sh 1:13”... , EU minim Jud .h. . , Jammy-h . n. i. c “any? :2. .u karma?! 3 a .151. a. io... WW I‘ 5.1.1.191i I103. .1 s y: .. kg??? {.flhd 5%» w”... h. ‘6». LIBRARIES MICHIGAN STATE UNIVERSITY EAST LANSING, MICH 48824-1048 This is to certify that the dissertation entitled A CANINE MODEL OF IMERSLUND-GRASBECK SYNDROME: GENETIC MAPPING, MUTATION IDENTIFICATION AND PhD FUNCTIONAL ANALYSIS presented by Qianchuan He has been accepted towards fulfillment of the requirements for the degree in the Department of Microbiology and Molecular Genetics IQLCLWL— Major ProT'éssor’s Signature 3 I 3 I I of Date MSU is an Affirmative Action/Equal Opportunity Institution —.—.--—u—.--.-.-.-.—.—. PLACE IN RETURN Box to remove this checkout from your record. “IO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2/05 cJCIRC/DateDuejmp. I 5 A CANINE MODEL OF IMERSLUND-GRASBECK SYNDROME: GENETIC MAPPING, MUTATION IDENTIFICATION AND FUNCTIONAL ANALYSIS Bv U Qianchuan He A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Molecular Genetics 2005 .5 ABSTRACT A CANINE MODEL OF IMERSLUND-GRASBECK SYNDROME: GENETIC MAPPING, MUTATION IDENTIFICATION AND FUNCTIONAL ANALYSIS By QIANCHUAN HE lmerslund-Grasbeck syndrome (I-GS, OMIM #261 100) is an autosomal recessive disease characterized by congenital selective intestinal malabsorption of cobalamin (vitamin BIZ) and proteinuria. Some human patients are known to carry mutations in the cubilin gene (CUBN), which encodes a multiligand receptor expressed on both enterocytes and kidney proximal tubule cells. A Giant Schnauzer (GS) kindred studied in our lab is a well-established animal model for human l-GS, and has contributed significantly to the research on l-GS and cubilin. Previous experiments demonstrated abnormality of cubilin in the affected dogs, but linkage analysis excluded the CUBN as the disease-causing gene. We performed a whole genome scan linkage analysis and linked the canine [—05 to a marker on dog chromosome 8, which is orthologous to human chromosome 14q. Type I markers were developed in the GS kindred to refine the mapping. Haplotype analysis narrowed the candidates region to be between markers EMLI and SIVA, a 5Mb interval predicted from the human genome. KNS2, a marker located in the interval, was in complete linkage disequilibrium with [-08, with a multipoint LOD score of 15.4. One gene in the interval, AMN ((u-nnionless), was a compelling candidate because it was demonstrated as the second l-GS gene in a recent human study. We cloned the cDNA of canine AMN based on human, rat and mouse AMN sequences. RT-PCR and genomic PCR disclosed an in-frame deletion of 33 bp in exon 10 ofAMN, which was predicted to abolish the hypothetical transmembrane domain of AMN. The deletion segregated with the disease in the GS kindred and was absent in 112 unrelated nomial dogs We also studied an Australian Shepherd (AS) kindred with similar clinical features to the GS kindred. Linkage analysis of this small pedigree to marker KNS2 gave a LOD score of 1.7, marginally suggesting linkage ofthe disease to AMN. Mutation screening in AMN identified a G>A transition in the start codon, which was predicted to abrogate translation initiation. The mutation segregated with the disease and was not seen in 112 chromosomes of unrelated normal dogs. A recent study showed that cubilin and AMN form a tight complex, cubam, which is crucial for endocytosis of certain ligands, such as intrinsic factor-Cobalamin (IF-Cbl). Our RT-PCR in multiple canine tissues demonstrated that both genes have high expression levels in ileum and kidney, although they have different expression profiles in some other tissues. In an lF-C bl pull-down assay, three AMN isofomis were present in the kidney homogenates ofa nomial dog but absent in the affected dogs of both kindreds. In a heterologous cell transfection system, wildtype canine AMN, but not the 33bp- deletion mutant, assists the membrane expression of cubilin. We therefore demonstrated in viva that the fundamental cause of [-68 is the failure to express functional cubam, and canine [-08 is an orthologue ofthe human disease. The two pedigrees harboring different AMN mutations provide us a unique opportunity to study the functions of AMN and cubilin directly in tissues of both affected and nomtal dogs, which is nearly impossible in the human study of l-GS patients. DEDICATION This dissertation is dedicated to: Ying Du, my wife, whose love and support give me the most important strength during my graduate studies; Dr. John Fyfe, a great mentor and scientist, who led me into the field of medical genetics, and shares with me the pleasure, difficulty and hope ofthe research. ACKNOWLEDGEMENTS I am grateful to my committee members. Dr. Karen Friderici, Dr. Ronald Patterson and Dr. John Wang, for their excellent guidance, insight and encouragement. I thank Dr. Friderici for providing me many opportunities to learn human genetics; I thank Dr. Patterson and Dr. Wang for allowing me to participate in their joint seminars, where I constantly learn new knowledge in biochemistry and RNA biology. I am thankful to all the members in my committee members” labs, for their help and enthusiasms. I cherish the friendship with this group of wonderful people. I appreciate .Dr. Patrick Venta and Dr. Donna Housley for invaluable suggestions to my experiments. I would like to thank Dr. Vilma Yuzbasiyan-Gurkan for the gift of many dog DNA samples. This thesis would not be possible without the help from Dr. Alejandro Schaffer at NIH. Dr. Paula Henthorn at University of Pennsylvania and Dr. Mette Madsen at University of Aarhus in Denmark. It is always a great pleasure to collaborate with these experts from different fields. Finally, I would like to thank Adam Kilkenney, Brittany Gregory and Meghan Drummond in our lab for their excellent technical assistance. \I TABLE OF CONTENTS List oftables ......................................................................................... viii List of figures ......................................................................................... ix List of abbreviations .................................................................................. xi Chapter I: Literature review ......................................................................... 1 Vitamin BIZ and its physiological importance .............................................. 1 Vitamin BIZ absorption review ............................................................... 2 Vitamin BIZ storage and distribution ........................................................ 5 Defects in cobalamin transport ................................................................ 5 Defects in cellular AdoCbl and MerI synthesis ........................................... 7 lmerslund-Grasbeck Syndrome ................................................................ 8 Cubilin/IFCIngZ8O .......................................................................... lZ AMN ............................................................................................ 18 Canine I-GS .................................................................................... 25 Summary ........................................................................................ Z8 Purpose and outline ........................................................................... 29 Chapter 2: Genetic mapping ofcanine lmerslund-Griisbeck Syndrome in a giant schnauzer family ..................................................................................... 30 Introduction .................................................................................... 31 Materials and methods ........................................................................ 34 Results .......................................................................................... 39 Discussion ...................................................................................... 48 vi Chapter 3: cDNA cloning of canine AMN and mutation screening in the Giant Schnauzer family .................................................................................................... 51 Introduction .................................................................................... 52 Materials and methods ........................................................................ 56 Results .......................................................................................... 61 Discussion ...................................................................................... 68 Chapter 4: Linkage analysis and mutation screening of an Australian Shepherd family with Imerslund-Griisbeck Syndrome ................................................................................. 72 Introduction .................................................................................... 73 Materials and methods ........................................................................ 76 Results .......................................................................................... 79 Discussion ...................................................................................... 88 Chapter 5: Functional studies ofcanine AMN mutations ................................................... 90 Introduction .................................................................................... 91 Materials and methods ........................................................................ 93 Results .......................................................................................... 98 Discussion .................................................................................... 108 Appendices ............................................................................................ 1 13 Future direction ..................................................................................... l 19 References ............................................................................................. lZZ vii LIST OF TABLES Table 1. Two-Point LOD score analysis ................................................... Table 2. Genes listed in the EMLI-SIVA interval ofhuman genome browser. Table 3. Primer list for genomic amplification ofcanine AMN. Table 4. Genotyping data ofthe Giant Schnauzer family Table 5. Genotyping data ofthe Australian Shepherd kindred. viii ......... 42 ...54 ....78 ..115 ...118 LIST OF FIGURES Figure 1.] Vitamin BIZ absorption in the gastrointestinal trath Figure 1.2 Intracellular transport and processing ofvitamin 812. 8 Figure 1.3 A model of AMN and cubilin assembly in the biosynthetic pathway and recycling in the endocytic apparatus ofpolarized epithelial cells.. .....22 Figure 2.1 Pedigree ofdogs of known I-GS status used for linkage analysis................40 Figure 2.2 Multipoint LOD score curve showing linkage to l-GS. 45 Figure 2.3 Haplotype analysis ofthe GS kindred” 46 Figure 2.4 An iterative strategy to develop new markers for linkage disequilibrium mapping .............................................................................................................................. 47 Figure 3.1 cDNA sequence ofcanine AMN64 Figure 3.2 Alignment ofhuman, dog, and mouse AMN protein.................................65 Figure 3.3 DNA sequencing revealed a 33bp in-frame deletion in canine AMN of the affected dogs. ................................................................................................ .66 Figure 3.4 Mutation analysis ofthe 33bp deletion in exon 10. 67 Figure 3.5 Structure prediction ofAMN by the TMHMM software. 71 Figure 4.1 KNSZ genotyping ofthe Australian Shepherd kindred. 75 Figure 4.2. Genomic sequence ofcanine AMN genc82 Figure 4.3 DNA sequencing revealed a G>A mutation in the start codon of canine AMN in the affected dogs. .................................................................................... 86 Figure 4.4 Mutation analysis ofthe G>A transition in start codon.............................87 Figure 5.1 RNA expression profiles ofcubilin and AMNIOZ Figure 5.2 Specificity test ofthe anti-AMN antibody. ........................................ 103 Figure 5.3 Cubilin and AMN expression in normal and I—GS affected dog kidneys in viva. ............................................................................................................. 104 Figure 5.4 Cubilin saturation test. ............................................................... 105 Figure 5.5 lmmunofluorescence of double transfectcd CHO cells expressing cubilin and Figure 5.6 Cubilin and AMN expression in CHO cells with wildtype or cl 1 13-1 l45del Figure 6 The pr polymorphism ofAMN identified in a komondor and a beagle ....... l 14 Ab AdoC bl AMN AS BBM BMP C aco Cbl CF A8 CUBN FISH GPI GS HDL IF IF -Cbl IFCR l-GS LDL LOD M erl MGA MMA MMAA MMAB MTRR OMIMTM OK RAP RER SAM TRAF3 UTR VE YAC LIST OF ABBREVIATIONS annbody adenosylcobalamin amnionless Australian shepherd brush border membrane bone morphogenesis protein colon adenocarcinoma Cobalamin canine chromosome 8 cubilin fluorescence in situ hybridization glycosyl—phosphatidyl-inositol Giant Schnauzer high-density lipoprotein intrinsic factor intrinsic factor-Cobalamin intrinsic factor-cobalamin receptor Imerslund Grasbeck Syndrome low-density lipoprotein logarithm of odds methylcobalamin megaloblastic anemia methylmalonyl acidemia MethylMalonic Acidemia linked to the cblA complementation group MethylMalonic Acidemia linked to the cblB complementation group methionine synthase reductase Online Mendelian Inheritance in Mann opossum kidney Receptor Associated Protein Rough endoplasmic reticulum S-adenosyl methionine TNF receptor associated factor 3 untranslated region visceral endoderm Yeast artificial chromosome '1 xi CHAPTER 1 LITERATURE REVIEW Vitamin 812 and its physiological importance Vitamin BlZ (cobalamin) has been known as an anti-pernicious factor since 19205, when it was initially called “extrinsic factor”. A unique feature of vitamin BIZ is that a cobalt atom is located in the center of a corrin ring (Banerjee and Ragsdale, 2003 review). Different upper ligands may be linked to the cobalt atom, as shown in vitamin BIZ and its derivatives, such as methylcobalamin (MerI) and adenosylcobalamin (AdoCbl) (Banerjee and Ragsdale, 2003 review). All of vitamin B12 in nature is of microbial origin (Rosenblatt and Fenton, 2001 review). The biosynthesis of B12 in microbes is quite complicated, involving more than 30 genes (Roth et al., 1993). Bacteria and archaea use vitamin BIZ for acetyl-CoA synthesis, methyl transfer and fem1entation. etc (Martens et al., 2002 review). Animals (including humans) need cobalamin but do not synthesize it. Therefore, vitamin 812 must be absorbed from the diet. In animals and man, BIZ is known to participate in only two enzyme reactions, which are methionine synthesis and methylmalonyl-CoA isomerization (Okuda, 1999 review). In the cytoplasm, 812 (in the form of MerI) is a cofactor for methionine synthase, which catalyzes the conversion of homocysteine to methionine. In the mitochondria, 812 (in the form of AdoCbl) is an essential coenzyme for methylmalonyl-CoA mutase. which converts methylmalonyl—COA to succinyI-COA. Defects in these biochemical pathways lead variously to methylmalonic acidemia (MMA), homocystinuria, megaloblastic anemia and neurological symptoms (Stabler and Allen, 2004 review). BIZ deficiency may also affect the health of humans indirectly. Since BIZ is required in the synthesis of methionine, B12 deficiency also leads to low level of methionine and subsequent S-adenosyl methionine (SAM) deficiency. This can cause genomic instability in cells. because a low level of SAM may diminish DNA methylation. (Fenech, 2001review; Brunaud et al., 2003). Vitamin BlZ absorption review DIETARY COBALAMIN Cbl PANCREATIC PROTEASES IF-Cb / ILEAL RECEPTOR Figure 1.1 Vitamin BIZ absorption in the gastrointestinal tract (Seetharam and Alpers, 1982a review). Cbl, cobalamin; IF, intrinsic factor; R, R protein. Cobalamin (Cbl) absorption is a highly specialized multi-step process in which several proteins are involved (Figure 1.1). Released from food, C bl is bound to R protein in the stomach to form a stable complex. R protein, also called TCI or haptocorrin, is a glycoprotein and has high affinity for Cbl at stomach pH (Allen and Majerus, 19723). R protein is widely distributed in the saliva, bile, blood and other cells (Sennett and Rosenberg, 1981 review). Cbl-TCI complex remains stable in the intestine until pancreatic proteases degrade TCI and release Cbl for binding to intrinsic factor (Allen et al., 19783). In patients with cobalamin malabsorption due to pancreatic insufficiency, a cobalamin analogue that binds tightly to R protein but not to intrinsic factor can correct the defect, because cobalamin can circumvent the binding to R protein (Allen et al., 1978b). Several individuals were known to have deficient or absent TCI, but they didn’t have Cbl malabsorption, and they had no signs of Cbl deficiency except low serum Cbl concentration (Cooper and Rosenblatt, 1987). Intrinsic factor (IF) was so named because it is produced by the body and was used to prevent pernicious anemia (vitamin BIZ was named the extrinsic factor). With R protein degraded by proteolytic enzymes, IF binds to Cbl to form the noncovalent IF-Cbl complex (Tang et al., 1992). IF is a 44 kDa glycoprotein and is highly resistant to proteolytic degradation (Allen and Mehlman, 1973; Allen et al., 1978a). Tissues that produce IF may not be the same in differeiit species. Human IF is synthesized by the gastric parietal cells of the stomach (Levine et al., 1980), whereas canine and feline IF are pancreatic in origin (Vaillant et al., 1990; Fyfe, 1993). The IF-Cbl complex is endocytosed into enterocytes via intrinsic factor-cobalamin receptor (IFCR), expressed on the apical membrane of the ileal enterocytes (Sennett and Rosenberg, 1981 review). IF-Cbl and IFCR have a disassociation constant of 2.5x10‘m M, but Cbl or IF alone has no affinity to IFCR, suggesting that Cbl binding to IF triggers a conformational change. The binding requires pH between 6 and 9 and is Ca2+ dependent (Mathan et al., 1974), explaining why IF doesn’t bind to Cbl in the stomach. IFCR was later named cubilin for its unique structural features (discussed later in Cubilin/IFCR /gp280). Almost all the absorption of IF-Cbl by cubilin occurs in the ileum, because cubilin is mainly expressed in the ileum (Booth and Mollin, 1959; Xu and Fyfe, 2000). A recent study reported that another protein, amnionless (AMN), is indispensable for the expression and function of cubilin (Fyfe et al., 2004). The transport of IF -Cbl across the ileal enterocyte to the portal venous plasma takes several hours (Doscherholmen and Hagen, 1959). Three steps are already known, including degradation of IF in lysosomes. binding of Cbl to TCII, and transport of Cbl- TCII into the blood (Kapadia et al., 1983; Ramasamy et al., 1989). Human colon adenocarcinoma (C aco-Z) cells were used to study IF-Cbl internalization because they can express both IFCR and TCII and are able to direct Cbl transcytosis (Dix et al., 1990; Ramanujam et al., 1991). Study with Caco-2 cells demonstrated that IF degradation is a prerequisite to release of Cbl into the cytosol (Dan and Cutler, 1994). The accurate location for the formation of TC lI-Cbl has not been solved because TCII was synthesized at a low level and didn’t accumulate inside the Caco-2 cells (Seetharam, 1999 review). Although the majority of plasma Cbl is bound to TCI in humans, it is TCII that mainly mediates uptake of Cbl by tissues (Rosenblatt and Fenton, 2001 review). TCII is a 60 kDa non-glycosylated protein (Allen and Majerus, 1972b), distributed in plasma and a host of other extracellular fluids (Cooper and Rosenbaltt, 1987 review). Using 1251- labeled TCII, Youngdahl-Turner et al. (1978) first described a specific TCII-Cbl receptor existing on the cell surface of human fibroblasts, and confirmed that TC 11 was degraded within lysosomes. TC 11 receptors were later observed on human leukemia K562 and HL- 60 cells, and the expression level of the receptors is related to the growth conditions of the cells (Amagasaki et al., 1990). The receptor was purified from placenta membranes and characterized to be a 58 kDa glycoprotein with a dissociation constant of 2.6x10'm M (Quadros et al., 1994). Once TCII-Cbl is endocytosed into the cell. Cbl is released from the lysosomes in a pH-dependent manner (Idriss and Jonas, 1991). Subsequently, Cbl is processed into Mer1 or AdoCbl, and serves as coenzymes for methylmalonyl-COA mutase or methionine synthase (Marsh, 1999 review). Vitamin BIZ storage and distribution Experiments with C000 labeled Cbl showed that liver is the reservoir for absorbed Cbl (Schloesser et al., 1958). Some ofthe Cbl stored in liver is secreted into the intestine through biliary excretion, so called enterohcpatic recirculation (Willigan et al., 1958). Other organs showing high concentration of C bl include pituitary, kidney and pancreas (Coopemian et al., 1960). Defects in cobalamin transport Defects in cobalamin transport may lie in problems with one or more of the following steps: gastric intrinsic factor synthesis, pancreatic secretion, intestinal absorption, one of the circulating cobalamin binding proteins (TCII), or cellular absorption. Defects in intrinsic factor (IF) cause a disease called congenital pernicious anemia (OMIM#261000). Intrinsic factor in these patients either is highly susceptible to degradation in the lumen ofthe gastrointestinal tract (Yang et al., 1985) or has extremely low affinity to the IFCR (Katz et al., 1974b). The IF gene was cloned and mapped to chromosome 1 1 (Hewitt et al., 1991). Five patients were found to carry a nonsynonymous variation near the start codon, which changes glutamine to arginine (Gordon et al., 2004). However, since the variant was also carried in two control populations, it is unlikely to be the direct cause of the disease (Gordon et al., 2004). The long awaited mutation in [F gene was identified by Yassin et al. (2004) as a 4 bp deletion in exon 2, resulting in frameshift and predicted loss ofthe protein. Ten more mutations were identified in the IF gene recently (Tanner et al., PNAS 2005, in press). C bl malabsorption due to intestinal defects is called Imerslund-Grr’isbeck syndrome and will be discussed later. Genetic defects in TCII result in severe megaloblastic anemia in infants. The TCII gene is located on chromosome 22 (Arwert et al., 1986), and the cDNA encodes an extracellular protein of 409 amino acids (Platica et al., 1991). Mutation screening in the TC 11 gene of an affected child disclosed a compound heterozygote with deletions in both alleles, which led to failed TCII protein expression (Li et al., 1994). However, the most popular TCII variation in Caucasians is a single amino acid substitution, which significantly lowers the concentration of bound BIZ and increases the methylmalonic acid concentrations (Miller et al., 2002). Interestingly. it was noticed that human TCII, TCI, and IF gene sequences share high homology at some exon-intron boundary sites, suggesting that these genes were originated from a common ancestral gene by duplication mechanism (Regec et al., 1995). However, no patients have been implicated with TC II-receptor deficiency so far. The reason could be that either the deficiency is extremely rare or, most likely, is embryonic lethal, or some compensatory mechanisms exist such that people having TC ll-receptor defects do not show any symptoms. Defects in cellular AdoCbl and MerI synthesis Following entry of Cbl-TCII into the target cells, TCII is digested by enzymes in the lysosomes, with free Cbl released into the cytosol (Figure 1.2). Distinct defects in enzymes required for intracellular transport and processing of BIZ have been named as different cbl groups, based on complementation tests (Gravel et al., 1975). CblG group is due to mutations in the gene encoding methionine synthase (Gulati et al., 1996; Leclerc et al., 1996; Watkins et al., 2002), whereas cblE is caused by mutations in the methionine synthase reductase (MTRR) gene (Wilson et al., 1999; Zavadakova et al., 2002). The AdoCbl metabolism in both groups is intact. CblA complementation type is linked to mutations in the MMAA (MethylMalonic Acidemia linked to the cblA complementation group) gene, which encodes a protein that transports Cbl from cytosol into mitochondria (Dobson et al., 2002a). CblH has similar clinical and biochemical features to cblA, but the molecular basis is unknown. Methylmaltmyl CoA Mutase L-mcthylmalonyl COA p Succinyl C‘oA AdoCbl T (I)! I} (‘bl l mitm‘lmm/rirm ch] .4. ('1)! H (11103111 Cbllll—p Chill ('h/ F ('h/ (I ('b/ 1) MS reductase ('l)! E . McC bl . Homocystcme ' lVlethiomne lVlethioninc synthase (Ii/(1' Figure 1.2 Intracellular transport and processing of vitamin B12 (modified from Dobson et al., 2002a) Mutated MMAB (MethylMalonic Acidemia linked to the cblB complementation group) gene, encoding the cobalamin adenosyltransferase. is responsible for cblB type (Dobson et al., 2002b). The genes responsible for cblF, cblC and cle remain unknown. Imerslund-Grasbeck Syndrome History and clinical features Imerslund-Grasbeck syndrome (I-GS) was first described by Imerslund (1960) in Norway and by Grasbeck et al. (1960) in Finland. The patients showed familial juvenile vitamin BIZ deficiency anemia, methylmalonic acid excretion and proteinuria. By Schilling test, which measures the absorption oforally administered radioactive BIZ, the cause of the vitamin B12 deficiency was identified as decreased intestinal absorption from the diet. Other possible causes ofthe disease were excluded, such as (i) depletion of the vitamin by noxious intestinal bacteria, (ii) general dysfunction ofthe gastrointestinal tract and (iii) autoimmune mechanisms (Imerslund and Bjomstad, 1963). Since the patients were unresponsive to intrinsic factor, the syndrome is distinct from the congenital pemicious anemia, which is due to the lack of intrinsic factor in the gastric juice. To be diagnosed as I-GS, a patient generally should be no more than five years of age, and not have intrinsic factor insufficiency. However, the most important criteria are that patients are responsive to parenterally administered BIZ and Schilling test confirms BIZ malabsorption (Mohamed et al., 1966). Most of the patients had mild proteinuria, which was resistant to the parenteral administration of BIZ. In addition, the clinical features of I-GS may include methylmalonic aciduria (Rubin et al., 1974), homocystinuria (Mohamed et al., 1966), minimal glomerulonephritis (Becker et al., 1977; Rumpelt and Michl, 1979), subacute combined degeneration of the cord (Ismail et al., 1997) and neuropathy (Ludvigsson et al., 1980). Other complications were seen in rare cases, such as dolichocephaly (Ben-Ami et al., 19,90), distinct skin lesions (Lin et al., 1994), recurrent urinary tract infection and genitourinary tract abnormalities (Sandoval et al., 2000). Whether these complications are directly related to vitamin BIZ deficiency is uncertain. The defect was deemed congenital, because the time of onset of the symptoms corresponded to the sustaining time of the child’s stored vitamin BIZ conveyed from the mother in utero (Hippe, 1966). With a single in vitro experiment, Mackenzie et al. (1972) stated that the vitamin BIZ absorption defect occurred at a stage after the binding of IF-Cbl to IFCR and before the fom1ation of Cbl-TCII complex, indicating that the IFCR was intact. In contrast to this finding, studies ofthe absorption of radioactive BIZ by ileal biOpsy tissues pointed to a defect in the ileal receptor for IF-Cbl binding (Burman et al., 1985). The latter statement was corroborated by experiments showing that I-GS patients have a sharply decreased IFCR activity (Gueant et al. 1995). Thus, I-GS was strongly suggested to be associated with a defect in IFCR. Epidemiology More than 200 human cases and familial clusters have been reported world wide (Rosenblatt and Fenton, 1999). Whereas the original patients were identified from Finland and Norway, 3 number of patients from the Middle East have been reported in the past twenty years (Bum1an et al., 1985; Ben-Ami et al., 1990; Salameh et al., 1991; Abdelaal and Ahmed, 1991; Altay et al., 1995; Ismail et al., 1997). Other sporadic cases have been reported from Gennany (Bienzle et al., 1976), Libya (e1 Mauhoub et al., 1989), China (Lin et al., 1994), United States (Sandoval et al., 2000), and other countries. It is worth noting that only a few new cases have been diagnosed in Finland and Norway during the past thirty years, which was speculated to be due to the change of population structure (lower inbreeding coefficient) and dietary habits (more meat and therefore more 812) (Aminoff et al., 1995; Kristiansen et al., 2000). Therapy Parenteral vitamin BlZ treatment resulted in complete clinical and hematological remission except proteinuria (Mohamed et al., 1966). This finding was confim1ed by a 10 long term follow-up study, in which the prognosis of the patients are excellent on intramuscular vitamin BIZ therapy, but those who previously had proteinuria symptom keep excreting an average of 750 mg of protein in the urine per day (Broch et al., 1984). A patient with severe neurological abnom1alities was reported to have a remarkable recovery from spasticity, truncal ataxia and brain atrophy, by parenteral vitamin BIZ therapy (Salameh et al., 1991). Genetics Since one of the two patients in the earliest finding was from a consanguineous family, it was speculated that there might be an inherent defect of an acceptor for B12 or transport system in the enterocytes (Grasbeck et al., 1960). The familial occurrence of the syndrome in other patients also indicated that l-GS was a genetic disease (Imerslund and Bjomstad, 1963). Study of 18 cases in 14 families in Israel demonstrated that both sexes were affected and the parents were unaffected and usually shared common ancestors, highly suggesting that I-GS was an autosomal recessive disease. The gene frequency among Israeli Jews of Tunisian origin was estimated to be 0.025 and the frequency of heterozygotes (carriers) 0.05 (Ben-Bassat et al., 1969). Furuhjelm and Nevanlinna (1973) mentioned the gene frequency of I-GS in Finns was between 0.002 and 0.003. They further argued that, while founder effect causes disease allele enrichment in Finland, consanguinity is the main reason causing I-GS in the Jewish families. Because only anemia responded to parenteral vitamin BIZ therapy, it was postulated that the two traits, anemia and proteinuria, are caused by a multi-functional gene or two genes somehow related (Rumpelt and Michl, 1979). Later experiments in canine I-GS 11 showed IFCR’s mistrafficking in both intestine and kidney, suggesting that these defects are associated with a single gene (Fyfe et al., I991b). Cubilin/IFCR/gp280 Terminology People have realized for a long time that IF-Cbl must bind to an ileal receptor before being absorbed by the enterocytes and, therefore, named it the intrinsic factor- cobalamin receptor (IFCR) (Okuda, 1962). cDNA cloning of the rat IFCR disclosed a protein containing 27 CUB domains, thus the receptor was renamed cubilin (Moestrup et aL,l998) In an independent study, a 280 kDa glycoprotein (gp280) located on the proximal tubule cells and visceral yolk sac membrane was found to be likely involved in endocytosis activities (Sahali et al., 1988). Functional and immunological analysis proved that gp280 is identical to cubilin (Seetharam et al., 1997). Although one group in France was arguing that IFC R is distinct from cubilin on some limited information (Gueant et al., 2001), it is now generally accepted that cubilin, IF CR, gp280 represent the same protein. Tissue distribution Ileum from different species has been used as a source to purify the IFC R (cubilin) in early studies (Marcoullis et a1, 1977; Yki-Jarvinen et al., 1979; Kouvonen 1980; Seetharam et al. 1981). Immunocytochemistry localized the IFCR in the crypt, mid- villus, and villus tip areas ofthe ileal mucosa (Levine et al., 1984). In addition, the IFCR Was f‘o und to have high expression in human, canine and rat kidneys and small amounts in rat placenta and liver (Seetharam et al., 1988). Immunomicroscopy revealed that the receptor is located on the apical surface of proximal tubular cells (Seetharam et al., 1988) - It is also expressed on rat yolk sac apical membranes (Sahali et al., 1988; 535th aram et al., 1997), and male and female rat reproductive tissues (Van Praet et al., 2003 t, Argraves and Morales, 2004). EQtnd composition The size and composition of IF CR (cubilin) have perplexed the field for more than 2 0 years. The porcine IFCR was erroneously claimed to be purified as an 80 kDa protein Cgmposed of two subunits (Marcoullis et al, 1977). The IFCR was later purified from C Elnine ileal mucosa by hog IF—C bl affinity chromatography, showing molecular weight of 1 80 kDa on SDS-PAGE, disassembling into two subunits at reducing conditions ( S eetharam et al. 1981). However, the IF CR purified from dog kidneys showed molecular W’ eight of 230 kDa on SDS-PAGE but 457 kDa by amino acid analysis (Seetharam et al., 1 988; Fyfe et al., 1991b). Furthermore, the kidney IFCR remains to be a single band even at strong reducing condition (Seetharam et al., 1988). The apparent inconsistency regarding the molecular weight and compositions of the IFCR was not well explained until the cDNA of the cubilin was cloned in the late 19903. IFCR (cubilin) can be affinity purified on a RAP column. Receptor Associated Protein (RAP) is a 39 kDa endoplasmic reticulum protein serving as a molecular chaperone for low density lipoprotein receptor-related protein (Bu et al, 1995). One of its binding partners is megalin, which was indicated to mediate the endocytosis of TCII-Cbl 13 Comp] exes in kidney proximal tubules (Moestrup et al., 1996). Through a RAP affinity column, the IFCR of rabbit kidney was purified as a 460 kDa protein (Bim H et al., 1997) - By immunoscreening of a AZap cDNA library constructed from yolk sac cells, the rat Cubilin cDNA was first cloned (Moestrup et al., 1998). Rat cubilin has 3623 amino aCIdS if deduced from the cDNA sequence, while the mature protein migrates as a 460 kDa band on SDS-PAGE. Iftreated with peptidyl N-glycosidase F, the size is reduced to 400 kDa, suggesting that the N-linked oligosaccharides constitute about 60 kDa of matUre cubilin (Moestrup et al., 1998). Subsequent cloning of human and canine cubilin CDI\IA both agree with the size of 400 kDa (Kozyraki et al., 1998; Xu et al., 1999). A likewise explanation for the earlier findings is that the 230 kDa migration band was 51% tually 460 kDa (markers >200 kDa were not used in those studies), and the two SL1 bunits detected in intestine were proteolytic cleavage products of cubilin. fimcture and modification An early study suggested that IFCR (cubilin) has most of its mass in extracellular Space and a small portion inserted in the membrane (Seetharam et al., 1982b). cDNA Cloning of cubilin helped to dissect the protein’s structure. Preceding the mature form of a 3589-aa protein, the human cubilin has a Z4-aa signal peptide predicted to be cleaved in the endoplasmic reticulum, and a 10-aa peptide predicted to be cleaved by the trans-Golgi enzyme furin. The most prominent feature of cubilin is that it contains 8 EGF repeats and . Z7 CUB domains, suggesting cubilin is a multi-ligand receptor. Unexpectedly, as a receptor it has no transmembrane domain or glycosyl-phosphatidyl-inositol (GPI) anchor (Kozyraki et al., 1998). CUB domain 5-8 was demonstrated to be the Cbl—IF binding site 14 while CU B domains 13-14 bind to the receptor associated protein (RAP) (Kristiansen et al., 1999) - Two ofthe EGF repeats contain a Ca2+ binding motif(Moestrup et al., 1998), suggested to reflect previous finding that the interaction between IF-Cbl and the receptor is Ca2+ dependent (Katz and Cooper, 1974a). Forty-two potential N-glycosylation sites are present in cubilin (Moestrup et al., 1998), but the actual glycosylated sites may vary in different tissues. Studies in dogs suggested that the cubilin in the ileum is more highly glycosylated than that in the kidney, which may help cubilin to resist the proteolytic enzyn‘les in the ileum (Xu and Fyfe, 2000). Study in opossum kidney (OK) cells showed that IFCR is also palmitoylated (Ramanujam et al., 1994). Electron microscopy, analytical ultracentrifugation and computer-assisted analysis of bovine cubilin demons trated that cubilin in vitro fom1s a trimer by a a-helix structure located at the N- terminu s (Lindblom et al., 1999). Whether the trimer reflects the in vivo structure of CUbllln remains unknown. F—UMLQHS and Binding partners I F-Cbl, but not free IF or Cbl, binds to IFCR (cubilin). Apical membrane exPreSsion of IFCR is required for [57C0]Cbl transcytosis across polarized cells, that is, from the apical membrane to the basolateral membrane (Ramanujam et al., 1991). During the trElliseytosis, IF is degraded in lysosomes and free C bl is released for further transport by TCII (Dan and Cutler, 1994). Defective IFCR apparently can cause Cbl malab Sorption. In a different perspective, injection of anti-cubilin to mid-gestation rats resulted in {etaI resorption or malformations. Indirect immunofluoreseence study suggested that 15 cubilin mi ght be involved in endocytosis activities (Sahali et al., 1988). Further analysis showed that in the Ab-treated rat yolk sac both the size and shape of the early endocytosis vesicles were abnormal, resulting in disordered intracellular traffic of internalized proteins ( Le Panse et al., 1994). The teratogenic rat phenotype presented above indicates that cubilin may play a role in development. As a matter of fact, cubilin activity was observed to rise enormously in the rat placental membranes during the third week of pregnancy (Ramanujam et al., 1993}. One study showed that cubilin can be affinity-purified by galectin-3, a protein SUSSCSted to play a role in embryo development, from homogenates of murine UteroPlzlcental complex. In addition, cubilin colocalized with galactin-3 on yolk sac membrane in the last week of pregnancy. However, the anti-cubilin doesn’t seem to Pmduce the teratogenic effects via galactin-3, because the timing of the interaction betwee 1‘1 the two proteins was suggested to be at the late pregnancy (Crider-Pirkle et al., 2002), Tested in rat visceral yolk sac cells, cubilin also mediates the internalization of immu ldoglobulin light chains purified from the urine of myeloma patients. However, the bindin g of cubilin with light chain can strongly inhibit the function of the endosome in Vim" Therefore, excessive light chain present in the urine may lead to proximal tubular cytoto X icity (Batuman et al., 1998). I\/legalin (gp330) is a 600 kDa membrane protein belonging to the low density lipopI‘Qtein receptor gene family, and mediates endocytosis of several small proteins (Christensen et al., 1992; Saito et al., 1994). Megalin can be purified together with cubII‘x 11 from a RAP-Sepharose column (Bim et al., 1997), and it was suggested that 16 cubilin binds to the extracellular domain of megalin (Moestrup et al., 1998). Cubilin binds and colocalizes with megalin during the endocytosis process, as supported by surface p1 asmon analysis and electron microscopy, respectively (Moestrup et al., 1998). Cubilin and megalin act together to reabsorb certain proteins from the kidney proximal tubules. One example is albumin. Cubilin and megalin colocalize predonlinantly on the apical membrane of kidney proximal tubule cells, and mediate the reabsorption of albumin from the glomerular ultrafiltrate (Bim et al., 2000; Zhai et al., 2000)- Albumin was suggested to bind to cubilin at the same site as IF-Cbl, but with ~500- 750 fold lower affinity (Yammani et al., 2001). Other groups reported that megalin acts together with cubilin to mediate the endocytosis of apolipoprotein A-I receptor, a protein belonging to the high-density lipoprotein (HDL) family (Kozyraki et al., 1999; Hamrn ad et al., 1999; Hammad et al., 2000). The list of cubilin’s binding partners is SIOWing gradually, including transferrin (Kozyraki et al., 2001), vitamin D-binding protein (Nykjaer et al., 2001), hemoglobin (Gburek et al., 2002) and myoglobin (Gburek et al., 2003). Thus cubilin-megalin mediated endocytosis in the proximal tubule appears to be Q rucial to reabsorb a wide range of proteins. Another partner of cubilin, AMN, will be discussed later in this chapter. with the rapid progress of molecular genetics and computation tools in the early 905’ D Qsition cloning loomed as a powerful tool to discover the underlying genetic cause 0f “1h erited diseases. With six Finnish I-GS families (30 members) and three Norwegian fam'1\ i es (1 1 members), a whole genome scan linkage analysis of human I-GS defined the minimal candidate interval between markers D108586 and D10S570, which are 16-cM apart. The highest LOD (Logarithm of the Odds) score 5.36 was observed around DIOSI47 ‘7 (Aminoff et al., 1995). On the other hand, chromosome mapping by FISH (fluorescence in situ hybridization), radiation hybrid, and YAC (Yeast Artificial Chromosome) screening localized the cubilin gene within a 6 cM region on chromosome 1013 12, Close to marker D10S1477 (Kozyraki et al., 1998). Combined with previous findings that the ileal biopsy specimens of I-GS patients have a nearly undetectable IFCR activity (Gueant et al. 1995), these data strongly indicated that cubilin gene (CUBN) defects might cause I-GS (Kozyraki et al., 1998). Further linkage disequilibrium mapping and rn utation screening identified two different mutations in Finnish patients: the first mutation (FMl) is a missense mutation in the IF-Cbl binding domain, while the an mUIatiOn (FMZ) is a point mutation in the intron, causing abnormal splicing and essenti ally truncated proteins (Aminoff et al., 1999). Functional analysis demonstrated that th e FMl mutation impaired the binding site of cubilin to IF -Cb1, which subsequently led to vitamin B12 malabsorption (Kristiansen et al., 2000). Four more different mmations were later identified in Finnish, Bedouin and Turkish (Tanner et al., 2004). HoweV'er, no mutations were found in cubilin gene in the Norwegian and Saudi Arabian fami 1 i es (Aminoff et al., 1999), suggesting genetic heterogeneity exists. AMN Gene t i CS -_\. In 1996, a research group created a new transgenic mouse line by random insertion OI ‘hQ human CD80: transgene construct (T81) on the C57BL/6J (B6) mouse background. 18 Mice homozygous for the T81 insertion in chromosome 12 showed prenatal lethal phenotype (Wang et al., 1996). The insertion locus was mapped by interspecific backcross and FISH. Further molecular cloning work recovered the junction site sequence, which helped to pinpoint the insertion site to be close to the TRAF3 (TNF receptor associated factor 3) gene. Since the mutant does not develop a normal amnion, it Was narned amnionless (AMN) (Wang et al., 1996). Under the B6 background the mutants stopped embryonic development at E9.5, whereas under the B6x129 background the nautants developed till E10.5 with a shortened trunk, suggesting the genetic backgro und may modify the phenotype. A detailed characterization demonstrated that the amnfl’- mutant mouse is unable to generate middle streak derivatives (Tomihara- Newbe rger et al., 1998). It was found that the T81 transgene fragment (~200 kb) inserted into ir‘ltron 7 of the Amn gene, potentially abrogating expression of the gene product. Targeted mutation of Amn and complementation tests confim1ed that Amn is the gene responsible for the amnionless phenotype in mice (Kalantry et al., 2001). Surprisingly, mmations in AMN cause a very different phenotype in humans. In 2003, AMN was discovered by positional cloning to harbor mutations in the I—GS patients from Norway and the Middle East. A 1 bp deletion in the 5‘h codon was found in 3 Norwegian families “Ving i n proximity, which apparently resulted in frame shift of the translation. In another Noeregian family, a missense mutation was identified in the 415' codon, which might affect the structure of the protein because it changed a polar amino acid to a nonpolar aminQ acid at an evolutionarily conserved site. The third mutation, identified in an Israeli fam‘uy, was a splice site mutation that caused the total loss of exon 4 in the mRNA, whtch led to early translation termination (Tanner et al., 2003). Since then a total of six 19 different /1. MN mutations have been found in I-GS families from Norway, Turkey, Israel, United States, and Belgium. According to the distribution pattern of AMN and CUBN mutations, the relatively high frequency of I—GS in Scandinavian region is due to founder effects, While in the Mediterranean region consanguinity is the reason for rare mutations in both genes to be homozygous (Tanner et al., 2004). Retrospectively, the previous linkage of Norwegian I-GS families to CUBN (Aminoff et al., 1995) was probably coincidental and essentially false positive, given that the AMN mutations have been identi fi ed in the Norwegian families. Mmcture and functions H Liman AMN (Locus ID: 81693 in Pubmed_locuslink) is located between TRAF3 and CDC4ZBPB, on chromosome Mg 32. The gene is 8.5 kb in length, with 12 exons. The rn RNA (NM_030943, Pubmed) is about 1.8 kb, including a 22 bp 5’UTR and a 512 bp 3,U “IR, encoding a 453 aa protein. The mRNA ofhuman AMN is expressed mainly in small i ntestine, colon, and kidney (Tanner et al., 2003). AMN is predicted to have one transr‘liembrane domain, which connects a ~360 aa extracellular domain and a ~70 aa intrac e1 lular domain. Amino acids from position 205 to 253 in the extracellular region of AMN constitute a cysteine-rich domain, which resembles some bone morphogenic protei 11 (BMP) inhibitors such as chordin (Kalantry et al., 2001; Tanner et al., 2003). The postu 1 ated BMP binding motifofAMN appears to support that AMN may play a role in the m Q Lise development. Since both AMN and cubilin mutations can cause I—GS, it is tempting to think that the ‘W0 proteins may have some interactions. This hypothesis led to the findings that 20 cubilin and AMN colocalize in the apical membrane of kidney proximal tubule cells and copurify during affinity purification (Fyfe et al., 2004). Denaturing and nondenaturing gel filtration chromatography suggested that the two proteins bind to each other with a high affinity- Cell-based studies showed that cubilin and AMN must act in concert to mediate the IF-Cbl endocytosis, while neither protein alone is able to intemalize the ligand. Immuno fluorescence and immunoelectron microscopy demonstrated that cubilin trafficks t0 the cell surface and endosomes in the presence of AMN, but not without. These data indicate that cubilin and AMN fom1 a complex (cu/1am) in the endoplasmic reticulum or GOISI tljat is essential for surface expression and endocytic functions (Figure 1.3) (Fyfe et al., 20 04). Tlae findings also help to solve an intriguing puzzle, that is, why cubilin as a receptc) 1‘ doesn’t have a transmembrane domain or GPI anchor (Moestrup et al., 1998). It was PO Stulated that cubilin might be anchored on the cell membrane by megalin, or less “1(6le cubilin might anchor itself on the membrane by its hydrophobic signal peptide (MoeStrup et al., 1998). Although a helix structure similar to the membrane-insertion moti F Was indeed identified in cubilin by computer analysis (Kristiansen et al., 1999), it is now believed that cubilin is anchored by AMN, given the close relationship of the two protei tjs (Fyfe et al., 2004)- 21 . Vitamin 812 Intrinsic factor (cobalamin) I 1111‘ jib, “1111.4111- Figure 1.3 A model of AMN and cubilin assembly in the biosynthetic pathway and recycling in the endocytic apparatus of polarized epithelial cells. AMN is required for the folding, trafficking and membrane anchoring of cubilin. It may also mediate the internalization of the ligand and recycling of cubilin (from F yfe et al., 2004). 22 Develgamental role of cubam? As mentioned earlier, cubilin is expressed on the epithelial cells of the rat yolk sac. When injected into rats at early pregnancy, anti-cubilin induced a high rate of embryonic resorption and fetal malformations in a dose-dependent manner. This teratogenic effect was deemed to be specific because injection of anti-megalin had no effect (Sahali et al., 1988). Tested with yolk sac visceral epithelial cells, biotin-labeled rat IgG present in the culture medium can be endocytosed into the lysosomes, whereas adding anti-cubilin to the medium resulted in an unspecific pattern of IgG in the cytoplasm, suggesting that cubilin is involved in an endocytotic pathway in the yolk sac epithelial cells (Le Panse et al., 1994). On the other hand, AMN was found to be expressed specifically in the mouse extra-embryonic visceral endodem1 (VE) during embryo gastrulation (Kalantry et al., 2001). AMN disruption in mouse leads to developmental arrest and embryo reabsorption between E95 and E10.5 (Wang et al., 1996). An interesting question is whether the two proteins play their roles independently in embryo development or function coordinately as cubam instead. It is unknown yet whether cubilin and AMN exist in the form of cubam in the rat yolk sac or mouse visceral endoderm. However, immunohistochemistry demonstrated that AMN and cubilin colocalized on the apical cell surface of the VE, while in amn'l' mouse embryo cubilin failed to reach the cell surface (Strope et al., 2004), suggesting that cubam may exist on the VE apical membrane. If this is also the case in the rat yolk sac, the two similar phenotypes may be fundamentally related, that is, disruption of cubam causes the development defect. 23 The second intriguing question is whether the developmental defect is due to a general malnutrition problem or due to dysregulation of development signals. Cubilin can mediate the endocytosis of a wide range of plasma proteins, vitamins and lipids (Moestrup and Verroust, 2001), thus dysfunction of cubam may cause malnutrition to the embryo and subsequent development arrest. On the other hand, cubilin has 27 CUB domains (the name of CUB originated from complement subcomponents Clr/Cls, Uegf, and Bone morphogenic protein-l ), and AMN has a cysteine-rich module similar to some BMP inhibitors. The structural features of cubilin and AMN suggest that cubam may be involved in the BMP pathway to regulate the embryo development. Since both growth of the embryo and assembly of the middle primitive streak were affected in amn'i' mouse, Strope et al. (2004) proposed a unifying model, in which AMN/cubn-mediated endocytosis not only provides necessary nutrients, but also participates in important signaling pathways during embryo development. Although it appears that cubam is crucial for the embryo development of rat and mouse, humans and dogs having mutations in either C UBN or AMN do not show any signs of embryo abnormality. The reason may be that cubam is redundant in humans and dogs for embryo development (Tanner et al., 2004). An alternative explanation is that different organizations of embryo-related tissues, as well as different mechanisms for nutrient absorption may have produced such a phenotypic difference in rodents versus human/dogs (Strope et al., 2004). At this point, none of the three questions have definitive answers. Further investigations are needed to sort out all the possibilities proposed above. 24 Canine I-GS Phenotype Canine I-GS was identified as an autosomal recessive trait in a Giant Schnauzer (GS) family in 19805. At the age of 2-3 months, the affected dogs became inappetent. Hematological tests demonstrated nonregenerative anemia, hypersegmented neutrophils and erythrocyte anisocytosis. Metabolic screening of the urine discovered extremely high methylmalonic acid (>4,000 mg MMA/g creatinine), suggesting vitamin BIZ deficiency. Affected dogs had ~ 10-fold less vitamin BIZ concentration in serum than normal dogs. Fecal excretion test of orally administered [57C0]CN-Cbl suggested that affected dogs did not recycle the Cbl through the enterohcpatic recirculation. Schilling test showed deficiency in intestinal vitamin B12 absorption, and mild proteinuria was observed. Parenteral, but not oral, vitamin BlZ resulted in disappearance of megaloblastic anemia, MMA and other symptoms except for proteinuria. IF and TCII were excluded as the potential causes of the disease. These phenotypes were strikingly similar to the human 1- GS. All the data strongly suggested the canine disease is a homologue of human I-GS (Fyfe et al., 1989; Fyfe et al., 1991a). Comparison of Bl 2 absorption in humans and dogs Dogs have a mechanism for vitamin BlZ absorption similar to humans. It was shown that both IF and IFCR are expressed in dogs (Hooper et al., 1973; Marcoullis et al., 1980). In vivo study in dogs with orally administered [57Co]-Cbl demonstrated that Cbl is bound to IF and R protein in the gastric juice and bound to IFCR in the ileum, 25 indicating that dog also depends on IF-Cbl for BIZ absorption (Marcoullis and Rothenberg, 1981). On the other hand, physiological differences do exist between dogs and humans. In humans, most plasma Cbl is bound to TCI (Gullberg, 1972). In contrast, it is TCII that binds all the plasma Cbl in dogs (Rappazzo and Hall, 1972). Another difference is seen in the biogenesis ofintrinsic factor. In human, IF is produced in gastric mucosa, whereas in dogs, the pancreas is the place where the IF is produced (Simpson et al.. 1993). However, these physiological differences do not invalidate our statement that the dog family described herein is an animal model of the human I-GS. given that the defect concerned is located at the intestine. IFCR study Study of canine I—GS has provided many insights of IFCR (cubilin) and the disease. In immunohistochemical examination of ileal biopsies using a rabbit anti-dog IFCR, cubilin was present on the apical membrane and intracellular spaces of enterocytes in nomial dogs, but was retained exclusively in the cytoplasm of enterocytes in affected dogs (Fyfe et al., 1991a). Detem1ined by the (NH4)3SO4 precipitation method using rat IF-[57Co]Cobalamin, the IFCR activity in the affected dog’s brush border membrane (BBM) was significantly lower in ileum and renal cortex compared to nonnal dogs. Although IFCR was present in the homogenates of both ileum and kidney, it was sensitive to Endo H digestion. Endo H insensitivity usually indicates that the protein has completed the Golgi-mediated processing of glycosylation, while Endo H sensitivity suggests that the protein is still in high mannose state. Thus, canine I-GS is caused by 26 failed BBM expression of IFCR because it did not complete correct folding and glycosylation processing and was detained in an early biosynthetic compartment (Fyfe et al., 1991b; Xu and Fyfe, 2000). These studies provided the first functional evidence that cubilin might be the disease-causing gene, and well explained why both intestine and kidney are affected by the cubilin defect. They also established a solid ground to further explore the fundamental link between human I-GS and canine I-GS. In addition, the canine I-GS has been an invaluable animal model in studying the binding ligands of cubilin, such as apolipoprotein A-I receptor, albumin, transferrin, 25(OH)vitamin D3 and hemoglobin, as mentioned previously in this chapter. Genetic study CUBN has been shown to harbor mutations in human I-GS patients in Finland (Aminoff et al., 1999). Therefore, both biochemical and genetic evidences suggested that CUBN was a compelling candidate gene in canine I-GS. The canine cubilin cDNA was cloned from a renal tubule cDNA library, encoding a 3,620 aa protein. The protein shares more than 70% identity with human and rat cubilin. Similar to the human and rat cubilin, dog cubilin has a furin cleavage site, 27 CUB domains and 8 EGF domains (Xu et al., 1999). However, linkage study with a 17 bp intronic marker in the cubilin gene elegantly excluded C UBN or any gene close to the C UBN locus as the disease-causing gene (Xu et al., 1999). Similarly, another plausible candidate gene, megalin, was also ruled out (Xu, unpublished data). These data provided the first convincing evidence that a gene other than CUBN is associated with the I-GS, which was confirmed by a human study later in a Saudi Arabian family (Al-Alami et al., 2002). With the biochemical evidences, the canine 27 I-GS study also pointed out that the yet to be identified gene product functions as an accessory factor to assist the folding and delivery of cubilin to the cell surface. Summary Vitamin B12 (cobalamin) is an essential micronutrient for humans and other higher animals, and can be obtained only from the diet. Cobalamin (Cbl) has a highly specialized absorption mechanism, which includes the binding of Cbl with R proteins in the stomach, replacement of the R protein with the intrinsic factor in the intestine, recognition and endocytosis of the IF-Cbl complex by IFCR (cubilin) on the ileal enterocytes. With the IF degraded in the lysosomes, C bl binds to TCII and is transported into the portal blood. The Cbl complex is recognized by TC 11 receptors expressed on target cells and endocytosed. Inside the cells, Cbl is processed into MerI and AdoCbl, which are cofactors for methionine synthase in cytoplasm and methylmalonyl-CoA mutase in the mitochondria. A defect in any of these steps may cause functional vitamin BIZ deficiency, but the selective malabsorption of BIZ in intestine due to the IFCR (cubilin) defect is named I-GS. Since cubilin also is expressed on the kidney proximal tubule cells, patients may have proteinuria as well. A canine model of I-GS, created by a natural mutation event, clinically and biochemically resembles the human I-GS. Study of the canine I-GS in the past twenty years has contributed significantly to the understanding of cubilin and the disease. The human I-GS is caused by mutations in either cubilin or AMN. Recent studies demonstrated that cubilin and AMN form a novel complex, called cubam, to endocytose vitamin Bl 2 in the intestine. In our canine model, 28 although biochemical studies have pointed to the defect in cubilin, the CUBN gene was ruled out by genetic study, indicating that another gene may be the disease-causing gene. Purpose and outline When this project started in the year of 2000, the cubilin gene was the only known gene that harbored mutations responsible for I-GS. Based on biochemical and genetic studies, we hypothesized that an accessory factor mutated in the affected dogs is required for normal cubilin folding, exit from the ER, and/or transport to the brush border. Therefore, the goal of this thesis work was to map the disease locus, identify the disease- causing gene. and characterize the mutation(s). With the exclusion of cubilin and megalin genes and lack of other candidate genes, we initiated a whole genome scan linkage analysis, with the hope to positionally clone the gene responsible for canine I-GS. Chapter 2 describes the genetic mapping of the Giant Schnauzer family with I-GS to a region orthologous to human chromosome 14q. Chapter 3 includes cDNA cloning ofthe canine AMN and mutation screening in the AMN gene. Chapter 4 contains the linkage analysis and mutation identification ofa second dog family (Australian Shepherd) with 1- GS, which strengthens our finding in chapter 3. Finally, in chapter 5 we did functional analysis to characterize the mutations identified in the Giant Schnauzer and Australian Shepherd families. 29 CHAPTER 2 GENETIC MAPPING OF CANINE IMERSLUND-GRASBECK SYNDROME IN A GIANT SCHNAUZER FAMILY Most of the studies described in this chapter have been published as (Canine Imerslund— Grasbeck syndrome maps to a region orthologous to HSAI4q) by (He Q, Fyfe JC, Schaffer AA, Kilkenney A, Werner P, Kirkness EF, Henthom PS) in (Mamm Genome. 2003;14(1 l):758-64). Henthom PS and Werner P did the genotyping with microsatellite markers. Kirkness EF provided some of the canine genomic sequences from The Institute for Genomic Research (TIGR). Schaffer AA did the computer-based linkage analysis. Kilkenney A did the genotyping with markers S T N2, EML and GZA. I designed the strategy to locate the minimal candidates region, developed makers EIF 5, KNSZ, SI VA and did the genotyping with these three markers. I also did the haplotype analysis and data interpretation. 30 Introduction Imerslund-Grasbeck syndrome (I-GS) is an autosomal recessive disorder, originally identified in Norway and Finland, that is characterized by selective intestinal cobalamin (vitamin BIZ) malabsorption and proteinuria (Grasbeck et al., 1960; Imerslund, 1960). Clinical features of the disease include megaloblastic dyshematopoiesis, neuropathy, methylmalonic acidemia and aciduria, and hyperhomocysteinemia. Severe cobalamin malabsorption can be life—threatening (Babior, 1975). Periodic parenteral, though not oral cobalamin administration restores normal metabolism and hematopoiesis, suggesting that the cellular BlZ absorption mechanism is intact. Subsequent to the initial reports, I-GS has been reported in many countries all over the world, but highly concentrated in northem African and Middle Eastern families (Al-Alami et al., 2002; Altay et al., 1995; Ben-Bassat et al.. 1969; Barman et al., 1985; Ismail et al., 1997; Mackenzie et al., 1972; Salameh et al., 1991). People have known for a long time that intrinsic factor-Cobalamin (IF-Cbl) complex must bind to an ileal receptor, later called cubilin, before being absorbed. Cubilin is a multiligand receptor expressed in apical membranes of ileal enterocytes and renal proximal tubular epithelial cells. In the small intestine, cubilin mediates absorption ofthe intrinsic factor-cobalamin complex (Bim et al., 1997; Seetharam et al., 1997). In proximal tubules, cubilin physiologically mediates reabsorption of several proteins from glomerular filtrate, such as apoAl, albumin, transferrin, and vitamin D-binding protein (Bim et al., 2000; Kozyraki et al., 1999, 2001; Nykjaer et al., 2001). Thus, one gene product has two distinct functions in two different tissues. The cubilin gene (C UBN) cDNA was cloned and the gene was mapped to a 6-cM region on Chromosome 10p 12.1 31 (Kozyraki et al., 1998). This was intriguing because a previous positional cloning project localized the Finnish I-GS to almost the same region on Chr 10p (Aminoff et al., 1995), suggesting that the cubilin gene was the disease-causing gene for I-GS. Two CUBN mutations were indeed found in Finnish patients (Aminoff et al., 1999), providing an excellent example that combines biochemical studies with genetic studies to locate the mutated gene. Surprisingly, no C UBN mutations were found in the Norwegian or Saudi Arabian patients examined, suggesting that mutations of genes other than C UBN also underlie I-GS (Aminoff et al., 1999; Kristiansen et al., 2000). Locus heterogeneity in 1- GS was confimred when the CUBN region was excluded by linkage analysis in a Saudi Arabian I-GS family (Al-Alami et al.. 2002) and recently with the demonstration of mutations of the amnionless gene (AMN) in I-GS patients from Norwegian and Israeli families (Tanner et al.. 2003). Locus heterogeneity in I—GS had previously been supported by findings in a canine model of I-GS. Canine I-GS is an autosomal recessive disorder characterized by juvenile onset of failure to thrive owing to selective intestinal cobalamin malabsorption with mild proteinuria, resembling the human I-GS (Fyfe et al., 1991a). Immunocytochemical and cell fractionation studies of intestinal mucosa and renal cortex demonstrated that, although cubilin has a normal affinity to IF-Cbl in affected dogs, the receptor did not fold properly and was not trafficked to the apical plasma membrane (Fyfe et al., 1991a, 1991b; Xu and Fyfe, 2000). Cubilin, as well as two cubilin-interacting proteins, megalin (LRPZ) and receptor-associated protein (RAP), were considered as candidate genes for canine I-GS. but each was excluded by linkage analysis (Xu et al., 1999; unpublished data). 32 To detemrine the gene underlying canine I-GS, we undertook a whole-genome scan for genetic linkage in the canine I-GS pedigree. Here, we demonstrate linkage of canine I-GS to markers on canine Chr 8 (CFA8) in a region that is orthologous to human Chr 14q32.2-ter in an interval that contains the AMN gene. 33 Materials and methods Mulls. All dogs were handled according to the principles outlined in the NIH Guide for the Care and Use of Laboratory Animals, with protocols approved by the MSU All University Committee for Animal Use and Care. All dogs were members of a large outbred family in which canine I-GS segregates as a fully penetrant, simple autosomal recessive trait. The I-GS disease phenotype of each dog was detemrined by monitoring puppies for growth and laboratory abnormalities previously described (Fyfe et al., 1991a) until 12ml6 weeks of age and for 3—4 weeks after parenteral cobalamin administration. DNA was available from 128 dogs of known I—GS phenotype. The relationships of these dogs are depicted in the pedigree in Figure 2.1. Blood or tissue samples were stored frozen at ~80°C for DNA isolation. The DNA of F100 was initially unavailable but was obtained later for the genotyping with the KNSZ marker. Markers and genotypiLg DNA was isolated by standard methods (Sambrook et al., 1989). Microsatellite markers for the whole-genome scan, primers, and PCR conditions were those recommended by Richman et al. (2001). Primers and PCR conditions for marker COSS were from Werner et al. (1999). Fluorescently labeled PCR products were electrophoresed on an ABI Prism 377 Sequencer (Applied Biosystems). Allele sizes and genotypes were determined by using GeneScan® and Genotyper® software (Applied Biosystems). 34 Based on previous evidence of conserved gene content and order with the distal long amr of human Chr 14 (HSAI4q), single nucleotide polymorphisms (SNPs) were sought within genes predicted to reside on distal CFA8. SNPs were developed in the canine orthologues of stoninZ (STNZ) (Martina et al., 2001), echinoderrn microtubule associated protein-like l (EML!) (Eudy et al., 1997), G protein-couple receptor 2A (GZA) (Weng et al., 1998), EIF5 (Si et al., 1996), and SIVA (Prasad et al., 1997). Canine gene sequences were obtained from a database of canine genomic sequence maintained by The Institute for Genomic Research (TIGR). Sequence data were originally obtained from plasmid libraries of small (2 kb) and medium-sized (10 kb) genomic DNA inserts prepared and sequenced at C elera Genomics, as described previously for the human genome (Venter et al., 2001). The finished sequence data consist of 6.2 million reads (average read length, 576 bases), representing approximately 1.2x coverage of the haploid canine genome whose size is estimated at 2.8 Gb (Vinogradov, 1998). To identify dog sequences for genes of positional interest, we used a three-step procedure. First, human genomic sequences were masked for repetitive elements and searched against the assembled canine genomic sequences by using BLASTN (http://www.ncbi.nlm.r1ih.gov/BLAST/). Second, dog sequences with the highest BLASTN scores were searched back against the current human genome sequence. Third, the matches were evaluated, and dog sequences that were most similar to the human gene originally used for searching were considered to be fragments of a putative orthologue. This procedure can effectively avoid unspecific assignment due to partial homology. Primers for PCR were chosen from exons that flank introns of moderate predicted size. 35 SNPs were identified by sequencing PCR products amplified from an affected (Hilary i.e. F284) and a carrier (Shorty i.e. F274) dog. STNZ primers were (JCFZ36) 5’-AGGTGCAGAGCTGGCTTAGGATGT-3‘ and (J CF249) 5’-GAGTTGAAGGCATGCTCGTACTTG-3’. The STNZ SNP altered a leI restriction enzyme recognition site. PC R products were digested with leI (New England Biolab), and DNA fragments were separated on 3% agarose gels. EML] primers were (JCF285) 5’-CCTGTAAGCAA GTCGTAAGTGTGG-3’ and (JCF286) 5’-GTCTGGCACAA CCTCCTATG-3’, and the SNP was detected by Alul (New England Biolab) digestion. 02A primers were (JCF230) 5’-GCCGTCTACCTCTTCTGCCTGTC-3’ and (JCF254) 5’-GCTAGGAAGC GGTC AC AGGAGAT-3’, and the SNP was detected by Tat] (Femrentas Life Science) digestion. KNSZ primers were (JCF294) 5’- GAG CCT CTG GAT GAC CTT TTC-3’ and (JCF295) 5’- CAC TGC TAT GCT GCT GTT GGA CT-3’. PCR reactions (50141): genomic DNA 200ng, 5ul of each primers (2.5pmol/ul), 5ul 10xbuffer, 5ul dNTP (2.5pmol/ul each), 0.5u1 Taq polymerase (5U/ul) and 25.5ul H20. PCR cycles: 94°C 3min; 94°C 30’, 58°C 30’, 72°C 1min for 35cycles; 72°C 10min. 36 For genotyping, the KNSZ PCR products were run 011 the 1% agarose gel, then purified by QIAEX'RII Gel Extraction Kit (Qiagen, Cat. No. 20021) and sequenced at the University of Michigan DNA Sequencing Core. EIF5 primers were (JCF298) 5’- GGC GCC ATT TCC TAC GAG-3’ and (JCF299) 5’- CTC TGC TTC CTT CAA CCA TTT TAT-3’. The EIF5 PCR products were gel purified and sequenced. SlVA primers were (JCF31 1) 5’-GCT GTG CCA TTG TTG ACC TGC C-3’ and (JCF3IZ) 5‘-GCT CTG GTC ACT GTC CCG GAG-3’. The SNP was detected by Sea] (New England Biolab) digestion or by direct sequencing. Linkage analysis Linkage analysis computations were performed with the FASTLINK software package (Cottingham et al., 1993; Lathrop et al.. 1984; Schaffer et al., 1994). Loop breakers were chosen by the method of Becker et al. (1998), as implemented in FASTLINK 4.1P. This method is particularly helpful for inbred pedigrees with numerous multiple matings, because it allows a multiply mated loop breaker to break more than one loop. One hundred twenty-eight dogs of the I-GS pedigree were genotyped at one or more CFA8 markers. Omission of dogs from particular marker data sets was because either 1) PCR failed for certain dog/marker pairs, 2) the genotype could be inferred from first-degree relatives, or 3) it became evident that the markers on the proximal portion of CFA8 were not close to the disease gene. LOD scores were calculated with assumptions of a disease allele frequency of 0.001, full penetrance of the disease trait, and equal marker allele frequencies. The order of microsatellite markers was based 011 canine linkage maps (Mellersh et al., 2000; Werner et al., 1999), and the order of the three genes, STNZ—EMLI-GZA, was based on their order in the orthologous portion of HSA14. The marker order used here is consistent with the most recently published canine radiation hybrid map (Guyon et al., 2003). Multipoint analysis indicates that the STN2 gene is most likely between FHZI44 and FHZI38 011 CFA8 but might also be between FHZI38 and C08.618. Attempts to use multipoint linkage analysis to place the C088 marker with respect to EML] and GJA yielded very flat LOD score curves, indicating that C088 is close to both EML and 02.4. To confirm that the LOD scores we report are qualitatively robust, we repeated the analyses by using a disease allele frequency of 0.01 , penetrance of 0.99, and marker allele frequencies estimated from the data. The latter calculations resulted in nearly identical positive LOD scores. Results A \V’ hole-genome scan for marker linkage was initiated in the canine I-GS pedigree W111) DN A from 88 dogs of known I-GS phenotype. The scan was abbreviated when a LOD SQO re >5 was obtained for marker C08.618 examined early in the process. While the high LOD score gave confidence that the I-GS phenotype was linked to CFA8, the peak LOD score was achieved ~10 centiMorgan away from the disease locus. Therefore, we PIOCQeded to genotype other rnicrosatellites from CFA8 and included additional dogs. We also identified polymorphisms (SNPs) in genes predicted from humanwdog c0111 parative mapping to be in the region of interest (Breen et al., 2001). Reciprocal chromosome painting studies demonstrated previously that the entirety of CFA8 is horljologous to the entire long arm of HSA14 (Breen et al., 1999; Sargan et al., 2000). Hi glrer resolution conservation of gene order between CFA8 and HSAl4q was derrnonstrated by comparison of loci placed on the most recent canine integrated RH /genetic linkage map (Breen et al., 2001) and Build 30 ofthe human genome sequence as displayed on the UCSC Genome Bioinformatics website (1111p://genome.cse.ucsc.edu/index.html ). One or more SNP markers were genotyped on 11103 t ofthe dogs depicted in Figure 2.1. ”Two-point linkage analysis was performed between the I-GS locus and 9 CFA8 markers, including six rnicrosatellites (FH2149, C08.410, FH2144, FHZI38. C08.618, COS-8) and three SNPs (in genes STNZ, EML], and 02.4). This type of analysis conSidcred the relationship between the disease locus and one of the 9 markers at each ““71 e. The results are shown in Table 1. Partial analysis ofthose genotypes indicated that the I~GS locus should be on the distal side of C086] 8, where only one other polymorphic 39 Figure 2.1 Pedigree of dogs of known l-GS status used for linkage analysis. All dogs were derived from two affected purebred Giant Schnauzers (F70 and F100) and three unrelated normal dogs of other breeds (A66, A323, and MS74). Dogs F70 and F100 were both inbred (inbreeding coefficients of 0.133 and 0.25 respectively) and related (the sire of F100 is a great grarrdsire of F70). Squares indicate males. and circles indicate females. Filled symbols indicate I-GS affected dogs, and open symbols indicate clinically normal dogs. Vertical lines descending from the symbol for an animal indicate one or more matings, offspring of which are arranged 011 horizontal lines connecting the parents. Numbers within symbols indicate the number of dogs of the indicated gender and phenotype that were available for analysis when there was more than one from the depicted mating. 40 A 323 M 874 Figu re 2.1 41 Table 1. Two-Point LOD score analysis. Two-point LOD score (Z) values at various recombination fractions (6) with respect to the canine I-GS locus. lMarkér 1 FH2149 . C08.410 FHZI44 w STA/2 ' FHZI38 C08.618 EML] COSS G291 ” 3.31 .0.0 “30.494193791433 41 1.20 —8.00 344.66 0.34 ;20.29 —12.’02 {—8.36 l ' gr 7.63 9—2.95 V 0.98" ' 3.65 7.60‘ 3.84 10.02 1.80 "3.73 '—9.44 —5.86 3.86 N 4.99 6.00 6.87 8.56 8.79 14.07 4.11 10.511052 96.07 1—3.65 ' 4.83 5.59 ' 7.29 8.79 4.08 f 1 0.3 8 42 ——3.79 {—1.51 1.05 " —1.510.55 5.75 6.32 6.03 6.15 7.52 7.38 8.58 7.97 13.97 3.68 10.03 924' 10.55 0.041 2.26 " f 6.36 6.17 7.54 8.81 4.11" 0.07 1.0.03.0.” ”0.07” 0.10 "0.15 Pea/t 200 e 35...]. 1 0.416 "0.169 ' ' 0.135” 0.110 0.344 0.293 n ‘ ' ' 0.059 0.048 marker, C088, was initially available (Mellersh et al., 2000). The COS8 marker is near both EML] and GZA, but attempts to more precisely detemrine its position were unsuccessful owing to limited infomrativeness of this marker in the I-GS pedigree and lack of obligate recombinants. We provisionally placed COS8 between EML] and GZA because this appeared marginally more likely than either extreme placement. We obtained peak LOD scores for the three most distal microsatellite markers (FHZI38, C08.618, and COS8) and the SNPs in STNZ, EML], and 02A that were well above the standard threshold of 3.3 used to declare significant linkage in whole-genome scans (Lander and Kruglyak, 1995), with the score for 62A exceeding 10. Peak LOD scores for the six linked markers varied significantly owing to varying informativeness in this pedigree. Therefore, the recombination fraction where the peak LOD score is achieved for each marker, and not the value of that peak, is the most important criterion to consider. By that criterion, EML]. C088, and GZA are the markers closest to the gene, since their peak LOD scores occur at recombination fractions of ~0.05 or less. The LOD score with COSS at a recombination fraction of0 is positive (and not — 00), indicating that there are no obligate recombinants between the disease and COS8 within the pedigree. However. the LOD score with COS8 does not peak at a recombination fraction of 0 because there are two distinct COS8 alleles closely linked to different haplotypes that entered the pedigree separately through dogs F 70 and F100. Using the ILINK program in FASTLINK, we estimated that the recombination fraction between EML] and GZA was ~0.075. Using that estimate and the LINKMAP program, we generated the multipoint LOD score values plotted in Figure 2.2. The peak LOD score of 11.74 occurs with the I-GS locus placed slightly nearer to GZA than 43 midway between the two markers. The maximum LOD score with the disease gene placed proximal to EML] was only 9.8, making it much less likely that the gene is proximal to EML]. The maximum LOD score generated with the I-GS locus placed distal to GZA was 11.07, obtained when the putative disease gene was placed approximately 4 CM from 02.4. The 0.6 lower LOD score suggested that this location for the I-GS gene is less likely than the location proximal to 62/1. The disease gene location between EML] and 02.4 was further supported by examination of haplotypes, as illustrated in Figure 2.3. Recombination events and disease status in the dogs shown are most parsirnoniously explained by the gene order ErilL/—l—GS—G2A. In order to further narrow down the disease region, we developed 3 more markers (KNSZ, EIF5 and SIVA) between EML] and (22.4 using an iterative strategy (shown in Figure 2.4). Compared to randomly developing new markers, our strategy is more efficient. We subjected the 3 markers to genotyping (Table 4 in Appendices). Haplotype analysis showed that the minimal candidate region was between EML] and SIVA (Figure 2.3). 44 12‘ LOD Score ‘° 1 9 -« 8 -l _ -0.10 -0.05 0.0 0.05 0.10 0.15 EML1 GZA Figure 2.2 Multipoint LOD score curve showing linkage to I-GS. The X-axis shows distance in centiMorgans with respect to the EML1 marker. 45 274 STNZ C C C T EML1 G A EIFS A A , A T KNSZ T T C SIVA C C C T GZA A A A _ G 381 384 W STNZ C T C C C T EIFS A A A A A A [-65 KNSZ T T T T T T T . S'VA . c.. .. . (I. . . . .. .. . . C. . . 1r. . .. .. . .. .. . ._ ,. . C. . c. «a ._ .. .... .1 ._ . . . ._ . aw...“ . .. . . ..2 . . . .. . . .. . .. . ..._ . -... -... GZA A A A G A A Figure 2.3 Haplotype analysis of the GS kindred. The white bar represents the normal allele, while the black bar represents the disease allele. The minimal candidate region for canine I-GS is between EML] and SIVA. 46 If homozygous Develop MCR forallaffected _5 flanking —> ‘ » dogs markers RCIIIIC IIIC Dcvclo _> p “1313131118 ’ new markers \ If heterozygous for some a ffectcd dogs I Figure 2.4 An iterative strategy to develop new markers for linkage disequilibrium mapping. MCR represents minimal candidate region. In order to further minimize the candidate interval already delimited by EML1 and (72.4, we first transformed the genetic distances of EML] and 02A versus l-GS into physical distances to locate the I~GS. At the [—05 position, we would develop a new marker. 1. If the new marker were homozygous in all the affected dogs and heterozygous for all the carriers, we would develop more markers flanking this marker, until we found the nearest heterozygous marker in affected dogs. This minimal homozygous interval would be assigned as the minimal candidate region (MCR). ll. Ifthe new marker were heterozygous for some affected dogs, we would calculate the recombination score or do haplotype analysis to assign the disease gene into one of the smaller intervals divided by the marker. Then, in this smaller interval, we would repeat the process of developing new markers until we defined the minimal candidate region. 47 Discussion Results of this investigation indicate that the genetic locus of canine I-GS is on distal CFA8, a canine chromosome that is orthologous to HSAI4q. This finding was consistent with our previous report that canine I-GS was not linked to CUBN (Xu et al., 1999) and that canine CUBN is located on canine Chr 2 (Breen et al., 2001). We obtained compelling LOD scores of >10 for the CFA8 location by using a large canine pedigree specifically bred for investigation of I-GS. Based 011 the multipoint LOD scores, the disease gene locus is distal to a marker in EML], and examination of haplotypes suggests that it is proximal to a marker in SIVA. Within the interval between EML] and SIVA, the KNS2 marker was in complete linkage disequilibrium with the disease allele in 20 affected dogs, 1 l obligate heterozygous carriers, and l unrelated nomral dog used for an outcross mating in the l-GS linkage family. The multipoint LOD for canine l-GS, KNSZ, and 02A, peaked at 15.4 with 0 at 0.0 between [-05 and KNSZ, indicating that the disease-causing gene is very close to KNSZ. The estimated 0 between KNSZ and 02/1 is 0.05. No recombinants were found for marker EIF5. However, it‘s not possible to accurately deduce the genetic distance between EIF and l—GS because the number of animals genotyped with EIF5 is too small. All genes currently mapped to CFA8, including those reported here, are in the same order as on HSAl4q (Breen et al., 2001; Guyon et al., 2003). The AMN gene, reported to be mutated in some I-GS patients while this article was in review (Tanner et al., 2003), resides in the 5—Mb interval between EML] and SIVA in the human genome (Build 30. UCSC Genome Browser http:/./genome.ucsc.edu/). 48 Prior to this discovery, there had been no previous suggestion that a gene on HSA14q had a role in cobalamin metabolism. Thus, it seems highly probable that the 1- GS dogs also have a mutation in the AMN gene. One would expect a single homozygous mutation in the affected dogs. Indeed, we hope to use the canine model to clarify the tissues and stages of development at which it is expressed. The eventual determination of the gene and mutation causing canine I-GS will yield new biological insights, whether the disease gene is AMN or not. AMN was discovered as the gene disrupted by an insertional mutagenesis event in mouse that caused a recessive prenatal lethal phenotype (Kalantry et al., 2001; Tomihara-Newberger et al., 1998; Wang et al., 1996). Affected mice die at around day 10 of gestation, having failed to develop a portion of the primitive streak during gastrulation. The AMN gene encodes a predicted type I transmembrane protein, with an N-temrinal signal peptide and a single transmembrane domain that is found on the apical surface of visceral endoderm cells early in development (Kalantry et al., 2001). At this time, essentially nothing is known about the proteins or pathways with which the AMN gene product interacts during early embryonic development. The involvement of the AMN gene in both early mouse development and in the selective malabsorption of cobalamin in otherwise normal humans was unexpected. A proposed explanation for this phenomenon is that the function of AMN in human and rnurine embryonic development may differ significantly. Additional work is needed to test this and other hypotheses concerning functions of the AMN gene. Examination of the AMN gene in canine I-GS affected dogs combined with the worldwide effort to annotate genes in all species will allow the detemrination ofthe gene 49 involved in canine selective cobalamin malabsorption with proteinuria. Immunocytochemical and cell fractionation studies of canine intestinal mucosa and renal cortex have already demonstrated that, while cubilin is expressed in the appropriate tissues of I-GS affected dogs, the receptor does not fold properly and is not trafficked to the apical membrane (Fyfe et al., 19913, 1991b; Xu and F yfe, 2000). The involvement of AMN in this pathology, whether or not it is defective in the affected dogs, can now be examined as the canine I-GS model continues to provide a valuable resource for the development of our understanding of cobalamin metabolism. 50 CHAPTER 3 cDNA CLONING OF CANINE AMN AND MUTATION SCREENING IN THE GIANT SCHNAUZER FAMILY 1 designed and performed all the experiments described in this chapter except that Gregory B provided some technical assistance in the genotyping of the unrelated normal dOgs 51 Introduction In the previous chapter, we mapped the canine l-GS to an interval between marker EML] and SIVA, which is syrrtenic in human, mouse 311d dog. Referring to the human genome map (http://genome.uses.edu/; build 31), the interval is about 5 Mb in length, containing at least 40 genes (Table 2). KNSZ. a single nucleotide polymorphism (SNP) marker in the 5 Mb interval, is in complete linkage disequilibrium with the disease allele in our GS kindred. The multipoint LOD score for KNSZ is 15.4, indicating that the disease-causing gene is very close to this marker. Genes ~ 1 Mb on either side ofKNSZ in the human and mouse genome databases were considered comparative positional- candidate genes 311d were further triaged by examining EST databases for genes expressed in the known cubilin expressing tissues. AMN stood out in this analysis with a vast preponderance of ESTs from kidney 311d lesser representation in intestine and colon (UniGcne Cluster Hs.236720. since retired 311d replaced by Hs.534494). During the course of this work, three different AMN mutations were demonstrated in human I-GS kindreds from Norway and Middle East (Tanner et al., 2003). Combined with the study of human patients, our data highly sug Iest that AMN is also the disease-causing gene for canine l-GS. AMN gene was originally identified as 311 essential gene for mouse gastrulation. The gene encodes a transmembrane protein exclusively expressed on the visceral- endodenn of mouse embryo. AMN has an extracellular cysteine-rich domain, which resembles several bone morphogenic protein (BMP) inhibitors (Kalantry et al., 2001). Since the (117111"; mouse was embryonic lethal, it is not known whether AMN defect also causes BIZ malabsorption and proteinuria in the mouse. Human (NCBI Nucleotide 52 accession no. NM_030943), mouse (accession no. BC087954) and rat (accession no. XM__234547) cAMN have been cloned, each with 1362 bp, 1377 bp, 1377 bp in length, respectively. The GC content of the three cAMN is between 65%~ 71%. Protein alignment of the three gene products suggests that the N-terminal part is evolutionarily more conserved than the C-temrinal part. At this point, no sequence information was available for dog AMN. In order to screen the canine AMN gene for potential mutations, we first cloned the cDNA of canine AMN from a cDNA library constructed from dog kidney proximal tubule cells. RT-PCR and genomic PCR in both affected dogs and normal dogs were subsequently perfonned to locate the mutation in the AMN gene. 53 Table 2. Genes listed in the EML1-$1144 interval of human genome browser. The Human EST hits data were retrieved from the NCBI__Unigene_homo sapiens database (http://www.ncbi.n|m.nih.gov/entrez/query.fcgi?db=unigene). Genes that have higher expression in kidney and colon are highlighted. Locus Name Positioanuman EST hits Funcfion SIVA 99.21 kidney 2; colon 3 CDZ7 binding protein1 FLJ38602 99.18 FLJ22056 99.16kidney 3; colon 1 MGC13251 99.14kidney O; colon 0 hypothetic C14orf2 98.191kidney 5; colon 0 6.8 kDa mitochondrialproteolipid PPP1R138 98.01 kidney 2; colon 2 protein phosphatase regulatory subunit MGCZ550 97.99 kidney 4; colon O XRCC3 97.98 kidney 0; colon 0 Xray repair cross complementing protein 3 KNS2 97.91 kidney 0; colon 1 kinesin 2 MGC2562 97.84 kidney 0; colon 1 BAGS 97.83 kidney 2; colon 1 Bcl2 associated athanogene 5 FLJ40452 97.81 CKB 97.8 kidney 4; colon 20 creatine kinase. brain MARK3 97.66 kidney 3; colon 1 MAP/microtubule affinity regulatingkinaseB EIF5 97.61 Kidney 12; colon 1 Eukfl/otic translation initigtion factor 5 TNFAIP2 97.41 kidney 2 ; colon 0 protein 2 induced by TNFalpha CDC423PB 97.21 kidney4 ; colon 0 CDC42 binding protein kinase beta Kidney 106; colon AMN 97.2 7 amnionless protein TRAF3 97.06 kidney 2; colon 1 TNF receptor-associated factor 3 RCOR 96.87 kidney 1; colon 0 REST-corepressor M6021 990 96.79 K/AA0071 KIAA0329 96.64 kidney 5; colon 3 HeLa eyclin dependent kinase 2-interacting CINP 96.63 kidney1; colon 1 protein FLJ11132 96.61 kidney 6; colon 1 hypothetic renal tumor antigen mapkinase super RAGE 96.51 kidney 2; colon 1 family WDR20 96.42 WD repeat protein Kidney 49; colon HSPCA 96.36 35 heat shock 90 kDa protein 1, alpha PPP2R5C 96.09 protein phosphatase 2, regulatory subunitB DHC1 kidney 15; colon 6 dynein heagchain cytosolic DIO3 95.84 thyroxine deiodinase 3 LOC64150 95.83 54 Table 2 (cont’d) DLK1 95.01 delta-like homologue K/AA1446 94.81 MGC4645 94.66 WARS 94.61 Tgptophanyl tRNA synthetase FLJ38975 94.57 LOC145604 similar to adaptor-related protein complex 1 Novel ?mitochondrial uncoupling protein novel ?solute carrier family 25? Ensemb|140109 YY1 94.52 YY1 transcription factor FLJ32960 unknown, alternatively spliced novel ?degenerative spermatocyte? RN86 94.34 kidney 0; colon 0 ?actin filament organization neurons? EML1 94.07 kidney 3; colon 1 echinoderm microtubule associated protein 55 Materials and methods Obtaining partial secLuence of canine CAMN Human, mouse, rat AMN cDNA sequences were aligned by the MultAlin software (http://prodes.toulouse.inra.fr/multalin/multalin.html). Three sets of primers were designed based 011 the homologous sequences ofthe three species. Set I: JCF321 5’-TSC TGC TGT GGC TGC AGC TCT G-3’ JCF322 5’-CCT TGT CCG CCG GGA ACT GRA C-3’ Sci]: JCF323 5’-TCT TCT YCG TGG ACG CCG AGC G-3’ JCF325 5’-GGT CCT CGT CGC GCG WGA ACG T-3‘ Sci 3: JCF324 5"-ACG TTC WCG CGC GAC GAG GAC C-3’ JCF3Z6 5’-CGG TAC CGC TCC AGG TCA ATT G-3’ Total RNA was extracted from an unaffected dog by TrizolR Reagent (Invitrogen). 511g RNA was used to perform the RT reactions (SuperScriptTM First-Strand Synthesis System for RT-PCR, Invitrogen), with Oligo(dT) 12-18mer (Invitrogen). RT-PC R was performed with the three sets of primers shown above. respectively. PC R conditions: RT product 2 ul Primerl 5 111 (2.5 pmol/ul) Primer2 5 111 (2.5 pmol/ul) 10xPCR buffer without Mgzi 5 ul 25mM MgCl; 6 ul DMSO 2.5 ul dNTP 5 111 (2.5 pmol/ul each) AmplyTaq GoldTM(Roche) 0.5 ul H30 19 Ill Total 50 111 PCR cvclcs: 95°C 10 min; 95°C 15’. 62°C 15’, 72°C 1 min for 35cycles; 72°C 5 min. The PCR product was gel purified and sequenced. 56 cDNA cloning of canine AMN The primers JCF323 and JCF325 were used for the cDNA cloning of canine AMN from a ZAPIITM phage cDNA library (Stratagene) constructed from a pool of dog kidney proximal tubule cell RNA. The PCR with JCF323 311d JCF325 was referred to “indicator PCR" in the following steps. The cDNA cloning method was based on a pool-PCR strategy, which was designed to quickly and efficiently clone the target cDNA (Israel, 1993). Preparing host cells: A single colony of XLl—Blue MRF cultured 011 a plate (Tetracycline 12.5 ug/ml) was inoculated into 20 ml LB plus 0.2 1111 20% maltose without antibiotics. After culturing at 33°C overnight, the bacteria were harvested by centrifugation at 3,200 rpm for 8 min. The pellet was resuspended in 20 ml 10 mM MgSO4. 7 1111 of the bacteria was transferred to a new tube and centrifuged briefly, with 6.9 ml of the supematant removed. The pellet was suspended by the residual 0.1 ml MgSO4. Infection: The cDNA library was titered as described in the Molecular Cloning: A Laboratory Manual, 211d ed. (Sambrook et al., 1989; section 2.60—2.61). 5ul ofthe cDNA library (3x107 phages) was diluted into 95 ul SMG buffer (50 111M Tris-Cl, 100 mM NaCl, 10 mM MgSO4, 0.01% gelatin), and then mixed with the host cells. The mix was incubated at 37°C for 20 min. The mix was diluted into 20 ml LB containing 10 mM MgSO4, and distributed into 64 wells (100111/wellx 64wells). Thus only 1x107 phages 57 were actually screened. The wells were sealed by GeNunc Tape (NUNC #BC2689) 311d incubated at 37°C for 18 hrs with shaking (215 rpm). Pooling. lysis and PCR: 15 ul of phage culture from each well of the 8x8 matrix was pipetted. 120 111 ofphage culture from each row/column was pooled (15 111 x 8:120111), resulting in 16 pools. 50 11] culture was taken from each pool. and then subjected to 50 111 100 mM NaOH at 95°C for 10 min, respectively. 20 111 of l M Tris-Cl (pH7.4) was used to neutralize each reaction. Finally. 2 ul ofthe solution in each tube was used as template in the indicator PCR. Anaivsis and rcscrcwzing: The PC R products were r1111 011 % agarose gel. A single well that contained the AMN clone was identified by the synthesis of a PCR product at the right size. The positive well was titered and diluted. About 1920 phages were used to infect a new 8x8 matrix of host cells, with each well containing 30 phages. The matrix was rescreened by the protocol shown above. A positive well was identified and subjected to the tertiary screening, with each well containing 2 phages before the culture. As a positive well was located, the phages in that well were plated 311d screened by the indicator PCR one by one. Finally, The target cDNA was PCR amplified with primers T3 311d T7 from the positive phage clone or rescued by helper phage into pBluescript plasmid. The PCR product or plasmid was sequenced. FUll length RT-PC R of canine cA MN Primers: JCF334 5‘-CGG GCG CGC GGC GGG ATG-3’ 58 JCF332 5’-CTG GCC AGC CCC GCG GTT GC-3’ Total RNA were extracted from 311 affected dog (F284) and an unrelated normal dog (DCCU61 10) by Trizol'fi: Reagent. RT reactions were performed with SuperScriptTM First- Strand Synthesis System. The Advantage-GCZ PCR kit (BD BioScienccs) was used for the PCR. PC R conditions: RT product 2 ul JCF334 8 111 (2.5 pmol/ul) JCF332 8 111 (2.5 pmol/ul) 5xPCR btrffer 10 ul GC-Melt 5 ul 50deTP l 111 GCZ-polymerase 1 ul H30 15 111 Total 50 ul PCR cycles: 94°C 3 min; 94°C 30’, 68°C 3 min for 35cycles; 68°C 3 min. PC R machine: PTC-100TM (MJ Research). 5’end RT-PCR for canine Alli/V Primers: JCF379 5’-CGG GCG CGC GGC GGG-3’ is located in 5’UTR. JCF339 5’- CGC TCG GCG TCC ACG GAG AAG A-3’ is located in exon 5. The PCR conditions are the same as those for JCF334+JCF332. Genomic PCR extraction from blood DNeasyiR‘; Tissue Kit (Qiagen) or QIAampCRDNA Blood Midi Kit (Qiagen) was used to extract blood DNA. 59 Genomic PCR for mutation detection Primers: JCF370 5’-TCC GTT GCA GGC GAA GCC CTC-3’ crosses intron 9 and exon 10, while JCF366 5’-CTG CGG GGT GCG TGG AAC CTA G-3’ crosses intron 11 and exon 12. The Thennal Ace"M DNA Polymerase Kit (Invitrogcn) was used for the PCR. PCR conditions: Genomic DNA 100-200 ng JCF370 ‘5 111 (2.5 pmol/ul) JCF366 5 111 (2.5 pmol/ul) 10xThermal Ace PCR buffer 5 ul 50x dNTP 1 ul Themral Ace polymerase 1 ul H30 11p to 50 111 Total 50 ul PCR cycles: 98°C 3 min; 98°C 30’, 65°C 30’. 72°C 1 min for 35cycles; 72°C 10 min. PCR machine: FTC-100TM (MJ Research). The PC R products were sequenced or run 011 the 2% agarose gel. 60 Results Cloning of canine AMN cDNA To obtain a fragment of canine AMN cDNA, we designed 3 sets of primers based on the homology of human, mouse and rat AMN cDNA sequences. Only the set ofJCF323 plus JCF325 gave a band at expected size in the RT-PCR. DNA sequencing confirmed that the PCR product was homologous to its counterparts in other species. We used this PCR as the indicator PCR to screen the canine cDNA library by a pool-PCR method. At the last round of screening, three colonies were picked from the plate 311d PCR amplified by T3 311d T7 primers. with DMSO added to the reactions. Each of them gave a band at 1.6 kb. Sequencing one of the PCR products demonstrated high homology to AMN in other species, indicating it is the cDNA sequence ofcanine AMN. The full-length canine AMN cDNA (deposited to GenBank accession no. AY368152) is 77% GC 311d comprised of a '15 bp 5’ untranslated region (UTR), a 1374 bp open reading frame, and a 152 bp 3’ UTR that includes a polyadenylation signal 11 bp 5’ of the poly A tract (Figure 3.1 ). The deduced amino acid sequence of 458 residues is 73 %. 66 %. 65%, 38%, 34%, 33%, and 21% identical to AMN of human (NP112205), mouse (NPZ91081), rat (XP234547). chicken (XP421397), Xenopus lavis (AAH74152), pufferfish (Fugu chrUn:215.972,629—Zl5,974,032; UCSC Genome Browser, Aug 2002 assembly) and fruittly (NP608515), respectively. Structural features conserved between these species include the predicted signal sequence cleavage site (after Ala 19); 12 cysteine residues in the mature extracellular domain, 9 of which are clustered between residues 205-253; a similarly placed, single predicted transmembrane domain (residues 363-387 of dog AMN); and 2 copies of F/YXNPXF/Y, a well-characterized AP-Z adaptor 61 protein-binding signal for ligand-independent receptor internalization via clathrin-coated pits (Boll et al., 2002) (Figure 3.2). Potential N-glycosylation sites (dog AMN residues 35 and 39) were found within the first 40 amino acid residues ofthe protein in all but the chicken and fruitfiy sequences. Mutation screening of canine AMN RT-PCR is a rapid method for mutation screening. although it may not be able to detect certain types of mutations, such as mutations in introns, 5’UTR or 3’UTR etc. AMN cDNA of the entire coding region was amplified by RT-PCR from kidney cortex RNA isolated from an I-GS affected dog of the Giant Schnauzer (GS) kindred. Sequencing the product revealed arr in-frame, 33 bp deletion in exon 10 (c.lll3_l l45del), predicting loss of 11 amino acid residues from the transmembrane domain somewhere between residues 370-382 (Figure 3.3). The deletion endpoints could not be determined exactly because they occurred within each of2 nearly perfect copies of 3 24 bp direct repeat sequence that are 79% GC 311d separated by 9 bp. However, 3 sequence variation of the repeats allowed us to place the 5' deletion endpoint on or between nucleotides 1 106-1 1 l3 ofthe cDNA sequence 311d the 3‘ deletion endpoint on or between nucleotides 1 139-1 145. The same 33 bp deletion in exon 10 was found in PCR products amplified from genomic DNA of affected dogs, and can be detected by simply running the PCR products 011 the 2% agarose gel (Figure 3.4). The deletion was homozygous in 18 affected dogs, heterozygous in 8 obligate carriers, and was not seen in 224 chromosomes of unrelated normal dogs of various breeds. It must be pointed out that the full-length RT-PCR did not cover the start codon. because primer JCF334 overlapped the ATG site. We thus did the 5’ end RT-PCR, which by sequencing confirmed that no mutation exists in the start codon (data not shown). 63 Figure 3.1 cDNA sequence of canine AMN (deposited to Genebank; Accession number AY368152 ). The ATG start codon is underlined. The shaded area represents 5’ and 3’ untranslated region. The polyA signal is underlined near the end ofthe sequence. 51 101 151 201 251 301 351 401 451 501 551 601 651 701 751 801 851 901 951 1001 1051 1101 1151 1201 1251 1301 1351 1401 1451 1501 cgggcgcgcggcggggggggcgcgctgggccgggccctgctgtggctgca gctgtgcgcgctggcccgggccgcctacaagctctgggtccccaccacgg acttcgaggccgccgccaactggagccagaaccggacgccgtgcgcgggc gccgtggtccagttccccgcggacaaggcggtgtcggtggtggtgcgggc cagccacggcttctcggacatgctcctgccgcgggacggggagttcgtcc tggcctcgggagccggcttcggggccgcggacgccggcagggacccggac tgcggcgcaggcgcccccgcgctcttcctcgaccccgaccgcttctcgtg gcacgacccgcgcctgtggcgctccggggacgcggcgcgcggcctcttct ccgtggacgccgagcgcgtgccctgccgccacgacgacgtcgtcttcccg cccgacgcctccttccgagtggggctcgggcccggcgcccgccccgcgcg cgtccgcagcgtccaggttctgggccagacgttcacgcgcgacgaggacc tggctgccttcctggcgtcccgcgccggccgcctgcgcttccacgggccg ggcgctctgcgcgtgggccccggggcctgcgccgacccgtcgggctgcgt ctgcggcgacgcggaggtgcagccctggatctgcgcggccctgctccagc ccctgggcggtcgctgcccgccggccgcctgccccgacgccctccggccc gaggggcagtgctgcgacctctgcggagccatcgtgtcgctgacccacgg ccccacctttgacatcgagcggtaccgggcgcggctgctgcgagccttcc tgccccagtacccggggctgcaggcggccgtgtccaaggtgcggcggcgg ccggggccgcacacggaggttcaggtggtgctggcggagaccgggcccca gccgggcggcgcggggcggctggcccgggccctcctggcggacgtcgcgg agcacggcgaagccctcggggtcctgtcggcgacagcccgggagtcgggc gcgcccgtcggggacggctcggcggcggggccgctcggctcgggttcgcg cgcggggctggcgggcggcgtggcggccgggctgctgctgctgctgctgg cgctggcggcgggcctgctgctgctgcgccgcgctccgaggctcaggtgg actaagcgcgagcgattggtcgccacgcccgtcgaggcgcccctgggctt ctccaacccggtgttcgacgtggcgggctccgtggggccggttccacgca ccccgcagcctcccccagcgcagcaggcgggaagcagcagcaccagccgc agctacttcgttaacccgctgttcgccgaggccgaggcctgagcaaccgc ggggctggccagcccctacctgcgcccgccgccgcccccgcgagatggcc ccggccttgcgaggtccccgccccctgccacgcacgccttgtccccccag cccaaggatagggtggctttgcccaataaagcgtttcctgc 64 hum dog mouse Consensus hum (log mouse Consensus hum dog mouse Consensus hunt dog mouse Consensus hum dog mouse Consensus hum dog mouse Consensus hum dog mouse Consensus hum dog mouse Consensus 1 10 20 30 40 50 60 I e e e v e I HGVLGRVLLHLOLCRLTORVSKLUVPNTDFDVRRNHSONRTPCRGGRVEFPHDKNVSVLV HGHLGRHLLHLOLCHLHRHBYKLHVPTTDFERRRNHSONRTPCHGHVVOFPRDKRVSVVV "GRLGRVLLHLOLCRHTRHRYKLNVPNTSFDTRSNHNONRTPCHGDHVOFPRDKHVSVLV HGaLGRvLLHLQLCRItnfiagKLHVPanF#.HaNHsONRTPCflG.aVRFPRDKnVSVIV 61 70 80 90 100 110 120 I e e e e t I OEGHRVSDHLLPLDGELVLflSfiRGFGVSDVGSHLDCGREEPRVFRDSDRFSHHDPHLHRS RRSHGFSDHLLPRUGEFVLHSGRGFGRRDRGRDPDCGRGRPRLFLDPDRFSHHDPRL"RS RDSHRISDHLLPLDGELVLRSGRHLSRHGGDSDPHCNPGRPLLFRNPDRFSHLDPHLNSS r.sHa.SDHLLPIDGEIVLRSGflgfgaad.gsdpngaGaPalFr8pDRFSHhDPhLNrS 150 170 l - - ¢ - ¢ I GDEHPGLFFVDRERVPCRHDDVFFPPSRSFRVGLGPGRSPVRVRSISHLGRTFTRDEDLR GDHRRGLFSVDRERVPCRHUDVVFPPDHSFRVGLGPGHRPRRVRSVOVLGOTFTRDEDLfl GTORPGLFSVURERVPCSYDDVLFPRDGSFRVHLGPGPNPVHVRSVSHVGOTFSRDEDLT Gd.RpGLFsVDnERVPCthDV.FdeaSFRVgLGPGa.PerRS!saleTFtRDEDLa 181 190 200 220 240 I : t v c v I VFLRSRHGRLRFHGPGHLSVGPEDCRDPSGCVCGNHEHQPQICRRLLOPLGGRCPORRCH RFLRSRHGRLRFHGPGRLRVGPGHCRDPSGCVCGDHEVOPNICRRLLQPLGGRCPPRRCP HFLHSREGRLRFHGSGRLRVGSOHCTDRSGCVEGNHEHLPNICRSLLOPLGGRCPOBRCO aFLRSRaGRLRFHGpGflLrVGp.aCaDpSGCVCfiaflE.qPHICflaLLOPLGGRCPqflflC. 121 130 140 160 180 210 230 241 250 270 300 I ¢ . v v v I FHLRPOGOCCDLCGHVVLLTHGPRFDLERYRRRILOTFLGLPOYHGLOVHVSKVPRSSRL DRLRPEGOCCULCGRIVSLIHGPTFDIERYRBRLLRHFL--POYPGLQHHVSKVRRRPG- DPLLPOGOCCDLCGHIVSLTHDPTFDLERYRRRLLDLFLKOPOYOGLOVRVSKVLRD--- daLrPSGUCCDLCGH!VsLTHgPtFDlERYRHRlLd.FL..POY.GLOVHVSKV.R.... 260 280 290 301 310 320 330 340 350 360 I c e e e c l REHDTEIOVULVENGPETGGRGRLRRHLLHDVRENGEHLGVLERTHRESGHHVHGSSHHG --PHTEVOVVLRETGPOPGGHGRLRRRLLHDVHEHGERLGVLSHTHRESGRPVGDGSRHG --RHTEIOVVLVETEHRTGRRGOLGHRLLODHVHOGSVLGIVSRTLROSGKPNTHDSELN ..ahTE!OVVLvEtgp.tGgflGrLarflLLaDvae.GeaLG!lsflT.R#SGapv...Saag 390 A 410 420 v v v v e I --------- LflGGVRRHVLLflLLVLLVHPPLLRRHGRLRURRHE--RRHPRGRPLGFRNP PLGSGSRHGLHGGVHRGLLLLLLHLHHGLLLLRRHPRLRHTKRERLVHTPVERPLGFSNP OSSSG--HGLHGUVHHLVLLRLLGTV—-LLLLHRSGRLRHRRHEDREPVSHGLPLGFRNP ...sg..agLflGGVflfl.vLLaLL.1...llLLrRagRLRHrrhE...a.pagaPLGFrNP 381 370 380 400 I A A A 421 430 460 467 I t - - t I VFDVTRSEELPLPRRLSLVP--KRHHDSTSHSYFVNPLFRG-flEREfl VFDVHGSVG-PVPRTPOPPPROORGSSSTSRSYFVNPLFEE-HER IF DRIVFKOOPSVELPDSHOKVDILDIDTKFGCFVNPLF flGEflEnEfl )fpv. .s. . .P.pr.p. . .p. . .a. . .sTs.sufy!€.f:|;Eflg.flEflea 440 450 Figure 3.2 Alignment of human, dog. and mouse AMN proteins. The sole transmembrane domain is underlined by a gray bar. Two AP-2 binding signals for ligand-independent receptor internalization are underlined with dashed lines in the cytoplasmic domain (near position 420 and 460). Two nearly tandem N-linked glycosylation motifs (NXS/T) close to the N-terminus are underlined with dots (right before position 40). The predicted signal peptide cleavage site is indicated by an arrowhead. ()5 1107 1117 1127 1137 1147 1157 nglkluL4llllllll lllllllll 1lIJIIJLLJLlLlLllJlllLlLllLll CGTGGCGGC --------------------------------- GGGCCTGCTGCTGCTGC Mutant allele CG GGCGGC --------------------------------- GGGCC GC GC GCTGC Normal allele CG‘GGCGGCCGGGCTGCiGCTGCTGCTGCTGGCGC.GGCGGCGGGCCKGC{GCTGCTGC ’ > 24bp repeat Figure 3.3 DNA sequencing revealed a 33 bp in-frame deletion in canine AMN ofthe affected dogs. The deletion is located in a region containing two 24 bp repeats. The deletion was predicted to delete l 1 amino acids in a transmembrane domain. ()6 'l jag-1": t '-_._-,-‘,. 7.. M C A C C A C A C 5()()bp —> 'l " I Tm ~..- . Normal allele Mutant allele Figure 3.4 Mutation analysis ofthe 33 bp deletion in exon 10 ofAMN. Genomic PCR (JCF370 + JCF 366) flanking the 33 bp region were perfomied in a number ofdogs. The PCR products were separated on a 2% agarose gel. The mutant allele migrated faster than the nomial allele because ofthe 33 bp deletion. M. marker. C, carrier. A, affected. N, nomial. ()7 Discussion: Pool-PCR based cDNA cloning method can help to quickly identify the target clone, if used with caution. The potential risk of the method is that the indicator PCR may amplify undesired cDNA, which may lead to wrong directions. We actually experienced undesired cloning of estrogen receptor gene before we cloned the AMN cDNA. Thus, a second indicator PCR or sequencing may be necessary to assure that the clone of interest was amplified. The clone we identified from the cDNA library was thought to be the nomtal cDNA of canine AMN. However, later experiments demonstrated that the cDNA clone was identical to the mutant cDNA, that is, it has the 33 bp deletion. Although it may due to a in viva mutation event during the phage proliferation, a more likely explanation is that the starting mRNA material for the library construction contained some mutant mRNA. This indicates that at least one of the dogs used for kidney pools was a carrier for the mutation. It has been well known that if significant homology exists, the repetitive sequence region of a chromatid may misalign with its corresponding region in a homologous chromatid during meiosis. As a result, deletion or insertion of repeat units may be generated. Here, the presence of the two 24 bp repeat sequences suggests that the mechanism of the 33 bp deletion was an unequal cross event after misalignment of the repeat sequences during meiosis I. At least two lines of evidence suggest that the predicted transmembrane domain is a hona fide domain. First, protein alignment among multiple species showed that the transmembrane domain is similarly placed near the C-temiinus. Second, immunofluorescence studies showed that AMN assisted the anchoring of cubilin on the ()8 cell surface membrane (Fyfe et al.. 2004). Since the 33 bp in frame deletion is located in the transmembrane domain, we hypothesized that the deletion might abolish the transmembrane domain. Computerized transmembrane domain prediction (TMHMM Server v. 2.0, Center for Biological Sequence Analysis, Technical University of Denmark), returned a probability of 1.0 for a transmembrane domain between amino acid residues 363 and 387 ofthe normal sequence, but < 0.05 for the deleted sequence (Figure 3.5). The analysis further indicated that if the mutant product were translated and sufficiently stable. the polypeptide would be entirely extracellular and, therefore, secreted. A recent study showed that AMN and cubilin forms a novel complex (named cubam) on the apical membrane ofepithelial cells, where cubilin is responsible for ligand binding while AMN directs membrane localization and endocytosis of cubilin with its ligands (Fyfe et al., 2004). This study provides the first functional evidence to explain why mutations in either AMN or CUBN cause clinically undistinguishable symptoms in human l-GS patients. Our experiments showed that AMN in the affected dog could be readily amplified at a comparable amount to the nomial dog in RT-PCR. Therefore the mutant AMN mRNA exists and may be translated. Depending on the stability ofthe mutant protein, it may be degraded or be secreted. In either way, it is tempting to think that the defect or loss of AMN may disrupt the function of cubam, which is essential for IF-Cbl absorption. Mutations located in transmembrane domains of membrane proteins have been widely reported, but most ofthem are short deletions or missense mutations (Gasparini et al., 1991; Kelley et al., 1998; Gomez Lira et al., 2000). Deletion of more than l0 bp in a ()9 transmembrane domain is rarely seen. It would be interesting to find out the structure of the 33 del mutant AMN in the future work. In conclusion, we have identified the mutation responsible for the [-65 in a Giant Schnauzer family as a 33 bp in frame deletion, which was predicted to abolish the transmembrane domain of AMN and disrupt the functional cubam. 70 probability probability Normal canine AMN TMHMM posterior probabilities for Sequence 450 1.2 - 1.0 f 5 _—_" "—55 '5— "—U ”5 ' '— .-.1 7“ ' T—l I'm!“ l 0.8 3‘ . i ' l, a 0.6 . f g 0.4 " l 0.2 9* E ' 0 - . A 1 . Lin] 5 50 100 150 200 250 300 350 400 450 transmembrane» inside —- _, outside Mutant canine AMN 1 2 TMHMM posterior probabilities for Sequence 1.0 f' ' — "*———*" _____-___ 0.8 § 0.6 i 0.4 i 0.2 i l 0 - - - . um." i 50 100 150 200 250 300 350 400 transmembrane" inside “ ' outside” 71 Figure 3.5 Structure prediction of AMN by the TMHMM software. The normal canine AMN (upper panel) was predicted to be a single-transmembrane protein, while the 33del mutant AMN (lower panel) was predicted to lose the transmembrane domain and be secreted out of the cells. Images in this dissertation are presented in color. CHAPTER 4 LINKAGE ANALYSIS AND MUTATION SCREENING OF AN AUSTRALIAN SHEPHERD FAMILY WITH IMERSLUND-GRASBECK SYNDROME In this chapter, I designed and perfomied all the experiments, except that Schaffer AA did the computer-based linkage analysis and Kilkenney A did the genotyping with C UBN. Introduction Imerslund-Grasbeck syndrome, characterized by cobalamin malabsorption and proteinuria, has been reported in different ethnic groups, such as Norwegian, Finnish, Jewish, Turkish, etc (Grasbeck et al., 1960; Imerslund 1960; Ben-Ami et al, 1990; Yetgin et al., 1978). The AMN c.l4delG mutation was commonly seen in Norwegian patients, while AMN c.208-2A>G mutation was carried by two Turkish families. Two other distinct mutations of AMN have been identified in a USA family and a single Belgium patient, respectively (Tanner ct al., 20045). It appears that multiple independent mutation events have occurred in the AMN gene during the human history in different subpopulations. Similarly, an inherited disease may be observed in different breeds of dogs. In the previous chapter, we described a 33 bp deletion in the AMN gene identified from a Giant Schnauzer family with I-GS. The deletion leads to loss of l 1 aa, which was predicted to abolish the sole transmembrane domain of AMN. This is devastating to the function of AMN, because biochemical and cell-based studies suggested that the transmembrane domain of AMN is essential for anchoring cubilin on the apical membrane, to endocytose certain ligands such as IF-Cbl (Fyfe et al.. 2004). In this chapter, we study a second canine I-GS family of purebred Australian Shepherd dogs (AS kindred; Figure 4.1). In this pedigree, three affected littemiates exhibited growth failure, methylmalonic aciduria, mild anemia, neutropenia, subnormal serum cobalamin concentrations, and low- molecular-weight proteinuria (Fyfe, unpublished data). Parenteral cobalamin administration produced complete clinical. hematologic, and metabolic remission even 73 though cobalamin malabsorption and proteinuria remained, suggesting that the AS kindred have the same genetic defect as the GS kindred. It is therefore intriguing to ask whether the AS kindred carry the same 33 bp deletion as the GS kindred. Ifnot, does another mutation exist in the AMN gene? Or does C UBN harbor the mutation? The reason to pursue such questions is that different mutations may help to better understand the structure and functions of a protein, as seen in the cases like methylmalonyl-CoA mutase (Janata et al., 1997) and LDL receptor (Jensen et al., 1997) defects. On the other hand, it is also helpful to provide a genetic test for dog breeders to eradicate the disease in the specific dog breed. 74 C/T C/C 100 121 122 C/C C/C C/C Figure 4.1 KNSZ genotyping of the Australian Shepherd kindred. Three affected dogs, 100, 121 and 122, are homozygous for the C allele of KNSZ; two carriers, 101 and 102, are in the heterozygous status of OT; most of the clinically normal dogs are heterozygous or homozygous for the T allele of KNS2. Therefore, the disease appears to segregate with the marker in an autosomal recessive manner. Since AMN gene is only ~6OO kb from the KNSZ marker, it suggests that AMN may be the disease-causing gene. 75 Materials and methods me All dogs were handled according to the principles outlined in the NIH Guide for the Care and Use of Laboratory Animals, with protocols approved by the MSU All University Committee for Animal Use and Care. Blood or buccal brush samples were collected from 28 dogs ofthe kindred each of which was determined to be affected or clinically normal. The relationships ofthese dogs are partially depicted in the pedigree in Figure 4.1. Blood or buccal brush were stored frozen at ——80°C for DNA isolation. Markers and genotyping DNA was isolated by standard methods (Sambrook et al. 1989). To test the linkage of the disease to C UBN, an intronic 17 bp variation in C UBN previously described (Xu et al., 1999) was chosen as a marker. On the other hand, the KNS2 marker was chosen to test if the disease is linked to AMN, because this marker was only about 600 kb away from the AMN gene. Both markers were shown to be informative in this pedigree. CUBN primers: (JCF119) 5’-GAT CAC AGG CCT ACA GCT CCA TT-3’ (JCF120) 5’-CCA GGC CAA CCA GAG ATC TTC TA-3’ The C UBN PCR products were run on the 4% agarose gel, on which two different alleles showed two distinct bands. KNSZ primers: JCF294 and JCF295 have been described in Chapter 2. 76 Linkage analysis Linkage analysis computations were performed with the FASTLINK software package as described in Chapter 2. LOD scores were calculated with assumptions ofa disease allele frequency of 0.001, full penetrance ofthe disease trait, and equal marker allele frequencies. Mutation screening DNA samples from affected dogs B122 and B100 and an unrelated normal dog (DCCU61 10) were used for mutation screening. The primers and PCR reagents used for genomic PCR amplification are listed in Table 3. Mutation detection Primers: JCF563 (5‘-GGC TTG GAA GGA AGG CCC CCA-3’) is located ~280 bp upstream of the ATG start eodon. JCF387 (5’-CAA GGC GGG GAG CCT CCG AA-3’) is located in intronl, ~90 bp downstream of the ATG start codon. The Thermal AceTM DNA Polymerase Kit (Invitrogen) was used for the PCR. The PC R products were digested with EMF 5 I at 65°C for 2 hours and run on the 2.5% agarose gel. The mutant allele was resistant to the digestion while the normal allele was not. 77 Table 3. Primer list for genomic amplification of canine AMN. Notably, JCF392 (reverse) partially overlaps with JCF422 (forward) located in intron 6. .1 CF 356 (reverse) partially overlaps with J CF 370 (forward), which crosses intron 9 and exon 10. J CF 334 overlaps with the ATG start codon and amplifies in both normal and affected dogs. Notations in the PCR reagent column: a, Thennal AceTM Kit (Invitrogen); b, Expand‘fi High Fidelity PCR System (Roche); c, Advantage Bi-GC Genomic polymerase mix (BD Biosciences); d, Taq DNA polymerase (Qiagen). PCR Set Primer ID Primer sequence location reagent set 1 JCF334 5’- CGG GCG CGC GGC GGG ATG-3’ exon1 a JCF368 5'- CGT CCG GTT CTG GCT CCA GTT G-3’ exon2 set 2 JCF388 5’— CTCCGCAGACCTCGTAGGAGTT-S’ intron1 b JCF391 5’- AGGCTGGGGAAAGGGTATGGG-B’ intron3 set 3 JCF416 5’- TGC AGT CCT GCT CCC TCG GCT T-3’ intron3 b JCF396 5’- TCTTCCTCGACCCCGACCGCTT -3’ exon5 set 4 JCF397 5’- CGGCTTCTCTGACCCTCGGACA-3’ intron3 b 1 JCF339 5’- CGC TCG GCG TCC ACG GAG AAG A-3' exon5 set 5 JCF328 5’- TCG ACC CCG ACC GCT TCT TGT G-3’ exon5 c JCF392 5’- AGC GGG GTG AGC GCG GAC AGT-3’ intron6 set 6 JCF422 5’- ACT GTC CGC GCT CAC CCC GCT T-3’ intron6 c JCF423 5’— TCA CCG CAG AGG TCG CAG CAC T-3’ exon7 set 7 JCF355 5’—GGA GGT GCA GCC CTG GAT CTG-3’ exon7 c JCF346 5'-GCC GCC GCC GCA CCT TGG ACA C-3’ exon9 set 8 JCF335 5’-CCA CGG CCC CAC CTT TGA CAT C-3’ exon8 d JCF330 5’- CAC CAC CTG AAC CTC CGT GTG C-3’ exon9 set 9 JCF362 5’-GCA GGC GGC CGT GTC CAA GGT G-3’ exon9 c JCF356 5’-CGA CAG GAC CCC GAG GGC TTC-3’ exon10 set 10 JCF371 5’- GGA GCA CGG TAA CCG CGG GTG—3’ intron9 a JCF366 5’- CTG CGG GGT GCG TGG AAC CTA G-3' intron11 set 11 JCF370 5’- TCC GTT GCA GGC GAA GCC OTC-3’ exon10 a JCF366 5’- CTG CGG GGT GCG TGG AAC CTA G-3’ intron11 set 12 JCF351 5’-GTG GAC TAA GCG CGA GCG ATT G-3’ exon11 c JCF332 5’- CTG GCC AGC CCC GCG GTT GC-3' exon12 78 Results: Exclusion of CUBN 8100, B121 and 8122 were affected littermates descending from BIOZ and 8101. Under the null hypothesis that CUBN was the disease-causing gene, B100, 8121 and 8122 should have identical genotypes for C UBN, because both of their parents were carriers and therefore each parent could transmit only one disease allele to the affected offspring. However, the genotyping data of CUBN showed that only 8100 and B121 shared the same genotype while 8122 had a distinct genotype (Table 5 in Appendices). Thus, the null hypothesis should be rejected and we concluded that CUBN could not be the disease-causing gene. Linkage analysis with KNS.? 11 members of the kindred, including 3 affected siblings were genotyped for a polymorphism in the KNSB gene that was also variable in the GS pedigree (Table 5 in Appendices). Assuming that the C allele ofKNSZ is not rare and the disease allele having low probability, two-point linkage analysis was performed and LOD score 1.7 (6:00) was obtained. This suggested, to a slight extent, linkage of the disease to the KNSZ marker and thereby to the AMN gene, because KNS2 was only ~600 kb from the AMN gene. Mutation screening As tissues were initially not available, we could not do the RT-PCR in the affected dogs. Although the 7.6X coverage of the dog genome was partially available during our 79 mutation screening, the AMN gene contains several sequencing gaps and the sequencing quality was very poor. We therefore generated most of the genomic sequence of AMN in our lab by overlapping PCR and sequencing (Figure 4.2). Sequences of intron 3 and 5’ UTR were not obtained because of either technical difficulty or lack of primer sequences. We amplified nearly all the exons and exon-intron boundaries of AMN (partial exon 1 was not amplified due to lack of 5’UTR sequence). However, no mutations were identified. Five variations were seen in intron 4, intron 5 and intron 8, but were not located at the 5’ or 3’ intron-exon boundaries. The variations did not seem to disrupt the site of branching point A either. RT-PCR was perfomted later with JCF334 and JCF332 as the kidney cortex was obtained from 8100. This PCR covered the full-length AMN cDNA except the ATG start codon, for JCF334 overlapped the ATG site. With no mutations found, we then did the RT-PCR with JCF379 and .lCF339, to include the ATG site in the PCR product. DNA sequencing identified a single nucleotide change (G>A) in the start codon (Figure 4.3). In order to confirm that the variation was at the DNA level, we amplified the 5’ region by genomic PCR. Facing a sequence gap at the 5" region of AMN, we first designed primers based on the UC SC Genome Browser on Dog Assembly (July 2004) (http://www.genome.uesc.edu/cgi-bin/thateway), and filled in the ~1 kb gap by sequencing the PCR products. With the 5’UTR sequence available, primers JCF563 and .lCF387 were designed and subsequently used to PCR the genomic region flanking the ATG site. The G>A variation was confirmed by sequencing, along with some other polymorphisms identified in the 5’ region upstream of ATG. 80 The G>A transition abolished a BstF5 l digestion site, which could be used for convenient genotyping (Figure 4.4). We thus used this method to genotype other dogs in the AS kindred, plus more than 50 unrelated nomial dogs. The G>A transition segregated with the disease allele ofthe AS kindred as expected for an autosomal recessive disorder in 3 affected (A/A). 2 obligate carriers (G/A), and 16 other clinically normal Australian Shepherd dogs (13 G/G and 3 GA) (Table 5 in Appendices). The mutation was not observed in 1 12 chromosomes of unrelated normal dogs of various breeds. 81 Figure 4.2. Genomic DNA sequence of canine AMN gene. The italic sequence is upstream of the known AMN DNA and may contain the transcription initiation site (the underlined italic sequence represents a polymorphism). The following upper case letters represent exons sequences, while the lower case letters represent intron sequences. The coding regions are shaded. The size of intron 3 has not been determined. CC C I CC CCC CC C CCC C CC CC GC TTTC CA TCA TTCTTCCCCA C C C AA G CA C C A G C C A CCC'GGCCTC‘TTGCA C‘CGCGGGGTC‘YCCTCGACC'C'GCCCCCGCCCCCGGGCCG (’CCCCA TCACTTCGCGGGGCCCC'CCAA CA CCC/1CTCGTGCTCCGCCACCAGCTCC GTGTGCAGC’CCGGGGGGCCTYGCCCCCTCCGCCCCCCACCGCCACAGGCCCA CT GCCGGC’GCC’CGCAGGGCA TCCAGGCGGGGGGGGGTGAGGGGGGCTTGGAAGG AxlGGCCCCCARCC’CAGGGGGGGCA GAGCA GGTGCAGCCCCCGGGCACAGGGCA CGGGAGC’IGGGA GCCCAGGGGGTCCCTCCCGCCCCCAGC’TCGCCC'CGCCTTGG AGCGGGTGGGCGGGC‘C‘CCAGGGCTGAGTGTGGGGGCAGGAGC’CGCCGCGGCG C’CCCGGGGCC’SGGGGAGGTGGTAACGCCCCGCC‘CCGCCCCgeeeegccr'C‘GCCCCG CCCGGTCZI GGTGGGC’C'C'CGCG(IGGO1AAGTC'CCGGTGGCGGCGACGGGCGCG CGGCGGszT’glGGCGCGCTGGGCCGGGCCCTGCTGTGGCTGCAGCTGTGCtha agggggccccggggcgcggcgggggcttcggaggctccccgccttggggcccggacccctcgggcgccggg m t ’ ggggcct gcgagccgcgcaccgcccccagttcccgctccctgcggggchggccgcagcctgactcagtttcccctccgccttchcc cggccggcgcggtcccccgggcccgggagcttggaaact"cauccccUcccccUccccncccccgaucecccg ggt cc cgcgga cacgagtcccca gg gg cc«ccctccc0cveccccuccccucgacccccttgtgccut gtccetcggttgtcgcg Uccttgccccgccgggccctctgctccctctgctccagcggcccgcaagcccgg ggcgaggcgac agcgatcttggc ggg cg ycctggagggggcacgggggggcgggctccgcagacctcgtagga gttcgcgccccgcg gc ggggctcccgcgggct Exon2 Hcccccggcthg I gccgccttccttccg gggggggggggcggcggggagcactcagtcgtgccctccccagCGC 9. o: n: U: TGGCC C GGGCCGCCTACAAGCTCTGGGTCCCCACCACGGACTTCGAGGCCGC C GCC AACTGGAGCC AGAAC CGGAC GC CGTGCGC GGGCGC CGTGGTC C AGTTC CCCGCGGACAAthgccccccgcgggctcgg Iaggcggactg ItI Iggggctccggg ttI It ggagt gctggt ggggctcg Exon3 ggggggcccgg IccaacccgctccttcctgcagGCGGTGTCGGTGGTGGTGCGGGCCAGCCACG GCTTCTCGGACATthgaggc cgggg ctgcggggggtggccccecggaaceccctctgggcctggggcgggct ((((( tgcccggtccgcgg tI tIt gaacgcggcgg Icg tIt ggggcgggaccgtggggcacccggg Iccaccgtgtgcctgt gctgcccgcagC Bk 4 TCCTGCCGCGGGACGGGGAGTTCGTCCTGGCCTCGGGAGCCGGCTTCGGGGC CGCGGACGCCGGCAGGGACCCGGACTGCGGCGCAthgaggggcgggg cggggcggggc gggaggCS‘a’SgCCCSSSSCSSWL’CugngggCCt‘a’g‘o’nggCL-l’SSCSSSSCC Sng 8‘: 8C8 gggcctgagc Nan—«vs, ExonS gggggcggggggcggggeccg Igcccggggg egg agctcag IggacgcgcgcccccgcagGCGCCCCCGCG CTCTTCCTCGACCCCGACCGCTTCTYGTGGCACGACCCGCGCCTGTGGCGCTC CGGGGACGCGGCGCGCGGCCTCTTCTCCGTGGACGCCGAGCGCGTGCCCTGC CGCCACGACGACGTCGTCTTCCCGCCCGACGCCTCCTTCCGAGTGGGGCTCG GGCCCGGCGCCCGCCCCGCGCGCGTCCGCAGCGTCCAGGTTCTGGGCCAthg agcggcgctcggt ccecctccccgacctgcccacgcgctItIctgccgggcgctcaggg ctgtcccctcctcgggcggcgc cgctcgtgccgcccctccccccgcagIZCnGTTCACGCGCGACGAGGACCTGGCTGCCTTCCTG GCGTCCCGCGCCGGCCGCCTGCGCTTCCACGGGCCGGGCGCTCTGCGCGTGG GCCCCGGGGCCTGCGCCGACCCGTCGGGCTGCGTCTGCGGCGACGCGGAthg aggg cggccggcgggggg Igcggaggg Igcggaggg Iggcggggggnnnnactgtecgcgctcaccccgcttcctccgc agtggagCTGCAGCCCTGGATCTGCGCGGCCCTGCTCCAGCCCCTGGGCGGTCGC TGCCCGCCGGCCGCCTGCCCCGACGCCCTCCGGCCCGAGGGGCAGTGCTGCG ACCTCTGCthgagcgcccectcccggcccggagagctgccctggctcgcccecagcetcagtttccctgacggccct gt It IctcgctItIchCgaccecttccctItctgtcgcagGAGCCATCGTGTCGCTGACCCACGGCCCC ACCTTTGACATCGAGCGGTACCGGGCGCGGCTGCTGCGAGCCTTCCTthaac Ig 83 gtgccgcgtccccgcccccgccctgcccccccccgcgg Iggeccgcctcttccgcgg Icggcgcccggg Iaccccactgccccc ccacgcagtcactgaccgcgcacctcccgtcacccgcecgccg gecgagg Igccaggtccc IggaCt gaccccgtctccctcc ccagCEGCnCAGTACCCGGGGCTGCAGGCGGCCGTGTCCAAGGTGCGGCGGCGGC CGGGGCCGCACACGGAGGTTCAGGTGGTGCTGGCGGAGACCGGGCCCCAGC CGGGCGGCGCGGGGCGGCTGGCCCGGGCCCTCCTGGCGGACGTCGCGGAGC ACthaaccgcgggtgcccctceccggccggcecgg CIcccgcgtgcgggagcctgagcccgcccetccgttgcagGC EXAIXOGCCCTCGGGGTCCTGTCGGCGACAGCCCGGGAGTCGGGCGCGCCCGTCG GGGACGGCTCGGCGGCGGGGCCGCTCGGCTCGGGTTCGCGCGCGGGGCTGGC GGGCGGCGTGGCGGCCGGGCTGCTGCTGCTGCTGCTGGCGCTGGCGGCGGGC CTGCTGCTGCTGCGCCGCGCTCCGAGGCTCAthccgcg ggg cgggg tIt gtcgggggcgggggc gggggcggggccgcgtggg gg Ictcac egg Iggc tItccttgttccccagG’loCiliACTAAGCGCGAGCGATTGG TCGCCACGCCCGTCGAGGCGCCCCTGGGCTTCTCCAACCCGGTGTTCGACGTG GCGGCTCCGTGGGGCCthgagg gtg gcgcgcggggg Iacctgccctcccgg Iccthcggccgccgacgcccc ttgactccgcgcccccgcecctagGTTCzCACGCACCCCGCAGCCTCCCCCAGCGCAGCAGG CGGGAAGCAGCAGCACCAGCCGCAGCTACTTCGTTAACCCGCTGTTCGCCGA GGCCGAGGCCIQAGCAACCGCGGGGCTGGCCAGCCCCTACCTGCGCCCGCCG CCGCCCCCGCGAGATGGCCCCGGCCTTGCGAGGTCCCCGCCCCCTGCCACGC ACGCCTTGTCCCCCCAGCCCAAGGATAGGGTGGCTTTGCCCAATAAAGCGTT TCCTGCacccggagtccgttgcccccagtggc ccctcctgtgectg ggg cttgagcgtggggagccggccagtgtgcccgte ag tIt gctgeccactt ggtg atIcgtgctcctcatIaatccttIccctcttItIccaccatIeaaagggcauctceaggcactaccgaaaatt gcagtctgcag agctgcacccagaatgtgg Iggaagt gctt gCIccctggIagcat ggggagcgaggngcgggtgagg ggctg [CCEIUCaCCCCCEIO’CaUCCa”((1521ch8chaUlCCCIlCCCCa”(logaaC21gHaltallCl(ata atgcagaggaatgacact 84 gctgcaggatatgcgggctgcggcggaaccacagcagggcagcetggatIacccagggagaggaagaggaggacaaggg V S 85 ''''''''''''' I V V V Y I V I’ I V V U I V I CGCGGCGGGATRGGCGCGCTGGGCCGGG Normal allele '1 I I I l l I ‘ ’ . x.’ ,1 l J . 5 7 f \ /1 I ~ ._ fl )1 rA‘ ‘ ‘ '1" CGCGGCGGGATGGGCGCGCTGGGCCGGG Mutant allele t i _/ / ale 3’, I" 1A.], fr“? .1.. CGCGGCGGGAT.GGCGCGCTGGGCCGGG l ‘ ' 5 l ’ 5V .‘ . ‘5 l‘ ’ ‘l l 1 ‘ l. I ’ Ll 1’5 1 I /I L ‘5 1‘ I 1 l i - .1 I ‘ I 5»: :j\ 11;.» A 1 Figure 4.3 DNA sequencing revealed a G>A mutation in the start codon (underlined) of canine AMN in the affected dogs. The next in-frame ATG in canine AMN is 204 bp downstream, but not within the Kozak consensus context. 86 CCCNNN Mt tll-le “a... ,uanae ~--~~ «alum—ah M 500 bp—> .7. - Nomial allele Figure 4.4 Mutation analysis of the G>A transition in start eodon. Genomic PCR flanking the start codon were performed in a number of dogs. The PCR products were digested by EMF 5 I for 2 hours and separated on the 2.5% agarose gel. The normal allele was digested into two fragments (284 bp and 95 bp), while the mutant allele was resistant to the enzyme digestion. M, marker; C, carrier; N, normal. 87 Discussion C UBN and AMN are the two known genes that underlie the l-GS disease. With the CLI’BN gene being ruled out by exclusion linkage analysis, we therefore focused on the AMN gene. Directly testing linkage of the AMN gene to the disease was not possible because no infomiative marker was available in the AMN gene. However, since the KNS2 marker is very close to the AMN, we decided to test the linkage indirectly by genotyping KNS2 in the AS family. The LOD score obtained was 1.7, which means that the likelihood of linkage over nonlinkage is 10"7 =50. The score of 1.7 is below the generally accepted score 2.0 for suggestive linkage. Two reasons may explain why the LOD score was not high enough to suggest linkage. First, the number of family members tested is relatively small (11 members), and only 3 affected dogs were in the family. Second, the KNSZ marker was not in complete linkage disequilibrium with the disease allele in this AS kindred. One unaffected dog B108, who is a carrier ofthe AMN mutation, is homozygous for the disease-associated C allele at the KNS2 polymorphism. Another unaffected dog B109 who does not carry the AMN mutation is heterozygous at the KNS2 polymorphism site. It appears likely that the C allele of the KNSZ polymorphism has a high enough frequency that it entered the pedigree on multiple haplotypes, only one of which is associated with I-GS. Nevertheless, AMN appeared to be a compelling candidate gene, for which we decided to do an extensive mutation screening. The AMN gene has a 70-80% GC content, which makes the standard PCR condition not useful. High denaturing temperature, GC-Melt reagent, DMSO and some high- temperature-resistant polymerase helped to amplify the sequence, but none of them 88 worked universally. Although the dog genome sequencing project has been completed recently, there are still 3 gaps in the AMN gene, and the sequencing quality of the AMN gene is very poor. This is likely due to the high GC-content, because we have also experienced difficulties in sequencing some of the PCR products. Thus, combining genomic PCR with RT-PC R to detect mutations was necessary. The G>A mutation is predicted to change the initiating methionine into isoleueine. A few rare cases of altemative initiation eodons have been described, including ACG (Thr), CUG (Leu) and GUG (Val) (Taira et al., 1990; Tailor et al., 2001), but AUA has never been reported to be able to initiate translation to the best of our knowledge. Eukaryotic translation initiation consensus has been defined as (A/G)CC_A_U§G. In vitro translation experiments showed the G at position +4 and the purine at position --3 are highly conserved (Kozak, 1986; Kozak, 1997). For the canine AMN gene, the next in- frame ATG is at position 205-7, but not within the Kozak consensus context. Thus the mutant mRNA is unlikely to be translated. Furthermore, even iftranslation were initiated in a very rare circumstance, amino-terminal truncation of 68 residues would eliminate the signal peptide sequence required for co-translational rough endoplasmic reticulum insertion and subsequent delivery to the plasma membrane. Therefore, the G>A transition in the start codon is essentially a null mutation. This finding strengthens our previous conclusion that AMN is responsible for the canine I-GS. The two canine pedigrees harboring different AMN mutations provide us a unique opportunity to study the functions of AMN and cubilin directly in tissues of both affected and normal dogs, which is nearly impossible in the human study of l-GS patients. 89 CHAPTER 5 FUNCTIONAL STUDIES OF CANINE AMN MUTATIONS The experiments described in this chapter were carried out as a collaborative investigation involving Mette Madsen and Erik Christensen. Erik Christensen did the immunohistochemistry. For the transfection and immunofluorescence studies of the CHO cells, I designed the experiments and constructed the plasmids, while Mette Madsen carried out cell transfection, Western blots and immunofluorescence of the transfected cells. I did all the other experiments shown in this chapter. The interpretations of the data described here are mine. 90 Introduction Mutations in either cubilin or AMN gene cause l-GS in humans. No clinical differences have been reported between patients carrying mutations in CUBN vs. AMN (Tanner et al., 2004). Cubilin is a 460 kDa, multiligand, apical membrane receptor protein that mediates endocytosis ofthe intrinsic factor (1F)-cobalamin complex in distal small intestine and of several proteins of the glomerular filtrate in renal proximal tubules (Christensen and Bim, 2002). AMN is an ~ 45 kDa apical membrane protein also expressed in intestinal and proximal tubule epithelia. but whose function is poorly defined (Kalantry et al., 2001; Tanner et al., 2003). Recent studies in human cadaver kidney and transfected Chinese hamster ovary (CHO) cells suggested that cubilin and AMN function as subunits of a receptor complex, now called cubam. The complex is Cay independent and does not separate under denaturing condition of 2 M urea. Cotransfection of both AMN and a mini-cubilin construct into CHO cells demonstrated .. ta .. . that cubilin was expressed on the cell surface and c\onfe_rr‘edmtorthe cells the ability to endocytose lF-Cbl. In CHO cells transfected with only the cubilin construct, immuno- confoeal and immunoelectron microscopic examination revealed that cubilin was retained intracellularly, and the internalization of ’BSI-IF-Cbl was considerably less than in the double transfectants (Fyfe et al.. 2004). However, because the disease is completely ameliorated by periodic parenteral cobalamin administration, and the tissues that express the gene products are not readily accessible, there have been no studies that directly assess the effect of CUBN or AMN mutations on cubam expression in I-GS patients. Canine I-GS is a well—established animal model of the human disorder (Fyfe et al., 1991a; Fyfe et al., 1991b; Xu and Fyfe, 2000) that has contributed significantly to 91 understanding of its biological basis as well as molecular aspects of CUBN expression and cubilin function (Nykjaer et al., 2001; Kozyraki et al., 2001; Bim et al., 2000; Kozyraki et al., 1999). Cubilin has been implicated in the etiology of canine I-GS for more than ten years. It was demonstrated that cubilin had an abnonnal conformation and failed to reach the apical microvillus surface membrane of villus tip enterocytes in the affected dogs (Fyfe et al., 1991a; Xu and Fyfe, 2000). With the mutations in AMN already identified in both kindreds, we explored how mutated AMN caused the mal- expression and dysfunction of cubilin. In order to detemiine the molecular basis of the observed abnomialities of cubilin expression in the canine I—GS model, we studied the expression pattern of both proteins in multiple tissues by RT-PCR, and investigated the mutations’ effects on both cubilin and AMN expression in viva. Additionally, comparison of these results to expression of mutant canine AMN in a heterologous mammalian cell transfection system validates the system for examining functional defects caused by human CUBN and AMN mutations at vi v0. 92 Materials and methods mRNA ewression mttem in multiple tissues JCF426 5’AGCCTGCGTGCTGGACATAGAC 3’ JCF427 5’ CCAGCCCAACCTGATTCACACTTA3’ JCF428 5’ TCAGGGTGGAGACTTCTCAAATC3’ .lC F429 5’ GTTGCAGCTTCAGTCTATCTGCT3’ JCF334 and JCF332 have been described in chapter 3. Total RNAs were extracted from multiple tissues of a nomial dog. RT—PCR for AMN (JCF334 plus JCF332) was perfomied with the Expand high fidelity PCR system (Roche). ,RT-PCR for CUBN(.1CF426 plus JCF427) and Cyclophilin gene (JCF428 plus JCF429) were perfomied at standard conditions. Five mieroliter of PCR products were electrophoresed on 1% agarose gels. Characterization of the anti—AMN aritiborLy Anti-canine AMN peptide antiserum was produced by immunization of rabbits with the synthetic peptide TARESGAPVGDGSA (amino acid residues 340-353) of canine AMN. Peptide was synthesized, immunizations performed, and serum harvested in the custom antibody production facilities of ProSci Inc. (Poway, CA) under NIH Animal Welfare Assurance (no. A4182-01). The anti-AMN antibody was affinity purified by the synthetic peptide. The cubam complex was enriched, in bulk, by IF-cobalamin affinity chromatography from pooled fresh frozen kidney of unrelated nomial dogs (Fyfe et al., 2004) (provided by Dr. Fyfe). However, this enriched cubam complex still contained many unspecific proteins because of technical limits. The enriched cubam complex was 93 applied to reducing SDS-PAGE (12%) and transferred to PVDF membrane, then incubated with the preimmune serum, affinity-purified anti-AMN, llowthrough of the peptide-affinity-column-absorbed-scrum or secondary antibody only, respectively. The secondary antibody is a goat-anti-rabbit IgG alkaline phosphatase conjugate. The immunoactive proteins were detected by incubating membranes with 5-bromo,4—chloro,3- indolylphosphate (BCIP) / nitroblue tetrazoliurn (NBT) for 10 minutes. lF-Cbl Pull down and Western blot Firs! pull-down: For analysis of AMN expression in biopsy quantities of tissue, 2.7 g kidney cortex was thawed and homogenized in 24 mL of cold 50 mM Tris-Cl (pH 7.4) containing 150 mM NaCl, 1 mM N-ethylmaleimide, 5 mM phenylmethylsulfonyl fluoride. 3-[(3-cholamidopropyl)dimethylammonio}l-propanesulfonate (CHAPS) was added to 0.6 %, and the homogenate was incubated at 4° C for 2 h with constant agitation. The detergent extracts were centrifuged at 4° C for 40 min at 40,000 g, and the supernatant was collected. The protein concentration was detemtined by the Lowry method. Rat lF-cobalamin beads were prepared as previously described (Xu and Fyfe, 2000). Supernatant aliquots containing 100 mg oftotal protein were incubated with 20 ul ofrat lF-cobalamin beads with 5 mM Ca2+ ovemight at 4°C and centrifuged for 15 min at 4,500 g. The beads were washed 3 times with 1 mL ofhomogenization buffer containing 0.6% CHAPS and 1 mM CaCl;, then resuspended with 7 ul 4x SDS-PAGE sample buffer and 3 ul 10x DTE. They were boiled for 8 min and subjected to SDS-PAGE (12%), followed by Western blot. The first antibody was affinity-purified rabbit-anti-canine 94 AMN peptide (residues 340-353 of AMN) (1:40,00()), and the secondary antibody was goat-anti-rabbit IgG conjugated with alkaline phosphatase (l:20,000). Second pull-down: The supernatant after the first pull-down was incubated with rat IF- C bl beads again. As the protocol described above, the beads were washed and applied to reducing SDS-PAGE (7.5%), followed by Western blot. The first antibody was a rabbit- anti-dog cubilin (1:20.000), and the secondary antibody was a goat—anti-rabbit IgG alkaline phosphatase conjugate (1:20.000). 1m munohistochemistry Sections of normal and affected dog kidney cortex were fixed briefly in phosphate buffered 4% paraformaldehyde, sliced to 1.5 mm thickness, and immersed for 48 h in phosphate buffered 1% parafomialdchyde. Thin sectioning and peroxidase-labeled irnmunostaining for light microscopy was as previously described (Bim et al., 2000). Primary antibody was a 1:4000 dilution of previously described rabbit polyclonal anti- canine cubilin serum that did not cross—react with AMN (Xu and Fyfe, 2000; Bim et al., 2000). Transfection study l. Plasmid construction JCF 381 5’-TTT AAA GCT TCG GGC GCG CGG CGG CAT G-3’ .lCF382 5’-TTA TAA GCT TGG CCT CGG CCT CGG CGA AC-3’ We did full-length RT-PCR for the normal AMN with primers JCF381 and JCF382 to introduce Hind/ll restriction sites, which are underlined above. With the same primers, 95 we amplified the 33del mutant AMN cDNA from a plasmid pGEX-33del-AMN. The PCR products were first cloned into the pCRtR)-2.l TOPO® cloning vector (Invitrogen), then subcloncd into the vector pcDNA-5xmyc (a gift from Dr. Tanner) at the Hind!!! sites. The plasmids harboring either normal or mutant canine CAM/V were sequenced to confirm integrity ofthe inserts before being transfected into the CHO cells. 11. Transfection The cDNA encoding amino acids 1-1389 of rat cubilin was ligated into the Xbul and Him/Ill sites of the expression vector pcDNA3.l/Zeo(-) (Invitrogen) and stable, single- transfected Zeoein-resistant Chinese Hamster Ovary (CHO) cell clones were established as described (Kristiansen et al., 1999). Establishment of stable, double-transfected CHO clones expressing both the cubilin construct and myc-labelled canine AMN was carried out by transfection with the plasmids described in 1, and selection with 1 mg/mL GeneticinR (Invitrogen). Expression products of all clones were analyzed by immunoblotting ofcell lysates using rabbit polyclonal antibody against rat cubilin and an anti-myc-antibody (Invitrogen) for AMN detection. For endo H digestion, CNBr- activated Sepharose 4B beads (Amersham Biosciences) coupled with porcine lF- cobalamin as previously described (Bim et al., 1997) were incubated with cell lysates of double transfected CHO cells for pull-down of recombinant cubilin. The beads were suspended in 50 mM sodium citrate, pH 5.5, 0.02 % SDS and incubated over night at 37° C with endo H (Roche) prior to SDS-PAGE and immunoblotting. lll. Confocal immunofluorescence microscopy lrnrnunofluoreseent analyses of CHO cells expressing cubilin/AMN(33del), and cubilin/AMN were perfomied on living non-permeabilized cells. Living cells were 96 incubated for 90 min at 4°C with a polyclonal antibody against rat cubilin diluted to 10 mg/mL in growth medium. Cells incubated first with the primary antibody at 4°C were fixed for 1 hour at 4°C. Finally, the cells were incubated for 1 hour at room temperature with the secondary antibody: Alexa Fluor 488 goat anti—rabbit IgG (Molecular Probes) diluted 1:200. Stained cells were examined by confocal fluorescence microscopy using a laser scanning confocal unit (LSM510, Carl Zeiss) attached to an Axiovert microscope. 97 Results mRNA expression pattern RT-PCR demonstrated full-length AMN transcripts as expected in tissues of known CUBN expression, including kidney cortex, jejunum, and ileum (Figure 5.1). Both genes were also expressed in thymus but at very low levels. In addition, AMN was expressed in colon, liver, pancreas, and pituitary tissues where CUBN cDNA was not detected, and CUBN was expressed in placenta where AMN was not detected. Shortened RT-PCR amplification products were detected in small amounts in tissues where AMN was expressed. Sequencing of one of these that was amplified from ileal RNA demonstrated absence of exons 9 and 10, exactly, and predicted a secreted translation product of 262 residues (~ 30 kDa). Characterization ofthe anti-AMN antibody Although the cubam complex was affinity purified and enriched, numerous other proteins still coexisted with the cubam complex (Fyfe, unpublished data). Therefore, the gel- loaded materials for Western blot not only contained cubilin and AMN, but also contained many unspecific proteins. Three bands between 37 kDa and 50 kDa were seen in the lane blotted against the affinity-purified anti-AMN antibody, but absent in all the other control lanes, such as preimmune serum, flowthrough of the peptide-affinity- column-absorbed-serum, or secondary Ab only (Figure 5.2). Independent study indicated that the highest band represented the full-length AMN, while the two lower bands represented two truncated AMN isofomis (Fyfe, unpublished data). Thus the antibody is specific to the epitope (residues 340-353) of canine AMN. 98 Pull down assay and Western Blot We first did anti—AMN Westem blot with the kidney cortex homogenate from a nomtal dog but failed to detect any specific bands, indicating that the AMN protein was expressed at a low level in vivo (data not shown). We thus used the IF-Cbl beads to indirectly pull down the AMN via cubilin, because cubilin has been shown to form a tight complex with AMN (Fyfe et al., 2004). Western blot ofthe pulled-down proteins with the anti—canine AMN peptide serum (Figure 5.3, upper left) demonstrated three immunospecific proteins between 37 kDa and 50 kDa in the normal dog. None of the 3 immunoreactive proteins were observed on Western blots of proteins from kidney cortex of an affected dog of the GS kindred or the AS kindred (Figure 5.3, upper middle and upper right). In order to test if the beads were indeed saturated with cubilin during the first pull-down, we incubated the supernatant of the pull-down assay with lF-Cbl beads for the second time, and subjected the material bound to the beads to Western blotting with anti-cubilin. The cubilin band was present in all the three samples, indicating that the beads in the first pull-down assay were saturated with cubilin in both normal and affected dogs (Figure 5.4). Renal proximal tubule expression of cubilin in the absence of detectable AMN expression was also examined by irnmunohistochernistry. As previously demonstrated (Bim et al, 2000; Kozyraki et al, 1999), cubilin immunoreactivity was found predominantly at the apical brush border of the plasma membrane in normal dog proximal tubule epithelial cells (Figure 5.3. lower left). In contrast, cubilin immunoreactivity was found entirely in an intracellular vesicular pattern, with no 99 detectable labeling ofthe lurninal epithelial cell plasma membrane in proximal tubules, in affected dogs of both kindreds (Figure 5.3, lower panel). Transfection study and innnunofluorescence To further investigate the effect of the canine AMN e.1113_1 145del mutation on cubilin expression, we cloned the full-length nomial and mutant canine AMN open reading frames into a previously described AMN-Mye expression plasmid (Tanner et al., 2003) and transfected Chinese hamster ovary (C HO) cells that already expressed a truncated construct of rat cubilin containing the membrane association and IF-cobalamin binding domains (Fyfe et al., 2004). In our efforts to clone the PCR products ofthe canine eAMN into the expression vector, a high mutation rate (>0.002) was observed by directly sequencing the recombinant plasmids. The mutations were deemed to be due to the Taq polymerase. because different plasmid clones showed different mutations while sequencing the total PCR products didn’t reveal any mutation. Interestingly, we noticed that about 65% of the mutations in the plasmids were G/C>A/T, indicating that the polymerase tended to incorporate more A/T into the products than it should be. Considering that the CAMN is GC-rich (78%), we decided to increase the GC content of the dNTP mix from 50% to ~70% for PCR ((G/C):(A/T)=2.5: l ). In addition, the ExpandR~ high fidelity PCR system (Roche) was used to decrease the error rate of the DNA polymerase. Consequently, the mutation rate was decreased to about 0.001 and the target plasmid clones were obtained. Stable double transfectants were selected and incubated with anti-cubilin antibody. The cells were subsequently incubated with fluorescently labeled secondary antibody, 100 and analyzed by confocal microscopy. Cells transfected with normal canine AMN cDNA exhibited surface expression of cubilin, but cells transfected with the mutant AMN cDNA did not (Figure 5.5). Presence ofthe normal and mutant AMN proteins in their respective cell lines was confimied by Western blot of cell homogenates (Figure 5.6, upper panel). Cubilin in each cell line was examined by Western blot. In cells expressing wildtype canine AMN, cubilin was detected as 2 distinct bands (Figure 5.6, lower panel), the larger of which was endo H resistant and the smaller of which was endo H sensitive (data not shown). In cells expressing the cl 1 13-1 145del mutant AMN, cubilin was detected as a single, endo H sensitive band that comigrated with the smaller species observed in the wildtype AMN cell line. 101 Control Figure 5.1 RNA expression profiles of cubilin and AMN. Full-length AMN and CUBN are co-expressed in kidney and ileum, and to a minor extent injejunum and thymus. C yclophilin D was used as a control. In the ileum, at least 2 minor bands were seen. Sequencing one ofthem revealed a product missing exon 9 and 10 exactly, indicating it is an alternative splicing product. .. ‘ ‘ SOkDa I 1 4 37kDa II 4 25kDa Figure 5.2 Specificity test of the anti-AMN antibody. Affinity-eolumn-eoncentrated cubam was subjected to reducing SDS-PAGE (12%) and Western blot with different antibodies or serum. It should be pointed out that the concentrated cubam still contained many unspecific proteins because of technical limits. Three bands are present in the anti- AMN lane (lane 2), but absent in the other three control lanes. Lane 1, fiowthrough of the affinity-column-absorbed-serum (1:40,000); Lane 2, anti- AMN (1:40,000); Lane 3, secondary antibody only (l:20,000); Lane 4, preimmune serum (l:20,000). AMN detected on blots of IF-cobalamin pull- downs F tam ‘37 ' 5;! g i . -. .‘ . ‘ ‘.. I ‘ ’ y u n ’.:‘ {:5 .5 ‘1’. ;,‘ 5 ‘l '.. ,- "’5 ,"" 9. t ' 3‘ t i -i‘ .. -' ~ ' - " . ' t- . o o ’. ’ ' . up. n 5 ‘ I 'c t 5 Cllbllln _,.I .4’ I .. ,' V.” .I‘.. q. .1 ,1 . c" immunoreactivity in l 11 ,9 , ' . '1 “1‘ '3 1‘5 v-, r ' ‘ ' ' ._ . o - ,‘ l . . k ‘4’ ' .1." renal proxnmal tubules . . r. .. , . . _ ' “firm“. . 9} .. I‘ ~_. I. 3 ‘2‘: :2 ‘q , ’ r; - t. It ,Q . Normal Giant Schnauzer Aus. shepherd Figure 5.3 Cubilin and AMN expression in normal and I-GS affected dog kidneys in vivo. Images in this dissertation are presented in color. Upper panel: Anti-AMN Western blot of proteins pulled-down by rat IF-Cbl beads from normal (lefi), GS kindred affected (center), and AS kindred affected (right) dog kidneys. Three AMN isoforrns were seen in the normal dog (Iefi), but absent in the affected dogs of both kindreds (middle and right). Lower panel: Cubilin staining of perrneabilized renal cortex in normal (left) and I-GS affected dogs of the GS (center) and AS (right) kindreds expressing 33bp-de1AMN mutant and G>A AMN mutant, respectively. Cubilin showed surface staining in the normal dog, but abnormally intracellular staining in the affected dogs of both kindreds. 104 2041a)" 1231a)“ 80km” 48km» Figure 5.4 Cubilin saturation test. The second pulled-down materials by IF-Cbl beads (described in the Materials and Methods) were subjected to SDS-PAGE (7.5%) and blotted against anti-cubilin. The cubilin band was present in all the three samples, indicating that the beads in the first pull-down were saturated by cubilin. M, marker; Lane 1, a normal dog; Lane 2, an affected dog from GS kindred; Lane 3, an affected dog from AS kindred. 105 WT AMNcDNA 33del AMNcDNA + mini-cubilin + mini-cubilin Transfected CHO 5 Normal Mutant (33del) Figure 5.5 Immunofluorescence of double transfected CHO cells expressing cubilin and AMN (the cells were not permeabilized). CHO cell lines expressing a ‘mini-cubilin’ construct of rat origin were additionally transfected with wildtype (WT) or c.1 113- ] l45del (mut) canine AMN cDNA constructs, and stable transfectants expressing cubilin and AMN were selected. Non-permeabilized cells were stained for fluorescence confocal microscopy by incubation at 4°C with anti-cubilin and subsequently with labeled anti- rabbit IgG. Abundant surface cubilin staining was observed in cells expressing wildtype, but not mutant AMN. Images in this dissertation are presented in color. 106 WT Mutant 64 ‘AMN-myc 50 “ WT Mutant 250 '- r 4- Cubilin (mature form) ‘ ‘ <' Cubilin (immature form) 98 - Figure 5.6 Cubilin and AMN expression in CHO cells with wildtype or cl 1 13-1145del AMN cDNA. Note the CHO cells already contain a mini-cubilin construct. Upper panel: Lysates of double transfectant cell lines were subjected to Western blotting with anti-myc confirming expression of wildtype and mutant AMN. Lower panel: Lysates of double transfectant cell lines were subjected to IF -Cbl pull-down and Western blotting with anti-cubilin. The upper band is Endo H resistant (data not shown) and observed only in cells expressing wild type AMN. The lower band is Endo H sensitive (data not shown) and seen in both cells. Endo H sensitivity usually means that the protein has not undergone Golgi-mediated oligosaccharides processing. The results suggest that cubilin could not complete the Golgi-medicated glycosylation without the assistance of AMN. 107 Discussion I-GS is a unique, recessively inherited disorder characterized by selective intestinal cobalamin malabsorption and proteinuria due to loss of specific receptor functions in intestinal and renal proximal tubule epithelia. Studies in human cadaver kidney suggested that the functional receptor for IF-cobalamin in intestine and for multiple protein ligands in renal proximal tubules is a complex of cubilin and AMN, now called cubam (Fyfe et al., 2004). Although more than 10 different mutations have been identified in the two genes. no studies of CUBN or AMN expression in human I-GS patient tissues have been reported. Our data confirmed that both AMN and cubilin have ample mRNA expression in kidney and intestine, where malfunctions of cubam cause detectable defects. However, the tissues that have AMN but not cubilin expression, such as liver, pituitary and pancreas, are all somehow involved in protein secretion. It is tempting to think, although very preliminary at this stage, that AMN may assist some proteins’ secretion in those tissues. In our pull-down assay, 3 isoforrns of the AMN proteins were found in the normal kidney, but absent in the two affected dogs’ kidney. Independent experiments by LC/MS/MS mass spectrometric mapping and Western blot indicated that the highest Mr band represents the full-length AMN, while the two smaller bands represent C-terminal truncated AMN (Fyfe, unpublished data). Our second pull-down assay assured that the beads of the first pull-down were saturated with cubilin. Previous data (Fyfe et al., 1991b) also indicates that renal cubilin of GS kindred affected dogs has normal affinity (Kd 21x10“ M) for IF-Cbl. Thus, the absence of AMN bands in the affected dogs was 108 not due to insufficient cubilin in the kidney homogenates, but due to lack of the cubam complex in the kidney homogenates. In conclusion, the pull—down assay demonstrates that no cubam is expressed in the kidney proximal tubule cells ofthe two affected dogs. An interesting fact ofthe canine I-GS disease is that cubilin is affected by mutations in AMN. It was demonstrated that cubilin failed to reach the apical microvillus surface membrane ofvillus tip enterocytes (Fyfe et al., 1991a). The specific activity of cubilin in the affected dogs’ brush border membrane (BBM) was significantly less in ileum and renal cortex than in BBM ofthe corresponding tissue ofnorrnal dogs. In addition, cubilin in kidney of affected dogs was endo H sensitive, indicating that it had not undergone Golgi—mediated N-glycosylation processing, and peptide mapping by trypsin and elastase suggested that the protein was incompletely folded (Fyfe et al., 1991b; Xu and Fyfe, 2000). Those in viva findings were consistent to the results presented here that cubilin expressed in CHO cells with AMN cl 1 13-1 l45del did not reach the plasma membrane and. as indicated by cubilin endo H sensitivity, was retained in an early biosynthetic compartment. The effects of the canine AMN mutation on cubilin expression in this heterologous expression system were identical to those observed when no AMN was expressed (Fyfe et al., 2004). When overexpressed in CHO cells, the AMN deletion mutant protein was detected but was apparently incapable of fomiing a required quaternary interaction with cubilin. Combined with the irnrnunohistochemistry data, these findings are compatible with an active quality-control function of the RER in proximal tubules and suggest that cubilin and AMN must interact together in the RER for protein stability, to become competent for RER exit, and for efficient apical membrane cubam expression. 109 We then ask, in the pull-down assay, what is the molecular basis for the absence of the three AMN bands in the two affected dogs? For the AS kindred carrying the start codon mutation, it is very likely that the AMN protein was not translated at all, due to loss of the translation initiation site. However, for the GS kindred carrying the 33 bp deletion, two hypotheses are currently in consideration. The first hypothesis is that the 33del AMN mutant is simply secreted out of the proximal tubule cells. The second hypothesis is that the 33del mutant did not fold properly and was targeted for rapid degradation. Based on the current paradigm of protein biosynthesis, a possible scenario is that the 33del mutant completely enters the lumen of ER (because it loses its anchor domain), then binds to immature cubilin but fails to help cubilin fold correctly, which results in rapid degradation. Previous experiments demonstrated that the IFCR activity in the kidney homogenates of affected dogs was significantly less than that in the normal dogs (Fyfe et al., 1991b). A pulse-chase labeling study in kidney slices indicated that irnrnunoprecipitable renal cubilin of affected dogs had a shortened half-life than that ofa normal dog (Fyfe, unpublished data), suggesting that the rnisfolded cubilin was subject to degradation. However, these facts do not exclude either of the two hypotheses. Further experiments are needed to make more conclusive statements. Unfortunately for this investigation, none of the anti-dog, human or mouse AMN peptide antibodies produced to date function in immunohistochemical detection of AMN in canine tissue. The three isofomis of AMN could originate from altemative splicing products, or be created by a posttranslational mechanism. Another possibility is that they are simply due to ex vivo proteolysis. Interestingly, AMN isofonns were also observed in the IF-Cbl affinity purified materials from human kidney homogenates (Fyfe et al., 2004). Further 110 study with the CH0 transfectants using anti-AMN antibody instead of anti-Mye antibody may provide some clues. In mice, AMN is expressed exclusively in an extra-embryonic tissue, visceral endoderm, during the early post-implantation stages. Amn knockout mice fail to assemble a nomral middle primitive streak and have developmental arrest after E7.0 (Tomihara- Newberger et al., 1998). This phenotype is strikingly different from the phenotypes in humans and dogs. Though unknown as yet, cubilin and AMN likely collaborate to perform an essential endocytic function in rodent embryonic tissue that is not required or is performed by redundant mechanisms during human and canine embryogenesis. On the other hand, the three species do share some traits in common. Recently studies with Amn’ ES cell <—>+/+ blastocyst chimeras demonstrated that cubilin is not properly localized to the cell surface in Amn‘5' tissues in the embryo and adult mouse (Strope et al., 2004), suggesting that in the enterocytes and proximal tubules cubam is a complex conserved across the mammal species. Because AMN and CUBN patients have indistinguishable phenotypes, AMN as an accessory factor may solely serve cubilin for its expression on the cell surface in intestine and kidney. Such a “receptor-accessory factor” concept has been recapitulated in another study, in which melanocortin 2 receptor accessory protein (MRAP), a 19-kDa single-transmembrane domain protein, is responsible for the familial glueocorticoid deficiency (FGD) which can be caused by mutated melanocortin 2 receptor as well (Metherell et al., 2005). In sum, in both kindreds, effectively null AMN mutations block trafficking of cubilin to the apical plasma membrane and, thereby, all cubam-mediated endocytic lll functions, confirming a previous prediction by Xu et al. (1999) that an accessory factor is required for cubilin’s surface expression and function. Although some questions have been answered in this study, many questions remain to be explored in the future. For example, what is the fate ofthe 33del mutant AMN? If stable. what is the structure ofthe 33del mutant AMN? What happens ifAMN alone is expressed in the CHO cells? Can we ascertain the essential role of AMN in dog embryonic development? Are there any other roles of AMN in tissues where no cubilin is expressed? If yes, what are those partners? We hope this canine model of I-GS will continue to help us understand the function of AMN, and hopefully, shed light on the developmental biology. Appendices A brief report of mutation screening in a beagle and a komondor dog with I-GS In addition to GS and AS kindreds, I-GS has also been found in two other breeds of dogs. One of them is a beagle (Shamus) showing typical symptoms of congenital selective cobalamin (Cbl) malabsorption, including failure to thrive, methylmalonic aciduria and anemia. The dog had low serum Cbl concentration. Periodic injections of parenteral vitamin B12 corrected the Cbl deficiency and obliterated all the clinical abnormalities (Fordyce et al., 2000). The other one is a komondor (K101) with similar clinical presentations (Fyfe, unpublished data). Without fresh tissues to extract mRNA, we isolated blood DNA from the two affected dogs. Genomic PCR were perfomred as described in chapter 3. Most exons, exon-intron boundaries were amplified and sequenced. Sequences of partial exon 6 and partial exon 10 were not obtained in both dogs either because of failure of the PCR or poor quality of the sequencing results. No mutations were identified. A noteworthy fact is that the 33 bp deletion in GS kindred and the G>A mutation in AS kindred were not observed in either of the two dogs. A 5 bp deletion was found in both dogs compared to an unrelated nonnal dog (DCCU6110). The deletion is located at the 3’ end of the intron 6, but the accurate deletion sites could not be determined because of the presence of two flanking AG dinucleotide (Figure 6). Two lines of evidence suggest that the deletion is not a mutation. First, the deletion does not change the conserved acceptor sequence (N(C/T)AG I G) for splicing (Mount, 1982). Second, the 5 bp deletion was also present in a normal komondor 113 dog (K100). We therefore concluded that the 5 bp deletion is a nondeleterious polymorphism. Allele I. gctcaccccgcttectccgcggtggagGTGCAGCCCTGGATCTGCGCGGCCC Allele II. gctcaccccgcttcctccgc ------- agGTGC AGC CC TGGATC TGC GC GGCCC Figure 6. The 5 bp polymorphism of AMN identified in a komondor and a beagle. The two ag dinucleotide are underlined. The lower case letters represent partial intron 6, while the capital letters represent partial exon 7 of AMN. At least three possibilities should be considered in future study. First, the mutation(s) may locate in the exon 6 or exon 10 for which we didn’t get complete sequences due to technical difficulties. Second, the unsequenced introns, 5’UTR or 3’UTR of AMN may harbor the mutation(s), which may introduce defects by altering the mRNA splicing product. Third, cubilin or another gene may be the disease-causing gene. RT-PCR and linkage analysis will help to clarify these possibilities. 114 Table 4 Genotyping data ofthe Giant Schnauzer family. In the “KNS2” column, the shaded genotypes were deduced from genotypes or haplotypes of their first-degree relatives. Notationszl) 1n the “I-GS” column: N, Nomial; C, Carrier; A, Affected. 2) “?” represents unknown or uncertain genotype. DOG 1D Sire Dam l-GS Stonin EML1 EIF5 KNS2 SIVA G2A F100 A TT Fl 14 C CT M721 N CC M811 N CT M874 M721 M811 N CT GG? CC GG F70 A CC AA F146 F114 F100 A CC AG AA TT CT AG F150 F70 A66 C CC GG TC CT AA F153 F70 A66 C F154 F70 A66 C CC AG F155 F70 A66 C CC AA F227 F150 F146 A CC GG TT CC AA F233 F150 F155 A CC TT F274 A323 F233 C CT AG AT TC CT AG F275 A323 F233 C CC AG TC AG F283 F150 F146 A CC AG TT CT AG F284 F150 F146 A CC GG AA TT CC AA F294 F227 F154 A CC AA 303 F227 F275 C AG AG F304 F227 F275 C AG AG F305 F227 F275 C GG? AG _F306 F227 F275 C AG AG F307 F227 F275 C AG AG F309 F274 F146 A CC F310 F274 F146 A CT F31 1 F274 F146 A CT F324 F274 F284 A CC GG TT AA i325 F274 F284 C CT AG TC AG __F326 F274 F284 C CT AG TC AG £27 F274 F284 A CC GG TT CC AA fl F274 F284 A CC GG TT AA F329 F274 F284 A CC GG TT AA fl F274 F284 C CT AG TC AG i3; F227 F275 C AG AG .13; IT “F3443 F294 F146 A CC GG TT AA fl F283 F275 A 115 Table 4 (cont’d) DOGID Sire Dam l-GS StoninlEMLl EIF5 KNSZ SIVA G2A F349 F283 F275 A TT F350 F283 F275 c F351 F283 F275 c F352? F283 F275 c F353 F283 F275 c F354 F283 F275 c F355 F283 F275 c F358 F274 F284 A cc GG TT AA F359 F274 F284 A cc GG TT AA F360 F274 F284 A cc GG TT AA F361 F274 F284 A cc GG TT AA F362 F274 F284 A cc GG TT AA F363 F274 F284 c CT AG CT AG F366 F274 F146 A cc F367 F274 F146 A cc F368 F274 F146 c CT F369 F274 F146 A CT F370 F274 F146 A cc F372 F274 F146 c CT F373 F274 F146 c F374 F274 F146 A CT F375 F274 F146 c CT F380 F274 F284 c F381 F274 F284 A CT GG TT AA F382 F274 F284 A cc GG TT AA F384 F274 F284 A cc GG TT CT AG 388 F274 F327 c cr AG? F389 F274 F327 c. cc AG AT CT AG F390 F274 F327 A cc GG TT AA F398 F274 F343 A cc GG TT AA F399 F274 F343 c CT AG er AG F400 F274 F343 A CT GG TT AA F411 F342 F343 TT F414 F342 F343 TT F415 F342 F343 TT 7428 F274 F284 A cc GG TT AA TF442 F274 F284 c CT AG CT AG F445 F342 F343 A cc AG TT CT AG 7715446 F342 F343 TT 116 Table 4 (cont’d) DOG 1D Sire Dam I-GS Stonin EML1 EIF5 KNSZ SIVA G2A F451 F274 F327 C CT AG CT AG F453 F274 F327 A CT AG AA TT AA F454 F274 F327 A CC GG TT AA F455 F274 F327 A CC GG TT AA F458 F274 F284 A CC GG TT AA F459 F274 F284 A CC GG TT AA F460 F274 F284 A CT AG AA TT AA F461 F274 F284 A CC GG TT AA F462 F274 F284 A CC GG TT AA F463 F274 F284 C CT AG CT AG F470 F445 F343 TT F472 F445 F343 TT F477 F274 F284 A CC AA F478 F274 F284 C CT AG CT AG F481 F274 F442 A CC GG AA F482 F2 74 F442 ? CT AG F483 F2 74 F442 ? CC AG F484 F2 74 F442 ? CT AG F485 F2 74 F442 ? CT AG AG F497 F445 M874 C CT AG TC GG F498 F445 M874 C CC GG TC AG F499 F445 M 874 C CT GG TC GG F500 F445 M874 C CT AG TC GG F501 F445 M874 C CC GG TC F505 F274 F442 A CC F 506 F2 74 F442 ? CC AG F507 F 274 F442 ? TT AG F508 F274 F442 ? ¥F509 F 274 F442 ? CT AG i510 F274 F442 ? _FS_1 1 F274 F442 ? F512 F274 F442 ? F532 F445 F497 C CT fl: F445 F497 C CT iii F445 F497 A CC @536 F445 F497 A CC F537 F445 F497 A CC AA F538 F445 F497 C CT F539 F445 F497 C CT GG F540 F445 F497 C CT 117 Table 5 Genotyping data of the Australian Shepherd kindred. The notation “c.normal” represents “clinically normal”. Lab 1D NAME Sex Disease status AMN KNS2 CUBN B100 Buddy Henry M affected AA CC 11 B101 H. Floxy F carrier AG CT 10 B102 H. Boo M carrier AG CT 11 B103 H. Bits of Clover F c. normal GG B104 Twiggy F c. normal AG B105 H. Red Chili M c. normal GG B106 H. Tux M c. normal GG B107 H. Cozy Kitten F c. nomial GG B108 H. Keeper F c. nomial AG CC 10 B109 H. Blue Clipper M c. nomral GG CT 10 B1 10 Rodeo M c. normal GG Bl 1 1 South Twenty's R. F c. normal GG BI 12 H. Kiss of Rain F c. nomial GG CT 00 B1 13 Champ M c. normal BI 14 H. Cheerio F c. normal B1 15 Snoopy F c. normal AG BI 16 H. Zackery M c. normal GG BI 17 H. Banjo M c. nomial GG CT 11 B1 18 CR Poke Weed M c. normal TT 10 B1 19 Trevor M c. normal GG B120 H. Cheshire Cat F e. normal TT [0 B 1 21 Sidney Daniels M affected AA CC 11 B 122 Maddie Martin F affected AA CC 10 B123 Sassy b: 6-30-03 F c. normal GG B124 Angel F c. normal GG B125 Pixie b: 1-20-04 F c. normal GG 8126 Karri b: 1-3-04 F c. normal AG 8127 H. Lacey Lady F c. normal AG 118 FUTURE DIRECTION The beagle and the komondor dog with l-GS will be screened more extensively for potential AMN mutations. At the same time, complementation tests (breeding tests) could be done to ascertain that AMN is the disease-causing gene. CUBN will be considered for mutation screening too. Should AMN and CUBN be ruled out, the komondor dog family will be expanded for a positional cloning project. The AMN peptide antibodies currently in our hands only worked under reducing conditions in Western blot, and do not work in the imrnunohistochernistry. Therefore, we need to develop new antibodies. More experiments could be done to characterize the 33del mutant AMN. Directly assessing the existence of the mutant AMN in the affected dogs’ urine may be difficult, because the amount ofthe AMN is probably too low. Instead, the cell culture medium of the CH0 double transfectants should be examined to see if the mutant AMN is secreted out ofthe cell. Crystallization could be pursued to study the structure of the normal and mutant AMN. Current data suggest that cubilin binds to the N-terminal of AMN, but the exact binding site is unknown. Different AMN cDNA mutants will be constructed in vitro and tested in the CH0 cell system to accurately locate the binding site of AMN to cubilin. These mutants may also help to identify the minimal length of AMN’s N-temiinus that is required for cubilin’s folding and surface expression, and the minimal length of AMN’s C -terminus for endocytosis. Should new anti-AMN antibody be available, the CHO cells transfected with wild type cubilin and wild type AMN-Smyc will be subject to Westem blot with anti-AMN 119 antibody. If three isofonns are seen as those in the kidney homogenates, further characterizations, such as cubilin-binding ability, glycosylation, cellular localization, will be pursued for the AMN isofomis. It is not known whether that AMN’s surface expression requires cubilin. Single AMN-transfectant of CHO will be subjected to both Westem blot and immunofluorescence surface staining, to detemtine the molecular weight and location of AMN in the absence of cubilin. Most ofthe studies of cubam so far were performed with kidney tissues. It may be necessary to directly address the role of cubam in the intestine tissues, where the vitamin BIZ malabsorption actually occurs. Affinity chromatography purification, IF-Cbl pull down, Westem blot, mass spectrometry mapping may be used to characterize the intestinal cubam complex. In the tissues where AMN is expressed but cubilin is not, a yeast two hybrid system could be used to look for potential partners for AMN. Briefly, the cDNA ofAMN would be cloned into the vector pGBKT7 (Clontech) as the bait, and mRNA from the target tissue, say, pancreas, would be reverse—transcribed into double-stranded cDNA. Subsequently the pGBKT7-bait, cDNA pool and another vector pGADT7-Rec (Clontech) are cotransforrned into the yeast for screening. Once new partners have been identified, at least three approaches should be considered. First, it should be confirmed that the new protein interacts with cubilin/AMN in vitro. Second, the interaction of the new protein and cubilin/AMN should be examined in appropriate cell-based studies. Third, knock out mouse ofthe new protein could be produced. Ifthe phenotype is embryo lethal, chimera mouse may be made to determine the function ofthe gene. 120 To assess the potential role of AMN or cubilin in the embryo development, extra- embryonic tissues (including visceral endodenn and yolk sac) of dogs would be screened for AMN and cubilin expression at RNA level and protein level. In the tissues that have both AMN and cubilin expressions, IF-Cbl pull down assay would be used to detect if cubam exists. If cubam does present in the tissue ofinterest, it would be intriguing to find out the partners of AMN, since the cysteine-rich domain of AMN has been suggested to be interacting with some BMP inhibitors. These experiments may give us more insights ofthe signaling pathways involved in embryo development. REFERENCES Abdelaal MA, Ahmed AF. Imerslund-Grasbeek syndrome in a Saudi family. Acta Paediatr Scand. 1991; 80(11):l 109-12. Al-Alami JR, Tayeh MK, Al-Sheyyab MY, El-Shanti HI. Linkage analysis ofa large inbred family with congenital megaloblastic anemia. Saudi Med J. 2002; Z3( lO):1251-6. Allen RH, Majerus PW. Isolation of vitamin B12-binding proteins using affinity chromatography. II. Purification and properties of a human granulocyte vitamin BlZ- binding protein. J Biol Chem. 1972a; 247(23):7702-8. Allen RH, Majerus PW. Isolation of vitamin BIZ-binding proteins using affinity chromatography. 3. Purification and properties of human plasma transcobalamin 11. J Biol Chem. 1972b; 247(23):7709-17. Allen RH, Mehlman CS. Isolation of gastric vitamin B 12 -binding proteins using affinity chromatography. 1. Purification and properties ofhuman intrinsic factor. J Biol Chem. 1973; Z48(10):3660-9. Allen RH, Seetharam B, Podell E, Alpers DH. Effect of proteolytic enzymes on the binding of cobalamin to R protein and intrinsic factor. J Clin Invest. 1978a; 61(1):47-54. Allen RH, Seetharam B, Allen NC, Podell ER, Alpers DH. Correction of cobalamin malabsorption in pancreatic insufficiency with a cobalamin analogue that binds with high affinity to R protein but not to intrinsic factor. J Clin Invest. 1978b; 61(6):]628-34. Altay C, Cetin M, Gumruk F, Irken G, Yetgin S, Laleli Y. Familial selective vitamin BIZ malabsorption (Imerslund-Grasbeck syndrome) in a pool of Turkish patients. Pediatr Hernatol Oncol. 1995; 12(1):19-28. Amagasaki T, Green R, Jacobsen DW. Expression of transcobalamin 11 receptors by human leukemia K562 and HL-60 cells. Blood. 1990; 76(7):]380-6. Aminoff M, Tahvanainen E, Grasbeck R, Weissenbach J, Broch H, de la Chapelle A. Selective intestinal malabsorption of vitamin B12 displays recessive mendelian inheritance: assignment of a locus to chromosome 10 by linkage. Am J Hum Genet. 1995; 57(4):824-31. AminoffM. Carter JE, Chadwick RB, Johnson C, Grasbeek R, Abdelaal MA, Broch H, Jenner LB, Verroust PJ, Moestrup SK, de la Chapelle A, Krahe R.Mutations in CUBN, encoding the intrinsic factor-vitamin BIZ receptor, cubilin, cause hereditary megaloblastic anaemia 1. Nat Genet. 1999; 21(3):309-13. Argraves WS, Morales CR. lmmunolocalization of cubilin, megalin, apolipoprotein J, and apolipoprotein A-l in the uterus and oviduct. Mol Reprod Dev. 2004; 69(4):419-27. lZZ Arwert F, Porck HJ, Frater-Schroder M, Brahe C, Geurts van Kessel A, Westerveld A, Meera Khan P, Zang K, Frants RR, Kortbeek HT, et a1. Assignment of human transcobalamin II (TCZ) to chromosome 22 using somatic cell hybrids and monosornic meningioma cells. Hum Genet. 1986; 74(4):378-81. Babior BM. Cobalamin Biochemistry and Pathophysiology. 1975; Pg3-Pg14. By John Wiley&Sons, Inc. Banerjee R and Ragsdale SW. The many faces of vitamin B12: catalysis by cobalamin- dependent enzymes. Annual Review of Biochemistry. 2003; 72: 209-247. Baturnan V, Verroust PJ, Navar GL, Kaysen JH, Goda FO, Campbell WC, Simon E, Pontillon F, Lyles M, Bruno J, Hammond TG. Myeloma light chains are ligands for cubilin (gp280). Am J Physiol. 1998; Z75(Z):F246-54. Becker A, Geiger D, Schaffer AA. Automatic selection of loop breakers for genetic linkage analysis. Hum Hered. 1998; 48(1):49-60. Becker M, Rotthauwe HW, Weber HP, Fischbach H. Selective vitamin B12 malabsorption (Imerslund-Grfisbeck syndrome). Studies on gastroenterologieal and nephrological problems. EurJ Pediatr. 1977; lZ4(Z):l39-53. Ben-Ami M, Katzuni E, Koren A. Imerslund syndrome with dolichocephaly. Pediatr Hematol Oncol. 1990; 7(2):177-81. Ben-Bassat l. Feinstein A, Rarnot B. Selective vitamin B 12 malabsorption with proteinuria in Israel: clinical and genetic aspects. Ier Med Sci. 1969; 5(1):62-8. Bienzle U, Olischlager A, Leupold D, Kohne E, Hamisch R, Kleihauer E. Megaloblastic anaemia in childhood due to vitamin BIZ deficiency, report of3 cases of congenital selective vitamin BlZ malabsorption. Klin Padiatr. 1976; 188(2):97-103. Bim H, Verroust PJ, Nexo E, Hager H, Jacobsen C, Christensen El, Moestrup SK. Characterization of an epithelial approximately 460-kDa protein that facilitates endocytosis of intrinsic factor-vitamin BIZ and binds receptor-associated protein. J Biol Chem. 1997; Z72(4Z):Z6497-504. Bim H, Fyfe JC, Jacobsen C, Mounier F, Verroust PJ, Orskov H, Willnow TE, Moestrup SK, Christensen E1. Cubilin is an albumin binding protein important for renal tubular albumin reabsorption. J Clin Invest. 2000; 105(10):1353-61. Boll W, Rapoport I, Brunner C, Modis Y, Prehn S and Kirchhausen T. The mu 2 subunit ofthe clathrin adaptor AP-Z binds to FDNPVY and YppO sorting signals at distinct sites. Traffic. 2002; 3(8): 590-600. 123 Booth CC, Mollin DL. The site ofabsorption of vitamin BIZ in man. Lancet. 1959; 1(7062):18-21. Breen M, Thomas R, Binns MM, Carter NP, Langford CF. Reciprocal chromosome painting reveals detailed regions of conserved synteny between the karyotypes of the domestic dog (Canis familiaris) and human. Genomics. 1999; 61(2):145-55. Breen M, Jouquand S, Renier C, Mellersh CS, Hitte C, Holmes NG, Cheron A, Suter N, Vignaux F, Bristow AE, Priat C, McCann E, Andre C, Boundy S, Gitsham P, Thomas R, Bridge WL, Spriggs HF, Ryder EJ, Curson A, Sampson J, Ostrander EA, Binns MM, Galibert F.Chromosome—specific single-locus FISH probes allow anchorage of an 1800- marker integrated radiation-hybrid/linkage map of the domestic dog genome to all chromosomes. Genome Res. 2001; l 1(10):1784-95. Broch H, lrnerslund O, Monn E, Hovig T, Seip M. Imerslund-Grasbeck anemia. A long- term follow-up study. Acta Paediatr Scand. 1984; 73(2):Z48-53. Brunaud L, Alberto J M, Ayav A, Gerard P, Namour F, Antunes L, Braun M, Bronowicki JP, Bresler L, Gueant JL. Effects of vitamin BIZ and folate deficiencies on DNA methylation and carcinogenesis in rat liver. Clin Chem Lab Med. 2003; 41(8): 1012-9. Review. Bu G, Geuze HJ, Strous GJ, Schwartz AL. 39 kDa receptor-associated protein is an ER resident protein and molecular chaperone for LDL receptor-related protein. EMBO J. 1995; 14(10):2269-80. Bumtan JF, Jenkins WJ, Walker-Smith JA, Phillips AD, Sourial NA, Williams CB, Mollin DL. Absent ileal uptake of IF-bound vitamin B12 in vivo in the Imerslund- Grasbeck syndrome (familial vitamin BIZ malabsorption with proteinuria). Gut. 1985: Z6(3):311-4. Christensen E1, Gliemann J, Moestrup SK. Renal tubule gp330 is a calcium binding receptor for endocytic uptake of protein. J Histochem Cytochern. 1992; 40(10):]481-90. Christensen El and Bim H. Megalin and cubilin: multifunctional endocytic receptors. Nat. Rev. Mol. Cell. Biol. 2002; 3(4): 256-266. Cooper BA and Rosenblatt DS. Inherited defects ofvitamin BlZ metabolism. Ann. Rev. Nutr. 1987. 7: 291-320. Coopemtan JM, Luhby AL, Teller DN, Marley JF. Distribution of radioactive and nonradioactive vitamin BIZ in the dog. J Biol Chem. 1960; 2352191-4. Cottingham RW Jr, Idury RM, Schaffer AA. Faster sequential genetic linkage computations. Am J Hum Genet. 1993; 53(1):252-63. Crider-Pirkle S, Billingsley P, Faust C, Hardy DM, Lee V, Weitlauf H. Cubilin, a binding partner for galectin-3 in the murine utero-placental complex. J Biol Chem. 2002; Z77(18):15904-12. Dan N, Cutler DF. Transcytosis and processing of intrinsic factor-cobalamin in Caco-2 cells. J Biol Chem. 1994; 269(29):]8849-55. Dix CJ, Hassan IF, Obray HY, Shah R, Wilson G. The transport of vitamin B12 through polarized monolayers of Caco-2 cells. Gastroenterology. 1990; 98(5):1272-9. Dobson CM, Wai T, Leclerc D, Wilson A, Wu X, Dore C, Hudson T, Rosenblatt DS, Gravel RA. Identification of the gene responsible for the cblA complementation group of vitamin B(12)-responsive methylmalonic acidemia based on analysis of prokaryotic gene arrangements. Proc. Nat. Acad. Sci.USA 2002a; 99(24): 15554-15559. Dobson CM, Wai T, Leclerc D, Kadir H, Narang M, Lemer-Ellis JP, Hudson TJ, Rosenblatt DS, Gravel RA. Identification of the gene responsible for the cblB complementation group of vitamin B12-dependent methylmalonic aciduria. Hum Mol Genet. 2002b; 11(26):3361-9. Doscherholrnen A, Hagen PS. Delay of absorption of radiolabeled cyanocobalarnin in the intestinal wall in the presence ofintrinsic factor. J Lab Clin Med. 1959; 54:434-9. cl Mauhoub M, Sudarshan G, Aggarwal V, Banerjee G. Imerslund-Grasbeck syndrome in a Libyan boy. Ann Trop Paediatr. 1989; 9(3):180-1. Eudy JD, Ma-Edmonds M, Yao SF, Talmadge CB, Kelley PM, Weston MD, Kimberling WJ, Surnegi J. Isolation ofa novel human homologue ofthe gene coding for eehinodemt rnicrotubule-associated protein (EMAP) from the Usher syndrome type la locus at 14q32. Genomics. 1997; 43(1):104-6. Fenech M. The role of folic acid and Vitamin 812 in genomic stability of human cells. Mutat Res. 2001; 475(1-2):57-67. Review. Fordyce HH, Callan MB, Giger U. Persistent cobalamin deficiency causing failure to thrive in ajuvenile beagle. J Small Anim Pract. 2000; 41(9):407-10. Furuhjelm U, Nevanlinna HR. Inheritance of selective malabsorption of vitamin B12. Scand J Haematol. 1973; 11(1):Z7-34. F yfe J C , Jezyk PF, Giger U, Patterson DF. Inherited selective malabsorption of vitamin BIZ in giant schnauzers. Journal of the American animal hospital association. 1989; 25:533-539. Fyfe JC, Giger U, Hall CA, Jezyk PF, Klumpp SA, Levine JS, Patterson DF. Inherited selective intestinal cobalamin malabsorption and cobalamin deficiency in dogs. Pediatr Res. 1991a; 29(1):Z4-3l. Fyfe JC, Ramanujam KS, Ramaswamy K, Patterson DF, Seetharam B. Defective brush- border expression of intrinsic factor-cobalamin receptor in canine inherited intestinal cobalamin malabsorption. J Biol Chem. 1991b; 266(7):4489—94. Fyfe JC. Feline intrinsic factor (1F) is pancreatic in origin and mediates ileal cobalamin (cbl) absorption. J Vet Intern Med. 1993; 7:133. Fyfe JC, Madsen M, Hojrup P, Christensen El, Tanner SM, de la Chapelle A, He Q, Moestrup SK. The functional cobalamin (vitamin B12)-intrinsic factor receptor is a novel complex of cubilin and amnionless. Blood. 2004; 103(5):1573-9. Gasparini P, Nunes V, Savoia A, Dognini M, Morral N. Gaona A, Bonizzato A, Chillon M, Sangiuolo F, Novelli G, et al. The search for south European cystic fibrosis mutations: identification of two new mutations, four variants. and intronic sequences. Genomics. 1991;10(1):l93-ZOO. Gburek J, Verroust PJ, Willnow TE, Fyfe JC, Nowacki W, Jacobsen C, Moestrup SK, Christensen E1. Megalin and cubilin are endocytic receptors involved in renal clearance ofhemoglobin. J Am Soc Nephrol. 2002; 13(2):4Z3-30. Gburek J, Bim H, Verroust PJ, Goj B, Jacobsen C, Moestrup SK, Willnow TE, Christensen EI. Renal uptake of myoglobin is mediated by the endocytic receptors megalin and cubilin. Am J Physiol Renal Physiol. 2003; 285(3):F451-8. Gomez Lira M, Benetazzo MG, Marzari MG, Bombieri C, Belpinati F, C astellani C, Cavallini GC, Mastella G, Pignatti PF. High frequency ofcystic fibrosis transmembrane regulator mutation L997F in patients with recurrent idiopathic pancreatitis and in newborns with hypertrypsinemia. Am J Hum Genet. 2000; 66(6):2013-4. Gordon M M, Brada N, Remacha A, Badell 1, del Rio E, Baiget M, Santer R, Quadros E V, Rothenberg SP, Alpers DH. A genetic polymorphism in the coding region of the gastric intrinsic factor gene (GIF) is associated with congenital intrinsic factor deficiency. Hum. Mutat. 2004; 23(1): 85-91. Grasbeck R, Gordon R, Kantero 1, Kuhlback B. Selective vitamin BIZ malabsorption and proteinuria in young people. Acta Med. Scand. 1960; 167: 289-296. Gravel RA, Mahoney MJ, Ruddle FH, Rosenberg LE. Genetic complementation in heterokaryons of human fibroblasts defective in cobalamin metabolism. Proc. Nat. Acad. Sci.USA 1975; 72(8): 3181-3185. Gueant JL, Saunier M, Gastin I, Safi A, Lamireau T, Duclos B, Bigard MA, Grasbeck R. Decreased activity of intestinal and urinary intrinsic factor receptor in Grasbeck- Imerslund disease. Gastroenterology. 1995; 108(6):1622-8. Gueant JL, Chery C, Namour F. Cublin and the hydrophobic intrinsic factor receptor are distinct molecules. Blood. 2001; 97(10):3316-7. Gulati S, Baker P, Li YN, Fowler B, Kruger W, Brody LC, Banerjee R. Defects in human methionine synthetase in cblG patients. Hum Mol Genet. 1996; 5(12): 1 859-65. Gullberg R. Vitamin B 12 -binding proteins in normal human blood plasma and serum. ScandJ Haematol. 1972; 9(6):639-47. Guyon R, Lorentzen TD, Hitte C, Kim L, C adieu E, Parker HG, Quignon P, Lowe J K, Renier C, Gelfenbeyn B, Vignaux F, DeFrance HB, Gloux S, Mahairas GG, Andre C, Galibert F. Ostrander EA. A l-Mb resolution radiation hybrid map of the canine genome. Proc Natl Acad Sci U S A. 2003; 100(9):5296-301. Hammad SM, Stefansson S, Twal WO, Drake CJ, Fleming P, Remaley A, Brewer HB Jr, Argraves WS. Cubilin, the endocytic receptor for intrinsic factor-vitarnin B(lZ) complex, mediates high-density lipoprotein holoparticle endocytosis. Proc Natl Acad Sci U S A. 1999; 96(18):10158-63. Hammad SM, Barth J L, Knaak C, Argraves WS. Megalin acts in concert with cubilin to mediate endocytosis of high density lipoproteins. J Biol Chem. 2000; Z75(16):12003-8. Hewitt JE, Gordon MM, Taggart RT, Mohandas TK, Alpers DH. Human gastric intrinsic factor: characterization of cDNA and genomic clones and localization to human chromosome 11. Genomics. 1991; 10(2):43Z-40. Hippe E. Malabsorption of vitamin 812. Report ofa case in a 1-year-old boy, including studies ofthe absorption ofBlZ. Acta Paediatr Scand. 1966; 55(5):510-6. Hooper DC, Alpers DH, Burger RL, Mehlman CS, Allen RH. Characterization ofileal vitamin BIZ Binding using homogeneous human and hog intrinsic factors. J Clin Invest. 1973; 52(12):3074-83. Idriss JM, Jonas AJ. Vitamin BlZ transport by rat liver lysosornal membrane vesicles. J Biol Chem. 1991; 266(15):9438-41. Imerslund O. Idiopathic chronic megaloblastic anemia in children. Acta Paediat. Scand. 1960; 49: 1-115. Imerslund O and Bjomstad P. Familial vitamin BlZ malabsorption. Acta Haernat. 1963; 30:1-7. 127 Ismail EA, AI Saleh Q, Sabry MA, Al Ghanim M, Zaki M. Genotypic/phenotypic heterogeneity of selective vitamin BIZ malabsorption (Grasbeck-lmerslund syndrome) in two Bedouin families. Acta Paediatr. 1997; 86(4):4Z4-5. Israel DI. A PCR-based method for high stringency screening of DNA libraries. Nucleic Acids Res. 1993; 21(1 I)22627-31. Janata J, Kogekar N, Fenton WA. Expression and kinetic characterization of methylmalonyl-CoA mutase from patients with the mut- phenotype: evidence for naturally occurring interallelic complementation. Hum Mol Genet. 1997; 6(9):I457-64. Jensen H K, Jensen TG, Faergeman O, Jensen LG, Andresen BS, Corydon MJ, Andreasen PH, Hansen PS, Heath F, Bolund L, Gregersen N. Two mutations in the same low- density lipoprotein receptor allele act in synergy to reduce receptor function in heterozygous familial hypercholesterolemia. Hum Mutat. I997; 9(5):437-44. Kalantry S, Manning S, Haub O, Tomihara-Newberger C, Lee HG, Fangman J, Disteche CM, Manova K, Lacy E. The amnionless gene, essential for mouse gastrulation, encodes a visceral-endoderm-specific protein with an extracellular cysteine-rich domain. Nat Genet. 2001; 27(4):4IZ-6. Kapadia CR, Serfilippi D, Voloshin K, Donaldson RM Jr. Intrinsic factor-mediated absorption of cobalamin by guinea pig ileal cells. J Clin Invest. 1983; 71(3):440-8. Katz M, Cooper BA. Solubilized receptor for intrinsic factor-Vitamin BIZ complex from guinea pig intestinal mucosa. J Clin Invest. 1974a; 54(3):733-9. Katz M. Mehlman CS, and Allen RH. Isolation and Characterization ofan Abnomial Human Intrinsic Factor. J Clin Invest. 1974b; 53(5): 1274 1283. Kelley PM, Harris DJ, Comer BC, Askew JW, Fowler T, Smith SD, Kimberling WJ. Novel mutations in the connexin 26 gene (GJBZ) that cause autosomal recessive (DFNBI) hearing loss. Am J Hum Genet. 1998; 62(4):792-9. Kouvonen I. Solubilization ofthe pig ileal intrinsic factor receptor with papain treatment and studies on the solubilized receptor. Biochim Biophys Acta. 1980; 6Z6(1):Z44-53. Kozak M. Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell. 1986; 44(2):283-92. Kozak M. Recognition of AUG and alternative initiator codons is augmented by G in position +4 but is not generally affected by the nucleotides in positions +5 and +6. EMBO J. 1997; 16(9):2482~9Z. Kozyraki R, Kristiansen M, Silahtaroglu A, Hansen C, .lacobsen C, Tommerup N, Verroust PJ, Moestrup SK. The human intrinsic factor-vitamin BIZ receptor, cubilin: 128 molecular characterization and chromosomal mapping of the gene to 10p within the autosomal recessive megaloblastic anemia (MGAI) region. Blood. 1998; 91(10):3593- 3600. Kozyraki R, F yfe J, Kristiansen M, Gerdes C, Jacobsen C, Cui S, Christensen El, Aminoff M, de la Chapelle A, Krahe R, Verroust PJ, Moestrup SK. The intrinsic factor- vitamin B12 receptor, cubilin, is a high-affinity apolipoprotein A-I receptor facilitating endocytosis of high-density lipoprotein. Nat Med. 1999; 5(6):656-661. Kozyraki R, F yfe J, Verroust PJ, Jacobsen C, Dautry-Varsat A, Gburek J, Willnow TE, Christensen El, Moestrup SK. Megalin-dependent cubilin-mediated endocytosis is a major pathway for the apical uptake of transferrin in polarized epithelia. Proc Natl Acad Sci U S A. 2001; 98(22):12491-12496. Kristiansen M, Kozyraki R, Jacobsen C, Nexo E, Verroust PJ, Moestrup SK. Molecular dissection of the intrinsic factor-vitamin BIZ receptor, cubilin, discloses regions important for membrane association and ligand binding. J Biol Chem. 1999; 274(29)220540-4. Kristiansen M, AminoffM, Jacobsen C, de La Chapelle A, Krahe R, Verroust PJ, Moestrup SK. C ubilin P1297L mutation associated with hereditary megaloblastic anemia 1 causes impaired recognition ofintrinsic factor-vitamin B(12) by cubilin. Blood. 2000; 96(2):405-9. Lander E, Kruglyak L. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat Genet. 1995; 11(3):Z41-7. Lathrop GM, Lalouel JM, Julier C, Ott J. Strategies for multilocus linkage analysis in humans. Proc Natl Acad Sci U S A. 1984; 81(11):3443-6. Le Panse S, Ayani E, Mulliez N, Chatelet F, Cywiner-Golenzer C, Galceran M, Citadelle D, Roux C, Ronco P, Verroust P. Antibodies to the 280-kDa coated pit protein, target of teratogenic antibodies, produce alterations in the traffic of internalized proteins. AmJ Pathol. 1994; I45(6):1526-36. Leclerc D, Campeau E, Goyette P, Adjalla CE, Christensen B, Ross M, Eydoux P, Rosenblatt DS, Rozen R, Gravel RA. Human methionine synthetase: cDNA cloning and identification of mutations in patients of the cblG complementation group of folate/cobalamin disorders. Hum. Molec. Genet. 1996; 5(12): 1867-1874. Levine JS, Nakane PK, Allen RH. Immunocytochemical localization of human intrinsic factor: the nonstimulated stomach.Gastroenterology. 1980; 79(3):493-502. Levine J S, Allen RH, Alpers DH, Seetharam B. Immunocytochemical localization of the intrinsic factor—cobalamin receptor in dog-ileum: distribution of intracellular receptor during cell maturation. J Cell Biol. 1984; 98(3):l 1 11-8. Li N, Rosenblatt DS, Kamen BA, Seetharam S, Seetharam B. Identification of two mutant alleles oftranscobalamin II in an affected family. Hum Mol Genet. 1994; 3(10)21835-40. Lin SH, Sourial NA, Lu KC, Hsueh EJ. Imerslund-Grasbeck syndrome in a Chinese family with distinct skin lesions refractory to vitamin BIZ. J. Clin. Path. 1994; 47(10): 956-958. Lindblom A, Quadt N, Marsh T, Aeschlimann D, Morgelin M, Mann K, Maurer P, Paulsson M. The intrinsic factor-vitamin BIZ receptor, cubilin, is assembled into trimers via a coiled-coil alpha-helix. J Biol Chem. 1999; Z74(IO):6374-80. Ludvigsson P, Hassink SE, Bemstein M, Widzer S, Grover WD. Ataxia, cerebral atrophy, and abnormal somatosensory evoked potentials in a 9-year-old Boy with vitamin BIZ deficiency. Annals of Neurology. 1980; 8: 234 only. Mackenzie IL, Donaldson RM Jr, Trier JS, Mathan V1. Ileal mucosa in familial selective vitamin B 12 malabsorption. N Engl J Med. 1972; 286(19):]021-5. Marcoullis G, Grasbeck R, Salonen EM. Identification and characterization ofintrinsic factor and cobalophilin from pig ileal and pyloric mucosa. Biochim Biophys Acta. 1977; 497(3):663-7Z. Marcoullis G, Rothenberg SP, Labombardi VJ. Preparation and characterization of proteins in the alimentary tract of the dog which bind cobalamin and intrinsic factor. J Biol Chem. 1980; 255(5):1824-9. Marcoullis G, Rothenberg SP. Intrinsic factor-mediated intestinal absorption of cobalamin in the dog. Am J Physiol. 1981; 241(4):GZ94-9. Marsh EN. Coenzyme BIZ (cobalamin)-dependcnt enzymes. Essays Biochem. 1999; 34: 139-54. Review. Martens JH, Barg H, Warren MJ, Jahn D. Microbial production ofvitamin B12. Appl Microbiol Biotechnol. 2002; 58(3):Z75-85. Martina JA, Bonangelino CJ, Aguilar RC, Bonifacino JS. Stonin 2: an adaptor-like protein that interacts with components of the endocytic machinery. J Cell Biol. 2001; 153(5):I 1 11-20. Mathan VI, Babior BM, Donaldson RM, Jr. Kinetics of the Attachment of Intrinsic Factor-Bound Cobamides to Ileal Receptors. J Clin Invest. 1974; 54(3): 598—608. Mellersh CS, Hitte C, Richman M, Vignaux F, Priat C, Jouquand S, Werner P, Andre C, DeRose S, Patterson DF, Ostrander EA, Galibert F. An integrated linkage-radiation hybrid map ofthe canine genome. Mamm Genome. 2000; I 1(2):]20-30. I30 Metherell LA, Chapple JP, Cooray S, David A, Becker C, Ruschendorf F, Naville D, Begeot M, Khoo B, Numberg P, Huebner A, Cheetham ME, Clark AJ. Mutations in MRAP, encoding a new interacting partner ofthe ACTH receptor, cause familial glueocorticoid deficiency type 2. Nat Genet. 2005; 37(2):166~70. Miller J W, Ramos Ml, Garrod MG, Flynn MA, Green R. Transcobalamin II 775G-to-C polymorphism and indices ofvitamin BIZ status in healthy older adults. Blood. 2002; 100(2): 718-720. Moestrup SK, Bim H, Fischer PB, Petersen CM, Verroust PJ, Sim RB, Christensen El, Nexo E. Megalin-mediated endocytosis oftranscobalamin-vitamin-B12 complexes suggests a role ofthe receptor in vitamin-BIZ homeostasis. Proc Natl Acad Sci U S A. 1996; 93(16):8612-7. Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D, Pontillon F, Goda FO, Christensen E1, Hammond TG, Verroust PJ. The intrinsic factor- vitamin BIZ receptor and target of teratogenic antibodies is a megalin-binding peripheral membrane protein with homology to developmental proteins. J Biol Chem. 1998; Z73(9):5235-4Z. Moestrup SK, Verroust PJ. Megalin- and cubilin-mediated endocydosis of protein-bound vitamins, lipids, and hormones in polarized epithelia. Annu Rev Nutr. 2001; 21:407-28. Mohamed SD, McKay E, Galloway WH. Juvenile familial megaloblastic anaemia due to selective malabsorption of vitamin B-12. A family study and a review ofthe literature. Q J Med. 1966; 35(139):433—53. Mount SM. A catalogue of splicejunction sequences. Nucleic Acids Res. 1982; 10(2): 459—472. Nykjaer A, Fyfe JC, Kozyraki R, Leheste JR, Jacobsen C, Nielsen MS, Verroust PJ, Aminoff M, de la Chapelle A, Moestrup SK, Ray R, Gliemann J, Willnow TE, Christensen E1. Cubilin dysfunction causes abnomial metabolism of the steroid hormone 25(OH) vitamin D(3). Proc Natl Acad Sci U S A. 2001; 98(24):]3895-900. Okuda K. Mucosal adsorption and absorption ofvitamin BIZ in the intestine ofthe rat. Proc. Soc. Exp. Biol. Med. 1962; 11113203. Okuda K. Discovery of vitamin BIZ in the liver and its absorption factor in the stomach: a historical review. J Gastroenterol Hepatol. 1999; I4(4):301-8. Platica O, Janeczko R, Quadros EV, Regec A, Romain R, Rothenberg SP. The cDNA sequence and the deduced amino acid sequence of human transcobalamin II show homology with rat intrinsic factor and human transcobalamin I. J Biol Chem. 1991; Z66(12):7860-3. Prasad KV, Ao Z, Yoon Y, Wu MX, Rizk M, Jacquot S ,Schlossman SF. CD27, a member of the tumor necrosis factor receptor family, induces apoptosis and binds to Siva, a proapoptotic protein. Proc. Natl. Acad. Sci. USA. 1997; 94 (12): 6346-6351. Quadros EV, Sai P, Rothenberg SP. Characterization of the human placental membrane receptor for transcobalamin II-cobalamin. Arch Biochem Biophys. 1994; 308(1):19Z-9. Ramanujam KS, Seetharam S, Ramasamy M, Seetharam B. Expression of cobalamin transport proteins and cobalamin transcytosis by colon adenocarcinoma cells. Am J Physiol. 1991; 260(3):G416-ZZ. Ramanujam KS, Seetharam S, Seetharam B. Regulated expression of intrinsic factor- cobalamin receptor by rat visceral yolk sac and placental membranes. Biochim Biophys Acta. 1993; 1146(2):243-6. Ramanujam KS, Seetharam S, Dahms NM, Seetharam B. Effect of processing inhibitors on cobalamin (vitamin BIZ) transcytosis in polarized opossum kidney cells.Arch Biochem Biophys. 1994; 315(1):8-15. Ramasamy M, Alpers DH, Tiruppathi C, Seetharam B. Cobalamin release from intrinsic factor and transfer to transcobalamin II within the rat enterocyte. Am J Physiol. 1989; 257(5):G79I-7. Rappazzo ME, Hall CA. Cyanocobalamin transport proteins in canine plasma. Am J Physiol. 1972; 222(1):202—6. Regec A, Quadros EV, Platica O, Rothenberg SP. The cloning and characterization ofthe human transcobalamin II gene. Blood. 1995; 85(10):27l 1-9. Richman M, Mellersh CS, Andre C, Galibert F, Ostrander EA. Characterization of a minimal screening set of 172 microsatellite markers for genome-wide screens of the canine genome. J Biochem Biophys Methods. 2001; 47(1-2):I37-49. Rosenblatt DS and Fenton WA. Inbom errors of cobalamin metabolism. In: Banerjee, R., (ed.), Chemistry and Biochemistry ofVitamin BIZ. John Wiley and Sons, New York. 1999; pp. 367-384. Rosenblatt DS and Fenton WA. In The Metabolic and Molecular Bases ofInherited Disease (Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D., eds) 8th Ed., 2001; pp. 3897-3933, McGraw-Hill, Inc., New York. Roth JR, Lawrence JG, Rubenfield M, Kieffer-Higgins S, Church GM. Characterization ofthe cobalamin (vitamin B12) biosynthetic genes of Salmonella typhimurium. J Bacteriol. 1993; 175(1 l):3303-16. Rubin HM, Giorgio AJ, Macdonald RR, Linarelli LG. Selective malabsorption ofvitamin B 12: report ofa case with metabolic studies. Am J Dis Child. 1974; 127(5):713-7. I32 Rumpelt HJ, Michl W. Selective vitamin BIZ malabsorption with proteinuria (Imerslund- Najman-Grasbeck-syndrome): ultrastructural examinations on renal glomeruli. Clin Nephrol. 1979; II(4):213-7. Sahali D, Mulliez N, Chatelet F, Dupuis R, Ronco P, Verroust P. Characterization of a 280-kDa 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. 1988; I67(1):ZI3—8. Saito A, Pietromonaco S, Loo AK, Farquhar MG. Complete cloning and sequencing of rat gp330/"megalin," a distinctive member of the low density lipoprotein receptor gene family. Proc Natl Acad Sci U S A. 1994; 91(21):9725-9. Salameh MM, Banda RW, Mohdi AA. Reversal of severe neurological abnormalities alter vitamin B12 replacement in the Imerslund-Grasbeck syndrome. J Neurol. 1991; 238(6):349-50. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual, 2nd ed. 1989 (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). Sandoval C, Bolten P, Franco 1, Freeman S, J ayabose S. Recurrent urinary tract infections and genitourinary tract abnomialities in the Imerslund-Grasbeck syndrome. Pediatr Hematol Oncol. 2000; 17(4):331-4. Sargan DR, Yang F, Squire M, Milne BS, O'Brien PC, Ferguson-Smith MA. Use of flow- sorted canine chromosomes in the assignment of canine linkage, radiation hybrid, and syntenic groups to chromosomes: refinement and verification of the comparative chromosome map for dog and human. Genomics. 2000; 69(2):]82-95. Schaf fer AA, Gupta SK, Shriram K, Cottingham RW Jr. Avoiding recomputation in linkage analysis. Hum Hered. 1994; 44(4):225-37. Schloesser LL, Deshpande P, Schilling RF. Biologic turnover rat of cyanocobalamin (Vitamin B12) in human liver. Ama Acrch Int. Med. 1958; 101: 306-309. Seetharam B, Alpers DH, Allen RH. Isolation and characterization of the ileal receptor for intrinsic factor-cobalamin. J Biol Chem. 1981; 256(8):3785-90. Seetharam B, Alpers DH. Absorption and transport of cobalamin (vitamin B12). Annu Rev Nutr. I98Za; 2:343-69. Seetharam B, Bagur SS, Alpers DH. Isolation and characterization of proteolytically derived ileal receptor for intrinsic factor-cobalamin. J Biol Chem. 1982b; 257(1):183—9. Seetharam B, Levine J S, Ramasamy M, Alpers DH. Purification, properties, and immunochemical localization of a receptor for intrinsic factor-cobalamin complex in the 133 rat kidney. J Biol Chem. 1988; 263(9):4443-9. Seetharam B, Christensen EI, Moestrup SK, Hammond TG, Verroust PJ. Identification of rat yolk sac target protein of teratogenic antibodies, gp280, as intrinsic factor-cobalamin receptor. J Clin Invest. 1997; 99(10):2317-22. Seetharam B. Receptor-mediated endocytosis of cobalamin (vitamin BIZ). Annu Rev Nutr. 1999; 19:173-95. Review. Sennett C, Rosenberg LE, Mellman IS. Transmembrane transport of cobalamin in prokaryotic and eukaryotic cells. Annu Rev Biochem. 1981; 50:1053-86. Review. Si K, Das K, Maitra U. Characterization of multiple mRNAs that encode mammalian translation initiation factor 5 (eIF-5). J Biol Chem. 1996; 271(28): 16934-38. Simpson KW, Alpers DH, De Wille J, Swanson P, Farmer S, Sherding RG. Cellular localization and homional regulation of pancreatic intrinsic factor secretion in dogs. Am J Physiol. 1993; Z65(1):GI78-88. Stabler SP, Allen RH. Vitamin BIZ deficiency as a worldwide problem. Annu Rev Nutr. 2004; 24:299-326. Review. Strope S, Rivi R, Metzger T, Manova K, Lacy E. Mouse amnionless, which is required for primitive streak assembly, mediates cell-surface localization and endocytic function of cubilin on visceral endodem1 and kidney proximal tubules. Development. 2004; l3l(I9):4787-95. Tailor CS, Marin M, Nouri A, Kavanaugh MP, Kabat D. Truncated forms of the dual function human ASCTZ neutral amino acid transporter/retroviral receptor are translationally initiated at multiple alternative CUG and GUG codons. J Biol Chem. 2001; Z76(Z9):Z7221-30. Taira M, Iizasa T, Shimada H, Kudoh J, Shimizu N, Tatibana M. A human testis-specific mRNA for phosphoribosylpyrophosphate synthetase that initiates from a non-AUG codon. J Biol Chem. 1990; 265(27):]6491-7. Tang LH, Chokshi H, Hu CB, Gordon MM, Alpers DH. The intrinsic factor (IF )- cobalamin receptor binding site is located in the amino-terminal portion of IF. J Biol Chem. 1992; Z67(32):22982-6. Tanner SM, Aminoff M, Wright FA, Liyanarachchi S, Kuronen M, Saarinen A, Massika O, Mandel H, Broch H, de la Chapelle A. Amnionless, essential for mouse gastrulation, is mutated in recessive hereditary megaloblastic anemia. Nat Genet. 2003; 33(3):426-9. Tanner SM, Li Z, Bisson R, Acar C, Oner C, Oner R, Cetin M, Abdelaal MA, Ismail EA, Lissens W, Krahe R, Broch H, Gréisbeck R, de la Chapelle A. Genetically heterogeneous I34 selective intestinal malabsorption of vitamin B12: founder effects, consanguinity, and high clinical awareness explain aggregations in Scandinavia and the Middle East. Hum Mutat. 2004; 23(4):327-33. Tomihara-Newberger C, Haub 0, Lee HG, Soares V, Manova K, Lacy E. The amn gene product is required in extraembryonic tissUes for the generation of middle primitive streak derivatives. Dev Biol. 1998; 204(1):34-54. Vaillant C, Horadagoda NU, Batt RM. Cellular localization of intrinsic factor in pancreas and stomach ofthe dog. Cell Tissue Res. 1990; Z60(l):1 17—22. Van Praet O, Argraves WS, Morales CR. Co-expression and interaction of cubilin and megalin in the adult male rat reproductive system. Mol Reprod Dev. 2003; 64(2):IZ9-35. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ et al. The sequence ofthe human genome. Science. 2001; 29l(5507):1304-51. Vinogradov AE. Genome size and GC-percent in vertebrates as detemiined by flow cytometry: the triangular relationship. Cytometry. I998; 31(2):100-9. Wang X, Bornslaeger EA, Haub O, Tomihara-Newberger C, Lonberg N, Dinulos MB, Disteche CM, Copeland N, Gilbert DJ, Jenkins NA, Lacy E. A candidate gene for the amnionless gastrulation stage mouse mutation encodes a TRAP-related protein. Dev Biol. 1996; 177(1):Z74-90. Watkins D, Ru M, Hwang HY, Kim CD, Murray A, Philip NS, Kim W, Legakis H, Wai T, Hilton JF, Ge B, Dore C, Hosack A, Wilson A, Gravel RA, Shane B, Hudson TJ, Rosenblatt DS. Hyperhomocysteinemia due to methionine synthetase deficiency, cblG: structure of the MTR gene, genotype diversity, and recognition of a common mutation, Pl I73L. Am. J. Hum. Genet. 2002; 71(1): 143-153. Weng Z, Fluckiger AC, Nisitani S, Wahl MI, Le LQ, Hunter CA, Femal AA, Le Beau MM, Witte ON. A DNA damage and stress inducible G protein-coupled receptor blocks cells in GZ/M. Proc Natl Acad Sci U S A. 1998; 95(21):]2334-9. Werner P, Mellersh CS, Raducha MG, DeRose S, Aeland GM, Prociuk U, Wiegand N, Aguirre GD, Henthom PS, Patterson DF, Ostrander EA. Anchoring of canine linkage groups with chromosome-specific markers. Mamm Genome. 1999; 10(8):814-23. Willigan DA, Cronkite EP, Meyer LM, Noto SL. Biliary excretion of C060 labeled vitamin BIZ in dogs. Proc Soc Exp Biol Med. 1958; 99(1):81-4. Wilson A, Leclerc D, Rosenblatt DS, Gravel RA. ‘Molecular basis for methionine synthetase reductase deficiency in patients belonging to the cblE complementation group ofdisorders in folate/cobalamin metabolism. Hum Mol Genet. 1999; 8(1 1):2009-16. I35 Xu D, Kozyraki R, Newman TC, Fyfe .IC. Genetic evidence of an accessory activity required specifically for cubilin brush-border expression and intrinsic factor-cobalamin absorption.Blood. I999; 94(10):3604-6. Xu D, Fyfe JC. Cubilin expression and posttranslational modification in the canine gastrointestinal tract. Am J Physiol Gastrointest Liver Physiol. 2000; Z79(4):G748-56 Yammani RR, Seetharam S, Seetharam B. Identification and characterization of two distinct ligand binding regions ofcubilin. J Biol Chem. 2001; 276(48):44777-84. Yang YM, Ducos R, Rosenberg AJ, Catrou PG, Levine JS, Podell ER, Allen RH. Cobalamin malabsorption in three siblings due to an abnormal intrinsic factor that is markedly susceptible to acid and proteolysis. J Clin Invest. 1985; 76(6):2057-65. Yassin F, Rothenberg SP, Rao S, Gordon MM, Alpers DH, Quadros EV. Identification of a 4-base deletion in the gene in inherited intrinsic factor deficiency. Blood. 2004; 103(4): 1515-1517. Yetgin S, Altay C. Laleli Y. Familial selective vitamin BIZ malabsorption in three families. TurkJ Pediatr. 1978; 20(1-2):44-50. Yki-Jarvinen H, Kouvonen I, Griisbeck R. Solubilization and preliminary characterization ofthe human ileal vitamin BIZ—intrinsic factor receptor. Scand J Clin Lab Invest. 1979; 39(5):46I-7. Youngdahl-Tumer P, Rosenberg LE, Allen RH. Binding and uptake oftranscobalamin II by human fibroblasts. J Clin Invest. 1978; 61(1):I33-4l. Zavadakova P, Fowler B, Zeman J, Suonnala T, Pristoupilova K, Kozich V, Zavad'akova P. C blE type of homocystinuria due to methionine synthetase reductase deficiency: clinical and molecular studies and prenatal diagnosis in two families. J Inherit Metab Dis. 2002; 25(6):46l-76. Zhai XY, Nielsen R, Bim H, Drumm K, Mildenberger S, Freudinger R, Moestrup SK, Verroust PJ, Christensen El, Gekle M. Cubilin- and megalin-mediated uptake of albumin in cultured proximal tubule cells of opossum kidney. Kidney Int. 2000; 58(4):1523-33. I36