Cancer Letters, 67 (1992) 139 - 144 Eisevier Scientific Publishers Ireland Ltd.
139
Differential responsiveness to agents which stimulate CAMP production in normal versus neoplastic mouse lung epithelial cells Carol A. Lange-Cartera,
“Molecular Colorado,
and Enuironmental Boulder,
Sciences Center,
(Received (Accepted
Kurt A. Dromsb, Jeffrey J. Vuillequeza and Alvin M. Malkinson”
Toxicology
CO. 80309-0297
Program
Cancer Center, School of Pharmacy, University of of Cell Biology and Anatomy, Texas Tech University, Health
and Colorado
and bDepartment
3601 4th St., Lubbock,
‘FX 79430
(USA)
16 August, 1992) 18 September, 1992)
Summary We examined the responsiveness of normal and neoplastic lung cells to agents which stimulate CAMP production. While their basal CAMP levels were similar, spontaneous in vitro
transformant E9 cells and PCC4 cells produced much
tumor-derived less CAMP in
response to 1 PM isoproterenol compared to non-tumorigenic Cl0 cells derived from normal mouse lung epithelium. lodocyanopindolol binding studies indicated that both neoplastic lines contained fewer fl-adrenergic
receptors
than normal Cl0 ce!ls. When recep-
tors were bypassed via treatment with 10 pM cholera foxin, the pattern of CAMPresponsiveness was reversed; both neoplasfic cell tines produced more CAMP than Cf 0 cells. Direct stimulation of adenylate cyclase with 100 PM forskolin greatly increased CAMP concentrations in all three cell lines. These anomalies at both the receptor and G-protein levels in neoplastic lung epithelial cells may
contribute
to their deregulated
growth.
Correspondence to: Alvin M. Malkinson, School of Pharmacy, University of Colorado, Boulder, CO 80309-0297, USA.
0304-3835/92/$05.00 Printed and Published
0 1992 Elsevier Scientific Publishers in Ireland
Keywords: adenylate cyclase; beta-adrenergic receptor; GTP-binding proteins; lung epithelial cell lines; growth regulation Introduction Neoplastic growth is maintained by several mechanisms, including over-production of positive growth factors and/or an overexuberant cellular response to them and by insensitivity to negative growth factors. CAMP is an important modulator of normal cellular growth and differentiation [l]. Addition of CAMP to cultured cells has demonstrated both positive [2] and negative [3] effects on proliferation. The mechanism of this dichotomy is unknown. CAMP may interact with other factors (i.e. growth factor receptors, oncogene or tumor-supressor gene products) which control growth [4]. The response to CAMP would then depend on the cell- or pathological statespecific milieu of factors driving growth and/or differentiation events. Chemical analogs of CAMP can induce differentiation and inhibit the growth of many neoplastic cell lines and tumors, including human colon and mammary carcinoma cell lines [5] and rat mammary tumors in vivo [6]. However, some tumors are Ireland Ltd
Isoproterenol
Receptor
Cholera Toxin
L
G-protein
Forskolin
L
Adenylate Cyclase
L
CAMP
ICYP Scheme
1.
unresponsive to CAMP treatment, probably due to aberrations in their intracellular CAMP receptors. Alterations in the CAMP signal transduction pathway are associated with mouse lung neoplasia. Susceptible strains of inbred mice develop primary adenomas and adenocarcinemas spontaneously and in response to carcinogens [7]. These tumors are an important model system of human bronchiole-alveolar carcinoma, that subclass of primary human lung adenocarcinoma in which incidence is increasing most rapidly [8]. A cell line called Cl0 was derived from one of the precursor cell types of these tumors, the alveolar type II pneumocyte [9]. A spontaneous transformant of similar histologic origin, called E9 [lo] and a cell line derived from a lung tumor, called PCC4 [ 111, can be compared to Cl0 to study mechanisms of their resistance to negative growth signals. These tumorigenic cells express decreased levels of the type I isozyme of CAMP-dependent protein kinase (PKA) [ 121. In addition, type II PKA regulatory subunits from mouse lung tumors lack high-affinity CAMP-binding sites [13]. Droms et al. [14] reported decreased adenylate cyclase responsiveness of neoplastic lung cells to isoproterenol. Herein, we address the mechanism of this decreased responsiveness using agents which interact with or bypass the fl-adrenergic receptor, specifically stimulating each step in the CAMP synthesis pathway, as shown in Scheme 1.
Materials and Methods
Cell culture The non-tumorigenic, immortal cell line (hereafter referred to as ‘normal’) ClO, is contact inhibited, while tumorigenic E9 and PCC4 cells do not display density-dependent inhibition of growth at confluence [ 10,111. C 10 and E9 cells were maintained using CMRL 1066 (Irvine) medium and PCC4 cells using McCoys 5A (Gibco) in Corning loo-mm tissue culture plates. Media were supplemented with 10% fetal bovine serum and cells were grown in a humidified atmosphere of 5% COz/95% air. CAMP radioimmunoassay
The CAMP concentration in ether-extracted soluble fractions of cell Iysates was assayed using the cAMP[‘*~I] assay system (Amersham) as described [14]. Total CAMP accumulation was normalized to the total amount of cellular protein contained in the insoluble fraction as quantified by the method of Lowry et al. [15]. Time courses indicated maximal stimulation of CAMP production at 1 min following exposure to isoproterenol, 120 min using cholera toxin and 5 min using forskolin. ‘251-labeled ICYP binding assay &Adrenergic receptors were quantified in ClO, E9 and PCC4 cell membranes by Scatchard analysis of 1251-labeledICYP binding as described [16], except that 10 - 30 pg cell membrane protein was used per assay tube. A
1-pM quantity of unlabeled DL-propranolol was included in some reactions to measure non-specific ‘251-labeleing ICYP binding. Scatchard analysis was done using the Ligand Binding Analysis Program [ 171.
0.8 k .
m
Results in isoproterenol-treated lung cells To confirm that normal and neoplastic mouse lung epithelial cells were differentially responsive to agents which stimulate CAMP production, non-tumorigenic Cl0 cells and tumorigenic E9 and PCC4 cells were treated for 1 min with 1 PM isoproterenol (Fig. 1). Normal Cl0 cells produced more CAMP in response to isoproterenol than either in vitro transformant E9 cells or tumor-derived PCC4 cells (Fig. 1). This normal-neoplastic difference is considerably magnified compared to that reported earlier [14], in which a 5-min exposure to isoproterenol was assayed.
0.4
CAMP responsiveness
normal
and neoplastic
E9
Cl0
Cell
PCC4
Line
Fig. 1. CAMPProduction in isoproterenol treated normal and neoplastic lung cells. A11cell lines had similar basal levels of CAMP. Duplicate cultures of each cell line were treated for 1 min with (striped bars) or without (solid bars) 1 mM isoproterenol, a /3-adrenergic agonist. Normal Cl0 cells were more responsive to isoproterenolstimulated CAMPproduction than the two tumorigenic cell lines, E9 and PCC4. Data represent the’ mean of duplicate measurements from a typical experiment l the range.
0.0
10
U
“‘ICYP
20
Bound
30
40
(fmol/mg)
Fig. 2. P-Adrenergic receptor binding in normal and neoplastic cell membranes. Cell membranes were isolated and subjected to [‘251]-labeled ICYP binding assays as described in Methods. Normal Cl0 (B) cells contained more fl-adrenergic receptors than neoplastic PCC4 (A) and E9 (UJ cells. Points represent the mean of duplicate determinations from a typical experiment. A comparison of B,,, from several ments indicated that P < 0.001.
independent
experi-
To examine the mechanism of decreased isoproterenol responsiveness in the tumorigenie cell lines, binding studies were done using ‘251-labeled ICYP, a ligand with high affinity and specificity for the fl-adrenergic receptor [18]. Scatchard analysis indicated fewer /3-adrenergic receptors in neoplastic E9 and PCC4 cell membranes (B,, = 6.5 and 12 fmol/mg protein, respectively) than in normal Cl0 cell membranes (B,,, = 37 fmol/mg protein) (Fig. 2). Membranes isolated from all three cell lines demonstrated similar binding affinities for 1251-labeled ICYP (Kd = 20-40 PM). CAMP responsiveness forskolin-treated cells
in cholera toxinnormal and neoplastic
and lung
To determine whether further alterations downstream of the @-adrenergic receptor contributed to differential CAMP responsiveness between normal and neoplastic cells, each cell line was treated with cholera toxin which
to 10 pM cholera toxin, while normal Cl0 cells were minimally responsive to this agent (Fig. 3A). A 100~PM quantity of forskolin superinduced CAMP production in E9 cells to ZOO-fold over control levels, while Cl0 and PCC4 cells responded similarly, undergoing approximately lo-fold increases (Fig. 3B). Discussion PCC4
E9
2500 -
B
20001500 lOOO500 0
t_-l E9
Cl0
Cell
PCC4
Line
Fis. 3. CAMP Production in cholera toxin (A) and forskolin (B) treated normal and neoplastic lung cells. (A) duplicate cultures of each cell line were treated for 120 min with (striped bars) or without (solid bars) 10 pM cholera toxin. Neoplastic E9 and PCC4 cell lines were more responsive to cholera toxin-stimulated CAMP production compared to normal Cl0 cells. (B) Duplicate cultures of each cell line were treated for 5 min with (striped bars) or without (solid bars) 100 mM forskolin, which stimulates adenylate cyclase direclly. E9 cells were extremely sensitive to forskolin-stimulated CAMP production, while Cl0 and PCC4 cells demonstrated similar responsiveness. Data represent the mean of duplicate measurements from a typical experiment * the range.
constitutively activates Gsar via ADPribosylation [ 191, or with forskolin, a direct activator of the adenylate cyclase catalytic subunit [ZO] (Fig. 3). Both neoplastic cell lines produced large amounts of CAMP in response
Alterations in the receptor-coupled adenylate cyclase system characterize many tumors (for review see Ref. 21). Interestingly, the tumorigenic mouse lung epithelial cells examined herein contain defects at both the fladrenergic receptor and Gs transducer element levels. Both neoplastic cell lines express fewer /3-adrenergic receptors than normal Cl0 cells. Our cholera toxin data suggests that functional alterations also reside in the stimulatory G-protein complex (Gs) coupled to adenylate cyclase. Droms et al. [14,22] identified decreased photoincorporation of 8N3-I”/32P]GTP into GSCX proteins in several tumorigenic mouse lung cell lines, including E9 and PCC4; this was not due to decreased Gscr protein concentration as measured by immunoblotting experiments [14]. The alteration/mutation which causes this functional change in tumor cell GSCYmay also confer increased cholera toxin responsiveness. Alternatively, Gs in normal Cl0 cells may be negatively controlled by endogenous factors lacking in the neoplastic cells and is therefore less affected by cholera toxin. The PX subunits of the Gs complex can attenuate o-subunit activity [23]. Perhaps the ADP-ribosylation site on GSCYin Cl0 cells is blocked or posttranslationally modified. In neoplastic PCC4 and E9 cells, this negative control over Gs may be lost, allowing its constitutive activation. In B16 mouse melanoma cells, changes in the cellular growth rate are correlated with functional changes in Gscu [24]. In cell types in which CAMP is mitogenic, alterations in GSCX can induce hyperplasia and contribute to neoplastic transformation (for review see Ref. 25). Mutations that result in
143
GTPase-deficient and therefore constitutively active, Gscr subunits are associated with tumors of the pituitary [26] and thyroid [27] glands. Modifications of Gscr have also been associated with transformation of tissues in which CAMP is not mitogenic, but may influence neoplastic cell growth. GTPdependent coupling of adrenergic receptors to adenylate cyclase is modified in several hepatomas compared to ‘normal liver tissue [28]. Gsar proteins isolated from B16 melanoma cells with high metastatic potential stimulated adenylate cyclase to a greater extent in reconstituted membranes as compared to Gscr obtained from cells with low metastatic potential [29]. Each lung cell line produced large amounts of CAMP in response to forskolin, indicating similarities in the adenylate cyclase catalytic subunit in normal and neoplastic cells. The superinduction of CAMP levels by forskolin in E9 cells compared to C 10 and PCC4 cells may reflect their altered Gscr proteins [14] and cholera toxin hyper-responsiveness (Fig. 3A). Increased forskolin-responsiveness has been attributed to altered Gs function in B16 melanoma cells [29] and human pituitary tumors [30]. That PCC4 cells do not respond similarly to E9 cells suggests that these two cell lines may contain Gs alterations/mutations which are qualitatively different. Alternatively, E9 cells may express increased levels of adenylate cyclase. That each neoplastic cell line produced high concentrations of CAMP in response to both agents acting downstream of @-adrenergic receptor stimulation indicates intact coupling of Gs to adenylate cyclase and suggests that the decreased isoproterenol responsiveness of these cells is due entirely to their lowered @-adrenergic receptor expression. Multiple anomalies in the CAMP signalling pathway may diminish the ability of CAMP to regulate growth. One role of CAMP in affecting tumor cell growth may be to modulate, rather than mediate, the activity of various oncogenes and growth factors via phosphorylation events (for review see Ref. 4). Addition of
cholera toxin to human small-cell lung cancer cell lines decreased their responsiveness to several diverse mitogenic stimuli [31]. CHO cell growth was inhibited by CAMP, while o-src expressing, RSV-transformed CHO cells were CAMP-resistant. Cholera toxin was mitogenic to Swiss 3T3 cells [32], but inhibited the growth of ras-transformed Swiss 3T3 cells [33]. Aberrations associated with mouse lung tumorigenesis include K-ras proto-oncogene activation [34] and p53 tumor suppressor gene mutation [35]. It will be interesting to dissect the role of CAMP signal transduction in cells containing various combinations of activated oncogenes and/or mutated tumor suppressor genes. Acknowledgments
We would like to thank Dr. G.L. Johnson (National Jewish Center for Immunology and Respiratory Medicine, Denver) for helpful comments on this manuscript. We are grateful to Ms. S. Leah Butler for excellent technical assistance. This work was supported by USPHS Grant ES02370. References Beebe, S.J. and Corbin, J.D. (1986). Cyclic nucleotidedependent protein kinases. In: The Enzymes, Vol. 17, pp. 43- 111. Editor: P.D. Boyer. Academic Press, Orlando, FL. Dumont, J.E., Jauniaux, J.C. and Roger, P.P. (1989). The cyclic AMP-mediated stimulation of celf proliferation. Trends Biochem. Sci., 14, 67-71. Russell, D.H. (1978). Type I cyclic AMP-dependent protein kinase as a positive effecter of growth. Adv. Cyclic Nut. Res., 9, 493-506. Gottesman, M.M. and Fleischmann, R.D. (1986). The role of CAMP in regulating tumor cell growth. Cancer Surv., 5, 291- 308. Taglfaferri, P., Katsaros, D., Clair, T., Ally, S., Tortora, G., Neckers, L., Rubalcava, B., Parandoosh, Z., Chang, Y., Revankar, G.R., Crabtree, G.W., Robins, R.K. and Cho Chung, Y.S. (1988) Synergistic inhibition of growth of breast and colon humena cancer cell lines by siteselective cyclic AMP analogues. Cancer Res., 48, 1642 - 1650. Cho-Chung, Y.S., Clair, T. and Shepheard, C. (1983). Anticarcinogenic effect of N6,02-dibutyryl cyclic adenosine 3’,5’-monophosphate on 7,12-dimethy-
144
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
benz(a)anthracene mammary tumor induction in the rat adenosine 3’5’to cyclic its relationship and monophosphate metabolism and protein kinase. Cancer Res., 43, 2736 - 2740. Malkinson, A.M. (1991). Genetic studies on lung tumor susceptibility and histogenesis in mice. Environ. Health Perspect., 93, 149 - 159. Gazdar, A.F. and Linnoila, R.I. (1988). The pathology of lung cancer - changing concepts and newer diagnostic techniques. Semin Oncol., 55, 215 - 225. Smith, G.J., LeMesurier, S.M., De Montfort, M.L. and Lykke, A.W.J. (1984). Development and characterization of type 2 pneumocyte-related cell lines from normal adult mouse lung. Pathology., 16, 401-405. Smith, G.J. and Lykke, A.W.J. (1985). Characterization of a neoplastic epithelial cell strain derived from dexamethazone treatment of cultured normal mouse type 2 pneumocytes. J. Pathol., 147, 165- 172. Nicks, K.M., Droms, K.A., Fossli, T., Smith, G.J. and Malkinson, A.M. (1989). Altered function of protein kinase C and CAMP-dependent protein kinase in a cell line derived from a mouse lung papillary tumor. Cancer Res., 49, 5191- 5198. Lange-Carter, C.A., Fossli, T., Jahnsen, T. and Malkinson, A.M. (1990). Decreased expression of the type I isozyme of CAMP-dependent protein kinase in tumor cell lines of lung epithelial origin. J. Biol. Chem., 265, 7814- 7818. Malkinson, A.M. and Butley, M.S. (1981). Alterations in cyclic adenoxine 3’,5’-monophosphate-dependent protein kinases during normal and neoplastic lung development. Cancer Res., 41, 1334- 1341. Droms, K.A., Haley, B.E., Smith, G.J. and Malkinson, A.M. (1989). Decreased SN,-[X-32P]GTP photolabeling of Gso in tumorigenic lung epithelial cell lines: association with decreased hormone responsiveness and loss of contact-inhibited growth. Exp. Cell. Res., 182, 330 - 339. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951). Protein measurement with the folin phenol reagent. J. Biol. Chem., 193, 265-275. Valverius, P., Hoffman, P.L. and Tabakoff, B. (1987). Effect of ethanol on mouse cerebral cortical betaadrenergic receptors. Mol. Pharmacol., 32, 217 - 222. Munson, P.J. and Rodbard, D. (1983). LIGAND: A versatile computerized approach for characterization of ligandbinding systems. Anal. Biochem., 107, 220 - 239. Burgisser, E. and Lefkowitz, R.J. (1984) Brain Receptor Methodologies, Part A, pp. 229-253. Editors: P.J. Marangos, I.C. Cambell and R.M. Cohen. Academic Press, Orlando, FL. Moss, J. and Vaughan, M. (1988). ADP-ribosylation of guanyl nucleotide-binding regulatory proteins by bacterial toxins. Adv. Enzymol., 61, 303-379. Downs, J.R.W. and Aurbach, G.D. (1982). The effect of forskolin on adenylate-cyclase in S49-wild type and CYC (-)-cells. J. Cyclic Nucleotide Res., 8, 235-242. Hunt, N.H. and Martin, T.J. (1980). Hormone receptors
22
23
24
25
26
27
28
29
30
31
32 33
34
35
and cyclic nucleotides - significance for growth and function of tumors. Mol. Aspects Med., 3, 59- 118. Droms, K.A., Haley, B.E. and Malkinson, A.M. (1987). Decreased incorporation of the photoaffinity probe 8N3-[h-32P]GTP into a 45 kDa protein in lung tumors. Biochem. Biophys. Res. Commun., 144, 591-597. Johnson, G.L. and Dhanasekaran, N. (1989). The Gprotein family and their interaction with receptors. Endocrinol. Rev., 10, 317-331. Lange-Carter, C.A., Droms, K.D., Lester, B.R. and Malkinson, A.M. (1989). Growth rate dependence of differential incorporation of the photoaffinity probe 8N,-[A. 32P]GTP into GSa from metastatic variants melanoma cells. Cancer Res., 49, 3173-3177. Gupta, S.K., Gallego, C. and Johnson, G.L.
of B16 (1992).
Mitogenic pathway regulated by G protein oncogenes. Mol. Biol. Cell., 3, 123- 128. Climenti, E., Malgaretti, N., Meldolesi, J. and Toramelli, R. (1990). A newly constitutively activating mutation of the Gs protein (Y subunit-gsp oncogene is found in human pituitary tumors. Oncogene, 5, 1059 - 1061. Suarez, H.G., du Villard, J.A., Caillou, B., Schumberger, M., Parmenteir, C. and Monier, R. (1991). GSP mutations in human thyroid tumors. Oncogene, 6. 677-679. Okamura, N. and Terayama, H. (1976). Comparison of epinephrine-mediated activation of adenylate-cyclase in plasma membranes from liver and ascites hepatomas of rats. Biochim. Biophys. Acta., 455, 297 -314. Lester, B., Stadel, J.M., Buscarino, C., Sheppard, J., Greig, R.G. and Poste, G. (1987). Reconstitution of the Gs protein from B16 melanoma lines of high and low metastatic potential into S49-CYCexperimental membranes. Biochem. Biophys. Res. Commun., 147, 443-451. Vallar, L., Spada, A. and Giannattasio, G. (1987) Altered GS and adenylate cyclase activity in human GH-secreting pituitary adenomas. Nature, (Lond.). 330, 566 - 568. Viallet, J., Sharoni, Y., Frucht, H., Jensen, R.T., Minna, J.D. and Sausville, E.A. (1990). Cholera toxin inhibits signal transduction by several mitogens and the in vitro growth of human small-cell lung cancer. J. Clin. Invest., 86, 1904- 1912. Rosengurt, E. (1986) Early signals in the mitogenic response. Science, 234, 161- 166. Spiegel, A. and Fishman, P.H. (1987). Human cDNA clones for four species of Go, signal transduction protein. Proc. Natl. Acad. Sci., U.S.A., 83, 8893 - 8897. You, M., Candrian, U., Maronpot, R.R., Stoner, G.D. and Anderson, M.W. (1989). Activation of the Ki-ras protooncogene in spontaneously occurring and chemically induced lung tumors of the strain A mouse. Proc. Natl. Acad. Sci., U.S.A., 86, 3070-3074. Anderson, M.W., You, M., Belinsky, S., Hegi, M., Maronpot, B., Wiseman, R. and Reynolds, S. (1992). Genetic alterations in lung tumors. Proc. Am. Assoc. Cancer Res., 33, 601- 602.
139
Differential responsiveness to agents which stimulate CAMP production in normal versus neoplastic mouse lung epithelial cells Carol A. Lange-Cartera,
“Molecular Colorado,
and Enuironmental Boulder,
Sciences Center,
(Received (Accepted
Kurt A. Dromsb, Jeffrey J. Vuillequeza and Alvin M. Malkinson”
Toxicology
CO. 80309-0297
Program
Cancer Center, School of Pharmacy, University of of Cell Biology and Anatomy, Texas Tech University, Health
and Colorado
and bDepartment
3601 4th St., Lubbock,
‘FX 79430
(USA)
16 August, 1992) 18 September, 1992)
Summary We examined the responsiveness of normal and neoplastic lung cells to agents which stimulate CAMP production. While their basal CAMP levels were similar, spontaneous in vitro
transformant E9 cells and PCC4 cells produced much
tumor-derived less CAMP in
response to 1 PM isoproterenol compared to non-tumorigenic Cl0 cells derived from normal mouse lung epithelium. lodocyanopindolol binding studies indicated that both neoplastic lines contained fewer fl-adrenergic
receptors
than normal Cl0 ce!ls. When recep-
tors were bypassed via treatment with 10 pM cholera foxin, the pattern of CAMPresponsiveness was reversed; both neoplasfic cell tines produced more CAMP than Cf 0 cells. Direct stimulation of adenylate cyclase with 100 PM forskolin greatly increased CAMP concentrations in all three cell lines. These anomalies at both the receptor and G-protein levels in neoplastic lung epithelial cells may
contribute
to their deregulated
growth.
Correspondence to: Alvin M. Malkinson, School of Pharmacy, University of Colorado, Boulder, CO 80309-0297, USA.
0304-3835/92/$05.00 Printed and Published
0 1992 Elsevier Scientific Publishers in Ireland
Keywords: adenylate cyclase; beta-adrenergic receptor; GTP-binding proteins; lung epithelial cell lines; growth regulation Introduction Neoplastic growth is maintained by several mechanisms, including over-production of positive growth factors and/or an overexuberant cellular response to them and by insensitivity to negative growth factors. CAMP is an important modulator of normal cellular growth and differentiation [l]. Addition of CAMP to cultured cells has demonstrated both positive [2] and negative [3] effects on proliferation. The mechanism of this dichotomy is unknown. CAMP may interact with other factors (i.e. growth factor receptors, oncogene or tumor-supressor gene products) which control growth [4]. The response to CAMP would then depend on the cell- or pathological statespecific milieu of factors driving growth and/or differentiation events. Chemical analogs of CAMP can induce differentiation and inhibit the growth of many neoplastic cell lines and tumors, including human colon and mammary carcinoma cell lines [5] and rat mammary tumors in vivo [6]. However, some tumors are Ireland Ltd
Isoproterenol
Receptor
Cholera Toxin
L
G-protein
Forskolin
L
Adenylate Cyclase
L
CAMP
ICYP Scheme
1.
unresponsive to CAMP treatment, probably due to aberrations in their intracellular CAMP receptors. Alterations in the CAMP signal transduction pathway are associated with mouse lung neoplasia. Susceptible strains of inbred mice develop primary adenomas and adenocarcinemas spontaneously and in response to carcinogens [7]. These tumors are an important model system of human bronchiole-alveolar carcinoma, that subclass of primary human lung adenocarcinoma in which incidence is increasing most rapidly [8]. A cell line called Cl0 was derived from one of the precursor cell types of these tumors, the alveolar type II pneumocyte [9]. A spontaneous transformant of similar histologic origin, called E9 [lo] and a cell line derived from a lung tumor, called PCC4 [ 111, can be compared to Cl0 to study mechanisms of their resistance to negative growth signals. These tumorigenic cells express decreased levels of the type I isozyme of CAMP-dependent protein kinase (PKA) [ 121. In addition, type II PKA regulatory subunits from mouse lung tumors lack high-affinity CAMP-binding sites [13]. Droms et al. [14] reported decreased adenylate cyclase responsiveness of neoplastic lung cells to isoproterenol. Herein, we address the mechanism of this decreased responsiveness using agents which interact with or bypass the fl-adrenergic receptor, specifically stimulating each step in the CAMP synthesis pathway, as shown in Scheme 1.
Materials and Methods
Cell culture The non-tumorigenic, immortal cell line (hereafter referred to as ‘normal’) ClO, is contact inhibited, while tumorigenic E9 and PCC4 cells do not display density-dependent inhibition of growth at confluence [ 10,111. C 10 and E9 cells were maintained using CMRL 1066 (Irvine) medium and PCC4 cells using McCoys 5A (Gibco) in Corning loo-mm tissue culture plates. Media were supplemented with 10% fetal bovine serum and cells were grown in a humidified atmosphere of 5% COz/95% air. CAMP radioimmunoassay
The CAMP concentration in ether-extracted soluble fractions of cell Iysates was assayed using the cAMP[‘*~I] assay system (Amersham) as described [14]. Total CAMP accumulation was normalized to the total amount of cellular protein contained in the insoluble fraction as quantified by the method of Lowry et al. [15]. Time courses indicated maximal stimulation of CAMP production at 1 min following exposure to isoproterenol, 120 min using cholera toxin and 5 min using forskolin. ‘251-labeled ICYP binding assay &Adrenergic receptors were quantified in ClO, E9 and PCC4 cell membranes by Scatchard analysis of 1251-labeledICYP binding as described [16], except that 10 - 30 pg cell membrane protein was used per assay tube. A
1-pM quantity of unlabeled DL-propranolol was included in some reactions to measure non-specific ‘251-labeleing ICYP binding. Scatchard analysis was done using the Ligand Binding Analysis Program [ 171.
0.8 k .
m
Results in isoproterenol-treated lung cells To confirm that normal and neoplastic mouse lung epithelial cells were differentially responsive to agents which stimulate CAMP production, non-tumorigenic Cl0 cells and tumorigenic E9 and PCC4 cells were treated for 1 min with 1 PM isoproterenol (Fig. 1). Normal Cl0 cells produced more CAMP in response to isoproterenol than either in vitro transformant E9 cells or tumor-derived PCC4 cells (Fig. 1). This normal-neoplastic difference is considerably magnified compared to that reported earlier [14], in which a 5-min exposure to isoproterenol was assayed.
0.4
CAMP responsiveness
normal
and neoplastic
E9
Cl0
Cell
PCC4
Line
Fig. 1. CAMPProduction in isoproterenol treated normal and neoplastic lung cells. A11cell lines had similar basal levels of CAMP. Duplicate cultures of each cell line were treated for 1 min with (striped bars) or without (solid bars) 1 mM isoproterenol, a /3-adrenergic agonist. Normal Cl0 cells were more responsive to isoproterenolstimulated CAMPproduction than the two tumorigenic cell lines, E9 and PCC4. Data represent the’ mean of duplicate measurements from a typical experiment l the range.
0.0
10
U
“‘ICYP
20
Bound
30
40
(fmol/mg)
Fig. 2. P-Adrenergic receptor binding in normal and neoplastic cell membranes. Cell membranes were isolated and subjected to [‘251]-labeled ICYP binding assays as described in Methods. Normal Cl0 (B) cells contained more fl-adrenergic receptors than neoplastic PCC4 (A) and E9 (UJ cells. Points represent the mean of duplicate determinations from a typical experiment. A comparison of B,,, from several ments indicated that P < 0.001.
independent
experi-
To examine the mechanism of decreased isoproterenol responsiveness in the tumorigenie cell lines, binding studies were done using ‘251-labeled ICYP, a ligand with high affinity and specificity for the fl-adrenergic receptor [18]. Scatchard analysis indicated fewer /3-adrenergic receptors in neoplastic E9 and PCC4 cell membranes (B,, = 6.5 and 12 fmol/mg protein, respectively) than in normal Cl0 cell membranes (B,,, = 37 fmol/mg protein) (Fig. 2). Membranes isolated from all three cell lines demonstrated similar binding affinities for 1251-labeled ICYP (Kd = 20-40 PM). CAMP responsiveness forskolin-treated cells
in cholera toxinnormal and neoplastic
and lung
To determine whether further alterations downstream of the @-adrenergic receptor contributed to differential CAMP responsiveness between normal and neoplastic cells, each cell line was treated with cholera toxin which
to 10 pM cholera toxin, while normal Cl0 cells were minimally responsive to this agent (Fig. 3A). A 100~PM quantity of forskolin superinduced CAMP production in E9 cells to ZOO-fold over control levels, while Cl0 and PCC4 cells responded similarly, undergoing approximately lo-fold increases (Fig. 3B). Discussion PCC4
E9
2500 -
B
20001500 lOOO500 0
t_-l E9
Cl0
Cell
PCC4
Line
Fis. 3. CAMP Production in cholera toxin (A) and forskolin (B) treated normal and neoplastic lung cells. (A) duplicate cultures of each cell line were treated for 120 min with (striped bars) or without (solid bars) 10 pM cholera toxin. Neoplastic E9 and PCC4 cell lines were more responsive to cholera toxin-stimulated CAMP production compared to normal Cl0 cells. (B) Duplicate cultures of each cell line were treated for 5 min with (striped bars) or without (solid bars) 100 mM forskolin, which stimulates adenylate cyclase direclly. E9 cells were extremely sensitive to forskolin-stimulated CAMP production, while Cl0 and PCC4 cells demonstrated similar responsiveness. Data represent the mean of duplicate measurements from a typical experiment * the range.
constitutively activates Gsar via ADPribosylation [ 191, or with forskolin, a direct activator of the adenylate cyclase catalytic subunit [ZO] (Fig. 3). Both neoplastic cell lines produced large amounts of CAMP in response
Alterations in the receptor-coupled adenylate cyclase system characterize many tumors (for review see Ref. 21). Interestingly, the tumorigenic mouse lung epithelial cells examined herein contain defects at both the fladrenergic receptor and Gs transducer element levels. Both neoplastic cell lines express fewer /3-adrenergic receptors than normal Cl0 cells. Our cholera toxin data suggests that functional alterations also reside in the stimulatory G-protein complex (Gs) coupled to adenylate cyclase. Droms et al. [14,22] identified decreased photoincorporation of 8N3-I”/32P]GTP into GSCX proteins in several tumorigenic mouse lung cell lines, including E9 and PCC4; this was not due to decreased Gscr protein concentration as measured by immunoblotting experiments [14]. The alteration/mutation which causes this functional change in tumor cell GSCYmay also confer increased cholera toxin responsiveness. Alternatively, Gs in normal Cl0 cells may be negatively controlled by endogenous factors lacking in the neoplastic cells and is therefore less affected by cholera toxin. The PX subunits of the Gs complex can attenuate o-subunit activity [23]. Perhaps the ADP-ribosylation site on GSCYin Cl0 cells is blocked or posttranslationally modified. In neoplastic PCC4 and E9 cells, this negative control over Gs may be lost, allowing its constitutive activation. In B16 mouse melanoma cells, changes in the cellular growth rate are correlated with functional changes in Gscu [24]. In cell types in which CAMP is mitogenic, alterations in GSCX can induce hyperplasia and contribute to neoplastic transformation (for review see Ref. 25). Mutations that result in
143
GTPase-deficient and therefore constitutively active, Gscr subunits are associated with tumors of the pituitary [26] and thyroid [27] glands. Modifications of Gscr have also been associated with transformation of tissues in which CAMP is not mitogenic, but may influence neoplastic cell growth. GTPdependent coupling of adrenergic receptors to adenylate cyclase is modified in several hepatomas compared to ‘normal liver tissue [28]. Gsar proteins isolated from B16 melanoma cells with high metastatic potential stimulated adenylate cyclase to a greater extent in reconstituted membranes as compared to Gscr obtained from cells with low metastatic potential [29]. Each lung cell line produced large amounts of CAMP in response to forskolin, indicating similarities in the adenylate cyclase catalytic subunit in normal and neoplastic cells. The superinduction of CAMP levels by forskolin in E9 cells compared to C 10 and PCC4 cells may reflect their altered Gscr proteins [14] and cholera toxin hyper-responsiveness (Fig. 3A). Increased forskolin-responsiveness has been attributed to altered Gs function in B16 melanoma cells [29] and human pituitary tumors [30]. That PCC4 cells do not respond similarly to E9 cells suggests that these two cell lines may contain Gs alterations/mutations which are qualitatively different. Alternatively, E9 cells may express increased levels of adenylate cyclase. That each neoplastic cell line produced high concentrations of CAMP in response to both agents acting downstream of @-adrenergic receptor stimulation indicates intact coupling of Gs to adenylate cyclase and suggests that the decreased isoproterenol responsiveness of these cells is due entirely to their lowered @-adrenergic receptor expression. Multiple anomalies in the CAMP signalling pathway may diminish the ability of CAMP to regulate growth. One role of CAMP in affecting tumor cell growth may be to modulate, rather than mediate, the activity of various oncogenes and growth factors via phosphorylation events (for review see Ref. 4). Addition of
cholera toxin to human small-cell lung cancer cell lines decreased their responsiveness to several diverse mitogenic stimuli [31]. CHO cell growth was inhibited by CAMP, while o-src expressing, RSV-transformed CHO cells were CAMP-resistant. Cholera toxin was mitogenic to Swiss 3T3 cells [32], but inhibited the growth of ras-transformed Swiss 3T3 cells [33]. Aberrations associated with mouse lung tumorigenesis include K-ras proto-oncogene activation [34] and p53 tumor suppressor gene mutation [35]. It will be interesting to dissect the role of CAMP signal transduction in cells containing various combinations of activated oncogenes and/or mutated tumor suppressor genes. Acknowledgments
We would like to thank Dr. G.L. Johnson (National Jewish Center for Immunology and Respiratory Medicine, Denver) for helpful comments on this manuscript. We are grateful to Ms. S. Leah Butler for excellent technical assistance. This work was supported by USPHS Grant ES02370. References Beebe, S.J. and Corbin, J.D. (1986). Cyclic nucleotidedependent protein kinases. In: The Enzymes, Vol. 17, pp. 43- 111. Editor: P.D. Boyer. Academic Press, Orlando, FL. Dumont, J.E., Jauniaux, J.C. and Roger, P.P. (1989). The cyclic AMP-mediated stimulation of celf proliferation. Trends Biochem. Sci., 14, 67-71. Russell, D.H. (1978). Type I cyclic AMP-dependent protein kinase as a positive effecter of growth. Adv. Cyclic Nut. Res., 9, 493-506. Gottesman, M.M. and Fleischmann, R.D. (1986). The role of CAMP in regulating tumor cell growth. Cancer Surv., 5, 291- 308. Taglfaferri, P., Katsaros, D., Clair, T., Ally, S., Tortora, G., Neckers, L., Rubalcava, B., Parandoosh, Z., Chang, Y., Revankar, G.R., Crabtree, G.W., Robins, R.K. and Cho Chung, Y.S. (1988) Synergistic inhibition of growth of breast and colon humena cancer cell lines by siteselective cyclic AMP analogues. Cancer Res., 48, 1642 - 1650. Cho-Chung, Y.S., Clair, T. and Shepheard, C. (1983). Anticarcinogenic effect of N6,02-dibutyryl cyclic adenosine 3’,5’-monophosphate on 7,12-dimethy-
144
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
benz(a)anthracene mammary tumor induction in the rat adenosine 3’5’to cyclic its relationship and monophosphate metabolism and protein kinase. Cancer Res., 43, 2736 - 2740. Malkinson, A.M. (1991). Genetic studies on lung tumor susceptibility and histogenesis in mice. Environ. Health Perspect., 93, 149 - 159. Gazdar, A.F. and Linnoila, R.I. (1988). The pathology of lung cancer - changing concepts and newer diagnostic techniques. Semin Oncol., 55, 215 - 225. Smith, G.J., LeMesurier, S.M., De Montfort, M.L. and Lykke, A.W.J. (1984). Development and characterization of type 2 pneumocyte-related cell lines from normal adult mouse lung. Pathology., 16, 401-405. Smith, G.J. and Lykke, A.W.J. (1985). Characterization of a neoplastic epithelial cell strain derived from dexamethazone treatment of cultured normal mouse type 2 pneumocytes. J. Pathol., 147, 165- 172. Nicks, K.M., Droms, K.A., Fossli, T., Smith, G.J. and Malkinson, A.M. (1989). Altered function of protein kinase C and CAMP-dependent protein kinase in a cell line derived from a mouse lung papillary tumor. Cancer Res., 49, 5191- 5198. Lange-Carter, C.A., Fossli, T., Jahnsen, T. and Malkinson, A.M. (1990). Decreased expression of the type I isozyme of CAMP-dependent protein kinase in tumor cell lines of lung epithelial origin. J. Biol. Chem., 265, 7814- 7818. Malkinson, A.M. and Butley, M.S. (1981). Alterations in cyclic adenoxine 3’,5’-monophosphate-dependent protein kinases during normal and neoplastic lung development. Cancer Res., 41, 1334- 1341. Droms, K.A., Haley, B.E., Smith, G.J. and Malkinson, A.M. (1989). Decreased SN,-[X-32P]GTP photolabeling of Gso in tumorigenic lung epithelial cell lines: association with decreased hormone responsiveness and loss of contact-inhibited growth. Exp. Cell. Res., 182, 330 - 339. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951). Protein measurement with the folin phenol reagent. J. Biol. Chem., 193, 265-275. Valverius, P., Hoffman, P.L. and Tabakoff, B. (1987). Effect of ethanol on mouse cerebral cortical betaadrenergic receptors. Mol. Pharmacol., 32, 217 - 222. Munson, P.J. and Rodbard, D. (1983). LIGAND: A versatile computerized approach for characterization of ligandbinding systems. Anal. Biochem., 107, 220 - 239. Burgisser, E. and Lefkowitz, R.J. (1984) Brain Receptor Methodologies, Part A, pp. 229-253. Editors: P.J. Marangos, I.C. Cambell and R.M. Cohen. Academic Press, Orlando, FL. Moss, J. and Vaughan, M. (1988). ADP-ribosylation of guanyl nucleotide-binding regulatory proteins by bacterial toxins. Adv. Enzymol., 61, 303-379. Downs, J.R.W. and Aurbach, G.D. (1982). The effect of forskolin on adenylate-cyclase in S49-wild type and CYC (-)-cells. J. Cyclic Nucleotide Res., 8, 235-242. Hunt, N.H. and Martin, T.J. (1980). Hormone receptors
22
23
24
25
26
27
28
29
30
31
32 33
34
35
and cyclic nucleotides - significance for growth and function of tumors. Mol. Aspects Med., 3, 59- 118. Droms, K.A., Haley, B.E. and Malkinson, A.M. (1987). Decreased incorporation of the photoaffinity probe 8N3-[h-32P]GTP into a 45 kDa protein in lung tumors. Biochem. Biophys. Res. Commun., 144, 591-597. Johnson, G.L. and Dhanasekaran, N. (1989). The Gprotein family and their interaction with receptors. Endocrinol. Rev., 10, 317-331. Lange-Carter, C.A., Droms, K.D., Lester, B.R. and Malkinson, A.M. (1989). Growth rate dependence of differential incorporation of the photoaffinity probe 8N,-[A. 32P]GTP into GSa from metastatic variants melanoma cells. Cancer Res., 49, 3173-3177. Gupta, S.K., Gallego, C. and Johnson, G.L.
of B16 (1992).
Mitogenic pathway regulated by G protein oncogenes. Mol. Biol. Cell., 3, 123- 128. Climenti, E., Malgaretti, N., Meldolesi, J. and Toramelli, R. (1990). A newly constitutively activating mutation of the Gs protein (Y subunit-gsp oncogene is found in human pituitary tumors. Oncogene, 5, 1059 - 1061. Suarez, H.G., du Villard, J.A., Caillou, B., Schumberger, M., Parmenteir, C. and Monier, R. (1991). GSP mutations in human thyroid tumors. Oncogene, 6. 677-679. Okamura, N. and Terayama, H. (1976). Comparison of epinephrine-mediated activation of adenylate-cyclase in plasma membranes from liver and ascites hepatomas of rats. Biochim. Biophys. Acta., 455, 297 -314. Lester, B., Stadel, J.M., Buscarino, C., Sheppard, J., Greig, R.G. and Poste, G. (1987). Reconstitution of the Gs protein from B16 melanoma lines of high and low metastatic potential into S49-CYCexperimental membranes. Biochem. Biophys. Res. Commun., 147, 443-451. Vallar, L., Spada, A. and Giannattasio, G. (1987) Altered GS and adenylate cyclase activity in human GH-secreting pituitary adenomas. Nature, (Lond.). 330, 566 - 568. Viallet, J., Sharoni, Y., Frucht, H., Jensen, R.T., Minna, J.D. and Sausville, E.A. (1990). Cholera toxin inhibits signal transduction by several mitogens and the in vitro growth of human small-cell lung cancer. J. Clin. Invest., 86, 1904- 1912. Rosengurt, E. (1986) Early signals in the mitogenic response. Science, 234, 161- 166. Spiegel, A. and Fishman, P.H. (1987). Human cDNA clones for four species of Go, signal transduction protein. Proc. Natl. Acad. Sci., U.S.A., 83, 8893 - 8897. You, M., Candrian, U., Maronpot, R.R., Stoner, G.D. and Anderson, M.W. (1989). Activation of the Ki-ras protooncogene in spontaneously occurring and chemically induced lung tumors of the strain A mouse. Proc. Natl. Acad. Sci., U.S.A., 86, 3070-3074. Anderson, M.W., You, M., Belinsky, S., Hegi, M., Maronpot, B., Wiseman, R. and Reynolds, S. (1992). Genetic alterations in lung tumors. Proc. Am. Assoc. Cancer Res., 33, 601- 602.
