Effects of prostaglandin E1 on collagen production and degradation in human fetal lung fibroblasts

ARCHIVESOF

ISIOCHEMISTRYANDBIOPHYSICS

Vol. 265, No. 2, September,

pp. 441-446,1988

EEffects of Prostaglandin El on Collagen Production Degradation in Human Fetal Lung Fibroblasts’ FRANK

A. BARILE,

and

CATHERINE RIPLEY-ROUZIER, ZAFAR-E-ALAM AND ROBERT S. BIENKOWSK12

Department of Natural Sciences, York Pediatric Research Center, Schneider Park, New York 110.42; and Department

Received

February

SIDDIQI,

College of the City University of New York, Jamaica, New York 11451; Children’s Hospital of Long Island Jewish Medical Center, New Hyde of Pediatrics, Health Sciences Center, State University of New York, Stony Brook, New York 11790 29,1988,

and in revised

form

May

11,1988

We examined the effects of prostaglandin El on the production and degradation of collagen in human fetal lung fibroblasts. Percentage collagen production was determined by incubating confluent cultures for 6 h with [3H]proline and either [14C]glycine or [‘4C]leucine and measuring the relative amounts of radioactivity incorporated into collagenasesensitive and collagenase-insensitive material. Percentage collagen degradation was determined by measuring hydroxy[14C]proline in a low-molecular-weight fraction relative to total hydroxy[14C]proline. Prostaglandin El, when present at a concentration as low as 0.25 pg/ml, reduced net collagen production by a factor of one-half, from 8 k 2 to 4 -t 1% (P < 0.05). In contrast, the change in percentage degradation was relatively gradual, rising steadily from the control value of 15 -t 2 to 33 k 2% at 4 p.g/ml (P < 0.05). The increase in degradation, while significant, could not account for the total decrease in collagen production. We conclude that prostaglandin El exerts its inhibitory effect on collagen production in two essentially independent ways: lowering the rate of synthesis and increasing intracellular degradation. However, the decrease in synthesis is greater than the increase in degradation. o 1988 Academic PRSS, I,,~.

different levels. Baum et al. showed that prostaglandin El (PGE1)3 increases the amount of newly synthesized collagen degraded intracellularly (1); Saltzmann et al. reported that collagen production in fibroblasts exposed to CAMP agonists is very sensitive to the length of exposure (2); and recently, Varga et al. found that at least one effect of prostaglandin E2 is to decrease the amount of collagen mRNA, thereby decreasing the amount of collagen produced (3). The objective of the investigation reported here was to determine the relation between synthesis, intracellular degradation, and net production of collagen in

The ability to regulate collagen production has ilmportant consequences for treatment of various fibrotic disorders. Pharmacologic agents such as p agonists and prostaglandins have profound inhibitory effects on collagen production in cultured human fibroblasts; these compounds also raise the level of cyclic AMP, and it has been speculated that there is a causal relation between CAMP levels and the modulation of collagen metabolism. The mechanisms whereby these agents exert their effects are not yet fully understood, but evidence from several groups indicates that collagen metabolism is regulated at ’ This work was supported by NIH Grant HL342’79. ’ To whom correspondence should be addressed at: Schneider Children’s Hospital, Long Island Jewish Medical Center, New Hyde Park, NY 11042.

3 Abbreviations used: PGEl, prostaglandin HFLl, human fetal lung fibroblasts. 441

0003-9861/88 Copyright All rights

El;

$3.00

M 1988 by Academic Press, Inc. of reproduction in any form reserved

442

BARILE

cells exposed to a CAMP agonist, prostaglandin El. In particular, we wished to gauge the contribution that the increase in degradation makes to the decrease in net production during an acute exposure (defined as less than 8 h). The terms production and synthesis are often used synonymously; however, in this paper we distinguish between them. Synthesis, production (or net production), and degradation are related by the equation: Production = synthesis - degradation. Our studies showed that the PGEl-mediated decrease in production is due to both a decrease in synthesis and an increase in degradation. The two mechanisms are independent, but the inhibition of synthesis is quantitatively more significant than the enhancement of degradation. METHODS Materials. All supplies for cell culture were purchased from GIBCO. The following isotopes were purchased from New England Nuclear: [U-“Clproline (280 mCi/mmol); [U-‘Qleucine (300 mCi/mmol); [Z14C]glycine (49 mCi/mmol); [2,3-3H]proline (29 Ci/ mmol); [4,5-aH]leucine (60 Ci/mmol). Human fetal lung fibroblasts (HFLl; CCL-153) were purchased from the American Type Culture Collection. PGEl was purchased from Upjohn. Purified bacteria1 collagenase was purchased from Advanced Biofactures. Cell culture. HFLl cells between population doublings 25 and 35 were used in these experiments. The cells were plated at a density of 1.2 X 104/cm2 and the cultures were visually confluent by 6 days; cultures were used 7 or 8 days after plating. Cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum; the gas phase was 10% COJ90% air. Experiments to measure protein production were conducted using cells grown in lo-cm’ wells of multiwell plates. Experiments designed to study collagen production and degradation used cells grown in 25-cmZ flasks. Media for the various metabolic labeling procedures are described below. PGEl was used at concentrations between 0.25 and 4 @g/ml. Prior to labeling, cells were preincubated for 1 h in medium lacking isotope but containing the appropriate additives. Protein productim. Protein production was studied by measuring the incorporation of radioactive amino acids into trichloroacetic acid-insoluble material. The cell layer and the medium compartments were analyzed together. The cells were incubated with isotope ([3H]proline, [‘?]glycine, or [3H]leucine; 1 &i/ml, 1.5 ml/well) for 6 h; the labeling medium

ET

AL.

consisted of Hanks’ balanced salt solution supplemented with sodium bicarbonate (to a final concentration of 3.2 g/liter), dialyzed fetal bovine serum (2%), and ascorbic acid (50 rg/ml). Collagen pro&&m. Percentage collagen production was determined by the collagenase digestion technique of Peterkofsky and Diegelmann (4, 5) as modified by Berg et al. (6). When [3H]proline was used, the labeling medium consisted of Eagle’s minimum essential medium without glutamine (GIBCO “Autoclavable MEM”) and supplemented with 2% dialyzed fetal bovine serum and ascorbic acid (50 pg/ml), and the concentration of sodium bicarbonate was increased to 3.0 g/liter. When [“Hlproline and either [‘“Clleucine or [i4C]glycine were used in the same experiment, the labeling medium consisted of Hanks’ solution supplemented as described above. Each isotope was used at a concentration of 1 &i/ml (3 ml/flask) and the labeling time was 6 h. Experiments were done using triplicate controls and PGEl-exposed cultures. At the end of the incubation, the cell layer and the medium in each culture were combined and processed as described by Berg et al. (6). Percentage collagen production was calculated according to the following formula, which is a generalization of a formula originally derived by Breul et ~11. (7) for use when radioactive proline is the isotopic label: Percentage

collagen

production

(collagenase released = A X (total dpm) ~ B X (collagenase

dpm)

X 100

released

[II dpm)

The term “total dpm” refers to total radioactivity in an acid-insoluble fraction. The factors A and B account for differences in frequencies of distribution of the specific amino acids in the collagenase-sensitive and -insensitive regions of the procollagen molecule and in noncollagen proteins. For proline these factors are A = 3.37, B = 2.71; for glycine, A = 3.40, B = 2.73; and for leucine, A = 0.25, B = 0.56. These numbers were derived using published data cited by Breul et al. (7) to determine the number of prolyl, leucyl, and glycyl residues in the al(I) and (u2(1) chains of human collagen, in the C-terminal extension peptides of human proal(1) and prool2(1) chains, and in the N-terminal extension peptides of bovine proal(1) and proLu2(1) chains. The foregoing analysis applies only to type I collagen. Approximately 10% of the collagen synthesized by HFLl cells is type III; however, as discussed by Breul et al. (7) the error introduced in the calculations is negligible. Analysis of the amino acid compositions of 53 proteins showed that the average frequency of leucine was approximately 7% and the average frequency of glycine was approximately 7.5% (8).

EFFECTS

OF

PROSTAGLANDIN

Collagen degradation. Percentage collagen degradation was measured by the method of Bienkowski and Engels (9). Cells were incubated with [‘%]proline (1 &i/ml; 5 m‘l/flask) for 6 h; the labeling medium consisted of Eagle’s minimum essential medium modified as described above. The medium and cell layer were combined. Highand low-molecularweight species were separated according to their solubilities in ethanol/O.1 N NaCl (3:l). Hydroxy[%]proline was measured in each fraction and percentage collagen de,gradation was calculated as 100 times the amount of hydroxy[%]proline in the low-molecular-weight fraction divided by total hydroxy[%]proline. RESULTS

Eflect of PGEl on protein production and collagen production. The effect of PGEl on the amount of radioactively labeled amino acid that HFLl cells incorporated into protein (defined operationally as acid-insoluble material) was very sensitive to the species of amino acid used. As shown in Fig. 1, incorporation of leucine was not significantly affected, but incorporation of both glycine and proline was inhibited by approximately 25% in cells exposed to 1 pg/ml PGEl. The most conservative interpretation of these data is that PGEl does not significantly affect the rate of total protein production, but rather the specific activities of the precursor pools of proline and glycine.

.9

60 L--0

1

PGEl

Concentration

2

(pg/ml)

FIG. 1. Effect of prostaglandin El on incorporation of various labeled amino acids into trichloroacetic acid-insoluble material. Each point is the average of three separate determinations and the error bars represent 1 SD. The three labeling experiments were conducted simultaneously on parallel sets of culture plates.

El

ON

443

COLLAGEN TABLE

I

EFFECTOF PGEl ONPERCENTAGECOLLAGEN SYNTHESISINHUMANFETALLUNG FIBROBLAST~ Control (%I Experiment Experiment Experiment Experiment

1 2 3 4

6.6 6.1 10.0 12.2

f f + +

PGEl (%) 0.6 0.6 2.0 1.2

3.2 3.4 5.0 6.2

L f * +

Pb 1.2 0.8 1.0 0.2

<0.05 to.05 <0.05 <0.05

“[3H]Proline was used as the isotopic label. Percentage collagen synthesis calculated using Eq. [l]. Each value is the mean t 1 SD for three separate cultures. *P values for differences between control and PGEl values were calculated using the two-tailed t test.

Several experiments employing [3H]proline as the label showed that PGEl, at a concentration of 2 pg/ml, significantly reduced percentage collagen production. Results for four independent experiments are shown in Table I. In each experiment, the difference between control and PGElexposed cultures is significant at P < 0.05. The extent of inhibition of percentage collagen production by PGEl was the same in all experiments, viz one-half. These results are in agreement with previous work reported by Crystal and colleagues (1, 2). The reason for the variation in values of percentage collagen production between experiments is not clear. The same line of human fetal lung fibroblasts was used throughout these experiments; cells were used between population doublings 25 and 35; and all cultures were in approximately the same state of confluence. These experiments were conducted over a period of 3 years, and the possibility that different batches of fetal bovine serum caused differences in expression of the collagen phenotype cannot be excluded. As already noted, all these experiments were carried out using rH]proline. Variations in utilization of proline with PGEl concentration should not affect the measurement of percentage collagen produc-

444

BARILE

tion because the assay is independent of changes in specific activity of the isotope or changes in hydroxylation of prolyl residues. To check this line of reasoning, experiments were carried out using [3H]proline and either [14C]glycine or [‘“Clleucine to measure percentage collagen production as a function of PGEl concentration (10). The results are shown in Fig. 2. Percentage collagen production declines in the same way regardless of the label used. The curves for [3H]proline and [‘“Clleucine are almost coincident. It is not clear why the curves for [3H]proline and [14C]glycine are not in closer agreement, but it may be due to our estimate of the frequency of glycine in noncollagen proteins; nevertheless, both curves drop to approximately one-half of the control value at the lowest dose of PGEl (0.25 bg/ml) used. An interesting aspect of the curves presented in Fig. 2 is that after the initial decrease, collagen production remains constant over a broad range of concentrations (i.e., 0.25 to 4 pg/ml). Proline and glycine are nonessential amino acids that are overrepresented in collagen, and proline equilibrates very slowly among the various extracellular and intracellular pools (11). In contrast, leucine is an essential amino acid that is underrepresented in collagen, and it equil-

0

a-0

[SH]Proline

O--O

[14C]Glycine

1

2

AL

ibrates rapidly among the various pools (12). We consider the close agreement between values of percentage collagen production obtained using [3H]proline, [‘“Clglycine, and [‘4C]leucine to label the newly synthesized proteins to be a strong indication that the decreases in collagen production observed in the presence of PGEl are not due to changes in the activity or utilization of a particular isotope, but rather indicate the cellular response to the prostaglandin. Effects of PGEl on intracellular collagen degradation. The effect of PGEl on intracellular collagen degradation was studied in several experiments. A dose-response curve, covering the range of concentrations between 0 and 4 pg/ml, is shown in Fig. 3. Percentage degradation increases with the concentration of PGEl, rising from the control value of approximately 15 + 2 to 33 & 3% at 4 pg/ml. The relatively gradual increase in degradation over this range is in marked contrast to the steep decline in collagen production between 0 and 0.25 pg/ml shown in Fig. 2. A consistent finding in a series of eight independent experiments has been that PGEl at 2 or 4 pg/ml causes an increase of approximately 1.5- to 2.5-fold in percentage degradation; in each case, the difference is significant at the P < 0.05 level.

0

PGEl

FIG. 2. Effect amino acids to separate sets of three separate deviations range protocol.

ET

1

Concentration

2

3

4

(w/ml)

of prostaglandin El on percentage collagen production assessed using different label newly synthesized proteins. (A) [3H]Proline and [‘4Clglycine were used in cultures that were incubated and processed in parallel. Each point is the average of determinations. The error bars have been omitted for clarity, but the standard from 0.1 to 0.7%. (B) [3H]Proline and [‘%]leucine were used in a double-label

EFFECTS

OF

PROSTAGLANDIN

El

ON

445

COLLAGEN

DISCUSSION

From the data presented here we conclude that tlhe decrease in net collagen production in cultured human fetal lung fibroblasts incubated with PGEl is due to both a decrease in synthesis and an increase in degradation, but the effect on synthesis is greater. Our findings are summarized in Fig. 4. Collagen accounts for approximately 8 to 9% of total protein produced by the HFLl cells under control conditions. Because 15% of newly synthesized collagen is degraded rapidly, collagen really accounts for about 10% [=8.5%/(1 -- 0.15)] of the total protein synthesized by the cells. At a concentration of 4 pg/ml, PGEl decreases percentage collagen production to approximately 4% and increases collagen degradation to approximately 33%. Making the conservative assumption that PGEl does not appreciably decrease the rate of total protein synthesis (but rather decreases the specific activity of certain precursor pools, see Fig. l), and even taking account of intracellular degradation, it can be seen that the major effect of PGEl is to decrease the absolute rate of collagen synthesis by approximately onehalf. The most likely explanation for the reduction in collagen synthesis is a reduction in the amount of collagen mRNA. As noted in the introduction, Varga et al. (3) have found this to be the case for human fibroblasts exposed to prostaglandin E2.

FIG. 3. Effect of prostaglandin degradation
El on intracellular collagen in fibroof four determina1 SD.

Control

PGEl

FIG. 4. Effect of prostaglandin El on synthesis, net production, and intracellular degradation of collagen. This figure integrates data presented in Figs. 2 and 3 for a prostaglandin concentration of 4 pg/ml. The combined height of the solid and cross-hatched bars (net production + intracellular degradation) represents the amount of collagen synthesized by the culture; it is expressed as a percentage of total protein produced.

(Clearly, total protein synthesis must decrease at least to the extent that collagen synthesis decreases; however, a change from 9 to 4 or 5% in total synthesis is within the experimental errors shown in Fig. 1.) Notwithstanding its relatively minor role in modulating collagen production in response to PGEl, the increase in percentage intracellular degradation is significant and of considerable biological interest in the context of the metabolic economy of synthesis and processing of secretory proteins. In this regard, it should be noted that comparison of Figs. 2 and 3 reveals that collagen production and degradation respond very differently to the prostaglandin: Production drops by one-half at very low concentrations (0.25 pg/ml) and remains constant over a 16fold increase in concentration. In contrast, degradation increases very slowly with changes in the concentration of PGEl; indeed, it is essentially unaffected at a very low dose that elicits a maximum drop in production. REFERENCES 1. BAUM, B. J., Moss, J., BREUL, AND CRYSTAL, R. G. (1980)

2843-2847.

S. D., BERG, R. A., J. BioL Chem. 255,

446

BARILE

2. SALTZMANN,L.E.,MOSS,J.,BERG,R. A.,HoM,B., AND CRYSTAL, R. G. (1982) Biochem J. 204, 25-30. 3. VARGA,J.,DIAZ-PEREZ,A.,ROSENBLOOM,J.,AND JIMENEZ, S. A. (1987) Biochem. Biophys. Rea Commun. 147.1282-1288. 4. PETERKOFSKY, B., AND DIEGELMANN, R. (1971) Biochemistry 10,988-994. 5. PETERKOFSKY,B., CHOJKIER, M., AND BATEMAN, J. (1982) in Immunochemistry of the Extracellular Matrix (Furthmayr, H., Ed.), Vol. II, pp. 19-47, CRC Press, Boca Raton, FL. 6. BERG,R. A.,Moss,J.,BAuM,B.J., AND CRYSTAL, R. G.(1981)J. Clin. Invest. 67,1457-1462. 7. BREUL, S. D., BRADLEY, K. H., HANCE, A. J.,

ET AL

8. 9. 10. 11. 12.

SCHAFER, M. P., BERG, R. A., AND CRYSTAL, R. G.(1980)J. Biol. Chem. 255,5250-5230. DOOLITTLE, R. F. (1979) in The Proteins (Neurath, H., and Hill, R. L., Eds.), Third ed., Vol. 4, pp. 1-118, Academic Press, New York. BIENKOWSKI, R. S., AND ENGELS, C. J. (1981) Anal. B&hem. 116,414-424. SCHEIN, J., HARSCH, M., CYWINSKI, C., AND RoSENBLOOM, J. (1980) Arch. B&hem. Biophys. 203,572-579. HILDEBRAN,J.N., AIRHART,J.,STIREWALT, W.S., AND Low, R. B. (1981) Biochem. J. 198,249-258. Low, R. B., HILDEBRAN, J. N., ABSHER, M., STIREWALT,~. S., AND ARNOLD,J.(~~~~)C~~nect. Tissue Res. 14, 179-185.