thr Protein Phosphatase Inhibition by Microcystin in the Little Skate Raja erinacea

Toxicology and Applied Pharmacology 161, 40 – 49 (1999) Article ID taap.1999.8783, available online at http://www.idealibrary.com on

Hepatic Toxicity and Persistence of ser/thr Protein Phosphatase Inhibition by Microcystin in the Little Skate Raja erinacea Maria Runnegar,* ,† ,1 David J. Seward,* ,‡ Nazzareno Ballatori,* ,§ James M. Crawford, ¶ and James L. Boyer* ,i *Mount Desert Island Biological Laboratory, Salsbury Cove, Maine 04672; †Department of Medicine and USC Center for the Study of Liver Disease, University of Southern California, Los Angeles, California 90033; ‡Williams College, Williamstown, Massachusetts 01267; §Department of Environmental Medicine, University of Rochester, Rochester, New York 14642; ¶Department of Pathology, Immunology, and Laboratory Medicine, University of Florida, Gainesville, Florida 32610; and iDepartment of Medicine and Liver Center, Yale University, New Haven, Connecticut 06510 Received June 17, 1999; accepted August 25, 1999

blooms in waters rich in nutrients releasing toxins in the environment. Mcyst (and related toxins) are a family of small cyclic peptides (seven or five amino acids) that are all hepatotoxic but differ in relative toxicity, with LD50 ranging between 50 and 800 mg/kg ip in mice (Stotts et al., 1993). Ingestion of water containing these toxins results in significant numbers of deaths of terrestrial and aquatic animals (Carmichael, 1994). Through an unfortunate accident that resulted in the death of 60 of 126 patients exposed to water contaminated with Mcyst, we now know that the effects of Mcyst in humans parallel those in animals, with “a novel acute toxic hepatitis, similar to that seen in animals exposed to microcystins” (Pouria et al., 1998; Jochimsen et al., 1998). The mode of action of Mcyst as a toxin at the cellular level is the potent specific inhibition of PP1 and PP2A (MacKintosh et al., 1990). This inhibition always precedes and is essential for any histological evidence of toxicity (Runnegar et al., 1993). The interaction between Mcyst and the PP catalytic subunit that results in the inhibition of the catalytic activity is followed by the formation of a secondary covalent bond between the enzyme and Mcyst (Runnegar et al., 1995c; MacKintosh et al., 1995; Craig et al., 1996). The consequence of this sustained PP inhibition is a change in the phosphorylation status of the cell, resulting in sustained cytoskeletal and metabolic disturbances (Hooser et al., 1991; Wickstrom et al., 1995; Runnegar et al., 1997). Most of the work characterizing the toxicity of Mcyst has been done in the rat and mouse, although some studies have shown Mcyst to be hepatotoxic to a number of other mammals, including sheep, cattle, pigs, as well as chickens. In these animals death follows hepatocyte necrosis, loss of hepatic integrity, and massive hemorrhage (Falconer et al., 1981). Because of the occurrence of these toxins in both freshwater and marine environments (more recently shown), vertebrate and invertebrate aquatic animals are at considerable risk of intoxication (Chen et al., 1993). The effects of these toxins already have significant economic impact in the commercial rearing of Atlantic salmon in seawater pens. In British Columbia and Washington State, yearly occurrences since 1986 of a

Hepatic Toxicity and Persistence of ser/thr Protein Phosphatase Inhibition by Microcystin in the Little Skate Raja erinacea. Runnegar, M., Seward, D. J., Ballatori, N., Crawford, J. M., and Boyer, J. L. (1999). Toxicol. Appl. Pharmacol. 161, 40 – 49. Microcystin-induced ser/thr protein phosphatase (PP) inhibition and toxicity were examined in the little skate (Raja erinacea), an evolutionarily primitive marine vertebrate. As in mammals, PP inhibition and toxicity were exclusively hepatocellular, but were much more persistent in the skate. A dose of 63 mg/kg given iv to adult male skates resulted in the near complete inhibition of hepatic PP activity at 24 h. PP activity was still 95% inhibited 7 days after dosing in skates given 125 mg/kg microcystin. Mortality occurred at doses of 500 mg/kg or more. Hepatic lesions were only seen in animals with fully inhibited PP activity in liver. The histological changes seen at 125 mg/kg were mild periportal inflammatory changes increasing in severity together with hepatocyte necrosis at higher doses of microcystin. Microcystin persisted and could be detected in plasma up to 7 days after dosing. This finding shows that, in the skate, as in mammals, the liver is the only organ capable of uptake of microcystin, since there was no significant inhibition of PP activity in the rectal gland and small decreases in PP activity of the kidney that were not time or dose dependent. In vitro microcystin caused dose-dependent inhibition of PP activity in isolated skate hepatocytes, while it was without effect in cultured rectal glands. Uptake of microcystin and the accompanying inhibition of PP activity in skate hepatocytes was prevented by the addition of a series of organic dyes and bile acids. The spectrum of inhibitors of microcystin uptake in skate is similar to that seen in the rat, indicating common features of the carrier(s) in these diverse species. © 1999 Academic Press Key Words: microcystin; Raja erinacea; protein phosphatase inhibition; hepatotoxicity; liver transport.

Microcystins (Mcyst) are potent hepatotoxic protein phosphatase (PP) inhibitors that are produced by natural blooms of freshwater and marine cyanobacteria (MacKintosh et al., 1990; Runnegar et al., 1993). These microorganisms form dense 1

To whom correspondence should be addressed

0041-008X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

40

PP INHIBITION AND MICROCYSTIN TOXICITY IN THE SKATE

liver disease (NLD, netpen liver disease), characterized by severe necrosis and hepatic megalocytosis of the fish have been shown to be caused by Mcyst (Andersen et al., 1993). Experimentally, radiolabeled Mcyst administered to salmon in vivo resulted in the hepatic accumulation of the toxin (Williams et al., 1995, 1997a). More recently bioaccumulation of Mcyst in the saltwater mussel Mytilus edulis has been demonstrated (Williams et al., 1997b). Hepatic and renal histopathological changes following dosing with Mcyst have been described in the young carp (Rabergh et al., 1991) and in yearling rainbow trout (Kotak et al., 1996). In the carp there was hydropic degeneration and necrosis in the liver. In addition, in Mcystdosed carp there was dilation of the Bowman capsule and degenerative changes in the epithelial cells and glomeruli of the kidney (Rabergh et al., 1991). In the rainbow trout hepatocytes became swollen and vacuolated followed by diffuse coagulative hepatocellular necrosis. There was also coagulative necrosis of the tubular epithelium with dilation of Bowman’s space in some of the fish (Kotak et al., 1996). The kidney damage following Mcyst exposure has been exclusively associated with fish and does not occur in mammalian intoxication (Falconer et al., 1981; Hooser et al., 1989; Runnegar et al., 1993). To date no studies have been done to address the effect of Mcyst on PP activity and its toxicity in any freeliving vertebrates including fish. The purpose of this study was to determine the toxicity of Mcyst in an outbred population of skates as a model marine species. Many aspects of their physiology and cell biology have been well explored. In particular, it has been established that there are differences in the specificity of hepatic transporters of skates compared to mammals (Fricker et al., 1987, 1994; Boyer et al., 1993; Blumrich et al., 1993). In mammals Mcyst uptake occurs by an as yet unidentified carrier that is exclusively found in hepatocytes. Uptake of Mcyst is inhibited by a well-characterized range of inhibitors. The presence and activity of this carrier determines the susceptibility to Mcyst hepatotoxicity (Runnegar et al., 1995a).

41

and 350 mM urea [Smith et al., 1987]). The skates were anesthetized with an injection of pentobarbital sodium (2.5 mg/kg) via the caudal vein before euthanasia by exsanguination at 24, 48, and 72 h up to 7 days after dosing. Three skates (of 11 total) injected with the higher dose of Mcyst died during the experimental period. Blood samples were taken daily for 3 days after dosing and for the 7-day time point. Study 2. Three skates per dose were injected iv with Mcyst and killed 24 h after injection; three controls were included in the study. The liver, kidneys, and rectal gland were removed and weighed. Samples of each organ were quickly frozen for the measurement of PP activity. Samples were also fixed in 4% neutral buffered formalin for routine processing for histology (embedded in paraffin and then cut into sections of 4 – 6-mm thickness, stained with hematoxylin and eosin (H & E), and examined by light microscopy). Cell Isolation Isolated skate hepatocytes and cell clusters were prepared by collagenase perfusion of isolated livers as previously described and then maintained on ice in an elasmobranch Ringer buffer. For experiments cells were resuspended in fresh elasmobranch Ringer and Mcyst and/or other reagents added followed by incubation for the indicated times (Smith et al., 1987). Shark rectal gland cultures were prepared as previously described (Valentich and Forrest, 1991; Silva et al., 1993). Mcyst or calyculin A (LC Laboratories) were added to the confluent monolayer of the cells in Dulbecco’s modified Eagle’s-Ham’s F-12 and incubated for 24 h before processing. Inhibition of PP Activity and Mcyst Uptake Skate hepatocytes in suspension were incubated for 30 min at 15°C with varying amounts of Mcyst to determine the dose response for PP inhibition. To determine the effect of the presence of various organic anions and bile acids on PP inhibition by Mcyst, hepatocytes were also preincubated for 1 min with the test compounds before the addition of 10 mM Mcyst followed by incubation for 30 min at 15°C. Uptake of Mcyst was determined by the incubation of hepatocytes with tracer amounts of 125I-radiolabeled Mcyst together with 10 mM unlabeled Mcyst (Runnegar et al., 1995a). For initial uptake, samples were taken 10, 30, and 60 s after the addition of Mcyst and diluted in cold elasmobranch Ringer followed immediately by centrifugation to separate cells from medium. The radioactivity associated with the cell pellet was determined by gamma counting of the pellet. Total radioactivity was determined on an aliquot of the incubations. PP Assays

MATERIALS AND METHODS Animals All studies were performed at the Mount Desert Island Biological Laboratory (Salsbury Cove, ME) during the month of July in 1996, 1997, and 1998. Male skates (Raja erinacea; 0.60 –1.0 kg) were netted off the coast of Mount Desert Island and maintained in large aerated tanks supplied with continuously flowing 15 6 2°C seawater. Dogfish sharks (Squalus acanthias) of either sex were captured by trawl by commercial fishermen and kept in large aerated tanks supplied with continuously flowing 15 6 2°C seawater. For in vivo dosing with microcystin the skates were first weighed and a sample of blood taken from the caudal vein. Mcyst (Mcyst-YM from LC Laboratories, Woburn, MA) diluted in elasmobranch Ringer buffer solution (0.5 ml/kg) was injected in the caudal vein. Study 1. Two to three skates per dose (125, 250, and 500 mg/kg) for each time point were injected iv with Mcyst as well as five controls that were injected with elasmobranch Ringer buffer only (270 mM NaCl, 4 mM KCl, 2 mM CaCl 2, 3 mM MgCl 2, 8 mM NaHCO 3, 1 mM KH 2PO 4, 0.5 mM Na 2SO 4,

Tissue extract preparation for PP assay. The tissue extraction procedure follows that of Cohen et al. (1988). Weighed portions of the liver, kidney, or rectal gland were homogenized using a Polytron at 4°C with 5 vol of 50 mM Tris/HCl, pH 7.0, containing 1 mM DTT, 0.1 mM EDTA, 1% Nonidet, and the protease inhibitors phenylmethylsulfonyl fluoride (0.1 mM), benzamidine (1 mM), pepstatin (1 mg/ml), and leupeptin (0.5 mg/ml) (lysis buffer). These suspensions were then diluted to 10% (w/v) with the homogenizing buffer and then centrifuged at 10,000g for 10 min. The supernatants (tissue extracts) were used in the PP assays after suitable dilution. Cell extract preparation for PP assay. Cell extracts were prepared essentially as for the tissue extracts. Following incubations, cell suspensions (hepatocytes) were gently but exhaustively washed by centrifugation with elasmobranch Ringer buffer solution to remove any remaining extracellular Mcyst. The pellets were then lysed in 0.3– 0.6 ml of lysis buffer. Cultured cells (shark rectal gland cells) were washed repeatedly with buffer while attached to the culture dish before lysing as for hepatocytes. Shark rectal gland cells that were treated with calyculin A detached from the culture dish and were washed and lysed in suspension.

42

RUNNEGAR ET AL.

Assay of protein phosphatases 1 and 2A (PP1 and PP2A) in tissue or cell extracts. The PP activity in tissue extracts or cell lysates was determined by measuring the release of inorganic [ 32P]phosphate from [ 32P]phosphorylase a at 30°C (Cohen et al., 1988) as described in Runnegar et al. (1995a,b). PP activity is reported as activity in units (nmol of inorganic phosphate released/min) per milligram of protein in the extract. Persistence of PP-Inhibiting Activity of Mcyst in Skate Plasma The fact that Mcyst is a potent inhibitor of PP activity was used to determine if any residual Mcyst remained in the circulation. Dilutions of a solution of pure Mcyst of known concentration were used to construct a standard curve for inhibition of PP activity by Mcyst of an extract from untreated liver. Serial dilutions of skate plasma were added to the liver extract and PP activity was determined. The standard curve for pure Mcyst was used to convert the degree of inhibition by plasma to concentration of Mcyst. Mcyst remaining in the circulation was determined in plasma collected at time 0, 24, 48, and 72 h, and 7 days after dosing. The plasma was diluted in 50 mM Tris/HCl, pH 7.0, containing 1 mM DTT; dilutions were adjusted to give inhibition of between 20 and 80% of control. Plasma from untreated skates had no effect on the PP activity assay. The plasma volume of skates was taken to be 37.5 ml/kg (total blood volume approximately 4.5% of body weight, hematocrit of 10 –30%) Statistical Analysis Data were analyzed from individual skates and each skate hepatocyte preparation (n 5 1). Data for the rectal gland cultures are obtained from a number of animals. Each serial skate plasma sample was also taken as n 5 1. For in vitro experiments duplicate or triplicate incubations were done for each condition. The mean of each duplicate from one experiment (one cell preparation) was considered n 5 1. The results from multiple experiments/animals for treatment and control groups were compared by either Student’s t test or one-way analysis of variance followed by Fisher’s least significance test. The level of significance was p , 0.05. Unless stated otherwise, results are shown as means 6 SE of n (the number of experiments or animals).

RESULTS

Whole Animals Study 1. The effect of varying doses of Mcyst on PP activity of skate liver, kidney, and rectal gland were investigated over a period of 7 days. Doses of 125, 250, and 500 mg/kg resulted in profound PP inhibition. Over 90% inhibition was seen by 24 h and PP activity of liver remained fully inhibited even after 7 days (Figs. 1A–1C). PP activity in the kidneys of these skates was slightly but significantly decreased without any detectable dose or time dependence. There was no change in PP activity in rectal gland extracts (Figs. 1A–1C). Three skates dosed with 500 mg/kg Mcyst died (after 48 and 96 h, and 6 days, respectively). Some of the skates dosed with 250 and 500 mg/kg Mcyst appeared to be more lethargic. Gross examination of these livers showed varying degrees of hemorrhage and necrosis. No gross changes were seen in other organs. Study 2. The Mcyst doses for Study 1 were derived from the limited information on toxicity of Mcyst in aquatic animals. Since the lowest dose used in Study 1 still resulted in full PP inhibition even 7 days after dosing, lower doses (5, 32, and 63 mg/kg) were administered iv and animals were euthanized after 24 h. The PP

FIG. 1. Time dependence of Mcyst inhibition of PP activity in skates. Mcyst was administered iv to free swimming skates at doses of 125 mg/kg (A), 250 mg/kg (B), and 500 mg/kg (C). PP activity in control livers 5 3.20 6 0.40 units/mg protein, kidneys 5 4.32 6 0.19 units/mg protein, and rectal glands 5 5.95 6 0.55 units/mg protein. n 5 5 for controls, for treated skates n 5 1–3. Values for n 5 2 or 3 animals are means 6 SE. PP activity for liver, kidney, or rectal gland at each time point and Mcyst dose was for n 5 3 except when indicated beneath the respective bars, *p # 0.05.

activity in these skate livers dosed with 63 mg/kg was only 15 6 7% of controls, whereas, at a dose of 32 mg/kg, no detectable inhibition of PP activity was noted (Fig. 2). PP activity in skate kidney was not inhibited. PP activity of the rectal gland of the skates was not determined in this study, as no effect of Mcyst on this activity was observed at higher doses (Fig. 1). Histology of Skate Liver and Kidney in Study 1 All skates treated with Mcyst showed histologic changes in liver compared to controls (Fig. 3A). At the lowest dose (125

PP INHIBITION AND MICROCYSTIN TOXICITY IN THE SKATE

43

Microcystin in Plasma Circulating Mcyst could still be detected in skate plasma 7 days after dosing (Fig. 4A). Mcyst was detected by measuring the ability of diluted plasma to inhibit PP activity of liver extracts in vitro. The initial dose determined the amount of Mcyst in the circulation for each time point. Residual Mcyst in plasma decreased with time but was still detectable even 7 days after dosing with 125 mg/kg. The percentage of the dose remaining in the circulation at 24 h was proportionally greater with increasing doses of Mcyst (Fig. 4B). PP Inhibition by Mcyst in Isolated Hepatocytes

FIG. 2. Dose dependence of inhibition of PP activity by Mcyst after 24 h in the liver and kidney of skates expressed as % of control; for Mcyst dose of 5 mg/kg the number of skates was n 5 2, for 32 and 63 mg/kg the number of skates was n 5 3. Values are means 6 SE, *p # 0.05.

mg/kg) at 24 h there was mild portal and periportal inflammatory changes (Fig. 3B). Dosing with 250 mg/kg Mcyst resulted in more pronounced portal and periportal inflammatory changes as well as some areas of hepatocyte necrosis (Fig. 3C). At 500 mg/kg there were widespread inflammatory changes and extensive hepatocyte necrosis (Fig. 3D). These findings were consistently observed in the skates tested. Hepatocyte necrosis and mononuclear and eosinophilic infiltrates correlated with increasing doses of Mcyst. Mononuclear and eosinophilic inflammation of the portal tracts was also observed, as well as areas of bile duct proliferation and interface hepatitis. Although none of these changes was present in control livers, there did not appear to be a correlation of portal tract changes with dose or time. Hepatocellular apoptosis and vascular congestion were seen in only a few of the animals. Extensive microvesicular and macrovesicular steatosis were also observed, but were also present to an equivalent degree in control skates. There was no evidence of cholestasis, no neutrophilic infiltrate in the parenchyma or in the portal tracts, and no detectable changes in the terminal hepatic veins of Mcyst-treated skates. Although no tubular necrosis was observed, about a third of the kidneys had visible degrees of tubular vacuolization that was not observed in kidneys from control skates. This abnormality did not appear to be dose or time dependent and did not correlate with death or a moribund state of the skate. There was no significant effect of Mcyst in rectal gland histology. Histology of Skate Liver and Kidney in Study 2 In skates dosed with 63 mg/kg or less there were no detectable changes in histology of the liver, kidney, and rectal gland.

Mcyst inhibited PP activity in cell suspensions of skate hepatocytes. The dose dependence of the inhibition is shown in Fig. 5A. After 30 min of exposure there was significant inhibition of PP activity with 2 mM Mcyst. PP inhibition by Mcyst was time as well as dose dependent. An incubation time of 1 h with 1 mM Mcyst reduced PP activity of skate hepatocytes to 38 6 18% of control, while after 2 or 3 h incubation 1 mM Mcyst reduced PP activity to about 10% of control value (Fig. 5C). In contrast, PP inhibition by calyculin A (a cell-permeant PP inhibitor whose potency and specificity is almost identical to that of Mcyst when added to enzyme solutions or tissue extracts) was not time dependent. The PP activity of hepatocytes incubated with 500 nM calyculin A for 1 to 4.5 h was fully inhibited, while a lower dose of calyculin A had no effect on the PP activity (Figs. 5B and 5C). Protection of Hepatotoxicity by Some Organic Solute Transport Inhibitors A number of substrates have been shown to protect mammalian hepatocytes from Mcyst inhibition of PP activity by diminishing Mcyst uptake into the hepatocyte (Runnegar et al., 1995a). These were also tested in skate hepatocytes. A variety of organic dyes as well as bile acids (cholate and taurocholate) protected hepatocytes from PP inhibition by Mcyst (10 mM) when added prior to exposure to the toxin (Fig. 6). This substrate protection resulted from inhibition of initial uptake (Fig. 7) and accumulation of Mcyst (Fig. 6) (measured as 125 I-Mcyst) in these cells when compared to hepatocytes incubated with Mcyst alone. Preincubation with glutamate had no effect on PP activity or Mcyst accumulation. Effects on Cultured Shark Rectal Gland Cells The effect of Mcyst on cultures of shark rectal gland cells was investigated to determine whether Mcyst PP inhibition was limited to hepatocytes. Addition of 5 mM Mcyst to shark rectal gland cells had no significant effect on PP activity even after 24 h incubation. PP activity in control cells was 7.28 6 1.94 units/mg of protein. After exposure to Mcyst (5 mM), the PP activity was 85 6 13% of control value (n 5 5) from two

44

RUNNEGAR ET AL.

PP INHIBITION AND MICROCYSTIN TOXICITY IN THE SKATE

45

FIG. 4. Persistence of Mcyst in plasma. Mcyst was determined by measuring the ability of diluted plasma to inhibit PP activity in control liver extracts. (A) Mcyst remaining in the circulation: time and dose dependence for skates dosed iv with 125, 250, And 500 mg/kg Mcyst. To allow comparisons between skates of different weight, results are shown as mg of Mcyst present in the total plasma volume of skates normalized to weigh 1 kg. (B) Mcyst remaining in plasma 24 h after dosing with 32 to 3000 mg/kg expressed as % of administered dose. For A and B, n 5 2–5 means 6 SE (n 5 individual plasma samples for each dose). In B, the values for c, d, e, and f (doses of Mcyst greater than 125 mg/kg) are all significantly different from a (32 mg/kg) and from b (63 mg/kg). Values for a to e (32–500 mg/kg) are all significantly different from f (3000 mg/kg).

separate cell isolations. Mcyst-treated cells remained well attached to the culture plates. Light microscopy revealed no detectable difference in morphology between control and Mcyst-treated cells. In contrast, incubations with calyculin A (500 nM) decreased PP activity to 13% of control values. Cells were all detached from the culture plate and rounded up, reflecting changes in the cytoskeleton that result from the profound inhibition of PP activity (Eriksson and Goldman, 1993). DISCUSSION

In mammals Mcyst inhibits PP activity of liver in a dosedependent manner. Similar findings were observed in the present study in livers from the marine skate. Intracellular concentrations of PP1 and PP2A in mammals have been estimated to be between 0.1 and 1.0 mM (Cohen et al., 1988). Mcyst concentrations of 0.3 mg/g liver (0.3 mM) are sufficient to inhibit 50% of hepatic PP activity in the mouse (Runnegar et al., 1993). Thus, approximately equimolar concentrations of Mcyst are enough to significantly inhibit PP activity, demonstrating the potency of this group of toxins. Since PP activity of control skate liver in units/g of tissue or as units/mg of protein is of the same order of magnitude as that of the mouse, we can

conclude that the concentration of PP1 and PP2A in skate liver is similar to that of mouse given the highly conserved nature of these enzymes (Cohen, 1989). Although concentrations of Mcyst in skate were not measured in this study, we can estimate them from the administered dose; 63 mg/kg (63 nmol/ kg) Mcyst (Fig. 2) was sufficient to inhibit PP activity of liver to 15% of control values. Skate livers weighed about 30 g and were 2.89 6 0.35% of body weight. PP concentrations of about 1 mM represent about 30 nmoles of enzyme/liver, so that the minimal amount of Mcyst needed to inhibit PP activity based on a 1:1 ratio of inhibitor (Mcyst) and enzyme (PP) would be 30 nmol, which is approximately 50% of the administered dose. Mcyst is a potent toxin because only hepatocytes can take up Mcyst and thus there is no dilution of the administered dose to other organs. Once inside the hepatocyte Mcyst binds covalently to the catalytic subunit of PP (Runnegar et al., 1995c; MacKintosh et al., 1995; Craig et al., 1996), effectively sequestering the enzyme-inhibitor complex until it is degraded. This phenomenon also explains the persistence of the PP inhibition over the 7 days of the experiment (Fig. 1). The toxicity that accompanies PP inhibition in liver has been well described in a number of studies. In mammals, extensive intrahepatic hemorrhage is found in lethal Mcyst intoxication,

FIG. 3. Microscopic findings in skate liver following the administration of Mcyst. (A) Control skate liver showing portal area with bile canaliculus. Pigmented melanocytes in the parenchyma are a normal finding. (B) Mild liver injury in skate liver 24 h after dosing with 125 mg/kg Mcyst; note the mild portal and periportal inflammatory changes (arrow). (C) More pronounced liver injury in a skate liver 24 h after dosing with 0.250 mg Mcyst. The portal and periportal inflammatory changes are more pronounced (arrow) and there is scattered hepatocyte necrosis. (D) Severe liver injury in a skate 48 h after dosing with 500 mg/kg Mcyst. There are widespread inflammatory changes (arrow) and extensive hepatocyte necrosis (small curved arrow). (H & E stained sections, magnification 2003).

46

RUNNEGAR ET AL.

FIG. 5. Dose response of PP activity in skate hepatocytes. (A) PP activity as % of control in skate hepatocytes 30 min after addition of increasing doses of Mcyst. PP activity of control hepatocytes was 4.39 6 0.63 units/mg of protein. (B) PP activity of skate hepatocytes incubated for 4.5 h with 100 nM to 5 mM Mcyst or 50 –500 nM Calyculin A (CalA). (C) PP activity of skate hepatocytes incubated for 1, 2, or 3 h with 1 or 5 mM Mcyst or 500 nM CalA. n 5 3– 4, means 6 SE, *p # 0.05.

resulting from the loss of sinusoidal integrity with the rounding up of hepatocytes because of toxin-induced cytoskeletal changes. In the mouse, liver weight as percentage of body weight can double and death is due to hemorrhagic shock (Runnegar et al., 1993). However, in the carp (Rabergh et al., 1991) and in the rainbow trout (Kotak et al., 1996) little hepatic hemorrhage was seen in the Mcyst-dosed fish. In the skate

localized hemorrhages were observed on visual examination of the liver, but there were no large accumulations of blood. Histologically, mild hepatic congestion was seen in a few of the skates, without any correlation with time or dose. There was no significant increase in liver weight as a percentage of body weight. Liver weight varied from 2.2 to 3.8% in the treated skates and overlapped with liver weight in control skates, which ranged from 2.3 to 4.2%. With 125 mg/kg Mcyst, the hepatic histology consisted mainly of mild portal and periportal inflammation (Fig. 3B). At higher doses mononuclear and eosinophilic infiltration of the parenchyma increased and hepatocyte loss or necrosis was seen. Hydropic changes were described in hepatocytes in the rainbow trout (Kotak et al., 1996) and hepatic megalocytosis was noted in salmon (Andersen et al., 1993), but these changes were not seen in the skate. Varying degrees of microvesicular and macrovesicular hepatic steatosis (Fig. 3) were also seen in control skates. Extensive hepatocyte necrosis, as seen with higher Mcyst doses in skate livers, has been described in all studies where close to lethal doses of Mcyst have been used and increases in hepatic enzymes in plasma have always been seen (Falconer et al., 1981; Rabergh et al., 1991). Up to 10-fold elevation in plasma lactate dehydrogenase levels were observed in a number of treated skates at 24 h when dosed with 250 and 500 mg/kg Mcyst and in some of the skates receiving the 125-mg/kg dose (results not shown). Mcyst remained in the circulation of skates and even 7 days later the PP inhibiting activity of Mcyst could still be detected in plasma (Fig. 4A). The accumulation of Mcyst in the circulation probably occurs because Mcyst can only be transported out of plasma into hepatocytes at the sinusoidal membrane. In rat hepatocytes, inhibition of PP activity changes the phosphorylation status of the cell and leads to inactivation of the transporter (Runnegar et al., 1995a). Thus, when sufficient Mcyst has entered the cell to substantially inhibit PP activity, the sinusoidal membrane transporter(s) for Mcyst are inhibited and prevent further entry of Mcyst into the hepatocyte. This explains why relatively greater percentages of the dose remain in the circulation at 24 h with increasing doses (Fig. 4B). Mcyst also inhibits bile flow in the perfused rat liver (Pace et al., 1991; Runnegar et al., 1995b). When bile flow is inhibited this potential route of Mcyst elimination is greatly decreased. The persistence of Mcyst in plasma also means that the skate is vulnerable to the toxic effects of Mcyst long after the initial exposure. In the rat, liver function is impaired at Mcyst doses that are not lethal and only partially inhibit PP activity resulting in decreased albumin production and in decreased clearance of bile acids (Solter et al., 1998). These changes and other metabolic disturbances caused by the persistence of Mcyst also make the animal more vulnerable to other potential insults. Numerous studies in mammals have shown that Mcyst has no effect on the kidney: PP activity is not inhibited and there are no primary pathological changes. However, in fish, histo-

PP INHIBITION AND MICROCYSTIN TOXICITY IN THE SKATE

47

FIG. 6. Organic solute dyes and bile acids (100 mM) prevented PP inhibition and Mcyst accumulation in skate hepatocytes when added 1 min before 10 mM Mcyst and incubated for 30 min. Values represent PP activity as % of control cells and Mcyst accumulation as % of that of cells treated with Mcyst alone. *Significant differences from PP activities with incubations with Mcyst alone; †Significantly less accumulation of Mcyst than for cells incubated with Mcyst alone. n 5 4 –5, means 6 SE; * ,†p # 0.05.

logical changes in kidney tissue have been noted in the carp and rainbow trout. In the initial study in the skate a partial decrease in kidney PP activity was observed together with local areas of tubular vacuolization that were not seen in controls. Rabergh et al. (1991) proposed that some fish are capable of transporting Mcyst into the kidney. However if this were the case for the skate, given that Mcyst remains in the circulation even at 7 days after dosing, we would expect a time-dependent increased PP inhibition in the kidney as more Mcyst is taken

FIG. 7. Addition of substrates (100 mM) to skate hepatocytes decreased the initial uptake of 10 mM Mcyst. Values are expressed as % of the uptake in cells treated with Mcyst alone, n 5 4 –5, means 6 SE *p # 0.05.

up. Figure 1 shows no such trend, making it more likely that the renal inhibition in the skate, although significant, results from some hepatic metabolic change. Although unlikely, we have not excluded the possibility that the kidney in lower vertebrates is capable of some Mcyst uptake. This issue remains to be resolved in future studies. In isolated skate hepatocytes, as in the mouse and rat, Mcyst causes a time- and dose-dependent inhibition of PP activity (Figs. 5A–5C) indicative of facilitated transport. This is in contrast to the cell permeant PP inhibitor calyculin A that showed no time dependence of PP inhibition (Fig. 5B–5C). Because even after 24 h there was no detectable effect of Mcyst in rectal gland cultures while lower doses of Calyculin A resulted in profound PP inhibition and cytoskeletal changes, we can again conclude the Mcyst uptake and therefore toxicity requires a specific transporter, which in the skate as in mammals is exclusively found in liver. Rat hepatocytes are protected from Mcyst by the addition of a number of dyes and bile acids (Runnegar et al., 1995a). These compounds inhibit the uptake of Mcyst by the hepatocytes. This protection extends to the in vivo situation since rifamycin can prevent accumulation in the liver and the toxicity of a lethal dose of Mcyst in mice over a period of 24 h (Runnegar et al., 1993). We have found that the PP activity of skate hepatocytes is also protected by preincubation with the same compounds, with decreased initial uptake of Mcyst (Fig. 7) and decreased accumulation of Mcyst (Fig. 6).

48

RUNNEGAR ET AL.

Mcyst uptake in the rat is not through the sodium-taurocholate transporting polypeptide (Ntcp) or the organic aniontransporting polypeptide (oatp1) or epoxide hydrolase, but rather through an as yet unidentified transporter (M. Runnegar, unpublished observations). The present studies with skate hepatocytes indicate that these elasmobranchs, although evolutionarily distant from the rat and mouse, already possess a sinusoidal transporter(s) that recognizes Mcyst as a substrate and is inhibited by a similar range of organic anions. These findings suggest that a yet to be identified sinusoidal organic solute transporter has evolved early in vertebrate evolution and has persisted to the present. Characterization of this transporter(s) at the molecular level will be important for understanding and preventing the hepatic toxicity of Mcyst and possibly of other compounds. ACKNOWLEDGMENTS We thank Dr. J. N. Forrest, Jr. (MDIBL and Yale University School of Medicine) for kindly providing the shark rectal gland cultures. This work was supported by National Institutes of Health Grant DK51788 (to M.R.), by a Pilot Project Award from the National Institute of Environmental Health Sciences (NIEHS)-supported Marine and Freshwater Biomedical Sciences Center at the Mount Desert Island Biological Laboratory (ES03828), and by the Imaging, and Cell Isolation, Culture and Organ Perfusion Cores of this NIEHS Center. Support for D.J.S. was provided by National Science Foundation Grant BIR9531348, Research Experiences for Undergraduates.

REFERENCES Andersen, R. J., Luu, H. A., Chen, D. Z. X., Holmes, C. F. B., Kent, M. L., Le Blank, M., Taylor, F. J. R., and Williams, D. E. (1993). Chemical and biological evidence links microcystins to salmon “netpen disease.” Toxicon 31, 1315–1323. Blumrich, M., Petzinger, E., and Boyer, J. L. (1993). Characterization of bumetanide transport in isolated skate hepatocytes. Am. J. Physiol. 265, G926 –G933. Boyer, J. L., Hagenbuch, B., Ananthanarajanan, M., Suchy, F., Stieger, B., and Meier P. J. (1993). Philogenic and ontogenic expression of hepatocellular bile acid transport. Proc. Natl. Acad. Sci. USA 90, 435– 438.

Liver pathology in mice in poisoning by the blue-green alga Microcystis aeruginosa. Aust. J. Biol. Sci. 34, 179 –187. Fricker, G., Dubost, V., Finsterwald, K., and Boyer, J. (1994). Characteristics of bile salt uptake in skate hepatocytes. Biochem. J. 29, 665– 670. Fricker, G., Hugentobler G., Meier, P. J., Kurz, G., and J. L. Boyer. (1987). Identification of a single sinusoidal bile salt uptake system in skate liver. Am. J. Physiol. (London) 253, G816 –G822. Hooser, S. B., Beasley, V. R., Lovell, R. A., Carmichael, W. W., and Haschek, W. M. (1989). Toxicity of microcystin-LR, a cyclic heptapeptide hepatotoxin from Microcystis aeruginosa to rats and mice. Vet. Pathol. 26, 246 –252. Hooser, S. B., Beasley, V. R., Waite, L. L., Kuhlenschmidt, M. S., Carmichael, W. W., and Haschek, W. M. (1991). Actin filament alterations in rat hepatocytes induced in vivo and in vitro by microcystin-LR, a hepatotoxin from Microcystis aeruginosa. Vet. Pathol. 28, 259 –266. Jochimsen, E. M., Carmichael, W. W., An, J. S., Cardo, D. M., Cookson, S. T., Holmes, C. E., Antunes, M. B., de Melo Filho, D., Lyra, T. M., Barreto, V. S., Azevedo, S. M., and Jarvis, W. R. (1998). Liver failure and death after exposure to microcystins at a hemodialysis center in Brazil. New Engl. J. Med. 338, 873– 878. Kotak, B. G., Semalulu, S., Fritz, D. L., Prepas, E. E., Hrudey, S. E., and Coppock, R. W. (1996). Hepatic and renal pathology of intraperitoneally administered microcystin-LR in rainbow trout (Oncorhyncus mykiss). Toxicon 34, 517–525. MacKintosh, C., Beattie, K. A., Klumpp, S., Cohen, P., and Codd, G. A. (1990). Cyanobacterial microcystin-LR is a potent and specific inhibitor of protein phosphatase-1 and phosphatase-2A from both mammals and higher plants. FEBS Lett. 264, 187–192. MacKintosh, R. W., Dalby, K. N., Campbell, D. G., Cohen, P. T. W., Cohen, P., and MacKintosh, C. (1995). The cyanobacterial toxin microcystin binds covalently to cysteine-273 on protein phosphatase 1. FEBS Lett. 371, 236 –240. Pace, J. G., Robinson, N. A., Miura, G. A., Matson, C. F., Geisbert, T. W., and White, J. D. (1991). Toxicity and kinetics of 3[H]Microcystin-LR in isolated perfused rat livers. Toxicol. Appl. Pharmacol. 107, 391– 401. Pouria, S., de Andrade, A., Barbosa, J., Cavalcanti, R. L., Barreto, V. T., Ward, C. J., Preiser, W., Poon, G. K., Nield, G. H., and Codd, G. A. (1998). Fatal microcystin intoxication in hemodialysis unit in Caruaru, Brazil. Lancet 352, 21–26. Rabergh, C. M. I., Bylund, G., and Eriksson, J. E. (1991). Histopathological effects of microcystin-LR, a cyclic peptide toxin from the cyanobacterium (blue-green alga) Microcystis aeruginosa on common carp (Cyprinus carpio L.). Aquat. Toxicol. 20, 131–136.

Carmichael, W. (1994). The toxins of cyanobacteria. Sci. Am. 270(1), 64 –72.

Runnegar, M. T., Kong, S., and Berndt. N. (1993). Protein phosphatase inhibition and in vivo hepatotoxicity of microcystins. Am. J. Physiol. (London). 265, G224 –G230.

Chen, D. Z. X., Boland, M. P., Smillie, M. A., Klix, H., Ptak, C., Andersen, R. J., and Holmes, C. F. B. (1993). Identification of protein phosphatase inhibitors of the microcystin class in the marine environment. Toxicon 31, 1407–1414.

Runnegar, M. T., Berndt, N., and Kaplowitz, K. (1995a). Microcystin uptake and inhibition of protein phosphatases: Effects of chemoprotectants and self-inhibition in relation to known hepatic transporters. Toxicol. Appl. Pharmacol. 134, 264 –272.

Cohen, P. S. (1989). The structure and regulation of protein phosphatases. Annu. Rev. Biochem. 58, 453–508.

Runnegar, M. T., Maddatu, T., DeLeve, L. D., Berndt, N., and Govindarajan, S. (1995b). Differential toxicity of the protein phosphatase inhibitors microcystin and calyculin A. J. Pharmacol. Exp. Ther. 273, 545–553.

Cohen, P. S., Alemany, S., Hemmings, B. A., Resink, T. J., Stralfors, P., and Tung H. Y. L. (1988). Protein phosphatase 1 and protein phosphatase 2A from rabbit skeleletal muscle. Methods Enzymol. 159, 390 – 408. Craig, M., Luu, H. A., McCready, T. L., Williams, D., Andersen, R. J., and Holmes, C. F. (1996). Molecular mechanisms underlying he interaction of motuporin and microcystins with type-1 and type-2A protein phosphatases. Biochem. Cell Biol. 74, 569 –578. Eriksson, J. E., and Goldman, R. D. (1993). Protein phosphatase inhibitors alter cytoskeletal structure and cellular morphology. Adv. Prot. Phosphatases 7, 335–357. Falconer, I. R., Jackson, A. R. B., Langley, J., and Runnegar, M. T. (1981).

Runnegar, M. T., Berndt, N., Kong, S., Lee, E. Y. C., and Zhang, L. (1995c). In vivo and in vitro binding of microcystin to protein phosphatases 1 and 2A. Biochem. Biophys. Res. Commun. 216, 162–169. Runnegar, M., Wei, X., Berndt, N., and Hamm-Alvarez, S. F. (1997). Transferrin receptor recycling in hepatocytes is regulated by protein phosphatase 2A, possibly through effects on microtubule-dependent transport. Hepatology 26, 176 –185. Silva, P., Epstein, F. H., Karnaky, K., Jr., Reichlin, S., and Forrest, J. N., Jr. (1993). Neuropeptide Y inhibits chloride secretion in the shark rectal gland. Am. J. Physiol. (London) 265, R439 –R446.

PP INHIBITION AND MICROCYSTIN TOXICITY IN THE SKATE Smith, D. J., Grossbard, M., Gordon, E. R., and Boyer, J. L. (1987). Isolation and characterization of a polarized isolated hepatocyte preparation in the skate Raja erinacea. J. Exp. Zool. 241, 291–296. Solter, P. F., Wollenberg, G. K., Chu, F. S., and Runnegar, M. T. (1998). Prolonged sublethal exposure to the protein phosphatase inhibitor microcystin-LR, results in multiple hepatotoxic effects. Toxicol. Sci. 44, 87–96. Stotts, R. R., Namikoshi, M., Haschek, W. M., Rinehart, K. L., Carmichael, W. W., Dahlem, A. M., and Beasley, V. R. (1993). Structural modifications imparting reduced toxicity in microcystins from Microcystis spp. Toxicon 31, 783–789. Valentich, J. D., and Forrest, J. N., Jr. (1991). Cl secretion by cultured shark rectal gland cells. I. Transepithelial transport. Am. J. Physiol. (London). 260, C813–C823. Wickstrom, M. L., Khan, S. A., Haschek, W. M., Wyman, J. F., Eriksson, J. E.,

49

Schaeffer, D. J., and Beasley, V. R. (1995). Alterations in microtubules, intermediate filaments, and microfilaments induced by microcystin-LR in cultured cells. Toxicol. Pathol. 23, 326 –337. Williams, D. E., Kent, M. L., Andersen, R. J., Klix, H., and Holmes, C. F. B. (1995). Tissue distribution and clearance of tritium-labeled dihydromicrocystin-LR epimers administered to Atlantic salmon via intraperitoneal injection. Toxicon 33, 125–131. Williams, D. E., Craig, M., Dawe, S. C., Kent, M. L., Holmes, C. F., and Andersen, R. J. (1997a). Evidence for a covalently bound form of microcystin-LR in salmon liver and Dungeness crab larvae. Chem. Res. Toxicol. 10, 463– 469. Williams, D. E., Dawe, S. C., Kent, M. L., Andersen, R. J., Craig, M., and Holmes, C. F. B. (1997b). Bioaccumulation and clearance of microcystins from salt water mussels, Mytilus edulis, and in vivo evidence for covalently bound microcystins in mussel tissues. Toxicon 35, 1617–1625.