Search

Prolyl hydroxylase was purified from human foetal skin and from a mixture of human foetal tissues by the affinity chromatography procedure using poly(l-proline). The enzyme from both sources was pure, when examined by polyacrylamide gel electrophoresis, as a native protein or in the presence of sodium dodecylsulphate, and enzyme activity recovery varied from 38% to 70% with seven enzyme preparations. The enzyme synthesized from 61.0μmol to 82.7 μmol hydroxyproline mg protein−1 h−1 at 37 °C with a saturating concentration of (Pro-Pro-Gly)5 as substrate. The molecular weight of the enzyme was identical with that of the chick prolyl hydroxylase when studied by gel filtration, and the molecular weights of the subunits of the enzyme were about 61000 and 64000 as determined by sodium dodecylsulphate-polyacrylamide gel electrophoresis. The amino acid composition of the human enzyme was very similar to that of the chick prolyl hydroxylase. Antisera to human and chick prolyl hydroxylases were prepared in rabbits. A single precipitin line was seen between the antiserum to human prolyl hydroxylase and the human enzyme in double immunodiffusion, and no cross-reactivity was detected between the human and chick enzymes by this technique. However, a distinct cross-reactivity was observed between the human and chick enzymes in inhibition experiments.

An affinity column procedure is reported for purifying prolyl hydroxylase. The procedure is based on the affinity of the enzyme for its competitive polypeptide inhibitor, and involves affinity chromatography in a column containing poly(L-proline) of molecular weight 30000 linked to agarose, and the elution of the enzyme from the column with poly(L-proline) of molecular weight 5700. The enzyme is finally separated from this polyproline by gel filtration. The procedure was employed for purifying prolyl hydroxylase from an ammonium sulphate fraction of chick embryo extract. The recovery of enzyme activity varied in ten enzyme preparations from 50 to 82%, and the purified preparations synthesized from 59.3 to 91.5 mumol hydroxyproline per mg enzyme per h at 37 degrees C with a saturating concentration of (Pro-Pro-Gly)5 as substrate. The enzyme was pure when examined by polyacrylamide gel electrophoresis as a native protein or in the presence of sodium dodecyl sulphate, and the amino acid composition of the enzyme agreed with that reported previously with only minor exceptions. The molecular weight of the enzyme measured by equilibrium in an analytical ultracentrifuge was 240000, indicating that the enzyme had been isolated in the tetramer form.

Metal activation of galactosylhydroxylysyl glucosyltransferase was studied with EDTA-treated highly purified enzyme from chick embryos and EDTA-treated citrate-soluble collagen substrate. Mn2+ activated the transferase at much lower concentrations than thought previously, and high Mn2+ concentrations were more effective than could be expected from the Michaelis-Menten equation. The data suggest that the enzyme can bind at least two Mn2+ ions, the first having a Kd of 3–5 μM (site I) and the second 50–70 μM (site II). The binding of the second Mn2+ decreases the Km of UDP-glucose from 25 μM to 6–7 μM, whereas the V at a saturating concentration of UDP-glucose is unaffected. The EDTA-treated enzyme possessed some activity even without added Mn2+, suggesting that either the transferase can be activated in the absence of any metal or that it has one additional metal binding site with high affinity. The kinetic data would be consistent with a reaction scheme involving an ordered binding of Mn2+, UDP-glucose and the collagen substrate at low Mn2+ and an ordered binding of Mn2+, Mn2+, UDP-glucose and the collagen substrate at high Mn2+ concentrations. In both cases the binding of the Mn2+, which occurs before that of UDP-glucose, is at thermodynamic equilibrium. Zn2+, Cu2+, Cd2+, Ni2+ and Ca2+ are inhibitors of the Mn2+-activated enzyme. Fe2+ (Kd 5–7 μM) and Co2+ (Kd 30 μM) also activate the enzyme at low concentrations, whereas no further activation, but instead an inhibition, was found at high concentrations. Ca2+ and Mg2+, which have previously been reported to serve as metal co-factors, did not activate the enzyme alone and did not stimulate the glucosyltransferase activated at low Mn2+ concentrations. The Kd values for Mn2+ at site II and for Co2+ are probably much higher than the physiological concentrations, whereas the activation by Mn2+ at site I and by Fe2+ may be of physiological significance.

Serum immunoreactive prolyl hydroxylase protein was measured in sixty-five patients with liver disease, and liver prolyl hydroxylase activity, immunoreactive prolyl hydroxylase protein and collagen hydroxyproline in forty of these patients. Serum immunoreactive prolyl hydroxylase protein was above the 95% confidence limit of the controls in most patients with primary biliary cirrhosis, portal cirrhosis, acute hepatitis and cancer with liver metastases, but below this in most patients with fatty liver, chronic active hepatitis, extrahepatic cholestasis, cholangitis, cancer without liver metastases and other malignant diseases. Elevated serum immunoreactive prolyl hydroxylase protein decreased rapidly with time in acute hepatitis but not in primary biliary cirrhosis. Liver prolyl hydroxylase activity and immunoreactive prolyl hydroxylase protein were elevated in the same diseases as serum immunoreactive prolyl hydroxylase protein, and correlated significantly with the latter whereas no correlation was found between serum immunoreactive prolyl hydroxylase protein and collagen hydroxyproline. Serum immunoreactive prolyl hydroxylase protein correlated highly significantly with serum alkaline phosphatase and weakly with serum aspartate aminotransferase in primary biliary cirrhosis, but not in any other disease. No correlation was found between serum immunoreactive prolyl hydroxylase protein and other tests of liver function. The results suggest that changes in serum immunoreactive prolyl hydroxylase protein in liver disease primarily reflect changes in this enzyme in the hepatic tissue, and that assays of serum immunoreactive prolyl hydroxylase in liver disease may give useful information on the actual hepatic collagen synthesis.

FOR several decades it has been apparent that many diseases directly or indirectly involve collagen, the tough fibrous material that is the most abundant protein in the body and the major constitue...

[ Lys - 14 C] Protocollagen was isolated from embryonic tendon cells, and examined as a substrate for protocollagen lysyl hydroxylase. The K m for denatured protocollagen was found to be about 10 −18 M, which expressed in terms of molar concentration of the peptide chain. This value is substantially lower than the K m for any of the synthetic peptides tested previously as a substrate for lysyl hydroxylase. The effect of helical conformation of protocollagen on hydroxylysine formation was studied by incubating native and heat-denatured protocollagen with lysyl hydroxylase at 17° C. An inhibiting effect of helical conformation was found in that lysyl residues in native protocollagen were hydroxylated at a markedly lower rate than those in denatured protocollagen. In addition, the relationship between the enzyme concentration and the synthesis of hydroxylysine with native protocollagen deviated from linearity, and the values seemed to be approaching a limit of less than one hydroxylysyl residue per α-chain. The data support th earlier suggestion that formation of a triple-helical structure may be one of the critical factors limiting the hydroxylation of lysyl residues during collagen biosynthesis. To study whether the amino-terminal peptide extensions of protocollagen had any effect on hydroxylation of lysyl residues, the extensions were cleaved off by limited pepsin digestion. The results of hydroxylation experiments were found to be similar whether protocollagen or pepsin-modified protocollagen was used as the substrate.

The kinetics of the prolyl hydroxylase reaction were studied with pure enzyme from chick embryos by varying the concentration of one substrate in the presence of different fixed concentrations of the second substrate, while the concentrations of the other substrates were held constant. Intersecting lines were obtained in double-reciprocal plots for all possible pairs of Fe2+, 2-oxoglutarate, O2 and the polypeptide substrate, whereas parallel lines were obtained for pairs involving ascorbate with each substrate. In addition, parallel lines were obtained when the polypeptide substrate concentration was varied at different fixed 2-oxoglutarate concentrations in the presence of saturating O2 concentration. Poly(L-proline) was a competitive inhibitor with respect to the polypeptide substrate, but uncompetitive with respect to Fe2+ and 2-oxoglutarate. High concentrations of the polypeptide substrate inhibited the reaction, this substrate inhibition being competitive with respect to Fe2- and 2-oxoglutarate. Succinate, CO2 and collagen were product inhibitors, succinate inhibiting the reaction competitively with respect to 2-oxoglutarate, but noncompetitively with respect to the other substrates, and collagen noncompetitively with respect to all substrates. The apparent Km and Ks values for the substrates and Ki values for the inhibitors are given. These and additional data would be consistent with a tentative reaction scheme involving an ordered binding of Fe2+, 2-oxoglutarate, O2 and the polypeptide substrate to the enzyme in this order, the binding of Fe2+ being at thermodynamic equilibrium. The enzyme can also react directly with the polypeptide substrate or its analogue poly(L-proline) under certain conditions, forming dead-end complexes. The products are released only after the hydroxylation, possibly in the order: the hydroxylated polypeptide, CO2 and succinate. Ascorbate may react either with enzyme·Fe before the release of Fe2+ or with free enzyme before the binding of Fe2+, but a reaction with ascorbate at any stage after the release of the first product is not excluded. The mechanism proposed is not entirely identical with either of the main two previous suggestions for the mechanism of 2-oxoglutarate dioxygenases.

Serum immunoreactive prolyl hydroxylase protein, galactosylhydroxylysyl glucosyltransferase activity, and the aminoterminal propeptide of type III procollagen (S-Pro(III)-N-P) were measured in twenty patients with cirrhosis and ninety with various infectious diseases, and the values were compared with those in sixty apparently healthy Nigerians. The means for all three markers were elevated significantly in the patients with cirrhosis (P less than 0.001), acute viral hepatitis (P less than 0.001), amoebic liver abscess (P less than 0.001) and the early stages of Schistosoma mansoni infection (P less than 0.001 for S-Pro(III)-N-P, P less than 0.005 for the two other markers). The mean S-Pro(III)-N-P was also distinctly elevated during the early stages of Schistosoma haematobium infection (P less than 0.01) and filariasis (P less than 0.001), whereas none of the three markers was elevated during an acute attack of malaria. Significant correlations were found between the values for the three markers within the groups of patients with cirrhosis, amoebic liver abscess and schistosomiasis, the correlations for the pooled group of all patients being highly significant (P less than 0.001). The data suggest that elevated hepatic collagen formation is found not only in cirrhosis but also in several infectious diseases. The three serum markers may be useful for showing the stages of active collagen formation in various liver diseases and for predicting the development of fibrosis in acute cases if the values remain elevated.

The glycosylations of hydroxylysine during collagen biosynthesis in isolated chick-embryo tendon cells were studied by using pulse-chase labelling experiments with [14C]-lysine. The hydroxylation of lysine and the glycosylations of hydroxylysine continued after a 5 min pulse label for up to about 10 min during the chase period. These data differ from those obtained previously in isolated chick-embryo cartilage cells, in which, after a similar 5 min pulse label, these reactions continued during the chase period for up to about 20 min. The collagen synthesized by the isolated chick-embryo tendon cells differed markedly from the type I collagen of adult tissues in its degree of hydroxylation of lysine residues and glycosylations of hydroxylysine residues. When the isolated tendon cells were incubated in the presence of L-azetidine-2-carboxylic acid, the degree of glycosylations of hydroxylysine during the first 10 min of the chase period was identical with that in cells incubated without thcarboxylic acid for at least 60 min, whereas no additional glycosylations took place in the control cells after the 10 min time-point. As a consequence, the collagen synthesized in the presence of this compound contained more carbohydrate than did the collagen synthesized by the control cells. Additional experiments indicated that azetidine-2-carboxylic acid did not increase the collagen glycosyltransferase activities in the tendon cells or the rate of glycosylation reactions when added directly to the enzyme incubation mixture. Control experiments with colchicine indicated that the delay in the rate of collagen secretion, which was observed in the presence of azetidine-2-carboxylic acid, did not in itself affect the degree of glycosylations of collagen. The results thus suggest that the increased glycosylations were due to inhibition of the collagen triple-helix formation, which is known to occur in the presence of azetidine-2-carboxylic acid.

A modified purification procedure, consisting of affinity chromatographies on concanavalin A-agarose, collagen-agarose and UDP-glucose-derivative-agarose and one gel filtration, is reported for galactosylhydroxylysyl glucosyltransferase. The enzyme obtained is entirely pure when studied by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. The enzyme protein was rich in glutamic acid + glutamine, aspartic acid + asparagine, glycine and alanine. The enzyme catalysed no significant glucose transfer to any of the glycoproteins tested, except for collagens. This included all the glycoproteins that have previously served as glucosyl acceptors for impure enzyme preparations, thus indicating a high degree of specificity of the enzyme for galactosylhydroxylysine. Galactosylsphingosine would act as a glucosyl acceptor, however. This compound has a close structural similarity to galactosylhydroxylysine in that they both have an unsubstituted amino group next to the hydroxy group to which the galactose is attached.

Kivirikko, K. I. and O. Laitinen. Effect of cortisone on the hydroxyproline in the serum and urine of young rats. Acta physiol. scand. 1965. 64. 356–360. –The effect of cortisone on the metabolism of collagen was studied by analysing the concentration of free hydroxyproline in the serum and the excretion of free and total hydroxyproline in the urine of 2- and 7-month-old rats. The younger rats, weighing 150 g at the beginning of the experiment, received 3 mg and the older rats, weighing 310 g at the beginning of the experiment, 6 mg of cortisone acetate daily. In the 2-month-old rats the concentration of free hydroxyproline in the serum and the excretion of free and total hydroxyproline in the urine decreased significantly after the administration of cortisone. By contrast, in the 7-month-old rats, which had considerably lower basal values than the 2-month-old rats, no changes or only very slight decreases were observed after the administration of cortisone. It is suggested that the decrease in the values observed after the administration of cortisone was due to a decreased content of soluble collagen fractions in the tissues. Probably the smallness of the changes in the 7-month-old rats was because a considerably smaller proportion of the hydroxyproline in the serum and urine of these older rats was derived from soluble collagen fractions.

The site of the hydroxylation of proline during collagen biosynthesis was studied by using cycloheximide to prevent the release of newly-synthesized polypeptides from ribosomal complexes. When [14C]proline and cycloheximide were added simultaneously to isolated cartilage from chick embryos, only about 1 % of the 14C incorporated into non-dialyzable polypeptides was converted to [14C]hydroxyproline. When the [14C]proline which was incorporated in the presence of cycloheximide was chased with multiple changes of fresh media containing [12C]proline and no cycloheximide, about 20 % of the incorporated [14C]proline was converted to [14C]hydroxyproline. Control experiments indicated that cycloheximide did not directly inhibit the hydroxylation of proline in protocollagen within the tissue. The question of whether peptides bound to ribosomes can be hydroxylated was examined further with purified protocollagen hydroxylase, the enzyme which hydroxylates proline in protocollagen and synthetic polypeptides similar to protocollagen. Protocollagen hydroxylase in vitro did not hydroxylate the non-dialyzable [14C]proline in 15000 × g supernatants of homogenates of tibiae which had been incubated with [14C]proline in the presence of cycloheximide. By contrast, when the labeled polypeptides in these homogenates were extracted from ribosomes with sodium dodecylsulfate or by boiling in urea, the protein-bound [14C]proline was hydroxylated by the enzyme.

The urinary excretion of total hydroxyproline was determined in 262 patients with various thyroid disorders and in 47 euthyroid patients with elevated serum protein-bound iodine (PBI) but without thyroid disease. The values were compared with normal values (limits of ± 2 S.D.) in 92 children and in 72 adults 18 to 55 years of age, reported previously from our laboratory. In addition, normal values were determined in 53 adult euthyroid patients over 55 years of age. In adult patients, the urinary excretion of hydroxyproline was increased in hyperthyroidism above the upper limit of normal values in 107 out of 111 patients 18 to 65 years of age and in 10 out of 14 patients over 65 years of age. In hypothyroidism the urinary excretion of hydroxyproline was decreased below the lower limit of normal values in 22 out of 33 patients, and PBI was decreased in 24 of these patients. In the euthyroid patients with various thyroid disorders the urinary excretion of hydroxyproline was within the limits of normal values in 63 out of 68 patients. In an additional group of 47 patients with elevated PBI without hyperthyroidism, the urinary hydroxyproline values were within the limits of normal values in 46 cases. In children, the hydroxyproline values were above the upper limit of normal values in 10 out of 13 patients with hyperthyroidism, and below the lower limit of normal values in all 23 patients with hypothyroidism. The results support the earlier suggestion that determination of urinary hydroxyproline excretion may be a useful supplement to existing laboratory tests for the diagnosis of disturbances of thyroid function. The reliability of this test is high in cases of hyperthyroidism in a dult patients, and in cases of hypothyroidism in children.

The metabolism of collagen was studied by injecting proline-14C into rats and determining the specific activities and total activities of hydroxyproline-14C in the collagen fractions of the skin and in the urine, (a) When proline-14C was injected into thyroxine-treated rats, the total activity of hydroxyproline-l4C in the collagen fractions of the skin was lower in the thyroxine-treated rats than in the controls. The total activity of hydroxyproline-14C in the urine was higher in the thyroxine-treated rats than in the controls, (b) When proline-14C was injected 30 days before the start of thyroxine injections, the administration of thyroxine caused an increase in the total activity of hydroxyproline-14C in the urine compared to the controls, (c) When proline-14C was injected into rats made previously hypothyroid by 2 131I injections, the total activity of hydroxyproline-14C in the soluble collagen of the skin and in the total collagen of bone was lower in the hypothyroid rats than in the controls. The tot...

Protocollagen proline hydroxylase was purified from chick embryos with an affinity-column procedure which yielded pure enzyme in a tetramer form. Electron microscopy of the enzyme after it was largely dissociated into monomers indicated that the monomers were rod-shaped with a diameter of 3.31 ± 0.07 nm (S.E.M.) and a length of 6.95 ± 0.07 nm. In preparations of the enzyme in which the protein was partially dissociated, dumb-bell and V-shaped structures were seen which morphologically consisted of two sub-structures each with dimensions similar to those of the isolated monomers. The results suggested that these structures were dimers in which the monomers were joined at one end with an angle between their longitudinal axes. The largest regular protein structures seen in preparations of the enzyme were about 8 × 8 nm and in some of these structures four sub-structures were seen, indicating that they were tetramers. On the basis of the data presented here we propose a model of the tetramer in which two V-shaped dimers were interlocked. Further studies demonstrated that the single-ring and four-ring structures seen by electron microscopy of enzyme preparations obtained by a previous purification procedure were not the enzyme.

The metabolism of collagen was studied with 14C-proline in organ cultures of embryonic ulnas. The addition of 0.01 ng-1.0 μg/ml of triiodothyronine caused a decrease in the incorporation of 14C-proline into protein. The amount of 14C-hydroxyproline was reduced even more, and a decrease was observed in the ratio of 14C-hydroxyproline to total 14C. In experiments with purified protocollagen proline hydroxylase, triiodothyronine did not inhibit the hydroxylation of protocollagen labeled with 14Cproline. Nor did cultivation of the ulnas with triiodothyronine have any effect on the total protocollagen proline hydroxylase activity in the bones. The degradative release of peptide-bound 14Chydroxyproline from the ulnas into the culture medium was increased in the presence of triiodothyronine, but there was no consistent increase in the release of free 14C-hydroxyproline. The results suggest that in our experimental conditions triiodothyronine caused a decrease in the accumulation of collagen in the ulnas. This ef...

Collagen degradation is essentially the only source of free and peptide hydroxyproline in plasma and urine. Normally over 95% of the hydroxyproline in urine is peptide-bound, and the excretion of peptide hydroxyproline has been used to follow collagen metabolism. Newborn infants 1–2 wk old excreted about 60 times more free hydroxyproline per kg than infants 6–12 months old. The excretion of hydroxyproline peptides per kg in the young infants was only slightly greater than the hydroxyproline peptide excretion in the older infants, suggesting that changes in collagen catabolism cannot be a major factor in explaining the large excretion of free hydroxyproline in young infants. In fullterm infants the ratio of free to peptide hydroxyproline in urine appeared to increase during the first few days of life, and then it gradually decreased to adult levels in about 5 months. The ratio of free to peptide hydroxyproline in urine was even higher in premature infants, and it remained high longer than in full-...

(1959). A Colorimetric Method for Determination of Hydroxyproline in Tissue Hydrolysates. Scandinavian Journal of Clinical and Laboratory Investigation: Vol. 11, No. 2, pp. 128-133.

Synthetic polytripeptides with the structure (l-Pro-Gly-l-Pro)n were used to study protocollagen hydroxylase, the enzyme which hydroxylates appropriate proline residues in protocollagen, the polypeptide precursor of collagen. With these synthetic substrates the synthesis of hydroxyproline could be followed quantitatively, and for the first time the reaction was studied under conditions where the substrate concentration was not rate-limiting. As previously observed with protocollagen as a substrate for the reaction, the hydroxylation of (Pro-Gly-Pro)n required α-ketoglutarate, ferrous iron, and ascorbate, but the Km values for the cofactors were slightly different from those observed with protocollagen. Hydroxylation of (Pro-Gly-Pro)n fractions of varying molecular weights indicated that a fraction with a molecular weight of about 1,800 had a Km value of 220 µg per ml, and four fractions ranging in molecular weight from 4,000 to 15,000 gave Km values of 110 µg per ml. All the Km values for (Pro-Gly-Pro)n preparations were considerably greater than that previously reported for proline-labeled protocollagen. No differences were observed in the maximal velocities with the different fractions of (Pro-Gly-Pro)n. A copolymer of glycine and proline (1:1.9) with a molecular weight of 4000 gave a Km of 440 µg per ml, and the maximal velocity was less than that observed with (Pro-Gly-Pro)n preparations. Poly-l-proline II was a competitive inhibitor of the enzyme, and the Ki value observed was 0.3 µg per ml. The results suggest that (Pro-Gly-Pro)n is not as appropriate a substrate for the hydroxylase as protocollagen, and that the higher Km values observed for preparations of (Pro-Gly-Pro)n were not explained by their smaller size. The results also suggest that the enzyme has a high affinity for proline in the extended single-stranded helix found in poly-l-proline II.

The activity of parotocollagen proline hydroxylase was assayed in punch biopsy specimens of human skin. The enzyme in human skin was similar to that from animal sources, in that it was recovered in the soluble protein fraction and in that it required atmospheric oxygen, ascorbate, α-ketoglutarate and ferrous iron. The activity of protocollagen proline hydroxylase was highest in foetuses and newborn infants, and decreased with advancing age. A slight elevation of the enzyme activity was found in three out of five patients with actively progressing scleroderma.

A 200-fold purification was obtained for the enzyme which hydroxylates proline residues in protocollagen, the polypeptide precursor of collagen. Essentially all of the enzymic activity was recovered in the soluble protein fraction of chick embryo homogenates. Ferrous iron, ascorbate, and α-ketoglutarate were cofactors for the reaction with purified enzyme. The enzyme did not hydroxylate free proline, and it did not hydroxylate either the tripeptide, glycyl-l-prolyl-l-proline, or poly-l-proline II with a molecular weight of about 15,000. The synthetic polytripeptide, (14C-l-prolylglycyl-l-prolyl)n, with a molecular weight of about 4000 was hydroxylated, and the results indicated that the enzyme is specific for proline in the second position after glycine, the position in which the hydroxyproline in collagen is found. A 30-fold purified preparation of the enzyme also hydroxylated lysine residues in protocollagen, and the cofactor requirements were similar to those for the hydroxylation of proline-labeled protocollagen. The question of whether a single enzyme performs both hydroxylations was not resolved, but the general characteristics of the enzyme or enzymes suggest that it might be appropriate to call them protocollagen hydroxylases.

Fractions of poly-l-proline Form II ranging in molecular weight from 1,600 to 21,000 were compared in their ability to serve as competitive inhibitors for the synthesis of hydroxyproline by protocollagen proline hydroxylase. The degree of inhibition of the enzyme increased with the size of the polymer, and the Ki values decreased progressively with molecular weight. The results indicated that poly-l-proline polymers with 150 to 200 residues have higher affinities for the inhibitor site or sites on the enzyme than smaller poly-l-proline fractions with the same structure in solution. The data could not be explained solely on the basis that large fractions of poly-l-proline act as multifunctional reactants with cooperative or noncooperative binding of several enzyme molecules per polypeptide. Accordingly, it appears that interaction between the enzyme and poly-l-proline involves some unusual feature such as an inhibitor site large enough to accommodate 150 to 200 prolyl residues or lateral movement of the enzyme along the polypeptide.

Protocollagen proline hydroxylase was purified from chick embryos, and it was examined by electron microscopy with negative staining techniques. Enzyme prepared in Helsinki consisted almost entirely of single ring structures. Enzyme prepared in Philadelphia under essentially the same conditions consisted almost entirely of tubular forms in which four rings were stacked together. Repeated freezing and thawing of the Helsinki preparation markedly increased the number of four-ring structures, but it was not possible to establish reproducible conditions for the interconversion of these two major forms of the enzyme. Each ring in the four-ring structure had a diameter of 100 A, a height of 30 to 35 A, and a central hole of about 20 A. The single ring structure was slightly more distended. The molecular weight of each ring was about 200,000, and therefore the molecular weight of the four-ring structure was about 800,000. The inner two rings of the four-ring structure were closer together than the outer two rings. Since the four rings appeared to be identical, the observations suggested that the rings were held together by a combination of heterologous and isologous bonds. No regular structures with more than four rings were seen, and the simplest bonding sequence consistent with these observations was (AB)—(AB)— (BA)—(BA) in which larger structures were not favored, because the formation of the inner B—B bond altered the A surfaces on the inner rings so that they were not equivalent to the A surfaces on the outer rings.

1. Attempts were made to isolate and characterize the protocollagen that accumulates in connective tissue when the hydroxylation of proline and lysine is inhibited. The term protocollagen has been used to describe the proline-rich and lysine-rich polypeptide or polypeptides that serve as substrates for the formation of hydroxyproline and hydroxylysine during the synthesis of collagen. 2. Both protocollagen and newly synthesized collagen from embryonic cartilage were isolated as complex aggregates, which contained sulphated mucopolysaccharides and other proteins or polypeptides from the same tissue. The complexes containing protocollagen were similar to those containing newly synthesized collagen when examined with several different techniques. 3. After the complexes were denatured and disaggregated, zone centrifugation and gel filtration indicated that the denatured protocollagen was similar to the denatured newly synthesized collagen obtained from cartilage in which the hydroxylation was not inhibited, and it was also similar to purified alpha-collagen. The results suggest that, when the hydroxylation is inhibited, most of the protocollagen polypeptides that accumulate are as large as complete alpha-chains of collagen. 4. Significant purification of the protocollagen polypeptides was obtained with a new technique for DEAE-Sephadex chromatography in which urea was used to prevent aggregation of the samples and the column was eluted with guanidine thiocyanate. 5. Protocollagen polypeptides were completely hydrolysed to diffusible peptides by a specific collagenase. 6. It is not entirely clear whether the hydroxylation normally begins while relatively short protocollagen molecules are still attached to polysomes, or whether protocollagen molecules of the size of alpha-collagen are synthesized even when the hydroxylation is not inhibited. 7. Results obtained with puromycin suggest that some hydroxylation occurs with smaller polypeptides, but polypeptide chains approaching the size of alpha-collagen are required to obtain complete hydroxylation of the appropriate amino acid residues of protocollagen.

Organ cultures of embryonic cartilage were prepared which synthesized collagen at a relatively constant rate for up to 72 h. When protocollagen hydroxylase in the tissue was inhibited by adding α,α′-dipyridyl to the medium, the incorporation of [14C]proline continued at 43 to 72% of the control rate but no collagen hydroxy[14C]proline was synthesized. Assays of tissue homogenates for protocollagen hydroxylase activity indicated that the inhibition of the enzyme by α,α′-dipyridyl became irreversible with time so that about half of the enzymatic activity was irreversibly lost in 4 to 6 h, and only 10% of the initial activity was recovered after 72 h. Digestion of tissue homogenates with purified collagenase indicated that under control conditions about half of the 14C incorporated from [14C]proline was in [14C]collagen, and in the presence of α,α′-dipyridyl about half of the 14C incorporated was in [14C]protocollagen. Pulse-labeling of the tissue indicated that a large part of the [14C]protocollagen synthesized early in the culture period was retained by the tissue for up to 72 h. Fractionation of tissues pulse-labeled for 6 h indicated that there was a progressive decrease in the amount of [14C]protocollagen recovered in the 100 000 × g supernatant fraction, and a complementary increase in the amount of [14C]protocollagen in the 100 000 × g sediment. Autoradiographs of tissues cultured with [3H]proline indicated that the protocollagen remained intracellular for up to 72 h whereas the collagen synthesized under control conditions was rapidly extruded into the matrix.

Sequential oligopeptides prepared by condensation of peptides with the sequence Ala-Pro-Gly or Gly-Pro-Gly were examined as possible substrates or inhibitors for the synthesis of hydroxyproline by protocollagen proline hydroxylase. The tripeptide BOC-Ala-Pro-Gly-OMe (where BOC-represents t-butoxycarboxyl- and Me represents methyl) was not hydroxylated, but oligopeptides ranging in size from BOC(Ala-Pro-Gly)2OMe to BOC(Ala-Pro-Gly)6OMe served as substrates for the synthesis of hydroxyproline. The results indicated that a peptide must contain a minimum of 4 to 6 amino acid residues in order to serve as a substrate, and they supported previous indications that a peptide of more than about 20 residues is required for optimal interaction with the enzyme. The oligopeptides H(Gly-Pro-Gly)2OH and H(Gly-Pro-Gly)4OH were not substrates for the enzyme, but H(Gly-Pro-Gly)4OH was a competitive inhibitor for the hydroxylation of proline in the polymer (Gly-Pro-Pro)n (synthesized as (Pro-Gly-Pro)n). Comparison of H(Gly-Pro-Gly)4OH with H(Pro)12OH indicated that H(Gly-Pro-Gly)4OH was a more effective inhibitor than the proline homopeptide with the same number of amino acid residues. A polymer with the permuted sequence Gly-Pro-Ala had a small inhibitory effect, and it also served as a substrate for the synthesis of hydroxyproline at a small but significant rate.

The action of growth hormone on the metabolism of collagen was studied with 14 C-proline. (a) When 14 C-proline was injected 20 days after the beginning of growth hormone injections, the total activity of 14 C-hydroxyproline in the collagen fractions of the skin and in the nonfractionated collagen of the femurs was higher in the rats treated with growth hormone than in the controls. The distribution of 14 C-hydroxy-proline between the soluble and insoluble collagen of the skin was similar in the treated rats and the controls. The total activity of urinary 14 C-hydroxyproline was higher in the treated rats than in the controls, and this increase was of the same magnitude as in the 14 C-hydroxyproline of the skin and bone. (b) When 14 C-proline was injected 12 days before the beginning of growth hormone injections, administration of growth hormone caused an increase in the total activity of urinary 14 C-hydroxyproline compared to the controls. The results of the present study suggest that the rate of collagen synthesis was increased by growth hormone, and this effect was followed by an increased urinary excretion of hydroxyproline. No marked changes were found in the rate of conversion of soluble collagen into insoluble collagen or in the rate of catabolism of soluble collagen when expressed per unit of soluble collagen. The rate of catabolism of insoluble collagen was increased by growth hormone, and this effect contributed to the increased urinary hydroxyproline excretion.

A method is described for the assay of protocollagen lysyl hydroxylase activity in human skin specimens of 100 to 300 mg wet weight. With the procedure reported here, about 90% of the enzyme activity could be extracted from the skin, and results for six skin samples analyzed separately from the same human subject agreed within a variation coefficient of 13%. The lysyl hydroxylase in human skin resembled that from chick embryos in requiring ascorbate, α-ketoglutarate, ferrous iron and atmospheric oxygen. In human skin the activity was highest in foetuses, and higher in infants than in adult subjects. The mean value in young foetuses was about 40 times, and the mean value in two infants of 2 and 4 months of age about 8 times that found in adult subjects.

A procedure for the isolation of relatively stable preparations of highly purified protocollagen proline hydroxylase is reported. The most purified enzyme preparation was homogeneous in the analytical eltracentrifuge, and showed only one minor contaminant in disc electrophoresis. This enzyme preparation synthesized about 600 μg of hydroxyproline per mg protein per h under the assay conditions used, with a saturating concentration of the synthetic polytripeptide (Pro-Gly-Pro) molecular weight 6600, as substrate. The enzyme had an S20w of 6.7. However, the elution pattern in gel filtration suggested a molecular weight of about 350 000. Amino acid analysis indicated that the enzyme preparation contained relatively large amounts of aspartic acid, glutamic acid, serine, glycine and lysine. The activity of the purified enzyme was strongly inhibited by mercuribenzoate-p , and the inhibition could be reversed by dithiothreitol. N-Ethylmaleimide was a less effective inhibitor. The results suggests that free sulphydryl groups are required for the activity of the enzyme. Although partially purified preparation of the enzyme hydroxylate lysine as well as proline residues in protocollagen, the highly purified enzyme preparations did not hydroxylate lysine residues. Thus two separate enzymes, protocollagen proline hydroxylase and protocollagen lysine hydroxylase, are probably involved in the hydroxylation of proline and in protocollagen.

The possible presence of protein-bound iron and of certain coenzymes in highly purified protocollagen proline hydroxylase was studied. All three enzyme preparations subjected to the study contained far less than 1 mole of iron/mole of enzyme. The results indicate that protocollagen proline hydroxylase does not contain stoichiometric amounts of iron bound firmly. The absorption spectrum of the enzyme showed no absorption maxima between 280 and 600 mμ, suggesting the absence of certain prosthetic groups, such as flavin nucleotides. Inhibition studies suggested that pyridoxal phosphate is not involved in the enzyme reaction, and direct assays for thiamine indicated the absence of thiamine and thiamine pyrophosphate. Thus, the decarboxylation of α-ketoglutarate coupled to the hydroxylation of peptidyl proline residues does not seem to be catalyzed by these prosthetic groups.

The kinetics of the protocollagen hydroxylase reaction were examined by the technique of measuring the initial velocities as a function of the concentration of one reactant with a series of fixed concentrations of a second reactant, and with a constant concentration of the remaining reactants. From the data a rate equation was developed for the interaction of the enzyme with ascorbate, α-ketoglutarate, oxygen, and polypeptide substrate. The kinetic data were consistent with a scheme involving consecutive bimolecular additions of ascorbate and α-ketoglutarate followed by the random order addition of oxygen and polypeptide substrate, but other kinetic schemes were not excluded. Studies with a purified enzyme system indicated that α-ketoglutarate was not consumed in stoichiometric amounts in the reaction. Further studies on the polypeptide substrate supported earlier indications that the enzyme can hydroxylate polypeptides without triple helical structure. When proline-labeled [14C]protocollagen was boiled and quenched, the synthesis of [14C]-hydroxyproline was 74% of the control value. Also, the polypeptide (Gly- l -Ala- l -Pro)n which apparently does not have triple helical structure in solution was found to serve as a substrate for the synthesis of hydroxyproline by the enzyme. The Km value and maximum velocity for (Gly- l -Ala- l -Pro)n were slightly lower than for ( l -Pro-Gly- l -Pro)n under the same conditions.

Lysyl hydroxylase catalyzes the formation of hydroxylysine in collagens by a reaction that involves oxidative decarboxylation of 2-oxoglutarate. Its binding site can be divided into two main subsites: subsite I consists of a positively charged side-chain which binds the C-5 carboxyl group, while subsite II consists of two coordination sites of the enzyme-bound Fe2+ and is chelated by the C-1-C-2 moiety. In order to identify subsite I, we converted Arg-697, Arg-700 and Ser-705 individually to alanine and Arg-700 also to lysine, and expressed the mutant enzymes in insect cells. Arg-700-Ala inactivated lysyl hydroxylase completely, whereas Arg-697-Ala and Ser-723-Ala had only a relatively minor effect. Arg-700-Lys produced 93% inactivation under standard assay conditions, the main effect being a 10-fold increase in the Km for 2-oxoglutarate, whereas the Vmax was unchanged. Arg-700 thus provides the positively charged residue that binds the C-5 carboxyl group of 2-oxoglutarate, whereas Ser-705 appears to be of no functional significance in this binding.

Lysyi oxidase activity in the culture medium of eight malignantly transformed human cell lines was very low compared with that in four control fibroblast lines, being 9–16% in five sarcoma cell lines and 7–11% in three other tumour cell lines. The low enzyme activity was probably due to deficient enzyme synthesis rather than impaired secretion into the cell medium, as low activity was also found in urea extracts of the cell pellets. Lysyi oxidase production thus appears to be closely regulated with deficient collagen gene expression in malignant transformation.

A two-step procedure is described for the purification of protein disulphide-isomerase (PDI). This procedure is based on the previous finding that the beta-subunit of the prolyl 4-hydroxylase tetramer (alpha 2 beta 2) is identical with PDI [Koivu, Myllylä, Helaakoski, Pihlajaniemi, Tasanen & Kivirikko (1987) J. Biol. Chem. 262, 6447-6449; Pihlajaniemi, Helaakoski, Tasanen, Myllylä, Huhtala, Koivu & Kivirikko (1987) EMBO J. 6, 643-649]. The procedure involves purification of the prolyl 4-hydroxylase tetramer by a simple affinity chromatography and subsequent isolation of the beta-subunit from the dissociated tetramer by ion-exchange chromatography.

The hydroxylation of proline and lysine residues by the collagen hydroxylases is coupled with a stoichiometric decarboxylation of 2-oxoglutarate. Ascorbate is virtually a specific requirement for these enzymes, but previous studies have demonstrated that it is not consumed during most catalytic cycles. Prolyl 4-hydroxylase and lysyl hydroxylase are known also to catalyze an uncoupled decarboxylation of 2-oxoglutarate in the absence of the peptide substrate. It is shown here that, unlike the complete hydroxylation reaction, the uncoupled decarboxylation reaction involves stoichiometric ascorbate consumption. This stoichiometric ascorbate consumption was also seen when the rate of the uncoupled prolyl 4-hydroxylase reaction was enhanced by the addition of poly(L-proline). Since collagen hydroxylases may catalyze occasional uncoupled reaction cycles even in the presence of the peptide substrates, the main function of ascorbate in these reactions in vivo is suggested to be that of reactivating the enzymes after such uncoupled cycles.

Isotopic studies have indicated that the hydroxyproline excreted in urine is derived primarily from the degradation of collagen (1, 2), and that the excretion of hydroxyproline can be used as a measure of collagen metabolism (for reviews, see 3, 4). Normally, 95% or more of the hydroxyproline excreted in the urine of children and adults appears as peptides, and less than 5% is present as the free imino acid. No significant change in the ratio of free to peptide hydroxyproline in urine has been demonstrated in most conditions in which collagen degradation and the excretion of peptide hydroxyproline are increased (3, 4). The relatively small amounts of free hydroxyproline in urine are probably explained by the higher renal threshold for free hydroxyproline than for peptide hydroxyproline (5, 6) and by the rapid enzymatic degradation of free hydroxyproline to Δ-pyrroline-3-hydroxy-5-carboxylic acid and γ-hydroxy-l-glutamic acid (see 7). Recently, Efron et al. (7) have observed high levels of free hydroxyprol...

Monoclonal antibodies against human prolyl 4-hydroxylase (EC 1.14.11.2), an intracellular enzyme of collagen biosynthesis, were produced by fusing spleen cells from BALB/c mice hyperimmunized with human prolyl 4-hydroxylase and mouse myeloma cells (P3/NS 1/1-AG 4-1). Hybridomas from 14 different primary microtiter-plate well cultures produced antibodies to human prolyl 4-hydroxylase; six of them with the highest antibody titer were cloned and antibodies produced by one clone from each of the six lines were further characterized. All of the six cloned hybrids produced antibodies of the IgG class as detected by immunodiffusion. The enzyme antigen used in the present study was a tetramer composed of two pairs of different subunit proteins, alpha and beta. Only one clone which produced antibodies to the alpha subunit was obtained, the other five antibodies being directed against the beta subunit. All the antibodies reacted with the tetramer form of the enzyme. Species cross-reactivity of the antibodies was tested using cultured human, mouse and chick fibroblasts and purified prolyl 4-hydroxylase from chick and mouse sources. None of the antibodies cross-reacted with chick or mouse fibroblasts, as determined by immunofluorescence, whereas one antibody reacted with purified chick and mouse prolyl 4-hydroxylase when examined by the western blotting technique. This antibody caused a strong inhibition of human prolyl 4-hydroxylase activity, but the other five antibodies had negligible inhibitory effect on the activity of the enzyme.

Prolyl 4-hydroxylase [procollagen-proline, 2-oxyglutarate 4-dioxygenase; procollagen-L-proline, 2-oxoglutarate:oxygen oxidoreductase (4-hydroxylating), EC 1.14.11.2], an alpha 2 beta 2 tetramer, catalyzes the formation of 4-hydroxyproline in collagens by the hydroxylation of proline residues in peptide linkages. We report here on the isolation of cDNA clones encoding the alpha-subunit of the enzyme from human tumor HT-1080, placenta, and fibroblast cDNA libraries. Eight overlapping clones covering almost all of the corresponding 3000-nucleotide mRNA, including all the coding sequences, were characterized. These clones encode a polypeptide of 517 amino acid residues and a signal peptide of 17 amino acids. Previous characterization of cDNA clones for the beta-subunit of prolyl 4-hydroxylase has indicated that its C terminus has the amino acid sequence Lys-Asp-Glu-Leu, which, it has been suggested, is necessary for the retention of a polypeptide within the lumen of the endoplasmic reticulum. The alpha-subunit does not have this C-terminal sequence, and thus one function of the beta-subunit in the prolyl 4-hydroxylase tetramer appears to be to retain the enzyme within this cell organelle. Interestingly, three of the cDNA clones for the alpha-subunit contained a 64-nucleotide sequence homologous but not identical to the corresponding 64-nucleotide sequence found in four other cDNA clones. Nuclease S1 mapping experiments demonstrated that this difference was due to the existence of two types of mRNA present in approximately equal amounts. Southern blot analyses of human genomic DNA with a cDNA probe for the alpha-subunit suggested the presence of only one gene encoding the two types of mRNA, which appear to result from mutually exclusive alternative splicing of primary transcripts of one gene.

Lysyl oxidase is an extracellular enzyme that initiates crosslink formation in the major connective tissue proteins, the collagens and elastin. This enzyme activity accumulated in a fresh medium of cultured human skin fibroblasts for at least 24 h, but the accumulation was distinctly non-linear after the first 12 h. Most of the total enzyme activity was present in the medium, the activity found in the cell layer representing about 30% of the total activity at 4 h, and about 10-15% at 24 h. The bulk of the cell-layer-associated activity appeared to be extracellular, as more than half was lost upon trypsinization. Culturing of the cells for 8 h in the presence of either monensin or nigericin, ionophores known to inhibit the secretion of many proteins at the level of the Golgi complex, markedly reduced the accumulation of lysyl oxidase activity in the medium. Monensin was particularly effective, as it produced a distinct inhibition even at a 10 nM concentration, reaching 50% at 30 nM. Both ionophores also reduced enzyme activity in the cell layer, whereas no definite decrease was seen in the activity of the trypsinized cells. The effect of monensin was evidently not due to any general toxicity on the part of the drug, since even a 500 nM concentration gave no inhibition of the incorporation of [3H]leucine into total protein. Tunicamycin also reduced lysyl oxidase activity in the medium and to a lesser extent in the cell layer, but the effective dose, 1-10 micrograms/ml, also inhibited the incorporation of [3H]leucine into total protein. The reduced enzyme activity may therefore not be due to a direct effect of tunicamycin on the glycosylation of the lysyl oxidase protein itself but may be mediated through other actions of the drug. Colchicine caused no inhibition in lysyl oxidase activity secretion even at a 10 microM concentration, although it has been reported to inhibit collagen secretion at doses more than one order of magnitude lower.

Prolyl 4-hydroxylase, an alpha 2 beta 2 tetramer, catalyses the formation of 4-hydroxyproline in collagens. The beta subunit is known to be identical with the enzyme protein disulphide-isomerase and to possess disulphide-isomerase activity even when present in the prolyl 4-hydroxylase tetramer. We here report that lysyl hydroxylase, a homodimer, and algal prolyl 4-hydroxylase, a monomer, do not contain detectable protein disulphide-isomerase activity. Since the hydroxylase reaction mechanisms are similar, the data suggest that the protein disulphide-isomerase activity of the vertebrate prolyl 4-hydroxylase beta subunit is unlikely to be involved in the catalytic mechanism of the hydroxylation reaction.

Prolyl 4-hydroxylase (Procollagen-L-proline, 2-oxoglutarate: oxygen oxidoreductase (4-hydroxylating), EC 1.14.11. 2.) catalyzes the hydroxylation of peptide-bound proline residues, exclusively at the N-terminus of glycine (1–3). The first step of the enzymatic reaction, the oxidative decarboxylation of 2-oxo-glutarate by one atom of a dioxygen molecule, has been suggested to proceed as a ligand reaction in the coordination sphere of enzyme-bound Fe2+ (4), generating succinate, carbon dioxide, and a highly reactive iron-oxygen atom complex. This ferryl ion subsequently hydroxy lates an appropriate proline residue, probably by an abstraction-recombination mechanism (4). The oxidative decarboxylation of 2-oxoglutarate can proceed without subsequent hydroxy lation in so-called uncoupled reaction cycles (5–7) in which ascorbate is utilized as a specific alternative oxygen acceptor (8,9). In the absence of this reducing agent, prolyl 4-hydroxylase is rapidly inactivated by self-oxidation (10,11).

The anthracyclines doxorubicin and daunorubicin were found to act as irreversible inhibitors of prolyl 4-hydroxylase. The reaction rate for enzyme from both chick and human origin was first order, the concentration of inhibitor giving 50% inhibition being 60 microM for both compounds after 1 h. The effect was dependent on the presence of iron ions in the reaction mixture. Inactivation could be prevented by addition of high concentrations of ascorbate, but not 2-oxoglutarate, before the inactivation period. The same results were obtained with competitive analogues of these cosubstrates. Lysyl hydroxylase from chick embryos was also susceptible to inactivation. Its activity was decreased by 50% after incubation for 1 h with a 150 microM concentration of the inhibitors. When chick-embryo prolyl 4-hydroxylase was incubated with [14-14C]doxorubicin, both enzyme subunits were radioactively labelled, about 70% of the total radioactivity being found in the alpha-subunit. Since the anthracyclines are known to undergo a redox reaction generating semiquinone radicals with Fe3+ only, the results suggest that the enzyme-bound iron ion is oxidized to a tervalent intermediate in uncoupled reaction cycles. The data also suggest that both enzyme subunits contribute to the catalytic site of prolyl 4-hydroxylase.

The concentration of immunoreactive prolyl hydroxylase was measured in human skin, serum and synovial fluid. In skin concentrations of the immunoreactive enzyme protein and the enzyme activity were highest in foetuses, at an intermediate level in 0–1-year-old children and lowest in adults. About 13–19% of the enzyme protein in foetal skin was in the active form compared with 2–4% in the adult. Concentrations of immunoreactive prolyl hydroxylase in serum were lower than in skin and no changes were found with age. The concentration of immunoreactive prolyl hydroxylase in synovial fluid of the knee joint in patients with traumatic effusion was about half that found in serum. The ratio of enzyme tetramer, which is the active form of the enzyme, to the total immunoreactive protein was studied by gel filtration. In foetal skin, about 15–20% of the total immunoreactive protein was present in tetramer form, and a good correlation was found between the ratio of enzyme tetramer to the total immunoreactive protein and the ratio of enzymic activity to the concentration of total immunoreactive enzyme. Small amounts of enzyme tetramer were found in the serum and synovial fluid. The data suggest that the activity of prolyl hydroxylase in human tissues may be partly controlled by a mechanism involving changes in the ratio of the active enzyme tetramer to total enzyme protein.

The collagen component of isolated glomerular basement membrane ws solubilized by limited pepsin digestion and further purified. The amino acid and carbohydrate composition of the final material was very similar to that described for the alpha1-chains of type IV collagen. However, the presence of several components was detected when the material was analyzed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis either without or after reduction. The molecular weights of the major components in the reduced material were about 140,000, 100,000, 80,000 and less than 65,000 (in the case of five components). The data thus do not support the previous suggestion that the human glomerular basement membrane contains only one type of collagen polypeptide chain. However, part of the heterogeneity may be due to the presence of more than one collagenous part in a procollagen-type polypeptide chain, and an additional reason may lie in a partial degradation of collagenous portions during pepsin digestion.