- Email: [email protected]
Iron and phosphorus content of rabbit liver ferritin fractions with different subunit composition
Int. J. Biochem. Vol. 17, No. 3, pp. 421-424, 1985
0020-711X/85 $3.00 + 0.00 Copyright (~) 1985 Pergamon Press Ltd
Printed in Great Britain. All rights reserved
IRON AND PHOSPHORUS CONTENT OF RABBIT LIVER FERRITIN FRACTIONS WITH DIFFERENT SUBUNIT COMPOSITION ERIK J. FRENKEL*,BEP VANDEN BELD and JOANNESJ. M. MARX Departments of Haematology and Internal Medicine, University Hospital Utrecht, The Netherlands [Tel. 030-372547] (Received 25 April 198,~)
Abstract--l. After isolation of subtypes of rabbit liver-ferritin, the phosphorus to iron ratio (P/Fe-ratio) in the samples was found to parallel the shift in subunit composition of the two types of subunits of which ferritins generally consist. 2. No relation was found with the amount of iron per ferritin molecule. 3. The increase of the P/Fe-ratio, in relation to subunit composition, is postulated to be a result of the change of total surface area of the iron microcrystallites inside the ferritin molecule. 4. This surface area depends on the number of nucleation points and thereby may be dependent on the subunit composition.
INTRODUCTION
The iron storage protein ferritin has an apparent molecular weight of 450,000 dalton. It is composed of 24 rather similar subunits (Drysdale, 1977), which can be divided into two types: a heavy type (H) subunit (mol.wt 21,000 dalton), predominantly found in heart tissue and a light type (L) subunit (mol.wt 19,000 dalton), mainly found in liver tissue. These subunits compose a hollow sphere that can accommodate up to 4500 iron atoms per molecule of ferritin (Clegg et al., 1980). The core is generally believed to consist of a polynuclear, ferric oxyhydroxidephosphate complex (Granick and Hahn, 1944; Spiro and Saltman, 1969; Harrison and HoT, 1973; Harrison et al., 1974), exhibiting a chemical formula: (FeOOH)8(FeO:OPO3H2). This formula is based on the molar ratio of phosphorus to iron (P/Fe-ratio), which was believed to reveal a constant and reproducible value of 0.11 (Granick and Hahn, 1944; Mazur et al., 1950). This value holds true for horse spleen ferritin, which is classically used for investigations on ferritin and which contains a high iron content, close to the maximum of 4500 ions per ferritin molecule. In recent years some doubts were thrown on the role of phosphorus in the iron core of the ferritin molecule. Myagkaya and de Bruijn (1982A) found various P/Fe-ratios of ferritins with different iron content, isolated from different organs of human and animal origin, using X-ray microanalysis in situ. They confirmed the P/Fe-ratio of 0. l 1 in horse spleen ferritin. In some ferritin samples, however, no phosphorus could be detected at all. This makes it very unlikely that the above mentioned formula of the ferric oxyhydroxide-phosphate complex is correct. It even has to be questioned whether phosphorus is part of the atomic structure of the microcrystalline iron* Present address: Medical Chemical Institute, Berne University, Postfach, 3000 Bern 9, Switzerland. Reprint requests should be sent to: Dr J. J, M. Marx, Department of Haematology, University Hospital, P.O. Box 16250, 3500 CG Utrecht, The Netherlands.
core of ferritin. This question was first raised by van Kreel et al. (1972), who suggested, based on liver perfusion studies, that the inorganic phosphate of the liver tissue is in a state of dynamic equilibrium with the phosphate of ferritin fractions with variable Fe/N ratio. In a more recent study Treffry and Harrison (1978) compared the properties of ferritin reconstituted with inorganic phosphate and iron together in the reaction mixture, with a reconstituted ferritin in which Pi was added after Fe in a separate step. Properties of these ferritins were compared with those of native ferritin and of native ferritin that had been incubated with [32p]pi. They concluded that much of the Pi of native ferritin is adsorbed on surfaces of ferritin iron-core crystallites. It is known that the number of iron crystallites is variable, probably depending on the amount of nucleation points for the formation of iron oxyhydroxide complexes (Clegg et al., 1980). Iron entering the ferritin core through the 6 different channels will attach to the surface of these micelles in a "first in, last out" fashion. Inorganic phoshate may be incorporated into the iron core during this process or may bind to the surfaces of the microcrystallites at a later stage when the iron content of the molecule has stabilized. We have investigated the influence of the molecular structure of ferritin, based on its subunit composition, on the amount of phosphorus relative to iron. After separation of rabbit liver ferritin into its composing molecular forms by high performance liquid chromatography on a chromatofocussing column (Frenkel et al., 1983B), we determined the iron content and the molar ratio of phosphorus to iron of the fractions obtained. The results of these investigations showed that the phosphorus to iron ratio of fractions which were highly enriched with distinct subtypes of ferritin, exhibits a direct dependence on the subunit composition of the ferritin molecule. The iron content itself was not dependent on the subunit composition. We believe that the amount of P~ bound to the iron depends on the concentration of inorganic phosphate of the medium and on the total surface of iron crystallites inside the ferritin molecule.
421
422
ERIKJ. FRENKELet al. described before (Frenkel et al., 1983A), using urea, ammonium sulphate and bovine serum albumin as standards. Phosporus was assayed according to a modification of the procedure of Fiske Subbarow (B0ttcher et al., 1961), using disodium hydrogen phosphate as a reference. All chemical assays were performed in duplicate. Means and standard deviations were calculated from at least three separate analysis runs on the same sample.
M A T E R I A L S AND M E T H O D S
Fresh rabbit livers were obtained from the National Institute of Public Health, R1V, Bilthoven, The Netherlands. All chemicals used were analytical grade and purchased either from BDH Chemicals Ltd, Poole, U.K., or from Merck, Darmstadt, F.R.G. Purification o f rabbit liver fi'rritin
Ferritin was isolated essentially as described before (Frenkel et al., 1983a). In short, fresh rabbit livers or rabbit livers which had been stored for not more than fourteen days at - 60°C were homogenized at 4°C. The ferritin was subsequently isolated at 4°C by: (i) centrifugation of the homogenate (20 rain; 3500 .8); (it) filtration of the supernatant through Whatman GF/D filters; (iii) ammonium sulphate precipitation (611'7osaturation) and centrifugation (10 rain; 3500 .8); (iv) dissolution of the precipitate in 0.01 M sodium phosphate buffer (pH 7.4), containing 11.15 M sodium chloride; (v) immunoaffinity chromatography using goat anti-serum to rabbit liver ferritin, coupled to CNBractivated Sepharose 4B; (vi) elution of the ferritin with 3 M potassium thiocyanate; (vii) extensive dialysis against 511 mM Tris, pH 7.4; (viii) size exclusion chromatography on Seraphose CL-6B, eluting with 50 mM Tris, pH 7.4. The purity of all preparation used was confirmed by gradient polyacrylamide gel electrophoresis, followed by staining identical gels for iron and for protein. Thereby the same bands were found to stain with both techniques. No other protein staining bands could be detected (Frenkel et al., 1983A). Isolation o f distinct forms of rabbit liver ferritin
Fractions with a high enrichment with distinct molecular forms of rabbit liver ferritin were obtained by chromatofocussing of a sample of purified ferritin on Mono P, using high performance liquid chromatography equipment from Pharmacia, Uppsala, Sweden (fast protein liquid chromatography - - FPLC) as described before (Frenkel et al., 1983B). The fractions obtained exhibited a subunit ratio, ranging from 87% L-subunit in the first fraction to 28% L-subunit in the last fraction. This gradual change in subunit composition also was reflected in the isoelectric point of the preparations as shown before (Frenkel et al., 1983B). Chemical iron, nitrogen and phosphorus determinations
Iron was measured spectrophotometrically in the samples at 530 nm using the reaction with bathophenanthroline (Schilt, 19691. As a standard ammonium ferrous sulphate was used. The nitrogen determinations were carried out as
RESULTS AND DISCUSSION
A f t e r purification of rabbit liver ferritin with i m m u n o a f f i n i t y c h r o m a t o g r a p h y and separation of the ferritin p r e p a r a t i o n into fractions with distinct subunit ratio, the a m o u n t s of iron, nitrogen and p h o s p h o r u s were d e t e r m i n e d in the starting material and all fractions. S u b s e q u e n t l y the m o l a r ratios of iron to n i t r o g e n , p h o s p h o r u s to iron, a n d p h o s p h o r u s to n i t r o g e n were calculated as s u m m a r i z e d in T a b l e 1. T h e table also shows the iron load per ferritin molecule calculated for the various subtypes of ferritin. It should be n o t e d that the diffcrences in percentages of L-subunits b e t w e e n F1 and F4 are much greater t h a n b e t w e e n F4 a n d F6. T h e different fractions were not quantitatively equivalent. T h e largest part of the original sample was c o n c e n t r a t e d in fractions F3 a n d F4. This explains why the calculated ratio of the starting material exhibited values in the range b e t w e e n F3 and F4. T h e differences in the a m o u n t of iron a t o m s per ferritin molecule, ranging from 670 to 3020 in fractions of the same liver ferritin p r e p a r a tion, were r e m a r k a b l e . T h e r e was no relation with the subunit c o m p o s i t i o n of the ferritin, a l t h o u g h the two fractions with the highest p e r c e n t a g e of L-subunits c o n t a i n e d the least iron. These results are not in a g r e e m e n t with results from B o m f o r d et al. (1978, 1981), w h o f o u n d that ferritin fractions with the highest iron c o n t e n t consist in general of the most basic ferritin forms with p r e d o m i n a n t l y L-subunits, which are mainly f o u n d in liver tissue. A n increasing iron c o n t e n t would correlate with a shift of ferritins rich in H - s u b u n i t s to those rich in L-subunits ( A r o s i o et al., 1977: B o m f o r d et al., 198l). O n the o t h e r h a n d the iron c o n t e n t of the protein has no effect on its (soelectric point, which is strongly related with the subunit composition of ferritin ( U r u s h i z a k i et al., 1971). It is possible that the purification p r o c e d u r e in the
Fable I. C o m p a r i s o n of the subunit composition, the iron load and the iron to nitrogen, phosphorus to iron and phosphorus to nitrogen molar ratios in a sample of purified rabbit liver fcrritin (S) and in six fractions obtained after separation of that sample into fractions, highly enriched to ferritins with a distinct subunit composition (FI -F6) Fraction
% 1."
S F1 F2 F3 F4 F5 F6
46.11 81 .l/ 67.3 52.6 41.2 35. I 32.4
Chemical analyses
Fe/t:T~'
I:c/N
St)
P/Fe
SI)
PiN
St)
11.55 I). 13 11.13 il.57 1t.54 0.34 0.41
IL01 0.04 1t.I11 11.112 0.01 0.02 0.115
0.090 I). 1311 II. 117 t).tl93 tl.0811 0.070 11.1)61)
1L007 11.1130 0.011 0.010 0.007 II.l) 12 0.036
I).11511 I).1)21/ 11,1114 [).1157 0,043 0.II27 0.029
I).1)t14 /I.006 1L003 11.[113 11.004 0.0116 0.012
29~2
67(I 67"; 3020 2892 183 I 2214
The values of all chemical analyses were obtained as m e a n wdues of at least three separate analysis runs, p e r f o r m e d in duplicate. The values for the relative subunit composition of the samples were obtained by gelscanning of polyacrylamide gradient gels, as described before (Frenkel et a l . , 1983b). " The relative subunit composition is expressed as the percentage L-subunits present in the sample. The other part is composed of tt-lvpc subnnits. ~' Expressed is the n u m b e r of iron atoms per fcrritin molecule (Ft)
Ferritin iron and phosphorus above mentioned studies, which included heat treatment, has led to erroneous results. We have shown before that the classical heat treatment influences some biochemical and physiological properties of ferritin (Frenkel etal., 1983a, 1983c); in particular a loss of H-subunit rich ferritin was noticed. In the present study a gentle purification method was used with omission of heat treatment. Before discussing the results of our calculation of the P/Fe-ratios some points have to be made. Because many authors use various definitions for the ratio of, for instance, phosphorus to iron in ferritin preparations, we feel the need for a clear definition of the ratio used in this report, to be able to prevent ambiguity. Because of its relationship with the molecular formula we use the molar ratio (not the weight ratio). We divide the amount of phosphorus present by the amount of iron to obtain numbers, lower than the unity. On speaking about an increasing ratio, we mean that the number, obtained by performing this division increases (not that the difference in absolute value of numerator and d e n o m i n a t o r increases). In the literature the iron-containing core of ferritin often is described as a polynuclear ferric oxyhydroxide-phosphate complex, with the chemical formula: ( F e O O H ) s ( F e O : O P O 3 H z ) , exhibiting a constant phosphorus to iron ratio of 0.11 (Mazur et al., 1950). As mentioned in our introduction this ratio was obtained from horse spleen ferritin, and the value of 0.11 was confirmed by several authors. The P/Fe-ratio, however, appeared to be by no means a constant value when determined in ferritin from other origin (Myagkaya and de Bruijn, 1982a,b). In our starting material we found a P/Fe-ratio of 0.090 + 0.007 (m + SD). The fractions from this ferritin preparation, however, exhibited P/Fe-ratios in the range from 0.060 to 0.130. We found a significant correlation between the percentage of L-subunits (x) and the P/Fe-ratio (y) in the isolated fractions with the following equation for linear regression: y = 0.00135x + 0.02309 (r = 0.71, P < 0.001). Assuming complete linearity of this curve the P/Fe-ratio would be 0.023 in ferritin containing exclusively H-subunits, and 0.158 in ferritin with 100% L-subunits. In our experiments conditions for the different fractions were identical, apart from the p H of the polybuffer during elution of fractions from the chromatofocussing column. Therefore it is evident that no differences have existed in contamination with inorganic phosphate between the distinct ferritin forms from the m o m e n t of their synthesis in the rabbit liver, during incorporation of iron, until the final samples were obtained. Because the ferritin iron core contains a variable amount of iron-microcrystallites, which grow from nucleation points on the inner surface of the apoferritin molecule, the total surface of these microcrystallites will vary considerably, and consequently the amount of inorganic phosphate, able to bind superficially to the iron complexes. In L-subunit rich ferritin the P/Fe-ratio is higher. This is compatible with a larger total surface of the microcrystallites inside the ferritin protein shell, implicating a greater number of microcrystals stored.
423 SUMMARY
Rabbit liver ferritin was isolated by immunoaffinity chromatography. After separation of the preparation into fractions with a high enrichment with distinct subtypes of ferritin, we found a gradual increase in the phosphorus to iron ratio (P/Fe-ratio) that paralleled the shift from high molecular weight (H) to low molecular weight (L) subunits in the isolated ferritin forms. The iron content of these fractions, ranging from 670 to 3020 iron ions per ferritin molecule, was not related to the subunit composition. Our results indicate that phosphorus is not a structural part of the micellar iron core of ferritin. We postulate that most of the inorganic phosphate (Pi) is reversibly attached to the surfaces of the iron microcrystallites inside the ferritin molecule. The increase of the P/Fe-ratio in relation to the relative proportion of L-subunits may be a result of an increase of the total surface area of the multiple iron microcrystallites. The extent of this surface depends on the number of intramolecular microcrystallites. L-subunit rich ferritin probably contains, therefore, a greater number of nucleation points for formation of iron complexes than H-subunit rich ferritin, which may lead to a larger number of iron microcrystallites to which phosphorus can bind.
Acknowledgements - - We wish to thank the National Institute of Public Health, RIV., for providing us with fresh rabbit livers, and we are grateful to Drs M. 1. Cleton and W. C. de Bruijn for their valuable advice. The foundation for Biological Research in The Netherlands, BION. which is supported by The Netherlands Organization for the Advancement of Pure Research, ZWO, is acknowledged for financial support (Grant No. 14-96--4)09).
REFERENCES
Arosio P., Yokata M. and Drysdale J. W. (1977) Characterization of serum ferritin in iron overload: possible identity of natural apoferritin. Br. J. Haematol. 36,199-207. Bomford A., Berger M., Lis Y. and Williams R. (1978) The iron content of human liver and spleen isoferritins correlates with their isoelectric point and subunit composition. Biochem. biophys. Res. Commun. 83,334-341. Bomford A., Conlon-Hollingshead C. and Munro H. N. (1981) Adaptive responses of rat tissue isoferritins to iron administration. Changes in subunit synthesis, isoferritin abundance and capacity for iron storage. J. biol. Chem. 256,748--955. B6ttcher C. J. F., van Gent C. M. and Pries C. (1961) A rapid and sensitive sub-micro phosphorus determination. Anal. chim. Acta 24, 203-204. Clegg G. A., Fitton J. E., Harrison P. M. and Treffry A. (1980) Ferritin: molecular structure and iron storage mechanisms. Progr. Biophys. molec. Biol. 36, 53-86. Drysdale J. W. (1977) Ferritin phenotypes: structure and metabolism. In Iron Metabolism, Ciba Foundation Symposium Series 51, pp. 41-47. New Series Elsevier/ Exerpta Medica/North Holland, Amsterdam. Frenkel E. J., van den Beld B., van Oost B. A. and Marx J. J. M. (1983a) Influence of heat treatment on rabbit liver ferritin. Biochim. biophys. Acta 745, 202-208. Frenkel E. J., van den Beld B., K6nig B. W. and Marx J. J.
424
ERIK J. FRENKELet al.
M. (1983b) Preparative isolation of distinct molecular forms of rabbit liver ferritin using high performance liquid chromatography. Analyt. Biochem. 135,489-494. Frenkel E. J., van den Beld B. and Marx J. J. M. (1983c) Influence of heat-treatment on rabbit liver-ferritin kinetics. Br. J. Haematol. 55,449-453. Granick S. and Hahn P. F. (1944) Ferritin: VIII. Speed of uptake of iron by the liver and its conversion to ferritin iron. J. biol. Chem. 155,661~69. Harrison P. M. and Hoy T. G. (1973) Ferritin. In Inorganic Biochemistry (Edited by Eichhorn G. G.) pp. 253-279. Elsevier, New York. Harrison P. M., Hoy T. G., Macara I. G. and Hoare R. J. (1974) Ferritin iron uptake and release. Structurefunction relationships. Biochem. J. 143,445-451. Van Kreel B. K., Pijnenburg A. M. C. M., Van Eiik H. G. and Leijnse B, (1972) The incorporation of ~gFe, 3H . leucine and 3-~P morgamc phosphate in ferritin during the perfusion of isolated rat livers. Biochim. biophys. Acta 273,243-247. Mazur A., Litt 1. and Shorr E. (1950) Chemical properties of ferritin and their relation to its vasodepressor activity.
J. biol. Chem. 187,473-484. Myagkaya G. L, and de Bruijn W. C. (1982a) X-ray microanalysis of cellular localization of ferritin in mammalian spleen and liver. Micron 13, 7-21. Myagkaya G. L. and De Bruijn W. C. (1982b) Ferritin ill chorionic villi of the sheep placcnta. Bihl. Anat. 22, 117-222. Schilt A. A. (1969) In Analytical Applieations of l, lO-Phenanthroline and Related compounds. International Series of Monographs in Ana@tical Chemistry 32 (Edited by Belcher R. and Freiser M.) pp. 59-62. Pergamon Press, Oxford. Spiro T. G. and Sattman P. (1969) Polynuclcar complexes of iron and their biological implications. Struct. Bond. 6, 116-156. Treffry A. and Harrison P. M. (1978) Incorporation and release of inorganic phosphate in horse spleen ferritin. Biochem. J. 171,313-320. Urushizaki I., Niitsu Y., Ishitani K., Matsuda M. and Fukuda M. (1971) Microheterogencity of horse spleen ferritin and apoferritin. Bioehim. biophys. Aeta 243, 187-192.
0020-711X/85 $3.00 + 0.00 Copyright (~) 1985 Pergamon Press Ltd
Printed in Great Britain. All rights reserved
IRON AND PHOSPHORUS CONTENT OF RABBIT LIVER FERRITIN FRACTIONS WITH DIFFERENT SUBUNIT COMPOSITION ERIK J. FRENKEL*,BEP VANDEN BELD and JOANNESJ. M. MARX Departments of Haematology and Internal Medicine, University Hospital Utrecht, The Netherlands [Tel. 030-372547] (Received 25 April 198,~)
Abstract--l. After isolation of subtypes of rabbit liver-ferritin, the phosphorus to iron ratio (P/Fe-ratio) in the samples was found to parallel the shift in subunit composition of the two types of subunits of which ferritins generally consist. 2. No relation was found with the amount of iron per ferritin molecule. 3. The increase of the P/Fe-ratio, in relation to subunit composition, is postulated to be a result of the change of total surface area of the iron microcrystallites inside the ferritin molecule. 4. This surface area depends on the number of nucleation points and thereby may be dependent on the subunit composition.
INTRODUCTION
The iron storage protein ferritin has an apparent molecular weight of 450,000 dalton. It is composed of 24 rather similar subunits (Drysdale, 1977), which can be divided into two types: a heavy type (H) subunit (mol.wt 21,000 dalton), predominantly found in heart tissue and a light type (L) subunit (mol.wt 19,000 dalton), mainly found in liver tissue. These subunits compose a hollow sphere that can accommodate up to 4500 iron atoms per molecule of ferritin (Clegg et al., 1980). The core is generally believed to consist of a polynuclear, ferric oxyhydroxidephosphate complex (Granick and Hahn, 1944; Spiro and Saltman, 1969; Harrison and HoT, 1973; Harrison et al., 1974), exhibiting a chemical formula: (FeOOH)8(FeO:OPO3H2). This formula is based on the molar ratio of phosphorus to iron (P/Fe-ratio), which was believed to reveal a constant and reproducible value of 0.11 (Granick and Hahn, 1944; Mazur et al., 1950). This value holds true for horse spleen ferritin, which is classically used for investigations on ferritin and which contains a high iron content, close to the maximum of 4500 ions per ferritin molecule. In recent years some doubts were thrown on the role of phosphorus in the iron core of the ferritin molecule. Myagkaya and de Bruijn (1982A) found various P/Fe-ratios of ferritins with different iron content, isolated from different organs of human and animal origin, using X-ray microanalysis in situ. They confirmed the P/Fe-ratio of 0. l 1 in horse spleen ferritin. In some ferritin samples, however, no phosphorus could be detected at all. This makes it very unlikely that the above mentioned formula of the ferric oxyhydroxide-phosphate complex is correct. It even has to be questioned whether phosphorus is part of the atomic structure of the microcrystalline iron* Present address: Medical Chemical Institute, Berne University, Postfach, 3000 Bern 9, Switzerland. Reprint requests should be sent to: Dr J. J, M. Marx, Department of Haematology, University Hospital, P.O. Box 16250, 3500 CG Utrecht, The Netherlands.
core of ferritin. This question was first raised by van Kreel et al. (1972), who suggested, based on liver perfusion studies, that the inorganic phosphate of the liver tissue is in a state of dynamic equilibrium with the phosphate of ferritin fractions with variable Fe/N ratio. In a more recent study Treffry and Harrison (1978) compared the properties of ferritin reconstituted with inorganic phosphate and iron together in the reaction mixture, with a reconstituted ferritin in which Pi was added after Fe in a separate step. Properties of these ferritins were compared with those of native ferritin and of native ferritin that had been incubated with [32p]pi. They concluded that much of the Pi of native ferritin is adsorbed on surfaces of ferritin iron-core crystallites. It is known that the number of iron crystallites is variable, probably depending on the amount of nucleation points for the formation of iron oxyhydroxide complexes (Clegg et al., 1980). Iron entering the ferritin core through the 6 different channels will attach to the surface of these micelles in a "first in, last out" fashion. Inorganic phoshate may be incorporated into the iron core during this process or may bind to the surfaces of the microcrystallites at a later stage when the iron content of the molecule has stabilized. We have investigated the influence of the molecular structure of ferritin, based on its subunit composition, on the amount of phosphorus relative to iron. After separation of rabbit liver ferritin into its composing molecular forms by high performance liquid chromatography on a chromatofocussing column (Frenkel et al., 1983B), we determined the iron content and the molar ratio of phosphorus to iron of the fractions obtained. The results of these investigations showed that the phosphorus to iron ratio of fractions which were highly enriched with distinct subtypes of ferritin, exhibits a direct dependence on the subunit composition of the ferritin molecule. The iron content itself was not dependent on the subunit composition. We believe that the amount of P~ bound to the iron depends on the concentration of inorganic phosphate of the medium and on the total surface of iron crystallites inside the ferritin molecule.
421
422
ERIKJ. FRENKELet al. described before (Frenkel et al., 1983A), using urea, ammonium sulphate and bovine serum albumin as standards. Phosporus was assayed according to a modification of the procedure of Fiske Subbarow (B0ttcher et al., 1961), using disodium hydrogen phosphate as a reference. All chemical assays were performed in duplicate. Means and standard deviations were calculated from at least three separate analysis runs on the same sample.
M A T E R I A L S AND M E T H O D S
Fresh rabbit livers were obtained from the National Institute of Public Health, R1V, Bilthoven, The Netherlands. All chemicals used were analytical grade and purchased either from BDH Chemicals Ltd, Poole, U.K., or from Merck, Darmstadt, F.R.G. Purification o f rabbit liver fi'rritin
Ferritin was isolated essentially as described before (Frenkel et al., 1983a). In short, fresh rabbit livers or rabbit livers which had been stored for not more than fourteen days at - 60°C were homogenized at 4°C. The ferritin was subsequently isolated at 4°C by: (i) centrifugation of the homogenate (20 rain; 3500 .8); (it) filtration of the supernatant through Whatman GF/D filters; (iii) ammonium sulphate precipitation (611'7osaturation) and centrifugation (10 rain; 3500 .8); (iv) dissolution of the precipitate in 0.01 M sodium phosphate buffer (pH 7.4), containing 11.15 M sodium chloride; (v) immunoaffinity chromatography using goat anti-serum to rabbit liver ferritin, coupled to CNBractivated Sepharose 4B; (vi) elution of the ferritin with 3 M potassium thiocyanate; (vii) extensive dialysis against 511 mM Tris, pH 7.4; (viii) size exclusion chromatography on Seraphose CL-6B, eluting with 50 mM Tris, pH 7.4. The purity of all preparation used was confirmed by gradient polyacrylamide gel electrophoresis, followed by staining identical gels for iron and for protein. Thereby the same bands were found to stain with both techniques. No other protein staining bands could be detected (Frenkel et al., 1983A). Isolation o f distinct forms of rabbit liver ferritin
Fractions with a high enrichment with distinct molecular forms of rabbit liver ferritin were obtained by chromatofocussing of a sample of purified ferritin on Mono P, using high performance liquid chromatography equipment from Pharmacia, Uppsala, Sweden (fast protein liquid chromatography - - FPLC) as described before (Frenkel et al., 1983B). The fractions obtained exhibited a subunit ratio, ranging from 87% L-subunit in the first fraction to 28% L-subunit in the last fraction. This gradual change in subunit composition also was reflected in the isoelectric point of the preparations as shown before (Frenkel et al., 1983B). Chemical iron, nitrogen and phosphorus determinations
Iron was measured spectrophotometrically in the samples at 530 nm using the reaction with bathophenanthroline (Schilt, 19691. As a standard ammonium ferrous sulphate was used. The nitrogen determinations were carried out as
RESULTS AND DISCUSSION
A f t e r purification of rabbit liver ferritin with i m m u n o a f f i n i t y c h r o m a t o g r a p h y and separation of the ferritin p r e p a r a t i o n into fractions with distinct subunit ratio, the a m o u n t s of iron, nitrogen and p h o s p h o r u s were d e t e r m i n e d in the starting material and all fractions. S u b s e q u e n t l y the m o l a r ratios of iron to n i t r o g e n , p h o s p h o r u s to iron, a n d p h o s p h o r u s to n i t r o g e n were calculated as s u m m a r i z e d in T a b l e 1. T h e table also shows the iron load per ferritin molecule calculated for the various subtypes of ferritin. It should be n o t e d that the diffcrences in percentages of L-subunits b e t w e e n F1 and F4 are much greater t h a n b e t w e e n F4 a n d F6. T h e different fractions were not quantitatively equivalent. T h e largest part of the original sample was c o n c e n t r a t e d in fractions F3 a n d F4. This explains why the calculated ratio of the starting material exhibited values in the range b e t w e e n F3 and F4. T h e differences in the a m o u n t of iron a t o m s per ferritin molecule, ranging from 670 to 3020 in fractions of the same liver ferritin p r e p a r a tion, were r e m a r k a b l e . T h e r e was no relation with the subunit c o m p o s i t i o n of the ferritin, a l t h o u g h the two fractions with the highest p e r c e n t a g e of L-subunits c o n t a i n e d the least iron. These results are not in a g r e e m e n t with results from B o m f o r d et al. (1978, 1981), w h o f o u n d that ferritin fractions with the highest iron c o n t e n t consist in general of the most basic ferritin forms with p r e d o m i n a n t l y L-subunits, which are mainly f o u n d in liver tissue. A n increasing iron c o n t e n t would correlate with a shift of ferritins rich in H - s u b u n i t s to those rich in L-subunits ( A r o s i o et al., 1977: B o m f o r d et al., 198l). O n the o t h e r h a n d the iron c o n t e n t of the protein has no effect on its (soelectric point, which is strongly related with the subunit composition of ferritin ( U r u s h i z a k i et al., 1971). It is possible that the purification p r o c e d u r e in the
Fable I. C o m p a r i s o n of the subunit composition, the iron load and the iron to nitrogen, phosphorus to iron and phosphorus to nitrogen molar ratios in a sample of purified rabbit liver fcrritin (S) and in six fractions obtained after separation of that sample into fractions, highly enriched to ferritins with a distinct subunit composition (FI -F6) Fraction
% 1."
S F1 F2 F3 F4 F5 F6
46.11 81 .l/ 67.3 52.6 41.2 35. I 32.4
Chemical analyses
Fe/t:T~'
I:c/N
St)
P/Fe
SI)
PiN
St)
11.55 I). 13 11.13 il.57 1t.54 0.34 0.41
IL01 0.04 1t.I11 11.112 0.01 0.02 0.115
0.090 I). 1311 II. 117 t).tl93 tl.0811 0.070 11.1)61)
1L007 11.1130 0.011 0.010 0.007 II.l) 12 0.036
I).11511 I).1)21/ 11,1114 [).1157 0,043 0.II27 0.029
I).1)t14 /I.006 1L003 11.[113 11.004 0.0116 0.012
29~2
67(I 67"; 3020 2892 183 I 2214
The values of all chemical analyses were obtained as m e a n wdues of at least three separate analysis runs, p e r f o r m e d in duplicate. The values for the relative subunit composition of the samples were obtained by gelscanning of polyacrylamide gradient gels, as described before (Frenkel et a l . , 1983b). " The relative subunit composition is expressed as the percentage L-subunits present in the sample. The other part is composed of tt-lvpc subnnits. ~' Expressed is the n u m b e r of iron atoms per fcrritin molecule (Ft)
Ferritin iron and phosphorus above mentioned studies, which included heat treatment, has led to erroneous results. We have shown before that the classical heat treatment influences some biochemical and physiological properties of ferritin (Frenkel etal., 1983a, 1983c); in particular a loss of H-subunit rich ferritin was noticed. In the present study a gentle purification method was used with omission of heat treatment. Before discussing the results of our calculation of the P/Fe-ratios some points have to be made. Because many authors use various definitions for the ratio of, for instance, phosphorus to iron in ferritin preparations, we feel the need for a clear definition of the ratio used in this report, to be able to prevent ambiguity. Because of its relationship with the molecular formula we use the molar ratio (not the weight ratio). We divide the amount of phosphorus present by the amount of iron to obtain numbers, lower than the unity. On speaking about an increasing ratio, we mean that the number, obtained by performing this division increases (not that the difference in absolute value of numerator and d e n o m i n a t o r increases). In the literature the iron-containing core of ferritin often is described as a polynuclear ferric oxyhydroxide-phosphate complex, with the chemical formula: ( F e O O H ) s ( F e O : O P O 3 H z ) , exhibiting a constant phosphorus to iron ratio of 0.11 (Mazur et al., 1950). As mentioned in our introduction this ratio was obtained from horse spleen ferritin, and the value of 0.11 was confirmed by several authors. The P/Fe-ratio, however, appeared to be by no means a constant value when determined in ferritin from other origin (Myagkaya and de Bruijn, 1982a,b). In our starting material we found a P/Fe-ratio of 0.090 + 0.007 (m + SD). The fractions from this ferritin preparation, however, exhibited P/Fe-ratios in the range from 0.060 to 0.130. We found a significant correlation between the percentage of L-subunits (x) and the P/Fe-ratio (y) in the isolated fractions with the following equation for linear regression: y = 0.00135x + 0.02309 (r = 0.71, P < 0.001). Assuming complete linearity of this curve the P/Fe-ratio would be 0.023 in ferritin containing exclusively H-subunits, and 0.158 in ferritin with 100% L-subunits. In our experiments conditions for the different fractions were identical, apart from the p H of the polybuffer during elution of fractions from the chromatofocussing column. Therefore it is evident that no differences have existed in contamination with inorganic phosphate between the distinct ferritin forms from the m o m e n t of their synthesis in the rabbit liver, during incorporation of iron, until the final samples were obtained. Because the ferritin iron core contains a variable amount of iron-microcrystallites, which grow from nucleation points on the inner surface of the apoferritin molecule, the total surface of these microcrystallites will vary considerably, and consequently the amount of inorganic phosphate, able to bind superficially to the iron complexes. In L-subunit rich ferritin the P/Fe-ratio is higher. This is compatible with a larger total surface of the microcrystallites inside the ferritin protein shell, implicating a greater number of microcrystals stored.
423 SUMMARY
Rabbit liver ferritin was isolated by immunoaffinity chromatography. After separation of the preparation into fractions with a high enrichment with distinct subtypes of ferritin, we found a gradual increase in the phosphorus to iron ratio (P/Fe-ratio) that paralleled the shift from high molecular weight (H) to low molecular weight (L) subunits in the isolated ferritin forms. The iron content of these fractions, ranging from 670 to 3020 iron ions per ferritin molecule, was not related to the subunit composition. Our results indicate that phosphorus is not a structural part of the micellar iron core of ferritin. We postulate that most of the inorganic phosphate (Pi) is reversibly attached to the surfaces of the iron microcrystallites inside the ferritin molecule. The increase of the P/Fe-ratio in relation to the relative proportion of L-subunits may be a result of an increase of the total surface area of the multiple iron microcrystallites. The extent of this surface depends on the number of intramolecular microcrystallites. L-subunit rich ferritin probably contains, therefore, a greater number of nucleation points for formation of iron complexes than H-subunit rich ferritin, which may lead to a larger number of iron microcrystallites to which phosphorus can bind.
Acknowledgements - - We wish to thank the National Institute of Public Health, RIV., for providing us with fresh rabbit livers, and we are grateful to Drs M. 1. Cleton and W. C. de Bruijn for their valuable advice. The foundation for Biological Research in The Netherlands, BION. which is supported by The Netherlands Organization for the Advancement of Pure Research, ZWO, is acknowledged for financial support (Grant No. 14-96--4)09).
REFERENCES
Arosio P., Yokata M. and Drysdale J. W. (1977) Characterization of serum ferritin in iron overload: possible identity of natural apoferritin. Br. J. Haematol. 36,199-207. Bomford A., Berger M., Lis Y. and Williams R. (1978) The iron content of human liver and spleen isoferritins correlates with their isoelectric point and subunit composition. Biochem. biophys. Res. Commun. 83,334-341. Bomford A., Conlon-Hollingshead C. and Munro H. N. (1981) Adaptive responses of rat tissue isoferritins to iron administration. Changes in subunit synthesis, isoferritin abundance and capacity for iron storage. J. biol. Chem. 256,748--955. B6ttcher C. J. F., van Gent C. M. and Pries C. (1961) A rapid and sensitive sub-micro phosphorus determination. Anal. chim. Acta 24, 203-204. Clegg G. A., Fitton J. E., Harrison P. M. and Treffry A. (1980) Ferritin: molecular structure and iron storage mechanisms. Progr. Biophys. molec. Biol. 36, 53-86. Drysdale J. W. (1977) Ferritin phenotypes: structure and metabolism. In Iron Metabolism, Ciba Foundation Symposium Series 51, pp. 41-47. New Series Elsevier/ Exerpta Medica/North Holland, Amsterdam. Frenkel E. J., van den Beld B., van Oost B. A. and Marx J. J. M. (1983a) Influence of heat treatment on rabbit liver ferritin. Biochim. biophys. Acta 745, 202-208. Frenkel E. J., van den Beld B., K6nig B. W. and Marx J. J.
424
ERIK J. FRENKELet al.
M. (1983b) Preparative isolation of distinct molecular forms of rabbit liver ferritin using high performance liquid chromatography. Analyt. Biochem. 135,489-494. Frenkel E. J., van den Beld B. and Marx J. J. M. (1983c) Influence of heat-treatment on rabbit liver-ferritin kinetics. Br. J. Haematol. 55,449-453. Granick S. and Hahn P. F. (1944) Ferritin: VIII. Speed of uptake of iron by the liver and its conversion to ferritin iron. J. biol. Chem. 155,661~69. Harrison P. M. and Hoy T. G. (1973) Ferritin. In Inorganic Biochemistry (Edited by Eichhorn G. G.) pp. 253-279. Elsevier, New York. Harrison P. M., Hoy T. G., Macara I. G. and Hoare R. J. (1974) Ferritin iron uptake and release. Structurefunction relationships. Biochem. J. 143,445-451. Van Kreel B. K., Pijnenburg A. M. C. M., Van Eiik H. G. and Leijnse B, (1972) The incorporation of ~gFe, 3H . leucine and 3-~P morgamc phosphate in ferritin during the perfusion of isolated rat livers. Biochim. biophys. Acta 273,243-247. Mazur A., Litt 1. and Shorr E. (1950) Chemical properties of ferritin and their relation to its vasodepressor activity.
J. biol. Chem. 187,473-484. Myagkaya G. L, and de Bruijn W. C. (1982a) X-ray microanalysis of cellular localization of ferritin in mammalian spleen and liver. Micron 13, 7-21. Myagkaya G. L. and De Bruijn W. C. (1982b) Ferritin ill chorionic villi of the sheep placcnta. Bihl. Anat. 22, 117-222. Schilt A. A. (1969) In Analytical Applieations of l, lO-Phenanthroline and Related compounds. International Series of Monographs in Ana@tical Chemistry 32 (Edited by Belcher R. and Freiser M.) pp. 59-62. Pergamon Press, Oxford. Spiro T. G. and Sattman P. (1969) Polynuclcar complexes of iron and their biological implications. Struct. Bond. 6, 116-156. Treffry A. and Harrison P. M. (1978) Incorporation and release of inorganic phosphate in horse spleen ferritin. Biochem. J. 171,313-320. Urushizaki I., Niitsu Y., Ishitani K., Matsuda M. and Fukuda M. (1971) Microheterogencity of horse spleen ferritin and apoferritin. Bioehim. biophys. Aeta 243, 187-192.