Scandinavian Journal of Clinical and Laboratory Investigation
ISSN: 0036-5513 (Print) 1502-7686 (Online) Journal homepage: http://www.tandfonline.com/loi/iclb20
Clinical Chemistry: Characterization of R-Type Vitamin B12-Binding Proteins by Isoelectric Focusing U.-H. Stenman To cite this article: U.-H. Stenman (1975) Clinical Chemistry: Characterization of R-Type Vitamin B12-Binding Proteins by Isoelectric Focusing, Scandinavian Journal of Clinical and Laboratory Investigation, 35:2, 147-155 To link to this article: http://dx.doi.org/10.1080/00365517509087218
Published online: 28 Aug 2009.
Submit your article to this journal
Article views: 2
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iclb20 Download by: [University of California Santa Barbara]
Date: 06 November 2015, At: 03:06
Characterization of R-Type Vitamin B,,-Binding Proteins by Isoelectric Focusing 11. Comparison of Cobalophilin (R Proteins) from Different Sources
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
U.-H. STENMAN The Minerva Foundation Institute for Medical Research, Helsinki, Finland
Stenman, U.-H. Characterization of R-Type Vitamin B,,-Binding Proteins by Isoelectric Focusing. 11. Comparison of Cobalopilin (R Proteins) from Different Sources. Scand. J. clin. Lab. Invest. 35, 147-155, 1975. The R-type vitamin BIZ-binding proteins, here called cobalophilin, in human plasma, serum, leukocytes, gastric juice, amniotic fluid, saliva, and milk, were characterized by isoelectric focusing and gel filtration. Cobalophilin from all these sources is microheterogeneous, consisting of several isoproteins with isoelectric points (PI) between 2.3 and 5.0. Isuproteins with PI values of 3.0, 3.3, 3.6, 3.9, and 4.2 were found in most of the cells and fluids studied, but in different proportions. Milk contains isoproteins with PI values mainly between 4.0 and 4.7, and such components also occur in considerable amounts in saliva. The cobalophilin in these fluids also has a smaller Stokes radius than that from other sources. On the basis of the isoelectric patterns and the Stokes radii the isoproteins of cobalophilin are tentatively divided into two populations, a glandular one occurring in milk and saliva and a myelogenic one occurring in all the cells and fluids studied but only occasionally in milk. Key-words: Cobalophilin; gel filtration; isoelectric focusing; Stokes radius; transcobalamin; vitamin B12 U.-H. Stenman. M.D., T h e Minerva Foundation Institute for Medical Research, P.O.Bon 819, SF-00101 Helsinki 10, Finland
The term binder R was originally introduced to denote the gastric non-intrinsic factor vitamin B,,-binding protein with rapid electrophoretic mobility (13,34). Proteins with similar immunological properties have been demonstrated in most body fluids and blood cells (33), and the names R-type vitamin Blz-binding protein and R protein have been used ,for this protein, irrespective of its origin (12, 33). Jn electrophoresis R protein from different sources has electrophoretic mobilities varying from alpha to beta (12, 14, 33, 41). Thus the denomination R (for rapid mobility) is somewhat misleading, and the name cobalophilin is now introduced to replace the name R protein. Cobalophilin has been isolated from saliva (17), granulocytes (2), and amniotic fluid (36). It is a glycoprotein with an Mr value of
about 60,000, but in gel filtration cobalophilin from most sources behaves like a protein with an Mr value of about 120,OOO j12, 24, 35, 40, 41), obviously because of its high carbohydrate content, 33 per cent (2, 36). Salivary cobalophilin has a smaller apparent molecular size in gel filtration, 60,OOO-88,OOO f16, 23) and a lower carbohydrate content (17). Several names have been used for cobalophilin in plasma and serum: transcobalamin (TC) I (22), ‘high’ molecular weight B,, binder (28), TC L (19), TC I-!- TC I11 (3), and TC I third vitamin B,,-binding protein (4) binder 2 binder 3 (25). That two R-type vitamin Blz-binding proteins are present in serum ((3, 4, 32) is dubious, however. By isoelectric focusing it has been shown that the separation of cobalophilin in serum into an
+
+
148
U.-H. Stenman
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
a- (TC
I) and a ,&globulin fraction (31) (TC 111) is artificial. TC I and TC 111 merely represent a partial separation of the isoproteins of the microheterogeneous cobalophilin into two overlapping fractions (37). Cobalophilin in amniotic fluid 1(39), saliva (17), and granulocytes l(37) is also microheterogeneous, containing at least five components isoproteins - most of which are common to all these materials. Isoelectric focusing has now been used to ascertain the reasons for the differences in electrophoretic mobility and molecular size of cobalophilin from different sources.
MATERIAL AND METHODS Blood samples were obtained from healthy laboratory personnel. Cord blood was taken from the placental part of the umbilical cord immediately after delivery. Serum was separated 2.5 h after sampling. Plasma was prepared by addition of 2 mg EDTA (K,) per millilitre blood and centrifugation within 1 h. Amniotic fluid was taken at term by amniotomy. Samples contaminated by blood were discarded. Milk was collected at different times after delivery. The milk fat was removed by centrifugation. Leukocytes were prepared from BDTA blood by lysis of the erythrocytes with 0.155 mol/l NH,OI (35). Gastric juice was collected after histalog stimulation and depepsinized by raising the pH to 10 (33). Saliva was obtained from healthy laboratory personnel immediately before use. The samples were frozen in small aliquots as soon as possible after sampling (unless used immediately) and stored at -20 "C. Thawed samples were not refrozen. Isoelectric focusing was performed in a 110ml column LKB 8101 (LKB, Bromma, Sweden) essentially as described '(39) but with the following modifications. The dense solution has the following composition: 0.5 ml AmpholineB (LKB), pH 2 . 5 4 (20 per cent w/v), 27.0 ml glycerol ( 1 0 0 per cent v/v), 25.0 ml distilled water, and 1.5 ml of a stock solution of acids. This solution contains 45 mmol/l malic acid, 185
mmol/J formic acid, 85 mmol/l succinic acid, 250 mmoYl acetic acid, and 200 mmoYl propionic acid. The less dense solution contains 1 ml Ampholine, pH 3.5-5 (40 per cent w/v), 0.5 ml Ampholine, pH 4-6 (40 per cent w/v), and 52.5 ml water. The density gradient was formed with gradient mixer. The addition of Ampholine, pH 2.5-4.0, makes a smoother pH gradient between pH 3 and 4, but the results are comparable with those earlier obtained. The sample was mixed with glycerol and introduced directly into the column when half the gradient had been formed. This modification of the original method (39) was introduced because the proteins tended partly to remain in the, upper part of the column when the sample was mixed with the dense or the less dense solution. A maximum of 0.2 ml serum or plasma was applied, but better separation was obtained when 0.1 ml or less was used. The focusing time was 5 days; temperature, 10 "C; and potential, 400 V. The column was emptied with a pump, at a rate of 60 ml/h. One-millilitre fractions were collected and radioactivity and pH measured (39). Cobalophilin treated with sialidase and that in milk were also focused in a pure ampholyte gradient covering the pH range 3.5-6; focusing conditions: 1 per cent Ampholine, pH 3-6; sucrose density gradient, Ck50 per cent; 450 V; temperature, 10 "C; focusing time: 2 days. Attempts were also made to use an ampholyte gradient without acid solution for the pH range 2 . 5 4 (42). Good separation of the protein peaks was only occasionally obtained with this method, and unsuccessful focusing was always accompanied by blurring or tilting of the visible zone of methyl red (routinely added), which normally forms a horizontal band 0.5 cm wide at pH 3.75 ((39). Gel filtration was performed on Sephadex G-150 (Pharmacia, Uppsala, Sweden), column size 1.6 X 90 cm, with 0.1 moYl sodium phosphate buffer, pH 7.4, containing 0.02 per cent sodium azide. The fraction volume was 3.5 ml and the temperature + 4 "C. The void volume of the column was determined with Blue Dextran 2000 fPharmacia). The molecular size of cobalophilin from different sources was com-
+
+
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
Vitamin BIZ-BindingProteins
pared with that in amniotic fluid by the double isotope technique (36), using vitamin B,, labelled with 57Co and 5 8 C o i(The Radiochemical Centre, Amersham, U.K.). The Stokes radius was calcukted (1) on the basis of the known value, 4.61 nm (36), for cobalophilin in amniotic fluid. The vitamin BIZ-binding capacity was determined with haemoglobin-coated charcoal (10) in all samples and additionally with gel filtration l(24) in samples containing more than one binding protein - plasma, serum, and gastric juice. One nanogram radioactive vitamin B,, was incubated for 15 min with a volume of plasma or serum (usually 0.1 ml) giving a five- to ten-fold excess of vitamin BIZ; the total incubation volume was 1 ml. The other fluids studied were incubated with 5 ng vitamin BIZ in the same volume. From the association constants of cobalophilin (23) it can be calculated that under these conditions more than 98 per cent of the binding capacity will be saturated. Separation of bound and free vitamin B,, with haemoglobin-coated charcoal (10) or gel filtration (24) gave the same results. Gel filtration separates the binding proteins in plasma, serum (24), and gastric juice (12), making calculation of the binding capacity of individual binding proteins possible. Identification of vitamin BIZ-binding proteins as cobalophilin was based on their ability to react with antisera against cobalophilin in saliva (17) and amniotic fluid ((36). These antisera did not react with other vitamin BIZbinding proteins. Samples saturated with radioactive vitamin B,, were incubated with antiserum and subjected to gel filtration. Reaction with anti-cobalophilin serum resulted in a changed elution volume (8). Sialic acid was removed from cobalophilin with sialidase (Behringwerke, Mahrburg-Lahn, Germany, EC 3.2.1.18). To a 0.1-ml sample saturated with vitamin B,, was added 50 units sialidase (0.1 ml) and 0.3 ml 0.05 mol/l sodium acetate buffer @H 5.9) containing 0.154 mol/l NaCl and 9 mmol/l CaCl,; this mixture was incubated for 24 h at 3 7 "C during continuous dialysis against the acetate buffer. Controls were incubated without sialidase.
149
Table I. The unsaturated vitamin B1,-binding capacity of the cobalophilin in the different body fluids studied Range
No. of samples
-_
Adult plasma (EDTA) Adult serum Foetal plasma (EDTA)
Foetal serum Amniotic fluid Milk Saliva Gastric juice
0.133 0.360
0.055-0.330 5 0.090 -0.910 5
0.097 0.40 15.3 41.0 34.4 28.3
0.012 - 0.270 6 0.067 - 1.180 6 1.0-32 9.0-85 20.7 -45 6.4-65
57 6 6 8
RESULTS Vitamin B,,-binding capacity o f cobalophilin
Data for the unsaturated vitamin B,,-binding capacity of cobalophilin in different body fluids are summarized in Table I. The samples of plasma and serum were taken simultaneously from the same patients, and the values are therefore directly comparable. In both adult and foetal serum the free binding capacity of cobalophilin was three to four times as high as in EDTA plasma. Since the binding capacity of amniotic fluid had not been determined previously, a larger number of samples were studied. The values for milk were in the same range as reported earlier, 80 pg/l f l 8 ) and 20-157 pg/l (20). Zsoelectric focusing
Fig. 1 shows schematically the percentage distribution of radioactive vitamin B,, between the isoproteins of cobalophilin from different sources. Cobalophilin was always microheterogeneous, consisting of five or more isoproteins. One sample with a low (A), one with an average (B) and one with a high (C) proportion of acid components are presented. Amniotic fluid, adult and foetal plasma, serum, and leukocytes all contained the same components with isoelectric points (PI) at 3.0, 3.3, 3.6, 3.9, and 4.2. In addition, isoproteins with PI values of 2.3 and 2.7 were occasionally seen in amniotic fluid. The proportion of the
U.-H. Stenman
I50
Amnlotlc lluld
Plasma
Mllk
50:ed"lt'
A
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
B
-
Serum
0
A
l
,
tl, ~
Fig. 1. Diagram of the percentage distribution of radioactive vitamin BIZ among the isoproteins of cobalophilin separated by isoelectric focusing. Each bar represents one isoprotein, and its place on the abscissa is determined by its isoelectric point. The isoelectric patterns of three samples of the cobalophilin from each source are shown above each other. One sample each of extreme cases: more acid pattern (C), less acid pattern (A), and one sample representing the average pattern (B). Ordinate: percentage of radioactive vitamin BIZ contained in each isoprotein; abscissa: PI values of the components.
!,,wr C
0 2
3
4
5
2
3
4
5
2
pl values of the
3
4
5
2
3
4
5
components
components varied with the source. Plasma had a more acid pattern than serum, the pattern of which was slightly more acid than that of leukocytes. The difference in pattern between plasma and serum is explained by release of cobalophilin from granulocytes during clotting (37). Cobalophilin in adult plasma had a more acid pattern than that in foetal plasma. A similar but smaller difference was observed between adult and foetal leukocytes. The cobalophilin in amniotic fluid had an isoprotein pattern very similar to that in leukocytes. The differences in pattern between individual samples were mostly small and were greatest among the serum samples, probably owing to the variety in the admixture of cobalophilin from granulocytes. The same isoproteins of cobalophilin as were found in the above samples were also seen in gastric juice and saliva. l n addition, these body fluids contained various amounts of isoproteins isoelectric between 4.2 and 5.0.. Be-
cause the PI values of these differed from sample to sample, typical isoproteins could not be defined. Efforts were made to study gastric juice and saliva immediately after sampling (although this was not possible with all samples of gastric juice). In spite of this, the isoprotein patterns of saliva and gastric juice varied greatly. One explanation for this is that the cobalophilin in saliva and gastric juice IS probably derived from at least two sources (see Discussion). Another possible reason is degradation, probably desialylation, which obviously starts in vivo immediately after secretion, especially at the low pH in gastric juice. Milk conltains isoproteins with PI values between 4 and 5 (Fig. 2), as do saliva and gastric juice. But the isoproteins of milk, unlike those of these latter fluids, typically have PI values #of 4.0, 4.1, 4.25-4.30, 4.32-4.39, 4.5, and, occasionally, 4.7. Only one of the eight milk samples studied (Fig. 1C) contained mea-
Vitamin B , -Binding Proreins
-
-
S-
7 -
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
I a
0
20
40
60
80
100
F r a c t l o n no.
Fig. 2. Isoelectric focusing of the cobalophilin in milk (thick line) and the sialidase-treated cobalophilin in saliva (dotted line). pH is indicated by a thin line. Results of two experiments superimposed on each other.
surable amounts of 'acid isoproteins' with PI values of 3.0, 3.3, and 3.6. These accounted for only 8 per cent of all the bound vitamin BIZ. After treatment with sialidase, cobalophilin from different sources had fairly uniform isoelectric patterns. In shallow gradients covering the pH range 4-6, three to fmive peaks were found at pH values of 4.8, 4.9, 5.0, 5.1, and 5.2 'pig. 2). In most samples, the main isoprotein was the one isoelectric at pH 5.0, but the proportions of the other components varied. Controls incubated without sialidase also underwent changes in isoelectric pattern that indicated loss of sialic acid. These changes were small, however, except for in the samples of saliva, gastric juice, and leukocyte extract (cf. Ref. 41). After incubation, these contained no isoproteins isoelectric below p H 4.3. The change in isoprotein pattern caused by incubation without sialidase indicates that sialidase activity was present in some of the materials studied. This activity obviously starts to affect the isoprotein pattern of cobalophilin immediately after it has been secreted; its action continues until the sample is frozen, and it may even have some effect during focusing. No attempts were made to measure the autolysis taking place in vivo, but the effects on cobalophilin of different methods of analysis and handling were studied in saliva, which was most prone to autolysis. Storage of saliva for
151
18 h at f 4 "C changed the proportion of isoproteins isoelectric below/above pH 4 from 60140 to 30/70. This change was largely inhibited by saturation of the sample with vitamin B,2. Addition of subsaturating amounts of vitamin B12 also changed the distribution of the vitamin, which was preferentially bound to the most acid isoproteins. Whether desialylation occurs during focusing was checked in samples of saliva focused in an acid-free gradient covering the pH range 4-6, in which focusing is completed in 1 day. This method gave the same proportions of isoproteins isoelectric below and above pH 4 as the pH gradient normally used. Possible desialylation or deamidation of acid iisoproteins at low pH was checked by refocusing of individual peaks. No change in the PI value was observed. Identification of focused vitamin B12-binding proteins as cobalophilin was no problem with samples containing only one type of binding protein - leukocytes, saliva, amniotic fluid, and milk. The radioactive vitamin bound by any of these materials showed a changed elution pattern in gel filtration after incubation with anti-cobalophilin serum. The non-cobalophilin vitamin B,,-binding proteins in plasma, serum (TC II), and gastric juice (intrinsic factor) were occasionally removed by gel filtration before isoelectric focusing. T h h step was not routinely performed because of the risk of denaturation. TC I1 is isoelectric at around pH 6 (37) and does not disturb the pattern of cobalophilin. The most acid isoprotein of intrinsic factor has a PI value of 4.84 (11) and may thus overlap with the least acid isoproteins of cobalophilin. This was observed only in one sample (Fig. 1, gastric juice sample B). In this case the identity of the binding protein was established by incubation with anti-cobalophilin serum and gel filtration.
Gel filtration The Stokes radius of cobalophilin from all sources except saliva and milk did not differ significantly from that of the reference protein. The Stokes radii of cobalophilin in saliva and milk were smaller, and the values of the
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
152
U.-H. Stenman
former varied greatly. Samples of saliva with a high proportion of acid isoproteins had a fairly large and samples with little acid components a small Stokes radius. Therefore the different isoproteins of cobalophilin in saliva were also studied. There was a trend towards smaller Stokes radius with increasing PI value, and isoproteins with PI values above 4.8 were clearly smaller than the others. The molecular size of cobalophilin from most sources was reduced by 2-4 per cent by treatment with sialidase. Desialylated salivary cobalophilin had a Stokes radius of 3.7-3.8 nm - that is, about 15 per cent smaller than the native protein and the same as the component with PI 5.0 I,Table 11) occurring in some samples of native saliva. DISCUSSION Isoelectric focusing has a high resolving power and permits accurate characterization of separated proteins in terms of their PI values. This made it possible to demonstrate that cobalophilin from different cells and fluids consists mainly of the same isoproteins in different proportions. These isoproteins are not all found in every source material, and it is actually possible to discern two populations of isoproteins with typical pI ranges and occurrence. One population, tentatively called myelogenic,
has a PI range of 2.3-4.2 and occurs in pure form in leukocytes, plasma, serum, and amniotic fluid. The other population has a PI range of 4.0-5.0 and may be called glandular because of its occurrence in milk. Saliva contains a mixture of the two populations, as does possibly gastric juice; glandular isoproteins found in the latter, however, may be the result of autolysis or contamination with saliva. A difference between the myelogenic and glandular isoprotein populations was also demonstrated by gel filtration, the latter having a smaller average molecular size. The mixed population of saliva obviously explains the variety in molecular-size values found in this and earlier studies (5, 16, 17, 23), but denaturation during storage or purification is probably partially responsible for the lowest values reported (16, 17). The Stokes radius of cobalophilin in milk is compatible with an earlier Mr estimation of 102,000 (9). A further reason for dividing the isoproteins of cobalophilin into two populations is that two cell types have been shown to synthesize cobalophilin in vitro - granulocytes (35) with their precursors (30) and the serous cells of the salivary glands (26). The denominations myelogenic and glandular are only tentative, however, and the possibility that other cell types can produce cobalophilin should be kept in mind. Nevertheless, the similarity between
Table 11. Stokes radius of cobalophilin from milk, saliva, and isoproteins of the salivary cobalophilin* Stokes radius (nm)
______
Milk
Saliva Salivary isoprotein, 3.6t Salivary isoprotein, 3.9 Salivary isoprotein, 4.1 Salivary isoprotein, 4.5 Salivary isoprotein, 4.8 Salivary isoprotein, 5.0
Mean
Range
4.31 4.36 4.49 4.40 4.37 4.39 4.06 3.81
4.11 4.39 4.05 4.56
-
No. of expts. 6 5 1 1 1 1 1 1
___
*Determined by gel filtration (1) on Sephadex G-150 by the double isotope technique. The cobalophilin in amniotic fluid was used as reference protein (Stokes radius, 4.61 nm; Ref. 36). Temperature f 4 "C. The Stokes radii of cobalophilin from other sources did not differ significantly from that of the reference protein. ?This value is the isoelectric point of the isoprotein.
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
Vitamin B,,-Binding Proteins
the isoproteins of cobalophilin in granulocytes and those in most of the body fluids studied supports the view that granulocytes are an impontant source of cobalophilin (35). The cobalophilin in plasma has a higher proportion of acid isoproteins than that in granulocytes, possibly because cobalophilin excreted in vivo contains more sialic acid than the intracellular form. Intrinsic factor in gastric juice is also more acid than the intracellular form ((29).The most acid isoproteins may also be derived from granulocyte precursors (30); the proportion of these isoproteins is increased in the plasma of many patients with chronic myelogenous leukaemia (38). The isoprotein pattern of cobalophilin in amniotic fluid is very similar to that in leukocytes of adults, but differences in the carbohydrate moiety (36) may indicate that cobalophilin in amniotic fluid is not derived from granulocytes. Variation in the carbohydrate composition of a glycoprotein derived from different ti ssues is common (36), and differences in sialic acid content have been suggested to be one cause of the variety in electrophoretic mobility of cobalophilin in different body fluids and cells (41). Sialidase has been shown to abolish the differences in electrophoretic mobility (41), and these results were confirmed by isoelectric focusing. However, treatment with sialidase did not transform the isoproteins to a single substance. This was possibly due to incomplete removal of sialic acid or to further degradation. The latter phenomenon obviously explains the considerable reduction in the size of the cobalophilin in saliva during this treatment. The variation in electrophoretic mobility of cobalophilin from various sources m(12, 14, 33, 41) can be correlated with differences in isoprotein pattern. Thus, cobalophilin with a high proportion of acid isoproteins has a higher mobility than cobalophilin with a low proportion of these. This is the case, for instance, with foetal and adult plasma, which have been considered to contain different species of vitamin B,,-binding proteins of R-type - TC I and foetal binder (27). According to the present study, these fluids contain the same isoproteins of I
153
cobalophilin but in different proportions. The nature of the polycythaemia Vera binder is similar (38). It is therefore advisable to describe vitamin B,,-binding proteins of cobalophilin (or R) type in terms of their isoprotein patterns rather than by the average electrophoretic mobility of their isoprotein population or by their behaviour in ion exchange chromatography with stepwise elution. No transport or enzyme function has yet been ascribed to cobalophilin, and two brothers suffering from lack of this protein had no overt disease (6). Thus the function, proposed by Gullberg (19), of aiding the organism to resist infections by depriving vitamin BIZdependent microorganisms of a growth factor seems to be the most logical one. This hypothesis would also explain the abundance of R protein in tears ((13,granulocytes, saliva, milk, and amniotic fluid. The useful classification of vitamin B12binding proteins into three groups - intrinsic factor, TC 11, and R proteins '(12, 41) - has gained acceptance (7, 21). However, the name R protein has now been changed to cobalophilin. This word has been coined by analogy with siderophilin and does not impute any function to the protein other than the ability to bind vitamin BIZ. ACKNOWLEDGEMENTS Supported by the Finnish National ,Research Council for Medical Sciences, the Finska Vetenskaps-Societeten, the Finska Lakaresallskapet, the Sigrid JusClius Foundation, and Nordisk Insulinfond. I thank Dr. Ralph Grasbeck and Miss Liisa Puutula, M. Sc., for valuable criticism of the manuscript and Miss Anneli Hangasoja for excellent technical assistance. REFERENCES 1. Ackers, G . K. Molecular exclusion and re-
stricted diffusion processes in molecular-sieve chromatography. Biochemistry 3, 723, 1964. 2. Allen, R. H. & Majerus, P. W. Isolation of vitamin BIZ-binding proteins using affinity chromatography. 11. Purification and properties of a human granulocyte vitamin BIZ-binding protein. J . biol. Chem. 247, 7702, 1972.
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
154
U.-H. Stenman
3. Bloomfield, F. J. & Scott, J. M. Identification of a new vitamin BIZ binder (transcobalamin 111) in normal human serum. Brit. J , Haemat. 22, 33, 1972. 4. Carmel, R. Vitamin BIZ-binding protein abnormality in subjects without myeloproliferative disease. 11. The presence of a third vitamin BIZ-binding protein in serum. Brit. J . Haemat. 22, 53, 1972. 5 . Carmel, R. The vitamin BIZ-binding proteins of saliva and tears and their relationship to other vitamin BIZ binders. Biochirn. biophys. Acra ( A m s t . ) 263, 747, 1972. 6 . Carmel, R . & Herbert, V. Deficiency of vitamin BIZ-binding alpha globulin in two brothers. Blood 33, 1, 1969. 7 . England, J . M., Clarke, H. G. M., Down, M. C. & Chanarin. I. Studies on the transcobalamins. Brit. 1. Haemat. 25, 737, 1973. 8 . Finkler, A. E., Green, P. D. & Hall, C. A. lmmunological properties of human vitamin BIZ binders. Biochirn. biophys. Acta (Amst.) 200, 151, 1970. 9. Finkler, A. E., Rappazzo, M. E. & Hall, C. A. Cyanocobalamin binding protein in human milk. Fed. Proc. 26, 695, 1967. 10. Gottlieb, C., Lau, K.-S., Wasserman, L. R. & Herbert, V. Rapid charcoal assay for intrinsic factor (IF), gastric juice unsaturated B l ~ binding capacity, antibody to IF, and serum unsaturated BIZ binding capacity. Blood 25, 875, 1965. 7 I . Grasbeck, R. Microheterogeneity of the complex between human intrinsic factor and cyanocobalamin demonstrated by isoelectric focusing. Acta cherii. scand. 22, 1041, 1968. 12. Grasbeck, R . Intrinsic factor and the other vitamin 812 transport proteins. Progr. Hernut. 6, 233, 1969. 13. Grasbeck, R., Simons, K. & Sinkkonen, I. Purification of intrinsic factor and vitamin BIZ binders from human gastric juicc. Ann. M e d . exp. Fenn. 40, Suppl. 6 , 1962. 14. Griisbeck, R., Stenman, U.-H., Puutula, L. & Visuri, K. A procedure for detaching bound vitamin BIZ from its transport proteins. Biochitn. biophys. Acta (Amst.) 158, 292, 1968. 15. Grasbeck, R. & Takki-Luukkainen, I.-T. Vitamin BIZ-binding substance in human tear fluid. Acta ophthal. ( K b h . ) 36, 860, 1958. 16. Grasbeck, R . & Visuri, K . Isolation of the vitamin BIZ binding protein of human saliva. Scand. 1. c h i . Lob. Invesr. 21, Suppl. 101, 13. 1968. 17. Grasbeck, R., Visuri, K. & Stenman, U.-H. The vitamin BIZ-binding protein in human saliva. Isolation, characterization and comparison with the unpurified protein. Biochini. biophys. Acta (Amst.)263, 721, 1972.
18. Gregory, M. E. & Holdsworth, E. S. The OCcurrence of a cyanocobalamin-binding protein in milk and the isolation of a cyanocobalinprotein complex from sow’s milk. Biochem. J. 59, 329, 1955. 19. Gullberg, R. Vitamin Biz-binding proteins in normal human blood plasma and senun. Scnnd. J . Haernat. 9, 639, 1972. 20. Gullberg, R. Possible influence of vitamin BIZbinding protein in milk on the intestinal flora in breast-fed infants. I. BIZ-binding proteins in human and bovine milk. Scond. 1. Gastroenf. 8, 497, 1973. 21. Hall, C. A. Vitamin Blz-binding proteins of man. Ann. intern. Med. 7 5 , 297, 1971. 22. Hall, C. A. & Finkler, A. E. The dynamics of transcobalamin 11. A vitamin BIZ binding substance in plasma. 1. Lab. clin. M e d . 65, 459, 1965. 23. Hippe, E. Studies on the cobalamin-binding protein of human saliva. Scand. 1. d i n . Lab. Irivcst. 29, 59, 1972. 24. Hom, B., Olesen, H. & Lous, P. Fractionation of vitamin BIZ hinders in human serum. 1. Lab. d i n . Med. 68, 958, 1966. 25.Hurlimann, I . & Zuber, C. Vitamin BIZbinders in human body fluids. I. Antigenic and physico-chemical characteristics. Clin.exp. Immunol. 4 , 125, 1969. 26. Hurlimann, I. & Zuber, C. Vitamin BIZ-binders in human body fluids. 11. Synthesis in vitro. Clin. r x p . Immunol. 4. 141, 1969. 27. Kumento, A. Studies on the serum binding of vitamin U12 in the newborn human infant. Arru puedior. scand., Suppl. 194, 1969. 2X.Lawrence. C. The heterogeneity of the high molecular weight 8 1 2 binder in serum. Blood 33, 899, 1969. 29. Marcoullis, G . & Grasbeck. R. Vitamin Bizbinding proteins in human gastric mucosa. General pattern and demonstration of intrinsic factor isoproteins typical of mucosa. Scand. 1. d i n . Lab. invest. 35, 5 , 1975. 30. Rachmilewitz, B., Rachmilewitz, M. & Gross, J. A vitamin 812 binder with transcobalamin I characteristics synthesized and released by human granulocytes in vitro. Brit. J . Haemat. 26. 557, 1974. 31. Retief, F. P., Gottlieb, C. W., Kochwa, S., Pratt, P. W. & Herbert, V. Separation of vitamin Bin-binding proteins of serum, gastric juice and saliva by rapid DEAE cellulose chromatography. Blood 29, 501, 1967. 32. Scott, J . M., Bloomfield, F. J.. Stebbins, R. & Herbert, V. Studies on derivation of transcobalamin 111 from granulocytes. Enhancement by lithium and elimination by fluoride of in vitro increments in vitamin B12-binding capacity. J . d i n . Invest. 53, 228, 1974.
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
Vitamin B,,-Binding Proteins 33.Simons, K . Vitamin BIZ binders in human body fluids and blood cells. Soc. Sci. Fenn. Commentat. B i d . 27, Fasc. 5, 1964. 34. Simons, K. & Grasbeck, R. Immunoelectrophoresis of human gastric juice. Clin. chim. Acta 8, 425, 1963. 35. Simons, K. & Weber, T. The vitamin BIZbinding protein in human leukocytes. Biochim. biophys. Acta (Amst.) 117, 201, 1966. 36. Stenman, U.-H. Amniotic fluid vitamin BIZbinding protein. Purification and characterization with isoelectric focusing and other techniques. Biochim. biophys. Acta (Amst.) 342, 173, 1974. 37. Stenman, U.-H. Characterization of R-type vitamin BIZ-binding proteins by isoelectric focusing. I. The relationship between transcobalamin I, transcobalamin I11 and the granulocyte R protein. Scand. J . Haemaf. 13, 129,1974.
Received 28 June 1974 Accepted 6 November 1974
155
38. Stenman, U.-H. Characterization of R-type vitamin BIZ-binding proteins by isoelectric focusing. 111. Cobalophilin (R proteins) in myeloproliferative states and leukocytosis. Scand. J. clin. Lab. Invest. 35, 157, 1975. 39. Stenman, U.-H. & Grasbeck, R. Gradients for isoelectric focussing at low pH. Biochim. biophys. Acta (Amst.) 286, 243, 1972. 40. Stenman, U.-H. & Nyberg, W. Vitamin BIZbinding proteins in fetal blood and amniotic fluid. Scand. J. clin. Lab. Invest. 23, Suppl. 108, 33, 1969. 41.Stenman, U.-H., Simons, K. & Griisbeck, R. Vitamin BIZ-binding proteins in normal and leukemic human leukocytes and sera. Scand. 1. clin. Lab. Invest. 21, 202, 1%8. 42. Vesterberg, 0. Isoelectric focusing of acidic proteins. Studies on pepsin. Acta chem. Scand. 27, 2415, 1973.
ISSN: 0036-5513 (Print) 1502-7686 (Online) Journal homepage: http://www.tandfonline.com/loi/iclb20
Clinical Chemistry: Characterization of R-Type Vitamin B12-Binding Proteins by Isoelectric Focusing U.-H. Stenman To cite this article: U.-H. Stenman (1975) Clinical Chemistry: Characterization of R-Type Vitamin B12-Binding Proteins by Isoelectric Focusing, Scandinavian Journal of Clinical and Laboratory Investigation, 35:2, 147-155 To link to this article: http://dx.doi.org/10.1080/00365517509087218
Published online: 28 Aug 2009.
Submit your article to this journal
Article views: 2
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iclb20 Download by: [University of California Santa Barbara]
Date: 06 November 2015, At: 03:06
Characterization of R-Type Vitamin B,,-Binding Proteins by Isoelectric Focusing 11. Comparison of Cobalophilin (R Proteins) from Different Sources
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
U.-H. STENMAN The Minerva Foundation Institute for Medical Research, Helsinki, Finland
Stenman, U.-H. Characterization of R-Type Vitamin B,,-Binding Proteins by Isoelectric Focusing. 11. Comparison of Cobalopilin (R Proteins) from Different Sources. Scand. J. clin. Lab. Invest. 35, 147-155, 1975. The R-type vitamin BIZ-binding proteins, here called cobalophilin, in human plasma, serum, leukocytes, gastric juice, amniotic fluid, saliva, and milk, were characterized by isoelectric focusing and gel filtration. Cobalophilin from all these sources is microheterogeneous, consisting of several isoproteins with isoelectric points (PI) between 2.3 and 5.0. Isuproteins with PI values of 3.0, 3.3, 3.6, 3.9, and 4.2 were found in most of the cells and fluids studied, but in different proportions. Milk contains isoproteins with PI values mainly between 4.0 and 4.7, and such components also occur in considerable amounts in saliva. The cobalophilin in these fluids also has a smaller Stokes radius than that from other sources. On the basis of the isoelectric patterns and the Stokes radii the isoproteins of cobalophilin are tentatively divided into two populations, a glandular one occurring in milk and saliva and a myelogenic one occurring in all the cells and fluids studied but only occasionally in milk. Key-words: Cobalophilin; gel filtration; isoelectric focusing; Stokes radius; transcobalamin; vitamin B12 U.-H. Stenman. M.D., T h e Minerva Foundation Institute for Medical Research, P.O.Bon 819, SF-00101 Helsinki 10, Finland
The term binder R was originally introduced to denote the gastric non-intrinsic factor vitamin B,,-binding protein with rapid electrophoretic mobility (13,34). Proteins with similar immunological properties have been demonstrated in most body fluids and blood cells (33), and the names R-type vitamin Blz-binding protein and R protein have been used ,for this protein, irrespective of its origin (12, 33). Jn electrophoresis R protein from different sources has electrophoretic mobilities varying from alpha to beta (12, 14, 33, 41). Thus the denomination R (for rapid mobility) is somewhat misleading, and the name cobalophilin is now introduced to replace the name R protein. Cobalophilin has been isolated from saliva (17), granulocytes (2), and amniotic fluid (36). It is a glycoprotein with an Mr value of
about 60,000, but in gel filtration cobalophilin from most sources behaves like a protein with an Mr value of about 120,OOO j12, 24, 35, 40, 41), obviously because of its high carbohydrate content, 33 per cent (2, 36). Salivary cobalophilin has a smaller apparent molecular size in gel filtration, 60,OOO-88,OOO f16, 23) and a lower carbohydrate content (17). Several names have been used for cobalophilin in plasma and serum: transcobalamin (TC) I (22), ‘high’ molecular weight B,, binder (28), TC L (19), TC I-!- TC I11 (3), and TC I third vitamin B,,-binding protein (4) binder 2 binder 3 (25). That two R-type vitamin Blz-binding proteins are present in serum ((3, 4, 32) is dubious, however. By isoelectric focusing it has been shown that the separation of cobalophilin in serum into an
+
+
148
U.-H. Stenman
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
a- (TC
I) and a ,&globulin fraction (31) (TC 111) is artificial. TC I and TC 111 merely represent a partial separation of the isoproteins of the microheterogeneous cobalophilin into two overlapping fractions (37). Cobalophilin in amniotic fluid 1(39), saliva (17), and granulocytes l(37) is also microheterogeneous, containing at least five components isoproteins - most of which are common to all these materials. Isoelectric focusing has now been used to ascertain the reasons for the differences in electrophoretic mobility and molecular size of cobalophilin from different sources.
MATERIAL AND METHODS Blood samples were obtained from healthy laboratory personnel. Cord blood was taken from the placental part of the umbilical cord immediately after delivery. Serum was separated 2.5 h after sampling. Plasma was prepared by addition of 2 mg EDTA (K,) per millilitre blood and centrifugation within 1 h. Amniotic fluid was taken at term by amniotomy. Samples contaminated by blood were discarded. Milk was collected at different times after delivery. The milk fat was removed by centrifugation. Leukocytes were prepared from BDTA blood by lysis of the erythrocytes with 0.155 mol/l NH,OI (35). Gastric juice was collected after histalog stimulation and depepsinized by raising the pH to 10 (33). Saliva was obtained from healthy laboratory personnel immediately before use. The samples were frozen in small aliquots as soon as possible after sampling (unless used immediately) and stored at -20 "C. Thawed samples were not refrozen. Isoelectric focusing was performed in a 110ml column LKB 8101 (LKB, Bromma, Sweden) essentially as described '(39) but with the following modifications. The dense solution has the following composition: 0.5 ml AmpholineB (LKB), pH 2 . 5 4 (20 per cent w/v), 27.0 ml glycerol ( 1 0 0 per cent v/v), 25.0 ml distilled water, and 1.5 ml of a stock solution of acids. This solution contains 45 mmol/l malic acid, 185
mmol/J formic acid, 85 mmol/l succinic acid, 250 mmoYl acetic acid, and 200 mmoYl propionic acid. The less dense solution contains 1 ml Ampholine, pH 3.5-5 (40 per cent w/v), 0.5 ml Ampholine, pH 4-6 (40 per cent w/v), and 52.5 ml water. The density gradient was formed with gradient mixer. The addition of Ampholine, pH 2.5-4.0, makes a smoother pH gradient between pH 3 and 4, but the results are comparable with those earlier obtained. The sample was mixed with glycerol and introduced directly into the column when half the gradient had been formed. This modification of the original method (39) was introduced because the proteins tended partly to remain in the, upper part of the column when the sample was mixed with the dense or the less dense solution. A maximum of 0.2 ml serum or plasma was applied, but better separation was obtained when 0.1 ml or less was used. The focusing time was 5 days; temperature, 10 "C; and potential, 400 V. The column was emptied with a pump, at a rate of 60 ml/h. One-millilitre fractions were collected and radioactivity and pH measured (39). Cobalophilin treated with sialidase and that in milk were also focused in a pure ampholyte gradient covering the pH range 3.5-6; focusing conditions: 1 per cent Ampholine, pH 3-6; sucrose density gradient, Ck50 per cent; 450 V; temperature, 10 "C; focusing time: 2 days. Attempts were also made to use an ampholyte gradient without acid solution for the pH range 2 . 5 4 (42). Good separation of the protein peaks was only occasionally obtained with this method, and unsuccessful focusing was always accompanied by blurring or tilting of the visible zone of methyl red (routinely added), which normally forms a horizontal band 0.5 cm wide at pH 3.75 ((39). Gel filtration was performed on Sephadex G-150 (Pharmacia, Uppsala, Sweden), column size 1.6 X 90 cm, with 0.1 moYl sodium phosphate buffer, pH 7.4, containing 0.02 per cent sodium azide. The fraction volume was 3.5 ml and the temperature + 4 "C. The void volume of the column was determined with Blue Dextran 2000 fPharmacia). The molecular size of cobalophilin from different sources was com-
+
+
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
Vitamin BIZ-BindingProteins
pared with that in amniotic fluid by the double isotope technique (36), using vitamin B,, labelled with 57Co and 5 8 C o i(The Radiochemical Centre, Amersham, U.K.). The Stokes radius was calcukted (1) on the basis of the known value, 4.61 nm (36), for cobalophilin in amniotic fluid. The vitamin BIZ-binding capacity was determined with haemoglobin-coated charcoal (10) in all samples and additionally with gel filtration l(24) in samples containing more than one binding protein - plasma, serum, and gastric juice. One nanogram radioactive vitamin B,, was incubated for 15 min with a volume of plasma or serum (usually 0.1 ml) giving a five- to ten-fold excess of vitamin BIZ; the total incubation volume was 1 ml. The other fluids studied were incubated with 5 ng vitamin BIZ in the same volume. From the association constants of cobalophilin (23) it can be calculated that under these conditions more than 98 per cent of the binding capacity will be saturated. Separation of bound and free vitamin B,, with haemoglobin-coated charcoal (10) or gel filtration (24) gave the same results. Gel filtration separates the binding proteins in plasma, serum (24), and gastric juice (12), making calculation of the binding capacity of individual binding proteins possible. Identification of vitamin BIZ-binding proteins as cobalophilin was based on their ability to react with antisera against cobalophilin in saliva (17) and amniotic fluid ((36). These antisera did not react with other vitamin BIZbinding proteins. Samples saturated with radioactive vitamin B,, were incubated with antiserum and subjected to gel filtration. Reaction with anti-cobalophilin serum resulted in a changed elution volume (8). Sialic acid was removed from cobalophilin with sialidase (Behringwerke, Mahrburg-Lahn, Germany, EC 3.2.1.18). To a 0.1-ml sample saturated with vitamin B,, was added 50 units sialidase (0.1 ml) and 0.3 ml 0.05 mol/l sodium acetate buffer @H 5.9) containing 0.154 mol/l NaCl and 9 mmol/l CaCl,; this mixture was incubated for 24 h at 3 7 "C during continuous dialysis against the acetate buffer. Controls were incubated without sialidase.
149
Table I. The unsaturated vitamin B1,-binding capacity of the cobalophilin in the different body fluids studied Range
No. of samples
-_
Adult plasma (EDTA) Adult serum Foetal plasma (EDTA)
Foetal serum Amniotic fluid Milk Saliva Gastric juice
0.133 0.360
0.055-0.330 5 0.090 -0.910 5
0.097 0.40 15.3 41.0 34.4 28.3
0.012 - 0.270 6 0.067 - 1.180 6 1.0-32 9.0-85 20.7 -45 6.4-65
57 6 6 8
RESULTS Vitamin B,,-binding capacity o f cobalophilin
Data for the unsaturated vitamin B,,-binding capacity of cobalophilin in different body fluids are summarized in Table I. The samples of plasma and serum were taken simultaneously from the same patients, and the values are therefore directly comparable. In both adult and foetal serum the free binding capacity of cobalophilin was three to four times as high as in EDTA plasma. Since the binding capacity of amniotic fluid had not been determined previously, a larger number of samples were studied. The values for milk were in the same range as reported earlier, 80 pg/l f l 8 ) and 20-157 pg/l (20). Zsoelectric focusing
Fig. 1 shows schematically the percentage distribution of radioactive vitamin B,, between the isoproteins of cobalophilin from different sources. Cobalophilin was always microheterogeneous, consisting of five or more isoproteins. One sample with a low (A), one with an average (B) and one with a high (C) proportion of acid components are presented. Amniotic fluid, adult and foetal plasma, serum, and leukocytes all contained the same components with isoelectric points (PI) at 3.0, 3.3, 3.6, 3.9, and 4.2. In addition, isoproteins with PI values of 2.3 and 2.7 were occasionally seen in amniotic fluid. The proportion of the
U.-H. Stenman
I50
Amnlotlc lluld
Plasma
Mllk
50:ed"lt'
A
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
B
-
Serum
0
A
l
,
tl, ~
Fig. 1. Diagram of the percentage distribution of radioactive vitamin BIZ among the isoproteins of cobalophilin separated by isoelectric focusing. Each bar represents one isoprotein, and its place on the abscissa is determined by its isoelectric point. The isoelectric patterns of three samples of the cobalophilin from each source are shown above each other. One sample each of extreme cases: more acid pattern (C), less acid pattern (A), and one sample representing the average pattern (B). Ordinate: percentage of radioactive vitamin BIZ contained in each isoprotein; abscissa: PI values of the components.
!,,wr C
0 2
3
4
5
2
3
4
5
2
pl values of the
3
4
5
2
3
4
5
components
components varied with the source. Plasma had a more acid pattern than serum, the pattern of which was slightly more acid than that of leukocytes. The difference in pattern between plasma and serum is explained by release of cobalophilin from granulocytes during clotting (37). Cobalophilin in adult plasma had a more acid pattern than that in foetal plasma. A similar but smaller difference was observed between adult and foetal leukocytes. The cobalophilin in amniotic fluid had an isoprotein pattern very similar to that in leukocytes. The differences in pattern between individual samples were mostly small and were greatest among the serum samples, probably owing to the variety in the admixture of cobalophilin from granulocytes. The same isoproteins of cobalophilin as were found in the above samples were also seen in gastric juice and saliva. l n addition, these body fluids contained various amounts of isoproteins isoelectric between 4.2 and 5.0.. Be-
cause the PI values of these differed from sample to sample, typical isoproteins could not be defined. Efforts were made to study gastric juice and saliva immediately after sampling (although this was not possible with all samples of gastric juice). In spite of this, the isoprotein patterns of saliva and gastric juice varied greatly. One explanation for this is that the cobalophilin in saliva and gastric juice IS probably derived from at least two sources (see Discussion). Another possible reason is degradation, probably desialylation, which obviously starts in vivo immediately after secretion, especially at the low pH in gastric juice. Milk conltains isoproteins with PI values between 4 and 5 (Fig. 2), as do saliva and gastric juice. But the isoproteins of milk, unlike those of these latter fluids, typically have PI values #of 4.0, 4.1, 4.25-4.30, 4.32-4.39, 4.5, and, occasionally, 4.7. Only one of the eight milk samples studied (Fig. 1C) contained mea-
Vitamin B , -Binding Proreins
-
-
S-
7 -
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
I a
0
20
40
60
80
100
F r a c t l o n no.
Fig. 2. Isoelectric focusing of the cobalophilin in milk (thick line) and the sialidase-treated cobalophilin in saliva (dotted line). pH is indicated by a thin line. Results of two experiments superimposed on each other.
surable amounts of 'acid isoproteins' with PI values of 3.0, 3.3, and 3.6. These accounted for only 8 per cent of all the bound vitamin BIZ. After treatment with sialidase, cobalophilin from different sources had fairly uniform isoelectric patterns. In shallow gradients covering the pH range 4-6, three to fmive peaks were found at pH values of 4.8, 4.9, 5.0, 5.1, and 5.2 'pig. 2). In most samples, the main isoprotein was the one isoelectric at pH 5.0, but the proportions of the other components varied. Controls incubated without sialidase also underwent changes in isoelectric pattern that indicated loss of sialic acid. These changes were small, however, except for in the samples of saliva, gastric juice, and leukocyte extract (cf. Ref. 41). After incubation, these contained no isoproteins isoelectric below p H 4.3. The change in isoprotein pattern caused by incubation without sialidase indicates that sialidase activity was present in some of the materials studied. This activity obviously starts to affect the isoprotein pattern of cobalophilin immediately after it has been secreted; its action continues until the sample is frozen, and it may even have some effect during focusing. No attempts were made to measure the autolysis taking place in vivo, but the effects on cobalophilin of different methods of analysis and handling were studied in saliva, which was most prone to autolysis. Storage of saliva for
151
18 h at f 4 "C changed the proportion of isoproteins isoelectric below/above pH 4 from 60140 to 30/70. This change was largely inhibited by saturation of the sample with vitamin B,2. Addition of subsaturating amounts of vitamin B12 also changed the distribution of the vitamin, which was preferentially bound to the most acid isoproteins. Whether desialylation occurs during focusing was checked in samples of saliva focused in an acid-free gradient covering the pH range 4-6, in which focusing is completed in 1 day. This method gave the same proportions of isoproteins isoelectric below and above pH 4 as the pH gradient normally used. Possible desialylation or deamidation of acid iisoproteins at low pH was checked by refocusing of individual peaks. No change in the PI value was observed. Identification of focused vitamin B12-binding proteins as cobalophilin was no problem with samples containing only one type of binding protein - leukocytes, saliva, amniotic fluid, and milk. The radioactive vitamin bound by any of these materials showed a changed elution pattern in gel filtration after incubation with anti-cobalophilin serum. The non-cobalophilin vitamin B,,-binding proteins in plasma, serum (TC II), and gastric juice (intrinsic factor) were occasionally removed by gel filtration before isoelectric focusing. T h h step was not routinely performed because of the risk of denaturation. TC I1 is isoelectric at around pH 6 (37) and does not disturb the pattern of cobalophilin. The most acid isoprotein of intrinsic factor has a PI value of 4.84 (11) and may thus overlap with the least acid isoproteins of cobalophilin. This was observed only in one sample (Fig. 1, gastric juice sample B). In this case the identity of the binding protein was established by incubation with anti-cobalophilin serum and gel filtration.
Gel filtration The Stokes radius of cobalophilin from all sources except saliva and milk did not differ significantly from that of the reference protein. The Stokes radii of cobalophilin in saliva and milk were smaller, and the values of the
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
152
U.-H. Stenman
former varied greatly. Samples of saliva with a high proportion of acid isoproteins had a fairly large and samples with little acid components a small Stokes radius. Therefore the different isoproteins of cobalophilin in saliva were also studied. There was a trend towards smaller Stokes radius with increasing PI value, and isoproteins with PI values above 4.8 were clearly smaller than the others. The molecular size of cobalophilin from most sources was reduced by 2-4 per cent by treatment with sialidase. Desialylated salivary cobalophilin had a Stokes radius of 3.7-3.8 nm - that is, about 15 per cent smaller than the native protein and the same as the component with PI 5.0 I,Table 11) occurring in some samples of native saliva. DISCUSSION Isoelectric focusing has a high resolving power and permits accurate characterization of separated proteins in terms of their PI values. This made it possible to demonstrate that cobalophilin from different cells and fluids consists mainly of the same isoproteins in different proportions. These isoproteins are not all found in every source material, and it is actually possible to discern two populations of isoproteins with typical pI ranges and occurrence. One population, tentatively called myelogenic,
has a PI range of 2.3-4.2 and occurs in pure form in leukocytes, plasma, serum, and amniotic fluid. The other population has a PI range of 4.0-5.0 and may be called glandular because of its occurrence in milk. Saliva contains a mixture of the two populations, as does possibly gastric juice; glandular isoproteins found in the latter, however, may be the result of autolysis or contamination with saliva. A difference between the myelogenic and glandular isoprotein populations was also demonstrated by gel filtration, the latter having a smaller average molecular size. The mixed population of saliva obviously explains the variety in molecular-size values found in this and earlier studies (5, 16, 17, 23), but denaturation during storage or purification is probably partially responsible for the lowest values reported (16, 17). The Stokes radius of cobalophilin in milk is compatible with an earlier Mr estimation of 102,000 (9). A further reason for dividing the isoproteins of cobalophilin into two populations is that two cell types have been shown to synthesize cobalophilin in vitro - granulocytes (35) with their precursors (30) and the serous cells of the salivary glands (26). The denominations myelogenic and glandular are only tentative, however, and the possibility that other cell types can produce cobalophilin should be kept in mind. Nevertheless, the similarity between
Table 11. Stokes radius of cobalophilin from milk, saliva, and isoproteins of the salivary cobalophilin* Stokes radius (nm)
______
Milk
Saliva Salivary isoprotein, 3.6t Salivary isoprotein, 3.9 Salivary isoprotein, 4.1 Salivary isoprotein, 4.5 Salivary isoprotein, 4.8 Salivary isoprotein, 5.0
Mean
Range
4.31 4.36 4.49 4.40 4.37 4.39 4.06 3.81
4.11 4.39 4.05 4.56
-
No. of expts. 6 5 1 1 1 1 1 1
___
*Determined by gel filtration (1) on Sephadex G-150 by the double isotope technique. The cobalophilin in amniotic fluid was used as reference protein (Stokes radius, 4.61 nm; Ref. 36). Temperature f 4 "C. The Stokes radii of cobalophilin from other sources did not differ significantly from that of the reference protein. ?This value is the isoelectric point of the isoprotein.
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
Vitamin B,,-Binding Proteins
the isoproteins of cobalophilin in granulocytes and those in most of the body fluids studied supports the view that granulocytes are an impontant source of cobalophilin (35). The cobalophilin in plasma has a higher proportion of acid isoproteins than that in granulocytes, possibly because cobalophilin excreted in vivo contains more sialic acid than the intracellular form. Intrinsic factor in gastric juice is also more acid than the intracellular form ((29).The most acid isoproteins may also be derived from granulocyte precursors (30); the proportion of these isoproteins is increased in the plasma of many patients with chronic myelogenous leukaemia (38). The isoprotein pattern of cobalophilin in amniotic fluid is very similar to that in leukocytes of adults, but differences in the carbohydrate moiety (36) may indicate that cobalophilin in amniotic fluid is not derived from granulocytes. Variation in the carbohydrate composition of a glycoprotein derived from different ti ssues is common (36), and differences in sialic acid content have been suggested to be one cause of the variety in electrophoretic mobility of cobalophilin in different body fluids and cells (41). Sialidase has been shown to abolish the differences in electrophoretic mobility (41), and these results were confirmed by isoelectric focusing. However, treatment with sialidase did not transform the isoproteins to a single substance. This was possibly due to incomplete removal of sialic acid or to further degradation. The latter phenomenon obviously explains the considerable reduction in the size of the cobalophilin in saliva during this treatment. The variation in electrophoretic mobility of cobalophilin from various sources m(12, 14, 33, 41) can be correlated with differences in isoprotein pattern. Thus, cobalophilin with a high proportion of acid isoproteins has a higher mobility than cobalophilin with a low proportion of these. This is the case, for instance, with foetal and adult plasma, which have been considered to contain different species of vitamin B,,-binding proteins of R-type - TC I and foetal binder (27). According to the present study, these fluids contain the same isoproteins of I
153
cobalophilin but in different proportions. The nature of the polycythaemia Vera binder is similar (38). It is therefore advisable to describe vitamin B,,-binding proteins of cobalophilin (or R) type in terms of their isoprotein patterns rather than by the average electrophoretic mobility of their isoprotein population or by their behaviour in ion exchange chromatography with stepwise elution. No transport or enzyme function has yet been ascribed to cobalophilin, and two brothers suffering from lack of this protein had no overt disease (6). Thus the function, proposed by Gullberg (19), of aiding the organism to resist infections by depriving vitamin BIZdependent microorganisms of a growth factor seems to be the most logical one. This hypothesis would also explain the abundance of R protein in tears ((13,granulocytes, saliva, milk, and amniotic fluid. The useful classification of vitamin B12binding proteins into three groups - intrinsic factor, TC 11, and R proteins '(12, 41) - has gained acceptance (7, 21). However, the name R protein has now been changed to cobalophilin. This word has been coined by analogy with siderophilin and does not impute any function to the protein other than the ability to bind vitamin BIZ. ACKNOWLEDGEMENTS Supported by the Finnish National ,Research Council for Medical Sciences, the Finska Vetenskaps-Societeten, the Finska Lakaresallskapet, the Sigrid JusClius Foundation, and Nordisk Insulinfond. I thank Dr. Ralph Grasbeck and Miss Liisa Puutula, M. Sc., for valuable criticism of the manuscript and Miss Anneli Hangasoja for excellent technical assistance. REFERENCES 1. Ackers, G . K. Molecular exclusion and re-
stricted diffusion processes in molecular-sieve chromatography. Biochemistry 3, 723, 1964. 2. Allen, R. H. & Majerus, P. W. Isolation of vitamin BIZ-binding proteins using affinity chromatography. 11. Purification and properties of a human granulocyte vitamin BIZ-binding protein. J . biol. Chem. 247, 7702, 1972.
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
154
U.-H. Stenman
3. Bloomfield, F. J. & Scott, J. M. Identification of a new vitamin BIZ binder (transcobalamin 111) in normal human serum. Brit. J , Haemat. 22, 33, 1972. 4. Carmel, R. Vitamin BIZ-binding protein abnormality in subjects without myeloproliferative disease. 11. The presence of a third vitamin BIZ-binding protein in serum. Brit. J . Haemat. 22, 53, 1972. 5 . Carmel, R. The vitamin BIZ-binding proteins of saliva and tears and their relationship to other vitamin BIZ binders. Biochirn. biophys. Acra ( A m s t . ) 263, 747, 1972. 6 . Carmel, R . & Herbert, V. Deficiency of vitamin BIZ-binding alpha globulin in two brothers. Blood 33, 1, 1969. 7 . England, J . M., Clarke, H. G. M., Down, M. C. & Chanarin. I. Studies on the transcobalamins. Brit. 1. Haemat. 25, 737, 1973. 8 . Finkler, A. E., Green, P. D. & Hall, C. A. lmmunological properties of human vitamin BIZ binders. Biochirn. biophys. Acta (Amst.) 200, 151, 1970. 9. Finkler, A. E., Rappazzo, M. E. & Hall, C. A. Cyanocobalamin binding protein in human milk. Fed. Proc. 26, 695, 1967. 10. Gottlieb, C., Lau, K.-S., Wasserman, L. R. & Herbert, V. Rapid charcoal assay for intrinsic factor (IF), gastric juice unsaturated B l ~ binding capacity, antibody to IF, and serum unsaturated BIZ binding capacity. Blood 25, 875, 1965. 7 I . Grasbeck, R. Microheterogeneity of the complex between human intrinsic factor and cyanocobalamin demonstrated by isoelectric focusing. Acta cherii. scand. 22, 1041, 1968. 12. Grasbeck, R . Intrinsic factor and the other vitamin 812 transport proteins. Progr. Hernut. 6, 233, 1969. 13. Grasbeck, R., Simons, K. & Sinkkonen, I. Purification of intrinsic factor and vitamin BIZ binders from human gastric juicc. Ann. M e d . exp. Fenn. 40, Suppl. 6 , 1962. 14. Griisbeck, R., Stenman, U.-H., Puutula, L. & Visuri, K. A procedure for detaching bound vitamin BIZ from its transport proteins. Biochitn. biophys. Acta (Amst.) 158, 292, 1968. 15. Grasbeck, R. & Takki-Luukkainen, I.-T. Vitamin BIZ-binding substance in human tear fluid. Acta ophthal. ( K b h . ) 36, 860, 1958. 16. Grasbeck, R . & Visuri, K . Isolation of the vitamin BIZ binding protein of human saliva. Scand. 1. c h i . Lob. Invesr. 21, Suppl. 101, 13. 1968. 17. Grasbeck, R., Visuri, K. & Stenman, U.-H. The vitamin BIZ-binding protein in human saliva. Isolation, characterization and comparison with the unpurified protein. Biochini. biophys. Acta (Amst.)263, 721, 1972.
18. Gregory, M. E. & Holdsworth, E. S. The OCcurrence of a cyanocobalamin-binding protein in milk and the isolation of a cyanocobalinprotein complex from sow’s milk. Biochem. J. 59, 329, 1955. 19. Gullberg, R. Vitamin Biz-binding proteins in normal human blood plasma and senun. Scnnd. J . Haernat. 9, 639, 1972. 20. Gullberg, R. Possible influence of vitamin BIZbinding protein in milk on the intestinal flora in breast-fed infants. I. BIZ-binding proteins in human and bovine milk. Scond. 1. Gastroenf. 8, 497, 1973. 21. Hall, C. A. Vitamin Blz-binding proteins of man. Ann. intern. Med. 7 5 , 297, 1971. 22. Hall, C. A. & Finkler, A. E. The dynamics of transcobalamin 11. A vitamin BIZ binding substance in plasma. 1. Lab. clin. M e d . 65, 459, 1965. 23. Hippe, E. Studies on the cobalamin-binding protein of human saliva. Scand. 1. d i n . Lab. Irivcst. 29, 59, 1972. 24. Hom, B., Olesen, H. & Lous, P. Fractionation of vitamin BIZ hinders in human serum. 1. Lab. d i n . Med. 68, 958, 1966. 25.Hurlimann, I . & Zuber, C. Vitamin BIZbinders in human body fluids. I. Antigenic and physico-chemical characteristics. Clin.exp. Immunol. 4 , 125, 1969. 26. Hurlimann, I. & Zuber, C. Vitamin BIZ-binders in human body fluids. 11. Synthesis in vitro. Clin. r x p . Immunol. 4. 141, 1969. 27. Kumento, A. Studies on the serum binding of vitamin U12 in the newborn human infant. Arru puedior. scand., Suppl. 194, 1969. 2X.Lawrence. C. The heterogeneity of the high molecular weight 8 1 2 binder in serum. Blood 33, 899, 1969. 29. Marcoullis, G . & Grasbeck. R. Vitamin Bizbinding proteins in human gastric mucosa. General pattern and demonstration of intrinsic factor isoproteins typical of mucosa. Scand. 1. d i n . Lab. invest. 35, 5 , 1975. 30. Rachmilewitz, B., Rachmilewitz, M. & Gross, J. A vitamin 812 binder with transcobalamin I characteristics synthesized and released by human granulocytes in vitro. Brit. J . Haemat. 26. 557, 1974. 31. Retief, F. P., Gottlieb, C. W., Kochwa, S., Pratt, P. W. & Herbert, V. Separation of vitamin Bin-binding proteins of serum, gastric juice and saliva by rapid DEAE cellulose chromatography. Blood 29, 501, 1967. 32. Scott, J . M., Bloomfield, F. J.. Stebbins, R. & Herbert, V. Studies on derivation of transcobalamin 111 from granulocytes. Enhancement by lithium and elimination by fluoride of in vitro increments in vitamin B12-binding capacity. J . d i n . Invest. 53, 228, 1974.
Downloaded by [University of California Santa Barbara] at 03:06 06 November 2015
Vitamin B,,-Binding Proteins 33.Simons, K . Vitamin BIZ binders in human body fluids and blood cells. Soc. Sci. Fenn. Commentat. B i d . 27, Fasc. 5, 1964. 34. Simons, K. & Grasbeck, R. Immunoelectrophoresis of human gastric juice. Clin. chim. Acta 8, 425, 1963. 35. Simons, K. & Weber, T. The vitamin BIZbinding protein in human leukocytes. Biochim. biophys. Acta (Amst.) 117, 201, 1966. 36. Stenman, U.-H. Amniotic fluid vitamin BIZbinding protein. Purification and characterization with isoelectric focusing and other techniques. Biochim. biophys. Acta (Amst.) 342, 173, 1974. 37. Stenman, U.-H. Characterization of R-type vitamin BIZ-binding proteins by isoelectric focusing. I. The relationship between transcobalamin I, transcobalamin I11 and the granulocyte R protein. Scand. J . Haemaf. 13, 129,1974.
Received 28 June 1974 Accepted 6 November 1974
155
38. Stenman, U.-H. Characterization of R-type vitamin BIZ-binding proteins by isoelectric focusing. 111. Cobalophilin (R proteins) in myeloproliferative states and leukocytosis. Scand. J. clin. Lab. Invest. 35, 157, 1975. 39. Stenman, U.-H. & Grasbeck, R. Gradients for isoelectric focussing at low pH. Biochim. biophys. Acta (Amst.) 286, 243, 1972. 40. Stenman, U.-H. & Nyberg, W. Vitamin BIZbinding proteins in fetal blood and amniotic fluid. Scand. J. clin. Lab. Invest. 23, Suppl. 108, 33, 1969. 41.Stenman, U.-H., Simons, K. & Griisbeck, R. Vitamin BIZ-binding proteins in normal and leukemic human leukocytes and sera. Scand. 1. clin. Lab. Invest. 21, 202, 1%8. 42. Vesterberg, 0. Isoelectric focusing of acidic proteins. Studies on pepsin. Acta chem. Scand. 27, 2415, 1973.
