- Email: [email protected]
Transfer of cobalamin from the cobalamin-binding protein of egg yolk to R binder of human saliva and gastric juice
GASTROENTEROLOGY
1990;98:1460-1466
Transfer of Cobalamin From the Cobalamin-Binding Protein of Egg Yolk to R Binder of Human Saliva and Gastric Juice ALBERT0 Department California
DEL CORRAL of Medicine,
and RALPH CARMEL
University of Southern California School of Medicine,
Patients may fail to absorb cobalamin (vitamin B,,) bound to food even when they have adequate intrinsic factor to absorb free cobalamin normally. We studied cobalamin transfer from egg yolk cobalaminbinding protein to human saliva and gastric juice as a model of this important first step in cobalamin assimilation. The cobalamin-binding protein of egg yolk eluted with human R binder on Sephadex gel chromatography and bound cobalamin with a comparable affinity, but it did not cross-react with R binder immunologically. Transfer of cobalamin from egg yolk to saliva or gastric juice R binder did not occur at neutral pH. Slight transfer (8%-Z% of the 57Cocobalamin bound to egg yolk) occurred when the saliva was acidified to pH 1.5. This minor transfer by acid was not inhibited by pepstatin A, a pepsin inhibitor. Acidification caused variable transfer to gastric juice R binder (12%-40%) that appeared to be partially due to residual gastric pepsin activity. Adding 1200 U of pepsin per milliliter enhanced cobalamin transfer to saliva or gastric juice R binders (39%-58% transfer). At no time was cobalamin transferred directly to intrinsic factor; R binder-deficient gastric juice failed to accept cobalamin from egg yolk. The transfer of cobalamin from egg yolk to human R binder requires both an acid pH and pepsin activity. While as little as 30 U of pepsin added per milliliter of saliva promoted transfer of cobalamin, the requirement for an acid pH was very strict. Virtually no transfer occurred when pH exceeded 2.0,regardless of the amount of pepsin present. Acid provided an optimal pH for pepsin activity and, to a lesser extent, affected transfer by a mechanism unrelated to pepsin. Our data suggest that compromised pepsin secretion and, probably even more importantly, compromised acid secretion interfere with transfer of food cobalamin to R binder.
Los Angeles,
T
he release of cobalamin [vitamin B,,) from food is a fundamental step in its assimilation. However, the process by which cobalamin is released and transferred to human cobalamin-binding proteins for further transport and eventual absorption has been little studied. Nearly all of the studies of cobalamin absorption, and the clinical tests to which they have given rise, have focused on the free vitamin. The potential clinical importance of the transfer of cobalamin from food was suggested by Doscherholmen and associates (l-7) when they showed that some patients absorb free cobalamin normally, as measured in such tests as the Schilling test, but cannot absorb foodbound cobalamin. This form of malabsorption is now a well-recognized phenomenon (8-20) and can be associated with cobalamin deficiency (9-11,15,17-20). The failure to release cobalamin from food is presumably the basis of this disorder. Sullivan et al. (21)showed that liver cobalamin was not taken up as well as free cobalamin in vitro by ileal homogenate and that it first had to be transferred to intrinsic factor. Several studies using dialyzability of food-bound cobalamin or in vivo tests of absorption have noted the importance of acid and/or pepsin in releasing cobalamin from food (22-25).Kittang and Schjonsby (26)extended these studies by examining transfer to gastric binding proteins directly rather than simply measuring release from food, although some of their liver cobalamin appears to have been free rather than bound. R binder from swallowed saliva, rather than intrinsic factor (27,281, binds free cobalamin preferentially in gastric juice at low pH because R binder has a higher affinity than intrinsic factor for cobalamin (29).
0 1990 by the American Gastroenterological Association 0016-5065/90/$3.00
COBALAMIN
June 199~0
However, what happens to cobalamin bound to food is not as clear in this setting as what has been described for free cobalamin. Our
Materials and Methods Extraction
of the Egg Yolk Binder
Egg yolk cobalamin-binding protein was extracted with ether, as done by Levine and Doscherholmen (30). Standard grade AA large chicken eggs were bought in a supermarket. The raw egg yolk was carefully separated from the egg white, and about 20 ml of yolk was diluted to 100 ml with saline. This mixture was added 21 (vol/vol) to ether in a glass funnel and mixed gently for 5 min. After 60 min at room temperature, the lower cloudy portion was extracted again by the same procedure. The clear extract was diluted 1:lO with saline. Aliquots were stored at -20°C until used.
Identification
of Cobalamin-Binding
Proteins
Unsaturated cobalamin-binding proteins were labeled with 57Co-cyanocobalamin, specific activity lo-20 &i/wg (Amersham Searle, Arlington Heights, Ill.]. Chromatography was performed on Sephadex G-200 gel (Pharmacia, Piscataway, N.J.] to separate and quantitate the binders, using 0.1 M phosphate-l.0 M NaCl buffer, pH 7.4. The eluted fractions were counted for radioactivity in a welltype gamma counter. The cobalamin-binding protein peaks were converted to picograms of 57Co-cobalamin bound. Human R binder was also identified immunologically with a rabbit antiserum to human saliva, which reacts with human R binder but not with intrinsic factor or transcobalamin II. Binding (type II] antibody to human intrinsic factor was obtained from a patient with pernicious anemia. Experimental samples were incubated with either type of antibody for 60 min at 37’C. A shift of the 57Co-cobalamin-labeled binding peak to the void volume on Sephadex gel chromatography indicated a positive reaction.
Human
Digestive
Secretions
Human gastric juices had been collected in the course of earlier studies from subjects with normal cobalamin status, pernicious anemia, or R binder deficiency (31). All the specimens were collected by continuous suction after betazole or pentagastrin stimulation. Aliquots of neutralized, depepsinized (321, filtered specimens were centrifuged at 410o xg for 20 min at 4°C and stored at -20°C. Wholemouth saliva was collected on 2 separate occasions from a
TRANSFER
FROM EGG YOLK TO R BINDERS
1461
normal subject, centrifuged at 4100 xg for 20 min at 4°C. and stored in aliquots at -20°C. Pepsin activity of the secretions was measured by a modification (33) of the method of Anson and Mirsky (34).
Cobalamin Transfer From Human
Digestive
Egg Yolk Binder
to
Secretions
Studies of cobalamin transfer from the egg yolk binder to R binder and intrinsic factor used either gastric juice or saliva. One milliliter of ether-extracted, diluted yolk was partially saturated with 50 ~1 (500 pg) 57Co-cobalamin and incubated for 60 min at room temperature. This preincubation allowed nearly complete binding of the cobalamin; only 17~ of the cobalamin remained unbound, as determined by Sephadex gel chromatography; 0.5 ml gastric juice or saliva was then added to the egg yolk-57Co-cobalamin mixture. After 60 min at 37”C, a 0.2-ml aliquot of the total mixture (yolk-57Co-cobalamin-gastric juice) was incubated with 25 11 anti-human R binder antiserum or 0.1 ml anti-human intrinsic factor antiserum for 60 min at 37°C and applied to Sephadex G-200 gel columns for chromatography. The amount of 57Co-cobalamin that was no longer bound to the egg yolk cobalamin-binding peak could thus be quantitated from the amount shifted to the void volume. This transferred amount was expressed as percent of the total 57Co-cobalamin. Anti-intrinsic factor antibody was not used in the experiments involving saliva because saliva does not contain intrinsic factor. In further experiments, the same procedure was followed except that gastric juice or saliva was acidified with 5 N or 1 N HCl to pH 1.2-1.3 before it was added to the egg yolk-57Co-cobalamin. After 60 min at 37”C, the mixture was reneutralized to pH 7 + 0.5 with NaOH immediately before the antibody was added, because the immunologic reaction requires a neutral pH. In other experiments, pepsin (double extracted from hog stomach mucosa; 3060 U/mg protein; Sigma Chemical Co., St. Louis, MO.) was added to the acidified gastric juice to a final pepsin concentration of 1200 U/ml specimen, which is very close to the physiological concentration of enzyme in normal gastric juice (35). Pepsin solutions were freshly prepared for each experiment. Two controls were included in all the initial experiments. Saline replaced the egg yolk extract in one set of control incubations. In the other control incubation, saline [with and without acidification or added pepsin, as appropriate) replaced the gastric juice or saliva. These controls allowed the results for the entire mixture to be compared with the effects of each manipulation on the egg yo1.k alone or on the digestive secretion alone. Effect of Time on Cobalamin
Transfer
The effect of time was analyzed in a preliminary study using acidified gastric juice to which pepsin was added. The transfer of cobalamin from egg yolk at 37°C was tested after 20, 40, and 60 min and, in another experiment, after 10 and 60 min and 6 and 24 h. Maximal transfer occurred at 60 min; no further increase in transfer was observed after this time. Therefore, 60 min was chosen as the incubation period for all experiments.
1462
DEL CORRAL AND CARMEL
Effect of pH on Cobalamin
GASTROENTEROLOGY
Transfer
The effect of pH was tested in saliva with and without added pepsin. Saliva was acidified with 5 N or concentrated HCI, avoiding excessive dilution of the samples. The pH of the final mixture (including pepsin, egg yolk extract, and 57Co-cobalamin) was always 0.4-0.5 units higher than that of the prepared acidified saliva.
Effect of Pepsin on Cobalamin
Transfer
The effect of pepsin concentration was tested by titration of the enzyme added to saliva. Pepsin was added to achieve given concentrations of pepsin per milliliter of saliva. The saliva was acidified to pH 1.0 in order to reach pH 1.5 in the final mixture. The role of pepsin in promoting transfer was further explored with pepstatin A (J-amino-3 hydroxy-6-methylheptanoic acid; Sigma Chemical Co.], an inhibitor of pepsin activity. Concentrations of pepstatin A that inhibited >90% of the assayable pepsin activity in gastric juice were used. Pepstatin A itself did not have any direct effect on cobalamin transfer.
Results Cobalamin
Binding
in Egg Yolk
Egg yolk extract incubated with excess 57Cocobalamin bound 4750 pg cobalamin per milliliter of the diluted extract or 14 pg/mg protein. Assuming no
loss in the extraction procedure, this is equivalent to a binding capacity of 238 ng/ml in the original egg yolk. On Sephadex gel filtration, all of the bound cobalamin eluted in a single peak in the same region as human R binder. This binding peak did not cross-react with either anti-human R binder or anti-human intrinsic factor antibody. The cobalamin-binding protein of egg yolk had similar avidity for free cobalamin as human R binder.
Table 1. Transfer
~f~~Co-Cob&min
Vol. 98. No. 6
Egg yolk extract and normal saliva, each adjusted to a cobalamin-binding capacity of about 450 pg 57Cocobalamin, were mixed and exposed together to a subsaturating amount (300 pg) of 57Co-cobalamin. In two experiments, 52%-54% of the cobalamin bound to the salivary R binder, as determined by the bound 57Co-cobalamin that was shifted to the void volume upon chromatography after exposure to anti-human R binder antibody. Scatchard plot analysis showed that egg yolk binder and salivary R binder had similar binding affinity constants, 0.4-0.6 x 10” L . M-’ and 0.1-0.2 x 10” L - M-l, respectively, in three separate determinations for each binder at room temperature and pH 7.4.
Cobalamin Transfer From Egg Yolk to Human Secretions Table 1 summarizes the effect of acid and pepsin on saliva and gastric juice specimens. No transfer of 57Co-cobalamin from egg binder to either saliva or gastric juice occurred at a neutral pH. The 57Co-cobalamin remained attached to egg yolk binder, which did not react with either anti-human R binder or anti-intrinsic factor antibody. When the saliva and gastric juice specimens were acidified to pH 1.5 before incubation with the egg yolk-57Co-cobalamin, transfer increased slightly (Table 11.All of the transfer was to R binder in the gastric juice, and not to intrinsic factor. The effect of acid alone was studied further by measuring transfer of cobalamin at various adjusted pH levels to normal saliva, which contains little or no pepsin activity (Table 2). A decrease in transfer of cobalamin from egg yolk was apparent when the pH of the saliva rose above 1.5 (equivalent to a pH of 2.0 in the final mixture).
From Egg Yolk Extract to Saliva and Gastric \uice % Transfer of 57Co-cobalamin from egg yolk Pepsin content
Specimen
Patient source
PH
Saliva A Saliva B Gastric juice Gastric juice Gastric juice Gastric juice
Normal volunteer’ Normal volunteer” “Normal” subjectb “Normal” subjectb Pernicious anemia R binder deficiency
8.9
1 2 3 4
(4.1)" (1.4p (7.5)
(U/ml] 6
Baseline (neutral pH) 0
-
0
8gd 95d 18d 34"
0 0 0
Acidification (PH 1.5) 9;7;10 8;12 25;25 40 12 0
Pepsin (1200 U/ml) (PH 1.51 43;49;58 45;46;49 48;39;41 45 50 0
“Saliva specimens A and B were obtained on different occasions from the same normal subject. bThe subjects were patients with normal serum cobalamin levels, normal intrinsic factor secretion, and normal Schilling test results. “The actual pH of the gastric juice used in the baseline study was 7 because all specimens had been depepsinized by brining the pH to 10 immediately after collection, neutralized after 20 min, and stored at -2O’C. The value given in parentheses is the original pH before depepsinization and neutralization.dThe pepsin levels in the gastric juices represent residual pepsin activity after depepsinization.
June 1990
COBALAMIN
TRANSFER
The Role of Pepsin in Cobalamin Transfer From Egg Yolk
1.0
Table 2. Effect of pH on Transfer of 57Co-Cobalamin From Egg Yolk Extract to Normal Saliva R Binder
0.5
Experiment Experilment
1 2
of %o-cobalamin 1.0
1.2
to saliva 1.5
EGG YOLK TO R BINDERS
pH of total
The addition of pepsin greatly enhanced the transfer of 57Co-cobalamin from egg yolk (Table 1). Transfer to the gastric juice specimens, including one from a patient with pernicious anemia, and to saliva was high when pepsin was added and did not differ significantly among these specimens. In all cases, the transfer was to R binder rather than to intrinsic factor. It is noteworthy that transfer to R binder-deficient gastric juice never occurred despite acidification and pepsinization (Table 1).This gastric juice’s intrinsic factor, although unsaturated with cobalamin (unsaturated ciobalamin-binding capacity was 19.3 ng/ml), did not accept any 57Co-cobalamin from egg yolk under these conditions. Pepsin was effective in promoting transfer of cobalamin at an acid pH only. When the pepsinized, acidified saliva B was reneutralized to pH 7 before it was added to the egg yolk-57Co-cobalamin, transfer decreased to 0% despite the high concentrations of pepsin. Figure 1 illustrates the dependence of pepsin effect on an acid pH. Transfer was markedly diminished when the pH of saliva reached 1.5-2.0, and no transfler occurred above pH 2 despite the presence of 1200 IJ pepsin per milliliter. The effect of pepsin was further explored. Saliva was clnosen for these studies because we showed that it, unlike gastric juice, contains virtually no endogenous pepsin. Although the gastric juice specimens had all belen previously depepsinized by alkalinization to pH 10 immediately after being collected, such depepsinization was never complete. For example, the depepsinized pernicious anemia gastric juice contained 18 U of residual pepsin per milliliter, while the other gastric juices still contained 31-95 U/ml. Only a small amount of pepsin was necessary for the promotion of transfer. Concentrations between 600 and 4800 U of added pepsin per milliliter of saliva produced similar degrees of transfer, and only a slight diminution was seen at lower pepsin concentrations (Table 3). Indeed, 42% transfer was evident even with as little as 30 U of added pepsin per milliliter.
% Transfer
FROM
A at pH”
2.0
2.5
14%
8%
12%
4%
0%
0%
8%
8%
-
9%
4%
2%
“The p!H is the pH to which the saliva specimen had been adjusted before incubation with egg yolk-57Co-cobalamin. The final pH of the total mixture of egg yolk-57Co-cobalamin and saliva was 0.4-0.5 pH units higher.
60
;
I
1.5
I
2.0
I
1463
mixture 2.5
I
3.0
1
3.5
I
50-& Xl 2
40--
W 2 ; % 6 2 *
30-20-lO-0
I
I
I
0.5
1.0
1.5
I
I
I
2.0
2.5
3.0
_I
pH of saliva
Figure 1. Dependence on acid of pepsin-mediated transfer of “Co-cobalamin from egg yolk to human saliva R binder. In these 3 experiments, using saliva A and B, the concentration of pepsin was 1200 U/ml saliva. The points and bars represent the mean t 1 SD; no SD is shown for means of 2 rather than 3 experiments. Both saliva and total mixture pH values are given in the abscissa.
Finally, experiments using pepstatin A were performed to determine whether the acid pH had any effect on cobalamin transfer other than by its activation of pepsin. As expected, pepstatin A significantly inhibited pepsin-mediated cobalamin transfer to saliva or gastric juice R binder at pH 1.5 (Table 4). However, pepstatin A did not reduce the small amount of transfer seen at acid pH when no pepsin had been added. Transfer was unchanged by pepstatin A in 2 of these 3 experiments at low pH, when saliva and pernicious anemia gastric juice were used; both of these specimens contained very little endogenous pepsin. In the third experiment, pepstatin A mildly decreased transfer from 18% to 8%; the gastric juice used had an endogenous pepsin content of 89 U/ml. Consistent with all the above results, pepstatin A never reduced even pepsin-mediated transfer in the presence of acid below this 8% level. Effect of Acid and Pepsin on Human Cobalamin-Binding Proteins The ability of human R binders to bind 57Cocobalamin was unaffected by acidification or by pepsin. Gastric juice R binder cobalamin-binding capacity was unchanged under such conditions. However, the affinity of gastric intrinsic factor for cobalamin was impaired at pH 1.5. Although intrinsic factor could bind cobalamin at low pH when 57Co-cobalamin was added to it in excess amounts (>lOO% of the gastric juice’s binding capacity), it failed to compete with R binder when subsaturating amounts (~30%) of 57Co-
1464
DEL CORRAL
Table 3.
GASTROENTEROLOGY
AND CARMEL
Titration of Pepsin Effect on Cobalamin Transfer From Egg Yolk Binder to Saliva R Binder % Transfer 4800
Experiment Experiment
Vol. 98. No. 6
1 2
49% 55%
2400 55% -
1200
of 57Co-cobalamin 600
49%
49% 50%
cobalamin were added. To a large extent, these findings explain the failure to demonstrate transfer of cobalamin to intrinsic factor at any time, and especially the previously mentioned finding in the R binder-deficient gastric juice. Egg yolk did not release its cobalamin at all at normal pH, whereas intrinsic factor was inactive at low pH. Discussion Egg yolk contains a saturable binding protein for cobalamin that has the same molecular size on Sephadex gel chromatography as the binder found in chicken serum, confirming the findings of Levine and Doscherholmen (30). Presumably this protein is equivalent to human R binder, although this has never been established. As shown here, it does not cross-react immunologically with human R binder. The function of this egg protein is unknown, but its ability to bind cobalamin presumably explains the lesser absorption of egg-bound cobalamin than of free cobalamin (30). Our data show that its avidity for cobalamin is similar to that of human saliva R binder. The experiments also demonstrate that egg yolk binder holds cobalamin and does not release it to human binders present in digestive secretions at neutral pH. It must be remembered that the conditions of our in vitro experiments may not represent the in vivo situation completely. We studied a lipid-free preparation of egg yolk and did not study it in a cooked form because the experiments required an aqueous phase.
Table 4. Effect of Pepstatin A on Acid- and Pepsin-Mediated Transfer of Cobalamin From Egg Yolk Binder to R Binder” 70 Transfer of 57Co-cobalamin from egg yolk Baseline
+ Pepstatin A
13 2 5"
10 + 3b
42 + 10”
9 * 2c
Acidification (pH 1.5); no added pepsin Acidification (pH 1.5) + added pepsin (600 U/ml]
“Three experiments were done, measuring transfer from egg yolk to gastric juice I, and pernicious R binder of saliva B, “normal” anemia gastric juice 3 (see Table 1). bThe difference between baseline transfer and transfer in the presence of pepstatin A was not significant. “The difference between baseline transfer and transfer in the presence of pepstatin A was significant [p < O.Ol].
at pepsin 300
46% -
concentration 180 48%
(U/ml
saliva) of:
150
90
30
41% -
46%
42%
0 10'70 9%
The 57Co-cobalamin was in cyanocobalamin form and was added exogenously, and therefore may not be fully equivalent to endogenous cobalamins. Furthermore, results with egg yolk may not be equally true for other food sources of cobalamin. Nevertheless, the model provides a starting point for analyzing the fate of food-bound cobalamin. Moreover, it reflects a form of cobalamin commonly used in testing absorption of food cobalamin. The exposure of egg yolk-bound cobalamin to saliva and gastric juice was examined next. The transfer of cobalamin to human cobalamin-binding proteins was found to depend on acid. The pH had to be t2 for transfer to occur. The addition of pepsin greatly increased the amount of cobalamin transferred. This increase can be seen most clearly in the experiments using saliva rather than gastric juice, which was never entirely free of pepsin activity. Saliva is also a more physiological target because it is the first digestive secretion to which food is normally exposed. Moreover, saliva rather than the stomach is the source of secretory R binder (36). Our data suggest that salivary enzymes do not promote the transfer of food cobalamin to salivary R binder, at least at neutral pH. Although an acid pH is obviously a necessary condition, primarily because optimal pepsin activity requires a low pH, it is not sufficient in itself for optimal transfer of food-bound cobalamin. Nevertheless, the minimal effect of acid on transfer of cobalamin to saliva, where barely measurable pepsin levels were present, suggests that acid pH also has a slight, possibly direct activity promoting such transfer that is independent of pepsin. The failure of added pepstatin A to inhibit such transfer supports this conclusion. Cobalamin from egg yolk was always transferred to R binder rather than to intrinsic factor. Our observations confirm previous findings that intrinsic factor has a much lower affinity than R binder for cobalamin at a low pH (29). We have further shown that pepsin does nothing to change this equilibrium between intrinsic factor and R binder. The transfer of cobalamin from food to R binder has important implications because many patients may have food cobalamin malabsorption despite normal intrinsic factor secretion (l-20). Our data support previous suggestions that the lack of acid and pepsin may be a major factor in such cases. However, loss of
June 1990
COBALAMIN TRANSFER FROM EGG YOLK TO R BINDERS
pepsin seems to be important only if it is virtually complete. As shown in our experiments, as little as 30 U of exogenously added pepsin per milliliter was adequate to promote transfer of cobalamin from egg yolk when the pH was 1.5. This concentration is only a small fraction of the normal pepsin content of gastric juice (35). The data in Table 1 suggest that endogenous gastric juice pepsin was slightly less active than pepsin added to saliva, because acidified gastric juices containing 89-95 U pepsin per milliliter promoted slightly less transfer than comparable added pepsin concentrations. The explanation for this minor discrepancy is not clear. In contrast to the relatively small pepsin requirement, even a mild impairment of acid secretion sufficed to almost completely inhibit transfer of cobalamin from egg yolk. No transfer was evident at pH >2. Thus, even mild hypochlorhydria should prove inhibitory to normal absorption of food cobalamin, regardless 0.f gastric pepsin content. The following sequence of events can be proposed for the initial phases of cobalamin assimilation in the normal individual. Food mixes with saliva in the mouth and enters the stomach without having released any of its cobalamin. The low pH and the pepsin encountered in the stomach then release the cobal,amin from its binders in food. This cobalamin becomes bound to the salivary R binders that had also enterled the stomach. The possible contribution of other gastric, duodenal, and biliary enzymes to this process of transfer remains to be defined. It is possible that food cobalamin malabsorption may not always be attributable solely to gastric dysfunction (1820). Full clarification of all potential mechanisms of food cobalamin malabsorption vvill require further investigation. As postulated by Okuda et al. (371 and demonstrated by A.llen et al. (29) and subsequently others (38-411, after cobalamin has bound to R binder, pancreatic enzymes degrade the R binder. This event sets the stage for the later binding of the released cobalamin by intrinsic factor in the upper small intestine. Some animal studies have also suggested a minor but direct absorptive role for pancreatic secretions (42,431, but intrinsic factor is the essential transport protein that is responsible for the eventual absorption of cobalamin across the ileal epithelium.
References 1. Doscherholmen
A, Swaim WR. Impaired assimilation of egg Co5’vitamin B,z in patients with hypochlorhydria and achlorhydria and after gastric resection. Gastroenterology 1973;64:913919. 2. Doscherholmen A, McMahon J, Ripley D. Vitamin B,, absorption from eggs. Proc Sot Exp Biol Med 1975;149:987-990. 3. Doscherholmen A, McMahon J, Ripley D. Inhibitory effect of
1465
eggs on vitamin B,, absorption: description of a simple ovalbumin 57CoB,, vitamin B,, absorption test. Br ] Haematol 1976;33: 261-272.
4. Doscherholmen A, McMahon J, Ripley D. Vitamin B,, assimilation from chicken meat. Am J Clin Nutr 1978;31:825-830. A, McMahon J, Economon P. Vitamin B,, 5. Doscherholmen absorption from fish. Proc Sot Exp Biol Med 1981;167:480-484. 6. Salom IL, Silvis SE, Doscherholmen A. Effect of cimetidine on the absorption of vitamin B,,. Stand J Gastroenterol1982:17:129131.
7. Doscherholmen A, Silvis S. McMahon J. Dual isotope Schilling test for measuring absorption of food-bound and free vitamin B,, simultaneously. Am J Clin Path01 1983;80:490-495. 8. Streeter AM, Duraiappah B, Boyle R, O’Neill BT, Pheils MT. Malabsorption of vitamin B,, after vagotomy. Am J Surg 1974:128: 340-343.
9. Streeter AM, Shum HY, Duncombe VM, Hewson JW, Thorpe MEC. Vitamin B,, malabsorption associated with a normal Schilling test result. Med J Austr 1976;1:54-55. 10. Carmel R. Nutritional vitamin B,, deficiency. Possible contributory role of subtle vitamin B,, malabsorption. Ann Intern Med 1978;88:647-649. 11. King CE, Leibach J, Toskes PP. Clinically significant vitamin Bj2 deficiency secondary to malabsorption of protein-bound vitamin B,Z.Dig Dis Sci 1979;24:397-402. 12. Steinberg WM. King CE, Toskes PP. Malabsorption of proteinbound cobalamin but not unbound cobalamin during cimetidine administration. Dig Dis Sci 1980;25:188-192. 13. Cattan D, Belaiche J, Zittoun J, Yvart J, Chagnon JP, Nurit Y. Role de la carence en facteur intrinseque dans la malabsorption de la vitamine B,, liee aux proteines, dans les achlorhydries. Gastroenterol Clin Biol 1982;6:570-575. 14. Streeter AM, Goulston KJ, Bathur FA, Hilmer RS, Crane GG, Pheils MT. Cimetidine and malabsorption of cobalamin. Dig Dis Sci 1982:27:13-16. 15. Dawson DW, Sawers AH, Sharma RK. Malabsorption of protein bound vitamin B,,. Br Med J 1984;288:675-678. 16. Hamborg B, Kittang E, Schjonsby H. The effect of ranitidine on the absorption of food cobalamins. Stand J Gastroenterol 1985; 20:756-758. 17. Jones BP, Broomhead AF, Kwan YL, Grace CS: Incidence and clinical significance of protein-bound vitamin B,, malabsorption. Eur J Haematoll987:38:131-136. 18. Carmel R, Sinow RM. Karnaze D. Atypical cobalamin deficiency. Subtle biochemical evidence of deficiency is commonly demonstrable in patients without megaloblastic anemia and is often associated with bound cobalamin malabsorption. J Lab Clin Med 1987;109:454-463. 19. Gozzard DI, Dawson DW, Lewis MJ. Experiences with dual protein-bound aqueous vitamin B,, absorption test in subjects with low serum vitamin B,, concentrations. J Clin Path01 1987:40:633-637. 20. Carmel R, Sinow RM, Siegel ME, Samloff IM. Food cobalamin malabsorption occurs frequently in patients with unexplained low serum cobalamin levels. Arch Intern Med 1988:148:17151719. 21. Sullivan LW. Herbert V, Reizenstein P. Evidence against preferential intestinal absorption of physiologic quantities of liver-bound vitamin B,, by patients with pernicious anemia. Am J Clin Nutr 1962;11:568-573. 22. Reizenstein PG. Effect of digestive enzymes on bound vitamin B,,. Acta Med Stand 1959;165:481-486. 23. Cooper BA, Castle WB. Sequential mechanisms in the enhanced absorption of vitamin B,, by intrinsic factor in the rat. J Clin Invest 1960;39:199-214. 24. Schade SC, Schilling RF. Effect of pepsin on the absorption of food vitamin B,, and iron. Am J Clin Nutr 1967;20:636-640,
1466 DEL CORRAL AND CARMEL
25. Adams JF, Kennedy EH. Thomson J, Williamson J. The effect of acid peptic digestion on free and tissue-bound cobalamins. Br J Nutr 1968;22:111-114. 26. Kittang E, Schjonsby H. Effect of gastric anacidity on the release of cobalamins from food and their subsequent binding to R protein. Stand J Gastroenterol1987;22:1031-1037. 27. Goldberg LS, Fudenberg HH. Effect of pH on the vitamin B,,-binding capacity of intrinsic factor. J Lab Clin Med 1969;73: 469-475. 28. Shum HY, O’Neill BJ, Streeter AM. Effect of pH changes on the binding of vitamin B12 by intrinsic factor. J Clin Path01 1971:24: 239-243. 29. Allen RH, Seetharam B, Podell E, Alpers DH: Effect of proteolytic enzymes on the binding of cobalamin to R protein and intrinsic factor. In vitro evidence that a failure to partially degrade R protein is responsible for cobalamin malabsorption in pancreatic insufficiency. J Clin Invest 1978;61:47-54. 30. Levine AS, Doscherholmen A. Vitamin B,, bioavailability from egg yolk and egg white: relationship to binding proteins. Am J Clin Nutr 1983;38:436-439. 31. Carmel R, Herbert V. Deficiency of vitamin B,,-binding alpha globulin in two brothers. Blood 1969;33:1-12. 32. Flood C, West R. Some properties of Castle’s intrinsic factor. Proc Sot Exp Biol Med 1936;34:542-543. 33. Skak-Nielsen T, Holst JJ, Nielsen OV. Role of gastrin-releasing peptide in the neural control of pepsinogen secretion from the pig stomach. Gastroenterology 1988;95:1216-1220. 34. Anson ML, Mirsky AE. The estimation of pepsin with hemoglobin. J Gen Physiol1932;16:59-63. 35. Hirschowitz BI. Electrolytes in human gastric secretions. Observations and theory. Am J Dig Dis 1961;6:199-228. 36. Kudo H, Inada M, Ohshio G, Wakatsuki Y, Ogawa K, Hamashima Y, Miyake T. Immunohistochemical localisation of vitamin B,, R-binder in the human digestive tract. Gut 1987;28:339-345. 37. Okuda K, Kitazaki T, Takamatsu M. Inactivation of vitamin B,,
GASTROENTEROLOGY
38.
39.
40.
41.
42.
43.
Vol. 98, No. 6
by a binder in rat intestine and the role of intrinsic factor. Digestion 1971;4:35-48. Andersen KJ, von der Lippe G. The effect of proteolytic enzymes on the vitamin B,,-binding proteins of human gastric juice and saliva. Stand J Gastroenterol1979;14:833-838. Marcoullis G. Parmentier Y, Nicolas JP, Jimenez M, Gerard P. Cobalamin malabsorption due to nondegradation of R proteins in the human intestine. J Clin Invest 1980;66:430-440. Carmel R, Abramson SB, Renner IG. Characterization of pure human pancreatic juice: cobalamin content, cobalamin-binding proteins and activity against human R binders of various secretions. Clin Sci 1983;64:193-205. Belaiche J. Zittoun J. Marquet J, Yvart J, Cattan D. In vitro effect of duodenal juice on R binders cobalamin complexes in subjects with pancreatic insufficiency: correlation with cobalamin absorption Gut 1987;28:70-74. Okuda K. Enhanced vitamin B,, absorption from rat intestine by proteases in the absence of intrinsic factor. Proc Sot Exp Biol Med 1967;124:79-83. Carmel R. Hollander D, Gergely HM, Renner IG, Abramson SB. Pure human pancreatic juice directly enhances uptake of cobalamin by guinea pig ileum in vivo. Proc Sot Exp Biol Med 1985:178:143-150.
Received May 23,1989. Accepted November 16.1989. Address requests for reprints to: Ralph Carmel, M.D., RMR 306, USC School of Medicine, 2025 Zonal Avenue, Los Angeles, California 90033. This study was supported by grant DK-32640 from the National Institutes of Health. Dr. de1 Corral was supported by a scholarship from Consorzio per 1’LJniversita degli Studi, Bari, Italy. The authors thank Dr. Sandra Schiffman for helpful suggestions, Rosemarie Nimo and Che Hung Yeh for technical assistance, and Ruth Ballard for manuscript preparation.
1990;98:1460-1466
Transfer of Cobalamin From the Cobalamin-Binding Protein of Egg Yolk to R Binder of Human Saliva and Gastric Juice ALBERT0 Department California
DEL CORRAL of Medicine,
and RALPH CARMEL
University of Southern California School of Medicine,
Patients may fail to absorb cobalamin (vitamin B,,) bound to food even when they have adequate intrinsic factor to absorb free cobalamin normally. We studied cobalamin transfer from egg yolk cobalaminbinding protein to human saliva and gastric juice as a model of this important first step in cobalamin assimilation. The cobalamin-binding protein of egg yolk eluted with human R binder on Sephadex gel chromatography and bound cobalamin with a comparable affinity, but it did not cross-react with R binder immunologically. Transfer of cobalamin from egg yolk to saliva or gastric juice R binder did not occur at neutral pH. Slight transfer (8%-Z% of the 57Cocobalamin bound to egg yolk) occurred when the saliva was acidified to pH 1.5. This minor transfer by acid was not inhibited by pepstatin A, a pepsin inhibitor. Acidification caused variable transfer to gastric juice R binder (12%-40%) that appeared to be partially due to residual gastric pepsin activity. Adding 1200 U of pepsin per milliliter enhanced cobalamin transfer to saliva or gastric juice R binders (39%-58% transfer). At no time was cobalamin transferred directly to intrinsic factor; R binder-deficient gastric juice failed to accept cobalamin from egg yolk. The transfer of cobalamin from egg yolk to human R binder requires both an acid pH and pepsin activity. While as little as 30 U of pepsin added per milliliter of saliva promoted transfer of cobalamin, the requirement for an acid pH was very strict. Virtually no transfer occurred when pH exceeded 2.0,regardless of the amount of pepsin present. Acid provided an optimal pH for pepsin activity and, to a lesser extent, affected transfer by a mechanism unrelated to pepsin. Our data suggest that compromised pepsin secretion and, probably even more importantly, compromised acid secretion interfere with transfer of food cobalamin to R binder.
Los Angeles,
T
he release of cobalamin [vitamin B,,) from food is a fundamental step in its assimilation. However, the process by which cobalamin is released and transferred to human cobalamin-binding proteins for further transport and eventual absorption has been little studied. Nearly all of the studies of cobalamin absorption, and the clinical tests to which they have given rise, have focused on the free vitamin. The potential clinical importance of the transfer of cobalamin from food was suggested by Doscherholmen and associates (l-7) when they showed that some patients absorb free cobalamin normally, as measured in such tests as the Schilling test, but cannot absorb foodbound cobalamin. This form of malabsorption is now a well-recognized phenomenon (8-20) and can be associated with cobalamin deficiency (9-11,15,17-20). The failure to release cobalamin from food is presumably the basis of this disorder. Sullivan et al. (21)showed that liver cobalamin was not taken up as well as free cobalamin in vitro by ileal homogenate and that it first had to be transferred to intrinsic factor. Several studies using dialyzability of food-bound cobalamin or in vivo tests of absorption have noted the importance of acid and/or pepsin in releasing cobalamin from food (22-25).Kittang and Schjonsby (26)extended these studies by examining transfer to gastric binding proteins directly rather than simply measuring release from food, although some of their liver cobalamin appears to have been free rather than bound. R binder from swallowed saliva, rather than intrinsic factor (27,281, binds free cobalamin preferentially in gastric juice at low pH because R binder has a higher affinity than intrinsic factor for cobalamin (29).
0 1990 by the American Gastroenterological Association 0016-5065/90/$3.00
COBALAMIN
June 199~0
However, what happens to cobalamin bound to food is not as clear in this setting as what has been described for free cobalamin. Our
Materials and Methods Extraction
of the Egg Yolk Binder
Egg yolk cobalamin-binding protein was extracted with ether, as done by Levine and Doscherholmen (30). Standard grade AA large chicken eggs were bought in a supermarket. The raw egg yolk was carefully separated from the egg white, and about 20 ml of yolk was diluted to 100 ml with saline. This mixture was added 21 (vol/vol) to ether in a glass funnel and mixed gently for 5 min. After 60 min at room temperature, the lower cloudy portion was extracted again by the same procedure. The clear extract was diluted 1:lO with saline. Aliquots were stored at -20°C until used.
Identification
of Cobalamin-Binding
Proteins
Unsaturated cobalamin-binding proteins were labeled with 57Co-cyanocobalamin, specific activity lo-20 &i/wg (Amersham Searle, Arlington Heights, Ill.]. Chromatography was performed on Sephadex G-200 gel (Pharmacia, Piscataway, N.J.] to separate and quantitate the binders, using 0.1 M phosphate-l.0 M NaCl buffer, pH 7.4. The eluted fractions were counted for radioactivity in a welltype gamma counter. The cobalamin-binding protein peaks were converted to picograms of 57Co-cobalamin bound. Human R binder was also identified immunologically with a rabbit antiserum to human saliva, which reacts with human R binder but not with intrinsic factor or transcobalamin II. Binding (type II] antibody to human intrinsic factor was obtained from a patient with pernicious anemia. Experimental samples were incubated with either type of antibody for 60 min at 37’C. A shift of the 57Co-cobalamin-labeled binding peak to the void volume on Sephadex gel chromatography indicated a positive reaction.
Human
Digestive
Secretions
Human gastric juices had been collected in the course of earlier studies from subjects with normal cobalamin status, pernicious anemia, or R binder deficiency (31). All the specimens were collected by continuous suction after betazole or pentagastrin stimulation. Aliquots of neutralized, depepsinized (321, filtered specimens were centrifuged at 410o xg for 20 min at 4°C and stored at -20°C. Wholemouth saliva was collected on 2 separate occasions from a
TRANSFER
FROM EGG YOLK TO R BINDERS
1461
normal subject, centrifuged at 4100 xg for 20 min at 4°C. and stored in aliquots at -20°C. Pepsin activity of the secretions was measured by a modification (33) of the method of Anson and Mirsky (34).
Cobalamin Transfer From Human
Digestive
Egg Yolk Binder
to
Secretions
Studies of cobalamin transfer from the egg yolk binder to R binder and intrinsic factor used either gastric juice or saliva. One milliliter of ether-extracted, diluted yolk was partially saturated with 50 ~1 (500 pg) 57Co-cobalamin and incubated for 60 min at room temperature. This preincubation allowed nearly complete binding of the cobalamin; only 17~ of the cobalamin remained unbound, as determined by Sephadex gel chromatography; 0.5 ml gastric juice or saliva was then added to the egg yolk-57Co-cobalamin mixture. After 60 min at 37”C, a 0.2-ml aliquot of the total mixture (yolk-57Co-cobalamin-gastric juice) was incubated with 25 11 anti-human R binder antiserum or 0.1 ml anti-human intrinsic factor antiserum for 60 min at 37°C and applied to Sephadex G-200 gel columns for chromatography. The amount of 57Co-cobalamin that was no longer bound to the egg yolk cobalamin-binding peak could thus be quantitated from the amount shifted to the void volume. This transferred amount was expressed as percent of the total 57Co-cobalamin. Anti-intrinsic factor antibody was not used in the experiments involving saliva because saliva does not contain intrinsic factor. In further experiments, the same procedure was followed except that gastric juice or saliva was acidified with 5 N or 1 N HCl to pH 1.2-1.3 before it was added to the egg yolk-57Co-cobalamin. After 60 min at 37”C, the mixture was reneutralized to pH 7 + 0.5 with NaOH immediately before the antibody was added, because the immunologic reaction requires a neutral pH. In other experiments, pepsin (double extracted from hog stomach mucosa; 3060 U/mg protein; Sigma Chemical Co., St. Louis, MO.) was added to the acidified gastric juice to a final pepsin concentration of 1200 U/ml specimen, which is very close to the physiological concentration of enzyme in normal gastric juice (35). Pepsin solutions were freshly prepared for each experiment. Two controls were included in all the initial experiments. Saline replaced the egg yolk extract in one set of control incubations. In the other control incubation, saline [with and without acidification or added pepsin, as appropriate) replaced the gastric juice or saliva. These controls allowed the results for the entire mixture to be compared with the effects of each manipulation on the egg yo1.k alone or on the digestive secretion alone. Effect of Time on Cobalamin
Transfer
The effect of time was analyzed in a preliminary study using acidified gastric juice to which pepsin was added. The transfer of cobalamin from egg yolk at 37°C was tested after 20, 40, and 60 min and, in another experiment, after 10 and 60 min and 6 and 24 h. Maximal transfer occurred at 60 min; no further increase in transfer was observed after this time. Therefore, 60 min was chosen as the incubation period for all experiments.
1462
DEL CORRAL AND CARMEL
Effect of pH on Cobalamin
GASTROENTEROLOGY
Transfer
The effect of pH was tested in saliva with and without added pepsin. Saliva was acidified with 5 N or concentrated HCI, avoiding excessive dilution of the samples. The pH of the final mixture (including pepsin, egg yolk extract, and 57Co-cobalamin) was always 0.4-0.5 units higher than that of the prepared acidified saliva.
Effect of Pepsin on Cobalamin
Transfer
The effect of pepsin concentration was tested by titration of the enzyme added to saliva. Pepsin was added to achieve given concentrations of pepsin per milliliter of saliva. The saliva was acidified to pH 1.0 in order to reach pH 1.5 in the final mixture. The role of pepsin in promoting transfer was further explored with pepstatin A (J-amino-3 hydroxy-6-methylheptanoic acid; Sigma Chemical Co.], an inhibitor of pepsin activity. Concentrations of pepstatin A that inhibited >90% of the assayable pepsin activity in gastric juice were used. Pepstatin A itself did not have any direct effect on cobalamin transfer.
Results Cobalamin
Binding
in Egg Yolk
Egg yolk extract incubated with excess 57Cocobalamin bound 4750 pg cobalamin per milliliter of the diluted extract or 14 pg/mg protein. Assuming no
loss in the extraction procedure, this is equivalent to a binding capacity of 238 ng/ml in the original egg yolk. On Sephadex gel filtration, all of the bound cobalamin eluted in a single peak in the same region as human R binder. This binding peak did not cross-react with either anti-human R binder or anti-human intrinsic factor antibody. The cobalamin-binding protein of egg yolk had similar avidity for free cobalamin as human R binder.
Table 1. Transfer
~f~~Co-Cob&min
Vol. 98. No. 6
Egg yolk extract and normal saliva, each adjusted to a cobalamin-binding capacity of about 450 pg 57Cocobalamin, were mixed and exposed together to a subsaturating amount (300 pg) of 57Co-cobalamin. In two experiments, 52%-54% of the cobalamin bound to the salivary R binder, as determined by the bound 57Co-cobalamin that was shifted to the void volume upon chromatography after exposure to anti-human R binder antibody. Scatchard plot analysis showed that egg yolk binder and salivary R binder had similar binding affinity constants, 0.4-0.6 x 10” L . M-’ and 0.1-0.2 x 10” L - M-l, respectively, in three separate determinations for each binder at room temperature and pH 7.4.
Cobalamin Transfer From Egg Yolk to Human Secretions Table 1 summarizes the effect of acid and pepsin on saliva and gastric juice specimens. No transfer of 57Co-cobalamin from egg binder to either saliva or gastric juice occurred at a neutral pH. The 57Co-cobalamin remained attached to egg yolk binder, which did not react with either anti-human R binder or anti-intrinsic factor antibody. When the saliva and gastric juice specimens were acidified to pH 1.5 before incubation with the egg yolk-57Co-cobalamin, transfer increased slightly (Table 11.All of the transfer was to R binder in the gastric juice, and not to intrinsic factor. The effect of acid alone was studied further by measuring transfer of cobalamin at various adjusted pH levels to normal saliva, which contains little or no pepsin activity (Table 2). A decrease in transfer of cobalamin from egg yolk was apparent when the pH of the saliva rose above 1.5 (equivalent to a pH of 2.0 in the final mixture).
From Egg Yolk Extract to Saliva and Gastric \uice % Transfer of 57Co-cobalamin from egg yolk Pepsin content
Specimen
Patient source
PH
Saliva A Saliva B Gastric juice Gastric juice Gastric juice Gastric juice
Normal volunteer’ Normal volunteer” “Normal” subjectb “Normal” subjectb Pernicious anemia R binder deficiency
8.9
1 2 3 4
(4.1)" (1.4p (7.5)
(U/ml] 6
Baseline (neutral pH) 0
-
0
8gd 95d 18d 34"
0 0 0
Acidification (PH 1.5) 9;7;10 8;12 25;25 40 12 0
Pepsin (1200 U/ml) (PH 1.51 43;49;58 45;46;49 48;39;41 45 50 0
“Saliva specimens A and B were obtained on different occasions from the same normal subject. bThe subjects were patients with normal serum cobalamin levels, normal intrinsic factor secretion, and normal Schilling test results. “The actual pH of the gastric juice used in the baseline study was 7 because all specimens had been depepsinized by brining the pH to 10 immediately after collection, neutralized after 20 min, and stored at -2O’C. The value given in parentheses is the original pH before depepsinization and neutralization.dThe pepsin levels in the gastric juices represent residual pepsin activity after depepsinization.
June 1990
COBALAMIN
TRANSFER
The Role of Pepsin in Cobalamin Transfer From Egg Yolk
1.0
Table 2. Effect of pH on Transfer of 57Co-Cobalamin From Egg Yolk Extract to Normal Saliva R Binder
0.5
Experiment Experilment
1 2
of %o-cobalamin 1.0
1.2
to saliva 1.5
EGG YOLK TO R BINDERS
pH of total
The addition of pepsin greatly enhanced the transfer of 57Co-cobalamin from egg yolk (Table 1). Transfer to the gastric juice specimens, including one from a patient with pernicious anemia, and to saliva was high when pepsin was added and did not differ significantly among these specimens. In all cases, the transfer was to R binder rather than to intrinsic factor. It is noteworthy that transfer to R binder-deficient gastric juice never occurred despite acidification and pepsinization (Table 1).This gastric juice’s intrinsic factor, although unsaturated with cobalamin (unsaturated ciobalamin-binding capacity was 19.3 ng/ml), did not accept any 57Co-cobalamin from egg yolk under these conditions. Pepsin was effective in promoting transfer of cobalamin at an acid pH only. When the pepsinized, acidified saliva B was reneutralized to pH 7 before it was added to the egg yolk-57Co-cobalamin, transfer decreased to 0% despite the high concentrations of pepsin. Figure 1 illustrates the dependence of pepsin effect on an acid pH. Transfer was markedly diminished when the pH of saliva reached 1.5-2.0, and no transfler occurred above pH 2 despite the presence of 1200 IJ pepsin per milliliter. The effect of pepsin was further explored. Saliva was clnosen for these studies because we showed that it, unlike gastric juice, contains virtually no endogenous pepsin. Although the gastric juice specimens had all belen previously depepsinized by alkalinization to pH 10 immediately after being collected, such depepsinization was never complete. For example, the depepsinized pernicious anemia gastric juice contained 18 U of residual pepsin per milliliter, while the other gastric juices still contained 31-95 U/ml. Only a small amount of pepsin was necessary for the promotion of transfer. Concentrations between 600 and 4800 U of added pepsin per milliliter of saliva produced similar degrees of transfer, and only a slight diminution was seen at lower pepsin concentrations (Table 3). Indeed, 42% transfer was evident even with as little as 30 U of added pepsin per milliliter.
% Transfer
FROM
A at pH”
2.0
2.5
14%
8%
12%
4%
0%
0%
8%
8%
-
9%
4%
2%
“The p!H is the pH to which the saliva specimen had been adjusted before incubation with egg yolk-57Co-cobalamin. The final pH of the total mixture of egg yolk-57Co-cobalamin and saliva was 0.4-0.5 pH units higher.
60
;
I
1.5
I
2.0
I
1463
mixture 2.5
I
3.0
1
3.5
I
50-& Xl 2
40--
W 2 ; % 6 2 *
30-20-lO-0
I
I
I
0.5
1.0
1.5
I
I
I
2.0
2.5
3.0
_I
pH of saliva
Figure 1. Dependence on acid of pepsin-mediated transfer of “Co-cobalamin from egg yolk to human saliva R binder. In these 3 experiments, using saliva A and B, the concentration of pepsin was 1200 U/ml saliva. The points and bars represent the mean t 1 SD; no SD is shown for means of 2 rather than 3 experiments. Both saliva and total mixture pH values are given in the abscissa.
Finally, experiments using pepstatin A were performed to determine whether the acid pH had any effect on cobalamin transfer other than by its activation of pepsin. As expected, pepstatin A significantly inhibited pepsin-mediated cobalamin transfer to saliva or gastric juice R binder at pH 1.5 (Table 4). However, pepstatin A did not reduce the small amount of transfer seen at acid pH when no pepsin had been added. Transfer was unchanged by pepstatin A in 2 of these 3 experiments at low pH, when saliva and pernicious anemia gastric juice were used; both of these specimens contained very little endogenous pepsin. In the third experiment, pepstatin A mildly decreased transfer from 18% to 8%; the gastric juice used had an endogenous pepsin content of 89 U/ml. Consistent with all the above results, pepstatin A never reduced even pepsin-mediated transfer in the presence of acid below this 8% level. Effect of Acid and Pepsin on Human Cobalamin-Binding Proteins The ability of human R binders to bind 57Cocobalamin was unaffected by acidification or by pepsin. Gastric juice R binder cobalamin-binding capacity was unchanged under such conditions. However, the affinity of gastric intrinsic factor for cobalamin was impaired at pH 1.5. Although intrinsic factor could bind cobalamin at low pH when 57Co-cobalamin was added to it in excess amounts (>lOO% of the gastric juice’s binding capacity), it failed to compete with R binder when subsaturating amounts (~30%) of 57Co-
1464
DEL CORRAL
Table 3.
GASTROENTEROLOGY
AND CARMEL
Titration of Pepsin Effect on Cobalamin Transfer From Egg Yolk Binder to Saliva R Binder % Transfer 4800
Experiment Experiment
Vol. 98. No. 6
1 2
49% 55%
2400 55% -
1200
of 57Co-cobalamin 600
49%
49% 50%
cobalamin were added. To a large extent, these findings explain the failure to demonstrate transfer of cobalamin to intrinsic factor at any time, and especially the previously mentioned finding in the R binder-deficient gastric juice. Egg yolk did not release its cobalamin at all at normal pH, whereas intrinsic factor was inactive at low pH. Discussion Egg yolk contains a saturable binding protein for cobalamin that has the same molecular size on Sephadex gel chromatography as the binder found in chicken serum, confirming the findings of Levine and Doscherholmen (30). Presumably this protein is equivalent to human R binder, although this has never been established. As shown here, it does not cross-react immunologically with human R binder. The function of this egg protein is unknown, but its ability to bind cobalamin presumably explains the lesser absorption of egg-bound cobalamin than of free cobalamin (30). Our data show that its avidity for cobalamin is similar to that of human saliva R binder. The experiments also demonstrate that egg yolk binder holds cobalamin and does not release it to human binders present in digestive secretions at neutral pH. It must be remembered that the conditions of our in vitro experiments may not represent the in vivo situation completely. We studied a lipid-free preparation of egg yolk and did not study it in a cooked form because the experiments required an aqueous phase.
Table 4. Effect of Pepstatin A on Acid- and Pepsin-Mediated Transfer of Cobalamin From Egg Yolk Binder to R Binder” 70 Transfer of 57Co-cobalamin from egg yolk Baseline
+ Pepstatin A
13 2 5"
10 + 3b
42 + 10”
9 * 2c
Acidification (pH 1.5); no added pepsin Acidification (pH 1.5) + added pepsin (600 U/ml]
“Three experiments were done, measuring transfer from egg yolk to gastric juice I, and pernicious R binder of saliva B, “normal” anemia gastric juice 3 (see Table 1). bThe difference between baseline transfer and transfer in the presence of pepstatin A was not significant. “The difference between baseline transfer and transfer in the presence of pepstatin A was significant [p < O.Ol].
at pepsin 300
46% -
concentration 180 48%
(U/ml
saliva) of:
150
90
30
41% -
46%
42%
0 10'70 9%
The 57Co-cobalamin was in cyanocobalamin form and was added exogenously, and therefore may not be fully equivalent to endogenous cobalamins. Furthermore, results with egg yolk may not be equally true for other food sources of cobalamin. Nevertheless, the model provides a starting point for analyzing the fate of food-bound cobalamin. Moreover, it reflects a form of cobalamin commonly used in testing absorption of food cobalamin. The exposure of egg yolk-bound cobalamin to saliva and gastric juice was examined next. The transfer of cobalamin to human cobalamin-binding proteins was found to depend on acid. The pH had to be t2 for transfer to occur. The addition of pepsin greatly increased the amount of cobalamin transferred. This increase can be seen most clearly in the experiments using saliva rather than gastric juice, which was never entirely free of pepsin activity. Saliva is also a more physiological target because it is the first digestive secretion to which food is normally exposed. Moreover, saliva rather than the stomach is the source of secretory R binder (36). Our data suggest that salivary enzymes do not promote the transfer of food cobalamin to salivary R binder, at least at neutral pH. Although an acid pH is obviously a necessary condition, primarily because optimal pepsin activity requires a low pH, it is not sufficient in itself for optimal transfer of food-bound cobalamin. Nevertheless, the minimal effect of acid on transfer of cobalamin to saliva, where barely measurable pepsin levels were present, suggests that acid pH also has a slight, possibly direct activity promoting such transfer that is independent of pepsin. The failure of added pepstatin A to inhibit such transfer supports this conclusion. Cobalamin from egg yolk was always transferred to R binder rather than to intrinsic factor. Our observations confirm previous findings that intrinsic factor has a much lower affinity than R binder for cobalamin at a low pH (29). We have further shown that pepsin does nothing to change this equilibrium between intrinsic factor and R binder. The transfer of cobalamin from food to R binder has important implications because many patients may have food cobalamin malabsorption despite normal intrinsic factor secretion (l-20). Our data support previous suggestions that the lack of acid and pepsin may be a major factor in such cases. However, loss of
June 1990
COBALAMIN TRANSFER FROM EGG YOLK TO R BINDERS
pepsin seems to be important only if it is virtually complete. As shown in our experiments, as little as 30 U of exogenously added pepsin per milliliter was adequate to promote transfer of cobalamin from egg yolk when the pH was 1.5. This concentration is only a small fraction of the normal pepsin content of gastric juice (35). The data in Table 1 suggest that endogenous gastric juice pepsin was slightly less active than pepsin added to saliva, because acidified gastric juices containing 89-95 U pepsin per milliliter promoted slightly less transfer than comparable added pepsin concentrations. The explanation for this minor discrepancy is not clear. In contrast to the relatively small pepsin requirement, even a mild impairment of acid secretion sufficed to almost completely inhibit transfer of cobalamin from egg yolk. No transfer was evident at pH >2. Thus, even mild hypochlorhydria should prove inhibitory to normal absorption of food cobalamin, regardless 0.f gastric pepsin content. The following sequence of events can be proposed for the initial phases of cobalamin assimilation in the normal individual. Food mixes with saliva in the mouth and enters the stomach without having released any of its cobalamin. The low pH and the pepsin encountered in the stomach then release the cobal,amin from its binders in food. This cobalamin becomes bound to the salivary R binders that had also enterled the stomach. The possible contribution of other gastric, duodenal, and biliary enzymes to this process of transfer remains to be defined. It is possible that food cobalamin malabsorption may not always be attributable solely to gastric dysfunction (1820). Full clarification of all potential mechanisms of food cobalamin malabsorption vvill require further investigation. As postulated by Okuda et al. (371 and demonstrated by A.llen et al. (29) and subsequently others (38-411, after cobalamin has bound to R binder, pancreatic enzymes degrade the R binder. This event sets the stage for the later binding of the released cobalamin by intrinsic factor in the upper small intestine. Some animal studies have also suggested a minor but direct absorptive role for pancreatic secretions (42,431, but intrinsic factor is the essential transport protein that is responsible for the eventual absorption of cobalamin across the ileal epithelium.
References 1. Doscherholmen
A, Swaim WR. Impaired assimilation of egg Co5’vitamin B,z in patients with hypochlorhydria and achlorhydria and after gastric resection. Gastroenterology 1973;64:913919. 2. Doscherholmen A, McMahon J, Ripley D. Vitamin B,, absorption from eggs. Proc Sot Exp Biol Med 1975;149:987-990. 3. Doscherholmen A, McMahon J, Ripley D. Inhibitory effect of
1465
eggs on vitamin B,, absorption: description of a simple ovalbumin 57CoB,, vitamin B,, absorption test. Br ] Haematol 1976;33: 261-272.
4. Doscherholmen A, McMahon J, Ripley D. Vitamin B,, assimilation from chicken meat. Am J Clin Nutr 1978;31:825-830. A, McMahon J, Economon P. Vitamin B,, 5. Doscherholmen absorption from fish. Proc Sot Exp Biol Med 1981;167:480-484. 6. Salom IL, Silvis SE, Doscherholmen A. Effect of cimetidine on the absorption of vitamin B,,. Stand J Gastroenterol1982:17:129131.
7. Doscherholmen A, Silvis S. McMahon J. Dual isotope Schilling test for measuring absorption of food-bound and free vitamin B,, simultaneously. Am J Clin Path01 1983;80:490-495. 8. Streeter AM, Duraiappah B, Boyle R, O’Neill BT, Pheils MT. Malabsorption of vitamin B,, after vagotomy. Am J Surg 1974:128: 340-343.
9. Streeter AM, Shum HY, Duncombe VM, Hewson JW, Thorpe MEC. Vitamin B,, malabsorption associated with a normal Schilling test result. Med J Austr 1976;1:54-55. 10. Carmel R. Nutritional vitamin B,, deficiency. Possible contributory role of subtle vitamin B,, malabsorption. Ann Intern Med 1978;88:647-649. 11. King CE, Leibach J, Toskes PP. Clinically significant vitamin Bj2 deficiency secondary to malabsorption of protein-bound vitamin B,Z.Dig Dis Sci 1979;24:397-402. 12. Steinberg WM. King CE, Toskes PP. Malabsorption of proteinbound cobalamin but not unbound cobalamin during cimetidine administration. Dig Dis Sci 1980;25:188-192. 13. Cattan D, Belaiche J, Zittoun J, Yvart J, Chagnon JP, Nurit Y. Role de la carence en facteur intrinseque dans la malabsorption de la vitamine B,, liee aux proteines, dans les achlorhydries. Gastroenterol Clin Biol 1982;6:570-575. 14. Streeter AM, Goulston KJ, Bathur FA, Hilmer RS, Crane GG, Pheils MT. Cimetidine and malabsorption of cobalamin. Dig Dis Sci 1982:27:13-16. 15. Dawson DW, Sawers AH, Sharma RK. Malabsorption of protein bound vitamin B,,. Br Med J 1984;288:675-678. 16. Hamborg B, Kittang E, Schjonsby H. The effect of ranitidine on the absorption of food cobalamins. Stand J Gastroenterol 1985; 20:756-758. 17. Jones BP, Broomhead AF, Kwan YL, Grace CS: Incidence and clinical significance of protein-bound vitamin B,, malabsorption. Eur J Haematoll987:38:131-136. 18. Carmel R, Sinow RM. Karnaze D. Atypical cobalamin deficiency. Subtle biochemical evidence of deficiency is commonly demonstrable in patients without megaloblastic anemia and is often associated with bound cobalamin malabsorption. J Lab Clin Med 1987;109:454-463. 19. Gozzard DI, Dawson DW, Lewis MJ. Experiences with dual protein-bound aqueous vitamin B,, absorption test in subjects with low serum vitamin B,, concentrations. J Clin Path01 1987:40:633-637. 20. Carmel R, Sinow RM, Siegel ME, Samloff IM. Food cobalamin malabsorption occurs frequently in patients with unexplained low serum cobalamin levels. Arch Intern Med 1988:148:17151719. 21. Sullivan LW. Herbert V, Reizenstein P. Evidence against preferential intestinal absorption of physiologic quantities of liver-bound vitamin B,, by patients with pernicious anemia. Am J Clin Nutr 1962;11:568-573. 22. Reizenstein PG. Effect of digestive enzymes on bound vitamin B,,. Acta Med Stand 1959;165:481-486. 23. Cooper BA, Castle WB. Sequential mechanisms in the enhanced absorption of vitamin B,, by intrinsic factor in the rat. J Clin Invest 1960;39:199-214. 24. Schade SC, Schilling RF. Effect of pepsin on the absorption of food vitamin B,, and iron. Am J Clin Nutr 1967;20:636-640,
1466 DEL CORRAL AND CARMEL
25. Adams JF, Kennedy EH. Thomson J, Williamson J. The effect of acid peptic digestion on free and tissue-bound cobalamins. Br J Nutr 1968;22:111-114. 26. Kittang E, Schjonsby H. Effect of gastric anacidity on the release of cobalamins from food and their subsequent binding to R protein. Stand J Gastroenterol1987;22:1031-1037. 27. Goldberg LS, Fudenberg HH. Effect of pH on the vitamin B,,-binding capacity of intrinsic factor. J Lab Clin Med 1969;73: 469-475. 28. Shum HY, O’Neill BJ, Streeter AM. Effect of pH changes on the binding of vitamin B12 by intrinsic factor. J Clin Path01 1971:24: 239-243. 29. Allen RH, Seetharam B, Podell E, Alpers DH: Effect of proteolytic enzymes on the binding of cobalamin to R protein and intrinsic factor. In vitro evidence that a failure to partially degrade R protein is responsible for cobalamin malabsorption in pancreatic insufficiency. J Clin Invest 1978;61:47-54. 30. Levine AS, Doscherholmen A. Vitamin B,, bioavailability from egg yolk and egg white: relationship to binding proteins. Am J Clin Nutr 1983;38:436-439. 31. Carmel R, Herbert V. Deficiency of vitamin B,,-binding alpha globulin in two brothers. Blood 1969;33:1-12. 32. Flood C, West R. Some properties of Castle’s intrinsic factor. Proc Sot Exp Biol Med 1936;34:542-543. 33. Skak-Nielsen T, Holst JJ, Nielsen OV. Role of gastrin-releasing peptide in the neural control of pepsinogen secretion from the pig stomach. Gastroenterology 1988;95:1216-1220. 34. Anson ML, Mirsky AE. The estimation of pepsin with hemoglobin. J Gen Physiol1932;16:59-63. 35. Hirschowitz BI. Electrolytes in human gastric secretions. Observations and theory. Am J Dig Dis 1961;6:199-228. 36. Kudo H, Inada M, Ohshio G, Wakatsuki Y, Ogawa K, Hamashima Y, Miyake T. Immunohistochemical localisation of vitamin B,, R-binder in the human digestive tract. Gut 1987;28:339-345. 37. Okuda K, Kitazaki T, Takamatsu M. Inactivation of vitamin B,,
GASTROENTEROLOGY
38.
39.
40.
41.
42.
43.
Vol. 98, No. 6
by a binder in rat intestine and the role of intrinsic factor. Digestion 1971;4:35-48. Andersen KJ, von der Lippe G. The effect of proteolytic enzymes on the vitamin B,,-binding proteins of human gastric juice and saliva. Stand J Gastroenterol1979;14:833-838. Marcoullis G. Parmentier Y, Nicolas JP, Jimenez M, Gerard P. Cobalamin malabsorption due to nondegradation of R proteins in the human intestine. J Clin Invest 1980;66:430-440. Carmel R, Abramson SB, Renner IG. Characterization of pure human pancreatic juice: cobalamin content, cobalamin-binding proteins and activity against human R binders of various secretions. Clin Sci 1983;64:193-205. Belaiche J. Zittoun J. Marquet J, Yvart J, Cattan D. In vitro effect of duodenal juice on R binders cobalamin complexes in subjects with pancreatic insufficiency: correlation with cobalamin absorption Gut 1987;28:70-74. Okuda K. Enhanced vitamin B,, absorption from rat intestine by proteases in the absence of intrinsic factor. Proc Sot Exp Biol Med 1967;124:79-83. Carmel R. Hollander D, Gergely HM, Renner IG, Abramson SB. Pure human pancreatic juice directly enhances uptake of cobalamin by guinea pig ileum in vivo. Proc Sot Exp Biol Med 1985:178:143-150.
Received May 23,1989. Accepted November 16.1989. Address requests for reprints to: Ralph Carmel, M.D., RMR 306, USC School of Medicine, 2025 Zonal Avenue, Los Angeles, California 90033. This study was supported by grant DK-32640 from the National Institutes of Health. Dr. de1 Corral was supported by a scholarship from Consorzio per 1’LJniversita degli Studi, Bari, Italy. The authors thank Dr. Sandra Schiffman for helpful suggestions, Rosemarie Nimo and Che Hung Yeh for technical assistance, and Ruth Ballard for manuscript preparation.