Limited Potential of Polystyrene Microplastic as a Vector of Microcystin-LR in Diluted Lysate of Microcystis aeruginosa Strain MASH01-A05 in Laboratory Freshwater and Brackish Water Conditions

Abstract

Microplastics (MPs) and microcystins (MCs) frequently occur together in eutrophic environments. However, their interaction in aquatic systems is poorly understood. This study aimed to examine how MP particle size and salinity influence the adsorption behaviour of the cyanotoxin MC-LR onto polystyrene MPs (PS-MPs). Two particle size groups (180–500 µm and 700–1000 µm diameter) were mixed with a microcystin-LR (MC-LR) producing Microcystis aeruginosa lysate in either freshwater (salinity ≤ 0.05 g L⁻¹) or brackish water (salinity 16.00 g L⁻¹) and incubated at 25 °C in an orbital shaker for 48 h. MC-LR bound to PS-MPs was extracted and measured using triple quadrupole LC-MS/MS. The MC-LR adsorption rate exhibited a degree of oscillation throughout time, with peak adsorption observed for the smaller-sized PS-MPs at 1.60% in freshwater after 4 h and 4.60% in brackish water after 6 h. For the larger particle size of PS-MPs, peak adsorption occurred after 4 h, reaching 0.1% in freshwater and 1.3% in brackish water. This study provides evidence that PS-MPs have limited potential as vectors of MC-LR in eutrophic freshwater and brackish environments.

1. Introduction

Plastic is a building block of many daily use commodities such as disposable drinking water bottles, plastic shopping bags and packages, car tyres and cosmetics, with >400 million tonnes produced annually [1] and approximately 11% entering aquatic environments [2]. Plastic debris can persist in the environment for hundreds of years [3,4]. Over time, physical, chemical and biological processes in an aquatic environment can cause plastic debris to degrade into smaller fragments referred to as microplastics (MPs), ranging from 1 μm to 5000 μm [5,6]. MPs are now present in all parts of the environment, from soil to oceans, and pose a significant threat to ecosystem functioning [7,8].

MPs can float on the water surface or be suspended in the water column and may offer sorption sites for a variety of concomitant environmental pollutants, e.g., toxic compounds, trace metals, hydrophobic organic pollutants, pathogens, etc. [9,10]. MPs can also have strong hydrophobicity and a large surface area, facilitating the adsorption of pollutants from surrounding environments [11,12]. For example, MPs can have 2- to 6-fold higher adsorption capacity towards organic pollutants than soil [13,14]. MPs have been frequently detected in freshwater and estuarine systems, including co-occurring alongside algal blooms [15,16]. High inputs of nutrients in these areas promote eutrophication and the excessive growth of harmful algae such as cyanobacteria, which produce toxins [17,18,19,20]. In an artificial lagoon in China, a maximum abundance of up to 5.8 × 10⁷ cells L⁻¹ was reported, and in an urban lake in Chile, a surface scum bloom peaked at 3.1 × 10⁶ cells L⁻¹ [21,22]. In recent years, Mannus Lake, a reservoir in southeast Australia, has undergone extremely dense cyanobacterial blooms (biovolumes > 80 mm³ L⁻¹), threatening recreational water use and downstream water supplies [23].

One of the dominant bloom-forming cyanobacteria is Microcystis aeruginosa, which produces microcystin (MC) [24,25]. MCs have cyclic heptapeptide structures comprising two variable L-amino acid (R1 and R2) functional groups and five amino acids (D-alanine-R1-D-MeAsp-R2-Adda-D-glutamate-Mdha) [26,27]. To date, 310 congeners of MCs have been characterised, and the most common ones are MC-LR, MC-YR and MC-RR (where Y, R and L indicate the tyrosine, arginine and leucine amino acids, respectively) [28]. Among all variants of MCs, MC-LR is known as the most common variant, and is reported to account for 46.0–99.8% of the total concentration of MCs in cyano-HABs [28]. MC concentrations generally remain below 100 µg L⁻¹, but concentrations can be greater than 1000 µg L⁻¹ in eutrophic water [29].

Microplastic contamination and eutrophication are often caused by factors such as proximity to urban and industrial areas [30,31]. Because they co-occur in eutrophic waters, MPs could act as vectors of MCs. Pestana et al. [32] found that about 28 times the amount of MC-LF analogue (142 µg g⁻¹) was found on polystyrene plastic as in the surrounding waters. Some studies have explored the interaction between MPs and pure MCs in freshwater conditions; for example, Moura et al. [33] studied the adsorption kinetics of eight MC analogues on polyethylene terephthalate (PET) and polypropylene (PP) microplastics. The authors reported that the small PP-MPs, due to their high affinity, were able to adsorb a significant proportion of the eight MCs. The proportion of microcystin adsorbed onto these small particles of PP was substantial (83–100%) for some analogues. MC-LW and MC-LF were the only analogues that adsorbed onto the larger-sized PP and PET microparticles [32,33]. Surface electrostatic interactions of microplastics are a dominant mechanism in microplastic-contaminant sorption [34,35]. Microplastics commonly carry a net negative surface charge and are likely to attract positively charged species and repel negatively charged species [36]. The MC-LR molecule has two negatively charged groups and a single positively charged group, which influence surface electrostatic attractions and influence the interactions with MPs. While acknowledging the potential for interactions between MPs and MC-LR, there is still a research gap on the interaction between naturally occurring microcystin in bloom material and polystyrene microplastics (PS-MPs) under varying salinity conditions on different size groups. Therefore, the aim of this study was to investigate the adsorption of MC-LR onto PS-MPs in freshwater and brackish environments to investigate the extent to which polystyrene MPs of different size ranges bind to extracted MC-LR from cultures of M. aeruginosa at environmentally relevant concentrations. We hypothesized that smaller-sized PS-MPs would adsorb more due to a greater surface area of hydrophobic sorption sites. Further, we also hypothesized that salinity would influence competition for the sorption sites between organic compounds, e.g., MC-LR and salt ions, and modify the chemical binding and partitioning of MC-LR onto MPs, resulting in differential adsorption.

2. Results

Initially, MC-LR showed rapid adsorption for all treatments. After reaching a maximum level, the MC-LR adsorption rate (%) declined to arrive at an equilibrium level (Figure 1). MP particle size, salinity change, and contact time showed an interaction effect for the amount of MC-LR adsorbed onto PS-MPs. Also, significant (F = 12.81, p ≤ 0.001) differences were observed in particle size, salinity, and time. Factorial ANOVAs also revealed significantly higher MC-LR adsorption (both % bound and concentration) onto PS-MPs in brackish water than in freshwater, and in the smaller size fraction of PS-MPs.

2.1. Microcystin (MC-LR) Adsorption Behaviour onto Polystyrene Microplastics (PS-MPs)

Maximum adsorption of MC-LR (1.60%) was found after 4 h of incubation in freshwater on the smaller-sized PS-MPs (180–500 µm). It reached approximate equilibrium after 8 h of incubation, with the lowest adsorption being 0.11% (Figure 1a, Table S1). A similar trend in MC-LR adsorption was found in the freshwater treatment, with the larger-sized having a maximum adsorption at 4 h of incubation (1.18%) and the lowest recorded adsorption (0.42%) at 48 h of incubation. It appeared to reach an equilibrium state after 8 h (Figure 1b, Table S1). In the brackish water treatment, the maximum adsorption of MC-LR on 180–500 µm PS-MPs occurred after 6 h (4.60%), and the lowest recorded adsorption rate (1.32%) was after 48 h (Figure 1c, Table S1). The larger-sized (700–1000 µm) MPs in the brackish water exhibited peak adsorption after 24 h of incubation (0.86%), and the lowest adsorption was 0.19% after 2 h (Figure 1d, Table S1).

Factorial analysis of variance (ANOVA) showed that environmental salinity changes and particle sizes significantly interacted with the MC-LR adsorption amount (%) onto PS-MPs (

Table 1. Output of factorial ANOVA, exploring the effect of environmental salinity and MP particle size on MC adsorption rate (%) on MP. Significant difference is denoted in bold where p < 0.05.

Dependent Variable	Factors	df	Sum of Squares (SS)	F Ratio	p
% Adsorption	Environmental salinity change	1	9.08	13.30	0.001
	MPs Particle size	1	12.36	18.10	≤0.001
	Environmental salinity change × MPs Particle size	3	11.34	16.61	≤0.001

; SS = 11.34, F3 = 9.58, p < 0.01). Further, adsorption (%) significantly differed between freshwater and brackish water treatments, with greater adsorption (%) observed in brackish water (Table 1; SS = 9.08, F1 = 13.30, p < 0.01). A significantly lower MC-LR adsorption (%) was observed on the larger-sized MPs compared to the smaller MPs (Table 1; SS = 12.36, F1 = 18.10, p < 0.01). Post hoc analysis showed the MC-LR adsorption rate (%) to be significantly higher on the smaller-sized fraction (180–500 µm) of PS-MPs in brackish water conditions (Table S2, LSM = 2.31).

2.2. Effect of Particle Size on the Adsorption of MC-LR onto Polystyrene Microplastics (PS-MPs)

The amounts of adsorbed MC-LR (ng g⁻¹) on two different size groups of PS-MPs at different sampling times in freshwater (FW) and brackish water (BW) conditions are presented in Supplementary Figure S1a,b. In freshwater conditions, no detectable MC-LR was found in the extract from both examined sizes of PS-MPs immediately after the addition of MC-LR. At later sampling times in freshwater, smaller-sized PS-MPs (180–500 µm) showed significantly higher MC-LR uptake than larger-sized PS-MPs (700–1000 µm) after 2 h (151.00 ng g⁻¹ > 44.00 ng g⁻¹; p < 0.01) and 4 h (160.00 ng g⁻¹ > 118.00 ng g⁻¹; p < 0.01) (Supplementary Figure S1a and Table S1). No statistical difference in LC-MR concentration on PS-MPs was observed after 6 h, 8 h, 24 h, and 48 h of PS-MP and MC-LR interaction, despite the smaller-sized PS-MPs showing a trend of greater adsorption (Figure 2a and

Table 2. Output of factorial ANOVA, exploring the effect of environmental salinity, MPs particle size, and contact time on the adsorbed MC amount bound on MPs. Significant difference is denoted in bold where p < 0.05.

Dependent Variable	Factors	df	Sum of Square (SS)	F Ratio	p
MC-LR (ng/g)	Environmental salinity change	1	68,759.48	50.54	≤0.001
	MPs Particle size	1	110,679.07	81.35	≤0.001
	Contact time	5	82,788.25	12.17	≤0.001
	Environmental salinity change × MPs Particle size	1	86,882.04	63.86	≤0.001
	Environmental salinity change × Contact time	5	101,347.71	14.90	≤0.001
	MPs Particle size × Contact time	5	39,414.88	5.79	≤0.001
	Environmental salinity change × MPs Particle size × Contact time	5	87,198.95	12.81	≤0.001

).

In brackish water conditions, adsorption of MC-LR was significantly higher on smaller PS-MPs (180–500 µm) than on larger particles (700–1000 µm) at 4 h (296.17 ng g⁻¹ > 56.79 ng g⁻¹; p < 0.01), 6 h (460.00 ng g⁻¹ > 64.00 ng g⁻¹; p < 0.01), and 8 h (216.00 ng g⁻¹ > 58.00 ng g⁻¹; p < 0.01) (Supplementary Figure S1b, Table 2 and Supplementary Table S1). Smaller PS-MP particles showed a greater concentration of MC-LR at 2 h, 24 h, and 48 h in the PS-MP extract, despite not being statistically significant (Supplementary Figure S1b and Table 2).

2.3. Effect of Salinity on the Adsorption of MC-LR onto Polystyrene Microplastics (PS-MPs)

The amount of adsorbed MC-LR (ng g⁻¹) on PS-MPs under FW and BW conditions is presented in Figure 2a,b. On the smaller-sized PS-MPs (180–500 µm) at most sampling times, there was a higher level of MC-LR in BW than in FW, e.g., 6 h (460.00 ng g⁻¹ > 12.00 ng g⁻¹; p < 0.01), and 8 h (216.00 ng g⁻¹ > 11.00 ng g⁻¹; p < 0.01) (Figure 2a, Table S3). At 24 and 48 h, the adsorption of MC-LR to smaller particles was greater in BW, but the results were not significantly different. On the larger-sized PS-MPs (700–1000 µm), there was no significant difference in adsorbed MC-LR between FW and BW (p > 0.05) (Figure 2b, Table S3). There was no detectable level of MC-LR observed at 0 h (Figure 3).

3. Discussion

Microplastic contamination and the occurrence of cyanobacterial blooms have been projected to increase globally due to climate change and anthropogenic activities [19,37]. MPs and cyanotoxins have known individual adverse effects on aquatic organisms. However, the co-existence of these two stressors and their interaction needs better understanding. A recent review focused on the ecological implications of MPs as a vector of microcystin (MCs) in the natural environment [35]. The authors noted that the current understanding of the co-transfer of MPs and MCs is insufficient and urged further experimental research to explore the role of MPs in the transport of MCs, and that research efforts should consider environmentally relevant experimental conditions. In this study, we have shown that lysates of M. aeruginosa containing MC-LR had relatively low adsorption onto PS-MPs. Adsorption varied across different size ranges of the PS-MPs, and smaller particle sizes (180–500 µm) generally had greater adsorption capacity in comparison to larger sizes (700–1000 µm) after 48 h exposure. Further, salinity influenced microcystin adsorption to MPs, with MC-LR adsorption to PS-MP being higher under saline water conditions.

We used a very high concentration of PS-MPs (5 g L⁻¹) in this experiment to explore the adsorption behaviour of MC-LR onto different size fractions of PS-MP under freshwater and brackish water conditions. However, the concentration of MC-LR (50 µg L⁻¹) was typical of a bloom and higher than the recreation water guideline of the National Health and Medical Research Council (NHMRC), Australia. The adsorption under these conditions was low (maximum on fine particles in brackish water was approximately 4.5% of available MC-LR). This suggests that the PS-MPs were saturated with MC-LR, with all available binding sites taken up. Furthermore, there are few reports of the concentration of MPs in river and estuarine waters as mg L⁻¹ for comparison with this study, but Zhang et al. [38] estimated that the concentration of MPs in estuarine and coastal waters was usually between 0 and 10 mg L⁻¹. PS-MPs at this concentration would take up negligible amounts of MC-LR from a bloom concentration of 50 µg L⁻¹ MC-LR, although there might be some effect at a lower background MC concentration of ~50 µg L⁻¹. In some highly polluted areas, however, Zhang et al. [38] estimated that MPs could exceed 1 g L⁻¹, and the Yellow River estuary could reach up to 8 g L⁻¹. Such high concentrations of MPs could be enough to remove significant amounts of MC-LR from the water, even if the rate of binding is low.

In this experiment, we utilised lysates of M. aeruginosa as a source of MC-LR. Lysates contain different types of organic byproducts, e.g., broken algal cells, extracellular polysaccharides, and intracellular organic matter. These organic compounds are usually found in eutrophic environments alongside microcystin during algal bloom formation and breakdown. They may modify MC-LR adsorption kinetics by competing for sorption sites, fouling, and altering the plastic surface chemistry [39,40]. Use of MC-containing cell lysates may be more realistic than using pure MCs, because it models the organic matter (including both microcystin and non-microcystin) in lake water, where possible dilution occurs, and that can compete with MCs for sorption sites on the surface of MPs.

The adsorption rate of MC-LR to PS-MS was found to be time-dependent. MC-LR adsorption reached the peak level within 4–6 h of incubation and then reduced to an equilibrium, indicating that desorption occurred. These findings agree with Hataley et al. [41], who reported that three different congeners of MCs, including MC-LR, exhibited maximum adsorption within 1 h of the interaction of MCs and polyethylene (PE), polyvinyl chloride (PVC), and polypropylene (PP) MPs and eventually showed a general trend of decline. The variation in adsorption rate (%) before reaching an equilibrium state might be due to the dryness of PS-MPs, which initially offer large numbers of hydrophobic sorption sites, but their affinity diminishes over interaction time in water due to increased competition with other organic compounds [42]. Another possible cause of desorption is bacterial degradation in a nutrient-rich environment. Despite the change, we believe bacterial degradation was low as we took precautions to control bacterial contamination, e.g., autoclaved glassware, working under laminar airflow. Further, organic matter (extracellular polysaccharides and intracellular organic matter) present in the lysate could be attached to the MP surface later during the incubation, which may lead to a reduction in the adsorption capacity of PS-MPs [39,42].

The maximum adsorption (%) of MC-LR onto PS-MPs ranged from 1.18% in the larger 700–1000 µm PS-MPs to 4.6% in the smaller 180–500 µm particles. In both the FW and BW environments, smaller-sized 180–500 µm PS-MPs showed greater adsorption of MC-LR. Moura et al. [32] and Pestana et al. [33] also demonstrated that the amount of MC-LR and MC-LF binding (or adsorbing) onto MPs decreases with increased particle size. Other plastic types, such as PVC, polyamide-6, PET, and PP, have been shown to adsorb 10%, 89%, 19%, and 88% of MC, respectively [16,43]. This suggests that PS-MPs may have limited potential as a vector for the transmission of MC-LR compared to other plastics. However, the surface property and zeta potential were not measured in this study. We recognised that adsorption of an MC-LR to plastics could be affected by the plastics’ glassiness, crystallinity, and polarity [44]. The rubber state is more amorphous than the glassy state, so rubber-state plastic can absorb more organic compounds, as intermolecular bonds are more flexible in the rubber state than in the glass state [45]. Polystyrene microplastic has high crystallinity compared to other plastic types reported; therefore, it is likely to have lower adsorption of MC-LR [46].

MC-LR adsorption was significantly higher in brackish water compared to freshwater for smaller-sized MPs in this study. This is consistent with a recent study, in which authors reported higher MC-LR concentration on MPs in coastal water than in freshwater [47]. They suggested this is influenced by the effects of pH and monovalent salt ions on the adsorption and desorption of MCs onto the PP and PE MPs. The adsorption of MC-LR onto PS-MPs observed in this study can be explained by the mechanisms involved in the adsorption of organic compounds on the hydrophobic MP surface. These mechanisms include complexation, van der Waals forces, hydrogen bonds, pore filling, partitioning, electrostatic interactions, π-π conjugation, and hydrophobic interaction [39,44,45,46,47,48]. Wan et al. [47] observed the surface of virgin and aged PE and PP-MPs and proposed that pore filling was involved in the adsorption process of MC-LR. A greater hydrophobic surface might play a significant role in the adsorption of MC-LR onto smaller PS-MPs. We used neutral pH water for the adsorption experiment. However, during incubation with microcystin, the pH increased slightly (7.1–7.2), which is still slightly basic, so the electrostatic conjugation of MC-LR and PS-MPs is expected to be weak. One possible reason for higher adsorption in brackish water conditions is the cation binding effect between MC-LR and PS-MPs, where salt ions may conjugate with MC-LR to form a complex compound, which favours the adsorption on PS-MPs [49,50]. Another explanation for greater MC-LR adsorption on PS-MPs in brackish water is that the mixture contains a higher concentration of ionic compounds that may facilitate polymer bridging during adsorption, thereby increasing MC-LR adsorption. Brewer et al. [51] reported that the adsorption of organic materials onto microplastics in seawater showed a higher dependence on Kow than in freshwater. Again, this indicates that hydrophobicity might play a more significant role in solutions with higher salt content. It has also been reported that the salting-out effect can occur, where increased salt concentration reduces a compound’s solubility and enhances hydrophobic interactions between organic content and microplastics [11]. Given that it is an important attempt to ascertain MC-LR adsorption behaviour on PS-MPs, though data and interpretation should be viewed cautiously, and caveats kept in mind, this experiment was done using single doses of MC-LR and only polystyrene microplastic. Thus, to disentangle the complex impact of water chemistry on the microcystin adsorption on microplastic polymers, further investigation with different microcystin congeners with multiple concentrations and their interaction with multiple microplastic polymers is recommended, and this was beyond the scope of this study.

4. Conclusions

This experiment investigated the adsorption behaviour of MC-LR from M. aeruginosa cell lysates onto PS-MPs with different particle sizes and salinity conditions. Maximum adsorption of MC-LR occurred after approximately 4 h of contact with PS-MPs in freshwater 1.6%, 160 ng g⁻¹) and after 6 h in brackish water (4.6%, 460 ng g⁻¹). Smaller-sized PS-MPs had higher adsorption, and this was greater under brackish water conditions. However, adsorption is low and transient, desorption is substantial, and therefore vector potential is limited under the tested conditions. But this study also suggests that PS-MPs could distribute MC-LR from eutrophic environments with cyanobacterial blooms and pose ecotoxicological bioaccumulation risks in eutrophic environments. Thus, the assessment of the ecotoxicological risk posed by both PS-MPs and microcystins should be considered when they coexist in eutrophic ecosystems.

5. Materials and Methods

5.1. Experimental Design

This experiment was conducted in 100 mL glass containers in a temperature-controlled (25 °C) laboratory room. The experimental factors were water salinity (2 treatments: freshwater (FW) (salinity ≤ 0.05 g L⁻¹) and brackish water (BW) (salinity 16 g L⁻¹)) and MP size fraction (2 treatments: 180–500 µm and 700–1000 µm). These were chosen as they represent salinity conditions where eutrophication and microplastic contamination occur [25]. Artificial freshwater (AFW) and artificial brackish water (ABW) were prepared in the laboratory by adding CaCl₂·2H₂O (58.5 mg L⁻¹), MgSO₄·7H₂O (24.7 mg L⁻¹), NaHCO₃ (12.0 mg L⁻¹), and KCl (1.2 mg L⁻¹) in ultrapure water (18.2 MΩ). In the case of brackish water preparation, NaCl was added to the solution until it reached a salinity of 16 g L⁻¹. The water pH was adjusted to 7.0 at room temperature (25 °C). The two MP size fractions were chosen to reflect the essential size classes of meso-zooplankton that juvenile and small-bodied fish feed on, so the particles may be ingested with their natural food sources [43].

A toxic strain of M. aeruginosa MASH01-AO5 (Australian National Algae Culture Collection, Hobart, Tasmania, Australia) was incubated in MLA medium under controlled environmental conditions [44]. The initial cell density was 10⁴ cells mL⁻¹, and the incubation conditions were at 22 °C under 20–25 µmol m⁻² s⁻¹ illumination with a 14–10 h light-to-dark cycle throughout the culture period. During the exponential growth period, 40 mL of Microcystis culture media was transferred into a Falcon tube and centrifuged at 3000× g at 4 °C for 15 min. The supernatant was removed, and the pellet was sonicated for 1 min for cell breakdown and diluted with 3 mL ultrapure water (Sartorius H₂O Pro-VF-B). This diluted cell lysate was centrifuged again, pellets were discarded, and supernatants containing MCs lysate were preserved at −80 °C. Microcystin congener (MC-LR, MC-LF, MC-RR) concentrations in the lysate were measured by LC-MS/MS (see details in Section 5.2); however, only the MC-LR congener was detected in the lysate. As a measure of quality control and to prevent contamination, all the jars, media and utensils used for algae culture and lysate isolation were sterilised by autoclaving and a laminar air flow cabinet was used for performing the experiment.

We sourced polystyrene packaging, i.e., container lids, to create the MPs for the study. This plastic was chosen as it is one of the most common polymers contaminating the environment. Polystyrene was broken into small pieces in the laboratory, washed with 75% methanol, and dried. A stainless-steel grinder was used to grind the plastic particles into smaller-sized materials. Then, plastics were soaked in liquid nitrogen and ground again. The grinder was operated for 20 s with 1–2 min cooling intervals to avoid any thermal alteration of the plastics. The microplastic particles were size fractionated using metal sieves to create two fractions: 180 to 500 µm and 700 to 1000 µm. The properties of the polymer material were determined by FT-IR spectrometry (Nicolet 6700, Thermo Electron Corporation, Melbourne, VIC, Australia) which showed that the spectra of the particles analysed matched the corresponding spectra for polystyrene (Supplementary Figure S2). Glass containers were filled with 50 mL artificial water, and PS-MPs were added at a concentration of 5 g L⁻¹ to ease the adsorption analysis and to attain enough plastic samples in each sampling time. However, particle number was different between the two microplastic size groups, with 2 × 10⁵ particles L⁻¹ for 180–500 µm and 1 × 10⁴ particles L⁻¹ for 700–1000 µm. The microcystin lysate collected above was added to the mixture at a concentration of 50 µg L⁻¹ MC-LR, which was a five-fold higher concentration than the recreational water guideline by WHO [52] of 10 µg L⁻¹. This experimental unit was replicated three times in separate glass beakers for each PS-MP size group and water salinity treatment. All replicated experimental units were placed on an orbital shaker at 180 rpm at room temperature (25 °C) for 48 h.

5.2. Sampling Procedures and Analysis

MPs sampling was carried out at 2, 4, 6, 8, 24, and 48 h. At each time of sampling, 5 mL of solution from each replicate of each treatment was pipetted out and filtered through a glass microfibre filter (0.45 µm). Filtration was carried out in a fume hood using a vacuum pump by maintaining a minimum suction pressure to drain the liquid. Filter papers containing MPs were dried at room temperature (25 °C) and washed with 75% methanol three times to extract adsorbed MC-LR. As procedural controls, filter paper without microplastics was also washed similarly with 75% methanol to test if the filter paper adsorbed any MC during filtration. Filter paper did not adsorb a detectable amount of MC-LR. The extract was evaporated to near dryness (<0.01 mL) by using a nitrogen evaporator (Ratek, Boronia, Australia) at 40 °C. The residue was resuspended in 40 µL LC-MS grade methanol and ultrapure water (75:25, v/v) and transferred to an LC insert and 2 mL amber vial (Agilent, Mulgrave, VIC, Australia) for liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.

The MC-LR concentrations were quantified using the method described by Pravadali-Cekic et al. [27] with some modifications. The samples were analysed using Shimadzu Nexera UC UHPLC coupled to a Shimadzu LCMS-8060 triple quadrupole mass spectrometer (Shimadzu, Rydalmere, NSW, Australia). Chromatographic separation was carried out in an Agilent reverse-phase Zorbax C18 column (2.1 × 100 mm, 1.8 µm RRHD Eclipse Plus, Agilent, Mulgrave, VIC, Australia) at 18 °C and a flow rate of 0.65 mL min⁻¹. The LC gradient started at 70% A (ultrapure water + 0.1% v/v formic acid (Sigma-Aldrich, Castle Hill, NSW, Australia)) and 30% B (acetonitrile (Sigma-Aldrich, Castle Hill, NSW, Australia) + 0.1% v/v formic acid), increasing to 32.5% B over 4 min (Figure 3). The mobile phase was held at 32.5% B until 4.5 min, when it was increased to 90% B and held for 1.5 min to wash the column. The system then returned to the initial condition of 30% B for an additional 2 min to equilibrate before the next injection. Each sample injection volume was 5 µL, and samples were run in duplicate.

The MS/MS was run in positive mode, with multiple reaction monitoring (MRM) parameters as detailed in

Table 3. MRM parameters for MCs. ‘*’ denotes quantifier ions.

	Retention Time (min)	Precursor Ion (m/z)	Product Ion (m/z)	Collision Energy (eV)
MC-RR	1.18	519.9	135.0 *	−29
			103.0	−55
			127.0	−43
MC-YR	3.21	1045.6 523.6	135.0 *	−55
			126.8	−51
			212.9	−53
			135.0	−14
MC-LR	3.54	498.7 995.6	135.0 *	−13
			112.0	−54
			135.0	−54
			213.2	−54

. The electrospray ionisation (ESI) interface had a temperature of 300 °C. The Nebuliser gas flow was set to 2.00 L/min, with a heating gas flow of 10.00 L/min, and a drying gas flow of 10.00 L/min. The heat block temperature was set to 400 °C, and the desolvation line was 250 °C. All other parameters were set as per the most recent instrument tuning file. Data acquisition and analysis were performed using Shimadzu’s Lab Solutions.

The concentration of MC-LR was determined by comparing against an 8-point calibration curve (microcystin RR-YR-LR) with standards ranging from 0.1 to 500 ng mL⁻¹. The linearity of the calibration curve was >0.999 for all analytes. The accuracy of the analysis was confirmed by using a series of spiked samples, and the recovery rate of spiked MC-LR was within the acceptable range (109.3%). The limit of detection (LOD) and limit of quantification (LOQ) were 0.03 ng mL⁻¹ and 0.1 ng mL⁻¹, respectively.

5.3. Data and Statistical Analysis

The amount of MC-LR adsorbed (Q_t ng g⁻¹) at any sampling time (t) on the microplastic was calculated by the equation ${\mathrm{Q}}_{\mathrm{t}}={\mathrm{K}}_{\mathrm{t}}\times {\displaystyle \frac{\mathrm{V}}{\mathrm{m}}}$; where K_t is the MC-LR concentration of the extracted solution (ng mL⁻¹) from sampled MPs, V is the total volume (mL) of extracted solution, and m is the mass of MPs sampled at time t.

The percent (%) adsorption of MC-LR on the MPs was calculated by using the equation:

$$\% \mathrm{o}\mathrm{f} \mathrm{M}\mathrm{C}-\mathrm{L}\mathrm{R} \mathrm{a}\mathrm{d}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{p}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=\left({\displaystyle \frac{{\mathrm{C}}_{\mathrm{t}}\times 100}{{\mathrm{C}}_{\mathrm{i}}}}\right)$$

where C_i (ng) is the initial added amount of MC-LR in the test solution, and C_t (ng) is the adsorbed amount of MC-LR.

Data and statistical tests were visualised in SigmaPlot version 15 and IBM SPSS Statistics version 28. All data were log₁₀ (X + 1) transformed before the statistical test, and the normality of data within the treatment group was tested by the Shapiro–Wilk test of normality. Factorial ANOVA using Generalised Linear Models (GLMs) examined the main effect and the interaction effect of independent factors, i.e., MP particle size, environmental salinity conditions, and incubation time on the dependent variable, adsorbed MC-LR concentration onto MPs. A post hoc (LSmeans Tukey HSD) test of the interaction effect was performed if there was an observed significant interaction.