Introduction

Zebrafish are a widely used research model in many domains of biology. Characterized by easy maintenance, high fecundity, capacity for high-throughput screening, rapid development of translucent embryos, and genetic tractability, zebrafish have been a research model in developmental biology and genetics for more than 40 years1. Due to their versatility, it is no surprise that zebrafish have become a central model to many more disciplines, including, but not limited to, fish biology and (comparative) physiology2,3, developmental biology and genetics4, evolution5, biomedical research6,7 and, increasingly, (eco)toxicology5,6,7. High-throughput zebrafish screening applications in pharmacology3 and toxicology8,9 have experienced rapid development in recent years. This is, at least in part, linked to the fact that zebrafish during early life stages prior to exogenous feeding (0–5 days post-fertilization (dpf)) are exempt from vertebrate model animal ethics regulations in some legislations8,10,11,12. As such, they represent one of the only whole animal vertebrate models suitable to reduce, refine, and replace the use of traditional animal use in modern toxicology testing mandated by the National Institutes of Health (NIH), United States Department of Agriculture (USDA), and the Food and Drug Administration (FDA)8,12.

Compared to long-standing mammalian research models, several aspects of zebrafish husbandry remain less characterized. As early life stages of zebrafish are increasingly utilized for high-throughput pharmaceutical screening13,14 and as new approach methodology (NAM) for regulatory toxicology testing9, efforts to standardize guidelines for zebrafish rearing protocols have become a focus in the field, especially amidst concerns surrounding reproducibility in science15. Originally developed as a dye in the late 1800s16,17, methylene blue (MB) is widely recommended for use in zebrafish husbandry as a component of zebrafish embryo medium due to its antifungal activity. However, guidelines regarding its usage vary across different scientific organizations. For instance, the Canadian Council of Animal Care (CCAC) guidelines for zebrafish currently recommend the use of 0.00005% MB in E3 medium10. Similarly, a review of the European Union (EU) guidelines suggests the same concentration for embryos prior to exogenous feeding and transfer to a larger housing system18. Along the same lines, the United States Cold Spring Harbor (CSH) Protocols recommends the use of MB, albeit at a two-fold higher concentration of 0.0001% MB19. On the contrary, neither the Organization for Economic Co-operation and Development (OECD) fish acute embryo toxicity test guideline (#236)20, the 8th edition of the Guide for the Care and Use of Laboratory Animals by the National Research Council21, nor the 4th edition of the Zebrafish Book22 mention the use of MB. The lack of consensus leads to inconsistencies in its application and reporting among zebrafish researchers.

Interestingly, in addition to its disinfectant, antibacterial, and antifungal properties in aquatic husbandry23, metabolic effects of MB have been reported for almost 100 years24,25. For example, increased oxygen consumption in tumor tissues and erythrocytes in vitro has been reported following methylene blue application at a concentration of 0.0005% as early as the 1930s24,25. MB has been applied as medical treatment and figures on the World Health Organization’s (WHO) list of essential medicines26. This is due to MB’s low cost of synthesis, and action as a reducing agent. For example, MB has been used to treat toxicant-induced methemoglobinemia through its ability to reduce methemoglobin to regenerate hemoglobin27. Similarly, MB can act on components of the mitochondrial electron transport chain (ETC)28,29,30,31,32,33, and has been used to treat cyanide poisoning characterized by inhibition of the electron transport chain via non-competitive binding to complex IV34,35. MB is, due to its described action on mitochondria, furthermore investigated as a potential treatment for aging and/or neurodegeneration36,37,38,39,40,41,42, and has been shown to mitigate pesticide effects on mitochondria in wildlife, such as bumblebees43. Despite the long-known metabolic effects of MB, it remains unknown whether currently recommended MB concentrations affect early zebrafish metabolism, and thus may act as a widespread confound in experimental approaches using early zebrafish. Indeed, possible effects of MB on zebrafish (eleuthero)embryos have only very recently begun to be investigated44,45,46,47,48. The few studies published to date focused on overt developmental and behavioral endpoints but yielded conflicting results. While one study reported no effects of MB exposure at concentrations between 0.0003–0.003% on morphology and locomotor behavior45, others reported reduced locomotor behavior48, and a reduction of the acoustic startle response and associated kinematic responses following MB exposures47. Given the MB’s long-described metabolic mode of action in other models, we here test the hypothesis that currently recommended MB concentrations exert repeatable and significant effects on early-developing zebrafish metabolism and energy balance. To test this hypothesis, we use described zebrafish-adapted metabolic assays (Seahorse assay49; Alamar Blue assay50) at the organismal level, and targeted metabolic gene expression analysis at the molecular level. Because zebrafish adapted medium-high-throughput metabolic assays, such as the Alamar Blue assay50, are based on oxidation-reduction dependent spectrophotometry50, we secondarily also test the hypothesis that MB at recommended concentrations, can directly interfere with the Alamar Blue zebrafish-specific metabolic assays.

Results

Methylene blue increases basal OCR and ΔΨm in 1 and 2 dpf zebrafish

Developmental exposure to MB at concentrations recommended by the CCAC and the Federation of European Laboratory Animal Science Associations (FELASA) (0.00005%), as well as the CSH Protocols (0.0001%) were assessed in a modified Seahorse assay for zebrafish embryos (Fig. 1A–C). Both MB concentrations significantly increased basal OCR in 1 dpf zebrafish embryos compared to control (Fig. 1D). For the CCAC (0.00005%) recommended MB concentration, this OCR increase was coupled to increased ATP production compared to controls (Fig. 1D), while the CSH (0.0001%) recommended MB concentration had a significant reduction in spare respiratory capacity compared to the control (Fig. 1D). Furthermore, in control larvae, a significantly positive relationship between basal and maximal OCR was found, which was no longer present in MB-exposed embryos (Fig. 1E). Significant positive relationships between basal respiration and ATP-linked OCR were found in controls and embryos exposed to CCAC (0.00005%) -recommended, but not CSH (0.0001%)-recommended MB concentrations (Fig. 1E). In a replicate experiment, MB exposure at CCAC-recommended concentrations confirmed an increase in OCR and reduction in spare capacity at 1 dpf (Supplementary Fig. 2A). In line with the increase in OCR, zebrafish embryos at 2 dpf show a significant and repeatable increase in larval surface layer ΔΨm after exposure to CCAC (0.00005%), CSH (0.0001%) and 0.0003% MB concentrations compared to the control (Fig. 1F and Supplementary Fig. 3).

Fig. 1: Methylene blue increases basal OCR and mitochondria membrane potential (ΔΨm) in 1 and 2 dpf zebrafish embryos.
Fig. 1: Methylene blue increases basal OCR and mitochondria membrane potential (ΔΨm) in 1 and 2 dpf zebrafish embryos.
Full size image

A Schematic representation detailing the experimental design. B Table showing the concentrations of MB with references. C Scheme of the MITO stress test OCR measurements. D OCR measurements of 1 dpf zebrafish embryo after MB exposure. n = 6–7 (Each biological replicate is represented by a single embryo). E Plots investigating relationships between basal respiration and different MITO stress test endpoints in control and different MB exposure concentrations. R2 and P values are provided. F The ratio of green/red (TRITC/GFP) fluorescent intensity and representative images of 2 dpf zebrafish embryos stained with JC-10 fluorescent stain after exposure to multiple concentrations of methylene blue. n = 4–5 (Each biological replicate is represented by a single embryo). The data were graphed as violin plots showing the mean (symbol: +), the median represented by a line in the middle, the upper and lower quartile range indicated by the dotted lines, and the outline of the violin plot illustrates the kernel probability density. Asterisks indicate *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. (Magnification 10x; scale bar, 300 μm).

Methylene blue decreases energy expenditure in post-hatched zebrafish

Following the observed changes in basal OCR in 1 dpf zebrafish, we conducted an indirect zebrafish energy expenditure assay using Alamar Blue at different stages of zebrafish development (Fig. 2). All embryos/larvae were washed three times prior to incubation with Alamar Blue reagent. While oxidative metabolism-based energy expenditure was not affected in CCAC (0.00005%), CSH (0.0001%), and a higher concentration recommended by Chambel et al. (0.0003%)51 when compared to controls after the first 48 h of exposure, the highest tested concentration of MB (0.001%) significantly decreased oxidative metabolism-based energy expenditure irrespective of incubation time (3 h P = 0.0003; 6 h P < 0.0001; 12 h P < 0.0001; 24 h P < 0.0001; 48 h P = 0.0009; Fig. 2B). Under standard embryo/larval zebrafish rearing protocols, zebrafish are kept in embryo/E3 media containing MB until 5 dpf before being transferred to system water. We, therefore, tested to see if longer exposure to MB exerts significant effects on zebrafish oxidative metabolism-dependent energy expenditure (Fig. 2C). Zebrafish exposed to recommended and higher concentrations of MB significantly and consistently exhibited decreased oxidative metabolism-related energy expenditure compared to controls measured at 24 h (one-way ANOVA; P < 0.0001) and 48 h (one-way ANOVA; P < 0.0001) (Fig. 2D). To determine whether these reductions in oxidative metabolism-dependent energy expenditure observed at 5 dpf are dependent on a preceding continuous 4-day steady state renewal exposure window or can be elicited by acute post-hatching exposure alone, we conducted a 1 day acute and short-term MB exposure at 3 dpf and quantified oxidative metabolism-related energy expenditure quantified at 4 dpf (Fig. 2E). Zebrafish acutely exposed to MB from 3–4 dpf zebrafish also exhibited a significant decrease in Alamar Blue fluorescence at 24 h (one-way ANOVA; P < 0.0001).

Fig. 2: Energy expenditure decreases after Methylene blue exposure in zebrafish larvae.
Fig. 2: Energy expenditure decreases after Methylene blue exposure in zebrafish larvae.
Full size image

Schematic representation detailing the experimental design for A 0–1 dpf, C 0–4 dpf, and E transient 3–4 dpf MB exposure. B Energy expenditure was measured using the Alamar Blue assay in 1 dpf embryos at 3, 6, 12, and 24 h. D Energy expenditure was measured using the Alamar Blue assay in 4 dpf larvae at 24 and 48 h. F Energy expenditure was measured using the Alamar Blue assay in 4 dpf larvae at 24 after transient 3–4 dpf MB exposure. The data were graphed as violin plots showing the mean (symbol: +), the median represented by a line in the middle, the upper and lower quartile range indicated by the dotted lines, and the outline of the violin plot illustrates the kernel probability density. n = 10–12 (Each biological replicate is represented by a single embryo), Asterisk indicates *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Methylene blue downregulates multiple mitochondrial respiratory complex gene expression in 1 and 4 dpf zebrafish larvae

To assess the effects of developmental MB exposure on mitochondrial respiratory complex gene expression at CCAC-recommended concentrations (0.00005%), 1 and 4 dpf zebrafish were sampled (Fig. 3A). At 1 dpf, the MB-exposed group exhibited a significant decrease in the relative abundance of mitochondrial respiratory complex transcripts sdhda (P = 0.0032), sdhdb (P = 0.0002), and cox4i1 (P = 0.0077) when compared to controls (Fig. 3B). Similarly, the 4 dpf MB-exposed zebrafish larvae exhibited a significant decrease in the relative abundance of transcripts of mitochondrial complex genes ndufa9a (P = 0.0051), sdhdb (P = 0.0023), uqcrc2a (P = 0.0202) and atp5f1e (P = 0.0355) (Fig. 3C).

Fig. 3: Methylene blue downregulates transcript abundance of mitochondrial respiratory complex genes in 1 dpf and 4 dpf zebrafish embryos.
Fig. 3: Methylene blue downregulates transcript abundance of mitochondrial respiratory complex genes in 1 dpf and 4 dpf zebrafish embryos.
Full size image

A Schematic representation detailing the experimental design. B Gene expression of 1 dpf embryos after MB exposure. C Gene expression of 4 dpf larvae after MB exposure. The data were graphed as violin plots showing the mean (symbol: +), the median represented by a line in the middle, the upper and lower quartile range indicated by the dotted lines, and the outline of the violin plot illustrates the kernel probability density. n = 8–10 (Each biological replicate is represented by 15 pooled embryos or 5 pooled larvae), Asterisk indicates *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. D Mode-of action of MB on the mitochondrial electron transport chain. E Images of 5 dpf zebrafish after continuous methylene blue exposure (0–5 dpf).

Developmental MB exposure advances hatching time but does not impact the stress axis

Developmental exposure of zebrafish embryos to MB at the CSH-recommended concentration (0.0001%) induced early hatching at 2 dpf (Fig. 4B, C). Furthermore, because we observed a consistent reduction in relative transcript abundance of genes linked to the endocrine stress axis function, including star, the rate-limiting mitochondrial cholesterol import protein important in regulating steroidogenesis (1 dpf P < 0.0001; 4 dpf P = 0.0007), and pck1, a known cortisol-responsive gluconeogenesis enzyme (1 dpf P = 0.0099; 4 dpf P = 0.0364), at 1 and 4 dpf exposed to 0.00005% MB (Fig. 4G, H), we directly assessed whether developmental MB exposure disrupts the endocrine stress. To this end, we subjected zebrafish larvae exposed to the CCAC-recommended concentration of MB (0.00005%) to a swirling stressor challenge at 5 dpf (Fig. 4D) and quantified whole-body cortisol concentrations over time. While a significant effect of post-stressor time on cortisol levels was observed (P < 0.0001; F3,92 = 41.77), no main MB- treatment or MB-treatment effect over time (interaction) were found (Fig. 4E, F).

Fig. 4: Methylene blue induces early hatching in zebrafish embryos.
Fig. 4: Methylene blue induces early hatching in zebrafish embryos.
Full size image

Schematic representation detailing the experimental design for A hatching rate and D larval stress response. B, C Hatching rate of zebrafish embryos at 2 and 3 dpf. The graphs represent Mean ± SEM. n = 5 (One replicate represents 44–48 embryos/larvae in a 48-well microplate), Bars with different letters are significantly different (P ≤ 0.05). E, F The cortisol stress response in 5 dpf larvae after a swirling stressor where larvae were sampled at 0 (unstressed), 5-, 10-, and 60-min post-stressor. The line graph represents mean and individual data points, whereas the violin plots represent the mean (symbol: +), the median represented by a line in the middle, the upper and lower quartile range indicated by the dotted lines, and the outline of the violin plot illustrates the kernel probability density. n = 5–23 (Each biological replicate is represented by 10 pooled larvae) Plots with different letters are significantly different (P ≤ 0.05). G Gene expression of 1 dpf embryos after MB exposure. H Gene expression of 4 dpf larvae after MB exposure. The data were graphed as violin plots showing the mean (symbol: +), the median represented by a line in the middle, the upper and lower quartile range indicated by the dotted lines, and the outline of the violin plot illustrates the kernel probability density. n = 8–10 (Each biological replicate is represented by 15 pooled embryos or 5 pooled larvae), Asterisk indicates *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Discussion

We here show that developmental exposure to MB, at concentrations and durations widely recommended in zebrafish husbandry guidelines and used in zebrafish research, causes acute and robust organismal metabolic changes in developing zebrafish embryos. These changes coincide with in vivo increases in mitochondrial membrane potential, and significant reduction of key transcripts associated with the mitochondrial respiratory complex and cytoplasmic metabolic pathways. Together, these findings show that MB affects oxidative metabolism in zebrafish embryos consistent with the compound’s reported mode-of-action in other in vitro and in vivo models. We furthermore report that MB, as a redox agent, has the potential to confound an increasingly used high-throughput zebrafish metabolic phenotyping assay, the Alamar Blue assay. Both findings have wider implications for zebrafish research.

Developmental exposure of zebrafish embryos to recommended MB concentrations resulted in visible blue stains in the larval head, mouth, olfactory epithelial area, and, at higher concentrations, throughout the whole body (Fig. 3E). Indeed, studies have reported that zebrafish embryos exposed to MB across multiple developmental stages revealed a dose- and time-dependent increase in MB uptake: initial uptake resulted in low embryo concentrations as early as 3 hpf, which increases to embryo concentrations similar to medium by 15 hpf52. This is in line with the high biodistribution reported for MB in other in vivo models, indicating its mobility across cell membranes53,54, and a reported plasma half-life of 5–6.5 h in rats55. Taken together, the chorion does not seem to act as an effective barrier to MB, but additional routes other than diffusion likely differentially affect and possibly accelerate the uptake kinetics of MB after hatching.

In line with uptake through the chorion, we observed repeatable MB-dependent increases in oxygen consumption rate (OCR) in unhatched 1 dpf embryos, which were linked to ATP generation. An in vivo mitochondrial mode of action of MB exposure at recommended husbandry concentrations was further corroborated by an increase in mitochondrial membrane potential (ΔΨm), assessed using a waterborne JC-10 probe at 2 dpf56. Because a higher ΔΨm is associated with an increased efficiency of the electron transport chain, it may subsequently promote oxidative phosphorylation57. Indeed, several studies have reported the capacity of MB to increase internal respiration, both in vivo and in vitro. The capacity of MB to increase OCR was initially reported nearly a century ago, where MB was shown to increase oxygen consumption in leukocytes and tumor tissue24,25. More recent studies have reported OCR-stimulating effects of low doses of MB in the CNS, in which it readily accumulates by crossing the blood-brain barrier58, as well as diabetic hearts28, and hepatocytes59. Several studies in multiple systems have linked MB-induced increases of OCRs to its role as an electron cycler in the mitochondrial ETC, acting to reduce cytochrome c and bypassing complexes I and III28,30,33,37,42,59,60,61. Given that the unique redox potential of MB enables the direct transfer of electrons to cytochrome c to enhance mitochondrial respiration in several model systems28,30,33,37,42,59,60,61, it is feasible to assume a conserved complex-specific mode of action in developing zebrafish embryos based on the evidence presented. While a detailed investigation of the conservation of a mitochondrial complex-specific mode of action of MB in developing zebrafish embryos (consisting of diverse cell types and tissues) remains to be formally tested, this is outside the scope of the current study. Nevertheless, all lines of evidence of metabolic MB effects across biological scales point to a mitochondrial mode of action. At the organismal level, MB increases OCR, and the reduced responsiveness to oligomycin, a blocker of the mitochondrial ATPase proton channel, reveals that the increase in OCR is linked to increased ATP production potential. This is further corroborated by the application of JC-10, a dye-based marker of mitochondrial membrane potential which reveals consistent and significant increases in response to MB exposure in vivo. Finally, whole embryo gene expression profiles reveal unidirectional decreases in gene expression of components of mitochondrial complexes at 1 and 4 dpf. Nuclear DNA-encoded transcripts of complex ii (sdhab) was consistently downregulated in 1 and 4 dpf embryos, while transcripts encoding for complex ii (sdhda) iv components (cox4l1) were significantly downregulated exclusively at 1 dpf. Conversely, transcripts with roles in complex i (ndufa9) iii (uqcrc2a) and ATP-Synthase F1 (atp51fe) were significantly downregulated exclusively at 4 dpf. Together, this whole embryo-targeted oxidative phosphorylation gene expression profile further confirms mitochondrial modes of action of MB. The unidirectional downregulation may suggest adaptive responses to MB as interacting agent of the mitochondrial transport chain, or changes in OCR. While MB has been reported to promote mitochondrial biogenesis60, the consistent decrease in additional transcripts with mitochondrial function, such as star, the rate-limiting transporter involved in mitochondrial cholesterol import, may suggest a general decrease in mitochondrial capacity.

However, it is also important to note that MB may affect other redox-dependent systems, such as xenobiotic response cytochromes62. Indeed, MBs mode-of action may not be restricted to its function on redox-dependent systems in zebrafish, and non-mitochondrial modes of action of MB may play important functional roles in the described phenotype. For example, MB has been reported to inhibit the serotonin-degrading enzyme monoamine oxidase A (MAO-A)63, shown to promote mitochondrial OCR in isolated mouse cortical mitochondria64, and to induce hypolocomotion and a hyperserotonergic phenotype in zebrafish65,66. Therefore, additional mode-of-action-based assessment of the consequences of MB exposure warrants future testing in developing zebrafish.

Since OCR measurements using the Seahorse assay are restricted to unhatched embryos due to the size limitation of the assay plates, we further explored oxidative metabolism-dependent energy expenditure using the Alamar Blue assay in post-hatched (eleuthero)embryos and larvae. In contrast to the initial OCR stimulation, we observed a consistent reduction in oxidative metabolism-dependent energy expenditure at 4 and 5 dpf following a prolonged MB exposure starting immediately after embryo collection. This decrease was not solely dependent on prolonged static renewal exposure from embryo to post-hatching, because acute post-hatching MB exposure alone recapitulated the reduced energy expenditure at 4–5 dpf. In contrast to chorionated embryos, post-hatching (eleuthero)embryos and larvae are free-swimming, and Alamar Blue assays conducted at these stages thus have the capacity to additionally capture locomotion-based energy expenditure. This should be tested formally in future experiments. Importantly, we also show that in addition to the reported metabolic effects of MB, the Alamar Blue assay itself appears to be sensitive to MB concentrations routinely used in zebrafish husbandry. The principle behind the Alamar Blue assay involves a reduction of the non-fluorescent indicator dye resazurin to its highly fluorescent form resorufin in the presence of reducing agents, especially NADH and NADPH (Supplementary Fig. 1). In line with its reported role as a redox-cycling substrate, MB transforms, in the presence of NADH or NADPH, into leucomethylene blue, which, in turn, is subsequently oxidized back to MB by O217. This redox-cycling results in the utilization of NADH/NADPH and O217, leading to an increase in O2 consumption and a general decline in the availability of NADH and NADPH. This, at least in part, may explain the decrease in energy expenditure observed after prolonged and acute post-hatching MB exposure at 4 and 5 dpf. When using the Alamar Blue assay to assess the consequences of MB exposure at recommended concentrations in unhatched embryos at 1 dpf, we observed no change in energy expenditure. However, exposure to a higher dose (0.001%) showed consistent reductions. In line with these MB dose-dependent effects observed in 1 dpf Alamar Blue assays, we also observed no increase in OCR at higher MB concentrations (Supplementary Fig. 2).

Overall, we thus identified organismal-level metabolic differences that depend on concentration and developmental time. Considering the well-defined dose-dependent effects of MB on mitochondrial energy metabolism in other model systems59,61, it is feasible that repeated steady-state exposure and/or post-hatching exposure resulted in higher internal MB concentrations, leading to differential metabolic effects. Importantly, a direct effect of MB on Alamar Blue is also supported by a small but significant reduction in Alamar Blue fluorescence after 24 h exposure with CCAC (0.00005%) and CSH (0.0001%) MB concentrations in the absence of zebrafish embryos (Supplementary Fig. 4). While these decreases in fluorescence readings are smaller than what is observed in post-hatching embryo assays, they nevertheless show that direct interaction between compounds affects, at least in part, Alamar Blue assay fluorescence readings.

Finally, we also observed that the CSH concentration (0.0001%) of MB induces early hatching. Previous studies reported conflicting results regarding the effect of MB exposure on hatching44,45,51. While chemical exposure-dependent alterations in hatching time have been reported as an endpoint in zebrafish toxicity testing67, future studies should investigate possible interactive effects of MB and test chemicals on this endpoint in toxicity screening studies.

In light of the fact that the teleost endocrine stress axis end product cortisol have demonstrable effects zebrafish development and metabolism68,69,70, we directly probed possible effects of MB on the acute stress response. Following the identification of consistent decreases in the transcript abundance of star, the rate-limiting transporter in mitochondrial cholesterol import and thus mitochondrial steroidogenesis capacity71, and the cortisol-responsive pck1, the rate-limiting cytoplasmatic enzyme involved in de novo gluconeogenesis72, we quantified the activity of the endocrine stress axis at baseline and in response to an acute stressor. However, developmental MB exposure did not affect cortisol response over time as tested in a two-way ANOVA. Thus, we conclude that the mitochondrial effects of developmental MB exposure were metabolic in nature, and did not extend to differences in cortisol production and, thus, endocrine stress axis function.

Overall, MB exposure at recommended CCAC10, FELASA18, and CSH19 guidelines induces robust metabolic effects linked to mitochondrial function. These effects are in line with the reported mode-of-action of MB in mammalian model systems. Compared to previous studies investigating the effects of MB exposure on more general aspects of zebrafish development45,47,48, these previously unmeasured, mode-of-action-based, robust and repeatable metabolic effects occur at exposure concentrations currently recommended and widely applied in zebrafish husbandry protocols. This is problematic, because zebrafish are, due to their long-standing use as developmental model73, their genetic tractability74,75, their potential for high-throughput applications76, and their exemption from vertebrate model animal ethics regulation in the (eleuthero)embryo stage before initiation of exogenous feeding in several legislations8,10,11, a widely and increasingly used research model. As such, the versatile zebrafish model has and continues to contribute to significant advances in comparative physiology2,5,77,78,79, biomedical science52, and toxicology7,80. While caution regarding the use of MB in zebrafish research is thus generally warranted to avoid currently underappreciated confounds on measured endpoints, the efficacy of MB as a therapeutic compound may be specifically tested in this model. Indeed, MB has been investigated as a potential therapeutic due to its neuroprotective properties37. As zebrafish are increasingly recognized as a valuable biomedical research model organism to study therapeutic compounds and their mechanisms, this model may be effectively utilized to evaluate the potential benefits of MB.

Here, we briefly review select examples of research, in which standard MB concentrations in E3 media have been used but may, at least partially, have affected experimental endpoints assessed in larval zebrafish. Zebrafish knock-out models in comparative physiology have contributed greatly to advances in whole organism functional assessment of key signaling components in response to environmental change. For example, the homozygous knock-out of paralogous hif1aa and hif1ab genes encoding for the hypoxia-induced factor 1 alpha, a key component of the evolutionarily conserved molecular oxygen sensing pathway81, has significantly contributed to assessing the functional involvement of this pathway in fish-specific physiological responses to environmental hypoxia2,77,82,83. These studies, conducted with MB concentrations (0.05%) higher than the CCAC and FELASA-guideline (0.00005%) MB concentrations shown here to elicit changes in OCR and/or energy expenditure, were used in embryo water to rear wildtype and mutants until their assessment of larval oxygen consumption and Pcrit at 7 dpf2. Since routine metabolic rate, among other variables, has been linked to Pcrit in a systematic review of factors affecting fish hypoxia tolerance84, it is possible that developmental MB exposure at least partially affected the routine metabolic rate and Pcrit measurements. Additionally, aside from mitochondrial effects on OCR, MB has been shown, through its action as MAO-A inhibitor63, to affect the serotonin system, which is critically involved in oxygen sensing in zebrafish85.

Due to their aforementioned capacity for high-throughput assays76, and their exemption from vertebrate animal research model ethical regulations8,10,11, early life stages of zebrafish have also been used for pharmacological compound screening86 and are receiving significant attention as new approach methodology (NAM) in toxicity testing87,88. As such, zebrafish (eleuthero)embryos represent biologically informative and economically viable, in vivo models under the mandate to replace, reduce, and refine (3Rs) animal use in toxicology80. The importance of models such as the zebrafish is evident in recent international policy mandates reflected by recent changes in the US Food and Drug Administration (FDA) legislation allowing toxicology testing of human drugs in alternative models88, the US Environmental Protection Agency’s (EPA) goal to reduce/replace traditional animal toxicity testing by 203589, and the EU’s Directive 2010/63 and REACH90, among others. Of critical importance to these endeavors is the establishment of standardized parameters91. Indeed, high-throughput toxicity assessment in zebrafish embryos is heterogeneous regarding the use of MB: while some studies employ it92, others do not, or do not specifically describe its possible inclusion in the embryo medium93. Given the observed effects on zebrafish embryo metabolism in the current study, we recommend against the inclusion of MB to avoid possible confounds. Indeed, MB has the potential to increase the toxicity of known aquatic contaminants. For example, because MB has been demonstrated to act on mitochondria28,30,33,37,42,59,60,61 and/or the serotonin system63, studies investigating the effects of developmental SSRI exposure in zebrafish at environmentally relevant levels while using MB should be probed for possible potentiation of 5-HT-dependent and or mitochondrial effects. In a similar vein, confounds of MB-containing embryo medium on PFAS-induced locomotor behavior have recently been reported in zebrafish embryos48. Thus, MB addition to embryo media should be reviewed in zebrafish toxicity studies to harmonize experimental protocols and avoid possible confounds in risk assessment. With regards to standardization, effects of genetic background and rearing conditions on larval and adult zebrafish physiology have been reported94,95,96,97 and are increasingly recognized. While we used AB strain due to its higher repeatability in some physiological endpoints, such as behavior98,99,100, strain choice may affect responses to MB. In the field of zebrafish research, systematic study and reporting of such effects are thus paramount to improve reproducibility.

Methods

Zebrafish husbandry

All procedures followed the CCAC guidelines and were approved by the University of Ottawa Animal Care and Veterinary Service (ACVS) under protocol number BL-2786. Adult wild-type zebrafish (AB strain, 6–8 months old, unless otherwise stated) stem from a line of AB fish originally been procured from the Zebrafish Core Facility, Life Science Research Institute (LSRI) at Dalhousie University, Halifax, Nova Scotia, Canada by ACVS in 2023. Zebrafish were housed on a recirculating system (Techniplast, Montreal, QC, Canada) held at a 14:10 light: dark cycle at the University of Ottawa’s Bioscience Aquatics Facility. Water temperature was maintained at 28 °C with pH and conductivity at 7.4 and 400 µS, respectively. Animals were fed using automatic feeders (Serenelife, Brooklyn, NY, USA) with Gemma micro 300 (Skretting, Westbrook, ME, USA) in the morning (10:00 h) and afternoon (16:00 h), respectively. An additional feed of live Artemia (Brine Shrimp Direct, Ogden, UT, USA) was fed at noon (12:00 h). Adult male and female zebrafish were transferred to breeding traps placed in 3 L tanks on a recirculating system the afternoon before spawning. The following morning, eggs were collected 60 min after lights turned on, embryos were placed in a fine nylon mesh strainer (700 microns) and gently rinsed (5x) with system water followed by a final rise (2x) with autoclaved E3 medium, prevented any fungal growth. Following this, fertilized and normally developing embryos at 1–2 hpf, assessed according to Sprague and colleagues101, were selected and incubated at 28 °C until further processing.

Methylene blue exposure

A stock concentration of 1% (w/v) MB (Cat. No. M9140; Sigma-Aldrich, Oakville, ON) was prepared using Milli-Q water and further diluted to its appropriate concentrations of 0.00005, 0.0001, 0.0003, and 0.001% in E3 medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2, 0.33 mM MgSO4 (Sigma-Aldrich)). An overview of recommended/suggested MB concentrations used in zebrafish embryo rearing is depicted in Fig. 1B. Sexually non-differentiated zebrafish embryos were exposed to MB solution immediately after embryo processing in 100 mm petri dishes (VWR International, Mississauga, ON, Canada), and maintained at a density of ~40–50 embryos per petri dish with a total media volume of 25 ml. Control embryos were only exposed to E3 medium without MB. Embryos were maintained at 28 °C in an incubator (Fisherbrand™, Fisher Scientific, Ottawa, ON, Canada) with a 14:10 light: dark cycle. Media was replaced three times (12:00, 15:00, and 18:00 h) at 0 dpf, and renewed daily (~50%) from 1 to 5 dpf. Using a Zeiss Discovery V8 Stereoscope images of zebrafish larvae were acquired, larvae were washed (3x) with E3 medium, anesthetized and immobilized using 200 mg/L of MS-222 and 4% Methylcellulose.

Hatching

Zebrafish embryos exposed to relevant concentrations of MB (0.00005 and 0.0001%) were assessed at 1 dpf using a light microscope (Olympus LS, Québec, QC, Canada), where only healthy developing embryos were selected and transferred to 48-well plates (Fisher Scientific), with each well containing 1 ml of E3 medium and its relevant concentration of MB. Media replacement (~50%) was done daily until 5 dpf. Hatching was evaluated at various time intervals (10:00, 13:00, and 17:00 h) from 2 dpf until 5 dpf, and quantified as a percentage of hatched embryos relative to the total number of viable embryos at 5 dpf. Each 48-well plate was treated as a single replicate.

O2 consumption rate (OCR) measurement

After overnight exposure to MB (0.00005, 0.0001, and 0.001%), the oxygen consumption rate (OCR) of 1 dpf zebrafish embryos were measured using a Seahorse XF 24 Extracellular Flux Analyzer (Agilent Technologies, Inc., Santa Clara, CA, USA) (Fig. 1A), as previously described102. Briefly, the Seahorse XF 24 Extracellular Flux Analyzer cartridge (Cat. No. 102342-100; Agilent Technologies) was hydrated with Milli-Q water (1 ml per well) at 28 °C the day before the assay. The XF Cell Mito Stress Test kit (Cat. No. 103015-100; Agilent Technologies) was used, and reagent concentrations of 100 μM Oligomycin A (Sigma-Aldrich), 45 μM Carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP), and 15 μM Rotenone/antimycin-A were prepared in E3 medium containing 0.1% DMSO. About 75 μl of each reagent was loaded to port A (oligomycin), port B (FCCP), and port C (Rotenone/antimycin-A) of the XFe24 sensor cartridge. Following sensor cartridge hydration, 1 ml of Seahorse XF calibrant solution (pH 7.4) was added for sensor cartridge calibration and incubated at 28 °C, after which the cartridge was inserted into the Seahorse XFe24 Extracellular Flux Analyzer. Before transferring embryos to each Islet Capture microplate, a thorough media replacement was done to remove the E3 medium containing MB. Subsequently, individual embryos were transferred into each well containing a total volume of 525 μl of E3 medium and stabilized using the capture screen. After sensor cartridge calibration, the microplate containing the embryos was inserted into the analyzer, and the XF Cell Mito Stress Test was measured over 50 cycles. Each cycle consisted of a 5 min duration, encompassing 1 min and 30 s for mixing, 1 min and 30 s wait, and a 2 min measurement period. Basal OCR was measured for a total of 12 cycles, followed by 16 cycles after oligomycin injection, eight cycles after FCCP injection, and 14 cycles after rotenone/antimycin-A injection. OCR measurements were carried out across two independent exposures, both of which contained unexposed control and CSH MB concertation of 0.0001%. While the first exposure also investigated the effect of CCAC concentration exposures (0.00005%) on OCR, the second exposure also investigated MB exposure effects on OCR at a concentration of 0.001%.

JC-10 mitochondrial membrane potential (ΔΨm) assay

The mitochondrial membrane potential (ΔΨm) was measured using JC-10 (Enzo Life Sciences Inc, Farmingdale, NY, USA), a more soluble iteration of the dichromatic JC-1 dye (5,5′,6,6′-Tetrachloro-1,1,3,3′-tetraethylbenzimidazolylcarbocyanine iodide). Initially, a 50 μM concentration of JC-10 was made from a 5 mM stock (solubilized in 100% DMSO) in an E3 medium. Following this, 100 μL of 50 μM JC-10 solution was placed in a black-walled, clear-bottom 96-well microplate (Invitrogen™, Eugene, OR, USA). Briefly, dechorionated 2 dpf embryos in 100 μL E3 medium were transferred to individual wells containing 100 μL of the 50 μM of JC-10 to achieve a final JC-10 concentration of 25 μM. Embryos were then incubated for 2 h at 28 °C shielded from light. After incubation, embryos were washed twice with E3 medium containing MS-222 (400 mg/L), where each wash step consisted of the removal of 100 μL of media from the wells, followed by adding 100 uL of media containing MS-222. Images of embryos were acquired using the BioTek Cytation C10 Image Reader (Agilent) with brightfield, GFP, and TRITC laser cubes. Fluorescent intensities of the GFP and TRITC channels were analyzed using Image J (NIH), background was subtracted from both the red and green, fluorescent images before the polarized mitochondrial percentage was calculated. The following equation was used to quantify fluorescence ratios observed on larval surface layers in the head region, including eyes, the gut region, and the yolk sac region, but not in the tail region, where the dye was not taken up efficiently: Polarized = Red Fluorescent Intensity – Background/((Red Fluorescent Intensity – Background) + (Green Fluorescent Intensity – Background)) JC-10 ΔΨm measurements were performed across two independent exposures.

Energy expenditure assay

The energy expenditure assay was performed using the Alamar Blue assay as previously described103,104,105, since the direct OCR measurement in the modified Seahorse assay is restricted to unhatched 1–2 dpf embryos. Alamar Blue stock solution containing 0.1% DMSO (AlamarBlue™ Cell Viability Reagent, Life Technologies, Eugene, OR, USA) was prepared in E3 medium. Briefly, exposed embryos/larvae were rinsed thoroughly (at least three times) with MB-free E3 medium before initiating the assay. Following this, zebrafish embryos/larvae were placed in black-walled, clear-bottom 96-well microplates (Invitrogen™, Eugene, OR, USA) with 200 μL of Alamar Blue stock solution per well. Fluorescence was measured in each well with embryo/larvae (unless otherwise stated) at 0, 3, 6, 12, 24, and 48 h using a SpectraMax Gemini™ fluorescence reader (Molecular Devices, San Jose, CA, USA) with excitation and emission wavelengths of 530 and 590 nm, respectively. Alamar Blue measurements were performed across four independent exposures and included different timepoints and exposure windows during development.

Transcript abundance

Total RNA extraction was performed on 1 and 4 dpf zebrafish after developmental exposure to 0.00005% MB. For 1 dpf, 15 embryos were combined, while for 4 dpf, a total of five larvae were pooled for a single replicate. Samples were stored at −80 °C after collection. Total RNA extraction was performed using TRIzol (Invitrogen™) according to the manufacturer’s instructions. Using the NanoDrop 2000C Spectrophotometer (Thermo Scientific, Waltham, MA, USA), the quality and quantity of the isolated RNA was analyzed, and 1 μg of total RNA was used to make cDNA utilizing the QuantiTech Reverse Transcription Kit (Qiagen, Toronto, ON, Canada) following the manufacturer’s instructions. To confirm the absence of genomic DNA contamination, a no reverse transcriptase control (NRT) was also generated. The qRT-PCR assays were carried out using a CFX96 PCR instrument (Bio-Rad, Mississauga, ON, Canada) with SsoAdvanced™ Universal SYBR® Green Supermix (Bio-rad), as previously described94,95,103,104,105. Using gene-specific primers and annealing temperatures (Supplementary Table 1), all samples were run in duplicates. The cycling conditions consisted of an initial activation step at 95 °C for 30 s, followed by 39 cycles consisting of two steps: an initial denaturation step at 95 °C for 10 s, and a combined annealing/extension step at 60 °C for 30 s (temperature based on primer annealing temperature). After each assay run, melt curves were generated by a gradual increase in temperature from 65 to 95 °C in 0.5 °C increments every 5 s, and melt curves were monitored for single peaks to verify the specificity of the assays. Each assay run consisted of an NRT and no template control (NTC). Finally, gene expression data were normalized using the NORMA-Gene approach, as previously described96,104,105. All data were normalized to control and shown as fold-change.

Larval stress response

Larval stress response was determined at 5 dpf after developmental exposure to 0.00005% MB. At 5 dpf, larval zebrafish were transferred (at 12:00 pm) to 1.5 ml tubes (ten larvae per 1 mL) and subjected to a brief swirling stressor for 30 s using a benchtop vortex at level 6 (Fisherbrand™ Analog Vortex Mixer Catalog No.02-215-414), experimental parameters for the stress test were chosen as previously described97. Larval zebrafish were sampled at 5-, 10-, and 60 min post stressor, and quickly euthanized on powdered dry ice (larvae were immobile in less than 60 s). For the assessment of basal cortisol concentrations, zebrafish larvae were directly euthanized in the petri dish. A total of ten larvae were pooled for a single replicate for cortisol quantification. Larval stress response samples were collected across three independent exposures.

Cortisol quantification

Whole-body cortisol was determined after diethyl ether extraction. Briefly, ten pooled larvae constituting a single replicate were homogenized in 120 μL of cortisol kit assay buffer using a sonicator (Fisher Scientific, Canada). Cortisol was extracted in glass tubes by adding 1 ml diethyl ether to 120 μL of samples, vortexed for 30 s, and left undisturbed for 10 min to allow for phase separation. The aqueous phase was frozen using ice-cold ethanol (95%) in dry ice, and the organic phase was decanted into a new glass tube. This extraction procedure was repeated two more times. Following a total of three extractions, the organic phase was evaporated under a gentle stream of N2 gas. After evaporation, an additional 1 ml of diethyl ether was introduced to the tubes while gently turning them in a circular motion to ensure thorough dissolution of any residues stuck on the wall of the glass tubes, and then evaporated again. Samples were then reconstituted with 240 μL of cortisol kit assay buffer and stored at −80 °C. Cortisol was quantified using a commercial Cortisol immunoassay Kit (DetectX® Cortisol Immunoassay Kit, Arbor Assays, Ann Arbor, MI, USA; sensitivity of 27.6 pg/ml). The plate-to-plate variation had an inter-plate CV of 6.7%.

Statistics and reproducibility

Care was taken to randomize sample positioning in assay replicates. All data were analyzed and graphed using GraphPad Prism Version 9 (GraphPad Software Inc, San Diego, CA, USA). Data were log-transformed when necessary to meet the requirements of normality and equal variance. In cases where this transformation was insufficient, a non-parametric test was used. Based on whether data met these assumptions, a parametric (ANOVA) or non-parametric (Kruskal–Wallis test) was used to assess significant differences in multi-group comparisons. In cases of significance, relevant post hoc tests (Tukey’s, Dunnett’s, or Dunn’s multiple comparisons) were used to resolve differences between specific treatment groups. All data were graphed as violin plots (unless indicated otherwise). The mean (indicated by the + symbol), median (represented by a line in the middle), upper and lower quartile range (indicated by the dotted lines), and kernel probability density (the outline of the violin plot) are illustrated. In cases where two groups were compared, a parametric (two-tailed unpaired t-test) or non-parametric (Mann–Whitney U-test) were used depending on whether the data met the assumptions stated above. A P value of less than 0.05 was considered significant in all cases.

Conclusions

We demonstrated that MB exposure of embryo-larval zebrafish at widely recommended concentrations robustly and repeatedly affects several organismal and molecular components of oxidative energy metabolism. MB furthermore has the potential to confound redox reaction-dependent high-throughput assays of energy expenditure. The impairment of oxidative energy metabolism in developing zebrafish is mediated, at least in part, through MB-induced mitochondrial dysfunction, in line with the known mode-of-action of MB in other model systems. Due to its unintended consequences in developing zebrafish larvae and its strong potential to confound metabolic endpoints, researchers should reconsider MB use as an antifungal agent in zebrafish research.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.