Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1143

Glycosaminoglycan Biosynthesis in Zebrafish

BEATA FILIPEK-GÓRNIOK

ACTA UNIVERSITATIS UPSALIENSIS ISSN 1651-6206 ISBN 978-91-554-9368-4 UPPSALA urn:nbn:se:uu:diva-264269 2015 Dissertation presented at Uppsala University to be publicly examined in C8:305, BMC, Husargatan 3, Uppsala, Friday, 27 November 2015 at 13:15 for the degree of Doctor of Philosophy (Faculty of Medicine). The examination will be conducted in English. Faculty examiner: Associate Professor Kay Grobe (Institute for Physiological Chemistry and Pathobiochemistry, University of Münster).

Abstract Filipek-Górniok, B. 2015. Glycosaminoglycan Biosynthesis in Zebrafish. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1143. 54 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-554-9368-4.

Proteoglycans (PGs) are composed of highly sulfated glycosaminoglycans chains (GAGs) attached to specific core proteins. They are present in extracellular matrices, on the cell surface and in storage granules of hematopoietic cells. Heparan sulfate (HS) and chondroitin/dermatan sulfate (CS/DS) GAGs play indispensable roles in a wide range of biological processes, where they can serve as protein carriers, be involved in growth factor or morphogen gradient formation and act as co-receptors in signaling processes. Protein binding abilities of GAGs are believed to be predominantly dependent on the arrangement of the sugar modifications, sulfation and epimerization, into specific oligosaccharide sequences. Although the process of HS and CS/DS assembly and modification is not fully understood, a set of GAG biosynthetic enzymes have been fairly well studied and several mutations in genes encoding for this Golgi machinery have been linked to human genetic disorders. This thesis focuses on the zebrafish N-deacetylase/N-sulfotransferase gene family, encoding key enzymes in HS chain modification, as well as glycosyltransferases responsible for chondroitin/dermatan sulfate elongation present in zebrafish. Our data illustrates the strict spatio-temporal expression of both the NDST enzymes (Paper I) and CS/DS glycosyltransferases (Paper II) in the developing zebrafish embryo. In Paper III we took advantage of the four preexisting zebrafish mutants with defective GAG biosynthesis. We could demonstrate a relation between HS content and the severity of the pectoral fin defects, and additionally correlate impaired HS biosynthesis with altered chondrocyte intercalation. Interestingly, altered CS biosynthesis resulted in loss of the chondrocyte extracellular matrix. One of the main findings was the demonstration of the ratio between the HS biosynthesis enzyme Extl3 and the Csgalnact1/Csgalnact2 proteins, as a main factor influencing the HS/CS ratio. In Paper IV we used the newly developed CRISPR/Cas9 technique to create a collection of zebrafish mutants with defective GAG biosynthetic machineries. Lack of phenotypes linked to null-mutations of most of the investigated genes is striking in this study.

Keywords: Heparan sulfate, chondroitin/dermatan sulfate, biosynthesis, development, N- deacetylase N-sulfotransferase, glycosyltransferases, morpholino, CRISPR-Cas9

Beata Filipek-Górniok, Department of Medical Biochemistry and Microbiology, Box 582, Uppsala University, SE-75123 Uppsala, Sweden.

© Beata Filipek-Górniok 2015

ISSN 1651-6206 ISBN 978-91-554-9368-4 urn:nbn:se:uu:diva-264269 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-264269)

Nothing that comes to easy has any significance.

Jacek Hugo-Bader, journalist and reporter

Supervisors: Lena Kjellén, Professor Department of Medical Biochemistry and Microbiology Uppsala University Uppsala, Sweden

Johan Ledin, Researcher Department of Organismal Biology Uppsala University Uppsala, Sweden

Faculty opponent: Kay Grobe, Associate Professor Institute for Physiological Chemistry and Pathobiochemistry University of Münster Münster, Germany

Examining Committee: Dan Larhammar, Professor Department of Neuroscience, Pharmacology Uppsala University Uppsala, Sweden

Hiroshi Nakato, Associate Professor Department of Genetics, Cell Biology and Development University of Minnesota Minnesota, US

Sally Stringer, Honorary Lecturer/Editorial team leader Cardiovascular Division University of Manchester/HealthCare21 Communications Ltd, Manchester, UK

Chairperson: Jin-ping Li, Professor Department of Medical Biochemistry and Microbiology Uppsala University Uppsala, Sweden

Cover: Adult female zebrafish. Drawing by Beata Filipek-Górniok, 2015.

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Filipek-Górniok B., Carlsson P., Haitina T., Habicher J., Ledin J., Kjellén L. (2015) The NDST gene family in zebrafish: role of NDST1b in pharyngeal arch formation. PLoS One, 10(3):e0119040. doi: 10.1371/journal.pone.0119040. II Filipek-Górniok B., Holmborn K., Haitina T., Habicher J., Oliveira M.B., Hellgren C., Eriksson I., Kjellén L., Kreuger J., Ledin J. (2013) Expression of chondroitin/dermatan sulfate glycosyltransferases during early zebrafish development. Developmental Dynamics, 242(8):964-75. doi: 10.1002/dvdy.23981. III Holmborn K., Habicher J., Kasza Z., Eriksson A.S., Filipek- Górniok B., Gopal S., Couchman J.R., Ahlberg P.E., Wiweger M., Spillmann D., Kreuger J., Ledin J. (2012) On the roles and regulation of chondroitin sulfate and heparan sulfate in zebrafish pharyngeal cartilage morphogenesis. The Journal of biological chemistry, 287(40):33905-16. IV Habicher J., Filipek-Górniok B., Varshney G., Ahlberg P.E., Burgess S., Kjellén L., Ledin J. (2015) Large-scale generation of zebrafish mutants with defective glycosaminoglycan biosynthesis. Manuscript.

Reprints were made with permission from the respective publishers.

Additional Publications

I Fisher S., Filipek-Górniok B., Ledin J. (2011) Zebrafish Ext2 is necessary for Fgf and Wnt signaling, but not for Hh signaling. BMC Developmental Biology, 11:53. doi: 10.1186/1471-213X-11-53. II Dagälv A., Lundequist A., Filipek-Górniok B., Dierker T., Eriksson I., Kjellén L. (2015) Heparan sulfate structure: methods to study N- sulfation and NDST action. Methods in molecular biology, 1229:189- 200. doi: 10.1007/978-1-4939-1714-3_17.

Contents

Introduction ...... 13 Background ...... 14 Proteoglycans and glycosaminoglycans ...... 14 Heparan sulfate proteoglycans...... 15 Chondroitin/dermatan sulfate proteoglycans ...... 16 Heparan sulfate biosynthesis...... 16 Chondroitin/dermatan sulfate biosynthesis...... 19 …and one to rule them all: PAPS...... 21 Regulation of the glycosaminoglycans biosynthesis – a puzzle not yet solved...... 22 Glycosaminoglycans in biological processes...... 23 Zebrafish as a model system...... 28 Zebrafish in glycosaminoglycan research...... 30 Method considerations...... 32 Overexpression studies ...... 32 Rise and fall of the morpholino technology...... 32 Selected methods for targeted genome editing in zebrafish ...... 33 Present investigation ...... 36 Aim ...... 36 Paper I ...... 36 Paper II...... 37 Paper III ...... 38 Paper IV ...... 39 Concluding remarks and future perspectives ...... 41 Populärvetenskaplig sammanfattning ...... 43 Acknowledgments...... 44 References ...... 46

Abbreviations

ADP Adenosine diphosphate APS Adenosine 5’-phosphosulfate ATP Adenosine triphosphate b3gat3 Glucuronyltransferase I BMP Bone morphogenetic protein C4ST Chondroitin sulfate 4-O-sulfotransferase C6ST Chondroitin sulfate 6-O-sulfotransferase 1 CS/DS2ST Chondroitin sulfate/dermatan sulfate 2-O-sulfotransferase CHPF/CHSY2 Chondroitin polymerization factor CHPF2 Chondroitin polymerization factor 2 CHSY1 Chondroitin synthase 1 CHSY3 Chondroitin synthase 3 CS Chondroitin sulfate CSGALNACT, GalNAcT Chondroitin sulfate N-acetyl galactosaminyl transferase DS Dermatan sulfate EXTL Exostosin-like FGF Fibroblast growth factor GAG Glycosaminoglycan Gal Galactose GalNAc N-Acetylgalactosamine GlcA Glucuronc acid GlcNAc N-Acetylglucosamine HME Hereditary multiple exostoses HS Heparan sulfate IdoA Iduronic acid KS Keratan sulfate MO Morpholino NDST N-Deacetylase/N-sulotransferase p53 Protein 53 (Tumor protein 53) PAPS 3’-Phosphoadenosine 5’-phosphosulfate PAPSS PAPS synthase PAPST PAPS transporter TGF Transforming growth factor uxs1 UDP-glucuronic acid decarboxylase 1 Wnt Wingless-related MMTV integration site Xyl Xylose Xylt1 Xylosyltransferase 1

Introduction

Glycosaminoglycans (GAGs), located in the extracellular matrix and on the cell surface, have a fundamental role in many aspects of sustaining vertebrate life. Due to their negative charge, they have an ability to interact with a large number of proteins, including matrix components, morphogens and signaling molecules, modifying their functions and availability. Although knowledge about these extended carbohydrate polymers is not as great as that of DNA and protein, the crucial role of GAGs in biological systems gives them a place among the most fascinating biomolecules. This work is focused on two types of glycosaminoglycans, heparan sulfate and chondroitin/dermatan sulfate, the complex enzymatic machinery needed for their biosynthesis, and the functions they play in early development. All investigations included in this thesis are based on studies of the zebrafish model.

13 Background

Proteoglycans and glycosaminoglycans Linear, sulfated and therefore negatively charged GAGs are produced by virtually all vertebrate cells. Sulfated GAG chains, covalently attached to certain serine residues within so called core proteins, form proteoglycans (PGs). Depending on their structure and degree of sulfation, GAGs can be further divided into heparan sulfate (HS), heparin, chondroitin sulfate (CS), dermatan sulfate (DS) and keratan sulfate (KS). Hyaluronan (HA), differs from the other GAG family members, as it consist of a single non-sulfated and core protein-free GAG chain (Fig. 1).

Figure 1. Glycosaminoglycan disaccharide structures. Heparin is a highly modified and sulfated version of heparan sulfate. Positions that may be sulfated are marked in red, except for the fully modified heparin disaccharide structure.

GAG-resembling polysaccharides, lacking core proteins, have been identified in some bacterial strains where their function seems to be connected to infection processes. In the animal kingdom, GAGs in the form of CS and HS are present already in the platyhelminthes (Yamada, Sugahara et al. 2011). PGs are present mostly on the cell surface and in the extracellular matrix, but also inside the storage granules of cells of hematopoietic origin and, tentatively, inside the nuclei of some cells (Kjellen and Lindahl 1991;

14 Richardson, Trinkaus-Randall et al. 2001). Some PGs are present in a wide range of tissues, whereas other members, such as neurocan and serglycin appear to be, tissue or cell type specific, respectively (Sugahara and Mikami 2007; Esko, Kimata et al. 2009). The various biological functions of PGs are possible to maintain due to their high diversity. While the largest PG, aggrecan, contains more than hundred CS chains, HSPGs at the cell surface often display 3-5 GAG chains. Complex sulfation patterns of the GAG molecules allow interactions with a range of cell surface and matrix proteins. Some interactions may be based on more specific sulfation motifs with high binding affinities, whereas, for others, only a certain degree of sulfation seems to be required (Gama, Tully et al. 2006).

Heparan sulfate proteoglycans The basic building blocks of HS are N-acetyl-D-glucosamine (GlcNAc) and glucuronic acid (GlcA) units. Similar to CS/DS, HS chains are linked to the core protein through the GlcA-Gal-Gal-Xyl tetrasaccharide sequence. HS can be further modified by epimerization of a portion of the GlcA to iduronic acid (IdoA), and sulfation of both glucosamine and IdoA at specific positions. Within HS chains, domains with various degree of sulfation can be distinguished. N-acetylated domains (NA-domains) are virtually unmodified and NS-domains, in most cases 4-12 residues long, are characterized by fully N-sulfated GlcN residues (Turnbull and Gallagher 1991; Gallagher, Turnbull et al. 1992). Epimerization, 3-O-, 6-O- and 2-O-sulfation are also typical for NS-domains, whereas within NA/NS-domains, modifications are present with lower density (Westling and Lindahl 2002). HSPGs can be divided into four main groups, including the two types of plasma membrane bound PGs syndecans and glypicans. Perlecan, agrin and collagen XVIII form another group of PGs. Members of this group are secreted and participate in extracellular matrix and basement membrane organization (Iozzo 1998). Serglycin, carrying heparin, and in some cases CS chains, is stored in the mast cell intracellular granules (Fig.2). In the extracellular matrix, HSPGs can bind growth factors and store morphogens designed for further use and membrane bound, they can serve as co-receptors, modulate both signaling events and cell adhesion (Bernfield, Gotte et al. 1999). Interestingly, HS chain composition and domain organization have been shown to be tissue specific, despite the lack of template for their biosynthesis (Ledin, Staatz et al. 2004).

15

Figure 2. Examples of PGs. PGs consist of a core protein (dark gray) and one ore more covalently attached GAG chain: CS (black), HS (gray) or heparin (light grey). The membrane bound proteoglycans syndecans, may contain both CS and HS chains. Serglycin with attached heparin and in some cases also CS chains is stored in granules of hematopoietic cells.

Chondroitin/dermatan sulfate proteoglycans CS/DSPGs are abundant in the extracellular matrix and are essential both for cartilage development and functionality. They are important players of various processes, including inflammatory response, neuronal development and cellular proliferation (Prabhakar and Sasisekharan 2006). Most of the functions of CSPGs are believed to involve the CS chains, with core proteins serving mostly as a scaffold (Sugahara and Kitagawa 2000; Sugahara, Mikami et al. 2003). CS chains are formed by 40, to more than 100 linearly aligned disaccharide units of glucuronic acid (GlcA) and N-acetyl-D-galactosamine (GalNAc). In DS chains, some of the GlcA residues are epimerized to IdoA (Prabhakar and Sasisekharan 2006). Sulfation of CS includes 4-O-sulfation and/or 6-O-sulfation of the GalNAc and occasional 2-O-sulfation or 3-O- sulfation of the GlcA residue. DS sulfation patterns can involve 4-O- sulfation of GalNAc residue, together with 6-O-sulfation of GalNAc and 2- O-sulfation of the IdoA (Sugahara, Mikami et al. 2003). Aggrecan contains CS side-chains, whereas DS is more common in versican and the small leucine-rich PGs (Iozzo 1999).

Heparan sulfate biosynthesis CS/DS and HS GAGs share a common linkage region (GlcAβ1-3Galβ1- 3Galβ1-4Xylβ1-O-Ser) composed of four monosaccharides and covalently

16 attached to serine residues of the core proteins (Fig. 1,3) (Sugahara and Kitagawa 2000; Sugahara and Kitagawa 2002). In some cases, when the same serine residue can carry either type of GAG chains, addition of the first residue to the linkage region will determine the type of synthesized GAG chain. This information seems to be core protein dependent and the reaction itself is accomplished by either CSGALNACT1/CSGALNACT2 resulting in CS biosynthesis, or EXTL3 resulting in HS biosynthesis (Lindahl and Jin-ping 2009). Although modifications (phosphorylation and sulfation) of the GAG linkage region have been proposed to influence the choice between HS or CS attachment, the exact mechanisms stay unclear (Prydz 2015).

Figure 3. Schematic outline of heparan sulfate and chondroitin/dermatan sulfate biosynthesis. Enzymes with sulfotransferase activity, marked with red lines, are dependent on the availability of 3’-phosphoadenosine 5’-phosphosulfate (PAPS). Single mutations in the PAPS synthase or transporter can therefore affect the activity of multiple enzymes. Adapted from (Häcker, Nybakken et al. 2005).

In mammals, the EXT1/EXT2 copolymerase is proposed to play the major role in the further HS chain polymerization, although, as many as five glycosyltransferases belonging to the exostosin family (EXT) have been reported (Fig. 3). The mutations within EXT1 and EXT2 genes are linked to

17 Hereditary Multiple Exostoses (HME). The remaining members of this family, EXTL1, EXTL2 and EXTL3, were identified based on sequence homology to the HME linked genes, and are also proposed to have roles in HS biosynthesis (Busse-Wicher, Wicher et al. 2014). Further modifications of the HS chain, including GlcA epimerization to IduA, and O-sulfation at various sites, occur in the N-sulfated regions (Lindahl, Kusche-Gullberg et al. 1998). Epimerization is carried out by the single glucuronyl C5-epimarase (GLCE). Iduronic acid residues, resulting from GLCE action, can be further 2-O-sulfated by the single 2-O- sulfotransferase. The 2-OST enzyme acts predominantly on IdoA residues, however 2-O-sulfation of GlcA units can also occur, but at much lower rates (Rong, Habuchi et al. 2001; Li, Gong et al. 2003). As many as three different 6-O-sulfotransferases and seven 3-O- sulfotransferases have been reported. 6-OST1 is the primary 6-O- sulfotransferase in the majority of the tissues (Habuchi, Nagai et al. 2007). Interestingly, the 3-O-sulfotransferases, the largest family among the HS modifying enzymes, are responsible for the least frequently occurring sulfation type (Thacker, Xu et al. 2014).

NDSTs In HS, 40-50%, and in heparin, 80-90%, of the glucosamine residues are N- sulfated. In mammals, the N-sulfation process is catalyzed by the four bifunctional N-deacetylase/N-sulotransferase (NDST1-4) enzymes, with the use of 3’-phosphoadenosine 5’-phosphosulfate (PAPS) as sulfate donor (Kusche-Gullberg, Eriksson et al. 1998; Aikawa and Esko 1999; Aikawa, Grobe et al. 2001). NDST is the first enzyme modifying the HS chain by removing acetyl groups and then adding sulfate groups to the previously deacetylated glucosamine residues. The NDSTs play a crucial role during the HS modification process, since most of the sulfate groups added by O- sulfotransferases will be located to residues in the N-sulfated regions of the HS chain where also epimerization occurs (Holmborn, Ledin et al. 2004). The four isoforms differ in their enzymatic activities. While murine recombinant NDST1, 2 and 3 have relatively high N-deacetylase activities, NDST4 seems to be primarily an N-sulfotransferase. In contrast, the N- sulfotransferase activty of NDST3 is very low, whereas the three other NDSTs have more comparable activities (Aikawa, Grobe et al. 2001). Both Drosophila and C. elegans express a single ndst gene. The Drosophila Ndst mutant sulfateless (sfl) displays a phenotype manifested as impaired segmentation during embryonic development, caused by production of unmodified HS chains, unable to fulfill their regular function (Lin, Buff et al. 1999). In mice and humans, NDST1 and NDST2 are expressed at different levels from early embryonic stages and through adulthood. The NDST3 and NDST4 isoforms on the other hand, are

18 expressed mostly during embryonic development and later in adult brain tissues (Grobe, Ledin et al. 2002). NDST1 null-mutant mice die prenatally or shortly after birth. They display lung and craniofacial defects which make their phenotype the most severe of all investigated NDST knockouts (Ringvall and Kjellen 2010). Interestingly, the phenotype of the NDST2-/- mouse is very mild. The lack of functional NDST2 results in reduced number of mast cells containing large, empty vacuoles, in contrast to wild type mast cells, where granules are fully packed with proteoglycans and inflammatory mediators (Grobe, Ledin et al. 2002). NDST3-/- mice are characterized by a minor undersulfation of HS and phenotypically, by small behavioral and hematological defects, suggesting that other isoforms might compensate for the lack of NDST3 (Pallerla, Lawrence et al. 2008). In a recent study, Reuter et. al have described patients carrying NDST1 missense mutations of conserved amino acids within the sulfotransferase domain, leading to changes in the substrate binding abilities and/or structural changes of this domain. These changes altering protein conformations, have been linked to several clinical features, ranging from intellectual disability, epilepsy, muscular hypotomia to deficiency in postnatal growth (Reuter, Musante et al. 2014). Additionally, point mutations within the regulatory region of NDST3 have been correlated with schizophrenia and bipolar disorder. Postmortem cerebellar tissues from patients carrying risk allele were found to express higher levels of NDST3 as compared with heterozygotes (Lencz, Guha et al. 2013).

Chondroitin/dermatan sulfate biosynthesis Similar to HS, CS/DS chain polymerization and modification take place in the Golgi apparatus. After the initial step of GalNAc transfer to the linkage region, further CS chain elongation is catalyzed by CS polymerases (see below). In mammals, modification of the CS chain includes sulfation of the GalNAc residue by the three chondroitin sulfate 4-O-sulfotransferases (C4ST1-3) and chondroitin sulfate 6-O-sulfotransferases 1 and 2 (C6ST1-2). As in HS biosynthesis, selected GlcA residues can undergo epimerization into IdoA as a result of the activity of the two DS epimerases (Pacheco, Malmstrom et al. 2009). The IdoA-GalNAc disaccharides can be further sulfated at the C-4 position of the GalNAc residue, carried out by the dermatan 4-O-sulfotransferase (Evers, Xia et al. 2001). CS/DS chain sulfation is considered to proceed at the same time as the chain undergoes elongation. It involves 2-O-sulfotransferase action, leading to C-2 GlcA/IdoA sulfation of the previously 4-O-sulfated or 6-O-sulfated disaccharides. Finally, the C-6 position of a previously C-4 sulfated GalNAc residue can become sulfated by the GalNAc4S-6-O-sulfotransferase (Prabhakar and Sasisekharan 2006). A

19 majority of the polymers obtained as a result of the elongation and modification processes consists of mixed GlcA and IdoA residues, and therefore is often referred to as CS/DS (Fig. 3, Tab. 1).

Chondroitin/dermatan sulfate glycosyltransferases Six CS glycosyltransferases have been identified until now. Each of them have been known under several names, all listed here; chondroitin sulfate N- acetylgalactosaminyltransferase-1 (GalNAcT1,CSGALNACT1) (Gotoh, Sato et al. 2002), chondroitin sulfate N-acetylgalactosaminyltransferase-2 (GalNAcT2, CSGALNACT2) (Sato, Gotoh et al. 2003), chondroitin synthase 1 (CHSY1, ChSy, ChSy-1) (Kitagawa, Uyama et al. 2001), chondroitin synthase 3 (CHSY3, CSS3, ChSy-2), chondroitin polymerization factor (CHPF) (Kitagawa, Izumikawa et al. 2003) and chondroitin polymerization factor 2 (CHPF2, CSGlcAT, ChSy-3) (Izumikawa, Koike et al. 2008). The specific role of each of these enzymes is still under debate, although some information regarding their activities is available. Although the CHSY1, CHPF2 and CHSY3 enzymes have been shown to possess both GlcAT-II and GalNAcT-II activities, it was also demonstrated that only co- expression of two synthases, or any of the synthases together with one of the polymerization factors, can lead to CS polymerization in vitro (Kitagawa, Izumikawa et al. 2003; Izumikawa, Uyama et al. 2007; Izumikawa, Koike et al. 2008). The remaining two enzymes, CSGALNACT1 and CSGALNACT2, are thought to be the main enzymes responsible for both CS chain initiation and elongation (Sato, Gotoh et al. 2003). The importance of CSPG glycosyltransferases have been shown in several models, starting from C. elegans, producing mainly the non-sulfated form of CS. Animals with a defect in the sqv-5 gene, the nematode ortholog of CHSY, exhibited defective vulva formation (Hwang, Olson et al. 2003). The same phenotype was observed in C. elegans with a mutation in mig-22 (pfc- 1), the nematode CHPF ortholog. Mutation in the human CHSY1 gene was reported to cause short posture and skeletal defects including brachydactyly and hearing loss (Tian, Ling et al. 2010). Zebrafish morpholino knockdown of Chsy1 results in defective development of the jaw, pectoral fins and eyes (Tian, Ling et al. 2010), whereas mouse with null-mutations in CHSY1 were reported to exhibit deformations of the joints and digits (Wilson, Phamluong et al. 2012). Taken together, this data would suggest that CHSY1 plays important roles in the skeletal development including digit patterning. Also CSGALNACT1 knockdown mice exhibit a cartilage phenotype, including shortened axial skeleton and limbs, as well as cartilage abnormalities. In these mice the CS content in cartilage is reduced to less than half of the wild type amount (Watanabe, Takeuchi et al. 2010). Also

20 human missense mutations in CSGALNACT1 have been linked to hereditary motor and sensory neuropathies (Saigoh, Izumikawa et al. 2011).

…and one to rule them all: PAPS 3’-phosphoadenosine 5’-phosphosulfate (PAPS) is synthesized in a two-step reaction carried out by PAPS syntase in the cell cytosol. PAPS is an universal sulfate donor for all sulfotransferase reactions both in the cytosol (estrogen, steroid hormone, neurotransmitter, drug and xenobiotic sulfation) and in the Golgi apparatus (tyrosine and GAG sulfation) (Besset, Vincourt et al. 2000; Venkatachalam 2003) Therefore all sulfation reactions are dependent on PAPS availability (Esko and Selleck 2002; Bishop, Schuksz et al. 2007; Nadanaka and Kitagawa 2008). Additionally, in vitro studies have shown that PAPS concentration during HS biosynthesis can affect NDST enzyme performance (Carlsson, Presto et al. 2008). Animal PAPS synthase combines ATP sulfurylase and APS kinase activity in a single protein, whereas in bacteria, fungi, yeast and plants these enzyme activities are found separately in two polypeptides chains. ATP sulfurylase catalyzes the reaction in which inorganic sulfate is combined with ATP to form APS. In the second step of the reaction, PAPS is formed from APS through phopshorylation by APS kinase (Besset, Vincourt et al. 2000). In D. melanogaster and C. elegans, only one PAPS synthase is present, whereas in human and mouse two PAPSS isoforms, PAPSS1 and PAPSS2, have been identified (Jullien, Crozatier et al. 1997; Dejima, Seko et al. 2006). The zebrafish genome contains a single papss1 gene and two gene copies encoding papss2a and papss2b. Interestingly, in vitro studies revealed nuclear localization of PAPSS1 suggesting that it has a role in the nucleus, whereas PAPSS2b was found to be located in the cytoplasm. It has also been suggested that PAPSS1 and PAPSS2 can form high-affinity heterodimers, allowing transport of PAPSS2 into the cell nucleus (Besset, Vincourt et al. 2000). However both PAPSS1 and PAPSS2 contain conserved nuclear localization motifs, the one present in PAPSS1 is presumed to be more potent. This finding is not in agreement with the common opinion that PAPS synthesis is exclusively localized to the cytosol, but could explain why these two isoforms do not appear to compensate for each other (Besset, Vincourt et al. 2000; Schroder, Gebel et al. 2012). PAPSS1 is ubiquitously expressed in adult human tissues, whereas PAPSS2 appears to be mainly expressed in the cartilage growth plate (Fuda, Shimizu et al. 2002). Murine Papss2 expression was observed in elements of forming cartilage, heart, tongue, kidney and neuronal tissue. Brachymorphic mice, identified as PAPSS2 mutants, are characterized by

21 dwarfism and undersulfation of chondroitin sulfate localized in the extracellular matrix (Orkin, Pratt et al. 1976; Stelzer, Brimmer et al. 2007). PAPS needed in the Golgi apparatus for sulfation of glycoproteins, proteoglycans and glycolipids is transported from the cytoplasm by 3’- phosphoadenosine 5’-phosphosulfate transporters (PAPST). Two of these specific PAPS transporters have been identified in vertebrates including zebrafish. Both are localized to the Golgi membrane (Kamiyama, Suda et al. 2003; Kamiyama, Sasaki et al. 2006). The zebrafish papst1 mutant (pinscher/slc35b2) has been identified based on its pectoral fin, cartilage and axon miss-sorting phenotype. Pinscher mutants have been reported with greatly reduced amounts of total GAGs (Wiweger, Avramut et al. 2011).

Regulation of the glycosaminoglycans biosynthesis – a puzzle not yet solved. One of the most challenging questions in the GAG field is the biosynthesis regulation. Assembly and modification of the GAG chains is carried out by a large set of enzymes. Spatiotemporally regulated expression, makes them capable of producing organ-specific, or presumably, even cell-specific GAGs epitopes (Ledin, Staatz et al. 2004). This finding suggests that similar to the genome and proteome, also the glycome might be tightly regulated. In spite of the intensive hunt for a model, which could explain the formation of the GAG fine structure, their specific “sulfation fingerprint” and chain length, this field still leaves a lot to be discovered (Little, Ballinger et al. 2008; Victor, Nguyen et al. 2009). The existence of a “GAGosome”, a set of Golgi membrane-bound enzymes arranged into a GAG biosynthetic machinery, is a popular theory within the proteoglycan biosynthesis field (Esko and Selleck 2002). Several studies supporting this hypothesis for HS biosynthesis have been published. Data presenting interactions between extostosin 1 (EXT1) and exostosin 2 (EXT2), C5 epimerase and 2-O-sulfotransferase and finally N- deacetylase N-sulfotransferase 1 and EXT2 are available (Kobayashi, Morimoto et al. 2000; McCormick, Duncan et al. 2000; Pinhal, Smith et al. 2001; Presto, Thuveson et al. 2008). An additional study from 1974 reports on the interaction between the linkage region enzymes, xylosyltransferase and galactosyltransferase-1 (Schwartz, Roden et al. 1974). Catalytic activity of the enzymes involved in the GAG biosynthetic pathway can be restricted by their substrate preferences. Experimental evidence suggests that HS modifications are tightly organized, where epimerization occurs after N-sulfation (Esko and Selleck 2002; Carlsson, Presto et al. 2008) and is followed by 2-O-sulfation, as N-sulfated and epimerized substrate tends to be favored by the 2OST (Rong, Habuchi et al. 2001). Also 6OSTs and 3OSTs have certain substrate preferences (Habuchi,

22 Tanaka et al. 2000). Sulfation by the 3OSTs is considered to be the last modification and 3OSTs preferentially act on previously modified residues, where 2-O-sulfation on the non-reducing side of the target glucosamine can be either inhibitory or preferred by different isoforms (Thacker, Xu et al. 2014). Although these are only a few examples of the substrate specificities, they exemplify how the diversity of GAG structures could be obtained. Tight spatiotemporal regulation of the enzyme expression, their preferences towards specific substrate, also in complexes with other enzymes, can tentatively explain the occurrence of tissue-specific “sulfation fingerprints”. Also, availability of PAPS and UDP-sugars, translational levels of GAG biosynthetic enzymes, posttranslational modifications of the enzymes and finally desulfation of HS by the extracellular endosulfatases capable of remodeling HS chains after biosynthesis, play roles in shaping the final GAG structure (Kreuger and Kjellen 2012).

Glycosaminoglycans in biological processes A growing number of GAG related studies reveals the importance of GAGs in a broad range of biological processes. GAGs are known to be involved in cell adhesion, lipid metabolism, angiogenesis and inflammatory responses (Bishop, Schuksz et al. 2007). In some cases, only particular modifications of the GAG oligosaccharide sequences can support binding between a GAG and its ligand, such as is case for the heparin/HS pentasaccharide interaction with antithrombin. For other interactions, the specificity is less obvious, with several possible oligasaccharides supporting a certain interaction. The best- studied example of this kind of GAG-protein interaction, is fibroblast growth factor (FGF) and FGF receptor binding with HS/heparin (Kreuger, Spillmann et al. 2006). Morphogen gradient formation is another intriguing role of GAGs. Binding to HS can protect proteins from degradation, but also regulate their availability. Growth factors such as Wnt, Hedgehog and BMPs are responsible for tissue patterning during development and HSPGs have been shown to bind them, restrict their diffusion, and therefore form morphogen gradients (Lander, Nie et al. 2002; Tabata and Takei 2004; Yan and Lin 2009). Additionally, HS can create signaling platforms where extended HS chains interact with cell-surface receptors and signaling molecules. Examples include FGF2 and its receptor, or BMP2 and BMP4 signal transduction (Quarto and Amalric 1994; Yan and Lin 2009; Kuo, Digman et al. 2010). Finally, GAGs have been linked to a number of pathological states. Table 1 contains a list of HS and CS/DS biosynthetic enzymes with associated human disorders, phenotypes of mouse and zebrafish mutants and zebrafish morpholino knock-downs.

23 Table 1. List of enzymes relevant for HS and CS/DS biosynthesis and known disorders connected to each of the genes in human, mouse and zebrafish. OMIM numbers represent entries listed at the Online Mendeian Inheritance in Men database.

Human (H) disease and Zebrafish gene(s), mutations Enzyme name Abbreviatio OMIM #, (M) mouse mutant (-/-), morpholino knock-down n phenotype. (MO) phenotypes 3’-phosphoadenosine PAPSS1 papss1 5’-phosphosulfate PAPSS2 H: Spondyloepimetaphyseal papss2a synthase 1 dysplasia, Pakistani type; #612847 M: Brachymorphic mouse; papss2b dwarfism, cartilage defects. (Lane and Dickie 1968) 3’-phosphoadenosine PAPST1 papst1/ slc35b2-/-: Cartilage 5’-phosphosulfate defects. (Clement, Wiweger et transporter 1 al. 2008) PAPST2 papst2/ slc35b3 UDP-xylose synthase UXS1 M: Prenatal death. uxs1-/-: Craniofacial cartilage 1 // (Blake, Bult et al. 2014) and bone defects, shorter UDP-glucuronate pectoral fins. (Eames, Singer et decarboxylase 1 al. 2010) Glycosaminoglycan FAM20B H: Point mutation associated fam20b-/-: Cartilage defects. xylosylkinase with the risk of psoriasis (Eames, Yan et al. 2011) (Yin, Cheng et al. 2014) 2-phosphoxylose PXYLP1 pxylp1 phosphatase Xylosyltransferase XYLT1 H: Pseudoxanthoma xylt1-/-: Cartilage defects. elasticum: #264800 (Eames, Yan et al. 2011) XYLT2 H: Pseudoxanthoma xylt2 elasticum; #264800 M: Polycystic liver and kidney disease at the age of 4-5 months. (Condac, Silasi- Mansat et al. 2007) Galactosyltransferase B3GALT7 H: Ehlers-Danlos syndrome, b4galt7-/-: Cartilage defects. progeroid type1; #130070 (Amsterdam, Nissen et al. 2004; Nissen, Amsterdam et al. 2006) B3GALT6 H: Spondyloepimetaphyseal b3galt6 dysplasia; #271640; Ehlers-Danlos syndrome progeroid type2; #615349 Glucuronyltransferase B3GAT3 H: Multiple joint b3gat3-/-; Craniofacial cartilage dislocations, short stature, bone defects, shorter pectoral craniofacial dysmorphism, fins. (Amsterdam, Nissen et al. and congenital heart defects; 2004; Wiweger, Avramut et al. #245600 2011) Exostosin EXT1 H: Chondrosarcoma, ext1a glycosyltransferase Hereditary multiple exostoses type1; #133700 ext1b M: Gastrulation failure, ext1c death at E7.5. (Lin, Wei et al. 2000)

24 EXT2 H: Hereditary multiple ext2-/-: Defects in craniofacial exostoses type2; #133701. cartilage, axon sorting, lack of Seizures-scoliosis- pectoral fins. (Lee, von der macrocephaly syndrome Hardt et al. 2004; Clement, (Farhan, Wang et al. 2015). Wiweger et al. 2008; M: Mutants – gastrulation Holmborn, Habicher et al. failure, death at E6-7.5. 2012) Heterozygous adults have skeleton abnormalities. (Stickens, Zak et al. 2005) Exostosin-like EXTL1 extl1 glycosyltransferase EXTL2 M: Normal development with extl2 impaired liver regeneration. (Nadanaka, Kagiyama et al. 2013; Nadanaka, Zhou et al. 2013; Nadanaka and Kitagawa 2014) EXTL3 M: Embryonic lethal at E9. extl3-/-: Craniofacial cartilage (Takahashi, Noguchi et al. defects, shorter pectoral fins. 2009) (Schilling, Piotrowski et al. 1996; Lee, von der Hardt et al. 2004; Clement, Wiweger et al. 2008) N-deacetylase/N- NDST1 H: Mental retardation; ndst1a MO: Aberrant blood sulfotransferase #616116 circulation and vessel M: Prenatal death – lung, development. brain and craniofacial (Harfouche, Hentschel et al. defects. 2009) (Fan, Xiao et al. 2000; ndst1b MO: Aberrant Ringvall, Ledin et al. 2000; craniofacial skeleton/ pectoral Grobe, Inatani et al. 2005) fins. (Filipek-Gorniok, Carlsson et al. 2015) NDST2 M: Reduced number of mast ndst2a cell with empty storage ndst2b vacuoles. (Forsberg, Pejler et al. 1999; Humphries, Wong et al. 1999) NDST3 H: Mutation in the regulatory ndst3 region, linked to schizophrenia and bipolar disorder. (Lencz, Guha et al. 2013) M: Mild behavioral and hematological changes. (Pallerla, Lawrence et al. 2008) NDST4 - Glucuronyl C5- Hsepi/ M: Lack of kidneys, lung glcea (MO): Expansion of epimerase GLCE failure, skeletal defects with ventral structures during aberrant chondrocyte development. (Ghiselli and proliferation (Li, Gong et al. Farber 2005) 2003; Dierker, Bachvarova et glceb (MO): Expansion of al. 2015) ventral structures during development. (Ghiselli and Farber 2005)

25 Hexuronyl 2-O- HS2ST M: Renal agenesis, over- hs2st1a (MO): Epiboly sulfotransferase mineralized skeletons, initiation failure. (Cadwalader, retardation of eye Condic et al. 2012) development, neonatal death. hs2st1b (Bullock, Fletcher et al. 1998) Glucosaminyl 6-O- HS6ST1 H: Hypogonadotropic hs6st1a sulfotransferase hypogonadism; #614880 hs6st1b M: Increased embryonic lethality and decreased growth (Habuchi, Nagai et al. 2007) HS6ST2 hs6st2 (MO): Abnormal vascular development (Chen, Stringer et al. 2005) HS6ST3 hs6st3a hs6st3b Glucosaminyl 3-O- HS3ST1 hs3st1 sulfortansferase hs3st1l1/hs3st7 (MO): Loss of the ventrical contraction (Samson, Ferrer et al. 2013) hs3st1l2 HS3ST2 hs3st2 HS3ST3a - HS3ST3b hs3st3x/ hs3st3b1a hs3st3z/ hs3st3b1a HS3ST4 hs3st4 HS3ST5 hs3st5 HS3ST6 hs3st3l Chondroitin Synthase CHSY1 H: Temtamy preaxial chsy1 (MO): Atrioventricular brachydactyly syndrome; # abnormalities (Peal, Burns et 605282 al. 2009); Impaired skeletal, M: Limb patterning and pectoral fin and retinal skeletal defects including development (Tian, Ling et al. chondrodysplasia (Wilson, 2010) Phamluong et al. 2012) CHSY3 chsy3 Chondroitin CHPF/ M: Normal. (Wilson, chpfa Polymerizing factor CHSY2 Phamluong et al. 2012) chpfb CHPF2 chpf2 N-acteylgalactos- CSGALNA H: A heterozygous mutation csgalnact1a (MO): In aminyltransferase CT1/ may be associated with combination with csgalnact2 GALNT1 peripheral neuropathies. MO decreased CS and (Saigoh, Izumikawa et al. increased HS amounts. 2011) (Holmborn, Habicher et al. M: Affected blood 2012) morphology. (Block, Ley et csgalnact1b al. 2012) CSGALNA csgalnact2 (MO): In CT2/ combination with csgalnact1a GALNT2 MO; decreased CS, increased HS amounts. (Holmborn, Habicher et al. 2012)

26 Dermatan sulfate DSE H: Ehlers-Danlos syndrome, dse epimerase Musculocontractural type 2; #615539 M: Viable, smaller, with no changes in major organs, alterations seen in the skin. (Maccarana, Kalamajski et al. 2009) Dermatan sulfate DSEL dsela epimerase-like dselb Carbohydrate C6ST1 H: Spondyloepiphyseal chst3a (chondroitin 6) dysplasia with congenital chst3b sulfotransferase 3 joint dislocation; #603799 M: Reduced cell number in the lymph nodes, reduced fertility (Orr, Le et al. 2013) Carbohydrate (N- C6ST2 chst7 acetylglucosamine 6- O) sulfotransferase 7 Carbohydrate C4ST1/ M: Severe dwarfism, death at chst11 (MO): Bent trunk, (chondroitin 4) C4ST11 6h after birth from curled and/or kinky tail, sulfotransferase respiratory distress. Bone and impaired muscle development cartilage defects. (Kluppel, and axon guidance. Wight et al. 2005) (Mizumoto, Mikami et al. 2009) C4ST2/ chst12a C4ST12 chst12b C4ST3/ chst13 C4ST13 Carbohydrate (N- D4ST1 H: Ehlers-Danlos syndrome, chst14 acetylgalactosamine musculocontractural type 1; 4-0) sulfotransferase #608429 14 Carbohydrate CHST15 chst15 (chondroitin 4) sulfotransferase 15 Uronyl-2- 2OST/ ust sulfotransferase UST

Notably, many of the listed human disorders involve either muscle, skeletal or neuronal defects. The extracellular matrix content of cartilage is close to 90% (dry weight), and is known to be particularly rich in PGs. These are known to greatly contribute to cartilage biomechanical properties, extracellular matrix water retention, which together encounters for the resistant, yet elastic load-bearing structure. The collagen-aggrecan-hyaluronan network is of a special importance, providing cartilage with osmotic properties, making it pressure resistant and giving it a turbid nature (Roughley and Lee 1994; Knudson and Knudson 2001). Agrrecans can form aggregates with hyaluronan, further stabilized by the cartilage link proteins (Iozzo 1998). Other proteoglycans like decorin,

27 biglycan, fibromodulin and perlecan are commonly found components of the cartilage, responsible for tissue integrity and metabolism (Roughley and Lee 1994; Knudson and Knudson 2001). The main signaling molecules involved in bone formation belong to growth factors families, including BMP, IGF, TGF-β, FGF, PDGH, EGF and Wnts. These growth factors participate in bone and cartilage growth and repair processes, and a majority of them is known to bind to HS (Bernfield, Gotte et al. 1999). As already described, knockout mice for either CHSY1, CHPF or CSGALNACT1, although viable and fertile, are characterized by reduced CS content, and display chondrodysplasia and digit-patterning defects. Also the C6ST-1 mutation in humans and the PAPS synthase mutation in mouse, lead to severe chondroysplasia and brachymorphism, respectively. These results may suggest that to fully support cartilage functionality, a specific GAG sulfation pattern is needed (Orkin, Pratt et al. 1976; Yada, Gotoh et al. 2003; Watanabe, Takeuchi et al. 2010; Wilson, Phamluong et al. 2012). Unlike cartilage, brain ECM tissue is presumed to be free from fibronectin and collagen, frequently occurring in other organs. In contrast, hyaluronan, HSPGs (including glypican and syndecan) and CS proteoglycans like versican, brevican and neurocan are present in large amounts in the intracellular spaces between neurons and glia (Bandtlow and Zimmermann 2000; Bonneh-Barkay and Wiley 2009). Either bound to the cell surface or as components of the ECM, proteoglycans can influence neuronal growth, migration and synaptic remodeling. CS proteoglycans seem to be of special importance also during development of the central nervous system, playing roles in the formation of the neuronal boundaries and limiting synaptic plasticity (Dyck and Karimi- Abdolrezaee 2015). Zebrafish, described in the chapter below, is one of the popular animal models, suitable for studies of the formation of cartilage and the nervous system patterning.

Zebrafish as a model system Zebrafish (Danio rerio), a teleost tropical freshwater fish, is a valuable vertebrate model organism used in biological research. Adult individuals can grow up to 6 cm (4 cm in captivity) and owe their name to five blue and silver vertical stripes present on the side of their body. Embryo fertilization and further development proceed externally in a transparent chorion, enabling their observation (Fig. 4). Furthermore, to facilitate imaging, pigmentation, normally occurring from about 24 hours post fertilization (hpf), can be chemically inhibited.

28

Figure 4. Early zebrafish development. A. Zygote period (0-0.75 hpf): the chorion is separated from the fertilized egg cell and the non-yolk cytoplasm undergoes movements towards the animal pole, B. Cleavage period (0.75-2.25 hpf): the cells undergo six cell divisions in 15 min intervals, reaching a total number of 64 cells, C. Blastula period (2.025-5.25 hpf): the epiboly process begins. The yolk syncytial layer forms and midblastula transition starts, D. Gastrula period (5.25-10 hpf): when as a result of involution, convergence and extension movements, primary germ layers and embryonic axis arise, E. Segmentation period (10-24 hpf): first body movements occur, groups of cells start to differentiate, formation of the primary organs is initiated. Somites develop starting from the anterior end of the embryo, continuing towards its posterior end, F. Pharyngula period (24-48 hpf): the embryo is organized bilaterally, somite segmentation and notochord formation is finished. Development of the pharyngeal arches, hatching gland, pigment cells, fins, heart with the first heart beats and vasculature take place and the nervous system is composed of five lobes, G. Hatching period (48-72 hpf): embryos will hatch, pectoral fins, jaw and gills develop from the first two and posterior five pharyngeal arches, respectively. H. Early larval period (3-7 days): the swimming bladder and the mouth appear. The larvae begin to actively swim and feed. Adapted with permission from Charles Kimmel. Reprinted with permission from (Kimmel, Ballard et al. 1995; Nusslein-Volhard and Dahm 2005), John Wiley and Sons.

The zebrafish genome is comprised of 25 chromosomes. For approximately 20% of all zebrafish genes, two homologues of a single mammalian gene are present (Howe, Clark et al. 2013). These gene paralogs arose as a result of a whole genome duplication which took place early in the evolution, most likely before the origin of teleosts, in the ray-fin fish lineage. Investigations carried out on paralogs in zebrafish suggest that ancestral gene function may be divided between the duplicated genes as both their time and localization of expression in some cases differ significantly (Nusslein-Volhard and Dahm

29 2005). The gene duplication may create some problems in the course of studies, but the fully sequenced zebrafish genome and careful genomic analyses diminish the risk of overlooking additional gene copies. As a consequence of the relatively close evolutionary relation between zebrafish and humans, the two species share many genetic and anatomical features. Zebrafish display most of the human organs systems and structures, making it a valuable model to study biological mechanisms underlying some of the human diseases like muscular dystrophaty, melanoma or tuberculosis (Bassett, Bryson-Richardson et al. 2003; Swaim, Connolly et al. 2006; Ceol, Houvras et al. 2011).

Zebrafish in glycosaminoglycan research A number of zebrafish mutants affecting GAG synthesis as well as proteoglycans core proteins have been studied. Previously described mutations alter proteoglycan integrity at many levels, affecting availability of sugar nucleotide precursor (uxs1/hi954, udgh/jek), GAG chain polymerization (ext2/dak, extl3/box, xylt1, b3gat3/hi307), enzymes important for GAG sulfation (papst1/pic) and finally core protein availability (gpc4/knypek) (Topczewski, Sepich et al. 2001; LeClair, Mui et al. 2009). The zebrafish genes ext2 and extl2 involved in HS polymerization were found to be important for optic-tract sorting as well as pectoral fin and jaw development (Schilling, Piotrowski et al. 1996; Lee and Chien 2004). ext2 and extl3 mutant phenotypes have been associated with alterations in the signaling pathways, in particular those responding to Wnt and Fgf (Norton, Ledin et al. 2005). Mutants in both ext2 and papst1 (PAPS transporter 1) have been shown to develop cartilage defects strongly resembling those observed in humans with Hereditary Multiply Exostoses disease (HME) (Clement, Wiweger et al. 2008). Furthermore, studies of a zebrafish mutant lacking the core protein of Glypican 4, knypek, have shown that this proteoglycan is involved in the modulation of convergent extension movement during gastrulation (Topczewski, Sepich et al. 2001). Also three available lines carrying mutations in UDP-glucuronic acid decarboxylase 1 (uxs1) producing UDP- xylose from UDP-glucuronic acid, xylotransferase 1 (xylt1) and glucuronyltransferase I (b3gat3) exhibit impaired cartilage and bone formation. These enzymes are all important for the formation of the GAG linkage region (Eames, Yan et al. 2011; Wiweger, Avramut et al. 2011; Holmborn, Habicher et al. 2012).

Zebrafish pharyngeal cartilage development Several aspects of the zebrafish pharyngeal cartilage development, including correct chondrocyte stacking and pharyngeal cartilage element maturation are GAG structure dependent.

30 Formation of cartilaginous elements is initiated with the precartilage condensation. Already at 2 days post fertilization, the first elements of the cartilaginous skeleton begin to appear. This process gives rise to the perichondrium filled with chondrocytes and extensive amounts of matrix. With time, chondrocytes undergo flattening and intercalation forming elongated, slender structures, whereas other chondrocytes assemble to form flattened single cell layer elements. Soon after, cartilage is replaced by bone, in contrast to dermal bones which develops without any intermediate cartilage stage. Development of the zebrafish cranial skeleton involves seven pharyngeal arches where the two first arches serve as supportive structures for the jaw and operculum. The remaining pharyngeal arches (3rd-7th) are known as branchial arches (1st-5th), in adulthood functioning as a skeletal base for gills and teeth (Kimmel, Miller et al. 1998) (Fig. 5).

Figure 5. Craniofacial cartilage of the zebrafish larvae at 5 dpf. A. A ventral view of the Tg(fli1:EGFP) head region of a living zebrafish larvae. Blood vessels and cartilage elements express GFP. B. Schematic outline of the craniofacial cartilage. Abbreviations used: bh, basihyal; ch, ceratohyal; hs, hyosymplectic; ih, interhyal; m, Meckels; pq, palatoquadrate.

Taken together, zebrafish are easy to maintain and culture, giving them a strong advantage over other popular vertebrate models like mouse or rat. Numerous clutches, reaching up to 300 embryos from a pair, can be obtained weekly. All these advantages, together with their rapid embryonic development, makes zebrafish suitable for many types of studies, including toxicology screens, neurobiological investigations, behavioral studies, cancer research or for generation of human disease models.

31 Method considerations During the past few years, immense progress in the zebrafish reverse genetic approaches has been made. Our work on glycosaminoglycan biosynthetic enzymes have taken advantage of both old and newly established technologies. Learning and applying these have been a considerable part of my PhD work. A short description of the principles, advantages and disadvantages of methods used during the course of my studies can be found below.

Overexpression studies One of the biggest advantages of using zebrafish in research is the well- established microinjection procedure, commonly used to manipulate zebrafish gene expression. Microinjection into fertilized eggs at the 1-2 cell stage is relatively fast, gives low mortality rate and highly reproducible results. Short-term expression of high levels of protein can be obtained by injection of either DNA constructs or mRNA. Additionally, DNA constructs driving tissue specific expression, can be integrated into the genome. When combined with advanced microscopic visualization methods, a wide range of dynamic processes taking place in the developing embryo can be analyzed (Nusslein-Volhard and Dahm 2005).

Rise and fall of the morpholino technology Morpholino antisense nucleotides (MO) have been widely used in studies of gene function at the early stages of zebrafish development since the beginning of the century (Corey and Abrams 2001). Transient “knockdown” of translation or miss-splicing of a gene of interest can be obtained by injecting syntetic MOs characterized by their high affinity for mRNA. Complete penetration of the MO phenotype in the morphant fish (MO- injected) is usually observed during the first 2 days post fertilization, as the concentration of the MO in the cells is getting lower with every cell division (Nusslein-Volhard and Dahm 2005; Eisen and Smith 2008). Translational blocking MOs result in inhibition of both the maternally contributed transcripts and of mRNA from genes acting early during development. Splice-inhibiting MOs, on the other hand, are designed to cause either skipping of the exon or inhibition of pre-mRNA splicing. Their big advantage is the possibility of quantification of MO efficiency, but unfortunately splice-inhibiting MOs will not eliminate maternally contributed transcripts. After the initial burst of articles involving MO as a main technique to achieve gene “knock-down” without fully established control protocol, a growing need for monitoring possible off-target effects and knock-down

32 efficiency have been extensively discussed in the research community (Robu, Larson et al. 2007; Eisen and Smith 2008). To make MO derived results more reliable, several control experiments are recommended. Multiple target sites (translational-blocking and splice-blocking MOs), allow for confirmation of an obtained phenotype. Injections of the same amounts of mismatch and control MO as phenotype-giving MO followed by test for MO toxicity is one of the control experiment. If available, the most valuable control is comparison of MO phenotype with mutants carrying a mutation in the investigated gene. Also, MO blocking the most commonly up-regulated off-target pathways (p53 MO) and finally, MO phenotype rescue with RNA encoding for gene-of-interest orthologues or isoforms are suggested. Use of MO has been linked to increased levels of apoptotic neural death and to craniofacial cartilage development defects caused by off-target activation of the p53 pathway. These false positive MO results are reported to include close to 20% of total MO used in zebrafish (Robu, Larson et al. 2007). This bar has been now raised even higher by a study claiming that only about 20% of the already reported MO phenotypes can be reproduced in mutant fish created with CRISPR-Cas9 or TALENs technology (Kok, Shin et al. 2015). Yet, the MO technique, used wisely, remains a powerful instrument in zebrafish developmental studies. A number of MO based functional studies concerning the role of GAG biosynthetic genes in zebrafish including glce, hs2st, ndst1b, chsy1, csgalnact1 and chst11 have been published. Zebrafish morphants for the zebrafish glce (glcea and/or glceb) exhibit mild to severe early ventralization defects (Ghiselli and Farber 2005), whereas the phenotype of Glce mouse null-mutants is different, as they lack kidneys, and their lungs are poorly inflated. Pups die immediately after birth. In addition, skeletal abnormalities, with prominent defects in chondrocyte proliferation, are present (Li, Gong et al. 2003; Dierker, Bachvarova et al. 2015). This discrepancy between two different model organisms, points out the need for a trustworthy genetic method to verify MO phenotypes, which would be an important point of reference for any future studies using the MO method.

Selected methods for targeted genome editing in zebrafish

Transcription Activator-Like Effector Nucleases (TALENs) In 2011 a new technique for targeted mutagenesis was reported (Huang, Xiao et al. 2011). Transcription Activator-Like Effectors (TALEs) were originally discovered as proteins produced by plant pathogens, where, after entering the host cell, they bind to genomic DNA and affect transcription. In genetic engineering, modified TALE peptides have the ability to selectively, and with high affinity bind to DNA. Such peptides combined

33 with the nonspecific FokI endonuclease domain, have been presented as promising tool to introduce locus-specific double strand DNA breaks (DSB) (Dahlem, Hoshijima et al. 2012; Joung and Sander 2013). Depending on the conditions, either non-homologous end-joining (NHEJ) or homology directed repair (HDR) can then be used by the cell to repair damages. Imprecise correction of the DSBs by the NHEJ system can lead to small, insertions or/and deletions (indels) of nucleotides at the DSB site. Mistakes created within open reading frames, can cause frame shift and result in the generation of a premature stop codon, leading to gene disruption and production of truncated protein (Joung and Sander 2013). FokI nucleases work only as a dimer. Therefore, TALENS are designed to bind two unique targets separated from each other by a short spacer. TALE consists of sequence blocks of 33-35 amino acid repeats followed by a final shorter repeat. The amino acids at positions 12 and 13 within each TALE building block, determines to which of the four DNA bases the TALE will bind (Cermak, Doyle et al. 2011). Typically, working on a small scale with this technique involves two rounds of complicated cloning steps and purification of as many as 60 different plasmids which can be both time consuming and troublesome.

Power of the simplicity, Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR associated protein 9 (Cas9) The latest innovation in the field of genome editing technologies came in 2013, when the first work regarding usage of the CRISPR-Cas system in C. elegans, Drosophila, zebrafish and many other systems, was published (Bassett, Tibbit et al. 2013; Friedland, Tzur et al. 2013; Hwang, Fu et al. 2013). CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) together with CRISPR-associated (Cas) genetic elements were first discovered as a form of unique immune system against viral infections, evolved in bacteria and archaea. The CRISPR-Cas9 system, used as a reverse genetic tool, is based on its ability to create targeted DSBs. In zebrafish, the CRISPR-Cas9 system is based on two RNA components which are injected into the zebrafish embryo at the 1-cell stage. The RNA molecules encode Cas9 endonuclease and a recombinant single guide RNA, directing the Cas9 nuclease to the specific site in the genome (Fig. 6). Valuable applications of the CRISPR-Cas9 method are not limited to generation of null-mutants. Combining the CRISPR-Cas9 system with single strand DNA carrying homology blocks towards DNA sequences framing the predicted DSB location, can be used for site specific DNA insertions or substitutions. These developments of the CRISPR-Cas9 technology illustrate

34 the possibilities to use it to model human genetic disorders and in the future for treatment of single-gene hereditary diseases (Doudna 2015).

Figure 6. The CRISPR-Cas9 system. Cas9 is guided by the RNA molecule containing sequence matching the targeted site within the genome. The PAM site is required by the CRISPR-Cas9 complex for target recognition. DBS are generated as a result of Cas9 action and are then repaired by the cell by either homologous recombination or by the NHEJ process. Adapted from (Charpentier and Doudna 2013).

Taken together, in comparison with other available targeted genome editing tools, the CRISPR-Cas9 technique has the advantage to be much easier when it comes to the target design and method implementation. High genome editing efficiency together with relatively low toxicity make the CRISPR- Cas9 system possible to use also for targeting of several sites in the genome at the same time, defined as multiplexing.

35 Present investigation

Aim The overall aim of this work was to use the zebrafish model to study the role of GAGs in embryonic development. Basic knowledge regarding several of the zebrafish GAG biosynthesis enzymes was lacking, including information about number of isoforms and their temporal and spatial expression during development. The first objective was therefore to fill this gap. The second objective to generate GAG biosynthesis zebrafish mutants for our studies was greatly aided by the recent development of the CRISPR-Cas9 genetic tool.

Paper I The NDST gene family in zebrafish: role of NDST1b in pharyngeal arch formation.

Much previous work in our group has concentrated on NDSTs, their enzymatic properties and interactions with other HS biosynthetic enzymes and the way by which the NDSTs affect GAG chain composition, length and amounts. In vivo functions of NDSTs have also been the subject of previous publications from our group including the characterization of mice lacking functional NDST1 and NDST2 enzyme (Forsberg, Pejler et al. 1999; Ringvall, Ledin et al. 2000; Ledin, Staatz et al. 2004). When the present study was initiated, only fragmentary information regarding NDST orthologues present in zebrafish was available. With our study we aimed to complete the analysis of the zebrafish heparan sulfate biosynthetic enzymes. We have identified, cloned and sequenced the five members of the zebrafish NDST gene family which includes two orthologues of NDST1 and NDST2, ndst1a, ndst1b, ndst1b, ndst2b, and a single ndst3 orthologue (Filipek-Gorniok, Carlsson et al. 2015). Our results indicate that NDST3 and NDST4, present in mouse and human and located in close proximity to each other, arose as a result of a local gene duplication. A knockout of the single ndst3 orthologue in zebrafish could possibly mimic the effect of a NDST3/NDST4 double knockout in mice (Filipek- Gorniok, Carlsson et al. 2015).

36 All zebrafish ndsts were shown to be maternally contributed and were expressed in the otic capsule and pectoral fins. Additionally, most of the genes were expressed in the different brain parts, whereas in the spinal cord, only ndst3 transcript was detected. Attempts were also made to knock-down each of the zebrafish isoforms to study their effects on zebrafish embryonic development. Despite our strenuous attempts, we were not able to obtain confirmed, stable phenotypes for all but the ndst1b morphants. Two transgenic zebrafish lines were used to better visualize craniofacial cartilage defects of the ndst1b morphants and the observed phenotype showed resemblance but was somewhat milder than that of the Ndst1-/- mouse. Ndst1b morphants had shortened pectoral fins and body axis. The most prominent phenotypic change was observed in the second pharyngeal arch, which was reverted in about 40% and perpendicular in approximately 10% of investigated morphant larvae.

Paper II Expression of chondroitin/dermatan sulfate glycosyltransferases during early zebrafish development.

In this work, we studied a previously not characterized group of enzymes involved in CS/DS chain polymerization. Protein sequence comparison of the CS/DS polymerizing enzymes revealed the existence of seven glycosyltransferases, which can be further divided into the CSGALNACT, CHSY and CHPF families. The single CHPF human gene was found to be duplicated in zebrafish. Whole mount in situ hybridization of the genes demonstrated their overlapping expression in the notochord, the pectoral fin, the otic capsule and in cranial cartilage including the branchial arches. These results suggest, that the cartilaginous elements of the developing zebrafish embryo may be especially rich in CS/DS. This hypothesis was confirmed by antibody staining with the CS-specific antibody (CS-56) as well as by staining with alcian blue, showing intense colouring in the pharyngeal cartilage elements and notochord. Most interestingly, the levels of HS were rather stable through all analyzed developmental stages (1hpf-6dpf), whereas CS/DS levels increased dramatically from 48hpf onwards, closely following the time course of cartilage formation. Our results indicate that HS and CS/DS present in the early embryo is produced by the zygote, rather than being of maternal origin.

37 Paper III On the roles and regulation of chondroitin sulfate and heparan sulfate in zebrafish pharyngeal cartilage morphogenesis.

In this study, we addressed the role of HS and CS proteoglycans in zebrafish development using existing mutants of two enzymes important for linkage region formation (uxs1, b3gat3) and two for HS chain initiation and polymerization (extl3, ext2). All investigated mutants exhibited severe defects and were lethal at larval stage. We focused on the pharyngeal cartilage structures, as they were the ones predominantly affected by altered GAG biosynthesis. To visualize and analyze early stages of the craniofacial cartilage development, mutant embryos of the fli:EGFP transgenic background were used. Interestingly, lack of functional uxs1 enzyme caused a dramatic effect on the development of the zebrafish larvae ceratohyal (pharyngeal cartilage element) leading to its nearly complete loss. We speculate that the craniofacial defects present in the mutant embryos might be a consequence of a disruption of the extracellular matrix physiochemical properties as a result of the loss of GAGs. The described mutations did not change the expression of PG core proteins except for biglycan, as analyzed by quantitative PCR. Main phenotypic changes observed in mutant embryos are summarized below (Tab. 2).

Table 2. Main findings. ‘+’; severity of the defect, ‘/’; not included in the study. Defects uxs1-/- b3gat3-/- extl3-/- ext2-/- ext2-/-; uxs1-/- Pharyngeal cartilage ++ ++ + ++ +++ Chondrocyte stacking + / / + ++ Reduction of the ECM + / / No ++ Shortened pectoral fin + + + Absent Absent GAG content HS; 52% HS; 55% HS; 44% HS; 18% HS; 13% (in % of the control) CS; 5% CS; 11% CS;147% CS; 88% CS; 16%

Additionally, based on the data obtained by analysis of the GAG content of the extl3 mutant and the csgalnact1;csgalnact2 morphant fish and data derived from the extl3 siRNA-treated cells, we propose that the balance between expression of these genes affects the HS/CS content. As shown in this work, functional ext2, extl3, uxs1 and b3gat3 enzymes are required to provide normal HS and CS/DS levels in the developing zebrafish larvae.

38 Paper IV Large-scale generation of zebrafish mutants with defective glycosaminoglycan biosynthesis.

In papers I and III our attempts to study the biological roles of the GAG biosynthetic enzymes in the developing zebrafish embryo were highly restricted by the lack of adequate tools to create new zebrafish mutants. Moreover, the MO technique, used to knock-down expression of GAG biosynthetic enzymes have several disadvantages, as discussed in the chapter on method considerations. In this ongoing study, we decided to take advantage of the rapidly developing CRISPR-Cas9 method, which in a short time has revolutionized the field of targeted mutagenesis in zebrafish. In close collaboration with Shawn Burgess group, (NIH, US) we used a CRISPR-Cas9 based approach, published by Varshney et. al, to create zebrafish mutants with defective HS and CD/DS biosynthesis (Varshney, Pei et al. 2015) (Fig. 7). I focused on the five zebrafish ndst genes as well as the two genes encoding zebrafish papss2. Combining recent advantages of the CRISPR- Cas9 system, including multiplexing and genotyping protocols, we have managed to identify 18 new zebrafish mutants of genes important for GAG biosynthesis. Lack of obvious morphological phenotypes linked to the null-mutations of most of the investigated genes is the main finding of this study. For example, no obvious morphological abnormalities were apparent in the adult homozygous chsy1 mutants, where earlier published zebrafish chsy1 morphants show strong malformations during development (Peal, Burns et al. 2009; Tian, Ling et al. 2010). By creating vertebrate models with alterations in GAG stucture, we hope to answer some of the questions regarding the roles of HS and CS/DS biosynthetic enzymes in zebrafish development and homeostasis. Studying function and effects of loss of one or more genes involved in GAG biosynthesis in a vertebrate model system such as zebrafish, have never before been so attainable.

39

Figure 7. Overview of the CRISPR/Cas9 zebrafish mutant pipeline. Two sgRNAs are recommended for each targeted gene. In our set up, we have targeted up to four genes at the same time by coinjection of maximally 8 sgRNAs with Cas9 mRNA into a 1-cell stage embryo. The chimeric embryos were raised to generate founder fish, incrossed and F1 offspring was screened for the mutations. Although identification of several individuals carrying uniform mutant alleles is possible, in the vast majority of cases, additional outcross steps were required to obtain a sufficient number of heterozygous individuals. Correlation between phenotype and genotype was tested in F3 embryos.

40 Concluding remarks and future perspectives

This thesis summarizes efforts made to elucidate functional aspects of GAG biosynthesis using the zebrafish as a model system. We demonstrate a strict spatiotemporal expression of both the members of the NDST gene family (Paper I) and the CS/DS glycosyltransferases (Paper II) in the developing zebrafish embryo. Paper III takes advantage of four preexisting zebrafish mutants, defective in GAG biosynthesis. Taken together, the results obtained in this paper demonstrate a correlation between HS content and the severity of the pectoral fin defect. Additionally, impaired HS biosynthesis was shown to be associated with altered chondrocyte intercalation. Interestingly, defect CS biosynthesis resulted in loss of the chondrocyte extracellular matrix in the investigated mutant fish. One of the main findings in this study is that the relative expression levels of Extl3 and Csgalnact1/Csgalnact2 may influence the HS/CS ratio. The importance of GAGs during embryonic development have been studied in several model organisms and a number of human genetic diseases have been linked to mutations in genes encoding HS and CS/DS biosynthetic enzymes. Intervention of HS and CS/DS biosynthesis may therefore be a future therapeutic approach to cure GAG-associated disorders. In Paper IV, we took advantage of the development of the CRISPR-Cas9 technique. Here we aimed to create a collection of zebrafish mutants with defected GAG biosynthetic machineries. The vast majority of the newly created null-mutants did not exhibit any obvious aberrant phenotype. The complexity of GAG structure in vertebrates may correlate with the increasing number of enzymes involved in GAG assembly. Gene duplications in the vertebrate lineage can, in some cases, be the reason why efficient reduction of GAG synthesis is not achieved with knockot/knockdown of only one of the gene copies Also, contribution of the mRNA encoding the correct version of the protein deposited in the embryo by the heterozygous mother could be sufficient for early development of zebrafish (Kok, Shin et al. 2015). Additionally, other poorly understood compensation mechanisms might be activated in the cell affected by the gene knock-out. In a recent study by Rossi et al. it is pointed out that transcriptome changes of the mutant, not present in morphant fish, can compensate for the null mutations (Rossi, Kontarakis et al. 2015). Such

41 changes would be very interesting to study in the CRISPR-Cas9 mutants we have generated. Two of the promising candidate genes linked to human genetic disorders and of special interest to me are NDST3 and PAPSS2, for which we are currently raising fish carrying null mutations. Our preliminary results from the morpholino studies of the papss2a gene, suggest that it is of importance for muscle fiber organization as well as for notochord development.

42 Populärvetenskaplig sammanfattning

Förutom DNA, RNA och proteiner kan också kolhydrater bära information som behövs för att möjliggöra liv. Proteoglykaner består av en proteinkärna till vilken en eller flera långa ogrenade kolhydratkedjor (s k glykosaminoglykaner) är fastsatta. Glykosaminoglykanena är negativt laddade, där ofta mer än hälften av laddningarna beror på sulfatgrupper utspridda längs glykosaminoglykanerna i mönster som känns igen av proteiner med positivt laddade aminosyror på sin yta. Olika mönster attraherar olika molekyler och ett sätt för celler att kommunicera med omgivningen är med hjälp av proteoglykanerna. Mitt arbete har fokuserat på två typer av glykosaminoglykaner, heparansulfat (HS) och chondroitin/dermatansulfat (CS/DS). Jag har intresserat mig för deras biosyntes och de roller de har under embryonalutvecklingen. Mina studier har gjorts på zebrafisk, som tillsammans med möss, råttor, fruktflugor (Drosophila melanogaster) och nematoder (C. elegans) är populära biologiska modellsystem. Zebrafiskar är ryggradsdjur och innehåller motsvarigheter till alla huvudsakliga organ hos människor. De har använts i många typer av studier bl a för att göra modeller för att efterlikna sjukdomar hos människa. Zebrafisken är en liten tropisk sötvattensfisk. I fångenskap blir den upp till 4 cm lång. Befruktning och fortsatt utveckling sker utanför honan och de många embryonerna utvecklas innanför en genomskinlig fosterhinna. Detta gör det lätt att följa embryots utveckling. Mitt första arbete i avhandlingen fokuserar på NDST-genfamiljen. Enzymer i denna familj deltar i biosyntes av HS och har en nyckelroll när det gäller att bestämma hur sulfatmönstret på den färdiga HS-kedjan ska se ut. Vi fann att zebrafiskar tillverkar fem sådana enzymer istället för människans fyra. Den andra artikeln i avhandlingen handlar om de sju enzymer i zebrafisk som bygger CS/DS-kedjorna. Dessa två första studier där vi kartlagt var och när enzymerna tillverkas utgör en viktig bas för fortsatta studier av glykosaminoglykaners roll under embryonalutvecklingen. I den tredje studien där vi använde oss av mutanter med defekt syntes av glykosaminoglykaner kunde vi bl a visa att CS/DS ibland kan kompensera för HS när dess syntes inte fungerar som den ska. I det fjärde arbetet som är en preliminär studie, använde vi oss av nyligen utvecklade genetiska metoder för att inaktivera zebrafiskgener. Detta har tills nyligen varit svårt att göra. Vi har generarat mutanter för många av de enzymer som deltar i biosyntes av både HS och CS/DS och påbörjat studier av hur detta påverkar embryonalutvecklingen.

43 Acknowledgments

Starting from the earliest years, my interest in biology was encouraged by the extraordinary teachers - Mr. Bogunia in the primary school; Ms. Dorda and Zamarska in the high school; a group of passionate academics, including Ms. Zięba, Mr. Szklarczyk and Mr. Jasienski at the University in Krakow; and finally Lena Kjellén and Johan Ledin at Uppsala University.

In my wildest dreams, I never expected to be enrolled in a PhD program. This work could never be completed without the support and guidance I have had over the years. Thus, I would like to express my sincere gratitude to all who made this exciting journey possible and enjoyable.

My supervisor Lena, thank you for all the support, time, attention and patience you gave me! I could not wish for a better supervisor. You permitted me to develop as a researcher, have shared your infinite knowledge and have been an excellent role model as a scientist, but also as a human being.

My co-supervisor Johan, without you, from the start until the end, none of this would have been possible. Your help and support are always of the greatest value, whether it regards professional or personal issues.

I would like to thank Jonas von Hofsten from Umeå University, Sweden, Shawn Burgess and Gaurav Varshney from NIH, US for giving me the great opportunity of getting the new perspectives by working in their labs.

To all present and former members of the Kjellen group, Inger, Pernilla, Anders & Anders (Lunde), Audrey, Dagmar, Anh-Tri, Cathy and all the project and master students, for sharing your knowledge, support, but also for creating such a friendly atmosphere throughout these years. Additionally, a special thanks to Tabea, who many times has proven to be an excellent travel companion and cat-caretaker.

My lovely zebrafish gang, Judith, Katarina and more recently Ling, you have been irreplaceable company whatever we did, or wherever we went. Thank you all for your support, advice and friendship! Tatjana and other

44 collaborators, thank you for fruitful cooperation and sharing your knowledge.

Office-team, you have filled my time with engaging conversations regarding politics, sport, history, geography, music, books, movies and science! If I needed help or company, there was someone to answer my questions or just grab a cup of coffee with. I missed you all this last year!

Members of the D9:4 corridor, IMBIM and EvoDevo EBC colleges, it was a pleasure and an inspiration to share my time with you.

Administrative workers in IMBIM and EBC, for solving all my requests, support and friendly attitude.

To all my dear Uppsala friends through the years, Riccardo & Niroj, Sebastian, Maja, Annette, Zsuzska & Juri, Prune, Zhenechka, Steffano, Nhi & Enzo, Nick & Lisa, Åsa & Matan, Gabi & Alina, Veronique & Benq, Mats & Harriet, Anna & Andrei. I am lucky to have such great people around me!

Special thanks to all of you involved in Sputnik-Katushka-Kalashnikov- Balkan-Gyspy-Punk-Ska-More Palinka-Dubstep Madness! As long as you will play, I will always be there.

Wszystkim byłym i obecnym członkom polskiej ekipy w Uppsali, za nasze spotkania, wycieczki, domowe obiadki, imprezy, wsparcie, pomoc i niańczenie Emilki i Mumrika. Ekipie campingowo-wypadowej, za zapewnianie wybornej zabawy na wyjazdach!

Kabot, Didi, Lala i Wronia – dziękuję, że jesteście! Nie potrafię wyobrazić sobie życia, bez waszej w nim obecności! Agatka, Maria i Ania – za stare, dobre czasy na Akademi Rolniczej, za to, że zawsze znajdujecie czas na krakowskie ploteczki i odwiedziny w Szwecji.

Moim rodzicom i rodzinie, za to, że nigdy nie zwątpili w moje możliwości. Chociaż nasze życiowe drogi są różne, zawsze akceptowaliście moje wybory i podróżniczą pasję. Dziękuję za wasze wsparcie i pomoc w każdej sytuacji! Dziękuję Zosi i Henrykowi, za przyjęcie mnie do swojej rodziny, wychowanie najlepszego w świecie męża i za pomoc w opiece nad Emilką.

Łukaszowi i Emilce, za to że jesteście Ziemią pod moimi stopami i Niebiem nad moją głową.

45 References

Aikawa, J. and J. D. Esko (1999). "Molecular cloning and expression of a third member of the heparan sulfate/heparin GlcNAc N-deacetylase/ N- sulfotransferase family." J Biol Chem 274(5): 2690-2695. Aikawa, J., K. Grobe, et al. (2001). "Multiple isozymes of heparan sulfate/heparin GlcNAc N-deacetylase/GlcN N-sulfotransferase. Structure and activity of the fourth member, NDST4." J Biol Chem 276(8): 5876-5882. Amsterdam, A., R. M. Nissen, et al. (2004). "Identification of 315 genes essential for early zebrafish development." Proc Natl Acad Sci U S A 101(35): 12792- 12797. Bandtlow, C. E. and D. R. Zimmermann (2000). "Proteoglycans in the developing brain: new conceptual insights for old proteins." Physiol Rev 80(4): 1267-1290. Bassett, A. R., C. Tibbit, et al. (2013). "Highly efficient targeted mutagenesis of Drosophila with the CRISPR/Cas9 system." Cell Rep 4(1): 220-228. Bassett, D. I., R. J. Bryson-Richardson, et al. (2003). "Dystrophin is required for the formation of stable muscle attachments in the zebrafish embryo." Development 130(23): 5851-5860. Bernfield, M., M. Gotte, et al. (1999). "Functions of cell surface heparan sulfate proteoglycans." Annu Rev Biochem 68: 729-777. Besset, S., J. B. Vincourt, et al. (2000). "Nuclear localization of PAPS synthetase 1: a sulfate activation pathway in the nucleus of eukaryotic cells." FASEB J 14(2): 345-354. Bishop, J. R., M. Schuksz, et al. (2007). "Heparan sulphate proteoglycans fine-tune mammalian physiology." Nature 446(7139): 1030-1037. Blake, J. A., C. J. Bult, et al. (2014). "The Mouse Genome Database: integration of and access to knowledge about the laboratory mouse." Nucleic Acids Res 42(Database issue): D810-817. Block, H., K. Ley, et al. (2012). "Severe impairment of leukocyte recruitment in ppGalNAcT-1-deficient mice." J Immunol 188(11): 5674-5681. Bonneh-Barkay, D. and C. A. Wiley (2009). "Brain extracellular matrix in neurodegeneration." Brain Pathol 19(4): 573-585. Bullock, S. L., J. M. Fletcher, et al. (1998). "Renal agenesis in mice homozygous for a gene trap mutation in the gene encoding heparan sulfate 2-sulfotransferase." Genes Dev 12(12): 1894-1906. Busse-Wicher, M., K. B. Wicher, et al. (2014). "The exostosin family: proteins with many functions." Matrix Biol 35: 25-33. Cadwalader, E. L., M. L. Condic, et al. (2012). "2-O-sulfotransferase regulates Wnt signaling, cell adhesion and cell cycle during zebrafish epiboly." Development 139(7): 1296-1305. Carlsson, P., J. Presto, et al. (2008). "Heparin/heparan sulfate biosynthesis: processive formation of N-sulfated domains." J Biol Chem 283(29): 20008- 20014.

46 Ceol, C. J., Y. Houvras, et al. (2011). "The histone methyltransferase SETDB1 is recurrently amplified in melanoma and accelerates its onset." Nature 471(7339): 513-517. Cermak, T., E. L. Doyle, et al. (2011). "Efficient design and assembly of custom TALEN and other TAL effector-based constructs for DNA targeting." Nucleic Acids Res 39(12): e82. Charpentier, E. and J. A. Doudna (2013). "Biotechnology: Rewriting a genome." Nature 495(7439): 50-51. Chen, E., S. E. Stringer, et al. (2005). "A unique role for 6-O sulfation modification in zebrafish vascular development." Dev Biol 284(2): 364-376. Clement, A., M. Wiweger, et al. (2008). "Regulation of Zebrafish Skeletogenesis by ext2/dackel and papst1/pinscher." PLoS ONE 4(7): 1-11. Clement, A., M. Wiweger, et al. (2008). "Regulation of zebrafish skeletogenesis by ext2/dackel and papst1/pinscher." PLoS Genet 4(7): e1000136. Condac, E., R. Silasi-Mansat, et al. (2007). "Polycystic disease caused by deficiency in xylosyltransferase 2, an initiating enzyme of glycosaminoglycan biosynthesis." Proc Natl Acad Sci U S A 104(22): 9416-9421. Corey, D. R. and J. M. Abrams (2001). "Morpholino antisense oligonucleotides: tools for investigating vertebrate development." Genome Biol 2(5): REVIEWS1015. Dahlem, T. J., K. Hoshijima, et al. (2012). "Simple methods for generating and detecting locus-specific mutations induced with TALENs in the zebrafish genome." PLoS Genet 8(8): e1002861. Dejima, K., A. Seko, et al. (2006). "Essential roles of 3'-phosphoadenosine 5'- phosphosulfate synthase in embryonic and larval development of the nematode Caenorhabditis elegans." J Biol Chem 281(16): 11431-11440. Dierker, T., V. Bachvarova, et al. (2015). "Altered heparan sulfate structure in Glce mice leads to increased Hedgehog signaling in endochondral bones." Matrix Biol. Doudna, J. A. (2015). "Genomic engineering and the future of medicine." JAMA 313(8): 791-792. Dyck, S. M. and S. Karimi-Abdolrezaee (2015). "Chondroitin sulfate proteoglycans: Key modulators in the developing and pathologic central nervous system." Exp Neurol 269: 169-187. Eames, B. F., A. Singer, et al. (2010). "UDP xylose synthase 1 is required for morphogenesis and histogenesis of the craniofacial skeleton." Dev Biol 341(2): 400-415. Eames, B. F., Y. L. Yan, et al. (2011). "Mutations in fam20b and xylt1 reveal that cartilage matrix controls timing of endochondral ossification by inhibiting chondrocyte maturation." PLoS Genet 7(8): e1002246. Eisen, J. S. and J. C. Smith (2008). "Controlling morpholino experiments: don't stop making antisense." Development 135(10): 1735-1743. Esko, J. D., K. Kimata, et al. (2009). Essentials of Glycobiology. 2nd edition. Cold Spring Harbor Laboratory Press. A. Varki, R. D. Cummings and J. D. Esko, Cold Spring Harbor Esko, J. D. and S. B. Selleck (2002). "Order out of chaos: assembly of ligand binding sites in heparan sulfate." Annu Rev Biochem 71: 435-471. Evers, M. R., G. Xia, et al. (2001). "Molecular cloning and characterization of a dermatan-specific N-acetylgalactosamine 4-O-sulfotransferase." J Biol Chem 276(39): 36344-36353.

47 Fan, G., L. Xiao, et al. (2000). "Targeted disruption of NDST-1 gene leads to pulmonary hypoplasia and neonatal respiratory distress in mice." FEBS Lett 467(1): 7-11. Farhan, S. M., J. Wang, et al. (2015). "Old gene, new phenotype: mutations in heparan sulfate synthesis enzyme, EXT2 leads to seizure and developmental disorder, no exostoses." J Med Genet 52(10): 666-675. Filipek-Gorniok, B., P. Carlsson, et al. (2015). "The ndst gene family in zebrafish: role of ndst1b in pharyngeal arch formation." PLoS One 10(3): e0119040. Forsberg, E., G. Pejler, et al. (1999). "Abnormal mast cells in mice de®cient in a heparin-synthesizing enzyme." Nature 400: 773-776. Friedland, A. E., Y. B. Tzur, et al. (2013). "Heritable genome editing in C. elegans via a CRISPR-Cas9 system." Nat Methods 10(8): 741-743. Fuda, H., C. Shimizu, et al. (2002). "Characterization and expression of human bifunctional 3'-phosphoadenosine 5'-phosphosulphate synthase isoforms." Biochem J 365(Pt 2): 497-504. Gallagher, J. T., J. E. Turnbull, et al. (1992). "Patterns of sulphation in heparan sulphate: polymorphism based on a common structural theme." Int J Biochem 24(4): 553-560. Gama, C. I., S. E. Tully, et al. (2006). "Sulfation patterns of glycosaminoglycans encode molecular recognition and activity." Nat Chem Biol 2(9): 467-473. Ghiselli, G. and S. A. Farber (2005). "D-glucuronyl C5-epimerase acts in dorso- ventral axis formation in zebrafish." BMC Dev Biol 5: 19. Gotoh, M., T. Sato, et al. (2002). "Enzymatic synthesis of chondroitin with a novel chondroitin sulfate N-acetylgalactosaminyltransferase that transfers N- acetylgalactosamine to glucuronic acid in initiation and elongation of chondroitin sulfate synthesis." J Biol Chem 277(41): 38189-38196. Grobe, K., M. Inatani, et al. (2005). "Cerebral hypoplasia and craniofacial defects in mice lacking heparan sulfate Ndst1 gene function." Development 132(16): 3777-3786. Grobe, K., J. Ledin, et al. (2002). "Heparan sulfate and development: differential roles of the N-acetylglucosamine N-deacetylase/N-sulfotransferase isozymes." Biochim Biophys Acta 1573(3): 209-215. Habuchi, H., N. Nagai, et al. (2007). "Mice deficient in heparan sulfate 6-O- sulfotransferase-1 exhibit defective heparan sulfate biosynthesis, abnormal placentation, and late embryonic lethality." J Biol Chem 282(21): 15578-15588. Habuchi, H., M. Tanaka, et al. (2000). "The occurrence of three isoforms of heparan sulfate 6-O-sulfotransferase having different specificities for hexuronic acid adjacent to the targeted N-sulfoglucosamine." J Biol Chem 275(4): 2859-2868. Häcker, U., K. Nybakken, et al. (2005). "Developmental Cell Biology: Heparan sulphate proteoglycans: the sweet side of development." Nature Reviews Molecular Cell Biology 6(7): 530-541. Harfouche, R., D. M. Hentschel, et al. (2009). "Glycome and Transcriptome Regulation of Vasculogenesis." Circulation 120(19): 1883-1892. Holmborn, K., J. Habicher, et al. (2012). "On the roles and regulation of chondroitin sulfate and heparan sulfate in zebrafish pharyngeal cartilage morphogenesis." J Biol Chem 287(40): 33905-33916. Holmborn, K., J. Ledin, et al. (2004). "Heparan sulfate synthesized by mouse embryonic stem cells deficient in NDST1 and NDST2 is 6-O-sulfated but contains no N-sulfate groups." J Biol Chem 279(41): 42355-42358. Howe, K., M. D. Clark, et al. (2013). "The zebrafish reference genome sequence and its relationship to the human genome." Nature 496(7446): 498-503.

48 Huang, P., A. Xiao, et al. (2011). "Heritable gene targeting in zebrafish using customized TALENs." Nat Biotechnol 29(8): 699-700. Humphries, D. E., G. W. Wong, et al. (1999). "Heparin is essential for the storage of specific granule proteases in mast cells." Nature 400(6746): 769-772. Hwang, H. Y., S. K. Olson, et al. (2003). "Caenorhabditis elegans early embryogenesis and vulval morphogenesis require chondroitin biosynthesis." Nature 423(6938): 439-443. Hwang, W. Y., Y. Fu, et al. (2013). "Efficient genome editing in zebrafish using a CRISPR-Cas system." Nat Biotechnol 31(3): 227-229. Iozzo, R. V. (1998). "Matrix proteoglycans: from molecular design to cellular function." Annu Rev Biochem 67: 609-652. Iozzo, R. V. (1999). "The biology of the small leucine-rich proteoglycans. Functional network of interactive proteins." J Biol Chem 274(27): 18843-18846. Izumikawa, T., T. Koike, et al. (2008). "Identification of chondroitin sulfate glucuronyltransferase as chondroitin synthase-3 involved in chondroitin polymerization: chondroitin polymerization is achieved by multiple enzyme complexes consisting of chondroitin synthase family members." J Biol Chem 283(17): 11396-11406. Izumikawa, T., T. Uyama, et al. (2007). "Involvement of chondroitin sulfate synthase-3 (chondroitin synthase-2) in chondroitin polymerization through its interaction with chondroitin synthase-1 or chondroitin-polymerizing factor." Biochem J 403(3): 545-552. Joung, J. K. and J. D. Sander (2013). "TALENs: a widely applicable technology for targeted genome editing." Nat Rev Mol Cell Biol 14(1): 49-55. Jullien, D., M. Crozatier, et al. (1997). "cDNA sequence and expression pattern of the Drosophila melanogaster PAPS synthetase gene: a new salivary gland marker." Mech Dev 68(1-2): 179-186. Kamiyama, S., N. Sasaki, et al. (2006). "Molecular cloning and characterization of a novel 3'-phosphoadenosine 5'-phosphosulfate transporter, PAPST2." J Biol Chem 281(16): 10945-10953. Kamiyama, S., T. Suda, et al. (2003). "Molecular cloning and identification of 3'- phosphoadenosine 5'-phosphosulfate transporter." J Biol Chem 278(28): 25958- 25963. Kimmel, C. B., W. W. Ballard, et al. (1995). "Stages of embryonic development of the zebrafish." Dev Dyn 203(3): 253-310. Kimmel, C. B., C. T. Miller, et al. (1998). "The shaping of pharyngeal cartilages during early development of the zebrafish." Dev Biol 203(2): 245-263. Kitagawa, H., T. Izumikawa, et al. (2003). "Molecular cloning of a chondroitin polymerizing factor that cooperates with chondroitin synthase for chondroitin polymerization." J Biol Chem 278(26): 23666-23671. Kitagawa, H., T. Uyama, et al. (2001). "Molecular cloning and expression of a human chondroitin synthase." J Biol Chem 276(42): 38721-38726. Kjellen, L. and U. Lindahl (1991). "Proteoglycans: Structures and Interactions." Annu. Rev. Biochem 60: 443-475. Kluppel, M., T. N. Wight, et al. (2005). "Maintenance of chondroitin sulfation balance by chondroitin-4-sulfotransferase 1 is required for chondrocyte development and growth factor signaling during cartilage morphogenesis." Development 132(17): 3989-4003. Knudson, C. B. and W. Knudson (2001). "Cartilage proteoglycans." Semin Cell Dev Biol 12(2): 69-78.

49 Kobayashi, S., K. Morimoto, et al. (2000). "Association of EXT1 and EXT2, hereditary multiple exostoses gene products, in Golgi apparatus." Biochem Biophys Res Commun 268(3): 860-867. Kok, F. O., M. Shin, et al. (2015). "Reverse genetic screening reveals poor correlation between morpholino-induced and mutant phenotypes in zebrafish." Dev Cell 32(1): 97-108. Kreuger, J. and L. Kjellen (2012). "Heparan sulfate biosynthesis: regulation and variability." J Histochem Cytochem 60(12): 898-907. Kreuger, J., D. Spillmann, et al. (2006). "Interactions between heparan sulfate and proteins: the concept of specificity." J Cell Biol 174(3): 323-327. Kuo, W. J., M. A. Digman, et al. (2010). "Heparan sulfate acts as a bone morphogenetic protein coreceptor by facilitating ligand-induced receptor hetero- oligomerization." Mol Biol Cell 21(22): 4028-4041. Kusche-Gullberg, M., I. Eriksson, et al. (1998). "Identification and expression in mouse of two heparan sulfate glucosaminyl N-deacetylase/N-sulfotransferase genes." J Biol Chem 273(19): 11902-11907. Lander, A. D., Q. Nie, et al. (2002). "Do morphogen gradients arise by diffusion?" Dev Cell 2(6): 785-796. Lane, P. W. and M. M. Dickie (1968). "Three recessive mutations producing disproportionate dwarfing in mice: achondroplasia, brachymorphic, and stubby." J Hered 59(5): 300-308. LeClair, E. E., S. R. Mui, et al. (2009). "Craniofacial skeletal defects of adult zebrafishGlypican 4 (knypek)mutants." Developmental Dynamics 238(10): 2550-2563. Ledin, J., W. Staatz, et al. (2004). "Heparan sulfate structure in mice with genetically modified heparan sulfate production." J Biol Chem 279(41): 42732- 42741. Lee, J. S. and C. B. Chien (2004). "When sugars guide axons: insights from heparan sulphate proteoglycan mutants." Nat Rev Genet 5(12): 923-935. Lee, J. S., S. von der Hardt, et al. (2004). "Axon sorting in the optic tract requires HSPG synthesis by ext2 (dackel) and extl3 (boxer)." Neuron 44(6): 947-960. Lencz, T., S. Guha, et al. (2013). "Genome-wide association study implicates NDST3 in schizophrenia and bipolar disorder." Nat Commun 4: 2739. Li, J. P., F. Gong, et al. (2003). "Targeted disruption of a murine glucuronyl C5- epimerase gene results in heparan sulfate lacking L-iduronic acid and in neonatal lethality." J Biol Chem 278(31): 28363-28366. Lin, X., E. M. Buff, et al. (1999). "Heparan sulfate proteoglycans are essential for FGF receptor signaling during Drosophila embryonic development." Development 126(17): 3715-3723. Lin, X., G. Wei, et al. (2000). "Disruption of gastrulation and heparan sulfate biosynthesis in EXT1-deficient mice." Dev Biol 224(2): 299-311. Lindahl, U. and L. Jin-ping (2009). "Interactions Between Heparan Sulfate And Proteins - Design And Functionl Implications." International Review of Cell and Molecular Biology 276: 105-159. Lindahl, U., M. Kusche-Gullberg, et al. (1998). "Regulated diversity of heparan sulfate." J Biol Chem 273(39): 24979-24982. Little, P. J., M. L. Ballinger, et al. (2008). "Biosynthesis of natural and hyperelongated chondroitin sulfate glycosaminoglycans: new insights into an elusive process." Open Biochem J 2: 135-142. Maccarana, M., S. Kalamajski, et al. (2009). "Dermatan sulfate epimerase 1- deficient mice have reduced content and changed distribution of iduronic acids

50 in dermatan sulfate and an altered collagen structure in skin." Mol Cell Biol 29(20): 5517-5528. McCormick, C., G. Duncan, et al. (2000). "The putative tumor suppressors EXT1 and EXT2 form a stable complex that accumulates in the Golgi apparatus and catalyzes the synthesis of heparan sulfate." Proc Natl Acad Sci U S A 97(2): 668-673. Mizumoto, S., T. Mikami, et al. (2009). "Chondroitin 4-O-sulfotransferase-1 is required for somitic muscle development and motor axon guidance in zebrafish." Biochem J 419(2): 387-399. Nadanaka, S., S. Kagiyama, et al. (2013). "Roles of EXTL2, a member of the EXT family of tumour suppressors, in liver injury and regeneration processes." Biochem J 454(1): 133-145. Nadanaka, S. and H. Kitagawa (2008). "Heparan sulphate biosynthesis and disease." J Biochem 144(1): 7-14. Nadanaka, S. and H. Kitagawa (2014). "EXTL2 controls liver regeneration and aortic calcification through xylose kinase-dependent regulation of glycosaminoglycan biosynthesis." Matrix Biol 35: 18-24. Nadanaka, S., S. Zhou, et al. (2013). "EXTL2, a member of the EXT family of tumor suppressors, controls glycosaminoglycan biosynthesis in a xylose kinase- dependent manner." J Biol Chem 288(13): 9321-9333. Nissen, R. M., A. Amsterdam, et al. (2006). "A zebrafish screen for craniofacial mutants identifies wdr68 as a highly conserved gene required for endothelin-1 expression." BMC Dev Biol 6: 28. Norton, W. H., J. Ledin, et al. (2005). "HSPG synthesis by zebrafish Ext2 and Extl3 is required for Fgf10 signalling during limb development." Development 132(22): 4963-4973. Nusslein-Volhard, C. and R. Dahm (2005). Zebrafish. Tubingen, Germany, Oxford University Press. Orkin, R. W., R. M. Pratt, et al. (1976). "Undersulfated chondroitin sulfate in the cartilage matrix of brachymorphic mice." Dev Biol 50(1): 82-94. Orr, S. L., D. Le, et al. (2013). "A phenotype survey of 36 mutant mouse strains with gene-targeted defects in glycosyltransferases or glycan-binding proteins." Glycobiology 23(3): 363-380. Pacheco, B., A. Malmstrom, et al. (2009). "Two dermatan sulfate epimerases form iduronic acid domains in dermatan sulfate." J Biol Chem 284(15): 9788-9795. Pallerla, S. R., R. Lawrence, et al. (2008). "Altered heparan sulfate structure in mice with deleted NDST3 gene function." J Biol Chem 283(24): 16885-16894. Peal, D. S., C. G. Burns, et al. (2009). "Chondroitin sulfate expression is required for cardiac atrioventricular canal formation." Dev Dyn 238(12): 3103-3110. Pinhal, M. A., B. Smith, et al. (2001). "Enzyme interactions in heparan sulfate biosynthesis: uronosyl 5-epimerase and 2-O-sulfotransferase interact in vivo." Proc Natl Acad Sci U S A 98(23): 12984-12989. Prabhakar, V. and R. Sasisekharan (2006). "The biosynthesis and catabolism of galactosaminoglycans." Adv Pharmacol 53: 69-115. Presto, J., M. Thuveson, et al. (2008). "Heparan sulfate biosynthesis enzymes EXT1 and EXT2 affect NDST1 expression and heparan sulfate sulfation." Proc Natl Acad Sci U S A 105(12): 4751-4756. Prydz, K. (2015). "Determinants of Glycosaminoglycan (GAG) Structure." Biomolecules 5(3): 2003-2022. Quarto, N. and F. Amalric (1994). "Heparan sulfate proteoglycans as transducers of FGF-2 signalling." J Cell Sci 107 ( Pt 11): 3201-3212.

51 Reuter, M. S., L. Musante, et al. (2014). "NDST1 missense mutations in autosomal recessive intellectual disability." Am J Med Genet A. Richardson, T. P., V. Trinkaus-Randall, et al. (2001). "Regulation of heparan sulfate proteoglycan nuclear localization by fibronectin." J Cell Sci 114(Pt 9): 1613- 1623. Ringvall, M. and L. Kjellen (2010). "Mice deficient in heparan sulfate N- deacetylase/N-sulfotransferase 1." Prog Mol Biol Transl Sci 93: 35-58. Ringvall, M., J. Ledin, et al. (2000). "Defective heparan sulfate biosynthesis and neonatal lethality in mice lacking N-deacetylase/N-sulfotransferase-1." J Biol Chem 275(34): 25926-25930. Robu, M. E., J. D. Larson, et al. (2007). "p53 activation by knockdown technologies." PLoS Genet 3(5): e78. Rong, J., H. Habuchi, et al. (2001). "Substrate specificity of the heparan sulfate hexuronic acid 2-O-sulfotransferase." Biochemistry 40(18): 5548-5555. Rossi, A., Z. Kontarakis, et al. (2015). "Genetic compensation induced by deleterious mutations but not gene knockdowns." Nature 524(7564): 230-233. Roughley, P. J. and E. R. Lee (1994). "Cartilage proteoglycans: structure and potential functions." Microsc Res Tech 28(5): 385-397. Saigoh, K., T. Izumikawa, et al. (2011). "Chondroitin beta-1,4-N- acetylgalactosaminyltransferase-1 missense mutations are associated with neuropathies." J Hum Genet 56(2): 143-146. Samson, S. C., T. Ferrer, et al. (2013). "3-OST-7 regulates BMP-dependent cardiac contraction." PLoS Biol 11(12): e1001727. Sato, T., M. Gotoh, et al. (2003). "Differential roles of two N- acetylgalactosaminyltransferases, CSGalNAcT-1, and a novel enzyme, CSGalNAcT-2. Initiation and elongation in synthesis of chondroitin sulfate." J Biol Chem 278(5): 3063-3071. Schilling, T. F., T. Piotrowski, et al. (1996). "Jaw and branchial arch mutants in zebrafish I: branchial arches." Development 123: 329-344. Schilling, T. F., T. Piotrowski, et al. (1996). "Jaw and branchial arch mutants in zebrafish I: branchial arches." Development 123: 329-344. Schroder, E., L. Gebel, et al. (2012). "Human PAPS Synthase Isoforms Are Dynamically Regulated Enzymes with Access to Nucleus and Cytoplasm." PLoS ONE 7(1): e29559. Schwartz, N. B., L. Roden, et al. (1974). "Biosynthesis of chondroitin sulfate: interaction between xylosyltransferase and galactosyltransferase." Biochem Biophys Res Commun 56(3): 717-724. Stelzer, C., A. Brimmer, et al. (2007). "Expression profile of Papss2 (3'- phosphoadenosine 5'-phosphosulfate synthase 2) during cartilage formation and skeletal development in the mouse embryo." Dev Dyn 236(5): 1313-1318. Stickens, D., B. M. Zak, et al. (2005). "Mice deficient in Ext2 lack heparan sulfate and develop exostoses." Development 132(22): 5055-5068. Sugahara, K. and H. Kitagawa (2000). "Recent advances in the study of the biosynthesis and functions of sulfated glycosaminoglycans." Curr Opin Struct Biol 10(5): 518-527. Sugahara, K. and H. Kitagawa (2002). "Heparin and heparan sulfate biosynthesis." IUBMB Life 54(4): 163-175. Sugahara, K. and T. Mikami (2007). "Chondroitin/dermatan sulfate in the central nervous system." Curr Opin Struct Biol 17(5): 536-545. Sugahara, K., T. Mikami, et al. (2003). "Recent advances in the structural biology of chondroitin sulfate and dermatan sulfate." Curr Opin Struct Biol 13(5): 612-620.

52 Swaim, L. E., L. E. Connolly, et al. (2006). "Mycobacterium marinum infection of adult zebrafish causes caseating granulomatous tuberculosis and is moderated by adaptive immunity." Infect Immun 74(11): 6108-6117. Tabata, T. and Y. Takei (2004). "Morphogens, their identification and regulation." Development 131(4): 703-712. Takahashi, I., N. Noguchi, et al. (2009). "Important role of heparan sulfate in postnatal islet growth and insulin secretion." Biochem Biophys Res Commun 383(1): 113-118. Thacker, B. E., D. Xu, et al. (2014). "Heparan sulfate 3-O-sulfation: a rare modification in search of a function." Matrix Biol 35: 60-72. Tian, J., L. Ling, et al. (2010). "Loss of CHSY1, a secreted FRINGE enzyme, causes syndromic brachydactyly in humans via increased NOTCH signaling." Am J Hum Genet 87(6): 768-778. Tian, J., L. Ling, et al. (2010). "Loss of CHSY1, a Secreted FRINGE Enzyme, Causes Syndromic Brachydactyly in Humans via Increased NOTCH Signaling." The American Journal of Human Genetics 87(6): 768-778. Topczewski, J., D. S. Sepich, et al. (2001). "The zebrafish glypican knypek controls cell polarity during gastrulation movements of convergent extension." Dev Cell 1(2): 251-264. Turnbull, J. E. and J. T. Gallagher (1991). "Sequence analysis of heparan sulphate indicates defined location of N-sulphated glucosamine and iduronate 2-sulphate residues proximal to the protein-linkage region." Biochem J 277 ( Pt 2): 297- 303. Varshney, G. K., W. Pei, et al. (2015). "High-throughput gene targeting and phenotyping in zebrafish using CRISPR/Cas9." Genome Res 25(7): 1030-1042. Venkatachalam, K. V. (2003). "Human 3'-phosphoadenosine 5'-phosphosulfate (PAPS) synthase: biochemistry, molecular biology and genetic deficiency." IUBMB Life 55(1): 1-11. Victor, X. V., T. K. Nguyen, et al. (2009). "Investigating the elusive mechanism of glycosaminoglycan biosynthesis." J Biol Chem 284(38): 25842-25853. Watanabe, Y., K. Takeuchi, et al. (2010). "Chondroitin sulfate N- acetylgalactosaminyltransferase-1 is required for normal cartilage development." Biochem J 432(1): 47-55. Watanabe, Y., K. Takeuchi, et al. (2010). "Chondroitin sulfate N- acetylgalactosaminyltransferase-1 is required for normal cartilage development." Biochemical Journal 432(1): 47-55. Westling, C. and U. Lindahl (2002). "Location of N-unsubstituted glucosamine residues in heparan sulfate." J Biol Chem 277(51): 49247-49255. Wilson, D. G., K. Phamluong, et al. (2012). "Chondroitin sulfate synthase 1 (Chsy1) is required for bone development and digit patterning." Dev Biol 363(2): 413- 425. Wilson, D. G., K. Phamluong, et al. (2012). "Chondroitin sulfate synthase 1 (Chsy1) is required for bone development and digit patterning." Developmental Biology 363(2): 413-425. Wiweger, M. I., C. M. Avramut, et al. (2011). "Cartilage ultrastructure in proteoglycan-deficient zebrafish mutants brings to light new candidate genes for human skeletal disorders." J Pathol 223(4): 531-542. Yada, T., M. Gotoh, et al. (2003). "Chondroitin sulfate synthase-2. Molecular cloning and characterization of a novel human glycosyltransferase homologous to chondroitin sulfate glucuronyltransferase, which has dual enzymatic activities." J Biol Chem 278(32): 30235-30247.

53 Yamada, S., K. Sugahara, et al. (2011). "Evolution of glycosaminoglycans: Comparative biochemical study." Commun Integr Biol 4(2): 150-158. Yan, D. and X. Lin (2009). "Shaping morphogen gradients by proteoglycans." Cold Spring Harb Perspect Biol 1(3): a002493. Yin, X., H. Cheng, et al. (2014). "Five regulatory genes detected by matching signatures of eQTL and GWAS in psoriasis." J Dermatol Sci 76(2): 139-142.

54

Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1143 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 Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine. (Prior to January, 2005, the series was published under the title “Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine”.)

ACTA UNIVERSITATIS UPSALIENSIS Distribution: publications.uu.se UPPSALA urn:nbn:se:uu:diva-264269 2015