Lipidomic and Neurobehavioral Alterations from Developmental Lead Exposure in Zebrafish

Abstract

Lead (Pb) exposure during early development remains a major public health concern, with no identified safe threshold for developmental neurotoxicity. This study examined the effects of embryonic Pb exposure at environmentally relevant low concentrations (0.01, 0.1, and 1 ppb; μg/L) on neurodevelopmental outcomes in zebrafish (Danio rerio). Embryos were exposed during embryogenesis from 1–72 hours post-fertilization (hpf), and larval zebrafish were evaluated for cognitive and behavioral performance, oxidative stress biomarkers, lipidomic alterations, and general developmental parameters. Pb-exposed larvae displayed significant cognitive impairments and anxiolytic-like behaviors, along with concentration-dependent disruptions in decision-making behaviors involving complex neural circuits. Developmental outcomes included altered spontaneous movement and heart rate variability. Biochemical assays revealed disrupted estradiol concentrations, late and improper modulation of antioxidant enzyme activities, and increased oxidative stress and lipid peroxidation. Lipidomic profiling showed early upregulation of triacylglycerol (TG) and cholesteryl esters (CE), accompanied by downregulation of sphingomyelin (SM) and phosphatidylcholine (PC) at 72 hpf. By 120 hpf, persistent reductions in phosphatidylethanolamine (PE), SM, and PC were observed, suggesting delayed metabolic reprogramming. Together, these results demonstrate even low Pb concentrations impair neurodevelopmental processes, alter lipid metabolism, and disrupt oxidative balance in zebrafish larvae. Importantly, the findings provide evidence of adverse effects below currently recognized exposure limits, underscoring the heightened sensitivity of the developing brain to Pb and raising concern for long-term cognitive and neurodevelopmental consequences in exposed populations.

Graphical Abstract

1. Introduction

Lead (Pb) is a pervasive environmental pollutant with well-documented neurotoxic effects, particularly in developing organisms [1–4]. Pb exposure, even at low concentrations during critical developmental windows, can produce lasting cognitive and behavioral deficits, with no safe threshold established to date [5]. Understanding the mechanisms of Pb developmental neurotoxicity at human-relevant levels is therefore essential for translational research and public health. Most laboratory studies have historically focused on higher Pb concentrations, whereas the neurodevelopmental effects of lower exposures aligned with contemporary exposure levels in children remain poorly understood, particularly regarding cognitive outcomes and lipid metabolism. Although environmental Pb levels in water are reported in some regions, the concentrations used in this study were selected to model early-life human exposure rather than environmental contamination. For context, the current blood lead reference value (BLRV) identified by the US Centers for Disease Control and Prevention (CDC) is 3.5 μg/dL (equivalent to 35 ppb), based on children aged 1–5 years in the top 2.5% of measured blood levels [6]. However, in utero exposure results in lower fetal blood Pb concentrations due to placental transfer and absorption dynamics. As such, in this study lower Pb concentrations were used in a zebrafish model to mimic these early developmental exposures. Even at these low levels, Pb has been associated with developmental delays, learning difficulties, and behavioral impairments in children [7–10], highlighting the need for studies investigating mechanisms of neurotoxicity at ultra-low exposures.

Mechanisms of Pb developmental neurotoxicity have been studied in multiple biological systems including rodents [11–14] and fish [3, 15, 16]. These studies support adverse neurological outcomes observed across species expanding the biological model toolbox by which mechanisms of Pb neurotoxicity can be studied. One major mechanism implicated in Pb neurotoxicity is oxidative stress, which arises from an imbalance between reactive oxygen species (ROS) and antioxidant defenses [17–19]. Pb interferes with both enzymatic (e.g., superoxide dismutase, catalase, and glutathione peroxidase) and non-enzymatic systems, depletes glutathione (GSH), and disrupts key enzymes such as glutathione reductase (GR) and glutathione S-transferase (GST) [20–23]. GST plays a crucial role in detoxification and in protecting neural cells from oxidative injury. Genetic polymorphisms in GST genes further modulate susceptibility to Pb, linking variability in antioxidant defenses to neurodevelopmental disorders in humans [24–27]. In addition to redox disruption, Pb alters lipid homeostasis, leading to lipid peroxidation and membrane dysfunction, processes tightly linked to neurotransmitter balance and brain development [28–34]. Such disturbances may contribute to behavioral impairments, including hyperactivity, anxiety-like behaviors, and cognitive deficits observed across species [35]. Pb also interferes with the expression of genes related to behavioral markers providing insight into molecular drivers of the neurobehavioral alterations observed in children and in laboratory animal models. For example, prior studies report Pb alters genes related to neuronal development [36, 37], which disrupt decision-making behavior in zebrafish larvae [38] as well as cause memory [39, 40] and learning [41, 42] impairments.

Pb environmental contamination has decreased significantly in the past several decades in many global regions given regulatory actions banning Pb in gasoline and paints. Modern Pb exposure in these global regions continues to occur but at much lower concentrations and varies depending on the source (e.g., soil, food, dust, water). Human exposure commonly occurs via ingestion of contaminated food or water. For example, drinking water in the US ranges from 1 to 3.88 ppb Pb depending on location and pollution source [43, 44]. Most zebrafish studies investigating Pb toxicity have used higher concentrations, typically in the range of 5–500 ppb, to induce observable developmental, behavioral, or biochemical changes [45–47]. However, environmental and translational relevant exposures are often lower than these experimental concentrations and the effects of such low-dose exposures remain poorly understood.

Given the need to further the understanding of Pb developmental neurotoxicity at lower exposure levels, this study investigated the effects of an embryonic Pb exposure (1–72 hours post fertilization, hpf) in developing zebrafish at 0.01, 0.1, and 1 ppb (μg/L; equivalent to 0.001, 0.01, and 0.1 μg/dL, respectively). Zebrafish (Danio rerio) are emerging as a particularly powerful model due to their genetic similarity to humans, transparent embryos, conserved antioxidant and metabolic systems [48–52], and well-characterized behavioral repertoire [15, 53]. Importantly, Pb can penetrate the zebrafish chorion [54, 55], underscoring the embryo’s vulnerability to even minimal exposure. Survival, hatching, anxiety-like, cognition, and decision-making behaviors, oxidative stress biomarkers, antioxidant enzyme activity, and lipidomic profiles were evaluated. To our knowledge, this is the first study to examine Pb developmental neurotoxicity at concentrations as low as 0.01 ppb representing a level below to those detected in the blood of pregnant women [56] and thereby, providing critical insight into the earliest thresholds of Pb-induced developmental neurotoxicity.

While Pb contamination in aquatic environments is also a recognized concern, the primary objective of this study was not to address impacts on aquatic organisms. Instead, we employ zebrafish as a translational model to investigate how Pb exposure during early development (1–72 hpf) affects neurobehavioral outcomes, oxidative stress, endocrine function, and lipid metabolism. Zebrafish offer conserved neurodevelopmental and molecular pathways that allow the study of mechanisms relevant to human early-life exposure. This approach enables the assessment of mechanistic endpoints in a controlled experimental setting, providing insights into potential risks for developmental neurotoxicity in humans and complementing studies completed in other biological translational models.

2. Material and Methods

2.1. Experimental design and exposure paradigm

Adult zebrafish were housed in a recirculating system (Iwaki Aquatic, Holliston, MA, USA) at 28°C, pH 7.0–7.2, conductivity at 530–550 μS, DO at 7.3–8.3 ppm, and on a 14:10 light-dark cycle. Water quality was assessed two times a day with adult fish fed a mixture of brine shrimp (Artemia International LLC., Fairview, TX, USA), Golden Pearls 500–800 μm (Artemia International LLC., Fairview, TX, USA), and Zeigler adult zebrafish food (Zeigler Bros Inc., Gardners, PA, USA). To attain embryos, seven-month-old wild-type zebrafish (Danio rerio, 5D-strain) were bred following established laboratory procedures [37]. Briefly, seven days prior to breeding, fish were separated by sex, kept in the recirculating system, and fed twice a day. On the day prior to breeding (i.e., 12 hours before), fish were placed in a breeding tank with a compartment for embryo collection at the bottom. Fish were kept sex-separated in the breeding tank overnight divided by a barrier at a density of 3 females to 1 male. Environmental enrichment was provided in the breeding tank to avoid stress in the new environment. On day of breeding, the barrier was removed one hour after the lights turned on and the fish were allowed to breed for 50 min. After 50 min, the fish were removed from the breeding tank and placed back into the recirculating system.

Embryos were collected at 2 to 16 cell stage, washed with ‘embryo water’ (i.e., RO water with pH at 7.4, conductivity at 530 – 550 μS, and with Na⁺, K⁺, Ca²⁺, Mg²⁺, Cl⁻, and HCO₃⁻ as dissolved cations and anions to ensure proper embryonic development), and used for experimental procedures by either random sorting into 100 × 20 mm polystyrene Petri dishes (i.e., 50 embryos per plate considered as subsamples submerged in 20 mL solution) or into 12-well polystyrene plates (i.e., 25 embryos per well considered as subsamples submerged in 10 mL solution). The experimental design was a randomized complete block design (RCBD) with Petri dish or well plate as the blocking factor; thus, fish within a Petri dish or well plate were considered as subsamples. Pb treatment solutions [as Pb(II) acetate trihydrate, CAS #6080–56-4, ≥99.99% purity, Sigma Aldrich, St. Louis, MO, USA] were prepared in embryo water with exposures consisting of 0 (negative control; embryo water only), 0.01, 0.1, or 1 ppb (μg/L). Pb concentrations in the treatment solutions were confirmed by Inductively Coupled Plasma - Mass Spectrometry (ICP-MS) (ELEMENT-2, Thermo Fisher Scientific, Waltham, MA, USA) at the Department of Chemistry, Purdue University, Research Instrumentation Center (Supplemental Table 1). The 0.1 and 1 ppb Pb treatment solutions were near the expected concentration, while the 0.01 ppb Pb treatment solution was below the level of detection (LOD). The 0 ppb Pb negative control treatment solution was also confirmed to be below the LOD.

Pb treatment was terminated at 72 hours post fertilization (hpf), corresponding to the completion of embryogenesis. By this stage, all major embryonic developmental processes are completed. The cessation of exposure was done by rinsing twice with embryo water to remove the exposure contact. Developmental stage of evaluation was optimized for each endpoint. For evaluations completed following exposure cessation (i.e., beyond 72 hpf), fish continued to develop in embryo water only. Specifically, spontaneous movement was measured at 24 hpf, heart rate at 48 hpf, and hatching and survival rates throughout the experiment. Biochemical parameters, estradiol concentrations, and lipidomics were completed immediately at exposure cessation (i.e., 72 hpf) and 48 hours after exposure cessation (i.e., 120 hpf). For the behavioral assays, the visual motor response test and the shoaling tests were completed at 120 hpf, the open field test at 144 hpf, and the aversive stimuli at 168 hpf (Supplemental Figure 1). Protocols were approved by the Purdue University Animal Care and Use Committee (no. 1110000088) with all fish treated humanely.

2.2. Developmental parameters for life-quality biomarkers

During the developmental exposure period, life-quality biomarkers including spontaneous movement (SM), heart rate (HR), hatching curves (HC), and survival curves (SC) were measured. At 24 hpf the embryo, although still inside the chorion, is moving and permits the evaluation of spontaneous movements. For the SM test, embryos were placed under a light stereoscope and the total SM was counted for 1 min. A total of 5 fish (subsamples) were evaluated in each treatment group per replicate with a total of 8 replicates assessed (n = 40) [57]. HR was evaluated at 48 hpf. By 48 hpf the fish have clear cardiac function but low movement, which facilitates visualization and evaluation of heart beats. For this assessment, embryos were placed under a light stereoscope, and the total HR was counted for 20 sec. The final HR was expressed in heart beats/min. A total of 5 fish (subsamples) were evaluated in each treatment group per replicate with a total of 8 replicates measured (n = 40) [57]. For SM and HR assessments, fish were exposed in Petri dishes as described above and maintained in the exposure solution during the tests.

The HC was evaluated through monitoring of the total hatching number through 72 hpf; at which time in all experiments hatching was complete for all fish. Hatching rates were collected every 24 hours, where the total number of hatchings was counted under a light stereoscope. In addition, survival rates were measured every 24 hours under a light stereoscope until the experimental procedures were completed to determine the SC. If dead embryos/larvae were observed, they were immediately removed following the daily evaluation. Similarly for HC and SC, the embryos were exposed in Petri plates as described above with 50 fish per plate. Petri plates are considered as the blocking factor with a total of 8 replicates per treatment group (n = 400) [57].

Behavioral and biochemical assessments were conducted at distinct developmental stages to capture different aspects of neurobehavioral maturation following embryonic Pb exposure. The 72 hpf time point corresponds to the end of embryogenesis at which biochemical responses were evaluated immediately following Pb cessation, providing insight into processes occurring during Pb contact [58, 59]. The 120 hpf stage represents a key neurodevelopmental phase when the central nervous system is functionally mature, glutamatergic signaling is fully excitatory, and GABAergic transmission has become inhibitory. At this stage, behaviors such as shoaling and visual motor responses are well established. The open-field test (144 hpf) and aversive stimuli test (168 hpf) were selected to assess anxiety-like and cognitive-related behaviors, respectively, which emerge during these later larval stages [60]. This design allowed isolation of the consequences of embryonic Pb exposure while avoiding confounding effects of post-embryonic exposure.

2.3. Cognition testing

For cognition testing, the shoaling test (ST) and the aversive stimuli test (AST) were used to evaluate behavioral parameters following the embryonic Pb exposure. The ST was performed to access early-life development of social behaviors as indicators of developing decision-making and sensory integration rather than fully mature shoaling. The ST was performed as described by Tamagno et al. [61] with some adaptations. Four 120 hpf larvae were placed in each well of a 6-well plate. After a 2 min adaptation, the larvae were recorded for 6 min using a camera (Ikegami model: I CD-49E, type: REV, 768 × 576 pixels (PAL), analog output). After filming, the video was cut into 30 second segments to produce images. The distance between each fish in the shoal was measured in each image using ImageJ software (ImageJ2, Fiji). The final distance per image was considered as mean of the distance between the fish in each well. A total of 16 replicates were performed with 3 technical replicates (in each technical replicate 4 fish were used for ST and 5 for AST) per replicate (n = 48).

The AST was employed to evaluate the larvae’s ability to visually detect potential threats from predators and to navigate away from perceived risk giving a clear pattern of the cognitive status of the larvae. To perform the AST, five 168 hpf larvae were placed in a well of a 6-well-plate filled with 10 mL of embryo water. The plates were positioned above an LCD monitor and after a 2-minute adaptation period, fish were exposed to a visual stimulus. This stimulus consisted of a red sphere measuring 1.35 cm, following a trajectory covering only half of the well [62]. The movement of the sphere was controlled using PowerPoint software, and larvae were subjected to the aversive stimulus for 5 min. The number of larvae remaining in the stimulus area was recorded at the end of the 5 min of continuous stimuli [63]. A total of 16 replicates was carried out per treatment group with 3 technical replicates per replicate (n = 48).

2.4. Anxiety-like and decision-making behavior

For decision-making behavior, the visual motor response test (VMRT) was applied, by evaluating the swimming response triggered by a sudden change in light [64]. For this test, 120 hpf fish were placed individually in a 96-square-well plate filled with 500 μL of embryo water. Fish were acclimated for a 10 min dark period in the Noldus DanioVision Chamber (Noldus, Leesburg, VA, USA) and then subjected to five alternating 10 min periods of dark and light (i.e., 10 min dark, 10 min light, 10 min dark, 10 min light, and 10 min dark) [65]. Infrared light that is not visible to zebrafish larvae was used for tracking movement [66]. During the light phase, a 5000 lX light was activated under the DanioVision Chamber. The infrared movement traces were recorded at a rate of 25 fps with a Basler GenICam acA 1300–60 gm camera. Tracks were smoothed via a minimum distanced moved profile set to >0.2 mm. The total distance traveled (mm), the time spent moving (s), and velocity (mm/s) were evaluated during each phase using the Ethovision XT 17.5 software (Noldus, Leesburg, VA, USA). 16 replicates with 12 subsamples per replicate (192 fish total per treatment group) were tested.

Anxiety-like behavior was evaluated using the open field test (OFT). At 144 hpf, larvae were placed individually in a well of a 6-well plate filled with 10 mL of embryo water. Subsequently, fish were recorded for a duration of 6 minutes in a Noldus DanioVision Chamber (Leesburg, VA, USA), consisting of a 1 minute acclimatization period followed by a 5 minute exploration phase, as described previously [67]. Behavioral variables were assessed by evaluating 16 replicates with 6 technical replicates each. The recording was analyzed using the automated tracking software Ethovision XT 17.5 (Noldus, Leesburg, VA, USA). To examine thigmotaxic behavior, the testing well was conceptually divided into a central and peripheral/wall zone [57]. The parameters evaluated were total distance traveled (mm), total time spent moving (s), distance traveled in the center (mm), time moving in the center (s), distance traveled in the periphery (mm), time moving in the periphery (s), and the latency to the center zone (s). A total of 16 replicates was carried out per treatment group with 3 technical replicates per replicate (n = 48).

2.5. Antioxidant system evaluation

2.5.1. Extract preparation

The antioxidant system evaluations were performed at two developmental stages: 72 hpf (the end of embryogenesis and time point of Pb exposure cessation) and 120 hpf (larval period at which all major organ development is complete and 48 hours after exposure cessation). The two developmental stages were included to capture immediate profiles following exposure cessation at end of embryonic development and to address modulation occurring to reestablish oxidative homeostasis following exposure cessation as the fish continues to develop. Extract preparations at 72 and 120 hpf were performed as previously described with some modifications [68]. To evaluate the first and second line of antioxidant defense, 45 fish from each Petri dish (as described above) at each time point were collected and pooled into a 1.5 mL tube. Any remaining solution was removed, 300 μL of ultrapure water added, and samples immediately flash frozen in liquid nitrogen. After flash freezing, samples were defrosted and ultrapure water removed. Fish were then resuspended in 250 μL of Tris HCl buffer (pH 7.4, 50 mM) and subjected to three cycles of flash freezing in liquid nitrogen and thawing. Following the third freezing cycle, samples were homogenized on an ice bath for 1 min using a RNAse-DNAse pellet pestle (Kimble Kontes). After homogenization, 750 μL of Tris HCl (pH 7.4, 50 mM) was added to the samples and mixed by vortexing. Samples were then placed in an ultrasound bath for 10 min and centrifuged at 1500 × g for 10 min at 4°C. After the first centrifugation cycle, 250 μL of the supernatant was transferred and stored in a new 1.5 mL tube at −80°C for lipid peroxidation and estradiol concentration evaluation. The remaining supernatant (750 μL) was transferred to a new 1.5 mL tube and centrifugation repeated for 10 min at 13000 × g at 4°C. After the second centrifugation cycle, 750 μL of the cytosolic sample was transferred and stored in a new 1.5 mL tube for further evaluation of first and second line of antioxidant defense. The remaining pellet from the second centrifugation contained mitochondrial sample, which was resuspended in 100 μL of Tris HCl (pH 7.4, 50 mM), vortexed, placed in an ultrasound bath for 10 min, immediately fresh frozen in liquid nitrogen, and kept at −80°C for mitochondrial portion evaluation. Protein quantification and quality monitoring for all assays were assessed using the NanoDrop One spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).

2.5.2. First line of antioxidant defense (FLAD)

All assays were performed using commercial kits following the manufacturers’ protocols. Proprietary reagents were used as supplied (please note, the exact compositions are not disclosed by the manufacturer). No modifications to the standard sample preparation procedures were performed to ensure reproducibility. Specific procedural details, including wavelengths, probes, and calibration steps, were performed as recommended in the kit manuals.

The first line of antioxidant defense was evaluated by measuring mitochondrial Mn-superoxide dismutase (Mn-SOD) activity, cytoplasmatic Cu-Zn superoxide dismutase activity (Cu/Zn-SOD), and catalase activity (CAT). All enzymes were evaluated using an ELISA kit with a 96-well plate design from Cayman Chemicals (Ann Arbor, Michigan, USA) including SOD assay kit (No. 706002) and CAT assay kit (No. 707002). For quantification of the mitochondrial Mn-SOD enzyme, in addition to the preparation of mitochondrial suspension described above, 10 μL of potassium cyanide was added to each reading well (i.e., final concentration in the well of 3 mM). The addition of potassium cyanide inhibited both Cu/Zn-SOD resulting in the detection of only Mn-SOD. To further confirm that T-SOD activity was directly affected by Pb²⁺, an in vitro assay was performed using a wider range of Pb²⁺ concentrations (0, 0.01, 0.1, 1, 5, and 10 ppb final concentration in the well). A control bovine erythrocyte T-SOD preparation (0.03 U/mL; Cayman Chemical, Kit No. 706002, Item No. 706005) was used. Pb²⁺ was added to each well at the respective concentrations, and T-SOD activity was measured to determine the extent of enzyme inhibition under increasing Pb²⁺ exposure. A total of 12 replicates (each consisting of a pool of 50 fish per Petri dish) per treatment group per time point with 3 technical replicates were performed for each enzyme (n = 12).

2.5.3. Second line of antioxidant defense (SLAD)

The second line of antioxidant defense was evaluated by measuring glutathione-S-transferase activity (GST), glutathione peroxidase activity (GPx), and glutathione reductase activity (GR). All enzymes were evaluated using an ELISA kit with a 96-well plate from Cayman Chemicals (Ann Arbor, Michigan, USA; GST, No. 703302; GPx, No. 703102; GR, No. 703202) following manufacturer recommendations. A total of 12 replicates (each consisting of a pool of 50 fish per Petri dish) per treatment group per time point with 3 technical replicates were performed for each enzyme (n = 12).

2.5.4. Lipid peroxidation (TBARS) and estradiol concentration

Lipid peroxidation was evaluated in cytoplasm by measuring Tiobarbituric Reactive Species (TBARS) using an ELISA kit with a 96-well plate design from Cayman Chemicals (Ann Arbor, Michigan, USA; No. 10009055) following manufacturer recommendations. Estradiol concentrations were quantified using an ELISA kit with a 96-well plate design from Cayman Chemicals (Ann Arbor, Michigan, USA; No. 501890). A total of 12 replicates (each consisting of a pool of 50 fish per Petri dish) per treatment per time point group with 2 technical replicates were performed for each enzyme (n = 12).

2.6. Lipidomic profiling

2.6.1. Lipid extraction

Lipid profiling analyses were performed through multiple reaction monitoring (MRM) profiling [69, 70]. For each concentration group, eight individual whole-larvae were processed and evaluated as follows. At 72 or 120 hpf, a single individual larva was collected and transferred to a 1.5 mL tube placed on ice to reduce its activity. After movement ceased, remaining water inside the tube was removed and 50 μL of ultrapure water was added. Tubes were flash frozen in liquid nitrogen and kept for a week at −80°C until processed. For processing, samples were defrosted and homogenized for 30 sec in ultrapure water using pellet pestles (Kimble Kontes). After homogenization, the Bligh & Dyer protocol was performed as previously described [71]. In summary, 90 μL of methanol (HPLC grade) and 50 μL of chloroform (HPLC grade) were added. After adding each reagent, samples were mixed by pipetting several times to generate a single-phase solution. Samples were incubated for 15 min at room temperature for lipid extraction. 50 μL of ultrapure water was then added and samples mixed. Next, 50 μL of chloroform was added and samples mixed. Samples were then centrifuged at room temperature for 5 min at 5000 × g. Following centrifugation, the bottom phase was transferred to a new tube and allowed to dry for 12 h at room temperature in a SpeedVac. After the drying process, samples were resuspended in 200 μL of 300 mM acetonitrile:methanol:amonium acetate (v:v:v 3:6.65:0.35) with the final concentration of ammonium acetate in the solvent at 10 mM. A total of 16 biological replicates per time point (individual larvae) per treatment group were analyzed for each enzyme with 2 technical replicates (i.e., duplicate injections) per sample (n = 16).

2.6.2. Instrumentation and injection

For data acquisition, samples were diluted seven times using the same solution in which they were resuspended. The total detected lipids in the MRM profiling exploratory experiment and conditions were the same as previously described [72]. In summary, data acquisition was performed using flow-injection (no chromatographic separation) from 10 μL of the diluted lipid extract stock solution which was delivered using a micro-autosampler (G1377A) to the ESI source of an Agilent 6410 triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA). A capillary pump was connected to the autosampler and operated at a flow rate of 8 μL/min and pressure of 150 bar. The capillary voltage on the instrument was 5 kV and the gas flow was 5.1 L/min at 300°C.

2.7. Data processing and statistical analysis for lipid profiling

Lipid structure is described at the species level [73]; as such the lipid class as well as the number of carbon and unsaturation in the fatty acyl chains was tentatively attributed. For example, PC(34:1) means a phosphatidylcholine containing 34 carbon and 1 unsaturation in the two fatty acyl chains esterified to the lipid glycerol backbone. The ion signal for each MRM or ion transition was filtered considering lipids that were 30% higher than the blank. These lipids were then normalized by lipid class with the sum of the total lipid MRM intensity per class (i.e., sample normalization). After normalization, the MRM data was transformed by square root for data transformation and by Pareto scaling for data scaling using MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/). The normalized data was then statistically analyzed using MetaboAnalyst (6.0) for one-way ANOVA (FDR<0.05 and Tukey post hoc, and FDR correction), outlier removal, and PLS-DA analysis generation. Further, for the interpretation of affected lipid pathways, the lipid ontology (LION) enrichment analysis web (LION/web) [74] was used to assess the changes in patterns of lipids between the negative control and Pb exposure groups. In LION, the Target List mode was used to find enriched pathways based on the significantly changed lipids (one way ANOVA with FDR<0.05). The fold change and Log2-fold change, at the same development stages were calculated by Excel (v 2016) (FDR<0.05 with Tukey post hoc test). Data is available at https://purr.purdue.edu/publications/4933/1.

2.8. Statistical analysis of developmental parameters, behavior tests, and biochemical assessments

For all behavioral assays, 16 biological replicates were performed per treatment and parameter with each replicate consisting of a Petri dish containing 50 larvae. The number of technical replicates varied depending on the assay and is specified in the relevant sections above. For biochemical and lipidomic analyses, 16 biological replicates per treatment were analyzed at each time point (i.e., 72 hpf and 120 hpf) to ensure robust statistical power and reproducibility.

Survival rates were compared using the Gehan-Breslow-Wilcoxon test. All other data for developmental parameters, behavior tests, and biochemical assessments were first assessed for outlier evaluation using the Interactive Grubbs (α = 0.05). After outlier removal, data was tested for homoscedasticity by the Bartlett’s test and for normality using the Kolmogorov-Smirnov test. Parametric data was analyzed blocked by replicates within the randomized complete block design using a one-way ANOVA with a post hoc Fisher’s least significant difference test of LSD (α=0.05) and represented as mean ± SEM. The LSD test was chosen due to its greater sensitivity in detecting differences between groups in balanced designs with a relatively small number of comparisons, which is commonly applied in developmental toxicology studies. The exception was the VMRT and hatching curve which were analyzed by a repeated measures ANOVA (α=0.05) by phase. Nonparametric data were analyzed by Kruskal-Wallis with post hoc Dunn test (α=0.05) and represented as median ± interquartile range. Specific statistical tests applied are detailed in the figure legends. Statistical differences among treatment groups are represented as different lowercase letters on the error or interquartile range bars.

3. Results

3.1. Developmental parameters

Pb exposure treatments during embryogenesis did not alter survival (Supplemental Figure 2A) or hatching rates (Supplemental Figure 2B). At 24 hpf, SM in the 0.1 ppb Pb treatment group was increased compared to the negative control and other Pb treatment groups (Figure 1A). HR fluctuated among the Pb treatment groups with an increase observed in the 0.1 ppb Pb treatment group and a decrease occurring in the 1 ppb Pb treatment group compared to the negative control (Figure 1B).

Figure 1 –

Spontaneous movements at 24 hpf (A) and heart rate at 48 hpf (B) of zebrafish exposed to 0, 0.01, 0.1, or 1 ppb of Pb during embryogenesis (1–72 hpf). SM was analyzed by a Kruskal-Wallis test with a post hoc Dunn test and is represented as median ± interquartile range of 8 independent biological replicates (each replicate with 5 technical replicates per concentration resulting in a total of 40 individuals evaluated). HR was analyzed by a one-way ANOVA with post hoc LSD test and is represented as mean ± SEM of 8 independent biological replicates (each replicate with 5 technical replicates per concentration resulting in a total of 40 individuals evaluated). Groups were considered statistically different when p<0.05. Statistically different treatment groups are represented by different lower-case letters on the bars.

3.2. Cognition testing

Cognition was accessed using the ST and AST. In the ST, the distance among each fish comprising a shoal (Figure 2A) was decreased in the 1 ppb Pb treatment group. In the AST, the percentage of fish remaining in the stimuli zone after 5 min of stimuli was increased in the 0.1 ppb Pb treatment group (Figure 2B)

Figure 2 –

Shoaling test (ST) at 120 hpf represented by the distance among the fish (A) and aversive stimuli test (AST) at 168 hpf represented by the percentage of fish remaining in the stimuli zone after 5 min of visual aversive stimuli (B). ST distance among fish was analyzed by a one-way ANOVA with post hoc LSD test and is represented as mean ± SEM of 16 independent biological replicates (each well contained 4 individuals, with 3 technical replicates per concentration and 16 biological replicates, resulting in a total of 48 wells and 192 individuals evaluated). In the AST percentage of fish remaining in stimuli zone was analyzed by a Kruskal-Wallis test with a post hoc Dunn test and is represented as median ± interquartile range of 16 independent biological replicates (each well contained 5 individuals, with 3 technical replicates per concentration and 16 biological replicates, resulting in a total of 48 wells and 240 individuals evaluated). Groups were considered statistically different when p<0.05 with differences indicated by different lower-case letters on the bars.

3.3. Anxiety-like and decision-making behavior

In the VMRT, the distance traveled (Figure 3A) was decreased in 0.1 ppb Pb treatment group in the first light phase when compared to the negative control group. In addition, in the second light phase, the distance was reduced in the 0.01 ppb Pb treatment group when compared to the negative control group. Furthermore, the distance traveled was decreased in the last dark phase (phase 3) in all the Pb treatment groups when compared to the negative control. The time spent moving was also changed in all Pb treatment groups compared to the negative control group but only in the second light phase (Figure 3B). Alterations in velocity were similar to total distance moved with a reduction in the second light phase and final dark phase in the 0.01 ppb treatment group compared to the negative control group (Figure 3C).

Figure 3 -.

Visual Motor Response Test (VMRT) at 120 hpf evaluated by calculating the distance moved (A), time moving (B), and velocity (C) in five different 10-min phases: three dark (D1, D2, and D3) and two light phases (L1 and L2). Data is represented as mean ± SEM of 16 independent biological replicates (each replicate consists of 12 individuals evaluated per concentration resulting in a total of 192 individuals). Data was evaluated by a repeated measures ANOVA blocked by replicate for each phase. Statistical difference among treatment groups within each phase are represented by different lower-case letters on the bars (p<0.05).

For OFT, the total distance traveled (Figure 4A) and distance traveled in the center zone (Figure 4C) was increased in the 0.1 ppb Pb treatment group. However, the distance traveled in the periphery zone (Figure 4E) was not changed in any treatment group. The total time moving was not changed in any exposed group (Figure 4B). However, when evaluating time within each zone, time in the center was increased (Figure 4D) and time in the periphery zone was decreased in the 0.1 ppb treatment group. Finally, the latency to center zone (Figure 4G) was reduced in the groups exposed to 0.1 ppb and 1 ppb Pb when compared to negative control group.

Figure 4 -.

Open Field Test (OFT) evaluated by calculating the total distance traveled (A), total time moved (s), total distance traveled in the center (C), total time in the center (D), total periphery distance traveled (E), total time in the periphery (F), and latency to center zone (G). Data in panels C, D, and F are represented as mean ± SEM of 16 biological replicates (each biological replicate consists of 6 individuals to total 96 fish per treatment group) and were analyzed by a one-way ANOVA with post hoc LSD test. Data in panels A, B, E, and G are represented as median ± interquartile range of sixteen independent biological replicates (each biological replicate consists of 6 individuals to total 96 fish per treatment group) and evaluated by a Kruskal-Wallis test with post hoc Dunn test. Statistical differences among treatment groups are represented by different lower-case letters on the bars (p<0.05).

3.4. First and second lines of antioxidant defense

Cu/Zn-SOD at 72 hpf was increased in the 0.01 and 0.1 ppb Pb treatment groups and decreased in the 1 ppb Pb treatment group (Figure 5A). At 120 hpf, Cu/Zn-SOD was increased in the 0.1 and 1 ppb treatment groups (Figure 5B). The activity of Mn-SOD was consistently increased in the 0.01 ppb Pb treatment group at 72 (Figure 5C) and 120 hpf (Figure 5D). CAT activity was reduced in the 0.1 and 1 ppb Pb treatment groups at 72 hpf (Figure 5E) but no changes were observed at 120 hpf (Figure 5F).

Figure 5 -.

Activity of the cytoplasmatic Cu/Zn superoxide dismutase (Cu/Zn-SOD) at 72 hpf (A) and 120 hpf (B), mitochondrial Mn superoxide dismutase (Mn-SOD) at 72 hpf (C) and 120 hpf (D), and cytoplasmatic catalase (CAT) at 72 hpf (E) and 120 hpf (F). Data is represented as mean ± SEM of 12 biological replicates (each replicate consists of three technical replicates). Data was evaluated using a one-way ANOVA with post hoc LSD test. Statistical differences among treatment groups are represented by different lower-case letters on the bars (p<0.05).

GPx activity showed different changes at the two developmental stages with an increase at 72 hpf in the 0.1 ppb Pb treatment group (Figure 6A) and an increase at 120 hpf in the 1 ppb Pb treatment group (Figure 6B). The GR activity also had an increase in the 1 ppb Pb treatment at 72 hpf (Figure 6C) but saw an increase in the 0.01 and 1 ppb Pb treatment groups at 120 hpf (Figure 6D). GST activity also increased in the 1 ppb Pb treatment group at 72 hpf (Figure 6E) but no changes were observed at 120 hpf (Figure 6F).

Figure 6 –

Cytoplasmatic activity of the enzymes glutathione peroxidase (GPX) at 72 hpf (A) and 120 hpf (B), glutathione reductase (GR) at 72 hpf (C) and 120 hpf (D), and glutathione-S-transferase (GST) at 72 hpf (E) and 120 hpf (F). Data from wild-type zebrafish larvae embryonically exposed (from 1 to 72 hpf) to 0, 0.01, 0.1, or 1 ppb of Pb. Data is represented as mean ± SEM of 12 independent biological replicates (each replicate consists of three technical replicates). Data was evaluated using a one-way ANOVA with post hoc LSD test. Statistical differences among treatment groups are represented by different lower-case letters on the bars (p<0.05).

To further confirm T-SOD activity was directly affected by Pb²⁺, an in vitro assay using a broader range of six Pb²⁺ concentrations (0, 0.01, 0.1, 1, 5, and 10 ppb) was completed. A control erythrocyte T-SOD (0.03 U/mL) was used and Pb²⁺ added to each well at the respective concentrations. The T-SOD activity in the 0.01, 0.1, and 1 ppb Pb treatments remained unchanged, while T-SOD activity decreased in the 5 and 10 ppb Pb treatments. (Supplemental Figure 3).

3.5. Lipid peroxidation and estradiol concentration

TBARS levels at 72 hpf were increased in the 0.01 and 1 ppb Pb treatment groups (Figure 7A) but at 120 hpf was only increased in the 1 ppb Pb treatment group (Figure 7B). At 72 hpf, estradiol concentration was increased only in the 1 ppb Pb treatment group (Figure 7C) and then no changes were observed at 120 hpf (Figure 7D).

Figure 7 -.

Quantification of lipid peroxidation by TBARS at 72 hpf (A) and 120 hpf (B) and estradiol at 72 hpf (C) and 120 hpf (D). Data is represented as mean ± SEM of 12 biological replicates (each replicate consists of 3 technical replicates). Data was evaluated using a one-way ANOVA with post hoc LSD test. Statistical differences among treatment groups are represented by different lower-case letters on the bars (p<0.05).

3.6. Lipidomic analysis

3.6.1. Lipid profiling

At 72 hpf, the PLS-DA plot (Supplemental Figure 4A) based on targeted lipid profiling analysis from the Lipid Maps database indicated zebrafish exposed to different Pb concentrations were more intrinsically interacting and related than at 120 hpf (Supplemental Figure 4B). At 120 hpf, the 0.01 and 0.1 ppb Pb treatment groups were more related to the negative control group to a lesser degree than the 1 ppb Pb treatment group. These results indicate changes in these two treatment groups are more dissimilar than in the 1 ppb Pb treatment group.

At 72 hpf, out of a total of 631 lipids detected, 79 lipids were dysregulated (Supplemental Table 1). For those significantly changed lipids, fold change was calculated in comparison to the negative control treatment group. Of those 79 lipids, 60 lipids were altered in the 0.01 ppb Pb treatment group, 65 lipids were altered in the 0.1 ppb Pb treatment group, and 68 lipids were altered in the 1 ppb Pb treatment group. A total of 39 lipids were dysregulated in all Pb concentrations. For those lipids, the Log₂ fold change was calculated to express the magnitude of change in each individual feature. From this assessment, 35 lipids were identified with a Log₂ fold change greater than 0.5 in at least one Pb treatment group (Figure 8A). Among these 35 lipids, 18 were triacylglycerides (TG) with similar upregulation observed, primarily in the 0.01 and 1 ppb Pb treatment groups. In addition, a single cholesterol ester (CE) was also upregulated within these same two treatment groups. Alternatively, the main downregulated class of lipids was the sphingomyelin group (SM) and were primarily observed in the 0.1 and 1 ppb Pb treatment groups, while the group of phosphatidylcholines (PC) was decreased mainly within the 0.01 ppb and 0.1 ppb Pb treatment groups.

Figure 8 –

Select lipids (i.e., lipids with Log2(FC) greater than 0.5 in at least one Pb concentration and ANOVA with FDR<0.05 in all three Pb concentrations when compared to control by Tukey post hoc test) at 72 hpf (A) and 120 hpf (B).

At 120 hpf, out of the 630 lipids detected at this developmental stage, 123 lipids were altered (Supplemental Table 2). As in the 72 hpf analyses, fold change was calculated in comparison to the negative control group for those significantly changed. Of the 123 lipids, 90 lipids were changed in the 0.01 ppb Pb treatment group, 119 lipids in the 0.1 ppb Pb treatment group, and 72 lipids in the 1 ppb Pb treatment group. Moreover, 51 lipids were significantly altered in all three Pb concentrations when compared to the negative control treatment group at 120 hpf. Log₂ fold change was calculated for these 51 selected lipids to quantify the magnitude of alterations in each feature for the 120 hpf endpoint. From this analysis, 30 lipids had changes greater than 0.5 (Figure 8B). All 30 of these lipids were downregulated with 5 SM, 24 phosphatidylethanolamine (PE), and 1 PC. Overall at 120 hpf, the 0.01 and 1 ppb Pb treatment groups were more similarly aligned for observed alterations.

3.6.2. Lipid enrichment and pathway analysis

Pathway enrichment analysis was performed on select lipids at 72 and 120 hpf (i.e., 35 at 72 hpf and 30 at 120 hpf) (Table 1). As expected from the comparison above on the specific lipids altered in each Pb treatment group at 72 hpf, pathways were more similar among the 0.01 and 1 ppb treatment groups including lipid droplet, triacylglycerols, triradylglycerols, and headgroup with negative charge. Lipid storage was the only pathway similar among all three Pb treatment groups. Pathways altered in only the 0.1 ppb treatment group at 72 hpf included endosome/lysosome, ceramide phosphocholines, sphingolipids, and plasma membrane. Similarities among the enriched pathways identified at 120 hpf were also observed with glycerophosphoethanolamines, mitochondrion, and negative intrinsic (membrane) curvature present in all three Pb treatment groups. Membrane component was identified in the 0.01 and 0.1 ppb Pb treatment groups, while diradylglycerols was present in the 0.1 and 1 ppb Pb treatment groups.

Table 1 –

Pathway enrichment analysis for Pb treatment groups at 72 and 120 hpf.

Collection point	Concentration	Description of the enriched pathway	Significantly changed molecules	p-value
72 hpf	0.01 ppb	lipid storage	30	2.74E-09
		lipid droplet	30	2.74E-09
		triacylglycerols [GL0301]	28	8.21E-09
		triradylglycerols [GL03]	28	8.40E-09
		headgroup with neutral charge	33	4.48E-08
	0.1 ppb	endosome/lysosome	16	0.000381
		ceramide phosphocholines (sphingomyelins) [SP0301]	14	0.000616
		sphingolipids [SP]	14	0.001456
		plasma membrane	14	0.020533
		lipid storage	20	0.020533
	1 PPb	lipid storage	30	2.70E-09
		lipid droplet	30	2.70E-09
		triacylglycerols [GL0301]	28	8.60E-09
		triradylglycerols [GL03]	28	1.10E-08
		headgroup with neutral charge	32	5.70E-07
120 hpf	0.01 ppb	glycerophosphoethanolamines [GP02]	44	6.40E-13
		mitochondrion	44	6.40E-13
		negative intrinsic curvature	44	4.20E-08
		membrane component	89	3.20E-07
		headgroup with positive charge / zwitter-ion	80	1.80E-05
	0.1 ppb	negative intrinsic curvature	60	5.20E-12
		membrane component	118	6.40E-10
		glycerophosphoethanolamines [GP02]	44	7.90E-08
		mitochondrion	44	7.90E-08
		diradylglycerols [GL02]	16	0.00026
	1 ppb	negative intrinsic curvature	44	1.79E-10
		diradylglycerols [GL02]	16	1.18E-05
		diacylglycerols [GL0201]	13	2.99E-04
		glycerophosphoethanolamines [GP02]	28	3.14E-04
		mitochondrion	28	3.14E-04

4. Discussion

In this study, the effects of an embryonic exposure (1–72 hpf) to Pb concentrations reflecting those well below health action levels for humans (i.e., equivalent to 0.1 μg/dL and below) were evaluated using the developing zebrafish as a model for a translational study. Analyses focused on behavioral, biochemical, lipid, and physiological changes to determine if a low threshold of Pb exposure may be identified. Developmental Pb exposure at these low levels down to 0.01 ppb (equivalent to 0.001 μg/dL) impaired cognition and induced either anxiolytic- or anxiogenic-like behavior depending on concentration and neural circuits, without altering survival or hatching rates. Developmental Pb exposure at concentrations used in this study also disrupted estrogen levels, promoted lipid peroxidation, and delayed antioxidant enzyme activation, leading to oxidative stress and lipid damage. Notably, lipid alterations at 72 hpf differed from those at 120 hpf, underscoring the dynamic response to Pb toxicity. The persistent downregulation of sphingomyelin at both stages suggests potential long-term impacts on physiological systems with possible late-onset consequences.

Furthermore, in the 0.1 ppb Pb treatment group spontaneous movement (SM) and heart rate (HR) were increased, while in the 1 ppb Pb treatment group HR was reduced, possibly reflecting central effects of Pb. Pb is reported to disrupt neurotransmitter systems by altering synthesis, release, and uptake of dopamine [75–77], glutamate [68, 78–80], and acetylcholine [78, 81]. Such disruptions can enhance excitability and spontaneous movement. Alternatively, these effects may be linked to Pb-induced oxidative stress through reactive oxygen species (ROS) generation and weakened antioxidant defenses, damaging neurons and altering activity [53, 82, 83]. In the present study, Mn-SOD activity increased at 72 and 120 hpf in the 0.01 ppb Pb treatment group but was unchanged at higher concentrations, indicating delayed or impaired activation. Cu/Zn-SOD showed fluctuations among treatment groups and timepoints with increased activity in the 0.1 ppb Pb treatment group and decreased activity in the 1 ppb Pb treatment group at 72 hpf. In comparison, increased activity was detected in the 0.1 and 1 ppb Pb treatment groups at 120 hpf. These patterns suggest delayed SOD activation, likely caused by Pb disrupting metalloenzyme function. As metalloenzymes, SODs require Mn²⁺, Cu²⁺, or Zn²⁺ for catalysis [84]. Pb²⁺, with a similar ionic radius, can bind active sites without enabling superoxide detoxification [85]; thereby impairing antioxidant defenses. Previous work from our laboratory [86] demonstrated a 10 ppb Pb embryonic exposure led to an intracellular accumulation of ~300 ng/g (300 ppb) and a 100 ppb Pb embryonic exposure resulted in an intracellular accumulation of ~1000 ng/g (1000 ppb). Here our findings suggest that although 0.1–1 ppb Pb may not downregulate SOD in vitro, intracellular accumulation of these same concentrations in vivo can impair activity. As such, observations from higher in vitro concentrations in this study confirm Pb inhibits T-SOD activity by direct interaction.

CAT activity is expected to decline when SOD is inhibited, as both enzymes act in tandem. A similar pattern was observed at 72 hpf in the 1 ppb Pb treatment group where Cu/Zn SOD and CAT activity were most suppressed. This same pattern though was not observed in the 0.1 ppb Pb treatment group, where CAT activity was reduced at 72 hpf but Cu/Zn SOD activity was increased. Similarly, Cu/Zn-SOD and Mn-SOD were both increased in the 0.01 ppb Pb treatment group at 72 hpf but no changes occurred for CAT activity. Furthermore, although increased Cu/Zn-SOD (i.e., at 0.1 and 1 ppb Pb) and/or Mn-SOD (i.e., at 0.01 ppb Pb) was observed among all treatment concentrations at 120 hpf, there were no differences in CAT activity. As a metalloenzyme, CAT contains a heme group with iron (Fe) at its core. Previous assays indicate Pb can bind directly to CAT, interfering with the heme group or surrounding protein structure disrupting catalysis [87]. This direct binding interaction may explain the lack of activity. Overall, these biochemical disruptions, particularly in first-line antioxidant defenses (FLAD) such as Mn-SOD, Cu/Zn-SOD, and CAT, highlight a mechanism of neurotoxicity. Even at low concentrations, Pb inhibited these enzymes, which are critical for neural cells that are highly vulnerable to oxidative stress [80]. Inhibition of FLAD exacerbates oxidative damage in the brain, potentially contributing to neurodegenerative processes and cognitive deficits [88]. In developing organisms, such inhibition may increase oxidative stress, impair brain development, and alter behavior [89], which aligns with the observed impairments in cognition and anxiogenic-like behavior observed in this study.

The impairment of FLAD appears to drive the lipid peroxidation observed across treatment groups at 72 hpf and in the 1 ppb Pb treatment group at 120 hpf. These results suggest antioxidant defenses were not adequately activated due to Pb interaction, enabling oxidative damage [51]. In addition, although Pb is not typically classified as an endocrine disruptor, Pb can disrupt hormonal pathways [91]. In this study, estradiol levels were elevated at 72 hpf in the 1 ppb Pb treatment group, possibly due to oxidative stress that showed strong enzymatic disruption. Alternatively, Pb may disrupt the homeostasis of essential metals such as zinc and copper, cofactors in steroidogenesis, indirectly enhancing estrogen synthesis [92, 93]. Such endocrine alterations can contribute to cognitive deficits commonly associated with exposure to endocrine-disrupting chemicals [50].

Behavioral changes observed in this study reflected both oxidative stress and endocrine disruption. Although canonical shoaling typically emerges later in development [94], the behavior observed here reflects early-stage social responses, where larvae begin responding to visual cues of conspecifics. In the ST, larvae exposed to 1 ppb Pb during embryogenesis showed increased shoal cohesion, which later in life this behavior is associated with defense or fear-like responses in natural environments [95]. In contrast, in the AST, fish in the 0.1 ppb Pb treatment group showed reduced cognitive performance. Since aversive responses depend on detecting visual cues, and higher Pb levels are known to impair optic nerve development [97–99], AST performance within the 1 ppb Pb treatment group likely reflects visual dysfunction rather than cognition alone.

Given possible visual impairment, the OFT provides a complementary assessment of anxiety-like behavior and cognition [100, 101]. Fish exposed to 0.1 ppb Pb displayed hyperactivity, including more line crossings and time spent in the center of the arena, indicative of anxiolytic-like behavior or disinhibition [102]. Such behavior suggests Pb alters neurotransmitter systems (e.g., serotonergic and GABAergic) involved in anxiety regulation. Low-concentration Pb exposure may impair inhibitory circuits, promoting disinhibition and risk-taking behavior [103]. Similar patterns of hyperactivity and reduced anxiety have been reported in zebrafish at other developmental stages [104, 105], supporting this mechanism but require further study. Overall, in the 1 ppb Pb treatment group, cognitive impairment appears linked to anxiogenic-like behavior and visual dysfunction, promoting risk avoidance and preference for safe environments. Alternatively, in the 0.1 ppb Pb treatment group, exposure induced disinhibition and anxiolytic-like behavior, likely reflecting altered frontal lobe-related circuits and impaired decision-making.

To further examine decision-making and anxiety-related responses, the VMRT was applied. Fish exposed to 0.01 or 0.1 ppb Pb displayed anxiogenic-like behavior in both light phases, contrasting with the anxiolytic-like effects observed in the OFT. Two explanations may account for these differences. First, Pb-induced disinhibition in the 0.1 ppb Pb treatment group increased spontaneous activity in the OFT, but when a specific stress response was required in VMRT, impaired neural circuits failed to generate appropriate responses, resulting in hypoactivity. Second, Pb may differentially affect motor pathways controlling spontaneous [106] versus stimulus-evoked behaviors [67]. While the OFT reflects general locomotion and exploratory activity, VMRT requires stimulus-driven decision-making, which may be selectively impaired. Pb may induce hyperactivity by affecting basal ganglia circuits regulating motor control [108], while simultaneously impairing the optic tectum or retinal pathways essential for visual processing [109]. These findings emphasize the need for multiple behavioral assays and highlight the dynamic nature of neural responses to Pb neurotoxicity but require further study to identify specific neural circuit disruptions leading to the observed behavioral changes.

In addition to FLAD changes, enzymes involved in glutathione metabolism were also evaluated. GPX, GR, and GST represent key components of the second line of antioxidant defense (SLAD), responsible for detoxifying reactive species generated by oxidative stress. GPX activity increased in the 0.1 ppb Pb treatment group at 72 hpf and in the 1 ppb Pb treatment group at 120 hpf suggesting a compensatory response to oxidative stress, particularly since CAT was inhibited. Like CAT, GPX reduces H₂O₂ to water, protecting cells from oxidative damage [110]. The later increase at 120 hpf may also reflect a response to prolonged dysfunction and to the Mn-SOD activation observed at the same concentration. Although direct Pb–GPX interaction is uncommon, GR and GST are highly metal-sensitive. GR maintains redox balance by regenerating reduced glutathione (GSH) from oxidized glutathione (GSSG), an essential process for antioxidant function under stress [111]. Proper GR activation is therefore critical for sustaining SLAD function. Furthermore, a reported mechanism of Pb neurotoxicity involves activation of the Nrf2 pathway, which upregulates GST transcription via antioxidant response elements (ARE). Increased GST activity observed in the 1 ppb Pb treatment group at 72 hpf indicates Pb induced oxidative stress is sufficient to trigger this adaptive response. Pb may also interact with thiol groups in proteins, including cysteine residues in GST, further altering enzyme function [27]. The observed GST upregulation thus reflects significant oxidative stress and activation of detoxification pathways at this concentration.

Lipidomic profiling provided additional insight into behavioral outcomes. Lipid alterations differed between 72 and 120 hpf with some similarities and differences noted among the Pb treatment groups. Triglyceride (TG) and cholesterol ester (CE) upregulation at 72 hpf suggests disrupted energy metabolism and membrane composition. Since cholesterol is essential for neurodevelopment, altered CE levels may affect synaptic membrane integrity and neurotransmission. In contrast, the downregulation of sphingomyelin (SM) and phosphatidylcholine (PC) at both 72 and 120 hpf indicates persistent membrane destabilization. SM is crucial for cell signaling, while PC supports membrane fluidity and neurotransmitter release [112]. Persistent downregulation suggests impaired neuronal signaling and possible long-term neurodevelopmental consequences [113, 114]. At 120 hpf, phosphatidylethanolamine (PE) was also downregulated. PE is vital for membrane curvature and stress responses [115]. A reduction in PE may impair neuronal growth, membrane dynamics, and cognitive function [116]. Together, these lipid disruptions support a mechanistic link between Pb exposure, impaired neurotransmission, and behavioral alterations [117]. Behavioral tests further revealed this link with OFT results suggesting reduced anxiety and impaired inhibitory control, while the VMRT indicated increased anxiety and impaired decision-making. This mismatch may reflect disrupted integration of sensory inputs due to lipid-mediated effects on membrane stability (especially SM, PC, and PE) [117]. Altered excitatory–inhibitory neurotransmission, particularly involving glutamatergic signaling, could also further explain these contradictions [80]. Dysregulation of SM, PC, and PE across both developmental stages points to structural and functional disturbances in neuronal membranes. Such changes likely impair neurotransmitter release, receptor function, and neuronal plasticity, weakening decision-making circuits. The divergent behavioral outcomes (i.e., anxiogenic responses in VMRT and anxiolytic-like behavior in OFT) may represent maladaptive regulation of stress responses. Given Pb’s established effects on calcium signaling and synaptic function, lipid alterations may exacerbate these neurotoxic mechanisms.

Several endpoints in this study displayed non-monotonic concentration-response patterns, with low and intermediate Pb concentrations producing distinct effects that were not always proportionally related. Such non-linear responses are consistent with previous reports of low-concentration/low-dose environmental toxicants and likely reflect complex homeostatic and compensatory mechanisms operating during early development of zebrafish [118, 119]. For instance, alterations in lipid metabolism, such as TG and CE upregulation at 72 hpf, may serve as protective responses to oxidative stress at certain concentrations but stabilize or reverse at higher exposures [120]. Similarly, behavioral and biochemical endpoints may exhibit threshold or hormetic effects depending on developmental stage and pathway-specific sensitivity. By explicitly acknowledging these non-monotonic patterns, we highlight that low-dose Pb toxicity cannot be fully captured by simple linear or threshold-based models. Integrating these observations across lipidomic, oxidative, and behavioral outcomes provides a more nuanced understanding of early-life Pb toxicity and underscores the importance of evaluating multiple endpoints to accurately assess environmental risk.

In addition, one of the most affected pathways was sphingomyelin biosynthesis, which is essential for signal transduction, cell recognition [121], and myelin sheath formation [122]. Myelin deficiency is linked to several neurodegenerative diseases, with oxidative stress considered a contributing factor [123–125]. This study provides evidence that low-concentration Pb exposure can downregulate sphingolipid metabolism, potentially predisposing organisms to myelin-related disorders. Both analyses completed at 72 and 120 hpf showed downregulated lipids in this pathway, indicating lasting dysregulation. Altered sphingomyelin levels may compromise myelination and neuronal signaling, ultimately contributing to cognitive impairments [126]. Further studies are needed to determine whether these effects persist across the lifespan, as they may represent a critical mechanism underlying Pb-induced demyelination and associated neurological diseases.

Additionally, this study highlights a stage-specific effects of low-dose Pb exposure on zebrafish lipid metabolism, reflecting the transition from yolk-dependent to independent energy sources. At 72 hpf, embryos rely largely on maternally supplied lipids from the yolk, which support growth, organogenesis, and early neuronal development [120]. Despite this maternal buffering, significant upregulation of triglycerides (TG) and cholesterol esters (CE) alongside downregulation of phosphatidylcholine (PC) and sphingomyelin (SM) were observed, suggesting that even low-dose Pb can disrupt lipid storage and membrane composition during early development. The increased TG and CE may serve as a protective response to oxidative stress, providing energy reserves or acting as sinks for lipid peroxidation products, whereas the reduction in PC and SM indicates potential alterations in membrane integrity and signaling, which are critical for proper neuronal function. By 120 hpf, when yolk-derived lipids are largely depleted and larvae rely on their own metabolism, persistent downregulation of PC, SM, and phosphatidylethanolamine (PE) was detected, while TG and CE levels returned closer to control values. This shift suggests Pb exposure disrupts intrinsic lipid homeostasis and membrane composition, which may, in turn, influence neuronal activity and behavioral outcomes. For example, altered membrane lipid composition could affect synaptic signaling and membrane excitability, contributing to the subtle anxiety-like and locomotor changes observed in our behavioral assays [117]). Importantly, these findings demonstrate stage-specific lipidomic shifts at lower Pb exposure concentrations that are largely underexplored. By linking lipidomic alterations with oxidative stress, membrane remodeling, and behavioral endpoints, this study provides mechanistic insight into how early-life exposure to lower concentrations of Pb may subtly perturb neurodevelopment and energy metabolism. These results underscore the need to consider developmental stage and maternal nutrient contributions when evaluating neurotoxic effects.

5. Conclusions

This study demonstrates an embryonic Pb exposure, even at concentrations as low as 0.01 ppb, induces cognitive impairments, anxiety-related behaviors, and profound biochemical alterations. To our knowledge, this is the first work to investigate such low levels of Pb, providing a critical step toward defining a toxicological threshold for early developmental exposure. Pb disrupted estrogen balance, delayed antioxidant enzyme activation, and promoted oxidative stress and lipid damage. These effects were reflected in altered spontaneous movement, heart rate, and enzyme activities, with GST upregulation suggesting an adaptive but insufficient response to oxidative burden. Lipidomic profiling revealed persistent disruptions. Most notably, the downregulation of sphingomyelin, phosphatidylcholine, and phosphatidylethanolamine which can compromise membrane integrity, neurotransmission, and neurodevelopment. These lipid alterations likely underpin the observed behavioral deficits and highlight lipid metabolism as a central target of Pb neurotoxicity. Overall, these findings underscore the extreme vulnerability of developing organisms to Pb exposure, even at concentrations far below current health safety advisories, and establish a critical foundation for understanding Pb-induced neurodevelopmental risks.

Environmental Implications

Developmental exposure to the heavy metal lead is a global challenge in public health. Significant questions remain regarding a lower threshold of exposure for developmental neurotoxicity. In this study zebrafish were applied as an integrative vertebrate model to address developmental toxicity at exposure concentrations of 0.01, 0.1, or 1 ppb (μg/L), which is substantially lower than most published studies to address the lower threshold question. Developmental outcomes, neurobehavior, biochemical assays, and lipidomic profiling identified alterations even at the lowest test concentration of 0.01 ppb including downregulation of sphingomyelin, phosphatidylcholine, and phosphatidylethanolamine which can compromise membrane integrity, neurotransmission, and neurodevelopment.

Highlights.

• Lipid disruption underpins neurobehavioral deficits from Pb exposure

• Pb disrupts estrogen balance, antioxidant defenses, and oxidative stress

• GST upregulation reflects adaptive yet insufficient detox response

• Lipidomics shows persistent SM, PC, and PE downregulation

• Endosome pathways altered at 72 hpf and shifts to mitochondrial effects at 120 hpf

Abbreviations:

ARE

Antioxidant Response Element

AST

Avesive Stimuli Test

CAT

Catalase

CE

Cholesteryl Ester

FLAD

First Line of Antioxidant Defense

GPX

Glutathione Peroxidase

GR

Glutathione Reductase

GSH

Reduced Glutathione

GSSG

Oxidized Glutathione (Glutathione Disulfide)

GST

Glutathione S-Transferase

H₂O₂

Hydrogen Peroxide

HC

Hatching Curve

hpf

Hours Post-Fertilization

HR

Heart Rate

OFT

Open Field Test

Pb

Lead

PC

Phosphatidylcholine

PE

Phosphatidylethanolamine

ppb

Parts Per Billion

SLAD

Second Line of Antioxidant Defense

SM

Sphingomyelin

SOD

Superoxide Dismutase

SC

Survival Curve

ST

Shoaling Test

TBARS

Thiobarbituric Acid Reactive Substances

TG

Triglyceride

VMRT

Visual Motor Response Test