An Integrative Toxicological Assessment of the Herbicide Tebuthiuron: Elucidating Biochemical and Behavioral Responses in Developing Zebrafish

1. Introduction

Since 2008, Brazil has been the leading consumer of pesticides globally, followed by the United States of America. Over the last decade, Brazil’s pesticide consumption increased by 190% compared to previous years, a growth rate more than double the global market’s 93% increase during the same period [1]. In the 2010/2011 season, Brazilian pesticide consumption reached 936,000 tons, resulting in $8.5 billion in financial transactions [2]. In 2014, the use of other pesticides accounted for the most proportion of total pesticides (53.84%), followed by herbicides (25.10%), fungicides & bactericides (12.06%), insecticides (7.50%), plant growth regulators (1.24%), and others [3] Determining how these chemicals enter aquatic systems and their associated risks to aquatic species is therefore paramount for their safe application and use. Tebuthiuron (TBH), chemically named N-[5-(1,1-dimethylethyl)-1,3,4- thiadiazol-2-y1]-N,N’-dimethylurea (C₉H₁₆N₄OS), has a molecular weight of 228.314 g/mol and a log Kow of 1.79 [4,5]. TBH is an herbicide employed for brush control in pastures, rangelands, rights-of-way, and industrial sites. This phenylurea herbicide, commercially introduced in the mid-1970s [6], enters weeds via root and foliar absorption. It subsequently kills weeds by inhibiting electron transport, primarily at the reducing side of photosystem II during photosynthesis [7]. The carryover of biotoxins through residual TBH leaching can lead to contamination and biodiversity loss due to food chain bioaccumulation or environmental exposure [8].

TBH exhibits soil persistence ranging from 11 to 25 months, potentially extending up to 7.2 years, and has a DT50 (Degradation Half-Life) ranging from 16 to 20 days. With high water solubility (2,500 mg/L), TBH is considered relatively persistent and possesses a high leaching potential [9,10]. As such, it has been detected in freshwater environments at concentrations ranging from 10 to 291 ng/L [11]. For instance, TBH was detected in Pardo River water, São Paulo, Brazil reaching a maximum concentration of 1.02 μg/L [12]. It was also measured in the Mogi Guaçu River Basin in Brazil at an alarming maximum concentration of 6.4 μg/L, with a detection frequency of 100% [13].

While some studies, based on theoretical criteria and mathematical models, describe the susceptibility of groundwater to TBH contamination, research conducted in laboratory and field settings using soil columns and soil and water sample collections indicate lower leaching of this herbicide [14]. However, questions remain regarding its contamination potential and effects on aquatic organisms. In the USA, a maximum of 4.5% of the TBH applied to watersheds (2.24 kg a.i./ha) was determined to run off into adjacent water bodies. After watershed treatment, TBH concentration in receiving lotic waters exhibited a range from nondetectable to 83 μg/L [15]. TBH therefore is present in aquatic ecosystems, posing a risk for adverse exposures to aquatic species.

Within this context of environmental exposure, the zebrafish (Danio rerio) serves as a premier vertebrate model for studying sublethal developmental toxicity. Early life stages (e.g., up to 5 days post-fertilization) are particularly vulnerable to environmental toxicants, as interference with rapid morphological and neurological maturation during this window provides highly sensitive indicators of sublethal disruption [16].

While the acute toxicity and morphological impacts of various photosystem inhibitor herbicides are well-documented, the sublethal molecular mechanisms of TBH remain poorly understood. A broader look at phenylurea toxicity in aquatic vertebrates reveals that this class of herbicides profoundly disrupts endocrine systems—particularly the hypothalamus-pituitary-gonadal (HPG) and thyroid (HPT) axes—and alters extensive reproductive, oxidative, and metabolic networks [17]. In line with these sublethal perturbations, recent studies indicate that structurally related phenylurea herbicides, such as linuron, induce severe metabolic depression and neurotoxic hypoactivity in early staged zebrafish [18]. Moreover, initial behavioral screening of TBH have reported acetylcholinesterase (AChE) inhibition and late-stage larval hyperactivity [19]. This indicates that TBH may act through complex, non-linear neurodevelopmental and bioenergetic cascades that have yet to be fully elucidated.

Given Brazil’s leading and increasing pesticide consumption, combined with the persistence of TBH in international watersheds, it is reasoned that this chemical represents a significant environmental concern. To address ecotoxicological knowledge gaps, this study utilizes the zebrafish model to conduct an integrative assessment of TBH toxicity. Based on previous research, we hypothesized that zebrafish would show impaired behavioral responses and evidence of oxidative stress at high concentrations. By evaluating a suite of behavioral, biochemical, and transcriptome responses, this research aims to define the precise sublethal mechanisms and neurodevelopmental risks this herbicide poses to aquatic ecosystems.

2. Materials and Methods

2.1. Chemical Preparation

Tebuthiuron (TBH) stock solutions were prepared and added to embryo rearing media (ERM) containing the zebrafish embryos. Exposure solutions were prepared to yield environmentally relevant concentrations previously detected in the Pardo River and Mogi Guaçu River basins in Brazil (e.g., 1.02 μg/L and 6.40 μg/L). To ensure chemical stability and prevent the buildup of metabolic waste, fresh stocks were prepared daily to replace the corresponding volume during media renewal.

2.2. Husbandry and Egg Production of Zebrafish

Adult zebrafish (AB x Tübingen, Danio rerio) were maintained under standard environmental conditions. Details on breeding and water quality have been outlined previously [20,21]. Adult breeders are monitored for humane endpoints that include notable infections, impaired swimming behavior, or damaged tissues. Breeders are between 6 months and 1 year old. Staging of embryos and general husbandry protocols followed established guidelines. The Institutional Animal Care and Use Committee approved all experiments (IACUC202100000080, Approval date: April, 2021). Details on breeding and water quality).

2.3. TBH Exposure Design

Fertilized and normally developing embryos were selected at ~6 hours post-fertilization (hpf) and randomly assigned to experimental groups: an ERM negative control and a range of TBH concentrations. Experiments were conducted in 25 mL Pyrex glass beakers, each containing 15-20 embryos in 10 mL of the respective test solution (n = 4 to 6 beakers per treatment group). Beaker was the biological unit. Beakers were maintained in an incubator at 27 ± 1 °C. A 90% media renewal was performed daily for a continuous 5-day exposure period. Embryos and larvae were assessed daily using an EVOS™ FL Auto Imaging System to record mortality, calculate hatch rates, and document morphological deformities (spinal lordosis, caudal tail malformations, and yolk sac/pericardial edema).

Heart rate analysis was conducted at 5 days post-fertilization (dpf) using 10 randomly selected individuals per group. Heartbeats were counted under the EVOS™ FL Auto Imaging System for a period of 10 seconds per larva, and the resulting values were multiplied by six to determine the total beats per minute (bpm). Larvae were euthanized using buffered MS-222 (250 mg/L), pooled within a 1.7 ml tube, snapped frozen in liquid nitrogen, and placed at −80 freezer for further ROS or RNA extraction.

2.4. Visual Motor Response Test

Behavioral responses were assessed in 5 dpf larvae utilizing a DanioVision™ Observation Chamber (Noldus Information Technology, Leesburg, VA). Individual larvae were transferred into a 96-well plate containing 200 μL of ERM per well. Following a 2-hour acclimation period to minimize handling stress, a 50-minute visual motor response (VMR) protocol was initiated, consisting of alternating 10-minute dark and light phases. The assay began with a dark period to specifically evaluate photokinesis. Total distance moved (mm) was tracked continuously using EthoVision^®XT software. Trajectories demonstrating poor tracking fidelity were systematically excluded. Three experiments were conducted and the data were combined into one graph that represents all three runs. Poorly tracked animals were removed from the analysis based on manual inspection of the tracking. Standard operating procedures have been published previously for our VMR assay [22].

2.5. Reactive Oxygen Species

At 5 dpf, surviving larvae from each treatment group were pooled (n= 3-5 biological replicates per group), transferred to microcentrifuge tubes, and homogenized in 200 μL of ice-cold phosphate-buffered saline (PBS). Homogenates were centrifuged at 12,000 x g for 20 minutes at 4 °C. For ROS quantification, 20 μL of the supernatant was incubated with 2′,7′-Dichlorofluorescin Diacetate (H2-DCFDA) in a black fluorescence microplate for 30 minutes at 37 ± 1 °C in the dark. Samples were added to the microplate in triplicates. Fluorescence intensity was measured (excitation 485 nm; emission 520 nm). Final ROS levels were normalized to total protein content (μg/mL) using a BCA assay.

2.6. Acridine Orange Staining

Apoptosis was evaluated using Acridine Orange (AO) staining. Surviving 5 dpf larvae (n= 17-22 larvae/ group) were incubated in 2 μg/ml of AO (CAS 65-61-2, Sigma-Aldrich) in ERM for 30 minutes in the dark at 28 °C. Larvae were subsequently rinsed three times and imaged using a fluorescence microscope (EVOS TM Fl Auto Imaging System (ThermoFisher, Scientific, USA) with GFP filter. The intensity and localization of the fluorescent signal were quantified using the histogram tool of the Image J software to determine tissue-specific apoptosis.

2.7. Real-Time PCR Analysis

Total RNA was extracted from pooled 5 dpf larval samples using 750 μL TRIzol^® Reagent (Life Technologies, Carlsbad, CA, USA) as per the manufacturer’s protocol. Samples were assessed for RNA concentration using the Qubit^®Fluorometer (#Q33216, Invitrogen). Complementary DNA (cDNA) synthesis was done using ~750 ng of RNA and ABScript Neo RT Master mix with gDNA remover (Abclonal Technology, CAT. No. RK20433) in a final sample volume of 20 μL. Samples were placed into a T100™ Thermal Cycler (BioRad, USA). The cDNA was generated using the following steps: 37 °C for 2 min, 55 °C for 15 min, 85 °C for 5 min, and 4 °C for 5 min. Prior to real-time PCR, cDNA stocks were diluted 1:25 in RNAse-DNase free water. Quantitative real-time PCR was conducted using a fluorescence-based master mix. Real-time PCR methods followed that in our previous studies [20,21]. The relative expression levels of target genes were normalized to the geometric mean of established housekeeping genes (Ribosomal subunit 18, rps18; Ribosomal subunit L13a, and beta-actin, bactin). Final relative mRNA abundance among the experimental groups was calculated applying the 2^-ΔΔCt method. Primer sequences used in this study are provided in Supplemental Methods as Table S1 [23,24,25,26,27,28,29,30].

2.8. Statistical Analysis

Statistical analysis and graphing were conducted using GraphPad Prism V9.4 (La Jolla, CA, USA). A Log-rank (Mantel-Cox) test was employed to analyze survival and hatch curves. Groups being compared are those with different concentrations of TBH to ERM control (no chemical). ROS data were assessed for normality using a Shapiro-Wilk test; normally distributed data were analyzed via a One-Way ANOVA followed by Dunnett’s or Holm-Šídák’s multiple comparisons tests, whereas non-normally distributed data were log-transformed prior to analysis (log10). VMR data within discrete time bins were logarithmically transformed then analyzed utilizing a One-Way ANOVA test followed by post-hoc Dunnett’s multiple comparisons test. VMR data were normalized to the ERM group for each independent run. The ERM group was standardized to a value of 1 and each treatment was subsequently compared in a relative manner. Simple linear regression was conducted to evaluate gene expression variance across treatments. The significance of difference was determined at p < 0.05.

3. Results

3.1. Survival and Deformity

Some initial mortality was observed within the first 24 hours across all groups, including the ERM control. This is due to non-viable eggs initially, which stabilize after 24 hours. Exposure to TBH at concentrations up to 5000 μg/L did not induce significant decreases in survival compared to the control group (Log-rank test, p > 0.05). This suggests that TBH does not induce acute lethal toxicity in developing zebrafish at the tested environmentally relevant and sub-lethal concentrations.

Sporadic malformations, including instances of severe spinal lordosis, pericardial edema and yolk sac edema were documented. However, these deformities were rare (less than 5%) across groups. TBH exposure did not induce significant alterations in hatching rates (F _{(4, 34)} = 0.67, P= 0.62) (Figure 2) across the tested concentrations (p > 0.05). Heart rate was not affected by exposure to TBH up to 5000 µg/L (F _{(4, 46)} = 1.14, P= 0.35) (Figure 3).

Figure 1. Survival of zebrafish embryos and larvae (0–5 dpf) exposed to TBH. Four experiments were conducted, with approximately 1, 350 larval fish used. Mortality was monitored daily throughout the continuous 120-h exposure period. Data are presented as cumulative percent survival. No statistically significant differences were observed between the control (ERM) and treated groups (p > 0.05; Log-rank Mantel-Cox test).

Figure 1. Survival of zebrafish embryos and larvae (0–5 dpf) exposed to TBH. Four experiments were conducted, with approximately 1, 350 larval fish used. Mortality was monitored daily throughout the continuous 120-h exposure period. Data are presented as cumulative percent survival. No statistically significant differences were observed between the control (ERM) and treated groups (p > 0.05; Log-rank Mantel-Cox test).

Figure 2. Hatching success of zebrafish embryos exposed to TBH from 0 to 120 hpf. Data represents the cumulative percentage of successfully hatched embryos per treatment group. Statistical significance was evaluated using One Way ANOVA followed by a Dunnett’s test (p < 0.05). No significant delays in hatching were observed across the tested concentrations. ns= not significant.

Figure 2. Hatching success of zebrafish embryos exposed to TBH from 0 to 120 hpf. Data represents the cumulative percentage of successfully hatched embryos per treatment group. Statistical significance was evaluated using One Way ANOVA followed by a Dunnett’s test (p < 0.05). No significant delays in hatching were observed across the tested concentrations. ns= not significant.

Figure 3. Heart rate in zebrafish larvae (5 dpf) after exposure to TBH (µg/L). Heart rate is expressed as beats per minute (bpm). Data are presented as mean ± SD (vertical line and horizontal line respectively) (n = 10 fish per treatment). No significant differences were detected between the control and treated groups (p > 0.05; One-Way ANOVA). ns = not significant.

Figure 3. Heart rate in zebrafish larvae (5 dpf) after exposure to TBH (µg/L). Heart rate is expressed as beats per minute (bpm). Data are presented as mean ± SD (vertical line and horizontal line respectively) (n = 10 fish per treatment). No significant differences were detected between the control and treated groups (p > 0.05; One-Way ANOVA). ns = not significant.

3.2. Visual Motor Response Test

Behavioral responses are presented in Figure 4a-e. During the initial sudden transition to light (Light 1) (F _{(4, 133)} = 6.24, P<0.0001) which naturally induces a hypoactivity in healthy zebrafish, larvae exposed to 10 μg/L TBH exhibited hyperactivity (P= 0.03), in line with the Light 2 phase (F _{(4, 133)} = 12.4, P<0.0001), which showed a high hyperactive response in zebrafish exposed to both 10 μg/L (p < 0.0001) and 1000 μg/L (p < 0.005). The total distance moved in these groups was significantly increased compared to the ERM control (F _{(4, 133)} = 6.241; P = 0.0001). In addition, the concentration of 1000 μg/L induced hyperactivity (p < 0.01) during the third dark period (F _{(4, 135)} = 3.5, P=0.01). Taken together, there is evidence that TBH induces hyperactivity in zebrafish larvae, suggesting neurotoxicity or impaired motor movements.

3.3. Reactive Oxygen Species

A statistically significant reduction in basal ROS levels was observed in larvae (F _{(4, 13)} = 4.82, P= 0.013) exposed to 1000 μg/L TBH compared to the ERM control (Figure 5). The relative fluorescence units in the other exposure groups, including the highest concentration (5000 μg/L), were statistically indistinguishable from the control.

3.4. Acridine Orange Staining

Exposure to TBH did not induce significant alterations in whole-body apoptosis in 5 dpf zebrafish larvae. Quantification of acridine orange (AO) staining revealed no statistically significant differences in mean fluorescence intensity across any of the tested concentrations (0.1, 10, 1000, and 5000 μg/L),(F _{(4, 96)} = 0.58, P=0.68) relative to the ERM negative control (Figure 6).

3.5. Real-Time PCR Analysis

Transcriptional analysis revealed alterations in the expression of genes associated with oxidative stress, apoptosis, and neurodevelopment. Rather than exhibiting a simple concentration-dependent linear response, the expression data demonstrated effects on transcripts at specific exposure thresholds.

Regarding oxidative stress markers, cat expression was significantly down-regulated (F _{(4, 16)} = 6.24; P= 0.003) at both 10 μg/L (p < 0.05) and 5000 μg/L (p < 0.001) (Figure7a). Similarly, sod1 exhibited significant downregulation (F _{(4, 17)} = 4.492; P=0.012) at 10 μg/L (p<0.05) and 5000 μg/L (p<0.05) (Figure 7c), whereas sod2 showed a significant decrease (F _{(4, 16)} = 5.64; P= 0.005)) only at 5000 μg/L (p=0.03) (Figure 7d). A corresponding spike was observed in the intrinsic apoptotic pathway, with casp9 demonstrating highly significant up-regulation (F _{(4, 18)} = 8.569; P= 0.0005)), at 5000 μg/L (p < 0.001) ((Figure 8.c), and down-regulation (F _(4,17) = 3, P=0.046) of nrf at 5000 μg/L (p=0.025) (Figure 8.f). Markers for neurodevelopment and synaptic function also displayed anomalous expression peaks; both maptb and syn2a were significantly up-regulated (F _{(4, 18)} = 2.6; P= 0.07) (p < 0.05) (Figure 9.g) only at the highest concentration of 5000 μg/L, (F _{(4, 18)} = 4.3; P= 0.013) (Figure 9.j). Conversely, cholinergic signaling was disrupted at a different threshold, with ache expression being significantly suppressed (F _{(4, 17)} = 4.02; P= 0.018) at 5000 μg/L (P=0.02) (Figure 9a). Other evaluated transcripts: atp5i (F_(4,17)= 1.5, P=0.25), bcl2 (F_(4,16)= 2.2, P=0.12), cas3 (F_(4,16)= 3, P=0.051), elav3 (F_(4,16)= 3.2, P=0.04), gap43 (F_(4,16)= 1.2, P=0.34), gfap (F_(4,17)= 2.7, P=0.07), gpx (F_(4,18)= 1.7, P=0.2), keap (F_(4,16)= 1.3, P=0.3), manf (F_(4,16)= 1.1, P=0.4), mbp (F_(4,17)= 0.18, P=0.9), nqo1(F_(4,17)= 1.9, P=0.15),), p53(F_(4,17)= 0.98, P= 0.4), shha (F_(4,16)= 1.3, P= 0.3), and tubb3 (F_(4,16)= 1.6, P= 0.2) did not show significant differential expression across the evaluated concentrations.

4. Discussion

The toxicological profile established in this study indicates that the herbicide TBH acts primarily at physiological and biochemical levels, rather than as a structural teratogen. The lack of acute lethality, cardiotoxicity, and gross teratogenicity observed at concentrations up to 5000 μg/L strongly corroborates the baseline developmental observations previously reported for TBH in zebrafish embryos [19]. From a toxicological standpoint, these stable physiological milestones are highly informative; they suggest that subsequent behavioral and transcriptomic alterations are specific molecular responses to the herbicide, rather than secondary consequences of a systemic physiological collapse. Similar findings have been documented for another phenylurea herbicide, linuron, where early developmental stages and heart rates remained largely unaffected at concentrations up to 623 µg/L and 4982 µg/L respectively, reinforcing the premise that phenylurea-induced toxicity predominantly targets molecular pathways over structural integrity [18].

Despite the absence of morphological defects, TBH exposure elicited severe, non-linear behavioral alterations in 5 dpf larvae, characterized by a progressive dysregulation of the natural photomotor response. The Visual Motor Response (VMR) assay captures the integration of sensory processing and motor execution; because larval locomotion is highly dependent on oxidative respiration efficiency, xenobiotic-induced metabolic reductions often manifest as hypoactivity and a failure to habituate to photic transitions [21]. Emphasizing statistical anomalies over simple linear dose-responses, while the initial sudden transition to darkness (Dark 1) did not yield significant deviations from the control, a paradoxical and severe hyperactive response emerged during the subsequent photic phases. Sudden transitions to light naturally induce basal hypoactivity in healthy zebrafish; however, larvae exposed to 10 μg/L TBH exhibited pronounced hyperactivity during the Light 1 phase. This anomalous hyper-arousal persisted and expanded during the Light 2 phase, manifesting significantly in both the 10 and 1000 μg/L exposure groups. Furthermore, this systemic failure to habituate eventually cascaded into the dark phases, with the 1000 μg/L concentration inducing significant hyperactivity during the final Dark 3 period. While de Oliveira et al. [19] previously documented persistent hyperactivity specifically during dark periods in older (6 dpf) larvae exposed to TBH, the current assessment uncovers a developmentally earlier and distinct photomotor shift. defined by light-induced hyperactivity that eventually disrupts dark-phase rest. This trajectory suggests that TBH actively dysregulates the larvae’s ability to acclimate to dynamic lighting conditions. Notably, this hyperactive manifestation contrasts with the behavioral phenotype reported for the structurally related herbicide linuron, which induced profound hypoactivity under similar VMR paradigms [18]. This underscores that while phenylureas severely disrupt sensory-motor integration, their specific behavioral manifestations—hyper-arousal versus motor depression—can be highly distinct.

This hypothesis of early-stage metabolic exhaustion is biochemically supported by a critical anomaly observed in the reactive oxygen species (ROS) assay. Typically, xenobiotic exposure triggers a ROS generation burst that overwhelms endogenous antioxidant defenses [31]. However, TBH exposure did not induce an oxidative burst; instead, a significant depletion of basal ROS levels was observed specifically at 1000 μg/L. This non-monotonic depletion correlates with the pronounced hypoactivity observed in the exact same treatment group during the photic phase of the VMR assay. This phenomenon aligns with recent toxicological assessments of other phenylurea herbicides. For instance, both linuron and its primary metabolite (3,4-dichloroaniline) have been shown to induce profound mitochondrial dysfunction and redox modulation [32], while diuron disrupts mitochondrial energy metabolism and oxidative phosphorylation in early developmental stages [33,34]. Consequently, a suppression in mitochondrial bioenergetics would directly reduce the natural byproduct of basal ROS. While direct bioenergetic quantification is required to confirm this, such mitochondrial suppression would theoretically limit ATP availability, severely compromising the larvae’s capacity to maintain basal locomotion at 5 dpf. Interestingly, at the highest concentration (5000 μg/L), ROS levels were statistically indistinguishable from the control. This phenomenon suggests that at extreme toxicity thresholds, secondary stress pathways—such as endoplasmic reticulum (ER) stress or the onset of inflammatory responses—may generate compensatory ROS, hypothesized to mask the underlying mitochondrial depletion and further illustrating the non-monotonic nature of TBH toxicity.

The transcriptional profiling of 5 dpf zebrafish larvae exposed to TBH revealed highly specific, non-linear alterations, challenging the classical dose-dependent paradigm. Regarding oxidative stress, the significant downregulation of cat at 10 and 5000 μg/L, alongside the isolated suppression of sod1 (10 and 5000 μg/L) and sod2 (5000 μg/L), is indicative of a Non-Monotonic Dose-Response (NMDR). Rather than mounting a sustained, linear antioxidant defense, the suppression of these critical enzymes at specific thresholds suggests an early exhaustion or breakdown of the endogenous adaptive pathways. This hypothesis of systemic antioxidant collapse is strongly corroborated by the significant downregulation of nrf2a exclusively at the highest concentration (5000 μg/L). The nrf2a transcription factor serves as the master regulator of the cellular antioxidant response, responsible for driving the expression of downstream detoxifying enzymes like SOD and catalase [35]. Its suppression at peak exposure levels indicates that TBH directly impairs the upstream regulatory machinery necessary to combat oxidative damage. Similar non-monotonic transcriptional profiles and antioxidant fatigue have been extensively documented in zebrafish exposed to other widespread contaminants, such as the PSII inhibitor atrazine [36,37] and dimethyl phthalate [38]. This lack of a compensatory transcriptional response aligns with the profound basal metabolic depression observed in the ROS and VMR assays, suggesting that TBH dysregulates energy allocation, which likely precipitates the premature fatigue of the antioxidant defense system before a linear response can be fully mounted.

This state of physiological stress also triggered severe compensatory mechanisms within the developing nervous system. Consistent with de Oliveira et al. [19], who reported that TBH exposure did not significantly affect the mRNA levels of four key neurodevelopmental marker genes (elavl3, gfap, gap43, and shha), basal neurodevelopmental transcripts evaluated in the present study similarly remained unaltered. However, the significant upregulation of maptb and syn2a exclusively at the highest concentration (5000 μg/L) highlights an acute hypercompensatory response. The maptb gene encodes for tau proteins essential for axonal microtubule stabilization [39], while syn2a is crucial for anchoring presynaptic vesicles [40]. The sudden spike in the transcription of these specific structural markers suggests that the larval nervous system is actively attempting to preserve cellular connectivity and stabilize disrupted synaptic architectures in the face of impending xenobiotic neurotoxicity, a transcriptional defense mechanism similarly observed under developmental neurotoxicity induced by difenoconazole [41], trichlorfon [42], and flame retardants [43,44]. This molecular “state of alert” provides a fundamental basis for the chaotic hyperactive motor responses observed during the light phases of the behavioral assay. Furthermore, this neurological dysregulation is corroborated by the significant transcriptional suppression of ache at 5000 μg/L. The downregulation of this critical cholinergic transcript at 5 dpf provides compelling molecular support for the hypothesis that the late-stage AChE enzymatic inhibition observed by de Oliveira et al. [19] at concentrations above 100 μg/L may originate as a transcriptional failure during earlier developmental windows.

Perhaps the most counterintuitive finding of this assessment is the discrepancy between the intrinsic apoptotic pathway transcription and phenotypic cell death. Real-time PCR revealed a massive, highly significant upregulation of casp9 at 5000 μg/L however, in vivo whole-body apoptosis assays using acridine orange (AO) staining showed no significant phenotypic cell death. This transcriptional-phenotypic decoupling is mechanistically supported by the ontogeny of the zebrafish. First, sub-lethal and compartmentalized casp9 activity is a known physiological requirement for non-apoptotic neurodevelopmental processes, specifically axon pruning, synapse elimination, and glial differentiation [45,46]. Furthermore, the caspase family in zebrafish plays extensive roles beyond classical cell death during embryogenesis [47,48]. While maptb and syn2a are primarily neuro-specific markers, the use of whole-body homogenates in this study limits the ability to restrict these casp9 signals exclusively to the central nervous system. However, the concurrent upregulation of these synaptic and microtubule transcripts strongly implies a neuro-centric repair mechanism. Therefore, it is plausible that the larval brain is leveraging casp9 to cleanly amputate damaged axonal projections without inducing whole-cell death, which directly correlates with the simultaneous maptb and syn2a repair efforts. In addition, this discrepancy highlights a kinetic and analytical latency. The extreme sensitivity of qPCR captures early, transient shifts in mRNA before post-transcriptional checkpoints—such as Inhibitor of Apoptosis Proteins (IAPs)—can sequester the zymogens. Conversely, the AO vital dye exclusively intercalates into nucleic acids during the terminal stages of chromatin condensation [49]. Therefore, TBH effectively triggers the molecular precursors of the apoptotic cascade, but endogenous biological circuit breakers appear to prevent the widespread execution of cellular degradation by 5 dpf.

In summary, the integration of behavioral, biochemical, and molecular endpoints in this study provides a comprehensive look at TBH ecotoxicity that fundamentally challenges classical, linear dose-response paradigms. While traditional macro-structural assessments could erroneously classify this herbicide as low-risk due to the absence of gross teratogenicity, this multi-tiered approach uncovers a complex sequence of sublethal injuries. The data points toward an ontogenetic cascade starting with an early stage bioenergetic collapse that physically limits larval locomotion, transitioning into a state of severe neurodevelopmental stress and a fully activated, yet post-transcriptionally gated, apoptotic trigger at higher thresholds. These non-monotonic mechanisms highlight that evaluating pesticide safety based solely on lethality or visible malformations severely underestimates the silent, systemic physiological disruptions that phenylurea herbicides pose to non-target aquatic vertebrates.