Journal of Clinical InvestigattonVol. 41, No. 2, 1962

STUDIES ON THE ABSORPTIONBY GUINEA PIG INTESTINEOF CYANOCOBALAMININCUBATEDWITH INTRINSIC

FACTOR*

By BERNARDA. COOPER,t WILLIAM PARANCHYCHANDLOUIS LOWENSTEIN

(From the Haematology Service, Department of Medicine, Royal Victoria Hospital, McGillUniversity Clinic, and the McGill-Montreal General Hospital Research Institute,

Montreal, Canada)

(Submitted for publication July 20, 1961; accepted October 12, 1961)

The absorption of cyanocobalamin from thegastrointestinal tract requires the mediation ofthe gastric intrinsic factor (1). Although recentstudies suggest that the cobamide coenzymes maybe the naturally occurring form of vitamin B12(2), these coenzymes also require intrinsic fac-tor for absorption from the gastrointestinal tract(3, 4).

The demonstration by Wilson and Strauss (5)that everted sacs of guinea pig intestine in vitrotake up cyanocobalamin in the presence of humangastric juice, has provided a tool for studyingthe action of human intrinsic factor in vitro. Wehave utilized this system to study the absorptionof cyanocobalamin from the gastrointestinal tract,in conjunction with studies of cyanocobalamintransport into normal and malignant cells (6).

Previous studies with perfused isolated rat in-testinal loops in vivo have indicated that theabsorption of cyanocobalamin is not affected bymetabolic poisons (7). The difficulties associatedwith studies with metabolic inhibitors in vivo haveprompted us to repeat these studies in the iso-lated sac system.

Herbert (8) has demonstrated that calciumions are required for the uptake of hog intrinsicfactor by rat intestine segments at 40 C in vitro,and that most of this radioactivity could be re-moved by subsequent rinsing with ethylenediaminetetraacetate (EDTA). This calcium dependenceof intrinsic factor has been observed in vivo inthe rat (7, 9) and in man (10). Although cal-cium-binding materials interfere with the in-

* Supported by Grants MA-802 and MA-1016, MedicalResearch Council of Canada, and U. S. Public HealthService Grant C-3596.

t Medical Research Associate, Medical Research Coun-cil of Canada.

testinal absorption of a number of substances(11), the observation that a portion of the cyano-cobalamin taken up in the presence of normalhuman gastric juice (NHGJ) by the perfusedisolated rat intestine loop in vivo is removed byrinsing with EDTA, but not with saline (7),further emphasizes the probable role of bivalentcations in the physiology of absorption of vitaminB12' Because of these observations we havecarried out further studies on the cationic re-quirements for vitamin B12 absorption.

MATERIALS AND METHODS

Cyanocobalamin labeled with Co' or Co"'1 (referredto as Co'6B,, or CoTB,,2) was adjusted to a specific activityof 2 to 3 uec per ,ug with nonradioactive cyanocobalamin.2The cyanocobalamin solutions then were diluted with wa-ter to a concentration of 46 mlAg per ml, and stored at-200 C in small aliquots until used. With each experi-ment a standard was counted, so that the radioactivityvalues could be converted to millimicrograms of cyano-cobalamin.

Normal human gastric juice was prepared as de-scribed previously (7) and stored at -20° C in smallaliquots. Rat gastric juice (RGJ) and guinea pig gas-tric juice (GPGJ) were collected from fasted animals bythe method of Shea (12), as described previously (7).

Male guinea pigs weighing 250 to 350 g were killedby decapitation; the lowest 80 cm of small intestine wasremoved, and 8 everted sacs of intestine were preparedas follows: The segments were turned inside out withthe aid of a plastic probe, and rinsed free of luminal ma-terial with cold 0.9 per cent saline. Each segment wasmade into a tied intestinal sac as described by Wilsonand Wiseman (13). These sacs were numbered 1-8, be-ginning with the segment adjacent to the cecum. Theywere injected with 0.7 ml of medium and then placed in25-ml Erlenmeyer flasks containing 5.0 ml of incubationsolution. The incubation solutions contained Co'5B,.,

1 Obtained from Merck & Co., Canada, Ltd.2 Generously supplied as "standard cyanocobalamin"

by Mr. I. Lesk, Merck & Co., Canada, Ltd.

370

INTESTINAL ABSORPTIONOF CYANOCOBALAMIN

and were adjusted so that the final ionic concentrationswere those of a modified Krebs-Ringer phosphate glu-cose medium with the following composition, in milli-equivalents per liter: Na 143, K 6, Mg 0.65, Ca 0.47, Cl150, P04 3.1; and glucose 4.7 mM. Flasks were incu-bated for 60 minutes in a Dubnoff metabolic shaking in-cubator in an atmosphere of 100 per cent oxygen at 370 C.

It has been reported previously (7) that, when radio-active cyanocobalamin bound to NHGJ is taken up byperfused intestinal loops of rats, a portion of the radio-activity taken up, which is not removed by saline rinsing,can be removed by rinsing with EDTA. We have de-fined the radioactive cyanocobalamin taken up by theintestinal sacs and not removed by saline rinsing as theS fraction, the radioactive cyanocobalamin not removedby EDTA as the E fraction, and the value obtained uponsubtracting the E from the S as the S-E fraction. Thus,the cyanocobalamin taken up by the intestine was differ-entiated into three fractions, which were defined as fol-lows: S, the value obtained after rinsing the intestinewith saline; E, the value obtained after rinsing the in-testine with EDTA; and S-E, the difference between Sand E (i.e., the cyanocobalamin which was washed fromthe intestine by rinsing with EDTA but not by rinsingwith saline). In most of the studies to be described, theuptake of cyanocobalamin into each of these fractionswas determined.

To study the uptake of Co5'B,2 into the two fractions,sacs were incubated in pairs, and rinsed for 40 minutesat 4° C at the conclusion of the incubation period. Oneof each pair was rinsed with a saline rinse solution (Na155 mEq per L, Cl 137 mEq per L, PO 5 mEq per L;pH 7.4) while the other was rinsed with an EDTArinse solution (Na 155 mEq per L, ethylenediaminetetraacetate 45 mEq per L; pH 7.4).

After rinsing, the sacs were blotted dry, emptied oftheir contents, and the sutured ends discarded; the re-mainder of each sac was measured, placed in a test tube,and the radioactivity determined in a well-type scintil-lation counter. The uptake was expressed as millimicro-grams of cyanocobalamin taken up per 10 cm of intestine.

In order to determine if more of the cyanocobalamintaken up. by the intestinal sacs could be removed by re-peated rinsing, sacs were incubated in a medium con-taining Co5'B,2 and NHGJ. After incubation, the sacswere rinsed with 10 ml of the saline or EDTA rinsingsolution for 20, 40, 60, or 80 minutes. After each 20 min-utes of rinsing, the sacs were transferred to flasks con-taining fresh rinsing solution. After 40 minutes of rins-ing, little additional radioactivity was washed from theintestinal sacs by either the saline or EDTA rinse, andthe S-E value remained constant. For this reason, allrinsing in subsequent experiments was carried out for40 minutes.

Digestion of the intestinal sacs in concentrated sul-furic acid before counting gave results no different fromthose obtained when the intestinal sacs were placed in atest tube and counted directly. For this reason, all ra-dioactive counting was carried out on undigested in-

testinal segments. Equilibrium dialysis was carried outas described previously (7).

OBSERVATIONS

In order to determine whether Co58B12 boundto intrinsic factor could be absorbed by intactguinea pigs, 9.2 m1ug of Co58B12 mixed with 0.3ml of NHGJ in a total volume of 1.0 ml wasadministered by stomach tube to guinea pigsanesthetized previously with pentobarbital. After5 hours the animals were killed by decapitationand the radioactivity in the liver and kidneys wasmeasured. This was compared with the radio-activity in the liver and kidneys of guinea pigstube-fed C058B12 mixed with 0.9 per cent saline.

Co58B19 administered with NHGJ was ab-sorbed at least as well as was the Co58B12 admin-istered with saline. The Co58B.2 found in theliver and kidneys of five animals receiving Co58B12+ saline was 0.75 ± 0.24 mlug (mean ± standarddeviation), whereas the Co58B12 in the liver andkidneys of six animals receiving Co58B1, + NHGJwas 0.98 ± 0.37 m/ug. The liver and kidneys oftwo animals tube-fed Co58B19 + NHGJ2.5 hoursbefore they were sacrificed contained only 0.05m,ug Co58B9. Thus, the majority of the absorbedCo58B12 appeared in the liver and kidneys be-tween 2.5 and 5 hours after tube feeding. Thisdelayed absorption of cyanocobalamin from theintestine is similar to the pattern of absorptionreported in the rat (14) and in man (15, 16).

Correction for biologic variation. When theuptake of Co58B12 mixed with NHGJ was meas-ured, variations in uptake were noted among in-testinal sacs obtained from different animals, andamong sacs obtained from a single animal. Asshown in Table I, the variation between theamount of Co58B1.> taken up by different sacs de-rived from each animal was considerably smallerthan the variation between sacs of the same num-ber obtained from different animals. In orderto correct for this biologic variation, so that theuptake under different conditions could be com-pared, all experiments were designed as follows.The number of different experimental conditionstested in each experiment was restricted to thenumber of sacs which could be obtained from oneanimal (eight). As described previously, eachsac was numbered, so that all sacs of the samenumber were prepared from intestinal segments

371

B. A. COOPER,W. PARANCHYCHAND L. LOWENSTEIN

TABLE I

Comparison of uptake by intestinal segments ofCo"5B12 incubated with NHGJ*

Uptake (myg/10 cm intestine)

Animal no.Segment Standard

no. 1 2 3 4 Mean error

1 0.47 0.73 1.58 1.95 1.18 0.212 0.75 0.78 1.66 1.42 1.15 0.183 0.79 0.90 1.44 1.76 1.22 0.204 0.51 1.09 1.65 2.49 1.43 0.375 0.76 0.93 1.78 1.95 1.35 0.266 0.43 0.96 1.61 2.07 1.27 0.317 0.59 0.79 1.94 1.77 1.27 0.298 0.34 0.70 1.38 2.32 1.18 0.38

Mean 0.58 0.86 1.63 1.84 1.26Standard 0.06 0.04 0.05 0.12 0.03

error

* Uptake by intestinal segments from 4 animals of Co68B12 incubatedfor 60 minutes with NHGJ. The segments were rinsed in saline rinsesolution after incubation. Correction is made for the large variationbetween animals, and between segments obtained from 1 animal, bycalculating the mean uptake of each segment for all 4 animals.

situated the same distance from the cecum. Eachexperiment was repeated in a sufficient number ofanimals so that each experimental condition couldbe tested in each of sacs 1-8. In order to accom-plish this, each sac was obtained from a differentanimal.

At the conclusion of the experiment, the meanvalue for each experimental condition was cal-culated. Since this mean value was a correctionfor the segment-to-segment and animal-to-animalvariation, it was found to be less variable than thevalues obtained when sacs obtained from oneanimal were used. In some experiments fewerexperimental conditions were tested, and so this"round-robin" scheme was completed by usingfewer than eight animals. In all of the experi-ments using everted intestinal sacs described be-low, results are expressed as the mean value,defined in this way.

Cyanocobalamin-binding capacity of normal hu-man gastric juice. To determine the concentrationof NHGJ which, when mixed with a standardquantity of Co58B12, would produce optimal up-take of cyanocobalamin by intestinal sacs, the sacswere incubated in different concentrations ofNHGJmixed with 13.8 mjug Co58B2. As moreNHGJ was added, the uptake by the intestinalsegments increased to a maximum and then de-creased. The maximum corresponded to thebinding capacity of the NHGJ for vitamin B12when measured by equilbrium dialysis. In all

subsequent experiments the concentration ofNHGJwas adjusted to bind all of the cyanocobal-amin in the medium.

The uptake by intestinal sacs of Co58B.2 in-cubated in a medium containing 9.2 mfg ofCo58B12, without intrinsic factor, was 0.25 to 0.38mAg per 10 cm. None of this could be removedby EDTA rinsing. Thus, no S-E fraction wasfound when Co5"B12 alone was incubated with in-testinal sacs.

Kinetics of uptake. The rate of uptake ofCo58B12 into the E and SUE fractions of guineapig intestine in vitro is illustrated in Figure 1.It is apparent that the pattern of uptake into bothfractions is linear with time for 60 minutes. Thecurves have been fitted by the method of leastsquares to the points determined experimentally.Different experiments designed to investigatemore completely the rate of uptake of Co55B1,incubated with NHGJ have revealed that therate of uptake is constant from 0 to 60 minutes.In these experiments 9.2 ,ug C058B12 and 0.4 mlNHGJ were present in each incubation vessel.

Intestinal sacs were incubated in different con-centrations of a Co58B 2-NHGJ mixture. As theconcentration of the B,2-NHGJ mixture was in-creased, the uptake per 10 cm of intestine in-creased in an exponential fashion and approacheda maximum, as shown in Figure 2. The pattern

I.2

Co B12 o10 E

UPTAKE0.8/ S-E

(mAg )0.6/

0.4

0.2-

0 .0 20 40 60

- TIME OF INCUBATION -min. )

FIG. 1. RATE OF UPTAKEBY GUINEA PIG INTESTINE OFCoB1,2 INCUBATED WITH NORMALHUMANGASTRIC JUICE.Intestinal sacs incubated in vitro for different periods inmedium containing 9.2 m/Ag Co'6B12 and 0.4 ml NHGJ.The uptake of Co'Bl2 into the E and S-E fractions is ex-pressed as m/Ag Co"B1, per 10 cm of intestine. The curveswere calculated by the method of least squares.

372

INTESTINAL ABSORPTIONOF CYANOCOBALAMIN

.2 7

co58 a l2

UPTAKE

m~ug )

1.0 -

0.8

0.6

0.4

0.2 -

4.5 9.0 1 8.0 31.5 36.0 4 5.0

Co58 B12 (mAg) + N HGJ

FIG. 2. EFFECT OF CONCENTRATIONOF Co B12-NHGJMIXTURE ON UPTAKEOF Co'B12 BY INTESTINAL SACS. In-testinal sacs were incubated for 60 minutes in vitro inmedium containing different concentrations of Co'B12and an equivalent quantity of NHGJ. The uptake intothe E and S-E fractions is compared. Curves were cal-culated by the method of least squares. In this experi-ment, the mixture contained 22.5 m/Ag Com'B,2 per mlNHGJ.

of uptake with increasing concentration was simi-lar for both the S-E and the E fractions.

Metabolic inhibitors. To study the metabolicrequirements of vitamin B12 uptake in the presence

of NHGJ, intestinal sacs were incubated for 60minutes in incubation solutions, under differentconditions. The sacs were incubated: at 40 C

TABLE II

Effect of metabolic inhibitors on uptake by guinea pigintestinal sacs of Co58B12 incubated with NHGJ*

Uptake (mgg/10 cm intestine)

Inhibitor E S-E

Control 0.92 0.58Incubated at 4° C 0.11 0.06

%Inhibition 88 90

Control 0.70 0.37N2 atmosphere 0.65 0.10

%Inhibition 7 73

Control 0.76 0.19DNP 0.35 0.05

%Inhibition 54 74

Control 0.74 0.72IA 0.49 0.22

%Inhibition 34 69

* Intestinal sacs were incubated in a medium containingCo55BI2 and NHGJ. The effect of incubation at 40 and at370 C, in an atmosphere of nitrogen, and with 1 mM2,4-dinitrophenol (DNP), or 1 mMsodium iodoacetate (IA),is compared with incubation at 37° C in oxygen (control).Uptake is divided into: E, residual Co58B12 after EDTArinsing; and S-E, the quantity of Co58B12 removed byrinsing with EDTAbut not with saline.

TABLE III

Effect of cyanide on uptake by guinea pig intestinal sacs ofCo88B12 incubated with human, rat, and guinea pig

gastric juice *

Uptake (mpg/10 cm intestine)

Intrinsic factor E S-E

NHGJ 1.00 0.55NHGJ+ NaCN 0.45 0.10

%Inhibition 55 82

RGJ 0.75 0.22RGJ + NaCN 0.23 0.03

%Inhibition 69 82

GPGJ 0.67 0.30GPGJ+ NaCN 0.24 0.13

%Inhibition 64 57

* Intestinal sacs were incubated in a medium contain-ing 9.2 m;&g Co55B12 and NHGJ, RGJ, and GPGJ. Theeffect on uptake of Co58B12 of 1 mMcyanide in the mediumis compared.

in oxygen; at 370 C under anaerobic conditions;at 370 C in oxygen, in the presence of 1 mM2,4-dinitrophenol (DNP); or at 370 C in oxygen, inthe presence of 1 mMsodium iodoacetate (IA).

As shown in Table II, marked inhibition ofCo55B12 uptake was noted at 40 C. At 370 C,under anaerobic conditions and in the presenceof 1 mMdinitrophenol or iodoacetate, some inhi-bition of uptake was noted. This inhibition wasmore striking in the S-E fraction than in the Efraction. The inhibition of uptake observed dur-ing incubation under anaerobic conditions wasmuch less marked than that observed by Straussand Wilson (17). The explanation for this isnot apparent.

As shown in Table III, 1 mMsodium cyanideinhibited the uptake of vitamin B12 incubatedwith normal human gastric juice, rat gastric juice,and guinea pig gastric juice. However, whereasuptake into the S-E fraction was more severelyaffected when NHGJ and RGJ were used, cy-anide appeared to affect uptake into both fractionsto an equal degree when GPGJ was used.Whether this represents a difference betweenhomologous and heterologous intrinsic factor isnot apparent at this time.

Cationic requirements. In order to determinethe cationic requirements of vitamin B12 uptakeby intestinal sacs in the presence of NHGJ, sacsivere incubated in solutions containing Co58B12and NHGJ in Krebs-Ringer phosphate glucose

373

B. A. COOPER, W. PARANCHYCHAND L. LOWENSTEIN

TABLE IV

Effect of EDTA on uptake of Co55B,2 by intestinal sacsincubated in vitro with NHGJ*

Uptake of Co58Bi2/1O cmintestine

EDTAconcentration E S-E

mM mipg0 0.77 0.230.045 0.82 0.310.09 0.38 0.000.225 0.28 0.03

* Uptake into E and S-E fractions of sacs of guinea pigintestine incubated in vitro with NHGJ in solutions con-taining different concentrations of EDTA.

(KRPG), in calcium- and magnesium-deficientKRPG, and in calcium- and magnesium-deficientKRPG containing different concentrations ofEDTA. The final sodium concentration of theseincubation solutions was adjusted to 155 mM.The omission of calcium and magnesium from theRinger's solutions did not affect uptake of Co55B12incubated with NHGJ. In one experiment meanE and S-E values were 1.04 and 0.68 mpeg, re-spectively, when the complete KRPGwas used,and 0.96 and 0.68 mpg, respectively, when cal-cium- and magnesium-deficient KRPGwere used.

As shown in Table IV, the addition of smallamounts of disodium EDTA abolished B12 up-take. Although "ion-free" water was used forthese experiments, the water was not analyzedto determine its cationic content.

To determine the relative importance of cal-cium, magnesium, manganous, and ferrous ionson vitamin B12 uptake in the presence of NHGJ,sacs were incubated in calcium- and magnesium-deficient KRPG containing 0.18 mM sodiumEDTA, calcium EDTA, magnesium EDTA, man-ganous EDTA or ferrous EDTA. All solutionswere adjusted to pH 7.4.

As shown in Table V, calcium, manganous, andferrous ions prevented the EDTA-induced inhi-bition of uptake of Co55B12 by the intestine,whereas magnesium ions did not. However, asshown in Table VI, when only 0.45 umole ofEDTAper vessel was added (final concentration,0.09 mM), the inhibition was corrected by eithercalcium or magnesium. Thus, in the presence of0.09 mMEDTA, calcium and magnesium wereequally effective in restoring B12 uptake. Whena higher concentration of EDTA was present

(0.18 mM), only calcium and the cations testedthat showed a stronger affinity for EDTA thancalcium (18) (viz, manganous and ferrous ions).restored the uptake.

Transport into serosal fluid. As observed byStrauss and Wilson (17), no significant transportof Co15B.2 into the serosal solutions was detectedat any time. When everted intestinal sacs wereobtained from animals tube-fed Co58B1.2 andNHGJ 2 hours previously, or from anesthetizedanimals in which Co58B12 had been instilled intoisolated intestine segments in vivo 1 to 2 hourspreviously, significant quantities of Co58B12 werenot found in the serosal fluid after incubationin vitro.

Relationship between E and S-E fractions.It has been suggested previously that vitamin B12taken up by the intestine may first enter the S-Efraction and then move into the E fraction (7).In order to test this hypothesis, and to determinethe relationship of these two fractions, the follow-ing experiment was carried out. Intestinal sacswere incubated for 40 minutes in the incubationmedium containing NHGJ and 9.2 m/Ag ofCo58B12. Then some of the sacs were rinsed with

TABLE V

Effect of ions on uptake, by everted intestinal sacs in vitro,of Co58BI2 incubated with NHGJ*

Uptake of CoH8Bis byintestine

Ions added to FractionsExpt. incubation

no. medium S E S-E

mpg mjpg mAg1 Control 1.60 1.02 0.58

Na EDTA 0.24 0.22 0.02CaEDTA 1.48 1.08 0.40MgEDTA 0.32 0.30 0.02

2 Control 1.22 0.84 0.38NaEDTA 0.21 0.21 0.00Ca EDTA 1.34 0.88 0.46MnEDTA 1.11 0.76 0.35

3 Control 1.72 1.17 0.55NaEDTA 0.17 0.17 0.00Ca EDTA 1.76 1.23 0.53Fe EDTA 1.88 1.19 0.69

* Mean uptake by guinea pig intestine of Co58B12 in-cubated with NHGJ. The uptake is divided into: S,residual radioactivity after saline rinsing; E, after EDTArinsing; and S-E, the difference between these. Eachincubation vessel contained 9.2 mp&g Co58B12, 0.3 mlNHGJ, and 9.0 IAM of the cationic chelate in 5 ml ofCa++- and Mg++-deficient Krebs-Ringer phosphate glucosemedium.

374

INTESTINAL ABSORPTIONOF CYANOCOBALAMIN

buffered saline and some with the EDTA rinsingsolution. After saline rinsing to remove theEDTA, the sacs were reincubated in fresh solu-tions containing 9.2 m/Ig Co57B12 and NHGJ foran additional 20 minutes. The results of thissequential incubation are shown in Figure 3. Itwill be noted that after saline rinsing and reincu-bation in the second medium the linear uptakeof cyanocobalamin with time into the E and S-Efractions was unchanged. The Co58B,2 taken up

during the first 40 minutes of incubation did notmove from the S-E to the E fraction during thesubsequent 20 minutes of incubation, whileCo57B,2 was being taken up. After EDTA rins-ing, Co57B19 was taken up into the S-E and Efractions at the same rate at which Co58 hadbeen taken up previously.

This experiment shows that the S-E and Efractions do not represent sequential phases ofuptake, that the two fractions are not in activeequilibrium, and that EDTA rinsing does notdamage the intestine, because the rate of uptakeof vitamin B12 after EDTA rinsing is identicalwith that before rinsing.

Effect of saline and EDTA rinsing on segmentsof intestine incubated in vivo. To determine ifan S-E fraction could be demonstrated in vivo,fasted guinea pigs were tube-fed 9.2 mjug of Co58-cyanocobalamin together with 0.3 ml of NHGJor saline. After 5 hours, the animals were de-capitated, and the small intestine was removedand divided into segments 10 cm in length. Usu-ally 12 segments were obtained from each animal.Alternate segments were rinsed in saline or

EDTA rinsing solutions. The results are sum-

marized in Table VII. Results are expressedas the sum of the C058B12 from all segments from

TABLE VI

Effect of calcium and magnesium on correction of the inhibi-tion of Co58B12 uptake induced by 0.09 mMEDTA in

KRPG

Uptake of Co58B12 by intestine

Ions added to Fractionsincubation

medium S E S-E

mpg mug mpg

Control 1.73 1.17 0.56Na EDTA 0.17 0.17 0.00Ca EDTA 1.76 1.23 0.53MgEDTA 1.88 1.19 0.69

1.0 -

B 12

UPTAKE

(mpug )

0.5

0

rinse

TOTAL

> S-E-

E

Co58 BipS-E

:-E C o5 BSE I

O 20 40 60- T ME OF N C U BA T I0 N -

( min. )

FIG. 3. SEQUENTIAL INCUBATION OF INTESTINAL SACS

IN Co'B,, AND NHGJ, AND IN Co57B,2 AND NHGJ. In-testinal sacs were incubated for 40 minutes in medium con-taining Co5"Bl! and NHGJ. They then were rinsed withsaline or EDTA rinsing solutions, and reincubated infresh medium containing Co57B12 and NHGJ. The uptakeof each isotope into the S and S-E fractions was com-pared. E values are indicated by 0, and S-E values by*. Solid lines indicate Co5'B12, and dotted lines Co51B,2taken up per 10 cm of intestine. In this experiment eachvessel contained 9.2 mpg Co58- or Co7B,, and 0.3 mlNHGJ.

each animal, which were rinsed in saline or EDTA.It is apparent that, in all animals, the sum ofthe Co58B12 in the segments rinsed in saline washigher than that in the segments rinsed in EDTA.Thus, even 5 hours after tube feeding, when con-

TABLE VII

Effect of saline and EDTA rinsing on C058B12 in guineapig intestine after 5 hours of incubation in vivo *

Total Co58B12 uptake by intestineFractions

Animalno. NHGJ S E S-E

mpg mpg mAg1 + 1.57 0.92 0.632 + 0.92 0.63 0.293 + 0.61 0.56 0.05

Mean + 1.04 0.71 0.33

4 0.89 0.73 0.165 0.54 0.44 0.10

Mean 0.72 0.59 0.13

* Mean uptake into S, E, and S-E fractions of the intes-tine of guinea pigs tube-fed Co58B12 with and withoutNHGJ. Five hours after tube feeding, the intestine wasdivided into 10-cm segments, and alternate segments wererinsed with saline or with EDTArinsing solutions. (Totalvalues are given for segments from each animal, rinsed witheither saline or EDTA.)

375

B. A. COOPER,W. PARANCHYCHAND L. LOWENSTEIN

siderable Co58B,2 was present in the liver, anS-E component was found in the intestine. Thissuggests that the division into E and S-E frac-tions is not peculiar to the in vitro incubation sys-tem, and that neither fraction represents an early,transient phase of vitamin B12 absorption.

DISCUSSION

These data indicate that the intact guinea pigabsorbs cyanocobalamin bound to human intrinsicfactor. Previous studies (19, 20) have demon-strated that the cyanocobalamin molecules whichare bound to intrinsic factor are absorbed in thegastrectomized rat and in man. Thus, the guineapig appears to be capable of absorbing cyanoco-balamin, using either human or guinea pig in-trinsic factor.

Wehave demonstrated again that the cyanoco-balamin taken up by intestinal sacs incubated withintrinsic factor may be separated into two frac-tions: one which is removed by EDTA rinsingbut not by saline rinsing (S-E), and one which isnot removed by EDTArinsing (E). The uptakeof cyanocobalamin into each fraction is linearwith time, and describes an exponential curvewith increasing concentrations of gastric juice-cyanocobalamin mixture. These fractions may bedifferentiated during the absorption of cyanoco-balamin in vivo, whether the cyanocobalamin isbound to human intrinsic factor or is free to com-bine with guinea pig intrinsic factor. They couldbe recognized 5 hours after tube feeding of cy-anocobalamin, at a time when cyanocobalamin wasleaving the intestine and being deposited in theliver and kidneys.

In previous experiments with the isolated, per-fused rat intestinal loop in vivo, an S-E fractionwas detected when human gastric juice was usedas a source of intrinsic factor, but no S-E waspresent when rat gastric juice was used (7). Inthe guinea pig intestinal preparation, an S-E frac-tion was present when human, rat, or guinea piggastric juice was used.

It is tempting to speculate that the S-E fractionmay be composed of cyanocobalamin loosely boundto the intestinal surface by partially competentintrinsic factor molecules, or, conversely, of cy-anocobalamin and intrinsic factor molecules at-tached to inefficient intestinal receptors by meansof bivalent cations. However, the inhibition of

uptake into the S-E fraction by metabolic inhibi-tors would suggest that this fraction may havephysiologic function. It is clear that, as cyanoco-balamin is absorbed, it does not progress throughone fraction into the other.

The apparent calcium dependence of uptake intoboth phases suggests that calcium ions may benecessary for absorption of cyanocobalamin boundto intrinsic factor. The demonstration that cal-cium ions may be necessary for the absorption ofmany substances (11) implies that calcium de-pendence may be a more general phenomenon thanhas been suggested previously (7). The failure ofAbels and associates (21) to demonstrate inhibi-tion of cyanocobalamin absorption with EDTA invivo is of interest, and may have been due to thedifficulty of chelating intestinal calcium in vivo.The calcium dependence of cyanocobalamin ab-sorption by way of the intrinsic factor mechanismin vivo would appear to be well established (7, 9,10, 22, 23). The replacement of calcium by mag-nesium in the presence of low concentrations ofEDTA (0.09 mM) may be due to a calcium-sparing effect. With higher concentrations ofEDTA (0.18 mM), only cations with greateraffinity for EDTAthan calcium (Mn and Fe) re-placed calcium and allowed normal uptake of cy-anocobalamin bound to human intrinsic factor.

The data available are insufficient to determinethe importance of the S-E and E fractions of ab-sorbed intestinal cyanocobalamin in the physio-logic absorption of cyanocobalamin from thegastrointestinal tract. However, the recognitionof these fractions requires that future studiesdifferentiate between them when the uptake of cy-anocobalamin by intestinal preparations is studied.

SUMMARY

1. The intact guinea pig absorbs cyanocobala-min bound to human intrinsic factor from thegastrointestinal tract.

2. The cyanocobalamin absorbed in vivo, andtaken up by intestinal sacs incubated with gastricjuice in vitro, can be differentiated into two frac-tions: one which is removed by EDTA rinsing(S-E), and one which cannot be removed byEDTA rinsing (E).

3. Uptake into the S-E and E phases in vitrois inhibited to different degrees by metabolic in-hibitors.

376

INTESTINAL ABSORPTIONOF CYANOCOBALAMIN

4. The S-E and E fractions do not representsequential phases of absorption of cyanocobalaminfrom the intestine.

5. The uptake of cyanocobalamin incubated withhuman gastric juice by intestinal sacs appears tobe dependent upon the presence of calcium ions inthe medium.

ACKNOWLEDGMENT

We are grateful to Professor J. H. Quastel for hisadvice and assistance, and to Miss Helle Puupill fortechnical assistance.

REFERENCES

1. Castle, W. B. Factors involved in the absorptionof vitamin B12. Gastroenterology 1959, 37, 377.

2. Toohey, J. I., and Barker, H. A. Isolation of coen-

zyme B,2 from liver. J. biol. Chem. 1961, 236,560.

3. Wasserman, L. R., Estren, S., Brody, E., and Her-bert, V. Intestinal absorption of vitamin B12.Lancet 1960, 1, 173.

4. Grasbeck, R., and Bj6rksten, F. Intestinal absorp-tion of vitamin B12 and the B,, coenzymes. Lancet1960, 1, 337.

5. Wilson, T. H., and Strauss, E. W. Some species dif-ferences in the intrinsic factor stimulation of B12

uptake by small intestine in vitro. Amer. J.

Physiol. 1959, 197, 926.6. Cooper, B. A., and Paranchych, W. Selective uptake

of specifically bound cobalt 58 vitamin B12 by hu-man and mouse tumour cells. Nature (Lond.)1961, 191, 393.

7. Cooper, B. A., and Castle, W. B. Sequential mecha-nisms in the enhanced absorption of vitamin B,2 by

intrinsic factor in the rat. J. clin. Invest. 1960, 39,199.

8. Herbert, V. Mechanism of intrinsic factor actionin everted sacs of rat small intestine. J. clin. In-vest. 1959, 38, 102.

9. Grasbeck, R., and Nyberg, W. Inhibition of radio-vitamin B,, absorption by ethylenediaminetetra-acetate (EDTA) and its reversal by calcium ions.Scand. J. clin. Lab. Invest. 1958, 10, 448.

10. Grasbeck, R., Kantero, I., and Siurala, M. Influenceof calcium ions on vitamin B,. absorption in

steatorrhoea and pernicious anaemia. Lancet 1959,1, 234.

11. Sognen, E. Intestinal absorption influenced by cal-cium-binding substances (abstract). Proc., Sec-ond Scand. Summer Meeting, Biochem., Med.,Chem., Pharmacol., and Physiol., Turku, 1959, p.67.

12. Shea, S. M. Inhibition of gastric secretion in ratsby some quaternary derivatives of atropine. Brit.J. Pharmacol. 1956, 11, 171.

13. Wilson, T. H. M., and Wiseman, G. Use of sacsof everted small intestine for study of transferenceof substances from mucosal to serosal surface. J.Physiol. (Lond.) 1954, 123, 116.

14. Booth, C. C., Chanarin, I., Anderson, B. B., andMollin, D. L. The site of absorption and tissuedistribution of orally administered MCo-labelledvitamin B,., in the rat. Brit. J. Haemat. 1957, 3,253.

15. Booth, C. C., and Mollin, D. L. Plasma, tissue andurinary radioactivity after oral administration of'Co-labelled vitamin B12. Brit. J. Haemat. 1956,2, 223.

16. Doscherholmen, A., and Hagen, P. S. A dual mecha-nism of vitamin B12 plasma absorption. J. clin.Invest. 1957, 36, 1551.

17. Strauss, E. W., and Wilson, T. H. Factors con-trolling B12 uptake by intestinal sacs in vitro.Amer. J. Physiol. 1960, 198, 103.

18. Handbook of Chemistry and Physics, 42nd ed.,C. D. Hodgman, Ed. Cleveland, Chemical Rub-ber Publishing Co., 1960, p. 1486.

19. Bishop, R. C., Toporek, M., Nelson, N. A., andBethell, F. H. The relationship of binding powerto intrinsic factor activity (abstract). J. Lab.clin. Med. 1955, 46, 796.

20. Nieweg, H. O., Shen, S. C., and Castle, W. B.Mechanism of intrinsic factor action in the gas-trectomized rat. Proc. Soc. exp. Biol. (N. Y.)1957, 94, 223.

21. Abels, J., Woldring, M. G., Nieweg, H. O., Gaber,J. G., and de Vries, J. A. Ethylenediamine tetra-acetate and the intestinal absorption of vitaminB12. Nature (Lond.) 1959, 183, 1395.

22. Okuda, K. Vitamin B,2 absorption in rats studiedby a "loop" technique. Amer. J. Physiol. 1960,199, 84.

23. Herbert, V. Studies on the role of intrinsic factorand vitamin B,, absorption, transport, and storage.Amer. J. clin. Nutr. 1959, 7, 433.

377