Age Specific Impacts of Nicotine and Withdrawal on Hippocampal Neuregulin Signaling

Abstract

Smoking remains the leading cause of preventable death in the United States, with 87% of smokers starting before the age of 18. Age of initiation is a major predictive factor for smoking frequency and successful smoking cessation. People who initiate smoking during adolescences are 2.33 times more likely to become heavy smokers and half as likely to quit compared to smokers who started during adulthood. Additionally, schizophrenia, a disease state linked to altered neurodevelopment during adolescence, is a major predictive factor for smoking status. Smoking rates among people suffering from schizophrenia are between 60–90%. Interestingly, the Neuregulin Signaling Pathway (NSP), which plays an important role in neurodevelopment, is implicated in both schizophrenia and nicotine use disorder. Specifically, SNPS in neuregulin 3 (Nrg3) and Erb-B2 Receptor Tyrosine Kinase 4 (ErbB4) have been associated with smoking cessation outcomes and schizophrenia. Here, we examine the effects of chronic nicotine (18 mg/kg/day) and 24-hour withdrawal on NSP gene expression in the hippocampus of adult (20-week-old) and adolescent (4-week-old) mice. We show that withdrawal from chronic nicotine decreased the expression of Erbb4 mRNA in the hippocampus of the adult mice but increased the expression of cytosolic Erbb4 protein in adolescent mice. Nrg3 mRNA and protein expression was not altered by chronic nicotine or withdrawal in the adult or adolescent cohorts, but Nrg3 mRNA and synaptosomal protein expression was lower in the adult withdrawal group when compared to their adolescent counterparts. These results highlight the age specific effects of nicotine withdrawal on the NSP and may contribute to the lower quit rate and higher cigarette consumption of smokers who initiation during adolescences.

Graphical Abstract

Adolescent mice undergoing forced withdrawal had a higher Nrg3, Nrg3os, ErbB4 mRNA than adult mice undergoing forced withdrawal. A proposed age dependent mechanism where a decrease in Ca2+ signaling decreases Nrg3 expression, but a higher Nrg3os mRNA expression in the withdrawal adolescents prevents degradation of Nrg3 leading to a higher level of synaptosomal NRG3 protein during withdrawal compared to the adults. The increased synaptosomal NRG3 binds to ErbB4 causing an endocytosis and effects downstream signaling which may lead to the increased risk of affective disorder in adolescents who smoke.

Introduction

Smoking remains the number one leading cause of preventable death in the United States [1]. Despite the known health risks, 20.8% of American adults currently use a tobacco product, the most popular being cigarettes, followed by e-cigarettes [2]. Cigarette usage has been falling among adults and adolescents, but vaping has grown from 1.5% to 16.0% amongst high schoolers during this period [3]. In fact, from 2017–2019 teen vaping has continued to increase rapidly and it has more than doubled among high schoolers [4]. While the negative health effects of smoking are well established, the negative effects of chronic vaping are not well understood. Despite these unknowns, we do know that adolescents smoking high nicotine concentration e-cigs are >2-fold more likely to report frequent daily cigarette smoking than adolescents who don’t vape [5].

The adolescent period is a critical period for nicotine exposure, as 87% of smokers start smoking before age 18 [1]. Smokers who start during adolescences are 2.33 times more likely to become a heavy smoker and half as likely to quit 10 years after initiation compared to smokers who started during adulthood [6]. One contributing factor could be differential responses to nicotine during development, as people 13 years and younger were significantly more likely to rate their first time smoking more pleasurable than those older than 13 [7].

Smoking is also significantly higher in certain psychiatric disorders that are linked to altered neurodevelopment during adolescence, most notably in schizophrenia. Smoking rates among schizophrenia patients are incredibly high, between 60–90%, compared to the to the approximately 18% prevalence rate in the general U.S. population [8–11]. Interestingly, there is both a genetic and molecular link between schizophrenia and smoking. The Neuregulin Signaling Pathway (NSP) is a neural-enriched molecular cascade that is especially important during critical developmental periods. This pathway modulates neuronal migration and differentiation during development (reviewed in ref. [12]) and continues to be important throughout life for stabilization of synapses [13]. Furthermore, multiple genes within the NSP have been implicated in both schizophrenia and tobacco use disorder [14–18]. The neuregulin family encodes for six genes (Nrg1–6), each of which each can selectively signal through receptors of the ErbB family (ErbB2–4). Neuregulin 3 (NRG3) is a part of this family and exclusively binds to ErbB4-containing receptors, triggering ErbB4 signaling within the cell [19, 20]. Previous studies have shown that blocking ErbB4 signaling through an antagonist or suppressing the expression of Nrg3 mRNA blocks the anxiogenic effects associated with nicotine withdrawal [16]. However, how nicotine may impact this signaling pathway differentially during developmental periods, such as adolescence, is currently unknown.

In this study, we examined the effect of chronic nicotine and withdrawal on NSP signaling in the hippocampus of adolescent and adult mice. Our findings indicate that naïve adolescent animals have higher expression of NSP genes in the hippocampus relative to their adult counterparts. Additionally, exposure to nicotine differentially regulates ErbB4 and Nrg3 mRNA expression in adolescent and adult mice, which could be detected at the protein level specifically at synapses in the hippocampus, which may be due to alterations in mRNA regulation by long non-coding RNA (lncRNAs). These observations were concordant with differential anxiety-like withdrawal behaviors in the adolescent and adult mice, suggesting involvement of this pathway in complex withdrawal symptomology.

Methods and Materials

Animals

Male and female B6/129S F1 mice (8–20 weeks of age) were bred in house and used for this study. The B6/129S F1 mouse strain is a hybrid of C57/B6 and 129S strains, which is routinely used for the background of many genetically modified mice. Use of the F1 mice allows for hybrid vigor and more widely applicable results from experimentation and minimizes founders’ effects. Those mice undergoing chronic treatment were 4 or 20 weeks of age at the beginning of saline or nicotine treatment. Mice were maintained on a 12-hour light-dark cycle (lights on at 7:00 AM), with ad libitum food and water in accordance with the University of Kentucky’s Institutional Animal Care and Use Committee.

Drugs and administration

(-)-Nicotine tartrate (MP Biomedicals, Solon, OH.) was dissolved in 0.9% saline. Nicotine was administered subcutaneously via osmotic minipumps (Alzet model 2002, Cupertino, CA) at a dose of 18 mg/kg/d for 14 days. This dose, reported as freebase weight and based off of previous work [16, 21–23], corresponds to plasma levels of ~0.3 μM [24], a concentration similar to that observed in human smokers consuming an average of 17 cigarettes a day (plasma levels between 0.06 and 0.31 μM) [24].

Osmotic minipumps surgeries

In all experiments, animals were implanted with osmotic minipumps to deliver administration of either nicotine (18 mg/kg/day) or saline. Pump implantation was performed as previously described [25]. Briefly, mice were anesthetized with 1–3% isofluorane and pumps containing either nicotine or saline were implanted subcutaneously via a small incision on their right flank and closed with 7mm stainless steel wound clips. Following 2 weeks of chronic administration, mice were anesthetized with an isoflurane/oxygen vapor mixture (1–3%), an incision was made above the pump at shoulder level and the pump was either removed (to initiate spontaneous withdrawal from either nicotine or saline) or left in place (to serve as sham surgical controls in the nicotine and saline groups). The incision was then closed with 7 mm stainless steel wound clips to induce spontaneous withdrawal.

Open Field Test and Tissue Collection

The OF test was performed as previously described for all mice [25]. Briefly, the OF test is an anxiety-related behavioral model, which also allows simultaneous assay of overall locomotor activity levels in mice [26]. The withdrawal cohort of mice was tested in this model at the 24-hour withdrawal time point. All mice were handled one week prior to the OF test. Every testing session lasted for 10 min, and the Top Scan (Clever Sys Inc., Reston, VA, USA) software was used to track and evaluate mouse movement for analysis. A background profile was generated, and the testing chamber was calibrated in arena design mode according to the manufacturer’s instructions. The software tracked and provided total distance traveled (in millimeters) and the precent time spent in the defined center zone. The boxes were 400 mm X 400 mm, and the center zone was generated by the Top Scan software to be 50% of the zone area, or approximately a 282 mm X 282 mm square in the center of the field. Following behavioral testing, animals were sacrificed and whole hippocampal tissues were microdissected and collected for biochemical analysis.

Quantitative PCR

Quantitative reverse transcriptase PCR was performed as previously described [27] on whole hippocampal samples across all treatment groups. Briefly, RNA was isolated using the RNeasy Mini kit (Qiagen) and qPCR reactions were assembled using Thermo Scientific Maxima SYBR Green master mix along with 100nM primers (Eurofins). The mRNA levels were determined using the 2^-ΔΔCT method [28] and target genes were normalized to the housekeeping gene Hypoxanthine Phosphoribosyltransferase (HPRT). All gene expression values of were normalized to their respective 4 week naïve or 4 week saline-treated controls. Primer sequences are shown in Table 1.

Table 1.

Sequence of primers used in qPCR

Gene	Forward Primer	Reverse Primer
Gomafu	TCTGGAGATAGGTGGTGTTGG	TCCTGCTCAAACAAGCCTAAA
Nrg3os	CAGCTGTGGTGTGTTGAAAGA	GGGGTTTGTCTCTCTTGAAGG
Creb	CATTGCCCCTGGAGTTGTTAT	TTCTCTTGCTGCCTCCCTGTT
Nrg3	CAGCTGTGGTGTGTTGAAAGA	GGGGTTTGTCTCTCTTGAAGG
ErbB4	ACAACCAGCACCATACCAGAG	TGTCATGCATTGGAGTCATGT
ErbB2	ATGAAGTCCTGTCCTCCCAGT	ACCTGGCGTGTGTTTATGTCC
Nrg1	CATGCTCCTATTCAGGCAGAG	TAAAGTGTGCGGAGAAGGAGA
Tbp	GCACAGGACTTACTCCACAGC	GTGGGTTGCTGAGATGTTGAT

Abbreviations: Gomafu, Myocardial Infarction Associated Transcript; Nrg3os Neuregulin 3, Opposite Strand; Creb, cAMP Response Element-binding Protein; Nrg3, Neuregulin 3; ErbB4, Erb-B2 Receptor Tyrosine Kinase 4; ErbB2, Erb-B2 Receptor Tyrosine Kinase 2; Nrg1, Neuregulin 1; Tbp, TATA-Box Binding Protein.

Synaptosomal preparation

To obtain synaptosomal and cytosolic fractions, frozen whole hippocampal tissues were weighed and dounced in a glass homogenizer in 10 vol (1:10, wt/vol) of Syn-PER synaptic protein extraction reagent (Thermo, Rockford, USA) supplemented with a protease and phosphatase inhibitor cocktail. Following manufacturer’s instructions, the homogenate was centrifuged at 1200g for 10 min at 4°C, and then the supernatant was centrifuged for a further 20 min at 1500g at 4°C. The supernatant (cytosolic fraction) was collected, and the synaptosome pellets were resuspended in Syn-PER reagent. The protein concentrations of synaptosomal and cytosolic fractions were determined by the BCA method (Pierce Protein Biology™).

Western blotting

Protein analysis was performed as described previously [29] on whole hippocampal samples of all treatment groups. Briefly, 20μg of protein were resolved in AnyKD™ precast polyacrylamide gel (Bio-Rad Laboratories Inc., Hercules, CA, USA) and transferred to nitrocellulose membranes. Membranes were incubated with LI-COR blocking buffer (LI-COR, Lincoln, NE, USA) for 1h at room temperature before reacting overnight at 4°C with primary antibodies: Neuregulin-3 (NRG3) (1:500, PA5–18552, Invitrogen, Carlsbad, CA), ErbB4 (1:500, NBPI-33120, Novus, Centennial, CO.), and Beta-tubulin (1:2000, 2128L, Cell Signaling Technology, Danvers, MA.). After washing in phosphate buffered saline-Tween-20, the blots were incubated in fluorescent secondary antibodies (1:20000, LI-COR) in LI-COR blocking buffer for 1 h at room temperature. Membranes were then washed, and immunolabeling detection and densitometry measurements were performed using the LICOR Odyssey System (LI-COR). Ratios of the proteins of interest (NRG3 and ErbB4) to the housekeeping protein (β-actin) densities were calculated for each sample and normalized to 4-week old saline-treated controls.

Data analysis

Statistical analyses were performed with GraphPad Prism 8.0 software package (GraphPad Software, CA). The naïve qPCR for mice 4 week through 20 week results were analyzed using one-way repeated measures ANOVA with age as the primary factor, followed by Sidak’s multiple comparison tests. The rest of the qPCR, WB, and behavioral results were analyzed using two-way repeated measures ANOVA with age and treatment as factors, followed by Sidak’s multiple comparison tests. All data are expressed as mean ±SEM.

RESULTS

Gene Expression Changes in Neuregulin Signaling Pathway Family Members in the Hippocampus Across Adolescence and Adulthood

To explore the role of age on the expression of key genes associated with the Neuregulin Signaling Pathway, mice were sacrificed at 4, 8, 12, 16, and 20 weeks. First we examined expression levels of the transcription factor CREB, which has been broadly implicated in substance use disorders as well as affective dysfunction [30]. Our lab has previously shown that deletion of the gene Creb in the ventral hippocampus attenuates nicotine induced anxiety like phenotypes [31], making it a potential driver of observed differences in adolescent and adult withdrawal symptoms in humans. We find no differences in Creb mRNA expression across the 5 age cohorts [Figure 1A], which is not unexpected, as there is tight homeostatic control of this transcription factor in the brain [for review [30]]. While the expression levels of a genomic regulator, such as CREB, may not change, altered activity patterns of the protein can have distinct effects on gene expression of it’s targets. For example, many of the Neuregulin Signaling Pathway members possess CREB binding sites near their promoters, including both Nrg1 and Nrg3. When examining expression patterns of Nrg1 and Nrg3, their mRNA expression was lower in the 20 week cohort compared to the 4 week (F(4,27) = 1.0007 P<0.05, F(4,27)= 0.1866 P<0.01) [Figure 1B–C]. These changes were observed in the expression patterns of their cognate receptors, ErbB4 and ErbB2, as well. Nrg3-ErbB4 signaling, specifically, plays an important role in mediating anxiety-like behaviors associated with nicotine withdrawal. ErbB4 mRNA expression was lower in the week 12 group compared to the 8 week or 4 week (F(4,34)=3.003 P<0.01, P<0.05), but the 20 week is not significantly different from the 4 week cohort [Figure 1E]. ErbB2 is a dimerization partner with ErbB4 and can bind both Nrg1 and Nrg3. The 20-week cohort had lower mRNA expression than the 12-week group (P<0.05), and the 20-week cohort was trending to be significantly lower than the 4-week (p=0.0789 [Figure 1D]). A table including sex specific expression of each gene of interest is in Supplemental table 1.

Fig. 1. Neuregulin Signaling Pathway gene expression decreases with age in the hippocampus.

(A) Bar graph shows qPCR analysis of transcription factor Creb mRNA expression in hippocampus of 4-, 8-, 12-, 16-, and 20-week mice. (B) Graphical representation of which neuregulin molecules bind to specific ErbB dimers. (C-F) Bar graphs show qPCR analysis of neuregulin signaling pathway genes mRNA expression in hippocampus of 4-, 8-, 12-, 16-, and 20-week mice. [n = 5 to 10 per age group; error bars are SE; * - P<0.05, **P<0.01]

Hippocampal Gene Expression Changes in Nrg3, ErbB4, and their Regulators Due to Chronic Nicotine Treatment and Withdrawal During Adolescence and Adulthood

While baseline differences across age may underpin critical developmental changes in the hippocampus, many psychiatric conditions, including nicotine use disorder and schizophrenia, are related to the drug exposure during adolescence [6, 32–34]. Because of this, we chose to investigate the effects of chronic nicotine and acute nicotine withdrawal during adolescence or adulthood on hippocampal mRNA expression changes in the Neuregulin Signaling Pathway. To evaluate age-related differences in mRNA expression during chronic nicotine and withdrawal, mice were administered saline or 18 mg/kg/day of nicotine via osmotic minipump implantation for 2 weeks. Following chronic treatment, a 24h withdrawal was initiated by surgical removal of implanted pumps.

We found that nicotine treatment or acute nicotine withdrawal (24h WD) had no significant effect on Nrg1 mRNA levels in either the adolescent- or adult-exposed cohorts, but there was a main effect of age (Figure 2A; F(2,24)=0.8537 P>0.05, F(1, 34)= 16.83 P<0.001). This contrasted with Nrg3 mRNA levels, which demonstrated a treatment x age effect at 24h WD, with animals exposed to nicotine during adulthood expressing significantly less Nrg3 mRNA during 24h WD (F(1,38)=8.781 P<0.01). While the treatment effects of nicotine and withdrawal were somewhat modest on Nrg1 and Nrg3 expression, chronic nicotine and 24h WD had distinct effects on ErbB4 mRNA expression levels. As shown in Figure 2C, while nicotine and withdrawal had little effect on hippocampal ErbB4 levels during adolescence, withdrawal from chronic nicotine significantly reduced ErbB4 expression in the adult hippocampus compared to saline and chronic nicotine (F(2,36)=5.775, P<0.01). These effects were also in contrast to ErbB2 expression, which remained unaltered by treatment at either age.

Fig. 2. Nicotine withdraw impacts hippocampal Neuregulin Signaling Pathway gene expression in age specific manner.

(A-F) Bar graphs show qPCR analysis of neuregulin signaling pathway genes mRNA expression in hippocampus of 4- and 20- week Sal, Nic, and WD mice. [n = 5 to 10 per treatment; error bars are SEM; Compared to 4 week: * - P<0.05, *** - P<0.001, **** P<0.0001; Compared to saline within age: # - P<0.05, Compared to nicotine within age: ! - P<0.05,]

Due to the circumscribed nature of changes due to nicotine treatment and withdrawal on Nrg3 and ErbB4, we investigated other potential transcriptomic regulators of these NSP family members. The gene Nrg3 possesses a long-non-coding RNA (lncRNA) on the opposite strand of the DNA, which is present in both our model system as well as in humans. lncRNAs in this configuration (termed opposite strand lncRNAs), have been found to impact the t_1/2 of the target mRNAs [for review, see REF [35]]. To query whether this mechanism may be involved in the regulation of Nrg3, we designed primers to the lncRNA Nrg3os. Our findings show that Nrg3os, similar to its target mRNA Nrg3, also demonstrated a treatment x age effect at 24h WD, with animals exposed to nicotine during adulthood expressing significantly less Nrg3os mRNA during 24h WD (Figure 2E; F(1,38)=11.83, P<0.01). While ErbB4 does not possess a known lncRNA in the opposite strand configuration, ErbB4 does undergo significant alternative splicing regulated by the lncRNA Gomafu. Gomafu has been shown to be induced during neuronal activity and it’s specific regulation of ErbB4 splice variants has been posited as a major mechanism underlying schizophrenia. In contrast to the lncRNA Nrg3os, however, we saw no significant changes in Gomafu expression due to age or treatment (Figure 2F). A table including sex specific expression of each gene of interest is in Supplemental table 2.

Protein Expression Changes in Nrg3 and ErbB4 Due to Chronic Nicotine Treatment and Withdrawal During Adolescence (4wk) and Adulthood (20+wk)

While changes in transcript expression can suggest alterations in neuronal signaling, protein analysis of cell lysates or synaptically enriched lysates can further validate any observed mRNA changes and ensure that these altered transcript levels can result in functional changes in the cell. Furthermore, while whole cell lysates can indicate important changes, synaptically-associated proteins, such as NRG3 and ErbB4, sometimes display selective enrichment only at the synapse. While both NRG3 and ErbB4 are synaptically enriched proteins, these proteins can also be detected in the cytosol as they are trafficked to the cell membrane, and ErbB4 activation can also result in cleavage fragments being detected in the cytosol. Therefore, our next steps were to evaluate protein levels of NRG3 and ErbB4 both at the synapse and in the cytosol.

To do this, we isolated synaptosomal fractions from the contralateral hippocampus and then extracted protein from both the synaptosomal and cytosolic fractions for western blot analysis. Synaptosomal NRG3 protein levels exhibited a main age effect (F(1,45)=15.06, P<0.001), with the adult treatment group possessing lower levels of NRG3 at the synapse. This was most apparent in the 24h WD treatment groups, where the withdrawal treatment effects on Nrg3 had opposite directionality (Figure 3A). These synaptic NRG3 protein levels mirror our Nrg3 qPCR data, which also displayed a decrease in Nrg3 in the 20 week withdrawal group (Figure 2B). However, no changes were detected in the cytosolic concentration of NRG3 (Figure 3B). Conversely, while ErbB4 protein levels were unchanged in the synaptic fraction (Figure 3C), cytosolic levels of ErbB4 were significantly increased in the adolescent group following 24h WD from nicotine, but unchanged in the adult treatment group (Figure 3D; F(2,43)=4.294, P<0.05). A table including sex specific expression of cytosolic and synaptosomal NRG3 and ERBB4 in Supplemental table 3.

Fig. 3. NRG3 and ERBB4 protein expression in isolated synaptosome and cytosol.

(A-B) Bar graph shows western blot analysis of protein expression in isolated synaptosomes of hippocampal neurons with representative images of western blots below. (C-D) Bar graph shows western blot analysis of protein expression in isolated cytosol of hippocampal neurons with representative images of western blots below. [n = 7 to 10 per treatment; error bars are SEM; * - P<0.05]

Affective Behavioral Changes Due to Chronic Nicotine Treatment and Withdrawal During Adolescence (4wk) and Adulthood (20+wk)

Affective dysfunction, in particular increased anxiety, is a hallmark of nicotine withdrawal in smokers [36, 37]. In rodents, the open field (OF) test is a well-validated, quantitative, and reproducible model to measure anxiety-like behaviors, which has been utilized in both adolescent and adult animals [26]. Because of these considerations, we chose to utilize the OF test in both age cohorts following chronic treatment with nicotine and 24h withdrawal. Both age (F(1,46)=28.19 P<0.0001) and treatment (F(2,46)=11.49 P<0.0001) had significant main effects in this test (Figure 4A). Overall, the adolescent cohort spent significantly less time in the center of the arena, indicating an anxiogenic-like response compared to the older animals. Additionally, all animals undergoing 24h WD, regardless of age, also displayed an anxiogenic-like behavior, with reduced time spent in the center compared to their saline or nicotine treated cage-mates. These effects were in the absence of any locomotor effects due to age or treatment (Figure 4B). Sex specific behavioral effects of chronic saline, nicotine, and withdrawal are included in Supplemental table 3. Additionally, we investigated locomotion in five-minute bins for the open field test. We found that in a three way repeated measures ANOVA that there was a main effect of time, with the mice moving more during the first 5 minutes, but time had no effect on precent time in the center of the field (Figure 1S; (F(1,48)=228.5, P<.0001); (F(1,48)=0.79, P>.05)

Fig. 4. Nrg3 and Erbb4 mRNA expression correlates with anxiety like phenotypes.

(A) OF: Bar chart showing precent time spent in the center of the OF arena by 4- and 20- week and Sal, Nic, and WD mice. (B) Locomotor activity: Bar chart showing average distance traveled in the OF arena by 4- and 20- week and Sal, Nic, and WD mice. (C) Correlation of Nrg3 mRNA expression to precent time spent in the center of the OF arena for 4- and 20-week-old Sal, Nic, and WD treated mice. (D) Correlation of Erbb4 mRNA expression to precent time spent in the center of the OF arena for 4- and 20-week-old Sal, Nic, and WD treated mice. [n = 8 to 10 per treatment; error bars are SEM; Compared to 4 week: **** - P<0.0001; Compared to saline within age: # - P<0.05, ## - P<0.01, Compared to nicotine within age: ! - P<0.05]

We next correlated OF test results with Nrg3 and ErbB4 mRNA expression to evaluate the relationship between anxiety-like behaviors and the NSP in both adult and adolescent mice. These analyses indicated that increasing percent time in the center of the open field test correlates to a decreased expression of both Nrg3 and Erbb4 (F(1,38)=5.762 R²=0.1317 P<0.05; F(1,39)=8.654 R²=0.1816 P<0.01) [Figure 4C–D]. This relationship would suggest that overexpression of Nrg3 or Erbb4 mRNA in the whole hippocampus contributes to anxiety-like behaviors. We did not have sufficient power to correlate males or females specifically to Nrg3 and ErbB4 mRNA expression, but we specified sex in Figure 4C–D.

Discussion

Nicotine dependence during adolescence is associated with numerous health-related outcomes later in life, including significantly increased levels of nicotine dependence and higher risk for other psychiatric disorders [6, 34, 38, 39]. However, the mechanistic processes underlying these associations is not well defined. In this study, we demonstrate that elements of the neuregulin signaling pathway are differentially impacted by nicotine exposure as a function of developmental age. Furthermore, given that NRG3-ErbB4 signaling in particular has been implicated in smoking cessation and schizophrenia in the patient population, these findings suggest that increased examination of this pathway during late development may lead to better therapeutics for individuals with co-morbid nicotine dependence and schizophrenia.

NSP Expression During Late Development and Adulthood

Adolescence marks a particularly vulnerable time for the development of substance use disorders or other psychiatric disorders due to the many changes in the developing brain. The Neuregulin Signaling Pathway (NSP) plays an important role in neurodevelopment and adulthood via regulating assembly of neural circuitry, myelination, neurotransmission, synaptic plasticity, and stabilizing synapses (for review see refs [12, 14]). In rodents, adolescence is typically defined as post-natal week 4 to 6, because there are distinct behavioral changes and sexual maturation during this period. However, neurobiological changes associated with adolescence can be identified as early as post-natal week 3 in females and as late as post-natal week 8 in males (for review see [40]). For example, hippocampal synapse formation in rodents has been found to continue postnatally and peak around postnatal week 8 during adolescence, which is followed a period of synaptic pruning into early adulthood [41]. Given that both NRG1 and NRG3 promote excitatory synapse formation and stabilization on GABAergic interneurons via ErbB4 [13, 42], the observed developmental decrease in Nrg1, Nrg3, Erbb4, and Erbb2 mRNA expression during late adolescence could in part be due to the decrease in hippocampal synapse formation.

Besides the role in synapse formation and stabilization, the NSP also plays a critical role in myelination. In humans, robust myelination has been reported during adolescence in the subiculum, which is a part of the extended hippocampal formation [43]. Additionally, postmortem tissues from individuals diagnosed with schizophrenia show decreased myelination across the entire hippocampal structure [44]. In rodents, hippocampal myelination reaches adult levels around postnatal week 4 [45]. This observation may also underlie the higher mRNA expression of NSP genes observed during adolescence compared to adulthood, given this pathway’s role in myelination. Oligodendrocytes express both ErbB2 and ErbB4 receptors, and signaling through these receptors promotes myelination of axons [46–49]. While NRG3 is not generally associated with myelination, in vitro studies have shown that NRG3 treatment of serum-deprived oligodendrocyte-precursor-enriched primary cultures decreased cell death by more than half, suggesting a role in oligodendrocyte health [50]. Therefore, increased mRNA expression of NSP members during early adolescence may reflect their impact on higher levels of myelination during postnatal week 4.

Differential Effects of Nicotine and 24h Withdrawal on NSP Expression between Adolescents and Adults.

Aside from baseline expression differences, 24h nicotine withdrawal had the most divergent effects on mRNA and protein expression levels of NSP members. For example, mRNA levels of Nrg3 were significantly different during WD in adolescents compared to adults and this extended to differential NRG3 protein expression in the synapse during nicotine withdrawal in both adolescents and adults [Figures 2B, 3A]. Similarly, ErbB4 expression was also dynamically altered during 24h WD from nicotine. In adolescent animals, Erbb4 mRNA was unchanged, but cytosolic protein levels of ErbB4 were significantly increased. In contrast, 24h nicotine WD in adults resulted in significantly decreased mRNA expression with no change in either synaptosomal or cytosolic ErbB4 protein levels, suggesting regulation of the steady state transcript levels only [Figures 2C, 3C–D]. While complex, these relationships may underlie differential effects of nicotine on future dependence behaviors. For example, the vast majority of smokers, 87%, begin during adolescence, and smokers who start during adolescence are more than twice as likely to become a heavy smoker while being half as likely to quit 10 years after initiation compared to smokers who started during adulthood [6]. The divergent effects of nicotine withdrawal on NRG3 and ErbB4 expression during adolescence may in part underlie the differential risk for dependence in adolescents and adults. This is supported by human genetic studies identifying single nucleotide polymorphisms in the genes ErbB4 and Nrg3, which would be present throughout development, that predict [16] or correlate [18] with smoking cessation outcomes.

The higher levels of NRG3 synaptosomal protein in the adolescent 24h WD cohort, relative to the adult 24h withdrawal cohort, may be driving the increase in cytosolic Erbb4 protein levels. The ErbB4 noncanonical forward signaling pathway suggests that the c-terminal intracellular domain of ErbB4 can be cleaved creating an 80 kda fragment that is translocated to the nucleus to regulate transcription.[51–53] While this mechanism is appealing, we did not observe any changes in 80 kda Erbb4 fragments (data not shown), suggesting this not the case. However, in cultured hippocampal neurons activation of ErbB4 via NRG1 exposure caused an internalization of the ErbB4 receptor.[54] Furthermore, blocking NRG1 mediated ErbB4 endocytosis in cultured hippocampal neurons prevents phosphorylation of Erk and Akt, established downstream pathways of ErbB4.[55] Perhaps during adolescent nicotine withdrawal NRG3 is mediating Erbb4 endocytosis and altering the Ras-MAPK and PI3k-Akt pathways in the hippocampus. The hippocampus, specifically the ventral hippocampus, mediates affective responding through projections to both the prefrontal cortex and amygdala, and perturbation in this circuit via increased ErbB4 endocytosis during adolescence could cause long term changes in affective responding. A previous study found that overactivation of ErbB4 via systemic NRG3-EGF in preadolescent mice produces increased anxiety like phenotypes during adulthood. [56] An increased activation and internalization of Erbb4 during the later developmental period of adolescence could underlie the increased risk for anxiety disorders observed in adults who began smoking during adolescence.

Regulatory mechanisms of the NSP: effects of nicotine and withdrawal (ncRNA)

LncRNA are a diverse type of non-coding RNA that can regulate translation, post-translational changes, and stability of mRNA (for review see reference [35]). Given the emerging emphasis on lncRNAs’ role in regulating gene expression we investigated Gomafu and Nrg3os expression, both of which have potential to impact NSP member expression and function. Down regulation of Gomafu expression has been observed in post-mortem tissue of individuals with schizophrenia, and cultured hippocampal cells modulating Gomafu levels alter total, cytosolic, and synaptosomal ErbB4 levels [57]. Studies have shown that Gomafu accomplishes this by directing alternative splicing of ErbB4 [57, 58]. These alternative Erbb4 splice variants can influence the downstream signaling cascade, such as mediating specific coupling to SH2 and WW domain-containing proteins [59]. However, while Gomafu was trending towards a main age effect (P=0.0622), we did not observe any significant treatment effects in Gomafu expression in adult or adolescent mice. The observed trend towards higher Gomafu expression in adolescent animals may partially underlie the increase in cytosolic Erbb4 protein levels in adolescent mice.

Additionally, we explored the expression of Nrg3os. Little is known about Nrg3os’ role, but other opposite strand lncRNAs are known to regulate their paired coding partner (ie, Nrg3). For example, when BDNFos lncRNA expression is upregulated, BDNF mRNA is down regulated [60]. However, other lncRNA partners have the opposite effect. For example, BACE1as expression leads to an increase in BACE1 mRNA expression, which occurs via BASE1as competitive blockade of miRNA-induced repression of BACE1 [61]. Given that expression of both Nrg3 and its antisense lncRNA, Nrg3os, have the same directionality across development, this suggests that Nrg3os lncRNA may play a similar regulatory role as BACE1as.

Nrg3 and ErbB4 mRNA expression in the Hippocampus correlates with withdrawal phenotypes in adult mice and adolescent mice

Previous work has shown that adolescent animals have a higher baseline level of anxiety-like behaviors compared to adults [62–64]. Our work similarly shows an overall effect of age, with younger animals spending less time in the center of the open field arena (Figure 4A), with no observed differences in locomotion based on age or treatment, suggesting that hyperlocomotion caused by age or psychostimulant effects of nicotine is not driving the observed differences. This observation also correlates with increased overall expression of Nrg3 and ErbB4 mRNA (Figure 4C–D). Our group previously found that inhibiting NRG3-ErbB4 signaling during acute nicotine withdrawal in adults ablates nicotine withdrawal phenotypes,[16] further supporting this association. This correlational finding can be extended to nicotine treatment and withdrawal in these age groups as well. We find that both adolescent and adult animals display affective disruption during nicotine withdrawal, as tested in the open field test [Fig 4A].

About 50% of adult smokers [65] and 31% of adolescent smokers [66] attempting to quit reported increased anxiety during nicotine withdrawal. However, previous findings from other groups have reported that adolescent rodents do not display increased anxiety-like phenotypes during nicotine withdrawal [67, 68]. It is important to note, though, that all of these studies utilized the precipitated model of nicotine withdrawal, which often gives divergent effects from the more translationally relevant spontaneous withdrawal methodology [24]. In our study, which utilized spontaneous nicotine withdrawal, we found that both the 4- and 20-week withdrawal cohorts demonstrated anxiety-like phenotypes compared to their saline controls in the open field test. While our results counter the previous rodent adolescent studies utilizing precipitated nicotine withdrawal via mecamylamine, our translational results utilizing spontaneous withdrawal are consistent with clinical data as well as epidemiological studies suggesting that nicotine dependence increases the risk of anxiety disorders and panic attack [69]. While the continuous release of nicotine via osmotic minipump can mimic peak nicotine levels in humans, this administration paradigm does not necessarily replicate daily nicotine intake in the human population. However, nicotine withdrawal symptoms, such as increased anxiety, begin 1 to 3 days after cessation, which we replicated in the open field test in both the adolescent and adult 24h nicotine withdrawal cohorts. [70] Furthermore, these behaviors were inversely related to message for Nrg3 and ErbB4, where subjects spending less time exploring the center of the arena had higher expression of Nrg3 and ErbB4 mRNA. Given the that our lab and others have previously demonstrated that disrupting the NSP ameliorates anxiety like behavior associated with nicotine withdrawal, [16, 71] our findings suggest that Nrg3 and ErbB4 expression in the hippocampus play a crucial role in mediating anxiety-like behaviors, particularly during nicotine withdrawal. Previous studies from our group suggest that CREB mediated transcription of NSP genes in the ventral, but not dorsal, hippocampus mediates anxiety-like phenotypes during nicotine withdrawal.[31] However, affective dysfunction caused by nicotine withdrawal can also include depressive symptoms, and adult male rats experienced decreased neurogenesis in the dorsal hippocampus following the learned helplessness model for depression.[72] Given the NSP’s role in neurogenesis via neuronal migration along glial fibers in the developing cerebellum, perhaps alterations of dorsal hippocampal NSP caused by chronic nicotine withdrawal underlie depressive symptoms experienced during nicotine withdrawal, but not anxiety related symptoms.[73, 74] Future studies are needed to examine the age- and regiospecific effects of chronic nicotine and withdrawal in the hippocampus.

Neuregulin signaling pathway as target for future smoking cessation therapies

Nicotine withdrawal is characterized by craving as well as many affective disfunctions, such as anxiety, depressions, anger and difficulty concentrating, with symptoms peaking during the first week of cessations but can last up to 4 weeks. [70] Additionally, nicotine use disorder has a high rate of co-morbidity with many affect disorders, such as 60–90% of people suffering from schizophrenia also smoke. [8–11] Furthermore, mutations in Nrg3 and ErbB4 have been implicated in smoking cessation outcomes and schizophrenia. [15, 16, 18, 75] Smoking during adolescence is linked to an increase risk of developing affective disorders such as anxiety and panic disorders. [69] While adolescence smoking rates have fallen in the United States, adolescent vaping has sharply risen, and adolescents that vape are twice as likely to report daily cigarette usage than those who don’t.[3–5] The mechanism underlying adolescent smokers’ risk of developing more sever nicotine dependance and anxiety related disorders is not well understood, however a previous study demonstrated systemic administration of Nrg3-EGF in mice, which only binds to Erbb4, during early development increases anxiety like phenotypes during adulthood. [56] The hippocampus has been heavily implicated in broader substance use disorder, as well an many neuropsychiatric disorders characterized by affective disfunction, including depression, schizophrenia, and post-traumatic stress disorder. The hippocampus is a locus for contextual memory and affective responding making it a prime target for associative learning in nicotine use disorder and nicotine withdrawal phenotypes. Smoking, especially during adolescence, may increase Erbb4 activation and endocytosis via Nrg3 in the hippocampus, altering LTP by activation of the Ras-MAPK and PI3k-Akt pathways, leading to the increased risk of adulthood nicotine dependence and development of affective disorders.

Nicotine exposure during adolescence produces numerous structural, neurochemical, and developmental changes not seen in adult (for review see ([76]). Genetic variation in Nrg3 in humans has been implicated in smoking and smoking cessation success [16, 18]. Layered above this genetic susceptibility is adolescent experimentation with and exposure to nicotine, which is on the rise due to widespread availability of e-cigarrettes [3, 4, 77]. Given that 87% of smokers start before 18 years old and the role the NSP plays (1) in adolescent brain development, (2) in nicotine withdrawal symptomology, and (3) in other psychiatric diseases with anxiety endophenotypes, the Neuregulin Signaling Pathway may play an important role in establishing nicotine dependance, especially among adolescents [1]. Further evaluation of the NSP within limbic structures, such as the hippocampus, may potentially lead to tailored pharmacotherapies that address anxiety-related withdrawal symptoms, which current FDA-approved therapies do not address [78, 79].

Data Availability Statement:

The data that supports these findings are available upon request.