Biochimica et Biophysica Acta, 1091 (1991) 231-235 © 1991 Elsevier Science Publishers B.V. (Biomedical Division) 0167-4889/91/$03.50 ADONIS 016748899100089X
231
BBAMCR 12872
The effect of transforrmng growth factor fl on rates of procollagen synthesis and degradation in vitro Robin J. McAnulty, Juan S. Campa, Alison D. Cambrey and Geoffrey J. Laurent Biochemistry Unit, National Heart and Lung Institute, University of London, London (U.K.)
(Received24 July 1990)
Key words: TGFfl;Procollagen;Collagen; Fibroblast; Collagensynthesis;Collagendegradation; Growthfactor
Transforming growth factor fl (TGFfl) is known to stimulate procollagen production and steady-state levels of procollagen mRNAs, but its ability to affect post-translational processing of procollagen has been little studied. This paper demonstrates the application of recently developed ultrasensitive methods for measuring hydroxyproline to assess rates of procoilagen synthesis and degradation in vitro with and without TGFfl. Foetal rat fibroblasts synthesized 8.63 + 0.21 pmol hydroxyproline//tg DNA per h, which corresponds to approx. 40 molecules of procollagen/cell per s. Additimt of TGFfl to cultures increased total amounts of procollagen synthesized and degraded by 112% and 82%, respectively, but there was a significant decrease in the proportion of procollagen degraded (control, 38.0 + 1.1%; TGFfl, 32.3:1: 0.9%; P < 0.005). This study demonstrates a novel mechanism which may contribute to the TGFfl-induced increase in procollagen production by fibroblasts.
Introduction TGFfl is a polypeptide with a molecular weight of approx. 25 kDa. It is known to be a potent stimulant of procollagen synthesis [1-5] associated with increased steady-state levels of procollagen mRNAs [5-8]. It is uncertain whether the increased levels of mRNAs are due to increased transcription, increased stability of the mRNA, or both [5,6,8]. There is also evidence that TGFfl may affect extracellular collagen degradation by decreasing collagenase, and increasing tissue inhibitor of metalloproteinase (TIMP) and a2-macroglobulin production [9-11]. However, there have been no reports on the effects of TGFfl on intracellular procollagen degradation. In this study we have used methods recently developed in our laboratory [12] for measuring picomolar amounts of hydroxyproline by reverse-phase HPLC, to estimate directly the ratcs of procollagen synthesis and degradation by cells in culture as well as the proportion Abbreviations: TGFfl, transforming growth factor fl; TIMP, tissue inhibitor of metalloproteinase;DMEM, Dulbecco'smodified Eagle's medium; NCS, newborncalf serum; PBS, phosphate-bufferedsaline; PCA, perchloricacid. Correspondence: R.J. McAnulty,BiochemistryUnit, National Heart and LungInstitute, Universityof London, EmmanuelKaye Building, Manresa Road, London, SW3 6LR, U.K.
degraded rapidly intracellularly. We have also examined the effects of TGFfl on these processes. Methods
Cell culture Foetal rat fibroblasts (Rat 2, obtained from ATCC, Rockville, MD, U.S.A.) were cultured in 2.4 cm diameter plates with DMEM +5% NCS in a humidified atmosphere containing 10% CO 2 at 37 ° C. When cells appeared confluent they were incubated for a further 24 h. The media was then removed and replaced with 1 ml of pre-incubation media, which contained 4 mM glutamine, 50 # g / m l ascorbic acid, 0.2 mM proline and 2~ (v/v) NCS and incubated for 24 h. The pre-incubation media was then removed and a further I ml pre-incubation media was added with or without porcine TGFfll (British Biotechnology, Cowley, Oxford) at a concentration of 5 ng/ml (200 pM) and incubated for a further 24 h before harvesting. For time-course experiments ce~ls wcre grown in 10 cm diameter petri dishes.
Sample preparation For cultures in which procollagen metabolism was to be assessed the cell layer was scraped into the media and aspirated. Each well was washed with 1 ml PBS and the washings were combined with the initial aspirate. The samples were then boiled for 15 rain and cooled, and proteins were precipitated by addition of ethanol to
232 give a final concentration of 67% (v/v) and left at 4 ° C overnight. The precipitated proteins were pelleted by centrifugation at 30000 × g for 30 min, the supernatant was retained and the pellet washed twice in ethanol (67%, v/v). The combined supernatants and the protein pellets were evaporated to dryness using a Dri-Block Sample Concentrator (Techne DB-3, SC-3) and a vacuum desiccator, respectively, prior to hydrolysis in 2 ml 6 M hydrochloric acid at 10°C overnight. Hydrolysates were mixed with approx. 70 mg charcoal and filtered (Millipore, 0.65/~m) prior to chromatography. Similar cultures were set up to assess cell numbers by counting and measurement of DNA. At the end of the incubation the medium was removed and discarded. For DNA analysis the cell layer was scraped into 1 ml PBS, aspirated into a microcentrifuge tube and centrifuged at 9000 x g for 5 min; the clear supernatant was discarded. Any remaining cells were washed from the well with a further 1 ml PBS, aspirated and combined with the initial cell pellet. After centrifugation the supematant was discarded and the pellet was stored at - 2 0 ° C prior to analysis. In order to estimate cell number, cells were removed from the wells by trypsinization, stained with crystal violet and counted in an Improved Neubauer haemocytometer.
Measurement of hydroxyproline and calculation of procollagen synthesis and the proportion degraded Hydroxyproline was isolated and measured by reverse-phase HPLC of 7-chloro-4-nitrobenzo-2-oxa-l,3diazole (NBD-C1)-derivatized hydrolysates as described previously [12]. Hydroxyproline content was determined by comparing peak area of samples from the chromatogram to that generated from standard solutions, derivatized and separated under similar conditions and run at the beginning and end of each day. As the cell monolayer contains a small amount of procoUagen and the serum contained hydroxyproline, the amount of hydroxyproline present in the combined culture medium and cell layer at the start of incubation was determined in both the ethanol-soluble and -insoluble fractions. This background level, which represented 1.08 :l: 0.01 nmol hydroxyproline/weU in the insoluble fraction and 1.73 + 0.12 nmol hydroxyproline/weU in the soluble fraction, was then subtracted from the sample values. The hydroxyproline present in the ethanol-soluble fraction from each culture was taken to represent hydroxyproline derived from the degradation of procollasen during the culture period, whilst the hydroxyproline found in the ethanol-insoluble fraction was taken to represent hydroxyproline in procollagen. The proportion of procoUagen degraded can be calculated by dividing the hydroxyproline content of the ethanol-soluble fraction by the sum of the hydroxyproline in the ethanol-soluble and -insoluble fractions. The rate of synthesis was obtained from the combined values for
ethanol-soluble and -insoluble fractions and rates of degradation and net production were obtained from values for ethanol-soluble and -insoluble fractions, respectively, and expressed per #g DNA per h.
DNA estimation Estimates of DNA were obtained as described previously [13] with modifications to allow estimates to be made from small numbers of cells. Briefly, assays were carried out in microcentrifuge tubes at 4 ° C. Samples were acidified with 1 ml 0.5 M PCA, allowed to stand for 10 rain, then centrifuged at 9000 x g for 5 min. The supernatant was discarded and the precipitate.washed twice in 1 ml 0.5 M PCA. The pellet was resuspended in 1.2 ml 0.8 M PCA and heated to 70°C for 45 min with occasional shaking. After standing on ice for 15 rain the samples were centrifuged (9000 >" r
5
4
8
12
16
20
24
Time in culture (hours)
Fig. 1. Hydroxyproline production with time. Total hydroxyproline production (i.e., medium +cell layer) by cultures of feotal rat fibroblasts was measured directly by reverse phase HPLC at various times up to 24 h. Each value represents the mean5:S.E. for the combined estimate of hydroxyproline in ethanol-soluble and ethanol-insoluble fractions from four replicate cultures. hydroxyproline compared with that in the corresponding control, whereas the hydroxyproline in the soluble fraction increased by about 80% compared with control values. This resulted in a significant decrease in the proportion of procollagen degraded. TABLE I
Proportion of newly symhesisedprocollagen degraded Percentage degradation of newly synthesised procollagen was calculated from the ratio of hydroxyproline in the ethanol-soluble fraction to total hydroxyproline (i.e., ethanol-soluble + ethanol-insoluble fractions) multiplied by 100. Data is given for fibroblasts cultured with and without TGFfl (5 ng/ml). Each value represents the mean 5: S.E. for the number of replicate cultures indicated (n).
Control (n -- 12) TGFfl (n =6)
Hydroxyproline (pmol) ethanolethanolinsoluble soluble 1557+ 57 948 + 24 36035:63 17245:62 (P < 0.001) (P < 0.001)
Degradation (%)
38.0+ 1.1 32.3+0.9 (P < 0.005)
TABLE 11 Effect of TGFfl on procollagensynthesis and degradation Rates of hydroxyproline synthesis (ethanol-soluble+ ethanol-insoluble fractions) and degradation (ethanol-soluble fraction) are expressed per #g DNA per h for fibroblasts cultured with and without TGFfl. Net production represents the difference between synthesis and degradation. Each value shows the mean+S.E, for the number of replicate cultures indicated (n).
Control (n -- 11) TGFfl (n =6)
pmol hydroxyproline//~g DNA per h synthesis degradation net production 8.63 + 0.21 3.26 5:0.08 5.365:0.19 18.33+0.31 5.93+0.21 12.40+0.22 P < 0.001 P < 0.001 P < 0.001
Table II shows the data presented as rates of synthesis and degradation with respect to DNA. A production rate of 8.63 + 0.21 pmol hydroxyproline/#g DNA per h was obtained for control cultures, a value almost identical to that obtained from the time-course experiment (Fig. 1). Assuming a molecular weight of 131 for hydroxyproline, that hydroxyproline represents 12.2% (w/w) of collagen, that collagen has a molecular weight of 300000 and Avagadro's number is 6.02252.10 23, this represents a production rate of approx. 40 molecules of procollagen/cell per s. Addition of TGFfl increased synthesis by approx. 112% and degradation by about 82%. Net production, which represents the difference between synthesis and degradation, increased by about 1307o. Discussion
Assessment of a novel method for measuring procollagen synthesis and degradation Procollagens of various types are synthesised by many cells, including fibroblasts, and after cleavage of the propeptides are secreted as collagens. Procollagen contains hydroxyproline, an imino acid which is not incorporated directly into proteins but is produced posttranslationally by hydroxylation of proline. Hydroxyproline is present in relatively few proteins and predominantly in procollagen/collagen and can therefore be measured both as a marker of procollagen/collagen synthesis in intact proteins and of procollagen degradation when measured in low-molecular-weight degradation products. Although simple in theory, the amounts of procollagen synthesised and degraded by cells in culture are often small, and until recently methods have not been sensitive enough to measure the hydroxyproline directly. This has resulted in the use of radiolabelled proline with a high specific activity followed by its subsequent measurement in hydroxyproline. To measure rates of procollagen synthesis and degradation using radiolabelled proline assumptions have to be made in relation to the specific activity of tile precursor pool for protein synthesis. For cells in culture this does not equilibrate with the specific activity of either the extracellular or intracellular free pools [14-16]. To avoid these problems, measurements of the specific activities of t R N A or of proline or hydroxyproline in procollagen have been used as an estimate of the precursor pool [14-16]. These methods are technically difficult, requiring relatively large numbers of cells. In the present studies we have measured hydroxyproline directly using an ultrasensitive HPLC method. This was a major advantage, since it allowed us to calculate rates of synthesis and degradation directly, thus avoiding problems associated with precursor pool estimates. Another assumption implicit in the methods used here is that free hydroxyproline is derived from de-
234 gradation of procollagen. It is generally believed that hydroxyproline is only produced in biological systems by enzymatic, post-translational hydroxylation of proline. However, it has been suggested that hydroxyproline may be formed non-enzymaticaUy from proline in the presence of active oxygen species [17]. This does not appear to occur in in vitro cell culture systems such as those employed here, since addition of protein synthesis inhibitors to cultures inhibits hydroxyproline production, demonstrating that it is derived from newly synthesized proteins [18-20]. The results given in Table I show measurements of the proportion of procollagen synthesized by foetal rat fibroblasts which is degraded. This degradation of procollagen has previously been shown to occur rapidly and intracellularly [21,22]. Values of about 38~ were obtained from measurements of molar amounts of hydroxyproline. Th~,se values are at the upper end of the range of values reported for fibroblasts which range from about 95 to 405 [23]. One possible explanation for these differences is the use of varying concentrations of proline in the culture medium. In the present studies, 0.2 mM proline was added to the cultures. Decreasing or increasing this amount of proline produces a concomit~mt decrease or increase in the proportion of collagen degraded (data not showa).
Effect of TGFt3 At doses similar to that used here TGF/3 has been reported to cause a 2-3-fold increase in procollagen production. However, there have been no studies examining its effect on intracellular procollagen degradation. In this study we report a 1125 increase in synthesis and an 825 increase in the rate of intracellular procollagen degradation. Thus, there was a decrease in the proportion of procollagen degraded. Without this decrease in the proportion of procollagen degraded, the TGF,&induced increase in net procollagen produced would have been about 1750 pmol compared with the 2050 pmol increase observed (i.e., if the proportion of procollagen degraded had remained at 385 rather than decreasing to 325), a difference which represents about 205 of the production in control cultures. This reduction in the proportion of procollagen degraded, although small in percentage terms, clearly has the potential to play an important role in the increased net production of procollagen induced by TGF-~. TGF/3 has been shown to stimulate endogenous prostaglandin E2 (PGE2) production by fibroblasts [24]. PGE2 is known to inhibit collagen synthesis and increase the proportion of newly synthesised procollagen degraded. If prostaglandins were produced by fibroblasts in the present studies this could have limited the effects of TGF/3 on collagen degradation observed here. This could be tested by adding inhibitors of PGE synthesis such as indomethacin to the cultures. The
studies of Diaz et al. [24] have already demonstrated that TGFfl-induced increases in collagen production are further enhanced by blocking PGE2 synthesis. The mechanism by which TGFfl reduces the proportion of procollagen degraded intracellularly is unclear, but several possible explanations exist. First, the proportion of procollagen degraded can be affected by the degree of proline hydroxylation. If procollagen is underhydroxylated it does not form a stable triple helix and is more susceptible to degradation. Therefore, if TGF~ increased hydroxylation of proline it may lead to a more stable molecule which would be less susceptible to degradation. However, the evidence available at present suggests that TGFfl does not affect prolyl hydroxylation [4] and is therefore unlikely to be a contributor to the changes in degradation observed here. Another possible mechanism by which TGFfl could inhibit the proportion of newly synthesised procollagen degraded is by inhibiting the production of ,ysosomal acid proteinases which are thought to be involved in degradation of procollagen intracellularly. There is no evidence for TGF~ affecting these particular enzymes at present, but it is known to be capable of decreasing production of other proteinases, such as collagenase [9,10] and elastase [25], which are capable of degrading collagen extracellularly. TGF~ also stimulates production of the anticollagenases, TIMP and ,~2-macroglobulin [9-11]. In summary, we have demonstrated that procollagen synthesis and degradation can be estimated directly in small cultures without the use of radiolabelled compounds. The use of these methods eliminate assumptions relating to precursor pool specific activities involved with incorportion of radioisotopic tracers. Addition of TGF/] to these fibroblast cultures increased rates of synthesis and degradation of procollagen with a decrease in the proportion degraded. This decrease could play a significant role in the TGF~-induced increase in collagen production.
Acknowledgements This work was supported in part by the Clinical Research Committee of the National Heart and Chest Hospitals. References 1 lgnotz, R.A. and Massague, J. (1986) J. Biol. Chem. 261, 43374345. 2 Roberts, A.B., Sporn, M.B., Assoian, R.K., Smith, J.M., Roche, N.S., Wakefield, L.M., Heine, U.I., Liotta, L.A., Falanga, V., Kehrl, J.H. and Fau¢i, A.S. (1986) Proc. Natl. Acad. Sci. USA 83, 4167-4171. 3 Varga, J. and Jimencz, S.A. (1986) Biochem. Biophys. Res. Commun. 138, 974-980. 4 Fine, A. and Ooldstein, H. (1987) J. Biol. Chem. 262, 3897-3902.
235 5 Raghow, R., Postlethwaite, A.E., Keski-Oja, J., Moses, H.L. and Kang, A.H. (1987) J. Clin. Invest. 79, 1285-1288. 6 Ignotz, R.A., Endo, T. and Massague, J. (1987) J. Biol. Chem. 262, 6443-6446. 7 Varga, J., Rosenbloom, J. and Jimenez, S.A. (1987) Biochem. J. 247, 597-604. 8 Pentinnen, R.P., Kobayashi, S. and Bornstein, P. (1988) Proc. Natl. Acad. Sci. USA 85, 1105-1108. 9 Edwards, D.R., Murphy, G., Reynolds, J.J., Whitman, S.E., Docherty, A.J.P., Angel, P. and Heath, J.K. (1987) EMBO J. 6, 1899-1904. 10 Overall, C.M., Wrana, J.L. and Sodek, J. (1989) J. Biol. Chem. 264, 1860-1869. 11 Shi, D.L., Savona, C., Cagnon, J., Cachet, C., Chambaz, E.M. and Feige, J.J. (1990) J. Biol. Chem. 265, 2881-2887. 12 Campa, J.S., McAnulty, R.J. and Laurent, G.J. (1990) Anal. Biochem. 186, 257-263. 13 Burton, K. (1956) Biochem. J. 62, 315-322. 14 Hildebran, J.N., Airhart, J., Stirewalt, W.S. and Low, R.B. (19Sl) Biochem. J, 198, 249-258. 15 Clark, J.G. and Hildebran, J.N. (1984) Anal. Biochem. 140, 478485.
16 Bellon, G., Monboisse, J., Randoux, A. and Borel, J. (I987) Biochim. Biophys. Acta 930, 39-47. 17 Trelstad, R.L., Lawley, K.R. and Holmes, L.B. (1981) Nature 289, 310-312. 18 Steinmann, B., Rao, V.H. and Gitzelman, R. (198,,1) FEBS Lett. 133, 142-144. 19 Imberman, M., Oppenheim, F. aild Franzblau, C. (1982) Biochim. Biophys. Acta 719, 480-487. 20 Bienkowski, R.S. (1984) J. Cell. Physiol. 121, 152-158. 21 Bienkowski, R.S., Cowan, M.J., McDonald, J.A. and Crystal, R.G. (1978) J. Biol. Chem. 253, 4356-4363. 22 Bienkowski, R.S., Baum, B.J. and C,"jstal, R.G. (1978) Nature 276, 413-416. 23 Rennard, S.I., Stier, L.E. and Crystal, R.G. (1982) J. Invest. Dermatol. 79 suppl. 1, 77s-82s. 24 Diaz, A., Varga, J. and Jimenez, S.A. (1989) J. Biol. Chem. 264, 11554-11557. 25 Redini, F., Lafuma, C., Pujol, J., Robert, L. and Hornebeck, W. (1988) Biochem. Biophys. Res. Commun. 155, 786-793.
231
BBAMCR 12872
The effect of transforrmng growth factor fl on rates of procollagen synthesis and degradation in vitro Robin J. McAnulty, Juan S. Campa, Alison D. Cambrey and Geoffrey J. Laurent Biochemistry Unit, National Heart and Lung Institute, University of London, London (U.K.)
(Received24 July 1990)
Key words: TGFfl;Procollagen;Collagen; Fibroblast; Collagensynthesis;Collagendegradation; Growthfactor
Transforming growth factor fl (TGFfl) is known to stimulate procollagen production and steady-state levels of procollagen mRNAs, but its ability to affect post-translational processing of procollagen has been little studied. This paper demonstrates the application of recently developed ultrasensitive methods for measuring hydroxyproline to assess rates of procoilagen synthesis and degradation in vitro with and without TGFfl. Foetal rat fibroblasts synthesized 8.63 + 0.21 pmol hydroxyproline//tg DNA per h, which corresponds to approx. 40 molecules of procollagen/cell per s. Additimt of TGFfl to cultures increased total amounts of procollagen synthesized and degraded by 112% and 82%, respectively, but there was a significant decrease in the proportion of procollagen degraded (control, 38.0 + 1.1%; TGFfl, 32.3:1: 0.9%; P < 0.005). This study demonstrates a novel mechanism which may contribute to the TGFfl-induced increase in procollagen production by fibroblasts.
Introduction TGFfl is a polypeptide with a molecular weight of approx. 25 kDa. It is known to be a potent stimulant of procollagen synthesis [1-5] associated with increased steady-state levels of procollagen mRNAs [5-8]. It is uncertain whether the increased levels of mRNAs are due to increased transcription, increased stability of the mRNA, or both [5,6,8]. There is also evidence that TGFfl may affect extracellular collagen degradation by decreasing collagenase, and increasing tissue inhibitor of metalloproteinase (TIMP) and a2-macroglobulin production [9-11]. However, there have been no reports on the effects of TGFfl on intracellular procollagen degradation. In this study we have used methods recently developed in our laboratory [12] for measuring picomolar amounts of hydroxyproline by reverse-phase HPLC, to estimate directly the ratcs of procollagen synthesis and degradation by cells in culture as well as the proportion Abbreviations: TGFfl, transforming growth factor fl; TIMP, tissue inhibitor of metalloproteinase;DMEM, Dulbecco'smodified Eagle's medium; NCS, newborncalf serum; PBS, phosphate-bufferedsaline; PCA, perchloricacid. Correspondence: R.J. McAnulty,BiochemistryUnit, National Heart and LungInstitute, Universityof London, EmmanuelKaye Building, Manresa Road, London, SW3 6LR, U.K.
degraded rapidly intracellularly. We have also examined the effects of TGFfl on these processes. Methods
Cell culture Foetal rat fibroblasts (Rat 2, obtained from ATCC, Rockville, MD, U.S.A.) were cultured in 2.4 cm diameter plates with DMEM +5% NCS in a humidified atmosphere containing 10% CO 2 at 37 ° C. When cells appeared confluent they were incubated for a further 24 h. The media was then removed and replaced with 1 ml of pre-incubation media, which contained 4 mM glutamine, 50 # g / m l ascorbic acid, 0.2 mM proline and 2~ (v/v) NCS and incubated for 24 h. The pre-incubation media was then removed and a further I ml pre-incubation media was added with or without porcine TGFfll (British Biotechnology, Cowley, Oxford) at a concentration of 5 ng/ml (200 pM) and incubated for a further 24 h before harvesting. For time-course experiments ce~ls wcre grown in 10 cm diameter petri dishes.
Sample preparation For cultures in which procollagen metabolism was to be assessed the cell layer was scraped into the media and aspirated. Each well was washed with 1 ml PBS and the washings were combined with the initial aspirate. The samples were then boiled for 15 rain and cooled, and proteins were precipitated by addition of ethanol to
232 give a final concentration of 67% (v/v) and left at 4 ° C overnight. The precipitated proteins were pelleted by centrifugation at 30000 × g for 30 min, the supernatant was retained and the pellet washed twice in ethanol (67%, v/v). The combined supernatants and the protein pellets were evaporated to dryness using a Dri-Block Sample Concentrator (Techne DB-3, SC-3) and a vacuum desiccator, respectively, prior to hydrolysis in 2 ml 6 M hydrochloric acid at 10°C overnight. Hydrolysates were mixed with approx. 70 mg charcoal and filtered (Millipore, 0.65/~m) prior to chromatography. Similar cultures were set up to assess cell numbers by counting and measurement of DNA. At the end of the incubation the medium was removed and discarded. For DNA analysis the cell layer was scraped into 1 ml PBS, aspirated into a microcentrifuge tube and centrifuged at 9000 x g for 5 min; the clear supernatant was discarded. Any remaining cells were washed from the well with a further 1 ml PBS, aspirated and combined with the initial cell pellet. After centrifugation the supematant was discarded and the pellet was stored at - 2 0 ° C prior to analysis. In order to estimate cell number, cells were removed from the wells by trypsinization, stained with crystal violet and counted in an Improved Neubauer haemocytometer.
Measurement of hydroxyproline and calculation of procollagen synthesis and the proportion degraded Hydroxyproline was isolated and measured by reverse-phase HPLC of 7-chloro-4-nitrobenzo-2-oxa-l,3diazole (NBD-C1)-derivatized hydrolysates as described previously [12]. Hydroxyproline content was determined by comparing peak area of samples from the chromatogram to that generated from standard solutions, derivatized and separated under similar conditions and run at the beginning and end of each day. As the cell monolayer contains a small amount of procoUagen and the serum contained hydroxyproline, the amount of hydroxyproline present in the combined culture medium and cell layer at the start of incubation was determined in both the ethanol-soluble and -insoluble fractions. This background level, which represented 1.08 :l: 0.01 nmol hydroxyproline/weU in the insoluble fraction and 1.73 + 0.12 nmol hydroxyproline/weU in the soluble fraction, was then subtracted from the sample values. The hydroxyproline present in the ethanol-soluble fraction from each culture was taken to represent hydroxyproline derived from the degradation of procollasen during the culture period, whilst the hydroxyproline found in the ethanol-insoluble fraction was taken to represent hydroxyproline in procollagen. The proportion of procoUagen degraded can be calculated by dividing the hydroxyproline content of the ethanol-soluble fraction by the sum of the hydroxyproline in the ethanol-soluble and -insoluble fractions. The rate of synthesis was obtained from the combined values for
ethanol-soluble and -insoluble fractions and rates of degradation and net production were obtained from values for ethanol-soluble and -insoluble fractions, respectively, and expressed per #g DNA per h.
DNA estimation Estimates of DNA were obtained as described previously [13] with modifications to allow estimates to be made from small numbers of cells. Briefly, assays were carried out in microcentrifuge tubes at 4 ° C. Samples were acidified with 1 ml 0.5 M PCA, allowed to stand for 10 rain, then centrifuged at 9000 x g for 5 min. The supernatant was discarded and the precipitate.washed twice in 1 ml 0.5 M PCA. The pellet was resuspended in 1.2 ml 0.8 M PCA and heated to 70°C for 45 min with occasional shaking. After standing on ice for 15 rain the samples were centrifuged (9000 >" r
5
4
8
12
16
20
24
Time in culture (hours)
Fig. 1. Hydroxyproline production with time. Total hydroxyproline production (i.e., medium +cell layer) by cultures of feotal rat fibroblasts was measured directly by reverse phase HPLC at various times up to 24 h. Each value represents the mean5:S.E. for the combined estimate of hydroxyproline in ethanol-soluble and ethanol-insoluble fractions from four replicate cultures. hydroxyproline compared with that in the corresponding control, whereas the hydroxyproline in the soluble fraction increased by about 80% compared with control values. This resulted in a significant decrease in the proportion of procollagen degraded. TABLE I
Proportion of newly symhesisedprocollagen degraded Percentage degradation of newly synthesised procollagen was calculated from the ratio of hydroxyproline in the ethanol-soluble fraction to total hydroxyproline (i.e., ethanol-soluble + ethanol-insoluble fractions) multiplied by 100. Data is given for fibroblasts cultured with and without TGFfl (5 ng/ml). Each value represents the mean 5: S.E. for the number of replicate cultures indicated (n).
Control (n -- 12) TGFfl (n =6)
Hydroxyproline (pmol) ethanolethanolinsoluble soluble 1557+ 57 948 + 24 36035:63 17245:62 (P < 0.001) (P < 0.001)
Degradation (%)
38.0+ 1.1 32.3+0.9 (P < 0.005)
TABLE 11 Effect of TGFfl on procollagensynthesis and degradation Rates of hydroxyproline synthesis (ethanol-soluble+ ethanol-insoluble fractions) and degradation (ethanol-soluble fraction) are expressed per #g DNA per h for fibroblasts cultured with and without TGFfl. Net production represents the difference between synthesis and degradation. Each value shows the mean+S.E, for the number of replicate cultures indicated (n).
Control (n -- 11) TGFfl (n =6)
pmol hydroxyproline//~g DNA per h synthesis degradation net production 8.63 + 0.21 3.26 5:0.08 5.365:0.19 18.33+0.31 5.93+0.21 12.40+0.22 P < 0.001 P < 0.001 P < 0.001
Table II shows the data presented as rates of synthesis and degradation with respect to DNA. A production rate of 8.63 + 0.21 pmol hydroxyproline/#g DNA per h was obtained for control cultures, a value almost identical to that obtained from the time-course experiment (Fig. 1). Assuming a molecular weight of 131 for hydroxyproline, that hydroxyproline represents 12.2% (w/w) of collagen, that collagen has a molecular weight of 300000 and Avagadro's number is 6.02252.10 23, this represents a production rate of approx. 40 molecules of procollagen/cell per s. Addition of TGFfl increased synthesis by approx. 112% and degradation by about 82%. Net production, which represents the difference between synthesis and degradation, increased by about 1307o. Discussion
Assessment of a novel method for measuring procollagen synthesis and degradation Procollagens of various types are synthesised by many cells, including fibroblasts, and after cleavage of the propeptides are secreted as collagens. Procollagen contains hydroxyproline, an imino acid which is not incorporated directly into proteins but is produced posttranslationally by hydroxylation of proline. Hydroxyproline is present in relatively few proteins and predominantly in procollagen/collagen and can therefore be measured both as a marker of procollagen/collagen synthesis in intact proteins and of procollagen degradation when measured in low-molecular-weight degradation products. Although simple in theory, the amounts of procollagen synthesised and degraded by cells in culture are often small, and until recently methods have not been sensitive enough to measure the hydroxyproline directly. This has resulted in the use of radiolabelled proline with a high specific activity followed by its subsequent measurement in hydroxyproline. To measure rates of procollagen synthesis and degradation using radiolabelled proline assumptions have to be made in relation to the specific activity of tile precursor pool for protein synthesis. For cells in culture this does not equilibrate with the specific activity of either the extracellular or intracellular free pools [14-16]. To avoid these problems, measurements of the specific activities of t R N A or of proline or hydroxyproline in procollagen have been used as an estimate of the precursor pool [14-16]. These methods are technically difficult, requiring relatively large numbers of cells. In the present studies we have measured hydroxyproline directly using an ultrasensitive HPLC method. This was a major advantage, since it allowed us to calculate rates of synthesis and degradation directly, thus avoiding problems associated with precursor pool estimates. Another assumption implicit in the methods used here is that free hydroxyproline is derived from de-
234 gradation of procollagen. It is generally believed that hydroxyproline is only produced in biological systems by enzymatic, post-translational hydroxylation of proline. However, it has been suggested that hydroxyproline may be formed non-enzymaticaUy from proline in the presence of active oxygen species [17]. This does not appear to occur in in vitro cell culture systems such as those employed here, since addition of protein synthesis inhibitors to cultures inhibits hydroxyproline production, demonstrating that it is derived from newly synthesized proteins [18-20]. The results given in Table I show measurements of the proportion of procollagen synthesized by foetal rat fibroblasts which is degraded. This degradation of procollagen has previously been shown to occur rapidly and intracellularly [21,22]. Values of about 38~ were obtained from measurements of molar amounts of hydroxyproline. Th~,se values are at the upper end of the range of values reported for fibroblasts which range from about 95 to 405 [23]. One possible explanation for these differences is the use of varying concentrations of proline in the culture medium. In the present studies, 0.2 mM proline was added to the cultures. Decreasing or increasing this amount of proline produces a concomit~mt decrease or increase in the proportion of collagen degraded (data not showa).
Effect of TGFt3 At doses similar to that used here TGF/3 has been reported to cause a 2-3-fold increase in procollagen production. However, there have been no studies examining its effect on intracellular procollagen degradation. In this study we report a 1125 increase in synthesis and an 825 increase in the rate of intracellular procollagen degradation. Thus, there was a decrease in the proportion of procollagen degraded. Without this decrease in the proportion of procollagen degraded, the TGF,&induced increase in net procollagen produced would have been about 1750 pmol compared with the 2050 pmol increase observed (i.e., if the proportion of procollagen degraded had remained at 385 rather than decreasing to 325), a difference which represents about 205 of the production in control cultures. This reduction in the proportion of procollagen degraded, although small in percentage terms, clearly has the potential to play an important role in the increased net production of procollagen induced by TGF-~. TGF/3 has been shown to stimulate endogenous prostaglandin E2 (PGE2) production by fibroblasts [24]. PGE2 is known to inhibit collagen synthesis and increase the proportion of newly synthesised procollagen degraded. If prostaglandins were produced by fibroblasts in the present studies this could have limited the effects of TGF/3 on collagen degradation observed here. This could be tested by adding inhibitors of PGE synthesis such as indomethacin to the cultures. The
studies of Diaz et al. [24] have already demonstrated that TGFfl-induced increases in collagen production are further enhanced by blocking PGE2 synthesis. The mechanism by which TGFfl reduces the proportion of procollagen degraded intracellularly is unclear, but several possible explanations exist. First, the proportion of procollagen degraded can be affected by the degree of proline hydroxylation. If procollagen is underhydroxylated it does not form a stable triple helix and is more susceptible to degradation. Therefore, if TGF~ increased hydroxylation of proline it may lead to a more stable molecule which would be less susceptible to degradation. However, the evidence available at present suggests that TGFfl does not affect prolyl hydroxylation [4] and is therefore unlikely to be a contributor to the changes in degradation observed here. Another possible mechanism by which TGFfl could inhibit the proportion of newly synthesised procollagen degraded is by inhibiting the production of ,ysosomal acid proteinases which are thought to be involved in degradation of procollagen intracellularly. There is no evidence for TGF~ affecting these particular enzymes at present, but it is known to be capable of decreasing production of other proteinases, such as collagenase [9,10] and elastase [25], which are capable of degrading collagen extracellularly. TGF~ also stimulates production of the anticollagenases, TIMP and ,~2-macroglobulin [9-11]. In summary, we have demonstrated that procollagen synthesis and degradation can be estimated directly in small cultures without the use of radiolabelled compounds. The use of these methods eliminate assumptions relating to precursor pool specific activities involved with incorportion of radioisotopic tracers. Addition of TGF/] to these fibroblast cultures increased rates of synthesis and degradation of procollagen with a decrease in the proportion degraded. This decrease could play a significant role in the TGF~-induced increase in collagen production.
Acknowledgements This work was supported in part by the Clinical Research Committee of the National Heart and Chest Hospitals. References 1 lgnotz, R.A. and Massague, J. (1986) J. Biol. Chem. 261, 43374345. 2 Roberts, A.B., Sporn, M.B., Assoian, R.K., Smith, J.M., Roche, N.S., Wakefield, L.M., Heine, U.I., Liotta, L.A., Falanga, V., Kehrl, J.H. and Fau¢i, A.S. (1986) Proc. Natl. Acad. Sci. USA 83, 4167-4171. 3 Varga, J. and Jimencz, S.A. (1986) Biochem. Biophys. Res. Commun. 138, 974-980. 4 Fine, A. and Ooldstein, H. (1987) J. Biol. Chem. 262, 3897-3902.
235 5 Raghow, R., Postlethwaite, A.E., Keski-Oja, J., Moses, H.L. and Kang, A.H. (1987) J. Clin. Invest. 79, 1285-1288. 6 Ignotz, R.A., Endo, T. and Massague, J. (1987) J. Biol. Chem. 262, 6443-6446. 7 Varga, J., Rosenbloom, J. and Jimenez, S.A. (1987) Biochem. J. 247, 597-604. 8 Pentinnen, R.P., Kobayashi, S. and Bornstein, P. (1988) Proc. Natl. Acad. Sci. USA 85, 1105-1108. 9 Edwards, D.R., Murphy, G., Reynolds, J.J., Whitman, S.E., Docherty, A.J.P., Angel, P. and Heath, J.K. (1987) EMBO J. 6, 1899-1904. 10 Overall, C.M., Wrana, J.L. and Sodek, J. (1989) J. Biol. Chem. 264, 1860-1869. 11 Shi, D.L., Savona, C., Cagnon, J., Cachet, C., Chambaz, E.M. and Feige, J.J. (1990) J. Biol. Chem. 265, 2881-2887. 12 Campa, J.S., McAnulty, R.J. and Laurent, G.J. (1990) Anal. Biochem. 186, 257-263. 13 Burton, K. (1956) Biochem. J. 62, 315-322. 14 Hildebran, J.N., Airhart, J., Stirewalt, W.S. and Low, R.B. (19Sl) Biochem. J, 198, 249-258. 15 Clark, J.G. and Hildebran, J.N. (1984) Anal. Biochem. 140, 478485.
16 Bellon, G., Monboisse, J., Randoux, A. and Borel, J. (I987) Biochim. Biophys. Acta 930, 39-47. 17 Trelstad, R.L., Lawley, K.R. and Holmes, L.B. (1981) Nature 289, 310-312. 18 Steinmann, B., Rao, V.H. and Gitzelman, R. (198,,1) FEBS Lett. 133, 142-144. 19 Imberman, M., Oppenheim, F. aild Franzblau, C. (1982) Biochim. Biophys. Acta 719, 480-487. 20 Bienkowski, R.S. (1984) J. Cell. Physiol. 121, 152-158. 21 Bienkowski, R.S., Cowan, M.J., McDonald, J.A. and Crystal, R.G. (1978) J. Biol. Chem. 253, 4356-4363. 22 Bienkowski, R.S., Baum, B.J. and C,"jstal, R.G. (1978) Nature 276, 413-416. 23 Rennard, S.I., Stier, L.E. and Crystal, R.G. (1982) J. Invest. Dermatol. 79 suppl. 1, 77s-82s. 24 Diaz, A., Varga, J. and Jimenez, S.A. (1989) J. Biol. Chem. 264, 11554-11557. 25 Redini, F., Lafuma, C., Pujol, J., Robert, L. and Hornebeck, W. (1988) Biochem. Biophys. Res. Commun. 155, 786-793.
