The intergenerational transmission of developmental nicotine exposure-induced neurodevelopmental disorder-like phenotypes is modulated by the Chrna5 D397N polymorphism in adolescent mice

Abstract

Maternal tobacco smoking during pregnancy constitutes developmental nicotine exposure (DNE) and is associated with nicotine dependence and neurodevelopmental disorders in both children and grandchildren as well as animal models thereof. Genetic variants such as the CHRNA5 single nucleotide polymorphism (SNP) rs16969968, which leads to an aspartic acid to asparagine substitution at amino acid position 398 (D398N) in the alpha-5 nicotinic acetylcholine receptor subunit, can also confer risk for nicotine dependence and neurodevelopmental disorders in the absence of DNE. However, the degrees to which, the consequences of maternal smoking on offspring outcomesare influenced by genetic variants and interactions therewith are not well understood. Addressing this void in the literature, the present study utilizes a DNE mouse model engineered to possess the equivalent of the human D398N SNP in CHRNA5 (D397N SNP in mice) to assess how the N397 risk allele impacts the induction and intergenerational transmission of a range of neurodevelopmental disorder-related behavioral phenotypes in first- and second-generation DNE offspring. Results reveal that offspring possessing the N397 variant in the absence of DNE as well as DNE offspring and grandoffspring possessing theD397 variant exhibit analogous neurodevelopmental disorder-like phenotypes including hyperactivity, risk-taking behaviors, aberrant rhythmicity of activity, and enhanced nicotine consumption. DNE amplified these behavioral anomalies in first-generation N397 progeny, but the severity of DNE-evoked behavioral perturbations did not significantly differ between first-generation D397 and N397 DNE mice for any measure. Remarkably, the behavioral profiles of second-generation N397 DNE progeny closely resembled DNE-naive D397 mice, suggesting that the N397 variant may protect against the intergenerational transmission of DNE-induced neurodevelopmental disorder-like behaviors.

Introduction

Ten percent of women in the United States report smoking traditional cigarettes during pregnancy (Centers for Disease Control and Prevention, 2011; Oncken et al., 2017), and fourteen percent disclose electronic cigarette consumption during pregnancy (Oncken et al., 2017). The prevalence of maternal smoking is greater still in other countries as well as within various states and provinces, geographical regions, and socioeconomic strata within and without the United States (Lange et al., 2018). Worrisomely, epidemiological studies indicate that the majority of those queried misperceive that electronic cigarettes are a benign surrogate for traditional cigarettes, and this misunderstanding is most common among women of reproductive age (Kahr et al., 2015; Nguyen et al., 2016; McCubbin et al., 2017; Wagner et al., 2017; Whittington et al., 2018). Contrary to this misconception, the use of both traditional and electronic cigarettes during pregnancy, which constitutes developmental nicotine exposure (DNE), is associated with a plethora of fetal abnormalities such as pre-mature birth, low birth weight, and Sudden Infant Death Syndrome (Salihu and Wilson, 2007; Knopik et al., 2016a; Ernst et al., 2001; Linnett et al., 2003; Button et al., 2007). Along with hindering fetal development, DNE is linked to neurodevelopmental disorders including ADHD, autism, and schizophrenia in developing children (Ernst et al., 2001; Linnett et al., 2003; Button et al., 2007; Knopik, 2009; Knopik et al., 2016b; He et al., 2017; Huang et al., 2018; Marceau et al., 2018; Golding et al., 2017; Niemela et al., 2016; Quinn et al., 2017). Of additional relevance to public health, DNE-elicited enhancement of autism risk is transmitted to not only the children but also the grandchildren of maternal smokers (Golding et al., 2017).

DNE and neurodevelopmental disorders including ADHD, autism, and schizophrenia are associated with behavioral anomalies such as hyperactivity, impulsivity (risk-taking), sleep disturbances, aberrant circadian rest-activity rhythms, and predisposition to substance use disorders, particularly nicotine dependence (Baird et al., 2012; Van Veen et al., 2009; Imeraj et al., 2012; Bijlenga et al., 2013; American Psychiatric Association, 2013; Snitselaar et al., 2017; Ajarem et al., 1998; Pauly et al., 2004; Paz et al., 2007; Heath et al., 2010; Ernst et al., 2001; Amsterdam et al., 2018; Milberger et al., 1997; Lambert and Hartsough, 1998; Kollins et al., 2005; Hong et al., 2011; Rhodes et al., 2016; Schuch et al., 2016; Elkins et al., 2018; Ohi et al., 2019; Barr et al., 2006; Tang et al., 2016; Mallet et al., 2017; Sagud et al., 2018). Similarly, nicotine-dependent neurodevelopmentally-disordered individuals report increased severity of nicotine withdrawal (Gray et al., 2010; Prochaska et al., 2014; Bidwell et al., 2018; Linnett et al., 2003; Button et al., 2007; Knopik, 2009; Knopik et al., 2016b; He et al., 2017; Huang et al., 2018; Cornelius et al., 2000; Buka et al., 2003; Agrawal et al., 2010). In support and extension of this literature, we recently reported that DNE precipitates multigenerational neurodevelopmental disorder-like nicotine preference, diurnal and nocturnal hyperactivity, aberrant circadian rhythmicity of activity, and increased risk-taking behaviors that are reversibly rescued by methylphenidate treatment and modulated by voluntary nicotine consumption (Buck et al., 2019). Taken together, the aforementioned evidence is consistent with the self-medication hypothesis for the predisposition to nicotine consumption in neurodevelopmental disorders, which postulates that a therapeutic effect of nicotine is responsible for the increased incidence of nicotine dependence in neurodevelopmentally-disordered individuals (Fang et al., 2019; McClernon and Kollins 2008; Lucatch et al., 2018; Logan et al., 2014; van Amsterdam et al., 2018; Kumari and Postma, 2005) and is additionally supported by research demonstrating that nicotine and other nicotinic compounds exert therapeutic-like effects in the treatment of inattention, impulsivity, and hyperactivity in ADHD, autism, schizophrenia, and animal models thereof (Gehricke et al., 2009; Arnold et al., 2012; Lippiello, 2006; Deutsch et al., 2015; Takechi et al., 2016; Koukouli et al., 2017; Amsterdam et al., 2018; Milberger et al., 1997; Lambert and Hartsough, 1998; Kollins et al., 2005; Hong et al., 2011; Rhodes et al., 2016; Schuch et al., 2016; Elkins et al., 2018; Ohi et al., 2019).

Studies in knockout mice suggest a role of Chrna5, the gene that encodes the nicotinic acetylcholine receptor alpha-5 subunit, in nicotine dependence (Morel et al., 2014;), neurodevelopment (Bailey et al. 2010; Tian et al., 2011) and in the effects of DNE on neurodevelopment (Bailey et al., 2014). Consistent with a role of CHRNA5 in both nicotine dependence and neurodevelopmental disorders, polymorphisms in CHRNA5, in particular the single nucleotide polymorphism (SNP) rs16969968, are associated with nicotine dependence (Bierut et al., 2008; Sarginson et al., 2011; Berrettini et al., 2012; Wen et al., 2016; Ohi et al., 2019) as well as neurodevelopmental disorders including ADHD (Schuch et al., 2016), autism (Forrest et al., 2018), and schizophrenia (Jackson et al., 2013; Forrest et al., 2018; Han et al., 2019; Ohi et al., 2019) in humans. Particularly relevant to DNE, the rs16969968 SNP is associated with increased smoking rates during pregnancy (Freathy et al., 2009).

Rodent models engineered to possess the rs16969968 CHRNA5 variant, which substitutes an asparagine (N) residue for an aspartic acid (D) residue at amino acid position 398 in humans (D398N) and at position 397 in rodents (D397N), support the role of this variant in nicotine dependence phenotypes (Morel et al., 2014) as well as behavioral and neurological phenotypes associated with neurodevelopmental disorders (Koukouli et al., 2017). Moreover, we recently demonstrated that DNE interacts with the N397 allele to both increase nicotine consumption and impair nicotine-evoked striatal dopamine release (O’Neill et al., 2018). These findings indicate that the D397N SNP interacts with DNE to influence behavior and neurochemistry in first-generation offspring and, coupled with our previous research demonstrating that DNE elicits multigenerational transmission of neurodevelopmental disorder-like phenotypes in wildtype C57BL/6J mice (Buck et al., 2019), warrant further research to determine whether the D397N SNP influences the intergenerational transmissibility of DNE-induced neurobehavioral anomalies.

In aggregate, previous studies suggest that the D398N SNP may interact with DNE to modulate the intergenerational transmission of neurodevelopmental and neurobehavioral deficits to the children and grandchildren of maternal smokers. Addressing this possibility, we herein characterized the multigenerational effects of DNE on activity in familiar and novel environments, the circadian rhythmicity of activity, risk-taking behaviors, voluntary nicotine consumption, and nicotine responsivity in adolescent D397N mice. We hypothesized that DNE and the D397N SNP would interact to mitigate the magnitude of DNE-induced hyperactivity, risk-taking behaviors, circadian rhythmic alterations, nicotine intake and preference, and behavioral responsivity to nicotine in first- and second-generation adolescent N397 DNE mice.

Materials & Methods

Reagents

Freebase nicotine and sodium saccharin were acquired from Sigma-Aldrich (St. Louis, MO) and ThermoFisher Scientific (Waltham, MA), respectively.

Animals

All housing and experimental conditions were approved by the Institutional Animal Care and Utilization Committee at the University of Colorado, Boulder and were compliant with the guidelines for animal care and use mandated by the NIH and the Guide for the Care and Use of Laboratory Animals (8th Ed.). Mice were maintained on a standard 12h light/dark cycle (lights on at 07:00), and food (Envigo Teklad 2914 irradiated rodent diet, Harlan, Madison, WI) and water were available ad libitum. Homozygous F1 and F2 D397N DNE mice were generated as previously described (Buck et al., 2019; O’Neill et al., 2018). Briefly, beginning thirty days prior to mating with nicotine-naïve sires, female D397N dams were administered 0.2% saccharin (vehicle) or 0.2% saccharin and nicotine (100 μg/mL) (DNE) in place of drinking water. All solutions were replaced bi-weekly. Treatment of breeders continued until weaning of progeny at PND 21, and water was provided as the sole fluid source for weaned offspring thereafter. At PND 60, female F1 DNE offspring were randomly selected for breeding with NIC-naïve male sires to produce F2 generation DNE offspring (Fig. 1). Pilot studies revealed no differences among F1 and F2 generation adolescent mice developmentally exposed to saccharin or plain water vehicle, and we therefore resolved to test only F1 developmental Veh-exposed mice as DNE-naïve controls. Behavioral testing of all mice commenced at PND 35, which corresponds to early adolescence in humans, a developmental stage that overlaps with the ages of onset of various neurodevelopmental disorders. For each group at least 15 total mixed-sex litters from at least 10 different breeder pairs were tested to control for between-litter and between-breeder differences within each group. Multifactorial ANOVAs revealed no covariation with breeder, litter, or season for any measure.

Figure 1. Procedural Schematics for breeding and behavioral testing.

(A) Procedural Schematic for breeding. Female C57BL/6J breeders were exposed to 0.2% saccharin (Veh) or 0.2% saccharin and nicotine (200 μg/mL) in drinking water for 30 days prior to mating with drug-naïve males. Veh or nicotine treatment of breeders continued until weaning of offspring at PND 21, after which water was provided to all offspring as the sole fluid source. Randomly selected F1 female developmental nicotine-exposed offspring were subsequently mated with drug-naïve male sires to obtain F2 generation maternal germline nicotine-exposed offspring. (B) Procedural timeline for nicotine administration in the four-bottle choice test and behavioral testing. Home cage activity (nF1Veh=30, nF1NIC=32, and nF2NIC=33) was recorded over 3d before (baseline, BL), over two continuous 4d intervals during (N1, N1 and N2, N2), and over 24h after (withdrawal, WD) voluntary nicotine intake in the four-bottle choice test (four-bottle choice test). Open field behavior (nF1Veh=30, nF1NIC=24, and nF2NIC=35) was tested before (BL), immediately following (nicotine, NIC), and 24h after (WD) voluntary nicotine intake. PND: post-natal day; Veh: 0.2% aqueous saccharin; nicotine: 200 μg/mL aqueous nicotine; F1 Veh: first-generation parental Veh-exposed offspring; F1 NIC: first-generation parental nicotine-exposed offspring; F2 NIC: second-generation parental nicotine-exposed offspring; four-bottle choice test: four-bottle choice test for voluntary oral nicotine intake and preference.

Importantly, it has previously been demonstrated that the 100 μg/mL oral nicotine regimen employed in the current study produces cotinine levels in C57BL/6J mice and their offspring similar to levels seen in habitual smokers and evokes neurobehavioral anomalies (Pauly et al., 2004; Zhu et al., 2012; Buck et al., 2019). It should further be noted that, once co-housed with pre-treated dams, nicotine-naïve sires were exposed to vehicle or nicotine drinking solutions. Thus, the DNE paradigm utilized for this study may be more appropriately described as parental rather than maternal DNE for DNE offspring, while DNE grandoffspring were exposed to nicotine exclusively via the maternal germline (oocytes).

Four-bottle choice test of voluntary nicotine intake and preference

Voluntary nicotine consumption and preference were assayed via a four-bottle choice test as previously described (Li et al., 2007; Buck et al., 2019). Briefly, solutions were supplied in Pyrex test tubes (9820, 18mm x 150 mm, Fisher Scientific) fitted with rubber septa (Suba-Seal septa 29; Sigma Aldrich) containing a standard 6.5mm stainless steel sipper tube. To enable concurrent top-down activity monitoring, nicotine and water bottles were placed in four equidistant holes fashioned in one side of conventional polycarbonate home cages, and food was placed directly in the cage bottom. The four-bottle choice test was conducted over two continuous four-day intervals (N1, N2) to assess intake from and preference for bottles containing water, 25 μg/mL nicotine, 50 μg/mL nicotine, and 100 μg/mL nicotine. Bottles were weighed before and after N1 and N2 and were rotated daily to eliminate confounding due to position bias. Daily nicotine consumption (mg/kg/day) was calculated as the outcome measure for voluntary nicotine intake, and preference ratios (volume consumed from each bottle divided by the total daily fluid consumed) were calculated as the outcome measure for nicotine preference at each concentration tested.

Passive home cage activity recordings

Home cage locomotor activity was recorded as previously described (Buck et al., 2019). Briefly, home cage activity of singly-housed mice was continuously monitored via cage top-mounted passive infrared motion sensors and VitalView Data Acquisition software (Starr Life Sciences Corporation, Oakmont, PA) during a 3-day baseline period and throughout each stage of four-bottle choice testing (N1, N2, and a 24-hour nicotine withdrawal period).

Open field testing

Locomotor activity in a novel environment and risk-taking behaviors were evaluated as previously described (Buck et al., 2019). Summarily, mice were subjected to 15-minute open field trials in a 40×40cm opaque plexiglass arena at baseline, immediately following the conclusion of the four-bottle choice test (NIC), and immediately following a 24-hour post-four-bottle choice test nicotine withdrawal period. Activity Monitor (Med Associates Inc., Fairfax, VT) was used for data collection and analysis. Total distance moved in the open field was evaluated as the outcome measure for locomotor activity in a novel environment. Risk-taking behavior was assessed by measuring the percent of total distance moved in the center 10×10cm area of the open field arena. Percent of total distance moved was used rather than center time in order to control for individual differences in OFA activity. It should be noted that although we interpret center time as a measure of risk-taking, center time also is often used/interpreted as a measure of anxiety.

Statistical Analyses

All datasets were initially analyzed by multifactoral ANOVA to identify possible main effects of sex or confounding by breeder or litter. Since no effects of sex, breeder, or litter were elucidated, these co-variates were removed and the collapsed data were analyzed by mixed effects ANOVAs as subsequently described. Prior to statistical analyses, all data were subjected to outlier screening via the ROUT test (Q=1%), and confirmed outliers were excluded from formal analyses. Where appropriate, Bonferroni’s multiple comparisons test was used to assess statistical significance (α = 0.05). Data analyses and visualization were conducted via R (https://cran.r-project.org), SPSS (IBM Analytics, Armonk, NY), and GraphPad Prism 7.04 (GraphPad Software, La Jolla, California, USA).

Baseline home cage locomotor activity data were analyzed by mixed ANOVA with the between-subjects factors group (F1 Veh_n=61, F1 NIC_n=63, or F2 NIC_n=73) and genotype (D397 or N397) and the within-subjects factor phase (active or inactive). Datasets for total distance moved and percent distance moved in the center of the open field were analyzed by two-way ANOVA with the between-subjects factors group (F1 Veh_n=59, F1 NIC_n=52, or F2 NIC_n=71) and genotype (D397 or N397).

Data for voluntary nicotine intake in the four-bottle choice test were analyzed by mixed ANOVA with the between-subjects factors group (F1 Veh_n=38, F1 NIC_n=33, or F2 NIC_n=46) and genotype (D397 or N397) and the within-subjects factor measure (total daily nicotine intake, preference for water, preference for 25μg/mL nicotine, preference for 50μg/mL nicotine, or preference for 100μg/mL nicotine.

To determine whether any behavioral effects of voluntary nicotine intake covaried with individual drug intake, home cage and open field data for four-bottle choice-tested mice were transformed to percentage differences from baseline. The resultant datasets were analyzed by repeated measures ANCOVA with nicotine intake as a covariate. Results indicated that the effects of nicotine on home cage activity, total distance moved in the open field, percent distance moved in the center of the open field, MESOR estimates, global amplitude estimates, and orthophase estimates did not significantly co-vary with individual nicotine intake (data not shown, but are available upon request), indicating that the behavioral effects of either drug were not dose-dependent. Therefore, unprocessed (non-transformed) datasets for four-bottle choice-tested mice were analyzed by mixed ANOVA as subsequently described.

Home cage activity of four-bottle choice-tested mice was analyzed by mixed ANOVA with the between-subjects factors group (F1 Veh_n=30, F1 NIC_n=32, or F2 NIC_n=33) and genotype (D397 or N397) and the within-subjects factors phase (active or inactive) and experimental stage (BL, N1, N2, or WD). Total distance moved and percent distance moved in the center of the open field were analyzed by mixed ANOVA with the between-subjects factors group (F1 Veh_n=30, F1 NIC_n=24, or F2 NIC_n=35) and genotype (D397 or N397) and the within-subjects factor experimental stage (BL, NIC, or WD).

Rhythmicity of home cage locomotor activity was assessed using a multi-harmonic cosinor model as previously described (Buck et al., 2019) and as detailed in the supplement (Fig. S1). Briefly, data for baseline, N1, N2, and withdrawal were collapsed into hourly bins and subjected to frequency decomposition via harmonic regression at Fourier frequencies. The fundamental period and each of its harmonics with significant non-zero amplitude that accounted for a significant proportion of the total variance (percentage of rhythm) were incorporated into the multi-harmonic cosinor model for each group at each experimental stage. The resulting non-linear regression models were then utilized to fit activity data separately for each mouse at each experimental stage. While this approach yielded parameter of rhythm estimates for the fundamental period and each of its harmonics, this transmittal exclusively reports global rhythmometric measures that parameterize composite home cage activity rhythms, namely the MESOR (midline of rhythm), global amplitude (0.5x peak – trough), and orthophase (latency to peak), as these metrics are more amenable to translational comparisons with available measures of activity rhythms in neurodevelopmentally-disordered and DNE children. A detailed description of the rhythmometric procedures and a schematization of the global parameters of rhythm utilized herein are provided in the supplement (Fig. S1). Representative multi-harmonic regression curve fits for four-bottle choice-tested mice are also provided in the supplement (Fig. S2). Parameter of rhythm estimates at baseline and in response to voluntary nicotine intake via the four-bottle choice were analyzed by mixed ANOVA with the between-subjects factors group (F1 Veh_n=30, F1 NIC_n=32, or F2 NIC_n=33) and genotype (D397 or N397) and the within-subjects factors experimental stage (baseline, N1, N2, or withdrawal) and measure (MESOR, global amplitude, or orthophase).

Results

The experimental procedures employed for the breeding of mice and the schedule of behavioral testing are depicted in Fig. 1 and detailed in the methods section. Reiteratively, dams designated as DF0 (F0 mice homozygous for the D397 variant of Chrna5) or NF0 (F0 mice homozygous for the N397 allele of Chrna5) received either nicotine (100 μg/mL in 0.2% saccharin) or vehicle (0.2% saccharin) as the sole fluid source beginning 30 days prior to mating and continuing until the weaning of pups. Both DF1 NIC (F1 mice homozygous for the D397 allele) and NF1 NIC (F1 mice homozygous for the N allele) mice were thereby exposed to vehicle and nicotine from conception until weaning, while DF1 Veh and NF1 Veh mice were exposed to vehicle alone from conception until weaning. DF2 NIC and NF2 NIC mice are the progeny of female DF1 NIC and NF1 NIC mice, respectively, bred with nicotine-naïve DF0 and NF0 male sires, respectively. Aside from the oocytes from which they were conceived, DF2 NIC and NF2 NIC mice were not directly exposed to nicotine. All behavioral testing commenced at PND 35 (early adolescence).

DNE and the D397N SNP interact to influence home cage activity, activity in a novel environment, and risk-taking behaviors among F1 and F2 D397N DNE mice

Baseline home cage activity was assayed during the active (dark phase) and inactive (light phase) phase to assess the effect of DNE, genotype and the potential interaction between these factors. Main effects of group (F_2,231=10.67; p=0.00004) and phase (F_1,231=3764.8; p=0.00003) and significant group x genotype (F_2,231=24.55; p=2.2e⁻¹⁰), group x phase (F_2,231=5.69; p=0.004), and group x genotype x phase (F_2,231=14.73; p=9.5e⁻⁷) were detected for baseline active (Fig. 2A) and inactive (Fig. 2B) phase home cage activity. Both DF1 NIC and DF2 NIC mice were more active than DF1 Veh mice during the active phase (p=0.0027 and p=0.000009, respectively) and the inactive phase (p=0.0054 and p=0.000005, respectively) at baseline. However, DNE did not appear to prevent or attenuate the home cage hyperactivity in NF1 NIC mice. NF1 NIC mice were more active than NF1 Veh mice during the inactive phases (p=0.027) and did not differ from NF1 Veh mice during the active phase. On the other hand, the hyperactive phenotype did not continue into the NF2 mice. NF2 NIC mice were less active than NF1 Veh, and NF1 NIC mice as well as DF2 mice during the active phase (p=0.0000006, p=0.0013, and p=0.0000008, respectively) and the inactive phase (p=0.031, p=0.0002, and p=0.000007, respectively) at baseline. Lastly, NF1 Veh mice were more active than DF1 Veh mice during the active phase (p=0.006) suggesting that the N397 variant alone leads to a hyperactive phenotype.

Figure 2. DNE and the D397N SNP interact to influence home cage activity, activity in a novel environment, and risk-taking behaviors among F1 and F2 D397N DNE mice.

Baseline (BL) data for home cage activity and open field behavior. (A) Baseline active phase home cage. DF1 NIC and DF2 NIC mice were more active than DF1 Veh mice during the active phase at baseline (p=0.0027 and p=0.000009, respectively), and DF2 NIC mice were more active than NF2 NIC mice during the active phase (p=0.0000008). NF1 Veh mice were more active than DF1 Veh mice during the active phase (p=0.006), and NF2 NIC mice were less active than both NF1 Veh (p=0.0013) and NF1 NIC (p=0.0000006) mice during the active phase. (B) Baseline inactive phase home cage activity. DF1 NIC and DF2 NIC mice were more active than DF1 Veh mice during the inactive phase at baseline (p=0.0054 and p=0.000005, respectively), and DF2 NIC mice were more active than NF2 NIC mice during the inactive phase (p=0.000007). NF1 NIC mice were more active than both NF1 Veh (p=0.027) and NF2 NIC (p=0.0002) mice during the inactive phase. NF2 NIC mice were less active than NF1 Veh mice during the inactive phase (p=0.031). (C) Total distance moved in a baseline open field trial. DF1 NIC and DF2 NIC mice were more active in the open field than DF1 Veh mice at baseline (p=0.0003 and p=0.000002, respectively), and DF2 NIC mice were more active in the open field than NF2 NIC mice (p=0.00009). NF1 Veh mice were more active in the open field than DF1 Veh mice (p=0.0039), and NF1 NIC mice were more active in the open field than NF1 Veh mice (p=0.047). NF2 NIC mice were less active in the open field than both NF1 Veh (p=0.038) and NF1 NIC (p=0.000001) mice. (D) Percent distance moved in the center of the open field at baseline. DF1 NIC and DF2 NIC mice moved a greater percentage of total distance in the center of the open field than DF1 Veh mice at baseline (p=0.014 and p=0.00001, respectively). NF1 Veh mice moved a greater percentage of total distance in the center of the open field than DF1 Veh mice (p=0.025), and NF1 NIC and NF2 NIC mice moved a greater percentage of total distance in the center of the open field than NF1 Veh mice (p=0.029 and p=0.009, respectively). (n_DF1Veh=61, n_DF1NIC=63, n_DF2NIC=73, n_NF1Veh=61, n_NF1NIC=63, and n_NF2NIC=73). All data are mean ± S.E.M. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

To assess the impact of DNE, genotype and the potential interaction between DNE and genotype on activity in a novel environment, baseline activity and risk-taking behaviors were measured in the open field arena. Main effects of group (F_2,264=11.47; p=0.00002) and measure (F_1,264=6927.0; p=0.00001) and significant group x genotype (F_2,264=19.21; p=1.6e⁻⁸), group x measure (F_2,264=11.39; p=0.00002), and group x genotype x measure (F_2,264=19.19; p=1.7e⁻⁸) interactions were detected for total distance moved (Fig. 2C) and percent distance moved in the center (Fig. 2D) in the open field. DF1 NIC and DF2 NIC mice were more active in the open field than DF1 Veh mice (p=0.0003 and p=0.000002, respectively) at baseline. Consistent with the home cage activity data, NF2 NIC mice were less active in the open field at baseline relative to NF1 Veh (p=0.038), NF1 NIC (p=0.000001) and DF2 NIC (p=0.00009) mice, and NF1 NIC mice were more active in the open field than NF1 Veh mice (p=0.047) but did not differ from DF1 NIC mice. Also comparable to the home cage activity results, NF1 Veh mice were more active in the open field than DF1 Veh mice (p=0.0039).

DF1 NIC and DF2 NIC mice moved a greater percentage of total distance in the center of the open field than DF1 Veh mice at baseline (p=0.014 and p=0.00001, respectively). In addition, NF1 NIC and NF2 NIC mice moved a greater percentage of total distance in the center of the open field at baseline relative to NF1 Veh mice (p=0.029 and p=0.009, respectively) and also did not differ from DF1 NIC or DF2 NIC mice. Again, NF1 veh mice differed from DF1 Veh mice as NF1 Veh mice moved a greater percentage of total distance in the center of the open field than DF1 Veh mice (p=0.025).

DNE and the D397N SNP interact to modulate the rhythmicity of home cage activity among F1 and F2 DNE mice

Rhythmometric analyses (Figs. 3A and 3B) were performed to assess the potential role of DNE, genotype and their potential interaction on daily rhythmicity in the home cage. To this end, we compared baseline composite parameters of rhythm including the MESOR (oscillatory mean), global amplitude (oscillatory range), and orthophase (oscillatory peak) of home cage activity. Main effects of group (F_2,226=5.73; p=0.004) and measure (F_2,226=3212.8; p=0.00002) and significant group x genotype (F_2,226=14.31; p=0.000001), group x measure (F_4,226=3.79; p=0.012), and group x genotype x measure (F_4,226=8.77; p=0.000001) interactions were detected for baseline MESOR (Fig. 3C), global amplitude (Fig. 3D), and orthophase (Fig. 3E) estimates.

Figure 3. DNE and the D397N SNP interactively influence the baseline rhythmicity of home cage activity among adolescent F1 and F2 D397N DNE mice.

Representative plots of average daily home cage activity over a 3-day baseline period collapsed into hourly bins (floating points depicting mean ± S.E.M) with superimposed multi-harmonic regression curve fits (mean ± 95% CI bands) for (A) D397 mice and (B) N397 mice. (C) Baseline MESOR estimates. DF1 NIC and DF2 NIC mice had increased MESOR estimates versus DF1 Veh mice at baseline (p=0.00009 and p=1.1e⁻¹⁰, respectively). NF1 Veh mice had increased MESOR estimates compared to DF1 Veh mice (p=0.011). NF2 NIC mice had decreased baseline MESOR estimates compared to DF2 NIC (p=2.5e⁻⁹), NF1 Veh (p=0.002), and NF1 NIC (p=0.000003) mice. (D) Baseline global amplitude estimates. DF1 NIC and DF2 NIC mice had increased global amplitude estimates versus DF1 Veh mice at baseline (p=0.004 and p=0.003, respectively), and DF2 NIC mice had increased global amplitude estimates versus NF2 NIC mice (p=0.005). NF1 Veh mice trended toward increased baseline global amplitude estimates compared to DF1 Veh mice (p=0.11). NF1 NIC mice had increased global amplitude estimates relative to NF2 NIC mice (p=0.001), and NF2 NIC mice had decreased global amplitude estimates versus NF1 Veh mice (p=0.025). (E) Baseline orthophase estimates. DF1 NIC and DF2 NIC mice had delayed orthophase estimates versus DF1 Veh mice at baseline (p=0.030 and p=0.001, respectively), and NF2 NIC mice had advanced orthophase estimates versus DF2 NIC mice (p=0.00002). NF1 Veh mice had delayed orthophase estimates compared to DF1 Veh mice (p=0.041). NF2 NIC mice had advanced orthophase estimates versus NF1 Veh and NF1 NIC mice (p=0.009 and p=0.006, respectively. n_DF1Veh=61, n_DF1NIC=63, n_DF2NIC=73, n_NF1Veh=61, n_NF1NIC=63, and n_NF2NIC=73. Baseline, 3 days pre-four-bottle choice test; MESOR, midline estimating statistic of rhythm (oscillatory mean); global amplitude, difference between peak and trough (oscillatory range); orthophase, clock time at (latency to) peak activity (oscillatory phase). Floating points depict mean ± S.E.M. Dotted bands around regression curves in (A) and (B) denote 95% confidence intervals. ^#p=0.11; *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

DF1 NIC and DF2 NIC mice exhibited increased baseline MESOR (p=0.00009 and p=1.1e⁻¹⁰, respectively), global amplitude (p=0.004 and p=0.003, respectively), and orthophase (p=0.030 and p=0.001, respectively) estimates compared to DF1 Veh mice and NF1 NIC mice did not differ in MESOR, global amplitude, or orthophase estimates from either DF1 NIC or NF1 Veh mice. However, the N397 allele did appear to exert a prophylactic effect in NF2 NIC mice, which displayed decreased baseline MESOR, global amplitude, and orthophase estimates compared to NF1 Veh (p=0.002, p=0.025, and p=0.009, respectively), NF1 NIC (p=0.000003, p=0.001, and p=0.006, respectively) and DF2 NIC (p=2.5e⁻⁹, p=0.005, and p=0.00002, respectively) mice. Relative to DF1 Veh mice, NF1 Veh mice displayed increased MESOR estimates (p=0.011), trended toward increased global amplitude estimates (p=0.11), and exhibited delayed orthophase estimates (p=0.041).

DNE and the D397N SNP differentially regulate voluntary nicotine intake and preference in F1 and F2 D397N DNE mice

The four-bottle choice test was administered to address the role of DNE, D397N SNP and their interaction on voluntary nicotine intake and preference in first- and second-generation N397 DNE mice relative to their D397 DNE counterparts. Main effects of group (F_2,224=2.98; p=0.05) and measure (F_4,224=455.7; p=1.5e⁻¹⁵) and significant group x measure (F_8,224=7.56; p=6.2e⁻⁹), genotype x measure (F_4,224=4.02; p=0.026), and group x genotype x measure (F_8,224=10.67; p=1.1e⁻¹²) interactions were detected for voluntary nicotine intake (Fig. 4A) and preference (Fig. 4B). DF1 NIC and DF2 NIC mice consumed more nicotine than DF1 Veh mice (p=0.0002 and p=0.003, respectively), consumed a greater proportion of total fluid from 25μg/mL (p=0.008 and p=0.001, respectively), 50μg/mL (p=0.013 and p=0.044, respectively), and 100μg/mL (p=0.026 and p=0.039, respectively) nicotine bottles than DF1 Veh mice, and consumed a smaller proportion of total fluid from water (0μg/mL nicotine) bottles than DF1 Veh mice (p=1.7e⁻⁹ and p=1.6e⁻⁸, respectively). Nicotine consumption and preference in NF1 NIC mice did not differ from either NF1 Veh or DF1 NIC mice. However, similar to our results for home cage and open field activity as well as home cage activity rhythm parameters, the N397 genotype appeared to interact with DNE to exert a protective-like effect against DNE-evoked intergenerational enhancement of voluntary nicotine consumption in NF2 NIC mice, which actually consumed less nicotine than NF1 Veh (0.042), NF1 NIC (p=0.0005), and DF2 NIC (p=0.013) mice. As anticipated given the association of the CHRNA5 D398N SNP with nicotine use in humans, NF1 Veh mice consumed more nicotine than DF1 Veh mice (p=0.037).

Figure 4. DNE and the D397N SNP cooperatively regulate voluntary nicotine intake and preference in F1 and F2 D397N DNE mice.

(A) Voluntary nicotine intake (mg/kg/day) in the four-bottle choice test. DF1 NIC and DF2 NIC mice consumed more nicotine than DF1 Veh mice (p=0.0002 and p=0.003, respectively). NF2 NIC mice consumed less nicotine than DF2 NIC (p=0.013), NF1 Veh (p=0.042), and NF1 NIC (p=0.0005) mice. NF1 Veh mice consumed more nicotine than DF1 Veh mice (p=0.037). (B)N Bottle preference ratios for the four-bottle choice test. DF1 NIC and DF2 NIC mice consumed a greater proportion of total fluid from 25μg/mL (p=0.008 and p=0.001, respectively), 50μg/mL (p=0.013 and p=0.044, respectively), and 100μg/mL (p=0.026 and p=0.039, respectively) nicotine bottles than DF1 Veh mice. DF1 NIC and DF2 NIC mice consumed a smaller proportion of total fluid from water (0μg/mL) bottles than DF1 Veh mice (p=1.7e⁻⁹ and p=1.6e⁻⁸, respectively). NF2 NIC mice consumed a greater proportion of total fluid from water (0μg/mL) bottles than NF1 Veh and NF1 NIC mice (p=0.007 and p=0.00001, respectively). Compared to DF2 NIC mice, NF2 NIC mice consumed a greater proportion of total fluid from water (0μg/mL) bottles (p=3.3e⁻¹⁰) but a smaller proportion of total fluid from 25μg/mL (p=0.00002) and 50μg/mL (p=0.033) bottles. n_DF1Veh=61, n_DF1NIC=63, n_DF2NIC=73, n_NF1Veh=61, n_NF1NIC=63, and n_NF2NIC=73. 0, 0 μg/mL nicotine; 25, 25 μg/mL nicotine; 50, 50 μg/mL nicotine; 100, 100 μg/mL nicotine. All data are mean ± S.E.M. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

DNE and the D397N SNP interact to affect the impacts of voluntary nicotine intake on home cage activity patterns among adolescent F1 and F2 D397N DNE mice

The impacts of voluntary nicotine consumption on active and inactive phase home cage activity were compared at baseline, during the first (N1) and second (N2) four-day intervals of the four-bottle choice test, and over a 24-hour nicotine withdrawal period to assess how differential activity patterns related to DNE, the D397N SNP, and the interaction thereof are affected by voluntary nicotine intake. Main effects of phase (F_1,200=3397.5; p=0.00001) and experimental stage (F_3,200=113.2; p=1.2e⁻¹⁵) and significant group x genotype (F_2,200=3.05; p=0.05), group x experimental stage (F_6,200=10.16; p=8.4e⁻¹⁰), phase x experimental stage (F_3,200=120.4; p=2.2e⁻¹⁵), group x genotype x experimental stage (F_6,200=6.71; p=0.000002), group x genotype x phase (F_2,200=3.16; p=0.044), group x phase x experimental stage (F_6,200=8.06; p=8.4e⁻⁸), and group x genotype x phase x experimental stage (F_6,200=3.92; p=0.001) interactions were detected for active (Figs. 5A and 5B) and inactive (Figs. 5C and 5D) phase home cage activity.

Figure 5. DNE and the D397N SNP collude to affect the impacts of voluntary nicotine intake on home cage activity patterns among adolescent F1 and F2 D397N DNE mice.

(A) Within-genotype contrasts of active phase home cage activity in four-bottle choice-tested mice. DF1 NIC and DF2 NIC mice were more active in the active phase during nicotine withdrawal than DF1 Veh mice (p=0.011 and p=0.0002, respectively). NF1 NIC mice were less active in the active phase during N2 than NF1 Veh mice (p=0.017), and NF2 NIC mice were less active than NF1 Veh mice during nicotine withdrawal (0.048). (B) Within-treatment group contrasts of active phase home cage activity in four-bottle choice-tested mice. DF1 Veh mice were more active in the active phase during N2 compared to baseline (p=0.007). Both DF1 Veh and NF1 Veh mice were less active in the active phase during nicotine withdrawal versus baseline (p=0.00005 and p=0.004, respectively), and NF1 Veh mice were more active in the active phase during nicotine withdrawal compared to DF1 Veh mice (p=0.009). DF1 NIC, DF2 NIC, and NF1 NIC mice were less active in the active phase during N1 (p=0.002, p=0.037, and p=0.0003, respectively), N2 (p=0.007, p=0.048, and p=0.0008, respectively), and nicotine withdrawal (p=1.7e⁻⁹, p=5.1e⁻⁷, and p=3.6e⁻¹¹, respectively) compared to baseline. NF2 NIC mice were more active in the active phase during both N1 and N2 compared to baseline (p=0.024 and p=0.0004, respectively), but were less active in the active phase during nicotine withdrawal compared to both baseline (p=0.0003) and DF2 NIC mice (p=0.045). (C) Within-genotype contrasts of inactive phase home cage activity in four-bottle choice-tested mice. Compared to DF1 Veh mice, DF1 NIC and DF2 NIC mice were more active in the inactive phase during nicotine withdrawal (p=0.003 and p=0.000001, respectively). NF1 NIC mice were more active in the inactive phase during nicotine withdrawal than NF1 Veh mice (p=0.021). (D) Within-treatment group contrasts of active phase home cage activity in four-bottle choice-tested mice. DF1 NIC, DF2 NIC, NF1 NIC, and NF1 Veh mice were less active in the inactive phase during N1 (p=0.00004, p=4.9e⁻⁸, p=0.000004, and p=0.0009, respectively), N2 (p=0.00004, p=0.0000007, p=0.0000005, and p=0.0006, respectively), and nicotine withdrawal (p=0.038, p=0.049, p=0.0008, and p=0.0001, respectively) compared to baseline. NF2 NIC mice were less active in the inactive phase than DF2 NIC mice during nicotine withdrawal (p=0.0004). Statistical analyses of baseline active and inactive phase home cage activity are detailed in Figure 2A and Figure 2B, respectively. n_DF1Veh=61, n_DF1NIC=63, n_DF2NIC=73, n_NF1Veh=61, n_NF1NIC=63, and n_NF2NIC=73. four-bottle choice test, four-bottle choice test; BL, baseline, 3 days pre-four-bottle choice test; N1, days 1–4 of the four-bottle choice test; N2, days 5–8 of the four-bottle choice test; WD, withdrawal, 1 day post-four-bottle choice test. All data are mean ± S.E.M. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

DF1 NIC and DF2 NIC were less active in the active phase during N1 (p=0.002 and p=0.037, respectively), N2 (p=0.007 and p=0.048, respectively), and nicotine withdrawal (p=1.7e⁻⁹ and p=5.1e⁻⁷, respectively) compared to baseline and did not differ from DF1 Veh mice for active phase activity during N1 or N2, whereas DF1 Veh mice exhibited increased activity in the active phase during N2 compared to baseline (p=0.007). DF1 NIC and DF2 NIC mice were more active in the active phase during nicotine withdrawal than DF1 Veh mice (p=0.011 and p=0.0002, respectively). Analogous to DF1 NIC and DF2 NIC mice, NF1 NIC mice were less active in the active phase during N1 (p=0.0003), N2 (p=0.0008), and nicotine withdrawal (p=3.6e⁻¹¹, respectively) compared to baseline. NF1 NIC mice were also less active in the active phase during N2 than NF1 Veh mice (p=0.017). Both DF1 Veh and NF1 Veh mice were less active in the active phase during nicotine withdrawal versus baseline (p=0.00005 and p=0.004, respectively), and NF1 Veh mice were more active in the active phase during nicotine withdrawal compared to DF1 Veh mice (p=0.009). Similar to DF1 Veh and NF1 Veh mice, NF2 NIC mice were more active in the active phase during both N1 and N2 compared to baseline (p=0.024 and p=0.0004, respectively) but were less active in the active phase during nicotine withdrawal compared to baseline (p=0.0003), NF1 Veh mice (p=0.048), and DF2 NIC mice (p=0.045).

DF1 NIC and DF2 NIC mice were less active in the inactive phase during N1 (p=0.00004 and p=4.9e⁻⁸, respectively), N2 (p=0.00004 and p=0.0000007, respectively), and nicotine withdrawal (p=0.038 and p=0.049, respectively) compared to baseline. DF1 NIC and DF2 NIC mice were more active in the inactive phase during nicotine withdrawal (p=0.003 and p=0.000001, respectively) compared to baseline. Similar to DF1 NIC and DF2 NIC mice, both NF1 NIC and NF1 Veh mice were less active in the inactive phase during N1 (p=0.000004 and p=0.0009, respectively), N2 (p=0.0000005 and p=0.0006, respectively), and nicotine withdrawal (p=0.0008 and p=0.0001, respectively) compared to baseline, and NF1 NIC mice were more active in the inactive phase during nicotine withdrawal relative to NF1 Veh mice (p=0.021). In contrast to all other DNE groups, the activity of NF2 NIC mice during the inactive phase was not altered during N1, N2, or withdrawal. Compared to DF2 NIC mice, NF2 NIC mice were less active in the inactive phase during nicotine withdrawal (p=0.0004).

DNE and the D397N SNP interact to bias the effects of voluntary nicotine intake on both activity in a novel environment and risk-taking behaviors in adolescent F1 and F2 D397N DNE mice

The impacts of four-bottle choice testing on total distance moved and percent distance moved in the center of the open field were assessed to evaluate how nicotine consumption alters the effect of genotype, DNE, and the interaction thereof on activity in a novel context as well as risk-taking behaviors. Main effects of group (F_2,242=5.58; p=0.004), measure (F_1,242=6554.0; p=0.00001), and experimental stage (F_2,242 =197.8, p=0.00002) and significant group x genotype (F_2,224=4.64; p=0.011), group x measure (F_2,224=5.58; p=0.004), group x experimental stage (F_4,242=4.68; p=0.001), measure x experimental stage (F_2,242=197.7; p=0.00001), group x genotype x measure (F_2,242=4.59; p=0.011), group x genotype x experimental stage (F_4,242=3.36; p=0.012), group x measure x experimental stage (F_4,242=4.70; p=0.001), and group x genotype x measure x experimental stage (F_4,242=3.37; p=0.012) interactions were detected for total distance moved (Figs. 6A and 6B) and percent distance moved in the center (Figs. 6C and 6D) of the open field.

Figure 6. DNE and the D397N SNP interact to bias the effects of voluntary nicotine intake on both activity in a novel environment and risk-taking behaviors in adolescent F1 and F2 D397N DNE mice.

(A) Within-genotype contrasts of total distance moved in the open field by four-bottle choice-tested mice. DF1 NIC and DF2 NIC mice moved a greater total distance in the open field compared to DF1 Veh mice immediately following the four-bottle choice test (p=0.0009 and p=0.00002, respectively), and DF2 NIC mice moved a greater total distance in the open field after nicotine withdrawal compared to DF1 Veh mice (p=0.038). (B) Within-treatment group contrasts of total distance moved in the open field by four-bottle choice-tested mice. DF1 Veh (p=0.000001), DF1 NIC (p=0.0000002), DF2 NIC (p=1.1e⁻⁸), NF1 Veh (p=0.00005), NF1 NIC (p=0.000008), and NF2 NIC (p=2.1e⁻¹⁴) mice moved a greater total distance in the open field immediately following the four-bottle choice test compared to baseline. NF2 NIC mice were more active in the open field after nicotine withdrawal compared to baseline (p=0.0002). (C) Within-genotype contrasts of percent distance moved in the center of the open field by four-bottle choice-tested mice. DF2 NIC mice moved a greater percent distance in the center of the open field immediately following the four-bottle choice test compared to DF1 Veh and DF1 NIC mice (p=0.001 and p=0.006, respectively). NF2 NIC mice moved a lesser percent distance in the center of the open field immediately following the four-bottle choice test relative to NF1 Veh and NF1 NIC mice (p=0.002 and p=0.009, respectively). (D) Within-treatment group contrasts of percent distance moved in the center of the open field by four-bottle choice-tested mice. Immediately following the four-bottle choice test, NF1 Veh mice moved a greater percent distance in the center of the open field compared to both baseline (p=0.013) and DF1 Veh mice (p=0.0008), and NF2 NIC mice moved a lesser percent distance in the center of the open field compared to DF2 NIC mice (p=0.0002). Compared to baseline, DF1 NIC (p=0.0000003), DF2 NIC (p=0.00002), NF1 Veh (p=0.003), NF1 NIC (p=0.0000005), and NF2 NIC (p=5.3e⁻⁸) mice moved a lesser total distance in the open field after nicotine withdrawal. Statistical analyses of baseline open field activity and risk-taking behaviors are detailed in Figure 2C and Figure 2D, respectively. n_DF1Veh=61, n_DF1NIC=63, n_DF2NIC=73, n_NF1Veh=61, n_NF1NIC=63, and n_NF2NIC=73. four-bottle choice test, four-bottle choice test; BL, baseline, 3 days pre-four-bottle choice test; N1, days 1–4 of the four-bottle choice test; N2, days 5–8 of the four-bottle choice test; WD, withdrawal, 1 day post-four-bottle choice test. All data are mean ± S.E.M. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

For all groups, total distance moved in the open field was significantly increased following the four-bottle choice test relative to their respective baseline levels (DF1 Veh (p=0.000001), DF1 NIC (p=0.0000002), DF2 NIC (p=1.1e⁻⁸), NF1 Veh (p=0.00005), NF1 NIC (p=0.000008), and NF2 NIC (p=2.1e⁻¹⁴)). DF1 NIC and DF2 NIC mice moved a greater total distance in the open field immediately following the four-bottle choice test compared to DF1 Veh mice (p=0.0009 and p=0.00002, respectively). Conversely, NF1 NIC and NF2 NIC mice did not differ in open field activity immediately following the four-bottle choice test relative to NF1 Veh mice. Following acute nicotine withdrawal, open field activity for all groups except NF2 NIC mice returned to baseline levels. Compared to baseline, open field activity of NF2 NIC mice was elevated following nicotine withdrawal (p=0.0002).

Subsequent to completion of the four-bottle choice test, risk-taking behaviors were not altered in DF1 Veh, DF1 NIC, or DF2 NIC mice relative to baseline. However, DF2 NIC mice moved a greater percent distance in the center of the open field immediately following the four-bottle choice test compared to DF1 Veh and DF1 NIC mice (p=0.001 and p=0.006, respectively), while DF1 mice no longer differed from DF1 Veh mice. In contrast, the impact of the four-bottle choice test on risk-taking behaviors in N397 mice was group-dependent. Compared to baseline, NF1 Veh mice exhibited an increase in risk-taking behaviors (p= 0.013), NF1 NIC mice exhibited no difference in risk-taking behaviors, and NF2 NIC mice displayed a decrease in risk-taking behaviors (p=0.002) immediately following the four-bottle choice test. Moreover, NF2 NIC mice moved a lesser percent distance in the center of the open field immediately following the four-bottle choice test versus NF1 Veh (p=0.002), NF1 NIC (p=0.009), and DF2 NIC (p=0.0002) mice. Acute nicotine withdrawal suppressed risk-taking behavior in all groups relative to baseline except for DF1 Veh mice (DF1 NIC (p=0.0000003), DF2 NIC (p=0.00002), NF1 Veh (p=0.003), NF1 NIC (p=0.0000005), and NF2 NIC (p=5.3e⁻⁸)).

DNE and the D397N SNP combinatorially influence the rhythmicity of home cage activity and the impacts of voluntary nicotine intake thereon among adolescent F1 and F2 D397N DNE mice

Composite parameters of home cage activity rhythms during voluntary nicotine consumption in the four-bottle choice-test and during acute nicotine withdrawal were compared within and among groups at baseline, N1, N2, and nicotine withdrawal. Main effects of genotype (F_1,187=4.96; p=0.027), measure (F_2,187=2625.7; p=0.00001), and experimental stage (F_3,187=50.84, p=0.00002) and significant group x genotype (F_2,187=3.59; p=0.029), group x measure (F_4,187=2.71; p=0.30), group x experimental stage (F_6,187=6.27; p=0.002), genotype x measure (F_2,187=6.13; p=0.010), measure x experimental stage (F_6,187=43.91; p=0.00002), group x genotype x measure (F_4,187=3.15; p=0.035), group x genotype x experimental stage (F_6,187=6.01; p=0.003), group x measure x experimental stage (F_12,187=4.27; p=0.015), and group x genotype x measure x experimental stage (F_12,187=4.16; p=0.017) interactions were detected for MESOR (Figs. 7A and 7B), global amplitude (Figs. 7C and 7D), and orthophase (Figs. 7E and 7F) estimates among four-bottle choice-tested mice.

Figure 7. DNE and the D397N SNP combinatorially influence the impacts of voluntary nicotine intake on the rhythmicity of home cage activity among adolescent F1 and F2 D397N DNE mice.

(A) Within-genotype contrasts of MESOR estimates for four-bottle choice-tested mice. DF1 NIC and DF2 NIC mice had elevated MESOR estimates during nicotine withdrawal relative to DF1 Veh mice (p=0.039 and p=0.008). NF1 NIC mice had reduced MESOR estimates during N2 compared to NF1 Veh mice (p=0.048). (B) Within-treatment group contrasts of MESOR estimates for four-bottle choice-tested mice. Relative to baseline, DF1 Veh and NF1 Veh mice trended toward elevated (p=0.09 and p=0.17, respectively) and exhibited reduced (p=0.00005 and p=0.002, respectively) MESOR estimates during N2 and nicotine withdrawal, respectively. NF1 Veh mice had increased MESOR estimates during nicotine withdrawal versus DF1 Veh mice (p=0.006). Compared to baseline, DF1 NIC, DF2 NIC, and NF1 NIC mice had reduced MESOR estimates during N1 (p=0.005, p=0.002, and p=0.00006, respectively), N2 (p=0.009, p=0.005, and p=0.00006, respectively), and nicotine withdrawal (p=2.3e⁻⁹, p=2.8e⁻⁸, and p=1.3e⁻¹², respectively). Relative to baseline, NF2 NIC mice had elevated (p=0.024) and reduced (p=0.0009) MESOR estimates during N2 and nicotine withdrawal, respectively. NF2 NIC mice also had decreased MESOR estimates during nicotine withdrawal compared to DF2 NIC mice (p=0.008) (C) Within-genotype contrasts of global amplitude estimates for four-bottle choice-tested mice. DF2 NIC mice had increased global amplitude estimates during nicotine withdrawal versus DF1 Veh mice (p=0.034). NF1 NIC mice had decreased global amplitude estimates during N2 versus NF1 Veh mice (p=0.029), and NF2 NIC mice trended toward decreased global amplitude estimates during nicotine withdrawal versus NF1 Veh mice (p=0.16). (D) Within-treatment group contrasts of global amplitude estimates for four-bottle choice-tested mice. DF1 Veh and NF1 Veh mice had increased global amplitude estimates during N2 compared to baseline (p=0.002 and p=0.018, respectively), and DF1 Veh mice had reduced global amplitude estimates during nicotine withdrawal versus baseline (p=0.035). NF1 Veh mice had increased elevated global amplitude estimates during N1 (p=0.041) and nicotine withdrawal (p=0.005) compared to DF1 Veh mice. DF1 NIC (p=0.001), DF2 NIC (p=0.033), and NF1 NIC (p=0.026) mice had decreased global amplitude estimates during nicotine withdrawal versus baseline. NF2 NIC mice had increased global amplitude estimates during both N1 (p=0.005) and N2 (p=0.0004) relative to baseline. (E) Within-genotype contrasts of orthophase estimates for four-bottle choice-tested mice. DF2 NIC mice had delayed orthophase estimates relative to DF1 Veh mice (p=0.008) during nicotine withdrawal. NF2 NIC mice had advanced orthophase estimates versus NF1 NIC mice during nicotine withdrawal (p=0.019). (F) Within-treatment group contrasts of orthophase estimates for four-bottle choice-tested mice. Compared to baseline, DF1 Veh (p=0.044 and p=0.023), DF1 NIC (p=0.0006 and p=0.0001), DF2 NIC (p=0.00005 and p=0.00005), NF1 Veh (p=0.004 and p=0.002), and NF1 NIC (p=0.0005 and p=0.0001) mice had advanced orthophase estimates during N1 and N2, respectively. NF1 Veh mice had delayed orthophase estimates during nicotine withdrawal relative to baseline (p=0.007). NF2 NIC mice had advanced orthophase estimates during nicotine withdrawal versus DF2 NIC mice (p=0.0002). n_DF1Veh=61, n_DF1NIC=63, n_DF2NIC=73, n_NF1Veh=61, n_NF1NIC=63, and n_NF2NIC=73. four-bottle choice test, four-bottle choice test; BL, baseline, 3 days pre-four-bottle choice test; N1, days 1–4 of the four-bottle choice test; N2, days 5–8 of the four-bottle choice test; WD, withdrawal, 1 day post-four-bottle choice test; MESOR, midline estimating statistic of rhythm (oscillatory mean); global amplitude, difference between peak and trough (oscillatory range); orthophase, clock time at peak activity (oscillatory phase). All data are mean ± S.E.M. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

DF1 NIC and DF2 NIC mice had reduced MESOR estimates during N1 (p=0.005 and p=0.002, respectively), N2 (p=0.009 and p=0.005, respectively), and nicotine withdrawal (p=2.3e⁻⁹ and p=2.8e⁻⁸, respectively) compared to baseline. Both DF1 NIC and DF2 NIC mice had elevated MESOR estimates relative to DF1 Veh mice (p=0.039 and p=0.008, respectively) during nicotine withdrawal. Similar to DF1 NIC mice, NF1 NIC mice exhibited reduced MESOR estimates during N1 (p=0.00006), N2 (p=0.00006), and nicotine withdrawal (p=1.3e⁻¹²) relative to baseline. Contrary to the lack of difference in MESOR estimates between DF1 NIC and DF1 Veh mice during N2, NF1 NIC mice displayed decreased MESOR estimates during N2 relative to NF1 Veh mice (p=0.048). Unlike DF1 NIC, DF2 NIC and NF1 NIC mice, NF2 NIC mice showed elevated MESOR estimates during N2 versus baseline (p=0.024). However, analogous to all other groups, NF2 NIC mice displayed decreased MESOR estimates during nicotine withdrawal (p=0.0009) compared to baseline. Nevertheless, MESOR estimates for NF2 NIC mice during nicotine withdrawal were reduced relative to DF2 NIC mice (p=0.008). Compared to baseline, DF1 Veh and NF1 Veh mice trended toward elevated MESOR estimates during N2 (p=0.09 and p=0.17, respectively) and, consistent with all other groups, exhibited reduced (p=0.00005 and p=0.002, respectively) MESOR estimates during nicotine withdrawal. In contrast to the decreased MESOR estimates during nicotine withdrawal compared to baseline detected for both DF1 Veh and NF1 Veh mice, NF1 Veh mice exhibited elevated MESOR estimates during nicotine withdrawal versus DF1 Veh mice (p=0.006).

Global amplitude estimates were unaltered in DF1 NIC and DF2 NIC mice during N1 and N2 but were decreased during nicotine withdrawal (p=0.001 and p=0.033, respectively). Analogous to DF1 NIC mice, relative to baseline NF1 NIC mice exhibited no alteration in global amplitude estimates during N1 or N2 but exhibited decreased global amplitude estimates during nicotine withdrawal (p=0.026). NF1 NIC mice also displayed decreased global amplitude estimates during N2 versus NF1 Veh mice (p=0.029). Whereas DF2 NIC mice exhibited increased global amplitude estimates during nicotine withdrawal versus DF1 Veh mice (p=0.034), NF2 NIC mice trended toward decreased global amplitude estimates during nicotine withdrawal versus NF1 Veh mice (p=0.16). In addition, DF1 Veh and NF1 Veh mice displayed increased global amplitude estimates during N2 compared to baseline (p=0.002 and p=0.018, respectively), and DF1 Veh but not NF1 Veh mice showed reduced global amplitude estimates during nicotine withdrawal relative to baseline (p=0.035). Despite the similar nicotine responsivity of activity rhythms in DF1 Veh and NF1 Veh mice, NF1 Veh mice exhibited increased global amplitude estimates during N1 (p=0.041) and nicotine withdrawal (p=0.005) compared to DF1 Veh mice. Comparable to DF1 Veh mice but unlike DF2 NIC mice, NF2 NIC mice displayed increased global amplitude estimates during both N1 (p=0.005) and N2 (p=0.0004) but did not differ during nicotine withdrawal versus baseline.

DF1 Veh (p=0.044 and p=0.023), DF1 NIC (p=0.0006 and p=0.0001, respectively) and DF2 NIC (p=0.00005 and p=0.00005, respectively) mice exhibited advanced orthophase estimates during N1 and N2 relative to baseline. In parallel with the orthophase-advancing effect of voluntary nicotine intake in DF1 NIC mice, NF1 NIC mice displayed advanced orthophase estimates during N1 (p=0.0005) and N2 (p=0.0001) compared to baseline. Whereas DF2 NIC mice displayed delayed orthophase estimates relative to DF1 Veh mice during nicotine withdrawal (p=0.008), NF2 NIC mice exhibited advanced orthophase estimates during nicotine withdrawal compared to both DF2 NIC (p=0.0002) and NF1 NIC (p=0.019) mice but did not differ from NF1 Veh mice. Similar to DF1 Veh mice, NF1 Veh mice showed advanced orthophase estimates during N1 (p=0.004) and N2 (p=0.002) versus baseline. While orthophase estimates returned to baseline levels in DF1 Veh mice during nicotine withdrawal, NF1 Veh mice exhibited delayed orthophase estimates during nicotine withdrawal relative to baseline (p=0.007).

Discussion

The multigenerational DNE paradigm utilized for the present study models and provides insight into the intergenerational impacts of maternal and grandmaternal smoking before and during pregnancy and nursing. Therein, F1 DNE mice (DF1 NIC and NF1 NIC groups) model nicotine exposure in the children of maternal smokers, and F2 DNE mice (DF2 NIC and NF2 NIC groups) underwent maternal germline nicotine exposure and model the grandchildren of maternal smokers.

All behavioral testing commenced at PND 35, an age that corresponds to early adolescence in humans and is a developmental stage that overlaps with the ages of onset of various neurodevelopmental disorders. Importantly, the behavioral profiles of D397 mice generated by the present study were highly consistent with those from our previous research with C57BL/6J mice (Buck et al., 2019) This was expected since the D397N mouse was generated on a C57BL/6 background and D397 is the native C57BL/6 allele. These results demonstrate the highly reproducible impact of DNE on behavioral outcomes in wild-type mice (C57BL/6 or D397 mice) and reveal that DNE precipitates a constellation of intergenerational behavioral perturbations which mimics the symptomologies of neurodevelopmental disorders such as ADHD, autism, and schizophrenia. Further, our findings are congruent with epidemiological studies documenting disordered behaviors and exacerbated risk for neurodevelopmental disorders in both the children and the grandchildren of maternal smokers (Baird et al., 2012; Van Veen et al., 2009; Imeraj et al., 2012; Bijlenga et al., 2013; American Psychiatric Association, 2013; Snitselaar et al., 2017; Ajarem et al., 1998; Pauly et al., 2004; Paz et al., 2007; Heath et al., 2010; Ernst et al., 2001; Amsterdam et al., 2018; Milberger et al., 1997; Lambert and Hartsough, 1998; Kollins et al., 2005; Hong et al., 2011; Rhodes et al., 2016; Schuch et al., 2016; Elkins et al., 2018; Ohi et al., 2019; Barr et al., 2006; Tang et al., 2016; Mallet et al., 2017; Sagud et al., 2018; Golding et al., 2017; Gustavson et al., 2017).

Considering that the N397 variant confers neurodevelopmental disorder-like behaviors as well as hypofrontality in the absence of DNE (Koukouli et al., 2017), we suspected that vehicle-exposed adolescent N397 mice may exhibit abnormalities relative to vehicle-exposed D397 mice for the behaviors assayed herein. Indeed, we observed that the behaviors of vehicle-exposed N397 mice were highly comparable to those of D397 DNE mice. The observation that vehicle-exposed N397 mice exhibit increased nicotine consumption and preference versus vehicle-exposed D397 mice is consistent with the association of the D398N SNP with several measures of nicotine use in humans (Bierut et al., 2008; Sarginson et al., 2011; Berrettini et al., 2012; Wen et al., 2016; Ohi et al., 2019). The novel findings of this study that the N397 variant confers hyperactivity, disrupted rhythmicity of activity, and increased risk-taking behaviors are consistent with a recent animal model study implicating the N397 variant in other brain and behavioral anomalies characteristic of neurodevelopmental disorders (Koukouli et al., 2017) and are further concordant with clinical studies implicating the N398 variant in neurodevelopmental disorders including ADHD (Schuch et al., 2016), autism (Forrest et al., 2018), and schizophrenia (Jackson et al., 2013; Forrest et al., 2018; Han et al., 2019; Ohi et al., 2019) as well as in altered functional connectivity of neural circuitry involved in psychiatric disorders (Hong et al., 2011).

The primary objective of the present study was to determine whether first- and/or second-generation DNE mice homozygous for the N397 SNP were differently impacted by DNE relative to mice homozygous for the D397 SNP. In light of previous studies demonstrating that Chrna5 deletion leads to abnormal cortical structure and function (Bailey et al. 2010; Tian et al., 2011) which appear to be rescued by DNE (Bailey et al., 2014) and that the N397 allele of Chrna5 is a hypomorph that evokes cortical dysfunction reminiscent of that produced by Chrna5 deletion (Bierut et al., 2008; George et al., 2012; Kuryatov et al., 2011; Tammimaki et al., 2012; Sciaccaluga et al., 2015; Morel et al., 2014), we postulated that, similar to Chrna5 KO DNE mice, N397 DNE mice would be resistant to the behavioral teratogenicity of DNE. Contrary to our prediction, however, DNE did not restore normative behavior in first-generation N397 DNE progeny. For slightly greater than half of the behaviors assessed herein, DNE neither reduced nor exacerbated abnormal behaviors in first-generation N397 DNE mice relative to the vehicle-exposed N397 mice. For three measures, namely active phase home cage activity, open field activity, and risk-taking behaviors, DNE actually exacerbated the severity of behavioral perturbations conferred by the N397 allele alone. In contrast, and as evinced in Figure 8, second-generation N397 DNE mice were almost entirely devoid of DNE-elicited behavioral perturbations and instead were nearly indistinguishable from vehicle-exposed D397 mice for the behaviors assessed., These data suggest that the N397 allele is protective against the intergenerational transmission of the vast majority of DNE-induced behavioral anomalies. The sole exception to this conclusion is the finding that the N397 allele does not exert a prophylactic effect on the intergenerational transmission of DNE-evoked enhancement of risk-taking behaviors, suggesting that this phenotype may be uniquely susceptible to the intergenerational impacts of DNE.

Figure 8. Survey of the discrete and combinatorial impacts of DNE and the D397N polymorphism on baseline behavioral phenotypes.

Dual-gradient heatmap depicting the percentage differences of first-generation NF1 Veh and first- and second-generation D397 and N397 DNE mice (horizontal axis) relative to DF1 Veh control mice for baseline behavioral measures (vertical axis). A percentage difference of zero versus indicates no difference (depicted in gray) relative to DF1 Veh control mice, whereas percentage differences of −10% and 40% indicate a 10% decrease (depicted in red) and a 40% increase (depicted in green), respectively, relative to DF1 Veh control mice. For the nicotine intake (mg/kg) measure, the percentage increases relative to D Veh mice observed for DF1 NIC (56.7%), DF2 NIC (47.8%), N Veh (37.8%), and NF1 NIC (57.9%) mice exceed the 40% maximum scale of the heatmap.

The apparent nonexistence of any prophylactic effects of the N397 variant in first-generation N397 DNE offspring implies that the prophylaxis observed in second-generation N397 DNE progeny is not mediated by direct effects of nicotine on first-generation N397 DNE mice that rescue deficits conferred by the N397 hypomorph allele. Rather, the findings of this study suggest that the protective effects of the N397 variant instead appear to be transmitted, via the maternal germline, from first-generation DNE progeny to second-generation DNE progeny via a presumptive epigenetic mechanism. As such, future studies endeavoring to elucidate the epigenetic alterations that differ between D397 and N397 DNE mice may provide valuable insight into mechanisms underlying the prophylactic effects of the N397 variant on the intergenerational transmission of DNE-induced behavioral perturbations.

It has been hypothesized that the increased prevalence of smoking among neurodevelopmentally-disordered individuals may constitute a form of self-medication (Fang et al., 2019; McClernon and Kollins 2008; Lucatch et al., 2018; Logan et al., 2014; van Amsterdam et al., 2018; Kumari and Postma, 2005). Consistent with this hypothesis, mice whose baseline home cage activity patterns and rhythms were disrupted by DNE (first- and second-generation D397 and first-generation N397 DNE mice) displayed significant normalization of these behaviors in response to nicotine. Conversely, second-generation N397 DNE mice showed home cage hyperactivity in response to nicotine analogous to that exhibited by vehicle-exposed D397 control mice. Considering that nicotine exerts therapeutic-like effects on hyperactivity in ADHD, autism, and schizophrenia (Gehricke et al., 2009; Arnold et al., 2012; Lippiello, 2006; Deutsch et al., 2015; Takechi et al., 2016; Koukouli et al., 2017) but induces hyperactivity in healthy individuals and drug-naïve rodents (Davies et al., 2001; Palmatier et al., 2003; Centers for Disease Control and Prevention, 2005; Harris et al., 2014; Fait et al., 2017; Domingues et al., 2019), these findings support the interpretation that first-generation D397 and N397 DNE offspring as well as second-generation D397 DNE offspring exhibit neurodevelopmental disorder-like behavioral responsivity to nicotine, while second-generation N397 DNE offspring exhibit stimulant-like behavioral responsivity to nicotine that mirrors that observed in vehicle-exposed D397 control mice.

Lastly, there were minimal impacts of genotype on nicotine withdrawal responses, with the noteworthy exception that all groups except vehicle-exposed D397 mice exhibited suppression of risk-taking behaviors relative to baseline. Coupled with the finding that nicotine withdrawal also induces repression of active phase home cage activity across all groups, these results suggest that an eight-day four-bottle choice testing paradigm is sufficient to induce measurable nicotine withdrawal symptoms but that neither DNE nor the N397 allele exacerbates or reduces said symptoms.

There are limitations to and possible confounds associated with the developmental exposure paradigm utilized for this study which warrant consideration when interpreting the findings reported herein. Namely, considering that F1 DNE animals (DF1 NIC and NF1 NIC mice) underwent behavioral testing beginning fourteen days following weaning (and therefore fourteen days following cessation of nicotine exposure via F0 dams), it is possible that the behavioral phenotypes documented in F1 DNE animals may be confounded by unmeasured effects of nicotine discontinuation (withdrawal). However, the findings in D397 DNE mice from the current study as well as those in C57BL/6J DNE mice from our prior work (Buck et al., 2019) that the preponderance of phenotypic changes exhibited by wild-type F1 DNE mice are transmitted to F2 DNE mice are incongruent with the possibility that the behavioral anomalies observed in F1 DNE animals stem from a nicotine discontinuation syndrome, since F2 DNE mice were never directly exposed to nicotine and thus were not subject to possible confounding of behavioral phenotypes by nicotine discontinuation syndrome following weaning. Moreover, for N397 Veh mice, it is possible that an interaction between the N397 allele and developmental saccharin (vehicle) exposure could contribute to the hyperactivity observed in N397 Veh compared to D397 Veh mice. Lastly, since the offspring we tested were either directly exposed to nicotine (F1 NIC) or arose from female germ cells that were exposed to nicotine (F2 NIC), we can only state that the effects of DNE are intergenerational. In order to determine if the effects are transgenerational, we would need to test F3 offspring.

Conclusions

The present study is, to the best of our knowledge, the first to indicate the impact of a SNP that is associated with nicotine use and neurodevelopmental disorders on the multigenerational transmission of DNE-induced neurodevelopmental disorder-like behavioral perturbations. Therein, the results of this study suggest that an interaction between the D397N SNP and DNE confers prophylaxis against the intergenerational transmission of DNE-induced neurodevelopmental disorder-like behavioral phenotypes. However, the molecular mechanisms underpinning these gene by environment interactions remain indeterminate. Building upon the foundational findings of this research, future epidemiological studies of neurodevelopmental disorder risk in the descendants of maternal smokers should consider genotype wherever possible and, more specifically, should endeavor to assess whether the N398 variant confers protection against the transmission of enhanced neurodevelopmental disorder risk to the grandchildren of maternal smokers. Moreover, future preclinical studies should attempt to delineate the neurobiological mechanisms underlying the protective effects of the N397 variant on the intergenerational transmission of DNE-induced behavioral alterations, which could in turn reveal novel targets for the development of treatments for and/or prophylactics against the multigenerational transmission of DNE-evoked neurodevelopmental disorder-like phenotypes.

Figure 9. Survey of the discrete and combinatorial impacts of DNE and the D397N polymorphism on behavioral responsivity to nicotine.

(A) Nicotine responsivity of home cage activity patterns and rhythms. Dual-gradient heatmaps depicting the percentage differences of first-generation NF1 Veh and first- and second-generation D397 and N397 DNE mice relative to DF1 Veh control mice for measures (vertical axis) of home cage activity at (horizontal axis) baseline (BL), during the first four-day interval of voluntary nicotine consumption in the four-bottle choice test (N1), during the second four-day interval of voluntary nicotine consumption in the four-bottle choice test (N2), and following a 24-hour post-four-bottle choice test acute nicotine withdrawal period (WD). (B) Nicotine responsivity of open field activity and risk-taking behaviors. Dual-gradient heatmaps depicting the percentage differences of first-generation NF1 Veh and first- and second-generation D397 and N397 DNE mice relative to DF1 Veh control mice for measures of activity and risk-taking behaviors (vertical axis) in the open field at (horizontal axis) baseline (BL), immediately following the conclusion of four-bottle choice testing (NIC), and following a 24-hour post-four-bottle choice test acute nicotine withdrawal period (WD). A percentage difference of zero versus indicates no difference (depicted in gray) relative to DF1 Veh control mice, whereas percentage differences of −20% and 40% indicate a 20% decrease (depicted in red) and a 40% increase (depicted in green), respectively, relative to DF1 Veh control mice.