- Email: [email protected]
Response of Drosophila embryonic cells to tumor promoters
TOXICOLOGY
AND
APPLIED
Response NICOLE Division
oj‘(l?mgerzetics
85,196-206
PHARMACOLOGY
of Drosophila
( 1986)
Embryonic
BOURNIAS-VARDIABASIS’ and Cytology,
Received
Cit)!
Jarmar)>
Cells to Tumor Promoters AND JOSEPHINE
of Hope
National
3. 1986; accepted
Medical
.4pril
Center,
C. FLORES Duarte.
California
91010
IK. 19M
Response of Drosophila Embryonic Cells to Tumor Promoters. BOURNIAS-VARDIABASIS, N. FLORES, J. C. (1986). Toxicol. .4ppl. Pharmacol. 85, 196-206. Several recent observations on tumor promoters point to the many developmental and embryonic characteristics associated with their mode of action. These observations have led us to investigate the effects of a series of tumor promoters on Drosophila embryonic cultures at both the morphological and molecular levels. The cultures have been used with some successby us to assessthe teratogenic potential of a large number of molecules, including drugs, chemicals. and environmental pollutants. In this culture system, 12-0-tetradecanoylphorbol 13-acetate (TPA), the most potent of the tumor promoters tested. disrupted normal muscle and neuron differentiation at concentrations ranging from 0.1 to 10 GM: 4-0-methylphorbol 12-myristate a weak stage I promoter. used at concentrations the same as or higher than those of TPA had no inhibitory effect on cell differentiation. A selected group of tumor promoters was also investigated at the molecular level for their effects on differentiating Drosophila cells. All tumor promoters tested induced synthesis of three heat shock proteins. On the basis of these two levels of effects (morphological and molecular) it is apparent that the tumor promoters tested act similarly as teratogens do in the Dmrophila embryonic cultures. This finding confirms some recent published reports suggesting that a large number of tumor promoters act as teratogens if the exposure interval is during the embryonic rather than the adult stage. We suggest that this system can be usefully employed to investigate some of the common mechanisms involved in tumor promotion and teratogenesis. (~1 1986 Academic AND
Press. Inc.
Various workers in the field of teratogenesis basis et al., 1983a,b; Bournias-Vardiabasis and Flores, 1983). The assay is based on the have proposed that chemicals which interfere principle that teratogenic effects can be with intercellular communication or cell-cell interactions or induce cell surface changes caused by abnormal cell death, failure of cell interaction. reduced biosynthesis, or imduring early organogenesishave the potential of being teratogens. If they are present in the peded morphogenetic movement (Wilson, adult form of the organism these chemicals 1977). Since tumor promoters can also act at the epigenetic level by inhibiting cell-to-cell have the potential of acting astumor promotcommunication. it is hypothesized that the ers (Ellinger, 1982; Huber and Brown, 1983a,b; Trosko et al., 1982; Yotti et al.. Drosophila embryonic cells undergoing morphogenesis and differentiation may be af1979). In our laboratory, we have developed an in fected by such a class of agents. A limited number of tumor promoters already have vitro teratogenesis assay which utilizes Drobeen shown to affect both frog embryos and sophila embryonic cultures (Bournias-Varblastomeresby alteration of cell surface propdiabasisand Teplitz, 1982; Bournias-Vardiaerties and inhibition of histogenesis(Ellinger, 1982), to interfere with bud production in ’ To whom reprint requests should be addressed. H.vdra (Shiba, 1981). and to inhibit early 0041-008X/86
$3.00
Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved
196
Drosophila
RESPONSE
TO TUMOR
morphogenesis in sea urchins (Bresch and Arendt, 1978). More recently Huber and Brown (1983a,b) in a series of experiments utilizing rat embryo cultures have shown that after exposure to TPA2 a disruption of the morphology and function of the embryonic visceral yolk sac occurs, although cellular differentiation appears normal. There has also been increasing evidence that several of the biologically effective chemicals which give negative results in the Ames-type assays but block cell-cell communication in a variety of metabolic cooperation cell culture systems, [i.e., phenobarbital, polybrominated biphenyls, Valium, and dilantin (Trosko et al.. 1982)] are also structural and/or behavioral teratogens (Shepard, 1983). While originally only neuron and muscle differentiation were used as morphological endpoints for assessing teratogenic potential, in the Drosophila assay, we recently extended the endpoints utilized to identify teratogens to include assessment of differences in the number and levels of embryonic proteins. Thus, we have examined, by two-dimensional gel electrophoresis, the effects of these drugs on protein synthesis in embryonic cells. Addition of a teratogen results in the induction of three proteins of molecular weight about 20,000 in addition to the normal proteins synthesized by untreated cells. These proteins have been identified as a subset of a group of stress-induced proteins, termed heat shock proteins (hsp). The three small heat shock proteins induced by teratogens have been identified as hsp 23, 22a, and 22b (Buzin and Boumias-Vardiabasis, 1982, 1984). These proteins are induced in a wide spectrum of organisms and in response to a variety of metabolic insults (heat, drugs, an-
’ Abbreviations used: TPA, 12-O-tetradecanoylphorbol 13-acetate; 4-0-MeTPA. 4-0-methylphorbol 12-myristate: 4-cu-PDD, 4-cu-phorbol 12,13-dideconate; PDA, phorbol 12,13-diacetate; PDD, phorbol 12,13-dideconate: PDB, phorbol 12,13-dibenzoate: ODC, omithine decarboxylase; hsp, heat shock protein(s); FCS, fetal calf serum: SDS, sodium dodecyl sulfate.
PROMOTERS
197
oxia, etc.) (Ashburner and Bonner, 1979; Schlesinger et al., 1982). The functional significance of the heat shock response is unknown although some recent results, plus our own experimental observations, suggest that the response might afford the organism thermal or stress protection (Buzin and Bournias-Vardiabasis, 1982; Schlesinger et al., 1982). There exist also reports that contradict this protective function of hsps (for a review see Nover, 1984). Obviously, much more information needs to be obtained before the functional significance of hsps is determined. Tumor promoters have been shown to induce embryonic proteins in the adult (Balmain, 1976) and to increase RNA, protein, and DNA synthesis (Diamond et al., 1980) and there is one recent report regarding induction of heat shock proteins in mouse epidermal cells after exposure to TPA (Gindhart et al., 1984). In the limited number of tumor promoters tested, we observed the induction of the same set of proteins as are induced after exposure to teratogens (Buzin and Boumias-Vardiabasis, 1982, 1984). The purpose of this study is to investigate the interactions of tumor promoters with invertebrate embryogenesis and examine the hypothesis that this class of chemicals affects embryonic cells in a fashion similar to that of some teratogens. METHODS Chemicals All chemicals were purchased from Sigma (St. Louis. MO.) except for the various phorbol esters used. which were purchased from LC Services Corporation (Wobum. Mass.). [%]Methionine was from New England Nuclear (Boston, Mass.). Coomassie blue and all electrophoresis reagents were purchased from Bio-Rad (Richmond. Calif.).
Dosage Since the Drosophila assay was originally developed for teratogen testing as a prescreen, it had to be able to respond to dose levels that are markedly lower than those
198
BOURNIAS-VARDIABASIS
that have proved to be toxic to the adult (i.e., testing for a developmental hazard only). All chemicals were initially fed to adult female Drosophila at doses where at least half ofthe adults died (LD50). Once the LD50 value was established for that particular agent, the initial dose tested in the in vitro assay was 0.0 1 ofthe LD50 dose (i.e., TPA has an LD50 value of 1000 PM in the adult female while the highest concentration used in identifying tumor promoters was 10 PM). By this means the lowest concentration producing the irreversible endpoint effect can be determined.
Preparation ofDrosophila Cell Cultures Drosophila eggs were collected for 2 hr, dechorionated, and sterilized at the early gastrula stage (3.5 hr after oviposition), before any overt morphologjcal or ultrastructural differentiation took place. The eggs were homogenized and the embryonic cells were collected after centrifugation. The number of cells was estimated by standard hemocytometer methods, and 2 ml of cells was plated out in 35-mm cell culture dishes at 8 X lo5 cells per milliliter of modified Schneider’s medium that was supplemented with 18% fetal calf serum (heat inactivated). After the cells were plated out and allowed to attach to the bottom of the dish ( 15-20 min), the medium covering the cells was replaced with medium in which the particular chemical to be tested had been dissolved. The solution was sterilized by being passed through a 0.45~pm Millipore filter. Control samples were treated the same way as experimental samples by removing the medium from the dish and replacing it with fresh medium. Cell and tissue differentiation could be scored at about 24 hr, at which time the living cells were initially observed under phasecontrast microscopy. Ifthe cells are left in the petri dishes for 48 or 72 hr instead of 24. more axon contacts will be made but there is no further quantifiable increase in the number of myotubes or ganglia (unpublished observation). The differentiated cultures. in preparation for scoring, were rinsed in 0.1% trypsin in saline solution to remove cellular debris. fixed in Zenker’s. stained with hematoxylin, and counterstained with Evans Blue (10% in saline). Scoring of the cultures, which entailed counting the number of myotubes and neuron clusters (ganglia) in the culture, was carried out with an automated image analysis system manufactured by Bausch and Lomb. The field of view is randomly chosen and its size is 25 mm2 (this represents only a fraction of the total surface area of the petri dish). The image analyzer is instructed to score myotubes as shapes that are at least 50 pm long and 10 pm thick. while ganglia are round cells of convex perimeters ranging between 60 and 100 pm. In these primary cultures. several cell types differentiate from their respective stem cells during a 24-hr period; of particular importance to this assay is the development of myoblasts and neuroblasts.
AND
FLORES
Myoblasts, which in Drosophila are stem cells and are not bipolar, divide synchronously once in culture at 5-6 hr after gastrulation. Elongation, aggregation, and alignment of the daughter myocytes begin at about 12 hr after gastrulation and continue for 4-6 hr; most of the aligned myocytes fuse to form myotubes, and spontaneous pulsations may occur by 24 hr. Muscle cells show thick and thin myofilaments, T tubules, and sarcoplasmic reticulum (Seecofet al., 1973b). Neuroblasts. in cultures as well as in vivo. undergo a series of eight unequal divisions, which begin shortly after the initiation of gastrulation. This is followed by a final round of division of the daughter cells, giving rise to clusters of about 18 neurons. The daughter neurons apparently recognize and adhere to each other, forming miniature ganglia with cell bodies at the periphery and neuropiles with synapses in the interior. Axons, 50 +rn or longer, are observed first in cultures at 7.5 hr after gastrulation and increase in length and number through about 16.5 hr. Acetylcholinesterase and choline acetyltransferase first appeared in embryonic cultures at 8 and 12.5 hours after gastrulation. Growing axons contact and recognize muscle cells in vitro and form functional neuromuscular junctions (Seecof et a/., 1973a). A chemical was classified as a positive if it resulted in a statistically significant reduction in the total number of myotubes and ganglia when compared to controls. Thus. interference with normal cell differentiation was inferred to be an indication of tumor promoter activity. A total of four dishes per trial were scored and the total number of myotubes and ganglia was ranked to the cell numbers obtained in parallel control cultures (Wilcoxon’s signed rank test). This ranking test was necessary since we wanted to compare matched groups whose numbers might vary markedly. For more details regarding the assay, see Bournias-Vardiabasis and Teplitz (I 982) and Boumias-Vardiabasis et(I/. (1983a).
Metabolicffrtivutic)r1 qj’Drux.s
Since some of the phorbol esters proved to be relatively inactive in the Drosophila embryonic culture, an S-27 Drosophila (microsomal) fraction was added prior to application. Details of S-27 microsome preparation have been described previously (Boumias-Vardiabasis and Flares, 1983). The reaction mixture was made up of I mM NADPH, 8 mM MgCIZ, IO mM glucose phosphate. and 0.4 ml ofthe microsomal fraction. The mixture plus the phorbol ester to be tested were combined in dialysis bags and incubated in 8 ml of medium plus 18% FCS for 4-6 hr. The medium outside the dialysis tubing containing the metabolites was then added to the cell cultures and its effect assessed after 18 hours. Controls were handled the same way with the omission of the phorbol ester.
Drosophila Radiolabeling
RESPONSE
TO
and Fluorography
Cells were labeled with [35S]methionine (200 @/ml) for 1 hr at 26°C and then pelleted in Drosophila saline and solubilized as described by Buzin and Seecof ( 198 1). All samples contained between 100,000 to 500,000 counts per minute. Two-dimensional gel electrophoresis was carried out as described by O’Farrell ( 1975). The isoelectric focusing gels containing a 5.0-7.0 pH gradient were run in 130 X 2.5-mm cylindrical tubes. After electrophoresis, gels were extruded, the pH gradient was determined, and then the gels were run on a SDS-polyacrylamide 9-IS% gradient slab gel for the second dimension Gels were then stained with Coomassie blue (0.05%). fluorographed, dried. and exposed to Kodak XAR iilm. Integrated optical densities ofhsp 22a and 22b were measured from fluorograms with an Omnicon image analysis system (Bausch and Lomb). Densities from treated cells were normalized to the density of actin II from the same sample. The normalized values for proteins 22a and 22b were then averaged and the results expressed as relative increases over control values, Protein 23 does not always separate well enough from protein N to provide consistent measurements. At least three independent labelings were performed.
RESULTS Of the 20 compounds tested, 13 have been previously identified as tumor promoters in various assays (Diamond et al., 1980: Trosko et al., 1982: Cat-r et al.. 1984). The Drosophila assay identified 10 of them as inhibitors of neuron and/or muscle differentiation and as such could be considered tumor promoters on the basis of their effects on cell differentiation (Table 1, Fig. 1). Anthralin and saccharin, under our conditions and concentrations used [both classified as weak tumor promoters (Diamond et al., 1980; Trosko et al., 1982)], did not have an affect on neuron and/ or muscle differentiation (Table 1). The rest of the compounds used were ones suggested for validation purposes since they are structurally related to tumor promoters but show no such action in standard tumor promotion assays (Sivak, 1982; Weinstein, 198 1). None of these seven compounds tested showed an inhibitory effect in the differentiating cell cultures (Table 1).
TUMOR
PROMOTERS
199
The doses used for the various compounds tested varied considerably, but, as mentioned under Methods, the concentrations used were determined from the LD50 values established from female adult Drosophila. From the dosages we could rank TPA, PDA, PDB, PDD, diphenylhydantoin, griseofulvin, mellitin, and 5azacytidine as having an inhibitory effect on cell differentiation stronger than that of coumarin, lithocholic acid, and phenobarbital, but this comparison would be appropriate only for Drosophila. Muscle and neuron differentiation is complete at around 24 hr; thus when a tumor promoter is added from 24 to 48 hr there is no effect on muscle and neuron number. Furthermore, if the tumor promoter is removed from the cultures and replaced with normal medium at 24 hr, a large population of the undifferentiated cells differentiates into neurons and myotubes (unpublished observation). Thus, the inhibition of differentiation by the tumor promoter is a reversible phenomenon as also observed in other cell systems. Our previous results in identifying human teratogens through the Drosophila assay indicate that these tumor promoters act as teratogens if there is an exposure during the embryonic-fetal stage. Certainly, comparison of our own data with in vivo findings of identified tumor promoters and teratogens showed a high degree of overlap (Table 1). To assessthe effects of tumor promoters at the molecular level, we examined the effects of several tumor promoters/teratogens on protein synthesis in embryonic cells by twodimensional gel electrophoresis. All proteins synthesized by control cells were also synthesized in treated cells. However, all of the compounds identified as tumor promoters when tested in the Drosophila cells induced the synthesis of three low molecular weight proteins [22a, 22b, and 23 (molecular weights about 22,000 and 23,000)]. These proteins are found in only trace amounts in untreated cells. The proteins migrate with identical electrophoretic mobilities on two-dimensional gels as the heat-shock proteins hsp 22a.
200
BOURNIAS-VARDIABASIS
AND
TABLE
FLORES
I
Assay finding (% of controls)* Concentration (PM)
Compound
I.000 1.000 30 2.000 1,000 1,000 500 100 500 100 50 10 100 10 I 250 150 10
Control Anthrahn Anthraquinone 5-Azacytidine Biphenyl Cholesterol Coumarin Diphenylhydantoin
Griseofulvin Lithocholic
Acid
Tumor promoter” + +c NT +’
Mellitin 0.5 0.1 500 10,000 1,000 10
Phenobarbital Phenylalanine Saccharin Phorbol Esters TPA
0.1 0.01 10 10 + S-27 fraction IO 10 + S-27 fraction 10 10 + S-27 fraction 10 10 + S-27 fraction 10 100 100 + S-27 fraction
PDA PDB PDD 4-0-MeTPA 4-a-PDD Phorbol S-27 fraction
alone
+
+ + * -
Ganglia
Myotubes
100 95 88 58 89 100 33* 48* 85 34* 45* 44* 49* 95 103 93 30* 52* 89 71 92 100 66 100 68 31* 67 80 99 100 74 100 65 90 90 100 95 65 100 90 101
100 72 72 50* 88 96 36* 64 87 30* 49* 83 88 6* 27* 106 6* 18* 93 34* 50* 58 31* 94 110 46* 42* 53* 87 89 46+ 92 40* 98 50* 98 90 77 96 I15 105
Teratogen as indicated in in viva assays’ NT” NT NT NT
NT Nl. + +’ + NT
NT NT NT N-lNT NT
0 As designed by Diamond et nl. ( 1980). Sivak ( 1982). and Trosko et ul. ( 1982). *Drosophila embryonic cell cultures were prepared as described by Bournias-Vardiabasis and Tephtz (1982) and Bournias-Vardiabasis er (11. ( 1983a). The cells were allowed to differentiate at 26’C in the presence of the compound for 24 hr. Cultures were stained and numbers of differentiated myotubes and ganglia were counted with a Bausch and Lomb automated image analysis system. Results were expressed as a percentage of the number of myotubes and neuron clusters in control cultures prepared on the same day. A total of four dishes per trial was scored and each compound was tested in three or more separate trials. The addition ofthe compound. on three separate trials, resulted in the average number of myotubes and/or neuron clusters being significantly lower than those of the controls (Wilcoxon’s signed rank tested (Y < 0.05, indicated by *). Some of these data were taken from Boumias-Vardiabasis et al. (1983a). ‘From Shepard (1983). d NT, not tested.‘ ’ 5-Azacytidine has recently been classified also as a tumor promoter or syncarcinogen (Carr et al., 1984). fDiohenvlhvdantoin has been classified as a tumor nromoter bv Trosko et al. (1982) on the basis of in vitro assays rather-than-from in vivo studies. gOnly as a result of phenylketonuria. Phenylafanine is a constituent of the Drosophila cell culture medium. h f, weak tumor promoter or weak teratogen.
Drosophila
RESPONSE
TO
TUMOR
PROMOTERS
201
FIG. 1. Differentiated Drosophila cells. Embryonic cell culture, 24 hr. N, neuron clusters derived from divisions of neuroblasts; M, myotube derived from fusion of myocytes; A, axons. (A) Cultures treated with TPA. Note incomplete fusion of myocytes and the reduced number of neurons (only 8 cells are visible in this cluster). SEM. X 1250 (each bar = 10 Fm). (B) Controls, showing fully differentiated embryonic cells. Neuron clusters (N) usually have 16 or more cells. SEM. X 1250 (each bar = IO pm).
!b, and 23 (Buzin and Bournias-Vardianis. 1982, 1984). To simplify the results we lmpared only the 22,000-Da proteins of the
treated and control groups. 5-Azacytidi ne (300 WM) [identified as both a teratog :en (Shepard, 1983) and a tumor promoter or
202
BOURNIAS-VARDIABASIS
AND F’LORES
TABLE 2 INDUCTIONOFHEATSHOCKPROTEINSBYTUMORPROMOTERS
Compound 5-Azacytidine Biphenyl Coumarin Diphenylhydantoin Griseofulvin Phenobarbital TPA 4-0-MeTPA PDB Saccharin
Concentration (PM) 30 300 2000 1000 100 10 500 10 10 10 1000
Class TP/T b -d TPJT TP/T TP/T TP/T TP/T’ TPfTP/T’ TP/p
22-kDa protein (relative increase over control)’ 7 L’ 86 I 64’ 21’ 4 25 21 5.5 4.5 0.8’
’ Integrated optical densities for proteins 22a and 22b were determined from fluorographs with a Bausch and Lomb automated image-analysis system. Values were normalized to the integrated optical density of actin II on the same film. The normalized values for proteins 22a and 22b were averaged and the results are expressed as the factors of increase over normalized untreated control values. ’ TP, tumor promoter; T, teratogen. ’ Data taken from Buzin and Bournias-Vardiabasis ( 1984. 1982). d p, negative. ’ Designated by Huber and Brown (1983a) as teratogens in an in vitro rat assay.
syncarcinogen (Carr et al., 1984)] increased the levels of the induced proteins by a factor of 86 and as such elicited the strongest response. Coumarin (1000 PM), phenobarbital (500 PM), TPA ( 10 ,uM), and diphenylhydantoin ( 100 PM) ail produced levels of induction ranging between 2 1 and 25. Griseofulvin ( 10 PM), PDB ( 10 PM), and 4-0-MeTPA ( 10 j&M) showed induction over control of approximately a factor of 4. Biphenyl showed no relative increase over control levels. Saccharin (1000 PM), a weak tumor promoter (Diamond et al., 1980) and also a weak teratogen (Shepard, 1983) did not induce any of these 3 hsp in Drosophila cells (Table 2, Fig. 2). DlSCUSSION Little is known yet about the molecular basis of tumor promoter action. Promoters are thought to have the cell membrane as their primary target while initiators are thought to be damaging to DNA (Weinstein, 198 1).
These promoters have not been shown to be mutagenic in bacterial or mammalian cells (Sivak, 1982) so Ames-type tests have failed to identify them. This class of chemicals has been shown to exhibit a variety of specific effects, such as a change in the cell phenotype, an increase in protein-specific synthesis (not attributable to increased cell division), cell surface changes, decrease in acetylcholine receptor levels, and inhibition of terminal differentiation, among others (Diamond ef al., 1980). All of these perturbation events indicate an epigenetic mode of action. Teratogens also can be thought of as a large class of chemicals. some acting by epigenetic means, others by their DNA damaging potential. Furthermore, a causal mechanistic link has been proposed for both sets of agents: a number of tumor promoters and teratogens have been shown to interfere with normal cellcell communication and cell-cell interacand Bournias-Vardiabasis, tions (Buzin 1984; Trosko et al., 1982). It has been pro-
l>r~&la
RESPONSE
TO
TUMOR
203
PROMOTERS
b
d
FIG. 2. Protein synthesis in Dro.&zif~ primary embryonic cells treated with tumor promoters. Drosophila cells were plated at 3.5 hr after oviposition at 5 X IO6 cells per dish. The indicated agent was added at about 4 hr after oviposition and the cells were allowed to differentiate at 25°C for 18 hr in its presence. Cells were then labeled for I hr with [35S]methionine (200 nCi/ml) in the presence of the particular agent. Cells were solubilized and proteins separated by two-dimensional gel electrophoresis and visualized by fluorography. Agents used and exposure times (cpm X days) were TPA (480,000) (a); PDD (380,000) (b); Coumarin (362.000) (c): Control (498,000) (d); N is a non-heat shock protein that migrates close to protein 23. IEF. isoelectric focusing.
posed that if such an event happens in the adult organism, it results in uncontrollable growth (i.e., cancer): during embryonic development such interference would lead to abnormal development (i.e., teratogenesis).
Our results on the action of tumor promoters in the Drosophilu assay, indicate that this class of compounds inhibit muscle and or neuron specific embryonic differentiation. This observation confirms other reports
204
BOURNIAS-VARDIABASIS
showing that TPA in cell culture systems inhibits differentiation of Friend leukemia cells (Rovera et al., 1977), adipose conversion of mouse preadipose cells (Diamond et al., 1978). formation of neuroblastoma cells (Ishii et al., 1978) and myotube formation of myoblasts (Cohen et al., 1977: Dlugosz et al., 1983). In fact, it has been suggested that the ability of these compounds to inhibit terminal differentiation may be central to their mode of action as tumor promoters (Diamond et al., 1980). Three of the phorbol esters (PDA, PDB, and PDD) required incubation with an S-27 Drosophila microsome fraction before their teratogenic activity was detected. Little information about the metabolism of phorbol diesters is currently available; while hamster fibroblasts have been shown to be able to metabolize TPA (O’Brien and Diamond, 1978a,b), HeLa cell cultures show no such activity (Kreibich et al., 1974). Diamond et al. (1978) have shown that TPA lost its ability to induce ornithine decarboxylase (ODC) after 2-3 days of exposure to either normal or transformed hamster cells. They also found that PDB and PDD retained ODC-inducing activity several days later than did TPA. O’Brien and Saladik (1980) have also reported that PDD had a slower metabolism than TPA and again noticed that while human fibroblasts did not metabolize either of the tumor promoters, hamster fibroblasts metabolized TPA and PDD, although the latter was again metabolized much more slowly. In the Drosophila system it was ascertained that these particular phorbol esters require incubation with a microsomal system in order to inhibit cell differentiation. Further studies are underway to determine rates of metabolism for these phorbol esters. Furthermore, the assay is obviously exhibiting a high degree of sensitivity to this particular mode of action of tumor promoters. As such, it can be argued that it could be developed as a potential in vitro assay to identify potential tumor promoters. Such a need for in vitro assays has been put forward in recent
AND
F-LORES
reports (Diamond et al., 1980; Mondal et al., 1976; Sivak, 1982). Diamond et al. ( 1978) have also suggested that promotion may be easier to analyze in vitro than in vivo where absorption is a pulse rather than a continuous exposure to target cells. We see the Drosophila assay as providing several good endpoints for such assessment. Furthermore, techniques are currently being developed in our laboratory to assess the response of Drosophila cells to tumor promoters in terms of acetylcholine receptor levels. In chick embryo muscle cells, TPA has already been reported to cause both an increase in the rate of degradation of acetylcholine receptors and a decrease in the rate of receptor synthesis (Miskin et al., 1978). Preliminary experiments in Drosophila with teratogens already suggest that such modulations will also be present in the embryonic cells after exposure to tumor promoters. More importantly, Drosophila embryonic cells can provide us with the much needed information on molecular mechanisms of tumor promotion. It is not yet known why only certain types of cells show inhibition of differentiation while other cell types show an induction of differentiation. Cell cycling has been suggested as a possible primary requirement in the mechanism of inhibition of differentiation (Slaga et al., 1978). Although Drosophila embryonic cells are far removed from mouse skin epidermis, the evolutionarily conserved nature of many genetic and molecular mechanisms suggests that Drosophila studies may be useful in aiding our understanding of two-step carcinogenesis. They offer a very well-developed model for studying differentiation processes and thus analyzing inhibition of differentiation. The vast knowledge available on Drosophila genetics, development, and molecular biology may provide us with many clues to the role of genetic and biochemical variables in the process of carcinogenesis. Already the data obtained on heat shock protein induction suggest that at the molecular level, the cells’ response to tumor promot-
Drosophila RESPONSE
TO
ers is similar to their response to other “stress” agents. Boutwell ( 1974) and others have already proposed that altered gene expression is fundamental to the action of tumor promoters. This would mean that promoters can change the pattern of protein synthesis qualitatively or quantitatively. This ability of tumor promoters to modify gene activity by altering differentiation processes makes them valuable in studying the role of alterations in cellular differentiation in carcinogenesis as well as in normal differentiation. ACKNOWLEDGMENTS We wish to thank Carolyn Buzin and R. L. Teplitz for discussions and comments on the manuscript, and Marlene McCarrey and Rochelle Hoylman for skillful and patient preparation ofthe manuscript. Scanning electron micrographs were prepared in the Shared Instrumentation Laboratory using the Philips SEM 505 purchased with NIH funds. This work has been supported by a BRSG grant from the National Institutes of Health and the Gerald I. Parisi Research fellowship.
REFERENCES ASHBURNER, M., AND BONNER, J. (1979). The induction of gene activity in Drosophila by heat shock. Cell 17,241-254. BALMAIN, A. (1976). The synthesis of specific proteins in adult mouse epidermis during phases of proliferation and differentiation induced by the tumor promoter TPA and in basal and differentiating layers of neonatal mouse epidermis. J. Invest. Dermatol. 67,246-253. BOURNIAS-VARDIABASIS, N., TEPLITZ, R. L., CHERNOFF, G. F.. AND SEECOF, R. L. (1983a). Detection of teratogens in the Drosophila embryonic cell culture test: Assay of 100 chemicals. Teratology 28, 109- 122. BOURNIAS-VARDIABASQ N., BUZIN. C. H., AND REILLY, J. G. (1983b). The effect of 5-azacytidine and cytidine analogs on Drosophila melanogaster cells in culture. Wilhelm Roux’s Arch. Dev. Biol. 192,299-302. BOURNIAS-VARDIABASIS, N., AND FLORES, J. (1983). Drug metabolizing enzymes in Drosophila melanogas/cr. Teratogenicity of cyclophosphamide in vitro. Ter-
atog. Carcinog. Mutagen. 3,255-262. BOURNIAS-VARDIABASIS, N., AND TEPLITZ. R. L. ( 1982). Use of Drosophila embryo cell cultures as an in vitro teratogen assay. Teratog. Carcinog. Mutagen. 2,333-341. BOUTWELL. R. K. (1974). Function and mechanism of
TUMOR
PROMOTERS
205
promoters of carcinogenesis. CRC Crit. Rev. Toxicol. 2,419-447. BRESCH. H., AND ARENDT, U. (1978). Disturbances of early sea urchin development by the tumor promoter TPA (phorbol ester). Natwwissenschajien 65, 660662. BUZIN, C. H., AND BOURNIAS-VARDIABASIS, N. (1984). Teratogens induce a subset of small heat shock proteins in Drosophila primary embryonic cell cultures. Proc. Natl. Acad. Sci. USAgl, 4075-4019. BUZIN, C. H., AND BOURN~AS-VARDIABASIS, N. (1982). The induction of a subset of heat shock proteins by drugs that inhibit differentiation in Drosophila embryonic cell cultures. In Heat Shock: From Bacteria to Man (M. J. Schlesinger, M. Ashbumer, and A. Tissieres, eds.), pp. 387-394. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. BUZIN, C. H., AND SEECOF, R. L. (198 1). Developmental modulation ofprotein synthesis in Drosophilu primary embryonic cell cultures. Dev. Genet. (N.Y.) 2, 237252. CARR, B. I., REILLY. J. G., SMITH, S. S., WINBERG, C.. AND RIGGS, A. (1984). The tumorigenicity of 5-azacytidine in the male Fischer rat. Curcinogenesis 5, 1583-1590. COHEN. R., PACIFICI, M.. RUBINSTEIN, N., BIEHL, J., AND HOLTZER. H. (I 977). Effect of a tumour promoter on myogenesis. Nature (London) 266,538-540. DIAMOND, L., O’BRIEN, T. G., AND BAIRD, W. M. (1980). Tumor promoters and the mechanism of tumor promotion. Adv. Cancer Res. 32, l-74. DIAMOND, L., O‘BRIEN, T. G., AND ROVERA, G. (1978). Tumor promoters: Effects on proliferation and differentiation of cells in culture. Life 23, 1979-1988. DLUGOSZ, A. A., TAPSCOTT, S. J.. AND HOLTZER. H. ( 1983). Effects of phorbol 12-my&ate 13-acetate on the differentiation program of embryonic chick skeletal myoblasts. Cancer Res. 43,2780-2789. ELLINGER. M. S. (1982). Responses of amphibian embryos and blastomeres to a tumor-promoting phorbol ester. Cancer Res. 42,2804-28 12. GINDHART. T. D.. STEVENS, L.. AND COPLEY, M. P. (1984) Transformation and tumor promoter sensitive phosphoproteins in JB-6 mouse epidermal cells: One is also sensitive to heat stress. Carcinogenesis 5, 1 I 151121. HUBER, B. E., AND BROWN, N. A. (1983a). Tumor promoter actions on rat embryonic development in culture. Cancer Res. 43,5544-555 I. HUBER. B. E., AND BROWN, N. A. (1983b). 12-O-tetradecanoylphorbol13-acetate actions on macromolecular synthesis, omithine decarboxylase, and cellular differentiation ofthe rat embryonic visceral yolk sac in culture. Cancer Res43,5552-5559. ISHII, D. N., FIBACH, E., YAMASAKI, H., AND WEINSTEIN. I. B. (1978). Tumor promoters inhibit
206
BOURNIAS-VARDIABASIS
morphological differentiation in cultured mouse neuroblastoma cells. Science 200,556-559. KREIBICH, G., Suss, R., AND KINZEL, R. (1974). On the biochemical mechanism of tumorigenesis in mouse skin. V. Studies of metabolism of tumor promoting and non-promoting phorbol derivatives in vivo and in vitro. 2. Krebsforsch. 81, 135-149. MISKIN, R., EASTON, T. G., MAELICKE, A., AND REICH, E. (1978). Metabolism of acetylcholine receptor in chick embryo muscle cells: Effects of RSV and PMA. Cell 15, 1287-1300. MONDAL, S.. BRANKOW, D., AND HEIDELBERGER, C. (1976). Two stage chemical oncogenesis in cultures of C3H/lOT$ cells. Cancer Res. 36,2254-2260. NOVER, L. (ed.) (1984). Heat Shock Response ofEukar.yotic Cells. Springer-Verlag, Berlin. O’BRIEN. T. G.. AND DIAMOND, L. ( 1978a). A cell culture bioassay to analyze metabolism of phorbol diester tumor promoters. Cancer Res. 38,2567-2572. O’BRIEN, T. G., AND DIAMOND, L. (1978b). Metabolism of tritium-labeled 12-O-tetradecanoylphorbol13-acetate by cells in culture. Cancer Res. 38,2562-2566. O’BRIEN, T. G., AND SALADIK, D. (1980). Differences in the metabolism of 12-O-[3H]tetradecanoylphorbol13-acetate and [3H]phorbol12,13-dideconate by cells in culture. Cancer Res. 40,4433-4437. O’FARRELL. P. (1975). High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250,40074021. ROVERA, G., O’BRIEN, T. G., AND DIAMOND, L. (1977). Tumor promoters inhibit spontaneous differentiation of Friend erythroleukemia cells in culture. Proc. Nat/.
Acad. Sci. USA 74,2894-2898. SCHLESINGER, M. F., ASHBURNER, M., AND TISSIERES. A.. (eds.) (1982). Heat Shock: From Bacteria to Man.
AND
FLORES
Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. SEECOF, R. L., DONADY, J. J., AND TEPLITZ, R. L. (1973a). Differentiation of Drosophila neuroblasts to form ganglion-like clusters of neurons in vitro. CeN D@er. 2, 143-149. SEECOF, R. L.. GERSON, I., DONADY, J. J., AND TEPLITZ. R. L. (1973b). Drosophila myogenesis in vitro: The genesis of small myocytes and myotubes. Dev. Biol. 35,250-26 1. SHEPARD, T. H. ( 1983). Catalog of Terutogenic Agents Johns Hopkins Press. Baltimore. SHIBA, Y. (198 1). Laboratory action of 12-O-tetradecanoylphorbol13-acetate on bud production in Hydra.
Wilhelm Rowz’s Arch. Dev. Biol. 190.305-307. SIVAK. A. (1982). Evaluation of assay procedures for detection of tumor promoters. Mutat. Res. 98,377-387. SLAGA. T. J.. SIVAK, A., AND BOUTWELL, R. K. (eds.) (1978). Carcinogenesis Mechanisms of Tumor Production and Cocarcinogenesis, Vol. 2, Raven Press, New York. TROSKO, J. E., CHANG, C. C., AND NETZLOF. M. (1982). The role of inhibited cell-cell communication in teratogenesis. Teratog. Carcinog. Mutagen. 2, 3 I 45. WEINSTEIN, I. B. (198 1). Current concepts and controversies in chemical carcinogenesis. J. Supramol. Struct. Cell. Biochem. 17,99- 120. WILSON, J. G. (1977). Current status of teratology. In Handbook c$ Teratology (J. G. Wilson and F. C. Fraser, eds.). pp. 309-355. Plenum, London. YOTTI, L. P.. CHANG, C. C.. AND TROSKO, J. E. ( 1979). Elimination of metabolic cooperation in Chinese hamster cells by a tumor promoter. Science 206. 10891091.
AND
APPLIED
Response NICOLE Division
oj‘(l?mgerzetics
85,196-206
PHARMACOLOGY
of Drosophila
( 1986)
Embryonic
BOURNIAS-VARDIABASIS’ and Cytology,
Received
Cit)!
Jarmar)>
Cells to Tumor Promoters AND JOSEPHINE
of Hope
National
3. 1986; accepted
Medical
.4pril
Center,
C. FLORES Duarte.
California
91010
IK. 19M
Response of Drosophila Embryonic Cells to Tumor Promoters. BOURNIAS-VARDIABASIS, N. FLORES, J. C. (1986). Toxicol. .4ppl. Pharmacol. 85, 196-206. Several recent observations on tumor promoters point to the many developmental and embryonic characteristics associated with their mode of action. These observations have led us to investigate the effects of a series of tumor promoters on Drosophila embryonic cultures at both the morphological and molecular levels. The cultures have been used with some successby us to assessthe teratogenic potential of a large number of molecules, including drugs, chemicals. and environmental pollutants. In this culture system, 12-0-tetradecanoylphorbol 13-acetate (TPA), the most potent of the tumor promoters tested. disrupted normal muscle and neuron differentiation at concentrations ranging from 0.1 to 10 GM: 4-0-methylphorbol 12-myristate a weak stage I promoter. used at concentrations the same as or higher than those of TPA had no inhibitory effect on cell differentiation. A selected group of tumor promoters was also investigated at the molecular level for their effects on differentiating Drosophila cells. All tumor promoters tested induced synthesis of three heat shock proteins. On the basis of these two levels of effects (morphological and molecular) it is apparent that the tumor promoters tested act similarly as teratogens do in the Dmrophila embryonic cultures. This finding confirms some recent published reports suggesting that a large number of tumor promoters act as teratogens if the exposure interval is during the embryonic rather than the adult stage. We suggest that this system can be usefully employed to investigate some of the common mechanisms involved in tumor promotion and teratogenesis. (~1 1986 Academic AND
Press. Inc.
Various workers in the field of teratogenesis basis et al., 1983a,b; Bournias-Vardiabasis and Flores, 1983). The assay is based on the have proposed that chemicals which interfere principle that teratogenic effects can be with intercellular communication or cell-cell interactions or induce cell surface changes caused by abnormal cell death, failure of cell interaction. reduced biosynthesis, or imduring early organogenesishave the potential of being teratogens. If they are present in the peded morphogenetic movement (Wilson, adult form of the organism these chemicals 1977). Since tumor promoters can also act at the epigenetic level by inhibiting cell-to-cell have the potential of acting astumor promotcommunication. it is hypothesized that the ers (Ellinger, 1982; Huber and Brown, 1983a,b; Trosko et al., 1982; Yotti et al.. Drosophila embryonic cells undergoing morphogenesis and differentiation may be af1979). In our laboratory, we have developed an in fected by such a class of agents. A limited number of tumor promoters already have vitro teratogenesis assay which utilizes Drobeen shown to affect both frog embryos and sophila embryonic cultures (Bournias-Varblastomeresby alteration of cell surface propdiabasisand Teplitz, 1982; Bournias-Vardiaerties and inhibition of histogenesis(Ellinger, 1982), to interfere with bud production in ’ To whom reprint requests should be addressed. H.vdra (Shiba, 1981). and to inhibit early 0041-008X/86
$3.00
Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved
196
Drosophila
RESPONSE
TO TUMOR
morphogenesis in sea urchins (Bresch and Arendt, 1978). More recently Huber and Brown (1983a,b) in a series of experiments utilizing rat embryo cultures have shown that after exposure to TPA2 a disruption of the morphology and function of the embryonic visceral yolk sac occurs, although cellular differentiation appears normal. There has also been increasing evidence that several of the biologically effective chemicals which give negative results in the Ames-type assays but block cell-cell communication in a variety of metabolic cooperation cell culture systems, [i.e., phenobarbital, polybrominated biphenyls, Valium, and dilantin (Trosko et al.. 1982)] are also structural and/or behavioral teratogens (Shepard, 1983). While originally only neuron and muscle differentiation were used as morphological endpoints for assessing teratogenic potential, in the Drosophila assay, we recently extended the endpoints utilized to identify teratogens to include assessment of differences in the number and levels of embryonic proteins. Thus, we have examined, by two-dimensional gel electrophoresis, the effects of these drugs on protein synthesis in embryonic cells. Addition of a teratogen results in the induction of three proteins of molecular weight about 20,000 in addition to the normal proteins synthesized by untreated cells. These proteins have been identified as a subset of a group of stress-induced proteins, termed heat shock proteins (hsp). The three small heat shock proteins induced by teratogens have been identified as hsp 23, 22a, and 22b (Buzin and Boumias-Vardiabasis, 1982, 1984). These proteins are induced in a wide spectrum of organisms and in response to a variety of metabolic insults (heat, drugs, an-
’ Abbreviations used: TPA, 12-O-tetradecanoylphorbol 13-acetate; 4-0-MeTPA. 4-0-methylphorbol 12-myristate: 4-cu-PDD, 4-cu-phorbol 12,13-dideconate; PDA, phorbol 12,13-diacetate; PDD, phorbol 12,13-dideconate: PDB, phorbol 12,13-dibenzoate: ODC, omithine decarboxylase; hsp, heat shock protein(s); FCS, fetal calf serum: SDS, sodium dodecyl sulfate.
PROMOTERS
197
oxia, etc.) (Ashburner and Bonner, 1979; Schlesinger et al., 1982). The functional significance of the heat shock response is unknown although some recent results, plus our own experimental observations, suggest that the response might afford the organism thermal or stress protection (Buzin and Bournias-Vardiabasis, 1982; Schlesinger et al., 1982). There exist also reports that contradict this protective function of hsps (for a review see Nover, 1984). Obviously, much more information needs to be obtained before the functional significance of hsps is determined. Tumor promoters have been shown to induce embryonic proteins in the adult (Balmain, 1976) and to increase RNA, protein, and DNA synthesis (Diamond et al., 1980) and there is one recent report regarding induction of heat shock proteins in mouse epidermal cells after exposure to TPA (Gindhart et al., 1984). In the limited number of tumor promoters tested, we observed the induction of the same set of proteins as are induced after exposure to teratogens (Buzin and Boumias-Vardiabasis, 1982, 1984). The purpose of this study is to investigate the interactions of tumor promoters with invertebrate embryogenesis and examine the hypothesis that this class of chemicals affects embryonic cells in a fashion similar to that of some teratogens. METHODS Chemicals All chemicals were purchased from Sigma (St. Louis. MO.) except for the various phorbol esters used. which were purchased from LC Services Corporation (Wobum. Mass.). [%]Methionine was from New England Nuclear (Boston, Mass.). Coomassie blue and all electrophoresis reagents were purchased from Bio-Rad (Richmond. Calif.).
Dosage Since the Drosophila assay was originally developed for teratogen testing as a prescreen, it had to be able to respond to dose levels that are markedly lower than those
198
BOURNIAS-VARDIABASIS
that have proved to be toxic to the adult (i.e., testing for a developmental hazard only). All chemicals were initially fed to adult female Drosophila at doses where at least half ofthe adults died (LD50). Once the LD50 value was established for that particular agent, the initial dose tested in the in vitro assay was 0.0 1 ofthe LD50 dose (i.e., TPA has an LD50 value of 1000 PM in the adult female while the highest concentration used in identifying tumor promoters was 10 PM). By this means the lowest concentration producing the irreversible endpoint effect can be determined.
Preparation ofDrosophila Cell Cultures Drosophila eggs were collected for 2 hr, dechorionated, and sterilized at the early gastrula stage (3.5 hr after oviposition), before any overt morphologjcal or ultrastructural differentiation took place. The eggs were homogenized and the embryonic cells were collected after centrifugation. The number of cells was estimated by standard hemocytometer methods, and 2 ml of cells was plated out in 35-mm cell culture dishes at 8 X lo5 cells per milliliter of modified Schneider’s medium that was supplemented with 18% fetal calf serum (heat inactivated). After the cells were plated out and allowed to attach to the bottom of the dish ( 15-20 min), the medium covering the cells was replaced with medium in which the particular chemical to be tested had been dissolved. The solution was sterilized by being passed through a 0.45~pm Millipore filter. Control samples were treated the same way as experimental samples by removing the medium from the dish and replacing it with fresh medium. Cell and tissue differentiation could be scored at about 24 hr, at which time the living cells were initially observed under phasecontrast microscopy. Ifthe cells are left in the petri dishes for 48 or 72 hr instead of 24. more axon contacts will be made but there is no further quantifiable increase in the number of myotubes or ganglia (unpublished observation). The differentiated cultures. in preparation for scoring, were rinsed in 0.1% trypsin in saline solution to remove cellular debris. fixed in Zenker’s. stained with hematoxylin, and counterstained with Evans Blue (10% in saline). Scoring of the cultures, which entailed counting the number of myotubes and neuron clusters (ganglia) in the culture, was carried out with an automated image analysis system manufactured by Bausch and Lomb. The field of view is randomly chosen and its size is 25 mm2 (this represents only a fraction of the total surface area of the petri dish). The image analyzer is instructed to score myotubes as shapes that are at least 50 pm long and 10 pm thick. while ganglia are round cells of convex perimeters ranging between 60 and 100 pm. In these primary cultures. several cell types differentiate from their respective stem cells during a 24-hr period; of particular importance to this assay is the development of myoblasts and neuroblasts.
AND
FLORES
Myoblasts, which in Drosophila are stem cells and are not bipolar, divide synchronously once in culture at 5-6 hr after gastrulation. Elongation, aggregation, and alignment of the daughter myocytes begin at about 12 hr after gastrulation and continue for 4-6 hr; most of the aligned myocytes fuse to form myotubes, and spontaneous pulsations may occur by 24 hr. Muscle cells show thick and thin myofilaments, T tubules, and sarcoplasmic reticulum (Seecofet al., 1973b). Neuroblasts. in cultures as well as in vivo. undergo a series of eight unequal divisions, which begin shortly after the initiation of gastrulation. This is followed by a final round of division of the daughter cells, giving rise to clusters of about 18 neurons. The daughter neurons apparently recognize and adhere to each other, forming miniature ganglia with cell bodies at the periphery and neuropiles with synapses in the interior. Axons, 50 +rn or longer, are observed first in cultures at 7.5 hr after gastrulation and increase in length and number through about 16.5 hr. Acetylcholinesterase and choline acetyltransferase first appeared in embryonic cultures at 8 and 12.5 hours after gastrulation. Growing axons contact and recognize muscle cells in vitro and form functional neuromuscular junctions (Seecof et a/., 1973a). A chemical was classified as a positive if it resulted in a statistically significant reduction in the total number of myotubes and ganglia when compared to controls. Thus. interference with normal cell differentiation was inferred to be an indication of tumor promoter activity. A total of four dishes per trial were scored and the total number of myotubes and ganglia was ranked to the cell numbers obtained in parallel control cultures (Wilcoxon’s signed rank test). This ranking test was necessary since we wanted to compare matched groups whose numbers might vary markedly. For more details regarding the assay, see Bournias-Vardiabasis and Teplitz (I 982) and Boumias-Vardiabasis et(I/. (1983a).
Metabolicffrtivutic)r1 qj’Drux.s
Since some of the phorbol esters proved to be relatively inactive in the Drosophila embryonic culture, an S-27 Drosophila (microsomal) fraction was added prior to application. Details of S-27 microsome preparation have been described previously (Boumias-Vardiabasis and Flares, 1983). The reaction mixture was made up of I mM NADPH, 8 mM MgCIZ, IO mM glucose phosphate. and 0.4 ml ofthe microsomal fraction. The mixture plus the phorbol ester to be tested were combined in dialysis bags and incubated in 8 ml of medium plus 18% FCS for 4-6 hr. The medium outside the dialysis tubing containing the metabolites was then added to the cell cultures and its effect assessed after 18 hours. Controls were handled the same way with the omission of the phorbol ester.
Drosophila Radiolabeling
RESPONSE
TO
and Fluorography
Cells were labeled with [35S]methionine (200 @/ml) for 1 hr at 26°C and then pelleted in Drosophila saline and solubilized as described by Buzin and Seecof ( 198 1). All samples contained between 100,000 to 500,000 counts per minute. Two-dimensional gel electrophoresis was carried out as described by O’Farrell ( 1975). The isoelectric focusing gels containing a 5.0-7.0 pH gradient were run in 130 X 2.5-mm cylindrical tubes. After electrophoresis, gels were extruded, the pH gradient was determined, and then the gels were run on a SDS-polyacrylamide 9-IS% gradient slab gel for the second dimension Gels were then stained with Coomassie blue (0.05%). fluorographed, dried. and exposed to Kodak XAR iilm. Integrated optical densities ofhsp 22a and 22b were measured from fluorograms with an Omnicon image analysis system (Bausch and Lomb). Densities from treated cells were normalized to the density of actin II from the same sample. The normalized values for proteins 22a and 22b were then averaged and the results expressed as relative increases over control values, Protein 23 does not always separate well enough from protein N to provide consistent measurements. At least three independent labelings were performed.
RESULTS Of the 20 compounds tested, 13 have been previously identified as tumor promoters in various assays (Diamond et al., 1980: Trosko et al., 1982: Cat-r et al.. 1984). The Drosophila assay identified 10 of them as inhibitors of neuron and/or muscle differentiation and as such could be considered tumor promoters on the basis of their effects on cell differentiation (Table 1, Fig. 1). Anthralin and saccharin, under our conditions and concentrations used [both classified as weak tumor promoters (Diamond et al., 1980; Trosko et al., 1982)], did not have an affect on neuron and/ or muscle differentiation (Table 1). The rest of the compounds used were ones suggested for validation purposes since they are structurally related to tumor promoters but show no such action in standard tumor promotion assays (Sivak, 1982; Weinstein, 198 1). None of these seven compounds tested showed an inhibitory effect in the differentiating cell cultures (Table 1).
TUMOR
PROMOTERS
199
The doses used for the various compounds tested varied considerably, but, as mentioned under Methods, the concentrations used were determined from the LD50 values established from female adult Drosophila. From the dosages we could rank TPA, PDA, PDB, PDD, diphenylhydantoin, griseofulvin, mellitin, and 5azacytidine as having an inhibitory effect on cell differentiation stronger than that of coumarin, lithocholic acid, and phenobarbital, but this comparison would be appropriate only for Drosophila. Muscle and neuron differentiation is complete at around 24 hr; thus when a tumor promoter is added from 24 to 48 hr there is no effect on muscle and neuron number. Furthermore, if the tumor promoter is removed from the cultures and replaced with normal medium at 24 hr, a large population of the undifferentiated cells differentiates into neurons and myotubes (unpublished observation). Thus, the inhibition of differentiation by the tumor promoter is a reversible phenomenon as also observed in other cell systems. Our previous results in identifying human teratogens through the Drosophila assay indicate that these tumor promoters act as teratogens if there is an exposure during the embryonic-fetal stage. Certainly, comparison of our own data with in vivo findings of identified tumor promoters and teratogens showed a high degree of overlap (Table 1). To assessthe effects of tumor promoters at the molecular level, we examined the effects of several tumor promoters/teratogens on protein synthesis in embryonic cells by twodimensional gel electrophoresis. All proteins synthesized by control cells were also synthesized in treated cells. However, all of the compounds identified as tumor promoters when tested in the Drosophila cells induced the synthesis of three low molecular weight proteins [22a, 22b, and 23 (molecular weights about 22,000 and 23,000)]. These proteins are found in only trace amounts in untreated cells. The proteins migrate with identical electrophoretic mobilities on two-dimensional gels as the heat-shock proteins hsp 22a.
200
BOURNIAS-VARDIABASIS
AND
TABLE
FLORES
I
Assay finding (% of controls)* Concentration (PM)
Compound
I.000 1.000 30 2.000 1,000 1,000 500 100 500 100 50 10 100 10 I 250 150 10
Control Anthrahn Anthraquinone 5-Azacytidine Biphenyl Cholesterol Coumarin Diphenylhydantoin
Griseofulvin Lithocholic
Acid
Tumor promoter” + +c NT +’
Mellitin 0.5 0.1 500 10,000 1,000 10
Phenobarbital Phenylalanine Saccharin Phorbol Esters TPA
0.1 0.01 10 10 + S-27 fraction IO 10 + S-27 fraction 10 10 + S-27 fraction 10 10 + S-27 fraction 10 100 100 + S-27 fraction
PDA PDB PDD 4-0-MeTPA 4-a-PDD Phorbol S-27 fraction
alone
+
+ + * -
Ganglia
Myotubes
100 95 88 58 89 100 33* 48* 85 34* 45* 44* 49* 95 103 93 30* 52* 89 71 92 100 66 100 68 31* 67 80 99 100 74 100 65 90 90 100 95 65 100 90 101
100 72 72 50* 88 96 36* 64 87 30* 49* 83 88 6* 27* 106 6* 18* 93 34* 50* 58 31* 94 110 46* 42* 53* 87 89 46+ 92 40* 98 50* 98 90 77 96 I15 105
Teratogen as indicated in in viva assays’ NT” NT NT NT
NT Nl. + +’ + NT
NT NT NT N-lNT NT
0 As designed by Diamond et nl. ( 1980). Sivak ( 1982). and Trosko et ul. ( 1982). *Drosophila embryonic cell cultures were prepared as described by Bournias-Vardiabasis and Tephtz (1982) and Bournias-Vardiabasis er (11. ( 1983a). The cells were allowed to differentiate at 26’C in the presence of the compound for 24 hr. Cultures were stained and numbers of differentiated myotubes and ganglia were counted with a Bausch and Lomb automated image analysis system. Results were expressed as a percentage of the number of myotubes and neuron clusters in control cultures prepared on the same day. A total of four dishes per trial was scored and each compound was tested in three or more separate trials. The addition ofthe compound. on three separate trials, resulted in the average number of myotubes and/or neuron clusters being significantly lower than those of the controls (Wilcoxon’s signed rank tested (Y < 0.05, indicated by *). Some of these data were taken from Boumias-Vardiabasis et al. (1983a). ‘From Shepard (1983). d NT, not tested.‘ ’ 5-Azacytidine has recently been classified also as a tumor promoter or syncarcinogen (Carr et al., 1984). fDiohenvlhvdantoin has been classified as a tumor nromoter bv Trosko et al. (1982) on the basis of in vitro assays rather-than-from in vivo studies. gOnly as a result of phenylketonuria. Phenylafanine is a constituent of the Drosophila cell culture medium. h f, weak tumor promoter or weak teratogen.
Drosophila
RESPONSE
TO
TUMOR
PROMOTERS
201
FIG. 1. Differentiated Drosophila cells. Embryonic cell culture, 24 hr. N, neuron clusters derived from divisions of neuroblasts; M, myotube derived from fusion of myocytes; A, axons. (A) Cultures treated with TPA. Note incomplete fusion of myocytes and the reduced number of neurons (only 8 cells are visible in this cluster). SEM. X 1250 (each bar = 10 Fm). (B) Controls, showing fully differentiated embryonic cells. Neuron clusters (N) usually have 16 or more cells. SEM. X 1250 (each bar = IO pm).
!b, and 23 (Buzin and Bournias-Vardianis. 1982, 1984). To simplify the results we lmpared only the 22,000-Da proteins of the
treated and control groups. 5-Azacytidi ne (300 WM) [identified as both a teratog :en (Shepard, 1983) and a tumor promoter or
202
BOURNIAS-VARDIABASIS
AND F’LORES
TABLE 2 INDUCTIONOFHEATSHOCKPROTEINSBYTUMORPROMOTERS
Compound 5-Azacytidine Biphenyl Coumarin Diphenylhydantoin Griseofulvin Phenobarbital TPA 4-0-MeTPA PDB Saccharin
Concentration (PM) 30 300 2000 1000 100 10 500 10 10 10 1000
Class TP/T b -d TPJT TP/T TP/T TP/T TP/T’ TPfTP/T’ TP/p
22-kDa protein (relative increase over control)’ 7 L’ 86 I 64’ 21’ 4 25 21 5.5 4.5 0.8’
’ Integrated optical densities for proteins 22a and 22b were determined from fluorographs with a Bausch and Lomb automated image-analysis system. Values were normalized to the integrated optical density of actin II on the same film. The normalized values for proteins 22a and 22b were averaged and the results are expressed as the factors of increase over normalized untreated control values. ’ TP, tumor promoter; T, teratogen. ’ Data taken from Buzin and Bournias-Vardiabasis ( 1984. 1982). d p, negative. ’ Designated by Huber and Brown (1983a) as teratogens in an in vitro rat assay.
syncarcinogen (Carr et al., 1984)] increased the levels of the induced proteins by a factor of 86 and as such elicited the strongest response. Coumarin (1000 PM), phenobarbital (500 PM), TPA ( 10 ,uM), and diphenylhydantoin ( 100 PM) ail produced levels of induction ranging between 2 1 and 25. Griseofulvin ( 10 PM), PDB ( 10 PM), and 4-0-MeTPA ( 10 j&M) showed induction over control of approximately a factor of 4. Biphenyl showed no relative increase over control levels. Saccharin (1000 PM), a weak tumor promoter (Diamond et al., 1980) and also a weak teratogen (Shepard, 1983) did not induce any of these 3 hsp in Drosophila cells (Table 2, Fig. 2). DlSCUSSION Little is known yet about the molecular basis of tumor promoter action. Promoters are thought to have the cell membrane as their primary target while initiators are thought to be damaging to DNA (Weinstein, 198 1).
These promoters have not been shown to be mutagenic in bacterial or mammalian cells (Sivak, 1982) so Ames-type tests have failed to identify them. This class of chemicals has been shown to exhibit a variety of specific effects, such as a change in the cell phenotype, an increase in protein-specific synthesis (not attributable to increased cell division), cell surface changes, decrease in acetylcholine receptor levels, and inhibition of terminal differentiation, among others (Diamond ef al., 1980). All of these perturbation events indicate an epigenetic mode of action. Teratogens also can be thought of as a large class of chemicals. some acting by epigenetic means, others by their DNA damaging potential. Furthermore, a causal mechanistic link has been proposed for both sets of agents: a number of tumor promoters and teratogens have been shown to interfere with normal cellcell communication and cell-cell interacand Bournias-Vardiabasis, tions (Buzin 1984; Trosko et al., 1982). It has been pro-
l>r~&la
RESPONSE
TO
TUMOR
203
PROMOTERS
b
d
FIG. 2. Protein synthesis in Dro.&zif~ primary embryonic cells treated with tumor promoters. Drosophila cells were plated at 3.5 hr after oviposition at 5 X IO6 cells per dish. The indicated agent was added at about 4 hr after oviposition and the cells were allowed to differentiate at 25°C for 18 hr in its presence. Cells were then labeled for I hr with [35S]methionine (200 nCi/ml) in the presence of the particular agent. Cells were solubilized and proteins separated by two-dimensional gel electrophoresis and visualized by fluorography. Agents used and exposure times (cpm X days) were TPA (480,000) (a); PDD (380,000) (b); Coumarin (362.000) (c): Control (498,000) (d); N is a non-heat shock protein that migrates close to protein 23. IEF. isoelectric focusing.
posed that if such an event happens in the adult organism, it results in uncontrollable growth (i.e., cancer): during embryonic development such interference would lead to abnormal development (i.e., teratogenesis).
Our results on the action of tumor promoters in the Drosophilu assay, indicate that this class of compounds inhibit muscle and or neuron specific embryonic differentiation. This observation confirms other reports
204
BOURNIAS-VARDIABASIS
showing that TPA in cell culture systems inhibits differentiation of Friend leukemia cells (Rovera et al., 1977), adipose conversion of mouse preadipose cells (Diamond et al., 1978). formation of neuroblastoma cells (Ishii et al., 1978) and myotube formation of myoblasts (Cohen et al., 1977: Dlugosz et al., 1983). In fact, it has been suggested that the ability of these compounds to inhibit terminal differentiation may be central to their mode of action as tumor promoters (Diamond et al., 1980). Three of the phorbol esters (PDA, PDB, and PDD) required incubation with an S-27 Drosophila microsome fraction before their teratogenic activity was detected. Little information about the metabolism of phorbol diesters is currently available; while hamster fibroblasts have been shown to be able to metabolize TPA (O’Brien and Diamond, 1978a,b), HeLa cell cultures show no such activity (Kreibich et al., 1974). Diamond et al. (1978) have shown that TPA lost its ability to induce ornithine decarboxylase (ODC) after 2-3 days of exposure to either normal or transformed hamster cells. They also found that PDB and PDD retained ODC-inducing activity several days later than did TPA. O’Brien and Saladik (1980) have also reported that PDD had a slower metabolism than TPA and again noticed that while human fibroblasts did not metabolize either of the tumor promoters, hamster fibroblasts metabolized TPA and PDD, although the latter was again metabolized much more slowly. In the Drosophila system it was ascertained that these particular phorbol esters require incubation with a microsomal system in order to inhibit cell differentiation. Further studies are underway to determine rates of metabolism for these phorbol esters. Furthermore, the assay is obviously exhibiting a high degree of sensitivity to this particular mode of action of tumor promoters. As such, it can be argued that it could be developed as a potential in vitro assay to identify potential tumor promoters. Such a need for in vitro assays has been put forward in recent
AND
F-LORES
reports (Diamond et al., 1980; Mondal et al., 1976; Sivak, 1982). Diamond et al. ( 1978) have also suggested that promotion may be easier to analyze in vitro than in vivo where absorption is a pulse rather than a continuous exposure to target cells. We see the Drosophila assay as providing several good endpoints for such assessment. Furthermore, techniques are currently being developed in our laboratory to assess the response of Drosophila cells to tumor promoters in terms of acetylcholine receptor levels. In chick embryo muscle cells, TPA has already been reported to cause both an increase in the rate of degradation of acetylcholine receptors and a decrease in the rate of receptor synthesis (Miskin et al., 1978). Preliminary experiments in Drosophila with teratogens already suggest that such modulations will also be present in the embryonic cells after exposure to tumor promoters. More importantly, Drosophila embryonic cells can provide us with the much needed information on molecular mechanisms of tumor promotion. It is not yet known why only certain types of cells show inhibition of differentiation while other cell types show an induction of differentiation. Cell cycling has been suggested as a possible primary requirement in the mechanism of inhibition of differentiation (Slaga et al., 1978). Although Drosophila embryonic cells are far removed from mouse skin epidermis, the evolutionarily conserved nature of many genetic and molecular mechanisms suggests that Drosophila studies may be useful in aiding our understanding of two-step carcinogenesis. They offer a very well-developed model for studying differentiation processes and thus analyzing inhibition of differentiation. The vast knowledge available on Drosophila genetics, development, and molecular biology may provide us with many clues to the role of genetic and biochemical variables in the process of carcinogenesis. Already the data obtained on heat shock protein induction suggest that at the molecular level, the cells’ response to tumor promot-
Drosophila RESPONSE
TO
ers is similar to their response to other “stress” agents. Boutwell ( 1974) and others have already proposed that altered gene expression is fundamental to the action of tumor promoters. This would mean that promoters can change the pattern of protein synthesis qualitatively or quantitatively. This ability of tumor promoters to modify gene activity by altering differentiation processes makes them valuable in studying the role of alterations in cellular differentiation in carcinogenesis as well as in normal differentiation. ACKNOWLEDGMENTS We wish to thank Carolyn Buzin and R. L. Teplitz for discussions and comments on the manuscript, and Marlene McCarrey and Rochelle Hoylman for skillful and patient preparation ofthe manuscript. Scanning electron micrographs were prepared in the Shared Instrumentation Laboratory using the Philips SEM 505 purchased with NIH funds. This work has been supported by a BRSG grant from the National Institutes of Health and the Gerald I. Parisi Research fellowship.
REFERENCES ASHBURNER, M., AND BONNER, J. (1979). The induction of gene activity in Drosophila by heat shock. Cell 17,241-254. BALMAIN, A. (1976). The synthesis of specific proteins in adult mouse epidermis during phases of proliferation and differentiation induced by the tumor promoter TPA and in basal and differentiating layers of neonatal mouse epidermis. J. Invest. Dermatol. 67,246-253. BOURNIAS-VARDIABASIS, N., TEPLITZ, R. L., CHERNOFF, G. F.. AND SEECOF, R. L. (1983a). Detection of teratogens in the Drosophila embryonic cell culture test: Assay of 100 chemicals. Teratology 28, 109- 122. BOURNIAS-VARDIABASQ N., BUZIN. C. H., AND REILLY, J. G. (1983b). The effect of 5-azacytidine and cytidine analogs on Drosophila melanogaster cells in culture. Wilhelm Roux’s Arch. Dev. Biol. 192,299-302. BOURNIAS-VARDIABASIS, N., AND FLORES, J. (1983). Drug metabolizing enzymes in Drosophila melanogas/cr. Teratogenicity of cyclophosphamide in vitro. Ter-
atog. Carcinog. Mutagen. 3,255-262. BOURNIAS-VARDIABASIS, N., AND TEPLITZ. R. L. ( 1982). Use of Drosophila embryo cell cultures as an in vitro teratogen assay. Teratog. Carcinog. Mutagen. 2,333-341. BOUTWELL. R. K. (1974). Function and mechanism of
TUMOR
PROMOTERS
205
promoters of carcinogenesis. CRC Crit. Rev. Toxicol. 2,419-447. BRESCH. H., AND ARENDT, U. (1978). Disturbances of early sea urchin development by the tumor promoter TPA (phorbol ester). Natwwissenschajien 65, 660662. BUZIN, C. H., AND BOURNIAS-VARDIABASIS, N. (1984). Teratogens induce a subset of small heat shock proteins in Drosophila primary embryonic cell cultures. Proc. Natl. Acad. Sci. USAgl, 4075-4019. BUZIN, C. H., AND BOURN~AS-VARDIABASIS, N. (1982). The induction of a subset of heat shock proteins by drugs that inhibit differentiation in Drosophila embryonic cell cultures. In Heat Shock: From Bacteria to Man (M. J. Schlesinger, M. Ashbumer, and A. Tissieres, eds.), pp. 387-394. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. BUZIN, C. H., AND SEECOF, R. L. (198 1). Developmental modulation ofprotein synthesis in Drosophilu primary embryonic cell cultures. Dev. Genet. (N.Y.) 2, 237252. CARR, B. I., REILLY. J. G., SMITH, S. S., WINBERG, C.. AND RIGGS, A. (1984). The tumorigenicity of 5-azacytidine in the male Fischer rat. Curcinogenesis 5, 1583-1590. COHEN. R., PACIFICI, M.. RUBINSTEIN, N., BIEHL, J., AND HOLTZER. H. (I 977). Effect of a tumour promoter on myogenesis. Nature (London) 266,538-540. DIAMOND, L., O’BRIEN, T. G., AND BAIRD, W. M. (1980). Tumor promoters and the mechanism of tumor promotion. Adv. Cancer Res. 32, l-74. DIAMOND, L., O‘BRIEN, T. G., AND ROVERA, G. (1978). Tumor promoters: Effects on proliferation and differentiation of cells in culture. Life 23, 1979-1988. DLUGOSZ, A. A., TAPSCOTT, S. J.. AND HOLTZER. H. ( 1983). Effects of phorbol 12-my&ate 13-acetate on the differentiation program of embryonic chick skeletal myoblasts. Cancer Res. 43,2780-2789. ELLINGER. M. S. (1982). Responses of amphibian embryos and blastomeres to a tumor-promoting phorbol ester. Cancer Res. 42,2804-28 12. GINDHART. T. D.. STEVENS, L.. AND COPLEY, M. P. (1984) Transformation and tumor promoter sensitive phosphoproteins in JB-6 mouse epidermal cells: One is also sensitive to heat stress. Carcinogenesis 5, 1 I 151121. HUBER, B. E., AND BROWN, N. A. (1983a). Tumor promoter actions on rat embryonic development in culture. Cancer Res. 43,5544-555 I. HUBER. B. E., AND BROWN, N. A. (1983b). 12-O-tetradecanoylphorbol13-acetate actions on macromolecular synthesis, omithine decarboxylase, and cellular differentiation ofthe rat embryonic visceral yolk sac in culture. Cancer Res43,5552-5559. ISHII, D. N., FIBACH, E., YAMASAKI, H., AND WEINSTEIN. I. B. (1978). Tumor promoters inhibit
206
BOURNIAS-VARDIABASIS
morphological differentiation in cultured mouse neuroblastoma cells. Science 200,556-559. KREIBICH, G., Suss, R., AND KINZEL, R. (1974). On the biochemical mechanism of tumorigenesis in mouse skin. V. Studies of metabolism of tumor promoting and non-promoting phorbol derivatives in vivo and in vitro. 2. Krebsforsch. 81, 135-149. MISKIN, R., EASTON, T. G., MAELICKE, A., AND REICH, E. (1978). Metabolism of acetylcholine receptor in chick embryo muscle cells: Effects of RSV and PMA. Cell 15, 1287-1300. MONDAL, S.. BRANKOW, D., AND HEIDELBERGER, C. (1976). Two stage chemical oncogenesis in cultures of C3H/lOT$ cells. Cancer Res. 36,2254-2260. NOVER, L. (ed.) (1984). Heat Shock Response ofEukar.yotic Cells. Springer-Verlag, Berlin. O’BRIEN. T. G.. AND DIAMOND, L. ( 1978a). A cell culture bioassay to analyze metabolism of phorbol diester tumor promoters. Cancer Res. 38,2567-2572. O’BRIEN, T. G., AND DIAMOND, L. (1978b). Metabolism of tritium-labeled 12-O-tetradecanoylphorbol13-acetate by cells in culture. Cancer Res. 38,2562-2566. O’BRIEN, T. G., AND SALADIK, D. (1980). Differences in the metabolism of 12-O-[3H]tetradecanoylphorbol13-acetate and [3H]phorbol12,13-dideconate by cells in culture. Cancer Res. 40,4433-4437. O’FARRELL. P. (1975). High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250,40074021. ROVERA, G., O’BRIEN, T. G., AND DIAMOND, L. (1977). Tumor promoters inhibit spontaneous differentiation of Friend erythroleukemia cells in culture. Proc. Nat/.
Acad. Sci. USA 74,2894-2898. SCHLESINGER, M. F., ASHBURNER, M., AND TISSIERES. A.. (eds.) (1982). Heat Shock: From Bacteria to Man.
AND
FLORES
Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. SEECOF, R. L., DONADY, J. J., AND TEPLITZ, R. L. (1973a). Differentiation of Drosophila neuroblasts to form ganglion-like clusters of neurons in vitro. CeN D@er. 2, 143-149. SEECOF, R. L.. GERSON, I., DONADY, J. J., AND TEPLITZ. R. L. (1973b). Drosophila myogenesis in vitro: The genesis of small myocytes and myotubes. Dev. Biol. 35,250-26 1. SHEPARD, T. H. ( 1983). Catalog of Terutogenic Agents Johns Hopkins Press. Baltimore. SHIBA, Y. (198 1). Laboratory action of 12-O-tetradecanoylphorbol13-acetate on bud production in Hydra.
Wilhelm Rowz’s Arch. Dev. Biol. 190.305-307. SIVAK. A. (1982). Evaluation of assay procedures for detection of tumor promoters. Mutat. Res. 98,377-387. SLAGA. T. J.. SIVAK, A., AND BOUTWELL, R. K. (eds.) (1978). Carcinogenesis Mechanisms of Tumor Production and Cocarcinogenesis, Vol. 2, Raven Press, New York. TROSKO, J. E., CHANG, C. C., AND NETZLOF. M. (1982). The role of inhibited cell-cell communication in teratogenesis. Teratog. Carcinog. Mutagen. 2, 3 I 45. WEINSTEIN, I. B. (198 1). Current concepts and controversies in chemical carcinogenesis. J. Supramol. Struct. Cell. Biochem. 17,99- 120. WILSON, J. G. (1977). Current status of teratology. In Handbook c$ Teratology (J. G. Wilson and F. C. Fraser, eds.). pp. 309-355. Plenum, London. YOTTI, L. P.. CHANG, C. C.. AND TROSKO, J. E. ( 1979). Elimination of metabolic cooperation in Chinese hamster cells by a tumor promoter. Science 206. 10891091.