- Email: [email protected]
Iron-dependent uptake of ascorbate into isolated microsomes
ARCHIVESOFBIOCHEMISTRY AND BIOPHYSICS Vol. 254, No. 1, April, pp. 282-289, 1987
Iron-Dependent
Uptake of Ascorbate
BEVERLY PETERKOFSKY,l Laboratory
of Biochemistry,
into Isolated Microsomes
GEORG TSCHANK:
AND
National
Bethesda, Maryland
Cancer Institute,
CHRISTINA LUEDKE3 243892
Received September 19,1986, and in revised form December 16, 1986
A preliminary study (J. M. Mata, R. Assad, and B. Peterkofsky (1981) Arch. Biochem Biophvs. 206,93-104) suggested that chick embryo limb bone microsomes took up and concentrated [14C]ascorbate in the presence of cofactors for prolyl hydroxylase. In the present study, we found that the apparent Km for ascorbate in the hydroxylation of intracisternal unhydroxylated procollagen by endogenous prolyl hydroxylase was approximately an order of magnitude less than the value obtained when enzyme solubilized from microsomes was used with an exogenous substrate. These results are compatible with a concentrative uptake of ascorbate into microsomes. The uptake of [14C]ascorbate into microsomes was confirmed and it required only iron, in either the ferrous or ferric form, and was time and temperature dependent, proportional to microsome concentration, and substrate saturable at 2-3 mM ascorbate. Iron-dependent ascorbate uptake also was observed with L-929 cell microsomes. [14C]Ascorbate seemed to be taken up without prior oxidation, since only unlabeled ascorbate, and not dehydroascorbate, competed for uptake into limb bone microsomes. A functional requirement for Fe2+ in ascorbate transport was demonstrated using the intracisternal proline hydroxylating system. L-929 cell microsomes were preincubated with ascorbate with or without the metal and then external ascorbate was oxidized to inactive dehydroascorbate using ascorbic acid oxidase, which cannot penetrate the microsomal membrane. Samples which did not receive iron during the preincubation received it, along with other requirements for prolyl hydroxylase, in a final incubation to measure hydroxylation. Significant hydroxylation was obtained only in samples incubated with iron prior to oxidase treatment, consistent with the conclusion that an iron-dependent process was required to translocate ascorbate and protect it from the oxidase. Q 1987 Academic press. I,,~.
Ascorbic acid must pass through several membrane barriers in order to participate in some cellular reactions. For example, prolyl (EC 1.14.11.2)and lysyl (EC 1.14.11.4) hydroxylases require ascorbate (1) and are localized in the rough endoplasmic reticulum (2-4) where they carry out posttranslational modifications of procollagens (4). Several other vesicular enzymes which require ascorbate in vitro are dopamine ,f3hydroxylase (EC 1.14.17.1),which catalyzes 1To whom correspondence should be addressed. ’ Present address: Laboratory of Biochemistry, Children’s Hospital, University of Mainz, Mains, FRG. ‘Laboratory of Medical Biochemistry, The Rockefeller University, New York, NY 10021. 0003-9861/87 $3.00 Copyright All rights
CC!198’7 by Academic Press. Inc. of reproduction in any form reserved.
formation of norepinephrin in chromaffin granules of the adrenal (5, 6) and vesicles of adrenergic neurons (7), and peptidylglytine a-amidating monooxygenase, which cleaves and amidates a polypeptide hormone precursor in pituitary secretory granules (8,9). Since the cofactor is acidic, it is unlikely to penetrate membranes by simple diffusion (10). Active transport across the cell membrane has been demonstrated in a variety of cell types or tissues, including chick embryo bone cells (ll), alveolar macrophages (12), adrenomedullary chromaffin cells (13), intestinal mucosa (14), and renal brush border tubules (15). These plasma membrane transport systems generally are sodium depen-
282
ASCORBATE
UPTAKE
INTO
CHICK
dent. In contrast, there are relatively few reports of ascorbate transport into subcellular organelles. The finding that a large proportion of the ascorbate in adrenal medulla was concentrated in the granule fraction suggested that there might be a transport system for ascorbate in these vesicles (16). Other studies, however, indicate that there is no direct uptake of ascorbate by isolated granules (17, 18). Aseorbate transport into isolated guinea pig mitochondria has been demonstrated, but it was suggested that the process occurred by diffusion (19). Preliminary results from our laboratory (20) indicated that isolated microsomes which contained prolyl and lysyl hydroxylases, took up and concentrated radioactive ascorbate. The uptake was temperature dependent and was stimulated by a mixture of the cosubstrate a-ketoglutarate and other factors required for optimal hydroxylase activity. In the present study, we have determined the exact requirements and other characteristics of the transport system. MATERIALS
AND
METHODS
Materials. L-[1-W]Ascorbate (7-8 &i/pmol), ‘HzO, and [4-3H]proline were purchased from New England Nuclear. Ascorbate oxidase was purchased from Boehringer-Mannheim. The sources of other chemicals have been described previously (20). Microsome preparations. The isolation and characterization of chick embryo limb bone microsomes have been described previously (2). In some cases, prior to preparation of the microsomes, the tissue was labeled with [4-‘Hlproline by the same procedures used to label frontal bones (21). L-929 cell microsome-enriched fractions, either unlabeled or containing [4aHJproline-labeled, unhydroxylated procollagen, were prepared from cells incubated in the presence of the iron chelator a,&-dipyridyl (20). [“C]Ascorbate transport assay. Microsome pellets that had been stored in liquid Nz, were thawed and pH resuspended in 0.25 M sucrose/O.05 M Tris-HCl, ‘7.6, and 40 ~1 was added to a loo-p1 assay. Microsomes stored under these conditions have been shown to have intact membranes (2, 3, 20). Conditions were essentially those described previously (20), except that the final concentration of ascorbate was 2 mM (1.0 &i/ pmol), unless otherwise indicated, the temperature was 24°C and MgzSOI and KC1 were not present. The other components were 40 mM Tris-HCI, pH 7.6, and 0.15 mM FeSO,, unless otherwise indicated. Incuba-
EMBRYO
LIMB
BONE
MICROSOMES
283
tions were carried out for 8 min and a zero-time control was used for each sample. The value from the unincubated control was subtracted from the 8-min value in order to correct for nonspecific adsorption. After the incubation, 0.9 ml of ice-cold 0.25 M sucrose/ pH 7.6/2 mM ascorbate was added, 0.05 M Tris-HCl, the samples were immediately centrifuged at 43,500g for 10 min at 4°C and the supernatant solutions were discarded. The recovered microsomes were resuspended in 0.5 ml of the same buffer and the suspension was centrifuged as before. The pellets were dissolved in 0.5 ml of 0.5% Triton X-100 and the solutions were transferred to counting vials. The tubes were rinsed with 0.5 ml of detergent which was added to the vials along with 9 ml of Hydrofluor scintillation fluid and the radioactivity was measured. All assays were performed on duplicate samples and the results presented are the averages of duplicate values corrected for unincubated controls. In most cases, results are expressed as picomoles of ascorbate taken up, which was calculated using the specific radioactivity of [i4C]ascorbate. In some cases, the specific transport activity was calculated as picomoles per microgram RNA, in order to compare experiments in which different amounts of microsomes were used. RNA, rather than protein, was used as a basis for comparison since it is found specifically in microsomal vesicles and thus eliminates the problem of varying protein contributions from other vesicles in different preparations. Generally, lo-30 pg of RNA per assay was used. RNA was measured in hot trichloroacetic acid extracts of microsomes (2,20). Proline hydroxylation in intact microscmes. Microsomes containing endogenous [4-8H]proline-labeled unhydroxylated procollagen and prolyl hydroxylase were incubated in a total volume of 100 ~1 with 0.2 mM a-ketoglutarate, 0.2 mM FeSOI, 0.1 mM thymol, 1.0 mM ascorbate, and 50 mM Tris-HCI, pH 7.6, unless otherwise indicated. The release of ‘Hz0 resulting from proline hydroxylation was measured as described previously (20,21). Soluble pro@ h&m$ase activity. Conditions were similar to those described above except that a purified preparation of [4-‘Hlproline-labeled unhydroxylated procollagen (21) was used and amounts of solubilized enzyme extracts were used which fell in the linear portion of a rate versus protein concentration curve. Preparation of cleh&oaswrbate. Ascorbate, 50 mM, was incubated with a large excess (150 units/ml) of ascorbic acid oxidase in 10 mM Tris-HCl, pH 7.6, for 15 min at 25°C. These conditions were shown to oxidize 100% of the ascorbate using either a calorimetric assay (22) or uv absorption at 265 nm to measure ascorbate. RESULTS
Our previous experiments (20) suggested that there was a temperature-dependent
284
PETERKOFSKY,
Intracisternal:
TSCHANK,
enzyme substrate
Solubilized Exogenous
Ascorbate,
enzyme substrate
mM
FIG. 1. Proline hydroxylation as a function of ascorbate concentration in chick embryo limb bone microsomes compared to solubilized prolyl hydroxylase. Assay procedures are described under Materials and Methods. Microsomes containing [4-Qlproline-labeled unhydroxylated procollagen were suspended in 0.25 M sucrose/O.05 M Tris-HCl, pH 7.6, and portions were incubated with varying concentrations of ascorbate for 20 min and assayed for proline hydroxylation by measuring8Hz0 release (e). An unlabeled microsome pellet was similarly resuspended and the suspension was frozen at -2O’C and then thawed to solubilize prolyl hydroxylase. Enzyme activity was measured with an exogenous unhydroxylated procollagen substrate labeled with [4-‘Hlproline using varying concentrations of ascorbate (0).
vectorial transport system for ascorbate in microsomes. Further evidence for such a system was sought by comparing the ascorbate concentrations required for proline hydroxylation in soluble and vesicular systems. In one case we used an exogenous substrate and microsomes permeabilized by freezing and thawing a suspension. In the second case, intact microsomes containing intracisternal [4-3H]proline-labeled unhydroxylated procollagen substrate and prolyl hydroxylase were used. The labeled microsomes were prepared by incubating tissue in the presence of a&-dipyridyl, which inhibits proline hydroxylation and procollagen secretion and thus causes an accumulation of unhydroxylated procollagen in the cisternae of the rough endoplasmic reticulum. In the permeabilized system, ascorbate was freely accessible to
AND
LUEDKE
the solubilized enzyme, whereas in the intact system ascorbate had to traverse the membrane for access to the enzyme. The Km for ascorbate in the solubilized system (Fig. 1) was approximately 0.45 mM, which is similar to that found for enzyme purified from various species (1). In contrast, the K,,, for ascorbate in the intact system was almost an order of magnitude lower (0.05 mM), which suggested that ascorbate had been concentrated within the vesicles. In our preliminary experiments, [i”C]ascorbate uptake into chick embryo limb bone microsomes was stimulated severalfold by the addition of the cosubstrate LYketoglutarate and the other factors required for optimal proline hydroxylation in intact microsomes. In our present studies, we tested each component separately and found that FeS04 alone, at the concentration required for the hydroxylation reaction (0.2 mM), accounted for the entire stimulation and similar results were obtained using microsomes from L-929 cells (data not shown). When experiments were carried out to determine the optimal Fe2+ concentration for ascorbate transport in limb bone microsomes, we observed a biphasic curve (Fig. 2). There appeared to be a plateau reached at about 0.1-0.2 mM (expanded in the inset) which was followed by
Fee2 ImMl FIG. 2. [“CJAscorbate bound to chick embryo limb bone microsomes at varying concentrations of Fe’+. Unlabeled microsome pellets were suspended and [“Clascorbate uptake was measured as described under Materials and Methods except that the concentration of FeSO, was varied as indicated in the figure. The region of the curve from 0 to 0.2 mM is expanded in the inset.
ASCORBATE
UPTAKE
TABLE
INTO
CHICK
EMBRYO
LIMB
BONE
285
MICROSOMES
I
DIFFERENTIAL HEAT STABILITY OF MICROSOMAL ASCORBATE UPTAKE AT Low AND HIGH FE’+ CONCENTRATIONS Fe2’ (mM)
Microsomes Control Boiled b Control Boiled
Ascorbate uptake’ (pmol/pg RNA)
0.2 0.2 2.0
17.6 5.8 248.6
2.0
248.6
(1Uptake of [Wlascorbate into chick embryo limb bone microsomes was measured as described under Materials and Methods except that FeSO, was used at the concentrations indicated. b Portions of microsome suspensions were added to assay tubes and heated at 100°C for 15 min. The tubes were cooled to 0°C and the remaining components of the transport assay were added.
an exponential increase in activity with increasing concentration. At concentrations above 0.2 mM, the microsomes contained a red precipitate which increased with concentration, suggesting that a nonbiological process might be involved. This idea was tested by comparing [‘4C]ascorbate uptake at 0.2 and 2 InM FeS04 using control and boiled microsomes (Table I). The results showed that the uptake at 2 mM was not TABLE SPECIFICITY OF THE METAL [14C]A~~~~~~~~
(mMl
affected by boiling, but at the lower concentration, uptake was inhibited by almost 70%, supporting the conclusion that the appearance of radioactivity in the microsomes at iron concentrations greater than 0.2 mM was not due to a specific transport system. A number of other metals were
II REQUIREMENT UPTAKE
FOR ‘)
[i4C]Ascorbate uptake b Relative activity
Expt.
Metal”
1
Fe” Fe’+ cl?+ None
106 114
Fe’+ Zn2+ Ca2+ None
100
2
Ascorbate
FIG. 3. [“C]Ascorbate uptake into microsomes as a function of ascorbate concentration. Unlabeled chick embryo limb bone microsomes were suspended and [‘“C]ascorbate uptake was measured as described under Materials and Methods except that the amount of r’C]ascorbate added was varied as indicated in the figure.
65
12 33
8 15
a Metals were present at 0.2 mM as either the sulfate (Fee+, Cua+, Zx?) or chloride (Fea+, Ca2’) salts. b[14CjAscorbate uptake into chick embryo limb bone microsomes was measured as described under Materials and Methods.
I, 1
2
I
I
3
4
I
I
I
I
I
5 6 Time Win.)
I
7
6
9
10
FIG. 4. [“CjAscorbate uptake as a function of time. Unlabeled chick embryo limb bone microsomes were suspended and uptake was measured as described under Materials and Methods except that incubations were carried out for varying time intervals as indicated in the figure. The results are from two separate experiments (0,O) using different intervals but with overlapping points (2 and 4 min).
286
PETERKOFSKY.
TSCHANK,
tested at 0.2 mM for stimulation of ascorbate transport activity (Table II), but only Fe(2+ or 3+) and Cu2’ exhibited significant activity while Zn2+ was about one-third as active as iron. Mg2+ (1 mM) and K+ (5 mM) were without effect (data not shown). In subsequent experiments, 0.15 mM Fe2+ was used routinely. Uptake of radioactive ascorbate was concentration dependent and reached saturation at 2-3 mM (Fig. 3). It also was time dependent (Fig. 4), with maximum uptake occurring at 6 min. No loss of label was observed even when samples were incubated for 60 min (data not shown). Uptake also was dependent on the concentration of microsomes, expressed as micrograms of ribosomal RNA (Fig. 5). The reason for the limitation of uptake at high concentrations of microsomes is not clear but it might be due to depletion of either Fe2+ or ascorbate. A dilution experiment using unlabeled ascorbate and dehydroascorbate was carried out to determine whether radioactivity I
I
Micmsomal
RNA 1~1~)
FIG. 5. [“ClAscorbate uptake as a function of microsome (rRNA) concentration. Chick embryo limb bone microsomes were suspended in sucrose/Tris buffer to give an RNA concentration of 1.16 mg/ml. Dilutions of the suspension were made and 40 ~1 of the undiluted suspension (46 pg RNA) and of the dilutions was added to the transport assay system and uptake was measured as described under Materials and Methods. Unincubated (0 min) and incubated (8 min) samples were processed for each microsome concentration.
AND
LUEDKE TABLE
III
ADDITIONOF~NLABELED AXORBATEAND DEHYDROASCORBATETOTHE[~%]ASCORBATE UPTAKEASSAYSYSTEM Unlabeled compound (n-u 0 2 4 10 28
Percentage Ascorbate 100 48.8 30.2 27.7
of controla Dehydroascorbate” 100 100 92.7 -
“r’C]Aseorbate uptake into chick embryo limb bone microsomes was measured as described under Materials and Methods. The concentration of ascorbate in the control was 2 man. b Dehydroascorbate was prepared from unlabeled ascorbate using ascorbic acid oxidase, as described under Materials and Methods.
was taken up by microsomes in the form of ascorbate. The results (Table III) showed that only unlabeled ascorbate effectively diluted the radioactivity taken up by microsomes, suggesting that there was no oxidation of ascorbate prior to uptake. Other reducing agents such as dithiothreitol and glutathione had no effect on [‘4C]ascorbate uptake (data not shown). To determine whether the Fe2+-dependent uptake of ascorbate was involved in providing external cofactor for proline hydroxylation within the microsomes, we used microsomes containing [4-3H]prolinelabeled, unhydroxylated procollagen and prolyl hydroxylase within the cisternae (20). An outline of the protocol and the results are shown in Fig. 6. Microsomes were preincubated with ascorbate in the presence or absence of Fe2+ (No. 1) and samples from each set (- and + Fe2+) were treated (No. 2) with buffer or with ascorbic acid oxidase to oxidize the ascorbate to dehydroascorbate. Dehydroascorbate has been shown previously to be inactive as a cofactor for prolyl hydroxylase (23). Preliminary experiments showed that there was no effect of any of the components of the incubation, including FeS04, on ascorbic acid oxidase activity (data not shown). Since the oxidase, like other proteins (24), would not
ASCORBATE
1
UPTAKE
INTO
CHICK
Fe+’
AA
FIG. 6. Dependence on Fe’+ for protection of ascorbate from ascorbic acid oxidase as assessed by intracisternal proline hydroxylation. L-929 cell microsomes containing [4-‘Hlproline-labeled unhydroxylated procollagen and prolyl hydroxylase activity were suspended as described in the legend to Fig. 1. Two portions were incubated for 10 min at 37°C (No. 1) with 1 mM ascorbate with (+) or without (-) 0.2 mM FeSO,. Samples were removed from each of those incubations and received either 25 units/ml of ascorbic acid oxidase in 50 mM Tris-HCl, pH 7.6 (black bars) or buffer alone (open bars) and they were further incubated for 15 min at 25°C (No. 2). In the final incubation for 5 min at 37°C (No. 3), the remaining factors required for hydroxylation were added. One set of samples (last column, hatched bar) was not subjected to preincubations in order to determine maximal hydroxylation activity. In a control sample not shown in the figure for the sake of simplicity, dehydroascorbate was prepared using ascorbic acid oxidase as described under Materials and Methods and was added in the place of ascorbate in the presence of FeSOd. The sample was carried through all three incubations but no additional ascorbic acid oxidase was added in incubation No. 2. No hydroxylation was obtained. Hydroxylation was determined by measuring release of tritiated water as described under Materials and Methods. Each value represents the average of duplicate assays.
be able to penetrate the microsomal membrane, ascorbate transported during the first preincubation period would be protected and support hydroxylation during a final incubation (No. 3) with the other co-
EMBRYO
LIMB
BONE
MICROSOMES
287
factors required for the reaction. In the final incubation, Fe2+ was supplied to those samples which had not received it during the preincubation. Samples not preincubated and containing all cofactors served as a control to indicate maximal hydroxylation activity. We confirmed the inactivity of dehydroascorbate in control samples run simultaneously. Ascorbate was oxidized with ascorbic acid oxidase prior to the initial incubation and no hydroxylation was obtained in the final incubation (data not shown). This control also established that dehydroascorbate could not be reduced back to ascorbate during the three incubations. The results showed that samples which had not been treated with the oxidase but which had received Fe2+ during either incubations No. 1 or No. 3, exhibited hydroxylation similar to that of the control, so that there was no loss of ascorbate, enzyme, or substrate during the preincubations. Of the samples treated with the oxidase, the ones which did not receive iron during the preincubation exhibited essentially no activity, in spite of the fact that Fe2+was added during the final incubation. In contrast, the oxidase-treated samples which had received Fe2+ only during the first preincubation period were able to support close to 50% of the maximum hydroxylation activity. The results suggest that an Fe2+-dependent process occurred during the preincubation which allowed a large proportion of the ascorbate to be protected from the action of ascorbic acid oxidase. DISCUSSION
The present study confirms our preliminary results (20) which suggested that there was a transport system for ascorbate in microsomal vesicles. The finding that the apparent Km for hydroxylation of proline in isolated, intact microsomes containing endogenous prolyl hydroxylase and unhydroxylated procollagen was about onetenth that observed using an exogenous substrate with enzyme released from the microsomes by freezing and thawing, was consistent with the existence of such a system. The results imply that ascorbate was
288
PETERKOFSKY,
TSCHANK,
concentrated within the microsomes to achieve concentrations similar to those required in the soluble system and for purified enzyme preparations (1). The Km for the microsomal system was still at least lo-fold higher than the half-maximal concentration of ascorbate required for proline hydroxylation or procollagen secretion in cell cultures (11,25). Since transport across the cell membrane led to a IO-fold concentration of ascorbate within cells (11) it is possible that the low extracellular concentration required might be a result of the combined concentrating effects of the cell and microsomal membrane transport systems. The uptake of [14C]ascorbate into microsomes was shown to be time and temperature dependent and substrate saturable. Transport was dependent on microsome concentration and heated microsomes were inactive, suggesting that a membrane protein might be involved in the process. The sole requirement for uptake of [14C]ascorbate by microsomes was the presence of iron in either the reduced or oxidized form. It is likely that ferric ion was reduced to the ferrous form by ascorbate, since this reaction proceeds readily (22). The concentration of ascorbate remaining would still be close to saturating since the ratio of ascorbate to iron was 1O:l. The functional requirement for Fe’+ in ascorbate transport was demonstrated using an intracisternal proline hydroxylating system. After preincubating microsomes and ascorbate with or without the metal, any ascorbate not transported was converted to inactive dehydroaseorbate by ascorbic acid oxidase, which cannot penetrate the microsomal membrane. The fact that significant hydroxylation was obtained only after preincubation with Fe2+ is consistent with the conclusion that the iron-dependent transport system is required for ascorbate to penetrate the microsomal membrane and participate in intracisternal enzyme reactions. The role of iron might be to bind to a membrane protein involved in the transport or to form an Fe2+-ascorbate chelate (26) which may be required for transport. Although the conversion of ascorbate to dehydroascor-
AND
LUEDKE
bate is required to obtain transport into some cells (27), the mierosomal system appears to take up ascorbate directly, based on the hydroxylation experiment and on the finding that only unlabeled ascorbate, and not dehydroascorbate, competed for uptake of [14C]ascorbate. The behavior of ascorbate with respect to biochemical reactions occurring within subcellular vesicles appears to be quite different in the rough endoplasmic reticulum of connective tissues compared to chromaffin granules of the adrenal medulla. Chromaffin granules contain unusually high concentrations of ascorbate, yet several studies have failed to reveal transport of exogenous ascorbate into isolated granules (17, 18). Nevertheless, exogenous ascorbate is required for epinephrine formation at the dopamine @-hydroxylase step in the granules (28). These results have led to the suggestion that in this system ascorbate functions as a shuttle for a transmembrane transfer of reducing equivalents, and evidence for such a system has been presented (29). The demonstration of uptake of ascorbate into connective tissue microsomes strongly suggests that ascorbate does not function through a shuttle mechanism during the post-translational hydroxylation of procollagen chains. Further support for this conclusion stems from the observation that L-929 cells grown in the absence of ascorbate, as they were in these experiments, do not contain any of this cofactor (30) so that exogenous ascorbate must function directly in proline hydroxylation in isolated microsomes. REFERENCES 1. CARDINALE, G. J., AND UDENFRIEND, S. (1974) Adv. Enzymd Relat. Areas Mol Bid 41,245~300. 2. PETERKOFSKY, B., AND ASSAD, R. (1976) J. Bid Chem 251,47’70-47’77. 3. PETERKOFSKY, B., AND ASSAD, R. (1979) J. Biol Chem 264,4714-4720. 4. PROCKOP, D. J., BERG, R. A., KIVIRIKKO, K. I., AND UITTO, J. (1976) in Biochemistry of Collagen (Ramachandran, G. N., and Reddi, A. H., Eds.), pp. 163-273, Plenum, New York. 5. H~RTNAGL, H., WINKLER, H., AND Lams, H. (1972) Bit&em. J. 129,187-195. 6. KIRSHNER, N. (1957) .J. Bid Chem. 226,821-825.
ASCORBATE
UPTAKE
INTO
CHICK
7. ST&&NE, L., ROTH, R. H., AND LISHAJKO, F. (1967) Biochem Pharm. l&1729-1739. 8. EIPPER, B. A., MAINS, R. E., AND GLEMBOTSKI, G. C. (1983) J. Bid Chem. 80,5144-5148. 9. GLEMBOTSKI, G. C. (1984) J. Biol. Chem 259,1304113048. 10. LOHMANN, W., AND WINZENBURG, J. (1983) 2. Naturforch C 38,923-929. 11. BLANCK, T. J. J., AND PETERKOFSKY,B. (1975) Arch. Biochewt. Biophvs. 171,259-267. 12. CASTRANOVA, V., WRIGHT, J. R., COLBY, H. D., AND MILES, P. R. (1983) J. Apple Physiol 54, 208214. 13. DILIBERTO, E. J., JR., HECKMAN, G. D., AND DANIEL& A. J. (1983) J. Bid Ch.em 258,12886-12894. 14. PATTERSON, L. T., NAHRWOLD, K. L., AND ROSE, R. C. (1982) Life Sci. 31,2783-2791. 15. TOGGENBURGER, G., HXUSSERMANN, M., MUTSCH, B., GERRONI, G., KESSLER, M., WEBER, F., HORNIG, D., O’NEILL, B., AND SEMENZA, G. (1981) Biochem. Biophys. Acta 646,433-443. 16. INGBRETSEN, 0. C., TERLAND, 0.. AND FLATMARK, T. (1980) Biochim Biophys. Acta 628.182-189. 17. LEVINE, M., MORITA, K., AND POLLARD, H. (1985) J. Bid Chem. 260,12942-12947. 18. TIRRELL, J. G., AND WESTHEAD, E. W. (1979) Neuroscience 4.181-186.
EMBRYO
LIMB
BONE
MICROSOMES
289
19. INGBRETSEN, 0. C., AND NORMANN, P. I. (1982) Biochim Biophys. Acta 684,21-26. 20. MATA, J. M., ASSAD, R., AND PETERKOFSKY, B. (1981) Arch Biochem Biophys. 206,93-104. 21. PETERKOFSKY, B., AND ASSAD, R. (1975) Anal B&hem 66,279-286. 22. ZANNONI, V., LYNCH, M., GOLDSTEIN, S., AND SATO, P. (1974) Biochem Med 11,41-48. 23. MYLLYLH, R., KUUTTI-SAVOLAINEN, E.-R., AND KIVIRIKKO, K. I. (1978) Biochem. Biophys. Res. Commun 83,441-448. 24. NILSSON, R., PETERSON,E., AND DALLNER, G. (1973) J. Cell Biol. 66.762-776. 25. PETERKOFSKY, B., BATEMAN, J. F., AND CHOJKIER, M. (1982) in Immunochemistry of the Extracellular Matrix (Furthmayr, H., Ed.), Vol. 2, pp. 19-47, CRC Press, Boca Raton, FL. 26. CONRAD, M. E., AND SCHADE, S. G. (1968) Gas&e enterologg 55,35-45. 27. BIGLEY, R., RIDDLE, M., LAYMAN, D., AND STANKOVA, L. (1981) Biochim Biophys. Acta 659,1522. 28. LEVINE, M. (1986) J. Biol. Chem. 261.7347-7356. 29. BEERS, M. F., JOHNSON, R. G., AND SCARPA, A. (1986) J. Biol Chem 261,2529-2535. 30. PETERKOFSKY, B., KALWINSKY, D., AND ASSAD, R. (1980) Arch Biochem Biophys. 199,362-373.
Iron-Dependent
Uptake of Ascorbate
BEVERLY PETERKOFSKY,l Laboratory
of Biochemistry,
into Isolated Microsomes
GEORG TSCHANK:
AND
National
Bethesda, Maryland
Cancer Institute,
CHRISTINA LUEDKE3 243892
Received September 19,1986, and in revised form December 16, 1986
A preliminary study (J. M. Mata, R. Assad, and B. Peterkofsky (1981) Arch. Biochem Biophvs. 206,93-104) suggested that chick embryo limb bone microsomes took up and concentrated [14C]ascorbate in the presence of cofactors for prolyl hydroxylase. In the present study, we found that the apparent Km for ascorbate in the hydroxylation of intracisternal unhydroxylated procollagen by endogenous prolyl hydroxylase was approximately an order of magnitude less than the value obtained when enzyme solubilized from microsomes was used with an exogenous substrate. These results are compatible with a concentrative uptake of ascorbate into microsomes. The uptake of [14C]ascorbate into microsomes was confirmed and it required only iron, in either the ferrous or ferric form, and was time and temperature dependent, proportional to microsome concentration, and substrate saturable at 2-3 mM ascorbate. Iron-dependent ascorbate uptake also was observed with L-929 cell microsomes. [14C]Ascorbate seemed to be taken up without prior oxidation, since only unlabeled ascorbate, and not dehydroascorbate, competed for uptake into limb bone microsomes. A functional requirement for Fe2+ in ascorbate transport was demonstrated using the intracisternal proline hydroxylating system. L-929 cell microsomes were preincubated with ascorbate with or without the metal and then external ascorbate was oxidized to inactive dehydroascorbate using ascorbic acid oxidase, which cannot penetrate the microsomal membrane. Samples which did not receive iron during the preincubation received it, along with other requirements for prolyl hydroxylase, in a final incubation to measure hydroxylation. Significant hydroxylation was obtained only in samples incubated with iron prior to oxidase treatment, consistent with the conclusion that an iron-dependent process was required to translocate ascorbate and protect it from the oxidase. Q 1987 Academic press. I,,~.
Ascorbic acid must pass through several membrane barriers in order to participate in some cellular reactions. For example, prolyl (EC 1.14.11.2)and lysyl (EC 1.14.11.4) hydroxylases require ascorbate (1) and are localized in the rough endoplasmic reticulum (2-4) where they carry out posttranslational modifications of procollagens (4). Several other vesicular enzymes which require ascorbate in vitro are dopamine ,f3hydroxylase (EC 1.14.17.1),which catalyzes 1To whom correspondence should be addressed. ’ Present address: Laboratory of Biochemistry, Children’s Hospital, University of Mainz, Mains, FRG. ‘Laboratory of Medical Biochemistry, The Rockefeller University, New York, NY 10021. 0003-9861/87 $3.00 Copyright All rights
CC!198’7 by Academic Press. Inc. of reproduction in any form reserved.
formation of norepinephrin in chromaffin granules of the adrenal (5, 6) and vesicles of adrenergic neurons (7), and peptidylglytine a-amidating monooxygenase, which cleaves and amidates a polypeptide hormone precursor in pituitary secretory granules (8,9). Since the cofactor is acidic, it is unlikely to penetrate membranes by simple diffusion (10). Active transport across the cell membrane has been demonstrated in a variety of cell types or tissues, including chick embryo bone cells (ll), alveolar macrophages (12), adrenomedullary chromaffin cells (13), intestinal mucosa (14), and renal brush border tubules (15). These plasma membrane transport systems generally are sodium depen-
282
ASCORBATE
UPTAKE
INTO
CHICK
dent. In contrast, there are relatively few reports of ascorbate transport into subcellular organelles. The finding that a large proportion of the ascorbate in adrenal medulla was concentrated in the granule fraction suggested that there might be a transport system for ascorbate in these vesicles (16). Other studies, however, indicate that there is no direct uptake of ascorbate by isolated granules (17, 18). Aseorbate transport into isolated guinea pig mitochondria has been demonstrated, but it was suggested that the process occurred by diffusion (19). Preliminary results from our laboratory (20) indicated that isolated microsomes which contained prolyl and lysyl hydroxylases, took up and concentrated radioactive ascorbate. The uptake was temperature dependent and was stimulated by a mixture of the cosubstrate a-ketoglutarate and other factors required for optimal hydroxylase activity. In the present study, we have determined the exact requirements and other characteristics of the transport system. MATERIALS
AND
METHODS
Materials. L-[1-W]Ascorbate (7-8 &i/pmol), ‘HzO, and [4-3H]proline were purchased from New England Nuclear. Ascorbate oxidase was purchased from Boehringer-Mannheim. The sources of other chemicals have been described previously (20). Microsome preparations. The isolation and characterization of chick embryo limb bone microsomes have been described previously (2). In some cases, prior to preparation of the microsomes, the tissue was labeled with [4-‘Hlproline by the same procedures used to label frontal bones (21). L-929 cell microsome-enriched fractions, either unlabeled or containing [4aHJproline-labeled, unhydroxylated procollagen, were prepared from cells incubated in the presence of the iron chelator a,&-dipyridyl (20). [“C]Ascorbate transport assay. Microsome pellets that had been stored in liquid Nz, were thawed and pH resuspended in 0.25 M sucrose/O.05 M Tris-HCl, ‘7.6, and 40 ~1 was added to a loo-p1 assay. Microsomes stored under these conditions have been shown to have intact membranes (2, 3, 20). Conditions were essentially those described previously (20), except that the final concentration of ascorbate was 2 mM (1.0 &i/ pmol), unless otherwise indicated, the temperature was 24°C and MgzSOI and KC1 were not present. The other components were 40 mM Tris-HCI, pH 7.6, and 0.15 mM FeSO,, unless otherwise indicated. Incuba-
EMBRYO
LIMB
BONE
MICROSOMES
283
tions were carried out for 8 min and a zero-time control was used for each sample. The value from the unincubated control was subtracted from the 8-min value in order to correct for nonspecific adsorption. After the incubation, 0.9 ml of ice-cold 0.25 M sucrose/ pH 7.6/2 mM ascorbate was added, 0.05 M Tris-HCl, the samples were immediately centrifuged at 43,500g for 10 min at 4°C and the supernatant solutions were discarded. The recovered microsomes were resuspended in 0.5 ml of the same buffer and the suspension was centrifuged as before. The pellets were dissolved in 0.5 ml of 0.5% Triton X-100 and the solutions were transferred to counting vials. The tubes were rinsed with 0.5 ml of detergent which was added to the vials along with 9 ml of Hydrofluor scintillation fluid and the radioactivity was measured. All assays were performed on duplicate samples and the results presented are the averages of duplicate values corrected for unincubated controls. In most cases, results are expressed as picomoles of ascorbate taken up, which was calculated using the specific radioactivity of [i4C]ascorbate. In some cases, the specific transport activity was calculated as picomoles per microgram RNA, in order to compare experiments in which different amounts of microsomes were used. RNA, rather than protein, was used as a basis for comparison since it is found specifically in microsomal vesicles and thus eliminates the problem of varying protein contributions from other vesicles in different preparations. Generally, lo-30 pg of RNA per assay was used. RNA was measured in hot trichloroacetic acid extracts of microsomes (2,20). Proline hydroxylation in intact microscmes. Microsomes containing endogenous [4-8H]proline-labeled unhydroxylated procollagen and prolyl hydroxylase were incubated in a total volume of 100 ~1 with 0.2 mM a-ketoglutarate, 0.2 mM FeSOI, 0.1 mM thymol, 1.0 mM ascorbate, and 50 mM Tris-HCI, pH 7.6, unless otherwise indicated. The release of ‘Hz0 resulting from proline hydroxylation was measured as described previously (20,21). Soluble pro@ h&m$ase activity. Conditions were similar to those described above except that a purified preparation of [4-‘Hlproline-labeled unhydroxylated procollagen (21) was used and amounts of solubilized enzyme extracts were used which fell in the linear portion of a rate versus protein concentration curve. Preparation of cleh&oaswrbate. Ascorbate, 50 mM, was incubated with a large excess (150 units/ml) of ascorbic acid oxidase in 10 mM Tris-HCl, pH 7.6, for 15 min at 25°C. These conditions were shown to oxidize 100% of the ascorbate using either a calorimetric assay (22) or uv absorption at 265 nm to measure ascorbate. RESULTS
Our previous experiments (20) suggested that there was a temperature-dependent
284
PETERKOFSKY,
Intracisternal:
TSCHANK,
enzyme substrate
Solubilized Exogenous
Ascorbate,
enzyme substrate
mM
FIG. 1. Proline hydroxylation as a function of ascorbate concentration in chick embryo limb bone microsomes compared to solubilized prolyl hydroxylase. Assay procedures are described under Materials and Methods. Microsomes containing [4-Qlproline-labeled unhydroxylated procollagen were suspended in 0.25 M sucrose/O.05 M Tris-HCl, pH 7.6, and portions were incubated with varying concentrations of ascorbate for 20 min and assayed for proline hydroxylation by measuring8Hz0 release (e). An unlabeled microsome pellet was similarly resuspended and the suspension was frozen at -2O’C and then thawed to solubilize prolyl hydroxylase. Enzyme activity was measured with an exogenous unhydroxylated procollagen substrate labeled with [4-‘Hlproline using varying concentrations of ascorbate (0).
vectorial transport system for ascorbate in microsomes. Further evidence for such a system was sought by comparing the ascorbate concentrations required for proline hydroxylation in soluble and vesicular systems. In one case we used an exogenous substrate and microsomes permeabilized by freezing and thawing a suspension. In the second case, intact microsomes containing intracisternal [4-3H]proline-labeled unhydroxylated procollagen substrate and prolyl hydroxylase were used. The labeled microsomes were prepared by incubating tissue in the presence of a&-dipyridyl, which inhibits proline hydroxylation and procollagen secretion and thus causes an accumulation of unhydroxylated procollagen in the cisternae of the rough endoplasmic reticulum. In the permeabilized system, ascorbate was freely accessible to
AND
LUEDKE
the solubilized enzyme, whereas in the intact system ascorbate had to traverse the membrane for access to the enzyme. The Km for ascorbate in the solubilized system (Fig. 1) was approximately 0.45 mM, which is similar to that found for enzyme purified from various species (1). In contrast, the K,,, for ascorbate in the intact system was almost an order of magnitude lower (0.05 mM), which suggested that ascorbate had been concentrated within the vesicles. In our preliminary experiments, [i”C]ascorbate uptake into chick embryo limb bone microsomes was stimulated severalfold by the addition of the cosubstrate LYketoglutarate and the other factors required for optimal proline hydroxylation in intact microsomes. In our present studies, we tested each component separately and found that FeS04 alone, at the concentration required for the hydroxylation reaction (0.2 mM), accounted for the entire stimulation and similar results were obtained using microsomes from L-929 cells (data not shown). When experiments were carried out to determine the optimal Fe2+ concentration for ascorbate transport in limb bone microsomes, we observed a biphasic curve (Fig. 2). There appeared to be a plateau reached at about 0.1-0.2 mM (expanded in the inset) which was followed by
Fee2 ImMl FIG. 2. [“CJAscorbate bound to chick embryo limb bone microsomes at varying concentrations of Fe’+. Unlabeled microsome pellets were suspended and [“Clascorbate uptake was measured as described under Materials and Methods except that the concentration of FeSO, was varied as indicated in the figure. The region of the curve from 0 to 0.2 mM is expanded in the inset.
ASCORBATE
UPTAKE
TABLE
INTO
CHICK
EMBRYO
LIMB
BONE
285
MICROSOMES
I
DIFFERENTIAL HEAT STABILITY OF MICROSOMAL ASCORBATE UPTAKE AT Low AND HIGH FE’+ CONCENTRATIONS Fe2’ (mM)
Microsomes Control Boiled b Control Boiled
Ascorbate uptake’ (pmol/pg RNA)
0.2 0.2 2.0
17.6 5.8 248.6
2.0
248.6
(1Uptake of [Wlascorbate into chick embryo limb bone microsomes was measured as described under Materials and Methods except that FeSO, was used at the concentrations indicated. b Portions of microsome suspensions were added to assay tubes and heated at 100°C for 15 min. The tubes were cooled to 0°C and the remaining components of the transport assay were added.
an exponential increase in activity with increasing concentration. At concentrations above 0.2 mM, the microsomes contained a red precipitate which increased with concentration, suggesting that a nonbiological process might be involved. This idea was tested by comparing [‘4C]ascorbate uptake at 0.2 and 2 InM FeS04 using control and boiled microsomes (Table I). The results showed that the uptake at 2 mM was not TABLE SPECIFICITY OF THE METAL [14C]A~~~~~~~~
(mMl
affected by boiling, but at the lower concentration, uptake was inhibited by almost 70%, supporting the conclusion that the appearance of radioactivity in the microsomes at iron concentrations greater than 0.2 mM was not due to a specific transport system. A number of other metals were
II REQUIREMENT UPTAKE
FOR ‘)
[i4C]Ascorbate uptake b Relative activity
Expt.
Metal”
1
Fe” Fe’+ cl?+ None
106 114
Fe’+ Zn2+ Ca2+ None
100
2
Ascorbate
FIG. 3. [“C]Ascorbate uptake into microsomes as a function of ascorbate concentration. Unlabeled chick embryo limb bone microsomes were suspended and [‘“C]ascorbate uptake was measured as described under Materials and Methods except that the amount of r’C]ascorbate added was varied as indicated in the figure.
65
12 33
8 15
a Metals were present at 0.2 mM as either the sulfate (Fee+, Cua+, Zx?) or chloride (Fea+, Ca2’) salts. b[14CjAscorbate uptake into chick embryo limb bone microsomes was measured as described under Materials and Methods.
I, 1
2
I
I
3
4
I
I
I
I
I
5 6 Time Win.)
I
7
6
9
10
FIG. 4. [“CjAscorbate uptake as a function of time. Unlabeled chick embryo limb bone microsomes were suspended and uptake was measured as described under Materials and Methods except that incubations were carried out for varying time intervals as indicated in the figure. The results are from two separate experiments (0,O) using different intervals but with overlapping points (2 and 4 min).
286
PETERKOFSKY.
TSCHANK,
tested at 0.2 mM for stimulation of ascorbate transport activity (Table II), but only Fe(2+ or 3+) and Cu2’ exhibited significant activity while Zn2+ was about one-third as active as iron. Mg2+ (1 mM) and K+ (5 mM) were without effect (data not shown). In subsequent experiments, 0.15 mM Fe2+ was used routinely. Uptake of radioactive ascorbate was concentration dependent and reached saturation at 2-3 mM (Fig. 3). It also was time dependent (Fig. 4), with maximum uptake occurring at 6 min. No loss of label was observed even when samples were incubated for 60 min (data not shown). Uptake also was dependent on the concentration of microsomes, expressed as micrograms of ribosomal RNA (Fig. 5). The reason for the limitation of uptake at high concentrations of microsomes is not clear but it might be due to depletion of either Fe2+ or ascorbate. A dilution experiment using unlabeled ascorbate and dehydroascorbate was carried out to determine whether radioactivity I
I
Micmsomal
RNA 1~1~)
FIG. 5. [“ClAscorbate uptake as a function of microsome (rRNA) concentration. Chick embryo limb bone microsomes were suspended in sucrose/Tris buffer to give an RNA concentration of 1.16 mg/ml. Dilutions of the suspension were made and 40 ~1 of the undiluted suspension (46 pg RNA) and of the dilutions was added to the transport assay system and uptake was measured as described under Materials and Methods. Unincubated (0 min) and incubated (8 min) samples were processed for each microsome concentration.
AND
LUEDKE TABLE
III
ADDITIONOF~NLABELED AXORBATEAND DEHYDROASCORBATETOTHE[~%]ASCORBATE UPTAKEASSAYSYSTEM Unlabeled compound (n-u 0 2 4 10 28
Percentage Ascorbate 100 48.8 30.2 27.7
of controla Dehydroascorbate” 100 100 92.7 -
“r’C]Aseorbate uptake into chick embryo limb bone microsomes was measured as described under Materials and Methods. The concentration of ascorbate in the control was 2 man. b Dehydroascorbate was prepared from unlabeled ascorbate using ascorbic acid oxidase, as described under Materials and Methods.
was taken up by microsomes in the form of ascorbate. The results (Table III) showed that only unlabeled ascorbate effectively diluted the radioactivity taken up by microsomes, suggesting that there was no oxidation of ascorbate prior to uptake. Other reducing agents such as dithiothreitol and glutathione had no effect on [‘4C]ascorbate uptake (data not shown). To determine whether the Fe2+-dependent uptake of ascorbate was involved in providing external cofactor for proline hydroxylation within the microsomes, we used microsomes containing [4-3H]prolinelabeled, unhydroxylated procollagen and prolyl hydroxylase within the cisternae (20). An outline of the protocol and the results are shown in Fig. 6. Microsomes were preincubated with ascorbate in the presence or absence of Fe2+ (No. 1) and samples from each set (- and + Fe2+) were treated (No. 2) with buffer or with ascorbic acid oxidase to oxidize the ascorbate to dehydroascorbate. Dehydroascorbate has been shown previously to be inactive as a cofactor for prolyl hydroxylase (23). Preliminary experiments showed that there was no effect of any of the components of the incubation, including FeS04, on ascorbic acid oxidase activity (data not shown). Since the oxidase, like other proteins (24), would not
ASCORBATE
1
UPTAKE
INTO
CHICK
Fe+’
AA
FIG. 6. Dependence on Fe’+ for protection of ascorbate from ascorbic acid oxidase as assessed by intracisternal proline hydroxylation. L-929 cell microsomes containing [4-‘Hlproline-labeled unhydroxylated procollagen and prolyl hydroxylase activity were suspended as described in the legend to Fig. 1. Two portions were incubated for 10 min at 37°C (No. 1) with 1 mM ascorbate with (+) or without (-) 0.2 mM FeSO,. Samples were removed from each of those incubations and received either 25 units/ml of ascorbic acid oxidase in 50 mM Tris-HCl, pH 7.6 (black bars) or buffer alone (open bars) and they were further incubated for 15 min at 25°C (No. 2). In the final incubation for 5 min at 37°C (No. 3), the remaining factors required for hydroxylation were added. One set of samples (last column, hatched bar) was not subjected to preincubations in order to determine maximal hydroxylation activity. In a control sample not shown in the figure for the sake of simplicity, dehydroascorbate was prepared using ascorbic acid oxidase as described under Materials and Methods and was added in the place of ascorbate in the presence of FeSOd. The sample was carried through all three incubations but no additional ascorbic acid oxidase was added in incubation No. 2. No hydroxylation was obtained. Hydroxylation was determined by measuring release of tritiated water as described under Materials and Methods. Each value represents the average of duplicate assays.
be able to penetrate the microsomal membrane, ascorbate transported during the first preincubation period would be protected and support hydroxylation during a final incubation (No. 3) with the other co-
EMBRYO
LIMB
BONE
MICROSOMES
287
factors required for the reaction. In the final incubation, Fe2+ was supplied to those samples which had not received it during the preincubation. Samples not preincubated and containing all cofactors served as a control to indicate maximal hydroxylation activity. We confirmed the inactivity of dehydroascorbate in control samples run simultaneously. Ascorbate was oxidized with ascorbic acid oxidase prior to the initial incubation and no hydroxylation was obtained in the final incubation (data not shown). This control also established that dehydroascorbate could not be reduced back to ascorbate during the three incubations. The results showed that samples which had not been treated with the oxidase but which had received Fe2+ during either incubations No. 1 or No. 3, exhibited hydroxylation similar to that of the control, so that there was no loss of ascorbate, enzyme, or substrate during the preincubations. Of the samples treated with the oxidase, the ones which did not receive iron during the preincubation exhibited essentially no activity, in spite of the fact that Fe2+was added during the final incubation. In contrast, the oxidase-treated samples which had received Fe2+ only during the first preincubation period were able to support close to 50% of the maximum hydroxylation activity. The results suggest that an Fe2+-dependent process occurred during the preincubation which allowed a large proportion of the ascorbate to be protected from the action of ascorbic acid oxidase. DISCUSSION
The present study confirms our preliminary results (20) which suggested that there was a transport system for ascorbate in microsomal vesicles. The finding that the apparent Km for hydroxylation of proline in isolated, intact microsomes containing endogenous prolyl hydroxylase and unhydroxylated procollagen was about onetenth that observed using an exogenous substrate with enzyme released from the microsomes by freezing and thawing, was consistent with the existence of such a system. The results imply that ascorbate was
288
PETERKOFSKY,
TSCHANK,
concentrated within the microsomes to achieve concentrations similar to those required in the soluble system and for purified enzyme preparations (1). The Km for the microsomal system was still at least lo-fold higher than the half-maximal concentration of ascorbate required for proline hydroxylation or procollagen secretion in cell cultures (11,25). Since transport across the cell membrane led to a IO-fold concentration of ascorbate within cells (11) it is possible that the low extracellular concentration required might be a result of the combined concentrating effects of the cell and microsomal membrane transport systems. The uptake of [14C]ascorbate into microsomes was shown to be time and temperature dependent and substrate saturable. Transport was dependent on microsome concentration and heated microsomes were inactive, suggesting that a membrane protein might be involved in the process. The sole requirement for uptake of [14C]ascorbate by microsomes was the presence of iron in either the reduced or oxidized form. It is likely that ferric ion was reduced to the ferrous form by ascorbate, since this reaction proceeds readily (22). The concentration of ascorbate remaining would still be close to saturating since the ratio of ascorbate to iron was 1O:l. The functional requirement for Fe’+ in ascorbate transport was demonstrated using an intracisternal proline hydroxylating system. After preincubating microsomes and ascorbate with or without the metal, any ascorbate not transported was converted to inactive dehydroaseorbate by ascorbic acid oxidase, which cannot penetrate the microsomal membrane. The fact that significant hydroxylation was obtained only after preincubation with Fe2+ is consistent with the conclusion that the iron-dependent transport system is required for ascorbate to penetrate the microsomal membrane and participate in intracisternal enzyme reactions. The role of iron might be to bind to a membrane protein involved in the transport or to form an Fe2+-ascorbate chelate (26) which may be required for transport. Although the conversion of ascorbate to dehydroascor-
AND
LUEDKE
bate is required to obtain transport into some cells (27), the mierosomal system appears to take up ascorbate directly, based on the hydroxylation experiment and on the finding that only unlabeled ascorbate, and not dehydroascorbate, competed for uptake of [14C]ascorbate. The behavior of ascorbate with respect to biochemical reactions occurring within subcellular vesicles appears to be quite different in the rough endoplasmic reticulum of connective tissues compared to chromaffin granules of the adrenal medulla. Chromaffin granules contain unusually high concentrations of ascorbate, yet several studies have failed to reveal transport of exogenous ascorbate into isolated granules (17, 18). Nevertheless, exogenous ascorbate is required for epinephrine formation at the dopamine @-hydroxylase step in the granules (28). These results have led to the suggestion that in this system ascorbate functions as a shuttle for a transmembrane transfer of reducing equivalents, and evidence for such a system has been presented (29). The demonstration of uptake of ascorbate into connective tissue microsomes strongly suggests that ascorbate does not function through a shuttle mechanism during the post-translational hydroxylation of procollagen chains. Further support for this conclusion stems from the observation that L-929 cells grown in the absence of ascorbate, as they were in these experiments, do not contain any of this cofactor (30) so that exogenous ascorbate must function directly in proline hydroxylation in isolated microsomes. REFERENCES 1. CARDINALE, G. J., AND UDENFRIEND, S. (1974) Adv. Enzymd Relat. Areas Mol Bid 41,245~300. 2. PETERKOFSKY, B., AND ASSAD, R. (1976) J. Bid Chem 251,47’70-47’77. 3. PETERKOFSKY, B., AND ASSAD, R. (1979) J. Biol Chem 264,4714-4720. 4. PROCKOP, D. J., BERG, R. A., KIVIRIKKO, K. I., AND UITTO, J. (1976) in Biochemistry of Collagen (Ramachandran, G. N., and Reddi, A. H., Eds.), pp. 163-273, Plenum, New York. 5. H~RTNAGL, H., WINKLER, H., AND Lams, H. (1972) Bit&em. J. 129,187-195. 6. KIRSHNER, N. (1957) .J. Bid Chem. 226,821-825.
ASCORBATE
UPTAKE
INTO
CHICK
7. ST&&NE, L., ROTH, R. H., AND LISHAJKO, F. (1967) Biochem Pharm. l&1729-1739. 8. EIPPER, B. A., MAINS, R. E., AND GLEMBOTSKI, G. C. (1983) J. Bid Chem. 80,5144-5148. 9. GLEMBOTSKI, G. C. (1984) J. Biol. Chem 259,1304113048. 10. LOHMANN, W., AND WINZENBURG, J. (1983) 2. Naturforch C 38,923-929. 11. BLANCK, T. J. J., AND PETERKOFSKY,B. (1975) Arch. Biochewt. Biophvs. 171,259-267. 12. CASTRANOVA, V., WRIGHT, J. R., COLBY, H. D., AND MILES, P. R. (1983) J. Apple Physiol 54, 208214. 13. DILIBERTO, E. J., JR., HECKMAN, G. D., AND DANIEL& A. J. (1983) J. Bid Ch.em 258,12886-12894. 14. PATTERSON, L. T., NAHRWOLD, K. L., AND ROSE, R. C. (1982) Life Sci. 31,2783-2791. 15. TOGGENBURGER, G., HXUSSERMANN, M., MUTSCH, B., GERRONI, G., KESSLER, M., WEBER, F., HORNIG, D., O’NEILL, B., AND SEMENZA, G. (1981) Biochem. Biophys. Acta 646,433-443. 16. INGBRETSEN, 0. C., TERLAND, 0.. AND FLATMARK, T. (1980) Biochim Biophys. Acta 628.182-189. 17. LEVINE, M., MORITA, K., AND POLLARD, H. (1985) J. Bid Chem. 260,12942-12947. 18. TIRRELL, J. G., AND WESTHEAD, E. W. (1979) Neuroscience 4.181-186.
EMBRYO
LIMB
BONE
MICROSOMES
289
19. INGBRETSEN, 0. C., AND NORMANN, P. I. (1982) Biochim Biophys. Acta 684,21-26. 20. MATA, J. M., ASSAD, R., AND PETERKOFSKY, B. (1981) Arch Biochem Biophys. 206,93-104. 21. PETERKOFSKY, B., AND ASSAD, R. (1975) Anal B&hem 66,279-286. 22. ZANNONI, V., LYNCH, M., GOLDSTEIN, S., AND SATO, P. (1974) Biochem Med 11,41-48. 23. MYLLYLH, R., KUUTTI-SAVOLAINEN, E.-R., AND KIVIRIKKO, K. I. (1978) Biochem. Biophys. Res. Commun 83,441-448. 24. NILSSON, R., PETERSON,E., AND DALLNER, G. (1973) J. Cell Biol. 66.762-776. 25. PETERKOFSKY, B., BATEMAN, J. F., AND CHOJKIER, M. (1982) in Immunochemistry of the Extracellular Matrix (Furthmayr, H., Ed.), Vol. 2, pp. 19-47, CRC Press, Boca Raton, FL. 26. CONRAD, M. E., AND SCHADE, S. G. (1968) Gas&e enterologg 55,35-45. 27. BIGLEY, R., RIDDLE, M., LAYMAN, D., AND STANKOVA, L. (1981) Biochim Biophys. Acta 659,1522. 28. LEVINE, M. (1986) J. Biol. Chem. 261.7347-7356. 29. BEERS, M. F., JOHNSON, R. G., AND SCARPA, A. (1986) J. Biol Chem 261,2529-2535. 30. PETERKOFSKY, B., KALWINSKY, D., AND ASSAD, R. (1980) Arch Biochem Biophys. 199,362-373.