Proline Analogues Inhibit Human Skin Fibroblast Growth and Collagen Production in Culture*
0022-202X/83/8004-0261$02.00/ 0 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 80:261-267, 1983 Vol. 80, No.4 Copyright © 1983 by The Williams & Wilkins Co. Printed in U.S.A. Proline Analogues Inhibit Human Skin Fibroblast Growth and Collagen Production in Culture*
ELAINE M . L. TAN, M.Sc., LASSE RYHANEN, M.D., PH.D.,·j AND JOUNI UITTO, M.D., PH.D.:j:
Division of Derm.alology, Department of Medicine, UCLA School of Medicine, Harbor-UCLA Medical Cen.te1 ~ Torrance, California, U.S.A.
Several structural analogues of proline have been carboxylic acid was noted with scleroderma cells as with shown to be incorporated into proteins in place of pro control fibroblast cultures. The results suggest, there line. As a consequence, the proliferation of cells in cul fore, that proline analogues may, in the future, prove ture and the extracellular deposition of collagen in ani useful in limiting excessive collagen deposition in scle mal systems are reduced. In this study, the effects of two roderma and other forms of dermal fibrosis. proline analogues, cis-4-hydroxy-L-proline and L-azeti dine-2-carboxylic acid, on the growth parameters and procollagen production by cultured normal human skin Several structmal analogues of proline, e.g., cis-4-hydroxy-L fibroblasts were examined. The results indicated that proline (cis-hydroxyproline), L-azetidine-2-carboxylic acid (aze incubation of the cells with the analogues reduced the tidine carboxylic acid), 3,4-dehydroproline, cis-4-bromo-L-pro rate of fibt·oblast proliferation and lowered the plating line, and cis -4-fluoro-L-proline, have been shown previously to efficiency. Further experiments demonstrated that fibro be incorporated into proteins in place of proline (see [2]). The blasts in the presence of L-azetidine-2-carboxylic acid incorporation of these analogues into polypeptides can dramat synthesized procollagen polypeptides which were not in ically alter the structme and metabolism of the protein. For a triple-helical conformation, as judged by limited pepsin example, incorporation of the analogues into collagen, a protein proteolysis. Also, a significantly increased fraction of relatively rich in imino acids, proline, and trans-4-hydroxy-L the newly synthesized collagenous peptides was in a proline (hydroxyproline), has been demonstrated to have sev dialyzable form, suggesting increased degradation of the eral consequences [2,3]. First, the newly synthesized polypep nonhelical chains. -The rate of translation of'collagenous tides of procollagen, pro-a chains, do not fold into a stable polypeptides and the prep~ocollagen messenger RNA triple-helical conformation if a critical number ofprolyl residues activity in the cells were not affected by the analogues. has been replaced by the analogue [ 4,5]. Since the triple-helical The proline analogues thus appear to inhibit the produc conformation is a prerequisite for normal secretion of procol tion of procollagen on the posttranslational level by lagen, the nonhelical pro-a chains are extruded from the cells preventing the polypeptides from folding into a stable into the extracellular m_ilieu at a markedly reduced rate [6]. triple-helical conformation. The nonhelical polypeptides The nonhehcal polypeptides are also susceptible to degradation are then readily susceptible to proteolysis leading to by intracellular as well as extracellular proteases which do not reduced deposition of extracellular collagen fibers. Sim degrade triple-helical collagen [ 4, 7]. Consequently, the net pro ilar experiments were also performed with fibroblasts duction of extracellular collagen is decreased in the presence of cultured from patients with active progressive systemic the analogues. Secondly, previous studies have demonstrated sclerosis. Quantitatively and qualitatively comparable that incubation of 17-day-old chick embryo tendon fibroblasts inhibition of procollagen production by L-azetidine-2- in primary culture with cis-hydroxyproline results in an inhi bition of growth of these cells [8]. Subsequently, it was also shown that this analogue retards the spreading and proliferation Manuscript received April 26, 1982; accepted for publication Septem- of normal murine connective tissue cells, but not of their ber 21, 1982. · counterparts rendered tumorigenic by viral, chemical, OI" spon This work was supported by U. S. Public Health Service, National taneous transformation [9,10]. On the basis of the latter obser Institutes of Health Grants AM-28450 and GM-28833, and by a grant vations, it was suggested that the inhibition of the growth of from March of Dimes-the Birth Defects Foundation. • A preliminary repo rt of part of this study was presented at the 39th normal cells by the analogues could be explained by the defi cient production of extracellular collagen substrate which may Annual Meeting of the American Academy of Dermatology, New York, December 1980 [1]. be required to support cell spreading and growth. -[Dr. Ryhiinen is the recipient of New Investigator Research Awa1·d In the present study, we have examined the effects of 2 AG-03172 from the National Institutes of Health. proline analogues, cis-hydroxyproline and azetidine carboxylic 4: Dr. Uitto is the rec ipient of Research Career Development Award acid, on growth pru·ameters and collagen production by human 5-K04-AM-0897 from the National Institutes of Health. Dr. Uitto was skin fibroblasts in cultme. also the winner of the Nelson Paul Anderson Memorial Essay Contest at the 33rd Annual Meeting of the Pacific Dennatologic Assoc iation, H.eno, Nevada, October 1981, with a presentation based on this study. MATERIALS AND METHODS Reprint requests to: Dr. E. M. L. Tan, Division of Dermatology, Human Shin Fibroblast Cultures Harbor-UCLA Medical Ce nter, Torrance, Ca li fornia 90509. Abbreviations: Control ce ll cultmes were initiated from punch biopsies or from skin azetid in e carboxylic ac id : L-azetidine-2-carboxyLic acid obtained from surgical procedures after obtaining informed consent. cis-hydroxyproline: cis-4-hyclroxy-L-proline Ce lls were maintained in Dulbecco's modified Eagle's medium DMEM: Dulbecco's modified Eagle's medium (DMEM) co ntaining glu tamine (KC Biologicals) supplemented with 30 Hepes: N- 2-hyru·oxyethylpiperazine-N'-2-ethenesulfonic acid mM Hepes buffer, pH 7.6, 200 U / ml of penicillin, 200 f.l.g/ ml of strepto hydroxyproline: trans-4-hydroxy-L-proline mycin, and 20% fetal calf serum. Confluent primary cultures were 2 mRNA: messenger RNA trypsinized and subcultured in 75-cm fl asks (Falcon Plastics) at 37°C Na"EDTA: disod iu m ethylenediaminetetraacetate [U.]. The ce lls were studied in passages 3-6. NEM: N-ethylma leimide Fibroblast cultures were initiated also from patients with active PMSF: phenylmethylsulfonylfluoride scleroderma. Clinically, these patients demonstrated acrosclerosis, pro SDS: sodium dodecyl sulfate gressive induration of the skin, and involvement of the lungs, esophagus, TCA: trich loroacetic ac id and kidneys; histologically, the skin was characterized by the accumu tRNA: transfer RNA lation of excessive co llagen in the lower dermis and in the subcutaneous 261 262 TAN, RYHAN E N , A ND U JTTO Vol. 80, No. 4 tissue; laboratory abnormalities incl uded positive ANA and elevated label, t he radioactive values were corrected by changes in the s pecific serum lgG. T hus, on the basis of cl inical, histopathologic, and labora radioactivity of ["H)prolyl-tRNA, as described below. tory observations, these patients were diagnosed as having active T o test the heli cal stability of the newly synthesized ["H]procollagen, progressive systemic sclerosis [12]. T he fibroblast cul tures from these ["H]proline-labeled proteins, synthesized in the presence of varying patients were initiated from punch biopsies obtained from the edge of concentrations of the analogue, were subjected Lo limited pepsin pro an actively spreading, indurated area of the forearm. T he cells were teolysis. T he cell plus mediu m f ractions were homogenized, dialyzed cultured and maintained the same way as t he control cul tures. against 0.5 M acetic acid, and pepsin (Sigma, 2 X crystalized), 330 Jig per ml, was added; the specimens were digested for 6 h at 24 oc Assays for Cell Growth [19,20]. T he digestions were stopped by adjusting the pH to 8.0 by the addition of NaOH. T he samples were then heated at 100°C fo r 5 min To study the proli feration of fibro bl asts in cul ture, single cell sus in the presence of 2% SDS, 20 mM Na,EDTA, 10 mM NEM, and 1 mM pensions were produced by brief trypsinization of monolayer cultures. P MSF. AliquoLs of the SDS-treated "H-labeled proteins were chromat Trypsin was removed by extensive washing with m ediu m containing ographed on 6% SDS agarose (Bio-Ge 2 l A-5M, 200-400 mesh, B io-Rad), 20% fetal calf serum. T he cells were plated on 25-cm plastic flasks, 5 equilibrated and elu ted with 0.1 % SDS, 0. 1 M sodium phosphate, pH X 10'' cells per flask. The cells were all owed to attach fo r 4 h ; the 7.4, as described previously [21,22]. Aliquots of the fractions were analogues in varying concentrations were then added to the medium. hydrolyzed in 6 M HCl for 24 h aL ll0°C, and ["H]hydroxyproline was T he medium, wit h or without t he a nalogue, was changed daily, and the assayed [14]. cell number in parallel fl asks was determined by the hemocytometer at time points indicated in Results. In the same cultures, the a mount of Assay of {'HJProlyl-tRNA Sp ecific Radioactivity cell protein was determined as an index refl ecting the cell number [13). For assay of the specific radioactivity of prolyl-tRNA, the cell layer 4 of the fibroblast cul tures, after incubation To study the effect of varying analogues on cell attachment, 5 X 10 with 30 J.!C i of ["H) proline in the cells were plated on 25-cm" tlasks in a medium containing varying presence of varying concentrations of azetidine carboxylic acid, were scraped in to 4 ml concentrations of the a nalogue. After a 4- or 24-h incubation, the of 0.09 M NaOAc, pH 5. 1, containing 1% SDS [23]. N ucleic acids medium containing t he unattached cells was decanted, and the amount were extracted wi th 4 ml of a mixture of phenol:c hloroform:isoamy of attached cells was measured by hemocytometer or by determination lalcohol (2 :2:0.04) . T he phenol l ayer was of the cell protein, as above. reextracted with 4 ml of 0.09 M NaOAc (pH 5. 1) and 1% SDS. Nucleic acids were precipitated from the combi ned aqueous layers w ith 16 ml of ethanol at -20°C for 18 h. T he precipitate was recovered by centrif Assays for Procollagen Production ugation at 17,000 g for 30 min at -20°C and d issolved in 4 ml of 50 mM For assay of co llagen synthesis, conflu ent monolayer cultures of T ris-HCl, pH 6.0. T he concentration of total RNA was determined, fibroblasts were rinsed with DMEM and then placed in fresh m edium a fter incubation with 1 M NaOH for 18 h at 37°C, by an orcin ol method supplemented with 25J.!g/ml ascorbic acid, 20 J.!g/ml P-aminopropioni using d-ribose as a standard [24 ]. tril e- HCI, and 20% dialyzed fetal calf serum [11). After a 4- h preincu T he amount of ["H]proline that was bound to prolyl-tRNA was bation, the cell cultm es were labeled with ["H]proline or [ 3H)glycine determined by incubation of the isolated total nucleic acid preparation (30 J.!Ci/flask)' for 20 h at 37°C. At the end of the labeling period, the with 1 M NaOH for 60 min at 37°C. T he incubation was terminated b y mediu m was removed, cooled at 4°C, and protease inhibitors were the addition of 10% ice-cold TCA, and t he sampl es were collected on added to give the following fin al concentrations: 20 mM Na,E DTA, 10 M illipore (AP 1502500) glass- fi ber fil ters. Aliquots of the nucleic acid mM NEM, and 1 mM PMSF. T he cell layer was rinsed 3 times with preparations were simultaneously processed without deacylaL ion by serum-free medium at 4°C, scraped with a rubber policeman into 5 ml NaOH . T he radioactivity retained on the fi lters was determined using of solu tion containing 0.4 M NaCl, 50 mM Tris-HCl (pH 7.5), a nd t he a Beckman LS 7500 liquid scin tillation counter after extraction in to 10 same protease i.nhibitors; the cells were then sonicated at 50 Hz fo r 30 ml of liquid scintill ation fl uid (3a70B; Research Products International s. To quantitate the synthesis of ["H) hydroxyproline in cultures la beled Corp. ). T he counting efficiency of ''H was 36%. T he difference in the with ["H]proline, aliquots of medium and homogenized cell fractions radioactivity between the alkali -treated and nontreated preparations of were dialyzed against running tap water, hydrolyzed in 6 M H Cl in each sample was taken as a measure of the incorporation of ["H) sealed tubes at ll0°C for 24 h a nd assayed fqr nondialyzable ["H) proline in to prolyl-tRNA. hydroxyproline using a specifi c radiochemical method [1 4). F w·ther independent quantitation of radioactively labeled collagen, synthesized Assay of m.RNA Activity by Cell-free Translations when the cells were incubated either with [ "H)proli ne or [''H]glycine, To isolate RNA, the fi broblasts were in cubated with varying concen was performed using a bacterial collagenase assay [15]. Aliquots of the trations of Lhe a nalogue, as descri bed above, but w ithout radioactive medium and cell homogenates were dialyzed extensively against 0.15 a mino acids. The cell layers were scraped into 4 ml of 10 mM Tris- H Cl, M NaCl, 50 mM T ris-HCl (pH 7.5) containing 1 mM Na,EDT A. P urified pH 7.5, containing 5 mM Na"EDTA, 1% SDS, 50 ••g/ml of protein ase K bacterial collagenase (Advanced Biofactors), fi nal concentration of 50 (Boehringer) a nd homogenized with a g lass-glass tissue grinder J.!g / ml in 0.15 M NaCl-Tris buffer (pH 7.5) , 10 mM CaCl,, 25 mM NEM , (Dounce). The homogenate was in cubated fo r 60 min at 45 °C. N ucle.ic and 1 mM PMSF, were added. Samples were incubated for 6 h at 30°C. acids we re isolated by extraction with phenol:chloroform:isoamylal Digestion was halted with 0.1 vo l of 0.25 M Na2EDTA a nd the samples cohol mi xture, as described above. Nucleic acids we re precipitated with were dialyzed against 0.15 M NaCl-Tris buffer (pH 7.5) containing 25 the addition of 2 vol of cold (-20°C) ethanol in Lhe presence of 0.1 M mM NazEDTA, 25 mM NEM , and 1 mM P MSF, at 4°C. T he amount of NaCI. P roteins a nd t races of phenol were removed by multipl e wash collagenous proteins wa.s measured by the release of r adioactive pep in gs, first wi th 6 M g uanidin e and subsequently wi th 66% ethanol tides by dialysis [16). containing 0.1 M NaCl [25,26]. T he sample was dissolved in distilled In some experiments, the amount of dialyzable ["H]hydroxyproline H"O and the concentration of RNA was deLe nn in ed by assuming t hat present in cultures incubated with ["H)proline was also determined. In the optical density of a 1 mg/ml solution was 20 at 260 nm. these experiments, the cells were incubated with radioactive proline Aliquots of the RNA solution were employed fo r cell-free tra nslation from which the contaminating radioactive hydroxyproline was sepa of the mRNAs by in cubation with nuclease-treated rabbit reticul ocyte rated by chromatography on a Dowex 50-X ion excha nge resin. After lysate [27]. T he translations were carried out for l20 min at 26°C with the incubation with ["H]proli ne, the cell a nd medium fractions, follow 1 /LC i of ["H]proline in a total volume o f 11 JLI. AL the e nd of the ing the addition of Lhe protease inhib itors (see above), were dialyzed incu bati on, the samples were pl aced on i ce, and aliquots from each against lO vol of distill ed water. After a 48-h dialysis, the retentates incubate were taken for para ll el incubation with and w ithout bacteria l were further dialyzed against ru nni ng Lap water. Aliquots of the reLen collagenase in a f'in al vo lume of 100 J.!l, as described above. After the taLes (representing nondialyzable ["H]hyclroxyproli ne) and t he dialy initial in cubation, 500 J.! l of 1 M NaOH, 100 J.!l of 40 mM proline, and 50 sates (representing dialyzable ["H]hydroxyproline) were hydrolyzed in J.!l of 50% hydrogen perox id e were a dded, and Lhe samples were incu (j M HCl for 24 h at ll0°C, and [''H)hydroxyproline was determined on bated for a n additional 15 min at 37°C. T he latter incubation r esul ts in a Beckman ]]9 CL amino acid analyzer column elu ted with 0.3 M deacylation of the ["H]prolyl-tHNA complex and decolori zation of the citrate buffe r, pH 3.25. samples containing endogenous hemog lobin . T he radioactivity in cor T he values fo r coll agen production measured as ["H]hydroxyproline porated in to polypeptide transla ti on products was recovered by precip or ·'H-labeled collagenase-digestible peptides were expresse<:I as dpm itation with 10% TCA at 4°C. T he total mRNA activity was expressed radioactivity in cell plus medium fraction per JLg DNA [17) or m g cell as incorporation of ["H]proline in to T CA-precipitabl e translation prod protein [18). In some experiments in which ["H]proline was used as the ucts per Jig HNA. April1983 HUMAN SKIN FIBROBLASTS IN CULTURE 263
RESULTS of cis-hydroxyproline to produce comparable effects. Essen Cell Proliferation and Attachment tially similaT results were obtained when t he amount of cells was estimated by assaying the cellular protein content of the In initial studies, the effects of cis-hydJ·oxyproline a nd azeti cultures (Fig 2B). The mechanism of this inhibition was further dine cru·boxylic acid on cell proliferation were examined. In studied in sepru·ate experiments under similru· conditions, but these experiments, t he cells were allowed to attach to the plastic the number of cell s was determined at 3, 5, 7, and 9 days after surface before the addition of analogues at varying concentra the initiation of the treatment. The results showed that after 3 tions. In the first experiment (Fig 1) , the number of cells was days of treatment with 5 p.g/ ml of azetidine carboxylic t he cell determined by hemocytom eter. The results indicated that a 6- number was already significantly lower than in control cultures day incubation with 2, 20, or 200 p.g/ m1 of cis-hydroxyproline (Fig 3) . It should be noted, however, that the number of cells, significantly reduced the number of cells that were present, as even in the presence of azetidine cru·boxylic acid, continued to compar ed to controls (Fig 1A). Sin1ilar results were obtained increase up to 9 days of incubation, t hus reflecting a reduced with 10, 25, a nd 50 ~t g/ ml of azetidine carboxylic acid ; in cultures rate of proliferation (Fig 3). incubated with 100 p.g/ml of this a nalogue only a few cells In fmther experiments, the effects of t he a nalogues on cell rem ained (Figs 1B, 2A). It should be noted that the concentra attachment were studied. In these experiments, the cells were tions of azetidine carboxylic acid were approximately half those plated in a medium containing t he a nalogue, and the number of cells which attached to the plastic was determined at 4 a nd 24 h after plating. The results showed that in control cultures, essentially all cells had attached after 4 h (Fig 4A). However, 8 .0 in cultmes plated in the presence of 25 or 100 p.g per m1 of A azetidine carboxylic acid, the number of attached cells was less than h alf that of the controls. A l esser effect was noted with 1 .>< p.g/ml of the "'0 a nalogue. Similarly, a reduced number of cells was present in cultmes containing 25 or 100 p.g/ ml azetidine cru·box ylic acid when examined 24 h after the seeding (Fig 4B). The resul ts indicate that the analogues tested can reduce the rate of (f) fibroblast proliferation in vitro a nd this effect may be explained -' -' partly by lower plating efficiency. w u u. 0 Procollagen Production 0:: w cc The effects of one of the analogues, azetidine carboxylic acid, ~ ::::J on procollagen production by the cult ured fibroblasts were also z studied. In one set of the experiments, the cells in confluent monolayer cultures were preincubated in the presence of the analogue a nd ascorbic acid for 4 h, a nd then eHJproline was Control 2 20 200 added. At the e nd of a 20-h incubation, the cell and medium C I S·HYDROXYPROLINE fractions were separately assayed for eHJhyru·oxyproline, as a (,AJg/ml) marker of newly synthesized collagen; t he contents of the cellulru· protein and DNA were also determined. T he results indicate that the amount of newly y nt hesized CHJhyru·oxypro 8.0 t- line in nondialyzable peptides was significantly reduced in cul B tm es incubated with 25 p.g/ ml of azetidine carboxylic acid (Table I) ; the reduction was of the same order of magnitude .>< r+- r-+-- whether the values were expressed as dpm radioactivity per p.g .2"' 6.0 f- DNA or mg cell protein . ~ 3 b Since azetidine carboxylic acid can compete with [ H]proline in the formation of prolyl-tRNA complexes (see Discussion), it ~ was important to perform control experiments to examine t (f) he -' possibility that reduced CHJhydroxyproline synthesis is a result -' 4.0 w u of isotope dilution by t he a nalogue itself. In the first set of u. experiments, the total intracellulru· pool of RNA was isolated, 0 and the specific radioactivity of CHJprolyl-tRNA was deter 0:: w mined, as described in Materials and Methods. The results cc 2.0 f- ~ r+- indicated that the specific radioactivity of the [aH]prolyl-tRNA ::::Jz complex was markedly reduced in cells incubated with 5 and 25 p.g/ ml of azetidine carboxylic acid, as co m~ ru·ed to the control .----. cultmes (Ta ble II). If the incorporation of ' H-radioactivity into the collagenase-digestible polypeptides in the same Con tro l 1 I 0 I 00 experiment was corrected b y the changes in the s pecific radioactivity of L- AZETIDINE- 2 -CARBOXYLIC ACID ( }Jg / ml ) ["H]prolyl-tRNA, the differences between the controls a nd the azetidine cru·boxylic acid-treated cult m es were lru·gely a bolished FIG l. Inhibition of the proliferation of human skin fibroblasts in (Table II) . In fmther experiments, the cell cultures were labeled cul ture by cis-hydroxyproline and azetidine cru·boxylic acid. Cells, with ["H]glycine, an amino acid which is not displaced by plated on plastic fl asks, were a llowed to attach for 4 h before the azetidine carboxylic acid from the specific tRNA complex. The addition of the analogues in concentrations indicated. The medium results indicated that the formation of ["H]glycine-labeled col contai11ing the ana logue was changed daily, and after a 6-day incubation lagenous polypeptides, measured by digestion with specific the cell number w as determined b y a hemocytometer. The valu es represent mean ± 2 SD of 3 para ll el fl asks. A, Cul tures incubated with bacterial collagenase, was only slightly decreased in cultures cis-hydroxyproline. B, Cul tw·es incubated with azetidine carboxylic incubated with 5 a nd 25 p.g/ ml of azetidine carboxylic acid. acid . To determine whether t he incubation of cells with azetidine 264 TAN, RYHANEN, AND UITTO Vol. 80, No. 4
presence of azetidine carboxylic acid is degraded posttransla tionally. 12 .0 1- To test the helical stability of the newly synthesized procol r-r- lagen, medium and cell al-I-la beled proteins were s ubjected to A pepsin proteolysis using conditions under which triple-helical gen resists proteolysis while the nonhelical polypeptides 10.0 1-r- colla ~ are degraded into small fragments. Examination ofthe digestion .><(. en products by gel filtration on SDS-agarose indicated that in ....0 control cultures, approximately 75% of the ["H) hydroxyproline on'- 8.0 1- containing peptides were in a triple-helical conformation which 'o resisted pepsin proteolysis and were recovered as intact a ~ chains (Fig 5A, T able IV) . In contrast, essentially all (3H) U) ...J hydroxyproline-containing polypeptides synthesized in the ....J w 6.0 1- presence of 25 and 50 p,g/ ml of azetidine carboxylic acid were u degraded by pepsin (Table IV, Fig 5B). It should be noted t hat lL. r- ~ prior to pepsin proteolysis, the pro-a: chains of procollagen 0 0:: synthesized both in control cultures and in the presence of the w 4.0 1- analo gue e luted with an apparent M,. of 140,000, when chro (I) ~ matographed on SDS-agarose following reduction with 2-mer ::::> z captoetha nol (see [ 4)) . The results thus indicate that the re duced procollagen production by the fibroblasts in the presence 2.0 I- ,...... f.- of the analogue is not a result of decreased rate of transcription or translation. Rather, the reduced procollagen production can be explained, at least in part, by increased d egradation of the
~ .><(. 0 11 "'0 ~ 00 ,...... +- .3 z 60 I- w 8 I- r+ 0 20.0 0:: 0.. -r- ~ ...J 40 ...J O)lg/ml w u ~ lL. 0 20 1- I- z .><:. ::::> 1/) 15.0 0 0 2 .5 }Jg/ml ~ '+- <[ ....--...... I.O}Jg/ml ' 0 I 10 25 50 0 L-AZETIDINE-2- CARBOXYLIC ACID (}Jg/ml) )(
FIG 2. Effect. of varying concentrations of azel.idine carboxylic acid tJ) on fibroblast. proliferation. The cell cul tures were treated as described ...J ...J in Fig l, and after a 6-day incubation the number of cells was deter w 10.0 mined by hemocytometer (A) or by assayin g the content of cell protein <...> (B). The valu es are mean ± 2 SD of 3 parallel fl asks. lL. 0 0::: carboxylic acid has any effect on the mR NA activity, cell-free w translation of the total mRNA was canied out in the nuclease CD treated rabbit reticujocyte lysate system. The total incorpora ~ 5.0)Jg/ml :::> tion of ["H)proline into TCA-precipitable translation products z 5.0 was s ignificantly stimulated b y the addition of isolated RNA, indicating a human fibroblast mRNA-dependent peptide syn thesis. No difference was noted, however, between the mRNA preparation isolated fTom fibroblast cultures incubated with or without azetidine carboxylic acid (25 p,g/ml) . Also, the procol lagen specific mRNA activity,- assayed by the synthesis of bacterial coll agenase-digestible translation products, was not altered by the incubation of cells with azetidine carboxylic acid. In further experiments, the a mount of e'H]hydroxyproline in 0 3 5 7 9 dialyzable form, reflecting degradation of collagenous polypep TIME (days) tides, was also measw·ed. The results indicated that in control cul tUTes, 15-18% of the radioactive [''H)hydl'Oxyproline was in FI G 3. Time-dependent s uppression of fibroblast prolifera tion by c acid. Cells were seeded the dialyzable form (Table III); the corresponding value in varying concentrations of azetidine carboxyli on pl astic fl asks, as described in Fig 1, and the medium conta ining t he cul tures incubated with 100 p,g/ ml azetidine carboxylic acid was analogue was changed da il y. At 3, 5, 7, and 9 days of in cubation, the 32-36%. These observations clearly indicate th at a relatively cell number was determined by hemocytometer. The values are mean large fraction of collagenous polypeptides synthesized in the ± 2 SD of 3 parallel nasks. April1983 HUMAN SKIN FIBROBLASTS IN CULTURE 265
we examined 2 scleroderma fibroblast lines, which were actively ---f- synthesizing collagen, for their response to 25 ~-tg / ml of azetidine carboxylic acid. In the presence of the analogue, a s imilar -"'en A ~ 4.0 ~ reduction in ["H]procollagen production was noted as with ...... control fibroblast lines . 'o DISCUSSION 3.0 - (/) .J . One of the critical steps e nsuring the correct amino acid .J w T sequence of th e newly synthesized polypep tides is the formation u 2.0 - 1 T of a mino-acyl-tRNA complexes. As a rule, the systems respon u_ .l 0 sible for a g iven amino acid are highly s pecific, and usually only a: a molecule with th e exact stereochemical configw-ation is rec ~ 1.0 - ognized by the amino acid activating e nzym es. A n exception to :::E :::> this general rule is proline; the amino acid activating enzymes z responsible for th e synthesis o f prolyl-tRNA complex are ap 0 parently less specific t ha n other amino acid activating systems.
-"' As a consequence, a relatively large number of structural a na en logues of proline have been shown to be incorporated into .g 16.0 f-- B ...... proteins by a variety of bacterial, plant, and a nimal cell systems b (see [2,3]) . ..::.12.0 f-- Collagen is the most abundant protein in higher organisms (/) T .J (for recent reviews on collagen, see [3,32,33]) . In vertebrate .J w 1 tissues, collagen is the m ajor fibrillar component of the connec u 8.0 1- tive tissues, a nd, for example in huma n skin, collagen comprises u_ T 0 T approximately 70% of the d1·y weight of the tissue [34]. Collagen a: -'- 1 is relatively rich in proline a nd, in addition, collagen contains ~ 4 .0 f- :::E ::::> TABLE II. Incorporation of {'H]prolin e into collagenous z polypeptides a.nd correction of the values by speci(l:c radioactivity ol { 'Hjprolyl-tRNA in cultures incubated with vwying concentrations 0 1.0 25.0 100.0 of azetidine carboxylic add L-AZETIDI NE -2- CARBOXYLIC ACID (JJQ/ml) Specific FIG 4. E ffect of varying co ncentrations of azetidine carboxylic acid Concentralion "H-Coll agen rad ~ oactivity " "H-Collagen (observed)" attachment of human skin fibroblasts on tissue cul ture plastic of the of f 'Hlprolyl- (corrected)·· on the analogue (cpm x lo- •;; tRNA (cpm X 10- '/ substratum. Cells, 5 X 10'', were plated on plastic fl asks in a medium (fLg/ ml) l'g DNA) (cpm x Jo-•; I'!: DNA) co ntaining l , 25, or 100 J.!g/ml of azetidine carboxylic acid. After a 4- or ltgHNA) 24- h incubation, the medium co ntaining the unattached cells was 0 3.3 0.93 3.5 removed. T he number of cell attached was determined b y hemocyto 5 2.8 0.82 3.4 meter. The values are mean ± 2 SD. A, The number of cells 4 h after 25 1.4 0.48 2.9 the pl ating. B , The number of ce lls 24 h after the plating. " Human skin fibroblasts were incubated wi th azetidine carboxyli c acid and ["H]proline, as indicated in Table I. The amount of radioactive TABLE I. Jn. cOJ/JO ration of { 'H]proline into protein and synthesis of collagen synthesized was then assayed by di gestion with a spec ifi c non dialyzable {'Hjhydro:t)•prolin.e in fibroblast cultures incubated bacteri al co llagenase, and the radioactive peptides released by the with varying concentrations of azetidine carboxylic acid enzy me digestion are expressed as cpm per cellular DNA in the same cultw·es. Concentra- Tota l f"l-!]1-! ydroxy- DNA ["1-!]1-! ydroxy- 1 tion of the "1-! -inco rpo- proline (ftg/ fl nRk) proline , ' The specific radioactivity of ["H]prolyl-tRNA pool was determined a na logue ration (dpm x lo-'·; (dpm x JO-··; in the same cul tures as described in Ma.teria.ls and Methods. (fLg/ ml) (dpm x w-'•; ll ask) N> DNA) ,. The observed va lues fo r "H-co llagen synthesis a re divided by the llask) specific radioactivi ties of' the ["H]prolyl-tRNA pool, thus correcting for 0 40.2 6.3 5.0 1.3 the differences in the spec ific activity of the precursor ["H]proline pool. 1 33.5 4.9 4.5 1.1 5 26.3 4.1 3.2 1.3 25 21.7 2.3 3.2 0.7 TABLE III. The effect of L -azetidine-2-carboxylic acid on the degradation of newly synthesized "H -la. beled collagen polypeptides Human skin fibrobl asts in confluent monolayer cul tures were incu bated in the presence of azetidine carboxy li c acid and ["H]proline, as f"H]Hydroxyproline indicated in Materials and Methods. After 20-h incubation, the incor (dpm x 10- '/ltg DNA) (rlpm X 10- "/mg protein) po ration of "H-radioaclivity and the synthesis of ["H]hydroxyproline in Culture the nondialyzable fraction were determined. The values of radioactivity Lost in Lost in Without With dialys is.-. W ithout With in ce ll plus medium fraction are the mean of 3 parall el fl asks; the valu es d ialysis dia lysis dialysis dia lysis dialysis" (%) (%) fo r ["H]hydroxyproline are also ex pressed as dpm per 1-1g cellular DNA. Co ntrol 16.7 14.2 15.0 101 .9 84 .0 17.6 Azetidine 9.8 6.3 35.7 24.6 1.6.7 32. 1 unstable procollagen polypeptides at the posttranslationallevel, carboxylic eith er before or a fter the secretion of t he pro-a chains. acid Human skin fibroblasts were incubated with azetidine carboxylic acid Effects of Azetidine Carboxylic Acid on Procollagen (100 J.!g/ ml) and [3H]proline, which was purified free of [3H]hydro Production by Scleroderma Fibroblasts xyproline, as indicated in Materials and Methods. After 20-h incuba tion, the radioactive ['1H]hydroxyproline with and without dialysis of Scleroderma, or progressive systemic sclero is, is a clinical the samples was assayed. The valu es are ex pressed per 11g DNA or mg llagen condition ch aracterized by excessive accumulation of co cell protein in the same cultures. in the s kin a nd other connective tissues [28]. The deposition of " The fraction of ["H) hydroxyproline lost in dialysis represents the coll agen in the s kin in most cases is due to increased synthesis degradation of newly synthesized collagen, expressed as percent of the of procoll agen by the fibroblasts [29-31]. In the present study, total. 266 TAN, RYHANEN, AND UITTO Vol. 80, No.4
Vo ct verted to hydJ·oxyproline [35]. The presence of h ydroxyproline, c: together wi th the unusually high con tent of imino acids (proline 0 ~ ~ A u plus hydJ·oxyproline), plays a c ri tical role in stabilizing th e 0 1.5 r- triple-helical conformation unique to collagen (see [35]). ....-- Previous work has demonstrated that at least 5 analogues of .,...... proline are incorporated into procollagen polypeptides syn th e b sized by chick embryo tendon cells [ 4,5,36). The incorporation of any one of these analogues prevents the procollagen poly E a. peptides from folding into a stable triple-helical conformation "'0 1.0 1- [ 4,5]. S ince the triple heli x is necessary for secr etion of procol w lagen at a normal rate, the nonhelical polypeptides accumulate z in the cell s. Furthermore, the triple-h elical conformation is _J required for proper fiber formation, and, therefore, polypeptides 0 a:: that contain the a nalogue ar e unable to form functional extra a.. >- cellula r fib ers. Instead, the nonhelical polypeptides are suscep X 0 .5 1- tible to degradation by a variety of proteinases and ar e likely to 0 a:: be digested rapidly. 0 >- In the present study we have examined t he effects of 2 of the I proline a nalogues, cis-hydJ·oxyproline and azetidine carboxylic 'I' acid, on huma n skin fibroblasts in culture. The r esults indicate ~ 1 that the proliferation of t he cell s is m arkedly reduced in the 0 '~ presence of the analogue. The attachment of the cells on th e c plastic substratum is also decreased, perhaps explaining, at 0 0.6 1- B least in part, t he c ompromised growth potential. It should be -u noted that the inhibition of cell growth w as achieved with 0 .... a na logue concentrations considerably lower than those affect ...... -- I() ing t he collagen metabolism. Thus, the compromised growth b potentia l may not be a consequence of r educed collagen pro duction, as has been suggested previously [8-10]. Rather, the E reduction in the rate of proliferation a nd attachment of the a. 0.4 1- "'0 fibroblasts may reflect altered functions of pericellular or cell membrane proteins, such as fibronectin, containing t he ana w logue. z In further experiment _J s, one of the a nalogues, azetidine car 0 boxylic acid, also mru·k edly reduced the production of extracel a:: a.. lular procoll agen. Examination of the procollagen production >- 0 .2 - by the fibroblasts at different levels of protein synthesis a nd X 0 degradation indicated that the site of inhibition of collagen a:: production occurred at 0 the postt1·anslationallevel. Specifically, >- t he t ra nscription of the procoll agen genes into mRNA was not I affected, since t he mRNA activity, measured b y cell-free trans 'I' lation in a rabbit reticuJ ocyte lysate system, was about the ~ _pJ sam e 0 ~ in cells incubated without a nd with 25 ~-t g/ mJ of the analogue. 30 40 50 60 AJso, th e t ranslational r ate of proco!J agen polypeptides in the FRACTION N O. cells was only slightly decreased, if the values were corrected F I G 5. Gel filtration clu·omatography on SDS-agarose of pepsin for changes in th e s pecific radioactivity of prolyl-tRNA com digested, ["H]hydroxyproline-containing proteins synthesized in the plexes, the immediate precursor of peptide-bound proline in the presence of aze tidine carboxylic acid. Human skin fibrobl asts in con molecule. In contrast, significant increases in the degradation !luenl cul tures were in cubated with aze tidine cru·boxyli c acid, 25 /-(g/ml; of newly synthesized collagen polypeptides, measm ed as the pru·allel control cultures were incubated without the analogue. After 4- fraction of CHJhydJ'Oxyproline in dialyzable form, were ob h preincubation, 30 /-IC i ["H]proline was added and the incubation was served. It should be noted t hat if the dialyzable fraction of the continued for 20 h. After the addi tion o f proteinase inhibitors (see coll agen p eptides synthesized in the presence of the a nalogue Materials and Methods), the cells plus medium were homogenized, dialyzed against 0. 5 N acetic acid, a nd subjected to proteolysis by was underhydJ'Oxylated, as compared with the nondialyzable pepsin (330 /-(g / ml) for 6 h at 24°C. The digestions were stopped by pep tides, the values for degradation would be an underestimate adjusting the pH to 8.0, the digests were treated with SDS, and of t he co ll agen breakdown. The collagen polypeptides synt he- chromatographed on a 2.5 X 90 em co lumn of 6% agru·ose (Bio-Gel A- 5M, 200-400 mesh; Bio-Rad) elu ted with 0.1 % SDS, 0.1 M sodium ph os phate buffer, pH 7.4 . Fractions of 6 ml were coll ected, and aliquots TABLE IV. Degradation of {'H]proline·labeled collagen synthesized we re hyd rolyzed and assayed for ["H]hydroxyprolin e, as indicated in in the presence of azetidine carboxylic acid by pepsin proteolysis Materiali; and Method8. The e lu tion position of a-chains of ty pe I Concentration T otal Triple-heli cal coll agen i s indicated b y a n w-row (a). Also, the void vo lume of the oft.he "H-coll agen "1-1-collagen "H-Coll agen analogue (dpm x 10-'•; (dpm x JO -'·; degraded colu mn , determined with Blu e Dexlrnn 2000 (Phru·macia), is indicated (%) ( V.,). The LoLa! vo lume of the co lu mn determined with phenol red (/tg/ ml ) mg D NA) mg DNA)" (Nutritional Biochemi cals) was in fraction 74 . A, Pepsin digest of 0 13.3 10.1 24.1 control culture incubated without the analogue. B, Pepsin digest of aH 25 6.5 0 100.0 protei,n in culture incubated with 25 /-(g/ ml azelidine carboxylic acid. 50 2.6 0 100.0 Ce ll cultures were in cubated with azetidine carboxylic acid in con ce ntrations indicated. Triple- helical "H-co llagen in the medium plus hydroxy proline, an amino acid rather specific for this protein cell fra ction was assayed by pepsin proteolysis, followed by ge l fi ltration [3]. HydJ·oxyproline is not incorporated into collagen from a chromatography, as in Fig 5. free hydroxyproline pool in the cells, but some proline residues, " The va lu es represent nondialyzabl e ["H]hydJ·oxyproline recovered which are already in peptide linkages, a re enzymatically con- by ge l filtration in fractions 36-44 in Fig 5. April1983 HUMAN SKIN FIBROBLASTS IN CULTURE 267
sized in t h e presence of t he a nalogue were also sh own to be in 13. Busiek DF, Bauer, EA: Envi.ronmentaf pH modul ation of coll agen ase in normal human fibroblast cultures. Biochim Biophys Acta non h elical conformation, as tested by limited pepsin proteolysis. ' 585:389-397, ] 979 Thus, th e increased posttranslational degradation of t h e a na 14. Juva K, Prockop DJ: Modified procedure for the assay of "H- or logu e-containing poly peptides appears to refl ect the degrada '·'C-Ia beled hydroxyproline. Anal Biochem 15:77-83, 1966 tion of ne wly synthesized pro-cv ch a ins eith er before or after 15. Peter k o~ ky B, D iegelmann, RF: Use of a mixture of proteinase secr etion. As a consequence, t h e production of extracellular fre e co llagenases fo r the specific assay of radioactive collagen in the presence of other proteins. Biochemistry 10:988-994, 1971 collagen is inhibited, leading to deficient fiber formation. 16. Uitto J, Lichtenstein JR, Bauer EA: Characterization of procolla Excessive accumulation of collagen in tissu es is t h e major gen synthesized by matrix-free cells isolated fTom chick embryo pathologic feature in several fibrotic processes, such as pulmo tendons. Biochemistry 15:4935-4942, 1976 17. Burton K: Determination of DNA concentration with diphenyla naTy fibrosis and liver cin·h osis [3]. Sever a l dermatologic con mine. Methods Enzymol12B:163-l66, 1968 ditions are a lso ch a r acterized by excessive accumulation of 18. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Protein meas collagen in t h e dermis and th e s ubcutaneous t issu e; t h ese urement with the Folin phenol reagent. J Bioi Chem 193:265- include, among oth ers, progressive systemic sclerosis, morphea, 275, 195 1 keloids, hyper trophic scars, a nd familial cutan eous collagenoma 19. Schofield JD, Uitto J , Prockop OJ: Formation of interchain disul fid e bonds and heli cal structure during biosynthesis of procolla [28,33,34,37]. The treatmen t of th ese conditions is, for the most gen by embryonic tendon cells. Biochemistry 13: 1801-1806, 1974 part, unsatisfactory, and curren tly t h ere is no treatment m o 20. Uitto J , Booth BA, Polak I