Next Article in Journal
The Muscle Function Deficit Concept and Inflammaging
Previous Article in Journal
Safety and Efficacy of Autologous Bone-Marrow-Derived Mesenchymal Stem Cells for Treating Partial Rotator Cuff Tears: Observational Study with 4-Year Follow-Up
Previous Article in Special Issue
Magnetic Resonance Imaging and Cerebrospinal Fluid Biomarker Clustering Defines Biological Subtypes of Alzheimer’s Disease
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Biomedicines
Volume 14
Issue 2
10.3390/biomedicines14020381
Due to scheduled maintenance work on our servers, there may be short service disruptions on this website between 11:00 and 12:00 CEST on March 28th.
biomedicines-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Differential Metabolic Changes in Zebrafish Embryos Are Induced by Discontinued Citalopram Exposure

by
Gwendolyn Cooper
1,
Ryan J. North
2,
Dana Auganova
1,3,
Christa S. Merzdorf
2,* and
Brian Bothner
1,*
1
Department of Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717, USA
2
Department of Microbiology and Cell Biology, Montana State University, Bozeman, MT 59717, USA
3
National Center for Biotechnology, Astana 010000, Kazakhstan
*
Authors to whom correspondence should be addressed.
Biomedicines 2026, 14(2), 381; https://doi.org/10.3390/biomedicines14020381
Submission received: 15 January 2026 / Revised: 31 January 2026 / Accepted: 3 February 2026 / Published: 6 February 2026
(This article belongs to the Special Issue State-of-the-Art Molecular and Translational Medicine in USA)
Browse Figures
Review Reports Versions Notes

Abstract

Background/Objectives: Citalopram is a selective serotonin reuptake inhibitor that is prescribed to relieve anxiety and depression. Widespread use has led to the contamination of freshwater systems downstream of wastewater treatment facilitates. Few studies have investigated the impact of citalopram on early embryonic development in humans or other species, despite the prevalence of intentional or unintentional exposure. Danio rerio (zebrafish) is a model organism for investigating effects of environmental exposure to xenobiotics on developmental outcomes in vertebrates. Methods: In this study, we examined the metabolome of developing zebrafish embryos exposed to citalopram hydrobromide concentrations (0.03–250 ng/mL) spanning environmental to therapeutic doses during the first 24 h post-fertilization. Exposure was followed by 24 h exposure-free development before harvest at 48 h. Results: Gross morphology of the embryos was normal, although changes were observed in the heart rates of citalopram-exposed embryos. Untargeted metabolomic and multivariate analyses revealed significant, nonlinear changes in the metabolome in response to citalopram exposure. Arginine and proline metabolism was significantly altered, potentially reflecting changes in serotonin signaling, nitric oxide metabolism, and polyamine synthesis. Conclusions: Together, these data demonstrate that transient exposure to citalopram can induce long-lasting metabolomic changes during development, including dose-dependent changes that include aberrant metabolic processes in the developing metabolome. As a result, this work reveals potential biomarkers for early developmental exposure.

1. Introduction

Selective serotonin reuptake inhibitors (SSRIs), including citalopram, fluoxetine, and sertraline have been administered as antidepressants and anxiolytic medications since the 1970s [1]. These compounds target the serotonin transporter (SERT), blocking reuptake of serotonin (5-HT), causing a subsequent accumulation of serotonin in the synaptic space that prolongs serotonin signaling [2]. The therapeutic benefits of citalopram for depression and anxiety are attributed to action in the central nervous system (CNS). There are also effects on the peripheral nervous system that can be significant and are often responsible for side effects [3,4]. Peripheral 5-HT cannot cross the blood–brain barrier (BBB); however, it can influence CNS function indirectly via regulation of gut motility, the immune system, and metabolism [5,6,7]. Tryptophan, which can cross the BBB, is utilized for 5-HT synthesis in the CNS [8]. Most 5-HT is stored in intracellular vesicles and upon neuron depolarization, it enters the synaptic cleft where it can bind to G-protein coupled receptors on either the pre- or postsynaptic neuron [9]. Autoreceptors on presynaptic neurons inhibit further 5-HT release, whereas postsynaptic receptors, depending on the receptor subtype, will propagate inhibitory or excitatory pathways via second messenger cascades [10]. These pharmacological actions raise important questions about how SSRI exposure affects the developing embryo, where serotonin plays fundamentally different roles.
SSRIs are the most prescribed antidepressants with roughly 18% of American women taking the medication [11]. Although citalopram is commonly prescribed during pregnancy and postpartum, information how prenatal exposure may affect embryonic development remains largely unknown. In humans the embryonic period comprises the first 8 weeks after conception. During this time window, key processes such as gastrulation, neurulation, segmentation, and the initiation of organogenesis occur [12,13,14]. Early presence of 5-HT during these embryonic periods suggests the involvement of serotonin as a signaling molecule for development prior to it becoming a neuromodulator/neurotransmitter during later brain development [15]. For example, 5-HT, serotonin receptors, and serotonin-degrading enzymes like monoamine oxidase (MAO) are involved in the patterning of the left-right axis [16]. Craniofacial development and neural tube development also appear to be significantly impacted by 5-HT signaling [17,18]. In addition to axon guidance and neurite extension, serotonin regulates the proliferation of neural progenitors and synaptogenesis during early mammalian development via its involvement in the Wnt, MAPK, and SHH signaling pathways [19,20,21]. These signaling pathways play important roles during early development and since 5-HT can readily cross the placenta [22], alterations in 5-HT levels due to SSRI exposure during the first 8 weeks of pregnancy are likely to influence embryonic development.
In addition to concerns surrounding the effects of citalopram use during pregnancy, recent studies have brought attention to the presence of the drug within waterways. As consumption of citalopram and other antidepressant pharmaceuticals has increased, so has their occurrence in waterways with concentrations ranging from ng/L to µg/L [23]. The primary route of citalopram into the environment is via excretion from the body via urine and feces, but improper disposal of expired and unused mediation has also increased the environmental prevalence [24]. Additionally, citalopram and other SSRIs are not effectively removed by sewage treatment plants resulting in increased prevalence of these compounds in surface waters and municipal wastewater across the globe, amplifying concern for unintended exposures [25,26,27].
Reported data examining the effects caused by citalopram exposure during development of zebrafish and other model organisms illustrate how timing of exposure, dose, and other environmental factors play important roles during developmental stages [19]. Zebrafish embryos exposed to citalopram for 5 days at concentrations typical in surface water downstream of a wastewater plant exhibited increased heart rate (HR), metabolomic alterations, abnormal neurotransmitter profiles, altered feeding, and abnormal locomotion [28]. Xenopus laevis tadpoles exposed to citalopram at 2.5 mg/mL had increased potassium currents and decreased excitability of optic tectum neurons [29]. Citalopram concentrations ranging from 1 to 10 µg/L have induced scoliosis, yolk sac abnormalities, and pericardial edema in zebrafish embryos [30]. Additionally, zebrafish embryos exposed to 0.1–1000 µg/L for 7 days post fertilization exhibited an increased presence of reactive oxygen species, decreased locomotion, and alterations in the transcriptome related to muscle function, inflammation, and thyroid function [31] confirming similar findings observed in mammalian model organisms [32]. Even studies analyzing the effects of citalopram exposure in adult zebrafish have noted altered locomotive behaviors likely resulting from reduced numbers of glutamatergic and dopaminergic neurons [33]. It is evident that citalopram exposure induces both behavioral and morphological alterations in developing zebrafish embryos even at low concentrations; however, the biochemical changes driving these abnormalities are poorly understood.
Studies assessing associations between prenatal citalopram exposure and lasting postnatal effects in humans remain inconsistent. Some investigations of cord blood have found relationships between overall DNA methylation and prenatal citalopram exposure [34,35,36], whereas another has only shown specific increases in DNA methylation of certain genes such as NR3C2, which codes for the mineralocorticoid receptor and is vital for the infant stress response [37]. Additionally, a systematic review suggests that the lack of standard methodology, design, genome coverage, and statistical modeling complicate how these results should be interpreted [36]. Similar inconsistencies have been documented for the downstream diagnosis of ADHD, delayed psychomotor development, and delayed communication development in children exposed to citalopram [38,39,40]. Though the data surrounding the downstream developmental changes via epigenetics and clinical diagnostics are ambiguous, evidence in multi-omic profiling shows more consistent results.
Human embryonic stem cells (hESCs) are a model for understanding differentiation towards telencephalic neurons during the first trimester of human development [41]. hESCs and placental cells can be altered by brief and low-dose citalopram exposure. Differences in the expression of genes related to depression etiology, amino acid metabolic and catabolic processes, neuronal differentiation, and genes associated with hippocampal volume were observed on days 1–13 of continuous citalopram exposure in hESCs [42]. Amino acid transport and redox balance were particularly altered [42], suggesting an underrecognized effect of developmental citalopram exposure on gene expression specifically related to neuronal function and differentiation. Lower levels of astroglia-specific calcium-binding protein S100B, an early marker of brain maturation and central serotonergic function, and glycoprotein reelin, a protein involved in neuron migration and position, have also been documented in SSRI-exposed neonates [43,44]. Metabolomic profiling of placental samples from biobanks has shown changes in redox chemistry related to glutathione metabolism and mitochondrial alterations [45], suggestive of a complex metabolic shift in response to citalopram exposure in the developing embryo. Together, these data demonstrate the value of utilizing high-resolution mass spectrometry-based metabolomic analysis to gain detailed insight into the underlying mechanisms responsible for citalopram-induced effects.
Most of these studies have relied on constant exposures, unable to differentiate between acute changes (due to presence of the toxicant at the point of harvest) and changes that persist after exposure has ceased. Such persistent changes after discontinued chemical exposure have potential to cause lasting developmental perturbations [46,47,48]. To determine if a transient exposure to citalopram could have long-term effects on development, we conducted a separate transcriptomic analysis of zebrafish embryos exposed to citalopram for 24 h and then allowed them to develop drug-free for another 24 h [48]. Even at the lowest concentration tested (0.03 ng/mL), the discontinuous exposure regime resulted in significant changes in gene expression [48]. RNA sequencing of whole embryos revealed persistent, dose-specific shifts in gene expression. In the lowest concentration, glycine, serine, and threonine metabolism were upregulated, and genes impacting the eye lens, extracellular matrix, and cellular adhesion were perturbed. These results suggest that even low concentrations can have broad developmental consequences for lens morphogenesis and tissue organization. Higher, therapeutically relevant doses of citalopram elicited changes in stress-related processes including ferroptosis and glutathione metabolism, demonstrating the drugs’ ability to induce oxidative stress during early development. In every exposure group, genes pertaining to synaptic architecture and neurotransmission were altered, clearly demonstrating the ability of citalopram to elicit some degree of genetic remodeling.
Given these transcriptomic perturbations, we sought to determine how a similar experimental regimen would affect the metabolome. While transcriptomic and metabolomics analyses both can provide unique insight into how citalopram exposure affects regulatory networks, comparing the two techniques can bridge the gap between genomic potential and the reality of cellular processes. Additionally, integration of these techniques provides a more holistic understanding of how an embryo responds to discontinued citalopram exposure and can enhance confidence in identified biomarkers of exposure or the implicated mechanisms. We hypothesized that a transient, low-dose citalopram exposure to zebrafish embryos would result in a metabolic shift similar to that observed in the aforementioned transcriptomic analysis. The work presented here utilized a liquid chromatography-mass spectrometry-based untargeted metabolomics approach in conjunction with gross measures of physiology to study persistent effects of discontinued citalopram exposure.

2. Materials and Methods

2.1. Zebrafish Maintenance and Treatments

Adult wild-type Danio rerio (zebrafish) at four to six months old (Carolina Biological Supply Co., Burlington, NC, USA) were housed in 10 gallon tanks in fish water at a temperature of 28 (+/−1 °C). Dissolved oxygen was maintained at 7.8 mg/L. Water pH ranged from 7 to 7.6 and conductivity from 400 to 800 µS/cm. Embryos were collected within 0–2 h post-fertilization (hpf), sorted, and rinsed using 0.3× Danieau solution prior to citalopram exposure as described below. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee at Montana State University–Bozeman.
At 0–2 hpf, clutches were separated into one control group and four citalopram exposure groups: 0.03 ng/mL (environmental surface water concentration), 0.9 ng/mL (wastewater treatment plant effluent), 50 ng/mL (therapeutic human serum concentration), and 250 ng/mL (supertherapeutic human serum concentration) [48]. Citalopram hydrobromide solution was prepared in 0.3× Danieau solution and added to each treatment dish to achieve the designated concentrations. All dishes were labeled and incubated overnight at 28 °C.
To assess developmental effects following cessation of citalopram exposure, embryos were removed from treatment solutions at 24 hpf, rinsed three times with fresh Danieau solution, and transferred to clean dishes containing 0.3× Danieau without citalopram. Embryos were then returned to the incubator for an additional 24 h. At 48 hpf, embryos were either collected for morphological assessment or processed for metabolomic analysis. The first 24 h of zebrafish embryo development correspond to the first month of human development. As such, 48 hpf is analogous to the first 8 weeks of human gestation [49].
For metabolomic analysis, we followed a previous protocol [46]. Thirty embryos were randomly selected from each control and treatment groups and placed into individual Eppendorf tubes. Residual citalopram solution was carefully removed using a gel-loading pipette tip before samples were stored at −80 °C until analysis. Each experimental condition consisted of four to five independent biological replicates derived from separate clutches, with 30 embryos included per replicate.

2.2. Embryonic Imaging

Developmental progression of randomly selected embryos (n = 10 per group) was followed from 24 to 72 hpf in accordance with criteria established by Kimmel et al. (1995) [13]. Observations were conducted using a Zeiss Discovery V8 microscope (Zeiss AG, Oberkochen, Germany) mounted with a JenOptics Arktur camera (Jenoptik, Jena, Germany). Heart rate of randomly selected embryos (n = 5 per group) was measured using live feed by counting the number of beats in 30 s and multiplying by two [49]. Embryo length was determined using ImageJ version 1.54d (https://imagej.nih.gov/ij/ (accessed on 1 January 2025)) following previously established protocols [50].

2.3. Sample Preparation for Metabolomic Analysis

Each biological replicate consisted of 30 whole embryos for metabolite extraction. To promote protein precipitation and improve metabolite extraction, 200 μL of cold acetone was added to the samples prior to homogenization. Samples were homogenized using a water bath sonicator (Cole-Parmer, Vernon Hills, IL, USA) for 15 min at 15 °C. Following homogenization, samples were stored at −80 °C overnight. The following morning, samples were centrifuged at 16,100× g for 10 min at 4 °C. The resulting supernatant was collected to isolate metabolites and subsequently dried via a vacuum concentrator to evaporate solvent. Dried extracts were then resuspended in 100 μL of 1:1 acetonitrile:water in preparation for mass spectrometry analysis. Pooled samples were then generated by combining 5 μL of each resuspended extract. To monitor for contamination, two quality control samples were included: one processed through the full extraction protocol using only solvents and another containing only mass spectrometry-grade water.

2.4. LC-MS/MS Metabolite Analysis

Samples were subjected to liquid chromatography–mass spectrometry (LC-MS) analysis using an Acquity I-Class UPLC system coupled via an electrospray ionization source to a Waters Synapt XS mass spectrometer (Waters, Milford, MA, USA). Metabolite separation was achieved with a Waters BEH-HILIC column (2.1 × 100 mm) operated at a flow rate of 400 μL/min. The mobile phase consisted of solvent A (95% water, 5% acetonitrile, 0.1% formic acid) and solvent B (95% acetonitrile, 5% water, 0.1% formic acid). The 19 min gradient progressed from 95% to 25% solvent B over 12 min, followed by a 5 min wash at 25% solvent B, with each run beginning with a 2 min equilibration period. Quality control blanks were introduced after every 12 injections to monitor instrument stability and detect potential spectral drift. All samples were analyzed using standard MS1 acquisition, while pooled samples were additionally subjected to LC-MS/MS with a fixed collision energy ramp ranging from 20 to 50 V. Data from each run were manually reviewed to ensure consistent instrument performance and identify any anomalies.

2.5. Global Metabolomic Profiling

LC-MS datasets, including mass-to-charge (m/z) values, ion intensities, and retention times, were processed using Progenesis QI (Nonlinear Dynamics, Newcastle, UK) in conjunction with MetaboAnalyst. To address deviations from normality, the dataset was subjected to quantile normalization, log transformation, and autoscaling.
MetaboAnalyst was used for statistical evaluation and visualization of metabolomic data. Volcano plot analysis was applied to evaluate both the magnitude and statistical significance of observed changes (Figure S1). Principal component analysis (PCA) and partial least squares–discriminant analysis (PLS-DA) were performed to examine similarities and differences between control and citalopram-exposed metabolomic profiles. Predictive performance of the PLS-DA models was assessed using Q2 cross-validation metrics (Figure S2) [51]. Hierarchical clustering analysis (HCA) was conducted to group metabolite features with similar regulation patterns. Together, these approaches enabled the identification of metabolite groups exhibiting differential regulation across treatments. Functional pathway enrichment analysis within MetaboAnalyst was used to infer biologically relevant metabolic networks associated with the detected features. Figures and graphical outputs were generated using GraphPad Prism (version 10.6.1), R (version 4.5.1), and Adobe Illustrator (version 30.0).
Metabolite identification based on LC-MS/MS data was carried out in Progenesis QI. Both MS1 and MS2 centroid data were imported for alignment and peak detection. Compound annotations were determined by comparing experimental fragmentation patterns with reference spectra from the Human Metabolome Database (HMDB) and an in-house standards library (Mass Spectrometry Library of Standards, IROA Technologies, Ann Arbor, MI, USA). Candidate metabolites were accepted only if they achieved a Progenesis confidence score above 30, taking into account mass accuracy, isotope distribution, and fragmentation agreement. Features with mass errors exceeding 20 parts per million (ppm) were excluded from further consideration.

3. Results

3.1. Morphological Analysis of Zebrafish Embryos After Citalopram Exposure

To assess the persistent effects of citalopram exposure on development, freshly fertilized zebrafish embryos were treated with citalopram for 24 h and then allowed to develop for an additional 24 h in citalopram-free conditions. At 48 h post-fertilization (hpf), embryos were collected for morphological and metabolomic analyses. This citalopram exposure regimen, subsequently referred to as discontinued citalopram exposure, is summarized in Figure 1.
To investigate the effects of the discontinued citalopram exposure on gross embryo development and physiology, a series of morphological characteristics were measured at 24–72 hpf. Heart rate (HR) was measured at 24, 48, and 72 hpf. At 24 hpf, HR was significantly different across the dose regime with the highest HR being measured in the 50 ng/mL citalopram-exposed embryos (ANOVA, p < 0.001) (Figure 2A). At 48 hpf, the average HR was higher in all exposure groups when compared to the controls; however, the response was not linear—rather U-shaped with large increases in both the 0.03 ng/mL and 0.9 ng/mL embryos but a smaller overall increase in the 50 ng/mL and 250 ng/mL embryos (p < 0.05). At 72 hpf, the citalopram-exposed embryos all exhibited significantly decreased HRs relative to the control embryos (p < 0.001). Complete statistical outputs for all HRs are included in Supplemental Table S1.
Body length of embryos at 72 hpf was measured to determine if any other significant differences in morphology arose in the citalopram-exposed embryos. No significant differences in embryo length at 72 hpf (Figure 2B) were present. Eye and tail development also appeared normal, suggesting that gross physiology and morphology did not exhibit any significant differences. Complete statistical outputs for body lengths are included in Supplemental Table S2.

3.2. Metabolomic Analysis

Global metabolomic profiling is a powerful tool for detecting an organismal response to environmental perturbations while providing insight into the molecular mechanisms governing the response [52]. Therefore, employing a highly sensitive metabolomic analysis can yield insight into subtle changes that may be overlooked with traditional physiological measurements. To investigate how discontinued citalopram exposure affected zebrafish embryonic development, an untargeted metabolomic analysis was performed on independent clutches of control (n = 4), 0.03 ng/mL (n = 4), 0.9 ng/mL (n = 5), 50 ng/mL (n = 5), and 250 ng/mL (n = 5) citalopram-exposed embryos. In total, 4813 metabolomic features were detected by liquid chromatography-mass spectrometry (LC-MS) across samples prepared from total embryos. Multivariate statistics were utilized to assess large-scale changes between the citalopram-exposed and control groups. A total of 88 metabolomic features with significant abundance changes (Figure 3A) were revealed by a one-way ANOVA (false discovery rate (FDR) adjusted p-value < 0.05). Abundance changes fluctuated from two- to six-fold.
To reduce the dimensionality of the data while preserving group-descriptive information, partial least squares-discriminant analysis (PLS-DA) was performed. Though PLS-DA revealed overlap between experimental groups, the impact of citalopram exposure is evident in component 1, which accounted for 27.5% of variability in the dataset (Figure 3B). By focusing on the top 25 most discriminatory metabolomic features using PLS-DA clustering and a heatmap, a distinct metabolic shift in the embryonic response to citalopram exposure was observed (Figure 3C). While some metabolomic features had a clear linear increase in abundance with increased citalopram exposure, such as the m/z 854.5693_3.68, some also clearly decreased in abundance like the m/z 503.4574_0.64 (Figure 3C). Non-linear behavior was also observed in the intensity of several metabolomic features including 623.5041_4.14. The diverse patterns of metabolomic abundance changes in response to the citalopram regime emphasizes the complex metabolic shifts that are occurring. Several metabolites that differentiated the citalopram exposure conditions were identified with high confidence, including (−)-Epigallocatechin, 1,3-octadiene, and hydroxy DHEA sulfate. Though it is not entirely clear how these molecules may influence embryo developments, they are potential biomarkers for citalopram exposure. Together, these data confirm that even an ultra-low, discontinued citalopram exposure causes persistent metabolic changes in the developing zebrafish metabolome.

3.3. Pairwise Analyses

To gauge the magnitude of the metabolomic shift in citalopram-exposed zebrafish embryos, pairwise analyses using PCA and volcano plots were conducted. Pairwise PCA was performed for each citalopram concentration versus the control. Each group had relatively good separation of citalopram and control embryos (Figure 4A–D). The lowest citalopram concentration (0.03 ng/mL) had the most overlap with control embryos, as expected. As citalopram concentration increased, samples within each group clustered tighter reflecting increased intragroup similarity and greater intergroup variation.
Given that the lowest citalopram concentration elicited the largest degree of intragroup variation (Figure 3B), and PLS-DA analysis suggested a possible linear response of the metabolome to higher citalopram concentrations, it was suspected that the 0.03 ng/mL group would have the least number of dysregulated metabolomic features via volcano plot analysis. Surprisingly, this was not found to be true. Volcano plots identified total increased/decreased metabolite features for each citalopram condition compared to controls: 143/89 for 0.03 ng/mL, 130/56 for 0.9 ng/mL, 264/343 for 50 ng/mL, and 84/92 for 250 ng/mL (Figure S1). To represent these differences visually, these data were plotted via a histogram to highlight differential regulation of features (Figure 4E). Taken together, these data not only demonstrate the ability of extremely low doses of citalopram to alter zebrafish embryonic metabolic status, but it also shows a significant and nonlinear response of the embryonic metabolome to the drug.

3.4. Pathway Analysis

To pursue pathways implicated in the persistent metabolic disruptions induced by discontinued citalopram exposure, a pathway enrichment analysis was performed. Identification and mapping of metabolites utilized Progenesis QI and functional pathway analysis from MetaboAnalyst. An abundance of dysregulated pathways were mapped to citalopram exposure groups (Table S3, Figure 5A); however, only one pathway, arginine and proline metabolism, was shared by all four groups (Figure 5A).
As citalopram exposure increased, a non-linear increase in the number of unique pathways associated with each concentration was observed. The 0.03 ng/mL group had perturbations in the following nine unique pathways: butanoate metabolism; alanine, aspartate, and glutamate metabolism; propanoate metabolism; cysteine and methionine metabolism; glycolysis/gluconeogenesis; lysine degradation; citrate cycle, phosphonate and phosphinate metabolism; and lipoic acid metabolism (Table S3). The 0.9 ng/mL group had unique perturbations in the following five pathways: riboflavin metabolism; terpenoid backbone biosynthesis; glycerolipid metabolism; one carbon pool by folate; and glycine, serine, and threonine metabolism (Table S3). The 50 ng/mL group had perturbations in the following four pathways: glutathione metabolism; caffeine metabolism; beta-alanine metabolism; and metabolism of xenobiotics by cytochrome P450 (Table S3). The 250 ng/mL group had perturbations in the following 16 unique pathways: N-glycan biosynthesis; putative anti-inflammatory metabolites formation from EPA; amino sugar metabolism; glycosphingolipid biosynthesis—neolactoseries; R group synthesis; sialic acid metabolism; aspartate and asparagine metabolism; vitamin B9 (folate) metabolism; leukotriene metabolism; polyunsaturated fatty acid biosynthesis; blood group biosynthesis; glycosphingolipid biosynthesis—globoseries; carnitine shuttle; glycosphingolipid metabolism; carbon fixation; and urea cycle/amino group metabolism (Table S3). The 250 ng/mL citalopram-exposed group was particularly striking, exhibiting the most extensive pathway disruptions despite having the lowest number of dysregulated features (Figure 4E and Figure 5A). This phenomenon is often attributed to the systems-level nature of complex toxicology and/or disease, which can induce subtle, coordinated changes across multiple pathways [53]. This may cause many small changes that are not detected at the feature level, but their cumulative effect within pathways is enough to disrupt pathway function.
Next, violin plots were generated to investigate how citalopram exposure levels affected metabolites in the arginine and proline metabolism pathway that was shared by all groups. Five metabolites from the pathway were identified that changed significantly in response to the citalopram exposure (Figure 5B). Non-linear behavior was observed in several metabolites including 4-aminobutyraldehyde, proline, and putrescine. N4-acetylaminobutanal decreased in abundance with increased citalopram exposure, whereas S-adenosylmethioninamine increased in abundance with an increase in citalopram concentrations (Figure 5B). Taken together, these data demonstrate that different concentrations of citalopram exposure produce differential effects on zebrafish embryonic metabolism. Additionally, they emphasize that an increase in exposure concentration does not necessarily result in a linear response of metabolites or pathway regulation.

4. Discussion

To our knowledge, there are no other reports that analyze the persistent metabolomic effects of a non-continuous citalopram exposure in zebrafish embryos (Figure 1). The majority of studies have harvested embryos directly from the citalopram-containing solutions, making it impossible to distinguish acute versus persistent effects. Here, we evaluated the persistent effects of discontinued citalopram exposure via untargeted metabolomic analysis and morphological assessments.

4.1. Heart Rate but Not Gross Morphology Was Altered by Environmentally or Biologically Relevant Citalopram Exposure

To investigate if discontinued citalopram exposure resulted in general physiological or morphological changes in developing zebrafish embryos, heart rate (HR), embryo length, and gross morphology were assessed from 24 hpf to 72 hpf. We found no difference between control and treatment groups with respect to body length or morphology (Figure 2B). Although one study showed that exposure to 1–10 ng/mL of citalopram results in developmental abnormalities like scoliosis, yolk sac edema, and pericardial edemas [30], those embryos were exposed for the entirety of their development. Our findings demonstrated that in contrast to persistent exposure, a discontinued citalopram exposure did not result in changes to the embryos’ gross morphology.
In contrast, HR from control and citalopram-exposed embryos were significantly different (Figure 2A). A study exposing zebrafish embryos to constant environmentally relevant citalopram concentrations (1–10 ng/mL) reported that HR measured at 48 hpf, 72 hpf, and 96 hpf decreased with increased citalopram exposure [28]. Embryos in our study underwent a discontinued citalopram exposure that could explain the more complex patterns observed. Increases in HR, specifically those observed at the 24 hpf and 48 hpf time points (Figure 2A), could include a compensatory response to removal of the citalopram after the 24 h exposure period and before HR determination. Compensatory feeding and phototactic behaviors in response to citalopram exposure have been observed and even linked to metabolic compensation, serotonergic disruption, and oxidative collapse in zooplankton [54]. Therefore, it is possible that citalopram exposure could result in unknown compensatory responses after toxicant removal, such as the increased HR observed at 24 hpf and 48 hpf (Figure 2A).

4.2. Non-Linear Metabolomic Changes Are Induced by Discontinued Citalopram Exposure

Developmental abnormalities often require prolonged exposures or severe stress and usually represent a cumulative outlook of metabolic disruptions [46,55]. Given that metabolic and physiological responses are far more rapid than morphological ones and this analysis included extremely low concentrations of citalopram exposure, we hypothesized that investigation of the metabolome would be more sensitive and could potentially elucidate mechanisms that could lead to long-term biological changes. We found that the metabolomes of 0.03 ng/mL (environmental surface waters), 0.9 ng/mL (wastewater treatment plant effluent), 50 ng/mL (therapeutic human serum concentration), and 250 ng/mL (supertherapeutic serum concentration) citalopram-exposed embryos underwent a global metabolomic shift compared to control embryos (Figure 3). Though differential metabolomic changes resulting from low-dose citalopram exposure of 12 ng/mL have been documented, these results were collected from a targeted analysis on embryos continuously exposed and harvested at 96 hpf [30]. Our data suggest that even an extremely low-dose, discontinuous exposure of 0.03 ng/mL citalopram is sufficient to induce metabolic changes in a developing zebrafish embryo (Figure 3).
Pairwise analyses examining the responses to the exposure highlight a differential response depending on citalopram concentration (Figure 4). PLS-DA analysis demonstrated that as citalopram concentration increases, biological replicates cluster tighter—suggesting that citalopram exposure may be eliciting a specific and coordinated biological response in the embryos. Although the highest concentration (250 ng/mL citalopram) had the tightest clustering when compared to other citalopram-exposed groups in the PLS-DA, the therapeutically relevant dose of 50 ng/mL had the largest number of differentially regulated metabolomic features via volcano plot analysis (Figure 4 and Figure S1). Tightness observed in PLS-DA clustering is a visual representation of how well the groups separate in a reduced dimensional space. Tight clustering does not necessarily reflect a greater number of distinct differential features, it can also reflect a strong cohesive response of a few features in the group [56,57]. In this case, the tight clustering is due to a limited number of metabolic features with a coherent and distinct pattern of abundance in the 250 ng/mL group. This is in contrast to the large number of metabolites driving differences in the other exposure groups. Therefore, the metabolomic features driving the separation of the 250 ng/mL group, such as epigallocatechin, may represent targets for biomarker development in screening for higher dose citalopram exposures during early development.
In a parallel study analyzing the transcriptome of zebrafish embryos exposed to the same discontinued citalopram regimen, we observed a hormetic response in the transcriptome of zebrafish embryos [48]. The lowest and highest citalopram concentrations, 0.03 ng/mL and 250 ng/mL, elicited the highest numbers of dysregulated genes. Although different cellular processes were affected at each concentration, several genes mapped to synaptic architecture and neurotransmission at both concentrations [48]. The embryos exposed to the middle concentrations of 0.9 ng/mL and 50 ng/mL had the fewest number of differentially expressed genes. The metabolomic analysis also showed a non-linear response to citalopram exposure; however, the metabolomics data appeared to have an inverted hormetic response compared to the hormetic response in the transcriptomic analysis. In the metabolic analysis, the two lower and the highest citalopram concentrations resulted in the lowest number of dysregulated metabolites, whereas the 50 ng/mL concentration resulted in the largest number of dysregulated metabolites. Metabolic alterations can act as both upstream drivers and downstream consequences of transcriptional regulation [58,59,60,61]. Thus, the inverse relationship between the metabolomic and transcriptomic changes in response to citalopram exposure likely reflects a complex crosstalk between metabolism and transcriptional activity.
Our transcriptomic analysis revealed two distinct metabolic disruptions based on citalopram dose. At low citalopram exposure concentrations, upregulation of genes belonging to glycine, serine, and threonine metabolism were observed, whereas at high citalopram concentrations, genes mapping to stress-related pathways (glutathione metabolism and ferroptosis) were observed [48]. In agreement with the transcriptomic analysis, the metabolites measured here mapped to multiple amino acid pathways including glycine, serine, and threonine metabolism, which were dysregulated in the low citalopram concentrations, 0.03 ng/mL and 0.9 ng/mL (Table S1). Glutathione metabolism emerged as a perturbed pathway in the 50 ng/mL citalopram-exposed embryos, supporting the transcriptomic results and implicating oxidative stress as a pathway in citalopram-induced metabolic dysregulation. Taken together, these datasets suggest that a transient citalopram exposure is sufficient to induce changes in amino acid metabolism and stress responses, emphasizing the value of multi-omics analyses in toxicology.
Very few pathways were shared among all citalopram concentrations in the transcriptomic analysis [48]. Similarly, only perturbations in one pathway, arginine and proline metabolism, were shared between the metabolomes of all citalopram exposure concentrations (Figure 5A). While the transcriptomic analysis demonstrated that all of the citalopram exposure concentrations resulted in impairments to pathways broadly related to synapse function, neurotransmission, and plasticity in the embryos [48], the metabolomics data did not reveal significant impairments of pathways related to neurotransmission. The metabolome and transcriptome can often exhibit seemingly discordant responses as they are capturing different layers of biological regulation and operate on different time scales [62,63,64,65]. While the metabolome measures what the organism is actively doing via the actual biological outcomes in the small molecules—which may respond within seconds of a stimulus—the transcriptome measures what the cell is planning on doing via mRNA and responds on time scales close to minutes or hours [66]. The metabolome is also subjected to substrate availability, cofactor ratio, and potentially, multiple sources of regulation (allosteric regulation, feedback inhibition, etc.), which could explain how metabolic output is decoupled from the gene expression levels [67]. These results underscore the necessity of integrating diverse analytical platforms to fully capture the biological impact of toxicant exposures.
Citalopram is known to exhibit a dose-response effect on the binding of serotonin transporters (SERTs) [68]. These proteins are responsible for 5-HT recycling by resorbing 5-HT from the synaptic cleft [69]. SERTs can be occupied rapidly by citalopram until a threshold is reached at which higher doses do not elicit any additional benefits, but increase the risk of side effects [68,70]. Our data may indicate the presence of feedback inhibition where citalopram binding of SERTs reached a threshold or became saturated at the 50 ng/mL dose. This could explain the apparent decrease in metabolic effects observed in the highest dose of 250 ng/mL.
These non-linear patterns may also implicate the activity of 5-HT autoreceptors. 5-HT autoreceptors play crucial roles in serotonergic neurotransmission, providing a negative feedback loop via the reduction in 5-HT synthesis and release in addition to modulating 5-HT reuptake by SERTs [71,72,73]. Low-to-moderate doses of citalopram can result in an incomplete activation of the 5-HT1A and 5-HT1B autoreceptors, thereby failing to activate desensitization [68]. In contrast, high exposures, such as the 250 ng/mL dose demonstrated in our experiment, may result in a robust autoreceptor activation leading to desensitization, ultimately producing less dramatic effects (Figure 4E) [71]. The dose-response relationships observed here are novel, providing avenues for understanding the consequences of citalopram exposure to the developing embryo.
Although there are apparent differences in the transcriptomic and metabolomic responses to citalopram exposure, both studies observed non-linear responses to discontinued citalopram exposure. In mice, rats, and other developmental models, dose- and time-dependent serotonin signaling [74,75] and other responses to citalopram exposure have been reported as exhibiting hormetic behavior [48,76,77,78]. Together, these data show that citalopram exposure elicits a strong and non-linear response from the genomic to the metabolomic scale in the developing zebrafish embryo.

4.3. Arginine and Proline Metabolism

Identifying the biochemical pathways perturbed by citalopram exposure is required for building a mechanistic model of toxicity, predicting adverse outcomes, assessment of risk, and biomarker development. We were surprised that only one pathway was shared by all citalopram exposure concentrations: arginine and proline metabolism (Figure 5A).
Arginine is an amino acid with multiple metabolic fates. The metabolism of arginine is modulated by numerous transporters responsible for movement of arginine and its metabolites across membranes and between cell types [79]. Not only is arginine metabolically interconvertible with proline and glutamate, it also is a precursor for the synthesis of proteins, polyamines, urea, creatine, agmatine, and notably, nitric oxide (NO) [79]. NO is synthesized from arginine and, in the CNS, is responsible for neuronal development, cell proliferation, synaptic plasticity, modulation of 5-HT, and long-term potentiation [80]. NO is also crucial for regulating the placental and fetal blood flow for oxygen and nutrient transfer between mother and fetus [81]. Regulation of NO is therefore vitally important during early development, given that NO overproduction has been shown to induce cell damage and cell death via ROS production [82].
Nitrergic signaling is morphologically and functionally tied to 5-HT signaling. Neurons responsible for the synthesis and release of these neurotransmitters colocalize in the hypothalamus and dorsal raphe nucleus of the brain [80,83]. A connection between NO and 5-HT signaling systems has been observed in young and adult animals, suggesting that the systems play roles in developing and maintaining the CNS. Alterations of 5-HT signaling via citalopram exposure have been shown to alter NO signaling in mice [84], indicative of changes in the biochemistry of these cycles. Here, the arginine pathway metabolites 4-aminobutyraldehyde, N4-acetylaminobutanol, proline, S-adenosylmethionine, and putrescine were all identified as being significantly perturbed by citalopram exposure (Figure 5B). This could influence CNS development and function through NO signaling and synthesis.
Arginine metabolism is also key in the production of polyamines. Polyamines, including putrescine, spermidine, and spermine, are recognized for their roles in DNA regulation, protein synthesis, suppression of inflammation, and cell proliferation and differentiation during early development [85,86,87]. By binding negatively charged macromolecules like DNA and RNA, polyamines help protect the cell against damage from ROS [88]. During blastocyst stages in mice, transcription of ODC1, the protein responsible for polyamine biosynthesis, increases [85] reflecting the demand for cellular proliferation. In our metabolomic analysis, putrescine abundance decreased as citalopram levels increased (Figure 5B), which could be reflective of persistent alterations in cell proliferation processes occurring during embryogenesis. Polyamine levels have been shown to be altered in humans taking therapeutic levels of citalopram, though the consequences of these alterations remain unknown [89]. The decrease in polyamines could also be reflective of a stress response induced by the citalopram exposure. In adult rat brains responding to stress, the polyamine stress response (PSR) is constructive. It is characterized by an increase in polyamine metabolism resulting in an increase in putrescine and a decrease in spermidine and spermine [90]. In contrast, the PSR in the developing rat brain results in a decrease in polyamine metabolism and is likely implicated in cell death and pathophysiological and behavior changes [91]. The decrease in putrescine observed here could be reflective of an embryonic polyamine stress response, suggesting subtle changes in physiology may be occurring in citalopram-exposed embryos. To further understand any potential persistent, maladaptive responses of the developing embryo to citalopram exposure, a more targeted analysis of polyamine and arginine pathway is necessary.

5. Conclusions

Zebrafish embryos exposed to a discontinued citalopram exposure regimen exhibited significantly different metabolomic profiles than control embryos. These data demonstrate that even relatively brief exposure to citalopram at concentrations found in freshwater habitats can induce metabolic changes in a developing embryo. Although gross changes to morphology were absent, significant alterations in heart rates and metabolomic profiles were observed. Changes in arginine and proline metabolism may reflect changes in neuronal development by altering nitric oxide and polyamine synthesis. These findings are suggestive of a nonlinear metabolic response resulting in increased redox stress and alterations in neural signaling. This study provides new insight into the biochemistry underpinning early embryonic citalopram exposure and supports the development of dose-dependent citalopram relationships.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/biomedicines14020381/s1: Figure S1: Volcano plots of dysregulated metabolites in citalopram-exposed and control embryos; Table S1: Complete statistical output for analysis of heart rate data; Table S2: Complete statistical output for analysis of body length; Table S3: Perturbed pathways in each citalopram exposure condition; Figure S2: All groups’ PLS-DA cross-validation calculated via 5-fold CV validation method.

Author Contributions

Conceptualization, C.S.M., B.B., R.J.N. and G.C.; methodology, G.C., R.J.N.; software, G.C. and D.A.; validation, G.C., B.B. and D.A.; formal analysis, G.C. and D.A.; investigation, G.C., R.J.N. and D.A.; resources, C.S.M. and B.B.; data curation, G.C. and R.J.N.; writing—original draft preparation, G.C.; writing—review and editing, G.C., R.J.N., D.A., C.S.M. and B.B.; visualization, G.C.; supervision, C.S.M. and B.B.; project administration, C.S.M. and B.B.; funding acquisition, C.S.M. and B.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by grants from Montana INBRE Technology Access, National Institute of General Medical Sciences (P20GM103474) and MSU’s College of Agriculture and by NSF grant HRD-2054276 to CSM.

Institutional Review Board Statement

The animal study protocol was approved by the IACUC Institutional Review Board of Montana State University (protocol number 2022-125-IA, approval date 10 August 2022).

Data Availability Statement

The raw data supporting the conclusions of this article can be found at the following DOI: 10.6084/m9.figshare.31268044.

Acknowledgments

We thank the Montana State University Mass Spectrometry facility members, Don Smith and Jesse Thomas, for their help with instrumentation and the members of the Animal Resource Center for their assistance with animal husbandry. We thank the students of the Trails to Research program, who inspired this project. This research was aided by the University of Montana Genomics Core and the Montana INBRE Data Science Core, which are funded by the National Institute of General Medical Sciences (P20GM103474), the Office of the Vice President for Research and Creative Scholarship at the University of Montana, and the M. J. Murdock Charitable Trust. The content is solely the responsibility of the authors and does not necessarily represent the official views of the UMGC or the National Institutes of Health.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Laux, G. Serotonin Reuptake Inhibitors: Citalopram, Escitalopram, Fluoxetine, Fluvoxamine, Paroxetine, and Sertraline. In NeuroPsychopharmacotherapy; Riederer, P., Laux, G., Nagatsu, T., Le, W., Riederer, C., Eds.; Springer International Publishing: Cham, Switzerland, 2020; pp. 1–13. [Google Scholar]
  2. Shoar, N.S.; Fariba, K.A.; Padhy, R.K. Citalopram. Available online: https://www.ncbi.nlm.nih.gov/books/NBK482222/ (accessed on 7 September 2025).
  3. Skop, B.P.; Brown, T.M. Potential Vascular and Bleeding Complications of Treatment with Selective Serotonin Reuptake Inhibitors. Psychosomatics 1996, 37, 12–16. [Google Scholar] [CrossRef] [PubMed]
  4. Anderson, G.M. Peripheral and central neurochemical effects of the selective serotonin reuptake inhibitors (SSRIs) in humans and nonhuman primates: Assessing bioeffect and mechanisms of action. Int. J. Dev. Neurosci. 2004, 22, 397–404. [Google Scholar] [CrossRef]
  5. Yabut, J.M.; Crane, J.D.; Green, A.E.; Keating, D.J.; Khan, W.I.; Steinberg, G.R. Emerging Roles for Serotonin in Regulating Metabolism: New Implications for an Ancient Molecule. Endocr. Rev. 2019, 40, 1092–1107. [Google Scholar] [CrossRef] [PubMed]
  6. Zhang, H.; Leitner, D.R.; Hasegawa, Y.; Waldor, M.K. Peripheral serotonergic neurons regulate gut motility and anxiety-like behavior. Curr. Biol. 2024, 34, R133–R134. [Google Scholar] [CrossRef]
  7. El-Merahbi, R.; Löffler, M.; Mayer, A.; Sumara, G. The roles of peripheral serotonin in metabolic homeostasis. FEBS Lett. 2015, 589, 1728–1734. [Google Scholar] [CrossRef] [PubMed]
  8. Best, J.; Nijhout, H.F.; Reed, M. Serotonin synthesis, release and reuptake in terminals: A mathematical model. Theor. Biol. Med. Model. 2010, 7, 34. [Google Scholar] [CrossRef]
  9. Hensler, J.G. Serotonin. In Basic Neurochemistry; Elsevier: Amsterdam, The Netherlands, 2012; pp. 300–322. [Google Scholar]
  10. Mohammad-Zadeh, L.F.; Moses, L.; Gwaltney-Brant, S.M. Serotonin: A review. J. Vet. Pharmacol. Ther. 2008, 31, 187–199. [Google Scholar] [CrossRef]
  11. Brody, D.J.; Gu, Q. Antidepressant Use Among Adults: United States, 2015–2018; NCHS Data Brief, no. 377; National Center for Health Statistics: Hyattsville, MD, USA, 2020. Available online: https://www.cdc.gov/nchs/products/databriefs/db377.htm (accessed on 15 January 2026).
  12. Rossant, J.; Tam, P.P. Early human embryonic development: Blastocyst formation to gastrulation. Dev. Cell 2022, 57, 152–165. [Google Scholar] [CrossRef]
  13. Kimmel, C.B.; Ballard, W.W.; Kimmel, S.R.; Ullmann, B.; Schilling, T.F. Stages of embryonic development of the zebrafish. Dev. Dyn. 1995, 203, 253–310. [Google Scholar] [CrossRef]
  14. Smith, N.L.; Kimelman, D. Establishing the body plan: The first 24 hours of zebrafish development. In The Zebrafish in Biomedical Research; Elsevier: Amsterdam, The Netherlands, 2020; pp. 81–88. [Google Scholar]
  15. Côté, F.; Fligny, C.; Bayard, E.; Launay, J.-M.; Gershon, M.D.; Mallet, J.; Vodjdani, G. Maternal serotonin is crucial for murine embryonic development. Proc. Natl. Acad. Sci. USA 2007, 104, 329–334. [Google Scholar] [CrossRef]
  16. Levin, M.; Buznikov, G.A.; Lauder, J.M. Of minds and embryos: Left-right asymmetry and the serotonergic controls of pre-neural morphogenesis. Dev. Neurosci. 2006, 28, 171–185. [Google Scholar] [CrossRef] [PubMed]
  17. Lauder, J.M. Neurotransmitters as morphogens. Prog. Brain Res. 1988, 73, 365–387. [Google Scholar] [CrossRef] [PubMed]
  18. Lauder, J.M.; Wallace, J.A.; Krebs, H. Roles for serotonin in neuroembryogenesis. Adv. Exp. Med. Biol. 1981, 133, 477–506. [Google Scholar] [CrossRef] [PubMed]
  19. Brummelte, S.; Mc Glanaghy, E.; Bonnin, A.; Oberlander, T.F. Developmental changes in serotonin signaling: Implications for early brain function, behavior and adaptation. Neuroscience 2017, 342, 212–231. [Google Scholar] [CrossRef]
  20. Trakhtenberg, E.F.; Goldberg, J.L. The role of serotonin in axon and dendrite growth. Int. Rev. Neurobiol. 2012, 106, 105–126. [Google Scholar]
  21. Higuchi, Y.; Arakawa, H. Serotonergic mediation of the brain-wide neurogenesis: Region-dependent and receptor-type specific roles on neurogenic cellular transformation. Curr. Res. Neurobiol. 2023, 5, 100102. [Google Scholar] [CrossRef]
  22. Bonnin, A.; Levitt, P. Fetal, maternal, and placental sources of serotonin and new implications for developmental programming of the brain. Neuroscience 2011, 197, 1–7. [Google Scholar] [CrossRef] [PubMed]
  23. Słoczyńska, K.; Orzeł, J.; Murzyn, A.; Popiół, J.; Gunia-Krzyżak, A.; Koczurkiewicz-Adamczyk, P.; Pękala, E. Antidepressant pharmaceuticals in aquatic systems, individual-level ecotoxicological effects: Growth, survival and behavior. Aquat. Toxicol. 2023, 260, 106554. [Google Scholar] [CrossRef] [PubMed]
  24. Sangkuhl, K.; Klein, T.E.; Altman, R.B. PharmGKB summary: Citalopram pharmacokinetics pathway. Pharmacogenetics Genom. 2011, 21, 769–772. [Google Scholar] [CrossRef] [PubMed]
  25. Lajeunesse, A.; Smyth, S.A.; Barclay, K.; Sauvé, S.; Gagnon, C. Distribution of antidepressant residues in wastewater and biosolids following different treatment processes by municipal wastewater treatment plants in Canada. Water Res. 2012, 46, 5600–5612. [Google Scholar] [CrossRef]
  26. Lajeunesse, A.; Gagnon, C.; Gagné, F.; Louis, S.; Čejka, P.; Sauvé, S. Distribution of antidepressants and their metabolites in brook trout exposed to municipal wastewaters before and after ozone treatment—Evidence of biological effects. Chemosphere 2011, 83, 564–571. [Google Scholar] [CrossRef] [PubMed]
  27. Schultz, M.M.; Furlong, E.T.; Kolpin, D.W.; Werner, S.L.; Schoenfuss, H.L.; Barber, L.B.; Blazer, V.S.; Norris, D.O.; Vajda, A.M. Antidepressant Pharmaceuticals in Two U.S. Effluent-Impacted Streams: Occurrence and Fate in Water and Sediment, and Selective Uptake in Fish Neural Tissue. Environ. Sci. Technol. 2010, 44, 1918–1925. [Google Scholar] [CrossRef]
  28. Romero-Alfano, I.; Prats, E.; Raldúa, D.; Blazquez, M.; Tauler, R.; Piña, B.; Gómez-Canela, C.; Barata, C. Assessing neurological and cardiovascular effects caused by pharmaceuticals in river water: Insights from Daphnia magna and Danio rerio embryos. Chemosphere 2025, 385, 144541. [Google Scholar] [CrossRef]
  29. Liu, K.; Garcia, A.; Park, J.J.; Toliver, A.A.; Ramos, L.; Aizenman, C.D. Early Developmental Exposure to Fluoxetine and Citalopram Results in Different Neurodevelopmental Outcomes. Neuroscience 2021, 467, 110–121. [Google Scholar] [CrossRef]
  30. Mohanthi, S.; Sutha, J.; Gayathri, M.; Ramesh, M. Evaluation of the citalopram toxicity on early development of zebrafish: Morphological, physiological and biochemical responses. Environ. Pollut. 2024, 357, 124399. [Google Scholar] [CrossRef]
  31. Kazi, K.J.; English, C.D.; Ivantsova, E.; Souders Ii, C.L.; Martyniuk, C.J. Transcriptome networks and physiology related to cardiac function and motor activity are perturbed in larval zebrafish (Danio rerio) following exposure to the antidepressant citalopram. Environ. Pollut. 2024, 361, 124767. [Google Scholar] [CrossRef] [PubMed]
  32. Zahra, A.; Du, L.; Jia, M.; Butt, M.U.; Wang, Q.; Wang, Y.; Wu, J. Prenatal exposure of citalopram elicits depression-like and anxiety-like behaviors and alteration of morphology and protein expression of medial prefrontal cortex in young adult mice. J. Integr. Neurosci. 2022, 21, 61. [Google Scholar] [CrossRef]
  33. Hong, X.; Chen, R.; Zhang, L.; Yan, L.; Xin, J.; Li, J.; Zha, J. Long-Term Exposure to SSRI Citalopram Induces Neurotoxic Effects in Zebrafish. Environ. Sci. Technol. 2022, 56, 12380–12390. [Google Scholar] [CrossRef] [PubMed]
  34. Kallak, T.K.; Bränn, E.; Fransson, E.; Johansson, Å.; Lager, S.; Comasco, E.; Lyle, R.; Skalkidou, A. DNA methylation in cord blood in association with prenatal depressive symptoms. Clin. Epigenetics 2021, 13, 78. [Google Scholar] [CrossRef]
  35. Viuff, A.-C.F.; Pedersen, L.H.; Kyng, K.; Staunstrup, N.H.; Børglum, A.; Henriksen, T.B. Antidepressant medication during pregnancy and epigenetic changes in umbilical cord blood: A systematic review. Clin. Epigenetics 2016, 8, 94. [Google Scholar] [CrossRef]
  36. Olstad, E.W.; Nordeng, H.M.E.; Gervin, K. Prenatal medication exposure and epigenetic outcomes: A systematic literature review and recommendations for prenatal pharmacoepigenetic studies. Epigenetics 2022, 17, 357–380. [Google Scholar] [CrossRef]
  37. Galbally, M.; Watson, S.J.; van IJzendoorn, M.; Saffery, R.; Ryan, J.; De Kloet, E.R.; Oberlander, T.F.; Lappas, M.; Lewis, A.J. The role of glucocorticoid and mineralocorticoid receptor DNA methylation in antenatal depression and infant stress regulation. Psychoneuroendocrinology 2020, 115, 104611. [Google Scholar] [CrossRef] [PubMed]
  38. Olstad, E.W.; Nordeng, H.M.E.; Sandve, G.K.; Lyle, R.; Gervin, K. Effects of prenatal exposure to (es)citalopram and maternal depression during pregnancy on DNA methylation and child neurodevelopment. Transl. Psychiatry 2023, 13, 149. [Google Scholar] [CrossRef] [PubMed]
  39. Hartwig, C.A.M.; Robiyanto, R.; de Vos, S.; Bos, J.H.J.; van Puijenbroek, E.P.; Hak, E.; Schuiling-Veninga, C.C.M. In utero antidepressant exposure not associated with ADHD in the offspring: A case control sibling design. Front. Pharmacol. 2022, 13, 1000018. [Google Scholar] [CrossRef] [PubMed]
  40. Suarez, E.A.; Bateman, B.T.; Hernández-Díaz, S.; Straub, L.; Wisner, K.L.; Gray, K.J.; Pennell, P.B.; Lester, B.; McDougle, C.J.; Zhu, Y.; et al. Association of Antidepressant Use During Pregnancy with Risk of Neurodevelopmental Disorders in Children. JAMA Intern. Med. 2022, 182, 1149–1160. [Google Scholar] [CrossRef]
  41. Li, X.-J.; Zhang, X.; Johnson, M.A.; Wang, Z.-B.; LaVaute, T.; Zhang, S.-C. Coordination of sonic hedgehog and Wnt signaling determines ventral and dorsal telencephalic neuron types from human embryonic stem cells. Development 2009, 136, 4055–4063. [Google Scholar] [CrossRef]
  42. Spildrejorde, M.; Leithaug, M.; Samara, A.; Aass, H.C.D.; Sharma, A.; Acharya, G.; Nordeng, H.; Gervin, K.; Lyle, R. Citalopram exposure of hESCs during neuronal differentiation identifies dysregulated genes involved in neurodevelopment and depression. Front. Cell Dev. Biol. 2024, 12, 1428538. [Google Scholar] [CrossRef]
  43. Brummelte, S.; Galea, L.A.M.; Devlin, A.M.; Oberlander, T.F. Antidepressant use during pregnancy and serotonin transporter genotype (SLC6A4) Affect newborn serum reelin levels. Dev. Psychobiol. 2013, 55, 518–529. [Google Scholar] [CrossRef]
  44. Pawluski, J.L.; Galea, L.A.M.; Brain, U.; Papsdorf, M.; Oberlander, T.F. Neonatal S100B Protein Levels After Prenatal Exposure to Selective Serotonin Reuptake Inhibitors. Pediatrics 2009, 124, e662–e670. [Google Scholar] [CrossRef]
  45. Itkonen, A.; Kärkkäinen, O.; Lehtonen, M.; Sahlman, H.; Keski-Nisula, L.; Rysä, J. Metabolomic Insights: LC-MS Profiling of Human Placental Tissue from SSRI-Treated Pregnancies. Eur. J. Med. Res. 2026. [Google Scholar] [CrossRef]
  46. Cooper, G.; North, R.; Hunt-Smith, T.; Larson, J.; Rennie, M.; Bailey, M.L.; Scarlata, S.; Merzdorf, C.S.; Bothner, B. Persistent Metabolic Changes Are Induced by 24 h Low-Dose Lead (Pb) Exposure in Zebrafish Embryos. Int. J. Mol. Sci. 2025, 26, 1050. [Google Scholar] [CrossRef]
  47. North, R.J.; Cooper, G.; Mears, L.; Bothner, B.; Dlakić, M.; Merzdorf, C.S. Persistent Transcriptome Alterations in Zebrafish Embryos After Discontinued Opioid Exposure. Int. J. Mol. Sci. 2025, 26, 4840. [Google Scholar] [CrossRef]
  48. North, R.J.; Cooper, G.; Mears, L.; Bothner, B.; Dlakić, M.; Merzdorf, C.S. Persistent and Dose-Dependent Neural and Metabolic Gene Expression Changes Induced by Transient Citalopram Exposure in Zebrafish Embryos. Int. J. Mol. Sci. 2025, 26, 11288. [Google Scholar] [CrossRef]
  49. Hoage, T.; Ding, Y.; Xu, X. Quantifying cardiac functions in embryonic and adult zebrafish. Methods Mol. Biol. 2012, 843, 11–20. [Google Scholar] [CrossRef]
  50. Schneider, C.A.; Rasband, W.S.; Eliceiri, K.W. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 2012, 9, 671–675. [Google Scholar] [CrossRef]
  51. Szymańska, E.; Saccenti, E.; Smilde, A.K.; Westerhuis, J.A. Double-check: Validation of diagnostic statistics for PLS-DA models in metabolomics studies. Metabolomics 2012, 8, 3–16. [Google Scholar] [CrossRef]
  52. Ramirez, T.; Daneshian, M.; Kamp, H.; Bois, F.Y.; Clench, M.R.; Coen, M.; Donley, B.; Fischer, S.M.; Ekman, D.R.; Fabian, E.; et al. Metabolomics in toxicology and preclinical research. Altex 2013, 30, 209–225. [Google Scholar] [CrossRef] [PubMed]
  53. Liu, K.-Q.; Liu, Z.-P.; Hao, J.-K.; Chen, L.; Zhao, X.-M. Identifying dysregulated pathways in cancers from pathway interaction networks. BMC Bioinform. 2012, 13, 126. [Google Scholar] [CrossRef] [PubMed]
  54. Yang, H.; Tan, J.; Jiang, H.; Xing, H.; Zhang, J.; Kong, D.; Chen, Z.; Kong, L. The Effect of the Antidepressant Citalopram on the Bioconcentration and Biomarker Response of Daphnia magna at Environmentally Relevant Concentrations. Toxics 2025, 13, 532. [Google Scholar] [CrossRef] [PubMed]
  55. Kuti, D.; Winkler, Z.; Horváth, K.; Juhász, B.; Szilvásy-Szabó, A.; Fekete, C.; Ferenczi, S.; Kovács, K.J. The metabolic stress response: Adaptation to acute-, repeated- and chronic challenges in mice. iScience 2022, 25, 104693. [Google Scholar] [CrossRef]
  56. Ma, S.; Kosorok, M.R. Identification of differential gene pathways with principal component analysis. Bioinformatics 2009, 25, 882–889. [Google Scholar] [CrossRef]
  57. Lever, J.; Krzywinski, M.; Altman, N. Principal component analysis. Nat. Methods 2017, 14, 641–642. [Google Scholar] [CrossRef]
  58. Choi, C.H.; Zimon, A.; Usheva, A. Metabolic stress regulates basic transcription through acetyl-coenzyme A. Cell. Mol. Life Sci. 2005, 62, 625–628. [Google Scholar] [CrossRef]
  59. Patil, K.R.; Nielsen, J. Uncovering transcriptional regulation of metabolism by using metabolic network topology. Proc. Natl. Acad. Sci. USA 2005, 102, 2685–2689. [Google Scholar] [CrossRef] [PubMed]
  60. Desvergne, B.; Michalik, L.; Wahli, W. Transcriptional Regulation of Metabolism. Physiol. Rev. 2006, 86, 465–514. [Google Scholar] [CrossRef] [PubMed]
  61. Hornisch, M.; Piazza, I. Regulation of gene expression through protein-metabolite interactions. npj Metab. Health Dis. 2025, 3, 7. [Google Scholar] [CrossRef]
  62. Jozefczuk, S.; Klie, S.; Catchpole, G.; Szymanski, J.; Cuadros-Inostroza, A.; Steinhauser, D.; Selbig, J.; Willmitzer, L. Metabolomic and transcriptomic stress response of Escherichia coli. Mol. Syst. Biol. 2010, 6, 364. [Google Scholar] [CrossRef] [PubMed]
  63. Lai, W.Y.; Otte, K.A.; Schlötterer, C. Evolution of Metabolome and Transcriptome Supports a Hierarchical Organization of Adaptive Traits. Genome Biol. Evol. 2023, 15, evad098. [Google Scholar] [CrossRef]
  64. Rawle, R.A.; Hamerly, T.; Tripet, B.P.; Giannone, R.J.; Wurch, L.; Hettich, R.L.; Podar, M.; Copié, V.; Bothner, B. Multi-omics analysis provides insight to the Ignicoccus hospitalis-Nanoarchaeum equitans association. Biochim. Biophys. Acta Gen. Subj. 2017, 1861, 2218–2227. [Google Scholar] [CrossRef]
  65. Maaty, W.S.; Steffens, J.; Heinemann, J.; Ortmann, A.C.; Reeves, B.D.; Biswas, S.K.; Dratz, E.A.; Grieco, P.A.; Young, M.J.; Bothner, B. Global Analysis of Viral Infection in an Archaeal Model System. Front. Microbiol. 2012, 3, 411. [Google Scholar] [CrossRef]
  66. Johnson, C.H.; Ivanisevic, J.; Siuzdak, G. Metabolomics: Beyond biomarkers and towards mechanisms. Nat. Rev. Mol. Cell Biol. 2016, 17, 451–459. [Google Scholar] [CrossRef] [PubMed]
  67. Metallo, C.M.; Vander Heiden, M.G. Understanding metabolic regulation and its influence on cell physiology. Mol. Cell 2013, 49, 388–398. [Google Scholar] [CrossRef]
  68. Sørensen, A.; Ruhé, H.G.; Munkholm, K. The relationship between dose and serotonin transporter occupancy of antidepressants—A systematic review. Mol. Psychiatry 2022, 27, 192–201. [Google Scholar] [CrossRef] [PubMed]
  69. Gradisch, R.; Schlögl, K.; Lazzarin, E.; Niello, M.; Maier, J.; Mayer, F.P.; Alves da Silva, L.; Skopec, S.M.C.; Blakely, R.D.; Sitte, H.H.; et al. Ligand coupling mechanism of the human serotonin transporter differentiates substrates from inhibitors. Nat. Commun. 2024, 15, 417. [Google Scholar] [CrossRef] [PubMed]
  70. Schrantee, A.; Solleveld, M.M.; Schwantje, H.; Bruin, W.B.; Mutsaerts, H.-J.M.M.; Adriaanse, S.M.; Lucassen, P.; Booij, J.; Reneman, L. Dose-dependent effects of the selective serotonin reuptake inhibitor citalopram: A combined SPECT and phMRI study. J. Psychopharmacol. 2019, 33, 660–669. [Google Scholar] [CrossRef]
  71. Hagan, C.E.; McDevitt, R.A.; Liu, Y.; Furay, A.R.; Neumaier, J.F. 5-HT(1B) autoreceptor regulation of serotonin transporter activity in synaptosomes. Synapse 2012, 66, 1024–1034. [Google Scholar] [CrossRef]
  72. Hoyer, D.; Middlemiss, D.N. Species differences in the pharmacology of terminal 5-HT autoreceptors in mammalian brain. Trends Pharmacol. Sci. 1989, 10, 130–132. [Google Scholar] [CrossRef]
  73. Hjorth, S.; Suchowski, C.S.; Galloway, M.P. Evidence for 5-HT autoreceptor-mediated, nerve impulse-independent, control of 5-HT synthesis in the rat brain. Synapse 1995, 19, 170–176. [Google Scholar] [CrossRef]
  74. Raote, I.; Bhattacharya, A.; Panicker, M.M. Serotonin 2A (5-HT(2A)) Receptor Function: Ligand-Dependent Mechanisms and Pathways. In Serotonin Receptors in Neurobiology; Frontiers in Neuroscience; Chattopadhyay, A., Ed.; CRC Press: Boca Raton, FL, USA; Taylor & Francis Group, LLC.: Boca Raton, FL, USA, 2007. [Google Scholar]
  75. Voronezhskaya, E.E. Serotonin as a volume transmission signal in the “simple nervous system” of mollusks: From axonal guidance to behavioral orchestration. Front. Synaptic Neurosci. 2022, 14, 1024778. [Google Scholar] [CrossRef]
  76. Suryanto, M.E.; Audira, G.; Uapipatanakul, B.; Hussain, A.; Saputra, F.; Siregar, P.; Chen, K.H.C.; Hsiao, C.-D. Antidepressant Screening Demonstrated Non-Monotonic Responses to Amitriptyline, Amoxapine and Sertraline in Locomotor Activity Assay in Larval Zebrafish. Cells 2021, 10, 738. [Google Scholar] [CrossRef]
  77. Mombereau, C.; Gur, T.L.; Onksen, J.; Blendy, J.A. Differential effects of acute and repeated citalopram in mouse models of anxiety and depression. Int. J. Neuropsychopharmacol. 2010, 13, 321–334. [Google Scholar] [CrossRef]
  78. Koek, W.; Mitchell, N.C.; Daws, L.C. Biphasic effects of selective serotonin reuptake inhibitors on anxiety: Rapid reversal of escitalopram’s anxiogenic effects in the novelty-induced hypophagia test in mice? Behav. Pharmacol. 2018, 29, 365–369. [Google Scholar] [CrossRef] [PubMed]
  79. Morris, S.M. Arginine Metabolism: Boundaries of Our Knowledge123. J. Nutr. 2007, 137, 1602S–1609S. [Google Scholar] [CrossRef]
  80. Tagliaferro, P.; Ramos, A.J.; Lopez-Costa, J.J.; López, E.M.; Brusco, A. Changes in the postnatal development on nitric oxide system induced by serotonin depletion. Dev. Brain Res. 2003, 146, 39–49. [Google Scholar] [CrossRef]
  81. Bird, I.M.; Zhang, L.; Magness, R.R. Possible mechanisms underlying pregnancy-induced changes in uterine artery endothelial function. Am. J. Physiol.-Regul. Integr. Comp. Physiol. 2003, 284, R245–R258. [Google Scholar] [CrossRef]
  82. Dawson, V.L. Nitric oxide: Role in neurotoxicity. Clin. Exp. Pharmacol. Physiol. 1995, 22, 305–308. [Google Scholar] [CrossRef] [PubMed]
  83. Kaehler, S.T.; Singewald, N.; Sinner, C.; Philippu, A. Nitric oxide modulates the release of serotonin in the rat hypothalamus. Brain Res. 1999, 835, 346–349. [Google Scholar] [CrossRef] [PubMed]
  84. Bahremand, A.; Payandemehr, B.; Rahimian, R.; Ziai, P.; Pourmand, N.; Loloee, S.; Ebrahimi, A.; Ghasemi, A.; Fakhfouri, G.; Ghasemi, M.; et al. The role of 5-HT3 receptors in the additive anticonvulsant effects of citalopram and morphine on pentylenetetrazole-induced clonic seizures in mice. Epilepsy Behav. 2011, 21, 122–127. [Google Scholar] [CrossRef]
  85. Hussain, M.M.; Jin, W.; Jiang, X.C. Mechanisms involved in cellular ceramide homeostasis. Nutr. Metab. 2012, 9, 71. [Google Scholar] [CrossRef]
  86. Igarashi, K.; Kashiwagi, K. Polyamines: Mysterious modulators of cellular functions. Biochem. Biophys. Res. Commun. 2000, 271, 559–564. [Google Scholar] [CrossRef]
  87. Watts, D.; Pfaffenseller, B.; Wollenhaupt-Aguiar, B.; Paul Géa, L.; Cardoso, T.D.A.; Kapczinski, F. Agmatine as a potential therapeutic intervention in bipolar depression: The preclinical landscape. Expert. Opin. Ther. Targets 2019, 23, 327–339. [Google Scholar] [CrossRef] [PubMed]
  88. Liu, J.H.; Wang, W.; Wu, H.; Gong, X.; Moriguchi, T. Polyamines function in stress tolerance: From synthesis to regulation. Front. Plant Sci. 2015, 6, 827. [Google Scholar] [CrossRef] [PubMed]
  89. MahmoudianDehkordi, S.; Ahmed, A.T.; Bhattacharyya, S.; Han, X.; Baillie, R.A.; Arnold, M.; Skime, M.K.; John-Williams, L.S.; Moseley, M.A.; Thompson, J.W.; et al. Alterations in acylcarnitines, amines, and lipids inform about the mechanism of action of citalopram/escitalopram in major depression. Transl. Psychiatry 2021, 11, 153. [Google Scholar] [CrossRef] [PubMed]
  90. Gilad, G.M.; Gilad, V.H.; Eliyayev, Y.; Rabey, J.M. Developmental regulation of the brain polyamine-stress-response. Int. J. Dev. Neurosci. 1998, 16, 271–278. [Google Scholar] [CrossRef]
  91. Gilad, G.M.; Gilad, V.H. Overview of the Brain Polyamine-Stress-Response: Regulation, Development, and Modulation by Lithium and Role in Cell Survival. Cell. Mol. Neurobiol. 2003, 23, 637–649. [Google Scholar] [CrossRef]
Biomedicines 14 00381 g001
Figure 1. Discontinued citalopram exposure. Zebrafish embryos were collected at 0–2 h post-fertilization (hpf). Embryos were exposed to 0.03 ng/mL (environmental surface water concentration), 0.9 ng/mL (wastewater treatment plant effluent), 50 ng/mL (therapeutic human serum concentration), or 250 ng/mL citalopram hydrobromide (supertherapeutic human serum concentration) from 2 to 24 hpf. Embryos were washed at 24 hpf and allowed to develop without citalopram until 48 hpf. At 48 hpf, embryos were collected for metabolomic analysis. Physiological measurements were taken at 24 hpf, 48 hpf, and 72 hpf.
Figure 1. Discontinued citalopram exposure. Zebrafish embryos were collected at 0–2 h post-fertilization (hpf). Embryos were exposed to 0.03 ng/mL (environmental surface water concentration), 0.9 ng/mL (wastewater treatment plant effluent), 50 ng/mL (therapeutic human serum concentration), or 250 ng/mL citalopram hydrobromide (supertherapeutic human serum concentration) from 2 to 24 hpf. Embryos were washed at 24 hpf and allowed to develop without citalopram until 48 hpf. At 48 hpf, embryos were collected for metabolomic analysis. Physiological measurements were taken at 24 hpf, 48 hpf, and 72 hpf.
Biomedicines 14 00381 g001
Biomedicines 14 00381 g002
Figure 2. Morphological data of citalopram-exposed and control embryos. (A) Violin plots displaying heart rate (bpm) of control, 0.03 ng/mL, 0.9 ng/mL, 50 ng/mL, and 250 ng/mL of citalopram bromide at 24 hpf, 48 hpf, and 72 hpf. ANOVA shows significant differences between heart rates for control and citalopram-exposed embryos at 24 hpf (n = 5) at 48 hpf (n = 5), and at 72 hpf (n = 5). (B) Violin plots showing embryo length (mm) of the control and citalopram-exposed embryos at 72 hpf. No significant differences are observed by ANOVA.
Figure 2. Morphological data of citalopram-exposed and control embryos. (A) Violin plots displaying heart rate (bpm) of control, 0.03 ng/mL, 0.9 ng/mL, 50 ng/mL, and 250 ng/mL of citalopram bromide at 24 hpf, 48 hpf, and 72 hpf. ANOVA shows significant differences between heart rates for control and citalopram-exposed embryos at 24 hpf (n = 5) at 48 hpf (n = 5), and at 72 hpf (n = 5). (B) Violin plots showing embryo length (mm) of the control and citalopram-exposed embryos at 72 hpf. No significant differences are observed by ANOVA.
Biomedicines 14 00381 g002
Biomedicines 14 00381 g003
Figure 3. Global multivariate statistics of LC-MS data. (A) One-way ANOVA of 4813 filtered metabolomics features shows that 88 change significantly (p-value < 0.05, red, yellow, and orange). (B) Partial least squares-discriminant analysis (PLS-DA) scores plot of metabolomics data (Orange = Control, Green = 0.03 ng/mL, Teal = 0.9 ng/mL, Purple = 50 ng/mL, Blue = 250 ng/mL). 95% confidence intervals are indicated by ovals. Plot describes PC1 (27.5% of total variance) and PC2 (13.6% of total variance) coordinate pairs from each sample. (C) Heatmap of the top 25 features, filtered by lowest p-value. Samples are clustered in columns, features are clustered in rows. Yellow reflects high abundance and purple reflects low abundance. Metabolites with high confidence identification are labeled. Metabolomic features with tentative, or no ID, are presented as mass-to-charge ratio and retention time (mz_rt).
Figure 3. Global multivariate statistics of LC-MS data. (A) One-way ANOVA of 4813 filtered metabolomics features shows that 88 change significantly (p-value < 0.05, red, yellow, and orange). (B) Partial least squares-discriminant analysis (PLS-DA) scores plot of metabolomics data (Orange = Control, Green = 0.03 ng/mL, Teal = 0.9 ng/mL, Purple = 50 ng/mL, Blue = 250 ng/mL). 95% confidence intervals are indicated by ovals. Plot describes PC1 (27.5% of total variance) and PC2 (13.6% of total variance) coordinate pairs from each sample. (C) Heatmap of the top 25 features, filtered by lowest p-value. Samples are clustered in columns, features are clustered in rows. Yellow reflects high abundance and purple reflects low abundance. Metabolites with high confidence identification are labeled. Metabolomic features with tentative, or no ID, are presented as mass-to-charge ratio and retention time (mz_rt).
Biomedicines 14 00381 g003
Biomedicines 14 00381 g004
Figure 4. Pairwise statistics of control versus citalopram-exposed embryos. (A) PCA plot of control (orange) and 0.03 ng/mL citalopram-exposed embryos. PC1 accounts for 38% of the total variation and PC2 covers an additional 21.1% of the total variation. The 95% confidence intervals are indicated by ovals. (B) PCA plot of control and 0.9 ng/mL (teal) citalopram-exposed embryos. PC1 and PC2 explain 38.4% and 15% of the total variation, respectively. (C) PCA plot of control and 50 ng/mL (purple) citalopram-exposed embryos. PC1 and PC2 explain 33.2% and 20.1% of the total variation, respectively. (D) PCA plot of the control and 250 ng/mL (blue) citalopram-exposed embryos. PC1 and PC2 explain 41.1% and 14.9% of the total variation, respectively. (E) Histogram showing the number of up and downregulated features in the citalopram exposures.
Figure 4. Pairwise statistics of control versus citalopram-exposed embryos. (A) PCA plot of control (orange) and 0.03 ng/mL citalopram-exposed embryos. PC1 accounts for 38% of the total variation and PC2 covers an additional 21.1% of the total variation. The 95% confidence intervals are indicated by ovals. (B) PCA plot of control and 0.9 ng/mL (teal) citalopram-exposed embryos. PC1 and PC2 explain 38.4% and 15% of the total variation, respectively. (C) PCA plot of control and 50 ng/mL (purple) citalopram-exposed embryos. PC1 and PC2 explain 33.2% and 20.1% of the total variation, respectively. (D) PCA plot of the control and 250 ng/mL (blue) citalopram-exposed embryos. PC1 and PC2 explain 41.1% and 14.9% of the total variation, respectively. (E) Histogram showing the number of up and downregulated features in the citalopram exposures.
Biomedicines 14 00381 g004
Biomedicines 14 00381 g005
Figure 5. Significantly dysregulated pathways. (A) UpSet plot comparing all the significantly dysregulated pathways for each citalopram exposure condition. Vertical blue bars indicate the total number of pathways in the exposure condition defined by the black dot present below. Single dots represent the number of unique pathways to an exposure group. Black lines connecting the dots reflect shared dysregulated pathways. Set size refers to the scale bar reflecting how many pathways are present. Only one pathway, arginine and proline metabolism, was shared by all four citalopram exposure groups. (B) Violin plots of metabolites (p < 0.05) in the arginine and proline metabolism pathway. Violin plots show the normalized abundance across control (orange), 0.03 ng/mL (green), 0.9 ng/mL (teal), 50 ng/mL (purple), and 250 ng/mL (blue). Reported p values from ANOVA.
Figure 5. Significantly dysregulated pathways. (A) UpSet plot comparing all the significantly dysregulated pathways for each citalopram exposure condition. Vertical blue bars indicate the total number of pathways in the exposure condition defined by the black dot present below. Single dots represent the number of unique pathways to an exposure group. Black lines connecting the dots reflect shared dysregulated pathways. Set size refers to the scale bar reflecting how many pathways are present. Only one pathway, arginine and proline metabolism, was shared by all four citalopram exposure groups. (B) Violin plots of metabolites (p < 0.05) in the arginine and proline metabolism pathway. Violin plots show the normalized abundance across control (orange), 0.03 ng/mL (green), 0.9 ng/mL (teal), 50 ng/mL (purple), and 250 ng/mL (blue). Reported p values from ANOVA.
Biomedicines 14 00381 g005
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Cooper, G.; North, R.J.; Auganova, D.; Merzdorf, C.S.; Bothner, B. Differential Metabolic Changes in Zebrafish Embryos Are Induced by Discontinued Citalopram Exposure. Biomedicines 2026, 14, 381. https://doi.org/10.3390/biomedicines14020381

AMA Style

Cooper G, North RJ, Auganova D, Merzdorf CS, Bothner B. Differential Metabolic Changes in Zebrafish Embryos Are Induced by Discontinued Citalopram Exposure. Biomedicines. 2026; 14(2):381. https://doi.org/10.3390/biomedicines14020381

Chicago/Turabian Style

Cooper, Gwendolyn, Ryan J. North, Dana Auganova, Christa S. Merzdorf, and Brian Bothner. 2026. "Differential Metabolic Changes in Zebrafish Embryos Are Induced by Discontinued Citalopram Exposure" Biomedicines 14, no. 2: 381. https://doi.org/10.3390/biomedicines14020381

APA Style

Cooper, G., North, R. J., Auganova, D., Merzdorf, C. S., & Bothner, B. (2026). Differential Metabolic Changes in Zebrafish Embryos Are Induced by Discontinued Citalopram Exposure. Biomedicines, 14(2), 381. https://doi.org/10.3390/biomedicines14020381

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop