STUDIES ON THE REGULATION OF PROLYL HYDROXYLASE

Chichester, Clinton Oscar. M.S., University of Rhode Island, 1976. Studies on the Regulation of Prolyl Hydroxylase. Major Professor: Dr. George C. Fuller The hydroxylation of proline is thought to be one of the critical cellular events necessary for the synthesis and secretion of structural collagen. Using antibody directed against prolyl hydroxylase it has been shown that there is an enzymatically inactive protein related to prolyl hydroxylase in mammalian tissue. This cross-reacting protein is always present in excess relative to active hydroxylase and it is not kriown whether it is a precursor or a degradation product 0£ prolyl hydroxylase. The turnover rates of prolyl hydroxylase and immunolo.gically related protein, CRP, were examined using labeled leucine as precursor or by measuring the decay of elevated prolyl hydroxylase and CRP back to basal levels. Prolyl hydroxylase and CRP were purified from neonatal rabbit skin at various times following the administration of 3 H-leucine. Prolyl hydroxylase was purified by affinity chromatography. CRP was purified by antibody precipitation from the dialyzed 70% (NH4)2S04 supernatants and subsequent electrophoresis on 10% SDS polyacrylamide slab gels. CRP was shown to migrate similarly to the two prolyl hydroxylase monomers which had molecular weights of 65,000 and 60,000. A smaller

there is an increase in the biosynthesis of collagen (Grant and Prockop, 1972). In order to effectively modify this pathological process the control of collagen synthesis must be elucidated.
Prolyl hydroxylase (EC i.14.11.2; praline, 2-oxoglutarate dioxygenase) converts specific prolyl residues in the peptide precursors of collagen to 4-hydroxyproline and is thought to be one of the critical cellular events necessary for th~ synthesis and secretion of structural collagen (Cardinale and Udenfriend, 1974). Although the importance of prolyl hydroxylase as a controlling factor in collagen synthesis is unclear, large increases in activity have been reported in a variety of tissues responding to injury induced damage which result in increased collagen synthesis (Grant and ~rockop, 1972) . This parallelism has resulted in the use of prolyl hydroxylase as a marker for the rate of collagen synthesis, and has helped generate interest in the cellular regulation of this hydroxylase. , obtained evidence with antibody against prolyl hydroxylase that an enzymatically inactive antigen is present in L-929 fibroblasts which may be a precursor to the active enzyme. Examination of animal tissues using antibody directed against rat (Stassen et al., 1974), rabbit (Fuller et al.,19 76) or human prolyl hydroxy lase (Tuderman et al., 1975) has confirmed the presence of both forms of antigen (active enzyme and immunologically crossreacting protein, CRP) in mammalian tissues with CRP always in excess relative to active hydroxylase. However, the demonstration of a precursor product relationship between these two proteins has not been reported.
This study examines the relationship between prolyl hydroxylase and CRP through the determination of the turnover rate of these two protein pools. Prolyl hydroxylase and CRP were purified from neonatal rabbit skin at various time periods following t!'Eadministration of 3 H leucine. To confirm the half-lives observed in the neonates, prolyl hydroxylase and CRP were elevated in the vasculature of rabbits by daily exposure to thyroxine and epinephrine injections Wuller and Langner, 1970), and the rate of decay back to basal levels was measured.  (Miller, 1974).
Collagen synthesis occurs by a series of steps starting with the synthesis of polypeptide 'precursors rich in proline and glycine called procollagen chains (Martin, 1975). While still on the polyribosomal complex susceptible proline and lysine residues are hydroxylated and glycosylation of the ~esulting hydroxylsyl residues occurs intracellularly. Triple helical aggregates of pro r;f... chains are secreted by the cells and subsequently a segment is cleaved from the non-helical . region on both the C and N terminal ends by the action of procollagen peptidases (Fessler et al., 1975;Goldberg, et al., 1975). Further extracellular processing occurs through the action of lysyl oxidase which oxidatively deaminates specific lysine and hydroxylysine residues leaving aldehyde moieties which form extra-and intramolecular cross-links through condensation reactions (Siegel et al., 1970). 5 It has been suggested that the hydroxylation of proline is one of the rate limiting steps in collagen synthesis (Udenfriend, 1966). Hydro x yproline is important for the structural integrity of collagen. The name prolyl hydroxylase is a descriptive title since it hydroxylates peptidyl bound praline. The enzyme was the first described member of the~-ketoglutarate requiring mixed-function oxygenases. It has been established for some time that molecular oxygen is the source of oxygen for the hydroxyl group, (Prockop, et al., 1962;Fujimoto and Tamiya, 1962). Alpha keto-glutaiate is an essential co-factor which is stoichiometrically decarboxylated to succinic acid in relation to the amount of hydroxyproline formed (Rhoado and Cardinale, et al., 1971). The "in vitro" r e action also requires a ferrous ion and ~scorbic aci d reduction-oxidation system.
Other reducing agents,such as the tetrahydropteridines or reductones  can substitute.

--
The cellular site at which hydroxylation of praline occurs wa s d ebated for some time. Miller and · Udenfriend (1971) · provided the first evidence that hydroxylation occurs on nascent chains by isolating ribosomes from guinea pig granu-14 loma minces which had been incubated with c praline. They showed that the ribosomes containe d p ep t i rl y l bound 1 4 c praline and 14c hydroxyproline was released from the ribosomes by puromycin treatment. Lazarides and Lukens (1971), confirmed that the site of hydroxylation occurs on nascent chains by 6 labeling 3T6 f ibrobLasts with 3H praline and isolating polysames and showed that they did contain radioactive hydroxypraline. They further showed that when hydroxylation was inhibited, underhydroxylated chains were still released.
The location of prolyl hydroxylase in membrane was suggested by the observation that enzyme activity is increased in homogenates by treatment with detergents (Guzman and Cutroneo, 1973;Harwood et al., 1974). Cutroneo (1974) reported the isolation of a microsomal fraction which contained both the highest specific activity of cellular prolyl hydroxylase and substrate which could be hydroxylated. EM studies using conjugated antibodies to prolyl hydroxylase .
have confirmed the microsomal localization of the enzyme 197 4;Olsen et al. ,197 3.) . Recently, Peterkofsky and Assad (1976) have shown that low concentrations (0.05%) of detergents such as Triton X-100 or Brij-35 can release prolyl hydroxylase from isolated microsomes~ In addition, prolyl hydroxylase, which could subsequently be released from the microsomes by was resistant to trypsin proteolysis at concentrations which removed 40% of the protein from the microsomes. These results suggest that prolyl hydroxylase is located within the cisternae, either bound to the inner membrane or freely soluble.

Prolyl Hydroxylase Substrates
The hydroxylation of praline "in vivo" occurs primarily on nascent procollagen c/.... chains. Thus, most substrates are fairly large macromolecules but the enzyme will hydroxylate susceptible proline residues of small peptides.
McGee, Rhoads, and Udenfriend (1971) studied the vasoactive peptide bradykinin and some of its substituted analogs as substrates for the enzyme. They showed that the minimum sequence required for prolyl hydroxylation was an X-Pro-Gly triplet. Adjacent amino acid residues to this triplet were shown to modify the rate at which hydroxylation occurred. Hutton et al. (1968) showed, using the synthetic peptide (Pro-Gly-Pro)n, that as the molecular weight of the substrate increased from 1,200 to 8,000 its effectiveness as a substrate also increased. This was due to a <lecrease in Km while the Vmax value remained constant.
There has been much work done on determining which susceptible praline residues get hydroxylated. Berg and Prockop (1973a) demonstrated, as others had deduced (Rhoads arid Udenfriend, 1968;Fujimoto and Prockop, 1968), that collagen chains can only be hydroxylated when they are in random-coil configuration and not when they are triple helical. Bornstein (1967) showed, using CNBr cleavage fragments of rat~l chains, that individual praline residues, which are susceptible to hydroxylation, were reproducibly, incompletely hydroxylated. The degree of hydroxylation saems to be tissue specific and is dependent on adjacent amino acids.

Purification of Prolyl Hydroxylase
Prolyl hydroxylase is a uniquitous enzyme being found in all mammalian tissue so far analyzed. The enzyme has been purified to homogeneity from several different sources having 8 high levels of activity including chick embryos (B e rg and Prockop 1973b), newborn rat skin, (Rhou.ds u.nd Ud c n f ricnd, 1970), and human fetal material (Kuutti et al., 1975). The molecular -weight of prolyl hydroxylase, obtained from chick and human tissue is 240,000 (Kuutti et al., 1975). The molecular weight of the subunits are 61,000 and 64,000 (obtained by dissociation of the enzyme) suggesting that the enzyme is a tetramer (Kuutti et al., 1975). Amino acid analysis shows that the protein contains a large amount of aspartic and glutamic acid which accounts for its acidic nature.
Several different methods have been used in the purif ication of the enzyme.
Initial purification methods involved (NH4)2S04 precipitation followed by ion:--exchange chromatography and gel filtration (Rhoads and Udenfriend, 1970). Subsequently the high affinity of prolyl hydroxylase for its native substrate (Km ':::t 2nM Berg and Prockop 197 3a) was utilized to develop an affinity column method for purif icat~on of the enzyme (Berg and Prockop, 1973b). This procedure involved affinity chromatography of prolyl hydroxylase on a column containing · Ascaris cuticle collagen linked to agarose and the elution of the enzyme from the column using a high concentration of a synthetic substrate, (Pro-Gly-Pro)n. Recently, a second affinity column procedure has been developed based on the high affinity of the enzyme for its competitive inhibitor, poly (L-proline) (Tuderman et al., 1975a).

--
The affinity column in this case consists of poly (L-proline), molecular weight 30,000, linked to agarose and the enzyme is eluted with poly (L-proline) which has a mole cular weight of 5,700.
As a logical consequence of enzyme purification, antibodies have been developed against rat (Roberts et al., 1973) rabbit (Fuller et al. ,197 6) , chick (Berg et al. ,197 2) , and human (Kuutti et al., 1975) prolyl hydroxylase. Using these antibodies, several assays have been developed to measure enzyme related antigen in tissues. McGee and Udenfriend (1972a) used an antibody to rat prolyl hydroxylase to identify the pr~sence of a protein immunologically related to prolyl hydroxylase (cross-reacting protein, CRP) in L-929 fibroblasts and reported the separation of CRP from enzyme. The main disadvantage of their enzyme immunoassay was that enzymatically inactive cross-reacting proteins could only be measured in extracts which contained little or no enzyme aetivity. Stassen et al. (1974) modified the original enzyme immunoassay so that it could be used in the presence bf large quantities of active prolyl hydroxylase. With the modified enzyme immunoassay, it was shown that tissues bf rat and mouse contain large amounts of CRP relative to the amount of active prolyl . hydroxylase. McGee and Udenfriend (1972b) were able to isolate and separate these two protein species from early log phase cultures of L-929 fibroblast cells by ion-exchange and gel £iltration chromatography. It was found that CRP from the fibroblasts has a molecular weight between 85,000 and 105,000 compared to prolyl hydroxylase which has a molecular weight between 260,000 and 300,000. The relationship of CRP and prolyl hydroxylase is still unclear.
Recently, a radioimmunoassay has been developed for human and chick prolyl hydroxylase (Tuderman et al., 1975b).

Regulation of Prolyl Hydroxylase
Activation of collagen synthesis occurs as a resp6nse to injury or as a result of rapid growth. The mechariisms for the regulation of collagen synthesis, however, are unclear at the present time. The question of regulation becomes very important in fibrotic diseases where there is an overproduction of connective tissue elements. As a result, prolyl hydroxylase has been investigated as a possible controlling factor in collagen synthesis. It is well established that prolyl hydroxylase activity is increased in tissue when collagen synthesis is stimulated. Siegel (1976) has studied the temporal relationship of the increases in the various enzymatic steps required for collagen biosynthesis in the carbon tetrachloride d.amaged liver. This data clearly indicates that increased prolyl hydroxylase activity is the first change observed and that this occurs even prior to an increase in collagen chain synthesis.
In experimental models £or disease states such as epinephrine-thyroxine induced atherosclerosis (Langner and Fuller, 1973}, _ where increases in prolyl hydroxylase occur . before there are detectable changes in collagen synthesis, CRP levels are also increased but to a smaller extent than prolyl hydroxylase (Fuller et al., 1976).
It has been proposed that proline hydroxylation is a rate limiting step in collagen biosynthesis (Udenfriend, 1966).
This is based on the fact that inhibition of prolyl hydroxylase

I'
by~'~ dipyridyl i~ chick tendon cells causes secretion of procollagen at a very reduced rate (Jiminez et al., 1973).
In a related study, Jiminez et al.,( 197 4) suggests that this is due to the failure of underhydroxylated collagen to form stable triple helices which may be required for secretion.
Additional evidence for the necessity of prolyl hydroxylation would include the fact that none of the identified human inheritable connective tissue diseases involves the loss of prolyl hydroxylase activity while there are genetic diseases accredited to each of the other enzymes in the collagen biosynthetic pathway (McKusick, 1972).
Cell culture has been a popular method for studying the regulation of prolyl hydroxylase. Studies with L-929 fibroblasts in culture show that the formation of peptidyl bound hydroxypropline increases toward the end of the logarithmic phase of growth which is accompanied by a sharp increase in prolyl hydroxylase activity (Green and Goldberg, 1963;Gribble et al., 1969). Enzyme activity in early log phase cells can be increased by concentrating the cells to a higher density ; or by the addition of sodium lactate , or sodium ascorbate (Stassen . et al., 1973) to the culture medium. Administration of protein synthesis inhibitors such as puromycin or cycloheximide do not inhibit these increases in prolyl hydroxylase activity and hydroxyproline formation (Peck et al., 1967;. Further suggesting that protein synthesis is not required for an increase in prolyl hydroxylase activity, The r 0 qui retnents f or "in v it ro" a ctiva tio n · a re identical to t hos e ne e d ed f o r the hydroxy lation rea ction i. e ., ol\ -ke t oglut ar ate, ascorbate, f errous i on and catalase. Duri ng a c t ivati o n hydr oxypro l i ne is forme d ( Ku ttan, 19 7 6). This woul d s ug g est that t here is a complex between activ e enzyme and ar. ~ u nderhydroxyla ted . f orm of c ollagen. 'fh us a cti va ti on occ urs. when the en zyme is fre e d due to hyd r oxy la tidn of the endoge nous substrate. It appears t h at "ir1 v ivo" act ivatio n i n t is s re cultur e is th e same as the " i n vitro" acti vat ion s :L nce the s an1e maximum level of enzyme activati o n is ef fec ted by both me thods. I n a dditi o n, o nc e . maximum acti v · t.y is a ch ieved " in vit::::-0 11 acti vatible fo rm of t he enzyme is differe n c: f rom t he 13 small molecular weight component of CRP since the activatible enzyme, CRP and active enzyme can be separated into three peaks on DEAE-Sephadex (Kuttan et al., 1975). Thus CRP is a heterogeneous pool, ~ut in all probability a portion of the pool is precursor to · active enzyme.
In experiments designed to elucidate the mechanisms of fibrosis, McGee. et al. (1973) have found a material fractionated from the liver of mice with acute carbon tetrachloride liver injury which stimulates prolyl hydroxylase and collagen synthesis in L-929 fibrqblasts. Three collagen stimulating factors were found with an approximate molecular weight of 5,000 which were not found in control livers. It is possible that these factors may control the synthesis of collagen ''in vivo".

Protein Turnover
It is now well established that all proteins are continually being turned over. Schimke (1974) has reviewed the subject and has described several common features of protein turnover. First, it appears that most intracellular proteins are degraded intracellularly. Secondly, there is heterogeneity between the rates of degradation of different proteins.
In fact, the rate of degradation of a given prqtein within a cell can change with respect to the metabolic state of that cell.
A multitude of methods has been used to measure protein half-lives but all are based either on time course of 14 chanqes in enzyme activity or on the use of isotopic tracers.

Rates of synthesis and degradation can be obtained by observ-
ing the time course of change in enzyme activity (increased or decreased) after the institution or withdrawal of a stirnulus (Segal and Kirn, 1963). Any change in enzyme content can be described by the following equation: Although saturation labeling is more precise, the most common method for measuring the. rate of degradation using isotopes is the single administration of a radioactive amino acid precursor due to the high cost of label. The loss of 15 specific activity in the protein is exponential, which allows for the calculation of Kd. The major limitation of this method is reutilization of labeled amino acid. The effect of label reutilization on the measurement of turnover, however, can be calculated (Poole, 1971). ) 2 so 4 . After centrifugation at 45,000 x g, this . enzym~ solution was placed on affinity columns consisting of reduced and carboxymethylated Ascaris collagen coupled to Sepharose 4B as previously described (Berg and Prockop, 1973).
Each individual sample was applied to a separate 1.5 x 4 cm column, and enzyme was eluted with 1 ml of buffer containing The antibody directed against rabbit prolyl hydroxylase required for these experiments was obtained from an immunized goat using conditions similar to those previously reported for . the antiserum directed against rat skin .prolyl hydroxylase (Roberts, et al., 1973). The antigen used was rabbit skin · prolyl hydroxylase purified by affinity chromatography. Be..;.. fore injection into goats the enzyme was separated from the (Pro-Gly-Pro)n used to elute the enzyme by electrophoresis in 7.5% polyacrylamide at 4°c (Davis, 1966). The tetramer form of the enzyme, which constituted the major band on each qel in these preparations, was identified by its catalytic activity and could be measured in undenatured form as a fluorescent b a nd i n ultravi ole t light oy s ta ini ng wi t h a n i lino n a phthalene Polvacrylamid.e Electrophoresis Initial experiments analyzing irmnunop recipitates were done on 7.5% acrylamide disc gels (Davis, 1 966). This system was also used to determine the purity of e nzyme preparations .

20
To quantitate the radioactivity present, the gels were sliced and the tritium content determined (in 10 ml Aquasol) following digestion in 0.1 ml H202.
The final system for studying the incorporation and decay of label in CRP was the 10% SDS-polyacrylamide slab sys-' tern (70 cm x 10 cm x 2.5 mm) described by Laemmli and Faure (1973

Determination of Free Leucine Specific Radioactivity in
Homogenates From each of the 20,000 x g supernatants 0.4 ml aliquot was taken and the protein precipitated with 50 ul of 50% TCA.

21
After 30 minutes on ice the samples were centrifuged and the supernates harvested. Each sample was placed on a Dowex 50W-8X column (5 ml bed volume) and was washed with 15 ml distilled water. The columns were eluted with 2 ml of lON NH 4 0H and 2 ml fractions were collected. The radioactive fractions were pooled, lyophilized and then resuspended in 100 ul of 0.15N lithium citrate buffer (pH 2.2). Tritium content was determined in 10 ml Aquasol and leucine · content determined by amino acid analysis on a Durrurn D-500 analyzer (Lee, 1974).

Amino Acid Analysis of Purified Enzyme Samples
The leucine content of labeled enzyme samples as well as standard prolyl hydroxylase pools were determined by amino acid analysis after hydrolysis. The standard enzyme pools were purified in large batches according to the previously described affinity method (Berg and Prockop, 1973), using a 1.5 x 30 cm column. After the enzyme was eluted from the column, with 10 ml of (Pro-Gly-Pro)n (10 mg/ml), it was concentrated 10-fold in an ultrafiltration chamber with a ' membrane _having a 30,000 MW cut-off. For amino acid analysis and electrophoresis enzyme was separated from (Pro-Gly-Pro)n using a Sephadex G-200 or G-150 column (0.9 x 30 cm). Enzyme pools were dialyzed exhaustively against distilled water before hydrolysis.
The dialyzed standard pools and the labeled enzyme sample were hydrolyzed at ll0°c in 6N HCl, 0.5% phenol for 20 hours in tubes sealed under a 25 millitore vacuum. The 22 samples were then evaporated to dryness, brought up in lithium citrate buffer and analysis was carried out on the Durrum D-500 analyzer (Lee, 1974).

Epinephrine-Thyroxine Induced Arteriosclerosis
Prolyl hydroxylase and CRP were elevated in the vasculature of male New Zealand rabbits by daily injections of epinephrine and thyroxine for 12 days (Fuller and Langner, 1970  The distillate (0.8 ml) was added to 10 ml of Aquasol (New England Nuclear) and the radioactivity determined.

Preparation of Substrate for Prolyl Hydroxylase Assay
The tritium labeled substrate was prepared using the method of Hutton et al. (1966). Five hundred 7 to 8 day chick embryos were removed, decapitated and the bodies placed in icecold Krebs-Ringer buffera (Stone and Meister, 1962). The intact embryos were washed twice with ice-cold Krebs and a mince was made in the presence of a small amount of ice-col~ Krebs.
After the tissue was washed in buffer and drained, 5 to 6 gr!1ID aliquots were placed in 50 ml beakers followed immediately by sufficient Krebs buffer to bring volume to 20 ml. After adding Student's "t" Test: Linear Re gress~0 n: y .::: The influence of reutilization of leucine on the apparent turnover of prolyl hydroxylase was calculated by the method of Poole (1971). Stepwise integration of the differential equation was obtained using a nonlinear regression program developed by Metzler et al . . (1974).

RESULTS
This investigation examined the turnover relationship of prolyl hydroxylase to an enzymatically inactive immunologically cross-reacting protein (CRP) found in all tissue.
Prolyl hydroxylase and CRP were purifi e d from the skin of 3 to 8 day rabbits. It was found that the prolyl hydroxylase activity of neonatal skin did not change significantly over this time period. Hydroxylase activity was determined to be + 81.l-7.3 cpm/ug protein (~ S.E., N=l3) in the 20,000 x g homogenate supernates. CRP, as assayed in the same enzyme + units, was 320.4-2.19 cpm/ug protein whil e total antigen + was 412.2-20.9 cpm/ug protein. Thus, in this study, as in earlier reports (Stassen et al., 1974;Tuderman et al., 1975), CRP levels were much higher than prolyl hydroxylase levels.
The recovery of total antigen and prolyl hydroxylase activity during purification of enzyme and CRP are shown in Table I  Enzyme was purified from the resuspended, dialyzed 70% pellets by affinity chromatography. The purity of the enzyme preparations was examined by disc electrophoresis on polyacrylamide gels (Davis, 1966). In each of the enzyme preparations, only a single band was seen which contained the enzyme activity. After incubation with urea and mercaptoethanol, two smaller molecular weight bands were obtained which correspond to the two monomeric subunits . of prolyl hydroxylase (Berg and Prockop, 1973). Immunoassay of enzyme preparations before and after dissociation into subunits showed no change in immuno reactivity. For amino acid analysis, the affinity purified enzyme preparations were freed from contaminating (Pro-Gly-Pro)n by Sephadex chromatography.
The amino acid composition of rabbit prolyl hydroxylase is ·very similar to that of human prolyl hydroxylase. Both contain a high proportion of qCidic amino acids (Table II) .
In preliminary experiments which attempted to determine  was removed, fractionated, and antibody was added to the dialyzed 70% supernatant (10 mls) and to the resuspended~ dialyzed 70% pellet (10 mls).
The resulting immunoprecipitates were harvested, . dissociated with urea and mercaptoethanol, and electrophoresed on 7.5% polyacrylamide disc gels. In both preparations 2 bands were seen of equal intensity that corresponded to the 2 subunits of prolyl hydroxylase. The gels were then sliced and the radioactivity determined. As can be seen in Figure I, the monomers derived from the 70% supernatant appear to be of much higher specific radioactivity than those from the 70% pellet. Similar experiments using short labeling periods ( < 2 hours) substantiated these results.
A more extensive study was undertaken to determine whether the differences in incorporation of label into these two proteins could be explained by differences in turnover.
The turnover rates of both CRP and prolyl hydroxylase were measured by . the decay of radioactivity from these protein species.
Prolyl hydroxylase and CRP were purified from neonatal rabbit skin at various time periods following the injection of ·5 mCi/60g of 3 H~leucine. Figure  (S) and from the 70%(NH4)zSQ4 pellet (P) of neonatal rabbit skin 2 hours · post injection with 2mCi 3H-leucine. Inununoprecipitates were prepared as described in the text, dissociated and electrophoresed on 7.5% polyacrylamide gels. The gels were then sliced and the radioactivity determined.  To determine the rate of turnover, the decay of radioactivity in a given protein is assumed to b~ first order and can be ~xpressed by ·~~(t) = -kP(t) where P(t) is the specific Specific ~adioactivity of leucine in skin prolyl hydroxylase as a function of time following the administration of 3H-leucine. Neonatal rabbits were injected with 5mCi/60g of 3 H-leucine and killed at the times indicated. The skins were removed, fractionated and prolyl hydroxylase was purified by affinity chromatography. The solid line represents the best fit line from a computer regression program. The dotted line represents the solution to the equation which corrects for reutilization. w V1 radioactivity in the protein and k is its rate of destruction.
Assuming this equation to be accurate, the points in Figure   III were regressed by computer from 12 to i32 hours and the solid line, repr~senting a T~ of 73 hours, was plotted. This was significant at the P<.OOS l~vel. Poole (1971), however, has suggested that in these types of labeling experiments there is a significant amount of reutilization of precursor and that the appropriate differential equation is ~~(t) = k {F(t)-P(tl._7 where P(t) represents the specific radioactivity of precursor in the protein and F(t) represents the specific radioactivity for the pool of the precursor.
In order to correct for reutilization, free leucine activity was measured in the same experiment. Figure Figure IV. Specific radioactivity of free skin leucine as a function of time following the administration of 3H-leucine (SmCi/60g). Aliquots (0.4ml) were taken from the 20,000 x g rabbit skin homogenate supernatants and the protein precipitated with. 50% TCA. The samples were centrifuged and the supernatants harvested. Each sample was placed on a Dowex 50W-8X column and subsequently eluted with lON NH40H. The radioactive fractions were pooled, lyophilized and resuspended in lithium citrate buffer. Tritium content was measured and leucine content was determined by amino acid analysis. _Figure V. Antibody precipitation curve where increasing amounts of antisera were added to a constant amount of CRP solution. Each point represents the mean of duplicate samples. Neonatal rabbits labeled with SmCi of 3H-leucine were killed after 2 hr and 70% supernatants prepared from the skin. The supernatants were then pooled, dialyzed, and concentrated to a uniform concentration of CRP (SxlOS CPM of enzyme related antigen/ml). Five ml aliquots were taken and antibody added. The resulting precipitates were collected and washed by centrifugation, collected on filters and the tritium content determined.   . Figure  Immunoprecipitates of CRP were prepared from concentrated, dialyzed 70% (NH4)2S04 fractions derived from the ~kins of neonatal rabbits. The imrnunoprecipitates were washed, collected on filters and the radioactivity determined.
(Best fit line from computer regressicnprograrn.) In other gels not shown, this band was more distinct. ' This -protein did not correspond to any of the subunits of prolyl hydroxylase but it was antigenic in the enzyme immunoassay.
The amount of radioactivity present in this peak was much less than the two major peaks.
The SOS gels were calibrated using known molecular  The two major subunits of CRP and prolyl hydroxylase migrated with a mobility corresponding to molecular weights of 65,000 and 60,000. These values are close to what has been reported previously for the monomers of human and chick prolyl hydroxylases (Berg and Prockop, 1973;Tuderman et al., 1975). The third antigenic component of CRP was shown to have a molecular weight of 45,000.
Radioactivity from the dissociated i:mmunoprecipitate which migrated with molecular weight hydroxylase was used to determine the turnover of CRP. The concentrated dialyzed 70% (NH 4 ) 2 so 4 supernatants were derived from the same skins from which prolyl hydroxylase was purified (see Figure III) .
Each sample was adjusted to a uniform concentration of CRP (5xl0 5 CPM of enzyme related antigen/ml) and 2 mls of antibody was added to 5 mls of CRP solution to give maximum precipitation of CRP-antibody complex. This precipitate was harvested and washed twice by centrifugation. The samples were then dissociated and electrophoresed on the 10% SDS gels. For turnover analysis, the radioactivity in the higher molecular weight bands of CRP were pooled (see Figure   VIID) and background counts were subtracted. Figure  Immunoprecipitates of CRP were prepared from con-centrated, dialyzed 70% ammonium sulfate fractions derived from the skins of neonatal rabbits. The immunoprecipitates were dissociated and electrophor sed on 10% SDS-acrylamide gels. The gels were then sliced and the radioactivity determined. Counts in the two major band of CRP were pooled and background substracted.
(Best fit line from computer regression program.) though this is slightly longer than that determined with the use of filters, it is still significantly shorter than that found for prolyl hydroxylase without correction for reutili-  (Kuutti et al., 1975;Berg and Prockop, 1973) . Amino acid analysis bf purified rabbit hydroxylase also demonstrated a close homology to chick and human hydroxylase (Kuutti et al., 1975;Tuderman et al., 1975a).
The CRP levels in the neonatal rabbit skin were found to be four times that of active enzyme. Stassen et al. (1974) showed that this same ratio of CRP to active enzyme was present in neonatal mouse skin. In newborn rat skin the ratio · was approximately 20 to 1. In the immunoassay used in this investigation, no change in total immunoreactive protein was observed with enzyme before or after dissociation by mercaptoethanol. The high levels of CRP present in tissue as comp~red to active enzyme is not the result of increased immunogenicity . upon breakdown of the tetrameric form of the enzyme.
Durinq the purification of prolvl hvdroxvlase a laroe Proportion of the total CRP pool was found in the 70% (NH 4 ) 2 so 4 52 supernatant while all active enzyme was precipitated. It should be noted that some CRP is also found in the 70% pellet (Stassen et al., 1974).  Thus, these data would support the hypothesis that a proportion of the CRP in cells is precursor to active enzyme. This is also supported by the rapid appearance of label in CRP as compared to prolyl hydroxylase.
The Kd's obtained from the fall of elevated prolyl hydroxylase activity in adult rabbit aorta and liver are similar to the corrected values obtained in the labeling experiment using neonatal rabbit skin (see Table IV). The turnover rates determined for prolyl hydroxylase in these experiments were _l. 7-1. 9 days which are similar to those determined for the proteins of the endoplasmic reticulum as a whole (Arias et al., 1969). Thus, prolyl hydroxylase appears to turn over at the same rate at which the membrane is being replaced. However, the Kd for CRP in the adult rabbit aorta was found to be similar to that of prolyl hydroxylase. Again, in this tissue, the total amount of CRP determin e d cannot b e accounted for by the 55 breakdown of active enzyme. The fact that the T~s for both CRP and prolyl hydroxylase were the same is significant. The discrepancy in CRP T!2 observed betwe e n the two biological systems may be related to the fact that in the arteriosclerotic animals we measured total CRP in the homogenates, not just the 70% supernatant CRP, as in the labeling experiment.
The possibility exists that a portion of the CRP pool is de.gradation production of prolyl hydroxylase. Stassen et al. (1974) were able to s e parate both CRP a nd prolyl hydroxylase from neonatal skin by gel filtration and ion-exchange chromatography, but the CRP peaks obtained all had shoulders. In the present study, several molecular weight species were identified in the CRP pool. Only the higher molecular weight subunits are found in active enzyme. Thus, the CRP pool as obtained by immunoprecipitation in thi.s study contains unidentified heterogeneity.
It is possible that the difference between CRP and prolyl hydroxylase in the time required for maximum label incorporation, is a reflection of the difference in the turnover · rates of these two proteins. Poole (1971) has shown that proteins with long T~s require a longer period of time to reach maximum label incorporation than proteins which turn over at a more rapid rate. A rapid initial d e crease of the incorporated radioactivity from liver microsome s has been report e d (Arias, 1969). Negishi and Omura (1972) demonstrated that this biphasic decay is a property of the microsomal membrane proteins themselves. They have suggested that the rapid loss of a considerable proportion of newly synthesized proteins from the microsomes, after their association with membrane, reflects the mechanism by which proteins are inserted in the membrane. Individual proteins of the microsomal membrane such as NADPH-cytochrome-c-reductase (Negishi and Omura, 1972) and cytochrome P-450 (Levin and Kuntzman, 1969) have also been shown to have this biphasic decay curve after administration of radioactive precursors.
The observation that prolyl hydroxylase, being a microsomal protein, does not exhibit a fast initial decline in specific radioactivity during the first few hours after label administration, and no evidence of biphasic decay, may have functional · importance. The rapid turnover of CRP present in the •70% . (NH 4 ) 2 so 4 supernatant may, however, be analogous to the fast decay phase of other microsomal proteins. This observation, when considered with the lack of delay in incorporation of label into CRP, may thus be relevant to the mechanism by which newly synthesized CRP or prolyl hydroxylase tetramers are incorporated into or onto the microsomal membrane as functional enzyme units. If CRP exists in cells as a cytoplasmic protein its turnover, compared to enzyme, could be predicted to be more rapid if membrane or ribosomal protein are degraded primarily as free cytoplasmic protein, as suggested by Dice and Schimke (1972).
Epinephrine-thyroxine treatment raised prolyl hydroxylase activity in both the aorta and the liver. Increased prolyl hydroxylase activity can be seen after five days of this treatment, with a redoubling after each subsequent 5 days of treatment, which is prior to the appearance of fibrous plaques (Fuller and Langner, 1970). CRP was also significantly elevated in the aorta. This finding corresponds to results previously reported (Fuller et al., 1976), and suppor~s the hypothesis that increases in prolyl hydroxylase activity "in vivo" are the result of new synthesis, with facilitiation of the conversion of CRP into prolyl hydroxylase as the active enzyme is inserted onto the microsomal membrane.
Comparison of enzyme decay curves with those for increased activity "in vivo" also suggest that active enzyme levels may ' be elevated by inhibition of prolyl hydroxylase degradation.
·Large increases in prolyl hydroxylcse activity have been reported in a variety of tissues responding to injury-induced damage (Hussain et al., 1976;Langner and Fuller, 1970;Mussini et al., 1967). These increases in enzyme activity occur over a period of days and are compatible with the slow turnover of prolyl hydroxylase. This stabilization of prolyl hydroxylase within the microsomes, leading to elevated enzyme levels, could not be the same as ascorbate activiation observed in tissue culture, which is maximal within hours after administration (Stassen et al., 1973) and appears to be the result of the dissociation of enzyme-substrate complexes (Kuttan et al., 1975).

1)
Rabbit skin prolyl hydroxylase was purified to homogeneity by affinity chromatography as judged by polyacrylamide electrophoresis. Molecular weight determination and amino acid analysis showed close homology to other vertebrate prolyl hydroxylases.
2) The turnover rates of prolyl hydroxylase and CRP were measured and the values obtained for active enzym~ were very similar in all tissues examined. The T~ for prolyl hydroxylase was found to be 38 hours in the adult rabbit liver and 42 hours in the adult rabbit aorta. Using neonatal skin, a corrected half-life of 45 hours was determined for the enzyme. In comparison, the apparent CRP T~ was significantly shorter than the apparent T~ of prolyl hydroxylase in the neonatal ~kin.
No difference was seen between CRP and prolyl hydroxylase turnover in the adult rabbit aorta. 3) The rates of synthesis of CRP in the adult aorta and neonatal skin were very close. In the aorta, CRP synthesis was 40 tim~s that of active enzyme synthesis. Using uncorrected data, the rate of synthesis of CRP in neonatal skin was approximately 6 times that of prolyl hydroxylase. thus, the total amount of CRP in tissue cannot be accounted 60 fo r by breakdown of th e e nzyme. The data suppor ts t he i dea that a proportion of the CRP pool is precur sor t o ac tive enzyme.

4)
I ·1 neonatal ski n , maxim um label incor po at ion .into CRP was c~J se rved and occurred 11 a much short e r p erio d of time e:.::impared to pre~ yl hyd :::-o x ylas e. Al tho ugh this d ifference

6)
In all tissues examined, CRP levels were much higher than those of prolyl hydroxylase. These high levels of CRP Present in tissue are not the result of increased immunegenicity upon breakdown of the enzyme.