AbstractArtificial sweeteners are broadly used as safe food additives and in pharmaceutical formulations. As these compounds are relatively stable and poorly removed by water treatment facilities, their environmental concentrations increase. Therefore, concerns about their potential risks to non-targeted aquatic biota have been raised. Although no lethal effects are usually observed in standard toxicity testing, recent effect studies suggest a possible neurotoxic mode of action. We tested the effects of commonly used sugar substitutes (sucralose and acesulfame, up to 1 mg/L) in a nontarget species Daphnia magna, by assessing biochemical (acetylcholinesterase activity, AChE), physiological (heart rate, HR) and behavioural (time spent on swimming and feeding) endpoints. We found a dose-dependent increase in AChE activity and inhibitory effects on HR and behavioural endpoints, with lower EC50 values observed for acesulfame than sucralose, although all these values were within the reported AS levels in the wastewater. Moreover, a biphasic response was observed for acesulfame-exposed daphnids, with AChE inhibition at the lower concentrations (< 1 µg/L ) and stimulatory effects at the higher concentrations. Thus, the response propagated across the biological organization levels, with swimming inhibition associated with AChE stimulation. Whereas HR and swimming were positively related across the ASs and all concentrations tested, the relationship between AChE and HR was substance-specific, suggesting possible differences in the mode of action. The observed LOEC values for acesulfame were as low as 0.1 µg/L , these results suggest that artificial sweeteners may exert adverse effects on nontarget biota at environmentally relevant concentrations, which calls for a critical evaluation of current thresholds used in their risk assessment.KeywordsSucralose; Acesulfame K; Aquatic toxicity; Acetylcholinesterase; Heart rate; Behaviour; Daphnia magnaIntroductionArtificial non-nutritive sweeteners (AS) are widely used in food and beverages for human consumption, animal feed, and pharmaceutical and personal care products to replace sucrose and decrease caloric uptake \cite{Hough_1993}. ASs can be divided into two categories: the first generation, which mainly includes saccharin, cyclamate, and aspartame; and the second generation, which includes acesulfame, sucralose, neotame and neohesperidin dihydrochalcone \cite{Wang_2023}. The survey from Euromonitor International (https://www.euromonitor.com/sugar-and-sweeteners; 2017) showed that aspartame was the most used AS (18.5 thousand metric tons), followed by saccharin (9.7 thousand metric tons), acesulfame (6.8 thousand metric tons), and sucralose (3.3 thousand metric tons). Moreover, over the past decades, the production and consumption of these synthetic organic compounds have been increasing steadily \cite{Daher_2022}. Due to their relatively high stability and persistence in nature, the release of these compounds in the environment is also increasing. As a result, ASs are now considered emerging contaminants \cite{Praveena_2019,Wang_2023}, even though from the human health aspect, ASs have been regarded as safe additives to foodstuff \cite{Kroger_2006} and pharmaceuticals \cite{Haroun_2018}. The most commonly detected in wastewater effluents are acesulfame, sucralose, cyclamate, aspartame, and saccharin, with some regional variations, reflecting human consumption and inadequate degradation by water treatment technologies \cite{Li2018,Tran2018}. Moreover, they are also among the popular anthropogenic trace contaminants \cite{Van2020}, with the highest concentrations in sewage, ground- and surface waters \cite{Lange2012}. The major part (>95%) of sucralose can pass through the human digestion system unchanged and reach domestic wastewater by excretion \cite{Buerge_2009}, and acesulfame can be excreted without chemical alterations \cite{Klug_2012,Castronovo2017}. Therefore, acesulfame and sucralose are the micropollutants introduced mainly by human intake via food and pharmaceutical products and with high persistence \cite{Lubick_2008} and broad occurrence in nature \cite{Perkola2014,Lange2012,Buerge_2009}. In addition to the frequent reports on their occurrence in freshwaters, these ASs have also been detected in marine systems, including offshore areas \cite{Mead_2009,Whitall_2021}, which indicates a much larger scale of dispersion, a broader range of species under exposure, and, hence, greater ecological impact. Their environmental concentrations vary widely among regions, depending on local consumption, regional geography and seasonal changes \cite{Sang2014,Karstadt_2006}. In rivers and lakes, concentrations of these ASs are in the range of nano- to micrograms per litre \cite{Lange2012}, with often higher levels of acesulfame than sucralose \cite{Scheurer_2009,Scheurer_2014}. In untreated and treated wastewater, acesulfame and sucralose generally have the highest concentrations, up to mg L−1 \cite{Arbel_ez_2015}. In the environment, these ASs are quickly dispersed due to their high solubility and taken up by biota \cite{Ma2021,Saucedo-Vence2017,Stoddard2014,Tollefsen2012,Zygler2012,Lillicrap2011}. Concerns have been raised that exposure to ASs of nontarget organisms may result in adverse effects we presently know little about \cite{Tollefsen2012}. Sucralose, for example, with its molecular structure resembling sucrose, may perturb sugar receptors and responses involved in the olfactory, behaviour/feeding of grazers and photosynthetic pathways of primary producers \cite{Kessler_2009}. Also, due to ASs persistence, chronic exposure to even low doses can lead to delayed effects not readily detectable by standard (eco)toxicity assays \cite{Wiklund2012}. Moreover, in the environment, ASs can transform with the products having elevated toxicity as shown using oxidative transformation of acesulfame by permanganate \cite{Yin2017}. In line with that, the acute toxicity of the metabolites generated by exposing acesulfame and sucralose to UV light was enhanced by factors 575 and 17, respectively, in the assays using Vibrio fischeri \cite{Sang2014}. In standard toxicity tests, the effects of AS, including sucralose and acesulfame, are usually negligible \cite{Stolte2013,Kerberov__2021}. For example, in the OECD-guided tests ('Reproduction inhibition assay with limnic green algae Scenedesmus vacuolatus', 'Acute immobilisation assay with Daphnia magna', and 'Growth inhibition assay with duckweed Lemna minor'), none of these substances at the test concentrations of up to 1 g/L induced significant effects on the plant growth (microalgae and duckweed; up to 7-d exposure), and no acute mortality was observed in water fleas \cite{Soh_2011,Stolte2013}. Similarly, no significant adverse effects on survival and reproduction in daphnids and mysid shrimps exposed to sucralose at concentrations up to 1.8 and 0.09 g/L, respectively, were observed \cite{Huggett_2011}. An excellent set of laboratory-generated ecotoxicological data for acute and chronic toxicity of acesulfame in fish, invertebrates, plants, and sludge microorganisms suggests the lowest chronic no observed effect concentration (NOEC) of 22 mg/L based on mortality, weight, and length changes versus controls in zebrafish \cite{Belton2020}; thus, this value was used for calculating predicted no-effect concentration (PNEC) for acesulfame. All the effect studies mentioned above reported NOEC values of sweeteners being much higher than their actual environmental concentration, suggesting a very low environmental risk. In effect studies using non-standard endpoints, the adverse effects are detected more frequently and at much lower concentrations. For example, behavioural changes related to food intake and locomotion in different zooplankton species exposed to low sucralose concentrations, e.g., down to nanogram levels, were reported \cite{Hjorth_2010,Eriksson_Wiklund_2014,Wiklund2012}. In fish exposed to acesulfame, the observed responses included behavioural changes \cite{Dong2020}, oxidative stress \cite{Cruz_Rojas_2019}, and embryo development aberrations \cite{Li2016,Colín-García2022} at the exposure levels within NOEC. Moreover, the effects were observed at concentrations much below NOEC, e.g., in the gill, brain, and muscle of common carp exposed to acesulfame, the change in oxidative status was detectable at concentrations of 0.05 and 149 µg/L \cite{Cruz_Rojas_2019}. Also, dietary acesulfame exposure in mice has been linked to genotoxic effects manifested as DNA strand breaks and chromosomal aberrations \cite{Bandyopadhyay_2008,Mukherjee_1997}. Finally, profound adverse effects on the gut microbiome have been linked to AS consumption in humans \cite{Yu2023,Richardson2022,Del2022}, mice \cite{Zheng2022}, and rats \cite{Zhang2021}. All these indications for specific mechanisms of action suggest potential exposure risks to these ASs and imply a high possibility of unknown effects. Thus, with respect to the general data shortage and contradicting evidence, more (eco)toxicity data based on a comprehensive selection of sensitive endpoints relevant to chronic exposure are needed to evaluate toxicity thresholds for these substances. The aim of this study was to investigate the toxicity of sucralose and acesulfame by assessing biochemical, physiological and behavioural responses in a model species Daphnia magna. Considering a plausible neurotoxic mode of action for sucralose \cite{Finn2000} and acesulfame \cite{Dong2020}, we employed a battery of endpoints related to the sensory physiology of these animals to perform food collection by motion, cardiac activity, and depending upon the stimuli received from the internal and external environments. Based on the basic understanding of the daphnid physiology, these endpoints included acetylcholinesterase activity (AChE; biochemical endpoint), heart rate (HR, physiological endpoint) and motion related to swimming and feeding (behavioural endpoints). Material and Methods Test organismsDaphnia magna was used as a test organism. This key herbivore in aquatic food webs is a well-established model species in ecotoxicology and stress ecology, with well-established tests for immobilisation and reproduction. In addition, D. magna has also been established as a model for quantifying cardiac arrhythmia in vivo and can be used for predicting cardiotoxicity \cite{M_Whiteoak_2017}, and various techniques are available to use behavioural endpoints \cite{Bownik2021} when assessing the neurotoxic effects of environmental chemicals in this species \cite{Maggio_2021}. The test daphnids originated from a single clone (Clone V) cultured in Elendt M7 media at a constant temperature (approximately 20℃) with a 16:8 light: dark photoperiod and fed with Pseudokirchneriella subcapitata and Scenedesmus subspicatus (105 cells/L) three times a week. Algal concentration was determined by using 10AU™ Fluorometer (Turner Designs). For the exposure experiments, newly hatched individuals were randomly picked from the source culture and grown on the same feed and the same media as the stock cultures in 3-L beakers (3-L beakers, 50 individuals /L) until they were 48-h old (Instar 2; 1.3-1.4 mm). By this age, D. magna has reached an appropriate size for handling, stabilised heart rate \cite{Spicer_2001} and behaviour tracking \cite{Parolini_2018}. Endpoints The choice of endpoints was based on the expected neurobehavioral effects induced by sucralose \cite{Finn2000,Eriksson_Wiklund_2014} and acesulfame \cite{Dong2020}. To target neurobehavioral responses, we used: (1) time spent swimming and feeding (behavioural endpoints for detecting movement disorders). The rationale is that in daphnids and other filter-feeders, the thoracic limb activity is the highest-level functional manifestation of integrated neurological functions and disruption of motor control functioning would inevitably decrease ventilation, food intake, growth and survival in these animals \cite{Bownik2021}.(2) heart rate (HR, cardiac function). In daphnids, the sinoatrial node is a collection of spontaneously active nerves in a body called the cardiac ganglion and comprises a globular heart with a myogenic heart beat \cite{Spicer_2001} responding to many drugs that affect human heart rate and rhythm \cite{Campbell_2004,Dzialowski_2006} and displaying varying arrhythmias on exposure to pro-arrhythmic agents \cite{M_Whiteoak_2017}. Alterations in heart rate result from the disruption of nerve cells and/or signal transmission in the cardiac muscle. Both stimulation and inhibition in heart beating may have physiological repercussions, such as disrupted energy balance, reduced hemolymph circulation, and damage to cardiac muscle and other body tissues \cite{Santoso_2020}. (3) the whole body acetylcholinesterase (AChE) activity, a neurotoxicity biomarker commonly used in ecotoxicology. AChE is an essential enzyme controlling the nervous system, and its correlations with locomotion and feeding in crustaceans have been shown \cite{Xuereb2009,Jensen_1997}. Under neurotoxic exposure, neurotransmitters at synapses cannot be completely hydrolysed by AChE, resulting in abnormal behaviours \cite{Ren_2015}. Thus, AChE activity is a suitable biochemical endpoint to combine with behavioural responses when testing a substance with a putative neurogenic mode of action. Experimental setupThe exposure experiments with different endpoints were conducted on different occasions (Table 1). All the tests were performed under the temperature and illumination used for growing the stock culture. In addition to the test substances (sucralose and acesulfame), we used negative (Elendt M7 media) and positive (daphnia exposed to haloperidol) controls were used. For haloperidol 0 to 3.2 mg/L test concentrations were used in behavioural assays to confirm the instrument performance and experimental settings. It is an antipsychotic drug with broad pharmaceutical actions (e.g., dopamine receptor blocker and AChE inhibitor), causing behavioural effects in various animal species, thus providing a rationale for its use as a positive control, i.e., substance inhibiting neuromotor activity. The selection of the test concentration range was based on the dose-dependent inhibitory effect of haloperidol on feeding in Daphnia with IC50 value of 1.6 mg/L \cite{Furuhagen2014}. Thus, we expected haloperidol exposure at 1.6 mg/L to cause a measurable decline in locomotion, AChE activity and HR in all trials. In addition, we conducted a pilot experiment establishing a dose-response to haloperidol in the behavioural assay. Table 1. Summary of the experiments used to measure acetylcholinesterase (AChE) activity, heart rate (HR) and behavioural endpoints. The test concentrations were 0.1, 1, 10, 100 µg/L and 1 mg/L for each substance. The exposure duration 24 h. Test parameters Biochemical response Behavioral response (activity) Physiological response Endpoint AChE activity time spent swimming and feeding HR Replicates per treatment 3 3 10 Replicates per control 3 5 10 Individuals per replicate 10 1 1
