Introduction 1.1 Post-Translational Modifications

Chapter I: Introduction

1.1 Post-translational modifications

Post-translational modifications (PTMs) are responsible for dynamic regulation of protein structure and function (Jensen et al., 2006). One example of the function of post-translational modifications in the dynamic regulation of protein structure and function is signal transduction; post-translational modification of key regulatory proteins in signal transduction pathways controls cellular proliferation, metabolism, gene expression, cytoskeletal organization and apoptosis (Jensen et al., 2006).

Post-translational modifications not only change the mass of the protein, they are also often responsible for changes in charge, structure and conformation. This leads to changes in the biological activity of proteins like enzymes, the binding affinity of proteins to receptors and other protein-protein interactions (Hoffmann et al., 2008) (Murray et al., 1998). For example, the interaction of the cytokine TGF-beta with its receptor complex results in the phosphorylation of SMAD family proteins which activates these proteins to regulate target gene expression (Massague et al., 2000). Post-translational modified amino acid residues can act as binding sites on proteins for a specific recognition domain on other proteins. For example, phosphorylated tyrosine residues can bind to SH2 or PTB domains (Pawson et al., 1997). Depending on the type of PTM they can be very abundant and have a large number of target proteins, whereas other PTM`s have only a few target proteins. Moreover, some PTM`s are connected with several different residues and not only with one residue. For example multi-site phosphorylation that primarily targets Serine, Threonine and Tyrosine residues is a crucial mechanism for the regulation of protein localization and functional activity (Cohen et al., 2000).

In the following table 1 most commonly studied post-translational modifications and their biological functions are highlighted.

1 Post-translational modification Function Acetylation Regulates protein-protein and protein-ligand interactions as well as protein function Leads to protein stability through the protection of the N-terminus from proteolytic degradation (Han et al., 1992) (Seo et al., 2004) Acylation Mediator of protein-protein and protein- membrane interaction Directs cytosolic proteins to membranes and activates them (Beck et al., 2006) Amidation Regulates intracellular signal transduction, DNA repair and replication, transcription and apoptosis Responsible for protein stability through the protection of the C-terminus and for ligand- receptor interactions (Han et al., 1992) Simple glycosylation Histone remodelling, cell growth and division, gene transcription, apoptosis and proteasomal degradation Regulates cellular responses to hormones and initiates a protective response to stress (Slawson et al., 2006) (Ohtsubo et al., 2006) Complex glycosylation Promotes or inhibits intra- or intermolecular protein interactions Involved in protein folding and intracellular trafficking Involved in cell-cell communication, molecular and cellular homeostasis, receptor activation, signal transduction and endocytosis (Walsh et al., 2006) Hydroxylation Mediates structural stability, hypoxic sensing and blood coagulation

2 (Zagorska et al., 2004) Methylation Methylation on lysine regulates chromatin structure, activation and repression of gene transcription Arginine methylation is involved in transcriptional regulation and signal transduction (Lee et al., 2005) Nitrosylation Mediates redox based signalling Modulates energy metabolism Can inhibit or promote apoptosis (Foster et al., 2003) (Spickett et al., 2006) Oxidation Associated with protein damage

Table 1: Common post-translational modifications and their functions. Table adapted from

Hoffman et al., 2008

Phosphorylation is the most commonly occurring form of post-translational modification with up to 30% of all proteins phosphorylated (Cohen et al., 2001). It is the objective of this work to study how phosphorylation of cellular proteins impact on the generation, loading and presentation of peptide antigens to cytotoxic T lymphocytes. Thus phosphorylation will be discussed in more detail in the next sections.

1.2 Phosphorylation

With the discovery of rabbit skeletal muscle glycogen phosphorylase, it is now half a century since phosphorylation has been identified as a post-translational modification in eukaryotes (Cozzone et al., 2004), (Krebs et al., 1956). However, the detection of phosphorylation in prokaryotes took another 20 years (Cozzone et al., 1988). Unlike to prokaryotes, where

3 phosphorylation is mostly at histidine, glutamic acid and aspartic acid residues, in eukaryotes, phosphorylation is based on serine, threonine and tyrosine. It is known to be an essential event of signal transduction, activation or inactivation of enzyme activity. The phosphorylation of serine and threonine takes place more abundantly than the phosphorylation of tyrosine with a ratio of 1800 pSer : 200 pThr : 1 pTyr in eukaryotic cells (Mann et al., 2002). Phosphorylation of peptides / proteins is reversible and catalytic where the phosphorylation / dephosphorylation are done by kinases and phosphatases respectively. There are about 500 kinases and around 100 phosphatases encoded in the human genome (Salih et al., 2005), (Cohen et al., 2001).

The role of phosphorylation within the cell is enormous. It is linked with regulating cell growth, survival, apoptosis, receptor activation and the corresponding signalling, regulating metabolic pathways and enzyme activity. Phosphorylation is able to regulate protein activities on several levels: a) as a structural element that is necessary for correct protein folding; b) by inducing conformational changes that can result in an increase or inhibition of enzyme activity; c) by acting as docking sites for the assembly of complexes with other proteins; d) as recognition signal for further modification and e) by changing subcellular localization (Lochhead et al., 2005) (Preisinger et al., 2008).

Signalling is one of the most important function where phosphorylation of proteins is involved. Here much attention has been focussed on receptor tyrosine kinases (RTK). Activated RTK autophosphorylates tyrosine residues in their intracellular kinase domain. These phosphorylation sites act as docking sites for adaptor proteins that bind to the phosphotyrosines through interaction domains: for instance Src homology domain 2 (SH2) or phosphotyrosine-binding (PTB) domains. The adaptors interact with further binding partners to build a multiprotein signalling complex at the receptor. This regulates distribution, strength and duration of the incoming signal (Kolch et al., 2000); (Pawson et al., 1994); (Seet et al., 2006). One well characterized example is the epidermal growth factor (EGF) receptor signalling pathway.

New tools based around phosphopeptide enrichment and high resolution mass spectrometry have provided tremendous new insights into phosphorylation of proteins following various stimuli and allowed studying phosphorylation events at both the local (a single protein or signalling pathway) and global level.

4 1.3 Enrichment and analysis of phosphopeptides / phosphoproteins

In recent years, liquid chromatography coupled with mass spectrometry (LC–MS/MS) has become the first choice of analysis in phosphoproteomics. Although 32P-labeling and Edman sequencing have also been used for the characterization of phosphorylation sites (Mann et al., 2002), these methods are more time consuming than MS. Mass spectrometry has a number of advantages over other methods including (a) high sensitivity, (b) the site of the phosphorylation can be unequivocally identified, (c) novel posttranslational modifications can be found, (d) phosphorylation sites in complex mixtures of proteins can be identified, and (e) relative changes in post-translational modification tenure at distinctive sites can be quantified (Larsen et al., 2006). However, prior to analysis it is necessary to enrich phosphopeptides / phosphoproteins bearing in mind the low stoichiometry of phosphorylated proteins / peptides (only a small portion is phosphorylated in the cell) and the reversibility of the modification (Mann et al., 2002). Outlined in this section are some of the methods of enrichment of phosphopeptides (Figure 1).

Figure 1. Enrichment methods for phosphopeptides/phosphoproteins.

Figure adapted from Reinders et al. (2005). State-of-the-art in phosphoproteomics. Proteomics 5: 4052-4061

5 1.3.1 Affinity chromatography

Affinity chromatography is based on resins that interact with the phosphate group of the peptide. Several resins exist, for example immobilized metal affinity chromatography

(IMAC), titanium dioxide (TiO2) and zirconium dioxide (ZrO2) chromatography (Jensen et al., 2007). Of these, IMAC is the method most commonly used to enrich phosphopeptides and proteins (Larsen et al., 2005) and was first set up by Porath et al. (Porath et al., 1975). Although the metal ions Fe3+, Ga3+ and Al3+ strongly bind with phosphopetides, Fe3+ -based techniques are more popular. Phosphopeptides bind to the metal ions that are coupled to the column material by iminodiacetic acid (IDA), nitriloacetic acid (NTA) or Tris- (carboxymethyl) - ethylendiamine (TED) linkers, via electrostatic interactions. They are then eluted by salt- and/or pH- gradients after removing non-phosphorylated peptides with a wash step. To avoid very acidic nonphosphorylated peptides from binding to metal ions, the acidic side chains are modified with HCL- saturated, dried methanol (methylesterification) prior to IMAC (Reinders et al., 2005).

Recently, TiO2 was established by Pinkse et al. (Pinkse et al., 2004) as an alternative to IMAC for enrichment of phosphopeptides (Thingholm et al., 2006). TiO2 can better tolerate low pH solutions, detergents, salts and other contaminants (Jensen et al., 2007).

Moreover, in a study it was found to be more efficient than IMAC (Cantin et al., 2007). Larsen et al. (Larsen et al., 2005) demonstrated that a higher percentage of nonphosphorylated peptides were detected after purification with IMAC rather than with the TiO2 chromatography. This indicates a more specific binding of the phosphorylated peptides on the

TiO2 column.

1.3.2 Antibodies

Antibodies that immunoprecipitate phosphoproteins / phosphopeptides, are commercially available. Antibody enrichment is mainly used to precipitate proteins that are phosphorylated on tyrosine residues. After precipitation, the phosphopeptides can be identified by reverse- phase liquid chromatography – tandem mass spectrometry (LC–MS/MS) (Rush et al., 2005). Immunoprecipitation can be carried out with the methylesterification/IMAC method to

6 increase the number of phosphopetides (Rush et al., 2005). Although antibodies for phosphoserine- and phosphothreonine containing peptides are also available, they are seldom used because of their high costs.

1.3.3 Chemical methods

These methods are based on the chemical modification of the phosphate group. Two notable techniques are ß-elimination and a transient carbodiimide catalyzed addition of cystamine to phosphate groups that was introduced by Zhou et al. (Zhou et al., 2001) (Mann et al., 2002). ß-elimination happens under strongly alkaline conditions and the arising dehydroalanine or dehydroaminobutyric acid residues can be identified by tandem mass spectrometry (MS/MS). Zhou’s method involves the modification of the phosphate group and permits the purification of phosphopeptides on glass beads that have immobilized iodoacetyl groups.

The disadvantage of these chemical methods is that they need a large amount of sample (Mann et al., 2002).

1.4 MHC molecules

A relatively new area of studies on phosphopeptides refers to their ability to act as peptide antigens by binding to MHC (major histocompatibility complex) molecules for presentation to T lymphocytes which is discussed in these work. MHC-peptide complexes are objects of the cell-mediated immune response where the peptide is presented on the surface of an antigen presenting cell to the T cell receptor of a cytotoxic T cell or T helper cell (Figure 2).

The first structure of a MHC molecule, HLA A2, was described in 1987 by Bjorkmann et al. (Bjorkmann et al., 1987).

There are two families of MHC molecules: MHC class I molecules which present antigens to cytotoxic T cells that express the CD8 cell-surface protein and MHC class II molecules which

7 present the antigen to helper T cells that express the CD4 cell-surface protein (Parnes et al., 1989); (Teh et al., 1988).

MHC class I molecules present antigens derived from endogenously degraded proteins, whilst MHC class II molecules present antigens from exogenously synthesized proteins and endogenous proteins that intersect the secretory pathway (Townsend et al., 1989); (Unanue et al., 1987).

In this thesis I will focus on phosphopeptide presentation by MHC class I molecules.

MHC class I molecules consist of a polymorphic heavy chain and a monomorphic light chain, known as beta 2 microglobulin. The heavy chain contains three domains, two similar folded domains α1 and α2 and a α3 domain. The α3 domain and beta 2 microglobulin have an immunoglobulin fold and lie proximal to the cell membrane. The α1 and α2 domains are not related to the immunoglobulin fold. They form a superdomain that sits on top of the α3 domain and the beta 2 microglobulin subunit. (Haimes and Glover, 1996)

Figure 2. Crystal structure of the peptide- (presented by the MHC molecule) T cell receptor complex. Figure from Hennecke et al. (2001). T Cell Receptor-MHC Interactions up Close. Cell 104: 1-4

8 The α1 and α2 domains consist of 2 helices, one from each domain, which form a groove approximately 30 Å long and 12 Å wide. The bottom of this groove is formed by an 8 stranded anti-parallel ß-sheet, contributed of 4 strands from each domain (Madden et al., 1995). This groove is responsible for the binding to antigenic peptides.

MHC class I molecules are highly polymorphic with the majority of the polymorphism focussed within the α1 and α2 domains. This polymorphism engenders different peptide binding specificities with each MHC class I allotype binding thousands of different peptides.

The peptide itself comprises so-called anchor residues which are identical among the subset of peptides for one MHC molecule. For example the preferred amino acids (primary anchor residues) of peptides for HLA A2 are Leucine at position 2 and Valine at position 9 (Madden et al., 1995). Peptide binding to the MHC molecule and subsequent presentation to T cells is well understood. There are well defined differences in the MHC class I – and MHC class II pathways but this work focuses on the MHC class I processing pathway.

1.5 MHC class I processing pathway

The activation of the T cell depends on the ability of the antigenic peptide to bind to the MHC class I molecule as well as the affinity of the peptide-MHC complex for the T cell receptor (TCR) expressed on the surface of the T lymphocyte (Parmiani et al., 2002). Typically peptides of 8-11 amino acids in length bind to the MHC class I molecule and are recognized by the TCR (Figure 3) of the cytotoxic T cell, although longer peptides are also known to bind (Purcell et al., 2003), (Tynan et al., 2005).

In the MHC class I pathway, peptides usually originate from intracellular proteins or proteins that find their way into the cytosol of the cell, e.g. viral antigens. The proteins are degraded through the proteolytic actions of the proteasome, translocated via TAP (Transporter associated with Antigen Processing) to the endoplasmic reticulum and loaded onto the MHC class I molecule. The MHC – peptide – complex is then transported to the cell surface where the peptide is presented to cytotoxic T lymphocytes. However, exogenous proteins can enter the MHC class I processing pathway via phagocytosis. In addition, endocytosed antigens may be transported out of the endoplasmic reticulum, and are degraded through the proteasome so that the peptides can enter the MHC class I pathway (Figure 3) (Anderson et al., 2006).

9 Much research has been performed over the last three decades to understand more about these peptide antigens and their connection with disease. Here the focus lies on cancer and tumor associated antigens connected with cancer and potential cancer immunotherapy.

Figure 3. MHC class I processing pathways. Usually peptides originate from intracellular proteins (a). Exogenous proteins also enter the MHC class 1 processing pathway through endocytosis/phagocytosis (b). Endocytosed antigens may be transported out of the endoplasmic reticulum and will be degraded through the proteasome and the peptides can enter the normal pathway as well. Figure from Andersen et al. (2006). Cytotoxic T cells. Journal of Investigative Dermatology 126: 32-41

1.6 Tumor Antigens and Cancer Immunotherapy

In recent years the fight against cancer has made use of immunotherapy. Immunotherapy is a method to combat cancerous cells by the stimulation of the patients´ immune system. In contrast to chemotherapy, immunotherapy takes advantage of the natural immune response and does not have the harmful side-effects seen with chemotherapeutics or radiation therapy.

10 The induction of an immune response is one way of killing a cancerous cell. While the humoral immune response is based on antibodies, the cell-mediated immune response is based on cytotoxic T cells and helper T cells. However, both immune responses have the same objective namely removal of the antigen bearing cancerous cell (Abbas et al., 2004).

Cancer immunotherapy that induces cytotoxic T lymphocytes, focuses on the use of full length protein antigens and/or peptides (epitopes) derived from tumor-associated antigens (TAA) to create vaccines to eradicate tumors. Many different tumor associated antigens have been discovered. They are divided into different classes such as (i) cancer testis antigens with Mage proteins which were the first characterized tumor associated antigens (cancer testis antigens are the most hopeful antigens for immunotherapy because of their broad tumor specific expression), (ii) tumor-specific shared antigens, (iii) tumor-specific unique antigens, (iv) overexpressed antigens such as human epidermal growth factor receptor 2 (HER2)/neu and mucin 1 found at low levels in normal tissues and high levels in tumors and (v) viral antigens, derived from oncogenic viruses such as papilloma virus (HPV), hepatitis B virus (HBV) and Epstein-Barr virus (EBV) (Paschen et al., 2004).

Tumor associated antigens can be discovered through several methods:

a) genetic approaches where bacterial generated pools of cDNA are cotransfected with a MHC I gene construct into a highly transfectable eukaryotic cell line. These cells are then screened for their ability to stimulate cytokine release from tumor reactive cytotoxic T lymphocytes (Rosenberg, Philadelphia:Lippincott 2000). From single clones, peptide epitopes derived from these antigens are identified through the use of HLA peptide binding motif algorithms. Examples of TAAs identified through this method are MAGE, BAGE, GAGE, MART-1, gp100, tyrosinase and beta-catenin.

b) Serologic analysis of recombinant tumor cDNA libraries (SEREX). For this method cDNA libraries are prepared from tumor specimens and cloned into λ phage expression vectors and then the recombinant phage is transfected into Escherichia coli (Tureci et al., 1997). Using patient sera, the cDNA libraries are immuno-screened for tumor antigens that elicit a high-titer IgG response. Positive clones are isolated, screened and evaluated with a DNA database. A high-titer IgG response is also a proof for T cell recognition of the antigens by helper T cells and indirectly suggests CTL induction (Bright et al., 2002); (Jager et al., 1998); (Lewis et al., 2003). TAAs identified by this method are from small cell lung carcinoma (Gure et al., 2000), breast cancer (Jager et al., 1999), renal cell carcinoma (Scanlan

11 et al., 1999), colon cancer (Scanlan et al., 1998) and esophageal carcinoma (Chen et al., 1997). For example, the TAA NY-ESO-1 cloned from esophageal carcinoma is shared between different tumors and was able to induce MHC I-restricted CTL (Jager et al., 1998). Other examples are NY-CO 1-48, SSX2, MAGE and tyrosinase.

c) Isolation of peptides from tumor cell surface MHC class I molecules where pools of peptides are produced, fractionated with high pressure liquid chromatography and sequenced (Slingluff et al., 1994); (Falk et al., 1991).

d) Serial analysis of gene expression (SAGE). With SAGE the expression pattern of many genes can be evaluated without sequence information. There are two main concepts with (i) a short sequence tag of 9-10 base pairs to identify a transcript; (ii) serial units of short sequence tags are sequenced within a single clone. The ability of candidate antigens of stimulating peripheral blood lymphocytes (PBL) to generate cytotoxic T lymphocytes in vitro that are capable of killing the tumor cells is tested by using the sequence information of the cloned differentially expressed genes (Lewis et al., 2003).

Peptides derived from tumor-associated antigens are created and presented to cytotoxic T lymphocytes via the MHC class I processing pathway. Experimental and clinical data have shown that cytotoxic T cells are able to kill cancerous cells and, therefore, these T cells might play a central role in immunotherapy. However, other immune effector cells such as natural killer (NK) cells, lymphokine activated killer (LAK) cells, activated macrophages and granulocytes are also known to have antitumor effects in vitro (Purcell et al., 2007), (Ioannides et al., 1993), (Paschen et al., 2004).

A common property of TAA is their poor immunogenicity. One solution to overcome this problem is modifications within the TAA which results in so called altered peptide ligands.

There are several approaches to modify peptide antigens to engender improved immunogenicity. One method involves substitutions of amino acids at anchor positions to enhance MHC binding. Valmori et al. used the peptide analogue Melan A26-35 A27L where the alanine residue was substituted with leucine at position 2. This peptide analogue showed a more stable complex formation with HLA A 0201 and was subsequently recognized more efficiently by Melan A – specific CTL clones (4-1000 fold) (Valmori et al., 1998).

Parkhurst et al. used the melanoma self antigen gp100 for their studies. There are 5 known gp100 epitopes of which 3 bind with intermediate affinity to HLA A 0201 molecules which may be due to the fact that each peptide contains only one preferred amino acid at one of the two anchor positions of HLA A 0201. For example with substitution of the secondary amino 12 acid at position 2 of G9154 (KTWGQYWQV) with Leucine, resulted in a high binding affinity for HLA A 0201 ( Parkhurst et al., 1996).

Other methods for enhancement of immunogenicity of tumor associated antigens are introduction of aromatic amino acids at position 1, 4 and 5 of the peptide, terminal modification, substitution of cysteine residues, modification of TCR interacting amino acid residues and modifications of individual amino acid residues (Tourdot et al., 2000) (Chen et al., 2000) (Parmiani et al., 2002).

Despite the widespread use of rational chemical modifications of peptide vaccine constituents (Purcell et al., 2007), not much is known about the role of natural post translational modifications and their impact on the immunogenicity of tumor- associated antigens.

1.7 Vaccine strategies

Recent vaccine strategies include DNA, recombinant virus based vaccines, tumor cell vaccines, oncoprotein vaccines, receptor targeted vaccines and peptide/protein based vaccines. DNA vaccines have demonstrated utility in animal models, however there has not been significant effectiveness in human clinical trials so far. For both recombinant virus based vaccines and whole tumor cells there is evidence of enhanced immune responses. The disadvantage of whole tumor cell vaccines is their risk to induce complex immune responses and/or hypersensitivity (autoimmune reactions) (Lazoura et al., 2005). An alternative to above mentioned vaccine possibilities is the design of peptide-based vaccines. One method using peptide-based vaccines is linked with the use of dendritic cells (DCs) which are very potent antigen presenting cells. For dendritic cells pulsed with peptide antigens an induction of antitumor immunity has been observed in several cases. A recent study showed that patients treated with dendritic cell pulsed peptides resulted in a tumor regression in 5 of 16 patients evaluated (Wang et al., 1999) (Nestle et al., 1998) (Ockert et al., 1999) (Timmermann et al., 1999). Some examples for current clinical trials of peptide-based vaccines are human epithelial growth factor receptor 2 (HER-2/neu), human telomerase reverse transcriptase (hTERT), melanoma associated antigens (MAGE3, Melan-A/MART1, tyrosinase, gp100), human papillomavirus (HPV) antigens, carcinoma embryonic antigen (CEA) and mucin-1 (MUC1)

13 (Apostolopoulos et al, 1999) (Jager et al., 2002) (Parmiani et al., 2002) (Rosenberg, 2001) (Waldmann et al., 2003).

1.8 Post- translational modified MHC peptides

With the application mass spectrometry to identify MHC bound peptide antigens, it became possible to accurately detect and define post-translational modified peptides. The first detected modification was deamidated asparagines that resulted in aspartic acid of a peptide derived from tyrosinase (Skipper et al., 1996) followed by peptides containing cysteine residues which have been changed by the addition of another cysteine residue via a disulfide bond, i.e. cysteinylation (Meadows et al., 1997); (Kittlesen et al., 1998); (Pierce et al., 1999). Post-translational modifications found in MHC class I- and MHC class II restricted peptides are amino-terminal acetylation and O-linked glycosylation whereas N-linked glycosylation, nitration, deamidation and deimination have been found only in MHC class II-restricted peptides (Engelhardt et al., 2006).

Much attention has been focussed on phosphorylation. One important factor for this is that in contrast to normal cells, phosphorylation of cellular proteins is significantly different in malignant cells (Cantley et al., 1991); (Hunter 1991). This is true not only for cancer but for other diseases such as diabetes and rheumatoid arthritis (Cohen et al. 2001). Because the phosphorylation of cellular proteins is altered in cancerous cells, the repertoire of peptides derived from these cellular proteins and presentation on the cell surface by MHC class I molecules is also different (Zarling et al., 2006). Phosphorylated peptides are processed and presented by the MHC class I molecule and recognized by cytotoxic T cells (Zarling et al., 2000), (Andersen et al., 1999). Zarling et al. (2006) identified MHC class I phosphopeptides derived from cellular phosphoproteins in cancerous cells using the IMAC-LC-MS/MS approach described above. In this study, three different cancer cell lines (two melanoma cells, DM331 and SLM2 and an ovarian carcinoma, COV413) and an Epstein-Barr virus- transformed B lymphoblastoid cell line (BLCL), JY were used as sources of malignant and “control” phosphopeptides respectively. A number of HLA A2 restricted phosphopeptides were found on one or more cancer cell line but not on the JY BLCL “control” cell line. The 14 source of these tumor specific phosphopeptides was mainly associated with proteins involved in cytoplasmic signalling and cellular transformation. Tumor specific peptide antigens are of special interest as potential peptide based vaccine targets. Moreover HLA A2 restricted CD8+ T lymphocytes were generated which recognize phosphopeptides derived from IRS2, ß- Catenin and CDC25b but not their nonphosphorylated counterparts (Zarling et al., 2006).

Based on these findings we and others have investigated the structural basis of phosphopeptide antigen presentation. In 2008 Mohammed et al. described the structure of phosphopeptides bound to the human MHC class I molecule HLA A2 (Mohammed et al., 2008). They solved structures of peptide ligands with an Arginine at position 1 and a phosphorylated serine at position 4 of the peptide. In all cases the phospho- group of the peptide was solvent exposed and therefore was directly accessible to the T cell receptor. They also found that the negatively charged phospshorylated serine residue forms stabilizing interactions with several MHC residues. Because of that results they introduced the term `phosphate surface anchor`. These interactions also resulted in an increase in binding of the phosphopeptides to the MHC molecule compared to their nonphosphorylated counterparts. This enhancement in class I binding was dependent upon the arginine at position 1 of the peptide.

Our study further investigates the binding of additional phosphopeptides to HLA A2 and importantly includes structures of both phosphorylated and nonphosphorylated peptide antigens with varying N- termini and p4 or p5 phosphoserine residues.

For this study phosphopeptides derived from IRS2, ß-Catenin and CDC25b were chosen for structural studies because of their specific recognition by T lymphocytes and their presence on three different cancer cell lines (Zarling et al., 2006). They also differ from the peptides examined by the earlier study by Mohammed et al. in that:

(i) They include phosphopeptides with different amino acid residues at position 1 (other than arginine) to test whether other amino acids at position 1 of the peptide may also have an influence on phosphopeptide binding to the MHC molecule.

(ii) One studied phosphopeptide is phosphorylated at position 5 of the peptide to test how a different position of the phospho moiety may result in enhanced phosphopeptide-MHC complex stability.

Another point for the discrimination of this study from Mohammed et al. is that not only the phosphopeptide-MHC complex structures but also the structures of their nonphosphorylated

15 counterparts have been solved to directly gain insights into potential structural differences due to the presence of the phospho- group.

In contrast to MHC class I molecules, which bind peptides of 8-12 amino acids in length derived from mainly endogenously proteins, MHC class II molecules bind peptides that are mostly derived from exogenously proteins. For class II molecules these peptides vary in length from 9-25 amino acids. Recent studies have found that MHC class II molecules also bind phosphopeptides which are presented to CD4+ T cells (Meyer et al., 2009) (Depontieu et al., 2009) . This is a remarkable finding because not only cytotoxic T cells which recognize antigens presented by MHC class I molecules but also CD4+ T cells play an important role in the cellular immune response. CD4+ T cells are responsible for the induction and maintenance of cytotoxic T cells through activation of the antigen presenting cell and secretion of cytokines. Therefore both cytotoxic T cells as well as CD4+ T cells are necessary for antitumor immunity. Especially the finding that 2 phosphoproteins, tensin 3 and insulin receptor substrate 2 (IRS2), are the source for MHC class I- as well as MHC class II restricted phosphopeptides may provide an opportunity of a combinatorial treatment strategy using MHC class I- and MHC class II phosphopeptides to generate tumor immunity. This would be congruent to clinical trials with MART-1, a melanoma antigen, which contains both MHC class I- and MHC class II restricted peptide epitopes (Depontieu et al, 2009) (Zarling et al., 2006). Also recently the structure of a phosphopeptide derived from melanoma antigen recognized by T cells-1 (pMART-1) in complex with the MHC class II molecule HLA-DR1 has been solved (Li et al., 2010). MART-1 has been target for clinical trials not only because peptides derived from MART-1 are both MHC class I- and MHC class II restricted but also because of its selective expression on melanoma cells. The crystal structure of pMART100-114 in complex with HLA-DR1 showed, as also reported in this thesis, a solvent exposure of the phosphorylated serine at position 5 of the peptide. But other than for the phosphopeptide- MHC class I structures there are no direct contacts between the phospho-group and the peptide or HLA-DR1. However, the phospho-group of the peptide is essential for T cell recognition: DR1-expressing antigen presenting cells presented the phosphorylated MART-1 peptide to CD4+ T cell clones specific for MART-1 which in response secreted the cytokine granulocyte-macrophage colony-stimulating factor (GM-CSF). But no secretion has been observed for the nonphosphorylated peptide (Li et al., 2010).

16 Chapter II: Materials and methods

2.1 Media and buffer

2.1.1 Media

2.1.1.1 Luria broth (LB) media

10g NaCl 5g Yeast extract 10g Tryptone Make it up to 1 l with Milli Q and autoclave

2.1.1.2 Terrific Broth media

12g Tryptone 24g Yeast extract 4ml Glycerol Adjust volume to 900 ml with Milli Q

90ml Milli Q

2.31g KH2PO4

12.54g K2HPO4 Adjust volume to 100 ml

Autoclave both solutions separately, allow to cool and mix

17 2.1.2 Buffer

2.1.2.1 Resuspension buffer

50mM Tris-HCl pH 8.0 25% Sucrose 1mM EDTA 10mM DTT Add DTT just before use

2.1.2.2 Lysis buffer

50mM Tris-HCl pH 8.0 1% v/v Triton X-100 1% w/v Sodium deoxycholate 100mM NaCl 10mM DTT Add DTT just before use

2.1.2.3 Wash buffer 1

50mM Tris-HCl pH 8.0 0.5% v/v Triton X-100 1mM EDTA 1mM DTT 1xPMSF 1xPepstatin A

18 Keep it at 4°C Add DTT, PMSF and Pepstatin A just before use

2.1.2.4 Wash buffer 2

50mM Tris-HCl pH8.0 2.5mM EDTA 1mM DTT 1.5xPepstatin A 1.5xPMSF Keep it at 4°C Add DTT, Pepstatin A and PMSF just before use

2.1.2.5 Urea buffer

25mM MES 8M Urea 10mM EDTA 1mM DTT Add DTT just before use

2.1.2.6 Refold buffer

100mM Tris-HCl pH8.0 2mM EDTA pH8.0 23.59g L-arginine-HCl

19 0.126g oxidised glutathione 0.63g reduced glutathione Make it up to 420 ml with MilliQ Store solution at 4°C prior to use

2.2 Methods

2.2.1 Transfection of E.coli BL21 with human ß2m and HLA A2 heavy chain and protein expression

E.coli BL21 cells were removed from -80°C and thawed on ice. 1µl of plasmid encoding either human ß2microglobulin or HLA A2 were added to 100 µl of E.coli BL21 cells, mixed well and incubated on ice for 30 minutes then heat shocked at 42 °C for 90 seconds and finally incubated on ice for 2 minutes. Then 1 ml of Luria Broth (LB) media was added to the cells, mixed well and incubated at 37°C for 60 minutes on a shaker. Cells are chloramphenicol resistant and the vectors are kanamycin resistant and 10 µl of chloramphenicol (34 mg/ml) and kanamycin (30 mg/ml) were added to the cells with LB media. 100µl and 200µl of suspension were added to LB agar plates, spread using a sterile glass pipette and incubated at 37°C overnight. Single colonies were picked from the plates and added to 10 ml of LB media containing 10 µl of 34 mg/ml chloramphenicol and 10 µl of 30 mg/ml kanamycin and incubated overnight at 37°C on a shaker. A 10 ml preculture was added per 1 l of LB media and to induce protein expression 1 ml of

1M IPTG was added to 1 l or 500 ml of E.coli culture at an optical density (OD) 600nm of 0.9. After 3 hour or overnight incubation cells were spun at 4000 rpm for 10 minutes at 4°C with a centrifuge, cell pellets of 1 l culture were resuspended in 10 ml of resuspension buffer and frozen at –80°C.

20 2.2.2 Inclusion body preparation

Cells were removed from -80°C and thawed. Lysis buffer (2.5 ml per 1ml culture) and DNAse (1mg per 1 l culture) were added to the cells and lysed for 20min at room temperature. After lysis 10 mM EDTA was added, homogenised for 30 sec and spun for 15 min at 10000 rpm and 4°C with a Jasco centrifuge. The supernatant was discarded and the pellet resuspended in wash buffer 1 (150ml per 1 l culture), homogenised and spun for 15 min at 10000 rpm and 4°C. This step was repeated twice. After the third wash with buffer 1 the pellet was resuspended in wash buffer 2 (100ml per 1 l culture), homogenised and spun again for 15 min at 10000 rpm and 4°C. Supernatant was discarded and the pellet was finally resuspended in Urea buffer (10 ml for 1 l culture), homogenised and spun for 30 min at 15000 rpm and 4°C. Supernatant was collected and 1xPMSF and 1x pepstatin A added and frozen at -80°C.

2.2.3 Refolding of the HLA A2 peptide complexes

The refold buffer was stored in the cold room at 4°C and 800µl of PMSF (stock is 500x) and 800µl of Pepstatin A (stock is 500x) were added immediately prior to use. In the late afternoon, while stirring the bottle gently, first 15mg of peptide (peptide powder dissolved in 1-2ml DMSO), secondly 10mg of beta2m (from inclusion body preparation) and finally 15mg of HLA A2 heavy chain (from inclusion body preparation) were added to the refold buffer. The solution was left stirring at 4°C for 2 days.

21 2.2.4 Dialysis

After 2 days of refolding of the MHC heavy chain, ß2m and the nonphosphorylated or phosphorylated peptide, dialysis was undertaken. The dialysis buffer, 10 mM Tris pH 8.0, was made in the morning (8l for 420ml refold) and stored in the cold room at 4 °C. Dialysis membrane was soaked in Milli Q water. The refolded material was then placed into the dialysis tubing, and dialysed for 2 days with gentle agitation.

2.2.5 Ion exchange chromatography with Q sepharose column

The column was stored in 20 % EtOH. First step was the washing of the column by gravity with filtered MilliQ water and subsequently with 10 mM Tris pH 8.0. The column was then washed with 1 M NaCl in 10 mM Tris pH 8.0 to remove residual protein from the column. Finally the column was reequilibrated with 10 mM Tris pH 8.0. Next the column was loaded with the refold by gravity at room temperature. After off-line loading of the refolded protein, the column was connected to an AKTA HPLC to perform fraction collection (10 ml fractions) and step wise gradient elution at 2 ml/min. Buffer A was 10 mM Tris pH 8.0 and buffer B was 1 M NaCl in 10 mM Tris pH 8.0. Once a stable baseline was achieved running 100% buffer A, a step elution to 10% buffer B was performed and fractions collected. After elution of one protein to which PMSF and pepstatin A were added and again a stable baseline was reached, buffer B was stepped to 40 % and the elution was performed until the baseline was stable again. Finally buffer B was stepped on 100% to elute all other protein. Column was washed with Milli Q and 20 % EtOH in Milli Q and stored at 4 °C.

22 2.2.6 Molecular sieve chromatography

A 1ml loop was used for the concentrated fractions picked after the ion exchange chromatography. Fractions were selected by an SDS Page gel and concentrated using centrifugal filtration. The column was attached to the Akta system. After the column was washed with 150 mM NaCl in 10mM Tris pH 8.0 at 0.2 ml/min overnight, the concentrated fractions were loaded onto the loop, the Akta was changed to “inject” and the chromatography was started at 0.8 ml/min. After 30min the collection of the fractions (2 ml fractions) was started. After the baseline was reached again the chromatography was finished and the system and column was washed with 100 ml MilliQ at 0.8 ml/min. The column was washed overnight with 20% EtOH in MilliQ at 0.2 ml/min. The fractions were checked with SDS-PAGE for the refolded protein and purity.

2.2.7 Ion exchange chromatography using Mono Q Sepharose

The chromatography was performed under nearly same conditions as for the ion exchange chromatography using Q sepharose. It only differed in using a 1 ml loop for online loading of the fractions from the molecular sieve chromatography. Furthermore 5 ml fractions were collected at 1.0 ml/min.

2.2.8 Competition based peptide binding assay

Principle of measurement: MHC molecules are exposed on the surface of a cell line. Through citric acid treatment bound peptides are removed from the MHC binding pocket and therefore the pocket is free for MHC restricted antigens. The cells are incubated in the presence of the peptide of interest (competitor peptide) and a fluorescence labelled reference peptide. According to its binding

23 affinity the competitor peptide competes with the reference peptide in its binding to the MHC molecule. The mean fluorescence of the sample is measured by flow cytometry. As higher the fluorescence as more reference peptide is bound to the MHC molecule and as lower is the binding affinity of the peptide of interest. The competition based peptide binding assay was performed and percentage of inhibition of reference peptide was calculated with a formula mentioned in Burg et al (Burg et al., 1995). Briefly, 25 µl of competitor peptide (different end concentrations) was mixed with 25 µl fluorescence labelled reference peptide [GILGK(FITC)VFTL, end concentration = 150 ng/ml] in a 96-well V-bottom plate (For 48 wells competitor peptide and for 3 wells positive control reference peptide was added). Approximately 2000µl of reference peptide (900 ng/ml) were necessary. Silver foil was finally putted around the plate because the fluorescence labelled reference peptide is light sensitive. After mild acid-treatment 100µl of JY cells (5x104/well) were added into each well and incubated at 4°C for 24 hours. After incubation, cells were washed with PBS containing 1% BSA to remove unbound peptides. Finally 10µl of propidium iodide (1mg/ml) solution) was added and the mean fluorescence (MF) was measured by FACScan (Becton Dickinson).

2.2.9 Thermal denaturation using CD Spectroscopy

Principle of measurement: Plane polarised light is comprised of left and right circularly components. CD Spectroscopy refers to the different absorption of these two components. It occurs when a chromophore is chiral (optical active) when it is a) intrinsically chiral because of its structure or b) covalently linked to a chiral centre or c) placed in an asymmetric environment (Kelly et al, 2005). In case of proteins the CD spectrum is an expression of their secondary structure due to peptide bond, aromatic amino acid side chains and disulphide bonds as chromophores of interest (Kelly et al., 2005).

Figure 4. CD spectra from various types of secondary structure of proteins. solid curve, α- helix; long dashes, anti-parallel ß-sheet; dots, type 1 ß-turn; cross dashed line, extended 31- helix or poly (Pro) II helixdots and short dashes, irregular structure. Figure from Kelly et al. (2005). How to study proteins by circular dichroism. Biochimica et biophysica Acta 1751: 119-139

Here we used CD spectroscopy for thermal denaturation of the phosphopeptide- and nonphosphopeptide- MHC complexes to test their stability. Thermal denaturation was performed on a Jasco 815 spectropolarimeter at a certain wavelength (minimum of normal scan) at a temperature range between 20 and 90 degree. Far-UV spectra were collected and analyzed as described (Webb et al., 2004). Due to decay of structural elements of the protein a shift in the spectrum appears. Denaturation temperature is the half maximum of the shift.

25 2.2.10 Phosphatase treatment

Basis of this experiment is the determination of dephosphorylation time of phosphopeptides and phosphopeptides bound to MHC molecules by alkaline phosphatase. Hypothesis is that the dephosphorylation of phosphopeptides bound to MHC molecules takes longer in comparison to dephosphorylation of single phosphopeptides because of the protection of the phospho group through interactions with the MHC molecule. 5 µg of phosphopeptide or phosphopeptide HLA A2 complex was incubated with 5µl of alkaline phosphatase for 5-30 min and measured by MALDI-TOF mass spectrometry as described previously (Purcell et al., 2001). Because of a certain mass of the phospho- group the percentage of dephosphorylation could be obtained in a mass spectrum.

2.3 Plasmids encoding HLA A2 heavy chain and ß2m

Plasmids encoding the HLA A2 heavy chain and ß2m were made and provided by Dr. Andrew Webb, Purcell laboratory, Bio21.

2.4 Peptides

All peptides were synthesized and purified to >85% at the Bio21 peptide synthesis facility. Peptide stocks were prepared and dissolved in DMSO to a final concentration of 10-100 mg/ml.

26 2.5 Crystallization and X-ray crystallography

Crystallization and X-ray crystallography was undertaken in collaboration with the Jamie Rossjohn laboratory at Monash University and the protocol was written by Jan Petersen. Crystals of the peptide- HLA complexes were grown by the hanging drop vapour diffusion method at 20°C using similar precipitant solutions with 2-4 mM MgCL2, 2-4 mM CdCl2, 0.1 M HEPES (pH 7.4), 100-200 mM NaCl, and 12-13% PEG3350 (vol/vol). For the IRS2 nonphospho-complex, CoCl2 was used instead of MgCl2. Streak seeding was required to nucleate crystals of both CDC25b complexes. Crystals appeared after 12-48 h and grew to maximal size in 7-14 days. Crystals were flash-cooled to 100 K before data collection using 20% glycerol (vol/vol). X-ray diffraction experiments were performed using a Rikagu RU- 3HBR rotating anode generator with helium-purged OSMIC focussing mirrors coupled to an

R-AXIS IV++ detector (Rikagu). All crystals belong to space group P212121, with very similar unit cell dimensions. The structures were solved by molecular replacement using the program Phaser (Storoni et al., 2004). A modified protomer of a previously solved HLA A2 structure with the peptide residues removed was used as the search probe. Refinement was monitored by the Rfree value (5% of the data), using the same set of reflections for all data sets. Rigid body refinement and restrained refinement were performed using the program Refmac (Murshudov et al., 1997). This was followed by simulated annealing and individual B-factor and total least squares (TLS) refinement in Phenix (Adams et al., 2002). Model building was performed using the program Coot (Emsley et al., 2004). Water molecules and peptides were built into unambiguous electron density during the refinement process. Cd and Co ions were modelled into strong spherical peaks of the 2Fo-Fc maps, and their occupancies were adjusted manually to fit the maps. Figures were generated with Pymol (DeLano Scientific LLC) and APBS (Baker et al., 2001).

27 2.6 SDS Polyacrylamide gel electrophoresis (SDS-PAGE)

SDS-PAGE was performed according to Schaegger and von Jagow (Schaegger et al., 1987). Before performing the SDS-PAGE 20 µl of sample with 2µl DTT (Dithiothreitol) and 5µl loading buffer (300mM Tris/HCL pH 6.8, 4% (w/v) SDS, 12% (v/v) Glycerine, 2% (v/v) Mercaptoethanol, 0.01 % (w/v) Bromphenolblue pH 6.8) was denatured at 100°C for 5 min. For calculating the sizes of the proteins molecularweight standards were used.

28 Chapter III: Expression and purification of the phosphopeptide- and nonphosphopeptide- MHC complexes and stabilization experiments

The induction of a T cell response against the cancerous cell depends on the ability of a peptide that is specific to or over-expressed in the cell to bind to the MHC class I molecule as well as the affinity of the peptide-MHC complex for the TCR. As mentioned in chapter I, tumor associated antigens are usually poorly immunogenic but some post translational modifications encourage the binding of peptides to MHC molecules as for example deamidation of a glutamine residue in a gliadin peptide, which builds a stronger anchor residue that increases peptide binding affinity for HLA-DQ2 (Engelhard et al., 2006), (Qiao et al. (2005), (Qiao et al., 2004). Deimination of arginine encourages the binding of a vimentin peptide to HLA- DR*0401 (Hill et al., 2003). For phosphopeptides it is known that the phospho-group of some peptides promotes the binding of phosphopeptides to the MHC molecule (Zarling et al., 2006) (Mohammed et al., 2008). It is known that phosphorylation in cancerous cells is significantly different than in normal cells (Zarling et al., 2006). Therefore phosphopeptides derived from intracellular phosphoproteins may serve as good targets for cancer immunotherapy. Phosphopeptides derived from cellular proteins in cancerous cells can be recognized by cytotoxic T lymphocytes, but it is not known how the cytotoxic T lymphocyte recognizes the phosphopeptides displayed by the MHC class I molecule (Williamson et al., 2006). Phosphopeptides derived from ß-Catenin, CDC25b and IRS2 (Figure 5) were identified in three different cancer cell lines (2 melanoma cells, DM331 & SLM2 and an Ovarian carcinoma, COV413) and are therefore of particular interest as immunotherapeutic targets. In this study we wanted to supplement the crystallographic data (chapter IV) with data stating the binding of the phosphopeptides to the MHC molecule. Therefore a competition based peptide binding assay and thermal denaturation using CD spectroscopy have been performed on recombinant produced HLA A2-peptide complexes incorporating both phosphorylated and wild type peptide sequences from ß-catenin, CDC25b and IRS2. Moreover, we assessed the relative stability of the phosphate moiety by examining the sensitivity of the free and MHC- bound peptide to phosphatase treatment. As mentioned above, some phosphopeptides and their correlated native peptides were previously tested for stabilization differences including the 3 peptides tested in our study (Zarling et al., 2006 (Mohammed et al., 2006). For example

29 Zarling et al. found that the nonphosphorylated IRS2 peptide bound to HLA A2 with an affinity 4 times lower than the phosphorylated peptide (Zarling et al., 2006). The data from the competiotn based peptide binding assay and thermal denaturation using CD spectroscopy in this study are largely consistent with the data of the previous studies (Zarling et al., 2006) (Mohammed et al., 2008). As described in chapter II it was first necessary to express the heavy chain of HLA A2 and ß2m in E.coli as inclusion bodies as described previously (Garboczi et al., 1996). After inclusion body preparation the protein was refolded in the presence of the phosphopeptides and nonphosphopeptides. Following the complexes were purified using ion exchange and molecular sieve chromatography.

. Figure 5. Phosphopeptides derived from ß – catenin, IRS2 and CDC25b. Letters in blue show the amino acid sequence.

30 3.1. Expression of the heavy chain of HLA A2 and ß2m

3.1.1 Optimization of HLA A2 expression

Initially the parameters employed to optimize HLA A2 and ß2m expression were: → LB media → addition of chloramphenicol while culturing

→ induction at optical density (OD) OD600nm= 0.9 → 3 hour expression after induction with IPTG → use of one l LB media in 2 l flask → no additional incubation time with Urea buffer while inclusion body preparation

kDa

250

75 50 37 ← heavy chain of HLA A2

← ß2m 10

1 2 3 4 5 6 7 8 9

Figure 6: Gel of HLA A2 heavy chain expression with 1l LB media and

induction at OD600= 0.9 after 3 hour culture. lane 1-3 show 1µl, 2µl, 5µl of protein after inclusion body preparation with the HLA A2 heavy chain band at ~ 30 kDa, lane 4-6 show 1µl, 2µl, 5µl of protein after inclusion body preparation with the ß2m band at ~ 12 kDa, lane 7-9 show 1µl, 2µl, 5µl of 1mg/ml BSA used to quantify the recombinant material

31 Initial expressions achieved HLA A2 heavy chain concentrations of 0.5 mg/ml in 10 ml buffer as shown in Figure 6, lane 1-3. For ß2m the yield after inclusion body preparation was much higher at approximately 2 mg/ml in 10 ml buffer as shown in Figure 6, lane 4-6. Since especially for later experiments a high amount of HLA A2 heavy chain was necessary, the expression of HLA A2 heavy chain needed to be optimized.

3.1.1.1 Change of volume of cultured media and change of expression time

To test whether the volume of culture media and the expression time had an influence in achieving a higher yield of HLA A2 heavy chain, the volume of media was reduced to 500 ml in a 2l flask for better aeration of the cells and the expression time was extended to an overnight culture. The result obtained is shown in Figure 7. Compared to the initial expression of HLA A2 heavy chain in Figure 6 the gel shows a one times lower expression level after changing these two parameters (Figure 7). Therefore for later HLA A2 heavy chain expression the initial parameters for volume of media (1l in a 2 l flask) and expression time (3 hours expression) were used as with these parameters a better yield of protein could be achieved.

32 kDa

250

75 50 37 ← heavy chain of HLA A2

15 10 1 2 3 4 5 6

Figure 7. Gel of HLA A2 heavy chain expression with half of

LB media (500 ml) and induction at OD600= 0.9 after overnight culture. Lane 1-3 show 1µl, 2µl, 5µl after inclusion body preparation. Lane 4-6 show 1µl, 2µl, 5µl of 1mg/ml Bovine serum albumin (BSA)

3.1.1.2 No use of chloramphenicol during expression

To test if chloramphenicol affected the yield of protein, one experiment was undertaken with no use of chloramphenicol during expression of the HLA A2 heavy chain. For transfection of E.coli with the plasmid, chloramphenicol and kanamycin were necessary to select only cells with the plasmid for the HLA A2 heavy chain. For the large scale culture these antibiotics are usually not necessary anymore. The expression level of HLA A2 heavy chain, shown in Figure 8, was higher than in the sections before (Figures 6 and 7). Lane 1-3 of Figure 8 show the expression level of the HLA A2 heavy chain at approximately 30 kDa. There was a significant 2 fold overexpression of the protein visible. The final yield of HLA A2 heavy was approximately 20 mg. But as for the amounts of HLA A2 heavy chain necessary for the refolding of the MHC complex in presence of the peptides it was still not high enough.

33 kDa 250

50 37 ← heavy chain of HLA A2 25

10 1 2 3 4 5 6

Figure 8: Gel of HLA A2 heavy chain expression without

chloramphenicol and induction at OD600= 0.9 after 3 hour culture. Lane 1-3 show 1µl, 2µl, 5µl of protein after inclusion body preparation, lane 4-6 show 1µl, 2µl, 5µl of 1mg/ml BSA

3.1.1.3 Use of terrific broth media instead of LB media

Finally a different media was used for potential improvement of the yield of HLA A2 heavy chain. Instead of LB media, terrific broth media was used for the culture. Terrific broth media is a richer media than LB media and allows a higher cell density. As shown in Figure 9 a 5 fold improvement in the yield of HLA A2 heavy chain (lanes 4-6) was observed with use of terrific broth media compared to the expression of HLA A2 heavy chain in LB media (lanes 1-3). The final yield of HLA A2 heavy chain was appr. 50 mg.

34 kDa 250

50 37 ← heavy chain of HLA A2

1 2 3 4 5 6 7 8 9

Figure 9. Gel of HLA A2 heavy chain expression after inclusion body

preparation. a) LB media at OD600= 0.9 after 3 hours expression with 1µl,

2µl and 5µl of protein (lane 1-3), b) Terrific broth media at OD600= 0.9 after 3 hours expression with 1µl, 2µl and 5µl of protein (lane 4-6) and c) 1 µl (lane 7), 2µl (lane 8), 5µl (lane 9) of 1 mg/ml BSA

3.2. Refolding of the heavy chain of HLA A2 and ß2m in presence of the phosphopeptides and nonphosphopeptides

Having optimised the expression of the heavy chain of HLA A2 the next task was to refold the expressed proteins in the presence of the phospho- and nonphospho-peptides. This section details the steps performed to optimize the refold of the proteins.

3.2.1 Optimization of HLA A2-peptide complex preparation

In the initial experiment the ratio of HLA A2 heavy chain and ß2m used was 1:1 (v/v).

35 After refolding the proteins for 24 hours initial ion exchange chromatography using Q sepharose and molecular sieve chromatography were performed.

kDa 250

50 37

15 10

3 4 5 6 7 lane

Figure 10. Fractions after initial ion exchange chromatography using Q sepharose- and molecular sieve chromatography. a) refolded protein (lane 3-4) and b) unbound ß2m (lane 5-7)

Figure 10 shows the refolded HLA A2 heavy chain and ß2m (appr 12 kDa) in lane 3 and 4 and a large amount of uncomplexed ß2m was apparent in the late eluting ion exchange fractions, reflecting poor levels of complex formation (lanes 5-7). Therefore instead of 1 aliquot of HLA A2 heavy chain in later experiments the HLA A2 heavy chain was added for a second and third time after 24 and 48 hours to drive further complex formation. After refolding of the proteins with a ratio 1:1 (v/v) for HLA A2 heavy chain and ß2m, purification and final concentration step using centrifugal concentration the refolded complex had a protein concentration of about 1 mg/ml with total yield of 0.5 mg of refolded protein (Figure 11 a). With changing the ratio of HLA A2 heavy chain and ß2m to 3:1 (v/v) the final concentration of refolded protein increased to 6-12 mg/ml (Figure 11 b) with a total yield of appr. 5 mg of refolded complex (Figure 11 b).

36 a) b)

1 2 3 1 2 3 Refold Refold

Figure 11. Differences in the amount of final refolded protein after all purification steps and concentration. a) before optimization of the refolding step with lane 1 1µl protein, lane 2 2µl protein, lane 3 5µl protein and b) after two additional injections of HLA A2 heavy chain while refolding with lane 1 1µl of protein, lane 2 2µl of protein and lane 3 5µl of protein

3.3. Purification of the MHC- phosphopeptide- and MHC- nonphosphopeptide- complexes

After optimization of the expressions of the heavy chain of HLA A2 and the refolding step the refolded complexes were purified to achieve pure protein for performing X-ray crystallography. The protocol for the purification of the proteins after refolding step and dialysis was the following:

• Ion exchange chromatography using Q sepharose column • Collection and concentration of the appropriate fractions • Molecular sieve chromatography using superdex 75 column • Collection of the appropriate fractions

37 • Ion exchange chromatography using Mono Q column • Final concentration using centrifugal filtration

3.3.1 Ion exchange chromatography using Q sepharose

The refolded complexes were loaded onto a Q sepharose column and the protein was eluted with a NaCl gradient as described in chapter II.

2 1 3

b) kDa 250

10 1 2 3 4 5 6 7 8 9

Figure 12. Ion exchange chromatography after optimization of the refolding step. a) chromatogram showing UV absorbance at 280 nm (blue line) and b) SDS Page gel with fractions after purification, lane 1-6 are fractions of peak 1, lane 7-8 are fractions of peak 2, lane 9 is fraction of peak 3

38 The chromatogram shown in Figure 12 had three significant peaks. Peak one contains relatively pure refolded MHC-peptide complexes. The MHC-peptide complexes are also in peak two but with much high molecular weight contaminations, suggesting this fraction also contained aggregated complexes. Therefore fractions of peak one were pooled and concentrated to perform the next purification step.

3.3.2 Molecular sieve chromatography

The concentrated fractions from the Q Sepharose chromatography were loaded onto the molecular sieve column and run as described in chapter II.

a) 1

b) kDa 250

15 10

1 2 3 4 5 6 7 8 9 lane

Figure 13. Molecular sieve chromatography after optimization of the refolding step and initial ion exchange chromatography. a) chromatogram and b) SDS Page gel with fractions after purification, lanes 2-7 of peak 1 and lanes 8-9 of peak 2

39 A representative chromatogram of this chromatographic separation is shown in Figure 13. The main early eluting peak (peak one) contains the refolded MHC-peptide complexes. This material was pooled and concentrated for the final ion exchange chromatography step.

3.3.3 Ion exchange chromatography using Mono Q

Pooled molecular sieve chromatography fractions containing HLA A2-peptide complexes were finally loaded onto a Mono Q column and the chromatography was performed as described in chapter II. As shown in Figure 14, two significant peaks were obtained in the chromatogram. Peak one at the beginning of the chromatography represented the flow through while peak 2, which corresponds to fractions 8 and 9 contained the refolded complex (Figure 14 a, b). As can be seen in Figure 14 b, fraction 9 did not appear to contain any high molecular weight contaminants but did have one prominent low molecular weight protein contaminant. This contaminant is commonly seen for HLA refolds and does not impair crystallization. Nevertheless because of high concentration of refolded MHC-peptide complexes, fractions 8 and 9 were taken for final concentration and in the end the purity of the MHC-peptide complexes was approximately 95% and therefore high enough to perform crystallization trials. The protein concentration after final concentration was approximately 8 mg/ml and a total yield of 4 mg could be achieved.

40 a ) 2 1

c) b) kDa kDa

250 250

75 75

50 50

25 25

15 15

10 10 1 2 3 4 5 6 1 8 9 Fractions

Figure 14. Final ion exchange chromatography after optimization of the refolding step, initial ion exchange chromatography and molecular sieve chromatography. a) chromatogram and b) SDS Page gel with fractions after purification, fraction 1 is the flow through of peak 1 and fractions 8 and 9 are congruent with peak 2 c) fractions 8 and 9 were finally concentrated and used for crystallization experiments with 1µl (lane 1), 2µl (lane 2) 5µl (lane 3) of refolded protein and lane 4-6 with 1µl, 2µl, 5µl of 1mg/ml BSA

41 3.4 Characterisation of the HLA A2 peptide complexes

At the time of this study little was known of the role of the phosphate head group of the phosphorylated peptide epitope in HLA binding and stabilisation of the HLA-peptide complex. We undertook two different experiments to investigate this - thermal denaturation of phosphorylated and nonphosphorylated HLA A2-peptide complexes using CD spectroscopy and a competition based peptide binding assay to examine the relative affinity of the modified and wild type peptides for HLA A2. In addition, the sensitivity of the phosphopeptides in complex with HLA A2 to phosphatase treatment was examined relative to the phosphopeptide free in solution.

Here thermal denaturation was undertaken to compare the stability of phosphopeptide-MHC- and nonphosphopeptide-MHC complexes. During denaturation the exposure of structural elements decays due to unfolding which is obtained as a shift in the spectrum. Half maximum of the shift is the denaturation temperature.

3.4.1 Competition based peptide binding assay

The first experiment undertaken to compare the affinity of the phosphopeptides and their nonphosphorylated counterparts to the MHC molecule was a competition based peptide binding assay. High affinity of the competitor peptides is congruent with a highly stable peptide-MHC complex. The peptides of interest (competitor peptides) competed with a fluorescence labelled reference peptide as described in chapter II. Before doing the assay, JY cells were treated with mild acid for 90s to remove HLA-bound peptides so that the competitor peptides (phosphopeptides and nonphosphopeptides derived from IRS2, ß-catenin and CDC25b) and reference peptides could bind to HLA A2 while performing the assay. After 24 hours cells were washed and fluorescence was measured by flow cytometry. For the peptides derived from CDC25b and ß-Catenin no differences in the HLA A2 affinity was observed between the wildtype- and phosphopeptides.

42 b a 100 100

80 80

60 60

40 40 percentage inhibition

percentage inhibition percentage 20 20

0 0 0 204060 0 204060

concentration competitor peptde (µg/ml) concentration competitor peptde (µg/ml) c 100

60 ------phosphopeptide 40 ------nonphosphopeptide

percentage inhibition 20

0 0204060 concentration competitor peptide (µg/ml)

Figure 15. Stabilzation experiment using competiton based peptide binding assay. a) Peptide GLLGpSPVRA derived from CDC25b, b) peptide RVApSPTSGV derived from IRS2, c) peptide YLDpSGIHSGA derived from beta Catenin

As shown in Figure 15, the binding curves for the phosphopeptides and nonphosphopeptides derived from CDC25b and ß-Catenin are similiar. For both peptides there is a high degree (approximately 80%) inhibition of reference peptide which indicates these peptides bind to HLA A2 with a relatively high affinity. In contrast the peptides derived from IRS2 displayed distinct binding to HLA A2. The affinity of the phosphorylated IRS2 peptide was higher than the affinity of the nonphosphorylated peptide which suggests that the phosphopeptide-MHC complex is more stable. The curve for the nonphosphopeptide is markedly reduced.

43 3.4.2 Thermal denaturation of HLA A2-peptide complexes using CD spectroscopy

A second experiment used to measure the stability of the phosphopeptide- MHC complexes compared to the nonphosphopeptide- MHC complexes was thermal denaturation using CD spectroscopy. The CD spectra were measured at a temperature range between 20 and 90°C. As shown in Figure 16, normalised spectral minima of the native complex is plotted against temperature. The denaturation of the peptide-MHC complexes appeared as a shift in the spectrum. As higher the denaturation temperature (half maximum of the shift) is as higher is the peptide-MHC complex stability. Overall the results correlated with the results of the competition based peptide binding assay described in 3.4.1. No differences of the denaturation/melting temperature (half maximum of the shift) were observed between the wildtype-MHC and phosphopeptide-MHC complexes with peptides derived from CDC25b and ß-catenin. The curves for MHC-phosphopeptide and MHC-nonphosphopeptide complexes were essentially identical. These curves show no rise to up to 40° Celsius. With 40° Celsius the thermal denaturation starts. A shift is obtained between 40° Celsius and 60° Celsius which is congruent with the exposure of structural elements. At 60° Celsius 100% denaturation is reached. For both the phosphopeptide- and nonphosphopeptide complexes derived from CDC25b the melting temperature is approximately 58°C and for the complexes with peptides derived from ß-Catenin it is approximately 56°C (Table 2). For the peptides derived from IRS2 the curves show no denaturation up to 40° Celsius. Between 40° Celsius and 60° Celsius a kink with up to 40% denaturation has been obtained. With 60° Celsius the shift starts and ends with 100% denaturation at 65° Celsius for the nonphosphopeptide-MHC complex and at 70° Celsius for the phosphopeptide-MHC complex. The difference in the shift is congruent with an increase of the melting temperature (half maximum of the shift) of approximately 6°C for the phosphopeptide-MHC complex with a melting temperature of 68°C compared to the nonphosphopeptide-MHC complex where the melting temperature is 62°C. Therefore the IRS2 phosphopeptide-MHC complex is more stable (Fig 16) (Table 2).

44 a b

10 0 100

50 50

percentage denaturation 0

percentage denaturation 20 40 60 80 20 40 60 80

temperature (degree) temperature (degree) c

100 ------phosphopeptide-MHC complex ------nonphosphopeptide-MHC complex

50 percentage denaturation

0 20 40 60 80 temperature (degree)

Figure 16. Stabilzation experiment using thermal denaturation with CD spectroscopy. a) Peptide GLLGpSPVRA derived from CDC25b and nonphosphopeptide in complex with MHC, b) peptide RVApSPTSGV derived from IRS2 and nonphosphopeptide in complex with MHC, c) peptide YLDpSGIHSGA derived from beta Catenin and nonphosphopeptide in complex with MHC

Peptide complexes Denaturation Temperature CDC25b phosphopeptide-MHC complex 58°C nonphosphopeptide-MHC complex 58°C IRS2 phosphopeptide-MHC complex 68°C nonphosphopeptide-MHC complex 62°C ß-Catenin phosphopeptide-MHC complex 56°C nonphosphopeptide-MHC complex 56°C

Table 2: Denaturation temperature of the phosphopeptide- and nonphosphopeptide-HLA A2 complexes derived from CDC25b, IRS2 and ß-Catenin

45 3.4.3 Stability of HLA A2-peptide complexes to phosphatase treatment

To test whether the phospho-group of the peptide is protected from phosphatase activity by the MHC molecule dephosphorylation of single phosphopeptides was compared with dephosphorylation of phosphopeptide-MHC complexes.

a b 10 0 100

50 50

0 0 Percentage dephosphorylation dephosphorylation Percentage Percentage dephosphorylation dephosphorylation Percentage 02040 02040 c time (min) time (min)

100

50 ------phosphopeptide ------phosphopeptide-MHC complex

Percentage dephosphorylation dephosphorylation Percentage 02040

time (min)

Figure 17. Stabilzation experiment using phosphatase treatment. a) Peptide GLLGpSPVRA derived from CDC25b and peptide-MHC complex, b) peptide RVApSPTSGV derived from IRS2 and peptide-MHC complex, c) peptide YLDpSGIHSGA derived from beta Catenin and peptide-MHC complex

Figure 17 shows the percentage of dephosphorylation over a period of 30 minutes with 5 minute time points. For the free phosphopeptide derived from CDC25b the percentage of dephosphorylation rises up to 50% whereas the phosphopeptide- HLA A2 complex shows no dephosphorylation after 30 minutes incubation with the alkaline phosphatase. For the other 2

46 phosphopeptides derived from IRS2 and ß-Catenin the percentage of dephosphorylation goes up to 80% whereas the complexes show again no dephosphorylation. In the figures it is shown that the free phosphopeptide derived from CDC25b is more stable in contrast to the other phosphopeptides. This was not further investigated since the aim of this experiment was to only show the dephosphorylation of the free phosphopeptides compared to no dephosphorylation for the phosphopeptide- MHC complexes. In all cases protection of the phosphate-group was observed for the phosphopeptides bound to the MHC complex (Fig 17). This correlates with the observations that the phospho-group undergoes several interactions with the MHC molecule (Chapter 4).

47 Chapter IV: Crystallization and X-ray crystallography of the MHC- phosphopeptide- and MHC-nonphosphopeptide complexes

In this chapter the structural clarification of the MHC- phosphopeptide complexes with phosphopeptides derived from insulin receptor substrate (IRS) 2 (RVApSPTSGV), ß-Catenin (YLDpSGIHSGA) and cell division cycle (CDC) 25b (GLLGpSPVRA) is described. To see whether there are potential structural differences, MHC-phosphopeptide complexes were compared with the structures of their nonphosphorylated counterparts. The phosphopeptides used for this study were chosen because they are found on different tumor cell lines (Zarling et al., 2006). Moreover these peptides have been chosen to distinguish this study from an earlier study (Mohammed et al., 2008) where phosphopeptides with mainly arginine at position 1 and the phospho group at position 4 of the peptide were studied. The primary goal was to ascertain the phospho-group of the peptides influenced binding to the MHC molecule and therefore makes the antigen more immunogenic. Two locations of the phospho – group are possible: either the phospho – group is involved in the binding to the MHC molecule or the phospho – group is exposed on the surface of the complex. In order to understand the influence of the phopho - group on MHC-peptide binding, the structure of three HLA A2– phosphopeptide complexes were compared to the corresponding MHC – nonphosphopeptide complexes.

4.1 Crystallization

Initial crystallization trials by a robot screen were performed using concentrations of approximately 1-3 mg/ml of protein. Only micro crystals appeared and according to less than 50% of precipitation in the wells the protein concentration was not considered high enough. In later crystallization experiments, protein concentrations of 8-13 mg/ml were used and very nice crystals in shape and size were found at condition 5mM CoCl2, 5mM CdCl, 5mM MgCl2,

5mM NiCl2, 0.1 M HEPES pH 7.5, PEG 3350 12% (Figure 18). This condition was nearly the same for all MHC-phosphopeptide- and MHC-nonphosphopeptide complexes. Only for one peptide-MHC (peptide derived from CDC25b) complex further optimization, using streak seeding, was necessary.

MHC - Beta catenin nonphsopho - MHC - Beta catenin phospho - complex MHC – CDC25b phospho - complex complex

MHC - IRS2 phospho - MHC - IRS2 nonphospho - MHC – CDC25b nonphospho - complex complex complex

Figure 18. Photos of the phosphopeptide- and nonphosphopeptide- HLA A2 complexes. From Jan Petersen, Monash university, Melbourne

4.2 X-ray crystallography

The crystal structures of the phosphopeptide- and nonphosphopeptide- HLA A2 complexes were determined to a resolution of 1.93 Å or better (Table 4). Moreover all peptide- MHC complexes had the same space group and unit cell dimensions as summarised in Table 4. For all MHC- phosphopeptide complexes the phosphate group was solvent exposed (Figure 19). This is consistent with the results from Mohammed et al. (Mohammed et al., 2008). Our study differed from this study however in that we were able to directly compare the structures of HLA A2 complexed with both the phosphopeptide and the corresponding unmodified peptide.

49 Phosphopeptide (derived from ß Phosphopeptide (derived CDC 25 b) Phosphopeptide (derived from IRS 2) catenin) – MHC complex, – MHC complex, Phosphopeptide in – MHC complex, Phosphopeptide in Phosphopeptide in yellow and MHC blue and MHC complex in magenta red and MHC complex in green complex in blue

Figure 19. Total view of the phosphopeptide- MHC complexes

4.2.1 Structure of the MHC-phosphopeptide- and MHC- nonphosphopeptide complexes derived from IRS2

Compared with its nonphosphorylated counterpart the phosphorylated peptide derived from IRS2 shows no significant differences when binding into the HLA A2 binding cleft. Only the phosphate brings a change in the electrostatic MHC-peptide surface with an addition of an electronegative charge and small changes in the position of Lys66 and Arg65 (Fig 20). Overall, the structure of the HLA A2-IRS2 complex indicated that the main interactions between the peptide and the antigen binding cleft involved the canonical anchor residues P2- Val and P9-Val. In addition P3-Ala and P6-Thr also point into the binding cleft. P4-Ser in the wildtype peptide complex is not involved in significant interactions with HLA A2 residues. In contrast, the phosphate group on P4-phosphoSer in the phosphopeptide complex was found to adopt two conformations indicating flexibility of this group and it forms stabilizing contacts with the antigen binding cleft. One conformer interacts with Lys66 on the α1-helix and with Gln155 on the α2-helix through a salt bridge and a water-mediated H-bond. The other conformer interacts with Lys66 and Arg65 of the HLA A2 heavy chain through a salt bridge (Fig 21) (Table 3). These interactions of the phosphate moiety with the HLA A2 heavy chain would explain the protection of the phosphate from dephosphorylation through alkaline phosphatase as mentioned in chapter III. A reason for the higher stability of the phosphopeptide-HLA A2 complex and higher affinity of the phosphopeptide observed by thermal denaturation using CD spectroscopy and the competition based peptide binding assay may be due to prominent salt bridges, which are very

50 strong interactions, between the phosphoserine and the HLA A2 binding cleft whereas no interactions have been observed for P4-nonphosphoSer.

Fig 20. Altered surface potential for TCR recognition. Surface representation of HLA A2 with bound peptides. (a) CDC25b, (b) IRS2, (c) CDC25b-phospho, (d) IRS2-phospho. Gray indicates α-chain with potential TCR contact residues in purple. The arrows indicate the peptide phosphorylation site. Electrostatic potentials are in blue=positive and red=negative. From Jan Petersen, Monash University, Melbourne

51 HLA A2 Intra-peptide Nature of interaction CDC25b Ser5: Oγ, O Arg8(O) Water mediated H-bond Oγ, Cα, Cβ Ala69(Cβ) Van der Waals CDC25b-phospho pSer5: Conformer A (50 % occupancy) Oγ Pro6(O) Water mediated H-bond pSer5: Conformer B (50 % occupancy) Oγ Pro6(O) Water mediated H-bond O2P Arg65(Nη2) Salt bridge IRS2 non-phospho Ser4: N/A IRS2 phospho pSer4: Conformer A (60 % occupancy) O1P Gln155(Oε1) pSer4(N) Water mediated H-bond O2P pSer4(N) Water mediated H-bond Lys66(Nζ) Salt bridge O3P Lys66(Nζ) Salt bridge pSer4: Conformer B (40 % occupancy) O2P Arg65(Nη2), Lys66(Nζ) pSer4 (N) Salt bridge O3P Arg65(Nη1) Salt bridge Oγ Lys66(Nζ) Salt bridge N pSer4 (O1P, O2P) Water mediated H-bond β-Catenin non-phospho Ser4: Conformer A (40 % occupancy) Oγ Leu2(O), Ser4(N), Water mediated H-bond N Ser4(Oγ), Leu2(O) Water mediated H-bond Asp3(Oδ1) H-bond Cα Lys66(Cδ) Van der Waals Ser4: Conformer B (60 % occupancy) N/A β-Catenin phospho pSer4: Conformer A (60 % occupancy) O2P Lys66(Nζ) Water mediated H-bond Tyr1(Oη), Leu2(O) Water mediated H-bond O3P Cd Water mediated H-bond N pSer4 (O2P), Leu2(O) Water mediated H-bond Asp3(Oδ1) H-bond pSer4: Conformer B (40 % occupancy) O2P Lys66(Nζ), Water mediated H-bond Arg65(Nη2) Salt bridge Oγ Lys66(Nζ) Water mediated H-bond Lys66(Cδ) Van der Waals N pSer4 (O2P), Leu2(O) Water mediated H-bond Asp3(Oδ1) H-bond Cα,Cβ Lys66(Cδ) Van der Waals His7(Nε2) H-bond

Table 3: Serine and phosphoserine contacts. From Jan Petersen, Monash university, Melbourne

Figure 21. Phosphorylation site in nonphosphopeptide- and phosphopeptide- HLA A2 complexes. Stick representation of peptides and of heavy-chain side chains that interact with the phsophorylation site. Yellow indicates nonphosphopeptide-HLA A2 complexes. Green indicates phosphopeptide-HLA A2 complexes. (a-b) peptide derived from CDC25b. (c-d) peptide derived from IRS2. (e-f) peptide derived from ß-Catenin. From Jan Petersen, Monash University

53 IRS2 IRS2 ß- ß- CDC25b CDC25b nonphospho Catenin Catenin nonphospho phospho nonphospho phospho phospho Resolution (Å) 23.7- 24.0- 23.7- 24.3- 24.9- 30-1.80 1.93 1.70 1.65 1.70 1.80

Space Group P212121 P212121 P212121 P212121 P212121 P212121 Cell dimensions (Å) 59.98 59.85 59.72 59.68 59.68 59.91 (a,b,c) 79.38 79.68 79.71 79.71 79.84 80.04 111.29 111.00 110.87 111.56 110.21 110.37 Total No. 38009 56258 63777 56745 49135 49810 observations Multiplicity 4.7 5.1 6.3 7.1 3.5 4.9 Data completeness 93.07 95.25 98.7 95.8 99.19 99.8 (%) (75.9) (73.9) (88.8) (76.3) (97.8) (100)

I/σI 22.1 26.1 27.3 26.9 18.6 23.9 (3.7) (3.9) (2.5) (4.3) (2.8) (3.3) 1 Rmerge (%) 5.6 4.4 5.0 5.1 5.0 5.2 (32.8) (31.8) (42.9) (30.4) (45.0) (46.6) 2 Rfactor (%) 16.85 17.10 17.92 16.85 17.53 17.04 3 Rfree (%) 19.63 19.61 19.95 18.86 20.34 20.18 Rms deviations from ideality Bond lengths (Å) 0.006 0.008 0.006 0.005 0.005 0.007 Bond angles (°) 1.022 1.115 1.058 0.998 0.967 1.142 Ramachandran angles (%) favoured 97.61 98.13 97.84 97.61 98.14 98.40 allowed 2.39 1.87 2.16 2.39 1.86 1.60 outliers ------

B-factors Peptide 28.7 29.0 28.0 26.4 28.0 38.3 Protein 27.3 28.5 27.3 24.4 29.4 27.6 water, ions (Cd), 36.5 41.3 40.3 38.0 40.1 38.7 glycerol

1 Rmerge = Σ |Ihkl - | / ΣIhkl 2 3 Rfactor = Σhkl | |Fo| - |Fc| | / Σhkl |Fo| for all data except for 5% which was used for the Rfree calculation. Numbers in parentheses refer to statistics in the highest resolution bin.

Table 4: Data collection and refinement statistics. From Jan Petersen, Monash University, Melbourne

54 4.2.2 Structure of the MHC-phosphopeptide- and MHC- nonphosphopeptide complexes derived from ß-catenin

The peptide derived from ß-catenin has 2 primary anchor residues at P2-Leu and P10-Ala. The nonphosphorylated peptide shows high flexibility in complex with the MHC molecule in the central region which is not based on poor crystallographic data. Here the main interactions of the P4-Ser are of intrapeptide nature through H-bonds. Only one interaction with Lys66 of the HLA A2 heavy chain, a van der Waals interaction, has been observed (Table 3). The mobility of the phosphorylated peptide-HLA A2 structure is reasonably reduced and the whole epitope is visible. This stabilization of the peptide by the phospho-group is due to intrapeptide interactions as well as interactions with Arg65 and Lys66 of the MHC molecule (Table 3). Intrapeptide interactions are mainly water-mediated H-bonds whereas the interaction of phosphoserine with Arg65 of the binding cleft is a salt bridge which makes the phosphopeptide-MHC complex more stable and protects the phospho- group from dephosphorylation (Table 3). Also here the phospho-group appears in two different conformations. In one conformer the phosphate group builds interactions with P7-His and Arg65 with P7-His “pulled in” toward the phospho-group compared to the nonphosphorylated complex. Furthermore the phosphate group interacts with P1-Tyr through a water-mediated H-bond (Fig 21).

4.2.3 Structure of the MHC-phosphopeptide- and MHC- nonphosphopeptide complexes derived from CDC25b

For the peptide derived from CDC25b, P2-Leu and P9-Ala are the main anchor residues and also P3-Leu contributes to peptide binding within the peptide binding cleft. P5-Ser and P8-Arg interact through a water-mediated H-bond and both seem to be potential TCR contact sites with pointing upwards from the MHC complex. For the phosphopeptide-MHC complex there are no conformational changes in the MHC heavy chain compared with the nonphosphopeptide-MHC structure. But, the phosphopeptide itself shows conformational changes around the phosphorylation site to avoid steric clashes

55 with Ala69 and Thr73 of the HLA A2 heavy chain. Therefore the peptide is pushing away from the α1-helix toward the center of the antigen-binding cleft, resulting in a 2 Å shift in the Cα position between P5-nonphospho- and P5-phosphoSer (Fig 20). Consistent with the 2 other phosphopeptide-MHC complexes the phospho-group is observed in 2 different conformations which concludes a general theme of a mobile phosphate group. One conformer interacts with Arg65 of the HLA A2 heavy chain through a salt bridge and both conformers build a water-mediated H-bond with P6-Pro (Fig 21). Moreover in the nonphosphorylated complex the P4-Ser builds a van der Waals interaction with Ala69 from the HLA A2 heavy chain and a water-mediated H-bond with P8-Arg (Table 3). In this example the peptide-MHC complexes define a case of altered self where the peptide shows significantly altered conformation. Here a post- translational modification alters a self- peptide-MHC complex.

56 Chapter V: Summary and Discussion

5.1 Summary of the results

This work provides an insight into the impact of phosphorylation of peptides epitopes on binding and stabilisation of MHC molecules. Moreover, our detailed structural analysis highlights the prominence of the phosphate moiety for T cell recognition and roles for the phosphate in aiding peptide binding and modulating peptide conformation. First it was necessary to generate peptide-MHC complexes of high purity and concentration appropriate for crystallization. Using an optimised expression protocol for the HLA A2 heavy chain that involved initial expression of the HLA A2 heavy chain in a rich media and 2 additional injection steps of HLA A2 heavy chain inclusion body preparations during the oxidative refolding of the complex improved the yield sufficiently to allow the production of highly pure HLA A2-peptide complexes at around 10 mg/ml. The crystallization was straight forward since very nice crystals in shape and size were obtained in the robot screen. Optimization of the crystallization conditions was required for only one MHC- phosphopeptide- and MHC- nonphosphopeptide complex with the peptide derived from CDC25b using streak seeding. The structures of the 6 peptide-MHC complexes, with phosphopeptides derived from IRS2 (RVApSPTSGV), ß-Catenin (YLDpSGIHSGA), CDC25b (GLLGpSPVRA) and their nonphosphorylated counterparts, were solved at 1.9 Å or better. The structures of the phosphopeptide-MHC complexes demonstrated distinctive epitope dependent structural features. Phosphorylation of P5-Ser of the peptide derived from CDC25b leads to an altered peptide backbone conformation through steric conditions of incorporating the phosphate head group. The phospho-group of the peptide derived from IRS2 formed stabilizing contacts with HLA A2 through Lys66 and Arg65, this contributed to a higher complex stability and peptide binding affinity for the phospho-IRS ligand yet the two peptides had very similar bound conformations. Finally the phosphorylation of the peptide derived from ß-Catenin stabilizes the mobile nature of the central region of the peptide from R3 to R6. Stabilization experiments were undertaken to show potential stabilization differences between the nonposphopeptide-MHC- and phosphopeptide-MHC complexes. For the peptide derived from IRS2, phosphorylation enhanced MHC binding and thermostability. But this was not

57 generally observed. Finally dephosphorylation experiments showed that the MHC complex protected the phosphopeptides from dephosphorylation.

5.2 Discussion

The main aim of this study was to examine the presentation of phosphopeptides by the common human class I MHC molecule HLA A2. In each complex examined here, the phospho- group was solvent exposed and involved in intrapeptide interactions as well as interactions with residues of the HLA A2 heavy chain. These findings are consistent with the observations of an earlier study where solvent exposure of the phospho-group has been observed (Mohammed et al., 2008). However in our study we were able to directly observe the role of the phosphate group in altering the presentation of the peptide ligand by direct comparison of the structures of the native- and corresponding phosphopeptide- HLA A2 complexes. Especially for the phosphopeptide derived from CDC25b the presence of the phosphorylated P5 serine of the peptide lead to a conformational change of the peptide in comparison to its nonphosphorylated counterpart due to steric constraints. Because of this fact, these phosphopeptide- HLA A2 complexes show a case of altered self in which the peptide demonstrates significantly altered conformation which may have an impact on T cell recognition. In contrast to Mohammed et al. (Mohammed et al, 2008) this is the first published example of a phosphopeptide-MHC complex with the peptide phosphorylated at position 5. Moreover, Mohammed et al. describe a common binding motif for a positively charged N- terminal residue with the phosphorylated serine residue at position 4 of the peptide. In this study one example of a peptide with these characteristics, the phosphopeptide derived from IRS 2, was tested. However, this common binding motif is only partly true for the phosphorylated IRS2 peptide because the salt bridge to the N-terminal arginine is absent in the phosphopeptide-HLA A2 complex and therefore no movement of arginine at position 4 of the peptide toward position 1 of the peptide into the position observed by Mohammed et al. (Mohammed et al., 2008) was visible. Furthermore for all three phosphopeptides bound to the MHC molecule the phospho- group shows 2 distinct conformations reflecting the flexibility of the phosphate headgroup. These

58 findings as well as the alterations of the electrostatic footprint of the phosphopeptide- MHC complexes indicate a strong influence of phosphorylation of peptide epitopes on T-cell recognition. Especially the flexibility of the phospho- group leads to the assumption that the phospho- group of the peptide is significantly involved in a strong binding to the T-cell receptor. This is consistent with the findings of Zarling et al. (Zarling et al., 2006) where phosphopeptide-specific CD8+ T lymphocytes were generated which do not recognize the native form of the peptide. Therefore phosphopeptides seem to be excellent candidates for tumor vaccination. Moreover, changes in peptide conformation and mobility between phosphorylated and nonphosphorylated version were observed in an epitope-dependent manner. This is an important progression from Mohammed et al., 2008 where only the structures of the MHC-phosphopeptide complexes have been solved. Another important observation from this study was the protection of the phosphopeptide antigens from dephosphorylation when bound within the MHC complex. This is consistent with the fact that phosphopeptide antigens were found on the surface of MHC class I molecules as well as their ability to generate phosphopeptide-specific CTLs in vivo (Zarling et al., 2006). As mentioned earlier for some post translational modifications it is know that they encourage the binding of peptides to MHC molecules as for example deamidation of a glutamine residue in a gliadin peptide creates a better anchor residue that increases peptide binding affinity for HLA-DQ2 (Engelhardt et al., 2006) (Qiao et al., 2005) and the deimination of arginine encourages the binding of a vimentin peptide to HLA-DR*0401 (Hill et al., 2003). In this study two experiments were chosen to test whether phosphorylation may also enhance the binding of the peptides to the MHC molecule. Thermal denaturation of the HLA A2 peptide complexes was followed using CD spectroscopy to assess complex thermostability and a competition based peptide binding assay was used to assess peptide ligand affinity. These assays demonstrated that for the phosphopeptide derived from IRS2, with an arginine at position 1, an enhanced phosphopeptide-MHC complex stability compared with the nonphosphopeptide-MHC complex. This is also consistent with the findings observed by Mohammed et al. (Mohammed et al., 2008) where they describe the concept of the phosphate acting as a surface anchor providing enhanced MHC binding of the phosphopeptides. For the other 2 phosphopeptides, which have a glycine and tyrosine at position 1 of the peptide, no enhanced complex stability was observed.

59 Our data suggest that preformed phosphopeptide-MHC complexes by using dendritic cells pulsed with the phosphopeptide would be more efficient than using only the peptide as vaccine. This is due to the fact that binding of the phosphopeptides within the HLA A2 heavy chain protects the peptides from dephosphorylation whereas the free phosphopeptides are dephosphorylated through the phosphatase. Moreover the use of peptide epitopes is more efficient than a whole protein vaccine because the whole protein also accommodates a high level of irrelevant or potentially deleterious components (Lee et al., 1999). One study, where a peptide epitope elicited a T cell response but no immune response was detected after immunization with the whole protein, proves this statement (Zanetti et al., 1987). In context to our study the use of whole protein vaccines is not possible because it is impossible to make sure that the protein gets phosphorylated at the appropriate position since kinases phosphorylate at many different positions within the protein. For peptide epitopes the peptide with its phosphorylation site can be synthetically designed. Finally, the overrepresentation of the phosphate at position 4 and 5 of the peptide may be due to stabilizing interactions of the phospho-group with the HLA A2 heavy chain as well as a prominent position for T cell receptor contacts. Also processing issues, e.g. degradation of the antigen through the proteasome may play a role. Overall these findings show insights into the mode of phosphopeptide presentation by HLA A2 and suggest that phosphopeptide antigens may serve as good targets for the design of tumor vaccines.

5.2.1 Future directions

To further investigate the importance of phosphopeptides in the immunopeptidome as well as their possible function as tumor vaccines it will be necessary to solve the structures of MHC class I- phosphopeptide- cytotoxic T cell receptor complexes as well as the structures of MHC class II- phosphopeptide- CD4+ T cell receptor complexes. Also the measurement of their biophysical properties is highly desirable and will give insights in how far phosphopeptide antigens, similar to MART-1 as mentioned in the introduction chapter I, will serve as good targets for immunotherapy of cancer.

60 Another point of interest will be the discovery of novel MHC class I restricted phosphopeptides phosphorylated at different positions within the peptide as well as phosphopeptides with a phosphorylated tyrosine residue which has not been found to date because of its rarity in the cell (mentioned in the introduction chapter). This would extend the spectrum of more potent phosphopeptide vaccine candidates. One important progression was made in the identification of MHC class II- restricted phosphopeptides by Depontieu et al. (Depontieu et al, 2009). As mentioned in the introduction chapter I, especially the finding that 2 phosphoproteins, tensin 3 and insulin receptor substrate 2 (IRS2), are the source for MHC class I- as well as MHC class II restricted phosphopeptides may provide an opportunity of a combinatorial treatment strategy using MHC class I- and MHC class II phosphopeptides for tumor immunity (Depontieu et al., 2009) (Zarling et al., 2006). A question hard to answer is how many phosphopeptides are necessary for vaccination trials. But there is need to discover more phosphopeptides to be able to look at MHC allele specificities and to establish algorithms. We only looked at phosphopeptides from one MHC allele. Therefore the discovery of phosphopeptide epitopes from other MHC alleles will be necessary to achieve population coverage. Ideally would be the detection of those phosphopeptides that have a broad repertoire for different MHC alleles. Another area of future studies will be the investigation in how far phosphorylation effects proteasomal processing or transport of the peptide antigen into the endoplasmic reticulum since little is known about that (Mohammed et al., 2008).

61 Abbas et al. (2004). Basic Immunology, Functions and Disorders of the Immune System. Second Edition. Elsevier Adams et al. (2002). PHENIX: Building new software for automated crystallographic structure determination. Acta crystallogr D 58: 1948-1954 Andersen et al. (1999). phosphorylated peptides can be transported by TAP molecules, presented by class I MHC molecules, and recognized by phosphopeptide-specific CTL. J Immunol 163(7): 3812-3818 Andersen et al. (2006). Cytotoxic T Cells. Journal of Investigative Dermatology 126: 32-41 Apostolopoulos et al. (1999). MUC1 and breast cancer. Curr Opin Mol Ther 1: 98-103 Baker et al. (2001). Electrostatics of nanosytems: Application to microtubules and the ribosome. PNAS 98: 10037-10041 Beck et al. (2006). quantitative proteomic analysis of posttranslational modifications of human histones. Mol Cell Proteomics 5:1314-1325 Bjorkmann et al. (1987). Structure of the human class I histocompatibility antigen HLA-A2. Nature 329: 506-512 Bright et al. 2002. SV40 Tag-specific cytotoxic T-lymphocytes generated from peripheral blood of malignant pleural mesothelioma patients. Cancer Immunol Immunotherapy 50: 682-690

Cantin et al. (2007). Optimizing TiO2-Based Phosphopeptide Enrichment for Automated Multidimensional Liquid Chromatography Coupled to Tandem Mass Spectrometry. Analytical Chemistry 79: 4666-4673 Cantley et al. (1991). Oncogenes and signal transduction. Cell 64: 281-302 Chen et al. (1997). A testicular antigen aberrantly expressed in human cancers detected by autologous antibody screening. PNAS 94: 1914-1918 Chen et al. (2000). Identification of NY-ESO-1 peptide analogues capable of improved stimulation of tumor reactive CTL. J Immunol 165: 948-955 Cohen (2001). The role of protein phosphorylation in human health and disease. European Journal of Biochemistry 268: 5001-5010 Cozzone et al. (1988). Protein phosphorylation in prokaryotes. Annual Review in Microbiology 42: 97-125 Cozzone et al. (2004). Protein phosphorylation on tyrosine in bacteria. Archives of Microbiology 181: 171-181 Depontieu et al. (2009). identification of tumor-associated, MHC class II-restricted phosphopeptides as targets for immunotherapy. PNAS 106( 29): 12073-12078

62 Emsley et al. (2004). Coot: Model-building tools for molecular graphics. Acta crystallogr D 60(Pt1): 2126-2132 Falk et al. (1991). Allele specific motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature 6324: 290-296 Foster et al. (2003). S-nitrosylation in health and disease. Trends Mol Med 9: 160-168 Gure et al. (2000). Sereological identification of embryonic neural proteins as highly immunogenic tumor antigens in small cell lung cancer. PNAS 97: 4198-4203 Haimes and Glover (1996). Molecular Immunology. Oxford University Press Han et al. (1992). Post translational chemical modifications of proteins. Int J Biochem 24: 19- 25 Hennecke et al. (2001). T cell receptor-MHC interactions close up. Cell 104:1-4 Hoffmann et al. (2008). Current approaches for global post-translational modification discovery and mass spectrometric analysis. Anal Chim Acta 627: 50-61 Hunter (1991). Cooperation between oncogenes. Cell 64: 249-270 Hurd et al. (2005). Disulphide formation on mitochondrial protein thiols. Biochem Soc Trans 33: 1390-1393 Ioannides et al. (1993). T Cell Recognition of Human Tumors: Implications for Molecular Immunotherapy of Cancer. Clinical Immunology and Immunopathology 66: 91-106 Jager et al. (1998). simultaneous humoral and cellular immune response against cancer testis antigen NY-ESO-1: Definition of human histocompatibility leukocyte antigen (HLA)- A2- binding epitopes. J. Exp Med 187(2): 265-270 Jager et al. (1999). Cancer-testis antigens and ING1 tumor suppressor gene product are breat cancer antigens: characterization of tissue specific ING1 transcripts and homologue gene. Cancer Res 59: 6197-6204 Jager et al. (2002). Clinical cancer vaccine trials. Curr Opin Immunol 14: 178-182 Jensen et al. (2006). interpreting the protein language using proteomics. Nat Rev Mol Cell Biol 7(6): 391-403 Jensen et al. (2007). Evaluation of the impact of some experimental procedures on different phosphopeptide enrichment techniques. Rapid Communications in Mass Spectrometry 21: 3635-3645 Kelly et al. (2005). How to study proteins by circular dichroism. Biochimica et biophysica Acta 1751: 119-139

63 Kittlesen et al. (1998). Human Melanoma patienst recognize an HLA-A1-restricted CTL epitope from tyrosinase containing two cysteine residues: implications for tumor vaccine development. J. Immunol 160:2099-2106 Kolch et al. (2000). meaningful relationships: the regulation of the Ras/Raf/MEK/ERK pathway by protein interactions. Biochem J. 351(2): 289-305 Krebs et al. (1956). The phosphorylase b to a converting enzyme of rabbit skeletal muscle. Biochimica et Biophysica Acta 20: 150-157 Larsen et al. (2005). Highly Selective Enrichment of Phosphorylated Peptides from Peptide Mixtures Using Titanium Dioxide Microcolumns. Molecular and Cellular Proteomics 4: 873-886 Larsen et al. (2006). Analysis of posttranslational modifications of proteins by tandem mass spectrometry. Biotechniques 40: 790-798 Lazoura et al. (2005). Rational peptide-based vaccine design for cancer immunotherapeutic applications. Curr Med Chem 12: 629-639 Lee et al. (1999). T-cell epitope mapping of the three most abundant extracellular proteins of mycobacterium tuberculosis in outbred guinea pigs. Infect Immun. 67(5): 2665-2670 Lee et al. (2005). Endocrin Rev 26: 147-170 Lewis et al. (2003). Tumor-associated antigens- from discovery to immunity. Int Rev Immunol 22:81-112 Li et al. (2010). Structural basis or the presentation of tumor-associated MHC class II- restricted phosphopeptides to CD4+ T cells. J Mol Biol 399: 596-603 Lochhead et al. (2005). activation-loop autophosphorylation is mediated by a novel transitional intermediate form of DYRKs. Cell 121: 925-936 Madden (1995). The three-dimensional structure of peptide-MHC complexes. Annu Rev Immunol 13: 587-622 Mann et al. (2002). Analysis of protein phosphorylation using mass spectrometry: deciphering the phosphoproteome. Trends in Biotechnology 20: 261-268 Massague et al. (2000). controlling TGF-beta signalling. Genes Dev 14(6): 627-644 Meadows et al. (1997). The HLA-A*0201-restricted HY antigen contains a post- translationally modified cysteine that significantly affects T cell recognition. Immunity 6: 273-281 Meyer et al. (2009). Identification of natural MHC class II presented phosphopeptides and tumor derived MHC class I phospholigands. J. Proteome Res 8: 3666-3674 Mohammed et al. (2008). Phosphorylation-dependent interaction between antigenic peptides

64 and MHC class I: a molecular basis for the presentation of transformed self. Nature immunol. 9: 1236-1243 Murray et al. (1998). The soluble granulocyte-macrophage colony-stimulating factor receptor`s carboxyl-terminal domain mediates retention of the soluble receptor on the cell surface through interaction with the granulocyte-macrophage colony-stimulating factor receptor beta-subunit. Biochemistry 37: 14113-14120 Murshudov et al. (1997). Refinement of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D 53(Pt3): 240-255 Nestle et al. (1998). Vaccination of melanoma patients with peptide- or tumor lysate-pulsed dendritic cells. Nat Med 4: 328-332 Ockert et al. (1999). Advances in cancer immunotherapy. Immunol Today 20: 63-65 Parkhurst et al. (1996). Improved induction of melanoma-reactive CTL with peptides from the melanoma antigen gp100 modified at HLA-A*0201-binding residues. J Immunol 175: 2539-2548 Parmiani et al. (2002). Cancer Immunotherapy With Peptide-Based Vaccines: What Have We Achieved? Where Are We Going?. Journal of the National Cancer Institute 11: 805- 818 Parnes et al. (1989). Molecular Biology and function of CD4 and CD8. Advances in Immunol 44: 265-311 Paschen et al. (2004). Identification of tumor antigens and T-cell epitopes, and its clinical application. Cancer Immunology Immunotherapy 53: 196-203 Pawson et al. (1994). SH2 and SH3 domains in signal transduction. Adv Cancer Res 64: 87- 110 Pawson et al. (1997). Signalling through scaffold, anchoring and adaptor proteins. Science 278(5346): 2075-2080 Pierce et al. (1999). Cutting edge: the HLA-A*0101-restricted HY minor histocompatibility antigen originates from DFFRY and contains a cysteinylated cysteine residue as identified by a novel mass spectrometric technique. J Immunol 163: 6360-6364 Pinkse et al. (2004). Selectivs Isolation at the Femtomole level of Phosphopeptides from Proteolytic Digests Using 2D-NanoLC-ESI-MS/MS and Titanium Oxide Precolumns. Analytical Chemistry 76: 3935-3943 Porath et al. (1975). Metal chelate affinity chromatography, a new approach to protein fractionation. Nature 258: 598-599 Preisinger et al. (2008). proteomics and phosphoproteomics for the mapping of cellular

65 signalling networks. Proteomics 8(24): 4402-4415 Purcell et al. (2001). Quantitative and qualitative influences of tapasin on the class I repertoire. J Immunol 166: 1016-1027 Purcell et al. (2003). Dissecting the Role of Peptides in the Immune Response: Theory, Practice and the Application to Vaccine Design. Journal of Peptide Science 9: 255- 281 Purcell et al. (2007). More than one reason to rethink the use of peptides in vaccine design. Nature Reviews 6: 404-41 Reinders et al. (2005). State-of-the-art in phosphoproteomics. Proteomics 5: 4052-4061 Rosenberg Philadelphia: Lipincott 2000 Rosenberg (2001). Progress in human tumor immunology and immunotherapy. Nature 411: 380-384 Rush et al. (2005). Immunoaffinity profiling of tyrosine phosphorylation in cancer cells. Nature Biotechnology 23: 94-101 Salih (2005). Phosphoproteomics by Mass Spectrometry and Classical Protein Chemistry Approaches. Mass Spectrometry Reviews 24: 828-846 Scanlan et al. (1998). characterization of human colon cancer antigens recognized by autologous antibodies. Int J Cancer 76: 652-658 Scanlan et al. (1999). antigens recognized by autologous antibody in patients with renal cell carcinoma. Int J Cancer 83: 456-464 Schaegger et al. (1987). Tricine-sodium dodecyl sulphate- polyacrylamide gel elctrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166: 368- 379 Seet et al. (2006). reading protein modifications with interaction domains. Nat Rev Mol Biol 7: 473-483 Seo et al. (2004). Post-translational modifications and their biological functions: proteomic analysis and systematic approaches. J Biochem Mol Biol 37: 35-44 Sherr (1996). Cancer cell cycles. Science 274: 1672-1677 Sitia et al. (2004). Sci STKE 2004: pe27 Skipper et al. (1996). An HLA-A2 restricted tyrosinase antigen on melanoma cells result from post-translational modification and suggests a novel processing pathway for membrane proteins. J Exp Med 183: 527-534 Slawson et al. (2006). O-GlcANC cycling: how a single sugar post-translational modification is changing the way we think about signalling networks. J Cell Biochem 97: 71-83

66 Slingluff et al. (1994). cytotoxic T-lymphocyte response to autologous human squamous cell cancer of the lung – epitope reconstitution with peptide extracted from HLA-AW68. Cancer Res 54(10): 2731-2737 Spickett et al. (2006). proteomic analysis of phosphorylation, oxidation and nitrosylation in signal transduction. Biochim Biophys Acta 1764: 1823-1841 Storoni et al. (2004). Likelihood-enhanced fast rotation functions. Acta Crystallogr D 60(Pt3) 432-438 Teh et al. (1988). Thymic major histocompatibility antigens and the alpha-beta-t-cell receptor determine the CD4/CD8 phenotype of T-cells. Nature 335: 229-233 Thingholm et al. (2006). Highly selective enrichment of phosphorylated peptides using titanium dioxide. Nature Protocols 1: 1929-1935 Timmermann at el. (1999). Dendritic cell vaccines for cancer immunotherapy. Annu Rev Med 50: 507-529 Tourdot et al. (2000). A general strategy to enhance immunogenicity of low-affinity HLA- A2.1 associated peptides: implication in the identification of cryptic tumor epitopes. Eur J Immunol 30: 3411-3421 Tureci et al. (1997). sereological analysis of human tumor antigens: molecular definition and implications. Mol Med Today 8:342-349 Tynan et al. (2005). T cell receptor recognition of a ‘super-bulged’ major histocompatibility complex class 1-bound peptide. Nature Immunology 6: 1114-1122 Unanue et al. (1987). The basis for the immunoregulatory role of macrophages and other accessory cells. Science 236: 551-557 Valmori et al. (1998). Enhanced generation of specific tumor-reactive CTL in vitro by selected Melan-A/MART-1 immunodominant peptide analogues. J Immunol 160: 1750-1758 Waldmann et al. (2003). Immunotherapy: past, present and future. Nature Medicine 9: 269- 277 Walsh et al. (2006). post-translational modifications in the context of therapeutic proteins. Nat Biotechnol 24: 1241-1252 Wang et al. (1999). Human tumor antigens: implications for cancer vaccine development. J Mol Med 77: 640-655 Webb et al. (2004). Functional and structural characteristics of NY-ESO-1- related HLA A2- restricted epitopes on the design of a novel immunogenic analogue. J Biol Chem 279: 23438-23446

67 Zagorska et al. (2004). HIF-1: the knowns and unknowns of hypoxia sensing. Acta Biochem Pol 51:563-585 Zanetti et al. (1987). The immunology of new gerneration vaccines. Immunol Today 8:18-25 Zarling et al. (2006). Identification of class I MHC-associated phosphopeptides as targets for cancer immunotherapy. PNAS 103(40): 14889-14894 Zarling et al. (2000). Phosphorylated peptides are naturally processed and presented by major histocompatibility complex class I molecules in vivo. J. Exp Med 192(2): 1755-1762 Zhou et al. (2001). A systematic approach to the analysis of protein phosphorylation. Nature Biotechnology 19: 375-378

Minerva Access is the Institutional Repository of The University of Melbourne

Author/s: Wurzbacher, Stephanie Julia

Title: Structural basis of tumor-associated phosphopeptides bound to the MHC molecule HLA A2 and stabilization studies

Date: 2011

Citation: Wurzbacher, S. J. (2011). Structural basis of tumor-associated phosphopeptides bound to the MHC molecule HLA A2 and stabilization studies. Masters Research thesis, Medicine, Dentistry & Health Sciences, Biochemistry and Molecular Biology and Bio 21 Institute, The University of Melbourne.

Persistent Link: http://hdl.handle.net/11343/36569

File Description: Thesis

Terms and Conditions: Terms and Conditions: Copyright in works deposited in Minerva Access is retained by the copyright owner. The work may not be altered without permission from the copyright owner. Readers may only download, print and save electronic copies of whole works for their own personal non-commercial use. Any use that exceeds these limits requires permission from the copyright owner. Attribution is essential when quoting or paraphrasing from these works.