Studies on the Role of UDP-Glucose Dehydrogenase in Polysaccharide Biosynthesis

Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1388

BY ELISABET ROMAN

ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2004 !"# $ % & $ $ '% % ( $ )* +% , -,%*

. /* * - % . $ '01 % & ' % %* 2 * !33* #3 * * 4-5 60## 07 3308

' % $ $ $ $ * +% 9 % $ % & * : % & '0& ,%% $ $ '0& % '0& % & (1;) 9* 1; % % & % % $ % * +% $ % % % , & % $ 1; % % % % $$ " % & , & +% $$ $ 1;0 % % ,% % '0& , & * 2 1; , $ < 52 * + $ % 9 1;0 * 5 % & & $ *7 < ,% % $ !* < $ *7 <* : $ % 1; % $ % & & = % % % % % , % % & & % * +% $$ & & % $ '0& , $ % % (; !) ,%% > '0& * : $ ; ! $ ; ! 1; %&% & $ % , % $$ & % $ % $ %* / $ % 1; $ $ '0& % $ $, ?# % % * : $ 1 $ $ 1; & , $* 2 $ % , % % , % * ' % '0& $ '0& *

! " '0& % & % % & & 2* % /* ?#

# $ % # & %' ()* +,(-*. !

@ / .

4--5 308 87 4-5 60## 07 3308 " """ 0 7#8 (% "AA *<*A B C " """ 0 7#8) "Science is not absolute truth but a journey." D. T. A. Lamport

List of Papers

This thesis is based on the following papers, in the text referred to by Roman numerals:

I Lind, Falk, Hjertson, Kusche-Gullberg, Lidholt (1999) "cDNA cloning and expression of UDP-glucose dehydrogenase from bovine kidney" Glycobiology, 9, (6): 595-600

II Roman, Roberts, Lidholt, Kusche-Gullberg (2003) "Overexpression of UDP-glucose dehydrogenase in Escherichia coli results in decreased biosynthesis of K5 polysaccharide" Biochem. J., 374, 767- 772

III Roman, Brinck, Heldin, Kusche-Gullberg "Glycosaminoglycan biosynthesis in UDP-glucose dehydrogenase overexpressing 293 cells", manuscript

IV Johannesson, Roman, Andersson, Åman, Lidholt, Engström "In- creased expression of a UDP-glucose dehydrogenase gene in Arabi- dopsis thaliana results in altered cell wall composition and dwarfism", submitted

Reprints of paper 1 were made with permission from the Society for Glyco- biology and reprints of paper 2 made with permission from the Biochemical Society.

Contents

Introduction and aim...... 11 Background ...... 12 UDP-Glucose dehydrogenase...... 12 General...... 12 Role of UGDH...... 13 Mechanism of UGDH...... 15 Nucleotide compounds...... 17 UDP-Glucose...... 17 UDP-Glucuronic acid ...... 18 UDP-Xylose...... 19 Measuring nucleotide sugar levels...... 20 NAD+/NADH...... 21 Sugar transport ...... 22 Sugar transport into the cell...... 22 Nucleotide sugar transport into the Golgi compartment...... 23 Mammalian glycosaminoglycans...... 24 Hyaluronan ...... 24 Heparan sulphate/Heparin ...... 25 Chondroitin sulphate/ Dermatan sulphate...... 26 Bacterial glycosaminoglycans...... 27 K5 polysaccharide ...... 27 Other glycosaminoglycan-producing bacteria...... 28 Plant polysaccharides ...... 29 Present investigation...... 31 Aims ...... 31 Results ...... 31 Paper I...... 31 Paper II ...... 32 Paper III (Manuscript) ...... 34 Paper IV (Manuscript) ...... 36 Discussion ...... 37 UGDH gene expression ...... 37 UDP-sugar levels...... 38 GAG biosynthesis...... 39 Other aspects of UGDH...... 40 Future perspectives...... 42 Acknowledgements ...... 43 Populärvetenskaplig sammanfattning på svenska (Summary in Swedish) ...... 45 References...... 48 Abbreviations

CS chondroitin sulphate ER endoplasmic reticulum GAG glycosaminoglycan Gal D-D-galactose GalNAc N-acetyl-D-D-galactosamine Glc D-D-glucose GlcA D-D-glucuronic acid GlcNAc N-acetyl-D-D-glucosamine HA hyaluronan Has hyaluronan synthase HS heparan sulphate NAD+ nicotinamide-ß-D-adenine dinucleotide (oxidised) NADH nicotinamide-ß-D-adenine dinucleotide (reduced) UDP uridine 5'-diphosphate UGDH UDP-glucose dehydrogenase UMP uridine 5'-monophosphate UTP uridine 5'-triphosphate Xyl D-D-xylose

Introduction and aim

Complex carbohydrates surround all living cells. The carbohydrates are produced by the cells and vary greatly in both monosaccharide composition (hence, the name complex) and structure. The diversity is perplexing. Great efforts have been put into elucidating the roles and functions of the carbohydrates, as well as the possible purposes of their diversity. Despite the efforts and notable advances made, many questions still remain unanswered. One group of the complex carbohydrates are the heteropolysaccharides, chains of at least twenty monosaccharide residues of at least two different types. Some heteropolysaccharides contain glucuronic acid (GlcA) as one of the constituent monosaccharides. The monomeric precursor of the GlcA residue in the polysaccharide is UDP-glucuronic acid (UDP-GlcA). UDP- GlcA is produced by UDP-glucose dehydrogenase (UGDH) which has been found in many forms of life. Due to different characteristics of UGDH, it has been proposed to be a key enzyme in the biosynthesis of heteropolysaccharides. The aim of the studies presented in this thesis was to elucidate the role of UGDH in the biosynthesis of different heteropolysaccharides, with an emphasis on glycosaminoglycans (GAGs). This knowledge could possibly provide tools in future research for manipulation of the respective biosyntheses, an interest ranging from the paper industry to the pharmaceutical industry. The investigations were performed by overexpressing UGDH in bacteria, plants and mammalian cell culture, and studying the effects on polysaccharide production. The introductory part of this thesis contains background information to the studies addressed in Present investigation and includes a large part of factors involved in the shaping of polysaccharides. Other parts have been left out, since they are not within the scope of this work. Therefore such topics as turnover and degradation of nucleotide sugars and polysaccharides, and most work performed in Drosophila, C. elegans and zebrafish have been omitted.

11 Background

UDP-Glucose dehydrogenase General UDP-glucose dehydrogenase (1.1.1.22) oxidises UDP-glucose (UDP-Glc) to UDP-GlcA in a two-step NAD-dependent reaction (Fig. 1).

CH OH + CHO COOH 2 NAD NADH NAD + NADH O O O

O UDP O UDP O UDP

+ + UDP-Glc + 2 NAD + H2O UDP-GlcA + 2 NADH + 2 H Figure 1. Reaction performed by UDP-glucose dehydrogenase. The hydroxyl groups on C2-C4 have been excluded for clarity; only their orientation is indicated (Gainey et al. 1975).

The enzyme was first purified from guinea pig liver in 1954 and from pea seedlings in 1957 by Strominger et al (Strominger et al. 1954; Strominger et al. 1957). Since then it has been detected in species ranging from viruses to human (Table 1). A bacterial isoform of the enzyme was crystallised in 2000. The charac- terisation was performed on native and mutated forms of the enzyme at the 2 Å level (Campbell et al. 2000). The cDNA and gene have been cloned from several species, and the chromosome location of the gene has been mapped for human and mouse (Marcu et al. 1999; Spicer et al. 1998). The amino acid sequences of different species have been found to have striking similarity, which is exemplified in figure 2. The human and bovine sequences are 99 % similar, human and bacteria (E. coli, KfiD) 58 % similar and finally, human and plant (Arabidopsis, UGD1) have 21% similarity. The enzyme is believed to be a single copy gene in the human genome, whereas Arabidopsis thaliana (mouse-ear cress) contains four genes, of which all are expressed (Reiter et al. 2001). Bacillus subtilis has been reported to have two UGDH genes (Mijakovic et al. 2003). The enzyme has no transmembrane part and is cytosolic. However, both bacterial and fungal UGHD enzymes have been reported to be membrane-associated (Griffith et al. 2004; Rigg et al. 1998).

12 Table 1. UGDH genes from selected species Species Gene Reference

VIRUSES Chlorella virus PBCV-1 a609l (Landstein et al. 1998)

BACTERIA E. coli (K5), (K4) ugd(kfiD), (Sieberth et al. 1995), (kfoF) (Ninomiya et al. 2002) Pasteurella multocida hyaC (Chung et al. 1998) Streptococcus pneumoniae cap3A, (Dillard et al. 1995) cps3D (Arrecubieta et al. 1996) Streptococcus pyogenes hasB (Dougherty et al. 1993),

FUNGI Cryoptococcus neoformans UGD1 (Bar-Peled et al. 2004), (Griffith et al. 2004)

PLANTS Arabidopsis thaliana Ugd (Seitz et al. 2000).

ANIMALS Caenorhabditis elegans sqv-4 (Hwang et al. 2002) Danio rerio (zebrafish) jekyll (Walsh et al. 2001) Drosophila melanogaster kiwi, (Binari et al. 1997) sugarless, (Hacker et al. 1997) suppen-kasper (Haerry et al. 1997) Homo sapiens UGDH (Spicer et al. 1998) Xenopus laevis xsgl TrEMBL: Q8UV25

Role of UGDH The most apparent role of UGDH is obviously to provide UDP-GlcA. UDP- GlcA is used in the biosynthesis of glycans, glycolipids and GAGs, as well as to produce further biosynthetic precursors. Nucleotide sugars are energy- rich compounds and costly for the cell to form. Therefore, it is vital that these compounds are channelled to where they are needed and are not wasted. UGDH has been suggested to be a regulatory or rate-limiting enzyme that control part of the polysaccharide biosynthesis in plants and animals (De Luca et al. 1976; Hickery et al. 2003; Robertson et al. 1995; Wegrowski et al. 1998). The definition of a regulatory enzyme is that it is an enzyme that defines the rate of a metabolic pathway. Thus, the rate is not primarily gov- erned by the amount of substrate but by the activity of the rate-limiting en-

13 zyme, which in turn is regulated by one or several factors. These factors include amount of enzyme, covalent and/or allosteric modification of the enzyme (Nelson et al. 2000b; Smith et al. 1997).

Human M F E I K ¡K ¡I C ¡C I ¡¡G A G Y V ¡G G P T ¡C S V I ¡ A H M C P E I R ¡V T ¡V V ¡D V N E S R I N A W N Bovin M F E I K K I C C I G A G Y V G G P T C S V I A H M C P E I R V T V V D I N E S R I N A W N UGD1 M V - - - K I C C I G A G Y V G G P T M A V I A L K C P D I E V A V V D I S V P R I N A W N KfiD M F G T L K I T V S G A G Y V G L S N G I L M A Q - - - N H E V V A F D T H Q K K V D L L N

Human S P T L P I Y E P G L K E V V E S C R G K N L F F S T N I D D A I K E A D L V F I S V N T P Bovin S P T L P I Y E P G L K E V V E S C R G K N L F F S T N I D D A I K E A D L V F I S V N T P UGD1 S D Q L P I Y E P G L D D I V K Q C R G K N L F F S T D V E K H V R E A D I V F V S V N T P KfiD D K L S P I E D K E I E N Y L S T - - - K I L N F R A T T N K Y E A Y K N A N Y V I I A T P

Human T K T Y G M G K G R A A D L K Y I E A C A R R I V Q N S N G Y K I V T E K S *T V P V R A A E Bovin T K T Y G M G K G R A A D L K Y I E A C A R R I V Q N S H G Y K I V T E K S T V P V R A A E UGD1 T K T T G L G A G K A A D L T Y W E S A A R M I A D V S V S D K I V V E K S T V P V K T A E KfiD T N - Y D P G S - N Y F D T S S V E A V I R D V T E I N P N - A I M V V K S T V P V G F T K

Human S I R R I F D A N T K P N L N L Q V L S N P *E F L A E G T A I K D L K N P D R V L I G G D E Bovin S I R R I F D A N T K P N L N L Q V L S N P E F L A E G T A I K D L K N P D R V L I G G D E UGD1 A I E K I L M H N S K - G I K F Q I L S N P E F L A E G T A I A D L F N P D R V L I G G R E KfiD T I K E H L G I N N ------I I F S P E F L R E G R A L Y D N L H P S R I I I G E C S

Human T P E G Q R A V Q A L C A V Y E H W V P R E K I L T T N T W S S E L S *K L A A *N A F L A Q R Bovin T P E G Q R A V Q A L C A V Y E H W V P R E K I L T T N T W S S E L S K L T A N A F L A Q R UGD1 T P E G F K A V Q T L K E V Y A N W V P E G Q I I T T N L W S A E L S K L A A N A F L A Q R KfiD E R A E R L A V L F Q E G A I K Q N I P - - - V L F T D S T E A E A I K L F S N T Y L A M R

Human I S S I N S I S A L C E A T G A D V E E V A T A I G M D Q R I G N K F L K A S V G F G G S *C Bovin I S S I N S I S A L C E A T G A D V E E V A T A I G M D Q R I G N K F L K A S V G F G G S C UGD1 I S S V N A M S A L C E S T G A D V T Q V S Y A V G T D S R I G S K F L N A S V G F G G S C KfiD V A F F N E L D S Y A E S F G L N T R Q I I D G V C L D P R I G N Y Y N N P S F G Y G G Y C

Human F Q K *D V L N L V Y L C E A L N L P E V A R Y W Q Q V I D M N D Y Q R R R F A S R I I D S L Bovin F Q K D V L N L V Y L C E A L N L P E V A R Y W Q Q V I D M N D Y Q R R R F A S R I I D S L UGD1 F Q K D I L N L V Y I C Q C N G L P E V A E Y W K Q V I K I N D Y Q K N R F V N R I V S S M KfiD L P K D T K Q L - - L A N Y Q S V P N - - - - - K L I S A I V D A N R T R - - K D F I T N V

Human F N T V T D K K I A I L G F A F K K D T G D T R E S S S I Y I S K Y L M D E G A H L H I Y D Bovin F N T V T D K K I A I L G F A F K K D T G D T R E S S S I Y I S K Y L M D E G A H L H I Y D UGD1 F N T V S N K K V A I L G F A F K K D T G D T R E T P A I D V C K G L L G D K A Q I S I Y D KfiD I L K H R P Q V V G V Y R L I M K S G S D N F R D S S I L G I I K R I K K K G V K V I I Y E

Human P K V P R E Q I V V D L S H P G V S E D D Q V ------S R L V T I S K D P Y E A C Bovin P K V P R E Q I V V D L S H P G V S K D D Q V ------A R L V T I S K D P Y E A C UGD1 P Q V T E E Q I Q R D L S M K K F D W D H P L H L Q P M S P T T V K Q V S V T W D A Y E A T KfiD P L I S G D T F ------F N S P L E R E L A I ------

Human D G A H A V V I C T E W D M F K E L D Y E R I H K K M L K P A F I F D G R R V L D G L H N E Bovin D G A H A V V I C T E W D M F K E L D Y E R I H K K M L K P A F I F D G R R V L D G L H N E UGD1 K D A H A V C V L T E W D E F K S L D Y Q K I F D N M Q K P A F I F D G R N I M N V - - N K KfiD ------F K G ------K A D I I I T N R - - - - - M S E E

Human L Q T I G F Q I E T I G K K V S S K R I P Y A P S G E I P K F S L Q D P P N K K P K V Bovin L Q T I G F Q I E T I G K K V S S K R I P Y A P S G E I P K F S L Q D M P N K K P R V UGD1 L R E I G F I V Y S I G K P L D P ------W L K D M P - - - A F V KfiD L N D V ------V D K V Y S R D L F K C D

Figure 2. Alignment of amino acid sequences of human, bovine, A. thaliana (UGD1) and E. coli K5 (KfiD) UGDHs. The sequences were collected from Swiss- Prot: Human, O60701; bovine, P12378; KfiD, Q47329 and from TrEMBL: UGD1, Q9FM01. The alignment was made using MegAlign: Clustal method (PAM 250 residue weight table). The circles indicate amino acids of the ADP-binding fold as described in (Wierenga et al. 1986) and the asterisks catalytic amino acids described in (Campbell et al. 2000); those in bold are the residues in Fig. 3 in this thesis.

14 The suggestions that UGDH may be regulatory are based on several ob- servations, indicative of a regulatory enzyme: UGDH has a key position and creates a branch from the ordinary UDP-Glc metabolism. The action of UGDH is irreversible and the activity depends on multimericity (in turn dependent on its concentration, (Jaenicke et al. 1986)). Furthermore, the enzyme is feedback inhibited both by its main product, UDP-GlcA, and coenzyme-product, NADH (Ordman et al. 1977), as well as the downstream product UDP-Xyl (Gainey et al. 1972; Neufeld et al. 1965). These characteristics of UGDH will be discussed further below. Additionally, the expression of UGDH has been found to increase in response to interleukin 1E (an in- flammatory cytokine), thus adding additional regulating factors that may influence the activity of UGDH (Spicer et al. 1998). Recently, there have been reports of tyrosine-kinases phosphorylating bacterial UGDHs. Phos- phorylation of UGDH, which has been observed in both gram-negative and gram-positive bacteria, seems to activate the enzyme and opens for yet another regulatory mechanism (Grangeasse et al. 2003; Mijakovic et al. 2003).

Mechanism of UGDH In most eukaryotes UGDH is a homohexamer of a300 kDa (Feingold et al. 1981). Rat together with Cryptococcus fungi, seem to be exceptions, being reported to have UGDH in a tetrameric (Sivaswami et al. 1972), and dimeric form, respectively (Bar-Peled et al. 2004). Irrespective of multimericity, the basic enzyme unit is a dimer that displays half-the-sites-reactivity, i.e. at saturation one substrate-molecule is bound per dimer (Feingold et al. 1981). But in prokaryotes it has been claimed to be active both as a 94 kDa dimer (Schiller et al. 1976) and a 45 kDa monomer (Campbell et al. 1997). In contrast, the mammalian enzyme is inactive in the monomeric state (Jaenicke et al. 1986). The action mechanism was extensively studied during the 70-es and has been proposed to be a "bi-uni-uni-ping-pong" mechanism (Campbell et al. 1997): First the substrate, UDP-Glc, and then the coenzyme, NAD+, bind to the enzyme. In the first half of the reaction, UDP-Glc is oxidized to an alde- hyde and the NAD+ is reduced to NADH and released. After that, a second NAD+ molecule binds to the enzyme, and is reduced to NADH, as the alde- hyde is oxidized to UDP-GlcA. UDP-GlcA is released after the second NADH has been released (Fig. 3). Also the plant UGDH seems to function in the same sequential order, as reported from a study in sugarcane (Turner et al. 2002).

15 Figure 3. Proposed action mechanism of human UGDH with key residues identified. Reprinted, with permission, from (Sommer et al. 2004).

Due to the striking similarity in primary amino acid sequence (Fig. 2) the UGDHs in different species probably have similar action mechanisms, although this remains to be established (Ge et al. 2004; Sommer et al. 2004). Both the NAD binding sequence, originally identified by Wierenga et al (Wierenga et al. 1986), and the catalytic amino acids (Campbell et al. 2000) are found in the sequences identified so far (Fig. 2). The catalytic cysteine (marked with one of the bold asterisks) is thought to form a covalent bond with the substrate, since no free reaction intermediates have been found (Zalitis et al. 1968). This covalent bond, in the form of a thiolester, could also explain the experimental finding that the reaction is irreversible, since thiolesters have high negative free energy values upon hydrolysis (Ordman et al. 1977). The pH optimum of the reaction has been established for both mammalian and bacterial isoforms and found to be around 9.0. The reaction is mostly performed at 8.7 due to the greater stability of the enzyme at this pH (Zalitis et al. 1972).

16 The reported KM-value of UGDH for UDP-Glc varies greatly, from 0.01 to 8.4 mM, likewise the KM-value for NAD varies between 0.02 and 0.4 mM (Schomburg et al. 2004, http://www.brenda.uni-koeln.de/). Interestingly, one group has reported differences with respect to kinetic characteristics in UGDH extracted from various tissues and suggested the existence of tissue specific forms of the enzyme (Bardoni et al. 1989; Castellani et al. 1986).

Nucleotide compounds Nucleotide sugars are activated forms of monosaccharides, ready to be in- corporated into oligo- or polysaccharides or transferred onto proteins or lip- ids. In mammals there are nine different nucleotide sugars. In plants, there are even more nucleotide sugars, of which most are UDP-sugars. In this thesis, only the UDP-sugars will be discussed. The relevant parts of sugar metabolism and nucleotide sugar interconversions are shown in figure 4a, b and c, for the mammalian, bacterial and plant systems respectively. It should be noted that the relative contribution of each pathway is unknown and that no difference has been made between de novo synthesising and recycling pathways. A presentation of three nucleotide- sugars follows below; the structures of the saccharide moieties are shown in figure 5.

UDP-Glucose UDP-Glc is formed intracellularly from Glc-1-P and UTP, by glucose pyrophosphorylase. From UDP-Glc all other UDP-sugars can be formed. UDP- Glc is used in glycosylation reactions, and serves as a precursor in mammalian glycogen synthesis (Fig. 4a). In plants, UDP-Glc is at the crossroads of several processes being a precursor for sucrose, cellulose and non-cellulosic polysaccharides (Fig. 4c). Note that ADP-Glc, used for bacterial glycogen production and plant starch biosynthesis has not been included in the figures 4b and c. Recently, an extracellular G-protein coupled receptor for UDP-Glc was discovered. The finding of this receptor suggests that there are still unknown roles of UDP-Glc (Abbracchio et al. 2003).

17 Gal Gal-1-P UDP-Gal KS Initiation of CS/HS-PG:s Glc Glc-6-P Glc-1-P UDP-Glc

Fru-6-P Glycogen UDP-GlcA CS, HA, HS

Initiation of GlcN GlcN-6-P Glycolysis UDP-Xyl CS/HS-PG:s

GlcNAc GlcNAc-6-P GlcNAc-1-P UDP-GlcNAc HA, HS, KS

GalNAc GalNAc-1-P UDP-GalNAc CS Figure 4a Interconversions of monosaccharides relevant for glycosaminoglycan biosynthesis in mammals (Freeze 1999; Mroz et al. 2003).

UDP-Glucuronic acid In bacteria and animals, UDP-GlcA is believed to be formed exclusively via the path of UGDH-dependent oxidation. This pathway exists in plants also, but is complemented by the myo-inositol pathway, where UDP-GlcA is formed from GlcA-1-P and UDP (Fig. 4c) (Loewus et al. 1973). In mammals the UDP-GlcA is utilised in GAG biosynthesis but also to glucuronidate xenobiotics (foreign molecules) such as drug metabolites (Tukey et al. 2000). Addition of a GlcA renders the molecule less active and more polar and is then easier excluded from the body. In plants, GlcA is a vital component of the polysaccharide groups hemicellulose and pectin, which constitute a good part of the cell wall. UDP- GlcA is also the origin of other nucleotide sugars used in pectin and hemicellulose biosynthesis: UDP-galacturonic acid, UDP-xylose, UDP-apiose and UDP-arabinose. In bacteria the GlcA-containing polysaccharides are extracellular, many of the GAG type (see below).

As a curiosity it can be mentioned that GlcA is a precursor in ascorbic acid biosynthesis in plants and many non-primate animals, although the pathways in plants and animals differ from each other (Dutton 1966; Nelson et al. 2000a).

18 Glc Glc-6-P Glc-1-P UDP-Glc

Fru-6-P Glycolysis UDP-GlcA K5

GlcN GlcN-6-P GlcN-1-P

GlcNAc GlcNAc-6-P GlcNAc-1-P UDP-GlcNAc K5

PEP UDP-MurNAc Peptidoglycan Figure 4b. Interconversions of monosaccharides relevant for K5 biosynthesis in E. coli ((N. 2002) : http://biocyc.org/), activity of KfiD in E.coli K5 added. PEP; transport from the periplasm to the cytosol via the phosphoenolpyruvate system.

UDP-Xylose UDP-Xylose is formed by enzymatic decarboxylation of UDP-GlcA, performed by UDP-glucuronate decarboxylase in both plants and animals (Kobayashi et al. 2002; Moriarity et al. 2002). This reaction is believed to occur in the cytosol, although there are reports of production of UDP-Xyl also in the ER (Kearns et al. 1993). It seems surprising that UDP-Xyl is formed from UDP-GlcA and not from UDP-Glc, since the reaction removes the C-6 anyway, and a decarboxylation of UDP-Glc would be an easier way to arrive at UDP-Xyl. Possibly, it is a way to control the production of both UDP-GlcA and UDP-Xyl at the same time. Plants have been found to recycle xylose by addition of UDP to Xyl-1-P (Kotake et al. 2004). UDP-Xyl donates Xyl to the biosynthesis of hemicellulose and pectin. In animals, the primary use of UDP-xylose is in the biosynthesis of the tetrasaccharide region, with which GAGs attach to their core protein, see below. UDP-Xylose is a powerful inhibitor of UGDH, in prokaryotes claimed to be a competitive (Schiller et al. 1973) and in eukaryotes to be an allosteric inhibitor (Ankel et al. 1966; De Luca et al. 1976) and has been claimed to have a regulatory role (Balduini et al. 1970). It also inhibits the activity of UDP-glucuronate decarboxylase and glucose pyrophosphorylase (Harper et al. 2002).

19 Gal Gal-1-P UDP-Gal

Glycolysis Fru

Photo- synthesis Fru-6-P Sucrose

Glc Glc-6-P Glc-1-P UDP-Glc Cellulose

Myoinositol Myoinositol-1-P

Rha Rha-1-P UDP-Rha Hemi- cellulose GlcA GlcA-1-P UDP-GlcA and pectic polymers Xyl-1-P UDP-Xyl

Ara Ara-1-P UDP-Ara

GalA GalA-1-P UDP-GalA

Figure 4c. Interconversions of monosaccharides into UDP-sugars relevant for cellulose- and hemicellulose-biosynthesis in plants. Note that paths of both de novo synthesis and of recycling from glycosides and sugar phosphates are present. (Feingold 1982; Kotake et al. 2004; Reiter et al. 2001)

Measuring nucleotide sugar levels It is difficult to measure the intracellular concentrations of UDP-sugars due to the low abundance and instability of most of them. Further, the concentrations vary quickly and depend on the energy status (i.e. ATP/ADP ratio) of the cell. Despite this, several groups have reported methods and measurements of intracellular nucleotide sugar levels. The methods for extraction have varied, as well as the methods of separa- tion. They have ranged from paper chromatography (Zhivkov et al. 1975) and thin layer chromatography (Feingold et al. 1990), to capillary electro- phoresis (Kneidinger et al. 2003), isotachophoresis (Eriksson et al. 1984) and HPLC chromatography, either in reverse-phase (Rabina et al. 2001) or by ion exchange (Handley et al. 1972a; Sweeney et al. 1993).

20 Detection of the nucleotide sugars has been performed mostly by spectro- photometry, measuring at 260 nm, but there are also reports where metabolic labelling has been used to detect nucleotide sugars. A few groups report have used enzymatic methods (Dörmann et al. 1998; Singh et al. 1980). The nucleotide sugars most interesting to the work presented in this thesis are UDP- GlcNAc and UDP-GlcA. Several studies have reported the UDP-GlcNAc concentration to be two to six times higher than the UDP-GlcA concentration in mammalian cells or cell culture (Handley et al. 1972b; Sweeney et al. 1993; Zhivkov et al. 1975).

CH2OH COOH O O O OH OH OH HO OH HO OH HO OH OH OH OH glucose (Glc) glucuronic acid (GlcA) xylose (Xyl)

CH OH 2 CH2OH O O O COOH HO OH OH O OH R R HO OH 1 OH NH OH NH

C O C O CH R-iduronic acid (IdoA)-R1 3 CH 3

N-acetylglucosamine (GlcNAc) N-acetylgalactosamine (GalNAc) Figure 5. Haworth representations of relevant monosaccharides (all a-D- configuration, except for a-L-iduronic acid). Iduronic acid is shown as in a GAG chain.

NAD+/NADH Nicotinamide dinucleotide (NAD) can be synthesised from niacin or trypto- phan. It is an electron-carrying coenzyme of many enzyme-catalysed redox reactions, partakes in respiration and is a substrate in some biosynthetic reactions. Here, only the role of NAD as coenzyme in the UGDH catalysed enzyme reaction will be further mentioned. Depending on the direction of the reaction NAD+ is reduced to NADH, to carry a hydride ion (i.e. a hydrogen and an electron pair) or oxidised from NADH to NAD+, having left the same entity (Fig. 1). The resulting changes in conformation of the benzenoid ring causes NADH to absorb light at 340 nm, whereas NAD+ does not, a characteristic utilised in spectrohotometric measurements. The ratio of

21 NADH/NAD+ in the cell is important for UGDH activity, since NADH is an inhibitor of UGDH, competitive with NAD+ (Schiller et al. 1973).

Sugar transport Sugar transport into the cell Animals derive carbohydrates from their diet and the most important are fructose, glucose, and galactose (Kutchai 1990). In humans, fructose is me- tabolised by the small intestine, liver, and kidney (Champe et al. 2005). Most animal cells take up glucose from the extracellular matrix by passive transport, also called facilitated diffusion. The transport from the blood into the cells is performed with a uniporter mechanism by one of the many sugar transporter proteins, a big family of transmembrane proteins. Once inside the cell, the monosaccharide is quickly phosphorylated by a hexokinase, thus preventing its transport out of the cell (Alberts et al. 1994). Galactose is taken up by the cells in an insulin-independent manner, and like glucose, phosphorylated upon entry (Champe et al. 2005). Prokaryotes use many different ways to take up sugars from their environment. Since nutrients generally have a lower concentration outside the bacteria the transport needs to be active, i.e. coupled to energy-consumption. This is the case in both active transport and in the PEP-PTS system. Active transport is carried out by symport proteins, which utilise the sodium gradient created by the electron transport. A sodium ion binds to the transport protein on the outside of the bacterial cell and changes the conformation of the protein. Subsequently, the nutrient binds the protein and both solutes are delivered to the inside of the cell. (Prescott et al. 1993). The PEP-PTS system is more complex. It is composed of several proteins, most of them enzymes. A cascade of phosphorylation starts with the transfer of a phosphate from phosphoenolpyruvate to the first enzyme in the complex. The phosphorylation signal is relayed throughout the complex and the final outcome is an imported and phosphorylated specific sugar or other nutrient. (Siebold et al. 2001) Plants however, produce their carbohydrates themselves. The photosyn- thesis takes place in the chloroplasts, subcellular organelles in the cells of the green plant-tissue. The carbohydrates are then either stored as starch or transported in the form of sucrose through the phloem (nutrient-conducting tissue of plants) to other parts of the plant. The sucrose seems to spread throughout the plant, down its concentration gradient, which is maintained by sucrose consumption at the receiving end. (Smith 1993)

22 Nucleotide sugar transport into the Golgi compartment Polysaccharide biosyntheses are initiated in the ER and completed in the Golgi apparatus. A majority of the nucleotide sugars utilised in these processes are synthesised in the cytosol, necessitating their transport into the organelles (exemplified by UDP-sugar in Fig. 6). The transport is attained by an antiporter mechanism with simultaneous transfer of a nucleotide sugar into the organelle and transfer of a corresponding nucleoside monophosphate out of the organelle.

GT Glycosylated structure or polysaccharide UDP-sugar UDP-sugar UDP AP UMP UMP

Pi ND

Pi Figure 6. Schematic view of the antiporter mechanism and the fate of nucleotide- sugars inside the mammalian and plant Golgi apparatus, as exemplified by UDP- sugar. UDP, uridinediphospho; UMP, uridinemonophosphate: AP, antiporter; GT, glycosyltransferase and ND, nucleoside diphosphatase. Ortophosphate, Pi, is proposed to be transported by a putative transporter. (Gibeaut 2000; Hirschberg et al. 1998)

The nucleotide sugar transporters are proteins of 6-10 transmembrane domains where both the N-and C-terminals face the cytosol and they are believed to function as homodimers (Berninsone et al. 2000). All of the transporters found to date have KM-values ranging in the low micromolar range and most are monospecific. However, three transporters originating from Leishmania donovani (pathogenic protozoan), C. elegans and human respectively, have been found to transport three different nuclotide sugars (Gerardy-Schahn et al. 2001). The transporters have been suggested to have a regulatory role in posttranslational modifications because of the possibility to control substrate concentrations in the organelles via the transporters. Once the nucleotide sugar is inside the organelle, the sugar moiety can be enzymatically transferred to a growing polysaccharide or other subject of glycosylation. The UDP-part of the nucleotide sugar forms the leaving group and facilitates transfer by preparing the anomeric carbon of the saccharide for a nucleophile attack, either from the enzyme or the acceptor, depending on the mechanism (Davies 2001; Unligil et al. 2000). After the sugar moiety

23 has been transferred to a target molecule by a glycosyltranferase, the released UDP is hydrolysed by a nucleoside diphosphatase to UMP and inorganic phosphate, thereby rendering the reaction irreversible. The uridine monophosphate exits the organelle via the nucleotide sugar transporter and the inorganic phosphate is believed to exit via a specific, still unidentified transporter (Fig. 6). All the plant cell wall polysaccharides, except cellulose, are believed to be synthesised in the Golgi apparatus. The polysaccharides destined for the cell wall are then transported in vesicles to the cellmembrane and incorpo- rated into the wall by specific enzymes. The proposal of UGDH having a rate-limiting role in plant cell wall polysaccharide biosynthesis is not agreed upon by all, some claim that the regulation is performed at the level of synthases (Bolwell 1993).

Mammalian glycosaminoglycans GAGs are heteropolysaccharides composed of a repeating disaccharide unit. This disaccharide is composed of a hexosamine and a hexuronic acid residue, except for in keratan sulphate, which will not be further mentioned in this thesis. Heparan sulphate (HS), chondroitin sulphate (CS) and dermatan sulphate are GAGs that appear to always be covalently bound to a protein, thus forming a proteoglycan. Hyaluronan (HA) on the other hand, does not form proteoglycans, but can act as a template and accomodate proteoglycans, bound via linker proteins. An overview of the topography of the GAG-biosynthesis discussed below is shown in (Fig. 7). There are many different kinds of core proteins and their expression pattern varies with different factors such as stage of embryonic development and in pathology (Iozzo 1998; Nybakken et al. 2002; Qiao et al. 2003).

Hyaluronan Hyaluronan, also called hyaluronic acid was first isolated in 1934. The name is derived from its original source of purification, the vitreous humour of the eye, in Greek hyaloid. It is a long polysaccharide, composed of between 2,000 to 25,000 saccharide residues. The repeated disaccharide is composed of alternating GlcA and GlcNAc (Fig. 5 and 8) and is not further modified, in contrast to the sulphated GAGs. HA is found mainly in connective tissue where it functions as a lubricant, regulates water balance and the penetrability of the extracellular matrix for cells. This last characteristic is especially important during embryogenesis (Spicer et al. 2004) and in the pathophysiology of cancer (Toole 2002).

24 Contrary to the other GAGs HA is not synthesised in the Golgi apparatus but at the plasma membrane (Fig. 7). HA of some species is, also in contrast to other GAGs, synthesised in the reducing end (Hoshi et al. 2004). Three mammalian hyaluronan synthases (Has) have been identified to date (Itano et al. 2002). Over the years, evidence has accumulated that HA can be found also intracellularly (Hascall et al. 2004).

UDP-Glc UGDH UDP-GlcA HS/CS- Hyaluronan synthesis synthesis NUCLEUS HS/CS GOLGI proteoglycan

Figure 7. Schematic view of UDP-GlcA utilisation in GAG-biosynthesis in a mammalian cell

Heparan sulphate/Heparin HS and heparin are molecular "siblings", heparin being a structurally more modified version of HS. Heparin was first discovered in 1916 and received its name in 1918 due to its assumed origin in liver; hepar is Greek for liver. After the discovery of heparin the inherent anticoagulant activity soon made it popular in medical treatment. The repeating disaccharide units in heparin and HS are composed of GlcNAc and GlcA (Figs. 5 and 8). Heparin is found in the granulae of mast cells in the form of the proteoglycan serglycin and functions as a "coat hanger" for the proteins in the granulae. HSs are found extracellulary and associated with virtually all mammalian cells. Both the heparin and HS backbones are synthesised by the EXT proteins. Thereafter a host of modifying enzymes change the fine- structure of the GAG by epimerisation, N- and O-sulfation (Fig. 8) (Sugahara et al. 2002). HS is implicated in growing numbers of cellular events, among them as an actor in growth factor signalling. From a patho- logical view, HS also has been found to play a role, e.g. in virus entry of cells (Liu et al. 2002) and in cancer (Sasisekharan et al. 2002).

25 H COR CH2OH COO 2 1 COO O O O O OH OH O OR OH O O 1 O

HNAc OH HNR2 OR1 GlcNAc GlcA GlcNR2 HexA K5 polysaccharide Heparin/Heparan sulphate

CH OR COO CH2OH COO 2 1 O O O O R1O

O OH O O OR O H O O 1

HNAc OH HNAc OR1 GlcNAc GlcA GalNAc HexA Hyaluronan Chondroitin/Dermatan sulphate Figure 8. Disaccharide structures of selected glycosaminoglycans. Heterogeneity in structure is indicated as: R1= H/SO3, R2= H/Ac/SO3 and arrow = possible epimerisation (Kresse 1997)

Chondroitin sulphate/ Dermatan sulphate The repeating disaccharide in CS is composed of GalNAc and GlcA (Fig. 5 and 8). A CS chain containing GlcA epimerised to iduronic acid is called dermatan sulphate. CS was first purified from cartilage in 1889, receiving its name from chondros, Greek for cartilage. CS polymerisation occurs on the same tetrasaccharide protein-linkage structure as the HS polymerisation but is performed by distinct enzymes (Silbert et al. 2002). Also the sulphation pattern of chondroitin sulphate dif- fers from that of HS, (see Fig. 8). The best established role of CS is that in cartilage, together with HA. Due to electrostatic interactions between the polysaccharide chains, they repel from each other. This interaction makes the cartilage able to withstand com- pressive forces. CS has been implicated in several biological processes e.g., in embryogenesis (Sugahara et al. 2003).

GAGs have been studied for very long but still many questions remain unanswered. What are the regulating factors determining whether a cell produces HS or CS? What determines the length of the GAGs? Are the diverse modifications seen in GAGs created by strict control or by chance?

26 Bacterial glycosaminoglycans The extracellular polysaccharides of bacteria fulfil several functions. They enable the bacteria to adhere to various surfaces, prevent the desiccation (Roberts 1995) of the bacteria and most importantly here: Disguise the bacteria so they are not discovered by the immune system of the host they have invaded (Roberts et al. 1989). This is accomplished by production of polysaccharides that resemble some of the polysaccharides on the host cell surfaces. Below, the bacterial K5 polysaccharide, K4 polysaccharide and HA will be discussed.

K5 polysaccharide K5 polysaccharide is so named because it belongs to the group of E. coli capsule (German: kapsel) polysaccharides that are produced inside the bacteria and subsequently transported out onto the surface of the bacteria. The structure of K5 polysaccharide is identical to the unmodified precursor backbone of HS (Vann et al. 1981) and it is believed that this similarity en- ables the E. coli K5 to evade the immune system of the host it has invaded. This type of bacteria mostly causes urinary tract infection and, in severe cases, meningitis. In wild type K5 bacteria the capsule is only produced when the bacteria are cultured in 37°C, making the connection between capsule role and infectivity even more apparent. Several gene products partake in the production of capsule and the genes are all gathered in a cluster divided into three regions (Roberts et al. 1988). The proteins of the genes from region one and three are common to all capsule gene clusters and are responsible for the export of polysaccharide to the outside of the bacterium. The proteins stemming from the genes of region two, however, are strain-specific and polymerise the polysaccharide that is specific for the strain. K5 region two is composed of four genes named kfiA, kfiB, kfiC and kfiD (kfi stands for K five). Their corresponding proteins are named the same way but not italicised: KfiA, KfiB, KfiC and KfiD (Petit et al. 1995). Kfi A and KfiC have been shown to have polymerising activity, but their precise relationship remains to be determined (Griffiths et al. 1998; Hodson et al. 2000). The function of KfiB is possibly one of stabilisation and its presence is vital to the production of polysaccharide. KfiD finally, is the E. coli version of UGDH, performing the last step in the production of one of the building blocks.

27 Other glycosaminoglycan-producing bacteria Pasteurella multocida Pasteurella multocida are gram-negative bacteria that also produce mammalian-mimicking GAGs. The bacteria mainly cause disease in domestic animals but also in humans. The strains are categorised by the Carter typing system, and the different strains, A, D and F, produce one type of GAG each (DeAngelis 2002). Type A produces HA by a dual-functioning synthase similarly with the mammalian HA synthases (Jing et al. 2003), but in opposition to the mammalian the P. multocida synthase performs polymerization at the nonreduc- ing end (Weigel 2002). Type D produces a polysaccharide identical to the K5 polysaccharide that sometimes goes by the name N-acetyl heparosan, due to its kinship with heparin/HS. Recently, the polymerase was cloned and shown to have dual-activity (DeAngelis et al. 2002). P. multocida Type F produces chondroitin, identical to non-sulphated CS. The polymerase has been cloned and seems to be bifunctional (DeAngelis et al. 2000) as all the above mentioned, with exception for the uncertainty of the E. coli K5 polymerase.

E. coli K4 K4 polysaccharide, produced by E. coli K4 is another nonsulphated bacterial analogue of CS. K4 polysaccharide has the same backbone structure as CS but carries an additional fructose bound covalently to every GlcA residue. Recently, the eight genes of region two of the K4 capsule cluster was cloned and shown to contain a gene kfoC, coding for a dual-activity polymerase (Ninomiya et al. 2002).

Streptococci The gram-positive Streptococci (group A and C) also produce HA by a dual- functioning synthase (Weigel 2002). The group A streptococci cause different diseases in humans whereas group C strictly infect non-human animals. The synthase of Streptococcus has been classified to belong to the Class I synthases, that polymerise the HA chain at its reducing end.

The GAGs are believed to be produced to improve the invasiveness of the bacteria, and a lack of capsule often renders them avirulent (Ashbaugh et al. 1998). However, there are studies reporting that mutants in capsule- production survived in hostile environment (Bahrani-Mougeot et al. 2002), showing that several factors interact to maintain virulence. A summary of the GAG-producing bacteria discussed here is shown in Table 2.

28 Table 2 GAGs produced by bacteria* Hyaluronan Streptococcus, group A and C Pasteurella multocida, type A

N-acetyl heparosan E. coli K5 Pasteurella multocida, type D

Chondroitin E. coli K4 (fructosylated) Pasteurella multocida, type F * table adapted from (DeAngelis 2002)

Plant polysaccharides In similarity with prokaryotes, plant cells have cell walls outside their plasma membrane. The plant cell wall confines the volume of the cell and in combination with internal water pressure achieves mechanical stability of the cell. Previously, the cell wall was considered to be an inert part of the plant cell, but accumulating evidence indicates that this organelle is active in signalling and in co-ordinating plant growth and development (Pilling et al. 2003; Reiter 1998). The cell wall consists of 90 % carbohydrates and the main portion of this is cellulose, an unbranched homopolymer of glucose (Heldt 1997). Most of the remainder of the cell wall carbohydrates is hemicellulose and pectin, collectively referred to as the non-cellulosic polysaccharides. Historically, hemicellulose was defined operationally, as that part of the cell wall polysaccharides that can be extracted by alkaline solution, and the name was chosen when hemicellulose still was thought to be a precursor of cellulose. Today, the definition is functional: the polysaccharides that form hydrogen- bonded complexes with cellulose are hemicellulosic (Zablackis et al. 1995). Hemicellulose consists of linear homopolymers with short side chains of varying composition. The other group of polymers, the pectins, are branched polysaccharides made up of sugar acids, mainly galacturonic acid. In the presence of Ca2+ and Mg2+ the pectic polymers are crosslinked with each other, via the free carboxyl groups present in the polysaccharides, and form a gel that holds the cells together. (Heldt 1997). The constitution and amount of the non-cellulosic polysaccharides vary between different species and the stage of development, and is made up of a multitude of sugar residues: glucose, mannose, galactose, glucuronic acid, galacturonic acid, and xylose, arabinose, apiose, rhamnose and fucose (Fry 1999). In the cell wall, the cellulose is gathered into parallel microfibrils that

29 are crosslinked by hemicellulose via hydrogen bonds, thus giving tensile strength (Carpita et al. 1993). The pectin in the cell wall ensures that the cell wall is not compressed due to the increased pressure. A growing plant cell has a primary cell wall. Once the cell has reached its final size the secondary cell wall is established, inside the primary, and prevents further growth. The compositions of these two cell walls differ from each other, the primary contains both cellulose and non-cellulosic polysaccharides, whereas the secondary cell wall contains mostly cellulose. In woody plants lignin is also part of the secondary cell wall and this compound gives mechanical strength also when the wood contains no water at all (Heldt 1997).

30 Present investigation

Aims The overall aim of the studies presented here was to study the role of UGDH activity in the biosynthesis of various GlcA containing polysaccharides. In order to approach the overall aim the more specific aims were as follows: x Clone and/or establish overexpression of UGDH, native to each system respectively, in human cell culture, E. coli and A. thaliana. x Investigate effects of UGDH-overexpression on the UDP-sugar levels in these systems. x Investigate effects of UGDH-overexpression on polysaccharide biosynthesis in the respective system.

Results Paper I cDNA cloning and expression of UDP-glucose dehydrogenase from bovine kidney

In order to simplify continued research on mammalian UGDH it was necessary to isolate a cDNA coding for the enzyme. To this end, a previously published protein sequence (Hempel et al. 1994) was used to screen a human EST database. A matching sequence was found. The sequence was ordered from NCBI and subsequently used as a probe to screen a bovine cDNA library and to isolate a cDNA clone covering the coding part of the bovine UGDH. The isolated cDNA sequence differed from the previously published bovine sequence in that it contained an extra internal leucine and an additional C-terminal tail of 25 amino acids. The full-length sequence was excised from its Ogt10 vector with EcoRI restriction enzymes. Due to the presence of an internal EcoRI-restriction site, the EcoRI cleavage generated two fragments, a shorter sequence coding for the C-terminal tail of the protein and a longer 5´end fragment, containing most of the coding region. Both the 5´ end fragment, and the full length form of the bovine UGDH sequence, were ligated into pcDNA3-(His/Flag) plas-

31 mid vectors, thus creating expression constructs fused to an N-terminal pep- tide containing the Flag epitope and the polyhistidine tag. COS-7 cells were transfected with the constructs by electroporation. The expression of the fusion protein was verified by western blotting, using a monoclonal antibody against the Flag-epitope. A new, sensitive method for analysis of UGDH-activity was developed. The method was as follows: UDP-[14C]Glc was incubated with cell lysate and coenzyme. The resulting product, UDP-[14C]GlcA was separated from the non-utilised substrate by anion-exchange chromatography and quantified by scintillation counting. This method proved to be very useful also in sub- sequent work described below. The UGDH-activity of the transfectants were compared with the activity of mock-transfected cells and found to have increased two- to threefold. A 295 bp cDNA probe derived from the full-length sequence was used to analyse UGDH mRNA expression levels in human and mouse tissues, using commercially available multiple tissue Northern blots. The mouse blot revealed an especially strong expression in liver tissue. This probably reflects a greater activity in glucuronosylation of xenobiotics in mouse liver. The human blot surprisingly showed two hybridising bands, throughout the tiss- sues. One band appeared at 2.6 kb and the other at 3.2 kb. The reason for the occurrence of two bands is unknown. x The bovine cDNA for UGDH was cloned and expressed in mammalian cell culture. x A new method for analysis of UGDH-activity was developed and used for final verification of the isolated bovine cDNA sequence. x Northern blot analyses performed on mouse tissues showed especially high contents of UGDH-mRNA in liver. All human tissues tested in Northern analyses expressed two mRNA-transcripts.

Paper II Overexpression of UDP-glucose dehydrogenase in Escherichia coli results in decreased biosynthesis of K5 polysaccharide

The bacterial K5 polysaccharide is identical with the sugar backbone of the precursor polysaccharide in eukaryote heparin/HS biosynthesis (for details on the E. coli K5 polysaccharide, see the Introduction). Moreover, the UGDH enzyme is highly conserved throughout the species. Therefore we decided to use the biosynthesis of K5 polysaccharide as a model system to investigate the role of UGDH in GAG-biosynthesis. A engineered version of E. coli K5, constructed by transformation of E. coli JM 109 (DE3) (Petit et al. 1995), was used in this study. The bacteria were transformed with two plasmid-constructs. One plasmid contained the 32 complete gene cluster necessary for synthesis and export of K5 polysaccharide, including a copy of kfiD (UGDH). The other plasmid was either the vector alone (control strain) or vector containing an extra, IPTG-inducible kfiD-copy (overexpressing strain). Transformants were maintained with an- tibiotic selection for the two plasmids. The transformants were tested for UGDH-activity with the method developed in paper I. The in vitro activity of the overexpressing strain was increased a15-fold over the control strain. A method was developed to measure UDP-GlcA content of the bacteria. Bacteria were metabolically labelled with [14C]Glc, harvested and lysed. The UDP-GlcA was purified from the lysate in several steps, and the difference in content of radioactive UDP-GlcA compared between the strains. The overexpressing strain contained three times more UDP-GlcA than the control strain. Attempts to measure UDP-GlcNAc were not successful (not mentioned in the paper). A method to measure K5-polysaccharide biosynthesis was developed, again using metabolic labelling with [14C]Glc. Bacterial cultures were incubated with IPTG and [14C]Glc. Cultures were harvested and lysed and the K5 polysaccharide purified through steps of anion-exchange chromatography and gel chromatography. Identity of the polysaccharide was ascertained by testing its susceptibility to degradation with a specific K5-phage lyase. The amount of K5 polysaccharide was quantified by scintillation counting. The overexpressing strain unexpectedly produced three times less K5 polysaccharide than the control. The finding was further supported by colorimetric and enzymatic measurements in a larger scale version of the same K5 polysaccharide purification protocol. The carbazole assay colorimetrically measures the glucuronic acid content of the polysaccharide (Bitter et al. 1962). The enzymatic quantitation spectrophotometrically measures the amount of unsaturated uronic acid, generated as a result of lyase cleavage of saccharides (Fig. 9). Heparitinase I cleavage of K5 polysaccharide resulted in degradation of the polysaccharide to disaccharides, proportional to the amount of polysaccharide, and the amount of unsaturated uronic acid residue was quantified at 232 nm. The size of the radiolabelled K5 polysaccharide was analysed by gel chromatography. No significant differences were found between the elution patterns of the K5-polysaccharide from overexpressing and control strains. These findings suggest that the decreased amount of K5 polysaccharide in the overexpressing strain was due to fewer, rather than shorter polysaccharide chains. x The kfiD (UGDH) was overexpressed in E. coli. Overexpression was verified by enzymatic measurement.

33 x A method for measuring UDP-GlcA content in E. coli was established. UDP-GlcA content was 3-fold higher in the overexpressor than in the control. x A protocol for radiolabelling and purifying K5 polysaccharide in bacteria was established. The overexpressor produced three times less K5 polysaccharide than the control in both radioactive and non-radioactive methods of measurement. x K5 polysaccharide from overexpressor and control was compared by gel chromatography and concluded to be of similar size. This indicates that the overexpressor probably makes fewer polysaccharide chains than the control.

CH2OH COO A O O R 1 OH OH O O R2

HNAc OH R1-GlcNAc GlcA-R2

CH2OH COO B O O R1 OH OH OH O R2

HNAc OH α/β R1- -GlcNAc 4,5-GlcA-R2 Figure 9. Schematic view of the action of lyases as exemplified by the K5 phage lyase, cleaving K5 polysaccharide. A, before cleavage; B, after cleavage; R1 and R2 denote continuation of polysaccharide (Hanfling et al. 1996; Yamagata et al. 1968).

Paper III (Manuscript) Glycosaminoglycan biosynthesis in UDP-glucose dehydrogenase overexpressing 293 cells

The full-length cDNA UGDH-sequence described in paper I was used to establish stabile UGDH expressing clones in human embryonic kidney cells (HEK) 293 cells. The method for UGDH-activity assessment used in paper I was again utilised. The highest expressing clone showed a 7-fold increase in activity. This clone was transfected with hyaluronan synthase 3 (Has3) to establish UGDH/Has3 double-overexpressing clones. Clones stably express-

34 ing Has3 alone and control clones containing the two original pcDNA3 vectors together completed the experimental set-up. The Has3 alone and the Has3/UGDH double overexpressors showed comparable increases in Has-activity. The UGDH-overexpression level was maintained after introduction of Has3 in the double-overexpressor. Northern blot analyses, with probes for the two mRNAs, were performed on two clones of each transfectant. Substantial increase in the UGDH and Has3 mRNA levels were observed in the respective transfectants, whereas the corresponding endogenous expression levels were undetected. The four differently transfected clones were used to investigate the effects of the respective overexpression on GAG biosyntheses. What effect would an increase in the supply of one of the building blocks have? Would the effect be replicated by an increase in the demand of building blocks? Would a combination of these changes show additional effects? The effects were in- vestigated by metabolic labelling of cell cultures, using [3H]GlcN. Cells were cultured in complete cell culture medium containing [3H]GlcN for 24 h. The labelled proteoglycans and HA were purified by standard protocols and identified by specific enzymatic cleavage, using hyaluronidase, chondroitinase ABC and heparin lyases (Table 3). The total amounts of GAGs had increased in all the overexpressors as compared to the mock-transfectant. This GAG increase ranged from 3-fold in the UGDH-overexpressor to 6-fold in the UGDH/Has3 double-overexpressor. The overexpressor that expressed Has3 alone increased the total GAG-amounts 5-fold, indicating that UGDH activity is not rate limiting in GAG-biosynthesis. The relative proportions of the GAG-species from the overexpressors were compared with those of the mock transfectant. Increased UGDH activity did not significantly influence the relative proportions of the different GAG-species. Furthermore, although, due to increased synthesis of HA, the relative proportions of HS and CS were greatly reduced, in the transfectants overexpressing Has3 (alone or with UGDH) the total amounts of HS and CS were not reduced. In fact, they were approximately unchanged. The change in relative GAG proportions, caused by the overexpression of Has3, remained also in the double- overexpressor; i.e. the UGDH overexpression displayed no "correcting" capacity. x Four variants of stably overexpressing clones were generated in 293 cells: singly UGDH- or Has3- overexpressing, double UGDH- and Has3- overexpressing and double mock-expressing. Overexpression was verified by Northern analyses and enzymatic measurements. x GAGs were purified from metabolically labelled cell cultures: 1. Total GAG-amounts increased in all overexpressing clones. 2. Overexpression of UGDH caused no changes in ratios between different GAGs. The change in GAG-ratios caused by Has3-overexpression

35 could not be corrected for by the UGDH-overexpression in the double overexpressors.

Table 3. Selected endolytic, GAG-specific enzymes Enzyme and Source Specificity Products K5 phage lyase, K5 polysaccharide/ hexa-, octa-, deca- (Coliphage K5A) a N-Acetylheparosan saccharides

Hyaluronidase, Hyaluronan tetra- and hexa- (Streptomyces saccharides Hyalurolyticus) b

Chondroitinase ABC, Chondroitin of all sulfation disaccharides (Proteus vulgaris) c degrees and hyaluronan

Heparinase I*, Heparin and highly di- and oligo- (Flavobacterium sulphated areas of saccharides Heparinum) d heparan sulphate

Heparinase II Less sulphated areas in as above /Heparitinase II, heparan sulphate as above

Heparinase III Low sulphated areas of as above /Heparitinase I, heparan sulphate and as above N-Acetyl heparosan a(Hanfling et al. 1996), (Gupta et al. 1982), b (Ohya et al. 1970), c (Yamagata et al. 1968), d(Yamada et al. 1998) *heparinase I = heparin lyase I etc.

Paper IV (Manuscript) Increased expression of a UDP-glucose dehydrogenase gene in Arabidopsis thaliana results in altered cell wall composition and dwarfism

During the search for UGDH sequences in various databases performed in paper I several Arabidopsis thaliana (mouse-ear cress) sequences also turned up. An EST containing the complete cDNA for a sequence later named UGD1 was ordered. UGD1 was used to produce a promoter-reporter fusion construct and a 35S-UGD1 construct (the latter a construct that induces con- stitutive expression of UGD1). A. thaliana plants were transformed with the constructs to analyse the enzyme expression pattern, and to create overexpressing plants, respectively.

36 The expression analysis showed that UGD1 was strongly and specifically expressed in the elongation zones of primary root and lateral roots. Some staining was also found in the flower organs. A purification protocol was developed in order to assay the enzyme activities of wild type and plants transformed with the 35S-construct. The spectrophotometric assay verified that overexpression had been established. The phenotype of the transformants was severely disturbed. The aerial parts of the transformants were shorter and the primary root longer than those of the wild type. Additionally, there were more secondary inflorescences in the transformants. Seeds of the transformants displayed a raisin-like outside, as compared to the smooth surface of the wild type. The composition of monosaccharide residues in the cell walls of rosette leaves was analysed with an established method. The differences were small but distinct. Most prominent was that the content of mannose and galactose had increased, whereas the arabinose, xylose and uronic acid contents had decreased. The data was analysed using the mathematical method principal component analysis, visually illustrating the differences. x A reporter gene construct showed that UGD1, one of the four A. thaliana UGDHs, is expressed mainly in the elongation zone of primary and lateral roots. x UGD1 was overexpressed in A. thaliana resulting in dwarfism and other changes in appearance. x A method for purification of UGDH activity in A. thaliana was established and overexpression verified by enzymatic measurements. x The composition of the cell wall polysaccharides showed small but distinct changes in the content of several saccharide residues.

Discussion The experimental approach of the studies presented in this thesis has been to study the effects of UGDH overexpression on polysaccharide biosynthesis in different systems. Several questions have been answered. However some of the original questions still wait for answers and many more questions have joined them.

UGDH gene expression Northern analysis of human tissues (paper I) revealed two transcripts; this has been reported in other studies as well (Lapointe et al. 1999; Peng et al. 1998; Spicer et al. 1998). The human UGDH seems to be encoded by a single gene (Bontemps et al. 2000). Two transcripts could indicate alternative splicing or alternative polyadenylation sites. There have been proposals for 37 tissue-specific UGDHs (Bardoni et al. 1989). If this is the case, and the UGDHs of different tissues differ already on the mRNA-level, there would most likely be only one signal in each tissue, but of different sizes in different tissues. The expression analysis in paper IV showed that the expression pattern of UGD1 is most distinct in lateral root tips. The result is comparable to that of Seitz et al (Seitz et al. 2000). The fact that four different UGDs have been found in EST-libraries establishes that they are transcribed and probably active in the plant. What use does the plant make of four genes of the same kind? It would be very interesting to see the expression patterns of all four genes. Are they expressed at different times of the plant development or are they expressed in different tissues? The sequence similarity of the genes is so high, 80-92 %, that it is quite probable that they have similar kinetics. With the establishment of the bovine UGDH cDNA sequence (paper I) further comparison of UGDHs from different species was possible. On comparison a high degree of sequence similarity was found (Fig. 2). The obvious use of a cDNA is in further in vivo and in vitro experiments, as described in paper III.

UDP-sugar levels Since the studies in this thesis are parallel to each other, in the sense that they have the same approach, they mostly have the same strengths and weaknesses. One of the weaknesses is the difficulty to measure UDP-sugars levels (see Introduction). Attempts to set up the method of Mason (Sweeney et al. 1993; Ysart et al. 1994) in the lab have not been successful, due to major obstacles in purification procedures and instability of the compounds. Metabolic labelling made it possible to obtain a relative measurement of UDP-GlcA content in bacteria (paper II). The fact that UGDH overexpression in cell culture can increase synthesis of all three GAGs does not necessarily mean that the concentration of UDP- GlcA is changed; the flux through the pool may be increased. There have been reports presenting evidence for both changed as well as unchanged UDP-sugar pools in relation to increased GAG-synthesis (Sweeney et al. 1993; Ysart et al. 1994). The lack of data on metabolite concentrations pre- cludes conclusions on the possible regulatory role of UGDH in polysaccharide biosynthesis. Due to the difficulties in measuring UDP-sugar levels, instead measurements of cell wall polysaccharide residues were made in Arabidopsis. The analyses showed a decrease in some of the saccharide residues. The saccharides that most distinctly decreased are all downstream products of UDP- GlcA (Fig. 4c). Whether this is a result of a decrease in UDP-GlcA alone or UDP-GlcA and related nucleotide sugars altogether, or a perturbation of the transferases using them is uncertain. This can only be settled by actual 38 measurement of the nucleotide sugars. Recently, another study was published, reporting alterations of cell wall polysaccharides as a result of UGDH expression in alfalfa (Samac et al. 2004).

GAG biosynthesis Amounts of GAG In bacteria, the increase in UGDH-amounts led to an increase in UDP-GlcA content (paper II). This is what would be expected. Less expected was the finding that the K5 polysaccharide decreased in amounts. There can be several reasons for this. Firstly, the assembly at the inner membrane of proteins involved in the biosynthesis could be disturbed by the overexpression of one of the enzymes. As mentioned in the Introduction, UGDH has been reported to be membrane associated in bacteria and fungi. Secondly, the increase in UDP-GlcA levels may have caused a decrease in UDP-GlcNAc, or at least a switch in the ratio between the nucleotide sugars. Any of the changes may disturb initiation and growth rate of the nascent chains. The bacterial system was intended to be a model system for the more complex, expensive and time-consuming mammalian cell culture system. Unexpectedly, the results of the two systems are contradictory: In bacteria, UGDH-overexpression caused a reduction in K5 polysaccharide amounts, whereas the corresponding overexpression in mammalian cells resulted in increased amounts of GAGs. The reason for this difference can lie both within the UGDH-enzymes themselves and within the control of the metabolism in the two systems. Possibly, part of the question could be answered by an ectopic expression of the two enzymes. If the bacterial enzyme could be expressed and active in mammalian cell culture and vice versa, clues could be found as to whether the difference lies in the enzymes or the systems.

Proportions of GAG-species It is known that the ratios and fine structure of different GAG-species in mammals vary between different cell types and with the age of the individ- ual (Feyzi et al. 1998; Ledin et al. 2004; Lindahl et al. 1998). But it is not known how these changes are achieved in mammalian cells and what causes the changes. The cell culture experiments (paper III) suggest that UGDH is not the point of regulation. An overexpression of UGDH results in production of more GAGs, but without any changes in the proportions of the GAG- species. Further experiments are needed to be able to confidently state how or if an increased need of building blocks (as in the Has overexpressor) af- fects the ratio between HS/CS. It seems however as if the overexpression of UGDH (in the UGDH/Has3 double overexpressor) is not correcting the aber- rations in GAG ratio caused by Has overexpression.

39 The mammalian cellular concentrations of UDP-GlcA, UDP-GlcNAc and UDP-GalNAc, are estimated to 10-400 µM (Gainey et al. 1972; Handley et al. 1972c). The antiporter system that transports nucleotide sugars into the Golgi apparatus has been shown to have KM-values in the low PM-range (Gerardy-Schahn et al. 2001). This fact may place the Golgi-situated GAG- synthases in the same position as the Has-synthase, in the competition for UDP-sugars. However, the antiporters that import UDP-sugars all use UMP for con- comitant export (Fig. 6). If the activity of the antiporter system is ruled by the UMP-concentration in the Golgi, an increase in cytosolic UDP-GlcA would not change the amount of UDP-sugars imported into the Golgi but probably only change the proportions of imported UDP-sugar species. The finding that overexpression of UGDH increases CS, HS and HA synthesis to the same extent (paper III) speaks against this theory. Furthermore, a recently published study reports that overexpression of one of the transporters, hfrc1, causes an increase in synthesis of HS (Suda et al. 2004). Hfrc1 is proposed to transport UDP-Glc and UDP-GlcNAc. Accordingly, they observed a change in HS, but not in CS, which utilises neither of these as building blocks. The finding of Suda et al speaks against UMP-concentration as a limiting factor, but supports the notion of the Golgi nucleotide-sugar transporters being specifically rate limiting for different GAG-syntheses.

Further experiments The results from the radio-labelling experiments in paper III would benefit from studies using unlabelled material, for the same reason as in paper II, i.e. to ascertain that the changes seen are true differences and not due to changes in specific activity of the sugar pool. Exchanging the scintillation counting for a method that measures total unlabeled cellular GAGs, e.g. a post- column fluorescence detection technique (Ledin et al. 2004) would ensure that no fluctuations in radioactive specific activity has occurred and influ- enced the interpretation of the data. Further, the study could be expanded with experiments using xylosides. Xylosides can substitute for the core protein upon which all sulphated GAGs are polymerised, thus releasing the GAGs from the possible restraining factor of the availability of the core protein. In that situation HS/CS may com- pete on more equal terms with HA for the nucleotide sugars.

Other aspects of UGDH

UGDH in pathology To date, no human diseases have been directly linked to UGDH (Marcu et al. 1999). However, UGDH knock-out experiments in Drosophila causes lethal- ity at a very early stage (Hacker et al. 1997) and mutations in the UGDH of 40 Cryptococcus cause decreased viability (Griffith et al. 2004). This suggests that absence of this gene, in animals, is incompatible with life. In humans, UGDH has been studied in the context of rheumatoid arthritis. The viscosity of the synovial fluid in arthritic joints is decreased, possibly due to a combination of increased degradation, and decreased production of HA in the joints. On the other hand the serum-levels of HA are increased (Laurent et al. 1992). Studies trying to establish whether UGDH has a role in this scenario, report that the UGDH activity in the synovial cells is decreased (Pitsillides et al. 1993; Uzuki et al. 1999). Nevertheless, firm conclusions are still lacking.

UGDH in biotechnology Attempts have been made to create bacteria or yeast capable of producing mammalian GAGs for pharmaceutical purposes. Lately, synthetic compounds have reached the market (Petitou et al. 2004). Microbiologists have suggested that an alternative to antibiotics could be created for UGDH-containing bacteria. The aim would be to abolish the adherent capacity of the bacteria by targeting the UGDH activity (Campbell et al. 2000). It is difficult however, to perceive how this would work, taking into account that the bacteria have invaded a host with a UGDH enzyme that is very similar to the bacterial. Again, a further knowledge of the mechanism becomes interesting. If the mechanisms of bacterial and mammalian UGDHs are different, target-specific treatment can be generated. Knowledge of the enzymes in nucleotide sugar metabolism and polysaccharide biosynthesis is interesting, not only to the pharmaceutical industry, but also to the paper industry. Production of paper is mainly dependent on cellulose, but the polysaccharides of hemicellulose are of growing interest. The hemicellulosic polysaccharides can be used to create different kinds of paper surfaces. Therefore, paper producers are also interested in knowing how to manipulate the polysaccharide composition in wood.

Naturally, generating synthetic compounds requires a good deal of knowledge of both chemistry and biosynthesis. The studies presented here add a few pieces to the knowledge of polysaccharide biosynthesis. Still, the most immediate use of the results presented here will be in basic research, a con- tinuous process constantly in need of new methods and more information.

41 Future perspectives

A fundamental principle in experimental science is to simplify as much as possible in order to get unambiguous answers. Hopefully, the answers will still be applicable to the complicated reality. But the interactions in a living cell are so many and so intricate that sometimes the simplified picture becomes fallacious. Therefore theories are always compared with experimental data and when necessary adjusted. Maybe such an adjustment is necessary for the simple concepts of regulatory and rate limiting? There are advocates for a more mathematical approach to understanding and predicting metabolism (Fell 1997). Possibly, in the future, this will replace the traditional, ver- bal theories of basic biochemistry. The last century experienced a surge in biochemical knowledge and prac- tice. In much, this was due to the advent of computers and the discovery of DNA. Although the area of glycobiology also has benefited greatly from these break-throughs, it still suffers from lack of fast, simple and specific methods. When will the laboratories of the world use an "Eastern blot" for carbohydrates as naturally as they use the Western blot for proteins? The last twenty years many scientists of other fields have been directed by their science into glycobiology. The resulting cross-boundary-mix of schooling and experiences will surely be beneficial for the field; which may very well have its greatest times ahead!

42 Acknowledgements

Scientific work is not done single-handedly and I am grateful to many per- sons. First, my supervisors. I am very grateful for all efforts, both personally and scientifically, that you have put into my education. Kerstin Lidholt, who started out with me when I was “brand new”. I will always remember your enthusiasm. Marion Kusche-Gullberg, my main-supervisor, who helped me in many ways to the very end. I do admire your endurance! Thank you especially, for teaching me how to write scientific texts (no jokes, no “flowers”). Ulf Lindahl, my assisting supervisor, who was always ready to help if problems arose and to help point out weak spots when my critical thinking was leaky. Thank you all for your encouragement along the way!

I am also grateful to all present and former members of the “Heparin- group”, for good company, help and advice in both practical, theoretical and ethical aspects of science. Especially, I want to thank Eva Gottfridsson and Eva Hjertson for being eager to add spice to the life of the lab in the form of outings, parties and laughter. What would we do without you?

I also want to thank my colleagues at the whole of IMBIM for friendship, pleasant lunch-company, and for willingly sharing your knowledge with me, when I had found yet another question. Special thanks to Örjan Zetterqvist, who was always willing to share his time, and who helped me improve my teaching and my understanding of biochemistry, and to Åke Engström who led me through the labyrinth of HPLC-repair. This department has a skilled secratariat in Barbro Lowisin, Erica En- ström, Inga-Lisa Karlsson-Lööf, Liisa Holmsäter and Maud Pettersson and also an outstanding technical/software staff in Kerstin Lidholt, Olav Nordli, Piotr Wlad, Ted Persson and Tony Grundin. I lost track of in how many ways you have helped me through the years, but I always appre- ciated it very much! I am also indebted to the GLIBS foundation for funding part of my education and for arranging courses and national glycobiology meetings. They have helped widen and deepen my insight in the field.

Gratitude is due also to our collaborators: Ian Roberts, Evi Heldin, Jonas Brinck, Peter Engström, Henrik Johannesson, Roger Andersson and Per Åman.

43 Special thanks to Gunilla and Margareta who helped me during the last year. Thanks also, to my former fellow-students in biomedicine who kept in touch with lunch-dates at “Gröne Orm”. I was always late, although I tried not to be! Additional friends, outside the lab, and relatives in blood or in faith, especially Charlotta Nilsson: thank you all for distracting me from science ever so often, and giving me new strength to carry on the work. Our friendship is very valuable to me! Thanks also, to my dear parents, Jill and Per Falk, for their love and support, and to my very special favourite, my husband Magnus Roman, a seemingly never-ending source of goodness. Lastly, thanks to Jesus Christ, who gave me life, and a mind to investigate part of it with.

44 Populärvetenskaplig sammanfattning på svenska (Summary in Swedish)

Studier av UDP-glukosdehydrogenasets roll i polysackaridbiosyntes Människokroppen består av många organ, i sin tur uppbyggda av celler med olika uppgifter och utformning. Men cellernas grundprincip är densamma. De har ett cellmembran som omsluter cellens övriga beståndsdelar, ungefär så som huden omsluter kroppen. Cellen är fylld av vätska, som kallas cytosol. I cytosolen finns det organeller som, precis som kroppens organ, har olika funktioner. Mitokondrierna utvinner energi ur molekyler från maten vi äter, endoplasmatiska retiklet är en tillverkningsplats för proteiner och fetter, Golgiapparaten är tillverkningsplats för olika slags sockermolekyler och i lysosomerna sker nedbrytning av gamla molekyler. I stort sett alla molekylära processer som sker i kroppen utförs av enzymer. Nästan alla enzymer är proteiner och kan liknas vid maskiner som utför en specifik uppgift. Beroende på sin uppgift befinner de sig i någon av de organeller som beskrivits ovan, eller i cytosolen. Cellen tillverkar själv de enzymer den behöver efter "ritningar", gener, som i de allra flesta fall finns i cellkärnan. Ute på cellmembranet eller i det nätverk som omger cellerna, extracellulä- ra matrix, finns polysackarider. De är långa kedjor av sockermolekyler. De sitter fast i cellmembranet via en ankarmolekyl, eller flyter omkring i extra- cellulära matrix. Alternativt kan de också sitta fästade vid ett protein som, på liknande vis, antingen är förankrat i cellmembranet eller flyter omkring i det extracellulära matrix. Även polysackariderna tillverkas av cellen själv och den processen kallas biosyntes, i motsats till kemisk syntes. När enzymerna bygger polysackarider aktiveras en hel maskinpark. En del enzymer förbereder enskilda sockermolekyler, monosackarider, så att de kan inkorporeras i en längre kedja. Förberedelsen innebär att monosackariden kopplas ihop med en energirik molekyl, tex UDP. UDP-delen avskiljs sedan när monosackariden sammanfogas med den växande polysackariden och på så sätt "betalas" den energi som krävs för att sammanfoga monosackariden med resten av kedjan. Särskilda enzymer utför själva sammanfogan- det av monosackarider och återigen andra kan modifiera polysackariden på olika sätt. Föremålet för studierna som presenteras i denna avhandling är ett enzym som heter UDP-glukos dehydrogenas (UGDH). UGDH förändrar monosack-

45 ariden glukos så att den blir en glukuronsyra, medan den fortfarande är bun- den till UDP, dvs UDP-glukos blir UDP-glukuronsyra. UGDHs placering längs en tänkt tillverkningslinje är efter det enzym som kopplar ihop glukos med UDP-del och före det enzym som sammanfogar monosackarider till en längre kedja. För att en hel grupp av enzymer ska kunna samarbeta i produktion av nå- gonting behövs signalsystem som kan koordinera enzymerna. Denna koordi- nering (reglering) säkerställer att de molekyler som cellen behöver produce- ras vid rätt tidpunkt och plats och att molekylerna får det utseende som be- hövs just då. Detta sker ständigt, i alla celler, på många nivåer och på olika sätt. Syftet med studierna i denna avhandling var att få veta mer om UGDHs roll i produktionen av vissa polysackarider, dvs att hitta några fler pusselbitar som kunde ge information om hur polysackaridbiosyntesen styrs i våra celler. En djupare kunskap om UGDH är värdefull både inom fortsatt grund- forskning och i klinisk (=tillämpad, medicinsk) forskning. En del sjukdomar är kopplade till polysackarider eller störningar i polysackaridbiosyntesen och då är det värdefullt att veta vilket/vilka enzym man ska angripa för att påver- ka situationen. Vidare finns det läkemedel som är polysackarider och i dags- läget utvinns några av dem fortfarande från slaktdjur. Med en bättre känne- dom om biosyntesen kanske det blir möjligt att tillverka dessa i tex bakterier. UGDH är ett enzym som finns i stort sett i alla livsformer, alltifrån virus till bakterier, svampar, växter och djur. Genen är i själva verket en kod som talar om ordningen för de aminosyror som tillsammans utgör protei- net/enzymet UGDH. Trots att generna kommer från så olika källor är de väldigt lika varandra. På grund av denna likhet bestämde vi oss för att under- söka UGDH i både bakterier, växter och djurceller. Efter att vi identifierat djur- (Artikel I) och växtgenen (Artikel IV) var tillvägagångssättet detsamma i de tre systemen. Vi överuttryckte UGDH, dvs behandlade bakterier, växter och djurceller så att de hade fler UGDH enzymer än normalt. Sedan under- sökte vi hur överuttrycket hade påverkat polysackaridbiosyntesen.

Artikel I beskriver hur vi gick till väga för att identifiera cDNA:t (kopia av genen) för UGDH från nötkreatur. Slutligen fördes cDNA:t in i celler i cell- kultur och enzymaktiviteten jämfördes mellan celler som fått UGDH och celler som inte fått det. Enzymmätningen gjordes med en ny metod som utarbetats i vårt laboratorium och visade att cDNA:t verkligen kodade för UGDH.

Arbetet i bakterier (Artikel II) gjordes i samarbete med en brittisk grupp som redan hade identifierat och överuttryckt ett bakteriellt UGDH i bakterier. Vi undersökte dessa bakterier och fann att överuttrycket resulterade i mer enzym och mer produkt (UDP-GlcA) men att mängden polysackarid blev läg- 46 re. Mängden polysackarid mättes på flera sätt, vilket styrker riktigheten i resultaten. Dessutom jämfördes längden på polysackaridkedjorna mellan bakterier med UGDH-överuttryck och kontroll-bakterier. Vi upptäckte ingen skillnad i längd och drog därför slutsatsen att minskad mängd polysackarid i överuttryckaren berodde på färre kedjor.

Arbetet i artikel III utfördes i samarbete med en annan grupp i Uppsala och är ännu inte publicerat. Där beskrivs samma sorts försök som i artikel II, men utförda på celler från människa. I denna studie överyttrycktes inte bara UGDH, utan även ett annat enzym, Has3. Has3 producerar hyaluronan, som är en av de polysackarider som innehåller glukuronsyra (Fig. 7). På så sätt kunde vi undersöka vad som händer när man ökar aktiviteten av antingen "producent" (UGDH), "konsument" (Has3) eller båda tillsammans. Även här ökade enzymmängderna av respektive enzym vid överuttryck, men också deras produkter. Försök gjordes att mäta UDP-sockerhalterna, men har hit- tills inte lyckats. När UGDH överuttrycktes ökade mängen av alla tre polysackarider som undersöktes, hyaluronan, heparan sulfat och kondroitin sulfat; och proportionerna mellan dem verkade inte förändras nämnvärt. När Has3 överut- trycktes ökade förstås mängden hyaluronan, men utan att de andra två mins- kade jämfört med de ursprungliga mängderna. När båda enzymerna överut- trycktes tillsammans kvarstod den förändring som Has orsakat, med det tillägget att alla polysackariderna ökade i mängd. Sammanfattningsvis verkar det som att UGDH-aktiviteten påverkar mängden, men inte proportionerna, av de polysackarider som innehåller glukuronsyra.

Arbetet med växter (Artikel IV) utfördes i samarbete med två Uppsala- grupper; även här återstår publicering. I växter finns fyra genvarianter av UGDH. En av dem, UGD1, identifierades i samband med sökningen efter gener i artikel I. Den fördes in i växter tillsammans med en s.k reportergen. Reportergenen visar på vilka ställen i växten genen är aktiv, dvs används för att göra protein. I det här fallet visade sig växten främst använda UGD1 i den första, vertikala roten, samt i spetsen på de rötter som skickas ut horisontellt från den. Överuttryck av UGD1 orsakade dvärgväxt av den del av plantorna som var ovan jord, medan rötterna blev längre än normalt. Möjligen beror detta på att UGD1 uttrycktes på ställen där den normalt inte är aktiv. Istället för mätning av UDP-sockernivåer analyserades sackaridsammansättningen i cellväggs-polysackariderna som visade små, men distinkta förändringar.

Detta är, i korthet, de pusselbitar som framkommit i arbetet som presenteras i den här avhandlingen. Förhoppningsvis kommer de att vara andra forskare till hjälp i den fortsatta kartläggningen av de komplexa samband som råder i biosyntesen av polysackarider, både när vi är sjuka och friska. 47 References

Abbracchio, M. P., J. M. Boeynaems, E. A. Barnard, J. L. Boyer, C. Ken- nedy, M. T. Miras-Portugal, B. F. King, C. Gachet, K. A. Jacobson, G. A. Weisman and G. Burnstock (2003). "Characterization of the UDP-glucose receptor (re-named here the P2Y14 receptor) adds diversity to the P2Y receptor family." Trends Pharmacol. Sci. 24(2): 52-5. Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts and J. S. Watson (1994). Carrier proteins and active membrane transport. Mol. Biol. Cell. R. Adams. New York, Garland publishing, Inc.: 512-22. Ankel, H., E. Ankel and D. S. Feingold (1966). "Biosynthesis of Uridine Diphosphate D-xylose. III Uridine diphosophate D-glucose Dehy- drogenase of Cryptococcus laurentii." Biochemistry 5(6 June): 1864-69. Arrecubieta, C., E. Garcia and R. Lopez (1996). "Demonstration of UDP- glucose dehydrogenase activity in cell extracts of Escherichia coli expressing the pneumococcal cap3A gene required for the synthesis of type 3 capsular polysaccharide." J. Bacteriol. 178(10): 2971-4. Ashbaugh, C. D., S. Alberti and M. R. Wessels (1998). "Molecular analysis of the capsule gene region of group A Streptococcus: the hasAB genes are sufficient for capsule expression." J. Bacteriol. 180(18): 4955-9. Bahrani-Mougeot, F. K., E. L. Buckles, C. V. Lockatell, J. R. Hebel, D. E. Johnson, C. M. Tang and M. S. Donnenberg (2002). "Type 1 fim- briae and extracellular polysaccharides are preeminent uropatho- genic Escherichia coli virulence determinants in the murine urinary tract." Mol. Microbiol. 45(4): 1079-93. Balduini, C., A. Brovelli and A. A. Castellani (1970). "Biosynthesis of Gly- cosaminoglycans in Bovine Cornea- the effect of uridine diphosphate xylose." Biochem. J. 120: 719-23. Bar-Peled, M., C. L. Griffith, J. J. Ory and T. L. Doering (2004). "Biosyn- thesis of UDP-GlcA, a key metabolite for capsular polysaccharide synthesis in the pathogenic fungus Cryptococcus neoformans." Bio- chem. J. Pt. Bardoni, A., M. P. Pagliula, G. Pallavicini, S. Rindi, A. A. Castellani and G. De Luca (1989). "Biosynthesis of glycosaminoglycan precursors: Evidences for different tissue specific forms of UDP-glucose dehydrogenase." Ital. J. Biochem. 38: 360-68. 48 Berninsone, P. M. and C. B. Hirschberg (2000). "Nucleotide sugar transporters of the Golgi apparatus." Curr. Opin. Struct. Biol. 10(5): 542-7. Binari, R. C., B. E. Staveley, W. A. Johnson, R. Godavarti, R. Sasisekharan and A. S. Manoukian (1997). "Genetic evidence that heparin-like glycosaminoglycans are involved in wingless signaling." Develop- ment 124: 2623-32. Bitter, T. and H. M. Muir (1962). "A modified uronic acid carbazole reaction." Anal. Biochem. 4: 330-34. Bolwell, P. G. (1993). Dynamic aspects of the Plant Extracellular matrix. Int. Rev. Cytol., Academic Press Inc. 146: 261-324. Bontemps, Y., F.-X. Maquart and Y. Wegrowski (2000). "Human UDP- glucose dehydrogenase gene: complete cloning and transcription start mapping." Biochem. Biophys. Res. Commun. 275: 981-85. Campbell, R. E., S. C. Mosimann, I. van de Rijn, M. E. Tanner and N. C. J. Strynadka (2000). "The first structure of UDP-Glucose dehydrogenase reveals the catalytic residues necessary for the two-fold oxidation." Biochemistry 39: 7012-23. Campbell, R. E., R. F. Sala, I. van de Rijn and M. E. Tanner (1997). "Proper- ties and kinetic analysis of UDP-glucose dehydrogenase from group A Streptococci. Irreversible inhibition by UDP-chloroacetol." J. Biol. Chem. 272(Feb 7): 3416-22. Carpita, N. C. and D. M. Gibeaut (1993). "Structural models of primary cell walls in flowering plants: consistency of molecular structure with the physical properties of the walls during growth." Plant J. 3(1): 1- 30. Castellani, A., A., G. De Luca, S. Rindi, R. Salvini and M. Tira, E. (1986). "Regulatory mechanisms of UDP-glucuronic acid biosynthesis in cultured human skin fibroblasts." Ital. J. Biochem.(Sept-Oct 35 (5)): 296-303. Champe, P. C., R. A. Harvey and D. R. Ferrier (2005). Metabolism of monosaccharides and disaccharides. Biochemistry S. Seigafuse. Balti- more, MD, Lippincott Williams & Wilkins: 135-42. Chung, J. Y., Y. Zhang and B. Adler (1998). "The capsule biosynthetic locus of Pasteurella multocida A:1." FEMS Microbiol. Lett. 166(2): 289- 96. Davies, G. J. (2001). "Sweet secrets of synthesis." Nat. Struct. Biol. 8(2): 98- 100. De Luca, G., P. Speziale, S. Rindi, C. Balduini and A. A. Castellani (1976). "Effect of some nucleotides on the regulation of glycosaminoglycan biosynthesis." Connect. Tissue Res. 4: 247-54. DeAngelis, P. L. (2002). "Microbial glycosaminoglycan glycosyltransferases." Glycobiology 12(1): 9R-16R. DeAngelis, P. L. and A. J. Padgett-McCue (2000). "Identification and molecular cloning of a chondroitin synthase from Pasteurella multocida type F." J. Biol. Chem. 275(31): 24124-9.

49 DeAngelis, P. L. and C. L. White (2002). "Identification and molecular cloning of a heparosan synthase from Pasteurella multocida type D." J. Biol. Chem. 277(9): 7209-13. Dillard, J. P., M. W. Vandersea and J. Yother (1995). "Characterization of the cassette containing genes for type 3 capsular polysaccharide biosynthesis in Streptococcus pneumoniae." J. Exp. Med. 181(3): 973-83. Dougherty, B. A. and I. van de Rijn (1993). "Molecular characterization of hasB from an operon required for hyaluronic acid synthesis in group A Streptococci Demonstration of UDP-glucose dehydrogenase activity." J. Biol. Chem. 268(10): 7118-24. Dutton, G. J., Ed. (1966). Glucuronic acid- free and combined. New York, Academic press. Dörmann, P. and C. Benning (1998). "The role of UDP-glucose epimerase in carbohydrate metabolism of Arabidopsis." Plant J. 13(5): 641-52. EcoCyc (2002). "Nucleic Acids Res." 30(1): 59. Eriksson, G., B. Särnstrand and A. Malmström (1984). "Equilibration of [3H]glucosamine and [35S]sulfate with intracellular pools of UDP-N- Acetylhexosamine and 3´-Phosphoadenosine-5´-phosphosulfate (PAPS) in cultured fibroblasts." Arch. Biochem. Biophys. 235(2 Dec): 692-98. Feingold, D. S. (1982). Aldo (and keto) hexoses and uronic acids. Plant Car- bohydrates I, Encyclopedia of Plant Physiology. F. A. Loewus and W. Tanner. Berlin, Springer Verlag. 13A: 3-76. Feingold, D. S. and G. A. Barber (1990). Nucleotide sugars. Methods in plant biochemistry. H. Dey. New York, Academic press limited. Vol. 2 Carbohydrates: 39-78. Feingold, D. S. and J. S. Franzen (1981). "Pyridine nucleotide-linked four- electron transfer dehydrogenases." Trends Biochem. Sci. vol 6(April): 103-05. Fell, D. (1997). Understanding the control of metabolism. London, Portland press. Feyzi, E., T. Saldeen, E. Larsson, U. Lindahl and M. Salmivirta (1998). "Age-dependent modulation of heparan sulfate structure and function." J. Biol. Chem. 273(May 29): 13395-98. Freeze, H. H. (1999). Monosaccharide metabolism. Essentials of glycobiology. A. Varki, R. Cummings, J. Esko, H. Freeze, G. Hart and J. Marth. New York, Coldspring Harbor Laboratory Press: 69-84. Fry, S. C. (1999, september). Plant cell walls, Nature encyclopedia of life sciences. 2004. Gainey, P. A. and C. F. Phelps (1972). "Uridine diphosphate glucuronic acid production and utilization in various tissues actively synthesizing glycosaminoglycans." Biochem. J. 128: 215-27.

50 Gainey, P. A. and C. F. Phelps (1975). "Interactions of uridine diphosphate glucose dehydrogenase with the inhibitor uridine diphosphate xylose." Biochem. J. 145: 129-34. Ge, X., L. C. Penney, I. Van De Rijn and M. E. Tanner (2004). "Active site residues and mechanism of UDP-glucose dehydrogenase." Eur. J. Biochem. 271(1): 14-22. Gerardy-Schahn, R., S. Oelmann and H. Bakker (2001). "Nucleotide sugar transporters: biological and functional aspects." Biochimie 83(8): 775-82. Gibeaut, D. M. (2000). "Nucleotide sugars and glycosyltransferases for synthesis of cell wall matrix polysaccharides." Plant physiol. biochem 38(1/2): 69-80. Grangeasse, C., B. Obadia, I. Mijakovic, J. Deutscher, A. J. Cozzone and P. Doublet (2003). "Autophosphorylation of the Escherichia coli protein kinase Wzc regulates tyrosine phosphorylation of Ugd, a UDP- glucose dehydrogenase." J. Biol. Chem. 278(41): 39323-9. Griffith, C. L., J. S. Klutts, L. Zhang, S. B. Levery and T. L. Doering (2004). "UDP-glucose dehydrogenase plays multiple roles in the biology of the pathogenic fungus Cryptococcus neoformans." J. Biol. Chem. Griffiths, G., N. J. Cook, E. Gottfridson, T. Lind, K. Lidholt and I. S. Rob- erts (1998). "Characterization of the glycosyltransferase enzyme from the Escherichia coli K5 capsule gene cluster and identification and characterization of the glucuronyl active site." J. Biol. Chem. 273(19): 11752-7. Gupta, D. S., B. Jann, G. Schmidt, J. R. Golecki, I. Orskov and K. Jann (1982). "Coliphage K5, specific for E. coli exhibiting the capsular K5 antigen." FEMS Letters 14: 75-78. Hacker, U., X. Lin and N. Perrimon (1997). "The Drosophila sugarless gene modulates Wingless signaling and encodes an enzyme involved in polysaccharide biosynthesis." Development 124(18): 3565-73. Haerry, T. E., T. R. Heslip, J. L. Marsh and M. B. O'Connor (1997). "De- fects in glucuronate biosynthesis disrupt Wingless signaling in Dro- sophila." Development 124(16): 3055-64. Handley, C. J. and C. F. Phelps (1972a). "The biosynthesis in vitro of chondroitin sulphate in neonatal rat epiphysial cartilage." Biochem. J. 126(2): 417-32. Handley, C. J. and C. F. Phelps (1972b). "The biosynthesis in vitro of keratan sulphate in bovine cornea." Biochem. J. 128(2): 205-13. Handley, C. J. and C. F. Phelps (1972c). "The concentrations of sugar nucleotides in bovine in corneal epithelium and endothelium." Bio- chem. J. 127: 911-12. Hanfling, P., A. S. Shashkov, B. Jann and K. Jann (1996). "Analysis of the enzymatic cleavage (ß elimination) of the capsular K5 polysaccharide of Escherichia coli by the K5-specific coliphage: a reexamina- tion." J. Bacteriol. 178(no 15 Aug): 4747-50.

51 Harper, A. D. and M. Bar-Peled (2002). "Biosynthesis of UDP-Xylose. Cloning and Characterization of a Novel Arabidopsis Gene Family, UXS, Encoding Soluble and Putative Membrane- Bound UDP- Glucuronic Acid Decarboxylase Isoforms." Plant Physiol. 130(4): 2188-98. Hascall, V. C., A. K. Majors, C. A. De La Motte, S. P. Evanko, A. Wang, J. A. Drazba, S. A. Strong and T. N. Wight (2004). "Intracellular hyaluronan: a new frontier for inflammation?" Biochim Biophys Acta 1673(1-2): 3-12. Heldt, H.-W. (1997). Plant biochemistry & molecular biology. Oxford, Ox- ford university press. Hempel, J., J. Perozich, H. Romovacek, A. Hinich, I. Kuo and D. S. Fein- gold (1994). "UDP-glucose dehydrogenase from bovine liver: Pri- mary structure and relationship to other dehydrogenases." Protein Sci. 3: 1074-80. Hickery, M. S., M. T. Bayliss, J. Dudhia, J. C. Lewthwaite, J. C. Edwards and A. A. Pitsillides (2003). "Age-related changes in the response of human articular cartilage to IL-1alpha and transforming growth factor-beta (TGF-beta): chondrocytes exhibit a diminished sensitivity to TGF-beta." J. Biol. Chem. 278(52): 53063-71. Hirschberg, C. B., P. W. Robbins and C. Abeijon (1998). "Transporters of nucleotide sugars, ATP, and nucleotide sulfate in the endoplasmic reticulum and Golgi apparatus." Annu. Rev. Biochem. 67: 49-69. Hodson, N., G. Griffiths, N. Cook, M. Pourhossein, E. Gottfridson, T. Lind, K. Lidholt and I. S. Roberts (2000). "Identification that KfiA, a protein essential for the biosynthesis of the Escherichia coli K5 capsular polysaccharide, is an alpha -UDP- GlcNAc glycosyltransferase. The formation of a membrane-associated K5 biosynthetic complex requires KfiA, KfiB, and KfiC." J. Biol. Chem. 275(35): 27311-15. Hoshi, H., H. Nakagawa, S. Nishiguchi, K. Iwata, K. Niikura, K. Monde and S. Nishimura (2004). "An engineered hyaluronan synthase: characterization for recombinant human hyaluronan synthase 2 Escherichia coli." J. Biol. Chem. 279(4): 2341-9. Hwang, H. Y. and H. R. Horvitz (2002). "The Caenorhabditis elegans vulval morphogenesis gene sqv-4 encodes a UDP-glucose dehydrogenase that is temporally and spatially regulated." Proc. Natl. Acad. Sci. U. S. A. 21: 21. Iozzo, R. V. (1998). "Matrix proteoglycans: from molecular design to cellular function." Annu. Rev. Biochem. 67: 609-52. Itano, N. and K. Kimata (2002). "Mammalian hyaluronan synthases." IUBMB Life 54(4): 195-9. Jaenicke, R., R. Rudolph and D. S. Feingold (1986). "Dissociation and in vitro reconstitution of bovine liver uridine diphosphoglucose dehydrogenase. The paired subunit nature of the enzyme." Biochemistry 25(23): 7283-87.

52 Jing, W. and P. L. DeAngelis (2003). "Analysis of the two active sites of the hyaluronan synthase and the chondroitin synthase of Pasteurella multocida." Glycobiology 13(10): 661R-71R. Kearns, A. E., B. M. Vertel and N. B. Schwartz (1993). "Topography of glycosylation and UDP-Xylose production." J. Biol. Chem. 268(No 15 May 25): 11097-104. Kneidinger, B., K. O'Riordan, J. Li, J. R. Brisson, J. C. Lee and J. S. Lam (2003). "Three Highly Conserved Proteins Catalyze the Conversion of UDP-N-acetyl-D-glucosamine to Precursors for the Biosynthesis of O Antigen in Pseudomonas aeruginosa O11 and Capsule in Staphylococcus aureus Type 5. Implications for the UDP-N-acetyl- L-fucosamine biosynthetic pathway." J. Biol. Chem. 278(6): 3615- 27. Kobayashi, M., H. Nakagawa, I. Suda, I. Miyagawa and T. Matoh (2002). "Purification and cDNA cloning of UDP-D-glucuronate carboxy- lyase (UDP-D-xylose synthase) from pea seedlings." Plant Cell Physiol. 43(11): 1259-65. Kotake, T., D. Yamaguchi, H. Ohzono, S. Hojo, S. Kaneko, H. K. Ishida and Y. Tsumuraya (2004). "UDP-sugar pyrophosphorylase with broad substrate specificity towards various monosaccharide 1-phosphates from pea sprouts." J. Biol. Chem. E-pub ahead of print Kresse, H. (1997). Proteoglycans- structure and function. Glycosciences status and perspectives. H. J. Gabius and S. Gabius. Weinheim, Chapman & Hall: 201-22. Kutchai, H. C. (1990). Digestion and absorption. Principles of physiology. R. M. Berne and M. N. Levy. London, Wolfe publishing limited: 398-414. Landstein, D., M. V. Graves, D. E. Burbank, P. DeAngelis and J. L. Van Etten (1998). "Chlorella virus PBCV-1 encodes functional gluta- mine: fructose-6-phosphate amidotransferase and UDP-glucose dehydrogenase enzymes." Virology 250: 388-96. Lapointe, J. and C. Labrie (1999). "Identification and cloning of a novel androgen-responsive gene, uridine diphosphoglucose dehydrogenase, in human breast cancer cells." Endocrinology 140(no 10): 4486-93. Laurent, T. C. and J. R. Fraser (1992). "Hyaluronan." FASEB J. 6(7): 2397- 404. Ledin, J., W. Staatz, J. P. Li, M. Gotte, S. Selleck, L. Kjellen and D. Spill- mann (2004). "Heparan Sulfate Structure in Mice with Genetically Modified Heparan Sulfate Production." J. Biol. Chem. 279(41): 42732-41. Lindahl, U., M. Kusche-Gullberg and L. Kjellén (1998). "Regulated diversity of heparan sulfate." J. Biol. Chem. 273, no 39(Sept 25): 24979- 82.

53 Liu, J. and S. C. Thorp (2002). "Cell surface heparan sulfate and its roles in assisting viral infections." Med. Res. Rev. 22(1): 1-25. Loewus, F., M.-S. Chen and M. W. Loewus (1973). The myo-inositol pathway to cell wall polysaccharides. Biogenesis of Plant Cell Wall Polysaccharides. Loewus. New York, Academic Press: 1-27. Marcu, O., D. G. Stathakis and J. L. Marsh (1999). "Assignment of the UGDH locus encoding UDP-glucose dehydrogenase to human chromosome band 4p 15.1 by radiation hybrid mapping." Cytogenet Cell Genet 86: 244-45. Mijakovic, I., S. Poncet, G. Boel, A. Maze, S. Gillet, E. Jamet, P. Decottig- nies, C. Grangeasse, P. Doublet, P. Le Marechal and J. Deutscher (2003). "Transmembrane modulator-dependent bacterial tyrosine kinase activates UDP-glucose dehydrogenases." EMBO J. 22(18): 4709-18. Moriarity, J. L., K. J. Hurt, A. C. Resnick, P. B. Storm, W. Laroy, R. L. Schnaar and S. H. Snyder (2002). "UDP-glucuronate decarboxylase, a key enzyme in proteoglycan synthesis: cloning, characterization, and localization." J. Biol. Chem. 277(19): 16968-75. Mroz, P. J. and J. E. Silbert (2003). "Effects of [3H]glucosamine concentration on [3H]chondroitin sulphate formation by cultured chondrocytes." Biochem. J. 376(Pt 2): 511-5. Nelson, D. L. and M. M. Cox (2000a). Biosynthesis of glycogen, starch, sucrose and other carbohydrates. Lehninger Principles of biochemistry. E. Geller. New York, Worth publishers: 735-46. Nelson, D. L. and M. M. Cox (2000b). Regulation of carbohydrate catabo- lism. Lehninger Principles of biochemistry. E. Geller. New York, Worth publishers: 551-57. Neufeld, E. F. and C. W. Hall (1965). "Inhibition of UDP-D-Glucose dehydrogenase by UDP-D-Xylose: a possible regulatory mechanism." Biochem. Biophys. Res. Commun. 19(no 4): 456-61. Ninomiya, T., N. Sugiura, A. Tawada, K. Sugimoto, H. Watanabe and K. Kimata (2002). "Molecular cloning and characterization of chondroitin polymerase from Escherichia coli strain K4." J. Biol. Chem. 277(24): 21567-75. Nybakken, K. and N. Perrimon (2002). "Heparan sulfate proteoglycan modulation of developmental signaling in Drosophila." Biochim Biophys Acta 1573(3): 280-91. Ohya, T. and Y. Kaneko (1970). "Novel hyaluronidase from streptomyces." Biochim Biophys Acta 198(3): 607-9. Ordman, A. B. and S. Kirkwood (1977). "UDPglucose dehydrogenase kinetics and their mechanistic implications." Biochim. Biophys. Acta 481(1): 25-32. Peng, H.-L., M.-D. Lou, M.-L. Chang and H.-Y. Chang (1998). "cDNA cloning and expression analysis of the human UDPglucose dehydrogenase." Proc. Natl. Sci. Counc ROC (B) 22(no 4): 166-72.

54 Petit, C., G. P. Rigg, C. Pazzani, A. Smith, V. Sieberth, M. Stevens, G. Boulnois, K. Jann and I. S. Roberts (1995). "Region 2 of the Es- cherichia coli K5 capsule gene cluster encoding proteins for the biosynthesis of the K5 polysaccharide." Mol. Microbiol. 17( (4)): 611- 20. Petitou, M. and C. A. A. van Boeckel (2004). "A synthetic antithrombin III binding pentasaccharide is now a drug! What comes next?" Angew. Chem. Int. Ed. Engl.(43): 3118-33. Pilling, E. and H. Höfte (2003). "Feedback from the wall." Curr. Opin. Plant Biol. 6: 611-16. Pitsillides, A. A., L. S. Wilkinson, M. T. Mehdizadeh, B. T. Bayliss and J. C. W. Edwards (1993). "Uridine diphosphoglucose dehydrogenase activity in normal and rheumatoid synovium: the description of a spe- cialized synovial lining cell." Int. J. Exp. Path. 74: 27-34. Prescott, L. M., J. P. Harley and D. Klein, A. (1993). Microbial nutrition. Microbiology. K. Kane. Dubuque, Wm. C. Brown publishers: 96- 111. Qiao, D., K. Meyer, C. Mundhenke, S. A. Drew and A. Friedl (2003). "Heparan sulfate proteoglycans as regulators of fibroblast growth factor-2 signaling in brain endothelial cells. Specific role for glypi- can-1 in glioma angiogenesis." J. Biol. Chem. 278(18): 16045-53. Rabina, J., M. Maki, E. M. Savilahti, N. Jarvinen, L. Penttila and R. Renkonen (2001). "Analysis of nucleotide sugars from cell lysates by ion-pair solid-phase extraction and reversed-phase high- performance liquid chromatography." Glycoconj. J. 18(10): 799- 805. Reiter, W.-D. (1998). "The molecular analysis of cell wall components." Trends in Plant Sci. 3(1 January): 27-32. Reiter, W. D. and G. F. Vanzin (2001). "Molecular genetics of nucleotide sugar interconversion pathways in plants." Plant Mol. Biol. 47(1-2): 95-113. Rigg, G. P., B. Barrett and I. S. Roberts (1998). "The localization of KpsC, S and T and KfiA, C and D proteins involved in the biosynthesis of the Escherichia coli K5 capsular polysaccharide: evidence for a membrane-bound complex." Microbiology 144: 2905-14. Roberts, I. S. (1995). "Bacterial polysaccharides in sickness and in health. The 1995 Fleming Lecture." Microbiology 141(Pt 9): 2023-31. Roberts, I. S., R. Mountford, R. Hodge, K. B. Jann and G. J. Boulnois (1988). "Common organization of gene clusters for production of different capsular polysaccharides (K Antigens) in Escherichia coli." J. Bacteriol. 170 no 3: 1305-10. Roberts, I. S., F. K. Saunders and G. J. Boulnois (1989). "Bacterial capsules and interactions with complement and phagocytes." Biochem. Soc. Trans. 17(3): 462-4.

55 Robertson, D., I. Beech and G. P. Bolwell (1995). "Regulation of the enzymes of UDP-sugar metabolism during differentiation of French bean." Phytochemistry 39(1): 21-28. Samac, D. A., L. Litterer, G. Temple, H. J. Jung and D. A. Somers (2004). "Expression of UDP-Glucose Dehydrogenase Reduces Cell-Wall Polysaccharide Concentration and Increases Xylose Content in Al- falfa Stems." Appl. Biochem. Biotechnol. 116(1-3): 1167-82. Sasisekharan, R., Z. Shriver, G. Venkataraman and U. Narayanasami (2002). "Roles of heparan-sulphate glycosaminoglycans in cancer." Nat. Rev. Cancer 2(7): 521-8. Schiller, J. G., A. M. Bowser and D. S. Feingold (1973). "Partial purification and properties of UDPG dehydrogenase from Escherichia coli." Biochim. Biophys. Acta 293: 1-10. Schiller, J. G., F. Lamy, R. Frazier and D. S. Feingold (1976). "UDP- Glucose dehydrogenase from Escherichia coli purification and subunit structure." Biochim. Biophys. Acta 453: 418-25. Schomburg, I., A. Chang, C. Ebeling, M. Gremse, C. Heldt, G. Huhn and D. Schomburg (2004). "BRENDA, the enzyme database: updates and major new developments." Nucleic Acids Res (Database issue): D431-3. Seitz, B., C. Klos, M. Wurm and R. Tenhaken (2000). "Matrix polysaccharide precursors in Arabidopsis cell walls are synthesized by alternate pathways with organ-specific expression patterns." Plant J. 21(6): 537-46. Sieberth, V., G. P. Rigg, I. S. Roberts and K. Jann (1995). "Expression and characterization of UDPGlc dehydrogenase (KfiD), which is encoded in the type-specific region 2 of the Escherichia coli K5 capsule genes." J. Bacteriol. 177(15): 4562-65. Siebold, C., K. Flukiger, R. Beutler and B. Erni (2001). "Carbohydrate transporters of the bacterial phosphoenolpyruvate: sugar phos- photransferase system (PTS)." FEBS Lett. 504(3): 104-11. Silbert, J. E. and G. Sugumaran (2002). "Biosynthesis of chondroitin/dermatan sulfate." IUBMB Life 54(4): 177-86. Singh, J., L. R. Schwarz and F. J. Wiebel (1980). "A rapid enzymic proce- dure for the determination of picomole amounts of UDP-glucuronic acid." Biochem. J. 189: 369-72. Sivaswami, A., S. M. Kelkar and G. B. Nadkarni (1972). "Uridine diphosphoglucose dehydrogenase from rat liver: purification and effect of pH on regulatory properties." Biochim. Biophys. Acta 276: 43-52. Smith, A. D., S. P. Datta, G. Howard Smith, P. N. Campbell, R. Bentley and H. A. McKenzie, Eds. (1997). Oxford dictionary of Biochemistry and molecular biology. Oxford, Oxford university press.

56 Smith, C. J. (1993). Carbohydrate chemistry. Plant biochemistry and molecular biology. P. J. Lea and R. C. Leegood. Chichester, John Wiley & sons: 73-111. Sommer, B. J., J. J. Barycki and M. A. Simpson (2004). "Characterization of human UDP-glucose dehydrogenase: Cys 276 is required for the second of two successive oxidations." J. Biol. Chem. E-pub ahead of print Spicer, A. P., L. A. Kaback, T. J. Smith and M. F. Seldin (1998). "Molecular cloning and characterization of human and mouse UDP-Glucose dehydrogenase genes." J. Biol. Chem. 273(Sept 25): 25117-24. Spicer, A. P. and J. Y. Tien (2004). "Hyaluronan and morphogenesis." Birth Defects Res Part C Embryo Today 72(1): 89-108. Strominger, J. L., H. M. Kalckar, J. Axelrod and E. S. Maxwell (1954). "En- zymatic oxidation of uridine diphosphate glucose to uridine diphosphate glucuronic acid." J. Am. Chem. Soc. 76: 6411-12. Strominger, J. L. and L. W. Mapson (1957). "Uridine diphosphoglucose dehydrogenase of pea seedlings." Biochem J. 66(Jan 14): 567-72. Suda, T., S. Kamiyama, M. Suzuki, N. Kikuchi, K. I. Nakayama, H. Nari- matsu, Y. Jigami, T. Aoki and S. Nishihara (2004). "Molecular cloning and characterization of a human multi-substrate specific nucleotide-sugar transporter homologous to Drosophila fringe connection." J. Biol. Chem. E-pub ahead of print Sugahara, K. and H. Kitagawa (2002). "Heparin and heparan sulfate biosynthesis." IUBMB Life 54(4): 163-75. Sugahara, K., T. Mikami, T. Uyama, S. Mizuguchi, K. Nomura and H. Kita- gawa (2003). "Recent advances in the structural biology of chondroitin sulfate and dermatan sulfate." Curr. Opin. Struct. Biol. 13(5): 612-20. Sweeney, C., D. Mackintosh and R. M. Mason (1993). "UDP-sugar metabolism in Swarm rat chondrosarcoma chondrocytes." Biochem. J. 290: 563-70. Toole, B. P. (2002). "Hyaluronan promotes the malignant phenotype." Gly- cobiology 12(3): 37R-42R. Tukey, R. H. and C. P. Strassburg (2000). "Human UDP- glucuronosyltransferases: metabolism, expression, and disease." Annu. Rev. Pharmacol. Toxicol. 40: 581-616. Turner, W. and F. C. Botha (2002). "Purification and kinetic properties of UDP-glucose dehydrogenase from sugarcane." Arch. Biochem. Bio- phys. 407(2): 209-16. Unligil, U. M., S. Zhou, S. Yuwaraj, M. Sarkar, H. Schachter and J. M. Rini (2000). "X-ray crystal structure of rabbit N- acetylglucosaminyltransferase I: catalytic mechanism and a new protein superfamily." EMBO J. 19(20): 5269-80. Uzuki, M., T. Watanabe, Y. Katsura and T. Sawai (1999). "Quantitative histochemical study of hyaluronic acid binding protein and the activ-

57 ity of uridine diphosphoglucose dehydrogenase in the synovium of patients with rheumatoid arthritis." Analytical and quantitative cy- tology and histology 21: 75-80. Vann, W. F., M. A. Schmidt, B. Jann and K. Jann (1981). "The structure of the capsular polysaccharide (K5 antigen) of urinary-tract-infective Escherichia coli 010:K5:H4." Eur. J. Biochem. 116: 359-64. Walsh, E. C. and D. Y. Stainier (2001). "Udp-glucose dehydrogenase required for cardiac valve formation in zebrafish." Science 293(5535): 1670-3. Wegrowski, Y., C. Perreau, Y. Bontemps and F.-X. Maquart (1998). "Uridine diphosphoglucose dehydrogenase regulates proteoglycan expression: cDNA cloning and antisense study." Biochem. Biophys. Res. Commun. 250: 206-11. Weigel, P. H. (2002). "Functional characteristics and catalytic mechanisms of the bacterial hyaluronan synthases." IUBMB Life 54(4): 201-11. Wierenga, R. K., P. Terpstra and W. G. J. Hol (1986). "Prediction of the occurrence of the ADP-bindingEDE-fold in proteins, using an amino acid sequence fingerprint." J Mol. Biol. 187: 101-07. Yamada, S. and K. Sugahara (1998). "Structure of oligosaccharides isolated from heparan sulfate/heparin and substrate specificities of the de- grading enzymes of bacterial origin." Trends in Glycosci. Gly- cotechn. 10 no 52(March 1998): 95-123. Yamagata, T., H. Saito, O. Habuchi and S. Suzuki (1968). "Purification and properties of bacterial chondroitinases and chondrosulfatases." J. Biol. Chem. 243(7): 1523-35. Ysart, G. E. and R. M. Mason (1994). "Serum factors, growth factors and UDP-sugar metabolism in bovine articular cartilage chondrocytes." Biochem. J. 303: 713-21. Zablackis, E., J. Huang, B. Muller, A. G. Darvill and P. Albersheim (1995). "Characterization of the cell-wall polysaccharides of Arabidopsis thaliana leaves." Plant Physiol. 107(4): 1129-38. Zalitis, J. and D. S. Feingold (1968). "The mechanism of action of UDPG dehydrogenase." Biochem. Biophys. Res. Commun. 31(5): 693-98. Zalitis, J., M. Uram, A. M. Bowser and D. S. Feingold (1972). UDP-glucose dehydrogenase from beef liver. Methods Enzymol. V. Ginsberg. New York, Academic press. 28: 430-35. Zhivkov, V., R. Tosheva and Y. Zhivkova (1975). "Concentration of uridine diphosphate sugars in various tissues of vertebrates." Comp. Bio- chem. Physiol. B. 51(4): 421-4.

58 Errata

p. 21, legend of figure 5. The configuration of the monosaccharides in should be read D-, not “a-”. p. 21, last paragraph, line five from the bottom. A hydride ion equals a pro- ton and an electron pair, not “hydrogen and an electron pair”. p. 23, legend of figure 6. UDP should be read as uridinephospho- only when conjugated to another molecule, but as uridinediphosphate when unbound. p. 34 figure 9. The C-6-carboxylgroup of the glucuronic acid should carry a minus sign in both part A and B.

Elisabet Roman, Dec. 4, 2004 Acta Universitatis Upsaliensis Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Comprehen- sive Summaries of Uppsala Dissertations from the Faculty of Medicine. (Prior to October, 1985, the series was published under the title “Abstracts of Uppsala Dissertations from the Faculty of Medicine”.)

Distribution: Uppsala University Library Box 510, SE-751 20 Uppsala, Sweden www.uu.se, [email protected]

ISSN 0282-7476 ISBN 91-554-6088-7