0022-202X/80/7505-0404$02.00/0 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 75:404-407, 1980 Vol. 75, No. 5 Copyright © 1980 by The Williams & Wilkins Co. Printed in U.S.A. Prolyl and Lysyl Hydroxylase Activities of Human Skin Fibroblasts: Effect of Donor Age and Ascorbate

SAOOD MuRAD, PH.D., ARUNTHATHY SIVARAJAH, B.Sc., AND SHELDON R. PINNELL, M.D. Division of Dermatology, Department of Medicine, Dulle University Medical Center, Durham, North Carolina, U.S.A.

Prolyl and lysyl hydroxylase activities in cultures of MATERIALS AND METHODS human skin fibroblasts from fetal to 94-yr-old donors Fetal human skin fibroblasts were obtained from the American Type were measured. In contrast to earlier studies with whole Culture Collection, Rockville, Maryland. Fibroblasts from donors of all skin, neither prolyl nor lysyl hydroxylase activity was other ages were obtained from the Human Genetic Mutant Cell Re­ found related to donor age. Prolyl hydroxylase activity pository, Camden, New Jersey. They were split at 1:3 ratio in Dul­ increased 3- to 6-fold when cell extracts were incubated becco's modified Eagle's medium (Grand Island Biological Company, with ascorbate and other hydroxylation cofactors before Grand Island, New York) supplemented with 10% heat inactivated assay. A similar increase in prolyl hydroxylase activity fetal calf serum (Irvine Scientific, Santa Ana, California) and main­ occurred when cells were incubated with ascorbate. Ly­ tained at 37°C in a humid atmosphere of 5% carbon diox.ide/95% air. On the 7th day the cells were harvested with 0.5% trypsin/0.54 syl hydroxylase activity remained unaltered under these M[ethylenedinitrilo]-tetraacetate and counted in a hemocytometer. conditions. The cell suspension was centrifuged at 500 Xg for 5 min and the pellet was washed 4 times with Hanks' balanced salt solution (without cal­ cium, magnesium, bicarbonate, and glucose), each wash step being Collagen is the major extracellular protein of connective followed by centrifugation. The pellet was then suspended in a medium tissues and its metabolism has been the subject of many age (1 ml per 5 x 106 cells) containing 0.2 M NaCI, 0.1 M glycine, 50 ILM related studies [reviewed in reference 1]. The biosynthesis of dithiothreitol, 20 mM Tris-(hydroxymethyl) aminomethane (pH 7.5), collagen is of particular interest in this regard as it involves, in and 0.1% Triton X-100. The cells were subjected 3 times to quick addition to the polypeptide synthesis, certain intra- and extra­ freezing and thawing. The lysate was centrifuged at 27,000 x g for 1 hr cellular modifications that are essential for collagen maturation and the supernatant was used for enzyme assays. [2]. Two of the intracellular modifications are hydroxylation of Prolyl and lysyl hydroxylase activities were measured by methods basfd on tritium release [16,17]. The substrate for prolyl hydroxylase prolyl and lysyl residues, giving rise to important reaction war, unhydroxylated procollagen containing [4- 3H]-L-proline and for products. Hydroxyproline is believed to stabilize the collagen lyf.yl hydroxylase was that containing [4,5-3H]-L-lysine [18]. The reac­ triple helix [3] and hydroxylysine is the site for glycosylation tion mixture (1.5 ml) contained either 127,000 cpm of prolyl substrate and intermolecular crosslinking in collagen [2]. The formation or 280,000 cpm of lysyl substrate, 50 mM tris-(hydroxymethyl)-amino­ of hydroxyproline and hydroxylysine in procollagen is catalyzed methane (pH 7.8), 0.5 mM a-ketoglutaric acid, 2 mM sodium ascorbate, by separate enzymes, prolyl [EC 1.14.11.2] and lysyl [EC 0.1 mM dithiothreitol, 0.1 mg per ml catalase, 1.5 mg per ml bovine 1.14.11.4] hydroxylases, which share the same cofactors [3]. serum albumin, and cell extract (approximately 100 ILg of protein for Prolyl hydroxylase occurs in inactive (monomeric) and active prolyl hydroxylase and 200 ILg of protein for lysyl hydroxylase). Incu­ (tetrameric) forins and is considered to play a key regulatory bation time was 1 hr for prolyl hydroxylase and 2 hr for lysyl hydrox­ ylase. Incubation temperature was 30°C. The reactions were terminated role in collagen biosynthesis [1,3]. by addition of 0.1 ml of 50% trichloroacetic acid and the tritiated water Age dependent changes in prolyl and lysyl hydroxylase activ­ formed was vacuum distilled in a custom made multisample apparatus. ities of human tissue have been reported [ 4-6]. Both enzyme One milliliter of the distillate was mixed with 10 ml of Aquasol-2 (New activities were high in fetal skin and rapidly declined with age England Nuclear Corporation, Boston, Massachusetts) and its radio­ until maturity. However, it is not clear from these studies if the activity was determined in an Intertechnique liquid scintillation spec­ aging effect reflects a decrease in the number of fibroblasts or trometer (efficiency 36%). Under these conditions assays without en­ in the enzyme activity per fibroblast, the cell mainly responsible zyme gave 50 to 100 cpm; the values with enzyme were at least 10 times for collagen production. The interpretation of these results is higher. Three or more cultures were pooled for measurement of enzyme further complicated by technical difficulties encountered in activities. Values from duplicate assays on the same cell extract agreed within 2%. The rate of reaction was found linearly related to the amount handling skin specimen for enzyme studies. These considera­ of cell extract up to approximately 400 ILg of protein for prolyl hydrox­ tions prompted us to measure prolyl and lysyl hydroxylase ylase and 300 }lg of protein for lysyl hydroxylase. activities in human skin fibroblasts from donors of various ages. Enzyme activation was performed either by incubating the cells with In the course of this study it became necessary to explore ascorbate or by preincubating the cell extracts with complete reaction possible activation of prolyl hydroxylase by ascorbate admin­ mixture minus substrate. The stimulated prolyl hydroxylase activity istration. Such an activation has been observed in some types was measured with half the usual amount of cell extract (approximately of cells [7-10] but not in others [11,12], and for one cell type 50 ILg of protein) in order to ensure that the activity fell on the linear different assay methods have given conflicting results [7,13, portion of the rate versus enzyme concentration curve. 14]. Moreover, in human skin fibroblasts ascorbate has been Protein was assayed using Bio-Rad kit with bovine gamma globulin found to depress prolyl hydroxylase activity [15). The present as the standard [19]. report provides evidence for preferential stimulation of prolyl RESULTS hydroxylase activity by ascorbate in cultured human skin fibro­ blasts. In human -skin fibroblasts prolyl hydroxylase activity was stimulated by ascorbate in a time dependent manner, reaching Manuscript received February 26, 1980; accepted for publication a maximum by hour 3 at which time the ascorbate-treated cells May 27 , 1980 . . had 3 times the activity of the untreated cells (Fig 1). As This work was supported by grants AM-17123 and AM-07093 from expected, lysyl hydroxylase activity was not affected by ascor­ the National Institutes of Health and from the Howard Hughes Medical Institute Laboratories of Duke University Medical Center. bate treatment. The observed stimulation ofprolyl hydroxylase This is publication number 56 from the Dermatological Research activity by ascorbate is in apparent conflict with the reported Laboratories of Duke University Medical Center. depression of this activity in human skin fibroblasts [15]. How­ Reprint requests to: Sheldon R. Pinnell, M.D., Box 3135, Duke ever, in this study the cells were cultured with ascorbate; the University Medical Center, Durham, NC 27710. long-term effect of ascorbate on prolyl hydroxylase activity may

404 Nov. 1980 ASCORBATE AND AGE EFFECTS ON COLLAGEN HYDROXYLASES 405

75 ,..,

c:: b PH >< 0 ElO -~ a. 0.. u C7' ::1.. 'E5o 0.. u

""0 ~ E Q.) ~ -0 ~ 3: ~ ""0 "025 Q.) 3: -0 ""0 -~ I-- LH -0 LH ~ ...... a~--~1----~2~--~3----~ Preincubation Time (hr) 1 2 3 4 FIG 2. Time course of activation of prolyl and lysyl hydroxy lases in Ascorbate Treatment ( hr) cell extracts. Extracts of human skin fibroblasts.from a 15-yr-old male Duration of were incubated with complete reaction mixture minus substrate for varying times at the indicated temperatures. Substrates were then FIG 1. Time course of activation of prolyl and lysyl hydroxy lases in added and incubations were continued at 30°C for additional! and 2 hr intact cells. Confluent cultures of human skin fibroblasts from a 15-yr­ for prolyl and lysyl hydroxylases, respectively. Other details are given old male were incubated for varying times with serum containing in Materials and Methods section. medium supplemented with 0.25 mM sodium ascorbate. A set of cultures was not exposed to ascorbate and this is referred to as "zero time." TABLE I. Cofactor requirement for activation ofprolyl hydroxylase Cells were then harvested and prolyl and lysyl hydroxylase activities in in cell extracts their extracts were measured as described in Materials and Methods Prolyl section. hydroxylase Additions during preincubation activity be different from the short-term effect observed in the present % study. None 100 Prolyl hydroxylase activity was also stimulated in cell ex­ Complete system" 361 (334)b tracts when these were preincubated with ascorbate and other Complete system minus ascorbate 86 hydroxylation cofactors. The time course of this activation is Complete system minus a-ketoglutaric acid 71 shown in Fig 2. The activation was complete in 2 hr at 37°C Complete system minus ferrous sulfate 445 (24l)b and nearly so in 4 hr at 30°C. As in intact cells, lysyl hydroxylase Complete system minus catalase 226 activity remained unchanged in cell extracts when these were Complete system minus bovine serum albumin 327 similarly preincubated with hydroxylation cofactors. Full acti­ Complete system minus dithiothreitol 403 vation of prolyl hydroxylase in cell extracts required all hy­ Extracts of human skin fibroblasts from a 15-yr-old male were droxylation cofactors except dithiothreitol and ferrous ion incubated for 4 hr at 30°C with hydroxylation cofactors as indicated. which were somewhat inhibitory (Table I) . The activation was Omitted component(s) and prolyl substrate were then added and in­ details are given in absolutely dependent on ascorbate and a-ketoglutarate and cubations were continued for additional! hr. Other Methods section. on catalase and bovine serum albumin, the former Materials and partially " Consisted of hydroxylation cofactors including 5 JiM ferrous sulfate. exerting a greater influence. Ferrous ion was needed for acti­ b Obtained with preparation which had been dialyzed against the vation in dialyzed extracts. However, considerable activation extraction medium. occurred in its absence, perhaps due to endogenous iron which .may have remained bound to the enzyme. The observed acti­ vation of prolyl hydroxylase in extracts of human skin fibro­ hydroxylase exhibited some variation in activity for reasons blasts is in agreement with the reported activation in sonicates that are not entirely clear. Several causes are suspected, chief of mouse skin fibroblasts [20]. This study also suggested that among them is the difference in growth characteristics between the activation in intact cells occurs by the same mechanism. In various strains, leading to variable final cell density and conse­ subsequent experiments therefore prolyl hydroxylase activation quently variable prolyl hydroxylase activity [21]. only. was carried out using cell extracts DISCUSSION Table II shows the levels of prolyl and lysyl hydroxylase activities in cultures of human skin fibroblasts from donors of A stimulation of prolyl hydroxylase activity of human skin various ages (fetal to 94 yr). Prolyl hydroxylase activity in cell fibroblasts was observed in intact cells as well as in cell extracts; extracts which had been preincubated with cofactors was mark­ the stimulation in intact cells was dependent on ascorbate while edly higher than in the controls. Unexpectedly, neither prolyl that in cell extracts was dependent on ascorbate, a-ketogluta­ nor lysyl hydroxylase activity depended on donor age. Prolyl rate, and ferrous ion. Lysyl hydroxylase activity did not change 406 MURAD, SIVARAJAH, AND PINNELL Vol. 75, No.5

TABLE II. Effect of donor age on fibroblast prolyl and lysyl behave differently in response to ascorbate treatment, such that hydroxylase activities stimulation ofprolyl hydroxylation [26] and activation ofprolyl Prolyl hydroxylase activity" Lysyl hydroxylase [9] occurs only in the young cells. Whatever may Age hydroxylase be the explanation, the studies described here would make it Unstimulated Stimulated activity" unnecessary to control donor age in experiments utilizing hu­ Fetal 29.5 160.5 4.9 man skin fibroblasts, at least as far as the 2 hydroxylases in 3 Days 36.1 101.8 6.2 collagen biosynthesis are concerned. 3Mo 23.6 90.2 6.3 10 Mo 21.7 82.9 7.3 71.8 5 Yr 28.5 7.1 REFERENCES 10 Yr 28.2 124.9 5.7 15 Yr 17.8 67.5 4.9 1. Kivirikko KI, Risteli L: Biosynthesis of collagen and its alterations 24 Yr 37.5 201.8 9.0 in pathological states. Med Bioi 54:159-186, 1976 40 Yr 28.4 169.8 7.4 2. Pinnell SR: Disorders of collagen, The Metabolic Basis of Inherited 64 Yr 38.4 210.8 6.2 Disease, 4th edition. Edited by JB Stanbury, JB Wyngaarden 94 Yr 31.7 101.1 8.3 and DS Fredrickson. McGraw-Hill, New York, 1978, pp 1366- 1394 Prolyl and lysyl hydroxylase activities were measured in extracts of 3. Prockop DJ, Berg RA, Kivirikko Kl, Uitto J: Intracellular steps in human skin fibroblasts from donors of various ages. Prolyl hydroxylase the biosynthesis of collagen, Biochemistry of Collagen. Edited by activity was also measured in cell extracts which had been preincubated GN Ramachandran and AH Reddi. Plenum Press, New York, with complete reaction mixture minus substrate for 2 hr at 37°C and 1976, pp 163-273 this is referred to as "stimulated activity." Details are given in Materials 4. Uitto J, Halme J, Hannuksela M, Peltokallio P, Kivirikko KI: hydroxylase activity in the skin of normal Cells were in passage 8 to 17. Protocollagen proline and Methods section. human subjects and of patients with scleroderma. Scan J Clin " Expressed as tritiated water cpm per p.g of protein. Lab Invest 23:241-247, 1969 5. Tuderman L, Kivirikko KI: Immunoreactive prolyl hydroxylase in human skin, serum and synovial fluid: changes in the content and under these conditions. These observations are consistent with components with age. Eur J Clin Invest 7:295-300, 1977 the postulation that prolyl hydroxylase occurs as an activatable 6. Anttinen H, Orva S, Ryhanen L, Kivirikko KI: Assay of protocol­ lagen lysyl hydroxylase activity in the skin of human subjects complex with unhydroxylated procollagen [20]. In the presence and changes in the activity with age. Clin Chern Acta 47:289-294, of ascorbate and other hydroxylation cofactors this complex is 1973 dissociated into free enzyme and hydroxylated pro collagen. The 7. Stassen FLH, Cardinale GH, Udenfriend S: Activation of prolyl failure of ascorbate to stimulate lysyl hydroxylase activity in hydroxylase in L-929 fibroblasts by ascorbic acid. Proc Nat! Acad here and in previous reports Sci 70:1090-1093, 1973 fibroblasts as demonstrated 8. Levene CI, Aleo JJ, Prynne CJ, Bates CJ: The activation of [9,15,22] may be taken to suggest that lysyl hydroxylase, unlike protocollagen proline hydroxylase by ascorbic acid in cultured prolyl hydroxylase, does not occur in an activatable form. The 3T6 fibroblasts. Biochim Biophys Acta 338:29-36, 1974 earlier consideration that ascorbate activates prolyl hydroxy­ 9. Chen KH, Evans CA, Gallop PM: Prolyl and lysyl hydroxylase WI-38 subunit aggregation of the enzyme [7,23] can activation and cofactor specificity in young and senescent lase by promoting fibroblast cultures. Biochem Biophys Res Commun 74:1631-1636, be disregarded in view of the finding that the activation in cell 1977 extracts was accompanied by little change in the amount of the 10. Evans CA, Peterkofsky BJ: Ascorbate-independent proline hydrox­ active tetramer [20]. ylation resulting from viral transformation of Balb 3T3 cells and l 89:355- An alternative explanation is that ascorbate deficiency leads unaffected by dibutyryl cAMP treatment. J Cell Physio 368, 1976 to accumulation of hydroxyproline-deficient procollagen [24,25] 11. Kao WW-Y, Berg RA, Prockop DJ: Ascorbate increases the syn­ which competes with the radioactive substrate in the assay, thesis of procollagen hydroxyproline by cultured fibroblasts from thus underestimating the prolyl hydroxylase activity. Treat­ chick embryo tendons without activation of prolyl hydroxylase. 1975 ment of cells with ascorbate or treatment of cell extracts with Biochim Biophys Acta 411:202-215, 12. Blank TJJ, Peterkofsky B: The stimulation of collagen secretion ascorbate and other cofactors results in hydroxylation of the by ascorbate as a result of increased proline hydroxylation in endogenous substrate which can no longer compete with the chick embryo fibroblasts. Arch Biochem Biophys 171:259-267, exogenous substrate. The endogenous substrate that accumu­ 1975 hydroxylase activity in in ascorbate deficiency is not deficient in hydroxylysine 13. Berg RA, Kao WW-Y, Prockop DJ: Prolyl lates L-929 fibroblasts incubated with and without ascorbate. Biochim [25] and therefore would not be a problem in lysyl hydroxylase Biophys Acta 444:756-764, 1976 assay. Although this explanation is generally favored, a recent 14. Kuttan R, Parrott DP, Kaplan SR, Fuller GC: Effect of ascorbic report [14] claimed similar results with c·c]-a-ketoglutarate acid on prolyl hydroxylase activity, collagen hydroxylation and Res Com­ and tritium release assays, making it unlikely collagen synthesis in human synovial cells in culture. decarboxylation mun Pathol Pharmacol 26:337-345, 1979 that the activation phenomenon is due only to substrate dilu­ 15. Quinn RS, Krane SM: Abnormal properties of collagerl lysyl hy­ tion. droxylase from skin fibroblasts of siblings with hyd.roxylysine­ The observation that human skin fibroblasts from donors of deficient collagen. J Clin Invest 57:83-93, 1976 ages had similar prolyl as well as lysyl hydroxylase 16. Hutton JJ Jr, Tappe! AL, Ud.enfriend S: A rapid assay for collagen various proline hydroxylase. Anal Biochem 16:384-394, 1966 activities differs from the known influence of age on these 17. Miller, RL: Rapid assay for lysyl-protocollagen hydroxylase activ­ enzyme activities in whole skin. In adult skin prolyl hydroxylase ity. Anal Biochem 45:202-210, 1972 activity was 10% of that in fetal skin [ 4] and lysyl hydroxylase 18. Murray JC, Lindberg KA, Pinnell SR: The in vitro inhibition of activity was 2.5% [6]. Furthermore, the reduction in prolyl chick embryo lysyl hydroxylase by homogentisic acid. J Clin Invest 59:1071-1079, 1977 hydroxylase activity with aging was accompanied by a decrease 19. Bio-Rad Laboratories: Bio-Rad protein assay. Technical Bulletin in the ratio of the active enzyme to the total immunoreactive 1051, 1977 protein [5]. . 20. Kuttan R, .Cardinale GJ, Udenfriend S: An activatable form of The present study clarifies the earlier observation in that the prolyl hydroxylase in fibroblast extracts. Biochem Biophys Res Commun 64: 947-954, 1975 decline in prolyl and lysyl hydroxylase activities in skin samples 21. Coinstock JP, Gribble TJ, Udenfriend S: Further study on the with aging may be attributed to a decrease in the number of activation of collagen proline hydroxylase in cultures of L-929 fibroblasts. On the other hand, it may be argued that conditions fibroblasts. Arch Biochem Biophys 137:115-121, 1970 of cell culture have modified the phenotypic expression of 22. Miller RM: The effect of ascorbic acid on lysyl and prolyl hydrox­ Biochem Biophys 170: lost the regulatory mecha­ ylase activity of cultured fibroblasts. Arch fibroblasts and that such cells have 341-344, 1975 nisms that control their activity in the tissue. It is noteworthy 23. McGee JO'D, Udenfriend S: Partial purification and characteriza­ in this regard that early and late passage human fibroblasts tion of oeotidyl proline hydroxylase precursor from mouse fibro- Nov.l980 ASCORBATE AND AGE EFFECTS ON COLLAGEN HYDROXYLASES 407

blasts. Arch Biochem Biophys 152:216-221, 1972 25. Quinn RS, Krane SM: Collagen synthesized by cultured skin fibro­ 24. Bates CJ, Prynne CJ, Levene CI: Ascorbate dependent differences blasts from siblings with hydroxylysi:ne-deficient collagen. in the hydroxylation of proline and lysine in collagen synthesized Biochim Biophys Acta 585:589-598, 1979 by 3T6 fibroblasts in culture. Biochim Biophys Acta 278:610-616, 26. Paz MA, Gallop PM: Collagen synthesized and modified by aging 1972. fibroblasts in culture. In Vitro 11:302-312, 1975

Preliminary Announcement

The Stratum Corneum-An International Symposium

An International Symposium on the stratum corneum will be held at the Welsh National School of Medicine, Heath Park, Cardiff, on October 29, 30, and 31, 1981. There will be invited lectures, free communications and poster papers. Abstracts should be submitted by June 26, 1981, in the style as submitted to the European Society for Dermatological Research. We believe this symposium will be of importance for dermatologists, toxicologists, cosmetic sc'ientists, the pharmaceutical industry and others interested in skin biology. Organizing committee: Dr. R. Marks Dr. G. Plewig Department of Medicine Dermatologische U niversitatsklinik Welsh National School of Medicine FrauenlobstraBe 9-11 Heath Park, Cardiff, CF4 4XN. D-8000 Munchen 2. Further information can be obtained by contacting Dr. R. Marks.