British Jorrrnal
of I-laernatology,
1975,30, 517.
Correspondence H U M A N ISOFERRITINS As I see it, Powcll et a1 (1973) take as prcniise what has yct to bc provcd, nanicly, that tlic bands produced by clectrofocusing of a protcin-in this case, ferritin-in polyacrylamide gcls are necessarily separate, covalently different, subspecies of that protcin. Only if this prcmisc be true, may tlic authors’ conclusioi~that they have found i~umcrousisoferritins in several liuiiiaii tissucs be warrantcd. One has to assume, further, that tlic samples of ferritin were sufficiently purc. Allow me to consider these points briefly. For the sake of brcvity, I shall refrain from citing many papers individually, and refer readers to recent reviews by Criclitoii (1973) and by Harrison ct a1 (1974). Tlic electrofocusiiig procedure employed by Powell et ul (1975) involved the use of 4% polyacrylaniide gel with sieving properties, of ampliolytes (polyaminopolycarboxylic acids) in order to produce the neccssary pH gradients, and of buffer ions. It is knowii that ferritin and apoferritiii exist not oiily as monomer inolccules, but as dimcrs, triiners and higher oligomers as well, being demonstrable as such by non-focusing electrophoresis in polyacrylamide or starch gels. It is iieccssary, therefore, to show that, in each sample aiialyscd, sieving effects play no role in the banding differences observcd on electrofocusiiig. Morc important, it is also known that interactions of proteins with other molecules, such as the anipholytes used, can result in various kinds of association products or adducts. Equilibria in multicoiiiponent systems such as those used in electrofocusing are complex (cf. Cam, 1970). The possibility that chemical changes, such as oxidation of amino groups or of metliioiiinc in some of tlie ferritin, took place during preparation must also be considered. hi any case, thc categorical naturc of any putative isoproteiii should be established before claims of structural peculiarities arc made for it. The results of the ‘inimuiiofixation’ procedure employed by Powell ct ul(1975) raise other questions. Firstly, bauds representing, for example, association products of ferritin aiid other constituents of the system would ‘fix’ antibodies to ferritin if tlie relevant antigenic group is available for reaction. Secondly, the antigenic purity of the imniunogcu used to raise tlic antibodies in rabbits (or goats) should have been established. The authors found that CNBr cleavage of ferritiii extracted, respectively, from human spleen aiid kidney produced peptides which, upon electrofocussing, gave banding patterns that differed quantitatively, but they were vague about the existeiice of qualitative differences in tlic pcptides (page 52 and Fig S).Yet, one would expect to see capital, unmistakable, qualitative differelices in the cleavage products if primary structures differed to the extent indicated in the authors’ interpretation of the electrofocusing patterns of‘ spleen and kidney ferritin in Fig 4. Although tlie existeiice of isoferritins in several situations seems quite probable on the basis of other work (see references in the reviews cited), it is of more than academic interest that putative isoferritins in man be defined with special care. Else we may witness a premature 517
$18
Correspondence
rush into quasi-diagnostic tests for ‘tissue-specific’, ‘cancer-associated’ or ‘carcinofetal’ ferritin s. Departt~icntof Pathology T h e Utiiversity of Rochestcv Roclzrstcr, N. Y. 14642, U.S.A.
G. W. RICHTER r7 April 1975 REFERENCES
CANN, J.R. (1970) Iiiferactbg Macrorirolccir?es. Academic Press, New York. CRICHTON, R.R. (1973) Ferritin. Striictrrre and Borzdirzg (Ed. by J. D. Duiiitz e t a / ) , Vol. 17, p 67. Springer, Berlin. P.M., HOARE,R.J., HOY,T.G. & MACARA, HARRISON, I.G. (1974) Ferritin and haemosiderin: structure
and function. Irori irz Biochtristry arid hledicirzc (Ed. by A. Jacobs and M. Worwood), p 73. Academic Press, London. POWELL,L.W., ALPERT,E., ISSELBACHER, F.J. & DRYSDALE, J.W. (1975) Human isoferritins: organ specific iron and apoferritin distribution. British Jorirnnl of Haenmtology, 30, 47.
Dr Drysdale and rolkagircs reply asfo/lows: Dr Richter offers two main objections to our payer, claiming that the inultiple isoferritins seen by gel electrofocusing arc either iiicthodological artifacts, or insignificant chemical modifications. W e disagree with both contentions. The fact that many tissue ferritiiis may bc resolvcd into multiple components is geiierally accepted by niost investigators who have iiiastered the art of isoelectric focusing (Righetti & Drysdalc, 1974). All of Dr Richter’s arguments suggesting that the multiple forms arc coiitamiiiaiits or artifacts resulting froin aggregates or ampholytc binding have already been fully answered in the past few years (Drysdale, 1970, 1974; Urushizaki et al, 1971, 1973). Regrettably, hc has failed to coiisidcr this iriiportaiit body of evidence. W e have shown that thc isoferritin profiles obtained in our system do, in fact, represent equilibrium banding patterns (Drysdale, 1974). Furthermore, similar patterns are also obtained by isoelectric focusing in sucrosc density gradients or in Scphadex bcds (Alpert et al, 1973 ; Righetti 8~ Drysdale, 1974; Drysdale et al, 1974). These isoferritins are discrete molecular entities that do not redistribute on refocusing (Drysdale et nl, 1974; Marcus & Zimberg, 1974). They do iiot represent differences in iron content or aggregate size since monomeric, i.e. single shell, apoferritiii gives cssentially the same pattern (Drysdalc, 1970; Urushizaki et al, 1971). Finally, since a similar lieterogencity may be observed by ioii cxchaiige chromatography (Urusliizaki et al, 1973 ; Drysdalc, 1974; Marcus & Ziinberg, 1974), arguments about ampholyte binding, or any other possible artifact of electrofocusiiig, become irrelevant. The demonstration of multiple forms of ferritiii led us to question whether thcy result from specific modifications of a single, virginal protein in different tissues,or from the synthesis of different ferritin gene products. Most current evidence favours the latter hypothesis. It is generally accepted that different tissues contain ferritins of different primary structure. Soine investigators bclievc each tissue ferritin to be different molecular species. In contrast, we believe that most tissue ferritiiis are composed of related families of isoferritins, some of which are common to several tissues. Subunit analyses indicate that the individual isoferritins also share common structural determinants. W c have found that most isoferritins are hybrid
of I-laernatology,
1975,30, 517.
Correspondence H U M A N ISOFERRITINS As I see it, Powcll et a1 (1973) take as prcniise what has yct to bc provcd, nanicly, that tlic bands produced by clectrofocusing of a protcin-in this case, ferritin-in polyacrylamide gcls are necessarily separate, covalently different, subspecies of that protcin. Only if this prcmisc be true, may tlic authors’ conclusioi~that they have found i~umcrousisoferritins in several liuiiiaii tissucs be warrantcd. One has to assume, further, that tlic samples of ferritin were sufficiently purc. Allow me to consider these points briefly. For the sake of brcvity, I shall refrain from citing many papers individually, and refer readers to recent reviews by Criclitoii (1973) and by Harrison ct a1 (1974). Tlic electrofocusiiig procedure employed by Powell et ul (1975) involved the use of 4% polyacrylaniide gel with sieving properties, of ampliolytes (polyaminopolycarboxylic acids) in order to produce the neccssary pH gradients, and of buffer ions. It is knowii that ferritin and apoferritiii exist not oiily as monomer inolccules, but as dimcrs, triiners and higher oligomers as well, being demonstrable as such by non-focusing electrophoresis in polyacrylamide or starch gels. It is iieccssary, therefore, to show that, in each sample aiialyscd, sieving effects play no role in the banding differences observcd on electrofocusiiig. Morc important, it is also known that interactions of proteins with other molecules, such as the anipholytes used, can result in various kinds of association products or adducts. Equilibria in multicoiiiponent systems such as those used in electrofocusing are complex (cf. Cam, 1970). The possibility that chemical changes, such as oxidation of amino groups or of metliioiiinc in some of tlie ferritin, took place during preparation must also be considered. hi any case, thc categorical naturc of any putative isoproteiii should be established before claims of structural peculiarities arc made for it. The results of the ‘inimuiiofixation’ procedure employed by Powell ct ul(1975) raise other questions. Firstly, bauds representing, for example, association products of ferritin aiid other constituents of the system would ‘fix’ antibodies to ferritin if tlie relevant antigenic group is available for reaction. Secondly, the antigenic purity of the imniunogcu used to raise tlic antibodies in rabbits (or goats) should have been established. The authors found that CNBr cleavage of ferritiii extracted, respectively, from human spleen aiid kidney produced peptides which, upon electrofocussing, gave banding patterns that differed quantitatively, but they were vague about the existeiice of qualitative differences in tlic pcptides (page 52 and Fig S).Yet, one would expect to see capital, unmistakable, qualitative differelices in the cleavage products if primary structures differed to the extent indicated in the authors’ interpretation of the electrofocusing patterns of‘ spleen and kidney ferritin in Fig 4. Although tlie existeiice of isoferritins in several situations seems quite probable on the basis of other work (see references in the reviews cited), it is of more than academic interest that putative isoferritins in man be defined with special care. Else we may witness a premature 517
$18
Correspondence
rush into quasi-diagnostic tests for ‘tissue-specific’, ‘cancer-associated’ or ‘carcinofetal’ ferritin s. Departt~icntof Pathology T h e Utiiversity of Rochestcv Roclzrstcr, N. Y. 14642, U.S.A.
G. W. RICHTER r7 April 1975 REFERENCES
CANN, J.R. (1970) Iiiferactbg Macrorirolccir?es. Academic Press, New York. CRICHTON, R.R. (1973) Ferritin. Striictrrre and Borzdirzg (Ed. by J. D. Duiiitz e t a / ) , Vol. 17, p 67. Springer, Berlin. P.M., HOARE,R.J., HOY,T.G. & MACARA, HARRISON, I.G. (1974) Ferritin and haemosiderin: structure
and function. Irori irz Biochtristry arid hledicirzc (Ed. by A. Jacobs and M. Worwood), p 73. Academic Press, London. POWELL,L.W., ALPERT,E., ISSELBACHER, F.J. & DRYSDALE, J.W. (1975) Human isoferritins: organ specific iron and apoferritin distribution. British Jorirnnl of Haenmtology, 30, 47.
Dr Drysdale and rolkagircs reply asfo/lows: Dr Richter offers two main objections to our payer, claiming that the inultiple isoferritins seen by gel electrofocusing arc either iiicthodological artifacts, or insignificant chemical modifications. W e disagree with both contentions. The fact that many tissue ferritiiis may bc resolvcd into multiple components is geiierally accepted by niost investigators who have iiiastered the art of isoelectric focusing (Righetti & Drysdalc, 1974). All of Dr Richter’s arguments suggesting that the multiple forms arc coiitamiiiaiits or artifacts resulting froin aggregates or ampholytc binding have already been fully answered in the past few years (Drysdale, 1970, 1974; Urushizaki et al, 1971, 1973). Regrettably, hc has failed to coiisidcr this iriiportaiit body of evidence. W e have shown that thc isoferritin profiles obtained in our system do, in fact, represent equilibrium banding patterns (Drysdale, 1974). Furthermore, similar patterns are also obtained by isoelectric focusing in sucrosc density gradients or in Scphadex bcds (Alpert et al, 1973 ; Righetti 8~ Drysdale, 1974; Drysdale et al, 1974). These isoferritins are discrete molecular entities that do not redistribute on refocusing (Drysdale et nl, 1974; Marcus & Zimberg, 1974). They do iiot represent differences in iron content or aggregate size since monomeric, i.e. single shell, apoferritiii gives cssentially the same pattern (Drysdalc, 1970; Urushizaki et al, 1971). Finally, since a similar lieterogencity may be observed by ioii cxchaiige chromatography (Urusliizaki et al, 1973 ; Drysdalc, 1974; Marcus & Ziinberg, 1974), arguments about ampholyte binding, or any other possible artifact of electrofocusiiig, become irrelevant. The demonstration of multiple forms of ferritiii led us to question whether thcy result from specific modifications of a single, virginal protein in different tissues,or from the synthesis of different ferritin gene products. Most current evidence favours the latter hypothesis. It is generally accepted that different tissues contain ferritins of different primary structure. Soine investigators bclievc each tissue ferritin to be different molecular species. In contrast, we believe that most tissue ferritiiis are composed of related families of isoferritins, some of which are common to several tissues. Subunit analyses indicate that the individual isoferritins also share common structural determinants. W c have found that most isoferritins are hybrid
