Human Plasma R-Type Vitamin B ,,-binding Proteins ROLE OF TRANSCOBALAMIN TRANSCOBALAMIN AND THE NORMAL GRANULOCYTE PROTEIN IN THE PLASMA TRANSPORT OF VITAMIN

The normal human granulocyte vitamin B,,-binding protein, transcobalamin I, and transcobalamin III, have been labeled with lZSI-labeled N-succinimidyl 3-(4-hydroxyphenyl)propionate and utilized for plasma clearance studies performed with rabbits. Both moieties of 1251-labeled granulocyte vitamin B,,-binding protein-[5’Co]vitamin B,, were cleared rapidly from the plasma (>90% by 5 min) by the liver. After 30 min, the bulk of the lZ51 reappeared in the plasma in small molecular weight ( < 1000) form and was rapidly excreted in the urine. After 60 min the bulk of the [“Colvitamin B,, reappeared in the plasma bound to rabbit transcobalamin II and was subsequently taken up by a variety of tissues. Approximately 15% of the “Y-labeled granulocyte vitamin B,,-binding protein. [57Co]vitamin B,, was excreted intact into the bile during the period from 10 to 80 min after injection. The uptake of the complex was blocked

The normal human granulocyte vitamin B,,-binding protein, transcobalamin I, and transcobalamin III, have been labeled with lZSI-labeled N-succinimidyl 3-(4-hydroxyphenyl)propionate and utilized for plasma clearance studies performed with rabbits. Both moieties of 1251-labeled granulocyte vitamin B,,-binding protein- [5'Co]vitamin B,, were cleared rapidly from the plasma (>90% by 5 min) by the liver. After 30 min, the bulk of the lZ51 reappeared in the plasma in small molecular weight ( < 1000) form and was rapidly excreted in the urine. After 60 min the bulk of the ["Colvitamin B,, reappeared in the plasma bound to rabbit transcobalamin II and was subsequently taken up by a variety of tissues. Approximately 15% of the "Y-labeled granulocyte vitamin B,,-binding protein.
[57Co]vitamin B,, was excreted intact into the bile during the period from 10 to 80 min after injection. The hepatic uptake of the protein*vitamin B,, complex was blocked by the prior injection of desialyzed fetuin but not by native fetuin. Similar results were obtained with Y-labeled transcobalamin 111. [5'Co]vitamin B,,. Approximately 90% of both moieties of Y-labeled transcobalamin I.
[57Co]vitamin B,, had prolonged plasma survivals similar to that of 13Y-labeled bovine serum albumin. After treatment with neuraminidase, both moieties of the Y-labeled transcobalamin 1.
["Colvitamin B,, complex were cleared rapidly from the plasma by the liver in a manner that was indistinguishable from that observed in the case of untreated granulocyte vitamin B,,-binding protein and transcobalamin III. These observations indicate that desialyzed transcobalamin I and the native forms of the granulocyte vitamin B,,-binding protein and transcobalamin III are cleared from plasma by the mechanism elucidated by Ashwell and Morel1 (Ashwell, G., and Morell, A. G. (1974) Ado. Enzymol. 41, 99-128) that is capable of clearing a wide variety of asialoglycoproteins.
These observations have implications concerning the function of the human R-type vitamin B,,-binding proteins, the nature of the enterohepatic circulation of vitamin B 12, the biologic significance of the mechanism described by Ashwell and Morel& and the etiology of the increased plasma concentration of human R-type protein that occurs frequently in chronic myelogenous leukemia and occasionally in hepatocellular carcinoma and other solid tumors.
Human plasma contains three vitamin B,,-binding proteins that are referred to as transcobalamin I, II, and III (l-5). Transcobalamin II is immunologically distinct from the other two proteins, has a molecular weight of 38,000 based on gel filtration (2,3), and is not a glycoprotein (5). Transcobalamin II facilitates the cellular uptake of vitamin B,, by a variety of cells (6)(7)(8)(9). Transcobalamin II contains only 10 to 25% of the total plasma vitamin B,, (1, lo), but [5'Co]vitamin B,, bound to this protein is cleared rapidly (11,12) from human plasma with a t, of 5 to 90 min. Recent studies (13)  and have apparent molecular weights in the range of 120,000 to 150,000 based on gel filtration.
Transcobalamin I contains more sialic acid (18 7708 uersus 11 residues) and less fucose (9 uersus 20 residues) than transcobalamin III (1). Recent studies (1,4) indicate that a high percentage of the transcobalamin III present in human plasma is an artifact in the sense that it is released in unaltered form from granulocytes in vitro after blood is collected. The function of the R-type proteins is unknown. Approximately 80% of the total plasma vitamin B,, is bound to transcobalamin I (l-3) but studies performed in humans indicate that ["Colvitamin B,, bound to this protein is cleared slowly from plasma with a t, of 9 to 12 days that is similar to that of albumin (11,12). Carmel  proteins; less than 10% release was observed when these proteins were incubated in the absence of neuraminidase. Other Methods-1*61, "'Co, and Is11 were assayed on a dual channel Packard Auto-Gramma counter as described elsewhere (13). All other methods were performed as described in the accompanying paper in this series (1).

Labeling
of Proteins with '*"I and 1311-When the '*"I-labeled human granulocyte vitamin B,,-binding protein was saturated with [6'Co]vitamin B,, and applied to a column of Sephadex G-150 the elution profile presented in Fig. L4  of transcobalamin I and transcobalamin III (data not presented). Greater than 98% of the "? and "Co eluted in an asymmetrical peak that centered on Fraction 28. Greater than 98% of the lZ61 and "'Co present in Fractions 25 to 29 were adsorbed by rabbit anti-human milk R-type vitamin B,,-binding protein-Sepharose; less than 5% were adsorbed by rabbit control serum-Sepharose, rabbit anti-human transcobalamin II-Sepharose, or by vitamin B,,-Sepharose. has not been established. 1311-Labeled preparations of apo-human granulocyte vitamin B,,-binding protein, apo-transcobalamin I and apo-transcobalamin III eluted from Sephadex G-150 as single symmetrical peaks of radioactivity with apparent molecular weights of 150,000. Greater than 90% of the 1311 in each preparation was selectively adsorbed by rabbit anti-R-type protein-Sepharose. From 60 to 70% of the "'1 was adsorbed by vitamin B,,-Sepharose; less than 10% was adsorbed when a 2-fold excess of free vitamin B,, was added to the proteins 30 min before they were applied to vitamin B,,-Sepharose. The inability to achieve 100% binding of apo-R-type protein to vitamin B,,-Sepharose is consistent with previous observations (1, 21) that indicate that a portion (30 to 50%) of all R-type proteins are denatured, i.e., they lose the ability to bind vitamin B,, when they are renatured from guanidine in the absence of vitamin Bl,.
Hepatic Uptake of 'T-Labeled Vitamin B,,-binding Proteins Containing Bound ["'ColVitamin B,,-The data presented in Table I Fig. 2A reveal that both moieties of the 1261-labeled granulocyte vitamin B,,-binding protein.
[57Co]vitamin B,, complex reappeared in the plasma after their initial uptake by the liver. The "'1 moiety began to reappear 30 min after injection and was rapidly excreted in the urine. The peak plasma level and the period of maximal urinary excretion of lz51 both occurred between 45 and 75 min. The "Co moiety did not begin to reappear in the plasma until 60 min after injection and did not reach its peak plasma level until 120 min after injection. The plasma level of "Co declined slowly after 120 min; only negligible amounts were excreted in the urine. Similar results were observed with the ""I-labeled transcobalamin III* [6'Co]vitamin B,, complex ( Fig. 2B) and with the desialyzed '*"I-labeled transcobalamin 1.
The data presented in Fig ["Colvitamin B,, complex was missing from the plasma 5 min after injection and that approximately 17% of the lZsI was excreted in the urine during the subsequent 180 min. The remaining material (85 to 90%) disappeared slowly at a rate equal to that of 1311-labeled bovine serum albumin. This similarity in the clearance rates of bovine serum albumin and transcobalamin I continued for at least 5 days (data not presented).
Properties of 1261-Labeled Material and ["'Co]Vitamin B,, Released from Liver-An 80-min plasma sample was obtained from a rabbit that had been administered lZSI-labeled granulocyte vitamin B,,-binding protein.
["'Colvitamin B,, and the sample was subjected to gel filtration on Sephadex G-150. The elution profile obtained is presented in Fig. 1C. The elution profiles of lzsI and [5'Co]vitamin B,, both differed markedly from their preinjection profiles (Fig. 1B) and from each other. Less than 2% of either isotype eluted in the 150,000 apparent molecular weight position of intact 1261-labeled granulocyte vitamin B,,-binding protein.
The nature of the 70,000 apparent molecular weight 1*51-labeled material is unknown although the Y appears to be covalently bound to some component, possibly rabbit albumin, since it is not dialyzable when it is dialyzed against 1% sodium dodecyl sulfate or 7.5 M remove sialic acid, the lzsI and "Co moieties were both cleared guanidine HCl in the presence of 1% 2-mercaptoethanol. The 7710  e Per cent of total amount administered. d Assumed to be 100% and used to calculate the plasma volume.
fact that similar material has been observed in experiments in which lZ51-labeled rabbit and human transcobalamin II were studied in rabbits (13) suggests that it is formed after the catabolism of a number of similarly Y-labeled proteins. The nature of the small molecular weight (< 1000) lz61 is also unknown although free lz61 and '261-labeled N-succinimidyl 3-(4-hydroxyphenyl) propionate, either free or attached to one or a few amino acids, are likely possibilities.
All of the 12'1 excreted in the urine in the experiments shown in Fig. 2 had a similar apparent molecular weight of less than 1000 (data not presented).
All of the [5'Co]vitamin B,, present in the 80-min plasma sample (Fig. 1C) eluted from Sephadex G-150 with an apparent molecular weight of 40,000. This value is the same as the apparent molecular weight of rabbit transcobalamin II which accounts for >80% of the total, and >98% of the unsaturated, vitamin B ,,-binding protein present in rabbit plasma (13). The ["Co Jvitamin B 11 also resembled rabbit transcobalamin II* ["Colvitamin B,, in that it was precipitated by chicken anti-human transcobalamin II sera but was not precipitated by rabbit anti-human transcobalamin II sera or rabbit antihuman R-type protein sera.
Gel filtration results similar to those described above were also observed with a plasma sample obtained 180 min after the injection of 1*61-labeled granulocyte vitamin B,,-binding protein. [ the liver at a faster rate than ["Colvitamin B,, and that the "'1 reaches its peak concentration in the kidney at the same time as its maximal excretion in the urine in the form of small molecular weight ( < 1000) fragments (see above). lZ51 does not accumulate in the heart or lung at any time period. ["'Co]-Vitamin B,, accumulates in the kidney, heart, and lungs, but only after its release from the liver and its subsequent binding to rabbit transcobalamin II in the plasma (see above). This late tissue distribution of [5'Co]vitamin B,, is the same, with the exception of the intestine (see below), as that observed when rabbit transcobalamin II.
The time course of appearance and distribution of "'1 and ["Colvitamin B L2 in the small intestine is distinct from that of the other organs (see Fig. 3). Both moieties appear in equal amounts in the proximal small intestine between 5 and 30 min after injection and reach their maximal values by 60 min. At later time periods (120 to 180 min) [57Co]vitamin B,, is present in excess of lZ51 and the bulk of both moieties are present in the distal small intestine, which is where vitamin B,, appears to be absorbed from the gastrointestinal tract in rabbits (22). The "'1 and [5'Co]vitamin B,, enter the intestine via the bile, since when the common bile duct was cannulated in rabbits the level of both moieties in the intestine at 60 and 180 min fell by over 90%. The time course of appearance of lZ51 and [57Co]vitamin B,, in the bile is shown in Fig. 4. When a pooled sample of bile collected from 0 to 180 min after injection was subjected to gel filtration on Sephadex G-150, greater than 90% of the "'1 and ["Colvitamin B,, eluted together in a single peak with an apparent molecular weight of approximately 150,000 (data not presented). Greater than 90% of both moieties were selectively adsorbed by rabbit anti-human R-type protein-Sepharose. Results indistinguishable from those described in the preceding paragraphs for the granulocyte protein were also obtained with 1251-labeled transcobalamin 111.
["Colvitamin B,, (data not presented). The results of similar early time period injected.
In addition to 'S'I-labeled bovine serum albumin, the injected samples contained: A, 1261-labeled granulocyte vitamin B,,-binding protein.
[ Plasma Survivals of Apo and Holo-R-type Proteins-The presence of bound vitamin B,, had no effect on the hepatic uptake of the human granulocyte vitamin B ,,-binding protein and transcobalamin III. The 1311 label originally present on these apo-proteins was excreted slightly more rapidly in the urine than was the lZ51 label originally present on the holoproteins. The significance of this difference has not been determined although it could reflect a difference in susceptability to lysosomal proteases within the hepatocyte. Apo-transcobalamin I was cleared from plasma twice as fast as holo-transcobalamin I, although both forms were cleared much more slowly than their granulocyte vitamin B,,-binding protein and transcobalamin III counterparts. It should be noted that all of the differences observed between the apo-and holo-forms of these three proteins could be related to the fact that 30 to 40% of each apo-protein preparation appears to be present in a denatured state since it does not bind to vitamin B,,-Sepharose (see above). DISCUSSION Ashwell and Morel1 and their associates (23)  Terminal galactose residues appear to be required for glycoprotein binding to common receptors that are present on hepatocyte plasma membranes since binding is abolished by modification or removal of galactose residues from the glycoprotein or by the attachment of sialic acid distal to galactose on the glycoprotein.
Within minutes after binding to plasma membrane receptors, asialoglycoproteins appear to enter hepatocytes intact, presumably by pinocytosis, and are degraded by lysosomal enzymes over the ensuing 30 to 90 min. The biological function of this phenomenon has been difficult to elucidate since all of the proteins listed above appear to be present normally in plasma in their fully sialated forms. It has been postulated that sialic acid is slowly released from these proteins in viuo by neuraminidase but this has not been demonstrated.
The studies presented here demonstrate that transcobalamin I is included among those glycoproteins whose desialyzed forms are cleared rapidly from plasma by the mechanism of Ashwell and Morell. The studies concerning the granulocyte vitamin B,,-binding protein and transcobalamin III provide additional evidence for the concept that the latter protein is secreted by granulocytes in unaltered form (1,4,24). This appears to be the first example of a glycoprotein that is secreted from cells in a form such that the native protein is cleared rapidly from plasma by the mechanism of Ashwell  intact form, although the mechanism by which this occurs also remains to be determined.
It is not known whether other glycoproteins are excreted intact into the bile after being taken up by hepatocytes but if this is a general phenomenon it appears that an examination of the glycoprotein composition of bile could provide important information concerning the kinds and amounts of naturally occurring glycoprotein that are cleared from plasma in uiuo by the mechanism of Ashwell and Morel1 (16).
The plasma survival data presented here appear relevant to what occurs in humans since significant species differences have not been observed in the process described by Ashwell and Morel1 (16). This view is supported by the fact that [5'Co]vitamin B,, bound to transcobalamin I has a prolonged (tH = 9 to 12 days) plasma survival in normal human subjects (12). Direct evidence that some human R-type proteins are cleared rapidly from human plasma by the liver is not available although several studies (27,28) can be interpreted in a way that supports this possibility.
The fact that human bile contains 3 to 9 Kg of vitamin B,, per day (29) and the fact that the unsaturated vitamin B,,-binding protein in human bile is an R-type protein (30) indicate that significant amounts of R-type protein.vitamin B,, may be cleared from human plasma by the liver in uiuo. Much if not all of the R-type protein could be derived from granulocytes since the granulocytes produced daily in man (31) contain enough R-type protein to bind 100 to 150 c(g of vitamin B,, (32).
The function of the human R-type proteins has been difficult to define although recent studies by Gilbert (33) suggest that the granulocyte vitamin B,,-binding protein may play an active role in the handling of intra-and extracellular bacteria by leaching vitamin B,,. A similar antibacterial function might also explain the presence of high levels of R-type vitamin B,,-binding protein in human secretions since R-type proteins might serve to bind vitamin B,, and prevent its utilization by bacteria that require the vitamin for growth. A similar function has been well documented for the iron-binding protein lactoferrin (34,35) which is present in granulocytes (36) and a number of secretions. The significance of our observation that the 7713 granulocyte vitamin B,,-binding protein is cleared rapidly from plasma by the liver has not been established although it indicates the presence of a mechanism by which vitamin B,, present in areas of cell necrosis and infection within the body can be delivered exclusively to the liver rather than to cells throughout the body as would occur if such vitamin B,, were bound by transcobalamin II (13). This mechanism could be important if the liver controls the amount or coenzyme form of vitamin B,, bound to transcobalamin II in plasma or if the liver contains a mechanism for distinguishing native vitamin B,, from vitamin B,, analogs which are synthesized by bacteria (37) and might be harmful to certain cells within the body. The latter mechanism is suggested by the fact that R-type proteins are capable of binding a wide variety of vitamin B,, analogs (38,39) and by the fact that the liver could conceivably dispose of such analogs by preferentially secreting them into the bile. The fact that intrinsic factor binds a much narrower range of vitamin B,, analogs (38,39) indicates that many such analogs would not be reabsorbed from the intestine. It is of interest, in regard to this type of protective mechanism, that one of the two brothers with congenital R-type protein deficiency has a poorly defined neurologic illness that is clinically similar to multiple sclerosis (15).
The existence of a granulocyte-mediated mechanism for the transport of vitamin B,, exclusively to the liver suggests a possible biologic function for the process described by Ashwell and Morel1 and suggests that similar mechanisms may exist for iron and for other vitamins and metals. Transport of this kind could serve to regulate various metabolic systems within the liver as well as serve a scavenger function.