How do freshwater microalgae and cyanobacteria respond to antibiotics?

Abstract Antibiotic pollution is an emerging environmental challenge. Residual antibiotics from various sources, including municipal and industrial wastewater, sewage discharges, and agricultural runoff, are continuously released into freshwater environments, turning them into reservoirs that contribute to the development and spread of antibiotic resistance. Thus, it is essential to understand the impacts of antibiotic residues on aquatic organisms, especially microalgae and cyanobacteria, due to their crucial roles as primary producers in the ecosystem. This review summarizes the effects of antibiotics on major biological processes in freshwater microalgae and cyanobacteria, including photosynthesis, oxidative stress, and the metabolism of macromolecules. Their adaptive mechanisms to antibiotics exposure, such as biodegradation, bioadsorption, and bioaccumulation, are also discussed. Moreover, this review highlights the important factors affecting the antibiotic removal pathways by these organisms, which will promote the use of microalgae-based technology for the removal of antibiotics. Finally, we offer some perspectives on the opportunities for further studies and applications. Graphical Abstract


Introduction
Antibiotics are extensively utilized in many fields, such as human and veterinary medicine, aquaculture, livestock farming, and crop production. Global antibiotic consumption rapidly increased 65% between 2000 and 2015 [1]. Antibiotic residue can enter aquatic ecosystems through several major pathways, including the dissolution of feed pellets in fish farming [2], excretion of residual antibiotics from cultured organisms [3] and humans [4], the discharge of waste from hospitals and pharmaceutical plants [5] (Figure 1). Within the aquatic environment, antibiotics can result in the spread of antibiotic-resistant bacteria, which is creating a global environmental crisis.
Antibiotics can affect the growth of non-target organisms, even though such effects remain poorly documented [4,6,7]. Among the non-target organisms, microalgae and cyanobacteria are very sensitive to germicidal compounds. Due to their low trophic level in the food chain, any change in the population of these primary producers can profoundly affect the ecosystem's health ( Figure 1). A range of antibiotics has been reported to impact photosynthesis and macromolecule metabolism in microalgae and cyanobacteria. For example, erythromycin (the most effective antibiotic used to treat various pathogenic bacteria) acts as a photosynthesis suppressor in Microcystis aeruginosa [8] and Selenastrum capricornutum [9]. In this review, we present an overview of the effects of antibiotics on microalgae and cyanobacteria, focusing on the freshwater environment.
Upon antibiotic exposure, there is a rise in reactive oxygen species (ROS) level, which induces oxidative stress to the cells [10][11][12][13][14]. Accordingly, alteration in antioxidant systems was mostly reported due to antibiotic exposure in microalgae and cyanobacteria. On the other hand, they have evolved several ways for antibiotic detoxification, including bioadsorption, bioaccumulation, and biodegradation. The detailed adaptive mechanisms of microalgae and cyanobacteria against antibiotics will be highlighted in the second part of this review. Last, but not least, factors driving their response to antibiotics and opportunities for further studies in environmental protection are discussed.

Effects of antibiotics on microalgae and cyanobacteria
Although antibiotics are designed to target pathogenic bacteria, several bacterial receptors and/or pathways have been found to be conserved in microalgae and cyanobacteria [14][15][16], which suggest a potential threat of antibiotics on these photosynthetic organisms. The half-maximal effective concentrations (EC 50 ) and target systems of individual antibiotics on freshwater microalgae and cyanobacteria are summarized in Tables 1, S1, and S2. The available references demonstrate the impacts of antibiotics on specific target systems in these organisms, however, the mode of action of the antibiotic (i.e., inhibit biosynthesis of protein, inhibit nucleic acid synthesis, etc.) is not always described. For the reader's convenience, the antibiotics are classified based on their mode of action in bacteria obtained from drug bank (https://www.drugbank. com/). It is important to note that most of the research on the effects of antibiotics on microalgae and cyanobacteria were conducted in laboratory settings, some of which examined the antibiotic concentrations that are much higher than in natural waters. Even so, such studies may provide starting points to understand the working mechanisms of antibiotics in these organisms. In this section, we discuss the effects of antibiotics on major biological processes in freshwater microalgae and cyanobacteria including: (i) photosynthesis, (ii) oxidative stress, and (iii) macromolecules alterations.

Photosynthesis
As with higher plants, microalgae and cyanobacteria have an oxygenic photosynthesis system comprising two components known as photosystem I (PSI) and photosystem II (PSII), which are localized in the thylakoid membranes and linked together by a chain of intermediate electron carriers [39]. Previous research has suggested that antibiotics could inhibit photosynthesis in these organisms via three main mechanisms: disintegration of thylakoid membrane [40,41], inhibition of electron transport [42], and reduction of photosynthetic pigments [12] (Figure 2). Thylakoid membranes have long been known as the site of photosynthetic light reactions. It has been noted that antibiotics can cause severe damage to thylakoid membranes, thereby negatively affecting the stability of photosystems [40,41]. Accordingly, reduced photosynthesis is often observed and considered a common feature of antibiotics exposure in microalgae and cyanobacteria. For instance, erythromycin reduced the synthesis of many essential proteins in the PSII reaction center and the cytochrome b 6 f complex in the freshwater green alga S. capricornutum [9]. Consequently, a lack of these thylakoid membrane-embedded proteins can lead to a collapse of the algal photosynthetic machinery. Florfenicol and thiamphenicol at high concentrations (>1 mg L À1 ) significantly reduced the chlorophyll content of the cyanobacterium Microcystis flos-aquae, which might be a result of the disintegration of thylakoid membranes by these antibiotics [40].
As the heart of photosynthesis, PSII serves as a light-driven water-plastoquinone oxidoreductase, which mediates the production of oxygen and highenergy electrons from water for all of the downstream photosynthetic reactions to occur [43]. As noted above, erythromycin can inhibit the biosynthesis of thylakoid membrane proteins such as D1 protein, which is a key factor for maintaining the stability of the secondary quinone electron acceptor (Q B ) protein in PSII. This change can lead to a significant decrease in the electron transport rate of S. capricornutum [9]. The harmful effects of antibiotics on PSII in cyanobacteria have also been described. Amoxicillin exposure results in a slowing down of electron transport at both the donor and acceptor sides of the PSII complex in Synechocystis sp., leading to accumulation of the highly oxidizing radical P680 þ , and eventually photoinhibition [42]. Besides, amoxicillin increases the proportions of inactive PSII centers and those with small antennae, which further reduces photosynthetic performance in this cyanobacterium. Compared with PSII, PSI appears to be less sensitive to antibiotics [8]. Erythromycin significantly reduces the quantum yield of PSII in M. aeruginosa at a concentration of 0.1 mg L À1 , but a higher dose (!5 mg L À1 ) is required for the inhibition of PSI activity [8]. It has been suggested that cyclic electron flow plays a crucial role in protecting PSI against the toxic effects of erythromycin at low concentrations [8]. At present, the influences of antibiotics on PSI and the interplay between the two photosystems are not fully understood and deserve further investigation.
Natural pigments such as phycobilins, chlorophylls, and carotenoids play important roles in photosynthetic metabolism in algae. Antibiotics can affect microalgae and cyanobacteria by interfering with the production of photosynthetic pigments. For example, the chlorophyll content of the green alga Desmodesmus subspicatus decreased from 2.4 units in the control to 1.67 units, and the total carotenoids concurrently decreased from 0.59 units in the control to 0.45 units in treated cells after a four-day exposure to tylosin at 57.26 mmol L À1 [15]. Tylosin, lincomycin, and trimethoprim were found to suppress the growth of Anabaena flos-aquae, a filamentous cyanobacterium, by inhibiting the biosynthesis of chlorophylls and carotenoids, and photosynthetic oxygen evolution [15]. High concentrations of chlortetracycline (!2.0 mg/L À1 ) caused inhibition of phycocyanin synthesis in Spirulina platensis [44]. (continued) The mechanism of antibiotic interference with pigment synthesis in microalgae remains poorly understood. However, it has been suggested that observed decreases in chlorophyll-a with antibiotic exposure may be related to the disintegration of the thylakoid membrane, blockage of chlorophyll synthesis by accumulated ROS, and the combination of antibiotics with some cellular ingredients that inhibit synthesis of the light-harvesting chlorophyll a/b protein complex in microalgae [12,45]. Taken together, it appears that the influence of antibiotics on the growth of microalgae and cyanobacteria is correlated with the integrity and biological function of the photosynthesis system.

Oxidative stress
In the cells grown under optimal conditions, ROS, such as singlet oxygen, superoxide radical, hydrogen peroxide, and hydroxyl radical play fundamental roles in regulating physiological processes. However, excessive ROS formation can damage cellular components, increase membrane permeability, and lead to cell death [10][11][12][13]. Apart from their inhibition of photosynthetic activity and pigment biosynthesis, antibiotics can induce ROS production, which causes oxidative stress to microalgae and cyanobacteria [46,47]. The intracellular ROS levels in Anabaena cylindrica were significantly increased when exposed to tigecycline (100 mg L À1 ), spiramycin (30 mg L À1 ), and amoxicillin (15 mg L À1 ) [46]. Although antibiotic-induced ROS formation in algae has been observed, there is still much to learn about the specific sequence for ROS production. Following streptomycin exposure at concentrations higher than 10 mg L À1 , the expression of psbA gene, which encodes an integral membrane protein D1 of PSII, was suppressed rapidly in Chlorella vulgaris. This resulted in the blockage of electron transport chain and the accumulation of electrons, leading to excessive ROS formation [48]. In another study, examining the effects of ciprofloxacin on C. vulgaris, the authors suggested that ROS can be generated through the biotransformation process of the antibiotic via the CYP450 system. The derived ROS may lead to the activation of glutathione biosynthesis enzymes and other antioxidant enzymes to assist the cells in coping with antibiotic-induced oxidative damage [49].

Macromolecules alterations
As typical macromolecules, carbohydrates, lipids, DNA, and proteins play essential roles in cells, such as providing structural support, contributing to cell signaling, storing and retrieving genetic information. When microalgae are exposed to antibiotics, the alterations in macromolecules composition and structures can occur and several biological processes can be disrupted [50].

DNA
Antibiotics are known to trigger DNA damage and genome instability in bacteria [51]. However, very few studies have been conducted to evaluate their effects on the DNA of microalgae. DNA conformational change upon antibiotic treatment has been reported in Pseudokirchneriella subcapitata, Scenedesmus quadricauda, Scenedesmus obliquus, and Scenedesmus acuminatus [23]. The exposure of those green algae to chloramphenicol and roxithromycin promotes DNA aggregation through the transformation of Besides, antibiotics also adversely affect the major macromolecules such as protein, DNA, lipid, and carbohydrate.
right-handed B-DNA into left-handed Z-DNA [23]. In addition, folic acid is an essential factor necessary for DNA synthesis and the maintenance of genome integrity [52]. It has been reported that the growth inhibitory effect of sulfadiazine on C. vulgaris was reversed by the supplement of folic acid to the medium [31]. This finding suggested that harmful effects of sulfadiazine are associated with the blockage of folate biosynthesis in C. vulgaris.

Protein
The effects of antibiotics on protein synthesis in different microalgae and cyanobacteria could be wide and even opposite. For example, oxytetracycline dihydrate and sulfamethoxazole promoted protein synthesis in M. aeruginosa and Chlamydomonas microsphaera [53]. The exposure of Raphidocelis subcapitata to erythromycin, clarithromycin, and ciprofloxacin resulted in elevated total protein [54]. The protein content of C. vulgaris was also increased significantly following treatment with sulfonamides [50]. The authors suggested that increased protein content may be attributed to the increase in enzyme synthesis, especially antioxidant enzymes or enzymes related to energy-producing fractions [50,54]. On the other hand, antibiotics have been shown to inhibit protein synthesis in the cyanobacterium Anacystis montana [55] and two green algae Dictyosphaerium pulchellum and Micractinium pusillum [56]. In addition, many studies have found the inhibitory effects of antibiotics on the synthesis of proteins and enzymes involved in photosynthetic machinery in microalgae and cyanobacteria, as mentioned earlier in the Photosynthesis section. The altered protein secondary structures in four species of green microalgae (P. subcapitata, S. quadricauda, S. obliquus, and S. acuminatus) were also observed under antibiotics exposure. For example, chloramphenicol and roxithromycin were found to reduce a-helix but increase b-sheet structures, suggesting possible conversion of a-helix to b-sheet in these microalgae [23].

Lipid and carbohydrate
Although research into the influence of antibiotics on lipid metabolism in microalgae and cyanobacteria has been limited, there is no doubt that antibiotics can induce alterations in fatty acid metabolism [29,44,57,58]. The mixture of ceftazidime and gentamicin sulfate increased the expression of AcoX1/AvoX3, corresponding to acyl-CoA dehydrogenase, thus enhancing fatty acid degradation in the green alga C. vulgaris [57]. On the other hand, chlortetracycline (1.0 mg/L À1 ) exposure significantly increased the proportion of c-linolenic acid (C 18:3 ), an essential polyunsaturated fatty acid with various clinical indications in S. platensis [44]. Another example is the increased monounsaturated fatty acids content of Synechocystis sp. PCC 6803 in response to sulfamethoxazole and tetracycline [59]. The saturated fatty acid content in S. obliquus was reduced by 12.5% when the concentration of sulfamethazine and sulfamethoxazole antibiotic mixture was increased, whereas the unsaturated fatty acid content increased by 13.8% [29]. The higher unsaturated fatty acid content appeared to induce membrane fluidity changes as a protective mechanism against sulfamethazine and sulfamethoxazole. Antibiotics can interfere with the metabolic processes related to energy storage and thereby affect the carbohydrate content of microalgae [29,44]. For example, the total carbohydrate content in S. obliquus was reduced from 25.0% to 20.1% upon treatment with sulfamethazine and sulfamethoxazole mixture [29]. This result is consistent with decreased photosynthetic activity caused by sulfamethazine and sulfamethoxazole, as carbon fixation by photosynthesis is the primary source of carbohydrates. In addition, tetracycline exposure can lead to cellular plasmolysis, thylakoid lamellae deformation, and starch granule deposition in C. vulgaris [13]. Further research on different antibiotics in a broader range of species is required to understand how these toxic compounds affect the lipid and carbohydrate metabolism.

Types of action of antibiotics in microalgae and cyanobacteria
As described above, antibiotics may affect the growth of microalgae and cyanobacteria by targeting their photosynthetic and other cellular systems. However, the underlying molecular mechanisms may be more complicated than previously thought. A single antibiotic may exhibit different biological effects depending on the concentrations used, known as hormesis or "low-dose stimulation and high-dose inhibition" [60]. Furthermore, a combination of two or more antibiotics would lead to more sophisticated effects, which can be synergistic or antagonistic [61]. Those dynamic mechanisms of action of antibiotics in microalgae will be discussed below.

Hormesis
The "hormetic effect" concept can be defined as the biphasic response of an organism to a stressor, with contrasting effects depending on the stressor concentrations. Many toxic chemicals can stimulate biological processes at low concentrations, but become harmful and reduce cell growth at higher concentrations [62]. The hormetic effects of antibiotics on microalgal growth have been reported [34,62,63]. Tilmicosin at low concentrations (0.01-2 mg L À1 ) promoted the growth of Chlorella PY-ZU1, but significantly inhibited photosynthesis and growth when the concentration increased above 5 mg L À1 [44]. The authors suggest that low levels of tilmicosin may stimulate repair and maintenance mechanisms (i.e., producing proteins related to cytoprotection and antioxidant enzymes), but high concentrations could irreversibly damage these protective systems. Furthermore, high concentrations of tilmicosin can also inhibit protein synthesis in chloroplasts by binding to the 23S rRNA on 50S ribosomal subunits, thereby reducing overall photosynthesis in this algal species.
Some antibiotics such as erythromycin, chlortetracycline hydrochloride, and amoxicillin exhibited hormetic effects on microcystin production in Microcystis [34,64,65]. For instance, low doses of erythromycin (10-60 mg L À1 ) stimulated the biosynthesis of intracellular and extracellular microcystins in M. aeruginosa, while higher doses (100-150 mg L À1 ) limited their production [34]. Low concentrations of chlortetracycline hydrochloride (2 and 5 mg L À1 ) also enhanced microcystin-LR production in this cyanobacterium [64]. Amoxicillin at an environmentally relevant concentration (0.1-1 mg L À1 ) stimulated cell growth and the synthesis and release of microcystin in M. aeruginosa [65]. Cyanotoxins play essential roles in grazing defense, allelopathy, nutrient uptake, oxidative stress response, and carbon-nitrogen metabolism [66]. Hence, the stimulated production of cyanotoxins may be a stress response caused by antibiotic exposure at low concentrations [64]. To date, studies regarding the hormetic effects of antibiotics on cyanotoxin production have focused only on microcystin. Thus, additional studies are necessary to reveal the mechanisms of action of antibiotics in the production and release of other cyanotoxins.
Antibiotics are generally detected in the order of "hospital effluents (higher mg L À1 range)" > "municipal wastewater (lower mg L À1 range)" > "sea, lake, and groundwater (ng L À1 range)" [67]. Based on the information presented in Table 1, the EC 50 values of several antibiotics are higher than mg L À1 levels. In other words, the low-dose growth-stimulating effects of some antibiotics on microalgae and cyanobacteria may frequently occur in aquatic ecosystems and therefore deserve greater public concern. Further studies should be conducted to explore how this issue may threaten public health and the environment.

Multiple antibiotic exposure
In aquatic ecosystems, microalgae and cyanobacteria are often exposed to a mixture of different antibiotics, rather than individual ones. Hence, the final impacts of antibiotics are the result of their combined effects. While the concentrations of individual antibiotics are relatively low in the environment, their coexistence can lead to potential threats to aquatic biota [68]. Depending on the specific mechanism of action of each antibiotic and their ratios in solution, the combined effects of antibiotic mixtures could be synergistic or antagonistic [69]. Synergism refers to a greater effect of two or more antibiotics given together than would be seen from the sum of individual effects. In contrast, antagonism refers to the inhibition of one drug by another [69]. Synergism can be divided into two groups: facilitating and complementary actions [22]. Facilitating action implies that one drug increases the pharmacological activity of another drug, whereas complementary action could occur when both drugs have the same target at different sites or overlapping sites, or different targets of the same pathway [22]. For example, joint treatment with erythromycin and enrofloxacin at low concentrations (0.01 mg L À1 ) synergistically inhibited chlorophyll biosynthesis in C. vulgaris, although neither erythromycin (0.02 mg L À1 ) nor enrofloxacin (0.03 mg L À1 ) affected the algal chlorophyll content when they were added separately [18].
Antagonistic effects of two antibiotics could be attributed to competition for uptake or the same binding sites, or suppression of one drug activity by another [68]. At very low concentrations, antagonism is the predominant interaction in most antibiotic binary mixtures (e.g., erythromycin, levofloxacin, norfloxacin, tetracycline, and amoxicillin) in the filamentous cyanobacterium Anabaena sp. CPB4337 [68]. However, the combined toxic effects of antibiotic mixtures are not consistent [18]. Interactions can shift from synergism to antagonism or vice versa according to the antibiotic doses and mixing ratios [37,70]. For example, the antagonistic interactions mentioned above between five antibiotic binary mixtures in Anabaena sp. CPB4337 became synergistic when the antibiotic concentrations increased [68]. The combined toxic effects of spiramycin and ampicillin on M. aeruginosa also varied from synergism to antagonism with an increasing proportion of spiramycin in the mixtures [37]. In another study, amoxicillin and spiramycin mixed at a 1:1 ratio exhibited a shift from synergism to antagonism with increasing concentrations of both antibiotics in M. aeruginosa [70]. This interchange between the action modes of combined antibiotics was also observed in microalgae. For example, a mixture of chlortetracycline and oxytetracycline acted antagonistically, while the combinations of chlortetracycline þ enrofloxacin and oxytetracycline þ enrofloxacin showed additive effects in the green alga R. subcapitata [71].
It should be emphasized that the nature of the interaction between combined antibiotics is strongly related to the investigated strains [37]. The results from one indicator organism may not apply to other organisms. Owing to the lack of molecular analysis, the detailed mechanisms underlying the combined effects of antibiotics on microalgae and cyanobacteria are still unknown. In some cases, a combination of two drugs could lead to additional risks. For example, when used in combination with ampicillin, spiramycin can promote the growth of M. aeruginosa and microcystin production, thus increasing potential harmful effects [37]. Hence, it is urgent to address the mechanisms of these joint toxic interactions [18]. Field experiments, such as microcosms and mesocosms would be necessary to obtain practical information.

Adaptation of microalgae and cyanobacteria to antibiotic exposure
Microalgae have developed multiple defense mechanisms against the toxic effects of antibiotics. As mentioned previously, antibiotic exposure has been linked to ROS accumulation, which can cause cellular damage and ultimately cell death in many organisms. Indeed, enhanced activities of the antioxidant system were observed in diverse microalgae and cyanobacteria, which play an essential role in removing the abundant ROS generated upon antibiotic stimulation [14,18,72]. In the meantime, these organisms have evolved different pathways to remove antibiotics from the environment and culture media, including: bioadsorption, bioaccumulation, and biodegradation. All four mechanisms will be described in this section.

Activation of cellular antioxidant system
In photosynthetic organisms, integrated antioxidant systems, including antioxidant enzymes ( [11,18,62]. As the first defensive barrier, SOD can convert superoxide to oxygen and H 2 O 2 . Then, CAT and GPx convert H 2 O 2 to oxygen and water [73]. For instance, the activities of SOD, POD, and CAT increased 1.27-, 1.55-, and 1.84-fold, respectively, in C. vulgaris exposed to erythromycin [18], suggesting that elevated activities of antioxidant enzymes are necessary for alleviating oxidative damage caused by ROS in response to antibiotic exposure. In line with this finding, C. pyrenoidosa showed at least a 1.82-fold increase in SOD and CAT activities when the cells were exposed to roxithromycin for 96 h (short-term) at a high dose (2 mg L À1 ) or 14-21 d (long-term) at a low dose (0.1-0.25 mg L À1 ) [72].
In addition to enzymatic systems, non-enzymatic antioxidants can remove ROS independently or serve as a cofactor for GPx [74]. To reduce the toxic effect of gentamicin, GSH entered ROS metabolism in Synechocystis sp. PCC 6803 and was oxidized to glutathione disulfide (GSSG) during the detoxification of peroxides [75]. In M. aeruginosa, the increase in GSH content, along with the increased activity of GST, reduced the toxic effects of amoxicillin [14]. This finding implies that, like GST, GSH also plays a vital role in removing excess ROS caused by antibiotic exposure. More recently, it has been found that microalgae exhibit different antioxidant defense mechanisms, depending on the type and concentration of antibiotics [76]. At a low concentration of clarithromycin, only the non-enzymatic antioxidant system was activated in C. vulgaris, whereas at a higher concentration, the activity of antioxidant enzymes was also increased [76]. Sulfamethoxazole, erythromycin, and clarithromycin induced SOD activity in R. subcapitata at different concentrations (1.85-8.3 mM, 17-40.8 nM, and 0.76 nM, respectively) [54].
In photosynthetic organisms, carotenoids are vital non-enzymatic antioxidants that protect photosystems by deactivating the excited chlorophylls, which may otherwise form triplet states and react with oxygen to form singlet oxygen and other ROS [77][78][79]. The conjugated double bond in the carotenoid structure allows them to accept electrons from reactive species and, therefore, negate the adverse effects of ROS [80][81][82]. Sulfamethazine and sulfamethoxazole mixture at concentrations from 0.15 to 0.35 mg L À1 increased the carotenoid content of S. obliquus approximately 1.2 fold [29]. The authors suggested that carotenoids could act as protective agents against excessive ROS produced in the chloroplasts under antibiotic exposure conditions. Taken together, there is no doubt that the antioxidant systems play a critical role in the survival and resistance of microalgae against antibiotics.

Bioadsorption
Microalgae have been found to possess several means for antibiotic detoxification including bioadsorption, bioaccumulation, and biodegradation [11,83] (Figure 3, Table S3). Bioadsorption is an extracellular process that involves the passive binding of substances to the cell surface [83,84]. The adsorption of antibiotics to microalgae is highly concentration-dependent [72]. In various microalgae (Chlorella sp. Cha-01, Chlamydomonas sp. Tai-03, and Mychonastes sp. YL-02), bioadsorption was found to occur within a few minutes of treatment at relatively high antibiotic concentrations ranging from 25 to 150 mg L À1 [33]. In another study, the adsorption of roxithromycin was observed on day 3 in C. pyrenoidosa cells exposed to low concentrations of 0.1 and 0.25 mg L À1 , whereas, bioaccumulation was only observed after 14 d of treatment [72]. Approximately 11% of trimethoprim and sulfamethoxazole removal, 13% of carbamazepine removal, and 27% of triclosan removal from lake water was attributed to microalgaemediated bioadsorption [85]. The microalgal cell wall has negative charges from dominant functional groups such as carboxyl and phosphoryl. Therefore, pollutants with cationic groups are readily adsorbed by the cell surface via electrostatic interactions [83]. The bioadsorption capacity of microalgae has been linked to the hydrophobicity, structure, and functional groups of pollutants [83]. Ciprofloxacin and sulfadiazine adsorbed on Chlamydomonas sp. Tai-03 cells mainly interacted with the carbonyl and amine groups of tryptophan-like substances and hydroxyl groups of extracellular polymeric substances (EPS) [84]. The authors also reported that microalgal EPS showed much higher adsorption potential for ciprofloxacin than for sulfadiazine. In addition, microalgal EPS can form a hydrated biofilm matrix that serves as an external digestive system by holding extracellular enzymes and allowing them to metabolize organic compounds [83]. Several cyanobacteria can produce large quantities of EPS [86]. However, to date, the adsorption behavior of antibiotics onto cyanobacteria have rarely been studied, though it has been shown that M. aeruginosa exhibited higher rates of bioadsorption of tetracycline than C. pyrenoidosa [87].

Bioaccumulation
While bioadsorption is an extracellular binding of antibiotics, bioaccumulation is defined as the intracellular uptake and accumulation of substances in an organism [83]. Several microalgal species can accumulate antibiotics in their cytosolic compartments [88][89][90]. Macrolides and other hydrophobic drugs can diffuse across the lipid bilayer membrane, whereas small hydrophilic drugs use pore-forming porins to penetrate cells [32,91]. Bioaccumulation is generally considered a pre-step for biodegradation [92]. The bioaccumulation of antibiotics in microalgae can vary depending on the exposure time and the antibiotic level. For instance, roxithromycin was not detected in C. pyrenoidosa cells exposed to concentrations ranging from 0.2 to 0.25 mg L À1 for the first 10 d. However, after 3 d of exposure to 1.0 mg L À1 of roxithromycin, the antibiotic was detected in algal cells and continuously increased during the 21-d treatment period [72].
Membranes play critical roles in the penetration of antibiotics [93,94]. The outer membrane of Gram-negative bacteria is an evolving barrier, which prevents several antibiotics from penetrating the cell [91,[94][95][96]. As a result, Gram-negative bacteria show a higher level of resistance to antimicrobial agents than Gram-positive bacteria [94]. The cyanobacterial cell wall was identified as a combination of Gram-positive and Gram-negative properties with a thick peptidoglycan layer and an outer membrane [97]. How this unique architecture affects the accumulation of antibiotics in cyanobacteria remains unclear. Additionally, it should be highlighted that the over-accumulation of antibiotics in algal cells can lead to overproduction of ROS and consequent damage to cellular components. Further research is necessary to determine the accumulation capacity and location of antibiotics in microalgae and cyanobacteria.
Since the intermediate and final metabolites of the biotransformation process may be less (detoxification) or more (bioactivation) toxic than the parental compounds, toxicological evaluation of the products is needed [92]. Kiki et al. [88] identified potential transformation products from ten antibiotics frequently detected in water bodies (sulfamerazine, sulfamethoxazole, sulfamono-methoxine, trimethoprim, clarithromycin, azithromycin, roxithromycin, lomefloxacin, levofloxacin, and flumequine), metabolized by four microalgae, namely, Haematococcus pluvialis, S. capricornutum, S. quadricauda, and C. vulgaris. The authors highlighted that most of the transformation products identified appeared to be less toxic than their corresponding precursors. Pan et al. [89] demonstrated that climbazole-alcohol, a degradation product of climbazole by S. obliquus, was less toxic than its parent compound. The algal treatment step also reduced the toxicity of ceftazidime from 93% to 55% (based on the death rate of rotifers) [100]. In contrast, anhydrotetracycline, a metabolite of tetracycline degraded by C. vulgaris, exhibited more potent toxicity toward algal cells than its parent compound [13].
In many algal species, biodegradation is considered the predominant mechanism to eliminate antibiotics from the dissolved fraction, whereas bioadsorption and bioaccumulation have minor and incomplete roles [72,[88][89][90]. Tetracycline was rapidly removed by M. aeruginosa (over 98% after 24 h) mainly via biodegradation [87]. The remediation mechanism of antibiotics in microalgae largely depends on their initial concentrations [11]. For example, the removal of florfenicol was attributed solely to biodegradation in Chlorella sp. L38 cells at an initial concentration of 46 mg L À1 . However, once the florfenicol concentration increased to 159 mg L À1 , the antibiotic was removed through biodegradation as well as bioaccumulation and bioadsorption [11]. Taken together, these results indicate that microalgae could be used to remove antibiotics and other pollutants from urban wastewater. Unlike fungi and bacteria that require organic carbon for growth, microalgae can use CO 2 for photosynthesis, thus creating a "zero-waste system" without further carbon requirements [83]. With the potential of co-producing valueadded products, microalgae-based technology has been highlighted as a promising method for antibiotic removal from wastewater [92]. However, the toxicological effects of the intermediate metabolites from antibiotic breakdown should be taken into consideration. Also, further studies on the mechanisms of antibiotics removal in cyanobacteria could enable us to take advantage of these photosynthetic organisms in mitigating antibiotic pollution.

Factors regulating the response of microalgae and cyanobacteria to antibiotics
The following sections summarize several abiotic (e.g., nutrients, light, temperature, and pH) and biotic (e.g., microbial communities) factors known to affect physiological responses of microalgae and cyanobacteria to antibiotic exposure.

Nutrients
As described earlier, antibiotics can stimulate cell death by inhibiting DNA replication, protein synthesis, and cell wall synthesis. It is worth noting that nutrients are key factors affecting cellular biological activity and, therefore, have a remarkable influence on microalgal metabolism of antibiotics. For growth, algae require specific amounts of essential nutrients such as carbon, nitrogen, phosphorus, and sulfur [101]. Of these, nitrogen, phosphorus, and carbon are fundamental macronutrients that are considered crucial regulators in controlling microalgal response to antimicrobial agents [102,103]. For instance, the growth inhibitory effect of spiramycin, a protein synthesis inhibitor, is nitrogendependent in M. aeruginosa [104,105]. At high levels of nitrogen (5-50 mg L À1 ), the increase in spiramycin concentration from 100 to 400 ng L À1 led to a >12% decrease in the algal growth rate. However, under nitrogen-deficient conditions, spiramycin showed lower toxicity and even stimulated algal growth at a concentration of 100 ng L À1 . The mode of action of spiramycin is to block protein synthesis and translocation, which is more sensitive in log-phase cells [106]. Its reduced toxicity at low nitrogen concentrations can be explained by the low algal growth due to starvation. In other words, nitrogen affected the toxicity of spiramycin on M. aeruginosa by regulating protein synthesis.
Moreover, nitrogen may also affect the biodegradation rate of spiramycin through the GSH pathway [104].
Phosphorus is also a key factor in cellular responses to antibiotics. In M. aeruginosa, low phosphorus concentrations (0.05-0.2 mg L À1 ) inhibited the synthesis of the cell wall and the penicillin-binding proteins located inside the cell wall, which are responsible for the transportation of amoxicillin into the cells [107]. Therefore, the toxic effects of amoxicillin were not significant at low phosphorus levels, due to the reduced entry of amoxicillin into the cells. In another study, the phosphorus level influenced the interaction between chloramphenicol and M. aeruginosa [108]. During the early period of chloramphenicol exposure, M. aeruginosa cells grown at a higher phosphorus concentration (5 mg L À1 ) exhibited greater antioxidant responses and higher protein synthesis, which could accelerate the subsequent removal of chloramphenicol via biodegradation. The exact mechanism by which phosphorus levels regulate cyanobacterial adaptation to antibiotics remains unclear, and further investigation is needed.
Microalgae and cyanobacteria take up and convert CO 2 into useful organic products such as carbohydrates, lipids, and other bioactive metabolites. A recent study highlighted that the removal efficiency of cefradine by M. aeruginosa and C. pyrenoidosa could be enhanced by the addition of 20% and 10% CO 2 , respectively, to the culture [109]. The addition of organic substrates also promoted the removal of antibiotics by microalgae [83]. Supplementing growth substrates can boost the activities of essential enzymes involved in the degradation of organic contaminants [83]. For example, the removal efficiency of sulfamethoxazole by C. pyrenoidosa was significantly improved from 6.05% to 99.3% with the addition of sodium acetate as an electron donor [17,110]. Similarly, sodium acetate promoted ciprofloxacin degradation by Chlamydomonas mexicana [17]. In contrast, sodium formate negatively affects the removal of sulfamethoxazole by C. pyrenoidosa and ciprofloxacin by C. mexicana [17,110]. The authors assumed that this result could be associated with carbon catabolite repression, in which readily degradable carbon sources such as sodium formate and glucose are preferentially consumed, leading to weaker degradation of the target compounds.

Light
Light is a fundamental energy source that determines the success or failure of microalgal cultivation. Some antibiotics, such as quinolones, tetracyclines, sulfonamides, tylosin, and nitrofuran, are light-sensitive [111].
Under simulated sunlight irradiation, 98% of sulfonamides were removed from the water after 24 h [112]. The photochemical decomposition of an antibiotic is related to its absorption spectrum and light intensity [111]. It should be emphasized that incomplete photochemical decomposition can lead to the formation of more or less toxic or stable products [111,113]. For example, chlortetracycline can be degraded by UV light into segments that are more toxic to S. obliquus than chlortetracycline itself [113]. Cefradine is stable and not highly susceptible to photodegradation under sunlight irradiation, but rapidly underwent photolysis when subjected to UV irradiation in the presence of Chlamydomonas reinhardtii [114]. C. reinhardtii might provide the surface area for the reaction, thereby improving the photodegradation of cefradine. In addition to photolysis, light can influence algal growth and enhance the efficiency of microalgae-mediated removal of antibiotics. The most efficient degradation of cefradine and amoxicillin by M. aeruginosa can be achieved at light intensities of 5500 and 8500 lx, respectively [109]; the optimum light intensity for cefradine and amoxicillin elimination by C. pyrenoidosa is 8500 lx, indicating that C. pyrenoidosa requires stronger light than M. aeruginosa for optimal performance. Taken together, the light intensity and wavelength are related to the removal efficiency of antibiotics, both by photolysis and microalgae-mediated degradation.

Temperature
The solubility and chemical reaction rate of antibiotics and the organism metabolic activity are closely related to temperature. Thus, the temperature is one of the main factors responsible for antibiotic toxicity [115]. In general, the growth rate of microalgae and cyanobacteria increases with increasing temperature until an optimum level is reached, and then declines [7]. Therefore, the water temperature can affect algal sensitivity to antibiotics by regulating cell growth and metabolism. Optimal temperatures (30 C) promoted both the growth and enrofloxacin uptake of M. aeruginosa, thereby increasing its sensitivity to enrofloxacin [7]. However, at 20 C, M. aeruginosa was less susceptible to enrofloxacin than at 30 C. An opposite trend was reported for the green alga S. obliquus, indicating that temperature and physiological differences influence the sensitivity of algae to antibiotics [7]. pH pH can affect the solubility and ionization of antibiotics as well as the physiology of algal cells. When dissolved CO 2 is depleted by microalgal photosynthesis, the pH of the culture increases [26]. The elevation of pH values can lead to increased ionization of acidic antibiotics and reduce their toxicity [30,116]. Alkaline conditions (pH 9) were more favorable for cefradine removal by C. reinhardtii, than acidic conditions (pH 5) [114]. Also, at low pH (pH 5) and low ionic strength (1 mM Na þ and 2 mM Mg 2þ ), the toxicity of kanamycin (0.1 mg mL À1 ) and tobramycin (0.1 mg mL À1 ) to S. elongatus and M. aeruginosa was higher than under higher pH and higher ionic strength [117].

Other environmental factors
The co-contamination of antibiotics and metals occurs naturally in aquatic ecosystems. Metals can alter antibiotic properties, enhancing the entry of antibiotics into cells [118,119]. A mixture of four antibiotics (amoxicillin, ciprofloxacin, sulfamethoxazole, and tetracycline) reduced the harmful effects of copper sulfate (CuSO 4 ) and enhanced the production of microcystin in M. aeruginosa [120]. Although the application of CuSO 4 effectively prevented the formation of cyanobacterial blooms, antibiotic contaminants might interfere with the bloom control effect of CuSO 4 , favoring bloom formation and creating a more severe threat to the environment.
Engineered nanoparticles are used extensively in several fields, such as pharmaceuticals, biomolecule detection, and cancer therapy. Engineered nanoparticles are defined as materials containing nanoscale structures ranging from 1 to 100 nm [121]. Once released into the aquatic ecosystem, they can enter the food web and affect microbial, plant, invertebrate, and vertebrate communities [121]. At a lower concentration of cerium oxide nanoparticles (1 mg L À1 ), S. obliquus was more effective at removing sulfonamides than at a higher concentration (50 mg L À1 ) [99]. This could be due to the upregulation of genes encoding hydrogenase and oxidoreductase by cerium oxide nanoparticles. A better understanding of this co-selection mechanism would enable developing a more efficient antibioticremoval process using microalgae and cyanobacteria.

Phytoplankton communities
The behavior of phytoplankton communities, as primary producers, toward antibiotics has rarely been explored in natural habitats. A microalgal consortium, consisting of five common freshwater algal species (S. obliquus, C. mexicana, C. vulgaris, Ourococcus multisporus, and Micractinium resseri) exhibited a higher sensitivity to enrofloxacin than individual microalgal species [10]. However, the removal efficiency of enrofloxacin by the microalgal consortium was comparable to that of C. mexicana and C. vulgaris, the most effective species. Another study demonstrated that the addition of S. obliquus cells to M. aeruginosa at a density ratio of 3:1 (M. aeruginosa:S. obliquus) made M. aeruginosa more sensitive to enrofloxacin than under the single species scenario [7]. Thus, species competition may act as a stress factor, leading to disproportionate effects of antibiotics on specific microalgal groups. These effects can result in population declines and affect the structure of primary producer communities [7].

Future perspectives and recommendations
Considerable efforts have been made to elucidate the effects of antibiotics on non-target aquatic organisms. However, most studies to date on the interaction between microalgae and antibiotics have been conducted at a laboratory scale but have not been elucidated in field research. There is still a long road ahead before a comprehensive picture can be obtained regarding the effects of antibiotics on microalgae and cyanobacteria, and vice versa. In this context, we will discuss some of the main unresolved questions about these bidirectional interactions.
Can microalgae and cyanobacteria become antibiotic-resistant?
As a natural phenomenon, antibiotic resistance in bacteria is induced under antibiotic selective pressures [122]. The acquisition of antibiotic resistance in microalgae and cyanobacteria after repeated exposure to the same antibiotic has been reported [34,123]. After an initial treatment with erythromycin at a high concentration of 60 mg L À1 , M. aeruginosa developed antibiotic resistance with the second exposure, specifically by enhancing the photosynthesis rate, antioxidant activity, and microcystin production, while reducing oxidation stress and growth inhibition [34]. In another study, the toxic effect of tetracycline on M. aeruginosa was relatively lower during the second exposure than that of the first exposure, suggesting that the cyanobacterium achieved some resistance to tetracycline [123]. Additionally, the 96 h EC 50 values of C. pyrenoidosa and M. aeruginosa after a second exposure to chlortetracycline were higher than those with the first exposure [124]. Using in silico modeling approaches, Sanderson et al. [125] ranked the overall relative order of vulnerability toward antibiotics to be daphnia > fish > algae.
The lower susceptibility of algae and cyanobacteria to some antibiotics implies that they are either naturally resistant to these compounds or might have acquired resistance through de novo mutation under the increasing selective pressure of antibiotics in the environment [126]. More research is required to provide insight into the mechanisms underlying the adaptation of these primary producers to contaminant antibiotics in the environment.
Can microalgae and cyanobacteria acquire and transfer antibiotic resistance genes (ARGs)?
The aquatic environment is a major reservoir of antibiotic-resistant bacteria and antibiotic resistance genes (ARGs) [127]. Antibiotic-resistant bacteria can spread their ARGs into other indigenous microbes [128,129]. As such, ARGs could become enriched, selected, or horizontally transferred between different bacterial taxa [127,130]. Dias et al. [131] suggested that freshwater cyanobacteria (Planktothrix agardhii strains LMECYA 153 A, LMECYA 280, and LMECYA 303) can acquire and transfer ARGs, including a class-1-type integrin (int 1) (a genetic element that can capture and express diverse resistance genes) and a sul1-type gene (resistance to sulfonamides). One question raised by this finding is how cyanobacteria contribute to the evolution of antibiotic resistance in bacteria. Although it is well-known that plasmids play a significant role in disseminating ARGs, non-plasmid-encoded genes associated with resistance to trimethoprim and quinolones were found in these cyanobacterial species [131]. So far, the related information for microalgae is not available.

Can antibiotics induce mutations in microalgae and cyanobacteria?
There is growing evidence that antibiotic resistance can emerge at low environmental concentrations of antibiotics [132]. As with other aquatic organisms, microalgae and cyanobacteria are often exposed to multiple antibiotics at a wide range of concentrations. The selection pressure caused by contaminant antibiotics in the aquatic ecosystem could lead to the evolution of antibiotic resistance in microalgae via de novo mutation [126]. As a consequence of DNA damage caused by zeocin, the DNA repair system and delayed cell cycle progression were activated in C. reinhardtii to maintain the integrity of genomic DNA [133]. Similarly, the addition of zeocin at the beginning of S. quadricauda growth cycle blocked cell cycle progression in the G2 phase [134]. These data suggest that mutations in microalgae can be induced by antibiotics or occur during DNA repair processes. Research efforts are needed to obtain a more comprehensive understanding of antibiotic effects on the mutation rate in microalgae and cyanobacteria.
Are antibiotics related to cyanobacterial blooms?
As one of the most harmful eutrophication impacts, cyanobacterial blooms create various problems for aquatic ecosystems and humans. In addition to traditional factors such as nutrients, temperature, and light, novel factors such as coexisting antibiotic contaminants should be investigated to better understand these compounds' potential in regulating cyanobacterial blooms [53,135]. There is considerable evidence suggesting that some antibiotics at environmentally relevant concentrations, which are lower than the toxic threshold, can promote cyanobacterial blooms [136]. Florfenicol, thiamphenicol, erythromycin, and levofloxacin at concentrations of 0.001-1 lg L À1 promoted the growth of M. flos-aquae [12,40,137]. Similarly, amoxicillin (600 ng L À1 ), spiramycin (100 ng L À1 ), ciprofloxacin (50-200 ng L À1 ), sulfamethoxazole (100-200 ng L À1 ), and tetracycline (0.05 mg L À1 ) enhanced the photosynthetic activity and growth of M. aeruginosa [123,138,139]. In the environment, the microbial community plays a vital role in the formation of cyanobacterial blooms [140]. Therefore, antibiotics may indirectly affect the extent of cyanobacterial blooms by changing the microbial community structure. Antibiotics could also contribute to the toxicity of cyanobacterial blooms as mentioned previously (see section Hormesis above). The influence of antibiotic contamination on cyanobacterial blooms may have been underestimated. Hence, further in-depth studies are required to elucidate the correlation between coexisting antibiotics and the extent of cyanobacterial blooms, as well as the synthesis and release rates of cyanotoxins in the environment.
How control methods for harmful algal blooms (HABs) may affect microalgae-antibiotics interactions?
Enormous efforts have been made to develop diverse strategies (physical, chemical, and biological methods) for harmful algal blooms (HABs) control. Many studies elucidated the advantages and disadvantages of these strategies. However, their effects on the physiological responses of phytoplankton to antibiotics have remained unknown. The lysis of microalgae by conventional methods may lead to the tradeoffs between controlling HABs and secondary pollution [141]. Microalgae can accumulate antibiotics in cytosolic compartments and break them down into intermediate metabolites. Additional research is required to explore whether the lysis of microalgal cells by physical disruption or chemicals accelerates the release of these intermediate metabolites into freshwater environments. Since the excessive nutrient is the main reason leading to the formation of HABs, the nutrient-load reduction using dredging [142], oxygen nanobubble [143], and phosphorus-binding clays [144] is considered a useful strategy to prevent HABs. Such treatment may affect the microbial community structure. For example, dredging reduces the abundance and diversity of the bacterioplankton community [145]. As mentioned above, the nutrient and phytoplankton communities affect the responses of microalgae and cyanobacteria to antibiotics (see section "Factors regulating the response of microalgae and cyanobacteria to antibiotics"). Accordingly, it will not be surprising if these methods impact the microalgae-antibiotics interactions. Flocculation is a convenient and cost-effective method to remove HABs [146]. Sand and clay are commonly used to flocculate algae, bacteria, and other particulates to the bottom layer of the water body [141,147]. Further studies are needed to determine whether these flocculants affect the bioadsorbption capacity of microalgae and cyanobacteria to antibiotics.

New insights from algomics
Algomics is the application of genomic and post-genomic approaches to elucidate the cellular physiology and metabolism of microalgae and cyanobacteria [148]. The use of integrated algomics is flourishing and provides a rich source of genetic information [148]. By combining various omics datasets, researchers have greater opportunities to recognize genes and metabolic pathways crucial for microalgal cells under certain conditions [149]. Hence, algomics is expected to be a useful approach that can inform a more complete picture of how microalgae and cyanobacteria respond to antibiotics.

Conclusions
Since the introduction of antibiotics into clinical use, the bulk of antibiotics has spread in the environment and driven microbial evolutionary dynamics. To better understand the adverse effects of antibiotics on ecosystems, non-target organisms should be considered besides the bacterial community. This review emphasized the impacts of antibiotics on freshwater microalgae and cyanobacteria and their responses. The main conclusions are as follows: 1. Antibiotics can affect several cellular pathways in microalgae and cyanobacteria including photosynthesis, carbohydrate and lipid metabolism, DNA and protein synthesis. 2. Microalgae and cyanobacteria have developed different adaptation mechanisms to survive under antibiotic selective pressure, such as antioxidant responses, bioadsorption, bioaccumulation, and biodegradation. With the potential of co-producing value-added products, microalgae-based technology can become a promising method for antibiotic removal from wastewater. 3. In the aquatic ecosystem, microalgae and cyanobacteria are often exposed to multiple antibiotics. The combination of two or more antibiotics could result in more sophisticated effects such as synergistic or antagonistic, compared to single antibiotic exposure, and these effects are also different depending on microalgal strains. In addition, the low-dose growth-promoting effects of some antibiotics on microalgae and cyanobacteria may occur and deserve greater attention. 4. There are several abiotic factors that may drive algal responses to antibiotics by affecting their physiology and altering the antibiotic properties. Also, biotic factors such as microbial communities can elevate the impact of antibiotics on algae. 5. Microalgae and cyanobacteria can acquire antibiotic resistance after repeated exposure to the same antibiotic. Questions remain to be answered whether this happens through de novo mutation, or they can receive and transfer ARGs from the environment and other species. The continuous development of omics technology will provide a more comprehensive overview of algal responsive mechanisms to antibiotics, further enhancing the application of these organisms in antibiotic removal.

Disclosure statement
No potential conflict of interest was reported by the authors.