Microcystins (MCs) are hepatotoxic cyanotoxins produced by cyanobacteria belonging to the genera Microcystis, Anabaena, Aphanizomenon, Planktothrix, Anabaenopsis, Hapalosiphon, and Nostoc (Janse et al., 2005; Zurawell et al., 2005). MCs have a cyclic heptapeptide structure comprising seven amino acids, with more than 250 known variants depending on the kinds of amino acids constituting the MCs (Bouaïcha et al., 2019). The main detected congeners of freshwater MCs are MC-LR, -YR, and -RR (Kim et al., 2010; Mowe et al., 2015; Beversdorf et al., 2017). Hydrophobic MCs such as MC-LA, -LW, and -LF are sometimes found in environmental samples, but rarely in higher concentrations (Graham et al., 2010; Faassen and Lürling, 2013; Beversdorf et al., 2017).

The toxicity of MCs is determined by the degree of their hydrophobicity by the type of substituted amino acid present (Vesterkvist et al., 2012; Santori et al., 2020). Depending on the hydrophobicity, the degree of transport into the cell and that of detoxification within the cell vary (Fastner and Humpage, 2021). The magnitude of hydrophobicity of the well-known MCs were in the order MC-LF > -LW > -LR > -YR > -RR (Santori et al., 2020). In an in vitro detoxification experiment using Caco-2 cells, detoxification of LF with high hydrophobicity did not occur well. Since MC-RR is relatively hydrophilic, organic acid transporter polypeptides (OATP)-mediated hepatic uptake is limited and shows the highest detoxification efficiency (Vesterkvist et al., 2012). The LD₅₀ (i.p., mouse) for MC-LR is 50 μg/kg body weight (bw), whereas those of MC-YR and MC-RR are reported to be 70 and 600 μg/kg bw, respectively (Table 1; Hitzfeld et al., 2000).

Table 1 Comparison of enzyme inhibition and toxicity of microcystin congeners between this study and previous studies

Parameter	Enzymed	Substratee	MC-LRf	MC-RR	MC-YR	MC-LA	References
IC50a (µg/L)	PP1	MUP	0.80	5.39 (0.149)h	3.12 (0.257)	2.05 (0.392)	This study
	PP1	phosvitin	0.8	0.9 (0.889)	1.9 (0.421)	-	Heresztyn and Nicholson (2001)
	PP1	DiFMUP	0.25g	0.71g (0.353)	1.04g (0.238)	-	Blom and Jüttner (2005)
	PP2A	DiFMUP	0.05g	0.10g (0.480)	0.27g (0.183)	-	Blom and Jüttner (2005)
	PP2A	pNPP	0.22	0.25 (0.880)	0.27 (0.815)	-	Rivasseau et al. (1999)
	PP2A	pNPP	0.46g	0.62g (0.742)	0.88g (0.523)	0.51g (0.902)	Robillot and Hennion (2004)
	mixed PPase	32P-histone H1	1.59g	3.53g (0.452)	1.46g (1.089)	-	Yoshizawa et al. (1990)
EC50b (µg/ml)	-	-	0.2	4.5 (0.044)	0.35 (0.571)	-	Fastner et al. (1995)
LD50c (µg/kg)	-	-	50	600 (0.083)	70 (0.714)	50 (1.000)	Zurawell et al. (2005)

^a Half inhibition concentration, estimated with protein phosphatase inhibition assay.

^b Half effective concentration, indicated cytotoxicity determined using rat hepatocytes.

^c Half lethal concentration, intra-peritoneal injection to mouse.

^d PPase, protein phosphatase; PP1, PPase type 1; PP2A, PPase type 2A.

^e pNPP, p-nitrophenyl phosphate; MUP, 4-methylumbelliferyl phosphate; DiFMUP, 6,8-difluoro-4-methylumbelliferyl phosphate.

^f MC, microcystin.

^g Converted from molar concentration.

^h Numbers in parenthesis indicated the TEF values to MC-LR.

The predominantly used methods for quantitative and qualitative analysis of MCs are liquid chromatography (LC), enzyme-linked immunosorbent assay (ELISA), and protein phosphatase inhibition assay (PPIA) (Msagati et al., 2006). MCs covalently bind with ser/thr protein phosphatase (PPase) type 1 (PP1) or type 2A (PP2A) to inhibit the dephosphorylation of these enzymes (MacKintosh et al., 1990; Massey and Yang, 2020). PPIA uses this mechanism to quantify the concentration of MCs (Almeida et al., 2006; Sassolas et al., 2011). This method is widely used to quantify the concentration of soluble and particulate MCs in water samples because it has a lower limit of quantitation (LOQ) compared to the LC method, does not require special equipment and analysis technology, and has a low analysis cost (McElhiney and Lawton, 2005). However, it is difficult to determine the concentration of each MC congener using PPIA, similar to the ELISA method (Fischer et al., 2001). Moreover, since PPase activity can be inhibited by substances such as okadaic acid, calyculin A, and tautomycin, the concentration of MCs may be overestimated due to the presence of such substances in the environmental samples (Albay et al., 2003; Lindholm et al., 2003; Swingle et al., 2007).

In 1996, more than 50 patients died due to the use of water contaminated with MCs at a hemodialysis center in Caruaru, Brazil (Jochimsen et al., 1998). In response to this incident, the World Health Organization (WHO) proposed 1 μg MC-LR/L as a provisional guideline for MCs in drinking water (Kuiper-Goodman et al., 1999). This guideline was recently updated to 1 and 12 μg/L for lifetime and short-term criteria, respectively (Fastner and Humpage, 2021). The new guidelines were established by reflecting the results of the toxicological data for MC-LR (Fawell et al., 1999; Heinze, 1999). However, MCs exist in the form of a mixture in the natural environment, and toxicological data for congeners other than MC-LR are insufficient. Therefore, the congeners of MCs were considered to have similar toxicity to that of MC-LR, and the concentration of total MCs was the reference used in the revised guideline (Fastner and Humpage, 2021).

Most countries have established guidelines for MCs in drinking water based on the reference substances such as MC-LR or total MCs. Brazil, Uruguay, the Czech Republic, France, and Spain have set guidelines based on MC-LR, while Australia, France, and Finland have set guidelines based on the total MCs (sum of all variants) (Chorus, 2012; Ibelings et al., 2015). In the United States, after the tap water of Toredo City (Ohio State) was reported as contaminated with MCs in August 2014, the US Environmental Protection Agency (US EPA) established a 10-day health adversary guideline value against total MCs, which was 0.3 μg/L for bottle-fed infants and young children of preschool age and 1.6 μg/L for school-age children and adults (Henrie et al., 2017).

In countries that use total MCs as the reference substances in setting guidelines, such as the United States, the differences in toxicity according to MC congeners are not considered, and it is assumed that the toxicity of congeners other than MC-LR is the same as that of MC-LR (Henrie et al., 2017). The toxicity of MCs differ largely among congeners, therefore, it is possible that the actual risk may be overestimated when the control guideline is set based on the sum of the concentrations of MC congeners.

Toxic equivalent quantity (TEQ) is sometimes used to establish guidelines for substances with several variants. TEQ is the emission control standard for dioxins, which is obtained by adding the product of its concentration and the toxic equivalent factor (TEF). TEF is the relative toxicity of each variant to that of the reference material of each congener, estimated based on the most toxic dioxin, 2,3,7,8-tetrachlorodibenzo-p-dioxin (Manning et al., 2013). The TEFs for each congener of polynuclear aromatic hydrocarbons (Zhong and Zhu, 2013) and polychlorophenols (Xing et al., 2012) have been also determined.

The LC and ELISA methods do not consider the degree of toxicity of MC congeners, and hence, provide the same measurement results for the same concentrations. However, in PPIA, there is a difference in the degree of enzyme inhibition depending on the MC congener, even at the same concentration (Ward et al., 1997; Blom and Jüttner, 2005; Mountfort et al., 2005; Covaci et al., 2012). Moreover, PPIA facilitates the measurement of the relative toxicity of each congener of MCs, and based on this, the TEF can be estimated. This would enable the establishment of a reasonable control guideline that reflects the actual toxicity of MC congeners, as in the case of dioxins.

In this study, to obtain TEFs of MC congeners, the enzyme inhibition was analyzed using the PPIA method at various concentrations for each congener following the protocol of a previous study (Rapala et al., 2002; Oh et al., 2022). The total volume of the reaction solution was 200 µl comprising 10 µl of enzyme solution (PP1, 0.005 U/µl), 30 µl of reaction buffer (6×), 20 µl of substrate solution, and 140 µl of sample solution. The reaction buffer solution (6×) comprised Tris-HCl buffer (0.3 M, pH 7.4), MgCl₂ (120 mM), MnCl₂ (1.2 mM), and bovine serum albumin (3 mg/ml). The 4-methylumbelliferyl phosphate (MUP) was used as substrate because it shows the lower LOQ than other substrates such as p-nitrophenyl phosphate (pNPP) (Bouaïcha et al., 2002; Oh et al., 2022). The substrate solution was prepared by dissolving MUP in a reaction buffer (1×) at a concentration of 400 µg/ml. The sample solution consisted of the authentic standards of MC-LR, -YR, -RR, and -LA (Enzo Life Sciences, Inc.) in 15% (v/v) methanol. The range of MC concentrations were as follows: MC-LR at 0.013–6.25 µg/L, MC-LA at 0.125–12.5 µg/L, and MC-RR and MC-YR at 0.125–25 µg/L.

The enzyme solution, reaction buffer solution, and sample were mixed and heated in a water bath at 37°C for 3 min to inhibit the enzyme activity. The substrate solution was then added, and the reaction was carried out at 37°C for 90 min. Finally, a reaction terminator, 1,800 µl of glycine buffer (0.01 M, pH 10.5), was added. The fluorescence of 4-methylumbelliferone (MUF), the hydrolysis product of MUP, was measured at the detection wavelength (excitation wavelength, 365 nm; emission wavelength, 445 nm) using a fluorometer (GloMax^® Multi Jr; Promega).

After measuring the enzyme inhibition according to the concentration of MCs for the four congeners (MC-LR, MC-RR, MC-YR, and MC-LA), which are widely found in freshwater (Kim et al., 2010; Oh et al., 2010; Faassen and Lürling, 2013; Mowe et al., 2015; Beversdorf et al., 2017; Lee et al., 2018; Fastner and Humpage, 2021), the TEF for each congener was calculated from the concentration of congeners at which the enzyme activity was inhibited by 50% (half inhibition concentration, IC₅₀). At the same concentration, the enzyme inhibition was in the order of MC-LR > MC-LA > MC-YR > MC-RR, which showed the same trend as that of the toxicity (half lethal dose, LD₅₀) of MC congeners reported by previous studies (Table 1). In a study by Covaci et al. (2012), the toxicity of MC varied in the order of MC-LR > MC-YR > MC-RR, as estimated by PPIA.

The log value of the concentration of MCs (ln[MCs]) and the degree of enzyme inhibition showed a linear positive correlation (Fig. 1); the regression equation for each congener is as follows:

Fig. 1. Extent of enzyme inhibition according to the concentration of four microcystin (MC) congeners in the protein phosphatase inhibition assay using protein phosphatase type 1 and 4-methylumbelliferyl phosphate as enzyme and substrate, respectively.

The IC₅₀ values for MC-LR, MC-RR, MC-YR, and MC-LA, calculated according to the regression equations presented above, were 0.80, 5.39, 3.12, and 2.05 µg/L, respectively. In a previous study, the IC₅₀ for MC-LR was 0.05–2.19 µg/L, which was different depending on the enzyme or substrate (Table 1). The IC₅₀ for MC-RR ranged 0.1–3.53 µg/L, and the IC₅₀ of MC-LR with respect to MC-RR (IC₅₀ [MC-LR]/IC₅₀ [MC-RR]) was found to be 0.353–0.889, and the mean was 0.642 (± 0.227) (Table 1). In this study, it was estimated to be 0.149. For MC-YR, the average IC₅₀ (MC-LR)/IC₅₀ (MC-YR) was 0.550 (± 0.344) in previous studies and 0.257 in this study. In most studies, the IC₅₀ (MC-LR)/IC₅₀ (MC-RR) was higher than the IC₅₀ (MC-LR)/IC₅₀ (MC-YR) (Table 1). These results are different from the results of acute toxicity (LD₅₀) to MCs measured using mice and those of cytotoxicity (EC₅₀) using hepatocytes of mice, where the LD₅₀ and EC₅₀ for MC-RR were higher than those of MC-YR, the toxicity of MC-RR was considered to be lower than that of MC-YR. In a study by Yoshizawa et al. (1990), the latter value was larger than the former one, similar to the results obtained in this study, which is consistent with the toxicity values of MC-RR and MC-YR. From the results obtained in this study, the IC₅₀ (MC-LR)/IC₅₀ (MC-RR) and IC₅₀ (MC-LR)/IC₅₀ (MC-YR) were calculated as 0.149 and 0.257, and these values were considered as the TEFs for MC-RR and MC-YR against MC-LR, respectively.

The concentration of MC-LA required to inhibit PP1 activity was measured to be lower than that of MC-YR and MC-RR and higher than that of MC-LR (Table 1). The IC₅₀ (MC-LR) to IC₅₀ (MC-RR) and the TEF of MC-LA were 0.392. Based on these results, the formula for converting the concentration of the four MC congeners measured from environmental samples into TEQ for MC-LR (MC-LR_TEQ) is presented as follows:

where C_n is the concentration of MC congeners, and TEF_n is the TEF of the corresponding congeners to MC-LR. Rewriting Eq. 5, using the TEF calculated in this study:

Gupta et al. (2003) reported the TEFs for MC-RR and MC-YR to MC-LR as 0.2 and 0.4, respectively, which are widely used to obtain the TEQ of a mixture containing these congeners (Lei et al., 2008; Li et al., 2008). These values are almost similar to those obtained in this study.

Considering the TEF and concentration, a mixed solution of MC-LR, MC-RR, MC-YR, and MC-LA was prepared such that the TEQ of MCs was approximately 1. In these solutions, the sum of the congener concentrations was 1–6.67 µg/L. The degree of enzyme inhibition was 51.4 (± 1.34)%, which was almost similar to the enzyme inhibition of 1 µg/L MC-LR, 52.3% (Table 2). Therefore, the TEFs for MCs suggested in this study are suitable to calculate the TEQ from environmental samples containing various MC congeners.

Table 2 Measurement of enzyme inhibition in a mixture adjusted for toxic equivalent quantity based on the toxic equivalent factor for four microcystin congeners

No	Concentration (µg/L)					TEQa	Inhibition (%)
	MC-LR	MC-RR	MC-YR	MC-LA	SUM
1	0.00	3.23	2.08	0.00	5.31	1.02	52.5
2	0.00	2.08	1.98	0.52	4.58	1.02	51.1
3	0.00	4.06	0.00	1.04	5.10	1.01	50.0
4	0.00	0.00	2.29	1.15	3.44	1.04	50.7
5	0.21	4.48	0.52	0.00	5.21	1.01	54.1
6	0.21	4.27	0.00	0.42	4.90	1.01	50.8
7	0.21	3.02	1.04	0.21	4.48	1.01	50.8
8	0.42	3.91	0.00	0.00	4.32	1.00	53.7
9	0.42	2.50	0.00	0.52	3.44	0.99	53.7
10	0.42	2.50	0.83	0.00	3.75	1.00	50.6
11	0.42	2.19	0.42	0.42	3.44	1.01	49.7
12	0.63	2.50	0.00	0.00	3.13	1.00	49.7
13	0.63	0.94	0.63	0.21	2.40	1.01	52.0
14	0.42	0.00	1.56	0.52	2.50	1.02	50.3
15	0.42	0.00	2.40	0.00	2.81	1.03	51.0
16	0.84	0.00	0.00	0.41	1.25	1.00	51.8
17	1.00	0.00	0.00	0.00	1.00	1.00	52.3
18	0.00	6.67	0.00	0.00	6.67	0.99	51.0
19	0.00	0.00	4.06	0.00	4.06	1.04	50.7

^a TEQ (toxic equivalent quantity) = [MC-LR] + 0.149 × [MC-RR] + 0.257 × [MC-YR] + 0.392 × [MC-LA]

Saxitoxins, also known as paralytic shellfish poisons, are produced by marine microalgae, dinoflagellates of the genera Alexandrium, Gymnodinium, and Pyrodinium, as well as freshwater cyanobacteria such as Anabaena, Aphanizomenon, and Cylindrospermopsis (Testai, 2021). Saxitoxins have 57 analogues, and their toxicity differs by approximately 160 times depending on their structure (Wiese et al., 2010; Testai, 2021). The European Food Safety Authority developed TEFs using the results obtained from mice (EFSA, 2009; Munday et al., 2013). Based on this, it was possible to obtain the equivalent for saxitoxin (STX-equivalents; STX-eq) by multiplying each toxin concentration by the TEF of corresponding saxitoxin analogue. The Oregon Health Authority of Oregon, USA set a health-based guideline for saxitoxins as 1 and 10 μg STX-eq/L for drinking and recreational water, respectively (Farrer et al., 2015).

In South Korea, MCs have been included in the Drinking Water Quality Surveillance List since 2013, and the provisional guideline for purified water is 1 μg/L for MC-LR (Yoon et al., 2016). According to previous studies investigating the production of MCs in isolates from Korean freshwaters, the proportion of MC-RR was significantly higher, and that of MC-RR and MC-LR was 89.3% and 10.6%, respectively (Kim et al., 2010). In the environmental samples, the concentrations of MC-RR were also the highest among MC congeners in most freshwaters in Korea (Kim et al., 1999; Oh et al., 2012; Lee et al., 2018). Therefore, it is necessary to monitor other MCs in addition to MC-LR in order to manage the risk of MCs in drinking water in Korea. In addition, since it is not appropriate to use MC-LR as a reference material to manage the risk of MCs, it is necessary to consider using the TEQ based on the TEF of each congener presented in this study.

음용수 또는 친수용수에서 남조류 독소인 microcystins (MCs)에 대한 관리기준은 MC 동족체의 독성 차이를 반영하지 않는다. 본 연구에서는 효소저해법을 사용하여 담수 환경에서 널리 발견되는 4종의 MC 동족체에 대한 독성등가지수(toxic equivalent factor, TEF)를 산출하였다. MC-LR에 대한 MC-RR, MC-YR, MC-LA의 TEF는 각각 0.149, 0.257, 0.392이었다. 이러한 결과를 바탕으로 환경 시료에서 측정한 4종의 MC 동족체 농도를 MC-LR에 대한 독성등가치(toxic equivalent quantity, TEQ_MC-LR)로 환산하는 식을 제시하였다. 본 연구에서 제시한 4종의 MC 동족체에 대해 TEF를 적용하여 TEQ_MC-LR을 기반으로 관리기준을 수립한다면 실질적이고 효율적인 MCs의 위해 관리가 가능할 것으로 판단된다.

This research was supported by Chungbuk National University Korea National University Development Project (2021). We would like to thank Editage (www.editage.co.kr) for English language editing.

The authors have no conflict of interest to report.