BPA binds to both ERα and ERβ (Rubin, 2011; Rochester, 2013), mimicking endogenous estrogen, though BPA is reported to have greater affinity for ERβ (Ben-Jonathan and Steinmetz, 1998; Vandenberg et al., 2009). Overall, for both ERs, the affinity for BPA is weak compared to E₂, though BPA binding can induce transcription of estrogen responsive genes (Kishida et al., 2001; Richter et al., 2007; Chung et al., 2011; Kinch et al., 2015; Cano-Nicolau et al., 2016; Qiu et al., 2016). Beyond this classic genomic pathway, BPA can also bind GPER (Mustieles et al., 2015), initiating rapid non-genomic responses (Lim et al., 2002; Gorelick et al., 2008; Bouskine et al., 2009). The ability of BPA to interact with estrogen receptors is suggested to be causal for some adverse effects in humans (Almeida et al., 2018).

The aromatase enzyme binds testosterone and catalyzes the formation of E₂ (Simpson et al., 2002; Nuzzi and Caselgrandi, 2022). Aromatase is also able to synthesize E₁ from androstenedione, with the resulting E₁ being converted to E₂ (Luu-The, 2013). Thus, EDCs that bind aromatase disrupt E₂ synthesis. TBT is one of these compounds. As an aromatase inhibitor, TBT reversibly binds to the active site on aromatase, preventing substrate (testosterone) binding (Cheng and Yang, 2023). Consequently, E₂ levels are reduced and testosterone levels are increased after TBT exposure (McGinnis and Crivello, 2011) which causes masculinization in gastropods (Horiguchi, 2006) and zebrafish (McAllister and Kime, 2003; Santos et al., 2006; McGinnis and Crivello, 2011). TBT concentrations as low as 0.003–0.005 μg/L have deleterious consequences for marine and freshwater organisms (Al-shatri et al., 2015). Filter and sediment feeding organisms are especially at risk and TBT has been detected in both target and non-target organisms (Hoch, 2001).

The general structure and function of the vertebrate retina is highly conserved (Baden et al., 2020). Rod and cone photoreceptors have their cell bodies in the distal outer nuclear layer (ONL) and synapse onto second-order bipolar and horizontal cells in the outer plexiform layer (OPL). Bipolar cell somata are in the inner nuclear layer (INL) and these cells are presynaptic to amacrine and ganglion cells in the inner plexiform layer (IPL). Ganglion cell axons leave the eye, forming the optic nerve. Signaling from photoreceptors to bipolar cells to ganglion cells is direct and glutamatergic; horizontal and amacrine cells, with cell bodies in the INL, are local circuit neurons that provide inhibitory feedback to the photoreceptor-bipolar-ganglion cell pathway in the OPL and IPL, respectively. Primate retinas, including humans, have three cone types (L, Long wavelength sensitive or “red” cones | M, Mid-wavelength sensitive “green” cones | S, Short wavelength sensitive or “blue” cones) (Mustafi et al., 2009), non-primate mammals have two cone types (M and S) (Lindenau et al., 2019), and zebrafish have four cone types (L, M, S, and a UV sensitive cone) (Robinson et al., 1993). Multiple bipolar, horizontal, amacrine, and ganglion cell types are reported in all species, contributing to complex and diverse levels of processing within the retina (Baden et al., 2020).

While circulating E₂ can reach the retina through the blood (Chaychi et al., 2015), it is also locally synthesized (Nuzzi et al., 2018). In rats, aromatase is found in distal retinal neurons (photoreceptors, horizontal and bipolar cells) within the ONL, OPL, and INL (Cascio et al., 2015; Valero-Ochando et al., 2024). Goldfish retina contains aromatase-positive horizontal, bipolar, and amacrine cells, as well as aromatase-containing ganglion cell axons (Gelinas and Callard, 1993; Callard et al., 1995). Aromatase transcripts have been identified in zebrafish (Sawyer et al., 2006), along with esr1 and esr2b (Cohen et al., 2025). Inhibition of aromatase changes the thickness of the IPL, INL, and OPL in both zebrafish and rats (Salyer et al., 2001; Hamad et al., 2007).

E₂ exposure affects photoreceptor gene expression (Hao et al., 2013; Cohen et al., 2022), suggesting ER are present in that retinal layer. ERα, ERβ, and GPER are present in mouse retina, with GPER (Jiang et al., 2019; Li and Li, 2020a; Li and Li, 2020b; Pinon-Teal and Ogilvie, 2024) and ERβ expressed on ganglion cells (Rodriguez-Ramirez et al., 2025). In rat and human retinas, ERα is expressed in the OPL and in amacrine and ganglion cells in inner retina and ERβ is expressed in the IPL, with colocalization of these ER types observed on some amacrine and ganglion cell types (Cascio et al., 2015). GPER has been identified in the retinas of zebrafish (Liu et al., 2009; Jayasinghe and Volz, 2012; Shi et al., 2013) and goldfish (Mangiamele et al., 2017).

While estrogen is present in both males and females, the higher circulating E₂ levels in females are suggested to underlie sex-differences in retinal function. ERG recordings from female Sprague-Dawley rats (age 2–6.5 months) are larger in amplitude than age-matched males (Chaychi et al., 2015). Similarly, multifocal ERGs recorded from men and women <50 years of age identified shorter implicit times in women participants (Ozawa et al., 2014). Increased E₂ levels in female Tungara frogs during the reproductive season is associated with an increase in retinal sensitivity (Leslie et al., 2021). Estrogen is also neuroprotective in retina (Bondesson et al., 2015; Cascio et al., 2015). E₂ helps regulate the blood-retinal-barrier and protect photoreceptors from glutamate-induced damage (Nuzzi and Caselgrandi, 2022). In a mouse retinopathy of maturity model, activation of GPER on ganglion cells and astrocytes decreased endoplasmic reticulum stress response to hypoxia (Li and Li, 2020a) and decreased apoptosis (Li and Li, 2020b). Activation of GPER on mouse ganglion cells was also protective against NMDA-mediated neurotoxicity (Jiang et al., 2019); while activation of ERβ was protective after optic nerve crush (Rodriguez-Ramirez et al., 2025). In vitro, cultured mouse Muller glial cells were protected from oxidative stress by E₂ treatment (Tawarayama et al., 2024). Given this latter role of E₂, it is not surprising that age-related changes in estrogen levels are associated with neurodegenerative retinal diseases in humans (Gupta et al., 2005; Nuzzi et al., 2018) and rodents (Rowe et al., 2025) or that estrogen modulation as a breast cancer treatment has been associated with a variety of retinal complications (Eisner et al., 2008; Eisner and Luoh, 2011; Moschos et al., 2012).

Developmental exposure to both BPA and TBT cause immediate effects on neurogenesis, with many of the studies performed in zebrafish. Early BPA exposure reduces zebrafish eye diameter (Crowley-Perry et al., 2021; Volz et al., 2024) and affects retinal layer thickness (Volz et al., 2024). Zebrafish larvae exposed to BPA showed changes to red (L) and UV cones within the retinal mosaic (Qiu et al., 2025). Chronic 8 days BPA exposure thinned the IPL and impaired responses to red and green (M cone) light (Volz et al., 2024). Similarly, adult male zebrafish exposed to BPA for 7 weeks showed altered color preference (Li et al., 2017). BPA increased ERα and ERβ expression, but decreased locomotor behavior in larvae (dos Santos et al., 2022). BPA analogs also alter visual function. Chronic BPS exposure until 6 dpf reduced thickness of the ganglion cell layer (Gu et al., 2019) and BPS exposure for 120 days disorganized the ONL, thinned the IPL and GCL, and altered adult spectral sensitivities (Liu et al., 2018). BPS exposure from 2 to 5 dpf disrupts spacing and alters signaling of red and UV cones (Qiu et al., 2023); differences in optic nerve structure were noted in a separate study that chronically exposed larvae until 6 dpf (Gu et al., 2019). Exposure to TBBPA until 5 dpf decreased both eye size and optokinetic responses in zebrafish (Baumann et al., 2016).

TBT exposure during early life stages causes abnormal and delayed eye development (Hano et al., 2007), thinning of the cornea (Wang and Huang, 1999), and retina-specific defects (Fent and Meier, 1992; Wang and Huang, 1999; Wester et al., 2004; Dong et al., 2006).Chronic 5 days TBT exposure (0–120 hpf) increased apoptosis in zebrafish retina, with the greatest difference observed at 60 hpf when there was a 2× increase in macrophages (Dong et al., 2006). In tropical guppies, a 7 days exposure to TBT decreased retinal pigment epithelium, disorganized the photoreceptor layer, and caused vacuoles in the IPL (de Paulo et al., 2020). In minnows, embryonic-larval exposure to TBT similarly reduced the amount of pigment and disrupted retinal layering (Fent and Meier, 1992) and a 7 days exposure decreased retinomotor responses in tiger perch (Wang and Huang, 1999). TBT exposure increased oxidative stress in the eyes of medaka juveniles (Shi et al., 2021) and disrupted the blood-brain-barrier in rats (Mitra et al., 2013; Mitra et al., 2015). Eyes of neonatal ICR mice exposed to trimethyltin (TMT), a related organotin compound, until P14 had thinner retinal layers, increased apoptosis, increased overall glutamate levels, and decreased micro-ERG b-wave amplitude, consistent with neurotoxicity (Kim et al., 2023). TMT exposure during zebrafish development caused concentration-dependent effects that reduced vision-based behaviors, thinned retinal layers, formed pyknotic nuclei, and distorted the boundaries of the IPL and OPL (Kim et al., 2019). A 14-day exposure to Triphenyltin (TPT) altered expression of genes involved in polarity during retinal development in zebrafish (Li and Li, 2020a).

Most published reports of BPA and TBT induced effects typically occurred immediately after exposure, and exposure was often chronic, lasting multiple days. In contrast, our approach assessed whether a transient, 24 h developmental exposure to either compound was sufficient to cause long-term changes in adult retinal function and anatomy. We also sought to determine if these compounds have opposite sequelae in retina, given their contrasting effects on estrogen signaling.

Compound exposure occurred when larvae were either 3 dpf or 7 dpf because these ages represent key time points in both visual system and estrogen pathway development, as noted above. Exposure lasted for 24 h. The concentrations used for each compound are environmentally relevant and in agreement with previous studies (Fent and Meier, 1992; Kolpin et al., 2002; Schmidt et al., 2004; el Hassani et al., 2005; Schmidt et al., 2005; Dong et al., 2006; Oliver et al., 2011; Saili et al., 2012; Wang et al., 2013; Borges et al., 2014; Hayashi et al., 2015; Kinch et al., 2015; Weber et al., 2015; Cano-Nicolau et al., 2016; Santangeli et al., 2016). Our BPA concentrations were 0.228 μg/L (0.001 µM) and 22.8 μg/L (0.1 µM), with DMSO (0.0003%) as the vehicle control (Cohen et al., 2025); TBT concentrations were 0.04 μg/L (0.12 nM) and 0.4 μg/L (1.2 nM), with 0.1% ethanol as the vehicle control (Jensen et al., 2025). After the 24 h exposure, treated larvae were returned to system water and allowed to grow undisturbed until adulthood (≥3–4 months of age). In adults, we examined retinal anatomy (using H&E staining of retinal sections) and function [using electroretinograms (ERGs)] to identify if there were persistent outcomes resulting from transient developmental exposure.

To determine if developmental exposure to BPA or TBT altered retinal function, we used ERGs. The ERG response components include an initial negative a-wave (photoreceptor response) at light ON, followed by a positive b-wave (ON-bipolar cell response), and a positive d-wave (OFF-bipolar cell response) at the end of the light pulse (Perlman, 2007; Nelson and Singla, 2009). We decided to focus on ERGs as differences in b- and d-waves (reflecting ON and OFF bipolar cell responses) have been associated with retinal diseases in animal models (Constable et al., 2023; Aviles et al., 2025; Medina Arellano et al., 2025) and in the clinic (Cornish et al., 2021; Chiang et al., 2022; Constable et al., 2023).

ERGs were recorded from adult zebrafish that had been developmentally exposed to either BPA or TBT when they were 3 dpf or 7 dpf. ERGs were recorded as described in (Nelson and Singla, 2009; Cohen et al., 2025; Jensen et al., 2025). Briefly, eye cups, with the cornea and lens removed, were placed onto a piece of 0.45 µm black filter paper in the recording chamber and perfused (0.3 mL/min inflow; 4 mL/min outflow) with MEM solution that had been equilibrated with 95%O₂/5%CO₂. A tungsten recording electrode was placed into the eye cup to record responses to 300 ms flashes of white light presented individually at 7 different brightness levels (ND 3.0 – ND 6.0 in 0.5 ND increments) on an infrared background (RG780 filter). Each flash was presented 4 times for each brightness level, and the entire protocol was repeated 10 times for a total of 280 retinal responses per eye. Recordings were amplified using a DAM80 amplifier (WPI) and Digidata 1440A (Axon Instruments). Data was collected using pCLAMP software (Axon) and analyzed in Origin. For each recording, b-wave amplitudes were measured from the a-wave trough to the peak of the b-wave (50–200 ms after stimulus onset); d-wave amplitude was measured as the peak occurring within 25–300 ms after stimulus offset. Technical replicates were excluded from analysis if b-waves were negative or the response occurred outside the time intervals. Amplitudes and implicit times (= time to peak amplitude) were analyzed using SPSS or R software, and graphs were made in Excel. One-way ANOVAs were used to assess differences in b- and d-wave responses for each age and compound. Results were considered significant if the p-value was ≤0.05.

For both compounds, exposure to the lower concentration (0.228 μg/L BPA; 0.04 μg/L TBT) at 3 dpf had greater effects than the higher concentrations (Figure 1), revealing non-monotonic effects. Exposure to 0.04 μg/L TBT at 3 dpf increased b-wave response amplitude (Figure 1A, p < 0.001) and b-wave peak time (Figure 1B, p = 0.019) compared to vehicle controls and 0.4 μg/L TBT exposed tissue. However, TBT exposure at 3 dpf did not alter OFF-bipolar d-wave responses (Figures 1C,D; N’s for the different groups: vehicle = 6 | 0.04 μg/L TBT = 6 | 0.4 μg/L TBT = 6). Exposure to 0.228 μg/L BPA at this age increased ON-bipolar cell b-wave (Figure 1E; p < 0.001) and OFF-bipolar cell d-wave (Figure 1G; p = 0.003) response amplitudes; d-wave response time was also faster than vehicle controls (Figure 1H; p = 0.005; vehicle = 10 | 0.228 μg/L BPA = 6 | 22.8 μg/L BPA = 18). Exposure to 22.8 μg/L BPA at 3 dpf decreased b-wave amplitude (p < 0.001).

Exposure to 0.04 μg/L TBT at 7 dpf also affected adult ERG responses (Figure 2). This TBT concentration increased b-wave (Figure 2A; p ≤ 0.001) and d-wave response amplitudes (Figure 2C; p ≤ 0.001) and delayed d-wave peak time (Figure 2D; p ≤ 0.001; vehicle = 4 | 0.04 μg/L TBT = 7 | 0.4 μg/L TBT = 5). For all three variables, responses measured after 0.04 μg/L TBT exposure were significantly different from both vehicle controls and the 0.4 μg/L treatment group. In contrast to TBT, it was the higher BPA concentration (22.8 μg/L) that significantly altered ERG responses in adult retinas when exposure occurred at 7 dpf. ERG b-wave amplitude was increased (Figure 2E; p < 0.001) compared to the 0.228 μg/L treatment group, whereas d-wave response amplitude was decreased (Figure 2E; p < 0.001) compared to both 0.228 μg/L BPA and vehicle controls. Exposure to 22.8 μg/L BPA at 7 dpf also quickened (reduced) the response time of the d-wave peak compared to the 0.228 μg/L BPA treatment group (Figure 2H; p = 0.02; vehicle = 6 | 0.228 μg/L BPA = 13 | 22.8 μg/L BPA = 5). Time to b-wave peak was not affected by either BPA (Figure 2F) or TBT exposure at 7 dpf (Figure 2B).

To determine if the observed physiological differences reflected compound-induced changes in retinal anatomy, we measured the thicknesses of the different layers in H&E stained adult retinal sections (Figure 3). The only retinal layer found to be sensitive to compound exposure was the IPL, where axon terminals of ON- and OFF-bipolar cells are located. The anatomical change in retinal layer thickness for the IPL was different for the two compounds. BPA exposure at 3 dpf, but not 7 dpf, significantly decreased IPL thickness in adult retinas (Figure 3A; p = 0.042). In contrast, TBT exposure at 7 dpf significantly increased IPL thickness (Figure 3B; p = 0.002).

We recently reported (Cohen et al., 2025) changes in estrogen signaling components in adult retinas following exposure to BPA at either 3 dpf or 7 dpf. Retinal homogenates collected from adults exposed to 0.228 μg/L BPA at 3 dpf showed a significant increase in mRNA expression of aromatase (cyp19a1b), and a significant decrease in protein levels of ERβ, p-ERK and p-JNK (Cohen et al., 2025) indicating effects on genomic/nuclear signaling and the activation of either membrane bound ER or GPER (Prossnitz and Barton, 2014). In contrast, retinal homogenates from adult fish exposed when they were 7 dpf did not show differences in aromatase, esr1, or esr2a expression though ERβ protein levels were increased. Exposure to 0.228 μg/L BPA at 7 dpf decreased p-ERK levels, while exposure to 22.8 μg/L BPA increased p-JNK levels. The two BPA concentrations also differentially effected p-AKT levels, with increased p-AKT protein observed in retinas treated with 22.8 μg/L BPA but decreased protein levels observed in the 0.288 μg/L treatment group (Cohen et al., 2025). BPA-induced changes in AKT levels are important to note, as this pathway is associated with GPER activation and neuroprotection in mouse retina (Jiang et al., 2019).

We have not completed a similar molecular analysis of retinal homogenates from adults developmentally exposed to TBT, and no similar analysis is reported in the literature. However, we can anticipate TBT-induced effects on estrogen signaling components, as a transcriptomic analysis of whole zebrafish embryos exposed from 2 to 5 dpf revealed TBT exposure altered genes involved in development and immune and inflammatory responses (Martinez et al., 2020). Further, given that TBT is an aromatase inhibitor (Matthiessen and Gibbs, 1998; McAllister and Kime, 2003; McGinnis and Crivello, 2011) and that aromatase and ER are present in zebrafish retina (Menuet et al., 2002; Sawyer et al., 2006; Cohen et al., 2025), it is likely that estrogen levels/signaling will be reduced by early life TBT exposure, as reported in juvenile Atlantic salmon (Lyssimachou et al., 2006).

Exposure to BPA and TBT, if only transiently, likely had a direct effect on retinal estrogen signaling. All estrogen receptor types are present in vertebrate retinas (Cascio et al., 2015) and aromatase is found in rat INL and OPL (Valero-Ochando et al., 2024) where bipolar cells and their dendrites are located. Importantly, estrogen signaling components are present in zebrafish retina at our exposure ages, including nuclear ER (Bardet et al., 2002; Lassiter et al., 2002; Tingaud-Sequeira et al., 2004), aromatase (Mouriec et al., 2009b; Cohen et al., 2025) and GPER (Shi et al., 2013). Developmental BPA exposure reduces eye size (Crowley-Perry et al., 2021; Volz et al., 2024) and affects brain development in zebrafish (Nakamura et al., 2006; Vandenberg et al., 2009; Zhou et al., 2009; Kim et al., 2011; Tse et al., 2013). Similarly, developmental TBT exposure delays eye development (Hano et al., 2007) and has retina-specific effects (Fent and Meier, 1992; Wang and Huang, 1999; Wester et al., 2004; Dong et al., 2006).

Our molecular analysis identified aromatase (cyp19a1b) and ER (esr1, esr2a) mRNA and/or protein (ERβ) in adult zebrafish retinal homogenates and found that developmental BPA exposure altered expression of some of these estrogen responsive genes (Cohen et al., 2025). We also observed differences in MAPK and AKT pathway activation (phosphorylation) suggesting BPA was also binding to GPER. In mammalian retina activation of GPER and/or ERβ is neuroprotective to retinal ganglion cells (Jiang et al., 2019; Rodriguez-Ramirez et al., 2025). BPA exposure (0.228 μg/L) at 3 dpf decreased ERβ, but increased p-ERK and p-JNK levels; while exposure to the same concentration at 7 dpf increased ERβ protein and decreased p-ERK, suggesting age-dependent differences ER/GPER activation by 0.228 μg/L BPA. BPA concentration-dependent effects were also observed for AKT pathway activation, with an increase in p-AKT (associated with pathway activation) found in retinas treated with 22.8 μg/L at 7 dpf and decreased levels in the 7 dpf 0.228 μg/L group (Cohen et al., 2025). GPER activation associated with PI3K/AKT pathway activation is neuroprotective (Jiang et al., 2019), suggesting a possible effect of BPA exposure. Together, this data suggests that a brief early life BPA exposure can cause latent effects on estrogen pathways in retina, with BPA concentration determining which specific ER/GPER is/are activated.

With regard to TBT, the presence of aromatase in zebrafish (Cohen et al., 2025) and mammalian (Cascio et al., 2015) retinas suggests a direct effect on estrogen synthesis. In his review of estrogenic signaling in mammalian retina, Cascio et al. (2015) indicates the retinal INL is the layer where most local E₂ is synthesized. Goldfish retina contains aromatase-positive bipolar, horizontal, and amacrine cells in the INL (Gelinas and Callard, 1993; Callard et al., 1995), suggesting E₂ synthesis also occurs in the INL in fish retina. Importantly, the INL is where bipolar cells are located. The changes in ERG b- and d-waves, which reflect bipolar cell responses, in reponse to TBT exposure suggest the INL may be targeted by compound exposure. Inhibition of aromatase changed the thickness of the IPL and INL in zebrafish and rats (Salyer et al., 2001; Orger and Baier, 2005) in support of this hypothesis. The IPL is where ERβ is located in mammals (Cascio et al., 2015) and our anatomical data indicates that the IPL is sensitive to TBT (and BPA) exposure, suggesting that TBT may selectively target the inner retina.

While BPA and TBT are classified as EDCs, both compounds have a variety of effects. TBT exposure induces oxidative stress (Zhang et al., 2017) and increases levels of glucose and creatine kinase in the blood (Li and Li, 2021). TBT exposure increased permeability of the blood brain barrier, lipid peroxidation, glial activation and autophagy in rats (Mitra et al., 2013; Mitra et al., 2015). TBT exposure also causes apoptosis in isolated mammalian cells (Correia et al., 2025) and disrupts thyroid signaling in mice (Sharan et al., 2014), rats (Rodrigues-Pereira et al., 2022) and zebrafish (Li and Li, 2021). BPA can binding to thyroid receptors (Almeida et al., 2018; dos Santos et al., 2022) and to thyroid hormone transport proteins such as transthyretin, displacing thyroxine (T4) (Lim et al., 2002; Fujimoto et al., 2004; Brent, 2023). BPA working through thyroid hormone receptor β (thrb), in particular, is required for proper development of cone photoreceptors in zebrafish (Suzuki et al., 2013; Qiu et al., 2025) and mice (Roberts et al., 2006).

Other reports indicate BPA and TBT exposure alter levels of the neurotransmitters glutamate, dopamine, and GABA (Aoshima et al., 2001; Tsunoda et al., 2006; Choi et al., 2007; Almeida et al., 2018; Liu et al., 2020; Tu et al., 2020; Hyun and Ka, 2024). BPA and TBT impact these neurotransmitter systems by either blocking compound synthesis or binding to receptors. In Xenopus oocytes (Aoshima et al., 2001) and dissociated rat CA3 neurons (Choi et al., 2007), BPA had concentration dependent effects on GABA_A receptor mediated responses. In rodent brain, BPA decreased TH-immunoreactivity (Ishido et al., 2007). GABA and dopamine synthesis decreased, and glutamate synthesis increased, in zebrafish larvae exposed to BPA from 5 to 8 dpf (Kim et al., 2020). Similarly, adult male zebrafish exposed to TBT had decreased expression of genes involved in dopamine, GABA and serotonin synthesis (Tu et al., 2020). Developmental TBT exposure blocked ligand (glutamate) binding to NMDA receptors in rodent brain membrane preparations (Konno et al., 2005).

As in the rest of the brain, retinal signaling depends on the release of neurotransmitters. Photoreceptor synapses onto bipolar and horizontal cells are glutamatergic, with glutamate release decreasing in response to light stimulation–the basis for ON and OFF bipolar cell responses (i.e., ERG b- and d-waves, respectively). Zebrafish retinal bipolar cells also express GABA and dopamine receptors (Connaughton, 2011). In zebrafish retina, GABA is localized to horizontal cells in outer retina and amacrine cells in inner retina (Connaughton et al., 1999; Yazulla and Studholme, 2001; Marc and Cameron, 2002). Dopamine is present in retinal amacrine and interplexiform cells (Connaughton et al., 1999; Yazulla and Studholme, 2001; Marc and Cameron, 2002). Amacrine and interplexiform processes are found in the IPL, the retinal layer found to be sensitive to BPA and TBT exposure. Thus, BPA/TBT exposure could be impacting bipolar cell responses by changing modulatory (GABA, dopamine) inputs or altering photoreceptor glutamate release.

Dopamine is synthesized in a 3-step sequence from the amino acid phenylalanine, with tyrosine hydroxylase (TH) the rate limiting enzyme in the pathway (Daubner et al., 2011). GABA is synthesized from glutamate via the enzyme glutamic acid decarboxylase (GAD) (Zhang et al., 2024). We observed changes in the protein levels of both enzymes in adult retinal homogenates that were exposed to 0.228 μg/L BPA exposure at 3 dpf. TH protein levels were increased, while GAD levels were decreased (Cohen et al., 2025). These enzymes are the rate limiting enzymes in the synthesis of dopamine and GABA, respectively, suggesting a direct BPA-induced effect on neurotransmitter synthesis in exposed retinas. Exposure to the higher BPA concentration (22.8 μg/L) at 3 dpf increased GAD levels in the same homogenates, identifying concentration dependent effects.

In amphibian retina, blocking GABA_A receptors increased ERG b- and d-wave amplitudes (Hirasawa et al., 2021). We observed a thinner IPL and a decrease in GAD protein levels in adult retinal homogenates developmentally exposed to 0.228 μg/L BPA at 3 dpf (Cohen et al., 2025), suggesting a BPA-induced decrease in GABA synthesis, which may contribute to the enhanced b- and d-wave responses that were observed. Adult retinas treated with the higher concentration of BPA (22.8 μg/L) at 3 dpf displayed reduced b-wave amplitude but increased GAD protein levels, consistent with an increase in GABAergic inhibition. Both concentrations of BPA increased TH protein levels in retinal homogenates exposed at 3 dpf (Cohen et al., 2025), suggesting a BPA-induced increase in dopamine levels. Elevated dopamine levels reduce gap junctional coupling of (McMahon, 1994) and block GABA release from retinal horizontal cells (Yazulla and Kleinschmidt, 1982; Calaza et al., 2006), contributing to the observed ERG responses. TBT exposure similarly influences both dopamine and GABA. We observed an increase in IPL thickness due to TBT exposure at 7 dpf. This could suggest an increase in GABA and/or dopaminergic inner neurons and their synaptic connections or an increase in bipolar cell synaptic contacts. TBT is reported to increase dopamine levels in zebrafish (Liu et al., 2020) but decrease GABA levels (Tu et al., 2020). The increase in ERG response amplitudes we observed in retinas exposed to the 0.04 μg/L TBT treatment suggest a decrease in inhibition, consistent with these reports.

Alternatively, BPA/TBT could be targeting photoreceptor responses, causing a downstream effect on the postsynaptic bipolar cells which would alter b- and d-wave responses. Examining ERG a-wave response amplitudes and timing suggest this possibility (Cohen et al., 2025; Jensen et al., 2025). For example, exposure to 0.228 μg/L BPA at 3 dpf increased both b-wave and d-wave amplitudes. Photoreceptor a-wave amplitude was also increased in this BPA treatment group, suggesting a downstream effect. Similarly, exposure to 22.8 μg/L BPA at 7 dpf quickened both a-wave and d-wave implicit times. Exposure to the higher TBT concentration (0.4 μg/L) at 3 dpf similarly quickened b-wave and a-wave peak times. The lower TBT concentration (0.04 μg/L) at 7 dpf increased and delayed adult a-wave and d-wave responses. These similarities suggest that, for some of the outcomes measured, the change in bipolar cell responses may be indirect and downstream of a direct impact on photoreceptors. BPA exposure in zebrafish larvae altered red (L) and UV cone morphology and disrupted the retinal cone mosaic (Qiu et al., 2025), suggesting a direct effect of BPA on photoreceptors. Photoreceptor sensitivity to TBT is highly likely, given the localization of aromatase to the ONL in some vertebrate retinas (Valero-Ochando et al., 2024) and the identification of aromatase (cyp19a1b) in zebrafish retina (Cohen et al., 2025).ERG.

Taken together, the changes in ERG b- and d-wave responses in adult retinas from fish developmentally exposed to BPA or TBT could be due to EDC-induced changes in neurotransmitter release/levels within the retina.

Finally, both BPA and TBT are known to directly alter neuronal activity, suggesting that the ERG differences observed may reflect a direct effect of these EDCs on retinal bipolar cells. TBT application can increase calcium levels in isolated DRG neurons (Fross et al., 2021) and reduce Na⁺/K⁺ pump activity in brain (Zhang et al., 2008; Li et al., 2015). BPA inhibited L-, N-, and T-type Ca⁺² channels in isolated GH3 cells and reduced L-type Ca⁺² channel amplitude in cardiac myocytes (Deutschmann et al., 2013) and rat aortic smooth muscle (Feiteiro et al., 2018). These actions of BPA are due to the compound binding to and stabilizing calcium channels in the resting state (Deutschmann et al., 2013). BPA application increases the amplitude of BK potassium channels in AD 293 cells by binding to both intracellular and extracellular sites on the channel (Rottgen et al., 2014), and inhibits TTX-sensitive and TTX-insensitive Na⁺ channels in DRG neurons via PKA and PKC dependent pathways (Wang et al., 2011). Zebrafish bipolar cells express depolarization elicited calcium (T-type or L-type) and potassium (sustained I_K or transient I_A) channels (Connaughton and Maguire, 1998). Patch clamp analysis of adult zebrafish bipolar cells revealed BPA-induced effects (Cohen et al., 2025). Developmental BPA exposure reduced L-type Ca⁺² current amplitude regardless of exposure age; whereas outward rectifying K⁺ currents (I_K) were reduced when developmental BPA exposure occurred at 3 dpf but increased when exposure occurred at 7 dpf. These differences in voltage gated currents measured in tissue collected from adults at the later age were mostly observed in response to 22.8 μg/L BPA exposure, suggesting concentration-dependent effects, as reported in other preparations (Wang et al., 2011; Deutschmann et al., 2013; Feiteiro et al., 2018). In isolated pancreatic β-cells, concentration dependent effects of BPA on R-type Ca⁺² channels were reported, with exposure to a low (1 nM) dose of BPA decreasing current amplitude (Villar-Pazos et al., 2017). In this preparation, BPA did not directly interact with the Ca⁺² channel, but induced effects through differential activation of ERβ and ERα (Villar-Pazos et al., 2017).

Changes in voltage gated currents across a population of retinal cells would affect ERG responses. Increases in b-wave amplitude, for example, could reflect increased internal calcium levels, as reported to occur after TBT exposure (Fross et al., 2021). A reduced current through voltage-gated calcium channels coupled with a reduced outward rectifying K⁺ current would lower overall bipolar cell responses. Alternatively, a reduced calcium current and enhanced outward K⁺ current would reduce and delay neuronal responses. These latter options were observed in BPA treated adult retinas.