ExperimentalGerontology.Vol. 27, pp. 287-300, 1992 Printedin the USA.All fightsreserved.

0531-5565/92$5.00 + .00 Copyright© 1992PergamonPressLtd.

M E C H A N I S M O F E N H A N C E D CYCLIC A M P S T I M U L A T I O N BY ISOPROTERENOL IN AGED HUMAN FIBROBLASTS

MICHAELF. ETHIER, ~ MAUREEN

MEDEIROS, 2 FRED

D.

ROMANO 2 and

JAMES G. DOBSON, JR.t'2 Departments of tMedicine and 2physiology, University of Massachusetts Medical School, Worcester, Massachusetts 01655

Abstract - - Human diploid lung fibroblasts (IMR-90) were used to investigate the

reported increase in B-adrenergic-stimulated cyclic adenosine 3', 5'-monophosphate (cAMP) levels in fibroblasts aged in culture. Under basal conditions cellular cAMP was 34.2 + 5.6 and 38.4 + 9. l pmol/mg protein in early (PDL 22-24) and late (PDL 4752) passage fibroblasts, respectively. Net release of cAMP from fibroblasts was 67.8 + 8.6 and 18.5 + 7.0 pmol/30 min/mg protein in early and late passage cultures, respectively. In confluent, early passage fibroblasts, cellular cAMP and net release of cAMP increased by 2.7-fold and 3.8-fold, respectively, after a 30 miD incubation in 2 ~,M isoproterenol. In confluent late passage fibroblasts, isoproterenol incubation increased cellular cAMP and net release of cAMP by 7.8-fold and 26. l-fold, respectively. Adenosine failed to inhibit isoproterenol-induced stimulation of cAMP in early or late passage fibroblasts. There was no passage-related difference in basal, isoproterenol, or forskolinstimulated adenylyl cyclase activity in crude fibroblast membrane preparations. The activity of cAMP-phosphodiesterase in sonicates of early and late passage IMR-90 was 9.61 + 1.15 and 5.81 + l . l l pmol/min/mg protein respectively. Measurements of cAMP in subconfluent early passage fibroblasts indicated that mechanisms related to the reduced cell density in confluent late passage IMR-90 may, in part, account for the enhanced isoproterenol-induced cAMP levels observed in these cultures. The results suggest that the remainder of the enhanced cAMP response to isoproterenol of in vitro aged fibroblasts may be due to a lower cAMP phosphodiesterase activity in these cells. Key Words: aging, human fibroblasts, cyclic AMP, adenylyl cyclase, phosphodiesterase

INTRODUCTION

THE OBSERVATIONthat human diploid fibroblasts have a limited proliferative capacity in culture (Hayflick, 1965) has led to the extensive use of these cells as an in vitro model for aging studies. As they approach the end of their in vitro life span these cells show a number of changes which are also found to occur in aged human cells in vivo (Hayflick, 1980). Correspondence to: M.F. Ethier, Department of Medicine, $6-720, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655. (Received 8 April 1991; Accepted 26 September 1991) 287

288

M.F. ETHIER et aL

Included in these changes is the augmentation of the B-adrenergic catecholamine-elicited increase of cellular cyclic adenosine 3',5'-monophosphate (cAMP) levels (Haslam and Goldstein, 1974; Manganiello and Breslow, 1974; Polgar et al., 1978). In our study, mechanisms which might enhance B-adrenergic stimulation of cAMP during in vitro aging were investigated in early and late (in vitro aged) passage cultures of human diploid lung fibroblasts (IMR-90). It has been postulated that cAMP release by cells may serve to modulate cellular levels of the cyclic nucleotide (Kelly et aL, 1978). If so, diminished cAMP release from in vitro aged fibroblasts could account for the enhanced levels of B-adrenergic-stimulated cellular cAMP reported to occur in late (compared to early) passage cultured human fibroblasts. In our study, cellular cAMP as well as cAMP released into the medium were measured in early and late passage human diploid lung fibroblasts after exposure to the B-adrenergic agonist isoproterenol. Enhancement of cAMP by B-agonists is dependent on binding of the agonist to a cell surface receptor. Both the number of B-receptors (Pochet et aL, 1982) and the cAMP response to B-agonists (Kelly and Butcher 1974) have been reported to diminish with increasing cell density in fibroblast cultures. Since a decreased population density at confluency is a characteristic feature of late versus early passage human fibroblasts (Simons, 1967; Schneider and Mitsui, 1976), increased stimulation of cAMP by isoproterenol might be expected in these late passage cells. To determine the extent of the effect of cell density on increased stimulation of cAMP in late passage fibroblasts, B-adrenergic stimulation of cAMP levels was determined and compared in early versus late passage IMR-90 cultures of equivalent cell densities. Potential changes with in vitro age of two major enzymes affecting cAMP levels were also explored as possible mechanisms for enhanced cAMP levels in response to B-agonists in late (compared to early) passage fibroblasts. Isoproterenol increases cellular cAMP by stimulating the activity of membrane-bound adenylyl cyclase (LaMonica et aL, 1985). Phosphodiesterase is responsible for degradation of cAMP within the cell (Thompson et al., 1979). In our study, adenylyl cyclase and cAMP phosphodiesterase activity were determined and compared in cultures of early versus late passage IMR-90 fibroblasts. Finally, the accumulation and release of cAMP in response to adenosine was determined as another indicator of receptor-mediated cAMP responses in aged human fibroblasts. Adenosine can inhibit or stimulate cAMP formation by binding to the Aj or A2 subclass of adenosine receptors, respectively (Daly, 1982). Adenosine has also been shown to inhibit B-adrenergic-induced cAMP formation (Dobson, 1978). Moreover, adenosine release by cultured human fibroblasts increases with age (Ethier et al., 1989). Thus, the effect of adenosine on cAMP and/or isoproterenol-induced stimulation of cAMP was determined in early and late passage IMR-90 fibroblasts. MATERIALS AND METHODS Cell culture

All experiments were performed using human lung fibroblasts (IMR-90) obtained from the NIA Aging Cell Repository, Institute for Medical Research, Camden, New Jersey. This well characterized diploid fibroblast strain was established from a 16-week normal female fetus for research purposes and has been used extensively for in vitro aging studies (Nichols

cAMP RELEASE AND ACCUMULATION IN AGED CELLS

289

et al., 1976). The fibroblasts were grown in Eagle's minimal essential medium (MEM) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 international units (U)/ ml penicillin, and 100 ug/ml streptomycin maintained at 37"C in a 5% CO2-95% air environment. At confluency, fibroblasts were subcultivated with 0.25% trypsin in calcium-and magnesium-free MEM. Fibroblasts were received at population doubling level (PDL) 12 and PDL 36. For subcultivation of IMR-90 fibroblasts, a trypsinized aliquot was removed for counting on a hemocytometer, and the fibroblasts were seeded into new vessels at a density of 1 × 10 4 cells/cm 2. The new PDL was calculated by comparing fibroblast counts per vessel at seeding with counts at confluence. All early passage experiments were on cells from PDLs 21-24, and all late passage experiments were on cells from PDLs 45-53. In our laboratory, the IMR-90 fibroblasts ceased dividing at PDL 55. For experiments, fibroblasts were seeded at 1 × l 0 4 cells/cm 2. Subconfluent early passage cells were used 3 days after seeding, and confluent early passage cells were used at least 7 days after seeding. Late passage cells grew more slowly and were used for confluent experiments at least 10 days after seeding. cAMP measurement

The cAMP content was determined in fibroblast extracts and in the bathing medium of fibroblast cultures. For these determinations, confluent or subconfluent IMR-90 fibroblasts grown in 60-mm culture dishes were gently washed three times with 5 ml of physiological saline (PS, 37"C) containing 118.4 mM NaCI, 4.69 mM KCI, 2.52 mM CaCIE, 25 mM NaHCO3, 1.18 mM MgSO4, 1.18 mM KH2PO4, 10 mM dextrose, and 0.1 mg/ml ascorbate. Representative dishes of confluent and subconfluent cells were trypsinized, and the viable cell number determined by staining cells with 0.4% trypan blue and counting on a hemocytometer. The remaining dishes were incubated at 37°C in a 5% CO2-95% air environment (pH 7.4) for 30 min in 2 ml of PS or in PS containing either 20 ~M adenosine, 2 ~M isoproterenol, or adenosine (20 ttM) plus isoproterenol (2 ~M). After incubation, the media was transferred to 12 × 75 mm glass tubes containing 2500 cpm of 3H-cAMP and placed in a boiling water bath for 5 min. After centrifugation at 1000 X g for 15 min, the supernatant was removed for determination of cAMP. Cell extracts were obtained by adding 2.0 ml of cold (0*C) 10% trichloroacetic acid (TCA) containing 2500 cpm 3H-cAMP to each dish immediately after removal of the incubation media. The cells were scraped from the growth surface with a rubber spatula and transferred with one washing of 1.0 ml TCA to ice-chilled centrifuge tubes. After centrifugation at 1000 X g for 15 min, the supernatant was transferred to glass tubes (l 6 X 150 mm), and the remaining pellet was assayed for protein content by the method of Lowry et al. ( 195 l), using bovine serum albumin as a standard. The supernatant was extracted four times with two volumes of H20-saturated ether. Ether from the final wash was evaporated by briefly raising the temperature of the supernatant to 800C, and an aliquot of the aqueous solution was used for cAMP measurement. The content of cAMP in media and cell extracts was determined by a ~25I-cAMP radioimmunoassay kit (Amersham, Arlington Heights, IL). This assay employs a cAMP-specific rabbit antibody, with the final separation being achieved by a second antibody (donkey) bound to a polymer, thereby permitting separation of the unbound labeled tracer by centrifugation. Extraction recoveries, typically greater than 85%, were used in calculating the cAMP levels. Levels of cAMP are expressed as pmol per mg of cell protein.

290

M.F. ETHIER et al.

Membrane preparation and adenylyl cyclase assay Adenylyl cyclase activity was determined using fibroblast membranes. Confluent early and late passage IMR-90 fibroblasts on 35-mm dishes were trypsinized (0.25%), and the detached cells were transferred to ice-chilled centrifuge tubes. Cell suspensions were centrifuged at 200 × g for 15 min, and the pellets were washed with 5 ml of 0*C Hanks buffered salt solution (HBSS) (GIBCO, Grand Island, NY). After repeating this procedure twice the cell pellet was suspended in 500 #1 ofa 0°C buffer containing 10 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES), 10 mM EDTA, 1 mM dithiothreitol (DTT), 0.1 mM benzamidine, 10% sucrose, and 0.01% phenylmethylsulfonyl fluoride. The cell suspension was homogenized for 45 s at half maximal speed with a Polytron PT10 generator (Brinkman, Westbury, NY) and centrifuged at 20 000 × g for 30 minutes at 4°C. Protein was determined on the cell pellet by the method of Bradford (1976), using bovine serum albumin as a standard. Each 35-mm dish yielded 100-200 ~g of protein. Adenylyl cyclase activity was determined by measuring the conversion of [a-32p]2 'deoxy-ATP to [a-32p]2'-deoxy-cAMP as described previously (LaMonica et al., 1985). The membranes (5-25 ug protein) were treated for 10 min with adenosine deaminase (5 U/ml) in a buffer containing 40 mM HEPES and 1.0 mM ascorbic acid (pH 7.4) at 30°C prior to assay. For the assay, membranes were incubated for 20 min at 30°C in 50 ul of a buffer containing 40 mM HEPES (pH 7.4), 5 mM MgC12, 1 mM DTT, 5.5 mM KC1, 0.1 mM 2'-deoxy-cAMP (dcAMP), 0.1 mM 2'-deoxy-ATP (dATP), 20 mM phosphoenolpyruvate, 2 U pyruvate kinase, 0.25 U adenosine deaminase, 1 mM ascorbic acid, 100 mM NaCI, 2 mM [ethylenebis(oxyethylene-nitrilo]tetraacetic acid (EGTA), 10 uM guanosine 5'-triphosphate (GTP), and approximately 2 × 10 6 cpm of[a-3Zp]dATP. Isoproterenol (1 ~M), phenylisopropyladenosine (PIA, 1 ~M), and forskolin (100 nM) were present as indicated. The reaction was stopped by adding 50 #1 of a solution containing 2% sodium dodecyl sulfate (SDS), 45 mM ATP, 1.3 mM cAMP, and 3H-dcAMP (approximately 4000 cpm) and boiling for 2 min. The [a-32p]dcAMP produced was separated from [a-32p]dATP by sequential column chromatography using columns of cation exchange resin AG 50W-X4 (200-400 mesh) and neutral alumina AG 7 (100-200 mesh) after the method of Salomon (1979). All results were corrected for 3H-dcAMP recovery, which ranged between 60 and 90%. The activity of the adenylyl cyclase is expressed as picomoles [a-32p]dcAMP formed per rain per mg of membrane protein.

Fibroblast homogenate preparation and phosphodiesterase assay The activity of cAMP phosphodiesterase was determined using fibroblast homogenates. The culture medium was removed by aspiration from confluent 60-mm dishes of early (24-27 PDLs) and late (47-53 PDLs) passage IMR-90 fibroblasts. The fibroblasts were briefly rinsed once with 2 ml of 0°C HBSS, and the wash solution was aspirated. Two ml of fresh HBSS was added to each dish, and the fibroblasts were scraped from the dish and harvested by centrifugation (200 × g for 5 min). The supernatant was removed, 600 #1 of 40 mM Tris buffer (pH 8) added to the fibroblast pellet, and the mixture sonicated at 0*C for 20 s using a microprobe of a Fisher Sonic Dismembrator (Model 300) at 10% power. Phosphodiesterase activity was determined immediately after cell sonication according to the methods of Thompson et al (1979). Briefly, 20-40 t~g offibroblast homogenate protein were incubated in 400 ul of a buffer containing 40 mM Tris pH 8, 125 nM cAMP, 2.2 × 105 cpm 3H-cAMP, and 20 mM MgC12 for 20 rain at 30°C. Isobutylmethylxanthine

cAMP RELEASEAND ACCUMULATIONIN AGEDCELLS

291

(IBMX), an inhibitor of phosphodiesterase activity (Kakashi and Dage, 1988) at 100 uM, was added where indicated. The reaction was stopped by placing the reaction tube in a boiling H20 bath for 45 s. The formed 3H-cAMP that resulted from the phosphodiesterase activity was converted to 3H-adenosine by the addition of 0.1 mg of snake venom (5'nucleotidase) for 20 min at 30"C. Immediately after the second incubation, 1 ml of an anion exchange resin (AG l-X8) in 50% methanol was added to precipitate the unreacted 3H-cAMP. The 3H-adenosine remaining in the supernatant was determined by scintillation spectrometry. The phosphodiesterase activity was expressed as pmol of cAMP converted to adenosine per min per mg fibroblast homogenate protein.

Statistical analysis All results are reported as the mean _+ 1 SE. Statistical analysis of the data was conducted using a two-tailed t test for paired samples (Zar, 1984). Differences were considered statistically significant when p < 0.05. RESULTS

Cellular cAMP content and release and fibroblast population doubling level The cellular content and net release of cAMP were determined in confluent cultures of early and late passage IMR-90 human fibroblasts (Table 1). Late passage IMR-90 fibroblasts were less dense than early passage IMR-90 fibroblasts at confluence. After incubation in PS for 30 min, cellular levels of cAMP based on cell protein were not different in early versus late passage cultures of confluent fibroblasts. However, the net release of cAMP from fibroblasts, as determined by accumulation of the cyclic nucleotide in the PS during a 30-min incubation period was 3.7 times greater in early versus late passage cultures. The total cAMP (sum of the fibroblast cellular and bathing medium cAMP content), based on cell protein, was 1.8 times greater in early passage cultures. Levels of cAMP were determined in confluent cultures of early and late passage IMR90 fibroblasts after 30-min incubations in the presence of adenosine, isoproterenol, or both (Fig. 1). In early passage fibroblasts, neither cAMP content nor net release was changed by incubation in 20 uM adenosine. In late passage fibroblasts incubated in 20 uM adenosine, cellular cAMP was 1.7 times greater and net release of cAMP 3.5 times greater than in TABLE 1. CELL PROTEIN, CELL NUMBER, AND c A M P IN CONFLUENT CULTURES OF EARLY AND LATE PASSAGE IMR-90 FIBROBLASTS

Cell protein (ug) Early Passage (PDLs 22-24) Late Passage (PDLs 47-52)

Cell number (10 -J)

Cyclic AMP (pmol/mg protein) Cellular (A)

Net release (B)

528.5 + 34(8)

6.9 + 0.46(4)

34.2 + 5.6(8)

67.8 + 8.6(8)

418.2 + 49.7(9)

3.2 + 0.20(4)*

38.4 + 9.0(9)

18.5 _+ 7.0(9)*

Total (A + B) 102.0 _+ 10.8(8) 57.0 + 6.6(9)*

Confluent fibroblasts in 6 0 - m m culture dishes were incubated for 30 m i n in 2 ml o f physiological saline. Cells a n d m e d i u m were subsequently analyzed for c A M P c o n t e n t as described in Materials a n d Methods. Values are the m e a n + 1 SE for the n u m b e r o f e x p e r i m e n t s in parentheses. P D L denotes population doubling level. *Statistically significant difference from the corresponding early passage value.

292

M.F. ETHIER et al.

EARLY PASSAGE

LATE PASSAGE

600

8 b

8

b e-

500

I

0 r..,

400 !

E

0,. Z U

i

300 200

8

T 6

t00

8

0 #EDIA

S

A

I

A+I

S

A

n

8

6

8

6

9

6

1 I

A+I

9

6

FIG. 1. The effectof adenosine and/or isoproterenol on cellular content and net releaseof cAMP in confluent cultures of early (PDLs 22-24) and late (PDLs 47-52) passageIMR-90 fibroblasts. Cells were incubated at 37"C for 30 min in either physiologicalsaline (S) or in saline with either 20/~M adenosine (A), 2 uM isoproterenol (I), or adenosine plus isoproterenol (A + I). Cyclic AMP was determined in the cell extracts (hatched bars) and the incubation medium (open bars) according to techniques described in Material and Methods. Each bar represents the mean + 1 SE for the number of experiments indicated (n). The letter a denotes a significant difference from the appropriate saline control value (S); b denotes a significant difference from the corresponding early passage value. cultures incubated in saline only. Incubation o f confluent fibroblasts in the presence of 2 u M isoproterenol increased cellular c A M P and net release o f c A M P in both early and late passage cultures. In early passage cultures, isoproterenol increased cellular c A M P by 2.7fold to 94 p m o l / m g protein, and net release o f c A M P by 3.8-fold to 255 p m o l / m g protein. The response to isoproterenol was significantly greater in late passage cultures, as cellular c A M P was increased by 7.8-fold to 299 p m o l / m g protein and net release o f c A M P was increased by 26. l-fold to 483 p m o l / m g protein. W h e n adenosine and isoproterenol were c o m b i n e d in the incubation m e d i u m o f early or late passage fibroblast cultures, cellular content and release o f c A M P were not different from cultures exposed to isproterenol in the absence o f adenosine.

Cellular cAMP content and release and culture density of early and late passage fibroblasts In the experiments described above, late passage fibroblast cultures displayed a reduced cell density at confluency c o m p a r e d to early passage cultures (Table l). In an attempt to determine whether fibroblast culture density influenced the enhanced response o f late pas-

293

cAMP RELEASE AND ACCUMULATION IN AGED CELLS

sage fibroblasts to adenosine and isoproterenol, the effects of these compounds on total cAMP were examined in late and early passage cultures at an equivalent cell density (Table 2). Low density cultures of early passage (PDL 24) IMR-90 fibroblasts used 3 days after seeding (subconfluent) were at a mean population density of 3.2 X l05 cells/dish or 225 ug protein/dish. Confluent cultures of late passage (PDL 52) IMR-90 had a similar number of cells per dish (3.1 X 105), but a greater amount of cell protein (398 ~tg/dish). cAMP levels were also determined in confluent, high density cultures of early passage (PDL 22) IMR-90, which served as a basis of comparison for the other studies reported herein. After a 30-rain saline incubation, total cAMP was not different in high and low density early passage cultures (Table 2). However, total cAMP in confluent late passage fibroblast cultures was significantly less. Incubation with 20 uM adenosine did not affect total cAMP in high density early passage fibroblast cultures. However, adenosine increased total cAMP 1.7-fold in low density early passage cultures and 4.4-fold in confluent late passage cultures. Incubation in 2 tzM isoproterenol resulted in a significant increase in total cAMP in all fibroblast cultures. Total cAMP increased 1.7-fold in high density early passage fibroblasts and 3.4-fold in low density early passage cultures. After incubation in 2 uM isoproterenol, total cAMP in late passage fibroblast cultures was greater than high or low density early passage cultures, increasing 38.4-fold above the baseline value. When adenosine and isoproterenol were combined in the incubation saline, the increase in total cAMP was similar to that observed with isoproterenol alone for high and low density early passage and late passage fibroblast cultures. The relative contribution of cellular cAMP and net cAMP release to the total cAMP values for high (confluent) and low (subconfluent) density early passage and confluent late passage IMR-90 cultures is presented in Fig. 2. There was no difference in cellular cAMP levels between the three groups offibroblast cultures. After a 30-min saline incubation, the

TABLE 2. EFFECTS OF ADENOSINE AND/OR ISOPROTERENOL ON c A M P IN CONFLUENT AND SUBCONFLUENT CULTURES OF EARLY PASSAGE AND CONFLUENT CULTURES OF LATE PASSAGE

1MR-90 Confluent Early Passage (PDL 22) Subconfluent Early Passage (PDL 24) Confluent Late Passage (PDL 52)

IMR-90

FIBROBLASTS

Total cAMP (pmol/mg protein)

Cell protein (ug/dish)

Cell number (10 5)

Saline

Adenosine

Isoproterenol

533 _+ 19.4"

7.1 + 0.38*

89 + 12.8

87 + 4.9*

155 + l l . 8 * t

140 + 5.7"t

225 _+ 11.1

3.2 + 0.17

93 _+ 14.5

160 _+ 25.2t

312 + 37.2#

235 + 29.5t

398 _+ 17.5"

3.1 _+ 0.24

34 _+ 9.2*

150 +_ 24.0t

1307 _+ 182.6"t

Adenosine plus isoproterenol

1207 _+ 184.9"t

Fibroblasts in 60-mm culture dishes were incubated for 30 min in the absence or presence of 20 uM adenosine and/or 2 uM isoproterenol. Cells and medium were subsequently analyzed for cAMP as described in Materials and Methods, and the values were combined to present total cAMP per dish. Values, except for protein, are the mean _+ 1 SE for 3 experiments. Protein values are the mean _+ 1 SE for 12 experiments. PDL denotes population doubling level. *Significantly different from the corresponding value in subconfluent, early passage cultures. tSignificantly different from the appropriate saline value.

294

M.F. ETHIER et al.

D

C

800

.I

b

1

600 O ¢.

01

a

8

400

0

II

0 U

200

a

il

CEP SCEP CLP

CEP SCEP CLP

CEP SCEP CLP

0

CEP SCEP CLP

FIG. 2. Cellular cAMP (hatched bars) and net cAMP released (open bars) in cultures of confluent early passage (CEP), subconfluent early passage (SCEP), and confluent late passage (CLP) IMR-90 fibroblasts. The cells were incubated at 37"C for 30 min in physiological saline (Panel A) or in saline with 20 uM adenosine (Panel B), 2 uM isoproterenol (Panel C), or 20 uM adenosine plus 2 ~M isoproterenol (Panel D). Each bar represents the mean +_ 1 SE for three experiments. Mean cell protein (ug/dish) was 533 for CEP, 225 for SCEP, and 398 for CLP. Representative cell number per dish was 7.1 X 105 for CEP, 3.2 X 105 for SCEP, and 3.1 X 105 for CLP. The letter a denotes a significant difference from the appropriate saline control in Panel A; b denotes a significant difference from the corresponding SCEP value.

net release of cAMP in confluent late passage cultures was 5.0 pmol/mg and significantly less than the net release from early passage fibroblast cultures at an equivalent density or at high density (Fig. 2A). Incubation in 20 #M adenosine did not affect cellular cAMP or cAMP release in high density early passage cultures. However, adenosine increased cellular cAMP by 1.5-fold and cAMP release by 1.9-fold in low density early passage cultures (Fig. 2B). In confluent late passage cultures, adenosine incubation resulted in a 2.6-fold increase in cellular cAMP and a 15-fold increase in cAMP release. After adenosine incubation, the absolute values of cellular cAMP or cAMP release were not different for early versus late passage fibroblasts of equivalent density. Incubation in 2 #M isoproterenol increased both cellular cAMP and release of cAMP in all three groups ofIMR-90 fibroblast cultures (Fig. 2C). Incubation of high density early passage cultures in the presence ofisoproterenol resulted in a 1.9-fold increase in cellular cAMP and a 1.6-fold increase in cAMP release. The isoproterenol-induced increase in cellular and released cAMP was greater in low density early passage cultures. Cellular cAMP was increased 3.2-fold, and the net release of cAMP was increased 3.4-fold. When confluent late passage cultures were incubated in 2 #M isoproterenol, both cellular cAMP and

295

cAMP RELEASE AND ACCUMULATION IN AGED CELLS

cAMP release were greater than in high or low density early passage fibroblasts. Cellular cAMP was increased 23.3-fold, and the net release of cAMP was increased 126.4-fold. When adenosine (20 um) and isoproterenol (2 um) were combined in the incubation medium, both cellular cAMP and net release of cAMP were similar to the levels observed with isoproterenol alone (Fig. 2D).

Adenylyl cyclase activity in early and late passage fibroblast membranes Since the isoproterenol-induced increase in cAMP content and release is much greater in late (compared to early) passage lung fibroblast cultures, the possibility that adenylyl cyclase activity could be involved in these differences was examined. The basal and stimulated levels of this enzyme, which is responsible for cAMP synthesis, were determined in membranes from confluent early and late passage IMR-90 fibroblasts (Table 3). There was no difference in basal adenylyl cyclase activity in crude membrane preparations from early versus late passage IMR-90 fibroblasts. There was no difference between early and late passage fibroblasts with respect to stimulation of adenylyl cyclase activity by either forskolin (2.6 to 2.8-fold) or isoproterenol ( 1.7 to 1.8-fold). Phenylisopropyladenosine (PIA), an A~-adenosine receptor agonist (Romano et al., 1989), had no effect on adenylyl cyclase and did not affect stimulation of adenylyl cyclase activity by isoproterenol in early or late passage IMR-90 fibroblasts.

cAMP phosphodiesterase activity of early and late passage fibroblasts Another possible explanation for the observed differences in cellular cAMP levels and cAMP release in early versus late passage cultures was explored by examining the activity of cAMP phosphodiesterase, an enzyme responsible for cAMP hydrolysis. Phosphodiesterase activity was determined in cell sonicates from confluent early and late passage IMR90 fibroblast cultures (Table 4). Activity of the enzyme in early passage fibroblast homogenates was 1.7-fold greater than homogenates of late passage fibroblasts. Greater than 95%

T A B L E 3. E F F E C T OF FORSKOLIN, ISOPROTERENOL, AND PHENYLISOPROPYLADENOSINE ON ADENYLYL CYCLASE ACTIVITY IN C R U D E MEMBRANE PREPARATIONS FROM CONFLUENT DISHES OF EARLY AND LATE

PASSAGEIMR-90 FIBROBLASTS

Adenylyl cyclase activity (pmol/min/mg protein) IMR-90

n

Basal

Forskolin (lO-ZM)

lsoproterenol (aM)

PIA (aM)

lsoproterenol + PIA (aM)

Early Passage (PDLs 21-22) Late Passage (PDLs 45-47)

11

19.0 + 0.8

53.9 + 3.4*

34.5 + 2.7*

20.3 _+ 2.0

32.1 _+ 1.91"

9

20.3 + 0.7

51.9 __+2.0*

35.5 + 1.6"

21.2 _+ 1.9

37.1 +_ 1.61"

Fibroblast membranes were prepared as described in Materials and Methods and incubated at 30* C for 20 min in the absence or presence of forskolin, isoproterenol, PIA (phenylisopropyladenosine),or isoproterenol plus P1A as indicated. Adenylyl cyclase activity was determined by measuring the conversion of [a-32p12'-deoxy-ATP to [a-32p]2'-deoxy-cAMP as described in Materials and Methods. Values are the mean + 1 SE of the indicated number of experiments (n). *Statistically different from appropriate basal value.

296

M.F. ETHIER et al.

TABLE 4. T O T A L A N D I S O B U T Y L M E T H Y L X A N T H I N E SENSITIVE c A M P P H O S P H O D I E S T E R A S E A C T I V I T Y IN H O M O G E N A T E S FROM C O N F L U E N T EARLY A N D LATE PASSAGE

IMR-90

FIBROBLASTS

Phosphodiesterase activity (pmol/min/mg protein) IMR-90

n

Total

n

I B M X inhibited

Early Passage (PDLs 24-27) Late Passage (PDLs 47-53)

10

9.61 + 1.15

7

9.17_ 1.18

11

5.81 _+ 1.11"

8

5.68 _+ 1.09"

Phosphodiesteraseactivity is the pmol of cAMP convertedto adenosine per min per mg fibroblast homogenate protein as described in Materialsand Methods. IBMX (isobutylmethylxanthine)-inhibited activity was determined by subtracting the phosphodiesterase activity measured in the presenceof 100 ~M IBMX from the total activity. Values are the mean _+ l SE of the indicated number of experiments (n). *Significantdifferencefrom the correspondingearly passage value.

o f the phosphodiesterase activity measured was inhibited by 100 uM isobutylmethylxanthine (IBMX), a potent inhibitor o f the enzyme (Kakashi and Dage, 1988). The IBMXinhibited activity in the early passage homogenates was 1.6-fold greater than in the homogenates from late passage fibroblasts. DISCUSSION Basal c A M P levels in our study were determined after incubating IMR-90 embryonic lung fibroblast cultures at 37°C for 30 min in physiological saline. There was no difference in the cellular c A M P levels o f confluent or subconfluent early passage or confluent late passage fibroblasts (Table 1 and Fig. 2). This is in contrast to an earlier study (Haslam and Goldstein, 1974) reporting an increase in cellular cAMP per mg o f protein in confluent cultures o f late versus early passage skin fibroblasts taken from adult h u m a n donors. It is possible that the different effects o f in vitro aging on cellular cAMP between the two studies is due to a difference in aging o f fetal versus adult tissue or reflects a difference due to the site o f fibroblast origin. Alternatively, methodological differences between the two studies may account for the apparent discrepancy. In our study, cAMP was determkled on cells incubated in physiological saline, while the experiments of Haslam and Goldstein were performed using a m i n i m u m essential media. It should also be pointed out that, when comparing basal concentrations in early versus late passage fibroblasts, the denominator o f the unit of expression can have a bearing on the results. For instance, when Haslam and Goldstein based their cellular c A M P values on cell size (rather than protein) and expressed their results as molar concentrations, they found no difference between early and late passage fibroblasts. This reflected a proportionately greater increase in cell size than protein content in their late passage fibroblasts. When IMR-90 fibroblasts were incubated for 30 min in physiological saline, the net

cAMP RELEASE AND ACCUMULATION IN AGED CELLS

297

release of cAMP from confluent late passage cells was less than that from confluent or subconfluent early passage cells (Table 1 and Fig. 2). This observation has not been reported previously. There is no clear physiological role for extracellular cAMP released by human fibroblasts. Cyclic AMP does not readily enter mammalian cells (Robison et al., 1965), and the existence of surface receptors for cAMP has not been clearly demonstrated. A potential role of cAMP release is to lower intracellular levels of cAMP (Kelly et al., 1978). In our study, release of cAMP was markedly less from late passage fibroblast cultures, but cellular cAMP was not different in late versus early passage cultures. It is possible that reduced cAMP release by late passage IMR-90 serves to maintain the intracellular level of cAMP in these cells. In our study, incubation in 2 ~M isoproterenol resulted in an increase in cellular cAMP of 2.7-fold in early passage and 7.8-fold in late passage cultures of confluent IMR-90 fibroblasts (Fig. l). The enhanced stimulation of cellular cAMP levels in late passage fibroblasts was not due to diminished release of the cyclic nucleotide, since isoproterenol incubation increased cAMP release by 26.1-fold in late passage fibroblasts, compared to 3.8-fold in early passage fibroblasts. The enhancement of/3-adrenergic-induced stimulation of cellular cAMP levels in late passage cultures confirms previous reports on confluent cultures of human skin and lung fibroblasts aged in vitro (Haslam and Goldstein, 1974; ManganieUo and Breslow, 1974; Polgar et al., 1978). In addition, the cAMP response to/3-adrenergic stimulation has been reported to diminish with increasing cell population density in early passage human fibroblasts (Haslam and Goldstein, 1974; Kelly and Butcher, 1974; Manganiello and Breslow, 1974). A possible explanation for this is the reported decrease in the number of/~-receptors associated with increased population density in early passage human fibroblasts (Pochet et al., 1982). Since a greater population density at confluency is a characteristic feature of early versus late passage human fibroblasts (Simons, 1967; Schneider and Mitsui, 1976), density differences might account for the comparatively weak cAMP response to isoproterenol in early versus late passage confluent cultures. In our study, the enhancement of cellular cAMP levels by isoproterenol was greater in subconfluent (low density) than in confluent (high density) early passage IMR-90 fibroblasts, but not as great as confluent late passage fibrobiasts at a similar density (Table 2 and Fig. 2). This suggests that the enhanced cAMP levels in response to isoproterenol in late versus early passage IMR-90 fibroblasts may be partially, but not solely, attributable to the greater fibroblast density at confluency in early passage cultures. In comparing early and late passage fibroblasts at equivalent densitites, it should be noted that confluent last passage fibroblasts are growth arrested, while subconfluent early passage fibroblasts are in an active growth phase. Although there is no evidence that growth per se of these early passage fibroblasts accounts for diminished/~-adrenergic stimulation of cAMP, the possibility cannot be ruled out. Isoproterenol increases cAMP by stimulating the activity of membrane-bound adenylyl cyclase. In our study, adenylyl cyclase activity was determined in crude membrane preparations from cultured human flbroblasts. The activity of the enzyme was stimulated by isoproterenol, indicating that the membrane preparations contained an intact and functional receptor-G protein-adenylyl cyclase complex. Also, a functional adenylyl cyclase catalytic subunit was demonstrated in these membrane preparations by stimulation of adenylyl cyclase activity by forskolin. In our study there was no passage-related difference in basal adenylyl cyclase activity or in stimulation of activity by isoproterenol or forskolin.

298

M.F. ETHIER et aL

This suggests that the greater isoproterenol-induced stimulation of cAMP accumulation and release in late (compared to early) passage IMR-90 fibroblasts was not due to enhanced adenylyl cyclase activity in the late passage fibroblasts. Since cyclic nucleotide phosphodiesterases are responsible for the degradation of intracellular cAMP, the activity of this enzyme was determined in sonicates of IMR-90 fibroblasts. The activity of cAMP-phosphodiesterase in confluent early passage fibroblasts was 1.7-fold greater than the activity in confluent late passage fibroblasts (Table 4). This passage-related difference in enzyme activity suggests that the enhanced levels of cAMP observed upon ~-adrenergic stimulation in late (compared to early) passage fibroblasts may be due to diminished hydrolysis of cAMP to 5'-AMP. The cAMP-phosphodiesterase activity assayed in our study was from a crude extract of cultured human IMR-90 fibroblasts. However, more than 95% of the measured phosphodiesterase activity was IBMX-inhibited. This indicates that the assay employed assessed primarily cAMP-dependent phosphodiesterase activity (Kakashi and Dage, 1988). A number of species of cAMP-phosphodiesterase have been shown to exist in a variety of tissues, and the total phosphodiesterase activity reported in our study likely represents several different species of the enzyme. A passage-related change in a particular species of the enzyme could explain the apparent contradiction between the present results and a previous study (Polgar et aL, 1978) that reported no difference in cAMP-phosphodiesterase activity in early versus late passage cultured human fibroblasts. Since adenosine has been shown to inhibit/~-adrenergic-induced formation of cAMP (Dobson, 1978) and to have an enhanced release from cultured human fibroblasts with age (Ethier et al., 1989), the effect of adenosine on isoproterenol stimulation of cAMP was also determined in early and late passage IMR-90 fibroblasts. When early or late passage fibroblasts were incubated in physiological saline containing isoproterenol plus adenosine, neither cellular cAMP nor release of cAMP was different from that of IMR-90 fibroblast cultures incubated in the presence of isoproterenol alone (Fig. 1). This indicates that the adenosine-induced stimulation of cAMP release observed in late passage IMR-90 cultures was not additive to cAMP release induced by isoproterenol (2 aM). This also suggests that isoproterenol-induced stimulation of cellular and released cAMP in IMR-90 fibroblasts cultures is not inhibited by adenosine. Attenuation of B-adrenergic response by adenosine is thought to occur via the A~ cell surface receptor (Romano et aL, 1989). Activation of this receptor reduces adenylyl cyclase activity. Exposure to PIA, a potent analogue for AI receptors, did not effect either fibroblast membrane adenylyl cyclase activity or isoproterenol stimulation ofadenylyl cyclase activity in membranes from early or late passage fibroblasts (Table 3). The lack of effect of PIA on early and late passage fibroblast adenylyl cyclase activity indicates that these cells possess little or no functional AI receptor activity and further suggests that adenosine does not inhibit the 3-adrenergic stimulation of cAMP levels in IMR-90 fibroblasts. In summary, under the basal conditions reported in our study, cAMP release was decreased in cultures of in vitro aged IMR-90 human fibroblasts. This decrease was not associated with a change in cellular cAMP or related to the reduced cell density displayed by in vitro aged IMR-90 fibroblasts. The isoproterenol-induced increase in cellular cAMP accumulation and net cAMP release was markedly greater in cultures of in vitro aged IMR-90 fibroblasts compared to early passage cultures. This increase was not associated with enhanced adenylyl cyclase activity in late passage fibroblasts but may be due to the diminished cAMP-phosphodiesterase activity observed in these cells. The reduced cell

cAMPRELEASEANDACCUMULATIONIN AGEDCELLS

299

density displayed by in vitro aged IMR-90 fibroblasts also contributed in part to the increased #-adrenergic stimulation of cellular cAMP formation observed in late passage cultures. The adenosine-induced increase in cAMP levels in confluent late passage IMR90 fibroblasts was not observed in confluent early passage cells and may be entirely due to the decreased cell density in late passage cultures. Adenosine has no effect on the isoproterenol-induced increase in cAMP accumulation and release in IMR-90 human fibroblast cultures. Acknowledgments - - The authors thank Dr. Richard A. Fenton for his suggestions and encouragement. The excellent technical assistance of Lynne G. Shea is gratefully acknowledged. This study was supported in part by Public Health Service Grants HL-22828 and HL-36964.

REFERENCES BRADFORD, M.M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254, 1976. DALY, J.W. Adenosine receptors: Targets for future drug use. J. Med. Chem. 25, 197-207, 1982. DOBSON, J.G., JR. Reduction by adenosine of the isoproterenol-induced increase in cyclic adenosine 3',5'monophosphate formation and glycogen phosphorylase activity in rat heart muscle. Circ. Res. 43, 785-792, 1978. ETHIER, M.F., HICKLER, R.B., and DOBSON, J.G., JR. Aging increases adenosine and inosine release by human fibroblast cultures. Mech. Aging Dev. 50, 159-168, 1989. HASLAM, R.J. and GOLDSTEIN, S. Adenosine 3',5'-cyclic monophosphate in young and senescent human fibroblasts during growth and stationary phase in vitro. Biochem. J. 144, 253-263, 1974. HAYFLICK, L. The limited in vitro lifespan of human diploid cell strains. Exp. Cell Res. 25, 585-621, 1965. HAYFLICK, L. Cell aging. In: Annual Review o f Gerontology and Geriatrics, Eisdorfer, C. (Editor), pp. 26-67, Springer, New York, NY, 1980. KAKASHI, T. and DAGE, R.C. Tissue distribution and selective inhibition of subtypes of high affinity cAMP phosphodiesterase. Biochem. Pharmacol. 37, 3267-3270, 1988. KELLY, L.A. and BUTCHER, R.W. The effects of epinephrine and prostaglandin E~ on cyclic adenosine 3':5' -monophosphate levels in WI-38 fibroblasts. J. Biol. Chem. 249, 3098-3102, 1974. KELLY, L.A., WU, C., and BUTCHER, R.W. The escape of cyclic AMP from human diploid fibroblasts: General properties. J. Cyclic Nucleotide Res. 4, 423-435, 1978. LAMONICA, D.A., FROHLOFF, N., and DOBSON, J.G., JR. Adenosine inhibition of catecholamine-stimulated cardiac membrane adenylate cyclase. Am. J. Physiol. 248, H737-H744, 1985. LOWRY, O.H., ROSEBROUGH, N.J., FARR, A.L., and RANDALL, R.J. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-274, 1951. MANGANIELLO, V.C. and BRESLOW, J. Effects ofprostaglandin E~ and isoproterenol on cyclic AMP content of human fibroblasts modified by time and cell density in subculture. Biochim. Biophys, Acta 362, 509-520, 1974. NICHOLS, W.W., MURPHY, D.G., CRISTOFALO, V.J., TOJI, L.H., GREENE, A.E., and DWIGHT, S.A. Characterization of a new human diploid cell strain, IMR-90. Science 196, 60-63, 1976. POCHET, R.P., GREEN, D.A., GOKA, T.J., CLARK, R.B., BARBER, R., DUMONT, J.E., and BUTCHER, R.W. ~-Adrenergic receptors and cyclic AMP responses to epinephrine in cultured human fibroblasts at various population densities. J. Cyclic Nucleotide Res. 8, 83-87, 1982. POLGAR, P., TAYLOR, L., and BROWN, L. Plasma membrane associated metabolic parameters and the aging of human diploid fibroblasts, Mech. Aging Dev. 7, 151-160, 1978. ROBISON, G.A., BUTCHER, R.W., OYE, I., MORGAN, H.E., and SUTHERLAND, E.W. The effect of epinephrine on adenosine 3',5'-phosphate levels in the isolated perfused rat heart. Mol. Pharmacol. 1, 168-177, 1965. ROMANO, F.D., MACDONALD, S.G., and DOBSON, J.G., JR. Adenosine receptor coupling to adenylate cyclase of rate ventricular myocyte membranes. Am. J Physiol. 257, H 1088-H 1095, 1989. SALOMON, Y. Adenylate cyclase assay. Adv. Cyclic Nucleotide Res. 10, 35-55, 1979.

300

M.F.ETHIERet aL

SCHNEIDER, E.L. and MITSUI, Y. The relationship between in vitro cellular aging and in vivo human age. Proc. Natl. Acad Sci. U.S.A. 73, 3584-3588, 1976. SIMONS, J.W.I.M. The use of frequency distributions of cell diameter to characterize cell populations in tissue cultures. Exp. CellRes. 45, 336-350, 1967. THOMPSON, W.J., TERASKI, W.L., EPSTEIN, P.M., and STRADA, S.J. Assay of cyclic nucleotide phosphodiesterase and resolution of multiple molecular forms of the enzyme. In: Advances in Cyclic Nucleotide Research, Greengard, G.B. and Robison, G.A. (Editors), pp. 69-92, Raven Press, New York, NY, 1979. ZAR, J.H. Biostatistical Analysis. Prentice-Hall, Inc., Englewood Cliffs, N J, 1984.