Biochimica et Biophysicu Acra, 1095 t1991) 39-,~5 © 1991 Elsevier Science Publishers B.V. All rights rese~'cd 0167-4889/91/$03.50 ADONIS 0167488991002606
39
BBAMCR 13035
Effects of calcium on hepatocyte iron uptake from transferrin, iron-pyrophosphate and iron-ascorbate T. N i l s e n Departmem of Clinical Chemistry, Unil'ersity of Trondheirn, Trondheim ~Nor~'uy) (Received 15 March lq9D
Key words: Hepatocyle; Iron-uptake; Calcium: Transmembrane transport
Calcium stimulates hepatocyte iron uptake from transferrin, ferric-iron-pyrophosphate and ferrous-iron-ascorbate. Maximal stimulation of iron uptake is observed at 1-1.5 mM of extra-cellular calcium and the effect is reversible and immediate. Neither the receptor affinity for transferrin, nor the total amounts of transferrin associated with the cells or the rate of transferrin endocytosis are significantly affected by calcium. In the presence of calcium the rate of iron uptake of non.transferrin bound iron increases abruptly at approximate I'PC and 2"FC and as assessed by Arrhenius plots, the adivation energy is reduced in a calcium dependent manner at approx. 27°C. At a similar temperature, i.e., between 2S°C and 28°C, calcium increases the rates of cellular iron uptake from transferrin in a way that is not reflected in "~le rate of transferrin endocytosis. By the results of this study it is concluded that calcium increases iron transport across the plasma membrane by a mechanism dependent on membrane fluidity.
Introduction
It has been shown that iron uptake from transferrin by hepatocytes [1,2] and retieulocytes [3] depends on calcium. In reticulocytes Morgan [3] found that the transferrin receptor is irreversibly impaired by calcium deprivation and that this impairment is responsible for reduced transferrin iron uptake in the absence of calcium. In the perfused rat liver, however, iron uptake from iron-ascorbate is also found to be calcium dependent [2]. In hepatocytes therefore mechanisms for iron uptake other than receptor binding of transferrin are calcium dependent. As yet, however, no detailed investigation of the effect of calcium on hepatocyte iron uptake has been presented. In hepatocytes iron uptake from receptor bound transferrin has been found to take place both prior to and after endoeytosis of the transferrin-receptor complex [1,4-6]. This is in contrast to iron uptake by
erythroid cells which depends on endocytosis of receptor bound transferrin [6--12]. Irrespective of the cellular localization for iron mobilization from transferrin, iron depends on a mechanism for transmembrane transport to reach cytosol. However, the properties of this transmembrane transport mechanism are at present largely unknown. The regulatory effects of calcium on various aspects of cellular metabolism have been extensively studied [13-15]. Extracellular calcium through its interaction with membrane lipids and the polar phosphate heads of the phospholipids induces substantial changes on membrane fluidity, membrane lipia arrangement and the lipid composition [16,17]. Changes in membrane fluidity have been shown to influence a great variety of membrane functions, including receptor binding, endocytosis, enzyme activity and transmembrane transport mechanisms [18,19]. To explore which step(s) in the process of hepatocyte iron uptake from transferrin depend on calcium, iron uptake of both transferrinand non-transferrin-bound ferric and ferrous iron was investigated.
Abbreviations: Tes, N-tris(hydroxymethylhnethYi-2-aminoethanesulphonle acid; Tricine, N-tris(hydrox'ymethyl)methylglycine; LD, lactate dehydrogenase; HPLC, high-performance liquid chromatography; BSA, bovine serum albumin.
Methods
Correspondence:T. Nilsen,Departmentof ClinicalChemistry,Universityof Trondheim,N-7006Trondheim,Norway.
Cell isolation and incubation. Hepatoeytes were isolated from male Wistar-M011 rats (250-350 g) by the
40 collagenase perfusion method [23]. After a 30 min pre-incubation period at 37°C to allow for re-expression of cell surface receptors and release of native transferrin, the cells were washed four times at 4"C in a calcium-free buffer (SPB) containing: 68.4 mM NaCI, 5.4 mM KCI, 0.6 mM MgCI 2 • 6H20, 1.1 mM KH2PO4, 0.7 mM NazSO4, 30.2 mM Hepes, 30.1 mM Tes, 36.3 mM Tricine, 40 mM NaOH (pH 7.40). The hepatocytes were re-suspended (5. l0 b cells/ml) and incubated in SPB containing 10 g/I dialyzed (see below) bovine serum albumin referred to as calcium free medium (see Results) or with the calcium concentrations supplied to the medium as given in legends to the figures. The incubations were performed in Erlenmeyer flasks in a shaking water bath. Following temperature equilibration for 15 rain, [59Fe]transferrin, [59Fe]pyrophosphate, [SgFe]ascorbate or t2Sl-labeled transferrin were added in the concentrations given in the legends to the figures. At the time intervals indicated, 1 ml aliquots of cell suspensions were withdrawn and .~ashed four times at 4"C with excess volumes of albumin-free buffer containing the same amount o~" calcium as during the incubation. The final cell pellets were counted for SgFe and ~ I activity in an LKB 1282 Compugamma Universal gamma counter (LKB, Turku, Finland). In each experiment the 'zero-time' value was subtracted from the subsequent determinations. Fractionation of cytosol proteins. After sonication of cells (100 W, three times 30 s) with a Braun Labsonic 1510 sonicator, cytosol was prepared by centrifugation at 100000 × g for 30 rain at 4°C. Cytosol proteins were separated by gel filtration, using the high-performance liquid chromatography (HPLC) equipment from Waters Associates (Milford, MA, U.S.A.) and a Zorbax GF-250 column from Dupont Company (Willmington, DE, U.S.A.). The gel filtration calibration kit was from Pharrnacia AB (Uppsala, Sweden). General. Protein was determined by the Bio-Rad protein assay with bovine serum albumin (BSA) as standard [21]. Plasma membrane ferrieyanide reduetase activity was monitored in an SLM Aminco DW-2C speetrophotometer (SLM Instruments, Urbana, IL, U.S.A.) (dual wavelength mode, wavelength pair 420500 nm) as described by Thorstensen et al. [1]. Calcium determinations were done on a Perkin Elmer 2380 atomic absorption speetrophotometer (Perkin Elmer, Norwalk, CT, U.S.A.) with HGA-300 graphite furnace equipment (sensitivity 2 nM at a maximal volume of 100/~!). The presence of the iron chelates, i.e., pyrophosphate (0.4 raM) or ascorbate (2 raM) did not affect cell viability as assessed by leakage of lactate debydro. genase (LD) [22], content of ATP [23] or trypan blue exclusion. Each experiment (n > 3) was run in duplicate and the data presented in the Results are given as mean + S.D. Tne figures presented are of representative exper-
iments when not otherwise stated in the legends to the figures. Straight lines within figures are given as the best fit by linear regression analysis. Reagents. Diferric transferrin (absorbance ratio A466/A2a o = 0.045 + 0.001 (mean + S.D., n = 4) [24] labeled with 59Fe (specific activity 100 epm/pmol iron) or tZSl (specific activity 1500 cpm/pmol transferrin) were prepared as described in Refs. 25 and 26, respectively. Radiolabeled transferrin was dialyzed overnight prior to uptake studies to remove any low-molecularweight labeled transferrin fragments produced by radiolysis. 59Fe-labeled pyrophosphate (the molar ratio iron/pyrophosphate was 1:20) and 59Fe-labeled ascorbate (the molar ratio iron/ascorbate was 1 : 20) were prepared in deionized water just prior to use (specific activity 5 epm/pmol iron). S9FeCI3 (6.5-43.7 MBq/ retool) and BoRon and Hunter reagent (74 TBq/mmol) were from Amersham International (Buckinghamshire, U.K.). Collagenase (type IV), human transferrin (98%, essentially iron-free), Hepes, Tes, Tricine, pyrophosphate and bovine serum albumin (BSA) (essentially fatty add-free) were from Sigma (St. Louis, MO, U.S.A.). BSA was dialyzed 72 h against six changes of excess volumes of SPB buffer (see above). The calcium content of BSA was thus reduced by 90%, i.e., to 0.0004% (w/w) as assessed by atomic absorption spectrophotometry. Results
Effects of calcium on time-dependent iron uptake Hepatoeyte uptake of iron from [59Fe]transferrin, [59Fe]pyrophosphate or [59Fe]ascorbate at increasing concentrations of extracellular calcium was measured (Fig. 1A-C). In accordance wi{h earlier reports [1,2], the stimulating effect of calcium reached a level at 1-1.5 mM of calcium. At all calcium concentrations, iron uptake from the different iron donors proceeded linearly (i.e., r ~ 98, four measurements within 30 rain). Addition of calcium to cells incubated in a calcium free medium immediately and completely restored the ability of the cells to take up iron from transferrin (Fig. 2). Some 10% of BSA-bound calcium was unremovable by dialysis (see Methods). However, eeUular iron accumulation in a medium with undialyzed BSA (i.e., approx. 90 /~M calcium in the incubation medium) did not significantly differ from that in a medium with dialyzed BSA (Fig. 2). It thus appeared that BSA-assodated calcium did not stimulate cellular iron uptake. Some iron was taken up by the hepatocytes even in the caicium free medium (see Methods) (Figs. 1 and 2) and as for the control cells in the presence of calcium, more than 70% of this iron was recovered in eytosolie ferritin (as assessed by HPLC (data not shown)).
4l
C3
z
60
m~ z ~40 if: o_
o
A
0 300
(3~ 7
t 0
100 200 ~2~I-TRANSFERRIN] (riM) Fig. 3. The effect of calcium on cellular accumulation of tmnsferrin
Y
at increasing transferrirt concentratious. T h e hepatocytes were pro-
B
o
r r 4000 30O0
2000 lOOO
q
S
incubated 15 rain at 37°C in a calcium-free medium with (o, o) or without ( zx, A) 100 .aM EDTA. Followingfour washes in SPB, ceils were re-incubated at 4°C (closed w~nbols)or 3"PC(open symbols)in the calcium free medium.Control cells (120 •)were incubated in the presence of 1.2 mM of calcium throughout the experiment, l~llabeled transferrin was supplied to the re-incubated cells at the concentrations indicated. After60rain incubationat 4"C and after 30 rain incubation at 37°C the amounts of transferrin accumulated the cells were determined.
q
0
,
i
,
i
,
C
t 2 [Ca2+] (rnM) Fig. 1. The effect of increasing concentrationsof calcium on hepatocyte iron uptake. Hepatocytes were incubated with (A) [SVFe]transferrin (3 /zM iron), (B) [SgFe]pyrophosphate(20 ttM iron) or (C) [SgFe]ascorbate(1130/~M iron) at 37"C in media cor4aining calcium at the concentrations indicated. After 30 rain u~ iucubation duplicate aliquots were withdrawn and cellular iron uptake was determined.
Effects of calcium on transfen4n binding and transferrin endocytosis In reticulocytes Morgan showed that the receptors irreversibly lost the ability to bind transferrin after the cells had been incubated with E D T A at temperatures above 2ff'C [3]. To test such effects of calcium depriva-
-0.97. The activation energies for iron uptake given by the slopes of the lines were (A) from transferrin: 19.7 kcalmot -~ with calcium and 13.4 kcalmol - ! without calcium; (B) from the iroo-pyrophosphate complex as given from
2ooc
1ooc ..,,u
3.6
TEMPERATURE (~ x 10a) Fig. 6. Arrhenius plots of temperature-dependentrates of iron uptake with or withoutcalcium.The iron uptakedata (Fig. 5) with
400C- 3'3~ A
,
the right to the left (i.e., by increasingtem~ratare): 8.1, 18.8 and 14.3 kcalmol-t with calcium and 9.5 and 4.1 kcalmol-z without
ol
calcium and (C) from the iron-ascotbate complex as given from the right to the left: 30.2, 18.5 and 11.0 kcalmol - I with calcium and 12.5 and 4,6 kealmoi -1 without calcium.
o
5 cl 40x10~
g 20xtO~ rc
i
iJJ rc
.
i
.
i
.
°icZ
0.8x10~ 0.6x10~
0.4x10~
,0.2x10~, 0
10
20
3O
40
TEMPERATURE ('C} Fig, 5. The effect of calcium on the rates of hepatocyt¢ iron accumulation at increasing temperature. The hepatocytes were incubated with (A) [59Fe]transferrin (3 #M iron), (B) [SgFe]pyrophosphate (20 /zM iron) or (C) [SgFe]ascorbate (100 ,~M iron) (see Methods). The rates of iron uptake were measdred with ( o ) and without (e) 1.2 mM calcium in the temperature range of 4°C-38°C. The results are given as mean+S.D. ( n = 3). When no) indicated the S,D. values are within the size of the symbols. For all lines [n q and C, r ~_0.97.
phosphate (Fig. 5B) a second inflection was found at 27.0 + 0.2°C. In the calcium free medium, on the other hand, the rate of iron uptake from non.transferrin bound iron increased linearly through the entire temperature range (Fig. 5B and C) and the slight deviation from linearity in the temperature dependent rates of iron uptake from transferrin (Fig. 5A) simply reflected iron taken up by calcium-independent transferrin endocytosis (see Fig. 4). Fig. 6A-C shows the data of Fig. 5A-C in Arrhenius plots [27,281. As shown by the slopes of the lines no temperature dependent changes in the activation energies for iron uptake from transferrin were found either in the presence or the absence of calcium (Fig. 6A). A calcium dependent change in the activation energies for iron uptake from pyrophosphatc and ascorbate was, however, found at 27.5 + 1.1°C and 26.9 _+.2.0°C, respectively (Fig. 6B and (2). In addition, a second change in the activation energies for iron
43 4000 uJ "
~-
~ 3000
:"
1oo(]
38
tr
0
6 T4
,
I
100
,
I
200
,
300
BATE OF TRANSFERRINENDOCYTOSIS (molecules/cell/s)
Fig. 7. Rates of iron uptake related to rates of transferrinendocytosis at different temperatures. The rates of cellulariron uptake at differenttemperatureswith(o) and without(e) 1.2 mM calcium(as given in Fig. 5) were plotted against the correspondingrates of transferrin endocytosis(as given in Fig. 4). The temperatures are indicatedby the numbersnextto the symbols. uptake from pyrophosphate and ascorbate was found both in the presence and absence of calcium, i.e., at 16.0 4- 1.18°C and 15.0 4- 1.34°C in the absence of calcium and at 15.6 4- 0.6°C and 17.4 + 2.(FC in the presence of calcium for iron uptake from pyrophosphate and ascorbate, respectively. Only for calcium-dependent iron uptake from pyrophosphate the activation energy was found to be reduced at low temperatures (Fig. 613). Even if not the general rule, reductions in activation energies have been reported also for other transmembrane transport mechanisms [27]. However, the reason for this reduction in the activation energy and the concomitant low rates of iron uptake from pyrophosphate at low temperatures (Fig. 5B) is not apparent. In spite of that no calcium dependent reduction in the activation energy for iron uptake from transferrin was found at approx. 27*(], an effect of calcium near this temperature was still dearly shown by Fig. 7; i.e., when plotting the rates of iron uptake from transferrin at the different temperatures (data from Fig. 5) against the rates of transferrin ¢ndoeytosis at identical temperatures (data from Fig. 4), this figure showed a sharp calcium dependent increase in the rates of iron uptake above 25"C which was not reflected by a similar increase in transferrin ¢ndocytosis.
Discussion The results presented in this study show that extracellular calcium stimulates hepatocellular iron uptake from transferrin as well as iron uptake of non-transfertin bound ferrous and ferric iron (Fig. 1A-C). Maximal stimulation of iron uptake from transferrin and astorbate is observed at 1-1.5 mM of extracellular calcium.
Chelation of some caIcium by a surplus of pyrophosphate in the iro~-pyrophosphate solution may explain why iron uptake from the iron-pyrophasphate complex levels off at a somewhat higher concentration ot calcium. In accordance with earlier reports [2], uptake of iron from the ferrous-iron-ascorbate complex (Fig. 1) is also stimulated by calcium. Thus, the effect of calcium apparently is not on the reduction of ferric iron. Consistent with this interpretation is the finding that calcium has no effect on the plasma membrane ferricyanide reductase activity (data not shown). The complete and immediate restoration of iron uptake from transferrin upon calcium repletion (Fig. 2) shows that the hepatocyte transferrin receptors in contrast to the reticulocyte transferrin receptors [3], are not irreversibly impaired by calcium depletion. In fact no effect of calcium on either the receptor affinity for transferrin (see Results) or on the amount of cellular transferrin accumulated can be found (Fig. 3). To fully exclude any effect of calcium on the process of hepatocellular transferrin handling, the effect of calcium on the rate of transferrin endocytosis was investigated. The initial linearity for cellular transferrin accumulation (see Resutts) shows that exocytosis of labeled transferrin has not yet started and that a single process, i.e., endocytosis of transferrin, is measured. Since it thus is found that calcium does not affect the kinetics of transferrin endocytosis (Fig. 4), it appears that no step in the process of hepatocellular transferrin handling is calcium dependent. The finding that extracellular calcium stimulates iron uptake from all the iron donors investigated, indicates the involvement of a mechanism in the process of iron uptake which is common for the different iron donors, Even if it has been found that calcium promotes the binding of pyrophosphate to the plasma membrane of hepatocytes [29], the finding that calcium does not increase the membrane binding of transferrin (Fig. 3) shows that the stimulating effect of calcium on iron uptake is not on a general effect on the membrane binding of the iron donor complexes. However, transmembrane transport of iron mobilized outside the cell, i.e., at the plasma membrane or within endosomes as described for transferrin iron and pyrophosphate bound iron [1,5-12,29], may well be exerted by a calcium dependent mechanism general for all extracellularfy mobilized iron. An effect of calcium on a thermotropie membrane activity, is demonstrated by the calcium dependent reduction in the activation energy for iron uptake from both the iron-pyrophusphate complex and from the iron-ascorbate complex at approx. 27"C (Fig. 613 and C). The temperature dependent inflections in the rates of calcium-dependent iron uptake from pyropbosphate and ascorbate (Fig. 5B and C) and the calcium depen-
44 dent changes in activation energies (Fig. 6B and (2) which all correspond closely with the upper and lower thermotropic lipid phase transitions reported for the hepatocyte plasma membrane [30,31], further support that we are dealing with a calcium-regulated iron transmembrane transport mechanism. The absence of sharp temperature dependent inflections in the rate of calcium-dependent iron uptake from transferrin and the concomitant calcium dependent change in the activation energy (Figs. 5A and 6A), may be explained by a more complex process of iron release from transferrin than from the other iron donors. As a rate limiting step the iron release from transferrin may thus conceal calcium regulating effects on the iron transmembrane transport mechanism [28]. A calcium dependent influence on iron uptake from transferrin at the upper thermotroplc lipid phase transition is, however, clearly demonstrated in Fig. 7. This figure shows that above 25°C, calcium primarily stimulates the iron uptake which is independent of transferrin endocytosis. This finding is interpreted to mean that despite the rate limitations by iron release from transferrin, calcium stimulates iron uptake from transferrin the same way as for iron uptake from pyrophosphate and ascorbate, i.e., by an effect of calcium on the thermotropic iron transmembrane transport mechanism. Two additional findings of this study require some further comments; firstly, the temperature dependent inflection in the rates of calcium-dependent iron uptake from pyrophosphate at the higher temperature (Fig. 5B) (also seen in iron uptake from NTA, data not shown), can not be found for calcium-dependent iron uptake from ascorbate (Fig. 5C) (or for uptake of cobalt, data not shown). Since preliminary studies show no concomitant change in the ferricyanide reductase activity, this discrepancy in iron uptake of ferrous and ferric iron donors ma~, well be in the transmembrane transport process. Despite the claim that biological membranes are impermeable to the ferric iron ion and iron uptake from ferric iron donors thus depends upon reduction [8], it is suggested that the abundance of free fatty acids in the hepatocyte plasma membrane may allow for an additional transmembrane transport of ferric iron [32]. Lipid dependent transloeation of ferric iron ions may, however, depend on the more optimal constellations of membrane components, only to be obtained at temperatures above the upper lipid phase transition [311 and in the presence of calcium [17]. Secondly, since the iron accumulated by hepatocytes in the calcium free medium (Figs. 1, 2 and 5) is recovered in ferritin (see Results), this iron apparently is transported over the plasma membrane. Because viable ceils inevitably will release some calcium to the medium, iron uptake in the complete absence of calcium is difficult to document. However, if the iron
accumulated in the calcium free medium is translocoted to cytosol by the calcium-dependent mechanism running slow on small amounts of calcium redistributed to the membrane from intracellular stores, calcium induced changes in rates of iron uptake (Fig. 5B and C) are expected. The absence of such changes indicates a discrete mechanism for calcium independent transmembrane trans~rt of iron. This interpretation is further supported by the observed differences in the activation energies in the presence and absence of supplied calcium (Fig. 6A-C). However, at physiological temperatures iron accumulation by hepatocytes is strongly dominated by the calcium-dependent iron uptake mechanism. Whether this mechanism for the calcium-dependent transmembrane transport of iron is equally expressed in cells of different tissue is not known. As a scavenger of iron in times of noxious iron overload this high capacity iron uptake mechanism of the liver, may well be a highly developed function of this tissue. An answer to this question depends on future investigations.
Acknowledgements This work was supported by grants from the University of Trondheim. The assistance of med. lab, tech. I. Arbo and the helpful discussions with Dr. L Romslo and Dr. K. Thorstensen are gratefully acknowledged.
References 1 Thorstensen,IL and Romsln,I. U988)L Biol.Chem.263,88448850. 2 Wright,T.L, Brissot,P., Ma, W-L.and Weisiger,R.A. (1986)J. Biol. Chem.261, 10909-10914. 3 Morgan,E.H. (1989)Biochim.Biophys.Acta 981, 121-129. 4 Thorstensen,K. and Romslo,1. (1987)Scand.J. Ctin.Invest.4"/, 837-846. 5 Morley,C.G.D. and Bezkorovainy,A. (1985)Int. L Biochem. 17, 553-564. 6 Thorstensen,K. 0988) J. Biol.Chem. 263, 16837-16841. 7 Cfichlon,R.R, (1990)Adv. ProteinChem, U S A 40, 281-363. 8 De Jong O., Van Dijk,J.P.and Van Eijk,H.O. (1990)Clin.Chim. Acta 190, 1-46. 9 Bezkorovainy,A (1989)Clin.PM'siol.Biochem. 7, 1-17. IO Sun, I.L,Navas,P.,Crane, EL, Mor#, DJ. and Low, H. (1987) J. Biol.Chem. 262, 15915-15921. 11 Baker, E., McArdle, H.I. and Morgan, E.H. (1985)in Proteinsof
Iron Storage and Transport(Spik, G., Montreuil,J., Crichton, R.R. and Mazurier,J., eds.),pp. 131-142,ElsevierSciencePublishers,Amsterdam. 12 Dautry-Varsat,A., Ciechanover,A. and Lodish,H.F.(1982t)Prec. Natl. Acad. Sci. USA 80, 2258-2262. 13 Eaton,J.E. (1987)KidneyInt. 32, S68-$76. 14 McCormack, J.O., Halestrap, A.P. and Denton, R,M. (1990) PltysioLRev.70, 421-509. 15 PaL C.Y.C. (1990)in Fluids and EleclroMes(Kokko,LP. and Tannen,R.L, eds.), pp. 596-630, SaundersCompany,Philadelphia, 16 Livingstone,C.J. and Schachter, D. (1980) Biochemistry 19, 4823-4827.
45 17 Benga, G. and Holmes, R.P. (19M) Prog. Biophys. Mot. Biol. 43, 195-257. 18 Speetor, A.A. and Yorek, M.A. (1985) J. Lipid Res. t.~, t015-1035. 19 Mills, P.R., Ide[er, P.J., Smith, D.J., Ballatori, N., Boyer, J.L. and Gordon, E,R. (1987) Hepatology 7, 61-66. 20 Seglen, P.O. (1976) Methods Cell Biol. 13, 29-83. 21 Bradford, M.M. (1976) Anal. Biechem. 7Z, 284-254. 22 Empfehlungen der Deutschen Gesellshaft fur tedinische Chemie (1977) J. Clin. Chem. Clin. Biochem. 15, 249-254. 23 Ingebretsen, O.C., Bakken, A.M,, Segadal, L. and Farstad, M. (1982) J. Chromatogr. 2,12, 119-126. 24 Huebers, H.A., Csiba, E., Huebers, E. and Finch, C.A. (t983) Prec. Natl. Aead. Sei. USA 80, 300-3t,~4.
25 Konopka. K. and Romslo. I. (1981) Eur. J. Biochem. 117. 239-244. 26 Bolton. A.E. and Hunter, W.M. (1973) Biochem. J. 133. 529-539. 27 McEIhaney. R.N. (1982) Current Top. Membr. Transport 17, 317-380. 28 Raison, J.K. (1973) Bioenergetics 4, 285-309. 29 Nilsen, T and Romslo, I. (1990) Seand. J. Clin. Invest. 50, |9-25. 30 Livingstone, C.J. and Schachter, D. (1980) J. Biol. Chem. 255. 10902-10908. 31 Schachter. D. (1984) Hepatolo~ 4, 140-151. 32 Peters, T.J., Raja, K.B.. Simpson, R.J. and Snape. S. (1988) Ann. N.Y. Acad. Sci. 526, 141-147.
39
BBAMCR 13035
Effects of calcium on hepatocyte iron uptake from transferrin, iron-pyrophosphate and iron-ascorbate T. N i l s e n Departmem of Clinical Chemistry, Unil'ersity of Trondheirn, Trondheim ~Nor~'uy) (Received 15 March lq9D
Key words: Hepatocyle; Iron-uptake; Calcium: Transmembrane transport
Calcium stimulates hepatocyte iron uptake from transferrin, ferric-iron-pyrophosphate and ferrous-iron-ascorbate. Maximal stimulation of iron uptake is observed at 1-1.5 mM of extra-cellular calcium and the effect is reversible and immediate. Neither the receptor affinity for transferrin, nor the total amounts of transferrin associated with the cells or the rate of transferrin endocytosis are significantly affected by calcium. In the presence of calcium the rate of iron uptake of non.transferrin bound iron increases abruptly at approximate I'PC and 2"FC and as assessed by Arrhenius plots, the adivation energy is reduced in a calcium dependent manner at approx. 27°C. At a similar temperature, i.e., between 2S°C and 28°C, calcium increases the rates of cellular iron uptake from transferrin in a way that is not reflected in "~le rate of transferrin endocytosis. By the results of this study it is concluded that calcium increases iron transport across the plasma membrane by a mechanism dependent on membrane fluidity.
Introduction
It has been shown that iron uptake from transferrin by hepatocytes [1,2] and retieulocytes [3] depends on calcium. In reticulocytes Morgan [3] found that the transferrin receptor is irreversibly impaired by calcium deprivation and that this impairment is responsible for reduced transferrin iron uptake in the absence of calcium. In the perfused rat liver, however, iron uptake from iron-ascorbate is also found to be calcium dependent [2]. In hepatocytes therefore mechanisms for iron uptake other than receptor binding of transferrin are calcium dependent. As yet, however, no detailed investigation of the effect of calcium on hepatocyte iron uptake has been presented. In hepatocytes iron uptake from receptor bound transferrin has been found to take place both prior to and after endoeytosis of the transferrin-receptor complex [1,4-6]. This is in contrast to iron uptake by
erythroid cells which depends on endocytosis of receptor bound transferrin [6--12]. Irrespective of the cellular localization for iron mobilization from transferrin, iron depends on a mechanism for transmembrane transport to reach cytosol. However, the properties of this transmembrane transport mechanism are at present largely unknown. The regulatory effects of calcium on various aspects of cellular metabolism have been extensively studied [13-15]. Extracellular calcium through its interaction with membrane lipids and the polar phosphate heads of the phospholipids induces substantial changes on membrane fluidity, membrane lipia arrangement and the lipid composition [16,17]. Changes in membrane fluidity have been shown to influence a great variety of membrane functions, including receptor binding, endocytosis, enzyme activity and transmembrane transport mechanisms [18,19]. To explore which step(s) in the process of hepatocyte iron uptake from transferrin depend on calcium, iron uptake of both transferrinand non-transferrin-bound ferric and ferrous iron was investigated.
Abbreviations: Tes, N-tris(hydroxymethylhnethYi-2-aminoethanesulphonle acid; Tricine, N-tris(hydrox'ymethyl)methylglycine; LD, lactate dehydrogenase; HPLC, high-performance liquid chromatography; BSA, bovine serum albumin.
Methods
Correspondence:T. Nilsen,Departmentof ClinicalChemistry,Universityof Trondheim,N-7006Trondheim,Norway.
Cell isolation and incubation. Hepatoeytes were isolated from male Wistar-M011 rats (250-350 g) by the
40 collagenase perfusion method [23]. After a 30 min pre-incubation period at 37°C to allow for re-expression of cell surface receptors and release of native transferrin, the cells were washed four times at 4"C in a calcium-free buffer (SPB) containing: 68.4 mM NaCI, 5.4 mM KCI, 0.6 mM MgCI 2 • 6H20, 1.1 mM KH2PO4, 0.7 mM NazSO4, 30.2 mM Hepes, 30.1 mM Tes, 36.3 mM Tricine, 40 mM NaOH (pH 7.40). The hepatocytes were re-suspended (5. l0 b cells/ml) and incubated in SPB containing 10 g/I dialyzed (see below) bovine serum albumin referred to as calcium free medium (see Results) or with the calcium concentrations supplied to the medium as given in legends to the figures. The incubations were performed in Erlenmeyer flasks in a shaking water bath. Following temperature equilibration for 15 rain, [59Fe]transferrin, [59Fe]pyrophosphate, [SgFe]ascorbate or t2Sl-labeled transferrin were added in the concentrations given in the legends to the figures. At the time intervals indicated, 1 ml aliquots of cell suspensions were withdrawn and .~ashed four times at 4"C with excess volumes of albumin-free buffer containing the same amount o~" calcium as during the incubation. The final cell pellets were counted for SgFe and ~ I activity in an LKB 1282 Compugamma Universal gamma counter (LKB, Turku, Finland). In each experiment the 'zero-time' value was subtracted from the subsequent determinations. Fractionation of cytosol proteins. After sonication of cells (100 W, three times 30 s) with a Braun Labsonic 1510 sonicator, cytosol was prepared by centrifugation at 100000 × g for 30 rain at 4°C. Cytosol proteins were separated by gel filtration, using the high-performance liquid chromatography (HPLC) equipment from Waters Associates (Milford, MA, U.S.A.) and a Zorbax GF-250 column from Dupont Company (Willmington, DE, U.S.A.). The gel filtration calibration kit was from Pharrnacia AB (Uppsala, Sweden). General. Protein was determined by the Bio-Rad protein assay with bovine serum albumin (BSA) as standard [21]. Plasma membrane ferrieyanide reduetase activity was monitored in an SLM Aminco DW-2C speetrophotometer (SLM Instruments, Urbana, IL, U.S.A.) (dual wavelength mode, wavelength pair 420500 nm) as described by Thorstensen et al. [1]. Calcium determinations were done on a Perkin Elmer 2380 atomic absorption speetrophotometer (Perkin Elmer, Norwalk, CT, U.S.A.) with HGA-300 graphite furnace equipment (sensitivity 2 nM at a maximal volume of 100/~!). The presence of the iron chelates, i.e., pyrophosphate (0.4 raM) or ascorbate (2 raM) did not affect cell viability as assessed by leakage of lactate debydro. genase (LD) [22], content of ATP [23] or trypan blue exclusion. Each experiment (n > 3) was run in duplicate and the data presented in the Results are given as mean + S.D. Tne figures presented are of representative exper-
iments when not otherwise stated in the legends to the figures. Straight lines within figures are given as the best fit by linear regression analysis. Reagents. Diferric transferrin (absorbance ratio A466/A2a o = 0.045 + 0.001 (mean + S.D., n = 4) [24] labeled with 59Fe (specific activity 100 epm/pmol iron) or tZSl (specific activity 1500 cpm/pmol transferrin) were prepared as described in Refs. 25 and 26, respectively. Radiolabeled transferrin was dialyzed overnight prior to uptake studies to remove any low-molecularweight labeled transferrin fragments produced by radiolysis. 59Fe-labeled pyrophosphate (the molar ratio iron/pyrophosphate was 1:20) and 59Fe-labeled ascorbate (the molar ratio iron/ascorbate was 1 : 20) were prepared in deionized water just prior to use (specific activity 5 epm/pmol iron). S9FeCI3 (6.5-43.7 MBq/ retool) and BoRon and Hunter reagent (74 TBq/mmol) were from Amersham International (Buckinghamshire, U.K.). Collagenase (type IV), human transferrin (98%, essentially iron-free), Hepes, Tes, Tricine, pyrophosphate and bovine serum albumin (BSA) (essentially fatty add-free) were from Sigma (St. Louis, MO, U.S.A.). BSA was dialyzed 72 h against six changes of excess volumes of SPB buffer (see above). The calcium content of BSA was thus reduced by 90%, i.e., to 0.0004% (w/w) as assessed by atomic absorption spectrophotometry. Results
Effects of calcium on time-dependent iron uptake Hepatoeyte uptake of iron from [59Fe]transferrin, [59Fe]pyrophosphate or [59Fe]ascorbate at increasing concentrations of extracellular calcium was measured (Fig. 1A-C). In accordance wi{h earlier reports [1,2], the stimulating effect of calcium reached a level at 1-1.5 mM of calcium. At all calcium concentrations, iron uptake from the different iron donors proceeded linearly (i.e., r ~ 98, four measurements within 30 rain). Addition of calcium to cells incubated in a calcium free medium immediately and completely restored the ability of the cells to take up iron from transferrin (Fig. 2). Some 10% of BSA-bound calcium was unremovable by dialysis (see Methods). However, eeUular iron accumulation in a medium with undialyzed BSA (i.e., approx. 90 /~M calcium in the incubation medium) did not significantly differ from that in a medium with dialyzed BSA (Fig. 2). It thus appeared that BSA-assodated calcium did not stimulate cellular iron uptake. Some iron was taken up by the hepatocytes even in the caicium free medium (see Methods) (Figs. 1 and 2) and as for the control cells in the presence of calcium, more than 70% of this iron was recovered in eytosolie ferritin (as assessed by HPLC (data not shown)).
4l
C3
z
60
m~ z ~40 if: o_
o
A
0 300
(3~ 7
t 0
100 200 ~2~I-TRANSFERRIN] (riM) Fig. 3. The effect of calcium on cellular accumulation of tmnsferrin
Y
at increasing transferrirt concentratious. T h e hepatocytes were pro-
B
o
r r 4000 30O0
2000 lOOO
q
S
incubated 15 rain at 37°C in a calcium-free medium with (o, o) or without ( zx, A) 100 .aM EDTA. Followingfour washes in SPB, ceils were re-incubated at 4°C (closed w~nbols)or 3"PC(open symbols)in the calcium free medium.Control cells (120 •)were incubated in the presence of 1.2 mM of calcium throughout the experiment, l~llabeled transferrin was supplied to the re-incubated cells at the concentrations indicated. After60rain incubationat 4"C and after 30 rain incubation at 37°C the amounts of transferrin accumulated the cells were determined.
q
0
,
i
,
i
,
C
t 2 [Ca2+] (rnM) Fig. 1. The effect of increasing concentrationsof calcium on hepatocyte iron uptake. Hepatocytes were incubated with (A) [SVFe]transferrin (3 /zM iron), (B) [SgFe]pyrophosphate(20 ttM iron) or (C) [SgFe]ascorbate(1130/~M iron) at 37"C in media cor4aining calcium at the concentrations indicated. After 30 rain u~ iucubation duplicate aliquots were withdrawn and cellular iron uptake was determined.
Effects of calcium on transfen4n binding and transferrin endocytosis In reticulocytes Morgan showed that the receptors irreversibly lost the ability to bind transferrin after the cells had been incubated with E D T A at temperatures above 2ff'C [3]. To test such effects of calcium depriva-
-0.97. The activation energies for iron uptake given by the slopes of the lines were (A) from transferrin: 19.7 kcalmot -~ with calcium and 13.4 kcalmol - ! without calcium; (B) from the iroo-pyrophosphate complex as given from
2ooc
1ooc ..,,u
3.6
TEMPERATURE (~ x 10a) Fig. 6. Arrhenius plots of temperature-dependentrates of iron uptake with or withoutcalcium.The iron uptakedata (Fig. 5) with
400C- 3'3~ A
,
the right to the left (i.e., by increasingtem~ratare): 8.1, 18.8 and 14.3 kcalmol-t with calcium and 9.5 and 4.1 kcalmol-z without
ol
calcium and (C) from the iron-ascotbate complex as given from the right to the left: 30.2, 18.5 and 11.0 kcalmol - I with calcium and 12.5 and 4,6 kealmoi -1 without calcium.
o
5 cl 40x10~
g 20xtO~ rc
i
iJJ rc
.
i
.
i
.
°icZ
0.8x10~ 0.6x10~
0.4x10~
,0.2x10~, 0
10
20
3O
40
TEMPERATURE ('C} Fig, 5. The effect of calcium on the rates of hepatocyt¢ iron accumulation at increasing temperature. The hepatocytes were incubated with (A) [59Fe]transferrin (3 #M iron), (B) [SgFe]pyrophosphate (20 /zM iron) or (C) [SgFe]ascorbate (100 ,~M iron) (see Methods). The rates of iron uptake were measdred with ( o ) and without (e) 1.2 mM calcium in the temperature range of 4°C-38°C. The results are given as mean+S.D. ( n = 3). When no) indicated the S,D. values are within the size of the symbols. For all lines [n q and C, r ~_0.97.
phosphate (Fig. 5B) a second inflection was found at 27.0 + 0.2°C. In the calcium free medium, on the other hand, the rate of iron uptake from non.transferrin bound iron increased linearly through the entire temperature range (Fig. 5B and C) and the slight deviation from linearity in the temperature dependent rates of iron uptake from transferrin (Fig. 5A) simply reflected iron taken up by calcium-independent transferrin endocytosis (see Fig. 4). Fig. 6A-C shows the data of Fig. 5A-C in Arrhenius plots [27,281. As shown by the slopes of the lines no temperature dependent changes in the activation energies for iron uptake from transferrin were found either in the presence or the absence of calcium (Fig. 6A). A calcium dependent change in the activation energies for iron uptake from pyrophosphatc and ascorbate was, however, found at 27.5 + 1.1°C and 26.9 _+.2.0°C, respectively (Fig. 6B and (2). In addition, a second change in the activation energies for iron
43 4000 uJ "
~-
~ 3000
:"
1oo(]
38
tr
0
6 T4
,
I
100
,
I
200
,
300
BATE OF TRANSFERRINENDOCYTOSIS (molecules/cell/s)
Fig. 7. Rates of iron uptake related to rates of transferrinendocytosis at different temperatures. The rates of cellulariron uptake at differenttemperatureswith(o) and without(e) 1.2 mM calcium(as given in Fig. 5) were plotted against the correspondingrates of transferrin endocytosis(as given in Fig. 4). The temperatures are indicatedby the numbersnextto the symbols. uptake from pyrophosphate and ascorbate was found both in the presence and absence of calcium, i.e., at 16.0 4- 1.18°C and 15.0 4- 1.34°C in the absence of calcium and at 15.6 4- 0.6°C and 17.4 + 2.(FC in the presence of calcium for iron uptake from pyrophosphate and ascorbate, respectively. Only for calcium-dependent iron uptake from pyrophosphate the activation energy was found to be reduced at low temperatures (Fig. 613). Even if not the general rule, reductions in activation energies have been reported also for other transmembrane transport mechanisms [27]. However, the reason for this reduction in the activation energy and the concomitant low rates of iron uptake from pyrophosphate at low temperatures (Fig. 5B) is not apparent. In spite of that no calcium dependent reduction in the activation energy for iron uptake from transferrin was found at approx. 27*(], an effect of calcium near this temperature was still dearly shown by Fig. 7; i.e., when plotting the rates of iron uptake from transferrin at the different temperatures (data from Fig. 5) against the rates of transferrin ¢ndoeytosis at identical temperatures (data from Fig. 4), this figure showed a sharp calcium dependent increase in the rates of iron uptake above 25"C which was not reflected by a similar increase in transferrin ¢ndocytosis.
Discussion The results presented in this study show that extracellular calcium stimulates hepatocellular iron uptake from transferrin as well as iron uptake of non-transfertin bound ferrous and ferric iron (Fig. 1A-C). Maximal stimulation of iron uptake from transferrin and astorbate is observed at 1-1.5 mM of extracellular calcium.
Chelation of some caIcium by a surplus of pyrophosphate in the iro~-pyrophosphate solution may explain why iron uptake from the iron-pyrophasphate complex levels off at a somewhat higher concentration ot calcium. In accordance with earlier reports [2], uptake of iron from the ferrous-iron-ascorbate complex (Fig. 1) is also stimulated by calcium. Thus, the effect of calcium apparently is not on the reduction of ferric iron. Consistent with this interpretation is the finding that calcium has no effect on the plasma membrane ferricyanide reductase activity (data not shown). The complete and immediate restoration of iron uptake from transferrin upon calcium repletion (Fig. 2) shows that the hepatocyte transferrin receptors in contrast to the reticulocyte transferrin receptors [3], are not irreversibly impaired by calcium depletion. In fact no effect of calcium on either the receptor affinity for transferrin (see Results) or on the amount of cellular transferrin accumulated can be found (Fig. 3). To fully exclude any effect of calcium on the process of hepatocellular transferrin handling, the effect of calcium on the rate of transferrin endocytosis was investigated. The initial linearity for cellular transferrin accumulation (see Resutts) shows that exocytosis of labeled transferrin has not yet started and that a single process, i.e., endocytosis of transferrin, is measured. Since it thus is found that calcium does not affect the kinetics of transferrin endocytosis (Fig. 4), it appears that no step in the process of hepatocellular transferrin handling is calcium dependent. The finding that extracellular calcium stimulates iron uptake from all the iron donors investigated, indicates the involvement of a mechanism in the process of iron uptake which is common for the different iron donors, Even if it has been found that calcium promotes the binding of pyrophosphate to the plasma membrane of hepatocytes [29], the finding that calcium does not increase the membrane binding of transferrin (Fig. 3) shows that the stimulating effect of calcium on iron uptake is not on a general effect on the membrane binding of the iron donor complexes. However, transmembrane transport of iron mobilized outside the cell, i.e., at the plasma membrane or within endosomes as described for transferrin iron and pyrophosphate bound iron [1,5-12,29], may well be exerted by a calcium dependent mechanism general for all extracellularfy mobilized iron. An effect of calcium on a thermotropie membrane activity, is demonstrated by the calcium dependent reduction in the activation energy for iron uptake from both the iron-pyrophusphate complex and from the iron-ascorbate complex at approx. 27"C (Fig. 613 and C). The temperature dependent inflections in the rates of calcium-dependent iron uptake from pyropbosphate and ascorbate (Fig. 5B and C) and the calcium depen-
44 dent changes in activation energies (Fig. 6B and (2) which all correspond closely with the upper and lower thermotropic lipid phase transitions reported for the hepatocyte plasma membrane [30,31], further support that we are dealing with a calcium-regulated iron transmembrane transport mechanism. The absence of sharp temperature dependent inflections in the rate of calcium-dependent iron uptake from transferrin and the concomitant calcium dependent change in the activation energy (Figs. 5A and 6A), may be explained by a more complex process of iron release from transferrin than from the other iron donors. As a rate limiting step the iron release from transferrin may thus conceal calcium regulating effects on the iron transmembrane transport mechanism [28]. A calcium dependent influence on iron uptake from transferrin at the upper thermotroplc lipid phase transition is, however, clearly demonstrated in Fig. 7. This figure shows that above 25°C, calcium primarily stimulates the iron uptake which is independent of transferrin endocytosis. This finding is interpreted to mean that despite the rate limitations by iron release from transferrin, calcium stimulates iron uptake from transferrin the same way as for iron uptake from pyrophosphate and ascorbate, i.e., by an effect of calcium on the thermotropic iron transmembrane transport mechanism. Two additional findings of this study require some further comments; firstly, the temperature dependent inflection in the rates of calcium-dependent iron uptake from pyrophosphate at the higher temperature (Fig. 5B) (also seen in iron uptake from NTA, data not shown), can not be found for calcium-dependent iron uptake from ascorbate (Fig. 5C) (or for uptake of cobalt, data not shown). Since preliminary studies show no concomitant change in the ferricyanide reductase activity, this discrepancy in iron uptake of ferrous and ferric iron donors ma~, well be in the transmembrane transport process. Despite the claim that biological membranes are impermeable to the ferric iron ion and iron uptake from ferric iron donors thus depends upon reduction [8], it is suggested that the abundance of free fatty acids in the hepatocyte plasma membrane may allow for an additional transmembrane transport of ferric iron [32]. Lipid dependent transloeation of ferric iron ions may, however, depend on the more optimal constellations of membrane components, only to be obtained at temperatures above the upper lipid phase transition [311 and in the presence of calcium [17]. Secondly, since the iron accumulated by hepatocytes in the calcium free medium (Figs. 1, 2 and 5) is recovered in ferritin (see Results), this iron apparently is transported over the plasma membrane. Because viable ceils inevitably will release some calcium to the medium, iron uptake in the complete absence of calcium is difficult to document. However, if the iron
accumulated in the calcium free medium is translocoted to cytosol by the calcium-dependent mechanism running slow on small amounts of calcium redistributed to the membrane from intracellular stores, calcium induced changes in rates of iron uptake (Fig. 5B and C) are expected. The absence of such changes indicates a discrete mechanism for calcium independent transmembrane trans~rt of iron. This interpretation is further supported by the observed differences in the activation energies in the presence and absence of supplied calcium (Fig. 6A-C). However, at physiological temperatures iron accumulation by hepatocytes is strongly dominated by the calcium-dependent iron uptake mechanism. Whether this mechanism for the calcium-dependent transmembrane transport of iron is equally expressed in cells of different tissue is not known. As a scavenger of iron in times of noxious iron overload this high capacity iron uptake mechanism of the liver, may well be a highly developed function of this tissue. An answer to this question depends on future investigations.
Acknowledgements This work was supported by grants from the University of Trondheim. The assistance of med. lab, tech. I. Arbo and the helpful discussions with Dr. L Romslo and Dr. K. Thorstensen are gratefully acknowledged.
References 1 Thorstensen,IL and Romsln,I. U988)L Biol.Chem.263,88448850. 2 Wright,T.L, Brissot,P., Ma, W-L.and Weisiger,R.A. (1986)J. Biol. Chem.261, 10909-10914. 3 Morgan,E.H. (1989)Biochim.Biophys.Acta 981, 121-129. 4 Thorstensen,K. and Romslo,1. (1987)Scand.J. Ctin.Invest.4"/, 837-846. 5 Morley,C.G.D. and Bezkorovainy,A. (1985)Int. L Biochem. 17, 553-564. 6 Thorstensen,K. 0988) J. Biol.Chem. 263, 16837-16841. 7 Cfichlon,R.R, (1990)Adv. ProteinChem, U S A 40, 281-363. 8 De Jong O., Van Dijk,J.P.and Van Eijk,H.O. (1990)Clin.Chim. Acta 190, 1-46. 9 Bezkorovainy,A (1989)Clin.PM'siol.Biochem. 7, 1-17. IO Sun, I.L,Navas,P.,Crane, EL, Mor#, DJ. and Low, H. (1987) J. Biol.Chem. 262, 15915-15921. 11 Baker, E., McArdle, H.I. and Morgan, E.H. (1985)in Proteinsof
Iron Storage and Transport(Spik, G., Montreuil,J., Crichton, R.R. and Mazurier,J., eds.),pp. 131-142,ElsevierSciencePublishers,Amsterdam. 12 Dautry-Varsat,A., Ciechanover,A. and Lodish,H.F.(1982t)Prec. Natl. Acad. Sci. USA 80, 2258-2262. 13 Eaton,J.E. (1987)KidneyInt. 32, S68-$76. 14 McCormack, J.O., Halestrap, A.P. and Denton, R,M. (1990) PltysioLRev.70, 421-509. 15 PaL C.Y.C. (1990)in Fluids and EleclroMes(Kokko,LP. and Tannen,R.L, eds.), pp. 596-630, SaundersCompany,Philadelphia, 16 Livingstone,C.J. and Schachter, D. (1980) Biochemistry 19, 4823-4827.
45 17 Benga, G. and Holmes, R.P. (19M) Prog. Biophys. Mot. Biol. 43, 195-257. 18 Speetor, A.A. and Yorek, M.A. (1985) J. Lipid Res. t.~, t015-1035. 19 Mills, P.R., Ide[er, P.J., Smith, D.J., Ballatori, N., Boyer, J.L. and Gordon, E,R. (1987) Hepatology 7, 61-66. 20 Seglen, P.O. (1976) Methods Cell Biol. 13, 29-83. 21 Bradford, M.M. (1976) Anal. Biechem. 7Z, 284-254. 22 Empfehlungen der Deutschen Gesellshaft fur tedinische Chemie (1977) J. Clin. Chem. Clin. Biochem. 15, 249-254. 23 Ingebretsen, O.C., Bakken, A.M,, Segadal, L. and Farstad, M. (1982) J. Chromatogr. 2,12, 119-126. 24 Huebers, H.A., Csiba, E., Huebers, E. and Finch, C.A. (t983) Prec. Natl. Aead. Sei. USA 80, 300-3t,~4.
25 Konopka. K. and Romslo. I. (1981) Eur. J. Biochem. 117. 239-244. 26 Bolton. A.E. and Hunter, W.M. (1973) Biochem. J. 133. 529-539. 27 McEIhaney. R.N. (1982) Current Top. Membr. Transport 17, 317-380. 28 Raison, J.K. (1973) Bioenergetics 4, 285-309. 29 Nilsen, T and Romslo, I. (1990) Seand. J. Clin. Invest. 50, |9-25. 30 Livingstone, C.J. and Schachter, D. (1980) J. Biol. Chem. 255. 10902-10908. 31 Schachter. D. (1984) Hepatolo~ 4, 140-151. 32 Peters, T.J., Raja, K.B.. Simpson, R.J. and Snape. S. (1988) Ann. N.Y. Acad. Sci. 526, 141-147.
