Introduction
Environmental exposures to persistent organic pollutants (POPs), including the toxicant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), pose a considerable hazard to public health. Reference Kirkok, Kibet, Kinyanjui and Okanga1 Released as an unintentional byproduct of fossil fuel combustion and waste incineration, TCDD is a polycyclic aromatic hydrocarbon and potent ligand of the aryl hydrocarbon receptor (AhR). Reference Van den Berg, Birnbaum and Denison2 Although generally detected below parts per trillion (ppt) levels in environmental compartments, the persistent, lipophilic properties of this compound result in bioaccumulation through the food chain; as such, human exposure to TCDD predominantly occurs via dairy and meat consumption. Reference Kirkok, Kibet, Kinyanjui and Okanga1,Reference Van den Berg, Birnbaum and Denison2 As TCDD is classified as an endocrine-disrupting compound, even low-level exposure can impact the large array of physiological systems cued by endocrine signaling. Reference Nicolopoulou-Stamati and Pitsos3 Per the developmental origins of health and disease (DOHaD) approach, such exposure during sensitive periods of development exacerbates these adverse effects, predisposing developing organisms towards lifelong and adult-onset disease outcomes. Reference Gluckman, Hanson and Beedle4
Developmental TCDD exposure has thus been associated with numerous well-characterized and persistent toxic effects across humans and animal models, including systemic disruption of skeletal development, metabolism, and neurobehavior. TCDD impairs osteogenesis in human cell lines, Reference Guo, Zhao, Zhao, Sun, Liu and Zhang5 and developmental exposure can induce congenital defects in skeletal and cartilaginous tissues, including spina bifida and cleft palate in humans and rodents. Reference Ngo, Taylor and Roberts6,Reference Yamada, Hirata and Sasabe7 The sensitivity of teleost fish to TCDD-induced skeletal dysmorphogenesis has also been well-documented, including alterations to craniofacial and axial skeletal structures; Reference Baker, Peterson and Heideman8,Reference Dong, Hinton and Kullman9 however, less is known about the developmental progression of such effects. Another outcome of acute TCDD exposure is pronounced metabolic dysregulation, including stunted growth, decreased body weight, and wasting syndrome as hallmark indicators of toxicity in rodent and fish models. Reference Pohjanvirta and Tuomisto10–Reference Kelling, Christian, lnhorn and Peterson12 However, accumulating epidemiological evidence suggests a more complex relationship with energy homeostasis and growth rate that may depend on the dose and latency to effect, Reference Lindén, Lensu, Tuomisto and Pohjanvirta13–Reference Iszatt, Stigum and Govarts19 variables that can be assessed using low-dose, longitudinal studies. Finally, TCDD exposure also targets nervous system development and function, impairing neuronal differentiation and neural network formation as well as altering brain size in rodents, fish, and humans. Reference Hill, Howard, Strahle and Cossins20–Reference Clements, Lawrence and Blank27 Both rodent and epidemiological studies have linked peri- and post-natal TCDD exposure to early-life deficits in neurocognition and behavior regulation, including impaired cognitive and language development, socioemotional functioning, and increased risk of neurodevelopmental disorders. Reference Nishijo, Kuriwaki, Hori, Tawara, Nakagawa and Nishijo22,Reference Latchney, Hein, O’Banion, DiCicco-Bloom and Opanashuk28–Reference Crépeaux, Bouillaud-Kremarik, Sikhayeva, Rychen, Soulimani and Schroeder33 Still understudied, however, is the persistence and heritability of such neurobehavioral outcomes. To model these multisystemic outcomes of early-life TCDD exposure, we have used zebrafish (Danio rerio), a freshwater teleost fish species, as an NIH-validated tractable vertebrate model. Due to factors including cost-effective housing, external fertilization, large clutch size, rapid development of major organ systems, and quick generation time, zebrafish are a prominent model in developmental and generational toxicology and an excellent sentinel species for the effects of waterborne endocrine disruption on aquatic populations. Reference Roper and Tanguay34 Previous work in our lab determined that sublethal juvenile exposure to 50 ppt (parts per trillion; pg/mL) TCDD at 3- and 7- weeks post fertilization (wpf) resulted in heritable adult-onset skeletal and reproductive health outcomes. Reference Baker, Peterson and Heideman8 Therefore, we aim to use this same paradigm to characterize the longitudinal progression of craniofacial and trunk morphometrics and growth rate in developmentally exposed zebrafish, and to explore the extent and heritability of neurobehavioral outcomes.
Methods
Fish husbandry
Wild-type AB zebrafish were housed at Wayne State University (Detroit, MI, USA) and maintained according to approved protocols (see “Ethical Standards”). Adult stock fish were raised on recirculating water systems (Aquaneering Inc.; San Diego, CA, USA) with reverse osmosis (RO) water supplemented with Instant Ocean Spectrum Brands salts (60 mg/L; Blacksburg, VA, USA) and buffered with sodium bicarbonate. Fish were maintained at 27–30°C on a 14h/10h light/dark cycle and were fed a dry food mixture 1–2x daily, supplemented with brine shrimp through 91 days post fertilization (dpf). Further details regarding fish dietary regimen are included in Supplemental File S1.
TCDD exposure
Chemicals
1 mg of unlabeled 98% 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) was purchased from Cambridge Isotope Laboratories (Tewksbury, MA, USA) and dissolved in dimethyl sulfoxide (DMSO).
Exposures
Similarly to our previous study, Reference Baker, Peterson and Heideman35 juvenile fish were exposed to 50 ppt (pg/mL) TCDD for 1 hr at both 3 and 7 wpf (Figure 1), a dose and exposure regimen we had previously identified as sublethal after dose-response experimentation (25 -400 ppt TCDD). These two developmental timepoints were chosen because they span the period of gonadal differentiation and maturation in zebrafish. Fish were exposed in 4 separate cohorts, each consisting of 72 fish per condition. All means and n-values for each endpoint, timepoint, and condition are reported in Supplemental File S2.
Figure 1. Experimental paradigm. Four cohorts of AB wild-type fish were exposed to DMSO control or 50 parts per trillion (ppt; pg/mL) 2,3,7,8-tetrachlorodibenzo-p-dioxin at 3 and 7 weeks post fertilization (wpf). *To determine the effect of exposure timing, one cohort of fish was divided into two groups: one exposed at both 3 and 7 wpf, and the other exposed at 3 wpf alone. At 5, 7, 9, and 12 wpf, juvenile fish are imaged for developmental morphometrics (AB). At 4, 7, and 12 months post fertilization (mpf), subsets of fish euthanized for a separate study were measured for length (mm) and wet weight (g). From 7 to 12 months post fertilization, fish were spawned in alternating weeks as either within-group incrosses or as outcrosses between treated and control fish. Offspring from these crosses were kept through 5 days post fertilization (dpf) to analyze larval abnormalities. These F1 offspring from incrosses only were used to analyze light/dark neuromotor behavior. After completion of spawning (∼12 mpf), F0 AB fish underwent novel tank behavior testing.
3 wpf exposure. Juveniles were added to 20 mL glass vials of fish water (300 mg/L Instant Ocean salts and 50 mg/L sodium bicarbonate dissolved in RO water). Four fish per vial were added at a density of 1 fish/mL fish water. Fish were dosed with either TCDD dissolved in DMSO (50 pg/mL end concentration) or equivalent volume of DMSO (0.0875% v/v) for 1 hour, then rinsed and transferred to fresh fish water.
7 wpf exposure. Juveniles were added to 50 mL beakers of fish water. Depending on the number of surviving fish, between 3 and 7 fish were added per beaker at a density of 1 fish/5 mL fish water. Fish were dosed and rinsed as previously described, then transferred to fresh fish water. A subset (24 fish/treatment group) of the fourth cohort did not undergo the 7 wpf exposure protocol in order to examine the effects of exposure window.
Mortality was recorded daily in the timespan between the initial exposure (21 dpf) and adulthood (91 dpf). See Supplemental File S1 for further details concerning dosing protocol and experimental fish husbandry.
F0 juveniles – morphometrics
At 5, 7, 9, and 12 wpf (Figure 1), fish were anesthetized with 0.15 g/L buffered tricaine methanesulfonate (MS-222). Fish were then imaged using a Nikon SMZ18 stereomicroscope with a Nikon DS-Qi2 camera. For every fish, ImageJ (http://imagej.nih.gov; v.1.53, 24 August 2022) was used to determine length (in µm), area (in µm2), and angle measurements of twelve morphometric endpoints associated with fish growth and development (Figures 2d, 3e, Supplemental Fig. S2): total body length measured snout to tail, snout – UGP (urogenital pore) and UGP – tail lengths, vertical UGP (distance between the urogenital pore and dorsal fin), horizontal and vertical eye widths, eye-snout lengths measured from upper and lower jaw, head-trunk angle, and areas of head, operculum, and swim bladder. The endpoints above are represented as mean ± SEM (standard error of the mean). AB developmental morphometrics were analyzed for each timepoint using Student’s t-test in Microsoft Excel. Significance level (α) was set at 0.05 for all hypothesis testing throughout this study. For further details on imaging protocol, see Supplemental File S1.
Figure 2. Developmental craniofacial morphometrics in F0 DMSO and TCDD-exposed fish from 5 -12 wpf. a . Eye width (horizontal, µm). b . Eye-snout length, upper jaw (µm). c . Eye-snout length, lower jaw (µm). d . Representative image of a 7 wpf fish with craniofacial measurements indicated. # indicates p-value < 0.1; * indicates significant p-value < 0.05. DMSO, dimethyl sulfoxide; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; wpf, weeks post fertilization.
Figure 3. Developmental trunk morphometrics in F0 DMSO and TCDD-exposed fish from 5 to 12 wpf. a . Vertical UGP (distance from urogenital pore to dorsal fin; µm). b . Head-trunk angle (snout – highest point of spine – tail). c . Area of operculum (µm2). d . Area of swim bladder (µm2). e . Representative image of a 7 wpf fish with trunk measurements indicated. # indicates p-value < 0.1; * indicates significant p-value < 0.05. DMSO, dimethyl sulfoxide; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; wpf, weeks post fertilization.
F0 adults – morphometrics
At 4, 7, and 12 months post fertilization (mpf; Figure 1), a subset of fish was euthanized with 0.4 g/L buffered MS-222 for tissue collection. After euthanization, fish were briefly blotted to remove excess solution, after which wet weight (g) and length (mm) were recorded. BMI (g/cm2) was further calculated as an index of body condition. The measures above are represented as mean ± SEM. Student’s t-test in Microsoft Excel was used to analyze adult morphometric data for each timepoint.
F0 adults – novel tank test
At 12 ± 1 mpf (Figure 1), a subset of treated (TCDD) and control (DMSO) fish were selected for the novel tank test (NTT). The setup for the NTT consisted of a 6L tank in the center of a darkened room, with two lamps positioned at upper left and lower right diagonals from the tank. Two cameras were set up to provide side (Arena 1) and top (Arena 2) views of the experimental tank, and Noldus Ethovision XT software (Noldus Information Technology, Wageningen, The Netherlands) was used to record videos tracking zebrafish movement within the tank. One week after the final spawning session, fish were individually netted into the tank, and the 5-minute recording period was initiated. Distance traveled was integrated every 1/6th second. Mean distance traveled over the test period was calculated separately by sex and analyzed between treated and control fish using one-way analysis of variance.
To measure exploratory parameters, Arena 1 (side view) was divided into a top and bottom zone, and Arena 2 (top view) was divided into an inner and outer zone. We manually recorded the timing and number of transitions between zones for each arena per video. The endpoints analyzed included the proportion of fish that transitioned between zones at any point during the test period, total count of transitions, total amount of time spent in the preferred zones (bottom/outer zones), latency to first transition, and total count of “rapid” transitions (<1 s spent in a zone between transitions). Fish were separated by sex for analysis, and Chi-square analysis was conducted to compare the proportion of fish that transitioned at any point between treated and control groups. All other variables were analyzed with a Student’s t-test in Microsoft Excel. For further details on experimental setup and zone analysis, see Supplemental File S1.
F1 larvae – light/dark transition test
Offspring of exposed fish were analyzed for neuromotor behavior in a light: dark transition test. Between 10 and 12 mpf (Figure 1), treated and control fish were spawned to produce the indirectly exposed F1 generation. Only offspring from incrosses were analyzed for F1 larval behavior. Procedures for spawn setup, egg collection, and raising of larval fish were carried out as described in Supplemental File S1, but eggs were collected and raised in an egg water solution which did not contain 0.05% methylene blue to prevent interference with neuromotor outcomes. Reference Vaccaro, Patten and Ciura36 At 120 hpf (hours post fertilization), fish with inflated swim bladders and without any visible abnormalities were selected for the light-dark transition test. Fish were individually added to wells of a 24-well plate containing 2 mL egg water/well and allowed to acclimate at 28°C for at least 1 hr. Testing took place starting at 14:00. Plated fish were placed into a 28.5 ± 0.5°C Noldus DanioVision Observation chamber and allowed to acclimate for 12 min prior to a 24-min series of alternating 3-min light-dark cycles. Larval movements were tracked throughout this series using an infrared camera, and distance traveled was integrated every 30 s. For quality control purposes, outliers were removed prior to analysis (Supplemental File S1). Mean distance traveled per 30 s period ± SEM was calculated for light and dark cycles separately and compared between treated and control fish using one-way analysis of variance.
F1 larvae – developmental abnormalities
Between 7 and 12 mpf (Figure 1), treated (TCDD) and control (DMSO) fish were spawned as both incrosses and outcrosses to produce the indirectly exposed F1 generation. Procedures for spawn setup and egg collection were carried out as described in Supplemental File S1. After collection, nonviable eggs were removed, and viable eggs were plated in egg water with methylene blue at a density of 75 eggs/100 mm petri dish and incubated at 28°C. At 24 hpf, dead eggs were removed and surviving eggs were divided into 6-well plates. Three replicates of 30 eggs/well were plated for each successful spawn. From 24 to120 hpf, debris and dead eggs were removed, and 50% water changes were conducted on each well. At 120 hpf, fish were screened for hatch rate from chorion, skeletal deformities, yolk sac edema, and heart edema. All measures are reported as proportions with 95% confidence intervals. Chi-square analysis was conducted to compare the proportions of fish with various abnormalities between control and treated F1 fish for groups exposed at 3 and 7 wpf, while Fisher’s exact test was used for groups exposed at 3 wpf only due to low sample size.
Results
The experimental schema is depicted in Figure 1. We tracked mortality daily from the initial exposure at 3 wpf through adulthood at approximately 13 wpf. While exposed fish trended towards decreased time to mortality, the overall difference in mortality from 3 to 13 wpf was not statistically significant (p > 0.05; Supplemental Fig. S1).
F0 juveniles – morphometrics
We analyzed craniofacial and trunk/body morphometrics of control and TCDD-exposed fish at four timepoints across juvenilehood: 5, 7, 9, and 12 wpf. At 5 wpf, two weeks after the initial exposure, TCDD exposure significantly decreased eye width by 6.8% and lower eye-snout length by 7.9% (p = 0.041; p = 0.003; Figure 2a,c). The upper eye-snout length of exposed fish similarly trended towards a 4.5% decrease (p = 0.088; Figure 2b). No body/trunk morphometrics were affected at 5 wpf (Figure 3). At 7 wpf, there were no significant changes in craniofacial parameters in exposed fish – only a trend towards a 3.6% decrease in eye width (p = 0.055; Figure 2a). However, TCDD exposure significantly decreased head-trunk angle by 0.8% (p = 0.002; Figure 3b).
At 9 wpf, two weeks after the second exposure, upper eye-snout length was the only craniofacial measure significantly affected, at a 10.6% decrease (p = 0.002; Figure 2b). Trunk/body morphometrics were most altered by TCDD exposure at 9 wpf: head-trunk angle and operculum area were significantly decreased by 0.7% and 13.0% respectively (p = 0.011; p = 0.014; Figure 3b,c). Similarly, both vertical UGP and swim bladder area trended towards respective decreases of 8.6% and 26.1% (p = 0.098; p = 0.056; Figure 3a,d). At 12 wpf, horizontal eye width trended towards a 2.48% decrease in exposed fish (p = 0.078; Figure 2a), while head-trunk angle significantly decreased by 1.15% (p = 0.004; Figure 3b). Swim bladder area was not assessed at 12 wpf, as increased pigmentation and body thickness prevented external visualization. Remaining morphometrics that were not altered by TCDD exposure at any timepoint are reported in Supplemental Fig. S2.
F0 adults – morphometrics
We analyzed fish length (mm), weight (g), and BMI (g/cm2) at three timepoints across adulthood: 4, 7, and 12 mpf. At 4 mpf, only male fish were assessed. Male fish exposed to TCDD at both 3 and 7 wpf weighed significantly more (+10.1% p = 0.029; Figure 4b) and had a significantly higher BMI (+7.0%, p = 0.037) than controls (Figure 4c). At 7 mpf, both males and females were assessed; females were not affected (Supplemental Fig. S3A–C). Again, exposed males weighed significantly more than controls (+11.1%, p = 0.022; Figure 4b) but were also significantly longer (+3.1%, p = 0.021; Figure 4a); thus, BMI was no longer affected (Figure 4c). Male fish that had only been exposed to TCDD at 3 wpf were also evaluated at 7 mpf. Similar to their counterparts exposed at both timepoints, singly exposed fish weighed significantly more than controls (+12.6%, p = 0.042; Supplemental Fig. S3E); however, they were not significantly longer (Supplemental Fig. S3D). BMI trended towards a 9.0% increase (p = 0.0501; Supplemental Fig. S3F). At 12 mpf, both males and females were assessed. No significant differences in weight, length, or BMI were found between exposed and control fish for either sex, suggesting that, for males, control growth had caught up during the 7–12 mpf interval (Figure 4; Supplemental Fig. S3a–c).
Figure 4. Adult morphometrics of TCDD-exposed and control (DMSO) fish at 4, 7, and 12 mpf. a . Mean length (mm) of male fish exposed at both 3 and 7 wpf. b . Mean weight (g) of male fish exposed at both 3 and 7 wpf. C. Mean BMI (g/cm2) of male fish exposed at both 3 and 7 wpf. * indicates significance at p-value < 0.05. wpf, weeks post fertilization; mpf, months post fertilization; DMSO, dimethyl sulfoxide; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin.
F0 adults – novel tank test
At 12 mpf, we used a novel tank test to investigate whether juvenile exposure altered fish behavioral tendencies in adulthood. Overall distance traveled (cm) was not influenced by exposure in either sex, as measured from the side or top view of the tank (Supplemental Fig. S4A). However, exposed fish of both sexes demonstrated overall trends of increased exploratory/anxiolytic behavior over controls. For exploratory parameters, we evaluated transition timing and latency between a preferred “bottom zone” and less preferred “top zone”, manually tracked from the side profile of the tank. Exposed females had a significantly shorter latency to enter the top zone (−42.3%, p = 0.016), while latency was not affected in males (Figure 5a). No other exploratory parameters were significantly affected by exposure as measured from the side profile (Supplemental Fig. S4B; S4C), but exposed males trended towards an increased number of transitions between zones (+129.5%, p = 0.095; Figure 5b) and a decreased amount of time spent in the preferred bottom zone (−8.1%, p = 0.065; Figure 5c). From the top profile of the tank, we tracked the same endpoints with reference to a preferred “outer zone” and a less preferred “inner zone”. Exposed males spent significantly less time in the preferred outer zone (−7.6%, p = 0.009), and exposed females trended similarly (−4.9%, p = 0.099; Figure 5c). While no other parameters were significantly affected (Figure 5b ; Supplemental Fig. S4B; S4C), exposed males also trended towards decreased latency to enter the inner zone (−77.9%, p = 0.080), while latency was not affected in females (Figure 5a).
Figure 5. Adult novel tank test behavioral measures in TCDD-exposed and DMSO control fish of each sex at 12 mpf, as tracked manually from side (left column) and top (right column) profiles. a . Latency (s) to enter the non-preferred zones (top/inner zones) of the tank. b . Count of total number of zone transitions. c . Total time (s) spent in preferred zones (bottom/outer) of the tank. # indicates p-value < 0.1; * indicates significant p-value < 0.05. DMSO, dimethyl sulfoxide; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; mpf, months post fertilization.
F1 larvae – light/dark transition test
To investigate whether the offspring of exposed fish also demonstrated behavioral dysregulation, we tracked total distance traveled during a series of alternating light: dark periods for 5 dpf F1 larvae. Parental exposure at both 3 and 7 wpf showed a significant increase in mean distance traveled for both light (+13.4%; p < 0.001) and dark (+6.6%; p < 0.001) periods (Figure 6a). However, parental exposure at only 3 wpf significantly decreased distance traveled for both light (−22.6%; p < 0.001) and dark periods (−12.3%; p < 0.001; Figure 6b).
Figure 6. Average distance traveled (cm) per 30 s intervals in dark and light periods for 5 dpf F1 lineage larvae of DMSO- and TCDD-exposed parents. a . F1 larvae whose parents were exposed at both 3 and 7 wpf. b . F1 larvae whose parents were exposed at 3 wpf only. Fish were evaluated for distance traveled in a 24 min series of alternating 3 min light and dark intervals. Statistical comparisons are made between DMSO and TCDD groups for dark and light intervals. *** indicates significant p-value < 0.001 . DMSO, dimethyl sulfoxide; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; abn., abnormality; dpf, days post fertilization; wpf, weeks post fertilization.
F1 larvae – developmental abnormalities
We examined the effects of juvenile TCDD exposure at 3 and 7 wpf on the prevalence of developmental abnormalities in the F1 offspring. F1 larval fish with both parents developmentally exposed to TCDD at both timepoints showed significant increases in skeletal abnormalities (+56.7%; p < 0.05), cardiac edema (+52.7%; p < 0.001), and yolk sac edema (+39.1%; p < 0.01; Figure 7) compared to controls. However, fewer exposure-lineage fish were unhatched by 5 dpf (−28.4%; p < 0.01; Figure 7). In contrast, when both parents were exposed at only 3 wpf, there was no effect on abnormalities in the offspring (Supplemental Fig. S5A). We also examined effects of developmental exposure of only one parent to TCDD. Significantly more offspring of exposed males were unhatched at 5 dpf than offspring of exposed females (+87.7%; p < 0.001; Figure 7). No other endpoints were affected by one-parent exposure at both timepoints (Figure 7). When the parent was exposed at only 3 wpf, offspring of exposed males trended towards an increase in cardiac edema compared to offspring of exposed females (+567.7%; p < 0.1; Supplemental Fig. S5A). There was no change otherwise in the prevalence of abnormalities.
Figure 7. Percentage of fish at 5 dpf with developmental abnormalities in F1 lineage of DMSO and TCDD-exposed fish. On alternating weeks, F0 DMSO and TCDD adults were either spawned as incrosses or outcrossed by sex between the two exposure groups. TCDD female indicates F1 fish had an exposed female parent (and control male parent), while TCDD male indicates an exposed male parent (and control female parent). F1 larvae were screened for presence of cardiac edema, skeletal abnormalities, hatch rate from chorion, and yolk sac edema. Statistical comparisons are made between DMSO and TCDD groups or between TCDD female and TCDD male groups. * indicates significant p-value < 0.05; ** indicates significant p-value < 0.01; *** indicates significant p-value < 0.001 . DMSO, dimethyl sulfoxide; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; abn., abnormality; dpf, days post fertilization.
Discussion
Here, we report multigenerational impacts on neurobehavior and body morphometrics across the lifespan as a consequence of sublethal TCDD exposure during juvenile development. To our knowledge, while others have investigated TCDD-induced cardiometabolic behaviors Reference Marit and Weber37 and the impact of AHR2-dependent signaling on adult neurobehavior, Reference Garcia, Bugel, Truong, Spagnoli and Tanguay38,Reference Shankar, Garcia and La Du39 ours is one of very few studies using the zebrafish model to characterize anxiety-related behavioral impacts of juvenile TCDD exposure, and the first of such to report neurobehavioral sequelae in offspring. Specifically, we found that juvenile TCDD exposure increased boldness and anxiolytic behavior in adult fish of both sexes, and that the larval offspring of these fish were hyperactive compared to controls.
Our outcomes correspond with previous findings: developmental exposure to dioxins and/or PCBs has been repeatedly associated with increased hyperactivity and other attention deficit hyperactivity disorder (ADHD)-linked phenotypes in rodents Reference Hojo, Kakeyama, Kurokawa, Aoki, Yonemoto and Tohyama40,Reference Endo, Kakeyama and Uemura41 and across cohorts of Vietnamese, Reference Tran, Pham-The, Pham, Vu, Luong and Nishijo31,Reference Pham-The, Nishijo and Pham42 German, Reference Neugebauer, Wittsiepe, Kasper-Sonnenberg, Schoneck, Scholmerich and Wilhelm43 American, Reference Sagiv, Thurston, Bellinger, Altshul and Korrick44 and Canadian Reference Sussman, Baker and Wakhloo45 children. Such phenotypes are thought to be mediated by multiple interconnected neurodevelopmental pathways also targeted by TCDD exposure, including dysregulation of thyroid and steroid hormone homeostasis, Reference Winneke, Walkowiak and Lilienthal46–Reference Karman, Basavarajappa, Craig and Flaws50 cholinergic dysfunction, Reference Xie, Ma and Fu51,Reference English, Hahn and Gizer52 altered dopaminergic metabolism, Reference Whalley53–Reference Akahoshi, Yoshimura, Uruno and Ishihara-Sugano55 and AhR-mediated mitochondrial dysfunction. Reference Marazziti, Baroni and Picchetti56,Reference Hwang, Dornbos and LaPres57
The zebrafish, an established model organism for neurodevelopmental disease due to shared core neural circuits and neurotransmission pathways, high-throughput, accessible early developmental periods, and an expanding behavioral repertoire, is an emerging tool to investigate the etiology of ADHD and related disorders. Reference Whalley53,Reference Fontana, Franscescon, Rosemberg, Norton, Kalueff and Parker58 Phenotypes such as hyperactivity and boldness/impulsivity can be measured using the larval light/dark assay, which reports changes in locomotion as stimulated by light/dark transitions, and the novel tank test (NTT), which reports both locomotion and anxiety-like behaviors. NTT interpretation is based on the innate tendencies of fish to exhibit anxiety-driven thigmotaxis and diving behavior in a novel environment. This response is attenuated over time, as fish gradually explore upper and central zones of the tank. Reference Levin, Cerutti and Buccafusco59,Reference Haghani, Karia, Cheng and Mathuru60 Thus, the decreases we observed in both latency to enter anxiogenic zones and time spent in anxiolytic zones indicate a pattern of TCDD-induced decreased anxiety/increased boldness, which can be a maladaptive phenotype in nature, increasing predation risk. Reference Réale, Reader, Sol, McDougall and Dingemanse61 This aligns with the hyperactivity observed in larval offspring, supporting a persistent hyperactive/impulsive phenotype due to TCDD exposure. While adult fish did not show an increase in overall locomotion, as might be expected in a hyperactive phenotype, recent work suggests a more complicated relationship between these variables, as limited correlation was found between overall adult locomotion and anxiety behaviors across multiple assays of anxiety-like behavior in zebrafish. Reference Johnson, Loh and Verbitsky62 Thus, future inclusion of a broader suite of behavioral assays, including startle response, light/dark adult movement, shoaling behavior, predator avoidance, attention/impulsivity, and avoidance learning tests, will complement this data, helping to pinpoint the specific ADHD-linked neurophenotypes affected.
While certain neurodevelopmental and cognitive outcomes of early exposure to TCDD and other PAHs have been reported in adulthood in humans, rodents, and zebrafish, Reference Vu, Pham and Yokawa23,Reference Crépeaux, Bouillaud-Kremarik, Sikhayeva, Rychen, Soulimani and Schroeder33,Reference Geier, James Minick and Truong63 less is known about the longitudinal persistence of TCDD-induced deficits in executive functioning, behavior regulation, and ADHD outcomes. Our findings in zebrafish suggest that these neurobehavioral phenotypes, well-documented during early developmental periods, are also capable of persevering into adulthood and the next generation. Epimutations, or aberrations in epigenetic profile, are likely candidates for mediating this persistence due to their relative stability and role in regulating transcriptome dynamics. Reference Gibney and Nolan64 Although the heritability of TCDD-induced ADHD and hyperactivity-linked phenotypes remains to be determined, developmental exposure of zebrafish to another AhR agonist, benzo[a]pyrene, resulted in transgenerational (to the F2 generation) hyperactivity partially mediated by changes to the methylome. Reference Knecht, Truong, Simonich and Tanguay65 Previously, we have also implicated methylomic epimutations in zebrafish gonads as a factor in transgenerational infertility caused by TCDD exposure. Reference Akemann, Meyer, Gurdziel and Baker66,Reference Akemann, Meyer, Gurdziel and Baker67 DNA methylation is increasingly considered a promising epidemiological biomarker of ADHD risk; Reference Cecil and Nigg68 thus, the dysregulation of DNA methylation and other epigenetic pathways likely contributes to our persistent neurobehavioral effects. Such effects may also continue into F1 adulthood and subsequent unexposed generations; future studies will couple multigenerational behavior assays with transcriptomic and epigenetic interrogation of zebrafish brains to clarify the extent and mechanisms of pathway disruption.
We observed divergent effects on offspring behavior dependent upon parental exposure window: fish with parents only exposed at 3 wpf were hypoactive compared to controls, contrasting with the hyperactivity seen in fish with parents exposed at both timepoints. While both patterns of exposure indicate neurodevelopmental dysregulation, this contrast further highlights the importance of exposure timing in determining effect, as the parental gonad is at different stages of differentiation and maturation at 3 and 7 wpf. Isolating the 7 wpf window in future will help to tease apart the degree to which the number and timing of exposures influences the shift between hypo- and hyperactivity in affected offspring.
The adverse effects of TCDD exposure on fish body morphometrics, including axial and craniofacial skeletal development, eye size, swim bladder formation, and operculum growth, have previously been well-characterized, Reference Burns, Peterson and Heideman69–Reference Yue, Peterson and Heideman74 albeit mostly during embryonic/larval development or adulthood. Considering that zebrafish undergo rapid, hormone-mediated growth and remodeling across juvenilehood, Reference Mork and Crump75–Reference McMenamin and Parichy77 the longitudinal effects of sublethal TCDD exposure on juvenile growth trajectories are comparatively understudied. Thus, we investigated growth measures across juvenile and early adult timepoints, reporting decreases in several body morphometrics as an effect of developmental TCDD exposure.
The most pronounced effect of exposure throughout the juvenile period was a persistent reduction in head-trunk angle across the later three timepoints, indicating a minor increase in spinal curvature. While nonlethal, such spinal curvature in fish can impair the efficiency of swimming, feeding, reproduction, and antipredation behavior across the lifespan. Reference Bengtsson and Brown78 This outcome in juvenilehood corresponds with, and is a likely precursor to, our previous observations of scoliosis-like vertebral kinks in adult fish exposed to the same juvenile TCDD dosage schema, Reference Baker, Peterson and Heideman8,Reference Baker, Peterson and Heideman35 as well as in Japanese medaka exposed to TCDD as embryos. Reference Watson, Planchart, Mattingly, Winkler, Reif and Kullman71 As observed in medaka and zebrafish larvae, TCDD-induced dysregulation of osteoblast differentiation may be a potential mechanism contributing to our findings of progressive, persistent spinal curvature; whole genome transcriptomic studies in zebrafish will help to further pinpoint affected pathways involved in skeletal development across juvenilehood. Reference Burns, Peterson and Heideman69,Reference Watson, Planchart, Mattingly, Winkler, Reif and Kullman71
While the effects of TCDD exposure on other juvenile morphometrics were generally transient, we observed a few distinct trends. Craniofacial measures, including snout length and eye size, were most consistently decreased (or trending thereto) following the initial 3 wpf exposure, with intermittent decreases at later timepoints. However, such effects on trunk/body morphometrics, including vertical fish height, operculum area, and swim bladder area, were only present after the second exposure at 7 wpf. The difference in timing suggests that craniofacial and trunk growth trajectories have time- and tissue-specific patterns of susceptibility to TCDD insult across juvenilehood, resulting in minor growth truncation that mostly disappeared by early adulthood (12 wpf). Such compensatory growth after removal of a suppressive insult is a well-studied phenomenon, particularly in aquatic species. Reference Alvarez and Farrell79,Reference Broekhuizen, Gurney, Jones and Bryant80
This pattern of accelerated growth was also observed in other body morphometrics assessed throughout early to mid-adulthood: weight and BMI of male fish exposed at both 3 and 7 wpf were increased at 4 mpf, while both weight and length were increased at 7 mpf. Male fish only exposed at 3 wpf also showed increased weight at 7 mpf, indicating that the initial exposure alone is sufficient to alter metabolism months later in adulthood. While acute, high-dose exposure to TCDD classically leads to reduced feed intake, wasting syndrome, reduced body mass, and delayed growth trajectories as overt indicators of developmental toxicity, studies of TCDD exposure at lower doses and over longer latencies to outcome reveal a more complex relationship with growth and metabolism. Reference Pohjanvirta and Tuomisto10,Reference Lindén, Lensu, Tuomisto and Pohjanvirta13,Reference Girer, Tomlinson and Elferink14,Reference Brulport, Le Corre and Chagnon81–Reference Gray, Ostby and Kelce84 These outcomes range from inverse correlations with adiposity and BMI to increased BMI and accelerated growth metrics in early childhood, often after an initial period of growth restriction. Reference Wohlfahrt-Veje, Audouze and Brunak15–Reference Iszatt, Stigum and Govarts19
Accelerated growth trajectories post developmental insult are a known risk factor for adverse metabolic health outcomes across the lifespan, due to early adaptive shifts in metabolic programing that persist into adulthood. Reference Cauzzo, Chiavaroli, Di Valerio and Chiarelli85 In fact, TCDD exposure early in life has been linked to later-life phenotypes including dyslipidemia and altered glucose homeostasis pathways Reference Brulport, Le Corre and Chagnon81,Reference van Esterik, Verharen and Hodemaekers86–Reference Pelclová, Urban and Preiss88 that we previously reported as dysregulated in the gonads of developmentally exposed adult zebrafish and their offspring. Reference Baker, Yee, Meyer, Yang and Baker89,Reference Meyer, Baker and Baker90 Underlying effects on metabolism may also contribute to the behavioral outcomes we observed in adults and F1 offspring, as metabolic rate has been positively correlated with boldness and activity level across bird, rodent, and fish models. Reference Biro and Stamps91 Our findings of accelerated growth throughout early adulthood thus suggest persistent, likely compensatory, metabolic dysregulation across the lifespan as a mechanism for adverse phenotypes in adulthood; longitudinal transcriptomic and epigenomic analysis will further determine the timing and progression of such pathway changes. Notably, TCDD only affected the growth rate of male fish in our study; this finding corresponds with the complex, often conflicting, sex-specific growth outcomes of TCDD exposure reported across humans and rodents. Reference Su, Chen, Chen and Wang16,Reference Warner, Rauch, Brambilla, Signorini, Mocarelli and Eskenazi17,Reference Brulport, Le Corre and Chagnon81,Reference van Esterik, Verharen and Hodemaekers86 While it is expected that innately sex-divergent patterns of hormone regulation would also be differentially disrupted by exposure, contradictory outcomes across the field suggest a currently undetermined multifactorial relationship between sex, dosage, environmental exposure history, available nutrition, and latency to outcome that deserves further investigation.
While direct exposure of larval teleost fish to TCDD and other PAHs is commonly linked to the development of blue sac syndrome, comprised of several cardiometabolic, edematous and skeletal phenotypes, Reference Billiard, Querbach and Hodson92,Reference Carney, Prasch, Heideman and Peterson93 our findings indicate that such outcomes clearly persist to the next generation, affecting larval fitness. In larvae spawned from exposed fish, we observed an increased incidence of 3 of 4 apical endpoints linked to decreased population recruitment, including cardiac edema, yolk sac edema, and skeletal deformities, similarly to previous studies that exposed parental larval and juvenile fish to TCDD weekly from 0 to 7 wpf. Reference King Heiden, Spitsbergen, Heideman and Peterson73 Several factors likely contribute to these findings in the F1 generation, including direct exposure to TCDD as susceptible developing germ cells, indirect longitudinal effects on F0 parental growth and metabolism across the lifespan, and persistent epimutations, as reported in our previous work. Reference Akemann, Meyer, Gurdziel and Baker67 Interestingly, the offspring of exposed fish were also more likely than controls to have hatched from the chorion membrane by 5 dpf. Decreased hatch rate is typically an indicator of general toxicant-induced developmental delay and/or direct defects in chorion toughness and hatching enzyme secretion. Reference Ong, Zhao and Thistle94–Reference Ahmad, Ali and Richardson96 However, a more rapid hatch rate compared to controls can also indicate dysregulation: as the hatching of zebrafish embryos is semiplastic and responsive to environmental stimuli, exposure to stressors or alarm cues from conspecifics can increase hatch rate, albeit with corresponding developmental and functional tradeoffs. Reference Wisenden, Paulson and Orr97 Another factor that influences hatching is the spontaneous movement of the embryo within the weakened chorion; Reference Kim, Sun and Yun98 our reports of larval hyperactivity in exposure-lineage fish may thus contribute to this increase in hatch rate. Previous work with endocrine-disrupting triazole fungicides implicated dopaminergic signaling in the dysregulation of both embryonic activity level and hatching enzyme secretion, suggesting a potential mechanism for future consideration. Reference De la Paz, Beiza, Paredes-Zúñiga, Hoare and Allende99
We investigated the impact of exposure window on F1 abnormalities: fish with parents only exposed at 3 wpf were not overtly affected, demonstrating that either at least two “hits” during juvenile development are necessary for persistent effects on the offspring, or that the impact to developing germ cells is more severe at 7 wpf, later in the process of gonad differentiation and maturation. Upon investigating outcomes driven by parental sex, we observed few effects overall in offspring with only one exposed parent, indicating that the contributions of both parents may be required to produce overt developmental phenotypes. In one exception, fewer offspring of exposed males hatched than that of exposed females, although when both parents were exposed, more offspring hatched in comparison with controls. Overall, we highlight the intergenerational persistence of TCDD-induced morphometric effects by reporting adverse developmental phenotypes in the offspring of exposed fish; future transcriptomic analysis of larval offspring will reveal pathways through which indirect germline exposure induces such outcomes.
In conclusion, we identified several transient morphometric outcomes: eye width, eye-snout length, and swim bladder area of exposed fish decreased intermittently across juvenilehood, while exposed males showed accelerated growth in early-mid adulthood (4 – 7 mpf) that tapered off by 12 mpf. Decreases in head-trunk angle in exposed fish were more persistent across juvenilehood, suggesting lasting changes in spinal curvature. These exposed fish had offspring with developmental abnormalities, including altered hatch rates and increases in edema and skeletal deformities. In adulthood, we report novel anxiolytic/increased exploratory behaviors of exposed fish; their larval offspring were likewise hyperactive, indicating heritable behavioral dysregulation as an outcome of early-life TCDD exposure. Parental exposure window influenced behavior and abnormality outcomes in F1 offspring: parental exposure at 3 wpf alone decreased activity of larval offspring and did not induce developmental abnormalities, contrasting with parental exposure at 3 wpf and 7 wpf. Overall, our findings of dysregulated morphometrics and behavior in a zebrafish model expand current knowledge of the long-term and intergenerational consequences of early-life dioxin exposure, with potential ramifications for exposed aquatic populations.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S2040174424000229.
Acknowledgments
We thank past and present members of the Warrior Aquatic, Translational, and Environmental Research (WATER) lab for their assistance with fish husbandry, laboratory maintenance, and their general advice and support on this project. Fish rearing, exposure, and data collection were performed at Wayne State University, and data analysis took place at the University of Florida.
Author contribution
Conceptualization: TRB, DNM; Data curation: CW, DNM, IS, BV, GG; Formal analysis: DNM, IS, GG, BV, EC, CCW, AH; Funding acquisition: TRB, DNM; Investigation: DNM, IS, BV, AP, JRB, SEG, AK, AMP; Methodology: TRB, DNM, CCW, BV; Project Administration/Supervision: TRB, DNM; Resources: TRB; Software: CCW, GG, DNM; Visualization: DNM, IS, GG, CCW; Writing- original draft: DNM, TRB; Writing -review & editing: all authors.
Financial support
Funding was provided by the National Institute of Environmental Health Sciences (R01 ES030722 to T.R.B., A.H., and D.N.M.; F31 ES030278 to D.N.M.), and the National Center for Advancement of Translational Sciences (K01 OD01462 to TRB). Additional funding included the University of Florida University Scholars Program (to G.G.).
Competing interests
None.
Ethical standard
The authors assert that all procedures contributing to this work comply with the ethical standards of the National Institutes of Health Guide for Care and Use of Laboratory Animals and have been approved by the Wayne State University Institutional Animal Care and Use Committee (protocol number 19-02-0938; approved 10/14/2020).
Open access
