Ecotoxic Effects of the Vehicle Solvent Dimethyl Sulfoxide on Aquatic Model Organisms

Dimethyl sulfoxide (DMSO) is widely used as a vehicle solvent in ecotoxicity bioassays. However, despite its frequent use, itcould be toxic for organisms at some concentrations. Hence, the aim of this study was to investigate the effectsof DMSO on the population growth rate of the microalgaeRaphidocelis subcapitata, the mobility of the microcrustacean Daphnia magna,and the reproduction of the rotiferBrachionus calyciorus. DMSO was applied to the organisms in concentrations ranging from 0.031–4%. For R. subcapitata signicant effects in growth inhibitionafter 72 h of exposure was 0.125% DMSO,being the lowest observed effectconcentration (LOEC). The 50% effective concentration (EC 50 ) was2.138 ± 0.372%. In D. magna,signicant differences in the mobility after 24 h or 48 h of exposure was 1% DMSO being 1.712± 0.207% and 1.167± 0.220%DMSO the EC 50 observed for 24 h and 48 h exposure, respectively. For B. calyciorus,it was not possible to validate the tests performed, as there were insucient animals alive in the control conditions at the end of the exposure period. Therefore, we recommended avoiding DMSO as a vehicle in assays using B. calyciorus,and to use nal DMSO concentrationsin experimental solution not exceeding 0.125% forR. subcapitata and 0.5% for D. magna. Data Graphical abstract is not provided with this version. Graphical Abstract - The effects of the vehicle solvent dimethyl sulfoxide (DMSO) was investigated in three aquatic model organisms, the microalgae Raphidocelis subcapitata, the microcrustacean Daphnia magna, and the rotifer Brachionus calyciorus. As a result, negative effects on growth rate for the algae population, immobilization of daphnids and inhibition of rotifer reproduction was observed obtaining EC50 close to 1% of DMSO for all these cases.


Introduction
Ecotoxicological laboratory bioassays often require solvents as vehicles to solubilize lipophilic test compounds and organic pollutants with low water solubility that need to be dissolved prior to addition to experimental systems (Jay et al.1996, David et al. 2012).It is important to note that although in the environment these solvents are not required to dissolve the compounds in question, in laboratory conditions where pure compounds are tested in isolation is unavoidable the use of them to make experiments possible. Nonetheless, their potential effects and the stress imposed on the test organisms by their application is of great concern, and has become a crucial issue in professional papers (Jay et al. 1996, Okumura et al. 2001).
Whilst international test guidelines (e.g. the O ce of Chemical Safety and Pollution Prevention [OCSPP] and the Organization for Economic Cooperation and Development [OECD]) and the US EPA (United Stated Environmental Protection Agency) de ne that solvents can be applied at their maximum acceptable concentrations, the nal concentrations used in aquatic assays are often higher than the typical recommended limits of 0.1 mL/L, i.e. 0.01% (v/v) (OECD 2019, Green & Wheeler 2013). However, this recommendation is not species-speci c, but a general statement and do not considered that different species respond differently to a given solvent exposure. In addition, usually there are difference also among strains of the same species. Therefore, in practice, it is not possible to be this rigorous, suggesting a review of perspective considering the work. Thus, the nature of the solvent and the exact concentration applied in a given study should be recorded and stated.
Dimethyl sulfoxide (DMSO) is the most common commonly used solvents in toxicology for the administration of chemicals (Christou et al. 2020). This sulphur-based compoundis produced in nature as a result of the oxidation of biogenic dimethylsulphide (DMS) by marine algae and bacteria (Deschaseaux et al. 2014, Kim & Lee 2021). Its popular use as a carrier vehicle is due to its low toxicity to living beings and its ability to permeate biological membranes without signi cant damage to their structural integrity (Kais et al. 2013). It is also an aprotic solvent with the ability to dissolve both polar and nonpolar compounds, and organic and inorganic chemicals (Huang et al. 2017).
In aquatic toxicology, the test organisms aresubjected to continuous exposure equivalent to the duration of the test protocol (David et al. 2012). Therefore,there has been a steadily growing concern among ecotoxicologists over the importance of understanding the effects of DMSO per seon experimental outcomes, as it in uencesbehavior, physiologic and biochemical parameters of the organism.
Thus, evaluation of the effects of DMSO itself is an important task.In this regard, the current study investigated the effects of DMSO on aquatic model organisms (green microalgae, microcrustacea, and rotifers) used for several common and standardized bioassays, at low concentrations (up to 4%) that are often employed in aquatic assays emphasizing either the differences in sensitivity among species. and zinc (prepared from ZnSO 4 and solutions containing 0.01 to 0.105 mg L -1 of Zn 2+ ) wereused as two separate positive controls and distilled water as negative control. The algal culture in the exponential growth phase was diluted in ISO medium for freshwater algae (pH 8.1 ± 0.2) to obtain the inoculum for the test with a cell density of 10,000 cells mL -1 .The test plates were incubated in a room with continuous illumination of 70 mmol m -2 s -1 (cool-white uorescent lamps) at 23±2ºC. After 72 h of incubation, the uorescence variation was measured at wavelength of 485 nm for excitation and 685 nm for emission with FLUOstar microplate reader (BGM Labtechnologies), to establish whether growth had been inhibited or stimulated in comparisonwith the control.

Microcrustacea: immobilization assay
The microcrustacean immobilization assay were performed according to ISO 6341 (2012). Brie y, three different experiments (with four replicates for each one) were performed. An assay tube with 10mL of culture medium (pH 7.8 ± 0.5, aerated overnight) plus the appropriate proportion of DMSO was prepared for each replicate. Five young D. magna (up to 24 hours of life) were then added to each tube. Potassium dichromate (K 2 Cr 2 O 7 ) (concentrations ranging 0.58 and 1.6 mg L -1 ) and zinc (prepared from ZnSO 4 with concentrations ranging from 0.05 to 19.79 mg L -1 of Zn 2+ ) were used as the positive controls and the distilled water as negative control. The test tubes were covered with aluminum foil to keep light out and were kept in an incubator at 20 ±2ºC. Inhibition of the mobility of the individual D. magna was determined visually after 24 and 48h of exposure.

Rotifers: reproduction assay
Cysts of the rotifer Brachionus calyci oruswere purchased from MicroBioTests Inc., Belgium. The cysts were placed in a Petri dish containing 5 mL of fresh ISO medium (ISO 20666, 2008) at pH 7.6 ± 0.3 and were incubated at 25 ± 1ºC for 18 to 24 h under continuous illumination of 1,600 Lux. An individual rotifer that had hatched from a cyst less than 2 hours previously was placed in each well of a 24-well microplate. Three different experiments (replicates) were conducted for each concentration of DMSO (0.063% to 4%). In each experiment, eight repetitions (wells) per concentration (or control) were made. A solution containing about 10 6 cells mL -1 of Raphidocelis subcapitata was added to each well in order to feed the individuals during the experiment. A copper solution (prepared from copper sulphate pentahydrate -CuSO 4 5H 2 Ocontaining Cu 2+ concentrations of between 1.1 and 121.22 mg L -1 ) was used as a positive control and distilled water as negative control. The plates were covered with aluminum foil and incubated at 25± 1ºC. After 48 hours, the total numbers of individuals were counted using a binocular microscope, to estimate the reproduction rate.

Statistical Analysis
The no observed effect concentration (NOEC) and the lowest observed effect concentration(LOEC) were determined through one-way analysis of variance (ANOVA) and followed by post-hoc comparisons using Williams's test Williams, (1972) with a signi cance level of 0.05. These analyses were performed with the PMCMRplus package version 1.9.0 in R 4.0.5 (Pohlert 2021).The treatment concentration that caused a 50% effect (EC 50 ) in comparison to the control treatment was derived from four-parameter log-logistic regression models using the DRCpackage version 3.0 in R, following Ritz et al. (2015).The regression curves graphs were prepared using ggplot2 version 3.3.3 in R.

Effects on the algae population
The half maximal effective concentration (EC 50 ) for potassium dichromate on the R. subcapitata strain was 0.45 ± 0.09 mg L -1 (Table 1).A positive control solution containing increasing concentrations of zinc (Zn 2+ ) was also tested. At LIEC (Laboratoire Interdisciplinaire de Environnements Continentaux, Université de Lorraine, France), where the experiments were conducted, ZnSO 4 is the preferred positive control for algal tests, as its response is reliable and a dose-response curve is well established, as demonstrated in this study ( Table 1). The lower concentration (0.01 mg L -1 of Zn 2+ ) tested was the NOEC, while the LOEC was a concentration of 0.014 mg L -1 .
The concentrations of DMSO applied to the culture medium ranged from 0.031% to 4%, with the algae being subjected to 72 h of exposure. Low concentrations (0.031% and 0.63%) did not exert a signi cant toxic effect on the growth of the freshwater algal population ( Figure 1A). However, increasing the concentration of DMSO to 0.125% and aboveresulted in a signi cantly (p<0.05)growth inhibition. Maximum inhibition was observed with 4% DMSO ( Figure 1A). Accordingly, the LOEC was 0.125% and the EC 50 value from the sigmoidal regression curve was 2.138 ± 0.372%( Figure 1B).

Effects on the mobility of daphnids
For the D. magna immobility assay,a gradual dose-dependency inhibition of the mobility of the individuals was observed after 24 and 48 h of exposure to K 2 Cr 2 O 7 (Table 2). Accordingly, the EC 50 values were found to be 1.11 (± 0.08) and 0.91(± 0.34) mg L -1 of K 2 Cr 2 O 7 after 24 and 48h of exposure, respectively.
The effects of zincwere also evaluated and concentrations from 0.05 to 1.52 mg L -1 were not toxic to the daphnids over 24h, but higher concentrations showed signi cant (p>0.05) toxicity ( Table 2). The observedEC 50 values were 2.32± 0.58 and 1.86 ± 0.21 mg L -1 for 24 and 48 h of exposure respectively ( Table 2).
The results for DMSO exposure demonstrated that concentrations ranging from 0.031% to 0.5% were not toxic after 24 h of exposure, while higher concentrations did exert signi cant (p<0.05) toxicity on daphnids ( Figure 2A). Observation of the behavior of the individuals after 48 h revealed that a DMSO concentration of 1% immobilized over 40% of the daphnids. The highest concentration applied (4%) completely inhibited the mobility of all individuals (Figure 2A). Hence, the LOEC after 24 and 48 h of exposure was 1% and the EC 50 observed for the mobility of D. magna was a DMSO concentration of 1.712± 0.207% for 24 h exposure and 1.167± 0.220% for 48 h ( Figure 2B).
All concentrations of DMSO tested inhibited somehow, the reproduction rate of B. calyci orus.Concentrations above 1% (0.031 -0.5%) inhibitedreproduction of B. calyci orusno more than 20%with reproduction being reduced by 70% with the highest concentration solution tested (4% DMSO) ( Figure 3A). Signi cant (p<0.05) inhibition on relation to control was detected only up 2% of DMSO and the EC 50 was calculated asa concentration of 1.624 ± 0.988% including the LOEC (2%) and NOEC (1%). However, this recommendation is general and non-dependent of the solvent applied, neither it is speci c for a given specie and do not considered the biological differences between living beings. A clear example of this different susceptibility is the result obtained with algae assays in the present study comparing the control positive substance with those found in the literature. The EC 50 values for potassium dichromate on the R. subcapitata strain was 0.45 ± 0.09 mg L -1 (Table 1)  Many bioassays employ DMSO as a carrier medium in concentrations of up to 10% (v/v) to achieve appropriate biological availability of hydrophobic toxicants (Huang et al. 2017) buteffects of DMSO at this level should be discussed as it could exert toxicity. Nonetheless, the recommended concentration (0.01%0 in the cited guidelines is very restrictive,and to gure the restriction of these guidelines as far as the importance to run concurrent solvent and negative controls as an experimental group to avoid misinterpretation of results and to assess the influence of the carrier in such assays (Bownik, 2019, Hu et al., 2017, DMSO was applied in aquatic models algae, daphnia and rotifer in the present study. Recently the same approach was applied in zebra sh embryo assay (Christou et al. 2020).

Discussion
The results achieved in the present work, for example, demonstrated that for the algae R. subcapitata the rst and second concentrations of DMSO tested (0.031 and 0.063%) do not exert any effects in the population growth (Figure 1). Inhibition effects on the growth were noticed in concentration 10 times more concentrated (0.125 %) than recommended in the guidelines (Figure 1).
In this sense, our result is in agreement with early experiments. Okumura et al. (2001) reported that DMSO did not affect the growth of nine species of marine microalgae at concentrations below 1%. Furthermore, Ma and Chen (2005) assessed the toxicity of seven solvents (acetone, ethanol, methanol, DMSO, N,Ndimethyl-formamide, furanidine, and acetidin) to green algae, and concluded that an adequate range of DMSO concentrations for the Chlorophyceaen Chlorella pyrenoidosa was 0.01% to 0.50%.They found an EC 50 concentration of 1.49 mg L -1 (or 1.49% DMSO) (Ma & Chen 2005), which is in agreement with the value we observed for R. subcapitata ( Figure 1B) considering the biological differences discussed above.
On the other hand, Andreani et al. (2017) found that DMSO up to 1% was not in itself toxic to the bacteria Vibrio scheri and the freshwater microalgae Raphidocelis subcapitata. This corroborates the work of Jay (1996), who tested the effects of organic solvents on the growth of two Chlorophyceae (Chlorella vulgaris and Selenastrum capricornutum), and no effect was detected in doses up to 1% DMSO. Considering that S. capricornutum is a synonym for R. subcapitata,in summary, our results is even more restrictive with this limit concentration as it should do not exceed 0.125% of DMSO, but still, more permissive than the limits recommendation in guidelines.
Daphnia a rst consumer in food chain demonstrated to be less sensitive than the alga, a producer in the food chain. It was rst observed comparing the positive control applied as the EC 50 value of K 2 Cr 2 O 7 found (1.11 (± 0.08) mg L -1 after 24 h of exposure) is in agreement with the ISO 6341 (2012) standard, which presents 1.12 mg L -1 as the EC 50 medium value for 24 h exposure. For DMSO the LOEC observed for D. magna after 24 h or 48 h of exposure was 1% DMSO. The above-mentioned LOEC concentration represent a solution 100 times more concentrated than the one established in the guidelines (Figure 2). Similar conclusions could be made for comparisons between our ndings for the acute assay in D. magnawith the results from the previous reports available. Barbosa et al. (2003) demonstrated that the EC 50 for acute toxicity of DMSO to D. magna after 48 h of exposure was 2.23 %, while ours was 1.167%. Huang et al. (2017)observed in an immobilization assay that concentrations of DMSO ranging between 0.1% and 1% were not toxic to Daphnia neonates when they were exposed for 24 h, but that higher concentrations (more than 2%) were toxic over both 24 and 48 h. The EC 50 values observed were 2.3% for 24-h exposure and 0.5% for 48-h exposure. Comparing these results to the ones we observed in our study (Figure 2A and B), we can emphasize that concentrations of DMSO higher than 0.5% should be avoided in the classic immobilization test standardized by ISO 6341 (2012). Therefore, for acute assay, concentrations of DMSO higher than 0.01% (recommended limits) could be used as long as 0.5% of DMSO is not exceeded.
Aside from these observations, Huang et al. (2017) also tested the effects of DMSO on behavioral traits in a time-dependent manner in chronic assays and demonstrated that DMSO can have signi cant consequences if it is used in concentrations ranging from 0.01 to 1% claiming cautious interpretation for lethal and behavioral parameters in these situations. According to Hutchinson et al. (2006), for acute test the tolerance of the organisms to the solvents is higher than in chronic assessments, in contrast, the effects of solvents for longer-term exposure in chronic ecotoxicity endpoints achieve sub-lethal levels (i.e. reproduction) and consequently minor doses is su cient to provoke adverse effects. This consideration is of increasingly importance for the data obtained in the present work for B. calyci orus exposed to DMSO.
However, it is important to note that according to the ISO 20666 standard (2008), the experiment could not be validated. Indeed, according to validity criteria, a reproduction should be observed in at least 7 out of the 8 replicates of the control wells by the end of the experiment, and the number of individuals per well should be three or more. In all three replicates conductedin our experiments, one or more wells had less than three B. calyci orus. Yet, every well containing 1% DMSO presented only the B. calyci orus specimensadded at the beginning of the assay, while some of them died at 2% DMSO, and all rotifers died in wells containing 4% DMSO. Only in concentrations below 0.5% DMSO we could observe B.calyci orus neonates in the wells. All in all, considering that it is a chronic assay, and that the rst DMSO concentration tested (0.031 %) is above the recommended limits for acute assays (0.01 %), we could a rm that in this case DMSO exert toxicity in the tested conditions we choose. Zhanget al. (2016) found no signi cant difference between the DMSO treatment (0.01%) and the control group in the species Brachionus plicatilis. Other works dealing with reproductive factors in B. calyci orus observed that 1% DMSO did not affect the population growth rate, the ratio of ovigerous females to non-ovigerous females, mitotic rate, or resting egg production (Ke et al. 2009, Snell & Carmona 1995, Snell & Janssen 1995. In brief, DMSO did not demonstrate any signi cant in uence on aquatic models such as microalgae and microcrustaceans, when applied according to the guidelines. In addition, the nonspeci c OECD recommendation, considering specie and type of solvent, of 0.01% is overprotective for algae and microscrucea model species. Hence, comparing our results with those from the literature, we can assume that the no-effect concentrations for the endpoints evaluated were up to 0.125% DMSO for algae R. subcapitata. This value is rather low compared to the observation reported previously by Jay et al. (1996) who demonstrated concentration up to 1% DMSO. In experiments with D. magna, the concentration of DMSO applied should not be above 0.5%, to avoid ecotoxic effects agreeing with Huang et al. (2018) for acute exposition.
Altogether, these results highlight the importance of considering the effect of DMSO on the experimental outcomes when designing assays and interpreting their results. If it is necessary to use solvents such as DMSO, rigorous monitoring procedures should be applied. It is also crucial to ensure that the solvent has no toxic effects on the evaluated parameters. When preparing reports, information on the concentration of solvent should be presented, together with the use of a parallel solvent control. Therefore, reporting of the toxicity of low doses of DMSO on aquatic organisms is of high relevance, as the interference of a carrier solvent should not be a confounding factor that can potentially affect the outcome of an assay.

Conclusion
Regarding our results, we comfort the recommendations of the respective guidelines of the assays: for experiments withR. subcapitatawe recommended that the nal concentration of DMSO in the tested solution should not be higher than 0.125% and 0.5% in experiments with acute toxicity in D. magna. For B. calyci orus,we were not able to identify a safety concentration value to recommend, thus, indicating to following the guidelines recommendations (0.01 % or even less).

Declarations
Funding: This study was funded by Brazilian funding agency "Coordination for the Improvement of Higher Education Personnel" (CAPES -Coordenação de Aperfeiçoamento de Pessoal de Nível Superior)" as post doctoral fellowship.
Financial Interest: The authors have non-nancial interests to disclose.
Con ict of interest: No con ict of interest has to be declared.
Availability of data and material: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.  Data presents the mean ± Standard Deviation. a All concentrations tested presented effects in comparison to control.

Graphical Abstract
Graphical abstract is not provided with this version.
Graphical Abstract -The effects of the vehicle solvent dimethyl sulfoxide (DMSO) was investigated in three aquatic model organisms, the microalgae Raphidocelis subcapitata, the microcrustacean Daphnia magna, and the rotifer Brachionus calyci orus. As a result, negative effects on growth rate for the algae population, immobilization of daphnids and inhibition of rotifer reproduction was observed obtaining EC50 close to 1% of DMSO for all these cases.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Additionalinformationtoeditor.docx