Effects of aflatoxin B1 in a liver cell line from rainbow trout (Oncorhynchus mykiss)

~

Toxic'. in Vitro Vol. 8, No. 3, pp. 317-328, 1994

Pergamon

0887-2333(94)E0006-F

Copyright © 1994 ElsevierScienceLtd Printed in Great Britain. All rights reserved 0887-2333/94 $7.00 + 0.00

EFFECTS OF AFLATOXIN B~ IN A LIVER CELL LINE FROM RAINBOW TROUT (ONCORHYNCHUS MYKISS) D. G. BECHTEL and L. E. J. LEE* Department of Veterinary Anatomy, Western College of Veterinary Medicine, University of Saskatchewan, Saskatoon, SK, Canada S7N 0W0 (Received 30 September 1993; revisions received 18 November 1993)

Abstract--The cytotoxic response to ariatoxin B 1 (AFB) was investigated in RTL-W1, a cell line derived from the normal liver of a mature rainbow trout (Oncorhynchus mykiss). AFB altered RTL-Wl morphology and ultrastructure and inhibited DNA synthesis. The effective concentration required for 50% inhibition (ECs0) of DNA synthesis, after 2 days of treatment was 0.04 #g/ml. This response was compared with that of two other salmonid, but non-liver, cell lines [rainbow trout gonad (RTG-2) and Chinook salmon embryo (CHSE-214)]. Although RTG-2 cells were as sensitive as RTL-W1 cells (ECs0 for inhibition of DNA synthesis was 0.05 pg/ml), CHSE-214 cells were unresponsive to AFB at concentrations as high as 2 # g/ml. After a single AFB exposure, RTL-W1 sublines were isolated that had phenotypic changes typical of malignant transformation. These were increased growth rate, reduced contact inhibition of growth, altered cellular morphology and growth in soft agar. In addition, RTL-Wl metabolized AFB: the major metabolites, aflatoxin (AFL) and aflatoxin MI (AFM), were detected by thin-layer chromatography and HPLC. The relative amounts of these metabolites, unlike those observed with RTG-2 cells, were in close agreement with those produced by trout liver in vivo. Thus, RTL-W 1 could provide a sensitive in vitro model system for studying the action of biotransformation requiring xenobiotics. Overall, the observed responses were similar to those reported for liver ceils in AFB-exposed trout, suggesting that RTL-Wl cells are suitable for studying cytotoxic effects and malignant transformation in vitro.

INTRODUCTION Cell cultures provide useful means for studying cytotoxicity without involvement of the complex control processes of the whole organism (Frazier, 1992; Paganuzzi-Stammati et al., 1981; Watson, 1992). Sentinel cells that could be readily and consistently used for monitoring environmental toxicants, and that maintain xenobiotic-metabolizing activity have been sought for their applicability in toxicology (Glatt et al., 1987). Usually, mammalian cell lines have been used for monitoring the effects that toxicants might have in those species and ultimately these are correlated with effects in humans. In contrast, relatively few efforts have been aimed at investigating the effect toxicants that are particularly abundant in aquatic environments have in the lower vertebrates such as fish, and few fish cell lines have *To whom correspondence should be addressed. Abbreviations: AFB=aflatoxin B6 AFB2=aflatoxin B2;

AFB2~ = ariatoxin B2~; AFL = ariatoxicol; AFL-M = aflatoxicol M 6 AFM = aflatoxin M~; AFP = ariatoxin P6 AFQ = aflatoxin Q6 B[a]P = benzo[a]pyrene; CHSE-214 = Chinook salmon embryo ceils; DMSO = dimethyl sulfoxide; ECs0= effective concentration required for 50% inhibition; FBS = foetal bovine serum; SPE = solid phase extraction; TEM = transmission electron microscopy; TLC = thin-layer chromatography; RTG-2 = rainbow trout gonad cell line; RTLW1 = rainbow trout liver-Waterloo 1 cell line.

been developed or used for ecotoxicity testing in these organisms (Babich and Borenfreund, 1987 and 1991). Established cell lines are advantageous for routine use in toxicology because of their immortality and phenotype stability, but they have limited applications because of their poor capacity to metabolize chemicals. Much effort has been spent in developing culture systems with metabolically competent cells such as hepatocytes (Kremers et al., 1990). Most toxicology studies performed with fish cells also use primary cultures of hepatocytes (Baksi and Frazier, 1990), as their functioning have been found to be consistent with the role of the liver in situ (Moon et al., 1985). But in general, primary cultures are cumbersome to prepare, are short lived and produce inconsistent results. Thus liver cell lines that maintain metabolic capability have been sought. Fish liver cell lines are few (Bols and Lee, 1991) and most have been derived from malignant tumours, which renders them unsuitable for correlating the effects toxicants might have in normal tissues as opposed to altered ceils. The R T L - W l cell line derived from rainbow trout liver, expressed cytochrome P4501A1 activity, was sensitive to benzo[a]pyrene, and possessed some attributes of the normal organ of origin (Lee et al., 1993). Thus the present study investigated whether the R T L - W I cell line could be useful as a model fish sentinel cell line

317

318

D.G. BECHTELand L. E. J. LEE

for monitoring the effects of other biotransformation requiring xenobiotics such as aflatoxin. Aflatoxin B~ (AFB) is a potent hepatotoxic, mutagenic and hepatocarcinogenic mycotoxin that may be present as a contaminant in human food and animal feed (Heathcote and Hibbert, 1978). Several animal species have been shown to be affected by AFB, and rainbow trout has been shown to be among the most sensitive species (Sinnhuber et al., 1977). Sensitivity of cells in culture to the toxicity of AFB varies and may depend on the metabolic capability of cells to form an active metabolite that ultimately binds to DNA (Yoneyama et al., 1987), although the relative affinity to DNA might not reflect on the carcinogenic or mutagenic potency of the compound (Loveland et al., 1988). Hepatic metabolism of AFB into the DNA-binding intermediate has been shown to be required for carcinogenicity and suggested for cytotoxicity in rainbow trout (Nunez et al., 1990). In vitro studies of AFB cytotoxicity have, in general, used cell lines which were not derived from liver, the target organ of AFB. In this paper we report the cytotoxic responses of the trout liver cell line, RTL-W1, to AFB and compare these with the responses of two other salmonid cell lines. MATERIALS AND M E T H O D S

Cell cultures. Three cell lines were used: rainbow trout liver, RTL-W1 (Lee et al., 1993), rainbow trout gonad, RTG-2 and Chinook salmon embryo, CHSE214, which were obtained from the American Type Culture Collection (Rockville, MD, USA). Cells were grown routinely in the basal medium, Leibovitz's L-15 supplemented with penicillin-streptomycin (100 IU/ml, 100 pg/ml), glutamine (2 mM), and foetal bovine serum (FBS) at 5% (v/v) for RTL-WI and CHSE-214, and at 9% (v/v) for RTG-2. All media and supplements were purchased from ICN Biomedicals (St Laurent, PQ, Canada). The cells were grown at room temperature in 75-cm2 Falcon flasks (Canlab, Mississauga, ON, Canada) in free gas exchange with the air. Mycoplasma monitoring was done routinely by the method of Chen (1977). Cells were removed from growth surfaces with 0.1% (w/v) bovine trypsin (Sigma, St Louis, MO, USA). For treatments, 1 x 105 cells were plated in 60-mm diameter Falcon plates (Canlab) or in six-well Falcon multidishes (Canlab). The experiments reported in this paper were done with cells between passages 44 to 64 for RTL-W1. Chemicals. Aflatoxins B~, B2, B2~, Ml, PI, QI and aflatoxicol (AFL) were obtained from Sigma, aflatoxicol M 1 was a gift from Dr G. Bailey (Oregon State University, OR, USA). Purity of all chemicals was 96% or greater as indicated by the supplier, and they were used without further purification. The test chemicals were dissolved in sterile dimethyl sulfoxide (DMSO; Sigma), divided into aliquots and stored at - 8 0 ° C . Chemicals for solid-phase extraction

(SPE), thin-layer chromatography (TLC) and HPLC were purchased from BDH Canada (Toronto, ON, Canada). Cytotoxicity tests. Cytotoxicity was determined in several ways. Changes in cellular morphology in response to AFB treatment were monitored in situ under an inverted phase contrast microscope (Nikon Diaphot, Nikon Canada, Mississauga, ON, Canada) or after staining with May-Grunwald and Giemsa (BDH). RTL-WI cell monolayers treated with AFB were fixed in situ in 60-mm diameter Falcon plates and ultrastructural changes were observed by transmission electron microscopy (TEM) as described previously (Lee et al., 1993). Another measure of cytotoxicity was a decline in [3H]thymidine incorporation into acid-insoluble material as the concentration of AFB was increased (test range from 0.001 to 2.5#g/ml). Cells in L-15/FBS were plated into six-well plates at 105 cells per well. After 24hr, the medium was changed to L-15/FBS with varying concentrations of AFB. Within an experiment, three wells were exposed to each concentration. The stock solution was made up in DMSO, and control wells received DMSO at concentrations never exceeding 0.1% (v/v). 24 hr after the addition of AFB, [3H]thymidine (2 #Ci/ml; ICN) was added to each well. As outlined previously (Lee and Bols, 1989), incorporation into acid-insoluble material was measured 24 hr after the addition of the radioactive DNA precursor in a Beckman LS 6000IC liquid scintillation counter. The results were expressed as mean dpm/105 cells for statistical analysis and as a percentage of the incorporation into the control wells for graphic presentation. For each experiment, a single factor analysis of variance (Zar, 1974) was used to test whether AFB had an effect on [3H]thymidine incorporation. Two-tailed hypotheses (null hypothesis: treatment had no effect) were tested and the maximum probability of a type-I error was set at 0.05. If a difference was detected, Dunnett's test (Zar, 1974) was used to compare mean incorporation by AFB-treated cultures with mean incorporation by control cultures. Growth in soft agar. Clonal growth in soft agar has been used as an indicator of neoplastic transformation in mammalian cells (Hamburger and Salmon, 1977; MacPherson, 1969). Agarose type VII (Sigma) was used because the low gelling point makes it more suitable than other agars and agaroses for the mixing in of salmonid cells, which are sensitive to temperatures above 26°C (Bols et al., 1992). Approximately 1 × 105 control or AFB-exposed cells were mixed with agarose to give a final agarose concentration of 0.6% (w/v), which at room temperature is a loose, semi-solid mixture. This was plated into petri dishes with or without a base of 0.6% (w/v) agar, which is solid at room temperature. The growth of RTL-W1 in dishes without agar indicated that being mixed in warm agarose was not lethal to the cells.

Aflatoxin effects in a trout liver cell line Metabolism o f A F B . Aflatoxin metabolites were collected from the media of fish cells treated with AFB at 0.5 #g/ml in 75-cm ~ flasks after 48 hr of exposure. Acetic acid (5 #1) was added to each 10 ml of media collected. Samples were frozen at - 8 0 ° C until extraction was peformed. Media were passed through a methanol-activated octadecyl (C18) SPE column (Burdick and Jackson, Canlab), washed with increasing concentrations of methanol, and the metabolites were eluted in 1 ml dichloromethaneacetonitrile (3: 1, v/v). Samples were evaporated and processed for TLC or HPLC. For TLC, sample volumes of 75-100/~1 in dichloromethaneacetonitrile (3 : 1, v/v) were spotted onto Silica gel 60 TLC plates (Merck, BDH, Canada) and allowed to dry. Separation and development of the samples in TLC plates were adapted from a protocol outlined by Chen et al. (1985). The TLC plates were sequentially developed a t room temperature in the dark, first in ether (55 min) then in chloroform-acetone-water (85: 15: 1, by vol) for 40 rain, and observed under a long-wave UV source. Standards run along with the samples included: aflatoxins BI, B2, B2~, M1, P~, Q~, aflatoxicol and aflatoxicol M~. For HPLC, samples were dissolved in 200#1 acetonitrile-methanoltetrahydrofuran (15 : 20: 6, by vol) and separated by a protocol modified from Loveland et al. (1988). Separation was carried out at room temperature without a guard column in a Spectra Physics SP8800 Ternary HPLC system (Spectra Physics, San Jose, CA, USA) fitted with two Partisphere CI8 columns (5 #m, 4.5 x 12.5 mm; Whatman Inc., Clifton, NJ, USA) and a 100-#1 injection loop. The mobile phase was set for a 15 min linear gradient at 1 ml/min of a 23-46% mixture of acetonitrile-methanoltetrahydrofuran (15:20:6, by vol) in 0.02 N potassium acetate, pH 5.0, after which flow was held constant for 10min and then returned to the initial setting for a subsequent 5min. Metabolites were detected at 345 nm using an Applied Biosystems 757 absorbance detector (Applied Biosystems Inc., Foster City, CA, USA) coupled with a Hewlett Packard 3392A Integrator (Hewlett Packard, Mississauga, ON, Canada). RESULTS

A F B effects on cell morphology

AFB at concentrations in the range of 0.001 to 2.5 # g/ml caused morphological changes in RTL-WI and RTG-2 but not in CHSE-214 cells. The usually epithelial RTL-W1 cells (Plate la) became more fibroblastic in shape (Plate lb) in response to low doses of AFB (less than 0.1 #g/ml). This was noticeable within 2 to 4 days of treatment. Exposure to prolonged and/or higher AFB concentrations led to cell granularity, blebbing, rounding off and death of the cells (Plate Id). RTG-2 cells, on the other hand, did not change their usual fibroblastic shape, but granules accumulated within the cytoplasm and cell

319

death ensued. This was observed after a shorter duration of exposure, at least 24 hr earlier, than with RTL-Wi cells. In contrast, CHSE-214 cells did not appear to be affected by AFB at any of the concentrations tested even after 6 days of treatment. The normal epithelial morphology was unchanged and mitotic figures were abundant in both control and treated cells. Ultrastructural changes associated with AFB treatment for RTL-W1 were in agreement with those described for AFB-induced hepatomas in vivo (Nunez et al., 1991) and AFB-exposed mammalian cells in vitro (Cole et al., 1986; Yoneyama et al., 1987). Dense or residual bodies occurred frequently within the cytoplasm, cytoskeletal filaments appeared disrupted, the nuclear outline was irregular, the chromatin was loosely packed and the nuclear envelope was swollen (Plate 2). Cole et al. (1986) reported that the nucleolar segregation was not very distinct in AFB-treated mammalian hepatocytes. In the present study the nucleolus of AFB-treated RTL-WI cells was less dense (Plate 2c) than in control cells. However, these changes were not consistent throughout the monolayer and were noted only in some cells. A F B effects on D N A synthesis. The effects of AFB on the growth of RTL-WI cells were investigated by monitoring [3H]thymidine incorporation into DNA during the replicative phase of the cells. [3H]Thymidine incorporation at various time intervals by RTL-WI cells after plating was monitored to determine the appropriate labelling conditions for detecting changes in DNA synthesis following the various treatments. If the synthetic phase of the cell cycle is determined, the time period during which inhibition by the test compound occurs should be readily detected. Pulse labelling of 6 hr with [3H]thymidine at 6-hr intervals for 84hr showed RTL-W1 cells to undergo a peak of synthesis around 48 hr after a change of medium after 24 hr of plating. Incorporation was markedly declined by 72 hr, which more or less reflected the time required for the cells to undergo one cell cycle (Fig. 1). This identified the optimal times for treatment, labelling and extraction which were then set at every 24 hr after plating, respectively. The effects of DMSO, the vehicle for the test compounds, on [3H]thymidine incorporation were investigated (Table !). At concentrations ranging from 0.001 to 0.1% on RTL-W1 and RTG-2, DMSO had no significant effect in [3H]thymidine incorporation. With CHSE-214 cells, DMSO did not significantly affect [3H]thymidine incorporation at concentrations as high as 0.5%. AFB at concentrations ranging from 0.01 to 0.2 #g/ml was tested on RTL-Wi and RTG-2, and from 0.01 to 2 #g/ml on CHSE-214, for its ability to affect [3H]thymidine incorporation (Table 2). AFB consistently inhibited [3H]thymidine incorporation by RTL-W1 cells in a dose-dependent manner. From these data, the effective concentration needed for a

32O

D. G. BECHTEL and L. E. J. LEE

50000

Table I. Effect of dimethyl sulfoxide on [3H]thymidine incorporation by three salmonid cell lines Incorporation (% of control) by: DMSO (%, v/v)

40000

0 0.001 0.01 0.1 0.5

09

0

u~ 30000

o~

RTL-Wl

RTG-2

100.0_+6.2 107.0 _+2.2 100.8_+4.0 100.0_+4.2

100.0-+ 15.3 106.6_+0.9 106.3_+4.2 113.6_+2.6

CHSE-214 100.0_+1.0 103.4 _+5.9 101.1_+13.2 102.3 _+2.8

DMSO added to triplicate wells of cells and incubated for 24 hr. [3H]Thymidine was then added and cells were incubated for another 24 hr. Acid-insoluble material was extracted and counted as indicated in Materials and Methods. Incorporation as measured by dpm/plate was not significantly different from the control for any of the cell lines. Mean dpm values expressed as a percentage of the control and the coefficient of variation for each DMSO concentration and cell line are presented.

n 20000 E3 10000

0

0

12

24

36

l Treat

48

60

72

84

T i m e (hr)

Fig. l. Schedule of [3H]thymidine incorporation by RTLW1 cells. R T L - W l cells were labelled with [3H]thymidine every 6 hr for 6 hr, following a change of growth medium, 24 hr after plating. Acid-insoluble material was extracted at the end of each labelling period and incorporation as indicated by d p m was measured for up to 84 hr. Bars indicate standard deviations (n = 3). Arrows indicate the appropriate times for routine treatment, labelling and extraction protocols deduced from this experiment and performed subsequently for the cytotoxicity assays.

5 0 % i n h i b i t i o n (ECs0) for t h e s e cells w a s c a l c u l a t e d to be 0.04 # g / m l . A l t h o u g h t h e i n h i b i t o r y r e s p o n s e to A F B by R T L - W 1 cells w a s slightly m o r e p r o n o u n c e d t h a n by R T G - 2 cells (ECs0 = 0.05 # g / m l ) , t h e differe n c e w a s n o t significant. F o r C H S E - 2 1 4 cells, at t h e c o n c e n t r a t i o n s tested, A F B h a d n o s i g n i f i c a n t effect on [3H]thymidine incorporation.

M e t a b o l i t e s o f A F B were also t e s t e d f o r t h e i r ability to i n f l u e n c e D N A s y n t h e s i s in t h e t h r e e s a l m o n i d cell lines (Fig. 2). T h e cells were e x p o s e d to A F M , A F L , A F Q a n d A F B 2 at c o n c e n t r a t i o n s o f 0.01, 0.05 o r 0.1 # g / m l . A F Q a n d A F B 2 h a d n o s i g n i f i c a n t effect o n t h e t h r e e cell lines tested, b u t AFM and AFL significantly inhibited [3H]thymidine i n c o r p o r a t i o n at 0.05 a n d 0.1 # g / m l in R T L - W 1 a n d R T G - 2 cells. A F M at 0.1 # g / m l also i n h i b i t e d D N A s y n t h e s i s in C H S E - 2 1 4 cells, b u t this i n h i b i t i o n w a s n o t as p r o n o u n e d as for R T L - W l o r R T G - 2 cells. T h e ECs0 f o r A F M w a s c a l c u l a t e d to be 0.07 # g / m l for R T L - W ! a n d 0.08 # g / m l for R T G - 2 cells. T h e s e v a l u e s d i d n o t differ s i g n i f i c a n t l y f r o m e a c h o t h e r , b u t were s i g n i f i c a n t l y different ( P < 0.05) f r o m t h e ECs0 s for A F B for e a c h cell line. A FB transformed sublines. S u r v i v i n g R T L - W l cells t h a t h a d b e e n e x p o s e d o n c e to A F B (0.1 o r 2.5 # g / m l ) for 4 d a y s were m a i n t a i n e d a n d f u r t h e r p a s s a g e d . T h e s e were t e r m e d R T L - W I s u b l i n e s 0 . 1 A a n d 2.5A. T h e s e cells s h o w e d t h e c h a r a c t e r i s t i c d e t e r i o r a t i o n

Table 2. Effect of aflatoxin Bt on [3H]thymidine incorporation by three salmonid cell lines [3H]Thymidine incorporation (% of control) by: AFB (#g/ml) 0 0.001 0.002 0.005 0.01 0.025 0.05 0.1 0.2 1.0 2.0

n 21 1 1 1 14 1 20 19 6

RTL-WI 100.00 103.91 99.43 98.56 87.73 _ 8.94 (6) 64.24 (1) 43.33 + 13.74 (20) 23.98 _+ 10.51 (19) 17.04+7.70 (6)

n

RTG-2

n

CHSE-214

18

100.00

8

100.00

4

98.45 + 2.56

7 7 4 1

93.16 + 3.28 98.20 + 8.57 109.34_+ 14.86 106.93 79.54

13 2 18 16 5

87.05 + 9.27 (4) 74.47 + 8.72 (2) 47.90 + 17.06 (18) 29.45 + 16.57 (16) 30.12_+8.96 (5)

AFB = aflatoxin B~ AFB was added to triplicate wells of cells and incubated for 24 hr. [3H]Thymidine was then added and cells were incubated for another 24 hr. Acid-insoluble material was extracted and counted as indicated in Materials and Methods. Values for incorporation as measured by dpm/plate, were analysed by a single factor analysis of variance. Dunnett's test was used to compare mean incorporation by AFB-treated cultures with mean incorporation by control cultures. Mean dpm values were expressed as a percentage of the control. The mean percentage incorporation and standard deviations for the indicated number of experiments are presented, n = number of separate experiments performed for each concentration of AFB. Numbers in parentheses indicate the number of experiments for which incorporation was significantly different from the control (P < 0.05).

Plate 1. Effect of AFB on the morphology of RTL-W1 cells. Monolayer appearance of RTL-W1 cells in situ at x 100 under an inverted phase contrast microscope (Nikon). (a) Subconfluent monolayer of control RTL-W1 cells. Note the abundance of epithelial-like cells. (b) RTL-W1 cells treated with 0.05 pg AFB/ml for 4 days. Note the predominance of fibroblast-like cells. (c) Semiconfluent monolayer of control RTL-Wl cells. Note the typical cobblestone appearance of the monolayer. (d) RTL-Wl cells treated with 1/~g AFB/ml for 6 days. Many cells have detached from the growth surface. Also note the granularity in the cytoplasm of the few remaining attached cells. Pictures were magnified × 5.

321

Plate 2. Effect of A F B on the ultrastructure of R T L - W l cells. Appearance of RTL-W1 cells at x 16,150 after T E M processing. (a) Typical appearance of control RTL-W1 cell. The nucleus (Nu) is ovoidal, and euchromatic with a prominent nucleolus (n). Intracellular filaments (IF) are abundant and oriented in an organized array along the length of the cell. Mitochondria (Mit) are numerous, indicating a high level of energy production by the cell. R o u g h endoplasmic reticulum (rER) are usually present in variable a m o u n t s around the perinuctear region; smooth endoplasmic reticulum (sER) is a b u n d a n t as well as free ribosomes (rib). Lysosomes (Lys) at various stages of activity can also be seen. Small vacuoles (Vac) are conspicuous around the periphery of the cell. The plasma membrane (PM) is coated with glycocalyx. Although not evident in this picture, the golgi apparatus and residual bodies as well as desmosomes are c o m m o n in these cells. (b,c) R T L - W I cells treated with 0.1 ~g/ml A F B for 2 days. In (b) note the irregular nuclear outline with swollen areas of nuclear envelope (arrowheads). Residual bodies (asterisk) with lamellar structures are abundant. Intracellular filaments (IF) are disrupted and unorganized. In (c) the nucleus was cut tangentially: note the swollen nuclear envelope (bar). Nuclear pores (small arrows) are a b u n d a n t and very conspicuous. Nucleolus (n) is less dense than in control cells and the nucleoplasm appear more euchromatic.

322

Plate 3. Morphology of AFB-transformed R T L - W I sublines. A once only exposure to A F B at 0.1 or 2 , 5 # g / m l for 4 days gave rise to RTL-W1 sublines 0.1A (b,e) and 2.5A (c,f). Control subline 0A is illustrated in a,d. Figures a to c show the confluent sublines in situ under phase contrast microscopy ( x 100) 1 m o n t h after the first passage. Note arrows indicating areas of high cell density, which are especially large in the transformed sublines. Figures d to f show the sublines ( x 40) 2 wk after a second passage following May Grunwald and Giemsa staining. Note increasing size of cell clumps. Pictures were magnified x 5.

323

..~2"

-.
~AFB2

A FQ AFL-M D-

-'~AFBT~

AFM~,.-

~AFP

Plate 4. Thin-layer chromatography profile of AFB metabolites. A C~8 SPE column was used to extract metabolites from the media of salmonid cells treated with 0.5/~g/ml AFB for 2 days. Samples containing standards for AFM, AFL-M, A F Q and AFB (lane a) and AFP, AFB2~, AFB 2 and AFL (lane e), and the extracted samples from CHSE-214 (lane b), RTL-W1 (lane c) and RTG-2 (lane d) were blotted on silica gel TLC plates, and the chromatogram was developed as indicated in Materials and Methods. 1" and 2* indicate solvent fronts of ether, and chloroform-acetone water, respectively. This is one of six representative chromatograms viewed under a long-wave UV source and photographed with a Wratten 2B gelatin filter.

324

Aflatoxineff~tsina trout liver cell line R T L - Wl

325

seen in RTL-W1 cells when exposed to AFB for 4 days. The number of cells remaining in the highest treatment was less than half of the control. Control 120 cells that had been exposed once to DMSO at 0.1% 100 (subline 0A) remained generally unchanged, and looked like normal RTL-WI cells. After a change of medium the treated cells started to recover and proliferated faster than the control cells. The sublines 8 60 were first passaged on 3 October 1991 and have since ~ 4o undergone 16 subcultivations. Several flasks of each subline after passage 5 have been frozen in liquid 2o nitrogen. Viability after thawing was greater than 0 80%. Although the number of cells in subline 2.5A AFM AFL AFQ AFB 2 was less than the control cells after the first passage, they surpassed the control cell number before the second passage. Sublines 0.1A and 2.5A showed a distinct monolayer morphology (Plate 3). FibroRTG - 2 blastic cells clumped together forming large mounds 140 • of cells. These cells proliferated faster than in subline 120 • 0A or in normal RTL-Wl cells. The fibroblastic cells of AFB-treated sublines appeared to have lost contact inhibition of growth as they grew in clumps and 1,0oo . ridges, forming large cell mounds. This was readily visible in situ (Plate 3b,c) or after May-Grunwald Giemsa staining (Plate 3e,f). Williams et al. (1973) observed the appearance of similar mounds, "piledup" cells, in AFB-transformed rat liver cells. In addition, some cells of subline 2.5A lost anchorage dependency of growth and could survive in suspen0 sion. Many of the floating cells could regain anchorAFM AFL AFQ AFB 2 age,dependency of growth when subcultured into a new culture plate. This was not observed with floating cells from subline 0A. AFB-induced neoplastic transformation of RTLW1 cells could account for the above changes in cell CHSE - 214 morphology and behaviour. Clonal growth in soft 140 agar, an in vitro assay for neoplastic transformation 120 (Hamburger and Salmon, 1977), was investigated to test whether these cells had undergone this change. After 3 wk of incubation, colonies were observed growing within the soft agar for subline 0.1A. Fewer colonies were observed for subline 2.5A; however, the number and size of the colonies increased by the end of 4 wk of incubation. The percentage of cells resulting in colony formation was approximately 5 and 1% 20 for 0.1A and 2.5A, respectively. No colony formation was observed in the control subline 0A or in the 0 untreated RTL-W1 cells. Growth in soft agar was AFM AFL AFQ AFB 2 performed with passage 5 or greater for the sublines, Fig. 2. Effect of AFB metabolites on [3H]thymidineincorporation by salmonid cells. RTL-W1, RTG-2 and CHSE- and clonal growth extended past 10 wk. A F B metabolites. Media from cells treated with 214 cells were plated at 1 x 105in each well of six-wellplates. 24 hr later, the regular growth media was changed to AFB at 0.1 or 0.5/~g/ml for I or 2 days were collected the treatment solutions which consisted of 0.01 (1~), 0.05 and analysed by TLC and HPLC. AFM and AFL ( I ) or 0.I0 (B)/~g/ml of AFM, AFL, AFQ or AFBJml. [~H]Thymidine was added 24 hr later, and after another were two major metabolites that could be detected by 24 hr of incubation, DNA was extracted and incorporation both methods. The relative amounts of these two as dpm was measured. The percentage incorporation rela- metabolites varied with the cell type (Fig. 3). AFL tive to each set of control cultures was calculated. Treat- was predominant over AFM in RTL-W1 and CHSEments were carried out in triplicate, and the bars indicate 214 cells, although the relative amounts of these two standard deviations. Asterisks indicate those treatments that metabolites was far less abundant in the latter cells. were significantly different from the control (*P <0.05; In contrast, AFM was more abundant than AFL in Dunnett's test). 140

i 100 80 c 6O

326

D.G. BECHTELand L. E. J. LEE 10

common environmental contaminant, it is critical that the cell line has metabolizing capability. RTL¢/-/j W1 was shown to be sensitive to the effects of B[a]P ¢/,/j (Lee et al., 1993), and appears to be sensitive to AFB V//, at concentrations as low as 0.01 pg/ml. Thus, RTL¢1/1 Wl could be useful as a model fish sentinel cell line ,~ 'VIA (I~ ¢##J for monitoring the effects of biotransformation..i: 6 rH~ (I~ ¢tH requiring xenobiotics. However, the usefulness of an assay depends on a clear understanding of the applicability and limitations of the system, and thus I-one should keep in mind that the cytotoxic response ,..; may vary with the cells and assay conditions used. CHSE-214 cells were reported to be sensitive to AFB ###J in terms of protein synthesis with an ECs0 of 7.5/lg/ml after 20 hr of exposure (Chen et al., 1985), whereas Vosdingh and Neff (1974) reported BB cells (a line of catfish cells) to be sensititive to AFB in RTL - Wl RTG - 2 C H S E - 214 terms of cytopathic effects at an ECs0 of 0.02/~g/ml Cell line after 14 days of exposure. Thus, considerations should be given in time, temperature and mode Fig. 3. Relative amounts o f A F M ( m ) and A F L (1~1) extracted from AFB-treated salmonid cells. Metabolites of measurement for the cytotoxicity assay, and a from the media of salmonid cells treated with 0.5#g comparison of the cytotoxicity data must be done AFB/ml for 2 days were extracted using a C~8 SPE column with due care. and separated by HPLC. The major metabolites detected Studies on the cytotoxic effects in terms of DNA were AFM and AFL. The relative amounts of these metabolites are presented as a percentage of total peak area from synthesis appear to be most suitable for chemicals the total integrated profile for each cell line. A single that interact with DNA and have carcinogenic effects. representative run from four separate experiments for RTL- Sensitivity to AFB was not significantly different WI and RTG-2, and from two experiments for CHSE-214, between the rainbow trout cell lines in terms of is illustrated. morphology and DNA synthesis, although the metabolic products of AFB were produced in different RTG-2 cells. The exposure time or the amount of ratios between the two cell lines. On the other hand, AFB treatment did not appear to have an effect on the Chinook salmon cell line was unresponsive to the the general metabolite profile of the individual cell effects of AFB at the concentrations and exposure types. However, other metabolites became apparent times tested. Perhaps differences in the metabolic with increasing time and concentration of the parent activation of AFB between the salmonid species compound. The type and amount of the metabolites might account for the observed differences in sensiwere distinct for each cell line (Plate 4). One metab- tivity. Despite the differential toxicity to AFB, as olite which eluted close to AFM had characteristics reported in a variety of biological systems (Chen et al., 1985; Coulombe et al., 1984; Yoneyama et al., of AFL-M. This compound was present in the TLC 1987; Zhang et al., 1990), changes in the morphology and HPLC profiles of RTL-Wl and RTG-2 media, but was not evident in CHSE-214. AFP, AFQ and of RTL-W1 and RTG-2 cells are in common with AFB2 were identified by TLC with given standards. reported responses in other cells (Chen et al., 1985; AFB2, was another AFB metabolite that was ident- Vosdingh and Neff, 1974), and the ultrastructural ified by TLC, but could not be quantitated because changes observed with RTL-W1 are in agreement it co-migrated with AFM. Various other metabolites to those reported with liver cells exposed to AFB could not be identified but appeared in distinct in vivo (Nunez et al., 1991), and to AFB-exposed patterns for each cell line. RTL-Wl showed two mammalian cells in vitro (Cole et al., 1986; Yoneyama distinct metabolites migrating close to the chloroform et al., 1987). A carcinogen such as AFB is specific for liver tissue solvent front and between AFL and AFB, while a metabolite migrating beyond AFM and AFB2, was in vivo and forms epoxides that covalently bind to more pronounced in RTG-2 extracts. Several other DNA of liver cells, killing or transforming them into unidentified polar metabolites close to the origin and neoplastic cells. The sublines of RTL-W1 obtained within the vicinity of AFP were seen in all three cell after a single exposure to AFB behaved like neoplastically transformed cells. These cells lost contact extracts. inhibition of growth, changed morphology, grew in suspension and in soft agar and had a higher DISCUSSION proliferation rate than the originating RTL-WI cells. The choice of cell line for in vitro testing is insignifi- Bausher and Schaeffer (1974) developed a rat liver cant for direct acting toxins, but for bioactivated cell line that showed sensitivity to AFB by cell toxins such as AFB or benzo[a]pyrene (B[a]P), a number, protein content and RNA synthesis after ¢111

e'l

¢111 H/J ¢1~,

rH~

¢HJ

Atiatoxin effects in a trout liver cell line 48 hr of exposure, but these cells were less sensitive than other established cell lines. They contend that because liver cells are target cells for carcinogenesis by AFB, these cells would be less sensitive to its cytotoxic effects. On that basis, Bausher and Schaeffer (1974) believe they had isolated the target cells susceptible to AFB carcinogenesis. In addition, Nunez et al. (1990) reported that cytotoxicity, in common with carcinogenicity, is dependent on the metabolism of AFB and that cytotoxicity contributes to, but is not required for, hepatocarcinogenesis. They suggest that oval cells, which have lower monooxygenase activity than hepatocytes, and have the ability to survive and proliferate, might be the critical target cells of carcinogens. Because RTL-Wl was quite sensitive to the cytotoxic effects of AFB, and showed a similar, if not more sensitive, response to that seen with RTG-2 cells, it appears that RTL-WI cells are a mixed population of liver cells (Lee et al., 1993), containing target and non-target cells for AFB's carcinogenic effects, and that RTL-W1 sublines 0. I A and 2.5A represent the selected target cells. This is in agreement with the findings by Toyoshima et al. (1970) with a newborn Wistar rat cell line, NLW, in which the fibroblastic cells within the cell population appeared to be transformed more readily than the more epithelial cells. Toyoshima et al. (1970) also found that lower AFB concentrations were more effective in producing malignancy in vitro than higher concentrations. This also appears to be the case with the RTL-W1 sublines in which subline 0.1A had a higher percentage of cells forming colonies in soft agar than the 2.5A subline. In mammals, AFB has also been used as a simple and efficient single-step selective method for obtaining variant hepatoma cells of a wide variety of phenotypes (Corcos and Weiss, 1988). Tumorigenicity was observed with AFB-treated TRL cells (derived from rat liver) when injected into syngeneic rats (Williams et al., 1973), with NLW cells (Toyoshima et al., 1970) and with BL8 rat liver cell line when injected into nude mice (Sinha et al., 1987). In future investigations, the AFB-transformed RTL-W1 sublines will be further characterized and biochemical comparisons will be made to investigate the mechanisms of carcinogenesis in fish. Levels of ~-glutamyl transpeptidase are currently being studied. The levels of this enzyme have been shown to alter with transformation in cultured liver cells (Huberman et al., 1979; Sinha et al., 1987) and its activity was increased in AFB-treated rat liver cells (Manson and Green, 1982). Whether the RTL-Wl sublines are tumorigenic remains to be investigated. The major metabolites of AFB in trout liver have been reported to be AFL > AFM (Loveland et al., 1988). Our studies showed that RTL-W1 produced these metabolites in the same relative amounts, while RTG-2 produced AFM > AFL, and CHSE-214 produced very low amounts of the same. The activation efficiency of AFB to mutagens, as detected TIV 8/3--B

327

by Salmonella typhimurium, was reported to be approximately three times greater in trout hepatocytes in comparison with coho salmon hepatocytes (Coulombe et al., 1984). This might account for the mutagenic potential of AFB and its metabolites which was AFB > AFL > AFM (Coulombe et al., 1982) for trout liver cells. On the other hand, DNA binding has been implied to directly correlate with sensitivity to AFB (Nunez et al., 1990). Loveland et al. (1988) showed the order of affinity of AFB and metabolites for DNA to be AFB > AFM > AFL, and although in terms of carcinogenic potency (AFB ~>AFL > AFM) and mutagenic potency (AFB > AFL > AFM) (Coulombe et al., 1982) there appears to be no correlation, Nunez et al. (1990) indicated a linear dose response in both DNA binding and cytotoxicity. The results with both RTL-WI and RTG-2 for inhibition of DNA synthesis with AFB and metabolites also agree with these findings. The ECs0 for AFB and AFM were 0.04 and 0.07 pg/ml, respectively, for RTL-W1 and 0.05 and 0.08 #g/ml, respectively, for RTG-2. The ECs0 for AFL was greater than 0.1/~g/ml for both cell lines. Thus, the cytotoxic potency of these compounds in RTL-WI and RTG-2 was A F B > AFM > A F L which correlates well with DNA affinity. Whether there is a correlation between cytotoxicity and carcinogenicity or mutagenicity remains to be established. CHSE-214 cells were unresponsive except to AFM at 0.1 #g/ml, which needs to be further investigated. Overall, RTL-WI cells were slightly more sensitive than RTG-2 cells and reflected more accurately the in vivo response to AFB. Thus compounds like AFB and B[a]P that affect the liver as primary targets could be better investigated with the RTL-W1 cell line. Most cells in culture lose a metabolic activation system, such as the cytochrome P-450 (Glatt et al., 1987), which means that the compounds requiring metabolic activation cannot readily exert their effects. In the study reported here and in our previous study (Lee et al., 1993), we have shown that the RTL-Wl cell line is a metabollically competent cell system and could be valuable in future mutagenicity and carcinogenicity studies, biochemical studies of metabolism, and hepatoxicity and general cytotoxicity screening. Acknowledgements--This work was supported by an oper-

ating grant from the Natural Sciences and Engineering Research Council of Canada to L. E. J. L. This paper represents part of the thesis research conducted at the University of Saskatchewan in partial fulfilment for the degree of Masters of Science in Toxicology for D.G.B. We thank J. Gibbons and S. Caldwell for technical and photographic help. REFERENCES

Babich H. and Borenfreund E. (1987) Cultured fish cells for the ecotoxicity testing of aquatic pollutants. Toxicity Assessment: An International Quarterly 2, 119-133.

328

D.G. BECHTEL and L. E. J. LEE

Babich H. and Borenfreund E. (1991) Cytotoxicity and genotoxicity assays with cultured fish cells: a review. Toxicology in Vitro 5, 91-100. Baksi S. M. and Frazier J. M. (1990) Isolated fish hepatocytes--model systems for toxicology research. Aquatic Toxicology 16, 229-256. Bausher J. and Schaeffer W. I. (1974) A diploid rat river cell culture: 1. Characterization and sensitivity to aflatoxin B~. In Vitro 9, 286-293. Bols N. C. and Lee U E. J. (1991) Technology and uses of cell cultures from the tissues and organs of bony fish. Cytotechnology 6, 163-187. Bols N. C., Mosser D. D. and Steels G. B. (1992) Temperature studies and recent advances with fish cells in vitro. Comparative Biochemistry and Physiology 103A, 1-14. Chen T. R. (1977) In situ detection of mycoplasma contamination in cell cultures by fluorescent Hoechst 33258 stain. Experimental Cell Research 104, 255 262. Chen S.-C. G., Lin C.-H. and Cheng M.-C. (1985) Action of carcinogenic mycotoxin AFB t on cultured cells of Salmonid species. Journal of the Chinese Biochemical Society 14, 28-38. Cole K. E., Hsu I.-C. and Trump B. F. (1986) Comparative ultrastructural effects of aflatoxin B~ on mouse, rat, and human hepatocytes in primary culture. Cancer Research 46, 1290-1296. Corcos L. and Weiss M. C. (1988) Efficient one step selection of hepatoma cell variants of a variety of phenotypes by use of aflatoxin B~. Differentiation 38, 134-139. Coulombe R. A., Bailey G. S. and Nixon J. E. (1984) Comparative activation of aflatoxin BI to mutagens by isolated hepatocytes from rainbow trout (Salmo gairdneri) and coho salmon (Oncohynchus kisutch). Carcinogenesis 5, 29 33. Coulombe R. A., Shelton D. W., Sinnhuber R. O. and Nixon J. E. (1982) Comparative mutagenicity of ariatoxins using a Salmonella/trout hepatic enzyme activation system. Carcinogenesis 3, 1261-1264. Frazier J. M. (1992) In Vitro Toxicity Testing. Marcel Dekker, Inc., New York. Glatt H., Gemperlein I., Turchi G., Heinritz H., Doehmer J. and Oesch F. (1987) Search for cell culture systems with diverse xenobiotic-metabolizing activities and their use in toxicological studies. Molecular Toxicology 1, 313-334. Hamburger A. W. and Salmon S. E. (1977) Primary bioassay of human tumor stem cells. Science 197, 461-463. Heathcote J. G. and Hibbert J. R. (1978) Aflatoxins: Chemical and Biological Aspects. Elsevier, Amsterdam. Huberman E., Montesano R., Drevon C., Kumki T., Saint Vincent L., Pugh T. D. and Goldfarb S. (1979) Gamma-glutamyl transpeptidase and malignant transformation of cultured liver cells. Cancer Research 39, 269-272. Kremers P., Negro L. and Gielen J. E. (1990) Rat fetal hepatocytes in cultures: a model for metabolic and toxicological studies. Acta Pharmacologica Jugoslavica 40, 383-393. Lee L. E. J. and Bols N. C. (1989) Action of cortisol on the proliferation of rainbow trout fibroblasts. Cell Tissue Kinetics 22, 291-301. Lee L. E. J., Clemons J. H., Bechtel D. G., Caldwell S. J., Han K. B., Pasitschniak-Arts M., Mosser D. D. and Bols

N. C. (1993) Development and characterization of a rainbow trout liver cell line expressing cytochrome p450-dependent monooxygenase activity. Cell Biology and Toxicology 9, 279-294. Loveland P. M., Wilcox J. S., Hendricks J. D. and Bailey G. S. (1988) Comparative metabolism and DNA binding of aflatoxin B~, ariatoxin M~, aflatoxicol and aflatoxicolM~ in hepatocytes from rainbow trout (Salmo gairdneri). Careinogeneis 9, 441-446. MacPherson I. (1969) Agar suspension in culture for quantitation of transformed cells. In Fundamental Techniques in Virology. Edited by K. Habel and N. P. Salzman. pp. 214-219. Academic Press. New York. Manson M. M. and Green J. A. (1982) Effect of microsomally activated AFB~ on GGT activity in three rat liver cell lines. British Journal of Cancer 95, 945-951. Moon T. W., Walsh P. J. and Mommsen T. P. (1985) Fish hepatocytes: a model metabolic system. Canadian Journal of Fisheries and Aquatic Sciences 42, 1772-1782. Nunez O., Hendricks J. D. and Duimstra J. R. (1991) Ultrastructure of hepatocellular neoplasms in aflatoxin B~ (AFBl)-initiated rainbow trout (Oncorhynchus mykiss). Toxicologic Pathology 19, 11-23. Nunez O., Hendricks J. D. and Fong A. T. (1990) Interrelationships among aflatoxin B l (AFB1) metabolism, DNA-binding, cytotoxicity, and hepatocarcinogenesis in rainbow trout Oncorhynchus mykiss. Diseases of Aquatic Organisms 9, 15-23. Paganuzzi-Stammati A. L., Silano V. and Zucco F. (1981) Toxicology investigations with cell culture systems. Toxicology 20, 91-153. Sinha S., Hockin L. J. and Neal G. E. (1987) A system for transformation of rat liver cells in vitro by acute treatment with aflatoxin. British Journal of Cancer 55, 595-598. Sinnhuber R. O., Hendricks J. D., Wales J. H. and Putnam G. B. (1977) Neoplasms in rainbow trout, a sensitive animal model for environmental carcinogenesis. Annals of the New York Academy of Sciences 298, 389-408. Toyoshima K., Hiasa Y., Ito N. and Tsubura Y. (1970) In vitro malignant transformation of cells derived from rat liver by means of aflatoxin B~. Gann 61, 557 561. Vosdingh R. A. and Neff M. J. C. (1974) Bioassay of aflatoxins by catfish cell cultures. Toxicology 2, 107-I 12. Watson R. R. (1992) In Vitro Methods of Toxicology. CRC Press, Boca Raton, FL. Williams G. M., Elliott J. M. and Weisburger J. H. (1973) Carcinoma after malignant conversion in vitro of epithelial-like cells from rat liver following exposure to chemical carcinogens. Cancer Research 33, 6064512. Yoneyama M., Sharma R. P. and Eisner Y. Y. (1987) Effects of mycotoxins in cultured kidney cells: cytotoxicity of aflatoxin bl in madin-darby and primary fetal bovine kidney cells. Ecotoxicology and Environmental Safety 13, 174-184. Zar J. H. (1974) Biostatistical Analysis. pp. 130 162. Prentice Hall Inc., Englewood Cliffs, NJ. Zhang S.-Z., Lipsky M. M., Trump B. F. and Hsu I.-C. (1990) Neutral Red (NR) assay for cell viability and xenobiotic-induced cytotoxicity in primary cultures of human and rat hepatocytes. Cell Biology and Toxicology 6, 219-234.