GASTROENTEROLOGY1995;108:1622-1628
Decreased Activity of Intestinal and Urinary Intrinsic Factor in
Gr sbeck-lmerslund Disease
JEAN-LOUIS GUleANT, * MONIQUE S A U N I E R , *
THIERRY LAMIREAU,* BERNARD DUCLOS, § MARC ANDRI e B I G A R D , * and RALPH G R A S B E C K II
• Laboratory of Cellular and Molecular Nutrition, INSERM Unit~ 308, and Department of Gastroenterology, University Hospital and Medical
Faculty of the University of Nancy I, Nancy, France; *Department of Pediatrics, University Hospital of Bordeaux, Bordeaux, France;
§Department of Gastroenterology, University Hospital of Hautepierre, Strasbourg, France; and IIMinerva Institute for Medical Research,
Helsinki, Finland
Background/Aims: The pathogenesis of inherited intes-
tinal cobalamin malabsorption (Gr~sbeck-lmerslund
disease) remains unknown. The authors studied
whether the disease corresponds to a defective expres-
sion and/or function of the intrinsic factor-cobalamin
receptor in the ileum. Methods: Intrinsic factor-cobala-
rain receptor activity was measured using radioisotope
assay and gel-filtration exclusion chromatography in il-
eal biopsy specimens and urine concentrates from 4
patients with Gr~sbeck-lmerslund disease and 5 con-
trols. Results: Receptor activity was 164 ± 13 fmol/
mg of protein in control biopsy specimens and < 2 . 6
fmol/mg protein in specimens from patients. The asso-
ciation constant was estimated to be 3.8 ± 0.4 (nmol/
L) -~ in controls. A dramatic decrease in receptor activ-
ity was also observed in urine concentrate from pa-
tients with an association constant of 1.9 and 3.3
(nmol/L) -~. Isoelectrofocusing of the cross-linked in-
trinsic factor-cobalamin receptor complex showed an
isoelectric point at 4.8 in a patient as well as in control
samples. Conclusions: It is concluded that Gr~sbeck-
Imerslund disease is related to decreased intrinsic fac-
tor-receptor activity in intestinal mucosa; the receptor
assay in urine can be helpful for diagnosis.
nherited cobalamin (vitamin B12 ; Cbl) malabsorptions
I are rare. They include lack of intrinsic factor (IF) or
transcobalamin II (TC II) synthesis, secretion of structur-
ally abnormal intrinsic factor, and selective intestinal
malabsorption. 1 The selective intestinal malabsorption,
also known as the Gr~isbeck-Imerslund syndrome, is an
autosomal recessive disease. 2'3 It is characterized by a
deficient absorption of Cbl not corrected by ingestion of
IF and often is associated w i t h proteinuria that is not
affected by correction of Cbl deficiency.2'3 Its pathogene-
sis remains unexplained. It is not known why Cbl malab-
sorption and proteinuria are inherited together. 4 The Cbl
malabsorption may correspond to various mechanisms,
including defective synthesis, abnormal intracellular traf-
ticking, abnormal function of the IF receptor, and intra-
ISABELLE G A S T I N , * A M A L SAFI,*
cellular failure to release Cbl from IF or to transfer Cbl
to transcobalamin. 4-6
Published findings on IF receptor activity in ileal bi-
opsy specimens from patients suggest the heterogeneous
nature of the receptor defect in this disorder. Mackenzie
et al. reported a lack of pathology by electron microscopy
of ileal biopsy specimens and similar in vitro uptake of
I F - C b l as in controls. Their data show that the disease
is caused by a defect in intracellular trafficking of Cbl
rather than a deficient synthesis of the receptor. 5 Burman
et al. have shown a defective uptake of Cbl by ileal
mucosa (in vivo) evidenced by subcellular fractionation
of ileal biopsy specimens after ingestion of labeled Cbl. 6
More recently, an inherited intestinal Cbl malabsorption
has been described in Schnauzer dogs, which has been
related to a defective brush-border expression of the re-
ceptor, despite unimpaired and even increased synthesis
of the receptor] '8 The dogs showed the same characteris-
tics as humans with the disorder and were studied in the
hope of shedding some light on the human disease.
In the present work, we report deficient I F - C b l recep-
tor activity in ileal homogenates from patients' biopsy
specimens. Receptor activity also was measured in urine
concentrates and found to be lower than in control urine.
It was concluded that our four cases of selective intestinal
malabsorption were related to diminished receptor ac-
tivity.
Materials and Methods
Patients
Four patients (V.E.R., V.I.S., G.I.N., and H.E.N.)were
studied. Two were from the same family; patient V.I.S. was
the father of patient V.E.R.V.I.S. was a 29-year-old man
whose parents were cousins. The Imerslund-Gr~isbeck disease
was diagnosed when he was 21 months old, after his admission
Abbreviations used in this paper: Cbl, cobalamin; IF, intrinsicfac-
tor; TC II. transcobalaminII.
© 1995 by the AmericanGastroenterologicalAssociation
0016-5085/95/$3.00
June 1995
Table 1. M e a n Biological Data on 4 Patients With G r ~ s b e c k - l m e r s l u n d D i s e a s e
Mean corpuscular
Hemoglobin volume
Patient (g/dL) ( fL)
V,E.R. 9.4 109
V.I,S. 4,7 120
H.E.N. 7.9 89
G.I.N. 6.2 121
Normal 13 - 16 75 - 9 5
to the hospital for megaloblastic anemia. Selective intestinal
malabsorption was established on the basis of a low Schilling
test not corrected by IF, physiological gastric IF secretion, and
proteinuria. His brother had presented the same syndrome
when he was 4 years old and died accidentally when he was
20 years old. Patient V.E.R. was a 20-month-old girl who was
admitted to hospital for anemia and poor growth. She had no
defective IF gastric secretion and mild proteinuria; the Schil-
ling test was low, even in the presence of IF. She was also the
daughter of cousins. Patient H.E.N. was a 2-year-old boy. He
was admitted to the hospital for megaloblastic anemia and
iron deficiency. Patient G.I.N. was a 37-year-old woman ad-
mitted to the hospital for the recurrence of megaloblastic ane-
mia. She had proteinuria and hypotrophy and was mentally
retarded. The main clinical and biological features of the pa-
tients are summarized in Table 1. Results of endoscopic and
histological examination of the stomach, duodenum, and ileum
were normal. TC II and R-binder blood levels were 440 pmolt
L and 345 pmol/L, respectively.
Schilling Test
The standard Schilling test was performed using the
Dicopac test (Amersham, England). The second stage (inges-
tion of IF-cyano [SVCo]Cbl) was started 4 days after the first
(ingestion of cyano [SSCo]Cbl). Urine specimens were collected
for 48 hours at each stage. The limit of normal values was
established to 10% and 11% of urinary excretion of the tracer
for the first and second stages, respectively.
IF-Cbl Receptor Activity in Ileal Mucosal
Homogenates
Biopsies were performed by intubation of the ileum
under colonic endoscopy after informed consent was obtained
from each patient's family. The biopsy specimens were stored
in liquid nitrogen. The mucosal extract was obtained by the
Potter homogenization of biopsy specimens in the presence of
50 ~L of 20 mmol/L Tris-HC1, pH 7.4, containing 0.15 moll
L NaC1, 0.01% (wt/vol) sodium azide, 0.4 mmol/L phenyl-
methylsulphonyl fluoride, and 0.5% Triton X-100. 9 After an
18-hour incubation period, the sample was centrifuged at
14,000g for 60 minutes and the supernatant was used for
binding capacity assay. The radioisotope assay of the solubi-
lized receptor was performed as described recently. 9':° An I F -
cyano [SVCo]Cbl tracer solution was prepared by incubating
0.7 pmol of human IF (purified from gastric juice) with 0.7
GRASBECK-IMERSLUND DISEASE 1623
Serum Schilling test
cobalamin Proteinuria
(pmol/L) Cbl (%) IF-Cbl (%) (g/L)
27 1,6 1.0 0.20
77 0.1 0.2 1.60
<75 1.0 1.0 0.63
<75 7.5 7.5 5.0
120-660 >10.0 >11.0 <0.05
pmol of CN [57Co]Cbl (8.1 Bq/gg, purchased from the Radio-
chemical Centre, Amersham, England) in 1 mL of 20 mmol/
L Tris-HC1 buffer, p H 7.4, containing 0.15 mol/L NaC1. Satu-
rated I F - C b l was purified from human gastric juice as pre-
viously described.:1 Aliquots (50 gL) of mucosal extracts were
incubated with 7 4 - 3 0 4 8 fmol of the I F - C N Cbl tracer solu-
tion (total volume, 1 mL) in the presence of either 1 mmol/L
CaC12 (first series) or 10 mmol/L ethylene diamine tetraacetic
acid (EDTA) (second series) for 1 hour while shaking. Then
200 gL of a suspension of phenyl-Sepharose (1 volume of
decanted and washed phenyl-Sepharose in 1 volume of Tris-
HC1 buffer, pH 7.4) and 300 gL of 20 mmol/L Tris-HC1
buffer, pH 7.4, containing 3.3 mol/L NaCI were added succes-
sively to each tube. The tubes were shaken for 30 minutes at
room temperature and then centrifuged at 1500g for 5 minutes.
The pellets were washed three times by repetitive addition of
3 mL of 20 mmol/L Tris-HCl buffer, p H 7.4, containing 0.3
mol/L NaC1. The radioactivity corresponding to IF receptor
activity was estimated by assessing the difference in radioactiv-
ity between the pellets obtained in the first and second series.
The association constant and concentration of IF receptor were
determined by Scatchard plot.
Receptor activity was also determined using gel filtration
as described recently. 9 Briefly, the mucosal homogenate (50
~£L) was incubated with 0.7 pmol of I F - C b l tracer solution
in the presence of either 1 mmol/L CaCI2 or 10 mmol/L EDTA
and filtered through a 0.5 × 15-cm Superose 6 minicolumn
at a flow rate of 0.1 mL/min with 20 mmol/L Tris-HC1, pH
7.4, containing 0.15 mol/L NaC1 and 0.05% Triton X-100.
IF-Cbl Receptor Activity in Urine
Concentrates
Urine samples were concentrated 50-fold in a 250-
mL Amicon ultrafiltration cell through a YM-5 membrane
(Amicon, Beverly, MA). IF-Cbl receptor activity was deter-
mined by phenyl-Sepharose radioisotope assay and by gel fil-
tration as described above using 50-~tL aliquots per test tube.
The receptor activity was determined by Scatchard plot. The
result was expressed in femtomoles of IF-binding activity per
milliliter of nonconcentrated urine.
Cross-linking of [57Co]CbI-IF to the Urinary
Receptor
The urinary receptor incubated with IF-Cbl was ad-
sorbed to 200 ~L of phenyl-Sepharose as described above. The
1624 GUEANT ET AL. GASTROENTEROLOGY Vol, 1 0 8 , No. 6

Table 2. IF Receptor Activity by R a d i o i s o t o p e Assay in Ileal Biopsy S p e c i m e n s and Urine Concentrates From 4 Patients With
G r & s b e c k - l m e r s l u n d Disease

Ileal homogenate Urine

IF receptor activity IF receptor activity

Kass a Kass a
Patients fmol/mg protein fmol/mg tissue (nmol/L) 1 fmol/mL fmol/mg creatinine (nmol/L) ~
V.E.R. <2.5 <0.4 ND 2.6 2.5 1.9
V,l.S. -- -- -- 3.9 ND 3.3
H.E.N. <2.5 <0.4 ND <2.5 ND ND
G.I.N. 81 3.3 3.4 13.7 6.5 2.2
Controls (n = 5) 164 + 13 6.3 4. 1.3 3.8 4. 0.4 26.1 4, 8.2 16.1 + 3.8 3.9 _+ 0.3
Pregnant women (n = 3) 158 ± 35.8 ND 3.3 4, 1.3
Tubular nephropathy (n = 3) 46.7 _~ 17.0 ND 3.02 4- 1.0

ND, not determined.

aAffinity constant.

suspension was washed in 0.1 mol/L sodium carbonate, pH
8.0, containing 0.2 mol NaC1. The pellet was incubated with
2 btmol of dithiobisuccinimidyl propionate for 15 minutes, as
described previously. 12 The reaction was stopped by adding
100 mmol/L Tris-HCl buffer, pH 8.0, containing 0.7 mol/L
NaC1 and 5 mmol/L EDTA, and the pellet was washed twice
in the same buffer. The CN [SVCo]Cbl-IF receptor complex
was eluted from phenyl-Sepharose with distilled water con-
taining 10% methanol (vol/vol). After evaporation, the sample
was dissolved in 20 mmol/L Tris-HCl buffer, pH 7.4, con-
taining 0.15 mol/L NaC1, 5 mmol/L EDTA, and 0.05% Triton
X-100. A second sample was dissolved in 0.5 mL of 0.05 moll
L sodium acetate buffer (pH 5.0) and incubated with 0.1 U
of neuraminidase (Arthrobacter; Boehringer-Mannheim, Ger-
many) for 18 hours at room temperature. Both samples were
then run on superose 6 B gel filtration. The column (1.0 ×
30 cm) was eluted with the Tris-HC1 buffer described above
at a flow rate of 1 mL/min. The IF-Cbl receptor peak was
collected and subjected to column isoelectrofocusing. Isoelec-
trofocusing was performed on a l l0-mL LKB column 8101
using a 0 % - 5 0 % sucrose gradient (wt/vol) in 3 mol/L urea
for 48 hours. The pH ranged from 2.5 to 8.
Results
Ileal biopsy specimens were obtained from 3 of the
4 patients and from 5 controls examined by endoscopy.
Inclusion criteria for controls was the absence of intestinal
lesions and vitamin malabsorption. The results are sum-
marized in Table 2. The receptor activity was approxi-
mately 6.3 + 1.3 fmol/mg of tissue in control biopsy
specimens. It was approximately twice as low in the
specimen from patient G.I.N. No activity was detected
in the specimens from patients V.E.R. and H.E.N. (Fig-
ure 1). The association constant of the receptor was simi-
lar in mucosal homogenates from patient G.I.N. and
controls; it was estimated at 3.4 and 3.8 - 0.4 (nmol/
L) -I, respectively (Figure 1). The receptor activity of ileal
homogenates was also studied in gel filtration and in a
colon biopsy specimen from a control (Figure 2). The
receptor peak was eluted as a high molecular mass aggre-
gate in the same position as Blue Dextran 2000. This
peak was approximately half as high in patient G.I.N.
as in a control extract. Again, no activity was found in
the extract from patient V.E.R. (Figure 2).
The receptor activity of urine concentrates was studied.
The I F - C b l receptor complex was also eluted as a high-
molecular-mass aggregate in gel filtration (Figure 2). Re-
ceptor binding of Cbl-IF was not inhibited by excess of
either free Cbl or R b i n d e r - C b l (data not shown). Cal-
cium was required for receptor binding of I F - C b l be-
cause no receptor peak was observed when the gel filtra-
tion receptor assay was performed in the presence of 10
mmol/L E D T A (Figure 2). The urine receptor activity
was estimated at 26.1 + 8.2 fmol/mL and at 16.1 + 3.8
fmol/g of creatinine in control urine concentrate (Table
2 and Figure 1). It was 7 and 10 times higher in urine
specimens from pregnant women and from patients with
tubular nephropathy, respectively (Table 2). The receptor
activity was decreased in patients with the G d i s b e c k -
Imerslund syndrome, but the affinity of IF-Cbl for the
receptor was not modified (Table 2). The isoelectric point
of the uroreceptor cross-linked to IF-(SVCo)Cbl was esti-
mated in column isoelectrofocusing at 4.84 and 4.84 in
patient V.E.R. and control, respectively (Figure 3). After
incubation with neuraminidase, the isoelectric points
shifted to 5.46 and 5.42, respectively.
Discussion
The diagnosis of Gr~/sbeck-Imerslund disease in
our 4 patients was established from Schilling test results,
low blood Bt2 level, megaloblastic anemia, and protein-
uria (Table 1).
Ileal endoscopy was performed in 3 patients. The mu-
cosal material taken from ileal biopsy specimens was
June 1995 GRP,SBECK-IMERSLUND DISEASE 1625

A 10. B 0,04.
/
o /
E / 0,03"
& /
IL
a
~. 5' ~= 0,02"
o

-1:1
- - ~ j m
m
0,01
O

0 ¢ ~ ¢ , • 0,00 i i I ! i i

1000 20o0 2 4 6 8 10 12
IF-Cbl(pmole/I) Bound (femlomole/assay}

C 80 D 0,4

E 60 0,3
&
a "-...
~. 4o 0,2

==
"~ 20 0,1
O

Jt & 0,0 i
0
1000 2000 3000 4000 20 40 60 8O
IF-Cbl (pmole/I) Bound (femtomole/assay)

Figure 1. IF receptor assay in intestinal homogenates and urine concentrates from patients and controls. (A) Saturation curve of binding of
[SrCo]Cbl-IF to the receptor from ileal homogenate of patient G.I.N. ([2]), patient V.E.R. (0), and control ([]). No IF receptor activity was detected
in ileum extract of patient V.E.R. (B) Scatchard analysis of saturation curves obtained with ileal homogenate from patient G.I.N. (E3) and control
(11). The affinity constant of IF receptor was not modified, whereas its activity was about twofold lower in the patient extract. (C) Saturation
curve of binding of [SrCo]Cbl-IF to the receptor from urine concentrates of patients G.I.N. (11) and V.E.R. (A) and of one control (E:]). (D)
Corresponding Scatchard analysis of the saturation curves.

not sufficient to allow subcellular fractionation and a
Scatchard study of I F - C b l binding to brush-border
membranes. Recently, we have developed a radioisotope-
binding assay to determine IF receptor activity in hog
ileal homogenate. In this assay, the free I F - C b l complex
was separated from I F - C b l bound to the receptor using
hydrophobic adsorption of the receptor complex to phe-
nyl-Sepharose. W e adapted this assay to reach a detection
limit as low as 2.5 fmol using only 50 btL of mucosal
homogenate per test tube. This allowed us to adapt the
assay to measure IF receptor binding activity in mucosal
homogenate obtained from 10 to 20 mg of tissue.
The receptor activity was not detectable in biopsy
homogenates from colonic mucosa. The activity was
6.3 + 1.3 fmol/mg tissue, and the association constant
was 3.8 -+ 0.4 (nmol/L)-* in ileal biopsy specimens from
5 controls. These values are close to those obtained by
Fyfe et al. with ileal homogenates from dogs 8 and those
obtained by us with ileal homogenate from a 25-week-
old human fetus. *° No I F - C b l receptor activity was de-
tected in the ileal homogenate from 2 of our patients
with Gr~isbeck-Imerslund disease. In patient G.I.N,, the
receptor activity was half that of controls and the associa-
tion constant was similar.
The IF receptor is synthesized not only by ileal entero-
cytes but also by epithelial cells of rat kidney. .3'14 In
addition, de novo synthesis of I F - C b l receptor has been
shown in renal epithelial cells. 15 The receptor was shown
to be exclusively targeted to the apical brush-border
membrane. ~6 Because the extramembrane domain of the
receptor is susceptible to proteolytic cleavage, *v'~s we
thought of searching for an I F - C b l receptor activity in
urine specimens. This soluble I F - C b l receptor was de-
tected in 50-fold concentrated control urine specimens.
The I F - C b l receptor complex behaved as the ileal recep-
tor in gel filtration because it was eluted as a high-
molecular-mass complex in the void volume and dissoci-
ated in the presence of EDTA. The soluble receptor was
a hydrophobic molecule as the I F - C b l receptor complex
was adsorbed to phenyl-Sepharose in our radioisotope
binding assay (Figure 1). The u r o - I F - C b l receptor was
detectable in the urine concentrates of patients and was
decreased with about the same factor as observed with
ileal homogenate (Table 2). The association constant was
close to that of the I F - C b l receptor from control urine
specimens. The decreased urinary excretion of I F - C b l
receptor was not related to proteinuria because no correla-
tion was found between the two parameters in patients
(Table 2). Our results suggest that ileal and renal I F -
Cbl receptor activities represent the same protein under
similar types of regulation because both are decreased in
Grgsbeck-Imerslund disease. This hypothesis is likely
1626 GUI~ANT ET AL. GASTROENTEROLOGY Vol. 108, No. 6

(57Co)-Cbl
CPM
F
I-(5"Tco)-Cb! fect concerns a gene that is expressed in both tissues. 8 There-
,0000 A
fore, the I F - C b l receptor activity in urine specimens can be
helpful in diagnosing Gr'~beck-Imerslund disease because
8000.
receptor activity is much higher in the kidney than in the
6000
ileum 14 and much easier to determine in urine than in
4000. intestinal biopsy specimens. The range of IF-Cbl receptor
activity of patients was very wide. It was very low in two
2000.
cases but only twice as low as control values in patient
0.
10 2O 3O 4O 5O G.I.N. (Table 2). This may explain that the disease can be
Fraction number diagnosed only in late infancy in some cases.4
(57Co)-Cbl IF-(b"J'Co)-Cbl Mackenzie et al. showed in vitro that the Cbl uptake
CPM
of ileal biopsy specimens was increased by IF in 1 patient.
60000"
B They concluded that the expression and function of the
157C~-Cbg
receptor were not altered) On the contrary, Burman et
40000" al. reported defective in vivo uptake of labeled Cbl in
biopsy specimens of ileal mucosa from two brothers. 6 In
20000"
the present study, we observed a decreased binding capac-
ity of I F - C b l receptor in the ileal mucosa. W e could
not conclude that the disease was related to decreased
10 20 30 40 50 6'0 synthesis of the receptor because the concentration of the
Fraction number receptor was not determined quantitatively by immuno-
assay. However, this hypothesis is plausible because the
receptor from patients had an affinity constant similar to
(57Co)-Cbcl
CPM
IF-(57Co)-Cbl

*
that of controls and a decreased activity. In addition, we
did not observe any modification of the physicochemical
p
60000"
properties of the urinary receptor cross-linked to the IF-

40000"
IF~-Nm ~ (57C~)-Cbl
20000'
10 20 30 40 50
Frsctlon number
Figure 2. (A) Superose 6 minigel filtration (column size, 0.5 × 10
cm) of mucosal extracts incubated with IF-labeled Cbl (IF-[~TCo]Cbl)
in the presence of calcium chloride. The IF receptor complex of ileal
homogenate of a control was eluted as a high-molecular-mass com-
plex (IF-rec) in the presence of calcium chloride ( • ). The complex was
not observed when the chromatographic assay was performed in the
presence of EDTA (data not shown). No receptor peak was observed
when the experiment was performed with ileal homogenate from pa-
tient V.E.R. (E3). (B) Superose 6 gel filtration (column size, 1 × 30
cm) performed with urine concentrate from a control incubated with
IF-[~TCo]CbI. The receptor peak is observed in the presence of cal-
cium chloride (D) but not with EDTA ( • ) . (C) Same experiment with
urine concentrate from patient G.I.N. The receptor peak is significantly
lower than that observed in B.
because Seetharam et al. have previously shown that I F -
Cbl receptor from ileum and kidneys are immunologi-
cally identical. S'13 In addition, the same group has ob-
served a defective brush border expression of I F - C b l
receptor in both kidney and ileal tissues in canine inher-
ited intestinal malabsorption and suggested that the de-
Cbl complex in gel filtration and isoelectrofocusing. W e
have previously shown that the ~ subunit of the receptor,
which shares the binding site for IF, is a sialylated N-
glycosylated protein. 19 The receptor sialylation was not
significantly modified in Gr~isbeck-Imerslund disease;
| the isoelectric point of the I F - C b l receptor complex was
00
not changed. The acidic isoelectric point was caused in
part by sialylation of the I F - C b l receptor complex as
neuraminidase digestion produced a shift from 4.82 to
5.42 to 5.46 in control and Gr~isbeck-Imerslund disease
sample, respectively.
Fyfe et al. recently described inherited selective intes-
tinal Cbl malabsorption in giant Schnauzer dogs and
concluded that this canine disorder resembles Griisbeck-
Imerslund disease] They observed, in ileal brush-border
membranes of affected dogs, an I F - C b l receptor activity
30 times lower than that of controls. 8 However, the I F -
Cbl receptor activity of ileal homogenate was three to
four times higher than in controls. They concluded that
the I F - C b l receptor was retained intracellularly. Such a
defect of intracellular trafficking does not fit with our
data; we observed decreased activity in both ileal homog-
enates and urine concentrates.
The present data represent the first report of soluble
I F - C b l receptor in urine. The presence of soluble I F -
Cbl receptor in urine apparently corresponds to the re-
June 1995 GR,~SBECK-IMERSLUND DISEASE 1627

(57Co)-Cbl
CPM 4 84 pH (57Co)-Cbl pH
• CPM
50 12 60 "1 ! ~ 5.42 r 14

40

30
10 50

t- f'F 12

20

10
6

2
30

20

1
1 tL,/ Ii°
0 ~ i

0 0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Fraction number Fraction number

(57Co)-Cbl (57Co)-Cbl
CPM pH CPM pH
30 C 80 14
4.84
12 12
60
20 10 10

40 8

10 6
2O

0 0 2
0 20 40 60 80 100 120 0 20 40 60 80 100 120

Fraction number Fraction number

Figure 3. Isoelectrofocusing of urinary IF receptor cross-linked to IFlabeled Cbl (IF-[~rCo]Cbl), The experiment was performed with receptor
from (A and B) a control and (C and D) patient V.E,R. in both cases, the isoelectric point was estimated at 4.84 in absence of neuraminidase
(A and C). The isoelectdc point was (B) 4.42 in control and (D) 4.46 in patient V.E.R. after treatment with neuraminidase,

moval by proteolytic cleavage of the extracellular domain
of the receptor from kidney epithelial cells; we observed
increased receptor activity in urine specimens from pa-
tients with tubular nephropathy (Table 2). This hypothe-
sis is consistent with previous data from the group of
Seetharam et al., who have detected and isolated the
receptor in mammalian kidney. 8'13 In addition, de novo
synthesis of the receptor has been shown in renal epithe-
lial cells with an exclusive targeting m the apical brush-
border membrane. .5 The determination of IF activity in
urine specimens may be an easy, convenient method of
diagnosing defective expression of IF receptor in epithe-
lial ceils. Additional studies are needed to evaluate its
diagnostic value.
1628 GUEANT ET AL.
References
1. Fenton WA, Rosenberg LE. Inherited disorders of cobalamin
transport and metabolism. In: Scriver CR, Beaudet AL, Sly SW,
VaNe DV, eds. The metabolic basis of inherited disease. 6th ed.
New York: McGraw-Hill, 1989:2065-2082.
2. Gr~sbeck R, Gordin R, Kantero I, Kuhlback B. Selective vitamin
B12 malabsorption and proteinuria in young people. Acta Med
Scand 1960; 167:289-296.
3. Imerslund O. Idiopathic chronic megaloblastic anemia in children.
Acta Paediat 1960;49(Suppl):1-115.
4. Gr~sbeck R. La malabsorption cong6nitale et s61ective de la
vitamine B12 avec prot6inurie. Cah CoN Med H6p Paris
1967;8:935-944.
5. Mackenzie IL, Donaldson RM, Trier JS, Mathan VI. Ileal mucosa
in familal selective vitamin B12 malabsorption. N Engl J Med
1972; 286:1021-1025.
6. Burman JF, Jenkins WJ, Walker-Smith JA, Phillips AD, Sourial NA,
Williams CB, Mollin DL. Absent ileal uptake of IF-bound vitamin
B12 in vitro in the Imerslund-Gr~sbeck syndrome (familial vita-
min B12 malabsorption with proteinuria). Gut 1985;26:311-
314.
7. Fyfe JC, Giger V, Hall CA, Jezyk PF, Klump SA, Levine JS, Pat-
terson DF. Inherited selective intestinal cobalamin malabsorption
and cobalamin deficiency in dogs. Pediatr Res 1990; 29:24-31.
8. Fyfe JC, Ramanujam KS, Ramaswany K, Patterson DF, Seeth-
aram B. Defective brush-border expression of intrinsic factor-
cobalamin receptor in canine inherited intestinal cobalamin mal-
absorption. J Biol Chem 1991;266:4489-4494.
9. Gu6ant JL, Schohn H, Brul6 H, Saunier M, Gr~sbeck R, Nicolas
JP. Binding assay and physicochemical characteristics of solubi-
lized intrinsic factor receptor in ileal mucosa homogenates using
phenyI-Sepharose to separate the saturated receptor from free
intrinsic factor. Biochim Biophys Acta 1991; 1073:614-618.
10. Schohn H, Gu6ant JL, Leheup B, Saunier M, Grignon G, Nicolas
JP. Intrinsic factor receptor during fetal development of the hu-
man intestine. Biochem J 1992;286:153-156.
GASTROENTEROLOGY Vol. 108, No. 6
11. Gu6ant JL, Jokinen O, Schohn H, Monin B, Nicolas JP, Gr&sbeck
R. Purification of intrinsic factor receptor from pig ileum using
as affinity medium human intrinsic factor covalently bound to
Sepharose. Biochim Biophys Acta 1989;992:281-288.
12. Gu~ant JL, Masson C, Schohn H, Girr M, Saunier M, Nicolas JP.
Receptor-mediated endocytosis of the intrinsic factor cobalamin
complex in HT 29, a human colon carcinoma cell line. FEBS Lett
1992;297:229-232.
13. Seetharam B, Levine JS, Ramasamy M, Alpers DH. Purification,
properties and immunochemical localization of a receptor for
intrinsic factor-cobalamin complex in the rat kidney. J Biol Chem
1988; 263:4443-4449.
14. Ramanujam KS, Seetharam S, Ramasamy M, Seetharam B. Re-
nal brush border membrane bound intrinsic factor. Biochim Bio-
phys Acta 1990; 1030:157-164.
15. Seetharam S, Ramanujam KS, Seetharam B. Synthesis and
brush border expression of intrinsic factor-cobatamin receptor
from rat renal cortex. J Biol Chem 1992;267:7421-7427.
16. Ramanujam KS, Seetharam S, Dahms N, Seetharam B. Func-
tional expression of intrinsic factor-cobalamin receptor by renal
proximal tubular epithelial cells. J Biol Chem 1991; 266:13135-
13140.
17. Kouvonen I. Solubilization of the pig ileal intrinsic factor receptor
with papain treatment and studies on the solubilized receptor.
Biochim Biophys Acta 1980;626:244-253.
18. Seetharam B, Bagur SS, AIpers DH. Isolation and characteriza-
tion of preteolytically derived ileal receptor for intrinsic factor-
cobatamin. J Biol Chem 1982;257:183-189.
19. Jokinen O, Gu~ant JL, Schohn H, Gr~sbeck R. Lectin binding to
the porcine and human ileal receptor of intrinsic factor-cobala-
min. Glycoconjugate J 1989;6:525-538.
Received February 3, 1994. Accepted February 1, 1995.
Address requests for reprints to: Jean-Louis Gu~ant, M.D., D.Sc.,
Laboratory of Cellular and Molecular Nutrition, Medical Faculty of
Nancy, BP 184, 54505 Vandoeuvre les Nancy Cedex, France. Fax:
(33) 83-15-38-26.