differentiatingcells. Control samples were always included that contained theameliorant alone, so that each experiment consisted of fourtreatment groups: control, TSE alone, ameliorant alone, amelio-rant + TSE. The final concentrations of the ameliorants were chosenfrom earlier studies that demonstrated the requisite effect of eachwith this model: nicotinic acetylcholine receptor blocker, 10 mMmecamylamine (Slotkin et al., 2007a); antioxidants,10 mM vitaminC (sodium ascorbate) (Slotkin and Seidler, 2010), 10 mM vitamin E(a-tocopherol) (Slotkin and Seidler, 2010), 5 mM NAC (Lee et al.,2012); methyl donors, 0.2 mM vitamin B12 (Savelkoul et al., 2012),15 mM folic acid (Savelkoul et al., 2012), 100 mM choline (Knappand Wurtman, 1999) (all agents from Sigma). Vitamin E wasdissolved in ethanol, yielding a final ethanol concentration of 0.05%in the medium; for experiments with this agent, all samplescontained this ethanol concentration, which by itself has no effecton PC12 cell neurodifferentiation (Slotkin et al., 2007a). The finalcholine concentration is approximately five times the normalconcentration already contained in the culture medium.

2.3. Assays

All of the techniques used in this study have appeared inprevious papers and accordingly, only brief descriptions ofprocedures will be given.

To assess DNA synthesis, undifferentiated cells were exposed tothe test agents for 24 h, and then, for the final hour of exposure, themedium was changed to include the agents along with 1 mCi/ml of[3H]thymidine (specific activity, 2Ci/mmol; PerkinElmer LifeSciences, Waltham, MA). After 1 h, the medium was aspiratedand cells were washed repeatedly to allow unincorporated label todiffuse out, leaving only [3H]thymidine that was incorporated intoDNA remaining within the cell. Cells were then homogenized and

T.A. Slotkin et al. / Toxicology 333 (2015) 63–75 65

an aliquot counted for radiolabel. This technique is a simplificationof an established procedure (Bell et al., 1986; Slotkin et al., 1984),and we conducted preliminary experiments to show that it gaveequivalent results. We corrected incorporation values to theamount of DNA present in each culture to provide an index of DNAsynthesis per cell (Winick and Noble, 1965) and the total DNAcontent was also recorded. The 24 h time point was chosen toensure that the cells were still in the log-phase of replication andwell short of confluence.

For measurements of cell number and cell size, we relied on thefact that neural cells contain a single nucleus, so that the DNAcontent provides a measure of cell number (Winick and Noble,1965). Cells were harvested, washed, and the DNA and proteinfractions were isolated and analyzed as described previously(Slotkin et al., 2007b). Since the DNA per cell is constant, cellgrowth entails an obligatory increase in the protein/DNA ratio(Qiao et al., 2003; Song et al., 1998).

To assess neurodifferentiation into dopamine and acetylcholinephenotypes, we assayed the activities of tyrosine hydroxylase (TH)and choline acetyltransferase (ChAT), respectively, using estab-lished techniques (Jameson et al., 2006a,b). Assessments weremade 6 days after initiating neurodifferentiation with NGF, withcontinuous coexposure to test agents.

Effects on viability were assessed after 24 h of exposure ofundifferentiated cells, or 4 days of exposure of differentiating cells.The cell culture medium was changed to include trypan blue(1 volume per 2.5 volumes of medium; Sigma) and cells wereexamined for staining under 400� magnification, counting anaverage of 100 cells per field in four different fields per culture.

Fig. 1. Comparison of TSE effects with nicotine in undifferentiated cells after 24 h of exundifferentiated cells; (B) DNA content in undifferentiated cells; (C) trypan blue exclusioexclusion in differentiating cells. Data represent means and standard errors of the numbethe top of each panel and asterisks denote values that differ significantly from the con

For evaluation of oxidative stress, we assessed the degree oflipid peroxidation in undifferentiated cells after 24 h of exposure totest agents, and in differentiating cells after 4 days of exposure. Theconcentration of malondialdehyde (MDA) was measured withthiobarbituric acid (Guan et al., 2003). To give the MDAconcentration per cell, values were calculated relative to theamount of DNA.

The 4 day time point for differentiating cells was chosenbecause effects on viability or oxidative stress would provide apresumptive, intervening mechanism for effects seen at the 6 dayendpoint of neurodifferentiation.

2.4. Data analysis

Each study was performed using 2–4 separate batches of cells,with 3–4 independent cultures for each treatment in each batch;each batch of cells comprised a separately prepared, frozen andthawed passage. Results are presented as mean � SE, withtreatment comparisons carried out by analysis of variance(ANOVA) followed by Fisher’s protected least significant differencetest for post-hoc comparisons of individual treatments. The initialcomparisons included factors of treatment and cell batch, and ineach case, we found that the treatment effects for each type ofexperiment were the same across the different batches of cells,although the absolute values differed from batch to batch.Accordingly, we normalized the results across batches prior tocombining them for presentation. Studies of amelioration of TSEeffects were analyzed by two-factor ANOVA: factor 1 = with orwithout TSE; factor 2 = with or without ameliorant. For these, the

posure, and in differentiating cells after 4 days of exposure: (A) DNA synthesis inn in undifferentiated cells; (D) DNA content in differentiating cells; (E) trypan bluer of determinations shown in parentheses. ANOVA across all treatments is shown attrol.

Fig. 2. Effects of mecamylamine on the response to TSE in the undifferentiated state (24 h exposure) and during differentiation (6 day exposure): (A) DNA synthesis inundifferentiated cells, (B) DNA content in undifferentiated cells, (C) DNA content in differentiating cells, (D) protein/DNA ratio in differentiating cells, (E) tyrosine hydroxylaseactivity in differentiating cells, and (F) choline acetyltransferase activity in differentiating cells. Data represent means and standard errors of the number of determinationsshown in parentheses. MANOVAs above each set of panels combine all the measurements for each differentiation state, and two-factor ANOVAs are shown at the top of eachpanel for each individual measure. Where there was a mecamylamine � TSE interaction (E), asterisks denote whether the TSE group is significantly different from thecorresponding control value. Abbreviation: NS, not significant.

66 T.A. Slotkin et al. / Toxicology 333 (2015) 63–75

T.A. Slotkin et al. / Toxicology 333 (2015) 63–75 67

initial ANOVA assessed the main effects of TSE and ameliorants andthe ameliorant � TSE interaction. Where the ameliorant � TSEinteraction was significant, we evaluated individual differencesbetween treatments, but in the absence of an interaction, only themain effects are reported without lower-order tests.

In addition to evaluating each individual set of measurements,we performed a global MANOVA to evaluate treatment effectsacross all measurements that were done on the same samples, soas to avoid an increased probability of type I errors that could arisefrom multiple tests of the same cultures. At the same time, this testalso increases statistical power by combining effects acrossmultiple measurement types. Data were log-transformed becauseof heterogeneous variance across the different measures, andsignificance was assessed by Wilks’ l. Significance for all tests wasassumed at p < 0.05 (two-tailed).

3. Results

3.1. Comparison of TSE with nicotine

In undifferentiated cells, 24 h of TSE exposure elicitedconcentration-dependent inhibition of DNA synthesis, with thehigh TSE concentration producing a 22% decrement (Fig. 1A). Incontrast, 10 mM nicotine by itself, equivalent to the nicotineconcentration at high TSE, produced only a slight, nonsignificantreduction. The TSE effect on DNA synthesis was accompanied by areduction in the total number of cells, assessed by measuring DNAcontent (Fig. 1B). Again, an equivalent concentration of nicotineelicited a smaller, nonsignificant effect. Although the highconcentration of TSE caused a significant increase in nonviablecells, this represented a small net change from about 3% of the totalin controls to about 6% in the TSE group, so viability remained at94%, compared to the control value of 97% (Fig. 1C). By itself, 10 mMnicotine was without significant effect on viability.

In differentiating cells, the high concentration of TSE reducedthe number of cells after 4 days of exposure, an effect that was notshared by nicotine (Fig. 1D). Unlike the case for undifferentiatedcells, there were no significant changes in viability (Fig. 1E). In ourprevious work in differentiating cells, we also found that TSE, butnot nicotine, induced TH activity, connoting a shift in neuro-differentiation toward the dopaminergic phenotype (Slotkin et al.,2014). Based on these results, we evaluated the effects ofameliorant treatments against the high TSE concentration in theremaining studies.

3.2. Mecamylamine (nicotinic receptor antagonist)

In undifferentiated cells, MANOVA across the two measure-ments (DNA synthesis, DNA content) showed a significant effect ofTSE but not mecamylamine, and there was no interaction ofmecamylamine � TSE, connoting a lack of significant antagonism(Fig. 2). This conclusion was borne out with the individualmeasures. TSE elicited a significant decrease in DNA synthesis thatwas unaffected by mecamylamine treatment (Fig. 2A). The sameresult was seen for DNA content in undifferentiated cells (Fig. 2B).

This was not the case for the four measures made indifferentiating cells, where the MANOVA revealed not only a maineffect of TSE, but also a significant mecamylamine � TSE interac-tion. For DNA content, TSE produced the expected reduction butthis was not blocked by mecamylamine (Fig. 2C). Likewise, TSEincreased cell size (protein/DNA ratio) but again, mecamylaminedid not alter the effect (Fig. 2D). However, for TH activity,mecamylamine significantly enhanced the induction caused byTSE, more than doubling the increase (Fig. 2E). In contrast, therewere no significant changes in ChAT with TSE or mecamylaminealone or in combination (Fig. 2F).

3.3. Antioxidants (vitamin C, vitamin E, NAC)

Unlike the situation for mecamylamine, MANOVA for DNAsynthesis and content in undifferentiated cells showed significantoverall main effects of TSE and the antioxidant cocktail alone, as wellas a significant interaction of antioxidants � TSE (Fig. 3). However,because of the small magnitude of TSE’s effects on undifferentiatedcells, there was insufficient power to show a significant interactionfor the two measures taken individually. For DNA synthesis, TSEcaused a significant overall reduction, with antioxidant treatmentreducing the effect from 18% inhibition to 13% inhibition (Fig. 3A). ForDNA content, the TSE effect was reduced from 9% to 3% (Fig. 3B).Notably, though the antioxidants by themselves significantlyincreased DNA content, pointing to unexpected effects of the rescueagents, a finding that was reinforced when they were given todifferentiating cells, described below.

In differentiating cells, MANOVA combining the four differenttypes of measurements revealed significant effects of TSE andantioxidants individually, as well as a significant interaction ofantioxidants � TSE (Fig. 3). For DNA content, antioxidants bythemselves evoked a deficit in cell number but at the same time,they blocked the decrement caused by TSE (Fig. 3C). Antioxidantsalso partially protected the cells from the increase in protein/DNAratio evoked by TSE (Fig. 3D). In contrast, antioxidants provided nodiscernible protection from the effects of TSE on neurotransmitterphenotypes. By itself, TSE elicited about a 25% increase in TH(Fig. 3E). In the presence of antioxidants, TSE produced the sameincrease over the antioxidant-exposed control group. Importantly,the antioxidants alone suppressed TH activity relative to thevehicle controls. The effect of antioxidants was even more notablefor ChAT (Fig. 3F). Whereas TSE by itself had little or no effect onChAT, antioxidants alone produced an enormous increase andsensitized the cells to TSE: in the presence of antioxidants, TSEelicited a robust decrease in activity, an effect not seen with TSE inthe absence of the antioxidants.

In light of the limited ability of antioxidants to protect againstthe effects of TSE, we next evaluated lipid peroxidation as anendpoint for oxidative stress. In undifferentiated cells, neither TSEconcentration caused an increase in MDA levels after 24 h ofexposure, although a positive control, monovalent silver (Powerset al., 2010b), elicited a robust elevation (Fig. 4A). In fact, the highconcentration of TSE actually reduced MDA levels. Similarly, indifferentiating cells, the high TSE concentration reduced MDAlevels after 4 days of exposure, whereas monovalent silver eliciteda clear-cut elevation (Fig. 4B).

The unexpected, large effects of the antioxidant cocktail onneurotransmitter phenotypes led to an additional study todelineate whether any individual component was responsible.Vitamin C and NAC reduced TH but vitamin E was withoutsignificant effect (Fig. 5A). The combination of all three antiox-idants was indistinguishable from additive effects of vitamin C andNAC. For ChAT, only NAC evoked a significant increase by itself butaddition of the other two antioxidants augmented the net effectabove that achieved with NAC (Fig. 5B).

3.4. Methyl donors (vitamin B12, folic acid, choline)

In undifferentiated cells, MANOVA identified a significant maineffect of TSE but failed to find a significant interaction of methyldonors � TSE (Fig. 6). TSE inhibited DNA synthesis (Fig. 6A) andreduced cell numbers (Fig. 6B). The methyl donors alone had nosignificant effect, nor did they prevent the effects of TSE on eithermeasure.

In differentiating cells, MANOVA identified main effects of TSEand methyl donors individually, but no significant interaction whenthe four measures were combined (Fig. 6). TSE, as expected, reduced

Fig. 3. Effects of a mixture of antioxidants (10 mM vitamin C + 10 mM vitamin E + 5 mM NAC) on the response to TSE in the undifferentiated state (24 h exposure) and duringdifferentiation (6 day exposure): (A) DNA synthesis in undifferentiated cells, (B) DNA content in undifferentiated cells, (C) DNA content in differentiating cells, (D) protein/DNA ratio in differentiating cells, (E) tyrosine hydroxylase activity in differentiating cells, and (F) choline acetyltransferase activity in differentiating cells. Data representmeans and standard errors of the number of determinations shown in parentheses. MANOVAs above each set of panels combine all the measurements for each differentiationstate, and two-factor ANOVAs are shown at the top of each panel for each individual measure. Where there was an antioxidants � TSE interaction, asterisks denote whetherthe TSE group is significantly different from the corresponding control value and daggers indicate whether the antioxidants alone are different from the vehicle control.Abbreviation: NS, not significant.

68 T.A. Slotkin et al. / Toxicology 333 (2015) 63–75

Fig. 4. Effects of TSE on MDA concentrations, compared to a positive control, 10 mM AgNO3: (A) undifferentiated cells after 24 h of exposure; (B) differentiating cells after4 days of exposure. Data represent means and standard errors of the number of determinations shown in parentheses. ANOVA across all treatments is shown at the top of eachpanel and asterisks denote values that differ significantly from the control.

T.A. Slotkin et al. / Toxicology 333 (2015) 63–75 69

DNA content, but methyl donors increased DNA significantly(Fig. 6C). The methyl donors did not provide any protection fromthe effects of TSE, and in fact, the TSE-induced decrement wasactually larger because of the fact that the methyl donors aloneincreased DNA. As a result, we found a significant methyl donors� TSE interactionforDNAcontentbut thisshould beinterpretedwithcaution, since there was no significant interaction in the overallMANOVA. Likewise, we found that methyl donors did not protectdifferentiating cells from the effects of TSE directed toward cell size(Fig. 6D), TH activity (Fig. 6E) or ChAT activity (Fig. 6F).

Fig. 5. Effects of individual antioxidants and the combination of all three, on TH (A) and Cshown in parentheses. ANOVA across all treatments is shown at the top of each panel

3.5. Antioxidants combined with methyl donors

In undifferentiated cells, the combination of antioxidants andmethyl donors gave overall results resembling those of theantioxidants alone. MANOVA combining the measurements ofDNA synthesis and DNA content showed main effects of TSE andthe ameliorants separately as well as a significant interactionbetween them (Fig. 7). The main effect of the antioxidants + methyldonor cocktail represented a significant increase for both DNAsynthesis (Fig. 7A) and DNA content (Fig. 7B), in contrast to the

hAT (B). Data represent means and standard errors of the number of determinations and asterisks denote values that differ significantly from the control.

Fig. 6. Effects of a mixture of methyl donors (0.2 mM vitamin B12 + 15 mM folic acid + 100 mM choline) on the response to TSE in the undifferentiated state (24 h exposure) andduring differentiation (6 day exposure): (A) DNA synthesis in undifferentiated cells, (B) DNA content in undifferentiated cells, (C) DNA content in differentiating cells, (D)protein/DNA ratio in differentiating cells, (E) tyrosine hydroxylase activity in differentiating cells, and (F) choline acetyltransferase activity in differentiating cells. Datarepresent means and standard errors of the number of determinations shown in parentheses. MANOVAs above each set of panels combine all the measurements for eachdifferentiation state, and two-factor ANOVAs are shown at the top of each panel for each individual measure. Where there was a methyl donors � TSE interaction (C), asterisksdenote whether the TSE group is significantly different from the corresponding control value and the dagger indicates whether the methyl donors alone are different from thevehicle control. Abbreviation: NS, not significant.

70 T.A. Slotkin et al. / Toxicology 333 (2015) 63–75

T.A. Slotkin et al. / Toxicology 333 (2015) 63–75 71

decreases caused by TSE. The ameliorant treatments reduced theeffect of TSE by about half, although the interaction was significantonly for the effect on DNA content.

In differentiating cells, MANOVA across the four types ofmeasurements identified significant main effects of TSE and theameliorants separately, as well as a significant ameliorants � TSEinteraction. The interaction connoted a shift in the response to TSEin the presence of the combination of antioxidants and methyldonors (Fig. 7). For DNA content, we saw a pattern just like thatobtained for antioxidants without the methyl donors (Fig. 7C): TSEreduced DNA, as did the ameliorant cocktail, and the combinationof antioxidants + methyl donors protected the cells from the effectsof TSE to about the same extent as seen with just the antioxidants.For the protein/DNA ratio, the combined ameliorants by them-selves produced a small but significant increase that had not beenseen just antioxidants or just methyl donors (Fig. 7D). Further-more, the combination of antioxidants + methyl donors completelyprotected the cells from the increase evoked by TSE, whereas theprotection was only partial with just the antioxidants (compare toFig. 3D). To verify this difference we compared the rescue causedby antioxidants + methyl donors to rescue attributable to justantioxidants or just methyl donors, using a three-factor ANOVA:factor 1 = with or without TSE; factor 2 = with or withoutameliorant; factor 3 = antioxidants vs. antioxidants + methyldonors. We found a significant three-factor interaction (p < 0.02),indicating that the protection from the TSE effect provided by theameliorants was greater for the antioxidants + methyl donors,connoting synergism.

For the neurotransmitter differentiation endpoints, the combi-nation of antioxidants + methyl donors gave results similar to thoseseen with just the antioxidants. The ameliorant treatment loweredTH activity but did not block the induction caused by TSE (Fig. 7E).The antioxidants + methyl donors elicited a large increase in ChATactivity (Fig. 7F); under those conditions, addition of TSE elicited asubstantial decrement in ChAT that was not seen when TSE wasgiven without the ameliorants, resulting in a significant ameliorant� TSE interaction.

4. Discussion

In our previous work, we found that the effects of TSE ondifferentiating PC12 cells could not be explained solely by theeffects of nicotine (Slotkin et al., 2014). TSE accelerated neuro-differentiation, enhancing cell growth at the expense of cellnumbers, augmenting neurite formation and promoting emer-gence of the dopaminergic phenotype. Whereas nicotine couldproduce some of these effects (reduced cell numbers, enhancedcell growth), much higher concentrations were required than werepresent in TSE, and nicotine was unable to switch neurotransmitterphenotypes or to promote neurite formation. In the present study,we extended these observations to cells in the undifferentiatedstate, showing that TSE had greater effects on DNA synthesis andformation of new cells than did nicotine. Additionally, TSE, but notnicotine, impaired cell viability, albeit that the effect involved onlya small overall percentage of the cells. Interestingly, the reducedviability evoked by TSE was not seen in differentiating cells,reinforcing the earlier conclusion that reductions in cell numbersreflect promotion of the transition from cell replication toneurodifferentiation, rather than cytotoxicity (Slotkin et al.,2014). This inference was reinforced in the current study by ourfinding that TSE uniformly enhanced cell growth (Figs. 2D, 3D, 6D,7D), whereas cytotoxicity would be expected to impair growth. Wealso looked for lipid peroxidation consequent to oxidative stress asa likely mechanism independent of nicotinic receptor activationbut failed to find any increase in MDA levels that would indicatesuch an effect. In fact, we found reductions in MDA at high TSE

concentrations, regardless of differentiation state. Hence, a mainconclusion of this study is that the effects of TSE comprise multiplemechanisms over and above any contribution of nicotine acting oncholinergic receptors, or of lipid peroxidation.

In addition to distinguishing between effects of TSE andnicotine on neural cell replication and differentiation, the centralissue explored in this study was to investigate whether identifyingthe underlying mechanisms could be used to design ameliorationstrategies. Here, we specifically chose PC12 cells because theirinsensitivity to nicotine would enable us to pursue the actions ofother TSE constituents operating through mechanisms over andabove nicotinic receptor activation (Abreu-Villaça et al., 2005;Avila et al., 2003; Gueorguiev et al. 2000). Thus, it is not surprisingthat a nicotinic receptor antagonist (mecamylamine) was ineffec-tive in preventing TSE effects on either undifferentiated ordifferentiating cells. Indeed, this can be taken as confirmationthat the observed effects in this particular model reflect thecontributions of TSE constituents other than nicotine. Unexpect-edly, though, we found that mecamylamine augmented the abilityof TSE to induce TH activity, indicating a synergistic promotion ofdifferentiation into the dopaminergic phenotype. This likelyreflects the fact that nicotinic receptor activation by endogenousacetylcholine promotes differentiation into the opposite, cholin-ergic phenotype (Avila et al., 2003), and in that regard, receptorstimulation can act as a substitute for nerve growth factor(Yamashita and Nakamura, 1996). The main point is that, ratherthan ameliorating the effect of TSE, mecamylamine actuallyworsened the imbalance of neurotransmitter phenotypes bypreventing the emergence of the cholinergic phenotype, super-imposed on the ability of TSE to promote the dopaminergicphenotype. As will be reinforced below, when ameliorationstrategies instead exacerbate the effects of TSE, or themselvescause abnormalities, they may be harmful instead of helpful.

In contrast to the ineffectiveness of mecamylamine inameliorating the effects of TSE, antioxidants produced clear-cut,but only partial protection, reducing the impact on DNA synthesisand content in undifferentiated cells, and on DNA content andprotein/DNA ratio in differentiating cells. On the other hand,antioxidants provided no protection whatsoever against the shiftin neurotransmitter phenotypes elicited by TSE and actuallyexacerbated the decrement in ChAT. This failure cannot beattributed just to an inadequate antioxidant dose, since thisconcentration of vitamin C alone protects against oxidativedamage caused by monovalent silver (Powers et al., 2010b), whichwe found here to have a much larger oxidative effect than TSE.Perhaps even more troubling, the antioxidants by themselves hadunexpected effects, decreasing the number of cells in thedifferentiated state and grossly shifting neurodifferentiationtoward the cholinergic phenotype at the expense of thedopaminergic phenotype. Analysis of the separate antioxidantcomponents showed a hierarchy of effects on neurodifferentiation,with NAC producing the greatest shift away from the dopaminergicphenotype and toward the cholinergic phenotype. Vitamin C alsoaffected TH but not ChAT, and the addition of vitamins C and Eaugmented the effect of NAC. This probably reflects the fact thatoxidative events play an obligatory role in neurodifferentiation(Katoh et al., 1997; Panchision, 2009; Vieira et al., 2011), and thus,excessive antioxidant treatment can perturb normal developmen-tal processes.

Even partial protection against TSE by antioxidants seemedpuzzling, since we were unable to demonstrate significant lipidperoxidation in either undifferentiated or differentiating cells, andactually found a reduction in MDA levels at the higher TSEconcentration. Nicotine alone causes oxidative stress and lipidperoxidation in PC12 cells (Qiao et al., 2005), so the current findingindicates that some additional components of TSE actually quench

Fig. 7. Effects of a combined antioxidants and methyl donors on the response to TSE in the undifferentiated state (24 h exposure) and during differentiation (6 day exposure):(A) DNA synthesis in undifferentiated cells, (B) DNA content in undifferentiated cells, (C) DNA content in differentiating cells, (D) protein/DNA ratio in differentiating cells, (E)tyrosine hydroxylase activity in differentiating cells, and (F) choline acetyltransferase activity in differentiating cells. Data represent means and standard errors of the numberof determinations shown in parentheses. MANOVAs above each set of panels combine all the measurements for each differentiation state, and two-factor ANOVAs are shownat the top of each panel for each individual measure. Where there was an [antioxidants + methyl donors] � TSE interaction, asterisks denote whether the TSE group issignificantly different from the corresponding control value and daggers indicate whether the antioxidants alone are different from the vehicle control. Abbreviation: NS, notsignificant.

72 T.A. Slotkin et al. / Toxicology 333 (2015) 63–75

T.A. Slotkin et al. / Toxicology 333 (2015) 63–75 73

that activity. Interestingly, though, much higher concentrations oftobacco extract (not combusted) can produce oxidative stress(Yildiz et al., 1999), so the net response can be in either directiondepending on dose, a not-unexpected consequence from acomplex chemical mixture. It is also possible that the MDAendpoint is insufficiently sensitive, since oxidative damage canoccur short of producing measurable accumulation of lipidperoxides. In particular, the high membrane lipid turnover inrapidly-growing cells could dilute effects on MDA (Araki andWurtman 1997), although that would hardly explain the decreaseseen at the high TSE concentration. Future studies focusing onmore sensitive endpoints, such as generation of reactive oxygenspecies or mitochondrial dysfunction, should be able to resolve thisissue. Certainly, the fact that antioxidants provide partial protec-tion argues in favor of oxidative damage that is not detected withthe MDA method. Nevertheless, our results suggest that, at theconcentrations used here, oxidative stress is at best, a minorcontributor to the net effects on neurodifferentiation.

In any case, though, our results for MDA and for antioxidantamelioration are surprising in light of the clear-cut oxidative stressexperienced in vivo with maternal smoking (Aycicek and Ipek,2008; Fayol et al., 2005; Orhon et al., 2009; Rossner et al., 2009),and the protection by antioxidants against fetal lung damagecaused by maternal tobacco exposure (McEvoy et al., 2014;Proskocil et al., 2005). In that regard, TSE does not include volatiletobacco-smoke chemicals like HCN and CO that interfere with fetaloxygen delivery and utilization (Carmines and Rajendran, 2008;Pettigrew et al., 1977; Robkin, 1997), nor does an in vitro modelincorporate hypoxic stress resulting from uteroplacental vasocon-striction and reduced placental oxygen transfer, which are majorcontributors to fetal hypoxia in vivo (Albuquerque et al., 2004;Bush et al., 2000; Clark and Irion, 1992). Accordingly, the resultsseen here may underestimate the contribution played by oxidativestress for prenatal tobacco smoke exposure in vivo.

Notwithstanding these limitations, the inability of antioxidantsto protect against the shift in neurotransmitter phenotype elicited byTSE indicates that this particular effect is unrelated to oxidativestress. A number of other oxidative stressors produce a similar shifttoward the dopaminergic phenotype in the PC12 model: fipronil,chlorpyrifos, diazinon, parathion, dieldrin, monovalent silver andsilver nanoparticles (Lassiter et al., 2009; Powers et al., 2010a,b;Slotkin et al., 2007b). However, divalent nickel, which is a reductant,also increases TH at the expense of ChAT (Slotkin et al., 2007b). It isthus obvious that multiple mechanisms can affect the balance ofdopaminergic and cholinergic phenotypes, over and above con-tributions from oxidative stress or nicotinic receptor stimulation. Asseen here, TSE elicited a profound shift despite its failure to increaselipid peroxidation, and the effects were not prevented by eitherantioxidantsormecamylamine.Alternativemechanismsare likely toinclude excitotoxicity triggered by changes in glutamate receptors(Slotkin et al., 2010; Slotkin and Seidler, 2009) and stimulationof protein kinase signaling (Adigun et al., 2010). Indeed, wehave already shown that combined exposure to two TSE products,benzo[a]pyrene and nicotine, alters neurotransmitter phenotypes ina manner unlike those of either compound given separately (Slotkinet al., 2013), pointing to complex interactions mediated by multiplemechanisms.

In contrast to the antioxidants, methyl donors provided nodiscernible protection against any of the effects of TSE, whether onundifferentiated or differentiating cells. The only significant effectwas a methyl donor-induced increase in DNA in differentiatingcells, which then led to a relative exacerbation of the loss caused byTSE. Combining the methyl donors with antioxidants producedresults resembling those of the antioxidants alone, with oneexception, namely a synergistic interaction in differentiating cells,promoting the protection against the increase in cell size evoked by

TSE. Although we did not dissect which component of the methyldonor mixture was responsible for this interaction, the mainfinding for methyl donors is their relative ineffectiveness inoffsetting the effects of TSE on neural cell replication ordifferentiation.

Any in vitro approach to developmental neurotoxicity has clearlimitations (Coecke et al., 2007; Qiao et al., 2001; Song et al., 1998).In this case, we took advantage of the insensitivity of PC12 cells tonicotine (Abreu-Villaça et al., 2005; Avila et al., 2003; Gueorguievet al., 2000), in order to evaluate additional mechanisms by whichTSE affects neurodifferentiation, providing cause-and-effect rela-tionships that would be much more difficult to establish in vivo.Notably, culture systems cannot model cell-to-cell interactions,architectural modeling of brain regions or more complex events inbrain assembly, and consequently, sensitivity is likely to be greaterin vivo than in vitro. The difference in sensitivity is exacerbated bythe necessity to detect effects within a short span of hours to daysin vitro, vs. weeks of exposure in vivo (Coecke et al., 2007). This isespecially true for transformed cell lines, such as PC12 cells, whichare less sensitive to toxicants than are primary neurons. However,for toxicants that act on the cell cycle and that target specificneurodifferentiation endpoints, primary neurons are problematic,since they do not divide in culture and undergo heterogeneousneurodifferentiation, whereas PC12 cells undergo mitosis anddifferentiate uniformly (Coecke et al., 2007; Radio et al., 2008).Most importantly, though, culture models do not addressmaternal-fetal pharmacokinetics. Although we can use culturemodels to identify the direct effects of TSE and some of itsunderlying mechanisms, we do not yet know which TSEcomponents pass through the placenta to enter the fetus, northeir relative concentrations in the fetal compartment. According-ly, the current approach points to the potential neurodevelop-mental effects that should be pursued with in vivo studies; theseare underway in our laboratory.

5. Conclusions

There are several important conclusions from the presentresults. First, although nicotine is a major contributor to theneurodevelopmental damage associated with maternal smokingduring pregnancy (Pauly and Slotkin, 2008; Slikker et al., 2005;Slotkin, 2004, 2008), the thousands of other components intobacco smoke are likely to play a part. This is especially criticalwith regard to second-hand smoke exposure, which could then bemore injurious than anticipated from measured levels of nicotineor its metabolites (DeLorenze et al., 2002; Koren et al., 1998; Lucket al., 1985). At the same time, the complex composition of tobaccosmoke renders the identification of specific underlying mecha-nisms problematic. Even a combination of just two components,nicotine and benzo[a]pyrene, interact to produce effects onneurodifferentiation unlike either agent alone (Slotkin et al.,2013). In the current study, we were unable to reverse the effects ofTSE on neurodifferentiation by using a nicotinic receptor antago-nist, or through use of antioxidants and methyl donors separatelyor together. At best, we saw only a partial protection byantioxidants against cell loss, and no amelioration of adverseeffects on emergence of neurotransmitter phenotypes. Even worse,we found that ameliorant strategies themselves had adverseeffects on neurodifferentiation, likely because of their interferencewith the critical roles of endogenous cholinergic stimulation andredox status in normal neuronal cell development (Dreyfus, 1998;Hohmann, 2003; Katoh et al., 1997; Lauder, 1985; Panchision,2009; Vieira et al., 2011). This becomes crucial in attempting tooffset the adverse effects of maternal smoking through inter-ventions targeting putative mechanisms individually and withoutregard for heterogeneity of responses in different tissues. For

74 T.A. Slotkin et al. / Toxicology 333 (2015) 63–75

example, recent studies have recommended maternal vitamin Csupplementation to protect lung development and respiratoryfunction in offspring of smokers (McEvoy et al., 2014; Proskocilet al., 2005). However, this beneficial result is likely to be offset byadverse effects on brain development, both from augmentation ofnicotine levels in the fetus (Slotkin et al., 2005) and frominterference with natural oxidative signals involved in neuro-differentiation, as seen here. Ultimately, there is no betterapproach than to limit, or preferentially eliminate, prenataltobacco smoke exposure.

Conflict of interest statement

TAS has received consultant income in the past three years fromthe following firms: Acorda Therapeutics (Ardsley NY), The CalwellPractice (Charleston WV), Carter Law (Peoria IL), Gutglass EricksonBonville & Larson (Madison WI), Alexander Hawes (San Jose, CA),Pardieck Law (Seymour, IN), Tummel & Casso (Edinburg, TX), andChaperone Therapeutics (Research Triangle Park, NC).

Acknowledgments

Research was supported by NIHES022831 and EPA83543701.The authors thanks Ashley Stadler for technical assistance. EPAsupport does not signify that the contents reflect the views of theAgency, nor does mention of trade names or commercial productsconstitute endorsement or recommendation for use.

References

Abbott, L.C., Winzer-Serhan, U.H., 2012. Smoking during pregnancy: lessons learnedfrom epidemiological studies and experimental studies using animal models.Crit. Rev. Toxicol. 42, 279–303.

Abreu-Villaça, Y., Seidler, F.J., Qiao, D., Slotkin, T.A., 2005. Modeling thedevelopmental neurotoxicity of nicotine in vitro: cell acquisition, growth andviability in PC12 cells. Dev. Brain Res. 154, 239–246.

Adigun, A.A., Seidler, F.J., Slotkin, T.A., 2010. Disparate developmentalneurotoxicants converge on the cyclic AMP signaling cascade, revealed bytranscriptional profiles in vitro and in vivo. Brain Res. 1316, 1–16.

Albuquerque, C.A., Smith, K.R., Johnson, C., Chao, R., Harding, R., 2004. Influence ofmaternal tobacco smoking during pregnancy on uterine, umbilical and fetalcerebral artery blood flows. Early Hum. Dev. 80, 31–42.

Araki, W., Wurtman, R.J., 1997. Control of membrane phosphatidylcholinebiosynthesis by diacylglycerol levels in neuronal cells undergoing neuriteoutgrowth. Proc. Natl. Acad. Sci. 94, 11946–11950.

Avila, A.M., Davila-Garcia, M.I., Ascarrunz, V.S., Xiao, Y., Kellar, K.J., 2003. Differentialregulation of nicotinic acetylcholine receptors in PC12 cells by nicotine andnerve growth factor. Mol. Pharmacol. 64, 974–986.

Aycicek, A., Ipek, A., 2008. Maternal active or passive smoking causes oxidativestress in cord blood. Eur. J. Pediatr. 167, 81–85.

Bell, J.M., Whitmore, W.L., Slotkin, T.A., 1986. Effects of a-difluoromethylornithine, aspecific irreversible inhibitor of ornithine decarboxylase, on nucleic acids andproteins in developing rat brain: critical perinatal periods for regionalselectivity. Neuroscience 17, 399–407.

Boeke, C.E., Gillman, M.W., Hughes, M.D., Rifas-Shimanh, S.L., Villamor, E., Oken, E.,2013. Choline intake during pregnancy and child cognition at age 7 years. Am. J.Epidemiol. 177, 1338–1347.

Bruijnzeel, A.W., Bauzo, R.M., Munikoti, V., Rodrick, G.B., Yamada, H., Fornal, C.A.,Ormerod, B.K., Jacobs, B.L., 2011. Tobacco smoke diminishes neurogenesis andpromotes gliogenesis in the dentate gyrus of adolescent rats. Brain Res. 1413,32–42.

Bush, P.G., Mayhew, T.M., Abramovich, D.R., Aggett, P.J., Burke, M.D., Page, K.R., 2000.Maternal cigarette smoking and oxygen diffusion across the placenta. Placenta21, 824–833.

Carmines, E.L., Rajendran, N., 2008. Evidence for carbon monoxide as the majorfactor contributing to lower fetal weights in rats exposed to cigarette smoke.Toxicol. Sci. 102, 383–391.

Clark, K.E., Irion, G.L., 1992. Fetal hemodynamic response to maternal intravenousnicotine administration. Am. J. Obstet. Gynecol. 167, 1624–1631.

Coecke, S., Goldberg, A.M., Allen, S., Buzanska, L., Calamandrei, G., Crofton, K.,Hareng, L., Hartung, T., Knaut, H., Honegger, P., Jacobs, M., Lein, P., Li, A., Mundy,W., Owen, D., Schneider, S., Silbergeld, E., Reum, T., Trnovec, T., Monnet-Tschudi,F., Bal-Price, A., 2007. Workshop report: incorporating in vitro alternativemethods for developmental neurotoxicity into international hazard and riskassessment strategies. Environ. Health Perspect. 115, 924–931.

Cornelius, M.D., Day, N.L., 2009. Developmental consequences of prenatal tobaccoexposure. Curr. Opin. Neurol. 22, 121–125.

Costa, L.G., 1998. Neurotoxicity testing: a discussion of in vitro alternatives. Environ.Health Perspect. 106 (Suppl. 2), 505–510.

DeLorenze, G.N., Kharrazi, M., Kaufman, F.L., Eskenazi, B., Bernert, J.T., 2002.Exposure to environmental tobacco smoke in pregnant women: the associationbetween self-report and serum cotinine. Environ. Res. 90, 21–32.

DiFranza, J.R., Lew, R.A., 1995. Effect of maternal cigarette smoking on pregnancycomplications and sudden infant death syndrome. J. Family Pract. 40,385–394.

Dreyfus, C.F., 1998. Neurotransmitters and neurotrophins collaborate to influencebrain development. Perspect. Dev. Neurobiol. 5, 389–399.

Fayol, L., Gulian, J.M., Dalmasso, C., Calaf, R., Simeoni, U., Millet, V., 2005. Antioxidantstatus of neonates exposed in utero to tobacco smoke. Biol. Neonate 87, 121–126.

Fujita, K., Lazarovici, P., Guroff, G., 1989. Regulation of the differentiation ofPC12 pheochromocytoma cells. Environ. Health Perspect. 80, 127–142.

Fuller, B.F., Cortes, D.F., Landis, M.K., Yohannes, H., Griffin, H.E., Stafflinger, J.E.,Bowers, M.S., Lewis, M.H., Fox, M.A., Ottens, A.K., 2012. Exposure of rats toenvironmental tobacco smoke during cerebellar development alters behaviorand perturbs mitochondrial energetics. Environ. Health Perspect. 120, 1684–1691.

Gaysina, D., Fergusson, D.M., Leve, L.D., Horwood, J., Reiss, D., Shaw, D.S., Elam, K.K.,Natsuaki, M.N., Neiderhiser, J.M., Harold, G.T., 2013. Maternal smoking duringpregnancy and offspring conduct problems: evidence from three independentgenetically-sensitive research designs. JAMA Psychiatry 70, 956–963.

Gitto, E., Reiter, R.J., Karbownik, M., Tan, D.X., Gitto, P., Barberi, S., Barberi, I., 2002.Causes of oxidative stress in the pre- and perinatal period. Biol. Neonate 81,146–157.

Golub, M.S., Slotkin, T.A., Pinkerton, K.E., 2007. Visual recognition memory andauditory brainstem response in infant rhesus monkeys exposed perinatally toenvironmental tobacco smoke. Brain Res. 1151, 102–106.

Gospe Jr., S.M., Joyce, J.A., Siebert, J.R., Jack, R.M., Pinkerton, K.E., 2009. Exposure toenvironmental tobacco smoke during pregnancy in rats yields less effect onindices of brain cell number and size than does postnatal exposure. Reprod.Toxicol. 27, 22–27.

Guan, Z.Z., Yu, W.F., Nordberg, A., 2003. Dual effects of nicotine on oxidative stressand neuroprotection in PC12 cells. Neurochem. Int. 43, 243–249.

Gueorguiev, V.D., Zeman, R.J., Meyer, E.M., Sabban, E.L., 2000. Involvement ofa7 nicotinic acetylcholine receptors in activation of tyrosine hydroxylase anddopamine b-hydroxylase gene expression in PC12 cells. J. Neurochem. 75,1997–2005.

Hohmann, C.F., 2003. A morphogenetic role for acetylcholine in mouse cerebralneocortex. Neurosci. Biobehav. Rev. 27, 351–363.

Jameson, R.R., Seidler, F.J., Qiao, D., Slotkin, T.A., 2006a. Adverseneurodevelopmental effects of dexamethasone modeled in PC12 cells:identifying the critical stages and concentration thresholds for the targeting ofcell acquisition, differentiation and viability. Neuropsychopharmacology 31,1647–1658.

Jameson, R.R., Seidler, F.J., Qiao, D., Slotkin, T.A., 2006b. Chlorpyrifos affectsphenotypic outcomes in a model of mammalian neurodevelopment: criticalstages targeting differentiation in PC12 cells. Environ. Health Perspect. 114,667–672.

Katoh, S., Mitsui, Y., Kitani, K., Suzuki, T., 1997. Hyperoxia induces the differentiatedneuronal phenotype of PC12 cells by producing reactive oxygen species.Biochem. Biophys. Res. Commun. 241, 347–351.

Knapp, S., Wurtman, R.J., 1999. Enhancement of free fatty acid incorporation intophospholipids by choline plus cytidine. Brain Res. 822, 52–59.

Koren, G., Eliopoulos, C., Klein, J., 1998. Measuring fetal exposure to nicotine. In:Benowitz, N.L. (Ed.), Nicotine Safety and Toxicity. Oxford University Press, NewYork, pp. 99–106.

Lassiter, T.L., MacKillop, E.A., Ryde, I.T., Seidler, F.J., Slotkin, T.A., 2009. Is fipronil saferthan chlorpyrifos? Comparative developmental neurotoxicity modeled inPC12 cells. Brain Res. Bull. 78, 313–322.

Lauder, J.M., 1985. Roles for neurotransmitters in development: possible interactionwith drugs during the fetal and neonatal periods. In: Marois, M. (Ed.),Prevention of Physical and Mental Congenital Defects. Alan R. Liss, New York,pp. 375–380.

Lee, J.E., Park, J.H., Shin, I.C., Koh, H.C., 2012. Reactive oxygen species regulatedmitochondria-mediated apoptosis in PC12 cells exposed to chlorpyrifos.Toxicol. Appl. Pharmacol. 263, 148–162.

Lobo Torres, L.H., Moreira, W.L., Tamborelli Garcia, R.C., Annoni, R., NicolettiCarvalho, A.L., Teixeira, S.A., Pacheco-Neto, M., Muscara, M.N., Camarini, R., deMelo Loureiro, A.P., Yonamine, M., Mauad, T., Marcourakis, T., 2012.Environmental tobacco smoke induces oxidative stress in distinct brain regionsof infant mice. J. Toxicol. Environ. Health A 75, 971–980.

Luck, W., Nau, H., Hansen, R., Steldinger, R., 1985. Extent of nicotine and cotininetransfer to the human fetus, placenta and amniotic fluid of smoking mothers.Dev. Pharmacol. Ther. 8, 384–395.

McEvoy, C.T., Schilling, D., Clay, N., Jackson, K., Go, M.D., Spitale, P., Bunten, C., Leiva,M., Gonzales, D., Hollister-Smith, J., Durand, M., Frei, B., Buist, A.S., Peters, D.,Morris, C.D., Spindel, E.R., 2014. Vitamin C supplementation for pregnantsmoking women and pulmonary function in their newborn infants: arandomized clinical trial. J. Am. Med. Assoc. 311, 2074–2082.

Orhon, F.S., Ulukol, B., Kahya, D., Cengiz, B., Baskan, S., Tezcan, S., 2009. The influenceof maternal smoking on maternal and newborn oxidant and antioxidant status.Eur. J. Pediatr. 168, 975–981.

Panchision, D.M., 2009. The role of oxygen in regulating neural stem cells indevelopment and disease. J. Cell Physiol. 220, 562–568.

T.A. Slotkin et al. / Toxicology 333 (2015) 63–75 75

Pauly, J.R., Slotkin, T.A., 2008. Maternal tobacco smoking, nicotine replacement andneurobehavioural development. Acta Paediatr. 97, 1331–1337.

Pettigrew, A.R., Logan, R.W., Willocks, J., 1977. Smoking in pregnancy: effect on birthweight and on cyanide and thiocyanate levels in mother and baby. Br. J. Obstet.Gynaecol. 84, 31–34.

Powers, C.M., Badireddy, A.R., Ryde, I.T., Seidler, F.J., Slotkin, T.A., 2010a. Silvernanoparticles compromise neurodevelopment in PC12 cells: criticalcontributions of silver ion, particle size, coating and composition. Environ.Health Perspect. 119, 37–44.

Powers, C.M., Wrench, N., Ryde, I.T., Smith, A.M., Seidler, F.J., Slotkin, T.A., 2010b.Silver impairs neurodevelopment: studies in PC12 cells. Environ. HealthPerspect. 118, 73–79.

Proskocil, B.J., Sekhon, H.S., Clark, J.A., Lupo, S.L., Jia, Y., Hull, W.M., Whitsett, J.A.,Starcher, B.C., Spindel, E.R., 2005. Vitamin C prevents the effects of prenatalnicotine on pulmonary function in newborn monkeys. Am. J. Respir. Crit. CareMed. 171, 1032–1039.

Qiao, D., Seidler, F.J., Slotkin, T.A., 2001. Developmental neurotoxicity of chlorpyrifosmodeled in vitro: comparative effects of metabolites and other cholinesteraseinhibitors on DNA synthesis in PC12 and C6 cells. Environ. Health Perspect. 109,909–913.

Qiao, D., Seidler, F.J., Slotkin, T.A., 2005. Oxidative mechanisms contributing to thedevelopmental neurotoxicity of nicotine and chlorpyrifos. Toxicol. Appl.Pharmacol. 206, 17–26.

Qiao, D., Seidler, F.J., Violin, J.D., Slotkin, T.A., 2003. Nicotine is a developmentalneurotoxicant and neuroprotectant: stage-selective inhibition of DNA synthesiscoincident with shielding from effects of chlorpyrifos. Dev. Brain Res. 147, 183–190.

Radio, N.M., Breier, J.M., Shafer, T.J., Mundy, W.R., 2008. Assessment of chemicaleffects on neurite outgrowth in PC12 cells using high content screening. Toxicol.Sci. 105, 106–118.

Robkin, M.A., 1997. Carbon monoxide and the embryo. Int. J. Dev. Biol. 41, 283–289.Rossner Jr., P., Milcova, A., Libalova, H., Novakova, Z., Topinka, J., Balascak, I., Sram, R.

J., 2009. Biomarkers of exposure to tobacco smoke and environmentalpollutants in mothers and their transplacental transfer to the foetus. Part II.Oxidative damage. Mutation Res. 669, 20–26.

Savelkoul, P.J.M., Janickova, H., Kuipers, A.A.M., Hageman, R.J.J., Kamphuis, P.J.,Dolezal, V., Broerson, L.M., 2012. A specific multi-nutrient formulation enhancesM1 muscarinic acetylcholine receptor responses in vitro. J. Neurochem. 120,631–640.

Sekizawa, S., Chen, C.Y., Bechtold, A.G., Tabor, J.M., Bric, J.M., Pinkerton, K.E., Joad, J.P.,Bonham, A.C., 2008. Extended second-hand tobacco smoke exposure inducesplasticity in nucleus tractus solitarius second-order lung afferent neurons inyoung guinea pigs. Eur. J. Neurosci. 28, 771–781.

Slikker, W., Xu, Z.A., Levin, E.D., Slotkin, T.A., 2005. Mode of action: disruption ofbrain cell replication, second messenger, and neurotransmitter systems duringdevelopment leading to cognitive dysfunction — developmental neurotoxicityof nicotine. Crit. Rev. Toxicol. 35, 703–711.

Slotkin, T.A., 2004. Cholinergic systems in brain development and disruption byneurotoxicants: nicotine, environmental tobacco smoke, organophosphates.Toxicol. Appl. Pharmacol. 198, 132–151.

Slotkin, T.A., 2008. If nicotine is a developmental neurotoxicant in animal studies,dare we recommend nicotine replacement therapy in pregnant women andadolescents? Neurotoxicol. Teratol. 30, 1–19.

Slotkin, T.A., Card, J., Seidler, F.J., 2013. Adverse benzo[a]pyrene effects onneurodifferentiation are altered by other neurotoxicant coexposures:interactions with dexamethasone, chlorpyrifos, or nicotine in PC12 cells.Environ. Health Perspect. 121, 825–831.

Slotkin, T.A., Card, J., Stadler, A., Levin, E.D., Seidler, F.J., 2014. Effects of tobaccosmoke on PC12 cell neurodifferentiation are distinct from those of nicotine orbenzo[a]pyrene. Neurotoxicol. Teratol. 43, 19–24.

Slotkin, T.A., Lobner, D., Seidler, F.J., 2010. Transcriptional profiles for glutamatetransporters reveal differences between organophosphates but similaritieswith unrelated neurotoxicants. Brain Res. Bull. 83, 76–83.

Slotkin, T.A., MacKillop, E.A., Ryde, I.T., Seidler, F.J., 2007a. Ameliorating thedevelopmental neurotoxicity of chlorpyrifos: a mechanisms-based approach inPC12 cells. Environ. Health Perspect. 115, 1306–1313.

Slotkin, T.A., MacKillop, E.A., Ryde, I.T., Tate, C.A., Seidler, F.J., 2007b. Screening fordevelopmental neurotoxicity using PC12 cells: comparisons oforganophosphates with a carbamate, an organochlorine and divalent nickel.Environ. Health Perspect. 115, 93–101.

Slotkin, T.A., Persons, D., Slepetis, R.J., Taylor, D., Bartolome, J., 1984. Control ofnucleic acid and protein synthesis in developing brain, kidney, and heart of theneonatal rat: effects of a-difluoromethylornithine, a specific, irreversibleinhibitor of ornithine decarboxylase. Teratology 30, 211–224.

Slotkin, T.A., Pinkerton, K.E., Seidler, F.J., 2006a. Perinatal environmental tobaccosmoke exposure in Rhesus monkeys: critical periods and regional selectivity foreffects on brain cell development and lipid peroxidation. Environ. HealthPerspect. 114, 34–39.

Slotkin, T.A., Pinkerton, K.E., Tate, C.A., Seidler, F.J., 2006b. Alterations of serotoninsynaptic proteins in brain regions of neonatal Rhesus monkeys exposed toperinatal environmental tobacco smoke. Brain Res. 1111, 30–35.

Slotkin, T.A., Seidler, F.J., 2009. Oxidative and excitatory mechanisms ofdevelopmental neurotoxicity: transcriptional profiles for chlorpyrifos, diazinon,dieldrin and divalent nickel in PC12 cells. Environ. Health Perspect. 117, 587–596.

Slotkin, T.A., Seidler, F.J., 2010. Oxidative stress from diverse developmentalneurotoxicants: antioxidants protect against lipid peroxidation withoutpreventing cell loss. Neurotoxicol. Teratol. 32, 124–131.

Slotkin, T.A., Seidler, F.J., Qiao, D., Aldridge, J.E., Tate, C.A., Cousins, M.M., Proskocil, B.J., Sekhon, H.S., Clark, J.A., Lupo, S.L., Spindel, E.R., 2005. Effects of prenatalnicotine exposure on primate brain development and attempted ameliorationwith supplemental choline or vitamin C: neurotransmitter receptors, cellsignaling and cell development biomarkers in fetal brain regions of Rhesusmonkeys. Neuropsychopharmacology 30, 129–144.

Slotkin, T.A., Seidler, F.J., Spindel, E.R., 2011. Prenatal nicotine exposure in rhesusmonkeys compromises development of brainstem and cardiac monoaminepathways involved in perinatal adaptation and sudden infant death syndrome:amelioration by vitamin C. Neurotoxicol. Teratol. 33, 431–434.

Song, X., Violin, J.D., Seidler, F.J., Slotkin, T.A., 1998. Modeling the developmentalneurotoxicity of chlorpyrifos in vitro: macromolecule synthesis in PC12 cells.Toxicol. Appl. Pharmacol. 151, 182–191.

Steegers-Theunissen, R.P., Twigt, J., Pestinger, V., Sinclair, V., 2013. Thepericonceptional period, reproduction and long-term health of offspring: theimportance of one-carbon metabolism. Hum. Reprod. Update 19, 640–655.

Teng, K.K., Greene, L.A., 1994. Cultured PC 12 cells: a model for neuronal functionand differentiation. In: Celis, J.E. (Ed.), Cell Biology: A Laboratory Handbook.Academic Press, San Diego, pp. 218–224.

Vieira, H.A., Alves, P., Vercelli, A., 2011. Modulation of neuronal stem celldifferentiation by hypoxia and reactive oxygen species. Prog. Neurobiol. 93,444–455.

Wakschlag, L.S., Lahey, B.B., Loeber, R., Green, S.M., Gordon, R.A., Leventhal, B.L.,1997. Maternal smoking during pregnancy and the risk of conduct disorder inboys. Arch. Gen. Psychiat. 54, 670–676.

Winick, M., Noble, A., 1965. Quantitative changes in DNA, RNA and protein duringprenatal and postnatal growth in the rat. Dev. Biol. 12, 451–466.

Yamashita, H., Nakamura, S., 1996. Nicotine rescues PC12 cells from death inducedby nerve growth factor deprivation. Neurosci. Lett. 213, 145–147.

Yildiz, D., Liu, Y.S., Ercal, N., Armstrong, D.W., 1999. Comparison of pure nicotine-and smokeless tobacco extract-induced toxicities and oxidative stress. Arch.Environ. Contam. Toxicol. 37, 434–439.