UNIVERSITY of CALIFORNIA Los Angeles From

Home , Phosphotransferase

UNIVERSITY OF CALIFORNIA

Los Angeles

From Vitamin C to Metabolite Repair:

The Role of Novel Sugar Nucleotide Phosphorylases

A dissertation submitted in partial satisfaction of the

requirements for the degree Doctor of Philosophy

in Biochemistry and Molecular Biology

Lital Noah Adler

2012

Lital Noah Adler

2012

2 ABSTRACT OF THE DISSERTATION

From Vitamin C to Metabolite Repair:

The Role of Novel Sugar Nucleotide Phosphorylases

Lital Noah Adler

Doctor of Philosophy in Biochemistry and Molecular Biology

University of California, Los Angeles, 2012

Professor Steven G. Clarke, Chair

The pcm-1 gene in the soil nematode Caenorhabditis elegans that encodes the L- isoaspartyl methyltransferase has roles in protein repair as well as in the worm’s insulin-like signaling pathway. Interestingly, the pcm-1 gene displays overlapping exon sequences with a previously undescribed gene (mcp-1, C10F3.4) that is transcribed from the opposite strand. This rare topology in the C. elegans genome (less than 0.05% of all genes) led us to ask whether the expression of the pcm-1 and mcp-1 genes may be co-regulated or whether their products may have a related function. PSI-BLAST analyses revealed that MCP-1 has homologs in plants, vertebrates and invertebrates and that all the homologs share a conserved HIT (histidine triad) motif. Mutations in the Arabidopsis thaliana homolog, vtc2, lead to vitamin C deficiency in these plants.

ii In this thesis, I have characterized the enzymatic activity of VTC2 from A. thaliana as a

GDP-L-galactose phosphorylase, the last missing step in the Smirnoff-Wheeler pathway for vitamin C biosynthesis in plants. Additional work has been done to show that the green alga

Chlamydomonas reinhardtii shares homologs with all of the genes in the Smirnoff-Wheeler pathway as well as the specific activity of GDP-L-galactose phosphorylase.

While vitamin C is an apparent important antioxidant and enzymatic cofactor in C. elegans, no reports had been made as to the ability of this nematode to produce it. In my work I found out that C. elegans extracts contain low levels of vitamin C. Nevertheless, as levels of vitamin C in the mcp-1 deletion strain are not different from the values in the control animals and as C. elegans lack key homologs in the biosynthesis pathway, I concluded that mcp-1 was not involved in vitamin C synthesis in C. elegans.

Enzymatic analysis revealed that C. elegans’ mcp-1 and its mammalian homologs encode

GDP-D-glucose phosphorylases. mcp-1 mutant worms that lacked this activity accumulated

GDP-D-glucose. This unique metabolite, which shares high similarity with GDP-D-mannose, does not have a reported role in invertebrates or mammals. With my collaborators, I suggested that GDP-D-glucose could compete with GDP-D-mannose for binding to glycosyltransferases in the first steps of protein N-glycosylation in the cell, thus potentially replacing mannose residues with glucose residues in a step that could lead to N-glycosylation defect. Thus, MCP-1 may serve as a metabolite repair enzyme that eliminates adventitiously formed GDP-D-glucose from the cytosol to ensure correct protein N-glycosylation. I have tested this hypothesis by observing morphological changes in mcp-1 mutant worms and by directly comparing the N-glycan profiles of mcp-1 and control animals.

iii The dissertation of Lital Noah Adler is approved.

______Catherine F. Clarke

______Stephen G. Young

______Wayne W. Grody

______Robert T. Clubb

______Steven G. Clarke, Committee Chair

University of California, Los Angeles 2012

iv I would like to dedicate this work to my family.

To my best friend and companion, Tahg K Adler, thank you for being such a supportive partner

in this journey. Thank you for pushing me and encouraging me all these years to pursue my

ambition. I love you and could not have done this without you.

To my amazing boys Shai, Ilai and Itai, I am so blessed to have you in my life.

To my beloved grandmother, our family matriarch, Klara Grinfeld, whose inspiring life story

and love made me believe that any obstacle can be overcome.

To my parents, Tzvi and Nusha Levidor, for their endless love and support.

To my sisters and their families Rinat, Orit, Aviad, Lavie, Avinoam and Elkana.

To all of my aunts and uncles and their families in Israel –

The Kuttners; Rimona, Avigdor, Yifat, Elad, Chaya, Namma and Neta

The Waksmans; Haim, Ruti, Yaron, Tamar, Vered and Nurit.

To my father in-law Dr. Norman Adler, probably the only one in my family who actually

understands this work.

To my mother in-law Shelley Adler for always believing in me.

I love you all and I am looking forward to starting a new chapter in my life together with all my

family.

v TABLE OF CONTENTS

Abstract of the Dissertation...... ii

List of Figures...... viii

List of Tables...... xi

Acknowledgements...... xiii

Vita...... xvi

CHAPTER 1 Plan of the Dissertation...... 1

References...... 9

CHAPTER 2 Arabidopsis VTC2 Encodes a GDP-L-Galactose Phosphorylase, the Last

Unknown Enzyme in the Smirnoff-Wheeler Pathway to Ascorbic Acid in

Plants…...... 13

References...... 20

CHAPTER 3 A Second GDP-L-Galactose Phosphorylase in Arabidopsis en Route to

Vitamin C. Covalent Intermediate and Substrate Requirements for the

Conserved Reaction...... 21

References...... 31

CHAPTER 4 Impact of Oxidative Stress on Ascorbate Biosynthesis in Chlamydomonas

via Regulation of the VTC2 Gene Encoding a GDP-L-Galactose

Phosphorylase...... 32

References...... 42

CHAPTER 5 Ascorbate is Present in Caenorhabditis elegans but its Origin is Unclear...... 53

References...... 79

vi CHAPTER 6 A Novel GDP-D-Glucose Phosphorylase Involved in Quality Control of the

Nucleoside Diphosphate Sugar Pool in Caenorhabditis elegans and

Mammals...... 84

References...... 96

CHAPTER 7 The Search for Phenotypes in Caenorhabditis elegans Mutants that Lack the

GDP-D-Glucose Phosphorylase Activity Encoded by the mcp-1 Gene...... 105

References...... 123

CHAPTER 8 Future Goals and Perspectives...... 127

References...... 137

APPENDIX I Glycan Analysis of Wild Type N2 and mcp-1 Mutant C. elegans...... 141

References...... 155

vii LIST OF FIGURES

2-1 Amino acid alignment of VTC2 homologs...... 15

2-2 Dependence of VTC2 activity on GME and Pi...... 16

2-3 VTC2-catalyzed phosphorolysis of GDP-L-Gal and formation of GDP...... 17

2-4 Smirnoff-Wheeler pathway from D-glucose to L-ascorbic acid...... 19

3-1 SDS-PAGE analysis of recombinant VTC5 (VTC2 homolog) and wild-type and

mutant VTC2 proteins ...... 24

3-2 Identification of a guanylylated enzyme intermediate formed during the catalytic

cycle of VTC2 that is dependent upon His-238……………....………………….……...25

3-3 Recombinant VTC2 is highly specific for Pi as the guanylyl acceptor used during

conversion of GDP-L-Gal to L-Gal-1-P……………...... ………………….…...... 27

3-4 Effect of Pi, D-Glc-1-P, and D-Man-1-P concentrations on the activity of

recombinant VTC2……………...... …………………….…...28

3-5 Effect of Pi, D-Glc-1-P, and D-Man-1-P concentrations on the activity of

recombinant VTC5………………………………….………....…………………….…..28

3-6 Effect of Pi, D-Glc-1-P, and D-Man-1-P concentrations on the GDP-L-Gal

phosphorylase or GDP-L-Gal-hexose-1-phosphate transferase activities in

partially purified extracts of A. thaliana (panel A) and Japanese mustard spinach

leaves (panel B) ………………………………….………....………………………...….29

4-1 Phylogenetic tree of VTC2-like proteins………………………………………………...36

4-2 The L-galactose pathway of ascorbic acid biosynthesis is conserved in green algae

4-3 Acceptor specificity of C. reinhardtii recombinant VTC2………………………..……...38

viii 4-4 VTC2 transcript levels are increased in response to oxidative stress..………...…..….....39

4-5 Ascorbate levels are elevated in response to oxidative stress..………...………..………39

4-6 Oxidative stress conditions result in up-regulation of GSH-ascorbate cycle enzymes.....40

4-7 Putative regulatory sites for ascorbate biosynthesis and ascorbate recycling in C.

reinhardtii..………………………………………………………………………...…….42

4-S1 Proposed biosynthetic pathways for ascorbate biosynthesis in higher

plants and algae…………………………………………………………..…………...….45

4-S2 Sequence alignment of VTC2-like sequences……………..………………………..…...46

4-S3 Tert-butyl peroxide is more stable in Chlamydomonas cultures than

hydrogen peroxide……………………………………………………………..………...48

5-1 HPLC measurements of AsA in wild type (N2) and the C10F3.4 mutant

strain (tm2679) ……………………………………………….…………………..……...67

5-2 LC-MS analysis of AsA in wild type (N2) and the C10F3.4 mutant strain (tm2679)…...69

5-3 C. elegans culture medium, food source, and extraction reagents do not

contain detectable AsA…………………………………….………………………….....71

5-4 HPLC measurements of AsA in LB and the E. coli OP50 extract…...………………….73

6-1 GDP-D-Glc phosphorylase activity in C. elegans C10F3.4 deletion strain

and in HEK293T cells treated with C15orf58 siRNA…………………………………...90

6-2 C10F3.4 expression pattern in C. elegans…………………………………………….....91

6-3 Tissue distribution of GDP-D-Glc phosphorylase activity in mouse…………...…….....92

6-4 Accumulation of GDP-D-Glc in C10F3.4 mutant C. elegans strain (tm2679)………….93

6-5 Possible role of GDP-D-Glc phosphorylase in preventing misincorporation

of glucose in place of mannose residues into glycoconjugates…………………………..95

ix 6S-1 Recombinant C. elegans C10F3.4 and human C15orf58 are highly specific for

inorganic phosphate as the guanylyl acceptor………………………………………….100

6S-2 mRNA tissue distribution of GDP-D-Glc phosphorylase (C15orf58 homolog)

and GDP-D-Man pyrophosphorylase α (Gmppa) in mouse…………………………...101

6S-3 Standard curves used for determining GDP-D-Glc and GDP-D-Man

concentrations in worm and mouse tissue extracts by IP-LC-ESI-MS analysis….....….102

7-1 Diagram of gonad morphology in wild-type hermaphrodites…....………………….….121

7-2 Morphological changes in the C. elegans mcp-1 mutant strain…...... ….122

AI-1 Flow chart for N- and O-glycan analysis of C. elegans pellets prepared at UCLA for

processing in the Stuart Haslam laboratory…...... …..149

AI-2 MS spectrum of PNGase-F released C. elegans N-glycans from the

35% acetonitrile fraction…....…………………………………………………….…….150

AI-3 MS spectrum of PNGase-F released C. elegans N-glycans,

50% acetonitrile fraction…………………………..…………………………………....151

AI-4 MS spectrum of PNGase-A released C. elegans N-glycans,

35% acetonitrile fraction………………………………………………………….….…152

AI-5 MS spectrum of C. elegans O-glycans released by reductive elimination,

35% acetonitrile fraction………………………………………………………………..153

AI-6 Symbolic nomenclature as outlined by the Consortium for Functional Glycomics

Nomenclature Committee (May 2004) Varki et al., 2009……………………………...154

x LIST OF TABLES

2-1 Substrate specificity of the A. thaliana VTC2 phosphorylase……………….…………18

3-1 Phosphorylase activities of wild-type and mutant VTC2 enzymes…………….……….24

3-2 Comparison of the substrate specificities of A. thaliana VTC2 and VTC5…………….24

3-3 Phosphorylase and transferase activities in various partially purified plant extracts…...29

4-1 Substrate specificity of recombinant C. reinhardtii VTC2………………………..…….37

4-2 Characterization of the GDP-hexose phosphorylase activities of the recombinant A.

thaliana and C. reinhardtii proteins………………………………………………..……38

4-S1 Enzymes of the Smirnoff-Wheeler vitamin C biosynthesis pathway have orthologs in

green algae………………………………………………………………………….……49

4-S2 Comparison of mRNA abundances for genes encoding enzymes of the L-galactose

ascorbate biosynthesis pathway in C. reinhardtii cells treated with 1 mM H2O2 for

30 and 60 minutes…………………………………………………………………...... 50

4-S3 Transcript abundances for genes encoding enzymes of the ascorbate-glutathione

cycle after exposure to 1 mM H2O2 for 30 and 60 minutes………………………...……51

4-S4 Primers sequences used for cloning of VTC2…………...……………………………….52

4-S5 Primers sequences used for real-time PCR……………...……………………………….52

5-1 AsA levels in C. elegans in wild type and tm2679 strains………………………….……75

5-2 Summary of AsA levels in different species………………………………………..……76

5-3 Search for C. elegans homologs of the enzymes of the Smirnoff-Wheeler and

the animal vitamin C biosynthesis pathways………………………………………….…77

xi 6-1 Characterization of GDP-hexose phosphorylase activities of recombinant human

C15orf58 and C. elegans C10F3.4 proteins……………………………………….……..89

6-2 Comparison of substrate specificity of human recombinant C15orf58

and of mouse brain GDP-D-Glc phosphorylase…………………………………...…….90

6-3 GDP-D-Man and GDP-D-Glc levels in C. elegans and various mouse tissues…………94

6-S1 Nucleotide specificity of recombinant C. elegans C10F3.4……………………..…..…103

6-S2 Comparison of the substrate specificity of C. elegans recombinant C10F3.4 and

of worm extract GDP-D-Glc phosphorylase activity……………………...……..……..104

7-1 Brood size analysis for N2 and mcp-1 mutant strains……………………...…………...119

7-2 The fraction of larvae with nuclei stained by Hoechst 33258 in wild type (N2),

bus-8, and mcp-1 mutants of C. elegans………………….…...………………………..120

xii

ACKNOWLEDGEMENTS

I am grateful to Professor Clarke for his guidance and support during my thesis work.

Professor Clarke is an avid educator with a passion for science that inspires everyone around him.

He has been an incredible mentor and advisor to which I will always be grateful. Thanks to his patience, guidance and commitment to excellence, I have been able to attain and incredible learning experience. I would like to thank him for his generosity in allowing me to participate in various conferences in France and the US. These experiences were fundamental in my academic growth and professional relationships. I also appreciate the fact that he has always been very supportive in allowing me to collaborate with several labs within the UCLA community and abroad. Thank you for accepting me into your prestigious lab and guiding to throughout my doctorate studies.

A special thank you to Dr. Carole Linster a former post-doctoral fellow in the Clarke lab and currently a research associate at the Luxembourg Centre for Systems Biomedicine at

Luxembourg University. When I joined the Clarke lab, Carole took me under her wings and I had the privilege to collaborate with her on many projects throughout my dissertation work.

Carole is a wonderful mentor and a great friend. Additionally, I would like to thank her for reviewing this work and for making significant suggestions.

I would like to thank all the past and present Clarke lab members that I had the privilege and the pleasure to work with throughout the years. It was wonderful to be part of such a supportive and collaborative group. Particularly, I would to acknowledge Brian D. Young, Dr.

Tara A. Gomez and Dr. Kris Webb with whom I collaborated on CHAPTERS two, three, five and six. I would especially like to thank my friends, Dr. Shilpi Khare, who introduced me to the

C. elegans world as well as to Drs. Cecilia Zurita Lopez and Jon Lowenson who were always

xiii there with good advice.

I would like to extend my appreciation to Dr. Alison Frand from the Department of

Biological Chemistry at UCLA for sharing her knowledge, ideas and expertise as well as for collaborating with me in the search for phenotypes in C. elegans’ mcp-1 mutants that is outlined in CHAPTER seven.

I would also like to thank Dr. Stuart Haslam and his team from the Faculty of Natural

Sciences Imperial College London, UK whose collaborative work to profile the differences between the glycans in mcp-1 mutant and wild type worms is detailed in APPENDIX I.

Finally, I would like to thank my committee members Dr. Catherine F. Clarke, Dr. Stephen

G. Young, Dr. Robert T. Clubb, Dr. Grody W. Wayne and Dr. Steven G. Clarke.

The work presented in this dissertation was supported by the UCLA Chancellor’s

Fellowship, the UCLA University Fellowship, the Whitcome Fellowship and by the Graduate

Division’s Dissertation Year Fellowship.

CHAPTER two of this dissertation is a reprint of the publication entitled “Arabidopsis

VTC2 Encodes a GDP-L-Galactose Phosphorylase, the Last Unknown Enzyme in the Smirnoff-

Wheeler Pathway to Ascorbic Acid in Plants” in the Journal of Biological Chemistry (Vol. 282,

18879-85). I would like to acknowledge the primary author Carole L. Linster and my co-authors

Steven G. Clarke, Charles Brenner, Tara A. Gomez, Brian D. Young and Kathryn C. Christensen.

CHAPTER three of this dissertation is a reprint of the publication entitled “A Second

GDP-L-Galactose Phosphorylase in Arabidopsis en Route to Vitamin C. Covalent Intermediate and Substrate Requirements for the Conserved Reaction” in the Journal of Biological Chemistry

(Vol. 283, 18483-92). I would like to acknowledge the primary author Carole L. Linster and my co-authors Steven G. Clarke, Charles Brenner, Kristofor Webb and Kathryn C. Christensen.

xiv CHAPTER four of this dissertation is a reprint of the publication entitled “Impact of

Oxidative Stress on Ascorbate Biosynthesis in Chlamydomonas via Regulation of the VTC2

Gene Encoding a GDP-L-Galactose Phosphorylase” in the Journal of Biological Chemistry (Vol.

287, 14234-45). I would like to acknowledge the primary author Eugen I. Urzica and my co- authors Steven G. Clarke, Sabeeha S. Merchant, Carole L. Linster, M. Dudley Page, Mark A.

Arbing, David Casero and Matteo Pellegrini.

CHAPTER six of this dissertation is a reprint of the publication entitled “A novel GDP-D-

Glucose Phosphorylase Involved in Quality Control of the Nucleoside Diphosphate Sugar Pool in Caenorhabditis elegans and Mammals” in the Journal of Biological Chemistry (Vol. 286,

21511-23). I would like to acknowledge my co-authors Steven G. Clarke, Carole L. Linster, and

Tara A. Gomez.

xv VITA

2000-2003 B.Sc. Medical Research Hebrew University Jerusalem, Israel

2001-2003 Research Assistant, Department of Anatomy and Cell Biology, Hadassah Medical School, Hebrew University, Jerusalem, Israel

2003-2005 Hebrew University Graduate School Scholarship for Academic Excellence

2003-2005 Graduate Student, Department of Radiobiology and Biotechnology Hadassah Medical School, Hebrew University, Jerusalem, Israel

2003-2005 Research Assistant HAPTO Biotech, Jerusalem, Israel

2005 Assistant to the Director, Microarray Unit Hadassah Medical School, Hebrew University, Jerusalem, Israel

2006-2008 Chancellor’s Fellowship UCLA

2008-2009 University Fellowship UCLA

2009-2010 University Fellowship UCLA

2009-2010 Whitcome Fellowship UCLA

2010-2011 Whitcome Fellowship UCLA

2011-2012 Graduate Division Award, Dissertation Year Fellowship UCLA

xvi PUBLICATIONS AND PRESENTATIONS

Urzica EI, Adler LN, Page MD, Linster CL, Arbing M, Casero D, Pellegrini M, Merchant SS, Clarke SG. Impact of Oxidative Stress on Ascorbate Biosynthesis in Chlamydomonas via Regulation of the VTC2 Gene Encoding a GDP-l-Galactose Phosphorylase. J Biol Chem. 2012 Apr 20;287(17):14234-45. Epub 2012 Mar 5

Adler LN, Gomez TA, Clarke SG, Linster CL. Sanitizing the Nucleoside Diphosphate Sugar Pool: Molecular Identification and Characterization of a GDP-d-Glucose Phosphorylase in Caenorhabditis elegans and Mammals. 22nd Joint Glycobiology Meeting , Lille Biology Institute, Lille , France. November 27-29 2011

Urzica EI, Adler LN, Casero D, Pellegrini M, Clarke SG and Merchant SS. RNA-seq of Iron- Limited Chlamydomonas Reveals Enzymatic and Non-Enzymatic Antioxidant Responses. Gordon Research Conference-Cell Biology of Metals, Salve Regina University, Newport, RI, USA July 31-August 5, 2011

Adler LN. From Vitamin C to Metabolic Repair: The Role of a Novel Sugar Nucleotide Phosphorylase. UCLA Worm Meeting, May 31, 2010

Adler LN, Gomez TA, Clarke SG, Linster CL. A Novel GDP-D-Glucose Phosphorylase Involved in Quality Control of the Nucleoside Diphosphate Sugar Pool in Caenorhabditis elegans and Mammals. J Biol Chem. 2011 June 17;286(24) 21511–23. Epub 2011 Apr 20

Adler LN. Protein repair, vitamin C and sugar nucleotide phosphorolysis: what is the connection? UCLA Worm Meeting, May 2009

Linster CL, Adler LN, Webb K, Christensen KC, Brenner C, Clarke SG. A second GDP-L- Galactose Phosphorylase in Arabidopsis en route to Vitamin C. Covalent Intermediate and Substrate Requirements for the Conserved Reaction. J Biol Chem. 2008 Jul 4;283(27):18483-92. Epub 2008 May 7

Linster CL, Gomez TA, Christensen KC, Adler LN, Young BD, Brenner C, Clarke SG. Arabidopsis VTC2 Encodes a GDP-L-Galactose Phosphorylase, the Last Unknown Enzyme in the Smirnoff-Wheeler Pathway to Ascorbic Acid in Plants. J Biol Chem. 2007 Jun 29;282(26):18879-85. Epub 2007 Apr 26

Makranz C, Cohen G, Baron A, Levidor L, Kodama T, Reichert F, Rotshenker S. Phosphatidylinositol 3-Kinase, Phosphoinositide-Specific Phospholipase-Cgamma and Protein Kinase-C Signal Myelin Phagocytosis Mediated by Complement Receptor-3 alone and Combined with Scavenger Receptor-AI/II in Macrophages. Neurobiol Dis. 2004 Mar;15(2):279- 86.

xvii

CHAPTER 1

Plan of the Dissertation

The protein L-isoaspartyl O-methyltransferase is evolutionarily conserved between many organisms and its chief role in higher organisms to repair age damaged aspartate and asparagine residues is well documented [1][2][3]. This enzyme in the roundworm Caenorhabditis elegans is encoded by the pcm-1 gene [4]. In C. elegans PCM-1 was found to have an additional role in lifespan extension via the insulin-like signaling pathway [5]. Interestingly, the genome sequence of pcm-1 revealed that an additional gene (mcp-1, C10F3.4) is transcribed in an antiparallel orientation on the opposite strand [6][7]. About six percent of the protein coding genes in C. elegans are overlapping genes [8]. However, only five gene pairs in the C. elegans genome were found to have a “gene pair with overlapping exons” [8] topology as seen for the pcm-1/ mcp-1 gene pair.

In general, overlapping genes were found to be more conserved compared to other genes in the C. elegans genome [8]. Additionally, gene pair analysis done by Chen and Stein discovered that while the expression pattern for overlapping genes encoded on opposite strands is generally correlated, overlapping gene pairs are generally not functionally related [8][9].

While the enzymatic activity and the role of PCM-1 in C. elegans were previously studied

[6][7][10], little was known about the activity and the function of MCP-1 [7]. The unique topology of the pcm-1/ mcp-1 gene pair raised many questions regarding the possible transcriptional correlation and the functional relationship between these genes.

The goal of my thesis project was to determine the activity and the role of MCP-1 in the soil nematode Caenorhabditis elegans and to understand whether its function is related to pcm-1 expression or activity. PSI-BLAST analysis revealed that MCP-1 had homologs in plants, vertebrates and invertebrates [7]. Specifically, the plant Arabidopsis thaliana homolog VTC2 displayed a 24% sequence identity with MCP-1 [7]. The amino acid alignment also revealed a

2 conserved HIT (histidine triad) motif along the homologs of A. thaliana, C. elegans and other species. The HIT motif (HxHxH) is characteristic of the HIT superfamily of enzymes and includes nucleotide hydrolases and transferases [11].

The enzymatic activity of VTC2 was not known at the time, but previous data demonstrated that Arabidopsis plants with a mutation in this gene display vitamin C deficiency

[12]. The main pathway for vitamin C biosynthesis in plants was proposed by Smirnoff and

Wheeler in 1999 [13]. In this ten-step pathway, D-glucose is converted to D-mannose 1- phosphate in four enzymatic steps followed by formation of GDP-D-mannose that is successively converted to GDP-L-galactose, L-galactose 1-phosphate, L-galactose, L-galactono-

1,4-lactone, and finally to L-ascorbate [13]. The only enzyme that had not been identified in this pathway was the one that catalyzes the conversion of GDP-L-galactose to L-galactose-1-P. In light of this information I decided to investigate the role of VTC2 in the route for vitamin C synthesis in plants.

CHAPTER 2 delineates the identification and characterization of VTC2 as a GDP-L- galactose phosphorylase that converts GDP-L-galactose to the next intermediate in the vitamin C biosynthesis pathway, L-galactose-1-P. This step is pivotal as intermediates produced in previous steps of the Smirnoff-Wheeler pathway can be used in other reactions in the plant while the phosphorolysis of GDP-L-galactose is the first dedicated or committed step in vitamin C biosynthesis [7][14]. In this project led by Dr. Carole Linster, I collaborated in the enzymological studies of the VTC2 activity.

Additional work outlined in CHAPTER 3 described the characterization of the

Arabidopsis VTC2 paralog VTC5. The enzymatic activity and substrate specificity of VTC5 was found to be similar to that of VTC2. This activity, which is dependent on phosphate as a

3 guanylyl acceptor, was also observed in extracts of seven additional plant species. Moreover, this work demonstrated that the HIT motif is essential for the catalytic activity of GDP-L-galactose phosphorylase as recombinant enzyme mutated in the second histidine of the HIT motif failed to complete the enzymatic reaction [15]. Formation of an enzyme intermediate covalently modified by GMP on this His residue was also demonstrated by mass spectrometry. In this study led by Dr.

Carole Linster, I collaborated with her in characterizing the enzymatic activity of VTC5, elucidating the catalytic mechanism of VTC2 and VTC5, and establishing the inorganic phosphate-dependency of the enzyme(s) converting GDP-L-galactose to L-galactose-1- phosphate in the path to vitamin C synthesis in higher plants.

While the biosynthesis of vitamin C in higher plants is relatively well established, this was not the case for green algae. The study delineated in CHAPTER 4 demonstrates that the

Smirnoff-Wheeler pathway is completely conserved in the green alga Chlamydomonas reinhardtii, and that GDP-L-galactose phosphorylase activity of the VTC2 homolog in this alga is similar to the one detected in higher plants [16]. This work also showed that following oxidative stress, C. reinhardtii cells were able to increase the total ascorbate content by elevating the mRNA transcript levels of the VTC2 homolog as well as the enzymes of the ascorbate- glutathione system that recycle oxidized ascorbate [16]. These results demonstrated the role of

VTC2 in regulating the de novo synthesis of vitamin C in the L-galactose pathway. In this project I collaborated with Dr. Eugen Urzica in the Sabeeha Merchant laboratory (Department of

Chemistry and Biochemistry, UCLA) to elucidate the enzymatic activity of the C. reinhardtii

VTC2 homolog and to measure the vitamin C levels in C. reinhardtii cells under various stress conditions.

4 After establishing the role of VTC2 in vitamin C synthesis in plants and algae, I moved on to ask whether the C. elegans homolog MCP-1 is also involved in vitamin C synthesis. The lack of published reports about the content (or the requirement for) ascorbate in C. elegans also raised questions regarding the presence of this important compound in this species.

The results of my studies to determine the levels of ascorbate in C. elegans and to investigate the possible role of MCP-1 in the biosynthesis of ascorbate in this nematode is delineated in CHAPTER 5. I was able to measure low levels of ascorbate in C. elegans extracts thus reporting for the first time the presence of ascorbate in this organism. Nevertheless, measurements of ascorbate levels in a strain with an mcp-1 deletion mutation (tm2679) showed that the ascorbate levels were not significantly different from the ones found for the wild type

(N2) strain. These results, together with bioinformatics analyses which revealed that the C. elegans genome lacks important homologs for the plant (or the animal) vitamin C synthesis pathway, led me to conclude that MCP-1 is not involved in vitamin C synthesis in C. elegans.

However, questions regarding the function of MCP-1 and its role in C. elegans still remained unanswered.

CHAPTER 6 of this thesis describes my work in characterizing the enzymatic activity of the C. elegans’ MCP-1 protein and its human homolog C15orf58. These two homologs were cloned, overexpressed, purified, and their enzymatic activities were determined. Both enzymes showed GDP-hexose phosphorylase activity like their plant and algae homologs. However, substrate specificity analysis revealed that unlike the plant and algae homologs, both the C. elegans and the human gene products clearly prefer GDP-D-glucose as a substrate over GDP-L- galactose, the plant precursor for vitamin C synthesis [17], thus supporting my previous conclusion that MCP-1 is not involved in vitamin C synthesis in C. elegans. In this work, I also

5 used a C. elegans transgenic strain expressing an MCP-1::GFP fusion protein to demonstrate that

MCP-1 is highly expressed in the neuronal and reproductive systems. Higher GDP-D-glucose phosphorylase activities were also measured in mouse brain and testis, thus showing that worms and mammals share similar tissue distribution of the protein. Further analysis revealed that mcp-

1 mutants, which lack GDP-D-glucose phosphorylase activity, accumulate GDP-D-glucose [17].

GDP-D-glucose production has been reported before in different mammalian tissues

[18][19][20][21], with GDP-D-mannose pyrophosphorylase A as the reported enzyme that catalyzes GDP-D-glucose synthesis [20]. While GDP-D-glucose could be used for cell wall synthesis in plants [22] and was also reported to be a precursor in the synthesis of the disaccharide trehalose in certain bacteria [23] [24], its physiological role in C. elegans and in mammals is unknown. Noting the high structural similarity between GDP-D-glucose and GDP-

D-mannose, I proposed that GDP-D-glucose could compete with GDP-D-mannose as a sugar donor in glycosylation reactions in the cell [25], therefore potentially inhibiting mannosyl incorporation or misincorporating glucose residues in place of mannose residues in the target glycoconjugates. GDP-D-glucose that accumulates under lack of GDP-D-glucose phosphorylase activity might be used as a substrate by mannosyltransferases that progressively add mannosyl units to the lipid-linked oligosaccharide (LLO) in the protein N-glycosylation process.

Whereas it is difficult to predict the consequences of incorporating glucose instead of mannose, some information can be gained from patients with congenital disorders of glycosylation type I (CDG-I) that have mutation in the genes involved in the assembly of the

LLO [26]. In these CDG-I patients it is typical to detect incomplete LLOs that result in hypoglycosylation of multiple glycoproteins [26].

6 I hypothesized that GDP-D-glucose synthesis is an unwanted side product of the GDP- mannose-pyrophosphorylase A activity and that in the absence of GDP-D-glucose phosphorylase the increased levels of GDP-D-glucose could potentially compete with GDP-D-mannose and lead to wrongly or incomplete glycosylation of glycoproteins and glycolipids (CHAPTER 6).

The assembly of LLO is conserved between mammals and C. elegans [29]. Nevertheless, unlike humans who display severe and lethal phenotypes as a result of mutations in genes that are involved in the N-glycosylation process [26][27], in C. elegans mutations in the homologous genes often display no apparent phenotype [28][29].

CHAPTER 7 of this thesis delineates my search for a quantifiable phenotype in the mcp-1 mutant strain that could be linked to a defect in the worms’ glycosylation process. I was able to find a few morphological changes in the mcp-1 mutants including dumpy appearance (Dpy phenotype), gonad migration (Mig phenotype), crushed pharynx, and egg-laying defect (Egl phenotype). Although these phenotypes can be linked to genes that affect the integrity of the glycosylation process in C. elegans (see CHAPTER 7), there is still a need to find a direct connection between the lack of GDP-D-glucose phosphorylase activity and a glycosylation defect that potentially led to the observed phenotypes. Further experiments are planned to investigate the role of MCP-1 as a metabolite repair enzyme and its contribution to the quality control of N-glycosylation in C. elegans.

APPENDIX I includes the work that has been done in collaboration with Dr. Stuart

Haslam and his colleagues at the Faculty of Natural Sciences Imperial College London, UK to discover changes in the glycome of mcp-1 mutant worms compared to control animals. The initial set of results revealed only subtle changes between the N-glycan profiles of the mcp-1 mutant and the wild type (N2) strain. Further work is in progress to confirm the initial findings

7 and to analyze the glycan profile of mcp-1 mutants that overexpress the worm homolog of GDP-

D-mannose pyrophosphorylase A.

REFERENCES

1. Clarke S. Aging as a war between chemical and biochemical processes: protein

methylation and the recognition of age-damaged proteins for repair. Ageing Res Rev

2003 Jul;2(3):263-85.

2. Kim E, Lowenson JD, MacLaren DC, Clarke S, Young SG. Deficiency of a protein-

repair enzyme results in the accumulation of altered proteins, retardation of growth, and

fatal seizures in mice. Proc Natl Acad Sci USA 1997 Jun 10:94 (12):6132-7.

3. Lowenson, J.D. and Clarke, S. Recognition of D-aspartyl residues in polypeptides by the

erythrocyte L-isoaspartyl/D-aspartyl protein methyltransferase. Implications for the repair

hypothesis. J Biol Chem, 1992 Mar 25;267(9):5985-95.

4. Kagan RM, Clarke S. Protein L-isoaspartyl methyltransferase from the nematode

Caenorhabditis elegans: genomic structure and substrate specificity. Biochemistry. 1995

Aug 29;34(34):10794-806.

5. Khare S, Linster CL, Clarke SG. The interplay between protein L-isoaspartyl

methyltransferase activity and insulin-like signaling to extend lifespan in Caenorhabditis

elegans. PLoS One. 2011;6(6):e20850. Epub 2011 Jun 13

6. Khare S, Gomez TA, Linster CL, Clarke SG. Defective responses to oxidative stress in

protein l-isoaspartyl repair-deficient Caenorhabditis elegans. Mech Ageing Dev 2009

Oct;130(10):670-80. Epub 2009 Aug 12

7. Linster CL, Gomez TA, Christensen KC, Adler LN, Young BD, Brenner C, Clarke SG.

Arabidopsis VTC2 encodes a GDP-L-galactose phosphorylase, the last unknown enzyme

9 in the Smirnoff-Wheeler pathway to ascorbic acid in plants. J Biol Chem. 2007 Jun

29;282(26):18879-85. Epub 2007 Apr 26

8. Chen N, Stein LD. Conservation and functional significance of gene topology in the

genome of Caenorhabditis elegans. Genome Res. 2006 May;16(5):606-17.

Epub 2006 Apr 10

9. Cutter AD, Dey A, Murray RL. Evolution of the Caenorhabditis elegans genome. Mol

Biol Evol. 2009 Jun;26(6):1199-234. Epub 2009 Mar 16

10. Gomez TA, Banfield KL, Clarke SG. The protein L-isoaspartyl-O-methyltransferase

functions in the Caenorhabditis elegans stress response. Mech Ageing

Dev. 2008 Dec;129(12):752-8. Epub 2008 Oct 14

11. Brenner C. Hint, Fhit, and GalT: function, structure, evolution, and mechanism of three

branches of the histidine triad superfamily of nucleotide hydrolases and transferases.

Biochemistry. 2002 Jul 23;41(29):9003-14.

12. Conklin PL, Saracco SA, Norris SR, Last RL. Identification of ascorbic acid-deficient

Arabidopsis thaliana mutants. Genetics. 2000 Feb;154(2):847-56.

13. Wheeler GL, Jones MA, Smirnoff N. The biosynthetic pathway of vitamin C in higher

plants. Nature. 1998 May 28;393(6683):365-9.

14. Linster CL, Clarke SG. L-Ascorbate biosynthesis in higher plants: the role of VTC2.

Trends Plant Sci. 2008 Nov;13(11):567-73. Epub 2008 Sep 27

15. Linster CL, Adler LN, Webb K, Christensen KC, Brenner C, Clarke SG. A second GDP-

L-galactose phosphorylase in arabidopsis en route to vitamin C. Covalent intermediate

and substrate requirements for the conserved reaction. J Biol Chem. 2008 Jul

4;283(27):18483-92. Epub 2008 May 7

10 16. Urzica EI, Adler LN, Page MD, Linster CL, Arbing MA, Casero D, Pellegrini M,

Merchant SS, Clarke SG. Impact of Oxidative Stress on Ascorbate Biosynthesis in

Chlamydomonas via Regulation of the VTC2 Gene Encoding a GDP-L-galactose

Phosphorylase. J Biol Chem. 2012 Apr 20;287(17):14234-45. Epub 2012 Mar 5

17. Adler LN, Gomez TA, Clarke SG, Linster CL. A novel GDP-D-glucose phosphorylase

involved in quality control of the nucleoside diphosphate sugar pool in Caenorhabditis

elegans and mammals. J Biol Chem. 2011 Jun 17;286(24):21511-23. Epub 2011 Apr 20

18. Verachtert,H.,Rodriguez,P, Bass,S.T.,and Hansen,R.G. Purification and properties of

guanosine diphosphate hexose pyrophosphorylase from mammalian tissues. J Biol

Chem. 1966 May 10;241(9):2007-13.

19. Szumilo T, Drake RR, York JL, Elbein AD. GDP-mannose pyrophosphorylase.

Purification to homogeneity, properties, and utilization to prepare photoaffinity analogs. J

Biol Chem. 1993 Aug 25;268(24):17943-50.

20. Ning, B., and Elbein, A. D. Cloning, expression and characterization of the pig liver

GDP-mannose pyrophosphorylase. Evidence that GDP-mannose and GDP-Glc

pyrophosphorylases are different proteins. Eur. J. Biochem. 2000 Dec;267(23):6866-74.

21. Carlson DM, Hansen RG. The isolation and synthesis of guanosine diphosphate glucose.

J Biol Chem. 1962 Apr;237:1260-5.

22. Liepman AH, Wilkerson CG, Keegstra K. Expression of cellulose synthase-like (Csl)

genes in insect cells reveals that CslA family members encode mannan synthases. Proc

Natl Acad Sci U S A. 2005 Feb 8;102(6):2221-6. Epub 2005 Jan 12

23. Silva Z, Alarico S, da Costa MS. Trehalose biosynthesis in Thermus thermophilus RQ-1:

biochemical properties of the trehalose-6-phosphate synthase and trehalose-6-phosphate

11 phosphatase. Extremophiles. 2005 Feb;9(1):29-36. Epub 2004 Sep 29Elbein AD, Pan YT,

Pastuszak I, Carroll D. New insights on trehalose: a multifunctional molecule.

Glycobiology. Apr;13(4):17R-27R. Epub 2003 Jan 22

24. Elbein AD, Pan YT, Pastuszak I, Carroll D. New insights on trehalose: a multifunctional

molecule. Glycobiology. Apr;13(4):17R-27R. Epub 2003 Jan 22

25. Struck DK, Lennarz WJ. Utilization of Exogenous GDP-Mannose for the Synthesis of

Mannose-containing Lipids and Glycoproteins by Oviduct Cells.J Biol Chem. 1976 Apr

25;251(8):2511-9.

26. Haeuptle MA, Hennet T. Congenital disorders of glycosylation: an update on defects

affecting the biosynthesis of dolichol-linked oligosaccharides. Hum Mutat. 2009

Dec;30(12):1628-41.

27. Jaeken J. Congenital disorders of glycosylation. Ann N Y Acad Sci. 2010 Dec;1214:190-

8. doi: 10.1111/j.1749-6632.2010.05840.x. Review.

28. Weston B Struwe, Bethany L Hughes, David W Osborn, Erica D Boudreau, Kristin M D

Shaw, and Charles E Warren. Modeling a congenital disorder of glycosylation type I

in C. elegans: A genome-wide RNAi screen for N-glycosylation-dependent loci.

Glycobiology. 2009 Dec;19(12):1554-62. Epub 2009 Sep 3

29. Gravato-Nobre MJ, Stroud D, O'Rourke D, Darby C, Hodgkin J. Glycosylation genes

expressed in seam cells determine complex surface properties and bacterial adhesion to

the cuticle of Caenorhabditis elegans. Genetics. 2011 Jan;187(1):141-55. Epub 2010 Oct

CHAPTER 2

Arabidopsis VTC2 Encodes a GDP-L-Galactose Phosphorylase, the Last

Unknown Enzyme in the Smirnoff-Wheeler Pathway to Ascorbic Acid in Plants

Arabidopsis VTC2 Encodes a GDP-L-Galactose Phosphorylase, the Last Unknown Enzyme in the Smirnoff-Wheeler Pathway to Ascorbic Acid in Plants*ࡗ Received for publication, March 9, 2007, and in revised form, April 25, 2007 Published, JBC Papers in Press, April 26, 2007, DOI 10.1074/jbc.M702094200 Carole L. Linster‡1, Tara A. Gomez‡, Kathryn C. Christensen§, Lital N. Adler‡, Brian D. Young‡, Charles Brenner§2, and Steven G. Clarke‡3 From the ‡Department of Chemistry and Biochemistry and the Molecular Biology Institute, UCLA, Los Angeles, California 90095 and the §Departments of Genetics and Biochemistry, Dartmouth Medical School, Lebanon, New Hampshire 03756

The first committed step in the biosynthesis of L-ascorbate reduction and lactonization of D-glucuronate to L-gulonolac-

from D-glucose in plants requires conversion of GDP-L-galac- tone and oxidation of the latter to L-ascorbate (reviewed in Ref. Downloaded from tose to L-galactose 1-phosphate by a previously unidentified 7). Deficiency of the enzyme catalyzing this last step (L-gulono- enzyme. Here we show that the protein encoded by VTC2, a gene lactone oxidase) is responsible for the loss of ascorbate synthe- mutated in vitamin C-deficient Arabidopsis thaliana strains, is a sis in the vitamin C-requiring vertebrates (8). In plants, the member of the GalT/Apa1 branch of the histidine triad protein ascorbate synthesis pathway has remained elusive until recently superfamily that catalyzes the conversion of GDP-L-galactose to and alternative pathways may exist (9). The Smirnoff-Wheeler www.jbc.org L-galactose 1-phosphate in a reaction that consumes inorganic pathway (10) has garnered strong biochemical and genetic sup- phosphate and produces GDP. In characterizing recombinant port (11–16) and appears to represent the major route to ascor- VTC2 from A. thaliana as a specific GDP-L-galactose/GDP-D- bate biosynthesis. In this pathway, GDP-D-mannose, formed D glucose phosphorylase, we conclude that enzymes catalyzing from -mannose 1-phosphate, is successively converted to at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 each of the ten steps of the Smirnoff-Wheeler pathway from GDP-L-galactose, L-galactose 1-phosphate, L-galactose, L-ga- glucose to ascorbate have been identified. Finally, we identify lactono-1,4-lactone, and finally to L-ascorbate. VTC2 homologs in plants, invertebrates, and vertebrates, sug- Screens for ozone-sensitive (17) or ascorbate-deficient (18) gesting that a similar reaction is used widely in nature. mutants in Arabidopsis thaliana led to the identification of four loci (VTC1, VTC2, VTC3, and VTC4) involved in the maintenance of the vitamin C pool. Characterization of the vtc1 (19) Vitamin C (L-ascorbic acid) is well known as an important and vtc4 (15) mutants, as well as biochemical studies (14), have antioxidant and enzyme cofactor in animals (1, 2) and in plants allowed the identification of two of the enzymes required for (3). Apparently, all plants are able to produce vitamin C and L-ascorbic acid synthesis through the Smirnoff-Wheeler path- mutants completely deficient in synthesis have not been way. VTC1 and VTC4 encode GDP-mannose pyrophosphory- 4 described, suggesting an essential role of ascorbate biosynthesis lase (19) and L-Gal-1-P phosphatase (15), respectively. The in these organisms (4). Although vertebrate vitamin C synthesis function of the VTC2 and VTC3 genes has not yet been eluci- is restricted to one organ (liver in mammals and kidney in fish, dated. The only step in the Smirnoff-Wheeler pathway that has amphibians, and reptiles) (5, 6), virtually all cells in plants can not been identified with a gene product is the conversion of form ascorbate (4). In the few vertebrate species, such as GDP-L-Gal to L-Gal-1-P. humans, which lack ascorbate biosynthesis, loss of the pathway Here we show that the product of the A. thaliana VTC2 gene, is compensated by dietary intake, particularly from plants. which has previously been cloned (20), is broadly conserved in Different pathways of ascorbate synthesis have evolved in plants and in animals. On the basis of sequence similarity and animals and plants. In animals, ascorbate is formed from UDP- enzymatic analysis, we characterize VTC2 as a histidine triad D-glucuronate in a pathway involving D-glucuronate formation, (HIT) enzyme of the GalT/Apa1 branch possessing a high GDP-L-Gal phosphorylase activity, which produces GDP and * This work was supported by National Institutes of Health Grants GM026020 the ascorbic acid precursor L-Gal-1-P. This finding completes (to S. G. C.; with a supplement for T. A. G.), AG018000 (to S. G. C.), and D CA75954 (to C. B.) and by National Science Foundation Grant the identification of the ten enzymatic steps from -glucose to MCB-0448533 (to S. G. C.). The costs of publication of this article were L-ascorbate in the Smirnoff-Wheeler pathway for vitamin C defrayed in part by the payment of page charges. This article must there- biosynthesis in plants. fore be hereby marked “advertisement” in accordance with 18 U.S.C. Sec- tion 1734 solely to indicate this fact. EXPERIMENTAL PROCEDURES ࡗ This article was selected as a Paper of the Week. 1 Supported by a research fellowship (bourse de formation-recherche) from Materials—ADP-D-Glc, GDP, GDP-D-Glc, GDP-D-Man, the Luxembourgish Government. UDP-D-Gal, UDP-D-Glc, D-Gal-1-P, and D-Man-1-P were from 2 To whom correspondence may be addressed. Tel.: 603-653-9922; Fax: 603- 653-9923; E-mail: [email protected]. 3 To whom correspondence may be addressed: Dept. of Chemistry and Bio- chemistry and the Molecular Biology Institute, UCLA, 607 Charles E. Young 4 The abbreviations used are: Gal-1-P, galactose 1-phosphate; GDP-L-Gul, Dr. East, Los Angeles, CA 90095. Tel.: 310-825-8754; Fax: 310-825-1968; GDP-L-gulose; Glc-1-P, glucose 1-phosphate; GME, GDP-D-mannose 3Ј,5Ј- E-mail: [email protected]. epimerase; HIT, histidine triad; Man-1-P, mannose 1-phosphate.

JUNE 29, 2007•VOLUME 282•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 18879

14 GDP-L-Galactose Phosphorolysis by VTC2 in Ascorbate Synthesis

GDP-D-Man with GME (46.5 ␮g/ml) for 5.5 h at 26 °C in 50 mM Tris-HCl, pH 7.5. After heating for 3 min at 98 °C, denatured GME was removed by centrifugation and the supernatant (containing 0.77 mM GDP-D-Man, 0.14 mM GDP-L-Gal, and 0.03 mM GDP-L-Gul) was used for determination of the kinetic constants of VTC2 for GDP-L-Gal. Cloning, Expression, and Purifica- tion of A. thaliana VTC2—The A. thaliana VTC2 coding sequence was amplified by PCR from a cDNA clone (U17921) provided by the Arabidopsis Biological Resource Downloaded from Center at Ohio State University. This clone contains the coding region of the At4g26850 locus on chromosome 4 (GenBankTM acces-

sion number BT006589) and was www.jbc.org prepared by the Salk Institute Genomic Analysis Laboratory (22). The forward and reverse primers were 5Ј-CACCATGTTGAAAAT- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 CAAAAGAG and 5Ј-TCACT- GAAGGACAAGGCACTCGG. The PCR product was cloned into the Champion pET100/D-Topo vector (Invitrogen), resulting in a construct adding a 36-amino acid N-terminal extension (MRGSHHHHHHGMA- SMTGGQQMGRDLYDDDDKD- HPFT) to the VTC2 protein. The DNA sequence of the insert was confirmed. The plasmid was transformed into E. coli BL21 Star (DE3) cells and grown until the optical density at 600 nm reached 0.5–0.6. Protein overexpression was induced with 0.4 mM isopropyl ␤-D-1-thiogalactopyranoside. After overnight incubation with shaking at 18 °C, cells were harvested, washed, fro- FIGURE 1. Amino acid alignment of VTC2 homologs. The A. thaliana VTC2 sequence is aligned with those of zen, resuspended in a pH 7.9 Oryza sativa, Homo sapiens, C. elegans and Drosophila melanogaster. Residues boxed in black are identical in buffer containing 20 mM Tris-HCl, three or more of the proteins. Residues boxed in gray are chemically similar. Asterisks mark the residues that are M M altered in the vtc2-2 and vtc2-3 A. thaliana mutant strains, namely G224D and S290F. The HIT motif is indicated. 500 m NaCl, 5 m imidazole, 0.1 Accession numbers include BT006589 for A. thaliana, NP_001066338 for O. sativa, BAC85370 for H. sapiens, mM phenylmethylsulfonyl fluo- NP_001023638 for C. elegans, and NP_648319 for D. melanogaster. ride, and 1 ␮g/ml leupeptin, and lysed by sonication. Clarified ϩ Sigma. All of the sugar nucleotides and sugar phosphates were lysate was loaded onto an Ni2 -charged column (Novagen, in the ␣-configuration. L-Gal-1-P, in the ␤-configuration, was San Diego, CA) and eluted using a gradient of the resuspen- purchased from Glycoteam (Hamburg, Germany). All other sion buffer containing 5 mM to 1 M imidazole. Peak VTC2 reagents were of analytical grade. Recombinant A. thaliana protein fractions were pooled and buffer was exchanged to GDP-D-mannose 3Ј,5Ј-epimerase (GME) containing an N-ter- 10 mM HEPES, pH 7.5, 100 mM NaCl, 1 mM dithiothreitol in minal His tag was prepared as described (21) from a construct a10kDacut-offAmiconUltraCentrifugalFilterUnit(Mil- provided by Prof. James H. Naismith (University of St. lipore, Billerica, MA). Protein concentration was deter- Andrews, UK). GDP-L-Gal was prepared by incubating 0.94 mM mined using a Lowry assay after precipitation with trichlo-

18880 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 26•JUNE 29, 2007 15 GDP-L-Galactose Phosphorolysis by VTC2 in Ascorbate Synthesis

respectively. To assay the enzymatic activity in the reverse direction (hexose 1-phosphate ϩ GDP 3 ϩ GDP-hexose Pi), GDP-hexose concentration was measured by the anion-exchange HPLC method after incubation of VTC2 with various hexose 1-phosphates and 5 mM GDP as described above, except that

sodium phosphate and MgCl2 were omitted. GDP and GDP-hexose concentrations were calculated by comparing the integrated peak areas with those of standard GDP or GDP-D-Man solutions. Mass Spectrometry—Matrix-as- Downloaded from sisted laser desorption/ionization mass spectrometry was performed on a Voyager-DE STR MALDI-TOF instrument. GDP-D-Man, GDP-L-

Gal and GDP-L-Gul were purified www.jbc.org by reverse-phase HPLC (21) using an Econosphere C-18 column (5 ␮m bead size, 4.6 ϫ 250 mm; Alltech Associates, Deerfield, IL). Fractions at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 (15 s) were collected, dried, and resuspended in a solution of 0.29 M 3-hydroxypicolinic acid with 9 ␮M GDP as an internal calibrant in ace-

tonitrile/H2O/trifluoroacetic acid

FIGURE 2. Dependence of VTC2 activity on GME and Pi. GDP-D-Man (A) is not a hydrolytic (B) nor phospho- (50/50/0.1, v/v/v). The samples rolytic (C) substrate of VTC2. The equilibrium reaction products of GME activity on GDP-D-Man (D), namely were analyzed in positive ion GDP-D-Man, GDP-L-Gal, and GDP-L-Gul are not hydrolytic substrates of VTC2 (E). However, VTC2 catalyzes a phosphate-dependent reaction that preferentially consumes GDP-L-Gal (F). In these experiments, samples mode and the mass accuracy was were preincubated in the absence of VTC2 for 20 min at 26 °C in the presence of 0.14 mM GDP-D-Man with or ϳ10–30 ppm. without 23 ␮g/ml GME in a pH 7.5 buffer containing 50 mM Tris-HCl, 2 mM MgCl2, 10 mM NaCl, and 1 mM dithiothreitol. When indicated, sodium phosphate was added at a concentration of 5 mM. Incubations were RESULTS then continued for 10 min at 26 °C with (B, C, E, F) or without (A, D) VTC2 at 6.5 ␮g/ml. The reactions were stopped by heating, and 60 ␮l of the deproteinized reaction mixtures were adjusted to 100 mM NaCl and In our study of aging in the analyzed by reverse-phase HPLC as described by Major et al. (21) with the exception that an Econosphere C-18 column (5 ␮m bead size, 4.6 ϫ 250 mm; Alltech Associates, Deerfield, IL) was used. The results shown are nematode worm Caenorhabditis representative of at least three independent experiments. Elution times for GMP and GDP were 21 and 43 min, elegans, we became interested in the respectively. C10F3.4 gene that partially overlaps the pcm-1 gene encoding the roacetic acid. Purified enzyme was stored at Ϫ80 °C in 10% L-isoaspartyl protein methyltransferase (23). We examined the glycerol. deduced amino acid sequence of C10F3.4,andasshownin Enzymatic Assays—Phosphorylase activity of purified Fig. 1, found apparent orthologs in invertebrates, verte- recombinant A. thaliana VTC2 was assayed by measuring GDP brates, and plants, including the VTC2 gene product of A. formation after incubation with various GDP-hexoses in a reac- thaliana.ByperformingPSI-BLASTanalyses(24)ofthe tion mixture at pH 7.5 containing 50 mM Tris-HCl, 5 mM C10F3.4 and VTC2 gene products, we detected a low level of

sodium phosphate, 2 mM MgCl2, 10 mM NaCl, and 1 mM dithi- similarity with members of the HIT protein superfamily. othreitol. Reactions (26 °C) were initiated with enzyme and HIT enzymes consist of nucleoside monophosphate hydro- stopped after 5 to 60 min by heating at 98 °C for 3 min. After lases related to Hint, Fhit, Aprataxin, and DcpS and nucleo- removal of precipitated protein by centrifugation, supernatants side monophosphate transferases related to GalT, the D-ga- were analyzed by anion-exchange HPLC on a Partisil SAX col- lactose 1-phosphate uridylyltransferase and Apa1, a umn (10 ␮m bead size, 4.6 ϫ 250 mm; Alltech Associates, Deer- diadenosine tetraphosphate phosphorylase (reviewed in Ref. field, IL) using a Hewlett Packard Series II 1090 liquid chro- 25). In A. thaliana,thevtc2-2 and vtc2-3 missense mutations

matograph. A gradient of 0.01–0.5 M NH4H2PO4, pH 3.7, was (20) map to either side of the HIT motif (Fig. 1). Since these used at a flow rate of 2 ml/min. Nucleotides were detected by mutants are vitamin C-deficient (18), we considered the pos- absorbance at 254 nm using a reference wavelength of 450 nm. sibility that a step in vitamin C biosynthesis is catalyzed by a GMP, GDP-hexoses, and GDP eluted at ϳ13, 17, and 24 min, HIT enzyme.

JUNE 29, 2007•VOLUME 282•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 18881 16 GDP-L-Galactose Phosphorolysis by VTC2 in Ascorbate Synthesis

nucleoside monophosphate moiety of substrates, forming a covalent nucleotidylylated enzyme. In the case of HIT hydrolases, the nucleotidylylated intermediate is labile to water, such that it rapidly decom- poses to enzyme plus nucleoside monophosphate. In contrast, the nucleotidylylated intermediate of HIT transferases such as GalT and Apa1 is stable to water and awaits reaction with phosphate (phosphorolysis) or a specific phosphorylated substrate (transfer). Given the lack of conservation of the third His residue in the HIT motif of the plant Downloaded from VTC2 sequence and the requirement of an intact VTC2 gene for maintenance of the ascorbate pool in A. thaliana, we hypothesized that

VTC2 may be a GDP-L-galactose www.jbc.org phosphorylase that would produce L-Gal-1-P plus GDP in a reaction requiring inorganic phosphate. Despite evidence for the activity, the at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 enzyme that forms L-Gal-1-P remained unknown (11), and it has been widely assumed that such an enzyme would simply hydrolyze GDP-L-Gal into GMP plus L-Gal- 1-P (10). To test the hypothesis that VTC2 catalyzes the phosphorolysis of GDP-L-Gal, we purified recombinant A. thaliana VTC2 as an N-terminal His-tagged protein from a bacterial expression system. Because the GDP-L-Gal substrate was not commercially available, we prepared it enzymatically from FIGURE 3. VTC2-catalyzed phosphorolysis of GDP-L-Gal and formation of GDP. GDP-D-Man was preincu- GDP-D-Man using the recombinant bated for 30 min with GME and 5 mM sodium phosphate as described in Fig. 2. After removal of heat-inactivated GDP-D-mannose 3Ј,5Ј-epimerase (3 min at 98 °C) GME, GDP-hexose consumption and GDP formation were monitored after incubation for the indicated times with VTC2 (6.5 ng/ml). Deproteinized reaction mixtures were analyzed by reverse-phase HPLC (GME) from A. thaliana (21) in a (A–C) as described in Fig. 2 and anion-exchange HPLC (100 ␮l samples, E–G) as described under “Experimental coupled assay. We measured the Procedures.” A small amount of GDP (less than 1% of the total GDP-hexose pool) was detected at 0 min as a consumption of GDP-D-Man in the VTC2-independent hydrolysis product. The HPLC traces shown are representative of three separate experiments. In D and H, the concentrations of GDP-D-Man, GDP-L-Gal, GDP-L-Gul, and GDP are given as means Ϯ S.D. presence of VTC2 without and with values calculated from these experiments. Only error bars exceeding the symbol sizes are represented. prior reaction with GME and with-

out and with Pi. In the absence of HIT hydrolases and HIT transferases can be distinguished GME, less than 2% of the initial GDP-D-Man was consumed

both by sequence and by enzymatic requirements. HIT hydro- after a 10-min incubation with VTC2, even when Pi was present lases typically contain an active site motif including His-␾-His- in the reaction mixture (Fig. 2C). GME catalyzes the reversible ␾-His, where ␾ is a hydrophobic amino acid. In contrast, the epimerization of GDP-D-Man to form GDP-L-Gal and, to a third His of the HIT motif is not conserved in HIT trans- lesser extent, GDP-L-gulose (GDP-L-Gul) (21, 26). The sub- ferases and is frequently a Gln residue (25). As shown in Fig. strate and two products can be separated by reverse-phase 1, plant VTC2 sequences have a His-Leu-His-Phe-Gln HPLC (21). We confirmed this chromatographic separation sequence, whereas animal VTC2 homologs contain a His- with comparable elution times and an equilibrium product Leu-His-(Leu/Phe)-His sequence. All known HIT enzymes distribution of 82:14:3 for GDP-D-Man, GDP-L-Gal, and use the second His residue of the HIT motif to attack the GDP-L-Gul (Fig. 2D), similar to that reported previously (80:

18882 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 26•JUNE 29, 2007 17 GDP-L-Galactose Phosphorolysis by VTC2 in Ascorbate Synthesis

15:5; Ref. 21). Furthermore, we confirmed the identification TABLE 1 of the GME products by UV absorption spectrum (identical Substrate specificity of the A. thaliana VTC2 phosphorylase to that of standard GDP-D-Man), as well as by mass spec- Km and Vmax values were obtained by fitting the experimental data to the Michaelis- Menten equation using the Km calculator of the BioMechanic program. Enzymatic trometry with the detection of a positive ion of m/z 606, turnover numbers were derived from the Vmax values by using a molecular weight of consistent with all three nucleotide sugars. 53.1 kDa for recombinant His-tagged VTC2 with the assumption that the enzyme preparation is pure. Incubation times and enzyme concentrations were adjusted to Addition of VTC2 in the absence of Pi had no effect on the obtain initial velocity data. Values are means Ϯ S.D. calculated from at least three concentration of the GDP-hexoses present in the GME reaction individual experiments for each substrate. Enzymatic activities were measured as described under “Experimental Procedures,” and substrate concentrations ranged mixture (Fig. 2E). However, in the presence of Pi, VTC2 addi- from 3.4–54 ␮M (GDP-L-Gal), 2.5–100 ␮M (GDP-D-Glc), 0.25–5 mM (GDP-D- tion led to a decrease of the three GDP-hexose peaks and, sig- Man), and 2.5–40 mM (hexose 1-phosphates). nificantly, a modification of the ratio of these peaks (Fig. 2F). A. Forward reaction k K k /K After a 10-min incubation with VTC2 and 5 mM Pi, GDP-D- cat m cat m Ϫ1 Ϫ1 Ϫ1 Man, GDP-L-Gal, and GDP-L-Gul were found in the ratio s mM s M Substrate 97:1:2, indicating that VTC2 preferentially acts on GDP-L-Gal. 6 GDP-L-Gal 64 Ϯ 8 0.010 Ϯ 0.001 6.3 Ϯ 0.9 ϫ 10 6 The Pi dependence of this enzymatic reaction is consistent with GDP-D-Glc 23 Ϯ 3 0.0044 Ϯ 0.0016 5.7 Ϯ 2.3 ϫ 10 2 the prediction from sequence analysis above that VTC2 is a GDP-D-Man 0.093 Ϯ 0.011 0.52 Ϯ 0.15 1.9 Ϯ 0.3 ϫ 10 phosphorylase rather than a hydrolase. B. Reverse reaction Downloaded from To confirm the activity of VTC2 with GDP-L-Gal, we incu- Substrate L-Gal-1-P 6.0 Ϯ 0.4 45 Ϯ 7 133 Ϯ 12 bated VTC2 in the presence of Pi with the GDP-hexose mixture D-Glc-1-P 12 Ϯ 0.5 29 Ϯ 4 412 Ϯ 50 Ϯ Ϯ Ϯ formed by reaction of GME with GDP-D-Man but under con- D-Man-1-P 0.0013 0.0001 54 8 0.025 0.001 D-Gal-1-P 0.023 Ϯ 0.002 67 Ϯ 3 0.34 Ϯ 0.02 ditions where GME was inactivated by heating prior to the

addition of VTC2. Under these conditions, VTC2 only affected www.jbc.org

the GDP-L-Gal concentration, which decreased linearly with sponding GDP-sugar derivatives. The kcat terms for the reverse incubation time (Fig. 3, A–D). Parallel analysis of the reaction reaction with these hexose 1-phosphates are, however, rela- mixtures by anion-exchange HPLC clearly showed that the tively high (2–10-fold lower than those found in the direct reac- consumption of GDP-L-Gal was correlated with the appearance tion) and exceed by 2 and 3 orders of magnitude those meas- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 of a peak eluting at ϳ24 min and identified as GDP by UV ured with D-Gal-1-P and D-Man-1-P, respectively. Thus, VTC2 spectrum and coelution with standard (Fig. 3, E–H). No forma- functions much better in the reverse reaction with L-Gal-1-P or tion of GMP was detected in these experiments and the pro- D-Glc-1-P than with D-Man-1-P or D-Gal-1-P, confirming the duction of GDP was precisely matched to the consumption of forward reaction products and the specificity. GDP-L-Gal. DISCUSSION To characterize the substrate specificity of VTC2, we incubated several nucleoside diphosphate hexoses at a concentra- Our demonstration that VTC2 catalyzes the conversion of

tion of 1 mM for 10 min (26 °C) in the presence of Pi with or GDP-L-Gal to L-Gal-1-P completes the identification of genes without VTC2 (32.5 ␮g/ml). NDP-sugar consumption (and encoding the enzymes of the Smirnoff-Wheeler pathway of NDP formation) were measured by anion-exchange HPLC. No ascorbate synthesis in plants (Fig. 4). We characterize VTC2 as phosphorylase activity was detected when UDP-D-Glc, UDP-D- a GDP-L-Gal/GDP-D-Glc phosphorylase. The phosphate Gal, or ADP-D-Glc were used as substrates and only a very low requirement of VTC2 and its production of GDP are consistent activity was found in the presence of GDP-D-Man. However, an with the presence of a Gln residue at position 5 of the HIT motif almost total conversion of the NDP-sugar to NDP was observed (25). Two well characterized HIT transferases are GalT and with GDP-D-Glc in the incubation conditions used (data not Apa1. GalT reacts with UDP-D-Glc to produce a uridylylated shown). enzyme intermediate plus D-Glc-1-P. The uridylylated enzyme Kinetic analysis of the VTC2 phosphorylase activity con- is then intercepted by D-Gal-1-P to produce UDP-D-Gal (27). In firmed that GDP-D-Glc is a good substrate for VTC2 (Table 1). the case of Apa1, the substrate is diadenosine tetraphosphate, A mixture obtained by GME incubation with GDP-D-Man (and which reacts to form adenylylated enzyme plus ATP. The ad- subsequent elimination of GME) was used to determine the enylylated intermediate, stable to water, is phosphorolyzed by

kinetic properties of VTC2 for the phosphorolysis of GDP-L- Pi to produce an ADP product (28). It can thus be predicted that Gal, under conditions where GDP formation reflects only this VTC2, in the presence of GDP-L-Gal or GDP-D-Glc, forms a

reaction. VTC2 also exhibited a low Km value and a high turn- guanylylated enzyme intermediate at His-238, which is phos- over rate with GDP-L-Gal, the values obtained resembling phorolyzed by Pi to yield free enzyme and GDP. closely those measured with GDP-D-Glc. GDP-D-Man, on the As shown in Fig. 4, conversion of D-Glc to L-ascorbic acid other hand, is a poorer substrate for VTC2 by a factor of through the Smirnoff-Wheeler pathway takes a remarkable 10 ϳ 30,000 in the kcat/Km term (Table 1). steps, 8 of which are required simply to convert D-Glc to L-Gal. VTC2 activity could also be measured in the reverse direc- L-Gal, representing the epimerization product of D-Gal at four tion by incubating the enzyme in the presence of high concen- carbon atoms, is rarely found in nature (29), emphasizing the trations of GDP and various hexose 1-phosphates, and assaying importance of this portion of the pathway. Because two of the

GDP-hexose formation in the absence of added Pi. Table 1 interconverted substrates and products of GME, namely GDP- shows that the Km values of VTC2 for L-Gal-1-P and D-Glc-1-P D-Man and GDP-L-Gal, are utilized in biosynthesis of cell wall are more than 3 orders of magnitude higher than for the corre- polysaccharides (30, 31) and/or protein glycosylation (32), the

JUNE 29, 2007•VOLUME 282•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 18883 18 GDP-L-Galactose Phosphorolysis by VTC2 in Ascorbate Synthesis

Wheeler pathway, a commodity sugar is converted to a specialty sugar by virtue of epimerization prior to substrate selection by the HIT enzyme (Fig. 4). VTC2 exhibits striking substrate specificity for GDP-L-Gal over GDP-D-Man and no apparent activity on GDP-L-Gul. This high selec- tivity of VTC2 for one of the GME products pulls the equilibrium of the GME-catalyzed reversible reaction in favor of GDP-D-Man conversion to GDP-L-Gal, which is made irreversible in the cell by the activity Downloaded from of the specific L-Gal-1-P phosphatase, VTC4 (14, 15). The specificity of these enzymes suggests that the main flux of ascorbate biosynthesis passes through the L-galactose

branch, represented in Fig. 4, and www.jbc.org not a putative L-gulose branch through formation of L-gulono-1,4- lactone, the vitamin C precursor in animals. The L-gulose branch has at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 been suggested as a possible alternative ascorbate biosynthesis pathway (26), but enzymes for it have not been identified in plants (9). While VTC2 exhibits strong dis- crimination against the GME equilibrium products, GDP-D-Man and GDP-L-Gul, and against ADP-D- Glc, UDP-D-Gal, and UDP-D-Glc, we were surprised at the lack of dis- crimination between GDP-L-Gal FIGURE 4. Smirnoff-Wheeler pathway from D-glucose to L-ascorbic acid. The forward, on-pathway reactions and GDP-D-Glc. In the forward are depicted, with sugars in the D-configuration unless otherwise designated. The pathway begins with six steps used to produce GDP-Man and GDP-L-Gal for protein glycosylation and/or cell wall biosynthesis. In the direction, both substrates are con- seventh and committing step for L-ascorbic acid synthesis, the GDP-L-Gal product of GME is recognized by verted to hexose 1-phosphates with VTC2, the GDP-L-Gal phosphorylase, to produce L-Gal-1-P. In the eighth step, L-Gal is generated by VTC4, the 6 ϩ ϳ ϫ L-Gal-1-P phosphatase. L-Gal is then oxidized to L-galactono-1,4-lactone and L-ascorbic acid by the NAD -de- a specificity constant of 6 10 Ϫ1 Ϫ1 pendent and ferric cytochrome c-dependent enzymes, L-galactose dehydrogenase and L-galactono-1,4-lac- s M . In the reverse direction, tone dehydrogenase. while the enzyme is nearly inert at converting D-Man-1-P, the C2 first committed step in the production of ascorbic acid is cata- epimer of D-Glc-1-P, or the D-form of Gal-1-P, the enzyme has lyzed by VTC2. Thus, VTC2 activity may be regulated to coor- robust activity on D-Glc-1-P. In fact, the enzyme is three times dinate cellular needs for ascorbate with needs for cell wall syn- more efficient at forming GDP-D-Glc than GDP-L-Gal. There thesis and protein glycosylation. are two potential explanations for the lack of specificity. First, In the Leloir pathway of galactose utilization, cells evolved a although GDP-D-Glc is apparently formed in plants (33) where pathway to convert D-Gal to D-Glc equivalents by coupling a it can be used for the synthesis of cell wall polysaccharides (34), HIT family transferase (GalT) with an epimerase (UDP-D-ga- this metabolite may be spatially separated from VTC2, such lactose 4-epimerase) (27). Similar enzymology is used in the that no evolutionary selection against D-Glc-1-P formation has Smirnoff-Wheeler pathway, in which an epimerase (GME) is occurred. In this regard it will be important to determine coupled to a HIT family phosphorylase (VTC2) to convert whether plant GDP-Man pyrophosphorylase (VTC1), which is D-Man to L-Gal. In the former case, a specialty sugar is con- responsible for GDP-D-Man formation in the Smirnoff- verted to a commodity sugar. The Leloir pathway moves for- Wheeler pathway, discriminates against GDP-D-Glc formation ward by supplying D-Glc as UDP-D-Glc, using GalT to convert as reported for the mammalian ortholog of VTC1, GDP-Man D-Gal-1-P plus UDP-D-Glc to D-Glc-1-P plus UDP-D-Gal, and pyrophosphorylase B (35). Alternatively, if VTC1 does not dis- epimerizing UDP-D-Gal to UDP-D-Glc. In the Smirnoff- criminate against D-Glc-1-P and significant GDP-D-Glc is

18884 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 26•JUNE 29, 2007 19 GDP-L-Galactose Phosphorolysis by VTC2 in Ascorbate Synthesis

formed, the activity of VTC2 on this metabolite could serve to 12. Wolucka, B. A., Persiau, G., Van Doorsselaere, J., Davey, M. W., Demol, H., recycle D-Glc-1-P such that there is a second chance to make Vandekerckhove, J., Van Montagu, M., Zabeau, M., and Boerjan, W. the fructose and mannose phosphates required for ascorbate (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 14843–14848 13. Gatzek, S., Wheeler, G. L., and Smirnoff, N. (2002) Plant J. 30, 541–553 synthesis. 14. Laing, W. A., Bulley, S., Wright, M., Cooney, J., Jensen, D., Barraclough, D., Of the four Arabidopsis genes whose mutations result in vita- and MacRae, E. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 16976–16981 min C deficiency, the roles of three have now been established. 15. Conklin, P. L., Gatzek, S., Wheeler, G. L., Dowdle, J., Raymond, M. J., The function of VTC3 remains unclear. Vitamin C turnover is Rolinski, S., Isupov, M., Littlechild, J. A., and Smirnoff, N. (2006) J. Biol. apparently not affected in the vtc3 mutant (11) suggesting that Chem. 281, 15662–15670 VTC3 is not involved in ascorbate catabolism. Furthermore, the 16. Qian, W., Yu, C., Qin, H., Liu, X., Zhang, A., Johansen, I. E., and Wang, D. 14 14 (2007) Plant J. 49, 399–413 conversion of D-[U- C]Man to [ C]ascorbic acid has been 17. Conklin, P. L., Williams, E. H., and Last, R. L. (1996) Proc. Natl. Acad. Sci. reported to be decreased in the vtc3 mutant (11). Thus, VTC3 U. S. A. 93, 9970–9974 may encode a protein involved in the regulation of the 18. Conklin, P. L., Saracco, S. A., Norris, S. R., and Last, R. L. (2000) Genetics Smirnoff-Wheeler pathway or participating in an alternative 154, 847–856 ascorbate biosynthesis pathway. 19. Conklin, P. L., Norris, S. R., Wheeler, G. L., Williams, E. H., Smirnoff, N., Finally, what can we say about the role of the C10F3.4 gene and Last, R. L. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 4198–4203 product of C. elegans that initiated our interest in this project? 20. Jander, G., Norris, S. R., Rounsley, S. D., Bush, D. F., Levin, I. M., and Last, Downloaded from While we detect small amounts of vitamin C in extracts of these R. L. (2002) Plant Physiol. 129, 440–450 21. Major, L. L., Wolucka, B. A., and Naismith, J. H. (2005) J. Am. Chem. Soc. worms (data not shown), we do not find orthologs of other 127, 18309–18320 distinctive enzymes of the Smirnoff-Wheeler pathway. The 22. Yamada, K., Lim, J., Dale, J. M., Chen, H., Shinn, P., Palm, C. J., Southwick, presence of a VTC2 homolog in humans, in which a role in A. M., Wu, H. C., Kim, C., Nguyen, M., Pham, P., Cheuk, R., Karlin-

vitamin C synthesis is unlikely, suggests that a different func- Newmann, G., Liu, S. X., Lam, B., Sakano, H., Wu, T., Yu, G., Miranda, M., www.jbc.org tion is conserved in animals, potentially in glucose metabolism. Quach, H. L., Tripp, M., Chang, C. H., Lee, J. M., Toriumi, M., Chan, Ongoing work is designed to further dissect the function of M. M., Tang, C. C., Onodera, C. S., Deng, J. M., Akiyama, K., Ansari, Y., Arakawa, T., Banh, J., Banno, F., Bowser, L., Brooks, S., Carninci, P., Chao, Arabidopsis VTC2 and to determine the functions of the appar- Q., Choy, N., Enju, A., Goldsmith, A. D., Gurjal, M., Hansen, N. F., Ha- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 ent animal orthologs. yashizaki, Y., Johnson-Hopson, C., Hsuan, V. W., Iida, K., Karnes, M., Khan, S., Koesema, E., Ishida, J., Jiang, P. X., Jones, T., Kawai, J., Kamiya, A., Acknowledgments—We thank Prof. James H. Naismith for providing Meyers, C., Nakajima, M., Narusaka, M., Seki, M., Sakurai, T., Satou, M., the GME clone and the Arabidopsis Biological Resource Center at Tamse, R., Vaysberg, M., Wallender, E. K., Wong, C., Yamamura, Y., Yuan, Ohio State University for providing the VTC2 clone. We are indebted S., Shinozaki, K., Davis, R. W., Theologis, A., and Ecker, J. R. (2003) Science to Kristofor Webb for assistance with mass spectrometry. Finally, we 302, 842–846 are grateful to Prof. Emile Van Schaftingen for critical reading of the 23. Kagan, R. M., and Clarke, S. (1995) Biochemistry 34, 10794–10806 manuscript. 24. Altschul, S. F., Madden, T. L., Scha¨ffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D. J. (1997) Nucleic Acids Res. 25, 3389–3402 25. Brenner, C. (2002) Biochemistry 41, 9003–9014 REFERENCES 26. Wolucka, B. A., and Van Montagu, M. (2003) J. Biol. Chem. 278, 1. Padayatty, S. J., Katz, A., Wang, Y., Eck, P., Kwon, O., Lee, J. H., Chen, S., 47483–47490 Corpe, C., Dutta, A., Dutta, S. K., and Levine, M. (2003) J. Am. Coll. Nutr. 27. Frey, P. A. (1996) FASEB J. 10, 461–470 22, 18–35 28. Plateau, P., Fromant, M., Schmitter, J.-M., and Blanquet, S. (1990) J. Bac- 2. Englard, S., and Seifter, S. (1986) Annu. Rev. Nutr. 6, 365–406 teriol. 172, 6892–6899 3. Smirnoff, N. (2000) Curr. Opin. Plant Biol. 3, 229–235 29. Leang, K., Maekawa, K., Menavuvu, B. T., Morimoto, K., Granstro¨m, T. B., 4. De Tullio, M. C., and Arrigoni, O. (2004) Cell Mol. Life Sci. 61, 209–219 Takada, G., and Izumori, K. (2004) J. Biosci. Bioeng. 97, 383–388 5. Chatterjee, I. B. (1973) Science 182, 1271–1272 30. Roberts, R. M. (1971) Arch. Biochem. Biophys. 145, 685–692 6. Moreau, R., and Dabrowski, K. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 31. Baydoun, E. A. H., and Fry, S. C. (1988) J. Plant Physiol. 132, 484–490 10279–10282 32. Rayon, C., Cabanes-Macheteau, M., Loutelier-Bourhis, C., Salliot-Maire, 7. Linster, C. L., and Van Schaftingen, E. (2007) FEBS J. 274, 1–22 I., Lemoine, J., Reiter, W. D., Lerouge, P., and Faye, L. (1999) Plant Physiol. 8. Nishikimi, M., Fukuyama, R., Minoshima, S., Shimizu, N., and Yagi, K. 119, 725–733 (1994) J. Biol. Chem. 269, 13685–13688 33. Barber, G. A., and Hassid, W. Z. (1964) Biochim. Biophys. Acta 86, 9. Valpuesta, V., and Botella, M. A. (2004) Trends Plant Sci. 9, 573–577 397–399 10. Wheeler, G. L., Jones, M. A., and Smirnoff, N. (1998) Nature 393, 365–369 34. Piro, G., Zuppa, A., Dalessandro, G., and Northcote, D. H. (1993) Planta 11. Smirnoff, N., Conklin, P. L., and Loewus, F. A. (2001) Annu. Rev. Plant 190, 206–220 Physiol. Plant Mol. Biol. 52, 437–467 35. Ning, B., and Elbein, A. D. (2000) Eur. J. Biochem. 267, 6866–6874

JUNE 29, 2007•VOLUME 282•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 18885 20

CHAPTER 3

A Second GDP-L-Galactose Phosphorylase in Arabidopsis en Route to Vitamin C.

Covalent Intermediate and Substrate Requirements for the Conserved Reaction

A Second GDP-L-galactose Phosphorylase in Arabidopsis en Route to Vitamin C COVALENT INTERMEDIATE AND SUBSTRATE REQUIREMENTS FOR THE CONSERVED REACTION* Received for publication, April 3, 2008, and in revised form, May 1, 2008 Published, JBC Papers in Press, May 7, 2008, DOI 10.1074/jbc.M802594200 Carole L. Linster‡1,LitalN.Adler‡,KristoforWebb‡,KathrynC.Christensen§,CharlesBrenner§,andStevenG.Clarke‡2 From the ‡Department of Chemistry and Biochemistry and the Molecular Biology Institute, UCLA, Los Angeles, California 90095 and the §Departments of Genetics and Biochemistry, Dartmouth Medical School, Lebanon, New Hampshire 03756

The Arabidopsis thaliana VTC2 gene encodes an enzyme that tion against oxidative damage and is used as a cofactor for sev-

catalyzes the conversion of GDP-L-galactose to L-galactose eral enzymes. It is synthesized via reactions initially proposed Downloaded from 1-phosphate in the first committed step of the Smirnoff- by Wheeler et al. (1) that are distinct from those used in vitamin Wheeler pathway to plant vitamin C synthesis. Mutations in C biosynthesis in animals. Although the Smirnoff-Wheeler VTC2 had previously been found to lead to only partial vitamin pathway, arising from GDP-D-mannose and involving L-galac- C deficiency. Here we show that the Arabidopsis gene tose formation, might not be the only route to vitamin C bio- At5g55120 encodes an enzyme with high sequence identity to synthesis in plants (alternative pathways arising from myo-ino- VTC2. Designated VTC5, this enzyme displays substrate speci- sitol and methylgalacturonate or involving L-gulose formation www.jbc.org ficity and enzymatic properties that are remarkably similar to have been proposed; for reviews, see Refs. 2 and 3), it is those of VTC2, suggesting that it may be responsible for residual undoubtedly the pathway that has received the strongest bio- vitamin C synthesis in vtc2 mutants. The exact nature of the chemical and genetic support in recent years (4–9). at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 reaction catalyzed by VTC2/VTC5 is controversial because of Of the four loci (VTC1–4) that have been found to be reports that kiwifruit and Arabidopsis VTC2 utilize hexose mutated in vitamin C-deficient Arabidopsis thaliana plants 1-phosphates as phosphorolytic acceptor substrates. Using liq- (10, 11), three are now known to encode enzymes involved in uid chromatography-mass spectroscopy and a VTC2-H238N the Smirnoff-Wheeler pathway. VTC1 encodes GDP-mannose 3 mutant, we provide evidence that the reaction proceeds through pyrophosphorylase (12), whereasVTC4 encodes L-Gal-1-P a covalent guanylylated histidine residue within the histidine phosphatase (8). In 2007 the VTC2 gene product was identified triad motif. Moreover, we show that both the Arabidopsis VTC2 as the enzyme that produces L-Gal-1-P from GDP-L-galactose, and VTC5 enzymes catalyze simple phosphorolysis of the gua- which completed the characterization of the 10 enzymatic steps nylylated enzyme, forming GDP and L-galactose 1-phosphate leading from D-glucose to L-ascorbic acid (13, 14). VTC2 is a from GDP-L-galactose and phosphate, with poor reactivity of member of the D-galactose-1-phosphate uridylyltransferase/ hexose 1-phosphates as phosphorolytic acceptors. Indeed, the Apa1 nucleoside monophosphate transferase branch of the his- endogenous activities from Japanese mustard spinach, lemon, tidine triad (HIT) protein superfamily (15). The function of the and spinach have the same substrate requirements. These VTC3 gene remains unknown. Significantly, all vtc mutants results show that Arabidopsis VTC2 and VTC5 proteins and identified so far are only partially deficient in vitamin C (11). their homologs in other plants are enzymes that guanylylate a In the wake of the identification of VTC2 as the L-Gal-1-P- conserved active site His residue with GDP-L-galactose, forming forming enzyme, three issues remained unresolved. First, are L-galactose 1-phosphate for vitamin C synthesis, and regenerate vtc2-2 and vtc2-3 null mutations? If so, is there a VTC2-inde- the enzyme with phosphate to form GDP. pendent pathway for making vitamin C that might be responsible for the viability of the mutant plants? Second, consistent with the sequence relationship of VTC2 as a HIT protein, does Vitamin C (L-ascorbic acid) is the most abundant soluble this enzyme proceed through a guanylylated histidine interme- antioxidant in plants, in which it plays crucial roles in protec- diate? Third, does the guanylylated intermediate of VTC2 become resolved by simple phosphorolysis (i.e. transfer to * This work was supported, in whole or in part, by National Institutes of Health phosphate) as we reported (13) or by transfer to a hexose Grants GM026020 (to S. G. C.), AG018000 (to S. G. C.), and CA75954 (to 1-phosphate as reported by others (14)? C. B.). This work was also supported by National Science Foundation Grant Here we find that vtc2-2 and vtc2-3 mutant proteins are MCB-0448533 (to S. G. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must there- nearly devoid of enzymatic activity and that a VTC2 homolo- fore be hereby marked “advertisement” in accordance with 18 U.S.C. Sec- gous enzyme, termed VTC5 (16), has nearly identical enzy- tion 1734 solely to indicate this fact. matic characteristics. We find that VTC2 activity and forma- 1 Supported by a research fellowship (bourse de formation-recherche) from the Luxembourgish Government. tion of a covalent guanylylated intermediate both depend on 2 To whom correspondence should be addressed: Dept. of Chemistry and Biochemistry and the Molecular Biology Institute, UCLA, 607 Charles E. Young Dr. E., Los Angeles, CA 90095. Tel.: 310-825-8754; Fax: 310-825-1968; 3 The abbreviations used are: Gal-1-P, galactose 1-phosphate; HIT, histidine E-mail: [email protected]. triad; HPLC, high performance liquid chromatography.

JULY 4, 2008•VOLUME 283•NUMBER 27 JOURNAL OF BIOLOGICAL CHEMISTRY 18483

22 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

His-238. Finally, we clearly show that there is no requirement mid was transformed into Escherichia coli BL21 Star (DE3) for a hexose 1-phosphate substrate for VTC2, VTC5, or for the cells. The DNA sequences of both the VTC2 and VTC5 inserts endogenous L-Gal-1-P-forming activities from extracts of four were confirmed. different plant species. Preparation of VTC2 Mutant Proteins—VTC2 mutants were constructed by the PCR and DpnI method (19) using Pfu poly- EXPERIMENTAL PROCEDURES merase (Stratagene, La Jolla, CA) and DpnI (New England Bio- Materials—ADP-D-Glc, GDP-D-Glc, GDP-D-Man, UDP-D- labs, Ipswich, MA). The pET100/D-Topo plasmid containing Gal, UDP-D-Glc, D-Gal-1-P, D-Glc-1-P, D-Man-1-P (all of these the VTC2 coding sequence was used as the mutagenesis tem- sugars are in the ␣-configuration), GDP-␤-L-Fuc, and GDP plate. The sense strand mutagenic primers used for the amino were from Sigma. L-Gal-1-P, in the ␤-configuration, was pur- acid substitutions were 5Ј-CATACTTCAGACTCGATTA- chased from Glycoteam (Hamburg, Germany). All other CAACAGCTTGGG for G224D, 5Ј-CTATCAATCATCT- reagents were of analytical grade. GDP-␤-L-Gal, synthesized CAACTTTCAGGCTTATTA for H238N, and 5Ј-CTATG- and purified as described (17), was provided by Prof. Shinichi CAAGAACTATTTGATACTGTTTCAGA for S290F. The Kitamura (Osaka Prefecture University). This preparation was underlined nucleotides indicate the mutations, which were further purified by the reverse-phase HPLC method described confirmed by DNA sequencing. Plasmids containing the Downloaded from in Linster et al. (13). Fractions containing GDP-L-Gal were mutant cDNAs were used to transform E. coli BL21 Star (DE3) Ϫ lyophilized, resuspended in H2O, and stored at 20 °C. cells for purification of mutant VTC2 proteins, designated A. thaliana (ecotype Wassilewskija) plants used in this study VTC2-G224D, VTC2-H238N, and VTC2-S290F. had either been grown on soil or under sterile conditions on Overexpression and Purification of Recombinant VTC2, agar-solidified Murashige and Skoog medium supplemented VTC2 Mutants, and VTC5—Recombinant wild-type and

with sucrose. Seeds were sown on soil and germinated for 2 days mutant VTC2 enzymes as well as wild-type VTC5 protein were www.jbc.org at 4 °C before being transferred to the UCLA greenhouse where overexpressed and purified as described in Linster et al. (13) they were grown at 22 °C for 21 days with 16 h of light (100–200 with the exception that the lysis buffer contained 0.5 mM phen- Ϫ Ϫ ␮mol m 2 s 1). Plants were then harvested and stored at ylmethylsulfonyl fluoride and 5 ␮g/ml leupeptin. Protein con- Ϫ80 °C until extracted for protein. For growth on agar-solidi- centration was determined using a Lowry assay after precipita- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 fied medium, seeds were surface-sterilized (50% (v/v) sodium tion with trichloroacetic acid. Purified enzyme was stored at hypochlorite, 0.1% (w/v) SDS for 10 min with gentle agitation) Ϫ80 °C in 10% glycerol. and then washed 5 times in sterile water. The seeds, resus- Activity Assays of Recombinant VTC2, VTC2 Mutants, and pended in sterile water, were vernalized for 2 days at 4 °C before VTC5—The phosphorylase activity of recombinant VTC2, being sown on Murashige and Skoog medium (Sigma) supple- VTC2 mutants, and VTC5 was assayed as described in Linster mented with 1% (w/v) sucrose and 0.7% (w/v) agar in Magenta et al. (13). In short, GDP formation was measured by anion- G-7 boxes (Sigma). After 10 days of growth at 22 °C under con- exchange HPLC after incubation of recombinant enzyme with Ϫ Ϫ tinuous light (ϳ40 ␮mol m 2 s 1), plants were harvested, flash- various GDP-hexose substrates in a reaction mixture, pH 7.5, frozen in liquid nitrogen, and then stored at Ϫ80 °C until used containing 50 mM Tris-HCl, 5 mM (unless otherwise indicated)

for protein extraction. Similar GDP-D-Glc and GDP-L-Gal sodium phosphate, 2 mM MgCl2, 10 mM NaCl, and 1 mM dithi- phosphorylase activities were measured in extracts obtained othreitol. Reactions (26 °C) were initiated by the addition of from plants grown under the two conditions described. Japa- enzyme and stopped after the indicated times by heating at nese mustard spinach (Brassica rapa var. komatsuna), lemon 98 °C for 3 min. To assay the enzymatic activity in the reverse tree (Citrus limon), and maize (Zea mays) leaves were freshly direction, GDP-hexose formation was measured by the same harvested from plants grown in the outside garden of the UCLA HPLC method after incubation of the enzyme with the indi- greenhouse, whereas tobacco (Nicotiana rustica) leaves were cated hexose 1-phosphate and 5 mM GDP as described above

collected from greenhouse grown plants. Kiwifruit leaves were except that sodium phosphate and MgCl2 were omitted from harvested from a female plant purchased at a local nursery. the reaction mixture. Finally, fresh spinach leaves were purchased at a local super- The GDP-L-Gal-hexose-1-phosphate guanylyltransferase market. All the leaves were washed with deionized water and activities of recombinant VTC2 and VTC5 were measured by then stored at Ϫ80 °C until used for protein extraction. replacing the sodium phosphate used in the phosphorylase Cloning of A. thaliana VTC2 and VTC5—The A. thaliana assay by the indicated concentrations of D-Glc-1-P or D-Man- VTC2 cDNA was cloned as described in Linster et al. (13). A 1-P and by monitoring GDP-D-Glc or GDP-D-Man formation similar approach was taken to clone the A. thaliana VTC5 by the same anion-exchange HPLC method than the one used cDNA. In short, the VTC5 coding sequence was PCR-amplified for the phosphorylase assay. from the U11937 clone containing the coding sequence of the GDP and GDP-hexose concentrations were calculated by At5g55120 gene (prepared by the Salk Institute Genomic Anal- comparing the integrated peak areas with those of standard ysis Laboratory (18) and provided by the Arabidopsis Biological GDP or GDP-D-Man solutions. The GDP-L-Gal (substrate of Resource Center at Ohio State University) using the following the transferase activity) and GDP-D-Glc (product of the trans- forward and reverse primers: 5Ј-CACCATGTTGTTGAA- ferase activity) peaks partially overlapped. To estimate the total GATCAAAAGAGTTCC and 5Ј-TCAATTAGAGACAGC- peak area of the GDP-D-Glc (which elutes just before GDP-L- CTCTTCTTTCACTG. The resulting DNA was cloned into the Gal) formed, we generally split the peak at the summit and Champion pET100/D-Topo vector (Invitrogen), and this plas- multiplied the left half-peak area by 2. However, because of an

18484 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 27•JULY 4, 2008 23 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

asymmetry of the peaks toward the right, this can lead to an underestimation of the GDP-D-Glc concentrations and, thus, the GDP-L-Gal-D-Glc-1-P guanylyltransferase activities by as much as 35%. Preparation and Assay of Partially Purified Plant Extracts— A. thaliana extracts were prepared from whole plants. All other plant extracts used in this study were derived from leaves. Plant tissue was ground to a fine powder under liquid nitrogen and resuspended in 3 volumes of extraction buffer (100 mM Tris- HCl, pH 7.5, 2 mM dithiothreitol, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, and 5 ␮g/ml leupeptin) per gram of extract powder. For extracts prepared from kiwifruit leaves, the extraction buffer was added in a 4:1 (v/w) ratio, and it additionally contained 2% (w/v) polyvinylpolypyrrolidone (Sigma). For

all plant tissues, the resuspended powder was then centrifuged Downloaded from for 20 min (4 °C) at 20,000 ϫ g, and the resulting supernatant FIGURE 1. SDS-PAGE analysis of recombinant VTC5 (VTC2 homolog) and filtered through one layer of Miracloth. This preparation was wild-type and mutant VTC2 proteins. Purified preparations of recombinant either stored at Ϫ80 °C or immediately subjected to ammo- wild-type VTC5 (WT) (4.9 ␮g of protein), wild-type VTC2 (5.1 ␮g of protein), nium sulfate fractionation. Protein precipitating between 35 VTC2-H238N (8.7 ␮g of protein), VTC2-S290F (6.9 ␮g of protein), and VTC2- G224D (6.8 ␮g of protein) were analyzed by SDS-PAGE using Coomassie stain- and 50% saturating ammonium sulfate (ICN Biomedicals, Ultra ing. The expected molecular weights for recombinant wild-type VTC2 and Pure) was resuspended in 50 mM Tris-HCl, pH 7.5, 1 mM dithi- www.jbc.org VTC5 proteins are 53.1 and 52.4 kDa, respectively. An arrow indicates the ␮ expected position of a 53-kDa polypeptide from the markers in the left and othreitol, 10% glycerol, and 1 g/ml leupeptin. These fractions right lanes, that include bovine phosphorylase a, bovine serum albumin, hen were desalted on protein desalting spin columns in 10 mM Tris, ovalbumin, bovine carbonic anhydrase, soybean trypsin inhibitor, and hen

lysozyme (Bio-Rad). pH 7.5 (Pierce) immediately before being used for activity at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 measurements. GDP-D-Glc/GDP-L-Gal phosphorylase and GDP-L-Gal-hexose-1-phosphate guanylyltransferase activities TABLE 1 were assayed as described for recombinant VTC2 and VTC5 Phosphorylase activities of wild-type and mutant VTC2 enzymes proteins, except that NaCl and MgCl2 were omitted from the Activities were measured as described under “Experimental Procedures” in the pres- reaction mixtures. ence of 5 mM Pi. Wild-type VTC2, VTC2-H238N, VTC2-S290F, and VTC2-G224D were assayed at final concentrations of 6.8 ng/ml, 23.2 ␮g/ml, 18.4 ␮g/ml, and 18.0 Liquid Column Chromatography Coupled to Electrospray ␮ g/ml, and reactions were stopped 10 or 30 min after the addition of wild-type or Ionization Mass Spectrometry—Recombinant wild-type VTC2 mutant VTC2 enzymes, respectively. Values are given as the means Ϯ S.D. calculated from at least three independent measurements. (0.28 mg/ml) and VTC2-H238N (0.48 mg/ml) proteins were Phosphorylase activity incubated for 5 min (26 °C) with 21 mM Tris-HCl, pH 7.5, and in Substrate 25 ␮M GDP-L-Gal 50 ␮M GDP-D-Glc the absence or presence of 60 ␮M GDP-D-Glc and then kept on Ϫ Ϫ nmol min 1 mg protein 1 ice until liquid chromatography-mass spectroscopy analysis. VTC2 (wild type) 15,100 Ϯ 2,100 14,200 Ϯ 2,100 The samples (100 ␮l) were fractionated by reverse-phase HPLC VTC2-H238N 0.51 Ϯ 0.07 0.27 Ϯ 0.05 VTC2-S290F 0.50 Ϯ 0.12 0.56 Ϯ 0.09 using a PLRP-S polymeric column with a pore size of 300 Å, a VTC2-G224D 0.054 Ϯ 0.006 0.013 Ϯ 0.128 bead size of 5 ␮m, and 150 ϫ 2.1-mm dimensions (Polymer Laboratories, Amherst, MA). The column was equilibrated in TABLE 2 Comparison of the substrate specificities of A. thaliana VTC2 and VTC5

Km and Vmax values were obtained by fitting the experimental data to the Michaelis-Menten equation using the Km calculator of the BioMechanic.org program. Enzymatic turnover numbers were derived from the Vmax values by using a molecular mass of 53.1 and 52.4 kDa for recombinant His-tagged VTC2 and VTC5, respectively, with the assumption that the enzyme preparations were pure. Incubation times and enzyme concentrations were adjusted to obtain initial velocity data. Values are the means Ϯ S.D. calculated from 2–4 individual experiments for each substrate. Enzymatic activities were measured as described under “Experimental Procedures,” and substrate concen- ␮ ␮ trations ranged from 2.5 to 50 M (GDP-L-Gal), 2.5 to 100 M (GDP-D-Glc), 0.25 to 5 mM (GDP-D-Man), 0.1 to 5 mM (Pi), and 2.5 to 40 mM (hexose 1-phosphates).

Km kcat kcat/Km Substrate VTC2 VTC5 VTC2 VTC5 VTC2 VTC5 Ϫ1 Ϫ1 Ϫ1 mM s M s Forward reaction 6 6 GDP-L-Gal 0.0079 Ϯ 0.0011 0.0083 Ϯ 0.0032 27 Ϯ 513 Ϯ 2 3.4 Ϯ 1.1 ϫ 10 1.6 Ϯ 0.3 ϫ 10 a 6 5 GDP-D-Glc 0.0044 Ϯ 0.0016 0.012 Ϯ 0.002 23 Ϯ 3 9.9 Ϯ 0.5 5.7 Ϯ 2.3 ϫ 10 8.6 Ϯ 1.6 ϫ 10 a 2 1 GDP-D-Man 0.52 Ϯ 0.15 1.3 Ϯ 0.3 0.093 Ϯ 0.011 0.024 Ϯ 0.004 1.9 Ϯ 0.3 ϫ 10 1.9 Ϯ 0.2 ϫ 10 b Ϯ Ϯ Pi 2.4 0.1 1.0 0.2 c Ϯ Ϯ Pi 0.76 0.06 0.22 0.04 Reverse reaction a 2 2 L-Gal-1-P 45 Ϯ 7 16 Ϯ 1 6.0 Ϯ 0.4 3.8 Ϯ 0.5 1.3 Ϯ 0.1 ϫ 10 2.5 Ϯ 0.2 ϫ 10 a 2 2 D-Glc-1-P 29 Ϯ 4 10 Ϯ 1 12 Ϯ 0.5 5.8 Ϯ 0.1 4.1 Ϯ 0.5 ϫ 10 5.8 Ϯ 0.8 ϫ 10 a For these substrates the values given for the VTC2 enzyme were taken from Linster et al. (13). b Measured in the presence of GDP-L-Gal 25 ␮M. c Measured in the presence of GDP-D-Glc 50 ␮M.

JULY 4, 2008•VOLUME 283•NUMBER 27 JOURNAL OF BIOLOGICAL CHEMISTRY 18485 24 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

80% solvent A (0.1% trifluoroacetic acid in water) and 20% solvent B (0.1% trifluoroacetic acid in acetonitrile). After a 5-min wash at 20% B, proteins were eluted using a gradient changing from 20% to 60% B in 40 min, with a final increase to 100% B during an additional 5 min, all at a flow rate of 100 ␮l/min. An API IIIϩ mass spectrometer (PE Sciex) was tuned and calibrated as described previously (20) to yield a mass accuracy of 0.02%. RESULTS VTC2 Point Mutants Lack Enzy- Downloaded from matic Activity—Three mutations in the VTC2 gene (vtc2-1, vtc2-2, and vtc2-3) are known to lead to partial vitamin C deficiency in A. thaliana

plants (11). Whereas the vtc2-1 www.jbc.org mutation consists in a G to A base change at the predicted 3Ј splice site of intron 5, the vtc2-2 and vtc2-3 mutations have been found to cor- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 respond to mis-sense mutations leading to G224D and S290F substitutions, respectively (21). To test the effect of the latter mutations on enzyme activity, we have introduced them in the VTC2 coding sequence by site-directed mutagenesis. We have also prepared a mutant allele in which the predicted catalytic residue of VTC2, His-238 (see below), has been replaced by an Asn residue. Overexpression and purification yielded relatively high amounts of the recombinant VTC2-H238N protein (Fig. 1). Production of the VTC2-G224D and VTC2-S290F mutants proved more difficult, and only low amounts of protein could be detected (Fig. 1). SDS-PAGE analysis of the pellets and supernatants obtained after centrifugation of the bacterial lysates showed that the VTC2-S290F mutant was mostly insoluble and that the VTC2-G224D mutant appeared to have been entirely degraded (data not shown). With the VTC2-H238N and VTC2-S290F preparations, we detected only very low residual phosphorylase activities both in the presence of GDP-L-Gal and GDP-D- Glc (Table 1). These residual activi-

18486 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 27•JULY 4, 2008 25 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

ties were, however, significantly higher than those detected in We then examined the affinity for phosphate as a substrate.

the presence of the VTC2-G224D preparation and, thus, do not VTC5 showed a 2.5- and 3.5-fold higher affinity for Pi than seem to be contributed by bacterial contaminants. The phos- VTC2 in the presence of GDP-L-Gal and GDP-D-Glc, phorylase activities of the VTC2-H238N mutant were at least respectively. 30,000-fold lower than the wild-type activities. Because of the Finally, we compared the ability of VTC5 and VTC2 to carry low level of soluble expression of the VTC2-S290F mutant, we out the reverse reaction (hexose-1-P ϩ GDP 3 GDP-hexose ϩ

were unable to calculate its activity relative to the wild-type Pi). Both enzymes are much more efficient in phosphorolyzing enzyme. GDP-L-Gal or GDP-D-Glc than in catalyzing the corresponding VTC5 Is a VTC2-paralogous Enzyme—The vtc2-2 and vtc2-3 reverse reactions (Table 2). However, this difference in effi- mutants had previously been shown to contain ϳ30 and 50%, ciency for the forward and reverse reactions is more pro- respectively, of the ascorbate present in wild-type plants (11). nounced in the case of VTC2. The specificity constants for the The very low activities detected with the VTC2-G224D and forward reactions with GDP-L-Gal and GDP-D-Glc are VTC2-S290F mutants, thus, raised the question of the origin of ϳ26,000- and 14,000-fold, respectively, higher than the ones these residual vitamin C contents. BLAST searches revealed for the reverse reactions in the case of VTC2, whereas for VTC5 that the A. thaliana genome encodes a protein sharing high these ratios amount to ϳ6400- and 1500-fold, respectively. sequence identity with VTC2. We cloned and overexpressed Taken together, these results suggest that VTC2 and VTC5 Downloaded from the corresponding gene (At5g55120) and purified the expres- catalyze similar reactions with similar kinetics. sion product (Fig. 1, lane 2) using the same procedures as those VTC2 Forms a His-238-dependent Guanylylated Inter- used to produce His-tagged VTC2 (13). We found that mediate—The alignment of plant VTC2 sequences and their At5g55120 encodes a GDP-L-Gal phosphorylase whose bio- vertebrate and invertebrate homologs revealed the presence of

chemical properties closely resemble those of VTC2 (Table 2) a conserved motif (His-␾-His-␾-His/Gln, where ␾ is a hydro- www.jbc.org and which might, thus, account for the residual vitamin C levels phobic amino acid) characteristic of the members of the HIT found in vtc2 mutants. During the course of these studies Dow- protein superfamily (13, 14). HIT enzymes consist of nucleoside dle et al. (16) also overexpressed At5g55120 and characterized monophosphate hydrolases and nucleoside monophosphate the gene product, which they named VTC5, as a GDP-L-Gal transferases which attack the ␣-phosphate of the monophos- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 phosphorylase. phonucleoside moiety of their substrates by the second His of We first characterized the substrate specificity of VTC5 for the HIT motif, forming a covalent nucleotidylylated intermedi- nucleotide sugars. As shown in Table 2, both VTC2 and VTC5 ate (15). While the nucleotidylylated intermediate is simply use GDP-L-Gal and GDP-D-Glc with high catalytic efficiencies, hydrolyzed in the case of HIT hydrolases, this intermediate is ϳ although the values obtained for VTC5 are 2- and 7-fold, stable to water in HIT transferases and awaits reaction with Pi respectively, lower than those calculated for VTC2. GDP-D- (phosphorolysis) or a specific phosphorylated substrate (trans- Man is a very poor substrate for both enzymes. We could, how- fer). We know that A. thaliana VTC2 requires the presence of

ever, detect significant phosphorylase activities with both Pi to convert its GDP-hexose substrates (GDP-L-Gal and GDP- enzymes in the presence of GDP-L-Fuc, which is the 6-deoxy D-Glc) to GDP and the corresponding hexose 1-phosphates derivative of GDP-L-Gal. About 10-fold lower phosphorylase (L-Gal-1-P and D-Glc-1-P, respectively) (13). However, given activities were measured in the presence of ϳ80 ␮M GDP-L-Fuc the relatively low sequence similarity with other members of Ϫ Ϫ (2.4 Ϯ 0.1 ␮mol min 1 mg of protein 1 for VTC2 and 1.1 Ϯ 0.1 the HIT protein superfamily, it was important to experimen- Ϫ Ϫ ␮mol min 1 mg of protein 1 for VTC5) than in the presence of tally confirm the predicted catalytic mechanism. We, therefore, a near-saturating (ϳ50 ␮M) concentration of the physiological prepared a point mutant in which the second His of the HIT Ϫ1 Ϫ1 substrate GDP-L-Gal (25 Ϯ 5 ␮mol min mg of protein for motif of Arabidopsis VTC2 (His-238) is replaced by an Asn Ϫ Ϫ VTC2 and 12 Ϯ 1 ␮mol min 1 mg of protein 1 for VTC5). residue. We have shown above that this substitution reduces Similar to the situation with VTC2 (13), no phosphorylase the catalytic activity by at least 30,000-fold (Table 1). activity could be detected with VTC5 when UDP-D-Glc, UDP- We then incubated wild-type VTC2 and VTC2-H238N with-

D-Gal, or ADP-D-Glc was used as a substrate (all tested at 0.5 out Pi in the presence or absence of GDP-D-Glc and analyzed mM; data not shown). Using a mixture obtained by incubation the four resulting preparations by liquid chromatography-mass of GDP-D-Man with GDP-D-mannose 3Ј,5Ј-epimerase and spectroscopy. As shown in Fig. 2, a decrease of the mass peak containing GDP-D-Man, GDP-L-Gal, and GDP-L-gulose in a corresponding to the non-modified enzyme and the appear- 82:15:3 ratio, we could not detect any GDP-L-gulose phospho- ance of a new peak of 344 higher mass, were observed after rolysis in the presence of VTC5 and under conditions where reaction with GDP-D-Glc when wild-type VTC2, but not about 80% consumption of GDP-L-Gal was observed (data not when VTC2-H238N, was used in the incubation. Because a shown); a similar result was found previously with VTC2 (13). 345-Da mass increase is expected for covalently bound GMP,

FIGURE 2. Identification of a guanylylated enzyme intermediate formed during the catalytic cycle of VTC2 that is dependent upon His-238. Recombi- nant A. thaliana wild-type (WT) VTC2 and mutant VTC2-H238N were incubated with GDP-D-Glc as described under “Experimental Procedures.” The two reactions along with enzyme only samples were separated by reverse-phase HPLC and directed to a SCIEX electrospray ionization mass spectrometer as described under “Experimental Procedures.” The figure displays the deconvoluted mass spectrum over the range of 52,500–54,000 Da. After incubation of wild-type VTC2, but not VTC2-H238N, with GDP-D-Glc, a new species was observed that is 344 Da larger as compared with the enzyme only spectrum. The expected molecular weights for wild-type VTC2, guanylylated wild-type VTC2, and VTC2-H238N are 53,095, 53,440, and 53,072 Da, respectively. All masses are given as an average mass based on natural isotopic abundance. cps, counts/s.

JULY 4, 2008•VOLUME 283•NUMBER 27 JOURNAL OF BIOLOGICAL CHEMISTRY 18487 26 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

these observations suggest formation of an enzyme intermediate guanylylated on the His-238 residue and strongly support the catalytic mechanism expected for a member of the D-galactose-1-phosphate uridylyltransferase/Apa1 nucleoside monophosphate transferase branch of the HIT protein superfamily (15).

A. thaliana VTC2 and VTC5 Are Both Highly Specific for Pi as the Guanylyl Acceptor—In our previous study on VTC2 (13), the specificity of the enzyme for the guanylyl acceptor had not been investigated. To test whether VTC2 could catalyze the conversion of its GDP-hexose substrate in the presence of gua-

nylyl acceptors other than Pi, we incubated the enzyme (at two different concentrations) with GDP-L-Gal and in the absence or

presence of Pi, PPi, D-Glc-1-P, and D-Man-1-P. As shown in Fig. 3G, only Pi gave rise to a significant GDP-L-Gal consumption when recombinant A. thaliana VTC2 was used at a concentra- Downloaded from tion of 0.026 ␮g/ml. At a 10-fold higher enzyme concentration, total conversion of GDP-L-Gal to GDP was measured in the

presence of Pi (Fig. 3B), whereas only a very low conversion (ϳ4%) of GDP-L-Gal to GDP-D-Glc could be detected in the

presence of D-Glc-1-P (Fig. 3E). Even at these high enzyme con- www.jbc.org centrations, no formation of GDP-D-Man or GTP (two compounds readily detectable with the HPLC method used; data not shown) could be measured in the presence of D-Man-1-P (Fig. 3D) or PPi (Fig. 3C), respectively. Very similar results were at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 obtained when VTC5 was used instead of VTC2 in these experiments (data not shown).

We then measured the effect of Pi and D-Glc-1-P concentration on the GDP-L-Gal phosphorylase and GDP-L-Gal-D-Glc- 1-P guanylyltransferase activities, respectively, of VTC2 (Fig. 4) and VTC5 (Fig. 5). At all acceptor concentrations tested, the VTC2 and VTC5 activities were considerably lower in the pres-

ence of D-Glc-1-P than in the presence of Pi. More particularly, at concentrations of 5 mM Pi or D-Glc-1-P (which are close to physiological concentrations reported for Pi in plants (22, 23) but are at least 100-fold higher than those measured for D-Glc- 1-P (24)), VTC2 activity was ϳ33-fold higher in the presence of

Pi than in the presence of D-Glc-1-P. For reasons that are yet unclear, the addition of Pi or D-Glc-1-P concentrations higher than 5 or 20 mM, respectively, leads to unexpectedly high enzyme activities when recombinant VTC2 or VTC5 enzymes are used (data not shown). However, in the concentration range used in Figs. 4 and 5, Michaelis-Menten kinetics were observed,

and kinetic parameters could be estimated. For VTC2, Km val- ␮ Ϫ1 ues of 1.8 and 26 mM and Vmax values of 19 and 2.6 mol min Ϫ1 mg of protein were calculated in the presence of Pi and D-Glc- 1-P, respectively, indicating that this enzyme is more than 100-

various potential guanylyl acceptors as described under “Experimental Pro- cedures.” The deproteinized samples were analyzed by anion-exchange HPLC as described in Linster et al. (13) to monitor GDP-L-Gal consumption and the possible production of GDP, GTP, GDP-D-Man, and GDP-D-Glc. The traces obtained after incubation with high VTC2 concentration (0.26 ␮g/ml) are shown (panels A–E). They are representative of three separate experiments. Panel F shows the separation of authentic GMP, GDP-D-Glc, GDP-L-Gal, and GDP by the HPLC method used. GDP-D-Man co-elutes with GDP-D-Glc, and GTP elutes at ϳ40 min (not shown). In panel G, the concentrations of GDP-L- FIGURE 3. Recombinant VTC2 is highly specific for Pi as the guanylyl Gal measured after incubation in the described conditions are given as acceptor used during conversion of GDP-L-Gal to L-Gal-1-P. VTC2 (0.026 or means Ϯ S.D. values calculated from three separate experiments. mAU, milli- 0.26 ␮g/ml) was incubated for 30 min in the presence of 25 ␮M GDP-L-Gal and absorbance units.

18488 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 27•JULY 4, 2008 27 Phosphorylase Activity of Arabidopsis VTC2 and VTC5 Downloaded from www.jbc.org at UCLA-Louise Darling Biomed. Lib., on March 7, 2012

FIGURE 4. Effect of Pi, D-Glc-1-P, and D-Man-1-P concentrations on the FIGURE 5. Effect of Pi, D-Glc-1-P, and D-Man-1-P concentrations on the activity of recombinant VTC2. GDP-L-Gal phosphorylase and GDP-L-Gal- activity of recombinant VTC5. GDP-L-Gal phosphorylase and GDP-L-Gal- hexose-1-phosphate guanylyltransferase activities were assayed as hexose-1-phosphate guanylyltransferase activities of VTC5 were assayed described under “Experimental Procedures” in the presence of 25 ␮M GDP-L- under the same conditions as those of VTC2 (see Fig. 4). The curves shown Gal and the indicated concentrations of Pi, D-Glc-1-P, or D-Man-1-P. Reactions represent ideal saturation curves obtained by fitting the activities measured in were stopped 10 min after the addition of recombinant VTC2. The curves the presence of Pi and D-Glc-1-P to the Michaelis-Menten equation. The axis shown represent ideal saturation curves obtained by fitting the activities scales in panels A and B have been drawn to highlight the high and low activities measured in the presence of Pi and D-Glc-1-P to the Michaelis-Menten equa- found. Activities are the means Ϯ S.D. values calculated from two (activity with tion. The axis scales in panels A and B have been drawn to highlight the high D-Man-1-P) or three (activities with Pi or D-Glc-1-P) separate experiments. and low activities found. Activities are the means Ϯ S.D. values calculated from at least three separate experiments. maize leaves but not of kiwifruit or tobacco leaves (Table 3). fold more efficient as a GDP-L-Gal phosphorylase than as a Except for the maize leaf extract, slightly higher phosphorylase GDP-L-Gal-D-Glc-1-P guanylyltransferase. In the case of activities were found when GDP-L-Gal (25 ␮M) was used ␮ VTC5, we found Km values of 0.96 and 6.8 mM and Vmax values instead of GDP-D-Glc (50 M). ␮ Ϫ1 Ϫ1 of 15 and 2.4 mol min mg of protein in the presence of Pi However, we found that the levels of phosphorylase activity and D-Glc-1-P, respectively, indicating that this enzyme is ranged widely among different plants (Table 3). The highest about 45-fold more efficient as a phosphorylase than as a trans- activities were measured in ammonium sulfate fractions of Jap- ferase. No GDP-L-Gal-D-Man-1-P guanylyltransferase activity anese mustard spinach leaves, lemon leaves, and A. thaliana could be detected with either VTC2 or VTC5 and up to 40 mM whole plants; much lower activities were found in similar D-Man-1-P (Figs. 4 and 5). extracts of spinach and maize leaves.

Confirmation of Pi as the Preferred Guanylyl Acceptor for In these same plant extracts we also measured transferase GDP-L-Gal to L-Gal-1-P Conversion Measured in Plant Tissue activity as D-Glc-1-P- and D-Man-1-P-dependent GDP-D-Glc Extracts—We next wanted to test whether our observations and GDP-D-Man formation, respectively, in the presence of concerning the acceptor specificity of recombinant VTC2 and GDP-L-Gal (Table 3). The guanylyltransferase activities found VTC5 could be confirmed with partially purified plant extracts in the various plant extracts in the presence of D-Glc-1-P were as the enzyme source. Using ammonium sulfate fractions of generally at least 10-fold lower than the GDP-L-Gal phospho- A. thaliana whole plant extracts, we measured Pi-dependent rylase activities (except for the maize leaf extract, where the GDP formation which correlated with Pi-dependent GDP-D- transferase activity was only about 2-fold lower than the phos- Glc or GDP-L-Gal consumption, reflecting the sum of the phorylase activity). No guanylyltransferase activity could be VTC2 and VTC5 activities of these preparations (Table 3). We detected in the presence of D-Man-1-P and GDP-L-Gal, except also found that GDP-D-Glc and GDP-L-Gal phosphorylase for the lemon leaf extract, where this activity was, however, activities could be detected in ammonium sulfate precipitates more than 100-fold lower than the corresponding GDP-L-Gal of extracts of Japanese mustard spinach, lemon, spinach, and phosphorylase activity. Finally, in tobacco or kiwifruit leaf

JULY 4, 2008•VOLUME 283•NUMBER 27 JOURNAL OF BIOLOGICAL CHEMISTRY 18489 28 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

TABLE 3 Phosphorylase and transferase activities in various partially purified plant extracts Activities were assayed as described under “Experimental Procedures.” GDP-L-Gal and GDP-D-Glc were added at concentrations of 25 and 50 ␮M, respectively. Protein concentrations and incubation times ranged from 0.04 to 1.6 mg/ml and 15 to 120 min, respectively, according to the plant extract used. Plant extracts, prepared as ammonium sulfate fractions as described under “Experimental Procedures,” were incubated in the absence or presence of 5 mM guanylyl acceptor (Pi, D-Glc-1-P, or D-Man-1-P) to measure Pi-dependent GDP (phosphorylase activity) or hexose 1-phosphate-dependent GDP-D-Glc or GDP-D-Man (transferase activities) formations, which were then used to calculate specific activities. Data are the means Ϯ S.D. values calculated from three separate measurements. Activity (GMP donor/GMP acceptor) Extracta GDP-D-Glc/Pi GDP-L-Gal/Pi GDP-L-Gal/D-Glc-1-P GDP-L-Gal/D-Man-1-P Ϫ Ϫ pmol min 1 mg of protein 1 Arabidopsis 424 Ϯ 225 698 Ϯ 194 74 Ϯ 49 Ϫ1.8 Ϯ 2.8 Japanese mustard spinach 2786 Ϯ 267 3491 Ϯ 122 284 Ϯ 49 2.0 Ϯ 3.1 Lemon 1312 Ϯ 329 3626 Ϯ 789 351 Ϯ 45 27 Ϯ 9 Spinach 45 Ϯ 8 88 Ϯ 19 7.1 Ϯ 2.8 Ϫ0.45 Ϯ 0.74 Maize 52 Ϯ 937 Ϯ 916 Ϯ 10 3.5 Ϯ 2.9 Kiwifruit 6.6 Ϯ 2.0 12 Ϯ 16 Ϫ17 Ϯ 25 Ϫ4.0 Ϯ 17 Tobacco Ϫ8.4 Ϯ 4.7 Ϫ1.8 Ϯ 5.1 3.8 Ϯ 0.8 0.26 Ϯ 1.0 a Except for Arabidopsis, all plant extracts were prepared from leaf tissue. Whole plants were used for Arabidopsis protein extraction.

Ϫ Downloaded from values of 4.8 (Ϯ0.1, n ϭ 2) and 0.99 (Ϯ0.11, n ϭ 3) nmol min 1 Ϫ1 mg of protein were estimated in the presence of Pi and D-Glc- 1-P, respectively. Similar Km values were found for Pi and D-Glc-1-P in A. thaliana extracts (1.1 Ϯ 0.3 and 18 Ϯ 2mM, n ϭ Ϯ 4, respectively), whereas the corresponding Vmax values (1.3 Ϯ Ϫ1 Ϫ1 ϭ 0.1 and 0.28 0.03 nmol min mg of protein , n 4, respec- www.jbc.org tively) were ϳ4 times lower in these extracts than in the Japanese mustard spinach extracts. It can, thus, be calculated

that the catalytic efficiencies (Vmax/Km) of the GDP-L-Gal phosphorylase activity of these extracts are about 45-fold (Jap- at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 anese mustard spinach leaf extract) and 75-fold (A. thaliana extract) higher than the catalytic efficiencies of the GDP-L-Gal- D-Glc-1-P guanylyltransferase activities in these same extracts. No GDP-L-Gal-D-Man-1-P guanylyltransferase activity could be detected in either of these extracts in the presence of D-Man- 1-P concentrations of up to 40 mM (Fig. 6). These results show that in both recombinant enzyme and in tissue extracts, the predominantly catalyzed reaction is phosphorolysis. DISCUSSION Of the three mutations in the Arabidopsis VTC2 gene (vtc2-1, vtc2-2, and vtc2-3) known to lead to vitamin C deficiency (11), two have been identified as mis-sense mutations leading to G224D (vtc2-2) and S290F (vtc2-3) substitutions (21). At the time these mutations were discovered the VTC2 function was still unknown. With the recent identification of FIGURE 6. Effect of P , D-Glc-1-P, and D-Man-1-P concentrations on the i L GDP-L-Gal phosphorylase or GDP-L-Gal-hexose-1-phosphate transfer- VTC2 as a GDP- -Gal phosphorylase (13), it became possible to ase activities in partially purified extracts of A. thaliana (panel A) and test the effect of the mis-sense mutations on this activity to Japanese mustard spinach leaves (panel B). The phosphorylase and trans- increase our understanding of the enzymology of VTC2 and ferase activities shown were assayed as described under “Experimental Pro- cedures” in the presence of 25 ␮M GDP-L-Gal and the indicated concentra- to confirm the physiological role of VTC2 as catalyzing a reactions of Pi and D-Glc-1-P or D-Man-1-P, respectively. Incubation times and tion involved in plant vitamin C biosynthesis. The S290F sub- protein concentrations were adjusted to obtain initial velocity data. The curves shown represent ideal saturation curves obtained by fitting the activ- stitution led to an enzymatic activity of the purified preparation ϳ ities measured in the presence of Pi and D-Glc-1-P to the Michaelis-Menten that was 30,000-fold lower than that of the wild-type VTC2 equation. The results shown are means Ϯ S.D. values calculated from at least preparation. However, the purity of the mutant protein being three (panel A) or two (panel B) separate measurements. much lower, we could not compare the specific activities of the wild-type VTC2 and VTC2-S290F enzymes. We were unable to extracts, transferase activities were close to or below the detec- stably express the VTC2-G224D protein; from our results it is, tion limit (Table 3). thus, unclear whether this enzyme is stable in cells or indeed Two of the partially purified plant extracts were used to esti- has any activity. During the course of this work, Dowdle et al. mate and compare the kinetic constants of their GDP-L-Gal (16) also characterized the VTC2-S290F and VTC2-G224D phosphorylase and GDP-L-Gal-D-Glc-1-P guanylyltransferase mutants. The authors found no activity with the latter mutant, activities (Fig. 6). In the Japanese mustard spinach leaf extracts, although the issue of its stability was not addressed. They were, Ϯ ϭ Ϯ ϭ Km values of 1.4 ( 0.2, n 2) and 13 ( 1, n 3) mM and Vmax however, able to show that the VTC2-S290F mutant displayed a

18490 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 27•JULY 4, 2008 29 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

10-fold higher Km for Pi and a 5-fold lower kcat than the wild- that VTC2 and VTC5 are both expressed in Arabidopsis leaf, type enzyme, resulting in a 50-fold lower catalytic efficiency for stem, root, flower, and silique tissue but that the mRNA expres- the mutant enzyme. Taken together, these results indicate that sion level of VTC5 is 100–1000-fold lower than the one of the relatively high residual vitamin C levels measured in vtc2-2 VTC2 in all these tissues. Exposure of Arabidopsis plants to (ϳ30% that of wild-type content) and vtc2-3 (ϳ50% of wild- high light led to increased ascorbate contents as well as type content) mutants (11, 16) are unlikely to be accounted for increased expression of VTC5 and, more importantly, VTC2 by the residual VTC2 activities. Although the existence of alter- (16). Jasmonate treatment also led to ascorbate accumulation in native pathways for vitamin C synthesis has been proposed in Arabidopsis as well as induction of both the VTC2 and VTC5 plants, the finding that the A. thaliana genome contains a gene genes (26). The expression of the VTC5 gene seemed, however, (At5g55120) sharing high similarity with the VTC2 gene indi- to be more responsive to jasmonate treatment as well as to cated the existence of an enzyme able to maintain flux through induction by ozone exposure (26), which might indicate that the Smirnoff-Wheeler pathway in the absence of VTC2. VTC5 only contributes significantly to ascorbate synthesis Accordingly, we found that the product of the At5g55120 under certain stress conditions. Importantly, with the identifi- gene is also highly efficient as a GDP-L-Gal/GDP-D-Glc phos- cation of VTC5, it has now become possible to find out whether phorylase. This partially confirms the results obtained during pathways other than the Smirnoff-Wheeler pathway signifi- the course of our studies by Dowdle et al. (16), who designated cantly contribute to plant vitamin C synthesis. During the Downloaded from this enzyme VTC5. However, the kinetic properties of the course of this study, Dowdle et al. (16) found that double GDP-L-Gal phosphorylase activities of recombinant VTC2 and mutants in VTC2 and VTC5 are unable to grow unless sup- VTC5 published by Dowdle et al. (16) differ quite substantially plemented with ascorbate, demonstrating not only that the from the ones found in this study. Compared with our results, Smirnoff-Wheeler pathway is the only physiologically signifi-

Dowdle et al. (16) found higher Km values for GDP-L-Gal (250 cant source of vitamin C, at least in A. thaliana, but also that www.jbc.org versus 7.9 ␮M for VTC2 and 667 versus 8.3 ␮M for VTC5), lower ascorbate is required for seedling viability. Ϫ1 Ϫ1 kcat values (2.0 versus 27 s for VTC2 and 2.7 versus 13 s for Members of the D-galactose-1-phosphate uridylyltrans- VTC5), and lower Km values for Pi (0.25 versus 2.4 mM for VTC2 ferase/Apa1 branch of the HIT protein superfamily transfer the at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 and 0.13 versus 1.0 mM for VTC5). The use of different expres- monophosphonucleoside moiety of their substrate either to Pi sion plasmids and/or different enzyme activity assays (direct or to a specific phosphorylated compound. D-galactose-1-phos- HPLC assay in this study and coupled assay in Dowdle et al. phate uridylyltransferase, for example, transfers UMP from (16)) might account for the discrepancy between the results UDP-D-Glc to D-Gal-1-P forming D-Glc-1-P and UDP-D-Gal

obtained. The fact that the Km values for GDP-L-Gal and Pi (27). Using a recombinant VTC2 homolog from kiwifruit, Laing obtained with our recombinant VTC2 and VTC5 preparations et al. (14) detected GDP-L-Gal transferase activities in the pres-

are closer to the values obtained with native pea enzyme (Km ence of Pi, PPi, and a series of hexose 1-phosphates including values of 18 ␮M and 1.1 mM for GDP-L-Gal and Pi, respectively D-Man-1-P and D-Glc-1-P. The highest activities were meas-

(16)) as well as with partially purified A. thaliana extracts (Km ured in the presence of the hexose 1-phosphates, and they pro- value of 1.1 mM for Pi; this study) suggests that the values found posed D-Man-1-P as the most likely in vivo guanylyl acceptor. in this study may reflect more closely the properties of the They reported that recombinant A. thaliana VTC2 also native enzymes. showed transferase activity with similar properties to the kiwi- Comparison of the kinetic properties of VTC2 and VTC5 did fruit enzyme (14). These results contrast with the characteriza- not reveal any fundamental differences between these two tion of VTC2 as a GDP-L-Gal phosphorylase by our group (13) enzymes. As for VTC2 (Ref. 13 and this study), we found that and by Dowdle et al. (16). Additionally, in this study we could VTC5 displays a strong substrate preference for GDP-L-Gal and not detect any formation of GDP-D-Man or GTP from GDP-L-

GDP-D-Glc over GDP-L-Fuc and GDP-D-Man and that UDP-D- Gal in the presence of VTC2 or VTC5 and D-Man-1-P or PPi, Glc, UDP-D-Gal, and ADP-D-Glc are not substrates at all. Fur- respectively. We measured a small transferase activity with thermore, as for VTC2, VTC5 does not seem to phosphorolyze D-Glc-1-P, but this activity was 100- and 45-fold less than the GDP-L-gulose, a compound formed, in addition to GDP-L-Gal, phosphorylase activity of VTC2 and VTC5, respectively. from GDP-D-Man by GDP-D-mannose 3Ј,5Ј-epimerase (25) in In plants, cytosolic concentrations have been estimated at ϳ ␮ the reaction catalyzed upstream of GDP-L-Gal phosphorolysis 3–7 mM for Pi (22, 23) and at 50 M for D-Glc-1-P (24). Given in the Smirnoff-Wheeler pathway. Neither VTC2 nor VTC5 the Km values of 0.96 to 1.8 mM found here for Pi and 6.8 to 26 does, thus, seem to participate in the putative vitamin C syn- mM for D-Glc-1-P, it seems clear that the phosphorylase reac- thesis pathway involving L-gulose and L-gulono-1,4-lactone tion will be responsible for almost all of the VTC2 and VTC5 formation, one of the alternative pathways that has been activities in plant cells. The striking preference of the VTC2 and proposed for vitamin C synthesis in plants (25). Both VTC2 VTC5 phosphorylase activities for GDP-D-Glc over GDP-D- and VTC5 catalyze the phosphorolysis of GDP-L-Gal and Man as the GDP-hexose donor indicates that these enzymes GDP-D-Glc much more efficiently than the corresponding have binding sites that can accommodate D-Glc-1-P, but not, or reverse reactions. much less efficiently, D-Man-1-P, which would explain the lack The reason for the conservation of two proteins with the of GDP-L-Gal-D-Man-1-P guanylyltransferase activity that we same biochemical function in Arabidopsis is not clear. VTC2 observe. and VTC5 might act in different subcellular compartments, but It is difficult to rationalize the differences between the stud- this has not yet been investigated. Dowdle et al. (16) showed ies supporting a transferase (14) rather than a phosphorylase

JULY 4, 2008•VOLUME 283•NUMBER 27 JOURNAL OF BIOLOGICAL CHEMISTRY 18491 30 Phosphorylase Activity of Arabidopsis VTC2 and VTC5

(13, 16) activity for VTC2/VTC5 (28). We find no evidence for 5. Wolucka, B. A., Persiau, G., Van Doorsselaere, J., Davey, M. W., Demol, H., differences in the acceptor specificity of the enzymatic reaction Vandekerckhove, J., Van Montagu, M., Zabeau, M., and Boerjan, W. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 14843–14848 in different plant species. GDP-L-Gal phosphorylase activities 6. Gatzek, S., Wheeler, G. L., and Smirnoff, N. (2002) Plant J. 30, 541–553 could be readily detected in partially purified extracts prepared 7. Laing, W. A., Bulley, S., Wright, M., Cooney, J., Jensen, D., Barraclough, D., from A. thaliana plants as well as Japanese mustard spinach and MacRae, E. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 16976–16981 and lemon leaves. These extracts also displayed some GDP-L- 8. Conklin, P. L., Gatzek, S., Wheeler, G. L., Dowdle, J., Raymond, M. J., Gal-D-Glc-1-P guanylyltransferase activity, which was, how- Rolinski, S., Isupov, M., Littlechild, J. A., and Smirnoff, N. (2006) J. Biol. ever, ϳ10-fold lower than the corresponding phosphorylase Chem. 281, 15662–15670 9. Qian, W., Yu, C., Qin, H., Liu, X., Zhang, A., Johansen, I. E., and Wang, D. activities. It is possible that the different assay methods may (2007) Plant J. 49, 399–413 have contributed to the divergent results. In this study, GDP 10. Conklin, P. L., Williams, E. H., and Last, R. L. (1996) Proc. Natl. Acad. Sci. and GDP-hexoses formed by the VTC2 and VTC5 phospho- U. S. A. 93, 9970–9974 rylase and transferase activities, respectively, were detected 11. Conklin, P. L., Saracco, S. A., Norris, S. R., and Last, R. L. (2000) Genetics directly by HPLC. Laing et al. (14) employed an indirect enzy- 154, 847–856 12. Conklin, P. L., Norris, S. R., Wheeler, G. L., Williams, E. H., Smirnoff, N., matic assay to measure release of the L-Gal-1-P product and Last, R. L. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 4198–4203 (formed in both the phosphorylase and transferase reactions).

13. Linster, C. L., Gomez, T. A., Christensen, K. C., Adler, L. N., Young, B. D., Downloaded from This enzymatic assay was based on the use of two E. coli-ex- Brenner, C., and Clarke, S. G. (2007) J. Biol. Chem. 282, 18879–18885 pressed coupling enzymes (L-Gal-1-P phosphatase and L-Gal 14. Laing, W. A., Wright, M. A., Cooney, J., and Bulley, S. M. (2007) Proc. Natl. dehydrogenase) and the spectrophotometric measurement of Acad. Sci. U. S. A. 104, 9534–9539 NADH formation. Although these enzymes have been claimed 15. Brenner, C. (2002) Biochemistry 41, 9003–9014 to be specific for their substrates (6, 7, 14), low activities found 16. Dowdle, J., Ishikawa, T., Gatzek, S., Rolinski, S., and Smirnoff, N. (2007) Plant J. 52, 673–689

L www.jbc.org with -Gal-1-P phosphatase and some other hexose 1-phos- 17. Watanabe, K., Suzuki, K., and Kitamura, S. (2006) Phytochemistry 67, phates (myo-inositol 1-phosphate, D-Gal-1-P, D-Man-1-P, 338–346 D-Glc-1-P (7)) as well as a possible inhibition of L-Gal-1-P phos- 18. Yamada, K., Lim, J., Dale, J. M., Chen, H., Shinn, P., Palm, C. J., Southwick,

phatase by Pi may complicate the coupled assay. A. M., Wu, H. C., Kim, C., Nguyen, M., Pham, P., Cheuk, R., Karlin- Based on the GDP-L-Gal-D-Man-1-P guanylyltransferase Newmann, G., Liu, S. X., Lam, B., Sakano, H., Wu, T., Yu, G., Miranda, M., at UCLA-Louise Darling Biomed. Lib., on March 7, 2012 Quach, H. L., Tripp, M., Chang, C. H., Lee, J. M., Toriumi, M., Chan, activity they measure, Laing et al. (14) proposed a VTC2 cycle in M. M., Tang, C. C., Onodera, C. S., Deng, J. M., Akiyama, K., Ansari, Y., which the biosynthesis of L-Gal-1-P from D-Man-1-P can be Arakawa, T., Banh, J., Banno, F., Bowser, L., Brooks, S., Carninci, P., Chao, sustained by the action of only two enzymes: VTC2 and GDP- Q., Choy, N., Enju, A., Goldsmith, A. D., Gurjal, M., Hansen, N. F., Ha- D-mannose 3Ј,5Ј-epimerase. In a subsequent review, Wolucka yashizaki, Y., Johnson-Hopson, C., Hsuan, V. W., Iida, K., Karnes, M., and Van Montagu (28) extended this initial proposal by includ- Khan, S., Koesema, E., Ishida, J., Jiang, P. X., Jones, T., Kawai, J., Kamiya, A., ing a putative GDP-D-mannose 2Ј-epimerase and by taking into Meyers, C., Nakajima, M., Narusaka, M., Seki, M., Sakurai, T., Satou, M., Tamse, R., Vaysberg, M., Wallender, E. K., Wong, C., Yamamura, Y., Yuan, L account the double specificity of VTC2 for GDP- -Gal and S., Shinozaki, K., Davis, R. W., Theologis, A., and Ecker, J. R. (2003) Science GDP-D-Glc, leading to a cycle that links photosynthesis with 302, 842–846 the biosynthesis of vitamin C and the cell-wall metabolism. 19. Weiner, M. P., Costa, G. L., Schoettlin, W., Cline, J., Mathur, E., and Bauer, Considering the very low transferase activity of VTC2 and J. C. (1994) Gene (Amst.) 151, 119–123 VTC5 that we measured in this study, it is unclear whether a 20. Whitelegge, J. P. (2004) Methods Mol. Biol. 251, 323–340 VTC2 cycle actually operates in plants. 21. Jander, G., Norris, S. R., Rounsley, S. D., Bush, D. F., Levin, I. M., and Last, R. L. (2002) Plant Physiol. 129, 440–450 22. Rebeille, F., Bligny, R., and Douce, R. (1984) Plant Physiol. 74, 355–359 Acknowledgments—We thank Prof. Shinichi Kitamura (Osaka Pre- 23. Lee, R. B., Ratcliffe, R. G., and Southon, T. E. (1990) J. Exp. Bot. 41, fecture University) for kindly providing GDP-L-galactose and the Ara- 1063–1078 bidopsis Biological Resource Center at Ohio State University for pro- 24. Farre´, E. M., Tiessen, A., Roessner, U., Geigenberger, P., Trethewey, R. N., viding the VTC2 and VTC5 clones. and Willmitzer, L. (2001) Plant Physiol. 127, 685–700 25. Wolucka, B. A., and Van Montagu, M. (2003) J. Biol. Chem. 278, 47483–47490 REFERENCES 26. Sasaki-Sekimoto, Y., Taki, N., Obayashi, T., Aono, M., Matsumoto, F., 1. Wheeler, G. L., Jones, M. A., and Smirnoff, N. (1998) Nature 393, 365–369 Sakurai, N., Suzuki, H., Hirai, M. Y., Noji, M., Saito, K., Masuda, T., Taka- 2. Valpuesta, V., and Botella, M. A. (2004) Trends Plant Sci. 9, 573–577 miya, K., Shibata, D., and Ohta, H. (2005) Plant J. 44, 653–668 3. Hancock, R. D., and Viola, R. (2005) Crit. Rev. Plant Sci. 24, 167–188 27. Frey, P. A. (1996) FASEB J. 10, 461–470 4. Smirnoff, N., Conklin, P. L., and Loewus, F. A. (2001) Annu. Rev. Plant 28. Wolucka, B. A., and Van Montagu, M. (2007) Phytochemistry 68, Physiol. Plant Mol. Biol. 52, 437–467 2602–2613

18492 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 27•JULY 4, 2008 31

CHAPTER 4

Impact of Oxidative Stress on Ascorbate Biosynthesis in Chlamydomonas

via Regulation of the VTC2 Gene Encoding a GDP-L-Galactose Phosphorylase

Impact of Oxidative Stress on Ascorbate Biosynthesis in Chlamydomonas via Regulation of the VTC2 Gene Encoding a GDP-L-galactose Phosphorylase*□S Received for publication, January 11, 2012, and in revised form, February 23, 2012 Published, JBC Papers in Press, March 5, 2012, DOI 10.1074/jbc.M112.341982 Eugen I. Urzica‡, Lital N. Adler‡, M. Dudley Page‡, Carole L. Linster‡§, Mark A. Arbing¶, David Caseroʈ, Matteo Pellegrini¶ʈ, Sabeeha S. Merchant‡¶**1, and Steven G. Clarke‡**2 From the Departments of ‡Chemistry and Biochemistry, ʈMolecular, Cell, and Developmental Biology, and ¶Institute of Genomics and Proteomics, **Molecular Biology Institute, UCLA, Los Angeles, California 90095 and the §de Duve Institute, Université Catholique de Louvain, BCHM 7539, Ave. Hippocrate 75, B-1200 Brussels, Belgium

Background: Ascorbate biosynthesis in plants occurs mainly via the L-galactose pathway. Results: Chlamydomonas reinhardtii VTC2 encodes a GDP-L-galactose phosphorylase whose transcript levels are induced in response to oxidative stress concurrent with increased ascorbate accumulation. Conclusion: Increased oxidative stress in C. reinhardtii results in an enzymatic and non-enzymatic antioxidant response. Significance: First characterization of C. reinhardtii ascorbate biosynthesis and recycling pathways.

The L-galactose (Smirnoff-Wheeler) pathway represents the L-Ascorbic acid plays an essential role in plants by protecting major route to L-ascorbic acid (vitamin C) biosynthesis in higher cells against oxidative damage. In addition to its antioxidant plants. Arabidopsis thaliana VTC2 and its paralogue VTC5 role, L-ascorbic acid is also an important enzyme cofactor, for function as GDP-L-galactose phosphorylases converting GDP-L- example, in violaxanthin de-epoxidase, required for dissipation galactose to L-galactose-1-P, thus catalyzing the first committed of excess excitation energy, and prolyl hydroxylases (1–3). In step in the biosynthesis of L-ascorbate. Here we report that the plants, several pathways have been proposed to function in L-galactose pathway of ascorbate biosynthesis described in L-ascorbic acid biosynthesis. The best described pathway, the higher plants is conserved in green algae. The Chlamydomonas Smirnoff-Wheeler pathway or the L-galactose pathway, reinhardtii genome encodes all the enzymes required for vitamin C involves 10 enzymatic steps to convert D-glucose to L-ascorbic biosynthesis via the L-galactose pathway. We have characterized acid via intermediate formation of GDP-D-mannose, GDP-L- recombinant C. reinhardtii VTC2 as an active GDP-L-galactose galactose, L-galactose-1-P, L-galactose, and L-galactono-1,4- phosphorylase. C. reinhardtii cells exposed to oxidative stress show lactone (4). Whereas the initial six steps are also involved in cell increased VTC2 mRNA and L-ascorbate levels. Genes encoding wall/glycoprotein biosynthesis, GDP-L-galactose phosphory- enzymatic components of the ascorbate-glutathione system (e.g. lase (VTC2/VTC5) catalyzes the first committed step in ascorbate peroxidase, manganese superoxide dismutase, and dehy- L-ascorbic acid biosynthesis forming L-galactose-1-P (5, 6). 3 droascorbate reductase) are also up-regulated in response to L-Galactose-1-P phosphatase (VTC4), L-galactose dehydro- increased oxidative stress. These results indicate that C. reinhardtii genase (L-Gal-DH), and L-galactono-1,4-lactone dehydrogen- VTC2, like its plant homologs, is a highly regulated enzyme in ase (GLDH) catalyze the final steps in the Smirnoff-Wheeler ascorbate biosynthesis in green algae and that, together with the pathway in higher plants such as Arabidopsis thaliana (7–9). ascorbate recycling system, the L-galactose pathway represents the The biosynthesis of L-ascorbic acid is not characterized in major route for providing protective levels of ascorbate in oxida- detail in the green algae. Unicellular green algae such as the tively stressed algal cells. chlorophytes Chlorella pyrenoidosa and Prototheca moriformis can synthesize L-ascorbate using the L-galactose pathway (10– 12). Two other photosynthetic unicellular protists (Euglena * This work was supported, in whole or in part, by National Institutes of Health gracilis and Ochromonas danica) (13, 14) and a diatom (Cyclo- Grant GM026020 (to S. G. C.), the Division of Chemical Sciences, Geosci- tella cryptica) utilize the inversion pathway commonly found in ences and Biosciences, Office of Basic Energy Sciences of the United States animals (supplemental Fig. S1) (15). Here we provide evidence Department of Energy Grant DE-FD02-04ER15529 (to S. S. M.), an Ellison that the Smirnoff-Wheeler pathway is completely conserved in Medical Foundation Senior Scholar Award (to S. G. C.), and UCLA Philip Whitcome and Graduate Division Dissertation Year Fellowships (to the green alga C. reinhardtii. The VTC2 protein from C. rein- L. N. A.). The preparation of recombinant VTC2 enzyme was supported by hardtii is highly similar to higher plant VTC2/VTC5, contain- the Department of Energy Grant DE-FC03-02ER63421 (to David Eisenberg). ing the HXHXH motif characteristic of members of the HIT □S This article contains supplemental Figs. S1–S3 and Tables S1–S5. The nucleotide sequence(s) reported in this paper has been submitted to the Gen- BankTM/EBI Data Bank with accession number(s) JQ246433. 1 To whom correspondence may be addressed. E-mail: [email protected]. 3 The abbreviations used are: VTC4, L-galactose-1-P phosphatase; L-Gal-DH, edu. L-galactose dehydrogenase; GLDH, L-galactono-1,4-lactone dehydrogen- 2 To whom correspondence may be addressed: Dept. Chemistry and Bio- ase; APX, ascorbate peroxidase; MDAR, monodehydroascorbate reduc- chemistry, University of California, Los Angeles, 607 Charles E. Young Dr. E., tase; DHAR, dehydroascorbate reductase; GSHR, glutathione reductase; Los Angeles, CA 90095. Tel.: 310-825-8300 or 310-825-8754; Fax: 310-206- MBP, maltose-binding protein; NDP, nucleoside diphosphate; HIT, histi- 1035 or 310-825-1968; E-mail: [email protected]. dine triad; tBuOOH, tert-butyl-hydroperoxide.

14234 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 17•APRIL 20, 2012

33 Ascorbate Biosynthesis and Regulation in Chlamydomonas

protein superfamily of nucleotide hydrolases and transferases binding protein (MBP) fusion using the Invitrogen protocol. (16). DNA sequencing (Genewiz) was used to confirm the sequence Higher plants facing increased oxidative stress exhibit, in of the expression construct. addition to increased VTC2 mRNA and activity levels, elevated VTC2 Expression and Purification—The expression plasmid transcript abundance for all the enzymes of the vitamin C recy- was transformed into Escherichia coli BL21-Gold (DE3) cells cling pathway (ascorbate-glutathione system) in the chloro- (Novagen). Cells were grown in LB medium at 37 °C to an

plast including ascorbate peroxidase (APX), monodehy- A600 nm of 0.6 at which point the temperature was shifted to droascorbate reductase (MDAR), dehydroascorbate reductase 18 °C and protein expression was induced by the addition of (DHAR), and glutathione reductase (GSHR) (2, 17). In this isopropyl 1-thio-␤-D-galactopyranoside to a concentration of 1 work, we found that Chlamydomonas reinhardtii cells facing mM. Cell growth was continued overnight and the cells were oxidative stress have increased abundance of VTC2 transcripts collected by centrifugation the following day. The cell pellet and all the enzymes of the ascorbate-glutathione system, as well was resuspended in wash buffer (20 mM Tris, pH 8.0, 300 mM as higher total ascorbate content. This suggests that C. rein- NaCl, 10 mM imidazole, 0.2% Nonidet P-40, 10% glycerol) sup- hardtii cells respond to oxidative stress by producing more plemented with protease inhibitor mixture (Sigma), PMSF (100 Ϫ1 L-ascorbic acid both via de novo synthesis through the L-galac- ␮M), DNase (20 ␮g ml ), a few crystals of lysozyme, and 10 mM tose pathway and via increased recycling. ␤-mercaptoethanol. Cells were lysed using a French press. The lysate was clarified by centrifugation (30 min at 35,000 ϫ g) and EXPERIMENTAL PROCEDURES the supernatant was incubated with nickel-nitrilotriacetic acid- Materials—ADP-D-Glc, GDP-D-Glc, GDP-D-Man, UDP-D- agarose beads (Qiagen) for 60 min at 4 °C. The beads were Gal, UDP-D-Glc (all in the ␣-configuration), GDP-␤-L-Fuc, and washed extensively with wash buffer and bound protein was GDP were from Sigma. GDP-␤-L-Gal, synthesized and purified eluted with elution buffer (wash buffer containing 300 mM as described (18) was provided by Prof. Shinichi Kitamura imidazole). VTC2 was further purified by size exclusion chro- (Osaka Prefecture University). This preparation was further matography using a HiLoad Superdex S-200 column (GE Life purified by the reversed-phase HPLC method as described in Sciences) equilibrated in 20 mM Tris, pH 8.0, 300 mM NaCl, and Ref. 5. Fractions containing GDP-L-Gal were lyophilized, resus- 10% glycerol. Peak fractions were analyzed by SDS-PAGE and Ϫ pended in H2O, and stored at 20 °C. Hydrogen peroxide those containing VTC2 were pooled and concentrated. Two (30%) and tert-butyl hydroperoxide (tBuOOH) (70%) were pur- peaks containing VTC2 MBP fusion proteins were obtained by chased from Fisher and Lancaster Synthesis, Inc., respectively. size exclusion chromatography. Both peaks contained pure Ascorbate oxidase from Cucurbita sp. (EC 1.10.3.3; A0157) was MBP-VTC2 fusion protein and were pooled separately and purchased from Sigma. concentrated. The fraction showing the highest activity was Strains and Culture Conditions—C. reinhardtii strains 2137 used for enzymatic analyses. (CC1021) and CC425 were obtained from the Chlamydomonas Nucleic Acid Analysis—Total RNA was extracted from expo- culture collection (Duke University) and grown in Tris acetate- nentially growing C. reinhardtii cells as previously described Ϫ phosphate (TAP) medium (19) at 24 °C and 50–100 ␮mol m 2 (21). RNA quality was assessed using an Agilent 2100 Bioana- Ϫ s 1 light intensity. lyzer and RNA blot hybridization for CBLP as described (22). Sequencing of C. reinhardtii VTC2—The VTC2 cDNA clone The probe used for detection was a 915-bp EcoRI fragment MXL096d05 (corresponding to EST BP098619) was completely from the cDNA insert (encoding CBLP) in plasmid pcf8-13 sequenced. It contains the entire predicted VTC2 open reading (23). frame, 1857 bp long, encoding a protein of 618 amino acids. The Quantitative Real-time PCR on cDNA—cDNA synthesis and open reading frame is flanked by 499 nt of 5Ј untranslated quantitative real-time PCR was performed on technical tripli- region and a 3Ј untranslated region of 1396 nt followed by a cates as described (22) using the gene-specific primers listed in 68-nt poly(A) tail. The complete VTC2 sequence has been supplemental Table S5. The data are presented as the fold- deposited in NCBI (GenBank accession JQ246433). change in mRNA abundance, normalized to an endogenous VTC2 Cloning—The VTC2 expression construct was gener- reference transcript (CBLP or UBQ2), relative to the sample

ated by nested PCR and the Gateway recombinational cloning grown before 1 mM H2O2 or 0.1 mM tBuOOH treatment (time system (Invitrogen) as described (20). Briefly, the coding 0). The abundance of the two reference transcripts did not sequence of VTC2 (amino acids D2-A618) was amplified with change under the conditions tested. Phusion polymerase (New England Biolabs) from plasmid Ascorbate Measurements—C. reinhardtii cells were grown in Ϫ MXL096d05 using gene-specific primers with 5Ј extensions TAP medium to a density of 3 ϫ 106 cells ml 1, collected by encoding a TEV protease cleavage site in the forward primer centrifugation at 2,500 ϫ g for 5 min, resuspended in extraction (VTC2.D2) and a C-terminal hexahistidine tag followed by a buffer containing 2% metaphosphoric acid, 2 mM EDTA, and 5 stop codon in the reverse primer (VTC2.A628) (supplemental mM DTT and stored at Ϫ80 °C. To prepare extracts for vitamin Table S4). The initial product was then amplified with a second C analysis, cells were lysed by freeze/thaw cycling and the sol- set of primers to introduce AttB1 and AttB2 recombination uble fractions were separated by centrifugation (16,100 ϫ g, 10 sites (PE-277 and PE-278) (supplemental Table S4). Amplifica- min at 4 °C). Vitamin C content was measured by reversed- tion products were gel purified and recombined into the donor phase HPLC on an Econosphere C-18 column (5 ␮m bead size, vector pDONR201 and subsequently into expression vector 4.6 ϫ 250 mm; Alltech Associates, Deerfield, IL) using a pKM596 (20) to produce an N-terminal His-tagged maltose- Hewlett Packard Series II 1090 liquid chromatograph. A

APRIL 20, 2012•VOLUME 287•NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 14235 34 Ascorbate Biosynthesis and Regulation in Chlamydomonas

mobile-phase gradient of 0–40% acetonitrile in 20 mM triethyl lihood method based on the Whelan and Goldman model (28). ammonium acetate, pH 6.0, was used at a flow rate of 1 ml The tree with the highest log likelihood is shown. Branch Ϫ min 1. The injection volume was 50–100 ␮l. Ascorbic acid was lengths reflect the number of substitutions per site. All posi- detected by monitoring the absorbance at 265 nm. The ascorbic tions containing gaps and missing data were eliminated. Align- acid peak was identified by comparison with the elution time of ment of putative VTC2 homologs was performed using MUS- an L-ascorbate standard and by demonstrating a decrease of the CLE, and evolutionary analyses were conducted in MEGA5 peak area after the samples were treated with ascorbate oxidase. (29). This treatment was performed by adding 2 units of ascorbate RESULTS oxidase from Cucurbita sp. (EC 1.10.3.3) to 60 ␮l of the extract in a final concentration 0.12 M monosodium citrate for 1 h at The C. reinhardtii Genome Encodes a Homolog of Plant GDP- 4 °C. The final pH of the reaction was about 5.6. The differences L-galactose Phosphorylase—Biosynthesis of vitamin C in higher in the peak areas measured before and after addition of ascor- plants occurs mainly via the L-galactose pathway (9). In A. thali- bate oxidase were used to calculate ascorbic acid levels based on ana, the first committed step in the sequence of 10 enzymatic a standard curve. The cellular concentration of L-ascorbate was reactions from D-glucose to L-ascorbate is conversion of GDP- determined using a cell volume of 140 femtoliters (24). L-galactose to L-galactose-1-P, a reaction catalyzed by the GDP- HPLC-based Nucleoside Diphosphate (NDP)-Hexose Phos- L-galactose phosphorylase VTC2. Therefore, we were inter- phorylase Assay—NDP-hexose phosphorylase activities of ested in finding homologs of VTC2 in C. reinhardtii and other recombinant VTC2 enzyme were assayed by measuring NDP green algae such as Volvox carteri, Chlorella sp. NC64A, Coc- formation after incubation with NDP-hexose in a reaction mix- comyxa sp. C169, Micromonas sp. RCC299, and Ostreococcus ture at pH 7.5 containing 50 mM Tris-HCl, 5 mM sodium phos- RCC809. BLASTp searches identified a VTC2 homolog phate, 10 mM NaCl, and 1 mM DTT. Reactions (26 °C) were (Cre13.g588150) in C. reinhardtii (supplemental Fig. S2). initiated by enzyme addition and stopped after 5–10 min by Cre13.g588150 exhibits 46% amino acid sequence identity to A. heating at 98 °C for 5 min. After removal of precipitated protein thaliana VTC2. The A. thaliana genome encodes a VTC2 para- by centrifugation, supernatants were analyzed by anion-ex- log, VTC5, which shows enzymatic properties similar to those change HPLC as described in Ref. 5. NDP and NDP-hexose of VTC2 (30, 31). The C. reinhardtii protein has 47% identity to concentrations were calculated by comparing the integrated VTC5 at the amino acid level. Because the C. reinhardtii peak areas with those of standard NDP or NDP-hexose solu- genome encodes only a single protein highly similar to A. thali-

tions. GraphPad Prism (La Jolla, CA) was used to calculate Km ana VTC2/VTC5, we termed this protein Cre13.g588150 and Vmax values. VTC2. The amino acid sequence of the VTC2 protein from C. RNA-Seq—Total RNA samples prepared from C. reinhardtii reinhardtii does not contain any transmembrane domains. Sev- strain 2137 grown photoheterotrophically in the presence of 1 eral subcellular localization prediction programs (ChloroP,

mM H2O2 for 30 and 60 min were sequenced on a GAIIx plat- TargetP, Psort, and PredSL) indicated that C. reinhardtii VTC2 form. cDNA libraries were made using the protocol from Illu- does not possess obvious organellar targeting sequences, sug- mina and sequenced as single-end 76-mers. Raw and processed gesting that, like the plant homologs, it is most likely a cytosolic sequence files are available at the NCBI Gene Expression protein. C. reinhardtii VTC2 contains a highly conserved histi- Omnibus (accession number GSE34826). Sequence reads were dine triad (HIT) motif (HXHXH, where X is a hydrophobic aligned using Bowtie (25) in single-end mode and with a maxi- residue) (supplemental Fig. S2). C. reinhardtii VTC2 is more mum tolerance of 3 mismatches to the Au10.2 transcript closely related to the Volvox VTC2 protein and among algal sequences corresponding to the version 4 assembly of the C. homologs it appears that the Micromonas sp. and Ostreococcus reinhardtii genome. Expression estimates were obtained for sp. proteins are more closely related to higher plant VTC2 pro- each individual run in units of RPKMs (reads per kilobase of teins than to the animal VTC2 homologs (Fig. 1). mappable transcript length per million mapped reads) (26), Enzymatic Components of L-Galactose Pathway to Vitamin C after normalization by the number of aligned reads and tran- Biosynthesis Are Conserved in Green Algae—Because higher script mappable length (27). Technical replicates were averaged plant VTC2 has orthologs in C. reinhardtii and other green to obtain per-sample expression estimates. Final expression algae, we investigated whether the green algae encode the rest estimates and fold-changes were obtained for each biological of the components of the Smirnoff-Wheeler pathway. BLASTp replicate. and tBLASTn searches identified orthologs (defined as mutual Sequence and Phylogenetic Analyses—To search for or- best BLAST hit) for almost all L-galactose pathway enzymes in thologs/homologs of A. thaliana members of the L-galactose six green algae (C. reinhardtii, V. carteri, Chlorella sp. NC64A, pathway in green algae, two BLAST searches were performed. Coccomyxa sp. C169, Micromonas sp. RCC299, and Ostreococ- In the first, protein sequences of the A. thaliana proteins were cus lucimarinus). Orthologs of A. thaliana phosphomannose used as queries to search against algal genomes databases. isomerase (PMI1), phosphomannomutase (PMM), GDP-D- Sequences of putative homologs were retrieved and used for the mannose 3Љ,5Љ-epimerase (GME1), L-galactose-1-P phospha- second query by performing BLASTp against the A. thaliana tase (VTC4), L-Gal-DH, and GLDH are present in all six species database. If in the second query the highest-scoring homolog in (Fig. 2 and supplemental Table S1). Interestingly, our sequence A. thaliana was exactly the original A. thaliana query sequence, analysis identified orthologs of GDP-D-mannose pyrophos- that protein was considered an ortholog (mutual best hit). Phy- phorylase (VTC1) in C. reinhardtii, V. carteri, Chlorella sp. logenetic relationships were inferred using the Maximum Like- NC64A, and Coccomyxa sp. C169, but not in Micromonas sp.

14236 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 17•APRIL 20, 2012 35 Ascorbate Biosynthesis and Regulation in Chlamydomonas

FIGURE 1. Phylogenetic tree of VTC2-like proteins. Protein sequences homologous to C. reinhardtii VTC2 were used to build the phylogenetic tree as described under “Experimental Procedures.” Bootstrap values are shown below the branches. Bar, 0.2 amino acid substitutions per site. C. reinhardtii (Cre), Coccomyxa sp. C-169 (Coc_C169), Chlorella sp. NC64A (ChlNC64A), Volvox carteri f. nagariensis (Vca), O. lucimarinus (Ost9901), Ostreococcus sp. RCC809 (OstRCC809), M. pusilla (MicpuC2), Micromonas sp. RCC299 (MicpuN3), A. thaliana (Ath), Physcomitrella patens (Ppa), Ricinus communis (Rco), Oryza sativa (Osa), Selaginella moellendorffii (Selmo), D. melanogaster (Dme), C. elegans (Cel), and H. sapiens (Hsa).

(RCC299 and Micromonas pusilla) nor in Ostreococcus sp. aldonolactonase (gulonolactonase) (35). SMP30 (senescence (Ostreococcus tauri, Ostreococcus lucimarinus, and Ostreococ- marker protein 30) has been recently identified to function as a cus RCC899) (Fig. 2 and supplemental Table S1). Micromonas glucono/gulonolactonase (36). The C. reinhardtii genome does sp. and Ostreococcus sp. both belong to the class Prasinophy- not encode a homolog to SMP30. Hence, it seems unlikely that ceae, which diverged at the base of the algal lineage and are C. reinhardtii would use this route as an alternate pathway to therefore more distantly related to the chlorophyte algae. It is vitamin C biosynthesis. possible that Micromonas spp. and Ostreococcus spp. synthe- It has also been proposed that biosynthesis of L-ascorbic acid size GDP-D-mannose using a different pathway (see “Discus- could occur via the galacturonate or salvage pathway (7, 9) (sup- sion”). Overall, we conclude that the L-galactose pathway to plemental Fig. S1). This pathway would involve conversion of L-ascorbate biosynthesis is conserved in the green algae. methyl-D-galacturonate to D-galacturonate. The enzyme cata- Because alternative L-ascorbate biosynthetic pathways have lyzing this reaction has not yet been identified. Formation of been proposed (7, 32), we searched for orthologs/homologs of L-galactonate from D-galacturonate is catalyzed in ripening the enzymes catalyzing the proposed steps in these alternate strawberry fruits by an aldo-keto reductase specific for D-galac- pathways (supplemental Fig. S1). First, the proposed L-gulose turonate (GalUR) (37). The C. reinhardtii genome encodes sev- pathway (33) involves the A. thaliana GDP-D-mannose 3Љ,5Љ- eral aldo-keto reductases with homology to strawberry D-galac- epimerase, which is orthologous to C. reinhardtii SNE1. This turonate reductase (supplemental Table S1), but none of them enzyme can form GDP-L-gulose, which, if converted to L-gu- is an ortholog of the plant enzyme. On the other hand, lono-1,4-lactone, would provide a substrate for an oxidase reac- orthologs of the strawberry GalUR are present in other algal tion leading directly to L-ascorbate. Although C. reinhardtii species such as Chlorella sp. NC64A, Micromonas sp. RCC299, GLDH demonstrates 30% amino acid sequence identity to the O. lucimarinus, or V. carteri (supplemental Table S1). The rat L-gulono-1,4-lactone dehydrogenase/oxidase (Gulo) (34), penultimate reaction in the galacturonate pathway (L-galacto- and two other similar proteins are present (Cre14.g611650 with nate to L-galactono-1,4-lactone conversion) would require the 26% amino acid identity and Cre03.g177600 with 22% amino function of an aldonolactonase, which has been recently char- acid identity), these putative enzymes have not been character- acterized in the protist E. gracilis (38). BLASTp and tBLASTn ized and no homologs of enzymes converting GDP-L-gulose to searches did not identify any homologs to E. gracilis aldonolac- L-gulono-1,4-lactone have been found. tonase in C. reinhardtii or V. carteri, but orthologs are present L-Gulono-1,4-lactone could also be potentially formed from in Chlorella sp. NC64A, Coccomyxa sp. C169, Micromonas sp. myo-inositol via D-glucuronate (supplemental Fig. S1, animal- RCC299, and O. lucimarinus (supplemental Table S1). like pathway). C. reinhardtii codes for a potential myo-inositol We conclude that essential components of the alternate oxidase (Cre01.g025850) that might be responsible for the for- pathways to L-ascorbate biosynthesis are missing in C. rein- mation of D-glucuronate and shows 31% amino acid identity to hardtii. We could identify homologs of L-gulono-1,4-lactone A. thaliana MIOX4 (supplemental Table S1). Conversion of dehydrogenase for the L-gulose pathway and galacturonate D-glucuronate to L-gulonate would require the action of a reductase for the salvage pathway, but there are no orthologs to glucuronate reductase, which has not been identified in plants. rat Gulo or strawberry GalUR and the sequence similarity is Formation of L-gulono-1,4-lactone from L-gulonate requires an poor. On the other hand, we show that all components of the

APRIL 20, 2012•VOLUME 287•NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 14237 36 Ascorbate Biosynthesis and Regulation in Chlamydomonas

TABLE 1 Substrate specificity of recombinant C. reinhardtii VTC2 Various sugar nucleotides (50 ␮M) were incubated in the presence of C. reinhardtii Ϫ1 VTC2 recombinant enzyme (0.025 ␮g ml ) for 10 min at 26 °C. With GDP-L-Gal as Ϯ a substrate,Ϫ the specific activityϪ of the C. reinhardtii VTC2 enzyme was 25.4 5.1 ␮mol min 1 mg protein 1 (mean Ϯ S.D., n ϭ 3); this value was taken as 100%. The phosphorylase activities found with the other sugar nucleotides are given as a percentage of the activity found with GDP-L-Gal Ϯ S.D. Values represent the means of 2–3 individual experiments for each substrate. Relative activity of recombinant Substrate C. reinhardtii VTC2 .S.D ؎ % GDP-L-Gal 100 GDP-D-Glc 87.4 Ϯ 29.2 GDP-L-Fuc 51.4 Ϯ 15.1 GDP-D-Man 0 Ϯ 0 UDP-D-Glc 0.5 Ϯ 0.8 UDP-D-Gal 2.5 Ϯ 4.1 ADP-D-Glc 3.7 Ϯ 5.8

D-Glc, UDP-D-Gal, and ADP-D-Glc (Table 1). Thus, our data indicate that C. reinhardtii VTC2 possesses a similar nucleotide sugar substrate specificity as A. thaliana VTC2. The conserved HIT motif (HXHXH) in C. reinhardtii VTC2 is typical of HIT hydrolases, whereas plant VTC2 proteins contain a HIT motif typical for HIT transferases/phosphorylases (HXHXQ) (16). Therefore, we tested the acceptor specificity of C. reinhardtii VTC2 by measuring the GDP-L-Gal or GDP-D- Glc consumption after incubation of the recombinant VTC2 enzyme with different possible acceptors. When recombinant

C. reinhardtii VTC2 was incubated in the absence of Pi, we detected no hydrolytic activity (Fig. 3). However, in the pres-

ence of Pi, we observed a dramatic increase in GDP-L-Gal and GDP-D-Glc consumption (Fig. 3). Incubation of the enzyme

with pyrophosphate (PPi), D-Glc-1-P (in the presence of GDP- FIGURE 2. The L-galactose pathway of ascorbic acid biosynthesis is con- L-Gal), or L-Gal-1-P (in the presence of GDP-D-Glc) did not served in green algae. Colored squares indicate the number of A. thaliana result in any significant nucleotide sugar substrate consump- orthologs present in each organism. The enzymes catalyzing the successive steps are hexokinase (HXK), phosphoglucose isomerase (PGI), phosphoman- tion (Fig. 3). Additionally, we did not detect the formation of nose isomerase (PIM), phosphomannomutase (PMM), GDP-L-mannose pyro- GMP or GTP, the expected products of hydrolase or pyrophos- phosphorylase (VTC1), GDP-D-mannose-3Ј,5Ј-epimerase (GME), GDP-L-galac- phorylase activity, under any of these conditions (data not tose phosphorylase (VTC2), L-galactose-1-P phosphatase (VTC4), L-Gal-DH, and GLDH. shown). These data clearly indicate that C. reinhardtii VTC2 is a phosphorylase like the Arabidopsis enzyme. C. reinhardtii VTC2 has similar, low micromolar, Michaelis plant Smirnoff-Wheeler pathway have orthologs in C. rein- constants for both GDP-L-Gal and GDP-D-Glc (Table 2). Inter- hardtii and in other algal species. These results point to a con- estingly, with C. reinhardtii VTC2 we have found at least 10

served L-galactose pathway to L-ascorbate biosynthesis, which times higher kcat values for both substrates compared with the might represent the major route to L-ascorbate biosynthesis in A. thaliana VTC2 recombinant enzyme, leading to about 10 algae, in particular C. reinhardtii and other Volvocales. times higher catalytic efficiencies for the former than for the Recombinant C. reinhardtii VTC2 Is a GDP-L-galactose/ latter enzyme (Table 2). GDP-D-glucose Phosphorylase—Previous studies demonstrated VTC2 Transcript Levels and Ascorbate Levels Are Increased that A. thaliana VTC2 and VTC5 are GDP-L-galactose phos- in Response to Oxidative Stress—Previous studies have indi- phorylases, converting GDP-L-galactose into L-galactose-1-P cated that A. thaliana VTC2 mRNA levels are increased in

and GDP in the presence of Pi (5, 30, 31). To test whether C. leaves subjected to high light (30) and in seedlings grown in reinhardtii VTC2 can catalyze this reaction, recombinant C. light compared with those grown in the dark (39). Therefore, reinhardtii VTC2 was purified as a His- and MBP-tagged pro- we tested whether transcript levels of C. reinhardtii VTC2 tein. Because A. thaliana VTC2 shows GDP-L-galactose and respond to oxidative stress. C. reinhardtii was grown photohet- 6 Ϫ1 GDP-D-glucose phosphorylase activities, we first determined erotrophically to 2 ϫ 10 cells ml , then challenged with 1 mM the activity of the C. reinhardtii enzyme on various sugar nucle- H2O2 or 0.1 and 0.2 mM tBuOOH for 30, 60, 120, and 240 min. otides in the presence of inorganic phosphate (Table 1). Similar Both H2O2 and tBuOOH (an organic peroxide capable of activity was seen with GDP-L-Gal and GDP-D-Glc, whereas a inducing lipid peroxidation) treatments enhance intracellular 2-fold lower activity was found with GDP-L-Fuc. No significant reactive oxygen species production. The concentrations of phosphorylase activity was measured with GDP-D-Man, UDP- H2O2 and tBuOOH and time points used in this study were

14238 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 17•APRIL 20, 2012 37 Ascorbate Biosynthesis and Regulation in Chlamydomonas

FIGURE 3. Acceptor specificity of C. reinhardtii recombinant VTC2. GDP-L-Gal and GDP-D-Glc consumption was measured by the HPLC assay described under “Experimental Procedures.” GDP-L-Gal or GDP-D-Glc were added to the reaction mixtures at a final concentration of 50 ␮M. The consumption of GDP-L-Gal (A) and GDP-D-Glc (B) was measured with a final enzyme concentration of 0.025 ␮g/ml (light gray bars) or 0.25 ␮g/ml (dark gray bars).

TABLE 2 Characterization of the GDP-hexose phosphorylase activities of the recombinant A. thaliana and C. reinhardtii proteins

Values for the A. thaliana enzyme were taken from Linster et al. (5). Km and Vmax values for the C. reinhardtii VTC2 homolog were obtained by fitting the initial rate data to the Michaelis-Menten equation using the GraphPad Prism program. Enzymatic turnover numbers were derived from the Vmax values by using a molecular mass of 110 kDa for His-MBP-tagged C. reinhardtii enzyme with the assumption that the enzyme preparations were pure. Incubation times and enzyme concentrations were adjusted to obtain initial velocity data. Enzymatic activities were measured by the HPLC assay as described under “Experimental Procedures.” Values are the mean Ϯ S.D. calculated from 2–3 individual experiments for each substrate.

kcat Km kcat/Km Substrate C. reinhardtii A. thaliana C. reinhardtii A. thaliana C. reinhardtii A. thaliana Ϫ1 Ϫ1 Ϫ1 s mM s M 7 6 GDP-L-galactose 615 Ϯ 3 64 Ϯ 8 0.008 Ϯ 0.001 0.010 Ϯ 0.001 9.2 Ϯ 1.3 ϫ 10 6.3 Ϯ 0.9 ϫ 10 7 6 GDP-D-glucose 813 Ϯ 277 23 Ϯ 3 0.0088 Ϯ 0.0029 0.0044 Ϯ 0.0016 8.0 Ϯ 1.3 ϫ 10 5.7 Ϯ 2.3 ϫ 10

previously shown to have no effect on cell growth in C. rein- phosphorylase step is potentially the key regulatory point of the hardtii and yet were high enough to induce the antioxidant L-ascorbate biosynthetic pathway in C. reinhardtii. defense mechanisms (40–43). A lower concentration of Next, we asked the question whether the increased VTC2

tBuOOH was used because it was more stable than H2O2 under mRNA levels correlate with a change in ascorbate content. our culture conditions (supplemental Fig. S3). VTC2 transcript Total ascorbate levels were measured in cell extracts from C.

abundance was assessed by real-time PCR. We found that reinhardtii grown under 1 mM H2O2 or 0.1 mM tBuOOH stress VTC2 mRNA transcript abundance increased 4-fold after 30 for 2, 4, 6, and 8 h. Total ascorbate content increased progres-

min and reached a maximum of 7-fold induction after 120 min sively after addition of H2O2, showing a slight increase after 2 h exposure to 1 mM H2O2 (Fig. 4A). When C. reinhardtii cells and reaching a maximum after 8 h, where we measured 7-fold were exposed to 0.1 mM tBuOOH, we observed a more dra- higher ascorbate concentrations than in untreated cells (Fig. matic induction in the VTC2 transcript levels (50-fold increase 5A). On the other hand, cells treated with 0.1 mM tBuOOH after 30 min with the highest induction of 155-fold after 240 displayed a 4-fold higher ascorbate content 2 h after addition of min). Increasing the tBuOOH concentration to 0.2 mM resulted tBuOOH, with a further increase after 4 h (5-fold). In contrast

in an even higher increase in the VTC2 mRNA abundance (150- to C. reinhardtii cells exposed to H2O2, after 6 h of tBuOOH fold after 30 min and 250-fold after 120 min) (Fig. 4A). treatment we noticed a drop in the total ascorbate levels (3-fold To assess the overall impact of peroxide stress on the Smir- more compared with untreated cells), which decreased even noff-Wheeler pathway, we quantified the abundance of tran- further after 8 h to levels similar to those observed for untreated

scripts for each gene in the pathway in H2O2-treated versus cells (Fig. 5B). The observation that tBuOOH treatment untreated cells. Changes in the VTC2 transcript levels after depletes cellular ascorbate has been made previously in rat

H2O2 exposure observed by RNA-Seq are very similar to those hepatocytes (44) and rat astrocytes (45). Altogether, our observed by real-time PCR (6.4-fold induction after 30 min and results indicate a correlation between the VTC2 mRNA lev- 8.6-fold induction after 60 min) (Fig. 4B). Other components of els and L-ascorbic acid content in C. reinhardtii cells exposed the pathway including MPI1, PMM, and GMP1 showed at best to oxidative stress. a 2-fold increase in their transcript abundance after 60 min of Genes Encoding Components of Ascorbate-glutathione Scav-

exposure to H2O2.ThetranscriptlevelsofSNE1, VTC4, enging System Are Induced in Response to Oxidative Stress— L-GalDH,andGLDH did not change significantly (Fig. 4B The ascorbate-glutathione cycle is a well known mechanism to

and supplemental Table S2)inresponsetoH2O2 treatment. scavenge H2O2 in various cell compartments (2), particularly in Together, the combined real-time PCR and RNA-Seq analyses plants (46) (and see Fig. 7). Therefore, we were interested in demonstrated that VTC2 mRNA levels are highly and selectively expression profiles of the genes encoding the ascorbate-gluta- induced by oxidative stress, indicating that the GDP-L-galactose thione system components in C. reinhardtii cells exposed to 1

APRIL 20, 2012•VOLUME 287•NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 14239 38 Ascorbate Biosynthesis and Regulation in Chlamydomonas

FIGURE 4. VTC2 transcript levels are increased in response to oxidative stress. A, fold-change (log10)ofVTC2 transcript assessed by real-time PCR in C. reinhardtii cells grown photoheterotrophically in the presence of 1 mM H2O2 (filled circles), 0.1 mM tBuOOH (open triangles), or 0.2 mM tBuOOH (filled inverted triangles)fortheindicatedtimes.EachdatapointrepresentstheaverageofthreetechnicaltriplicatesforquantitativePCRfromonebiological replicate of treated cells. The symbol for the untreated cultures (time 0) represents the overlap of all three symbols used. B,transcriptabundanceof genes encoding enzymes of the L-galactose pathway quantified by RNA-Seq in C. reinhardtii cells grown in the presence of 1 mM H2O2 for 30 min (open circles)or60min(filled circles). Fold-changes were calculated relative to the transcript abundances in untreated cells for both quantitative PCR and RNA-Seq experiments.

FIGURE 5. Ascorbate levels are elevated in response to oxidative stress. Vitamin C concentration was measured in extracts of C. reinhardtii cells exposed to 1mM H2O2 (A) or 0.1 mM tBuOOH (B) for 2, 4, 6, and 8 h. Error bars represent the S.D. from three biological replicates.

mM H2O2 or 0.1–0.2 mM tBuOOH for 30, 60, 120, or 240 min. The resulting oxidized GSSG is converted back to GSH by glu- In plants, and most likely also in C. reinhardtii, Photosystem I is tathione reductases (GSHR1/2 in C. reinhardtii). DHAR1 tran- . the major site for superoxide anion production (O2) (17), which script abundance was progressively up-regulated after expo-

is disproportionated to H2O2 by the action of one or several sure to 1 mM H2O2 (from 2–3-fold after 30 min to 50-fold after superoxide dismutases. Here we found that in C. reinhardtii, 240 min). A similar trend of DHAR1 overexpression was MSD3 transcript levels (encoding plastid-localized MnSOD3) observed under tBuOOH treatment (Fig. 6A). The transcript are highly induced in response to peroxide treatment (Fig. 6A). levels of the key enzyme involved in glutathione synthesis, Treatment of C. reinhardtii cells with 1 mM H2O2 resulted in a ␥-glutamylcysteine synthetase (GSH1), and GSHR1 were 2–15-fold induction of this gene over the 4-h exposure period. induced only in response to 1 mM H O (Fig. 6A). Interestingly, An even higher level of up-regulation (100-fold after 60 min) 2 2 neither of those transcripts changed in abundance during was reached when C. reinhardtii cells were exposed to 0.1 mM the first 60 min after 0.1 or 0.2 mM tBuOOH addition and in tBuOOH. H O produced by MnSOD3 is reduced to H O by 2 2 2 fact they even decreased after 120 min (Fig. 6A). RNA-Seq ascorbate in a reaction catalyzed by APX1. The mRNA abun- analysis of C. reinhardtii cells exposed to 1 mM H O for 30 dance of C. reinhardtii APX1 was induced 2–4-fold after expo- 2 2 and 60 min indicated up-regulation of all the genes encoding sure to 1 mM H2O2, whereas 0.1 mM tBuOOH treatment resulted in a 10–15-fold induction of APX1 transcript levels the enzymes of the ascorbate-glutathione cycle (Fig. 6B and (Fig. 6A). Ascorbate peroxidase oxidizes ascorbate to monode- supplemental Table S3). The increase in their transcript hydroascorbate, which is either reduced to ascorbate by the abundance was higher after 60 min and, in agreement with action of MDAR1, or spontaneously disproportionates to dehy- the real-time PCR results, MSD3 and DHAR1 were the most droascorbate. MDAR1 mRNA abundance was induced in highly induced genes. We conclude, based on the transcript

response to 1 mM H2O2 (5–6-fold after 120 min), whereas 0.1 abundance changes observed in this study in response to mM tBuOOH treatment resulted in a more subtle 2–3-fold up- peroxide stress, that the ascorbate-glutathione system plays regulation of this gene. Dehydroascorbate can be reduced back an important role in the oxidative stress response in C. to ascorbate by DHAR1. The reaction requires reduced GSH. reinhardtii.

14240 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 17•APRIL 20, 2012 39 Ascorbate Biosynthesis and Regulation in Chlamydomonas

FIGURE 6. Oxidative stress conditions result in up-regulation of GSH-ascorbate cycle enzymes. A, relative transcript levels of genes coding for ascorbate- glutathione cycle components were quantified by real-time PCR in C. reinhardtii cells grown in the presence of 1 mM H2O2 (filled circles), 0.1 mM tBuOOH (open triangles), or 0.2 mM tBuOOH (filled inverted triangles) for 30, 60, 120, and 240 min. Fold-changes were calculated relative to the transcripts abundance in untreated cells and data are represented in the log10 scale. Relative transcript abundance was calculated as described in the legend of Fig. 4A. Each data point represents the average of three technical triplicates from one biological replicate. B, RNA-Seq was also used to measure transcript levels of genes encoding for enzymes of the ascorbate-glutathione system in C. reinhardtii cells exposed to 1 mM H2O2 for 30 and 60 min.

DISCUSSION other green algal genomes encode functional plant VTC2 Higher plants synthesize L-ascorbic acid using primarily the homologs. Our sequence analyses identified orthologs of all the Smirnoff-Wheeler pathway (4, 9), in which VTC2 catalyzes a Smirnoff-Wheeler pathway enzymes in C. reinhardtii. More- rate-limiting step by converting GDP-L-galactose to L-galac- over, with the exception of GDP-D-mannose pyrophosphory- tose-1-P (5, 8). Here we provide evidence that C. reinhardtii and lase (VTC1), which appears to be missing in Prasinophyceae

APRIL 20, 2012•VOLUME 287•NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 14241 40 Ascorbate Biosynthesis and Regulation in Chlamydomonas

like Micromonas spp. or Ostreococcus spp., all other enzymes of Here we provide evidence suggesting a role of L-ascorbic the L-galactose pathway are conserved in divergent green algae. acid in protecting C. reinhardtii cells against oxidative The absence of VTC1 in Prasinophyceae might be compen- stress. Reactive oxygen species-inducing chemicals like

sated by the operation, in those species, of VTC2 cycles such as H2O2 and tBuOOH resulted in increased VTC2 mRNA lev- those proposed by Laing et al. (6) or Wolucka and Van Montagu els, which are 10–15 times more abundant after exposure to

(47), where L-galactose-1-P would be formed by a GDP-L-galac- tBuOOH compared with H2O2 treatment. This might be tose transferase activity of VTC2 (using D-Man-1-P or D-Glc- explained by the fact that tBuOOH persists for a longer time

1-P as guanylyl acceptors instead of Pi) and GDP-D-mannose than does H2O2 in liquid cultures. In addition, H2O2 can formation would be ensured by a hypothetical 2Ј-epimerase produce highly reactive hydroxyl radicals, whereas tBuOOH from GDP-D-glucose (8). can decompose to other alkoxyl and peroxyl radicals. Pro-

HIT proteins are members of a superfamily of nucleotide oxidant effects of H2O2 treatment resulted in persistent ele- hydrolases and transferases, which, based on sequence, sub- vated levels of total ascorbate, whereas, after an initial strate specificity, structure, evolution and mechanism, are increase, the total ascorbate levels dropped back to wild-type classified into the Hint, Fhit, Aprataxin, scavenger decap- levels after exposure to tBuOOH for 8 h. This is not surpris- ping protein, and GalT branches (16). The first four ing because exposure of astrocytes, hepatocytes, or Hep2G branches, characterized by a HXHXHmotif,containnucle- cells to tBuOOH had previously been shown to lead to otide hydrolases, whereas GalT branch members, generally decreased levels of intracellular L-ascorbic acid and GSH (44, possessing a HXHXQmotif,arenucleotidephosphorylases 45, 57). An A. thaliana line (ppr40-1)thathasimpairedelec- or transferases. The best-characterized member of the GalT tron flow at complex III showed decreased levels of total branch is galactose-1-phosphate uridylyltransferase, which ascorbate and enhanced activity of GLDH and ascorbate- represents the second enzyme in the Leloir pathway of galac- glutathione cycle enzymes (58). Similarly, inhibition of mito- tose utilization. The HIT motif in A. thaliana and other plant chondrial respiratory electron transport at the levels of com- VTC2 proteins (HXHXQ) and the enzymatic properties of A. plex I, complex II, or complex IV resulted in a 50% decrease thaliana VTC2, which is a GDP-L-Gal/GDP-D-Glc phos- in total ascorbate levels in A. thaliana (59). It is well known phorylase, would place this protein in the GalT branch of the that plant mitochondria are the place where the last step of HIT superfamily. Interestingly, C. reinhardtii VTC2 pos- vitamin C biosynthesis occurs in plants. GLDH is an inner sesses the HXHXHmotiffoundinmembersofthehydrolase membrane mitochondrial flavin enzyme that uses oxidized branches of the HIT superfamily (16). Animal homologs of cytochrome c as an electron acceptor (60) and recently has plant VTC2 also have the HXHXHmotifandhavebeen been shown to be required for accumulation of complex I in shown to act as specific GDP-D-glucose phosphorylases A. thaliana (61). On the other hand, tBuOOH has been needed for quality control of the nucleoside diphosphate shown to inhibit mitochondrial respiratory chain enzymes in sugar pool (48). This work provides an additional example to rat hepatocytes (62). Therefore the higher VTC2 transcript suggest that the HXHXH versus HXHXQmotifsdonot levels and depletion of intracellular ascorbate content in C. always predict the biochemical reaction catalyzed by the cor- reinhardtii exposed for longer times to tBuOOH might at responding HIT enzyme (48–50). In this study we indeed least in part be a result of oxidatively damaged mitochondria showed that the recombinant C. reinhardtii VTC2 enzyme and impaired respiratory electron transport.

has a GDP-L-Gal/GDP-D-Glc phosphorylase activity as do Our RNA-Seq analysis of H2O2 stressed C. reinhardtii the land plant homologs. The algal enzyme can use both cells indicates a significant increase in mRNA levels only for GDP-L-Gal and GDP-D-Glc as substrates and requires inor- VTC2 and only a small increase (1.5–2-fold) for the other ganic phosphate as acceptor. The recombinant purified components of the L-galactose pathway. A similar pattern of enzyme displayed an about 10-fold higher catalytic effi- expression for all genes encoding L-galactose pathway ciency with both nucleotide sugar substrates relative to A. enzymes has been observed in A. thaliana exposed to high thaliana VTC2. The latter was previously found to exhibit light (30). Our results suggest that VTC2 might be the regu- some transferase activity (31), and we also detected a minor latory point controlling L-ascorbate biosynthesis in C. rein- GDP-L-galactose transferase activity with D-Glc-1-P as a hardtii.Supportingevidenceforthisalsocomesfromstudies guanylyl acceptor for recombinant C. reinhardtii VTC2. in A. thaliana where it has been demonstrated that supple- This transferase activity was at least 100-fold lower than its mentation with L-ascorbate decreases VTC2 mRNA abun- phosphorylase activity (data not shown). dance, possibly indicating a feedback inhibition at the tran- L-Ascorbic acid is a major antioxidant in plants and ani- scriptional level (30). Moreover, the increased L-ascorbate mals (46). In plants, cellular L-ascorbic levels are increased in content after exposure to high-light resulted in higher GDP- response to environmental stresses such as high light (1, 51), L-galactose phosphorylase activity (30).

high temperature (52), and exposure to UV radiation (53, 54) The ascorbate-glutathione cycle is the major H2O2 scav- or ozone (55, 56). L-Ascorbic acid plays an important role in enging system in photosynthetic organisms (2, 17, 46) (Fig. . photosynthesis where it acts by scavenging superoxide and 7). In C. reinhardtii the superoxide anion (O2)formedatthe ␣ H2O2,participatesinregenerationof -tocopheryl radicals site of Photosystem I is converted to H2O2 by superoxide ␣ produced by -tochopherol during reduction of lipid per- dismutases MnSOD3 and FeSOD. The H2O2 is reduced to oxyl radicals, and functions as cofactor for violaxanthin de- water by ascorbate in a reaction catalyzed by APX1 (63). epoxidase (1) and prolyl hydroxylases (2, 3). Oxidation of ascorbate produces monodehydroascorbate,

14242 JOURNAL OF BIOLOGICAL CHEMISTRY 41 VOLUME 287•NUMBER 17•APRIL 20, 2012 Ascorbate Biosynthesis and Regulation in Chlamydomonas

FIGURE 7. Putative regulatory sites for ascorbate biosynthesis and ascorbate recycling in C. reinhardtii. Enzymes depicted in red indicate the proteins whose mRNA levels are increased in response to oxidative stress (1 mM H2O2 and 0.1 mM tBuOOH) and those in magenta that show changes in mRNA levels only in response to H2O2 treatment. APX1 (ascorbate peroxidase), MSD3 (Mn superoxide dismutase), MDAR1 (monodehydroascorbate reductase), DHAR1 (dehydroascorbate reductase), GSH1 (␥-glutamylcysteine synthetase), GSH2 (glutathione synthetase), GSHR1/2 (glutathione reductase), GLDH (L-galactono-1,4-lactone dehydrogenase), VTC2 (GDP-L-galactose phosphorylase), ASC (ascorbate), MDA (monodehydroascorbate), and DHA (dehydroascorbate) are indicated.

which either can be reduced to ascorbate by MDAR1 or selenoenzymes, including a thioredoxin reductase prototype can spontaneously disproportionate to dehydroascorbate. (69, 70). DHAR1 uses GSH to regenerate ascorbate from dehydroascorbate and GSHR1/2 regenerates GSH from GSSG. It Acknowledgments—We thank Prof. Shinichi Kitamura (Osaka Pre- has been demonstrated that overexpression of A. thaliana or fecture University) for kindly providing GDP-L-galactose and Annie tomato (Lycopersicon esculentum Mill) monodehydroascor- Shin for help with the expression and purification of recombinant VTC2. bate reductase (64, 65) results in increased ascorbate levels. Similarly, overexpression of dehydroascorbate reductase had the same effect in enhancing the plant vitamin C con- REFERENCES tent, conferring increased tolerance to oxidative stress (66, 1. Smirnoff, N. (2000) Ascorbate biosynthesis and function in photoprotec- 67). In this study, oxidatively stressed C. reinhardtii cells tion. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355, 1455–1464 showed enhanced mRNA abundance for all transcripts 2. Gill, S. S., and Tuteja, N. (2010) Reactive oxygen species and antioxidant encoding the ascorbate-glutathione components. An inter- machinery in abiotic stress tolerance in crop plants. Plant Physiol. Biochem. 48, 909–930 esting observation was that exposure of C. reinhardtii cells to 3. Smirnoff, N., and Wheeler, G. L. (2000) Ascorbic acid in plants. Biosyn- tBuOOH did not induce glutathione synthesis (GSH1)or thesis and function. Crit. Rev. Biochem. Mol. Biol. 35, 291–314 GSSG reduction (GSHR1), suggesting that under these con- 4. Wheeler, G. L., Jones, M. A., and Smirnoff, N. (1998) The biosynthetic ditions, another (glutathione-independent) mechanism is pathway of vitamin C in higher plants. Nature 393, 365–369 required for dehydroascorbate reduction. A similar mecha- 5. Linster, C. L., Gomez, T. A., Christensen, K. C., Adler, L. N., Young, B. D., Brenner, C., and Clarke, S. G. (2007) Arabidopsis VTC2 encodes a GDP- nism has been observed to be functional in rat liver where a L-galactose phosphorylase, the last unknown enzyme in the Smirnoff- selenoenzyme thioredoxin reductase reduces dehydroascor- Wheeler pathway to ascorbic acid in plants. J. Biol. Chem. 282, bate to ascorbate (68). C. reinhardtii,unlikelandplants,has 18879–18885

APRIL 20, 2012•VOLUME 287•NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 14243 42 Ascorbate Biosynthesis and Regulation in Chlamydomonas

6. Laing, W. A., Wright, M. A., Cooney, J., and Bulley, S. M. (2007) The 30. Dowdle, J., Ishikawa, T., Gatzek, S., Rolinski, S., and Smirnoff, N. (2007) missing step of the L-galactose pathway of ascorbate biosynthesis in plants, Two genes in Arabidopsis thaliana encoding GDP-L-galactose phospho- an L-galactose guanyltransferase, increases leaf ascorbate content. Proc. rylase are required for ascorbate biosynthesis and seedling viability. Plant Natl. Acad. Sci. U.S.A. 104, 9534–9539 J. 52, 673–689 7. Hancock, R. D., and Viola, R. (2005) Biosynthesis and catabolism in L- 31. Linster, C. L., Adler, L. N., Webb, K., Christensen, K. C., Brenner, C., and ascorbic acid in plants. Crit. Rev. Plant Sci. 24, 167–188 Clarke, S. G. (2008) A second GDP-L-galactose phosphorylase in Arabi- 8. Linster, C. L., and Clarke, S. G. (2008) L-Ascorbate biosynthesis in higher dopsis en route to vitamin C. Covalent intermediate and substrate require- plants. The role of VTC2. Trends Plant Sci. 13, 567–573 ments for the conserved reaction. J. Biol. Chem. 283, 18483–18492 9. Smirnoff, N., Conklin, P. L., and Loewus, F. A. (2001) Biosynthesis of 32. Ishikawa, T., and Shigeoka, S. (2008) Recent advances in ascorbate biosyn- ascorbic acid in plants. A renaissance. Annu. Rev. Plant Physiol. Plant Mol. thesis and the physiological significance of ascorbate peroxidase in pho- Biol. 52, 437–467 tosynthesizing organisms. Biosci. Biotechnol. Biochem. 72, 1143–1154 10. Di Matteo, A., Hancock, R. D., Ross, H. A., Frusciante, L., and Viola, R. 33. Wolucka, B. A., and Van Montagu, M. (2003) GDP-mannose 3Ј,5Ј-epi- (2003) Characterization of Chlorella pyrenoidosa L-ascorbic acid accumu- merase forms GDP-L-gulose, a putative intermediate for the de novo bio- lating mutants. Identification of an enhanced biosynthetic enzyme activity synthesis of vitamin C in plants. J. Biol. Chem. 278, 47483–47490 and cloning of the putative gene from Arabidopsis thaliana. Comp. 34. Radzio, J. A., Lorence, A., Chevone, B. I., and Nessler, C. L. (2003) L-Gu- Biochem. Physiol. A 134, S155 lono-1,4-lactone oxidase expression rescues vitamin C-deficient Arabi- 11. Renstrøm, B., Grün, M., and Loewus, F. (1982) Biosynthesis of L-ascorbic dopsis (vtc) mutants. Plant Mol. Biol. 53, 837–844 acid in Chlorella pyrenoidosa. Plant Cell Lett. 28, 299–305 35. Linster, C. L., and Van Schaftingen, E. (2007) Vitamin C. Biosynthesis, 12. Running, J. A., Burlingame, R. P., and Berry, A. (2003) The pathway of recycling, and degradation in mammals. FEBS J. 274, 1–22 L-ascorbic acid biosynthesis in the colourless microalga Prototheca mori- 36. Kondo, Y., Inai, Y., Sato, Y., Handa, S., Kubo, S., Shimokado, K., Goto, S., formis. J. Exp. Bot. 54, 1841–1849 Nishikimi, M., Maruyama, N., and Ishigami, A. (2006) Senescence marker 13. Helsper, J. P., Kagan, L., Hilby, C. L., Maynard, T. M., and Loewus, F. A. protein 30 functions as gluconolactonase in L-ascorbic acid biosynthesis, (1982) L-Ascorbic acid biosynthesis in Ochromonas danica. Plant Physiol. and its knockout mice are prone to scurvy. Proc. Natl. Acad. Sci. U.S.A. 69, 465–468 103, 5723–5728 14. Shigeoka, S., Nakano, Y., and Kitaoka, S. (1979) The biosynthetic pathway 37. Agius, F., González-Lamothe, R., Caballero, J. L., Muñoz-Blanco, J., Bo- of L-ascorbic acid in Euglena gracilis Z. J. Nutr. Sci. Vitaminol. 25, tella, M. A., and Valpuesta, V. (2003) Engineering increased vitamin C 299–307 levels in plants by overexpression of a D-galacturonic acid reductase. Nat. 15. Grün, M., and Loewus, F. A. (1984) L-Ascorbic acid biosynthesis in the Biotechnol. 21, 177–181 euryhaline diatom Cyclotella cryptica. Planta 160, 6–11 38. Ishikawa, T., Nishikawa, H., Gao, Y., Sawa, Y., Shibata, H., Yabuta, Y., 16. Brenner, C. (2007) in Encyclopedia of Life Sciences, pp. 1–6, John Wiley & Maruta, T., and Shigeoka, S. (2008) The pathway via D-galacturonate/L- Sons, Ltd., Chichester, UK galactonate is significant for ascorbate biosynthesis in Euglena gracilis. 17. Asada, K. (1999) The water-water cycle in chloroplasts. Scavenging of Identification and functional characterization of aldonolactonase. J. Biol. active oxygens and dissipation of excess photons. Annu. Rev. Plant Physiol. Chem. 283, 31133–31141 Plant Mol. Biol. 50, 601–639 39. Müller-Moulé, P. (2008) An expression analysis of the ascorbate biosyn- 18. Watanabe, K., Suzuki, K., and Kitamura, S. (2006) Characterization of a thesis enzyme VTC2. Plant Mol. Biol. 68, 31–41 GDP-D-mannose 3Љ,5Љ-epimerase from rice. Phytochemistry 67, 338–346 40. Allen, M. D., Kropat, J., Tottey, S., Del Campo, J. A., and Merchant, S. S. 19. Harris, E. H. (2009) The Chlamydomonas Sourcebook, A Comprehensive (2007) Manganese deficiency in Chlamydomonas results in loss of photo- Guide to Biology and Laboratory Use, 2nd Ed., Academic Press, San Diego system II and MnSOD function, sensitivity to peroxides, and secondary 20. Fox, J. D., and Waugh, D. S. (2003) Maltose-binding protein as a solubility phosphorus and iron deficiency. Plant Physiol. 143, 263–277 enhancer. Methods Mol. Biol. 205, 99–117 41. Fischer, B. B., Eggen, R. I., and Niyogi, K. K. (2010) Characterization of 21. Quinn, J. M., and Merchant, S. (1998) Copper-responsive gene expression single oxygen-accumulating mutants isolated in a screen for altered oxi- during adaptation to copper deficiency. Methods Enzymol. 297, 263–279 dative stress response in Chlamydomonas reinhardtii. BMC Plant Biol. 10, 22. Allen, M. D., del Campo, J. A., Kropat, J., and Merchant, S. S. (2007) FEA1, 279 FEA2, and FRE1, encoding two homologous secreted proteins and a can- 42. Ledford, H. K., Chin, B. L., and Niyogi, K. K. (2007) Acclimation to singlet didate ferrireductase, are expressed coordinately with FOX1 and FTR1 in oxygen stress in Chlamydomonas reinhardtii. Eukaryot. Cell 6, 919–930 iron-deficient Chlamydomonas reinhardtii. Eukaryot. Cell 6, 1841–1852 43. Long, J. C., and Merchant, S. S. (2008) Photooxidative stress impacts the 23. Schloss, J. A. (1990) A Chlamydomonas gene encodes a G protein ␤ sub- expression of genes encoding iron metabolism components in Chlamy- unit-like polypeptide. Mol. Gen. Genet. 221, 443–452 domonas. Photochem. Photobiol. 84, 1395–1403 24. Craigie, R. A., and Cavalier-Smith, T. (1982) Cell volume and the control 44. Glascott, P. A., Jr., Gilfor, E., and Farber, J. L. (1995) Relationship of the of the Chlamydomonas cell cycle. J. Cell Sci. 54, 173–191 metabolism of vitamins C and E in cultured hepatocytes treated with 25. Langmead, B., Trapnell, C., Pop, M., and Salzberg, S. L. (2009) Ultrafast tert-butyl hydroperoxide. Mol. Pharmacol. 48, 80–88 and memory-efficient alignment of short DNA sequences to the human 45. Daskalopoulos, R., Korcok, J., Tao, L., and Wilson, J. X. (2002) Accumula- genome. Genome Biol. 10, R25 tion of intracellular ascorbate from dehydroascorbic acid by astrocytes is 26. Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., and Wold, B. decreased after oxidative stress and restored by propofol. Glia 39, (2008) Mapping and quantifying mammalian transcriptomes by RNA- 124–132 Seq. Nat. Methods 5, 621–628 46. Foyer, C. H., and Noctor, G. (2011) Ascorbate and glutathione. The heart 27. Castruita, M., Casero, D., Karpowicz, S. J., Kropat, J., Vieler, A., Hsieh, S. I., of the redox hub. Plant Physiol. 155, 2–18 Yan, W., Cokus, S., Loo, J. A., Benning, C., Pellegrini, M., and Merchant, 47. Wolucka, B. A., and Van Montagu, M. (2007) The VTC2 cycle and the de S. S. (2011) Systems biology approach in Chlamydomonas reveals connec- novo biosynthesis pathways for vitamin C in plants. An opinion. Phyto- tions between copper nutrition and multiple metabolic steps. Plant Cell chemistry 68, 2602–2613 23, 1273–1292 48. Adler, L. N., Gomez, T. A., Clarke, S. G., and Linster, C. L. (2011) A novel 28. Whelan, S., and Goldman, N. (2001) A general empirical model of protein GDP-D-glucose phosphorylase involved in quality control of the nucleo- evolution derived from multiple protein families using a maximum-like- side diphosphate sugar pool in Caenorhabditis elegans and mammals. lihood approach. Mol. Biol. Evol. 18, 691–699 J. Biol. Chem. 286, 21511–21523 29. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., and Kumar, S. 49. Guranowski, A., Wojdyła, A. M., Zimny, J., Wypijewska, A., Kowalska, J., (2011) MEGA5. Molecular evolutionary genetics analysis using maximum Jemielity, J., Davis, R. E., and Bieganowski, P. (2010) Dual activity of certain likelihood, evolutionary distance, and maximum parsimony methods. HIT-proteins. A. thaliana Hint4 and C. elegans DcpS act on adenosine Mol. Biol. Evol. 28, 2731–2739 5Ј-phosphosulfate as hydrolases (forming AMP) and as phosphorylases

14244 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 17•APRIL 20, 2012 43 Ascorbate Biosynthesis and Regulation in Chlamydomonas

(forming ADP). FEBS Lett. 584, 93–98 involved in vitamin C biosynthesis. FEBS J. 275, 713–726 50. Mori, S., Shibayama, K., Wachino, J., and Arakawa, Y. (2010) Purification 61. Pineau, B., Layoune, O., Danon, A., and De Paepe, R. (2008) L-Galactono- -and molecular characterization of a novel diadenosine 5Ј,5ٞ-P(1),P(4)- 1,4-lactone dehydrogenase is required for the accumulation of plant res tetraphosphate phosphorylase from Mycobacterium tuberculosis H37Rv. piratory complex I. J. Biol. Chem. 283, 32500–32505 Protein Expr. Purif. 69, 99–105 62. Drahota, Z., Kriváková, P., Cervinková, Z., Kmonícková, E., Lotková, H., 51. Müller-Moulé, P., Golan, T., and Niyogi, K. K. (2004) Ascorbate-deficient Kucera, O., and Houstek, J. (2005) tert-Butyl hydroperoxide selectively mutants of Arabidopsis grow in high light despite chronic photooxidative inhibits mitochondrial respiratory chain enzymes in isolated rat hepato- stress. Plant Physiol. 134, 1163–1172 cytes. Physiol. Res. 54, 67–72 52. Larkindale, J., Hall, J. D., Knight, M. R., and Vierling, E. (2005) Heat stress 63. Takeda, T., Ishikawa, T., and Shigeoka, S. (1997) Metabolism of hydrogen phenotypes of Arabidopsis mutants implicate multiple signaling pathways peroxide by the scavenging system in Chlamydomonas reinhardtii. in the acquisition of thermotolerance. Plant Physiol. 138, 882–897 Physiol. Plant 99, 49–55 53. Conklin, P. L., Williams, E. H., and Last, R. L. (1996) Environmental stress 64. Eltayeb, A. E., Kawano, N., Badawi, G. H., Kaminaka, H., Sanekata, T., sensitivity of an ascorbic acid-deficient Arabidopsis mutant. Proc. Natl. Shibahara, T., Inanaga, S., and Tanaka, K. (2007) Overexpression of mono- Acad. Sci. U.S.A. 93, 9970–9974 dehydroascorbate reductase in transgenic tobacco confers enhanced tol- 54. Filkowski, J., Kovalchuk, O., and Kovalchuk, I. (2004) Genome stability of erance to ozone, salt and polyethylene glycol stresses. Planta 225, vtc1, tt4, and tt5 Arabidopsis thaliana mutants impaired in protection 1255–1264 against oxidative stress. Plant J. 38, 60–69 65. Li, F., Wu, Q. Y., Sun, Y. L., Wang, L. Y., Yang, X. H., and Meng, Q. W. 55. Chen, Z., and Gallie, D. R. (2005) Increasing tolerance to ozone by elevat- (2010) Overexpression of chloroplastic monodehydroascorbate reductase ing foliar ascorbic acid confers greater protection against ozone than in- enhanced tolerance to temperature and methyl viologen-mediated oxida- creasing avoidance. Plant Physiol. 138, 1673–1689 tive stresses. Physiol. Plant 139, 421–434 56. Sanmartin, M., Drogoudi, P. A., Lyons, T., Pateraki, I., Barnes, J., and 66. Chen, Z., Young, T. E., Ling, J., Chang, S. C., and Gallie, D. R. (2003) Kanellis, A. K. (2003) Overexpression of ascorbate oxidase in the apoplast Increasing vitamin C content of plants through enhanced ascorbate recy- of transgenic tobacco results in altered ascorbate and glutathione redox cling. Proc. Natl. Acad. Sci. U.S.A. 100, 3525–3530 states and increased sensitivity to ozone. Planta 216, 918–928 67. Wang, Z., Xiao, Y., Chen, W., Tang, K., and Zhang, L. (2010) Increased 57. Alía, M., Ramos, S., Mateos, R., Bravo, L., and Goya, L. (2005) Response of vitamin C content accompanied by an enhanced recycling pathway con- the antioxidant defense system to tert-butyl hydroperoxide and hydrogen fers oxidative stress tolerance in Arabidopsis. J. Integr. Plant Biol. 52, peroxide in a human hepatoma cell line (HepG2). J. Biochem. Mol. Toxicol. 400–409 19, 119–128 68. May, J. M., Mendiratta, S., Hill, K. E., and Burk, R. F. (1997) Reduction of 58. Zsigmond, L., Tomasskovics, B., Deák, V., Rigó, G., Szabados, L., Bánhegyi, dehydroascorbate to ascorbate by the selenoenzyme thioredoxin reduc- G., and Szarka, A. (2011) Enhanced activity of galactono-1,4-lactone de- tase. J. Biol. Chem. 272, 22607–22610 hydrogenase and ascorbate-glutathione cycle in mitochondria from com- 69. Novoselov, S. V., and Gladyshev, V. N. (2003) Non-animal origin of animal plex III-deficient Arabidopsis. Plant Physiol. Biochem. 49, 809–815 thioredoxin reductases. Implications for selenocysteine evolution and 59. Yabuta, Y., Maruta, T., Nakamura, A., Mieda, T., Yoshimura, K., Ishikawa, evolution of protein function through carboxyl-terminal extensions. Pro- T., and Shigeoka, S. (2008) Conversion of L-galactono-1,4-lactone to L- tein Sci. 12, 372–378 ascorbate is regulated by the photosynthetic electron transport chain in 70. Novoselov, S. V., Rao, M., Onoshko, N. V., Zhi, H., Kryukov, G. V., Xiang, Arabidopsis. Biosci. Biotechnol. Biochem. 72, 2598–2607 Y., Weeks, D. P., Hatfield, D. L., and Gladyshev, V. N. (2002) Selenopro- 60. Leferink, N. G., van den Berg, W. A., and van Berkel, W. J. (2008) L-Galac- teins and selenocysteine insertion system in the model plant cell system, tono-␥-lactone dehydrogenase from Arabidopsis thaliana, a flavoprotein Chlamydomonas reinhardtii. EMBO J. 21, 3681–3693

APRIL 20, 2012•VOLUME 287•NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 14245 44

!"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

!"##$%&%'()$*&)(%+,)$!*-.+/* " #$%&'(")*")+,-&(,./"0(1/00")2"&0')13&(/"3,)042(5/0,0",2"657&$4-)$)2&0".,&"1/897&(,)2")*"(5/"5637*8/2/" /2')-,28"&":;<=>=8&7&'()0/"%5)0%5)147&0/" " %0123*,4*"56789:*$7;9<*'4*)=<25:*&4*>0=<2?*#912:*@95A<2*$4*$73B;25:*&95C*)4*)5D731:*>9E7=*@9B25A:* &9;;2A*#2<<215737:*!9D22F9*!4*&258F93;*93=*!;2E23*G4*@<95C2"

!"##$%&%'()$* -,G"+%* !H4* #5AIAB2=* D7AB?3;F2;78* I9;FJ9?B* KA5* 9B8A5D9;2* D7AB?3;F2B7B* 73* F71F25*I<93;B*93=*9<1924*38*%)+-.'%/(++")1(5)7)80")*"(5/"$&?)1"&0')13&(/"3,)042(5/(,'"1)9(/"/2@4$/0"A>= 8&7&'()0/" %&(5B&4C" &1/" -/%,'(/-" ,2" 1/-D" E<#!" A$&22)0/=F=%5)0%5&(/" ,0)$/1&0/CG" 61/!HD8FIFJKK" A%5)0%5)$&22)$9(&0/CG" :E=$&22)0/" %41)%5)0%5)147&0/CG" LMN!" A098&1" 29'7/)(,-/" /%,$/1&0/CG" OP6I" A:;<=>=8&7&'()0/" %5)0%5)147&0/CG" 61/!QD8RJSQRK" A>=8&7&'()0/=!=<" %5)0%5&(&0/CG" 61/!HD8FQKHKK" A>=8&7&'()0/" -/54-1)8/2&0/C" &2-" :>;T" A>=8&7&'()2)=!GH=7&'()2/" -/54-1)8/2&0/CD" U7(/12&(,./"&0')13&(/"3,)042(5/(,'"%&(5B&40"5&./"3//2"%1)%)0/-"A,2"379/CG"39("$&24"1/&'(,)2"0(/%0"5&./" /,(5/1"2)("3//2",-/2(,*,/-"AV9/0(,)2"$&1W0C")1"(5/"38*%)+-.'%/(++"8/2)$/"-)/0"2)("%)00/00")1(5)7)80")*" %1/.,)9074"'5&1&'(/1,@/-"8/2/0"A5)$)7)80"&1/"-/%,'(/-",2"$&8/2(&CD"

45!" !"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

!!!!!!!!!!!!!!!!!!!!!!!!!!!"!!!!!!!##$!!!!!!!!!"!!!!!!!#%$!!!!!!!!!"!!!!!!!#&$!!!!!!!!!"!!!!!!!#'$!!!!!!!!!"!!!!!!!($$!!!!!!! )*+,-.(//#!!!!!!0!1111111111111111111111123454678397.:;<969=;245:>)<[email protected]<9743@<3:[email protected]>.A6;79=3):=:4:::@<118B739!0!C%C! -+D,-C$A(!!!!!!!0!11111111111111111111111B67=.=3=?74)744;94799?7:>7=;4<61=;A=6@9=1-216:-9).46)932<9=B7:.79::4@2?A411=B!0!C#'! ?EF,-C/GHI/'!!!!0!11111111111111111111111>@:=;@.B@);9@<3@A);;9-3;>4<:5791.9A:6:9:79<187):A;;1.9A:6)58;[email protected];.B8.658);37A):B9B3;>7):A;.1.9A:6)588531845B7.84465;7):A<:1.9A:6)58;[email protected];<9=7.:?1944:@87A:5)45!0!C#R! ;PQ,58-/!!!!!!!!0!11111111111111111111111=9)[email protected]>7):A<:1.9A:6)58;[email protected];<9=7.:?1944:@87A:5)45!0!C#C! -H+,58-#!!!!!!!!0!7A5:2;::..;.;;;;;8;53;;4;;;)B..;75@.:3997.552;9>7):;):1.2A:6)53<-718:59@.@;.A5;<9=7.:;1844:@87A3;):5!0!#N&! 5GDKF,/'&&'!!!!!0!1111111111111111111111111111.45975@3:39973;9289>7):;):1.9A:6)589-@18:[email protected]:.AB;<9=7.:;1844:@87589):5!0!!N/! -QD=-&%;,/RC(N!!0!))7;539).1111111111111133)3.;3;<55@<:359);999.7>7):;7;1.9A:6)58;-@[email protected];<-=7.:91344:@87A:5)95!0!CN&! -GK,-C&N,C('/CC!0!1111111111111111111111111111;@33)3.;8399)8599;7>7):;7<1.9A:6)58;-@18455@.;6.A5;.?5995@7@;:<9@<:99@.;9:!0!C'R! 2JKLM=(,&%/'#!!!0!5<;A)..4A=A84;)4;7B9A;A7:11111.):;2433;1116=3;48B11111111111733@=5129B=53@B76.?599-@:58)-9@<:B3@7999!0!#(N! OEP:--'$N,/'//'!0!-<=A)@44A=A84;)94759A3A849;137;)4=53:3B111A74);;511111111111)73@85159B=53@B76.?599-@:58)29@<@6)3)9B=:4-7@91111111111111111111111111?)5@=565B:@@5>B-!0!#&R! ?EF,-C/GHI/'!!!!0!;.B7;[email protected]:5.A=39..9;35=!9!9!.669;?:9@57<;@37@9111111111111111111111111111)@..?9?99<)9@;!0!#/#! 2JKLM=(,&%/'#!!!0!@@962;;73:=@611111111A:5.6=39.;6;8B=!9!A";66927;A@B7:;<88:9@<:5111111111111111111644:?:?.5;5=<58.6@5!0!(C(! OEP:--'$N,/'//'!0!@;962;;73:=@611111111A:5.6=39.;6;8B=!9!A";66927;A@B7:;38;7B11111111111111111111111A3.8?.7-85635=.6@5!0!#N&! ;PQ,58-#!!!!!!!!0!9;5?2;;7;;=@611111111A:9.6=39.;A;8B=!9!A";669;2@A@974;@844B11111111111111111111111188853.54B379936@5!0!#R/! ;PQ,58-/!!!!!!!!0!9;9<2;;7;)=@611111111A:9.6=39.;A;8B=!9!A";669;2

46 !"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

!!!!!!!!!!!!!!!!!!!!!!!!!!!"!!!!!!!#$%!!!!!!!!!"!!!!!!!#&%!!!!!!!!!"!!!!!!!#'%!!!!!!!!!"!!!!!!!#(%!!!!!!!!!"!!!!!!!'%%!!!!!!! )*+,-./##$!!!!!!0!12.3-345626)7)28492:4)8;<=;9=>=-:66=?@6;5;18)AA?.:7.BBBBBBBB63:45459A72<-46:);BBBBBBBBBBB92564.5-2;9!0!/#$! -+C,-D%5/!!!!!!!0!?BBB93458;)7;)B6<285:=63<:-42@;127689@6;5;9A98?3A11.92:22)AA.2A?8;4B1-B<45?A=6=3.9:??BBBB7656?9962;9!0!/#E! @FG,-D#HIJ#(!!!!0!?BBB.5;5<1A.?.1);27;37A4-A91)<;1)@239@6;541A.9??.:17?779BB;3.4A43;>9A:B775.3:)BB.2BBBBBBBB95649;-2;9!0!//'! =KLMN6/,'&#($!!!0!ABBB-;-52BBA:)91529;9);;.69-2A;82A63?56;;39)@BBBBBBBBBBBBBBB.9A45;3?84B57@A49:BB.23?2)4991.46?945237!0!/O%! PFQA--(%O,#(##(!0!ABBB-;-52BB4.)B7522;91;3.2--4:=87963?563;39)@BBBBBBBBBBBBBBB.9A45;3?84B573A396BB6:[email protected]?945237!0!/E$! 9QR,41-$!!!!!!!!0!ABBB7;;52BB..7B7=82;7)147)--4-;86663?563;37)-BBBBBBBBBBBBBBB.A835;=?8-B<92:89;BB.247?24;21846?94>237!0!/#D! 9QR,41-#!!!!!!!!0!ABBB.;;42BB..6B13:);9)147)974-;86663?563;37)7BBBBBBBBBBBBBBB.:A35;;?8-B<92:89;BB.24771;;)1846?94>2=7!0!/&(! -I+,41-$!!!!!!!!0!ABBB75442929.9B9;29499449A99)9=89968?563397).BBBBBBBBBBBBBBB.AA45;5?8-B<92A899BB.24?22;;)1.46?975239!0!#E/! 4HCLG,#(''(!!!!!0!6BBB95432BBBBBB9;17;217;<)-92A;89968?56;;37)9BBBBBBBBBBBBBBB.AA45;5?8-B592A899BB.;3??2;;21.46?995239!0!//'! -RC6-'&9,#ED/O!!0!ABBB7;452BB9.)B7;A2492;4.19-8A;19964?@6;539)-BBBBBBBBBBBBBBB.9A35;5?6-B592::9ABB.83?2)4;2184)?99>239!0!&&#! -HL,-D'O,D/(#DD!0!-BBB.;452BB97)B73)2;9744.29-8A;79963?@6;544)-BBBBBBBBBBBBBBB.8A35;5?69B59A9:9:BB.;4?2);;)784)?995237!0!/$&! !!!!!!!!!!!!!!!!!!!!!!!!!!!"!!!!!!!'$%!!!!!!!!!"!!!!!!!'&%!!!!!!!!!"!!!!!!!''%!!!!!!!!!"!!!!!!!'(%!!!!!!!!!"!!!!!!!E%%!!!!!!! )*+,-./##$!!!!!!0!.<3?;?)9):=26;12;84;5A3A131.69?:94<2BBBB8317=42.82)8?;13BBBBBBBBBBBBBBBB!0!&%(! -+C,-D%5/!!!!!!!0!.6;17<13A5527962B8743A3322297;)))15A7;-5);9)4;3.A74.17A?BBBBBBBBBBBBBBBBBBB8);)1;9.;17?23)2;A)75875=!0!&/E! @FG,-D#HIJ#(!!!!0!.@;?4:178)577;12B99949;38B)-A;??7892BBBB)4899;49;=78228BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB!0!/(#! =KLMN6/,'&#($!!!0!.@;;<:88))<)9-18B)995:=;9B-97;.2225A;;9B297;722A5:BBBB2419;95293.-76BBBBBBBBBBBBBBBBBBBBB822);[email protected]:76A!0!&$&! 9QR,41-#!!!!!!!!0!.@=4;:A:2)<2.972B2:9>A;;9B247;722A5ABBBB2461=35)93.577BBBBBBBBBBBBBBBBBBBBB@2222222;22867=6..BBB7513!0!&D(! -I+,41-$!!!!!!!!0!.@;4;:A92)59;9)2B9>99A;;7B.47;722A5=BBBB2496=-5.779BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB!0!'D(! 4HCLG,#(''(!!!!!0!.@;4;:A18)<12912B);9=8;;9B897;722A5=BBBB7;9A;-5..99.9)BBBBBBBBBBBBBBBBBBBBB3-.99;8BBBBBBBBBBBBBBBBBB!0!/OD! -RC6-'&9,#ED/O!!0!.@34;:A88)<29478B279>A;;2B597-722A59BBBB249A;9;2.;3BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB!0!&O%! -HL,-D'O,D/(#DD!0!.@33<:A78)<2@478B)94>A;;7B<97<729)4;BBBB2=9A:4;94BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB!0!/'E!

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

" " "

!"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

" "

! "#$$%&'&()*%! +,-#.&! "/0! !"#$123456! 789:;<=8! <>! ?:98! >4@268! ! D364398>! 4C@A!C5=9:E8A!789:;<=80"#$"%&'()*"++!,-".(/"0'1.*")*"2#3"&1","4"567"89::";<=5">')1'"&1"(??)&)1*"1@" A,B,"C5";D$E"F9/G:&/"@1'"&H'99"I)1:10)8(:"'9>:)8(&9/"('9"/H1.*E"J$"+9::/".9'9"0'1.*"(/"@1'"C#$">')1'"&1" (??)&)1*" 1@" &JGBBA" (&" 81*89*&'(&)1*/" 1@" 9)&H9'" 6E5" ;D" CŶK" Ƒ$" 1'" 6E," ;D" CƔK" ż$E" F9/G:&/" @1'" &.1" I)1:10)8(:"'9>:)8(&9/"('9"/H1.*E" ! ! ! ! ! ! ! ! !

48!"

!"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

! "#$$%&'&()*%!)*+%&!",-!&./0123!45!672!"189.455:;722<29!=86>[email protected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í" #/%*5/%83))%8,&3*'" #::" M$'N>K2SOSTPP" VOQ" >TX>!" S>S>T" VOPSN" XTT>X" NPT!OY" 7'*" ZD#1D183))%*'"57$%5/%*5/%$743*'" LIMN" M$'NSK2SVOYPPWZ:#N" S!Q" NXXNPO" OSYV>" !T!XS" TST" YNTY>" 7'*3" ZD#1D183))%*'"X[9![1'5(8'$3*'" Z:\N" M$'PNK2PNTO!PWH]\N" V>Q" N>PX!P" !O>>N" TOSYX" NYVPN" VVNNO" 7'*" LIMO" M$'NXK2!YYN!PWLIMO" >SQ" !VNXT" NXY!N" S>!YO" XVPO" !YSSY" í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í"NV>!T" í" í"TSXVV")%" Z,4%)%43-&%)3*'"A?'((1"*-$%@)0+#1"C" í" í""NXSVXP">V>YP"YYVNX">NV!N"í"7'*" 3>D2!A>67B>0" GINZ!TTSP" M$'POK2NXPT!PW"0M;OY" XSQ" XO>O!" !SSPT" YSXOT" OYSVP" TONVP" D12343-&,$%)3&'"$'@,-&3*'" GINZ!TT!P" M$'NSK2STOYPPWGÛ>" X!Q" NXY!SP" >Y!TO" TYOTX" YYY>!" NPYX!!" )%" A6%'0'%+'*7*'-'-'""'C" GIOZXVVTP" M$'NPK2>XOTPPWGÛX" XXQ" NTOXY" N>SNO" YP>>V" !!TY" !!NYO" GIOZXVVVP" M$'NTK2V!O>!PWGÛN" XNQ" N>ONPN" !SSNP" Y>PON" XV!ON" YP!OS" 34@%)%43-&%)3*'"A8102)-'*0%'#+2+"C" í" í""NXSVXP">V>YP"YYVNX">NV!N"í"7'*" " "

!"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

!"##$%&%'()$*()+$%*!,-*./012345/6*/7*08')*29:6;26<=5*7/3*>=6=5*=6*=6?@0=5*/7*AB=*$C>2D2

0I?JK.L" 1%7$.)&"=>" 1CD$7E7:." H.&."I#:." >.86)&." 0123#"M76-"(C#&N." ?,@A1J>J:#&&7'."*D%7*C7'*C7%D6#'." QS" !!" @AO" @3?* H>1JN#6#($7'.UN6,(7'.J@J*C7'*C#$." BA" @OA" @GQ" !J#%#+)&7&7J@/QJ6#($7&."79)-#'." " " " " " *

!"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

!"##$%&%'()$*()+$%*!,-*(./012.345*/6708/0291*:;.*<9091*902;830<*90=>?91*;:*5@9*/12;.6/59A2B9*/:59.*9C4;17.9*5;*D* ?&*E G *:;.*,H*/08*IH*?307591-*#$%#&'(%)*$"#+,&-#&(./"0123"4%.#-'"*.%"5)67+#'."78".97&":7-.6"*.%":)66)7&":#**.-"%.#-';<" F F " 0H?IJ.K" 1%7$.)&"=>" 1CD$7E7:." G.&."H#:." >.86)&." 0123#"L76-"(C#&M." ?,@A.CD-%7#'(7%+#$."%.-,($#'." Q9=7* G6,$#$C)7&."'D&$C.$#'." SO" !A" @NN" @9=?* G6,$#$C)7&."'D&$C.$#'." @B" R9=<7* G6,$#$C)7&."%.-,($#'." BB" B!" SN" @9=

51 !"#$%&'()*&+$",-(.)"+"*'-/*%)012'(+$-*+-*3.2'4,/$4$-'"*

!"##$%&%'()$* ()+$%* !,-* #./01.2* 213415612* 4217* 89.* 6:95/5;* 98* <(=>#" $%&'#('" )*+" $%&'#,-'!" ./012/3" 42/2" 532+" 67" 8&9" )1.:0;<" 6=2" $%&'" >7+0*?" 32@52*>2" )*+" 67" 0*32/6" 6=2" %A$" ./762)32" >:2)B)?2" 3062" )*+" )" &C62/10*):" D03-C6)?#" 8AC'EE" )*+" 8AC'E!" ./012/3" 42/2" 532+" 67" 0*6/7+5>2" ,66FG")*+",66F'"/2>71H0*)607*"30623#")8/012/"32@52*>23")/2"4/0662*"IJ"67"KJ#" "

'?01* #./01.*!1341561?* $%&'#('#L7/4)/+" !"!""##$!$"#$$##"!!!$!"$$#$#$$#"!!#!#$!#% $%&'#,-'!#92B2/32" "$$"!$!"$!"$!!$!!$!"$!"!#!#$!!"!##!""!% 8AC'EE#L7/4)/+" !!!!"#""!$$$!$"#""""""!#"!!#$#!!"!""##$!$"#$$##"!% 8AC'E!#"92B2/32" !!!!"##"#$$$!$"#""!"""!#$!!!$$"$$"!$!"$!"$!!$!!$!"$!% * * !"##$%&%'()$*()+$%*!@-*#./01.2*213415612*4217*89.*.1?:AB/01*#=C#")8/7620*"M(3">7//23.7*+" 67"6=2",5?53653"GN#'")**76)607*"7;"6=2"B2/307*"O")3321H:<"PLQOR#H8/012/"32@52*>23")/2"4/0662*"IJ"67"KJ#" "

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

52 !"

CHAPTER 5

Ascorbate is Present in Caenorhabditis elegans but its Origin is Unclear

53 INTRODUCTION

L-Ascorbate (AsA) is a robust antioxidant as well as an essential cofactor in many enzymatic reactions including collagen hydroxylation [1]. While plants and most vertebrate animals have the ability to synthesize it from glucose, primates (including humans) as well as a few other animal species have lost this ability due to deficiency in the last enzyme of the biosynthetic pathway [2][3][4].

In the animal pathway for vitamin C synthesis, D-glucuronate, derived from myo-inositol or UDP-glucuronate, is reduced to L-gulonate. The latter is then converted to L-gulono-1,4- lactone followed by oxidation by L-gulonolactone oxidase to L-ascorbate [5]. In the plant biosynthesis pathway, D-glucose is converted to D-mannose 1-phosphate in four enzymatic steps followed by formation of GDP-D-mannose that is successively converted to GDP-L-galactose,

L-galactose 1-phosphate, L-galactose, L-galactono-1,4-lactone, and finally to L-ascorbate [6].

The plant 10-step pathway from glucose to AsA was proposed over a decade ago by Smirnoff and Wheeler [7]. However, the identification of all the enzymes in the pathway was completed only a few years ago, with the characterization of VTC2 [8], the enzyme that catalyzes a rate- limiting reaction in the pathway by phosphorolysis of GDP-L-galactose to L-galactose-1-P. The identification of VTC2 in Arabidopsis led to the discovery of homologs of this enzyme in other plants, vertebrates and invertebrates, including the product of the C10F3.4 gene in the nematode worm Caenorhabditis elegans. This finding suggested that worms might synthesize vitamin C by the plant pathway.

While the routes for AsA synthesis in plants and vertebrates are well characterized, as of today, there are no published results that indicate the ability of invertebrates, and specifically C.

54 elegans, to synthesize this compound. Moreover, there are no published data on AsA levels in this nematode. In this study, I report for the first time the presence of vitamin C in C. elegans.

Ascorbate levels were quantified in C. elegans extracts and the role of C10F3.4 as a possible enzyme for vitamin C biosynthesis in worms was investigated.

EXPERIMENTAL PROCEDURES

C. elegans Strains and Culture Conditions

Standard procedures used to maintain C. elegans strains were adapted from Sulston and

Hodgkin [9]. M9 and S media were prepared as described by Lewis and Fleming [10]. The N2 strain was obtained from the Caenorhabditis Genetics Center (St. Paul, MN). The C10F3.4 mutant strain (tm2679) lacking coding sequences between exons 5 and 8 of the C10F3.4 gene

(Worm-Base) was provided by the Caenorhabditis elegans Core Facility at Tokyo Women’s

Medical University. C. elegans strains were cultured at 20 °C according to standard procedures

[11]. All strains were maintained on nematode growth medium plates streaked with Escherichia coli OP50 [9]. Liquid cultures of these worm strains were grown at 20 °C in S medium supplemented with concentrated E. coli OP50. Concentrated OP50 was prepared from overnight cultures of OP50 grown at 37 °C in LB medium. Following the overnight incubation, OP50 cells were spun at 4,500 g for 10 min at 4 °C. The pellet was resuspended in M9 medium to about

100 fold concentrated cell suspension and frozen at -20 °C until its addition to the S-medium.

Worms grown in liquid cultures were harvested, washed, and sucrose-floated as described

55 previously [12] except that M9 medium was used instead of 0.1 M NaCl as a resuspension and washing solution.

AsA Extraction from C. elegans

Worms, harvested as described above, were resuspended in two ml of extraction solution per gram of wet weight pellet of animals spun at 2000 g for 3 min at 4 °C. The extraction solution consisted of 2% metaphosphoric acid, 2 mM sodium EDTA (pH 8.0), and 5 mM DTT.

The worm suspensions were freeze-dropped in liquid nitrogen and ground to a fine powder in a pre-cooled mortar and pestle. The homogenates were allowed to thaw on ice and centrifuged at

21,000 g for 30 min at 4° C. The supernatants from this spin (defined here as the "extract") were stored at -80 °C until assayed for vitamin C content.

AsA Quantitation by HPLC

Vitamin C content was measured by reversed-phase HPLC on an Econosphere C-18 column (5 µm bead size, 4.6 x 250 mm; Alltech Associates, Deerfield, IL) using a Hewlett

Packard Series II 1090 liquid chromatograph. A mobile-phase gradient of 0 to 40% acetonitrile in 20 mM triethylammonium acetate (TEAA), pH 6.0 was used at a flow rate of 1 mL min−1.

The injection volume was 50-100 µL. Ascorbic acid was detected by monitoring the absorbance at 254 nm. The ascorbic acid peak was identified by comparison with the elution time of an AsA standard and by demonstrating decrease of the peak area after treatment of the samples with AsA oxidase. This treatment was performed by adding 2 U of ascorbate oxidase from Cucurbita sp.

56 (EC 1.10.3.3; Sigma Life Science product A0157, one U will oxidize one micromole of L- ascorbate to dehydroascorbate per min at pH 5.6 at 25 °C) to 60 µL of the extract in a final concentration of 0.12 M monosodium citrate for 1 h at 4 °C. The final pH of the reaction was about 5.6 using pH paper. The differences in the peak areas measured before and after addition of AsA oxidase was used to calculate ascorbic acid levels based on a standard curve.

Preparation of the E. coli Extracts and LB for AsA Analysis

Concentrated E. coli (see above) was centrifuged at 21,000 g for 15 min at 4 °C. The pellet was resuspended in eight volumes per gram of wet weight cell pellet using extraction solution (2% metaphosphoric acid, 2 mM sodium EDTA (pH 8.0), 5 mM DTT) and subsequently broken by six 10 s sonicator pulses (50% duty; setting 4) on ice with a Sonifier cell disruptor W-

350 (SmithKline Corp). The lysates was centrifuged for 30 min at 13,000 g at 4° C, followed by additional centrifugation of the supernatant for 15 min at 20,000 g at 4° C. The supernatant pH was adjusted to pH 5.6 by adding monosodium citrate to a final concentration of 0.13 M. 160 µL of the E. coli supernatant were either untreated, treated with 4 U of ascorbate oxidase from

Cucurbita sp. as described above or supplemented with 5 µL of 8 mM AsA. 120-128 µL of each

E. coli-containing sample were loaded onto the C-18 HPLC column and analyzed as described below. The LB pH was adjusted to pH 5.6 by adding monosodium citrate to a final concentration of 0.12 M. 40 µL of LB were either untreated, treated with 3 U of ascorbate oxidase from Cucurbita sp. (EC 1.10.3.3) or supplemented with 5 µL of 8 mM AsA. 33-40 µL of each LB sample were loaded onto the C-18 HPLC column and analyzed as described below.

57 AsA Detection and Quantitation by Liquid Chromatography-Electrospray Ionization Mass

Spectrometry (LC-ESI-MS)

The experiments were carried out on an Agilent 1200 HPLC system coupled to a 4000 QTRAP

MS/MS hybrid triple quadrupole/linear ion trap mass spectrometer from Applied Biosystems

(Foster City, CA). The HPLC system was equipped with a PAL autosampler with thermostated tray holders and stack (LEAP Technologies, Carrboro, NC). The turbo electrospray ionization source of the 4000 QTRAP system was operated in the negative ion mode. Q1 and Q3 were operated in unit resolution. All system control, data acquisition, and mass spectral data evaluation were performed using Analyst software version 1.4.2 from Applied Biosystems. A

Luna 3u HILIC 200A column (150 x 4.6 mm, Phenomenex) was used with a HILIC security guard cartridge (4 x 3.0mm, Phenomenex). Mobile phase A consisted of 90% Acetonitrile with

10% 100mM ammonium acetate adjusted to pH 4.7. Mobile phase B consisted of 10% 100mM ammonium acetate adjusted to pH 4.7. A mobile phase gradient was used with a flow rate of 0.6 ml/min for the first 9.5 min and was increased to 0.8ml/min from this point till the end of the run.

The run started with a linear gradient from 100% to 35% A over 9 min followed by a linear gradient from 35 to 100% A over 0.5 min and finally holding for 5.5 min at 100% A. The injection volume was 10 µl. Mass detection was carried out in the negative mode under the following conditions: spray voltage at -4.5kV, curtain gas at 15, and collision gas at the medium setting. The system was tuned using a stock solution of AsA (Sigma Life Sciences) (10 µM) in water. The precursor-to-product ion transitions in multiple reaction monitoring mode used to detect AsA were as follows: m/z 175>87, m/z 175>115.

58 RESULTS

The C. elegans C10F3.4 Gene Product is a Homolog of the Plant GDP-L-Galactose

Phosphorylase Involved in Vitamin C Synthesis.

The Arabidopsis thaliana VTC2 gene encodes a GDP-L-galactose phosphorylase, the enzyme that catalyzes a rate-limiting step in the plant vitamin C biosynthesis pathway (Linster

2007-8). In this reaction, GDP-L-galactose is phosphorolyzed to produce L-galactose-1- phosphate and GDP. Apparent orthologs for the VTC2 gene can be found in other plants, vertebrates and invertebrates, including the nematode worm C. elegans [8].

The C. elegans C10F3.4 protein has a 24% sequence identity with Arabidopsis VTC2 and contains the conserved HIT (histidine triad) motif [8] that characterizes enzymes with a nucleotide hydrolase or a nucleotide transferase activity [13]. This sequence conservation, together with the lack of information concerning AsA content and synthesis in C. elegans encouraged us to measure vitamin C levels in C. elegans as well as to investigate the possible involvement of C10F3.4 in vitamin C synthesis.

AsA levels in C. elegans extracts.

C. elegans strains that were either wild type (N2) or had a deletion mutation for the

C10F3.4 gene (tm2679) [14][15] were cultured, harvested and the levels of AsA in both strains were determined using an HPLC method as described in the "Experimental Procedures" section.

59 These results demonstrate for the first time that C. elegans extracts do contain AsA with a calculated mean concentration in the worm of about 28 µM in the wild type strain (Table 1A). In this assay, the samples were treated with ascorbate oxidase to confirm AsA identity. However, since ascorbate oxidase has additional substrates such as 6-deoxy-L-ascorbate, D-araboascorbate and L-galactoascorbate [16], the presence of AsA in these worm extracts were also analyzed using a liquid chromatography coupled to mass spectrometry assay (Table 1B). Here, AsA was also detected with a calculated mean concentration of about 33 µM, a very similar value as that found using the HPLC method.

The AsA concentrations measured here for C. elegans are low compared to those that have been reported in previous studies for plants, green algae, and mammals, but are comparable to those of the yeast 5-carbon AsA analogue erythroascorbate measured in Saccharomyces cerevisiae (Table 2).

Surprisingly, the results demonstrate that there is no statistically significant difference in the AsA levels between the wild type and the C10F3.4 deletion strain, tm2679 (Table 1A). In extracts from both strains we were able to detect a peak that eluted at about 3.2 - 4 min that disappeared after treatment with ascorbate oxidase and increased in size following supplementation of the extracts with authentic AsA (Figure 1). The chromatograms presented here from extracts done in parallel demonstrate that the mutant strain AsA levels were perhaps somewhat higher compared to the wild type N2 strain, although the difference was not statistically significant (Table 1A).

Additionally, we measured AsA levels in worm extracts using mass spectrometry. Here, the AsA precursor ion (175 m/z in the negative ion mode) yielded two product ions: 175>115 m/z and 175>87 m/z that were derived from the ascorbate peak eluting at 6.5 min (Figure 2, top

AsA concentrations obtained in the HPLC analysis and confirmed that wild type and C10F3.4 mutant extracts contain similar amounts of this antioxidant (Table 1B).

Presence of Orthologs of Vitamin C Synthetic Enzymes in C. elegans

Bioinformatic analyses performed by BLAST [17] revealed that the C. elegans genome lacks homologs of several of the important enzymes in the L-galactose pathway for AsA synthesis in plants. These enzymes include the GDP-D-mannose 3”,5”-epimerase that catalyzes

GDP-L-galactose formation from GDP-D-mannose, as well as the last enzyme in the pathway,

L-galactono-1,4-lactone-dehydrogenase that converts L-galactono-1,4-lactone to AsA (Table 3).

Additionally, homologs for the last two enzymes in the animal pathway, gluconolactonase and L- gulonolactone oxidase are missing from the C. elegans genome (Table 3). These results suggest that a complete biosynthetic pathway of neither the animal nor plant type is present in C. elegans.

Can C. elegans Obtain Ascorbate from their E. coli Food Source?

These findings led us to ask whether C. elegans worms may obtain vitamin C from their diet

[18][19][20]. Under laboratory culture conditions, C. elegans worms are fed with the non- pathogenic E. coli strain OP50 in S-medium containing salt buffers and cholesterol as described in the "Experimental Procedures" section. HPLC analysis of the C. elegans S medium as well as

61 of OP50 extracts and chemicals used in the extraction solutions did not reveal the presence of

AsA (Figures 3 and 4). Since some peaks in these samples overlapped with the AsA which eluted at 3.2 min, a three- dimensional (time versus wavelength versus absorbance) representation of the chromatogram were generated to provide additional evidence for the identification of these “non-AsA” peaks. AsA absorbs spectrum in the range 244–265 nm [21], subject to the pH of the solution [22]. Only the 20 nmol AsA standard (Figure 3A) showed this spectrum with a peak at 255 nm that was absent in the other samples (Figure 3B-H). Luria Broth

(LB), the culture medium for the OP50 cells, as well as E. coli OP50 extracts were additionally analyzed by HPLC after spiking the samples with AsA or together with ascorbate oxidase

(Figure 4A and 4B). The HPLC chromatograms at 265nm revealed that the OP50 samples did not contain Vitamin C, as “OP50 alone” chromatogram was similar to the OP50 with AO chromatogram (Figure 4A top and lower right panel). For LB chromatograms, some peaks in

“LB alone” chromatogram overlapped with peaks that were seen in the sample spiked with AsA.

To confirm AsA absence from these samples, three- dimensional UV spectrums were produced and revealed that only samples that were spiked with AsA displayed the typical AsA spectrum with a peak at 255 nm (Figure 4B). These experiments suggest that the AsA detected in C. elegans extracts was of endogenous origin, although the possibility that very low (and undetectable under our conditions) concentrations of AsA in the medium might be highly concentrated by endogenous ascorbate transport proteins in C. elegans cannot be eliminated at this point.

DISCUSSION

In the present study we examined AsA levels in the free-living nematode C. elegans.

Additionally, we studied the possible involvement of C10F3.4, the worm homolog of plant

VTC2, in vitamin C synthesis in C. elegans. As of today, there are no studies that demonstrate the capability of invertebrates to synthesize AsA. However, the significance of AsA in invertebrates has been reported previously. AsA has been shown to be an essential nutrient for the growth and development of plant-eating invertebrates [18]. Specifically, the role of AsA as an antioxidant in insects was reported by Felton and Summers who measured the enzymatic activity of ascorbate peroxidase in the moth Helicoverpa zea [19]. A role of AsA as a cofactor for the hydroxylation of the insect molting promoting hormone ecdysone has been suggested as well [19]. Rajan and colleagues investigated the dietary requirements of the human filarial parasite nematode Brugia malayi and reported that AsA is essential to complete the L3 to L4 molting in this species [23]. Although there are no direct reports showing requirement for AsA in

C. elegans diet, there are a number of observations that demonstrate the need for AsA both as a an antioxidant as well as an enzyme cofactor. The support for AsA playing the role of an antioxidant in C. elegans comes from studies that showcased the property of this compound to mitigate oxidative stress. Supplementation of culture media with AsA has been shown to alleviate the oxidative stress caused by a mutation in the C. elegan’s iron–sulfur flavoenzyme, succinate dehydrogenase (SDH) [24]. Another study by Khare et al. demonstrated that AsA could rescue the Egl phenotype inflicted by the oxidizing agents juglone and paraquat in C. elegans [14]. In both studies, the culture media had been supplemented with an exogenous source of AsA, suggesting that C. elegans is capable of taking up and utilizing AsA. Support for

63 this idea comes from the analysis of the C. elegans genome, where six homologs for the human

AsA transporter NAT have been identified [25]. These observations also indicate that the nematode is unable to produce enough endogenous AsA to effectively cope with the oxidative stress that was experimentally inflicted in the above studies.

In addition to its ability to use AsA as an antioxidant, C. elegans can recycle the oxidized form (dehydroascorbate) to AsA. The omega class glutathione transferase (GSTO-1) that is present in C. elegans has a glutathione-dependent dehydroascorbate reductase activity that was upregulated under stress conditions while its overexpression increased resistance to the oxidative stress inducing agents juglone, paraquat, and cumene hydroperoxide [26].

The role of AsA as a cofactor has been mainly reported in relation to the activation of prolyl 4-hydroxylase [27]. The hydroxylation activity of this enzyme, required for the synthesis of the C. elegans cuticle [28] and egg shell [29] collagens as well as for controlling the stability of the hypoxia-inducible transcription factor (HIF), has been shown to depend on AsA as a cofactor [28][30]. Additionally, tyramine beta-hydroxylase, which is required for the synthesis of the neurotransmitter octopamine in C. elegans is an ortholog of the mammalian dopamine beta hydroxylase, a copper and AsA dependent enzyme [31][32].

In our results, HPLC and LC/MS/MS methods that were developed to assay AsA in C. elegans demonstrated that this species contains low, yet detectable levels of AsA. The measured

AsA levels were highly variable, ranging from 7 to 60 µM in the wild type strain. This variability may have resulted from slight differences in the culturing conditions that led to variable oxidative stress levels in these worms. I note here that all measurements were made of extracts containing DTT that reduces dehydroascorbate to AsA, thus the values given here reflect the contributions of both AsA and dehydroascorbate. Measurements of AsA levels before and after

64 addition of oxidizing agents such as juglone and paraquat could be useful in testing the hypothesis that the variability in oxidative stress the worms were sensing affected the AsA levels that were measured in their extracts.

The mean AsA levels measured here in C. elegans (about 30 µM) are low compared to other organisms (Table 2). This finding could indicate that the chief function of AsA in C. elegans may be as a cofactor for enzymatic reactions rather than as an antioxidant. On the other hand, it is possible that the low levels of AsA reflect the low levels of oxidative stress that C. elegans experience under laboratory conditions.

In light of the plausible requirement of AsA in C. elegans together with the sequence similarity of C. elegans C10F3.4 to Arabidopsis VTC2, we hypothesized that C10F3.4 could be involved in a plant like pathway for AsA synthesis. Nevertheless, the AsA measurements in both the wild type and C10F3.4 deletion strains did not yield any significant difference, showing that

AsA levels in this species are independent of C10F3.4 activity. Bioinformatics analyses also did not support the presence of a plant-like AsA synthesis pathway in C. elegans as its genome lacks key enzymes of this pathway (Table 3). As a conclusion, while we could detect AsA in worm extracts, our other findings did not support an AsA biosynthetic activity in C. elegans. To try to resolve the issue of the origin of AsA in C. elegans extracts, we measured AsA levels in exogenous sources. We could not detect AsA in any of the culture media or the E. coli cells that were used to grow the worms in the laboratory (Figure 3, 4A and 4B). One possible explanation for this result is that C. elegans has a high-affinity transport system for AsA and is capable of taking up low levels of AsA from the media that were not detectable in our HPLC assay.

65 The C. elegans genome encodes for proteins that either require AsA as a cofactor

[28][29][30][31], recycle the reduced cofactor from its oxidized form dehydroascorbate [26], or serve as putative transporters of this molecule into the cells [25]. However, the fact that there was no difference in the levels of AsA between the wild type and the C10F3.4 deletion strain, together with the lack in C. elegans of key homologs of the plant L-galactose pathway or of the animal AsA synthesis pathway indicates that these nematodes do not synthesize AsA via the

Smirnoff Wheeler pathway or the animal pathway.

Future work is designed to develop a more sensitive assay for AsA and DHA that will allow us to test whether C. elegans can take up AsA or DHA from exogenous sources.

Additionally, It would be interesting to test whether C. elegans strains with mutation in the putative AsA transporters exhibit reduced levels of AsA. AsA levels in C. elegans can be measured in response to oxidative stress and to determine the role of the AsA recycling genes in these conditions.

ACKNOWLEDGEMENTS

I would like to thank Brian Young and Madeline Young, previous members of the Clarke lab, for their collaboration in sharing their HPLC data for AsA levels in C. elegans wild type and

C10F3.4 deletion strains presented in Table 1 and in performing HPLC assays to test the presence of AsA in C. elegans growth medium, E. coli OP50 extracts and chemicals used in the extraction solutions presented in Figure 3 and 4. I would also like to thank Dr. Beth

Marbois from the Department of Chemistry and Biochemistry at UCLA for her helpful advice in developing LC-ESI-MS method for AsA measurements in C. elegans extracts.

Figure 1 – HPLC measurements of AsA in wild type (N2) and the C10F3.4 mutant strain

(tm2679). C. elegans extracts prepared from N2 (left column) or tm2679 mutant strain (right column) were analyzed by the HPLC method described under “Experimental Procedures.” Worm extracts were injected alone (top panel), together with 2 U of ascorbate oxidase (AO) (middle panel) or spiked with 0.5 nmol AsA (lower panel). Arrows indicate the AsA peak.

67 Chromatograms obtained with one set of wild type and mutant extracts are shown. Four to seven other sets of extracts were analyzed; the AsA levels found in these worm extracts were quantified and the values obtained are given in Table 1A.

69 Figure 2 – LC-MS analysis of AsA in wild type (N2) and the C10F3.4 mutant strain (tm2679).

AsA standard (top panel), wild type N2 (middle panel) and mutant tm2679 (lower panel) extracts were analyzed by the LC-MS method described under “Experimental Procedures.” Extracted ion chromatograms of product ions m/z 175 > 115 (blue) and m/z 175 > 87 (red) were obtained after injection of a solution containing 0.1 nmol of AsA (top panel), wild-type N2 extract (middle panel) and C10F3.4 mutant extract (lower panel). Chromatograms obtained with one set of wild type and mutant extracts are shown. The AsA levels found in these worm extracts were quantified, and the values obtained are given in Table 1B. cps, counts/s.

71 Figure 3 – C. elegans culture medium, food source, and extraction reagents do not contain detectable AsA. The HPLC method used in Fig. 1 was employed to analyze AsA content in buffers that were used in worm culture media or to process worms extracts including extraction solution (B), S-medium (C), M9 medium (E), 12 nmol sucrose (G), 19 nmol monosodium citrate

(H), an extract of the worm’s food source, the E. coli OP50 strain (F), and a buffered preparation of the LB medium used to grow the E. coli strain (D). The injection volumes for all tested solutions were 40 µL with the exception of E. coli extract with injection volume of 120 µL. The three-dimensional UV spectrum (time in min on x-axis, absorbance in mAu on the y-axis, and wavelength in nm on the z-axis) of HPLC experiments presented here show that none of the solutions B-H produced a peak with the retention time and a UV similar spectrum to the 20 nmol

AsA standard (A) that resulted in a peak at a retention time of 3.2 min with maximal UV absorbance at about 255 nm. All the solutions and the E. coli extract were prepared as described in the "Experimental Procedures" section.

72 A

Figure 4 - HPLC measurements of AsA in LB and the E. coli OP50 extract. Extracts of the E. coli OP50 cells and its growth media LB were prepared and analyzed by the HPLC method described under “Experimental Procedures”. The absorbance at 265nm (A) or the three dimensional UV spectrum (time versus wavelength versus absorbance as described in Fig. 3) (B) for each sample are presented. For E. coli cell extract, 120 µL of the untreated sample (right top panel A and B), 124.5 µL of ascorbate oxidase (AO) treated sample (right middle panel A and B) or 128.25 µL of E. coli supernatant supplemented with 30 nmol of AsA (right lower panel A and

B) were injected onto the C18 column. For LB, 40 µL of untreated sample (left top panel A and

B), 33.75 µL of ascorbate oxidase (AO) treated sample (left middle panel A and B) and 33 µL of

LB supplemented with 30 nmols of AsA (left lower panel A and B) were injected onto the C18 column. Arrows indicate the AsA peak.

74 Table 1 - AsA levels in C. elegans in wild type and tm2679 strains. AsA concentrations were measured in extracts of C. elegans using either HPLC (A) or LC-MS/MS (B) as an analytical method. In each case, the number of nmol of ascorbate in the sample was determined from a calibration curve where known amounts of AsA were injected into the HPLC and their amounts quantitated by either the area of the absorbance at 254 nm or the integrated cps for the LC-

MS/MS. Concentrations were determined based on the wet weight of the worm pellet and the dilution factor of the extraction buffer and other reagents leading to the chromatography. This calculation assumed that the one gram of the pellet wet weight was equal to 1 ml of worm volume. The values measured with HPLC represent the means ± S.D. calculated from 4-7 individual experiments for each strain. Statistical analysis using a two-tailed unpaired Student’s t test with unequal variance shows that there is no statistically significant difference in AsA concentration between the N2 and tm2679 C. elegans strains (p value = 0.34). The values measured with LC/MS represent results of one repeat per strain. n = number of biological replicates.

AsA Strain Analytical method nmol/g worms (µM)

N2 HPLC 28.1± 22.6 (n=7)

tm2679 HPLC 49.1± 35.6 (n=4)

N2 LC-MS/MS 33.3

tm2679 LC-MS/MS 30.9

75 Table 2 - Summary of AsA levels in different species. The concentrations of AsA or the yeast 5- carbon AsA analog L-erythroascorbate in different species including plants, yeast invertebrates and vertebrates is shown.

AsA or erythroascorbate Species References (mM)

Arabidopsis thaliana 2 - 25 [33]

Mus musculus 0.3 - 4 [34]

Chlamydomonas 0.1-1.1 [35] reinhardtii

Saccharomyces 0.01 - 0.4 [36] cerevisiae

Caenorhabditis elegans 0.01 - 0.05 This chapter

76 Table 3 - Search for C. elegans homologs of the enzymes of the Smirnoff-Wheeler and the animal vitamin C biosynthesis pathways. Protein sequences of the A. thaliana L-galactose pathway components, A. thaliana MIOX4, rat gulono-lactone oxidase and gluconolactonase

(animal-like pathway) were used as query sequences in a BLAST search against the C. elegans protein database to identify homologs. MBH, mutual best hit.

MBH, mutual Protein ID/Name best hit

% amino A. thaliana C. elegans MBH acid identity

Phosphomannose PMI1 ZK632.4 34% yes isomerase PMI2 ZK632.4 36% - Phosphomannomutase PMM F52B11.2 54% yes GDP-D-mannose VTC1 TAG-335 55% yes pyrophosphorylase GDP-D-mannose GME1 - - no 3”,5”-epimerase GDP-L-galactose VTC2 C10F3.4 24% − phosphorylase VTC5 C10F3.4 32% yes L-galactose-1-P- inositol-1(or 4)- VTC4 40% yes phosphatase monophosphatase L-galactose l-Gal-DH MEC-14 37% yes dehydrogenase L-galactono-1,4- lactone- GLDH - - no dehydrogenase animal-like pathway myo-inositol MIOX4 C54E4.5 45% no oxygenase (MIOX1)

77 Gulonolactonase (SMP30) (Rattus − - - no norvegicus) Gulono-lactone AT3G47930 - - oxidase (GLDH) no (Rattus norvegicus) AT2G46750 - - AT2G46760 - -

REFERENCES

1. Pinnell SR. Regulation of collagen biosynthesis by ascorbic acid: a review. Yale J Biol

Med. 1985 Nov-Dec;58(6):553-9.

2. Burns JJ, Peyser P & Moltz A. Missing step in guinea pigs required for the biosynthesis

of l-ascorbic acid. Science 1956 Dec 7;124(3232):1148-9.

3. Nishikimi M, Fukuyama R, Minoshima S, Shimizu N & Yagi K. Cloning and

chromosomal mapping of the human nonfunctional gene for l-gulonogamma-lactone

oxidase, the enzyme for l-ascorbic acid biosynthesis missing in man. J Biol Chem 1994

May 6;269(18):13685-8.

4. Nishikimi M, Kawai T & Yagi K. Guinea pigs possess a highly mutated gene for l-

gulono-gammalactone oxidase, the key enzyme for l-ascorbic acid biosynthesis missing

in this species. J Biol Chem 1992 Oct 25;267(30):21967-72.

5. Linster CL, Van Schaftingen E. Vitamin C. Biosynthesis, recycling and degradation in

mammals. FEBS J. 2007 Jan;274(1):1-22. Review.

6. Linster CL, Clarke SG. L-Ascorbate biosynthesis in higher plants: the role of VTC2.

Trends Plant Sci. 2008 Nov;13(11):567-73. Epub 2008 Sep 27

7. Wheeler GL, Jones MA, Smirnoff N. The biosynthetic pathway of vitamin C in higher

plants. Nature. 1998 May 28;393(6683):365-9.

8. Linster CL, Gomez TA, Christensen KC, Adler LN, Young BD, Brenner C, Clarke SG.

Arabidopsis VTC2 encodes a GDP-L-galactose phosphorylase, the last unknown enzyme

in the Smirnoff-Wheeler pathway to ascorbic acid in plants. J Biol Chem. 2007 Jun

29;282(26):18879-85. Epub 2007 Apr 26.

79 9. Sulston, J.E. and J. Hodgkin, Methods, in The Nematode Caenorhabditis elegans, W.S.

Wood, Editor 1988: Cold Spring Harbor: Cold Spring Harbor Laboratory. p. 587-606.

10. Lewis, J.A., and Fleming, J.T. Basic culture methods. Methods Cell Biol. 1995;48:3-29.

11. Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974 May;77(1):71-94.

12. Kagan RM, Clarke S. Protein L-isoaspartyl methyltransferase from the nematode

Caenorhabditis elegans: genomic structure and substrate specificity. Biochemistry. 1995

Aug 29;34(34):10794-806.

13. Brenner C. Hint, Fhit, and GalT: function, structure, evolution, and mechanism of three

branches of the histidine triad superfamily of nucleotide hydrolases and transferases.

Biochemistry. 2002 Jul 23;41(29):9003-14.

14. Khare S, Gomez TA, Linster CL, Clarke SG. Defective responses to oxidative stress in

protein l-isoaspartyl repair-deficient Caenorhabditis elegans. Mech Ageing Dev 2009

Oct;130(10):670-80. Epub 2009 Aug 12.

15. Adler LN, Gomez TA, Clarke SG, Linster CL. A novel GDP-D-glucose phosphorylase

involved in quality control of the nucleoside diphosphate sugar pool in Caenorhabditis

elegans and mammals. J Biol Chem. 2011 Jun 17;286(24):21511-23. Epub 2011 Apr 20

16. White GA, Smith FG. Substrate specificity of the myrothecium ascorbic acid oxidase.

Nature. 1961 Apr 8;190:187-9.

17. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped

BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic

Acids Res. 1997 Sep 1;25(17):3389-402.

18. Chippendale GM. Ascorbic acid: an essential nutrient for a plant-feeding insect, Diatraea

grandiosella . J Nutr. 1975 Apr;105(4):499-507.

80 19. Felton GW, Summers CB. Antioxidant systems in insects. Arch Insect Biochem Physiol.

1995;29(2):187-97.

20. Dadd RH. Insect nutrition: current developments and metabolic implications. Annu Rev

Entomol. 1973;18:381-420.

21. Franco Van de Velde, María E. Pirovani, María S. Cámara, Daniel R.

Güemes and Cecilia M. del H. Bernardi. Optimization and Validation of a UV–HPLC

Method for Vitamin C Determination in Strawberries (Fragaria ananassa Duch.), Using

Experimental Designs. Food Analytical Methods, Methods (28 December 2011), pp. 1-8.

22. Hewitt EJ, Dickes GJ. Spectrophotometric Measurements on Ascorbic Acid and Their

Use for the Estimation of Ascorbic Acid and Dehydroascorbic Acid in Plant Tissues.

Biochem J. 1961 Feb;78:384-91.

23. Rajan TV, Paciorkowski N, Kalajzic I, McGuiness C. Ascorbic acid is a requirement for

the morphogenesis of the human filarial parasite Brugia malayi. J

Parasitol. 2003 Aug;89(4):868-70.

24. Huang J, Lemire BD. Mutations in the C. elegans succinate dehydrogenase iron-sulfur

subunit promote superoxide generation and premature aging. J Mol Biol. 2009 Apr

3;387(3):559-69. Epub 2009 Feb 20.

25. Gournas C, Papageorgiou I, Diallinas G. The nucleobase-ascorbate transporter (NAT)

family: genomics, evolution, structure-function relationships and physiological role. Mol

Biosyst. 2008 May;4(5):404-16. Epub 2008 Mar 11. Review.

26. Burmeister C, Lüersen K, Heinick A, Hussein A, Domagalski M, Walter RD, Liebau E.

Oxidative stress in Caenorhabditis elegans: protective effects of the Omega class

glutathione transferase (GSTO-1). FASEB J. 2008 Feb;22(2):343-54. Epub 2007 Sep 27.

81 27. Friedman L, Higgin JJ, Moulder G, Barstead R, Raines RT, Kimble J. Prolyl 4-

hydroxylase is required for viability and morphogenesis in Caenorhabditis elegans. Proc

Natl Acad Sci U S A. 2000 Apr 25;97(9):4736-41.

28. Myllyharju J. Prolyl 4-hydroxylases, key enzymes in the synthesis of collagens and

regulation of the response to hypoxia, and their roles as treatment targets. Ann

Med. 2008;40(6):402-17.

29. Riihimaa P, Nissi R, Page AP, Winter AD, Keskiaho K, Kivirikko KI, Myllyharju J. Egg

shell collagen formation in Caenorhabditis elegans involves a novel prolyl 4-hydroxylase

expressed in spermatheca and embryos and possessing many unique properties. J Biol

Chem. 2002 May 17;277(20):18238-43. Epub 2002 Mar 12.

30. Keskiaho K, Kukkola L, Page AP, Winter AD, Vuoristo J, Sormunen R, Nissi

R, Riihimaa P, Myllyharju J.Source . Characterization of

a novel Caenorhabditis elegans prolyl 4-hydroxylase with

a unique substrate specificity and restricted expression in the pharynx and excretory duct.

J Biol Chem. 2008 Apr 18;283(16):10679-89. Epub 2008 Feb 13

31. Alkema MJ, Hunter-Ensor M, Ringstad N, Horvitz HR.Tyramine Functions

independently of octopamine in the Caenorhabditis elegans nervous system. Neuron.

2005 Apr 21;46(2):247-60.

32. Diliberto EJ Jr, Daniels AJ, Viveros OH. Multicompartmental secretion of ascorbate and

its dual role in dopamine beta-hydroxylation. Am J Clin Nutr. 1991 Dec;54(6

Suppl):1163S-1172S.

33. Smirnoff N, Conklin PL, Loewus FA. Biosyntesis of ascorbic acid in plants:

Areniassance. Annu Rev Plant Physiol Plant Mol Biol. 2001 Jun;52:437-467.

82 34. Tsao CS, Leung PY, Young M. Effect of Dietary Ascorbic Acid Intake on Tissue

Vitamin C in Mice. J Nutr. 1987 Feb;117(2):291-7.

35. Urzica EI, Adler LN, Page MD, Linster CL, Arbing MA, Casero D, Pellegrini M,

Merchant SS, Clarke SG. Impact of Oxidative Stress on Ascorbate Biosynthesis in

Chlamydomonas via Regulation of the VTC2 Gene Encoding a GDP-L-galactose

Phosphorylase. J Biol Chem. 2012 Apr 20;287(17):14234-45. Epub 2012 Mar 5

36. Spickett CM, Smirnoff N, Pitt AR. The biosynthesis of erythroascorbate in

Saccharomyces cerevisiae and its role as an antioxidant. Free Radic Biol Med. 2000 Jan

15;28(2):183-92.

CHPTER 6

A Novel GDP-D-Glucose Phosphorylase Involved in Quality Control of the

Nucleoside Diphosphate Sugar Pool in Caenorhabditis elegans and Mammals

84 Supplemental Material can be found at: http://www.jbc.org/content/suppl/2011/04/20/M111.238774.DC1.html

A Novel GDP-D-glucose Phosphorylase Involved in Quality Control of the Nucleoside Diphosphate Sugar Pool in Caenorhabditis elegans and Mammals*□S Received for publication, March 9, 2011, and in revised form, April 15, 2011 Published, JBC Papers in Press, April 20, 2011, DOI 10.1074/jbc.M111.238774 Lital N. Adler‡1, Tara A. Gomez‡, Steven G. Clarke‡2, and Carole L. Linster‡§3 From the ‡Department of Chemistry and Biochemistry and the Molecular Biology Institute, UCLA, Los Angeles, California 90095 and the §de Duve Institute, Universite´Catholique de Louvain, 1200 Brussels, Belgium

The plant VTC2 gene encodes GDP-L-galactose phosphorylase, The last missing enzyme of the plant vitamin C biosynthesis arate-limitingenzymeinplantvitaminCbiosynthesis.Genes pathway was recently identified as a GDP-L-galactose phosphor-

encoding apparent orthologs of VTC2 exist in both mammals, ylase encoded by the Arabidopsis VTC2 gene (1, 2). In this Downloaded from which produce vitamin C by a distinct metabolic pathway, and in 10-step metabolic pathway from D-glucose to L-ascorbate, 4 the nematode worm Caenorhabditis elegans where vitamin C bio- GDP-L-Gal phosphorylase catalyzes the formation of L-Gal- synthesis has not been demonstrated. We have now expressed 1-P in the first committed (and highly regulated) step to vitamin cDNAs of the human and worm VTC2 homolog genes (C15orf58 C biosynthesis (2–5). Intriguingly, the VTC2 gene appears to be and C10F3.4,respectively)andfoundthatthepurifiedproteinsalso conserved in vertebrates, which are known to synthesize vita- www.jbc.org display GDP-hexose phosphorylase activity. However, as opposed min C via a different metabolic pathway than plants that does to the plant enzyme, the major reaction catalyzed by these enzymes not include GDP-L-Gal as an intermediate (6), and in inverte- is the phosphorolysis of GDP-D-glucose to GDP and D-glucose brates such as Caenorhabditis elegans that may lack the ability

1-phosphate. We detected activities with similar substrate specific- to synthesize vitamin C (1). The physiological functions of these at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 ity in worm and mouse tissue extracts. The highest expression of apparent orthologs are thus unclear. One clue may be that the GDP-D-glucose phosphorylase was found in the nervous and male Arabidopsis VTC2 enzyme can also use GDP-D-glucose as a reproductive systems. A C. elegans C10F3.4 deletion strain was substrate (1). Although the role of this activity in plants is found to totally lack GDP-D-glucose phosphorylase activity; this unclear, such an activity may be important in other species. activity was also found to be decreased in human HEK293T cells Whereas GDP-L-Gal is most probably absent from mammalian transfected with siRNAs against the human C15orf58 gene. These tissues, the presence of GDP-D-Glc in bovine mammary gland observations confirm the identification of the worm C10F3.4 and and the existence of a mammalian GDP-D-Glc pyrophosphory- the human C15orf58 gene expression products as the GDP-D-glu- lase have been reported (7, 8). More recent observations suggest cose phosphorylases of these organisms. Significantly, we found an that GDP-D-Glc formation can be catalyzed by the mammalian accumulation of GDP-D-glucose in the C10F3.4 mutant worms, GDP-D-mannose pyrophosphorylase (9, 10). No metabolic role suggesting that the GDP-D-glucose phosphorylase may function to of GDP-D-Glc in mammalian tissues has been proposed, how- remove GDP-D-glucose formed by GDP-D-mannose pyrophos- ever, and nothing is known about this metabolite in C. elegans. phorylase, an enzyme that has previously been shown to lack spec- The Arabidopsis VTC2 protein and its invertebrate and ver- ificity for its physiological D-mannose 1-phosphate substrate. We tebrate homologs are members of the histidine triad (HIT) propose that such removal may prevent the misincorporation of superfamily of proteins, which includes nucleotide hydrolases glucosyl residues for mannosyl residues into the glycoconjugates of and transferases (11). Known substrates of this superfamily worms and mammals. of enzymes include AMP-lysine, diadenosine polyphosphate, AMP-sulfate, and nucleotide sugars such as UDP-D-Glc and ADP-D-Glc in addition to GDP-L-Gal (12). HIT enzymes are characterized by the presence of a HXHX(H/Q) motif (X rep- * This work was supported, in whole or in part, by National Institutes of resents a hydrophobic amino acid), the second histidine residue Health Grants GM026020 and AG032303 (to S. G. C.). This work was also of which generally attacks the ␣-phosphate of the nucleotide supported by a Senior Scholar in Aging Award from the Ellison Medical Foundation (to S. G. C.). substrate. This leads to release of the leaving group and nucle- □S The on-line version of this article (available at http://www.jbc.org) contains otidylation of the enzyme. The nucleotidylated enzyme is then supplemental Figs. S1–S3 and Tables S1 and S2. resolved by either hydrolysis or, in the case of transferases and 1 Supported by a UCLA Philip Whitcome Fellowship. 2 To whom correspondence may be addressed: Dept. of Chemistry and Bio- chemistry and the Molecular Biology Inst., UCLA, 607 Charles E. Young Dr. 4 The abbreviations used are: GDP-L-Gal, GDP-␤-L-galactose; ADP-D-Glc, E., Los Angeles, CA 90095. Tel.: 310-825-8754; Fax: 310-825-1968; E-mail: ADP-␣-D-glucose; D-Gal-1-P, ␣-D-galactose 1-phosphate; L-Gal-1-P, ␤-L- [email protected]. galactose 1-phosphate; GDP-D-Glc, GDP-␣-D-glucose; GDP-D-Man, 3 Supported by the Fonds de la Recherche Scientifique (FNRS) and by the GDP-␣-D-mannose; GDP-L-Fuc, GDP-␤-L-fucose; D-Glc-1-P, ␣-D-glucose European Union Seventh Framework Programme (FP7/2007-2013) under 1-phosphate; HIT, histidine triad; IP-LC-ESI-MS, ion pair liquid chroma- grant agreement 276814. To whom correspondence may be addressed: de tography-electrospray ionization mass spectrometry; D-Man-1-P, ␣-D- Duve Inst., Universite´catholique de Louvain, BCHM 7539, Ave. Hippocrate mannose 1-phosphate; NDP, nucleoside diphosphate; UDP-D-Gal, 75, B-1200 Brussels, Belgium. Tel.: 32-2-7647561; Fax: 32-2-7647598; UDP-␣-D-galactose; UDP-D-Glc, UDP-␣-D-glucose; LLO, lipid-linked oli- E-mail: [email protected]. gosaccharide; CDG-I, congenital disorders of glycosylation type I.

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21511

85 Molecular Identification of GDP-D-glucose Phosphorylase

phosphorylases, transfer to a second phosphorylated substrate Facility (University of Utah). The C10F3.4::GFP construct, placing or inorganic phosphate, respectively (4, 12). the C10F3.4 gene as well as the 3-kb genomic region immediately In the present work, we overexpressed and purified the C. upstream of the C10F3.4 transcription start site in-frame with elegans and human VTC2 homolog gene products (C10F3.4 GFP, was made via a PCR fusion procedure described previously and C15orf58, respectively) and determined their kinetic prop- (18). The primer sequences used for this fusion procedure were as erties as well as their substrate specificities. We found that both follows: 5Ј-TGATCAGTTGGGACCCACCATGAG-3Ј,5Ј-CTT- genes encode GDP-D-Glc phosphorylases with properties very TGGCCAATCCCGGGGATCTTTCGATCCCTCGTCAGAT- similar to each other and to GDP-D-Glc phosphorylase activity TGAGC-3Ј,5Ј-GCTCAATCTGACGAGGGATCGAAAGATC- detected in C. elegans and mouse tissue extracts. We confirmed CCCGGGATTGGCCAAAG-3Ј,and5Ј-AGCACGAACAACT- the molecular identity of this enzyme by using a C10F3.4 worm CGGGGTGA-3Ј.TheC10F3.4::GFPfusionconstructwas deletion strain and by performing siRNA experiments in injected at 5 ng/␮lalongwith80ng/␮lpRF4andpBlueScriptKSat HEK293T cells. Based on the accumulation of GDP-D-Glc in C. 35 ng/␮lintothegonadarmsofN2adulthermaphrodites.Trans- elegans worms lacking the VTC2 homolog, we propose a phys- genic animals of the N2 xtEX85(C10F3.4::GFP, pRF4) genotype iological role for GDP-D-Glc phosphorylase in removing the were identified by the Rol phenotype, and stable lines were iso- GDP-D-Glc formed by the nonspecific enzyme GDP-D-man- lated. The isolated line used in this study was designated UZ119. nose pyrophosphorylase to limit its misincorporation into gly- Cloning and Expression of C. elegans C10F3.4—The C. Downloaded from coproteins and other glycoconjugates. elegans C10F3.4a coding sequence was PCR-amplified from an Invitrogen ProQuest C. elegans cDNA library in the vector EXPERIMENTAL PROCEDURES pPC86 (catalog number 11288-016) using forward primer Materials—ADP-D-Glc, GDP-D-Glc, GDP-D-Man, UDP-D- 5Ј-CACCATGGAGCCTTTTCCGCGAATC-3Ј and reverse Ј Ј Gal, UDP-D-Glc, D-Gal-1-P, D-Glc-1-P, D-Man-1-P (all of these primer 5 -TCATTTCGATCCCTCGTCAGATTG-3 . The www.jbc.org nucleotide sugars and hexose phosphates were in the ␣- ProQuest cDNA library was prepared from RNA extracted configuration), GDP-␤-L-Fuc, GDP, glucose-6-P dehydrogen- from a mixed population of adult and larval nematode worms of ase from Leuconostoc mesenteroides, and ␤-nicotinamide ade- the N2 strain. The PCR product was purified and cloned into nine dinucleotide (NAD) were from Sigma. L-Gal-1-P was from the Champion pET100/D-Topo vector (Invitrogen), and the at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 Glycoteam (Hamburg, Germany). Phosphoglucomutase (␣-D- recombinant plasmid was transformed into E. coli BL21 Star glucose-1,6-bisphosphate:␣-D-glucose-1-phosphate phospho- (DE3) cells. The transformed E. coli cells were grown at 37 °C in transferase) from rabbit muscle was purchased from OYC LB medium containing 50 ␮g/ml carbenicillin (Sigma) until the Americas (Andover, MA). All other reagents were of analytical optical density at 600 nm reached a value of about 0.6. Recom- grade. GDP-␤-L-Gal, synthesized and purified as described binant protein production was induced by the addition of 0.4 (13), was provided by Prof. Shinichi Kitamura (Osaka Prefec- mM isopropyl ␤-D-thiogalactopyranoside. After overnight incu- ture University). This preparation was further purified by the bation with shaking at 18 °C, cells were harvested, washed, and reverse-phase HPLC method described previously (1). Frac- frozen at Ϫ80 °C. Cells were then thawed; resuspended in a pH tions containing GDP-L-Gal were lyophilized, resuspended in 8.0 buffer containing 50 mM sodium phosphate, 500 mM NaCl, Ϫ ␤ H2O, and stored at 20 °C. 5mM -mercaptoethanol, 5% glycerol, Roche Applied Science C. elegans Strains and Culture Conditions—Standard proce- Complete protease inhibitor mixture (1 tablet/50 ml), and 10 dures used to maintain C. elegans strains were adapted from units/ml Benzonase nuclease (Novagen); and lysed by sonica- Sulston and Hodgkin (14). M9 and S media were prepared as tion as described (19). The lysate was centrifuged for 60 min at described by Lewis and Fleming (15). The N2 strain was 13,000 ϫ g at 4 °C. The resulting supernatant was loaded onto a obtained from the Caenorhabditis Genetics Center (St. Paul, 5-ml HisTrap HP nickel affinity column (GE Healthcare), and MN). The C10F3.4 mutant strain (tm2679) lacking coding His-tagged C10F3.4 protein was eluted using a gradient of sequences between exons 5 and 8 of the C10F3.4 gene (Worm- 20–500 mM imidazole according to the manufacturer’s specifi- Base), was provided by the C. elegans Core Facility at Tokyo cations. Peak protein fractions containing the C10F3.4 protein Women’s Medical University. We also used a transgenic strain were pooled, and buffer was exchanged to 10 mM HEPES, pH expressing a C10F3.4::GFP fusion protein under the endoge- 7.5, 100 mM NaCl, and 1 mM dithiothreitol in a 10-kDa-cutoff nous C10F3.4 promoter (see “Transgenic Worm Construction Amicon Ultra centrifugal filter unit (Millipore, Billerica, MA). and Analysis”). C. elegans strains were cultured at 20 °C accord- Protein concentration was determined using a Lowry assay ing to standard procedures (16). All strains were maintained at after precipitation with 10% trichloroacetic acid. SDS-PAGE 20 °C on nematode growth medium plates streaked with Esch- analysis of the purified enzyme preparation used throughout erichia coli OP50 (14). Liquid cultures of these worm strains this study revealed a single band at the expected molecular were grown at 20 °C in S medium supplemented with concen- weight for the His-tagged C10F3.4 protein after Coomassie trated E. coli OP50, and worms were harvested, washed, and Blue staining. The purified enzyme was stored at Ϫ80 °C in 10% sucrose-floated as described previously (17) except that M9 glycerol. medium was used instead of 0.1 M NaCl as a resuspension and Cloning and Expression of Human C15orf58—The Human washing solution. ORFeome Collaboration Clone 100016258, a Gateway Entry Transgenic Worm Construction and Analysis—The C10F3.4:: vector containing the predicted complete cDNA coding GFP fusion construct and the transgenic lines expressing the sequence of the human C15orf58 gene (GenBankTM accession C10F3.4::GFP fusion protein were generated at the Worm Core number BC153017), was purchased from Open Biosystems

21512 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 24•JUNE 17, 2011 86 Molecular Identification of GDP-D-glucose Phosphorylase

(Huntsville, AL). This Entry clone was used in LR recombina- RNA in a total volume of 20 ␮l. We used 5 ␮l of 20-fold diluted tion reactions (Invitrogen) with the T7 promoter expression reverse transcription reaction in subsequent 25-␮l quantitative vector pDEST17 to obtain a construct for production of recom- real time PCR reactions run in a MyiQTM thermal cycler with binant C15orf58 protein with an N-terminally fused hexahisti- SYBR Green as a fluorescent dye (Bio-Rad), HotStart Taq dine tag as described by the manufacturer. This construct was polymerase (Eurogentec, Seraing, Belgium), and 0.3 ␮M gene- transformed into E. coli BL21 Star (DE3) cells for protein over- specific forward and reverse primers. PCR programs were 95 °C expression. Recombinant human C15orf58 protein was pro- for 5 min, 40 cycles at 95 °C for 10 s, and 60 °C for 1 min with duced and purified as described for recombinant C. elegans acquisition of fluorescence information followed by a melting C10F3.4 protein with the exception that a 5–500 mM imidazole curve. Experiments shown represent averages for four biologi- gradient was used for elution of His-tagged C15orf58. Protein cal replicates. Primers used for amplification can be obtained concentration was determined using a Lowry assay after pre- upon request. cipitation with 10% trichloroacetic acid. Protein purity was Protein and Small Molecule Extraction from C. elegans and determined by SDS-PAGE analysis using Coomassie Blue stain- Mouse Tissues—All steps of the described extraction proce- ing followed by densitometry using the NIH ImageJ software; dures were carried out at 4 °C. Brain, heart, liver, kidney, thy- recombinant C15orf58 protein was estimated to represent 17% mus, and testes of C57BL6 mice were removed as quickly as of the total protein present in the enzyme preparation used possible from sacrificed animals; frozen in liquid nitrogen; and Downloaded from throughout this study. The contaminating proteins present in stored at Ϫ80 °C until further extraction. For protein extrac- this preparation did not interfere with our enzymatic assays tion, the frozen tissues were placed in 3 volumes (v/w) of lysis (absence of significant hydrolytic activities acting on GDP-sug- buffer (50 mM Tris-HCl, pH 7.5, 1 mM dithiothreitol, 2 mM ars or GDP), and recombinant C15orf58 was therefore not puri- EDTA, 10% glycerol, Roche Complete protease inhibitor mix-

fied further. The purified enzyme preparation was stored at ture (1 tablet/50 ml), and 1 mM phenylmethylsulfonyl fluoride) www.jbc.org Ϫ80 °C in 10% glycerol. and homogenized in a Potter-Elvehjem apparatus. The homo- siRNA Transfection and Extraction of HEK293T Cells— genates were centrifuged at 21,000 ϫ g for 40 min, and the HEK293T cells were maintained in Dulbecco’s modified Eagle’s supernatants were stored for enzymatic assays. For small mol- medium (DMEM) supplemented with 10% heat-inactivated ecule extraction, the frozen tissues were homogenized in a Pot- at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 fetal calf serum, 1% penicillin/streptomycin, and 2 mM Ultra- ter-Elvehjem apparatus after addition of 2 volumes (v/w) of 10%

Glutamine in an atmosphere of 5% CO2 in air. For transfection perchloric acid. After centrifugation of the homogenates at of cells with siRNAs, HEK293T cells were plated at a density of 21,000 ϫ g for 40 min, the supernatants were neutralized with 2 about 800,000 cells/9.6-cm well. Cells were transfected the fol- K2CO3. lowing day using 50 nM control siRNA (Dharmacon ON- Worms, harvested as described above, were resuspended in 2 TARGETplus non-targeting pool from Thermo Scientific volumes (v/w) of lysis buffer or 1 volume (v/w) of 10% perchlo- (Epsom, UK)) or 50 nM C15orf58 siRNA (Dharmacon ric acid. The worm suspensions were freeze-dropped in liquid ON-TARGETplus SMARTpool from Thermo Scientific) and 6 nitrogen and ground to a fine powder in a precooled mortar and ␮l of the transfection reagent Lipofectamine 2000 (Invitrogen) pestle. The homogenates were allowed to thaw on ice and cen- in the absence of antibiotics in the culture medium according trifuged at 21,000 ϫ g for 30 min. The supernatants of the per-

to the manufacturer’s instructions. The culture medium was chloric acid extracts were neutralized with K2CO3. All protein replaced by fresh medium containing antibiotics about 6 h after and small molecule extracts were stored at Ϫ80 °C until assayed addition of the RNA-Lipofectamine 2000 complexes. After for enzymatic activity or GDP-hexose content. Protein concen- 48–72 h, the cells were either washed with cold phosphate- tration was determined using the Lowry assay after protein pre- buffered saline (PBS) and resuspended in lysis buffer containing cipitation with 10% trichloroacetic acid. 25 mM HEPES, pH 7.1, 5 ␮g/ml antipain, and 5 ␮g/ml leupeptin HPLC-based Nucleoside Diphosphate (NDP)-hexose Phos- for protein extraction or resuspended in TriPure Isolation Rea- phorylase Assay—NDP-hexose phosphorylase activities of gent (Roche Applied Science) for total RNA extraction. For recombinant enzymes, worm and mouse tissue extracts, and protein extraction, cells were lysed by one cycle of freeze-thaw- HEK293T cell extracts were assayed by measuring NDP forma- ing and centrifuged. The resulting supernatant was used to tion after incubation with NDP-hexose in a reaction mixture at assay GDP-D-Glc phosphorylase activity by the HPLC method pH 7.5 containing 50 mM Tris-HCl or HEPES, 5 mM sodium

described below. Protein concentration in these extracts was phosphate, 0 or 2 mM MgCl2, 0 or 1 mM EDTA, 10 mM NaCl, determined by the Lowry assay after protein precipitation with and 1 mM dithiothreitol. Recombinant enzymes, worm extracts, 10% trichloroacetic acid. Total RNA was isolated from the Tri- and HEK293T cell extracts were generally assayed without

Pure cell extracts after phase separation using chloroform MgCl2 and always without EDTA in the reaction mixture. according to the manufacturer’s instructions. The RNA con- Mouse tissue extracts were assayed in the absence of MgCl2 and centration was determined by measuring the absorbance at 260 in the presence of 1 mM EDTA. The salt and small molecule nm, and the samples were stored at Ϫ80 °C. content of worm, mouse tissue, and HEK293T cell extracts was Quantitative Real Time PCR—To measure C15orf58 tran- minimized by processing them on spin desalting columns script levels in siRNA-transfected HEK293T cells, cDNA was (Thermo Scientific, Rockford, IL) immediately before using synthesized using RevertAidTM H Minus reverse transcriptase them for activity measurements. Desalting, omission of exter- ϩ (Fermentas, St. Leon-Rot, Germany) and random hexamers nally added Mg2 , and/or addition of EDTA allowed minimi- according to the manufacturer’s protocol and adding 1 ␮g of zation of NDP-hexose hydrolysis by contaminating nucleotide

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21513 87 Molecular Identification of GDP-D-glucose Phosphorylase pyrophosphatase activities. Additionally, desalting of the pro- methanol, 10% 10 mM ammonium acetate, and 7.5 mM hexyl- tein extracts allowed the removal of compounds prone to inter- amine adjusted to pH 8.9 with ammonium hydroxide. A mobile fere with the enzymatic assay (inorganic phosphate, nucleo- phase gradient was used with a flow rate of 0.4 ml/min starting tides, and nucleotide derivatives). Reactions (at 26 °C for C. at 100% A for 15 min followed by a linear gradient from 100 to elegans enzyme preparations and at 31 °C for mammalian 65% A over 18 min, then holding at 65% A for 2 min followed by enzyme preparations) were initiated by enzyme addition and a linear gradient from 65 to 100% A in 2 min, and finally holding stopped after 5–30 min by heating at 98 °C for 3 min. After for 13 min at 100% A. The column temperature was maintained removal of precipitated protein by centrifugation, supernatants at 40 °C, and the injection volume was 10 ␮l. Mass detection were analyzed by anion exchange HPLC as described (1). NDP was carried out in the negative mode under the following con- and NDP-hexose concentrations were calculated by comparing ditions: spray voltage at Ϫ4.5 kV, heated capillary temperature the integrated peak areas with those of standard NDP or NDP- at 550 °C, curtain gas at 25, and collision gas at the medium hexose solutions. setting. The system was tuned using a stock solution of GDP- Spectrophotometric GDP-D-Glc Phosphorylase Assay—The D-Glc or GDP-D-Man (10 ␮M) in water. The precursor-to-prod- GDP-D-Glc phosphorylase activity of worm and human recom- uct ion transitions in multiple reaction monitoring mode used binant enzymes was also measured by a spectrophotometric to detect GDP-D-Glc and GDP-D-Man were as follows: m/z assay modified from Nihira et al. (20). Briefly, a reaction mix- 604 3 362, m/z 604 3 158, m/z 604 3 210, and m/z 604 3 343. Downloaded from ture containing 60 mM Tris-HCl, pH 7.5, 2.4 mM NAD, 0–20 Both GDP-hexoses yielded the same precursor ion, m/z 604, mM sodium phosphate, 12 mM MgCl2, 12 mM NaCl, 1.2 mM which produced the same four product ions (m/z 158, 210, 343, dithiothreitol, 0–100 ␮M GDP-D-Glc, 6 IU/ml phosphogluco- and 362). However, the relative abundances at which these mutase, and 6 IU/ml glucose-6-P dehydrogenase was prepared, product ions were generated differed between the two GDP- ␮ 3 and 250 l of this mixture were added per well in a 96-well hexoses, and we chose product ion m/z 604 362 to detect and www.jbc.org plate. The latter was transferred to a plate reader (Molecular quantify GDP-D-Glc, whereas product ion m/z 604 3 158 was Devices, Sunnyvale, CA) that monitored the absorbance at 340 used to detect and quantify GDP-D-Man. nm. Recombinant enzyme was added after an ϳ5-min preincu-

RESULTS at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 bation at 26 or 31 °C, and linear production of NADH was ⌬ measured by A340 over a 10-min time period to determine the Characterization of C. elegans and Human Homologs of Ara- rate of conversion of GDP-D-Glc to D-Glc-1-P. The light path bidopsis VTC2—The Arabidopsis VTC2 gene has recently been length for each well was measured via the plate reader SoftMax shown to encode GDP-L-Gal phosphorylase, a highly regulated Pro5 software before addition of the enzyme. enzyme that catalyzes the first committed step in the plant vita- Microscopy—Nematodes were anesthetized using levamisole min C synthesis pathway (4). The presence of VTC2 homologs (2 mg/ml) and mounted on 2% agarose pads. Cells were identi- in species that synthesize vitamin C via a metabolic pathway fied by comparing the position of GFP-positive areas viewed by different from that of plants (most non-primate mammals), in simultaneous fluorescence and differential interference con- species that do not synthesize vitamin C (e.g. primates including trast microscopy. Images were captured using a Zeiss Axio- humans), or in species where vitamin C synthesis has not been scope (Thornwood, NY) at 40ϫ power magnification with an generally established (invertebrates) (6, 22, 23) stimulated us to attached Hamamatsu Orca-ER camera (Bridgewater, NJ) and characterize the human and C. elegans proteins to understand Volocity software (Improvision, Lexington, MA). All images their physiological function. These proteins share 25–30% shown in this study were prepared by using Adobe Photoshop sequence identity over their enzyme length with VTC2 from Elements 8.0 software (San Jose, CA). Arabidopsis thaliana (1). cDNAs of the human VTC2 homolog GDP-D-hexose Detection and Assay by Ion Pair Liquid gene (C15orf58) and the C. elegans VTC2 homolog gene Chromatography-Electrospray Ionization Mass Spectrometry (C10F3.4) were overexpressed in a bacterial expression system (IP-LC-ESI-MS)—All experiments were carried out on an Agi- as N-terminally His-tagged proteins and affinity-purified on ϩ lent 1200 HPLC system coupled to a 4000 QTRAP MS/MS Ni2 columns. hybrid triple quadrupole/linear ion trap mass spectrometer Interestingly, the Gln residue of the conserved HXHXQ from Applied Biosystems (Foster City, CA). The HPLC system motif in the plant VTC2 protein is replaced by a His residue in was equipped with a PAL autosampler with thermostated tray the mammalian and invertebrate homolog proteins including holders and stack (LEAP Technologies, Carrboro, NC). The human C15orf58 and worm C10F3.4. The resulting HXHXH turbo electrospray ionization source of the 4000 QTRAP sys- motif constitutes the histidine triad motif generally found in the tem was operated in the negative ion mode. Q1 and Q3 were nucleotide hydrolase branch of the HIT protein superfamily, operated in unit resolution. All system control, data acquisi- whereas HXHXQ is the signature motif of the nucleotide trans- tion, and mass spectral data evaluation were performed using ferase branch of this protein superfamily (12). To find out Analyst software version 1.4.2 from Applied Biosystems. The whether the human and worm proteins act as nucleotide GDP-D-hexose assay was based on the procedure of Coulier hydrolases or transferases, we tested the activity of the recom- ␮ ϫ et al. (21). A Luna 3- mC18(2) column (150 3.0 mm, Phe- binant enzymes on GDP-D-Glc, the preferred substrate for nomenex) was used with a Luna C18 security guard cartridge plant VTC2 after GDP-L-Gal, in the absence or presence of (4 ϫ 3.0 mm, Phenomenex). Mobile phase A consisted of 10 mM various potential nucleotide acceptors including inorganic hexylamine as an ion pairing agent (Aldrich) in water adjusted phosphate (Pi), inorganic pyrophosphate (PPi), L-Gal-1-P, and to pH 6.3 with acetic acid. Mobile phase B consisted of 90% D-Man-1-P (supplemental Fig. S1). The enzymatic activities

21514 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 24•JUNE 17, 2011 88 Molecular Identification of GDP-D-glucose Phosphorylase

TABLE 1 Characterization of GDP-hexose phosphorylase activities of recombinant human C15orf58 and C. elegans C10F3.4 proteins

Km and Vmax values were obtained by fitting the experimental data to the Michaelis-Menten equation using the Km calculator of the BioMechanic program. Enzymatic turnover numbers were derived from the Vmax values by using a molecular mass of 45 and 55.7 kDa for His-tagged C15orf58 and C10F3.4, respectively, with the assumption that the enzyme preparations were pure. Incubation times and enzyme concentrations were adjusted to obtain initial velocity data. Enzymatic activities were measured by the HPLC assay described under “Experimental Procedures” except for the GDP-D-Glc phosphorylase activity of the C. elegans enzyme where the spectrophotometric assay described under “Experimental Procedures” was used. The worm and human recombinant enzymes were assayed at 26 and 31 °C, respectively. Km and Vmax values for the GDP-hexoses were obtained in the presence of 5 mM inorganic phosphate, whereas those for inorganic phosphate were obtained in the presence of 50 ␮M GDP-D-Glc. Values are the means Ϯ S.D. calculated from three to five individual experiments for each substrate except for inorganic phosphate for which the kinetic parameters of the human enzyme were only determined twice.

kcat Km kcat/Km Substrate Human C. elegans Human C. elegans Human C. elegans Ϫ1 Ϫ1 Ϫ1 s mM s M 7 7 GDP-D-Glc 26.0 Ϯ 7.4 265 Ϯ 15 0.0020 Ϯ 0.0009 0.010 Ϯ 0.004 1.5 Ϯ 0.8 ϫ 10 3.0 Ϯ 1.2 ϫ 10 4 3 GDP-D-Man 2.1 Ϯ 0.2 0.89 Ϯ 0.37 0.042 Ϯ 0.003 0.33 Ϯ 0.16 4.9 Ϯ 0.7 ϫ 10 2.9 Ϯ 0.4 ϫ 10 5 4 GDP-L-Gal 5.1 Ϯ 0.3 0.40 Ϯ 0.28 0.012 Ϯ 0.002 0.011 Ϯ 0.008 4.3 Ϯ 0.9 ϫ 10 4.0 Ϯ 1.8 ϫ 10 Inorganic phosphate 28.1 Ϯ 0.1 380 Ϯ 27 2.9 Ϯ 0.6 1.2 Ϯ 0.2 9.9 Ϯ 2.1 ϫ 103 312 Ϯ 37 ϫ 103

were measured by HPLC-based methods that simultaneously The high substrate affinity of C15orf58 and C10F3.4 for Downloaded from monitor the consumption of GDP-D-Glc and the formation of GDP-D-Glc made it relatively difficult to accurately determine the possible products (GMP, GDP, GTP, GDP-L-Gal, and GDP- the kinetic parameters of their GDP-D-Glc phosphorylase D-Man, respectively). Despite the replacement of the Gln resi- activity by the HPLC method; the product (GDP) concentra- due of the VTC2 HIT motif by a His residue in the HIT motifs of tions to be measured under initial velocity conditions at GDP-

the human and worm proteins, the latter did not hydrolyze D-Glc concentrations below and around the Km values of these www.jbc.org GDP-D-Glc. In fact, as for the VTC2 enzyme, the human and enzymes for GDP-D-Glc were close to the detection limit of the worm VTC2 homologs only acted on GDP-D-Glc in the pres- assay. We therefore developed an enzymatic spectrophotomet-

ence of Pi (supplemental Fig. S1), and GDP-D-Glc consumption ric assay to monitor D-Glc-1-P formation from GDP-D-Glc by was accompanied by a concomitant formation of stoichiomet- coupling this reaction to those catalyzed by phosphoglucomu- at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 ϩ ric amounts of GDP (data not shown). These results showed tase and glucose-6-P dehydrogenase in the presence of NAD . that the mammalian and invertebrate VTC2 homologs are, like In this assay, the D-Glc-1-P formed is successively converted to the plant enzyme, HIT nucleotide phosphorylases and not D-Glc-6-P and then to 6-phospho-D-glucono-␦-lactone with ϩ hydrolases. concomitant reduction of NAD to NADH, which is detected We then determined the kinetic parameters of these human by its UV absorbance at 340 nm. The Km values for GDP-D-Glc and worm phosphorylases for GDP-D-Glc, GDP-D-Man, and obtained by this method for C15orf58 and C10F3.4 (3.4 Ϯ 1.0 GDP-L-Gal using the HPLC and/or the spectrophotometric and 10 Ϯ 4 ␮M, respectively) were similar to the values obtained assays described under “Experimental Procedures” (Table 1). with the HPLC method (2.0 Ϯ 0.9 and 18 Ϯ 11 ␮M, respec-

For both enzymes, we found low micromolar Km values and tively). Similar kcat values were also obtained for GDP-D-Glc high catalytic efficiencies with GDP-D-Glc as a substrate. As with human C15orf58 using the spectrophotometric assay Ϫ1 Ϫ1 opposed to plant VTC2, which phosphorolyzes GDP-D-Glc and (16 Ϯ 0.4 s ) and the HPLC method (26 Ϯ 7s ). However, GDP-L-Gal with similar catalytic efficiencies (1), the catalytic using the HPLC method, a non-Michaelis-Menten decrease in efficiencies of the human and worm enzymes for GDP-L-Gal activity was observed at higher GDP-D-Glc concentrations with were 35- and 750-fold lower, respectively, than for GDP-D-Glc. the worm C10F3.4 enzyme (but not with the human C15orf58 GDP-D-Man was, as for plant VTC2 (1), a poor substrate for the enzyme). This inhibitory effect was not observed using the human and worm phosphorylases. Also, as for plant VTC2 (24), spectrophotometric assay, suggesting that it might be due to affinities in the low millimolar range were found with the product inhibition and making the spectrophotometric assay

human and worm enzymes for the nucleotide acceptor Pi. our preferred method for kinetic analyses of the GDP-D-Glc These results show that the human C15orf58 and worm phosphorylase activity of C10F3.4. Furthermore, the use of C10F3.4 proteins act as specific and highly efficient GDP-D-Glc these two different methods allowed us to confirm the identity

phosphorylases. The enzymatic activity of C10F3.4 was also of the two products formed by the Pi-dependent conversion of measured in the reverse direction by incubating the protein in GDP-D-Glc by C15orf58 and C10F3.4 as D-Glc-1-P (spectro- the presence of D-Glc-1-P and GDP, ADP, UDP, or CDP and photometric assay) and GDP (HPLC method).

assaying NDP-D-Glc formation in the absence of added Pi (sup- GDP-D-Glc Phosphorylase Activity in Mouse Tissue and plemental Table S1). Significant NDP-D-Glc formation was Worm Extracts—GDP-D-Glc phosphorylase activity was also measured in these reactions in the presence of GDP. More than detected in mouse tissue and C. elegans extracts. The activity

10-fold lower activities were detected in the presence of UDP, levels measured in these extracts as the Pi-dependent GDP for- Ϫ1 Ϫ1 and no activity was found in the presence of CDP. Relatively mation from GDP-D-Glc (0.5–3 nmol min mg of protein ; high ADP-D-Glc levels were measured after incubating D-Glc- see Table 2 and Figs. 1 and 3) were of the same magnitude as 1-P in the presence of ADP, but this reaction was not dependent the GDP-L-Gal phosphorylase activity measured in Arabidopsis on the presence of the C10F3.4 enzyme (supplemental Table whole plant extracts (24). Under the assay conditions described S1). These results demonstrate the specificity of the enzyme for under “Experimental Procedures,” GDP formation measured in the guanosine nucleotide at least in the presence of D-Glc-1-P. the desalted protein extracts in the presence of GDP-D-Glc but

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21515 89 Molecular Identification of GDP-D-glucose Phosphorylase

in the absence of externally added Pi was low (less than 0.2 nmol These results indicate that mammalian and worm tissues con- Ϫ1 Ϫ1 min mg of protein ). Recapitulating the substrate specific- tain a GDP-D-Glc phosphorylase activity with properties simi- ity for the recombinant C15orf58 and C10F3.4 proteins, a pref- lar to those of the recombinant C15orf58 and C10F3.4 enzymes. erence for GDP-D-Glc over GDP-L-Gal and GDP-D-Man was Confirmation of Identity of C10F3.4 and C15orf58 as C. also found in mouse brain extracts (Table 2) and in C. elegans elegans and Human GDP-D-Glc Phosphorylases—To confirm extracts (supplemental Table S2) when measuring Pi-depen- that the human and C. elegans GDP-D-Glc phosphorylases are dent GDP formation from these GDP-hexoses. No (or very low) encoded by the C15orf58 and C10F3.4 genes, respectively, we phosphorylase activities were detected for human C15orf58 measured this enzymatic activity in extracts of cells where the (Table 2) or worm C10F3.4 (supplemental Table S2) recombi- expression of these genes was eliminated or reduced. We used a nant proteins with other NDP-hexoses including GDP-L-Fuc, mutant C. elegans strain (tm2679) that has a 444-bp deletion UDP-D-Glc, UDP-D-Gal, and ADP-D-Glc. Similarly, little or no within the C10F3.4 gene. This deletion eliminates exons 6 to 7 phosphorylase activity toward GDP-L-Fuc, UDP-D-Glc, UDP- and parts of exons 5 and 8 and introduces a premature stop D-Gal, and ADP-D-Glc was detected in mouse brain extracts codon. The mutated allele encodes a predicted truncated 246- (Table 2) and C. elegans extracts (supplemental Table S2). amino acid protein that lacks the C-terminal moiety of the 458-

TABLE 2 amino acid wild-type C10F3.4 protein (isoform a). Whereas Downloaded from Comparison of substrate specificity of human recombinant C15orf58 GDP-D-Glc phosphorylase activity could readily be measured and of mouse brain GDP-D-Glc phosphorylase in wild-type C. elegans extracts, no such activity could be Various sugar nucleotides (50 ␮M) were incubated in the presence of C15orf58 detected in extracts prepared from mutant tm2679 worms (Fig. recombinant enzyme (0.011 ␮g/ml) for 25 min or of desalted mouse brain extracts (0.3–0.5 mg of protein/ml) for 20 min at 31 °C. NDP-hexose consumption and NDP 1A). These mutant worms, however, did not present any appar- production were measured by HPLC as described under “Experimental Proce-

ent abnormal phenotype in the standard culture conditions www.jbc.org dures.” Phosphorylase activities were calculated based on the Pi-dependent NDP D used throughout this study. formation. With GDP- -Glc as a substrate, the specificϪ activity of theϪ human recombinant C15orf58 enzyme was 29.4 Ϯ 3.2 ␮mol min 1 mg of protein 1 (100%); Ϯ For the human enzyme, we knocked down the C15orf58 gene the correspondingϪ 100% specificϪ activity of the mouse brain extract was 1.9 0.7 nmol min 1 mg of protein 1 (specific activities are means Ϯ S.D., n ϭ 2). The in HEK293T cells by transfection with an siRNA pool specifi- phosphorylase activities found with the other sugar nucleotides are given as a per- cally targeting this gene. As shown in Fig. 1B, a reduction of at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 centage of the activity found with GDP-D-Glc. Values represent the means of two individual experiments for each substrate. about 50% in C15orf58 transcript level was obtained in cells Relative specific activity treated with C15orf58 siRNAs compared with cells treated with Recombinant Mouse non-targeting control siRNA. This decrease in mRNA expres- Substrate human enzyme brain extract sion level was accompanied by an ϳ2-fold decrease in GDP-D- % Glc phosphorylase activity in extracts derived from these cells GDP-D-Glc 100 100 GDP-L-Gal 23 27 (Fig. 1B). These findings establish the molecular identity of the GDP-D-Man 2 5 C. elegans and human GDP-D-Glc phosphorylases as corre- GDP-L-Fuc 0 0 UDP-D-Glc 1 5 sponding to the expression products of the C10F3.4 and UDP-D-Gal 0 3 C15orf58 genes, respectively. ADP-D-Glc 0 1

FIGURE 1. GDP-D-Glc phosphorylase activity in C. elegans C10F3.4 deletion strain and in HEK293T cells treated with C15orf58 siRNA. Extracts were prepared from wild-type (N2) or C10F3.4 mutant (tm2679) C. elegans strains (A) and from siRNA-transfected HEK293T cells (B), and GDP-D-Glc phosphorylase activity was measured by the HPLC assay described under “Experimental Procedures.” GDP-D-Glc was added to the reaction mixtures at a final concentration of 50 ␮M. The worm and HEK293T cell extracts were diluted in the reaction mixtures to final protein concentrations of 0.5–1.4 and 0.3–0.4 mg/ml, respectively. Control reactions were run in the absence of sodium phosphate, and GDP-D-Glc phosphorylase activities were calculated based on the Pi-dependent formation of GDP. A, worms were grown for 4–7 days at 20 °C in liquid culture and harvested for protein extraction. B, HEK293T cell cultures were stopped 48–72 h after siRNA transfection. In addition to GDP-D-Glc phosphorylase activity assays (black bars), quantitative RT-PCR was performed on cDNA derived from these cells. C15orf58 mRNA levels were normalized to those of ACTB (␤-actin; gray bars) or GAPDH (white bars), and -fold changes in C15orf58 expression in knockdown versus control cells were calculated using the 2ϪddCt method. The results shown represent the means Ϯ S.D. of two to three (A) or four (B) biological replicates.

21516 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 24•JUNE 17, 2011 90 Molecular Identification of GDP-D-glucose Phosphorylase Downloaded from www.jbc.org at UCLA-Louise Darling Biomed. Lib., on May 20, 2012

FIGURE 2. C10F3.4 expression pattern in C. elegans. A transgenic C. elegans strain (UZ119) expressing a C10F3.4::GFP fusion protein was generated as described under “Experimental Procedures.” The differential interference contrast images are presented in the left panels, GFP images are in the middle panels, and merged images are in the right panels. C10F3.4::GFP expression is seen in head neurons (arrows)(A), in neuronal cells throughout the ventral nerve cord and in tail (lumbar) ganglia (arrow)(B), in the spermatheca (arrow)(C), and in anterior hypodermis cells (arrow in inset representing a magnified region) (D). The bottom panels show a worm that lost the extrachromosomal array containing the C10F3.4::GFP transgene (E, negative control). Transgenic animals were analyzed for GFP expression at 40ϫ power magnification.

Tissue Distribution of C. elegans and Mouse GDP-D-Glc rior hypodermal cells. The high neuronal expression was found Phosphorylases—As a first step in elucidating the physiological in all the normal larval stages observed. C10F3.4 expression was role of the GDP-D-Glc phosphorylase enzyme in worm and also detected in developing embryos. mammalian tissues, we determined the tissue distribution of To examine the tissue distribution of the GDP-D-Glc phos- this protein in live worms as well as of the corresponding activ- phorylase in mammals, we measured enzyme activity in ity in mouse extracts. A transgenic C. elegans strain (UZ119) extracts of various mouse tissues. All of the analyzed tissues expressing a C10F3.4::GFP fusion protein under the endoge- contained GDP-D-Glc phosphorylase activity, but the highest nous C10F3.4 promoter was used to determine in which cell levels were reproducibly measured in brain and testis (4–8 types and tissues this protein is expressed in nematodes. times higher activities than in the remaining tissues; Fig. 3). We Remarkably, significant expression of this protein seemed to be also measured mRNA expression of the C15orf58 gene (desig- mostly confined to neurons and parts of the reproductive sys- nated D330012F22Rik in mouse) in various mouse tissues using tem. As shown in Fig. 2, cytosolic expression of the C10F3.4 real time quantitative PCR. The mRNA distribution pattern protein was detected throughout the neuronal system. In addi- revealed an up-regulation of the mouse C15orf58 homolog tion, expression was observed in the spermatheca and in ante- mRNA in brain and testis relative to kidney, liver, and heart

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21517 91 Molecular Identification of GDP-D-glucose Phosphorylase

intensity ratio found for GDP-D-Man and GDP-D-Glc corre- sponded to 3.4 and 0.13, respectively, over the range of GDP- hexose concentrations analyzed in this study by the IP-LC- ESI-MS method. For quantification, standard curves were generated for GDP-D-Man and GDP-D-Glc by injecting 10–100 and 0.1–4 pmol of these compounds, respectively, in a volume of 10 ␮l and by determining the corresponding product ion 604 3 158 (GDP-D-Man) and 604 3 362 (GDP-D-Glc) peak areas of the extracted ion current (supplemental Fig. S3). We found that amounts of GDP-D-Glc corresponding to as low as 0.05 pmol could readily be detected by this method. We then quantified GDP-D-Man and GDP-D-Glc in neutralized perchloric acid extracts prepared from wild-type (N2) and tm2679 mutant C. elegans worms grown in liquid culture and FIGURE 3. Tissue distribution of GDP-D-Glc phosphorylase activity in from a series of mouse tissues as described under “Experimen- Downloaded from mouse. Mouse tissue extracts were prepared, and GDP-D-Glc phosphorylase tal Procedures” (Table 3). Similar amounts of GDP-D-Man were activities were measured by HPLC as described under “Experimental Proce- dures.” GDP-D-Glc and mouse tissue extracts were added to the reaction mix- found in tm2679 and N2 worms (7.7 versus 8.9 nmol/g of tures at final concentrations of 50 ␮M and 0.3–1.5 mg of protein/ml, respec- worms), whereas more than 5-fold higher amounts of GDP-D- tively. Control reactions were run in the absence of sodium phosphate, and Glc were detected in the C10F3.4 deletion strain versus wild- GDP-D-Glc phosphorylase activities were calculated based on the Pi-dependent formation of GDP. The results shown represent the means Ϯ S.D. of three type worms (0.11 versus 0.02 nmol/g of worms). It should be

to five biological replicates. noted that the GDP-D-Glc levels found in wild-type worms www.jbc.org were actually close to the detection limit of our assay. This (supplemental Fig. S2), thus supporting the distribution pattern observation indicated that the best substrate found in vitro for obtained via GDP-D-Glc phosphorylase activity measurements. the C10F3.4 phosphorylase is a physiological substrate for this Taken together, these results also demonstrated that, in addi- enzyme in worms. Concerning wild-type mouse tissues, GDP- at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 tion to displaying similar substrate specificity and kinetic prop- D-Man concentrations varying from about 7 to 25 nmol/g of erties, the C. elegans and mammalian GDP-D-Glc phosphory- tissue (wet weight) were measured; liver was the organ with the lases share similar expression profiles with highest occurrence highest levels of this GDP-hexose (Table 3). These values were in the nervous and male reproductive systems. similar to GDP-D-Man concentrations found previously in rat GDP-D-Glc Is Present at Low Levels in Nematode and tissues (26). GDP-D-Glc levels were very low in all of the mouse Mouse Tissues but Accumulates in Absence of GDP-D-Glc tissues analyzed. The highest GDP-D-Glc levels were measured Phosphorylase—Data on the presence and formation of GDP- in liver (0.15 nmol/g of tissue) where GDP-D-Glc concentra- D-Glc in living organisms are scarce, but GDP-D-Glc has been tions were found to be more than 100-fold lower than GDP-D- isolated from mammary gland (7). Furthermore, a GDP-D-Glc Man concentrations. The GDP-D-Glc levels detected in brain, pyrophosphorylase activity, forming GDP-D-Glc from GTP and kidney, testis, and heart were 15–30-fold lower than in liver. D-Glc-1-P in the reverse reaction, has been measured in Except for liver, all of the GDP-D-Glc levels found in mouse extracts prepared from various mammalian (7, 8) and plant tissues were near or below the detection limit of our assay. sources (25). Although this activity could be clearly distin- Taken together, these results suggest that GDP-D-Glc levels are guished from UDP-D-Glc pyrophosphorylase, it has most often normally maintained at very low levels in cells. The only situa- been found in association with GDP-D-Man pyrophosphorylase tion in which we found more significant GDP-D-Glc levels was activity (8–10) and may thus result from a lack of substrate in extracts of a C. elegans mutant of the C10F3.4 GDP-D-Glc specificity of the enzyme that forms GDP-D-Man for glycocon- phosphorylase activity. We conclude that a physiological func- jugate synthesis. We thus developed methods to directly mea- tion of GDP-D-Glc phosphorylase in nematode cells is to sure GDP-D-Glc and GDP-D-Man in cell extracts. degrade GDP-D-Glc to D-Glc-1-P and hypothesize that such an We first set up a specific and sensitive assay of these GDP- activity is needed to prevent misincorporation of glucose for hexoses that was adapted from a previously published ion pair mannose into glycoconjugates (Fig. 5). The situation in mam- liquid chromatography coupled to electrospray ionization mass malian cells may be similar. spectrometry method (21) as described under “Experimental DISCUSSION Procedures.” This method allowed us to distinguish between GDP-D-Man and GDP-D-Glc both by elution time and by the In an effort to understand the function of apparent orthologs relative abundance of product ions m/z 604 3 158 and m/z of the plant VTC2 enzyme in non-plant organisms, we found 604 3 362 (both generated from the common precursor ion that C. elegans and mammalian tissues contain a specific GDP- m/z 604) (Fig. 4A). GDP-D-Man reproducibly eluted earlier D-Glc phosphorylase enzyme encoded by the C10F3.4 gene in from the LC column than GDP-D-Glc and consistently gener- worms and the C15orf58 gene in humans. As opposed to the ated higher levels of product ion 604 3 158 than of product ion plant VTC2 enzyme, the C10F3.4 and C15orf58 proteins dis- 604 3 362 in the range of GDP-D-Man concentrations ana- play a marked preference for GDP-D-Glc over GDP-L-Gal as a lyzed. GDP-D-Glc on the other hand generated higher levels of substrate. However, like the VTC2 enzyme, the C10F3.4 and the product ion 604 3 362. The 158/362 product ion peak C15orf58 enzymes are also completely dependent on the pres-

21518 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 24•JUNE 17, 2011 92 Molecular Identification of GDP-D-glucose Phosphorylase Downloaded from www.jbc.org at UCLA-Louise Darling Biomed. Lib., on May 20, 2012

FIGURE 4. Accumulation of GDP-D-Glc in C10F3.4 mutant C. elegans strain (tm2679). GDP-D-Man and GDP-D-Glc standards (A) and wild-type N2 and mutant tm2679 small molecule extracts (B) were analyzed by the IP-LC-ESI-MS method described under “Experimental Procedures.” A, extracted ion chromatograms of product ions m/z 604 3 158 (middle panel, white area) and m/z 604 3 362 (bottom panel, black area) obtained after injection of a solution containing 0.2 pmol of GDP-D-Man and 0.2 pmol of GDP-D-Glc. GDP-D-Man and GDP-D-Glc produced the same precursor ion, m/z 604, and the same product ions, m/z 604 3 158 and m/z 604 3 362. These GDP-hexoses, however, displayed different retention times with GDP-D-Man eluting first and different relative peak intensities for the product ions. The top panel shows an overlay of the extracted ion chromatograms of product ions m/z 604 3 158 (white area) and m/z 604 3 362 (black area) presented separately in the middle and bottom panels. B, extracted ion chromatograms of product ions m/z 604 3 158 (middle panels, white areas) and m/z 604 3 362 (bottom panels, black areas) after injection of equal volumes of C. elegans N2 and tm2679 perchloric acid extracts. The top panels show overlays of the corresponding middle and bottom panels. The arrows indicate the peak identified as GDP-D-Glc and accumulating in C10F3.4 mutant worms. Representative chromatograms obtained with one set of wild-type and mutant extracts are shown. Four other sets of wild-type and tm2679 mutant extracts were analyzed with similar results. The GDP-D-Man and GDP-D-Glc levels found in these worm extracts were quantified, and the values obtained are given in Table 3. cps, counts/s.

ence of inorganic phosphate to act on GDP-D-Glc, thereby con- thermore, A. thaliana Hint4 and C. elegans DcpS, two HIT verting it into D-Glc-1-P and GDP. The identity of the reaction proteins containing the HXHXH motif, have been shown to products was confirmed by HPLC (for GDP) and by an enzy- display a phosphorolytic activity toward adenosine 5Ј-phos- matic spectrophotometric assay (for D-Glc-1-P). phosulfate in addition to their hydrolytic activity on this sub-

The Pi dependence of the C10F3.4 and C15orf58 enzymes strate (28). The phosphorylase activity of the latter two was somewhat unexpected given that the Gln residue of the enzymes, however, was minor compared with their hydrolase VTC2 “HIT motif” (HXHXQ) was replaced by a His residue in activity at neutral pH. Taken together, these results suggest that the HIT motifs of both the worm and human VTC2 homolog critical residues other than the C-terminal residue of the HIT sequences (HXHXH). So far, the HXHXH motif has been pro- motif (His or Gln) must be involved in determining whether the posed to be characteristic of HIT nucleotide hydrolases, nucleotidylated enzyme intermediate that is formed during whereas the HXHXQ motif has been considered as predictive of HIT enzyme-catalyzed reactions is resolved via hydrolysis or a nucleotide transferase or phosphorylase activity (11, 12). The phosphorolysis. present study and two other recent reports indicate that this Concerning the nucleotide moiety of the nucleotide sugar “rule” has exceptions. In addition to C10F3.4 and C15orf58, a substrate, recombinant C10F3.4 and C15orf58 appear to be mycobacterial protein (Rv2613c) containing a HXHXH motif specific for GDP as none of the ADP- or UDP-hexoses tested has also recently been reported to act as a phosphorylase (on were phosphorolyzed by either of these proteins and as GDP diadenosine tetraphosphate) and not as a hydrolase (27). Fur- could not be replaced in the reverse reaction by ADP, UDP,

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21519 93 Molecular Identification of GDP-D-glucose Phosphorylase

TABLE 3 clude that if ascorbate is formed in this species it is not via a GDP-D-Man and GDP-D-Glc levels in C. elegans and various mouse L tissues plant-like -galactose pathway involving the C10F3.4 enzyme. Our characterization of the substrate specificity and kinetic GDP-D-Man and GDP-D-Glc levels were determined in neutralized perchloric acid extracts of wild-type (N2) and C10F3.4 mutant (tm2679) worms as well as of various properties of the C10F3.4 and C15orf58 proteins and the mouse tissues by using the IP-LC-ESI-MS method described under “Experimental observed accumulation of GDP-D-Glc in C10F3.4-deficient Procedures.” GDP-hexose concentrations were determined based on peak areas of selected product ions (604 3 362 for GDP-D-Glc and 604 3 158 for GDP-D-Man) worms indicate that GDP-D-Glc is a physiological substrate of and calibration curves established with standard GDP-D-Man and GDP-D-Glc solu- these enzymes. Why would invertebrates and vertebrates have tions. Metabolite concentrations were normalized against wet worm or mouse tissue weight. Values are means Ϯ S.D. of five (C. elegans) or two (mouse tissue) conserved an enzyme that breaks down GDP-D-Glc into D-Glc- biological replicates. Statistical analysis using a two-tailed unpaired Student’s t test 1-P and GDP? The literature on GDP-D-Glc is scarce, but the shows a statistically significant difference in GDP-D-Glc concentration and GDP- D-Glc/GDP-D-Man ratio between the N2 and tm2679 C. elegans strains as demon- presence of this GDP-hexose was reported in the early 1960s in strated by the indicated p values. lactating bovine mammary gland (7), and an enzymatic activity C. elegans forming GDP-D-Glc from GTP and D-Glc-1-P (GDP-D-Glc Strain GDP-D-Man GDP-D-Glc GDP-Glc/GDP-Man pyrophosphorylase) has also been found in various plant (25) Ϫ nmol/g worms nmol/g worms ϫ10 3 and animal tissues (8, 9). At least in mammals, however, this N2 8.9 Ϯ 5.4 0.020 Ϯ 0.009 2.3 Ϯ 0.7 Ϯ Ϯ Ϯ activity has so far always been found in association with GDP- tm2679 7.7 4.6 0.11 0.07 13.8 2.2 Downloaded from (p ϭ 0.029) (p ϭ 0.000004) D-Man pyrophosphorylase. Nothing is known about possible Mouse functions of GDP-D-Glc in mammals or invertebrates, but Tissue GDP-D-Man GDP-D-Glc GDP-Glc/GDP-Man GDP-D-Glc acts as a glucosyl donor for the synthesis of cell wall Ϫ nmol/g tissue nmol/g tissue ϫ10 3 polysaccharides in various plant species (29–32) and for treha- Liver 25 Ϯ 2 0.15 Ϯ 0.01 6.1 Ϯ 0.1 lose synthesis in certain bacterial species (33). Brain 13 Ϯ 0.2 0.0077 Ϯ 0.0012 0.6 Ϯ 0.02 Mammalian GDP-D-Man pyrophosphorylase has been www.jbc.org Kidney 6.8 Ϯ 0.4 0.0096 Ϯ 0.0002 1.4 Ϯ 0.06 Testis 11 Ϯ 0.1 0.0052 Ϯ 0.0016 0.48 Ϯ 0.14 purified from rat mammary gland and calf liver (8), from pig Heart 7.8 Ϯ 0.5 0.0055 Ϯ 0.0018 0.70 Ϯ 0.27 thyroid (34), and from pig liver (9). Although the thyroid enzyme was specific for GDP-D-Man, both the mammary or CDP for transfer to D-Glc-1-P. As to the sugar moiety of gland and liver enzymes have been shown to use GDP-D-Man at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 the GDP-sugar substrate, the C10F3.4 and C15orf58 as well as GDP-D-Glc as substrates. Remarkably, the native enzymes displayed a marked preference for D-Glc over L-Gal, pig liver enzyme was actually about 2 times more active in D-Man, and L-Fuc. A similar substrate specificity was found forming GDP-D-Glc than in the synthesis of GDP-D-Man for the GDP-hexose phosphorylase activity detected in C. (10). This enzyme is composed of two different subunits (␣- elegans and mouse tissue extracts. This finding as well as the and ␤-subunits) encoded by two different genes. Only the absence of GDP-D-Glc phosphorylase activity in mutant gene encoding the ␤-subunit has so far been cloned and worms lacking a functional C10F3.4 gene and decreased expressed (10). The expression product of this gene was GDP-D-Glc phosphorylase activity in C15orf58 siRNA shown to be a functional GDP-D-Man pyrophosphorylase knockdown human cells led us to conclude that C10F3.4 and that, unlike the native enzyme, is about 6 times more active C15orf58 encode for the C. elegans and human GDP-D-Glc in forming GDP-D-Man than GDP-D-Glc. Based on this, it phosphorylase enzymes, respectively. was suggested that the ␣-subunit could be the GDP-D-Glc A link of the mammalian VTC2 homolog to vitamin C syn- pyrophosphorylase or a regulatory subunit that affects the thesis seemed very unlikely because mammals (except for a few substrate specificity of the ␤-subunit. It is of interest to note non-ascorbate-synthesizing species including humans) synthe- here that genes encoding homologs of the ␣-and␤-subunits size ascorbate via a pathway that is different from the plant of GDP-D-Man pyrophosphorylase are also found in the C. pathway and that does not involve GDP-L-Gal as an intermedi- elegans genome. On the contrary, no gene encoding an ate (6). However, the presence of a VTC2-homologous gene in ␣-subunit is found in the genome of Saccharomyces cerevi- C. elegans raised questions concerning a possible existence of a siae,andtheGDP-D-Man pyrophosphorylase of this species, plant-like ascorbate biosynthesis pathway in this species. The which presents high similarity with mammalian GDP-D- finding that the C10F3.4 enzyme was more than 700-fold less Man pyrophosphorylase ␤,hasbeenreportedtobequite efficient in phosphorolyzing GDP-L-Gal than GDP-D-Glc, how- specific for GDP-D-Man (10). The S. cerevisiae genome does ever, was a first indication that this enzyme is not likely to par- not encode a VTC2 homolog either. Taken together, these ticipate in a plant-like vitamin C synthesis pathway in C. results suggest a link between the formation of GDP-D-Glc elegans. Interestingly, we could actually detect significant by a second activity of the enzyme that normally makes amounts of ascorbate in C. elegans worms, but similar levels GDP-D-Man and the presence of a GDP-D-Glc phosphoryl- 5 were found in wild-type and C10F3.4 mutant worms, again ase that specifically degrades the glucose-containing arguing against C10F3.4 involvement in ascorbate formation in nucleotide sugar. In favor of this, we found a similar tissue worms. Combining these observations with the absence of any distribution pattern of GDP-D-Glc phosphorylase and GDP- obvious orthologs of critical enzymes of the plant vitamin C D-Man pyrophosphorylase ␣ (Gmppa) at the mRNA level in biosynthesis pathway (e.g. GDP-D-mannose 3,5-epimerase and mice (supplemental Fig. S2). L-galactonolactone dehydrogenase) in C. elegans (1), we con- GDP-D-Man is the major mannosyl donor for the synthesis of N-linked glycoproteins and glycosylphosphatidylinosi- 5 B. D. Young and L. N. Adler, unpublished results. tol membrane anchors (35). It is also the precursor of GDP-

21520 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 24•JUNE 17, 2011 94 Molecular Identification of GDP-D-glucose Phosphorylase Downloaded from www.jbc.org at UCLA-Louise Darling Biomed. Lib., on May 20, 2012

FIGURE 5. Possible role of GDP-D-Glc phosphorylase in preventing misincorporation of glucose in place of mannose residues into glycoconjugates. In the presence of GTP and D-Man-1-P, GDP-D-Man pyrophosphorylase forms GDP-D-Man, a major mannosyl donor for glycoconjugation. The native pyrophosphorylase is composed of ␣- and ␤-subunits (GMPPA and GMPPB) and has been shown to also catalyze the formation of GDP-D-Glc when D-Man-1-P is replaced by D-Glc-1-P. However, we show in this study that mammalian tissues contain only very low levels of GDP-D-Glc. In addition, this nucleotide sugar accumulates in mutant worms lacking GDP-D-Glc phosphorylase, the enzyme described in the present work. We therefore propose that GDP-D-Glc phosphorylase functions to compensate for the lack of specificity of GDP-D-Man pyrophosphorylase by degrading the apparently useless GDP-D-Glc. This may be important to prevent misincorporation of glucose residues for mannose residues into oligosaccharide chains such as those linked to dolichol (this figure) and eventually transferred to proteins. GDP-D-Glc phosphorylase might thus be necessary to preserve functional protein N-glycosylation or other glycoconjugation processes. ER, endoplasmic reticulum.

L-Fuc, another important glycosyl donor for glycoprotein that GDP-D-Glc phosphorylase functions as a metabolite biosynthesis (35). In various bacterial, lower eukaryotic, and repair enzyme. plant species, GDP-D-Man donates mannosyl residues for The concept of metabolite repair is based on the observa- cell wall polysaccharides (32, 36, 37). Given the close struc- tion that certain enzymes are not as specific as generally tural similarity of GDP-D-Man and GDP-D-Glc (they only stated and sometimes form “unwanted” side products by differ by the configuration of their hydroxyl group on C2 of acting on cellular compounds different from their main sub- the sugar moiety), it can be hypothesized that some of the strate (39). A metabolite repair enzyme functions to specifi- enzymes using GDP-D-Man are not entirely specific for their cally remove such unwanted side products. L-2-Hydroxy- substrate and can use GDP-D-Glc instead (or be competi- glutarate dehydrogenase is the first example of a metabolite tively inhibited by the latter). In favor of this, glycosyltrans- repair enzyme that has been molecularly identified; it acts to ferases that use nucleotide sugars as the glycosyl donors remove the toxic product formed by L-malate dehydrogenase seem to be often particularly specific for the nucleotide por- when this enzyme acts on ␣-ketoglutarate instead of its main tion of their substrate while having some tolerance for dif- substrate oxaloacetate (40). The physiological importance of ferent sugars (38). This lack of specificity could be indirectly this enzyme is demonstrated by the severe consequences of corrected by keeping the concentrations of GDP-D-Glc low its deficiency, which leads to a neurometabolic disease enough for not entering in competition with GDP-D-Man. known as L-2-hydroxyglutaric aciduria (39). GDP-D-Glc As described in this study, GDP-D-Glc concentrations are phosphorylase could serve to compensate for the lack of indeed very low in wild-type worm and mouse tissues, which, specificity of GDP-D-Man pyrophosphorylase, which, as in addition to the apparent absence of any physiological described above, is quite active in forming GDP-D-Glc in function of GDP-D-Glc in these species, supports the idea vitro (10). It seems surprising that such an activity would

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21521 95 Molecular Identification of GDP-D-glucose Phosphorylase

have been conserved if there is no need for GDP-D-Glc in the In conclusion, we identified a novel enzyme, GDP-D-Glc cell. It is therefore tempting to speculate that this enzyme is phosphorylase, potentially involved in metabolite repair in regulated intracellularly in a way that this GDP-D-Glc-form- nematodes and mammals (Fig. 5). Ongoing work is designed to ing activity of GDP-D-Man pyrophosphorylase is inhibited to better understand the molecular basis of the GDP-D-Glc-form- acertainextent,potentiallyviaitscompositionin␣-and ing activity of GDP-D-Man pyrophosphorylase and to test ␤-subunits. The remaining D-Glc-1-P guanylyltransferase whether GDP-D-Glc phosphorylase is needed for functional activity could then be indirectly corrected by GDP-D-Glc glycoprotein biosynthesis. phosphorylase through removal of the GDP-D-Glc formed (see Fig. 5). It can be noted here that the activity levels Acknowledgments—We are indebted to Dr. Colin Thacker of the reported for D-Glc-1-P guanylyltransferase in mammalian Worm Core Facility (University of Utah) for generating the transgenic tissues (8) are in the same range as the activity levels found in C10F3.4::GFP fusion C. elegans strain. We sincerely thank Dr. Shohei Mitani (Tokyo Women’s Medical University, Tokyo, Japan) for pro- this study for GDP-D-Glc phosphorylase in mammalian tis- Ϫ1 viding the C10F3.4 mutant strain (tm2679) and Prof. Shinichi Kita- sues; in liver for example, a value of about 0.6 nmol min mg Ϫ mura (Osaka Prefecture University, Osaka, Japan) for kindly provid- of protein 1 has been found for both of these activities. ing GDP-L-galactose. We also thank our UCLA colleagues Drs. Alison D Combined with the high affinity for GDP- -Glc that we Frand and Lars Dreier for sharing expertise and equipment for Downloaded from determined for GDP-D-Glc phosphorylase, this enzyme has microscopy, Dr. Jonathan Lowenson for help in preparing mouse the required properties to efficiently remove the apparently extracts, and Dr. Jonathan Wanagat for advice on quantitative PCR useless GDP-hexose. data analysis. Finally, we thank Prof. Emile Van Schaftingen (de Duve In this metabolite repair hypothesis, a deficiency in GDP- Institute, Brussels, Belgium) for critical reading of the manuscript. D-Glc phosphorylase activity could lead to GDP-D-Glc accumulation as we observed in worms deficient in this en- www.jbc.org zyme and subsequent erroneous incorporation of glucosyl REFERENCES units instead of mannosyl units into oligosaccharide chains 1. Linster, C. L., Gomez, T. A., Christensen, K. C., Adler, L. N., Young, B. D., Brenner, C., and Clarke, S. G. (2007) J. Biol. Chem. 282, 18879–18885 (Fig. 5). If GDP-D-Glc is formed in vivo by GDP-D-Man pyro- 2. Laing, W. A., Wright, M. A., Cooney, J., and Bulley, S. M. (2007) Proc. Natl. at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 phosphorylase, the cytosolic mannosyltransferases (includ- Acad. Sci. U.S.A. 104, 9534–9539 ing ALG1, ALG2, and ALG11) involved in the protein N- 3. Dowdle, J., Ishikawa, T., Gatzek, S., Rolinski, S., and Smirnoff, N. (2007) glycosylation pathway should be most prone to encounter Plant J. 52, 673–689 accumulating GDP-D-Glc molecules. These cytosolic man- 4. Linster, C. L., and Clarke, S. G. (2008) Trends Plant Sci. 13, 567–573 5. Bulley, S. M., Rassam, M., Hoser, D., Otto, W., Schu¨nemann, N., nosyltransferases successively add five mannosyl residues Wright, M., MacRae, E., Gleave, A., and Laing, W. (2009) J. Exp. Bot. from GDP-D-Man to dolichol-PP-(GlcNAc)2.Thedolichol- 60, 765–778 PP-(GlcNAc)2Man5 intermediate is then flipped to the endo- 6. Linster, C. L., and Van Schaftingen, E. (2007) FEBS J. 274, 1–22 plasmic reticulum lumen where this lipid-linked oligosac- 7. Carlson, D. M., and Hansen, R. G. (1962) J. Biol. Chem. 237, 1260–1265 charide (LLO) is further extended and finally transferred to 8. Verachtert, H., Rodriguez, P., Bass, S. T., and Hansen, R. G. (1966) J. Biol. Chem. 241, 2007–2013 Asn residues on newly synthesized polypeptide chains (35). 9. Szumilo, T., Drake, R. R., York, J. L., and Elbein, A. D. (1993) J. Biol. Chem. It is difficult to predict the consequences of the replacement 268, 17943–17950 of mannosyl residues by glucosyl residues in LLO, but, as for 10. Ning, B., and Elbein, A. D. (2000) Eur. J. Biochem. 267, 6866–6874 the numerous inherited deficiencies in LLO assembly (con- 11. Brenner, C. (2002) Biochemistry 41, 9003–9014 genital disorders of glycosylation type I (CDG-I)) already 12. Brenner, C. (2007) Encyclopedia of Life Sciences, John Wiley & Sons, Ltd., Chichester, UK, doi: 10.1002/9780470015902.00020545 identified, this could lead to the production of structurally 13. Watanabe, K., Suzuki, K., and Kitamura, S. (2006) Phytochemistry 67, incomplete LLOs, resulting in hypoglycosylation of multiple 338–346 glycoproteins (35, 41). It will therefore be of great interest to 14. Sulston, J., and Hodgkin, J. (1988) in The Nematode Caenorhabditis el- screen for mutations in the C15orf58 gene in CDG-I patients egans (Wood W. B., ed) pp. 587–606, Cold Spring Harbor Laboratory for which the basic defect has yet to be elucidated (CDG-Ix Press, Cold Spring Harbor, NY 15. Lewis, J. A., and Fleming, J. T. (1995) Methods Cell Biol. 48, 3–29 patients; Ref. 42). The high expression of GDP-D-Glc phos- 16. Brenner, S. (1974) Genetics 77, 71–94 phorylase in the nervous system suggests a greater vulnera- 17. Kagan, R. M., and Clarke, S. (1995) Biochemistry 34, 10794–10806 bility of this organ to GDP-D-Glc accumulation, which is 18. Hobert, O. (2002) BioTechniques 32, 728–730 compatible with the pronounced neurological symptoms 19. Webb, K. J., Lipson, R. S., Al-Hadid, Q., Whitelegge, J. P., and Clarke, S. G. that are very commonly observed in CDG-I patients. As LLO (2010) Biochemistry 49, 5225–5235 is synthesized in C. elegans by the common eukaryotic path- 20. Nihira, T., Nakajima, M., Inoue, K., Nishimoto, M., and Kitaoka, M. (2007) Anal. Biochem. 371, 259–261 way (43), the C10F3.4 mutant strain might provide a good 21. Coulier, L., Bas, R., Jespersen, S., Verheij, E., van der Werf, M. J., and model for the identification of potential glycosylation Hankemeier, T. (2006) Anal. Chem. 78, 6573–6582 defects caused by GDP-D-Glc phosphorylase deficiency. It 22. Chatterjee, I. B. (1973) Science 182, 1271–1272 should be noted, however, that worms appear to be more 23. Massie, H. R., Shumway, M. E., Whitney, S. J., Sternick, S. M., and Aiello, resistant to such defects than humans as no overt pheno- V. R. (1991) Exp. Gerontol. 26, 487–494 24. Linster, C. L., Adler, L. N., Webb, K., Christensen, K. C., Brenner, C., and types have been observed upon RNA interference against Clarke, S. G. (2008) J. Biol. Chem. 283, 18483–18492 certain genes of the LLO synthesis pathway in worms grown 25. Barber, G. A. (1985) FEBS Lett. 183, 129–132 in standard culture conditions (43). 26. Spiro, M. J. (1984) Diabetologia 26, 70–75

21522 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 24•JUNE 17, 2011 96 Molecular Identification of GDP-D-glucose Phosphorylase

27. Mori, S., Shibayama, K., Wachino, J., and Arakawa, Y. (2010) Protein Expr. 35. Varki, A., Cummings, R. D., Esko, J. D., Freeze, H. H., Stanley, P., Purif. 69, 99–105 Bertozzi, C. R., Hart, G. W., and Etzler, M. E. (eds) (2009) Essentials of 28. Guranowski, A., Wojdyła, A. M., Zimny, J., Wypijewska, A., Kowalska, J., Glycobiology,2ndEd.,pp.50–56,104–112,586–592,ColdSpring Jemielity, J., Davis, R. E., and Bieganowski, P. (2010) FEBS Lett. 584, 93–98 Harbor Laboratory Press, Cold Spring Harbor, NY 29. Elbein, A. D. (1969) J. Biol. Chem. 244, 1608–1616 36. Roberts, I. S. (1996) Annu. Rev. Microbiol. 50, 285–315 30. Smith, M. M., Axelos, M., and Pe´aud-Lenoe¨l, C. (1976) Biochimie 58, 37. Behrens, N. H., and Cabib, E. (1968) J. Biol. Chem. 243, 502–509 1195–1211 38. Zea, C. J., and Pohl, N. L. (2005) Biopolymers 79, 106–113 31. Piro, G., Zuppa, A., Dalessandro, G., and Northcote, D. H. (1993) Planta 39. Van Schaftingen, E., Rzem, R., and Veiga-da-Cunha, M. (2009) J. Inherit. 190, 206–220 Metab. Dis. 32, 135–142 32. Liepman, A. H., Wilkerson, C. G., and Keegstra, K. (2005) Proc. Natl. Acad. 40. Rzem, R., Vincent, M. F., Van Schaftingen, E., and Veiga-da-Cunha, M. Sci. U.S.A. 102, 2221–2226 (2007) J. Inherit. Metab. Dis. 30, 681–689 33. Elbein, A. D., Pan, Y. T., Pastuszak, I., and Carroll, D. (2003) Glycobiology 41. Haeuptle, M. A., and Hennet, T. (2009) Hum. Mutat. 30, 1628–1641 13, 17R–27R 42. Jaeken, J. (2010) Ann. N.Y. Acad. Sci. 1214, 190–198 34. Smoot, J. W., and Serif, G. S. (1985) Eur. J. Biochem. 148, 83–87 43. Struwe, W. B., and Warren, C. E. (2010) Methods Enzymol. 480, 477–493 Downloaded from www.jbc.org at UCLA-Louise Darling Biomed. Lib., on May 20, 2012

JUNE 17, 2011•VOLUME 286•NUMBER 24 JOURNAL OF BIOLOGICAL CHEMISTRY 21523 97 SUPPLEMENTAL MATERIALS "A Novel GDP-D-glucose Phosphorylase Involved in Quality Control of the Nucleoside Diphosphate Sugar Pool in Caenorhabditis elegans and Mammals"

Lital N. Adler, Tara A. Gomez, Steven G. Clarke, and Carole L. Linster

Supplementary Figure Legends

FIGURE S1. Recombinant C. elegans C10F3.4 and human C15orf58 are highly specific for inorganic phosphate as the guanylyl acceptor. C10F3.4 (A) and C15orf58 (B) were incubated with 50 !M GDP-D-Glc at 26 °C and 31 °C, respectively, at a low enzyme concentration (striped bars) and at a

ten-fold higher enzyme concentration (black bars) in a reaction mixture at pH 7.5 containing 50 mM Tris- Downloaded from HCl, 10 mM NaCl, 1 mM dithiothreitol, and the indicated potential guanylyl acceptor (5 mM Pi or 5 mM PPi or 1 mM L-Gal-1-P or 1 mM D-Man-1-P). The reactions were stopped after 30 min by heating at 98 °C for 3 min and GDP-D-Glc was measured in the deproteinized samples by anion-exchange HPLC as described in the “Experimental Procedures” section, except when D-Man-1-P was tested as an acceptor.

Given the coelution of GDP-D-Glc and GDP-D-Man (one of the expected products when D-Man-1-P is www.jbc.org used as an acceptor) in the anion-exchange HPLC method, the samples resulting from the incubations in the presence of D-Man-1-P were analyzed by reverse-phase HPLC using a C18 column (250 " 4.6 mm, Grace Alltech, Deerfield, IL) with a Grace Alltech C18 security guard cartridge (7.5 x 4.6 mm). Here, mobile phase A consisted of 7.5 mM hexylamine as an ion pairing agent in water adjusted to pH 6.3 with at UCLA-Louise Darling Biomed. Lib., on May 20, 2012 acetic acid. Mobile phase B consisted of 90% methanol/10% 10 mM ammonium acetate and 5 mM hexylamine adjusted to pH 8.5 with ammonium hydroxide. A mobile-phase gradient was used with a flow rate of 0.4 ml/min starting at 100% A for 6 min, followed by a linear gradient from 100% to 40% A over 30 min, then holding at 40% A for 1 min, followed by a linear gradient from 40% to 100% A in 4 min, and finally holding for 10 min at 100% A. Control incubations without added enzyme (white bars) or without added guanylyl acceptor were also performed. The values represent the results of one measurement for each tested condition.

FIGURE S2. mRNA tissue distribution of GDP-D-Glc phosphorylase (C15orf58 homolog) and GDP- D-Man pyrophosphorylase ! (Gmppa) in mouse. Total RNA was isolated from adult male tissues after homogenization (two cycles of 30 s) with a Polytron homogenizer (Kinematica, Bohemia, NY) set at level 3 in TRI reagent (Sigma Chemical Co.), according to the manufacturer's instructions. RNA was treated with TURBO DNase I (Applied Biosystems /Ambion, Austin, TX) and cDNA was synthesized using random hexamers with the RETROscript kit (Ambion) according to the manufacturer's instructions. Quantitative real-time PCR reactions (20 !l) were carried out in an Opticon 2 Real-Time Cycler (MJ Research) using 4 !l of 20-fold diluted reverse transcription reaction, 0.3 !M gene-specific forward and reverse primers and the Takara SYBR Premix Ex Taq II reaction mix (Takara Bio, Madison,WI, USA) according to the manufacturer's recommendations. Primers used for amplification can be obtained upon request. In panel A, the individual Ct values obtained for the mouse genes of interest D330012F22Rik (homolog of the human C15orf58 gene) and Gmppa (GDP-mannose pyrophosphorylase !), as well as for the reference mouse genes Actb (beta-actin) and Gapdh (glyceraldehyde-3-phosphate dehydrogenase) in the indicated tissues for three technical replicates and three to four biological replicates are shown along with error bars which represent the means ± S.D. of all the data points obtained for each gene. In panel B, the data are presented as the fold changes in mRNA expression levels of the mouse C15orf58 homolog and the Gmppa genes, normalized to the endogenous reference genes Gapdh and Actb, relative to the normalized mRNA levels of those genes found in kidney. The results shown represent the means of three to four biological replicates. Normalized fold changes and statistical significance were calculated using the Qiagen (Valencia, CA) REST 2009 program (*p < 0.01).

! "!

FIGURE S3. Standard curves used for determining GDP-D-Glc and GDP-D-Man concentrations in worm and mouse tissue extracts by IP-LC-ESI-MS analysis. Standard curves were generated by injecting 0.1 to 4 pmol GDP-D-Glc (A) and 10 to 100 pmol GDP-D-Man (B) in a volume of 10 !l and by determining the corresponding product ion m/z 604>158 (GDP-D-Man) and m/z 604>362 (GDP-D-Glc) peak areas of the extracted ion current using the IP-LC-ESI-MS method described in the “Experimental Procedures” section.

Downloaded from

www.jbc.org

at UCLA-Louise Darling Biomed. Lib., on May 20, 2012

! "!

99 Downloaded from

Figure S1 www.jbc.org at UCLA-Louise Darling Biomed. Lib., on May 20, 2012

60 C10F3.4 0.03µg/ml A C15orf58 0.01µg/ml B C10F3.4 0.3 g/ml 60 µ C15orf58 0.1µg/ml No Enzyme No Enzyme ) ) M M 40 µ µ ( (

c c l l 40 G G - - D D - - P P D D 20 G G 20

0 0 r r o i i -P P o i P i -P t P P 1 1- t P 1 p P l- - p P l- ce a n ce a c G a c G a - M a - o L - o L N D N

100 Figure S2

35 5 A Kidney B

l Brain e * v

e Testis l

30 4 n Liver o i s

s Heart e r ) p y x e 3 * 25 e n

d A i

t k N

C15orf58 s R u s

homolog m

r e Gmppa n * i

20 v 2

( e g n

Actb a h * c

15 l 1 Gapdh o F -

10 0 C15orf58 homolog Gmppa BrainTestis Liver Heart BrainTestis BrainTestis Heart BrainTestis LiverHeart Kidney Kidney LiverHeart Kidney Liver Kidney

101 Figure S3

80000 800000 A B

60000 600000 ) ) s s p p c c ( (

y y t t i 40000 i 400000 s s n n e e t t n n I I 20000 200000

0 0 0 1 2 3 4 5 0 25 50 75 100 GDP-D-Glc (pmol) GDP-D-Man (pmol)

102 TABLE S1 Nucleotide specificity of recombinant C. elegans C10F3.4 Recombinant C10F3.4 was assayed in the reverse direction by incubating the enzyme (1.5 !g/ml) with various NDPs (1 mM or 5 mM) at 26°C in a reaction mixture at pH 7.5 containing 50 mM Tris-HCl, 10 mM NaCl, 40 mM D-Glc-1-P, and 1 mM dithiothreitol. The reactions were initiated by enzyme addition and stopped after 10 min by heating at 98 °C for 3 min. After removal of precipitated protein by centrifugation, NDP-D-Glc was measured in the supernatants by anion-exchange HPLC as described in the “Experimental Procedures” section. NDP-D-Glc concentrations were calculated by comparing the integrated peak areas with those of standard NDP-D-Glc solutions. Control incubations without added enzyme were conducted in the presence of 5 mM GDP or 5 mM ADP. Substantial non-enzymatic NDP- D-Glc formation was observed in the presence of ADP, but not in the presence of GDP. Values represent the result of one measurement for each of the indicated conditions.

NDP added C10F3.4 [NDP] [NDP-D-Glc] after incubation (mM) (!M)

GDP + 1 85

ADP + 1 21 UDP + 1 1.3

CDP + 1 0 GDP + 5 96

ADP + 5 85 UDP + 5 6.9 CDP + 5 0 GDP - 5 0 ADP - 5 85

! "! 103 TABLE S2 Comparison of the substrate specificity of C. elegans recombinant C10F3.4 and of worm extract GDP-D-Glc phosphorylase activity Various sugar nucleotides (50 !M) were incubated in the presence of C10F3.4 recombinant enzyme (0.23 !g/ml) or of desalted C. elegans wild-type (N2) extract (0.76 mg protein/ml) for 20–30 min at 26 °C. NDP production was measured by HPLC as described in the “Experimental Procedures” section. Phosphorylase activities were calculated based on the Pi-dependent NDP formations. With GDP-D-Glc as a substrate, the specific activity of the recombinant C. elegans enzyme was 8.4 !mol min-1 mg protein-1 (100%); the corresponding 100% specific activity of the worm extract was 0.28 nmol min-1 mg protein-1. The low GDP-D-Glc phosphorylase activity found here with the recombinant enzyme is due to the use of an excess of enzyme and subsequent substrate depletion. The phosphorylase activities found with the other sugar nucleotides are given as a percentage of the activity found with GDP-D-Glc. Values represent the results of one experiment for each substrate.

Relative specific activity (%) Substrate Recombinant C. elegans enzyme C. elegans extract GDP-D-Glc 100 100 GDP-L-Gal 1 0 GDP-D-Man 5 0 GDP-L-Fuc 1 0 UDP-D-Glc 0 0 UDP-D-Gal 0 0 ADP-D-Glc 0 0

7 104

CHAPTER 7

The Search for Phenotypes in Caenorhabditis elegans Mutants that Lack the

GDP-D-Glucose Phosphorylase Activity Encoded by the mcp-1 Gene

105 SUMMARY

This chapter includes details of the search for phenotypes in the mcp-1 mutant strain that might result from the accumulation of abnormal sugar nucleotide precursors and possible abnormal glycoconjugates. No defect was found in the ability of mcp-1 worms to produce viable eggs by quantitating the brood size of wild type and mutant worms, suggesting that the reproductive system was functional. Additionally, possible defects in the integrity of the cuticle in mcp-1 worms were investigated using Hoechst staining. Again, no difference was seen between wild type and mcp-1 mutants. Interestingly, the mcp-1 mutant strain did demonstrate a number of morphological abnormalities including the accumulation of eggs in the gonad, a gonad migration developmental abnormality, a "dumpy" appearance phenotype, and a "crushed" pharynx phenotype. Future work is designed to quantify these morphological changes and to confirm the role of the mcp-1 mutation in the manifestation of these abnormalities.

INTRODUCTION

The C. elegans mcp-1 mutant that lacks GDP-D-glucose phosphorylase activity has been characterized in CHAPTER 6 [1]. These mutants accumulate the unusual nucleotide sugar GDP-

D-glucose. Due to the lack of a physiological role of GDP-D-glucose in C. elegans (and in higher species) and its high similarity to GDP-D-mannose, we suggested that GDP-D-glucose, if not eliminated, could compete with GDP-D-mannose that serves as a mannosyl donor in glycosylation reactions in the cell, thus potentially inhibiting mannosyl incorporation or misincorporating glucose residues in place of mannose residues into the target glycoconjugates.

106 GDP-D-glucose is likely to be synthesized and accumulated in the cytosol as a product of GDP-

D-mannose pyrophosphorylase A [2][3]. GDP-D-glucose might be used as a substrate for mannosyltransferases that progressively add mannosyl units to the lipid-linked oligosaccharide

(LLO) in the protein N-glycosylation process. The consequences of incorporating glucose instead of mannose are difficult to predict. In humans, many of the mutations in the genes involved in the assembly of the LLO have been detected in patients with congenital disorders of glycosylation type I (CDG-I) [4]. Some of these can lead to the production of structurally incomplete LLOs, resulting in hypo-glycosylation of multiple glycoproteins [4]. My hypothesis is that GDP-D-glucose synthesis is an unwanted side product of the GDP-mannose- pyrophosphorylase A activity and that in the absence of GDP-D-glucose phosphorylase, the increased levels of GDP-D-glucose could potentially compete with GDP-D-mannose and lead to wrong or incomplete glycosylation of glycoproteins and glycolipids.

In contrast to the severe effects that CDG-I mutations have in humans, RNA interference

(RNAi) knockdown or deletional knockouts in C. elegans of the genes involved in glycosylation have generally relatively minor effects [5][6]. Many of the C. elegans genes with products involved in the glycosylation process have been identified based on screens that revealed a resistance to infection by various bacterial pathogens. Such resistance apparently results from the alteration in the structure of the worm’s surfaces that are exposed for bacterial adhesion such as the cuticle and the intestine [5]. Some examples include strains with mutations in the bre, bus and srf genes. The five bre genes are required for the synthesis of glycosphingolipids on the luminal surface of the intestine - mutations in these genes lead to resistance to the lethal effects of certain Cry toxins produced by Bacillus thuringiensis [7]. The bus mutants show resistance to

Microbacterium nematophilum infection. The majority of them encode for glycosyltransferases

107 that are predicted to change the ability of these pathogens to adhere to the worm’s cuticle [8][9].

Mutants in the srf-3 gene, which encodes for a UDP-sugar transporter, display resistance to infection by Yersinia pseudotuberculosis, Yersinia pestis, and Microbacterium nematophilum

[8][10][11]. In all of these cases, the altered surface glycoconjugates may prevent the binding of the bacteria [8] or their biofilm [10][11]. These groups of mutant strains also present an array of non-overlapping phenotypes such as molting defects (bus-8), fragile cuticle (bus-8, bus-17, srf-3), lower brood size (bre-1), and increased cuticle permeability (bus-8, bus-17) that are likely to reveal more specific roles of these proteins in the glycosylation process of the worms

[5][9][10][12].

My initial attempt to identify phenotypes in the mcp-1 mutant under control culture conditions did not yield any obvious differences compared to control wild type N2 animals. In view of some of the phenotypes that have been reported for other mutants with glycosylation defects, I attempted to find similar or related phenotypes in the mcp-1 mutant. This chapter includes a more detailed search for a quantifiable phenotype in the mcp-1 mutant strain that could be linked to a defect in the worm glycosylation process.

EXPERIMENTAL PROCEDURES

C. elegans Strains and Maintenance:

Standard procedures used to maintain C. elegans strains were adapted from Sulston and

Hodgkin [13]. The N2 strain was obtained from the Caenorhabditis Genetics Center (St. Paul,

MN). The bus-8 mutant strain was a donation from Dr. Allison Frand in the Department of

108 Biological Chemistry at UCLA. The C10F3.4 mutant strain (tm2679) lacking coding sequences between exons 5 and 8 of the C10F3.4 gene (WormBase) was provided by the C. elegans Core

Facility at Tokyo Women’s Medical University. Before further work was done to identify new phenotypes in the C10F3.4 (mcp-1) mutant strain (tm2679), this strain was backcrossed into N2 worms four times. PCR analysis was used to confirm the deletion mutation in the backcrossed strain, now named LA1. The forward and reverse primers that were used for the PCR assay are

5'-TGCGACACGGATCCTATCAG-3' and 5'-GTTCGACGGAGGTTTTGC-3' respectively.

Unless indicated otherwise, nematodes were grown on nematode growth media plates at 20 °C with the addition of OP50 bacteria.

Microscopy

Nematodes were anesthetized using 30 mM sodium azide in M9 buffer and immediately mounted on 2% agarose pads. Images were captured using a Zeiss Axioscope (Thornwood, NY) with an attached Hamamatsu Orca ER camera (Bridgewater, NJ) and Volocity software

(Improvision, Lexington, MA). All images were prepared with Adobe Photoshop and Adobe

Illustrator (San Jose, CA).

Hoechst Staining

Mixed stage worms from either N2 or mcp-1 strain were washed off plates with M9 buffer followed by incubation in M9 buffer containing 1 µg/ml Hoechst 33258 (Sigma Life

Sciences) at room temperature for 15 min with gentle agitation. The Hoechst dye was dissolved

109 in M9 medium at a concentration of 1 mg/ml to make a stock solution. Finally, the worm pellets were washed several times with M9 buffer. Worms were anesthetized and mounted on agarose pads as described above and the larvae were scored for staining of their nuclei by imaging on simultaneous fluorescence and differential interference contrast microscope.

Brood Size Analysis

Brood size was assayed as reported in Barrows et al. 2007 [12]. Briefly, single N2 or mcp-1 (tm2679) worms in L4 stage were plated. Worms on control plates and plates containing either 2.5 µg/ml or 5 µg/ml tunicamycin (Sigma Life Science) were incubated at 20 °C or 25 °C for 24-hour periods. After each period, the originally picked worms would be picked to a new plate. Progeny from the original parent worms were allowed to grow an additional 24 h at 25 °C or 20 °C before they were counted. This process was continued every 24 h until the original parents ceased to produce additional progeny. Live progeny were scored regardless of their developmental stage.

RESULTS AND DISSCUSSION

Morphological Changes in mcp-1 Mutants and their Relation to Glycosylation Defects

The deletion mutation in the mcp-1 strain had been shown to eliminate the GDP-D- glucose phosphorylase activity (CHAPTER 6) [1]. Nevertheless, these mutants do not portray an obvious on-plate phenotype. In collaboration with Dr. Alison Frand (Department of Biological

110 Chemistry, UCLA), I was able to identify several morphological changes in the mcp-1 mutant strain. These mcp-1 mutants display a gonad migration defect, an aberrantly shaped pharynx, an egg-laying defect (Egl) phenotype and a dumpy (Dpy) phenotype as shown in Figures 1 and 2 and discussed below.

Gonad Migration in mcp-1 and the Glycosylation Defects in other mig Mutants.

The shape of the hermaphrodite gonad is dictated by migration of two leader cells, the distal tip cells (DTCs) [14]. The DTCs migrate in a controlled and well-established time and space manner that results in two U-shaped gonad arms synchronized with the onset of adulthood

(Figure 1) [15]. Mutant worms with any variation in the orderly movement of a cell from one site to another fall under the definition of the Mig (cell migration variant) phenotype. Previous studies showed that two C. elegans proteins encoded by mig-17 and mig-23 control DTC migration and mutation in either one of them results in abnormal gonad positioning [16]. The authors reported that mig-17 encodes for ADAM, a secreted disintegrin and metalloprotease which requires correct N-glycosylation to ensure its proper secretion and binding to the somatic gonad. mig-23 encodes a nucloside diphosphatase that is necessary for the N-glycosylation of

ADAM. Additional genes that are involved in glycosylation processes in C. elegans and display mutant gonad migration abnormalities (Mig phenotype) are srf-3 and its paralog nstp-4, which encode UDP-N-acetylglucosamine Golgi transporters. srf-3 mutants treated with nstp-4 RNAi display oocyte accumulation and abnormal gonad morphology [17].

These well characterized Mig phenotypes that result from abnormal N-glycosylation provide suggest that a possible N-glycosylation defect in the mcp-1 mutant would also display

111 the Mig phenotype. In Figure 2A, we show that the mcp-1 mutant indeed shows a Mig phenotype where the gonad arms overlap each other. We find this phenotype in most of the scored animals

(about 95%). Another type of gonad migration abnormality I found includes posterior gonad arms that did not reach the dorsal muscle, similar to the phenotype seen for mig-17 and mig-23 mutants [16] (data not shown). However, further analysis is required to quantify the penetrance of the Mig phenotype in the mcp-1 mutants and to assess the distribution of the different types of gonad migration defects.

Crushed Pharynx in mcp-1 and the N-glycosylation Defects in ten-1 Mutants

The 80-cell C. elegans pharynx (foregut) is a neuromuscular tube that functions to pump and crush bacteria, and then to inject them into the worm’s intestine [18]. Many different phenotypes are associated with aberrant pharynx ranging from severe phenotypes as “no pharynx”

[19] to milder ones as pathogen-infected pharynx variant [20]. Based on the “crushed” looking pharynx we repeatedly saw in our mcp-1 mutants (Figure 2B), I propose to classify it under the

“pharynx disorganized phenotype” category. Initial screening for abnormal pharynx in the mcp-1 mutant strain revealed that about 10% of the worms display this morphological defect. A literature search for C. elegans mutants with disorganized pharynx phenotypes revealed that worms mutated in the ten-1 gene display a similar abnormal pharynx morphology as seen in our mcp-1 mutants [21]. TEN-1 is a transmembrane epidermal receptor for basement membranes collagens that is required for pattern formation and cell migration [21]. Interestingly, the ten-1 mutant also displays aberrant gonad migration [21][22] with gonads that do not form a tube-like structure but grow into a disorganized mass [21]. This is a more severe phenotype compared to

112 what I observed for the mcp-1 mutants in Figure 2B. Though there is no direct evidence showing that TEN-1 is glycosylated, its role as a receptor suggests that this may be the case. Additionally, the human homolog for TEN-1, Teneurin-3 has 17 predicted N-linked glycosylation sites

(UniProtKB/Swiss-Prot). In light of the similar morphological phenotypes and likely glycosylation requirement, it would be interesting to look into the glycosylation status of TEN-1 in the mcp-1 mutant and to test its possible role in the manifestation of the gonad migration and disorganized pharynx phenotypes.

Egg-Laying Defect and Dpy Phenotype in Relation to Glycosylation Defect

Worms with egg-laying defect phenotype (Egl) display variations in the developmental stage of eggs laid, egg-laying cycle, number of eggs or egg laying in response to stimuli compared to control [23]. The mcp-1 mutants we characterized retain eggs in the uterus for a longer period as evidenced from the progressed developmental stage of these eggs (Figure 2C).

Approximately 80% of the mcp-1 mutants display this phenotype. The mcp-1 mutant also displays a slight Dumpy (Dpy) phenotype (not shown) with a shorter and stouter morphology compared to control animals at the same developmental stage [24][25]. A literature search for mutants of C. elegans that display similar phenotypes in glycosylation-defective backgrounds revealed that both morphological changes (egg-laying defect and dumpy appearance) are present in egl-15 mutants that lack the first glycosylation site in the protein encoded by the egl-15 gene

[26]. EGL-15 is a fibroblast growth factor receptor, which has important roles in cell division, migration, and differentiation to cell survival and control of metabolic homeostasis [17].

Mutations in genes encoding fibroblast growth factor receptors that abolish conserved N-

113 glycosylation sites result in a diverse group of skeletal disorders in humans [26]. By looking for hypoglycosylation or aberrant glycosylation of EGL-15 in the background of mcp-1 mutant that display the Dpy or/and Egl phenotype I could possibly link the enzymatic defect in this mutant to deficient N-glycosylation and the observed morphological changes.

Future work will concentrate on quantification of these abnormalities to assess their penetration in the mcp-1 mutant worms. I also want to confirm the contribution of the GDP-D- glucose phosphorylase deficiency to the morphological changes we observed in the mcp-1 mutant. This could be accomplished by rescue experiments that will restore the GDP-D-glucose phosphorylase activity or by finding a mutant strain with a different allele that displays the same defects. Finally, I want to look for evidence that supports the role of GDP-D-glucose phosphorylase in the quality control of glycosylation in C. elegans. This goal could be achieved by showing an increase in the severity of the phenotypes upon addition of tunicamycin, which blocks the initial step of N-glycosylation [6]. Additionally, I can test for interaction with other glycosylation genes by creating double mutants of mcp-1 with other known glycosylation genes such as the bus mutants [5] or the srf-3 mutant [17] and look for new or enhanced phenotypes.

Brood Size in mcp-1 Mutant Worms

The transgenic UZ119 strain expresses a GFP-MCP-1 fusion protein under the mcp-1 promoter [1]. In this strain, GFP-MCP-1 expression was reported in the worm’s neuronal system as well as in the spermatheca, the site for sperm maturation and storage and in specific anterior hypodermis cells [1]. A different strain, with a mutation in the bre-1 gene was isolated in a screen for Bacillus thuringiensis toxin-resistance (bre). Barrows et al. reported that the bre-1

114 mutant lacks the ability to synthesize GDP-L-fucose, which consequently affects the strain’s ability to produce glycans. Interestingly, the bre-1 mutant also has a dramatic reduction in brood size at 25 °C [12]. Since the mcp-1 strain is predicted to display a glycosylation defect and given the expression of MCP-1 in the spermatheca as well as the newly discovered Egl phenotype (see above), we were encouraged to look into possible defects in brood size.

My results for brood size analysis revealed that these worms did not display any significant difference compared to the wild type strain at 20 ºC or at 25 ºC (Table 1).

Tunicamycin is an inhibitor of the enzyme GlcNAc phosphotransferase that catalyzes the first step of glycoprotein synthesis in eukaryotes; this agent thus inhibits N-linked glycoprotein synthesis. Tunicamycin has been used in an RNA interference screen to reveal and enhance phenotypes caused by genes involved in N-linked glycoprotein synthesis that were otherwise unnoticed or ignored [6]. We tested the effect of two different concentrations of tunicamycin on the C. elegans brood size (2.5 µg/ml at 20 ºC and 5 µg/ml at 25 ºC). The lower 2.5 µg/ml tunicamycin concentration did not yield a change in brood size of either of the strains. The higher 5 µg/ml concentration of tunicamycin did result in a lower brood size (about 13% lower compared to the control at 25 ºC) for both strains, demonstrating that this result was independent of the mcp-1 mutation.

It is worth noting that tunicamycin addition to the plates caused a developmental delay.

In this brood size assay parental worms were given 24 hours to lay eggs before being transferred to a new plate while the progeny were given additional 24 hours before live progeny number was recorded. At this time point (about 48 hours from the time egg laying had started), I expected to score animals that were L3 or L4 larvae depending on the time the eggs were laid and on the incubation temperature (20 ºC or 25 ºC). While that was the case with worms grown on control

115 plates, worms that were incubated on either 2.5 µg/ml or 5 µg/ml of tunicamycin were mostly at

L2 stage (with 2.5 µg/ml of tunicamycin at 20 ºC) or mostly at L1 stage (with 5 µg/ml of tunicamycin at 25 ºC). In the set up of this assay only the number of the hatched worms had been recorded regardless of their developmental stage therefore this observation was not quantified.

Future experiments will aim to test whether this developmental delay caused by tunicamycin is more severe in the mcp-1 mutant strain.

Interestingly both BUS-2, a predicted galactosyltransferase, and SRF-3, a UDP-sugar transporter, are also expressed in the hermaphrodite spermatheca [28][29] though mutation in these genes did not affect self-fertility. These results suggest that spermathecal function is normal. It would be interesting to test brood size in double mutants of mcp-1 and bus-2 or srf-3.

Another experiment in mated C. elegans can test the effect of male mcp-1 mutants on brood size.

Hoechst Staining in mcp-1 Mutant Worms as a Measure of Cuticle Integrity

bus-8 is predicted to encode for a mannosyltransferase, an enzyme that catalyzes the transfer of the sugar moiety mannose from the cytosolic-residing substrate GDP-D-mannose to the a dolichol phophate carrier in the lumen of the endoplasmic reticulum where it will be added to the growing lipid link oligosaccaride (LLO). BUS-8 was found to be required post- embryonically for epidermal organization, molting and production of a normal cuticle surface [9].

The role of BUS-8 in the cuticle integrity was demonstrated via a Hoechst staining experiment that revealed a higher permeability of the DNA-binding dye in the bus-8 mutant strain [9].

My previous work showed that mcp-1 mutant worms accumulate GDP-D-glucose. I hypothesized that accumulation of this metabolite could compete with GDP-D-mannose as a

116 substrate for the mannosyltransferase, thus inhibiting its activity or affecting the integrity of the newly assembled LLO [1]. The morphological changes that I detected in the mcp-1 mutant strain, such as the phenotypes of defective egg laying, dumpy appearance and crushed pharynx could also indicate that there is an effect on the extracellular matrix integrity. Together, these observations led me to test the permeability of the mcp-1 mutant to the Hoechst dye.

The bus-8 mutant strain was used as a positive control. In this strain, Hoechst dye is expected to stain 40-50% of the larvae [30]. The wild type N2 strain was used as a negative control, where

Hoechst dye is expected to stain about 10% of the larvae (probably due to unsynchronized molting) [30][31]. Our results demonstrated that there is no difference in the Hoechst staining between the wild type and the mcp-1 mutant suggesting that there is no change in cuticle permeability for these mutants. Nevertheless, it is possible that the cuticle integrity in mcp-1 mutants is affected in a less severe manner, not detectable by the Hoechst staining method.

Increasing the glycosylation defect stress in the mcp-1 mutant worms with tunicamycin could test this theory. Alternatively, as bus-8 and bus-17 mutants are known to exhibit cuticle permeability [5], evaluating cuticle integrity in bus-8 /mcp-1 or bus-17/mcp-1 double mutants compared to bus-8 and bus-17 mutants could be helpful in detecting a slight effect due to the lack in MCP-1 activity.

ACKNOWLEDGEMENTS

I am grateful to our collaborator Dr. Alison Frand from the UCLA Department of

Biological Chemistry for her collaboration in identifying morphological changes in the mcp-1 mutants and for sharing expertise and equipment for microscopy. Additionally, I would like to

117 thank Dr. Vijaykumar Meli, and Beatriz Osuna from Dr. Alison Frand’s laboratory for their help in performing Hoechst staining experiments.

118

Table 1 - Brood size analysis for N2 and mcp-1 mutant strains.

The brood size of wild type (N2) or the mcp-1 mutant strain (tm2679) at 20 ºC or 25 ºC was measured in the presence or absence of tunicamycin (TN) (Sigma Life Sciences, dissolved in dimethyl sulfoxide at a stock concentration of 5-10 mg/ml). The brood size assay was performed and analyzed as described in the "Experimental Procedures" section. n = number of parent worms analyzed. S.D. is the standard deviation. Statistical analysis using a two-tailed unpaired

Student’s t test with unequal variance shows that there is no statistically significant difference in total progeny between the N2 and mcp-1 C. elegans strains at 20 ºC (p value = 0.79) or at 25 ºC

(p value = 0.79) under control conditions. Additionally, the same type of statistical analysis showed that there is no statistically significant difference in total progeny between the N2 and mcp-1 C. elegans strains at 20 ºC that were treated with 2.5 µg/ml TN (p value = 0.54) or at 25

ºC that were treated with 5 µg/ml TN (p value = 0.76).

Total viable TN progeny/parent at 20 S.D for n Strain treatment ºC replicates

N2 (n = 9) 0 284 60 mcp-1 (n = 8) 0 285 51 - N2 (n = 2) 2.5 µg/mL 280 4 mcp-1 (n = 4) 2.5 µg/mL 283 8

Total viable TN progeny/parent at 25 S.D for n treatment ºC replicates

N2 (n = 10) 0 235 43 mcp-1 (n = 8) 0 229 55

N2 (n = 4) 5 µg/mL 197 43 mcp-1 (n = 4) 5 µg/mL 206 40

119 Table 2 - The fraction of larvae with nuclei stained by Hoechst 33258 in wild type (N2), bus-8, and mcp-1 mutants of C. elegans.

Mixed staged larvae of wild type (N2), the bus-8 mutant (e2887), or the mcp-1 mutant (tm2679) were stained with Hoechst 33258 (Sigma Life Sciences) as described in the "Experimental

Procedures" section. The results presented here represent one set of results obtained with staining of 62-69 worms per strain.

Strain Stained Unstained % Stained (+) (-)

N2 4 63 6%

bus-8 (e2887) 25 37 40%

mcp-1 (tm2679) 5 64 7%

120

Figure 1- Diagram of gonad morphology in wild type hermaphrodites. DTC-distal tip cells

Pharynx Gonad arms

Vulva DTCs

121 Figure 2- Morphological changes in the C. elegans mcp-1 mutant strain. Nomarski images of wild type (N2) gonads and of overlapping gonad arms in mcp-1 mutants demonstrating the Mig phenotype (panel A); N2 pharynx and “Crushed” disordered pharynx of the mcp-1 mutant (panel

B); eggs in the gonad of an N2 animal and mcp-1 mutant with egg-laying defect that causes retention of eggs in the gonad (panel C). Scale bar = 20 µm.

122 REFERENCES

1. Adler LN, Gomez TA, Clarke SG, Linster CL. A novel GDP-D-glucose phosphorylase

involved in quality control of the nucleoside diphosphate sugar pool in Caenorhabditis

elegans and mammals. J Biol Chem. 2011 Jun 17;286(24):21511-23. Epub 2011 Apr 20.

2. Ning, B., and Elbein, A. D. Cloning, expression and characterization of the pig liver

GDP-mannose pyrophosphorylase. Evidence that GDP-mannose and GDP-Glc

pyrophosphorylases are different proteins. Eur. J. Biochem. 2000 Dec;267(23):6866-74.

3. Szumilo T, Drake RR, York JL, Elbein AD. GDP-mannose pyrophosphorylase.

Purification to homogeneity, properties, and utilization to prepare photoaffinity analogs. J

Biol Chem. 1993 Aug 25;268(24):17943-50.

4. Jaeken, J. Congenital disorders of glycosylation. Ann N Y Acad Sci. 2010 Dec;1214:190-

8. doi: 10.1111/j.1749-6632.2010.05840.x.

5. Gravato-Nobre MJ, Stroud D, O'Rourke D, Darby C, Hodgkin J. Glycosylation genes

expressed in seam cells determine complex surface properties and bacterial adhesion to

the cuticle of Caenorhabditis elegans. Genetics. 2011 Jan;187(1):141-55. Epub 2010 Oct

26.

6. Struwe, W. B., and Warren, C. E. High-throughput RNAi screening for N-glycosylation

dependent loci in Caenorhabditis elegans. Methods Enzymol. 2010;480:477-93. Review.

7. Griffitts JS, Huffman DL, Whitacre JL, Barrows BD, Marroquin LD, Müller R, Brown

JR, Hennet T, Esko JD, Aroian RV. Resistance to a bacterial toxin is mediated by

removal of a conserved glycosylation pathway required for toxin-host interactions. J. Biol.

Chem. 278: 45594–45602.

123 8. Hodgkin, J., P. E. Kuwabara and B. Corneliussen. A novel bacterial pathogen,

Microbacterium nematophilum, induces morphological change in the nematode C.

elegans. Curr Biol. 2000 Dec 14-28;10(24):1615-8.

9. Partridge FA, Tearle AW, Gravato-Nobre MJ, Schafer WR, Hodgkin J. The

C. elegans glycosyltransferase BUS-8 has two distinct and essential roles in epidermal

morphogenesis. Dev Biol. 2008 May 15;317(2):549-59. Epub 2008 Mar 14

10. Darby C, Chakraborti A, Politz SM, Daniels CC, Tan L, Drace K. Caenorhabditis elegans

mutants resistant to attach- ment of Yersinia biofilms. Genetics. 2007 May;176(1):221-30.

Epub 2007 Mar 4.

11. Cipollo JF, Awad AM, Costello CE, Hirschberg CB. srf-3, a mutant of Caenorhabditis

elegans, resistant to bacterial infection and to biofilm binding, is deficient in

glycoconjugates. J Biol Chem. 2004 Dec 17;279(51):52893-903. Epub 2004 Sep 27.

12. Barrows BD, Haslam SM, Bischof LJ, Morris HR, Dell A, Aroian RV. Resistance to

Bacillus thuringiensis toxin in Caenorhabditis elegans from loss of fucose. J Biol

Chem. 2007 Feb 2;282(5):3302-11. Epub 2006 Nov 29.

13. Sulston, J.E. and J. Hodgkin, Methods, in The Nematode Caenorhabditis elegans, W.S.

Wood, Editor 1988: Cold Spring Harbor: Cold Spring Harbor Laboratory. p. 587-606.

14. Hubbard EJ, Greenstein D. The Caenorhabditis elegans gonad: a test tube for cell and

developmental biology. Dev Dyn. 2000 May;218(1):2-22.

15. Meighan CM, Schwarzbauer JE. Control of C. elegans hermaphrodite gonad size and

shape by vab-3/Pax6-mediated regulation of integrin receptors. Genes Dev. 2007 Jul

1;21(13):1615-20.

124 16. Nishiwaki K, Kubota Y, Chigira Y, Roy SK, Suzuki M, Schvarzstein M, Jigami Y,

Hisamoto N, Matsumoto K. An NDPase links ADAM protease glycosylation with organ

morphogenesis in C. elegans. Nat Cell Biol. 2004 Jan;6(1):31-7. Epub 2003 Dec 21.

17. Caffaro CE, Hirschberg CB, Berninsone PM. Functional redundancy between two

Caenorhabditis elegans nucleotide sugar transporters with a novel transport mechanism. J

Biol Chem. 2007 Sep 21;282(38):27970-5. Epub 2007 Jul 25.

18. Pilon M, Mörck C. Development of Caenorhabditis elegans pharynx, with emphasis on

its nervous system. Acta Pharmacol Sin. 2005 Apr;26(4):396-404.

19. Bowerman B, Eaton BA, Priess JR. skn-1, a maternally expressed gene required to

specify the fate of ventral blastomeres in the early C. elegans embryo. Cell. 1992 Mar

20;68(6):1061-75.

20. Haskins KA, Russell JF, Gaddis N, Dressman HK, Aballay A. Unfolded protein response

genes regulated by CED-1 are required for Caenorhabditis elegans innate immunity. Dev

Cell. 2008 Jul;15(1):87-97.

21. Trzebiatowska A, Topf U, Sauder U, Drabikowski K, Chiquet-Ehrismann R.

Caenorhabditis elegans teneurin, ten-1, is required for gonadal and pharyngeal basement

membrane integrity and acts redundantly with integrin ina-1 and dystroglycan dgn-1. Mol

Biol Cell. 2008 Sep;19(9):3898-908. Epub 2008 Jul 16.

22. Drabikowski, K., Trzebiatowska, A., and Chiquet-Ehrismann, R. ten-1, an essential gene

for germ cell development, epidermal morphogenesis, gonad migration, and neuronal

pathfinding in Caenorhabditis elegans. Dev Biol. 2005 Jun 1;282(1):27-38.

23. Trent C, Tsuing N, Horvitz HR. Egg-laying defective mutants of the nematode

Caenorhabditis elegans. Genetics. 1983 Aug;104(4):619-47.

125 24. Page AP, Johnstone IL. The cuticle. Worm Book. 2007 Mar 19:1-15.

25. Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974 May;77(1):71-94.

26. Polanska UM, Duchesne L, Harries JC, Fernig DG, Kinnunen TK. N

Glycosylation regulates fibroblast growth factor receptor/EGL-15

activity in Caenorhabditis elegans in vivo. J Biol Chem. 2009 Nov 27;284(48):33030-9.

Epub 2009 Oct 2.

27. Eswarakumar VP, Lax I, Schlessinger J. Cellular signaling by fibroblast growth factor

receptors. Cytokine Growth Factor Rev. 2005 Apr;16(2):139-49. Epub 2005 Feb 1

28. Höflich J, Berninsone P, Göbel C, Gravato-Nobre MJ, Libby BJ, Darby C, Politz

SM, Hodgkin J, Hirschberg CB,Baumeister R. Loss of srf-3-encoded nucleotide sugar

transporter activity in Caenorhabditis elegans alters surface antigenicity and prevents

bacterial adherence. J Biol Chem. 2004 Jul 16;279(29):30440-8. Epub 2004 Apr 30.

29. Palaima E, Leymarie N, Stroud D, Mizanur RM, Hodgkin J, Gravato-Nobre

MJ, Costello CE, Cipollo JF. The Caenorhabditis elegans bus-2 mutant reveals a new

class of O-glycans affecting bacterial resistance. J Biol Chem. 2010 Jun

4;285(23):17662-72. Epub 2010 Apr 12.

30. Meli VS, Osuna B, Ruvkun G, Frand AR. MLT-10 defines a family of DUF644 and

proline-rich repeat proteins involved in the molting cycle of Caenorhabditis elegans. Mol

Biol Cell. 2010 May 15;21(10):1648-61. Epub 2010 Mar 24.

31. Moribe H, Yochem J, Yamada H, Tabuse Y, Fujimoto T, Mekada E. Tetraspanin protein

(TSP-15) is required for epidermal integrity in Caenorhabditis elegans. J Cell Sci. 2004

Oct 15;117(Pt 22):5209-20. Epub 2004 Sep 28.

126

CHAPTER 8

Future Goals and Perspectives

127 Ascorbate Levels in C. elegans

The L-galactose pathway is the main route for vitamin C (L-ascorbic acid, AsA) biosynthesis in higher plants [1][2][3] as well as in green algae [4]. In this pathway, GDP-L- galactose is converted to L-galactose-1-phosphate [2]. The enzyme that catalyzes this reaction in

Arabidopsis thaliana is VTC2. VTC2 has homologs in vertebrates and invertebrates including the nematode C. elegans [2]. BLAST analysis showed that the C. elegans homolog, C10F3.4

(now designated MCP-1), shares 24% sequence identity with VTC2. However, a literature search did not yield any information about the content of AsA or the possible biosynthetic pathway of

AsA in C. elegans.

In CHAPTER 5, I reported that C. elegans worm extracts do contain low yet detectable levels of vitamin C. This finding was in agreement with the apparent need for AsA in C. elegans as an antioxidant [5][6] and as a cofactor for enzymatic reactions [7][8][9]. I also investigated the role of C10F3.4, the worm homolog of plant VTC2. HPLC as well as mass spectrometry analysis revealed that the AsA levels in the C10F3.4 deletion strain tm2679 are not significantly different from those of the wild type N2 strain. Additionally, the C. elegans genome lacks homologs for other key enzymes in the plant or the animal vitamin C synthesis pathways.

Altogether, these finding led us to conclude that while C. elegans worms contain vitamin C, they do not synthesize this antioxidant via either the plant or the animal pathway. To test the possibility that C. elegans can uptake AsA from an exogenous source, the worm’s food source, E. coli cells, as well as other media and buffers that were used to grow and prepare C. elegans extracts were tested for the presence of AsA. However, at least with our detection limit, HPLC analysis showed that none of these contain AsA. On the other hand, two studies showed that supplementing the worm’s culture media with an exogenous source of AsA was able to rescue

128 oxidative stress that was inflicted by oxidizing agents [5] or specific mutation [6], thus suggesting that C. elegans is capable of taking up and utilizing AsA. Additional support for this idea comes from the identification of six homologs for the human AsA transporter NAT in the C. elegans genome [10].

Further work can be aimed at determining whether C. elegans worms have the ability to concentrate minute amounts of AsA or dehydroascorbate (DHA) from their environment. To answer this question, more sensitive analytical assays are required that will allow detecting low levels of AsA or DHA. Additionally, C. elegans strains with defects in the expression of the genes encoding for the putative AsA transporters can be tested for their AsA content in comparison to control animals.

The Role of GDP-D-glucose Phosphorylase as a Quality Control Enzyme in N-glycosylation

We recently reported of the biochemical activity of the plant VTC2 homologs in C. elegans worms (mcp-1) and humans (C15orf58) as GDP-D-glucose phosphorylases [17]. These enzymes catalyze the phosphorolysis of GDP-D-glucose to glucose-1-P. Support for the physiological relevance of these findings came from mass spectrometry analyses that revealed accumulation of GDP-D-glucose in a C. elegans strain with an mcp-1 deletion mutation. The preferred substrate specificity towards GDP-D-glucose as opposed to that of the plant VTC2 enzyme which does not discriminate between GDP-L-galactose and GDP-D-glucose was surprising especially as there is no known function for GDP-D-glucose in mammals and invertebrates including C. elegans. GDP-D-glucose synthesis in liver, kidney and mammary glands of mammals has been reported previously [11][12][13][14] and this activity was

129 attributed to a side reaction of GDP-D-mannose pyrophosphorylase, the enzyme that produces

GDP-D-mannose from mannose-1-phosphate and GTP [13]. The most recent report by Ning and

Elbein reveled that GDP-D-mannose pyrophosphorylase in pig liver contains two subunits

(designated α and β), which are the products of two distinct genes. Recombinantly expressed

GDP-D-mannose pyrophosphorylase β was found to preferentially produce GDP-D-mannose while the native pig liver enzyme with both subunits was found to be twice as active in forming

GDP-D-glucose using glucose-1-phosphate and GTP as substrates. Thus, GDP-D-mannose pyrophosphorylase α was proposed to be promoting GDP-D-glucose production either independently or as a modifier of the GDP-D-mannose pyrophosphorylase β substrate specificity

[13]. As of today, there are no published data concerning the enzymatic activity of GDP-D- mannose pyrophosphorylase α that would confirm this hypothesis.

Future work is needed to characterize the enzymatic activity and substrate specificity of recombinant GDP-D-mannose pyrophosphorylase α from mammals and C. elegans to determine its possible participation in GDP-D-glucose synthesis.

GDP-D-mannose is the mannosyl donor for the N-glycosylation process in the cell.

During the initial assembly step of protein N-glycosylation, 14 sugar units are progressively added to assemble the lipid-linked oligosaccharide (LLO); the oligosaccharide is then transferred to an asparagine residue of newly translated proteins. Five mannosyl units are transferred from

GDP-D-mannose to the growing oligosaccharide chain on the LLO by cytosolic mannosyltransferases. The LLO is then flipped from a cytosolic to an ER lumen orientation, where mannosyl units continue to be added to the oligosaccharide chain by transfer from dolichol-phosphate-mannose. Congenital Disorders of Glycosylation (CDG) is a group of inherited human disorders caused by impaired glycosylation of proteins and lipids. In a subgroup

130 of patients named CDG-I the mutations that lead to the disorder cause defects in LLO assembly or its transfer to proteins, which results in hypoglycosylation [15]. GDP-D-mannose and GDP-D- glucose vary only in the configuration of the hydroxyl group on C2. This structural similarity makes it plausible that the cytosolic mannosyltransferases that recognize their substrates and are dependent to a greater or lesser extent on recognizing the nucleotide moiety [16], will not discriminate between GDP-D-mannose and GDP-D-glucose, thus inserting glucose instead of mannose into the LLO if GDP-D-glucose concentrations increase. Though it is hard to predict the consequences for replacing mannose for glucose in the assembly stage of N-glycosylaytion, hypoglycosylation or aberrant glycosylation may occur. We suggested that GDP-D-glucose phosphorylase works as a metabolite repair enzyme to eliminate GDP-D-glucose from the cytosolic sugar nucleotide pool, thus preventing competition with GDP-D-mannose for the binding to mannosyltransferases in the N-glycosylation pathway [17]. In this hypothesis, the

‘erroneous’ GDP-D-glucose synthesis via GDP-D-mannose pyrophosphorylase is potentially toxic to the cells and GDP-D-glucose phosphorylase helps to prevent additional errors by the mannosyltransferase that could lead to an N-glycosylation defect. Further work here can focus on finding evidence to support the role of GDP-D-glucose phosphorylase as a metabolite repair enzyme in the N-glycosylation process.

Interestingly, though mutations in the glycosylation pathway in humans and other mammals can lead to severe diseases [18][19], in nematodes, mutation or RNAi feeding experiments that reduce the activity of known genes in the protein glycosylaytion pathway typically result in a mild phenotype or in the lack of apparent phenotype [20][21]. As reported in

CHAPTER 7, work done in collaboration with Dr. Alison Frand (Department of Biological

Chemistry at UCLA) has allowed us to identify an array of morphological changes displayed in

131 the nematode mcp-1 mutant including a gonadal migration defect, an egg-laying defect, dumpy appearance, and a crushed pharynx phenotype. The next steps can include quantifying these changes, assessing their penetration in the mcp-1 mutant strain and verifying the contribution of the GDP-D-glucose phosphorylase to the reported morphological changes. Rescue experiments that restore GDP-D-glucose phosphorylase activity to the mcp-1 mutants or analyzing the phenotypes produced by a mutation with a different allele could be employed to confirm this connection. Additionally, although I was able to point out a few known glycosylation mutants that display one or two of the phenotypes attributed to the mcp-1 strain (see CHAPTER 7), the role of GDP-D-glucose phosphorylase in the glycosylation process in worms is yet to be determined. New or enhanced phenotypes as a result of tunicamycin addition to block the initial step of N-glycosylation [21] or by creating a double mutant with one or more known glycosylation defective strains, will support the involvement of GDP-D-glucose phosphorylase in the N-glycosylation process.

In summary, future work can be aimed at confirming the recently discovered morphological changes in the mcp-1 mutants and at establishing a connection between the lack of GDP-D-glucose phosphorylase activity and deficient glycosylation.

The C. elegans’ glycocalyx is an extracellular coat consisting of glycans and glycoconjugates, which can be found on the cuticle as well as on the luminal surface of the nematode intestine [22][20]. Mutation in glycosylation genes that were linked to change in the properties of the glycocalyx often resulted in a resistance to infection by various pathogens [20].

Mutations in bre genes (1-5) lead to resistance against the Cry5B toxin secreted by Bacillus thuringiensis [23][24]. Four of the bre genes are glycosyltransferases involved in synthesis of glycosphingolipids at the intestinal lumen that are utilized by Bacillus thuringiensis as a receptor

132 to infect the cells [23]. The bus mutants were isolated in a screen that tested resistance to

Microbacterium nematophilum, an organism that can adhere to the rectal cuticle [25]. Worms infected by M. nematophilum showed a swollen tail, while the resistance mutants display the Bus

(Bacterially Un-Swollen) phenotype. The majority of the already identified bus genes (bus-8, bus-17, bus-4 and bus-2) are predicted to encode for glycosyltransferases whereas bus-12 is predicted to function as an UDP-sugar transporter [20]. Interestingly, some of the bus mutants present additional resistance to infection by Yersinia pseudotuberculosis [26]. Infection by this pathogen results in creation of biofilm on the head area of the worms. Worms with resistance to the formation of the biofilm by this pathogen present a Bah (Biofilm Absent on Head) phenotype

[26][27]. A different mutant strain that displays resistance to both Microbacterium nematophilum and Yersinia pseudotuberculosis is srf-3. srf-3 is a UDP-N-acetylglucosamine and UDP- galactose transporter and the mutated strain was reported to have aberrant N- and O- glycosylation [28]. As mcp-1 mutants are predicted to synthesize wrongly glycosylated or hypoglycosylated N-glycans, it would be interesting to test whether this mutation could affect glycocalyx properties and specifically change the resistance of the mutant strain to pathogenic bacterial infection.

Future work can be aimed to test the possibility that the mcp-1 mutation results in a resistance to one or more of the pathogenic bacteria as reviewed above.

Analysis of mcp-1 Mutant Glycans

As discussed above, one implication for the increase in the GDP-D-glucose pool in the mcp-1 mutant animals is that GDP-D-glucose is competing with GDP-D-mannose for the

133 binding to mannosyltransferases, thus potentially leading to hypoglycosylation or aberrant glycosylation. Alternatively, as GDP-D-mannose is the substrate for GDP-mannose-4,6- dehydratase in the only pathway for GDP-L-fucose synthesis in C. elegans [24], GDP-D-glucose could potentially compete with GDP-D-mannose for the recognition by GDP-mannose-4,6- dehydratase and as a result inhibit GDP-L-fucose production. As the target molecules for many of the glycosylation genes are unknown, a glycome analysis that compares the different glycan structures released from proteins and lipids of wild type and mutant strains can be useful in elucidating the role of these genes in the glycosylation process.

In APPENDIX I, I described the search for changes in the glycome of C. elegans mcp-1 mutants in comparison to control animals. Briefly, following preparation of mcp-1 and control

N2 worm strains pellets in our lab, they were sent to our collaborators at the Haslam laboratory at the Imperial College London for processing and mass spectrometry analysis. The N-glycans and the O-glycans from the different strains were released by enzymatic and chemical activity respectively and subsequently derivatized by permethylation. Different fractions of glycans were purified and separated in a reverse phase chromatography using an acetonitrile gradient. The resulting fractions were loaded into a MALDI system for glycan structure analysis.

Only minor changes were found in the N-glycan fractions analyzed from the first set of mcp-1 and N2 samples. On the other hand, no changes were detected in the O-glycan profiles of the two strains. These subtle differences between the glycan profiles of the two strains could be artifactual as a result of a natural glycome variations within the worm population, processing procedures, or due to instrument handling. On the other hand, these alterations could be real and due to the GDP-D-glucose phosphorylase mutation.

134 Future work can be aimed at repeating these experiments with additional sets of mcp-1 mutants and wild type strains to verify whether these differences in glycome profile are reproducible. Additionally, it is possible that the mcp-1 mutation alone is not sufficient to elicit an effect severe enough that can be seen in a glycome analysis. Increasing the glycosylation stress by incubating worms in tunicamycin or by overexpressing the predicted worm homolog of

GDP-D-mannose pyrophosphorylase A (Y47D9A.1) that is predicted to increase the GDP-D- glucose pool in the cell could be helpful in testing this hypothesis.

Looking for Mutations in the Human Homolog of C. elegans mcp-1 in CDG-1 Patients

The diagnosis of N-glycosylation disorders in human patients is challenging as the clinical picture is multisystemic and can vary between patients in these rare disorders [19]. The basic characterization of patients as having a CDG is based on an isoelectric focusing (IEF) test of the highly glycosylated plasma protein transferrin (Tf) [29]. This test can also help classify the disorder as CDG-Type I when the mutations affect the LLO assembly (in cytosol and endoplasmic reticulum) or as Type II for mutations that affect the processing stage of the oligosaccharide after transfer to the protein following the LLO assembly (in endoplasmic reticulum and Golgi) [19]. Nevertheless, a normal test result does not exclude the possibility that the patient has a CDG. When the IEF shows abnormal results but the basic mutation is unknown, the disease is labeled CDG-x. Specifically, the disorder is labeled as CDG-Ix when the transferrin IEF shows a type 1 profile, and as CDG-IIx when the profile is type 2.

The mcp-1 mutants that accumulate GDP-D-glucose [17] and were predicted to display glycosylation defects are viable and did not display severe phenotypes under lab conditions (See

135 CHAPTER 7). This doesn’t exclude that mutations in the human homolog of this gene could cause deleterious effects in humans [21].

A recently initiated project to sequence the human C15orf58 gene exonic sequences in unidentified CDG-Ix patients was launched in collaboration with Dr. Carole Linster from the

University of Luxembourg, Dr. Elsa Wiame from the de Duve Institute and Université

Catholique de Louvain (Brussels, Belgium), and Valérie Race and Gert Matthijs from the Centre for Human Genetics, Katholieke Universiteit Leuven (Leuven, Belgium). I designed the primers for the DNA samples provided by Valérie Race and Gert Matthijs and Dr. Elsa Wiame preformed the sequencing analyses. The goal of this project was to identify mutations in

C15orf58, the human homolog of mcp-1, which could cause glycosylation defects leading to the development of CDG-I. Since the C15orf58 gene does not include an intron, the DNA coding sequence was divided into four fragments and specific primers were designed to sequence each fragment in both directions. DNA samples from 47 patients diagnosed as type I CDG patients were sequenced. Most of the samples (45) were successfully sequenced and yielded informative data while two samples failed to yield any product due to a low quality of the DNA. Analysis of the C15orf58 sequences from the CDG patients did not reveal any mutation but a series of frequent SNPs that were all registered in the NCBI SNP database.

Future work can be directed towards sequencing the C15orf58 gene in additional CDG patients. It may also be expected that mutations in this gene might be uncovered by exon sequencing efforts that are now performed in an increasing number of clinical research centers.

136 REFERENCES

1. Linster CL, Clarke SG. L-Ascorbate biosynthesis in higher plants: the role of VTC2.

Trends Plant Sci. 2008 Nov;13(11):567-73. Epub 2008 Sep 27

2. Linster CL, Gomez TA, Christensen KC, Adler LN, Young BD, Brenner C, Clarke SG.

Arabidopsis VTC2 encodes a GDP-L-galactose phosphorylase, the last unknown enzyme

in the Smirnoff-Wheeler pathway to ascorbic acid in plants. J Biol Chem. 2007 Jun

29;282(26):18879-85. Epub 2007 Apr 26

3. Wheeler GL, Jones MA, Smirnoff N. The biosynthetic pathway of vitamin C in higher

plants. Nature. 1998 May 28;393(6683):365-9.

4. Urzica EI, Adler LN, Page MD, Linster CL, Arbing MA, Casero D, Pellegrini M,

Merchant SS, Clarke SG. Impact of Oxidative Stress on Ascorbate Biosynthesis in

Chlamydomonas via Regulation of the VTC2 Gene Encoding a GDP-L-galactose

Phosphorylase. J Biol Chem. 2012 Apr 20;287(17):14234-45. Epub 2012 Mar 5

5. Khare S, Gomez TA, Linster CL, Clarke SG. Defective responses to oxidative stress in

protein l-isoaspartyl repair-deficient Caenorhabditis elegans. Mech Ageing Dev 2009

Oct;130(10):670-80. Epub 2009 Aug 12

6. Huang J, Lemire BD. Mutations in the C. elegans succinate dehydrogenase iron-sulfur

subunit promote superoxide generation and premature aging. J Mol Biol. 2009 Apr

3;387(3):559-69. Epub 2009 Feb 20

7. Alkema MJ, Hunter-Ensor M, Ringstad N, Horvitz HR.Tyramine Functions

independently of octopamine in the Caenorhabditis elegans nervous system. Neuron.

2005 Apr 21;46(2):247-60.

137 8. Keskiaho K, Kukkola L, Page AP, Winter AD, Vuoristo J, Sormunen R, Nissi

R, Riihimaa P, Myllyharju J.Source . Characterization of

a novel Caenorhabditis elegans prolyl 4-hydroxylase with

a unique substrate specificity and restricted expression in the pharynx and excretory duct.

J Biol Chem. 2008 Apr 18;283(16):10679-89. Epub 2008 Feb 13.

9. Myllyharju J. Prolyl 4-hydroxylases, key enzymes in the synthesis of collagens and

regulation of the response to hypoxia, and their roles as treatment targets. Ann

Med. 2008;40(6):402-17.

10. Gournas C, Papageorgiou I, Diallinas G. The nucleobase-ascorbate transporter (NAT)

family: genomics, evolution, structure-function relationships and physiological role. Mol

Biosyst. 2008 May;4(5):404-16. Epub 2008 Mar 11. Review.

11. Verachtert,H.,Rodriguez,P, Bass,S.T.,and Hansen,R.G. Purification and properties of

guanosine diphosphate hexose pyrophosphorylase from mammalian tissues. J Biol

Chem. 1966 May 10;241(9):2007-13.

12. Szumilo T, Drake RR, York JL, Elbein AD. GDP-mannose pyrophosphorylase.

Purification to homogeneity, properties, and utilization to prepare photoaffinity analogs. J

Biol Chem. 1993 Aug 25;268(24):17943-50.

13. Ning, B., and Elbein, A. D. Cloning, expression and characterization of the pig liver

GDP-mannose pyrophosphorylase. Evidence that GDP-mannose and GDP-Glc

pyrophosphorylases are different proteins. Eur. J. Biochem. 2000 Dec;267(23):6866-74.

14. Carlson DM, Hansen RG. The isolation and synthesis of guanosine diphosphate glucose.

J Biol Chem. 1962 Apr;237:1260-5.

138 15. Haeuptle MA, Hennet T. Congenital disorders of glycosylation: an update on defects

affecting the biosynthesis of dolichol-linked oligosaccharides. Hum Mutat. 2009

Dec;30(12):1628-41.

16. Zea CJ, Pohl NL. Unusual sugar nucleotide recognition elements of mesophilic vs.

thermophilic glycogen synthases. Biopolymers. 2005 Oct 5;79(2):106-13.

17. Adler LN, Gomez TA, Clarke SG, Linster CL. A novel GDP-D-glucose phosphorylase

involved in quality control of the nucleoside diphosphate sugar pool in Caenorhabditis

elegans and mammals. J Biol Chem. 2011 Jun 17;286(24):21511-23. Epub 2011 Apr 20.

18. Freeze HH. Genetic defects in the human glycome. Nat Rev Genet. 2006 Jul;7(7):537-51.

Epub 2006 Jun 6

19. Jaeken J. Congenital disorders of glycosylation. Ann N Y Acad Sci. 2010 Dec;1214:190-

8. doi: 10.1111/j.1749-6632.2010.05840.x. Review.

20. Gravato-Nobre MJ, Stroud D, O'Rourke D, Darby C, Hodgkin J. Glycosylation genes

expressed in seam cells determine complex surface properties and bacterial adhesion to

the cuticle of Caenorhabditis elegans. Genetics. 2011 Jan;187(1):141-55. Epub 2010 Oct

21. Weston B Struwe, Bethany L Hughes, David W Osborn, Erica D Boudreau, Kristin M D

Shaw, and Charles E Warren. Modeling a congenital disorder of glycosylation type I

in C. elegans: A genome-wide RNAi screen for N-glycosylation-dependent loci.

Glycobiology.Dec;19(12):1554-62. Epub 2009 Sep

22. McGhee JD. The C. elegans intestine. WormBook. 2007 Mar 27:1-36.

23. Griffitts JS, Huffman DL, Whitacre JL, Barrows BD, Marroquin LD, Müller R, Brown

JR, Hennet T, Esko JD, Aroian RV. Resistance to a bacterial toxin is mediated by

139 removal of a conserved glycosylation pathway required for toxin-host interactions. J. Biol.

Chem. 278: 45594–45602.

24. Barrows BD, Haslam SM, Bischof LJ, Morris HR, Dell A, Aroian RV. Resistance to

Bacillus thuringiensis toxin in Caenorhabditis elegans from loss of fucose. J Biol

Chem. 2007 Feb 2;282(5):3302-11. Epub 2006 Nov 29

25. Hodgkin, J., P. E. Kuwabara and B. Corneliussen. A novel bacterial pathogen,

Microbacterium nematophilum, induces morphological change in the nematode C.

elegans. Curr Biol. 2000 Dec 14-28;10(24):1615-8.

26. Darby C, Chakraborti A, Politz SM, Daniels CC, Tan L, Drace K. Caenorhabditis elegans

mutants resistant to attach- ment of Yersinia biofilms. Genetics. 2007 May;176(1):221-30.

Epub 2007 Mar 4

27. Höflich J, Berninsone P, Göbel C, Gravato-Nobre MJ, Libby BJ, Darby C, Politz

SM, Hodgkin J, Hirschberg CB,Baumeister R. Loss of srf-3-encoded nucleotide sugar

transporter activity in Caenorhabditis elegans alters surface antigenicity and prevents

bacterial adherence. J Biol Chem. 2004 Jul 16;279(29):30440-8. Epub 2004 Apr 30

28. Cipollo JF, Awad AM, Costello CE, Hirschberg CB. srf-3, a mutant of Caenorhabditis

elegans, resistant to bacterial infection and to biofilm binding, is deficient in

glycoconjugates. J Biol Chem. 2004 Dec 17;279(51):52893-903. Epub 2004 Sep

29. van Eijk HG, van Noort WL, Dubelaar ML, van der Heul C. The microheterogeneity of

human transferrins in biological fluids. Clin Chim Acta. 1983 Aug 15;132(2):167-71.

140

APPENDIX I

Glycan Analysis of Wild Type N2 and mcp-1 Mutant C. elegans

141 SUMMARY

Here I describe the search for differences in the structure of N-linked and O-linked protein glycans of the mcp-1 mutant strain of C. elegans that lacks GDP-D-glucose phosphorylase activity. This analysis of the worm protein glycans was done using MALDI-TOF mass spectrometry. This project has been launched in collaboration with the Stuart Haslam lab at the Faculty of Natural Sciences Imperial College London, UK. Following the extract preparation and processing, mass spectrometry analysis revealed only subtle differences between the wild type and the mcp-1 mutant strains. Future work can confirm the initial findings and analyze the glycan profile of mcp-1 mutants that overexpress the worm homolog of GDP-D-mannose pyrophosphorylase A.

INTRODUCTION

GDP-D-glucose can be used in the biosynthesis of the disaccharide trehalose in bacteria

[1] [2] and in the synthesis of cell wall glycoproteins in plants [3]. However, its function in vertebrates and invertebrates is unknown. Interestingly, previous studies revealed that this metabolite can be synthesized in some mammalians tissues [4][5] while GDP-D-mannose pyrophosphorylase A was identified as the potential enzyme that catalyzes this reaction [5]. Our previous work at the Clarke lab revealed that C. elegans mcp-1 mutant worms that have lost the ability to phosphorolyze GDP-D-glucose now accumulate it [6]. Unlike GDP-D-glucose, the role of GDP-D-mannose in the cell as a mannosyl donor for N-glycosylation is well established [7].

Due to the high similarity in structure of GDP-D-glucose and GDP-D-mannose, an increase in

142 the GDP-D-glucose pool could potentially compete with GDP-D-mannose for recognition by mannosyltransferases and as a result either inhibit the normal addition of D-mannose or lead to incorporation of the wrong hexose (D-glucose) into the lipid-linked oligosaccharides (LLO) formed in the N-glycosylation pathway, thus affecting the structure of the newly synthesized glycans. Additionally, it is possible that GDP-D-glucose could bind to GDP-D-mannose 4,6- dehydratase and displace the normal substrate and as a result block GDP-L-fucose biosynthesis.

This appendix includes a collaborative mass spectrometry analysis of the C. elegans glycome in search for abnormalities in the glycan profile of the mcp-1 mutant worms that might be explained by GDP-D-glucose phosphorylase deficiency.

EXPERIMENTAL PROCEDURES

Preparation of C. elegans Pellets at UCLA

Mixed-stage worms from either N2 or mcp-1 deletion strains were fed for 4-5 days in a

50 ml liquid culture in S-medium with 0.4 g of wet weight HB101 E. coli cells (provided by the

Dr. Lars Dreier laboratory at UCLA). The worms were pelleted, washed with M9 buffer and E. coli cells were removed with a sucrose float, followed by additional washes in M9. The washed worms were incubated in M9 with gentle agitation for 45 minutes at room temperature to allow the digestion of the gut residing bacteria. The cleaned worm pellets (about 0.2 g wet weight) were washed twice with phosphate-buffered saline and following the final wash the pellets were quickly frozen in liquid nitrogen and stored at -80 °C until shipped on dry ice to Imperial College in London.

143 Processing of C. elegans Pellets for Glycan Analysis at Imperial College

N2 and mcp-1 worms pellets were homogenized on ice in a homogenization buffer consisting of 1% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane sulfonate (CHAPS)

(v/v) in 25 mM Tris, 150 mM sodium chloride (NaCl), 5 mM EDTA in water, pH 7.4 (adjusted with dilute acetic acid). The homogenates were transferred to snakeskin dialysis tubing with a cut off point of 7 kDa (Pierce, Prod # 68700) and left to dialyze over a period of 48 h at 4 ºC against Dialysis buffer consisting of 50 mM ammonium hydrogen carbonate, pH 7.4 (adjusted with dilute acetic acid. Reduction of C. elegans proteins was carried out by adding 1-2 ml of 2 mg/ml dithiothreitol (DTT) in Tris–GuHCl Buffer to the samples for 60 min at 37 ºC. Addition of 1 ml of 12 mg/ml iodoacetic acid (IAA) in 0.6 M Tris buffer, pH 8.5 and incubating for 90 min at room temperature in the dark was preformed to protect the modify the free sulfhydral groups of the proteins. 1.5 µg of a stock of 1 µg/µL porcine pancreas trypsin in 50 mM ammonium hydrogen carbonate, pH 8.4 solution was added for every 1 mg of homogenate and incubated at 37 ºC for 14 h following a C18 Sep-Pak purification of the glycopeptides as described in North et al., 2010 [8]. After trypsin digestion each sample was divided into two: one half was used for analysis of glycans not subjected to exoglycosidase digestion, while the second half was kept for future analysis of glycans to be subjected to α-mannosidase degradation. N- glycosidase F (PNGase-F) digestion was carried out with addition of 5 U every 12 h of PNGase-

F in glycerol (Roche applied science) into each sample, followed by incubation at 37 ºC for a total of 24 h. The remaining glycopeptide fraction was subjected to N-glycosidase A (PNGase-

A) digestion by adding 0.2 mU of PNGase-A from sweet almond meal (Roche applied science) every 12 h and incubating at 37 ºC for a total of 24 h.

144 Permethylation of the samples was done by adding sodium hydroxide dissolved in dimethylsulphoxide (DMSO) and 0.6 ml of methyl iodide followed by agitation for 20 min and termination of the reaction by the addition of ultra-pure water. The permethylated samples were purified on a C18 Sep-Pak column as described above and extracted by chloroform before subjecting them to MS analysis.

MALDI-TOF and MALDI-TOF-TOF Analysis at Imperial College

The samples were dissolved in 2,5- dihydroxybenzoic acid (DHB) matrix and the screening analysis of glycans was performed on an ABI Voyager-DE STR MALDI-TOF instrument. All mass spectra were acquired in the positive ion mode.

RESULTS AND DISCUSSION

Glycan analysis of C. elegans strains was done in collaboration with graduate student

Grigorij Sutov from Dr. Stuart Haslam’s lab at London Imperial College, UK. Worm pellets from either the wild type or the mcp-1 mutant strains were prepared and frozen in our lab and sent for processing and mass spectrometry analysis to Dr. Haslam’s lab.

The work plan for the N- and O-glycan analyses of the C. elegans lysates is presented in

Figure 1. Briefly, for N-glycan analysis, the sugars were sequentially released by PNGase-F and

PNGase-A enzyme digestions, which are known to have differential activities towards structures bearing the fucose at the 3-position of reducing end GlcNAc, where PNGase F is inhibited while

PNGase A is not [9]. The differential substrate specificity between these enzymes indicates that

145 fucosylated glycans released by PNGase-A would have a fucose attached to the 3-position of reducing end GlcNAc. Further confirmation by MS/MS experiments is in progress. After derivatization by permethylation, which increases sensitivity, the glycans were subjected to a final C18 reverse phase chromatography purification with resultant 35%, 50% and 75% acetonitrile fractions. Most of the glycans usually elute in the 35% and/or 50% acetonitrile fractions. At this initial analysis step, glycans with different hexose compositions (mannose, galactose or glucose) could not be segregated. An additional step that includes digestion with α- mannosidase is needed to separate the glycans based on the identity of their hexoses (Figure 1).

The O-glycans were released from the extracts by reductive elimination followed by separation on Sep-Pak C18 and derivatization as described in the "Experimental Procedures" section.

One possible consequence of the increase in the GDP-D-glucose pool in the mcp-1 mutant animals is that GDP-D-glucose is competing with GDP-D-mannose for the binding to mannosyltransferases. As a result glucose units might be placed instead of mannose units on the growing LLOs, which could lead to either hypoglycosylation or aberrant glycosylation. This perspective is also appealing as high-mannose structures were reported to dominate the C. elegans glycome [10]. Alternatively, as GDP-D-mannose is the substrate for GDP-D-mannose

4,6-dehydratase in the only pathway for GDP-L-fucose synthesis in C. elegans [11], the presence of GDP-D-glucose in the cell could potentially compete with GDP-D-mannose for the recognition by GDP-D-mannose 4,6-dehydratase and thus inhibit GDP-L-fucose production.

The C. elegans bre-1 mutant which is deficient in GDP-D-mannose 4,6-dehydratase activity demonstrated an interesting mass spectrometric glycome profile that completely lacks any fucosylated structures [11].

146 The mass spectra for the N-glycans digested with PNGase F and PNGase A revealed subtle changes between the wild type and the mcp-1 mutant extracts (Figures 2-4) while the O- glycan profiles did not display any difference between the two strains (Figure 5).

The profile of N-glycans digested with PNGase F eluted with 35% acetonitrile shows that most glycan species in the mcp-1 mutant spectrum are slightly more abundant than in the wild type spectrum (Figure 2), while the spectra of N-glycans that were eluted with 50% acetonitrile show the reverse trend with higher peaks for the wild type strain (Figure 3). Moreover, one of the

N-glycan species at 2081 m/z is missing from the mcp-1 profile (Figure 3). The profile of N- glycans digested with PNGase A eluted with 35% acetonitrile also reveals that while all the N- glycan species exist in both spectra, the intensity of the peaks is lower for the mcp-1 mutant

(Figure 4).

These subtle differences between the glycan profiles of the two strains could be a result of a small variation within the worm population, extract processing or due to instrument handling.

On the other hand, these alterations could be real and due to the GDP-D-glucose phosphorylase mutation. One indication that these differences for N-glycans are not artefactual is the absence of differences detected for O-glycans. However, since the data sets acquired here represent only one set of extracts, more work is needed to confirm the results and analyze for statistical significance.

Two additional sets of extracts that include the wild type and the mcp-1 mutants were prepared and analysis is ongoing.

Future work will focus on finding growth conditions that may enhance differences in the glycome of wild type and mcp-1 mutant worms. We will try to stress the mcp-1 mutant worms by either increasing GDP-D-glucose levels internally by overexpressing the putative GDP-D- glucose forming enzyme GMPPA (see below) or by adding tunicamycin. The latter compound

147 blocks the first step of the protein N-glycosylation pathway and has been used to reveal and increase phenotype severity in worms with glycosylation defects [12].

With the help of Dr. Yu Sun from Dr. Lars Dreier’s laboratory at the UCLA Department of Neurology we were able to create transgenic worm strains that overexpress Y47D9A.1, the worm homolog of the mammalian GDP-D-mannose pyrophosphorylase A in the mcp-1 mutant background. These strains are expected to have a higher GDP-D-glucose accumulation due to the over production of GDP-D-glucose by GDP-D-mannose pyrophosphorylase A that cannot be eliminated from the cells in the mcp-1 mutants. We hypothesize that such a further increase in the GDP-D-glucose pool might elicit a more severe glycosylation phenotype that could be more readily detectable by mass spectrometry analysis. I have prepared fresh worm extracts of the new transgenic strain and mass spectrometry analyses of these extracts are in progress. In additional worm protein and small molecule extracts were prepared to assay GDP-D-mannose pyrophosphorylase activity and to measure GDP-D-glucose levels accumulated in wild type and transgenic strains, respectively.

ACKNOWLEDGEMENT

I would like to thank our collaborators Dr. Stuart Haslam, Dr Anne Dell and Grigorij

Sutov from London Imperial College at UK for the mass spectrometry analyses presented in

Figures 1-5. I would also like to thank Drs. Yu Sun and Lars Dreier at the UCLA Department of

Neurology for their help in creating the overexpressing worm strains.

148

Figure 1 - Flow chart for N- and O-glycan analysis of C. elegans pellets prepared at UCLA for processing in the Stuart Haslam laboratory. The preparation of the N2 and mcp-1 mutant strain pellets as well as the sample processing and mass spectrometry analysis is described in the

"Experimental Procedures" section. Figure provided by Dr. Stuart Haslam.

149

Figure 2 - MS spectrum of PNGase-F released C. elegans N-glycans from the 35% acetonitrile fraction. The spectrum above represents material from the wild type N2 strain; the spectrum below from the mcp-1 mutant strain. Configurations of monosaccharides have been assumed to be as illustrated based on references [8] [13]. For monosaccharide designations see Figure 6.

Figure provided by Dr. Stuart Haslam.

150

Figure 3 - MS spectrum of PNGase-F released C. elegans N-glycans, 50% acetonitrile fraction.

The spectrum above represents wild type N2 strain, spectrum below mcp-1 mutant strain.Configurations of monosaccharides have been assumed to be as illustrated based on references [8] [13]. For monosaccharide configuration key see Figure 6. Figure provided by Dr.

Stuart Haslam.

151

Figure 4 - MS spectrum of PNGase-A released C. elegans N-glycans, 35% acetonitrile fraction.

The spectrum above represents wild type N2 strain, spectrum below mcp-1 mutant strain.

Configurations of monosaccharides have been assumed to be as illustrated based on references

[8] [13]. For monosaccharide configuration key see Figure 6. Figure provided by Dr. Stuart

Haslam.

152

Figure 5 - MS spectrum of C. elegans O-glycans released by reductive elimination, 35% acetonitrile fraction. The spectrum above represents wild type N2 strain, spectrum below mcp-1 mutant strain. Configurations of monosaccharides have been assumed to be as illustrated based on references [8] [13]. For monosaccharide configuration key see Figure 6. Figure provided by

Dr. Stuart Haslam.

153

Figure 6 - a) Symbolic nomenclature as outlined by the Consortium for Functional Glycomics

Nomenclature Committee (May 2004) Varki et al., 2009 [13]. b) The mass of a permethylated sugar is calculated by adding the sum of the increment (or residue) masses of the sugars (a) to the sum of the masses of the reducing and non-reducing ends. This figure was taken from North et al., 2010.

154

REFERENCES

1. Silva Z, Alarico S, da Costa MS. Trehalose biosynthesis in Thermus thermophilus RQ-1:

biochemical properties of the trehalose-6-phosphate synthase and trehalose-6-phosphate

phosphatase. Extremophiles. 2005 Feb;9(1):29-36. Epub 2004 Sep 29

2. Elbein AD, Pan YT, Pastuszak I, Carroll D. New insights on trehalose: a multifunctional

molecule. Glycobiology. Apr;13(4):17R-27R. Epub 2003 Jan 22

3. Liepman AH, Wilkerson CG, Keegstra K. Expression of cellulose synthase-like (Csl)

genes in insect cells reveals that CslA family members encode mannan synthases. Proc

Natl Acad Sci U S A. 2005 Feb 8;102(6):2221-6. Epub 2005 Jan 12

4. Verachtert,H.,Rodriguez,P, Bass,S.T.,and Hansen,R.G. Purification and properties of

guanosine diphosphate hexose pyrophosphorylase from mammalian tissues. J Biol

Chem. 1966 May 10;241(9):2007-13.

5. Ning, B., and Elbein, A. D. Cloning, expression and characterization of the pig liver

GDP-mannose pyrophosphorylase. Evidence that GDP-mannose and GDP-Glc

pyrophosphorylases are different proteins. Eur. J. Biochem. 2000 Dec;267(23):6866-74.

6. Adler LN, Gomez TA, Clarke SG, Linster CL. A novel GDP-D-glucose phosphorylase

involved in quality control of the nucleoside diphosphate sugar pool in Caenorhabditis

elegans and mammals. J Biol Chem. 2011 Jun 17;286(24):21511-23. Epub 2011 Apr 20

7. Struck DK, Lennarz WJ. Utilization of Exogenous GDP-Mannose for the Synthesis of

Mannose-containing Lipids and Glycoproteins by Oviduct Cells.J Biol Chem. 1976 Apr

25;251(8):2511-9.

155 8. North SJ, Jang-Lee J, Harrison R, Canis K, Ismail MN, Trollope A, Antonopoulos

A, Pang PC, Grassi P, Al-Chalabi S, Etienne AT, Dell A, Haslam SM.

Mass spectrometric analysis of mutant mice. Methods Enzymol. 2010;478:27-77.

9. Tretter V, Altmann F, März L. Peptide-N4-(N-acetyl-beta-glucosaminyl) asparagine

amidase F cannot release glycans with fucose attached alpha 1–3 to the asparagine-linked

N-acetylglucosamine residue. Eur. J. Biochem. 1991 Aug 1;199(3):647-52.

10. Haslam SM, Gems D, Morris HR, Dell A. The glycomes of Caenorhabditis elegans and

other model organisms. Biochem Soc Symp. 2002;(69):117-34. J.Biol. Chem. 241, 2007–

2013.

11. Barrows BD, Haslam SM, Bischof LJ, Morris HR, Dell A, Aroian RV. Resistance to

Bacillus thuringiensis toxin in Caenorhabditis elegans from loss of fucose. J Biol Chem.

Feb 2;282(5):3302-11. Epub 2006 Nov 29

12. Weston B Struwe, Bethany L Hughes, David W Osborn, Erica D Boudreau, Kristin M D

Shaw, and Charles E Warren. Modeling a congenital disorder of glycosylation type I

in C. elegans: A genome-wide RNAi screen for N-glycosylation-dependent loci.

Glycobiology. 2009 Dec;19(12):1554-62. Epub 2009 Sep 3

13. Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Marth JD, Bertozzi CR, Hart

GW, Etzler ME. Symbol nomenclature for glycan representation. Proteomics. 2009

Dec;9(24):5398-9.

156