Monophosphate-Dependent Protein Kinases During the Progression of Urethan-Induced Mouse Lung Tumors1

Home , Autophosphorylation

[CANCER RESEARCH 45, 3677-3685, August 1985] Changes in Cyclic Adenosine 3':5'-Monophosphate-dependent Protein Kinases during the Progression of Urethan-induced Mouse Lung Tumors1

Martin S. Butley, Gary D. Stoner, David G. Beer, Deborah S. Beer, Robert J. Mason, and Alvin M. Malkinson2

School of Pharmacy, University of Colorado, Boulder, Colorado 80309 [M. S. B., D. G. B., D. S. B., A. M. M.]; Department of Pathology, Medical College of Ohio, Toledo, Ohio 43699 [G. D. S.]; and Department of Medicine, National Jewish Hospital/National Asthma Center, Denver, Colorado 80262 [R. J. M.]

ABSTRACT allow significant synergism of cAMP and Ca2+ on protein kinase activation. The cyclic adenosine 3':5'-monophosphate (cAMP)-dependent protein kinases in lung adenomas are functionally different INTRODUCTION from those of normal lung. The relevance of this change to Protein kinases have been implicated as agents controlling neoplastic conversion was examined by comparing tumor ki cellular growth and neoplastic transformation in a variety of nases with those obtained from the normal cell of origin and by systems (1-3). We have previously described functional changes studying the kinases at different stages of tumor growth. Lung in cAMP3-dependent protein kinases which occur during normal tumors were collected from A strain mice at different times after and neoplastic lung development (4-9). There are 2 isozymes of a single injection of urethan. These tumors are predominantly of cAMP-dependent protein kinases, each of which is composed of alveolar type two cell origin, and cAMP-binding proteins in ex a regulatory subunit dimer and 2 catalytic subunits (C). The type tracts from isolated type two cells and from lung adenomas at 1and type II isozymes differ only in their regulatory subunits, R, various stages of tumor progression were compared. Both the and R,,(10). Both R-subunits can be covalently radiolabeled with incorporation of the cAMP photoaffinity analogue, cyclic 8-azidoadenosine 3':5'-[32P]monophosphate (8-N3-[32P]cAMP), into 8-N3-[32P]cAMP, a photoaffinity analogue of cAMP (11, 12). In the regulatory subunits of the type I (R,) and type II (RM)cAMP- addition, the RMsubunit can be readily phosphorylated using either magnesium [7-32P]ATP or zinc [-y-32P]ATP as substrates dependent protein kinases and the autophosphorylation of R,, (10, 13). This autophosphorylation reaction is regulated by were similar in extracts from whole normal lung and from type cAMP, and the nature of this cAMP effect is dependent upon two cells. Altered protein kinases are thus not characteristic of which divalent cation is present (6, 14). By radiolabeling R, and normal type two cells. Lung tumors showed a decrease in R,, with 8-N3-[32P]cAMP or with [7-32P]ATP and subsequently photodetectable R,, which correlated in degree with tumor size separating proteins by denaturing gel electrophoresis, both pho- and extent of anaplasticity. This decreased RM photolabeling toincorporation into R-subunits and the effects of cAMP on R,, during tumor growth was associated with increased RH auto autophosphorylation can be independently monitored. In addi phosphorylation. In contrast, decreased R,, photolabeling in ex tion, DEAE-cellulose chromatography and sucrose gradient sedi tracts from neonatal lung is accompanied by a substantial de mentation can be used to obtain relative measures of the dis crease in R,, autophosphorylation. The characteristics of RM sociation state of the type I and type II isozymes (7, 9). during normal development thus clearly differ from those during By comparing these parameters using extracts from adult and neoplastic development. An increase in the amount of an M, neonatal lung and from urethan-induced lung adenomas, we 37,000 proteolytic fragment derived from R-subunits was also noted as a function of tumor progression. DEAE-cellulose chro- have previously established that both neonatal lung and lung adenomas are characterized by a decrease in the amount of matography of tumor cytosol showed that the increase in the photodetectable RMand by an increase in the amount of an M, amount of M, 37,000 protein was accompanied by increased 37,000 proteolytic fragment (6, 8). These earlier studies com subunit dissociation of the type I isozyme. The dissociated R, pared extracts from whole lung with extracts from lung adeno subunit has been shown to be more sensitive to cleavage by a Ca2+-dependent neutral protease than when R, was in the holo- mas. The lung is a cellularly heterogeneous tissue consisting of up to 43 different cell types (15). Mouse lung adenomas fall into enzyme form. This protease is present in both normal lung and 2 main histological classes. An alveolar form is composed of lung adenomas, and its activity increases during the later stages compactly arranged cuboidal cells and is thought to arise from of tumor progression. A comparison of cAMP binding and the light-induced covalent incorporation of 8-N3-[32P]cAMP showed alveolar type 2 cells (16, 17), whereas a papillary form is com posed of columns of cuboidal and/or columnar cells and is that, for both R, and RM,photoincorporation was about 75% as derived from the bronchiolar nonciliated Clara cells (18). The efficient as noncovalent binding. In contrast, although the M, relative proportion of alveolar to papillary tumors may vary with 37,000 fragment can be photolabeled with low concentrations of 8-N3-[32P]cAMP, noncovalent cAMP binding to the endoge the carcinogen used, the strain of mouse, and the route of administration (19). Approximately 85% of all tumors observed nous M, 37,000 fragment could not be demonstrated with a 14 weeks after a single injection of urethan into A strain mice standard filtration assay. Such altered cAMP binding character are of type 2 cell origin.4 As such, biochemical differences in R- istics following Ca2+-dependent proteolysis of R-subunits would 3The abbreviations used are: cAMP, cyclic AMP; R,,regulatory subunit of type I cAMP-dependentprotein kinase isozyme; R,,.regulatory subunit of type II cAMP- 1Supported by USPHS Grants ES02370, HL29891, and CA33497. dependentprotein kinase isozyme;S-Na-^PJcAMP,cyclic 8-azidoadenosine3':5'- 2 Recipient of Research Career Development Award CA00939. To whom re [^PJmonophosphate;RÂ»P,phosphorylatedform of the regulatory subunit of type II quests for reprints should be addressed. isozyme. Received 11/20/84; revised 4/24/85; accepted 4/30/85. 4D. G. Beer and A. M. Malkinson, unpublishedresults.

CANCER RESEARCH VOL. 45 AUGUST 1985 3677

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1985 American Association for Cancer Research. CAMP RECEPTORS IN LUNG TUMORS subunits between adenoma and whole lung extracts may reflect lation (6, 14). The amount of 32P incorporated into proteins by either unique properties of normal type 2 cells and not be related to procedure was quantified following protein separation by denaturing gel electrophoresis. Ca2+-dependent protease activity was measured using neoplastic conversion. To test this possibility, we herein compare [14C]casein as substrate, as previously described (7). The noncovalent 8-N3-[32P]cAMP photoincorporation in extracts from isolated nor binding of either [3H]cAMP or 8-N3-[32P]cAMP to proteins was assessed mal type 2 cells and from cell lines established from urethan- using the Millipore filtration technique of Oilman (27), as previously induced adenomas. modified (28). To further test the link between these biochemical changes Chromatography. Protein kinase isozymes were separated by DEAE- and neoplastic conversion, 8-N3-[32P]cAMP photoincorporation, cellulose Chromatography as described previously (7, 9). R,,autophosphorylation, and R-subunit proteolysis are examined at different stages of lung adenoma growth. R-subunit proteol ysis in mouse lung is primarily due to a Ca2+-dependent neutral RESULTS protease, and R, becomes more susceptible to Ca2+-stimulated proteolysis when dissociated free of the C-subunit (7). We have 8-N3-[32P]cAMP Photolabeling of R,,in Type 2 Cell Extracts. previously demonstrated an increased amount of photolabeled The absence of high-affinity 8-N3-[32P]cAMP photolabeling of Rn M, 37,000 fragment in neonatal lung relative to adult lung and in extracts from lung adenomas (6, 8) may result from neoplastic shown that this is very likely due to an increase in the dissociation conversion or may be a property characteristic of R,, from type state of the type I isozyme (7). In this study, we show that R, 2 cells. Type 2 cells can be isolated and purified from rat lungs from lung tumor cytosol is also primarily in the dissociated state. (24), and we have photolabeled extracts from these cells using This M, 37,000 proteolytic fragment possesses unique cAMP a range of 8-N3-[32P]cAMP concentrations (Chart 1). Since part binding characteristics, which suggest a model for Ca2+-modu- of the cellular isolation procedure involves plating cells in plastic lated activation of cAMP-dependent protein kinases. culture dishes, extracts from LA-4 cell lines derived from cloned type 2 cell-derived adenomas (25) were used to control for any effects that exposure to plastic and/or an absence of cellular MATERIALS AND METHODS interactions might cause on photoincorporation. The resulting RM binding curves show that normal rat type 2 cells contain both a Sample Preparation. Lung tumors were induced with a single i.p. high- and low-affinity component, which is identical to what is injection of urethan (1 mg/g) into male and female strain A/J mice that observed using whole mouse lung extracts (6, 8). This suggests were 6 to 12 weeks of age (6). Mice were killed by cervical dislocation at times ranging from 4 to 18 months after injection. Small tumors (<3- that a lack of high-affinity binding is not characteristic of the sq mm diameter) were obtained 4 to 6 months after urethan injection, normal type 2 cell but is the result of neoplastic conversion. medium tumors (3 to 5 sq mm) were obtained 12 months after injection, Consistent with this hypothesis, cultured adenoma cells at both and large tumors (>5 sq mm) and adenocarcinomas were obtained 15 passages 36 and 71 exhibited only low-affinity binding to RM to 18 months following injection. The large tumors and adenocarcinomas (Chart 1). were distinguished on the basis of the degree of anaplasticity and of Photolabeling and Autophosphorylation during Lung Ade invasiveness into adjacent tissue. Time-dependent growth of mouse lung noma Progression. Lung tumors analyzed in previous reports tumors has been described previously (20-23). Tumors which were to (4,6,8) were obtained 14 to 16 weeks following urethan injection be examined histologically were placed in 10% buffered formalin along with normal adjacent tissue. Tumors to be used for later biochemical ioo analysis were dissected free of adjacent tissue, and tumors of similar sizes were pooled. These pooled samples were either maintained as frozen tissue samples at -70Â°C until use or were homogenized in 5 volumes (w/v) of 10 mM potassium phosphate buffer, pH 7.0, containing 1 mÂ«EDTA, dispensed into aliquots, and kept at -70Â°C. Alveolar type 2 cells were isolated from adult male Sprague-Dawley rats by a procedure which utilizes elastase digestion of lungs, purification on a metrizamide gradient, and differential adherence to plastic culture dishes as described by Dobbs ef a/. (24). The cells are 90% type 2 cells as determined by the modified Papanicolaou stain (24). The isolated cells were frozen at -70Â°C, thawed, and sonicated in phosphate buffer, and a postmitochondrial supernatant was obtained by centrifugation at 20,000 x g for 45 min. LA-4 cell lines, isolated and cloned from alveolar lung adenomas, and passaged either 36 or 71 times as described by Stoner et al. (25), were kindly provided by Dr. Lyle Arnold, University of California at San Diego. After growth in Ham's F-12 medium supplemented with 10% fetal calf serum, the cells were scraped from plates into 0.32 M sucrose, and the pellets were frozen on dry ice. Extracts were prepared as described above. 500 IOOO [8-N3-S2P-cAMP] Assays. The technique of Lowry ef al. (26) was used to determine protein concentration, with crystalline bovine serum albumin as the Chart 1. Photoincorporation into RÂ«fromisolated alveolar type 2 cells (â€¢)and standard. The photoaffinity labeling of R-subunits from cell and tissue from the LA-4 cell line (O).Cytosolic extracts were photolabeledover a rangeof 8- Ns-^PjcAMP concentrations. Proteins were then separated on denaturing gels, extracts with S-Na-^PjcAMP was performed as described previously and radioactivitywas detected with autoradiographyas described in "Materialsand (8). Endogenous proteins were phosphorylated with [-y-^PJATP to mea Methods." Quantitative estimates of R,, photoincorporation were made by densi- sure the effects of cAMP on the initial rate of R-subunit autophosphory tometric analysisof the autoradiogram.

CANCER RESEARCH VOL. 45 AUGUST 1985 3678

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1985 American Association for Cancer Research. CAMP RECEPTORS IN LUNG TUMORS and were <3 mm in diameter. With increased time following o 200 injection, some tumors become larger and can involve an entire I- < lobe of lung, and a few show morphological and functional criteria of cancer (20-22). Fig. 1/\ shows 2 small lung adenomas ob K served 5 months after urethan injection, one of which exhibits O EL the histological characteristics of an alveolar type 2 cell tumor, Vi ~ and the other, of a papillary Clara cell tumor. A medium-sized 0 z 1 O ioo alveolar adenoma (2 to 4 mm) obtained 12 months after injection is shown in Fig. ÃŒB.Thelarge tumors (4 to 6 mm) obtained 13 II to 15 months following injection are usually papillary in structure (Fig. 1C). Papillary adenomas may have a greater malignant potential than alveolar adenomas (19), and portions of large (>7 z a: mm) invasive adenocarcinomas often retain a papillary structure o o (Fig. 1D). Tumors of similar size were pooled to form 4 different size classes designated small, medium, large, and adenocarci- < t- noma (Fig. 1, A to D). Each class thus represents a mixture of o < tumors, with the small class containing 85% alveolar adenomas while the larger size classes have both an increased malignant potential and may also contain a greater proportion of Clara cell- o " -50 derived tumors. tÃ LÃœ The composite of 4 autoradiograms in Fig. 2 is derived from u cytosols from each tumor class which have been photolabeled with varying concentrations of 8-N3-[32P]cAMP. Five proteins show photoincorporation and correspond to RnP, a phosphoryl- -I001â€” < ated form of Rn which has a molecular weight of 56,000 on < denaturing gels, Rn at M, 54,000, R, at M, 49,000, and 2 proteo- Z i O O? lytic fragments derived from R-subunits at M, 45,000 and 37,000. O Z Z (j Ã»LU (t UJ lÃ¼(C As the tumors progress from small to medium size, the most Lu O

CANCER RESEARCH VOL. 45 AUGUST 1985 3679

>-> 6mLiJ

^*V)O

UjQ.'Q. 4

^0?Lu

OS"*2_i toZ)l-

2zP2C/51â€”ur=C-,(-HIriiLiLu.'--â€¢-\<<

21 < < Â°-IQ 55 2 > Chart 3. Relative levels of Rn autophosphorylation in extracts from tumors at OÂ±-IO ^O uJO various stages of progression. The phosphorylation reaction was as described in (_)uz -IZ =Z OZ Z Chart 2; MgCI2 (â€¢):ZnCI2(D). UJQ:Q~<_2E^ui QUJ orui rfcn

is almost absent, and most R, is in a dissociated state (Chart 46). There is also a large peak of M, 37,000 proteolytic fragment in fractions from the tumor cytosol. This simultaneous demonstration of a large peak of free R, together with more M, 37,000 fragment in lung tumor extracts may be more than fortuitous. We have previously demonstrated that the large amount of M, 37,000 fragment in neonatal lung is due to the sensitivity of free R, (versus R, in holoenzyme form) to cleavage by an endogenously occurring Ca2+-dependent neu SMALL LUNG 3h ADENOMA tral protease (7). A quantitative measure of Ca2+-dependent protease activity in lung tumor extracts was obtained using [14C]casein as a substrate (Chart 5). Cytosol from small lung adenomas contains even less Ca2+-dependent protease activity than does cytosol from normal adult lung. This emphasizes that the dissociation state of the type I isozyme, and not only the amount of protease present, is an important factor affecting the oo amount of M, 37,000 fragment present. Since dissociation of the type I isozyme is nearly complete in the small adenomas, further dissociation cannot explain the continued increase in the relative proportion of this proteolytic fragment at later stages of tumor progression. These more advanced tumors do contain higher levels of Ca2+-stimulated protease activity, however, which could 10 20 30 40 FRACTION NUMBER well account for the further increase in the amount of M, 37,000 Chart 4. DEAE-cellulose chromatography of cAMP-binding proteins in normal protein (Chart 5). lung and lung tumor cytosol. Chromatographie separation was achieved with a 10 Comparison of Noncovalent [3H]cAMP and 8-N3-[32P]cAMP to 400 miui linear NaCI gradient in 4 HIM Tris buffer at pH 7.4, as previously described (2). A 20-^1 aliquot from each column fraction was preincubated with 125 Binding with 8-N3-[32P]cAMP Photoincorporation. Many of nM S-Na-C^PJcAMP and irradiated with UV light. Covalent photoincorporation into these conclusions regarding functional changes in R-subunits R. (O), RÂ«(â€¢),and an M, 37,000 fragment (A) was determined by separating proteins on denaturing gels, locating radioactive bands by autoradiography, and have been obtained from experiments examining the photoin then counting relevant portions of the gel in a liquid scintillation counter. corporation of 8-N3-[32P]cAMP. The cellular processes about which we make inferences involve noncovalent binding rather R, associated with C-subunit in the form of a type I holoenzyme, than photoincorporation, however, and cAMP rather than 8-N3- while the second peak contains dissociated, or free, R, (29). In [32P]cAMP. It is therefore important to ascertain the extent to the small lung adenoma, however, the type I holoenzyme peak which 8-N3-[32P]cAMP photoincorporation parallels cAMP bind-

CANCER RESEARCH VOL. 45 AUGUST 1985 3680

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1985 American Association for Cancer Research. CAMP RECEPTORS IN LUNG TUMORS ing in the mouse lung system. Adult and neonatal lung proteins kinase isozymes in normal neonatal and adult lung, in isolated were separated by DEAE-cellulose chromatography, and the alveolar type 2 cells, and in lung tumors at various stages of various fractions were assayed for noncovalent [3H]cAMP bind progression, we have determined whether such functional alter ing, noncovalent 8-N3-[32P]cAMP binding, and 8-N3-[32P]cAMP ations are correlated with normal versus neoplastic growth, and photoincorporation. The lung extracts were preincubated with whether they are unique to a particular cell type or reflect magnesium ATP prior to chromatography to reassociate the type properties of the lung as a whole. Although the cause and effect I isozyme, as described earlier (9). Noncovalent binding studies relationships between these parameters remain conjectural, cer were done using the standard filtration assay (27,28) at both pH tain patterns can be suggested. During both normal and neo 4.0 and pH 6.5. Binding at pH 4.0 effectively eliminates phos- plastic growth, the affinity of RM,but not R,, for 8-N3-[32P]cAMP phodiesterase activity and thus assures a constant concentration decreases (6,8). This is accompanied by a decrease in the ability of [3H]cAMP, while binding at pH 6.5 makes the assay directly comparable to the 8-N3-[32P]cAMP photoincorporation studies. of bound cAMP to cause dissociation of the type II holoenzyme 8-N3-[32P]cAMP is resistant to mammalian phosphodiesterases (8). These changes in the ability of cAMP to activate the type II isozyme have also been described in the rat mammary tumor (30). pH does not significantly affect [3H]cAMP binding in adult and system (34). They do not appear to involve structural changes in the R,, subunit, but more likely, they result from altered neonatal lung extracts (Chart 6, A and B). Binding to the type I interactions of the type II isozymes with other molecules in the holoenzyme (Fractions 11 to 15) is slightly reduced at pH 4.0, cell (8). Reduced Ru photoincorporation during normal growth is while the reverse is true for the type II holoenzyme (Fractions 33 to 43). Noncovalent 8-N3-[32P]cAMP binding was also measured accompanied by a very significant reduction in RMautophosphor- ylation (6), while during neoplastic growth, RMautophosphoryla- using the standard filtration assay at pH 6.5, and it closely parallels [3H]cAMP binding at the same pH (Chart 6, A to D). tion actually increases (Chart 3). As such, reduced Ru photoin Covalent 8-N3-[32P]cAMP photoincorporation was determined by corporation may be indicative of the proliferative status of the separating the radiolabeled proteins on denaturing gels and cell, while Rn autophosphorylation may regulate the proliferative measuring the amount of radioactivity in the various bands. response and help determine whether growth will be normal or Proteins corresponding to Ft,, FtÂ«,andthe M, 37,000 proteolytic neoplastic. To perhaps compensate for the decreased type II fragment were the only ones labeled. If more than one radiola isozyme activity, the type I isozyme in both neonatal lung and beled protein was present in a given fraction, they were pooled lung adenoma appears to be in a highly active state. This is to yield a single value for total photoincorporation (Chart 6, C shown by an increase in the dissociation state of the type I and D). From the data, it is evident that, for each holoenzyme, isozyme as determined both by sucrose gradient sedimentation photoincorporation is about 75% as efficient as noncovalent (8) and by DEAE-cellulose chromatography (Chart 4). Because binding of 8-N3-[32P]cAMP. Surprisingly, however, there is a free R, is more sensitive to proteolytic attack (7, 35), increased region between the 2 holoenzyme peaks (Fractions 25 to 33) in type I isozyme dissociation is accompanied by an increase in the which photoincorporation is actually more efficient than nonco amount of M, 37,000 fragment. Consistent with our studies on valent 8-N3-[32P]cAMP binding. This third peak of photoincorpor RI proteolysis using lung extracts, other investigators have es ation activity is especially apparent in the neonate (Chart 60), tablished that R, turnover is markedly stimulated by kinase but the protein responsible for this activity cannot be detected dissociation in intact cultured cells (36). An increased amount of with the filtration assay using either [3H]cAMP or 8-N3-[32P]cAMP free R, has also been reported in a hamster pancreatic islet cell as ligands. When photoincorporation into R,, RM, and the M, tumor (37) and in human adenovirus-transformed rat cells (38). 37,000 fragment is individually assessed (Chart 6, Â£and F), it The unique binding characteristics of the M, 37,000 fragment can be seen that the third peak of photoincorporation (Fractions suggest a new interpretation of its role in kinase activation. It 25 to 33) corresponds to the M, 37,000 protein. For this proteo has been previously proposed that this fragment, which does lytic fragment, 8-N3-[32P]cAMP photoincorporation does not cor not bind C-subunit, would act as a "cAMP sink" by binding free relate with the ability to bind [3H]cAMP. cellular cAMP, thus making it unavailable for holoenzyme acti There are 2 cAMP binding sites per R monomer, and the vation (39). This would effectively increase the amount of cAMP filtration assay measures cAMP bound to only one of the 2 sites required to stimulate kinase activity. This hypothesis presup (9, 31, 32). If this site were cleaved during formation of the Mr 37,000 fragment, this would explain the apparent lack of [3H]- poses, however, that the M, 37,000 fragment has a normal, cAMP binding. However, 8-N3-[32P]cAMP photoincorporation intact cAMP binding site. This does not appear to be the case. Incubation of both R, and R,, with various proteases produces also occurs at only one of the 2 binding sites, and that site which M, 37,000 fragments, and examination of [3H]cAMP binding to retains [3H]cAMP after filtration and the site of photoincorpora these fragments using an ammonium sulfate precipitation assay tion are probably one and the same (32, 33). The photoincorpo allows detection of both cAMP binding sites on each R-monomer ration without binding shown in Chart 6 cannot be explained by (39). Proteolytic cleavage of R, results in a 50 to 70% decrease proteolytic removal of one of the 2 cAMP binding sites. Photoin in [3H]cAMP binding, suggesting a loss or modification of one or corporation into the M, 37,000 fragment occurs with high affinity, exhibits saturation (Fig. 2), and can be completely inhibited by both of the R, binding sites. This is consistent with our hypothesis the addition of an excess of unlabeled cAMP to the reaction that the endogenous M, 37,000 fragment in both neonatal lung mixture. and lung adenoma is derived from free Rt. In contrast, the M, 37,000 fragments generated from Râ€žretainfull cAMP binding potential. These results, together with the data we present in DISCUSSION Chart 6, suggest that cAMP is only weakly bound to the M, Several functional changes in cAMP-binding proteins occur 37,000 protein and thus may be readily released or transferred during lung growth and neoplasia. By characterizing protein to intact R-subunits within the cell. Cleavage of the free R, subunit

CANCER RESEARCH VOL. 45 AUGUST 1985 3681

ADULT NEONATE

C 80 o E

Â§60

mO o. 40 Z

ADULT NEONATE o e so

"20

NEONATE

co 20 30 40 90 20 30 40 50 FRACTION NUMBER FRACTION NUMBER Chart 6. cAMP binding and S-Ns-^PjcAMP photoincorporation following DEAE-cellulose chromatography of neonatal and adult mouse lung cytosol. Extracts were first preincubated with 0.15 mw ATP and 3.5 HIM MgClj for 30 min at 30Â°Cto facilitate association of the type I isozyme, as previously described (5). Lung proteins were then separated by DEAE-cellulose chromatography using a 10 to 400 rtiM linear NaCI gradient in 10 mM potassium phosphate buffer at pH 7.0. In A and B, column fractions were assayed for [3H]cAMP binding at pH 4.0 (â€¢)andpH 6.5 (O) using a filtration assay as described (28). In C and D, the same column fractions were assayed for noncovalent 8-N3-[3:!P]cAMP binding (O), using the same filtration assay as in A and ÃŸabove, and for covalent S-Ns-l^PJcAMP photoincorporation (â€¢).Photoincor poration was quantified by separating proteins on denaturing gels, locating radioactive proteins by autoradiography, and then counting relevant portions of the gel in a liquid scintillation counter. Points, total photoincorporation into R,, RH,and the M, 37,000 protein. For E and F, photoincorporation into R-subunits was determined as in C and D above. In this case, however, covalent binding to Ri (O), R,, (â€¢),andthe M, 37,000 protein (A) is plotted separately. to form M, 37,000 fragment would thus both prevent that R, from even though it shows saturation, the efficiency of photoincor recombining and inactivating free C-subunit and allow the release poration is unknown. Given its unusual binding characteristics, or transfer of loosely bound cAMP to other intact R-subunits. the photoincorporating efficiency of the M, 37,000 protein may The M, 37,000 protein may thus act as a cAMP storage molecule, be significantly less than that obtained with intact RI or RM maintaining cAMP at high concentrations in specific intracellular subunits. locations where it can be continually recycled to more effectively Ca2+-dependent protease has been strongly implicated as the activate protein kinase holoenzymes. The amount of M, 37,000 enzyme responsible for R-subunit degradation. It is the major protein may well be underestimated in mouse lung tissue, for neutral protease activity detectable in mouse lung cytosol (7)

CANCER RESEARCH VOL. 45 AUGUST 1985 3682

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1985 American Association for Cancer Research. CAMP RECEPTORS IN LUNG TUMORS and has recently been shown to copurify with cAMP-dependent 16. Stewart, H. L., Dunn, T. B., Snell, K. C., and Deringer, M. K. Tumours of the respiratory tract. In: V. S. Turusov (ed.). Pathology of Tumours in Laboratory protein kinase from rat skeletal muscle (40). Such a close asso Animals,pp. 251-288. Lyon: l. A. R. C., 1979. ciation could allow for a highly synergistic effect of cAMP and 17. Svoboda, D. J. Ultrastructure of pulmonary adenomas in mice. Cancer Res., Ca2+ on kinase activation. In the absence of cAMP, Ca2+ would 22: 1197-1201,1962. 18. Kauffman, S. L., Alexander, L., and Sass, L. Histologie and ultrastructural have little effect on the kinase, since R, in the holoenzyme is features of the Clara cell adenomaof the mouse lung. Lab. Invest., 40: 708- relatively resistant to proteolysis (7). In the presence of even 716,1979. small amounts of cAMP, however, an increase in Ca2+ concen 19. Beer, D. G., and Malkinson,A. M. Localization of specific [3H)dexamethasone binding in urethan-inducedmouse lung tumors. Cancer Res., 44: 3546-3553, tration could result in a long-lasting activation of large amounts 1984. of cAMP-dependent kinase. If bound cAMP is indeed released 20. Shimkin, M. B., and Polissar/M. J. Some/quantitative observations on the induction and growth of primary pulmonar^ tumors in strain A mice receiving when free R, is cleaved to an M, 37,000 protein, then a positive urethan, J. Nati. Cancer Inst., 76: 75-9771955. feedback loop has been created by which a few molecules of 21. Kimura, I. Progression of pulmonary tumor in mice. 1. Histological studies of primary and transplanted pulmonary/umors. Acta Pathol. Jpn., 27: 13-56, cAMP could be continually recycled, acting catalytically to cause 1971. the continued release and cleavage of more R-subunits. This 22. Shimkin, M. B., and Stoner, G-. D. Lung tumors in mice: application to would allow the Ca2+-dependent protease to greatly magnify the carcinogenesisbioassay.Adv. Cancer Res., 27: 1-58, 1975. 23. Kauffman, S. L. Histogenesis of the papillary Clara cell adenoma. Am. J. effects of even small fluctuations in the intracellular concentration Pathol., 703: 174-180, 1981. of cAMP. 24. Dobbs, L. G., Geppert, E. F., Williams,M. C., Greenleaf,R. D., and Mason, R. J. Metabolic properties and ultrastructure of alveolar type II cells isolated with elastase. Biochim. Biophys. Acta, 678: 510-523, 1980. 25. Stoner, G. D., Kikkawa, Y., Kniazeff,A. J., Miyai, K., and Wagner, R. M. Clonal REFERENCES isolation of epithelialcells from mouse lung adenoma. Cancer Res., 35:2177- 2185, 1975. 1. Cho-Chung,Y. S. On the mechanismof cyclic AMP-mediatedgrowth arrest of 26. Lowry, O. H., Rosebrough, N. J., Fan-, A. L., and Randall, R. J. Protein solid tumors. Adv. Cyclic Nucleotide Res., 72:111-121,1980. measurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275, 2. Castagna, M., Takai, Y., Kaibuchi, K., Sano, K., Kikkawa, U., and Nishizuka, 1951. Y. Direct activation of calcium-activated, phospholipid-dependentprotein ki 27. Gilman, A. G. A protein-binding assay for adenosine 3',5'-cyclic monophos- nase by tumor-promoting phorbol esters. J. Biol. Chem., 257: 7847-7851, phate. Proc. Nati. Acad. Sci. USA,67: 305-312,1970. 1982. 28. Malkinson,A. M., Gharrett,A. J., and Hogy, L. Microheterogeneityof adenosine 3. Cooper, J. A., and Hunter, T. Regulationof cell growth and transformation by cyclic monophosphate-dependentproteinkinasesfrom mouse brainand heart. tyrosine-specific protein kinases: the search for important cellular substrate Biochem.J., 775: 365-375, 1978. proteins. Curr. Top. Microbiol. Immunol., 707:125-161,1983. 4. Malkinson,A. M., Gunderson, T. J., and McSwigan, C. E. Protein phosphoryl- 29. Walter, U., Costa, M. R. C., Breakfield, X. 0., and Greengard, P. Presenceof free cyclic AMP receptor protein and regulation of its level by cyclic AMP in ation in normal and neoplastic development. Cyclic AMP-dependent protein neuroblastoma-gliomahybridcells. Proc. Nati.Acad. Sci. USA,76:3251-3255, kinase activity in urethane-induced pulmonary tumours. Biochem. J., 768: 319-322,1977. 1979. 30. Hoyer, P. B., Owens, J. R., and Haley, B. E. Use of photaffinity probes to 5. Malkinson, A. M., Hogy, L, Gharrett, A. J., and Gunderson,T. J. Ontogenetic elucidatemolecularmechanismsof nucleotide-regulatedphenomena.Ann. NY studies of cyclic AMP-dependent protein kinase enzymes from mouse heart Acad. Sci., 346: 280-301, 1980. and other tissues. J. Exp. Zool., 205; 423-432, 1978. 31. Builder, S. E., Beavo, J. A., and Krebs, E. G. The mechanismof activation of 6. Malkinson, A. M., and Butley, M. S. Alterations in cyclic adenosine 3',5'- bovine skeletal muscle protein kinase by adenosine 3':5-monophosphate. J. monophosphate-dependentprotein kinases during normal and neoplasticlung Biol. Chem., 255: 3514-3519,1980. development.Cancer Res., 41: 1334-1341,1981. 32. Ogreid, D., and Doskeland, S. 0. Protein kinase II has two distinct binding 7. Beer, D. G., Butley, M. S., and Malkinson, A. M. Developmentalchanges in the endogenous Ca2*-stimulatedproteolysis of mouse lung cAMP-dependent sites for cyclic AMP, only one of which is detectable by the conventional membrane-filtrationmethod. FEBS Lett., 727: 340-344,1980. protein kinases. Arch. Biochem. Biophys., 228: 207-219, 1984. 33. Ogreid, D., and Doskeland,S. 0. Activation of protein kinase isozymes under 8. Butley, M. S., Beer, D. G., and Malkinson, A. M. Functional changes in the regulatory subunit of the type II cyclic adenosine 3':5'-monophosphate-de- new physiological conditions. Evidence that both types (A and B) of cAMP binding sites are involved in the activation of protein kinase by cAMP and 8- pendent protein kinase isozyme during normal and neoplastic lung develop Na-cAMP.FEBS Lett., 750: 161-166, 1982. ment. Cancer Res., 44: 2689-2697,1984. 34. Cho-Chung,Y. S., Clair, T., Yi, P. N., and Parkison, C. Comparative studies of 9. Butley, M. S., and Malkinson, A. M. Co-elution of type I and type II cAMP- cyclic AMP binding and protein kinase in cyclic AMP-responsive and -unre dependent protein kinase holoenzymes in high salt fractions from DEAE- sponsive Walker 256 mammarycarcinomas.J. Biol. Chem., 252; 6335-6341, cellulose.J. Biol. Chem., 259: 6001-6007, 1984. 1977. 10. Hofmann, F., Beavo, J. A., Bechtel, P. J., and Krebs, E. G. Comparison of 35. Potter, R. L., and Taylor, S. S. The structural domains of cAMP-dependent adenosine 3',5'-monophosphate-dependent protein kinases from rabbit skel protein kinase I. J. Biol. Chem., 255: 9706-9712, 1980. etal and bovine heart muscle. J. Biol. Chem., 250:7795-7801,1975. 36. Steinberg, R. A., and Agaro, D. A. Turnover of regulatory subunit of cyclic 11. Haley, B. E. Adenosine 3',5'-cyclic monophosphate binding sites. Methods AMP-dependent protein kinase in S49 mouse lymphoma cells. Regulation by Enzymol.,46: 339-346, 1977. catalytic subunit and analogs of cyclic AMP. J. Biol. Chem., 256: 10731- 12. Walter, U., Uno, I., Liu, A. Y-C., and Greengard, P. Identification, characteri 10734,1981. zation, and quantitative measurement of cyclic AMP receptor proteins in 37. Schubart, U. K., and Rosen,O. M. Characterizationof the protein kinases in a cytosol of various tissues using a photoaffinity ligand. J. Biol. Chem., 252: transplantable islet cell tumor of the Syrian hamster. Mol. Cell Endocrinol.,4: 6494-6500,1977. 167-182,1976. 13. Eriichman, J., Rosenfeld, R., and Rosen, O. M. Phosphorylation of a cyclic 38. Ledinko, N., and Chan, I-J. D. Increase in type I cyclic adenosine 3':5'- adenosine 3',5'-monophosphate-dependent protein kinase purified from bo monophosphate-dependentprotein kinase activity and specific accumulation vine cardiac muscle. J. Biol. Chem. 249: 5000-5003,1974. of type I regulatory subunits in adenovirus type 12-transformedcells. Cancer 14. Malkinson, A. M., Krueger, B. Â«.,Rudolph,S. A., Casnellie,J. E., Haley, B. E., Res., 44: 2622-2627, 1984. and Greengard, P. Widespread occurrence of a specific protein in vertebrate 39. Weber, W., and Hilz, H. Stoichiometryof cAMP binding and limited proteolysis tissues and regulation by cyclic AMP of its endogenous phosphorylation and of protein kinase regulatory subunits R, and RÂ».Biochem.Biophys. Res. dephosphorylation. Metab. Clin. Exp., 24: 331-341, 1975. Commun., 90: 1073-1081,1979. 15. Sorokin, S. P. The cells of the lungs. In: P. Nettesheim, M. G. Hanna, and J. 40. Zimmerman, U-J. P., and Schlaeffer, W. W. Kinase activities associated with W. Deatherage(eds.), Morphology of Experimental Respiratory Carcinogene- calcium-activatedneutral proteases. Biochem. Biophys. Res. Commun., 720: sis, pp. 3-43, U. S. A. E. C. Conf. 700501,1970. 767-774,1984.

CANCER RESEARCH VOL. 45 AUGUST 1985 3683

CANCER RESEARCH VOL. 45 AUGUST 1985 3684

SMALL ADENOMA *37K

-R, MEDIUM ADENOMA -37K

LARGE ADENOMA

ADENO- CARCINOMA

8 16 31 62 125 250 375 500 750 1000 [8N3-32P-cAMP] (nM) 2

Fig. 2. 8-N3-[Å“P]cAMPphotoincorporation into lung tumor cylosol. Cytosols from lung tumors at various stages of tumor progression were photolabeted using a range of 8-N3-[Å“P]cAMPconcentrations.Proteins were separated on denaturing gels, and radioactivity was detected autoradiographicallyas described in "Materials and Methods." Bands identified as Ri and RÂ«wereshown to comigrate with radiolabeled R-subunits from protein kinase isozymes previously separated by OEAE-cellulose chromatography (6).The RÂ»Pbandcomigrates with an M, 56,000 protein obtained after incubationof DEAE-cellulosePeak IIfractions with [7-Å“P]ATP(8).Two proteotytic fragments with molecularweights of 45,000 and 37,000 are also photolabeled.

CANCER RESEARCH VOL. 45 AUGUST 1985 3685

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1985 American Association for Cancer Research. Changes in Cyclic Adenosine 3′:5′-Monophosphate-dependent Protein Kinases during the Progression of Urethan-induced Mouse Lung Tumors

Martin S. Butley, Gary D. Stoner, David G. Beer, et al.

Cancer Res 1985;45:3677-3685.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/45/8/3677

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/45/8/3677. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.