An acute dose of glyphosate alters novel object exploration and hippocampal cFos expression in a sex-dependent manner in wildtype mice

Abstract

Glyphosate (GLY) is the active ingredient in most herbicides, including off-the-shelf weed killers such as Roundup^®. GLY crosses the blood–brain barrier, increases oxidative stress and genotoxicity, and impacts reproduction, but the extent of its effects remains unclear. Previous research reports conflicting evidence on sex-specific susceptibility to GLY’s effects, and very few investigate the effects of a single, acute dose on learning, memory, and neuronal activation. In vitro studies have found GLY interferes with gene expression and is uniquely capable of inducing DNA double strand breaks (DSBs) compared to other herbicides. DSBs can induce expression of immediate early genes (IEGs), which are important for synaptic plasticity, learning, and memory. However, a clear connection between GLY, IEGs, and learning and memory has yet to be made. To explore this, we tested male and female wildtype mice in novel object recognition after they received an acute, oral dose of 0, 250, or 500 mg/kg of GLY and assessed hippocampal DSB and IEG levels. We hypothesized that a single dose of GLY would impair memory by disrupting IEG expression and would affect males more than females. We did not find robust evidence that GLY impaired memory, though females that received 500 mg/kg did not explore the novel object more than the familiar. Hippocampal DSBs were decreased following 500 mg/kg in both sexes, yet hippocampal IEG immunoreactivity was decreased in GLY-exposed males only, revealing a complex sex-dependent relationship. These data add to the literature that GLY is potentially detrimental, highlighting the need for further investigations.

Graphical Abstract

Introduction

Since its introduction in 1996, glyphosate (N-(phosphonomethyl) glycine) has become the most widely used active ingredient in commercial herbicides globally (Benbrook, 2016 Dec). Glyphosate (GLY) inhibits the enzyme enolpyruvylshiki-mate-3-phosphate synthase (EPSPS) in the shikimate pathway, reducing aromatic amino acids vital for plant growth (Anderson et al., 1990 Feb). While originally thought to be safe for human consumption and exposure, recent research highlights the detrimental and potentially fatal effects of GLY ingestion (Caballero et al., 2018 Dec 16; Glyphosate: where is it banned or restricted phys.org [Internet]., 2023; Bali et al., 2019 Mar). Recent studies have highlighted connections between GLY and breast cancer (Panis et al., 2024 Jun 18), reproductive dysfunction (Kishore et al., 2026 Jan; Ren et al., 2025 Aug 5), and inflammation in the brain (Shu et al., 2025 Aug). Notably, levels of GLY detected in human urine samples appears to be increasing over time, with the general population excreting 0.16 to 7.6 μg/L (Gillezeau et al., 2019 Dec). The Center for Disease Control found that more than 80% of people had detectable levels of GLY in their urine, which appears to come via food (Ospina et al., 2022 Dec).

Multiple countries have banned the use of products containing GLY due to potentially detrimental effects on human health (Glyphosate: where is it banned or restricted phys.org [Internet]., 2023; Panis et al., 2024 Jun 18). However, use continues to be high in the United States, with over 113 million kilograms used annually for agricultural purposes (Benbrook, 2016 Dec). Studies have shown that GLY and glyphosate-based herbicides (GBHs) have detrimental cognitive and behavioral effects in rodents and other animals (Ait Bali et al., 2017 Aug; Baier et al., 2017 Nov; Winstone et al., 2022 Dec). For example, both subchronic (6 weeks) and chronic (12 weeks) oral ingestion of GBH led to a decrease in recognition and retention memory, while chronic exposure impacted working memory in male mice (Ait Bali et al., 2017 Aug). Additionally, repeated oral GBH exposure resulted in hyperactivity of the amygdala and medial prefrontal cortex (mPFC) in mice, regions that are critical for learning, memory, and emotional regulation (Bali et al., 2019 Mar; Winstone et al., 2022 Dec; Hao et al., 2019 Sep). Inhalation of GLY has also been shown to affect brain functionality, altering neurotransmitter pathways critical for locomotion, anxiety, and memory (Luna et al., 2021 Jun). However, there are very few studies that explore the effects of a single, acute GLY exposure, and minimal investigation into potential differences between males and females.

The exact mechanisms for GLY’s adverse neurological effects are unknown. GLY is able to cross the blood–brain barrier in a dose-dependent fashion, altering its integrity (Winstone et al., 2022 Dec). This leads to genetic dysregulation, increased levels of inflammatory markers (e.g., TNFa), and reduced cell viability. GLY and GBH have been found to induce mitochondrial death (Hao et al., 2019 Sep), promote dendritic damage (Luna et al., 2021 Jun), and increase DNA double-strand breaks (DSBs) (Woźniak et al., 2018 Oct; Suárez-Larios et al., 2017). Synaptic plasticity, and thereby learning and memory, is dependent on the transient expression of a family of genes known as Immediate Early Genes (IEGs) (Tischmeyer and Grimm, 1999 Apr 1). Recent studies have shown that DSBs increase after a learning event—as measured by the marker, γH2Ax—and increase the expression of IEGs, suggesting that DSBs are a crucial part of learning and memory (Suberbielle et al., 2013 May; Madabhushi et al., 2015 Jun; Boutros et al., 2023 Feb 2; Weber Boutros et al., 2022 Jul 28). Alterations to DSB regulation result in learning impairments (Boutros et al., 2023 Feb 2). Because GLY increases DSBs (Woźniak et al., 2018 Oct; Suárez-Larios et al., 2017), the adverse cognitive and behavioral outcomes that come with GLY exposure may be related to abnormal IEG expression in areas crucial for learning and memory, such as the hippocampus.

Both DSBs and IEGs have very quick, transient peaks (Turinetto and Giachino, 2015 Mar 11; Bulthuis et al., 2025 Oct 9; Lamprecht and Dudai, 1996; Gaigé et al., 2013 Jan). DSBs have a faster rise and fall than IEGs: DSB markers peak roughly 10–60 min after induction (Kinner et al., 2008 Oct), while IEGs peak 30–60 min after cellular activation (Fowler et al., 2011 Nov; Herschman, 1991 Jun). To understand effects on DSB and/or IEG expression, then, it is necessary to test a single, acute, and closely timed stimulus. Thus far, the effects of a single exposure to GLY on IEG expression have not been reported. Exploring a single event of GLY exposure in vivo is needed to clarify the possible connection between GLY, IEGs, and/or DSBs, as experiments with chronic exposure are not suitable for assessing this relationship.

With increased human exposure and consumption, it is critical to better understand the effects of GLY in mammals. We hypothesized that an acute dose of GLY would impair hippocampus-dependent learning and memory and alter IEG expression in C57Bl/6J (WT) mice. Additionally, we predicted that we would observe sex differences in sensitivity to GLY.

Methods

Animals

A total of 81 male (n = 40) and female (n = 41) C57Bl/6J (WT) mice were obtained from Jackson Labs (Bar Harbor, ME). The Biomedical Research Vivarium at Boise State University received the mice at 5 weeks of age. The mice were housed n = 4 per cage by sex, and one cage of 5 females. Food and water were provided ad libitum; lights were on a 12:12 light/dark cycle. Testing was conducted from 12 to 15 weeks of age. Mice were weighed prior to treatment delivery on days one, six, and ten. All work followed the ARRIVE guidelines and was reviewed and approved prior to the start of experiments by the Boise State University IACUC. Our facility is AAALAC certified, and all methods were in line with the U.S. Animal Welfare Act.

Glyphosate administration

Animals were randomly assigned to one of three groups. Flavored gelatin cubes (“jellies”) were used to administer treatment in lieu of oral gavage (Krenik et al., 2022 Apr). Experimental groups received “jellies” (Krenik et al., 2022 Apr) with either 250 mg/kg GLY, or 500 mg/kg GLY (N-(Phosphonomethyl)glycine, Sigma Aldrich #P9556–5G), in line with previous research (Bali et al., 2019 Mar; Panis et al., 2024 Jun 18; Suárez-Larios et al., 2017); the control group received jellies with saline (n = 9–14 mice/sex/dose). These doses, while higher than typically seen in human populations, are well below the no-observable adverse effect level (NOAEL) for acute dosing (>1000 mg/kg) (Henderson, n.d; Williams et al., 2012 Jan). GLY was diluted with saline and pH adjusted to 7. Blue food coloring was added to the GLY and saline solutions to visually ensure the mice consumed the entire dose.

To increase the likelihood that mice would consume the jellies, they underwent habituation to the jellies (i.e., no GLY added). For 5 days, mice were placed in individual cages for one hour and given untreated jellies to acclimate, then returned to their home cages. Behavioral testing ran from days 6–10; jellies were administered 2 h before the start of testing, as this is the time that circulating GLY levels are highest (Brewster, 1991 Jul; Wester, 1991 May) and rising in the brain (Winstone et al., 2022 Dec). On days 6–8, mice were given untreated jellies; on day 9, mice were given saline- or GLY-jellies; on day 10, mice were given untreated jellies for consistency pre-behavior.

On the administration day (day 9), researchers carefully examined any remaining pieces of jellies to ensure that the blue-dyed section had been consumed. Two mice in the saline groups (1 male, 1 female) only ate 60% of the blue area. One male in the 250 mg/kg group never ate any of the jellies, and was excluded from analysis.

Researchers were blinded to experimental groups throughout the duration of experimentation and analysis.

Open field and novel object recognition

Open field (OF) and novel object recognition (NOR) were performed as previously described (Boutros et al., 2021 Aug 2; Boutros et al., 2021 Nov). In a dimly lit room separate from the housing area, the mice were placed in an opaque 40 cm × 40 cm × 40 cm box for 5 min over 3 consecutive days. Trials were video recorded from above. Boxes were cleaned with 70% ethanol between each animal.

Following 3 days of OF, two identical objects (blue cylinders) were placed in the box for day 1 (training) of NOR. Mice were allowed to explore for 15 min. Twenty-four hours later, one object was replaced with a new object (yellow triangular prism) and mice allowed to explore for another 15 min (testing). A graphic of this can be seen in Fig. 2F.

Fig. 2.

Performance in novel object recognition following an acute dose of GLY. A) Total exploration on Day 1 (2 of the same object) of NOR. Females explored more than males (p = 0.011). B) Total exploration on Day 2 (one distinct, novel object) of NOR. Females explored more than males (p < 0.001). C) Percent of time spent with the objects on day 2 in female mice. Saline and 250 mg/kg females showed a preference for the novel object (p = 0.0068 and p = 0. 0166, respectively); 500 mg/kg females did not (p = 0.1095). D) Percent of time spent with the objects on day 2 in male mice. All showed a preference (saline: p = 0.0266; 250 mg/kg: p = 0.0007; 500 mg/kg: p = 0.0421). E) Discrimination index for NOR day 2. There were no group differences detected. F) Graphical depiction of the NOR set up. Two blue cylinders were used on day 1; 24 h later, a yellow triangular prism replaced one of the cylinders. ns = non-significant, *p < 0.05, **p < 0.01, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

All videos from NOR and OF were scored using EZ-track (Pennington et al., 2019 Dec 27) to evaluate the distance traveled, velocity, and time spent in the center. Two lab members manually scored time spent exploring the objects on NOR day 1 and day 2 using Button Box (Boutros et al., 2021 Nov). Interrater reliability was high (intraclass correlation coefficient = 0.932). One video file for day 2 NOR (n = 4 females in the 500 mg/kg group) was corrupted, and thus unable to be analyzed. For consistency in our repeated measures ANOVA, we did not include the data from these mice on either day 1 or day 2 for analysis.

Tissue collection

One cohort of mice were euthanized by cervical dislocation immediately following completion of NOR day 1 (n = 4/sex/dose). This was done to assess the DSB repair marker, γH2Ax, which rapidly accumulates at the sight of a DSB (Turinetto and Giachino, 2015 Mar 11). Brains were removed, and the hippocampus dissected in cold saline, then flash-frozen in liquid nitrogen and stored at −80°C until use. These mice did not undergo NOR day 2 (testing).

Mice that underwent NOR day 2 were euthanized by a lethal dose of a ketamine-xylazine cocktail 45–60 min after completing novel object recognition training to capture peak cFos expression (Cullinan et al., 1995 Jan). Mice were perfused intracardially with 15 mL of ice-cold phosphate buffer saline (PBS) followed by 15 mL of 4% paraformaldehyde (PFA). Brains were removed and kept in 4% PFA overnight, then sunk in 30% sucrose (n = 8–9/sex/dose).

Mice were chosen randomly for these distinct tissue collection procedures.

Western blots

Hippocampal samples were prepared as previously described (Boutros et al., 2021 Aug 2). Briefly, samples were homogenized and sonicated in lysis buffer (1 M Tris-Cl, pH 7.5; 6 M NaCl; 10%SDS; 0.5 M EDTA; Triton-X 100; Phosphatase Inhibitor, Thermo Fisher, #PIA32957; Protease Inhibitor III, Thermo Scientific, #AAJ64283LQ). A BCA kit (Pierce, Thermo Scientific, #23225) was used to determine protein concentrations. 10 μg of each sample were loaded into 4–20% Tris-Glycine Gels (Novex, #XP04205BOX); blots were then transferred to nitrocellulose membranes and stained with REVERT^™ Total Protein Stain (Li-Cor Biosciences, Lincoln, NE). Total protein was captured using a Li-Cor Odyssey CLx. Membranes were then washed in TBS, blocked in Odyssey blocking buffer, incubated in primary antibody overnight at 4°C (γH2Ax, 1:1000: rabbit-anti-phospho-histone H2Ax, Cell Signaling, #9718), followed by incubation in a donkey anti-rabbit secondary (IR-800, 1:20,000) at room temperature for 2 h. Images were again captured with a Li-Cor Odyssey; densitometry analysis was done using ImageStudio^™ (Li-Cor). Bands of interest were normalized to total protein for analysis.

Full blots can be seen in Supplemental Fig. 2.

Immunohistochemistry and microscopy

Whole brain samples were embedded in OTC and sectioned coronally at 40 μm. For each animal, 4 dorsal hippocampal sections (AP −1.5 to −2.2) spaced apart by 240 μm were selected for immunohistochemistry, as the dorsal hippocampus has been shown to be critical for NOR (Boutros et al., 2022 Dec; Cinalli et al., 2020 Oct; Goulart et al., 2010 Jun). Sections were rinsed in PBS 3× 5 min, then blocked in 5% normal goat serum in 0.3% triton-X solution. Sections were incubated in primary antibody (rabbit anti-cFos, 1:1000, Cell Signaling #2250S) overnight. The following day, sections were rinsed in PBS 3× 5 min, then incubated in secondary antibody for 2 h (goat anti-rabbit, AlexaFluor594, 1:1000, ThermoFisher # A-11012). Sections were rinsed and wet mounted onto slides and cover-slipped with Molecular Probes^™ SlowFade^™ Diamond Antifade Mountant (ThermoFisher #S36972). Images were taken at 10x magnification using STELLARIS Confocal Microscope (Leica, Wetzlar, Germany) in the Boise State Biomolecular Research Core.

The number of cFos + cells were counted in the dentate gyrus (DG), CA1, and CA3 by researchers blinded to treatment groups using Fiji (ImageJ, NIH, Bethesda, MD). Counting was limited to the granule layers of each of these regions. A universal threshold was determined for increasing visibility of cFos + cells, then cells manually counted within the defined regions.

Statistics

Data were assessed for normality and homogeneity before proceeding with standard parametric tests. Behavior from the open field tests (distance traveled, velocity, and percent time in the center) and time spent with the objects were analyzed with repeated measures ANOVA with dose and sex as between-group variables. Effects of dose were assessed using Bonferonni’s post hoc testing. When statistical differences were found between males and females, we split sexes for subsequent analyses of effects of GLY dose.

To analyze hippocampus-dependent memory, we first calculated the percent time spent with the objects on day 1 and day 2 (% Time = Time with Object / Total Time Exploring). We then analyzed the percent time spent with the novel object on day 2 compared to chance levels (value of 50%) with a one-sample t-test (Akkerman et al., 2012 Jul) in order to determine which groups discriminated between the familiar and the novel object. To assess group differences in object recognition, we calculated the discrimination index (DI = Time with Novel – Time with Familiar / Total Time Exploring). DI was compared using a 2-way ANOVA with sex and dose as between-group variables; we also used a one-sample t-test to compare to chance levels (value of 0).

The number of cFos + cells was averaged across the sections and analyzed using a 2-way ANOVA with sex and dose as between-group variables in the DG, CA1, and CA3. Males and females were then split, and cFos + cells analyzed with a 1-way ANOVA followed by Dunnett’s post-hoc testing. One male that received 250 mg/kg of GLY was detected as an outlier and removed from this analysis.

Results

Sex affects movement and anxiety-like behavior in the open field

Jellies, behavior, nor GLY affected weight over the course of our experiment (Supplemental Fig. 1A,D). All mice except for one male in the 250 mg/kg group habituated rapidly to the jellies, eating most of the Jell-O square within 3 days (Supplemental Fig. 1B,E). On treatment day, all mice except for 1 male and 1 female in the saline group ate > 75% of the entire jelly. The treatment itself (visually apparent with blue dye) was entirely consumed by n = 65 (of 81) mice, >90% consumed by n = 13, >85% consumed by n = 1 (in the saline group). One male in the 250 mg/kg group did not eat it at all; this animal was excluded from subsequent analysis (Supplemental Fig. 1C,F).

Mice were habituated to the open field for three days prior to receiving an acute dose of GLY and novel object recognition. We confirmed there were no preexisting differences based on future dose in total distance moved (p = 0.830), velocity (p = 0.826), and percent time spent in the center (p = 0.507). Thus, dose was removed from the statistical model to compare sexes. Females moved significantly more (p = 0.002, F(1,79) = 10.708; Fig. 1B), were faster than males (p = 0.006, F (1,79) = 7.904; data not shown), and spent more time in the center of the open field (p < 0.001, F(1,79) = 16.001; Fig. 1C).

Fig. 1.

Motor behavior in the open field and novel object recognition tests. A) Experimental design. Mice were habituated to jellies for 5 days. On days 6–9, mice received jellies prior to testing the open field for 3 consecutive days (5 min each day). On day 10, mice were given a single dose of glyphosate (0, 250, or 500 mg/kg), then trained in novel object recognition. Some mice were euthanized immediately to collect brain tissue (n = 4/sex/dose). Twenty-four hours later, some mice were tested for object recognition memory (n = 9–10/sex/dose). B) Motor behavior in the open field. Females moved more than males (p = 0.002, n = 40–41/sex). C) Anxiety-like behavior in the open field. Females spent more time in the center than males (p < 0.001). D) Motor behavior of females in the novel object tests. No differences were detected (day 1: n = 13–14/dose; day 2: n = 8–9/dose; 1 video for the 500 mg/kg group, reducing the n by 4). E) Motor behavior of males in the novel object tests (day 1: n = 13/dose; day 2: n = 8/dose). The 500 mg/kg group moved less than the saline group after receiving GLY. *p < 0.05, **p < 0.01, ***p < 0.001. Day 1, n = 9–14/sex/dose; Day 2, n = 9–10/sex/dose.

All mice habituated to the open field, demonstrated by a decrease in distance moved (p < 0.001, F(1.812,143.118) = 272.966, Greenhouse-Geisser corrected), and percent time spent in the center (p < 0.001, F (1.812,143.118) = 272.966, Greenhouse-Geisser corrected).

An acute dose of GLY affects movement and impairs novel object recognition in females

On day 4, mice were fed a jelly containing 0, 250, or 500 mg/kg of GLY, then placed in the OF with two identical objects; 24 h later, they were placed in the same open field, with one familiar and one novel object. We first assessed movement. We found a significant dose by day interaction (p < 0.001, F(1,46) = 12.698), a significant main effect of dose (p = 0.002, F(2,46) = 7.152), and a trend towards a main effect of sex (p = 0.054; Fig. 1D-E). Post hoc testing with Bonferroni corrections indicated that mice receiving 500 mg/kg moved less than mice receiving saline (p < 0.001) and 250 mg/kg (p = 0.001). This was driven by males during NOR day 1: Bonferroni-correct multiple comparisons revealed that males receiving 500 mg/kg moved significantly less than males that received saline (p = 0.0315). This dose-dependent difference was not seen in females on NOR day 1, and no differences were found on NOR day 2 in either sex.

We assessed total time spent exploring the objects on NOR day 1 and 2 (Fig. 2A-B). We found a significant main effect of sex (p < 0.001, F(1,46) = 18.909) and a dose by sex interaction (p = 0.026, F(2,46) = 3.953). Thus, we split sex for analysis of dose effects. In females, there was a day by dose interaction (p = 0.024, F(2,21) = 4.480), but no effect of dose on total exploration (p = 0.313). Conversely, there was a main effect of dose in males (p = 0.019, F(2,25) = 4.658). Dunnett’s post hoc testing did not reveal differences compared to the saline animals. All animals met the minimum threshold of exploring objects for 5 s on both days.

We next analyzed preference for the objects. As expected, there was no preference for either object on day 1 when both objects were the same. On day 2, control males (p = 0.0266, t(9) = 2.647) and females (p = 0.0068, t(9) = 3.489), 250 mg/kg males (p = 0.0007, t(8) = 5.386) and females (p = 0.0166, t(8) = 3.017), and 500 mg/kg males (p = 0.0421, t(8) = 2.416) showed a preference for the novel object, suggesting intact hippocampus-dependent memory. However, females that received 500 mg/kg failed to show a preference (p = 0.1095) Fig. 2C-D). Analysis of the DI did not reveal any dose- or sex-based differences (Fig. 2E).

Hippocampal γH2Ax and cFos are impaired by the high GLY dose in males, but not females

We assessed DSB repair in the hippocampus with Western blots stained for γH2Ax. A 2-way ANOVA of γH2Ax normalized to total protein revealed a main effect of GLY dose (F(2,18) = 6.722, p = 0.0066, Fig. 3A). There was no effect of sex (p = 0.412) nor interaction (p = 0.405). Dunnett’s post hoc testing indicated that mice receiving 500 mg/kg of GLY had significantly less γH2Ax than saline mice (p = 0.025). A representative blot is shown in Fig. 3B.

Fig. 3.

γH2Ax levels and cFos immunoreactivity in the hippocampus. A) Hippocampal γH2Ax levels normalized to total protein. 500 mg/kg of GLY decreased γH2Ax compared to saline (p = 0.025). n = 3–4/sex/dose. B) Representative blot showing γH2Ax and total protein stains, imaged on an Odyssey LiCor. C) cFos + cells in the DG. GLY decreased cFos immunoreactivity in males (p < 0.001). D) cFos + cells in the CA1. There were no effects of GLY dose. E) cFos + cells in the CA3. GLY decreased cFos immunoreactivity in males (p = 0.003). F) Representative images of the hippocampus from male mice, imaged on a STELLARIS Confocal Microscope. Yellow arrows are examples of cFos + cells. n = 4–8/sex/dose *p < 0.05, **p < 0.01, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

We conducted immunohistochemistry to assess neuronal activation in the hippocampus, as GLY has been implicated in disrupting typical cell signaling. Like the behavioral results, we detected significant dose-by-sex interactions (p = 0.042); thus, we split males and females for analysis.

There were no dose differences detected in female mice in the DG (p = 0.945, Fig. 3C), CA1 (p = 0.676, Fig. 3D), or CA3 (p = 0.920, Fig. 3E). Conversely, we found a significant main effect of dose in males in the DG (p < 0.001, F(2,17) = 13.294); Dunnett’s post hoc comparison indicated that 500 mg/kg of GLY significantly reduced cFos compared to saline (p < 0.001; Fig. 3C). This same effect was found in the CA3 of males (p = 0.005, F(2,17) = 7.477), again driven by a reduction in the 500 mg/kg group vs. saline (p = 0.003; Fig. 3E). There was no main effect found in the CA1 of males (p = 0.732; Fig. 3D). Representative images from the hippocampi of males are shown in Fig. 3F.

Discussion

In this study, we assessed the impact of an acute dose of GLY in male and female WT mice on hippocampus-dependent learning and memory and cFos expression. Our findings indicate a sex-specific effect of an acute dose of GLY, though this differed between our behavioral and molecular measures. Females were more affected than males in NOR, with female mice receiving 500 mg/kg of GLY failing to show a preference for the new object. Conversely, cFos immunoreactivity in the hippocampus was decreased by 500 mg/kg of GLY in male mice only. Hippocampal DSB repair was similarly decreased by our high dose in both sexes, shown by less γH2Ax. These data suggest that hippocampal function is affected by GLY, though the full picture is not clear with our dataset and highlights the need for more in-depth investigations that include biological sex as a variable. Nevertheless, our data contribute to the growing literature that GLY and GBH are detrimental to mammals.

Previous works have reported on the effects of GBH following subchronic (6 weeks) and chronic (12 weeks) exposure. One- to four-month old male Swiss mice exposed to 250 mg/kg and 500 mg/kg GBH for 6 and 12 weeks moved less, were slower, and spent less time in the center of the open field (Ait Bali et al., 2017 Aug). These exposures also reduced dopaminergic and serotonergic cells. A follow-up study also showed that subchronic and chronic exposure impaired hippocampus-dependent learning and memory and reduced acetylcholinesterase activity (Bali et al., 2019 Mar). Notably, an acute exposure to both doses of GBH impaired novel object recognition when percent time spent with the novel object was analyzed, similar to our results (Bali et al., 2019 Mar; Ait Bali et al., 2017 Aug). Assessment of neurotransmitter levels was not conducted in the acutely dosed groups in these previous experiments, however, and females were not assessed.

More recent investigations into chronic GLY exposure have found similar results. Sixteen weeks of GLY exposure (up to 200 mg/kg/day) increased anxiety-like behavior in C57Bl/6J mice, which was linked to a leaky blood–brain barrier, neuronal damage, and microglial activation (Shu et al., 2025 Aug). Notably, chronic 200 mg/kg is quite high, exceeding the no observable adverse effect levels (NOAEL) for chronic exposure in rodents (Henderson, n.d). Further research identified that 6 months of 1.75 and 175 mg/kg/day GLY exposure in C57Bl/6J mice had mild effects on anxiety-like behavior (Hsiao et al., 2024 Jan) yet more pronounced effects on gut metabolism. It is unclear which sex was used in this study, though, and there was no acute dose group.

Others have investigated the effects of an acute dose of GLY in 2.5-month-old female rats (Chávez-Reyes et al., 2025 Feb). Female Sprague-Dawley rats were given a single intraperitoneal injection of saline, 10 or 100 mg/kg of GLY, then tested in the Barnes Maze. Rats that received GLY did not reduce their number of mistakes, speed to the target, or orientation towards the target, indicating impaired spatial learning. Only 100 mg/kg of GLY decreased the time spent in the target zone during the probe trial. These data are reflective of our data: our high dose impaired novel object recognition memory in females, but not our low dose (Chávez-Reyes et al., 2025 Feb). No assessment of brain tissue was reported in this study, and males were not included.

There are some reports identifying sex-specific sensitivity to GLY/GBH, though male vs. female susceptibility conclusions are mixed. Exposure to 0.075% w/v GBH in drinking water from gestation to postnatal day 80 induced repetitive behaviors in both male and female Swiss mice, and impaired social interest in males (Del Castilo et al., 2022 May 17). These results indicate sex-specific susceptibility to GBH, with males slightly more affected. However, other studies suggest that females are more sensitive to lower doses (Gallegos et al., 2016 Mar; Gallegos et al., 2018 Oct). Male and female Wistar rats exposed to 0.2% or 0.4% GBH water in the perinatal period displayed delayed sensorimotor development later in adolescence. GBH-exposed males and females also moved less and spent less time in stressful areas of the open field and plus mazes. Notably, this occurred in females slightly earlier than males, appearing at postnatal day 45 in females but not until postnatal day 90 in males (Gallegos et al., 2016 Mar). A subsequent study identified that 3-month-old females were impaired in novel object recognition at both doses (0.2% and 0.4%) while males were impaired only at the high dose (Gallegos et al., 2018 Oct). Antioxidant and acetylcholinesterase activity were altered in both sexes by GBH exposure. These results suggest a slightly higher sensitivity to GBH in female rats, reflective of our behavioral results.

At the cellular level, our findings align with reports that GLY impacts neural signaling pathways essential for learning and memory. Previous work has shown that GLY and GBH induce downregulation of several IEGs in primary cortical neurons, such as Ier2, Dusp1, and Fos (Winstone et al., 2022 Dec). Conversely, immunohistochemical analysis of cFos and pCREB in the amygdala, prelimbic and infralimbic cortices of adult male mice revealed an increase following 6- and 12-week exposure to GBH (Ait bali Y, Kaikai N eddine, Ba-M’hamed S, Sassoè-Pognetto M, Giustetto M, Bennis M. Anxiety and Gene Expression Enhancement in mice Exposed to Glyphosate-based Herbicide Toxics, 2022). While conflicting, these reports point towards dysregulation of genes critical for synaptic plasticity, though neither explored potential sex-dependent effects. Other measures of synaptic plasticity have revealed that exposure to high levels of GLY or GBH during development (3% GLY in food from gestational day 6 to postnatal day 21) in rats reduced long-term potentiation (LTP) and the number of synapses in the hippocampus (Cresto et al., 2024 Mar). Males were more affected than females, demonstrating decreased LTP maintenance and altered synaptic density. These findings suggest that hippocampal function may be disrupted by GLY exposure via mechanisms that are both dose- and sex-specific.

It is also important to note that we only used one marker of cellular activation—it is entirely possible that other IEGs or measures of hippocampal function are disrupted in females or in both sexes. For example, the IEG Arc has distinct patterns of expression compared to cFos following fear conditioning (Bulthuis et al., 2025 Oct 9; Boutros et al., 2022 Dec). cFos expression does not always reflect behavioral changes, either (Shoji et al., 2023 Mar 29; Pattwell et al., 2012 Oct 2; Galvis-Alonso et al., 2010 Apr). It is not uncommon to identify behavioral changes, such as alterations in anxiety-like behavior, in the absence of cFos changes in brain regions that contribute to anxiety (Galvis-Alonso et al., 2010 Apr). It is also common to discover changes in cFos expression despite little-to-no behavioral changes (Shoji et al., 2023 Mar 29). Intriguingly, a marker of DSB repair, γH2Ax, decreased in the hippocampus following 500 mg/kg in both sexes. Several studies have shown correlations between γH2Ax and IEG expression (Madabhushi et al., 2015 Jun; Boutros et al., 2022 Dec), supported by our observation that hippocampal cFos and γH2Ax decrease in males receiving 500 mg/kg. While we expected that DSBs, and subsequently IEGs, would increase, this decrease still demonstrates a dysregulation of critical functions for proper learning and memory.

One explanation for the discrepancies in sex-specific sensitivity to GLY/GBH is in the range of ages tested. The behavioral experiments on only male mice occurred between postnatal day 30 and postnatal day 121 (~1 to 4 months) (Bali et al., 2019 Mar; Ait Bali et al., 2017 Aug); those on only female rats occurred at 2.5 months (Chávez-Reyes et al., 2025 Feb); those on male and female mice occurred at postnatal day 80 (~3 months) (Del Castilo et al., 2022 May 17); and those on male and female rats occurred at postnatal day 45 and 90 (~1.5 and 3 months) (Gallegos et al., 2016 Mar; Gallegos et al., 2018 Oct). Here, we tested mice at 3 months of age (13 weeks). Taken together, these studies suggest a slight female-specific impact of GLY/GBH. GBH exposure does appear to disrupt the endocrine system via the estrogen system, specifically affecting the female reproductive system (Ingaramo et al., 2020 Dec), and induces changes in cells that are reflective of endometrial cancer via estrogen pathways (Gastiazoro et al., 2020 Jun). It is entirely possible that there are distinct susceptibilities to GLY/GBH dependent on both sex and age. Further studies are warranted to more closely assess potential critical periods.

Another factor that differs between the previous and current studies is dosage. The NOAELs are defined as 1,546.5 mg/kg/day for males and 1,630.6 mg/kg/day for females (Henderson, n.d); average daily human exposure is estimated to be 0.0000114–0.0163 mg/kg/day (Gillezeau et al., 2019 Dec). For rodents, the chronic NOAEL has been identified as 100–400 mg/kg/day while the acute LD₅₀ is more than 10,000 mg/kg (Henderson, n.d). We used doses of 250 mg/kg and 500 mg/kg like others (Bali et al., 2019 Mar; Ait Bali et al., 2017 Aug; Ait bali Y, Kaikai N eddine, Ba-M’hamed S, Sassoè-Pognetto M, Giustetto M, Bennis M. Anxiety and Gene Expression Enhancement in mice Exposed to Glyphosate-based Herbicide Toxics, 2022)—higher than typical environmental exposures in humans, but still well below the NOAEL thresholds and lower than other studies (Gallegos et al., 2016 Mar; Gallegos et al., 2018 Oct; Cresto et al., 2024 Mar). Others have reported that these doses induce measurable increases in GLY and AMPA in the brain, such that an oral dose of 250 mg/kg leads to ~ 25 ng/mg and 500 mg/kg leads to ~ 40 ng/mg of GLY in brain samples (Winstone et al., 2022 Dec). These are metabolized quite quickly: AMPA’s elimination half-life measured in rat plasma was just over 15 h following oral exposure (Anadón et al., 2009 Oct).

Our experiments have some notable limitations, including sample size. Our a priori power analysis indicated that we had enough mice for a predicted effect size of 0.5, however technical malfunction resulted in loss of data that has reduced the power. We also were unable to measure GLY directly in the brains of our mice—when collecting brains, we prioritized collecting them for immunohistochemistry and Western blot. Despite this, we are confident that GLY crosses the blood–brain barrier following oral ingestion, as this has been shown in experiments utilizing radiolabeled GLY (Brewster, 1991 Jul; Wester, 1991 May), and, more recently (Winstone et al., 2022 Dec), in conjunction with increased inflammatory signals. Another limitation of our current study is that we did not analyze γH2Ax and cFos in the same animals. Due to the transient nature of both markers, we again had to choose when and what to prioritize during tissue collection. DSB markers peak ~ 10–60 min after induction, with γH2Ax being on the earlier side of this time window (Turinetto and Giachino, 2015 Mar 11; Kinner et al., 2008 Oct). IEGs peak ~ 30–60 min after cellular activation (Fowler et al., 2011 Nov; Herschman, 1991 Jun). We chose to collect tissue samples 1) immediately follow NOR Day 1 to capture γH2Ax, or 2) 45 min following NOR day 2 to capture cFos. While it is possible to analyze these markers in both sets of tissue, it would not be the most accurate representation of their peak levels.

An acute, high dose of GLY affected hippocampus-dependent memory in female mice, reduced a hippocampal DSB repair marker in both males and females, and decreased hippocampal neuronal activation in male mice. While historically GLY has been considered safe for mammals, the increasing evidence suggests the potential for adverse effects even in acute exposures. Our findings contribute to the growing body of literature that raises concerns about the effects of GLY exposure, indicating a need for further investigation into the potential long-term, sex-, age-, and dose-dependent effects of GLY and/or GBH exposure on brain and behavior.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.neuroscience.2026.05.031.