Dickinson AJG et al. (2022), E-liquids and vanillin flavoring disrupts retin...

XB-ART-58466

Dev Biol 2022 Jan 01;481:14-29. doi: 10.1016/j.ydbio.2021.09.004.

Show Gene links Show Anatomy links

E-liquids and vanillin flavoring disrupts retinoic acid signaling and causes craniofacial defects in Xenopus embryos.

Dickinson AJG , Turner SD , Wahl S , Kennedy AE , Wyatt BH , Howton DA .

???displayArticle.abstract???
Environmental teratogens such as smoking are known risk factors for developmental disorders such as cleft palate. While smoking rates have declined, a new type of smoking, called vaping is on the rise. Vaping is the use of e-cigarettes to vaporize and inhale an e-liquid containing nicotine and food-like flavors. There is the potential that, like smoking, vaping could also pose a danger to the developing human. Rather than waiting for epidemiological and mammalian studies, we have turned to an aquatic developmental model, Xenopus laevis, to more quickly assess whether e-liquids contain teratogens that could lead to craniofacial malformations. Xenopus, like zebrafish, has the benefit of being a well-established developmental model and has also been effective in predicting whether a chemical could be a teratogen. We have determined that embryonic exposure to dessert flavored e-liquids can cause craniofacial abnormalities, including an orofacial cleft in Xenopus. To better understand the underlying mechanisms contributing to these defects, transcriptomic analysis of the facial tissues of embryos exposed to a representative dessert flavored e-liquid vapor extract was performed. Analysis of differentially expressed genes in these embryos revealed several genes associated with retinoic acid metabolism or the signaling pathway. Consistently, retinoic acid receptor inhibition phenocopied the craniofacial defects as those embryos exposed to the vapor extract of the e-liquid. Such malformations also correlated with a group of common differentially expressed genes, two of which are associated with midface birth defects in humans. Further, e-liquid exposure sensitized embryos to forming craniofacial malformations when they already had depressed retinoic acid signaling. Moreover, 13-cis-retinoic acid treatment could significantly reduce the e-liquid induced malformation in the midface. Such results suggest the possibility of an interaction between retinoic acid signaling and e-liquid exposure. One of the most popular and concentrated flavoring chemicals in dessert flavored e-liquids is vanillin. Xenopus embryos exposed to this chemical closely resembled embryos exposed to dessert-like e-liquids and a retinoic acid receptor antagonist. In summary, we determined that e-liquid chemicals, in particular vanillin, can cause craniofacial defects potentially by dysregulating retinoic acid signaling. This work warrants the evaluation of vanillin and other such flavoring additives in e-liquids on mammalian development.

???displayArticle.pubmedLink??? 34543654
???displayArticle.pmcLink??? PMC8665092
???displayArticle.link??? Dev Biol
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: aldh1a2 crabp1 crabp2 cyp1a1 cyp26c1 rarg
GO keywords: response to toxic substance [+]

???displayArticle.disOnts??? orofacial cleft [+]

???attribute.lit??? ???displayArticles.show???

	Graphical Abstract.
	Fig. 1. Comparison of the faces of embryos exposed to e-liquid F and RAR antagonist. A) Schematic of the experimental design. B-J) Lateral and frontal views of embryos from one representative clutch that were imaged using a stereoscope. Pink dots outline the embryonic mouth. K-M) Quantification of the intercanthal distance and dorsal mouth angle (K) of embryos from B-J. Box and whiskers plots showing the median and the 5th and 95th percentiles (dots) of intercanthal distance (L) and dorsal mouth angle (M). Statistical analysis using ANOVA followed by Tukey pairwise comparisons. P-values are ∗ = p < 0.001, + = p = 0.016. All images and quantification are based on 2 biological replicates (n = 48).
	Fig. 2. A-C) Comparison of craniofacial and neural structures in embryos exposed to e-liquid F and RAR antagonist. Transverse section through the oral cavity with E-cadherin (green) labeling of the epidermis and counterstained with Dapi (blue), stage 43. Pal = palate region, nos = nostril, br = brain. D-F) Lateral view of embryos stained with Alcian blue to show cartilage at stage 45–46. Abbreviations: b = brachial, c = ceratohyal, eth = ethmoid, M = Meckel's. Arrow indicates the missing ethmoid cartilage. G-I) Lateral view of Phalloidin labeling (green) to show the muscles (counterstain Dapi) at stage 45–46. Abbreviations; lm = levator mandibular longus, a = angularis muscles, oh = orbitohyoideus. J-L) Dorsal view of embryos labeled with alpha-tubulin (red), Phalloidin (green), and counterstained with Dapi at stage 45–46. Abbreviations; olf + cer = olfactory nerve and cerebrum, nos = nostril. The percentage of embryos with the same phenotype as shown in the image is indicated. Each assay is based on 2 biological replicates (n = 20).
	Fig. 3. E-liquid F exposure alters the expression of RA metabolism and signaling pathway genes. A) Schematic of the experimental design. B) Expression of RA associated genes. Quantitative RT-PCR based on 3 biological replicates (n = 60 face dissections) compared to the RNA-seq data based on 2 biological replicates (n = 200 face dissections). Error bars are derived from standard error and all data was determined to be statistically significant. Student t-tests were performed on qRT-PCR data using delta-delta CT values and p values are: rarg.L = 0.009, aldh1a2.L = 0.016, crabp1.S = 0.02, crabp2.S = 0.026, cyp26c1.S = 0.015. Pvalues for the RNA-seq data are rarg.L = 0.0248, aldh1a2.L = 0.00252, crabp1.S = 0.0284, crabp2.S = 2.01E-09, cyp26c1.S = 5.72E-08 (also reported in Table 1) C) Overlap in genes decreased by e-liquid F (VE) exposure and RAR antagonist treatment.
	Fig. 4. Sensitized assays with e-liquid F with RAR antagonist or 13-cis RA. A-G) Frontal views of representative embryos at stage 44 after a e-liquid and RAR antagonist assay based on 4 biological replicates (n = 98). The shape of the mouth is outlined in pink dots. E) Box and whiskers plots showing the median and the 5th and 95th percentiles (dots) of the dorsal mouth angle. The lightest bars represent measurements from embryos treated with both e-liquid and RAR antagonist. F) Stacked bar graph of the same data as in E showing the distribution of dorsal mouth angles. G) Box and whiskers plots showing the median and the 5th and 95th percentiles (dots) of intercanthal distance. The lighter bar represents measurements from embryos treated with both e-liquid and RAR antagonist. H) Stacked bar graph of the same data as in G showing the distribution of dorsal mouth angles. E-H is based on 2 biological replicates (n = 48). Statistical analysis using ANOVA followed by pairwise Tukey tests. P-values below 0.001 are denoted with ∗. P-values deemed non-significant are denoted with and ns and are ns1 = 0.552, ns2 = 0.810, ns3 = 0.301 and ns4 = 0.282. I-L) Frontal views of representative embryos at stage 43 after e-liquid and 13-cis RA (13cRA) assay. The shape of the mouth is outlined in pink dots. M) Box and whiskers plots showing the median and the 5th and 95th percentiles (dots) of the dorsal mouth angle. The light bar represents measurements from embryos treated with e-liquid alone. N) Stacked bar graph of the same data as in M showing the distribution of dorsal mouth angles. O) Box and whiskers plots showing the median and the 5th and 95th percentiles (dots) of intercanthal distance. The lighter bar represents measurements from embryos treated with both e-liquid and RAR antagonist. P) Stacked bar graph of the same data as in O showing the distribution of dorsal mouth angles. M-P is based on 3 biological replicates (n = 72). Statistical analysis using ANOVA followed by pairwise Tukey tests. P-values below 0.001 are denoted with ∗. P-values deemed non-significant are ns5 = 0.739, ns6 = 0.285.
	Fig. 5. Neat vs vaporized extract, dessert-like flavors, and vanillin. A-C) The faces of representative embryos (stage 44) that were exposed to either the vaporized extract or neat e-liquid F from one representative clutch (derived from the same parents). Based on 4 biological replicates (n = 96). Pink dots outline the embryonic mouth. D-H) Representative embryos (stage 43) exposed to neat e-liquids, (1:1000, Birthday Cannoli and 1:1000 Vanilla Ice Cream) and 30 μg/ml vanillin from one representative clutch (derived from the same parents –but different than A-C). Based on 2 biological replicates (n = 48). Pink dots outline the embryonic mouth. I) Quantitative RT-PCR of genes associated with RA metabolism or signaling in embryos exposed to vanillin. Error bars are derived from standard error and all data was determined to be statistically significant. Student t-tests were performed on qRT-PCR data using delta-delta CT values and p values are: rarg.L = 0.009, aldh1a2.L = 0.016, crabp1.S = 0.02, cyp26c1.S = 0.015. Based on 3 biological replicates.
	Fig. 6. Models. A) Schematics of a simple hypothetical RA signaling pathway that could be present in cells of the normal developing face and cells exposed to e-liquid F. Genes that are reduced in facial tissues in response to e-liquid F exposure are indicated by blue and those that were increased are indicated in pink. The effects on craniofacial development are shown in the accompanying face cartoons. B) Hypotheses of rescue and synergy assays. On both of the graphs the dark grey area indicates a zone of RA signaling which results in a apparently normal craniofacial region. Outside of this zone (light grey) the embryos have craniofacial defects due to changes in RA signaling. In the first graph, embryos exposed to low 13-cis-RA (blue dot) fall into the normal zone while those exposed to high e-liquid F (pink dot) concentrations would be in an abnormal zone. When high e-liquid F and low 13-cis-RA are combined the RA signaling is increased and the craniofacial development is close to normal (purple dot). In the second graph, embryos exposed to low RAR antagonist (blue dot) or low e-liquid (pink dot) each have minorly reduced RA signaling but still within the zone resulting in normal craniofacial development. However, when these are combined the reductions in RA signaling are compounded and craniofacial defects are observed (purple dot).
	figs1.
	figs2.

References [+] :

Abraham, Vanillin, a potential agent for the treatment of sickle cell anemia. 1991, Pubmed