Characterization of cyclic adenosine 3':5'-monophosphate-dependent protein kinase isozymes in normal and neoplastic fetal rat brain cells

[CANCER RESEARCH 45, 2578-2583, June 1985]

Characterization of Cyclic Adenosine 3':5'-Monophosphate-dependent Protein

Kinase Isozymes in Normal and Neoplastic Fetal Rat Brain Cells1

Roald Ekanger, Dagfinn 0greid,2 Ola Evjen, Olav Vintermyr, Ole Didrik Laerum, and Stein Ove Doskeland

Cell Biology Research Group, Institute of Anatomy, University of Bergen, Ã•rstadveien 19, N-5000 Bergen [R. Â£,D. 0., O. E., O. V., S. 0.0.], and Department of Pathology,The Gade Institute, University of Bergen, Haukeland Hospital, N-5016 Bergen [0. D. L], Norway

ABSTRACT

Fetal brain cells from rats given a transplacental pulse of W-ethyl-N-nitrosourea progressively acquire malignant characteris

tics and differentiate when grown in vitro. One aspect of thisdifferentiation is a decreased morphological response to cyclicadenosine 3':5'-monophosphate (cAMP). In the present study,

we have characterized and compared the isozymes (I, II) ofcAMP-dependent protein kinase in fetal brain cells and in the

neoplastically transformed, dedifferentiated BT5C glioma cellline. This is a first approach to find the mechanism behind thesubresponsiveness of such cells towards cAMP. It is also partof a broader investigation of the cAMP effector system in cellsshowing various rates of normal and malignant growth.

We found the regulatory and catalytic subunits of cAMP-

dependent protein kinase to be expressed to a similar degree inboth cell types. Sixty % of the enzyme was located in the 30,000x g supernatant. The glioma cell line had a significantly higherratio (1.2) between protein kinase I and II than did the normalfetal cells (0.5). This difference in isozyme distribution was notapparent using conventional methods for enzyme separation anddetection, the use of specific antibodies being essential for thatpurpose. Of the chromatographically separated forms (a, b) ofprotein kinase II, Form Ila was selectively decreased in the gliomacell line.

The alterations of the protein kinases in the glioma cell linedescribed above may be of importance for some of the neoplastiaproperties of these cells. However, the subdued response ofSuch cells towards cAMP is not explained since the concentrations of cAMP or its analogues required for activation of thekinases were similar for the enzymes from normal and neoplastically transformed cells.

INTRODUCTION

A number of studies point to cAMP3 as an important regulator

of cellular growth and differentiation. Analogues of cAMP (oragents which increase the intracellular level of the nucleotide)cause profound effects on growth properties and morphologicalcharacteristics of cultured neoplastic cells (1-7). Moreover, ma

lignant transformation in vitro has been associated with analtered cAMP metabolism (8-11).

The first step in the action of cAMP in eukaryotic cells involves

1This work was supported by the Norwegian Cancer Society (LMK).2To whom requests for reprints should be addressed.'The abbreviations used are: cAMP, cyclic adenosine 3':5'-monophosphate;

cAKI, cAMP-dependent protein kinase type I; cAKII, cAMP-dependent proteinkinase type II; R1, regulatory subunit of cAKI; R", regulatory subunit of cAKII; C,catalytic subunit of either isozyme; HEPES, 4-(2-hydroxyethyl)-1-piperazineeth-

anesulfonic acid.Received 11/27/85; revised 2/13/85; accepted 2/15/85.

stimulation of cAMP-dependent protein kinase (12-16). Most

mammalian cells contain 2 types of this enzyme (17), isozymetype I (cAKI) and isozyme type II (cAKII). Both forms are tetra-meric with 2 cAMP-binding regulatory (R) subunits and 2 C-

moieties. Upon binding of cAMP, the enzyme dissociates, releasing the active catalytic subunits (13). The difference between theisozymes has been located to their R-subunits, the C-subunitsapparently being identical (15-18). Little is known about the

separate biological functions of cAKI and cAKII, but the highproportion of cAKI in several physiological and pathological conditions associated with rapid growth (19-24) has led to the

suggestion that cAKI may be a positive effector of growth atleast in certain cells (25), whereas cAKII may be important in theprocesses of cellular differentiation and growth inhibition (26,27).

The addition of exogenous cyclic nucleotides to cultured neuroblastoma cells and some glioma cell lines results in a more"differentiated" morphology and decreased growth rate (2,6,7).

The present study deals with BT5C glioma cells, transformed invitro after an in vivo transplacental pulse of the carcinogen A/-ethyl-W-nitrosourea. This cell line responds considerably less toanalogues of cAMP than do untreated fetal brain cells.4 In

addition, neoplastic transformation in these cells in vitro is accompanied by less response to glia maturation factor, a proteinwhich promotes astrocytic differentiation by activation of theadenylate cyclase system (5). This resistance could be due tomutational alterations of regulatory or catalytic subunits of thekinase as described for Chinese hamster ovary cells and severalmutant S49 mouse lymphoma cell lines (28-32), or it could be

distal to the level of kinase action as suggested for a few of themutant S49 lymphoma cell lines (33).

The aim of the present study was to characterize the cAMP-

dependent protein kinase isozymes from normal fetal brain cellsand BT5C glioma cells. The conventional Chromatographie separation of cAKI and cAKII was supplemented by recently developed immunoseparation techniques to allow the separate quantification of the 2 isozymes.

MATERIALS AND METHODS

Chemicals. Cyclic [5':8-3H]AMP (56 Ci/mmol) and [y-*P]KTP (4000

Ci/mmol) were from the Radiochemical Centre, Amersham, United Kingdom. Protein A-Sepharose CL-4B was purchased from Pharmacia, Uppsala, Sweden. Agarose-ethane-cAMP (for the affinity purification of R1)and agarose-hexane-cAMP (for R") were from PL Biochemicals, St. Goar,West Germany. cAMP, A/B-monobutyryl-cAMP, A/6,O2'-dibutyryl-cAMP,

8-bromo-cAMP, 8-amino-cAMP, 8-(6-aminohexyl)amino-cAMP, mixedhistones (type IIS), 1-methyl-3-isobutylxanthine, benzamidine, soybeantrypsin inhibitor, aprotinin, and kemptide (phosphate acceptor heptapep-

4O. D. Laerum and M. F. Rajewsky, unpublished observations.

CANCER RESEARCH VOL. 45 JUNE 1985

2578

Research. on November 13, 2015. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

PROTEIN KINASE ISOZYMES IN CULTURED BRAIN CELLS

tide; leucylarginylarginylalanylserylleucylglycine) were from Sigma Chemical Co., St. Louis, MO. Tissue culture supplies were from Costar,Cambridge, MA whereas the growth media were obtained from FlowLaboratories, Irvine, Scotland. Other reagents were from sources de

scribed earlier (34).Cell Cultures. Primary cultures of fetal brain cells were prepared from

untreated 18th-day BD IX fetuses as described earlier (35). Procedures

for establishment and maintenance of BT5C glioma cells have been givenelsewhere (36), and various biological and morphological properties ofthis and similar cell lines were studied extensively (37). The cells weregrown in plastic dishes at 37Â°Cin humidified air containing 5% CO2. TheEagle-Dulbecco's medium used was fortified with 10% calf serum and a

4-fold concentration of nonessential amino acids and contained penicillin

(50 units/ml) and streptomycin (50 ^g/ml). The BT5C cell line wasregularly screened for Mycoplasma infection.

For passaging, the cells were harvested by trypsinization (0.25%trypsin in phosphate-buffered saline for 15 min) and diluted 1:3 to 1:4.

The glioma cells analyzed in the present study were passages > 100.Preparation of Cell Extracts. Unless otherwise stated, cells at an

early stationary phase of growth were used. Monolayer cultures wererinsed 3 times at room temperature with phosphate-buffered saline anddrained well. All subsequent steps were performed at 0Â°C.Homogeni-zation buffer (1 ml/107 cells) containing 15 mw Tris-HCI (pH 7.4), 4 mm

EDTA, 0.25 mM sucrose, 20 HIM benzamidine, soybean trypsin inhibitor(0.1 mg/ml), aprotinin (0.04 mg/ml), and 20 mm 2-mercaptoethanol wasadded, and the cells were wiped from the culture dishes with a rubber-

tipped glass rod. Homogenization was for 10 s in a Polytron homogenizerat a setting of 4. The high-speed supernatant fraction was prepared bycentrifugation at 30,000 x gmaagefor 15 min. The "paniculate" fraction

was prepared by resuspending the pellet to the volume of the originalhomogenate in homogenization buffer (see Table 1).

DEAE-Cellulose Chromatography. High-speed supernatants from

homogenates of normal and transformed brain cells were prepared inparallel, and 2 volumes of ice-cold distilled water and 1 volume of Tris-buffer [15 mw Tris-HCI (pH 7.4), 5 mM EDTA, and 20 HIM 2-mercaptoethanol] were added. Each solution was passed through a DEAE-

cellulose column (1 x 7.5 cm) preequilibrated with the same Tris buffer.After a washing with 0.5 liter of Tris buffer, the enzyme activities wereeluted by a linear gradient formed from 0.025 liter of Tris buffer and0.025 liter of Tris buffer with 0.3 M NaCI. Fractions (1.5 ml) were collectedinto tubes with 50 /Â¿Iof a solution containing bovine serum albumin andsoybean trypsin inhibitor (both at 5 mg/ml). During the whole procedure,care was taken to keep the Chromatographie conditions similar forsamples from normal and malignant cells.

Preparation of Antisera against the Regulatory Moieties of cAKIand cAKII. Partially purified cAKI and cAKII were prepared from ratskeletal muscle by a procedure similar to that described for rabbit muscle

Table 1cAMP-binding capacity and protein kinaseactivity in paniculate andpostparticulate fractions of normal and neoplastiafetal brain cells

The cAMP-binding capacity and protein kinase activity (in the absence andpresenceof cAMP) were assayed as described in "Materials and Methods."

Fetal brain cells (n = 8)Homogenate30,000 x g sediment30,000 x g supernatant

BT5C cells (n = 10)Homogenate30,000 x g sediment30,000 x g supernatantcAMP-binding

capacity (prnol/mg of homogenateprotein)5.3

Â±0.54a

2.0 Â±0.363.0 Â±0.605.7

Â±0.872.2 Â±0.443.2 Â±0.62Protein

kinase activity (pmol Å“Ptransferred into histone/min/mg

of homogenateprotein)-cAMP22

Â±3.28.2 Â±1.6

12.1Â±2.727

Â±4.710 Â±2.214 + 3.1+

CAMP150Â±14

56 Â±8.189Â±9.2152

Â±1749 Â±7.294 Â±12

'Mean Â±SE for 8 and 10 cultures, respectively. Samples were assayed in

triplicate, and the results are expressed per mg of homogenate protein.

(34) and applied tp affinity columns (containing 0.5 ml of cAMP-agaroseresin). After a washing with 100 ml of 15 mM HEPES-NaOH (pH 7.0)containing 5 mM EDTA, 20 mM 2-mercaptoethanol, 2 M NaCI, and 0.5

mM AMP, the columns were stoppered, 2 ml of 20 mw cAMP in HEPESbuffer were added, the tube contents were mixed, and the gel wasallowed to settle and left for 14 h at 2Â°Cbefore the elution of R-subunit.

R' or R", affinity purified as described above, was given to rabbits as

multiple s.c. injections (0.1 to 0.3 mg of R-subunit) with intervals of 3

weeks. For the first 2 injections, the antigen was emulsified in completeFreund's adjuvant (Difco Laboratories, Detroit, Ml), for later injections in

incomplete adjuvant. The animals were bled when their sera ceased toshow increased titers against the antigen, generally after 5 to 8 immunizations. The binding capacity was about 0.1 mg R-subunit/ml serum.The cross-reactivity (binding of R' by antiserum against R" and vice versa)was 0.2 to 2% using R1and R" isolated from rat muscle.

Antibody specificity was also demonstrated by electrophoretic transferof samples from sodium dodecyl sulfate-polyacrylamide gels to nitrocel

lulose filters, followed by an immunolabeling procedure. With this method,anti-R1 antibody specifically recognized a M, 48,000 protein, whereas aM, 54,000 protein was specifically labeled by anti-R" antibody (data not

shown).Assay of [3H]cAMP Binding and Quantification of R1and R" Isopro-

teins. Samples (0.2 ml) of high-speed supernatants or DEAE-cellulosefractions were first incubated for 90 min at 22Â°C in 0.8 ml of 15 mM

HEPES-NaOH (pH 7.0) containing 20 mw EDTA, 0.4 mM 1-methyl-3-

isobutylxanthine, 0.1 mM adenosine, 0.1 mM AMP, 10 mM benzamidine,bovine serum albumin (0.5 mg/ml), 20 HIM 2-mercaptoethanol, 0.5 mMdithiothreitol, and 40 nM [3H]cAMP to ensure exchange of endogenouscAMP for [3H]cAMP. Then histone (final concentration, 0.7 mg/ml) was

added to ensure dissociation of cAKII which increases the apparentaffinity of R" for cAMP (22), and the incubation was continued for another90 min at 0Â°C.One aliquot (0.2 ml) was removed for the determinationof total protein-bound [3H]cAMP using an earlier described ammonium

sulfate precipitation method (38). Two other aliquots (0.2 ml) were eachmixed with 0.1 ml of a suspension containing 0.1 mg of Protein A-

Sepharose that had been preincubated (for 1 h Â¡nHEPES buffer) withantiserum (dilution, 1:10) against R1and R". The suspensions were gentlyrocked for 2 h at 2Â°Cand transferred to (empty) polypropylene columns

(0.8 x 10 cm, with a 10-ml reservoir, from Bio-Rad), which by filterretained the R â€¢[3H]cAMP complex bound to Protein A-Sepharose via theantibody bridge. Free isotope as well as R-[3H]cAMP not bound to

Protein A-Sepharose were removed by washing the columns with, successively, 10 and 3 ml of 15 mM Tris-HCI (pH 8.5) containing 10 mMEDTA, 2.4 M glycerol, 1 M (NH4)zS04,20 HIM2-mercaptoethanol, and 0.5HIM dithiothreitol, followed by a rinse with 3 ml of 5 mM HEPES-NaOH(pH 7.0) containing 1 mM EDTA. The [3H]cAMP retained on the column(representing the cAMP-binding capacity of R1or R") was quantitatively

eluted with 0.6 ml of 1 M acetic acid, and the amount of isotope wasdetermined by liquid scintillation counting.

Assay of Protein Kinase Activity and of the Potency of cAMPAnalogues as Kinase Activators. Samples to be tested for phospho-transferase activity were desalted by passage through Sephadex G-25columns (0.8 x 6 cm) equilibrated with 15 mM HEPES-NaOH (pH 7.0),0.3 mM EDTA, 20 mM 2-mercaptoethanol, and 0.5 mM dithiothreitol. Theincubations were at 30Â°Cin 15 mM HEPES-NaOH (pH 7.0) with 10 mM

magnesium acetate, 0.3 mM ethylene glycol bis(/<-aminoethyl ether)-A/,W,A/',A/'-tetraacetic acid, 0.1 HIM EDTA, 5 ,M [f-^PJATP (1 ^Ci/ml),

20 mM 2-mercaptoethanol, 0.1 mM heptapeptide (leucylarginyl-

arginylalanylserylleucylglycine; kemptide) and bovine serum albumin (0.5mg/ml) in the absence or presence of 2 Â¿IMcAMP. After 5 and 10 min ofincubation, samples were withdrawn, and |32P]phosphoheptapeptide

was determined by the procedure of Roskoski (39). In some cases,histone (0.7 mg/ml) was the substrate, and [^PJphosphohistone was

determined according to the method of Fossberg et al. (22).The ability of analogues of cAMP to activate protein kinase from

DEAE-cellulose fractions was tested at concentrations of cAKI or cAKII


2579




of 0.01 to 0.5 RM with respect to cAMP-binaing capacity. Cyclic nucfeo-

tides were present at 9 different concentrations covering a 256-fold

concentration range. The concentration of cyclic nucteotide required tohalf-maximaily activate the kmase (K.) was determined from plots of

fractional kmase activity (activity ratio) versus log cyclic nucleotide concentration (40). At least 3 separate determinations were performed foreach analogue.

Autophosphorylation of cAK. The incubation conditions were essentially as described (22). Samples (desalted fractions from the DEAE-cellulose chromatography) were diluted in 15 HIM HEPES-NaOH (pH 7.0)with 0.1 mw ethylene glycol bis(,-i-aminoethyl ether)-N.W.N'.A/'-tetra-

acetic add, 30 Â«MEDTA, 5 mw magnesium acetate, bovine serumalbumin (0.5 mg/ml), and 20 mM 2-mercaptoethanol. Incubations werecarried out at 0Â°C,and the reactions were started by adding h -3<!PIATP

to a final concentration of 2 MM (total volume, 0.5 ml). Aliquota (80 pi)were removed after different periods of incubation, and mixed with 1.12ml of ice-cold 25 mM sodium phosphate buffer (pH 7.0) with 20 mw

EOTA, 100 mM NaF, and 0.25 mM ATP. Then 200 n\ of a suspension ofcharcoal (20 mg/ml) were added. The supernatant obtained after centrif-

ugation (7000 x g,.aaqe for 10 min, 2 times) was precipitated with 10%(w/v) ice-cold trichioroacetic acid containing 20 mM phosphate and 20

mM pyrophosphate. The precipitate was collected on filters and countedas described (22).

RESULTS AND DISCUSSION

The submits (R and C) of the CAMP-dependent protein kinase

are normally expressed at a proportion of 1:1, and their concentration is fairly constant between tissues (41). However, severaltumor cell lines show a disproportionately high expression of theregulatory subunit of protein kinase (42-46). The biological sig-

Chart 1 Protein kmase activities in DEAE-cellulose fractions of normal braincells {A) and glioma cells (B). Extracts of normal fetal brain cells or glioma cellswere chromatographed on DEAE-cellulose as described in "Materials and Methods." The phosphotransferase activity of the fractions (left ordinate) was measured

in the absence (O) and the presence (â€¢)of 2 /M CAMP The salt gradient (rightordinate) is indicated by the straight diagonal line.

a 2

a.< 1

O.2

Uo

0.1

1O 20

Fraction number

30 4O

Chart2. Effect of incubation conditions on the apparent |3H|cAMP-bindingcapacity of DEAE-cellulose fractions. Extract of glioma cells was chromatographedas detailed in "Materials and Methods." One senes of samples of the DEAE-cellulose fraction was incubated with 40 nu pHJcAMP for 90 nun at 22Â°C. Next,

historie was added to a final concentration of 0.7 mg/ml, and the incubation wascontinued for another 90 min at 0Â°C(â€¢).In other series, the incubation step at22Â°Cwas omitted (A). The amount of f^HJcAMP bound (left ordinate) was determined by the ammonium sulfate precipitation method (see "Materials and Methods"

for details). The salt gradient (right ordinate) te indicated by the straight diagonalline.

nificance of such free R-subunit is unknown, but it may serve asa "sink" for cAMP and thus make the cells less sensitive to the

cyclic nucleotide. Alternatively, tumor cells may lose or decreasetheir sensitivity to cAMP by a mutational loss or decrease of thecAMP-dependent protein kinase (3,28,29). The data of Tabte 1show no difference in cAMP-dependent protein kinase activity orcAMP-binding capacity between fetal brain cells and the BT5C

glioma cell line. Furthermore, the relation between protein kinaseactivity and cAMP-binding capacity was the same as for purified

protein kinase holoenzymes from rat muscle (data not shown).We therefore conclude that the malignant, rapidly growing gliomacells have a normal level of cAMP-dependent protein kinase, the

subunits of which are equally expressed. The decreased response of these cells toward cAMP4 cannot be thus explained

by a reduced content of cAMP-dependent protein kinase or byan excess of free R-subunit acting as a "sink" for cAMP. Also,

the fraction of the cAMP-dependent protein kinase associated

with the particulates was the same (40%) in normal fetal cellsand the glioma cell line (Table 1).

The biological significance of the presence of 2 distinct iso-zyme forms (I, II) of cAMP-dependent protein kinase (17,18) still

remains enigmatic. One of the more consistent findings in thesearch for separate biological functions for the isozymes hasbeen the increased ratio between kinases I and II in rapidlyproliferating versus quiescent cells (20, 22, 24,25, 47). Recentlyexceptions to this rute have been observed (48), and methodological difficulties in separating the isozymes have been pointedout (22, 49, 50). In the case of the fetal brain cells and theirmalignant, more rapidly dividing counterpart, the conventionalanalysis of isozyme distribution (measurement of the proteinkinase activity of the DEAE-cellulose fractions) showed 2 major

peaks (I and II) of activity (Chart 1). Peak I eluted as a single


2580




~ O.8

I

â€¢a

Q.Z

O.2

B

(XI

Ooz

-0.2

-O.1

10 20 3O 40

Fraction numberCharts. Immunotogical separation of [*H]cAMP bound to Ff and R" in samples

of DEAE-cellulose fractions. Extracts of normal fetal eels (A) and glioma eels (B)

were chromatographed as described in the legend to Chart 2. Samples of thefractions were sequentially incubated with 40 ran 13H|cAMP and historiÃ©as de

scribed in the legend to Chart 1. ABquots were tested for total pHJcAMP-bindingcapacity by the ammonium sulfate precipitation method (â€¢)or incubated withantiserum against R1and R" (see 'Materiate and Methods' for delate) to determinethe amount of pHJcAMP-R1 <D) and [Â»HJcAMP-R1(A). The salt gradient (right

ordinate) is indicated by the straight diagonal line.

entity at 0.05 M Nad, whereas Peak II was composed of 2 partlyoverlapping subpeaks (Ila, lib) editing at 0.15 and 0.18 M NaC!,respectively (Chart 1). This conventional analysis thus did notreveal any difference in isozyme distribution between the 2 cellstypes. However, an analysis of the pHJcAMP binding in theDEAE-cellulose fractions disclosed a peak of activity (Ib) that

eluted between the peaks of kinase activity and was morepronounced for the glioma extract (Charts 2 and 3). This mightrepresent (a) the free regulatory moiety of one of the cAMP-dependent protein kinases or (b) a hotoenzyme with a "mute"

catalytic subunit, as described recently by Reed era/. (51). Sincethe apparent binding capacity of this peak increased underconditions facilitating the exchange of bound cAMP for [3H|cAMP

(Chart 2), it probably represented free R-subunit complexed withendogenous cAMP rather than a "mute' holoenzyme.

To find whether this free R-subunit (Peak Ib) represented R1

and R", an immunoassay was used. This assay (see "Materialsand Methods') is modified from those described by Fleischer ef

al. (52) and Weber ef al. (46). The salient difference from theassay of Ref. 46 is that the amount of the R â€¢[3H]cAMP complex

is measured directly and is not inferred from the amount ofpHJcAMP not pelleted by antibody-Protein A-Sepharose and

active charcoal. The blank value of the present assay is extremelytow, i.e., less than 0.01% of the added free [3H|cAMP is elutedin the fraction representing the |'H|cAMP bound to R-subunit.Using this immunoassay, Peak Ib was found to represent R1(Chart 3). The amount of R' and R* could be determined with thismethod in DEAE-cellulose fractions as well as directly in unfrac-

tionated 30,000 x g supernatant. Several such experiments gavea R':R" ratio in the ranges of 1.1 to 1.4 for the glioma cells and

0.4 to 0.6 for the fetal brain cells. This is another example of anincreased proportion of cAKI in a neoplastically transformed cellline. It has not been answered if this increase is ascribable to themore rapid proliferation only (for details on doubling times, see

15 25

Fraction number

Chart 4. Chromatographie profiles (DEAE-cellulose) of cerebral cortex extractsfrom neonatal, juvenile, and adult rats. High-speed supematants (30,000 x gr-iji}from cerebral cortex of newborn rats (A), 4-week-old animals (fl). or 15-month-oldanimals (C) were prepared and chromatographed on DEAE-cellulose essentiallylike the extracts from cultured cete (see 'Materials and Methods' for dÃ©faite).Samples of the fractions were incubated with |3H|cAMP, historie and appropriate

antiserum as described in the legend to Chart 3. The total prflcAMP-bindingcapacity (â€¢)as well as the amount of fHJcAMP bound to R1 (D) and R* (A) is

shown. The salt gradient (right ordinate) te indicated by the straight diagonal One.


2581




TableaPotency of cyclic nucleatidesas activators of protein kinase isozymes from skeletal muscle, cerebral cortex,

and BT5C glioma cellsProtein kinase isozymes from rat skeletal muscle, neonatal rat cerebralcortex, and BT5C glioma cells were

assayed for phosphotransferase activity with various concentrations of the cAMP analogue listed. Theconcentration of analogue required for half-maximalkinase activation (KJ is indicated. Each value representsthe mean of 3 experiments (range, Â±15%).

Rat skeletal muscle Rat cerebral cortex BT5C glioma cells

CyclicnudeotideCAMP

N", 02'-Dibutyryl-cAMPW-Monobutyryl-cAMP8-Bromo-cAMP8-Amino-cAMP8-(6-Aminohexyl)ammo-cAMPK.cAKI65

6,600753126

125K.

cAKII78

16,00023065

1101400K.cAKI72

24,000240

455172K.CAKII79

16,000330

6698

440K.cAKI82

14,000130

458737K.

cAKII84

20,000520

49280350

1.0

05

B

1O-' 1O-Â» 1O-Â»

Concwitrtrtlon of onoloQ (M)

TO-7 10-Â«

Chart 5. Activation by cAMP and by commonly used cAMP analoguesof kinaseisozymes from glioma cells and cerebral cortex from neonatal animals.The kinaseactivity is shown as a function of increasingconcentration of cAMP (A). 8-bromo-cAMP (O), and A/'-monobutyryl-cAMP (G).A and B, activation of protein kinase Ifrom neonatal cortex and glioma cells, respectively; C and D. activation of proteinkinase II from normal and malignantcells, respectively.

Ref. 37) or if it is a "basic" property of the chemically induced

malignant state. The stepwise in vitro transformation of BT5Ccells [cells from each of the stages are available (37)], makes itpossible to answer that question by determining the ratio between cAKI and cAKII and to correlate it to doubling time andmalignant index.

The finding of Chromatographie microheterogeneity of cAKII inthe glioma cell extract (see Subpeaks Ha and lib of Charts 1 to3) indicates that more than one subtype of cAKII may exist inone cell line. Subtype Ma was relatively more abundant in fetalbrain cells than in the glioma cells (Charts 1 and 3). As shown inChart 4, protein kinase Ila was also the dominating species inrat cerebral cortex of variously aged rats (neonatal, juvenile,adult). The lib subclass of cAKII eluted at higher ionic strengththan did subclass Ha and was presumably not a result of prote-

olysis of Peak Ha, since the omission of the protease inhibitorspresent in the homogenization medium shifted the Chromatographie elution of cAKII towards lower ionic strength. In fact, forglioma cell extracts prepared in the absence of protease inhibitors, R' and R" showed a considerable Chromatographie overlap.

The extent and rate kinetics of the autophosphorylation of peakHa and lib from the glioma cells was similar to that of proteinkinase II from rat cerebral cortex and rat skeletal muscle (data

not shown).Minimally different forms of R" may have separate biological

functions. Thus, Cho-Chung ef al. (53) observed minimal chargedifferences between R" in a mutant versus a "wild-type" cell line

of a chemically induced mammary carcinoma.In order to know if the species of protein kinase in the glioma

cells differed qualitatively from those of normal cells, we compared their ability to be activated by various concentrations ofcAMP as well as by some analogues of cAMP, including those(A/6,O2'-dibutyryl-cAMP, W6-monobutyryl-cAMP, 8-bromo-cAMP)

most commonly used in cell culture experiments. The activationconstants (Ka) were not significantly different between proteinkinases Haand lib whether the enzyme was derived from gliomacells or from cerebral cortex; therefore, only the means of thevalues for cAKIIa and cAKIIb are given in Table 2. In no case didthe Ka for any glioma isozyme differ by more than a factor of 3from the corresponding value for the isozyme from rat cerebralcortex (Table 2). Not only the K, but also the steepness of theplot of kinase activity versus cyclic nucleotide concentration wassimilar for enzyme prepared from the glioma cell line and othersources (Chart 5). An apparent Hill coefficient of about 1.5 wascalculated for the glioma enzymes, indicating that the positivecooperativity in the response to cAMP (12) was intact.

Although it cannot be excluded that an extended characterization (e.g., peptide mapping or sequencing) may disclose differences between the cAMP-dependent protein kinases in the

BT5C glioma cells and fetal brain cells, the data presented in thisstudy show that the glioma enzymes are present in normalamount and show normal cAMP responsiveness. The decreasedresponse of these cells towards cAMP is therefore presumablydue to an alteration "distal" to the activation of cAMP-dependent

protein kinase. Further studies will be carried out in search forthe step(s) in the chain of events, starting with the activation ofthe protein kinase by cAMP and ending up in altered morphologyand reduced proliferation rate, which is deficient in the gliomacell line.

REFERENCES

1. Cho-Chung, Y. S., Clair, T., Bodwin, J. S., and Berghotfer. B. Growth arrestand morphological change of human breast cancer cells by dibutyryl cAMPand L-arginine.Science (Wash. DC), 214:77-79,1981.

2. Edstrwn, A., Kanje, M., and Walum, E. Effects of dibutyryl cyclic AMP andprostaglandinE, on cultured humanglioma cells. Exp. Cell Res., 85:217-223,1974.

3. Gottesman, M. M., Lecam,A., Bukowski, M., and Pastan, I. Isolationof multipleclasses of mutants of CHO cells resistant to cyclic AMP. Somat Cell Genet..6: 45-61, 1980.


2582




4. Hamprecnt. B. Structural, electrophysiological, biochemical and pharmacological properties of neuroblastoma-glioma cell hybrids in cell culture. Int. Rev.Cytol., 49: 99-170,1977.

5. Haugen, A., Laerum, O. D., and Bock, E. Responsiveness of fetal rat braincells to glia maturation factor during neoplastia transformation in cell culture.Acta Patho!. Microbio!. Scand. Sect. A., 89:393-402,1981.

6. Prasad, K. N., and Hsie. A. W. Morphological differentiation of mouse neuroblastoma cells induced in vitro by dibutyryl adenosine 3':5'-cyclic monophos-

phate. Nat. New Btol., 233:141-142,1971.7. Prasad, K. N. Differentiation of neuroblastoma cells in culture. Biol. Rev., 50:

129-165,1975.8. Cho-Chung. Y. S. Mode of cyclic AMP action in growth control. In: B. S. Leung

(ed.), Hormonal Regulation of Experimental Mammary Tumours. Lancaster,England: MTP Press, Ltd., 1982.

9. Christoffersen, T., and Brenstad, G. O. A commentary on the role of cyclicnucleotides in cell growth and malignancy. Prog. Pharamacol., 4: 117-135,

1980.10. Friedman, D. L. Role of cyclic nucleotides in cell growth and differentiation.

Phystol. Rev., 56: 652-708,1976.

11. Pastan, l. H., Johnson, G. S., and Anderson, W. B. Role of cyclic nucleotidesin growth control. Annu. Rev. Biochem.. 44: 491-522,1975.

12. Daskeland, S. 0., and 0greid, D. Binding proteins for cAMP in mammaliantissues. Int. J. Biochem., 73:1-19,1981.

13. Flockhart, D. A., and Corbin, J. D. Regulatory mechanisms in the control ofprotein kinases. CRC Crit. Rev. Biochem., 9:133-186,1982.

14. Krebs, E. G. Protein kinases. Curr. Top. Cell. Regul., 5: 99-133,1972.15. Krebs, E. G., and Beavo, J. A. Phosphorylation-dephosphorylation of enzymes.

Annu. Rev. Biochem., 48: 923-960,1979.16. Nimmo, H. G., and Cohen, P. Hormonal control of protein phosphorylation.

Adv. Cyclic Nudeotide Res., 8:145-266,1978.17. Corbin, J. D., Keely, S. L., and Park, C. R. The distribution and dissociation of

cyclic adenosine 3':5'-monophosphate-dependent protein kinases in adipose,

cardiac and other tissues. J. Biol. Chem., 250: 218-225,1975.18. Hofmann, F., Beavo, J. A., Bechtel, P. J., and Krebs, E. G. Comparison of

adenosine 3':5'-monophosphate-dependent protein kinases from rabbit skel

etal and bovine heart muscle. J. Biol. Chem., 250: 7795-7801,1975.19. Costa, M., Gemer, U. W., and RÃ¼ssel,D. H. Cell cycle-specific activity of type

I and type II cyclic adenosine 3':5'-monophosphate-dependent protein kinases

in Chinese hamster ovary cells. J. Btol. Chem., 257:3313-3319,1976.20. Doskeland, S. 0., Kvinnsland, S., and Ueland, P. M. Protein kinases activated

by cAMP in the genital tract of spayed mice treated with oestradiol-17/3. J.Reprod. FÃ©rtil.,44: 207-216,1975.

21. Eppenberger, U., Biedermann, K., Handschin, J. C.. Fabbro, D., Kung, W.,Huber, P. R., and Reos, W. Cyclic AMP-dependent protein kinase type I andtype II and cyclic AMP binding in human mammary tumours. Adv. CyclicNudeotide Res., 72:123-128,1980.

22. Fossberg, T. M., Doskeland. S. O., and Ueland, P. M. Protein kinases in humanrenal cell carcinoma and renal cortex. A comparison of isozyme distributionand of responsiveness to adenosine 3':5'-cychc monophosphate. Arch.

Biochem. Biophys., 789:372-381,1978.

23. Handschin, J. C., and Eppenberger, U. Altered cellular ratio of type I and typeII cyclic AMP-dependent protein kinase in human mammary tumours. FEBSLett., 706:301-304,1979.

24. Lee, P. C., Radioff, D., Schwappe, J. S., and Jungmann, R. A. Testicularprotein kinases: characterization of multiple forms and ontogeny. J. Btol.Chem., 257: 914-921,1976.

25. Russell, D. H. Type I cyclic AMP-dependent protein kinase as a positiveeffector of growth. Adv. Cyclic Nudeotide Res.. 9: 493-506.1978.

26. Cho-Chung, Y. S. On the mechanisms of cyclic AMP-mediated growth arrestof solid tumours. Adv. Cydte Nudeotide Res., 72:111-121,1980.

27. McClung, J. K., and Kletzien. R. F. The effect of growth state on the activityof the protein kinase isoenzymes. An increase in type II cydic AMP-dependentprotein kinase is correlated with growth arrest in G1 phase. Biochim. Biophys.Acta, 678: 106-114,1981.

28. Coffino, P., Bourne, H. R., Insel, P. A., Melmon, K. L., Johnson, G., and Vkjne,J. Studies of cydic AMP action using mutant tissue culture cells. In Vitro(Rockvilte), 74: 140-145,1978.

29. Evain, D., Gottesman, M., Pastan, I., and Anderson, W. B. A mutation affectingthe catalytic subunit of cydic AMP-dependent protein kinase in CHO cells. J.Btol. Chem., 254: 6931-6937,1979.

30. Insel, P. A., Bourne, H. R., Coffino, P., and Tomkins, G. M. Cyclic AMP-

dependent protein kinase: Pivotal role in regulation of enzyme induction and

growth. Science (Wash. DC), 790:896-898,1975.

31. Singh, T. J., Roth, C., Gottesmann, M. M., and Pastan, I. Characterization ofcAMP-resistant Chinese hamster ovary mutants lacking type I protein kinase.J.Btol. Chem., 256:926-932,1981.

32. Steinberg, R. A., O'Farrell, P. H., Friedrich, U., and Coffino, P. Mutations

causing charge alterations in regulatory subunits of the cAMP-dependentprotein kinase of cultured S49 rymphoma cells. Cell, 70:381 -391,1977.

33. Lemaire, I., and Coffino, P. cAMP-induced cytolysis in S49 cells: selection ofan unresponsive "deathless" mutant. Cell, 77:149-155,1977.

34. Deskeland, S. O., and 0greid. D. Characterization of the interchain andintrachain interactions between the binding sites of the free regulatory moietyof protein kinase I. J. BkÂ».Chem., 259: 2291-2301, 1984.

35. Haugen, A., and Laerum, 0. D. Induced glial differentiation of fetal rat braincells in culture: an ultrastructural study. Brain Res., 750:225-238,1978.

36. Laerum, 0. D., and Rajewsky, M. F. Neoplastic transformation of fetal rat braincells in culture after exposure to ethylnitrosurea in vivo. J. Nati. Cancer Inst,55: 1177-1187,1975.

37. Laerum, 0. D Rajewsky, M. F., Schachner, M., Stavrou. D., Hagelid, K. G.,and Haugen, A. Phenotypic properties of neoplastic cell lines developed fromfetal rat brain cells in culture after exposure to ethylnitrosurea in vivo. Z.Krebsforsch., 89: 273-295,1977.

38. Deskeland, S. O., Ueland, P. M., and Haga, H. J. Factors affecting the bindingof H adenosine 3':5'-cyclic monophosphate to protein kinase from bovine

adrenal cortex. Biochem. J., 767:653-665,1977.39. Roskoski, R., Jr. Assay of protein kinases. Methods Enzymol., 99: 3-6,1983.40. Ogreid. D., Dsskeland, S. O., and Miller, J. P. Evidence that cydic nucleotides

activating rabbit musde protein kinase I interact with both types of cAMPbinding sites associated with the enzyme. J. Biol. Chem., 258: 1041-1049,

1983.41. Hofmann, F., Bechtel, P. J., and Krebs, E. G. Concentrations of cydic AMP-

dependent prtoein kinase subunits in various tissues. J. Btol. Chem., 252:1441-1447,1977.

42. Uu, A. Y-C., Chan, T. H., and Chen, K. Y. Induction of the regulatory subunitof type I adenosine cydic 3':5'-monophosphate-dependent protein kinase in

differentiated N-18 mouse neuroblastoma cells. Cancer Res., 47:4579-4587,

1981.43. Lohmann, S. M., Schwoch, G., Reiser, G., Pott, S., and Walter, U. Dibutyryl

cAMP treatment of neuroblastoma-glioma hybrid cells results in selectiveincrease in cAMP-receptor protein (R-l) as measured by monospecitic antibodies. EMBO J., 2:153-159,1983.

44. Prashad, N. Induction of free cyclic-AMP-binding protein by dibutyryl cydic

AMP in neuroblastoma cells. Cold Spring Harbor Conf. Cell Proliferation, 8:159-178,1980.

45. Walter, U., Costa, M. R. C., Breaketield. X. O., and Greengard, P. Presence offree cydic AMP receptor protein and regulation of its level by cydic AMP inneuroblastoma-glioma hybrid cells. Proc. NatJ.Acad. Sd. USA, 76:3251-3255,

1979.46. Weber, W., Schroder, H., and Hilz, H. Quantitation of heteroiogous protein

kinase subunits R I and R II with the aid of type specific antibodies. Biochem.Biophys. Res. Commun., 99: 475-483, 1981.

47. Gharrett. A. J., Malkinson, A. M., and Sheppard, J. R. Cydic AMP-dependentprotein kinases from normal and SV40-transformed 3T3 cells. Nature (Lond.),264: 673-675,1976.

48. Wittmaack, F. M., Weber, W., and Hilz, H. Isoenzymes of cAMP-dependentprotein kinase in developing rat liver and in malignant hepatic tissues. Eur. J.Biochem., 729:669-674,1983.

49. Butiey. M. S., and Malkinson, A. M. Co-elution of type I and type II cAMP-dependent protein kinase hotoenzymes in high salt fractions from DEAE-cellutose. J. Btol. Chem., 259:6001-6007,1984.

50. Malkinson, A. M., Beer, D. S., Wehner, J. M., and Sheppard, J. R. Elution ofthe regulatory subunit of cAMP-dependent protein kinase type I isozymederived from epididyrnal fat within the type II isozyme Chromatographie peak.Biochem. Biophys. Res. Commun., 772:214-220,1983.

51. Reed, J., Gagelmann, M., and Kinzel, V. Isolation and elucidation of somefunctional properties of the "mute" catalytic subunit of cAMP-dependent protein

kinase. Arch. Biochem. Biophys., 222:276-284,1983.52. Fleischer, N., Rosen, 0. M., and Reichlin, M. Radtoimmunoassay of bovine

heart protein kinase. Proc. Nati. Acad. Sd. USA, 73:54-58.1976.53. Cho-Chung, Y. S., Clair, T., Schwimmer, M., Steinberg, L., RegÃ³, J., and

Grantham, F. Cydic adenosine 3':5'-monophosphate receptor proteins in

hormone-dependent and -independent rat mammary tumors. Cancer Res., 47:1840-1846,1981.


2583



1985;45:2578-2583. Cancer Res Roald Ekanger, Dagfinn Øgreid, Ola Evjen, et al. and Neoplastic Fetal Rat Brain Cells-Monophosphate-dependent Protein Kinase Isozymes in Normal

′:5′Characterization of Cyclic Adenosine 3

Updated version

http://cancerres.aacrjournals.org/content/45/6/2578

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications


http://cancerres.aacrjournals.org/content/45/6/2578

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

mailto:[email protected]


Documents

Characterization of cyclic adenosine 3':5'-monophosphate-dependent protein kinase isozymes in normal and neoplastic fetal rat brain cells