sub-threshold levels of chemicals. They may also differ in the level of one or two major inducers as well as in nutrient composition, which can similarly affect MFO activity. These authors have drawn the important conclusion that the presence of low levels of several chemicals in a diet, even though independently they are found to be without toxicological significance, may nevertheless be additive. 0 1. A. H. Conney: Pharmacological Implications of Microsomal Enzyme Induction. Pharmacol. Revs. 19: 317-366, 1967 2. L. W. Wattenberg: Effects of Dietary Constituents on the Metabolism of Chemical Carcinogens. Cancer Res. 35: 3326-3331, 1975 3. L. G. Hart and J. R. Fouts: Effects of Acute and Chronic DDT Administration on Hepatic Microsomal Drug Metabolism in the Rat. Proc. Soc. Exp. Biol. Med. 114: 388392, 1963 4. T. C. Campbell and J. R. Hayes: Role of Nutrition in the Drug Metabolizing Enzyme System. Pharmacol. Revs. 26: 171- 197, 1974
5. F. P. Guengerich: Separation and Purification of Multiple Forms of Microsomal Cytochrome P-450. Activities of Different Forms of Cytochrome P-450 Towards Several Compounds of Environmental Interest. J. Biol. Chem. 252: 3970-3979, 1977 6. L. G. Hart and J. R. Fouts: Studies of the Possible Mechanisms by which Chlordane Stimulates Hepatic Microsomal Drug Metabolism in the Rat. Biochem. Pharmacol. 14: 263-272, 1965 7. J. R. Gillette: Factors that Affect the Stimulation of the Microsomal Drug Enzymes Induced by Foreign Compounds. Adv. Eng. Regulation 1: 215223, 1963 8. D. Gilbert and L. Golberg: Liver Response Test. 111. Liver Enlargement and Stimulation of Microsomal Processing Enzyme Activity. Food Cosmet. Toxicd. 3: 417-432, 1965 9. A.E.M. McLean and H. E. Driver: Combined Effects of Low Doses of DDT and Phenobarbital on Cytochrome P450 and Amidopyrine Demethylation. Biochem. Pharm. 26: 1299-1302, 1977
THE FUNCTION OF ASCORBIC ACID IN COLLAGEN FORMATION Ascorbic acid is essential for secretion of collagen from collagensynthesizing chick embryo fibroblasts. It acts as a cofactor for hydroxylationof peptide-boundproline; unhydroxylated collagen cannot be secreted.
Key Words: ascorbic acid, collagen, proline, hydroxyproline, prolyl hydroxylase, fibroblasts, a,a'dipyridyl
In 1965 Peterkofsky and Udenfriendl established that ascorbate acts as a cofactor for the enzyme collagen-prdyl hydroxylase. This enzyme hydroxylates peptide-bound proline to hydroxyproline in the formation of collagen. Subsequently, two other functions for ascor110
NUTRITION REVIEWS I VOL. 36. NO. 4 APRIL 1978
bate were postulated in connection with collagen formation: (a) the stimulation of secretion of collagen from the cells in which it is made;* and (b) the activation of the enzyme prdyl hydroxylase.3 Recent work proves that both these additional functions are consequences of the cofactor function. A recent paper by Blanck%nd Peterkofsky4 addresses itself to the question of secretion: using chick embryo fibroblasts in culture at
log-phase of growth, they studied collagen reached 30 percent, after which there was a formation and secretion into the culture me- rapid, straight-line rise of secretion with hydium by measuring uptake of labeled prdine droxylation, showing direct proportionality into collagenase-degradable protein. Intra- between hydroxylation of collagen-proline cellular collagen synthesis ceased in absence and collagen secretion into the medium. There of ascorbate after two hours, no doubt because was, therefore, a direct dependence of secrethe unhydroxylatedcollagen made was rapidly tion on hydroxylation after 30 percent of prodegraded (it should be noted that in all the line residuas had been hydroxylated. This work here reported, “collagen” refers to un- represents60percentof all the prdine residues hydroxylated as well as fully hydroxylated col- available for hydroxylation. It is important to lagen). Collagen formation continued for four note that a minimum of 35 percent hydroxyhours in cells in presence of ascorbate. The lation is needed if the collagen molecule is to presence of 1 x 10-SM ascorbate greatly stim- aggregate into the triple-helix configuration,s ulated secretion of collagen into the medium. the only way in which collagen can be secThere was, of course, no effect of ascorbate reted. Therefore, the sequence of events is: on noncollagen protein secretion. Next, the 1) synthesis of unhydroxylatedsingle polypepauthors investigated how ascorbateconcentra- tide chains; 2) their hydroxylation, dependent tion would affect prdine hydroxylation on the on a minimum concentration of ascorbate and one hand, and collagen secretion on the other. inhibited by a,a’-dipyridyl; 3) aggregation of They determined hydroxylation by assay of three hydroxylated polypeptide chains into labeled hydroxyproline in hydrdysates of cd- the triple helix of collagen; and 4) secretion of lagenasedigests, after incubation of their cells the triple-helical collagen. Hence, secretion is with labeled proline. They found a low level dependent on hydroxylation, which in turn deof hydroxylation, both in intracellular and me- pends on ascorbate. dium collagen, at low levels of ascorbic acid. It was noted by the authors that the apparent With increasing ascorbate concentration, a Km of ascorbate with respect to hydroxylation sudden rise occurred at about 5 x 10-7M as- in their fibroblast cells was 700-fold lower than that for pure prdyl hydroxylaseenzyme. Using corbate. Similarly, collagen secretion into the medium was low at a low level of ascorbate labeled ascorbate, they determined that and showed an abrupt rise at about the same fibroblasts could concentrate the vitamin ten concentration of ascorbate as hydroxylation fold over that in the medium. This would still (Kas for hydroxylation and secretion, 4.58 x leave a 70-fold discrepancy, and the authors 10-7M). Therefore, both hydroxylation and speculate that ascorbate may be further consecretion appear to require the same level of centrated within a subcellular compartment, ascorbate. If hydroxylation was inhibited by possibly the endoplasmic reticulum, where it other means than lack of ascorbate, such as functions as a cofactor for prdyl hydroxylase. by the presence of the hydroxylation inhibitor The second postulatednon-cofactorfunction a,a’-dipyridyl, at increasing levels (in presence for ascorbate is the activation of collagenprolyl hydroxylase. Stassen, Cardinale and of adequate ascorbate, of course), again hydroxylation and secretion went hand in hand:, Udenfriend,B using mouse skin fibroblasts in both declined abruptly at the same concentra- culture, and an antibody to the enzyme prdyl hydroxylase, showedthat these cells contained tion of dipyridyl (Kdipy for hydroxylation and secretion, 5.6 x 10-5). There was, therefore, a proteinwhich cross-reactedwith the antibody a direct relationship between collagen-proline but had no enzyme activity. They found that, with fibroblasts in the log-phase of growth, ashydroxylation and collagen secretion. When the above data were combined and corbate greatly activated the enzyme prdyl graphed as collagen secreted against total hydroxylase, as it seemed then, by causing proline hydroxylated, it was found that a low %ggregation of inactive subunits of the enzyme level of secretion occurred until intracellular hy- (the cross-reacting protein) into an active endroxylation of the collagen molecules had zyme. More recent reports from the same labNUTRITION REVIEWS I VOL. 36, NO. 4 APRIL 1971)
119
oratory’ have come to a different condusion. It is now known that the enzyme prdyl hydroxylase is a tetramer, consisting of two pairs of non-identical subunits, one pair of molecular weight 64,000, the other pair of 60,OOO.The larger subunit contains a carbohydrate moiety. The authors, using their antibody to the active (tetrameric)enzyme, found an excess of crossreacting protein without enzymatic activity present in many tissues. They purified this enzymatically inactive protein from 800 newborn rat skins by affinity chromatography, using their antibody bound to a Sepharose column, and subsequent gel filtration. For purposes of comparison, they dissociated the active (tetrameric) enzyme into the large and small subunits, and then showed by gel electrophoresis, amino acid analyses and immunodiffusion experiments, that the inactive, cross-reacting protein present in many tissues is identicalwith the small subunit of the enzyme. They suggest that this subunit is always present in excess and combines with newly-synthesized large subunits to form the active (tetrameric) enzyme, whenever hydroxylation of collagen is needed. Another paper by the same group8 shows that, contrary to what had been suggested earlier396 the enzymatically inactive, crossreacting smaller enzyme subunit cannot be activated by ascorbate, nor does ascorbate cause its aggregation with the larger subunits to yield the active tetramer. The authors found, however, that there existed an inactive form of the tetramer which could be extracted from mouse skin fibroblasts, and could be activated many-fold by ascorbate. Not only ascorbate, but the other prolyl hydroxylation cofactors were required for this activation of the enzyme, e.g., Fe++ and ketoglutarate. They suggest that, apart from the inactive small subunit of the enzyme, and the enzymatically active tetramer, there is present in these fibroblasts a form of the enzyme in which the substrate (i.e., unhydroxylatedcollagen) is tightly bound to the enzyme (in tetrameric form), thus inactivating the enzyme and that the unhydroxylated collagen substrate has a high affinity for the enzyme. When converted to the hydroxylated product by prdyl hydroxylation, the affin-
ity is reduced and the enzyme can then react with more substrate. Therefore, ascorbate reacts merely as a cofactor for prdyl hydroxylase; it thereby increases the available active enzyme, by converting substrate bound to the enzyme into unbound product and free enzyme. The activation of prdyl hydroxylase in log-phase mouse skin fibroblasts by ascorbate was confirmed by Blanck and Peterkofsky,4 and by Miller.9 Recently, however, Berg et a1.10 reported that they were unable to activate prolyl hydroxylase by ascorbate in these same cells. They ascribe the previously reported activation to an inaccuracy in the assay for prolyl hydroxylase used by all the researchers quoted above who found activatable enzyme. These workers used as substrate for their assay unhydroxylated collagen in which the praline residues were labeled with 3H. Upon hydroxylation by the enzyme, 3H2O release is determined as a measure of enzyme activity. Since the amount of added labeled unhydroxylated collagen is very small, it would become diluted with endogenousunlabeledunhydroxylated collagen, thus lowering its spe afic radioactivity and consequently that of the 3H20 released. Thus, a falsely low assay is obtained. This would not happenif, in presence of ascorbate, all the endogenous unlabeledunhydroxylated collagen was fully hydroxylated. Berg et aI.10 use a different assay: they add an excess of a synthetic polypeptide substrate, and measure hydroxylation chemically, rather than radiometrically.They daim that this assay gives an accurate measure of prdyl hydroxyiase activity. With this assay, they found no activation of the enzyme by ascorbate. In summary, whichever way the controversy concerning the activatability of prdyl hydroxylase by ascorbate is resolved, it seems dear that aswrbate has only one function in collagen formation: to act as cofactor for prdyl hydroxylase. The mechanism of this cofactor function is, of course, still shrouded in mystery and awaits further research. 0 1.
B. Peterkofsky and S. Udenfriend: Enzymatic Hydroxylation of Proline ipMicrosomal Polypeptide Leading to Formation of Collagen. Proc. Nat. Aced. Sci. USA 53: 335342, 1965
2 B. Peterkofsky: Regulation of Collagen Secretion by Ascorbic Acid in 3T3 and Chick Embryo Fibroblasts. Biochem. Biophys. Res. Commun. 49: 1343-1350, 1972
7.
3. Activation of Prolyl Hydroxylase by Ascorbic Acid. NutMon Reviews 31: 255257, 1.973 4. T.J.J. Blanck and B. Peterkofsky: The Stimulation of Collagen Secretion by Ascorbate as a Result of Increased Proline Hydroxylation in Chick Embryo Fibroblasts. Arch. Biochem. Biophys. 171: 259-267, 1975
5. J. Rosenbloom, M. Harsch and S. Jimenez: Hydroxyproline Content Determines the Denaturation Temperature of Chick Tendon Collagen. Arch. Biochem. Biophys. 158: 478-484, 1973 6. F.L.H. Stassen, G.J. Cardinale and S. Udenfriend: Activation of Prolyl Hydroxylase in
8.
9.
10.
L-929 Fibroblasts by Ascorbic Acid. Proc. Nat. Acad. Sci. USA 70: 1090-1093, 1973 S. Chen-Kiang, G.J. Cardinale and S. Udenfriend: Homology Between a Prolyl Hydroxylase Subunit and a Tissue Protein that Crossreacts Immunologically with the Enzyme. Proc. Nat. Acad. Sci. USA 74: 4420-4424, 1977 R. Kuttan, G.J. Cardinale and S. Udenfriend: An Activatable Form of Prolyl Hydroxylase in Fibroblast Extracts. Biochem. Biophys. Res. Commun. 64: 947-954, 1975 R.L. Miller: The Effect of Ascorbic Acid on Lysyl and Prolyl Hydroxylase Activity of Cultured Fibroblasts. Arch. Biochem. Biophys. 170: 341344, 1975 R.A. Berg, W. Whei-Yang Kao and D.J. Procklop: Prolyl Hydroxylase Activity in L-929 Fibroblasts IncubatedWith and Without Ascorbate. Biochem. Biophys. Acta 444: 756764, 1976
NUTRITION RIVIEWS IVOL. 36,NO. 4 APRIL 1978
121
5. F. P. Guengerich: Separation and Purification of Multiple Forms of Microsomal Cytochrome P-450. Activities of Different Forms of Cytochrome P-450 Towards Several Compounds of Environmental Interest. J. Biol. Chem. 252: 3970-3979, 1977 6. L. G. Hart and J. R. Fouts: Studies of the Possible Mechanisms by which Chlordane Stimulates Hepatic Microsomal Drug Metabolism in the Rat. Biochem. Pharmacol. 14: 263-272, 1965 7. J. R. Gillette: Factors that Affect the Stimulation of the Microsomal Drug Enzymes Induced by Foreign Compounds. Adv. Eng. Regulation 1: 215223, 1963 8. D. Gilbert and L. Golberg: Liver Response Test. 111. Liver Enlargement and Stimulation of Microsomal Processing Enzyme Activity. Food Cosmet. Toxicd. 3: 417-432, 1965 9. A.E.M. McLean and H. E. Driver: Combined Effects of Low Doses of DDT and Phenobarbital on Cytochrome P450 and Amidopyrine Demethylation. Biochem. Pharm. 26: 1299-1302, 1977
THE FUNCTION OF ASCORBIC ACID IN COLLAGEN FORMATION Ascorbic acid is essential for secretion of collagen from collagensynthesizing chick embryo fibroblasts. It acts as a cofactor for hydroxylationof peptide-boundproline; unhydroxylated collagen cannot be secreted.
Key Words: ascorbic acid, collagen, proline, hydroxyproline, prolyl hydroxylase, fibroblasts, a,a'dipyridyl
In 1965 Peterkofsky and Udenfriendl established that ascorbate acts as a cofactor for the enzyme collagen-prdyl hydroxylase. This enzyme hydroxylates peptide-bound proline to hydroxyproline in the formation of collagen. Subsequently, two other functions for ascor110
NUTRITION REVIEWS I VOL. 36. NO. 4 APRIL 1978
bate were postulated in connection with collagen formation: (a) the stimulation of secretion of collagen from the cells in which it is made;* and (b) the activation of the enzyme prdyl hydroxylase.3 Recent work proves that both these additional functions are consequences of the cofactor function. A recent paper by Blanck%nd Peterkofsky4 addresses itself to the question of secretion: using chick embryo fibroblasts in culture at
log-phase of growth, they studied collagen reached 30 percent, after which there was a formation and secretion into the culture me- rapid, straight-line rise of secretion with hydium by measuring uptake of labeled prdine droxylation, showing direct proportionality into collagenase-degradable protein. Intra- between hydroxylation of collagen-proline cellular collagen synthesis ceased in absence and collagen secretion into the medium. There of ascorbate after two hours, no doubt because was, therefore, a direct dependence of secrethe unhydroxylatedcollagen made was rapidly tion on hydroxylation after 30 percent of prodegraded (it should be noted that in all the line residuas had been hydroxylated. This work here reported, “collagen” refers to un- represents60percentof all the prdine residues hydroxylated as well as fully hydroxylated col- available for hydroxylation. It is important to lagen). Collagen formation continued for four note that a minimum of 35 percent hydroxyhours in cells in presence of ascorbate. The lation is needed if the collagen molecule is to presence of 1 x 10-SM ascorbate greatly stim- aggregate into the triple-helix configuration,s ulated secretion of collagen into the medium. the only way in which collagen can be secThere was, of course, no effect of ascorbate reted. Therefore, the sequence of events is: on noncollagen protein secretion. Next, the 1) synthesis of unhydroxylatedsingle polypepauthors investigated how ascorbateconcentra- tide chains; 2) their hydroxylation, dependent tion would affect prdine hydroxylation on the on a minimum concentration of ascorbate and one hand, and collagen secretion on the other. inhibited by a,a’-dipyridyl; 3) aggregation of They determined hydroxylation by assay of three hydroxylated polypeptide chains into labeled hydroxyproline in hydrdysates of cd- the triple helix of collagen; and 4) secretion of lagenasedigests, after incubation of their cells the triple-helical collagen. Hence, secretion is with labeled proline. They found a low level dependent on hydroxylation, which in turn deof hydroxylation, both in intracellular and me- pends on ascorbate. dium collagen, at low levels of ascorbic acid. It was noted by the authors that the apparent With increasing ascorbate concentration, a Km of ascorbate with respect to hydroxylation sudden rise occurred at about 5 x 10-7M as- in their fibroblast cells was 700-fold lower than that for pure prdyl hydroxylaseenzyme. Using corbate. Similarly, collagen secretion into the medium was low at a low level of ascorbate labeled ascorbate, they determined that and showed an abrupt rise at about the same fibroblasts could concentrate the vitamin ten concentration of ascorbate as hydroxylation fold over that in the medium. This would still (Kas for hydroxylation and secretion, 4.58 x leave a 70-fold discrepancy, and the authors 10-7M). Therefore, both hydroxylation and speculate that ascorbate may be further consecretion appear to require the same level of centrated within a subcellular compartment, ascorbate. If hydroxylation was inhibited by possibly the endoplasmic reticulum, where it other means than lack of ascorbate, such as functions as a cofactor for prdyl hydroxylase. by the presence of the hydroxylation inhibitor The second postulatednon-cofactorfunction a,a’-dipyridyl, at increasing levels (in presence for ascorbate is the activation of collagenprolyl hydroxylase. Stassen, Cardinale and of adequate ascorbate, of course), again hydroxylation and secretion went hand in hand:, Udenfriend,B using mouse skin fibroblasts in both declined abruptly at the same concentra- culture, and an antibody to the enzyme prdyl hydroxylase, showedthat these cells contained tion of dipyridyl (Kdipy for hydroxylation and secretion, 5.6 x 10-5). There was, therefore, a proteinwhich cross-reactedwith the antibody a direct relationship between collagen-proline but had no enzyme activity. They found that, with fibroblasts in the log-phase of growth, ashydroxylation and collagen secretion. When the above data were combined and corbate greatly activated the enzyme prdyl graphed as collagen secreted against total hydroxylase, as it seemed then, by causing proline hydroxylated, it was found that a low %ggregation of inactive subunits of the enzyme level of secretion occurred until intracellular hy- (the cross-reacting protein) into an active endroxylation of the collagen molecules had zyme. More recent reports from the same labNUTRITION REVIEWS I VOL. 36, NO. 4 APRIL 1971)
119
oratory’ have come to a different condusion. It is now known that the enzyme prdyl hydroxylase is a tetramer, consisting of two pairs of non-identical subunits, one pair of molecular weight 64,000, the other pair of 60,OOO.The larger subunit contains a carbohydrate moiety. The authors, using their antibody to the active (tetrameric)enzyme, found an excess of crossreacting protein without enzymatic activity present in many tissues. They purified this enzymatically inactive protein from 800 newborn rat skins by affinity chromatography, using their antibody bound to a Sepharose column, and subsequent gel filtration. For purposes of comparison, they dissociated the active (tetrameric) enzyme into the large and small subunits, and then showed by gel electrophoresis, amino acid analyses and immunodiffusion experiments, that the inactive, cross-reacting protein present in many tissues is identicalwith the small subunit of the enzyme. They suggest that this subunit is always present in excess and combines with newly-synthesized large subunits to form the active (tetrameric) enzyme, whenever hydroxylation of collagen is needed. Another paper by the same group8 shows that, contrary to what had been suggested earlier396 the enzymatically inactive, crossreacting smaller enzyme subunit cannot be activated by ascorbate, nor does ascorbate cause its aggregation with the larger subunits to yield the active tetramer. The authors found, however, that there existed an inactive form of the tetramer which could be extracted from mouse skin fibroblasts, and could be activated many-fold by ascorbate. Not only ascorbate, but the other prolyl hydroxylation cofactors were required for this activation of the enzyme, e.g., Fe++ and ketoglutarate. They suggest that, apart from the inactive small subunit of the enzyme, and the enzymatically active tetramer, there is present in these fibroblasts a form of the enzyme in which the substrate (i.e., unhydroxylatedcollagen) is tightly bound to the enzyme (in tetrameric form), thus inactivating the enzyme and that the unhydroxylated collagen substrate has a high affinity for the enzyme. When converted to the hydroxylated product by prdyl hydroxylation, the affin-
ity is reduced and the enzyme can then react with more substrate. Therefore, ascorbate reacts merely as a cofactor for prdyl hydroxylase; it thereby increases the available active enzyme, by converting substrate bound to the enzyme into unbound product and free enzyme. The activation of prdyl hydroxylase in log-phase mouse skin fibroblasts by ascorbate was confirmed by Blanck and Peterkofsky,4 and by Miller.9 Recently, however, Berg et a1.10 reported that they were unable to activate prolyl hydroxylase by ascorbate in these same cells. They ascribe the previously reported activation to an inaccuracy in the assay for prolyl hydroxylase used by all the researchers quoted above who found activatable enzyme. These workers used as substrate for their assay unhydroxylated collagen in which the praline residues were labeled with 3H. Upon hydroxylation by the enzyme, 3H2O release is determined as a measure of enzyme activity. Since the amount of added labeled unhydroxylated collagen is very small, it would become diluted with endogenousunlabeledunhydroxylated collagen, thus lowering its spe afic radioactivity and consequently that of the 3H20 released. Thus, a falsely low assay is obtained. This would not happenif, in presence of ascorbate, all the endogenous unlabeledunhydroxylated collagen was fully hydroxylated. Berg et aI.10 use a different assay: they add an excess of a synthetic polypeptide substrate, and measure hydroxylation chemically, rather than radiometrically.They daim that this assay gives an accurate measure of prdyl hydroxyiase activity. With this assay, they found no activation of the enzyme by ascorbate. In summary, whichever way the controversy concerning the activatability of prdyl hydroxylase by ascorbate is resolved, it seems dear that aswrbate has only one function in collagen formation: to act as cofactor for prdyl hydroxylase. The mechanism of this cofactor function is, of course, still shrouded in mystery and awaits further research. 0 1.
B. Peterkofsky and S. Udenfriend: Enzymatic Hydroxylation of Proline ipMicrosomal Polypeptide Leading to Formation of Collagen. Proc. Nat. Aced. Sci. USA 53: 335342, 1965
2 B. Peterkofsky: Regulation of Collagen Secretion by Ascorbic Acid in 3T3 and Chick Embryo Fibroblasts. Biochem. Biophys. Res. Commun. 49: 1343-1350, 1972
7.
3. Activation of Prolyl Hydroxylase by Ascorbic Acid. NutMon Reviews 31: 255257, 1.973 4. T.J.J. Blanck and B. Peterkofsky: The Stimulation of Collagen Secretion by Ascorbate as a Result of Increased Proline Hydroxylation in Chick Embryo Fibroblasts. Arch. Biochem. Biophys. 171: 259-267, 1975
5. J. Rosenbloom, M. Harsch and S. Jimenez: Hydroxyproline Content Determines the Denaturation Temperature of Chick Tendon Collagen. Arch. Biochem. Biophys. 158: 478-484, 1973 6. F.L.H. Stassen, G.J. Cardinale and S. Udenfriend: Activation of Prolyl Hydroxylase in
8.
9.
10.
L-929 Fibroblasts by Ascorbic Acid. Proc. Nat. Acad. Sci. USA 70: 1090-1093, 1973 S. Chen-Kiang, G.J. Cardinale and S. Udenfriend: Homology Between a Prolyl Hydroxylase Subunit and a Tissue Protein that Crossreacts Immunologically with the Enzyme. Proc. Nat. Acad. Sci. USA 74: 4420-4424, 1977 R. Kuttan, G.J. Cardinale and S. Udenfriend: An Activatable Form of Prolyl Hydroxylase in Fibroblast Extracts. Biochem. Biophys. Res. Commun. 64: 947-954, 1975 R.L. Miller: The Effect of Ascorbic Acid on Lysyl and Prolyl Hydroxylase Activity of Cultured Fibroblasts. Arch. Biochem. Biophys. 170: 341344, 1975 R.A. Berg, W. Whei-Yang Kao and D.J. Procklop: Prolyl Hydroxylase Activity in L-929 Fibroblasts IncubatedWith and Without Ascorbate. Biochem. Biophys. Acta 444: 756764, 1976
NUTRITION RIVIEWS IVOL. 36,NO. 4 APRIL 1978
121
