Adenyl Cyclase in Normal and Transformed Fibroblasts in Tissue Culture ACTIVATION BY PROSTAGLANDINS

Abstract The activity of the enzyme adenyl cyclase has been measured in various lines of mouse, hamster, and rat fibroblasts maintained in tissue culture. All cell lines have detectable adenyl cyclase activity. With respect to the activity of the enzyme in the presence of fluoride ions: (a) mouse embryo fibroblasts and L cells have relatively high levels of enzyme, whereas 3T3 cells derived from mouse embryo fibroblasts have low levels; (b) 3T3 cells transformed by SV40 and MSV/MuLV viruses have elevated enzyme levels, but 3T3 cells transformed by polyoma virus have slightly decreased levels; (c) hamster embryo fibroblasts and the established BHK line have relatively high levels and transformation of BHK cells by polyoma virus only slightly decreases enzyme activity; (d) rat embryo fibroblasts have the highest enzyme activity found. With respect to stimulation of the enzyme by prostaglandins, prostaglandin E1 is the most effective when tested on L cells, followed in effectiveness by prostaglandin E2, prostaglandin F2α, and prostaglandin B1. Prostaglandin A2 produces only a small response. Two cell lines derived from 3T3 cells by transformation with polyoma or SV40 virus are unresponsive to all prostaglandins tested. With most lines, adenyl cyclase activity remained constant during culture for 2 to 3 months. However, in rat embryo fibroblasts and SV40 (3T3) cells, the fluoride-stimulated enzyme levels (but not basal levels) fell with successive passages.


IRA PASTAN
From the Laboratory of Molecular Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20014 SUMMARY The activity of the enzyme adenyl cyclase has been measured in various lines of mouse, hamster, and rat fibroblasts maintained in tissue culture. All cell lines have detectable adenyl cyclase activity.
With respect to the activity of the enzyme in the presence of fluoride ions: (a) mouse embryo fibroblasts and L cells have relatively high levels of enzyme, whereas 3T3 cells derived from mouse embryo fibroblasts have low levels; (b) 3T3 cells transformed by SV40 and MSV/MuLV viruses have elevated enzyme levels, but 3T3 cells transformed by polyoma virus have slightly decreased levels; (c) hamster embryo fibroblasts and the established BHK line have relatively high levels and transformation of BHK cells by polyoma virus only slightly decreases enzyme activity; (d) rat embryo fibroblasts have the highest enzyme activity found.
With respect to stimulation of the enzyme by prostaglandins, prostaglandin Ei is the most effective when tested on L cells, followed in effectiveness by prostaglandin Ez, prostaglandin Fza, and prostaglandin Bi. Prostaglandin AZ produces only a small response.
Two cell lines derived from 3T3 cells by transformation with polyoma or SV40 virus are unresponsive to all prostaglandins tested. With most lines, adenyl cyclase activity remained constant during culture for 2 to 3 months. However, in rat embryo fibroblasts and SV40 (3T3) cells, the fluoride-stimulated enzyme levels (but not basal levels) fell with successive passages.
The ubiquitous occurrence and the diverse effects of cyclic adenosine 3',5 '-monophosphate (1, 2) have prompted the consideration of a possible role for this molecule in the regulation of cellular growth and morphology.
Indeed, previous studies have indicated that cyclic AMP1 or N6,0Z'-dibutyryl cyclic AMP inhibits the growth of various lines of transformed fibroblasts and changes the abnormal morphology of these cells towards normal (3). The intracellular level of cyclic AMP is controlled in part by the activity of the enzyme adenyl cyclase, which catalyzes the formation of cyclic AMP from ATP.
The prostaglandins, which are of widespread natural occurrence (4, 5), are known to stimulate adenyl cyclase activity in vitro in multiple systems (6)(7)(8) and may possibly function in vivo to regulate intracellular cyclic AMP levels (5, 6,[9][10][11][12][13]. Therefore, as part of an investigation into the possible role of altered cyclic AMP metabolism in cancer, we have studied the levels of adenyl cyclase in several normal and transformed cell lines maintained in tissue culture.
The activity of the enzyme was measured in the basal state and in the presence of fluoride or various prostaglandins.
The mouse, rat, and hamster embryo fibroblast secondary cultures were planted at 1.5 x 106 cells per loo-mm plastic tissue culture dish. The rat and mouse embryo fibroblasts were split 1: 2 at weekly intervals during the period that they were maintained in culture. The remaining cells were planted at approximately lo6 cells per loo-mm plastic tissue culture dish and were harvested 2 to 4 days after planting.
All cell lines were incubated at 37" in a 5% CO2 atmosphere.
During the course of the study, the activity of the adenyl cyclase did not vary during repetitive transfer of the cells except as noted in the text.
Preparation of Enzyme-The medium was removed, and the cells were washed four times with an ice-cold solution containing 0.15 M NaCl, 0.01 M sodium phosphate (pH 7.4) followed by four washes with ice-cold homogenizing medium (sucrose, 0.33 M; dithiothreitol, 1 mM; Tris-HCl buffer, pH 7.8, 50 mM; MgC12, 1 MM). All subsequent steps were carried out at O-4" unless otherwise specified.
The cells were scraped from the culture dish wit,h a clean rubber stopper, and then homogenized in 0.5 to 2.0 ml of homogenizing medium (protein 2.5 to 10 mg per ml) in a glass Dounce homogenizer (tight pestle; 10 strokes). The resulting homogenate was used directly as the enzyme component in the assay procedure.
Enzyme Assay-Adenyl 4 pg per ml of pyruvate kinase (rabbit muscle), and 40 mM Tris-HCl, pH 8.3. When fluoride was added, its concentration was 8 mM. Other substances tested for effect on enzyme activity were present as specified.
The reaction was started by the addition of [oL-~~P]ATP (2 to 3 X lo6 cpm), and incubation was carried out for 10 min at 37" with shaking.
The reaction was terminated by the addition of 0.5 ml of an ice-cold solution containing a large excess of ATP, cyclic AMP, and cyclic [3H]AMP (26). Cyclic AMP was isolated from other 32P labeled components by chromatography on Dowex AG 5OW-X8 (100 to 200 mesh), followed by two BaS04 precipitations (recoveries averaged 30 to 40% in all assays). The supernatants were mixed with Bray's solution and their content of radioactivity measured in a liquid scintillation spectrometer.
The blank for each experiment was prepared by terminating the incubation at 0" before the addition of [&*P]-ATP.
The blank values were subtracted from each experimental value.
Protein measurements were made by the method of Lowry et al. (27) with bovine serum albumin as the standard.
The rate of cyclic AMP production was linear for at least 10 min and proportional to protein concentration over a range of 0.4 to 2.5 pg per ml. In all the five cell lines studied, the fluoridestimulated enzyme has a pH optimum at approximately pH 8.3, as measured before the reaction was begun (Fig. 1) ; the final pH following the 10..min incubation was approximately pH 8.1. We chose a Mg* concentration equal to that of ATP since this value was optimal in one other adenyl cyclase system Issue of September 25,1971 C. V. Peery, G. S. Johnson, and I. Pastan 5787 1800  Results are expressed as picomoles of cyclic AMP formed per mg of protein per 10 min.
, and doubling the magnesium concentration had no effect on the assay. Raising the potassium concentration to 30 rnM was also without effect. The addition of P-enolpyruvate and pyruvate kinase improved the amount of cyclic AMP formed by about 20% and therefore was routinely added. Each experimental value reported is the mean of closely agreeing duplicate values and each experiment was repeated at least twice and frequently 5 to 10 times. In a typical experiment, 1 pmole of cyclic AMP formed per mg of protein per 10 min corresponded to 1 cpm of 32P. The adenyl cyclase from both lines, when grown in monolayer, had low basal activity (75 and 65 pmoles per mg of protein per 10 min) and responded in a similar fashion to fluoride and prostaglandin E1. We used line B in the remainder of our studies.
Cells taken directly from spinner culture and cells which were allowed to grow to high densities in monolayer also had similar adenyl cyclase activities.
Neither glucagon (10e5 M) nor epinephrine (10m5 M) produced significant stimulation of the enzyme in vitro. Fig. 2 shows the response of L-929 cell adeayl cyclase to varying concentrations of several prostaglandins.
The enzyme responded to aslittle as 0.1 pg per ml of PGEI and a 20-fold rise was observed at 5 pg per ml. PGEz was much less effective and large concentrations of PGBl and PGF2, (50 to 100 pg per ml) produced only a 5-fold stimulation.
PGAQ was the least effective of all. In the presence of sodium fluoride (8 m&f), PGEl (5 pg per ml) produced no additional increase in enzyme activity. In a control experiment, dihydroxystearic acid (5 and 50 pg per ml) did not alter enzyme activity.
Untransformed Mouse Cells-Mouse embryo fibroblasts and their derivatives are cells that are also commonly used for tissue culture work.
Enzyme levels in NIH-Swiss mouse embryo fibroblasts (measured after 2 to 3 transfers) are shown in Table  II. The basal level is somewhat higher than L-929 cells, but the activity in the presence of fluoride is comparable.
However, in 3T3 cells derived from either NIH-Swiss or Balb/c mouse embryo fibroblasts, a 750/, decrease in adenyl cyclase was found.
SV-40 and MXV/MuLV Transformation-Cell lines derived from 3T3 cells by transformation with SV-40 or MSV/MuLV had fluoride-stimulated adenyl cyclase levels much higher than the 3T3 cells and close to that of the mouse embryo fibroblasts (Table II) was also unresponsive to PGE1. During the 2-month period that these cells were maintained in our laboratory, their level of fluoride-stimulated adenyl cyclase decreased, but basal levels remained unchanged.
Polyoma Transformation-The effect of polyoma transformation on enzyme levels in mouse and hamster cells is shown in Table III.
In line Py-11 (3T3), polyoma transformation results in a rather small decrease in fluoride-stimulated adenyl cyclase activity, and a loss of responsiveness to PGE1.
However, when line Py-11 (3T3) was subsequently transformed by SV-40 virus (SV-Py-11 (3T3)), the enzymatic activity measured in the presence of fluoride was elevated and the enzyme became very responsive to PGE1. Btik (28) has previously reported that BHK cells transformed by polyoma virus had about a 50% decrease in basal adenyl cyclase activity; however, he did not assay the enzyme in the presence of fluoride.
Prostaglandins-The response of the adenyl cyclases from several mouse cell lines to increasing concentrations of PGEl is shown in Fig. 3. Relative to fluoride-stimulated levels, good responses are seen in mouse embryo fibroblasts and 3T3 cells with a lessened response to PGEr noted in the MSV/MuLVtransformed line. No significant response was found in either Py-11 (3T3) or SV-40 (3T3) cells. It is apparent that transformation can result in a decreased or absent sensitivity of the enzyme to PGEI.
The response of various mouse cell lines to prostaglandins other than PGEl is shown in Table IV. A small response to PGEs is observed with mouse embryo fibroblasts and 3T3 cells, and a small response to PGBl with mouse embryo fibroblasts. NO response to PGA? is found when tested on mouse embryo fibroblasts, SV-40 (3T3), MSV/MuLV (3T3), or Py-11 cells (data not shown).
Rat Cells-We measured adenyl cyclase activity in XC cells, a transformed rat cell line originally isolated from a rat inoculated with Rous sarcoma virus (17), and in rat embryo fibroblasts which were carried in culture for as long as 11 transfers (Table  V). The adenyl cyclase activity of the rat embryo fibroblasts measured either in the presence of fluoride or PGE1 decreased over this period, but a substantial response to both substances remained.
The rather high basal activity was unchanged. Issue of September 25, 1971 C. V. Peery, G. S. Johnson, and I. Pastan 5789 XC cells also had a high basal adenyl cyclase activity but the enzyme responded poorly to fluoride and less well to PGE1.

DISCUSSION
When various transformed and normal fibroblasts are treated with cyclic AMP, N6, 02'-dibutyryl cyclic AMP, or theophylline, their growth is slowed, and the cells become more elongated and adhere more tightly to the substratum (3,29). Similar cell elongation was found with treatment of Chinese hamster cells (30). These findings suggest that cyclic AMP normally regulates cell growth and morphology.
Since intracellular cyclic AMP levels are determined, in part, by the levels of adenyl cyclase, we measured the activity of adenyl cyclase in various normal and transformed cell lines.
these cells by SV-40 restored prostaglandin responsiveness. The detection of cell lines unresponsive to PGEl suggests that the prostaglandin receptor is not present or that it has been structurally altered.
A similar model has been utilized by Pennington et al. (34) to explain the divergence in epinephrine responsiveness of adenyl cyclase in liver and Morris hepatoma cells. The isolation of cell lines with altered responses to prostaglandins promises to be very useful for biochemical studies on the mechanism of prostaglandin action, and these lines also offer another marker for use in cell fusion studies. Further, a number of SV-40 and polyoma transformed lines need to be studied to ascertain how frequently this loss of responsiveness to PGEl occurs, and the responsive lines need to be cloned to determine whether there is a mixture of responsive and unresponsive cells.
The simplest model of the adenyl cyclase enzyme complex divides the enzyme into a catalytic subunit which is exposed to the interior of the cell and one or more regulatory subunits which are exposed to the cell exterior (1, 2). Presumably, fluoride, which activates the enzyme only in broken cell preparations, acts on the catalytic subunit, whereas prostaglandins act on one or more regulatory subunits.
The effects of transformation on the activity of the enzyme will be considered in terms of this model.

Since prostaglandin
El activates adenyl cyclase in homogenates of L-929 cells, this agent would be expected to elevate cyclic AMP levels in intact cells, and the elevated cyclic AMP levels, in turn, would be expected to restore morphology towards normal and to inhibit growth.
In recent studies, we have confirmed both of these predictions2v a Acknowledgment-We acknowledge the expert technical help of Mrs. Patricia Middleton.
The 3T3 and 3T6 cell lines are stable lines derived from mouse embryo fibroblasts (23). 3T3 cells were selected for their tendency not to overgrow, whereas the 3T6 cells and the parent mouse embryo fibroblasts do overgrow.
We found that mouse embryo fibroblasts had high adenyl cyclase activity when measured with or without fluoride, or with PGEI.
In contrast, two different lines of 3T3 cells had low basal, fluoride, and PGEl-stimulated adenyl cyclase activity.
The 3T6 cells had intermediate levels of enzyme activity.
Makman (31) has previously reported somewhat higher fluoride-activated levels in 3T6 cells than those found in this study.
Thus, the selection for contact-inhibited cells appears to result in cells with low adenyl cyclase activity. The low adenyl cyclase in 3T3 cells is probably due to a low level of catalytic subunit activity, since both basal and fluoride-dependent activities are diminished.
However, this decrease in adenyl cyclase activity does not appear to be simply a property of established cell lines since the established BHK-21 hamster cell line has high adenyl cyclase activity (Table III).
Virus Transformation-Transformation of 3T3 cells by various oncogenic viruses leads not only to cells with altered growth properties (19), but also to changes in cell membrane proteins and glycoproteins (32,33). Therefore, it seemed possible that transformed cells might also have alterations in their adenyl cyclase levels. We found that 3T3 cells transformed by SV-40 and MSV/MuLV viruses did have an elevation in fluoridestimulated activity, whereas both 3T3 and BHK cells transformed by polyoma virus had a slight decrease in activity.