Novel Inborn Error of Vitamin B12 Metabolism Caused by Mutations in ABCD4

Novel inborn error of vitamin B12 metabolism caused by mutations in ABCD4

Jaeseung Kim

Department of Human Genetics

McGill University

Montréal, Québec, Canada

June 2012

A thesis submitted to McGill University in partial fulfillment of the requirements

of the degree of

Master of Science

Vitamin B12 (cobalamin, Cbl) is an essential cofactor for two human enzymes: methylmalonyl-CoA mutase (MUT) and methionine synthase (MTR).

MUT utilizes 5’-deoxyadenosylcobalamin (AdoCbl) to convert methylmalonyl-

CoA to succinyl-CoA in the mitochondria, whereas MTR utilizes methylcobalamin (MeCbl) to convert homocysteine to methionine in the cytoplasm. To date, eight complementation groups (cblA-G and mut), each the result of mutations at a different gene, have been discovered to be involved in the intracellular metabolism of cobalamin. A patient presented at birth, following an abnormal newborn screen, with hypotonia, lethargy, poor feeding and bone marrow suppression. There were elevated levels of methylmalonic acid and homocysteine, suggestive of a defect in vitamin B12 metabolism. Studies of cultured fibroblast showed decreased function of the cobalamin-dependent enzymes, MTR and MUT. There was increased uptake of labelled cyanocobalamin (CNCbl) but decreased synthesis of the cobalamin cofactors

MeCbl and AdoCbl, with accumulation of “free” (i.e. non-protein bound) CNCbl in the cells. The cellular phenotype mimicked that of the cblF disorder caused by mutations in the LMBRD1 gene encoding the lysosomal membrane protein

LMBD1 that is thought to play a role in transfer of cobalamin across the lysosomal membrane into the cytoplasm. However, cells from the patient

2 complemented those from all known complementation groups, including cblF, and no mutations in LMBRD1 were found. Whole-exome sequencing led to the identification of two mutations in the ABCD4 gene: c.956A>G (p.Y319C) and c.1746_1747insCT (p.E583LfsX9). Two additional patients with deleterious

ABCD4 mutations were later found. Transfection of patient fibroblasts with wild type ABCD4 led to rescue of all abnormal cellular phenotypes. This thesis reports that this novel disorder, named cblJ, is an autosomal recessive disorder caused by mutations in ABCD4. The findings suggest that ABCD4, an ABC half-transporter, is another essential component of intracellular cobalamin metabolism.

RÉ SUME

La vitamine B12 (cobalamine, Cbl) est un cofacteur essentiel pour deux enzymes de l'homme: la méthylmalonyl-CoA mutase (MUT) et la méthionine synthase (MTR). La MUT utilise 5'-deoxyadenosylcobalamin (AdoCbl) pour convertir la méthylmalonyl-CoA en succinyl-CoA dans la mitochondrie, alors que

MTR utilise la méthylcobalamine (MeCbl) pour convertir l'homocystéine en méthionine dans le cytoplasme. À ce jour, huit groupes de complémentation

(cblA-G et mut) ont été découverts à être impliqués dans le métabolisme intracellulaire de la cobalamine. Chacun est le résultat de mutations au niveau d'un gène différent. Un patient s’est présenté à la naissance, suite à une anomalie sue le dépistage nouveau-né, avec l’hypotonie, la léthargie, la mauvaise alimentation et la suppression de la moelle osseuse. Le patient avait des niveaux

élevés d'acide méthylmalonique et d'homocystéine, suggérant un défaut dans le métabolisme de la vitamine B12. Les études de fibroblastes cultivés ont démontré une diminution de la fonction des enzymes dépendants sur la cobalamine, MTR et

MUT. Il avait aussi une augmentation de l’absorption de la cyanocobalamine

(CNCbl), mais une diminution de la synthèse de cofacteurs cobalamine la MeCbl et AdoCbl, avec une accumulation de CNCbl “libre” (c'est à dire non liée aux protéines plasmatiques) dans les cellules. Le phénotype cellulaire imitait celle de la maladie cblF, causée par des mutations dans LMBRD1, le gène codant pour la

4 protéine membrane lysosomale LMBRD1, qui semble jouer un rôle dans le transfert de la cobalamine à travers la membrane lysosomale dans le cytoplasme.

Cependant, les cellules des deux patients complémentaient celles de tous les groupes de complémentation connus, y compris cblF, et aucune des mutations dans LMBRD1 ont été trouvés. Le séquençage de l’exome a mené à l'identification de deux mutations dans le gène ABCD4: c.956A>G (p.Y319C) et c.1746_1747insCT (p.E583LfsX9). Deux autres patients avec des mutations dans

ABCD4 ont été retrouvés. Toutes les mutations ont été prévues d’être nocives. La transfection de fibroblastes de patients avec ABCD4 de type sauvage a conduit à sauver tous les phénotypes cellulaires anormaux. Cette thèse rapporte que ce trouble inédit, nommé cblJ, est une maladie autosomique récessive causée par des mutations dans ABCD4. Les résultats suggèrent qu’ABCD4, un demi-ABC transporteur, est un autre élément essentiel du métabolisme de la cobalamine intracellulaire.

TABLE OF CONTENTS

ABSTRACT ...... 2 RÉ SUME ...... 4 TABLE OF CONTENTS ...... 6 LIST OF ABBREVIATIONS ...... 11 LIST OF FIGURES ...... 13 LIST OF TABLES ...... 14 ACKNOWLEDGEMENTS ...... 15

CHAPTER 1: INTRODUCTION ...... 16 1.1 Vitamin B Group ...... 16

1.2 Vitamin B12 ...... 16 1.2.1 Historical Aspects ...... 16

1.2.2 Structure ...... 18 1.2.3 Biosynthesis ...... 20 1.2.4 Function ...... 20 1.2.4.1 Cofactor for Three Enzyme Classes ...... 20 1.2.4.2 Two Human Enzymes ...... 23

1.3 Absorption and Transport of Vitamin B12 ...... 23 1.3.1 Carrier Proteins ...... 23 1.3.2 Absorption and Transport Pathway ...... 24

1.3.3 Inborn Errors of Vitamin B12 Absorption and Transport ...... 25 1.3.3.1 Haptocorrin Deficiency ...... 26 1.3.3.2 Inherited Intrinsic Factor Deficiency ...... 26 1.3.3.3 Imerslund-Gräsbeck Syndrome ...... 28 1.3.3.4 Transcobalamin Deficiency ...... 29

1.3.3.5 Transcobalamin Receptor Deficiency ...... 29 1.3.4 Related Proteins ...... 30 1.3.4.1 Megalin/LRP2 ...... 30 1.3.4.2 ABCC1/MRP1 ...... 31

1.4 Intracellular Metabolism of Vitamin B12 ...... 32 1.4.1 The Patients ...... 32 1.4.2 Somatic Cell Complementation Analysis ...... 33 1.4.3 Discoveries of Eight Complementation Groups ...... 34 1.4.3.1 Four Complementation Groups ...... 35 1.4.3.2 A Complementation Group with Three Phenotypes ..... 36 1.4.3.3 Isolated HC with Megaloblastic Anemia ...... 37

1.4.3.4 Failure of Lysosomal Release of Vitamin B12 ...... 38 1.4.3.5 Heterogeneity Among Patients with HC ...... 38

1.4.4 Inborn Errors of Vitamin B12 Metabolism ...... 39 1.4.4.1 cblF, cblC, and cblD ...... 40 1.4.4.2 cblB, cblA, and mut ...... 44 1.4.4.3 cblE and cblG ...... 47 1.4.5 Other Causes of MMA ...... 49 1.4.6 Other Causes of HC ...... 49 1.5 cblF Disease ...... 50 1.5.1 Background on Lysosome ...... 50 1.5.1.1 Role of Lysosome ...... 50 1.5.1.2 Lysosomal Soluble Proteins ...... 51 1.5.1.3 Lysosomal Membrane Proteins ...... 51 1.5.1.4 Disorders of Lysosomal Export ...... 52 1.5.2 LMBRD1 Gene ...... 53 1.5.3 Pathophysiology and Treatment ...... 54 1.6 Undiagnosed Patients and Gaps in the Pathway ...... 56

1.7 Evolution of Gene Discovery Approaches ...... 57 1.7.1 Limitations of Traditional Approaches ...... 57 1.7.2 First Applications of Exome Sequencing ...... 58 1.7.3 New Paradigm of Disease Gene Discovery ...... 59 1.7.4 Exome Sequencing of Undiagnosed Patients ...... 60 1.8 Peroxisomal ABC Half-Transporters ...... 61 1.8.1 ATP-Binding Cassette Transporters ...... 61 1.8.2 ATP-Binding Cassette, Subfamily D ...... 62 1.8.3 ABCD4 Gene ...... 64 1.8.3.1 Discovery and Characterization ...... 64 1.8.3.2 Subcellular Localization ...... 65

RATIONALE AND OBJECTIVES OF STUDY ...... 67

CHAPTER 2: MATERIALS AND METHODS ...... 68 2.1 Case Reports ...... 68 2.1.1 Patient WG4066 ...... 68 2.1.2 Patient WG4140 ...... 69 2.1.3 Patient WG3630 ...... 71 2.2 Cell Culture ...... 73 2.3 Selection of Fibroblast Cell Lines ...... 74 2.4 Somatic Cell Complementation Analysis ...... 76 2.5 Exome Sequencing ...... 76 2.6 Mutation Analysis ...... 77 2.6.1 Polymerase Chain Reaction (PCR) ...... 79 2.7 Immortalization of Fibroblasts with E7 and Telomerase ...... 79 2.8 Transfection of Fibroblasts with Wild Type ABCD4 cDNA ...... 80 2.8.1 LR Recombination Reaction ...... 80

2.8.2 Colony PCR of LR Recombinants ...... 80 2.8.3 Transfection ...... 81 2.9 Transfection of Fibroblasts with Wild Type LMBRD1 cDNA ...... 82 2.9.1 Gateway Cloning ...... 82 2.9.2 BP Recombination Reaction ...... 82 2.9.3 Colony PCR of BP Recombinants ...... 83 2.9.4 LR Recombination Reaction ...... 83 2.9.5 Colony PCR of LR Recombinants ...... 83 2.9.6 Transfection ...... 84 2.10 Labelled MethylTHF and Propionate Incorporation Assays ...... 84 2.11 Cobalamin Derivative Distribution Assay ...... 85 2.12 Superose 12 Analysis of TC-Bound, Free and Enzyme-Bound Cbl ...... 86

CHAPTER 3: RESULTS ...... 87 3.1 Identification of Three Patients ...... 87 3.2 Somatic Cell Complementation Analysis ...... 89 3.3 Discovery of Causative Gene in Each Patient ...... 92 3.3.1 Patient WG4066 ...... 92 3.3.2 Patient WG4140 ...... 95 3.3.3 Patient WG3630 ...... 95 3.4 Transfections and Assessments of Biochemical Phenotypes ...... 97 3.4.1 Labelled MethylTHF and Propionate Incorporation Assays ...... 97 3.4.2 Cobalamin Derivative Distribution Assay ...... 101 3.4.3 Superose 12 Analysis ...... 104

CHAPTER 4: DISCUSSION ...... 108

4.1 Novel Inborn Error of Vitamin B12 Metabolism ...... 108

4.2 Role of ABCD4 in Vitamin B12 Metabolism ...... 113

ORIGINAL CONTRIBUTIONS TO SCIENCE ...... 116 BIBLIOGRAPHY ...... 117 APPENDIX A: List of Publications and Presentations ...... 132 APPENDIX B: Published Abstract ...... 134

LIST OF ABBREVIATIONS

ABC transporter: ATP-binding cassette transporter ABCD4: ATP-binding cassette, subfamily D, member 4 ABCX#: ATP-binding cassette, subfamily X, member # (nomenclature) AdoCbl: 5’-deoxyadenosylcobalamin BME: β-mercaptoethanol Cbl: cobalamin CBS: cystathionine β-synthase cDNA: complementary DNA CNCbl: cyanocobalamin DMB: 5,6-dimethylbenzimidazole DTT: dithiothreitol E. coli: Escherichia coli Enzyme-Cbl: MUT or MTR-bound cobalamin ER: endoplasmic reticulum FAD: flavin adenine dinucleotide FMN: flavin mononucleotide GWAS: genome-wide association studies HC: homocystinuria HgB: hemoglobin IF: intrinsic factor IGS: Imerslund-Gräsbeck syndrome LIMR: lipocalin-1 interacting membrane receptor LMBD1: LMBR1 domain-containing protein 1 LMBR: limb region 1 LMP: lysosomal membrane protein LRP2: low-density lipoprotein receptor-related protein 2

MCV: mean corpuscular volume MeCbl: methylcobalamin Met: methionine MMA: methylmalonic aciduria MMAA: methylmalonic aciduria cblA type MMAB: methylmalonic aciduria cblB type MMACHC: methylmalonic aciduria cblC type, with homocystinuria MMADHC: methylmalonic aciduria cblD type, with homocystinuria MRP1: multidrug resistance-associated protein 1 M6P: mannose-6-phosphate MTHFD1: methylenetetrahydrofolate dehydrogenase 1 MTHFR: 5,10-methylenetetrahydrofolate reductase MTR: methionine synthase MTRR: methionine synthase reductase MUT: methylmalonyl-CoA mutase MW: molecular weight NBD: nucleotide-binding domain OHCbl: hydroxocobalamin PA: pernicious anemia PEG: polyethylene glycol 1000 RT-PCR: reverse transcription polymerase chain reaction SNP: single-nucleotide polymorphisms TC: transcobalamin TCblR: transcobalamin receptor THF: tetrahydrofolate TMD: transmembrane domain VLCFA: very long chain fatty acids

LIST OF FIGURES

Figure 1. Structure of vitamin B12 ...... 19

Figure 2. Vitamin B12 metabolism ...... 41

Figure 3. Putative membrane topology of the LMBD1 protein ...... 55

Figure 4. Putative membrane topology of an ABC, Subfamily D transporter ..... 63

Figure 5. Protein sequence identity and similarity among ABCD proteins ...... 66

Figure 6. DNA sequencing chromatograms of heterozygous mutations in TPRG1, LRP2, and ABCD4 in the family of WG4066 ...... 93

Figure 7. (A) Pedigree of the family of WG3630 showing plasma homocysteine levels and genotypes (B) DNA sequencing chromatograms of the ABCD4 c.423C>G (p.N141K) mutation ...... 96

Figure 8. Cobalamin derivative distributions of 18 cell lines ...... 103

Figure 9. Elution patterns of Superose 12 analyses ...... 106

Figure 10. Superose 12 analyses of 18 cell lines ...... 107

LIST OF TABLES

Table 1. Fibroblast cell lines used in this study ...... 75

Table 2. Primers used in this study ...... 78

Table 3. Biochemical profiles of three patient fibroblasts ...... 88

Table 4. Somatic cell complementation of WG4066 ...... 90

Table 5. Somatic cell complementation of WG4140 ...... 91

Table 6. Segregation analysis of TPRG1, LRP2 and ABCD4 mutations in the family of WG4066 ...... 94

Table 7. Labelled methylTHF incorporations of 24 cell lines ...... 99

Table 8. Labelled propionate incorporations of 24 cell lines ...... 100

ACKNOWLEDGEMENTS

I express my greatest gratitude to Dr. David Rosenblatt, my research supervisor. I thank him for allowing me to take on an excellent scientific project, guiding me through it, and making the experience enjoyable and memorable. He is an inspiring scientist and a pioneer and I will remember his lessons when I delve into new areas of science in the future. I cannot thank Dr. David Watkins enough for mentoring me, editing my writings, and troubleshooting with me through all my experimental hurdles. I always knew who to ask when I accidently skipped a step in an experiment or needed a reagent we did not have. I am thankful to Drs. Eric Shoubridge and Jacek Majewski for their guidance and scientific input as supervisory committee members. I also thank Gail Dunbar for teaching me how to grow and maintain cell cultures, and Thomas Leslie and Kandace Springer of the HG office for their efforts to answer all my inquiries and aid me in every way possible. And I appreciate all the members and collaborators of our laboratory for being great friends and helping me with any problems: Laura Dempsey-Nunez, Peg Illson, Alison Brebner, Maria Plesa, Wayne Mah, Stephen Fung, Laura Benner, Isabelle Miousse, Justin Deme, Jackie Chung, Ross Mackay, and summer students Francis, Selim, Tracy, Dylan, and Kush. Especially, I will remember Laura and Peg for the time we shared, lattes we consumed, and random matters we contemplated on. Lastly, I am grateful to my family and friends for showing me their support and faith over the past two years of my graduate studies. They wished for my success and cheered for my accomplishments. I will treasure the special relationship I have with each and every one of them.

Due to the collaborative nature of this project, the following acknowledgement is given to important contributors. Patient WG4066 was referred to our laboratory by Dr. Nicola Longo, WG4140 by Dr. Brian Fowler, and WG3630 by Dr. Ni-Chung Lee. Somatic cell complementation was performed by Jocelyne Lavallée of the Rosenblatt Laboratory. Exome capture sequencings of WG4066 and WG3630 were performed in collaboration with the laboratory of Dr. Jacek Majewski. Exome capture sequencing and Sanger sequencings of patient WG4140 were performed by the research team led by Dr. Brian Fowler and Dr. Matthias Baumgartner. Immortalization and transfection of fibroblasts were performed by Timothy Johns and Stephen Fung, respectively, of the Eric Shoubridge Laboratory. The success of this project would not have been possible without their amazing work.

CHAPTER 1: INTRODUCTION

1.1 Vitamin B Group

Vitamins are essential human nutrients and are comprised of a broad range of organic compounds. Since human cells are not equipped with de novo pathways for their syntheses, vitamins must be exogenously obtained through diet or supplements to sustain health.1 The vitamin B group, once thought of as a single vitamin, is now a name for a group of eight chemically distinct, water- soluble organic compounds: thiamin (B1), riboflavin (B2), niacin (B3), pantothenic

1 acid (B5), pyridoxine (B6), biotin (B7), folic acid (B9), and cobalamin (B12). Once absorbed, the vitamins must be assimilated into their active forms via enzymes present in the body to perform various functions. For example, riboflavin is converted into flavin mononucleotide (FMN) or flavin adenine dinucleotide

(FAD), which are electron carriers in redox reactions, such as those in the oxidative phosphorylation pathway.2

1.2 Vitamin B12

1.2.1 Historical Aspects

The discovery of vitamin B12 in human physiology has special and specific ties with pernicious anemia (PA). The disease is a type of megaloblastic anemia and presents with low blood cell count, low hemoglobin concentration,

16 hypochlorhydria, and neuropathy, such as spinal cord degeneration in later stages.3 It was considered “pernicious” when its etiology was not understood and no treatments were available to prevent its certain fatality. Although the importance of diet in patients was argued occasionally, the first dietary therapy was conceived only when Minot and Murphy discovered the curative effect of a special diet containing cooked liver.4 However, they were incorrect in implying that the diet rich in complete proteins and iron was improving patients’ health.

Subsequent studies led to the isolation and purification of the anti-pernicious anemia factor from liver extracts and demonstrated a very high biological activity in treatment of PA.5, 6 It was confirmed to be the same material as the LLD factor, which was necessary for growth of Lactobacillus lactis Dorner and was postulated to be the active compound in liver extracts used for treating PA.7 This compound

6 was then properly named vitamin B12 to indicate its nutritional significance. This new B group vitamin was thoroughly studied and was revealed to function as a cofactor for numerous metabolic enzymes in bacteria and two enzymes in human, which will be further described in Section 1.2.4.8 The complete pathophysiology of PA has later been elucidated to be the destruction of stomach parietal cells by autoimmune antibodies and ensuing abolishment of intrinsic factor production and

3 vitamin B12 absorption. Current treatment for PA varies among countries but the recommendation is daily intramuscular injection of 5 mg of cyanocobalamin

(CNCbl) for 5 days followed by intramuscular injection of 5 mg of CNCbl every

3 3 months to maintain vitamin B12 storage. Oral administration of cobalamin is recommended in certain countries.

1.2.2 Structure

The complete three-dimensional structure of vitamin B12 was solved by

9 X-ray crystallography technique in 1956. Vitamin B12, also termed cobalamin

(Cbl) for the presence of a central cobalt ion, is an organometallic cofactor and one of the largest and most structurally complex cofactors in nature.1 The central cobalt is supported by six coordination sites, four of which are provided by nitrogen ligands of a planar corrin ring (Figure 1).10 5,6-dimethylbenzimidazole

(DMB) is the lower axial ligand appended to a side chain of the corrin ring and provides the fifth coordination site to the cobalt ion. Protonation of DMB detaches it from cobalt and the cobalamin switches from “base-on” to “base-off” state.

On the other hand, the upper axial ligand, which serves as the sixth coordination site, is diverse in its properties and can be a hydroxyl, cyano, 5’- deoxyadenosyl or methyl group. As a result, cobalamin can exist as four major derivatives depending on the upper axial ligand. The oxidation state of cobalt atom ranges from +1 to +3, and the preferred number of coordinates around it increases correspondingly from four to six (1). Cobalamin derivatives are metabolized into methylcobalamin (MeCbl) or 5’-deoxyadenosylcobalamin

(AdoCbl) to become active cofactors in human cells (19).

Figure 1. Structure of vitamin B12. Six coordination sites around the cobalt atom can be envisioned as six vertices of an octahedron. Four coordinates are provided by nitrogen atoms of pyrrole-like rings of the corrin ring, the lower axial ligand by the DMB moiety, and the upper axial ligand by various chemical groups, represented by R. Depending on the chemical nature of the R group, the vitamin

B12 molecule can exist as four cobalamin derivatives, cob(II)alamin, or cob(I)alamin. Adapted from Motwani et al, 2011.

1.2.3 Biosynthesis

Human cells are not equipped with enzymatic pathways for the de novo synthesis of corrinoids, which refers to corrin-containing compounds and includes cobalamin.11 As a result, cobalamin is an essential nutrient that must be obtained exogenously through supplements or animal products in the diet to sustain human health.12 The ability to produce cobalamin is restricted to bacteria and archaea.

Synthesis of cobalamin from uroporphyrinogen III, the precursor molecule, is a multi-step process requiring actions of more than 30 enzymes and encompassing the synthesis of corrin ring, incorporation of cobalt ion, and synthesis of DMB moiety.13

Interestingly, not all bacteria can produce cobalamin and not all bacteria require cobalamin for survival. For example, the Gram-negative bacterium,

Escherichia coli (E. coli), depends on cobalamin but has lost the genes to produce it during evolution; in order to survive, it must scavenge the cobalamin from the environment and transport it across the outer and inner membranes into the cytoplasm.13

1.2.4 Function

1.2.4.1 Cofactor for Three Enzyme Classes

Three classes of enzymes, methyltransferases, isomerases, and reductive dehalogenases, require cobalamin as a cofactor and exploit its unique chemistry in

20 various ways.8 The organometallic bond (Co-C bond) between central cobalt ion and carbon atom in the methyl or adenosyl group has a weak bond disassociation energy of ~39 kcal/mol or ~31 kcal/mol, respectively, which can break by heterolysis or homolysis to create a methyl cation or a free radical, respectively.14

Methyltransferases transfer a methyl group from a methyl donor (X) to a

- methyl acceptor (Y), as shown by a generalized chemical equation, Y + CH3-X

- 8 → CH3-Y + X . Methanol, methyl amines and methyltetrahydrofolate are exemplary methyl group donors and compounds such as homocysteine and coenzyme M are possible methyl group acceptors. Because cobalamin-dependent methyltransferases utilize MeCbl as a methyl carrier, the reaction is divided into two half reactions; first the methyl group is transferred from the donor to cob(I)alamin, and second, the methyl group is transferred from MeCbl to the acceptor. For that reason, a methyltransferase must have the capacity to bind its substrate, cobalamin, and methyl donor(s).8

Isomerases form a large subfamily of cobalamin-dependent enzymes.

These enzymes catalyze 1,2-rearrangements involving carbon, nitrogen, or oxygen in various types of substrates; the common characteristic of these isomerases is that they exploit the deoxyadenosyl radical formed by homolysis of the labile Co-C bond in AdoCbl.8 Some of the cobalamin-dependent isomerases are methylmalonyl-CoA mutase, glutamate mutase, ethanolamine ammonia lyase,

β-lysine 5,6-aminomutase, diol dehydrase, glycerol dehydratase, and

21 ribonucleotide reductase. The general reaction mechanism involves, in the following order, homolysis of AdoCbl, abstraction of a hydrogen atom from substrate by deoxyadenosyl radical, 1,2-rearrangement, and abstraction of a hydrogen atom by the intermediate substrate radical.8 For example, the glutamate mutase catalyzes the isomerization of L-glutamate to and from L-threo-3- methylaspartate.15 In fact, the glutamate mutase was the first enzyme discovered to be dependent on vitamin B12 as a coenzyme.

Reductive dehalogenases dependent on cobalamin and iron-sulfur clusters have been discovered in anaerobic microbes. These enzymes, although absent in humans, are found in around 20 strains of bacteria and serve to remove halogen atoms from polyhalogenated compounds; some strains have been shown to link dehalogenation reaction with the electron transport pathway for anaerobic respiration.8 Each dehalogenase can catalyze dehalogenation of a specific group of substrates and has a preference toward which position on the chemicals it removes a halogen atom from. For instance, Desulfitobacterium chlororespirans contains the 3-chloro-4-hydroxybenzoate dehalogenase which removes a chlorine located ortho to a hydroxyl group.16 The mechanism of action and the cobalamin forms used by the dehalogenases are yet unclear. In Dehalospirillum multivorans, the tetrachloroethene reductive dehalogenase was found to use a novel

17 norpseudovitamin B12 as its cofactor.

1.2.4.2 Two Human Enzymes

In human, cobalamin is required for the activity of two intracellular enzymes: methionine synthase, which is also termed 5-methyltetrahydrofolate- homocysteine methyltransferase (MTR), and methylmalonyl-CoA mutase

(MUT).18 MeCbl is used as a single-carbon carrier in the transfer of a methyl group from 5-methyltetrahydrofolate (5-methylTHF) to homocysteine to synthesize methionine in the cytoplasm. This is an important reaction controlling levels of several metabolites.19 On the other hand, AdoCbl acts as a radical intermediate for the isomerization of L-methylmalonyl-CoA to succinyl-CoA in the mitochondria.18 This reaction is required for the complete metabolism of odd- numbered chain fatty acids and branched-chain amino acid and feeding succinyl-

CoA into the Krebs cycle. It is notable that approximately 95% of intracellular cobalamins are bound to MTR in the cytosol or MUT in the mitochondria to form holoenzymes.18, 20 Defects in MTR or MUT enzymes or deficiencies in the metabolism of Cbl affecting synthesis of MeCbl or AdoCbl can lead to the accumulation of precursor molecules in these enzymatic reactions, namely homocysteine and methylmalonyl-CoA.21

1.3 Absorption and Transport of Vitamin B12

1.3.1 Carrier Proteins

Cobalamin molecules depend heavily on cobalamin-binding carrier

23 proteins for three main reasons; first, cobalamin intermediates are highly reactive and labile, so they must be protected from the aqueous environment; second, cobalamin, due to its large size, cannot cross the plasma membrane independently and requires carrier proteins to shuttle it via receptor-mediated endocytosis; and third, carrier proteins provide an efficient means of delivery for cobalamins which exist in low concentrations in the environment and in the body. Three carrier proteins in human are haptocorrin, intrinsic factor (IF), and transcobalamin

(TC).18 Given the similarity of genomic structures of the three genes, it was proposed that they were created by duplication events of an ancestral gene during evolution. IF diverged from a duplicate of TC, and haptocorrin then diverged from a duplicate of IF. The cobalamin-binding domain is relatively conserved among the three proteins while receptor binding specificity has differentiated.22 After much ambiguity on the genetic basis of the three carriers, it has been determined that haptocorrin is encoded by TCN1, IF is encoded by GIF, and TC is encoded by

TCN2.18

1.3.2 Absorption and Transport Pathway

In the mouth, cobalamin binds to haptocorrin, a salivary glycoprotein, once it is freed from other food components. Cobalamin is in complex with haptocorrin in the mouth and stomach until the glycoprotein is degraded by pancreatic proteases in the acidic environment of the small intestine. Then,

24 cobalamin binds to IF, which is secreted by gastric parietal cells that also produce

HCl-containing gastric acid. Absorption of IF-Cbl complex at the distal ileum is mediated by a membrane receptor termed cubam, which is a heterodimer of cubilin and amnionless proteins. Inside the enterocyte, IF is degraded and cobalamin finally binds to TC and is released into the blood plasma.18, 23

Approximately 20% of circulating cobalamin, which are bound to TC, represents functional cobalamin while the remaining 80%, bound to haptocorrin, is not known to have a clear function.24 The transcobalamin receptor (TCblR) is specialized to sequester the TC-Cbl complex into most cell types in the body; attachment of TC-Cbl complex triggers a receptor-mediated endocytosis of the supercomplex, and here on starts the intracellular metabolism of cobalamin.25

Inside the somatic cells, an adequate level of cobalamin is required for the normal activities of cobalamin-dependent enzymes.

1.3.3 Inborn Errors of Vitamin B12 Absorption and Transport

Defects in aforementioned proteins result in decreased absorption of vitamin B12 by the human body as illustrated by pernicious anemia, which is caused by a decrease in production and secretion of IF into the stomach. Inherited cobalamin malabsorption disorders can be caused by germline mutations in genes encoding haptocorrin, IF, cubam, TC, and TCblR and are inherited in autosomal recessive manner. They are individually described below.

1.3.3.1 Haptocorrin Deficiency

Haptocorrin is a cobalamin-binding protein and has previously been known as TCI, TCIII, cobalophilin, R-binder, and salivary binder.24 Since the cloning and mapping of the TCN1 gene, the field has started to consistently refer to its 433-amino acid gene product as haptocorrin.26, 27 Other than the fact that

80% of circulating cobalamin is bound to haptocorrin, its function and role in pathophysiology is not known.24 Low level of serum cobalamin has been associated with haptocorrin deficiency in a single study.28 Two mutations, c.270delG and c.315C>T, have been identified in three families and severe or mild haptocorrin deficiency was linked to compound heterozygous or heterozygous genotypes of patients, respectively.29 It must be emphasized that haptocorrin deficiency is not definitively proven to cause any clinical manifestations in patients. Absence of clinical outcomes in patients will extend the hypothesis that normal level of TC-Cbl is necessary and sufficient for the uptake of cobalamin by cells.18

1.3.3.2 Inherited Intrinsic Factor Deficiency

The gastric intrinsic factor is encoded by the GIF gene and specifically produced by the parietal cells. IF is a 417-amino acid long protein and it is essential for carrying cobalamin and binding the cubam receptor to trigger endocytosis into enterocytes of the ileum.30 The rare, inherited IF deficiency is

26 caused by mutations in the GIF gene, such as the c.183_186delGAAT founder mutation.31 Mutations may abolish the production of IF, decrease its affinity for cobalamin, decrease its affinity for cubam, or increase susceptibility to proteolysis.32, 33 Symptoms present between age of 1 to 5 years in affected patients and include megaloblastic anemia, developmental delay, decreased serum cobalamin level, and mild methylmalonic aciduria (MMA) and homocystinuria

(HC).18 Inherited IF deficiency is unlike pernicious anemia in that the physiology of parietal cells is normal and autoimmune antibodies targeting those cells are not detected. In the past, the Schilling test was used to measure intestinal absorption of radiolabelled cobalamin; oral administration of labelled cobalamin is followed by intramuscular injection of a large quantity of unlabelled cobalamin to flush the radiolabel into the urine. Patients with inherited IF deficiency show decreased excretion of radiolabel in the urine compared to reference subjects because cobalamin is not absorbed into the blood and filtered by the kidneys.32, 33 The test can be repeated with the addition of an exogenous source of IF to observe an increase in the absorption of radiolabelled cobalamin. An effective medical treatment involves intramuscular injection of hydroxocobalamin (OHCbl) or

CNCbl to bypass the gastric cobalamin absorption until internal storage is restored and then lower doses to maintain blood cobalamin at an appropriate level.18

1.3.3.3 Imerslund-Gräsbeck Syndrome

Imerslund-Gräsbeck syndrome (IGS) is a disorder of defective transport of IF-Cbl complex by enterocytes. IGS is a more common form of cobalamin malabsorption than IF deficiency and is genetically heterogeneous as it can be caused by mutations in CUBN and AMN.34, 35 The CUBN and AMN genes encode cubilin and amnionless, respectively, and heterodimerization leads to the formation of cubam, which is an enterocytic receptor for IF-Cbl. Cubilin is a very large membrane protein of 460 kDa that is responsible for binding cobalamin while amnionless is a smaller membrane protein of 45-50 kDa that is responsible for trafficking of cubilin and anchoring it to the enterocyte membrane.36 The disease has primarily been studied from Finnish and Norwegian families and its prominent symptoms are megaloblastic anemia, low serum vitamin B12 level, and proteinuria possibly due to decreased cubilin-mediated uptake of albumin.36, 37 A high prevalence of the IGS in the Scandinavia has been attributed to founder effects of mutations in the two genes.38 Previously, IGS was distinguished from inherited IF deficiency based on Schilling test which was negative even with the addition of an exogenous source of IF. Since this test has been discontinued from the medical practice, mutation screening is the most accurate method of molecular diagnosis for patients with hereditary cobalamin malabsorption.39 Nonetheless, it has been noted that screening for mutations in CUBN and AMN is not an easy task because CUBN has 67 exons and AMN has a very high GC-content.40

1.3.3.4 Transcobalamin Deficiency

Transcobalamin is a 43 kDa plasma protein encoded by the TCN2 gene.41

TC-bound cobalamin account for only ~20% of the serum cobalamin but this represents the functional subgroup that can be taken up by most cell types in the body.42 Patients with the TC deficiency therefore do not show a noticeable decrease in serum cobalamin level; however, they present with megaloblastic anemia, failure to thrive, vomiting, and weakness within the first few months of life.18 Mainly protein truncating mutations, such as frameshifting insertions/deletions, nonsense mutations, and splice site mutations, have been identified in the TCN2 gene.43 These mutations either decrease the synthesis of TC protein or disrupt the ability of mutant TC to bind cobalamin.42

1.3.3.5 Transcobalamin Receptor Deficiency

The transcobalamin receptor is encoded by the CD320 gene which translates into a native protein of 282 amino acids.25 TCblR is heavily glycosylated and cleavage of the putative N-terminal signal peptide leaves a 252- amino acid long plasma membrane protein. It is composed of an extracellular domain of 199 amino acids, which contains two LDL-receptor class A domains, a transmembrane region of 21 amino acids, and a cytoplasmic domain of 32 amino acids.25 Six patients were positive for elevated MMA on newborn screening test and were referred to medical genetics laboratories. The c.262_264delGAG

29 mutation on a common haplotype was identified in 10 mutant alleles and the c.297delA mutation was identified in the other two alleles.44, 45 Cultured fibroblasts indicated diminished uptake of TC-Cbl complex as expected. Whether

CD320 mutations can lead to severe disease outcomes must be determined by follow-up studies on the reported patients who were at the time asymptomatic.

1.3.4 Related Proteins

There are two important human proteins that are neither directly involved in the assimilation of cobalamin for utilization nor involved in the intracellular metabolism of cobalamin. Megalin is a multi-ligand binding receptor mainly responsible for reabsorption of filtered TC-Cbl complex in the kidney,46 and

ABCC1 is an ABC transporter responsible for efflux of cobalamin across the basolateral membrane of intestinal epithelial cells among others.47 Presumably, both proteins function to maintain the homeostasis of cobalamin distribution.

1.3.4.1 Megalin/LRP2

Megalin/low-density lipoprotein receptor-related protein 2 (LRP2), located mainly on the apical surface of intestinal and renal epithelia, is a multiligand binding receptor and can take up various ligands, such as lipoproteins, sterols, vitamin-binding proteins and hormones.48 The TC-Cbl complex, among other proteins, filters through the kidney glomeruli and ends up in the ultrafiltrate,

30 and megalin is crucial in reabsorbing the complex at the apical brush border of proximal tubule.46 Mutations in the LRP2 gene have been identified to cause the

Donnai-Barrow and Facio-oculo-acoustico-renal syndrome.48 Clinical phenotypes common to majority of the patients are agenesis of corpus callosum, congenital diaphragmatic hernia, proteinuria, facial dysmorphology, ocular anomalies, sensorineural hearing loss, and developmental delay.49 Since these patients do not have elevated MMA or HC, megalin does not seem to have a critical impact on the cobalamin metabolism in the body.

1.3.4.2 ABCC1/MRP1

Recently, an ABC transporter named, ATP-binding cassette, subfamily C, member 1 (ABCC1)/multidrug resistance-associated protein 1 (MRP1), was identified as the exporter of cobalamin from the cells.50 As its name suggests,

ABCC1 was initially discovered as a drug efflux pump in a multidrug-resistant lung cancer cell line,51 and has been shown to confer resistance to anthracyclines,

Vinca alkaloids, epipodophyllotoxins, and heavy metal oxyanions.52 While searching for a potential ABC transporter responsible for cellular efflux of free cobalamin, Beedholm-Ebsen et al. observed 50% reduction of cobalamin efflux in human HELA cells by siRNA-mediated knockdown of ABCC1 mRNA.47 In

Mrp1(−/−) mice, a disturbance in the cobalamin homeostasis was observed; cobalamin levels were decreased in the plasma, liver, and kidney and increased in

31 the ileum and colon. It was thus hypothesized that ABCC1 governs the distribution of cobalamin throughout the body and is particularly necessary for the release of cobalamin across the basolateral membrane of polarized ileum enterocytes.47

After that study, a cohort of 18 unrelated patients and mothers of 2 additional patients, who were presumed to have hereditary cobalamin malabsorption but carried no mutations in AMN, CUBN, or GIF, was screened for sequence changes in the ABCC1 gene.53 A total of 27 changes were detected but none were deemed deleterious; 4 intronic insertions/deletions, 2 missense, 6 silent, and 15 intronic single-nucleotide polymorphisms (SNPs). Even the 2 missense mutations, creating p.G671V and p.R723Q substitutions, were reported as naturally-occurring SNPs in the population. As a result, it still remains to be seen whether mutations in ABCC1 can lead to a human disease. Given the difference between men and mice, it is possible that ABCC1 mutations manifest with phenotypes different from those seen in mice and/or different from other inborn errors of cobalamin absorption.

1.4 Intracellular Metabolism of Vitamin B12

1.4.1 The Patients

Early biochemical studies on the function of vitamin B12 led to the discovery of many cobalamin-dependent enzymes in microbes and mammals.54

Although it was then known that the human MTR and MUT enzymes depend on two forms of cobalamin, it was in 1967 that Oberholzer et al. reported the first patient with inborn error of metabolism leading to methylmalonic aciduria.55 It was soon followed by more reports of methylmalonic aciduria in children without vitamin B12 deficiency; in these children, methylmalonate excretion could be

56, 57 lowered by parenteral administration of vitamin B12. Patients were often diagnosed early in life and symptoms included failure to thrive, lethargy, hypotonia, vomiting, and severe acidosis.20

Subsequently, enzymatic studies on intact fibroblasts and cell-free liver extracts revealed the existence of vitamin B12-responsive and vitamin B12- unresponsive patients, and it was speculated that the former has a defect in metabolizing or synthesizing AdoCbl while the latter has a defect in the MUT apoenzyme.58, 59 As the number of patient reports increased, so did our understanding of the biology of vitamin B12 in humans.

1.4.2 Somatic Cell Complementation Analysis

Because most patients present at an early age and are born to unaffected parents, the inborn errors of vitamin B12 metabolism are genetic diseases with autosomal recessive mode of inheritance.55 One functional copy of a gene, mutated in patients with a recessive disorder, is enough to produce a normal phenotype, as seen in heterozygous carriers.

Somatic cell complementation has been an essential technique in assigning patient cell lines into specific complementation groups. In this experiment, two cell lines with defects in cobalamin metabolism are fused together by exposure to Sendai virus or polyethylene glycol to create heterokaryons, which are multinucleated cells carrying the genetic materials of both cell lines.60, 61 Heterokaryons are then assessed by in vitro measurements of

MTR and MUT functions to observe whether the original cells’ biochemical phenotypes can be rescued or not. If restoration does not occur, they have mutations at the identical genetic locus and are assigned to the same complementation group. If restoration occurs, then they harbour distinct genetic defects and are assigned to two different complementation groups.18

1.4.3 Discoveries of Eight Complementation Groups

Prior to the present study, eight complementation groups, cblA-cblG and mut, have been discovered in the intracellular metabolism of cobalamin.23 This thesis reports the description of patients with a novel inborn error of cobalamin metabolism, the designation of a new complementation group, and the identification of the responsible gene. Thus, I will describe the discoveries of eight complementation groups in the following subsections, and then provide the genetic and functional information pertinent to each gene in Section 1.4.4.

1.4.3.1 Four Complementation Groups

First inborn errors of vitamin B12 metabolism were described from patients with MMA due to decreased activity of the MUT enzyme. When a patient with MMA, HC, and cystathioninemia was reported, it was deduced that the patient’s metabolic block was located before the cobalamin metabolism branches into AdoCbl synthesis and MeCbl synthesis, leading to decreased activities of both MUT and MTR.62, 63 While the patients shared MMA as a common phenotype, patient cell lines could be biochemically differentiated into four groups: mut mutants produced defective MUT apoenzyme; cblA mutants failed to synthesize AdoCbl in intact fibroblasts but had normal synthesis in crude cell extracts; cblB mutants failed to synthesize AdoCbl in both intact fibroblasts and crude cell extracts; cblC mutants failed to synthesize both AdoCbl and MeCbl.64

Genetic heterogeneity was confirmed by somatic cell complementation experiments which demonstrated enhancement of MUT enzyme function, as shown by [14C]-labelled propionate incorporation, when cells from different groups were fused together (see Section 2.4). As a consequence, the existence of four distinct gene loci was proposed.61

Additionally, the mut defect was subdivided into two categories; muto mutants had undetectable enzyme activity (~0.1%) and were refractory to increasing doses of OHCbl, and mut− mutants had detectable residual enzyme activity (0.5-50%), decreased binding affinity to AdoCbl, and displayed dose-

35 dependent responsiveness to OHCbl.65, 66

1.4.3.2 A Complementation Group with Three Phenotypes

Two patients, who were two brothers in a sibship of seven, were reported to have both MMA and HC.67 This biochemical phenotype was comparable to cblC but dissimilar from cblA, cblB, or mut due to the presence of HC. They were not included in the previous study that classified the initial four complementation groups. Later, during an endeavour by Rosenberg et al. to confirm and assign 21 new patients into the four complementation groups, they observed that cell lines from the two brothers complemented with cells from all four complementation groups. Accordingly, they were assigned into a new group, cblD.68

A puzzling situation arose when a patient, initially suspected of the cblA defect, was found to complement with 28 cblA lines.60, 69 The answer came when three patients, classified as cblD by somatic cell complementation, presented with two biochemical phenotypes that were different from the original description; two patients presented with isolated HC and one patient presented with isolated MMA.

These subgroups were then designated as cblD-variant 1/cblD-HC and cblD- variant 2/cblD-MMA, respectively.70 The classic cblD first identified in the sibship has been referred to as cblD-combined or cblD-MMA/HC in later reports.

1.4.3.3 Isolated HC with Megaloblastic Anemia

Since the early discoveries of patients with inborn errors of cobalamin metabolism, it was presumed that there may be patients with HC, but not MMA, due to a specific deficiency in the synthesis of MeCbl or activity of MTR apoenzyme.71 The first case of isolated HC was eventually identified when an infant was described with severe developmental delay, megaloblastic anemia, and

HC. The patient, whose defect was designated cblE, showed no elevation of methylmalonic acid or deficiency of folate and cobalamin, and he was responsive to injections of OHCbl but not to folate.72 From the observation that MeCbl synthesis and 14C-labelled 5-methylTHF incorporation were decreased in intact fibroblasts but methionine synthase activity was normal in crude cell extracts, the authors inferred that the defect was in intracellular cobalamin metabolism to synthesize MeCbl. It was noted that β-mercaptoethanol (BME), a reducing agent, in the reaction mixture might have obscured the findings since the reduction of cob(III)alamin to cob(I)alamin as a prerequisite to MeCbl synthesis had been well documented.72

Interestingly, when methionine synthase activity was measured without

BME, the patient’s cell extract showed significantly lower enzyme activity than the control.73 The MTR activity could be increased by adding increasing concentrations of dithiothreitol (DTT), another reducing agent, into the reaction mixture. Therefore, it was postulated that cblE is a defect in a reduction step

37 required for MTR activity. The authors remarked that E. coli employs two flavoproteins as a coupled reducing system to maintain the reduced state of MTR- bound cobalamin.73

1.4.3.4 Failure of Lysosomal Release of Vitamin B12

A defect in translocation of cobalamin from lysosome to cytoplasm was reported in a single patient with MMA, developmental delay, and no HC or megaloblastic anemia.74 After 4-day incubation with 25 pg/mL of [57Co]-labelled

CNCbl, cultured fibroblasts showed an accumulation of free cobalamin as exogenous, unmetabolized CNCbl, suggesting an inability to transfer cobalamin from the lysosome to the cytoplasm after release from TC. When cell extracts were fractionated by Percoll gradient, 93% of radioactive labels in patient cells were located in the lysosomal and mitochondrial fraction while 62% of labels in control cells were located in the lower density, cytoplasmic fraction. A defect in endocytosis of TC-Cbl complex was excluded because addition of chloroquine, which arrests lysosomal proteolysis, resulted in accumulation of TC-Cbl in control and patient fibroblasts alike.74 The patient cell line complemented with all five known complementation groups with MMA and was designated cblF.75

1.4.3.5 Heterogeneity Among Patients with HC

After the first identification of a patient with the cblE defect, more

38 patients with isolated HC and megaloblastic anemia were reported. However, a subgroup of patients were reported to be biochemically distinct from the earlier patients. The methionine synthase activities in patients’ cell extracts were decreased when suboptimal or optimal concentrations of reducing agent, BME or

DTT, were added.76, 77 Somatic cell complementation and measurement of methionine synthase activity demonstrated genetic heterogeneity among patient cell lines; the defect in patients with normal MTR activity was retained as cblE, and the new defect in patients with decreased MTR activity was named cblG.

Furthermore, the authors raised the possibility that the cblG mutations may affect the MTR enzyme itself.78

1.4.4 Inborn Errors of Vitamin B12 Metabolism

Based on the current clinical, biochemical, and genetic knowledge of the genes involved in the vitamin B12 metabolism, a model pathway has been constructed (Figure 2).18 After TC-Cbl-TCblR supercomplex enters the cell by receptor-mediated endocytosis, the late endosome fuses with the lysosome, TC is rapidly degraded by lysosomal proteases, and free cobalamin is released inside the lysosomal lumen.79 Cobalamin in the cytoplasm is further modified by a multifunctional chaperone protein, and it either remains in the cytoplasm for

MeCbl synthesis and MTR activity or gets transported into the mitochondria for

AdoCbl synthesis and MUT activity.18 Identifications of eight genes in the

39 pathway have further elucidated the intricate biology of genes and gene products in cobalamin metabolism. Nevertheless, information on the functions and structures of proteins is still very limited and remains to be studied.

The known eight inborn errors, each the result of mutations at a unique genetic locus, have been divided into three clinical phenotypes depending on the location of the defect in the cobalamin pathway. Combined MMA and HC is seen in the cblF, cblC, and cblD-combined defects which block the early passage of cobalamin through the cell. Isolated MMA is the result of cblD-variant 2, cblB, cblA, and mut defects which affect the synthesis and utilization of AdoCbl. On the other hand, isolated HC is caused by cblD-variant 1, cblE, and cblG defects which affect the synthesis and utilization of MeCbl.18

1.4.4.1 cblF, cblC, and cblD

Gene identification for the cblF defect involved homozygosity mapping of 12 unrelated patients and microcell-mediated chromosome transfer.80 In the latter technique, microcells carrying a single copy of each human chromosome were fused in sequence with a patient cell line until the specific chromosome that corrected the biochemical phenotype was detected. The success of both techniques led to the discovery of mutations in the LMBRD1 gene, which encodes the LMBD1 (LMBR1 domain-containing protein 1) protein.80 Fifteen unrelated patients have been reported with the disease and a common c.1056delG mutation

Figure 2. Vitamin B12 metabolism. This pathway outlines known steps of intracellular B12 metabolism carried out by genes responsible for each complementation class. In this simplified diagram, circles represent cobalamins, arrows represent metabolic steps in the pathway, and italicized names represent disease names/complementation classes. Proteins that bind to cobalamin molecules inside the cell, such as MMACHC, MTR, and MUT, are not depicted.

Each disease in cobalamin metabolism results in the blockage of a metabolic step(s) shown by the arrow(s). TC, transcobalamin; TCblR, transcobalamin receptor.

41 occurred in 13 out of 15 cases.81, 82 Based on the protein’s strong homology to

LMBR1 (limb region 1) and LIMR (lipocalin-1 interacting membrane receptor),

LMBD1 was predicted to be a lysosomal membrane protein with nine transmembrane regions (Figure 3).80 Because the cblF defect disrupts the cobalamin transport from lysosomes to cytoplasm,83 LMBD1 was characterized as a putative lysosomal cobalamin exporter. Various aspects of cblF are discussed in greater detail in Section 1.5.

The cblC defect is the most common inborn error of cobalamin metabolism with more than 500 patients from all over the world reported to date.

The gene commonly mutated in the patients was identified by linkage analysis, homozygosity mapping, and haplotype analysis and was named MMACHC

(methylmalonic aciduria cblC type, with homocystinuria).84 The primary sequence of MMACHC protein does not show any resemblance to known protein families, but residues 181-282 of MMACHC display homology to residues 152-239 of bacterial TonB protein, which has a role in cobalamin transport across the outer membrane.84 This cblC locus has long been speculated to encode a reductase of cob(III)alamin to cob(II)alamin before conversion into MeCbl and AdoCbl and/or a chaperone.2, 85, 86 Although the primary function of MMACHC protein is still uncertain, this seemingly multifunctional, cytosolic protein is shown to induce decyanation and dealkylation of cobalamin derivatives and conversion of newly

42 internalized cobalamin into the “base-off” conformation which may be necessary for binding to MTR and MUT.87, 88 This molecular chaperone interacts with

MMADHC for an unclear purpose.89 Structural study of MMACHC reported that it is a divergent member of the NADPH-dependent flavin reductase family and uses FMN or FAD to catalyze decyanation of CNCbl.90 AdoCbl binds to the protein’s nitroreductase scaffold in a base-off, five-coordinate configuration and glutathione binds to an adjacent arginine-rich pocket for the dealkylation of various alkylcobalamins. Moreover, cobalamin binding triggers a dimerization event and the PNRRP loop of each monomer provides capping for the upper axial ligand.91

The genetic etiology of the cblD defect was studied by microcell- mediated chromosome transfer and refined genetic mapping, and it was found to be caused by mutation in the MMADHC (methylmalonic aciduria cblD type, with homocystinuria) gene.92 A unique characteristic of the cblD defect is that it can present with three distinct biochemical findings. As a result, it has since been sub- classified into cblD-combined (MMA and HC), cblD-variant 2 (MMA), and cblD- variant 1 (HC).70, 93 A total of 17 patients have been reported; five patients with cblD-combined, six with cblD-variant 2, and six with cblD-variant 1.94

Investigations into the genotype-phenotype correlation of three variants found that truncating mutations close to the N-terminus in cblD-variant 2 allow for re-

43 initiation of translation at Met62 or Met116 and cytoplasmic MeCbl synthesis is unaffected while mitochondrial activity is compromised. Meanwhile, missense mutations toward the C-terminus, mostly between residues 246 and 259, in cblD- variant 1 result in particular abrogation of cytoplasmic activity and leaves mitochondrial AdoCbl synthesis intact.94 This domain is one of the five predicted interaction sites between MMADHC and MMACHC and missense mutations may affect proper binding of MMADHC for its cytoplasmic activity.89 Coupled decrease in AdoCbl and MeCbl synthesis is caused by truncating mutations downstream of Met116 which remove large portions of the C-terminal domain.

Initial analysis of the protein sequence identified a putative cobalamin-binding motif, a putative mitochondrial targeting sequence, and a stretch of 91 amino acids with similarity to the ATPase component of a putative ATP-binding cassette transporter although the significance of this similarity has not been validated.92

Despite the lack of knowledge regarding MMADHC function, it can be projected that this protein has a role in target-directed distribution of cobalamin to mitochondrial or cytosolic compartments for the usages by MUT or MTR, respectively.

1.4.4.2 cblB, cblA, and mut

The MMAB (methylmalonic aciduria cblB type) gene encodes for

ATP:cob(I)alamin adenosyltransferase or MMAB protein, and mutations in this

44 gene are the causes of the cblB defect.95 The gene was identified by selecting human orthologs of bacterial genes in operons containing the methylmalonyl-CoA mutase gene and verifying by finding mutations in cblB patients. Knowing that the gene of interest may encode an adenosyltransferase was the key to this success and its 45% similarity to adenosyltransferase PduO from Salmonella provided further confirmation. MMAB contains a mitochondrial leader sequence and converts the reduced cob(I)alamin to AdoCbl in the mitochondria.95 Crystal structure of MMAB demonstrated that it exists as a homotrimer and sterically favours the binding of ATP over GTP.96 Although the authors did not produce a structure with a bound cobalamin, they hypothesized that an oval-shaped opening near the ATP-binding active site may allow positioning of cobalamin for catalysis.

Functional analysis of mutations in patients has suggested that majority of mutations, clustered in exon 7, affect highly conserved residues and decrease the catalytic activity of the enzyme.96-98 The mechanism by which cob(II)alamin is reduced to cob(I)alamin before adenosylation remains unsolved. Evidence suggests that one or more unknown proteins may bind cobalamin in the

20 mitochondria and have a role in B12 metabolism.

The gene responsible for the cblA defect was discovered by Dobson et al. in parallel with MMAB identification using the same approach of homology searching.99 It was appropriately named MMAA (methylmalonic aciduria cblA

45 type) and deleterious mutations were detected in five cblA patients. Subsequently, many patients with the cblA defect have been reported and most of the mutations are located in exons 2 to 4 including the common c.433C>T (~43%) mutation.100-

102 Although it was described as a member of the G3E family of P-loop GTPases and was speculated to be an accessory protein in mitochondrial translocation of cobalamin, unraveling its exact function has long been elusive.99 Pathobiology of

MMAA mutations was initially studied by mapping point mutations onto the bacterial MeaB counterpart and mutations were predicted to damage the folding and stability of MMAA.103 The recently solved crystal structure of MMAA provided greater insight into its function and revealed it to form a homodimer.104

Interestingly, it also demonstrated that MMAA and MUT interact with each other preferably when MMAA is bound to GMPPNP, a nonhydrolyzable GTP analog, and when MUT is not bound to its cofactor. In the mitochondrial milieu, MMAA plays a gatekeeping role in which it channels AdoCbl to MUT and maintains

AdoCbl bound to MUT holoenzyme in its active form.104, 105

The mut defect is caused by mutations in the MUT enzyme itself, encoded by the MUT gene.106, 107 Identified by screening human liver and placenta cDNA expression libraries with anti-MUT antibody, this was the first gene in the vitamin B12 pathway to be cloned. The mut defect represents the most common cause of isolated MMA and mutations are frequently found in exons 2, 3, 6, and

11.23, 108 More patients have been reported to have the muto defect with undetectable enzyme activity than the mut− defect with residual enzyme activity.100, 108 The MUT enzyme has a 32-amino acid mitochondrial leader sequence and catalyzes the isomerization of methylmalonyl-CoA to succinyl-CoA, an intermediate molecule in the Krebs cycle which can then be completely metabolized. Structural and size exclusion chromatography studies estimate MUT homodimers to interact with MMAA homodimers in 1:2 ratio to form the core structure of cobalamin assimilation machinery.104

1.4.4.3 cblE and cblG

The cblE defect is caused by mutations in the MTRR gene which encodes methionine synthase reductase (MTRR).109 Based on enzymatic assays on patient fibroblasts and comparison to flavodoxin/flavodoxin reductase system in E. coli, it was deduced that the cblE complementation group is a disorder of reductive reactivation of MeCbl bound to MTR.78 Searching for FMN, FAD, and NADPH binding sites, Leclerc et al. discovered MTRR to be a dual flavoprotein reductase capable of binding all three cofactors. On the primary protein sequence, the flavodoxin domain is positioned near the N-terminus and the flavodoxin reductase domain near the C-terminus.109 Various sequence changes lead to truncation, inclusion of a pseudoexon, or missense mutations affecting the cofactor-binding sites.110, 111 The cblE patients often present early in life with neurological findings

47 and megaloblastic anemia.18 Cob(I)alamin, which is always bound to the active holoenzyme, can be inactivated by accidental oxidation once in approximately every 2000 catalytic cycles.8 MTRR performs reductive methylation of inactive cob(II)alamin using S-adenosylmethionine as a methyl donor to regenerate MeCbl and maintain the activity of MTR.109

The cblG defect results from mutations in the MTR gene encoding the

MTR enzyme.19 The MTR cDNA was cloned by performing RT-PCR with primers to four homologous regions of methionine synthases from different organisms and reconstituting the human cDNA. MTR catalyzes methyltransferase reactions during which a methyl group is transferred from 5-methylTHF to MeCbl and then to homocysteine, producing THF and methionine. For its function, MTR is divided into a homocysteine-binding domain, a 5-methylTHF-binding domain, a cobalamin-binding domain, and an activation domain.112 The enzyme recycles the bound cob(I)alamin and accidental oxidation of cob(I)alamin to cob(II)alamin can be restored by MTRR.109 Mutations have been identified throughout the entire gene in over 30 patients.112, 113 Patients suffer from megaloblastic anemia, as in cblE, because inadequate amount of THF is reverted back to 5,10-methyleneTHF for the production of thymidylate. Megaloblastic anemia seen in cblE and cblG is indistinguishable from that seen in folate deficiency.114

1.4.5 Other Causes of MMA

It is important to note that the eight defects described above are not the only causes of methylmalonic aciduria. For instance, mild methylmalonic aciduria can be caused by mutations in methylmalonyl-CoA epimerase (MCEE), which converts D-methylmalonyl-CoA into L-methylmalonyl-CoA,115 or by mutations in the succinyl-CoA ligase (SUCLG1, SUCLA2), which metabolizes succinyl-CoA through the Krebs cycle.116 Recent exome sequencing endeavours discovered mutations in the putative malonyl-CoA and methylmalonyl-CoA synthetase

(ACSF3) that result in combined malonic and methylmalonic aciduria;117, 118 the more common etiology for this metabolic phenotype is deficiency of the malonyl-

CoA decarboxylase (MLYCD), which converts malonyl-CoA into acetyl-CoA.119

1.4.6 Other Causes of HC

Furthermore, accumulation of homocysteine can be the result of decreased activities of three enzymes: MTR, 5,10-methyleneTHF reductase

(MTHFR), and cystathionine β-synthase (CBS).114 MTR, as described earlier, converts homocysteine to methionine, MTHFR converts 5,10-methyleneTHF to

5-methylTHF which is the methyl donor for MTR,120 and CBS catalyzes condensation of homocysteine and serine to form cystathionine.121 Mutations affecting the three proteins have been identified and they lead to varying degrees of homocystinuria in patients. Although phenotypically overlapping, MTHFR

49 deficiency and CBS deficiency are not inborn errors of cobalamin metabolism.

1.5 cblF Disease

This thesis, although not focused on the cblF disease, has tight connections with it. Therefore, the exact biology of this disease, including the passage of cobalamin through the lysosomal compartment of the cell, must be addressed to provide a proper context to succeeding sections.

1.5.1 Background on Lysosome

1.5.1.1 Role of Lysosome

Lysosomes are ubiquitous membrane-bound organelles found in nearly all human cell types, except erythrocytes, and they mainly function as the primary degradative compartment. Substrates enter the lysosomes via endocytosis, phagocytosis, or autophagy, and are digested by the acidic environment and hydrolase enzymes in intra-lysosomal vesicles. It has been estimated that lysosomes contain ~50 soluble hydrolases and ~25 integral lysosomal membrane proteins (LMPs), some of which carry particles between lysosomal lumen and cytosol. The acidic pH is essential for the activity of hydrolases and is maintained by vacuolar-type H+-ATPase, a proton pump.122, 123

1.5.1.2 Lysosomal Soluble Proteins

Lysosomal soluble proteins are synthesized at the rough endoplasmic reticulum (ER) and are translocated to the ER lumen by the N-terminal targeting sequence. Inside the ER, the targeting sequence is cleaved and proteins undergo

N-glycosylation before they are transported to the Golgi apparatus.122

Glycosylation is performed by oligosaccharyltransferase which attaches an oligosaccharide chain to the asparagine residue of “Asn-X-Ser/Thr” tripeptide motif where X is any amino acid but proline.124 Lysosomal hydrolases, with few notable exceptions, depend on the mannose-6-phosphate (M6P) tag for transport from the Golgi network to the lysosomes. The M6P moiety is added to the hydrolases by the actions of N-acetylglucosaminyl-1-phosphotransferase and N- acetylglucosamine-1-phosphodiester α N-acetylglucosaminidase (diesterase).

Hydrolases are then recognized by the M6P receptors on the Golgi membrane and are transported to the lysosomes.125 M6PR-independent transport pathways have been reported also; LIMP2/SCARB2 serves as a transport receptor for β- glucocerebrosidase and sortilin serves as a transport receptor for neurotensin, lipoprotein lipase, and the precursors of sphingolipid activator proteins.125

1.5.1.3 Lysosomal Membrane Proteins

LMPs do not acquire the M6P tag and are thus not transported through the

M6PR-dependent pathway. Studies have suggested that LMPs can either travel

51 directly from the Golgi network to the lysosome or travel first to the plasma membrane and then reach the lysosomes by a retrograde endocytic pathway.123

Although our understanding of the transport mechanism for LMPs is still limited, the LAMP/LIMP family of proteins and several other LMPs contain the tyrosine- based (YXXØ ) or dileucine-based ([DE]XXXL[LI]) motifs, which are common

Golgi sorting and endocytic signals. The lysosomal targeting seems to be conferred by the placement of these motifs near the transmembrane domain and generally also near the C-terminal domain, but the exact relationship has not been proven.125 LMPs are glycosylated on the luminal domain while in the Golgi network through the same N-glycosylation mechanism as the soluble proteins.123

1.5.1.4 Disorders of Lysosomal Export

Mutations in genes encoding lysosomal hydrolases or LMPs result in various metabolic diseases known as lysosomal storage diseases. They are characterized by the intra-lysosomal accumulation of specific substrates.122 For example, Tay-Sachs disease is caused by mutations in β-N-acetylhexosaminidase

A enzyme and an accumulation of GM2 ganglioside and related glycolipids is observed.126 The cell biology and pathophysiology of LSDs have been extensively reviewed elsewhere.122, 126

A subset of these diseases results from a failure of lysosomal export of a substrate by a transporter on the lysosomal membrane. Two disorders of

52 lysosomal export are well-characterized; cystinosis is caused by mutations in cystinosin which transports cystine, a dimeric amino acid, out of the lysosome, and Salla disease is caused by mutations in sialin which transports sialic acids and acidic hexoses out of the lysosome.123 Similarly, the cblF defect and the novel defect identified in this thesis are caused by failures of lysosomal export of cobalamin. However, the pathology of these defects in cobalamin transport is caused by the deficiency of substrate in the cytoplasm and mitochondria while the pathology of cystinosis and Salla disease is attributed to the accumulation of substrates in the lysosome.127

1.5.2 LMBRD1 Gene

The cblF disease is a rare disorder of lysosomal export caused by mutations in the LMBRD1 gene. The gene discovery was achieved by a combination of microcell-mediated chromosome transfer with homozygosity mapping. Twelve unrelated patients in the study were from diverse ethnic backgrounds, but a common haplotype of 1.34 Mb was identified on chromosome

6q13. The high frequency of the c.1056delG mutation, which occurs in 20 out of

30 mutant alleles in 15 patients, is predicted to be the result of a founder effect.80,

82 Eight other mutations have been identified in these patients; c.515_516delAC and c.1405delG are homozygous mutations in two patients and c.404delC, c.712_713delAC, c.842_845delAGAG, c.916-1G>T, c.1339-1G>T, and c.70-

4298_246+2311del6785 are compound heterozygous mutations with the common c.1056delG.82 All reported mutations are either frameshifting short deletions, splice acceptor site mutations, or a large deletion, and therefore truncate the encoded protein.

LMBD1 is a lysosomal membrane protein with nine transmembrane regions and 6 N-glycosylation sites on the intra-lysosomal face (Figure 3). Its primary sequence shares 15.9% and 13.7% identity, or 31.5% and 27.8% similarity, with LMBR1 and LIMR proteins, respectively.80 The function of LIMR is to internalize lipocalins, which are extracellular carriers of lipophilic compounds. The mechanism by which LMBD1 mediates cobalamin transport across the lysosomal membrane remains unclear.

1.5.3 Pathophysiology and Treatment

Clinical and pathological findings in an LSD can often be attributed to the accumulation of a specific substrate in a cell type/ tissue.126 Unlike other LSDs, the disease phenotypes in cblF are not direct consequences of cobalamin accumulation in the lysosome, but outcomes of decreased availability of active cofactors to MTR and MUT enzymes. Enlargement of cells or tissues is not observed most likely because cobalamin is not endogenously synthesized and does not need to be broken down.

Common clinical findings in the cblF patients, aside from MMA and HC,

Figure 3. Putative membrane topology of the LMBD1 protein.

Transmembrane regions, numbered from 1 to 9, are schematically shown in grey across the lysosomal membrane. Small numbers inside a rectangle represent the start and end residues of each transmembrane region. Six putative N-glycosylation sites are shown in green with numbers corresponding to the positions of asparagine residues. Adapted from Rutsch et al., 2009.

55 include failure to thrive, developmental delay, low serum cobalamin level, abnormal Schilling test results, small for gestational age, feeding difficulties, C3 carnitine elevation secondary to MMA, stomatitis/glossitis, megaloblastic anemia, hypotonia, and cardiac defects.128 Symptoms often present during the first year of life and they strongly suggest the importance of cobalamin metabolism in early child development. Patients diagnosed with cblF are treated by parenteral administration of 1.0 to 1.5 mg of OHCbl per day until homocysteine and methylmalonic acid levels normalize. The amount of OHCbl is then titrated to a minimal dose with optimal biochemical response in the patient.128

1.6 Undiagnosed Patients and Gaps in the Pathway

Patients are suspected to have an inborn error of cobalamin metabolism when they present with methylmalonic aciduria and/or homocystinuria among other symptoms. If dietary causes are ruled out, their skin fibroblasts are obtained for somatic cell complementation and biochemical assays, and their blood- extracted DNA are screened for mutations in known genes. Over the years, certain patients could not be genetically diagnosed with any of the known defects. They were either determined to carry no mutations in all genes associated with cobalamin metabolism, showed normal phenotype upon biochemical assays with fibroblasts, or could not be successfully classified by somatic cell complementation. Focused investigation into each unsolved case could lead to the

56 identification of mutations, but even with these limited successes, no patient has been correctly assigned into a new complementation group since cblG was designated as the eighth in 1988.78

Nevertheless, it has always been a possibility that there are more genes involved in vitamin B12 metabolism. Mutations in new genes may not have been reported due to reasons ranging from functional redundancy to severity of mutation that is incompatible with life.23 Presently, there are question marks scattered across the vitamin B12 metabolism pathway and the clinicians and scientists in the field are devoted to finding answers. Two proteins, seemingly necessary to complete the cobalamin metabolism pathway, have not been accounted by the known eight genes; one is a mitochondrial membrane transporter of cobalamin and the other is an adapter or energy-provider to LMBD1 for lysosomal egress of cobalamin.23, 129 If determined to be true, they would both be involved in the transport of cobalamin across organellar membranes.

1.7 Evolution of Gene Discovery Approaches

1.7.1 Limitations of Traditional Approaches

Genetic diseases in human can be roughly divided into rare, monogenic, simple diseases and common, multigenic, complex diseases. The former were mostly investigated by linkage analyses and Sanger sequencing of candidate genes130 while the latter were subjected to genome-wide association studies

(GWAS).131 This dichotomy by no means encapsulates the entirety of genetic diseases, and the other diseases that fall between the two classifications were studied by combinations of various gene discovery approaches. Over the many successes with linkage analysis and GWAS, scientists have witnessed their limitations also; linkage analysis required well-characterized families and were very time-consuming, while GWAS required the recruitment of a large cohort of cases and controls and only a few associations could be consistently reproduced.132

1.7.2 First Applications of Exome Sequencing

Rare, monogenic diseases, also termed Mendelian diseases for their classical pattern of inheritance, often result from major changes to protein sequences encoded by the mutated genes. Hence, the human exome, which refers to all protein-coding sequences called exons in the genome, represents an enriched source of disease-causing genetic variants.133 Given that the exome is

~1% of the human genome134 and that missense, nonsense, small insertion/deletion/indel, and splicing mutations account for ~88% of disease- associated mutations (Human Gene Mutation Database; http://www.hgmd.org), whole-exome sequencing presented a far more cost-effective method of disease gene discovery than whole-genome sequencing. With the development of next- generation sequencing technologies, high-throughput DNA sequencing is

58 becoming faster, cheaper, and more practical than ever before.

In 2009, Ng et al. published a seminal work in modern human genetics with their proof-of-principle paper on disease gene discovery by exome capture and massively-parallel sequencing. They sequenced exomes of twelve individuals, including four unrelated patients with Freeman-Sheldon syndrome, and confirmed that all four patients carried mutations in the MYH3 gene, without utilizing any additional data.134 Subsequently, the same team demonstrated gene discovery in

Miller syndrome, for which the genetic cause was unknown, by exome sequencing and it was approved as an accurate and efficient approach.135 Their method was quickly taken up and improved by the scientific community to identify the causes for numerous Mendelian disorders.136 Although the name derives from exome capture and massively-parallel sequencing, it is unarguable that a critical component of exome sequencing is really the bioinformatics. The complex process of using bioinformatics tools to align DNA sequences, detect variants, filter out false positives and false negatives, and algorithmically select damaging variants is what differentiates a successful exome sequencing project from one that is not.

1.7.3 New Paradigm of Disease Gene Discovery

In traditional approaches, it was not a simple task to identify a disease- causing gene for a rare Mendelian disorder from a small cohort of unrelated

59 patients. Reasons for such difficulty included availability of small number of cases, uncertainty of degree of penetrance, locus heterogeneity, and low reproductive fitness.136 As an extreme example, the identification of heterozygous de novo mutations in a few unrelated patients with an autosomal dominant disorder was nearly impossible. But whole-exome sequencing is an efficient, unbiased method of identifying a disease-causing gene commonly mutated in patients with an inherited disorder.137, 138 An efficient workflow for disease gene discovery using exome sequencing would be to first select patients with definitive, textbook phenotypes, sequence their exomes, find a commonly mutated gene, and then follow-up by Sanger sequencing the gene in other patients with the disorder.139

1.7.4 Exome Sequencing of Undiagnosed Patients

Patients with methylmalonic aciduria and/or homocystinuria that could not be diagnosed before are now candidates for exome sequencing approach to find novel mutations in genes presently not included in the vitamin B12 pathway.

Although many are unrelated patients with unique, disparate phenotypes, it was hypothesized by Dr. David Rosenblatt that exome sequencing of a single proband with an autosomal recessive disease could lead to the identification of the disease- causing gene by meticulously filtering for variants, selecting candidate genes, and validating by Sanger re-sequencing.140

Affirming the validity of this method, exome sequencing was performed on the genomic DNA of a patient with megaloblastic anemia, severe combined immunodeficiency, hyperhomocysteinemia, and methylmalonic aciduria.141 It resulted in the discovery of mutations in the MTHFD1 gene which encodes a trifunctional enzyme that acts as methyleneTHF dehydrogenase, methenylTHF cyclohydrolase, and formylTHF synthetase. Because the MTHFD1 protein is strictly involved in the folate metabolism including the generation of 5- methylTHF, the presence of methylmalonic aciduria in the patient was considered incidental. The present study, as will be discussed later, was also initiated by exome sequencing of an unusual patient for whom a genetic diagnosis could not be made by traditional methods.

1.8 Peroxisomal ABC Half-Transporters

1.8.1 ATP-Binding Cassette Transporters

ATP-binding cassette (ABC) transporters have diverse structural designs, membrane orientations, mechanisms of action, and transported substrates in prokaryotes and eukaryotes.142 Their common characteristic is that they all harvest

ATP hydrolysis for energy production and, in most cases, a full, functional transporter is constituted by two transmembrane domains (TMD) and two nucleotide-binding domains (NBD). The diversity in transport and regulation mechanisms originates from protein sequence differences, structural variability,

61 and existence of additional domains in certain cases.

Human ABC transporters have been grouped into seven subfamilies,

ABCA to ABCG, based on structural and functional properties. The most studied transporters have often been the ones tightly associated with human diseases; P- glycoprotein/MDR1/ABCB1 in multidrug-resistant cancer;143 CFTR/ABCC7 in cystic fibrosis;144 ALD/ABCD1 in X-linked adrenoleukodystrophy;145 and

CERP/ABCA1 in Tangier disease.146

1.8.2 ATP-Binding Cassette, Subfamily D

The peroxisomes are ubiquitous membrane-bound organelles with various specialized metabolic functions. These include β-oxidation of fatty acids, especially very long chain fatty acids (VLCFA), hydrogen peroxide detoxification, synthesis of bile acid, plasmalogen, and cholesterol, glyoxylate detoxification, and lysine catabolism.147, 148 This organelle utilizes hydrogen peroxide to perform some of its reactions, and the generation and degradation of hydrogen peroxide are tightly regulated to prevent this strong oxidizing agent from inflicting damage to the cell itself.

The ABC subfamily D is a group of four ABC half-transporters containing one TMD and one NBD (Figure 4); ABCD1/ALDP,145

ABCD2/ALDR,149 ABCD3/PMP70,150, 151 and ABCD4/P70R.152, 153 Among them,

ABCD1, ABCD2, and ABCD3 are located on the peroxisomal membrane and

Figure 4. Putative membrane topology of an ABC, Subfamily D transporter.

Members of the ABC subfamily D each contain one TMD and one NBD, so are half the size of a full ABC transporter. The TMD contains six transmembrane regions in the N-terminal domain of the transporter, and the NBD contains Walker

A, Walker B, and ABC signature motifs in the C-terminal domain. Adapted from

Morita & Imanaka, 2012.

63 they function to translocate VLCFA from the cytosol to the peroxisomes.148 It has been suggested that they homo- or heterodimerize to form a full, active transporter of VLCFA with varying substrate specificities.154, 155 ABCD1 is the most-studied peroxisomal ABC transporter because mutations in the gene are responsible for causing X-linked adrenoleukodystrophy.145 The disease is biochemically characterized by elevated levels of saturated VLCFA in the plasma and tissue, and very long chain acyl-CoA synthetase was considered a candidate gene for this disease until it was proven otherwise.

1.8.3 ABCD4 Gene

1.8.3.1 Discovery and Characterization

The fourth member of the ABC subfamily D was identified by Shani et al. and Holzinger et al. independently by searching the human expressed sequence tags database for a cDNA clone with homology to ABCD1 and ABCD3.152, 153 The

ABCD4 (ATP-binding cassette, subfamily D, member 4) gene is located on chromosome 14 and maps to q24.3 region.152 The gene consists of 19 coding exons, which produce a 1821-base pair mRNA transcript, which in turn produces a 606-amino acid polypeptide with a predicted molecular mass of 68.6 kD.153

Although ABCD4 was initially reported as a peroxisomal half-transporter as a result of its homology to the others, it is the most divergent member of the subfamily (Figure 5) and its N-terminus begins with a hydrophobic

64 transmembrane region while the others begin with a hydrophilic cytosolic segment.156 The ABCD4 protein, albeit not well-characterized, contains the two principle domains of ABC transporters: a TMD (residues 39-332) and an NBD

(residues 389-603) (http://www.uniprot.org/uniprot/O14678).

1.8.3.2 Subcellular Localization

Although initially reported to be peroxisomal,152 a more recent study suggested that ABCD4 is localized to the endoplasmic reticulum.156 Moreover, proteomic studies aimed at characterizing all peroxisomal proteins in rat and mouse peroxisomes failed to detect ABCD4 under conditions where the three other ABCD proteins were detected.157-159 In addition, an in vitro study reported that PEX19p, a peroxisomal biogenesis protein, binds ABCD1, ABCD2 and

ABCD3, but not ABCD4.160 Its lack of N-terminal hydrophilic region has been suggested to be the cause of its deviation from peroxisomal targeting.156

Figure 5. Protein sequence identity and similarity among ABCD proteins.

NCBI blastp (http://blast.ncbi.nlm.nih.gov) was operated by inputting the primary sequence of each protein (vertical list) and retrieving the percentage identity and percentage similarity (shown in brackets) against the other transporters (horizontal list). Four proteins always aligned against each other with lowest E-values amongst all human proteins in the Swissprot database. Protein sequence identity is the number of identical residues divided by the length of overlapping sequence.

Protein sequence similarity is the sum of number of identical residues and number of conserved residues divided by the length of overlapping sequence.

RATIONALE AND OBJECTIVES OF STUDY

Despite recent advances in understanding various aspects of vitamin B12 metabolism, many questions still remain to be answered. Intracellular metabolism of vitamin B12 in humans has been found to involve at least eight genes with each gene product responsible for one step of the pathway. Recently, our laboratory investigated three patients with phenotypes resembling the cblF defect but not belonging to any of the known eight complementation groups. My objective was to identify mutations in the gene causing this novel inborn error of vitamin B12 metabolism and confirm its role by transfecting cultured fibroblasts with a wild type copy of the gene. Gene identification was accomplished by leveraging the strength of exome capture and massively-parallel sequencing (exome sequencing), which had become a new paradigm for gene identification in rare Mendelian diseases. Correction of biochemical phenotypes was demonstrated by performing assays designed to measure different parameters of normal versus abnormal vitamin B12 metabolism. Enzymatic functions of MTR and MUT, synthesis of active cofactors from labelled CNCbl, and proportions of free and protein-bound state of cobalamin were assessed. All experiments, except those acknowledged to be performed by others, were performed by the author.

CHAPTER 2: MATERIALS AND METHODS

2.1 Case Reports

2.1.1 Patient WG4066

Patient WG4066 was the second child of healthy, non-consanguineous parents of northern European ancestry. She was noted to have hypotonia, lethargy, feeding difficulties, periodic breathing, and episodes of posturing. There was evidence of bone marrow suppression requiring red blood cells and platelet transfusions. Newborn screening showed an elevated C3 (propionyl) carnitine at

11.9 (reference <5) µmol/L. Initial laboratory testing failed to show metabolic or lactic acidosis or ketonuria. The infant was started on OHCbl and transferred to a tertiary facility at 2 weeks of age. On admission, plasma amino acids indicated low methionine of 5 (reference 17-53) µmol/L, absent free homocystine, and elevated total plasma homocysteine at 20 (reference <12) µmol/L. The serum methylmalonic acid was 20.25 (reference <0.4) µmol/L. Urine organic acids indicated an isolated increase in methylmalonic acid at 154 (reference <5) mmol/mol creatinine. Above findings, including hypotonia and feeding difficulties, were strongly suggestive of the cblF inborn error of cobalamin metabolism. Blood counts were significant for neutropenia (absolute neutrophil count 0.3 K/µL , reference 1.5-10 K/µL) and thrombocytopenia (55-106 K/µL, reference 150-400

K/µL). The patient was started on Colony-Stimulating Factor which corrected

68 neutrophil and platelet counts. She was continued on a regular diet (breast milk), intramuscular injections of OHCbl (2 mg per day), oral MeCbl (2 mg twice a day),

5-methylTHF (3 mg twice a day), pyridoxal phosphate (35 mg twice a day), and betaine (100 mg/kg per day divided into 3 doses). With this treatment, the plasma amino acids normalized completely, as did total plasma homocysteine and serum methylmalonic acid levels.

Following discharge from the hospital at 36 days of age, the patient did well with no other hospitalizations, except one at 7 months of age for the elective repair of a left inguinal hernia. She is normal from a developmental standpoint.

The therapy consists of intramuscular injections of OHCbl (2.5 mg twice a week), oral MeCbl (2 mg twice a day), 5-methylTHF (3 mg twice a day), and pyridoxal phosphate (35 mg twice a day). Betaine was stopped at 6 months of age. With this therapy, plasma amino acids and total plasma homocysteine are persistently within the reference range, while methylmalonic acid is mildly elevated in serum

(0.68-1.45 µmol/L, reference <0.4 µmol/L) and urine (6-13 mmol/mol creatinine, reference <5 mmol/mol creatinine).

2.1.2 Patient WG4140

Patient WG4140 was the second child of non-consanguineous German parents. Physical examination revealed hypertelorism, micrognathia, wide inter- mamillary distance, a bell-shaped thorax, horizontal ribs and short extremities.

Cardiac catheterization showed atrial septal defect, coarctation of the aorta, small left ventricle, enlarged right ventricle and pulmonary hypertension. He was treated with antibiotics, diuretics and digoxin.

Newborn screening was reportedly unremarkable and chromosome analysis revealed a normal male karyotype. The clinical course during the first months of life was complicated by feeding difficulties, generalized hypotonia and developmental delay. Cerebral ultrasound revealed mild cerebral atrophy at 4 months of age. Repeated cardiac catheterization at 6 months revealed increased narrowing of the aortic isthmus and a hemodynamically relevant type II atrial septal defect. Therefore, corrective surgery of the aorta and closure of a patent ductus arteriosus was performed. At the age of 13 months, Noonan syndrome was suspected based on symptoms but the mutation analysis of the PTPN11 gene was negative. At this time, his weight was 6400g (below 1st centile), length 64 cm

(below 1st centile), and head circumference 43 cm (above 97th centile). Skeletal

X-rays of pelvis, hand, knee and spine revealed a delayed bone age.

At the age of 16 months, methylmalonic aciduria was noted. Plasma homocysteine was 50 (reference <15) µmol/L, and plasma cobalamin was 815

(reference 190-880) pg/mL. These findings, in addition to cardiac defects, feeding difficulties, hypotonia, developmental delay, and etc, were suggestive of the cblF defect. Moreover, fibroblast studies revealed findings consistent with this hypothesis. Therefore, at the age of 18 months, he received his first intramuscular

70 injection of 1.0 mg OHCbl and the homocysteine levels dropped to 15 µmol/L.

Monthly injections with OHCbl were started to restore body storage of cobalamin.

The patient showed some progress in psychomotor development and started to speak two word sentences at the age of 20 months. At the age of 22 months, he had a bilateral inguinal hernia repair and a bilateral orchiopexy for cryptorchidism.

At the age of 2.5 years, he had a complicated febrile convulsion. At the age of 5 years, he was treated with intramuscular injections of 1.0 mg OHCbl every 4 to 5 weeks. He showed gradual progress in his psychomotor development, but still had low weight (13.0 kg, below 1st centile) and short stature (92 cm, below 1st centile). Developmental delay and small for gestational age are common findings among patients with the cblF defect.

At the age of 7 years, he developed a pneumococcal septic arthritis of the right hip, during an episode of neutropenia (leukocyte count 3700/µL, absolute neutrophil count 1300/µL), requiring surgery of the hip and prolonged systemic antibiotic therapy. Intramuscular injections with OHCbl were continued at a dose of 1.0 mg every three weeks. Currently, at the age of seven and a half years, both plasma homocysteine (15.5 µmol/L) and serum methylmalonic acid (1.13

µmol/L) levels remain elevated.

2.1.3 Patient WG3630

Patient WG3630 was a second child of non-consanguineous Han Chinese

71 parents. At the age of 4 years, he was noted to have hyperpigmentation and gray hair. He complained of dizziness and headache occasionally for 2-3 years. At the age of 7 years, he had a transient ischemic attack with 1-2 months of left limping gait which recovered spontaneously. MRI/MRA of the brain revealed decreased flow in the left middle cerebral artery (M2) region.

At the age of 8 years, serial workup showed elevated plasma homocysteine level (71.47 µmol/L, reference 3.4-15.6 µmol/L), relatively low methionine level (17.3 µmol/L, reference 18-42 µmol/L), low serum cobalamin level (152 pg/ml, reference 179-1132 pg/ml), and relatively high serum folate level (>15 ng/mL, reference 3-12 ng/mL). Mild microcytic anemia with MCV

(mean corpuscular volume) of 70 fL and HgB of 11 g/dL was noted. In addition, urine organic acid analysis revealed presence of methylmalonic acid and methylcitrate, suggesting an inborn error of cobalamin metabolism.

The patient has been responding well to treatment with oral administration of MeCbl (0.5 mg three times a day) over the last 6 years. After the first week of treatment, plasma homocysteine level dropped to 24.15 µmol/L and a further drop to 15.77 µmol/L was seen after two weeks with addition of folate.

Serum cobalamin increased from 152 to 407 pg/mL, but MCV surprisingly decreased from 70 to 62.7 fL. Meanwhile, the patient’s elder sister was also observed to have elevated homocysteine level (21.08 µmol/L) and serum cobalamin level (228 pg/mL) at the low end of the reference range. Her

72 homocysteine level was 14.76 µmol/L at the second clinical visit before treatment and it dropped to 5-6 µmol/L after treatment with folate and cobalamin.

A family history of thalassemia was reported by the parents. Blood tests on the father, aunt, sister, and proband all showed small RBC with MCV around

64-70 fL and HgB around 11 g/dL. Moreover, a high level of plasma homocysteine was observed in the father (29.22 µmol/L) and aunt (51.02 µmol/L) in addition to the sister and proband. Screening the MTHFR gene for the c.677C>T polymorphism revealed that the father, aunt, and mother were heterozygotes while the proband and sister were homozygotes (Figure 7). This polymorphism had been known to be associated with increase in plasma homocysteine levels.114

2.2 Cell Culture

Human skin fibroblasts grown from skin tissue biopsy is the conventional

64 cell culture type used for biochemical assays of vitamin B12 metabolism. Cell lines were provided by the Repository of Mutant Human Cell Strains located at the Montreal Children’s Hospital. At the Repository, which undertakes the long- term storage and quality control of cell lines, cells are designated with a code number in the following manner; control cells are named as MCH followed by a

2-digit number, and patient cells are named as WG followed by a 4-digit number.

For instance, MCH64 is a control and WG4066 is a patient cell line.

Cells were routinely maintained in minimum essential medium plus non- essential amino acids (Gibco) supplemented with 5% fetal bovine serum

(Intergen) and 5% iron enriched calf serum (Intergen) (Watkins, 2000). Cells transfected with pBABE retroviral expression vector, as described in Section 2.8 and 2.9, were maintained in medium with 1 µg/mL puromycin (InvivoGen).

2.3 Selection of Fibroblast Cell Lines

In addition to the fibroblast cell lines of WG4066, WG4140, and

WG3630, two wild type controls (MCH64, MCH46), two cblF patient cell lines

(WG3365, WG3377), one cblD cell line (WG3646), and one cblC cell line

(WG4095) were selected (Table 1). For the control and the cblF defect, two cell lines were selected for each to demonstrate consistency in each type. The cblD and cblC defects acted as negative controls in transfection experiments and one cell line was selected for each.

Eight fibroblast cell lines, except WG3630, were used for immortalization, transfection, and biochemical assays to measure vitamin B12 metabolism. Cells were immortalized because immortalization both increased growth rates and prevented senescence. It has been shown previously that cellular cobalamin metabolism is not affected by immortalization (also see section 2.7).84 The primary cell line of WG3630 was used for biochemical assays without immortalization or transfection.

Cell Line Type Gene Mutated1 MCH64 Control - MCH46 Control - WG4066 Unknown Unknown WG4140 Unknown Unknown WG3365 cblF LMBRD1 WG3377 cblF LMBRD1 WG3646 cblD MMADHC WG4095 cblC MMACHC WG3630 Unknown Unknown

Table 1. Fibroblast cell lines used in this study. All cell lines except WG3630 were immortalized and transfected. 1Mutations in these genes are responsible for the inborn error of cobalamin metabolism in each patient. The gene mutated in

WG4066, WG4140, and WG3630 was unknown at this point in time.

2.4 Somatic Cell Complementation Analysis

Somatic cell complementation has been an essential tool in establishing complementation groups and accurately diagnosing patients with a defect in vitamin B12 metabolism. Equal numbers of patient fibroblasts and fibroblasts from known complementation groups were mixed and fused by exposure to 40% (w/v) polyethylene glycol 1000 (PEG). Propionate incorporation, described in Section

2.10, was compared in parallel fused and un-fused cultures. Incorporation was increased by two-fold or greater in fused than un-fused cultures from different complementation groups because each fibroblast provided the gene that the other was defective in. Conversely, no complementation took place if cultures came from the same complementation group.60 Somatic cell complementation was performed by Jocelyne Lavallée of the Rosenblatt Laboratory.

2.5 Exome Sequencing

Exome capture sequencings of WG4066 and WG3630 were performed in collaboration with the laboratory of Dr. Jacek Majewski of McGill University. A total of 3 µg of genomic DNA was used for exome capture using the Agilent

SureSelect All Exon v1 kit and massively-parallel sequencing using Illumina

GAIIx 76 nucleotide reads, as previously described,117 to generate a mean 30X coverage of the targeted regions. Variants were compared against a pool of in- house exomes and those previously seen in two or more individuals were

76 discarded. Novel variants were defined as those absent from dbSNP and having an allele frequency of less than 0.005 in the 1000 Genomes Project. This allowed us to consider variants observed in 1000 Genomes at very rare frequencies.

Potentially damaging variants included non-synonymous substitutions caused by missense and nonsense single-nucleotide polymorphisms (SNP), splice-site SNP, and frameshift changes due to indels.117 Candidate genes were selected as those containing either a homozygous or two potentially compound heterozygous variants in the same gene, satisfying the above criteria.

2.6 Mutation Analysis

For patient WG4066 and family members, primer design, PCR, and

Sanger sequencing of candidate genes were performed by Aurelie Masurel at the

McGill University and Genome Quebec Innovation Centre. Mutation analysis was performed by the author. For patient WG4140, all steps were performed by the group under the supervision of Dr. Brian Fowler and Dr. Matthias Baumgartner.

For patient WG3630 and family members, primer design, PCR, and mutation analysis of the ABCD4 gene were performed by the author. Sanger sequencing was performed at the McGill University and Genome Quebec Innovation Centre.

Sequencing primers and PCR protocol used for the sequencing of ABCD4 exon 4 in patient WG3630 and family members are described below.

Usage Name Sequence from 5’ to 3’ ABCD4 exon 4 ABCD4_Exon4_primer_F GGGATGGGTCAGCGGAGGCT A sequencing (Section 2.6.1) ABCD4_Exon4_primer_R CCCCCATGACTCTGGGCAGC ABCD4_seq_primer_F AAGTGGGGCTTGGACACACCCCC ABCD4 cDNA ABCD4_seq_primer_R GCTCCCGAAGGGTCCCGTCAGTG B sequencing (Section 2.8.2) ABCD4_int_primer_5 ATACAGCAGGTTGCAGGTGAACTG ABCD4_int_primer_3 GCAGGCCTGTCCAACTTGGTGGCA GGGGACAAGTTTGTACAAAAAAGCAG LMBRD1 attB1/2 LMBRD1_attB1_Forward GCTTCACCATGGCGACTTCTGGCGCG C cloning GGGGACCACTTTGTACAAGAAAGCTG (Section 2.9.1) LMBRD1_attB2_Reverse GGTGTTAAGCAGAATAGACAGAGGG LMBRD1_seq_primer_5_R ATATAACAGAATATAAAGTATAGTA LMBRD1 cDNA LMBRD1_seq_primer_3_F AATATGTTATGTATGGAAGCCAAAA D sequencing (Section 2.9.3) LMBRD1_int_primer_5_F AAAAAGCAAAGATGGTCGACCTTT LMBRD1_int_primer_3_R TCCACCAGCTGTTTTCAATGAATT

Table 2. Primers used in this study. For LMBRD1 attB1/2 cloning primers, the

red-coloured bases represent regions overlapping with 5’ and 3’ ends of the

LMBRD1 cDNA for forward and reverse primers, respectively.

2.6.1 Polymerase Chain Reaction (PCR)

Primers were designed using the online program Primer-BLAST, which is a combination Primer3 and BLAST (http://www.ncbi.nlm.nih.gov/tools/primer- blast) to sequence the exon 4 of the ABCD4 gene (Table 2A). The forward primer was positioned 76 bp upstream and the reverse primer was positioned 117 bp downstream of the exon to cover the entire exon 4. PCR was performed to amplify the exon 4 from gDNA of patient WG4066 and family members. Standard

PCR conditions were used and Taq polymerase was replaced with HotStarTaq

(Qiagen). The PCR protocol consisted of the following: initial denaturation at

96ºC for 10 min, 40 cycles of {melting at 96ºC for 30 sec, annealing at 62ºC for 1 min, and amplification at 72ºC for 40 sec}, and final elongation at 72ºC for 10 min. PCR amplification was verified by gel electrophoresis in 1.2% agarose gels.

2.7 Immortalization of Fibroblasts with E7 and Telomerase

Eight fibroblast cell lines (Table 1) were immortalized with the E7 gene from human papilloma virus type-16 and human telomerase as previously described.161 Immortalization of cell lines did not affect the cellular phenotype as assessed by cellular incorporation of [14C]methylTHF and [14C]propionate substrates (data not shown). Immortalization of fibroblasts was performed by

Timothy Johns of the Eric Shoubridge Laboratory.

2.8 Transfection of Fibroblasts with Wild Type ABCD4 cDNA

2.8.1 LR Recombination Reaction

Wild type human ABCD4 cDNA in pENTR221 vector (Kanamycin resistance; DQ892847.2) was purchased from DNAFORM (Yokohama, Japan) and cloned into the Gateway-modified retroviral expression vector pBABE

(Ampicillin resistance)161 using LR Clonase II Enzyme Mix (Invitrogen). The LR recombination reaction was performed by mixing 9 µL of pENTR221-ABCD4 (50 ng/µL), 1 µL of pBABE (150 ng/µL), 2 µL of TE buffer pH 8.0, and 3 µL of LR

Clonase and incubating overnight at 25°C. The reaction was terminated by adding

1 µL of Proteinase K solution and incubating for 10 minute at 37°C. OneShot

TOP10 E. coli (Invitrogen) was transformed with 1 µL of reaction mixture and transformants were plated on LB/Amp 100 µg/mL and LB/Kan 30 µg/mL agar plates. Colonies positive for pBABE-ABCD4 and negative for pENTR221-

ABCD4 were selected for colony PCR.

2.8.2 Colony PCR of LR Recombinants

Colony PCR was performed to check the presence of ABCD4-harbouring vector in the selected colonies. Colonies were picked and resuspended in 50 µL of milliQ water and boiled for 10 minutes. For the PCR reaction, 10 µL of boiled colony was used as the template, and primers “ABCD4_seq_primer_F” and

“ABCD4_seq_primer_R” were used to create a PCR product of 316 bp (Table

2B). Standard PCR conditions for Taq polymerase (Qiagen) were used. The PCR protocol consisted of the following: initial denaturation at 94ºC for 3 min, 40 cycles of {melting at 94ºC for 30 sec, annealing at 58ºC for 1 min, and amplification at 72ºC for 30 sec}, and final elongation at 72ºC for 10 min. PCR amplification was verified by gel electrophoresis in 1.2% agarose gels. Positive colonies were grown in LB broth overnight at 37°C and plasmids were extracted using QIAprep Spin Miniprep Kit (Qiagen). The fidelity of the retroviral pBABE-

ABCD4 constructs was validated by Sanger sequencing with primers

“ABCD4_int_primer_3” and “ABCD4_int_primer_5”, which are designed to check the 5’ and 3’ ends of the ABCD4 insert for correct reading frame (Table 2B).

2.8.3 Transfection

The retroviral pBABE-ABCD4 construct was transiently transfected into a Phoenix Amphotrophic packaging cell line via using the HEPES-buffered saline

/Ca3(PO4)2 method (http://www.stanford.edu/group/nolan/retroviral_systems/ phx.html). After 48-hour incubation, virus-containing medium was collected, supplemented with 4 μg/ml polybrene and used to infect eight immortalized fibroblast cell lines (Table 1). Fibroblasts were grown for 2 weeks in medium containing 1 μg/mL of puromycin (InvivoGen) for selection.84 Transfection of fibroblasts was performed by Stephen Fung of the Eric Shoubridge Laboratory.

2.9 Transfection of Fibroblasts with Wild Type LMBRD1 cDNA

2.9.1 Gateway Cloning

Wild type human LMBRD1 cDNA in pBlueScriptR vector (BC047073.1) was purchased from Open Biosystems (Lafayette, CO). PCR was performed to produce the LMBRD1 insert flanked by attB1/attB2 sites using Gateway-cloning primers (Table 2C). Standard PCR conditions for Taq polymerase (Qiagen) were used. The PCR protocol consisted of the following: initial denaturation at 94ºC for 3 min, 40 cycles of {melting at 94ºC for 30 sec, annealing at 58ºC for 1 min, and amplification at 72ºC for 2 min}, and final elongation at 72ºC for 10 min.

PCR amplification was verified by gel electrophoresis in 1.2% agarose gels. The attB1-LMBRD1-attB2 PCR product was extracted using QIAquick Gel Extraction

Kit (Qiagen).

2.9.2 BP Recombination Reaction

The attB1-LMBRD1-attB2 PCR product was cloned into the empty pDONR221 vector using BP Clonase II Enzyme Mix (Invitrogen). The BP recombination reaction was performed by mixing 5 µL of attB1-LMBRD1-attB2

PCR product (35 ng/µL), 1 µL of pDONR221 (150 ng/µL), 2 µL of TE buffer pH

8.0, and 2 µL of BP Clonase and incubating overnight at 25°C. The reaction was terminated by adding 1 µL of Proteinase K solution and incubating for 10 minutes at 37°C. Chemically-competent E. coli DH5α was transformed with 1 µL of

82 reaction mixture and transformants were plated on LB/Kan 30 µg/mL and

LB/Amp 100 µg/mL agar plates. Colonies positive for pENTR221-LMBRD1 were selected for colony PCR.

2.9.3 Colony PCR of BP Recombinants

The procedure was identical to Section 2.8.2, except that primers

“LMBRD1_attB1_Forward” and “LMBRD1_attB2_Reverse” were used to create a PCR product of 1687 bp (Table 2C), and amplification at 72ºC was for 1 minute and 50 sec. The fidelities of the pENTR221-LMBRD1 constructs were validated by Sanger sequencing with primers “LMBRD1_seq_primer_5_R”,

“LMBRD1_seq_primer_3_F”, “LMBRD1_int_primer_5_F”, and

“LMBRD1_seq_primer_3_R”, which are designed to cover the entire LMBRD1 cDNA insert (Table 2D).

2.9.4 LR Recombination Reaction

The procedure was identical to Section 2.8.1, except that 9 µL of pENTR221-LMBRD1 (56 ng/µL) was used for LR recombination reaction, and

OneShot Mach1-T1 E. coli (Invitrogen) was used for transformation.

2.9.5 Colony PCR of LR Recombinants

The procedure was identical to Section 2.9.3.

2.9.6 Transfection

The procedure was identical to Section 2.8.3.

2.10 Labelled MethylTHF and Propionate Incorporation Assays

Fibroblast cell lines with a defect in cobalamin metabolism pathway can be tested for their ability to utilize methionine synthase and methylmalonyl-CoA mutase enzymes. Measurement of incorporation of label from [14C]methylTHF or

[14C]propionate into cellular macromolecules provides an indirect method of measuring the function of MTR or MUT, respectively, in intact cells. In the former,

MTR catalyzes the transfer of labelled carbon from methylTHF to cobalamin and then to homocysteine to produce methionine, which is incorporated into macromolecules. In the latter, propionate is converted to propionyl-CoA, (S)- methylmalonyl-CoA, and then (R)-methylmalonyl-CoA, which is catalyzed by

MUT into succinyl-CoA and incorporated into macromolecules. Incorporation of radiolabel is decreased if MTR or MUT function is diminished. Fibroblasts were plated in 35 mm tissue culture dishes at a density of 400,000 cells/dish. Cultures were incubated for 18 hours in a medium containing 8.6 µM [14C]methylTHF or

0.1 µM [14C]propionate.60, 162 After incubation, cellular macromolecules were precipitated in 5% trichloroacetic acid and radioactivity of precipitate resuspended in 0.2 N NaOH was determined by liquid scintillation counting. Protein values were measured by the Lowry assay to account for the difference in number of

84 cells between each P35 tissue culture dish. Each cell line was assayed in triplicate.

2.11 Cobalamin Derivative Distribution Assay

This assay measured the ability of fibroblast lines to synthesize AdoCbl and MeCbl from CNCbl. Confluent fibroblasts in T175 culture flasks were used for the assay, and 20 µL of [57Co]-CNCbl (MP Biomedicals) was pre-incubated with 2 mL of human serum for 30 min at 37°C to allow binding of cobalamin to transcobalamin. Serum with [57Co]-CNCbl was mixed with 23 mL of serum-free

MEM and filter-sterilized. Fibroblasts were incubated for 96 hours in this final medium containing 25 pg/mL of [57Co]-CNCbl. Cells were harvested by trypsinization and disrupted by freezing and thawing, cobalamins were extracted in hot ethanol (80°C), and cobalamin derivatives were separated by high performance liquid chromatography using a LiChrosorb RP-C8 column

(Phenomenex).93 Because this is a reverse phase chromatography, polar compounds eluted faster than non-polar compounds. Consequently, cobalamin derivatives were obtained in the order of OHCbl, CNCbl, AdoCbl, and MeCbl.

Radioactivity of fractions, co-eluting with each cobalamin derivative, was quantitated by gamma-counting, and distribution of cobalamin was expressed as a percentage of radioactivity in all fractions.

2.12 Superose 12 Analysis of TC-Bound, Free and Enzyme-Bound Cbl

This assay measured the ability of fibroblast lines to take up TC-bound cobalamin and deliver it to cellular compartments for association with enzymes.

Confluent fibroblasts in T175 culture flasks were used for the assay, and 20 µL of

[57Co]-CNCbl (MP Biomedicals) was pre-incubated with 2 mL of human serum for 30 min at 37°C to allow binding of cobalamin to transcobalamin. Serum with

[57Co]-CNCbl was mixed with 23 mL of serum-free MEM and filter-sterilized.

Fibroblasts were incubated for 96 hours in this final medium containing 25 pg/mL of [57Co]-CNCbl. Cells were harvested by trypsinization, resuspended in 0.1M potassium phosphate buffer (pH 7.4) and broken open by sonication using

Soniprep 150 (MSE Scientific Instruments). Disrupted cell membranes were removed by ultracentrifugation at 171,500 x g, 5°C for 30 min, and free Cbl, TC- bound Cbl (TC-Cbl) and MUT or MTR-bound cobalamin (enzyme-Cbl) were separated by fast protein liquid chromatography using a Superose 12 column (GE

Healthcare Life Sciences). Because this is a size exclusion chromatography, high

MW compounds eluted faster than low MW compounds. Consequently, enzyme-

Cbl eluted first, TC-Cbl eluted next, and free Cbl eluted last. Radioactivity of collected fractions was quantitated by gamma-counting, and distribution of cobalamin was expressed as a percentage of radioactivity in all fractions.

CHAPTER 3: RESULTS

3.1 Identification of Three Patients

Three patients were suspected to have inborn errors of cobalamin metabolism upon detection of methylmalonic aciduria and hyperhomocysteinemia by the referring clinicians. Ethnic background, age at diagnosis, and clinical course of each patient greatly varied from one another. For WG4066 and WG4140, fibroblast cell studies revealed a combined decrease in incorporation of labelled methylTHF and propionate, an increase in uptake of labelled CNCbl, nearly absent synthesis of AdoCbl and MeCbl, and an accumulation of large amounts of unmetabolized CNCbl (Table 3). Their cellular phenotypes mimicked those of the cblF disorder but no mutations in the LMBRD1 gene were found. Somatic cell complementation and exome sequencing were performed on WG4066 and

WG4140 independently to identify the causes of their diseases. WG3630, referred to our laboratory before the other two, was very difficult to diagnose because incorporations of labelled propionate and methylTHF by fibroblasts were within reference ranges and too high to allow for complementation analysis (Table 3). It could not be determined whether it belonged to any known complementation groups or not. Nonetheless, decreased synthesis of both AdoCbl and MeCbl and moderate accumulation of unmetabolized CNCbl were observed. Eventually, exome sequencing was performed on the genomic DNA of WG3630.

Reference Range WG4066 WG4140 WG3630 cblF Control [14C]MethylTHF − OHCbl 43.6 26.4 98.9 43 ± 11 191 ± 112 Incorporation (pmoles/mg + OHCbl 111.8 68.1 231.0 297 ± 165 312 ± 207 protein/18h) [14C]Propionate − OHCbl 1.0 2.1 15.3 1.9 ± 1.7 13.1 ± 4.1 Incorporation (nmoles/mg + OHCbl 4.7 4.4 16.8 7.2 ± 3.0 14.2 ± 4.5 protein/18h) Cobalamin Distribution

(%) OHCbl 2.8 2.6 10.2 8 ± 4 6 ± 3 CNCbl 91.2 90.2 58.5 81 ± 7 15 ± 13 AdoCbl 0.7 1.6 6.8 2.4 ± 0.6 15 ± 3 MeCbl 1.0 0.2 13.4 1.3 ± 1.1 56 ± 8

Table 3. Biochemical profiles of three patient fibroblasts. Biochemical

measurements were obtained from the initial assays performed on primary cell

lines at our medical genetics laboratory. Reference ranges are based on 12

measurements of cblF cell lines and around 200 measurements of control cell

lines in our laboratory over the years.

3.2 Somatic Cell Complementation Analysis

WG4066 fibroblasts were fused with cells from all complementation groups known to cause combined MMA and HC, namely cblC (WG3897,

WG4053), cblD (WG3646), and cblF (WG3365). Complementation was performed with and without the addition of PEG to observe the effect of PEG- induced cell fusion on increase of [14C]propionate incorporation. Three-fold or greater increase in incorporation was observed when WG4066 was fused with each of the four cell lines. This meant that the defect in WG4066 is neither cblC, cblD, nor cblF (Table 4).

WG4140 fibroblast cells were fused with cells from cblC (WG4136), cblF (WG3365, WG3377), and WG4066. Complementation was performed as above. A two to three-fold increase in incorporation was observed when WG4140 was fused each of the cblC or cblF cell lines. No complementation was observed when WG4140 and WG4066 were fused together. This meant that the defect in

WG4140 is neither cblC nor cblF but identical to the one in WG4066 (Table 5).

Cell Line Complementation Cell line [14C]propionate/propionate Group propionate incorporation (nmoles/mg protein/18h) incorporation Complementation Complementation before without cell fusion with cell fusion complementation (w/o PEG) (+ PEG) 0.7 WG4066 0.7 0.7 0.8 0.6 1.0 3.0 WG3897 cblC 0.6 0.6 0.7 1.0 1.0 1.0 3.0 3.2 2.9 1.9 1.2 3.6 WG4053 cblC 1.9 1.9 1.8 1.1 1.1 1.2 3.4 3.8 3.7 0.7 1.1 4.5 WG3646 cblD 0.6 0.7 0.7 1.1 1.1 1.1 4.4 4.5 4.5 1.8 1.3 4.9 WG3365 cblF 1.8 1.7 2.0 1.3 1.3 1.4 5.0 4.8 5.0

Table 4. Somatic cell complementation of WG4066. Complementation analysis was performed by pairing WG4066 cell line with two cblC, one cblD, and one cblF cell lines. Baseline [14C]propionate incorporation was measured before complementation for each cell line. Complementations were performed and level of [14C]propionate incorporations in those without cell fusion and those with cell fusion were compared. Bold numbers represent averages of triplicates for each condition.

Cell Line Complementation Cell line [14C]propionate/propionate Group propionate incorporation (nmoles/mg protein/18h) incorporation Complementation Complementation before without cell fusion with cell fusion complementation (w/o PEG) (+ PEG) 1.9 WG4140 1.9 1.8 1.9 0.2 1.1 2.5 WG4136 cblC 0.2 0.2 0.2 1.2 1.0 1.1 2.3 2.5 2.5 1.5 1.6 4.8 WG3365 cblF 1.4 1.5 1.5 1.6 1.6 1.7 4.4 4.9 5.0 1.0 1.4 4.8 WG3377 cblF 0.9 1.0 1.0 1.4 1.4 1.3 5.0 4.7 4.8 1.5 1.6 1.9 WG4066 Unknown 1.5 1.5 1.5 1.5 1.6 1.5 1.9 2.0 1.9

Table 5. Somatic cell complementation of WG4140. Complementation analysis was performed by pairing WG4140 cell line with one cblC, two cblF, and

WG4066 cell lines. Baseline [14C]propionate incorporation was measured before complementation for each cell line. Complementations were performed and level of [14C]propionate incorporations in those without cell fusion and those with cell fusion were compared. Bold numbers represent averages of triplicates for each condition.

3.3 Discovery of Causative Gene in Each Patient

3.3.1 WG4066

The exome sequencing and variant filtering process identified three candidate genes, TPRG1, LRP2, and ABCD4, each containing two heterozygous, potentially damaging mutations (Figure 6). The LRP2 gene was initially noted as a possible candidate because of its role in cobalamin absorption. Nonetheless, the three genes were re-sequenced in the father, mother, elder sister, and proband to check the fidelity of massively-parallel sequencing and eliminate candidate genes based on segregation pattern of mutations (Table 6).

Both mutations in TPRG1 were detected in the paternal DNA by Sanger sequencing, meaning that they were present in cis and passed down to the patient on a single allele. In fact, the c.422G>T mutation did not affect the gene product because the upstream c.183_192del (p.Tyr61Stop) mutation created a premature stop codon. Since the patient must have inherited the maternal, wild type allele,

TPRG1 mutations do not cause the patient’s disorder. For LRP2, the father and mother were heterozygous for each of the mutant alleles, but the asymptomatic sibling was found to be a compound heterozygote. Therefore, LRP2 mutations are also not responsible for the defect. For ABCD4, the father and mother were again heterozygous for each of the mutant alleles, c.956A>G and c.1746_1747insCT, respectively. However, unlike LRP2, the sibling did not carry either of the mutations. This pinpointed ABCD4 as the disease-causing gene.

Homozygous Heterozygous TPRG1 wild type mutant

c.183_192del p.Tyr61Stop

c.422G>T p.Ser141Ile

LRP2

c.3932G>A p.Arg1311His

c.10937G>A p.Arg3646His

ABCD4

c.956A>G p.Tyr319Cys

Glu C G G C A G A G C C T T G A G A A G C C A G A G C C T T C T G A G A A G c.1746_1747insCT Leu p.Glu583LeufsX9

Figure 6. DNA sequencing chromatograms of heterozygous mutations in

TPRG1, LRP2, and ABCD4 in the family of WG4066. For each chromatogram, the black box indicates the affected codon and the black arrow indicates the mutant base. For the frameshift insertion in ABCD4, wild type GAG, coding for glutamic acid, and mutant CTG, coding for leucine, are sequenced out of frame by

2 nucleotides because the exon was sequenced in the reverse direction. The red box indicates the dinucleotide insertion. Left column: homozygous wild type; right column: heterozygous mutant.

TPRG1 LRP2 ABCD4 Father c.183_192del c.422G>T WT c.10937G>A c.956A>G WT Mother WT WT c.3932G>A WT WT c.1746_1747insCT Sibling WT WT c.3932G>A c.10937G>A WT WT Patient c.183_192del c.422G>T c.3932G>A c.10937G>A c.956A>G c.1746_1747insCT

Table 6. Segregation analysis of TPRG1, LRP2 and ABCD4 mutations in the family of WG4066. Three candidate genes were re-sequenced to validate the presence of each heterozygous mutation in the father, mother, asymptomatic sibling, and/or proband. Nucleotide numbering is based on the cDNA sequence of each gene with 1 corresponding to the A of the ATG translation initiation codon.

WT, wild type.

3.3.2 WG4140

Concurrent with our efforts to uncover the molecular basis of the defect in

WG4066, the research team led by Dr. Brian Fowler investigated the patient

WG4140 to find disease-causing mutations. Microcell-mediated chromosome transfer identified chromosome 14 as the site of the mutated gene, and exome sequencing of chromosome 14 led to the identification of nine variants in eight genes. Further evaluation of the variants allowed them to narrow down to two variants in the ABCD4 gene, c.542+1G>T and c.1456G>T, as the best candidates.

3.3.3 WG3630

Exome sequencing discovered potentially damaging variants in six candidate genes, MUTYH, IL17RD, TECPR1, VPS13B, ABCD4, and ANGEL1, each with compound heterozygous or homozygous mutations. A homozygous mutation in ABCD4 was instantly recognized as the best candidate because mutations in this gene had already been identified and confirmed to be causative in WG4066 and WG4140. Exon 4 of the ABCD4 gene, containing the c.423C>G mutation, was sequenced in the aunt, father, mother, elder sister, proband, and younger brother to confirm the correct segregation pattern of mutation (Figure 7).

A. I-A I-B I-C

Hcy 51.02 μmol/L Hcy 29.22 μmol/L Hcy 7.24 μmol/L MTHFR C/T MTHFR C/T MTHFR C/T ABCD4 C/G ABCD4 C/G ABCD4 C/G

II-A II-B II-C

Hcy 21.08 μmol/L Hcy 71.47 μmol/L Hcy 4.71 μmol/L MTHFR T/T MTHFR T/T MTHFR C/C ABCD4 C/G ABCD4 G/G ABCD4 C/G

B. Homozygous Heterozygous Homozygous wild type carrier mutant

Exon 4 Intron4 Exon 4 Intron4 Exon 4 Intron4

Figure 7. (A) Pedigree of the family of WG3630 showing plasma homocysteine levels and genotypes. For the MTHFR c.677C>T polymorphism, C is the major allele and T is the minor allele. For the ABCD4 c.423C>G mutation, C is the wild type allele, and G is the mutant allele. I-A, aunt; I-B, father; I-C, mother; II-A, elder sister; II-B, proband; and II-C, younger brother. (B) DNA sequencing chromatograms of the ABCD4 c.423C>G (p.N141K) mutation. For each chromatogram, the black box indicates the affected codon, the black arrow indicates the mutant base, and the red line indicates the exon-intron boundary. Left column: homozygous wild type (MCH23 control); middle column: heterozygous carrier; and right column: homozygous mutant.

3.4 Transfections and Assessments of Biochemical Phenotypes

Eight cell lines in Table 1 were immortalized and transfected with wild type ABCD4 and LMBRD1 constructs. The transfection with ABCD4 was performed to see correction of biochemical phenotypes in WG4066 and WG4140 patient cell lines and confirm ABCD4 as the gene responsible for the new defect.

The other six cell lines were selected as various controls as described in Section 2.3.

The transfection with LMBRD1 was performed to see correction of biochemical phenotypes in the WG3365 and WG3377 cblF cell lines and assess whether

WG4066 and WG4140 cells were biochemically affected or not. MethylTHF incorporation, propionate incorporation, cobalamin derivative distribution assay, and Superose 12 analysis of TC-bound, free, and enzyme-bound cobalamin were assayed.

3.4.1 Labelled MethylTHF and Propionate Incorporation Assays

A total of 24 immortalized cell lines were assayed; MCH64, MCH46,

WG4066, WG4140, WG3365, WG3377, WG3646, and WG4095 were either not transfected, transfected with wild type ABCD4, or transfected with wild type

LMBRD1 (Tables 7, 8). For both assays, non-transfected MCH64 and MCH46 showed incorporation levels within the reference range for controls as expected.

Six patient cell lines, including WG4066 and WG4140, showed incorporation levels within the reference range for cblF patients; WG3646 and WG4095 do not

97 have the cblF defect but they showed low incorporation levels expected of cblD- combined and cblC defects, respectively.

Transfection with ABCD4 or LMBRD1 did not affect incorporation levels of positive controls, MCH64 and MCH46, and negative controls, WG3646 and

WG4095. Slight increases or decreases in incorporation levels were within our expectation of variability in incorporation assay results. As a result, expression of wild type ABCD4 and LMBRD1 were shown to have no effects on WG3646 and

WG4095 since the two cell lines carried defects in metabolic steps downstream to lysosomal cobalamin export.

Transfection of WG4066 and WG4140, which carry compound heterozygous mutations in ABCD4, with wild type ABCD4 resulted in the correction of incorporation levels to within the control reference range. Similarly, transfection of WG3365 and WG3377, which carry compound heterozygous mutations in LMBRD1, with wild type LMBRD1 resulted in the correction of incorporation levels. Unexpectedly, when cblF cell lines were transfected with

ABCD4, a similar restoration was observed. However, when WG4066 and

WG4140 cell lines were transfected with LMBRD1 in a reciprocal experiment, no restoration of incorporation levels was observed.

[14C]methylTHF incorporation (pmoles/mg protein/18h)

Transfected with Transfected with Not transfected ABCD4 LMBRD1 170.7 (12.5) 145.5 (3.8) 154.3 (8.3) MCH64 166.3 170.3 198.1 146.2 143.0 144.7 165.7 151.0 146.3 162.3 163.1 164.2 146.8 152.4 140.0 94.3 (7.4) 95.4 (7.5) 142.6 (2.1) MCH46 92.2 103.1 105.5 97.8 92.1 80.5 140.6 141.8 145.5 90.3 85.1 89.4 103.1 98.4 100.8 46.4 (8.2) 105.7 (3.3) 64.3 (0.8) WG4066 49.6 57.9 54.9 101.3 106.5 109.3 65.4 64.3 63.3 36.5 39.3 39.9 29.4 (2.2) 80.2 (4.1) 44.8 (0.9) WG4140 29.9 34.0 27.9 85.4 75.3 80.0 46.2 44.0 44.4 27.9 27.5 29.0 WG3365 37.1 (0.8) 95.6 (2.8) 121.8 (19.3) (cblF) 37.5 35.9 37.9 91.7 98.4 96.8 109.7 106.5 149.1 WG3377 46.1 (1.4) 95.0 (15.7) 180.1 (12.2) (cblF) 44.4 46.2 47.8 87.1 116.9 81.0 165.6 179.2 195.4 WG3646 40.8 (2.2) 38.7 (2.3) 21.2 (2.0) (cblD) 43.8 38.7 39.9 35.5 41.0 39.6 21.7 18.6 23.4 WG4095 48.0 (4.2) 45.0 (1.8) 40.5 (7.7) (cblC) 42.6 53.0 48.5 43.9 43.5 47.6 34.3 51.4 35.9

Table 7. Labelled methylTHF incorporations of 24 cell lines. [14C]methylTHF incorporation experiments were performed in triplicate and were repeated for a few cell lines. Averages are written in bold numbers and standard deviations are written beside them in brackets. Reference range for the cblF patients is 43 ± 11, and for controls is 191 ± 112.

[14C]propionate incorporation (nmoles/mg protein/18h)

Transfected with Transfected with Not transfected ABCD4 LMBRD1 8.3 (0.1) 7.6 (0.2) 8.4 (0.6) MCH64 8.4 8.2 8.4 7.4 7.5 7.9 7.6 9.0 8.7 9.4 (0.9) 9.5 (0.2) 9.9 (0.6) MCH46 8.4 8.7 9.0 9.4 9.8 9.4 9.5 9.5 10.8 11.2 9.5 9.8 1.2 (0.0) 13.1 (0.3) 1.1 (0.0) WG4066 1.2 1.2 1.2 13.0 13.4 12.8 1.0 1.1 1.1 1.6 (0.0) 11.9 (0.3) 2.7 (0.7) WG4140 1.5 1.6 1.6 12.1 12.2 11.5 2.5 2.3 2.2 4.2 2.7 2.0 WG3365 2.2 (0.2) 14.7 (0.7) 15.0 (0.4) (cblF) 2.0 2.2 2.4 15.7 14.2 14.1 15.1 14.5 15.3 WG3377 1.1 (0.0) 10.3 (0.1) 15.0 (1.1) (cblF) 1.1 1.1 1.1 10.4 10.2 10.4 14.6 14.0 16.5 WG3646 1.6 (0.0) 1.1 (0.1) 2.0 (0.8) (cblD) 1.6 1.6 1.5 1.1 1.0 1.2 3.2 1.5 1.4 1.5 (0.1) 1.1 (0.0) 2.0 (0.2) WG4095 1.5 1.6 1.4 1.2 1.1 1.2 1.7 2.0 2.0 (cblC) 2.0 2.3 1.8

Table 8. Labelled propionate incorporations of 24 cell lines. [14C]propionate incorporation experiments were performed in triplicate and were repeated for a few cell lines. Averages are written in bold numbers and standard deviations are written beside them in brackets. Reference range for cblF patients is 1.9 ± 1.7, and for controls is 13.1 ± 4.1.

100

3.4.2 Cobalamin Derivative Distribution Assay

A total of 18 immortalized cell lines were assayed; MCH64, MCH46,

WG4066, WG4140, WG3365, and WG3377 were either not transfected, transfected with wild type ABCD4, or transfected with wild type LMBRD1

(Figure 8). Non-transfected MCH64 and MCH46 showed normal syntheses of

AdoCbl and MeCbl with percentages within reference ranges. Control cell lines were not affected by transfection with ABCD4 or LMBRD1, and AdoCbl and

MeCbl levels remained within reference ranges. WG4066, WG4140, WG3365, and WG3377 cell lines showed nearly absent synthesis of AdoCbl and MeCbl, which is commonly observed in the cblF fibroblasts; percentages of AdoCbl were

1.7 - 3.1% while percentages of MeCbl were 0.4 - 1.4%. Importantly, they showed very high percentages of unmetabolized CNCbl ranging from 76.1 to 82.1%.

Transfection of WG4066 and WG4140 with wild type ABCD4 resulted in the increases in AdoCbl and MeCbl syntheses; AdoCbl levels were not as high as normal values but MeCbl levels were within the accepted normal range. This change was accompanied by corresponding decreases in CNCbl levels.

Interestingly, when WG3365 and WG3377 were transfected with ABCD4, slightly greater increases in AdoCbl synthesis and slightly lower increases in MeCbl synthesis were observed when compared to WG4066 and WG4140. They also displayed corresponding decreases in CNCbl levels.

In parallel, transfection of WG3365 and WG3377 with wild type

101

LMBRD1 resulted in the increases of AdoCbl and MeCbl syntheses with both levels reaching reference ranges. As a result, CNCbl levels were drastically decreased. However, when WG4066 and WG4140 were transfected with

LMBRD1 in a reciprocal experiment, no increases in AdoCbl and MeCbl syntheses were observed and CNCbl levels remained high.

102

Not transfected

80 OHCbl CNCbl 60 AdoCbl MeCbl

20 % of % cobalamin total 0 MCH64 MCH46 WG4066 WG4140 WG3365 WG3377 Transfected with ABCD4

20 % % of totalcobalamin 0 MCH64 MCH46 WG4066 WG4140 WG3365 WG3377 Transfected with LMBRD1

% % of total cobalamin 0 MCH64 MCH46 WG4066 WG4140 WG3365 WG3377 Figure 8. Cobalamin derivative distributions of 18 cell lines. Six cell lines were assayed before transfection and after transfection with ABCD4 and LMBRD1.

Each bar represents a value obtained by combining gamma radioactivity counts at the elution peak of each cobalamin derivative. A legend is located at the top left corner.

103

3.4.3 Superose 12 Analysis

Identities of three peaks in elution patterns were confirmed by comparing three samples; MCH64, a control cell line, was expected to produce mostly enzyme-Cbl; WG3630, a patient cell line, was predicted to produce mostly free

Cbl; and human serum-Cbl control, a mixture of 1.6 µL of [57Co]-CNCbl, 160 µL of human serum, and 840 µL of potassium phosphate buffer, was expected to produce mostly TC-Cbl (Figure 9). After comparing the patterns, the first peak

(fractions 13-17) was confirmed to be enzyme-Cbl, the second peak (fractions 19-

22) was confirmed to be TC-Cbl, and the third peak (fractions 27-34) was confirmed to be free Cbl. MUT and MTR enzymes had been reported to migrate in a single peak at ~150 kDa,74 and the identity of this peak in the human serum-

Cbl control was in fact haptocorrin-bound cobalamin which migrates at ~150 kDa due to heavy glycosylation.163 The MW of TC is 43 kDa and cobalamin is ~1.3 kDa.20

A total of 18 immortalized cell lines were assayed; MCH64, MCH46,

WG4066, WG4140, WG3365, and WG3377 were either not transfected, transfected with wild type ABCD4, or transfected with wild type LMBRD1

(Figure 10). Non-transfected MCH64 and MCH46 showed normal levels of enzyme-Cbl at 73.6 - 78.7% and free Cbl at 1.9 - 6.0% because cobalamins were properly delivered to target compartments and associated with MTR and MUT.

Control cell lines were not affected by transfection with ABCD4 or LMBRD1, and

104 enzyme-Cbl levels remained between 76.4 - 82.0%. WG4066, WG4140, WG3365, and WG3377 cell lines showed very low levels of enzyme-Cbl at 2.3 - 3.8% and high levels of free Cbl at 62.4 - 82.6% because unbound cobalamins were trapped in the lysosomes. This finding corresponded to the high levels of unmetabolized

CNCbl in the cblF and two patient cell lines assessed by cobalamin derivative distribution assay.

Transfection of WG4066 and WG4140 with wild type ABCD4 resulted in the increases in enzyme-Cbl levels to 79.1 and 70.5%, respectively. This change was accompanied by corresponding decreases in free Cbl levels to 2.8 and 4.1%.

Interestingly, when WG3365 and WG3377 were transfected with ABCD4, moderate increases in enzyme-Cbl levels to 42.7 and 56.1%, respectively, were observed. They also displayed corresponding decreases in free Cbl levels to 43.1 and 20.8%.

Similarly, transfection of WG3365 and WG3377 with wild type LMBRD1 resulted in the increases in enzyme-Cbl levels to 78.1 and 80.3%, respectively.

Consequently, free Cbl levels dropped to 9.2 and 3.4%. However, when WG4066 and WG4140 were transfected with LMBRD1 in a reciprocal experiment, no increases in enzyme-Cbl levels were observed as they remained at 4.8 and 2.6%, respectively. Free Cbl levels also remained high at 74.5 and 83.8%.

105

A. MCH64 5000 78.7%

4000

3000

2000

1000 9.1% 6.0% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 B. WG3630 9000 8000 76.2% 7000 6000 5000 4000 3000 2000 8.8% 1000 6.5% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 C. Human serum-Cbl control 2500 56.0% 2000

1500

1000 25.9% 500 5.7% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Enzyme-Cbl TC-Cbl Free Cbl

Figure 9. Elution patterns of Superose 12 analyses. Three elution patterns of

(A) MCH64, (B) WG3630, and (C) Human serum-Cbl control are shown as examples. For each sample, 40 fractions of decreasing MW were collected.

Identity of each peak is shown below a dashed line. Percentage of each peak in total cobalamin radioactivity is shown above the peak.

106

Not transfected

80 TC-Cbl Free Cbl Enzyme-Cbl

20 % % of totalcobalamin 0 MCH64 MCH46 WG4066 WG4140 WG3365 WG3377 Transfected with ABCD4

% % of total cobalamin 0 MCH64 MCH46 WG4066 WG4140 WG3365 WG3377 Transfected with LMBRD1

% % of total cobalamin 0 MCH64 MCH46 WG4066 WG4140 WG3365 WG3377

Figure 10. Superose 12 analyses of 18 cell lines. Six cell lines were assayed before transfection and after transfection with ABCD4 and LMBRD1. Each bar represents a value obtained by combining gamma radioactivity counts at the elution peak corresponding to the size of the cobalamin form. A legend is located at the top right corner.

107

CHAPTER 4: DISCUSSION

4.1 Novel Inborn Error of Vitamin B12 Metabolism

This thesis reports the description of patients with a novel inborn error of cobalamin metabolism, the designation of a new complementation group, the identification of the responsible gene, and the correction of biochemical phenotypes by expression of wild type ABCD4 cDNA in patient cells.

Three patients with MMA and HC were referred to our laboratory for suspected inborn errors of cobalamin metabolism. Based on biochemical findings from cultured fibroblasts, WG4066 and WG4140 were identified as cblF- phenocopies with failure of lysosomal cobalamin transport but without mutations in the LMBRD1 gene. Cells clearly showed accumulation of non-protein bound, unmodified cyano form of cobalamin and decreased synthesis of MeCbl and

AdoCbl associated with MTR and MUT, respectively. WG3630 was a very unusual case because methylTHF and propionate incorporations were assessed to be normal but moderate decreases in MeCbl and AdoCbl syntheses and accumulation of CNCbl were observed (Table 3).

Somatic cell complementation revealed that WG4066 and WG4140 did not belong to any of the known eight complementation groups. Based on the absence of complementation when cells from the two patients were fused together, it was confirmed that WG4066 and WG4140 suffer from the same defect in

108 cobalamin metabolism (Tables 4, 5). The designation of cblJ has been proposed for this novel complementation group.

To identify the genetic defect in WG4066 and WG4140, exome capture and massively-parallel sequencing were independently employed by our laboratory and the laboratories of Dr. Brian Fowler and Dr. Matthias Baumgartner, respectively. Exome sequencing of WG3630 was performed after the other two patients without the expectation that this patient carried the same defect.

Comparing genetic variants in patients to dbSNP and 1000 Genomes databases was beneficial in finding private mutations and narrowing down to a small number of candidate genes with compound heterozygous or homozygous mutations. In patients WG4066, WG4140, and WG3630, three, eight, and six candidate genes, respectively, were discovered after the filtering process.

Therefore, the current report demonstrated the efficacy of exome sequencing in identifying a disease-causing gene from unrelated patients.

Combining the exome variants data of WG4066 and WG4140, it was discovered that both patients carried compound heterozygous mutations in the

ABCD4 gene, which was selected as the top candidate gene by each laboratory even before merging the independent exome sequencing efforts (Table 6).

Subsequently, the exome of WG3630 was sequenced and a homozygous mutation in ABCD4 immediately stood out as the best candidate. To further confirm the causality of ABCD4 mutations in the patients, I transfected the fibroblasts of

109

WG4066 and WG4140, along with other control cell lines, with a wild type

ABCD4 cDNA. A retroviral expression vector system, previously established to be highly effective, was utilized.84, 161 It is worthwhile noting that the success of this project owes much to the development of next-generation sequencing techniques.

Exome sequencing of a single proband, identification of a candidate gene, and confirmation by transfection with a wild type copy of the gene were similarly accomplished before with complex I deficiency.164

The transfection experiment was successful in demonstrating that ABCD4 is a new player in vitamin B12 metabolism and is defective in the three patients

(Tables 7, 8 & Figures 8, 10). Transfections with ABCD4 and LMBRD1 did not have any unexpected effects on cellular cobalamin metabolism as assessed by four biochemical assays on the control cell lines. MethylTHF and propionate incorporations were assayed for transfected WG3646 and WG4095 cell lines and they did not show increases in incorporations since defects in MMADHC and

MMACHC could not be restored. Expression of wild type ABCD4 could correct cobalamin metabolism in both WG4066 and WG4140 patient cells, meaning that

ABCD4 was the gene causing the cblJ defect. In the meantime, expression of wild type LMBRD1 could correct cobalamin metabolism in the two cblF cells,

WG3365 and WG3377.

A serendipitous finding was that cblF cells with ABCD4-transfection, still lacking a functioning copy of the LMBRD1 gene, were capable of lysosomal

110 cobalamin export resulting in roughly 50% correction of cobalamin cofactor synthesis. On the other hand, cblJ cells with LMBRD1-transfection, still lacking a functioning copy of the ABCD4 gene, were incapable of lysosomal cobalamin export resulting in no enhancement of cobalamin cofactor synthesis. Given the observation that overexpression of ABCD4 could correct the cblF defect while overexpression of LMBD1 could not correct the cblJ defect, I speculate that

ABCD4 may be the actual transporter of cobalamin while LMBD1 has a regulatory or accessory role.

In retrospect, the phenotypes of three patients were not outwardly different from those of cblF patients. Clinical findings overlapping with those seen in more than one third of cblF patients were hypotonia and feeding difficulties seen in WG4066, cardiac defects, hypotonia, feeding difficulties, developmental delay, failure to thrive, and small for gestational age seen in

WG4140, and low serum cobalamin level seen in WG3630. Some of the rare findings seen in a few cases of cblF were lethargy, neutropenia, and thrombocytopenia seen in WG4066, facial dysmorphism, mild cerebral atrophy, and neutropenia seen in WG4140, and hyperpigmentation and mild microcytic anemia seen in WG3630. Most cblF patients were diagnosed in the first year of age, some following positive newborn screening, although one was not diagnosed until age of 11 years. Similarly, WG4066, WG4140, and WG3630 were diagnosed at one month, 16 months, and eight years of age, respectively.

111

In the unusual WG3630 patient, Superose 12 analysis of fibroblasts was helpful in unveiling an accumulation of free, non-protein bound cobalamin and validate that this patient was suffering from a mild case of failure of lysosomal cobalamin export (Figure 9B). Increased homocysteine levels in four members of the WG3630 family were also peculiar (Figure 7A). Segregation analysis of the

MTHFR c.677C>T polymorphism suggested that the hyperhomocysteinemia in the sister may be attributed to a homozygous MTHFR polymorphism which is known to result in ~65% decrease in the enzyme activity.114 However, the homocysteine levels could not be exclusively explained by the MTHFR polymorphism and the ABCD4 mutation because the mother had identical genotypes with the father and the aunt but showed normal plasma homocysteine and cobalamin levels. Therefore, their homocysteine levels may be affected by combinations of the MTHFR polymorphism, the ABCD4 mutation, and other genetic/environmental/dietary factors.

Based on our current knowledge, it would be hard to distinguish cblF and cblJ patients on clinical and laboratory grounds alone. The cblF defect itself causes highly variable phenotypes in 15 reported patients. If a patient presented with a phenotype suggestive of cblF or cblJ defects, a definitive diagnosis would require somatic cell complementation analysis and/or sequencing of the LMBRD1 and ABCD4 genes. By identifying more cblF and cblJ patients, we may be able to catalogue the spectrum of phenotypes and discover differences between the two.

112

4.2 Role of ABCD4 in Vitamin B12 Metabolism

There has been little effort to investigate the ABCD4 protein compared to the other three proteins of the ABC transporter subfamily D, so the function and structure of it remains unexplored. While initially reported as a peroxisomal membrane protein, a later investigation reported ABCD4 as an ER membrane protein. However, our collaborating research team led by Dr. Frank Rutsch and Dr.

Matthias Baumgartner disputed the previous findings and observed a strong co- localization of ABCD4 with lysosomal markers, including LMBD1, using the fluorescence microscopy technology (data not shown). It could be argued that, through evolution, ABCD4 acquired a role in cobalamin metabolism and localized to the lysosomes unlike its peroxisomal homologues.

Another example of a heterodimer formed by two ABC half-transporters is the antigen peptide transporter formed by dimerization of TAP1/ABCB2 and

TAP2/ABCD3 which is responsible for transporting antigens from cytoplasm to

ER lumen for association with MHC class I molecules.142, 165 Given that ABCD4 is the only member of its subfamily localized to the lysosomes, it perhaps forms a homodimer for its function as the transporter of cobalamin.

In cobalamin synthesizing bacteria and archaea, cobalamin is synthesized in a biosynthetic pathway partly shared by heme, chlorophyll, and siroheme syntheses. It is predicted that the pathway first evolved to produce cobalamin.13

Knowing such ancient origin of cobalamin in life forms, we continue to discover

113 intricate steps involved in the metabolism of cobalamin in humans. To date, ABC importers have not been reported in eukaryotes but have been limited to prokaryotes.142, 148 With further studies, ABCD4 may be proven to be the first

ABC importer reported in the empire eukaryota because ABCD4 transports cobalamin into the cytoplasm, not out of it. In bacteria, the ABC transporter

BtuCD specifically transports cobalamin across the inner membrane and into the cytoplasm.166

In WG4066, the c.956A>G mutation resulted in a tyrosine to cysteine substitution (p.Y319C) predicted to be “probably damaging” by the PolyPhen-2 program (http://genetics.bwh.harvard.edu/pph2). On the other hand, the c.1746_1747insCT mutation resulted in a frameshifting mutation (p.E583LfsX9) affecting 8 downstream residues and then creating a premature stop codon (UGA).

By RT-PCR, the c.542+1G>T and c.1456G>T mutations in WG4140 were found to result in in-frame skipping of exon 5 (p.D143_S181del) and of exons 13 and 14

(p.G443_S485del) (data not shown). In WG3630, the homozygous c.423C>G mutation caused a asparagine to lysine substitution (p.N141K) also predicted to be

“probably damaging” by the PolyPhen-2. Moreover, both p.Y319C and p.N141K missense mutations occurred at residues evolutionarily conserved all the way down to zebra fish (http://genome.ucsc.edu). Therefore, all five mutations could severely damage the ABCD4 protein.

ABCD4 has been broadly divided into the N-terminal TMD (residues 39-

114

332) and the C-terminal NBD (residues 389-603), which are two fundamental domains of an ABC transporter (Figure 4). Collectively, three mutations, p.N141K, p.D143_S181del, and p.Y319C, occurred in the TMD while the other two mutations, p.G486C and p.E583LfsX9, occurred in the NBD. Based on these five mutations, no preference was observed for the location of mutations. It is possible that they affect domains or regions essential for cobalamin binding and transporting or protein-protein interaction.

In conclusion, the novel disorder, named cblJ, is an autosomal recessive disorder caused by mutations in the ABCD4 gene. Patients present with methylmalonic aciduria, hyperhomocysteinemia, and other symptoms also found in patients with the cblF defect. I suggest that ABCD4, an ABC transporter, is an essential component of intracellular cobalamin metabolism and interacts with

LMBD1 to allow transport of vitamin B12 out of the lysosome.

115

ORIGINAL CONTRIBUTIONS TO SCIENCE

 Identification and biochemical characterization of the cblJ disease, a novel

inborn error of vitamin B12 metabolism

 Identification of ABCD4 as the causative gene in the cblJ disease

 Discovery of overexpression of ABCD4 mildly correcting the cblF defect and

overexpression of LMBD1 not correcting the cblJ defect

116

BIBLIOGRAPHY

1. Voet D and Voet JG (2004) Biochemistry, 3rd ed. John Wiley & Sons, Inc, Hoboken, NJ 2. Banerjee R (2006) B12 trafficking in mammals: A case for coenzyme escort service. ACS Chem Biol 1 (3):149-159 3. Annibale B, Lahner E, and Fave GD (2011) Diagnosis and management of pernicious anemia. Curr Gastroenterol Rep 13 (6):518-524 4. Minot GR and Murphy WP (1926) Treatment of pernicious anemia by a special diet. J Am Med Assoc 87 (7):470-476 5. Smith EL (1948) Purification of anti-pernicious anaemia factors from liver. Nature 161 (4095):638 6. Rickes EL, Brink NG, Koniuszy FR, Wood TR, and Folkers K (1948) Crystalline Vitamin B12. Science 107 (2781):396-397 7. Shorb MS (1948) Activity of Vitamin B12 for the Growth of Lactobacillus lactis. Science 107 (2781):397-398 8. Banerjee R and Ragsdale SW (2003) The many faces of vitamin B12: catalysis by cobalamin-dependent enzymes. Annu Rev Biochem 72:209-247 9. Hodgkin DC, Kamper J, Mackay M, Pickworth J, Trueblood KN, and White JG (1956) Structure of vitamin B12. Nature 178 (4524):64-66 10. Motwani HV, Qiu S, Golding BT, Kylin H, and Tornqvist M (2011) Cob(I)alamin reacts with sucralose to afford an alkylcobalamin: relevance to in vivo cobalamin and sucralose interaction. Food Chem Toxicol 49 (4):750- 757 11. Krautler B (2005) Vitamin B12: chemistry and biochemistry. Biochem Soc Trans 33 (Pt:4):4-10 12. Watanabe F (2007) Vitamin B12 sources and bioavailability. Exp Biol Med (Maywood ) 232 (10):1266-1274 13. Roth JR, Lawrence JG, and Bobik TA (1996) Cobalamin (coenzyme B12): synthesis and biological significance. Annu Rev Microbiol 50:137-181 14. Randaccio L, Geremia S, Demitri N, and Wuerges J (2010) Vitamin B12: unique metalorganic compounds and the most complex vitamins. Molecules 15 (5):3228-3259 15. Barker HA, Rooze V, Suzuki F, and Iodice AA (1964) The Glutamate Mutase System. Assays And Properties. J Biol Chem 239:3260-3266 16. Krasotkina J, Walters T, Maruya KA, and Ragsdale SW (2001) Characterization of the B12- and iron-sulfur-containing reductive dehalogenase from Desulfitobacterium chlororespirans. J Biol Chem 276 (44):40991-40997 17. Kräutler B, Fieber W, Ostermann S, Fasching M, Ongania KH, Gruber K, Kratky C, Mikl C, Siebert A, and Diekert G (2003) The Cofactor of

117

Tetrachloroethene Reductive Dehalogenase of Dehalospirillum multivorans Is Norpseudo-B12, a New Type of a Natural Corrinoid. HCA 86 (11):3698- 3716 18. Watkins D and Rosenblatt DS (2011) Inborn errors of cobalamin absorption and metabolism. Am J Med Genet C Semin Med Genet 157 (1):33-44 19. Leclerc D, Campeau E, Goyette P, Adjalla CE, Christensen B, Ross M, Eydoux P, Rosenblatt DS, Rozen R, and Gravel RA (1996) Human methionine synthase: cDNA cloning and identification of mutations in patients of the cblG complementation group of folate/cobalamin disorders. Hum Mol Genet 5 (12):1867-1874 20. Moras E, Hosack A, Watkins D, and Rosenblatt DS (2007) Mitochondrial vitamin B12-binding proteins in patients with inborn errors of cobalamin metabolism. Mol Genet Metab 90 (2):140-147 21. Qureshi AA, Rosenblatt DS, and Cooper BA (1994) Inherited disorders of cobalamin metabolism. Crit Rev Oncol Hematol 17 (2):133-151 22. Li N, Seetharam S, and Seetharam B (1995) Genomic structure of human transcobalamin II: comparison to human intrinsic factor and transcobalamin I. Biochem Biophys Res Commun 208 (2):756-764 23. Froese DS and Gravel RA (2010) Genetic disorders of vitamin B12 metabolism: eight complementation groups - eight genes. Expert Rev Mol Med 12:e37 24. Morkbak AL, Poulsen SS, and Nexo E (2007) Haptocorrin in humans. Clin Chem Lab Med 45 (12):1751-1759 25. Quadros EV, Nakayama Y, and Sequeira JM (2009) The protein and the gene encoding the receptor for the cellular uptake of transcobalamin-bound cobalamin. Blood 113 (1):186-192 26. Johnston J, Bollekens J, Allen RH, and Berliner N (1989) Structure of the cDNA encoding transcobalamin I, a neutrophil granule protein. J Biol Chem 264 (27):15754-15757 27. Johnston J, Yang-Feng T, and Berliner N (1992) Genomic structure and mapping of the chromosomal gene for transcobalamin I (TCN1): comparison to human intrinsic factor. Genomics 12 (3):459-464 28. Carmel R (2003) Mild transcobalamin I (haptocorrin) deficiency and low serum cobalamin concentrations. Clin Chem 49 (8):1367-1374 29. Carmel R, Parker J, and Kelman Z (2009) Genomic mutations associated with mild and severe deficiencies of transcobalamin I (haptocorrin) that cause mildly and severely low serum cobalamin levels. Br J Haematol 147 (3):386-391 30. Hewitt JE, Gordon MM, Taggart RT, Mohandas TK, and Alpers DH (1991) Human gastric intrinsic factor: characterization of cDNA and genomic clones and localization to human chromosome 11. Genomics 10 (2):432-440

118

31. Yassin F, Rothenberg SP, Rao S, Gordon MM, Alpers DH, and Quadros EV (2004) Identification of a 4-base deletion in the gene in inherited intrinsic factor deficiency. Blood 103 (4):1515-1517 32. Yang YM, Ducos R, Rosenberg AJ, Catrou PG, Levine JS, Podell ER, and Allen RH (1985) Cobalamin malabsorption in three siblings due to an abnormal intrinsic factor that is markedly susceptible to acid and proteolysis. J Clin Invest 76 (6):2057-2065 33. Tanner SM, Li Z, Perko JD, Oner C, Cetin M, Altay C, Yurtsever Z, David KL, Faivre L, Ismail EA, Grasbeck R, and de la Chapelle A (2005) Hereditary juvenile cobalamin deficiency caused by mutations in the intrinsic factor gene. Proc Natl Acad Sci U S A 102 (11):4130-4133 34. Aminoff M, Carter JE, Chadwick RB, Johnson C, Grasbeck R, Abdelaal MA, Broch H, Jenner LB, Verroust PJ, Moestrup SK, de la Chapelle A, and Krahe R (1999) Mutations in CUBN, encoding the intrinsic factor-vitamin B12 receptor, cubilin, cause hereditary megaloblastic anaemia 1. Nat Genet 21 (3):309-313 35. Tanner SM, Aminoff M, Wright FA, Liyanarachchi S, Kuronen M, Saarinen A, Massika O, Mandel H, Broch H, and de la Chapelle A (2003) Amnionless, essential for mouse gastrulation, is mutated in recessive hereditary megaloblastic anemia. Nat Genet 33 (3):426-429 36. Fyfe JC, Madsen M, Hojrup P, Christensen EI, Tanner SM, de la Chapelle A, He Q, and Moestrup SK (2004) The functional cobalamin (vitamin B12)- intrinsic factor receptor is a novel complex of cubilin and amnionless. Blood 103 (5):1573-1579 37. Broch H, Imerslund O, Monn E, Hovig T, and Seip M (1984) Imerslund- Grasbeck anemia. A long-term follow-up study. Acta Paediatr Scand 73 (2):248-253 38. Tanner SM, Li Z, Bisson R, Acar C, Oner C, Oner R, Cetin M, Abdelaal MA, Ismail EA, Lissens W, Krahe R, Broch H, Grasbeck R, and de la Chapelle A (2004) Genetically heterogeneous selective intestinal malabsorption of vitamin B12: founder effects, consanguinity, and high clinical awareness explain aggregations in Scandinavia and the Middle East. Hum Mutat 23 (4):327-333 39. Carmel R (2007) The disappearance of cobalamin absorption testing: a critical diagnostic loss. J Nutr 137 (11):2481-2484 40. Ament AE, Li Z, Sturm AC, Perko JD, Lawson S, Masterson M, Quadros EV, and Tanner SM (2009) Juvenile cobalamin deficiency in individuals of African ancestry is caused by a founder mutation in the intrinsic factor gene GIF. Br J Haematol 144 (4):622-624 41. Regec A, Quadros EV, Platica O, and Rothenberg SP (1995) The cloning and characterization of the human transcobalamin II gene. Blood 85

119

(10):2711-2719 42. Seligman PA, Steiner LL, and Allen RH (1980) Studies of a patient with megaloblastic anemia and an abnormal transcobalamin II. N Engl J Med 303 (21):1209-1212 43. Schiff M, Ogier de BH, Bard G, Barlogis V, Hamel C, Moat SJ, Odent S, Shortland G, Touati G, and Giraudier S (2010) Should transcobalamin deficiency be treated aggressively? J Inherit Metab Dis 33 (3):223-229 44. Anastasio, N., Watkins, D., Vezina, L., Dempsey-Nunez, L., Reichman, L., Quadros, E. V., and Rosenblatt, D. S. Mutations in TCblR, the gene for the transcobalamin receptor, result in decreased cellular uptake of vitamin B12 and methylmalonic aciduria; Program #615. Presented at the 11th International Congress of Inborn Errors of Metabolism, August 29- September 3, 2009, San Diego, California. 45. Quadros EV, Lai SC, Nakayama Y, Sequeira JM, Hannibal L, Wang S, Jacobsen DW, Fedosov S, Wright E, Gallagher RC, Anastasio N, Watkins D, and Rosenblatt DS (2010) Positive newborn screen for methylmalonic aciduria identifies the first mutation in TCblR/CD320, the gene for cellular uptake of transcobalamin-bound vitamin B(12). Hum Mutat 31 (8):924-929 46. Moestrup SK and Verroust PJ (2001) Megalin- and cubilin-mediated endocytosis of protein-bound vitamins, lipids, and hormones in polarized epithelia. Annu Rev Nutr 21:407-428 47. Beedholm-Ebsen R, van de Wetering K, Hardlei T, Nexo E, Borst P, and Moestrup SK (2010) Identification of multidrug resistance protein 1 (MRP1/ABCC1) as a molecular gate for cellular export of cobalamin. Blood 115 (8):1632-1639 48. Kantarci S, Al-Gazali L, Hill RS, Donnai D, Black GC, Bieth E, Chassaing N, Lacombe D, Devriendt K, Teebi A, Loscertales M, Robson C, Liu T, MacLaughlin DT, Noonan KM, Russell MK, Walsh CA, Donahoe PK, and Pober BR (2007) Mutations in LRP2, which encodes the multiligand receptor megalin, cause Donnai-Barrow and facio-oculo-acoustico-renal syndromes. Nat Genet 39 (8):957-959 49. Pober BR, Longoni M, and Noonan KM (2009) A review of Donnai-Barrow and facio-oculo-acoustico-renal (DB/FOAR) syndrome: clinical features and differential diagnosis. Birth Defects Res A Clin Mol Teratol 85 (1):76- 81 50. Beedholm-Ebsen R, van de Wetering K, Hardlei T, Nexo E, Borst P, and Moestrup SK (2009) Identification of multidrug resistance protein 1 (MRP1/ABCC1) as a molecular gate for cellular export of cobalamin. Blood 51. Cole SP, Bhardwaj G, Gerlach JH, Mackie JE, Grant CE, Almquist KC, Stewart AJ, Kurz EU, Duncan AM, and Deeley RG (1992) Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line.

120

Science 258 (5088):1650-1654 52. Deeley RG, Westlake C, and Cole SP (2006) Transmembrane transport of endo- and xenobiotics by mammalian ATP-binding cassette multidrug resistance proteins. Physiol Rev 86 (3):849-899 53. Shah NP, Beech CM, Sturm AC, and Tanner SM (2011) Investigation of the ABC transporter MRP1 in selected patients with presumed defects in vitamin B12 absorption. Blood 117 (16):4397-4398 54. Hogenkamp HP (1968) Enzymatic reactions involving corrinoids. Annu Rev Biochem 37:225-245 55. Oberholzer VG, Levin B, Burgess EA, and Young WF (1967) Methylmalonic aciduria. An inborn error of metabolism leading to chronic metabolic acidosis. Arch Dis Child 42 (225):492-504 56. Rosenberg LE, Lilljeqvist A, and Hsia YE (1968) Methylmalonic aciduria: metabolic block localization and vitamin B12 dependency. Science 162 (3855):805-807 57. Rosenberg LE, Lilljeqvist AC, and Hsia YE (1968) Methylmalonic aciduria. An inborn error leading to metabolic acidosis, long-chain ketonuria and intermittent hyperglycinemia. N Engl J Med 278 (24):1319-1322 58. Rosenberg LE, Lilljeqvist AC, Hsia YE, and Rosenbloom FM (1969) Vitamin B12 dependent methylmalonicaciduria: defective B12 metabolism in cultured fibroblasts. Biochem Biophys Res Commun 37 (4):607-614 59. Morrow G, III, Barness LA, Cardinale GJ, Abeles RH, and Flaks JG (1969) Congenital methylmalonic acidemia: enzymatic evidence for two forms of the disease. Proc Natl Acad Sci U S A 63 (1):191-197 60. Watkins D, Matiaszuk N, and Rosenblatt DS (2000) Complementation studies in the cblA class of inborn error of cobalamin metabolism: evidence for interallelic complementation and for a new complementation class (cblH). J Med Genet 37 (7):510-513 61. Gravel RA, Mahoney MJ, Ruddle FH, and Rosenberg LE (1975) Genetic complementation in heterokaryons of human fibroblasts defective in cobalamin metabolism. Proc Natl Acad Sci U S A 72 (8):3181-3185 62. Mudd SH, Levy HL, Abeles RH, and Jennedy JP, Jr. (1969) A derangement in B12 metabolism leading to homocystinemia, cystathioninemia and methylmalonic aciduria. Biochem Biophys Res Commun 35 (1):121-126 63. Mahoney MJ, Rosenberg LE, Mudd SH, and Uhlendorf BW (1971) Defective metabolism of vitamin B12 in fibroblasts from children with methylmalonicaciduria. Biochem Biophys Res Commun 44 (2):375-381 64. Mahoney MJ, Hart AC, Steen VD, and Rosenberg LE (1975) Methylmalonicacidemia: biochemical heterogeneity in defects of 5'- deoxyadenosylcobalamin synthesis. Proc Natl Acad Sci U S A 72 (7):2799- 2803

121

65. Willard HF and Rosenberg LE (1977) Inherited deficiencies of human methylmalonyl CoA mutase activity: reduced affinity of mutant apoenzyme for adenosylcobalamin. Biochem Biophys Res Commun 78 (3):927-934 66. Willard HF and Rosenberg LE (1980) Inherited methylmalonyl CoA mutase apoenzyme deficiency in human fibroblasts: evidence for allelic heterogeneity, genetic compounds, and codominant expression. J Clin Invest 65 (3):690-698 67. Goodman SI, Moe PG, Hammond KB, Mudd SH, and Uhlendorf BW (1970) Homocystinuria with methylmalonic aciduria: two cases in a sibship. Biochem Med 4 (5):500-515 68. Willard HF, Mellman IS, and Rosenberg LE (1978) Genetic complementation among inherited deficiencies of methylmalonyl-CoA mutase activity: evidence for a new class of human cobalamin mutant. Am J Hum Genet 30 (1):1-13 69. Cooper BA, Rosenblatt DS, and Watkins D (1990) Methylmalonic aciduria due to a new defect in adenosylcobalamin accumulation by cells. Am J Hematol 34 (2):115-120 70. Suormala T, Baumgartner MR, Coelho D, Zavadakova P, Kozich V, Koch HG, Berghauser M, Wraith JE, Burlina A, Sewell A, Herwig J, and Fowler B (2004) The cblD defect causes either isolated or combined deficiency of methylcobalamin and adenosylcobalamin synthesis. J Biol Chem 279 (41):42742-42749 71. Levy HL, Mudd SH, Schulman JD, Dreyfus PM, and Abeles RH (1970) A derangement in B12 metabolism associated with homocystinemia, cystathioninemia, hypomethioninemia and methylmalonic aciduria. Am J Med 48 (3):390-397 72. Schuh S, Rosenblatt DS, Cooper BA, Schroeder ML, Bishop AJ, Seargeant LE, and Haworth JC (1984) Homocystinuria and megaloblastic anemia responsive to vitamin B12 therapy. An inborn error of metabolism due to a defect in cobalamin metabolism. N Engl J Med 310 (11):686-690 73. Rosenblatt DS, Cooper BA, Pottier A, Lue-Shing H, Matiaszuk N, and Grauer K (1984) Altered vitamin B12 metabolism in fibroblasts from a patient with megaloblastic anemia and homocystinuria due to a new defect in methionine biosynthesis. J Clin Invest 74 (6):2149-2156 74. Rosenblatt DS, Hosack A, Matiaszuk NV, Cooper BA, and Laframboise R (1985) Defect in vitamin B12 release from lysosomes: newly described inborn error of vitamin B12 metabolism. Science 228 (4705):1319-1321 75. Watkins D and Rosenblatt DS (1986) Failure of lysosomal release of vitamin B12: a new complementation group causing methylmalonic aciduria (cblF). Am J Hum Genet 39 (3):404-408 76. Rosenblatt DS, Thomas IT, Watkins D, Cooper BA, and Erbe RW (1987)

122

Vitamin B12 responsive homocystinuria and megaloblastic anemia: heterogeneity in methylcobalamin deficiency. Am J Med Genet 26 (2):377- 383 77. Hallam LJ, Sawyer M, Clark AC, and Van der Weyden MB (1987) Vitamin B12-responsive neonatal megaloblastic anemia and homocystinuria with associated reduced methionine synthase activity. Blood 69 (4):1128-1133 78. Watkins D and Rosenblatt DS (1988) Genetic heterogeneity among patients with methylcobalamin deficiency. Definition of two complementation groups, cblE and cblG. J Clin Invest 81 (6):1690-1694 79. Youngdahl-Turner P, Rosenberg LE, and Allen RH (1978) Binding and uptake of transcobalamin II by human fibroblasts. J Clin Invest 61 (1):133- 141 80. Rutsch F, Gailus S, Miousse IR, Suormala T, Sagne C, Toliat MR, Nurnberg G, Wittkampf T, Buers I, Sharifi A, Stucki M, Becker C, Baumgartner M, Robenek H, Marquardt T, Hohne W, Gasnier B, Rosenblatt DS, Fowler B, and Nurnberg P (2009) Identification of a putative lysosomal cobalamin exporter altered in the cblF defect of vitamin B12 metabolism. Nat Genet 41 (2):234-239 81. Gailus S, Suormala T, Malerczyk-Aktas AG, Toliat MR, Wittkampf T, Stucki M, Nurnberg P, Fowler B, Hennermann JB, and Rutsch F (2010) A novel mutation in LMBRD1 causes the cblF defect of vitamin B12 metabolism in a Turkish patient. J Inherit Metab Dis 33 (1):17-24 82. Miousse IR, Watkins D, and Rosenblatt DS (2011) Novel splice site mutations and a large deletion in three patients with the cblF inborn error of vitamin B12 metabolism. Mol Genet Metab 102 (4):505-507 83. Vassiliadis A, Rosenblatt DS, Cooper BA, and Bergeron JJ (1991) Lysosomal cobalamin accumulation in fibroblasts from a patient with an inborn error of cobalamin metabolism (cblF complementation group): visualization by electron microscope radioautography. Exp Cell Res 195 (2):295-302 84. Lerner-Ellis JP, Tirone JC, Pawelek PD, Dore C, Atkinson JL, Watkins D, Morel CF, Fujiwara TM, Moras E, Hosack AR, Dunbar GV, Antonicka H, Forgetta V, Dobson CM, Leclerc D, Gravel RA, Shoubridge EA, Coulton JW, Lepage P, Rommens JM, Morgan K, and Rosenblatt DS (2006) Identification of the gene responsible for methylmalonic aciduria and homocystinuria, cblC type. Nat Genet 38 (1):93-100 85. Mellman I, Willard HF, and Rosenberg LE (1978) Cobalamin binding and cobalamin-dependent enzyme activity in normal and mutant human fibroblasts. J Clin Invest 62 (5):952-960 86. Mellman I, Willard HF, Youngdahl-Turner P, and Rosenberg LE (1979) Cobalamin coenzyme synthesis in normal and mutant human fibroblasts.

123

Evidence for a processing enzyme activity deficient in cblC cells. J Biol Chem 254 (23):11847-11853 87. Kim J, Gherasim C, and Banerjee R (2008) Decyanation of vitamin B12 by a trafficking chaperone. Proc Natl Acad Sci U S A 105 (38):14551-14554 88. Hannibal L, Kim J, Brasch NE, Wang S, Rosenblatt DS, Banerjee R, and Jacobsen DW (2009) Processing of alkylcobalamins in mammalian cells: A role for the MMACHC (cblC) gene product. Mol Genet Metab 97 (4):260- 266 89. Plesa M, Kim J, Paquette SG, Gagnon H, Ng-Thow-Hing C, Gibbs BF, Hancock MA, Rosenblatt DS, and Coulton JW (2011) Interaction between MMACHC and MMADHC, two human proteins participating in intracellular vitamin B(1)(2) metabolism. Mol Genet Metab 102 (2):139- 148 90. Koutmos M, Gherasim C, Smith JL, and Banerjee R (2011) Structural basis of multifunctionality in a vitamin B12-processing enzyme. J Biol Chem 286 (34):29780-29787 91. Froese DS, Krojer T, Wu X, Shrestha R, Kiyani W, von DF, Gravel RA, Oppermann UC, and Yue WW (2012) Structure of MMACHC reveals an arginine-rich pocket and a domain-swapped dimer for its B12 processing function. Biochemistry 92. Coelho D, Suormala T, Stucki M, Lerner-Ellis JP, Rosenblatt DS, Newbold RF, Baumgartner MR, and Fowler B (2008) Gene identification for the cblD defect of vitamin B12 metabolism. N Engl J Med 358 (14):1454-1464 93. Miousse IR, Watkins D, Coelho D, Rupar T, Crombez EA, Vilain E, Bernstein JA, Cowan T, Lee-Messer C, Enns GM, Fowler B, and Rosenblatt DS (2009) Clinical and molecular heterogeneity in patients with the cblD inborn error of cobalamin metabolism. J Pediatr 154 (4):551-556 94. Stucki M, Coelho D, Suormala T, Burda P, Fowler B, and Baumgartner MR (2012) Molecular mechanisms leading to three different phenotypes in the cblD defect of intracellular cobalamin metabolism. Hum Mol Genet 21 (6):1410-1418 95. Dobson CM, Wai T, Leclerc D, Kadir H, Narang M, Lerner-Ellis JP, Hudson TJ, Rosenblatt DS, and Gravel RA (2002) Identification of the gene responsible for the cblB complementation group of vitamin B12-dependent methylmalonic aciduria. Hum Mol Genet 11 (26):3361-3369 96. Schubert HL and Hill CP (2006) Structure of ATP-bound human ATP:cobalamin adenosyltransferase. Biochemistry 45 (51):15188-15196 97. Lerner-Ellis JP, Gradinger AB, Watkins D, Tirone JC, Villeneuve A, Dobson CM, Montpetit A, Lepage P, Gravel RA, and Rosenblatt DS (2006) Mutation and biochemical analysis of patients belonging to the cblB complementation class of vitamin B12-dependent methylmalonic aciduria.

124

Mol Genet Metab 87 (3):219-225 98. Jorge-Finnigan A, Aguado C, Sanchez-Alcudia R, Abia D, Richard E, Merinero B, Gamez A, Banerjee R, Desviat LR, Ugarte M, and Perez B (2010) Functional and structural analysis of five mutations identified in methylmalonic aciduria cblB type. Hum Mutat 31 (9):1033-1042 99. Dobson CM, Wai T, Leclerc D, Wilson A, Wu X, Dore C, Hudson T, Rosenblatt DS, and Gravel RA (2002) Identification of the gene responsible for the cblA complementation group of vitamin B12-responsive methylmalonic acidemia based on analysis of prokaryotic gene arrangements. Proc Natl Acad Sci U S A 99 (24):15554-15559 100. Merinero B, Perez B, Perez-Cerda C, Rincon A, Desviat LR, Martinez MA, Sala PR, Garcia MJ, Aldamiz-Echevarria L, Campos J, Cornejo V, Del TM, Mahfoud A, Martinez-Pardo M, Parini R, Pedron C, Pena-Quintana L, Perez M, Pourfarzam M, and Ugarte M (2008) Methylmalonic acidaemia: examination of genotype and biochemical data in 32 patients belonging to mut, cblA or cblB complementation group. J Inherit Metab Dis 31 (1):55-66 101. Martinez MA, Rincon A, Desviat LR, Merinero B, Ugarte M, and Perez B (2005) Genetic analysis of three genes causing isolated methylmalonic acidemia: identification of 21 novel allelic variants. Mol Genet Metab 84 (4):317-325 102. Lerner-Ellis JP, Dobson CM, Wai T, Watkins D, Tirone JC, Leclerc D, Dore C, Lepage P, Gravel RA, and Rosenblatt DS (2004) Mutations in the MMAA gene in patients with the cblA disorder of vitamin B12 metabolism. Hum Mutat 24 (6):509-516 103. Hubbard PA, Padovani D, Labunska T, Mahlstedt SA, Banerjee R, and Drennan CL (2007) Crystal structure and mutagenesis of the metallochaperone MeaB: insight into the causes of methylmalonic aciduria. J Biol Chem 282 (43):31308-31316 104. Froese DS, Kochan G, Muniz JR, Wu X, Gileadi C, Ugochukwu E, Krysztofinska E, Gravel RA, Oppermann U, and Yue WW (2010) Structures of the human GTPase MMAA and vitamin B12-dependent methylmalonyl- CoA mutase and insight into their complex formation. J Biol Chem 285 (49):38204-38213 105. Padovani D and Banerjee R (2009) A G-protein editor gates coenzyme B12 loading and is corrupted in methylmalonic aciduria. Proc Natl Acad Sci U S A 106 (51):21567-21572 106. Jansen R, Kalousek F, Fenton WA, Rosenberg LE, and Ledley FD (1989) Cloning of full-length methylmalonyl-CoA mutase from a cDNA library using the polymerase chain reaction. Genomics 4 (2):198-205 107. Ledley FD, Lumetta M, Nguyen PN, Kolhouse JF, and Allen RH (1988) Molecular cloning of L-methylmalonyl-CoA mutase: gene transfer and

125

analysis of mut cell lines. Proc Natl Acad Sci U S A 85 (10):3518-3521 108. Worgan LC, Niles K, Tirone JC, Hofmann A, Verner A, Sammak A, Kucic T, Lepage P, and Rosenblatt DS (2006) Spectrum of mutations in mut methylmalonic acidemia and identification of a common Hispanic mutation and haplotype. Hum Mutat 27 (1):31-43 109. Leclerc D, Wilson A, Dumas R, Gafuik C, Song D, Watkins D, Heng HH, Rommens JM, Scherer SW, Rosenblatt DS, and Gravel RA (1998) Cloning and mapping of a cDNA for methionine synthase reductase, a flavoprotein defective in patients with homocystinuria. Proc Natl Acad Sci U S A 95 (6):3059-3064 110. Wilson A, Leclerc D, Rosenblatt DS, and Gravel RA (1999) Molecular basis for methionine synthase reductase deficiency in patients belonging to the cblE complementation group of disorders in folate/cobalamin metabolism. Hum Mol Genet 8 (11):2009-2016 111. Zavadakova P, Fowler B, Suormala T, Novotna Z, Mueller P, Hennermann JB, Zeman J, Vilaseca MA, Vilarinho L, Gutsche S, Wilichowski E, Horneff G, and Kozich V (2005) cblE type of homocystinuria due to methionine synthase reductase deficiency: functional correction by minigene expression. Hum Mutat 25 (3):239-247 112. 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, and Rosenblatt DS (2002) Hyperhomocysteinemia due to methionine synthase deficiency, cblG: structure of the MTR gene, genotype diversity, and recognition of a common mutation, P1173L. Am J Hum Genet 71 (1):143-153 113. Wilson A, Leclerc D, Saberi F, Campeau E, Hwang HY, Shane B, Phillips JA, III, Rosenblatt DS, and Gravel RA (1998) Functionally null mutations in patients with the cblG-variant form of methionine synthase deficiency. Am J Hum Genet 63 (2):409-414 114. Carmel R and Jacobsen DW (2001) Homocysteine in Health and Disease. Cambridge University Press, Cambridge, United Kingdom 115. Gradinger AB, Belair C, Worgan LC, Li CD, Lavallee J, Roquis D, Watkins D, and Rosenblatt DS (2007) Atypical methylmalonic aciduria: frequency of mutations in the methylmalonyl CoA epimerase gene (MCEE). Hum Mutat 28 (10):1045 116. Ostergaard E (2008) Disorders caused by deficiency of succinate-CoA ligase. J Inherit Metab Dis 117. Alfares A, Nunez LD, Al-Thihli K, Mitchell J, Melancon S, Anastasio N, Ha KCH, Majewski J, Rosenblatt DS, and Braverman N (2011) Combined malonic and methylmalonic aciduria: exome sequencing reveals mutations in the ACSF3 gene in patients with a non-classic phenotype. J Med Genet

126

48 (9):602-605 118. Sloan JL, Johnston JJ, Manoli I, Chandler RJ, Krause C, Carrillo-Carrasco N, Chandrasekaran SD, Sysol JR, O'Brien K, Hauser NS, Sapp JC, Dorward HM, Huizing M, Barshop BA, Berry SA, James PM, Champaigne NL, de LP, Valayannopoulos V, Geschwind MD, Gavrilov DK, Nyhan WL, Biesecker LG, and Venditti CP (2011) Exome sequencing identifies ACSF3 as a cause of combined malonic and methylmalonic aciduria. Nat Genet 43 (9):883-886 119. Salomons GS, Jakobs C, Pope LL, Errami A, Potter M, Nowaczyk M, Olpin S, Manning N, Raiman JA, Slade T, Champion MP, Peck D, Gavrilov D, Hillman R, Hoganson GE, Donaldson K, Shield JP, Ketteridge D, Wasserstein M, and Gibson KM (2007) Clinical, enzymatic and molecular characterization of nine new patients with malonyl-coenzyme A decarboxylase deficiency. J Inherit Metab Dis 30 (1):23-28 120. Goyette P, Sumner JS, Milos R, Duncan AM, Rosenblatt DS, Matthews RG, and Rozen R (1994) Human methylenetetrahydrofolate reductase: isolation of cDNA, mapping and mutation identification. Nat Genet 7 (2):195-200 121. Kraus JP, Le K, Swaroop M, Ohura T, Tahara T, Rosenberg LE, Roper MD, and Kozich V (1993) Human cystathionine beta-synthase cDNA: sequence, alternative splicing and expression in cultured cells. Hum Mol Genet 2 (10):1633-1638 122. Vellodi A (2005) Lysosomal storage disorders. Br J Haematol 128 (4):413- 431 123. Saftig P and Klumperman J (2009) Lysosome biogenesis and lysosomal membrane proteins: trafficking meets function. Nat Rev Mol Cell Biol 10 (9):623-635 124. Mellquist JL, Kasturi L, Spitalnik SL, and Shakin-Eshleman SH (1998) The amino acid following an asn-X-Ser/Thr sequon is an important determinant of N-linked core glycosylation efficiency. Biochemistry 37 (19):6833-6837 125. Braulke T and Bonifacino JS (2009) Sorting of lysosomal proteins. Biochim Biophys Acta 1793 (4):605-614 126. Futerman AH and van MG (2004) The cell biology of lysosomal storage disorders. Nat Rev Mol Cell Biol 5 (7):554-565 127. Ruivo R, Anne C, Sagne C, and Gasnier B (2009) Molecular and cellular basis of lysosomal transmembrane protein dysfunction. Biochim Biophys Acta 1793 (4):636-649 128. Oladipo O, Rosenblatt DS, Watkins D, Miousse IR, Sprietsma L, Dietzen DJ, and Shinawi M (2011) Cobalamin F disease detected by newborn screening and follow-up on a 14-year-old patient. Pediatrics 128 (6):e1636-e1640 129. Gailus S, Hohne W, Gasnier B, Nurnberg P, Fowler B, and Rutsch F (2010) Insights into lysosomal cobalamin trafficking: lessons learned from cblF

127

disease. J Mol Med (Berl) 88 (5):459-466 130. Gusella JF, Wexler NS, Conneally PM, Naylor SL, Anderson MA, Tanzi RE, Watkins PC, Ottina K, Wallace MR, and Sakaguchi AY (1983) A polymorphic DNA marker genetically linked to Huntington's disease. Nature 306 (5940):234-238 131. Klein RJ, Zeiss C, Chew EY, Tsai JY, Sackler RS, Haynes C, Henning AK, SanGiovanni JP, Mane SM, Mayne ST, Bracken MB, Ferris FL, Ott J, Barnstable C, and Hoh J (2005) Complement factor H polymorphism in age-related macular degeneration. Science 308 (5720):385-389 132. Altshuler D, Daly MJ, and Lander ES (2008) Genetic mapping in human disease. Science 322 (5903):881-888 133. Ng PC, Levy S, Huang J, Stockwell TB, Walenz BP, Li K, Axelrod N, Busam DA, Strausberg RL, and Venter JC (2008) Genetic variation in an individual human exome. PLoS Genet 4 (8):e1000160 134. Ng SB, Turner EH, Robertson PD, Flygare SD, Bigham AW, Lee C, Shaffer T, Wong M, Bhattacharjee A, Eichler EE, Bamshad M, Nickerson DA, and Shendure J (2009) Targeted capture and massively parallel sequencing of 12 human exomes. Nature 461 (7261):272-276 135. Ng SB, Buckingham KJ, Lee C, Bigham AW, Tabor HK, Dent KM, Huff CD, Shannon PT, Jabs EW, Nickerson DA, Shendure J, and Bamshad MJ (2010) Exome sequencing identifies the cause of a mendelian disorder. Nat Genet 42 (1):30-35 136. Bamshad MJ, Ng SB, Bigham AW, Tabor HK, Emond MJ, Nickerson DA, and Shendure J (2011) Exome sequencing as a tool for Mendelian disease gene discovery. Nat Rev Genet 12 (11):745-755 137. Campeau PM, Kim JC, Lu JT, Schwartzentruber JA, Abdul-Rahman OA, Schlaubitz S, Murdock DM, Jiang MM, Lammer EJ, Enns GM, Rhead WJ, Rowland J, Robertson SP, Cormier-Daire V, Bainbridge MN, Yang XJ, Gingras MC, Gibbs RA, Rosenblatt DS, Majewski J, and Lee BH (2012) Mutations in KAT6B, encoding a histone acetyltransferase, cause Genitopatellar syndrome. Am J Hum Genet 90 (2):282-289 138. Hoischen A, van Bon BW, Rodriguez-Santiago B, Gilissen C, Vissers LE, de VP, Janssen I, van LB, Hastings R, Smithson SF, Newbury-Ecob R, Kjaergaard S, Goodship J, McGowan R, Bartholdi D, Rauch A, Peippo M, Cobben JM, Wieczorek D, Gillessen-Kaesbach G, Veltman JA, Brunner HG, and de Vries BB (2011) De novo nonsense mutations in ASXL1 cause Bohring-Opitz syndrome. Nat Genet 43 (8):729-731 139. Gilissen C, Hoischen A, Brunner HG, and Veltman JA (2012) Disease gene identification strategies for exome sequencing. Eur J Hum Genet 20 (5):490-497 140. Rosenblatt DS (2011) A RaDiCAL approach to gene discovery. J Med Genet

128

48 (9):577-578 141. Watkins D, Schwartzentruber JA, Ganesh J, Orange JS, Kaplan BS, Nunez LD, Majewski J, and Rosenblatt DS (2011) Novel inborn error of folate metabolism: identification by exome capture and sequencing of mutations in the MTHFD1 gene in a single proband. J Med Genet 48 (9):590-592 142. Davidson AL, Dassa E, Orelle C, and Chen J Structure, function, and evolution of bacterial ATP-binding cassette systems. Microbiol Mol Biol Rev 72 (2):June 143. Roninson IB, Chin JE, Choi KG, Gros P, Housman DE, Fojo A, Shen DW, Gottesman MM, and Pastan I (1986) Isolation of human mdr DNA sequences amplified in multidrug-resistant KB carcinoma cells. Proc Natl Acad Sci U S A 83 (12):4538-4542 144. Rommens JM, Iannuzzi MC, Kerem B, Drumm ML, Melmer G, Dean M, Rozmahel R, Cole JL, Kennedy D, Hidaka N, and . (1989) Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 245 (4922):1059-1065 145. Mosser J, Douar AM, Sarde CO, Kioschis P, Feil R, Moser H, Poustka AM, Mandel JL, and Aubourg P (1993) Putative X-linked adrenoleukodystrophy gene shares unexpected homology with ABC transporters. Nature 361 (6414):726-730 146. Brooks-Wilson A, Marcil M, Clee SM, Zhang LH, Roomp K, van DM, Yu L et al (1999) Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency. Nat Genet 22 (4):336-345 147. Theodoulou FL, Holdsworth M, and Baker A (2006) Peroxisomal ABC transporters. FEBS Lett 580 (4):1139-1155 148. Morita M and Imanaka T (2012) Peroxisomal ABC transporters: Structure, function and role in disease. Biochim Biophys Acta 149. Holzinger A, Kammerer S, Berger J, and Roscher AA (1997) cDNA cloning and mRNA expression of the human adrenoleukodystrophy related protein (ALDRP), a peroxisomal ABC transporter. Biochem Biophys Res Commun 239 (1):261-264 150. Gartner J, Moser H, and Valle D (1992) Mutations in the 70K peroxisomal membrane protein gene in Zellweger syndrome. Nat Genet 1 (1):16-23 151. Kamijo K, Kamijo T, Ueno I, Osumi T, and Hashimoto T (1992) Nucleotide sequence of the human 70 kDa peroxisomal membrane protein: a member of ATP-binding cassette transporters. Biochim Biophys Acta 1129 (3):323- 327 152. Shani N, Jimenez-Sanchez G, Steel G, Dean M, and Valle D (1997) Identification of a fourth half ABC transporter in the human peroxisomal membrane. Hum Mol Genet 6 (11):1925-1931 153. Holzinger A, Kammerer S, and Roscher AA (1997) Primary structure of

129

human PMP69, a putative peroxisomal ABC-transporter. Biochem Biophys Res Commun 237 (1):152-157 154. Liu LX, Janvier K, Berteaux-Lecellier V, Cartier N, Benarous R, and Aubourg P (1999) Homo- and heterodimerization of peroxisomal ATP- binding cassette half-transporters. J Biol Chem 274 (46):32738-32743 155. Hillebrand M, Verrier SE, Ohlenbusch A, Schafer A, Soling HD, Wouters FS, and Gartner J (2007) Live cell FRET microscopy: homo- and heterodimerization of two human peroxisomal ABC transporters, the adrenoleukodystrophy protein (ALDP, ABCD1) and PMP70 (ABCD3). J Biol Chem 282 (37):26997-27005 156. Kashiwayama Y, Seki M, Yasui A, Murasaki Y, Morita M, Yamashita Y, Sakaguchi M, Tanaka Y, and Imanaka T (2009) 70-kDa peroxisomal membrane protein related protein (P70R/ABCD4) localizes to endoplasmic reticulum not peroxisomes, and NH2-terminal hydrophobic property determines the subcellular localization of ABC subfamily D proteins. Exp Cell Res 315 (2):190-205 157. Kikuchi M, Hatano N, Yokota S, Shimozawa N, Imanaka T, and Taniguchi H (2004) Proteomic analysis of rat liver peroxisome: presence of peroxisome-specific isozyme of Lon protease. J Biol Chem 279 (1):421-428 158. Islinger M, Luers GH, Li KW, Loos M, and Volkl A (2007) Rat liver peroxisomes after fibrate treatment. A survey using quantitative mass spectrometry. J Biol Chem 282 (32):23055-23069 159. Wiese S, Gronemeyer T, Ofman R, Kunze M, Grou CP, Almeida JA, Eisenacher M, Stephan C, Hayen H, Schollenberger L, Korosec T, Waterham HR, Schliebs W, Erdmann R, Berger J, Meyer HE, Just W, Azevedo JE, Wanders RJ, and Warscheid B (2007) Proteomics characterization of mouse kidney peroxisomes by tandem mass spectrometry and protein correlation profiling. Mol Cell Proteomics 6 (12):2045-2057 160. Gloeckner CJ, Mayerhofer PU, Landgraf P, Muntau AC, Holzinger A, Gerber JK, Kammerer S, Adamski J, and Roscher AA (2000) Human adrenoleukodystrophy protein and related peroxisomal ABC transporters interact with the peroxisomal assembly protein PEX19p. Biochem Biophys Res Commun 271 (1):144-150 161. Yao J and Shoubridge EA (1999) Expression and functional analysis of SURF1 in Leigh syndrome patients with cytochrome c oxidase deficiency. Hum Mol Genet 8 (13):2541-2549 162. Watkins D (1998) Cobalamin metabolism in methionine-dependent human tumour and leukemia cell lines. Clin Invest Med 21 (3):151-158 163. Allen RH (1976) The plasma transport of vitamin B12. Br J Haematol 33 (2):161-171

130

164. Haack TB, Danhauser K, Haberberger B, Hoser J, Strecker V, Boehm D, Uziel G, Lamantea E, Invernizzi F, Poulton J, Rolinski B, Iuso A, Biskup S, Schmidt T, Mewes HW, Wittig I, Meitinger T, Zeviani M, and Prokisch H (2010) Exome sequencing identifies ACAD9 mutations as a cause of complex I deficiency. Nat Genet 42 (12):1131-1134 165. Powis SH, Mockridge I, Kelly A, Kerr LA, Glynne R, Gileadi U, Beck S, and Trowsdale J (1992) Polymorphism in a second ABC transporter gene located within the class II region of the human major histocompatibility complex. Proc Natl Acad Sci U S A 89 (4):1463-1467 166. Locher KP, Lee AT, and Rees DC (2002) The E. coli BtuCD structure: a framework for ABC transporter architecture and mechanism. Science 296 (5570):1091-1098

131

APPENDIX A: List of Publications and Presentations

Publications

1. Plesa M, Kim J, Paquette SG, Gagnon H, Ng-Thow-Hing C, Gibbs BF, Hancock MA, Rosenblatt DS, and Coulton JW (2011) Interaction between MMACHC and MMADHC, two human proteins participating in intracellular vitamin B12 metabolism. Molecular Genetics and Metabolism 102 (2):139-148

2. Campeau PM*, Kim JC*, Lu JT, Schwartzentruber JA, Abdul-Rahman OA, Schlaubitz S, Murdock DM, Jiang MM, Lammer EJ, Enns GM, Rhead WJ, Rowland J, Robertson SP, Cormier-Daire V, Bainbridge MN, Yang XJ, Gingras MC, Gibbs RA, Rosenblatt DS, Majewski J, and Lee BH (2012) Mutations in KAT6B, encoding a histone acetyltransferase, cause Genitopatellar syndrome. American Journal of Human Genetics 90 (2):282-289 * These authors contributed equally to this work

3. Deme JC, Miousse IR, Plesa M, Kim JC, Hancock MA, Mah W, Rosenblatt DS, Coulton JW (2012) Structural features of recombinant MMADHC isoforms and their interactions with MMACHC, proteins of mammalian vitamin B12 metabolism. Molecular Genetics and Metabolism, doi:10.1016/j.ymgme.2012.07.001 (Article in press)

4. Coelho D*, Kim JC*, Miousse IR, Fung S, du Moulin M, Buers I, Suormala T, Burda P, Frapolli M, Stucki M, Nürnberg P, Thiele H, Robenek H, Höhne W, Longo N, Pasquali M, Mengel E, Watkins D, Shoubridge EA, Majewski J, Rosenblatt DS, Fowler B, Rutsch F, and Baumgartner MR (2012) Mutations in ABCD4 cause a new inborn error of vitamin B12 metabolism. Nature Genetics (Article in press) * These authors contributed equally to this work

Presentations

1. Mutations in ABCD4 Cause a New Inborn Error of Vitamin B12 Metabolism (Poster). Human Genetics Graduate Research Day, McGill University. June 2011. Montreal, Canada.

132

2. Mutations in ABCD4 Cause a New Inborn Error of Vitamin B12 Metabolism (Oral). 12th International Congress of Human Genetics. October 2011. Montreal, Canada. (Abstract published)

3. Novel Inborn Error of Vitamin B12 Metabolism Caused By Mutations in ABCD4 (Poster). RMGA Journées Génétiques. May 2012. Montreal, Canada.

4. Novel Inborn Error of Vitamin B12 Metabolism Caused By Mutations in ABCD4 (Poster). Human Genetics Graduate Research Day, McGill University. June 2012. Montreal, Canada.

5. Novel Inborn Error of Vitamin B12 Metabolism Caused By Mutations in ABCD4 (Oral). FASEB: Folic Acid, Vitamin B12, and One-Carbon Metabolism. July 2012. Crete, Greece.

133

APPENDIX B: Published Abstract

12th International Congress of Human Genetics

Mutations in ABCD4 cause a new inborn error of vitamin B12 metabolism

J.C. Kim1, D. Coelho2,3, I.R. Miousse1, S. Fung1, M. du Moulin4, I. Buers4, T. Suormala2,3, M. Stucki2, P. Nürnberg5, H. Thiele5, N. Longo6, M. Pasquali6, E. Mengel7, D. Watkins1, E.A. Shoubridge1, F. Rutsch4, J. Majewski1,8, M. Baumgartner2, B. Fowler2,3, D.S. Rosenblatt1

1Department of Human Genetics, McGill University, Montreal, Canada, 2Division of Metabolism, University Children's Hospital, Zurich, Switzerland, 3Metabolic Unit, University Children's Hospital, Basel, Switzerland, 4University Children's Hospital, Münster, Germany, 5Cologne Center for Genomics, Cologne, Germany, 6University of Utah and ARUP Laboratories, Salt Lake City, United States, 7University Children's Hospital, Mainz, Germany, 8McGill University and Genome Québec Innovation Centre, Montreal, Canada

Two patients presented with methylmalonic aciduria and hyperhomocysteinemia. Patient 1 presented at birth following an abnormal newborn screen with hypotonia, lethargy, poor feeding and bone marrow suppression. Patient 2 presented in the newborn period with poor feeding, macrocytic anemia and heart defects. Studies of cultured fibroblasts from both patients showed decreased function of the cobalamin (vitamin B12) dependent enzymes methionine synthase and methylmalonyl-CoA mutase. There was increased uptake of labelled cyanocobalamin (CNCbl) but decreased synthesis of the cobalamin coenzymes methylcobalamin (MeCbl) and adenosylcobalamin (AdoCbl), with accumulation of “free” (i.e. non- protein bound) CNCbl in the cells. The cellular phenotype mimicked that of the cblF disorder caused by mutations in the LMBRD1 gene encoding the lysosomal membrane protein LMBD1 that is thought to play a role in transfer of cobalamin across the lysosomal membrane into the cytoplasm. However, cells from both patients complemented those from all known complementation classes, including cblF, and no mutations in LMBRD1 were found. Whole exome capture (patient 1) and microcell- mediated chromosome transfer and exome capture of chromosome 14 (patient 2), led to the identification of two mutations in the ABCD4 gene in each patient: c.956A>G (p.Tyr319Cys) and c.1746_1747insCT (p.Glu583LeufsX9) in patient 1 and c.542+1G>T and c.1456G>T (p.Gly486Cys) in patient 2. All mutations were predicted to be deleterious. Transfection of patient fibroblasts with wild type ABCD4 led to rescue of the abnormal cellular phenotype. Transfection with c.1456G>T did not rescue function confirming the functional significance of this mutation. We conclude that this novel disorder, tentatively named “cblJ”, is an autosomal recessive disorder caused by mutations in ABCD4. We suggest that ABCD4, a presumed ABC transporter, is another essential component of intracellular cobalamin metabolism and might interact with LMBD1 to allow transport of vitamin B12 out of the lysosome.

134