Nicotine metabolites generate craniofacial defects that are worsened by ethanol exposure

Abstract

Nicotine and alcohol are two of the most consumed substances among adolescents and adults. There is great concern regarding the deleterious effects of exposure to these substances during pregnancy. Complex interactions between genes and the environment are thought to influence the severity of birth defects and can occur by altering the metabolism of substances. In humans, nicotine is mainly metabolized by CYP2A6 and variants within this enzyme associate with nicotine dependence and metabolism. Previously, we showed that embryonic nicotine exposure reduced the size of craniofacial cartilages in zebrafish. Here, we show that zebrafish metabolize nicotine similarly to humans. We demonstrate that it is not nicotine itself, but nicotine metabolites, particularly 4HPBA followed by nicotine-N′-oxide (NOX), that disrupt craniofacial development. We found that cyp2y3 mutants (the zebrafish homolog of CYP2A6) have reduced levels of nicotine metabolites and are partially protected against the craniofacial defects produced by nicotine. We also examined the effect of nicotine and ethanol co-exposure in zebrafish embryos. Surprisingly, while embryos co-exposed to nicotine and alcohol had reduced craniofacial elements, they were significantly less affected than those exposed to nicotine alone. We found that nicotine metabolites levels were decreased in embryos coexposed to nicotine and ethanol. However, alcohol worsened the craniofacial defects in embryos exposed to metabolites. Our results collectively show that nicotine is metabolized into teratogenic substances and that multifactorial interactions can alter this metabolism and subsequent teratogenicity.

Significance Statement.

Using a zebrafish model for embryonic nicotine exposure, we show that it is nicotine metabolites, not nicotine itself that cause craniofacial defects. We demonstrate that zebrafish metabolize nicotine into all the major metabolites found in humans and at levels comparable to humans. We find that mutation of cyp2y3 slows nicotine metabolism and is protective against nicotine exposure. We find that coexposure to alcohol also alters nicotine metabolism and renders exposure to the metabolites more deleterious. Our findings help explain the variability in defects caused by embryonic nicotine exposure across human populations and provide insight into the risks of the use of tobacco products, including vaping and nicotine patches, during pregnancy.

Introduction

Nicotine and alcohol are the most used legal drugs among adolescents and adults and are commonly consumed together. A global concern is the use of these substances during pregnancy and the potential toxicological effects on the developing embryo and fetus. Frequently, pregnant women use tobacco and alcohol in their first trimester before recognizing they are pregnant (1). According to the National Survey on Drug Use and Health (NSDUH, 2015–2018), 19.6% of pregnant women in the first trimester reported drinking, and 10.5% reported binge drinking (2). Likewise, the rate of the co-use of these substances is 5% or 1 in 20 pregnant women (3). Although the prevalence of smoking during pregnancy has decreased in recent years (4), there is a corresponding increase in vaping (5) and nicotine is the addictive substance present in both.

Nicotine can cross the placenta, reach the embryo, and accumulate in amniotic fluid and fetal blood (6, 7). Maternal smoking and secondhand smoke exposure increase the risk of stillbirth, miscarriage, preterm birth, low birth weight, sudden infant death syndrome, and congenital malformations (8–12). Also, smoking during pregnancy has been associated with hyperactivity, Attention Deficit hyperactive Disorder (ADHD) and negative and externalizing behaviors in children (13–17). Alcohol consumption during pregnancy can lead to Fetal Alcohol Spectrum Disorders (FASD). FASD comprises the range of disorders caused by alcohol exposure, including cognitive impairments, hyperactivity, learning disabilities, and ADHD (18). On the severe end of FASD is Fetal Alcohol Syndrome (FAS), which, in addition to neurological abnormalities, includes craniofacial defects (19–21). Reports indicate that children exposed to nicotine during development can have similar cognitive and behavioral problems as prenatal alcohol exposure (22–26). Not all nicotine and alcohol exposures during pregnancy result in deleterious effects on the embryo. This suggests that other factors, such as genetics and gene-environment interaction, influence the embryo's susceptibility to these substances. Particularly, genetic variation that alter the metabolism of teratogens could mediate developmental outcomes.

Nicotine is the main psychoactive component of tobacco smoke (27). In humans, it is primarily metabolized in the liver, CYP2A6 is the major enzyme involved and cotinine is the predominant metabolite. Briefly, 90% of nicotine is oxidized to cotinine by CYP2A6 with the remaining 10% being metabolized by CYP2B6. Subsequently, cotinine is metabolized to trans-3′-hydroxycotinine (3HC) solely by CYP2A6. Nicotine and cotinine are also N′-oxidized to nicotine-N′-oxide (NOX) by the flavin monooxygenase (FMO)1 and FMO3 enzymes, and to cotinine-N′-oxide (COX) by CYP2A6 and CYP2C19, respectively (28–30). Another nicotine metabolite is 4-hydroxy-4-(3-pyridyl) butanoic acid (4HPBA). It has been suggested that CYP2A6 contributes to this reaction, however, its involvement is not definitive (31). Nicotine, cotinine, and 3HC are glucuronized to nicotine-glucuronide, cotinine glucuronide, and 3HC-glucuronide by UDP-glucuronosyltransferases (UGTs) enzymes (28) (Fig. 1). In all, several steps of nicotine metabolism depend on CYP2A6. Thus, altering CYP2A6 enzyme activity is expected to change the levels of many metabolites either directly or indirectly.

Figure 1.

Nicotine metabolism in humans. Brackets indicate that CYP2A6 has been implicated but not conclusively demonstrated to convert nicotine to 4HPBA.

Genetic variation in nicotine metabolizing enzymes modifies enzyme activity, as well as the physiological levels of nicotine and its metabolites, modulating their psychoactive effects (32–35). Genome-wide association studies have demonstrated an association of the CYP2A6 gene with smoking-related behaviors (36, 37). CYP2A6 is a highly polymorphic gene and these variants influence nicotine metabolism and tobacco consumption (38, 39). A considerable number of alleles have been identified (40). These variants have been phenotypically classified as normal, slow, intermediate, and rapid metabolizers (41, 42). In particular, rapid metabolizers, are more likely to become heavy smokers and experience more difficulty in cessation attempts (43). Slow metabolizers smoke fewer cigarettes per day; however, they are at greater risk of developing tobacco dependence during adolescence (44–46).

While smoking is a risk factor for birth defects, it is unknown whether it is nicotine itself or a metabolite that is responsible for the teratogenic effect. Variations in metabolic enzymes could alter the concentrations of metabolites to which embryos are exposed and may alter the vulnerability to the toxic effects of nicotine. In addition, most exposures are polysubstance and one of the most common coexposures is between alcohol and smoking during pregnancy (47, 48). It has been suggested that the co-use of both substances might have a synergistic effect that could lead to an increased risk of birth defects (49, 50). However, little is known about the combined effects of these two substances. Understanding the pharmacological interaction between these substances and how genetics influences environmental factors is crucial for understanding the cause of birth defects.

Previously, we established zebrafish as an animal model for nicotine teratogenesis research (51). We demonstrated that zebrafish metabolized nicotine to cotinine and that nicotine causes craniofacial and behavioral defects (51). Here, we used zebrafish embryos to analyze the effect of Cyp2y3 loss of function, the teratogenesis of nicotine metabolites, and the interaction between nicotine and alcohol during embryonic development. Our results showed that the loss of Cyp2y3 function reduced nicotine metabolites and the teratogenic effect of nicotine, these metabolites caused craniofacial defects, and that alcohol reduced nicotine metabolism but elevated the teratogenic effect of its metabolites.

Results

Daily media changes prevent nicotine-induced craniofacial defects

In our previous analyses (51), we used a single chronic nicotine exposure without changing the embryo media (EM). Thus, the concentration of nicotine could be reduced, and metabolites could accumulate over time. To test if this effected exposure outcomes, embryos were exposed to 50 µM nicotine at 6 hours postfertilization (hpf) and the nicotine solution was replaced every day for 4 days. We measured the length of the viscerocranial and neurocranial cartilages as shown in Fig. 2E and F using ImageJ. As we reported previously, embryos exposed to a single nicotine dose for 4 days showed decreased cartilage sizes (Fig. 2C, C′, D, D′, and G–N). However, no significant differences were found in cartilage size in the daily nicotine-exposed group versus the control group (Fig. 2A, A′–B, B′, and G–N). Only the inter-ocular distance (IOD) was decreased in both daily and single dosed embryos (P < 0.0001) (Fig. 2N) (Table S1 shows full ANOVA results). This result would suggest that nicotine is not the cause of craniofacial defects at the dose that we use, but that there is a chronic buildup of teratogenic metabolites in the EM which combines with newly synthesized metabolites to disrupt craniofacial development. Thus, we quantified the levels of nicotine metabolites in the EM at 4 days after chronic exposure, and in the group with daily EM changes. We found all of the nicotine metabolites at quantifiable levels in the embryo media of those exposed to a chronic dose for 4 days and significant decreased levels of all nicotine metabolites, in the daily medium changes group (Fig. 2O–S). This suggests that nicotine metabolites are teratogenic. Unexposed embryos were tested to set the baseline for our analyses. There was no chromatographic signal for any of the metabolites in unexposed embryos.

Figure 2.

Daily media changes restore normal craniofacial development and reduce the levels of nicotine metabolites in the media. A–D) Neurocranium images. A′–D′) Viscerocranium images. A, A′, B, and B′) Embryos were treated with a fresh solution every day for 4 days. C, C′, D, and D′) Embryos were treated with a single exposure for 4 days. E) Scheme showing viscerocranium and F) neurocranium cartilages analyzed. G–N) Craniofacial measurements. O–S) Nicotine metabolite levels in embryo media in daily media changes and full exposure at 4 dpf. E and F) Craniofacial scheme. M, Meckel's cartilage; C, ceratohyal; EPW, ethmoid plate width; N, neurocranium; PN, posterior neurocranium; AN, anterior neurocranium; IOD, interocular distance. EPW, ethmoid plate width; PN, posterior neurocranium; AN, anterior neurocranium; IOD, interocular distance. Sample sizes: DMSO n = 24; Single dosing n = 25; Daily dosing n = 23. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Scale bar 100 μm.

Nicotine metabolites cause craniofacial defects

To determine which nicotine metabolites were teratogenic, we exposed embryos to 50 µM of each metabolite separately from 6 hpf to 4 days postfertilization (dpf) (Fig. 3A, A′–F, and F′). Embryos treated with 50 µM 4HPBA were the most affected with significant reductions in the length of Meckle's cartilage (P < 0.0001 (Fig. 3G), the neurocranium (P < 0.0001) (Fig. 3K), the anterior neurocranium (P < 0.0001) (Fig. 3M), the posterior neurocranium (P = 0.0329) (Fig. 3L), and the width of the ethmoid plate (P = 0.0168) (Fig. 3I). The next most teratogenic metabolite was NOX, with reductions to Meckle's cartilage (P = 0.0029) (Fig. 3G), ethmoid plate width (P = 0.0003) (Fig. 3I), and the ceratohyal (P = 0.0301) (Fig. 3H). However, no significant differences, relative to control, were found in the groups exposed to cotinine, 3HC, and cotinine-oxide. Also, no differences were found in the length of trabeculae and the IOD in any group (Fig. 3J and N). Table S1 shows full ANOVA results. We tested co-exposures of 4HPBA and NOX at both 25 and 50 µM concentration. Surprisingly, we did not find the co-exposure to be more teratogenic than the single exposure (Figs. S1A–L and S2A–L). We do not understand why this is the case, but it would suggest that the overall teratogenicity of nicotine is more complex than just these two metabolites.

Figure 3.

The nicotine metabolites 4HPBA and NOX causes craniofacial defects. A–F) Neurocranium images. A′–F′) Viscerocranium images. Arrows indicate reduced Meckel's cartilage and ceratohyal in 4HPBA (E) and NOX (F) treated embryos. G–N) Craniofacial measurements of nicotine metabolites exposed embryos. EPW, ethmoid plate width; PN, posterior neurocranium; AN, anterior neurocranium; IOD, interocular distance. Sample sizes: DMSO n = 28; Cotinine n = 27; 3HC n = 29; 4HPBA n = 30; COX n = 29; NOX n = 22. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Scale bar 100 μm.

We next assessed the window of sensitivity to the most teratogenic nicotine metabolite, 4HPBA. We exposed embryos to 50 µM 4HPBA from 6 hpf to 24 hpf, 6 hpf to 48 hpf, and from 40 hpf to 4 dpf. We observed that only the 40 hpf to 4 dpf exposure resulted in decreased cartilage length (Fig. S3A–L). We also quantified nicotine metabolites in fish at different windows of exposure following nicotine exposure. As expected, we found that absolute metabolite levels were higher in embryos with longer exposures (Fig. S3M–Q). We also calculated the concentration of each metabolite normalized to total nicotine equivalents (TNE), calculated as the molar sum of cotinine, 3HC, 4HPBA, NOX, and COX and expressed as the percentage of TNE (Fig. S3R–V). Normalized data showed a significant elevation of 4HPBA and NOX in the group of 40 hpf to 4 dpf and 6 hpf to 4 dpf. Thus, relative to other metabolites, the teratogenic metabolites (4HPBA and NOX) are the most abundant during the sensitive time window of exposure. These results support the idea that metabolite accumulation in the chronic exposure play a role in the teratogenic effect of nicotine. Given our findings, we tested if 4HPBA produces craniofacial defects when the media was replaced daily. Cartilage lengths were reduced in embryos exposed daily to 4HPBA, similarly to those exposed to a single 4HPBA dose for 4 days (P < 0.0001) (Fig. S4). No differences were observed in the IOD in embryos exposed to daily doses as well as those exposed to a single dose of 4HPBA (Fig. S4J). Collectively, these results demonstrate that nicotine does not cause craniofacial defects itself at the doses used but rather it is nicotine metabolites, particularly 4HPBA and NOX that cause these defects. Somewhat surprisingly, the effect on in IOD is independent of this effect and may be due to a direct effect of nicotine.

Generation of cyp2y3 mutant zebrafish

In humans, nicotine metabolism is mainly mediated by CYP2A6 enzyme. In zebrafish, there is no annotated cyp2a6 gene, so to identify potential nicotine metabolizing enzymes we searched for zebrafish proteins with homology to human CYP2A6. The top hit from a BLAST search of zebrafish using the CYP2A6 protein was Cyp2y3 with 50% identity and 68% positives. Phylogenic analyses grouped Cyp2y3 with CYP2A6, CYP2A7, and CYP2A13 human proteins (Fig. S5A) and all share between 91–94% identity with one another and between 48% and 50% identity with Cyp2y3 (Fig. S5B), this strongly suggests that the zebrafish homolog of Cyp2A6 is Cyp2y3. Zebrafish embryos express cyp2y3 as early as bud stage (https://zfin.org/ZDB-FIG-050630-12968). Nawaji et al. (52), found higher cyp2y3 expression at 48 and 72 hpf. Here, we show cyp2y3 expression by in situ hybridization at 14-somites to 4 dpf (Fig. S6). Prior to the formation of the liver, we find that the periderm expresses cyp2y3 (Fig. S6A–C). It is then strongly expressed in the liver (Fig. S6D). We generated a 68 bp deletion, designated au131, using CRISPR-Cas9 (Fig. S7B). This deletion is predicted to induce a frameshift in the coding sequence at amino acid 86 and a STOP codon at position 90, disrupting the translation of the second exon of cyp2y3 and deleting most of the p450 domain (amino acids 40–487) (Fig. S7C). Thus, the mutation is predicted to abolish enzyme activity and result in a functional null mutation. Details of the generation and validation of the cyp2y3  ^au131 allele is shown Fig. S7.

Nicotine metabolite concentrations are reduced in cyp2y3 mutants

To verify the involvement of Cyp2y3 in nicotine metabolism, we quantified the levels of the most abundant metabolites (cotinine, 3HC, NOX, COX, and 4HPBA) in cyp2y3 mutants and wild types at two different time windows. Embryos were exposed to 50 µM nicotine at 4 dpf and collected after a 6-h exposure to assess short metabolism that is likely driven by the liver. We also quantified the levels in fish treated from 6 hpf to 4 dpf, to match our treatment window. We find that zebrafish tissue contains metabolites found in humans and that nicotine metabolism requires cyp2y3. In cyp2y3 mutants exposed to nicotine for 6 h, we found decreased levels of 3HC (P = 0.0172) and increased cotinine levels (P = 0.0035). Likewise, in mutants, we observed increased levels of COX and 4HPBA (P = 0.0161 and P < 0.0001, respectively). No differences were found in the NOX levels after 6 h of nicotine exposure (Figs. 2E–4A). The long exposure to nicotine resulted in decreased levels of all nicotine metabolites in mutants, relative to wild-type fish (Figs. 4G–7K). Thus, unlike short exposure, long nicotine exposure results in reduced overall metabolism of nicotine in cyp2y3 mutants.

Figure 4.

Loss of cyp2y3 reduces metabolite levels in fish with prolonged nicotine exposure and partially protects against nicotine-induced craniofacial defects. A–E) Levels of nicotine metabolites in mutant and wild-type zebrafish embryos exposed to nicotine at 4 dpf for 6 h, and G–K) from 6 hpf to 4 dpf. F) Nicotine metabolite ratio after 6 h of nicotine exposure and (L) from 6 hpf to 4 dpf. M, N, O, P) Wild-type embryos. M′, N′, O′, P′) cyp2y3 mutant embryos. M, M′, O, O′) Control embryos. N, N′, P, P′) Nicotine-treated embryos. Q–X) Craniofacial measurements. Red arrows indicate reduced Meckel's cartilage and ceratohyal in nicotine exposed wild-types and less altered in mutants. Sample sizes: DMSO n_wt = 14; DMSO n_mut = 13; 50 µM Nicotine n_wt = 13; 50 µM Nicotine n_mut = 13. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Scale bar 100 μm.

Figure 7.

Summary model of craniofacial defects produced by nicotine and ethanol exposure. A) Loss of function of cyp2y3 does not affect development. B) Nicotine-exposed zebrafish embryos produce nicotine metabolites that affect craniofacial development. C) cyp2y3 mutant embryos exposed to nicotine produced fewer nicotine metabolites, which partially prevent craniofacial defects. D) Wild-type embryos exposed to nicotine and 1% ethanol produce low levels of nicotine metabolites, which mitigates defects caused by nicotine. E) Nicotine metabolites combined with 1% ethanol caused a more significant effect than nicotine metabolites exposed alone.

In humans, the nicotine metabolite ratio (3HC/Cotinine) is used to measure CYP2A6 activity. A higher 3HC/Cotinine ratio indicates greater CYP2A6 enzyme activity and a faster rate of nicotine metabolism (53). We applied this measure to our zebrafish mutants to calculate the rate of nicotine metabolism. As expected, we found a lower nicotine metabolite ratio in cyp2y3 mutants compared to wild types when exposed to nicotine from 6 hpf to 4 days (P = 0.0001) (Fig. 4L), as well as after 6 h of nicotine exposure in 4 dpf embryos (P = 0.0045) (Fig. 4F). These data show that Cyp2y3 plays a similar role in zebrafish as CYP2A6 does in humans and provide biochemical validation of the cyp2y3 mutant zebrafish line.

Loss of cyp2y3 partially protects against craniofacial defects produced by nicotine

To assess the effect of cyp2y3 loss of function on nicotine teratogenesis, zebrafish embryos were exposed to 50 µM nicotine from 6 hpf to 4 dpf. Previously, we have shown that embryonic nicotine exposure decreases the length of craniofacial cartilages in fish exposed to nicotine from 6 hpf to 4 dpf, without affecting the overall size of the fish. We also found that embryonic nicotine exposure induces ventral curvature of the body (51). Interestingly, here we find that this curvature is reduced in nicotine-exposed cyp2y3 mutants (Fig. S8C–F). In controls, we found no differences between mutants and wild-types, demonstrating that cyp2y3 is not normally required for craniofacial development (Fig. 4Q–X). Likewise, the overall body length was not affected in cyp2y3 mutants (Fig. S8A and B). Consistent with our previous findings, nicotine exposure reduced cartilage size in wild-type embryos. However, exposed mutant embryos were significantly less affected than exposed wild types in the Meckel's cartilage (P = 0.0406) and the ceratohyal (P = 0.0450) (Fig. 4Q and R, respectively). Trabecula and posterior neurocranium length were reduced in nicotine-exposed wild-types versus the control group (P = 0.0408) but no differences were observed in nicotine-exposed mutants (P = 0.060 and P = 0.098, respectively) (Fig. 4S and V, respectively), suggesting that these cartilages were less affected by nicotine in the mutants. These results suggest that nicotine metabolites generated by Cyp2y3 participate in the genesis of craniofacial defects following nicotine exposure. It is unclear why loss of cyp2y3 was not uniformly protective, although we do note the trend. It is likely that other enzymes are involved in nicotine toxicity. In zebrafish, Meckel's cartilage and ceratohyal are ventral elements of the jaw and jaw support and defects in these are associated with feeding difficulties. They are generated in the first and second pharyngeal arches, respectively. In humans, Meckle's cartilage is also part of the embryonic lower jaw and contributes to the formation of the jaw and inner ear. There is no ceratohyal in humans, but the second pharyngeal arch contributes to the hyoid bone, which is important in the movement of the tongue and larynx.

Nicotine and ethanol coexposure affect the craniofacial skeleton in zebrafish embryos

Here, we analyze the craniofacial defects resulting from embryonic co-exposure to nicotine and alcohol. We exposed embryos to 1% ethanol, since our previous studies indicate that it does not cause gross craniofacial defects in zebrafish embryos (54–56). Embryos were exposed to different nicotine concentrations and were coexposed to nicotine and 1% ethanol. Consistent with our previous findings, embryonic nicotine exposure at 25 and 50 µM decreased the sizes of craniofacial cartilages in a dose-dependent manner (P < 0.0001) (Fig. 5A–D). Also consistent with our previous findings, all craniofacial measures were significantly reduced in embryos exposed to 50 µM nicotine (Fig. 5E–L). In the nicotine and ethanol coexposed groups, we observed a significant reduction in the length of craniofacial cartilages versus the control (P < 0.0001). Surprisingly though, the 25 and 50 µM nicotine and ethanol coexposed groups were significantly less affected than those exposed to nicotine alone (P < 0.001). In the case of the IOD both nicotine and nicotine coexposed groups were similarly affected at 50 µM, but the coexposed at 12.5 µM was more affected than the nicotine-exposed groups (Fig. 5L). No significant differences were observed between the DMSO and 1% ethanol control groups, except for the IOD. Thus, this suggests that nicotine and ethanol interact, but ethanol partially mitigates the biological consequences of nicotine exposure. (Table S3 shows full two-way ANOVA results.)

Figure 5.

A subteratogenic dose of ethanol improves craniofacial outcomes in nicotine-exposed fish and reduces the levels of teratogenic nicotine metabolites. A, B, C, D) Nicotine-exposed embryos. A′, B′, C′, D′) Nicotine and ethanol coexposed embryos. E–L) Craniofacial measurements of nicotine and nicotine and ethanol coexposed embryos. The X axis is the concentration of nicotine in the fish water. Gray dots refer to nicotine exposure alone and blue dots refer to nicotine coexposed with 1% ethanol. M–Q) Nicotine metabolites levels in fish exposed for 6 h. S–W) Levels of nicotine metabolites in fish exposed from 6 hpf to 4 days. R) Nicotine metabolite ratio after 6 h of nicotine and ethanol coexposure. X) Nicotine metabolite ratio of after nicotine and ethanol co-exposure from 6hpf to 4dpf. EPW, ethmoid plate width; PN, posterior neurocranium; AN, anterior neurocranium; IOD, inter-ocular distance. Sample sizes: DMSO n = 32; DMSO + 1% OH n = 31; 12.5 µM Nicotine n = 30; 12.5 µM Nicotine + 1% OH n = 30; 25 µM Nicotine n = 30; 25 µM Nicotine + 1% OH n = 33; 50 µM Nicotine n = 34; 50 µM Nicotine + 1% OH n = 33 *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Scale bar 100 μm.

We next tested the effect of higher ethanol concentration and its coexposure with nicotine. We found that embryos exposed to 1.5% ethanol developed edema and cartilage breaks, particularly trabecula breaks (Fig. S9A–K). We observed significantly more cartilage breaks in the group coexposed to 1.5% ethanol and 50 µM nicotine, relative to the 1.5% ethanol group (Fig. S9L). We observed that the craniofacial effects of nicotine and ethanol exposure are partially distinct, while both nicotine and ethanol affect cartilage lengths, 1.5% ethanol induces cartilage breaks and edema, nicotine reduces the IOD. We also find that coexposure to higher doses of both substances (1.5% ethanol and 100 µM nicotine) amplifies their toxicity, and the embryos do not survive beyond 3 dpf. Thus, the protective effect of ethanol is only at exposure levels that do not disrupt normal craniofacial development.

Nicotine metabolite concentrations are modified in embryos coexposed to nicotine and ethanol

Previous reports have suggested pharmacokinetic interaction between nicotine and alcohol (57, 58). In order to analyze the pharmacological effect of nicotine and ethanol co-exposure in embryos, we quantified nicotine metabolites and calculated the nicotine metabolic rate at two different time windows. Embryos were exposed to 50 µM nicotine at 4 dpf and collected after a 6-h exposure and from fish with a chronic exposure from 6 hpf to 4 dpf. We found significantly decreased levels of nicotine metabolites in embryos coexposed to nicotine and ethanol relative to nicotine alone in the shorter windows exposure, except for 3HC (Fig. 5M–Q). For the chronic coexposure, we found a significant decrease in the levels of all nicotine metabolites (Fig. 5S–W). We observed no significant differences in cyp2y3 expression during 1% ethanol exposure across developmental stages (Fig. S10), demonstrating that this effect is not due to altered transcription of cyp2y3. We calculated the nicotine metabolite ratio for each time point to determine if there was an effect on enzyme activity. Interestingly, the 6-h coexposure to nicotine and ethanol significantly increased the nicotine metabolite ratio (P = 0.0017) (Fig. 5R). This would indicate high enzyme activity and, therefore, high clearance rate during a short exposure. However, in the chronic exposure we found a decrease in the ratio in the coexposed group relative to the nicotine-exposed group (Fig. 5X) (P < 0.0001). Thus, a prolonged ethanol–nicotine coexposure resulted in a reduction of nicotine metabolites, which could explain the partially protective effect of ethanol.

Coexposure to nicotine metabolites and ethanol is more detrimental than exposure to nicotine metabolites alone

We have shown that nicotine metabolites are critical players in the development of craniofacial defects in nicotine-exposed embryos. We tested whether these metabolites combined with ethanol would have a more detrimental consequence than the metabolites alone. We exposed embryos from 6 hpf to 4 dpf to each nicotine metabolite alone or in combination with 1% ethanol. Consistent with our findings above, we found that 4HPBA was the most teratogenic metabolite. When coexposed with ethanol the length of Meckle's cartilage (P < 0.0001), ceratohyal cartilage (P < 0.0001), neurocranium (P = 0.0414), and IOD (P = 0.0013) were significantly reduced relative to 4HPBA alone. We also found that 3HC coexposed with ethanol reduces the length of EPW (P < 0.0001) and PN (P = 0.0463); COX coexposed with ethanol reduces the length of Meckle's cartilage (P = 0.0271), EPW (P < 0.0001), and ceratohyal (P = 0.0074), and NOX coexposed with ethanol affected the width of the EP (P = 0.0371) (Fig. 6). Thus, ethanol exacerbates the teratogenic effects of nicotine metabolites and renders some metabolites teratogenic that were not in single exposures. A full table of two-way ANOVA results can be found in Table S4.

Figure 6.

Nicotine metabolites coexposed with ethanol produce more detrimental effects than nicotine metabolites alone. A–F) Viscerocranium. (A′–F′) Neurocranium. A, A′) Control DMSO + 1% OH. B, B′, C, C′, D, D′, F, F′). Embryos co-exposed to nicotine metabolites and 1% OH. G–N) Craniofacial measurements of embryos exposed to nicotine metabolites (blue dots) and coexposed with ethanol (pink dots). EPW, ethmoid plate width; PN, posterior neurocranium; AN, anterior neurocranium; IOD, interocular distance. DMSO n = 23; DMSO + 1% OH n = 21; Cotinine n = 24; Cotinine + 1% OH n = 27; 3HC n = 24; 3HC + 1% OH n = 24; 4HPBA n = 24; 4HPBA + 1% OH n = 25; COX n = 24; COX + 1% OH = 18; NOX n = 18; NOX + 1% OH = 24. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Scale bar 100 μm.

Collectively, we find that cyp2y3 loss of function in zebrafish embryos does not affect development (Fig. 7A). Zebrafish embryos can generate nicotine metabolites and these metabolites cause craniofacial defects (Fig. 7B). Reducing the enzymatic activity of Cyp2y3, through mutation or ethanol exposure results in the generation of fewer nicotine metabolites, reducing the extent of craniofacial defects (Fig. 7C and D). However, nicotine metabolites combined with ethanol generate a more detrimental effect than the metabolite alone (Fig. 7E). Thus, the ability of an embryo to clear the metabolites would be a critical determinant of the teratogenicity of nicotine.

Discussion

Despite significant efforts to reduce cigarette consumption, smoking remains a global health problem. Most women continue smoking during pregnancy and find it hard to quit. Here, we use zebrafish to study nicotine teratogenesis and its pharmacological interaction with ethanol.

Zebrafish as a model of nicotine metabolism

We have provided the first characterization and quantification of nicotine metabolites in zebrafish. We demonstrate that zebrafish metabolize nicotine into the most abundant nicotine metabolites found in humans. These metabolites were found both in the zebrafish tissue and the embryo medium, demonstrating that zebrafish excrete the metabolites. In humans, these metabolites are typically measured in urine. Cotinine and 3HC have been measured in the plasma of smokers, which more accurately reflect tissue levels. Studies analyzing the levels of plasma cotinine have detected average levels of 174.9 to 229.7 ng/mL (59), and 146 ng/mL (60), with ranges of 56–538.1 ng/mL (61) and 17–499 ng/mL (60). 3HC levels averaged 44 ng/mL (60) with ranges of 0.5–138 ng/mL (60) and 15–236 ng/mL (61). The levels of cotinine and 3HC that we found in zebrafish tissue (171.4 and 1.975 ng/mL, respectively), is well within the range found in human smokers. Thus, following nicotine exposure, the tissues of the zebrafish embryo are exposed to physiologically relevant levels of the metabolites.

In humans, CYP2A6 is responsible for 90% of the metabolism of nicotine to cotinine and CYP2B6 for the remaining 10%. CYP2B6 contribution is more significant when levels of nicotine are higher (62). CYP2A13 have also been shown to metabolize nicotine, however CYP2A13 is more abundant in extrahepatic tissues like the lungs (63). CYP2A6 exclusively converts cotinine to 3HC. Genetic variants that reduce CYP2A6 enzyme activity result in elevated levels of cotinine and reduced levels of 3HC, due to the reduced capacity to convert cotinine to 3HC (28, 46, 64, 65). In zebrafish cyp2y3 mutants, we found that after a 6 h nicotine exposure, 3HC levels were reduced and cotinine levels were increased, consistent with it playing a role similar to CYP2A6. The levels of 3HC are still detectable in cyp2y3 mutants though, suggesting that in zebrafish there are additional enzymes capable of converting cotinine to 3HC. These findings demonstrate that Cyp2y3 serves a function similar to CYP2A6. Also, we observed elevated levels of COX and 4HPBA in cyp2y3 mutants exposed to nicotine. In humans, there are two primary enzymes that oxidize cotinine to COX, CYP2A6, and CYP2C19. Additionally, CYP2B6, CYP2E1, and CYP3A4 can also participate in COX formation, albeit with low affinities (66). Since cotinine levels were elevated in cyp2y3 mutants, we would expect to find elevated levels of COX if cotinine is similarly metabolized by multiple enzymes in zebrafish. However, higher levels of COX have not been reported in human populations deficient in CYP2A6 activity (65, 67). The effect of reduced CYP2A6 activity on 4HPBA levels is unclear. However, a recent study suggested that other enzymes, yet to be defined but not CYP2A6, participate in the production of 4HPBA (68). If this is the case, reduced CYP2A6 function would be expected to elevate 4HPBA levels, similar to what we see in our zebrafish cyp2y3 mutants. In contrast to the 6 h exposure, a long nicotine exposure resulted in the reduction of all nicotine metabolites in cyp2y3 mutants. We cannot currently explain this effect. It may be due to compensatory responses by the embryo or due to the chronic nature of the exposure saturating enzymes, according to properties of enzyme kinetics and the Michaelis–Menten model.

Variation in nicotine metabolism influences the outcomes of embryonic development

Genetic variation in CYP2A6 affects nicotine metabolism and nicotine metabolite levels (30, 69, 70). These variants are associated with interindividual differences in smoking behaviors, nicotine dependence, and tobacco-related diseases (71–74). Variation in nicotine metabolism could also have an impact on the developing embryo. Here, we found that conditions that decrease the levels of nicotine metabolites, cyp2y3 mutation or 1% ethanol exposure, partially protected embryos against the deleterious effects of nicotine. This ethanol coexposure significantly improves every craniofacial measure, relative to nicotine exposure alone, although rescue was never complete. In cyp2y3 mutants, only Meckel's cartilage and the ceratohyal were partially rescued. One possible reason for this difference could be how nicotine metabolism is altered in each of these treatments. We found that all metabolites were reduced in coexposed embryos treated with short and long exposures, with the exception of 3HC which was not altered after 6 h of exposure. In cyp2y3 mutants, the most teratogenic metabolites 4HBPA and NOX are only reduced in the longer exposure, with 4HBPA elevated after 6 h of exposure.

Our results could suggest that genetic variants that alter the enzymatic activity of CYP2A6, or other nicotine metabolizing enzymes, may modify the risk of developing embryos to craniofacial defects caused by nicotine. While few studies have been performed, Xie et al. (75), found that pregnant women carriers of the wild-type CYP2A6 and the deletion of GSTT1 are at risk of second hand smoke-related small gestational age, suggesting that GSTT1 protects against the metabolites produced in phase I of nicotine metabolism. Some reports have shown an increased risk of adverse birth outcomes (birth weight, length, and head circumference) in pregnant women who smoke and carry null alleles of phase 2 detoxifying enzymes, suggesting that the risk of fetal growth restriction is modulated by maternal metabolic gene polymorphisms (76–79). In addition, Stroud et al. found that a slower nicotine metabolite ratio was associated with decrements in birth weight but only among pregnant women who smoked a moderate number of cigarettes per day (<15). No significant associations were found between nicotine metabolite ratio and birth weight among women who smoked more (80). This study would seem to be in contrast to our findings, where a decreased ratio was protective. However, only 3HC and cotinine levels were quantified by Stroud and colleagues, so it is unclear if the concentrations of the metabolites that we have found to be teratogenic was altered. Additional research is needed to determine how the rate of nicotine metabolism is related to birth defects, both heavy and light.

Nicotine metabolites cause craniofacial defects

We found that direct exposure to the metabolites 4HPBA or NOX caused craniofacial defects. When we replaced the nicotine-containing media daily, maintaining the nicotine concentration while reducing the concentration of the metabolites, we found that the craniofacial skeleton was not affected at all. These results suggest that nicotine itself is not the primary cause of craniofacial defects in the exposed fish. However, the IOD was still significantly reduced in fish with daily medium changes. Thus, while nicotine appears to have limited impact on the skeleton, it is likely exerting effects on the brain. Additionally, none of the nicotine metabolites reduced the IOD, even when combined, suggesting a distinct mechanism of teratogenesis between nicotine and its metabolites.

How might these metabolites exert their effect? There is limited information about the toxicological effects of NOX. 4HPBA is an aromatic nitrosamine that results from the hydroxylation of nicotine. Nitrosamines are potent carcinogens that, when metabolically activated, bind to DNA and cause mutations (81, 82). Studies in animals have shown that embryonic exposure to nitrosamines increases the levels of 8-oxo-2-deoxyguanosine, a marker of DNA oxidative damage in maternal and fetal tissues (83). In zebrafish embryos, 4HPBA was detected as a metabolic product of NNK (a potent carcinogen from tobacco smoke and nicotine metabolism). Embryonic exposure to NNK activates the base excision repair pathway, which removes small DNA lesions (84). In humans, elevated nitrosamine levels in pregnant women are significantly associated with low birth weight, small gestational age, and an increased risk of preterm birth (85–87), similar outcomes to those seen with prenatal nicotine exposure. It is possible that 4HPBA and the subsequent oxidative stress and DNA damage may cause nicotine-induced craniofacial defects. Analysis of DNA damage and cellular repair mechanisms in nicotine-exposed fish could shed light on this possibility.

Coexposures of nicotine metabolites and alcohol enhance teratogenicity

The adverse effects of nicotine and alcohol use during pregnancy are well described; however, much less is known about how combined use effects the developing embryo. It has been found that preterm labor, low birth weight, and growth restriction occurred more frequently in women who drank and smoked during pregnancy with the odds ratio greater than the sum of the two substances, suggesting that nicotine and alcohol may have a synergistic effect (49). Also, the concomitant use of tobacco and alcohol increased the risk of delayed acquisition of fine motor skills than the use of either substance alone (88). However, in a study by Washington et al., a morphometric analysis was performed in newborns exposed to multiple substances. They observed that when alcohol and nicotine are consumed together, there is a decrease in phenotypic changes compared to alcohol alone, suggesting that nicotine could then be related to the inhibition of the teratogenic effects of alcohol (89). Thus, the results are mixed in human populations regarding the extent to which alcohol and nicotine coexposures are deleterious.

We found a complex interaction between alcohol and nicotine in the development of the zebrafish skull. Zebrafish coexposed to nicotine and alcohol did have reductions to the craniofacial skeleton, relative to the control or ethanol alone group. However, the defects in the coexposed group were less pronounced than in the group exposed to nicotine alone. In contrast, coexposure to nicotine metabolites and alcohol greatly worsened the craniofacial phenotypes relative to either single exposure, even nicotine metabolites that did not disrupt craniofacial structures did so in combination with ethanol. Thus, the conflicting findings in human populations could relate to total nicotine and alcohol consumption and activity of metabolizing enzymes.

It has been suggested that alcohol may alter nicotine metabolism (57, 58). We found that nearly all metabolites were reduced following a short or long exposure, with the exception of 3HC at the short exposure, there were different effects on the metabolite ratio. Following a short exposure the ratio was increased, denotating a more rapid enzymatic activity. The ratio was decreased in fish with a longer exposure, indicating slowed activity. Therefore, alcohol modifies nicotine metabolism but it is likely that the mechanism depends on the duration of alcohol exposure. It is possible that prolonged exposure to both substances leads to inhibition of nicotine metabolism and a consequent reduction in the levels of its metabolites while short exposure accelerates nicotine metabolism and the clearance of nicotine metabolites. There is evidence in humans that ethanol increases the nicotine metabolite ratio (57, 58). How the pattern of alcohol use may alter nicotine metabolism will be important to study to aid in understanding coexposure risks.

Together, our results indicate that nicotine metabolites play a role in nicotine teratogenesis and that there is a gene–environmental interaction between nicotine and metabolizing enzymes that modify developmental outcomes. We provide new insights into the pharmacological interaction between nicotine and alcohol; however, further studies are needed to characterize the teratogenic effects of this interaction. We demonstrate that zebrafish embryos are a useful model for the study of nicotine pharmacogenetics, enabling an understanding of biochemical and molecular processes that may have translational impact on human health and potentially supporting the development of new therapeutic strategies to mitigate the effects of prenatal nicotine exposure.

Materials and methods

Zebrafish maintenance and care

All animal experiments in this study were performed according to an animal use protocol approved by the University of Texas at Austin Institutional Animal Care and Use Committee. Zebrafish were raised and housed according to recommendations in The Zebrafish Book, 5th edition, and the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health.

Chemicals

Nicotine hydrogen tartrate salt (CAS: 65-31-6) and trans-3′-hydroxycotinine-D3 (3HC-D3) (CAS: 159956-78-2) were purchased from Sigma-Aldrich. Nicotine-D3 (CAS: 69980-24-1), cotinine (CAS: 486-56-6), cotinine-D3 (CAS: 110952-70-0), and trans-3′-hydroxycotinine (3HC) (CAS: 34834-67-8) were purchased from Cayman Chemical. Rac-nicotine-1′-oxide (NOX), rac-trans-nicotine-1′-oxide-D3 (NOX-D3) (CAS: (1R,2S)-isomer: 51020-67-8; (1S,2R)-isomer: 29419-55-4), (S)-cotinine N-oxide (COX) (CAS: 36508-80-2), (R,S) cotinine-D3-N-Oxide (COX-D3)(CAS: 1215561-37-7), rac 4-hydroxy-4-(3-pyridyl)butanoic acid (4HPBA) (CAS: 15569-97-8), rac 4-hydroxy-4-(3-pyridyl)butanoic-D3 acid (4HPBA-D3) (CAS: 1185129-36-5) were purchased from Toronto Research Chemicals, Inc. Dimethyl sulfoxide (DMSO) (CAS: 67-68-5), methanol LC/MS grade (CAS: 67-56-1), formic acid (CAS: 64-18-6), and acetonitrile (CAS: 75-05-8) were purchased from Fisher Scientific.

Generation of cyp2y3 mutant zebrafish line

The cyp2y3 mutant zebrafish line was created using CRISPR/Cas9. The target site for the guide RNA (gRNA) was designed using CHOPCHOP software (https://chopchop.cbu.uib.no/). The gRNA (5′-CCTCGGGCCTCAGAGAATGG-3′, crRNA IDT), targeted exon 2 of cyp2y3. One cell stage-embryos were injected with 2 µL of a mixture containing 500 ng/µL Cas9 protein (IDT) and 250 ng/μL of gRNA in water with 0.2% phenol red. Mutants were identified in the F1 offspring using Sanger sequencing. An individual with a 68 base pair deletion (cyp2y3^au137) was selected to backcrossed to AB to produce a stable line. The expression of cyp2y3 was analyzed in mutants using PCR and fluorescent in situ hybridization following the manufacturer's protocol.

RT-PCR expression analyses of cyp2y3

Total RNA was extracted from ten zebrafish embryos at 48 hpf using TRIzol reagent (Ambion, Austin, TX, USA) following the manufacturer's protocol. SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA, USA) was used for cDNA synthesis. Two sets of primers were designed to analyze cyp2y3 expression in mutant and wild-type embryos. For the wild-type allele, the forward primer was located at 173 pb upstream of the deletion and the reverse primer was located in the deletion (Forward 5′—TGCTGGTTTTGGTCTTGACTGTG—3′; Reverse 5′—GGCCCGAGGTGCACCGTCAT—3′) and amplifies a 193 bp product. A second pair of primers flanking the deletion was designed (Forward 5′- CACTGGACAAAAGTGCCCCA—3′; Reverse 5′—AGCAGATGACGTTGGACACA—3′) to amplify 408 bp in wild-type fish and 340 bp in mutants.

Chemical exposures

Nicotine and nicotine metabolite stock solutions were prepared in DMSO at 100 mM concentration and diluted in embryo media (EM) to the final concentrations. Unless otherwise stated, all exposures were from 6 hpf to 4 dpf.

Dual bone and cartilage staining

Cartilages and bones were stained using a previously described Alizarin Red and Alcian Blue staining protocol (90). Briefly, embryos were fixed for 1 h in 2% paraformaldehyde in phosphate-buffered saline, followed by dehydration into 100% ethanol and overnight staining with Alcian Blue. Embryos were rehydrated, bleached with 3% H₂O₂/0.5% KOH, and stained with Alizarin Red. Embryos were then washed and stored in 50% glycerol and 0.2% KOH. Imaging was performed on a Zeiss Axio Imager. All images were captured at 10× magnification. Neurocranial and viscerocranial cartilages were measured using ImageJ software for Windows version 1.8.0.

Sample preparation and extraction for LC/MS nicotine metabolites quantification

To measure nicotine metabolites in fish, 10 nicotine-exposed fish were homogenized per replicate in 50 µL of lysis buffer (1 M Tris pH = 8.3, 1 M KCl, 1 M MgCl2, 10% NP40). The extraction method was adapted from a published protocol (70). Each 50 µL zebrafish homogenate was fortified with 8 µL of 125 ng/mL deuterated internal standard solution (final concentration of 20 ng/mL), which contained nicotine-D3, trans-3′-hydroxycotinine-D3, cotinine-D3, cotinine-oxide-D3, 4HPBA-D3, and nicotine-1′-oxide-D3. The homogenates were mixed for 30 s. To precipitate proteins, 125 µL of 40 mmol/L NH₄AC (pH 6.7) and 400 µL of acetonitrile:methanol (75%:25%, v/v) were added. After vortexing and centrifugation at 12,000×g for 10 min at 4 °C, the supernatant was transferred to a clean 2 mL Eppendorf tube. Protein precipitation was repeated three times, and the supernatants were collected in the same tube. Supernatants were evaporated to dryness at 35 °C under a gentle stream of nitrogen and reconstituted with 50 µL of mobile phase (1% formic acid). Samples were resuspended by vortex and centrifuged at 16,000×g for 10 min. Forty-five microliters were transferred into injection microvials for analysis by liquid chromatography (LC) mass spectrometry (MS). To measure nicotine metabolites in EM, 500 µL of EM were centrifuged at 12,000×g for 10 minutes at 4 °C. Forty-five microliters were transferred into injection microvials for analysis. Samples were analyzed in triplicate or quadruplicate.

LC-MS conditions

Zebrafish samples were analyzed using a Shimadzu LC-MS/MS system consisting of a dual pump (LC40D XR: pumps A & B) coupled to LCMS-8060 triple quadrupole mass spectrometer with an electrospray ionization interface (ESI). Chromatographic separation was accomplished on a Raptor Biphenyl Column, Restek part # 9309A12 2.7u × 100 mm × 2.1 mm. LC was performed at a flow rate of 0.5 mL/min. The mobile phase consisted of a gradient elution of 0.1% formic acid with 5 mM ammonium formate (solvent A) and MeOH with 0.1% Formic acid (solvent B). Initial conditions were 90% of solvent A and 10% of solvent B. Concentrations of solutions A and B changed over 5.1 min as shown in Table S5. The injection volume of each prepared sample was 2 μL. The mass spectrometer was operated in positive electrospray mode using multiple reaction monitoring mode (MRM) data acquisition. Two MRM transitions were monitored for each analyte. Fragment number 1 was used for quantification, and fragment number 2 was used as a reference ion. Nitrogen was used as the nebulizer and collision gas. The following ESI conditions were applied: nebulizing gas flow 3 L/min, heating gas flow 10 L/min, interface temperature 300 °C, desolvation temperature 526 °C, DL temperature 250 °C, heat block temperature 400 °C, and drying gas flow 10 L/min. Analyte-specific collision energies and MRM transitions are provided in Table S6.

Statistical analysis

One-way ANOVA with Tukey's and Dunnett's multiple comparisons correction was performed for the analysis of craniofacial defects in control and nicotine-exposed zebrafish embryos. Two-way ANOVA with Tukey's multiple comparisons correction was used for comparison of craniofacial defects in the nicotine and nicotine and ethanol co-exposed embryos and between cyp2y3 mutants and wild-types. An unpaired, two-tailed t-test was used to calculate significance of nicotine metabolite levels between cyp2y3 mutants and wild-types and between nicotine and nicotine and ethanol coexposed groups. Data are presented as mean ± SEM; P < 0.05 was considered statistically significant and significant meaningful comparisons are shown in all graphs whereas non-significant comparisons are not shown. GraphPad Prism 9.4.0 (San Diego, CA, USA) was used for statistical analysis and graphing.

Supplementary Material

Supplementary material is available at PNAS Nexus online.

Competing Interest

The authors declare no competing interests.

Funding

This work was supported by the National Institutes of Health/National Institute of Dental and Craniofacial Research (R35DE029086) and the National Institutes of Health/National Institute on Alcohol Abuse and Alcoholism (R01AA023426). Both grants are to Johann K. Eberhart.

Author Contributions

Gissela Borrego-Soto (Conceptualization, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing—original draft, Writing—review & editing), Ian M. Riddington (Formal analysis, Investigation, Methodology, Supervision), Esther Y Maier (Investigation, Methodology, Supervision, Validation), Aneesha Baral (Investigation), Luis Compean (Investigation), Radhika Amin (Investigation), and Johann K. Eberhart (Conceptualization, Funding acquisition, Project administration, Supervision, Writing—review & editing)

Data Availability

All data and results of statistical analyses are reported in this article and included in Supplementary material. We welcome any specific inquiries.