NATURAL TOXINS 3~119-128 (1995)

Comparative Pathology of Microcystin-LR in Cultured Hepatocytes, Fibroblasts, and

Renal Epithelial Cells Safdar Ali Khan, Shushmita Ghosh, Mark Wickstrom, Lou Ann Miller, Rex Hess,

Wanda M. Haschek, and Val R. Beasley Departments of Veterinary Biosciences (S.A.K., S.G., M.W., L.A.M., R.H., V.R.B.) and Veterinary Pathobiology, (W.M.H.), College of Veterinary Medicine, University of Illinois, Urbana, Illinois

ABSTRACT The cyanobacterial toxin microcystin-LR (MCLR) is a potent inhibitor of protein phosphatases 1 and 2A, and is selectively toxic to the liver in vivo and to isolated hepatocytes in vitro. This selectivity is believed to be due to toxin uptake via bile acid carriers. We investigated at the light and ultrastructural levels the effects of high concentrations of MCLR and long incubation times to determine in vitro whether fibroblasts and kidney cells (non-target cells) respond in the same manner as do hepatocytes (target cells) at low concentrations and short incubation times. Cultured rat skin fibroblasts (ATCC 1213) and rat kidney epithelial cells (ATCC 1571) were incubated with MCLR at 133 pM for 1-24 hr. Lesions in these cells were compared with those in cultured hepatocytes incubated with MCLR at 13.3 pM from 1 to 32 min. Lesions in hepatocytes, kidney cells, and fibroblasts were noted at 4 min, 1 hr, and 8 hr, respectively, after initial exposure to MCLR. Lesions in all three cell types progressed and included plasma membrane blebbing, loss of cell-to-cell contact, clumping and rounding of cells, cytoplasmic vacuolization, and redistribution of cytoplasmic organelles. Loss of microvilli, whorling of rough endoplasmic reticulum, dense staining and dilated cristae in mitochondria, and pinching off of membrane blebs were noted only in hepatocytes. Nuclear changes typical of apoptosis were obsewed only in fibroblasts and kidney cells. Similarities in responses of different cell types to MCLR exposure probably reflect a common biochemical mechanism of action, i.e., inhibition of protein phosphatases 1 and 2A as described by others. The observed differences in the responses of the cell types examined in this study may reflect differences in the proteins phosphorylated and the severity of hyperphosphorylation. o 1995 wdey-~~ss, Inc.

Key Words: Blue-green algae, Pathology, Apoptosis, Cytoskeleton, Plasma membrane blebbing, Hepatotox- ins, In vitro

INTRODUCTION Microcystins are cyclic heptapeptide hepatotoxins pro-

duced by fresh water cyanobacteria, including species of Microcystis, Oscillatoria, Anabaena, and Nostoc. These tox- ins have caused outbreaks of acute liver damage and death in a range of vertebrate species [Beasley et al., 19891. The hepatic specificity of these compounds is believed to be due to their selective uptake via a bile acid carrier system present in hepatocytes [Runnegar et al., 19811. Light microscopic lesions in the livers of rats following intraperitoneal (i.p.) administration of microcystin-LR (MCLR) at 160 pg/kg included hepatocyte disassociation, rounding and degenera- tion, centrilobular necrosis, progressive disruption of he- patic sinusoids, and intrahepatic hemorrhage [Hooser et al., 19901.

Ultrastructural alterations in the livers of sheep dosed intraruminally with M. aeruginosa included aggregation and whorl formation of endoplastic reticulum, mitochondria1 swelling, and nuclear degeneration [Jackson et al., 19841.

0 1995 Wiley-Liss, Inc

Intraperitoneal administration of a sublethal dose of MCLR (10 pg/kg) to mice resulted in vesiculation of rough endo- plastic reticulum (RER), degranulation of RER, and swollen mitochondria [Dabholkar and Carmichael, 19871. In fasted rats dosed i.p. with MCLR at 200 pg/kg, hydropic, frag- mented mitochondria were observed as early as 15 min after treatment, and whorling of endoplasmic reticulum, dilation and vesiculation of smooth endoplasmic reticulum (SER) and RER, widening of the space of Disse with leakage of red blood cells into the space, loss of sinusoidal architecture, loss of desmosomes, and disrupted canaliculi were seen at 3 M O min [Miura et al., 19891. Similarly, when Hooser et al. [1990] administered MCLR to rats i.p. at 160 pg/kg,

Received March 10,1994; accepted for publication December 22,1994. Address reprint requests to Val R. Beasley, Department of Veterinary

Biosciences, College of Veterinary Medicine, University of Illinois, 2001 S. Lincoln Avenue, Urbana, IL 61801.

120 KHAN El AL.

ultrastructural lesions included invaginations of hepatocyte plasma membranes with vacuole formation, loss of microvilli along the sinusoidal face, hepatocyte separation, whorling of endoplasmic reticulum, widening of the space of Disse with leakage of red blood cells into the space, loss of sinu- soidal architecture, separation and necrosis of hepatocytes, and intact hepatocytes and hepatocellular debris in pulmo- nary capillaries, as well as hepatocellular debris in peritubu- lar capillaries of the renal cortex. Eriksson et al. I19891 reported ultrastructural findings in cultures of rat primary hepatocytes treated with MCLR at 1 pM for 20 min. Lesions included the absence of SER and Golgi bodies, numerous cytoplasmic protrusions, a band of condensed microfila- ments at the base of plasma membrane blebs, and disappear- ance of microvilli.

Morphologic changes were not noted in the following non-hepatocytes exposed to microcystins in vitro: Vero and Chinese hamster ovary (CHO) cells [Thompson et al., 19871, Swiss 3T3 mouse fibroblasts, primary mouse thymocytes and mammary alveolar cells [Falconer and Runnegar, 19871, bovine pulmonary artery endothelial cells and mouse perito- neal macrophages [Adams et al., 19881, mouse NIH-3T3 fibroblasts, human SH-SY5Y neuroblastoma cells and hu- man Hep G2 hepatoma cells [Eriksson et al., 1990a1, and hepatic non-parenchymal cells (primarily sinusoidal endo- thelial, and Kupffer cells) [Hooser et al., 19911. When incu- bated with tritiated dihydro-MCLR at 4 pM for 24 hr, NIH- 3T3 fibroblasts, human SH-SYSY neuroblastoma cells, and human Hep G2 hepatoma cells took up only 11.5, 2.5, and 11.0 pmol of the toxin derivative/mg of protein, respec- tively. By contrast, hepatocytes incubated with the labeled dihydro-MCLR at the same concentration took up 90 pmol/mg of protein in 30 min [Eriksson et al., 1990al. Microinjection of a 670 +M solution of microcystin-YR (MCYR) caused human skin fibroblasts to change from the normal spindle-like to a rounded form within 45 min after treatment [Matsushima et al., 19901. It appears, therefore, that the limited uptake of microcystins by non-hepatocytes results in their relatively high tolerance to exposure to these compounds.

Recently, microcystins have been shown to be potent inhibitors of protein phosphatases 1 and 2A [Honkanen et al., 1990; MacKintoshet al., 1990; Yoshizawaet al., 1990; Eriksson et al., 1990b; Falconer and Yeung, 19921. Inhibi- tion of protein phosphatases is believed to result in increased phosphorylation of several cytosolic and cytoskeletal pro- teins resulting in disruption of normal cytoskeletal architec- ture, cell rounding, bleb formation, loss of cell-to-cell adhe- sions, and necrosis. Recently, Ohta et al. 119921 identified certain hyperphosphorylated proteins in hepatocytes treated with MCLR as being cytokeratins 8 and 18. Like other inhibi- tors of protein phosphatases (okadaic acid, calyculin A, and dinophysistoxin [Sassa et a]., 1989; Suganuma et al., 1988, 1989, 1990, 19921), MCLR has been shown to be a potent liver tumor promoter [Nishiwaki-Matsushima et al., 19921.

In this study, we exposed fibroblasts and renal epithelial cells to MCLR at high concentrations for comparatively long periods of time and examined them using light and electron microscopy to assess whether such exposure would result in the development of lesions. We compared these responses to lesion development in hepatocytes briefly ex- posed to the toxin at much lower concentrations.

MATERIALS AND METHODS Toxin Purification

MCLR was isolated from lyophilized algal bloom mate- rial and purified to >95% purity as determined by high- performance liquid chromatography (HPLC) and fast atom bombardment mass spectrometry. Isolation and purification were performed using the method of Harada et al. [ 19881 as adapted by Stotts et al. (submitted for publication).

Verification of Biological Activity In Vio Biological activity of the purified MCLR was determined

by administration of the toxin i.p. at 100 pg/kg to 3 mice. Two control mice were given an equal volume of saline i.p. The toxin-treated mice rapidly became moribund and died within 2 hr. Livers were enlarged, engorged with blood, and friable. On histologic examination, the livers of the toxin- treated animals exhibited massive hemorrhage and centri- lobular hepatocellular necrosis. These lesions are consistent with MCLR toxicosis in the mouse. Livers of control mice were normal on histologic examination.

Cell Culture and Toxin Exposure Primary rat hepatocytes were isolated using a modifica-

tion of the two-step collagenase perfusion method described by Seglan [ 19761. Briefly, 200-250 g male Sprague-Dawley rats (Harlan Sprague-Dawley, Indianapolis, IN) were anes- thetized with pentobarbital and the portal vein was catheter- ized. The liver was perfused with the calcium-free buffer following by the same buffer with added CaCl, and collagen- ase. The hepatocytes were isolated, filtered, and centri- fuged, and viability was determined by trypan blue dye exclusion. Hepatocytes were grown in 24-well plates on Matrigelm (Collaborative Biomedical, Bedford, MA)-coated plastic coverslips (Thermonoxm, Electron Microscopy Sci- ences, Fort Washington, PA) in Ham’s F-10 medium sup- plemented with fetal bovine serum (16.7%), insulin (100 unitdml), and antibiotics (1 .O ml gentamicin [50 mg/ml, Sigma, St. Louis, MO], 6.0 ml Pen-Strep [ lO ,O00 units penicillin/ml, 10 mg streptomyciniml, Sigma]). Approxi- mately 30,000 hepatocytes were transferred to each well. Hepatocytes were used within 24 hr of isolation.

Rat skin fibroblasts (ATCC 1213) or rat kidney epithelial cells (ATCC 1571) were similarly grown in 24-well plates on MatrigeWcoated plastic coverslips. Fifty thousand to 1 X lo5 rat skin fibroblasts or rat kidney epithelial cells were seeded in each well. Fibroblasts were grown in Dul- becco’s modified Eagle’s medium supplemented with 10%

COMPARATIVE PATHOLOGY OF MICROCYSTIN-LR 121

fetal bovine serum and antibiotics as described above. Kid- ney cells were grown in the same medium supplemented with 5% fetal bovine serum. The cells were allowed to grow for 3-5 days until they reached confluence. The medium was changed every 3 days.

In pilot studies, we examined the effects on fibroblasts and kidney cells of MCLR at concentrations of 40, 100, and 200 pM following exposures ranging from 24 to 72 hr. Based on the absence of light microscopic changes at the 40 pM concentration after 72 hr, the observation of only mini- mal changes at 100 pM, and severe changes at 200 pM after 24 hr of exposure, we chose a dose of I33 pM for our formal trials with these cell types.

In formal studies, hepatocyte cultures were incubated for 0, 1, 2, 4, 8, 16, and 32 min in culture medium with the addition of MCLR in phosphate buffered saline (PBS; pH 7.4) such that the final toxin concentration was 13.3 pM. Control wells received an equal volume of PBS and were incubated for the same time periods. Kidney cells and fibro- blasts were incubated with MCLR at 133 pM for 0, 1, 2,4 , 8, 16, or 24 hr. After treatment, the medium was aspirated and the cells were washed twice with PBS. The cells were then fixed for light or electron microscopy with 10% neutral buffered formalin or 2.6% glutaraldehyde in 0.1 M cacody- late buffer (pH 7.25), respectively. Each study was done twice. At least 18 cells of each cell type were examined at each time point.

Light Microscopy The coverslips were washed and stained with hematoxy-

lin and eosin, dehydrated in 95% ethanol, 100% ethanol twice, and 100% xylene. The coverslips were mounted on glass slides and examined using a brightfield microscope. To estimate the number of cells attached and floating in the medium, cells were examined under an inverted microscope 30 sec before treatment -and at the end of each observation time.

Transmission Electron Microscopy (TEM) The cells for TEM were prepared using a modification of

the method described by Miller [1982]. The fixative was removed and coverslips were rinsed twice with 1:l cacody- late buffer in double distilled water. The coverslips were placed in 2: 1 : 1 osmium tetroxide/KCN/cacodylate buffer for 15-20 min. After rinsing with cacodylate buffer again, the cells were dehydrated in a graded ethanol series of 25- 100%. Final dehydration was followed by infiltration with 100% propylene oxide and epoxy. Each plastic coverslip with attached cells was covered with a second coverslip. The epoxy resin was allowed to polymerize at 90°C for 1 hr. Both coverslips were removed and the resulting cell-epoxy membrane was cut into small pieces and put into a beam capsule filled with 100% epoxy which was then allowed to polymerize at 90°C for 5 hr. Thin sections were cut and

Fig. 1. Cultured hepatocytes. Hernatoxylin and eosin. A: Control. Polyhe- dral cells with prominent nuclei are well spread out and present in large groups which have extensive cell-to-cell contacts (arrows). Bar = 10 pm. 6: After 32 min of incubation with MCLR at 13.3 kM. Most cells have multiple, variable sized blebs on the plasma membrane surface. Cells have lost normal cell-to-cell contact (arrows) and appear rounded and clumped together. Bar = 10 pm.

stained with uranyl acetate and lead citrate and examined using TEM on a JEOL l00CX microscope.

RESULTS Hepatocytes

By light microscopy, the attached hepatocytes in control cultures were polyhedral with prominent, centrally located spherical nuclei (Fig. 1A). These cells were well spread out in large, subconfluent groups. Some binucleated cells were present. After 8 min of exposure to MCLR at 13.3 pM, a few attached hepatocytes demonstrated plasma membrane blebbing, however, most of the cells maintained their nor- mal morphology. At 16 and 32 min, severe changes were seen. Most cells had multiple large plasma membrane blebs (Fig. IB). Loss of normal cell-to-cell attachments between hepatocytes often resulted in formation of cell clusters. Cells became rounded and appeared smaller. Most of the cells became detached from the extracellular matrix.

TEM of control cultured hepatocytes showed a large, centrally located nucleus and numerous organelles in the

122 KHAN El AL.

2 A

Fig. 2A

COMPARATIVE PATHOLOGY OF MICROCYSTIN-LR 123

2 D

Fig. 2. Ultrastructure of cultured hepatocytes. A Control. A prominent centrally located nucleus (n), numerous mitochondria (m), peroxisomes (p), RER (I), SER (s), and glycogen vesicles (v) are visible. Microvilli are present on both surfaces. Bar = 1 pm. B: After 4 min of incubation with MCLR at 13.3 pM MCLR. Note the presence of a plasma membrane bleb (anow). Bar = 0.5 pm. C: After 16 min of incubation with MCLRat 13.3 pM. Densely stained mitochondria with dilated cristae (large arrows). One concentric myelin body is present in the RER (small arrow). Bar = 0.5 pm.

cytoplasm (Fig. 2A). Granular RER, SER, numerous mito- chondria, peroxisomes, and lipid and glycogen vesicles were observed in the cytoplasm. Microvilli, highly variable in size and numbers, were present on the cell surface. Differ- ences from controls were detected in MCLR-treated cells at 4 and 8 min after toxin introduction when plasma membrane blebbing and membrane-bound vacuoles, respectively, were observed (Fig. 2B). At 16 min, vacuolization and plasma membrane blebbing were more prominent. Densely stained mitochondria had dilated cristae (Fig. 2C). A few myelin bodies were seen in the cytoplasm of treated cells. Similar but more severe changes were observed at 32 min. At that time, membrane-bound plasma membrane blebs were ob- served to pinch off from the cell surface. These pinched off blebs contained redistributed and compacted cytoplasmic organelles. Some of the affected cells disintegrated into

D, E: After 32 min of incubation with MCLR at 133 p,M D A plasma membrane bleb with closely compacted mitochondria is apparent as indicated by the arrow A second bleb (left) is disintegrating, and the adjacent cytoplasm has no mitochondria present Some cytoplasmic vac- uoles are also present Bar = 1 pm E Whorling of RER at the periphery of a plasma membrane bleb and redistribution of organelles are present Bar = 1 pm

small membrane-bound fragments (Fig. 2D). RER formed whorls around mitochondria or at the periphery of plasma membrane blebs (Fig. 2E). Microvilli were absent from most of the cells.

Kidney Cells On light microscopy, control kidney cells were irregu-

larly shaped and formed a loose monolayer (Fig. 3A). Fine plasma membrane extensions joined neighboring kidney cells. Each cell had a prominent centrally located nucleus. Occasional cells were binucleated. After 8 hr of treatment with MCLR, there was an increase in the number of de- tached cells. At 16 and 24 hr, areas of cell loss were evident in the monolayer (Fig. 3B). The cells adjacent to areas of cell loss became spindle-shaped. Although plasma mem- brane blebs were evident in the detached cells (not shown),

124 KHAN E l AL.

Fig. 3. Cultured kidney cell monolayers at 24 hr of incubation. Hematoxy- lin and eosin. A: Control. Irregularly shaped cells have prominent nuclei and fine plasma membrane extensions between neighboring cells. Bar = 10 pm. B: Incubated with MCLR at 133 pM. Areas of cell loss are observed in the cell monolayer (arrows). Bar = 10 pm.

such blebs were not observed in the attached kidney cells on light microscopy.

Examination of the ultrastructure of control kidney cells on TEM revealed large, centrally located nuclei with well- delineated nuclear membranes and, in the cytoplasm, nor- mal appearing organelles, such as mitochondria, SER, RER, and Golgi bodies, were observed (Fig. 4A). In MCLR- treated cells, at 1 hr, mild cytoplasmic vacuolization was observed in most of the cells. Some cells exhibited mild blebbing, disrupted plasma membranes, hyperconvoluted nuclear membranes, and condensed or fragmented chro- matin material. Similar changes as well as moderate num- bers of membrane-bound vacuoles and blebs were seen at 2 hr (Fig. 4B). At 4 and 8 hr, these changes were more severe. At 16 hr, lysosomes were prominent and appeared to be increased in number. Some lysosomes were located at the periphery of blebs. Several myelin bodies were also visible in the cytoplasm (Fig. 4C). By 24 hr, plasma membrane blebbing and nuclear fragmentation were severe, and lyso- somes were prominent (Fig. 4D).

Fibroblasts On light microscopy, control fibroblasts were generally

spindle-shaped with a large centrally located spherical nu- cleus surrounded by a small amount of cytoplasm. In the attached cells, no significant changes were observed in re- sponse to MCLR treatment. After 16 and 24 hr of toxin exposure, cells floating in the medium were rounded and had plasma membrane blebs (not shown).

Unlike control fibroblasts (Fig. 5A), on ultrastructural examination, MCLR-treated cells showed morphologic changes after 8 hr of treatment, consisting of severe plasma membrane blebbing, detachment of blebs, hyperconvoluted nuclear membranes, condensation of chromatin, and cyto- plasmic vacuolization (Fig. 5B). Similar changes were ob- served at 16 and 24 hr. At 24 hr, condensation of chromatin and nuclear fragmentation were severe (Fig. 5C).

DISCUSSION Many of the morphologic changes observed in kidney

cells and fibroblasts following exposure to a high concentra- tion of MCLR for prolonged periods of time were quite similar to the changes in hepatocytes exposed briefly to a much lower toxin concentration. Despite the prolongation of exposure times from minutes with hepatocytes to several hours with non-hepatocytes, the concentration required to induce similar morphologic changes in the latter was 10-fold of that necessary for the former. This finding is consistent with the report of Eriksson et al. [ 1990a], who showed that uptake of tritiated dihydromicrocystin by 3T3 fibroblasts, SH-SYSY neuroblastoma, and Hep G2 cells after 24 hr was roughly one-tenth of the uptake by primary rat hepatocytes after 30 min. In our study, ultrastructural changes observed following MCLR exposure of kidney cells and fibroblasts occurred much later than with hepatocytes. In kidney cells, changes were not seen until after 1 hr of toxin exposure, and in fibroblasts, changes were first evident at 8 hr. This differ- ence in response to MCLR exposure by different cell types is most likely due to differences in the ability to internalize the toxin. Cells in organs other than the liver seem to lack the bile acid transport system which enables rapid uptake of algal peptide hepatotoxins with the possible exception of renal tubular epithelial cells [Frimmer and Ziegler, 19881.

In a previous study in our laboratory, the addition of MCLR at 10 FM to freshly isolated hepatocytes in suspen- sion resulted in blebbing in most cells within 30 min of the onset of exposure. However, addition of the same toxin at 10 FM to 72 hr cultured hepatocytes did not cause blebbing in most of the cells until 6 hr post-dosing [Hooser et al., 19911. In our studies with less than 24-hr-old hepatocyte cultures exposed to MCLR at 13.3 pM, blebbing was ob- served as early as 8 min after initial toxin exposure when there was also an increase in the number of detached cells, and by 16 min, most of the cells had numerous blebs and were detached from the matrix. The difference between our findings and those of Hooser et al. [1991] is probably a

COMPARATIVE PATHOLOGY OF MICROCYSTIN-LR 125

4 A

Fig. 4. Ultrastructure of cultured kidney cells. A: Control. A large nucleus (n) with well-delineated membrane (nm), RER (I), and a few mitochondria (m) are present. Bar = 1 pm. B: After 2 hr of incubation with MCLR at 133 pM. Plasma membrane blebbing (large arrows) and cytoplasmic vacu- olization (small arrows) are present. Note the condensation of chromatin. Bar = 1 pm. C: After 16 hr of incubation with MCLR at 133 pM. Cytoplasmic

function of the reduction in hepatocyte bile acid carriers available for the uptake of microcystins [Runnegar et al., 19811. A reduction in bile acid carriers has been shown to occur over time after hepatocytes are isolated from the liver [Petzinger and Frimmer, 19881.

We observed plasma membrane blebbing, loss of cell-to- cell contacts, detachment from the extracellular matrix, rounding of cells with cell clumping, and cytoplasmic vacu- olization in all cell types examined. Our findings with hepa- tocytes are similar to those reported by Thompson et al. [ 19871 following exposure of 24-hr-old cultured rat hepato- cytes for 5 min to MCLR at 10 pM. Pinching off of mem- brane-bound blebs and fragmentation may be similar to the situation observed in vivo in rats given a lethal dose of MCLR. Both intact hepatocytes and hepatocellular debris released from the liver were seen in the pulmonary vascula- ture as early as 1 hr post-dosing, and hepatocellular debris was observed in peritubular capillaries of the renal cortex at

vacuolization, myelin bodies (large arrows), and prominent lysosomes are present (small anows). Bar = 0.5 pm. D: After 24 hr of incubation with MCLR at 133 pM. A large plasma membrane bleb (small arrow), cytoplas- mic vacuolization, nuclear fragmentation (large arrow), and a prominent lysosome ( I ) is present. Bar = 1 pm.

9 hr [Hooser et al., 19901. Bleb formation in hepatocytes seen after MCLR exposure is believed to be related in part to reorganization of actin filaments [Eriksson et al., 1989; Hoo- ser et al., 19911. MCLR, in addition to causing effects on actin filaments, has also been shown to result in the collapse of intermediate filaments [Falconer and Yeung, 19921, and microtubules in hepatocytes and non-hepatocytes [Wick- stromet al., 1993, 19941.

Recent work by Runnegar et al. (19931 has shown that, although protein phosphatases from rat kidney cytosol are inhibited by the addition of microcystin-YM (MCYM) in vitro, a lethal i.p. injection of MCYM to mice did not cause inhibition of protein phosphatases in the kidneys in vivo. This seems to suggest that the kidney may be unlikely to accumulate nephrotoxic concentrations of the microcystins in vivo. Although kidney lesions have been documented in some studies involving lethal dosing of animals with micro- cystins [Hooser et al., 1990; Love11 et al., 19891, these may

126 KHAN ET AL.

5c t

Fig. 5. Ultrastructure of cultured fibroblasts. A: Control. A nucleus (n), RER (r), and a few mitochondria (m) are present. Bar = 1 pm. B: After 8 hr of incubation with MCLR at 133 wM. Multiple plasma membrane blebs (ar- row), hyperconvoluted nuclear membrane, condensed Chromatin, and a few cellular fragments are present. Bar = 1 pm. C: After 24 hr of incubation with MCLR at 133 wM. Multiple blebs (large arrow) and nuclear fragmenta- tion (small arrow) are present. Bar = 1 pin.

be mainly a result of circulatory shock [Theiss et al., 19881 which is potentially aggravated by hepatocellular debris in the capillaries [Hooser et al., 19901. In this study, cultured

renal epithelial cells were more sensitive to MCLR than fibroblasts, as reflected in the more rapid onset in the former of plasma membrane blebs, hyperconvoluted nuclear mem- branes, chromatin condensation, nuclear fragmentation, and cytoplasmic vacuolization.

Cytoplasmic vacuolization was evident in MCLR-treated hepatocytes as well as in fibroblasts and kidney cells at various time points after initial exposure to the toxin. Cyto- plasmic vacuolization and formation of membrane-bound vacuoles in cells are non-specific responses to injury [Alden and Frith, 19911. The effects of MCLR on hepatocyte mito- chondria observed in our study may be secondary to the severe metabolic stress associated with the acute toxicosis experience by these cells, since incubation of MCLR with isolated hepatocyte mitochondria did not cause functional alteration [Miura et al., 1989; Eriksson et al., 19891.

The formation of “apoptotic bodies” has been reported following treatment of both cultured hepatocytes and non- hepatocytes with okadaic acid (0.5 pM for 3 hr) and cul- tured hepatocytes with MCLR (0.8 p M for 3 hr) [Boe et al., 19911. These “apoptotic bodies” were characterized in part by a hyperconvoluted nuclear membrane, condensed chro- matin, and nuclear fragmentation, which are classical fea- tures of apoptosis [Wyllie et al., 19801. In contrast to the findings of Boe et al. [1991], in hepatocytes treated with MCLR at 13.3 pM for up to 32 min, we did not see evidence of hyperconvolution of the nuclear membrane, nuclear frag- mentation, or chromatin condensation. Although nuclear changes compatible with apoptosis were evident in MCLR- treated kidney cells (Fig. 4D) and fibroblasts (Fig. 5C) be- ginning at 1 and 8 hr, respectively, in most cells, these changes were not severe. In our study, the absence of nu- clear changes in hepatocytes in response to MCLR exposure was probably a consequence of the short exposure times used. Another difference between our study and that of Boe et al. [I9911 is that, on TEM, we examined only those treated cells that remained attached to the coverslips, i.e., the less severely affected cells, whereas Boe et al. [1991] examined primarily the detached, i.e., the more severely damaged cells.

Plasma membrane blebbing is a classical feature of apop- tosis [Wyllie et al., 19801. Our results and previous studies [Runnegar et al., 1981; Hooser et al., 19911 indicate that plasma membrane blebbing may be the earliest detectable change seen on light microscopy in MCLR-treated cells. Other changes noted in hepatocytes, fibroblasts, and kidney epithelial cells in this study, which are compatible with apoptosis, included rounding of cells, vacuolation, detach- ment from the substrate, absence of microvilli, redistribu- tion and compacting of cytoplasmic organelles, pinching off of membrane-bound apoptotic bodies which frequently con- tained intact organelles, hyperconvolution of the nuclear membrane, and myelin body formation. It is considered likely that severe nuclear fragmentation and chromatin con- densation would follow the above-mentioned changes with

COMPARATIVE PATHOLOGY OF MICROCYSTIN-LR 127

sufficiently prolonged exposure and uptake of MCLR. Stud- ies, including quantification of DNA fragmentation, assess- ment of the activity of transglutaminases which are involved in crosslinking of cytoskeletal proteins and formation of the apoptotic envelope, and assessment of phosphorylation of proteins associated with apoptosis [Boe et al., 19911 would be of value in an effort to confirm the occurrence of apopto- sis in MCLR-treated cells. Once internalized, MCLR seems to act through the inhibi- tion of phosphatases 1 and 2A in a variety of cell types. Support for a common intracellular mechanism is provided by our earlier findings that MCLR induces similar cytoskel- eta1 changes in both hepatocytes and non-hepatocytes, and the final stages of cytoskeletal collapse are similar in the cell types examined to date [Wickstrom et al., 19931. There were, however, also differences between cell types in re- sponse to MCLR. These included the dilation of mitochon- drial cristae, which occurred only in hepatocytes; the more severe redistribution of organelles in hepatocytes; and the development of myelin bodies in kidney cells and hepato- cytes, but not fibroblasts. Also, we have recently observed differences between hepatocytes and non-hepatocytes with regard to the sequence of cytoskeletal changes in MCLR- treated cells in vitro [Wickstrom et al., 19951. Microtubule changes have been documented as a feature of some forms of apoptosis [Tsukidate et al., 19931, and we have noted high sensitivity of hepatocyte microtubules in MCLR- exposed cells [Wickstrom et al., 1993, 19951. The differ- ences between hepatocytes and non-hepatocytes observed on light and electron microscopy and in the sequence of cytoskeletal changes are likely to represent cell-specific dif- ferences in protein phosphorylation in response to MCLR.

ACKNOWLEDGMENTS This work was supported in part by grant ES 05552-01

from the National Institute of Environmental Health Sci- ences. The authors thank Bill Hollis and Mary Monhaut for their assistance in the cell culture and electron microscopy studies.

REFERENCES Adams WH, Stone JP, Sylvester B, Stoner RD, Slatkin DN, Tempe1 NR,

Siegelman HW ( 1988): Pathophysiology of cyanoginosin-LR: In vivo and in vitro studies. Toxicol Appl Pharmacol96:248-257.

Alden CA, Frith CH (1991): Urinary system. In Haschek WM, Rousseaux CG (eds): “Handbook of Toxicologic Pathology.” New York: Academic Press, p 373.

Beasley VR, Cook WO, Dahlem AM, Hooser SB, Lovell RA, Valentine WM (1989): Algae intoxication in livestock and waterfowl. Vet Clin N Am 5:345-361.

Boe R, Gjertsen BT, Vintermyr OK, Houge G, Lanotte M, Doskeland SO (1991): The protein phosphatase inhibitor okadaic acid induces morpho- logical changes typical of apoptosis in mammalian cells. Exp Cell Res

Dabholkar AS, Carmichael WW (1987): Ultrastructural changes in the mouse liver induced by hepatotoxin from the fresh water Microcystis aeruginosa strain 7820. Toxicon 25:285-292.

195:237-246.

Eriksson JE, Paatero GIL, Meriluoto JAO, Codd GA, Kass GEN, Nicotera P, Orrenius S (1989): Rapid microfilament reorganization induced in isolated rat hepatocytes by microcystin-LR, a cyclic peptide toxin. Exp Cell Res 185:8&100.

Eriksson JE, Gronberg L, Nygard S, Slotte JP, Meriluoto JOA (1990a): Hepatocellular uptake of ‘H-dihydromicrocystin-LR, a cyclic peptide toxin. Biochim Biophys Acta 1025:60-66.

Eriksson JE, Toivola D, Meriluoto JAO, Karaki H, Han Y-G, Hartshorne D ( 1990b): Hepatocyte deformation induced by cyanobacterial toxins reflects inhibition of protein phosphatases. Biochem Biophys Res Com- mun 173: 1347-1353.

Falconer IR, Runnegar MTC (1987): Effects of the peptide toxin from Microcystis aeruginosa on intracellular calcium, pH and membrane in- tegrity in mammalian cells. Chem Biol Interact 63:215-225.

Falconer IR, Yeung DSK (1992): Cytoskeletal changes in hepatocytes induced by Micocystis toxins and their relation to hyperphosphorylation of cell proteins. Chem Biol Interact 81:181-196.

Frimmer M, Ziegler K (1988): The transport of bile in liver cells. Biochim Biophys Acta 947:75-99.

Harada K-I, Suzuki M, Dahlem AM, Beasley VR, Carmichael WW, Rine- hart KL Jr (1988): lmproved method of purification of toxic peptides produced by cyanobacteria. Toxicon 26:433439.

Honkanen RE, Zwiller J , Moore RE, Daily SL, Khatra BS, Dukelow M, Boynton AL (1990): Characterization of microcystin-LR, a potent inhib- itor of type 1 and type 2A protein phosphatases. J Biol Chem 265: 19401- 1 9404.

Hooser SB, Beasley VR, Basgall EJ, Carmichael WW, Haschek WM ( 1990): Microcystin-LR induced ultrastructural changes in rats. Vet Pathol 27:9-15.

Hooser SB, Beasley VR, Waite LL, Kuhlenschmidt MS, Carmichael WW, Haschek WM (1991): Actin filament alterations in rat hepatocytes in- duced in vivo and in vitro by microcystin-LR, a hepatotoxin from the blue-gree alga, Microcystis aeruginosa. Vet Pathol28:259-266.

Jackson ARB, McInnes A, Falconer IR, Runnegar MTC (1984): Clinical and pathological changes in sheep experimentally poisoned by the blue- gree alga Microcystis aeruginosa. Vet Path012 1 : 102-1 13,

Lovell RA, Schaeffer DJ, Hooser SB, Haschek WM, Dahlem AM, Car- michael WW, Beasley VR (1989): Toxicity of intraperitoneal doses of microcystin-LR in two strains of male mice. J Environ Pathol Toxicol Oncol9:221-238.

MacKintosh C, Beattie KA, Klumpp S, Cohen P, Codd GA (1990): Cyano- bacterial microcystin-LR is a potent and specific inhibitor of protein phosphatases 1 and 2A from both mammals and higher plants. FEBS

Matsushima R, Yoshizawa S, Watanabe MF, Harada K, Furusawa M, Carmichael WW, Fujiki H (1990): In vitro and in vivo effects of protein phosphatase inhibitors, microcystins and nodularin, on mouse skin and fibroblasts. Biochem Biophys Res Commun 171:867-874.

Miller LA (1982): Practical rapid embedding procedure for transmission electron microscopy. Lab Med 12:752-756.

Miura GA, Robinson NA, Geisbert KA, Bostian JDW, Pace JG (1989): Comparison of in vivo and in v i m toxic effects of microcystin-LR in fasted rats. Toxicon 27:1229-1240.

Nishiwaki-Matsushima R, Ohta T, Nishiwaki S, Suganuma M, Kohyama K, lshikawa T, Carmichael WW, Fujiki H (1992): Liver tumor promo- tion by the cyanobacterial cyclic peptide toxin microcystin-LR. J Cancer Res Clin Oncol 118:42C-424.

Ohta T , Nishiwaki R, Yatsunami J, Komori A, Suganuma M, Fujiki H (1992): Hyperphosphorylation of cytokeratin 8 and 18 by microcystin- LR, a new liver tumor promoter, in primary cultured rat hepatocytes. Carcinogenesis 13:2443-2447.

Petzinger E, Frimmer M (1988): Comparative investigations on the uptake of phallotoxins, bile acids, bovine lacteroperoxidase, and horseradish peroxidase into rat hepatocytes in suspension and in cell cultures. Bio- chim Biophys Acta 937:135-144.

264: 187- 192.

128 KHAN ET AL.

Runnegar MT, Falconer IR, Silver J (1981): Deformation of isolated rat hepatocytes by a peptide hepatotoxin from the blue-green alga Microcys- tis aeruginosa. Naunyn-Schmiedeberg’s Arch Pharmacol3 17:268-272.

Runnegar MT, Kong S, Berndt N (1993): Protein phosphatase inhibition and in vivo hepatotoxicity of microcystins. Am J Physiol 265:G224- G230.

Sassa T, Richter WW, Uda N, Suganuma M, Suguri H, Yoshizawa S, Hirota M, Fujiki H (1989): Apparent “activation” of protein kindses by okadaic acid class tumor promoters. Biochem Biophys Res Commun 159:939-944.

Seglen PO (1976): Preparation of isolated rat liver cells. In Prescott D (ed): “Methods in Cell Biology.” New York: Academic Press, vol XIII, pp 29-81.

Suganuma M, Fujiki H, Suguri H, Yoshizawa S, Horita M, Nakayasu M, Ojika M, Wakamatsu K, Yamada K (1988): Okadaic acid: An additional non-phorbol-l2-tetradecanoate-l3-acetate-type tumor promoter. Proc Natl Acad Sci USA 85:1768-1771.

Suganuma M, Suttajit M, Suguri H, Ojika M, Yamada K, Fujiki H (1989): Specific binding of okadaic acid, a new tumor promoter in mouse skin. FEBS Lett 250:615-618.

Suganuma M, Fujiki H, Suguri H, Yoshizawa S, Yasumoto S, Kato Y, Fusetani N, Sugimura T (1990): Calyculin A, an inhibitor of protein phosphatases, a potent tumor promoter on CD- 1 mouse skin. Cancer Res 50:352 1-3525.

Suganuma M, Fujiki H, Okabe S, Nishiwaki S, Brautigan D, Ingebritsen TS, Rosner M (1992): Structurally different members of the okadaic acid

class selectively inhibit protein serinelthreonine but not tyrosine phos- phatase activity. Toxicon 30:873-878.

Theiss WC, Carmichael WW, Wyman J, Brunner R (1988): Blood pressure and hepatocellular effects of the cyclic heptapeptide toxin produced by the freshwater cyanobacterium (blue-green alga) Microcystis aeruginosa strain PCC-7820. Toxicon 26:603413.

Thompson WL, Allen MB, Bostian KA (1987): The effects of microcystin on monolayers of primary rat hepatocytes. In Gopalakrisnakone P, Tan CK (eds): “Progress in Venom and Toxin Research.” Singapore: Na- tional University of Singapore, pp 725-73 1.

Tsukidate K, Yamamoto K, Snyder JW, Farber JL (1993): Microtubule antagonists activate programmed cell death (apoptosis) in cultured rat hepatocytes. Am J Pathol 143:918-925.

Wickstrom M, Khan SA, Haschek W, Beasley V (1993): Cytoskeletal changes in microcystin-LR exposed fibroblasts and kidney cells. Toxi- cologist 13: 1793.

Wickstrom ML, Khan SA, Haschek-Hock WM, Wyman JF, Eriksson JE, Schaeffer DJ, Beasley VR (1995): Alterations in microtubules, interme- diate filaments, and microfilaments induced by microcystin-LR in cul- tured cells. Toxicol Pathol (in press).

Wyllie AH, Ken JFR, Cume AR (1980): Cell death: The significance of apoptosis. Int Rev Cytol 68:251-306.

Yoshizawa S, Matsushima R, Watanabe MF, Harada K, Ichihara A, Car- michael WW, Fujiki H (1990): Inhibition of protein phosphatases by microcystis and nodularin associated with hepatotoxicity. J Cancer Res Clin Oncol I16:609414.