_.

afiwm?
"Eng 5”“

J”. .54. 4. I»! «

.4: in... L3
xunuanrnmnwflflrs.
. . id‘sl. .-
fc?{r Ihufmg
V . . ulnflr‘w‘fi:ltafl.
, y . ‘ \l, A 1v
[A1, . . ‘
. .....xhuwh... V
humufitn
.. ‘

.
. . . «.1
aduFCJ‘ia

v: 5%“:
0...,

in

:«I 51., I . a: 4
.53.}. :5: t”.

Lama? than" r... .15... LA. I.

an...

 

..
a a.
:nae.

: 2.. ..
H.351; ,
15.13:,"-

.32
if...

 

a
i 3.???
3.
,7.-. .5:

713.1:

I! I... .61..
:1 ‘ ..
. “Pb:- a

... t.
.2;
L.

13,133.:
I 3 .

 

 

1.1.12}?
:33 5...... LI-

l:¥ll$.....5‘ )1?!

l? 1 I}...
. 0..

:5: .1!
1 3E.

 

 

 

 

:._ ..A I» no: u .l ~4m
‘ .. , . r . .fi. .. HF team"; ~35...
Enxxfikr. E agghczfiwaaé E {FR 6

 

THESIS

K7
V

2550

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

CLUM/vé. AND CHARACTZ>RIZATIDM 0F CANINE
TNTRJN§lC FACTDR- (Loam/Maw RECEPTOR/cueiLw

presented by

’DANBt/V XH

has been accepted towards fulfillment
of the requirements for

‘DDCTDR degreein PHILOSOPHY

 

 

SL—C/W

V Major p‘rofessg

{1- (#00

Date

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepes-p. 15

 

CLONING AND CHARACTERIZATION OF CANINE INTRINSIC FACTOR-
COBALAMIN RECEPTOR I CUBILIN

By

Danbin Xu

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Microbiology

2000

ABSTRACT

CLONING AND CHARACTERIZATION OF CANINE INTRINSIC FACTOR-
COBALAMIN RECEPTOR I CUBILIN
By

DANBIN XU

Cubilin is a high molecular weight multiligand receptor that mediates
intestinal absorption of intrinsic factor-cobalamin and selective protein
reabsorption in renal tubules. The genetic basis of selective intestinal cobalamin
malabsorption with proteinuria was investigated in a canine model closely
resembling human lmerslund-Grasbeck syndrome caused by cubilin mutations.
Canine cubilin CDNA of 11,282 bp was cloned and sequenced, including 73 bp of
5’ and 349 bp of 3’ untranslated sequence. The open reading frame of 10,860 bp
corresponds to a deduced amino acid sequence of 3,620 residues and
demonstrates an overall amino acid identity of 83% with human and 70% with rat
cubilin. Further characterization of tissue-specific cubilin expression and post-
translational modification with emphasis on the gastrointestinal tract was carried
out. Intrinsic factor-cobalamin binding activity, cubilin immunoreactivity, and
cubilin mRNA levels were determined in multiple segments of canine
gastrointestinal mucosa and other tissues. The parallel aspects of cubilin
expression suggested that the major determinant of regional cubilin expression in

the gastrointestinal tract was modulation of cubilin mRNA. Cell fractionation

indicated that ileal cubilin is not strongly membrane-associated. An ~185 kDa
brush-border specific and two >400 kDa precursor forms of cubilin were identified
in ileum, but only an ~460 kDa form in renal cortex. Asparagine-linked
oligosaccharide modifications characterized by differential glycosidase digestion
of affinity purified cubilin from ileal mucosa and renal cortex. Neal and renal
intracellular cubilin comigrated on SOS-PAGE at ~4OO kDa after oligosaccharide
removal, suggesting that the size differences were caused by different
glycosylation.

lntragenic markers were identified within introns for both cubilin and
megalin genes in the canine family and used to test the hypothesis of genetic
linkage of the disease to cubilin or megalin loci. Megalin, a transmembrane
receptor interacting with cubilin in vitro and colocalizing with cubilin in vivo, is
speculated to anchor cubilin onto apical membranes of ileal enterocytes. Linkage
of either gene was rejected, indicating that the canine disorder resembling
Imerslund-Grasbeck syndrome is caused by defect of a gene product other than
cubilin or megalin. These results imply that there may be locus heterogeneity
among human kindreds with selective intestinal cobalamin malabsorption and
proteinuria and that normal brush-border expression of cubilin requires the

activity of an accessory protein other than megalin.

This dissertation is dedicated to:

Qiang Tian, my husband and best friend, for his deep love, care and support
through all my scientific growth.

John Fyfe, my mentor and friend, for his patience, encouragement and
understanding during my graduate studies.

iv

ACKNOWLEDGMENTS

I am forever grateful to the large number of people who gave me advice,
assistance and encouragement during the completion of my Ph.D. I feel fortunate
to have had the support and friendship of so many wonderful people.

I am especially grateful to the members of my graduate committee, Dr.
Jerry Dodgson, Dr. Ronald Patterson, Dr. Richard Schwartz and Dr. Laura
McCabe, for their guidance, instruction and encouragement throughout this
research work, for the valuable time and comments on this dissertation.

I want to thank Rebeccah Kurzhals and Dr. Mary Lassaline for their
friendship and the help for taking care of the dogs. I am also thankful to Dr.
Karen Friderici and people in her lab for the use of laboratory equipment,
supplies, and technical knowledge. I also owe thanks to several members of the
Department of Microbiology, Dr. Walt Esselman, Angie Zell, and Cheryl Akers for
their help during my graduate studies. I also thank my fellow students for their
friendship, and support during my years at Michigan State University.

I like to thank Dr. Pierre Verroust, Dr. Renata Kozyraki in France for
providing the rat cubilin CDNA, and Dr. Paula Henthorn, Dr. Jianlong Liu in
University of Pennsylvania for providing the technical support.

Special thanks also go to my wonderful family, my parents and my sister
for their unconditional support and encouragement throughout all my endeavors
in life. A very special thank due to my lovely daughter, Elaine, for the cheerful

moment she gave me when I had hard time for the research.

TABLE OF CONTENTS

LIST OF FIGURES .......................................................................................... viii
CHAPTER 1
LITERATURE REVIEW ........................................................................................ 1
Overview of normal cobalamin absorption ................................................... 1
Cobalamin malabsorption and Imerslund-Gréisbeck syndrome .................. 7
Human lmerslund-Grasbeck syndrome .......................................................... 7
Canine model with selective cobalamin malabsorption ................................... 9

Characterization of intrinsic factor-cobalamin receptor (IFCR)Icubilin ....11

General characteristics of IFCR I cubilin ....................................................... 11
Molecular cloning of human and rat cubilin ................................................... 14
Functional domains of cubilin ......................... - ............................................... 1 9
Regulation of cubilin expression in small intestine .................................... 21
Protein folding in the secretory pathway ..................................................... 24
General outline for this thesis ...................................................................... 28
References ..................................................................................................... 30
CHAPTER 2
CLONING OF CANINE CUBILIN ....................................................................... 43
Introduction .................................................................................................... 44
Materials and Methods .................................................................................. 46
Results ............................................................................................................ 50
Discussion ...................................................................................................... 55
References ..................................................................................................... 59
CHAPTER 3
CUBILIN EXPRESSION AND POST-TRANSLATIONAL MODIFICATIONS IN
THE CANINE GASTROINTESTINAL TRACT .................................................... 61
Introduction .................................................................................................... 62
Materials and Methods .................................................................................. 65
Results ............................................................................................................ 74
Discussion ...................................................................................................... 94

vi

References ................................................................................................... 103

CHAPTER 4

LINKAGE ANALYSIS OF CANINE SELECTIVE COBALAMIN

MALABSORPTION WITH CUBN AND MEGALIN LOCI ................................. 107
Introduction .................................................................................................. 108
Materials and Methods ................................................................................ 1 1 1
Results .......................................................................................................... 1 16
Discussion .................................................................................................... 123
References ................................................................................................... 127

Future Direction .............................................................................................. 130

vii

LIST OF FIGURES

Figure 1.1 Canine intrinsic factor-mediated cobalamin absorption .............. 2
Figure 1.2 Predicted protein structure of cubilin ........................................... 15

Figure 1.3 Schematic representation of the interaction

between cubilin and megalin ...................................................... 19
Figure 2.1 CDNA cloning of canine kidney cubilin ........................................ 51
Figure 2.2 Primary sequence of canine cubilin

as deduced from CDNA clones ................................................... 52
Figure 3.1 Tissue specificity of cubilin mRNA expression ............................ 75

Figure 3.2 Codistribution of intestinal cubilin function,
immunoreactive proteins, and mRNA ......................................... 76

Figure 3.3 Codistribution of quantitated intestinal
cubilin ligand-binding activity and mRNA .................................... 80

Figure 3.4 lmmunoreactive and lF-Cbl binding cubilin
species in fractionated canine ileal mucosa
and renal cortex .......................................................................... 83

Figure 3.5 Western blot of canine cubilin in reducing
and non-reducing conditions ...................................................... 85

Figure 3.6 Distribution of immunoreactive cubilin forms
in different centrifugation fractions of dog
ileal homogenate ........................................................................ 87

Figure 3.7 Oligosaccharide analysis of affinity purified
ileal and renal cubilin .................................................................. 88

Figure 3.8 Northern blot analysis of canine cubilin and

CRP-ductin along the crypt-villus axis in dog ileum .................... 90
Figure 3.9 Alkaline phosphatase activity along the crypt-

villus axis in dog ileum ................................................................ 91
Figure 4.1 Independent segregation of canine I-GS and CUBN loci ............ 117

Figure 4.2 lntragenic cubilin marker is located in a
unique region of the canine genome .......................................... 118

viii

Figure 4.3 Independent segregation of canine l-GS and megalin loci ......... 120

ix

CHAPTER 1
LITERATURE REVIEW

Overview of normal cobalamin absorption

Cobalamin (Cbl) is a complex organometallic substance consisting
of a corrin ring, a central cobalt atom, and various axial ligands. The basic
structure, known as vitamin B12, is synthesized exclusively by microorganisms. ln
non-ruminant animals, bacterial synthesis takes place only in the large bowel and
the cecum, and from these sites absorption cannot take place. Therefore,
absorption of vitamin B12 is entirely from dietary sources (Booth and Mollin,
1959), including fecal contamination or coprophagy. Cbl is an essential
micronutrient required for the health and well being of all higher animals,
including human. All of these animals are capable of converting the vitamin into
the two required coenzyme forms, adenosyl-cobalamin (AdoCbl) and
methylcobalamin (Merl), which are essential cofactors for methylmalonyl-CoA
mutase and methionine synthetase, respectively, in intermediate metabolism.

Cobalamin absorption, which is a complex process comprised of

sequential protein-binding events, begins with a digestive phase (Figure 1.1).
The vitamin is bound predominantly by haptocorrin (HC, R protein) in gastric
juice and becomes bound to intrinsic factor (IF) after the HC moiety is partially
degraded by pancreatic proteases in the duodenum (Allen et al., 1978; Carmel et
al., 1983). IF is a Cbl-specific-binding protein classically synthesized in man by
the gastric parietal cells of the fundus and body of the stomach (Christensen et

al., 1973; Levine et al., 1980). IF is synthesized by various cells and glands

Target Tissues
Cbl

II"

IFCR

TC-II

ac Stomach

D

b7 IF-Cbl Complex
Y

V

O

 

Bi] Degraded IF

. D 3 Pancreas

Protease:

  
  

Q9 Degraded ac D

’ a ’ 1' Undefined Hechanim

 

 

 

 

Blood Side

Figure 1.1 Canine intrinsic factor-mediated gastrointestinal cobalamin absorption.
Provided by Dr. John Fyfe. '

derived from foregut depending on the different species (Gordon, 1989; Festen,
1991). IF is produced by the gastric mucosa of humans, pigs, and rats but is
produced mainly in pancreatic duct cells in dogs (Batt and Horadagoda, 1989;
Simpson et al., 1989; Vaillant et al., 1990; Simpson et al., 1993) and pancreas in
cats (Fyfe, 1993). IF synthesized by different species showed different cross-
species immunoreactivity (Hooper et al., 1973). Functional IF secretion and IF
mRNA levels are developmentally regulated (Ramaswamy et al., 1989;
Dieckgraete et al., 1988). IF secretion is also tissue specifically regulated, being
activated by pentagastrin in the human stomach and by cholecystokinin in the
canine pancreas after eating (Simpson et al., 1993), but the correlation between
IF secretion and the expression of other Cbl-binding proteins is still unknown.
The lF-Cbl complex is highly resistant to proteolysis. This enables it to
pass intact through the small bowel. Physiological cobalamin absorption takes
place in the distal small intestine (Booth and Mollin, 1959), which has specific
receptors for the IF-Cbl complex on the microvillous membranes of mucosal cells
(Donaldson et al 1967; Levine et al., 1982). Attachment of the IF-Cbl complex to
the receptor requires the presence of divalent cations (notably calcium)
(Mackenzie et al., 1972) and a pH between 6 and 9 (Cooper and Castle, 1960;
Herbert and Castle, 1961) but is independent of temperature and energy
(Donaldson et al., 1967). Intrinsic factor-cobalamin receptor (IFCR), a membrane
protein of the ileal brush border, mediates intestinal uptake and initiates
transcytosis of dietary and biliary Cbl across the mucosal barrier. The number of

receptors in the intestine seems to be very low. The greatest density of ileal

receptor sites is exhibited in dogs and pigs, being approximately 0.3 x1012 to 4.9
x 1012 binding sites/ gram mucosa (Hooper et al., 1973). The restricted
absorption of cobalamin is partly explained by the small number of receptors.
IFCR is specific for IF-Cbl complex while free cobalamin, free IF and Cbl
attached to other Cbl binding proteins cannot bind the receptor (Hooper et al.,
1973; Katz and cooper, 1974). There is an enterohepatic cycle of Cbl absorption.
The daily amount of Cbl excreted in bile constitutes approximately 0.1 to 0.19%
of the total amount stored in the human body (Nicolas JP and Guéant JL, 1995).
Pancreatic enzymes are responsible for the degradation of biliary R protein
leading to the transfer of the released Cbl to biologically active IF and eventually
to Cbl absorption in the terminal ileum (El Kholty et al., 1991).

Following the adsorption of the IF-Cbl complex to the ileal receptor there is
a delay of 3 to 4 hours before the vitamin appears in the blood (Doscherholmen
et al., 1959; Fyfe et al., 1991a). The processes involved in the translocation of
cobalamin from the ileal receptor to the blood are still largely unknown, though
some knowledge has accumulated. The receptor mediated endocytosis of IF-Cbl
complex occurred within 30 min after the complex attached to its ileal surface
receptor (Kapadia et al., 1983). In vivo and in vitro studies with 57Co-Iabeled
cyanocobalamin and [35$]methionine-Iabeled IF showed that following
endocytosis, the IF-cobalamin complex was released from the receptor and
transferred to the lysosomal or prelysosomal acidic vesicles, where IF was
degraded with a half-life of 4 hours (Kapadia et al., 1983; Ramanujam et al.,

1991a, Guéant et al., 1992; Dan at al., 1994; Bim et al.,1997). Chloroquine or

cycloheximide treatment reduced the transport of Cbl suggesting release of Cbl
from IF required acid pH, that there was a series of compartments involved in the
transport of Cbl, and that new protein synthesis was required (Robertson JA.,
1985a; Robertson JA, 1985b; Ramaswamy et al., 1989) The receptor of lF-Cbl
complex is rapidly recycled onto the brush-border membrane after endocytosis
(Dan and Cutler, 1994; Le Panse et al., 1997), and the liberated Cbl exported
from Iysosomes via a specific transporter ( that malfunctions in Cbl F) to
transcobalamin ll (TC II). The location for this transport is still an open question.
Some evidence suggested that Cbl released from IF and transfered to TC II
within the rat enterocytes (Robertson and Gallagher, 1985b; Ramaswamy et al.,
1989; Ramanujam et al., 1991a). But other researchers questioned if the crude
homogenate used for TC lI-[57Co]CbI immunoprecipitation can provide sufficient
discrimination between the tissue components of the ileum (Quadros et al.,
1999). TC II is a specific Cbl transport protein which has been detected in many
cell types in culture (Hall et al., 1985) and in human and rat tissues (Li et al.,
1994). In dogs, all the plasma Cbl is bound to TC-Il (Rappazzo and Hall, 1972).
In contrast, in human, only newly absorbed Cbl is bound to TC-ll, and most
circulating Cbl is bound to TC-l which is haptocorrin (HC) (Gullberg R., 1972).
This distinction is important since, in dogs, TC-II deficiency would be expected to
have extremely low serum Cbl concentration, but in human TC-II deficiency,
serum Cbl concentrations are similar to normal due to Cbl binding to HC (Fenton

and Rosenberg, 1989). The canine TC II has been studied

and showed similar molecular size , cobalamin binding activity, and immunologial
reactivity as in human (Rappazzo and Hall, 1972).

TC Il-mediated uptake of plasma Cbl has been shown in a number of cells
(Hall 1975; Retief et al., 1966; YoungdahI-Turner et al., 1978), and Cbl is
transported into cells via a plasma membrane TC II-Cbl receptor (TC ll-R)
expressed in many tissues (Bose et al., 1995a; Bose et al., 1995b). After
endocytosis, the TC Il-Cbl complex dissociates in endosomes or Iysosomes, and
free Cbl is released (Idriss et al., 1991). The free Cbl then enters the cytoplasm
where it is reduced and converted to methyl-Cbl, or in mitochondria, where it is
reduced and converted to adenosyl-Cbl. Methyl-Cbl and adenosyI-Cbl are used
as cofactors for methionine synthase and methylmalonyl CoA mutase,
respectively (Millman et al., 1977, Kolhouse and Allen, 1977). In addition to TC II-
R, whose role in plasma transport of Cbl is well established, another protein
known as megalin has been shown to bind to TC Il-Cbl in vitro and mediate its
endocytosis in BN/MSV cells (derived from yolk sac carcinoma) and perfused
renal proximal tubules (Moestrup et al., 1996). The physiological significance of
megalin-mediated TC II-Cbl uptake system in the plasma transport of Cbl is not
certain. Due to the high expression in the apical membrane of kidney proximal
tubules, renal megalin may function in the tubular reabsorption of TC II-Cbl, thus

preventing Cbl loss in the urine.

Cobalamin malabsorption and lmerslund-Grfisbeck syndrome

Due to the widespread distribution of cobalamin in animal products
(Sullivan LW, 1970; Smith A, 1962), Cbl deficiency usually occurs secondary to
inherited or acquired defects of gastrointestinal Cbl absorption rather than dietary
insufficiency. Inherited disorders of vitamin B12 include those in which the vitamin
is not utilized by target cells, and those which are defective in absorption of the
vitamin from the gut and transport to the appropriate tissues. The former includes
a defect in the release of free vitamin B12 from Iysosomes (CbIF) (LaFramboise et
al., 1992) and defects in the formation of one or both vitamin B12 cofactors (CbIA-
E). Defects of absorption and transport include intrinsic factor abnormalities
(Spurling et al., 1964; Katz et al., 1974b; Yang et al., 1985), deficiencies of TC II

(Hakami et al., 1971) Cbl F, and selective intestinal Cbl malabsorption.

Human lmerslund-Grfisbeck syndrome

Selective intestinal Cbl malabsorption is also called Imerslund-Grasbeck
syndrome (I-GS) which was independently described by Imerslund and Grasbeck
in 1960 (Grasbeck et al., 1960, Imerslund 0., 1960). lmerslund’s ten patients
were from six different families in south-east Nonlvay, while Grasbeck et al.
reported two cases from Finland. The patients examined showed Cbl-responsive
anemia and most of them also had mild proteinuria. Following thorough
investigation, all known causes of Cbl deficiency were eliminated, including

pathological intestinal flora, inactive intrinsic factor, and deficiency of

transcobalamin I or II. Both sexes were affected, and the inheritance was
deemed recessive. The similarity of the two sets of cases suggested that they
represented the same disease. Since then, totally 38 cases in Finland and 15
cases in Norway have been identified with selective Cbl malabsorption (Aminoff
et al., 1995). More cases have also been reported in other countries. By 1969, 20
cases with l-GS had been diagnosed in Israel, and proteinuria showed
predominantly albumin (Ben-Bassat et al., 1969). Other scattered cases were
also described from the united state (Colle et al., 1961; Spurling et al., 1964),
Denmark (Sievens CJ, 1964), Australia (Marsden et al., 1979), Turkey (Yetgin et
al., 1983; Altay et al., 1995; 1999; Celep et al., 1996), and Saudi Arabia
(Abdelaal and Ahmed, 1991). More recently a single case has also been
reported in an Africa (Stones and Ferreira, 1999). By 1999, there were over 250
cases reported in the literature. The syndrome occurs worldwide, but its
prevalence is higher in several Middle Eastern countries and Norway, and
highest in Finland. The onset of signs of this disease is between 0 to 5 years of
age, which mainly corresponds to the expected period of depletion of the child’s
vitamin B12 which was stored prenatally by transfer from the maternal blood. The
clinical features of this syndrome include failure to thrive, infections,
megaloblastic anemia, neuropathy and mild general malabsorption. Mild
proteinuria is common. All the clinical features except Cbl malabsorption and
proteinuria can be eliminated by parenteral, but not oral, Cbl administration
(Broch et al., 1984). Linkage analysis of 38 patients diagnosed in Norway and

Finland localized the unknown autosomal recessive megaloblastic anemia (MGA

1) gene to the short arm of chromosome 10 between markers D108548 and
D1OS466 (Aminoff et al., 1995).

Due to the clinical heterogeneity of this syndrome among the affected
families and complexity of the process of enterocyte transcytosis of Cbl, different
putative molecular natures of Imerslund-Grasbeck syndrome have been
reported. One study showed lack of pathology by electron microscopy of ileal
biopsies and normal lF-Cbl binding in mucosal homogenates and suggested the
cause of this disease has a defect in intracellular trafficking of Cbl rather than a
deficient synthesis of the receptor (Mackenzie et al., 1972). Another study in a
different family showed defective uptake of Cbl by ileal mucosa in vivo (Burman
et al., 1985). The difference between these two studies may be due to the
different genetic defects, but is also possibly due to the same genetic defect but

different study design.

Canine model with selective cobalamin malabsorption

A similar recessively inherited syndrome in giant schnauzer dogs has
been reported, in which affected puppies exhibited inappetance and failure to
thrive beginning between eight and twelve weeks of age. (Fyfe et al., 1989).
Clinical laboratory tests demonstrated megaloblastic anemia, low serum Cbl
concentrations, and mild proteinuria. Parenteral, but not oral, administration of
cyanocobalamin corrected all clinical and laboratory abnormalities except
proteinuria and Cbl malabsortption. Gastrointestinal cobalamin malabsorption

was demonstrated by oral radiolabeled cobalamin administration, and

biochemical characterization of IF and TC II were normal (Fyfe et al., 1991a).
Expression of IFCR was investigated in ileum and renal cortex of affected dogs
by immunoelectron microscopy and cell fractionation, and the results
demonstrated that there was no expression of IFCR in the apical brush-border
membrane of either tissue (Fyfe et al., 1991b). Receptor-ligand binding assay
showed IFCR activity in total homogenates was 3-4 fold higher than normal in
ileal mucosal and one-tenth of normal in kidney of affected dogs. These data
suggested that inefficiently expressed receptors accumulate in ileum but were
rapidly degraded in kidney of affected dogs (Fyfe et al., 1991 b).

Brush-border expression of IFCR is critical for intestinal IF-Cbl absorption.
The defective functional expression of IFCR onto the apical membrane raised the
consideration that IFCR may be a candidate gene for MGA 1 in this family of
dogs. The most common reason for disfuntional membrane expression of a
membrane protein is mutation of the coding sequence, which results in abnormal
protein folding (discussed later). Based on this phenomenon, we hypothesized
that mutation of IFCR was the cause of selective intestinal Cbl malabsorption in
this family of dogs.

The similarity of the clinical and laboratory features between human
patients and the dog family with selective intestinal Cbl malabsorption suggested
that this would be a good animal model for study of this syndrome. Due to the
easy treatment of the disorder in humans, experimental samples are extremely
hard to obtain from patients, greatly limiting investigation of the molecular

mechanism of this disease. Biochemical study of IFCR in the affected dogs first

demonstrated the candidate gene for MGA 1. Further study with this model will
enhance understanding of this syndrome and elucidate mechanisms of the

cellular processing pathway of newly synthesized proteins.

Characterization of intrinsic factor-cobalamin receptor (IFCR)Icubilin

General characteristics of IFCR I cubilin

Clearly IFCR plays a key role in gastrointestinal Cbl absorption. Research
has been done during the past over 25 years to further understand the function
and molecular nature of this receptor. Solubilization of IFCR from distal small
intestine has been carried out in guinea pig and pig (Katz et al., 1974a; Cotter
and Rothenberg et al., 1976). Solubilized receptor had similar IF-Cbl binding
characteristics to membrane-bound receptor, suggesting the functional domain of
this receptor resided on the Iumenal side of the enterocyte membrane.

Canine ileal IFCR purified by affinity chromatography revealed a 180 kDa
protein on non-reducing SDS-PAGE, and 2-mercaptoethanol treatment produced
two bands with apparent molecular weights of 59 and 42 kDa (Seetharam et al.,
1981). Papain digestion prior to purification resolved IFCR into smaller subunits
of the receptor which maintained IF-Cbl binding activity, both for dogs
(Seetharam et al., 1982) and pigs (Kouvonen I, 1980; Kouvonen and Grasbeck,
1980). Further cell fractionation of canine ileal enterocytes revealed an ~185 kDa
lF-Cbl binding protein on brush-border membrane (Fyfe et al., 1991b). However,
functional IFCR in the Caco-2 cell line, derived from a human colon carcinoma,

was reported as a 230 kDa protein (Ramanujam KS, 1991a) which does not

dissociate into subunits upon disulfide bond reduction. This suggested that,
instead of having a subunit structure, IFCR undergoes proteolytic cleavage once
exposed to the ileal lumen with the resulting fragments being held together by
disulfide bonds.

Besides the distal part of the small intestine, high levels of receptor for IF-
Cbl were detected in human, canine, and rat kidney. The ratio of specific activity
for kidney relative to intestine in these species was 116, 20 and 797, respectively
(Seetharam et al., 1988). In vivo (Seetharam et al., 1992) and in vitro
(Ramanujam et al., 1991b, 1992) functional studies of renal IF-Cbl receptor
showed the same binding activity and immunoreactivity as the ileal receptor.
Both activities were detected in the apical membrane of proximal tubule cells but
not in distal tubule cells of the kidney. The size determined from SDS-PAGE for
the renal receptor was reported as 230 kDa and was not reduced into subunits
by reductive alkylation (Seetharam et al., 1988; Ramanujam et al., 1993a; Fyfe et
al., 1991b). Quantitative amino acid analysis of the renal receptor gave a value of
457,310 g of amino acid / mol of IF-Cbl binding activity but was thought to
represent a dimer (Seetharam et al., 1988).

All the ileal and renal IFCR data together suggested that IFCR undergoes
proteolytic cleavage once exposed to the ileal lumen rather than having a subuit
structure. This hypothesis is based on the assumption that ileal and renal IFCR
are the same gene products. lF is not thought to enter the circulation system
during ileal endocytosis. The presence of IFCR in the kidney suggested that

renal IFCR had functions other than renal lF-Cbl reabsorption.

12

In the meantime, the study of an independent group described a 280-kD
protein (gp280), defined by monoclonal antibodies raised against rat renal brush-
border, which was found in the intennicrovillar domain of the proximal tubule cells
of kidney and rat yolk sac epithelium (Sahali et al., 1988). Antibodies to gp280
induced fetal malformations via trapping of the target antigen in the early
endocytic compartment thus preventing its normal function in lysosomal transfer
(Le Panse et al., 1994; 1995). The function of gp280 was thought to be a
receptor, but its ligand(s) was not known until 1997. Rat renal gp280, like the
IFCR, binds to the IF-Cbl complex with an association constant of 0.3 x 109 M'1
and mediates its internalization (Seetharam et al., 1997). In addition, antibodies
against purified gp280 and IFCR can inhibit the binding of lF-[57Co]Cbl complex
to intestinal, renal, and yolk sac apical membranes and revealed a single
identically sized protein on immunoblotting of renal membranes, suggesting
functional and immunological identity of the two proteins. The similar mobility on
SDS-PAGE and restricted tissue distribution of gp280 and IFCR suggested the
possibility that IFCR and gp280 were similar or identical proteins.

IFCR/gp280 was then considered an over 200 kDa protein facilitating the
intestinal endocytosis of IF-Cbl and renal reabsorption of some unidentified
proteins. In late 1997, using receptor-associated protein (RAP) affinity
chromatography, an lF-Cbl binding renal epithelial protein of ~460 kDa was co-
purified with the TC Il-Cbl binding 600 kDa receptor, megalin (Birn et al., 1997).
The IF-Cbl binding activity was further confirmed by IF-Cbl affinity

chromatography, which eluted the same ~460 kDa protein from rabbit and human

13

kidney homogenate. In vitro binding analysis demonstrated a calcium-dependent
high affinity binding of lF-Cbl to a site distinct from the RAP binding site.
Electrophoretic mobility and immunoreactivity was compared between rat gp280
and rabbit ~460 kDa IF-Cbl-binding protein, and the two proteins showed
identical SDS-PAGE migration and immunoreactivity. Microscopic
autoradiography revealed the ~460 kDa protein in both kidney and terminal ileum
and in vivo microinjection demonstrated the facilitation of the endocytosis of IF-

Cbl by this membrane protein.

Molecular cloning of human and rat cubilin

The molecular characterization of this ~460 kDa protein (IFCR/gp280) was
not solved until its recent cDNA cloning from rat and human (Moestrup et al.,
1998; Kozyraki et al., 1998). The cDNA clone corresponds to a single transcript
of 11.6 kb in intestine, kidney and yolk sac tissues in rat, where IFCR/gp280 is
expressed. Using fluorescence in situ hybridization, radiation hybrid mapping and
screening of YAC clones, the human cDNA clone was mapped on the short arm
of chromosome 10 (10p12-p14) (Kozyraki et al., 1998), which is in the same
region as human MGA 1. These results suggested that this ~460 kDa protein
was the candidate disease-causing gene product for selective intestinal Cbl
malabsorption in some patients. All these data showed that the newly purified
protein had the same ligand binding activity, tissue distribution, and
immunoreactivity, but had different molecular weight compare to IFCR/gp280.

Since two very well characterized markers bigger than 460 kDa were used in the

14

most recent study (Birn et al., 1997), the previous over 200 kDa molecular weight
for IFCR/gp280 was possibly misjudged by relying on migration on SDS-PAGE
with 200 kDa as the highest molecular weight marker.

The assembled cDNA revealed an uninterrupted open reading frame of
10.8 kb and a deduced 3621-3623 amino acid protein for human and rat,
respectively. The overall homology between the encoded proteins of these two
species is 69%. The predicted domain organization of this protein contains an N-
terrninal stretch of approximately 110 amino acids with no apparent homology to
known proteins, a cluster of eight EGF (epithelial growth factor) type B repeats
and 27 contiguous CUB domains (Moestrup et al., 1998; Kozyraki et al.,
1998)(Fig 1.2). Due to the unique tandem arrangement of CUB domains in the
protein sequence, the ~460 kDa protein (IFCR/gp280) has been designated as

cubilin, which will be used for the remainder of this thesis.

8 EGF Repeats 27 CUB Domains

WI

Figure 1.2 Predicted protein structure of cubilin

 

More evidence suggests that the protein expressed in intestine and kidney
are the same gene product. An early study (Robin et al., 1974) reported that a
patient with IG-S had an inherent renal proximal tubular defect in amino acid
reabsorption which was not Cbl deficiency related. The patients with selective

Cbl malabsorption with proteinuria have reduced urinary IFCR activity compared

15

to the normal control (Dugue et al., 1998). Further genetic studies with 17 Finnish
families with selective intestinal Cbl malabsorption identified two independent
disease-specific mutations in the cubilin gene (CUBN) (Aminoff et al., 1999) .
One was a missense mutation (FM1) apparently in the IF-Cbl binding site
(Kristinasen et al., 1999). The other mutation was a C to G mutation activating a
cryptic intronic splice site and introducting multiple stop codons in the mutant
transcript (FM2). These mutations indicated that defects of cubilin in the intestine
caused Cbl malabsorption. The Finnish patient with the FM2 mutation also
secreted dramatic amounts of urinary alipoprotein A-l (apoA-l) which was not
detected in the normal control or FM1 patient (Kozyraki et al.,1999). ApoA-l was
also detected in urine of dogs (five of six) with selective intestinal Cbl
malabsorption, caused by disfunctional expression of IFCR on both intestinal and
renal epithelial cell membranes, but was not detected in normal control dogs.
These results elucidated that a null mutation in CUBN (FM2) affects cubilin
functions in both intestine and kidney and elegantly supports the hypothesis that
these two proteins are the same gene product.

The patient with apoA-1 proteinuria indicated that besides lF-Cbl complex,
apoA-1 is another ligand for cubilin which can be reabsorbed in the kidney. In
vitro studies, such as cubilin affinity chromatography and surface plasmon
resonance analysis, further demonstrated the high-affinity binding of apoA-I and
high-density Iipoprotein (HDL) to kidney cubilin (Kozyrak et al., 1999; Hammad et
al., 1999; 2000). Cubilin-expressing yolk sac cells showed efficient 125l-HDL

endocytosis that could be inhibited by antibodies against apoA-l and cubilin.

l6

Other ligands have also been defined for renal cubilin recently, such as myeloma
light chains (Batuman et al., 1998) and albumin (Birn et al., in press).

Both human and rat cubilin open reading frame begin with a 24-20 amino
acid signal peptide which facilitates binding of the translating ribosome to bind
the endoplasmic reticulum (ER) and starts the membrane protein biosynthetic
pathway. An Arg-X-Arg/Lys-Arg consensus sequence can be found following the
signal peptide in both sequences, which is a site cleaved by the trans-Golgi
proteinase furin (Kozyraki et al., 1998; Moestrup et al., 1998). Cubilin N-terminal
sequence starts just after furin site. This indicates that the receptor is
synthesized as a precursor undergoing further proteolytic processing in the trans-
Golgi. Two of the EGF repeats contain the discontinuous consensus sequence
for Cay-binding, which may be required for a particular conformation for ligand
binding. The CUB domains, named as a structural feature common to
complement subcomponents C1rlC1s, the Xenopus patterning protein, Uvs.2,
and Bone morphogenic protein-1, have ~110 residues with two disulfide bridges,
and bind a diverse array of protein, carbohydrate, and phospholipid ligands (Bork
and Beckmann, 1993). CUB domains are common to an increasing number of
developmentally expressed proteins, but none of the known proteins contains
more than 5 CUB domains. The extensive repetition of CUB domains provides
another indication that this ~460 KDa is a multiligand receptor with multiple
potential binding sites.

Except for the leader peptide, no sequence compatible with a

transmembrane domain or gchosylphosphatidylinositol (GPI) anchor-addition

17

signal could be identified in cubilin. This excludes the possibility that the protein
is a transmembrane protein or GPI anchored protein that is synthesized with a
cleavable hydrophobic C terminus. Early studies with guinea pig ileum
homogenate without detergent treatment showed the receptor is soluble and
maintained lF-Cbl binding activity (Cotter et al., 1976). Later study confirmed that
association between the cubilin and membranes is weak since approximately
50% of cubilin could be released into the fluid phase during simple mechanical
grinding without detergent or during incubation of membranes with heparin or
EDTA (Moestrup et al., 1998). All these data suggested that cubilin is a
peripheral membrane protein.

As a peripheral membrane protein without internalization signals, how
cubilin can facilitate the endocytosis of lF-Cbl complex is an open question. Co-
internalization of a receptor that lacks internalization signal(s) by means of
another receptor has been shown previously. Thus, the
gchosylphosphatidylinositol-anchored urokinase receptor is endocytosed by
coupling of urokinase receptor-bound urokinase-inhibitor complex to low density
Iipoprotein receptor-related protein (LRP) (Nykjaer et al., 1997, Herz et al., 1992).
Megalin, which is a closely related to LRP (both are low-density Iipoprotein
receptor family members), also binds the urokinase-inhibitor complex (Moestrup
et al., 1993), and may perform a similar function in regard to both the urokinase
receptor and cubilin-ligand complexes. In vitro study showed high affinity
interaction (Kd~7nM) between cubilin and megalin which does not interfere with

subsequent lF-Cbl binding to cubilin (Moestrup et al., 1998; Hammad et al.,

2000). In vivo immunohistochemistry of kidney, intestine and yolk sac in rat
showed colocalization of these two proteins in these tissues (Sahali et al., 1988,
1993, Elm et al., 1997, Moestrup et al., 1998). Megalin has a single
transmembrane domain and cytoplasmic domain motifs that signal for coated pit-
mediated internalization. These data suggest that megalin might mediate cubilin
endocytosis and recycling to the plasma membrane. Since the in vitro cubilin-
megalin complex is stable at pH 5, the two receptors might remain in complex

during the entire recycling pathway (Fig 1.3).

Megalin

  
 

IF-Cbl

IFR

iiiil till

Plasma membrane
PXY

Figure 1.3 Schematic representation of the interaction between
peripheral membrane protein cubilin and
transmembrane protein megalin.

Functional domains of cubilin
To better understand the relationship between cubilin function and its
structure, a mammalian expression approach has been used to identify the

cubilin regions important for membrane association and ligand binding

(Kristiansen et al., 1999). The results showed the N-terminal cubilin region is
crucial for membrane association. This region has no classical transmembrane
segment, but helical plotting demonstrated a conserved amphipathic helix motif
that may be involved in the assembly of cubilin (rat amino acid 74-109, human
amino acid 73-108). A recent electron microscopic study of purified bovine cubilin
has shown the formation of trimers in solution without detergent or lipid
(Lindblom et al., 1999). A region with four-heptad repeats, actually a part of the
putative amphipathic helix structure, has been proposed to account for the
assembly. The only cysteine located in the 110-amino acid N-terminal sequence,
outside of the CUB domains and EGF repeats, might also account for the partial
disulfide bond-dependent dimerization of a minor part of purified receptors.

The binding site for lF-Cbl has been located in the region encompassing
CUB domains 5-8 (Kristiansen et al., 1999). This establishes the CUB domain
structure as a ligand-binding domain. The two mutations identified in the human
cubilin gene both relate to the critical region now known to harbor the lF-Cbl
binding region. The FM2 mutation, suggested to activate a cryptic intronic splice
site leading to truncation of the receptor in CUB domain 6, may lead to an
inactive or unstable translation product, whereas the P/L substitution (FM1) in
CUB domain 8 might cause a structural change specifically impairing lF-Cbl
binding. The CUB domain is a barrel-like structure containing two layers of five-
stranded B-sheets with the B-turns in a surface-exposed position as in the
antigen binding regions of immunoglobulins (Menendez et al., 1995; Dias et al.,

1997). In view of the existence of 27 CUB domains in cubilin, a high number of

20

putative sites for various protein interactions may then be predicted.

A site for binding of RAP was localized within CUB domains 13-14. The
physiological importance of this interaction remains to be determined, but it is
tempting to speculate that RAP may assist the processing/folding of cubilin in line
with its suggested chaperone function which is important for normal processing
of the LDL receptor-related protein and megalin (Willnow et al., 1995). RAP, a
40-kDa endoplasmic reticulum protein, serves as a novel kind of chaperone or
escort protein and protects multiple ligand binding sites of processed low density
lipoprotein receptor family proteins, in particular the two giant receptors megalin
and LRP. The association with RAP indirectly indicates that cubilin is a multiple

ligand binding protein.

Regulation of cubilin expression in small intestine:

Most gastrointestinal functions are not evenly distributed along the
intestinal tract. The remarkable degree of coordination between the development
of various aspects of gastrointestinal function suggests that the process may be
triggered by a single or a few central mechanisms, such as weaning and (or)
hormones (Hermiston and Gordon, 1995). Early studies showed the gradient of
the enzymatic activity, or even the absorptive capacities for sugar and amino
acids, correlated with the gradient of mucosa mass of the small intestine
(Laganiere et al., 1984). However, later studies on the regulation of sucrase
isomaltase (SI) and aminopeptidase expression along small intestine showed the
translational or post-translational regulation of these enzymes, but no change of

the mRNA level (Hoffman and Chang, 1991; Danielsen et al., 1986). Different

21

genes show different regional specificity and are under different regulatory
mechanisms.

During fetal development of the human intestine, cubilin activity can be
detected along the entire length of intestine in early fetal development, but is
located to the distal ileum by the end of fetal development (Schohn et al., 1992).
Studies in rats also suggested that cubilin expression was developmentally
regulated (Ramasamy et al., 1989; Ramanujam et al., 1993b). The functional
expression of cubilin in distal small intestine was confirmed by oral administration
of radio-labeled Cbl in dogs and humans (Baker et al., 1958; Fleming et al.,
1962; Hagedorn and Alpers, 1977). The results of surgical transposition of ileal
segments from distal to proximal part of the canine intestine further confirm the
ileal location for lF-Cbl absorption in dogs (Drapanas et al., 1963). The
distribution of cubilin along the gastrointestinal tract is against the gradient of the
mucosa mass for the dog (Laganiére et al., 1984). The regional specific
expression of cubilin along the small intestine must be regulated at some level.
There is some in vivo evidence to suggest that the ileal capacity for Cbl
absorption is regulated by modulation of cubilin expression during pregnancy
(Brown et al., 1977; Robertson and Gallagher, 1983), but whether it is regulated
by certain hormones needs to be investigated.

From another point of view, the lining epithelium of the small intestinal
mucosa is in a constant and rapid state of renewal (Weinstein WM, 1974;
Leblond CP, 1980; Stappenbeck et al., 1998). Newly formed enterocytes arise in

the crypt region and migrate up the villus to be shed into the intestinal lumen at

22

or near the villus tip. The overall process of cell renewal depends on the number
of cells of the proliferative compartment; the rate of cell proliferation; and the
migration time. During this migration, enterocytes mature and differentiate for a
host of specific functions. Immunoelectrophoretic studies on human small-
intestinal brush-border proteins showed the amounts of lactase, sucrase,
maltase, microvillus aminopeptidase and dipeptidyl peptidase IV from jejunum
biopsies had their maximum activities near the mid-region of the villi and their
lowest activities at the base of the crypts (Skovbjerg H, 1981 ). How different
gene expression along crypt to villus axis is regulated is still an unsolved
question. One of the most well characterized gene regulatory mechanisms along
the crypt-villus axis is the SI gene for which transcriptional regulation at the
promoter depends on at least two groups of transcriptional proteins: hepatocyte
nuclear factor 1(HNF-1) and caudal-related homeodomain proteins (Cdx)(Wu et
al., 1994; Suh et al., 1996).

An immunoelectron microscopic study (Levine et al., 1984) of canine
cubilin distribution during enterocyte maturation from crypt to villus tip of ileal
mucosa showed that synthesis of cubilin can be easily detected in the mid-villus
cells, with the majority distributed intracellularly. Expression of cubilin
dramatically increases in the villus tips, where also the majority of cubilin can be
detected in the microvillus pits. The expression of cubilin is extremely low in the
crypt cells. Later study (Kapadia and Essandoh, 1988) with enriched villus and
crypt cell populations also showed IF-Cbl binding was about 30-fold greater in

ileal villous cells than in crypt cells. From the above, it is evident that during cell

23

differentiation from crypt to villus, cubilin expression is up regulated during
differenciation (maybe just slow). Although the particular signals affecting cubilin
expression remain unknown, we can evaluate normal cubilin expression in
different tissues, different regions of the same tissue, and the same region at
different stages of cell differentiation to determine how cubilin expression is
regulated. This will provide the basis for further studies of cubilin regulation and

functions.

Protein folding in the secretory pathway

Being a membrane-associated protein, cubilin, like other secretory and
membrane proteins, is likely carried to the plasma membrane via a specific
membrane protein biosynthetic transport pathway. The first and most important
compartment of this pathway is the endoplasmic reticulum (ER) which is
responsible for protein synthesis, early posttranslational modifications,
maturation and export of the fully folded protein to the destined location. Among
all of these functions of ER, one of the most notable are mechanisms to monitor
the fidelity of early biosynthetic events in the protein export pathway, also known
as "ER quality control" (Kopito RR, 1997). This quality control is critical for normal
cell function since it prevents premature export of incompletely or improperly
folded proteins from ER.

In recent years, numerous inborn errors of metabolism have been
identified which affect secretory or plasma membrane proteins (Medeiros-Neto et

al., 1996; Repaske et al., 1996; Kuivaniemi et al., 1991; Lehrman et al., 1987;

24

Yang et al., 1993). In many instances, mutations causing minor changes in
proteins that result in a folding defect lead to abnormal protein trafficking,
suggesting that proper conformation is critical for protein transport as well as
biological activity. During protein biogenesis, a peptide chain can self-assemble
into a stable, low free-energy conformation based on information contained within
the primary structure. For many proteins, additional factors are required for
proper folding. All these factors reside within the ER and are known as ER
chaperones. According to their specific functional mechanism, some of them
have been defined as primary controls and apply to all newly synthesized
proteins. The others, defined as secondary controls, are specific for a particular
protein or protein family (Ellgaard et al., 1999).

The synthesis of primary molecular chaperones, including Bip, GRP94,
and calnexin, etc., can be induced in response to the accumulation of unfolded
secretory proteins (Chapman et al., 1998). This induction can be at the
transcriptional level or posttranslational level. Because ER chaperone
associations are based on recognition of features enriched in incompletely folded
versions of exportable proteins, the Ca2+ and ATP dependent associations of ER
chaperones are usually at their highest levels immediately upon nascent chain
translocation into the ER, and terminated before export of the folded proteins
from their compartment (Sambrook JF, 1990; Clairmont et al., 1991 ). The fact
that molecular chaperones act on “substrate” proteins does not affect self-

assembly. By interacting with nascent chains, chaperones prevent undesirable

25

protein-protein interactions and increase the chances that newly made proteins
will have the opportunity to achieve their native structure.

Primary control also includes disulfide isomerase and prolyl isomerase.
They are the true foldases, in that they catalyze thioI-disulfide interchange with
broad substrate specificity (Freedman, 1995). As the nascent chain enters the
ER lumen, the rapid interaction within the hydrophobic domains is accompanied
by the ordered formation of intramolecular disulfide bounds, which stabilize
secondary and tertiary structures and can be critical for maintaining a biologically
active conformation. Vlfithin the oxidizing ER environment, reactive thiols also can
form mispaired disulfide bonds, and subsequent correction of aberrant disulfide
bonds may represent one of the rate-limiting steps in protein folding. Specific
cysteine residue mutations can also cause severe disulfide bond misformation
which eventually blocks protein transportation (Reddy and Corley, 1998).

Another important primary control mechanism involves oligosaccharide
trimming (Parodi AJ, 1999). A 14-saccharide unit, which contains two N-acetyl
glucosamine (GlcNac) residues, nine mannoses and three glucose resides, is
initially added to asparagine-linked consensus acceptor sites (NXS/T) in
glycoproteins. During ER enzyme-mediated trimming, the interaction between
nascent proteins and ER chaperones (calnexin) are regulated by trimming
enzymes, such as glucosyltransferase and glucosidase (Trombetta and Parodi,
1992; Sousa et al., 1992), which serve as folding sensors to direct the unfolded
glycoprotein to ER chaperones. Thus, proper N-linked glycosylation and

oligosaccharide trimming are also important for nascent protein folding.

26

Quite a number of protein specific factors that are responsible for effecting
folding and assembly of nascent proteins have been identified in different species
(Ellgaard et al., 1999). A well-studied factor, classified as an escort protein is
receptor associate protein (RAP), the function of which was first recognized by its
interaction with LRP molecules through the whole biosynthetic pathway, even
remaining associated on the cell surface (Bu et al., 1995; Farquhar et al., 1995).
Later study showed that RAP can also interact with other membrane receptors,
like megalin (Moestrup et al., 1996) and cubilin (Moestrup et al., 1998). Absence
of RAP can cause the defective transport of LRP through the ER (Willnow et al.,
1995). With RAP association, both RAP and LRP can exit the ER, which
suggests that the interaction maintains a favorable conformation and prevents
premature association of LRP ligands in the ER, which could lead to receptor
retention and [or degradation. RAP dissociation occurs when LRP reaches the
cell surface and LRP ligands can displace the escort. Surface expression of RAP
has been questioned in a study which demonstrated that RAP relocates to RAP
binding site in vitro when tissues are snap frozen (Abbate et al., 1993).

Cubilin is a large protein which contains 76 disulfide bonds and 47
potential N-glycosylation sites in human (Kozyraki et al., 1998). The assembly
process within the ER must be time consuming and require a lot of ER
chaperones. The in vitro interaction with RAP indicates that RAP or other
proteins may act as chaperones to facilitate the intracellular maturation. It is not

known whether cubilin-specific chaperones are involved in its processing.

27

Biochemical studies of affected dogs with selective intestinal Cbl
malabsorption demonstrated the accumulation of cubilin in the ileal enterocytes,
but rapid degradation in the kidney epithelial cells (Fyfe et al., 1991b, Fyfe,
unpublished data). The different fates of the same protein in different tissues
suggests that there is tissue-specific machinery for the protein assembly and
export.

The mechanism of defective expression of cubilin on the apical brush-
border membrane of both ileal and kidney epithelial cells of affected dogs with
selective intestinal Cbl malabsorption is the major focus of this project. The most
common reason for an ER export disorder is mutation in the coding sequence of
the exportable protein, eventhough defects of other accessory proteins may also
be involved. The latter case has been demonstrated only in abetalipoproteinemia
(Wetterau et al., 1992), a defect of the apo B-100 specific processing enzyme,

microsomal triglyceride transfer protein.

General outline for this thesis

A dog family with selective intestinal cobalamin malabsorption, which
closely resembles the human Imerslund-Grasbeck syndrome, provides a
valuable genetic model to further investigate the mechanism of this inherited
disease. The general goals of this thesis were to clone the candidate disease-
causing gene, cubilin / intrinsic factor-cobalamin receptor (IFCR), and
characterize its gene product. To achieve these goals, three closely related

projects were conducted and are described in this thesis: (1) Cloning of the full-

28

length cDNA of canine cubilin, providing the molecular basis for further study of
this candidate gene; (2) Studies on cubilin expression and post-translational
modification in the canine gastrointestinal tract, which helped to better
understand this newly cloned gene and the relationship between this gene and
the disease in the dog family; and (3) Linkage analysis between the disease
locus and candidate gene loci, which efficiently allowed determination of whether

two candidate loci habored a disease-causing mutation.

29

References

Abbate M, Bachinsky D, Zheng G, Stamenkovic I, McLaughlin M, Niles JL,
McCluskey RT, and Brown D. (1993) Location of gp330/a2-m receptor-
associated protein (az-MRAP) and its binding sites in kidney: distribution of
endogenous az-MRAP is modified by tissue processing. EurJ Cell Biol 61: 139-
149.

Abdelaal MA and Ahmed AF. (1991) Case report: Imerslund-Grasbeck syndrome
in a Saudi family. Acta Paediatr Scand 80: 1109-1112.

Allen RH, Seetharam B, Podell E, and Alpers DH. (1978) 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 61: 47-54.

Altay C, Cetin M, Gtimrflk F, Irken G, Yetgin S. and Laleli Y. (1995) Familial
selective vitamin B12 malabsorption (Imerslund-Grasbeck syndrome) in a pool of
Turkish patients. Pediat Hematol Oncol 12: 1 9-28.

Altay C. and Cetin M. (1999) Vitamin Biz absorption test and oral treatment in 14
children with selective vitamin B12 malabsorption. Pediatr. Hematol. Oncol.
16(2):159-163.

Aminoff M, Tahvanainen E, Grasbeck R, Weissenbach J, Broch H, and de la
Chapelle A. ( 1995) Selective intestinal malabsorption of vitamin B12 displays
recessive mendelian inheritance: assignment of a locus to chromosome 10 by
linkage. Am J Hum Genet 57: 824-31.

Aminoff M, Carter JE, Chadwick RB, Johnson C, Grasbeck R, Abdelaal MA,
Broch H, Jenner LB, Verroust PJ, Moestrup SK, Chapelle A, and Krahe R. (1999)
Mutations in CUBN, encoding the intrinsic factor-vitamin B12 receptor, cubilin,
cause hereditary megaloblastic anaemia 1. Nat Genet 21: 309-313.

Baker SJ, Mackinnon NL, and Vasudevia P. (1958) The site of absorption of
orally administered vitamin B12 in dogs. Ind J Med Res 46: 812-817.

Batt RM, and Horadagoda NU. (1989) Gastric and pancreatic intrinsic factor-
mediated absorption of cobalamin in the dog. Am J Physiol 257: G344-G349.

Batuman V, Verroust PJ, Navar GL, Kaysen JH, Goda FO, Campbell WC, Simon

E, Pontillon F, Lyles M, Bruno J, and Hammond TG. (1998) Myeloma light chains
are ligands for cubilin (gp280) Am J Physiol 275(R44): F246-F254.

30

Ben-Bassat I, Feinstein A. and Ramot B. (1969) Selective vitamin B12
malabsorption with proteinuria in Israel-«Clinical and genetic aspects. Israel J
Med Sci 5:62-68.

Bim H, Verroust PJ, Nexo E, Hager H, Jacobsen C, Christensen El, and
Moestrup S. (1997) Characterization of an epithelial ~460-kDa protein that
facilitates endocytosis of intrinsic factor-vitamin B12 and binds receptor-
associated protein. J Biol Chem 272(42): 26497-26504.

Birn H, Fyfe JC, Jacobsen C, Mounier F, Veroust PJ, Orskov H, Willnow T,
MoestrupSK, and Christensen El. (2000) Cubilin is an albumin binding protein
important for renal albumin reabsorption. J Clin Invest (in press).

Booth CC and Mollin DL. (1959) The site of absorption of vitamin B12 in man.
Lancet 1: 18-21.

Bork P and Beckmann G. (1993) The CUB domain. A widespread module in
developmentally regulated proteins. J Mol Biol 231: 539-545.

Bose S, Seetharam 8, Hammond TG, and Seetharam B. (1995a) Regulation of
expression of transcobalamin Il-receptor in the rat. Biochem J 310: 923-929.

Bose S, Seetharam S, and Seetharam B. (1995b) Membrane expression and
interactions of human transcobalamin II receptor. J Biol Chem 270: 8152-8157.

Broch H, Imerslund O, Monn E, Hovig T, and Seip M. (1984) Imerslund-Grasbeck
anemia, A long-term follow-up study. Acta Paediatr Scand 73: 248-253.

Brown J, Robertson J, and Gallagher N. (1977) Humoral regulation of vitamin B12
absorption by pregnant mouse small intestine. Gastroentenology 72: 881-885.

Bu G, Geuze HJ, Strous GJ, and Schwartz AL.(1995) 39 kDa receptor-
associated protein is an ER resident protein and molecular chaperone for LDL
receptor-related protein. EMBO J 14: 2269-2280.

Burman JF, Jenkins WJ, Walker-Smith JA, Phillips AD, Sourial NA, Williams CB,
and MollinDL. (1985) Absent ileal uptake of IF-bound vitamin B12 in vivo in the
Imerslund-Grasbeck syndrome (familial vitamin B12 malabsorption with
protienuria). Gut 26: 311-314.

Carmel R, Abramson SB, and Renner IG. (1983) Characterization of pure human

pancreatic juice: cobalamin content, cobalamin-binding proteins and activity
against human R binders of various secretions. Clin Sci 64: 193-205.

31

Celep F, Karagfizel A, Aynaci FM. and Erduran E. (1996) A case report of 46,
XX, del (21) (q22) de novo deletion associated with Imerslund-Grasbeck
syndrome. Clin Genet. 50:248.

Chapman R, Sidrauski C, and Walter P. (1998) Intracellular signaling from the
endoplasmic reticulum to the nucleus. Annu Rev Cell Dev Biol 14: 459-485.

Christensen JM, Hippe E, Olesen H, Rye M, Haber E, Lee L, and Thomsen J
(1973) Purification of human intrinsic factor by affinity chromatography. Biochim
Biophys Acta 303, 319-332.

Clairmont CA, De Maio A, and Hirschberg CB. (1992) Translocation of ATP into
the lumen of rough endoplasmic reticulum-derived vesicles and its binding to
luminal proteins including BiP (GRP 78) and GRP 94. J Biol Chem 267: 3983-
3990.

Colle E, Greenberg L, and Krivit W. (1961) Studies of a patient with selective
deficiency in absorption of vitamin B12. Blood 18: 48.

Cooper BA and Castle WB. (1960) Sequential mechanisms in the enhanced
absorption of vitamin 812 by intrinsic factor in the rat. J Clin Invest 39: 199-214.

Cotter R and Rothenberg SP. (1976) Solubilization, partial purification and
radioassay for the intrinsic factor receptor from the ileal mucosa. Br J Haematol
34: 477-487.

Dan N and Cutler DF. (1994) Transcytosis and processing of intrinsic factor-
cobalamin in Caco-2 cells. J Biol Chem 269(29): 18849-18855.

Dias JM, Carvalho AL, Kolln l, Calvete JJ, Topfer-Petersen E, Varela PF,
Romero A, Urbanke C, and Romao MJ. (1997) Crystallization and preliminary X-
ray diffraction studies of aSF P, a bovine seminal plasma protein with a single
CUB domain architecture. Protein Sci 6: 725-727.

Dieckgraete BK, Seetharam B, and Alpers DH. (1988) Developmental regulation
of rat intrinsic factor mRNA. Am J Physiol 254: G913-919.

Donaldson RM, Mackenzie l, and Trier JS. (1967) Intrinsic factor-mediated
attachment of vitamin B12 to brush borers and microvillous membranes of
hamster intestine. J Clin Invest 46: 1215-1228.

Doscherholmem A, Hagen PS, and Olin L. (1959) Delay of absorption of
radiolabeled cyanocobalamin in the intestinal wall in the presence of intrinsic
factor. J Lab & Clin Med 54: 434-439.

32

Drapanas T, Williams JS, McDonald JC, Heyden W, Bow T, and Spencer RP.
(1963) Role of the ileum in the absorption of vitamin B12 and intrinsic factor (NF).
JAMA 184: 337-341.

Dugue B, Aminoff M, Aminoe—Gastin I, Leppanen E, Grasbeck R, and Gueant JL.
(1998) A urinary radioisotope-binding assay to diagnose Grasbeck-Imerslund
disease. J Pediatr Gastroenterol Nutr 26(1): 21 -25.

El Kholty S, Guéant JL, Bressler L, Djalali M, Boissel P, Gerard P, Nicolas JP.
(1991) Portal and biliary phases of entrohepatic circulation of corrinioids in
humans. Gastroenterology 101: 1399-1408.

Ellgaard L, Molinari M, and Helenius A. (1999) Setting the standards: Quality
control in the secretory pathway. Science 286: 1882-1888.

Farquhar MG, Saito A, Kerjaschki D, and Orlando RA. (1995) The Heymann
nephritis antigenic complex: megalin (gp330) and RAP. J Am Soc Nephrol 6: 35-
47.

Fenton WA and Rosenberg LE. (1989) Inherited disorders of cobalamin transport
and metabolism. In Scriver CR, Beaudet AL, Sly WS, Valle D. (eds) The
Metabolic Basis of Inherited Disease McGraw-Hill, New York.

Festen HP. (1991) Intrinsic factor secretion and cobalamin absorption.
Physiology and pathophysiology in the gastrointestinal tract. Scand J
Gastroenterol 1 88 (supplement): 1-7.

Fleming WH, King R, Galloway RA, and Roche JJ. (1962) The site of absorption
of orally administered Com-labeled vitamin B12 in dogs: the effect of dose.
Gastroenterology 42: 164-168.

Freedman RB (1995) The formation of protein disulfide bonds. Curr Opin Struct
Biol 5: 85-91.

Fyfe JC, Jezyk PF, Giger U, and Patterson DF. (1989) Inherited selective
malabsorption of vitamin B12 in giant schnauzers J Am Anim Hosp Assoc25:
533-539.

Fyfe JC, Giger U, Hall CA, Jezyk PF, Klumpp SA, Levine JS, and Patterson DF.
(1991a) Inherited selective intestinal cobalamin malabsorption and cobalamin
deficiency in dogs. Pediatr Res 29: 24-31.

Fyfe JC, Ramanujam KS, Ramaswamy K, Patterson DF, and Seetharam B.
(1991b) Defective brush-border expression of intrinsic factor-cobalamin receptor
in canine inherited intestinal cobalamin malabsorption. J Biol Chem 266: 4489-
4494.

33

Fyfe JC. (1993) Feline intrinsic factor (IF) is pancreatic in origin and mediates
ileal cobalamin (cbl) absorption. J Vet Intern Med 7: 133 (abstract).

Gordon Jl. (1989) Intestinal epithelial differentiation: new insights from chimeric
and transgenic mice. J cell Biol 108: 1187-1 191.

Grasbeck R, Gordin R, Kantero I, And Kuhlback B. (1960) Selective vitamin B12
malabsorption and proteinuria in young people. Acta Med Scand 167:289-296.

Guéant JL, Masson C, Schohn H, Girr M, Saunier M, and Nicolas JP. (1992)
Receptor-mediated endocytosis of the intrinsic factor-cobalamin complex in HT
29, a human colon carcinoma cell line. FEBS 297(3): 229-232.

Gullberg R. (1972) Vitamin B 12 -binding proteins in normal human blood plasma
and serum. Scand J Haematol 9(6):639-647.

Hagedorn CH and Alpers DH. (1977) Distribution of intrinsic factor-vitamin B12
receptors in human intestine. Gastroenterology 73: 1019-1022.

Hall CA. (1975) Transcobalamin I and II as natural transport proteins of vitamin
B12. J Clin Invest 56: 1125-1131.

Hall CA, Green-Colligan PD, and Begley JA. (1985) Synthesis of transcobalamin
II by cultured human hepatocytes. Biochim Biophys Acta 838(3): 387-389.

Hakami N, Neiman PE, Canellos GP, and Lazerson J. (1971) Neonatal
megaloblastic anemia due to inherited transcobalamin || deficiency in two
siblings. N Engl J Med 285: 1163-1170.

Hammad SM, Stefansson S, Twal WO, Drake CJ, Fleming P, Remaley A, Brewer
HB Jr, and Argraves WS. (1999) Cubilin, the endocytic receptor for intrinsic
factor-vitamin B(12) complex, mediates high-density Iipoprotein holoparticle
endocytosis. Proc Natl Acad Sci U S A 96(18): 10158-63 .

Hammad SM, Barth JL, Knaak C, and Argraves WS. (2000) Megalin Acts in
Concert with Cubilin to Mediate Endocytosis of High Density Lipoproteins.
J Biol Chem 275: 12003-12008.

Herbert V and Castle GBJ. (1961) Divalent cation and pH dependence of rat
intrinsic factor action in everted sacs and mucosal homogenates of rat small
intestine. J Clin Invest 40, 1978-1983.

Herz J, Clouthier DE, and Hammer RE. (1992) LDL receptor-related protein

internalizes and degrades uPA-PAI-1 complexes and is essential for embryo
implantation. Cell 71: 41 1-421.

34

Hoffman LR and Chang EB. (1991) Determinants of regional sucrase-isomaltase
expression in adult rat small intestine. J Biol Chem 266: 21815-20

Hooper DC, Alpers DH, Burger RL, Mehlman CS, and Allen RH. (1973)
Characterization of ileal vitamin B12 binding using homogeneous human and hog
intrinsic factors. J Clin Invest 52: 3074-3083.

ldriss JM and Jonas AJ. (1991) Vitamin B12 transport by rat liver lysosomal
membrane vesicles. J Biol Chem 266: 9438-9441.

Imerslund O. (1960) Idiopathic chronic megaloblastic anemia in children. Acta
Paediat Scand Suppl 49: 1-115.

Imerslund O and Bjornstad P. (1963) Familial vitamin B12 malabsorption. Acta
Haematol 30: 1-7.

Kapadia CR, Serfilippi D, Voloshin K, and Donaldson RM. (1983) Intrinsic factor-
mediated absorption of cobalamin by guinea pig ileal cells. J Clin Invest 71 : 440-
448.

Kapadia CR and Essandoh LK. (1988) Active absorption of vitamin B12 and
conjugated bile salts by guinea pig ileum occurs in villous and not crypt cells.
Digest Dis Sci 33: 1377-1382.

Katz M and Cooper BA. (1974a) Solubilized receptor for intrinsic factor-vitamin
B12 complex from guinea pig intestinal mucosa. J Clin Invest 54: 733-739.

Katz M, Mehlman CS, and Allen RH. (1974b) Isolation and characterization of an
abnormal human intrinsic factor. J Clin Invest 53: 1274-1283.

Kojima T, Miyaishi O, Saga S, lshiguro N, Tsutsui Y, and Iwat. (1998) The
retention of abnormal type | procollagen and correlated expression of HSP 47 in

fibroblasts from a patient with lethal osteogenesis imperfecta. J Pathol 184(2):
212-8.

Kolhouse JF and Allen RH. (1977) Recognition of two intracellular cobalamin
binding proteins and their identification as methyl-malonyl-CoA mutase and
methionine synthase. Proc Natl Acad Sci USA 74:921-925.

Kopito RR. (1997) ER quality control: the cytoplasmic connection. Cell 88(4):
427-30.

Kozyraki R, Kristiansen M, Silahtaroglu A, Hansen C, Jacobsen C, Tommerup N,

Verroust PJ, and Moestrup SK. (1998) The human intrinsic factor-vitamin B12
receptor, cubilin: molecular characterization and chromosomal mapping of the

35

gene to 10p within the autosomal recessive megaloblastic anemia (MGA 1)
region. Blood 91(10): 3593-3600.

Kozyraki R, Fyfe J, Kristiansen M, Gerdes C, Jacobsen C, Cui S, Christensen El,
Aminoff M, Chapelle A, Krahe R, Verroust PJ, and Moestrup S. (1999) The
intrinsic factor-vitamin B12 receptor, cubilin, is a high-affinity apolipoprotein A-I
receptor facilitating endocytosis of high-density lipoprotein. Nat Med 5: 656-661.

Kristiansen M, Kozyraki R, Jacobsen C, Nexe E, Verroust PJ, and Moestrup SK.
(1999) Molecular dissection of the intrinsic factor-vitamin B12 receptor, cubilin,
discloses regions important for membrane association and ligand binding. J Biol
Chem 274(29): 20540-20544.

Kuivaniemi H, Tromp G, and Prockop DJ. (1991) Mutations in collagen genes:
causes of rare and some common diseases in humans. FASEB J 5: 2052-2060.

Kuovonen I. (1980) Solubilization of the pig ileal intrinsic factor receptor with
papain treatment and studies on the solubilized receptor. Biochim Biophys Acta
626: 244-53.

Kuovonen l and Grasbeck R. (1980) A simplified technique to isolate the porcine
and human ileal intrinsic factor receptors and studies on their subunit structure.
Biochim Biophy Res Comm 86: 358-364.

LaFramboise R, Cooper BA, and Rosenblatt DS. (1992) Malabsorption of vitamin
812 from the intestine in a child with cbIF disease: evidence for lysosomal-
mediated absorption. Blood 80: 291 -292.

Laganiere S, Berteloot A, and Maestracci D. (1984) Digestive and absorptive
functions along dog small intestine: comparative distributions in relation to
biochemical and morphological parameters. Comp Biochem Physiol 79: 463-
472.

Leblond CP. (1980) The life history of cells in renewing systems. Am J Anat 160:
1 14-158.

Lehrman MA, Schneider WJ, Brown MS, Davis CG, Elhammer A, Russell DW,
and Goldstein JL. (1987) The Lebanese allele at the low density lipoprotein
receptor locus. Nonsense mutation produces truncated receptor that is retained
in endoplasmic reticulum. J Biol Chem 262: 401-410.

Levine JS, Nakane PK, and Allen RH. (1980) lmmunocytochemical localization of

human intrinsic factor: the nonstimulated stomach. Gastroenterology 79: 493-
502.

36

Levine JS, Nakane PK, and Allen RH. (1982) lmmunocytochemical localization of
intrinsic factor-cobalamin bound to the guinea pig ileum in vivo. Gastroenterology
82: 284-290.

Levine JS, Allen RH, Alpers DH, and Seetharam B. (1984) lmmunocytochemical
localization of the intrinsic factor-cobalamn receptor in dog-ileum: distribution of
intracellular receptor during cell maturation. J Cell Biol 98: 11 11-1118.

Le Panse S, Ayani E, Mulliez N, Chatelet F, Cywiner-Golenzer C, Galceran M,
Citadelle D, Roux C, Ronco P, and Verroust P. (1994) Antibodies to the 280-kd
coated pit protein, target of teratogenic antibodies, produce alterations in the
traffic of internalized proteins. Am J Pathol 145: 1526-1536 .

Le Panse S, Galceran M, Pontillon F, Lelongt B, van de Putte M, Ronco PM, and
Verroust PJ. (1995) Immunofunctional properties of a yolk sac epithelial cell line
expressing two proteins gp280 and gp330 of the intermicrovillar area of proximal
tubule cells: inhibition of endocytosis by the specific antibodies. EurJ Cell Biol
67: 120-129 .

Le Panse S, Ayani E, Nielsen S, Ronco P, Verroust P, and Christensen El.
(1997) Internalization and recycling of glycoprotein 280 in epithelial cells of yolk
sac. EurJ Cell Biol 72: 257-67.

Li N, Seetharam S, Rosenblatt DS, and Seetharam B. (1994) Expression of
transcobalamin ll mRNA in human tissues and cultured fibroblasts from normal
and TC ll deficient patients. Biochem J 301:585-590.

Lindblom A, Quadt N, Marsh T, Aeschlimann D, Morgelin M, Mann K, Maurer P,
and Paulsson M. (1999) The intrinsic factor-vitamin B12 receptor, cubilin, is
assembled into trimers via a coiled-coil alpha-helix. J Biol Chem 274: 6374-6380.

Mackenzie IL and Donaldson RM. (1972) Effect of divalent cations and pH on
intrinsic factor-mediated attachment of vitamin B12 to intestinal microvillous
membranes. J Clin Invest 51: 2465-2471.

Mackenzie IL, Donaldson RM, Trier JS, and Mathan VI. (1972) Ileal mucosa in
familial vitamin B12 malabsorption. New Engl J Med 286: 1021-1025.

Marcoullis G, Rothenberg SP, and Labombardi VJ. (1980) Preparation and
characterization of proteins in the alimentary tract of the dog which bind
cobalamin and intrinsic factor. J Biol Chem 255: 1824-1829.

Marsden KA, Newman NM, and Marsden DE. (1979) Imerslund’s syndrome --- A
case from Australia and review of the literature. Aust Paediatr J 15:49-52.

37

Medeiros-Neto G, Kim PS, Yoo SE, Vono J, Targovnik HM, Camargo R, Hossain
SA, and Arvan P. (1996) Congenital hypothyroid goiter with deficient
thyroglobulin. Identification of an endoplasmic reticulum storage disease with
induction of molecular chaperones. J Clin Invest 98: 2838-2844.

Menendez M, Gasset M, Laynez J, Lopez-Zumel C, Usobiaga P, Topfer-
Petersen E, and Calvete JJ. (1995) Analysis of the structural organization and
thermal stability of two spermadhesins. Calorimetric, circular dichroic and
Fourier-transfomi infrared spectroscopic studies. Eur J Biochem 234: 887-896.

Mellman IS, Youngdah-Turner P, \Mllard HF, and Rosenberg LE. (1977)
Intracellular binding of radioactive hydroxocobalamin to cobalamin-dependent
apoenzymes in rat liver. Proc Natl Acad Sci USA 74:916-920.

Moestrup SK, Nielsen S, Andreasen P, Jorgensen KE, Nykjaer A, Roigaard H,
Gliemann J, and Christensen El. (1993) Epithelial glycoprotein-330 mediates
endocytosis of plasminogen activator-plasminogen activator inhibitor type-1
complexes. J Biol Chem 268: 16564-16570

Moestrup SK, Birn H, Fischer PB, Peterson CM, Verroust PJ, Sim RB,
Christensen El, and Nexo E. (1996) Megalin-mediated endocytosis of
transcobalamin-vitamin B12 complexes suggests a role of the receptor in vitamin
B12 homeostasis. Proc Natl Acad Sci USA 93: 8612-8617.

Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D,
Pontillon F, Goda F, Christensen El, Hammond TG, and Verroust PJ. (1998) The
intrinsic factor-vitamin B12 receptor and target of teratogenic antibodies is a
megalin-binding peripheral membrane protein with homology to developmental
proteins. J Biol Chem 273(9): 5235-5242.

Nicolas JP and Guéant JL. (1995) Gastric intrinsic factor and its receptor. BaiII
Clin Haematol 8(3): 515-531.

Nykjaer A, Conese M, Christensen El, Olson D, Cremona O, Gliemann J, and
Blasi F. (1997) Recycling of the urokinase receptor upon internalization of the
uPAzserpin complexe. EMBO J 16: 2610-2620.

Okuda K and Fujii T. (1977) Solubilization of the ileal receptor intrinsic factor-
vitamin B12 comples in the rat. J Lab Clin Med 89: 172-180.

Parodi AJ. (1999) Reglucosylation of glycoproteins and quality control of
glyc0protein folding in the endoplasmic reticulum of yeast cells. Biochim Biophys
Acta 1426(2): 287-95.

38

Quadros EV, Regec AL, Khan KM, Quadros E, and Rothenberg SP. (1999)
Transcobalamin lI synthesized in the intestinal villi facilitates transfer of
cobalamin to the portal blood. Am J Physiol 277(1 Pt 1): 6161-166.

Ramanujam KS, Seetharam S, Ramasamy M, and Seetharam B. (1991a)
Expression of cobalamin transport proteins and cobalamin transcytosis by colon
adenocarcinoma cells. Am J Physiol 260(23): G416-G422.

Ramanujam KS, Seetharam S, Dahms NM, and Seetharam B. (1991b)
Functional expression of intrinsic factor-cobalamin receptor by renal proximal
tubular epithelial cells. J Biol Chem 266: 13135-13140.

Ramanujam KS, Seetharam S, and Seetharam B. (1992) Leupeptin and
ammonium chloride inhibit intrinsic factor mediated transcytosis of
[57Co]cobalamin across polarized renal epithelial cells. Biochem Biophys Res
Comm 182: 439-446.

Ramanujam KS, Seetharam S, and Seetharam B (1993a) Intrinsic factor-
cobalamin receptor activity in a marsupial, the American opossum (Didelphis
virginiana). Comp Biochem Physiol Comp Physiol 104: 771-775.

Ramanujam KS, Seetharam S, and Seetharam B. (1993b) Regulated expression
of intrinsic factor-cobalamin receptor by rat visceral yolk sac and placental
membranes. Biochimi Biophysi Acta 1146: 243-246.

Ramaswamy M, Alpers DH, Triuppathi C, and Seetharam B. (1989) Cobalamin
release from intrinsic factor and transfer to transcobalamin ll within the rat
enterocyte Am J Physiol 257: 6791-797.

Rappazzo ME and Hall CA. (1972) Cyanocobalamin transport proteins in canine
plasma. Am J Physiol 222(1): 202-206.

Reddy PS and Corley RB. (1998) Assembly, sorting, and exit of oligomeric
proteins from the endoplasmic reticulum. Bioessays 20: 546-554.

Repaske DR, Summar ML, Krishnamani MR, Gultekin EK, Arriazu MC, Roubicek
ME, Blanco M, Isaac GB, and Phillips JA 3rd (1996) Recurrent mutations in the
vasopressin-neurophysin II gene cause autosomal dominant neurohypophyseal
diabetes insipidus. J Clin Endocrinol Metab 81: 2328-2334.

Retief FP, Gottlieb CW, and Herbert V. (1966) Mechanism of vitamin B12 uptake
by erythocytes. J Clin Invest 45: 1 907-1 91 5.

Robertson JA, and Gallagher ND. (1983) Increased intestinal uptake of

cobalamin in pregnancy does not require synthesis of new receptors. Biochimi
Biophysi Acta 757: 145-150.

39

Robertson JA and Gallagher ND. (1985a) In vivo evidence that cobalamin is
absorbed by receptor-mediated endocytosis in the mouse. Gastroenterology
88(4): 908-12.

Robertson JA and Gallagher ND. (1985b) Intrinsic facotr-cobalamin accumulates
in the ileum of mice treated with chloroquine. Gastroenterology 89: 1353-1359.

Rubin HM, Giorgio AJ, Macdonald RR, and Linarelli LG. (1974) Selective
malabsorption of vitamin B12: Report of a case with metabolic studies. Am J Dis
Child 127: 713-717.

Sahali D, Mulliez N, Chatelet F, Laurent-Nnter C, Citadelle D, Sabourin JC,
Roux C, Ronco P, and Verroust P. (1993) Comparative immunochemistry and
ontogeny of two closely related coated pit proteins. The 280-kd target of
teratogenic antibodies and the 330-kd target of nephritogenic antibodies. Am J
Pathol142: 1654-1667.

Sahali D, Mulliez N, Chatelet F, Dupuis R, Ronco P, and Verroust P. (1988)
Characterization of a 280-kD protein restricted to the coated pits of the renal
brush border and the epithelial cells of the yolk sac. Teratogenic effect of the
specific monoclonal antibodies. J Exp Med 167: 213-218.

Sambrook JF. (1990) The involvement of calcium in transport of secretory
proteins from the endoplasmic reticulum. Cell 61: 197-199.

Schohn H, Gueant L, Leheup B, Saunier M, Grignin G, and Nicolas JP. (1992)
Intrinsic factor receptor during fetal development of the human intestine. Biochem
J 286: 153-156.

Seetharm B and Alpers DH. (1981) Isolation and characterization of the ileal
receptor for intrinsic factor-cobalamin. J Biol Chem 256(8): 3785-3790.

Seetharam B, Bagur SS, and Alpers DH. (1981) Isolation and characterization of
proteolytically derived ileal receptor for intrinsic factor-cobalamin. J Biol Chem
257(1): 183-189.

Seetharm B, Levine JS, Ramasamy M, and Alpers DH. (1988) Purification,
properties, and immunochemical localization of a receptor for intrinsic factor-
cobalamin complex in the rat kidney. J Biol Chem 263(9): 4443-4449.

Seetharam S, Ramanujam KS, and Seetharam B. (1992) Synthesis and brush

border expression of intrinsic factor-cobalamin receptor from rat renal cortex. J
Biol Chem 267: 7421-7427.

40

Seetharam B, Christensen El, Moestrup SK, Hammond TG, and Verroust PJ.
(1997) Identification of rat yolk sac target protein of teratogenic antibodies,
gp280, intrinsic factor-cobalamin receptor. J Clin Invest 99(10): 2317-2322

Segrest JP, De-Loof H, Dohlman JG, Brouillette CG, and Anantharamaiah GM.
(1990) Amphipathic helix motif: classes and properties. Proteins 8: 103-107.

Sievens CJ. (1964) Megaloblastic anemia in a 3 year old child. Blood. 23: 547.

Siggaard C, Rittig S, Corydon TJ, Andreasen PH, Jensen TG, Andresen BS,
Robertson GL, Gregersen N, Bolund L, and Pedersen EB (1999) Clinical and
molecular evidence of abnormal processing and trafficking of the vasopressin
preprohorrnone in a large kindred with familial neurohypophyseal diabetes
insipidus due to a signal peptide mutation. J Clin Endocrinol Metab 84:2933-41.

Simpson KW, Morton DB, and Batt RM. (1989) Effect of exocrine pancreatic
insufficiency on cobalamin absorption in dogs. Am J Vet Res 50: 1233-1236.

Simpson KW, Alpers DH, De Ville J, Swanson P, Farmer S, and Sherding RG.
(1993) Cellular localization and hormonal regulation of pancreatic intrinsic factor
secretion in dogs. Am J Physio! 265: 178-188.

Skovbjerg H. (1981) lmmunoelectriophoretic studies on human small-intestinal
brush-border proteins -- Relation between enzyme activity and immunoreactive
enzyme along the villus-crypt axis. Biochem J 193: 887-890.

Smith A (1962) Veganism: A clinical survey with observations of vitamin B12
metalbolism. Br Med J 1: 1655-1659.

Sousa MC, Ferrero-Garcia MA, and Parodi AJ. (1992) Recognition of the
oligosaccharide and protein moieties of glycoproteins by the UDP-
Glc:glycoprotein glucosyltransferase. Biochemistry 31: 97-105.

Spurling CL, Sacks MS, and Jiji RM. (1964) Juverile pernicious anemia. New
Engl J Med. 271:995-1003.

Stappenbeck TS, Wong MH, Saam JR, Mysorekar IU, and Gordon JI. (1998)
Notes from some crypt watchers: regulation of renewal in the mouse intestinal
epithelium. Curr Opin Cell Biol 10: 702-709.

Stones DK and Ferreira M. (1999) Imerslund-Grasbeck syndrome in an African
patient. J Trop Pediatr 45(2): 106-107.

Suh E-R, Chen L, Taylor J, and Traber PG. (1994) A homeodomain protein

related to caudal regulates intestine-specific gene transcription. Mol Cell Biol 14:
7340-7351.

41

Sullivan LW. (1970) Vitamin B12 metabolism and megaloblastic anemia.
Seminars Hematol 7: 6-22.

Trombetta SE and Parodi AJ. (1992) Purification to apparent homogeneity and
partial characterization of rat liver UDP-glucosezglycoprotein glucosyltransferase.
J Biol Chem 267: 9236-9240.

Vaillant C, Horadagoda NU, and Batt RM. (1990) Cellular localization of intrinsic
factor in pancreas and stomach of the dog. Cell Tissue Res 260: 117-122.

Weinstein WM. (1974) Epithelial cell renewal of the small intestinal mucosa. Med
Clin N Am 58: 1375-1386.

Wetterau JR, Aggerbeck LP, Bouma ME, Eisenber C, Munck A, Herrnier M,
Schmitz J, Gay G, Rader DJ, and Gregg RE. (1992) Absence of microsomal
triglyceride transfer protein in individuals with abetalipoproteinemia. Science. 258
(6): 999-1001.

VVillnow TE, Armstrong SA, Hammer RE, and Herz J. (1995) Functional
expression of low density lipoprotein receptor-related protein is controlled by
receptor-associated protein in vivo. Proc Natl Acad Sci USA 92: 4537-4541.

Wu GD, Chen L, Forslund K, and Traber PG. (1994) Hepatocyte nuclear factor
1a (HNF-1a) and HNF-1B regulated transcription via two elements in an
intestine-specific promoter. J Biol Chem 269: 17080-17085.

Yang Y, Ducos R, Rosenberg AJ, Catrou PG, Levine JS, Podell ER, and Allen
RH. (1985) Cobalamin malabsorption in three siblings due to an abnormal

intrinsic factor that is markedly susceptible to acid and proteolysis. J Clin Invest
76: 2057-2065.

Yang Y, Janich S, Cohn JA, and Wilson JM. (1993) The common variant of cystic
fibrosis transmembrane conductance regulator is recognized by hsp70 and
degraded in a pre-Golgi nonlysosomal compartment. Proc Natl Acad Sci USA 90:
9480-9484.

Yetgin S, Ozsoylu S, and Zamani VP. (1983) Imerslund-Grasbeck syndrome and
generalized malabsorption. Turk J Pedia 25:193-196.

Youngdahl-Turner P, Rosenberg LE, and Allen RH. (1978) Binding and uptake of
transcobalamin II by human fibroblasts. J Clin Invest 61 2133-141

42

CHAPTER 2

CLONING OF CANINE CUBILIN

43

Introduction

Gastrointestinal absorption of vitamin 812 (also called cobalamin,CbI) is
essential for animals, including humans, to maintain normal hematopoiesis and
integrity of the central nervous system. Cbl absorption is a complex process and
requires many Cbl binding proteins. A defect of any one of these proteins can
cause Cbl deficiency. Cubilin, previously known as intrinsic factor-cobalamin
receptor (IFCR), plays a critical role in endocytosis of the lF-Cbl complex in distal
small intestine and reabsorption of various proteins in the kidney.

A family of dogs with selective intestinal Cbl malabsorption (Fyfe et al.,
1989) has been identified that has similar clinical and laboratory features as seen
in humans with a disorder known as Imerslund-Grasbeck syndrome (l-GS)
(Grasbeck et al., 1960; Imerslund 0., 1960). The affected puppies exhibited
failure of gastrointestinal absorption of oral radiolabeled Cbl. Immunoelectron
microscopy and cell fractionation studies demonstrated a defect in cubilin
expression on apical brush border membranes of ileum and renal proximal
tubules, whereas other brush-border proteins were expressed normally (Fyfe et
al., 1991a, b). Purified normal and affected dog renal cubilin showed similar
migration on SDS-PAGE, but affected dog cubilin had abnormal proteolytic
peptide profiles and the asparagine-Iinked oligosaccharides were
endoglycosidase H-sensitive. The cubilin of affected dogs accumulated in ileal
enterocytes but was rapidly degraded in the renal epithelial cells. All these
findings suggested that affected dog cubilin did not fold properly and did not

reach the mid-Golgi compartment of the biosynthetic pathway, most likely being

44

retained in the endoplasmic reticulum (ER). Many such so-called ER storage
diseases which selectively inhibit cell surface expression of a plasma membrane
or secretory protein have been described (Moolenaar et al., 1997; Medeiros-Neto
et al., 1996; Repaske et al., 1996; Kuivaniemi et al., 1991; Lehrman et al., 1987;
Yang et al., 1993). With the exception of abetalipoproteinemia (Wetterau et al.,
1992), these disorders have been attributed to mutations in the coding sequence
of the exportable protein.

Rat (Moestrup et al., 1998) and human (Kozyraki et al., 1998) cubilin have
recently been cloned and showed a unique tandem arrangement of 8 EGF
domains and 27 CUB domains. Two independent mutations have been identified
in CUBN in 17 Finnish families with IG-S (Aminoff et al., 1999). To further
characterize the canine model with selective intestinal Cbl malabsorption and to
understand the multiple functions of the gene, we cloned the canine cubilin cDNA
from dog kidney proximal tubule epithelial cells and compared the canine cubilin

sequence to human and rat sequences.

45

Materials and Methods

Peptide sequencing of purified canine renal cubilin

Canine renal cubilin was purified by lF-Cbl affinity chromatography as
previously described (Fyfe et al., 1991b) and transferred from SDS-PAGE gel to
PVDF membrane. In situ endoLys-C or trypsin digestions, separation of peptides
by microbore HPLC, and peptide sequencing were performed by the Protein
Chemistry Facility of the Worcester Foundation for Experimental Biology

(Shrewsbury, MA).

Canine kidney cDNA library screening

A canine CDNA library was constructed from RNA isolated from normal
canine pooled kidney proximal tubular epithelial cells, using Trizol (Life
Technologies, Bethesda, MD) according to the manufacturer’s instructions. Total
RNA was sent to Stratagene (La Jolla, CA) for construction of the library in the
bacteriophage lambda vector, Lambda Zap ll, using both random octamer and
oligo dT priming of the cDNA synthesis reaction. This library was screened with a
digoxigenin-Iabeled probe derived from the rat cubilin cDNA (a kind gift of Pierre
Verroust, INSERM, Pairs, FR). Hybridizing plaques were purified, and the 1.6 kb
cDNA insert of one (DX2) was sequenced, revealing 85% identity with the human
cubilin sequence. DX2 was labeled with [32P1a-dATP to high specific activity
using a random primer labeling kit (Boehringer Mannheim) and used as a probe
to hybridize to 6 HybondTM-N nylon membrane (Amersham Pharmacia Biotech)

lifts totaling 3X105 pfu from the canine kidney proximal tubule epithelial cDNA

46

library. Duplicate membrane sets were prehybridized in 6X SSC, 0.5% SDS, 5X
Denhardts solution, 20mM NaH2P04, and 100 jig/ml sonicated sperm DNA for 4
hrs at 65°C. Hybridization was performed in the same solution after adding the
denatured probe for 16 hours at 65°C. Membranes were washed twice with 6X
SSC at room temperature and twice with 2XSSC and 0.5% SDS for 30 min at
68°C. Positively hybridizing plaques were picked and rescreened using the
same probe for three more rounds. The true positive plaques were purified, and
in vivo excision of the phagemid pBluescript (SK-) was performed according to
Stratagene protocol. XL1-Blue bacteria were infected with the positive
phagemids and grown on LB agar plates containing 50 pg/ml ampicillin. Plasmid
DNA was purified from overnight MR 2000 cultures (MacConnell Research, San
Diego CA) using QIAGEN’s plasmid purification kit (QIAGEN Inc. Santa Clarita
CA).

PCR amplification of cubilin cDNA probes for sequential screening:

Sequence information obtained from the cDNA clones was used to design
primers, F3922, 5’-CAGGCAACAACTGGCAACACGGT-B’, R4907, 5’-
GCAGGGAACAATGGAGAGGATA-3’, for use in PCR reactions to amplify a
cDNA probe for sequential rounds of screening. The fifty pl final reaction volume
contained: 1X PCR buffer (Gibco BRL), 0.25 mM each dNTP, 2 mM MgClz, 0.25

uM each primer, 0.4 ng plasmid DNA template and 2.5 units Taq DNA
polymerase (Gibco/BRL). After a hotstart denaturation at 95°C for 5 min, the

reactions were carried out for 10 cycles of denaturation at 94°C for 1 min,

47

annealing at 62°C for 1 min, and extension at 72°C for 5 min. Then the annealing
temperature was decreased to 58°C for 26 more cycles. The amplified fragments
were size fractionated by agarose gel electrophoresis in 1X TAE and were gel
purified to use as probes for the next round of screening. The screening was
carried out using the same protocol and same amount of cDNA library as above.
For a third round of screening, 5 micrograms of canine kidney cortex total

RNA were reversed transcribed with oligo dT primer (Gibco) at 42°C to
synthesize first strand cDNA. 2 pl of the RT product was used as template to

amplify a cDNA probe using primers F5404, 5’-GGTCACTTGGTGGGGC-
GATACT-3’ and R6066, 5'-GTCCAGGGAAAGGATGTTGAGTT-3’ designed from
the 3’ end sequence of a newly cloned fragment. The PCR reactions were

performed under the same conditions as above for 31 cycles except using 67°C
annealing for 203 and 72°C extension for 2 min. The PCR product was gel

purified and used as probe to screen the canine cDNA library again.

In multiple rounds of screening with newly isolated insert cDNAs as
probes in each round, 10 clones (DX2-DX11) containing overlapping inserts were
selected and partially or fully sequenced on both stands. Most were sequenced
by primer walking, but DX11, containing an ~5 kb insert was sequenced using an
in vivo transposon-based kit for random insertions of primer sites (GPS-Genome
Priming System, New England Biolabs, Beverly, MA). All sequences were
generated by dideoxy termination cycle sequencing methods on an ABI 373A
Automated Sequencer (Applied Biosystems, Inc, Foster, CA). cDNA sequences

were assembled and analyzed using DNASTAR (DNASTAR, Madison, WI) or

48

SEQUENCHER software (Gene Codes Corp, Ann Arbor, MI) and have been
submitted to GenBank (accession #AF137068). Position 1 of the canine cubilin
cDNA refers to the first nucleotide of the full-length cloned cDNA. By this system,

the A residue of the first ATG is at position 74.

49

Results

Cloning of Canine Cubilin cDNA
Using a single rat cubilin cDNA fragment to screen a kZap cDNA library

derived from dog kidney proximal tubule epithelial cells, an initial 1.6 kb clone
encoding a portion of canine cubilin was identified. Two independent clones
which covered upstream and downstream sequences were identified using this
1.6 kb fragment as probe to screen the same canine cDNA library. A number of
other clones were identified using polymerase chain reaction-generated probes
for further screening. Fig. 2.1 schematizes five overlapping clones used to
construct the final cDNA. Canine cubilin cDNA contains11,282 bp including 73
bp of 5’ and 349 bp of 3’ untranslated sequence. There was a poly A addition
signal sequence 320 bp 3’ of the stop codon and 21 bp 5’ of the beginning of the
poly A tail. The nucleotide sequence was 85 % identical to the human and 74 %

identical to the rat cubilin sequences in GenBank.

Primary structure of canine cubilin:

The open reading frame (ORF) of 10,860 bp had a deduced amino acid
sequence of 3,620 residues representing a 397,435 Da protein (Fig. 2.2).
Alignment of the deduced amino acid sequence demonstrated overall amino acid
identity of 83 % with human and 70 % with rat cubilin. The primary structure of

canine cubilin, which contains a 110 amino acid N-terminal sequence; 8 EGF

50

.xcmmcow 5 30533.0. 5.53 am. 9: 9 .8352 ea: new
:25: 9: 9 30:52 e\emm mm; 8538 258.03: of. .33 < boa 2: Co @5563 9: Co .m an em ucm :88
noun 9.: Go .m an own 8533 .956 .5525 < 23 m mm; 99:. .8533 8529.5 .m ho an mvm ucmb ho

an MK mEmEoo 95.0 <200 of. .NmNF Tvvmm H3x0 cam ovmuoémmm H85 ”334.58 ”38.33 Uaxe Hmmomémxxwxnu
80:20 02» >9 omnooco 8:303 (200 an Nam. E. on... 6:33 >963 ocEno ho m:_:o_o <zno 3d 232".

8 Nam: :98

 

 

 

E: .m 8 ext. E: .m 8 E
e
.m E4 .m
9cm: 9:an :80 an 08.3
Pg
Boom TV $me
88
89%: ~83

 

0X0

 

.58
58

 

<zoo 5:330 @558 59.2 :2 Lot. 3985 9.2.20

51

Figure 2.2: Primary sequence of canine cubilin as deduced from cDNA
clones. Deduced amino acid sequence of 3,620 residues representing a
397,435 Da protein. Alignment of the deduced amino acid sequence
demonstrated overall amino acid identity of 83% with human and 70% with rat
cubilin. The arrow shows the furin cleavage site. The underlined sequence
represents the consensus sequence Arg-X-Arg/Lys-Arg for cleavage by the
trans-Golgi proteinase furin. The sequences verified by protein sequencing of
purified canine kidney cubilin are shaded.

52

MSSPFLWSLIILLTFAESNGEAGGFELQRQKRSiDEQQPBMATERGNLVFLVGSAQNIEFRTGSLGKIKLE
EEEEGECIHQIQKNKEDITDLKRSAVNVPQNISSQIHQLNsgyvopgREFQSLooTvoKchsSNPCQNGG
TCLNLHDSFFCICPSQWKGPLCSVDVNECQIYSGTPLGCQNGATCENTAGSYSCLCSPETHGPQCASKYDD
CEGGSKALCVHGICEDLVRVKADEPKYNCICDAGWTSPLNSSACVLDIDECNLQHAPCSPLVQCFNTQGSF
YCGACPTGWQGNGYSCQDIDECKINNGGCSVVPPVMCVNTLGSYHCQACPPGYQGDGRVCTVIDICSVNNG
GCHPEASCSSVLGSLPLCTCLPGYTGNGYGPNGCAQLSDTCLSHPCLNGQCIETVSGYLCKCESGWAGINC
TENINECLSNPCFNGGTCVDGVNAFSCECTRFWTGFLCQIPQQVCGGSLSGMDGSFSYMSPDVGYVHDVNC
FWVIRTEDRKVLRITFTFFQLESVNNCPHEFLQIHDGDSSAALQLGRFCGSVLPHELLSSNNALYFHLYSE
HFRSGRGFTIRWETQQPECGGILMGTYGSIKSPGYPGNYPPXRDCVWQVVTSPDLLITFTFGTLSLEHHDD
CSKDYLEIRDGPLYQDPSLGKFCTTLSVPPLQTTGPFARVHFHSDNQINDQGFHITYLTSPSDLHCGGNYT
DPEGLLSSDLSGPFTHNRQCIYIIKQPLGEQIQVNFTHVBLEGQSSCSQSHIEVRDDKILLGKVCDNETLP
HIKSIRNHIWIRLKIDASLVRASFRAVYQVACGGELTGEGVIRSPFYPNVYBGEgICRWTIHQPQSQVVIL
NFTAFGIESSAHCDTDYIEIGSSSILGSPENKKYCGTDIPLFITSVYNFLYVIFVKSSSTENHGFMAKFSA
ADLACGEILTESTGIIQSPGHPNIYPHGINCTWHILVQPGHLIHLIFRKFHLEFHYNCTNDYLEVYDTGSN
TYLGRYCGKSIPPSLTSSTNSLKLIFVADSDLAYEGFLINYEATDASSACMEDYTENSGTFTSPNFPNNYP
NNWKCIYRITVETSQQIALHFTNFALEEAIGGQCVADFVEIRDGGYETSPPLGTYCGSIPPPRIISHSNKL
WLQFTSDFLGSGPGFSAYWDGSLTGCGGNITTPTGVFTSPSYPMPYYHSSECYWLLKASHGSPFELEFEDF
HLEHHPNCTLDYLAVYDGPSTSSHLLSQLCGNEKPPVIRSTGDSMFLKFRTDEDQQGGGFLAKYQQTCRNV
VIVNRNYGILESIHYPNPYSDNQRCNWTIQATTGNTVNYTFLAFELENHINCSTDYLELYDGPRRMGRYCG
ADMPPTGSTTGSKLQVLFYTDGVGHQEKGFQMQWFIHGCGGELSGTTGSFSSPGYPNTYPPNKECIWYITT
APGSSIQLTIHDFDVEYHARCNFDVLEVYGGPDFHSPRITQLCSQRSSENPMQVSSTGNELAIRFKTDSSI
NGRGFNASWQAVPGGCGGIFQAPNGEIHSPNYPSPYRGNTDCSWVIRVERNHRILLNFTDFDLEPQDSCIT
AYDGLSSTTTRLASVCGRQQLTNPITSSGNSLFLRFQSGPSRQGRGFRAQFNQVCGGHILTNSFDTISSPL
FPAKYPNNQNCSWVIQAQPPFNHITLSFDHFGLESSTTCTQDFLEILDGDYDDAPLRGRYCGHSMPHPITS
FSSALTLRFVSDSRVNSDGFHATYAASSSACGGTFHMAEGIFNSPGYPEVYPSNVECVWNIVSSPGNRLQL
SFITFQLEDSQDCSRDFVEVREGNATGHLVGRYCGNVLPLNYSSIVGHILWIRFVSDGSGSGTGFQATFTK
IFGNDNIVGTHGKIASPLWPGRYPHNSNYQWIVNVNATQVIHGRILEIDIEGAQSCYYDKLRVYDGLGIHS
RLIGTYCGTQTTSFSSSRNSLTFQFSSDSSITGKGFLLEWFAVNASGGPLPTIATGACGGFLRTGDAPVFL
FSPGWPESYSNSADCTWLIQAPDSTVELNILSLDIEAQRTCDYDKLVIRDGDSNLAPQLAVLCGREIPGPI
RSTGEYMFIRFTSDFSITGAGFNASFHKSCGGYLHADRGIITSPQYPETYSPNLNCSWHVLVQSGLTIAVH
FEQPFQIPSGDSSCSQGDYLVLKNGPDIYSPPLGPYGRNGHFCGSRPSSTLFTSDNQMFVQFISDGSNGGQ
GFKIKYEAKSLACGGNIYIHDVNSAGYVTSPGHPNNYPQHADCNWLIAAPPGKLIRVQFEDQFNIEETPNC
VSNYLELRDGVDSNAPLLAKLCGRSLPSSQLSSGEVMYLRFRSDNSSTQVGFKIKYAIAQCGGRVTGQSGE
:ESSGYPILPYRDNSFCEWHLKGPSGHYLTIHFEDFHLQNSSGCEKDFVEIWENHTSGNLLGRYCGNTIPD
SIDTSSNVALVRFVTDGSVTASGFRLRFESSMEACGGELQGPTGTFTSPNYPNPNPHGRVCEWRIMVQEGR
RITLTFNNLRLEAHPSCYSEHVTIFNGIRNNSPQLEKLCGSVNASSEIKSSGNTMKVVFFTDGSRPFGGFS
ATYTSSEDAVCGGSLTHFPEGNFTSPGYNGVSNYSRNLNCEWTLSNPNQGNSSIYIHFEDFYLESHQDCQF
DVLEFRYGHADGPLMWELCGPSKPIVPLVIPYPEVWIHFVTNEHVEHVGFHAEYSFTDCGGIQLGESGVIA
SPNYPASYDSLTHCSWLLEAPQGFTITLTFSDFDIEDHATCAWDSVSVRNGGSPGSPIIGQYCGTSNPRTI
QSGSNQLVVIFNSDHSVQNGGFYATWNTQTLGCGGILHSDNGTIRSPHWPQNFPENSRCSWTVITHESKQL
EISFDNNFRIPSGDGQCQNSFVKVWAGTEEVAESLLATGCGNVAPGSILTPRNVFIAVFQSQETPAQGFSA
SFVSRCGGNFTNPSGYILSPNYPRQYDNNMNCTYIIEADPLSVVLLTFESFHLEARSAITGSCANDGVHII
RGSNLSSTPFATVCGNEILSPVTILGPVLLNFYSNAHTTDLGFKFNYKITSCGGVFNVSTGVIKSPAYSYS
DYPNNIYCLYTIVGRDDRVVQLKFSDFDVVPSTFCSQDYLAIYDGSNISDPLLGKFCGSNLPPNIKSSNHS
MLLVFKTDSFQTARGWKITFQQTLGPQQGCGGYLTGSDNTFASPDSDSNGRYDKNLNCVWFIIAPVNKLIK
LTFNTFALEAQSAMQRCIYDYVKLYDGDSENANLAGTFCGSTVPAPFISSGNFLTVQFVSDVTLEREGFNA
TYTTVDMPCGGTYNATWTPQSISSPNSSNPEVPLSMCMWFLEAPPHQQVKITVWALELHSQDCDQNYLEFR
DSPESNGSPGPQICGRNASATPTFYSSRSTAIVIFKSEVLNRNSRVGFTYQIAGCNREYNKAFGNLKSPGW
PDNYDNNLDCTVILTAPQNHTISLFFHSFGIBDSSECTHDFLEVRNGSDSSSPLFGTYCGTLLPDPIFSRN
NKLYLRFKTDSATSNRGYEIVWTSSPSGCGGTLYGDSGSFTSPGYPGTYPNNTDCEWAIIAPAGRPVTVTF
YFISIDDPGDCVQNYLILYDGPDANSPSFGPYCGADTNIAPFVASSHRVFIKFHAEYAVYPSAIRLTWDS.

53

domains and 27 CUB domains, is the same as seen in human and rat. All of the
other structural features of human and rat cubilin were conserved in the dog
protein, including a furin cleavage site after Ar932, N-terminal heptad repeats
recently suggested to mediate trimer formation, 154 Cys residues involved in the
77 disulfide bridges of 8 EGF and 27 CUB domains, and the first two cysteines
were missing in CUB domain 13. The only cysteine outside of the CUB domains
and EGF repeats is located in the 110-amino acid N-terminal sequence. Six
peptides derived by protein sequencing of purified canine cubilin were found to
match identically the deduced amino acid sequence. Three peptides derived by
endoLys-C digestion, KIKLNEEDLGEXLHQ, IDFQQPRMATERG, and
KLVDLERK, were all in the first 120 amino acids N-terminal to EGF domain 1.
Three tryptic peptides, PFYPNWPGER, VTGQSGIIESSGYPT,
VGNADGPLMXR were in CUB domains 4, 17, and 19, respectively. These
results confirmed the clones under study correspond to canine kidney cubilin

cDNA.

54

Discussion

At the start of this porject, no sequence information on canine cubilin
(IFCR) was available except those 6 peptides derived from the amino acid
sequencing of purified canine kidney cubilin. We began with immunoscreening by
using polyclonal rabbit anti-dog cubilin antibody derived from the immunization of
rabbits with IF-Cbl affinity column purified canine renal cubilin (Fyfe et al.,
1991a). The cDNA library used for screening was constructed in the Lambda

ZAP II® expression vector, and protein expression can be induced by
isopropylthio-B-D-galactoside (IPTG). Nylon membranes presoaked with IPTG

were used to lift the phage plaques and induce the expression of inserts. Then
the membranes were screened with rabbit anti-dog cubilin antibody and sheep
anti-rabbit 196 antibody alkaline phosphotase conjugate sequentially (Broome
a‘nd Gilber, 1978). After four rounds of screening, positive clones were selected
and excised into pBluescript phagemids and characterized.

Totally, 12 individual clones were purified and sequenced. Sequencing
results showed that they corresponded to 8 different genes. Using these
sequencing to do Genbank database searches, three matched with a previously
characterized genes, such as mitosin, novel GTPase Rab 28 and HIV “TATA”
element modulatory factor. Four of them matched anonymous genes with
different degrees of similarity, and one had no match to any known gene.
Subsequently, alignment between the clones identified by immunoscreening and
the canine cubilin cDNA clone showed no similarity of any of those genes to

cubilin.

55

Cubilin is a giant receptor with unique and important secondary and
tertiary structures which provide multiple potential binding functions. These
secondary and tertiary structures are also critical for anti-dog cubilin antibody
recognition since the antibody bound poorly to 2-mercaptoethanol reduced
cubilin in either ileum or kidney western blots (see chapter 3 for detail). The
maximum cDNA length of the inserts in our cDNA library is about 5-6 kb which is
only half of the full length of cubilin. The epitopes formed by these partial protein
fragments might not be proper enough to be recognized by our anti-cubilin
antibody. This may be why immunoscreening of canine cubilin from that canine
kidney cDNA library failed. In addition, the inserts had been randomly ligated to
the vector during cDNA library construction. Library directionality is also critical
for immunoscreening.

The present data provide the molecular characterization of the primary
structure of canine cubilin, previously known as the intestinal receptor for IF-Cbl.
The predicted canine cubilin structure showed the same unique tandem domains
as human and rat cubilin, which suggests that they are critical for its functions.
The 8 EGF domains at the N-terminus of canine cubilin showed the conserved
EGF-like domain signature which is critical for protein-protein interactions and
biological function (Blomquist et al., 1984; Rao et al., 1995). Two of the EGF
domains contain the discontinuous consensus sequence for calcium-binding.
Calcium depletion by EDTA can completely prevent IF-Cbl binding to cubilin
which indicates that the conformational change induced by calcium binding to the

EGF-like domains may be important for the IF-Cbl interacting with cubilin.

56

The CUB domain consists of 110 amino acids defining a characteristic
hydrophobicity pattern predicted to form anti-parallel B-barrels (Bork and
Beckmann, 1993). The four conserved cysteines, generally thought to form two
disulfide bridges, are found in all but domain 13 of cubilin that lacks the first two
cysteines. This feature can be found in all three species and suggests that this
domain may have a more relaxed structure than the other CUB domains. In
human, but not in dog or rat, the last two cysteines in domain 6 were also
missing, which may indicate the unique structure in this region for human.

Molecular dissection of rat cubilin showed that the N-terrninal cubilin
region conveys membrane association, and the CUB domain region convey
ligand binding (Kristiansen et al., 1999). Further dissection of active binding
fragments localized the binding site for IF-Cbl to CUB domains 5-8. Two
mutations identified in cubilin genes of human patients with I-GS are both related
to the critical region defined for lF-Cbl binding (Aminoff et al., 1999). The FM2
mutation, suggested to activate a cryptic intronic splice site leading to truncation
of the receptor, was defined in CUB domain 6. Most likely the FM2 mutation does
not produce a stable cubilin protein and is functionly null. A P/L subsitution (FM1)
may cause a structural change specifically impairing IF-Cbl binding and is
located in CUB domain 8.

The canine cubilin cDNA we cloned in this study provided the basis for
mutation identification in the affected dog cubilin gene and further

characterization of cubilin protein, including in vivo expression of large quantities

57

of the protein. The full-length cubilin cDNA also provided the basis for genomic

structure analysis for canine cubilin.

58

References

Aminoff M, Carter JE, Chadwick RB, Johnson C, Grasbeck R, Abdelaal MA,
Broch H, Jenner LB, Verroust PJ, Moestrup SK, Chapelle A, and Krahe R. (1999)
Mutations in CUBN, encoding the intrinsic factor-vitamin B12 receptor, cubilin,
cause hereditary megaloblastic anaemia 1. Nat Genet 21: 309-313.

Bork P and Beckmann G. (1993) The CUB domain. A widespread module in
developmentally regulated proteins. J Mol Biol 231: 539-545.

Blomquist MC, Hunt LT, and Barker WC. (1984) Vaccinia virus 19-kilodalton
protein: relationship to several mammalian proteins, including two growth factors.
Proc Nat! Acad Sci USA 81: 7363-7367.

Broome S and Gilbert W. (1978) Immunological screening method to detect
specific translation products. Proc Nat! Acad Sci U S A 75: 2746-27499.

Fyfe JC, Jezyk PF, Giger U, and Patterson DF. (1989) Inherited selective
malabsorption of vitamin B12 in giant schnauzers J Am Anim Hosp Assoc 25:
533-539.

Fyfe JC, Giger U, Hall CA, Jezyk PF, Klumpp SA, Levine JS, and Patterson DF.
(1991a) Inherited selective intestinal cobalamin malabsorption and cobalamin
deficiency in dogs. Pediatr Res 29: 24-31.

Fyfe JC, Ramanujam KS, Ramaswamy K, Patterson DF, and Seetharam B.
(1991b) Defective brush-border expression of intrinsic factor-cobalamin receptor
in canine inherited intestinal cobalamin malabsorption. J Biol Chem 266: 4489-
4494.

Kozyraki R, Kristiansen M, Silahtaroglu A, Hansen C, Jacobsen C, Tommerup N,
Verroust PJ, and Moestrup SK. (1998) The human intrinsic factor-vitami B12
receptor, cubilin: molecular characterization and chromosomal mapping of the
gene to 10p within the autosomal recessive megaloblastic anemia (MGA 1)
region. Blood 91 (10): 3593-3600.

Kozyraki R, Fyfe J, Kristiansen M, Gerdes C, Jacobsen C, Cui S, Christensen El,
Aminoff M, Chapelle A, Krahe R, Verroust PJ, and Moestrup S. (1999) The
intrinsic factor-vitamin B12 receptor, cubilin, is a high-affinity apolipoprotein A-I
receptor facilitating endocytosis of high-density lipoprotein. Nat Med 5: 656-661.

Kristiansen M, Kozyraki R, Jacobsen C, Nexo E, Verroust PJ, and Moestrup SK.
(1999) Molecular dissection of the intrinsic factor-vitamin B12 receptor, cubilin,
discloses regions important for membrane association and ligand binding. J Biol
Chem 274(29): 20540-20544.

59

Kuivaniemi H, Tromp G, and Prockop DJ. (1991) Mutations in collagen genes:
causes of rare and some common diseases in humans. FASEB J 5: 2052-2060.

Lehrman MA, Schneider WJ, Brown MS, Davis CG, Elhammer A, Russell DW,
and Goldstein JL. (1987) The Lebanese allele at the low density lipoprotein
receptor locus. Nonsense mutation produces truncated receptor that is retained
in endoplasmic reticulum. J Biol Chem 262: 401-410.

Medeiros-Neto G, Kim PS, Yoo SE, Vono J, Targovnik HM, Camargo R, Hossain
SA, and Arvan P. (1996) Congenital hypothyroid goiter with deficient
thyroglobulin. Identification of an endoplasmic reticulum storage disease with
induction of molecular chaperones. J Clin Invest 98: 2838-2844.

Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D,
Pontillon F, Goda F, Christensen El, Hammond TG, and Verroust PJ. (1998) The
intrinsic factor-vitamin B12 receptor and targek of teratogenic antibodies is a
megalin-binding peripheral membrane protein with homology to developmental
proteins. J Biol Chem 273(9): 5235-5242.

Moolenaar CE, Ouwendijk J, Wittpoth M, Vlfisselaar HA, Hauri HP, Ginsel LA,
Naim HY, and Fransen JA. (1997) A mutation in a highly conserved region in

brush-border sucrase-isomaltase and lysosomal alpha-glucosidase results in
Golgi retention. J Cell Sci 110: 557-567.

Rao Z, Handford P, Mayhew M, Knott V, Brownlee GG, and Stuart D.(1995) The
structure of a Ca(2+)-binding epidermal growth factor-like domain: its role in
protein-protein interactions. Cell 82: 131-141.

Repaske DR, Summar ML, Krishnamani MR, Gultekin EK, Arriazu MC, Roubicek
ME, Blanco M, Isaac GB, and Phillips JA 3rd (1996) Recurrent mutations in the
vasopressin-neurophysin II gene cause autosomal dominant neurohypophyseal
diabetes insipidus. J Clin Endocrinol Metab 81: 2328-2334.

Wetterau JR, Aggerbeck LP, Bouma ME, Eisenber C, Munck A, Hermier M,
Schmitz J, Gay G, Rader DJ, and Gregg RE. (1992) Absence of microsomal
triglyceride transfer protein in individuals with abetalipoproteinemia. Science. 258
(6): 999-1001.

60

CHAPTER 3

CUBILIN EXPRESSION AND POST-TRANSLATIONAL MODIFICATIONS IN

THE CANINE GASTROINTESTINAL TRACT

61

Introduction

Cobalamin (vitamin B12, cbl) is an essential micronutrient which all higher
animals, including humans, must obtain from dietary sources (Kapadia, 1995). In
the proximal gastrointestinal tract, cbl binds to intrinsic factor (IF), a 50 kDa
glycoprotein produced by gastric mucosa in many species but mainly by
pancreas in dogs and cats (Batt et al., 1989; Simpson et al., 1993; Fyfe, 1993).
Endocytosis of the IF-cbl complex is mediated by a specific receptor expressed
in intermicrovillar clefts of apical membrane brush-borders on villus tip
enterocytes of the distal small intestine (Hammond et al., 1994; Levine et al.,
1984). Binding of cbl by IF and brush-border expression of the receptor on ileal
enterocytes are required for gastrointestinal cbl absorption.

In addition to ileum, lF-cbl binding activity and immunoreactive receptor
protein are expressed in renal proximal tubule, yolk sac, and placental epithelia,
tissues exposed to very little, if any, lF-cbl. Two lines of investigation converged
recently when the receptor mediating lF-cbl endocytosis was shown to have
immunologic and functional identity with gp280 (Seetharam et al., 1997), a
glycoprotein known as the "target of teratogenic antibodies" and demonstrated to
be a recycling plasma membrane receptor in rat yolk sac epithelium (Le Panse et
al., 1997).

Cubilin cDNAs of 11.3-11.6 kb encoding a highly conserved 3620 amino
acid sequence were cloned, first from rat yolk sac and kidney (Moestrup et al.,
1998) and subsequently from human (Kozyraki et al., 1998) and dog (Xu et al.,

1999) kidney. The encoded intrinsic factor-cobalamin binding receptor, previously

62

called IFCR, has been given the name cubilin in recognition of its unique protein
structure, including a tandem arrangement of 27 CUB domains, and recognition
that it is a multiligand receptor. In addition to lF-cbl, recently identified cubilin
ligands include apolipoprotein-A1 and high-density lipoproteins (Hammad et al.,
1999; Kozyraki et al., 1999), albumin (Birn et al., in press), and immunoglobulin
light chains (Batuman et al., 1998). Northern and western blots of rat tissues
demonstrated cubilin expression in kidney, small intestine, and yolk sac, but not
in liver (Moestrup et al., 1998). Renal cubilin appears to be a peripheral
membrane protein because it lacks a transmembrane domain or
glycophospholipid anchor addition signal in the deduced amino acid sequence.
Consistent with this, purified canine intestinal cubilin behaved as a peripheral
membrane protein when reconstituted in artificial liposomes (Seetharam et al.,
1981b). Published evidence suggests that an association with megalin (gp330),
an apical plasma membrane receptor of the low-density lipoprotein receptor
family, may mediate cubilin, and therefore IF-cbl, endocytosis and subsequent
cubilin recycling to the apical membrane (Birn et al., 1997; Moestrup et al., 1998,
Hammad et al., 2000). The size of detergent solubilized and affinity purified
canine intestinal IFCR has been determined variously as 180 to 230 kDa by gel
filtration, SDS-PAGE, or amino acid analysis (Seetharam et al., 1981a;
Seetharam et al., 1982). In contrast, the cDNA of canine renal cubilin suggests a
molecular mass of at least 400 kDa (Xu et al., 1999). For this reason, some
investigators have questioned the identity of intestinal IFCR and renal cubilin

(Guéant et al., 1999).

63

Intestinal cbl absorption is limited by cubilin expression in various normal
and disease states, but the molecular nature of cubilin regulation is unknown. As
an initial step to investigate regulation of intestinal cubilin expression, we have
examined differential expression of cubilin along the crypt to villus axis in dog
ileum and regional expression of cubilin/IFCR ligand-binding activity,
immunoreactivity, and mRNA along the longitudinal axis of the canine
gastrointestinal tract. The longitudinal study demonstrated that constitutive cubilin
expression in intestine is determined by mechanisms controlling cubilin mRNA
levels. Cell fractionation studies indicate that canine intestinal cubilin behaves as
a peripheral membrane protein and that apparent differences in size of intestinal

and renal cubilin are due to tissue-specific post-translational modifications.

64

Materials and Methods

Reagents The following were purchased from commercial sources: [57Co]
radiolabeled cyanocobalamin (300pCi/nmol) from Amersham International
(Buckinghamshire, England); anti-rabbit IgG-agarose, anti-rabbit lgG-alkaline
phosphatase conjugate, p-nitro blue tetrazolium (NBT), and 5-bromo-4-chloro-3-
indolyl phosphate (BCIP) from Sigma lmmunochemicals (St. Louis, MO);
disodium 2-chloro-5-(4-methoxyspiro{1,2-dioxethane-3,2'-(5'-
chloro)tricyclo[3.3.1 .13'7]decan}-4-yl)-1-phenyl phosphate (CDP-Starm) from
Roche Molecular Biochemicals (Indianapolis, IN); TRIzol® reagent from Life
Technologies/GibcoBRL (Gaithersburg, MD); and endoglycosidase H (endo H)
and peptide: N-glycosidase F (PNGase F) from New England Biolabs (Beverly,
MA). Rat stomach IF was prepared and radiolabeled by formation of lF-[57Co]cbl

complexes as described (34).

Animals and tissue isolation Animal husbandry was provided by MSU University
Laboratory Animal Resources, and animal-use protocols were approved by the
All-University Committee for Animal Use and Care. Twelve-20 kg body weight
dogs were fasted overnight. Surgical anesthesia was induced by intravenous
injection of sodium thiopental (15 mg/kg) and maintained by isoflurane inhalation.
Five cm segments of small intestine were resected at sequential 30 cm intervals
beginning 10 cm distal to the pyloroduodenal junction in some dogs and 10 cm
proximal to the ileocolic junction in others. A 5-cm segment of descending colon

and 25 cm2 of gastric fundus were also resected. The mucosa of each segment

65

was scraped from the underlying muscularis and snap frozen in liquid nitrogen. In
the distal small intestine of some dogs, samples of mucosa overlying Peyer's
patches were scraped and frozen separately from the adjacent absorptive
mucosa. Other tissues were removed simultaneously with interruption of the
blood supply of each and snap frozen. After surgery, each dog was euthanized,
without recovery from anesthesia, by intravenous overdose of pentobarbital

sodium (85 mg/kg body wt).

Isolation of dog intestinal epithelial cells The methods for intestinal epithelial cell
isolation used in this study were modifications of the method described by Weiser
(Weiser et al., 1973). Fifteen cm-long dog ileum sections were resected starting
at 40 cm proximal to the ileocecal junction and were immediately rinsed with

37°C phosphate buffered saline, pH 7.4, containing 40 mM EDTA and 2 mM

DTI’. The intestinal segments were filled with the same buffer, clamped at each
end, and incubated with shaking for 10 min in a bath of PBS at 37°C. The
Iumenal solutions were collected and refilled at intervals of 10 min. After 3
fractions, the DTT concentration in the PBS buffer was reduced to 0.5 mM and 7
more fractions were collected. The collected cell suspension of each fraction was
aliquoted into two tubes and a cells were pelleted. The cells in one of the tubes
were used for protein and alkaline phophatase assays. The cells were washed
with the PBS buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF) and 2
mM benzamidine and frozen as cell pellet. Total protein was determined by the

method of Lowry (Lowry et al., 1951), and the alkaline phosphatase activity was

66

determined at pH 9.2 using p-nitrophenyl phosphate as substrate (Garen A. and
Levinthal C. 1960). The other aliquot of cells was used for RNA assay. They
were washed with PBS buffer and Iysed in the Trizol reagent (GibCo)

immediately.

Ligand-binding activity determination Resected samples of gastrointestinal
mucosa were thawed and homogenized on ice in 9 volumes of cold 20 mM Tris-
HCI, pH 7.5, containing 150 mM NaCl (TBS), 5 mM KCI, protease inhibitor
cocktail (2 mM each of benzamidine and N-ethyl maleimide, 1 mM
phenylmethylsulfonyl fluoride (PMSF)) and 7.5 mM NaN3. Total protein
concentrations were determined by the method of Lowry (Lowry et al., 1951).
Ca2*-specific receptor binding of a saturating amount (370 fmol) of lF-[57Co]cbl
was determined in aliquots of mucosal homogenates containing 1.25-25 mg of
total protein by a rapid microcentrifuge method essentially as described
(Hagedorn and Alpers, 1977). Under these conditions, binding was never greater
than 95 fmol (25% of added ligand), and both free and bound ligand were

measured after the incubation period.

Immunoprecipitation and westem analysis Cubilin from canine renal cortex was
used to produce anti-cubilin antiserum for 2 reasons: 1) cubilin is far more
abundant in renal cortex than in ileal mucosa (Fyfe et al., 1991), and 2) in order
to examine the cross immunoreactivity of canine ileal and renal cubilin. Canine

renal cubilin was affinity purified to homogeneity as previously described (Fyfe et

67

al., 1991), and rabbits were immunized by subcutaneous injection of the protein
in oil/water emulsion. Rabbit serum was harvested, and anti-cubilin activity was
characterized. Ten pL of the polyclonal serum blocked 86% of lF-cbl binding to
700 fmol of the purified renal cubilin preparation used for immunizations. The
antiserum detected a single major band on western blots of renal cortex
homogenates (see Fig 4).

Resected intestinal mucosal samples were thawed and homogenized on
ice in 4 volumes of cold 50 mM Na phosphate buffer, pH 7.4, containing 100 mM
NaCl, 20 mM ethylenediaminetetraacetic acid (EDTA), and protease inhibitor
cocktail. Triton-X-100 and heparin were added to the homogenates to 2 % and
250 U/mL, respectively. They were incubated at 4°C overnight with constant
agitation and were then centrifuged at 20,000 x g for 40 min. Aliquots of the
supernatants containing 6 mg of total protein were incubated for 4 h at 4°C after
sequential additions of 20 iiL preimmune rabbit serum and 30 pL anti-rabbit lgG
agarose beads and then centrifuged at 12,000 x g for 2 min to remove
nonspecific binding proteins. Cubilin was immunoprecipitated from the precleared
supernatants by sequential addition of 20 pL anti-canine cubilin rabbit serum and
30 uL anti-rabbit IgG agarose beads, again with 4 h incubation at 4°C after
addition of each reagent. The protein-agarose bead complexes were washed 3
times in TBS containing 0.4 % Na deoxycholate and 1 mM PMSF and boiled in
20 pL non-reducing SDS-PAGE loading buffer. Solubilized proteins were
separated by SDS-PAGE on 5 % gels and transferred to polyvinylidene difluoride

membranes. Membranes were blocked in TBS with 0.2 % Triton-X-100 (TBST)

68

and 2% gelatin at 22°C for 4 h and incubated at 4°C overnight in TBST with 0.4
% BSA and anti-canine cubilin rabbit serum (1:20,000). After copious washing in

TBST, membranes were incubated at 4°C for 3 h in TBST with 0.4 % BSA and

anti-rabbit IgG-alkaline phosphatase conjugate (1 :30,000). After copious washing
in TBST, immunoreactive proteins were detected by incubating membranes in
NBT/BCIP. In preliminary re-immunoprecipitation experiments, it was determined
that the homogenate supernatants were entirely immunodepleted of detectable
cubilin by the method described above. Proteins from renal cortex homogenates

were detected on western blots as above, but without prior immunoprecipitation.

Northem analysis Frozen tissue samples and cell fractionation pellets were

homogenized in TRlzol® reagent and RNA was isolated according to the
manufacturer's protocol. Thirty pg of total RNA from each sample were

electrophoresed on 1% agarose-formaldehyde gels and transferred to nylon
membranes by standard protocols (Sambrook et al., 1989). A loading control
probe of canine gcheraIdehyde-3-phosphate dehydrogenase (GAPDH) cDNA
was produced by reverse transcription coupled polymerase chain reaction (RT-
PCR) of canine testicular RNA as previously described (Fyfe et al., 1999).
Similarly, an endogenous ~15 kb marker was produced by hybridizing blots to a
random-prime labeled probe of partial canine megalin (gp330) cDNA produced
by RT-PCR of canine kidney cortex RNA. The probe was amplified using PCR
primers 5'-GGCTCCTACATCTGTAAGTGTGC-3' and 5'-
CACAGAC'I'I'GTTGGTTCCATCC-B', corresponding to sequences at nucleotide

69

positions 9618-9640 and 10157-10178 of the human megalin cDNA sequence,
respectively (GenBank accession no. U33837). The PCR product was confirmed
to be megalin cDNA by sequencing and demonstrated 88 % identity with the
corresponding portion of the human megalin cDNA. The crypt specific control
probe was derived from RT-PCR of dog ileal RNA by primers, F5’-
CAAGCCAGTGTGAGCAGGAGCTA-3’ and R5’-AAGCGGTTCAGGAA-
GTGGAAGG-3’, designed from the conserved region of mouse CRP-ductin and
human DMBT1. Blots were hybridized to random-prime labeled, partial canine
cubilin cDNA, extending between positions 3922 and 4907 of the sequence
(Genbank accession no. AF 137068), marker, and/or control probes and washed
by standard methods (Sambrook et al., 1989). Hybridizing transcripts were
detected by exposing blots to autoradiographic film or to a storage phosphor
screen. The relative intensity of probe hybridization between samples was
determined quantitatively from the storage phosphor screen with the

ImagquantTM program (Molecular Dynamics, Inc, Sunnyvale, CA).

Tissue fractionation, purification of IF-cbl binding proteins, and oligosaccharide
analysis Frozen ileal mucosa was thawed and homogenized in cold TBS with
protease inhibitors, and an aliquot was centrifuged at 40,000 x g for 1 hr. The
resulting supernatant was again centrifuged at 150,000 x g for 2 hr, and lF-cbl
binding proteins were affinity purified from each fraction and western blotted as
described below. In separate experiments, frozen canine ileal mucosa or renal

cortex was thawed in 9 vol of cold 2 mM Tris-HCI, pH 7.4, containing 50 mM

70

mannitol and 2 mM benzamidine (TMB) and homogenized in an iced blender on
high in four 15 sec bursts. An aliquot was centrifuged at 4°C at 40,000 x g for 1 h.

The pellet comprising total membranes was rehomogenized in TBS with 1 mM
PMSF and 2 mM benzamidine (TBSPI). The supernatant was made 10 mM Tris-
HCI, 140 mM NaCl, and 1 mM PMSF by addition of concentrated solutions of
each. The remaining 10% TMB homogenate of ileal mucosa was diluted to 1 %
by addition of 9 vol of cold TMB and homogenized again. The homogenate was
made 10 mM CaCIz by addition of 0.01 vol of 1 M CaClz, stirred, incubated on ice
for 10 min, and centrifuged at 2,500 x g for 15 min. The membrane pellet was
homogenized in 9 vol of cold TBSPI, and the supernatant was recentrifuged at
16,000 x g for 40 min. The resulting brush-border enriched pellet was
homogenized with 10 strokes in a prechilled Potter-Elvehjem in 4 vol of cold TBS
with 2 mM benzamidine. The final supernatant was made 10 mM Tris-HCI, 140
mM NaCl, and 1 mM PMSF by addition of concentrated solutions of each. Total
protein and alkaline phosphatase activity of each fraction were determined as
described previously (Fyfe et al., 1991).

Each of the above fractions was made 1 % in Triton-X-100 and agitated

continuously at 4°C overnight. Each was centrifuged at 40,000 x g for 1 h, and

the supernatants were made 5 mM C3012. IF-cbl binding proteins were purified
from each fraction by affinity chromatography on individual 0.4 mL columns of rat
gastric lF-cbl-agarose bead matrix, essentially as described for purification of
canine renal cubilin (Fyfe et al., 1991). The proteins were eluted from the

columns in TBS, pH 5.0, containing 5 mM EDTA and 10 mM CHAPS. CaClz and

71

Tris base were added to each eluted fraction to 5 mM and pH 7.4, respectively,
and the eluted proteins were dialyzed against 1 mM benzamidine at 4°C
overnight. Aliquots of the dialyzed proteins were reduced in volume by vacuum
centrifugation, separated on 5% SDS-PAGE gels and western blotted as above,
except that immunoreactive proteins were detected by incubation of the blot with
CDP-Starm, a chemiluminescent alkaline phosphatase substrate, and exposure
to autoradiographic film. Control blots included both substitution of the primary
antibody with nonimmune rabbit serum and elimination of the primary antibody
altogether. Analysis of asparagine-Iinked (N-linked) oligosaccharides was by
endo H or PNGase F digestion of the dialyzed proteins as previously described
(Fyfe et al., 1991), electrophoretic separation by SDS-PAGE, and western
blotting. The only change from the manufacturer's protocol was that samples
were denatured by boiling in 0.5 % SDS without 2-mercaptoethanol prior to

glycosidase digestion.

Data analysis For comparison of intestinal tracts of different lengths, the position
of each small intestine mucosal sampling site was expressed as the percentage
that the length of gut from pyloroduodenal junction to sampling site was of the
entire small intestinal length measured between the pyloroduodenal and ileocolic
junctions. Because phosphoimager analysis of northern blots was not in
physiologically relevant units, quantitation of cubilin probe hybridization to RNA
from different mucosal samples was normalized to equal GAPDH hybridization

values and expressed as a fraction of the maximum hybridization determined in

72

each dog intestine. In order to compare specific lF-cbl binding activity and cubilin
mRNA expression, the IF-cbl binding data from each sample site were expressed
similarly in figure 3. Small intestinal distributions of cubilin ligand binding and
mRNA between the most proximal nonzero values and the maximal values were
examined by linear regression analysis. Densitometry of autoradiographic film
exposed within the linear response range was performed on a Gel Doc 2000 with

Quantity One® quantitation software from Bio-Rad (Hercules, CA).

73

Results

Gastrointestinal cubilin expression

Tissue specificity of cubilin mRNA expression was evaluated in multiple
canine tissues by northern blot hybridization (Fig 3.1) using a 1 kb portion of the
canine cubilin cDNA cloned from renal proximal tubule cells as probe (Xu et al.,
1999). Specific hybridization of the cubilin probe was detected in ileum but not in
any other portion of gastrointestinal mucosa, salivary gland, or liver. Cubilin
probe hybridization was strong in kidney cortex but was nearly undetectable in
kidney medulla. This pattern of expression was consistent with findings in rat
tissues (Moestrup et al., 1998), although the canine yolk sac equivalent was not
examined. In ileum and kidney cortex a strongly hybridizing transcript of 11-12 kb
and a larger but more faintly hybridizing transcript were observed consistently.
On overexposed blots of ileal RNA one or two even larger transcripts were
sometimes observed.

Sequential portions of gastrointestinal mucosa were examined for
expression of cubilin ligand-binding activity, immunoreactivity and mRNA
expression. Specific activity of EDTA inhibited IF-cbl binding in multiple segments
of gastrointestinal mucosa, from stomach to colon, of 5 dogs is shown in Fig
3.2A. In each dog, lF-cbl binding activity was detected in the distal one-half of
small intestine, with highest specific activity in the mucosa at 90 % of the
distance from pyloroduodenal to ileocolic junctions, corresponding to 30-40 cm
proximal to the ileocolic junction. Ligand-binding activity was reduced to 0.5-0.7

of the peak value in the most distal segment of small intestine in each dog. In 4

74

Saliv. gland
Kid. medulla

Esophagus

Brain
Lung

Liver
Spleen
:_' Kid. cortex

Stomach
‘ Duodenum

_, Ileum
Colon

  

Cubilin

Figure 3.1: Tissue specificity of cubilin mRNA expression. Thirty pg of total
RNA from each of several canine tissues were electrophoresed and blotted.
Membranes were hybridized simultaneously to cubilin and GAPDH cDNA
probes. Equality of loading was determined by examination of the 28 S
ribosomal RNA on the ethidium bromide stained gel. The ~15 kb position (top
arrow) was determined from subsequent hybridization to a canine megalin
(gp330) cDNA probe. The lower arrow represents a 9.5 kb marker.

75

Figure 3.2: Codistribution of intestinal cubilin function, immunoreactive
proteins, and mRNA. The data in panels A, B and C are aligned vertically. In
each panel, the mucosal sampling site is designated by the percentage that the
distance from pyloroduodenal junction to sampling site was of the full small
intestinal length. Panel A. Mucosal IF-cbl binding specific activity was
determined in intestinal segments of 5 dogs. Panel B. Cubilin proteins were
immunoprecipitated from mucosal homogenates, separated by SDS-PAGE, and
detected on blots with anti-renal cubilin antibody. The western blot of a
representative dog is shown. IC indicates cubilin determined to be intracellular
forms at different stages of oligosaccharide processing, and BB indicates a
proteolytically modified brush-border form. Panel C. Thirty pg of total RNA from
each mucosal segment were electrophoresed, blotted and hybridized
simultaneously to cubilin and GAPDH cDNA probes. The northern blot of a
representative dog is shown. St indicates stomach mucosal RNA.

76

h
0

  

lF-cbl binding

(fmollmg prot.)
N a
O O

.L
O

 

O

200 kDa

 

C

Cubilin

 

7 18 30 41 52 63 74 85 96 St Colon

 

GAPDH

77

other dogs, lF-cbl binding activity of mucosa overlying distal small intestinal

Peyer's patches (24 i 6.6 fmoI/mg protein) was compared to that of the adjacent
absorptive mucosa (47 i 1.4 fmol/mg protein). Thus, reduced ligand-binding

activity in the most distal ileal segment was probably due to modification of the
epithelium overlying the nearly-circumferential Peyer's patch observed in the last
10-15 cm of canine small intestine. Cubilin ligand-binding activity was not
measurable in stomach or colonic mucosa.

Western blots of proteins immunoprecipitated with anti-renal cubilin
antiserum from multiple segments of gastrointestinal mucosa from each dog
demonstrated immunoreactive cubilin only in the fractions exhibiting ligand-
binding activity and demonstrated a pattern of cubilin immunoreactivity which
paralleled the observed pattern of ligand-binding (Fig 3.2B). Four
immunoreactive proteins were observed on nonreducing gels in mucosa of distal
small intestinal segments, but none were observed in proximal small intestinal,
stomach, or colonic mucosa. There was a band observed just under 200 kDa
(labeled BB form) and 2 bands observed >400 kDa (labeled IC forms). The band
observed at high apparent Mr (labeled ?) was also observed in canine renal
cortex (Fig 3.4, lanes 4 and 5). None of these 4 cubilin species was observed
when the immunoprecipitation or immunodetection on western blots was
performed with nonimmune serum (not shown). Thus, small intestinal expression
of immunospecific cubilin corresponded to expression of cubilin ligand-binding

activity.

78

Cubilin mRNA expression in multiple segments of gastrointestinal mucosa
was examined in each dog by northern blot (Fig 3.20). As demonstrated in Fig
3.1, no hybridizing CUBN transcripts were observed in stomach (St) or colonic
mucosa. Hybridization of the cubilin probe to mucosal RNA was observed only in
samples isolated from the distal half of small intestine, and highest expression
was found in mucosa at 85-95% of the distance from pyloroduodenal to ileocolic
junctions. In all samples in which hybridization of the probe was observed, a
strongly hybridizing transcript at 11-12 kb and a more faintly hybridizing transcript
at 13-14 kb were seen, just as in the multiple tissue blot.

In order to compare the small intestinal distributions of the cubilin IF-cbl
binding function and mRNA in a quantitative way, both sets of data for each dog
were expressed as fractions of the peak activity observed (Fig 3.3). In the
proximal to distal direction, lF-cbl binding (upper panel) and cubilin mRNA (lower
panel) in each dog were first detectable in mucosa at 45-50 % and both
parameters were maximal at 85-90 % of the small intestinal length. Regression
analysis of the combined data of all 5 dogs in the interval from first detectable to
maximal indicated that both parameters increased linearly (slopes of 2.1 and 2.0,
respectively) with comparable variance (R2 values of 0.82 and 0.78,
respectively). Thus, in this group of dogs, cubilin function and mRNA expression
in the small intestine were highly correlated, both in distribution of detectable

expression and the position of maximal expression.

79

Fraction of Peak
IF-cbl Binding Activity

 

 

Fraction of Peak
mRNA Level

 

 

 

 

I
20 40 60 80 100

Figure 3.3: Codistribution of quantitated intestinal cubilin ligand-binding
activity and mRNA. Top panel. The IF-cbl binding specific activities
represented in Fig 3.2A, were recalculated as fractions of the peak value in the
intestine of each dog. Linear regression of the data between the first nonzero
values and the peak values of all 5 dogs had a correlation coefficient of 0.82.
Bottom panel Relative cubilin mRNA hybridization in each mucosal sample
was determined by exposure of northern blots to a storage phosphor screen.
The data are plotted as fractions of the peak value found in each dog. As in the
top panel, linear regression of the data between the first nonzero values and the
peak values of all 5 dogs had a correlation coefficient of 0.78.

80

Ileal cubilin membrane association

To determine whether endogenous cubilin in ileal mucosa behaves as a
weakly bound peripheral membrane protein in vitro, as has been reported for
renal cubilin (Moestrup et al., 1998), we examined the distribution of cubilin after
homogenization and fractionation of ileal mucosa without detergent. Ileal mucosa
was homogenized in neutral hypotonic buffer, and an aliquot was centrifuged at
40,000 x g separating soluble and total membrane fractions. Apical microvillus
brush-border membranes were enriched from the remaining homogenate by
CaCIz aggregation of membranes followed by differential centrifugation, with
recovery of brush-border membranes in the 16,000 x g pellet (Kessler et al.,
1978). In a typical fractionation of 1 g of ileal mucosa, lF-cbl binding activity was
3.1 and 3.9 pmol (0.10 and 0.18 pmol/mg protein), in the soluble and total
membrane fractions, respectively, and 0.59 pmol (0.41 pmol/mg) in the brush-
border enriched fraction, indicating 2.7 fold enrichment of cubilin ligand binding
over the homogenate and recovery of 8.4 % in the brush-border fraction. In
contrast, enrichment of alkaline phosphatase (ALP) specific activity, a tightly
associated apical brush-border membrane marker, was ~ 8.5 fold and recovery
was ~ 45 % in the brush-border fraction. This suggested that some portion of
cubilin may have been redistributed from the membrane to soluble fractions.

In order to examine this more closely, lF-cbl binding proteins in each of
the 40,000 x g soluble, total membrane, and brush-border-enriched fractions
were recovered by Triton-X-100 solubilization and affinity chromatography. They

were detected on western blots of nonreducing SDS-PAGE gels with polyclonal

81

rabbit anti-renal cubilin serum (Fig 3.4). Cubilin from the 40,000 x g supernatant
and total membrane fractions are shown in lanes 1 and 2, respectively, and
cubilin recovered from the brush-border membrane enriched fraction is shown in
lane 3. The IF-cbl binding proteins recovered in parallel from canine kidney
cortex homogenate are shown on the same gel in lane 4. The same single major
band was detected in lane 5 by western blot of canine kidney homogenate
without prior fractionation or affinity chromatography, indicating that the anti-
cubilin serum used as primary antibody was also immunospecific in whole tissue
homogenates.

Three major cubilin bands were found in each ileal fraction (Fig 3.4). Two
bands migrated just above and below the position of kidney cubilin with size
estimates of 550 and 450 kDa, respectively (labeled IC), and a diffuse band
migrated at ~185 kDa (labeled BB). There were also 2 faint bands of very high
apparent Mr in each fraction (labeled ?). In 2 independent experiments, the
proportion of total cubilin in each fraction represented by each size cubilin form
was determined by densitometry of exposed films. The diffuse BB band
comprised 30 % of recovered cubilin proteins in the soluble fraction, 55 % in the
total membrane fraction, and 60 % in the brush-border enriched fraction,
indicating an enrichment in the brush-border fraction of 1.5-2 fold relative to the
total homogenate. In addition, the smaller 10 form in the brush-border fraction
(lane 3) was depleted about 4 fold relative to that in the soluble fraction (lane 1)

and about 2 fold relative to the total membrane fraction (lane 2). The larger IC

82

kDa
ff} 4 ~460 p M

 

4 200 I

 

d 116

Figure 3.4: lmmunoreactive and lF-cbl binding cubilin species in
fractionated canine ileal mucosa and renal cortex. Cubilin species were
partially purified by IF-cbl affinity chromatography from various non-detergent
fractions of ileal mucosa (lanes 1-3) or renal cortex homogenate (lane 4) and
separated by non-reducing SDS-PAGE on 5 % gels. Ileal mucosa fractions
were 40,000 x g supernatant of total homogenate (lane 1), membranes pelleted
at 40,000 x g from total homogenate (lane 2), and brush-border membranes
enriched from homogenates by CaCl2 precipitation (lane 3). For comparison,
immunoreactive cubilin of renal cortex homogenate without lF-cbl affinity
purification is also shown (lane 5). Cubilin species were detected on western
blots with polyclonal anti-canine renal cubilin serum and an anti-rabbit IgG-
alkaline phosphatase conjugate, either by chemiluminescent exposure of
autoradiographic film (lanes 1-4) or by incubation in alkaline NBT/BCIP solution
(lane 5).

form represented about 20 % of cubilin in all 3 fractions. The 145 kDa band was
judged to be nonspecific crossreactivity because it was always present in the
chemiluminescent detection system as the only band detected and no decrease
was observed in signal on films of control blots developed with nonimmune rabbit
serum as the primary antibody, with the anti-rabbit IgG-ALP conjugate only, or
with [12511-protein A. Detection with the latter 2 control reagents suggested that
the 145 kDa band represented crossreacting, nonreduced canine lgG from
lymphocytes in ileal mucosa. The seeming enrichment of the very high apparent
MIr form (?) in lane 3 was not a consistent finding and often occurred in the other
fractions. Inclusion of disulfide bond reducing agents in the gel loading buffer in
preliminary experiments resulted in an increased apparent Mr, but 10-20 fold
reduction in signal, of renal cubilin and complete failure to detect ileal cubilin
bands (Fig 3.5).

The same set of 3 cubilin bands was also recovered from the final
supernatant after pelleting the brush-border fraction at 16,500 x g. In order to
examine whether cubilin had redistributed to a truly soluble fraction, ileal mucosa
was again homogenized without detergent, but in neutral isotonic buffer without
CaClz, and centrifuged at 40,000 x g for 1 hr, and the resulting supernatant was
centrifuged again at 150,000 x g for 2 hr. IF-cbl binding proteins were affinity
purified from each fraction and western blotted as above. All 3 of the cubilin
forms identified above were present in each of these fractions, but there was no
enrichment of the BB form or diminution of the smaller IC form apparent in any

fraction (Fig 3.6). lmmunoreactive cubilin was roughly equally distributed

84

’1"? Mi

 

e— 200 kDa

Figure 3.5: Western blot of canine cubilin in reducing and non-
reducing condition. Canine cubilin purified by affinity chromatography
from normal dog kidney (lanes 1 and 2) and normal dog ileum (lanes 3
and 4) was loaded with 100 mM DTT (lanes 1 and 3) and without DTT
(lanes 2 and 4) onto 5% polyacrylamide gel. The proteins were detected
by sequential incubation in polyclonal anti-canine renal cubilin serum,
anti-rabbit lgG-alkaline phosphatase conjugate, and alkaline NBT/BCIP
solution.

85

between the 150,000 x g supernatant and pellet fractions. Thus, none of the

identified ileal cubilin forms appeared to be tightly membrane bound.

Oligosaccharide analysis

To further characterize the apparent size and tissue-specific differences of
cubilin species, proteins purified from the 40,000 x g soluble fraction of ileal
mucosa and renal cortex homogenates by IF-cbl affinity chromatography were
subjected to differential glycosidase digestion prior to nonreducing
electrophoretic separation and western blotting (Fig 3.7). Renal (lanes 1-3) and
ileal cubilin (lanes 4-9) were digested with endo H (lanes 2, 5, and 8) or PNGase
F (lanes 3, 6, and 9). Lanes 1, 4, and 7 were mock digestions with no enzyme
added. The renal and ileal proteins in lanes 1-6 were from normal dog tissues,
and the ileal proteins in lanes 7-9 were from mucosa of a dog affected with
inherited selective cobalamin malabsorption due to failure of brush-border cubilin
expression (Fyfe et al., 1991). There was nonspecific protein degradation
apparent in both mock and real digestions of renal cubilin that was not apparent
in digestions of ileal cubilin. Endo H digestion of renal cubilin shifted its migration
only slightly, whereas PNGase F digestion reduced the apparent size to about
400 kDa. Endo H digestion of ileal proteins reduced the apparent size of the
~450 kDa band to ~400 kDa but had no affect on gel migration of the other major
ileal cubilin species. In contrast, PNGase F digestion of ileal cubilin reduced the
apparent size of both the ~550 and ~450 bands to ~400 kDa and the ~185 BB

form to 145 kDa. As previously described (Fyfe et al., 1991), the BB form was

86

IC form (

   

BB form-O :3

Figure 3.6: Distribution of immunoreactive cubilin forms in different
centrifugation fractions of dog ileal homogenate. Ileal mucosa was
homogenized without detergent or CaCI2 (lane H), and centrifuged at 40,000 x g
for 1 hr. The resulting supernatant (lane 8,) was centrifuged again at 150,000 x
g for 2 hr to produce a final supernatant (lane 82) and membrane pellet (lane p2)
lF-cbl binding proteins were affinity purified from each fraction and western
blotted as described for Fig 3.4.

87

 

IEndo H
IPNGaseF

 

 

 

 

 

 

 

 

 

 

 

 

kDa

~550
~450
~400

..,'.V.V

200 P.

 

116 D

Figure 3.7: Oligosaccharide analysis of affinity purified ileal and renal
cubilin. Renal cubilin of a normal dog (lane 1-3), ileal cubilin of a normal dog
(lanes 4-6), and ileal cubilin of a dog affected with an inherited defect blocking
brush-border expression of cubilin (lanes 7-9) were concentrated on IF-cbl-
agarose beads. Eluted proteins were dialyzed and mock digested (lanes 1, 4,
and 7), digested with endo H (lanes 2, 5, and 8), or digested with PNGase F
(lanes 3, 6, and 9) prior to separation by non-reducing SDS-PAGE on 5 % gels.
Proteins were detected by chemiluminescent western blot as described for Fig
3.4.

88

missing from affected dog ileal cubilin species, but the remaining species
exhibited similar shifts in size as normal dog ileal cubilin after both glycosidase

digestions.

Cubilin expression along crypt-villus axis in dog ileum

Cell fractions isolated from ileum during sequential intervals of incubation
in EDTA represent cell populations of various stages of differentiation along the
crypt to villus axis. The cell populations in early fractions contain mainly apical
villus cells at a late stage of differentiation. With longer incubation, the cell
populations encompass the villus base and crypt cells in early stages of
differentiation (Weiser et al., 1973). Northern blot analysis of these sequential cell
fractions, from two dogs, showed that the cubilin transcripts could be only
detected in the first 4-5 fractions with the highest level expression occurring in
the first or second fractions. In these two fractions, the majority of the cells were
assumed to be villus tip cells. The expression levels gradually decreased after
the fourth fraction. No cubilin transcript could be detected in the last 4 fractions.
GAPDH control mRNA could be detected in all cell fractions (Figure 3.8).

To confirm the reliability of the method of cell fractionation, the cell
fractions from the same two dogs were tested for villus tip and crypt specific
markers. Alkaline phosphatase (ALP) which is only expressed in the mature villus
cells was used as a tip specific marker (Traber et al., 1992), and the canine

homolog of mouse cryptidin related protein-ductin (CRP-ductin), which is only

89

Fraction 1 2 3 4 . 5 6 7 “8 9

Dog 1 cubilin '

CRP-ductin

 

GAPDH

   

GAPDH

Figure 3.8: Northern blot analysis of canine cubilin and CRP-ductin along
the crypt-villus axis in dog ileum. Thirty pg of total RNA isolated from dog
ileal epithelial cell fractions were used for the northern blots and hybridized
sequentially to dog cubilin cDNA, dog CRP-ductin cDNA amplified by RT-PCR,
and loading control probe GAPDH.

90

Figure 3.9: Alkaline phosphatase activity along the crypt-villus axis in dog
ileum. Total homogenate of dog ileal epithelial cell fractions were used for
alkaline phosphotase (ALP) activity analysis after protein quantitation. The
horizontal axis represents the cell fractions from villus to crypt. The vertical axis
represents the ALP activity in U/mg protein. The solid and dashed line represent
the duplicate experiment for the same dog.

91

unltlmg protein

Unltlrng protein

ALP activity In Ileal epItheIaII celh of 6091

0.1

 

0.09 __-_ _

 

 

 

 

 

 

 

 

 

 

 

 

0.01 ——- -— .-....";a;

fraction 1 traction 2 fraction 3 traction 4 fraction 5 fraction 6 fraction 7 fraction 8 fraction 9
cell fraction from vlllua to crypts

 

 

 

ALP actIvIty In Ileal epithelial cells of dog 2

 

 

 

 

 

 

O1

 

 

 

008,

0%

 

 

 

0.04

 

 

002 ,

 

 

 

 

 

Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Fraction 6 Fraction 7 Fraction 8 Fraction 9 Fraction 10
Fractions of Iaolated cells from villus to crypts

92

expressed in the immature crypt cells, was used as a crypt specific marker
(Cheng et al., 1996). For ALP activity in each dog (Figure 3.9), the peak ALP
activity was detected in the first fraction, and the activity decreased gradually in
the subsequent fractions. These results suggested that the cell populations in the
first three fractions represented mainly the villus tip cells, that cells in the middle
three fractions represented a mixture of villus and crypt cells, and that cells in the
last three fractions were mainly crypt cells. However, the same membrane
rehybridized with the crypt specific probe (CRP-ductin) showed that the canine
CRP-ductin transcript could be detected in almost all the fractions rather than just
the last 4-5 fractions. High level expression of CRP-ductin was detected in
fractions 2-3 for one dog and in fractions 3-5 for the other dog (Figure 3.8).
These data did not support our assumption that the fractions with cubilin

expression contained mainly villus tip cells.

93

Discussion

Functional expression of a plasma membrane receptor may be regulated
at one or more of multiple biosynthetic steps. The magnitude of changes in
intestinal cubilin mRNA expression from segment to segment demonstrated here
accounted fully for the observed changes in cubilin ligand-binding activity and
immunodetectable protein, indicating that the most important molecular
mechanism regulating regional distribution of cubilin expression in the canine
gastrointestinal tract is modulation of cubilin mRNA levels. The sharp decline of
both cubilin mRNA expression and ligand-binding activity in the most distal
segment of small intestine was most probably due to sampling from areas of
modified epithelium overlying a high density of Peyer's patches. The
codistribution of cubilin function, immunoreactive protein, and mRNA expression
correlated well with previous in vivo localizations of lF-cbl binding and absorption
in canine small intestine by oral radiolabeled cbl administration (Baker et al.,
1958; Drapanas et al., 1963; Fleming et al., 1962; Marcoullis et al., 1981) and is
similar to that demonstrated by mucosal ligand binding measurements in humans
(Hagedorn and Alpers, 1977). In the present study, however, we measured the
IF-cbl binding function of cubilin in homogenates of whole mucosa and did not
differentiate the activity of cell-surface receptor from that of intracellular cubilin.
We cannot comment, therefore, on whether endocytic functions mediated by
cubilin on the enterocyte surface may be regulated also by receptor translocation
to the plasma membrane, as has been described for glucose transporters (Kahn,

1992), or by other post-translational mechanisms.

94

In examining cubilin mRNA expression by northern blot with a canine renal
cubilin cDNA probe, two principal hybridizing transcripts were detected in distal
small intestinal mucosa and renal cortex. The 11-12 kb transcript corresponds in
size to the length of cloned renal CUBN cDNA (Xu et al., 1999) and hybridized ~
4 fold more strongly than did the larger transcript. Because the blotted samples
included nuclear RNA, the multiple hybridizing transcripts could represent mature
cubilin mRNA and incompletely processed forms of the primary CUBN transcript
still retaining one or more introns. As previously reported, Southern blots of
canine DNA digested with several restriction endonucleases indicated only a
single hybridizing CUBN locus in the canine genome (Xu et al., 1999), thus
eliminating the possibility that we observed hybridization to multiple closely
related transcripts of different genes. An alternative explanation which has not
been examined is that the CUBN transcript may be alternatively spliced.

The polyclonal antibody used in this study was raised against purified
renal cubilin of a single apparent Mr but specifically detected 3 major proteins on
western blots of ileal mucosal proteins. This indicated that the multiple
immunoreactive forms of cubilin observed in ileal mucosa were not simply
members of a protein complex obtained by co-immunoprecipitation or IF-cbl
binding of one member of the complex. In preliminary experiments, it was
determined that some form of receptor concentration, either immunoprecipitation
or affinity chromatography, was necessary for reliable immunodetection of ileal
cubilin on western blots. This may explain why cubilin was not detected on

western blots of ileal mucosa from a pair of Imerslund-Grasbeck patients

95

purported to overexpress cubilin ligand binding activity (Eaton et al., 1998). The
very high apparent Mr bands observed on western blots of ileal and renal cortex
(labeled ? in Fig 2B and 4) were most likely aggregates of cubilin that survived
boiling in SDS. We have observed these also on silver stained gels of affinity
purified ileal and renal cubilin so they are not a western blotting artifact. The
aggregate forms in ileal mucosa were higher apparent Mr than in renal cortex
(Fig 3.4), and they shifted migration on SDS-PAGE after glycosidase digestions
(Fig 3.7). Both observations were consistent with the more extensive N-Iinked
glycosylation on the 550 kDa form of ileal cubilin than that of renal cubilin evident
from the relative shifts in gel migration after PNGase F digestion (Fig 3.7). Heavy
glycosylation of ileal cubilin may be an adaptation conferring some protection
from Iumenal proteases, and may explain the greater stability of ileal vs renal
cubilin apparent during glycosidase digestions in vitro.

Physiochemical evidence of cubilin aggregation under various in vitro
conditions has been reported previously (Birn et al., 1997; Kristiansen et al.,
1999; Lindblom et al., 1999; Seetharam et al., 1981a; Seetharam et al., 1982). In
a recent report, trimers of bovine renal cubilin of ~1,500 kDa were resistant to
disruption in 6 M guanidine but were separated into 440 kDa monomer after
disulfide bond reduction (Lindblom et al., 1999). The N-terminal region
conjectured to mediate trimer formation has also been identified as a region
required for membrane association of cubilin fragments expressed in transfected
cells (Kristiansen et al., 1999). In the latter study, aggregate forms were

observed on nonreducing gels only in the constructs expressing the membrane

96

binding N-terminal sequence. This region of cubilin also has the only unpaired
cysteine in the entire protein that would be available for dimer formation. The
presence of aggregate forms is, however, an inconsistent observation, so it
remains unclear whether cubilin multimer formation occurs in vivo or is merely an
in vitro artifact.

We concluded that the ~185 kDa band, labeled BB in Fig 32B and 3.4, is
a brush-border form of cubilin based on 1) its absence from ileal mucosa of dogs
exhibiting an inherited defect that blocks brush-border cubilin expression and 2)
its enrichment in fractions enriched for brush-border membranes. It is likely that
the ~185 kDa form arises by proteolytic cleavage of the ~550 kDa form (see
below) when exposed to proteases in the intestinal lumen. This idea is consistent
with the previous finding by gel exclusion chromatography that in vitro trypsin
treatment converted high molecular weight canine ileal cubilin to lF-cbl binding
forms of 150-180 kDa (Seetharam et al., 1982). Enrichment and recovery of
cubilin ligand binding activity and immunoreactivity of the ~185 kDa form was
considerably less than that of the tightly-associated brush-border membrane
marker, ALP, but it was clear that a portion of brush-border cubilin redistributed
to other fractions during homogenization. This study confirmed that endogenous
cubilin is only loosely membrane associated in canine ileum, similar to what was
previously noted with endogenous cubilin in rabbit kidney (Moestrup et al., 1998)
and canine ileal cubilin reconstituted in artificial liposomes (Seetharam et al.,
1981 b). These are in vitro behaviors consistent with the designation of cubilin as

a peripheral membrane protein.

97

What we have designated IC forms in Fig 3.2B and 3.4 appear to be
intracellular cubilin precursors at different stages of maturation. The N-linked
oligosaccharides of the smaller IC form, estimated to be 450 kDa, were cleaved
by endo H, indicating that they are high-mannose and derived from cubilin which
has not been transported from the RER to the cis-Golgi. Furthermore, the 450
kDa band was found largely in the 40,000 x g soluble fraction after
homogenization in TMB without detergents and was depleted from the brush-
border enriched fraction, consistent with an association with RER derived
microsomes. Oligosaccharides of the larger IC form, estimated to be 550 kDa,
were endo H resistant but PNGase F susceptible, indicating they are complex or
hybrid types that had been exposed to Golgi processing enzymes. The same
was true of oligosaccharides on the 185 kDa form. We have not determined
which of these forms is directly responsible for lF-cbl endocytosis in vivo. The
185 kDa and 550 kDa forms may coexist for some time on the brush-border, and
it is possible that the endocytic function of cubilin is inactivated by a proteolytic
cleavage to 185 kDa.

Detection of the ~450 kDa form was evidence that ileal cubilin gained IF-
cbl binding activity and immunoreactivity at a step in biosynthesis prior to
modification of high mannose oligosaccharides to endo H resistant forms and
prior to expression in a position on the brush-border appropriate to its endocytic
function. Previous data indicated that the same was true of renal cubilin (Fyfe et
al., 1991). At the steady states of tissues examined in this report, renal cubilin

was mostly endo H resistant, but a large proportion of ileal cubilin was endo H

98

sensitive. However, this does not reflect a true tissue-specific difference of cubilin
partitioning between the RER and post-RER compartments because renal cells
that express cubilin are largely a population of fully differentiated cells whereas
ileal enterocytes that express cubilin in a homogenate of mucosa are a mixture of
proliferating and immature crypt cells, partially differentiated mid-villus cells, and
fully differentiated villus tip cells. Thus, in contrast to kidney, homogenates of ileal
mucosa are relatively enriched in cubilin forms produced early in the biosynthetic
pathway and the mature forms are underrepresented.

Removal of N-linked oligosaccharides from both IC forms of ileal cubilin
resulted in migration at ~400 kDa, consistent with the polypeptide molecular
weight deduced from the canine renal cubilin cDNA (Xu et al., 1999). Similar to
most glycoproteins destined for the plasma membrane, it appears that canine
ileal cubilin is translated as a 400 kDa polypeptide, that it is N-Iinked glycosylated
in the RER to ~450 kDa, and thereafter, that oligosaccharide modifications
produce a mature ~550 kDa glycoprotein. However, as shown in Fig 3.7, dogs
exhibiting an inherited defect blocking brush-border expression of cubilin still
produce both IC forms of cubilin in the intestine, including the endo H resistant,
~550 kDa form. Genetic linkage evidence indicates that the disease locus in
these dogs is not the cubilin locus (Xu et al., 1999), and therefore, some type of
cubilin-specific transport to the Iumenal brush-border surface must exist. The
present data suggest that it operates after cubilin matures to the endo H
resistant, ~550 kDa form in ileal mucosa, but a study in rat yolk sac cells

suggests that newly synthesized cubilin arrives on the brush-border with endo H

99

sensitive oligosaccharides (Baricault et al., 1995). It is possible that these
phenomena are tissue or species specific.

Tissue-specific differences of N-linked oligosaccharides may also have
contributed to confusion in the literature about ileal cubilin protein size and
lingering thought that intestinal and renal cubilin ligand-binding activity might be
due to expression of different but closely related genes (Guéant et al., 1999). The
present data indicated that ileal and renal cubilin in dogs have the same protein
backbone and that differences of apparent Mr on SDS-PAGE are due to differing
N-linked glycosylation. Moreover, both a cDNA probe and polyclonal antibody
derived from renal cubilin detected gastrointestinal cubilin in the same pattern as
determined by lF-cbl binding activity and as previously reported for canine
gastrointestinal binding and absorption of oral cobalamin (Baker et al., 1958;
Drapanas et al., 1963; Fleming et al., 1962; Marcoullis and Rothenberg, 1981).
Conversely, antibody raised against purified canine ileal cubilin demonstrated
cross immunoreactivity with canine renal cubilin in a previous study (Fyfe et al.,
1991). However, recent reports of a null mutation in the cubilin gene of a human
individual exhibiting both intestinal IF-cbl malabsorption and urinary loss of
cubilin ligands provide the most elegant and conclusive demonstration that ileal
and renal cubilin functions are expressions of the same gene (Aminoff et al.,
1999; Kozyraki et al., 1999). The findings reported here will aid in determining the
basis of the analogous disorder in dogs.

Previous immunoelectron microscopic study (Levine et al., 1984) of

intracellular cubilin distribution during enterocyte maturation from crypt to villus

100

tip of ileal mucosa showed that synthesis of cubilin protein could be easily
detected in the mid-villus cells. The expression level of cubilin dramatically
increased in the villus tips, and expression of cubilin was extremely low in the
crypt cells. Northern blot of canine cubilin mRNA expression in this study showed
that cubilin transcripts could only be detected in the early cell fractions assumed
to represent villus cell populations. These results were supported by the parallel
ALP activity which is known to be expressed only in villus cells. However, the
northern blot of crypt specific canine CRP-ductin showed a conflicting result.
CRP-ductin was isolated from cDNA libraries of mouse intestinal crypt cells
(Cheng et al., 1996). In situ hybridization in the same study showed that in the
small intestine, CRP-ductin mRNA is expressed in crypt cells at all stages of
differentiation from the stem cells to the terminally differentiating cells of the crypt
top, but not in the mature cells of the villus. A recent study (Takito et al., 1999)
suggested that mouse CRP-ductin, human DMBT1, rabbit hensin and ebnerin
are the products of alternative splicing of a single gene. The transcript of this
gene for all four family members could be detected in the small intestine, but the
crypt specificity has only been determined in mouse.

The conflicting results for the ALP activity and the canine CRP-ductin
expression may have been due to incomplete cell fractionation. The incubation
condition may have been over-extensive which caused both villus and crypt cells
to fall-off in the early fractions. The crypt specificity of canine CRP-ductin also
needs to be confirmed. To eliminate all these concerns, other crypt specific

markers could be used to determine the origin of cell population in each fraction,

101

such as thymidine kinase. Alternatively, in situ hybridization of canine cubilin
directly on the dog ileum sections could be used to determine the cubilin mRNA

expression along the crypt to villus axis.

102

References

Aminoff M, Carter JE, Chadwick RB, Johnson C, Grasbeck R, Abdelaal MA,
Broch H, Jenner LB, Verroust PJ, Moestrup SK, Chapelle A de la, and Krahe R.
(1999) Mutations in CUBN, encoding the intrinsic factor-vitamin B12 receptor,
cubilin, cause hereditary megaloblastic anaemia 1. Nat Genet 21: 309-313.

Baker SJ, MacKinnon NL, and Vasudevia P. (1958) The site of absorption of
orally administered vitamin B12 in dogs. Ind. J. Med. Res. 46: 812-817.

Baricault L, Galceran M, Ronco PM, Trugnam G, and Verroust PJ. (1995)
Unusual processing of gp280, a protein associated with the intennicrovillar areas
of yolk sac epithelial cells: plasma membrane delivery of immature protein.

Biochem. Biophys. Res. Comm. 212: 353-359.

Batt M, and Horadagoda NU. (1989) Gastric and pancreatic intrinsic factor-
mediated absorption of cobalamin in the dog. Am. J. Physiol. 257 (Gastrointest.
Liver Physiol. 20) G344-G349.

Batuman V, Verroust PJ, Navar GL, Kaysen JH, Goda FO, Campbell WC, Simon
E, Pontillon F, Lyles M, Bruno J, and Hammond TG. (1998) Myeloma light chains
are ligands for cubilin (gp280). Am. J. Physiol. 275 (Renal Physiol. 44): F246-
F254.

Birn H, Fyfe J, Jacobsen C, Mounier F, Verroust PJ, erkov H, Willnow T,
Moestrup SK, and Christensen El. (In press) Cubilin is an albumin binding protein
important for renal albumin reabsorption. J. Clin. Invest.

Bim H, Verroust PJ, Nexo E, Hager H, Jacobsen C, Christensen El, and
Moestrup SK. ( 1997) Characterization of an epithelial ~460-kDa protein that
facilitates endocytosis of intrinsic factor-vitamin B12 and binds receptor-
associated protein. J. Biol. Chem. 42: 26497-26504.

Cheng H, Bjerknes M and Chen H. (1996) CRP-ductin: a gene expressed in
intestinal crypts and in pancreatic and hepatic ducts. Anat Rec 244: 327-343.

Drapanas T, Williams JS, McDonald JC, Heyden W, Bow T, and Spencer RP.
(1963) Role of the ileum in the absorption of vitamin B12 and intrinsic factor (NF).
Clin. Sci. 184: 337-341.

Eaton DM, Livingston JH, Seetharam B, and Puntis JWL. (1998) Overexpression

of an unstable intrinsic factor-cobalamin receptor in Imerslund-Grasbeck
syndrome. Gastroenterology 115: 173-176.

103

Fleming WH, King ER, Galloway RA, and Roche JJ. (1962) The site of
absorption of orally administered Cow-labeled vitamin B12 in dogs: the effect of
dose. Gastroenterology 42: 164-168.

Fyfe JC. (1993) Feline intrinsic factor (IF) is pancreatic in origin and mediates
ileal cobalamin (cbl) absorption. J. Vet. Intern. Med. 7: 133 (abstract).

Fyfe JC, Kurzhals RL, Lassaline ME, Henthorn PS, Alur PRK, Wang P, Wolfe JH,
Giger U, Haskins ME, Patterson DF, Sun H, Jain S, and Yuhki N. (1999)
Molecular basis of feline B-glucuronidase deficiency: an animal model of
mucopolysaccharidosis VII. Genomics 58: 121-128.

Fyfe JC, Ramanujam KS, Ramaswamy K, Patterson DF, and Seetharam B.
(1991) Defective brush-border expression of intrinsic factor-cobalamin receptor in
canine inherited intestinal cobalamin malabsorption. J. Biol. Chem. 266: 4489-
4494.

Guéant J-L, Chery C, Namour F, Aimone—Gastin I, and Wustinger M (Reply,
Seetharam B). (1999) Decreased affinity of urinary intrinsic factor-cobalamin
receptor in a case of Grasbeck-Imerslund syndrome. Gastroenterology116:
1274-1276.

Hagedorn CH and Alpers DH. (1977) Distribution of intrinsic factor-vitamin B12
receptors in human intestine. Gastroenterology 73: 1019-1022.

Hammad SM, Stefansson S, Twal WO, Drake CJ, Fleming P, Remaley A, Brewer
HB, and Argraves WS. (1999) Cubilin, the endocytic receptor for intrinsic factor-
vitamin B12 complex, mediates high-density lipoprotein holoparticle endocytosis.
Proc. Natl. Acad. Sci. USA 96: 10158-10163.

Hammond TG, Verroust PJ, Majewski RR, Muse KE, and Oberley TD. (1994)
Heavy endosomes isolated from the rat renal cortex show attributes of
intennicrovillar clefts. Am. J. Physiol. 267 (Renal Fluid Electrolyte Physiol. 36:
F516-F527.

Kahn BB. (1992) Facilitative glucose transporters: regulatory mechanisms and
dysregulation in diabetes. J. Clin. Invest. 89: 1367-1374.

Kapadia CR. (1995) Vitamin B12 in health and disease: part l--inherited disorders
of function, absorption and transport. Gastroenterologist 3: 329-344.

Kessler M, Acuto O, Storelli C, Murer H, Miiller M, and Semenza G. (1978) A

modified procedure for the preparation of efficiently transporting vesicles from
small intestinal brush border membranes. Biochim. Biophys. Acta 506: 136-154.

104

Kozyraki R, Fyfe J, Kristiansen M, Gerdes C, Jacobsen C, Cui S, Christensen El,
Aminoff M, Chapelle A de la, Krahe R, Verroust PJ, and Moestrup SK. (1999)
The intrinsic factor-vitamin B12 receptor, cubilin, is a novel high affinity
apolipoprotein A-I receptor facilitating endocytosis of high-density Iipoprotein.
Nat. Med. 6: 656-661.

Kozyraki R, Kristiansen M, Silahtaroglu A, Hansen C, Jacobsen C, Tommerup N,
Verroust PJ, and Moestrup SK. (1998) The human intrinsic factor-vitamin B12
receptor, cubilin: Molecular characterization and chromosomal mapping of the
gene to 10p within the autosomal recessive megaloblastic anemia (MGA 1)
region. Blood 91: 3593-3600.

Kristiansen M, Kozyraki R, Jacobsen C, Nexo E, Verroust PJ, and Moestrup SK.
(1999) Molecular dissection of the intrinsic factor-vitamin B12 receptor, cubilin,
discloses regions important for membrane association and ligand binding. J. Biol.
Chem. 274: 20540-20544.

Le Panse S, Ayani E, Nielsen S, Ronco P, Verroust P, and Christensen El.
(1997) Internalization and recycling of glycoprotein 280 in epithelial cells of yolk
sac. Europ. J. of Cell Biol. 72: 257-267.

Levine JS, Allen RH, Alpers DH, and Seetharam B. (1984) lmmunocytochemical
localization of the intrinsic factor-cobalamin receptor in dog-ileum: distribution of
intracellular receptor during cell maturation. J. Cell Biol. 98: 111 1-1118.

Lindblom A, Quadt N, Marsh T, Aeschlimann D, Mergelin M, Mann K, Maurer P,
and Paulsson M. (1999) The intrinsic factor-vitamin B12 receptor, cubilin, is
assembled into trimers via a coiled-coil a-helix. J. Biol. Chem. 274: 6374-6380.

Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ. (1951) Protein
measurement with the folin phenol reagent. J. Biol. Chem. 193: 265-275.

Marcoullis G and Rothenberg SP. (1981) Intrinsic factor-mediated intestinal
absorption of cobalamin in the dog. Am. J. Physiol. 241 (Gastrointest. Liver
Physiol. 4) GZ94-G299.

Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D,
Pontillon F, Goda FO, Christensen El, Hammond TG, and Verroust PJ. (1998)
The intrinsic factor-vitamin B12 receptor and target of teratogenic antibodies is a
megalin-binding peripheral membrane protein with homology to developmental
proteins. J. Biol. Chem. 273: 5235-5242.

Sambrook J, Fritch EF, and Maniatis J. (1989) Molecular Cloning: A Laboratory
Manual, 2nd ed. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press.

105

Seetharam B, Alpers DH, and Allen RH. Isolation and characterization of the ileal
receptor for intrinsic factor-cobalamin. J. Biol. Chem. 256: 3785-3790.

Seetharam B, Bagur SS, and Alpers DH. (1981) Interaction of receptor for
intrinsic factor-cobalamin complex with synthetic and brush-border lipids. J. Biol.
Chem. 256: 9813-9815.

Seetharam B, Bagur SS, and Alpers DH. (1982) Isolation and characterization of
proteolytically derived ileal receptor for intrinsic factor-cobalamin. J. Biol. Chem.
257: 183-189.

Seetharam B, Bakke JE, and Alpers DH. (1983) Binding of intrinsic factor to ileal
brush border membrane in the rat. Biochem. Biophys. Res. Comm. 115: 238-
244.

Seetharam B, Christensen El, Moestrup SK, Hammond TG, and Verroust PJ.
(1997) Identification of rat yolk sac target protein of teratogenic antibodies,
gp280, as intrinsic factor-cobalamin receptor. J. Clin. Invest. 99: 2317-2322.

Simpson KW, Alpers DH, Wille J De, Swanson P, Farmer S, and Sherding RG.
(1993) Cellular localization and hormonal regulation of pancreati S c intrinsic
factor secretion in dogs. Am. J. Physiol. 265 :(Gastrointest. Liver Physiol. 28)
G178-G188.

Takito J, Yan L, Ma J, Hikita C, Vijayakumar S, Warburton D, and Al-Awqati Q.
(1999) Hensin, the polarity reversal protein, is encoded by DMBT1, a gene
frequently deleted in malignant gliomas. Am J Physiol 277 (Renal Physiol 46):
F277-F289.

Traber PG, Yu L, Wu GD, and Judge TA. (1992) Sucrase-isomaltase gene
expression along crypt-villus axis of human small intestine is regulated at level of
mRNA abundance. Am J Physiol 262 (Gastrointest Liver Physiol 25): G123-
G130.

Weiser MM. (1973) Intestinal epithelial cell surface membrane glycoprotein
synthesis. 1. An indicator of cellular differentiation. J.Biol. Chem. 248: 2536-2541.

Xu D, Kozyraki R, Newman TC, and Fyfe JC. (1999) Genetic evidence of an

accessory activity required specifically for cubilin brush-border expression and
intrinsic factor-cobalamin absorption. Blood 94: 3604-3606.

106

CHAPTER 4

LINKAGE ANALYSIS OF CANINE SELECTIVE COBALAMIN

MALABSORPTION WITH CUBN AND MEGALIN LOCI

107

Introduction

Hereditary megaloblastic anemia 1, (MGA 1, OMIM #261100), also known
as Imerslund-Grasbeck syndrome (lG-S), is a rare autosomal recessive disorder
characterized by selective intestinal cobalamin (vitamin B12) malabsorption
leading to juvenile-onset severe megaloblastic anemia and/or neurologic
manifestations. Most patients also exhibit mild proteinuria. The disorder has been
described in families in Nonivay (Imerslund 1960), Finland (Grasbeck et al.,
1960), Israel (Ben-Bassat et al., 1969), Saudi Arabia (Abdelaal and Ahmed,
1991), Turkey (Altay et al., 1995; Altay and Cetin, 1999) and sporadically
worldwide. MGA 1 was mapped to chromosome 10p12.1 in Finnish, Norwegian
and a Saudi Arabian kindred (Aminoff et al., 1995, 1999).

One candidate gene for lG-S was cubilin, a 460 kDa protein found in the
apical membrane domain of epithelial cells of the distal small intestine, and renal
proximal tubules (Seetharam et al., 1997; Birn et al., 1997). Human (GenBank
accession # AF034611; Kozyraki et al., 1998) and rat (GenBank accession
#AF022247; Moestrup et al., 1998) cubilin were recently cloned. The human
CUBN locus was mapped within the MGA 1 locus by fluorescence in situ
hybridization, radiation hybrid mapping, and screening of YAC clones (Moestrup
et al., 1998). Two mutations, a missense mutation in CUB domain 8 and an
intronic point mutation activating a cryptic splice site within CUB domain 6, were
recently demonstrated in Finnish kindreds (Aminoff et al., 1999).

A canine model with selective intestinal Cbl malabsorption was identified

with clinical and laboratory features similar to those seen in human (Fyfe et al.,

108

1991a). Immunoelectron microscopy and cell fractionation studies showed failure
of cubilin expression on the apical brush-border membrane of either tissue (Fyfe
et al., 1991b). Characterization of the urine proteins of affected dogs showed
ApoA-1 (Kozyraki et al., 1999), albumin (Birn et al., in press), haptoglobin, lgG
light chains and other unidentified proteins (Fyfe et al., unpublished data), which
are the same as the urine proteins identified in the patient with a cryptic splice
site (null) mutation of cubilin. To further determine whether defective function of
canine cubilin is caused by a mutation within its coding sequence, which is the
most common reason for a protein trafficiking disorders, we perfomed the linkage
analysis to determine whether the cubilin gene locus and the disease locus are
genetically linked.

The predicted protein structure of human and rat cubilin showed no
sequence compatible with a transmembrane domain nor a signal for
glycosylphospho-tidylinositol anchor addition. Parallel with the predicted
structure, the protein showed only a weak association with the plasma
membrane (Moestrup et al., 1998). All these data suggested that cubilin is a
peripheral membrane protein. As a peripheral membrane protein without
internalization signals, cubilin-mediated endocytosis of the lF-Cbl complex may
be facilitated by another membrane protein. Megalin, a transmembrane protein
with an endocytic signal sequences in the cytoplasmic domain, interacts with
cubilin in vitro (Moeatrup et al., 1998), colocalizes with cubilin in vivo (Sahali et
al., 1988, 1993, Birn et al., 1997, Moestrup et al., 1998) and exhibits coregulation

of expression (hammad et al., 2000). Megalin, therefore, has been suggested to

109

play an important role in facilitating the endocytosis of the IF-Cbl complex.
Therefore, a defect in the interaction between cubilin and megalin may be
another reason for failure of cubilin expression on apical brush border
membranes. To determine whether the defective interaction between megalin
and cubilin caused the Cbl malabsorption in this canine model, we also
conducted linkage analysis between canine megalin and selective intestinal

cobalamin malabsorption in this dog family.

110

Materials and Methods

Cubilin linkage study

Dogs in the linkage studies were members of a breeding colony
maintained at Michigan State University; protocols for husbandry and tissue
collection were approved by the MSU All-University Committee for Animal Use
and Care. The parents and 23 offspring, including 13 affected and 10 clinically-
normal carriers, of matings between an affected female and an obligate carrier
male were studied. The disease phenotype of each dog was determined by the
following criteria: affected dogs became inappetent and failed to gain weight after
10-14 weeks of age; they had low serum cobalamin concentrations,
methylmalonic aciduria, mild nonregenerative anemia, and neutropenia. And all
hemotologic, metabolic, and growth abnormalities were reversed by parenteral
cobalamin administration. Due to the carrier X affected matings strategy,
offspring determined as not affected were considered obligate carriers of
selective cobalamin malabsorption.

Genomic DNA of normal dogs was isolated from frozen liver samples.
Approximately 0.5 g of frozen tissue from each dog was ground into fine power in
liquid nitrogen and gently spread onto extraction buffer (10 mM Tris-HCI, pH 8.0,

0.1 M EDTA, 20 pglml pancreatic RNAase, 0.5% SDS), allowing the powder to
become submerged. Tissue suspensions were incubated with 100 pglml
proteinase K at 55°C overnight and were extracted in 0.5 M Tris-HCI (pH 8.0)

equilibrated phenol, phenol and chloroform mixture, and chloroform sequentially

and precipitated by 10 M ammonium acetate (0.2 vol/vol) and 100% ethanol (2

111

vol/vol). Isolated DNA was washed in 70% ethanol and resuspended in TE pH
8.0, and quantified by spectrophotometry.

Genomic DNA was amplified by PCR using various combinations of the
primers 838F, 5’-AGCCTGCGTGCTGGACATCGAC-S’; 947F, 5’-
GGATGGCAAGGAAATGGATATAGT—3’; 1150R, 5’-TGGGTGGCAGCCTC-
CATTAT'I'GA-3’; and 1341 R 5’-CCAGCCCAACCTGATTCACACTTA-B’ designed
from the canine cubilin cDNA sequence. 50 pL reaction mixtures contained 25
mM TAPS (pH 9.3), 50 mM KCI, 2 mM MgClz, 1 mM 2-mercaptoethanol, 0.4 mM
each deoxynucleotide, 0.5 pM each primer, 500 ng of DNA template, and 2.5 U
of TaKaRa LA Taq polymerase (PanVera, Corp., Madison, WI). After a hotstart
denaturation at 94°C for 2 min, reactions were carried out for 35 cycles of
denaturation at 98°C for 20 s and annealing / extension at 68°C for 20 min.
Reaction products were gel purified and partially sequenced using the above
PCR primers.

According to the sequencing results, a 0.9 kb intron of both an obligate
carrier male and an affected female were amplified using primer 947F and
1150R. The PCR products were gel purified and cloned into the plasmid PCRTM ll
(Invitrogen®, San Diego, CA) at 16°C overnight. Recombinant plasmids were
transformed into DH5 a competent cells. Six individual clones from each dog
were purified by QIAGEN plasmid purification kit and sequenced to identify
intronic sequence variations.

DNA templates used for PCR amplification of a cubilin intron variation

(CIV) were isolated from snap frozen liver or from fresh blood samples. The DNA

112

from frozen tissues was isolated by the same protocol as previously described.
The DNA from whole blood cells was isolated by the following procedure. The
whole blood cells were Iysed by ST buffer (0.32 M sucrose 10 mM Tris, pH7.6, 5
mM MgCl2 and 1% Triton X-100) and the leukocyte nuclei were resuspended in
NE buffer (75mM NaCl and 25 mM EDTA pH 8.0) and 0.1% SDS. After
sequential incubation with proteinase K and RNAse A, the genomic DNA was
extracted and precipitated as from frozen tissue described previously. To PCR
amplify the intron variation, the primers (CIVF, 5’-GATCACAGGCCTACAGCTC-
CATT-3’ and CIVR, 5’-CCAGGCCAACCAGAGATCTTCTA-3’) were designed to
flank a 17 base insertion sequence. PCR produced allele-specific amplification
products of 199 and 182 bp. PCR reactions were 50 pL, containing 50 mM KCI,
10 mM Tris-Cl (pH 8.3), 2 mM MgClz, 0.37 mM each deoxynucleotide, 0.25 pM
each primer, and 100 ng of DNA template. Taq DNA polymerase (2.5 U) was
added after a 5 min hotstart at 95°C, and PCR reactions were carried out for 28
cycles of denaturation for 30 s at 95°C, annealing for 60 s at 65°C, and extension
for 120 s at 72°C. The reaction products were electrophoresed in TAE buffer on
4% agarose gels and visualized by ethidium bromide staining.

Southern blotting was performed on genomic DNA of carrier and affected
dogs isolated from frozen liver as described above. Genomic DNA was digested
by Barn HI, EcoRl, Hind Ill, Pst l, and Xba I and 5 pg of each digested DNA was
electrophoresed on a 1% agarose gel. DNA was transferred to a nylon
membrane, and the blot was hybridized by standard methods (Sambrook et al.,

1989) to a canine cubilin cDNA probe (bp 947-1150).

113

Megalin linkage study:

The primers used for canine megalin DNA amplification were designed
from a conserved region of human and rat megalin and human low-density
lipoprotein related protein (LRP1). The primer: Meg fonivard, 5’-GGCTCCT-
ACATCTGTAAGTGTGC-3’; Meg reverse, 5’-CACAGACTTG‘ITGGTI’CCATCC-
3’ were designed from human LRP1 exon 56 and exon 60 respectively. Reverse
transcribed polymerase chain reaction (RT-PCR) was performed using above
primers to amplify canine megalin cDNA from total RNA of dog kidney which was
isolated from snap frozen dog kidney tissue using Trizol® reagent (GibCo
protocol). The PCR product was gel purified and sequenced using the same pair
of primers.

The same pair of primers was used to amplify the canine genomic megalin

fragment. The 50 pl reaction mixture contained 300 mM Tris-HCI (pH 8.5), 75
mM ammonium sulfate, 2.5 mM MgClz, 0.25 mM each deoxynucleotide, 0.5 pM
each primer, and 300 ng of dog genomic DNA template. 2.5 U of AmpliTaq
polymerase (Perkin Elmer, Foster city, CA) was added after a hotstart
denaturation at 95°C for 5 min, then reactions were carried out for 32 cycles with
30s denaturation at 95°C, 305 annealing at 58°C and 2 min extension at 72°C.
The single PCR product from an obligate carrier dog was gel purified and
sequenced. In order to separate two PCR products (both are around 2.1 kb) from
an affected dog, a new fonivard primer, 1501F, 5’-

CATTGCCTCTCTGTCTCTGACCT-3’, was designed from the middle exon and

114

used for PCR amplification under the same reaction conditions, but without the
hotstart, to produce smaller PCR products (about 700 bp). The two smaller PCR
products of the affected dog were gel purified and cloned individually in the
plasmid pCRTM ll (Invitrogen, San Diego, CA). Two purified plasmids containing
each PCR product were sequenced separately.

The primers, 1553F, 5’-GAGCACTGTGTGGATGTCAA-3’; 2009R, 5’-
CATGGAGAGTCCAGTATAGG-3’ were designed to flank an intronic insertion of
68 bp defined from canine megalin sequence. PCR amplification of this canine
megalin intron variation was performed on DNA isolated from frozen liver or fresh
blood samples using the same methods as above. Totally 24 dog samples with
confirmed phenotype were used for this study, including one normal dog, 9
carrier dogs and 14 affected dogs for selective cobalamin malabsorption. The
conditions for the 50 pl PCR reaction were as same as for the canine megalin
genomic DNA amplification except using 1 min extension time instead of 2 min.
Allele specific PCR products were separated on a 4% agarose gel in 1X TAE

buffer and visualized by ethidium bromide staining.

115

Results

The intron/exon structure of the canine CUBN was defined in the region of
EGF domains 4-7 by amplification of genomic DNA by PCR, using primer
sequences from the cDNA (Figure 4.1, panel A). Sequencing of the PCR
products demonstrated concensus splice donor and acceptor sequences at the
points where the genomic sequence diverged from the cDNA sequence. In the 5’
to 3’ direction, 4 introns (2.2, 0.9, 2.0 and 0.2 kb) and 3 included exons (cDNA
sequences 948-1079, 1080-1175, and 1176-1294; Gly292-Pro335, Gly336-
Leu367, and Gly368-Ile407) were defined. The intron/exon boundaries did not
correspond to the protein structural boundaries of the EGF domains.

A canine CUBN variation was determined within the 0.9 kb intron between
nucleotides 1079 and 1080 of the cDNA sequence in EGF domain 5 (Panel A).
The intron between PCR primers 947F and 1150 R of a heterozygous male dog
(F274 Figure 4.1) was amplified, and the product was cloned. Sequencing of six
individual clones demonstrated a 17 bp sequence (5’-CAGAACA'I‘I'G'I'T'I'ATGC-
3’) in three of the clones inserted 187 bp 5’ of the 3’ intron/exon boundary.
Sequencing of 6 clones of the intron similarly amplified from a homozygous
female (F284, Figure 4.1) demonstrated that she was likely (P>0.98)
homozygous for the allele containing the 17 bp insert. Single hybridizing bands of
5-10 kb, identical in carrier and affected dog DNA, were observed on Southern
blots of canine genomic DNA digested with 6 different restriction enzymes and

hybridized to a probe of CUBN cDNA sequence between bp 947 and 1150, thus

116

 

838F 947.:
- ’
2.2 kb 0.9 2.0 kb
I ’ ‘ ‘
a,” 1150a ‘~‘ 1341a
’ ’ ~ ‘
e ’ ’ crvr= ~ 2
fi
61" AG I
J "
B 17 bp variation C'VR
F274 F284
* * ‘k * 1k

Figure 4.1: Independent segregation of canine I-GS and CUBN loci. (A) Four
introns (horizontal lines) and included exons (vertical boxes) were defined

by PCR amplification using CUBN cDNA primers and sequencing the products.
A 17—bp variation (vertical line) was found in the 0.9-kb intron for which dog
F274, an obligate carrier of canine l-GS, was heterozygous and dog F284, an
affected dog, was homozygous. (B) Solid symbols indicate l-GS affected dogs,
half-solid symbols indicate obligate carriers, squares are males, and circles are
females. DNA from offspring of matings between dogs F274 and F284 was
amplified by PCR using primers flanking the 17-bp variation, producing allele-
specific products of 199 and 182 bp. Results of 10 offspring are shown, with the
stars indicating recombinants.

117

BamHl EcoRl Hind Ill Pstl Xbal
.C A . C} A C.

’7 Wind Ill

 
   

2

Figure 4.2: lntragenic cubilin marker is located in a unique region of the
canine genome.Southern blots of canine genomic DNA digested with 6 bp
recognition-site restriction endonucleases were prepared and hybridized to a
canine cubilin cDNA probe flanking the intron containing an identified variation
(see figure 4.1). Single hybridizing bands of 5—10 kb, identical in carrier and
affected dog DNA were observed, thus confirming that the identified 17-bp

marker variation was in a unique region of the canine genome. C: carrier dog,
A: affected dog.

118

confirming that the identified 17-bp variation was in a unique region of the canine
genome (Figure 4.2).

A CUBN allele-specific PCR test was designed to examine the 17 bp
insertion variation in affected and carrier offspring of affected X carrier matings
(phase-known double backcross matings) performed between dogs F284 and
F274 (Figure 4.1, Panel B). The affected offspring from these matings are
homozygous and clinically normal Iittermates which are heterozygous at the
disease locus. The PCR test confirmed that the clinically affected dam was
homozygous and the clinically normal, obligate carrier sire was heterozygous for
the 17 bp insert. However, 13 recombinants were detected among 23 offspring of
these matings, a recombination fraction (0.56) that did not differ from the
recombination fraction (0.5) expected under the hypothesis of independent
segregation of the CUBN and disease loci (X2=0.39, df=1, P=0.53).

To test another candidate gene for this canine model with selective
intestinal cobalamin malabsorption, a 2290 bp-canine megalin genomic DNA
fragment was amplified by PCR using primer pairs designed from conserve
regions of human and rat megalin. The alignment between dog megalin cDNA
amplified from dog kidney total RNA by RT—PCR using the same pair of primers
and this canine genomic DNA fragment defined the partial intron/exon structure
of canine megalin. This region contained two introns (1.1 kb and 0.6 kb) and 1
included exon (156 bp) (Figure 4.3, Panel A). The canine megalin cDNA
fregment showed 88% identity to human megalin, which confirmed the sequence

specificity.

119

.3000. 000508000: or 00265 octamto cm :0 0:300: 0:... .0: 03 0:0 000 :0

$0302: 050000.220 050309: .:o=0:0> 00-00 0:: 0:2:0: 0.0:..5 050: mun. >0 005.9:0 003 9.50:0 E0: <20
00.080: 00 00.2.0 0:0 .00.0E 0.0 00035 .000 .950: 0 00:00.05 .onE>m->EEO 05 0:0 .90.:00 00.09.00 0020:.
0.00.50 0:00.20: .0000 00000.00 m0. 0:00.05 0.88% 0_.ow Am. .0.0>.0:0 0000.5. :0.— .0xL0E .0E0E. 00 000: 002,

0:0 :22. 9.0.0 05 :_ 00:050. 002, 080 0:00 :o.~0.:0> 3.00 < 20:00:: 05 050:0:000 0:0 0:060: 0020050
5.000.: :0: 0:0 :0E2. E0: 00:0.000 0.0800 (200 5.000;: 0.50: 50005380 mun. .5 005000 0:03 068 :90
00020:. 0:0 .005. .0.:oN_:o£ 0:055 02:. 2. ._00_ 5.80:. 0:0 m0. 05:00 no 5000.500 E00:0:00:. ”0.0 0.50.“.

120

 

 

.4. a.
E
Home
I \
, It \ LIIIJ
MT 8. m... rflT e. E i
m 52: u_ 82 u. .55

121

A canine megalin intron variation was determined within the 0.5 kb intron.
Subclone and sequence results of the PCR amplified 0.5 kb intron of an affected
dog genomic DNA showed that one allele contained a 68 bp insertion
(AGCTGGTCTGTITTCTT(T28)TGT(T15)TT), but no insertion was found in the
other allele. This result was in conflict with the disease locus, which is
homozygous for the affected dogs. In contrast, the obligate carrier dog, which
was heterozygous for the disease locus, showed homozygousity for the megalin
intron variation, with no insertion in either allele. At this point, the data suggested
that the disease locus was not genetically linked with the megalin locus in this
family of dogs with selective intestinal Cbl malabsorption. Further family linkage
study with 20 offspring (12 affected dogs and 8 carrier dogs) (Figure 4.3), using
allele specific PCR amplification with primer 1553F and 2009R, producing a 390
bp product for the normal allele and a 456 bp product for the insertion allele,
showed 10 recombinants (0.5 recombination fraction) and confirmed independent
segregation of the disease locus and the megalin locus in this family of dogs with

selective intestinal Cbl malabsorption.

122

Discussion

Cubilin is a biologically compelling candidate gene for selective intestinal
cobalamin malabsorption with proteinuria. Cubilin is a multi-ligand receptor
expressed in the small intestinal epithelium where it mediates absorption of
cobalamin via binding of lF-Cbl complexes (Seetharam et al. 1997) Cubilin is
also expressed in the renal proximal tubule epithelium where it mediates
reabsorption of specific proteins filtered by the glomerulus (Batuman et al., 1998;
Birn et al., 2000; Fyfe et al., unpublished data). CUBN mutations causing
selective intestinal cobalamin malabsorption with proteinuria in Finnish kindreds
have been defined (Aminoff et al., 1995; Kozyraki et al., 1998).

The canine model of selective intestinal cobalain malabsorption with
proteinuria is a well-defined, simple autosomal recessive disorder for which
CUBN was also an attractive candidate gene. Intrinsic factor and transcobalamin-
ll functions of the affected dogs were demonstrated to be normal, and
immunoelectron microscopy revealed lack of cubilin in the apical membrane of
distal intestinal epithelial cells (Fyfe et al., 1991a). Measurements of cubilin
ligand-binding activity and western blotting of fractionated cells demonstrated
that the cubilin protein was expressed in affected dog ileal and proximal tubule
epithelial cells but did not reach the apical brush border membrane. Cubilin
purified by detergent solubilization and affinity chromatography from normal and
affected dog renal cortex comigrated on nonreducing SDS-PAGE gels but
yielded different peptides upon proteolytic digestion. Additionally, Asn-linked

oligosaccharides on affected dog cubilin were in an Endo H sensitive form,

123

suggesting retention in the endoplasmic reticulum (ER) and failure of the nascent
protein to make its way through the biosynthetic pathway (Fyfe et al., 1991b).

These findings were analogous to findings in a number of disorders of
secretory and membrane protein expression in which mutations of specific genes
cause failure of the translation product to fold properly and to remain associated
with ER chaperone proteins, to self-aggregate, or to be degraded rapidly (Kopito
1999; Bross et al., 1998; Kim and Arvan 1998). With the exception of
abetalipoproteinemia (MTP, MIM #200100), all inherited ER
retention/degradation disorders described to date are attributed to mutations in
the coding sequence of the particular secreted or membrane protein. The
similarities of the cloned canine CUBN cDNA sequence with those of other
species and the match of cubilin peptides to the deduced amino acid sequence
of the canine clone confirmed its identity (Chapter 2). Elimination of linkage of an
lntragenic CUBN marker to the inherited disorder in this canine family indicates
that failure of apical membrane expression of cubilin and the resultant selective
cobalamin malabsorption is caused by dysfunction of a gene product whose
locus maps 2 50 cM from CUBN in the canine genome. Because cellular
physiology of humans and dogs is highly homologous, these findings suggest
that human l-GS may exhibit locus heterogeneity in addition to allelic
heterogeneity at the CUBN locus. Human l-GS remains to be mapped in
kindreds of various ethnic backgrounds.

Genes unlinked to the human MGA 1 locus which may explain our findings

include the gene encoding the endocytic receptor, megalin (gp330, LRP2),

124

mapped to 2q31 (Chowdhary et al. 1995) which has been demonstrated to
associate and co-localize with cubilin (Birn et al. 1997; Moestrup et al. 1998). It
has been postulated that megalin may anchor cubilin to the apical membrane
and mediate its vesicular trafficking because cubilin lacks a transmembrane
domain or other obvious membrane anchoring signal. Megalin, which is closely
related to LRP also binds the urokinase-inhibitor complex (Nykjaer et al., 1997;
Herz et al., 1992), and may perform a similar function in regard to both the
urokinase receptor and cubilin-ligand complexes. The rejection of linkage
between the canine megalin locus and the disease locus of this canine model
with selective cobalamin malabsorption indicates that the failure of membrane
expression of cubilin in affected dogs was caused by a defect of a gene other
than megalin.

To locate the disease-causing gene for this canine model with selective
cobalamin malabsorption, instead of genome-wide screening, linkage analysis of
candidate genes were performed. Another candidate gene product is
an ER chaperone, Iipoprotein related protein-associated protein 1 (LRPAP 1)
or RAP. RAP interacts with cubilin in vitro and maps to 4p16.3 (VanLeuven et al.
1995) of humans. By analogy to the function of association with megalin (LRP2)
and the low density lipoprotein receptor-related protein 1/02-macroglobulin
receptor (LRP1), RAP may mediate proper folding of cubilin and prevent
premature ligand binding to cubilin within the ER and later compartments of the
biosynthetic pathway (Bu and Schwartz 1998). RAP is essential for the proper

folding and disulfide bond formation of LRP1, findings supported by ER retention

125

 

and rapid degradation of LRP1 and other members of the LDL-receptor gene
family in RAP-knockout mice (Willnow TE et al., 1996; Oberrnoeller et al., 1998).
Future investigations of canine selective intestinal cobalamin malabsorption in

this family will focus initially on the LRPAP1 or RAP locus.

126

References

Abdelaal MA and Ahmed AF. (1991) Case report: Imerslund-Grasbeck syndrome
in a Saudi family. Acta Padiatr Scand 80: 1109-1112.

Altay C, Cetin M, Giimri’ik F, Irken G, Yetgin S. and Laleli Y. (1995) Familial
selective vitamin B12 malabsorption (Imerslund-Grasbeck syndrome) in a pool of
Turkish patients. Pediat Hematol Oncol 12:19-28.

Aminoff M, Tahvanainen E, Grasbeck R, Weissenbach J, Broch H, and de la
Chapelle A. (1995) Selective intestinal malabsorption of vitamin B12 displays
recessive mendelian inheritance: assignment of a locus to chromosome 10 by
linkage. Am J Hum Genet 57: 824-31.

Aminoff M, Carter JE, Chadwick RB, Johnson C, Grasbeck R, Abdelaal MA,
Broch H, Jenner LB, Verroust PJ, Moestrup SK, Chapelle A, and Krahe R. (1999)
Mutations in CUBN, encoding the intrinsic factor-vitamin B12 receptor, cubilin,
cause hereditary megaloblastic anaemia 1. Nat Genet 21: 309-313.

 

Batuman V, Verroust PJ, Navar GL, Kaysen JH, Goda FO, Campbell WC, Simon
E, Pontillon F, Lyles M, Bruno J, and Hammond TG. (1998) Myeloma light chains
are ligands for cubilin (gp280) Am J Physiol. 275(R44): F246-F254.

Ben-Bassat I, Feinstein A. and Ramot B. (1969) Selective vitamin B12
malabsorption with proteinuria in lsraeI-—Clinical and genetic aspects. Israel J
Med Sci 5: 62-69.

Birn H, Verroust PJ, Nexo E, Hager H, Jacobsen C, Christensen El, and
Moestrup S. (1997) Characterization of an epithelial ~460-kDa protein that
facilitates endocytosis of intrinsic factor-vitamin B12 and binds receptor-
associated protein. J Biol Chem 272(42): 26497-26504.

Birn H, Fyfe JC, Jacobsen C, Mounier F, Veroust PJ, Qrskov H, Willnow T,
MoestrupSK, and Christensen El. (2000) Cubilin is an albumin binding protein
important for renal albumin reabsorption. J Clin Invest (in press).

Bork P and Beckmann G. (1993) The CUB domain. A widespread module in
developmentally regulated proteins. J Mol Biol 231: 539-545.

Bross P, Andresen BS, and Gregersen N. (1998) Impaired folding and subunit
assembly as disease mechanism: the example of medium-chain acyl-CoA
dehydrogenase deficiency. Prog Nucleic Acid Res Mol Biol 58: 301-337.

Bu G and Schwartz AI (1998) RAP, a novel type of ER chaperone. Trends Cell
Biol 8: 272-276.

127

Chowdhary BP, Lundgren S, Johansson M, Hjalm G, Akerstrom G, Gustavsson I,
and Rask L. (1995) In situ hybridization mapping of a 500-kDa calcium-sensing
protein gene (LRP2) to human chromosome region 2q31->q32.1 and porcine
chromosome region 15q22—>q24. Cytogenet Cell Genet 71: 120-123.

Fyfe JC, Giger U, Hall CA, Jezyk PF, Klumpp SA, Levine JS, and Patterson DF.
(1991a) Inherited selective intestinal cobalamin malabsorption and cobalamin
deficiency in dogs. Pediatr Res 29: 24-31.

Fyfe JC, Ramanujam KS, Ramaswamy K, Patterson DF, and Seetharam B.
(1991b) Defective brush-border expression of intrinsic factor-cobalamin receptor
in canine inherited intestinal cobalamin malabsorption. J Biol Chem 266: 4489-
4494.

Grasbeck R, Gordin R, Kantero I, And Kuhlback B. (1960) Selective vitamin B12
malabsorption and proteinuria in young people. Acta Med Scand 167:289-296.

Herz J, Clouthier DE, and Hammer RE. (1992) LDL receptor-related protein
internalizes and degrades uPA-PAl-1 complexes and is essential for embryo
implantation. Cell 71 : 411-421.

Imerslund O. (1960) Idiopathic chromic megaloblastic anemia in children. Acta
Paediat Scand Suppl 49: 1-115.

Kim PS and Arvan P (1998) Endocrinopathies in the family of endoplasmic
reticulum (ER) storage diseases: disorders of protein trafficking and the role of
ER molecular chaperones. Endocr Rev 19: 173-202.

Kopito RR (1999) Biosynthesis and degradation of CFTR. Physiol Rev 79: S167-
S173.

Kozyraki R, Kristiansen M, Silahtaroglu A, Hansen C, Jacobsen C, Tommerup N,
Verroust PJ, and Moesth SK. (1998) The human intrinsic factor-vitami Biz
receptor, cubilin: molecular characterization and chromosomal mapping of the
gene to 10p within the autosomal recessive megaloblastic anemia (MGA 1)
region. Blood 91 (10): 3593-3600.

Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D,
Pontillon F, Goda F, Christensen El, Hammond TG, and Verroust PJ. (1998) The
intrinsic factor-vitamin B12 receptor and targek of teratogenic antibodies is a
megalin-binding peripheral membrane protein with homology to developmental
proteins. J Biol Chem 273(9): 5235-5242.

Nykjaer A, Conese M, Christensen El, Olson D, Cremona O, Gliemann J, and

Blasi F. (1997) Recycling of the urokinase receptor upon internalization of the
uPAzserpin complexes. EMBO J 16: 2610-2620.

128

Obermoeller LM, Chen Z, Schwartz AL, and Bu G. (1998) Caz+ and receptor-
associated protein are independently required for proper folding and disulfide

bound formation of the low density Iipoprotein receptor-related protein. J Biol
Chem 273: 22374-22381.

Sahali D, Mulliez N, Chatelet F, Laurent-Winter C, Citadelle D, Sabourin JC,
Roux C, Ronco P, and Verroust P. (1993) Comparative immunochemistry and
ontogeny of two closely related coated pit proteins. The 280-kd target of
teratogenic antibodies and the 330-kd target of nephritogenic antibodies. Am J
Pathol142: 1654-1667.

Sahali D, Mulliez N, Chatelet F, Dupuis R, Ronco P, and Verroust P. (1988)
Characterization of a 280-kD protein restricted to the coated pits of the renal
brush border and the epithelial cells of the yolk sac. Teratogenic effect of the
specific monoclonal antibodies. J Exp Med 167: 213-218.

Sambrook J, Fritch EF, and Maniatis J. (1989) Molecular cloning: A laboratory
manual, 2"d ed. Cold Spring Harbor, NY, Cold Spring Laboratory Press, 1989.

Seetharam B, Christensen El, Moestrup SK, Hammond TG, and Verroust PJ.
(1997) Identification of rat yolk sac target protein of teratogenic antibodies,
gp280, intrinsic factor-cobalamin receptor. J Clin Invest 99(10): 2317-2322

Van Leuven F, Hilliker C, Serneels L, Umans L, Overbergh L, De Strooper B,
Fryns JP, and Van den Berghe H. Cloning, characterization, and chromosomal
localization to 4p16 of the human gene (LRPAP1) coding for the alpha 2-
macroglobulin receptor-associated protein and structural comparison with the
murine gene coding for the 44-kDa heparin-binding protein. Genomics 25: 492-
500.

V\fillnow TE, Armstrong SA, Hammer RE, and Herz J. (1995) Functional

expression of low density lipoprotein receptor-related protein is controlled by
receptor-associated protein in vivo. Proc Natl Acad Sci USA 92: 4537-4541.

129

Future Direction

The major focus of future projects will be to identify the molecular
mechanism of this canine model of selective intestinal cobalamin malabsorption
which closely resembles the human Imerslund-Grasbeck syndrome. In this
thesis, we used linkage analysis of the disease locus with cubilin and megalin
gene loci and rejected the hypothesis that cubilin or megalin was the disease
causing gene product. The specific goal for future projects will be to generate
new candidate genes. From this point, two approaches can be used to identify
new candidate genes, which are functional analysis and positional mapping.

The rejection of genetic linkage of canine l-GS with canine cubilin and
megalin gene suggested that a gene product other than megalin may physically
interact with cubilin and be required in the biosynthesis or apical delivery of
cubilin to the plasma membrane. Identification of proteins which interact cubilin
will generate functional candidate genes.

Receptor associate protein (RAP), which interacts with cubilin in vitro
(Moustrup et al., 1999) will be studied first. RAP is an ER chaperone-like protein
(Bu et al., 1995) and may mediate proper folding of cubilin and prevent
premature ligand binding to cubilin within the ER. Human and rat RAP genes are
cloned and the intron/exon structure of each were identified (van Leuven et al.,
1995; ). An lntragenic marker within the canine RAP gene can be identified by
using the same approach as for megalin. Genetic linkage will be used to test

linkage to the disease condition.

130

If the RAP gene is not linked to the disease locus, other methods, like co-
immunoprecipitation, affinity chromatography, and 2-D gel electrophoresis, will
be used to identify the proteins associate with cubilin. By comparing the different
pattern between normal and affected dogs, association aspects of the proteins
specific to the disease condition can be idetified. The identity of each candidate
proteins can be determined by protein microsequencing and database searches.
Similar genetic linkage tests will be performed to each candidate gene by
lntragenic markers.

An alternative approach will be a whole canine genome scan to identify
positional candidate genes. 341 highly polymorphic markers of all types have
been placed on the 3rd generation dog genome maps with average resolution of
9.0 cM and 95% coverage of the genome. Recently, a high resolution radiation
hybrid map with 400 markers has been produced and partially integrated with the
linkage map. Informative type I markers will be tested for linkage to canine I-GS.
Candidate genes will be identified by comparing the linked region to respective
human and mouse genome maps and can be further screened by examining the
function and tissue expression of each gene. Candidates will be tested for
linkage to the canine disease locus with lntragenic markers.

The linked gene identified by either method mentioned above will be
cloned and examined for the disease-causing mutations by comparing the gene
sequence between the normal and affected dogs. The defined mutation(s) will be
further confirmed by checking for the mutation(s) in the normal dog population

and functional studies on the gene from both normal and affected dogs.

131

   

iiiijjjjijjjjjljiiiiii