Modeling Congenital Disorders of Glycosylation in Caenorhabditis Elegans: Genetic Influences and Structural Consequences of N- Linked Glycosylation

University of New Hampshire University of New Hampshire Scholars' Repository

Doctoral Dissertations Student Scholarship

Spring 2009

Modeling congenital disorders of glycosylation in Caenorhabditis elegans: Genetic influences and structural consequences of N- linked glycosylation

Weston Booth Struwe University of New Hampshire, Durham

Follow this and additional works at: https://scholars.unh.edu/dissertation

Recommended Citation Struwe, Weston Booth, "Modeling congenital disorders of glycosylation in Caenorhabditis elegans: Genetic influences and structural consequences of N-linked glycosylation" (2009). Doctoral Dissertations. 488. https://scholars.unh.edu/dissertation/488

This Dissertation is brought to you for free and open access by the Student Scholarship at University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected]. MODELING CONGENITAL DISORDERS OF GLYCOSYLATION IN C. ELEGANS: GENETIC INFLUENCES AND STRUCTURAL CONSEQUENCES OF A/-LINKED GLYCOSYLATION

WESTON BOOTH STRUWE

B.S., UNIVERSITY OF WISCONSIN, 2002

DISSERTATION

Submitted to the University of New Hampshire

in Partial Fulfillment of

the Requirements for the Degree of

Doctor of Philosophy in Biochemistry

May, 2009 UMI Number: 3363732

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

UMI®

ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 This dissertation has been examined and approved.

'LQ_Xst*-**x_ L Dissertation Director, Vernon N. Reinhold Research Professor of Biochemistry

W. Kelley Thomas, Professor of Biochemistry

c*— Deena J. Small, Assistant Professor of Biochemistry

David J. Asfertfne, Research Scientist

Steven B. Levery, Research Associate Professor of Cellular and Molecular Medicine, University of Copenhagen

12, 2001 Date ' DEDICATION

FOR MY DAD

m ACKNOWLEDGEMENTS

I would like to sincerely thank my advisor Vernon Reinhold for his support and guidance. I would also like to thank Andy Hanneman, David Ashline and everyone in the UNH Glycomics Center for their time, tutelage and friendship.

Special thanks go out to my colleagues Will Wiswall and Justin Prien for their insightful words, discussions and collaboration in social endeavors.

I would like to thank my committee members Kelley Thomas, Steve

Levery and Deena Small for all their encouragement and help throughout my time at UNH. I also want to thank James Dennis and Wendy Johnston from the

Mount Sinai Hospital and University of Toronto for their support.

My sincerest gratitude and adoration goes to my mentor and friend

Charles Warren who, through his devotion, passion and kindness, changed my life forever and showed me the beauty and astonishment of scientific research. I am grateful for the short time I spent with Charles and will never forget him.

IV TABLE OF CONTENTS

DEDICATION iii

ACKNOWLEDGEMENTS iv

LIST OF TABLES vii

LIST OF FIGURES : viii.

ABSTRACT xii

CHAPTER PAGE

1. PERSPECTIVES OF GLYCOBIOLOGY 1

1.1 /V-linked Glycan Biosynthesis: Endoplasmic Reticulum 3

1.2 AMinked Glycan Biosynthesis: Golgi 8

1.3 Biosynthetic Complexity of /V-linked Glycans 13

1.4 A/-glycan Structure-Function Relationship 19

2. CONGENITAL DISORDERS OF GLYCOSYLATION.. 23

3. C. ELEG4NSAS A MODEL TO STUDY A/-GLYCOSYLATION 31

4. PROJECT AIMS 40

4.1 Genome-wide RNAi screen to Identify Tunicamycin Hypersensitive Loci in C. elegans 40

4.2 Comparative A/-glycan Structural Analysis of the C. elegans N2 (Bristol) and NF299 cogc-1(k179) strains 41

5. GENOME-WIDE RNAi SCREEN TO IDENTIFY TUNICAMYCIN HYPERSENSITIVE LOCI IN C. ELEGANS 43

5.1 Materials and Methods 50

5.2 Results 53

v 5.3 Discussion 63

6. COMPARATIVE A/-GLYCAN STRUCTURAL ANALYSIS OF THE

C. ELEGANSN2 (BRISTOL) AND NF299 cogc-1(k179) STRAINS 71

6.1 Materials and Methods 75

6.2 Results and Dissussion 79

6.2.1 Carbohydrate Structural Analysis Through Mass Spectrometry 81 6.2.2 Native Molecular Compositional Analysis via MALDI-TOFMS 83

6.2.3 Permethylated Molecular Compositional Analysis via MALDI-TOFMS 87

6.2.4 Structural Characterization of /V-glycans via MSn 92

6.2.5 Structural Characterization of GlcNAc2Man5 93

6.2.6 Structural Characterization of GlcNAc2Man5Fuc 104

6.2.7 Structural Characterization of GlcNAc2Man5Fuc2 118

6.2.8 Structural Characterization of GlcNAc2Man5Fuc3 130

6.2.9 Structural Characterization of GlcNAc2Man5Fuc4 139

7. CONCLUSIONS 144

REFERENCES 148

APPENDICES 159

APPENDIX A TUNICAMYCIN HYPERSENSITIVE LOCI 160

APPENDIX B NATIVE AND PERMETHYLATED MALDI-TOF SPECTRA 174

APPENDIX C RELATIVE ABUNDANCE OF NATIVE A/-GLYCANS 180

APPENDIX D ADDITIONAL MSnSPECTRA 183

vi LIST OF TABLES

Table 1. Common monosaccharides 2

Table 2. Congenital disorders of glycosylation by type and subtype 24-25

Table 3. Tunicamycin incudes a pleotropic condition in C. elegans 54

Table 4. RNAi of LLO genes with 2u.g/ml tunicamycin..... 57

Table 5. RNAi of "maturation" genes with 2|ig/ml tunicamycin 59

Table 6. Tunicamycin treatment does not alter RNAi effectiveness

in C. elegans 60

Table 7. Relative quantities of native O-methylated A/-linked glycans 86

Table 8. Relative quantities of permethylated AMinked glycans 89

Table 9. Relative intensity of MS/MS B-ion fragments from

GlcNAc2Man5Fuc 117 Table 10. Relative intensity of MS/MS B-ion fragments from GlcNAc2Man5Fuc2 129

Table 11. Relative intensity of MS/MS B-ion fragments from

GlcNAc2Man5Fuc3 138

Table A1. RNAi screen results 162

Table C1. Relative abundance of native A/-glycans 182

vn LIST OF FIGURES

Figure 1. Synthesis of the lipid-linked oliosaccharide precursor

in the endoplasmic reticulum 4

Figure 2. The calnexin/calreticulin cycle 6

Figure 3. Common A/-linked glycan types 9

Figure 4. Processing of the precursor LLO 12

Figure 5. Anterograde vesicular transport and cisternal maturation:

two models of Golgi morphology 15

Figure 6. CDG location and type 28

Figure 7. C. elegans A/-glycans 33

Figure 8. MS profiles of C. elegans A/-glycans released with

PNGase Fand hydrazine 35

Figure 9. Biosynthetic glycosylation pathway in C. elegans 37

Figure 10. Lipid-linked oligosaccharide pathway in the

endoplasmic reticulum 45

Figure 11. A/-glycosylation pathway from synthesis to localization 48

Figure 12. Tunicamycin induces a dose-dependent lethality in the N2 C. elegans strain 55 Figure 13. Tunicamycin effects postembryonic development among

C. elegans strains 56

Figure 14. RNAi phenotype penetrance and expressivity 62

Figure 15. Eukaryotic Orthologous Gene (KOG) gene assignemtns 66

Figure 16. MALDI-TOF spectra of native A/-linked glycans (1 of 3) 84

Figure 17. MALDI-TOF spectra of permethylated A/-glycans 88

Figure 18. Prevalence of A/-glycans by subtype 90 viii Figure 19. Comparative analysis of fucosylated /V-glycans organized

by subtype 91

Figure 20. GlcNAc2Man5 and the chitobiose core 94

3 Figure 21. MS of m/z 1302 from GlcNAc2Man5 (m/z 1595) 96

Figure 22. GlcNAc2Man5 Structural Isomer #2 97

4 Figure 23. MS of m/z880 ion from Man5 Isomer #2 98

Figure 24. MS5 spectra and fragmentation assignments of

m/z 667 from GlcNAc2Man5 100

5 Figure 25. MS of m/z 866 from GlcNAc2Man5 (m/z 1595) 102

6 Figure 26. MS of m/z 648 from GlcNAc2Man5 (m/z 1595) 103

Figure 27. Structural Isomers of GlcNAc2Man5 in N2 and NF299 103

2 2+ Figure 28. MS of GlcNAc2Man5Fuc(m/z896 ) 105

4 2+ Figure 29. MS of m/z 880 from GlcNAc2Man5Fuc (m/z 896 ) 107

5 2+ Figure 30. MS of m/z662 from GlcNAc2Man5Fuc (m/z896 ) 109

3 2+ Figure 31. MS of m/z 694 from GlcNAc2Hex5Fuc (m/z 896 ) 110

5 2+ Figure 32. MS of m/z 866 from GlcNAc2Man5Fuc (m/z 896 ) 112

Figure 33. MS5 spectra and fragmentation assignments of

m/z 667 from Man5Fuc 113 Figure 34. MS3 spectra and fragmentation assignments of

m/z 1476 from GlcNAc2Man5Fuc... 114

Figure 35. Structural Isomers of GlcNAc2Man5Fuc 116

2 2+ Figure 36. MS of GlcNAc2Man5Fuc2(m/z983 ) 119

3 Figure 37. MS of m/z894from GlcNAc2Man5Fuc2 121

n Figure 38. MS of m/z 1098 from GlcNAc2Man5Fuc2 123

n Figure 39. MS of m/z 1302 from GlcNAc2Man5Fuc2 124

ix 3 Figure 40. MS of m/z1272 from GlcNAc2Man5Fuc2 126

3 Figure 41. MS of m/z 1476 from GlcNAc2Man5Fuc2 128

2 2+ Figure 42. MS of GlcNAc2Man5Fuc3 (m/z 1070 ) 131

3 Figure 43. MS of m/z 1272 from GlcNAc2Man5Fuc2.. 132

Figure 44. MS4 spectra and fragmentation assignments of

m/z 667 from GlcNAc2Man5Fuc3 133

4 Figure 45. MS of m/z 1054 from GlcNAc2Man5Fuc2 (10702+-*1272^1054) 134

3 Figure 46. MS of m/z 1476 from GlcNAc2Man5Fuc3 135

3 Figure 47. MS of m/z 1446 from GlcNAc2Man5Fuc3 (10702+-^1446)inN2 136

3 Figure 48. MS of m/z 1650 from GlcNAc2Man5Fuc3

(10702+^1650)inN2 137

2 2+ Figure 49. MS of GlcNAc2Man5Fuc4 (m/z 1157 ) 140

n Figure 50. MS of m/z 1446 from GlcNAc2Man5Fuc4 140

3 Figure 51. MS of m/z 868 from GlcNAc2Man5Fuc4 141

Figure 52. MS3 of m/z 490 and m/z 690 from 11572+ 142

Figure 53. MS3 of m/z 894 from 11572+ 143

Figure B1. MALDI-TOF spectra of native AMinked glycans released by hydrazinolysis (2 of 3) 175 Figure B2. MALDI-TOF spectra of native AMinked glycans released by hydrazinolysis (3 of 3) 176

Figure B3. MALDI-TOF spectra of permethylated AMinked glycans released by hydrazinolysis (1 of 3) 177

Figure B4. MALDI-TOF spectra of permethylated AMinked glycans released by hydrazinolysis (2 of 3) 178

Figure B5. MALDI-TOF spectra of permethylated AMinked glycans released by hydrazinolysis (3 of 3) 179

x 6 Figure D1. N2 MS spectra of m/z 563 from Man5 (m/z1595-»1302-»1084-^667^563) 184

4 Figure D2. MS spectra of m/z880 from GlcNAc2Man5Fuc (8962+^1302-*880) 185

3 Figure D3. MS spectra of m/z 490 from GlcNAc2Man5Fuc (8962+->490) 186

3 Figure D4. MS spectra of m/z 490 from GlcNAc2Man5Fuc2 (9832+^490) 187

3 Figure D5. MS spectra of m/z694 from GlcNAc2Man5Fuc2 (9832+-+694) ,. 188

3 Figure D6. MS spectra of m/z 664 from GlcNAc2Man5Fuc2 (9832+-+664) 189

3 Figure D7. MS spectra of m/z868 from GlcNAc2Man5Fuc2 (9832+-+868) 190

3 Figure D8. MS spectra of m/z 1072 from GlcNAc2Man5Fuc2 (9832+-+1072) 191

3 Figure D9. MS spectra of m/z868 from GlcNAc2Man5Fuc3 (10702+^868) 192

3 Figure D10. MS spectra of m/z664 from GlcNAc2Man5Fuc3 (10702+^664) 193

3 Figure D11. MS spectra of m/z 694 from GlcNAc2Man5Fuc3 (10702+-^694) 194

3 Figure D12. MS spectra of m/z490 from GlcNAc2Man5Fuc3 (10702+^490) 195

XI ABSTRACT

MODELING CONGENITAL DISORDERS OF GLYCOSYLATION IN C. ELEGANS: GENETIC INFLUENCES AND STRUCTURAL CONSEQUENCES OF A/-LINKED GLYCOSYLATION

Weston Booth Struwe

University of New Hampshire, May 2009

The attachment of oligosaccharides to the amide nitrogen of asparagine side chains in proteins is a fundamental process occurring in all metazoans. This process, known as A/-glycosylation, is complex and is achieved by the precise interactions of various cellular components. The initial stage of A/-glycosylation occurs in the endolasmic reticulum and is preserved among eukaryotes. Glycans are further developed in the Golgi and the structural complexity depends greatly on the animal species, tissue and developmental stage. Oligosaccharides are unique biomolecules because unlike DNA or proteins, no primary sequence exists nor is its' synthesis template driven. A major goal of glycobiologist is to understand how glycan structures arise and their impact in biological systems.

Defective enzymes or components in this pathway cause congenital disorders of glycosylation (CDG) in humans. This disease is very rare, but exceedingly life-threatening. The CDGs are inherited in an autosomal recessive manner and clinical manifestations range from severe to mild. Most commonly, the disorders begin in infancy; manifestations range from severe developmental

xii delay and hypotonia with multiple organ system involvement to hypoglycemia and protein-losing enteropathy with normal development. Infants and children with CDG require nutrition supplements for maximal caloric intake and/or nasogastric tube or gastrostomy tube feedings. Orthopedic issues in adults require physical therapy, wheel chairs, transfer devices, and surgical treatment of scoliosis as needed.

The goals of researchers today are to identify further types of CDG, identify the defects in each type, and find means for management of the disease.

Paramount to CDG treatment and care is an understanding of the mechanisms of

/V-glycosylation and factors that influence the pathology of the disease.

Ultimately CDG treatment will stimulate correct glycosylation at the cellular level and potentially compensate for defects and restore normal glycosylation levels.

Xlll CHAPTER 1

PERSPECTIVES OF GLYCOBIOLOGY

The concept of the central dogma of molecular biology is no longer adequate to explain or understand the spectacular complexity of biological systems. Today, the field of systems biology adds to the notion of the central dogma which states that three major classes of biopolymers exist to describe the normal flow of biological information. In addition to the paradigm of DNA-RNA- protein, systems biology researchers focus on the systematic study of complex interactions in biological systems that include DNA, RNA, proteins, lipids, carbohydrates, non-coding RNAs and small molecule metabolites.

The discipline of glycomics is one within the large field of systems biology that aims to understand glycan structures and functions. Functional glycomics aims to tackle three general questions: 1) how glycans function in cellular communication; 2) what is the basis for specificity between proteins and glycans;

3) how glycan diversity and microheterogeneity results as a function of biology in development and disease.

In contrast to DNA, RNA and proteins, carbohydrates form branching structures and as a result, a relatively small set of monosaccharides (Table 1) can form considerably complex structures. Furthermore, glycans are not template generated biomolecules such as DNA, RNA and proteins. Sythesis relies on the transcription and translation of glycosylation enzymes that are

1 localized throughout the cell. The dynamics and complexity of glycan synthesis, coupled with the well established extrinsic factors makes glycobiology one of the most challenging disciplines in research today.

Carbohydrates, when attached to proteins or lipids, form large complex biomolecules termed glycoconjugates. The glycan portion of glycoconjugates fall into three catagories: those Suqar Type Abbreviation attached to lipids, and those Galactose Hexose Gal attached to proteins either Glucose Hexose Glc through a nitrogen atom {N- Mannose Hexose Man linked) or through an oxygen N-Acetylneuraminic acid Sialic acid NeuAc atom (O-linked). The Fucose Deoxyhexose Fuc attachment of glycans N-acetylgalactosamine Aminohexose Gal N Ac influences protein structure N-Acetylglucosamine Aminohexose GlcNAc and function, as well as the Xylose Pentose Xyl localization of cell surface Table 1. Common monosaccharides by type and secreted glycoproteins.

Glycans can confer cell-type specificity and are critical components of cell-to-cell signaling (Varki 1993). Carbohydrates are also involved in the immune response and host-pathogen interactions (Griffitts, Haslam et al. 2005). Moreover, changes in A/-glycan biosynthesis have been identified as key components in tumor progression in mice and humans (Fuster and Esko 2005).

Carbohydrate protein posttranslational modifications are a functionally important process that is vital in multiple biological systems. A/-linked glycans

2 are highly diverse branched structures that are attached to asparagines residues of polypeptide chains within the Asn-X-Ser/Thr sequon (Marshall 1974; Hubbard and Ivatt 1981). /V-glycosylation is universal of eukaryotes and between 25% and 50% of proteins encoded by eukaryotic genomes are estimated to be N- glycosylated (Apweiler, Hermjakob et al. 1999; Dennis, Granovsky et al. 1999;

Hirabayashi, Hayama et al. 2002). The mammalian glycome is estimated to range between hundreds and thousands of glycan structures suggesting it could be larger than the proteome (Ohtsubo and Marth 2006). Despite great advances in glycobiology, there remains much to be known about the roles of glycoconjugates, their metabolism and influence in diseases.

1.1 AMinked Glycan Biosynthesis: Endoplasmic Reticulum

The structural diversity of A/-glycans is determined through a complex biosynthetic pathway that begins on the surface of the endoplasmic reticulum

(ER). The early stages of A/-glycan biosynthesis is conserved among S. cerevisiae, C. elegans and vertebrates (Huffaker and Robbins 1982; Altmann,

Fabini et al. 2001). The biochemical pathway for A/-linked glycan synthesis occurs in four distinct stages: (1) formation of a lipid-linked precursor oligosaccharide; (2) en bloc transfer of the oligosaccharide to a nascent polypeptide; (3) trimming of oligosaccharides in the endoplasmic reticulum and

Golgi; (4) addition of new sugars in the medial and frans-Golgi. All eukaryotes share the first three steps of A/-glycosylation, but greater glycan complexity seen

3 in higher species is due to extensive processing that occurs in the Golgi. The formation of the dolichol-linked oligosaccharide precursor, GIC3Man9GlcNAc2 which is constructed on both sides of the ER, is the building block on which N- glycans become more elaborate (Figure 1).

0 Glucose (Glc)

® Mannose (Man)

| /V-acetylglucosamine (GlcNAc)

P-Dol • • • UDP—| J UDP— • T * T * T * ? P-P-Dol ± GDP-® ± Dol-P-® X Dol-P-0 ± P-P-Dol „_ P-P-Dol P-P-Dol • Fhppase • Cytosolic side of endoplasmic reticulum Lumen of endoplasmic reticulum

Figure 1. Synthesis of the lipid-linked oligosaccharide precursor in the endoplasmic reticulum.

The assembly of the 14 residue LLO molecule relies on the step-by-step action of substrate specific enzymes that build the precursor structure. The transfer of the 14 residue precursor oligosaccharide begins during the assembly of the polypeptide in the lumen of the ER and is facilitated by the oligosaccharyltransferase complex. The yeast and vertebrate oligosaccharyltransferase complex are significantly complex and consist of seven and eight subunits respectively (Kelleher and Gilmore 2006). Once the LLO is

4 transferred to the polypeptide in the lumen of the ER, the glycopeptide undergoes immediate quality control processing before moving to the Golgi.

As the nascent glycoprotein begins transport from the ER, glucosidase I removes the outer most glucose residue. Glucosidase II then removes the second proximal glucose leaving a glycoprotein with a single terminal gjucose.

Subsequently, two quality control lectins, calnexin and calreticulin ensure proper folding by binding to the monoglucosylated glycopeptide (Figure 2). Calnexin and calreticulin interact with ERp57, a thiol-disulfide oxidoreductase, which assists in the folding of luminal ER glycoproteins. Glucosidase II acts again by cleaving the third glucose and the calnexin/calreticulin complex dissociates from the glycoprotein.

At this point the glycoprotein can follow two distinct routes depending on proper protein folding. If the calnexin/calreticulin complex determines the glycoprotein to be properly folded, it will continue towards the Golgi in the intermediate compartment (IC) for further processing alone or with the aid of an additional ER lectin, ERGIC-53. If the glycoprotein is partially unfolded, it is retained in the ER and reglucosylated by a luminal glucosyltransferase and will once again interact with calnexin/calreticulin and ERp57 to achieve proper folding. If the protein is folded correctly, the ER a1,2-mannosidase I cleaves a specific terminal mannose residue from the Man9GlcNAc2 oligosaccharide to generate Man8GlcNAc2 before the glycoprotein exits the ER. However, if correct folding is not achieved, the glycoprotein will trigger the endoplasmic reticulum-

5 associated protein degradation (ERAD) pathway and will be translocated to the cytoplasm and digested in the proteasome (Jones, Krag et al. 2005).

If unfolded proteins accumulate in the endoplasmic reticulum, the cell can initiate the unfolded protein response (UPR). Transit from the ER to the Golgi complex is the rate-limiting step in secretion for many glycoproteins and if the

CRT

Glucnsidase II

Folded Glycoprotein / (ilucosylTransfcrasc < u exit cycle Glncosidase II

Glucosidase II Glucosidase 1 / ER mannosidas* I Jw _ exit cycle mm •ass j&a*»i ^ SlSSSllaSl Will -S*

Transition complex ERAD Figure 2. The calnexin/calreticulin cycle. Adapted from Jones et. al. 2005

influx of nascent, unfolded polypeptides exceeds the folding and/or processing capacity of the ER, the normal physiological state of the ER is perturbed. A high concentration of partially folded and unfolded proteins predisposes protein- folding intermediates to aggregate. The UPR provides a mechanism by which

6 cells can rapidly adapt to alterations in protein-folding loads in the ER lumen by expanding the capacity for protein folding. A variety of factors that disrupt protein folding in the ER lumen also activate the UPR. These include changes in intralumenal calcium, altered glycosylation, nutrient deprivation, pathogen infection, expression of folding-defective proteins, and changes in redox status

(Malhotra and Kaufman 2007). Additionally, perturbations of ER function can be caused by overexpressing large and heavily modified proteins or small structurally simple proteins (Schroder and Kaufman 2005).

In response to ER stress, the ER-localized transmembrane signal transducers are activated to initiate adaptive responses. These transducers are activating transcription factor 6 (ATF6), inositol requiring kinase 1 (IRE1), and double-stranded RNA-activated protein kinase (PKR)-like endoplasmic reticulum kinase (PERK). The most immediate response to ER stress is the reversible attenuation of mRNA translation, thereby slowing the influx of newly synthesized polypeptides. This response is signaled through PERK by phosphorylation of the eukaryotic translation initiation factor 2 on the alpha subunit (elF2a) which inhibits the formation of the active elF2 complex. elF2 is required for the ternary complex elF2-GTP-tRNA that recognizes AUG codons and is responsible for 60S ribosomal subunit assembly. The second response to ER stress is up-regulation of UPR genes that increase the folding capacity within the ER. IRE1 is a protein kinase that when activated undergoes frans-autophosphorylation to activate its'

RNase activity. Active IRE1 removes a 26 nucleotide intron from XBP1 mRNA.

Spliced XBP1 is a transcriptional activator of a wide variety of UPR target genes

7 such as ERAD and chaperones. The third sensor in the UPR, ATF6, acts by activating UPR-responsive genes after being localized to the Golgi. The two active forms, ATF6a and ATF6(3, translocate to the nucleus from the Golgi where it binds to the ATF/cAMP response element (CRE) and the ER stress responsive element (ERSE-1) to activate UPR chaperones. If the UPR fails to resolve the protein-folding defect and ER stress, apoptosis is activated through mitochondrial-dependent and independent pathways. However the relationship between ER stress and apoptosis remain to be defined as well as the frequency of this event. Seemingly, the unfolded protein response is cell-type specific and subject to modulation according to the needs of individual cells.

1.2 /V-linked Glvcan Biosynthesis: Golgi

During healthy secretion of nascent proteins, newly synthesized glycoproteins exit the ER by forming the intermediate compartment (IC) or vesicular-tubular clusters (VTCs) on its way to the Golgi apparatus, the hub of the secretory pathway. ER export of immature glycoproteins is mediated by COP

II coated vesicles and all AMinked glycoproteins at this stage possess the

Man8GlcNAc2 precursor oligosaccharide. The components involved in COP II vesicle formation are essential for cell viability in yeast and are conserved among eukaryotes (Duden 2003). The assembly of the COP II coat on the ER membrane is initiated by guanine nucleotide exchange of the GTPase Sarlp

(Futai, Hamamoto et al. 2004). The presence of Sar1p-GTP leads to the

8 recruitment of additional coat components, namely two hetero-dimeric protein complexes; Sec23/24p and Sec13/31p (Matsuoka, Orci et al. 1998). Once the

GTP-rich Sari p is embedded in the bilayer of the ER membrane and Sec23/24p and Sec13/31p are recruited, budding and vesicle formation occurs.

COP II vesicular budding has been observed in both mammalian and yeast cells by time lapse fluorescence microscopy,

® NeuAc using chimeras of a COP II • Gal • GlcNAc subunit with green fluorescent • Man fV Polypeptide protein (Duden 2003). After

High-Mannose Hybrid-type Complex-type COP II buds are released Figure 3. Common AM inked glycan types. from the ER the vesicles uncoat, allowing them to fuse with other COP II vesicles which forms the intermediate compartment (IC). The IC moves towards the Golgi along microtubule tracks and docks to the c/s-Golgi where it unloads its cargo.

Subsequently, COP II coat proteins are recruited back to the ER to repeat the cycle. In addition to COP II coat proteins, COP I is a second class of evolutionary conserved coat proteins that is also crucial for directing the transfer of material within the Golgi.

In COP II vesicles deliver Man8GlcNAc2 containing glycoproteins in the c/s-Golgi where three additional a-mannosidases (IA, IB and IC) reduce the oligosaccharide to Man5GlcNAc2. This structure serves as the substrate for complex, hybrid and high mannose /V-glycans (Figure 3) formed in the medial

9 and trans-Go\g\ by means of assorted glycosyltransferases. Next the glycoprotein may move forward to a high-mannose type or by the action of N- acetylglucosaminyltransferase-l (GnT-l), it may proceed to the complex and hybrid assembly pathway (Figure 4). GnT-l adds a A/-acetylglucosamine residue in a (31-2 linkage to the terminal end of the a1,3-linked mannose residue. After the action of GnT-l the glycan may enter hybrid glycan synthesis where the remaining mannose residues remain and extensions build upon the

GlcNAci Man5GlcNAc2 oligosaccharide.

For complex structure synthesis, a-mannosidase II removes the two terminal mannose residues on the crt ,6 arm and distinct GlcNAc-transferases extend the structure of the glycan. These GnT enzymes, II through V, are found in the medial Golgi and define different branching patterns common of complex

/V-glycans. GnT II adds a GlcNAc to the a1,3 arm of the Man3GlcNAc2 core.

GnT III adds a bisecting (31 -4 linked GlcNAc to the core mannose residue.

Likewise, GnT IV and GnT V add GlcNAc residues to the 1,3 arm and 1,6 arm to generate tri-antennary and tetra-antennary forms, respectively (Taylor and

Drickamer 2003). Complex branched structures are typically extended by the addition of a single galactose and sialic residue to each GlcNAc residue. The enzymes responsible for these additions are found further along in the secretory pathway in the trans Golgi and the frans-Golgi network (TNG). Additionally, fucosyltransferases and other GnT enzymes add to the complexity of hybrid and complex glycan structures at different locations on the glycan core and terminal residues. After generating glycan structures in the medial and trans Golgi, the

10 maturing glycopeptide enters the frans-Golgi network and is localized to the cell surface or within the cell.

Despite an understanding of the principal biosynthesis of /V-linked glycosylation, the extent of diversity and number of glycan structures present on proteins is still unclear. In spite of the structural information that has been gathered to date, the mechanisms in secretory pathway that contribute to the diversity of glycoconjugates are considerably less understood.

11 Endoplasmic Reticulum C/s-Golgi I

Mannosidase

GlcNAc-transferase I High Mannose (GnT-l) Biosynthesis

Additional Galactosyltransferase * p GlcNAc- ® f> Modifications Sialyltransferase ¥ transferase II Golgi » < i (GnT-ll) Mannosidase I Fucosyltransferase m . \ / Hybrid Glycan Medial Golgi Biosynthesis Trans Golgi

i" •••

Figure 4. Processing of the precursor LLO. The LLO is trimmed and modified to generate high-mannose, hybrid or complex type oligosaccharides in the Golgi. Three mannose residues are trimmed from the GlcNAc5Man8 containing polypeptide in the cis- Golgi. This structure is the scaffold for high-mannose biosynthesis. Alternatively, GlcNAc transferase I (GnT-l) can add a A/-acetylglucosamine residue on the 3-arm of the oligosaccharide. This structure is a precursor for complex and hybrid biosynthesis and after the trimming of two additional mannose residues by Golgi mannosidase II, the glycan can undergo complex type biosynthesis. During hybrid and complex assembly various glycosyltransferases, such as galactosyltransferases, sialyltransferases and fucosyltransferases, can modify the structure at different locations.

12 1.3 Biosynthetic Complexity of A/-Iinked Glycans

In addition to the inherent complexity of glycoconjugates, a particular oligosaccharide of a glycoprotein may occur in forms that differ in one or more of its oligosaccharide residues, a trend known as microheterogeneity. Many factors influence a protiens' glycomer distribution such as transport rates from the ER to the Golgi, the duration glycoconjugates are in the Golgi, sugar nucleotide metabolism, and localization of glycosyltransferases in the Golgi (Hossler,

Mulukutla et al. 2007). By and large the diversity of glycan structures is due to over 200 different glycosyltransferases located in the Golgi (Opat, van Vliet et al.

2001). On the other hand, the precise localization of only a small number of these glycosyltransferase enzymes within the Golgi has been determined through sub-cellular fractionation and immuno-labelling of thin sections followed by electron microscopy (Eisner, Hashimoto et al. 2003).

The gene expression of glycosyltransferases, glycosidases and high- energy sugar transporters alter the glycan assembly machinery in the Golgi and ultimately affect the structure of A/-linked glycans. Golgi resident glycosylation enzymes are present in a nonuniform steady-state distribution throughout each cisternae (Ungar, Oka et al. 2002). To generate this non-uniform distribution, the enzymes and associated substrates transporters must be first transported to a specific Golgi cisternae, after which their localization must be maintained through a combination of retention, retrieval or replacement (Ungar, Oka et al. 2006).

Glycosylation enzymes are transported to one of three discrete cisternae, or

13 stacks, in the Golgi and are retained in that location unless they inadvertently move. For example, A/-acetylglucosaminyltransferase I and Mannosidase II are localized in the c/s/medial Golgi, while galactosyltransferase and sialyltransferase are mostly found in the frans-Golgi (Nilsson, Pypaert et al. 1993; Rabouille, Hui et al. 1995). If an enzymes' location becomes disoriented then it is retrieved by vesicular transport mechanisms and returned to the appropriate site by means of intra-Golgi COP I vesicles (Harris and Waters 1996; Martinez-Menarguez,

Prekeris et al. 2001).

What determines A/-glycan microhicroheterogeneity is the transcription, translation and localization of these glycosylation enzymes as well as structural dynamics of the Golgi, which has yet to be fully defined. The structural diversity and complexity of giycans is a distinguishing characteristic of the Golgi and its non-template-based glycan biosynthesis. Even more intriguing is the manner in which the Golgi functions. Two theories exist concerning the nature of transport of glycoproteins across the Golgi stack. One called vesicular transport is carried out by budding and fusing among static cisternae by means of anterograde transport vesicles. The second, called cisternal maturation, suggests that the

Golgi cisternae themselves are constantly being formed at the cis face of the

Golgi by the fusion of ER-derived vesicles, particularly COP II vesicles (Figure 5)

(Pelham and Rothman 2000). Currently no data exists on which model is correct, although the cisternal maturation model is gaining favor. Additionally, some propose that neither model is mutually exclusive and that the true nature of

14 Golgi morphology is a combination of the two models (Pelham and Rothman

2000).

/^•^« rir- n n T\

Figure 5. Anterograde vesicular transport and cisternal maturation: two models of Golgi morphology, (a) In anterograde transport cargo from the ER (green) approaches and fuses with the c/s-Golgi. This cargo with subsequently bud and fuse (yellow) with each cisternae until it exits the frans-Golgi as mature glycoconjugates (brown). Retrograde vesicles (blue) return from downstream cisternae and return Golgi enzymes to their proper site, (b) In cisternal maturation, the Golgi cisternae are constantly being formed at the cis face of the Golgi from cargo vesicles from the ER (green). At the trans surface, cisternae are forming into vesicles (brown) intended for secretion or trafficking. Similarily to anterograde transport, cisternal maturation mechanisms require retrograde transport of Golgi proteins (blue) but at a much greater rate.

The fundamental difference between the two models is the differential movement of resident and cargo proteins. It is interesting to note that newly synthesized glycoproteins are segregated away from the resident proteins throughout the pathway in both models. What is known of the Golgi is that each cisternae functions specifically and is a fully functional unit, but not why or how these stacks aggregate and connect to form the complete functional Golgi

15 complex. During anterograde vesicular transport, cargo molecules

(glycoproteins) move forward through the Golgi while resident proteins

(glycosyltransferases) are specifically retained.

Alternatively, the cisternal maturation model proposes that the cargo molecules move passively as the cisternae mature towards the frans-Golgi and resident proteins are recycled by retrograde transport to establish the non uniform concentration across the Golgi. Inherent to both models is the dependency of localization, or retrograde transport, of resident proteins by COP I vesicles. COP I coat proteins mediate Golgi to ER and intra-Golgi traffic and help maintain the normal structure of the Golgi complex through retrograde localization of machinery proteins (Harris and Waters 1996; Love, Lin et al. 1998;

Lanoix, Ouwendijk et al. 2001). In vesicular transport retrograde transport is only required if resident proteins unintentionally move from their precise cisternae.

However, in the maturation model where resident enzymes travel with the cargo towards the trans-Golgi network, retrograde transport is required at a greater rate.

Before live cell imaging and the advent of green fluorescent protein (GFP) tags, both the ER and Golgi were thought as relatively stable structures, an assumption that led to the vesicular transport model. This was shown by

Jamieson and Palade, that in the absence of protein synthesis, the secretory pathway remained intact (Jamieson and Palade 1968). The number of Golgi cisternae remained constant and the Golgi as a whole appeared indistinguishable to conditions when protein synthesis occurred. In spite of this,

16 the fact that vesicular transport could not explain how macromolecular complexes, which are too large to fit in an intra-Golgi vesicle, could progress between the Golgi cisternae was ignored (Brown 1969; Marchi and Leblond

1984). Additionally, during the construction of complex glycans, it was assumed that glycosyltransferases were not allowed to mix with one another. This belief further reinforced the idea that strict compartmentalization was required which fitted with the vesicular transport model (Eisner, Hashimoto et al. 2003). In reality resident glycosyltransferases exist in mixtures in the Golgi stacks and compete for substrates (Nilsson, Pypaert et al. 1993).

The case for vesicular transport became increasingly vulnerable when large biomolecules (i.e. collagen) are considered to traverse the secretory pathway. Given their sheer size, it is difficult that such molecules could move forward in the pathway based on the 50 nm diameter of COP I and COP II vesicles. Studies have shown that collagen matures into fibrinous material in the lumen of the Golgi (Marchi and Leblond 1984) which suggest an additional scenario where larger cargo could be excreted through the Golgi as a collection of cisternae that mature with the constant input of new material. What was not know at this time was that the Golgi is subject to extensive recycling of resident components and have long half-lives (Eisner, Hashimoto et al. 2003). A typical glycosyltransferase has a half-life of -20 hours, in dividing cell tissue culture, which typically divide every 20 hours. Additionally, the rate of glycosylation enzyme(s) synthesis is low, which causes the cell to recycle them efficiently and effectively. In addition to these studies, it was shown through brefeldin A

17 disruption that the Golgi could completely be absorbed into the ER. When BFA was removed, the Golgi would reassemble within a short period of time (Misumi,

Misumi et al. 1986; Lippincott-Schwartz, Yuan et al. 1989).

The plasticity of the Golgi, the limiting size of its transport vesicles and the observation that glycosylation enzymes are not strictly compartmentalized has mired the legitimacy of the vesicular transport model. Despite this evidence, the maturation model is not without faults. A maturation model situation would infer that cargo entering the Golgi would move through and exit the pathway at the same rate. This is not the case, as seen with mucin and other highly glycosylated proteins, that are know to spend longer periods in the Golgi. Some answer that COP vesicles mediate retrograde transport to counter the cisternal maturation when needed (Eisner, Hashimoto et al. 2003). The extent of evidence that supports both models of intracellular transport highlights the lack of knowledge about the secretory pathway and even more so the dynamics of glycosylation in which this organelle is paramount.

The connection between Golgi morphology and glycosylation is significant and understanding cellular activity of these principles is perhaps the next step in glycan metabolism. The ability of the Golgi to mediate alterations in glycan structure is based on many factors and although the assembly pathway is somewhat known, glycan function remains obscure.

18 1.4 JV-glycan Structure-Function Relationship

Deciphering the association between a particular glycan structure and its function is an essential question that challanges glycobiologists. Presently, more than 7000 glycan structures have been determined, but their significance in cellular function remains to be established (Srivastava 2008). Some suggest that glycomics is at least an order of magnitude more difficult than proteomics (Dove

2001). Considering the complex diversity of glycans and how they influence proteins, it is no surprise that Schachter questions, "will it ever be possible to determine the role that a specific post-translational modification plays in the function of a specific protein for every protein in the genome" (Schachter, Chen et al. 2002). Nonetheless, notable advances have been made in elucidating the role of /V-linked glycans. It is known that glycans influence cell growth and development, tumor growth and metastasis, anticoagulation, immune recognition and response, cell-cell communication and microbial pathogenesis (Raman,

Raguram et al. 2005).

The goal of functional glycomics is to assign a specific glycan(s) to a particular protein and determine its function. There is increasing interest in functional glycomics and several international collaborative effects have been established, striving to explain the biological roles of glycans. This collaboration, which includes the Consortium for Functional Glycomics, EuroCarb and the

Japanese Consortium for Glycomics, has collected a list of 200 human genes corresponding to glycosyltransferases. Much of this data has been used to

19 generate cell culture and whole-organism knockouts to understand how genotype influences phenotype. In addition to the functional genetics methodology of glycomics, considerable advances have been made in the development of bioinformatics platforms, lectin databases, glyco-gene microarrays, structure databases and tools to investigate the conformational aspects of glycans and their corresponding protein component. The complexity in the field of glycomics requires a systems approach that investigates biosynthesis, structural analysis as well as glycan-protein interactions to delineate glycan-structure relationships.

The many experimental approaches taken to understand the roles of glycans include inhibition of glycosylation, alterations to processing mechanisms, elimination of glycosylation sites, enzymatic or chemical de-glycosylation of complete glycan chains and the study of glycosylation mutants (Varki 1999). The consequences of altering glycosylation are highly unpredictable and the effects can range from virtually undetectable to lethal. Moreover the same studies or changes in specific aspects of glycosylation can change depending on the animal model used. In many cases, investigating the role or structure of glycans through knockout models is beneficial, but the universal presence of glycoconjugates makes understanding cellular phenotypes as a function of its gene and protein components difficult.

On the most basic level, glycans alter proteins on an intrinsic (structural) level or extrinsically where the carbohydrate modulates the function of the underlying peptide. The external location of glycans on proteins can serve as a shield, protecting the protein from proteases or antibodies. Carbohydrates are

20 exceedingly hydrophilic which alters the conformation and solubility of proteins.

Protein folding is driven by its folding energy landscape initiated by hydrophobic collapse. The free energy of each possible conformation is largely determined by the primary sequence and by the contacts of its nonpolar groups. The addition of carbohydrates during translation in the ER greatly alters the energy landscape of proteins. Depending on the size, glycoform and extent of occupancy, a protein will fold until a native structure is formed and the lowest free energy is reached.

Individual protein motifs (a-helices and (3-turns) fold within microseconds, which is why quality control measures are in place to determine proper folding before any protein exits the ER (Schroder and Kaufman 2005).

The presence or absence of glycans can affect the primary function or activity of a protein. For example, (3-human chorionic gonadotrophin (P-HCG) can bind with similar affinity to its receptor with and without its glycan component.

However, in its deglycosylated form (3-HCG fails to stimulate adenylate cyclase

(Varki 1999). Glycans can also influence the longevity of proteins to which they are attached. In the case of human erythropoietin, the presence of sialic acid on its terminal branched increase the half-life, but decreases the activity in vitro.

The extent of branching can determine binding of erythropoietin to its receptor in specific tissues (Takeuchi and Kobata 1991). The tuning effect of glycan sequences act in protein function, although the effect may be a change in the binding mechanism seen through changes in glycan/protein structure.

Similarily, glycans act as specific ligands for endogenous and exogenous receptors. The role of glycans and ligands for lectins is perhaps the most

21 common functional aspect of carbohydrates in cellular systems. For example, the glycoprotein hemagglutinin (HA) on the surface of the avian influenza virus is responsible for binding to the cell being infected. Avian HA binds specifically to a2-3 sialylated glycans, which are absent in the respiratory tract of humans. It is thought that a switch in HA binding from a2-3 sialylated glycans to a2-6 sialylated glycans, which are present in humans, enables infections in humans

(Chandrasekaran, Srinivasan et al. 2008). The future directions will aim to define not only the nature of binding interactions of carbohydrates to lectins, but the elucidation of the nature of the biological functions of these molecules.

Currently functional glycomics lacks any high-throughput strategies for analyzing the structure and function of each carbohydrate moiety. Structural characterization of the vast heterogeneity of glycans generated from their non- template and poorly understood synthesis is only one important aspect of functional glycomics. The challenges of glycomics include understanding glycan structure as a function of extracellular signaling, determining the basis for glycan- protein specificity and interactions, and elucidating how glycan diversity is generated as a function of its biosynthesis. Furthermore the biology and biosynthesis of glycosylation on a cellular level remains unclear. Addressing the fundamental biology of glycosylation is vital in order to link all facets of functional glycomics.

22 CHAPTER 2

CONGENITAL DISORDERS OF GLYCOSYLATION

Protein A/-glycosylation is an essential and pervasive event in eukaryotic development. Mild defects in this process lead to deficiencies in cellular growth and function and are associated with disease. Most notable are a group of inherited autosomal recessive human disorders called congenital disorders of glycosylation (CDG). CDGs are categorized as type I or type II depending on the location in the biochemical pathway where the defect occurs. Type I disorders are characterized as defects of enzymes in the lipid-linked oligosaccharide assembly pathway. Currently there are 12 known defects, or subtypes, in this category (Table 2). Type II disorders are deficiencies of enzymes involved in the trimming and processing of the protein bound oligosaccharide in the Golgi. To date there are eight defective enzymes in type II CDG (Jaeken and Matthijs

2007; Zeevaert, Foulquier et al. 2008).

23 Disorder Gene Enzyme Features Developmental delay, hypotonia, esotropia, lipodystrophy, CDG-la PMM2 Phosphomannomutase II cerebellar hypoplasia, stroke like episodes, seizures Hepatic fibrosis, protein losing CDG-lb PMI Phosphomannose Isomerase enteropathy, coagulopathy, hypoglycemia Glucosyltransferase I Moderate developmental delay, CDG-lc ALG6 Dol-P-GIc: Man9GlcNAc2-PP- hypotonia, esotropia, epilepsy Dol Glucosyltransferase Profound psychomotor delay, Dol-P-Man: Man5GlcNAc2- optic atrophy, acquired CDG-ld ALG3 PP-Dol Mannosyltransferase microcephaly, iris colobomas; hypsarrhythmia Profound psychomotor delay, severe developmental delay, Dol-P-Man Synthase I optic atrophy, acquired CDG-le DPM 1 GDP-Man: Dol-P- microcephaly, epilepsy, Man nosyltransf erase hypotonia, mild dysmorphism, coagulopathy Short stature, icthyosis, CDG-lf MPDU 1 MPDU1/Lec35 psychomotor retardation, pigmentary retinopathy Hypotonia, facial dysmorphism, Dol-P-Man: Man7GlcNAc2PP- psychomotor retardation, CDG-lg ALG12 Dol Mannosyltransferase acquired microcephaly. Frequent infections Glucosyltransferase II Hepatomegaly, protein-losing Dol-P-GIc: enteropathy, renal failure, CDG-lh ALG8 Glc1 Man9GlcNAc2-PP-Dol hypoalbuminemia, edema, Glucosyltransferase ascites Normal at birth; developmental Mannosyltransferase II delay, hypomyelination, CDG-li ALG2 GDP-Man: Man1GlcNAc2-PP- intractable seizures, iris Dol Mannosyltransferase colobomas, hepatomegaly, coagulopathy UDP-GlcNAc: dolichol Severe developmental delay, phosphate CDG-lj DPAGT1 hypotonia, seizures, N-acetylglucosamine 1- microcephaly, exotropia phosphate transferase Severe psychomotor retardation, Mannosyltransferase I hypotonia, acquired CDG-lk ALG1 GDP-Man: GlcNAc2-PP-Dol microcephaly, intractable Mannosyltransferase seizures, fever, coagulopathy, nephrotic syndrome, early death Mannosyltransferase Dol-P-Man: Man6 and Severe microcephaly, hypotonia, CDG-IL ALG9 8GlcNAc2-PP-Dol seizures, hepatomegaly Mannosyltransferase GlcNAc-Transferase 2 (GnT developmental delay, CDG-lla MGAT2 N) dysmorphism, seizures Dysmorphism, hypotonia, CDG-llb GLS1 Glucosdase I seizures, hepatomegaly, hepatic fiborsis

24 Recurrent infections, persistent SLC35C1 neutrophilia, developmental CDG-llc GDP-Fucose Transporter /FUCT1 delay, microcephaly, hypotonia (normal Transferrin) Hypotonia (myopathy), CDG-lld B4GALT1 b1,4 galactosyltransferase spontaneous hemorrhage, Dandy-Walker malformation Fatal in early infancy; dysmorphism, hypotonia, Conserved oligomeric Golgi intractable seizures, CDG-lle COG7 complex subunit 7 hepatomegaly, progressive jaundice, recurrent infections, cardiac failure Thrombocytopenia, no CDG-llf SLC35A1 CMP-Sialic acid transporter neurologic symptoms, normal Tf, abnormal platelet glycoproteins Failure to thrive, hypotonia, growth retardation, cardiac CDG- COG1 Conserved oligomeric Golgi failure, microcephaly, mild I l/COG 1 complex subunit 1 hepatosplenomegaly, mild psychomotor retardation, cerebellar and cerebral atrophy Facial dysmorphism, encephalopathy, hypotonia, CDG-llh status epilepticus (recurrant (CDG- COG8\ Conserved oligomeric Golgi seizures), ataxia, progressive II/COG8) complex subunit 8 microcephaly, esotropia and eyelid pseudoptis, cerebellar atrophy

Table 2. Congenital disorders of glycosylation by type and subtype.

The importance of protein glycosylation is dramatically illustrated by the severity of the disease, although the history of the CDG is comparatively concise.

Congenital disorders of glycosylation (formally known as carbohydrate-deficient glycoprotein syndromes) was initially discovered in 1980 by the observation of serum protein abnormalities in monozygous twin sisters with psychomotor retardation, decreased thyroxine-binding globulin levels and increased arylsulphatase A activities (ASA) (Jaeken, Vanderschuerenlodeweyckx et'al.

1980).

25 However it was not until 1983 when the nature of the problem was identified by a shift in the isoelectric focusing (IEF) profile of serum transferrin.

This newly developed technique for IEF of transferrin pinpointed to an absence of the terminal, negatively charged sialic acid residue (Vaneijk, Vannoort et al.

1983; Jaeken, van Eijk et al. 1984). The breakthrough provided both insight into the biological nature of glycosylation deficient diseases and a powerful diagnostic test that remains the benchmark for screening today. Isoelectric focusing and the measurement of serum ASA activity is still part of the metabolic screening program for the disease. IEF of transferrin can not only indicate the occurrence of CDG, but based on the IEF shift pattern, it can identify a type I or type II disorder. Additionally, western blot analysis and mass spectrometry of serum transferrin are additional screening techniques employed, but are not as effective. Ultimately, measurement of specific enzyme activities and genetic screening confirm diagnosis (type and subtype), although clinical identification of the disease is exceedingly difficult due to unusual and random clinical presentation. Even with a CDG diagnosis, a high number of biochemical unresolved cases, where the particular enzyme involved is unknown, (CDG-lx) remain problematic (Morava, Wosik et al. 2008).

The number of identified CDG subtypes (or syndromes) is increasing exponentially, with most of the known subtypes having been discovered in the past decade. Most patients diagnosed with CDG suffer from a defective gene that encodes phosphomannomutase (PMM), the enzyme responsible for conversion of Man-6-phosphate to Man-1-phosphate. Deficiencies of PMM is the

26 basis for the most common form of CDG (type la), where the most recurrent mutation, R141H, has a carrier frequency of 1/60 (Kjaergaard, Schwartz et al.

2001). The calculated frequency of homozygotes for this mutation is 1/3600, although there is total lack of homozygotes in the population. The expected homozygote R141H/R141H genotype is 45% to 60% of CDG-la patients. Those affected with CDG-la are compound heterozygotes for R141H. Given that

R141H homozygotes are probably lethal, CDG-la with this genotype is a compelling candidate for neonatal deaths (Schollen, Kjaergaard et al. 2004).

The second most common CDG subtype is CDG-lc with more than 20 patients. These patients are deficient for glycosyltransferase I activity which adds a glucose residue to the growing LLO carbohydrate. For reasons that are unknown, these patients experience milder symptoms compared to CDG-la.

Some speculate that defects that occur later in the biosynthetic pathway have a smaller impact on neuronal function (Eklund and Freeze 2006). The only treatable form of the disease is CDG-lb which is due to a deficiency in phosphomannose isomerase (PMI). This enzyme catalyzes the formation of

Man-6-phospahte for fructose-6-phosphate. This enzyme is the first step in the

A/-glycosylation pathway and defects are efficiently treatable with oral mannose

(4 x 150 mg/kg per day).

27 the lack of CDG patients diagnosed, either CDG illnesses are severely underdiagnosed or those afflicted die before birth (Matthijs, Schollen et al. 2000; de Lonlay, Seta et al. 2001). Although neither theory has been established, is it plausible that each possibility is mutually exclusive depending on the severity of the biosynthetic defect (i.e. type, subtype or specific genetic mutation within subtype).

transfer ER M > CDG type I

trim

branch CDG type II

trans Golgi network

terminate trans

Figure 6. CDG cellular location and type. CDG diseases are characterized as type I if the defect is located in the endoplasmic reticulum and type II if the defects are in the stacks of the Golgi.

Strong losses-of-function are embryonic lethal, patients with partial losses- of-function are occasionally born but are very ill, presenting with defects in virtually every body system. Individuals born with CDG are inundated with

28 defective glycosylation enzymes leading to absent and/or aberrant glycan structures (Mills, Mills et al. 2001). CDG patients suffer from psychomotor retardation, low muscle tone, incomplete brain development, visual problems, seizures, stroke-like episodes, coagulation disorders, endocrine abnormalities and overall failure to thrive (Freeze and Westphal 2001; Eklund and Freeze

2006). Based on clinical features, CDG illnesses are split into two clinical subtypes: a neurological form and a multivisceral form with neurological manifestations. In spite of this, the reason why CDG patients develop neuronal problems is not fully known.

Even with an identical genotype, CDG-I clinical expression varies considerably and ranges from very mild to extremely severe. The biosynthetic pathway for glycoprotein formation is relatively understood and has provided a basis for establishing the etiology of CDG (Jaeken and Matthijs 2001). However, there are still cases of CDG with unknown etiology, where strong clinical evidence (i.e. isoelectric focusing of serum transferrin) points to CDG but the defective gene has yet to be identified (Prietsch, Peters et al. 2002). As well as the problematic genetics factors of CDG, the mechanistic consequences of defective glycoprotein synthesis are very poorly understood which confounds the clinical identification of this disease. Incorrect or under-glycosylation is the underlying basis in CDG syndromes but what puzzles physicians and researchers is pathogenesis. Future research in CDG will emphasize structural elucidation of glycans at a tissue specific level and aim to alleviate the stresses of hypo-or aberrant glycosylate . Aside from CDG-Ib, the only treatable form of

29 the disease, supportive therapy is the only option for all other CDG conditions and early diagnosis is critical. CHAPTER 3

C. ELEGANS AS A MODEL TO STUDY A/-GLYCOSYLATION

The free-living nematode Caenorhabditis elegans is a powerful resource for studying complex developmental topics and provides a robust model system to examine various human diseases. In addition to the ease of maintenance and manipulability C. elegans provides, it is the most understood multicellular eukaryotic animal to date. The complete cell lineage has been mapped and cell fates remain invariant between individuals (Lambie 2002). C. elegans is a simple organism comprised of only 959 cells, but possess feeding, nervous, muscle and reproductive physiological systems found in higher eukaryotes. One of the most useful features of C. elegans is its susceptibility to genetic silencing. Most notably is RNA mediated interference (RNAi). Double stranded RNA that is complementary to a gene of interest can be easily directed through feeding, soaking or injection and results in decreased gene expression (Fire, Xu et al.

, 1998). The use of systematic reverse genetics has increased the capacity to investigate functional genomics and biological functions of glycosylaton in C. elegans. The genome sequence of C. elegans was first published in 1998 and contains approximately 20,000 genes (C. elegans Sequencing Consortium,

C. e. S. 1998). More recently, the glycome of C. elegans has been studied and characterized. Considering the simple anatomy of C. elegans, its A/-glycan composition is extensive with over 100 structures present in wild-type N2 strains

31 (Paschinger, Guttemigg et al. 2008). The annotated glycome of C. elegans, in addition to its well characterized physiology, provides a fundamental starting point to compare glycan structure with biological function.

Much of the progress in our understanding the etiology of CDG has come from genetic model systems, particularly S. cerevisiae and mice (Aebi and

Hennet 2001). C. elegans share many sequence similarities to mammalian genes involved in the assembly, processing and modifications of a variety of glycans (Schachter 2004). Specifically, the early stages of A/-glycoprotein biosynthesis are conserved from yeast to man (Lehle, Strahl et al. 2006).

However, the later stages of glycoconjugate synthesis are much more complex in vertebrates, especially mammals. C. elegans do share similarities with vertebrate N- and O-glycans in terms of their core structures, but the majority of the differences are found in molecular size and terminal elaborations. The most notable difference between C. elegans and higher animals is the lack of sialic acid in worms (Paschinger, Guttemigg et al. 2008).

C. elegans have some biosynthetic components required for the synthesis of complex mammalian-type complex glycans, namely GnT-l (N- acetylglucosamine transferase-l), GnT-ll, and GnT-V (Chen, Zhou et al. 1999;

Chen, Tan et al. 2002; Warren, Krizus et al. 2002). Despite the identification of such enzymes, the presence of complex and hybrid /V-glycans are either absent or present at low levels in C. elegans (Cipollo, Costello et al. 2002; Natsuka,

Adachi et al. 2002). On the other hand, the occurrence of complex glycan has been found at different developmental stages and may be only present to govern

32 development (Cipollo, Awad et al. 2005). Additionally, C. elegans synthesize N- glycans having terminal fucose residues, similar to Lewis antigens found in humans.

The C. elegans glycome predominately contains high-mannose (Man5.

9GICNAC2), paucimannose (Man3.4GlcNAc2), fucose-rich (attached at the antennae or core), truncated complex (GICNAC3), O-methylated (O-Me) and phosphorlycholine (PC) substituted glycans (Haslam and Dell 2003; Cipollo,

Awad et al. 2004; Cipollo, Awad et al. 2005; Hanneman, Rosa et al. 2006)

(Figure 7). The most abundant class of glycans in C. elegans, and most eukaryotes, is the high-mannose type. This is followed by paucimannose and fucose-rich type structures.

.. T

Figure 7. C. elegans A/-glycans. From left to right are high mannose (Man6-Man9), fucose rich (up to 4 fucose), truncated complex (up to 3 GlcNAc) and paucimannose (Man3-Man4) type A/-glycans.

33 Six research groups have been involved in the analysis of the A/-glycans of C. elegans with mostly overlapping conclusions (Paschinger, Guttemigg et al.

2008). Only minor amounts of truncated complex structures were found by all groups and PC substituted glycans were only observed in some studies. The disparity in the results is partly due to the method of release and the sensitivity of the analysis method. Some of these studies reported structures that were only identified in glycosylation synthesis mutants (mutants lacking mannosidase II, hexosaminidase and fucosyltransferase activity). Most of the 150 published N- glycan structures represent only molecular compositions and do not consider structural isomers. Simply put, a hexose residue can either be a mannose, glucose or galactose. Additionally, a HexNAc can represent a N- acetylglucosamine or /V-galactosamine. More common, even considering monomer identification, is that the structural complexity of carbohydrates is further complicated with different linkage possibilities that change the overall fine structure. Furthermore, comprehensive structural analysis is especially difficult due to the low amounts of many glycan compositions.

Most investigations of A/-glycans in C. elegans use enzymes, namely peptide A/-glycosidase F (PNGase F) and endoglycosidase H (Endo H) for the release of A/-glycans. The chemical hydrazine has also been employed to release A/-glycans from (Natsuka, Adachi et al. 2002). With hydrazinolysis, hexose extensions on the 3- and 3,6-fucose positions on the penultimate and reducing GlcNAc residues were detected in C. elegans (Hanneman, Rosa et al.

2006). Enzymatic release fails to liberate many A/-glycan structures in C.

34 elegans, namely fucose-rich structures (Figure 8). Complex core modifications include a galactose residue linked (31,4 to the core a1,6-fucose and a1,3-fucose on the chitobiose core (Hanneman, Rosa et al. 2006). These structures account for the fucose-rich type A/-glycans found in C. elegans where the majority of fucose residues are found at the core with some residues (up to two) at the terminus. Some suggest that the fucose-rich glycans in C. elegans are the equivalent of complex type /V-glycans found in vertebrates (Altmann, Fabini et al.

2001).

A) 100-

Hex4Gn2Fuc c n Man* \ Hex Gn Fuc 4 2 2 He%Gn2Fuc2 HexsGrtjFuc3 JD Mfs \ Hex4Gn2Fuc3 / CO ?Hex5Gn2Fuc3 Hex6Gr»2Fuc4

Man* / | 4- LL u iLjJL >^ 1000 1230 1400 ^ 1600 1800 2000 2200 B) 100

Hex5Gn2Fuc3 Hex4Gn2Fuc3 03 Hex Gn Fuc u 4 2 \Hex Gn FuC2\\ c s 2 so xxHex4GnjFuc2 \ \ \ HexeGn2FuC3 HeXgGr^Fuc^ •co 2 . HexeGn2Fuc4 \ m \Hex7Gn2Fuc^ \ / \ Maru \ / \ UUftf^Ww^»..Mw 1000 1200 1400 , ieoo 2200

Figure 8. MS profiles of C. elegans A/-glycans released with PNGase F (top) and hydrazine (bottom). More /V-glycan peaks are detectable with hydrazinolysis than an enzymatic release. Adapted from Hanneman et. al. 2006.

35 The high-fucose content of A/-glycans is reflected by the finding of one core crt ,6-fucosyltransferase homologue (FUT-8), one characterized a1,2- fucosyltransferase and five a1,3/4 fucosyltransferases in the genome

(Paschinger, Gutternigg et al. 2008). One of the a1,3-fucosyltransferases (FUT-

1) was proven to modify the /V-glycan core, but is inhibited in the presence of the

GlcNAc residue added by GnTI. Of the remaining four a1,3-fucosyltransferases, two were found to produce Lewis X structures using Gal(31,4GlcNAc-pNP as a substrate, whereas a third one was using a LacdiNAc structure as its receptor; however, FUT-4 was not active with any substrate tested (Nguyen, van Die et al.

2007). The presence of enzymes in C. elegans with Lewis-type activities contrasts with the lack of Lewis-type epitopes; thus the actual role of these enzymes in vivo needs to be determined.

One of the most beneficial methods of studying glycomics is to modulate gene expression by mutation or interference. To uncover the function of a gene, process or structural consequence, the use of mutants is invaluable. Using a mutant with a known function and comparing its glycan profile to a wild-type standard, aids in the definition of the biochemical functions of these genes with respect to the substrate specificities and their involvement in the glycosylation pathways. There are several examples of comparative analyses that have shed light on the biological function of glycosylation genes.

The C.elegans genome contains three forms of GnTI, which is responsible for the addition of a (31,2-GlcNAc to the crt ,3-antenna of the chitobiose core. The triple-knock worm, has a normal phenotype, but fails to synthesize 31

36 paucimannose, complex and fucosylated wild-type structures (Zhu, Hanneman et al. 2004). This mutant only carries a maximum of two fucose residues, which demonstrates the GnTI dependent fucosyltransferases.

OHgem#nrwsJdit N-glycans Other flycart t&bf^awortK

•j ixihn GnGn AM tens m p~m-m fHUBT ifot-n GnGrtF6 m-m

MGn )nnir •"•Ml Man5

GnTIl P*"•*- • 3 JHHI GoMF-W—'GnM <-—— Qnt»r» «——~ MGn —## MM •* MMF Vtf ^ X iftrt-fj itesJ,-j;i , (#>sW? tilnx-2,-3) SfcM) j(fuf« lawyer »Mt)

JHHI GnMF64 ; GflGnf 6*^ MGnF *•—* MGnf%al —» MMF6Gal >-#-#

I Goigl taM«tw*ia»es i?>« i, -J I •Ml f 6 Ifc 'J' h 1 3 3 6 • MMF F*«-«*M0F F mmA •HUB iAjf-1) V** « I •k. f * 1HMI MOP

Figure 9. Biosynthetic glycosylation pathway in C. elegans. Green circle = mannose, yellow circle = galactose, red triangle = fucose, blue square = A/-acetylglucosamine and PC = phosphorylcholine. Adapted from Paschinger et. al. 2008.

37 However, five fucosylated oligomannose A/-glycan structures were present in the mutant, but were not in the wild-type strain, indicating the presence of

GnTI-independent fucosyltransferases. Additionally, the absence of the o> mannosidase II gene (amam-2) which removes one a1,3-linked and one a1,6- linked mannose from Man5GlcNAc2 results in non-wild-type A/-glycans

(Paschinger, Hackl et al. 2006). As expected, the glycome of this strain lacks paucimannosidic truncated structures. Comparative analyses of knock-out C. elegans strains has provided the framework for understanding the complex nature and substrate specificity of glycosylation in the Golgi, little of which was previously known (Figure 9 ). This is only a piece of the pathway and the presence of more complex structures observed in higher organisms does not have such summarized pathways.

The fine structures of C. elegans /V-glycans remains unresolved. Most published data is generally based on mass spectrometry and does not account for structural isomers or even glycan topology. C. elegans do not synthesize sialic acid residues, which distances itself from higher organisms and.may be a shortfall as a model. The more complex A/-glycan moieties such as Lewis x, fucosylated lacdiNAc, bisecting GlcNAc or tetraantennary glycans are not expressed in the worm. Despite the difference in complexity with mammals, our understanding of the role of glycoconjugates in development, disease and overall significance is largely due to the identification and comparative analysis of mutations in genes in C.elegans involved in biosynthesis. The C. elegans sequenced genome contains an abundance of glycosyltransferases,

38 glycosidases and lectins which highlights the potential of the worm for investigating basic roles of glycans in development. CHAPTER 4

PROJECT AIMS

The focus of this work will examine two aspects of CDG illnesses modeled in C. elegans. The first approach will explore genetic factors that may contribute to CDG pathology. The second portion of this study will investigate the structural consequences of a specific CDG type II illness.

4.1 Genome-wide RNAi Screen to Identify Tunicamycin Hypersensitive Loci in C. elegans

For any CDG type I, in which only one genetic defect is involved, there exists no consistent clinical manifestation. The biosynthetic pathway for glycoprotein formation is quite well understood and has provided a basis for establishing the etiology of CDG (Schollen, Kjaergaard et al. 2004). However, there are still cases of CDG with unknown etiology and the mechanistic consequences of defective glycoprotein synthesis are very poorly understood.

Given the diversity of glycoproteins made by animals, it is likely that there are many diseases to which this is relevant; CDG is simply paradigmatic. In broad

terms, clinical heterogeneity among CDG patients could have four origins. An individual may have inadequate genotypes in: 1) polypeptides that require

40 glycans for function, 2) polypeptides that require glycans for structure, 3) genes responsible for glycoprotein maturation (i.e. unfolded protein response, secretion or Golgi function, 4) genes that are involved in the lipid-linked oligosaccharide assembly pathway. It is plausible that an individual's genetic background may alter or aggravate the glycosylation defect. To explore this hypothesis, RNAi, in combination with a glycosylation inhibitor, will be used in a genome-wide screen to identify genes that make CDG worse.

4.2 Comparative N-qlvcan Structural Analysis of the C. eleaans N2 (Bristol) and NF299 coac-1(k179) strains

The second portion of this study will investigate the structure-function relationship of a specific CDG type II illness. Defects in the conserved oligomeric

Golgi complex-subunit 1 (COGC-1) cause CDG-II illnesses in humans. Recently a C. elegans knock of the COGC-1 mammalian ortholog was made available through the Caenorhabditis Genetics Center. It is believed that COGC-1 defects compromise the function of the Golgi and trigger incorrect glycan assembly. This strain will enable comprehensive glycan comparative analysis of a CDG- ll/COGC-1 state verses a wild-type condition via mass spectrometry. The work will analyze the implications of defective glycosylation machinery and will seek to relate Golgi function to aberrant glycan structures. The overall goal is to determine glycosylation trends in CDG-II/ COGC-1 type conditions modeled in C. elegans. Structural analysis of the COG complex defective C. elegans strain will

41 help in understanding the molecular basis of CDG and other glycosylation- dependent diseases. CHAPTER 5

GENOME-WIDE RNAi SCREEN TO IDENTIFY TUNICAMYCIN HYPERSENSITIVE LOCI IN C. ELEGANS

Glycosylation (CDG). CDGs are categorized as type I or type II depending on the location in the biochemical pathway where the defect occurs. Type I disorders are characterized as defects of enzymes in the lipid-linked oligosaccharide assembly pathway. Currently there are 12 known defects, or subtypes, in this category. Type II disorders are deficiencies of enzymes involved in the trimming and processing of the protein bound oligosaccharide. To date there are eight defective enzymes within type II CDGs (Jaeken and Matthijs

2007; Zeevaert, Foulquier et at. 2008).

CDG is a rare disease with roughly 900 documented cases (Vodopiutz and Bodamer 2008). In addition to the infrequency of CDG, there is evidence suggesting a clear disequilibrium with the frequency of the disease in the population (Schollen, Kjaergaard et al. 2000). Two possibilities exist to explain the lack of CDG patients, either CDG illnesses are severely underdiagnosed or those afflicted die before birth (Matthijs, Schollen et al. 2000; de Lonlay, Seta et al. 2001). Although neither theory has been established, is it plausible that each

43 possibility is mutually exclusive depending on the severity of the biosynthetic defect (i.e. type, subtype or specific genetic mutation within subtype).

Individuals born with CDG are inundated with defective glycosylation enzymes leading to absent and/or aberrant glycan structures (Mills, Mills et al.

2001). CDG patients suffer from psychomotor retardation, low muscle tone, incomplete brain development, visual problems, seizures, stroke-like episodes, coagulation disorders, endocrine abnormalities and overall failure to thrive

(Eklund and Freeze 2006). The biosynthetic pathway for glycoprotein formation is quite well understood and has provided a basis for establishing the etiology of

CDG (Jaeken and Matthijs 2001) (Figure 10). However, there are still cases of

CDG with unknown etiology, where strong evidence points to CDG but the defective gene has yet to be identified (Prietsch, Peters et al. 2002). As well as the problematic genetics factors of CDG, the mechanistic consequences of defective glycoprotein synthesis are very poorly understood which confounds the clinical identification of this disease.

44 /V-Glycoprotein Biosynthesis LLO assembly

F6P I PMI40 I CDG-1b M6P"— M SEC53 i CDG-1a Tn M1P UMP UDP-N I UDP-N CTP GDP-M UDP-G

I DppN, «— DppN ^ Dp ^-s£D5- Dp 5^, DpM DpG ^ Dp

GDP-M -v^ CDG.1k_ CDG-1] CDG-1e_

Pep-N2M9G3 DppN2M GDP-M —>. ALG2 DpM DpG STT3 0ST1 WBP1 ) CDG-H OST3 0ST6 SWP1 OST2 OST5 OST4 DppNgM2 dad-1 GDP-M—>. I MPDU1 DppN2M9G3 CDG-1f ALG10\ DppNgMg DppN M G GDP-M - 2 9 2 ALG8 J CDG-1 h

DPPN2M4 DppN2M9G GDP-M • T T t ALG6 A CDG-1c RFT1 ALG3 ALG9— ALG12 ALG9 jy-cbWid^P^^e ^LPPP^Myc-^gDppNsMsj^ Dpp^M, ER

DppN2M5 Cytosol

Figure 10. Lipid-linked oligosaccharide pathway in the endoplasmic reticulum. CDG type diseases are illustrated. S. cerevisiae homologues are also shown (alg: asparagine-linked glycosylation).

Since all CDG-I etiologies alter the effectiveness and substrate availability of oligosaccharyltransferases, it might be expected that the disease would manifest itself consistently from patient to patient. This is not the case, as clinical

CDG symptoms within types and subtypes vary considerably between individuals making diagnosis difficult. Additionally, in any specific CDG illness, where the genetic mutation is the same, no consistent clinical manifestation exists and the severity of the illness ranges from very mild to severe (Aebi and Hennet 2001).

45 Given the ubiquity of glycoproteins and the ability of /V-glycans to modulate protein function, it would seem likely that defects in glycosylation would lead to diverse and pleiotropic phenotypes, such as the multisystemic syndromes associated with CDG-I. Since glycoprotein oligosaccharides can only exert their effects in the context of the polypeptide to which they are attached, variation in the glycoprotein polypeptide gene in outbred populations could interact with a

CDG-I etiology to modify the consequences for specific glycoproteins. This collectively could trigger the broad clinical spectrum seen in an individual patient.

To date clinical evidence fails to explain the exceeding inconsistency between the etiology and pathology in all CDG conditions.

It is plausible that an individual's genetic background may alter or aggravate a CDG glycosylation defect. Given the diversity of glycoproteins made by eukaryotes and the extent of A/-glycosylation, additional biochemical defects directly or indirectly involved with a glycosylation defect could elicit variable phenotypes characteristic of CDG diseases. For example, the clinical severity of

CDG-la, caused by phosphomannomutase deficiency, is enhanced by a defect in a1,3-glucosyltransferase, an enzyme also found in the lipid-linked oligosaccharide (LLO) pathway (Westphal, Kjaergaard et al. 2002). Defects in a1,3-glucosyltransferase alone result in CDG-lc illnesses. In addition to this example, previous reports allude to the possibility of genetic and/or environmental factors that contribute to the clinical spectrum observed in CDG-I patients (Freeze and Westphal 2001).

46 In broad terms, clinical heterogeneity among CDG patients could have four origins. An individual may have mutations in genes encoding: 1) polypeptides that require /V-glycans for function, 2) polypeptides that require N- glycans for structure, 3) genes responsible for glycoprotein maturation (i.e. unfolded protein response, secretion or Golgi function), 4) genes that are involved in the lipid-linked oligosaccharide assembly pathway (Figure 11). Of the four theoretical types of mutations, two fall in the "effector" category, where the protein component of a glycoprotein is not synthesized correctly. A third, or

"etiologic" type, are flawed loci that result in defective glycosylation machinery

(i.e. CDG-type mutations). The fourth, or "maturation" type, are gene products responsible for secretion and quality control of immature glycoproteins. Any individual with CDG may have additional genetic defects anywhere in their genome that rely on A/-linked glycosylation to function correctly. It is this interaction between the CDG glycosylation defect and a secondary genetic mutation or flaw that is the grounds for the variable pathology in CDG cases.

47 Maturation genotypes Effector genotypes « » •4 Underlying polypeptide genes v Trafficking j®^ Localization I i I Tunicamycln Oligosaccharyl ^ Nascent ^ Glycan ^ alvcoDrotains -transferase glycoprotein remodeling 9y p Dolichol T Lipid-linked oligosaccharide t I I assembly E* Oligomer Phenotypes Monosaccharides ' ^uality assembly control t > A Etiologic genotypes Unfolded protein response

Figure 11. /V-glycosylation pathway from synthesis to localization. Plausible interactions that enhance defects in LLO biosynthesis that cause CDG I diseases to fall into three categories based on their role in /V-glycosylation.

In this study genetic factors that influence CDG pathogenesis are explored. The notion is that for any cellular function that depends on N- glycosylation, a protein component must be involved (the underlying polypeptide of a glycoprotein, a lectin receptor etc). Because so many structures and functions are compromised in /V-glycosylation diseases, the gene products that contribute to lethality or particular symptoms cannot be resolved. This experiment demonstrates a novel approach that combines RNA mediated interference (RNAi) in combination with a /V-glycosylation inhibitor to systematically dissect this problem on a gene-by-gene basis.

Much of the progress in understanding the etiology of CDG-I has come from genetic model systems, particularly S. cervisiae and mouse (Aebi and

Hennet 2001). By exploiting C. elegans' susceptibility to genetic silencing, it is

48 possible to conduct a genome-wide RNAi screen to identify genes that are dependant upon A/-linked glycosylation.

To accomplish this, a sub-phenotypic dose of the nucleoside antibiotic tunicamycin was used to mimic a very mild CDG-l-like condition by partially inhibiting the first step in the LLO pathway. Tunicamycin inhibits the action of the

UDP-GlcNAc dolichol-phosphate A/-acetylglucosamine 1-phosphate transferase

(DPAGT1) in the first step of the lipid-linked oligosaccharide biosynthetic pathway (Zhu, Zeng et al. 1992). Deficiencies of this enzyme in humans cause

CDG-lj illnesses. CDG-lj is caused by hypomorphic mutations in the human

ALG7 gene, which is conserved in eukaryotes and encodes the enzyme responsible for dolichol pyrophosphate /V-acetylglucosamine formation (Barnes,

Hansen et al. 1984; Wu, Rush et al. 2003). Patients diagnosed with CDG-lj suffer from severe developmental delays, hypotonia, seizures, microcephaly, exotropia and death. Without this reaction, no donor for oligosaccharyltransferase can be formed and protein A/-glycosylation is absent.

In addition to tunicamycin treatment, the expression of an individual glycan-dependant protein was decreased via RNAi, which will render it hypersensitive to reduced glycosylation. Because this interaction is largely specific, glycan-dependant gene products can be identified. Under these conditions, it is assumed that gene-products that control glycosylation efficiency or glycoprotein maturation, or encode key glycoproteins that require glycosylation for their synthesis or activity, will manifest synthetic enhancer or suppressor phenotypes. The genome-wide RNAi screen with the ORFeome v1.1 library and

49 its analysis in silico have revealed 512 tunicamycin-hypersensitive loci, some of which are characterized and cloned but not previously recognized to be N- glycosylation dependent.

5.1 Materials and Methods

The C. elegans strains NL2099 rrf-3 (pk1426), VC569 tag-179 (ok809),

RE666 ire-1 (v33) and N2 (Bristol) were obtained from the Caenorhabditis

Genetics Center, University of Minnesota, USA. General methods used for cultivating, handling and genetic manipulation of C. elegans are as described

(Brenner 1974). Genome-wide RNAi screening protocols were performed by feeding in NL2099 rrf-3 (pk1426), with minor adaptations from previous works

(Simmer, Moorman et al. 2003). The ORFeome v1.1 RNAi library was supplied from the Vidal lab at the Dana Farber Caner Institute (Reboul, Vaglio et al. 2003;

Rual, Ceron et al. 2004). Tunicamycin (Calbiochem) was dissolved in DMSO to

50 mg/ml stock concentration. To obtain synchronous populations of larval stage

1 hatchlings, eggs were acquired by digesting populations containing gravid hermaphrodites with alkaline hypochlorite and washing the egg pellet three times by centrifugation with M9 buffer (Hope 1999). Eggs hatched overnight in M9 buffer or were transferred to 35mm petri dishes. C.elegans strains were cultivated at 20°C unless otherwise indicated. Observations were made using a

Leica MS 5 stereomicroscope at 10x-40x magnification (Leica Microsystems).

50 To establish the effect of tunicamycin on larval lethality, individual hatchlings (n=144) were placed in each well of a 24-well cluster plate containing

NGM supplemented with tunicamycin (0-10 ug/ml). On day three and day six, plates were scored for death (no pumping, twitching or movement when prodded with platinum wire) of the founder hermaphrodite or the appearance of hatched progeny.

The larval phenotype assay was performed following synchronization via egg prep. 15 ul of hatchlings (-200 L1 Larvae) in buffer solution were pipetted onto seeded 60mm NGM plates containing 0-5 ug/ml tunicamycin. Animals were counted and phenotypes were scored after three days.

Similarly to the larval lethality assay of N2, separate populations of

NL2099 rrf-3 (pk1426), RE666 ire-1 {v33), VC569 tag-179 {ok809) and N2

(Bristol) were tested in increasing concentrations of tunicamycin. Each strains was synchronized by alkaline hypochlorite egg prep method. 5 ul of hatchings

(-150 L1 Larvae) in M9 buffer were placed on 35mm Petri dishes containing 0-10

|ag/ml tunicamycin. Animal survival was scored based on the ability of larvae to

reach gravid hermaphrodites after three days.

All RNAi was performed as observed by Simmer, Moorman et al. The

ORFeome v1.1 library of recombinant E. co//strains, consisting of 11,942 constructs, was arrayed in microtiter plates stored at -80°C. Bacteria were cultured in 2xYT containing 50 ug/ml ampicillin and 12.5 u.g/ml tetracycline for 72 hours at 22°C to ensure the presence of the plasmid (ampicillin selection) and the

DE3 lysogen carrying the IPTG inducible T7 RNA polymerase (tetracycline

51 selection). From these, 50 uL liquid cultures were grown overnight in 2xYT supplemented with 50 u,g/ml ampicillin at 37°C with shaking. 15 uL aliquots of the overnight culture were used to grow bacterial lawns on 24-well clusters of

NGM supplemented with 50 u,g/ml ampicillin and 1mM IPTG and 2 u,g/ml tunicamycin. These cultures were grown at room temperature for 48 hours to create a bacterial food source expressing double-stranded RNA.

In both phases of the screen, 24 well clusters were used as the platform for C. elegans analysis. Each column, consisting of 4 wells, consists of one

RNAi gene. In this manner, each gene tested provides four observable wells.

Initially, 7 to 10 NL2099 rrf-3 (pk1426) animals were placed in the top row of each plate to generate progeny that have only existed in the presence of each

RNAi construct and 2 u.g/ml tunicamycin. After two days of "priming", a single gravid hermaphrodite from the top row was transferred to each of the other lawns and allowed to produce F1 progeny. During the triage screen, genes where at least two of the three replicates displayed concordant phenotypes were regarded as candidates. Any phenotypes observed in the 'priming' well were also scored and data collected was weighed with lesser significance when selecting phase II candidates. Phenotypes in the three remaining wells, including P0 and Fi progeny, were given a value of 10. Phenotypes in the "priming" well was marked a value of 5.

For the genome-wide screen the preparation of media and bacterial culture were followed as stated previously. 24 well cluster plates were used for both the triage and follow-up screens (phase I & II). Each cluster yielded results

52 for 6 genes. Every column corresponded to a specific RNAi gene (4 wells total).

7 to 10 larval stage 3 NL2099 rrf-3(pk1426) animals were plated by hand to each well in the top row of each cluster. After 48 hours at 15°C to "prime" these parents, single gravid hermaphrodites were transferred to each of the three replicate wells in the column and allowed to segregate progeny at 22°C for 72 hours. All wells were scored for any visible phenotypes in P0 and F-i generations and noted using a controlled vocabulary.

5.2 Results

The phenotypic consequences of a CDG-l-like defect were characterized in C. elegans using 0, 3 and 5 u.g/ml tunicamycin (Table 3). Like CDG-I in humans, tunicamycin treatment in C. elegans results in a pleiotropic phenotype.

In the complete absence of tunicamycin, nearly all animals reach adulthood and exhibit no aberrant phenotype. At 3 (ig/ml concentration, the consequences of tunicamycin treatment become more evident. Only -76% of the population is wild type and reach the adult stage. Nearly 10% of the larvae die. Interesting to

note is the range of phenotypes present with the mild dose of drug; characteristic of CDG-I diseases. When the tunicamycin concentration is increased to 5 ng/ml almost all animals remain in the larval stage, of which -75% are lethal. Only 1% of animals scored grow to become wild type adults.

53 [Tunicamycin] (ug/mL) Stage Phenotype 0 3 5 W.T. 99.81% 75.63% 1.01% Dpy 0.00% 3.39% 1.72% Unc 0.00% 2.35% 0.20% Sma 0.00% 2.04% 0.00% Adult Vab 0.00% 0.63% 0.00% Egl 0.00% 0.57% 0.00% Clr 0.00% 0.05% 0.00% Dpy Unc 0.00% 0.05% 0.00% Dpy Egl 0.05% 0.05% 0.00% Gro 0.14% 1.51% 14.34% Larval Dpy Gro 0.00% 1.10% 7.95% Let 0.00% 9.71% 74.77% n= 2119 1916 1974

Table 3. Tunicamycin incudes a pleotropic condition in C. elegans. Tunicamycin concentrations are directly correlated with an increase in observable phenotypes and longevity. Tunicamycin concentrations at 3 u,g/ml trigger a wide variety of phenotypes among N2 adults and larvae, emulating a CDG-l-like pathology. Concentrations at 5u.g/ml and above result predominantly in lethality. (Clr, clear patches; Dpy, dumpy; Egl, egg-laying defective; Gro, slow growth; Let, lethality; Pvl, protruding vulva; Sck, sick; Sma, small; Unc, uncoordinated locomotion; Vab, variably abnormal morphology; W.T., wild type.)

The success of the genome-wide RNAi screen relies on a low dose of the antibiotic tunicamycin to emulate a mild form of CDG type Ij. In C. elegans high doses of tunicamycin cause adult and embryonic lethality, but partial inhibition by lower doses results in a variable phenotype. Tunicamycin at a concentration of 2 ug/ml or less does not significantly affect larval development, but at greater concentrations survival rates decrease (Figure 12). However, the presence of 2

54 |jg/ml tunicamycin emulates a lesser "sub-phenotypic" form of CDG-lj in C. elegans.

01 23456789 10 Tunicamycin (ng/ml)

Figure 12. Tunicamycin induces a dose-dependant lethality in the N2 C. elegans strain.

Tunicamycin dose-dependent lethality in C. elegans can vary from one mutant strain to another (Figure 13). Four strains were grown in the presence of increasing concentrations of tunicamycin. The four test-strains included: 1) N2

(Bristol) wild-type strain, 2) VC569 tag-179 (ok809) which is a complete loss of function of glucosyltransferase activity required to add the terminal glucose to the

LLO formed in the endoplasmic reticulum, 3) RE666 ire-1 (v33) which is required for the unfolded protein response (UPR) that can be caused by incomplete glycosylation of proteins and result in their accumulation in the endoplasmic reticulum, 4) NL2099 rrf-3 (pk1426) which encodes an RNA-directed RNA polymerase that inhibits somatic RNAi and renders C. elegans hypersensitive to

55 RNAi gene knockdown. The C. elegans VC569 strain is exceedingly sensitive to tunicamycin possibly due to a decrease in substrate availability caused by blockage of DPAGT1 in the ER. Additionally, the strain RE666 is more sensitive to tunicamycin than N2 (Bristol) or NL2099 rrf-3(pk1426) strains. Elevated concentrations of tunicamycin can trigger the UPR and is evident in this assay

(Iwata and Koizumi 2005). Since the NL2099 strain was used in the genome- wide screen, it is necessary to note that the effect of 2 |ig/ml tunicamycin between N2 and NL2099 is comparatively minor.

@ NL2099 rrt-3(pk1426)\\ o T'24D1 A{ok809)\ • RE666ire-1(v33)ll A N2 (Bristol) -a

-1.0 [Tm] lg(n.g/mL)

Figure 13. Tunicamycin affects postembryonic development among C. elegans strains. The percent of animals that reach fertile adulthood (% gH) were measuered as a function of tunicamycin concentration.

A series of RNAi trials were conducted to validate the legitimacy of the screen and the existence of tunicamycin-hypersensitive loci. Based on previous studies, a small number of genes that were either known or suspected to be

56 hypersensitive to tunicamycin were assayed on 2 pg/ml tunicamycin (Shen, Ellis et al. 2001). These included 6 "etiologic" genes in-the LLO pathway (Table 4) and 6 "maturation" genes in endoplasmic reticulum-Golgi functions (Table 5). To substantiate tunicamycin hypersensitivity of both genotypes, genes from the ER luminal phase of the LLO pathway and various genes in secretion were knocked- down via RNAi in an rrf-3 mutant background.

Gene Product Wormbase -Tm +Tm

K09E4.2 DppN2M6 W.T. Gro Emb

C14A4.3 DppN2M7/9 W.T. W.T. Ste,Emb

ZC513.5 DppN2M8 not tested W.T. Emb,Lvl,Lva

C08B11.8 DppNaMgGi W.T. W.T. Gro

C08H9.3 DppN2M9G2 W.T. W.T. Emb,Lvl,Lva

T24D1.4 DppN2M9G3 W.T. W.T. Brd.Gro Vector W.T. W.T.

Table 4. RNAi of LLO genes with 2pg/ml tunicamycin. RNAi of loci in the LLO pathway presents no overt phenotype on NGM alone. In the presence of 2 ug/ml tunicamycin, RNAi silencing of etiologic genotypes enhances hypo-glycosylation that causes multiple phenotypes. Corresponding RNAi phenotype experiments from WormBase were are also shown for comparison.

Our results were also compared to previously described RNAi phenotypes with an rrf-3(pk1426) mutant background of the same genes annotated in

WormBase (http://www.wormbase.org). All RNAi phenotypes tested correlated with previous works except in the case of sequence name C08H9.3 which

encodes the glucosyltransferase responsible for the second glucose addition on the N2M9Gi LLO precursor (Rual, Ceron et al. 2004). RNAi phenotypes in our

57 hands produced a slow growth phenotype, whereas previous reports found larval arrest, embryonic lethality and reduced brood size. Additionally, K09E4.2 generated a "Gro" RNAi phenotype in the absence of tunicamycin. Published

reports found no observable phenotype via RNAi (Kamath, Fraser et al. 2003;

Rual, Ceron et al. 2004). However, the slow growth phenotype was enhanced to embryonic lethal in the presence of 2 u.g/ml tunicamycin.

To verify the "maturation" hypothesis, trials with RNAi library plasmids

representing all suspected and known tunicamycin-hypersensitive loci were tested (Table 5). Once again, results were consistent with those made by others with the exception of sel-9 which gave no phenotype in the absence of tunicamycin. By and large the RNAi assays were consistent with published

results and also demonstrate a 10-30% replication variability of the genes tested which is common among laboratory-to-laboratory RNAi experiments in C.

elegans (Simmer, Moorman et al. 2003).

58 Function Gene Locus Worm base -Tm +Tm ERAD K08E3.7 pdr-1 W.T. W.T. Ste,Emb,Gro UPR C41C4.4 ire-1 W.T. W.T. Emb,Lva UPR F46C3.1 pek-1 W.T. W.T. Gro GVT W02D7.7 sel-9 Gro,Unc,Pvl W.T. Ste.Gro GVT Y60A3A.9 srf-4 not tested W.T. Let, Gro GVT F47G9.1 srf-8 not tested W.T. W.T. Vector W.T. W.T.

Table 5. RNAi of "maturation" genes with 2ug/ml tunicamycin. RNAi silencing of genes involved in three aspects of ER-Golgi networking exhibit no phenotype on NGM alone. However in the presence of 2 ug/ml tunicamycin, synthetic effects are evident in a range of phenotypes. ERAD = ER associated degradation, UPR = unfolded protein response and GVT = Golgi vesicle trafficking.

These tests establish that RNAi, in combination with a low dose of tunicamycin, can generate synthetic phenotypes to identify genes that are tunicamycin hypersensitive. All of the above genes were included as positive controls in the genome-wide screen and at the culmination of the screen 8 of 12 control genes remained present in the final results.

Moreover, we cannot exclude the possibility that 2 u.g/ml tunicamycin causes a generalized increase in the effectiveness of the RNAi by feeding method. It is important to show that the presence of tunicamycin does not alter the effectiveness of RNAi, or a significant number of genes that control visible phenotypes (particularly neuronal ones) that are refractory to RNAi will become obvious as screen results are generated. If this was the case, then our approach

59 would result in numerous false positives. To exclude this possibility, loci known to be refractory to RNAi (i.e. genes localized to neuronal cells) and that have observable mutant phenotypes were tested on 2 u.g/ml tunicamycin (Asikainen,

Vartiainen et al. 2005). It is expected that the RNAi phenotype of each locus would be unaltered by drug treatment. These results proved the legitimacy of the screen showing that tunicamycin-hypersensitive genes can be identified in the presence and absence of tunicamycin (Table 6).

Gene Locus Mutant Worm base +Tm

F11C3.2 unc-122 UNC W.T. W.T. W03A3.1 ceh-10 UNC W.T. W.T. C52B9.9 mec-18 UNC W.T. W.T. C17D12.2 unc-75 UNC W.T. W.T. C52E12.3 spv-7 LET W.T. W.T. D2096.4 spv-1 LET W.T. W.T. ZC64.3 ceh-18 LET W.T. W.T.

Table 6. Tunicamycin treatment does not alter RNAi effectiveness in C. elegans. Seven C. elegans strains that exhibit both a mutant phenotype and wild-type RNAi phenotype were chosen to test RNAi in the presence of tunicamycin. All strains remained refractory to RNAi when tunicamycin was present, demonstrating that tunicamycin does not modify the efficiency of RNAi.

60 The genome-wide screen required strict criteria for candidate selection for the second phase to limit the accumulation of false positives. A set of 100 genes that gave rise to phenotypes in 0,1,2,3 and 4 observatory wells out of 4 in the triage screen were tested in the presence and absence of 2 ng/ml tunicamycin

(n=500 total genes tested).

The penetrance was assessed by subtracting the number of wells showing any phenotype(s) with tunicamycin from the number in its absence (A). The larger the value of A, the more dependent the phenotype penetrance is on tunicamycin; i.e. a locus that induced phenotypes in all 4 wells containing tunicamycin but none in its absence would have A=4. The delta values were plotted against the percent total in each of the 5 sets (100 genes per set) tested.

The effect of expressivity was assessed similarly except that A is calculated using the number of wells showing non-viable phenotypes under each condition. In this case, the larger value of A, the more severe the phenotype becomes when tunicamycin is present; i.e. a locus that induced a "dumpy" phenotype in all 4 wells in the absence of tunicamycin but "lethal" in all wells when drug was present would have A=4. Under this analysis, false positives and loci causing single-gene phenotypes score A=0, but those genes that are tunicamycin-hypersensitive score A>0 (Figure 14).

61 Penetra nee number of phase I Expressivity . f\ positive w ells number of phase I / \ ...... 4 80- positive w ells 60- ...... 4 / A\ -o-- 2 —a—2 (0 60- (0 ; 0 — 40- W JL 40-

m - o f clas s "5 s s« BjFoSa 3 - . M» 20- 20- I 1* l ^ \-u a,^ \\ W»Z^*^JF ' ^^**s*n»^!~2 0 'T V i " 1 ' 1" ™"l 1 T" -4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 12 3 4 A A Figure 14. Calculation of phenotype penetrance and expressivity. These graphs represent the criteria for genes that were retested in the second phase of the genome-wide screen. Delta values were calculated by subtracting the number of wells showing any phenotype (penetrance) and wells showing non viable phenotypes from the number in its absence. For penetrance, the larger delta values correspond to a greater dependence of the phenotype penetrance on tunicamycin. For expressivity, the larger the delta values correspond to an increase in phenotype severity when tunicamycin is present.

Loci causing phenotypes in <1 well during the triage screen behave randomly during retesting and the group results fitted to a Gaussian curve centered on 0 as expected. However, loci that caused phenotypes in >2 wells during the triage screen behaved non-randomly in the second round with a substantial fraction scoring A>2 which was two standard deviations from the mean and represented tunicamycin-hypersensitive loci.

The genome-wide RNAi screen was divided into two phases: 1) a triage screen of the ORFeome v1.1 RNAi library for all genes that cause a visible synthetic phenotype in the presence of 2 u,g/ml tunicamycin and 2) a follow-up screen where candidate genes are retested in the presence and absence of 2

62 |ig/ml tunicamycin to separate tunicamycin-hypersensitive loci from single-gene phenotypes and false positives. During the triage screen, criterion for selecting candidates for progression to the second phase was experimentally established

(i.e. phenotypes in 2 or more wells out of 4).

The completion of the phase I triage screen required 13,190 RNAi assays in the presence of 2 u.g/ml tunicamycin. Of those genes tested in phase I, 4,459 genes were candidates for the phase II follow-up screen based on the criteria mentioned above. Ultimately, 512 genes were found to be tunicamycin- hypersensitive. These results were further analyzed to determine function location and presence of Asn-X-Ser/Thr sequon for /V-linked glycosylation (see

Discussion & Supplementary Data). Additionally, qualified results were classified according to their eukaryotic orthologous group (KOG) assignments to determine cellular functions (see Discussion).

5.3 Discussion

The core of this analysis is that mutations in "etiologic", "maturation" or

"effector" loci will interact genetically and that this interaction is the basis of the clinical diversity and heterogeneity in CDG-I. Much of the progress in understanding the etiology of CDG-I has come from genetic model systems, particularity S. cerevisiae and mouse (Aebi and Hennet 2001). By utilizing RNAi silencing in C. elegans, we can identify tunicamycin-hypersensitive loci, effectively asking what genes enhance the severity of a CDG-I condition? The

63 widespread understanding of C. elegans and its' susceptibility to genetic silencing it enables the investigation of functional glycomics and allows us to systematically investigate CDG interactions on a gene-by-gene basis. Overall, each gene that was determined to be hypersensitive was tested at least three times; two assays with tunicamycin and one without. A list of 512 genes was found to be decidedly tunicamycin-hypersensitive.

It is also important to consider that some tunicamycin-hypersensitive genes may be involved in drug metabolism or and not directly interrelated with N- glycosylation. Effectively, knocking down these genes may increase the absorption or decrease detoxification of tunicamycin and will result in an increased drug effect. However, the quality of this experiment relies on precise drug treatment and only in the absence of drug can this possibility be excluded.

That is to say future experiments would include a double mutant containing both the RNAi sensentive rrf-3(pk1426) mutant background with an additional loss-of- function mutation in the LLO pathway, thereby effectively replacing drug treatment with a genetic condition.

Currently it is difficult to characterize each gene into its respective functional class, but some results can be drawn from previous works. This study cannot definitively prove the assignment of specific genes to particular functions without confirmatory follow up. However this screen does provide researchers with a list of potential genes that could be used to screen human CDG patients in

hopes of identifying enhancer loci that contribute to the severity of the CDG-I diseases.

64 Furthermore, results were evaluated by comparing our results to eukaryotic orthologous groups of proteins (KOGs) from sequenced genomes

(Tatusov, Fedorova et al. 2003). This analysis classified our tunicamycin- hypersensitive genes into functional protein sets. Of the 512 tunicamycin- hypersensitive genes confirmed, 308 had KOG assignments and as a result 204 were omitted from analysis (Figure 15). The two largest known groups of KOG protein sets in our results were involved in translation, ribosomal structure and biogenesis (n=49) and signal transduction mechanisms (n=33). Unsurprisingly, the subsequent two groups were posttranslational modification, protein turnover, chaperones (n=27) and intracellular trafficking, secretion, and vesicular transport

(n=20). Unfortunately, almost two-thirds of the results are uncharacterized and have no known function. This analysis helps to reasonably classify results in

"effector", "maturation" or "etiologic" gene classes.

65 Translation, ribosomal structure and biogenesis jHaaaaaaHaHHHaBHaaMBHMBHaai 49 Transcription ••••••••i 18 Signal transduction mechanisms (•••••••••••••••• 33 Secondary metabolites biosynthesis, transport and catabolism • 1 Replication, recombination and repair ••••• 9 RNA processing and modification •••••••••i 18 Posttranslational modification, protein turnover, chaperones (•••••••••••• 27 Nucleotide transport and metabolism Mi 3 Nuclear structure • 1 Lipid transport and metabolism !•••• 8 Intracellular trafficking, secretion, and vesicular transport PBMHBMHBI 20 Inorganic ion transport and metabolism pm 4 General function predictions only wmmmmmmmmmmmmmmmm 37 Function unknown Extracellular structures mmm 7 Energy production and conversion Defense mechanisms ••• 5 Cytoskeleton paM 7 Coenzyme transport and metabolism an 3 Chromatin structure and metabolism ••• 6 Cell wall/membrane/envelope biogenesis m» 3 Cell cycle control, cell division, chromosome partitioning ••••• 10 Carbohydrate transport and metabolism •• 4 Amino acid transport and metabolism • 2 6 10 20 . 30 40 50 60

Figure 15. RNAi gene assignments via KOG analysis. Analysis of eukaryotic orthologous groups (KOGs) classify tunicamycin hypersensitive genes into functional sets. Of 512 genes found to be hypersensitive, 308 had KOG assignments.

In addition to KOG analysis, cellular location was determined with

Proteome Analyst (http://path-a.cs.ualberta.ca), a web based tool that can analyze the function and subcellular location of each sequence based on

BLASTS for sequence alignment. Brief identifications and/or concise descriptions

of each gene product were annotated manually with the aid of WormBase.

Results were also examined with NetNglyc

(http://www.cbs.dtu.dk/services/NetNGIyc), an online server that predicts N-

glycosylation sites in proteins using artificial networks that examine the sequence

66 context of Asn-Xaa-Ser/Thr sequons supplied in FASTA format (supplementary data).

As mentioned above results were expected to fall into three categories;

"effector", "etiologic" and "maturation". "Effector" gene products can be further classified into two subclasses. "Formation" subclass members require N- glycosylation to fold, pass quality control or to be trafficked within the cell. Those in the "function" subclass require A/-glycosylation for their extracellular function; their catalytic, ligand-receptor, localization or stability properties are modulated by the presence and structure of the glycans. Again stating whether the glycan function is a structural or functional role is not possible at this point, but these two subclasses can be assessed when examined together.

There are several examples of "effector" class genes found in the results.

The gene, eff-1 (epithelial fusion failure), is a known type I transmembrane glycoprotein required for epithelial cell fusion (Mohler, Shemer et al. 2002). RNAi knock-down of eff-1 in the presence of 2 u.g/ml tunicamycin caused larval arrest, sterility, uncoordinated movement, slow growth and clear phenotypes. RNAi treatment in the absence of tunicamycin remained wild-type phenotypes. A second "effector" type result is the mom-2 (more mesoderm) gene which is a

member of the Wnt family of secreted signaling glycoproteins. In addition to

being a glycoprotein, MOM-2 may also function as a ligand in Wnt signaling during embryonic development to specify the production of endoderm cells in the gut (Thorpe, Schlesinger et al. 2000). Gene knock-down of mom-2 with tunicamycin generates embryonic lethality, small brood size and sick

67 phenotypes. In the absence of tunicamycin, the mom-2 RNAi phenotype is wild- type.

The less-than expected amount of "etiologic" class genes in the results may be due to a general reduction in the number of possible candidate genes present in the C. elegans genome or their absence in the ORFeome v1.1 RNAi library. There are 17 genes in the C. elegans genome that correspond to enzymes in LLO biosynthesis (Berninsone 2006). Of those 17 genes, 11 are present in the RNAi library. 4 genes of 11 were found to be tunicamycin- hypersensitive and 2 genes produced robust phenotypes with and without drug.

Only 4 of the 11 genes failed to qualify from the phase I triage screen, which is well within the false negative rate that is typical in genome-wide RNAi screens

(Kamath, Fraser et al. 2003; Simmer, Moorman et al. 2003).

In addition to the 17 genes mentioned above, 4 genes were identified in the screen that are involved directly in /V-linked glycosylation and not part of LLO biosynthesis. One such result is the gene T09A5.11, which encodes a 48 kD subunit of the oligosacchyltransferase complex responsible for the en bloc transfer of the LLO oligosaccharide to nascent polypeptides in the endoplasmic reticulum. The RNAi phenotype of this gene knock-down was identical in the presence and absence of tunicamycin. However when drug was present, the expressivity of the embryonic lethality phenotype increased from 25% to 75%.

The other 3 genes found to be tunicamycin-hypersensitive in the final results encode 2 mannosyltransferases and 1 glucosyltransferase that are found in LLO

68 biosynthesis. Moreover, an additional UDP-glucosyltransferase was identified that may also act in LLO biosynthesis.

"Maturation" gene classifications, defined by genes in ER-Golgi maturation and secretion, were highly abundant in the 512 gene results based on KOG analysis. The phenotype penetrance caused by tunicamycin on the s/f (surface antigenicity abnormal) RNAi control experiments demonstrates that maturation of nascent glycoproteins can influence the phenotype outcome of defects in N- glycosylation in C. elegans. Likewise, many lectins may act in intracellular systems necessary for the transport of immature glycoproteins through the ER-

Golgi secretory pathway. Additionally, components of the Golgi glycosylation machinery responsible for the redistribution and proper localization of glycosylation enzymes were found in the results. For example, a subunit of the coatamer complex protein (COPI) which is involved in retrograde transport of

Golgi enzymes was found to be hypersensitive to tunicamycin. RNAi of this gene in the presence of tunicamycin caused phenotypes in all 4 observable wells whereas in the absence of drug, only phenotypes in 2 wells were scored.

A second example tunicamycin-hypersensitivity in Golgi trafficking was found in the SNAP-25 (synaptosome-associated protein) component of the

SNARE complex. The SNARE complex mediates fusion of Golgi transport vesicles and may interact with COPI vesicles directly during intercistemal Golgi transport (Ungar, Oka et al. 2006). RNAi of the SNAP-25 component with tunicamycin caused embryonic lethality and sterile progeny and produced no phenotypes in the absence of drug treatment.

69 Currently the remarkable variability of CDG diseases is generally believed to be a consequence of insufficient glycosylation among many glycoproteins in different cells. In this study evidence shows that variable pathology may also be caused by interaction between the primary CDG-I etiologic defect and modifier loci in the genetic background. Also this is a demonstration of a novel approach that systematically dissects the complex glycosylation phenotype into specific glycan-dependent proteins and functions. As a result the conserved

mechanisms that underlie the type of pathology observed in type I congenital disorders of glycosylation and identify genes that contribute to the severity of the disease are successfully delineated.

70 CHAPTER 6

COMPARATIVE A/-GLYCAN STRUCTURAL ANALYSIS OF THE C. ELEGANS N2 (BRISTOL) AND NF299 cogc-1(k179) STRAINS

The Golgi apparatus serves not only as the hub of the secretory pathway, but is vital in the glycosylation and trafficking of glycoproteins, glycolipids and proteoglycans. Glycosylation of proteins in the Golgi accounts for the diversification of AMinked glycans and the synthesis of O-linked glycans.

Transcriptional control of glycosylation enzymes involved in synthesis, modification and catabolism are believed to be the driving force in glycan diversity and abundance (Nairn, York et al. 2008). However, these studies only confirm the major glycoforms and simply confirm the correlation between glycan structures and the transcripts that encode the relevant enzymes. The fine details and changes in glycan structure cannot be accounted for by transcriptional

regulation of glycosyltransferases or glycosidases alone. It is known that sequential modifications of glycoproteins by glycosyltransferases also depends on the precise nonuniform steady-state distribution of resident glycosylation enzymes within the Golgi (Ungar, Oka et al. 2006).

Over thirty years ago Whaley stated, "No cellular organelle has been the subject of as many, as long lasting or as diverse polemics as the Golgi apparatus" (Whaley 1975). Still this intracellular membrane stack remains an

71 enigma and how proteins traverse through this pathway remains to be answered.

As discussed in Chapter 1, two models of Golgi morphology exist, both of which depend on the localization of glycosidases, glycosyltransferases and other resident proteins. In both cases, the Golgi membrane system undergoes continuous changes but its influence on glycan structure is uncertain

(Shestakova, Zolov et al. 2006). The various cistemae of the Golgi maintain different balances of enzymes and transporters amid constant movement of incoming and maturing cargo. Responsible for the movement of cargo vesicles are a group of multisubunit protein assemblies termed 'tethering complexes'

(Munson and Hughson 2002; Whyte and Munro 2002; Dong, Hutagalung et al.

2005). Among these complexes is the Conserved Oligomeric Golgi complex

(COG), what is strongly implicated in retrograde trafficking of COP I vesicles within the Golgi (Cavanaugh, Chen et al. 2007).

The conserved oligomeric Golgi complex is an eight subunit protein consisting of two lobes: COG 1 -4 form lobe A and COG 5-8 form lobe B (Ungar,

Oka et al. 2002; Loh and Hong 2004). The COG complex is vital for normal

Golgi morphology and structure. It is believed that COG complexes are involved in membrane trafficking and/or compartment function (Ungar, Oka et al. 2002).

Defects in COG function can cause abnormalities in intracellular protein sorting, protein secretion, cell growth and glycoconjugate synthesis (Oka, Vasile et al.

2005). Investigations of COG began in 1981 with the identification of low-density lipoprotein receptor (LDLR) deficient CHO cells mutants (Krieger, Brown et al.

1981). In this study, the mutant CHO cells showed changes in LDLR electrophoretic mobility that was attributed to the absence of glycans. The

72 mutation in the LDLR-deficient CHO cells were later found to be defects that occurred in subunit 1 and 2 in lobe A of the COG complex and resulted in alterations of cell surface glycoconjugates (Kingsley, Kozarsky et al. 1986; Reddy and Krieger 1989). These CHO cells (IdlB and IdlQ lack galactose and sialic acid residues on the /V-glycan terminus and show heterogeneity in sialic acid content on O-glycans. Additionally, the mutant cells also displayed dilated

Golgi cisternae. Abnormal glycosylation attributed to Golgi function have also been reported in yeast mutants with COG defects (Whyte and Munro 2001;

Conde, Pablo et al. 2003; Corbacho, Olivero et al. 2004). Deletions of any subunit in lobe A has severe effects in yeast growth, whereas deletions in subunits within lobe B have only mild consequences (Whyte and Munro 2001;

Ram, Li et al. 2002). Despite growing understanding of the COG complex, the overall implications involved in glycosylation remain unknown. That is to say that structural analysis of glycoconjugates of COG deficient cells has yet to be explored.

Mutations of the COG complex subunit 7 and 8 cause a congenital disorder of glycosylation type lie and llh in humans (Wu, Steet et al. 2004;

Spaapen, Bakker et al. 2005). Additionally, mutations in subunit 1 also cause a

CDG type ll/COGC-1 illnesses (Foulquier, Vasile et al. 2006). The first patient diagnosed to date with CDG-ll/COGC-1 suffers from hypotonia, ventricular

hypertrophy, growth retardation, progressive microcephaly and failure to thrive.

Mutations in the COG subunits impair the integrity of the whole complex and lead to disruption of Golgi trafficking and glycosylation. Since all COG-based CDG disorders ultimately alter Golgi function through a single complex, it would be

73 expected that the disease would manifest consistently form case to case. In fact similarly to CDG-I, clinical presentation is very heterogeneous. The COG complex has been studied considerably at the cellular level, but its role in animal development remains unknown. As such, it is important to understand the global impact that the COG complex has on N- and O-glycans throughout the body.

Recently, a C. elegans strain containing a mutation in the first subunit of the COG complex was created using a Tc1 transposon (Kubota, Sano et al.

2006). In C. elegans, the COG complex acts in the glycosylation of MIG-17, a member of the ADAM family of proteases that is required for gonadogenesis.

Worms lacking the cogcA gene show altered gonad morphology due to misdirection of the distal tip. RNAi knockdown of all 8 subunits of the COG complex in C. elegans also trigger gonadal migration defects suggesting a common pathway that regulates gonad development (Kubota and Nishiwaki

2006). Nishiwaki et al. prepared a model for the role of the COG complex- dependent glycosylation of MIG-17 ADAM protease. In brief, the COG complex is required for sorting/stabilizing the MIG-23/Golgi NDPase along the body wall muscles. As a result, MIG-23/Golgi NDPase hydrolyzes/ UDPs to UMPs. UMP export from the Golgi is vital for the higher energy UDP-sugar to be transported into the Golgi. The UDP-sugar is required for MIG-17 glycosylation. Therefore the COG complex is indirectly required for glycosylation of MIG-17 (Kubota and

Nishiwaki 2006). It was also observed that the cogc-1(k179::Tc1) mutants exhibit more robust phenotypes than mig-17(k113) mutants suggesting that glycoproteins other than MIG-17 are affected by the COG complex.

74 All eight mammalian COG components are conserved in C. elegans

(Kubota and Nishiwaki 2006). The C. elegans COGC-1 deficient strain presents an opportunity to study COG function in glycosylation. Currently, C. elegans is the only viable model organism lacking COG expression. Although defects in

COG-1, COG-7 and COG-8 were found to affect the trafficking of glycosylation machinery, the molecular mechanisms linking glycosylation defects to clinical presentations remain to be elucidated.

6.1 Materials and Methods

N2 (Bristol) and NF299 cogc-1 (k179) strains used were obtained from the

Caenorhabditis Genetic Center at the University of Minnesota. Mixed stage N2 and NF299 C. elegans strains were separately grown on large trays (6 trays per strain) containing peptone rich nematode growth media (NGM) and a bacterial food source of the E. coll OP50 strain as described (Hope 1999). Prior to the clearing of the bacterial lawns, 3 aliquots of 250 ml_ cold distilled de-ionized

(DDI) water was used to rinse animals from the tray. The worm solutions were centrifuged at 800 RCF (relative centrifugal force) for 4 minutes at 4°C. The supernatant was removed and the worms were washed with cold (DDI) water, spun at 1000 RCF for 60 seconds at 4°C until the supernatant was clear. The worm pellet was re-suspended in 20 ml_ cold DDI water. To this solution, 30 ml of cold 50% sucrose solution was added and mixed by inversion. Each tube was centrifuged at 50 g for 5 minutes at 4°C. The top layer of the solution, containing live animals, were aspirated and collected in a fresh tube containing 20 ml_ cold

75 DDI water. Tubes containing live sample were centrifuged at 1000 g for 2 minutes at 4°C. This step was repeated once with DDI water and once with M9 solution (Ausubel 1999) to remove excess sucrose. The worm solution was centrifuged again at 1000 g for 2 minutes at 4°C and the supernatant was

removed. The pellet was re-suspended in 5 volumes of M9, agitated at room temperature to allow the gut contents to be digested. A final series of three DDI water washes was performed to remove the gut contents.

Roughly, 1g of wet weight worms (~1.5 ml_), thawed from a frozen pellet, was added to 1.5 ml_ CHAPS lysis buffer containing 4% CHAPS, 8M urea, 40 mM Tris-HCL pH=8, 65 mM dithiothreitol (DTT). Worms were lysed using a

Pressure Biosciences Barocycler NEP2320 (South Easton, MA) with 20 cycles, each consisting of 40 seconds at 35,000 PSI, followed by 10 seconds at ambient pressure. Whole lysate protein concentration was quantified via Bradford assay

(Pierce, Rockford, IL) using a UV-Viz spectrophotometer (Shimadzu Corporation,

Kyoto, Japan). Sufficient lysate to yield 5 mg protein was transferred to a glass culture tube (13 x 100mm) and protein was precipitated using 15% trichloroacetic acid (TCA) for one hour on ice. Following precipitation, the sample was centrifuged at low speed for 15 minutes. After the supernatant was discarded, the protein pellet was washed and vortexed with 2 x 1 ml_ cold acetone, and 2 x 1

ml chloroform : methanol : water (10:10:1), with centrifugation between each wash. The protein pellet was dried under N2 flow and lyophilized overnight.

Prior to A/-glycan release, the protein pellet was vacuum-dried over P205 for 48 hours. 500 u.L of anhydrous hydrazine (Sigma-Aldrich, St. Louis, MO) was added to the dried protein pellet and heated at 100°C for 6 hours. Samples were

76 cooled to room temperature at which time the hydrazine was removed under N2 flow. Samples were then lyophilized overnight. The released glycans were re-A/- acetylated using 150 u.L acetic anhydride in 300 pL of saturated sodium bicarbonate on ice for 30 minutes, followed by 1 hour at room temperature. The samples were desalted on a DOWEX 50 W X8-400 cation-exchange resin

(Sigma-Aldrich, St. Louis, MO). Free A/-glycans were purified using a hand- packed column containing 500 u.L of medium fibrous cellulose. The column was washed with 4:1:1 n-butanol: ethanol: water to remove the peptide components.

A/-glycans were eluted with 1:1 ethanol: water. AMinked glycans were reduced with a 200 pL solution of 10 mg/mL sodium borohydride (NaBH4) in 0.01 M

NaOH for 1 hour on ice followed by room temperature overnight. To stop the reduction reaction, glacial acetic acid was added drop-wise until the sample pH =

5. Borate was removed by coevaporation of 3 ml_ ethanol, 2 X 2mL of 1% acetic acid in methanol, and 2 X 2 ml_ toluene. The reduced A/-glycans were dissolved in water and desalted on a pre-conditioned 1.5 ml_ column of porous graphitized carbon (Alltech Associates, Deerfield, IL). After washing the column with 3 ml_ of water, A/-glycans were eluted using 3 mL of 25% acetonitrile/water. Purified N- glycans in solution were dried in a SpeedVac® Concentrator (Savant

Instruments, Holbrook, NY).

After reduction, A/-glycans were permethylated as described (Kerek 1984).

Briefly, dried glycans were dissolved in 500 uL of dimethyl sulfoxide (DMSO)

(HPLC grade, Sigma-Aldrich, St. Louis, MO), followed by addition of powdered sodium hydroxide (99.999%, Sigma-Aldrich, St. Louis, MO). After vortexing, 100 mL of iodo-methane (99.9%, Sigma-Aldrich, St. Louis. MO) was added. Sample

77 tubes were flushed with argon, capped and wrapped in aluminum floil and vortexed for 1 hour. Samples were placed on ice and 1 ml_ of water was added to each sample to stop the reactions followed by 1 mL of HPLC grade dichloromethane. The sample mixtures were vortexed and the bottom organic phase (containing A/-glycan) was removed and put into a clean test tube. The dichloromethane addition was repeated twice and each extract was pooled. The new tube containing the organic phase was washed 5 times with HPLC grade water, removing the water layer after each wash. The dichloromethane was evaporated under N2flow, lyophilized overnight and stored at -20°C.

Both non-reduced and reduced A/-glycans spectra were generated using a

MALDI-CFR mass spectrometer (Kratos-Shimadzu Analytical, Manchester, UK) equipped with an ultraviolet 337 nm wavelength nitrogen laser. In the case of non-reduced samples, 100 pL of HPLC grade water was added and vortexed.

One pL of reconstituted sample solution was mixed with 1 pL of 2,5-dihydroxyl benzoic acid matrix (12 mg/mL in 50% (v/v) acetonitrile aqueous solution) on a stainless steel MALDI target plate. For methylated samples, 75 pL of methanol and 25 pL was used in place of water alone. All samples were ionized with 50-

60% of the maximum power while rastering on the plate surface. Mass spectra were acquired by averaging -200 profiles generated by 5 laser shots per profile with the post extraction parameter optimized at m/z 2500. Spectra processing was performed with Kratos Launchpad software.

Spectral peak annotation was performed initially by using the online program Glycomod Tool, a package under the ExPASY (Export Protein Analysis

System) proteomics server of the Swiss Institute of Bioinformatics (Gasteiger,

78 Gattiker et al. 2003). GlycoMod can predict the possible oligosaccharide compositions from their experimentally determined masses and on compositional constraints applied by the user (i.e. mass tolerance).

MSn A/-glycan mass spectra were obtained with a linear ion trap LTQ mass spectrometer (ThermoFinnigan, San Jose, CA) equipped with a TriVersa

Nanomate® automated nanoelectrospray ion source supported with automation and related computational tools. Dried samples were diluted in 150 u.L methanol and 50 u.L water. Samples were infused (10 uL) at a low flow rate of 0.30 uL/min and the spectra were collected using Xcaliber 2.0 software (ThermoFinnigan,

San Jose, CA). Signal averaging was accomplished by adjusting the number of microscans within each scan, generally ranging between 3-5. Collision parameters were left at default values with normalized collision energy set to

35%. Activaiton Q was set at 0.25, and activation time for 30 ms. All ions were sodium adducts.

6.2 Results and Discussion

The consequence of aberrant glycan structures, specifically in CDG patients, is not known. The human glycome is equally indefinite which complicates our understanding of carbohydrates in human development and disease. Current glycobiology efforts need to be two fold. Firstly, like the human genome sequencing initiative, the human proteome and glycome need to be precedent. This is an infinitely more complex challenge than the genome due to the nature of posttranslational modifications, which is both time and tissue

79 dependent. The second effort is more fundamental and perhaps an easier task, which should aim to understand the cellular biosynthesis and structure of glycoconjugates. Comparatively glycan microheterogenieity observed in metazoan cells is loosely defined by a small set of glycosylation enzymes. The complexity of glycan structure (topology, linkage, anomericity and monomer identification) is in most cases, assumed. Assuming carbohydrate structure could have dramatic effects in the future of functional glycomics as well in clinical treatment for glycosylation deficient diseases. To tackle the former goal, glycobiologist need to define a strategy for comprehensive carbohydrate structural analysis and establish it as fact. In this study, an established mass spectrometry-based method (Reinhold, Reinhold et al. 1996; Sheeley and

Reinhold 1998; Ashline, Singh et al. 2005; Hanneman, Rosa et al. 2006) is utilized to understand the global glycomic consequence of a CDG-ll/COGC-1 condition modeled in C. elegans.

Understanding the processes involved in the A/-glycosylation pathway is essential to understand the biochemical basis of known CDG as well as of types still to be discovered. In discovering the functions of glycan structures as modulators of glycoprotein activity, comparative analyses of mutated genes that target glycosylation enzymes have been the most useful. Null mutations of glycosyltransferases, glycosidases and sugar-nucleotide transporters have provided extensive information contributing to structural diversity, but the early conjecture of one gene, one enzyme is exceedingly unlikely. That questons how so many glycoforms exist when the biology seems to be so limited. Most likely, the answer lies in the complexity of the Golgi.

80 To resolve this problem, a comparative analysis was performed with a C. elegans mutant that not only modeled a CDG-ll/COGC-1 illness, but also investigated the effects of a component of Golgi transport. A comparative analysis in C. elegans enables this investigation to examine the complete glycome and thus establish a deficient-function to structure relationship not possible in higher model organisms. The investigation is a two pronged approach: (1) identify and determine the prevalence of each glycan composition in each sample's glycome (N2 and NF299); and (2) characterize the structure of specific compositions to determine any glycosylation patterns unique or absent in the mutant strain. This systematic approach will establish a global glycome pattern that is unique to COG deficiency and will elucidate aberrantly expressed structures observed in CDG-II.

6.2.1 Carbohydrate Structural Analysis through Mass Spectrometry

The initial step towards investigating alterations in glycosylation is identifying the total glycome (i.e. glycan compositions) in each sample. This can be performed by MS profiling of native and permethylated samples on a MALDI-

TOF instrument to determine molecular compositions. MALDI-TOF-MS is a useful tool that requires minimal sample and is highly sensitive, accurate and reproducible (Harvey 1999; Wada, Azadi et al. 2007). Permethylation derivitization of glycans is a useful step that not only aids in purification, but also enhances sensitivity up to an order of magnitude and offers predictable fragmentation patterning in both tandem mass spectrometry (MS/MS) and

81 sequential mass spectrometry (MSn), which is described below. (Hakomori 1964;

Reinhold, Reinhold et al. 1996; Ciucanu and Costello 2003).

MALDI-TOF MS is used not to define structure, but as a guide to establish the presence of compositions or even patterns of glycan compositions. In a comparative analysis, the ideal goal is to discover structures that are unique to a particular sample. This may be observed as a complete absence of a molecular composition (i.e. a missing peak in one sample and its presence in the other) or as a structural isomer that may be as subtle as a change in topology or linkage.

In either case, the application of both MALDI-TOF and electrospray ionization-ion trap (ESI-IT) mass spectrometry is a key dual technique for glycan sequencing.

ESI-IT provides the sensitivity and specificity required for carbohydrate structural mining.

Sequential mass spectrometry with an ion trap enables a robust and powerful method for oligosaccharide analysis not capable by tandem MS/MS alone. This method, using gas-phase isolation coupled with and multiple rounds of collision induced dissociation (CID), requires no chromatographic separation and is an unexcelled instrument for structural elucidation of complex oligosaccharides (Reinhold, Reinhold et al. 1995; Reinhold, Reinhold et al. 1996;

Reinhold and Sheeley 1998; Sheeley and Reinhold 1998; Ashline, Singh et al.

2005; Hanneman, Rosa et al. 2006). The strength of MSn analysis of permethylated glycans is its ability to track precursor masses through sequential

CID disassembly steps. Permethylation is crucial for disassembly because it converts all native hydroxyl functional groups into methyl ethers (Kerek 1984;

Ciucanu and Costello 2003). Principally, upon disassembly, the methylated

82 product ions leave "scars", or an hydroxide, at the point of cleavage that is helpful to assign glycosidic linkages.

6.2.2 Native Molecular Compositional Analysis via MALDI-TOF MS

MALDI-TOF-MS profiling of native-reduced N2 wild-type samples confirmed the presence of all /V-glycans previously reported released via hydrazinolysis (Hanneman, Rosa et al. 2006; Paschinger, Gutternigg et al. 2008).

Firstly, glycans were released by hydrazinolysis, reduced and profiled in their native species (Appendix B). This step is a comprehensive carbohydrate analysis by chemically releasing those /V-linked core modified structures that prevent endoglycosidase action. C. elegans express endogenous O-methylated glycans whose signature is transparent by the methylation protocols used.

However, such structures can be identified by a CD3-methylation. The biological function of glycan O-methylation is not completely understood, but it is known that O-methylated glycans act as a highly specific targets of host antibody recognition in the parasitic helminth Toxocara canis (Schabussova, Amer et al.

2007).

MALDI-TOF profiling of native /V-glycans from N2 and NF299 confirm the presence of /V-glycan compositions, with the predominant peaks representing pauci- and oligomannose structures (Figure16). This analysis was performed with the aid of the search tool Glycomod, which matches molecular ion masses

(m/z) to oligosaccharide compositions. Although the various fucosylated and truncated complex structures were present, their intensity levels were low.

83 100

90- N2 Bristol

1565.38

1177.19 1551.33 117E19 T74J3.15 1711.98 i ilk iLLLu, I 2020.87 2143.36 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 IVbss/Charge

100.

90; NF299 {cogc-1)k179

1243.31

B28P.20

93|15 K itrtjlit/i. I 1860.19 I 207D.72 1900 2000 2100 2200

Figure 16. MALDI-TOF profiles of native /V-linked glycans released by hydrazinolysis (1 of 3). (see Table C1 for molecular compositions). Peak m/z 1015 is a contaminant. Currently, there is no established standard approach for carbohydrate quantification, however a multi-institutional study has been completed in an effort to meet the demand for mass spectrometry based analysis (Wada, Azadi et al.

2007). Likewise, in this study, comparative quantitative analysis consisted of determining the percent composition of each molecular composition in a single spectrum. In short, the percent composition for each glycoform was defined as the intensity of each individual molecular composition m/z values normalized to the intensity sum of all the glycans ions between the appropriate mass-to-charge range in both samples (Appendix C). This breakdown allows for individual, as well as /V-glycan types, to be compared as well as to identify glycosylation patterns distinct to N2 or NF299. For example the relative abundance of endogenous 0-methylation was less in the NF299 cogc-1(k179) strain (Table 7).

Moreover, high-mannose structures are slightly more abundant in the mutant when compared to N2 Bristol.

85 N2 NF299 Molecular % /V-glycan % /V-glycan [m + Na]+ Composition Composition Composition

N2H2F2 -OMe 1079.8 1.22±0.23 1.27±0.30

N2H3F2 -OMe 1241.8 1.45±0.29 1.06±0.31

N2H4F -Ome 1257.8 3.31±0.81 1.02±0.34

N2H3F3 -OMe 1387.8 1.04±0.49 0.74±0.31

N2H3F3 -Ome2 1401.7 0.91±1.43 0.59±0.35

N2H4F2 -OMe 1403.8 3.58±0.91 0.98±0.31

N2H4F2 -Ome2 1417.8 1.49±0.52 0.74±0.30

N2H5F, -OMe 1419.8 1.71 ±0.21 0.69±0.28

N2H3F4 -OMe 1533.7 0.67±0.25 0.60±0.27

N2H4F3 -Ome 1548.7 0.76±0.47 0.56±0.32

N2H4F3 -Ome2 1563.5 1.15+0.51 0.66±0.31

N2H5F2 -OMe 1565.7 2.55±0.70 0.84±0.37

N2H4F4 -Ome 1695.7 0.87±0.23 0.57±0.27

N2H4F4 -Ome2 1709.7 0.82±0.42 0.64±0.37

N2H5F3 -Ome 1711.7 2.12±0.58 0.66±0.37

N2H5F4 -Ome 1857.7 1.10±0.28 0.56±0.34

N2H6F3 -Ome 1873.7 1.40±0.37 0.62±0.28

N2H6F4 -Ome 2019.6 0.84±0.12 0.50±0.22

N2H7F3 -Ome 2035.4 0.70±0.23 0.57±0.30

N2H7F4 -Ome 2183.0 0.56±0.49 0.64+0.44 % of the total 28.24±0.43 14.53±0.32 glycome

Table 7. Relative quantities of native O-methylated A/-linked glycans from N2 and NF299 strians. Molecular compositions along with the theoretical mass of the singly charged sodium adducts are shown. The first number under % composition represents the relative intensity to the sum of all measured glycan intensities and the second number is the standard deviation. The percent of the total glycome that represent O-methylated structures is also provided

86 6.2.3 Permthvlated Molecular Compositional Analysis via MALDI-TOF MS

After native analysis, A/-glycans were permethylated and profiled (Figure

17). Similarily to native glycan analysis, the search tool Glycomod was used to match molecular ion masses (m/z) to oligosaccharide compositions. The search parameters were strict; the experimental values were compared to hypothetical mass values within a ±0.3 Dalton threshold. After the presence of A/-glycans was confirmed, the experiment was repeated twice from worm growth to MS analysis maintaining all experimental constraints (Appendix B). The two replicates confirmed the molecular compositions of hydrazine-released A/-glycans form C. elegans as well as the profile trends seen in the first spectra.

As in the native compositional analysis, the relative abundances of paucimannose, high mannose and fucosylated structures were calculated by dividing each molecular composition intensity by the intensity sum of all the glycans (Table 8). Phosphorylcholine-rich and truncated complex structures were not included in the comparative because they were not detected in significant amounts by previous hydrazinolysis-based reports (Hanneman, Rosa et al. 2006) and only represent a minority of molecular compositions in the C.

elegans glycome respectively (Paschinger, Guttemigg et al. 2008).

87 N2S«i ;G~i5-0? Reduced Perr.iwtft^sled

Data. N? 6ft? 10-ia-O? ReaucgaPemolRslsHedCOOl.Fl i5 0«2Q[j? 13,13 Cai P3w0 lGOet2G0J 13:51 KratosPCA>imaCfRp!gsV2 3

N,H4F3 ,„• N2H4F2

N2H5F NLHL

MAJUJUMHV^A^UU 2200 2300

ogcl 5tf> 10-15-07 Reduced Permsihylated

Data coge-1 50s i 0-1 s-07 Reduced Perme!h*1*!i;dG0SJ1 F 3 IS Oct 2002 12 13G=.rOav1<3 10 Get 2007 13 51 KiatasPCftsimiteFf?f»lusV2 3.4 Modereaeeti'OT.^ower M.pext ®1?8Diftiri1i1) %it>t 2SmV(sgm=51SaHiV] Profits 1 -200 Smooth S* 6 -S-*setine 1 GQ NL299 cogc-1(k179) NJi,F.

NAF, N2H5F2 N,H5F N,H,F3 N2H7F2 7^6 N,HJF, 2/6 2 N2H6F , N H N,H7 NHF / 29 N22M797FA3 N-vr-12H1(0 ! jyyj JUAMOJ k^jUuJWI»*AJ>M^|lWWW\JMVVAillS«A»>w««

Figure 17. MALDI-TOF profiles of permethylated /V-glycans. The spectra mass ranges from 1700-2750 m/z. Molecular compositions that appear to be absent in the NF299 strain are marked with an * and represent composistions with 4 fucose residues. N=HexNAc (e.g. N-acetylglucosamine, N-acetylgalactosamine), H=Hexose (e.g. galactose, mannose, glucose), F=Deoxyhexose (e.g. fucose)

88 N2 NF299 Molecular % AA-glycan % JV-glycan [m + Na]+ [m + 2Na]2+ Composition Composition Composition

N2H3 1187.62 605.31 7.58±2.96 8.12±1.87

N2H2F2 1331.69 677.35 1.21 ±0.24 2.03±0.46

N2H3F 1361.70 692.35 5.92±2.50 8.06±2.36

N2H4 1391.71 707.36 5.10±1.75 8.95±1.35 N3H3 1432.74 727.87 1.75±0.27 2.46+0.75

N2H3F2 1535.79 779.40 3.01 ±0.88 4.89±1.48

N2H4F 1565.80 794.40 8.13±2.75 7.82±1.95

N2H5 1595.81 809.41 7.41 ±1.51 10.70±2.61

N2H3F3 1709.88 866.44 1.47±0.36 1.72±0.54

N2H4F2 1739.89 881.45 7.09±2.14 8.18±0.82

N2H5F 1769.90 896.45 3.80+1.48 4.04±0.85

N2H6 1799.91 911.46 2.62±0.18 2.69±0.41

N2H4F3 1913.98 968.49 4.28±0.78 3.82+1.52

N2H5F2 1943.99 983.50 6.47±1.07 6.12±1.73

N2H6F 2147.99 1085.50 1.09±0.27 1.21 ±0.40

N2H7 2004.10 1013.55 1.89±1.02 2.06±0.77

N2H4F4 2088.07 1055.54 1.48±0.37 0.49±0.13

N2H5F3 2118.08 1070.54 7.37±3.11 3.53±2.10

N2H6F2 2148.09 1085.55 2.44±1.10 2.20±0.67

N2H8 2208.11 1115.56 2.08±1.17 2.51+1.10

N2H5F4 2292.17 1157.59 3.15±1.43 0.37±0.10

N2H6F3 2322.18 1172.59 4.83+2.48 1.85±1.24

N2H7F2 2352.28 1187.64 0.99±0.36 0.58±0.26

N2H9 2412.21 1217.61 2.67±1.60 3.92±1.92

N2H6F4 2496.27 1259.64 3.36±1.43 0.31 ±0.11

N2H7F3 2526.28 1274.64 1.13±0.62 0.47±0.25

N2H10 2616.31 1319.66 0.60±0.36 0.65±0.26

N2H7F4 2700.37 1361.69 1.08±0.38 0.26±0.12

Table 8. Relative quantities of permethylated AZ-linked glycans from N2 and NF299 strians. Molecular compositions along with the theoretical mass of the single and double charged sodium adducts are provided. The first number under % composition represents the relative intensity to the sum of all measured glycan intensities and the second number is the standard deviation. Tetra-fucosylated compositions, which are significantly reduced in the mutant strain, are highlighted in grey. N=HexNAc (e.g. N-acetylglucosamine, N-acetylgalactosamine), H=Hexose (e.g. galactose, mannose, glucose), F=Deoxyhexose (e.g. fucose)

89 Mass spectrometric analysis of permethylated oligosaccharides showed an absence of tetra-fucosylated glycans as well as a substantial decrease in tri- fucosylated glycans (Figure 18). Comparatively, the frequency of mono- and di- fucosylated compositions did not differ significantly between strains (Figure 19).

This qualitative change in glycan distribution (namely fucosylated structures) was not detected in the native-reduced mode and can possibly be due the permethylation step itself which also serves as a purification step. The AA-glycans which are present in low abundances in native profiles became more detectable as permethylated oligosaccharides. Additionally, all tetra-fucosylated compositions with 0, 1 or 2 endogenous O-methyl groups will be observed as a single composition (m/z) after permethylation. Permethylation will have converted all the remaining hydroxyl groups to methyl ethers, making endogenous O-methylation undetectable.

1% 9% 12% NFL'99 N-glv can Compositio N2 N-glycan Composition

' ' \40% ^High Mannose " \31% B High Mannose • 1-Fucose • 1-Fucose •—-—*^>" *,-^.V - 25% • 2-Fucose D 2-Fucose • 3-Fucose D 3-Fucose • 4-Fucose ••/ • 4-Fucose P-:/ 22% \ I 19% 22% Figure 18. The prevalence of /V-glycans by subtype. Relative abundances of each glycan type are compared illustrating the decrease of tetra- and tri-fucosylated structures in the NF299 strain. Additionally, the occurance of high-mannose compostions is elevated in the mutant.

90 Mono-fucosvlated Di-fucosvlated 12.0 10.0

N2H3F N2H4F N2H5F N2H6F N2H2F2 N2H3F2 N2H4F2 N2H5F2 N2H6F2 N2H7F2 Molecular Composition Molecular Composition

Tri-fucosvlated Tetra-f ucosvlated 6.0 -p o £ 5.0-r •D = 4.0 -- SI

™ 2.0-n

® 1 0-- N2H3F3 N2H4F3 N2H5F3 N2H6F3 N2H7F3 cc ^- 0.0-- N2H4F4 N2H5F4 N2H6F4 N2H7F4 Molecular Composition Molecular Composition

N2 NF299

Figure 19. Comparative analysis of fucosylated /V-glycans from C. elegans organized by subtype. The vertical axis represents the percent relative abundance of each composition the glycome. The horizontal axis represents the molecular compositions (N=N-acetylglucosamin, H=Hexose and F=Fucose). Error bars represent the coefficient of variation, standard deviation/mean x 100, which indicates the percent variance of molecular compositions peak intensities between the individual samples.

The molecular compositional analysis of both native and permethylated samples revealed four major trends of /V-linked glycosylation in the NF299 strain.

First, endogenous O-methylation, which is prevalent in wild-type C.elegans, is decreased in COGC-1 deficient animals. The percent of O-methylated structures

91 in the total glycome of N2 and NF299 is 28.24±0.43 and 14.53+0.32 respectively.

Second, in native and permethylated profiles, the occurance of high-mannose compositions is elevated in the NF299 strain. In the native profiling mode the percent of the total glycan composition in NF299 is 44.98±1.07%, compared to

35.66±1.17% in N2. After permethylation the percent of high-mannose in NF299 is 39.61 ±1.23 and 29.96±1.32 for N2. Third, tetra-fucosylated structures are to a large extent absent in NF299 strains. Tetra-f ucosylated compositions consist of

1.43±0.91% of the total permethylated NF299 glycome. Alternatively, the N2 glycome is 9.07±0.91% composed of tetra-fucosylated compositions. Fourth, tri- fucosylated structures are decreased in the NF299 strain. The amount of tri- fucosylated structures is 11.39±1.13% in NL229 and 19.07±1.47 in N2. The glycomic patterns at this stage are merely a useful tool to determine the fine structures present in the comparative anlysis. Clearly, the NF299 strain indicates a discrepancy in fucosylation, specifically fucose-rich structures, but the exact nature of the defect is unclear. Structural characterization by MSn of the N- glycans will help elucidate the specific glycosylation patterns seen in CDG- ll/COGC-1 type conditions.

6.2.4 Structural Characterization of A/-qlvcans via MS"

Single MS and MS/MS analysis of carbohydrates provides the basis for compositional assignments. Compositions however cannot define the fine structure of glycans, namely linkage and topology. Within a single peak of a MS1 profile, several isomers and isobars may exist. Isomers are defined as structures

92 with identical atomic compositions, but arranged in different stereo and structural configurations. A mass profile is transparent to all isomer types and to uncover such detail fragments must be generated, and sometimes multiple times. In the current study, a differential analysis of N2 Bristol and NF299 cogc-1(k179) worms was carried out by sequential MS for a comprehensive evaluation of structural detail.

The goal of this project was to determine the function of the COG complex and its role in glycosylation in C. elegans. As found in the MALDI-TOF profiling, the degree of fucosylation in the NF299 strain appears to be below normal or absent compared to the wild-type. Detailed glycan sequencing will answer three key questions: 1) is the COG complex required for synthesis of wild-type N- glycans; 2) does the NF299 strain assemble fucose-rich glycans (FUC4) and 3) is the location of fucosylation (i.e. topology) different than wild-type structures .

Optimistically, this report will shed light on the role of the COG complex in CDG-

ll/COGC-1 conditions and in doing so will be the first comparative analysis where a glycosylation structural defect is found that is not directly linked or caused by a defect in a glycosyltransferase, glycosidase or carbohydrate-nucleotide transporter.

6.2.5 Structural Characterization of GlcNAcpMans

To discern the pattern and extent of fucosylation, first N2 strains with

molecular compositions of zero, one, two, three and four fucose residues were examined via MSn disassembly to establish A/-glycan wild-type standards.

93 Additionally, only certain molecular compositions are present with these compositions. These glycans are GlcNAc2Man4Fuci.4, GlcNAc2Man5Fuci.4,

GlcNAc2Man6Fuci-4 and GlcNAc2Man7Fuci-4(the penultimate and reducing end

GlcNAc residues were components of these compositions). Since

GlcNAc2Man5Fuci-4 structures are present in higher abundances in both strains, it was used for comparative analyses. Also, higher levels of abundance are helpful for isolation, fragmentation and analysis of molecular ions by MSn techniques. Additionally, structural isomers are more prevalent in Man5 than in

Man6 compositions (Prien 2009). Limitations and low levels of detection in

Man7Fuci-4 compositions make Man5Fuci-4 more suitable for examination. After the N2 wild-type structures were established, structural elucidation and MS" of the NF299 strain was carried out.

The first composition to be discussed in the analysis is Man5GlcNAc2.

Generally, the structure of GlcNAc2Man5 is described as having two mannose residues on the 6-arm of the conserved chitobiose core structure (1,4-GlcNAc(31,4-GlcNAc) (Figure 20).

Until recently this was the only Man5 structure chitobiose believed to be formed (Magnelli, Cipollo et al. core

2008). Cipollo et al. detected one additional Figure 20. GlcNAc2Man5and isomer in the MS3 spectrum but its structural the chitobiose core. The chitobiose core structure is detail was not further analyzed. In this study the present in all A/-linked glycans. presence of 4 Man5 structural isomers was

94 demonstrated in the wild-type N2 and NF299 strains. The MS2 spectra of

Man5GlcNAc2 provides little structural information due to the facile neutral loss of the terminal GlcNAc residue providing an abundant product ion at m/z 1302

(Figure 21).

95 C. e/egans_N2 C. elegans_cogc-1 2 j MS m/z 1595 100 IMS2 m/z 1595 50 scans 1302.7 150 scans 1302.6 NL: 4.35E4 1084.5 / NL: 4.92E3 1075.6 /

1057.6 / 1145.6 8805 1075.6 1084.5 709.4 \ 1377.7 1377.6 1057.5 \ / 1145.5 T/T/ \ // / / ti ) , , 1 i U i «4, L„ 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 m/z mfc 1159 MeO

MeCh /„866 JMeO-i „.. HD >-OMe ' 0-<_^0Me MeO 't, i '• MeO N- S61 y"° : >o

I Isolation and fragmentation of m/z 1302

C. elegans_H2 C. elegans_cogc-1 MS3 m/z 1595 IMS'm/z 1595 —•1302--R ->1302-"R 1075.5 1145.6 ^50 scans 50 scans 1057.5 \ / NL: 7.68E3 NL: 1.06E3 1260.6 1057.5 709.4 866.4 8 5 737,4 857.4/ / 431.3 676.4 f 676.3 880.4 ' 51.9.4 66,4\™-8rte, 458.2 839.4 | 737.4 927.5 519.3 667.4 / / / . ,... )|| I ,\LU ,iy>..Ul. V 400 500 600 700 800 900 1000 1100 1200 1300 1400 400 500 600 700 800 900 1000 1100 1200 1300 1400 m/z m/z

MeCK 64S667 MeO-T^V --OA MeO OMe 7.37 Na* MeO -Of / 1075 1J45 MeCK OMe /1057 i MeO-(_ } J |709 ' | • MeO-i MeO / "7MeCH OMe 866 J[231J DMei .kj.260|

MeO OMe

3 Figure 21. MS of m/z 1302 from GlcNAc2Man5(/n/z1595). The top two profiles are 2 MS spectra of GlcNAc2Man5 from N2 Bristol and NF299 cogc-1 (k179) C. elegans strains. The bottom profiles are MS3 of the m/z 1302 ion, which is the neutral loss of the reducing end GlcNAc. Corresponding structures are shown below each profile set and all ions are sodium metal ion adducts.

96 3 + + From the MS profile of GlcNAc2Man5 (m/z 1595 -+ m/z 1302 ) two structural isomers are immediately apparent (Figure 21). In addition to the classical Man5 structure, a second structure was indicated by the presence of the m/z880 ion (Figure 22). It is improbable to generate an m/z880 fragmentation of the Man5 structure in Figure 21, but could suggest an isomeric structure having a terminal non-reducing disaccharide element like that below.

1377 445

3 Figure 22. GlcNAc2Man5 Structural Isomer #2. The MS profile of m/z 1302+ ion shows the presence of a structural isomer from the m/z 880 peak. The accurate terminal hexose linkages are not determined and are shown as 1,4 linked.

The Hex-Hex loss could occur from either the 6- or 3-position of the central mannose within the /V-linked core and this was supported with the detection of the m/z 533 and m/z 547 ions in the MS4 spectra (Figure 23). The

MS4 spectrum of m/z 880 confirms the presence of this isomer but only with further disassembly can one discern the linkage of the terminal Hex-Hex disaccharide. Both N2 and NF299 strains exhibit this structure in equal amounts based on abundance. The neutral loss of the Hex-Hex disaccharide appears to occur more readily on the 3-arm of the core, but quantification could not be

97 considered without knowledge of each ions' pathway and ionization cross section.

C. e/egans_N2 C. elegans_cogc-1 100q 635.4 100 MS"/77/z1595 MS4 m/z 1595 635.2 1302--R 1302-» —•880 oo- -•880 oo 653.4 723.4 100 scans 100 scans 653.2 723.3 NL: 2.19E1 662.4 NL: 7.07E-2 838.5 662.1 533.3 838. 4 809.5 4150 '-. 415.3 / 547.4 533 2 417.2 ,£.,, 848.2 4441 547.3 X4'5 25°5-2 1561.2 '458.3 / I / 579.2 463.3/ | / '605.2 / fU- 400 500 600 700 800 900 "400 500 600 700 800 900 m/z m/z

Fragmentation Assignments of m/z 880

662 458 463 MeO—j , / 653 MeO—{ V- i- \/5.3.3 723 MeO OMe MeO OMe: }\ mi 445J505 i-i l| 809 838 HO 0Me i| 635j 1h

4 4 Figure 23. MS of m/z 880 ion from Man5 Isomer #2. MS spectra and fragment assignments from the /7?/z880 ion confirms the presence of an additional Hex-Hex disaccharide from the m/z 445 and m/z 463 ions.

What is not evident in the MS3 spectra in both strains is the structure of the m/z 667 C-type ion. The topology of the m/z 667 ion is presumably a branched structure consisting of three mannose residues. However, when the m/z667 ion is fragmented from the MS4 spectra after the neutral loss of a terminal hexose

98 (m/z 1595-* 1302-* 1084—>-667), the fragment ions demonstrate the presence of an additional linear topological isomer (Figure 24). The m/z227, m/z519 and m/z 563 ions can only be created from a linear precursor structure. Although it does appear that the branched structure is more commonly formed in both strains based on the relatively more abundant m/z 301 and m/z 375 ions.

Although further MSn disassembly would be required to define linkage for the linear structure, the signal strength from the NF299 strain did not permit additional fragmentation. However, MS6 spectra has been generated from the N2 strain and the m/z 563 fragment suggests that the terminal and penultimate hexose residues are connected by a 1,2 and 1,3 linkage respectively and not 1,4 as illustrated below. (Appendix D; Figure D1).

99 C. elegans_N2 C. elegans_cogc-1 100 449.2 100 S 375.1 MS51595 MS m/z 1595 1302-.* 375.1 1302-R 431.2 1084> 431.2 1084O \ 667 301.1 667 / 5.J*0 scan s ,301.1 jj45 scans a 50 NL1.75 NL: 635.3 < 315.1 563.2 593.3 5.58E-1 329.0 563.3 \ 6^53 ra 9227.0 345.1 \ 649.2 259.3 3)99.0\ \

400 ^ 500

MeO-i 241! '/315\ h°s !i A. MeO OMe MeO—( ) :;0 \

MeO OMe : """' 449

5 Figure 24. MS spectra and fragmentation assignments of m/z 667 from Man5. Isolation and fragmentation of the m/z 667 C-type ion (three terminal mannose residues) after the neutral loss of one terminal hexose (m/z 1084 precursor) reveals the presence of three isomers based on the presence of the m/z 563 peak that can only be created from a linear structure.

A fourth Man5 structural isomer was detected from the neutral loss of three terminal and one branched hexose residues. This fragmentation pathway tracks the following ions: m/z 1595->m/z 1302^m/z 1084-»m/z 866^m/z 648. The

MS4 spectra from the m/z 1084 parent ion contains many peaks that can be formed from the neutral loss of the reducing end GlcNAc and one terminal

mannose residue from all three of the isomers above. Isolating the m/z866 peak from an additional neutral loss of a terminal hexose still produces fragment ions that can be generated from these isomers (Figure 25). So it is necessary to

isolate the m/z 648 ion and examine the core GlcNAc and Man residues.

100 Only one of the isomers described above has three terminal hexose

residues and can produce an m/z 648 ion in the MS5 spectra. MS6 spectra of this

ion illustrates not only the presence of the branched crt ,3 and a1,6 core mannose

residue (m/z 458), but additionally shows trace amounts of an crt,3, a1,6 and crt ,4 bisecting core mannose structure from the m/z AAA peak, which has not

previously been reported (Figure 26). Because the chitobiose core is considered to be conserved amoung all N-glycan structures, any implications this elaboration would have on glycoprotein structure or function is speculative. It is possible that the bisecting mannose is not a product of the LLO pathway, but is constructed in the Golgi during Akjlycan processing. The extent of core modifications that are

unique to C. elegans would suggest the latter scenario and perhaps the

existence of a core-specific crt ,4-mannosyltransferase (Figure 27).

101 C. elegans_cogc-1 927.4 C. elegans N2 100 q, 100 927.5 MS" m/z 1595 MS'm/z 1595 839.5 1302-"R 1302 -"R \ / 1042.5 -•1084 > —•1084C- „.. 866.5 1042.6 48 scans u 48 scans ,709 4 / 1013.6 NL1.12E2 NL: 1 38E2 , i , 649.4 I 737.5 857.6 \ 449.2 S21.3\ 667.4 5 50 449.2 4912 !50 765.4 / 445.2 \ "58.3 \ \ \ / 445.2 I 458.2 » I 783.5 < i : 431.2 \\ I 491.3 \ \ M 431.2 \ I I / 519.2 i ; 880.5 385.2 1/ 1 "wr i\ i / 0; , .^*,ui. i i.M...i^ij 1 1 . , ivt i, 300 400 500 U700 900 1000 1100 1100 m/z

MeO

737 Na-

MeOK )—f-0

1013| MeO -N-"i042| 1Isolatio n and fragmentation of m/z 866

C. elegans_N2 621.3 C. elegans_cogc-1

5 Figure 25. MS of m/z 866 from GlcNAc2Man5 (m/z 1595). The top two profiles 4 are MS of GlcNAc2Man5 after the neutral loss of the reducing GlcNAc and one terminal Hex residue (m/z 1084). The corresponding structure with fragments is shown. Isolation of the m/z 866 ion which is an additional neutral loss of a terminal Hex generates the bottom MS5 profiles. Additionally, each profile set and all ions are sodium metal ion adducts.

102 C. elegans_N2 C. elegan$_cogc-1 1QQMS°m/z 1595 1302-»R 1084 c- 866 0- 403.1 491.1 8 —•648 O- \ 458.3 / c 130 scans 421.3 I •S NL: 268.1 § : 1.77E-1 1444.0 5 50: l / 606.2 < 212.9 1 > 301.1 575.2 1 Jlat i / l \ cc I / \ \ 1 : J .,,.],- _rtj 0 J 1:, i Li ,,l > 1 !»,. L„ 450

Na+ Na+ MeO—I 44i HO—| ?13 301 421 494 MeO—/ \ 1 —O 1 301 421, 491, 458 \ "7 423 ' r\'-m \ / HO OMe MeO OMe \.:0 Vq \ i )-/-o HO—-y )— J MeO-t'' )\ i268: / N 606J HO OMe MeO HO N 606) Jo-

6 Figure 26. MS of /n/z 648 from GlcNAc2Man5 (m/z 1595). The top spectra 6 represent MS pathwayss Man5 GlcNAc2dissasembly following the neutral loss of the reducing GlcNAc and three terminal Hex residues. Two isomers are apparent at this stage and constitute the normal branched chitobiose core mannose as well as a bisecting chitobiose core mannose. Each profile set and all ions are sodium metal ion adducts.

Figure 27. Structural Isomers of GlcNAc2Man5 in N2 and NF299

103 6.2.6 Structural Characterization of GlcNAc?MansFuc

The second composition analyzed was GlcNAc2Man5Fuc which has a single sodium adduct mass of 1769.9. However when using the electrospray ion trap instrument, this molecular ion appears primarily as a double sodium adduct

(m/z8962+). The comparative analysis of the two strains aims to determine the aberration of fucosylation in the COGC mutant. The location of the fucose in the

N2 wild-type is known to vary throughout the structures, from position on the chitobiose core to mannose residues on the outer arm (Hanneman, Rosa et al.

2006). With only a single fucose in the composition, the possibility for structural variation between the strains is limited. The relative abundance of mono- fucosylated structures remains mostly equal, with the N2 and NF299 glycome

having 18.9 ± 1.8% and 21.2 ± 1.3% respectively.

2 MS profiles of GlcNAc2Man5Fuc provide some topological information with regard to fucose location. As stated previously, a common fragmentation

pattern observed in the MS2 profiles is a result of labile glycosidic bond cleavage

between the chitobiose residues. This generates primary B- and Y-

complementary ions that can help locate a single fucose (i.e. on the reducing end

GlcNAc or on a terminal hexose). Cross ring cleavage fragment ions (A- type) that are instrumental to distinguish isomers and linkage positions are exposed

only with smaller sized oligomers where the number of oscillators are limiting to

dissipate the collision energy.

The MS/MS profiles of N2 and NF299 strains differ slightly (Figure 28).

The B/Y ion complementation suggests three locations where fucose is added.

104 The three positions are based on the following ion fragment pairs: m/z 316/1476, m/z490/1302 and m/z694/1098. However, the NF299 strain only contains trace amounts of the m/z 316 and m/z 1476 ions, suggesting that outer arm fucosylation is decreased in the mutant. From the N2 data set it can be inferred that, under normal conditions, fucose is either added on; 1) a terminal hexose or less likely the penultimate GlcNAc, 2) the reducing end GlcNAc with no other modifications or 3) the reducing end GlcNAc with an additional hexose residue.

The MS2 spectra provide no additional structural information. It is crucial to note that in the figure below the cartoon structures do not represent the exact location of fucose or the accurate overall topology. In fact in N2 wild type C. elegans, the

GlcNAc2Man5Fuc composition includes 7 different topologies.

694.4 100 1098.6

J749.9 1 I 787.5 490.3 I | 853.5 316.2 880.5 1769.8 939.5 1302.6 ' 433.3) 667.4 n^

JLm4m#i), 400 600 800 1000 1200 1400 1600 1800 400 600 800 1000 1200 1400 1600 1800 m/z m/z C. e/egans_N2 MS28962* C. e/egans_cogc-1 MS2 8962* 50 scans NL: 2.04E2 45 scans NL: 3.00E1

2 2+ Figure 28. MS of GlcNAc2Man5Fuc (m/z 896 ). B/Y ion complementation from MS/MS profiles of m/z 8962+ reveals the relative location of fucose. The cartoon structures represent compositions only. Open circle, glactose; filled circle, mannose; filled triangle, fucose; filled square, /V-acetylglucosamine.

105 The first compositions to be evaluated comprise the m/z 694 and m/z

1098 B/Y ion complements. The most predominant peak in the MS3 spectra of the m/z 1098 ion is m/z 880, which is the neutral loss of a terminal hexose

(Figure 29). Surprisingly, unlike the GlcNAc2Man5 isomer that generated a similar m/z 880 peak corresponding to a branched structure (Figure 23), these spectra lack the indicative m/z 445 peak characteristic to a linear hex-hex disaccharide unit. The m/z 445 ion is usually the most common ion peak generated from a precursor that does exhibit such a structure. However, there are two peaks that can only be created from a linear branched structure. These are m/z 676, which is the Y-ion complement of the sodiated B-ion m/z 445 disaccharide loss, and m/z 431, which is formed from the neutral loss of a hex-

hex disaccharide and a terminal hexose. Additionally there are two peaks that correspond to an ion created from a cross ring cleavage that contains a linear

hex-hex unit. These two peaks are the m/z 635 and the m/z 533 peak. These ions can be generated from a different structural isomer, but from the spectra do not suggest the presence of isomeric structures.

106

|5o: 2 ; 4 3 4 2+

C.elegans_N2 u u 100 p §50 i > 880 ion,whichistheneutra l lossofaterminahexose. indicated atthetoplefonspectra.Thbotto m profilesareMSofthm/z profiles areMSafterthlosofGlcNAc/Fuc/Ga l reducingendstructure Figure 29.MSofm/z880fromGlcNAcMan Fuc (m/z896).Thetoptwo 25 2

300 Rela ance ; ^MS"m/z896*o 635.3 J 1098-SR\ 5333 D 400 5060 NL: 124E2 57 scans 327.1 \444.0' 32 3431 ..- i.....,,,,i,,Vv.L MS m/z896*9 NL: 9.86E2.I 50 scans871.4 C. elegansN2 I,- \«M 8534 7 »\\\I i \.I MeO- MeO- -•-880 o MeO OMe"«0°lfc 561.3]605.3 41?2 -•1098-SR 576 579.3 \f-I 6623 662 617.3 \-825.4 662 MeO—( 635.3 \ 500 MeO MeO- -0 OMe 579.3 700 80 / 723.4\ ,.| I.. :880 :-H-0 Mi / 600 m/z / \ 458 rn/z MeO-/' .6.05/ HO 676.3 653.3 533 431! 662.3 / ,x 700 -OMe; o . 4.17] 635. I 900 1001104050607080 880.4 / 653 941.5 1056. -0-! 1 809.4 I MeO— 1027.5 838.4 723 \ 7"° \, *_ 809 900 30 838 I 107 Isolation andfragmentatioofm/z880 3 50 100 C.elegans_cogc-1 100 3: 2 82 scans MS m/z896 ]MS* m/z896*2 NL: 3.94E2 C. elegans_cogc-1 431.2 579.3 NL9.10E1 327.1 >444.\ 106 scans 0 MeO ()—:-:-0 \ 343.2458. I i'•^ / *I505.31 MeO OMe:444 533.3 x 5333 -••880 o- 417.2- i 400 M«G—^ MeO—< MsD—( MeO—i MeO MeO— 1098-SR 635 MeO- 896' Q MftO 1098-SR<3 MeO 617.3 I662.825.5 VIM L -1/ 635.3 OMe \ Wj 1:662 OMe ; 500 6070 561.3 723.4 579.3 ,f \ /MeO—,.v / OMei;iweo*' 533 \m/ m/z 605.3 853i:871?' / m/z 635 .72? 853.4 \ 871.5 Na* 653.4 / 676.3 ,{<• / / 809.4 662.4 / / 941.5 1056.6 I 723.4 \ 1027.5 \ \ 838.4 900 Next, the m/z 880 ion in the MS3 spectra from the neutral loss of an non- defined terminal hexose was isolated and fragmented. The MS4 spectra created reveals inconsistencies with the proposed branched structure above. In addition to a branched isomer, a bisecting structure is also evident, similar to the bisecting

GlcNAc2Man5 structure described in the previous chapter. For example, the m/z

327 or m/z 343 peak cannot be created from a branched structure. The most abundant peak in the MS2 spectra is the m/z 662 ion that is caused by the neutral loss of a terminal hexose. The MS5 spectra created difinitively confirms the bisecting mannose structure from the m/zAAA peak (Figure 30). The hydroxyl group at the 4-position on this mannose is evident not only from the m/z AAA, but also from the m/z 301 peak which is absent in the branched isomer.

108 C. elegans_N2 100= C. elegans_cogc-1 MS5 m/z 8962*o ,4172 100 5 2 MS m/z896 *o 417.2 1098*. |/ 1098-*.. / 880 o- 444.2 880^ 444.2 I : -•662 o / 505.3 • 662 C- 505.2 620.3 I :101 scans 620 3 J108 scans |50=NL:1.51E1 435.2 :NL: 8.61 591.2 435.2 591.3 1 2 .1 ]301.2, = 3P - 361.2 ' I / 361.2 387.1 = ,'313.1 \ 387.2 '\ 458.2 \ i I 313.0 \ 458.2 j/H \ i I / ,'315.2\ n=WT'*T-^rVrX>v^^TlwJv-JiXj ,. ,...!> 350 450 550 650 350 JU 450 550 650 m/z

MOO—1 444

MeO (~ )—j-i 417 MeO OM8 315 505

Na+ [301 )-.-0 i--q J591 M ••£- I620 HO/ ;bMe 435] 361/387' /

5 2+ Figure 30. MS of m/z 662 from GlcNAc2Man5Fuc {m/z 896 ). After the neutral loss of the reducing GlcNAc/Fuc/Gal and two terminal Hex residues two structural isomers are confirmed in both strains. The m/z 458 and m/z AAA ions correspond to a branched and bisecting chitobiose core mannose.

The corresponding Y-ion of m/z 1098 is m/z 694 and corresponds to a

reducing end GlcNAc with a fucose and hexose addition. Previous reports identified the hexose as a terminal galactose attached in a (31,4 linkage to an

internal fucose (Hanneman, Rosa et al. 2006). In this report, a sample of

released glycan was treated with and without (31,4-galactosidase. The MS

profiles from treated sample lacked complexity at the core, namely terminal

hexose residues that were present in the untreated sample. It was concluded that the galactose residue was attached in p1,4 linkage to both a Fuccd ,6-

GlcNAc or FuccM ,3-GlcNAc reducing end structure. Both N2 and NF299 strains synthesize the (31,4Gal-Fuca1,6-GlcNAc structure and do not contain any

109 structural variations (Figure 31). Comparatively, the analysis of the m/z 694 and m/z 1098 B/Y complements are identical between the wild-type and mutant strains.

676.4 100 C. elegans_cogc-1 100 C. elegans_N2 3 2 I MS3 8962* ] MS m/z 896 * -•694 -•694 50 scans 662. 50 scans 662.4 • NL: 3.53E1 NL1.14E3 476.3 634.4 <| 50 476.3 <$50 ^ 547.3 I 519.3 \ 433.2 \ ?3 634.4 433.2 589.3 4 519.3 4 5 2 415.2 \ 58.3 589.3 \ \ - \ 458.3 < \ 602.3 1 302.2 \ 602.3 : / •\ v \ \ \ I V' \l 0%T- ,.l,,i.., I t. ,• IJ ,1 ,1 i.j.U.. 300 400 500 600 700 300 40C 500 600 700 m/z m/i

476 MeO—| ^ 433 -oi: r _cl MeO—( 1 —o / \_ |_0 < >— / 3.02\ "A ' MeO OMe MeO OMe 547 \ 458 602 \ ]_ -H20 , R7R -OMe 519: Na+ OH-'.-C-; /—c Me -MeOH _„ ;589 \Y:d,,. MeO H 0+Ac V O - ? »634

3 Figure 31. MS of m/z 694 from GlcNAc2Hex5Fuc (m/z 896^).

The second B-ion to be analyzed from the GlcNAc2Hex5Fuc is the m/z

1302 which is composed of hexose residues, presumably all mannose. Based on published reports on the C. elegans glycome, the two types of hexose

residues that are present on A/-glycans are either galactose linked to a fucose which is bound to the penultimate or reducing end A/-acetylglucosamine or

mannose residues added to the chitobiose core in several conformations. The

structural variation of the m/z 1302 structure contains five mannose residues and

110 the structural detail closely resembles the isomeric structures of GlcNAc2Man5

(Figure 32). MS2 spectra of N2 and NF299 show the most prevalent peak to be m/z 880 which corresponded to a branched structure. MS3 of m/z 880 from both strains do not exhibit any change in the distribution or intensitiy of peaks generated (Figure D2). Represented in trace amounts in the MS2 spectra is the

C-type m/z 667 ion peak that contained isomers #1 and #3 described in the previous chapter as having either a branched or linear tri-mannose structure

(Figure 33). The N2 and NF299 spectra of the m/z 667 from the m/z 1302 precursor are relatively identical with the exception of the m/z463 peak cooresponding to a C-type hex-hex ion in the the N2 spectrum. Moreover, the m/z 445 which is the the B-type hex-hex ion is more detectable in the N2 profile.

This may suggest that the linear structure is less abundant in the NF299 strain.

However, the indicative m/z 563 ion of a linear structure is present in relatively equal amounts between the two. It is importatnt to note that the signal strength of the m/z 667 ion in the MS2 of m/z 1302 in the NF299 strain is exceeding weak and required over 400 scans to generate a single acceptable spectrum. This may contribute to the absence or undetectable ions in the NF299 profiles. Most

importantly is that the MSn comparison of the two strains does not reveal different structural isomers. This includes the m/z 490 Y-fragment ions, which are the fucosylated reducing end GlcNAc structures (Figure D3). The abundance of strucutres may vary, but may or may not depend on the COG complex.

Ill 1084.5 C. elegans cogc-1 880.5 100n C. elegans N2 1145.6 100 MS3m/z8962* 880.5 MS3 m/z 8962* -•1302-4* 1057.5 -•1302-4* 126,0.6 c e 50 scans 52 scans 1084.6 -,260.6 NL: 5.33 NL: 4.62E-1 da r \ 1145.6 \ ndanc e

1029 6 0 1057.5 605.3 83g4 1231.6 463 1 \ K / 723.4 ;' \ Relativ e Ab u \ - \ , 723.3 \ \ / 533.3\ I \ \ 533.3 ' ! I \ / I I i ' ; W j |l i ,i !. 0 4 l.i..,.. .ill.,. ,J.J Jj I ,i, I 0- '.,;,! J LlJ ! ,1,7 .1-r ^ , I J '. 400 500 600 700 800 900 1000 1100 1200 1300 1400 400 500 600 700 800 900 1000 1100 1200 1300 1400 m/z m/z

Isolation and fragmentation of m/z 1084

C. elegans_N2 i 2 100 C. elegans_cogc-1 6624 1A 1042 5 °7MSWz 896 * 4 2< 0.5 P- - 1042.6 [MS m/z896 1302*) 662.3 1302-*R -•1084 o —• 1084o 839.3 199 scans X 1013.4 50 scans 857.5 j 3NL: 2.53E-1 NL: 2.90 \ i 50: 550 809.5 417.0 417.1 667.4 445.3 445.3 635.1 709.5 1052.7 I 737.2 463.3 463.3 ,505'2 \

-iii .. ...j.ii.u. Hi. I Mi.,. 700 . 800 900 1000 700 800 900 1000 1100 m/z m/z IIsolatio n and fragmentation of m/z 866 100 C. elegans_N2 100 C. elegans_cogc-1 639.3 662.3 824.4 5 2 :MS m/z896 * ]MS5 m/z 8962* 621.3 1302-in 1302-IR 1 1084O 1084 648.3 ! = -•866 » -•866 c /709.3 305 scans 3212 scans 444.3 150 NL7.63E-2 4443 . 150 3, '458.3 565 3 NL: 2.38E-1 795.3 268.1 449.2 463.2 547.3 286.6 417.1 J 463.1 341.1 "31.3 491.2 / I 329.2 \ / / 519.2 268-2 I 417.1 \ (519.3/591.3 ! 1358.0 ! .Wo] /315.0|40i.3\\ \ 7 I : l.ii.l 0]V i'il)|l)'> M M,i 300 300 400 500 600 700 800 900 m/z

648 MeO-

MeO- "n*!tf«J 533 Me0 <_>:7\./15g2l 709 MeO OMS V-•?.!-•? 5.1.9 MS i 7 /417: MeO-i / ,4?i 7°,. j: J-/-o o 7 / i 7*""" MeO-{ V-;-:-:-0 : HO OMe :

5 2+ Figure 32. MS of m/z 866 from GlcNAc2Man5Fuc (m/z 896 ). These profiles follow the loss of the fucosylated reducing end and two terminal hexose losses. The MS5 profile of m/z 866 shows the presence of branched and bisecting isomers from the m/z 458, 444 and 417 ions, similar to the structural isomers of GlcNAc2Man5. Fragmentation patterns are shown on the m/z 866 precursor structure.

112 00 C. elegans_H2 449.2 100 C. elegans_cogc-1 449.1 2 MS"896 * ]MS'm/z8962< 1302-4R 1302-*R 375.2 -••667 463.2 -•667 101 scans 431.2 / S ^403 scans 375.1 11 NL: 3° NL: 301.1 431.2 50^1.61E1 150^ 635.2 359.1 593.2 8.02E-2 399.0 \ 315.0 | 315.0 \417.3 593.2 259.2 199.2 \ 563.1 \ 329.01 445.2 549.2 329.3 ,345.2 \417.2 635.3 259.3 345.1 241 .o\ 56 2 519.3 \ I \ 519.1 \ ^- \ \\ ,4-tUJ U •M,, 300 400 500 600 300JU L 400 500 600 700 m/z m/z

593 241 Na+ MeO- 241 MeO 431 MeO—(••-..v-OH : 593 •449 ./ 32? )-/<•••-. 635: 227 Med 'OMe ; OMe '•-, MeO- 259/ MeO OMe-449 563 / ..ko.;..3qi; •'; Na MeO—(•-.. •)—OH ; 315 [375 ; MeO^,:'--'V-^~o'i i •yyf:/ OMe"'"-., -o 301 MeO- | 227 MeO—f. T^OH 7345 ^MecT^Me i 563 ...; 399 MeO—

The third B/Y ion complement from GlcNAc2Man5Fuc is m/z 1476 and m/z

316. The m/z 1476 contains an outer arm fucose, the only such fragment ion in

2 the GlcNAc2Man5Fuc MS/MS spectra. The NF299 MS spectrum has a significant decrease in the peak intensities of this ion which suggests a decrease in outer arm fucosylation. Unfortunatly, the signal strength of the m/z 1476 peak

2 in the MS profile of GlcNAc2Man5Fuc in NF299 is too weak to analyze further.

MS3 spectra from both strains were obtained, but further fragmentation steps is not possible due to the poor signal (Figure 34). Comparatively, the two MS3

113 profiles do provide structural information and show a disparity of isomers between samples.

100 C. elegans_H2 100 C. elegans_cogc-1 871.4 1258.5 3 3 2 1434.6 ]MS m/z 896* 1054.6 1084.51258.6 1319.6 %AS mlz896 * 10545 10845 —•1476 -"R 11288.6/ -•1476--R \ / 1319.6 1231.6 I 50 scans 1405.6 200 scans NL: 2.46E1 1013.4 i 1434.7 NL: 7.84E-1 809.4 / 5 50 662.3 753.41 821.4 635.4 853.4 587.2 \ 707.4 433.2 1027.4 1231.5 1288.6 646.3 445.1 \ / 433.2 941.4 \ 519.1 \ / 823.3 1405.6 / I \ 587.3 \ \ Jut , Lj.i.ulku.iii-MiUi .i^l,,juilljlLM.iill.i.jiJtiiJiL,jMil> 400 600 800 1000 1200 1400 600 800 1000 1200 1400 m/z m/z

1.288 433 519

MeO— 707.. M 71231 1319 MsO OMe M«CS OMe '809 MeO- :1084 ;1405 445; '- 11434 MeO—: MeO—( )—;- -?° MeO OM« 1258 d—

1288 1084 MeO—I MeO-i l bb2 .; -°s h 3 /-% ! • > MeO OMe MeO OMe MeO OMe : 1231 1319 433 MeO—1 I MeO—1 / 0 MeO—( V~| 0 (~ -o ff'°) 1405

Na+ MeO OMe / MeO—I "r MeO—/ \ -0

MeO OMe 1258

Figure 34. MS3 spectra and fragmentation assignments of m/z 1476 from GlcNAc2Man5Fuc. Isolation and fragmentation of the m/z 1476 B-type ion reveals the presence of three isomers. The NF299 strain predominantly expresssed the isomer on the right, which does not have terminal fucosylation on a either arm. Terminal f ucosylated isomers are more common in the N2 strain.

114 The decrease of the m/z 1476 peak in NF299 suggests that

GlcNAc2Man5Fuc compositions are deficient with A/-glycans having terminal fucose residues. /V-glycans in NF299 with a GlcNAc2Man5Fuc composition position the fucose on the reducing end GlcNAc exclusively with trace amount of terminal-fucosylated structures. The wild-type N2 strain has a considerably more terminal-fucosylated A/-glycans from the increased peak intensity of m/z 1476.

Based on the MS/MS profiles of GlcNAc2Man5Fuc, the NF299 strains mainly contains fucose on the reducing end GlcNAc. In the N2 strain the location of fucose appears to be on the terminal ends and not attached to the 3-position on the penultimate GlcNAc, which is a common glycoform in C. elegans. This is based on the lack of GlcNAc-Fuc neutral loss or its presence as a sodium adduct

(m/z 646) in the spectrum. The m/z 1231 peak corresponds to a neutral loss of an internal GlcNAc and the m/z 1288 peak is formed from the neutral loss of a terminal fucose.

One preliminary conclusion that can be made from the decrease in the m/z peak in the NF299 strain is that the COG complex is required for outer arm fucosylation in C. elegans and that the NF299 strain does not lack nor create structural isomers from the N2 strain. MS" analysis at this stage does not find isomeric structures or peaks indicating different structures. What is evident from the MS/MS spectra is the change in fucosylation in the NF299 strain. The N2 strain has at least 7 structural isomers within the GlcNAc2Man5Fuc compositon all of which are present in the NF299 strains. Additionally, the NF299 strain expresses a structure that is found in the N2 strain in trace amounts (Figure 35).

115 Figure 35. Structural Isomers of GlcNAc2Man5Fuc.

The extent of outer arm fucosylation was quantified by measuring the relative intenisties of fucosylated outer-arm modifications in the MS2 spectra. In other words, the individual intensity of the non-fucosylated B-fragment ions (m/z

1098 and m/z 1302), and single fucosylated B-ion composition (m/z 1476) was normalized to the sum of B-ion peaks in the spectra and calculated as a percentage (Table 9). Three individual trial experiments and MS2 spectra were generated and used to ensure a robust analysis. The table below demonstrates that the NF299 strain has a decrease in terminal fucosylated structures. On average only 2.4 ± 1.2% of GlcNAcaMansFuc compositons in NF299 have outer- arm fucose residues, where in the N2 strain this structure is present 24.3 ± 7.6% of the time. The non-fucosylated terminal m/z 1302 B-ion are similar between strains with NF299 having 15.3 ± 4.8% and N2 having 14.1 ± 5.4% compositions.

The relative intensities of non-fucosylated terminal m/z 1098 B-ions are dissimilar between strains with NF299 having 82.3 ±6.0% and N2 having 61.6 ± 12.8% compositions.

116 N2 (Bristol) NF299 (k179) £•

Relative Intensity m/z B-ion Composition N2 (Bristol) NF299 {k179)

[ | 1098 61.6 ±12.8% 82.3 ± 6.0% •>- g 1302 •V*"" 14.1 ±5.4% 15.3 ±4.8% 1476 n *V*~" 24.3 ± 7.6% 2.4 ± 1.2%

Table 9. Relative intensity of MS/MS B-fragments from GlcNAc2Man5Fuc. These charts show the deviation of outer-arm fucosylation the NF299 strian compared to wild-type N2. The percent relative intensity each composition was calculated based on the measured intensity of both fucosylated and non-fucosylated outer arm B-ions from three individual experiments.

117 6.2.7 Structural Characterization of GlcNAcpMansFuc?

Of all the GlcNAc2Man5(Fuci-4) compositions analyzed, GlcNAc2MansFuc2 contains the most isomeric diversity. The relative abundance of di-fucosylated structures in the C. elegans glycome in the N2 strain is 21.5 ±1.0%. Similarily the A/-glycome of the NF299 strain is 24.9 ± 0.9% di-fucosylated compositions.

2+ The MS/MS spectra of the doubly charged Man5Fuc2 composition (m/z 983 ) illustrates the presence of six B/Y complement ion fragments in both strains

(Figure 36). These compositions range from GlcNAcMan4 and GlcNAcMan2Fuc2

B/Y complements (m/z 894 and m/z 1072) to a GlcNAc2Man5Fuc2 B-ion with single reducing end GlcNAc Y-ion (m/z316 and m/z 1650). Despite the presence of these ions in the MS/MS profiles, the abundance of these ion differ between the two strains, named the abundance of the m/z 1650 and 1476 in the

NF299 strain. The trend of decreased outer arm fucosylated observed in

GlcNAc2Man5Fuc continues in GlcNAc2Man5Fuc2 compositions as well. This is evident from the decrease in the aformentioned ions which contain outer-arm fucose residues.

118 100 694.4 J767.5 | 690.4 1272.6 868.5 *>« 676.4\ | 880.5 1098.5 1072.6 />- ' «3« 150-316.2* 664.4 894.5 •«„ 1476.6 :>- J 1302.6 490.3 1072.5 13026 939.5 /l533.| 7 1650.»7? 1*72.6 / .,533.7 I433.3 662.4 976.5 1551.7/ *£»./1551.7 \ \589.4 / ' 1476.7 M \ | , ,, 1850.8 Mi lil.„.,.,„ij|, I...IJ .ijl,/,,,, I 400 600 800 1000 1200 1400 1600 1800 2000 400 600 800 1000 1200 1400 1600 1800 2000 m/z m/z C. e/egans_N2 MS2* 9832* C. e/egans_cogc-1 MS29832* 50 scans NL: 2.39E2 53 scans NL: 6.31 E1

2 2+ Figure 36. MS of GlcNAc2Man5Fuc2(m/z 983 ). B/Y ion complementation from MS/MS profiles of m/z 9832+ reveals the increase complexity and increased fucosylation on the A/-linked core. The cartoon structures represent compositions only and not the exact location of residues. Open circle, galactose; filled circle, mannose; filled triangle, fucose; filled square, A/-acetylglucosamine.

From the absence of convincing isomeric structures in the two previously analyzed compositions, the investigation will focus on the location and changes in fucosylation between the two strains. Additionally, the comparative analysis is limited by the signal strength of the weakest sample and as such it is improbable that detailed MSn analysis is likely to establish linkage or detect the presence or absence of structural isomers only detectable in MSn>3 spectra. The limitations of sample signal is due to the presence of 6 topologies spread out in the

GlcNAc2Man5Fuc2 composition. Despite the relatively strong abundance of this composition in the MS1 profiles of NF299 (Figure 17), the wide distribution of fragments in the MS/MS profile limit the rounds of ion isolation and fragmentation. To establish a structure-function relationship of the COGC-1 complex in C. elegans, extensive MSn is not neccesary at this point. Preliminary data shows that the COG protein complex is required for outer-arm fucosylation

119 in GlcNAc2Mari5Fuc compositions. Additionally, the presence of tri- and tetra- fucosylated structures are significantly decreased in this strain (Figures 18 and

19). In order to comfirm this relationship, the tendencies of fucosylation via MS3 in NF299 will be explored, unless further fragmentation is possible.

Of the six B/Y ion complement fragments generated in the MS/MS profile from N2, all but one (m/z 1650) is present in sufficient amounts to generate a

MS3 spectrum in the NF299 strain. Three B-ions (m/z 894, m/z 1098 and m/z

1302) do not contain fucose residues and are the most abundant complementary peaks in NF299. Two fucosylated B-ions (m/z 1272 and m/z 1476) are present in moderate amounts in both strains. The MS3 spectra from the corresponding reducing end Y-ions do not exhibit isomeric structures between the strains

(Figure D4 and D5).

MS3 spectra of non-fucosylated B-ions are alike and suggest that the N- linked glycan structures are similar between N2 and NF299. The MSn profiles and corresponding structures of m/z 894, m/z 1098 and m/z 1302 ions are comparable to similar ions analyzed in detail in the previous two chapters. The

3 MS profiles of m/z 894 from GlcNAc2Man5Fuc2 do not show any significant variations (Figure 37).

120 676.4 100 C. elegans_N2 100, C. elegans _cogc-1 lMS3m/z9832n MS3m/z9832n -»-894| ->894|

54 scans ° c e 67 scans ° NL7.18 CD NL: 2.45 667.2 852.5 550 649.21 £50i 667.4 301.2 634.3 431.2 649.4 \ < 431.1 519.1 737,2 737.4 / 329.0 621.2 301.2 \ 519.3 634.4\ \ ' / 375.1 \ 458.2 847.3 / 329.2 458.2 I 621.4 \ \ \

Relative / / 375.2 | \ \3\'i3.i! I j \ 383.2 \\ ' Ml I 0- ,,!,,•..'.I. f.I. .-iLn.L I'M' V'I-In itiVvM 1 1 1 300 .M'M'I'H'IMIVM400 - 500 < 60!•0 700 800 300 400 60J!0 - 700 800 900 m/z m/z

676 .667 MeO-i r 649 .737, .329,' MeO^ V-i—;-0, 5.1.9,/ /431 ! MeO 30OM1e MeO—(/ ; >\—Oi Na+ 0 MeO V* V' ...B52

MeO OMe 458 3 Figure 37. MS of m/z894 from GlcNAc2Man5Fuc2. N2 and NF299 strains do not show any clear structural differences.

Similarity, the MSn profiles of m/z 1098 (Figure 38) and m/z 1302 (Figure

3 39) from GlcNAc2Man5Fuc2 remain the same between the strains. The MS spectra from the corresponding reducing end Y-ions do not exhibit isomeric structures between the strains (Figure D6 and D7). However the fragmentation of m/z ->1302 from 9832+ differs slightly from the fragmentation of m/z 1302 from

GlcNAc2Man5. The GlcNAc2Man5 composition contained an abundant 667 peak, indicative of either a branched or linear trisaccharide, but this peak is

undetectable in GlcNAc2Man5Fuc2 compositions. The spectra of m/z 1302 suggest the presence of two structural isomers. One isomer contains two disaccharide hexose residues linked to the core hexose at the 6- and 3- position.

The second isomer has three individual hexose residues linked to the core

hexose at the 2-, 4- and 6- position with an additional hexose linked to one of

121 those residues. Again the spectra between the two strains do not differ and signal strengths are comparable. 880.5 ,C. elegans_N2 100q C. elegans cogc-1 100 3 2 . jMS3m/z9832*o MS m/z 983 *o -•1098$ -••1098$ ^50 scans S 50 scans iNL: 8.32 NL1.37E1 ida n

1501 617.3 871.5 871.5 635.3 941.5 1056.6 a 941.5 1056.6 I 653.4 853.5^ 617.4 662.3 853.5 \ [431.2 J / 662.3 8' S\\ > / 676.4 1027.5 S 1027.6 / 723.4 : \ I \\\ //r\ \ „', , I ,>.l JU l„ o- i. --, 400 500 600 700 800 900 1000 1100 40..,0 . 50,.0 , J .,. 60i.,.lUl,.0 700J 800 J( '90 0 1000 1100 m/z m/z

662 §16 MeO—i \ MBO- MeO—i ---"-•. Na+ MeO—( )—Vo- .941. M 635 -72? MBO bMe \ »°' Me0 533, 723 / OMe \ 533 i W 11027 MeO—| \ / MeO—i /'/ MeO—(/ ) i-:tO—: }?\J\ ;_M J1027 MeO—( V—;-0 - 11056 p / OMeiji : MeO OMa : j-(\ '•• :M-"'^ hose 1 MeO-, ;l , / 853jjj87l| T° Pi/ -'>° Me0 : 853:1871/ / MeO— ~Lo ^ / MeO—Y ^> If;—O MeO 'OMe MeO OMei;:662 i880 431: i Isolation and fragmentation of m/z 880

100 C. elegans_H2 635.2 662.2 100 C. elegans_cogc-1 635.2 ^662.2 JMS*m/z 9832*o IMS4 m/z 9832*o 1098$ 1098| -»-880o -+880O- 51 scans 54 scans NL: 1.60 § 3NL2.81 723.2 838.3 723.3 838.3 §50^ 533.1 <| 50 \ 809.3 ! 327.1 417.1 533.2 \ \ 9> ^327.1 417.1 \ 561.1 343 0 S 444.1 \ 561.2 .653.2 809.2 343.1 \ 444.1 \ I 579.1 653.3 \ ' 458.1 \ 1579.1 \ 458.2 \ I 1605.2 848.4 399.01 | / i I 1605.1 848.2 676.2 I i \ rv\ 505.2; p / 505.0 676.2 11 \ / -; i/ \i W ii Or l,U.1l,,,lV, JX TfX 'ill ,-M. .'M .'.i>fJ,u60A0 700 800 900 400 600 700 800 100 400 500 60m/z0 m/z

662 458 635 MeO—i i MeO-^ 417 | i H0-)/';r:;]H u"0_i/ MeO Ofcte "eO OMe 5.33 723 _j.:-0 -/^L-ljiioH—. . ':: MeO^< //~':\— 809

HO /' OMs j 838 6.Q5.-'' 4-j 7: i T° 635J 653:

Figure 38. MS" of m/z 1098 from GlcNAc2Man5Fuc2. N2 and NF299 strains do not show any clear structural differences based on the profiles. The structural isomers shown are those identified in from m/z 1098 in GlcNAc2Man5Fuc compositions.

123 C. elegans_N2 100 C. elegans _cogc-7 1084.6 3 2 nn- MS m/z983 * 3 2 1057.6 MS m/z983 * 880.5 -M302 $ 1145.6 -•1302$ 1260.6 56 scans 880.4 100 scans NL. 1.18 NL: 1.89E-1 1145.6 809.3 1075.5 587.3 1057.6 !50 635.3 910.4 ^ 50: 57 8 587.2 t 662.4 839.4 445.4 f ? ™ , 839.4 | 1054.6 662.3 i 1029.6 \ I 463.3 \ 533.3 \ 1723.4 1114.5 I 697.5 910.5 \ \ \ 753.4 1029.5^ | 533.3\ 445.0 \ I ] 723.3 | (737.4 , 463.3' - I i i I \l \\ Q-.H,.i....,IJ, LiUkl J,IV,4MU 44*. W-A lujilk. 0- U,iii.-.l.i.J itiliU.Jl, .1 ILLMIJ.,,, J.J.I i 4*i i^tlUju.!, 400 500 600 700 800 900 1000 1100 1200 1300 400 500 600 700 800 900 1000 1100 1200 1300 m/z m/z

662 880 1084 MeO—i \ MBO—I ' •~-QV—G"K -* lm u«( \>M. \ "»° W \ 533 72.3.,/ MeO OMe MBO OMe i 11.45 ias. / 445] ''\737 u *°—I //•' MeO- UeO—| .' J587 ^tjH/"|0™"hiQ :< Meo- \ 463 \ 7 bM#! !1075| ,.«=^ ..£T-" o_J' 7 OMe: j ^ 1260 Med OMe M.O-, U..O-. "\. ;;, KiBSZi I / / ° I—o /-Q '•. !- i • / • / MB0_/ \ ;.0 / \ :-,;-;-;-0 : / MeO- 1057!| 1075:

MeO OM» MeO OMe \ [.-.".".'."""'.'"':":""".".".'' MeO— L™i 445 i MaO' OMe

635.2 100 ,C. elegans_N2 100 C. eleganscogc-1 635.2 3 MS4 m/z 9832* IMS4 m/z 9832* 1302$ j 1302$ -»• 880 cx> ! 662.2 8 j ->-880 oo 201 scans § 3201 scans 533.2 653.3 NL: 8.43E-2 1 3NL: 8.83E-2 838.4 5332 \ 4 505 662.2 327.1 723 2 838.2 561.3 653.1 6 1327.1 531.1 \ 561.1 \ 343.0 444.3 \ 809.4 j 579.1 723.1 i 343 1 444.5 \ 579.2 417.0 \ 458.4 I 805.1 I „„, I 458.1 605.2 709.3 \ ^MM.MMHII, 1,1,1!,,, XX 300 400 500 600 700 800 300 400 500 600 L70 0 860 900 m/z m/z

Isolation and fragmentation of m/z 880 662 458 i

MeO—\)—I'

MeO OMe MeO' OMe

Figure 39. MS" of m/z 1302 from GlcNAc2Man5Fuc2. N2 and NF299 strains do not show any clear structural differences based on the profiles.

124 3 The MS spectra of the m/z 1272 B-ion (Man4FucGlcNAc) differ considerably between strains (Figure 40). The MS3 spectra from the corresponding reducing end Y-ions do not exhibit isomeric structures between the strains (Figure D8). The most noticeble difference is the m/z 667 peak in the

NF299 spectrum with is consistant with a tri-mannosyl C-type ion which was characterized in detail in the Man5GlcNAc2 chapter. In that composition the topology was linear and branched. The fragment ions in this profile are not sufficient to make definative structural assignments, however there are some diagnostic ions to suggest the presence of only the branched type m/z 667 composition. These diagnostic ions are m/z 709, m/z 621, m/z 593, m/z 449 and m/z 431. These peaks, which correlate to two putative strucutures in Figure 39, are more prevalent in the NF299 and some are not detectable in the N2 strain.

Of the structures unique to NF299, the location of fucose remains close to the core and not on the outer arms. The location of fucose in NF299 is linked to the penultimate GlcNAc with an additional hexose (presumably a galactose) or linked to the core branched mannose. The latter structure is novel and most likely less plausible, but the m/z 621 and m/z 709 peaks can only be derived from this structure. It is possible that these ions are resulting unique fragmentation of the established core fucosylated strucuture. Unfortunately further MSn is not achievable. The m/z 667 peak is considerably decreased in the N2 strains and the most widespread structure in N2 is a bi-antennary structure.

An additional diagnostic peaks in the NF299 spectrum is the decreased m/z 1027 and m/z 1045, which are the B-type and C-type ions from the neutral

125 loss of an internal GlcNAc. This suggests that the GlcNAc residue in NF299 is more commonly modified with a fuc-gal disaccharide. Also the m/z 1084 peak generated form the neutral loss of a terminal fucose is significantly reduced in the NF299 strain further suggesting the prevalance of the complex core structure and decrease of outer arm fucose additions.

C. elegans_cogc-1 100 C. elegans_U2 1054.6 1054.4 100, 3 JMS3 m/z 9832* Q MS 2 6672 -•1272 •*. m/z 983 * Q - 1084.5 -M272-** 880.5 50 scans 1027.5 1115.6 880.5 \ 101 scans 1198.4 NL: 2.09E1 1230.6 NL: 3.69E-1 850.5 \ 1230.4 649.2 150 809.4 S50 862.2 1073.3 \ 415 2 505.3 646.2' 709.1 | 1045.5^ 415.1 621.3 1045.3 1075.3 \ \ 737.4 433.3 VY&S.3™* , 999.6 |431.1 619.2 \ 1027.3\ 1084.4> 475.3 \ I 662.31753.41 850.4 941.3\ \ 1201.6 .433.2 \\ 1115.3 I 591.3\\ \667.4 897.4 \ 449.2 \\ \ / I \ / 593.4 I..! \lil.l,l J.,.. ,„!,•< . . :-.U,H>l^ ^„.,.ll>lllj. . j .• J.UH MML 400 500 600 700 800 90\ 0 4»100. 0 1100 1200 1300 400 500 600 700 800 900 1000 1100 1200 1300 m/z m/z

1.054 850 MeO^ i MeO— 1045

WeO OMe MeO OMe \

:1201 'H<^ 11230

662 880 I 433 UeO-i \-\—O (~ V— •k~7 UeO OMe MaO OMe / 7 IM111P5

\ / MeO- / Me"tO v 201 Med OMe Med 'tf J1230

MaO—1 809I )~° r°s I * 1045 MeO—( ) -O MeO OMe 836

3 Figure 40. MS of m/z 1272 from GlcNAc2Man5Fuc2. The spectra of N2 and NF299 strains differ notably. The NF299 strain generates a greater amount of the structures shown at the top right and bottom, whereas N2 predominatly expresses the structure at the top left.

The spectra of the m/z 1476 B-ion have similar profile trends to the m/z

1272 spectra and the variation of fucosylation continues between strains. This

126 ion has one additional hexose residue compared to m/z 1272. The location of this hexose is on the outer arm of the branched m/z 667 structure in the NF299 strain and is found as an internal residue in the N2 strain (Figure 41). The

NF299 strain preserves the Gal-Fuc-GlcNAc ion peak (m/z646) that is mostly absent in N2. Additionally, the peak that correlates to a neutral loss of a terminal fucose (m/z 1288) is scarcely detectable in the NF299 spectrum. The majority of peak ions in the NF299 profile strongly point to the expression of the core modified structure over the bi-antennary and bisecting core structures found in

N2.

Consistently the NF299 strain illustrates a reduction of terminal fucosylation of A/-glycans. Similar to the analysis in the previous

GlcNAc2Man5Fuc chapter, the extent of fucosylation was quantified by observing the abundance of fucosylated B-ions in the MS/MS spectra. This analysis demonstrates the decrease of terminal fucose in NF299. The most abundant B- ion peak in the wild-type N2 was m/z 1272 (GlcNAcMan4Fuc) and encompasses

32.7% of GlcNAc2Man5Fuc2 structures. Conversely the abundance of this ion in the NF299 strain is 9.1%. Fucose residues in NF299 are found primarily on the core GlcNac residues. This conclusion is drawn from two observations: 1) the B-

ion quantification in Table 10 shows that the sum of B-ions in only 12.8% for

NF299 compared to 41.1 % in N2. As a result the location of fucose must be found on the reducing end GlcNAc in the corresponding Y-ion. 2) In B-ions from

NF299 that do have fucose, the ions generated in the MS3 spectra indicate the fucose is linked to the penultimate GlcNAc with an additional hexose residue and

less likely on terminal residues common in N2.

127 1054.6 100C. elegans_U2 100 C- elegans_cogc-1 jMS3/77/z 9832" MS3 m/z 9832* j/ >1476-*R -•1476 A. 350 scans 1434.6 = NL:2.74 809.4 l 821.5 1084.5 ^433.2 662-3 \ 1871.5 [463.2 \7°7'3 / 1897.5

1405.6

400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 m/z m/z

1288 1084 662 1258 1.0.5.4 MeO-i

MeO OMe Med bMe tkrl 433I MeO—| :

MeO—( V:

MeO OMe / OMe

MeO-

1.288 433 519

h \ 707, 71231 1319 MeO—( \-\—O MeO OMe 809 MeO OMe :1084 ;1405 445 - 11434 MeO-i MeO—( )—: -f° Mao bMe! MeO OMe 1054 1258

3 Figure 41. MS of m/z 1476 from GlcNAc2Man5Fuc2. The spectra of N2 and NF299 strains differ notably. The NF299 strain generates a greater amount of the structures shown at the top right, whereas N2 predominatly expresses the structure at the top left and bottom.

128 N2 (Bristol) NF299 (k179) 3 ^ \l"-r-'-^-'*-; •"'.

B-ion Relative Intensity m/z Composition N2 (Bristol) NF299 (k179)

Q 894 18.4 ±6.9% 27.2 ± 1.6% :«

•| 1098 33.5 ±12.3% 51.8 ± 2.2% •>-

| | 1272 <•*"• 32.7 ±15.1% 9.1 ±2.1% • 1302 •V*"" 6.9 ± 2.0% 8.2 ± 2.0% •j 1476 •v*~" 7.3 ± 3.0% 3.4 ± 0.4% 1650 H 4-i**~" 1.1 ±0.5% 0.3 ±0.1%

Table 10. Relative intensity of MS/MS B-fragments from GlcNAc2Man5Fuc2. This table shows the deviation of outer-arm fucosylation the NF299 strain compared to the wild-type N2 strain. The percent relative intensity each composition was calculated based on the measured intensity of both fucosylated and non-fucosylated outer arm B- ions from three individual experiments.

129 6.2.8 Structural Characterization of GlcNAcpMansFucg

The most noticable observation of tri-fucosylated compositions in the

NF299 strain is the considerable decreased abundance compared to N2. The N- glycome of the N2 strain boasts 19% tri-fucosylated compositions whereas

NF299 only contains 12%. Of all the tri-fucosylated structures in the N2 glycome, GlcNAc2Man5FuC3 is the most abundant. The MS/MS profile of this composition has four B/Y ion compositions (m/z490/1651, m/z 664/1476, m/z

694/1446 and m/z 868/1272). The m/z 1651 B-ion peak is absent in NF299 and the m/z 1446 B-ion peak is notably reduced, both of which contain two fucose

residues (Figure 42). Similar to the analysis of B-ion compositions with two fucose residues in the GlcNAc2Man5Fuc2, the GlcNAc2Man5Fuc3 analysis in

NF299 is also limited due to the small abundance of these structures.

Additonally, the tri-fucosylated structures are overall less abundant in the N- glycome than di-fucosylated compositions. As a result the structural mining is restricted to fewer rounds of MSn.

130 10Ch 1272.6 100 961.1 865.5 868.5 / :> 1027.5 868.I 5 1272.6 \ \ *U" 1054.5 1446.4 <*. 433.3 664 4 1476.6 1707.3 662.4 i 1317.6) 1476.7 $ ^>* ^2.6 | 17257 150 1707.8 a 50 433.2 gg44 1054.5 1446.6 589.3 \ I I /445.2 1 | 1696.7 / < ' i 649/ 9 + \ \ | 694.4 1725.8 1419.6 1593.21 I / 1589.31 i / / ^ \ITA i 04 ilujilajkyii, r—* ujj i„Jl„i 400 600 800 1000 1200 1400 1600 1800 2000 400 600 800 1000 1200 1400 1600 1800 2000 m/z m/z 2 2 C. elegans_H2 MS2 m/z 107CT C. elegans_cogc-1 MS m/z 1070 ' 50 scans NL: 2.92E1 50 scans NL: 5.21 E1

2 2+ Figure 42. MS of GlcNAc2Hex5Fuc3 (m/z 1070 ). B/Y ion complementation from MS/MS profiles of m/z 10702+ from the two strains reveals the decrease of outer arm fucosylation in NF299 compared to N2. This is evident from the decrease of m/z 1651 and m/z 1446. Open circle, glactose; filled circle, mannose; filled triangle, fucose; filled square, A/-acetylglucosamine.

3 The MS spectra of the m/z 1272 B-ion peaks (GlcNAcMan4Fuc) have noticable differences between strains (Figure 43). The MS3 spectra of the m/z Y- ion peaks are similar between strains (Figure D9). The trend of the NF299 to either lack or have decreased terminal fucose continues and is evident from the m/z 723, m/z 1027, m/z 1045, m/z 1084 and m/z 1115 peaks. Also, the presence of other peaks (m/z 646, m/z737, m/z 649 and m/z 667), which are largley absent in N2, point to an increase in penultimate GlcNAc modifications.

Interestingly, is the presence of the m/z 667 peak that in previous compositions was established as a linear and branched tri-mannosyl structure.

131 1054.5 1054.5 100; C. elegans cogc-1 100_C. elegans N2 3 2 3 MS m/z1070 § =MS m/z1070o 1084.5 -•1272* „_, = -+1272* 1115.5 * 667.4 8 $0 scans S804 / 1230.6 50 scans / | 1NL7.04 1045.5 danc e C ; NL1.61 ! 880.5 c 850.4 1027.5\ 1230.6 350; •15.2 6174 809.4 1012.5U §50; 433.3 4312 605.41 862.4 753.4 897.6 \ \ 850.4 433.2 \ \ 707.3 | 1201.5 649.4 999.6 \ I 802.5 \ 1198.5 L 3 4 646.4\

Relativ e 1115.5 \ Relati v \ 621.4 \\ 1045.6 \ lr \ ^ \7 , \ 737.5 \ \ ,j| N \

otL y/r.. .J,. 1. H,,,|J|UlU,iJ.U> i / v -V, ••iJtl i •iJi H.JU f-r^r". o ^ ij i iiMiU -M.. l< MU-, 400 500 600 700 800Jlu .90 0 1000 1100 1200 1300 500 600 700 800 900 1000 1100 1200 1300 mfz m/z 1.Q54

3 Figure 43. MS of m/z 1272 from GlcNAc2Man5Fuc2. The spectra of N2 and NF299 strains differ notably. The NF299 strain generates a greater amount of the structures shown at the top right and bottom, whereas N2 predominatly expresses the structure at the top left.

132 In this case, MSn of m/z667 from NF299 shows evidence for both the linear and branched structures (Figure 44). The m/z 563 peak can only be created from a linear structure. However, the remaining peaks can be generated from both structures and this MSn spectrum is not conclusive to rule out any isomer. The m/z 399 peak that is unique to the branched structure is not detected. Likewise, the m/z227 peak that is unique to the linear structure and is present in previous m/z 667 peak fragment profiles, is not represented in this spectrum.

MeO- 241 375.1 449.2 100 MeO—{ r 301.1 MeO -cOMe[449\ \ \ 315N 359.3 \J-o 301 • 445.2 Me0_^Jp0H 399 3241.3 593.2 313.0 \ 431.3 \ 473.4 2592 1345.2 I 510.6 563.0 MeO—i 259 \ / \SMB ;635 417.1 \ 549.4 ') /\ \375 W MeO—( \ \ / \ \431 MeO OMe 400 .,•,,50, 0I !,• i.I 600 m/z C. etefifans_cogc-1 50 scans MS'm/z10702o NL:1.61 1272 5 593 -•667 MeO—| 0 593, :/- ./ 241 Na+ MeO—t---...^-OH 635] Na+ MeO MeO OMe '•• 431 MeO—i 227: 563 MeO 7 M80^t^^o T345i •449/ / 329 "J?" 635! MeO' OMe 227 259'. 259/ \ 563 MeO-^ 241! •>(315;;, / ...k'D,-..301j •'] |375 '• MeO-^V--''^ 6] MeO—( ) ; o \ \4.3.1 : 345..-Meo' ^0Me MeO OMe ' .449

Figure 44. MS4 spectra and fragmentation assignments of m/z 667 from GlcNAc2Man5Fuc3. Isolation and fragmentation of the m/z 667 C- type ion (three terminal mannose residues) reveals the presence of three isomers based on the presence of the m/z 563 peak that can only be created from a linear structure.

133 Further fragmentation of the the m/z 1272 B-ion (10702+^1272-+1054), illustrates the significant difference between strains (Figure 45). The loss of a terminal hexose from m/z 1272 in N2 results primarily in a m/z 809 peak, which is not observed in NF299. This peak is the neutral loss of a non-modified

GlcNAc residue. The structure of the NF299 m/z 1272 B-ion is primarily the hex- fuc modified GlcNAc, evident from peaks m/z 433, m/z 646 and m/z 667. The m/z 880 peak results from the neautral loss of the hex-fuc disaccharide, which is not illustrated in Figure 45.

433.2

827.3 836.4 662.4 836.2 707.3 •850; | 866.2 \ \ 662 2 735 1 897.3 329.0 298.9 4332 - \ - 449.2 646 3 •343.2 1475.1 591.3 \ \l"3_3 I 415.2 \ 1012.6 983.41 1022.4 \431.3 507.2 \ (431.2 1501.3 \ \ W™$ 644.3 667.0 OJU / 794.1 \ / \ |4,I I,. I., 0W* .,iu .in,i !i •>,; i,,!,)) ,0-^4 300 400 500 600 700 800 900 1000 1100 300 400 500 600 700 800 900 1000 1100 m/z m/z C. elegans_N2 51 scans C. elegans_U2 101 scans 2 MS*m/z1070 o NL1.07 MS4 m/z 1070% NL: 1.12E-1 1272f 1272$ -+1054O- -•1054^ 646

§Q9 897 MeOn / }/-o j983

MeO^ 433 012 MeO—( ) O 0 .501 MeO OMef MeO OMe Na+ 866 662 \ / MeO' OMe MeO OMe 866 ,162

.0 ^y 809

MeO OMe MeO OMe \ / CQ-| 897

1475 ' 1983 Mrf s-n' J1012 10 w4 MeO OMe 836

4 2+ Figure 45. MS of m/z 1054 from GlcNAc2Man5Fuc2(1070 ->1272^1054). The NF299 strain generates a greater amount of the structures shown at the top right, whereas N2 predominantly expresses the structure at the top left and bottom.

134 Isolation and fragmentation of the m/z 1476 B-ion peak (GlcNAcMan5Fuc) shows similarity to the other mono-fucosylated B-ion analyzed above (m/z 1272)

(Figure 46). That is terminal fucosylation in the NF299 strain is not equivalent to the N2 strain. Similar peaks are present in NF299 that are not in N2 (m/z646 and m/z 911). The NF299 strain lacks higher m/z peaks that correspond to the neutral loss of an internal GlcNAc (-245). Particularly observed from the m/z

1231 and m/z 1319 ions. Also, the MS3 spectra of the m/z Y-ion peaks are similar between strains (Figure D9).

10oqC. elegans_H2 1434.4 100 C. elegans_cogc-1 JJ71.3 |MS3 m/z 1070* 1084.4 1258.3 I MS3 m/z 10702* 1084.3 '1476$ 1231.4 1319.3 -•1476$ 1054.5 J 1258.5 = 101 scans 201 scans I \ 14345 ^NL: 3.73E-1 NL:8.61E-2 c 853.4 5*50 871.3 853.4 941.5 1319.5 1068.6 \ 1402.6 <5 ^431.3 587.1 |431.4 646.3 825.3 \ 983.4 1187.4 w -_ 1433.1 1662.1 795.M3 1249.5 433.2 \ 653.6 \1216.5l 927.2 1446.7 n \ 463.3 \ \707.2\ 1402.3] 463.2 \ 1693.4 \ 1203.3 / K I /533.0\ \ \ \ \ 588.9 \ 1795 3 •Uthlif iUJl iliUiJ A* lUi'iti^ini'iifi'^*^. 0^ In • .1 illl,., , ll.,|i,ULl..luil

1288 1084 662 1.25.8 1.0.54 MeO-i 87.1 mo—(~\-\ 941 853 MeO otM MeO' bMs 445j 'MeO—( \ : £) J1405 MeO—> ) ( MeO- 7-° 433',/

MeO—( ) j °, A

MeO OMe 1288

1.288 433 51g j MeO—| ;MBO-| "";;-"-'

MeO- o 707 1319 71.231 MoO OMe •' .809 ;1084 MeO- J1405 445] | or* - 11434 MeOn MeO- ""r

MeO OMe: MeO 0l*ilfVa. 11258 "•-•

3 Figure 46. MS of m/z 1476 from GlcNAc2Man5Fuc3. NF299 generates a greater amount of the structures shown at the top right, whereas N2 predominatly expresses the structure at the top left and bottom.

135 The m/z 1446 and m/z 1650 peaks, which are mostly absent in the NF299 strain, have fucose residues linked to the penultimate GlcNAc residue or at the terminus of the structure in the N2 strain (Figures 47 and 48). The MS3 spectra from NF299 are too poor to make any accurate conclusions and therefore it can be concluded that these structures are generated in insignificant quantities. The

MS3 spectra of the m/z Y-ion peaks are similar between strains, but the weak signal of NF299 is evident in these profiles (Figure D11 and D12). It also appears in N2 that the penultimate GlcNAc is always modified with a hex-fuc disaccharide (m/z 646) from the absence of a m/z 1201 and m/z 1405 (245 neutral loss of an internal GlcNAc) in the m/z 1446 and m/z 1650 profiles

respectfully. That is to say that the structures in N2 do not have two terminal fucose residues. One fucose is attached to the GlcNAc and the second is on the terminus of one antenna. The NF299 strain does not add fucose to the terminus of /V-glycans (Table 11). Specifically, NF299 does not readily generate /V-glycans with two fucose residues distal from the reducing end GlcNAc.

841.5 662 10o C. elegans_N2 1258 11 3 2 «3 ^ MS m/z1070 *, 1054.5 MeO-i —| 82B3 -•1446 i MeO-( V —<\\ 503 911 *i 50 scans Med' bfi/te Meo' bfcle \ / MeO- ~7 NL: 1.26E1 1475 1/b ro 72 28.6 MeO-

136 C. elegans_m 1045.5 1462 1045 MS3m/z10702* 1258.6 M601 . 1027 —•1650-*R MeO-y ) 503 1115 g . 77 scans MaO bMe MeO OMe MeO-, / § :NL:2.05 1228.6 1608.7 ffi5 \ 1432 MCKX,/ -o (T) 11576 MBO-I 1228 7^J1608 .Q50 605.3 \ 1576.6 < \ 616.0 1087.5 l 646 / -^o 1027.5 1493.7 \ \ / i | 549.2| 1646.3 11115.6 Meo OMe Med bM* B 475.3\ 1462.6 \ 999.6 : 11180.5 1432.8 1432 ^ \ 503.1 : MeO-i 971.5 \ MM MeO-i )— v; 11258 400 500 600 700 800 900 10001100120013001400150016001700 m/z

3 2+ Figure 48. MS of m/z 1650 from GlcNAc2Man5Fuc3(1070 ^1650) in N2. This structure is largely absent in NF299.

137 N2 (Bristol) NF299 (k179)

i." • » - • '*• r \

t .'•-'-•.^"»-,»--• .vc:'-:.;'. =•

B-ion Relative Intensity m/z Composition N2 (Bristol) NF299 (k179)

B- 4**~" 79.4 ± 4.4% 80.4 ± 1.0% II 1446 4**~" 14.1 ±2.5% 3.1 ±0.5% I I 1476 ^i**"" 4.6 ± 1.5% 15.6 ±1.5% 1650 • Jw*~" 1.9 ±1.2% 0.9 ± 0.3%

Table 11. Relative intensity of MS/MS B-fragments from GlcNAc2Man5Fuc3. This table shows the increased decrease of outer-arm fucosylation the NF299 strian compared to the wild-type N2 strain. The percent relative intensity each composition was calculated based on the measured intensity of both fucosylated and non- fucosylated outer arm B-ions from three individual experiments.

138 6.2.9 Structural Characterization of GlcNAcpMan^Fuc^

The final composition to be analyzed is GlcNAc2Man5Fuc4. From the complete A/-glycome analysis obtained from the MALDI-TOF profiles, the existence of tetra-fucosylated compositions in NF299 cogc-1(k179) is roughly

1%. It is important to note that baseline peaks inherent in the spectrum may account for this percentage and not necessarily a true glycan. MS/MS of any peak confirms the presence of carbohydrates based on the peak generated.

Comparatively tetra-fucosylated structures comprise 9% in N2 strains. MS" of

GlcNAc2Man5Fuc4 in the N2 strain confirms the presence of one B/Y ion complement (m/z 1446 and 868). However, the MS/MS profile of NF299 does not contain these peaks or any other ions B/Y ions that could be generated from

2+ GlcNAc2Man5Fuc4 (m/z 1157 or 2291) (Figure 49). There are three peaks in both spectra that correspond to carbohydrate compositions {m/z 490, m/z 690 and m/z 894). MSn of the m/z 1446 B-ion and m/z 868 Y-ion confirms the structure and its likeness to m/z 1446 from the GlcNAc2Man5Fuc3 composition in the previous section (Figures 50 and 51).

139 690.4

894.4 864.4 1006.8 B51.5. 1156.7 /1081.2 •m 605.3 490.3 I 1383

i.nliliil,iiiiiiii.WMwlw4* i|in|il|H |Hn>wW» * iii|i>iii>Mftj4'iMi I ^ 400 600 800 C. e/egans_N2 MS2 m/z 11572* C. etegans_cogc-1 MS2 m/z 11572* 50 scans NL: 5.20E2 85 scans NL: 3.45E2

2 2+ Figure 49. MS of GlcNAc2Man5Fuc4 {m/z 1157 ). B/Y ion complementation from MS/MS profiles of m/z 11572+ shows only one N- glycan composition in N2. The NF299 does not contain any obvious GlcNAc2Man5Fuc4 structure. Open circle, galactose; filled circle, mannose; filled triangle, fucose; filled square, A/-acetylglucosamine.

1258 433 1003C. elegans_H2 4 \M£ m/z 11572' S^- 1054.5 646 911 _>1446 MeO Cy^l^-O (~ ) :

350 scans Meo QMe MeO OMe ; 1NL2.27E1

646.2 605.3 | 795.3 1372 5 976.3 433.1 \ \ 549.3 911.5 i

,i,. M..WM,1|I.O.U ,..M.„y)t l,ulL|J„.,„ii.|iA ,^|,,,U|.>ll»„4..»l«rt I Isolation and fragmentation of m/z 841 iooqC. elegans_N2 4 2 ^MS m/z1157 * 549.3 1446 433. -•841 i 50 scans 810.0

MeO OMe MeO OMe ir.0 'x / MeO-^ MeO—1 °Ue$9^\ Y767 Meb-^ \ 623 MeO OMe 500 600 800 m/z

Figure 50. MS" of m/z 1446 from GlcNAc2Man5Fuc4. Fragmentation of the m/z 841 peak from m/z 1446 B-ion illustrates the presence of a branched terminal FucMan3 structure.

140 100 C. elegans_H2 662.3 MS3 m/z 11572* -•868 50 scans NL: 1.25E2 589.3 ' 501 650.5 Ms2a-*836 398.3 458.2 680.4

433.2 519.3

0*f Jjw, 500 S0O 700 800

3 Figure 51. MS of m/z 868 from GlcNAc2Man5Fuc4. Fragmentation of the m/z 868 peak confirms the structure of the reducing end Y-ion (complement of the m/z 1446 B-ion).

The three peaks that correspond to a carbohydrates m/z composition in the NF299 spectrum are m/z 490, m/z 690 and m/z 894. The MS3 spectra demonstrate that the fragments of m/z 490 and m/z 690 can be assigned to carbohydrates. The m/z 490 ion corresponds to a fucosylated reducing end

GlcNAc and the m/z 690 is a Man2GlcNAc composition. Interestingly, the sum of the m/z 490 and m/z 690 peaks is 1157. It is likely that these two ions are B/Y complements of the singly charged m/z 1157 GlcNAc2Man2Fuc composition that

2 is inherently included from the isolation of the doubly charged (115772+

GlcNAc2Man5Fuc4) species in the mass spectrometer.

141 472.2 100: C. elegans cogc-1 A 2 100-C. etegans_N2 MS m/z 1157 * 3 2 ^MS 1157 * -+490 -»490 c e 52 scans 50 scans 2!*i 458.3 NL: 1.28 NL: 5.68 / 458.1 430.3

o 284.2 343.2 211.1 \ 302.0 371.3 430.1 302.2 i 385.2 veAbunda r 211.0 343.1 at i 224.0 \ \ \ 385.1 \ I 229.1 I 229.1 \ I R e \ w ^v* oJL 9 ! -4 i) I I II I, .„.d , „ j,,U 500 200 300 400 500 200 300 m/z 302

211. '284 2.29/ (-H20) 472 (-MeOH) 458 (-H20 + Ac) 430

iooqC. elegans_N2 100 C. elegans_cogc-1 1 2 472.3 IMS m/z 1157 * / EMS4 m/z 11572* I -+690 648.3 •690 8 344 scans 8 ^51 scans | 1NL1.07E2 | = NL:3.48 463.2 533.1 isol 463.2 445.2 \ 268.1 \ 329.2 &• -- 227.1

,I»|.T iW, nlmili 200 300 400 500 200 300 400 500 600 700 m/z m/z

MeO—(/ ) ;•:—O—i- . 6481 MeO OMei: 463i 445! "7"°

Figure 52. MSJof m/z 490 and m/z 690 from 1157 .

The presence of the m/z 894 peak in both sample spectra is also a product of a low molecular weight glycan carried over during the isolation of the

2+ doubly charged GlcNAc2Man5Fuc4 composition (1157 ). The m/z 894 peak with the addition of a reducing end GlcNAc residue, has a sodiated m/z value of

1187. This ion would not be selected from the m/z 1157 isolation window speculated on the ion trap. However if this composition (GlcNAc2Man3), which is

142 a documented A/-glycan in C. elegans (Cipollo, Costello et al. 2002), is under- permethylated then its mass would be 1157. In any case, the NF299 strain and

N2 strain differ considerably with NF299 lacking any GlcNAc2Man5Fuc4 composition in its glycome. The absence of this structure in NF299 further supports the role of the conserved oligomeric Golgi complex in terminal fucosylation in C. elegans.

676.2 676.4 100 C. elegansHl i oo - C. elegans_ cogc-1 3MS3m/z11572* MS3 m/z 11572* 1 -*• 894 -•894 8 51 scans 8 ^52 scans NL: 6.63 NL: 1.28 667.4 413.1 667.2 301.1 431.2 649.3 \ 301.2 •"?;,', S19.1 649.2\ 14311 / 413.3 J 458.3 621.4\ \ 329.2 ' 621.3 / 383.1 H49-2 / 329.2 \ .I I 519.3 \ \ 0%J^^XU ..i.iJ.Lii i 300 400 500 600 700 800 900 300 400 500 600 700 800 900 m/z m/z

676 MeOn 667 -o i . „™ 649 .7.37, MeO-/ .329, N 519./ MeO OMe : -v /431 ! ^301 i ,U ii • MeiW/ y1 :-:—O Na+ / OMe MeO—i .£52 -0 '•.:'•/ MeO—/

MeO °Me !>IEQ

Figure 53. MS3 of m/z 894 from 11572 +

143 CHAPTER 7

CONCLUSIONS

Glycosylation is a complex process that involves a substantial fraction of cellular components. The correlation between glycan related gene transcription, biosynthesis and glycan structure is one aspect of glycobiology that remains poorly understood. Even more confounding is the connection between a particular glycan structure and its biological function. It is known that defects in the glycosylation pathway are severely life threatening and the occurrence of these diseases is exponential. Recent developments have enabled more robust and accurate screening and diagnoses, but treatment and cure are wanting.

There is urgent need for new strategies to address the treatment of CDG and understand the pathology and inheritance, both of which are virtually unknown.

One goal of the CDG Family Network is to look for means of improving the symptoms of CDG illness. The most puzzling feature of CDG diseases is the range of symptoms and severity of disease. This holds true between individuals who share identical gene defects. Not considering possible environmental aspects that may influence the spectrum of clinical phenotypes, the severity of

CDG is a product of partial defects in two or more genes in the same pathway.

This phenomenon, also known as synergistic heterozygosity, is rarely considered among CDG patients. Although no cure is available, a list of potential genes to

144 screen and treat could reduce the severity of the disease to a state where supportive therapy would increase the quality of life. The list of 512 tunicamycin

hypersensitive genes represents only one half of the C. elegans genome. Future work would include a second screen of previously untested genes not

represented in the version 1.1 ORFeome RNAi library. Optimistically, this list of. tunicamycin-hypersensitive genes could be used as a routine screen to categorized additional disorders and reduce the clinical severity of CDG diseases.

A second goal of the CDG Family Network is to identify further types of

CDG and determine their symptoms and outcomes. CDG awareness is rapidly expanding and the possibility of subtypes is only limited by the number of genes

involved in the biosynthesis of glycans. Currently this number is roughly 200, but the number of known CDGs is 18. The identification of subtypes is difficult, but

relating the biological consequence to its gene defect is considerably complicated. Most recently, there has been a rapidly expanding group of

patients with novel CDG type II defects involving subunits of the conserved

oligomeric Golgi (COG) complex (CDG-ll/COG-1, CDG-lle and CDG-llh). Little

is known about the role of the COG complex and its implication in N- and O-

glycan synthesis. Hence, understanding the role of the COG complex in

glycosylation is the first step in determining an acceptable clinical treatment or

course of action. Unfortunately only two viable forms of COG deficient CDG

models exist. Both are single gene knockouts in C. elegans and Chinese

145 hamster ovary cells resulting in CDG-ll/COG-1 like illness. Although there are

CDG diseases of subunits 7 and 8, no animal models exist.

The comparative analysis of N2 (Bristol) and NF299(/cf 73) has led to the identification of defects of terminal fucosylation of A/-linked glycans on all glycoproteins in CDG-ll/COG-1 conditions. The biological significance of fucosylation is fairly understood and ranges from tissue development, angiogenesis, fertilization, cell adhesion inflammation, cancer metastasis to invasion of bacteria, parasites and viruses (Listinsky, Siegal et al. 1998; Ma,

Simala-Grant et al. 2006). An increase in fucosylation is implicated in inflammation and cancer (Miyoshi, Moriwaki et al. 2008). Additional studies found that outer arm fucosylation (attached to GlcNAc via an alpha 1-3/4 linkage) is significantly increased in the serum of hepatocellular carcinoma patients

(Tanabe, Deguchi et al. 2008). Notch signaling relies heavily on Ofucosylation.

Defects of Ofucosylation of Notch is embryonic lethal in mice (Chen, Lu et al.

2006).

Moreover, defects of the GDP-fucose transporter on the Golgi cause

Leukocyte Adhesion Defficiency (LAD)/CDG-llc (Sturla, Rampal et al. 2003).

LAD/CDG-llc is characterized by reduced expression of fucosylated glycoconjugates, specifically blood group H and Lewis antigens (Etzioni,

Frydman et al. 1992). Hypofucosylated structures include ligands (leukocytes) for the selectin family of adhesion molecules. The result caused a reduction of leukocyte rolling on endothelial cells which is mediated by selectins. This inability of selectins to identify their glycan ligands impairs the recruitment of

146 leukocytes into inflamed tissues. Data supports the concept that A/-glycans are primarily affected in LAD/CDG-llc patients, where fucose additions are found on the reducing end GlcNAc or as terminal modifications.

Traditionally the regulation of fucosylation depends on fucosyltransferases, GDP-fucose synthetic enzymes, GDP-fucose transporters and lysosomal fucosidases. In addition to these components, this study showed that the COG complex acts as non-direct factor of fucosylation in C. elegans.

Consequently the COG complex is required for proper function of the glycosylation machinery via intracellular trafficking. The lack of fucosylation in C. elegans has an effect on its development and brood size, but interestingly the strain NF299 cogc-1(k179) remains viable.

As the discovery of glycan structural variations increases and the number of enzymes enlisted to create these structures remains the same, it is crucial to consider other molecular mechanisms by which the Golgi functions that gives rise to glycan diversity. Furthermore, the pathogenic mechanisms linking glycosylation defects to clinical presentations in CDG cases remains to be elucidated. The key to understanding the systems that underlie and control glycosylation lies in animal models of CDG genetic defects. This approach not only found trends of fucosylation in CDG-ll/COG-1, but also provides a target for pharmacological intervention where a disruption in outer arm fucosylation is warranted, such as tumor metastasis, inflation or graft rejection.

147 REFERENCES

Aebi, M. and T. Hennet (2001). "Congenital disorders of glycosylation: genetic model systems lead the way." Trends Cell Biol 11 (3): 136-41.

Altmann, F., G. Fabini, et al. (2001). "Genetic model organisms in the study of N- glycans." Biochimie 83(8): 703-12.

Apweiler, R., H. Hermjakob, et al. (1999). "On the frequency of protein glycosylation, as deduced from analysis of the SWISS-PROT database." Biochim Biophvs Acta 1473(1): 4-8.

Ashline, D., S. Singh, et al. (2005). "Congruent strategies for carbohydrate sequencing. 1. Mining structural details by MSn." Anal Chem 77(19): 6250-62.

Asikainen, S., S. Vartiainen, et al. (2005). "Selective sensitivity of Caenorhabditis elegans neurons to RNA interference." Neuroreport 16(18): 1995-9.

Ausubel, F. M. (1999). Short protocols in molecular biology : a compendium of methods from Current protocols in molecular biology. New York, Wiley.

Barnes, G., W. J. Hansen, et al. (1984). "Asparagine-linked glycosylation in Saccharomyces cerevisiae: genetic analysis of an early step." Mol Cell Biol 4(11): 2381-8.

Beminsone, P. M. (2006). "Carbohydrates and glycosylation." WormBook: 1-22.

Brenner, S. (1974). "The genetics of Caenorhabditis elegans." Genetics 77(1): 71-94.

Brown, R. M., Jr. (1969). "Observations on the relationship of the golgi apparatus to wall formation in the marine chrysophycean alga Pleurochrysis scherffelii Pringsheim." J Cell Biol 41(1): 109-23.

Cavanaugh, L. F., X. Chen, et al. (2007). "Structural analysis of conserved oligomeric Golgi complex subunit 2." J Biol Chem 282(32): 23418-26.

Chandrasekaran, A., A. Srinivasan, et al. (2008). "Glycan topology determines human adaptation of avian H5N1 virus hemagglutinin." Nat Biotechnol 26(1): 107-13.

Chen, J., L. Lu, et al. (2006). "Expression of Notch signaling pathway genes in mouse embryos lacking beta4galactosyltransferase-1." Gene Expr Patterns 6(4): 376-82.

148 Chen, S., J. Tan, et al. (2002). "UDP-N-acetylglucosamine:alpha-3-D-mannoside beta-1,2-N-acetylglucosaminyltransferase I and UDP-N- acetylglucosamine:alpha-6-D-mannoside beta-1,2-N- acetylglucosaminyltransferase II in Caenorhabditis elegans." Biochim Biophvs Acta 1573(3): 271-9.

Chen, S., S. Zhou, et al. (1999). "Expression of three Caenorhabditis elegans N- acetylglucosaminyltransferase I genes during development." J Biol Chem 274(1): 288-97.

Cipollo, J. F., A. Awad, et al. (2004). "Biosynthesis in vitro of Caenorhabditis elegans phosphorylcholine oligosaccharides." Proc Natl Acad Sci U S A 101(10): 3404-8.

Cipollo, J. F., A. M. Awad, et al. (2005). "N-Glycans of Caenorhabditis elegans are specific to developmental stages." J Biol Chem 280(28): 26063-72.

Cipollo, J. F., C. E. Costello, et al. (2002). "The fine structure of Caenorhabditis elegans N-qlycans." J Biol Chem 277(51): 49143-57.

Ciucanu, I. and C. E. Costello (2003). "Elimination of oxidative degradation during the per-O-methylation of carbohydrates." J Am Chem Soc 125(52): 16213-9.

Conde, R., G. Pablo, et al. (2003). "Screening for new yeast mutants affected in mannosylphosphorylation of cell wall mannoproteins." Yeast 20(14): 1189- 211.

Consortium, C. e. S. (1998). "Genome sequence of the nematode C. elegans: a platform for investigating biology." Science 282(5396): 2012-8.

Corbacho, I., I. Olivero, et al. (2004). "Identification of low-dye-binding (Idb) mutants of Saccharomyces cerevisiae." FEMS Yeast Res 4(4-5): 437-44. de Lonlay, P., N. Seta, et al. (2001). "A broad spectrum of clinical presentations in congenital disorders of glycosylation I: a series of 26 cases." J Med Genet 38(1): 14-9.

Dennis, J. W., M. Granovsky, et al. (1999). "Protein glycosylation in development and disease." Bioessavs 21 (5): 412-21.

Dong, G., A. H. Hutagalung, et al. (2005). "The structures of exocyst subunit Exo70p and the Exo84p C-terminal domains reveal a common motif." Nat Struct Mol Biol 12(12): 1094-100.

Dove, A. (2001). "The bittersweet promise of glycobiology." Nat Biotechnol 19(10): 913-7.

149 Duden, R. (2003). "ER-to-Golgi transport: COP I and COP II function (Review)." Mol Membr Biol 20(3): 197-207.

Eklund, E. A. and H. H. Freeze (2006). "The congenital disorders of glycosylation: a multifaceted group of syndromes." NeuroRx 3(2): 254-63.

Eisner, M., H. Hashimoto, et al. (2003). "Cisternal maturation and vesicle transport: join the band wagon! (Review)." Mol Membr Biol 20(3): 221-9.

Etzioni, A., M. Frydman, et al. (1992). "Brief report: recurrent severe infections caused by a novel leukocyte adhesion deficiency." N Engl J Med 327(25): 1789-92.

Fire, A., S. Xu, et al. (1998). "Potent and specific genetic interference by double- stranded RNA in Caenorhabditis elegans." Nature 391(6669): 806-11.

Foulquier, F., E. Vasile, et al. (2006). "Conserved oligomeric Golgi complex subunit 1 deficiency reveals a previously uncharacterized congenital disorder of glycosylation type II." Proc Natl Acad Sci U S A 103(10): 3764- 9.

Freeze, H. H. and V. Westphal (2001). "Balancing N-linked glycosylation to avoid disease." Biochimie 83(8): 791-9.

Fuster, M. M. and J. D. Esko (2005). "The sweet and sour of cancer: glycans as novel therapeutic targets." Nat Rev Cancer 5(7): 526-42.

Futai, E., S. Hamamoto, et al. (2004). "GTP/GDP exchange by Sec12p enables COPII vesicle bud formation on synthetic liposomes." Embo J 23(21): 4146-55.

Gasteiger, E., A. Gattiker, et al. (2003). "ExPASy: The proteomics server for in- depth protein knowledge and analysis." Nucleic Acids Res 31(13): 3784-8.

Griffitts, J. S., S. M. Haslam, et al. (2005). "Glycolipids as receptors for Bacillus thuringiensis crystal toxin." Science 307(5711): 922-5.

Hakomori, S. (1964). "A Rapid Permethylation of Glycolipid, and Polysaccharide Catalyzed by Methylsulfinyl Carbanion in Dimethyl Sulfoxide." J Biochem 55: 205-8.

Hanneman, A. J., J. C. Rosa, et al. (2006). "Isomer and glycomer complexities of core GlcNAcs in Caenorhabditis elegans." Glycobiology 16(9): 874-90.

Harris, S. L. and M. G. Waters (1996). "Localization of a yeast early Golgi mannosyltransferase, Ochlp, involves retrograde transport." J Cell Biol 132(6): 985-98.

150 Harvey, D. J. (1999). "Matrix-assisted laser desorption/ionization mass spectrometry of carbohydrates." Mass Spectrom Rev 18(6): 349-450.

Haslam, S. M. and A. Dell (2003). "Hallmarks of Caenorhabditis elegans N- glycosylation: complexity and controversy." Biochimie 85(1-2): 25-32.

Hirabayashi, J., K. Hayama, et al. (2002). "Affinity capturing and gene assignment of soluble glycoproteins produced by the nematode Caenorhabditis elegans." J Biochem 132(1): 103-14.

Hope, I. A. (1999). C. elegans : a practical approach. Oxford ; New York, Oxford University Press.

Hossler, P., B. C. Mulukutla, et al. (2007). "Systems analysis of N-glycan processing in mammalian cells." PLoS ONE 2(1): e713.

Hubbard, S. C. and R. J. Ivatt (1981). "Synthesis and processing of asparagine- linked oligosaccharides." Annu Rev Biochem 50: 555-83.

Huffaker, T. C. and P. W. Robbins (1982). "Temperature-sensitive yeast mutants deficient in asparagine-linked glycosylation." J Biol Chem 257(6): 3203-10.

Iwata, Y. and N. Koizumi (2005). "Unfolded protein response followed by induction of cell death in cultured tobacco cells treated with tunicamycin." Planta 220(5): 804-7.

Jaeken, J. and G. Matthijs (2001). "Congenital disorders of glycosylation." Annu Rev Genomics Hum Genet 2: 129-51.

Jaeken, J. and G. Matthijs (2007). "Congenital disorders of glycosylation: a rapidly expanding disease family." Annu Rev Genomics Hum Genet 8: 261-78.

Jaeken, J., H. G. van Eijk, et al. (1984). "Sialic acid-deficient serum and cerebrospinal fluid transferrin in a newly recognized genetic syndrome." Clin Chim Acta 144(2-3): 245-7.

Jaeken, J., M. Vanderschuerenlodeweyckx, et al. (1980). "Familial Psychomotor Retardation with Markedly Fluctuating Serum Prolactin, Fsh and Gh Levels, Partial Tbg-Deficiency, Increased Serum Arylsulfatase-a and Increased Csf Protein - New Syndrome." Pediatric Research 14(2): 179- 179.

Jamieson, J. D. and G. E. Palade (1968). "Intracellular transport of secretory proteins in the pancreatic exocrine cell. IV. Metabolic requirements." J Cell Biol 39(3): 589-603.

151 Jones, J., S. S. Krag, et al. (2005). "Controlling N-linked glycan site occupancy." Biochim Biophvs Acta 1726(2): 121-37.

Kamath, R. S., A. G. Fraser, et al. (2003). "Systematic functional analysis of the Caenorhabditis elegans genome using RNAi." Nature 421(6920): 231-7.

Kelleher, D. J. and R. Gilmore (2006). "An evolving view of the eukaryotic oligosaccharyltransferase." Glvcobioloqy 16(4): 47R-62R.

Kerek, I. C. a. F. (1984). "A simple and rapid method for the permethylation of carbohydrates " Carbohydrate Research 131(2): 209-217.

Kingsley, D. M., K. F. Kozarsky, et al. (1986). "Three types of low density lipoprotein receptor-deficient mutant have pleiotropic defects in the synthesis of N-linked, O-linked, and lipid-linked carbohydrate chains." J Cell Biol 102(5): 1576-85.

Kjaergaard, S., M. Schwartz, et al. (2001). "Congenital disorder of glycosylation type la (CDG-la): phenotypic spectrum of the R141H/F119L genotype." Arch Pis Child 85(3): 236-9.

Krieger, M., M. S. Brown, et al. (1981). "Isolation of Chinese hamster cell mutants defective in the receptor-mediated endocytosis of low density lipoprotein." J Mol Biol 150(2): 167-84.

Kubota, Y. and K. Nishiwaki (2006). "C. elegans as a model system to study the function of the COG complex in animal development." Biol Chem 387(8): 1031-5.

Kubota, Y., M. Sano, et al. (2006). "The conserved oligomeric Golgi complex acts in organ morphogenesis via glycosylation of an ADAM protease in C. elegans." Development 133(2): 263-73.

Lambie, E. J. (2002). "Cell proliferation and growth in C. elegans." Bioessavs 24(1): 38-53.

Lanoix, J., J. Ouwendijk, et al. (2001). "Sorting of Golgi resident proteins into different subpopulations of COPI vesicles: a role for ArfGAPI." J Cell Biol 155(7): 1199-212.

Lehle, L, S. Strahl, et al. (2006). "Protein glycosylation, conserved from yeast to man: a model organism helps elucidate congenital human diseases." Anaew Chem Int Ed Enal 45(41): 6802-18.

Lippincott-Schwartz, J., L. C. Yuan, et al. (1989). "Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER." Cell 56(5): 801-13.

152 Listinsky, J. J., G. P. Siegal, et al. (1998). "Alpha-L-fucose: a potentially critical molecule in pathologic processes including neoplasia." Am J Clin Pathol 110(4): 425-40.

Loh, E. and W. Hong (2004). "The binary interacting network of the conserved oligomeric Golgi tethering complex." J Biol Chem 279(23): 24640-8.

Love, H. D., C. C. Lin, et al. (1998). "Isolation of functional Golgi-derived vesicles with a possible role in retrograde transport." J Cell Biol 140(3): 541-51.

Ma, B., J. L. Simala-Grant, et al. (2006). "Fucosylation in prokaryotes and eukarvotes." Glvcobiology 16(12): 158R-184R.

Magnelli, P., J. F. Cipollo, et al. (2008). "Unique Asn-linked oligosaccharides of the human pathogen Entamoeba histolytica." J Biol Chem 283(26): 18355- 64.

Malhotra, J. D. and R. J. Kaufman (2007). "The endoplasmic reticulum and the unfolded protein response." Semin Cell Dev Biol 18(6): 716-31.

Marchi, F. and C. P. Leblond (1984). "Radioautographic characterization of successive compartments along the rough endoplasmic reticulum-Golgi pathway of collagen precursors in foot pad fibroblasts of [3H]proline- injected rats." J Cell Biol 98(5): 1705-9.

Marshall, R. D. (1974). "The nature and metabolism of the carbohydrate-peptide linkages of glycoproteins." Biochem Soc Svmp(40): 17-26.

Martinez-Menarguez, J. A., R. Prekeris, et al. (2001). "Peri-Golgi vesicles contain retrograde but not anterograde proteins consistent with the cisternal progression model of intra-Golgi transport." J Cell Biol 155(7): 1213-24.

Matsuoka, K., L. Orci, et al. (1998). "COPII-coated vesicle formation reconstituted with purified coat proteins and chemically defined liposomes." Cell 93(2): 263-75.

Matthijs, G., E. Schollen, et al. (2000). "Mutations in PMM2 that cause congenital disorders of glycosylate, type la (CDG-la)." Hum Mutat 16(5): 386-94.

Mills, P., K. Mills, et al. (2001). "Congenital disorders of glycosylation type I leads to altered processing of N-linked glycans, as well as underglycosylation." Biochem J 359(Pt 2): 249-54.

Misumi, Y., Y. Misumi, et al. (1986). "Novel blockade by brefeldin A of intracellular transport of secretory proteins in cultured rat hepatocytes." J Biol Chem 261 (24): 11398-403.

153 Miyoshi, E., K. Moriwaki, et al. (2008). "Biological function of fucosylation in cancer biology." J Biochem 143(6): 725-9.

Mohler, W. A., G. Shemer, et al. (2002). "The type I membrane protein EFF-1 is essential for developmental cell fusion." Dev Cell 2(3): 355-62.

Morava, E., H. Wosik, et al. (2008). "Congenital disorder of glycosylation type Ix: review of clinical spectrum and diagnostic steps." J Inherit Metab Pis 31(3): 450-6.

Munson, M. and F. M. Hughson (2002). "Conformational regulation of SNARE assembly and disassembly in vivo." J Biol Chem 277(11): 9375-81.

Nairn, A. V., W. S. York, et al. (2008). "Regulation of glycan structures in animal tissues: transcript profiling of glycan-related genes." J Biol Chem 283(25): 17298-313.

Natsuka, S., J. Adachi, et al. (2002). "Structural analysis of N-linked glycans in Caenorhabditis elegans." J Biochem 131(6): 807-13.

Nguyen, K., I. van Die, et al. (2007). "Molecular cloning and characterization of the Caenorhabditis elegans alphal ,3-fucosyltransferase family." Glvcobioloav 17(6): 586-99.

Nilsson, T., M. Pypaert, et al. (1993). "Overlapping distribution of two glycosyltransferases in the Golgi apparatus of HeLa cells." J Cell Biol 120(1): 5-13.

Ohtsubo, K. and J. D. Marth (2006). "Glycosylation in cellular mechanisms of health and disease." Ceil 126(5): 855-67.

Oka, T., E. Vasile, et al. (2005). "Genetic analysis of the subunit organization and function of the conserved oligomeric golgi (COG) complex: studies of COG5- and COG7-deficient mammalian cells." J Biol Chem 280(38): 32736-45.

Opat, A. S., C. van Vliet, et al. (2001). "Trafficking and localisation of resident Golgi glycosylation enzymes." Biochimie 83(8): 763-73.

Paschinger, K., M. Gutternigg, et al. (2008). "The N-glycosylation pattern of Caenorhabditis elegans." Carbohvdr Res 343(12): 2041-9.

Paschinger, K., M. Hackl, et al. (2006). "A deletion in the golgi alpha- mannosidase II gene of Caenorhabditis elegans results in unexpected non-wild-type N-glycan structures." J Biol Chem 281(38): 28265-77.

Pelham, H. R. and J. E. Rothman (2000). "The debate about transport in the Golgi-two sides of the same coin?" CeN 102(6): 713-9.

154 Prien, J. M., Ashline, D.J., Lapadula, A.J., Zhang, H., Reinhold, V.N. (2009). "The High Mannose GLycans from Bovine Ribonuclease B; Isomer Characterization by Ion Trap MS." J. Am. Soc. Mass Spectrom. 20(in press).

Prietsch, V., V. Peters, et al. (2002). "A new case of CDG-x with stereotyped dystonic hand movements and optic atrophy." J Inherit Metab Pis 25(2): 126-30.

Rabouille, C, N. Hui, et al. (1995). "Mapping the distribution of Golgi enzymes involved in the construction of complex oligosaccharides." J Cell Sci 108 ( Pt4): 1617-27.

Ram, R. J., B. Li, et al. (2002). "Identification of Sec36p, Sec37p, and Sec38p: components of yeast complex that contains Sec34p and Sec35p." Mol Biol CeJ 13(5): 1484-500.

Raman, R., S. Raguram, et al. (2005). "Glycomics: an integrated systems approach to structure-function relationships of glycans." Nat Methods 2(11): 817-24.

Reboul, J., P. Vaglio, et al. (2003). "C. elegans ORFeome version 1.1: experimental verification of the genome annotation and resource for proteome-scale protein expression." Nat Genet 34(1): 35-41.

Reddy, P. and M. Krieger (1989). "Isolation and characterization of an extragenic suppressor of the low-density lipoprotein receptor-deficient phenotype of a Chinese hamster ovary cell mutant." Mol Cell Biol 9(11): 4799-806.

Reinhold, V. N., B. B. Reinhold, et al. (1996). "Carbohydrate sequence analysis by electrospray ionization-mass spectrometry." Methods Enzymol 271: 377-402.

Reinhold, V. N., B. B. Reinhold, et al. (1995). "Carbohydrate molecular weight profiling, sequence, linkage, and branching data: ES-MS and CID." Anal Chem 67(11): 1772-84.

Reinhold, V. N. and D. M. Sheeley (1998). "Detailed characterization of carbohydrate linkage and sequence in an ion trap mass spectrometer: glycosphingolipids." Anal Biochem 259(1): 28-33.

Rual, J. F., J. Ceron, et al. (2004). "Toward improving Caenorhabditis elegans phenome mapping with an ORFeome-based RNAi library." Genome Res 14(10B): 2162-8.

Schabussova, I., H. Amer, et al. (2007). "O-methylated glycans from Toxocara are specific targets for antibody binding in human and animal infections." Int J Parasitol 37(1): 97-109.

155 Schachter, H. (2004). "Protein glycosylation lessons from Caenorhabditis elegans." Curr Opin Struct Biol 14(5): 607-16.

Schachter, H., S. Chen, et al. (2002). "Functional post-translational proteomics approach to study the role of N-glycans in the development of Caenorhabditis elegans." Biochem Soc Svmp(69): 1-21.

Schollen, E., S. Kjaergaard, et al. (2000). "Lack of Hardy-Weinberg equilibrium for the most prevalent PMM2 mutation in CDG-la (congenital disorders of glycosylation type la)." Eur J Hum Genet 8(5): 367-71.

Schollen, E., S. Kjaergaard, et al. (2004). "Increased recurrence risk in congenital disorders of glycosylation type la (CDG-la) due to a transmission ratio distortion." J Med Genet 41 (11): 877-80.

Schroder, M. and R. J. Kaufman (2005). "The mammalian unfolded protein response." Annu Rev Biochem 74: 739-89.

Sheeley, D. M. and V. N. Reinhold (1998). "Structural characterization of carbohydrate sequence, linkage, and branching in a quadrupole Ion trap mass spectrometer: neutral oligosaccharides and N-linked glycans." Anal Chem 70(14): 3053-9.

Shen, X., R. E. Ellis, et al. (2001). "Complementary signaling pathways regulate the unfolded protein response and are required for C. elegans development." CeH 107(7): 893-903.

Shestakova, A., S. Zolov, et al. (2006). "COG complex-mediated recycling of Golgi glycosyltransferases is essential for normal protein glycosylation." Traffic 7(2): 191-204.

Simmer, F., C. Moorman, et al. (2003). "Genome-wide RNAi of C. elegans using the hypersensitive rrf-3 strain reveals novel gene functions." PLoS Biol 1(1): E12.

Spaapen, L. J., J. A. Bakker, et al. (2005). "Clinical and biochemical presentation of siblings with COG-7 deficiency, a lethal multiple O- and N-glycosylation disorder." J Inherit Metab Pis 28(5): 707-14.

Srivastava, S. (2008). "Move over proteomics, here comes glycomics." J Proteome Res 7(5): 1799.

Sturla, L, R. Rampal, et al. (2003). "Differential terminal fucosylation of N-linked glycans versus protein O-fucosylation in leukocyte adhesion deficiency type II (CDG lie)." J Biol Chem 278(29): 26727-33.

Takeuchi, M. and A. Kobata (1991). "Structures and functional roles of the sugar chains of human erythropoietins." Glvcobioloqy 1(4): 337-46.

156 Tanabe, K., A. Deguchi, et al. (2008). "Outer arm fucosylation of N-glycans increases in sera of hepatocellular carcinoma patients." Biochem Biophys Res Commun 374(2): 219-25.

Tatusov, R. L, N. D. Fedorova, et al. (2003). "The COG database: an updated version includes eukaryotes." BMC Bioinformatics 4: 41.

Taylor, M. E. and K. Drickamer (2003). Introduction to qlycobioloqy. Oxford ; New York, Oxford University Press.

Thorpe, C. J., A. Schlesinger, et al. (2000). "Wnt signalling in Caenorhabditis elegans: regulating repressors and polarizing the cytoskeleton." Trends Cell Biol 10(1): 10-7.

Ungar, D., T. Oka, et al. (2002). "Characterization of a mammalian Golgi- localized protein complex, COG, that is required for normal Golgi morphology and function." J Cell Biol 157(3): 405-15.

Ungar, D., T. Oka, et al. (2006). "Retrograde transport on the COG railway." Trends Cell Biol 16(2): 113-20.

Vaneijk, H. G., W. L. Vannoort, et al. (1983). "The Microheterogeneity of Human Transferrins in Biological-Fluids." Clinica Chimica Acta 132(2): 167-171.

Varki, A. (1993). "Biological roles of oligosaccharides: all of the theories are correct." Glvcobioloqy 3(2): 97-130.

Varki, A. (1999). Essentials of qlvcobioloqy. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press.

Vodopiutz, J. and O. A. Bodamer (2008). "Congenital disorders of glycosylation-a challenging group of lEMs." J Inherit Metab Pis.

Wada, Y., P. Azadi, et al. (2007). "Comparison of the methods for profiling glycoprotein glycans-HUPO Human Disease Glycomics/Proteome Initiative multi-institutional study." Glycobioloqy 17(4): 411-22.

Warren, C. E., A. Krizus, et al. (2002). "The Caenorhabditis elegans gene, gly-2, can rescue the N-acetylglucosaminyltransferase V mutation of Lec4 cells." J Biol Chem 277(25): 22829-38.

Westphal, V., S. Kjaergaard, et al. (2002). "A frequent mild mutation in ALG6 may exacerbate the clinical severity of patients with congenital disorder of glycosylation la (CDG-la) caused by phosphomannomutase deficiency." Hum Mol Genet 11 (5): 599-604.

Whaley, W. G. (1975). The Golgi apparatus. New York, Springer-Verlag.

157 Whyte, J. R. and S. Munro (2001). "The Sec34/35 Golgi transport complex is related to the exocyst, defining a family of complexes involved in multiple steps of membrane traffic." Dev Cell 1(4): 527-37.

Whyte, J. R. and S. Munro (2002). "Vesicle tethering complexes in membrane traffic." JCellScM15(Pt 13): 2627-37.

Wu, X., J. S. Rush, et al. (2003). "Deficiency of UDP-GlcNAc:Dolichol Phosphate N-Acetylglucosamine-1 Phosphate Transferase (DPAGT1) causes a novel congenital disorder of Glycosylation Type Ij." Hum Mutat 22(2): 144-50.

Wu, X., R. A. Steet, et al. (2004). "Mutation of the COG complex subunit gene COG7 causes a lethal congenital disorder." Nat Med 10(5): 518-23.

Zeevaert, R., F. Foulquier, et al. (2008). "Deficiencies in subunits of the Conserved Oligomeric Golgi (COG) complex define a novel group of Congenital Disorders of Glycosylation." Mol Genet Metab 93(1): 15-21.

Zhu, S., A. Hanneman, et al. (2004). "Caenorhabditis elegans triple null mutant lacking UDP-N-acetyl-D-glucosamine:alpha-3-D-mannoside betal ,2-N- acetylglucosaminyltransferase I." Biochem J 382(Pt 3): 995-1001.

Zhu, X., Y. Zeng, et al. (1992). "Evidence that the hamster tunicamycin resistance gene encodes UDP-GlcNAc:dolichol phosphate N- acetylglucosamine-1-phosphate transferase." J Biol Chem 267(13): 8895- 902.

158 APPENDICES

159 APPENDIX A

TUNICAMYCIN HYPERSENSITIVE LOCI

The genome-wide RNAi screen resulted in 512 genes that are hypersensitive to tunicamycin (Table A1). The C. elegans gene sequence name and gene name are given. As stated in Chapter 5.1, the effects of RNAi penetrance and expressivity were measured. The values below (Exp and Pen) are the difference between drug treated scores from non-drug treated scores.

Additionally, the presence of A/-glycosylation sites and cellular location were predicted as stated in Chapter 5.3. Short descriptions, when available, were also provided.

160 WS1« CGC Name Exp Pen NetNGlyc location Description AC8.6 10 20 No uncharacterized AH10.1 10 30 No cytoplasm AMP-dependent synthetase and ligase B0035.11 25 -5 No nucleus Yeast LE01 protein like B0035.16 0 20 No cytoplasm tRNA methyl transferase B0035.8 his-48 10 20 No nucleus H2B histone B0205.6 25 5 No aminotransferase B0280.5 0 20 No extracellular protein with six chitin-binding peritrophin-A domains and three mucin-like regions B0284.1 30 30 No cytoplasm Uncharacterized protein B0285.1 0 20 No nucleus serine/threonine kinase B0285.9 ckb-2 5 15 No cytoplasm encodes an isoform of choline kinase B0302.4 0 20 No uncharacterized B0334.4 20 15 No Protein similar to predicted member of the intramitochondrial sorting protein family B0361.9 25 25 Yes extracellular uncharacterized B0365.3 eat-6 20 30 No ER ortholog of the alpha subunit of a sodium/potassium ATPase B0391.5 20 30 No cytoplasm F-box protein B0412.1 dac-1 20 20 No extracellular yeast NUD1 protein B0464.1 drs-1 30 35 No cytoplasm putative aspartyl(D) tRNA synthetase B0464.4 bre-3 10 25 No plasma membrane protein similar to beta-glycosyltransferases B0464.7 baf-1 10 20 No nucleus novel protein that binds double-stranded DNA nonspecifically B0491.5 0 30 No uncharacterized B0545.1 tpa-1 20 20 Yes cytoplasm serine/threonine protein kinase C01F1.2 0 20 No mitochondrion orthologous to the human gene SCO (CYTOCHROME OXIDASE DEFICIENT YEAST) C01G10.6 10 20 No uncharacterized C01G5.5 0 15 Yes cytoplasm aldehyde reductase C01G8.6 35 15 Yes extracellular uncharacterized C01H6.9 30 25 No cytoplasm homolog of haploid germ cell-specific nuclear protein kinase C02B10.2 20 20 Yes orthologous to Yeast Kinetochore-associated protein required chromosome segregation C04C3.3 0 30 No pyruvate dehydrogenase C04G2.2 0 20 No cytoplasm Serine/threonine protein kinase C04H5.6 mog-4 20 15 No nucleus encodes a DEAH helicase C05D9.1 snx-1 30 30 No cytoplasm Membrane coat complex Retromer and related PX domain-containing proteins C06A1.3 10 20 No cytoplasm serine/threonine protein phosphatase C06A8.4 skr-17 10 15 No nucleus homolog of Skp1 in S. cerevisiae, a core component of the SCF (Skplp, Cullin, F-box) C06A8.5 20 20 No cytoplasm uncharacterized C06C3.1 mel-11 10 25 No cytoplasm ortholog of the vertebrate smooth muscle myosin-associated phosphatase regulatory subunit C06C6.7 10 20 No ER uncharacterized C06E1.4 glr-1 10 30 No plasma membrane encodes an AMPA (non-NMDA)-type ionotropic glutamate receptor subunit C06E4.6 0 15 No peroxisome dehydrogenase C06H2.4 folt-1 10 20 No plasma membrane encodes a folate transporter required for folate uptake C06H5.7 10 20 No plasma membrane uncharacterized j5> o O CO o •o it: CD cr ! O D) CD

•a te ) c CO 'CD "co -D O O o is 3 E Q. CO in s c ns p CD a CO CD CO CD I O CO CD CO c < CO n > 0) c o> 'o 1*: B , 1 E c CO c CD a. O 73 _Q. iq u c .a O co CO 3 CM CD C CM O li CO E CO CM CO Q. CD c CL c 73 73 c C £ CO o 3 S '55 CD -rj X O £1 CO £2 3 CO •§o 9I- I CO co p CD CO CO O I m CO >• ram E CD CO C3) 2 => i. CO CD CD t * -D 0- o CD ro c E o > CD t t 3 CD Q. *- o £:§ to < CO CD C I F 2> O 3 I Q."0 si CD o xa o V CD Q CD S CO 8" 2 <« P o. c ° c co O o o CO CD -g •¥ E Q.QO- c £ s co •= 5 CD CO CO CD 0> -O « & S o CO 'co p CD CrjD> CO m & (0 -r- 3 g co — fD E o. CD C•tO? -i c cr o. CO O Q. o o T> J2 'CD "O 3 73 CO Q. T3 •D Q. CD CO CD CD X3 CD E .N LX.SJ 3 .Ri 2 ° .N o ^Z CO a. 3 3 CO C E o f'l -C CO >_ CD CD cr c c CO ^ i*S rpro t CD i= _ ° ° cE terize an d u terize ilco — l-ft •=; ill CO O CD U S CD o -o o It c> E o CD CO D) CO si o u I o to co oi a en o .•= E S CD -S D) SO _ <» 11 CD Jo t: O O O 3 CO ~ CO co ^ o co CO CD CX 8 S o -c 2 8- O o CO « T3 C MM CO •§ o* Cl)_ — w O O CO ^ K: TJ 5? •- "- co • ° ^ 5 o * 73 co o -° §9 §•5 CD CO j-i •c c CD CO CD "o . ' O a) el) -C £hSS £ S c- .<= c o -c •C c XJ £ S CD > N Is- b -c CO 3 3 N 3 J= < O CD 'C 3 CL £ 3 co Q g> = E>E>g _> ° a O 3 CD 2 O Z 6§ _j !c rai? iS 3 ' CD CD 3 CD C CO CO CO CO co E E E -1 CO EB| E CO T3 ~i CO CO CO 3 E 3 co t u — co J 1 s CO C CO CO Q C J3) CO 5; Q- _co _co _co -^ CD CD _ CD 0 <*> EC JD — ^ QO Q. C O O Q. ^ O -S

jeleu s acellul a Em a Jo E •~ SP . C_ ll CO c c co =" <= O %%%-O O X E 5- o E .t; o E E co p CO co t JO CD-

So 2 oooooooifioooooooooooooKSSoooiSooiSooSooSoK

2 _ooooooooooooooooooo coooooocomoooooooooo CM CO CO COCM CO •^•^•^COCO CMCMT- «*••>- Nr.-rT- CM ^- (M CM r- CM CM r- m

m i U5 CM Cp N. CM c V ci co ra c f •a c f co S- o uo

Mac CMU) CDOJT.- — - T- CM i- t- CM CO ^ TTCO CM IT) CMS lCMCVjCMCOlOj^^-i-CM^-COlO- OO^CO^CNJCNJcM^dCD^CM^^CDrooo^^^^CMCM™^ CO Ôidî-^^Lris^ôS^^COCViCOT-r-T-^^T-iom^T-T-T-t-T-CjjCDCDlOi-T-COlO 1UUIIDIOL1JD[I1<

162 WS112 CGCName Exp Pen NetNGIyc location Description C32E8.5 0 20 No nucleus Transcriptional regulator SNIP1, contains FHA domain C32F10.1 obr-4 0 20 No cytoplasm oxysterol-binding protein C33A12.13 sru-2 5 15 Yes plasma membrane uncharacterized C33F10.2 5 35 No cytoplasm Cell division control protein C33F10.4 0 30 Yes Serine/threonine kinase TIP30/CC3 C34B2.10 30 No ER uncharacterized C34E10.6 atp-2 0 30 No mitochondrion beta subunit of the soluble, catalytic F1 portion of ATP synthase C34E11.1 rsd-3 15 5 Yes extracellular uncharacterized C35B1.1 ubc-1 0 30 No nucleus E2 ubiquitin-conjugating enzyme C35B8.2 vav-1 0 20 No cytoplasm uncharacterized C35D10.13 15 15 No uncharacterized C35D10.2 10 20 No uncharacterized C35D6.4 0 30 Yes nucleus CCGH-type Zn-finger protein C36B1.4 pas-4 -5 15 No nucleus 20S proteasome C37C3.6 ppn-1 15 -5 No extracellular encodes three isoforms of a large, novel multidomain glycoprotein C37H5.1 nex-4 5 15 No plasma membrane encodes a predicted annexin. C41C4.4 ire-1 0 20 No cytoplasm transmembrane serine/threonine protein kinase required for the unfolded protein response C42C1.14 rpl-34 15 25 No cytoplasm large ribosomal subunit L34 protein that affects growth and body coloration C43E11.7 ndx-7 10 30 No cytoplasm NUDIX hydrolase predicted to function as a pyrophosphatase C46E10.7 srh-99 10 20 No 7TM chemoreceptor C47A4.5 0 20 No nucleus uncharacterized C47B2.6 20 20 No UDP-glucose 4-epimerase/UDP-sulfoquinovose synthase C47D12.8 30 30 No nucleus ortholog to the human gene DNA Repair Protein C47E12.4 0 35 No cytoplasm inorganic pyrophosphatase C47E12.7 20 20 No nucleus Nucleolar protein NOP52/RRP1 C48D5.1 nhr-6 25 25 Yes nucleus nuclear recepter of the NR4A4 subfamily C48E7.2 10 30 No uncharacterized C49A9.6 10 15 No uncharacterized C49H3.11 rps-2 10 20 No cytoplasm small ribosomal subunit S2 protein C49H3.4 10 20 No like-Sm ribonucleoprotein C49H3.8 20 30 No cytoplasm Heat shock protein 70 C50A2.3 30 30 No uncharacterized C50B6.2 15 15 No nucleus histone binding protein C50F7.4 5 25 No mitochondrion succinyl-CoA synthetase beta chain C52D10.13 0 20 No extracellular cuticular collagen C52D10.2 10 20 No uncharacterized C52E4.3 snr-4 10 15 No nucleus small nuclear ribonucleoprotein C53D5.6 imb-3 0 25 No cytoplasm nuclear transport factor that regulates nuclear import of ribosomal proteins C53H9.1 rpl-27 0 20 No cytoplasm large ribosomal subunit L27 protein. C53H9.2 30 30 No cytoplasm small ribosomal subunit S2 protein WS112 CGCName Exp Pen NotNGIyc location Description C54G4.1 10 20 No cytoplasm uncharacterized C55A6.7 0 30 No cytoplasm alcohol dehydrogenase C56C10.6 0 20 No Casein kinase C56C10.8 icd-1 0 25 No nucleus ortholog of the beta-subunit of the nascent polypeptide-associated complex (betaNAC) C56C10.9 15 15 Yes uncharacterized D1007.1 ceh-17 0 20 No nucleus phox-2-like homeodomain protein D1007.13 20 20 No nucleus uncharacterized D1009.2 cyn-8 20 20 Yes cytoplasm peptidyl-propyl cis-trans isomerase of the cyclophilin family implicated in protein folding D1081.7 10 25 No uncharacterized D2021.8 -10 20 Yes cytoplasm Ankyrin repeat and DHHC-type Zn-finger domain containing proteins D2024.3 elo-3 -10 25 No ER paralog of elo-1 and elo-2, which encodes a polyunsaturated fatty acid (PUFA) elongase D2030.10 aex-1 20 20 No novel C2 domain protein; C2 is a proposed calcium binding domain D2030.3 0 20 Yes nucleus uncharacterized DY3.1 tin-13 0 20 No mitochondrion uncharacterized E01G4.2 0 30 No uncharacterized E02A10.1 15 15 Yes mitochondrion mitochondrial 40S ribosomal protein S5 E04A4.4 hoe-1 10 20 Yes cytoplasm ortholog to human ELAC2 associated with prostate cancer E04F6.9 0 30 No uncharacterized F01D4.3 0 20 No cytoplasm SH2 domain tyrsine protein kinase (FFES/FPS subfamily) F01 D5.5 0 20 No extracellular uncharacterized F02A9.3 far-2 20 20 No encodes a protein similar to a class of secreted fatty acid and retinol-binding proteins F02D10.5 flr-1 15 15 Yes plasma membrane sodium channel like F02E8.1 asb-2 20 30 No mitochondrion ATP synthase B homolog F07A5.7 unc-15 0 20 No cytoplasm paramyosin ortholog that interacts with MHC A, one isoform of myosin heavy chain (MHC) F07E5.2 fbxb-35 0 15 Yes protein containing an F-box, a motif predicted to mediate protein-protein interactions F07H5.6 20 10 No nucleus uncharacterized F08B4.5 10 20 No nucleus uncharacterized F08C6.2 0 25 No ER uncharacterized F08F3.3 rhr-1 30 30 No plasma membrane orhtolog of human Rhesus Blood-group associated glycoprotein F08G12.1 5 15 No cytoplasm ras-related like protein F09E5.1 pkc-3 15 15 Yes cytoplasm Protein kinase C F09G8.3 30 -5 Yes mitochondrion Mitochondrial ribosomal protein S9 F10C2.6 drs-2 0 20 No cytoplasm putative aspartyl-tRNA synthetase F10D11.1 sod-2 20 20 No mitochondrion iron/manganese superoxide dismutase F10E9.4 20 20 No nucleus uncharacterized F10G7.11 0 30 No uncharacterized F11A10.1 lex-1 10 20 No cytoplasm TAT-binding homolog like F11A3.2 10 20 No cytoplasm uncharacterized F11A5.10 glc-1 20 20 Yes plasma membrane encodes an alpha subunit of a glutamate-gated chloride channel F11C1.2 15 15 No uncharacterized WS112 CGC Name Exp Pen NetNGlyc location Description F11C1.3 10 20 No lysosome membrane glycoprotein F12B6.3 0 30 Yes plasma membrane Integral membrane ankyrin-repeat protein Kidins220 (protein kinase D substrate) F14E5.5 20 30 No Triacylglycerol lipase F16B12.6 0 20 No uncharacterized F16G10.6 10 20 No uncharacterized F16H6.2 0 15 No Golgi C-type lectin domain F19B6.2 ufd-1 20 20 Yes cytoplasm Ubiquitin fusion-degradation protein F20D6.8 20 10 No plasma membrane RAS-like protein F20H11.6 10 20 No extracellular uncharacterized F21D12.1 nhr-21 0 15 No nucleus modulating N-terminal of steroid/thyroid/retinoic nuclear hormone receptors F21 E9.3 20 20 No extracellular uncharacterized F22B7.1 20 25 Yes uncharacterized F22F7.1 0 20 No Saccharopine dehydrogenase F23B12.7 30 0 No nucleus CCAAT binding factor 1 protein like F23B12.9 egl-1 10 20 No programmed cell death activator F25B3.6 10 15 No nucleus uncharacterized F25B4.6 30 20 No mitochondrion hydroxymethlglutaryl-CoA synthase F25C8.4 0 20 No cytoplasm AMP-binding enzyme F25G6.3 acr-16 0 20 No plasma membrane alpha-7-like homomer-forming subunit of the nicotinic acetylcholine receptor F25H2.9 pas-5 10 35 No nucleus proteasome zeta chain F25H5.6 20 0 No uncharacterized F26A1.1 10 35 No uncharacterized F26B1.2 0 20 No nucleus RNA-binding protein F26D11.11 let-413 0 20 No encodes a protein that is required for the assembly of adherens junctions F26E4.9 cco-1 30 20 No mitochondrion Cytochrome c oxidase, subunit Vb/COX4 F26F2.6 0 30 Yes Cysteine rich C-type lectin F26F4.10 rrt-1 0 15 No cytoplasm arginyl-tRNA synthetase that affects embryonic viability and fertility F26H11.3 20 20 No uncharacterized F26H9.6 rab-5 25 35 No plasma membrane rab related protein of the Ras GTPase F27C8.2 10 20 No uncharacterized F27D4.2 0 30 No nucleus uncharacterized F27D4.6 20 20 Yes zinc-finger protein F28A10.3 0 20 No uncharacterized F28C12.5 sra-21 20 30 Yes plasma membrane 7TM chemoreceptore, sra family F28H6.1 akt-2 20 20 No cytoplasm homolog of the serrine/threonine kinase Akt/PKB F29C12.4 10 35 No mitochondrion Translation elongation factor EFG/EF2 F29D10.1 10 30 No nucleus uncharacterized F29D10.5 0 20 No uncharacterized F30A10.3 10 20 No cytoplasm Inositol polyphosphate kinase F31C3.5 0 25 Yes nucleus GINS complex, Psf2 component WS112 CGC Name Exp Pen NetNGIyc location Description F31C3.6 20 20 Yes uncharacterized F32A11.3 35 35 No nucleus uncharacterized F32D8.6 emo-1 20 20 No ER ortholog of Sec61p, required for translocation of secreted and membrane proteins into the ER F33D11.10 25 25 No intiation factor/helicase F33D11.3 col-54 0 20 No extracellular N-terminal collagen triple helix repeat F35G2.3 0 20 No cytoplasm uncharacterized F35H8.3 20 20 No nucleus Zinc finger, C2H2 type F36A2.6 rps-15 20 20 No cytoplasm small ribosomal subunit S15 protein F36A2.7 0 30 No uncharacterized F36F2.2 10 20 No uncharacterized F36H1.2 tag-144 10 20 Yes cytoplasm Integral membrane ankyrin-repeat protein Kidins220 (protein kinase D substrate) F36H1.5 0 20 No uncharacterized F37A8.1 25 25 No uncharacterized F37B4.7 folt-2 10 20 No plasma membrane encodes a putative folate transporter F37C12.11 rps-21 0 20 No cytoplasm small ribosomal subunit S21 protein F37E3.3 15 35 No Protein kinase-like F38A1.6 0 20 No C-type lectin superfamily member F38A3.2 ram-2 0 30 No extracellular cuticle collagen that affects ray cell migration F38E1.7 mom-2 10 35 No extracellular member of the Wnt family of secreted signaling glycoproteins F39E9.5 5 15 No mariner transposase F40B1.2 20 10 No cytoplasm uncharacterized F40G12.11 20 20 No Yeast orthologous involved in the translocation of macromoiecules between the nucleoplasm F41C3.8 20 10 No uncharacterized F41E7.7 0 20 No uncharacterized F42A10.4 efk-1 0 35 No cytoplasm calcium/calmodulin-dependent protein kinase F42A10.5 15 15 No uncharacterized F42A6.7 hrp-1 10 20 No nucleus RNA-binding protein F42G4.6 10 20 No uncharacterized F42G8.10 10 20 No uncharacterized F44A2.1 tag-153 20 30 No nucleus uncharacterized F44F4.2 0 20 No extracellular Protein-tyrosine phosphatase F44G3.10 20 10 No claudin homolog required for cohesion of apical junctions in epithelia F45E4.11 20 30 No plasma membrane permease of the major facilitator superfamily F45E4.7 10 20 No extracellular uncharacterized F46A9.3 twk-29 0 15 Yes uncharacterized F46C3.1 pek-1 10 20 No ER protein kinase orthologous to human eukaryotic translation initiation factor 2-alpha kinase 3 F46E10.9 dpy-11 20 25 No ER membrane-associated thioredoxin-like (TRX) protein F46F11.4 ubl-5 0 30 Yes ortholog of the human ubiquitin-like gene UBL5 F46F3.4 ape-1 10 20 No cytoplasm encodes an ortholog of inhibitory p53-interacting protein (iASPP) F47B10.8 20 20 No uncharacterized WS11? CGCName Exp Pen NetNGIyc location Description F47B3.2 0 20 No cytoplasm Protein tyrosine phosphatase F47B7.5 0 15 No nucleus uncharacterized F47G4.2 20 0 No plasma membrane uncharacterized F47G9.1 srf-8 10 20 No Golgi uncharacterized F47G9.3 20 20 No uncharacterized F47H4.7 0 20 No cytoplasm predicted to mediate protein-protein interactions either with homologs of yeast Skp-1p F48C1.5 20 35 No uncharacterized F48E8.5 paa-1 20 0 No cytoplasm protein phosphatase F49C12.2 0 30 No uncharacterized F49C12.4 0 30 No uncharacterized F49C12.5 0 20 Yes DUF23 type extracellular protein with conserved cysteines F49C12.8 rpn-7 25 35 No cytoplasm non-ATPase subunit of the 19S regulatory complex of the proteasome F52B11.3 noah-2 -10 15 No uncharacterized F52B5.6 rpl-25.2 0 15 No cytoplasm large ribosomal subunit L23a protein F52C6.1 10 20 No cytoplasm uncharacterized F52C6.11 25 35 No cytoplasm uncharacterized F52E10.5 ifa-3 30 10 Yes cytoplasm coiled intermediate filament protei F52H2.3 10 20 No uncharacterized F53A3.3 rps-22 20 20 Yes cytoplasm ribosomal subunit S15a protein F53C11.7 swan-2 10 20 No nucleus Yeast hypothetical protein YPL247C like F53G12.10 rpl-7 30 30 No cytoplasm ribosomal protein L30 F54C8.4 15 15 No nucleus protein-tyrosine phosphatase F54C8.5 15 15 No plasma membrane Ras small GTPase F54C9.2 stc-1 0 20 No ER Heat shock 70Kd protein F54D10.3 0 30 Yes uncharacterized F54D10.6 10 20 No uncharacterized F54E7.2 rps-12 10 20 No cytoplasm ribosomal subunit S12 protein F54E7.5 20 20 No uncharacterized F55A12.3 ppk-1 25 35 No cytoplasm Phosphatidylinositol-4-phosphate 5-kinase F55D10.2 rpl-25.1 15 25 No cytoplasm Ribosomal protein L23 F55G11.5 dod-22 0 15 No uncharacterized F55H2.2 vha-14 10 30 No mitochondrion ortholog of subunit D of the cytoplasmic (V1) domain of vacuolar proton-translocating ATPase F56A12.1 unc-39 0 20 No nucleus homeodomain transcription factor that belongs to the Six4/5 family F56B3.10 gst-40 35 35 No cytoplasm glutathione S-transferase F56E10.1 0 20 Yes nucleus uncharacterized F56F10.2 0 20 No uncharacterized F57B1.2 sun-1 0 20 No nucleus nuclear envelope receptor for CED-4 during apoptosis, and is bound by CED-4 in vitro F57B9.6 inf-1 15 5 No nucleus uncharacterized ^ F58B4.1 nas-31 0 20 No zinc metalloprotease F58B4.5 20 20 No ER uncharacterized WS112 CGCName Exp Pen 1YetNGly c location Description F58G1.2 25 25 No nucleus Zinc finger, C2H2 type F59A1.13 10 20 No plasma membrane uncharacterized F59A1.4 str-89 0 20 No plasma membrane 7-transmembrane olfactory receptor F59A3.3 20 10 No uncharacterized F59B2.11 20 20 Yes nucleus uncharacterized F59C6.5 30 5 No mitochondrion NADH ubiquinone oxidoreductase F59E10.3 -10 15 No Golgi subunit of the coatomer (COPI) complex H02I12.6 his-66 30 20 No nucleus H2B histone H04J21.3 gip-1 10 20 No gamma-tubulin-binding protein H06H21.3 15 25 No cytoplasm translation initiation factor H12I19.3 10 35 No uncharacterized H13N06.3 gob-1 25 25 Yes trehalose-6-phosphatase H23N18.4 10 30 Yes ER Hexosyltransferase H39E20.1 10 20 No uncharacterized JC8.11 0 20 No uncharacterized K01A6.4 0 20 No nucleus uncharacterized K01C8.6 30 0 No cytoplasm uncharacterized K01C8.9 nst-1 0 30 No nucleus encodes a homolog of human FLJ10613 and nucleostemin K02D10.5 20 20 No plasma membrane SNAP-25 (synaptosome-associated protein) component of SNARE complex K02D3.1 0 25 No uncharacterized K02F2.3 tag-203 25 25 Yes nucleus Splicing factor 3b, subunit 3 K04A8.10 30 30 No ER UDP-glucuronosyltransferase K04G2.1 20 20 No cytoplasm translational initiation factor 2 beta subunit K04G7.11 20 0 No cytoplasm uncharacterized K05C4.1 pbs-5 35 35 No cytoplasm 20S proteasome, regulatory subunit beta type PSMB5/PSMB8/PRE2 K06H7.6 apc-2 0 20 No cytoplasm subunit 2 of APC, a cyclin-specific E3 RING ubiquitin ligase K06H7.8 0 20 No cytoplasm Casein kinase (serine/threonine/tyrosine protein kinase) K07C5.8 35 35 No cytoplasm nuclear autoantigen K07E8.3 25 25 Yes nucleus Single-stranded DNA-binding replication protein A (RPA) K07H8.1 10 20 No cytoplasm uncharacterized K08C9.2 20 20 No uncharacterized K08E3.1 tyr-2 0 20 No extracellular tyrosinase K08E3.6 cyk-4 0 20 No cytoplasm GTPase-activator protein for Rho-like GTPases K08E3.7 pdr-1 0 20 No nucleus E3 ubiquitin-protein ligase K08F4.5 5 35 No nucleus uncharacterized K09A9.5 gas-1 10 20 No mitochondrion NADH-ubiquinone oxidoreductase 49KD subunit K09D9.11 0 20 No nucleus uncharacterized K09E10.2 10 20 No mitochondrion integral membrane ankyrin-repeat protein Kidins220 (protein kinase D substrate) K09E4.2 10 30 No ER mannosyltransferase K10C2.2 25 35 No uncharacterized O £> JD

•o E c CD < CO D) C CD Q. < en J5 a < CD Q_ co n > CD co T3 •^^ CD CM —- co CD CO -C CO O •* sz C O aj Q. co co o c c CD If) 1- ~= W CD CD >- O CD CO xz .E o •E g c S Q. E 3 0 £ co — o 2 2 **-•*- CO 2 ai o jo 0> . le x ! ex E ^1- Q. su b o CO Q. 3-1 °-> CO E co a) o> g f a. 0 0 ia I .E o .E DC 0 i CD CO 1151 CD c o 0,-2 E S c O O ,_ £• C CD LU 0 I >. o > <3. ^ 18 O CD CD co FO CO m £ CO 3 .<» = 2 if co E o-co I CO C31 is •£, E S CO 1 2- O O CM i c CD CD Sfo s E CD C c 5 co = g-' O cn 2 co CO CO O -~ CO „ SI CD T3 T3 0 g-c 5 "O a o) a •Q tj fW tj T3 •& a> "O S5 O CO - 1 £ "o CD CD •O .N .N CO ~ SP — ^ CO CD S) 3 fS CD CD >. CD Rj ^ 8 « SJ .N .N SI CD C C CO -g CO '£; .«£ CO J g CO j= CD »- cp (a — ±= 4=i iP C.fjO . j CO CD 10 _ S 5J ~0 M_ 5> J> c: c CD CD ° ° ° t> o V o « i° CO co _o> _co (u _co _co Q i5 « "CD 2> « CD g -co -2>iS Q. O." 73 73 O Q. Q. O Q. Q. Q. o O o O Ori a. g E w o o 330S O O co CD 5c 5c 55 o co CD C I 5 55 o o O o o o !1 ^ 2 CD o 5.8 «S ox C CO CD o o CD fc CO

jooo|oooo|oo|ooooo| nl 00000W U U W V 0 W ni fl. 0000000 zzzzz^z^

LnooLnLnooocnoomoooomooomLnooooooLnomoLnLOomoooin £C ^ CMCOCVICMCMCM COCNT-CNIC0^CNcr)mc>ICNIC0CNCMmc>JOJCOCNjT-CVJCMC0OJCOC\IC0^C\JC\lcr) -inoooooooui oincnoooooomocnoooouoooocn g.f- CM CO CM CO CMCMCMCOT-T- OJ ^ T- CM CO COCMCM CO IU

« CM ^_ CM £^ "? g CM 00 CO E •*J- 40 CO _!. CD n O E z E ° O II CD 8 o II CO CM • m- ~.~,"!l' c^1 _ CM T- in-* iv. •* co J^CO CM o O CM COro^^cNQlv-^CD^CO^^^QCMCqcD-ÔOCMCO • ^ •* <3> • iv. ^ _; 00 w > ,-,, AS^-w^S^SÊ^cbcrincoivri^â^îriLn^^âjoi^^tnT-^co

•^c2gOO^^mgOOOOOOOOOOOO-r--'--i--r--'--t-->-oOOOOOOOOOO ^SSSSSSSLtClLlirLllLtlLlirLriLrLlLtLtLtLlIhHhhhhhhHhh

169 WS112 CGC Name Exp Pen 1fetNGIyc location Description T06E6.2 cyb-3 5 15 No nucleus member of the cyclin B family that is required for embryonic viability T06G6.9 vbp-1 0 20 No cytoplasm Human VHL binding protein like T07D4.2 10 20 No cytoplasm calcineurin-like metallophosphoesterase T07F10.4 -5 25 No plasma membrane Predicted membrane protein T08B1.1 0 30 No plasma membrane sugar transporter T08D2.1 35 35 No Golgi uncharacterized T09A5.11 20 20 No ER N-oligosaccharyl transferase 48kd subunit T09D3.3 0 20 No extracellular uncharacterized T10B5.5 0 30 No cytoplasm uncharacterized T10B5.6 knl-3 10 25 No cytoplasm KNL-3 activity is essential for formation of a functional kinetochore T10B9.8 cyp-13A1 0 20 No ER Cytochrome P450 T11F8.3 rme-2 0 15 No extracellular LDL-like receptor T11G6.8 15 15 No nucleus uncharacterized T12A2.1 15 15 No cytoplasm uncharacterized T12B3.2 20 30 No plasma membrane uncharacterized T13A10.5 nlp-16 0 30 No ER predicted neuropeptide T13F3.1 str-183 10 20 No plasma membrane 7TM receptor T14B4.3 10 30 No uncharacterized T14G12.2 20 0 No plasma membrane synaptosomal associated protein 25A (SNAP-25A) T15B7.3 0 20 No extracellular cuticle collagen T16G1.9 10 20 No extracellular uncharacterized T17H7.4 gei-16 25 25 Yes nucleus protein with similarity to the B20 antigen of the parasitic nematode Onchocerca volvulus T19A5.3 0 20 No nucleus uncharacterized T19B10.2 0 25 No uncharacterized T20B12.3 10 30 No Predicted nucleolar protein involved in ribosome biogenesis T20B3.2 tni-3 0 35 No cytoplasm Troponin I T20D4.19 20 20 No uncharacterized T20D4.5 -5 15 Yes uncharacterized T21B10.2 20 20 Yes cytoplasm ortholog of human ENOLASE 1 T21C9.12 scpl-4 20 20 No nucleus REV protein (anti-repression transactivator protein) T21D12.12 0 20 No extracellular protease inhibitor T22D1.10 ruvb-2 20 -5 No nucleus uncharacterized T22F3.4 rpl-11.1 20 35 No cytoplasm large ribosomal subunit L11 protein T22F7.1 20 20 Yes plasma membrane protease inhibitor T23B12.7 dnj-22 15 15 No nucleus encodes a protein containing a DnaJ ('J') domain that is predicted to be mitochondrial T24D1.1 sqv-5 20 20 No chondroitin synthase that both initiates and elongates chondroitin chains T24D5.3 10 20 No uncharacterized T24F1.4 20 -10 No extracellular uncharacterized T26A5.8 20 20 Yes nucleus DNA polymerase epsilon, subunit D T26E3.7 10 25 No mitochondrion ATP synthase alpha and beta subunits CO c c ••- q o c DC ~ CD o ° E co 2 a. j2 0 0) ^ co •£ o "- •D o .E o E *- CO 1 X I •o Z y: o a ~° o S CO o > o o £ CO OCT CD ^ <» ">-S CO Q I CO CD = Q-'co •e co o S ^ IE I I! V W = 3 CO CD ; «g 9 ® o a E •gg CD > o i= o a3 c o CO CO c CO iCO t>< Q. .E O -5 CO 3 CD CD CD o3 S C 3 CD T3 = CO O %-CO I co S .c o £ CO CD „ • o CO c : O) D) 2 2 § CO 3 ex CO : c c T3 cy l T3 T3 T3 T3 •Q 0 T3 t> CD O. O jO CD CO qj < TS ran s r- CD CO CD 'c 'c o CD S> CO S 9 CD CD S> P CO D) Q. *~ x: N Ns> N CO -g N N CD -9 R co RJ •o *= C 2> "a.^ e C a .3 C C o S 3 •TT O) CO CB ^, CD CD CO CD « c £ c E c "5 •S3 •2 to o S •8 S -8 •8 5 CD _ o o ro o O o O o o O cj CJ ~ E ^ CJ "D O O) CO CM i? ca 50 CO CO CO CO o o O.T3 O fO CD CJ 2 CO o 2 55 S 5 E a c c 5 a C: C CD CO C 3 Si X3 1- CO CO CO £ cB CO CO CO CO CO CO "O CO O c c CO O CO co 0: .o -C -C CD CD <'•?- •c --s -c c •c •c O ^ CO o co CJ CJ o o CJ o O ^ D. "co U m "

nch nch E omo l elt a nch CO c CO c c c c c c c 2 2 DC I- p C £ C S O) MY S 3 •D me m 3 3 3 3 3 3 3 CD .c DC 3 p 3 Q. Q. 3 Q. 3 E(5» CO 3 CD CD E 3 C C CO CO c CD F i5 E w c E-§ E E E E co CO E E £ co -a CO CO CO CO ^ CO CO sf CO § « = S 3 3 O 3 CD 3 CD DC !9> co c 0 « oi * 0 CO CO CO ^ 0 0 CO CO CD CD E IS .0 •i •Q 9- O Q. "o Q. Q. Q. Q.73 O Q. Q. 2-5 3 O CO O 3 OOO O 3 3 O O go 0 3 5.-S.-S, 5.C C OX" o OOO O %O %O to B 0 _cc ^° Q.

0 00000 000000 00000000000000000000 ^zzzzz^z^zzzzzz^rzzzzzzzzzzzzzzzzzzzz

_oooooooooooococnoooOLooocoomooooooooooooOLooo g. CMmCM -i-CM-^ CM T--I-T- -^CM T-T-CO T-CO T-WT-COCVINx-T-lMWT-t- Ul

E to ^- O) CO T- ^ CO CO O 0 s| t il o o CO CO CM CO *tf •»*" CO CT>-, - CM LO r- r^ t- CM CMCoI]cncM^^CM^cDcn-r-OCQXXX< CO n LnoocococD|^.|^ooooco f>J •* CVJ Q X CD C

171 WS112 CGC Name Exp Pen NetNGIyc location Description Y24D9A.8 15 10 No cytoplasm ortholog of the human gene TRANSALDOLASE 1 Y37H9A.6 ndx-4 0 15 Yes cytoplasm homodimeric diadenosine tetraphosphate (Ap(4)A) hydrolase Y38H6C.3 10 30 No uncharacterized Y39A1A.14 20 0 No nucleus protein required for 18S rRNA maturation and 40S ribosome biogenesis Y39C12A.1 25 25 No cytoplasm SAM domain; Ankyrin repeat Y40B1A.1 0 20 No uncharacterized Y43H11AL3 pqn-85 20 0 Yes nucleus protein predicted to contain a glutamine/asparagine (Q/N)-rich ('prion') domain Y45F10C.4 10 20 Yes uncharacterized Y45F3A.3 20 20 No mitochondrion very-long-chain acyl-CoA dehydrogenase Y45G5AM.9 20 20 No nucleus uncharacterized Y46H3A.4 10 35 No uncharacterized Y47D7A.1 skr-7 25 35 Yes nucleus homolog of Skp1 in S. cerevisiae that is required for posterior body morphogenesis Y47G6A.24 mis-12 0 15 No protein required for proper attachment of chromosomes to the mitotic spindle Y47H9C.5 dnj-27 0 20 No ER uncharacterized Y48A6C.2 10 20 No uncharacterized Y48A6C.4 30 20 No ortholog of S. cerevisiae IPI1 that may suppress tumorous growth in the germline Y48B6A.1 30 10 No nucleus ortholog of human BOP1 (overexpressed in colon cancer); may suppress tumorous growth Y48C3A.14 10 20 No nucleus uncharacterized Y49F6B.1 0 30 No nucleus uncharacterized Y51A2D.7 25 -5 No uncharacterized Y51H1A.3 0 30 No mitochondrion uncharacterized Y51H4A.22 10 20 No extracellular uncharacterized Y51H4A.3 rho-1 20 0 No plasma membrane GTP-binding protein that is a member of the Rho family of GTPases Y53C12B.2 30 10 No cytoplasm predicted RNA-binding protein Pnolp involved in 26S proteasome assembly Y53F4B.14 20 20 No uncharacterized Y53F4B.3 0 20 No nucleus transcription factor CBF/NF Y53G8AR.3 ral-1 5 15 No plasma membrane Ras small GTPase Y54E10BR.5 35 35 No ER signal peptidase I Y54G2A.18 35 35 No Golgi B-cell receptor-associated protein 31 -like Y54H5A.2 10 20 Yes cytoplasm uncharacterized Y54H5A.3 35 0 Yes cytoplasm RNA-binding protein Y55F3C.6 0 30 No uncharacterized Y56A3A.1 ntl-3 25 25 Yes nucleus ortholog of NOT3/NOT5, a member of a protein complex also predicted to contain ccf-1 Y56A3A.4 taf-12 20 20 No nucleus homolog of transcription initiation factor TFIID Y57A10A.19 25 25 Yes nucleus splicing coactivator SRm160/300, subunit SRm300 Y57G11C.23 10 30 Yes plasma membrane synaptic vesicle transporter Y58A7A.1 20 20 No plasma membrane Copper transporter Y62E10A.13 0 20 No cytoplasm uncharacterized Y62F5A.1 mdt-8 25 15 Yes nucleus Uncharacterized conserved protein Y64G10A.5 25 35 No uncharacterized E c E (D to c E co l Q. S (0 O 10 a) 5 > a. 15 Q) a> to (0 O .Q O (D O X: a; CO o to 0) to c 0> . tl> .g E c JO tD o o > o c to D) •a o o tD to o E C 0! o O) X: c o 6 E co !c o EC CD en to CD 55 •D to CO o <= = I CD X: ai

to l< S-S T3 = B a> o 9> a> o ® § c ,£ a "- I tO (O 2 to Z §: = o a. e Q- W tb ^ to . g.2 C tD toI 2 c — >> O ~2 o 2 -o .2 £ CD ,_ S CD to to SL o 9) CD .Nj -g O .£J £ ^ 9> o .N ,N IIIS! - g I» c c o OJ'C C C Q. Co tb co XI tD a> 2 <° s:C D to c m o co '= to c: CD co S < 3> " 'S t» ^ o .c co CO CO to jo & p — to u £ £ « I X> CO to to £ CO Jill -§! p-s "r? § 8 P ^ S E 11_ 1c c 2 <= c | to c c So? 3 D)Z> 3 O.I3 3 6 § 1$Q . 'to §§!§ I! fc > 3 a ~~ O.Z

XI JO CO E E CD E E E E to to 3 CO CO 3 D) m CO las m las m las m eu s eu s eu s eu s g CD a) JO i? DC el l o to *9 o E g-LU O o 10 Q. Q. Q. to o to &&&E 1 C3 1 S o o 1 O nuc l yto p %.%.S,c c CO -tr c o E O %.o O tj O i £ %O %.C_> to * CO a.

OoooooooooooooooooSSooooofSoooooooo

— OOooooomoooOLooocnoooooocooomoooomo 2B CNJC0CNIC^JC\immc>JC\lC0-^CNC>lt>10JC\IC>JC\IC\lC0t>JC\l COOJCOCMCMCOtM OLXJOOOOOLOOOOOLOOOOOOOOOOOOtnLOOOOOLDO ft OJCOCCM CO CMM 1-tMtMCO CM I-T-I- »- T- CM CM CXI T-I-C\JCOCDCMC0I-CMCMT- IU

I 1- m ^ 1- I-. CM CO IT) n ji; ^L xi co co tb C XI z X! o g>§ > c o o o

CM O CM <;cocqc\i,_cM''-c\ico : 0 -* N. •r- CQCQO<<<'-. T '^'R ? ^oco ^f i- •* Is- 00 £r>CO •<j* CujT-cDWcococoCM'taiLncoD s CO i- m tM O Lf) 00 gCDCQCOQLLXOLuLLXpcogj-^f— --„ ^^ ^_--_. -- . T— O O T— CO CO O)oo'ciocMCMcriaicboia i LOUOCQi— t— -i— COCOCOCDCMOJCO^^LDr- i-i— T-T— COT-COCOT-COCDOLOC350 > tptppf-NNSf-sspppopp^bcyyy^ •q-Tf^-LOCDCDCOOOCOCO >->->->->->->>->->-NNNNNNNNNNNNNNNNNNNNN^ *: ^ ¥ ¥ •* ¥ >* ><>£N

173 APPENDIX B

NATIVE AND PERMETHYLATED MALDI-TOF SPECTRA

The spectra in this section represent MALDI-TOF profiles of both native and permethylated A/-linked glycans from N2 Bristol (wild-type) and NF299 (cogc-

1)k179 C. elegans strains. All glycans were released with hydrazine. The comparative analysis consisted of three individual trials from synchronized L1 larvae from each strain to the final permethylation step of reduced oligosaccharides. The native profiles show the presence of endogenous O- methylated A/-glycans that are not detectable after the permethylation step.

All figures show the N2 strain profiles above the NF299 samples. The mass range (m/z) for native samples is 900-2200 and 1100-2800 for permethylated samples.

174 100 90; N2 Bristol 80;

70:

50-

40:

187P.B0 a *«>IJVWOW»'I%I»II*'MH wW'N'wiWii mniSi ii 1500 1600 2100 2200 l^ss/Oiarge

93B.52

NF299 (cogc-1)k179

50-

40;

30;

20;

10; 2133.01 0: ••*|W^W|I^ •WW^TW'lWMl »t*V H>MTmSil(iihWi» 1500 1600 2000 2100 2200 f\fess/Charge Figure B1. MALDI-TOF profile of native AMinked glycans released by hydrazinolysis (2 of 3). 100

90; N2 Bristol

80;

70:

60; 1081.63 60: ion?, eb

40;

30;

2<>

10;

0: 1500 1600 2100 22001 [<=]•( Kfess/Charge

100

90; NF299 (cogc-1)k179

1097.71 ZO;

60-

50:

30;

20;

15fc.-M 1663.86 1714.53 10: ^fiifkm^ 0 1500 1600 2100 22002[c].i (\fess/Charge Figure B2. MALDI-TOF profile of native A/-linked glycans released by hydrazinolysis (3 of 3). 1321.2 100

90; N2 Bristol

80'

70:

60:

50;

40- 2498.2 30; 24! 8.9 20- 24! 9.7

10- 25 0.5 2702.4 0 VkrtfMrji Uft*"yvii,'ii»Hi*nw(U 1100 1200 1900 2000 2700 28O01 W-l

%lnl. 1595.7 100'

90: NF299 {cogc-1)k179

80- 1361.6

70;

60:

50;

40:

30;

20; 2413.3 10: 2323.7 24U.8

0 i 2 10 1100 1200 1300 1400 600 1700 1800 1900 2000 . . iMniimifiiWrttflivifliiymn'TYI r:V. i 270i ^0 28001M-1 NtassVCharge Figure B3. MALDI-TOF profile of permethylated A/-linked glycans released by hydrazinolysis (1 of 3). N2 Bristol

80-

7<> 11*6.4

2682.9

1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700 28001M-1 IVbss/Charge

-J 00 NF299 (cogc-1)k179

2617.7 10:

0: nifbin*iib«iiMy--*• — "*---•"' -»•>--.. 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700 280Cl[c].<| Wbss/Charge Figure B4. MALDI-TOF profile of permethylated A/-linked glycans released by hydrazinolysis (2 of 3). The NF299 sample contains a non-carbohydrate contaminant at m/z 2413.1. %lnt. 1106.0 100' N2 Bristol so- 1497.6

80- 1702.5 70: 1498.6 2681.5 60; 1147.0 26f 1.0 5fr

40; 26*3.0

30; 2630.0 11*8.0

20; 2684.6

10;

0 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700 28001WI Wbss/Charge

%lnt. 1105.9 100' NF299(cogc-1)k179 90-

80:

: 70 1186.9

1146.9 132o.l 40; 1498.6 1187.9 13Q^.8

30; 1188.9 12^

20: 2682.3 268P.8 10; 26 &5.5 o: 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700 28001M-1 IVkss/Charge Figure B5. MALDI-TOF profile of permethylated A/-linked glycans released by hydrazinolysis (3 of 3). The both samples contain a non-carbohydrate contaminant at m/z 2682.3. APPENDIX C

RELATIVE ABUNDANCE OF NATIVE A/-GLYCANS

Comparative analysis of native A/-glycans form N2 and NF299. The analysis is based on the population mean of three individual samples for each strain run simultaneously. Compositions present Table C1 include pauci-40 nnose, high-mannose, fucose rich, truncated complex and endogenous O- methylated /V-glycans. N=HexNAc (e.g. N-acetylglucosamine, N- acetylgalactosamine), H=Hexose (e.g. galactose, mannose, glucose),

F=Deoxyhexose (e.g. fucose), OMe=0-linked CH3.

180 N2 NF299 Molecular % /V-glycan % W-glycan [m + Na]+ Composition Composition Composition

N2H3 935.7 12.87±4.86 12.53±3.54

N2H2F2 1065.8 1.06±0.41 1.38±0.27

N2H2F2 -OMe 1079.8 1.22±0.23 1.27±0.30

N2H3F 1081.8 7.19±1.88 8.46±1.92

N2H4 1097.8 5.03+1.06 8.61 ±0.84 N3H3 1138.8 2.19±0.24 2.80±0.37

N2H3F2 1227.8 1.52±0.21 3.24±0.51

N2H3F2 -OMe 1241.8 1.45±0.29 1.06±0.31

N2H4F 1243.8 4.81 ±1.56 5.92±1.95

N2H4F -Ome 1257.8 3.31 ±0.81 1.02±0.34

N2H5 1259.8 9.33±0.83 11.37±0.89

N2H3F3 1373.8 0.72±0.58 1.26±0.33

N2H3F3 -OMe 1387.8 1.04±0.49 0.74±0.31

N2H4F2 1389.8 2.53±0.17 4.00±0.78

N2H3F3 -Ome2 1401.7 0.91 ±0.43 0.59±0.35

N2H4F2 -OMe 1403.8 3.58±0.91 0.98±0.31

N2H5F 1405.8 3.37±1.35 2.72±1.03

N2H4F2 -Ome2 1417.8 1.49±0.52 0.74±0.30 NgHsFi -OMe 1419.8 1.71 ±0.21 0.69±0.28

N2H6 1421.8 2.80±0.55 2.81 ±0.66

N2H3F4 -OMe 1533.7 0.67±0.25 0.60±0.27

N2H4F3 1535.7 1.33±0.33 1.71 ±0.29

N2H4F3 -Ome 1548.7 0.76±0.47 0.56±0.32

N2H5F2 1551.8 2.50±0.29 2.88±0.69

N2H4F3 -Ome2 1563.5 1.15+0.51 0.66±0.31

N2H5F2 -OMe 1565.7 2.55±0.70 0.84±0.37

N2H7 1583.7 1.55±0.47 2.26±0.53

N2H4F4 1681.7 0.75±0.30 0.62±0.38

N2H4F4 -Ome 1695.7 0.87±0.23 0.57±0.27

N2H5F3 1697.7 1.48±0.38 0.92±0.40

N2H4F4 -Ome2 1709.7 0.82±0.42 0.64±0.37

N2H5F3 -Ome 1711.7 2.12±0.58 0.66±0.37

N2H6F2 1713.7 2.06±0.37 1.20±0.08

181 N2H8 1745.7 1.54±0.46 2.37+0.54

N2H5F4 1843.7 0.81 ±0.36 0.61+0.23 N2H5F4 -Ome 1857.7 1.10±0.28 0.56+0.34

N2H6F3 1859.7 1.30±0.14 0.98+0.28

N2H6F3 -Ome 1873.7 1.40±0.37 0.62±0.28

N2H9 1907.6 1.80±0.62 4.08+1.37

N2H6F4 2005.7 0.74±0.26 0.62±0.28

N2H6F4 -Ome 2019.6 0.84±0.12 0.50±0.22

N2H7F3 2021.6 1.13+0.11 0.57+0.28

N2H7F3 -Ome 2035.4 0.70±0.23 0.57±0.30

N2H10 2069.5 0.74±0.46 0.95±0.17

N2H7F4 2169.0 0.60±0.30 0.61+0.30

N2H7F4 -Ome 2183.0 0.56±0.49 0.64±0.44

182 APPENDIX D

ADDITIONAL MSn SPECTRA

n 3 MS profiles from Man5Fi-4. The majority of spectra are the MS of Y-ion complements not included in Chapter 6. These ions are reducing end GlcNAc residues with fucose and hexose (galactose) additions.

183 Celegans(N2)MS6_1595_1302_1084_667_563_#2.1 #1-300 RT: 0.00-13.15 AV: 300 NL3.01E-2 T: nti/IS + p NSI Full ms6 [email protected] [email protected] [email protected] [email protected] [email protected] [155.00-600.00] 345.09 100n

327.09

301.00

285.09

259.00

433.09 545.18 447.09 324.18 532.09 279.27 358.73 476 64 489-18 519'?7 227.09 240.82 386.18 405.18 429.27 • j I 499.00 | , \ \ , , ill i . ,1 _L I Ji i i'"| i i Y"i y r i n r'i I"I"TI [,nv i i "i ii i i1 r i n T i "i i • i vr i l!i L r-i i r i i U-r - i r i y i 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 m/z 6 Figure D1. N2 MS spectra of m/z 563 from Man5 (m/z 1595-* 1302-^1084-^667^563). Several peaks confirm that the terminal and penultimate hexose residues are connected by a 1,3 or 1,6 and 1,3 linkage respectively C eleqans (N2> MS4_896 1302 880#1-102 RT:0.00-1.61 AV:102 NL: 4.36 T; ITMS -t-p NSIFull ms4 896.00®cid35.00 [email protected] [email protected] [240.00-900.00] „-- 635.27

838.45

561.27 579.18 60S.36 848.45 ZIZCX7 •-?? Tci.^,18!,, I ^J^63^il,i,li

C eleqans (cog) MS4 896 1302_8S0 #1-62 RT: 0.00-1 .91 AV: 62 NL: 5.17 T: ITMS +• p NSI Full ms4 896.0Oiacid35.OO 1 302.45<8>cid35.00 880.270cid35.00 [240.00-900.00] 635.27

fl 561 O 458.09 761 .00 788.27 I 863 4L ),U,,* I <.,.,,1,1 i] °^ 500 ^Ji 5S1,,.,,0 ^ 600.M^dl-MlLL 650 , 700 7BO 800 ,ly~* ?4 „___*._ ~t ^r,/-, OOA (r^r~ r^l«MA« n/i«« c •« /«,/, one2++ Figure D2. MS spectra of m/z880 from GlcNAc2Man5Fuc (m/z 896* -» 1302^880). N2 and NF299 spectra are identical. C eleqans (N2JMS3 896 490 <51h) #1-50 RT: 0.00-0.81 AV: 50 NL: 5.68 T: fTMS + p NSIFullms3 896.64®Cld35.00 490.27 @cid35.00 [130.00-500.00] 100- 472.27

95f

90= B&i

80z

75z

70z 65= 302

R 60- 284.09 I 55: 3 50- .A- \ 284 S : « 45: MeO 0Me \/ 1 «>[ 458.27 229Y 35: 211 -H20 472 30: H 25: °-<^-OMe

20: Na+ -MeOH 430,27 •458 15: 10: 21 .00 22.4.09 302.18 r° -(H20 + Ac)^ 34 B 5: 229.09 V 385.18 44827 I 315.27 371 27 39 6 18 430 | | 234.00 252.00 270.09 28.7.27 I ^ 340.09 359.09 i I l - 415.18 , 445,18 | 1 1 487,36 i' | '\ i 'I |" i if' "l I ' I I ' i i* I l | l l i I.,V i 'i

C elegans (COg)MS3 896 490 (5th)#2#1-113 RT: 0.00-4.65 AV: 113 NL: 5.78E-2 T: ITMS + [email protected] [email protected] [130.00-500.00]

Figure D3. MS3 spectra of m/z 490 from 2+ GlcNAc2Man5Fuc (896 ^490). This ion corresponds to the fucosylated reducing end GlcNAc. The presence of peaks are largely 9 1 45- identical, but the abundance of ions vary, possibly due to the poor signal in the NF299 strain. The top spetrum is N2 and the bottom is NF299.

398.27 »48.3( 408.8 ar I 287.18 fl j 25252,22 7 275.27 t.l.li.l 1 f'-tl'f iftl.^Jiifil.JLi4.,ii|k|j|,liiii. iljilh lliiy 300 320 380 400 I C elegans(N2)MS3 983 490 #1-50 RT: 0.00-0.72 AV: 50 NL: 9.36 T: fTMS + p NSI Fullms3 [email protected] 490.27 @cid35.00 [130.OO-5O0.OO]

302

•H 55r

302.18 448.27 229.O0 426,18 43.3-27 J[ 430

38! S f"l I*' i I 315.18 34343.00 9 360 >' 288.27^ | ,|. | 328.18 ^ m/^ z359.0 8 37.1.09 |, ^ Celeqansfcog) 983 490 #1-51 RT:0.0O-1.57 AV: 51 NL: 8.12E-1 T: (TMS + p NSI Full ms3 [email protected] [email protected] I130.0O-500.00] 3 Figure D4. MS3 spectra of m/z 490 2+ from GlcNAc2Man5Fuc2 (983 ^490). This ion corresponds to the fucosylated reducing end GlcNAc. The top spetrum is N2 and the bottom is NF299. '8 45-

35-E

30^ 2&i

20E

302-1 395.18 M~* I 10^ 323.09 343.09 355.00 397.09 295.09 L r • .fc I) i. • i ••ill - .fl- iflJli ,i.J .HI . 340 360 nVz C elegans(N2)MS3 983 694#1-51 RT: 0.00-2.12 AV: 51 NL: 3.26E1 T: fTMS + p NSI Full ms3 [email protected] [email protected] 1190.00-700 00]

476 302

676

19.27 1 662 458.27 589.27 I 444.18 I 3291 371.18 387.18 0+Ac • „L„. I... ^ J634 a i I I 652.36 ,^7„te I C T3fl ,....!.,

Figure D5. MS3 spectra of m/z 694 2+ from GlcNAc2Man5Fuc2 (983 ^694). This ion corresponds to the Hex-Fuc reducing end GlcNAc. The top spetrum is N2 and the bottom is NF299.

45-

4D"5

35T

15.09 I , 332.91 357.09 393.18 505.1B i 533.GO I. K *I^M. II^H.^! jlrt*ll|,t,^l|J,., Jy,..IU. > >(N2)MS3_983_ 664 #1-53 RT: O.O0-O.84 AV: 53 NL: 3.25 T: ITMS -i p NSI Full ms3 [email protected] [email protected] [180.00-7O0.O0J 458.18 100n

60-3 55-3

(-HjO+Ac) 385

(-H20) 646

-H20+Ac) 398 (-MeOH) 426 315.09 631.18 62Q.91 [ I319.09 353-0° 368.09 488.91 II 535.18 j 663.27 U.,1 I»l iXJu ._!_ ft | l] it*..*. •) i ,<„„l

C elegans(cog) 983_ 664 #1-51 RT: 0.00-1.6S AV: 51 NL: 1.30 T: (TMS + p NSI Full ms3 9a3.00©Cid35.00 664.00©cid35.00 J180.00-700.00] 45B.09

Figure D6. MS3 spectra of m/z 664 from 2+ GlcNAc2Man5Fuc2 (983 ->664). This ion corresponds to the di-fucosylated reducing end GlcNAc. The top spetrum is N2 and the bottom is NF299. 650

45d 850 40=j •"836 •-808

385-18 808.45 i \ 547.27 371,09 I 1 415.18 721 36 767 g, ?ga27 III] r-r-W- -k4- •1 f "t-'i r •!• 1 '1' i**! *r-P'

85-3 3 80-1 Figure D7. MS spectra of m/z 868 z+ from GlcNAc2Man5Fuc2 (983 ->868). This ion corresponds to the Fuc2Hex reducing end GlcNAc. The top spetrum is N2 and the bottom is NF299. I

385.09

398.09 808.,27 j 1 I 415.18 -4-^4- 400 U^-fa Celegans(N2)MS3_983_1072 #1-50 RT: 0.00-1.60 AV: 50 NL: 4.87 T: ITMS + p NSI Full ms3 [email protected] 1072.55®cid35.00 [295.00-110000]

1 463 1040

1012

385.09 I 444 o 557.1 f.,1.. II A

C efegans(cog)MS3 i RT: 0.00-2.54 AV: 100 NL: 1.74 T: rTMS+pWSIwFiilnr D 1072.64 @cid35.001295.00-1100.00] 662.36

90-H

85-3

3 70-: Figure D8. MS spectra of m/z 1072 2+ from GlcNAc2Man5Fuc2 (983 -+1072). 60^ This ion corresponds to the Fuc2Hex2 I 5(>3 reducing end GlcNAc. The top spetrum is * -. IS 45- N2 and the bottom is NF299.

35-:

20^

15-E 10r 854.45 1040.45 385.18 444.27 925.45 86_6.36 | l L 1073.55 .1. , • ^ i..M.» ——1-«~. Celeoans fN2j MS3 1070 868 #1-50 RT: 0.00-0.66 AV: 50 Nl_: 3.10E1 T: ITMS + pESIw Full ms3 [email protected] [email protected] [235.00-900.00]

650 60=

50£ Meo

—£ ••850 Me0H .336 [-(H.0 + AC)_ •808

458.27 630.27 476.18 I 650.27 71.18 I I 503.27 867.64 897.27 AA r1. r?. A •P-?^

C eteaans(cog)MS3 1070_868. (5th) #1-113 RT: 0.00-2.00 AV: 83 NL:1.94 T: rTMS + pNSIFUIms3 [email protected] 868.55<&cid35 00(235.00-900.00]

Figure D9. MS3 spectra of m/z 868 from 2+ GlcNAc2Man5Fuc3 (1070 ^868). This ion corresponds to the Fuc2Hex reducing end GlcNAc. The top spetrum is N2 and the bottom is NF299.

20^

153 651..27 8.55 476.27 415.18 21 27 J 721.45 757.91 795,45 -A •M •. I, ,. I rrfif 100z

90f

356

3(>|

(-H20+Ac) 398 (-MeOH) 426

341.00 359,18 37291" hX\ M'p.l.i.W,.! f*kr,rW?h-fa-*>lfo,- A«U^

s(cog) 1070.664S1-51 RT: 0.00-1.71 AV: 51 NL758E-1 T: rTMS + pNSIFulims3 1070 [email protected] [email protected][180.00-700.00] 458.18

Figure D10. MS3 spectra of m/z 664 from 2+ GlcNAc2Man5Fuc3 (1070 -+664). This ion corresponds to the di-fucosylated reducing end GlcNAc. The presence of peaks are largely identical, but the abundance of ions vary, possibly due to the poor signal in the NF299 strain. The top spetrum is N2 and the bottom is NF299.

697.45

700 Celeqans(N2)MS3_1070 694#1-50 RT; 0.00-0.64 AV: 50 NL: 4.34 T: ITMS + p ESIw Full ms3 [email protected] [email protected] [190.00-700.00]

m 60- | 55: I 503 S -, 1 45-

415.18

85.09 631.45 329.09 355.09 bOb.18 U 532-! 619 36 '398.36 | I - ll 676.64 694.18 »,>U 1| I,,,, JL,,M^,i JkJt kJ Jn...jMi.4 ..mH.i,iiL,.ni.,|U,1,14,1 1 flu, ,l„i, A, Celegans (cog) 1070_694 #1-201 RT: 0.00-5.99 AV: 201 NL: 1.04E-1 •4^ T: FTMS + p NS! Full ms3 [email protected] [email protected] [190 00-700.00}

Figure D11. MS3 spectra of m/z 694 from 2+ GlcNAc2Man5Fuc3 (1070 ^694). This ion corresponds to the Hex-Fuc reducing end GlcNAc. The presence of peaks are largely identical, but the abundance of ions vary, possibly due to the poor signal in the NF299 strain.The top spetrum is N2 and the bottom is NF299. Celegans

95f

90-

85^

80^ 75-E 70-=

65T 8 60- 302 •8 56^ 4 50^ « 45- 458.18 » 405

35r 284.09

30^ 25-^ *472 430.18 20-E 15^ *458 10- | 302.18 385.09 448.00 22 27 252.00 I l 315.27 210,73 | 34 27 398 9 414.91 5^ s8 91 270.O9 f 286.82 I 1 328.91 t 359.,8 376.27 1 ,° 439.91 1 474.09 JI \i ii , . Ail., .Lli .. II I .11 .ii . I- ,k - J. 1 l.l,i,.|fll., L^Jl 1.1 i . .Ift. . , J- ' I" 0 ^ r 1—1 "i "i—i— '1 1 ""i—1 iifii r if T" V t | 1 ... (a, f I P 430 340 360 m/z C elepans (cog) 1070 490*1-201 RT: 0.0O-7.15 AV: 201 NL:1.17E-1 T: fTMS + p NS1 Fullms3 107O.OO@cid35 OO [email protected] [130.00-500.00]

Figure D12. MS3 spectra of m/z 490 2+ from GlcNAc2Man5Fuc3 (1070 ^490). This ion corresponds to the fucosylated reducing end GlcNAc. The top spetrum is N2 and the bottom is NF299.

45- 40" 35f 30^ 25^

393.09 315.00 33491 401.00 415.00 302.00 j J . 352.91 373.09 I j"

200