Nonalcoholic fatty liver disease alters microcystin-LR toxicokinetics and acute toxicity

Abstract

Microcystin-LR (MCLR) is a cyanotoxin produced by blue-green algae that causes liver and kidney toxicities. MCLR toxicity is dependent on cellular uptake through the organic anion transporting polypeptide (OATP) transporters. Nonalcoholic fatty liver disease (NAFLD) progresses through multiple stages, alters expression of hepatic OATPs, and is associated with chronic kidney disease. The purpose of this study was to determine whether NAFLD increases systemic exposure to MCLR and influences acute liver and kidney toxicities. Rats were fed a control diet or two dietary models of NAFLD; methionine and choline deficient (MCD) or high fat/high cholesterol (HFHC). Two studies were performed in these groups: 1) a single dose intravenous toxicokinetic study (20 μg/kg), and 2) a single dose intraperitoneal toxicity study (60 μg/kg). Compared to control rats, plasma MCLR area under the concentration-time curve (AUC) in MCD rats doubled, whereas biliary clearance (Cl_bil) was unchanged; in contrast, plasma AUC in HFHC rats was unchanged, whereas Cl_bil approximately doubled. Less MCLR bound to PP2A was observed in the liver of MCD rats. This shift in exposure decreased the severity of liver pathology only in the MCD rats after a single toxic dose of MCLR (60 μg/kg). In contrast, the single toxic dose of MCLR increased hepatic inflammation, plasma cholesterol, proteinuria, and urinary KIM1 in HFHC rats more than MCLR exposed control rats. In conclusion, rodent models of NAFLD alter MCLR toxicokinetics and acute toxicity and may have implications for liver and kidney pathologies in NAFLD patients.

1. Introduction

Nonalcoholic fatty liver disease (NAFLD) is the most common liver disease in the western world and encompasses a spectrum of conditions, including simple fatty liver (steatosis), nonalcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma (1,2). NAFLD is defined as the presence of hepatic steatosis in ≥5% of hepatocytes that cannot be explained by other liver disease etiologies such as extensive alcohol consumption, viral hepatitis, or drug-induced liver injury (1,2). One clinical implication of NAFLD is the effect the disease can have on xenobiotic metabolism and disposition (3,4). NAFLD is highly associated with obesity, insulin resistance, and metabolic syndrome and its incidence has steadily increased in recent years (1,2,5). NAFLD is also associated with increased prevalence and incidence of chronic kidney disease (CKD) (6–9). Evidence is growing worldwide that exposure to environmental toxins and toxicants can drive progression of NAFLD and contribute to the overall incidence of CKD (9–13).

Microcystin-LR (MCLR) is a cyanotoxin produced by cyanobacteria (blue-green algae) and is known to cause both hepatic and renal toxicities. MCLR is classified as a high-priority drinking water contaminant by the Environmental Protection Agency, is the focus of several National Toxicology Program studies, and is the only microcystin that has been given a WHO tolerable daily intake value (0.04 μg/kg or 2.4 μg/day for a 60 kg person) (14,15). Consumption of microcystin-contaminated water and aquatic animals is the most frequent route of exposure and is well documented to occur around the world (16). MCLR is primarily a hepatotoxin, due to being a substrate for organic anion transporting polypeptide (OATP)1B isoforms that are specifically expressed on the sinusoidal membrane of hepatocytes (17,18). MCLR-mediated hepatotoxicity has been suggested through multiple epidemiological studies. A recent case-control study in China linked serum microcystin levels with the risk of hepatocellular carcinoma (19). Another study in China showed that children may be an at-risk population for liver damage because aspartate aminotransferase (AST) and alkaline phosphatase markers of liver damage were higher in children who were exposed to drinking water with high levels of MC compared to those exposed to water with low levels of microcystin (20). Several other studies have linked microcystin in drinking water to primary liver cancer (21–23). In a cohort of fisherman with an estimated MC exposure between 2.2–3.9 μg/day, serum MC concentrations correlated with the plasma levels of ALT and AST (24). To date, a link between MC exposure and kidney disease has not been demonstrated, although substantial preclinical data indicate that renal toxicity is also associated with MC exposure (25–33).

Some data suggest a link between NAFLD pathogenesis and MC exposure. A novel satellite imaging technique was recently used to correlate blooms with county level incidences of NAFLD in the U.S. (34). Several preclinical studies also suggest that prolonged intraperitoneal or oral exposure (25–100 μg MCLR/kg) in rodent models alters hepatic lipid content, suggestive of steatosis (35–37). Another study reported that mice exposed to 40 μg/kg MCLR orally every other day for 90 days had a NASH phenotype as indicated by increased steatosis and hepatic inflammation (38). Finally, recent data showed that OATP1B/Oatp1b isoforms are altered in both NASH patients and rodent models of NASH, suggesting that MCLR disposition and toxicity may be altered in NAFLD (39,40). The present study was designed to determine the impact of NAFLD on MCLR toxicokinetics and whether underlying NAFLD alters acute MCLR-induced hepatic and renal toxicity profiles.

2. Materials and Methods

Toxicokinetic study

Male Sprague-Dawley rats (8-week old) were purchased from Harlan Laboratories (Indianapolis, IN) and maintained in 12-hour light and 12-hour dark cycles in a University of Arizona animal facility. Animal treatments were conducted in accordance with protocols approved by the University of Arizona’s Institutional Animal Care and Use Committee (protocol #08–117, approval date 6 Jan, 2014). Animals were divided into three groups and fed a control diet (D518754; Dyets Inc.,Bethlehem, PA), methionine and choline deficient (MCD) diet (D518810; Dyets Inc.), or a high fat/high cholesterol (HFHC) diet (D06061401; Research Diets Inc., New Brunswick, NJ) for 8 weeks. These diets were selected based on evidence from previous publications that they can be used to represent healthy animals and NAFLD animals, with some notable strengths and weaknesses of each. Strengths of the MCD diet include replication of NAFLD liver pathology and it is a good model to study NAFLD progression because it causes steatosis, inflammation, oxidative stress and fibrosis in less than 6 weeks (41–43). A major weakness of the MCD dietary model of NAFLD is that it causes weight loss and metabolic abnormalities that are not reflected in clinical NAFLD. The HFHC diet also produces a robust replication of NAFLD liver pathology (44–46). A previous report indicates that Sprague Dawley rats fed the MCD or HFHC diet for 8 weeks have NASH activity scores ≥ 4, indicating a positive NASH diagnosis (47). Finally, these dietary models of NAFLD replicate changes in transporters observed in human NAFLD livers, specifically Oatp1b transporters important for MCLR uptake (47). Surgical placement of cannula into the carotid artery, jugular vein, and common bile duct were performed as previously described (39). MCLR (Enzo Life Science Inc., Farmingdale, NY) was diluted in 0.9% saline (0.09% ethanol final) and administered into the jugular vein (20 μg/kg, 5 mL/kg), and blood was collected from the carotid artery at 2, 10, 20, 40, 80, and 120 minutes. Whole blood was centrifuged at 10,000 × g for 1 minute to collect plasma. Bile was collected in 15-minute increments over the course of the experiment. Rats were euthanized via exsanguination, and a portion of liver was fixed in formalin and paraffin embedded for histopathological analysis. The remainder of the liver was snap frozen in liquid nitrogen and stored at −80°C until further analysis.

Toxicokinetic analysis

The toxicokinetics of MCLR were obtained via noncompartmental methods using Phoenix WinNonlin (version 7.0, Certara, Princeton, NJ). Area under the plasma concentration versus time curve (AUC) from 0 to 120 min (AUC_0–120) was determined using the logarithmic trapezoidal method. Terminal slope (λ_z) was calculated via linear regression of at least the last three data points. AUC from 0 to infinite time (AUC_0-inf) was calculated as the sum of AUC_0–120 and the ratio of the concentration at 120 min to λ_z. Systemic clearance (Cl) was calculated as the ratio of dose to AUC_0-inf. Volume of distribution at steady state (V_ss) was calculated as the product of Cl and mean residence time (MRT), where MRT is the ratio of area under the moment curve from zero to infinite time (AUMC_0-inf) to AUC_0-inf. Terminal half-life (t_1/2) was calculated as the ratio of 0.693 to λ_z. Biliary clearance (Cl_bil) was calculated as the ratio of the cumulative amount of MCLR recovered in bile from 0 to 120 min (A_bil,0–120) to AUC_0–120.

MCLR quantification

MCLR-glutathione (MCLR-GSH) and MCLR-cysteine (MCLR-Cys) reference standards were prepared following a published method with minor modifications (48). Briefly, 2 mg of MCLR was reacted with either L-GSH or L-Cys in 4 mL of 5% potassium carbonate aqueous solution while stirring for 2 hours at room temperature. The reaction was neutralized with 4 mL of 0.04 M hydrochloric acid and extracted on a Strata-X 33μm polymeric reversed phase column (1 g) (Phenomenex, Torrance, CA). Purity of MCLR-GSH and MCLR-Cys was assessed by LC-MS/MS. The methods for MCLR, MCLR-GSH, and MCLR-Cys quantification in plasma and bile were adapted from previously published methods (49). The mobile phase consisted of 0.1% formic acid (A) and acetonitrile (B) and was pumped at a flow rate of 0.5 mL/min through a Waters Acquity UPLC BEH C18 column (1.7 μm, 2.1 × 50 mm) (Milford, MA). The UPLC gradient began at 50% B and increased to 100% B over 3.5 min, then was equilibrated to 50% B for 1 min before the next injection. Samples were analyzed using the QTRAP 6500 UHPLC-MS/MS system (AB Sciex, Framingham, MA). Multiple reaction monitoring in positive mode was used to detect MCLR at m/z 498.3 > 102.9, MCLR-GSH at m/z 651.9 > 104.9, and MCLR-Cys at m/z 558.9 > 102.8. Sixty μL of either plasma or bile were mixed with 60 μL of acetonitrile and vortexed. Then 560 μL of HPLC-grade water were added, and the sample was vortexed again. The samples were centrifuged at 12,000 × g for 10 minutes at 4°C. Solid phase extraction was performed on Oasis HLB μElution plates from Waters. Each well was conditioned with 200 μL of methanol, equilibrated with 200 μL of water, and 610 μL of diluted plasma or bile were loaded onto the sorbent. Wells were washed with 400 μL of 5% methanol and eluted with 200 μL of methanol. Samples were dried under a nitrogen stream and reconstituted in 60 μL of 25% acetonitrile with 0.1% formic acid. Fifty μL were injected onto the BEH column.

Single exposure toxicity study

Eight-week old male Sprague-Dawley rats were purchased from Envigo (Huntingdon, Cambridgeshire, United Kingdom). Handling, care, and maintenance of the animals took place in the Association for the Assessment of Laboratory Animal Care International accredited Program of the Laboratory Animal Resources facility of Washington State University, Spokane. All animals were maintained in 12-hour light and dark cycles for the duration of the study. The experimental protocol was approved by the Institutional Animal Care and Use Committee at Washington State University (protocol #04937–003, approval date 27 Jan, 2017). Similar to the toxicokinetics study above, animals were divided into three groups and fed a control diet (D518754), MCD diet (D518810), or HFHC diet (D06061401) for 6 weeks. Blood was collected from the tail vein under isoflurane anesthesia at the end of the dietary phase, and animals were acclimated in metabolic cages 24 hours prior to MCLR exposure. MCLR was diluted in 0.9% saline (0.09% ethanol final) and administered via intraperitoneal injection (60 μg/kg, 5 mL/kg). Urine was collected from 0–6 hours, 6–12 hours, and 12–24 hours, after which animals were returned to their standard housing until euthanasia 48 hours after MCLR exposure. Animals were euthanized by carbon dioxide asphyxiation. Whole blood collected before MCLR exposure and terminally was centrifuged at 10,000 × g for 1 minute to collect plasma. Portions of liver and kidney were fixed in 10% neutral buffered formalin, paraffin embedded and sectioned for histopathological analysis, and the remainder of the liver and kidney was snap frozen in liquid nitrogen and stored at −80°C until further analysis.

Histopathology

Liver and kidney sections were stained with hematoxylin and eosin and scored by a ACVP board certified veterinary pathologist. Liver was scored for lipid accumulation, necrosis, inflammation, and apoptosis, while kidney was scored for necrosis, degeneration, regeneration, tubule dilation, casts, and glomerular changes. Pathology scores were as follows: 0, no significant lesions (0%); 1, minimal (<10%); 2, mild (10–25%); 3, moderate (25–40%); 4, marked (40–50%); 5, severe (>50%).

Plasma chemistry and urine analyses

Commercially available kits were used to quantify plasma triglycerides, glucose, cholesterol, (all from Cayman Chemical, Ann Arbor, MI), and insulin (Millipore, Burlington, MA) according to the manufacturer’s protocols. Urine protein was quantified using the Pierce BCA protein assay kit from ThermoFisher Scientific (Waltham, MA), and urinary KIM1 protein was quantified using an ELISA (Boster biotechnology, Pleasanton, CA) according to the manufacturer’s protocol.

Western blots

Liver whole cell lysates were collected, and western blots were performed as previously described (39) with the following antibodies and conditions: Oatp1b2 (sc-376904, Santa Cruz Biotechnology, Santa Cruz, CA), 90°C with 5% 2-mercaptoethanol for 10 minutes; PP2A (05–421, EMD Millipore), 37°C with 5% 2-mercap toethanol for 30 minutes; MCLR (MC10E7, Enzo), 90°C for 10 minutes. Bio-Rad stain- free whole protein loading normalization was used as a loading control for each lane. Densitometry was performed using ImageJ software (National Institutes of Health, Bethesda MD).

Statistics

All results are presented as mean ± standard deviation (SD). One-way ANOVA was used for histopathology scoring, log transformed TK data, western blots, urine volumes, proteinuria, and urinary KIM1. One-way ANOVA p-values and Tukey’s multiple comparisons post-tests are reported with each analysis. Two-way ANOVA was performed on all plasma chemistries (glucose, insulin, triglycerides, cholesterol), with p-values and Sidak’s multiple comparison post-tests reported for each analysis.

3. Results

3.1. NAFLD alters MCLR toxicokinetics

MCLR (20 μg/kg) was administered intravenously to rats fed a control, MCD, or HFHC diet to determine MCLR toxicokinetics in healthy versus NAFLD animals. Blood was collected from the carotid artery at 2, 10, 20, 40, 80, and 120 minutes, and bile was collected in 15-minute increments to 120 minutes. The MCD and HFHC diets caused similar levels of steatosis, necrosis and apoptosis (Table 1). Both NAFLD diets caused inflammation, although it was more severe in HFHC than in MCD rats (Table 1). Steatosis predominately presented as macrovesicular in MCD rats and as microvesicular in HFHC rats (Figure 1).

Table 1:

Liver histopathology scoring from the intravenous 20 μg/kg MCLR toxicokinetic experiment in control, MCD, and HFHC rats.

	Control	MCD	HFHC	p-value
Steatosis	0.0 (0.0)	4.2 (0.3)*	3.9 (0.1)*	<0.0001
Inflammation	0.0 (0.0)	0.7 (0.2)*	1.4 (0.2)*†	<0.0001
Necrosis	0.0 (0.0)	0.8 (0.2)*	1.0 (0.0)*	<0.0001
Apoptosis	0.0 (0.0)	0.5 (0.2)	0.4 (0.2)	<0.0001

Data represent mean ± SD. n=6 for each group. One-way ANOVA p-values are shown in the final column.

^*p-value <0.05 versus control

^†p-value <0.05 versus MCD according to Tukey’s multiple comparison post-test.

Figure 1:

Liver photomicrographs from the intravenous 20 μg/kg MCLR toxicokinetic experiment in control, MCD, and HFHC rats.

Systemic plasma exposure to MCLR, as assessed by AUC_0–120, in MCD animals was approximately twice that of both the control and HFHC animals (Table 2; Figure 2). Because extrapolated AUC accounted for >30% of AUC_0-inf in four of five MCD rats and three of five HFHC rats, robust estimates of AUC_0-inf could not be obtained for the majority of the NAFLD animals; consequently, robust estimates of Cl, V_ss, and t_1/2 could not be obtained. As such, values for one (MCD) or two (HFHC) rats are reported (Table 2). Although the mean A_bil,0–120 was always <0.001% of the administered dose, absolute values for both NAFLD models were approximately 50% higher than control (Table 2, Figure 2). Compared to control, mean Cl_bil for the MCD rats did not differ, whereas that in the HFHC rats was approximately doubled, but a significant difference was not detected because of the large inter-individual 236 variation in the HFHC rats.

Table 2:

Toxicokinetics of MCLR after intravenous administration (20 μg/kg) to control, MCD, and HFHC rats.

	Control	MCD	HFHC
AUC0–120 (min* ng/mL)	968 ± 119	1846 ± 263*	920 ± 211†
AUC0-inf (min* ng/mL)	1106 ± 160	1622	762, 1124
Cl (mL/min/kg)	18.4 ± 2.9	12.3	26.3, 17.8
Vss (mL/kg)	857 ± 277	626	1558, 1412
t1/2 (min)	49.0 ± 17.6	43.3	60.5, 80.0
Abil,0–120 (pg/kg)	18.6 ± 4.11	27.5 ± 11.1	29.8 ± 12.9
Clbil (μL/min/kg)	0.019± 0.0021	0.016 ± 0.0066	0.036 ± 0.020

Data represent mean ± SD for control rats or individual values. n=6 control rats. n=5 for AUC_0–120 and n=4 for A_bil,0–120 and Cl_bil in MCD animals. n=5 for AUC_0–120, A_bil,0–120, and Cl_bil in HFHC animals. Individual values from one and two rats, respectively, in MCD and HFHC animals are presented for Cl, V_ss, and t_1/2.

^*p-value <0.05 versus control rats

^†p-value <0.05 versus MCD rats according to a one-way ANOVA and Tukey’s multiple comparison post-test.

Figure 2:

MCLR (A) plasma time course and (B) accumulated bile time course after 20 μg/kg MCLR intravenous injection to control, MCD, and HFHC rats. Data represent mean ± SD. Control n=6, MCD n=5, HFHC n=5 for each group.

Both MCLR-GSH and MCLR-Cys were below the detection limit for plasma and bile (0.5 ng/mL). Oatp1b2 is the main hepatic transporter responsible for MCLR uptake into hepatocytes from the plasma in rodent models, and we observed Oatp1b2 protein expression decreased in both dietary models of NAFLD (Figure 3). Protein phosphatase 2A (PP2A) is a key molecular target for MCLR-induced toxicity (covalent binding), and MCLR bound to PP2A acts as a surrogate for tissue exposure (50). PP2A and MCLR blots were run separately because both are detected at ~35 kDa. The bands shown in the MCLR western blot were the only proteins detected in MCLR-exposed liver samples (Figure 3). PP2A increased in MCD livers, while the amount of MCLR bound to PP2A decreased compared to control (Figure 3).

Figure 3:

Western blot analysis of Oatp1b2, PP2A, and MCLR bound to PP2A 2 hours after 20 μg/kg MCLR intravenous injection to control, MCD, HFHC rats. Data represent mean ± SD. n=6 for each group. One-way ANOVA p-values are shown below each graph. *p-value <0.05 versus control according to Tukey’s multiple comparison post-test.

3.2. Less hepatotoxicity in MCD-induced NAFLD

MCLR (60 μg/kg) was administered intraperitoneally to rats fed a control diet, an MCD diet, or an HFHC diet to determine single-dose MCLR toxicity in healthy versus NAFLD rats. After injection, rats were euthanized at 48 hours. Histopathology incidence and severity scoring for inflammation and necrosis were significantly different amongst the three groups, suggesting that MCD may have had less severe liver toxicity than control and HFHC (Table 3 and Figure 4). No significant differences in plasma glucose, insulin, and triglycerides were observed after MCLR exposure in control, MCD, or HFHC rats (Figure 5). MCLR exposure significantly increased plasma cholesterol in HFHC animals 48 hours after MCLR exposure (Figure 5). Oatp1b2 and PP2A protein expression were not different after a single high-dose of MCLR (Figure 6). The amount of MCLR bound to PP2A was significantly lower in the MCD group compared to the control group (Figure 6).

Table 3:

Liver histopathology scoring from the intraperitoneal 60 μg/kg MCLR toxicity experiment in control, MCD, and HFHC rats.

	Control	MCD	HFHC	p-value
Steatosis	0.0 (0.0)	4.0 (0.8)*	3.0 (0.0)*	<0.0001
Inflammation	2.3 (0.3)	1.5 (0.5)	3.0 (0.0)	0.0107
Necrosis	2.7 (0.3)	1.0 (1.0)*	3.3 (0.3)	0.0045
Apoptosis	2.3 (0.3)	1.3 (0.3)*	2.0 (0.0)	0.0355

Data represent mean ± SD. Control and HFHC, n=3 per group; MCD, n=4. One-way ANOVA p-values are shown below each pathology score graph.

^*p-value <0.05 versus control and according to Tukey’s multiple comparison post-test.

Figure 4:

Liver photomicrographs from the intraperitoneal 60 μg/kg MCLR toxicity experiment in control, MCD, and HFHC rats. Data represent mean ± SD. Control and HFHC, n=3 per group; MCD, n=4. Two-way ANOVA p-values are shown in the table. *p-value <0.05 versus 0hr according to Sidak’s multiple comparison post-test.

Figure 5:

Plasma glucose, insulin, triglycerides, and cholesterol levels 48 hours after 60 μg/kg MCLR intraperitoneal injection to control, MCD, and HFHC rats. Black bars represent before MCLR injection and white bars represent 48 hours after MCLR injection. Data represent mean ± SD. Control and HFHC, n=3 per group; MCD, n=4. Two-way ANOVA p-values shown in the table below each graph. *p-value <0.05 versus 0hr according to Sidak’s multiple comparison post-test.

Figure 6:

Western blot analysis of Oatp1b2, PP2A, and MCLR bound to PP2A 48 hours after 60 μg/kg MCLR intraperitoneal injection to control, MCD, HFHC rats. Data represent mean ± SD. n=3 for each group. One-way ANOVA p-values are shown below each graph.

3.3. Higher proteinuria and urinary KIM1 in HFHC-induced NAFLD compared to healthy

Urine was collected for 24 hours after the single 60 μg/kg MCLR exposure in three increments (0–6, 6–12, and 12–24 hours) to measure markers of renal toxicity, and kidneys were collected at euthanasia for histopathology. Kidney histopathology incidence and severity scoring for necrosis, degeneration, regeneration, tubule dilation, casts, and glomerular change indicated no difference between control, MCD, and HFHC rats 48 hours after MCLR exposure (Table 4). MCD rats had the lowest average urine volume, proteinuria, and urinary KIM1 amongst the groups (Figure 7). Proteinuria (0–6 hour and 6–12 hour) and urinary KIM1 (0–6 hr) were higher in HFHC animals compared to control (Figure 7).

Table 4:

Kidney histopathology scoring from the intraperitoneal 60 μg/kg MCLR toxicity experiment in control, MCD, and HFHC rats.

	Control	MCD	HFHC	p-value
Overall score	1.3 (0.9)	0.5 (0.6)	1.7 (0.3)	0.2491
Necrosis	0.7 (0.3)	0.0 (0.0)	0.3 (0.3)	0.2039
Degeneration	0.7 (0.3)	0.0 (0.0)	0.7 (0.3)	0.1278
Regeneration	0.3 (0.3)	0.0 (0.0)	0.7 (0.3)	0.2039
Tubule dilation	1.0 (0.6)	0.0 (0.0)	1.3 (0.3)	0.0299
Casts	1.3 (0.9)	0.0 (0.0)	1.7 (0.3)	0.0338
Glomerular change	1.0 (1.0)	0.8 (0.9)	1.7 (0.3)	0.5388

Data represent mean ± SD. Control and HFHC, n=3 per group; MCD, n=4. One-way ANOVA p-values are shown below each pathology score graph.

Figure 7:

(A) Urine volume, (B) proteinuria, and (C) urinary KIM1 48 hours after 60 μg/kg MCLR intraperitoneal injection to control, MCD, HFHC rats. Data represent mean ± SD. Control and HFHC, n=3 per group; MCD, n=4. Two-way ANOVA p-values shown in the table below each graph. *p-value <0.05 versus 0hr according to Sidak’s multiple comparison post-test.

4. Discussion

Cyanobacteria blooms are common in bodies of water worldwide, posing a threat to public health (15,16,51–53). This problem is expected to grow because the frequency and magnitude of cyanobacterial blooms are predicted to increase as anthropogenic eutrophication of water systems and global climate change worsen (54). MCLR produced by cyanobacteria is readily soluble in water and contaminates surface water and aquatic species. One study reported that people consuming fish and water from temperate lakes in North America and tropical lakes in Africa could exceed the total dietary intake of 2.4 μg/day (16). With the increasing prevalence of NAFLD, people with NAFLD likely will experience acute exposures to MCLR. The present study is the first to explore MCLR toxicokinetics and acute liver and kidney toxicity in the context of NAFLD, and the data suggest that NAFLD may be a risk factor for altered MCLR exposure and may alter susceptibility to specific hepatic and renal toxicities.

Concentration and duration of toxin exposure are two important factors that determine the site and degree of toxicity. For each toxin, there is a unique set of metabolizing enzymes and/or transporters within each tissue that are responsible for detoxification and elimination. MCLR is a large molecular weight organic compound (995 Da) that is dependent on OATP1B/Oatp1b isoforms for hepatic uptake and toxicity (17,18). Compared to control rats, hepatic Oatp1b2 protein expression decreased in both MCD and HFCF rats, consistent with the prolonged elimination of MCLR in the two NAFLD models (Figure 2). Although robust estimates of systemic Cl could not be obtained for all MCD rats, the significant increase in AUC_0–120 likely reflected a decrease in systemic Cl. In contrast, the AUC_0–120 for HFHC rats was unchanged, suggesting systemic Cl was unchanged. Based on two HFHC rats, V_ss roughly doubled. Because biliary Cl represents <0.01% of systemic Cl (Table 2), the MCD diet decreased other routes of MCLR Cl, including renal and metabolic Cl. A longer collection period (several hours) for plasma, as well as urine, is required to delineate the contribution of these different clearance pathways. Despite that biliary Cl represents a very small fraction of total Cl, biliary Cl in HFHC rats was approximately twice that for control and MCD rats, which could reflect induction of a canalicular efflux transporter by the HFHC diet. Results from these two unique NAFLD models collectively suggest that impaired hepatic uptake, as well as higher body weight and/or body composition, may contribute to the altered distribution and elimination of MCLR.

The differences between the two NAFLD models are important for the interpretation and potential clinical implication of the current results. For example, it has previously been reported that the MCD model recapitulates NAFLD liver pathology and the alterations in drug metabolizing enzymes and transporters observed in clinical NAFLD with the highest fidelity, but the MCD model does not produce obesity (47). In contrast, animals on HFHC diet gain more weight than controls and typically better recapitulate the metabolic features of NAFLD (55). Thus, although only the MCD rats exhibited increased MCLR exposure, factors such as obesity, severity of NAFLD, and the presence of metabolic abnormalities may have unique effects on MCLR toxicokinetics and toxicities.

Evidence for the detrimental role of environmental contaminants in the development and progression of NAFLD continue to grow (13,56,57). The mechanism behind how these contaminants affect NAFLD include differential exposures and exacerbation of pathological features in NAFD. Data from the present study indicate that liver inflammation, necrosis and apoptosis were lowest in the MCD group after high dose MCLR exposure, suggesting that preexisting NAFLD attenuated MCLR-induced liver pathology compared to MCLR exposed controls. This observation may be explained by less hepatic MCLR exposure in the MCD group, as suggested by the toxicokinetics data and decreased MCLR bound to PP2A in the liver. Liver inflammation is a key feature in the diagnosis of NASH and in the present study HFHC had the highest level of inflammation after MCLR exposure, which was above the level of inflammation observed in the toxicokinetics study, suggesting that diet-induced NAFLD and MCLR toxicity contributed to higher inflammation in HFHC. Recent Aroclor-1260 and perchloroethylene data indicate that exposure to these hepatotoxicants did not cause significant liver damage in controls but exacerbated liver disease in rodent NAFLD, which is contrast to the current observations for MCLR toxicity in NAFLD (13,56). These differences likely reflect different duration of exposure (single MCLR exposure versus sub-chronic Aroclor and perchloroethylene exposure) and/or the specific mechanisms of hepatic exposure and toxicity. For example, sub-chronic exposure to MCLR has been reported to increase hepatic lipid content (36,37,58,59) and prolonged oral exposure produced a disease phenotype similar to NAFLD (38). In addition, MCLR has been reported to negatively affect insulin signaling via inhibition of glycogen synthase, potentially contributing to higher glucose levels (60). In the current single exposure study, MCLR did not change plasma glucose, insulin, or triglycerides, again, potentially due to single versus repeated MCLR exposure. Interestingly, MCLR toxicity increased plasma cholesterol levels only in the HFHC group, indicating that MCLR toxicity decreased the ability of the liver to process dietary cholesterol. Together, these data support the important role of hepatic exposure in MCLR-induced hepatotoxicity and suggest that increased inflammation caused by MCLR toxicity in HFHC-induced NAFLD may increase disease severity.

CKD is characterized by the progressive loss of kidney function, specifically decreased glomerular filtration and/or increased urinary albumin excretion that can progress to kidney failure. The prevalence of CKD worldwide is estimated to be between 8–16%, but increasing evidence suggests that patients with NASH have increased risk of developing CKD (9,61). One study reported the prevalence of CKD to be 21% among NASH patients but only 6% in non-NASH controls (62). Another study reported that ~85% of CKD patients had NAFLD as determined by elastography (8). Currently, the mechanisms for development of CKD in NAFLD are unclear, but it has been speculated that factors such as diabetes, obesity, dyslipidemia, inflammation, and oxidative stress are involved (8,63). Exposure to environmental contaminants are also expected to contribute to the overall burden of CKD (9). Although differences in kidney MCLR-induced pathology between healthy and NAFLD groups were not observed in the present study, the data showed higher proteinuria and urinary KIM1 in the HFHC model of NAFLD compared to controls, suggesting that the combination of MCLR and HFHC-induced NAFLD have a greater effect on glomerular filtration and/or tubule cell damage than MCLR toxicity in control animals. Currently, the renal transporters involved in MCLR uptake and toxicity are not known, hence why expression of renal Oatps was not assessed in the current work. Together, these data indicate that MCLR exposure in NAFLD could be a contributing factor to CKD in this population.

This work is the first to show altered MCLR toxicokinetics in NAFLD, resulting in higher and longer systemic exposure to MCLR. The potential resultant decreased hepatic exposure in MCD-induced NAFLD suggests that NAFLD patients with impaired OATP1B function may be partially protected from MCLR-induced toxicity. The HFHC model had higher plasma cholesterol, proteinuria, and urinary KIM1 in the MCLR exposed HFHC model compared to MCLR exposed controls, suggesting that MCLR exposure in NAFLD may alter cholesterol metabolism and increase the risk of developing CKD. Future work is needed to investigate the effect of repeated exposure on MCLR toxicity in NAFLD to determine if sub-chronic exposure alters the toxicity profile.

Highlights:

• Systemic exposure to MCLR increased approximately two-fold in MCD-induced NAFLD.

• Biliary clearance of MCLR increased approximately two-fold in HFHC-induced NAFLD.

• MCD-induced NAFLD had the lowest hepatic MCLR levels and hepatic toxicity.

• HFHC-induced NAFLD exposed to MCLR had the highest plasma cholesterol, proteinuria, and urinary KIM1.

Abbreviations:

AUC

area under the plasma concentration versus time curve

AST

aspartate aminotransferase

Cl_bil

Biliary clearance

CKD

chronic kidney disease

HFHC

high fat/high cholesterol

MCD

methionine and choline deficient

MCLR

microcystin-LR

MCLR-Cys

microcystin-LR-cysteine

MCLR-GSH

microcystin-LR-glutathione

NAFLD

nonalcoholic fatty liver disease

NASH

nonalcoholic steatohepatitis

OATP

organic anion transporting polypeptide

PP2A

protein phosphatase 2A

Cl

systemic clearance

t_1/2

terminal half-life

V_ss

volume of distribution at steady state