The Murine Cytomegalovirus Immunoevasin Gp40 Binds to MHC Class I Molecules to Retain Them in the Early Secretory Pathway

The murine cytomegalovirus immunoevasin gp40 binds to MHC class I molecules to retain them in the early secretory pathway

Linda Ellen Janßen

A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biochemistry

Approved: Dissertation Committee

Prof. Dr. Sebastian Springer Jacobs University Bremen

Dr. Susanne Illenberger Jacobs University Bremen

Prof. Dr. Matthias Ullrich Jacobs University Bremen

Prof. Dr. Hartmut Hengel Universitätsklinikum Freiburg

Date of Defense: November 4th, 2015

Statutory Declaration

(Declaration on Authorship of a Dissertation)

I, Linda Ellen Janßen, hereby declare, under penalty of perjury, that I am aware of the consequences of a deliberately or negligently wrongly submitted affidavit, in particular the punitive provisions of § 156 and § 161 of the Criminal Code (up to 1 year imprisonment or a fine at delivering a negligent or 3 years or a fine at a knowingly false affidavit).

Furthermore I declare that I have written this PhD thesis independently, unless where clearly stated otherwise. I have used only the sources, the data and the support that I have clearly mentioned.

This PhD thesis has not been submitted for the conferral of a degree elsewhere.

Bremen October, 1st

Place Date

Signature

This work was funded by Tönjes Vagt Foundation of Bremen (grant XXIX to S.Sp.).

Parts of this work will be published in:

Janßen L, Ramnarayan V, Aboelmagd M, Iliopoulou M, Majoul I, Fritzsche S, Halenius A, Springer S. The mCMV protein m152/gp40 retains MHC class I molecules in the early secretory pathway by direct interaction. Manuscript in revision at Journal of Cell Science.

Abstract

To recognize the viral infection of a cell, the adaptive immune system depends on Major Histocompatibility Complex (MHC) class I molecules, which present viral peptides (antigens) at the surface of an infected cell to patrolling immune cells. MHC class I molecules are assembled and loaded with viral peptides in the endoplasmic reticulum (ER), and, after passing a thourough quality control, are transported to the cell surface.

Viruses have evolved mechanisms to impair the so-called antigen presentation, as this reduces viral dissemination. Herpesviruses, which usually chronically infect their host, dedicate a big part of their ample genome to manipulate the antigen presentation pathway at all possible levels.

In the presence of the murine cytomegalovirus (mCMV) gp40 (m152) protein, murine MHC class I molecules do not reach the cell surface but are retained in an early compartment of the secretory pathway by an unknown mechanism.

In this work, I found that gp40 does not hijack any known cellular factors to retain MHC class I molecules, but rather binds to them and most likely circulates with them in the early secretory pathway, which consists of the endoplasmic reticulum (ER), the ER-Golgi intermediate compartment (ERGIC), and the cis- Golgi.

A flexible sequence in the lumenal domain of gp40 appears to be responsible for the circulation, as destruction of this sequence releases both MHC class I molecules and gp40 from the early secretory pathway without impairing their interaction.

Furthermore, I could show that the expression of gp40 influences the transcription of genes of the antigen presenting machinery (APM) in some, but not in all, cell lines, and that factors that influence the folding and export speed of MHC class I molecules (protein sequence, β2m abundance) decide on the effectiveness of gp40 function.

Acknowledgments

After almost five years working at this inspiring small university, I finish the adventurous journey of my PhD thesis. It was a truly great time and I will miss it.

First of all, I want to thank Sebastian Springer for giving me the opportunity to do this exciting project. Thank you so much for all your support, experience, and, most importantly, your undying passion for good science.

I want to also thank my thesis committee, Dr. Susanne Illenberger, Prof. Dr. Hartmut Hengel and Prof. Dr. Matthias Ullrich, for your intellectual input throughout this project and for reviewing my research work.

Apart from the science, the former and present members of the Springer laboratory contributed largely to make this time a particularly joyful one: many thanks to Britta Borchert, Cindy Kroll, Esam Abualrous, Peter Reinink, Sebastiàn Montealegre, Sunil Kumar, Sujit Verma, Susi Fritzsche, Venkat Ramnarayan, and Zeynep Hein, for being so positive, supportive, and smart. I am very spoiled now.

Special thanks to ‘Super-Uschi’ Ursula Wellbrock – this lab would not function without you, you are certainly irreplaceable.

I want to also thank the students that have worked with me during these years, Ina Huppertz, Maria Bottermann, Sharon Versteeg, Nikki Dimitrova Atanasova, Vaishnavi Venugopalan, Mohamed Aboelmagd, and Maria Iliopoulou, for contributing to this thesis with your hard work and good ideas.

Many thanks to Anne Halenius, for inspiring discussions and for investing so much time and energy into teaching me the methods that made many of our data possible.

I would like to thank the Tönjes-Vagt-Foundation for their funding in this project.

Ich möchte mich bei meiner Familie, meiner Mutter Rita, meiner Schwester Evelyn, meinem Vater Kalle, und Marion, dafür bedanken, dass ihr mich bedingungslos bei all meinen Vorhaben unterstützt.

Many thanks to Mehdi, my partner, for being the awesomest and most supportive person. Kheili dusset daram!

Table of Contents

1 Introduction ...... 11 1.1 The immune response to viruses ...... 11 1.1.1 The innate immune system ...... 11 1.1.2 Natural killer cells ...... 12 1.1.3 The adaptive immune system ...... 13 1.1.4 Interferons ...... 13 1.2 MHC class I trafficking ...... 15 1.2.1 Synthesis and assembly ...... 16 1.2.2 MHC class I retention and retrieval ...... 18 1.3 Transcription of APM genes ...... 21 1.4 ER stress ...... 22 1.5 Viral inhibition of the immune response ...... 24 1.5.1 General examples and herpesviruses ...... 24 1.6 Cytomegalovirus ...... 27 1.6.1 HCMV from a medical perspective ...... 27 1.6.2 Infection and latency ...... 27 1.7 gp40 ...... 28 1.7.1 Discovery of gp40 ...... 28 1.7.2 Background information about gp40 ...... 28 1.7.3 Comparison of gp40 sequence of different MCMV strains ...... 30 1.7.4 gp40 in the viral context ...... 32 1.7.5 The gp40 mystery ...... 35 1.8 The aim of this study ...... 36 2 Materials and Methods ...... 38 2.1 Constructs ...... 38 2.2 Domain swaps ...... 43 2.3 Cell culture ...... 44 2.4 Protein expression in mammalian cells ...... 45 2.4.1 Lentiviral transduction ...... 45 2.4.2 Preparation of lentiviral particles ...... 46 2.4.3 Transduction of mammalian cells ...... 47 2.4.4 Selection of transduced mammalian cells with puromycin ...... 48 2.4.5 Tunicamycin treatment ...... 48

2.4.6 IFN-α and IFN-γ treatment ...... 49 2.5 Immunostaining ...... 50 2.5.1 Immunofluorescence for microscopy ...... 50 2.5.2 Immunofluorescence for FACS ...... 51 2.6 Biochemical experiments ...... 52 2.6.1 Radioactive pulse-label ...... 52 2.6.2 Pulse-chase ...... 52 2.6.3 Immunoprecipitation ...... 52 2.6.4 Co-immunoprecipitation ...... 53 2.6.5 Re-immunoprecipitation ...... 54 2.6.6 EndoF1 digest ...... 54 2.6.7 PNGase digest ...... 55 2.6.8 Folding of MHC class I ...... 55 2.6.9 Thermostability of MHC class I ...... 55 2.6.10 Surface-immunoprecipitation using SL8-biotin ...... 56 2.6.11 SDS-PAGE ...... 56 2.6.12 Autoradiographic analysis ...... 56 2.7 Transcription ...... 58 2.7.1 Primer design ...... 58 2.7.2 mRNA isolation ...... 59 2.7.3 Determining RNA concentration ...... 60 2.7.4 DNase digest ...... 60 2.7.5 Reverse transcription of mRNA ...... 61 2.7.6 Quantitative PCR ...... 61 3 Manuscript ...... 64 3.1 The mouse cytomegalovirus protein m152/gp40 retains MHC class I molecules in the early secretory pathway by direct interaction...... 65 3.2 Abstract ...... 66 3.3 Introduction ...... 66 3.4 Results ...... 67 3.4.1 Gp40 retains MHC class I in the early secretory pathway ...... 67 3.4.2 Gp40 does not impair class I maturation or peptide binding ...... 70 3.4.3 Gp40-mediated retention of class I does not use the peptide loading complex ...... 73 3.4.4 Gp40 binds directly to Db and Kb ...... 73 3.4.5 The class I – gp40 interaction persists in the early secretory pathway ...... 74

3.4.6 A sequence in the linker of gp40 is essential for gp40–class I complex retention ...... 75 3.5 Discussion ...... 77 3.6 Materials and Methods ...... 79 3.7 Acknowledgements ...... 81 3.8 References ...... 86 3.9 Additional data on MHC class I interaction with gp40 and other proteins ...... 89 3.9.1 MHC class I binds weakly to gp40 linker mutants and the complex is exported together . 89 3.9.2 Prolonged interaction of MHC class I with the PLC is specific to wild type gp40 ...... 93 3.9.3 Identification of bands that coimmunoprecipitate with MHC class I ...... 94 3.9.4 Discussion ...... 95 4 Transcription of APM genes in gp40-expressing cells ...... 102 4.1 Transcription of ER stress markers after tunicamycin treatment ...... 102 4.2 Does the expression of gp40 cause ER stress? ...... 103 4.3 Is APM transcription altered in gp40-expressing cells? ...... 104 4.3.1 APM transcription in MEF CNX wt cells ...... 104 4.3.2 APM transcription in K41 cells ...... 106 4.4 Treatment with IFN-α and IFN-γ ...... 108 4.4.1 Surface class I levels in MEF CNX wt cells after IFN treatment ...... 108 4.4.2 Surface class I levels in K41 cells after IFN treatment...... 109 4.4.3 APM transcription in MEF CNX wt cells after IFN treatment ...... 110 4.4.4 APM transcription in K41 cells after treatment with IFN ...... 111 4.4.5 gp40 transcription in MEF CNX wt and K41 cells after IFN treatment ...... 113 4.4.6 gp40 expression in RMA cells ...... 114 4.4.7 Discussion ...... 115

5 The influence of β2m on gp40 function ...... 119 5.1 The effect of gp40 on human class I surface expression in HEK293T cells ...... 119 5.1.1 Murine class I molecules are not transported to the surface in HEK293T cells ...... 120

5.1.2 Murine class I molecules become resistant to gp40 with β2m overexpression ...... 121

5.1.3 Dose-response effect of hβ2m cotransfection on murine MHC class I surface levels in the presence of gp40 ...... 122

5.1.4 With little β2m available, murine MHC class I molecules are sensitive to gp40 ...... 124

5.1.5 Additional β2m in murine cells protects MHC class I from gp40 in an allotype-specific way 125 5.1.6 Discussion ...... 127 6 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40 ...... 134

6.1 Establishment of Db/Kb domain swaps in gp40-expressing L cells ...... 134 6.1.1 Are the endogenous molecules Dk and Kk in L cells retained by gp40? ...... 137 6.1.2 Discussion ...... 138 7 Outlook ...... 139 8 Additional publications ...... 141 9 References ...... 142

List of Figures

Figure 1-1 JAK/STAT pathway after interferon binding ...... 14 Figure 1-2 Structure of an MHC class I molecule ...... 16 Figure 1-3 MHC class I export and recycling ...... 18 Figure 1-4 Gene structure of human and murine MHC ...... 21 Figure 1-5 Unfolded protein response in the ER...... 23 Figure 1-6 gp40 crystal structure ...... 29 Figure 1-7 gp40 docked to H-2Dd ...... 30 Figure 1-8 Sequence alignment of gp40 of different MCMV strains ...... 32 Figure 2-1 Sequence of the gp40-HA construct used ...... 39 Figure 2-2 gp40-HA in puc2Cl6IEGwo ...... 40 Figure 2-3 gp40-HA in puc2Cl6IPwo ...... 40 Figure 2-4 gp40-KKSQ-HA in puc2Cl6IPwo ...... 41 Figure 2-5 gp40-GFP fusion in puc2Cl6IPwo ...... 42

Figure 2-6 hβ2m in puc2Cl6INwo ...... 42 Figure 2-7 Kb and Db domain swap constructs ...... 43 Figure 2-8 Principle of lentiviral transduction ...... 46 Figure 3-1 Gp40 intracellularly retains MHC class I...... 68 Figure 3-2 The PLC is not involved in gp40-mediated retention...... 72 Figure 3-3 Gp40 interacts with class I...... 74 Figure 3-4 The gp40-class I complex cycles through the early secretory pathway...... 75

Figure 3-5 Gp40-(G4S)9 rapidly exits the early secretory pathway, and does not retain class I...... 76 Figure 3-6 Proposed mechanism for gp40-mediated MHC class I retention...... 79 Figure 3-7 In the presence of gp40, Kb and Db localize to juxtanuclear compartments, and Db additionally to the ER...... 82 Figure 3-8 Characterization of retained MHC class I molecules...... 84 Figure 3-9 Retrieval by a KKXX signal leads to partial EndoF1 resistance...... 84 Figure 3-10 gp40 interacts with PLC components but not human MHC class I...... 85 Figure 3-11 Only little gp40-Gp40-LDel binds to MHC class I and is exported together with it ...... 89

Figure 3-12 Only little gp40-(G4S)9 binds to MHC class I and is exported together with it ...... 91

Figure 3-13 MHC class I binds weakly to gp40-(G4S)9 and is exported together with it ...... 92

Figure 3-14 Prolonged interaction of MHC class I with the PLC is specific to wild type gp40 ...... 93 Figure 3-15 Identification of bands that coimmunoprecipitate with MHC class I ...... 94 Figure 4-1 Transcription of most APM genes is reduced in gp40-expressing MEF CNX wt cells ...... 105 Figure 4-2 Transcription of APM genes is slightly increased in gp40-expressing K41 cells ...... 106 Figure 4-3 Surface Kb in RMA cells is not reduced by gp40 expression ...... 114 Figure 4-4 Surface Kb in RMA cells is slightly reduced by very high gp40 levels ...... 114 Figure 4-5 Overconfluency increases MHC class I surface expression in K41 cells ...... 117 Figure 5-1 gp40 expression does not alter human MHC class I surface expression in HEK293T cells ...... 119 Figure 5-2 Murine MHC class I molecules are not expressed at the cell surface of HEK293T cells ...... 120

Figure 5-3 gp40 does not retain murine MHC class I in HEK293T overexpressing hβ2m...... 121

Figure 5-4 Murine MHC class I surface levels increase linearly with increasing cotransfected hβ2m plasmid ...... 123

Figure 5-5 Low levels of β2m restore the gp40 phenotype on murine MHC class I surface levels ...... 124 b b Figure 5-6 Supertransduction of hβ2m slightly increases surface D and K ...... 126 b Figure 5-7 Supertransduction of hβ2m protects K in murine K41 from gp40 inhibition ...... 126 Figure 5-8 Alignment of murine and human MHC class I sequences...... 129 Figure 5-9 Residues in murine MHC class I that might play a role in interaction with gp40 ...... 130 Figure 6-1 Sequence alignment of H-2Db and H-2Kb shows 84% identity ...... 135 Figure 6-2 Kb and Db domain swap constructs ...... 135 Figure 6-3 The lumenal domain of MHC class I determines the effectivity of gp40 ...... 136 Figure 6-4 Dk is more affected by gp40 than Kk ...... 137

List of Tables

Table 2-1 Primers used for cloning of the gp40 gene ...... 38 Table 2-2 Mammalian cell lines ...... 44 Table 2-3 Primary antibodies for microscopy ...... 50 Table 2-4 Primary antibodies for flow cytometry ...... 51 Table 2-5 Antibodies used for immunoprecipitation ...... 53 Table 2-6 Primers used for real time PCR...... 58 Table 2-7 cDNA synthesis reaction ...... 61 Table 2-8 qPCR synthesis reaction...... 62 Table 2-9 Thermocycling conditions for qPCR reaction ...... 62 Table 3-1 Comparison of our results with the results of Koszinowski and coworkers ...... 100 Table 4-1 ER stress markers are enhanced after treatment with ER stressor tunicamycin ...... 102 Table 4-2 ER stress markers are not enhanced in cells expressing gp40 ...... 103 Table 4-3 Transcription of most APM genes is reduced in gp40-expressing MEF CNX wt cells ...... 105 Table 4-4 Transcription of APM genes is slightly increased in gp40-expressing K41 cells ...... 107 Table 4-5 MFI of Kb and Db surface expression in MEF CNX wt cells treated with IFN-α and IFN-γ ...... 108 Table 4-6 MFI of Kb and Db surface expression in K41 cells treated with IFN-α and IFN-γ ...... 109 Table 4-7 APM gene transcription in gp40-expressing MEF CNX wt cells can be rescued by treatment with IFN-α and IFN-γ ...... 110 Table 4-8 Constitutively expressed proteasome subunits are not increased by treatment with IFN ...... 111 Table 4-9 APM gene transcription in gp40-expressing K41 cells is increased after treatment with IFN-α and IFN-γ ...... 112 Table 4-10 Constitutively expressed proteasome subunits and calnexin are not increased by treatment with IFN ...... 112 Table 4-11 gp40 transcripts are strongly reduced after IFN treatment ...... 113 Table 5-1 MHC class I allotypes and cell lines tested with gp40 ...... 127

Table 5-2 Conserved residues in the murine MHC class I α1 and α2 sequence that are different in human ...... 131

Abbreviations and Symbols

APC Allophycocyanin APM antigen presenting machinery

β2m beta-2 microglobulin µ micro Bap31 B-cell receptor-associated protein 31 BFA brefeldin A coIP co-immunoprecipitation COPI coat protein I complex CT cytosolic tail DNA deoxyribonucleic acid EndoF1 Endoglycosidase F1 EndoH Endoglycosidase H ER endoplasmic reticulum ERGIC ER-Golgi intermediate compartment FACS fluorescence activated cell sorting FCS fetal calf serum H-2 histocompatibility-2 HLA human leukocyte antigen HCMV human cytomegalovirus IAA iodoacetic acid IFN interferon IP immunoprecipitation kDa kilodalton L liter MFI mean fluorescence intensity MHC major histocompatibility complex min minute NK cells natural killer cells PAGE polyacrylamide gel electrophoresis PBS phosphate buffered saline PCR polymerase chain reaction PEI polyethylenimin PMSF phenylmethylsulfonylfluoride PVDF polyvinylidene fluoride qPCR quantitative PCR reIP re-immunoprecipitation RNA ribonucleic acid SDS sodium dodecyl sulfate SL8 SIINFEKL, Kb specific peptide from ovalbumin (357-64) TMD transmembrane domain UGT1 UDP-glucose:glycoprotein glucosyltransferase x g fold gravity

10 Introduction

1 Introduction

1.1 The immune response to viruses

The body is continually invaded by organisms or viruses that can have pathological consequences for the host, such as disease or the formation of tumors. Therefore, the need for self defense has evolved, on the basic principle of recognizing non-self structures and destroying them. This is called an immune response.

In higher vertebrates, immunity is mediated by two closely interacting parts: the innate and the adaptive immune system. Briefly, the innate immune response is faster, evolutionarily older, and it reacts to conserved patterns on or inside the invaders. The adaptive immune system reacts more slowly, but very specifically to individual pathogens, and it can form a long-lasting immunity against a certain pathogen, the so-called immunological memory.

1.1.1 The innate immune system

The innate immune system is highly conserved and also found in simple organisms (see also (Janeway et al., 2001b)). Due to the complexity of this system this text will focus only on some points.

Generally, the innate immune system is the first to react to an infection and serves to either clear it or keep it under control until the adaptive immune response sets in. The innate immune response reacts to patterns that are generally found on pathogens – such as bacterial carbohydrates, bacterial or viral DNA or RNA, bacterial peptides and many more – but not on self tissue.

At the site of an infection, the first response is usually inflammation, stimulated by cytokines (such as interferon (IFN), see 1.1.4) that are released by injured cells. Inflammation alerts the neighboring cells and restricts the spread of the infection. The secreted cytokines stimulate cells of the innate immune system such as mast cells, macrophages, neutrophils, NK cells (see 1.1.2), and dendritic cells. These cells play different roles in clearing the infection. The neutrophils, for example, which reach the site of infection first, secrete more cytokines, engulf microorganisms, and release antimicrobial proteins.

11 Introduction

Natural killer (NK) cells play a very important role in the recognition and elimination of virally infected cells (see 1.1.2).

Cells of the innate immune system also bear different forms of extracellular or intracellular toll-like receptors, which are able to identify conserved pathogenic patterns.

Another cell-free factor is the complement system, which circulates in an inactive form in the blood, but upon recognition of a pathogen, it forms a cytotoxic membrane attack complex, leading to the destruction of the pathogen.

1.1.2 Natural killer cells

Cytotoxic T cells, which belong to the adaptive immune system, recognize infected cells by antigen presentation via major histocompatibility complex (MHC) class I molecules. Viruses have therefore evolved strategies to reduce the surface levels of MHC class I molecules (see 1.5), which prevents killing of the infected cell by cytotoxic T cells. NK cells fill this surveillance gap by attacking cells that have low MHC class I surface levels, i.e., by recognition of the ‘missing self’ signal.

The mechanism of killing by NK cells is similar to that of cytotoxic T cells. The cytotoxic granules that they release contain perforins and effector proteins; the former help the latter to enter the target cell and induce its apoptosis.

NK cells display two kinds of receptors on their surface: activating receptors, which trigger killing by the NK cells, and inhibiting receptors, which prevent killing. Activating receptors recognize stress-induced ligands such as MICA in humans or RAE-1 in mice (Bauer et al., 1999; Cerwenka et al., 2000). Inhibitory receptors recognize classical MHC class I molecules or non-classical HLA-E in humans or Qa1 in mice (these bind the signal peptides of classical MHC class I molecules and thereby help indirectly to monitor MHC class I expression). In a viral infection, when transcription of classical MHC class I or their surface transport is impaired, the reduced surface levels of these inhibitory ligands switch off the inhibitory signal in the NK cell (Kärre et al., 1986; Ljunggren and Kärre, 1990). Whether an NK cell activates the killing mode is decided after integration of inhibiting and activating signals.

NK cells are activated by cytokines such as IL-15, IL-12, or IL-18 and by type I IFN (see 1.1.4), and they secrete IFN-γ and TNF, which in turn activates T cells, dendritic cells, and macrophages.

12 Introduction

1.1.3 The adaptive immune system

The adaptive immune system responds more slowly to an infection than the innate immune system, especially when it meets the pathogen for the first time. The recognition of a specific antigen leads to the massive activation of the immune cells that react to this antigen. Later, an immune memory is formed: Some of the immune cells that reacted to the pathogenic signal are kept in a differentiated but dormant state, and can react quickly and powerfully should the pathogen recur.

CD8+ T cells recognize a virally infected cell by the display of a short viral peptide (the antigen) in complex with a protein called an MHC class I molecule on the surface of the infected cell. For this recognition, the immune cell needs a receptor that fits perfectly to the presented antigen/MHC class I complex, even if a virus has infected the organism for the first time.

This principle is realized by a phenomenon called V(D)J recombination, which happens in the early stages of the life of a T cell. During this process, gene segments of the receptor genes are randomly rearranged, leading to an enormous variety in T cell receptors. After the removal of those immature T cells that react to self antigens, the remaining T cells monitor the surface of antigen presenting cells. In the case of a specific recognition (the target of which is assumed to be non-self and therefore potentially pathogenic), the T cells proliferate to produce an effective response against the intruders.

CD8+ T cells kill cells that present non-self antigen on MHC class I molecules, whereas CD4+ T cells (which recognize antigens on MHC class II molecules) do not kill but regulate the immune response by secreting cytokines. A subset of CD4+ T cells activates B cells, which then proliferate into plasma cells to secrete antibodies. These bind pathogens and render them harmless or opsonize them for phagocytosis by macrophages.

1.1.4 Interferons

Cytokines play an important role in the communication between cells in the body. Interferons (IFN) are cytokines that warn neighboring cells of an infection and enable them to prepare their defense (Platanias, 2005). Interferons can be sorted into three types (type I, type II and type III), but only the first two will be discussed here.

13 Introduction

The type I interferons comprise several forms of IFN-α and one IFN-β, which are secreted by virally infected fibroblasts and leukocytes. Both interferons bind the same cellular receptor called IFNAR, which activates the JAK/STAT pathway (see Figure 1-1): JAK, which is attached to IFNAR, first autophosphorylates and then phosphorylates the cytosolic factors STAT1 and STAT2 (van Boxel-Dezaire et al., 2006; Calò et al., 2003). Phosphorylated STAT1 and STAT2 heterodimerize and travel into the nucleus, where they activate transcription via the IFN-stimulated response elements (ISREs). Alternatively, STAT1 can homodimerize and bind to IFN-γ-activated site (GAS), which can also be activated by IFN-γ.

Figure 1-1 JAK/STAT pathway after interferon binding Type I or type II IFN bind to their respective receptors and the receptor associated JAK kinases phosphorylate STATs, which dimerize. The STAT complexes then localize into the nucleus to activate transcription from the ISRE or GAS elements, resulting in the transcription of antiviral genes. Author’s own work, based upon (Platanias, 2005).

14 Introduction

The type II interferon, IFN-γ, is secreted by T cells and NK cells when they recognize infected cells. IFN-γ binds to a different receptor called IFNGR, which is expressed on almost all cell types (Valente et al., 1992). Activation of this receptor leads to homodimerization of STAT1, which in turn activates GAS (Decker et al., 1997).

IFN stimulation leads to the activation of many different genes that impair viral reproduction (Der et al., 1998), including MHC class I and some antigen presenting machinery (APM) genes (see 1.3) as well as the MHC class II transactivator CIITA (Martin et al., 1997). Other pathways are – for example – the activation of oligoadenylate synthetase, an antiviral enzyme that degrades viral RNA, and the downregulation of the translation of both host and viral proteins by the phosphorylation of elF2α (Der et al., 1998). Both type I and type II interferons also plays an important role in regulating the adaptive immune system (González-Navajas et al., 2012; Schoenborn and Wilson, 2007).

1.2 MHC class I trafficking

MHC class I molecules are surface glycoproteins that are present on every nucleated cell. The extremely polymorphic heavy chains (about 10,000 alleles known in humans) form a trimer together with the invariant beta-2 microglobulin (β2m) and the antigenic peptide that is presented to the cytotoxic T cells. The antigenic peptides are generated by degradation of cellular proteins, and therefore the MHC class I molecules display a sample of the cellular protein composition on the cell surface. This enables patrolling immune cells to recognize a viral infection or tumor formation.

In humans, the nomenclature for the MHC is as follows: the MHC is called the HLA (human leukocyte antigen) region, and the three different gene loci are called HLA-A, -B, and -C. The individual alleles as well as the proteins they encode are further described by numbers, for example HLA-A*02:01. In mice, the MHC is called the H-2 region, with its loci H-2D, K, and L. The different alleles are denoted with a superscript, for example H-2Db.

15 Introduction

1.2.1 Synthesis and assembly

MHC class I heavy chains are synthesized in the ER, where they are also glycosylated. Heavy chains either have one (human) or two or three (murine) glycosylation sites, which improves their folding and stability (Shental-Bechor and Levy, 2008). The heavy chain comprises two domains, the α3 domain, which is membrane proximal, and the α1/α2 superdomain (see Figure 1-2), which is relatively conformationally unstable and only folds after association with β2m. The α1/α2 superdomain forms the peptide binding groove, into which the antigenic peptide is bound and presented at the surface.

Figure 1-2 Structure of an MHC class I molecule The MHC class I molecule is a trimer consisting of the heavy chain, which forms the α1/α2 superdomain and the α3 domain (red), β2m (pink), and the antigenic peptide (blue). The peptide is bound in the so-called peptide binding groove, which is formed by the α1/α2 superdomain. Drawn by the author with VMD.

After cotranslational insertion into the ER, the heavy chain first associates with the chaperone calnexin

(Ou et al., 1993). Afterwards, when β2m binds, calnexin dissociates (Degen and Williams, 1991; Degen et al., 1992; Nössner and Parham, 1995) (see Figure 1-3)1. After dissociation of calnexin, calreticulin binds to the glycosylation site and incorporates MHC class I into a multi-chaperone complex called the peptide

1 In mice, calnexin and β2m are sometimes bound to MHC I class at the same time, possibly because of the additional glycosylation sites on murine MHC class I.

16 Introduction loading complex (PLC) (Li et al., 1999; Ortmann et al., 1994; Suh et al., 1994). Calreticulin is bound to the b and b′ domains of ERp57 (Frickel et al., 2004; Hughes and Cresswell, 1998), which in turn forms a disulfide bond with tapasin (Dong et al., 2009; Peaper et al., 2005). ERp57 is a disulfide isomerase that assists MHC class I in forming its disulfide bonds (Antoniou et al., 2002; Dick et al., 2002; Lindquist et al., 2001; Tector et al., 1997), and in ERp57 knockout mice, the assembly and loading of MHC class I molecules is impaired (Garbi et al., 2006). The central component of the PLC is tapasin, which has two functions: it stabilizes suboptimally loaded MHC class I, and it assists in iterative peptide exchange (Garstka et al., 2011; Lehner et al., 1998; Praveen et al., 2010; Tan et al., 2002; Williams et al., 2002). After successful binding of a high-affinity peptide, the MHC class I molecule dissociates from tapasin (and thus the PLC) and moves on (Rizvi and Raghavan, 2006; Wearsch and Cresswell, 2007). Tapasin itself associates with the transporter associated with antigen processing (TAP), which translocates the antigenic peptides (generated in the cytosol) into the ER lumen (Lankat-Buttgereit and Tampé, 1999; Neefjes et al., 1993). These peptides originate from defective ribosomal products (DRiPs) (Yewdell and Nicchitta, 2006) and from mature proteins (Farfán-Arribas et al., 2012), which are both digested by the cytosolic proteasome. TAP apparently delivers mainly N-terminally extended peptides, which, after import into the ER, are further trimmed by an ER-resident aminopeptidase (ERAAP or ERAP1) (Saric et al., 2002; Serwold et al., 2002). Only peptides of the right length of eight to ten amino acids are ever tightly bound by MHC class I molecules.

Whether bound to the PLC or not, suboptimally loaded class I molecules are held, at least for some time, in a pre-medial Golgi compartment, i.e., their glycans do not become resistant to EndoH or EndoF1 digestion (see 2.6.6) (Garstka et al., 2007). After binding a high-affinity peptide, the MHC class I molecule is exported to the plasma membrane (Abe et al., 2009; Spiliotis et al., 2000). The rate at which different class I allotypes travel varies considerably (Fritzsche et al., 2015; Williams et al., 1985).

It is so far not well understood how class I molecules are recognized for export out of the ER (if at all). Briefly, a protein called Bap31 was thought to play a role, since it interacts with MHC class I. The effect of a Bap31 knockout on MHC class I export is minimal, which speaks against an indispensable role of it in MHC class I export (Abe et al., 2009; Fritzsche, 2014; Ladasky et al., 2006; Paquet et al., 2004).

17 Introduction

Figure 1-3 MHC class I export and recycling The MHC class I free heavy chain (bottom left; red) first associates with calnexin (CNX, light blue, left) until it folds and binds to β2m (pink). Afterwards, it is incorporated into the PLC, consisting of TAP (light pink), tapasin (yellow), ERp57 (light green), and calreticulin (CRT, light blue, right). Here, it binds a peptide, which is either a high-affinity ligand (dark blue pentagon) or a low-affinity ligand (dark green triangle). The MHC class I molecule is released from the PLC and exported to the cis-Golgi, where it either proceeds to the cell surface or is retrieved to the ER for another round of peptide binding. Adapted from Prof. Springer.

1.2.2 MHC class I retention and retrieval

MHC class I molecules are trimers and generally only resistant to denaturation when all components are assembled. Due to allotypic variation in the sequence of the peptide binding groove, not all peptides bind with the same affinity to all allotypes, but some peptides show a higher affinity to certain allotypes than others. The higher the affinity of a peptide to binding groove, the higher the resistance of the MHC class I trimer to denaturation (Cerundolo et al., 1990; Townsend et al., 1990).

It is important for the functioning of class I antigen presentation that the MHC class I trimers are stable, mainly for two reasons: unstable MHC class I molecules on the cell surface are quickly endocytosed (Montealegre et al., 2015; Zagorac et al., 2012), and therefore they might not be able to perform their

18 Introduction job of presenting antigen long enough for the adaptive immune system to recognize them. Second, an unstable MHC class I molecule might lose its peptide on the surface, and then re-bind another peptide from the extracellular medium. In that case, the cytotoxic T cells might not recognize an infected cell, or they might falsely assume that a certain cell is infected and kill it.

The cellular quality control has many means to check for the conformational stability of a secreted protein and to maximize the chances of protein folding. Additionally, unstable proteins are mostly held back in the ER to give them another attempt to fold and assemble properly. These general mechanisms are also employed to survey class I molecules for folding and high-affinity peptide binding.

Binding of β2m to the heavy chain is the first precondition for successful surface transport. Most MHC class I allotypes will not be exported without β2m, but rather get degraded; some allotypes may not fold in the absence of β2m in the first place (Myers et al., 1989; Williams et al., 1989; Zijlstra et al., 1990).

Some allotypes apparently reach the cell surface without β2m (Allen et al., 1986), but they are most likely immediately endocytosed. Also, cytotoxic T cells will probably not see them, since their binding groove is probably largely unfolded (Hansen et al., 1988). Individual exceptions such as Db and Ld exist (Myers et al., 1989).

Calnexin is the first chaperone that associates with the MHC class I free heavy chain at its glycosylation site and prevents its misfolding. It is capable of retaining misfolded MHC class I molecules, as shown in insect cells (Jackson et al., 1994). Other chaperones such as BiP might take over its function if the supply of calnexin is limited. The disulfide bond in the α3 domain is formed rapidly (Tector et al., 1997), probably with the help of the protein disulfide isomerase PDI.

The PLC is the next step in quality control. It monitors proper peptide binding to MHC class I. Tapasin associates with MHC class I and helps it exchange its peptides. It dissociates from MHC class I when a high affinity-peptide is bound (see 1.2.1). Tapasin also retrieves suboptimally loaded MHC class I molecules with the help of its dilysine motif (Grandea et al., 2000; Paulsson et al., 2006); however, in tapasin-deficient cells, suboptimally loaded class I molecules can still be retrieved to the ER (Garstka et al., 2007; Zernich et al., 2004).

Calreticulin binds at the glycosylation site that was previously occupied by calnexin. Calreticulin is also a very important chaperone for retaining misfolded MHC class I molecules, since in its absence, class I exits quickly but loaded with mainly low-affinity peptides (Gao et al., 2002; Howe et al., 2009). Possibly, calreticulin and tapasin work together to retrieve misfolded MHC class I molecules.

19 Introduction

While tapasin is an MHC class I-specialized chaperone, calreticulin takes care of many glycoproteins, which it binds with its lectin site (Bedard et al., 2005). It works in concert with two other proteins called glucosidase II and UDP-glucose:glycoprotein glucosyltransferase (UGT1) (Caramelo et al., 2003; Wearsch et al., 2011) as parts of the so-called UGT cycle (Hammond et al., 1994; Sousa and Parodi, 1995). In this cycle, the lectin calreticulin binds to the monoglucosylated glycosylation site of a glycoprotein and dissociates after a while. Glucosidase II then removes the terminal glucose, and calreticulin cannot rebind. Next, UGT1 investigates the glycoprotein to look for partially unfolded patches; if the protein is not satisfactorily folded, it will reattach the terminal glucose residue. Calreticulin will now bind again and the protein has another attempt at folding properly. UGT1 nevertheless seems to play only a minor role in MHC class I quality control, at least as long as the PLC is intact (Zhang et al., 2011; and V. Ramnarayan, data not shown). Suboptimally loaded MHC class I molecules in PLC-defective cells circulate between the ER and the cis-Golgi in an attempt to fold well (Garstka et al., 2007; Hsu et al., 1991).

If a glycoprotein does not manage to fold after several attempts, it is labeled as terminally misfolded by the slow removal of its mannose residues by the EDEM (ER degradation mannosidase-like) protein. The protein is then retro-translocated to the cytosol through Sec61 and afterwards degraded by the proteasome (Hughes et al., 1997; Tsai et al., 2002). This process is called ER-associated degradation (ERAD) (Vembar and Brodsky, 2008).

Experiments in insect cells show the significance of chaperones and the quality control, since in these cells, MHC class I molecules do not fold well but are exported anyways in an unfolded state, unless chaperones such as calnexin or tapasin are co-expressed (Jackson et al., 1994; Schoenhals et al., 1999; Vassilakos et al., 1996).

Even in mammals, the quality control is not entirely efficient. Especially in murine cells, it appears that around 50% of surface MHC class I are loaded with low-affinity peptide and have escaped the quality control nevertheless (Su and Miller, 2001); in TAP-deficient cells, ER export of H-2Db and H-2Kb appears almost undiminished compared to wild-type cells (Fritzsche et al., 2015).

20 Introduction

1.3 Transcription of APM genes

The proper peptide loading of MHC class I molecules is a crucial process in the defense against viruses and tumors. Several proteins are involved in this process, and together they form the so-called antigen presenting machinery (APM)2. The genes of the APM are found in the MHC locus. This is located on chromosome 6 in humans and chromosome 17 in the mouse and extends over about 4 × 106 base pairs (see Figure 1-4). The genes that encode the heavy chains of MHC class I molecules and the α and β chains of MHC class II molecules are close together within the complex; the genes for β2m are outside the MHC, on different chromosomes (chromosome 15 in humans and 2 in the mouse).

Figure 1-4 Gene structure of human and murine MHC In humans, the MHC class I heavy chain genes are in the class I region, whereas the MHC class II genes as well as tapasin (TAPBP), TAP and LMP genes are in the class II region. In mice, the MHC class I heavy chain genes are also in the class I region, which - in contrast to humans - is split into two parts. It also includes the gene for tapasin. MHC class II genes, TAP and LMP genes are in the class II region. Author’s own work, based upon (Janeway et al., 2001a).

The transcription of APM genes is either at a basal or an activated level (Howcroft et al., 2003; Lee et al., 2010). Treatment with IFN-α, IFN-β or IFN-γ increases the transcription of MHC class I heavy chains, but only IFN-γ increases β2m, tapasin (Abarca-Heidemann et al., 2002; Herrmann et al., 2003), TAP (Ma et al., 1997) and ERAAP (Serwold et al., 2002), but not ERp57 (Peaper et al., 2005). In case of the proteasome,

2 APM proteins: immunoproteasome subunits LMP2, LMP5, and LMP7; TAP1, TAP2, and tapasin.

21 Introduction

IFN-specific subunits are synthesized and exchanged for the constitutive subunits, rendering the proteasome more active (the so called immuno-proteasome) (Cascio et al., 2001; Groettrup et al., 2001; Niedermann, 2002; Vigneron and Van den Eynde, 2014). The thyroid-stimulating hormone (TSH) can repress MHC class I transcription (Singer et al., 1997). TAP1 and LMP2 are regulated from a shared bidirectional promoter (Beck et al., 1992; Brucet et al., 2004; Wright et al., 1995), but their transcription is regulated differentially (Chatterjee-Kishore et al., 1998). TNF and IFN-γ stimulation together have a synergistic effect of MHC class I surface expression (Johnson and Pober, 1990).

IFN-γ stimulates the MHC class II transactivator (CIITA) in MHC class II-expressing cells, which regulates the transcription of MHC class II associated genes (Harton and Ting, 2000). NLRC5 seems to be a MHC class I transactivator, as it is also stimulated by IFN-γ and enhances transcription of MHC class I and associated genes (Biswas et al., 2012; Meissner et al., 2010).

In tumors, MHC class I transcription is often downregulated, for example by an impaired Erk5 MAP kinase pathway (Charni et al., 2009, 2010), by IGF1 overexpression (Pan et al., 2013), or by HER2 overexpression (Inoue et al., 2012). This serves, of course, the immune evasion of the tumor.

1.4 ER stress

The unfolded protein response (UPR) is an evolutionarily conserved mechanism that is activated in case of ER stress, when misfolded proteins accumulate in the ER (Hetz, 2012). The cell, in order to support its survival, normalizes cell function by downregulating translation, removing misfolded proteins, and increasing the amount of chaperones that support protein folding. Expansion of the ER membrane is also generally observed in case of ER stress (Bernales et al., 2006).

ER stress can be caused by several factors such as short supply of oxygen or glucose, high temperature, acidosis and others. As these factors fluctuate constantly, the UPR machinery needs to monitor the situation and adapt to it by a feedback mechanism.

The main sensor for ER stress is BiP (or Grp78), a chaperone that helps ER proteins fold properly. In an unstressed state, some BiP is bound to unfolded proteins, and some is bound to proteins of the different UPR signal cascades, namely PERK, IRE1, and ATF6 (see Figure 1-5), keeping them in an inactive state. In a stress situation, more BiP is needed to chaperone misfolded proteins and is then no longer available to

22 Introduction bind to the UPR proteins. These then perform different actions: PERK homodimerizes and phosphorylates elF2α, which suppresses translation of all proteins apart from the transcription factor ATF4. ATF4 controls the expression of proteins that play a role in autophagy, protein folding, and redox balance (Ameri and Harris, 2008). IRE1 also homodimerizes and splices the mRNA for XBP1, which leads to the expression of the transcription factor XBP1. This activates expression of proteins for e.g. ERAD and protein folding (Acosta-Alvear et al., 2007; Lee et al., 2003a). Finally, ATF6 is transported from the ER to the Golgi, where it is processed by S1 and S2 proteases to release a cytoplasmic fragment, which also serves as a transcription factor (Haze et al., 1999; Yamamoto et al., 2007).

Figure 1-5 Unfolded protein response in the ER BiP generally is partly bound by UPR activators PERK, IRE1 and ATF6 (top left) and keeps them in an inactive state. In case of ER stress, the accumulation of unfolded proteins forces BiP away from the UPR activators, which then switch to their active state and start the different UPR pathways, like phosphorylation of eIF2α, XBP1 splicing and cleavage of ATF6. Author’s own work, based upon (Cao and Kaufman, 2012).

23 Introduction

1.5 Viral inhibition of the immune response

1.5.1 General examples and herpesviruses

The virus and its host have contradictory interests during a viral infection. While the host would preferably clear the virus entirely, the main goal of the virus is its spreading, i.e., its transfer to other hosts. During many generations of host/virus interaction, an evolutionary process has led to the complex and fine-tuned immune system that we have, and on the other side, to innumerable viral strategies to circumvent it (Tortorella et al., 2000; van de Weijer et al., 2015). Many viruses impair antigen presentation through MHC class I molecules; this prevents infected cells from being destroyed by cytotoxic T cells. Viruses attack all stages of the MHC class I antigen presentation pathway; some examples are listed below.

Transcription. The adenovirus type 12 E1A protein reduces the expression of the MHC class I gene (Friedman and Ricciardi, 1988; Guan et al., 2008; Proffitt et al., 1994) as well as TAP and immunoproteasome subunits (Berhane et al., 2011; Proffitt and Blair, 1997; Vertegaal et al., 2003). HIV decreases MHC class I and β2m transcription (Carroll et al., 1998; Howcroft et al., 1993) and differentially regulates LMP2 (down) and LMP5/7 (up) transcription (Remoli et al., 2006), thereby modifying proteasome subunit composition. This can influence the peptide generation, leading to production of fewer immunodominant peptides and more subdominant peptides that are not recognized by cytotoxic T cells (Gavioli et al., 2004; Zanker et al., 2013). Epstein Barr virus encodes a secreted viral IL-10 ortholog that efficiently reduces the transcription of TAP1 and LMP2 (but not TAP2 or LMP7), which leads to suboptimally loaded MHC class I molecules (Yoon et al., 2012; Zeidler et al., 1997). Bovine papillomavirus both retains MHC class I intracellularly and reduces its expression (Ashrafi et al., 2002; Marchetti et al., 2009). The rhesus CMV protein Rh178 blocks translation of MHC class I molecules in a signal sequence-dependent manner, as the signal sequence of class I is conserved among allotypes and even species (Powers and Früh, 2008; Richards et al., 2011).

Peptide generation. The EBNA1 protein of the Epstein Barr virus has a glycine–alanine repeat sequence between the amino- and carboxy-terminal domains of the protein, which prevents its proteasomal degradation (Levitskaya et al., 1995). This reduces generation of virus-specific peptide, which then cannot efficiently be presented to cytotoxic T cells. The human cytomegalovirus (HCMV) structural protein pp65 impairs presentation of immediate early proteins by phosphorylation of their threonine residues (Gilbert et al., 1996). Frequent mutations in the parts of the viral genome that lead to the

24 Introduction generation of immunodominant peptides result in the so-called antigenic drift: cytotoxic T cells do not recognize their target anymore. This happens frequently in influenza virus (Nelson and Holmes, 2007) and HIV (Goudsmit et al., 1991).

Peptide transport. Generally blocking the peptide import by TAP is the easiest way to impair antigen presentation, and many members of the herpesvirus family have evolved their own strategies to achieve this (Abele and Tampé, 2006). Herpes simplex ICP47, for example, blocks peptide transport from the cytosolic side (Früh et al., 1995; Hill et al., 1995; York et al., 1994), whereas HCMV US6 works from the ER side (Ahn et al., 1997; Hengel et al., 1997; Lehner et al., 1997). Other viral proteins that impair TAP function are bovine herpesvirus UL49.5 (Koppers-Lalic et al., 2005), Epstein Barr BNLF2 (Hislop et al., 2007), and cowpox CPXV12 (Alzhanova et al., 2009).

MHC class I export. Many virally encoded proteins act directly on MHC class I molecules, with very different effects. In HCMV, for example, US2 and US11 work together to relocate MHC class I molecules into the cytosol and cause its degradation by the proteasome (Schust et al., 1998; Wiertz et al., 1996a, 1996b). The MHV-68-encoded K3 protein not only degrades MHC class I molecules, but also TAP and tapasin (Boname et al., 2004; Yu et al., 2002). In the murine cytomegalovirus (MCMV), several proteins are dedicated to manipulating MHC class I molecules: MCMV m06/gp48 binds to MHC class I molecules and leads them to lysosomal degradation (Reusch et al., 1999), while MCMV m04/gp34 binds to MHC class I molecules and escorts them to the cell surface (Kleijnen et al., 1997). This is apparently used to pacify NK cells (Babid et al., 2010), as the reduction of surface MHC class I by gp40 and gp48 might activate NK due to the missing self principle (see 1.1.2). Depending on the mouse strain, some activating NK cell receptors recognize the MHC class I/gp34 complex and react to infected cells (Pyzik et al., 2011). HCMV US3 and MCMV m152/gp40 retain MHC class I molecules intracellularly by an unidentified mechanism (Gruhler et al., 2000; Lee et al., 2003b; Ziegler et al., 1997, 2000) (see 1.7). Adenovirus E3/19K binds to MHC class I molecules and prevents their export with the help of its cytoplasmic dilysine motif, which works as an retrieval sequence (Pääbo et al., 1987). HIV Nef works from the other end of the antigen presentation pathway, as it enhances endocytosis of MHC class I molecules (Schwartz et al., 1996), while bovine papillomavirus E5 works on both transcription and export (Ashrafi et al., 2002; Marchetti et al., 2002).

IFN signaling. Treatment with IFN activates many antiviral strategies in the cell; therefore, viruses have evolved strategies to inhibit IFN related signaling at all possible levels (García-Sastre and Biron, 2006; Hengel et al., 2005). For example, paramyxovirus V proteins and hepatitis C virus NS3-4A protein inhibit

25 Introduction

IFN induction with viral dsRNA (Andrejeva et al., 2004; Meylan et al., 2005). Poxviruses sequester IFN-α and IFN-β with the help of a viral soluble IFN receptor (Symons et al., 1995). Kaposi sarcoma-associated herpesvirus LANA-1 protein blocks the IFN-β promoter (Cloutier and Flamand, 2010). Herpes simplex ICP0 impairs IRF3 mediated IFN expression (Paladino et al., 2010), and ICP27 downregulates STAT1 phosphorylation (Johnson et al., 2008). HCMV itself has a large set of viral IFN signaling modulators (Miller et al., 1999) to control the pathway at different levels.

NK cells. Cytomegaloviruses escape NK cell recognition by encoding MHC class I homologues to pacify NK cells (Revilleza et al., 2011), for example HCMV UL18 (Browne et al., 1990) and UL16 (Dunn et al., 2003) or MCMV m144 (Cretney et al., 1999). Another strategy is the selective upregulation of HLA-E by HCMV, which also silences NK cells (Tomasec et al., 2000).

Apoptosis. The organism uses apoptosis to get rid of aberrant cells which could turn out harmful; virally infected cells often apoptose. Since this inhibits viral spread, viruses have evolved different mechanism to prevent apoptosis, reviewed here (Everett and McFadden, 1999; Roulston et al., 1999). microRNA (miRNA). miRNAs are short noncoding RNA sequences that regulate gene expression by binding to certain mRNAs (the so called target mRNAs). Host miRNAs play an important role in fine- tuning the maturation of the immune system, such as lineage commitment, differentiation, proliferation, and cell migration (Xiao and Rajewsky, 2009). Viral miRNAs were first discovered in EBV (Pfeffer et al., 2004). DNA viruses often use them to increase the lifespan of infected cells, for immune evasion, and for regulating the switch to lytic infection, i.e., they are important in persistent infections (Kincaid and Sullivan, 2012). For the majority of viral miRNAs, the function has not yet been identified. Herpesviruses generate multiple miRNAs (Pfeffer et al., 2005; Umbach et al., 2008); this is especially well investigated in cytomegalovirus (Dölken et al., 2009; Grey et al., 2005). There are 14 miRNAs identified in HCMV and 21 miRNAs in MCMV. Some of them were shown to modify viral gene expression (Dölken et al., 2009), while others regulate host protein expression. HCMV-miR-UL112-1 apparently downregulates MICB expression (Stern-Ginossar et al., 2007), while HCMV-miR-UL112-3p targets TLR2 (Landais et al., 2015), and HCMV-miR-US4-1 targets ERAP1 (Kim et al., 2011). In MCMV, miR-M23-2 and miR-m21-1 knockout selectively impairs the CD4+ T cell- and NK cell-dependent viral production in salivary glands of infected mice (Dölken et al., 2010), but the precise mechanism is unknown.

26 Introduction

1.6 Cytomegalovirus

Cytomegaloviruses are species-specific beta herpesviruses and have an particular large genome, around 230 kbp for both HCMV and MCMV (Dolan et al., 2004; Rawlinson et al., 1996). Like all herpesviruses, cytomegaloviruses can switch to a latent infection and are able to reactivate the lytic cycle at suitable conditions (Sissons et al., 2002). A cytomegalovirus infection is asymptomatic in immunocompetent individuals.

1.6.1 HCMV from a medical perspective

CMV infection can lead to severe medical difficulties in immunocomprimised individuals such as newborns (Hamilton et al., 2014), people suffering from AIDS (Silva et al., 2010), and recipients of organ transplants (Bij and Speich, 2001). In the last years, the roles of CMV in the development of glioblastoma (Cobbs et al., 2002; Schuessler et al., 2014), in the phenomenon of immunosenescence (Caruso et al., 2009; Sansoni et al., 2014), and in atherosclerosis (Grahame-Clarke, 2005) have been discussed.

1.6.2 Infection and latency

The infection of a cell with CMV is a multi-step process: after infection, the DNA is uncoated and transported to the nucleus. Here, the transcription of the immediate early genes starts that control the activation of early phase genes. Early phase proteins are required for viral DNA synthesis, which is followed by the synthesis of structural proteins during the late phase of infection.

The assembly of the viral core and capsid takes place within the nucleus, and the particle is enveloped at the nuclear membrane and transported through the ER and Golgi. The release of progeny virus leads to the death of the infected cell (Mocarski Jr., 2007). After a relatively short lytic infection, latency is established (Landais and Nelson, 2013; Sinclair, 2008; Sissons et al., 2002).

27 Introduction

1.7 gp40

1.7.1 Discovery of gp40

In 1992, Koszinowski and coworkers found that an early phase protein of the murine cytomegalovirus impaired MHC class I surface expression (del Val et al., 1992). The MHC class I molecules in MCMV infected cells were retained intracellularly and were folded and β2m- and peptide-bound. Other surface proteins were unaffected by the infection. In 1997, Koszinowski and coworkers identified the responsible gene as m152 and termed the gene product gp40 (Ziegler et al., 1997). In 2002, it was realized that gp40 has an additional target besides MHC class I, namely the NK cell activating ligand RAE-1 (Krmpotid et al., 2002). In 2012, Margulies and collaborators managed to cocrystallize gp40 with RAE-1γ (Wang et al., 2012) (see Figure 1-6 and Figure 1-7).

1.7.2 Background information about gp40 gp40 is a transmembrane glycoprotein of 378 amino acids including the signal sequence (Ziegler et al., 2000). The bulk of the protein is in the ER lumen, and only a very short tail shows into the cytosol. In gel electrophoresis, gp40 appears as two glycan isoforms, one at approximately 40 kilodalton (kDa) and the other slightly lower, around 37 kDa (Ziegler et al., 1997). Its structure is MHC class I-like, as it forms an

α1/α2 superdomain with a “binding groove”. In contrast to MHC class I, the binding groove of gp40 is narrow and does not bind peptides (see Figure 1-6) (Wang et al., 2012). Also unlike in MHC class I, the very N terminus of gp40 forms a beta sheet that folds into the α3 domain (see Figure 1-6, arrow), whereas the N terminus of MHC class I begins directly in the α1/α2 superdomain. The gp40 polypeptide chain then continues with the α1/α2 superdomain and only later forms the rest of the α3 domain. In contrast to MHC class I and other class I-like proteins (including other immunoevasins in MCMV), there is no disulfide bond inside the α3 immunglobulin domain of gp40, only a single unpaired cysteine. There are two disulfide bonds formed in the protein at positions that are non-homologous to class I, one connecting C38 with C192, and another connecting C121 with C128. Removal of either of these disulfide bonds renders gp40 functionless and unable to be exported from the ER, which means that it probably is too misfolded to pass the ER quality control (Linda Janßen, data not shown). Koszinowski and coworkers showed that only the lumenal domain of gp40 is necessary for retention of MHC class I molecules and that the transmembrane domain and cytosolic tail can be exchanged with the one of CD4 or gp48, respectively, without loss of function (Ziegler et al., 2000). gp40 reduces the cell surface levels of murine

28 Introduction

MHC class I and RAE-1 (Krmpotid et al., 2002; Lodoen et al., 2003), but not of human MHC class I (Ziegler et al., 1997). gp40 directly binds to RAE-1, and this is most likely how it retains RAE-1 (Wang et al., 2012).

With respect to localization and trafficking of gp40 and class I, the Koszinowski group made the following observations: gp40 is itself located in the ER-Golgi intermediate compartment (ERGIC), and so are the retained MHC class I molecules (Ziegler et al., 1997). The retained MHC class I is folded and bound to β2m and high-affinity peptide, but it does not become EndoH-resistant even after many hours of chase, i.e., it is not exported out of the early secretory pathway (Ziegler et al., 2000). In contrast to MHC class I, gp40 itself is eventually exported (Ziegler et al., 2000). The effect of gp40 on MHC class I molecules is allotype specific (Wagner et al., 2002). gp40 cooperates with gp48 (m06) to reduce MHC class I surface expression even further in a synergistic manner (Wagner et al., 2002), while the immunoevasin gp34 (m04) does not reduce MHC class I surface levels (Wagner et al., 2002). The immunoevasins gp40, gp48 and gp34 are located on the outer ends of the linear MCMV genome (Rawlinson et al., 1996).

Figure 1-6 gp40 crystal structure A: The structure of gp40 is similar to that of MHC class I, but it does not bind to β2m. A beta sheet from the very N- terminus folds into the α3 domain (arrow). B: Like MHC class I molecules, gp40 forms an α1/α2 superdomain with a ‘binding groove’, which is visible when the structure is rotated 90°. The gp40 binding groove is too narrow to bind peptides. gp40 structure taken from the cocrystal made by (Wang et al., 2012). Drawn by the author with VMD.

29 Introduction

Figure 1-7 gp40 docked to H-2Dd gp40 was cocrystalized with RAE-1γ. Margulies and coworkers docked it onto Dd accordingly to show how it might interact with MHC class I molecules. Modified from (Wang et al., 2012). Drawn by the author with VMD.

1.7.3 Comparison of gp40 sequence of different MCMV strains

For research, usually the MCMV strains Smith and K181 are used, but the genome of several strains found in wild mice has also been sequenced. When these sequences are compared, some genes show considerable variation among the strains, but gp40 is a very conserved gene with only few amino acid changes in the sequence (see Figure 1-8) (Smith et al., 2006). The different MHC class I allotypes are very diverse proteins, and it is therefore surprising that an immunoevasin that retains them (and possibly even by binding to their most variable part, the α1/α2 superdomain) is so conserved in its sequence. It is possible that binding to MHC class I is only a useful side effect of its true role, the retention of RAE-1

(which are in their α1/α2 superdomain much more conserved than MHC class I). It is known that NK cells, which are activated by RAE-1, play a very important in combating MCMV infection (see 1.7.4.2). The question of the real target of gp40 is not easily answered, as in an MCMV infection both the virus and

30 Introduction the host use redundant strategies to attack or defend, and the function of one protein is often masked by other proteins (see also 1.7.4.1). The gp40 sequence used in this study originates from the Smith strain.

CLUSTAL 2.1 multiple sequence alignment

I signal sequence II—---- α3 ------II------α1/α2 ------K181 MLGAITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFISYREGPEMFRHSIN 60 Smith MLGAITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFISYREGPEMFRHSIN 60 G4 MLGTITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFISYREGPEMFRHSIN 60 C4A MLGAITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFISYREGPEMFRHSIN 60 N1 MLGAITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFTSYREGPEMFRHSIN 60 Wirral MLGAITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFISYREGPEMFRHSIS 60 C4C MLGAITYLLLSVLTNRGETAGSSYMDVRIFEDERVDICQDLTATFVSYREGPEMFRHSIN 60 C4D MLGAITYLLLSVLTNRGETAGSSYMDVRIFEDERVDICQDLTATFVSYREGPEMFRHSIN 60 AA18d MLGAITYLLLSVLINRGETAGSSYMDVRIFEDERVDICQDLTATFVSYREGPEMFRHSIN 60 WP15B MLGAIIYLLLSVLTNRGETAGSSYMDVRIFEDERVDICQDLTATFISYREGPEMFRHSIN 60 ***:* ******* ******************************* *************.

------+---- α1/α2 ------K181 LEQSSDIFRIEASGEVKHFPWMNVSELAQESAFFVEQERFVYEYIMNVFKAGRPVVFEYR 120 Smith LEQSSDIFRIEASGEVKHFPWMNVSELAQESAFFVEQERFVYEYIMNVFKAGRPVVFEYR 120 G4 LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYIMDVFKAGRPVVFEYR 120 C4A LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYIMNVFKAGRPVVFEYR 120 N1 LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYIMNVFKAGRPVVFEYR 120 Wirral LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYIMNVFKAGRPVVFEYR 120 C4C LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYLMNVFKAGRPIIFEYR 120 C4D LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYLMNVFKAGRPIIFEYR 120 AA18d LEQSSDIFRIEASGEVKHFPWMNVSELTQESAFFVEQERFVYEYLMNVFKAGRPIIFEYR 120 WP15B LEQSYDIFRIEASGEVKHFPWMNVSDLAQESAFFVEQEKFIYEYIMNVFKAGRPLIFEYR 120 **** ********************:*:**********:*:***:*:*******::****

------α1/α2 ------K181 CKFVPFECTVLQMMDGNTLTRYTVDKGVETLGSPPYSPDVSEDDIARYGQGSGISILRDN 180 Smith CKFVPFECTVLQMMDGNTLTRYTVDKGVETLGSPPYSPDVSEDDIARYGQGSGISILRDN 180 G4 CKFVPFECTVLQMMDGNTLTRYTVDKGVETLGSPPYSPDVSEDDIARYGRGSGISILRDN 180 C4A CKFVPFECTVLQMMDGNTLTRYTVDKGVETLGSPPYSPDVSEDDIARYGRGSGISILRDN 180 N1 CKFVPFECTVLQMMDGNTLTRYTVDKGVETLGSPPYSPDVSEDDIARYGRGSGISILRDN 180 Wirral CKFVPFECTVLQMMDGNTLTRYTVDKGVETLGSPPYSPDVSEDDIARYGRGSGISILRDN 180 C4C CKFVPFECTVLQMLDGNTLTRYTVDKGVETLGSPPYSPDVSDDDIARYGQGSGISILRDN 180 C4D CKFVPFECTVLQMLDGNTLTRYTVDKGVETLGSPPYSPDVSDDDIARYGQGSGISILRDN 180 AA18d CKFVPFECTVLQMLDGNTLTRYTVDKGVETLGSPPYSPDVSDDDIARYGQGSGISILRDN 180 WP15B CKFVPFECTVLQMLDGLTLTRYTVDKGVETLGFPPYSPDVSEDDIARYGRGSGISILRDN 180 *************:** *************** ********:*******:**********

------α1/α2 ----II------α3 ------+------K181 AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMRTQVPLAYNISLANGVDIY 240 Smith AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMRTQVPLAYNISLANGVDIY 240 G4 AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMRTQVPLAYNISLANGVDIY 240 C4A AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMHTQVPLAYNVSLANGVDIY 240 N1 AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMHTQVPLAYNVSLANGVDIY 240 Wirral AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMHTQVPLAYNISLANGVDIY 240 C4C AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMRTQVPLAYNISLANGVDIY 240 C4D AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMHTQVPLAYNISLTNGVDIY 240 AA18d AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMRTQVPLAYNISLANGVDIY 240 WP15B AALLQKRWTSFCRKIVAMDNPRHNEYSLYSNRGNGYVSCTMHTQVPLAYNVSLANGVDIY 240 *****************************************:********:**:******

31 Introduction

------+------α3 ------II—Linker-seq K181 KYMRMYSGGRLKVEAWLDLRDLNGSTDFAFVISSPTGWYATVKYSEYPQQSPGMLLSSID 300 Smith KYMRMYSGGRLKVEAWLDLRDLNGSTDFAFVISSPTGWYATVKYSEYPQQSPGMLLSSID 300 G4 KYMRMYSGGRLKVEAWLDLRDLNGSTDFAFVISSPTGWYATVKYSEYPQQSPGMLLSSID 300 C4A KYMRMYSGGRLKVEAWLDLRDLNGSTDFAFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 N1 KYMRMYSGGRLKVEAWLDLRDLNGSTDFAFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 Wirral KYMRMYSGGRLKVEAWLDLRDLNGSTDFEFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 C4C KYTRMYSGARLKVEAWLDLRDLNGSTDFEFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 C4D KYTRMYSGARLKVEAWLDLRDLNGSTDFEFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 AA18d KYTRMYSGARLKVEAWLDLRDLNGSTDFEFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 WP15B KYMRMYSGGRLKVEAWLDLRDLNGSTDFEFVISSPTGWYATVKYSEYPRQSPGMLVSSID 300 ** *****.******************* *******************:******:****

I—Linker-s Linker-seq------II—transmembrane-seq----I K181 GQFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFMYLVKLRVVWIN 360 Smith GQFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFMYLVKLRVVWIN 360 G4 GQFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFMYLVKLRVVWIN 360 C4A GKFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 N1 GKFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 Wirral GKFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 C4C GKFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 C4D GKFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 AA18d GKFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 WP15B GQFESSAVVSWHRGHGLKHAPPVSAEYSIFFMDVWSLIAIGVVFVIVFTYLVKLRVVWIN 360 *:********************************************** ***********

K181 RVWPRMRYRLVYINCRVW------378 Smith RVWPRMRYRLVYINCRVW------378 G4 RVWPRMRYRLVYINCRVW------378 C4A RVWPRMRYRLVYINCRVWGCSCDRHTGQCQIVGCSRSSRPSGHIH 405 N1 RVWPRMRYRLVYINCRVW------378 Wirral RVWPRMRYRLVYINCRVW------378 C4C RVWPRMRYRLVYINCRVW------378 C4D RVWPRMRYRLVYINCRVW------378 AA18d RVWPRMRYRLVYINCRVW------378 WP15B RVWPRMRYRLVYINCRVW------378 ******************

Figure 1-8 Sequence alignment of gp40 of different MCMV strains The different domains of gp40 are indicated. The yellow + signs highlight glycosylation sites. An * (asterisk) indicates positions which have a single, fully conserved residue. A : (colon) indicates conservation between groups of strongly similar properties - scoring > 0.5 in the Gonnet PAM 250 matrix. A . (period) indicates conservation between groups of weakly similar properties - scoring =< 0.5 in the Gonnet PAM 250 matrix. Created by the autor with clustalw (Goujon et al., 2010; Larkin et al., 2007).

1.7.4 gp40 in the viral context

1.7.4.1 T cells

The importance of CD8+ T cells in an MCMV infection is not clear; apparently the immune cells in mice play redundant roles, since mature mice control an MCMV infection even when CD8+ T cells are depleted (Jonjid et al., 1990; Polid et al., 1998). Infection of neonatal, i.e., immunocompromised, mice

32 Introduction with a gp40-deleted virus leads to increased killing of infected cells by T cells and a CD8+ T cell- dependent drop in viral titer (Krmpotic et al., 1999).

The absence or presence of immunoevasins gp34, gp40, and gp48 did not affect the size and population of virus-specific CD8+ T cells and memory T cells or viral replication and clearance kinetics in infected mice (Gold et al., 2004; Munks et al., 2007). In both studies, the virus was reactivated by immune suppression. gp40 retains MHC class I molecules also in dendritic cells; therefore one would expect that in mice infected with a gp40-deleted virus, the T cell specifity should be different because the dendritic cells present different antigens to the naive T cells. So far, there is no experimental indication for that (Gold et al., 2002). It rather seems that cross-presentation3 plays a large role in forming CD8+ T cell specificity in MCMV infection. This is supported by the fact that most MCMV specific T cells recognize antigens from early viral proteins and not immediate early proteins. gp40, which impairs antigen presentation (also in professional antigen presenting cells) is an early protein and reduces presentation of antigens of early and late viral proteins, but not immediate early proteins (Campbell et al., 1989; Del Val et al., 1989; Holtappels et al., 1998; Reddehase et al., 1984). Without cross-presentation, one would expect that most T cells react to immediate early antigens, as these are the only antigens presented on an infected cell in the early hours of infection. The cross-presentation actually seems to mislead the adaptive immune system, since it stimulates T cells against antigens that are later not presented on infected cells due to gp40 (Holtappels et al., 2004).

Why does the deletion of immunoevasins have such little effect on viral titers and CD8+ T cell killing? Probably the interaction between virus and host with their redundant attack and defense strategies is too complex to narrow down to a few proteins and is far from being understood in detail. Additionally, the experimental conditions of infecting inbred laboratory mice probably do not truly mimic the situation outside the laboratory, where the attack and defense strategies are shaped by immense evolutionary pressure. It is possible that the immunoevasins do not actually play a big role during an acute or latent infection but only in very specific situations, for example in the reactivation of the virus in the salivary glands for spreading to another host. Interestingly, infected salivary gland cells appear even more resistant to CD8+ T cell killing than other tissues (Jonjid et al., 1989). Another possibility is that the immunoevasins exist to lessen the risk of immunopathology such as tissue damage which might

3 Cross-presentation describes the phenomenon where professional antigen presenting cells take up and present exogenous antigen to activate naive cytotoxic T cells; this is for example of importance in case of an infection with a virus, which does not infect professional antigen presenting cells; without cross-presentation, the T cells could not be activated.

33 Introduction happen when infected tissue is faced by a large T cell population. CMV has a great interest in keeping its host healthy, because it usually goes into latency after the acute infection. Still, in experiments using immunoevasin deletion viruses, there was so far no indication for immunopathology (Gold et al., 2004).

1.7.4.2 RAE-1 and NK cells

The first line of defense a virus has to deal with in case of an infection is the innate immune system. The importance of the role of the innate immune system in the defense against MCMV is undisputed (Jackson et al., 2011; Tabeta et al., 2004).

NKG2D ligands such as RAE-1, MULT1, or H60 in mice are upregulated in stressed, transformed, or infected cells to signal an aberrant state to patrolling immune cells. This way, they can induce killing of the infected cell either directly or act as costimulatory receptors for T cells (Jamieson et al., 2002; Raulet, 2003). There are five RAE-1 proteins, α, β, γ, δ, and ε (Cerwenka et al., 2000). Activation of NK cells via their NKG2D ligands apparently plays an important role in an MCMV infection, as the virus uses as many as four immunoevasins to control NKG2D ligand expression. In addition to gp40, m145 reduces MULT1 expression (Krmpotic et al., 2005), m155 reduces H60 expression (Lodoen et al., 2004), m138/fcr downmodulates both MULT1 and H60 as well as RAE-1ε (Arapovid et al., 2009). gp40 reduces the cell surface expression of all forms of RAE-1 (α, β, γ and ε), except for RAE-1δ (Arapovic et al., 2009). This might have something to do with its higher stability at the cell surface, since its maturation pattern appears similarly impaired as for the other RAE-1 allotypes (Arapovic et al., 2009).

Some mouse strains, for example BALB/c, express RAE-1α, β, and γ, whereas BL/6 express RAE-1δ and ε (Lodoen et al., 2003). While BALB/c show a weak NK cell response towards MCMV infection (Krmpotic et al., 2005; Lodoen et al., 2003), BL/6 mice show a strong NK cell response. If the NKG2D ligands (RAE-1) are blocked with an antibody, which should eliminate RAE-1 dependent killing of infected cells, the viral titer rises only a little. The host can therefore control the viral infection even when the RAE-1 signaling function is impaired. This result speaks against an important role of NKG2D ligands in the defense against MCMV, and therefore against an important role of gp40 in evading NK cells. The small effect of NKG2D blocking on viral titers is probably due to another gp40-unrelated characteristic factor that is found on the NK cells in these mice, which is described in the following paragraph.

34 Introduction

Different inbred mouse strains react differently to infection with high doses of MCMV: some die quickly, but others manage to keep the viral titers under control. Especially when comparing BL/6 and BALB/c mice it was found that an MHC class I-unrelated locus determines the susceptibility of mice to MCMV. This locus was called Cmv-1 (Scalzo et al., 1990). It encodes an NK cell receptor that is different among the laboratory mouse strains. After interaction of this receptor with m157, another MCMV glycoprotein which is displayed at the cell surface of infected cells, the NK cell either kills or ignores the infected cell, depending on the mouse strain (Arase et al., 2002; Smith et al., 2002). In the resistant BL/6 mice, the NK cell receptor Ly49H recognizes m157 and activates the NK cell, whereas in the susceptible BALB/c mice, the NK cell receptor does not bind to m157, and therefore the NK cells are not activated. In 129/J mice, the NK cells bind to m157 via Ly49I, which in contrast to Ly49H is an inhibitory receptor and also does not lead to killing of the infected cell (Arase et al., 2002). m157 probably evolved as a class I decoy for NK cells, since it structurally looks much like MHC class I and can pacify certain NK cells. BL/6 mice use this strategy against the virus; but since the m157 sequence is very variable in different MCMV strains found in wild mice, it is to be expected that there are MCMV strains that are not recognized by the Ly49H receptor of BL/6 (Smith et al., 2006).

1.7.5 The gp40 mystery

Why should we investigate gp40? From a medical point of view, it seems uninteresting. While HCMV is a medically relevant virus (see 1.6.1), there is no gp40 ortholog in HCMV, and gp40 is apparently dispensable for a successful infection even in mice (see 1.7.4.1). So, from a medical perspective, gp40 appears not very interesting indeed.

From a cell biological point of view, this is very different. What is known about gp40 function so far is mysterious: It is apparently degraded very rapidly, it could not be shown to bind to MHC class I, but somehow it retains MHC class I in the ERGIC, while other secretory proteins are exported normally (Ziegler et al., 2000). gp40 does not have a known retention/retrieval signal in its sequence, and with its MHC class I-like structure it looks very different from other viral immunoevasins that affect MHC class I export. Usually, MHC class I-like viral proteins are used to pacify NK cells, whereas immunoevasins that impair MHC class I export have an MHC class I-unrelated structure, which enables them to directly bind to MHC class I molecules or another protein involved in antigen presentation (McCoy et al., 2012).

35 Introduction

How does gp40 work? Does it work directly or indirectly on MHC class I? The data that show that MHC class I remains in the ERGIC, even after all visible gp40 is degraded, speak against a mechanism that involves direct interaction between gp40 and MHC class I (Ziegler et al., 2000). On the other hand, the fact that gp40 only retains murine class I molecules but not human class I molecules, regardless of whether they are expressed in a murine or human cell line, speaks for a function that is based on an interaction.

And what is the molecular mechanism of the retention? Does gp40 hijack the cellular or some MHC class I-specific quality control in order to retain MHC class I? Compared to cellular quality control mechanisms, gp40 is actually far more efficient in retaining MHC class I molecules. This can be observed in Figure 3-2C, where gp40 reduces the anyways comparably low MHC class I surface levels in PLC knockout cells even further.

Does gp40 perhaps covalently modify MHC class I molecules themselves? So far, there is no indication for this, since the retained MHC class I molecules bind β2m and high affinity peptide and are recognized by conformation specific antibodies. Also, on SDS-PAGE, they run at the same apparent molecular weight as MHC class I molecules in control cells (after removing the glycans, since MHC class I molecules in gp40-expressing cells do not become EndoH resistant). And from its structure, gp40 does not look like an enzyme that could actively modify its target.

Whatever the secret of gp40 is, it must be an exciting one, since it might help us discover completely new retention or retrieval factors, localization sequences, or generally additional players in the network of the secretory pathway.

1.8 The aim of this study

The Springer group has worked on MHC class I export and quality control for around ten years and has a lot of expertise, reagents, and methods to investigate MHC class I maturation. We planned to use these resources to solve unanswered questions about gp40 and possibly understand more about its mechanism.

One of the possible mechanisms that came to our mind was the hijacking of the cellular quality control by gp40. We used PLC knockout cell lines to try to find out if gp40 is still functional in these cells.

36 Introduction

Additionally, we wanted to learn more about the MHC class I allotype specificity of gp40. For this approach, we used MHC class I domain swaps, where the domains of different allotypes are exchanged to understand which domain is decisive for the observed effect.

We investigated a possible direct interaction between gp40 and its target MHC class I, as well as interactions with possible unknown retrieval/retention factors. In the future, we plan to use immunoprecipitation, crosslinking, and MALDI-TOF for this purpose.

37 Materials and Methods

2 Materials and Methods

2.1 Constructs

The gp40 gene was amplified out of the baculovirus construct based on the MCMV strain Smith (Messerle et al., 1997), a gift from Prof. Hartmut Hengel. The primers used for amplification-introduced restriction sites and affinity tags or localization tags to the construct are shown in Table 2-1. The sequence of the HA-tagged gp40 construct is shown in Figure 2-1.

The polymerase chain reaction (PCR) products were first blunt-end ligated into the pJET2.1 plasmid (Thermo Fisher), excised with XhoI and BamHI, and ligated into the lentiviral transfer vectors pucCl6IEGwo (see Figure 2-2) and puc2Cl6IPwo (see Figure 2-3). With the oLJKKSQstAgeBam reverse primer, a KKSQ-tag was introduced between the gp40 sequence and the HA-tag (see Figure 2-4). The gp40-GFP fusion construct was created by removing the STOP codon with the help of the oLJgp40/GFP reverse primer, reintroducing BamHI and ligating into the GFP-fusion puc2Cl6IPwo plasmid (see Figure 2-5). All lentiviral transfer vectors were gifts from Helmut Hanenberg.

Table 2-1 Primers used for cloning of the gp40 gene The underlined sequence identify the restriction site, the gray sequence aligns with gp40, the pink sequence codes for the HA tag, the light blue sequence codes for the KKSQ tag, the red sequences indicate STOP codons.

Used for Restriction Primer ID construction Orientation 5’ to 3’ sequence site of CC CTC GAG ATG CTG GGC GCT ATC oLJ 1F gp40-HA4 XhoI forward AC GG GGA TCC TCA AGC ATA ATC AGG GAC ATC ATA AGG ATA GGC GTA oLJ 1R gp40-HA BamHI reverse GTC GGG CAC GTC GTA GGG GTAGTT GAT GTA GAC CAG GCG gp40-KKSQ- BamHI, GG GGA TCCACC GGTTCACTG TGA oLJKKSQstAgeBam reverse HA5 AgeI CTT CTTAGCATA ATC AGG G oLJgp40/GFP R CCC GGA TCCGTT GAT GTA GAC CAG gp40-GFP6 BamHI reverse GCG

4 gp40-HA construct used throughout the study 5 gp40-KKSQ-HA used in Figure 3-9 6 gp40-GFP used in Figure 6-3

38 Materials and Methods

1 atgctgggcgctatcacctacttgctcctctcggttctcataaaccgaggcgagacggcg 60 1 M L G A I T Y L L L S V L I N R G E T A 20 61 ggcagcagctatatggacgtgcgcatattcgaggatgagcgggtggacatctgtcaagac 120 21 G S S Y M D V R I F E D E R V D I C Q D 40 121 ctgacggcgacgttcatctcgtacagagaaggtccggagatgttccgccacagtatcaat 180 41 L T A T F I S Y R E G P E M F R H S I N 60 181 ctagagcagtcgtctgatatctttcggatcgaagcctccggagaggtgaaacattttcct 240 61 L E Q S S D I F R I E A S G E V K H F P 80 241 tggatgaacgtgagcgagctggcgcaggagagtgcgttcttcgtggagcaggagaggttc 300 81 W M N V S E L A Q E S A F F V E Q E R F 100 301 gtatacgagtacattatgaatgtcttcaaagccggacggccggtagtcttcgaatataga 360 101 V Y E Y I M N V F K A G R P V V F E Y R 120 361 tgcaagttcgttccattcgaatgtaccgtacttcagatgatggacggcaatacgttgaca 420 121 C K F V P F E C T V L Q M M D G N T L T 140 421 cgttacaccgtagacaaaggcgtcgaaacgctcgggtctccgccgtactctcccgacgta 480 141 R Y T V D K G V E T L G S P P Y S P D V 160 481 tccgaggatgacatcgcgcgctacggacaagggtccggaatctctatcttgagggacaac 540 161 S E D D I A R Y G Q G S G I S I L R D N 180 541 gctgctctactccagaaacgctggacgtccttctgtcggaagatcgtcgccatggacaac 600 181 A A L L Q K R W T S F C R K I V A M D N 200 601 cccagacacaacgaatactcgctgtacagtaatcgaggcaacggctacgtgtcctgtacg 660 201 P R H N E Y S L Y S N R G N G Y V S C T 220 661 atgcgcactcaggttccgttggcgtacaacatcagtctcgcgaacggagtggacatctac 720 221 M R T Q V P L A Y N I S L A N G V D I Y 240 721 aagtacatgcgcatgtattctggtggacgattgaaggtggaagcgtggctcgatctcaga 780 241 K Y M R M Y S G G R L K V E A W L D L R 260 781 gacctgaacggtagtaccgacttcgcgttcgtgatttcttccccgacgggatggtacgct 840 261 D L N G S T D F A F V I S S P T G W Y A 280 841 acggtcaagtattctgagtaccctcaacagagtcccggcatgctgttgtcgtcgatcgat 900 281 T V K Y S E Y P Q Q S P G M L L S S I D 300 901 gggcagttcgagtcgtccgcggtcgtctcgtggcacaggggacacggtctcaagcacgct 960 301 G Q F E S S A V V S W H R G H G L K H A 320 961 cctcccgtctccgcggagtactccatcttcttcatggacgtgtggtccttgatcgcgatc 1020 321 P P V S A E Y S I F F M D V W S L I A I 340 1021 ggagtcgtgttcgtgatcgtcttcatgtatctggtgaagttacgggtggtgtggatcaat 1080 341 G V V F V I V F M Y L V K L R V V W I N 360 1081 cgggtctggcctcgtatgcggtatcgcctggtctacatcaacTACCCCTACGACGTGCCC 1140 361 R V W P R M R Y R L V Y I N Y P Y D V P 380 1141 GACTACGCCTATCCTTATGATGTCCCTGATTATGCTTGA 1179 381 D Y A Y P Y D V P D Y A * 392

Figure 2-1 Sequence of the gp40-HA construct used The HA tag is shaded in pink, the stop codon is shaded in red, and the linker sequence is shaded in green. The linker sequence was either deleted (gp40-LDel) or exchanged with a (G4S)9 sequence (gp40-(G4S)9). These constructs were made by V. Ramnarayan and used in Figure 3-5, Figure 3-11, and Figure 3-12. Drawn by the author.

39 Materials and Methods

Figure 2-2 gp40-HA in puc2Cl6IEGwo Expression of this construct can be verified by GFP expression. Drawn by the author with ApE.

Figure 2-3 gp40-HA in puc2Cl6IPwo Cells expressing this construct can be selected with puromycin. Drawn by the author with ApE.

40 Materials and Methods

Figure 2-4 gp40-KKSQ-HA in puc2Cl6IPwo This construct was also cloned into puc2Cl6IEGwo with the same restriction sites (not shown). Drawn by the author with ApE.

41 Materials and Methods

Figure 2-5 gp40-GFP fusion in puc2Cl6IPwo This construct creates a gp40-GFP fusion protein without an HA-tag. Expressing cells can be selected with puromycin. Drawn by the author with ApE.

Figure 2-6 hβ2m in puc2Cl6INwo This construct expresses hβ2m, expressing cells can be selected with neomycin. Drawn by the author with ApE.

42 Materials and Methods

2.2 Domain swaps

The cloning of the Db and Kb domain swaps was described in detail (Fritzsche, 2014). A schematic depiction shows the swapped domains (see Figure 2-7). The MHC class I constructs, which all had a STOP codon at the end of their open reading frame, were ligated into the pEGFP-N1 expression plasmid, which was then transfected into L cells (see Table 2-2). L cells expressing the constructs were enriched by flow cytometry after staining of surface MHC class I (since the cells did not express GFP).

Figure 2-7 Kb and Db domain swap constructs The nomenclature of the constructs is K for Kb sequences, D for Db sequences. Denoted are the signal sequences with a small letter, the lumenal domain (α1, α2, α3), the transmembrane domain and the cytosolic tail with a capital letter, respectively. All constructs have the signal sequence for Kb (SS), to ensure similar ER import. The lumenal domains of MHC (α1, α2, α3) are swapped as one domain, the transmembrane domain (TMD) and cytosolic tail (CT) also. The kKKK (Y256N) construct has an additional glycosylation site at position 256, as Db has (orange circles = glycosylation sites). The color coding is green for Kb, red for Db. Drawn by the author.

43 Materials and Methods

2.3 Cell culture

Mammalian cells (see Table 2-2) were grown on 10 cm or 15 cm dishes at 37 °C in a 5% CO2 atmosphere in DMEM (PAA, GE Healthcare) supplemented with 10% fetal calf serum (FCS) (Biochrom AG), 2 mM L- Glutamine, 100 U/mL penicillin, and 100 g/mL streptomycin (PAA, GE Healthcare). Every two days, cells were subcultured by washing them once in phosphate buffered saline (PBS) and detaching them with 1x Trypsin-EDTA (PAA, GE Healthcare) at 37 °C for 5 min. After detaching, cells were split according to their growth rate.

Table 2-2 Mammalian cell lines List of mammalian cell lines and their characteristics used in this work.

Name Species Type Characteristics MHC class I Source Tim Elliott (Southampton, UK) K41 m fibroblast wild type H-2b (Nakamura et al., 2001) Calreticulin Tim Elliott (Southampton, UK) K42 m fibroblast H-2b deficient (Nakamura et al., 2001) MEF m fibroblast TAP1 deficient H-2b Tim Elliott (Southampton, UK) TAPd MEF Tapasin- Luc van Kaer (Nashville, USA) m fibroblast H-2b TPNd deficient (Grandea et al., 2000) Jody Gronendyk (Edmonton, MEF Calnexin- m fibroblast H-2b Canada) CNXd deficient (Kraus et al., 2010) MEF m fibroblast wild type H-2b see above CNX wt Tatiana Soldà (Bellinzona, MEF m fibroblast UGT1-deficient H-2b Switzerland) UGT1d (Soldà et al., 2007) Frank Momburg (Heidelberg, L cells m fibroblast wild type H-2k Germany)

B Alain Townsend (Oxford, UK), RMA m wild type H-2b lymphocyte (Kärre et al., 1986) A*02/02, B*07/07, ATCC CRL-3216 293T h fibroblast wild type C*07/07

(Vogel et al., 2013)

44 Materials and Methods

2.4 Protein expression in mammalian cells

2.4.1 Lentiviral transduction

Lentiviral transduction uses the ability of retroviruses to integrate their genome into the host genome. In the process of transduction, the gene of interest is integrated into the genome of the transduced cell, and therefore, the expression is not lost after several cell divisions. The transduction event is rather gentle for the cells, compared to the usual transfection strategies, and works also for hard-to-transfect cells.

The lentiviral system that we used consists of three different plasmids: the transfer plasmid (for example puc2CL6IEGwo (see Figure 2-2), which bears the gene of interest, the envelope plasmid, which encodes an envelope protein, and the helping plasmid, which encodes viral structural proteins and enzymes as the reverse transcriptase and the integrase. These three plasmids are cotransfected into packaging cells (usually cells containing the large T antigen such as HEK293T cells, since this will increase the yield) (Figure 2-8). The packaging cells transcribe the transfer plasmid into an RNA strand, and produce the proteins that are encoded on the packaging and envelope plasmid. The proteins build the lentiviral particles, which incorporate the RNA strand containing the gene of interest. The viral particles are used to infect the target cells, where the RNA is retrotranscribed into DNA by the reverse transcriptase and integrated into the host genome with the help of the integrase. These viral particles are replication- incompetent, since the DNA that is integrated into the host genome only encodes the gene of interest, but not any viral proteins. The infected cell therefore cannot produce new viral particles.

The transfer plasmid carries the gene of interest flanked with long terminal repeats (LTRs) as well as other factors which are necessary for the successful export of the RNA out of the nucleus, its incorporation into the viral particle, its retrotranscription, and for the integration of the gene of interest into the host genome. On the transfer plasmid, the only protein coding sequence is the gene of interest (and the selection markers such as GFP or puromycin), all other sequences are cis-acting elements.

The envelope plasmid carries the gene for VSV-G, the vesicular stomatitis virus glycoprotein. The use of this envelope plasmid renders the viral particle amphotropic, which means it can bind to and fuse with virtually every cell type, in contrast to ecotropic envelope plasmids, which restrict the viral particle for the infection of cells of a certain species. VSV-G-coated particles are classified as biosafety level S2 in Germany.

45 Materials and Methods

Packaging cells

Retroviral particles

Target cells

Figure 2-8 Principle of lentiviral transduction The three plasmid, transfer plasmid, helper plasmid, and envelope plasmid are cotransfected into the packaging cells (orange). The packaging cells produce replication-defective retroviral particles, which are used to infect the target cells (blue). Drawn by the author.

2.4.2 Preparation of lentiviral particles

The packaging cells (HEK293T cells), which were in an exponential growth phase, were seeded onto a 10 cm dish the previous evening at approximately 30% confluency. At the day of the transfection, the cells were around 70% confluent.

The transfection medium was prepared as follows:

Solution A: 955 µL DMEM without FCS (warmed to 37 °C) were pipetted into a 1.5 mL tube. 45 µL Polyethylenimine (PEI) (1 mg/mL; Sigma Aldrich, No. 40,872-7) were added and mixed well; the solution was incubated for 5 minutes (min) at room temperature.

46 Materials and Methods

Solution B: 1,000 µL DMEM without FCS (warmed to 37 °C) were pipetted into a 2 mL tube. Six µg each of the transfer plasmid, the helper plasmid, and the envelope plasmid were added and mixed well; the solution was incubated for 5 min at room temperature.

Transfection medium: Solution A was added to solution B and mixed well; the solution was incubated at RT for 30 min.

The old medium on the packaging cells was removed, and 5 mL of fresh medium was added. The transfection medium was added onto the cells (the cells were now considered to be biosafety level S2) and incubated for 12 hours. The old medium was discarded, and 10 mL of fresh medium was added. The cells were incubated for 24 hours, and the supernatant (which contains the lentiviral particles) was collected. The supernatant was filtered through a 0.45 µM filter and stored at 4 °C, where the viral supernatant was stable for many weeks.

The viral titer depended much on the quality of the plasmids used and on the health and density of the packaging cells.

2.4.3 Transduction of mammalian cells

The transduction of cells was done in an S2-designated laboratory, since the viral supernatant was classified as S2. The target cells, which were in an exponential growth phase, were seeded onto a 6-well- plate. At the time of transduction they had a low confluency (around 15%), since this increased transduction efficiency.

The old medium was removed, and the undiluted viral supernatant was added to the cells and incubated for 24 hours. The next day, the viral supernatant was removed, and the cells were provided with fresh medium and split if necessary. The cells were kept in the S2 laboratory for three days after removal of the viral supernatant and were transferred onto a fresh dish before declaring them S1.

Usually, the transduction marker (for example GFP) was detectable three to four days after transduction, depending on the doubling time of the transduced cells.

47 Materials and Methods

2.4.4 Selection of transduced mammalian cells with puromycin

Puromycin is a good selection marker, since it is a fast-working antibiotic and separates very well the expressing from the non-expressing cells. The correct puromycin concentration needed for selection was determined with a killing curve. For this, cells were seeded into a 6-well-plate at a confluency of around 15-20%, and puromycin was added at different concentrations. At the correct concentration (for fibroblasts usually 5 µg/mL), around 50% of the cells were dead after 24 hours, and after another 24 hours, almost all cells should be dead. With slow-growing cells it took longer, but after 5 days at the latest, all cells were dead.

The so-determined correct concentration of puromycin was then added to the transduced cells. The majority of the cells transduced with the empty control plasmid usually survive (50-70%, depending on the viral titer), but the majority of the cells transduced with an insert containing lentiviral plasmid die (around 90-95%). This is probably due to the IRES site: if an ORF is present before the IRES, the ORF after the IRES (in this case, the puromycin resistance) is transcribed less efficiently.

Cells should be kept under puromycin for two days, and then kept in normal medium to recover their confluency. After recovery, the puromycin treatment was repeated. After two weeks, the puromycin treatment was repeated once more to remove the last untransduced cells.

2.4.5 Tunicamycin treatment

Tunicamycin (Sigma Aldrich, T7765) blocks protein glycosylation in the ER, which results in ER stress (Hung et al., 2004). The tunicamycin was dissolved in DMSO at a concentration of 1 mg/mL, aliquoted, and stored at -20 °C.

To cause ER stress in mammalian cells, cells were grown on a 10 cm dish to a confluency of around 60%. Tunicamycin was added in a final concentration of 2 μg/mL for 12 hours. Afterwards, cells were processed for mRNA isolation (see 2.7.2).

48 Materials and Methods

2.4.6 IFN-α and IFN-γ treatment

IFN-α (Boehringer Ingelheim) was used in a final concentration of 400 U/mL and IFN-γ (Peprotech, No. 315-05) was used in a final concentration of 200 U/mL for 48 hours. Afterwards, cells were either harvested for flow cytometry (see 2.5.2) or mRNA isolation (see 2.7.2).

49 Materials and Methods

2.5 Immunostaining

2.5.1 Immunofluorescence for microscopy

For immunofluorescence microscopy, cells were seeded on 10 mm round glass cover slips, so that they reached a confluency of around 25% on the day they were used. Cells were washed once in PBS and fixed with 3% PFA in PBS for 15 min at room temperature. The PFA was removed, and cells were stored in PBS at 4 °C until further use.

Cells were permeabilized in 0.1% Triton X-100 for 5 min at room temperature. Afterwards, the Triton buffer was removed, and cells were incubated with 50 µl of the primary antibody (see Table 2-3) per cover slip for 1 hour at room temperature. Then, the antibody was removed and the cells were washed three times in 100 µl PBS. 50 µl of the secondary antibody (Cy2, Cy3 or Cy5 against murine or rabbit, Abcam) was added and incubated for 30 min at room temperature in the dark. Afterwards, the antibody solution was removed, and the cells were washed three times in 100 µl PBS. The cover slip was briefly rinsed in ddH2O and glued cell side down onto an object slide with 7 µl of MOWIOL (Fluka Chemica) heated to 50 °C.

The MOWIOL was allowed to harden for at least overnight in the dark, and microscopy slides were observed with a Zeiss LSM 510 confocal microscope.

Table 2-3 Primary antibodies for microscopy List of primary antibodies and antisera that were used for microscopy staining.

Name Target Host Working Dilution Source/reference Ted T. Hansen B22.249 Db, β m- mouse, undiluted, hybridoma 2 (Washington, USA)/(Lemke et (B22) bound monoclonal supernatant al., 1979) Tim Elliott Kb, β m- mouse, undiluted, hybridoma Y3 2 (Southampton, bound monoclonal supernatant UK)/(Hämmerling et al., 1982) HA (clone mouse, 1:20 in PBS, hybridoma HA (Niman et al., 1983) 12CA5) monoclonal supernatant mouse, BD transduction Laboratories, GM130 cis-Golgi 1:300 in PBS monoclonal No. 610822 rabbit, Calnexin Calnexin 1:300 in PBS D. Williams (Toronto, Canada) polyclonal mouse, HA-FITC HA 1:60 in PBS Roche, No 11988506001 monoclonal

50 Materials and Methods

2.5.2 Immunofluorescence for FACS

Cell were harvested at a subconfluent state, washed once in PBS and put on ice. Cells were counted and pipetted into a 96-well-plate (V-bottom) with 106 cells/well. Cells were pelleted at 400 x g for 5 min at 4 °C, and the supernatant was aspirated. Cells were resuspended in 50 µl primary antibody solution/well (see Table 2-4) with precut tips to reduce cell damage, and incubated on ice for 30 min. Afterwards, cells were pelleted at 400 x g for 5 min at 4 °C, and the antibody was aspirated. Cells were washed three times in 100 µl ice-cold PBS/well. Cells were then resuspended in 50 µl secondary antibody/well (either Alexa488 or APC, Dianova) and incubated in the dark for 30 min. Cells were washed three times in ice- cold buffer, and were afterwards kept on ice. Just before counting, cells were resuspended in 1 ml PBS and counted with a CyFlow space (Partec, Görlitz, Germany).

Table 2-4 Primary antibodies for flow cytometry

Name Target Host Working Dilution Source/reference B22.249 mouse, undiluted, hybridoma Db, β m-bound see Table 2-3 (B22) 2 monoclonal supernatant mouse, undiluted, hybridoma Y3 Kb, β m-bound see Table 2-3 2 monoclonal supernatant β m-bound Alain Townsend 2 mouse, undiluted, hybridoma W6/32 human MHC (Oxford, UK)/(Barnstable monoclonal supernatant class I et al., 1978)

51 Materials and Methods

2.6 Biochemical experiments

2.6.1 Radioactive pulse-label

Cells, which were in an exponential growth phase, were trypsinized, washed twice in PBS, and resuspended in starvation medium (modified Met-free RPMI1640, Sigma; supplemented with 2% FCS, 2 mM L-Glutamine, 100 U/mL penicillin, and 100 g/mL streptomycin) at a density of around 106 cells/1 ml starvation medium. Cells were starved at 37 °C for twenty minutes. Afterwards, cells were pelleted and resuspended in 106 cells/500 µl labeling medium (starvation medium containing 50-100 µCi 35S methionine/cysteine (PerkinElmer) per mL). Cells were labeled for the indicated time at 37 °C, and the labeling process was stopped by placing the tubes on ice. For co- and reimmunoprecipitations, more 35S methionine/cysteine was used than for direct immunoprecipitations.

2.6.2 Pulse-chase

Cells were pulse-labeled as described above. Afterwards, cells were pelleted, the radioactive medium was removed, and cells were resuspended in the excess chase medium (cell culture medium containing additional 500 µg/mL cysteine and 100 µg/mL methionine). Cells were incubated at 37 °C for the indicated chase time. At each time point, a sample was taken and put on ice to stop the chase.

2.6.3 Immunoprecipitation

Antibodies used for immunoprecipitation (IP) were prebound to protein A agarose beads (Calbiochem). For a direct IP, generally, 20 µl of beads and either 200 µl hybridoma supernatant or 2 µl antiserum were used, but for a co-immunoprecipitation (coIP), the amounts were doubled. When using antiserum, the mixture was filled up with PBS so that the beads could move freely. The beads and the antibodies were allowed to rotate in the cold room for at least 3 hours. Afterwards, the beads were washed three times in cold PBS before use and split onto 1.5 mL tubes according to the number of samples.

Cells were washed once in ice-cold PBS and lysed in native lysis buffer (approx. 200 µl buffer/106 cells; 50 mM Tris pH 7.5, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100 supplemented with protease inhibitors (5 mM iodoacetic acid (IAA), 1 mM phenylmethylsulfonylfluoride (PMSF)) for 30 min at 4 °C rotating.

52 Materials and Methods

After cell lysis was complete, nuclei and cell debris were pelleted at 16,000 x g for 10 min at 4 °C. The supernatant was added to the prebound protein A agarose beads and incubated at 4 °C rotating for 1 hour. Afterwards, beads were washed three times in 1 mL Triton wash buffer/sample (50 mM Tris pH 7.5, 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100). The agarose beads were dried with a 22 gauge needle and boiled in 20 µl 1% SDS solution containing 40 mM DTT for 10 min at 95 °C. Beads were either stored at -20 °C or directly processed further.

Table 2-5 Antibodies used for immunoprecipitation List of antibodies and antisera used for immunoprecipitation

Name Target Host Source/reference mouse, Y3 Kb, β m-bound see Table 2-3 2 monoclonal mouse, B22.249 (B22) Db, β m-bound see Table 2-3 2 monoclonal mouse, HA HA see Table 2-3 monoclonal Ted T. Hansen binds β m-bound and β m-free mouse, 28-14-8s 2 2 (Washington, USA ) / (Ozato forms of Ld, Lq, Db, Dq monoclonal et al., 1980) rabbit, Db/Kb serum cytosolic peptides of Db and Kb see (Fritzsche, 2014) polyclonal rabbit, (Yu et al., 1999)/ Tapasin serum N-terminus of mouse tapasin polyclonal Dr. Xiaoli Wang PA3-900 rabbit, Calreticulin from several species Pierce Antibodies (calreticulin) polyclonal rabbit, Calnexin serum Calnexin from several species see Table 2-3 polyclonal

2.6.4 Co-immunoprecipitation

CoIP uses a lysis buffer with a milder detergent than normal IP to keep protein-protein interactions intact during lysis and IP.

Protein A agarose beads were prepared as in (2.6.3). Cells were washed once in ice-cold PBS and lysed in

6 digitonin lysis buffer (approx. 100 µl buffer/10 cells; 140 mM NaCl, 20 mM Tris pH 7.6, 5 mM MgCl2, and 1% digitonin supplemented with protease inhibitors (5 mM IAA, 1 mM PMSF)) for 30 min at 4 °C rotating.

53 Materials and Methods

The digitonin buffer was prepared freshly every time. For this, the digitonin powder was put into the buffer solution and left at room temperature for 10 min to dissolve, afterwards boiled in the microwave three times and left to cool down to room temperature. Digitonin buffer was prone to precipitate, therefore the buffer was kept at room temperature and cooled down on ice only just before use.

After cell lysis was complete, nuclei and cell debris were pelleted at 16,000 x g for 10 min at 4 °C. The supernatant was added to the prebound protein A agarose beads and incubated at 4 °C rotating for 1 hour. Afterwards, beads were washed in 500 µl digitonin wash buffer (once in buffer B (10 mM Tris pH 7.5, 150 mM NaCl, 2 mM EDTA, 0.2% digitonin), twice in buffer C (10 mM Tris pH 7.5, 500 mM NaCl, 2 mM EDTA, 0.2% digitonin) and once in buffer D (10 mM Tris pH 7.5)). The agarose beads were dried with a 22 gauge needle and boiled in 20 µl 1% SDS solution containing 40 mM DTT for 10 min at 95 °C. Beads were either stored at -20 °C or directly processed further.

2.6.5 Re-immunoprecipitation

Re-immunoprecipitation (reIP) is identical to coIP until after the beads were dried with a 22 gauge needle. The beads then were boiled in 1% SDS with 2 mM DTT for 10 min at 95 °C. After cooling down, the supernatant was transferred to a fresh tube and diluted with 1 mL of PBS containing 0.1% Triton X- 100. The tubes were centrifuged at 1,000 x g and 4 °C for a couple of minutes to pellet beads that were accidentally transferred, and 900 µl of the supernatant were pipetted onto the prebound agarose beads (prepared as in (2.6.3)) and incubated at 4 °C rotating for at least one hour. Beads were then washed twice in 1 mL ice-cold PBS 0.1% Triton X-100 and dried with a 22 gauge needle. Beads were boiled in 20 µl 1% SDS containing 40 mM DTT and either stored at -20 °C or directly processed further.

2.6.6 EndoF1 digest

Endoglycosidase F1 (EndoF1) cleaves glycans on glycoproteins until the enzyme Golgi alpha- mannosidase II removes two mannose subunits. All later oligosaccharide structures are resistant to EndoF1, and therefore it is commonly used to investigate the maturation state of a glycoprotein (Maley et al., 1989). EndoF1 and the more commonly used endoglycosidase H (EndoH) have the same substrate specificity, they hydrolyze only high mannose glycans (Trimble and Tarentino, 1991). Upon treatment

54 Materials and Methods with brefeldin A (BFA), the cis-Golgi and its enzymes relocate to the ER, and all glycoproteins become EndoF1 resistant, irrespective of their maturation status (Pagny et al., 2000).

4 µl of 10x G5 buffer (0.5 M Sodium Citrate (pH 5.5 at 25 °C)) and 6 µl of 20% Triton X-100 were added to samples prepared as in 2.6.3, 2.6.4 and 2.6.5, mixed, and centrifuged briefly. To this mixture, 1.5 µl of EndoF1 enzyme (described before (Fritzsche, 2014)) was added and mixed well. Samples were digested overnight at 37 °C. The following day, 10 µl of 6x LSB (350 mM Tris pH 6.8, 36% glycerol, 10% SDS, 0.6 M DTT, 0.012% bromphenol blue) was added to each sample, and the samples were boiled at 95 °C for 5 min.

2.6.7 PNGase digest

Peptide-N-Glycosidase (PNGase) is an amidase that cleaves between the innermost N-acetyl- glucosamine and the asparagine residues of N-linked glycans (Maley et al., 1989). The reaction is independent of the maturation state of the glycoprotein, in contrast to EndoF1 (see 2.6.6).

4 µl of 10x G7 buffer (500 mM Sodium Phosphate pH 7.5) and 6 µl of 20% Triton X-100 were added to samples prepared as in 2.6.3, 2.6.4 and 2.6.5, mixed, and centrifuged briefly. To this mixture, 1.5 µl of homemade PNGase enzyme was added and mixed well. Samples were digested over night at 37 °C. The following day, 10 µl of 6x LSB was added to each sample and the samples were boiled at 95 °C for 5 min.

2.6.8 Folding of MHC class I

Cells were labeled for 5 min as in (2.6.1) and chased for the indicated times as in (2.6.2). Immunoprecipitation was done as in (2.6.3) using the conformation-specific antibodies Y3 and B22, which only bind to folded MHC class I molecules. Samples were PNGase digested and separated on an SDS-PAGE gel.

2.6.9 Thermostability of MHC class I

The stability of the folded state of MHC class I molecules is dependent on the quality (i.e., binding affinity) of the peptide that is bound into the binding site. The thermostability method makes visible the

55 Materials and Methods differences in the stability of MHC class I against heat-induced denaturation under different conditions (Williams et al., 2002); this allows a judgement of the average affinity of the peptide pool bound to a radiolabeled cohort of class I molecule at the time of cell lysis.

For this experiment, cells were radioactively labeled and lysed. Afterwards, nuclei and cell debris were pelleted by centrifugation at 16,000 x g at 4 °C for 10 min. The supernatant was heated to the indicated temperatures for 10 min and then returned to 4 °C. Lysates were then pre-cleared with protein A for 30 min, and MHC class I molecules were precipitated with the conformation-specific antibodies Y3 and B22. Samples were PNGase digested and separated on an SDS-PAGE gel.

2.6.10 Surface-immunoprecipitation using SL8-biotin

Cells were labeled for 5 min and chased for 120 min in 300 µl chase medium/sample containing 5 µM biotinylated SL8 peptide (SL8 = SIINFEKL, a Kb specific peptide from ovalbumin (357-64); biotin was attached to the lysine side chain by a 6-aminohexanoic acid linker to allow simultaneous binding of the peptide to Kb and to streptavidin). After the chase, cells were washed twice in ice-cold PBS and lysed in 500 mL lysis buffer/sample containing 10 µM unconjugated SL8 peptide to prevent post-lysis binding. Lysates were centrifuged at 16,000 x g for 5 min to remove nuclei and cell debris, precleared with protein A for 30 min, and immunoprecipitated with 20 µl streptavidin agarose beads (Thermo Fisher, No. 29201). Samples were then EndoF1 or PGNase digested and separated on an SDS-PAGE gel.

2.6.11 SDS-PAGE

Protein samples (as generated in the previous chapters, see 2.6.1 to 2.6.10) were separated on a discontinuous SDS-PAGE gel system with a 5% acrylamide stacking gel and an 11% acrylamide resolving gel. The electrophoresis was performed at 25 mA for 1.5 hours in 250 mM Tris, 2 M glycine, and 1% SDS. As a size reference, a 14C-labelled molecular weight marker (CFA626; Amersham) was used.

2.6.12 Autoradiographic analysis

After separation on an SDS-PAGE gel, the radiolabeled proteins were transferred onto a polyvinylidene fluoride (PVDF) membrane for 2 hours at 60 V (or overnight at 14 V) in a TE22 tank (GE Healthcare) in

56 Materials and Methods transfer buffer (192 mM glycine, 25 mM Tris, and 15% methanol). The dry membrane was then exposed for up to seven days to an imaging plate and scanned with the Fluorescent Image Analyzer FLA 3000 (Fujifilm). The images were analyzed using the ImageJ software (http://rsbweb.nih.gov/ij/). For direct IPs, the transfer onto a PVDF membrane could be skipped, and the fixed and dried gel was directly exposed to an imaging plate. For coIPs and reIPs, the transfer onto a PVDF membrane is advisable, since this makes the generally weak signals better visible.

57 Materials and Methods

2.7 Transcription

2.7.1 Primer design

Primers were designed automatically by entering the ID of the gene of interest on the homepage of qPrimerDepot (http://mouseprimerdepot.nci.nih.gov/). This software automatically finds primers with a comparable melting temperature that span several introns, so only the cDNA but not the genomic DNA will result in PCR products. The amplicons are between 75 and 150 nucleotides long. The primers used in this study are listed in Table 2-6.

Table 2-6 Primers used for real time PCR. *Primers did not perform well; °GapDH 3 reverse primer was paired with GapDH 2 forward primer; 1primers for genes of interest; 2primers for housekeeping genes for normalization (table continues on the next page).

Gene name Gene ID Forward primer: Primer ID; Reverse primer: Primer ID; sequence sequence oLJ qPCR 1 F; GAA GAG GAA oLJ qPCR 1 R; GTT TCT TGG CGA Calnexin1 NM_007597.3 GAA GAG GAG AAG CT TCT GTT CAA AAT T NM_009318 oLJ qPCR 2 F; CCA TCG AGG oLJ qPCR 2 R; GAC TTC AGG AAT TAPBP1 ACG GCA TCG GC GGT CCA GTA GGC AGC oLJ qPCR 3 R; CTC CTC CCA CAG oLJ qPCR 3 F; AAA GAG TTC BiP (Hspa5) 1 NM_022310.3 TTT CAA TAC CAA TTC AAT GGC AAG GAG C

oLJ qPCR 4 F; GCT ACG TGA oLJ qPCR 4 R; CAG CAC ATT CTT Calreticulin1 NM_007591.3 AGC TGT TTC CGA GCC CTT GTA GTT AA oLJ qPCR 5 F; GTG CAG TGC oLJ qPCR 5 R; TGC AAA TGG CAG GapDH 1*2 NM_008084.3 CAG CCT CGT CC CCC TGG TGA C oLJ qPCR 6 F; GCT CCT AGC oLJ qPCR 6 R; CTC CTG CTT GCT Actb 1*2 NM_007393.4 ACC ATG AAG ATC AA GAT CCA CA oLJ qPCR 7 F; CCT GTG CTG oLJ qPCR 7 R; GCC AGA CGT TGC TAP21 NM_011530.3 TTC TCG GGT TC TTC TGT CC oLJ qPCR 8 F; GTG GCT TCG oLJ qPCR 8 R; TTG AAG GTG ACG Wars1 NM_011710.3 ACA TCA ACA AA TGC TTC TG P58IPK 1*1 oLJ qPCR 9 F; GGT GGA CCT oLJ qPCR 9 R; GTC CGG CTG CGA NM_008929.3 (Dnajc3) GCA GTA CGA A GTA ATT T oLJ qPCR 10 F; AGC TCA ACC oLJ qPCR 10 R; AAG GCA TCA EDEM1 1*1 NM_138677.2 CCA TCT ACT GC ACC AGA GTC AGA oLJ qPCR 11 F; TT CTT TGC oLJ qPCR 11 R; ATG GAG GGG Actb 22 NM_007393.4 AGC TCC TTC GTT AAT ACA GCC C oLJ qPCR 12 F; CGT CCC GTA oLJ qPCR 12 R; TTG ATG GCA GapDH 2* NM_008084.3 GAC AAA ATG GT ACA ATC TCC AC

58 Materials and Methods

Gene name Gene ID Forward primer: Primer ID; Reverse primer: Primer ID; sequence sequence oLJ qPCR 13 R; TCA ATG AAG GapDH 3°2 NM_008084.3 GGG TCG TTG AT oLJ qPCR 14 F; GAC GAG CTC oLJ qPCR 14 R; TGT ATC CAA EDEM1 21 NM_138677.2 AAC CCC ATC TA GGC ATC AAC CA P58IPK 21 oLJ qPCR 15 F; CTG GTG GAC oLJ qPCR 15 R; CTG CAG CGT NM_008929.3 (Dnajc3) CTG CAG TAC G GAA ACT GTG AT oLJ qPCR 16 F; TCA AGA GAG oLJ qPCR 16 R; CTC CCT CTG TTG RPII (Polr2a) 2 NM_001291068.1 TGC AGT TCG GA TTT CTG GG oLJ qPCR 17 F; AGT AAT CGA oLJ qPCR 17 R; AGA AAT CAC gp401 U53330.1 GGC AAC GGC TA GAA CGC GAA GT oLJ qPCR 18 F; TCT GTC GGC oLJ qPCR 18 R; GCG CTC TGG H-2Db1 NM_001267808.1 TAT GTG GAC AA TTG TAG TAG CC oLJ qPCR 19 F; CCC ACA CTC oLJ qPCR 19 R; GCG CTC TGG H-2Kb1 NM_001001892.2 GCT GAG GTA TT TTG TAG TAG CC oLJ qPCR 20 F; GGA AGC CAC oLJ qPCR 20 R; CTT GGG GCT TAP11 NM_013683 TCC TGC TTA TC CTC ATA CAG GA oLJ qPCR 21 F; GGA CCC GGG oLJ qPCR 21 R; ACC AAA GGA LMP71 NM_010724 ACA CTA CAG TT CCT CAG GAA TG oLJ qPCR 22 F; CTC TGC TGA oLJ qPCR 22 R; GTC AAA CTC LMP21 NM_013585 GAT GCT GCG CAC TGC CAT GA oLJ qPCR 23 F; TGG TGC TTG oLJ qPCR 23 R; TTC AGT ATG beta2m1 NM_009735 TCT CAC TGA CC TTC GGC TTC CC

2.7.2 mRNA isolation

Work with mRNA was made complicated by the presence of RNases, which are very stable and active enzymes that can hardly be inactivated. Minor amounts of RNase are sufficient to destroy isolated mRNA. The only way to minimize this risk is to use sterile filter tips, RNase-free water and tubes, and to perform the work at a place that is reserved for RNA work, in the best case under a fume hood. Places where frequent DNA preparations are done should be avoided, since the protocols include the use of highly concentrated RNase. All tools and places should be wiped with RNase inactivating spray (Applichem, No. A7153,0100).

For the isolation of mRNA, the Qiagen kit RNeasy Mini was used (No. 74104) according to the company protocol. One confluent 10 cm dish with K41 cells (3-4*106 cells, either expressing the empty plasmid or gp40 wt) was used for the isolation.

59 Materials and Methods

To the buffer RLT, β-mercaptoethanol was added (10 µl into 1 mL RLT), and to the buffer RPE, four times the volume of ethanol was added (non-denatured ethanol).

The medium was removed from the cell culture dish, and 600 µl of buffer RLT were directly added onto the dish. Cell lysates was collected with a pipette, transferred to a 1.5 mL tube, and carefully pipetted until all cells were lysed. To the lysates, 1 volume of 70% ethanol (non-denatured) was added and mixed by pipetting. The mRNA was collected on an RNeasy spin column by centrifuging the lysates through it in two steps of 700 µl at 8,000 x g for 15 sec at room temperature. The column was washed with 700 µl RW1 at 8,000 x g for 15 sec at room temperature and afterwards with 500 µl RPE at 8,000 x g for 15 sec at room temperature. Then, 500 µl RPE was added onto the column and centrifuged at 8,000 x g for 2 min at room temperature to dry the column. The column was transferred to a 1.5 mL tube and 50 µl RNase free ddH2O was added onto the column membrane. The sample was centrifuged at 8,000 x g for 1 min at room temperature to collect the mRNA. The mRNA was kept on ice and proceeded with immediately or stored at 80 °C.

2.7.3 Determining RNA concentration

The concentration of the isolated RNA was determined before DNase digest, since the buffers used for the digest influence the spectrophotometric absorption of the RNA sample. The purification with the RNeasy mini kit removed most of the genomic DNA, and therefore DNA did not influence the absorption much.

The isolated RNA was diluted 1/10 in TE buffer pH 8.0, and RNA concentration was determined with a nanodrop device (Thermo Fisher). Usually, isolated mRNA has a concentration of 0.2-1 μg/µl in 50 µl volume. Both the 260/280 and 260/230 ratio should be above or close to 2 in pure RNA.

2.7.4 DNase digest

To avoid impairment of the subsequent reactions by genomic DNA carry-over, a DNase digest (TURBO DNA-freeTM Kit, Ambion, No. AM1907) was done before reverse transcription. 5 µl of the 10x TURBO DNase Buffer and 1 µl TURBO DNase were added to the RNA and mixed gently, and the reaction was incubated at 37 °C for 20-30 min. Afterwards, the DNase Inactivation Reagent was resuspended by flicking or vortexing the tube, and 5 µl of it were added to the reaction and mixed well. The reaction was

60 Materials and Methods incubated for 5 min at 26 °C, and the tube was flicked two to three times during the incubation period. Afterwards, the tube was centrifuged at 10,000 x g for 90 sec, and the supernatant was carefully transferred to a fresh tube.

2.7.5 Reverse transcription of mRNA

The mRNA was reverse transcribed into cDNA using the AffinityScript QPCR cDNA Synthesis Kit (No. 600559). The cDNA synthesis reaction was prepared according to Table 2-7.

Table 2-7 cDNA synthesis reaction

Component Amount in µl

RNase free ddH2O variable Oligo(dT) primer (170 ng) 1.7 Random primer (30 ng) 0.3 Affinity Script RT/Block enzyme 1 RNA (0,5 μg) variable Total volume 20

Before use, each component of the kit was vortexed and centrifuged briefly. The cDNA synthesis reaction was prepared according to Table 2-7 in an RNase-free 1.5 mL tube or an RNase-free PCR tube. The reaction was then incubated at 25 °C for 5 min for primer annealing, at 42 °C for 15 min for cDNA synthesis, and at 95 °C for 5 min for the inactivation of the reverse transcriptase.

2.7.6 Quantitative PCR

Quantitative PCR (qPCR) determines the amount of a certain DNA template by monitoring the increase of PCR product after each cycle. This is made possible for example with the help of a dye (SYBR green), which only fluoresces strongly when it is incorporated into double stranded DNA; therefore, the fluorescence intensity increases exponentially with each PCR cycle. A certain threshold is set, and the lower the cycle number at which the fluorescence crosses it during a PCR reaction, the higher the DNA amount is concluded to be. The qPCR reaction was prepared according to Table 2-8.

61 Materials and Methods

Table 2-8 qPCR synthesis reaction. *The reference dye was prepared fresh every time by adding 1 µl of the dye stock to 499 µl of RNase free ddH2O.

Component Amount in µl

RNase free ddH2O 3.7 2x SYBR Green QPCR master mix 10 Forward primer (2 pmol/µl) 2 Reverse primer (2 pmol/µl) 2 Reference dye (diluted)* 0.3 cDNA 2 Total volume 20

The components were pipetted into qPCR tubes (Brand, No. 781333) using filter pipettes. After closing the lids, samples were mixed by flicking the tubes and briefly centrifuged. The tubes were inserted into the qPCR cycler (Stratagene Mx3000P), and the qPCR program was started (see Table 2-9).

Table 2-9 Thermocycling conditions for qPCR reaction

Step No. Temperature Duration Notes 1 95 °C 3:00 initial denaturation 2 95 °C 0:10 denaturation 3 58 °C 0:20 annealing and extension repeat step 2 and 3 45 times 4 95 °C 1:00 denaturation 5 55 °C 0:30 annealing of complementary DNA strands 6 95 °C 0:30 DNA melting curve

Before starting with the actual investigation, the specificity of the primers and the purity of the cDNA preparation had to be tested. For this, several controls for a certain primer were made in duplicate: no template control (ddH2O instead of cDNA template; controls for contamination in the compounds); no reverse transcriptase control (mRNA preparation before cDNA synthesis as template; controls for DNA contamination); as well as three to four ten-fold dilution steps for the standard curve to determine the efficiency of the PCR reaction.

After the PCR reaction, there were two data sets to investigate: the amplification plots and the melting curves. The amplification plots show how the PCR product increases and at what cycle number they cross the threshold, which gives information about the abundance of the template. The melting curve

62 Materials and Methods shows at what temperature the double stranded DNA denatures (the fluorescence disappears). This gives information about the specificity of the primers. Specific primers should result in PCR products of only one length, which shows as one sharp peak in the melting curve. If the primers are unspecific, there are multiple peaks or very broad peaks.

There should be no or only very little PCR product in the control wells (no reverse transcriptase and no template). If the negative controls resulted in PCR products and unspecific peaks in the melting curve, the primers were deemed unsuitable and new primers were designed.

The standard curve gives information about the efficiency of the PCR. The resulting cycle numbers are plotted against the dilution factor, and the slope x of the standard curve is used to calculate the efficiency E with the following equation:

The efficiency should be in between 90 to 100%, but not above 100%.

All these controls verify primer specificity, exclude contaminations and control for good performance of the tools and the components.

To investigate the relative abundance of a gene of interest, it is amplified together with two or three unrelated housekeeping genes (see Table 2-6), all in duplicate. The term housekeeping is relative, as genes such as Actinb or Gapdh can also change during a certain treatment. Among the tested housekeeping genes, the one or two are chosen that show no or minimal change after treatment. To calculate the relative abundance R of a gene of interest GOI in a treated (tr) versus untreated (untr) sample, the cycle number Ct of the GOI and the housekeeping gene HG are used in the following equation:

63 Manuscript

3 Manuscript

In this section, I show that MHC class I retained by gp40 are properly folded and peptide loaded and that the PLC members are not required for gp40 function. Furthermore, I show for the first time with co- immunoprecipitation that gp40 and MHC class I interact and that retention of gp40 itself is necessary for retention of MHC class I.

The first part of this chapter (3.1.-3.8.) is a submitted manuscript, and the second part (3.9) presents additional data regarding the interaction of gp40 and MHC class I molecules.

The experimental work for Figure 3-1A, Figure 3-2C-D, Figure 3-3, Figure 3-4, Figure 3-8A, and Figure 3-10C was done by me, all other figures (apart from the microscopy) were generated by V. Ramnarayan. The microscopy images were generated by M. Aboelmagd. S. Fritzsche and A. Halenius supported us in terms of experimental design, interpretation of the data, and general discussion of the project. The manuscript was written by Sebastian Springer and me.

64 Manuscript

3.1 The mouse cytomegalovirus protein m152/gp40 retains MHC class I molecules in the early secretory pathway by direct interaction

Linda Janßen1, Venkat Raman Ramnarayan1, Mohamed Aboelmagd1, Susanne Fritzsche1, Anne Halenius2,

and Sebastian Springer1*

1Department of Life Sciences and Chemistry, Jacobs University Bremen, Germany;

2Institute of Virology, University of Freiburg, Germany

Correspondence: Jacobs University Bremen Campus Ring 1 28759 Bremen, Germany [email protected] +49 421 200 3243

Short title: mCMV gp40 – MHC class I interaction

65 Manuscript

3.2 Abstract

In the presence of the murine cytomegalovirus (mCMV) gp40 (m152) protein, murine major histocompatibility complex (MHC) class I molecules do not reach the cell surface but are retained in an early compartment of the secretory pathway. We find that gp40 does not impair folding or high-affinity peptide binding of class I molecules but binds them to retain them in the endoplasmic reticulum (ER), the ER-Golgi intermediate compartment (ERGIC), and the cis-Golgi, most likely by retrieval from the cis- Golgi to the ER. We identify a novel sequence in gp40 that is required for both its own retention in the ER–cis-Golgi cycle and that of class I molecules.

3.3 Introduction

To escape from the cellular adaptive immune system, viruses inhibit almost every step of major histocompatibility complex (MHC)7 class I mediated antigen presentation (Ambagala et al., 2005). Herpesviruses, as large DNA viruses, encode multiple interfering proteins (Basta and Bennink, 2003). In murine cytomegalovirus (mCMV) infection, the glycoprotein 40 kDa (gp40), encoded by the m152 gene, was the first inhibitor of MHC class I-mediated antigen presentation and natural killer cell function to be described in CMV (Krmpotid et al., 2002; Lu et al., 2006). Gp40 inhibits the transport of peptide-loaded class I molecules (proteins) to the cell surface and retains them in the early secretory pathway, but its molecular mechanism of action is still unknown (Pinto and Hill, 2005). The group of Koszinowski has investigated gp40 and found that retained class I molecules are bound to high-affinity peptide but fail to proceed beyond the ERGIC. In contrast, gp40 itself was shown to progress to the lysosomes for degradation. The authors found an interaction of gp40 with calnexin but not with class I (Ziegler et al., 2000; Ziegler et al., 1997).

Gp40 also retains intracellularly another protein, the class I-related stress marker RAE-1. Recently, the group of Margulies published a crystal structure of gp40 in complex with RAE-1 (Zhi et al., 2010). This has prompted us to re-examine the question of gp40-class I interaction. We show here that gp40 binds

7 b b Abbreviations: β2m, beta-2 microglobulin; D , H-2D ; EndoF1, endoglycosidase F1; ER, endoplasmic reticulum; ERGIC, ER-Golgi intermediate compartment; gp40, glycoprotein 40 kDa; HA, (influenza) hemagglutinin; Kb, H-2Kb; mAb, monoclonalantibody; mCMV, murine cytomegalovirus; MHC, major histocompatibility complex; PLC, peptide loading complex; TAP, transporter associated with antigenprocessing; UGT1, UDP-glucose:glycoprotein glucosyltransferase.

66 Manuscript directly to class I molecules to retain them in the early secretory pathway, most likely by retrieval from the ERGIC/cis-Golgi, and we demonstrate that a sequence in the linker between the folded lumenal domain of gp40 and the transmembrane sequence is required for this retention.

3.4 Results

3.4.1 Gp40 retains MHC class I in the early secretory pathway

To assess the effect of gp40 on murine class I molecules, we expressed m152 in K41 cells (murine fibroblasts) by lentiviral transduction. The surface levels of the endogenous class I allotypes H-2Db (Db) and H-2Kb (Kb) were reduced to background levels in most cells, as observed by flow cytometry with the

b b allotype-specific beta-2 microglobulin (β2m)-dependent antibodies B22.249 (for D ) and Y3 (for K ) (Figure 3-1A). Kb, but not Db, was resistant to gp40 in some cells (arrow), especially in confluent cultures. We conclude that while gp40-mediated retention of class I can be highly effective, it varies between class I allotypes and growth conditions.

67 Manuscript

Figure 3-1 Gp40 intracellularly retains MHC class I. A: Gp40 abolishes class I surface expression. Gp40 was expressed in K41 cells, and surface expression of endogenous class I was determined by cell surface staining with B22.249 (Db) and Y3 (Kb) and flow cytometry. Shadow: 2nd antibody control; solid line: empty vector control; dashed line: gp40-expressing cells. One experiment out of five is shown. B: In the presence of gp40, Kb and Db localize to juxtanuclear compartments, and Db additionally to the ER. gp40- expressing K41 cells were fixed, permeabilized and stained with HA-FITC (gp40), B22 (Db) or Y3 (Kb), and Cy3 as secondary antibody. More images are shown in Figure S1. C-E: Pulse-chase analysis of class I and gp40. Cells were pulse labeled for 5 (C, E) or 10 minutes (D), chased for the indicated times, and lysed. Proteins of interest were immunoprecipitated with anti-HA (gp40), B22 (Db), or Y3 (Kb), digested with EndoF1 as indicated, and separated on SDS-PAGE. Sia = sialylated band; black arrowheads = partial EndoF1 resistance; s = EndoF1 sensitive band; asterisks = unspecific bands. One experiment out of three is shown. Drawn by the author. Adapted with permission (Janßen et al., 2016).

68 Manuscript

To find the intracellular steady-state location of the gp40-retained class I, we performed fluorescence microscopy with the same antibodies, and we also stained gp40 with an antibody against its carboxy- terminal hemagglutinin (HA) tag (Figure 3-1B and Figure 3-7). Gp40 was observed in a compact juxtanuclear location that colocalized well with the Golgi marker GM130 (Figure 3-7B). Kb was found in the same location (Figure 3-1B, top row, and Figure 3-7A). In some cells (about 30%), Db localized exclusively to the same juxtanuclear region (Figure 3-1B, bottom row, and Figure 3-7A), but in the majority of cells, Db exhibited in addition a spread-out reticular pattern reminiscent of the ER (Figure 3-1B, center row, and Figure 3-7A).In agreement with earlier data for H-2Kd, (Ziegler et al., 1997), we conclude that Db and Kb are indeed retained in an intracellular compartment where they colocalize with gp40, and we also find that the exact subcellular steady-state location of a gp40-retained class I molecule depends on the allotype (Kavanagh et al., 2001).

To understand the influence of gp40 on the kinetics of class I transport from the ER to the Golgi apparatus, we next performed radiolabeling and pulse-chase experiments with endoglycosidase F1 digests (EndoF1; see materials). In wild type cells, Db and Kb glycans become resistant to digestion with EndoF1 (and EndoH) in the medial Golgi due to the action of mannosidase II. When they reach the trans- Golgi, further carbohydrates including sialic acids are added; this sialylation results in an upward shift of the EndoF1-resistant band in SDS-PAGE (Fritzsche and Springer, 2013). In gp40-expressing cells, Db showed no sialylation after two hours of chase (Figure 3-1C, bottom row); thus, it had not passed the trans-Golgi. In contrast, Kb showed a small amount of EndoF1 resistance and sialylation (Figure 3-1C, second row, sia). This agrees with the above observation that in the presence of gp40, Kb can still reach the surface in some cells (Figure 3-1A).

Remarkably, at the two-hour time point, both Db and Kb acquired partial (incomplete) EndoF1 resistance, visible as two (Kb) or three (Db; corresponding to the number of glycosylation sites for each allotype) intermediate bands between the EndoF1-resistant and –sensitive forms (Figure 3-1C, arrowheads). This was never seen in the gp40-free control samples, where Db and Kb progressed directly from the EndoF1- sensitive form to the fully EndoF1-resistant and sialylated form. At longer chase times of up to 16 hours, the partially EndoF1-resistant forms of Db and Kb persisted (Figure 3-1D, arrowheads), and sialylation still did not occur (Db), or only for a small fraction of the molecules (Kb). To determine whether the partially EndoF1-resistant forms were still intracellular or had progressed to the cell surface, we pulse-labeled gp40-expressing cells and chased for 120 minutes. We then incubated the intact cells with biotinylated Kb-specific peptide for 120 minutes, washed, lysed the cells, and found that only the sialylated, but not

69 Manuscript the EndoF1-sensitive or the partially resistant forms of Kb precipitated with streptavidin agarose (Figure 3-8A). Thus, the partially EndoF1-resistant forms of class I in gp40-containing cells were not cell surface forms but trapped in the cell interior, just like the partially EndoF1-resistant class I proteins in TAP- deficient cells, which circulate between the ER and the cis-Golgi (Fritzsche et al., 2015). Since at least Db does not become EndoF1-resistant at all in gp40-containing cells, we conclude that the compact juxtanuclear steady-state location observed for Db (and Kb) in Figure 3-1B must be a pre-medial-Golgi compartment, most likely the ERGIC and/or the cis-Golgi, as suggested before (del Val et al., 1992). In agreement with earlier reports, Db and Kb were not rapidly degraded during the chase but instead slightly more stable than in wild type cells (Figure 3-8B and (Ziegler et al., 2000)), suggesting that they are not transported to lysosomes in the presence of gp40.

Together with the microscopy, the results of the pulse-chase experiments suggest that in the presence of gp40, Db and Kb do not become transported to the medial Golgi but are present in the ERGIC/cis-Golgi (Kb) and the ER (Db) at steady state, most likely circulating between these compartments.

We next wondered whether gp40 travels from the ER to the Golgi in synchrony with class I, and we followed it in a pulse-chase experiment under the same conditions as above. Intriguingly, a form of gp40 that was fully EndoF1 resistant and sialylated (Figure 3-1E, sia) and several partially EndoF1-resistant forms, resembling those of class I (arrowheads), appeared after 15 minutes, and both fully and partially resistant forms persisted throughout the two-hour chase. Partial EndoF1 resistance during the chase was also visible for a variant of gp40 with a C-terminal cytosolic –KKSQ sequence for ER retention (Figure 3-9A). Thus, in our system, at least part of gp40 is retained in a pre-medial-Golgi compartment, most likely in an ER-Golgi cycle just like the class I molecules that it retains. Interestingly, the gp40–KKSQ variant very efficiently retained both Db and Kb, suggesting that gp40 does not need to progress beyond the cis-Golgi to be effective (Figure 3-9B).

3.4.2 Gp40 does not impair class I maturation or peptide binding

Many viral proteins interfere with class I synthesis and folding or with peptide loading onto class I, for example by inhibiting or destroying the transporter associated with antigen processing (TAP) (Loch and Tampe, 2005; Momburg and Hengel, 2002). Since our results suggested that gp40 acts in the early secretory pathway (i.e., the ER, ERGIC, and/or the cis-Golgi), we first asked whether it inhibits class I folding and peptide binding. We applied a five-minute radioactive pulse and followed the folding of the

70 Manuscript

class I heavy chains by immunoprecipitating with β2m-dependent antibodies at different chase times and quantifying total Db or Kb. For neither allotype, gp40 caused a significant change in the folding kinetics (Figure 3-2A). To assess the binding affinity of the overall peptide load of Db and Kb, we then performed thermal denaturation experiments by heating cell lysates to different temperatures and immunoprecipitating with Y3 and B22 (Gao et al., 2002; Garstka et al., 2011). There was no evidence of any impairment of peptide loading in the presence of gp40 for either allotype such as it is seen for tapasin or TAP deficiency (Figure 3-2B; compare Figures S2D and S4D in (Fritzsche et al., 2015)) Thus, in agreement with earlier reports (del Val et al., 1992; Ziegler et al., 2000), gp40 does not impair class I folding, maturation, or peptide binding.

71 Manuscript

Figure 3-2 The PLC is not involved in gp40-mediated retention. A: Gp40 does not impair class I maturation. K41 cells with or without gp40 were labeled for 5 min, chased for the indicated times, and lysed. Proteins were immunoprecipitated with conformation-specific antibodies B22.249 (Db) or Y3 (Kb), digested with PNGase, separated by SDS-PAGE, and quantified. The signal intensity was normalized to the start of the chase. Average of three independent experiments; error as standard deviation. B: Gp40 does not impair class I peptide binding. K41 cells with or without gp40 were labeled for 5 min, chased for the indicated times, and lysed. Lysates were heated to the indicated temperatures for 10 minutes; then, proteins were immunoprecipitated with conformation-specific antibodies B22.249 (Db) or Y3 (Kb), digested with PNGase, separated by SDS-PAGE, and quantified. The signal intensity was normalized to signal intensity of 4 °C. Average of two independent experiments; error as standard deviation. C: Gp40 does not use any PLC components to retain class I. gp40-expressing murine fibroblasts with a functional deficiency in the indicated protein were surface-stained for flow cytometry with B22.249 (Db) and Y3 (Kb). Cells were analyzed in flow cytometry. Lines as in Figure 1A. One experiment out of at least two is shown. D: Gp40 does not force permanent class I association with the PLC. K41 cells with or without gp40 were labeled for 10 min, chased for the indicated times, and lysed in 1% digitonin buffer. Tapasin was immunoprecipitated from the lysate (1st IP), and co-precipitated Db and Kb were re-precipitated (reIP) as described in Materials and Methods and Figure S4A, treated with PNGase, and separated by SDS-PAGE. One experiment out of two is shown. Drawn by the author. Reproduced with permission (Janßen et al., 2016).

72 Manuscript

3.4.3 Gp40-mediated retention of class I does not use the peptide loading complex

Class I molecules loaded with suboptimal peptide ligands are retained in the early secretory pathway of wild type cells by two different mechanisms (Springer, 2015): first, by the class I-specific chaperone tapasin (Paulsson et al., 2002), and second, by the lectin calreticulin (Howe et al., 2009), probably in concert with the UDP-glucose:glycoprotein glucosyltransferase (UGT1) (Zhang et al., 2011). Both tapasin and calreticulin are members of the class I peptide loading complex (PLC). To investigate whether gp40 somehow appropriates these or similar class I retention mechanisms that already exist in the cell, we studied the effect of gp40 in cell lines that lack functional calreticulin, tapasin, TAP, calnexin, or UGT1 (Figure 3-2C). In every case, gp40 caused retention of both Db and Kb to the same extent as in wild type cells. We conclude that no single known class I-associated protein of the early secretory pathway is required for gp40-mediated retention.

Class I is released from the PLC after binding high-affinity peptide (Ortmann et al., 1994; Springer, 2015). We next asked whether this interaction is prolonged by gp40 to achieve class I retention, and so we performed a pulse-chase experiment, precipitated tapasin from the cell lysate, and analyzed the class I molecules bound to it in a re-precipitation experiment (Figure 3-2D; schematic Figure 3-10A). Some Db was indeed bound to tapasin throughout the chase, but this was the case both in the presence and absence of gp40. Binding of Kb to tapasin appeared stronger in the presence of gp40, but the Kb band intensity clearly decreased over time, which suggests eventual dissociation of Kb from tapasin and not an irreversible association. We conclude that gp40 does not use the PLC, or any of its proteins individually, to retain class I in the early secretory pathway.

3.4.4 Gp40 binds directly to Db and Kb

Direct binding to class I is a hallmark of many viral immunoevasins (Bennett et al., 1999; Furman et al., 2002; Jones et al., 1996). To test whether gp40 binds directly to class I, we immunoprecipitated gp40-HA from lysates of radiolabeled cells and re-precipitated with an antiserum against a common sequence in the cytosolic tails of Db and Kb (Db/Kb serum). Both Db and Kb co-precipitated with gp40; the interaction was also seen when we first immunoprecipitated with Db/Kb serum and then re-precipitated with anti- HA (Figure 3-3A). Since Db and gp40 have an almost identical molecular weight on SDS gels, we demonstrated in HeLa cells that the Db/Kb serum does not cross-react with gp40 (Figure 3-10B). Gp40 also co-precipitated with tapasin, calnexin, and calreticulin (Figure 3-10C), but gp40 interaction with

73 Manuscript class I did not require tapasin, since in tapasin-deficient cells, it was still observed (Figure 3-3B). Intriguingly, first-round immunoprecipitations with the mAbs B22.249 and Y3 co-precipitated much less gp40 than the Db/Kb serum. This suggests that gp40 might mask the epitopes of these antibodies, which

b b lie in the α1/α2 superdomains of D and K . In contrast, the mAb 28-14-8S, which binds to the α3 domain of Db, efficiently co-precipitated gp40 (Figure 3-3C). Our data thus suggest that gp40 binds directly to the

b b d α1/α2 superdomain of D and K , as proposed previously for L (Wang et al., 2012).

Figure 3-3 Gp40 interacts with class I. A: Gp40 coprecipitates with Db and Kb. Gp40-expressing K41 cells were radiolabeled for 30 min and lysed. Proteins of interest were immunoprecipitated with serum against the cytosolic tail of Db and Kb or anti-HA MAb (for HA- gp40). Immunoprecipitates were dissociated, and associated proteins were re-precipitated as indicated, PNGase digested, and separated by SDS-PAGE. B: Gp40-class I interaction does not require tapasin. Experiment as in Figure 3A, but with gp40-expressing tapasin-deficient mouse fibroblasts. C: Gp40 binds to the α1/α2 domain. As in Figure 3A, but for the first immunoprecipitation conformation specific antibodies Y3 (Kb), B22 (Db), 28-14-8s (Db), and Db/Kb antiserum were used; re-precipitation was done with anti-HA mAb or Db/Kb antiserum as indicated. Proteins were PNGase digested and separated by SDS-PAGE. Drawn by the author. Reproduced with permission (Janßen et al., 2016).

3.4.5 The class I – gp40 interaction persists in the early secretory pathway

We next hypothesized that gp40 and the class I molecules bound to it might be retained together in the early secretory pathway, and so we decided to investigate the EndoF1 resistance pattern of class I- associated gp40 in pulse-chase experiments with re-precipitations (Figure 3-4). Association of gp40 with Db and Kb was observed at the start of the chase, suggesting that it occurs shortly after synthesis of the three proteins. Those gp40 molecules that were bound to class I (panel D) became partially EndoF1- resistant over time, but they showed little sialylation (as compared to the entire cohort of gp40 at the same time point, panel C), suggesting that most did not progress beyond the ER-Golgi cycle. Likewise, in

74 Manuscript the class I molecules bound to gp40, the sialylated band of Kb was very weak (panel B, arrow; compare panel F, arrow), suggesting that gp40 associates only with the intracellular forms of class I. The simplest explanation of these data is that gp40 and class I form a complex that is retained in the ER/ERGIC/cis- Golgi, probably by cycling through these compartments.

Figure 3-4 The gp40-class I complex cycles through the early secretory pathway. Gp40-expressing K41 cells were labeled for 10 min, chased for the indicated times and lysed in 1% digitonin lysis buffer. Proteins of interest were immunoprecipitated with conformation specific antibodies Y3 (Kb, F), B22 (Db, E), Db/Kb antiserum (A, D) or HA (gp40) (B, C) antibody; after denaturation in 1% SDS, either gp40 (C, D) or Db/Kb (A, B) were re-precipitated, EndoF1 digested as indicated, and separated by SDS-PAGE. Drawn by the author. Reproduced with permission (Janßen et al., 2016).

3.4.6 A sequence in the linker of gp40 is essential for gp40–class I complex retention

We next decided to investigate the mechanism of retention of gp40 in the early secretory pathway. Since the sequence of gp40 contains no known retention signal, we tested a panel of mutants (not shown). In one such mutant, we replaced the 43 amino acid linker between the folded luminal domain and the transmembrane domain (Wang et al., 2012) by a (glycine4-serine)9 sequence to yield the gp40-

(G4S)9 mutant (Figure 3-5A). In a pulse-chase, gp40-(G4S)9 was much more strongly sialylated then wild

75 Manuscript type gp40, suggesting fast progress to the trans-Golgi, and much shorter-lived, suggesting rapid degradation in lysosomes (Figure 3-5B). Gp40-(G4S)9 no longer decreased the steady-state surface levels of Db or Kb (Figure 3-5C, D), even though direct binding was still detectable (Figure 3-5E). The simplest explanation of these data is that retention of the gp40/class I complex in the early secretory pathway is achieved through a sequence in the linker of gp40.

Figure 3-5 Gp40-(G4S)9 rapidly exits the early secretory pathway, and does not retain class I. A: Schematic representation of gp40. Indicated are the lumenal domain, the linker sequence, the transmembrane domain and cytoplasmic tail (TMD/CT), the HA tag at the C terminus, and the residue numbers of the mature protein. B: Gp40 (G4S)9 rapidly exits the early secretory pathway. K41 cells expressing gp40 (G4S)9 were pulse labeled for 5 minutes and chased for the indicated times, and lysed with 1% Triton-TX100 buffer. Gp40 was immunoprecipitated with HA antibody, digested with EndoF1 as indicated, and separated by SDS-PAGE. Sia = sialylated band; black arrowhead = EndoF1 resistant band; s = EndoF1 sensitive band. One experiment out of three is shown. C: Gp40 (G4S)9 does not retain class I. K41 cells expressing empty vector, gp40 wt or gp40 (G4S)9 were surface stained with B22.249 (Db) and Y3 (Kb). Cells were analyzed in flow cytometry. Shadow: 2nd antibody control; solid line: empty vector control; dashed line: gp40 wt expressing cells. Long-dashed line: gp40 (G4S)9. A representative experiment is shown. D: Gp40 (G4S)9 does not retain class I. Average of eight independent experiments as in Figure 5C, MFI normalized to empty vector-expressing cells. Error as SEM. Drawn by the author. Adapted with permission (Janßen et al., 2016).

76 Manuscript

E: Gp40 (G4S)9 binds weakly to class I. K41 cells expressing gp40 wt or gp40 (G4S)9 were labeled for 15 min and lysed in 1% digitonin buffer, and proteins were immunoprecipitated with Db/Kb serum or anti-HA (for gp40-HA) mAb; after denaturation in 1% SDS, gp40 or Db/Kb were reprecipitated as indicated, PNGase digested, and separated by SDS-PAGE. Drawn by the author.

3.5 Discussion

We demonstrate here for the first time the physical interaction of murine class I molecules with mCMV gp40 that causes their retention in the cell interior. From a crystal structure of gp40 with the class I-like protein RAE-1, Margulies and collaborators have indeed predicted that gp40 binds to the top of the

α1/α2 superdomain of class I (Wang et al., 2012). Our data in Figure 3-3C support this hypothesis: both

MAbsY3 and B22.249, which bind to the α1/α2 superdomain of class I, preclude gp40 co-precipitation, whereas 28-14-8S, which binds to the α3 domain, co-precipitates gp40 (Allen et al., 1984; Nathenson et al., 1989). In the model of Margulies, there is space for class I-bound peptide, which agrees with the observation that gp40 does not inhibit peptide binding (Figure 3-2A, B; (del Val et al., 1992)). In previous co-precipitation experiments by Koszinowski and collaborators, a gp40/class I interaction was not found (Ziegler et al., 2000). Since the authors of that study did not re-precipitate, they could not use the heavy chain signal to detect co-precipitated class I (it appears at the same molecular weight as gp40); thus, they depended on the β2m signal, which might have been too weak to observe in their system. We have also found that gp40 does not bind to the human class I molecules in HeLa cells, which might explain why gp40 does not retain human class I ((Ziegler et al., 1997); data not shown).

We think that that in the cell, the class I/gp40 complex is restricted to a compartment prior to the medial Golgi, since class I-bound gp40 and gp40-bound class I both acquire little sialylation over time (which would signify arrival in the trans-Golgi) and do not even become completely EndoF1 resistant (Figure 3-4). By microscopy, in the presence of gp40, Db is mostly found in the ER, whereas Kb is mostly in the ERGIC/cis-Golgi. Since Kb is exported from the ER with greater efficiency than Db (Fritzsche et al., 2015), this suggests a scenario of class I retention by retrieval of gp40/class I complexes from the cis- Golgi to the ER; in such an export and retrieval cycle, the higher ER exit rate of Kb might shift its steady- state location further towards the ERGIC/cis-Golgi. The faster anterograde transport of Kb might also cause the escape of Kb from gp40-mediated retention that we observe in some cells (Figure 3-1A, arrow). When Kb is forced by gp40 to circulate between ER and Golgi, its increased concentration in the early secretory pathway might cause the tighter association with the PLC that we observe (Figure 3-2D).

77 Manuscript

The localization of gp40 in the ERGIC/cis-Golgi at steady state that we find agrees with the microscopy results of Koszinowski and collaborators, who showed excellent co-localization with p58 (Ziegler et al., 1997). They demonstrated that the steady state location of gp40 shifts to the lysosomes if protein degradation there is inhibited with leupeptin (Ziegler et al., 2000); this suggests that eventually, most gp40 leaves the ER-Golgi cycle to become degraded in the lysosomes. In our experiments, when expressed from a retroviral vector, gp40 is rather long-lived (Figure 3-1E), whereas in the work of the Koszinowski group, where cells are transfected with an expression plasmid, the bulk of gp40 is degraded after two hours (without cycloheximide) or EndoH-resistant, suggesting transit through the medial Golgi (with cycloheximide; (Ziegler et al., 2000)). We think that overexpression of gp40 might saturate the retention mechanism that holds gp40 in the early secretory pathway and thus lead to the transport of bulk excess gp40 to the lysosomes for degradation. Even in our expression system, retention of gp40 is not complete, since we observe some sialylated gp40 in the pulse-chase (Figure 3-1E).

Taken together, our data and those from the literature suggest that gp40 is temporarily held in the early secretory pathway by a saturable retention mechanism. This retention clearly depends on an amino acid sequence in the gp40 linker (Figure 3-5), but its molecular mechanism is not obvious since gp40 lacks any known retention or retrieval motif. As we and others have shown, gp40 interacts with calnexin and calreticulin, which have ER retention signals ((Ziegler et al., 2000) and Figure 3-10C), but neither protein is required for gp40-mediated class I retention (Figure 3-2C). Perhaps gp40 can bind to several different proteins to remain in the early secretory pathway, in analogy to the 'dynamic matrix' model (Nehls et al., 2000).

So how does gp40 retain class I molecules? Our data suggest that gp40 associates with class I in the ER, very soon after synthesis (Figure 3-4). The complex then circulates in the early secretory pathway for several hours (Figure 3-4, Figure 3-1D). Retention of gp40 itself in this cycle is necessary for class I retention (Figure 3-5, Figure 3-6). Eventually, the complex of class I and gp40 travels to the lysosomes for rapid degradation, perhaps by direct transfer from the trans-Golgi, avoiding the cell surface. In earlier work, the large amount of overexpressed gp40 that was rapidly transported to the lysosomes may have obscured those gp40 molecules that remained in the ER/cis-Golgi to retain class I.

78 Manuscript

Figure 3-6 Proposed mechanism for gp40-mediated MHC class I retention. In the absence of gp40 (left), MHC class I heavy chains (dark red) bind to β2m (pink) and high-affinity peptide (blue) before being exported to the cell surface. Once they pass the medial Golgi, they become fully EndoF1 resistant. In gp40-expressing cells (right), fully assembled and peptide-loaded MHC class I molecules are bound in the ER by gp40 (orange), which itself is bound by an unknown retrieval factor (green) that returns the complex to the ER. Cycling in the early secretory pathway leads to a slow maturation of the glycans of both gp40 and MHC class I and to their partial EndoF1 resistance. Drawn by the author. Adapted with permission (Janßen et al., 2016).

3.6 Materials and Methods

Antibodies, peptides, and reagents. Chemicals were purchased from AppliChem (Darmstadt, Germany) or Carl Roth (Karlsruhe, Germany). Mouse monoclonal hybridoma supernatants Y3 (Hammerling et al., 1982), B22.249 (Lemke et al., 1979) and HA 12CA5 (Niman et al., 1983) were described before. Tapasin serum 2668 was kindly donated by Dr. Xiaoli Wang. Rabbit anti-calnexin serum was kindly provided by D. Williams (Toronto, Canada). PA3‐900 antiserum (Pierce Antibodies) and MAb GM130 (BD transduction Laboratories, No. 610822) were purchased. Rabbit antiserum against H-2Db and H-2Kb was generated by Charles Rivers Laboratories (Kisslegg, Germany) against the peptide C-RRRNTGGKGGDYALAPGSQ corresponding to the membrane proximal C-terminal cytosolic tail of both H-2Db and H-2Kb (residues

79 Manuscript

331 - 349, UniProtKB/Swiss-Prot accession number P01901); Biotinylated SIINFEKL was synthesized by Genecust.

Cells. K41 and K42 cells (Gao et al., 2002) kindly provided by Tim Elliott, MEF TPNd (Grandea et al., 2000) and MEF TAPd were kindly provided by Luc van Kaer, MEF CNXd (Kraus et al., 2010) were a gift from Jody Groenendyk, MEF UGT1d (Solda et al., 2007) were a gift from Tatiana Soldà. Cells were grown at 37 °C and 5% CO2 in high glucose (4,5 g/L) DMEM medium (GE Healthcare Europe, Freiburg, Germany) supplemented with 10% FCS (Biochrom, Berlin, Germany), 2 mM glutamine, 100 u/mL penicillin, and 100 µg/mL streptomycin. For flow cytometry experiments with MEF TAP and MEF TPN deficient cells, cells were incubated before staining at 25 °C in CO2-independent medium (Life Technologies, Darmstadt, Germany) supplemented as above to accumulate MHC class I at the cell surface (Ljunggren et al., 1990).

Retroviral expression, microscopy and flow cytometry were performed as in (Hein et al., 2014).

Pulse chase experiments were performed as in (Fritzsche and Springer, 2014). For Figure S2A, 5 µM Kb- specific biotinylated peptide (ovalbumin 357-64 peptide, sequence SIINFEKbioL, with biotin attached to the lysine side chain by a 6-aminohexanoic acid linker) was added to the cells, followed by lysis in the presence of 10 µM non-biotinylated peptide (to prevent post-lysis binding), precipitation with streptavidin agarose, and SDS-PAGE (lane 4). For the other lanes, biotinylated peptide was added only after lysis (without non-biotinylated peptide) in order to precipitate all forms of Kb. Precipitation was with streptavidin agarose or MAb Y3 and protein A sepharose, as indicated.

Co-immunoprecipitation and re-immunoprecipitation. Labeling, pulse chase and co- immunoprecipitation were performed as in (Halenius et al., 2011) except that protein A agarose (Merck Millipore) was used instead of protein A sepharose. Precipitated proteins were then eluted from the agarose beads by boiling in 50 µL denaturation buffer (1% SDS, 2mM DTT) at 95 °C for 10 min. Samples were cooled on ice, and SDS was neutralized with the 20-fold volume (1 mL) of 0.1% Triton X-100 in PBS. Samples were centrifuged at 1000 x g for 10 min, and 900 µL were transferred to protein Agarose beads prebound to the respective antibody for the re-immunoprecipitation and incubated for 1 hour at 4 °C rotating. The beads were washed twice in PBS with 0.1% TX-100, and precipitated proteins were eluted by boiling in 20 µL denaturation buffer at 95 °C, 5 min, for SDS-PAGE.

80 Manuscript

3.7 Acknowledgements

We would like to thank Hartmut Hengel for advice on the project and the manuscript; Constanze Wiek and Helmut Hanenberg for the retroviral expression system; Peter Reinink for computational biology support; those mentioned in the Materials and Methods for donated reagents; Uschi Wellbrock for excellent technical assistance; Ina Huppertz, Maria Bottermann, Andrei Iosif Șmid, and Florin Tudor Ilca for excellent preparatory and additional laboratory work on the project; and for funding, the Tönjes Vagt Foundation of Bremen (grant XXIX to S.Sp.).

81 Manuscript

Figure 3-7 In the presence of gp40, Kb and Db localize to juxtanuclear compartments, and Db additionally to the ER. A: Subcellular localization of gp40, Db and Kb. Gp40-HA-expressing K41 cells were fixed, permeabilized, and stained with anti-HA-FITC, B22 (Db), or Y3 (Kb), and with anti-calnexin serum. Drawn by the author.

82 Manuscript

B: Gp40 colocalizes with the Golgi marker GM130. Gp40-HA-expressing K41 cells were fixed, permeabilized and stained with anti-HA-FITC and anti-GM130 serum. Drawn by the author.

83 Manuscript

Figure 3-8 Characterization of retained MHC class I molecules. A: The partially EndoF1-resistant forms of Kb are intracellular. EndoF1-sensitive and partially EndoF1-resistant bands represent intracellular Kb molecules. Cells were pulse labeled for 5 minutes and chased for 120 minutes. Then, Kb-specific biotinylated peptide was added to the intact cells to capture only the surface class I, followed by lysis of the cells, or to the detergent lysate to capture total class I, as indicated. Kb/peptide complexes were precipitated with streptavidin agarose or with MAb Y3 and then treated with PNGase or EndoF1 as indicated prior to SDS-PAGE. One out of two experiments is shown. B: Db and Kb are not rapidly degraded in cells expressing gp40. K41 expressing gp40 or the empty vector control were pulse labeled for 10 minutes and chased for the indicated times and lysed in 1% Triton-TX100 buffer. Proteins of interest were immunoprecipitated with B22 (Db) or Y3 (Kb), digested with PNGase, and separated by SDS-PAGE. One out of two experiments is shown. Drawn by the author.

Figure 3-9 Retrieval by a KKXX signal leads to partial EndoF1 resistance. A: Gp40 with a KKXX retention signal shows partial EndoF1 resistance. K41 cells expressing gp40-HA-KKSQ were pulse labeled for 10 minutes and chased for the indicated times and then lysed in 1% Triton-TX100 buffer. gp40- HA-KKSQ was immunoprecipitated with anti-HA MAb, digested with EndoF1 and PNGase as indicated, and separated by SDS-PAGE. One out of two experiments is shown. B: Gp40-KKSQ retains Db and Kb. K41 cells expressing empty vector or gp40-HA-KKSQ were surface stained with MAbs B22.249 (Db) or Y3 (Kb) and analyzed by flow cytometry. Shadow: 2nd antibody control; solid line: empty vector control; dashed line: gp40 KKSQ-expressing cells. Drawn by the author.

84 Manuscript

Figure 3-10 gp40 interacts with PLC components but not human MHC class I. A: Schematic representation of the re-immunoprecipitation protocol. B: The Db/Kb serum does not cross-react with gp40. HeLa cells transduced with gp40 or the empty vector only and K41 expressing gp40 were radiolabelled for 10 minutes, lysed in 1% TX-100 buffer and immunoprecipitated with Db/Kb serum. C: Gp40 binds to tapasin, calnexin, and calreticulin. K41 cells expressing gp40 were labeled for 30 minutes, lysed in 1% digitonin buffer, and proteins were immunoprecipitated with antisera against calnexin, calreticulin, and tapasin. After denaturation in 1% SDS, calnexin, calreticulin and tapasin were reprecipitated with their respective antisera, or gp40 was reprecipitated with anti-HA MAb. Proteins were PNGase digested and separated by SDS- PAGE. Drawn by the author.

85 Manuscript

3.8 References

Allen, H., Wraith, D., Pala, P., Askonas, B., and Flavell, R.A. (1984). Domain interactions of H-2 class I antigens alter cytotoxic T-cell recognition sites. Nature 309, 279-281.

Ambagala, A.P., Solheim, J.C., and Srikumaran, S. (2005). Viral interference with MHC class I antigen presentation pathway: the battle continues. Vet Immunol Immunopathol 107, 1-15.

Basta, S., and Bennink, J.R. (2003). A survival game of hide and seek: cytomegaloviruses and MHC class I antigen presentation pathways. Viral Immunol 16, 231-242.

Bennett, E.M., Bennink, J.R., Yewdell, J.W., and Brodsky, F.M. (1999). Cutting edge: adenovirus E19 has two mechanisms for affecting class I MHC expression. J Immunol 162, 5049-5052. del Val, M., Hengel, H., Häcker, H., Hartlaub, U., Ruppert, T., Lucin, P., and Koszinowski, U.H. (1992). Cytomegalovirus prevents antigen presentation by blocking the transport of peptide-loaded major histocompatibility complex class I molecules into the medial-Golgi compartment. The Journal of Experimental Medicine 176, 729-738.

Fritzsche, S., Abualrous, E.T., Borchert, B., Momburg, F., and Springer, S. (2015). Release from Endoplasmic Reticulum Matrix Proteins Controls Cell Surface Transport of MHC Class I Molecules. Traffic.

Fritzsche, S., and Springer, S. (2013). Investigating MHC class I folding and trafficking with pulse-chase experiments. Mol Immunol 55, 126-130.

Fritzsche, S., and Springer, S. (2014). Pulse-chase analysis for studying protein synthesis and maturation. Curr Protoc Protein Sci 78, 30 33 31-30 33 23.

Furman, M.H., Dey, N., Tortorella, D., and Ploegh, H.L. (2002). The human cytomegalovirus US10 gene product delays trafficking of major histocompatibility complex class I molecules. Journal of Virology 76, 11753-11756.

Gao, B., Adhikari, R., Howarth, M., Nakamura, K., Gold, M.C., Hill, A.B., Knee, R., Michalak, M., and Elliott, T. (2002). Assembly and antigen-presenting function of MHC class I molecules in cells lacking the ER chaperone calreticulin. Immunity 16, 99-109.

Garstka, M.A., Fritzsche, S., Lenart, I., Hein, Z., Jankevicius, G., Boyle, L.H., Elliott, T., Trowsdale, J., Antoniou, A.N., Zacharias, M., et al. (2011). Tapasin dependence of major histocompatibility complex class I molecules correlates with their conformational flexibility. FASEB J 25, 3989-3998.

Grandea, A.G., Golovina, T.N., Hamilton, S.E., Sriram, V., Spies, T., Brutkiewicz, R.R., Harty, J.T., Eisenlohr, L.C., and Van Kaer, L. (2000). Impaired assembly yet normal trafficking of MHC class I molecules in Tapasin mutant mice. Immunity 13, 213-222.

Halenius, A., Hauka, S., Dolken, L., Stindt, J., Reinhard, H., Wiek, C., Hanenberg, H., Koszinowski, U.H., Momburg, F., and Hengel, H. (2011). Human cytomegalovirus disrupts the major histocompatibility complex class I peptide-loading complex and inhibits tapasin gene transcription. J Virol 85, 3473-3485.

86 Manuscript

Hammerling, G.J., Rusch, E., Tada, N., Kimura, S., and Hammerling, U. (1982). Localization of allodeterminants on H-2Kb antigens determined with monoclonal antibodies and H-2 mutant mice. Proc Natl Acad Sci U S A 79, 4737-4741.

Hein, Z., Uchtenhagen, H., Abualrous, E.T., Saini, S.K., Janssen, L., Van Hateren, A., Wiek, C., Hanenberg, H., Momburg, F., Achour, A., et al. (2014). Peptide-independent stabilization of MHC class I molecules breaches cellular quality control. J Cell Sci 127, 2885-2897.

Howe, C., Garstka, M., Al-Balushi, M., Ghanem, E., Antoniou, A.N., Fritzsche, S., Jankevicius, G., Kontouli, N., Schneeweiss, C., Williams, A., et al. (2009). Calreticulin-dependent recycling in the early secretory pathway mediates optimal peptide loading of MHC class I molecules. EMBO J 28, 3730-3744.

Jones, T.R., Wiertz, E.J., Sun, L., Fish, K.N., Nelson, J.A., and Ploegh, H.L. (1996). Human cytomegalovirus US3 impairs transport and maturation of major histocompatibility complex class I heavy chains. Proc Natl Acad Sci U S A 93, 11327-11333.

Kavanagh, D.G., Gold, M.C., Wagner, M., Koszinowski, U.H., and Hill, A.B. (2001). The multiple immune- evasion genes of murine cytomegalovirus are not redundant: m4 and m152 inhibit antigen presentation in a complementary and cooperative fashion. J Exp Med 194, 967-978.

Kraus, A., Groenendyk, J., Bedard, K., Baldwin, T.A., Krause, K.H., Dubois-Dauphin, M., Dyck, J., Rosenbaum, E.E., Korngut, L., Colley, N.J., et al. (2010). Calnexin deficiency leads to dysmyelination. J Biol Chem 285, 18928-18938.

Krmpotid, A., Busch, D.H., Bubid, I., Gebhardt, F., Hengel, H., Hasan, M., Scalzo, A.A., Koszinowski, U.H., and Jonjid, S. (2002). MCMV glycoprotein gp40 confers virus resistance to CD8+ T cells and NK cells in vivo. Nature Immunology 3, 529-535

Lemke, H., Hammerling, G.J., and Hammerling, U. (1979). Fine specificity analysis with monoclonal antibodies of antigens controlled by the major histocompatibility complex and by the Qa/TL region in mice. Immunol Rev 47, 175-206.

Ljunggren, H.G., Stam, N.J., Ohlen, C., Neefjes, J.J., Hoglund, P., Heemels, M.T., Bastin, J., Schumacher, T.N., Townsend, A., Karre, K., et al. (1990). Empty MHC class I molecules come out in the cold. Nature 346, 476-480.

Loch, S., and Tampe, R. (2005). Viral evasion of the MHC class I antigen-processing machinery. Pflugers Arch 451, 409-417.

Lu, X., Pinto, A.K., Kelly, A.M., Cho, K.S., and Hill, A.B. (2006). Murine cytomegalovirus interference with antigen presentation contributes to the inability of CD8 T cells to control virus in the salivary gland. J Virol 80, 4200-4202.

Momburg, F., and Hengel, H. (2002). Corking the bottleneck: the transporter associated with antigen processing as a target for immune subversion by viruses. Curr Top Microbiol Immunol 269, 57-74.

Nathenson, S.G., Kesari, K., Sheil, J.M., and Ajitkumar, P. (1989). Use of mutants to analyze regions on the H-2Kb molecule for interaction with immune receptors. Cold Spring Harb Symp Quant Biol 54 Pt 1, 521-528.

87 Manuscript

Nehls, S., Snapp, E.L., Cole, N.B., Zaal, K.J., Kenworthy, A.K., Roberts, T.H., Ellenberg, J., Presley, J.F., Siggia, E., and Lippincott-Schwartz, J. (2000). Dynamics and retention of misfolded proteins in native ER membranes. Nat Cell Biol 2, 288-295.

Niman, H.L., Houghten, R.A., Walker, L.E., Reisfeld, R.A., Wilson, I.A., Hogle, J.M., and Lerner, R.A. (1983). Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc Natl Acad Sci U S A 80, 4949-4953.

Ortmann, B., Androlewicz, M.J., and Cresswell, P. (1994). MHC class I/beta 2-microglobulin complexes associate with TAP transporters before peptide binding. Nature 368, 864-867.

Paulsson, K.M., Kleijmeer, M.J., Griffith, J., Jevon, M., Chen, S., Anderson, P.O., Sjogren, H.-O., Li, S., and Wang, P. (2002). Association of tapasin and COPI provides a mechanism for the retrograde transport of major histocompatibility complex (MHC) class I molecules from the Golgi complex to the endoplasmic reticulum. The Journal of Biological Chemistry 277, 18266-18271

Pinto, A.K., and Hill, A.B. (2005). Viral interference with antigen presentation to CD8+ T cells: lessons from cytomegalovirus. Viral Immunology 18, 434-444

Praveen, P.V., Yaneva, R., Kalbacher, H., and Springer, S. (2010). Tapasin edits peptides on MHC class I molecules by accelerating peptide exchange. Eur J Immunol 40, 214-224.

Solda, T., Galli, C., Kaufman, R.J., and Molinari, M. (2007). Substrate-specific requirements for UGT1- dependent release from calnexin. Mol Cell 27, 238-249.

Springer, S. (2015). Transport and quality control of MHC class I molecules in the early secretory pathway. Curr Opin Immunol 34, 83-90.

Wang, R., Natarajan, K., Revilleza, M.J., Boyd, L.F., Zhi, L., Zhao, H., Robinson, H., and Margulies, D.H. (2012). Structural basis of mouse cytomegalovirus m152/gp40 interaction with RAE1gamma reveals a paradigm for MHC/MHC interaction in immune evasion. Proc Natl Acad Sci U S A 109, E3578-3587.

Zhang, W., Wearsch, P.A., Zhu, Y., Leonhardt, R.M., and Cresswell, P. (2011). A role for UDP-glucose glycoprotein glucosyltransferase in expression and quality control of MHC class I molecules. Proc Natl Acad Sci U S A 108, 4956-4961.

Zhi, L., Mans, J., Paskow, M.J., Brown, P.H., Schuck, P., Jonjid, S., Natarajan, K., and Margulies, D.H. (2010). Direct interaction of the mouse cytomegalovirus m152/gp40 immunoevasin with RAE-1 isoforms. Biochemistry 49, 2443-2453

Ziegler, H., Muranyi, W., Burgert, H.G., Kremmer, E., and Koszinowski, U.H. (2000). The luminal part of the murine cytomegalovirus glycoprotein gp40 catalyzes the retention of MHC class I molecules. The EMBO Journal 19, 870-881.

Ziegler, H., Thale, R., Lucin, P., Muranyi, W., Flohr, T., Hengel, H., Farrell, H., Rawlinson, W., and Koszinowski, U.H. (1997). A mouse cytomegalovirus glycoprotein retains MHC class I complexes in the ERGIC/cis-Golgi compartments. Immunity 6, 57-66.

88 Manuscript

3.9 Additional data on MHC class I interaction with gp40 and other proteins

3.9.1 MHC class I binds weakly to gp40 linker mutants and the complex is exported together We found that gp40 binds to MHC class I molecules, and since gp40 is itself retained in the early secretory pathway itself, it prevents the export of MHC class I molecules as well. We also identified a non-functional mutant of gp40, gp40-(G4S)9, which fails to reduce MHC class I surface expression. The mutation does not abolish the interaction of gp40 with MHC class I, but it changes the export dynamics of gp40 itself. In contrast to the wild type, the gp40-(G4S)9 exits the early secretory pathway rapidly.

Figure 3-11 Only little gp40-Gp40-LDel binds to MHC class I and is exported together with it 4*106 K41 cells per lane expressing either gp40 wt or gp40-LDel were labeled with 35S for ten minutes and chased for the indicated times. Samples were lysed in 1% digitonin and precipitated with either HA antibody (panel A) or Db/Kb antiserum and reprecipitated with HA antibody (panel B and C). The samples were denatured, digested with EndoF1 and resolved on an 11% SDS-PAGE gel. Drawn by the author.

gp40 wt and MHC class I remain bound together for a relatively long time, showing a similar partial EndoF1 resistance pattern (compare Figure 3-4B and D). To find out how the complex of MHC class I and the gp40 linker mutant travels, we followed the maturation pattern of the gp40-LDel mutant8 bound to Db and Kb. The signal of the bound gp40-LDel is much weaker than that of wild type gp40, but the

8 gp40-LDel lacks the entire linker sequence (seeFigure 3-5 A). it has the same phenotype as gp40-(G4S)9with respect to function and export.

89 Manuscript maturation pattern looks very similar to what we observe in a direct immunoprecipitation (see Figure 3-11, panel A).

The gp40-LDel/MHC class I complex appears stable even outside the early secretory pathway, since it is detected after up to four hours of chase. After this time, the signal of gp40-LDel in the direct IP (compare four hours with eight hours) has almost disappeared, suggesting degradation of gp40-LDel, probably in the lysosomes. The disappearance of the gp40-LDel signal in the reIP after four hours therefore indicates degradation of the complex rather than dissociation. We believe that the weak signal of the gp40-LDel bound to MHC class I compared to gp40 wt (see Figure 3-5E) is due to the faster export of gp40-LDel, which leaves it less time to interact with MHC class I. In conclusive support of this interpretation, we showed that when gp40-LDel or gp40-(G4S)9 are forced to remain in the ER for an extended time by treatment with BFA, the interaction signal becomes as strong as that with gp40 wt (V. Ramnarayan, unpublished observations). Our results indicate that the linker region is responsible for the retention of gp40 in the early secretory pathway, but not for its interaction with MHC class I molecules.

90 Manuscript

Figure 3-12 Only little gp40-(G4S)9 binds to MHC class I and is exported together with it 6 35 4*10 K41 cells per lane expressing either gp40 wt or gp40-(G4S)9 were labeled with S for ten minutes and chased for the indicated times. Samples were lysed in 1% digitonin and precipitated with either HA antibody and reprecipitated with HA antibody (panel A) or Db/Kb antiserum and reprecipitated with HA antibody (panel B and C). The samples were denatured, digested with EndoF1 and resolved on an 11% SDS-PAGE gel. The arrow shows the sialylated gp40 band. Drawn by the author.

The same result is observed in the case of gp40-(G4S)9 (Figure 3-12). The signal for the bound gp40-(G4S)9 is even weaker than that of gp40-LDel (Figure 3-11). The reason for this is probably experimental variability rather than a real difference between gp40-LDel and gp40-(G4S)9.

91 Manuscript

Figure 3-13 MHC class I binds weakly to gp40-(G4S)9 and is exported together with it 6 35 4*10 K41 cells per lane expressing either gp40 wt or gp40-(G4S)9 were labeled with S for ten minutes and chased for the indicated times. Samples were lysed in 1% digitonin and precipitated with either B22, Y3, HA antibody or Db/Kb antiserum and reprecipitated with Db/Kb antiserum as indicated. The samples were denatured, digested with EndoF1, and resolved on an 11% SDS-PAGE gel. Drawn by the author.

When the reIP is done in the reverse fashion (Figure 3-13 panel D and E), we observe – as expected –

b b that the D and K bound to gp40-(G4S)9 become sialylated in a similar fashion as they do in the direct IP (Figure 3-13 panel C), without apparent partial EndoF1 resistance, as observed in the case of gp40 wt.

We conclude that both gp40 linker mutants (LDel or (G4S)9) can bind to MHC class I, and that the complex is exported out of the early secretory pathway and most likely degraded in the lysosomes.

92 Manuscript

3.9.2 Prolonged interaction of MHC class I with the PLC is specific to wild type gp40

The presence of gp40 does not prevent the incorporation of MHC class I molecules into the PLC (Figure 3-2D). To the contrary, the interaction persists even longer than in the control cells. This is most noticeable for Kb, which generally leaves the PLC quickly, but in gp40 wt-expressing cells, the Kb signal disappears much more slowly, and after four hours it is still relatively strong (Figure 3-14). We assume that the increased local concentration of MHC class I due to gp40 retention leads to this effect. Eventually, the molecules dissociate from the PLC, and therefore, the reduced MHC class I surface levels are not caused by a stronger interaction with the PLC. Another effect of gp40 expression is that the interaction of Db with the PLC is maximal already at 0 min chase, whereas in control cells, the signal becomes much stronger after 30 min chase.

Db needs more time to fold properly than Kb (Fritzsche et al., 2015), and this is probably why it is not immediately available for incorporation into the PLC. Why these dynamics change in gp40-expressing cells is so far unknown. We hypothesize that gp40 is probably much more abundant in the cell than MHC class I molecules due to the expression from a viral promoter. As a glycoprotein with an MHC class I-like structure, it probably binds to the same folding chaperones as Db does, such as calnexin and calreticulin. This might make these chaperones less available for Db, which then binds to the PLC earlier.

To test this hypothesis, we repeated the PLC interaction experiment with the non-functional gp40 linker mutant gp40-LDel. This mutant has been shown to fold and export well, and we assume it uses the same set of chaperones as gp40 does.

The interaction dynamics of Db and Kb with the PLC in gp40-LDel-expressing cells is identical to control cells (Figure 3-14). Therefore, it seems that the observed phenotype really is due to gp40 function and not an artifact of glycoprotein overexpression.

Figure 3-14 Prolonged interaction of MHC class I with the PLC is specific to wild type gp40 2*106 K41 cells per lane expressing either the empty plasmid, gp40 wt, or gp40-LDel were labeled with 35S for ten minutes and chased for the indicated times. Samples were lysed in 1% digitonin, and proteins were precipitated with a tapasin antiserum and reprecipitated with Db/Kb antiserum. The samples were denatured, digested with PNGase, and resolved on an 11% SDS-PAGE gel. N = 3. Drawn by the author.

93 Manuscript

3.9.3 Identification of bands that coimmunoprecipitate with MHC class I

In the mild detergent digitonin, MHC class I-associated proteins can be co-immunoprecipitated. They show up as additional bands above or below the MHC class I band. Most of these bands are chaperones and parts of the PLC. To identify some of these bands, we compared the coIP bands with those from individually immunoprecipitated chaperones and PLC components (Figure 3-15).

Figure 3-15 Identification of bands that coimmunoprecipitate with MHC class I 2*106 K41 cells per lane were labeled with 35S for two hours. Samples were lysed in 1% digitonin or 1% TX-100 as indicated and precipitated with antisera against tapasin, Db/Kb, calnexin or TAP2, or with a monoclonal antibody against calreticulin, as indicated. 1 = calreticulin, 2 = tapasin, 3 = TAP2, and 4 = calnexin. The samples were denatured and resolved on a 10%-12% gradient SDS-PAGE gel. Drawn by the author.

All tested chaperones were found in Db/Kb coIPs: 1 = calreticulin, 2 = tapasin, 3 = TAP2, and 4 = calnexin (Figure 3-15). Even TAP2, which is an integral membrane protein, gives a strong signal. Both the calnexin and the calreticulin antibodies pull down a fuzzy band that runs between TAP2 and calnexin, but its identity is unknown. It is also faintly visible in the Db/Kb coIPs. Calreticulin, tapasin and calnexin additionally pull down a band that runs just below the undigested Db/Kb band, and in the Db/Kb coIPs it is also faintly visible underneath the strong Db/Kb signal. This band is unidentified as well. These proteins also co-precipitated in a TX-100 lysis buffer, and they reproduce in the other lanes, which indicated that they are true interaction partners with MHC class I or associated chaperones that are independent of gp40. A mass spectroscopy experiment might identify these bands.

94 Manuscript

3.9.4 Discussion

We have shown that gp40 binds to MHC class I molecules (Figure 3-3, Figure 3-4), most likely directly, i.e., without an intermediate protein. A co-immunoprecipitation using cell lysate cannot show if the observed interaction is indeed direct; for this, both interaction partners would have to be expressed as recombinant proteins and their interaction investigated.

We assume that the MHC class I molecules are retained intracellularly by this interaction with gp40, but to formally show this, one needs to construct a non-MHC-class-I-binding mutant of gp40 by point mutagenesis that still folds and is retained intracellularly. We do not currently have such a mutant.

We do not know in detail how the gp40-class I interaction behaves over the lifetime of the two proteins. We observe that the complex remains stable for up to four hours, but at four hours, we already see a decrease in signal of gp40 (Figure 3-4D). Does that mean that the complex dissociates then, or do the antibodies in IP simply bind more and more unlabeled MHC class I, which is bound to unlabeled gp40? We know that gp40 exits the early secretory pathway at some point in time, whereas MHC class I molecules are not exported for up to sixteen hours (Figure 3-1D). This can be explained by gp40 handing over its bound MHC class I to another gp40. It would be interesting to perform a long pulse chase with cycloheximide like the one Figure 8 in (Ziegler et al., 2000), to observe the interaction without fresh protein synthesis. Upon long incubation with cycloheximide, however, protein transport patterns in the cell might be unnatural.

How does gp40 actually travel? We observe partial EndoF1 resistance as well as sialylation (Figure 3-1E). Is the partially resistant form a precursor of the sialylated form, or does the gp40 population split in two forms, one that exits quickly (because of saturation of the retention factor) and one that cycles until it is degraded in the ER by ERAD? To address this question, one could inhibit ERAD degradation. If the partial resistant and the sialylated gp40 populations are two separate populations, one would expect that the partial resistant band become totally resistant (but not sialylated) over time, without loss of signal, while the sialylated band should disappear over time. If ERAD does not play a role in removal of gp40, the result should be similar to the untreated sample. Alternatively, one could inhibit lysosomal degradation. If the partial resistant gp40 population is degraded by ERAD, their signal should disappear, and the signal of the sialylated gp40 population should remain unchanged.

Another unsolved question: What percentage of MHC class I molecules is bound at steady state by gp40? The complexes we pull down in a coIP are perhaps only a small fraction of all complexes present, since

95 Manuscript others might dissociate during lysis or during the immunoprecipitation process. Since there is most likely sufficient gp40 present due to the overexpression from a viral promoter, it is possible that in the early secretory pathway, every MHC class I molecule is bound to a gp40 molecule at all times. But if that is so, then why can those molecules of Db and Kb that are retained in the ER be immunoprecipitated with B22 and Y3, even though binding of these antibodies is incompatible with gp40 binding (Figure 3-3C)? There are two simple explanations, which are mutually non-exclusive: Either, only a fraction of Db and Kb is actually bound to gp40 at any given time; or, the complex of class I and gp40 falls apart during cell lysis (although it is quite stable apparently (Figure 3-4))? A third, more far-fetched, explanation is that the IP with the Db/Kb serum might stabilize a conformation that binds better to gp40. It was observed before that antisera reactive against the cytoplasmic tail do sometimes bind a distinct conformation of MHC class I only (Capps and Zúñiga, 1993; Little et al., 1995; Smith et al., 1986). For now, I cannot think of a good experiment to distinguish between these hypotheses. gp40 and associated MHC class I molecules accumulate in the ERGIC. Is this due to static retention in the ERGIC or due to dynamic retrieval to the ER that leads to a steady state ERGIC accumulation? The fact that in the presence of gp40, Db is found both in the ER and in the ERGIC rather supports the retrieval hypothesis. Since Kb is exported faster from the ER than Db (Fritzsche et al., 2015), these are the subcellular localizations one would expect for these two allotypes in case of gp40-mediated ERGIC-to-ER retrieval. Live-cell microscopy at 20 °C, where retrograde transport out of Golgi and ERGIC to the ER is blocked, might solve this question (Lippincott-Schwartz et al., 1990; Saraste and Svensson, 1991).

With deletion constructs, we have identified a 43 amino acid sequence that is responsible for the retention of gp40 (Figure 3-5). Is this the recognition site for the cellular retention factor? If yes, is the entire sequence necessary for the retention, or is the actual sequence hidden in this linker sequence somewhere? Is the linker sufficient for retention? All these questions are very interesting and important to answer in the next steps of the project. Possible experiments are a linker transplant onto CD4, to see if it leads to retention of CD4, or swapping parts of the linker to pinpoint the responsible sequence. If we find that the whole sequence is necessary, we can try to determine whether the linker has a structure, and perhaps try to manipulate this structure by introducing prolines (to break helices) and cysteines (to rigidify by disulfide crosslinking).

Even more interesting than the question which part of the linker is recognized is the question of what cellular protein recognizes the linker. What is the retention factor that keeps gp40 inside? Is it a known protein or is it a so far undescribed protein? This is clearly an important question. To answer it,

96 Manuscript interaction studies followed by mass spectroscopy as well as a genetic screen in haploid cells both would be very helpful (Timms et al., 2013).

Does gp40 bind directly to tapasin? We do observe a strong interaction signal (Figure 3-10C), but since MHC class I also binds to tapasin (Figure 3-2D), it is likely that in any case, gp40 binds to tapasin via MHC class I. We might test whether gp40 binds to tapasin in MHC class I-negative cells, because we know we have interaction between them in MHC class I-expressing cells. If we cannot detect tapasin interaction with gp40, it is very likely that in wild type cells, gp40 binds to tapasin via MHC class I, most likely in the context of the PLC. Alternatively, one might use human cells, since we did not observe an interaction between gp40 and human MHC class I molecules (V. Ramnarayan, unpublished observations).

How does the expression of gp40 cause faster incorporation of Db into the PLC (3.9.2)? One possibility is that the interaction of gp40 with Db conformationally stabilizes a halfway folded Db and/or helps it to fold in the first place, thereby decreasing its folding time. This hypothesis is supported by the finding that gp40 most likely binds to the 1/2 superdomain (Figure 3-3C), which is the least tightly folded domain of class I (Garstka et al., 2011); and there is most likely plenty of gp40 present. To test this hypothesis, one might try to overexpress a chaperone such as calnexin or BiP to help Db fold faster and

b then observe the effect on PLC interaction with D . Alternatively, one could generate a gp40-(G4S)9-KKXX construct. This should, in contrast to gp40-(G4S)9, function like wild type gp40 does, and therefore also cause a faster incorporation of Db into the PLC.

Another explanation for the rapid incorporation of Db into the PLC is that the overexpressed gp40 occupies chaperones such as calnexin (3.9.2) that would normally bind to partially unfolded Db. As long as Db is in a complex with calnexin, it cannot enter the PLC. In Figure 3-14 we used the construct gp40-

(G4S)9 as a negative control, since it presumably binds to the same chaperones as wild type gp40 does. Still, it might not the appropriate control in this case, since it is exported more quickly than wild type gp40 and its local concentration is therefore probably much lower, which means it would bind to fewer chaperones in steady state than wild type gp40. To test this, one would have to use a non-binding gp40 mutant that is retained in the early secretory pathway. Interestingly, MHC class I molecules that are affected by gp40 seem to bind normally to the PLC, or even more tightly (Figure 3-2D and Figure 3-14). Do they need to dissociate from gp40 to be incorporated into the PLC? Or is it sterically possible that gp40 binds to PLC-bound MHC class I?

97 Manuscript

Does gp40 interact with MHC class I while class I is bound by the PLC? Probably yes, since the interaction of the PLC with MHC class I molecules is enhanced and prolonged in gp40-expressing cells, and gp40 can be precipitated with tapasin (see above). If gp40 indeed binds directly to MHC class I, it becomes very likely that in gp40-expressing cells, there is a supercomplex consisting of PLC, MHC class I molecules, and gp40. Is that sterically possible? There is, so far, no co-crystal of the PLC bound to MHC class I, only docked structures (Dong et al., 2009; Hulpke and Tampé, 2013; Turnquist et al., 2004). In the crowded environment of the PLC, where tapasin and calreticulin both bind to class I directly as well as to each other (via Erp57), there might not be much space left for gp40. It would probably have to bind to MHC class I from the top (Wang et al., 2012). This would match our immunoprecipitation data (Figure 3-3C), which suggest binding of gp40 to the 1/2 peptide binding superdomain of class I. But, if the complex is a little flexible, gp40 might also squeeze in from the side, possibly moving calreticulin a bit aside in the process. Who knows, perhaps this tightly packed supercomplex consisting of PLC, MHC class I, and gp40 is even more tightly associated than the wild type PLC without gp40. Though this would be very interesting to know, it is not the most urgent question. Still, some docking might be done with the existing structures, and it could tell us something about the spatial requirements of the proteins in this supercomplex.

Are gp40-retained MHC class I molecules loaded with better peptides than class I molecules in the absence of gp40? In the wild type situation, only about 50% of murine surface MHC class I molecules are optimally loaded (Su and Miller, 2001). Since gp40-retained MHC class I remains in the early secretory pathway longer than it normally does, it might have the chance to bind better peptides. Apart from just prolonging the stay in the early secretory pathway, gp40 might also, like tapasin, stabilize the binding groove and thereby help MHC class I bind and exchange peptide, perhaps as an unintended side effect of tight binding to the peptide-bound state of class I. One might check, in tapasin-deficient cells, how thermostability of MHC class I molecules change after introduction of gp40, as in tapasin-deficient cells the class I thermostability is low (Hein et al., 2014; Williams et al., 2002). To investigate the effect of prolonged stay in the ER only, without the presence of gp40, one could additionally express KKXX-tagged and untagged MHC class I in tapasin-deficient cells and compare their thermostability. If an extended stay in the ER indeed allows the binding of higher-affinity peptides, then the KKXX-tagged class I should be more thermostable at steady state.

When we first saw the gp40/MHC class I docked model (Wang et al., 2012), we imagined that gp40 needs its long linker to be able to fold over MHC class I, which would still stand in a 90° angle from the

98 Manuscript

membrane. The gp40-(G4S)9 construct, although the linker length remained the same, was unable to retain MHC class I, but it did still bind to it (Figure 3-5E). We also see that gp40-Ldel, which has no linker at all, still binds to MHC class I, albeit weakly (Figure 3-11). This interaction can probably enhanced with

BFA, since that technique worked for gp40-(G4S)9 (3.9.1; V. Ramnarayan, unpublished observations). What does that tell us about the orientation of MHC class I on membrane? If gp40 binds to MHC class I like the group of David Margulies suggest (Wang et al., 2012), then this can only work if MHC class I lies down, with the 1/2 peptide-binding superdomain almost touching the membrane, as suggested previously (Mitra et al., 2004). It would be very interesting to find out whether gp40-LDel coprecipitates with tapasin as gp40-wt does. We believe that gp40-wt interacts with tapasin via MHC class I. This means it should be able to bind to MHC class I which is incorporated in the PLC, which probably needs to stand perpendicular to the membrane, according to the published model (Dong et al., 2009). An interaction of gp40-LDel with tapasin would therefore indicate that either gp40 binds to tapasin without the help of MHC class I, or gp40-LDel can bind to MHC class I even when it is incorporated into the PLC.

Why did MCMV evolve gp40? Binding to the usually quite variable binding groove of MHC class I seems to unnecessarily narrow down the number of targets for gp40; and indeed, the effect of gp40 greatly varies between murine allotypes (Figure 3-1A - C). Normally, immunoevasins bind at conserved patches, often contacting α3 and β2m in a manner similar to tapasin (proposed) or Ly49; this is seen nicely with HCMV US2 or cowpox virus CPXV203 (McCoy et al., 2012). With this strategy, the immunoevasin can even sometimes downregulate both classical and non-classical MHC class I molecules. gp40, in contrast, probably binds to the variable α1/α2 superdomain. It has, in contrast to the other immunoevasins, an additional target called RAE-1, which consists only of the α1 and α2 domain and does not bind to β2m (1.7.4.2). RAE-1 acts as an activating NK cell receptor ligand, and in the course of a cytomegalovirus infection, it might be even more important than MHC class I. Perhaps it is the prime target of gp40? If so, then the gp40 target interaction site might have evolved out of the necessity to retain RAE-1, and class I retention might come as an additional, but not essential, feature. It would be interesting to find out which other animals have NK cell ligands that look like RAE-1, and if their respective cytomegaloviruses have a protein similar to gp40.

Our results do not always coincide with the results of the Koszinowski group, but for most deviations, there is a good explanation (see Table 3-1).

99 Manuscript

Table 3-1 Comparison of our results with the results of Koszinowski and coworkers (Ziegler et al., 2000)

Observations by Koszinowski Explanation for the apparent Subject and colleagues Our observations discrepancy (Ziegler et al., 2000)

Fig. 2C: Gp40 mutant without 9 Gp40 linker TM/CT and missing 2/3 of the We believe the linker to be a retrieval as a linker is secreted into the Fig. 5B: Gp40 wt leaves the signal, which is probably not 100% retrieval/ supernatant. secretory pathway more efficient and which might work less well retention Gp40 mutant without TM/CT slowly than linker mutants. if gp40 is not anchored to the membrane signal with (almost) intact linker is also anymore. secreted, but more slowly. Fig. 5B: Gp40 wt shows partially The gp48-CT contains a lysosomal sorting Export EndoH-resistant bands after 2.5 Fig. 1E: After 2 hrs, gp40-wt motif (Reusch et al., 1999) , which might time/partial hrs chase. shows partial and some total be dominant over the linker sorting resistant Gp40Δ353-gp48ct (linker and TM EndoF1-resistant bands and motif and pull gp40 out of the ERGIC bands of intact, but CT = gp48) becomes little sialylation. resulting in EndoH resistance and gp40 totally EndoH resistant and sialylation. sialylated (but not degraded). Fig. 3A: Interaction of gp40 Fig. 6: No interaction of gp40 with MHC class I, although Maybe the signal of bound MHC class I is with MHC class I (i.e. no β m too weak to see the β m signal. The MHC Interaction of 2 the MHC class I signal can be 2 gp40 with signal) detected in coIP (antibody quite weak. class I signal itself is only visible after a MHC class I for both lumenal and cytosolic reIP, since MHC class I and gp40 run at domains). We have never checked for the same height. presence of the β2m band. We have not yet done microscopy after lysosome inhibition. I expect that some gp40 will still be found in the ERGIC, but Fig. 7: MHC class I is mostly seen Fig. 1B: Gp40 is almost this might also depend on the cell type Steady state in the ERGIC (and a little ER), exclusively in the ERGIC/cis- and expression method. localization gp40 (after inhibition of Golgi; MHC class I location In the previous publication without b of gp40 and lysosomes with leupeptin) is only allotype-dependent (K lysosome inhibition, both gp40 and MHC MHC class I seen in the lysosomes. No mainly cis-Golgi, Db cis-Golgi class I are shown in the ERGIC ((Ziegler et colocalization of gp40 and MHC. and some ER). al., 1997) (Fig. 6)). In the Ziegler publications, the different murine class I allotypes are not cleanly differentiated.

9TM = transmembrane domain; CT = cyplasmic tail

100 Manuscript

Observations by Koszinowski Explanation for the apparent Subject and colleagues Our observations discrepancy (Ziegler et al., 2000)

We observe a much longer half-life for gp40, more than two hours. Fig. 1CDE: Gp40 is exported Fig. 8A: Short gp40 half-life, In the same paper in Fig. 5, gp40 is not slowly, there is almost no complete degradation after 2 hrs. degraded after two hours, in contrast to degradation after 2 hrs. Fig. 8A (expression vaccinia and Gp40 is At the same time, MHC class I MHC class I becomes partially transfection, resp.), therefore the exported remains EndoH-sensitive for EndoF1-resistant over time. expression system might play a role. This faster than many hours, therefore the is not discussed by the authors. MHC class I, authors claim that MHC class I is Fig. 4: Interacting MHC class I MHC class I is retained even without gp40 being and gp40 both acquire partial Alternatively, the cells might have died retained present (which suggests a EndoF1 resistance. after a couple of hours under without gp40 permanent modification of class I Gp40 leaves the ER cycloheximide treatment, and this is why present by gp40). eventually, which means that MHC class I does not become EndoF1- resistant at the late time points. MHC class I is probably handed over to a new gp40. Another possibility is that the cycloheximide did not work as expected and gp40 was present, just not visible. The authors claim, as in Fig. 8, that constant interaction between gp40 and MHC class I is not required for class I retention. Fig. 9: MHC class I and gp40 first Optical pulse We have not done an 'optical colocalize, and after 24 hrs, MHC Their microscopy shows the positive chase: MHC pulse chase' yet, but we colocalization with markers, but there is class I colocalizes with p58 but class I and believe that for retention of gp40 colocalizes with DAMP, a possibility that in these cells after gp40 go MHC class I, gp40 needs to be many hours of cycloheximide treatment, therefore gp40 does not need to separate in the same place. be close to MHC class I to retain the organelles rearrange and different ways markers stain the same organelle. It it. would be helpful to see, that gp40 and MHC class I do not colocalize with the markers of an organelle where they are not supposed to be. Fig. 1B: We have never done BFA Fig. 10: After BFA treatment, a BFA treatment in This fits with our findings that gp40- treatment: some MHC class I relocates to the combination with retained MHC class I cycles between ER MHC class I ER; after washout of BFA, MHC microscopy, but at steady and ERGIC/cis-Golgi and is therefore relocates to class I is mainly found back in the state, MHC class I is found in found in all compartments of the early ERGIC ERGIC/cis-Golgi region. the ER and the cis- secretory pathway. Golgi/ERGIC region.

101 Transcription of APM genes in gp40-expressing cells

4 Transcription of APM genes in gp40-expressing cells

In gp40-expressing cells, MHC class I molecules and gp40 accumulate in the ERGIC (Figure 3-1B). We wondered whether these cells react with an ER-specific stress response to this accumulation.

4.1 Transcription of ER stress markers after tunicamycin treatment

ER stress is caused by an abundance of unfolded proteins in the ER, which might lead to aggregation and impairment of the secretory protein export (see 1.4). Several proteins are upregulated as a consequence, such as EDEM, WARS and P58IPK (Gupta et al., 2010). We decided to use quantitative PCR to investigate the expression of these ER stress markers.

To verify that this method does indeed detect ER stress, we treated MEF CNX wt cells with tunicamycin for 12 hours. This drug prevents the glycosylation of secretory proteins and results in their accumulation in the early secretory pathway and subsequent ER stress (Hung et al., 2004). Afterwards, we lysed the cells, isolated the mRNA, reverse transcribed it into cDNA, and performed quantitative PCR with primers targeted at the ER stress marker genes.

Table 4-1 ER stress markers are enhanced after treatment with ER stressor tunicamycin The table shows the fold increase of the investigated transcripts in tunicamycin treated cells as determined by quantitative PCR. The increase ranges from 7.75 fold to almost 17 fold. N = 1. Transcript EDEM (a) EDEM (b) WARS P58IPK relative mRNA expression 9.51 9.03 7.75 16.97 Tunicamycin treated:untreated

The treatment with tunicamycin led to a substantial increase in ER stress marker transcripts, as described before (Gupta et al., 2010) (Table 4-1), up to a 17-fold increase in the case of P58IPK. Therefore, the quantitative PCR detects the transcriptional changes caused by the tunicamycin treatment.

102 Transcription of APM genes in gp40-expressing cells

This experiment was done only once, in MEF CNX wt cells. To confirm this result, it needs to be repeated, ideally including another cell line (for example K41). It would also be interesting to experiment with different ER stress indicators such as induction of CHOP, spliced XBP1, or phosphorylated eIF-2α, as well as other stressors such as thapsigargin or BFA (Gupta et al., 2010), to reach to a better understanding about the dynamics and the dimension of ER stress in our cell lines.

4.2 Does the expression of gp40 cause ER stress?

After we confirmed that the quantitative PCR can be used to investigate ER stress response, we determined the expression of ER stress markers in gp40-expressing cells. We were wondering whether the expression of gp40 and the resulting accumulation of gp40 and MHC class I in the ERGIC might cause ER stress and lead to enhanced transcription of ER markers.

MEF CNX wt cells expressing either the empty vector or gp40 were lysed, the mRNA was isolated and reversely transcribed to cDNA, and a quantitative PCR was performed to determine the transcription of ER stress markers.

Table 4-2 ER stress markers are not enhanced in cells expressing gp40 The table shows the fold change of the investigated transcripts in control cells compared to gp40-expressing cells as determined by quantitative PCR. The transcription is generally slightly decreased in gp40-expressing cells, down to 0.43 in gp40-expressing cells. N = 1. Transcript BiP WARS10 P58IPK11 EDEM12 relative mRNA expression 0.67 0.77 0.79 0.43 gp40 cells:control cells

In contrast to the treatment with tunicamycin, gp40 expression does not lead to an increase in the expression of ER stress markers. Rather in the opposite, most transcripts are moderately decreased, and in the case of EDEM, the amount of transcript was reduced by more than half. As this experiment has been done only once, it needs to be repeated, also in a different cell line (K41). This would tell us how reproducible the reduction is, what its dimension really is, and whether it is cell line specific. While the

10tryptophanyl tRNA synthetase (WARS) (Gupta et al., 2010; Han et al., 2013; Samali et al., 2010) 11 inhibitor of the interferon-induced double-stranded RNA-activated protein kinase (P58IPK) (Gupta et al., 2010; Huizen et al., 2003; Samali et al., 2010) 12 ER degradation enhancing alpha-mannosidase-like protein (EDEM) (Hosokawa et al., 2001)

103 Transcription of APM genes in gp40-expressing cells induction of ER stress markers is well described, there is not much known about their reduction, so it is not easy or useful to speculate about the meaning until we know whether this effect is reproducible.

So far, there is no indication from the work of any group, including the data presented here, that the expression of gp40 leads to ER stress in the cells. We have also observed that gp40-expressing cells are, apart from the reduced MHC class I surface expression, undistinguishable from control cells according to shape, growth rate, and ER amount (judged by fluorescence microscopy).

4.3 Is APM transcription altered in gp40-expressing cells?

In gp40-expressing cells, MHC class I surface levels are drastically reduced, and gp40 and MHC class I molecules accumulate in the ERGIC. We have not yet observed any indication that this causes stress in the cell, but cells might still register a proteostatic imbalance via some unknown feedback system and try to repair this imbalance by other mechanisms than the ER stress response.

One method to increase the export of MHC class I molecules to the cell surface might be the induction of genes that belong to the antigen presentation machinery (APM; for example the PLC with tapasin,

TAP, ERp57, β2m, but also general chaperones such as calreticulin and calnexin and immunoproteasome subunits) (Leone et al., 2013).

We wondered whether in gp40-expressing cells the transcription of APM genes was increased to counteract the reduced surface transport of MHC class I molecules.

4.3.1 APM transcription in MEF CNX wt cells

We investigated the APM transcription in MEF CNX wt cells with the help of quantitative PCR. For this purpose, we lysed cells expressing the empty vector (control) or gp40-expressing cells, isolated the mRNA, reverse transcribed to cDNA, and performed quantitative PCR with primers that target APM genes.

104 Transcription of APM genes in gp40-expressing cells

Figure 4-1 Transcription of most APM genes is reduced in gp40-expressing MEF CNX wt cells Fold change of APM gene transcription in cells that express gp40 compared to control cells. Most transcripts are decreased in gp40-expressing cells, down to 0.09. The general chaperones calreticulin (CRT) and calnexin (CNX) are only slightly reduced, the fold change for tapasin (TPN) is unclear. N = 3, N = 1 (for calreticulin and calnexin). Drawn by the author.

Table 4-3 Transcription of most APM genes is reduced in gp40-expressing MEF CNX wt cells This table shows the data that are depicted in Figure 4-1.

b b Transcript CRT CNX TPN β2m D K TAP1 TAP2 LMP2 LMP7 relative mRNA expression 0.82 0.88 1.01 0.37 0.22 0.30 0.19 0.18 0.16 0.09 gp40 cells:control cells

In gp40-expressing MEF CNX wt cells, most investigated APM genes are transcribed less than in control cells, with the lowest value of 0.09 for LMP7 (see Figure 4-1 and Table 4-3). Calreticulin and calnexin appear only slightly reduced with 0.82 and 0.88, respectively, but these transcripts were investigated only once. The result for tapasin remains unclear: in two of three experiments the transcription was reduced, the average resulting in 0.44, while in the third experiment the transcription was 1.58 fold higher. Perhaps the sample was inadvertently exchanged with another? The error bars of the other samples of three independent experiments show that in general the data are very reproducible. Therefore, it is most likely that the transcription of tapasin is also reduced. All other transcripts have fold values of around 0.4 to 0.1, which means the transcription is substantially reduced in these cells.

105 Transcription of APM genes in gp40-expressing cells

Since gp40 is a transmembrane protein with the major domain in the ER lumen, there is no obvious way in which it might directly influence the transcription of the APM genes. Transcription factors are usually soluble cytosolic proteins that can cycle between the cytosol and the nucleus, and factors that can modulate them usually work in the cytoplasm.

To investigate whether this effect is generally seen in gp40-expressing cells, we decided to include another murine cell line, K41.

4.3.2 APM transcription in K41 cells

Expression of gp40 in MEF CNX wt cells resulted in the unexpected reduction of APM transcription (4.3.1). We decided to repeat the experiment of 4.3.1 in another cell line called K41, which is a murine fibroblast cell line as MEF CNX wt.

Figure 4-2 Transcription of APM genes is slightly increased in gp40-expressing K41 cells Fold change of APM gene transcription in cells that express gp40 compared to control cells. All transcripts are increased in gp40-expressing cells, up to 2.43. The highest transcription was observed for TAP1 and TAP2. Only Db appears unchanged. N = 3. Drawn by the author.

106 Transcription of APM genes in gp40-expressing cells

Table 4-4 Transcription of APM genes is slightly increased in gp40-expressing K41 cells This table shows the data that are depicted in Figure 4-2.

b b Transcript TPN β2m D K TAP1 TAP2 LMP2 LMP7 relative mRNA expression 1.14 1.42 1.06 1.53 2.38 2.43 1.79 1.69 gp40 cells:control cells

Surprisingly, the transcription pattern in K41 cells is the opposite of that in the MEF CNX wt cells (see Figure 4-2 and Table 4-4). All tested transcripts are increased in gp40-expressing cells compared to control cells. The greatest increase was observed for TAP1 and TAP2, which was around 2.4 fold higher. Most other transcripts have values of around 1.4-1.8 fold, only tapasin and Db did not increase remarkably.

Apparently, the observed effects of gp40 on APM transcription are primarily determined by the identity of the cell line. This is surprising in the case of MEF CNX wt and K41, since they are both fibroblast from BL/6 mice, both immortalized with adenovirus transformation ((Garbi et al., 2006) and personal communication). We observed before in K41 that the level of confluency when culturing the cells modulates the MHC class I surface levels (see 4.4.7.3), a phenomenon which also might alther the APM transcription. We have not yet carefully investigated if this also occurs in MEF CNX wt cells. The confluency at which the cells were grown for mRNA isolation (around 80%), might have led to a higher overall APM transcription, which masked the gp40-mediated reduction.

It would be interesting to repeat this experiment in more cell lines, maybe also in a human cell line, to investigate APM transcription there. Also, one should determine APM transcription in K41 at different confluency levels. To find out if this effect is really gp40-related, one should try a non-functional gp40.

In these experiments (as in all other experiments as well), the control cells were K41 cells transduced with the empty vector, to exclude the possibility that we were investigating artefacts that are caused by the transduction event and not by our gene of interest. We have not observed that the transduction itself has an effect on cell phenotype, growth levels, MHC class I surface levels, or MHC class I maturation. We did not investigate whether in general, the transcription of APM genes is changed in transduced cells compared to untransduced cells.

107 Transcription of APM genes in gp40-expressing cells

4.4 Treatment with IFN-α and IFN-γ

Treatment of mammalian cells with IFN-α and IFN-γ results in increased transcription of MHC class I molecules and APM genes (see 1.1.4). Does this treatment rescue the MHC class I surface expression in gp40-expressing cells? Do the increased levels of folded MHC class I molecules out-titrate gp40? Or does gp40 block some kind of bottleneck in the MHC class I export pathway, such that the ratio of MHC class I molecules over gp40 proteins does not matter?

4.4.1 Surface class I levels in MEF CNX wt cells after IFN treatment

We investigated the effect of IFN-α and IFN-γ treatment on the MHC class I surface expression in gp40- expressing MEF CNX wt cells. Cells were treated with 400 U/mL IFN-α and 200 U/mL IFN-γ for 48 hours, and where then surface stained for Db and Kb.

Table 4-5 MFI of Kb and Db surface expression in MEF CNX wt cells treated with IFN-α and IFN-γ The treatment with IFN-α and IFN-γ partially rescues MHC class I surface expression.

IFN treated: gp40 cells: raw MFI untreated control cells

Kb Db Kb Db Kb Db empty untreated 1.23 2.55 1 1 vector IFN-α 3.75 4.16 3.05 1.63 control IFN-γ 8.03 8.87 6.53 3.48 untreated 0.06 0.17 1 1 0.05 0.07 gp40 IFN-α 2.43 2.45 40.5 14.41 0.65 0.59 IFN-γ 6.68 7.65 111.33 45 0.83 0.86

The treatment with IFN-α and IFN-γ does indeed rescue the MHC class I surface expression, to around 60% of the control with IFN-α, and to around 80% of the control with IFN-γ (see Table 4-5, right column). The gp40 effect is still visible, as the surface levels do not reach 100% of the control, so gp40 is still functional to some extent. The fold increase in MHC class I surface expression in gp40-expressing cells after IFN treatment is very large (more than 100-fold, see Table 4-5, middle column), and does not mirror the comparably mild increase in the control cells (up to 6.5-fold, see Table 4-5, middle column). It does not look as if IFN-α and IFN-γ compete with gp40, but rather as if gp40 was suddenly switched off

108 Transcription of APM genes in gp40-expressing cells in the cells. This experiment has been done only once, repeats are necessary to make a detailed statement about the change in MHC class I surface expression.

4.4.2 Surface class I levels in K41 cells after IFN treatment

We have seen before that the effects of gp40 expression can vary considerably between different cell lines. Therefore, we included the K41 cell line in the IFN-α and IFN-γ experiment. K41 cells were treated with 400 U/mL IFN-α and 200 U/mL IFN-γ for 48 hours and then surface stained for Db/Kb.

Table 4-6 MFI of Kb and Db surface expression in K41 cells treated with IFN-α and IFN-γ The treatment with IFN-α and IFN-γ partially rescues MHC class I surface expression.

IFN treated: gp40 cells: raw MFI untreated control cells untreated 1.99 1 empty vector IFN-α 3.95 1.98 control IFN-γ 5.81 2.92 untreated 0.18 1 0.09 gp40 IFN-α 3.77 20.94 0.95 IFN-γ 4.38 24.33 0.75

The MHC class I molecules on the cell surface where stained with an antibody that recognizes both Db and Kb (in contrast to the previous experiment, where Db and Kb where stained individually, see Table 4-5).

As seen before with MEF CNX wt cells, the IFN treatment rescues MHC class I surface expression (see Table 4-6). With IFN-α, class I surface levels reach almost 100% of the control levels (see Table 4-6, right column). With IFN-γ, the levels are a bit lower (75%), but this might be due to experimental variation, because this experiment has been done only once. As with the MEF CNX wt cells before, the class I surface levels of the gp40-expressing cells remain a bit lower even after IFN treatment, meaning that at least some gp40 is still functional. The fold change of treated over untreated cells is around 2 (IFN-α) and 3 (IFN-γ) in control cells, and 21 (IFN-α) and 24 (IFN-γ) in gp40-expressing cells. Although the dimensions are different (which might be due to the cell line and antibody), the observation that gp40 function seems to be switched off after IFN treatment is also true in K41 cells.

109 Transcription of APM genes in gp40-expressing cells

4.4.3 APM transcription in MEF CNX wt cells after IFN treatment

IFN increases MHC class I surface levels through increased transcription of MHC class I and APM genes. We observed before that gp40 expression reduces APM transcription in one but not in another cell line (compare 4.3.1 and 4.3.2), and that IFN treatment restored MHC class I surface expression in both cell lines expressing gp40 (4.4.1, 4.4.2). To understand how IFN treatment changes the APM transcription in gp40-expressing MEF CNX wt cells, we treated these cells with 400 U/mL IFN-α and 200 U/mL IFN-γ for 48 hours, isolated mRNA, and performed quantitative PCR.

Table 4-7 APM gene transcription in gp40-expressing MEF CNX wt cells can be rescued by treatment with IFN-α and IFN-γ Treatment with IFN increases expression of all APM genes, also in gp40-expressing cells. Expression of transcripts relative to untreated control cells.

b TPN β2m K TAP1 TAP2

untreated 1 1 1 1 1 empty vector IFN-α 2.67 2.03 2.05 3.69 2.43 control IFN-γ 6.95 2.70 6.48 17.25 8.44 untreated 1.58 0.44 0.33 0.28 0.22 gp40 IFN-α 2.71 1.17 1.08 1.88 0.57 IFN-γ 6.57 2.70 5.21 11.58 5.37

LMP2 LMP7 LMP10 ERAAP13 PA28alpha PA28beta14

untreated 1 1 1 1 1 1 empty vector IFN-α 2.78 3.02 1.17 1.86 1.17 1.62 control IFN-γ 37.39 15.91 4.05 3.94 2.32 3.57 untreated 0.19 0.08 0.52 0.85 0.44 0.58 gp40 IFN-α 1.12 0.54 1.05 1.38 0.84 1.06 IFN-γ 33.14 9.33 4.54 3.36 1.96 3.57

13 ERAAP is an ER amino peptidase, trims antigenic peptides, is IFN inducible (Serwold et al., 2002) 14 PA28alpha and PA28beta are IFN-inducible proteasome regulators (Schwarz et al., 2000)

110 Transcription of APM genes in gp40-expressing cells

Table 4-8 Constitutively expressed proteasome subunits are not increased by treatment with IFN Treatment with IFN reduces expression of non-APM genes in control cells but leaves them largely unchanged in gp40-expressing cells. Expression of transcripts relative to untreated control cells.

PSX PSY PSZ15

untreated 1 1 1 empty vector IFN-α 0.6 0.67 0.95 control IFN-γ 0.45 0.50 0.67 untreated 0.3 0.42 0.44 gp40 IFN-α 0.36 0.45 0.54 IFN-γ 0.36 0.6 0.64

In MEF CNX wt cells, the IFN treatment restores the APM transcription almost to control levels (see Table 4-7).

4.4.4 APM transcription in K41 cells after treatment with IFN

The treatment with IFN can restore the APM transcription in MEF CNX wt cells (see 4.4.3). We performed the same experiment in K41 cells, where the APM transcription is not reduced by gp40 expression.

15 PSX, PSY and PSZ are constitutively expressed proteasome subunits; they are not IFN-inducible (Akiyama et al., 1994; Kasahara et al., 1996).

111 Transcription of APM genes in gp40-expressing cells

Table 4-9 APM gene transcription in gp40-expressing K41 cells is increased after treatment with IFN-α and IFN-γ Treatment with IFN increases expression of all APM genes, also in gp40-expressing cells. Expression of transcripts relative to untreated control cells.

b TPN β2m K TAP1 TAP2

untreated 1 1 1 1 1 empty vector IFN-α 0.90 1.14 1.08 1.57 1.05 control IFN-γ 1.37 1.97 1.69 3.64 2.41 untreated 0.96 1.08 1.49 1.42 1.2 gp40 IFN-α 0.69 1.50 1.96 2.18 1.24 IFN-γ 1.83 2.36 2.95 6.77 3.24

LMP2 LMP7 LMP10 ERAAP PA28alpha PA28beta

untreated 1 1 1 1 1 1 empty vector IFN-α 2.82 2.32 1.85 0.75 1.55 1.10 control IFN-γ 13.51 4.71 4.94 1.48 3.08 2.18 untreated 1.26 1.16 0.77 0.94 1.11 1.13 gp40 IFN-α 3.73 3.03 2.06 0.93 2.18 1.73 IFN-γ 23.23 6.51 5.31 1.53 3.67 2.78

Table 4-10 Constitutively expressed proteasome subunits and calnexin are not increased by treatment with IFN Treatment with IFN does not increase expression of non APM genes in both control and gp40 cells. Expression of transcripts relative to untreated control cells. PSX PSY PSZ CNX

untreated 1 1 1 1 empty vector IFN-α 0.72 1.48 0.70 0.36 control IFN-γ 0.82 1.37 0.87 0.5 untreated 1.42 1.34 1.07 1.13 gp40 IFN-α 1.01 2.06 0.98 0.40 IFN-γ 1.04 1.47 1.11 0.57

The expression of gp40 does not reduce the transcription of APM genes, instead, it increases them slightly, as we have seen before (see 4.3.2). Therefore, the induction of transcription after IFN treatment

112 Transcription of APM genes in gp40-expressing cells looks very similar in both control and gp40 cells (see Table 4-10). Calnexin transcription seems downregulated after IFN treatment, but unchanged after gp40 expression.

4.4.5 gp40 transcription in MEF CNX wt and K41 cells after IFN treatment

IFN treatment in gp40 cells restores the MHC class I surface levels back to almost normal (see Table 4-5 and Table 4-6), it almost seems as if the effect of gp40 is switched off. Does the expression of gp40 change after IFN treatment? To answer this question, gp40-expressing cells were treated with 400 U/mL IFN-α and 200 U/mL IFN-γ for 48 hours, their mRNA was isolated, and the transcription of gp40 analyzed with quantitative PCR.

Table 4-11 gp40 transcripts are strongly reduced after IFN treatment Treatment with IFN does strongly reduce gp40 transcription in both cell lines

MEF K41 untreated 1 1 IFN-α 0.25 0.06 IFN-γ 0.27 0.15

The quantitative PCR shows that gp40 transcription is drastically reduced in both cell lines after treatment with IFN (see Table 4-11). This explains why this treatment restores MHC class I surface levels in gp40-expressing cells. The transcription of genes with viral vectors is often reduced by cytokines such as IFN (see 4.4.7.4); we do not know whether this is due to reduced transcription or to reduced stability of the transcript. To investigate how gp40 behaves in IFN-treated cells, one needs to use a promoter that does not react to these cytokines (such as most cellular promoters, but not APM promoters). The most likely result of such an experiment is that gp40 can be out-titrated by large amounts of class I, since in MEF CNX wt cells, there are still considerable amounts of gp40 transcript (see Table 4-11), but most MHC class I still make it to the cell surface (see Table 4-5). It is therefore unlikely that gp40 blocks a bottleneck pathway of MHC class I or acts as an enzyme to covalently modify class I molecules, and we conclude that it most likely functions in a stochiometric fashion.

113 Transcription of APM genes in gp40-expressing cells

4.4.6 gp40 expression in RMA cells

We were unable to test the effect of IFN-α and IFN-γ on gp40 function, since they drastically reduce gp40 transcription (4.4.5). Therefore, we decided to investigate the effect of gp40 expression in the cell line RMA, which are murine B lymphocytes (genotype H-2b). As immune cells, these cells cannot be further induced by IFN-γ treatment (Boname et al., 2004), the transcription of APM genes is already maximal in these cells. We wondered whether gp40 can reduce MHC class I surface expression in these cells.

Gp40-expressing and control RMA cells were stained for surface Db and Kb and analyzed in the flow cytometer.

Figure 4-3 Surface Kb in RMA cells is not reduced by gp40 expression RMA cells expressing gp40-IRES-GFP were stained for surface Db and Kb. Cells were gated for GFP positive, the histograms show the signal intensity for MHC class I surface expression. Shadow: 2nd antibody control; green line: empty vector control; orange line: gp40. Drawn by the author.

Figure 4-4 Surface Kb in RMA cells is slightly reduced by very high gp40 levels RMA cells expressing gp40-IRES-GFP were stained for surface Db and Kb. GFP signal was divided into different gates, and the APC MFI (for surface MHC class I) was determined for each gate. Values were normalized once for the first value, and then against the values of the control cells. Blue line: Kb; pink line: Db. The 'kink' in both curves at MFI 1 is probably an experimental artefact; N=1. Drawn by the author.

114 Transcription of APM genes in gp40-expressing cells

In RMA cells, Kb appears quite resistant to gp40, whereas Db surface is reduced at higher levels of gp40. It seems that in these cells the MHC class I molecules are slightly more resistant to the function of gp40 than in the fibroblasts that were investigated before. This is difficult to compare though, and we do not know how much gp40 in relation to how much MHC class I is expressed in RMA cells or fibroblasts. This experiment needs to be repeated.

4.4.7 Discussion

4.4.7.1 ER stress

We wondered whether the accumulation of MHC class I and gp40 in the ERGIC might lead to ER stress in gp40-expressing cells. We found no indication for ER stress (Table 4-2), which probably means that the amount of accumulated MHC class I molecules and gp40 is but a small percentage of the total amount of protein in the ER, not enough to disturb the balance in the secretory pathway and to induce ER stress. In order to test this further, one could express a C-terminally KKXX-tagged form of MHC class I to see whether retention in the secretory pathway itself will lead to ER stress, but I believe that this is unlikely to be the case. One might also carefully measure the size of the ER to see if it has increased to deal with the larger protein load, but so far there is no indication that the gp40-expressing cells are stressed in any way. In a related line of reasoning, one reviewer wondered whether accumulation of class I and gp40 in the ERGIC would lead to relocalization of cytosolic COP I proteins to this compartment; this is also not expected to be the case for a single protein, and we could indeed find no sign of any change of COP I localization by microscopy (Rainer Duden and Irina Majoul, data not shown).

4.4.7.2 APM transcription after gp40 expression

We investigated the APM gene transcription in gp40-expressing cells and realized that the result depends a lot on the cell line used. In MEF CNX wt cells, the transcription of APM genes was drastically reduced (Figure 4-1), but in K41 it was unchanged or slightly increased (Figure 4-2).

The reduction of APM gene transcription in MEF CNX wt cells is not understood. It might be the result of a feedback mechanism, with which the cell notices reduced MHC class I export; anyways, teleologically speaking, this ought to lead to enhanced transcription and not to reduced transcription. Another possibility is that gp40 directly influences APM transcription. This is an exciting thought, but it is not

115 Transcription of APM genes in gp40-expressing cells immediately obvious how gp40 would do that. As a secretory transmembrane protein, it probably cannot interact with cytosolic transcription factors, especially because its cytosolic tail is very short. Still, it might be interesting to test a non-functional mutant of gp40 to find out whether this transcription effect is dependent on gp40 function or not. If yes, one could test deletion mutants of gp40 to figure out which is the responsible domain and maybe draw some conclusions about how this mechanism works. gp40 has an effect on APM transcription but not on the transcription of general chaperones like calnexin and calreticulin (see Figure 4-1). There is some indication that gp40 also reduces the amount of transcripts of constitutive proteasome subunits, PSX, PSY, and PSZ (see Table 4-8). It might be interesting to check transcription of other (non-APM-related) genes in gp40-expressing cells, to understand how they are affected.

To understand why this effect is so different in two similar cell lines, it is necessary to investigate how the APM transcription in other murine cell lines changes after gp40 expression. This might help to find out why K41 and MEF CNX wt cells regulate their APM expression differently in response to gp40. Also, especially if the non-functional gp40 turns out to be able to reduce transcription, it might be interesting to try it in a human cell line or an MHC class I-negative cell line.

4.4.7.3 Enhanced surface MHC class I in K41 grown at high confluency

We observed that K41 cells show increased MHC class I surface levels (and increased resistance to gp40) when grown to overconfluency (see Figure 4-5).

116 Transcription of APM genes in gp40-expressing cells

Figure 4-5 Overconfluency increases MHC class I surface expression in K41 cells Cells were grown at the indicated confluency, harvested, and surface stained against Kb and Db. A: MHC class I surface levels in control cells at 50% and 100% confluency. B: MHC class I surface levels in gp40-expressing cells at 50% and 100% confluency. C: Percentage of MHC class I expressing cells in gp40-expressing cells at 50% and 100% confluency. Db N = 4; Kb N = 6; error bars as SEM. Data by V. Ramnarayan, figure drawn by the author.

This is especially visible for Kb in gp40-expressing cells (see Figure 4-5B and C). We have observed this phenomenon only recently, and we do not know its mechanism at all. Perhaps MHC class I endocytosis is reduced, leading to higher steady state surface levels. Or else, class I expression and/or export are increased. We do not know whether the amount of MHC class I protein and related APM proteins is higher in these cells or not, and whether the transcription is changed. These experiments need to be done (MHC class I endocytosis, export, protein amount and transcription).

This phenomenon has been described before in several publications, but its molecular reason was never investigated: MHC class I surface levels are increased in overconfluent BALB/c -3T3 cells (Offermann and Faller, 1990), with autocrine induction of MHC class by IFN-β (Offermann and Faller, 1989), and also in B16 cells (Calorini et al., 1994). Another publication mentions an increase in human cells arrested at G1/S phase but does not show it (Erusalimsky et al., 1987). There are more observations concerning cell density, such as the one that surface MHC class I lateral mobility is impaired in dense cell cultures (Wier

117 Transcription of APM genes in gp40-expressing cells and Edidin, 1986), and that the internalization of LDL is reduced (Vlodavsky et al., 1978). Also, the expression of MHC class II genes in macrophages is cell cycle dependent (Xaus et al., 2000) and also regulated by cell density (Romieu-Mourez et al., 2007). The event of a transfection apparently also increases the MHC class I expression temporally (Park et al., 1998).

Perhaps the contact inhibition in dense cultures leads to signaling inside the cell which either directly increases MHC class I surface expression or leads to a release of cytokines, which modifies MHC class I surface levels.

This could be found out by growing a confluent culture and transferring the supernatant to a subconfluent culture. If a soluble cytokine is responsible for the induction, the subconfluent cells should also show an increase in MHC class I surface levels. The identification of the exact pathway might be difficult, and the pathway might also differ between different cell lines.

4.4.7.4 Reduction of transcripts from viral promoters after IFN treatment

We realized that treatment of gp40-expressing cells with IFN leads to a drastic reduction of gp40 mRNA (see Table 4-11).

IFN treatment activates transcription of APM genes, and reduces translation in general, through phosphorylation of eIF2α (Balachandran et al., 2000; Meurs et al., 1990; Panniers and Henshaw, 1983). This does not change the amount of transcript though, since it works at the level of translation, and does therefore not explain the reduction in gp40 transcript. gp40, which was transduced and therefore integrated into the host genome, is transcribed from an SFFV promoter, a promoter from the spleen focus-forming virus (Joyner et al., 1982). The host has evolved ways to recognize viral promoters and inhibit their transcription.

When investigating the gene expression of genes from a viral promoter (CMV and SV40) or the MHC class I promoter under IFN treatment, researchers found that protein production from viral vectors is impaired in many cell types (Harms and Splitter, 1995), but they did not determine the amount of mRNA.

In retroviral vectors, there might be an additional site that is targeted by IFN, in the 3’ LTR region (Ghazizadeh et al., 1997). The mRNA is destabilized by factors that are synthesized after IFN treatment. Therefore, it might not help to change the promoter, as long as one still uses a retroviral vector.

118 The influence of β2m on gp40 function

Interestingly, the authors found a reduction only by transduction and not when the same vector was introduced by transfection.

Methylation and chromosomal position of the introduced retroviral vector also seem to play a role (Hoeben et al., 1991). In comparison to the CMV and SFFV promoters, the EF1α or β-actin promoters seem to work better in the presence of cytokines (Perez-Galarza et al., 2014; Qin et al., 1997).

It would be interesting to figure out how our expression system reacts to IFN treatment. Is it only the promoter? Is a lentiviral vector recognized by other sequences, too? Does it matter whether the vector is transfected or transduced?

5 The influence of β2m on gp40 function

5.1 The effect of gp40 on human class I surface expression in HEK293T cells gp40 was reported to retain murine but not human MHC class I molecules (Ziegler et al., 1997). We decided to test this observation in our system, and we transduced gp40-IRES-GFP into HEK293T cells. The cells were then surface stained for human MHC class I molecules.

Figure 5-1 gp40 expression does not alter human MHC class I surface expression in HEK293T cells HEK293T cells were transduced with empty vector or gp40-IRES-GFP. Cells were surface stained for human MHC class I molecules, counted in the flow cytometer and gated for GFP positive. The MHC class I signal is shown. Shadow: 2nd antibody; green: empty vector; orange: gp40. N = 1. Data by M. Bottermann, figure drawn by the author.

The expression of gp40 in HEK293T cells does not change the MHC class I surface levels (see Figure 5-1). It seems indeed as if human MHC class I molecules are resistant to gp40 action. According to Gutermann

119 The influence of β2m on gp40 function

and Ziegler, all tested human MHC class I molecules were resistant (Gutermann, 2001; Ziegler et al.,

1997). From swapping experiments, the 1/2 domain of murine class I was reported to be responsible for gp40 susceptibility (Gutermann, 2001).

5.1.1 Murine class I molecules are not transported to the surface in HEK293T cells

Ziegler et al. also claimed that murine molecules in human cells can be retained by gp40 (Ziegler et al., 1997). We decided to repeat this experiment in our system. We thus transduced gp40-IRES-GFP into HEK293T and transfected these cells with Db-IRES-DsRed or Kb-IRES-DsRed. These cells were surface stained for either Db or Kb, with APC as fluorophore.

Figure 5-2 Murine MHC class I molecules are not expressed at the cell surface of HEK293T cells HEK293T cells were transduced with empty vector or gp40-IRES-GFP and transfected with either Kb or Db in pIRES2- DsRed-Express. Cells were surface stained for Db or Kb, counted in the flow cytometer, and gated for GFP-positive and DsRed-positive. The APC signal (surface MHC class I) is shown. Shadow: 2nd antibody; green: empty vector; orange: gp40. N = 10. Data by M. Bottermann, figure drawn by the author.

In HEK293T cells, the signal of surface Db and Kb is hardly higher than the background signal in both control and gp40-expressing cells (see Figure 5-2). The gp40 effect is hardly visible with these low MHC class I signals, whereas the endogenous human MHC class I molecules gave a good signal.

Apparently, the murine MHC class I molecules are unable to reach the cell surface of the human cells, or perhaps they are endocytosed after a short time at the cell surface.

If the endocytosis of murine class I in HEK cells is rapid, this would indicate that MHC class I molecules

form dimers with β2m but are suboptimally loaded, just like in PLC-deficient murine cells. One would expect that incubation at low temperatures or with high-affinity peptide could accumulate the murine

120 The influence of β2m on gp40 function

MHC class I molecules at the cell surface (Ljunggren et al., 1990; Montealegre et al., 2015), but we have not yet done that experiment.

If the export is impaired, what is the reason? One reason might be that in HEK293T cells, genes on a transfected vector are overexpressed because the SV40 large T antigen in these cells leads to a replication of expression vectors with an SV40 origin of replication (DuBridge et al., 1987). Possibly, the export machinery is overwhelmed with the amount of protein, and most of the murine MHC class I does not fold properly in the first place and cannot be exported. Another potential reason is insufficiency of

β2m. We have observed before that murine MHC class I molecules export better in Vero cells when β2m is cotransfected (Montealegre et al., 2015). It seems therefore that the most limiting factor for folding and export is β2m.

5.1.2 Murine class I molecules become resistant to gp40 with β2m overexpression

We observed that murine MHC class I molecules are hardly detectable on the cell surface when transfected into HEK293T cells (see Figure 5-2). We reasoned that cotransfection of β2m might improve folding and export of the exogenous MHC class I molecules and increase their cell surface levels, so that we are able to investigate the function of gp40 under these conditions. HEK293T cells expressing gp40-

b b IRES-GFP were cotransfected with either D -IRES-DsRed or K -IRES-DsRed and with hβ2m with the same amount of plasmid DNA. The following day, the cells were stained against surface Db or Kb with APC.

Figure 5-3 gp40 does not retain murine MHC class I in HEK293T overexpressing hβ2m HEK293T cells were transduced with empty vector or gp40-IRES-GFP and transfected with either Db-IRES-DsRed or b b b K -IRES-DsRed and with hβ2m. Cells were surface stained for D or K , counted in the flow cytometer, and gated for GFP positive and DsRed positive. The APC signal (surface MHC class I) is shown. Gray: 2nd antibody; green: empty vector; orange: gp40. N = 5. Data by M. Bottermann, figure drawn by the author.

After cotransfection with hβ2m, the murine MHC class I molecules showed very high surface levels, both in control and in gp40-expressing cells (see Figure 5-3). The additional β2m seemed to improve folding

121 The influence of β2m on gp40 function and export of MHC class I to a large degree, ten- to twentyfold higher than in the experiment without extra β2m. There was no indication of any gp40 function in these cells.

The fact that the gp40 effect was not visible in these cells might have two reasons: either, in our system gp40 cannot retain murine MHC class I molecules inside human cells; or, the overexpression of MHC class I molecules and hβ2m by transfection lead to such an abundance of folded MHC class I/β2m dimers that gp40 was simply overwhelmed, and its actions were masked. A reduction in the cotransfected β2m should reduce the amount of folded MHC class I molecules and should make the gp40 effect visible again.

5.1.3 Dose-response effect of hβ2m cotransfection on murine MHC class I surface levels in the presence of gp40

Cotransfection of murine MHC class I molecules with hβ2m leads to a massive export of MHC class I molecules to the surface of HEK293T cells, which might mask the gp40 effect. We thus wondered how the MHC class I surface levels would change if we titrated the hβ2m plasmid into the transfections. gp40- b b IRES-GFP expressing cells were cotransfected K -IRES-DsRed and with hβ2m. While the amount of K

DNA was kept constant (18 μg for a 10 cm dish), the amount of hβ2m plasmid was titrated to different ratios compared to the MHC class I DNA (ranging from a ratio from 0 to 0.8 (w/w)). Cells were surface stained for Kb and analyzed in a flow cytometer.

122 The influence of β2m on gp40 function

Figure 5-4 Murine MHC class I surface levels increase linearly with increasing cotransfected hβ2m plasmid HEK293T cells were transduced with empty vector or gp40-IRES-GFP and transfected Kb-IRES-DsRed and with b different amounts of hβ2m. Cells were surface stained for K , counted in the flow cytometer and gated for GFP positive and DsRed positive. A: The APC MFI signal (surface MHC class I) is shown. B: The percentage of cells that express detectable surface MHC class I is shown. N = 1. Data by M. Bottermann, figure drawn by the author.

In the titration experiment, one can clearly see that the surface levels of MHC class I depend in an almost linear fashion on the amount of β2m available (see Figure 5-4 A). We do not know whether further titration of β2m (over the ratio of 1:1, i.e., more β2m than MHC class I) increases the levels further, and where we would reach a plateau. The 1:1 data from the experiment before (see Figure 5-3) lie actually below the 0.8:1 data in this experiment (see Figure 5-4) (MFI 11 for Kb in Figure 5-3 and MFI 13 for Kb in Figure 5-4); due to the experimental variation, it is thus not possible to compare across experiments. When we take a look at the percentage of cells that express any amount of detectable surface MHC class I levels (see Figure 5-4B), we see that even small amounts of β2m plasmid results in surface MHC class I in almost all cells, while without β2m not only the levels are lower, but only around 85% of the cells express surface MHC class I molecules at all.

123 The influence of β2m on gp40 function

When looking at the data, it is also important to take into account the endogenous human MHC class I molecules that are present in the cells, which might have a higher affinity to the hβ2m. If the endogenous human β2m is not sufficient for the endogenous human MHC class I molecules, it might be that these profit first from the transfected hβ2m, and that the actual amount of β2m available for the murine MHC class I molecules is much lower than expected. The amount of PLC components does not change, so they might pose a natural export limit, especially when a lot of exogenous MHC class I is expressed. The Db and Kb surface molecules are very likely suboptimally loaded, and the quality of the peptides bound might improve or worsen with more β2m available.

We decided to continue with a ratio of MHC class I: β2m of 1:0.2 (see 5.1.4), as at this ratio, almost 100% of the cells expressed surface MHC class I in sufficient levels to detect the gp40 effect.

5.1.4 With little β2m available, murine MHC class I molecules are sensitive to gp40

The surface levels of murine MHC class I molecules in human HEK293T cells depend on the amount of

β2m available. Without extra β2m, there are no surface MHC class I molecules (see Figure 5-2); with plenty of β2m present, MHC class I molecules return to the surface and are rendered insensitive to gp40 at the same time (see Figure 5-3), probably by overwhelming gp40. It was reported before that murine MHC class I molecules expressed in human cells are retained by gp40 (Ziegler et al., 1997). This might be cell line-dependent and not be true for HEK293T cells. To find out whether gp40 is able to retain murine

MHC class I molecules in these cells, we decided to use a small amount of β2m, just enough to get it to the surface, but not enough to mask the gp40 function. HEK293T cells were transduced with gp40-IRES-

b b GFP, and cotransfected with either D -IRES-DsRed or K -IRES-DsRed and hβ2m in a ratio of 1:0.2. The cells were then stained for surface Db or Kb and analyzed in a flow cytometer.

Figure 5-5 Low levels of β2m restore the gp40 phenotype on murine MHC class I surface levels HEK293T cells were transduced with empty vector or gp40-IRES-GFP and transfected with either Db-IRES-DsRed or b b b K -IRES-DsRed and with hβ2m in a ratio 1:0.2. Cells were surface stained for D or K , counted in the flow cytometer and gated for GFP-positive and DsRed-positive. The APC signal (surface MHC class I) is shown. Gray: 2nd antibody; green: empty vector; orange: gp40. N = 1. Data by M. Bottermann, figure drawn by the author.

124 The influence of β2m on gp40 function

The addition of low levels of hβ2m renders the murine MHC class I molecules fully gp40-sensitive again (see Figure 5-5). This tells us, as observed before, that the mechanism of gp40 retention is somehow connected with the levels or the conformational states of MHC class I molecules themselves, and so we think that it is unlikely that gp40 causes a general block in the secretory pathway. This result also indicates that the mechanism of gp40-mediated class I retention is not catalytic but rather stochiometric, i.e., for a certain number of folded MHC class I dimers one needs a certain number of gp40 proteins to retain them.

The dependence of MHC class I molecule trafficking on β2m and peptides and on their relative abundance in a particular cell line make it sometimes hard to investigate this protein. This experiment has told us that with this complex machinery, one needs to take additional care with the choice of the expression systems and with the abundance of each of the components of the system to obtain useful data.

5.1.5 Additional β2m in murine cells protects MHC class I from gp40 in an allotype-specific way

In human cells, gp40 retains murine MHC class I molecules, but it can be out-titrated ('overwhelmed') by large amounts of folded MHC class I molecules, which mask its effect (see 5.1.2). This happens when

β2m is overexpressed additionally. Without extra β2m, in human cells most of murine MHC class I molecules remain unfolded and will never reach the cell surface in the first place, regardless of the presence of gp40 (see 5.1.1).

What effect does the overexpression of β2m have on endogenous MHC class I in murine cells, where the

murine MHC class I molecules are already supplied with endogenous β2m and fold much better then when expressed exogenously in human cells? Will the MHC class I molecules also become protected from gp40 or is there no change? For this, we transduced K41 cells (which already express endogenous

murine β2m) with hβ2m-IRES-Neo and selected them (done by Zeynep Hein), and then we supertransduced them with gp40-IRES-GFP. The surface was then stained for Db and Kb.

125 The influence of β2m on gp40 function

b b Figure 5-6 Supertransduction of hβ2m slightly increases surface D and K b b K41 cells were transduced with hβ2m-IRES-Neo, selected and stained for surface D and K . N = 3. Error as standard deviation. Data by V. Ramnarayan, figure drawn by the author.

b b After overexpression of hβ2m in K41 cells, we observe a slight increase of surface D and K , which could not be proven to be significant (see Figure 5-6). There are two possible reasons for this increase: either, the endogenous β2m is not sufficient, and few MHC class I free heavy chains are unable to form dimers with the endogenous 2m and are eventually degraded; the extra hβ2m rescues them and enables their surface transport. Alternatively, murine MHC class I molecules fold faster and better with human β2m because of their increased affinity (Achour et al., 2006; Pedersen et al., 1995; Shields et al., 1998), and this effect helps to export some more MHC class I molecules than with murine β2m alone. This could be tested by the addition of murine β2m, which in this case should not lead to an increase of surface MHC class I.

b Figure 5-7 Supertransduction of hβ2m protects K in murine K41 from gp40 inhibition K41 cells were transduced with hβ2m-IRES-Neo and selected, and supertransduced with gp40-IRES-GFP. Cells were surface stained for Db and Kb, and analyzed in a flow cytometer. The GFP MFI was divided into different gates, and for each an APM MFI was taken and depicted as a curve. A: Kb surface levels; B: Db surface levels. N = 4, error as SEM. Data by V. Ramnarayan, figure drawn by the author.

126 The influence of β2m on gp40 function

After supertransduction of gp40, we observe that Kb is less affected by gp40 than before (see Figure 5-7A). This is not true for Db, where the curves for the surface levels are identical with or without extra

β2m (see Figure 5-7 B). This experiment tells us that the addition of β2m has only a slight affect of

endogenous MHC class I surface levels. Examining the effect of β2m on endogenous MHC class I b b molecules allows us to see an allotype-specific difference, as K profits more from the extra β2m than D does.

5.1.6 Discussion

5.1.6.1 Murine and human MHC class I and gp40

gp40 retains murine but not human MHC class I molecules (Ziegler et al., 1997). In Table 5-1, the tested molecules and cell lines are listed.

Table 5-1 MHC class I allotypes and cell lines tested with gp40

Allotype Cell line Retention Publication (Ziegler et al., 1997); H-2d, H-2k, H-2b, H-2q their respective cells yes mentioned, not shown Ld, Kk, Kd their respective cells yes (Ziegler et al., 1997); shown Kd human LC-5 cells16 yes (Ziegler et al., 1997); shown 17 B27 + hβ2m 1T 22 cells no (Ziegler et al., 1997); shown HLA-B7 HeLa no (Gutermann, 2001); shown human LC-5 cells, HLA-A2 no (Gutermann, 2001); shown transfected with Α2 human LC-5 cells, H-2Kb yes (Gutermann, 2001); shown transfected with Kb HLA-A2 murine Ltk- cells18 yes, partial (Gutermann, 2001); shown HLA-A2 murine L929 yes, partial (Gutermann, 2001); shown

gp40 retains all tested murine allotypes (see Table 5-1), and that most likely means it binds to all of them, but it does not bind to human MHC class I (see Figure 3-10B). Possibly there are residues that are conserved in the murine MHC class I sequence but different in the human sequence. To identify those sequences, an alignment of several human to several murine sequences was done (see Figure 5-8).

16 LC-5 cells express homozygously the alleles HLA-A*6802, HLA-B*1503, and HLA-C*1203. 17 Th cell line isolated from Swiss 3T3 cells (Arnold et al., 1984). 18 Ltk- cells are generated from L929 cells by deletion of the thymidine kinase (Kit et al., 1963).

127 The influence of β2m on gp40 function

Assuming gp40 binds to MHC class I as proposed by Margulies and coworkers (Wang et al., 2012), only the residues that are in the α1 and α2 domain and point outwards were selected. Thirteen residues were identified (see Table 5-2 and Figure 5-9), of which seven had differences in charge, hydrophobicity or size. The role of these residues in class I interaction with gp40 could be tested by mutagenesis of class I.

In addition, it would be interesting to try if gp40 can retain endogenous MHC class I in monkey (e.g. Vero) or hamster (e.g. CHO) cells. Depending on the outcome, one could research the MHC class I allotypes in these cell lines and include their sequences into the analysis. If gp40 functions in either of these cell lines, it would also be interesting to investigate if the correlation between function of gp40, its interaction with MHC class I, and the importance of the maturation speed of MHC class I can be observed here as well (Figure 6-3).

------1---5----10---15---20---25---30---35 HLA-A2 MAVMAPRTLVLLLSGALALTQTWAGSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRF 60 HLA-A29 MAVMAPRTLLLLLLGALALTQTWAGSHSMRYFTTSVSRPGRGEPRFIAVGYVDDTQFVRF 60 HLA-A1 MAVMAPRTLLLLLSGALALTQTWAGSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRF 60 HLA-B7 MLVMAPRTVLLLLSAALALTETWAGSHSMRYFYTSVSRPGRGEPRFISVGYVDDTQFVRF 60 HLA-B8 MLVMAPRTVLLLLSAALALTETWAGSHSMRYFDTAMSRPGRGEPRFISVGYVDDTQFVRF 60 HLA-B15 MRVTAPRTVLLLLSGALALTETWAGSHSMRYFYTAMSRPGRGEPRFIAVGYVDDTQFVRF 60 HLA-B27 MRVTAPRTLLLLLWGAVALTETWAGSHSMRYFHTSVSRPGRGEPRFITVGYVDDTLFVRF 60 HLA-B44 MRVTAPRTLLLLLWGAVALTETWAGSHSMRYFYTAMSRPGRGEPRFITVGYVDDTLFVRF 60 H-2Db MGAMAPRTLLLLLAAALAPTQTRAGPHSMRYFETAVSRPGLEEPRYISVGYVDNKEFVRF 60 H-2Ld MGAMAPRTLLLLLAAALAPTQTRAGPHSMRYFETAVSRPGLGEPRYISVGYVDNKEFVRF 60 H-2Lq MGAMAPRTLLLLLAAALAPTQTRAGPHSMRYFETAVSRPGLGEPRYISVGYVDNKEFVRF 60 H-2Kb ---MVPCTLLLLLAAALAPTQTRAGPHSLRYFVTAVSRPGLGEPRYMEVGYVDDTEFVRF 57 H-2Kk ---MAPCMLLLLLAAALAPTQTRAGPHSLRYFHTAVSRPGLGKPRFISVGYVDDTQFVRF 57 H-2Kd ---MAPCTLLLLLAAALAPTQTRAGPHSLRYFVTAVSRPGLGEPRFIAVGYVDDTQFVRF 57 H-2Dk MGAMVPRTLLLLLAAALAPIQTRAGPHSLRYFETVVSRPGLGKPRFISVGYVDNTEFVRF 60 H-2Dd MGAMAPRTLLLLLAAALGPTQTRAGSHSLRYFVTAVSRPGFGEPRYMEVGYVDNTEFVRF 60 .* ::*** .*:. :* **.**:*** * :**** :**:: *****:. ****

---40------50------60------70------80------90----- HLA-A2 DSDAASQRMEPRAPWIEQEGPEYWDGETRKVKAHSQTHRVDLGTLRGYYNQSEAGSHTVQ 120 HLA-A29 DSDAASQRMEPRAPWIEQEGPEYWDLQTRNVKAQSQTDRANLGTLRGYYNQSEAGSHTIQ 120 HLA-A1 DSDAASQKMEPRAPWIEQEGPEYWDQETRNMKAHSQTDRANLGTLRGYYNQSEDGSHTIQ 120 HLA-B7 DSDAASPREEPRAPWIEQEGPEYWDRNTQIYKAQAQTDRESLRNLRGYYNQSEAGSHTLQ 120 HLA-B8 DSDAASPREEPRAPWIEQEGPEYWDRNTQIFKTNTQTDRESLRNLRGYYNQSEAGSHTLQ 120 HLA-B15 DSDAASPRMAPRAPWIEQEGPEYWDRETQISKTNTQTYRESLRNLRGYYNQSEAGSHTLQ 120 HLA-B27 DSDAASPREEPRAPWIEQEGPEYWDRETQICKAKAQTDREDLRTLLRYYNQSEAGSHTLQ 120 HLA-B44 DSDATSPRKEPRAPWIEQEGPEYWDRETQISKTNTQTYRENLRTALRYYNQSEAGSHIIQ 120 H-2Db DSDAENPRYEPRAPWMEQEGPEYWERETQKAKGQEQWFRVSLRNLLGYYNQSAGGSHTLQ 120 H-2Ld DSDAENPRYEPQAPWMEQEGPEYWERITQIAKGQEQWFRVNLRTLLGYYNQSAGGTHTLQ 120 H-2Lq DSDAENPRYEPQAPWMEQEGPEYWERITQIAKGQEQWFRVNLRTLLGYYNQSAGGTHTIQ 120 H-2Kb DSDAENPRYEPRARWMEQEGPEYWERETQKAKGNEQSFRVDLRTLLGYYNQSKGGSHTIQ 117 H-2Kk DSDAENPRYEPRVRWMEQVEPEYWERNTQIAKGNEQIFRVNLRTALRYYNQSAGGSHTFQ 117 H-2Kd DSDADNPRFEPRAPWMEQEGPEYWEEQTQRAKSDEQWFRVSLRTAQRYYNQSKGGSHTFQ 117 H-2Dk DSDAENPRDEPRVRWMEQEGPEYWERETQIAKGNEQSFRVDLRTLLRYYNQSEGGSHTIQ 120 H-2Dd DSDAENPRYEPRARWIEQEGPEYWERETRRAKGNEQSFRVDLRTALRYYNQSAGGSHTLQ 120 **** . : *:. *:** ****: *: * . * * .* . ***** *:* .*

128 The influence of β2m on gp40 function

---100------110------120------130------140------150---- HLA-A2 RMYGCDVGSDWRFLRGYHQYAYDGKDYIALKEDLRSWTAADMAAQTTKHKWEAAHVAEQL 180 HLA-A29 MMYGCHVGSDGRFLRGYRQDAYDGKDYIALNEDLRSWTAADMAAQITQRKWEAARVAEQL 180 HLA-A1 IMYGCDVGPDGRFLRGYRQDAYDGKDYIALNEDLRSWTAADMAAQITKRKWEAVHAAEQR 180 HLA-B7 SMYGCDVGPDGRLLRGHDQYAYDGKDYIALNEDLRSWTAADTAAQITQRKWEAAREAEQR 180 HLA-B8 SMYGCDVGPDGRLLRGHNQYAYDGKDYIALNEDLRSWTAADTAAQITQRKWEAARVAEQD 180 HLA-B15 RMYGCDVGPDGRLLRGHDQSAYDGKDYIALNEDLSSWTAADTAAQITQRKWEAAREAEQW 180 HLA-B27 NMYGCDVGPDGRLLRGYHQDAYDGKDYIALNEDLSSWTAADTAAQITQRKWEAARVAEQL 180 HLA-B44 RMYGCDVGPDGRLLRGYDQDAYDGKDYIALNEDLSSWTAADTAAQITQRKWEAARVAEQD 180 H-2Db QMSGCDLGSDWRLLRGYLQFAYEGRDYIALNEDLKTWTAADMAAQITRRKWEQSGAAEHY 180 H-2Ld WMYGCDVGSDGRLLRGYEQFAYDGCDYIALNEDLKTWTAADMAAQITRRKWEQAGAAEYY 180 H-2Lq RMYGCDVGSHWRLLRGYEQYAYDGCDYIALNEDLKTWTAADMAAQITRRKWEQAGAAEHY 180 H-2Kb VISGCEVGSDGRLLRGYQQYAYDGCDYIALNEDLKTWTAADMAALITKHKWEQAGEAERL 177 H-2Kk RMYGCEVGSDWRLLRGYEQYAYDGCDYIALNEDLKTWTAADMAALITKHKWEQAGDAERD 177 H-2Kd RMFGCDVGSDWRLLRGYQQFAYDGRDYIALNEDLKTWTAADTAALITRRKWEQAGDAEYY 177 H-2Dk RLSGCDVGSDWRLLRGYEQYAYDGCDYIALNEDLKTWTAADMAALITKHKWEQAGAAERD 180 H-2Dd WMAGCDVESDGRLLRGYWQFAYDGCDYIALNEDLKTWTAADMAAQITRRKWEQAGAAERD 180 : **.: .. *:***: * **:* *****:*** :***** ** *::*** **

---160------170------180 HLA-A2 RAYLEGTCVEWLRRYLENGKETLQRTDAPKTHMTHHAVSDHEATLRCWALSFYPAEITLT 240 HLA-A29 RAYLEGTCVEWLRRYLENGKETLQRTDAPKTHMTHHAVSDHEATLRCWALSFYPAEITLT 240 HLA-A1 RVYLEGRCVDGLRRYLENGKETLQRTDPPKTHMTHHPISDHEATLRCWALGFYPAEITLT 240 HLA-B7 RAYLEGECVEWLRRYLENGKDKLERADPPKTHVTHHPISDHEATLRCWALGFYPAEITLT 240 HLA-B8 RAYLEGTCVEWLRRYLENGKDTLERADPPKTHVTHHPISDHEATLRCWALGFYPAEITLT 240 HLA-B15 RAYLEGLCVEWLRRYLENGKETLQRADPPKTHVTHHPISDHEATLRCWALGFYPAEITLT 240 HLA-B27 RAYLEGECVEWLRRYLENGKETLQRADPPKTHVTHHPISDHEATLRCWALGFYPAEITLT 240 HLA-B44 RAYLEGLCVESLRRYLENGKETLQRADPPKTHVTHHPISDHEVTLRCWALGFYPAEITLT 240 H-2Db KAYLEGECVEWLHRYLKNGNATLLRTDSPKAHVTHHPRSKGEVTLRCWALGFYPADITLT 240 H-2Ld RAYLEGECVEWLHRYLKNGNATLLRTDSPKAHVTHHPRSKGEVTLRCWALGFYPADITLT 240 H-2Lq KAYLEGECVEWLHRYLKNGNATLLRT------206 H-2Kb RAYLEGTCVEWLRRYLKNGNATLLRTDSPKAHVTHHSRPEDKVTLRCWALGFYPADITLT 237 H-2Kk RAYLEGTCVEWLRRYLQLGNATLPRTDSPKAHVTRHSRPEDKVTLRCWALGFYPADITLT 237 H-2Kd RAYLEGECVEWLRRYLELGNETLLRTDSPKAHVTYHPRSQVDVTLRCWALGFYPADITLT 237 H-2Dk RAYLEGTCVESLRRYLELGNATLLHTDSPKAHVTHHPRSKVEVTLRCWALGFYPADITLT 240 H-2Dd RAYLEGECVEWLRRYLKNGNATLLRTDPPKAHVTHHRRPEGDVTLRCWALGFYPADITLT 240 :.**** **: *:***: *: .* ::

Figure 5-8 Alignment of murine and human MHC class I sequences. The top row indicates the residue number excluding the signal sequence. The bottom row indicates the similarity between the sequences. An * (asterisk) indicates positions which have a single, fully conserved residue. A : (colon) indicates conservation between groups of strongly similar properties - scoring > 0.5 in the Gonnet PAM 250 matrix. A . (period) indicates conservation between groups of weakly similar properties - scoring =< 0.5 in the Gonnet PAM 250 matrix. Created by the author with clustalw (Goujon et al., 2010; Larkin et al., 2007). The residues highlighted in red are different between human and murine MHC class I molecules and might play a role in binding to gp40. The residues highlighted in light green indicate the glycosylation sites.

129 The influence of β2m on gp40 function

Figure 5-9 Residues in murine MHC class I that might play a role in interaction with gp40 d A: Murine MHC class I allotype D in blue, β2m in orange and peptide in red, view from top. Highlighted in gray are residues that are different compared to human MHC class I but similar in size and charge. Highlighted in yellow are residues that are different compared to human MHC class I in size and charge. Residues are labeled by number and three letter code. B: The MHC class I is as in A, and the docked gp40 is shown on top in transparent orange. Drawn by the author with VMD.

130 The influence of β2m on gp40 function

Table 5-2 Conserved residues in the murine MHC class I α1 and α2 sequence that are different in human The thirteen residues are listed with a description of their size and charge. The residues in gray are similar in size and charge between mice and human, so they might not be important. The residues shaded in yellow are different in size and charge. In violet, the gp40 residues that might contact MHC class I according to the docking are shown.

Residue No. Mouse Human gp40 docked 17 L hydrophobic R positively charged, large 41 E negatively charged A hydrophobic, small 42 N polar uncharged S polar uncharged 61 E negatively charged D negatively charged 69 G small hydrophobic A small hydrophobic 71 E negatively charged A small hydrophobic 28 E (YREGP) E negatively charged 76 V hydrophobic 113 D (MMDGN) A small hydrophobic 90 G small hydrophobic A small hydrophobic 131 K large, positively charged R large, positively charged 132 T polar uncharged S polar uncharged 149 Q large polar uncharged A small hydrophobic H positively charged 151 G small hydrophobic R large, positively charged 176 A small hydrophobic E negatively charged

5.1.6.2 The role of β2m in gp40 function

b It is known that β2m is absolutely indispensable for MHC class I export, with the exception of D , which is hardly functional as a free heavy chain (Allen et al., 1986). Without β2m, MHC class I heavy chains are unable to achieve the fold in their α1 and α2 domain in the first place (Hansen et al., 1988), and cannot bind to any antigenic peptides (Allen et al., 1986). Both β2m and high affinity peptide help to stabilize the MHC class I molecule, the so-called cooperative effect (Parker et al., 1992).

We actually do not know the stochiometry between MHC class I and β2m levels in the ER lumen in the natural situation. A 1:1 ratio would make sense, since one heavy chain binds one β2m. Presumably, not all MHC class I heavy chains manage to fold, even if there is enough β2m around. This is especially likely when the allotype is generally conformationally unstable, for example Ld (Myers et al., 1989). This would leave some unbound β2m in the cell. In serum, depending on the concentration, β2m can form amyloid deposits (Eichner and Radford, 2011). Apparently, formation of β2m amyloid is possible even inside the

131 The influence of β2m on gp40 function

cell (García-García et al., 1999). Therefore, an overproduction of β2m might pose a greater disadvantage to the cell than a slight underproduction. We do not know if overexpression of β2m leads to amyloid formation in the cell when it is not bound to class I heavy chains, but this can be investigated with amyloid specific antibodies.

We do not really know what the effect of β2m overexpression is on MHC class I folding and export. Is the export time different or the same, and does the peptide population bound to MHC class I molecules change? If yes, are the peptides of higher or lower affinity? This could be tested by pulse-chase experiments, peptide elution, and thermostability experiments (Figure 3-2B) (Fritzsche et al., 2015).

The sequence of β2m and its partnering MHC class I allotype also play an important role. Even though

β2m is quite conserved between mouse and human (69% identity) (Achour et al., 2006), murine and human β2m differ in their affinities to the heavy chains. For example, human β2m has a higher affinity to murine heavy chains than murine β2m (Achour et al., 2006; Pedersen et al., 1995; Shields et al., 1998). 19 Additionally, in mice, there are several β2m allotypes (Gasser et al., 1985), which can even change the peptide binding specificity of the heavy chain, to make the story even more complex (Hermel et al.,

1993). It can, therefore, also be expected that small changes in β2m expression levels or in the sequence of β2m cause profound changes in MHC class I folding and surface transport. Not surprisingly, also, xenogeneic β2m can influence the peptide specificity of an MHC class I molecule (Smith et al., 1993).

Interestingly, bovine β2m stabilizes murine MHC class I molecules as well as hβ2m does (Shields et al., 1998). This would mean that in a cell culture containing FCS, many class I molecules might have exchanged their β2m and we are possibly observing a heterogeneous population of β2m bound to MHC class I on the cell surface. Murine and human β2m, although they have the same number of amino acids, 20 have apparent molecular weights. This might also be true for bovine β2m . The identity of the bound

β2m could therefore be investigated by separating them on a gradient gel after surface biotinylation. One could also incubate the cells over night in serum-free medium and observe whether the surface levels or the surface degradation changes or investigate the identity of the β2m directly by staining the cells with β2m-specific antibodies and analyze the cells in flow cytometry.

b b Why does K in K41 cells profit more from the extra β2m than D (see Figure 5-6) in reverting the gp40 retention effect? This might be because Kb folds faster than Db, and therefore is able to bind more of the

19 See also Uniprot entry P01887 for murine β2m. 19 See Uniprot entry P61769 for human β2m, and P01888 for bovine β2m.

132 The influence of β2m on gp40 function

b additional β2m than D . Still, this should also be visible in the surface of control cells (see Figure 5-6), but here we see that levels of both allotypes are elevated to the same extent. It might also be that both

b b allotypes bind the same amounts of hβ2m, but that K exits faster than D (Fritzsche et al., 2015), and b b that therefore it is less affected by gp40. Or else, K has a higher affinity towards human β2m than D .

This could be tested on a gradient gel, as human and murine β2m run differently fast. The ratio with which the endogenous and the extra β2m are bound should tell us something about the affinity and interaction dynamics. We have observed before that Db and Kb react differently to gp40, but the real reason is so far not known. It might be the folding speed, the interaction strength with gp40, or a mixture of both.

This allotype effect upon β2m overexpression could only be detected in K41 (see Figure 5-7), but not in HEK293T cells (see Figure 5-3). In these, both Db and Kb always behaved the same, depending on how much β2m was supplied. This difference between the cell lines might be due to the levels of other proteins, such as chaperones, that differentially affect the folding or peptide binding of the class I allotypes. Possibly, with a certain amount of β2m, this fine difference might be visible even in the human

HEK293T cells. It would be interesting to repeat the β2m titration in gp40-expressing cells that express both Db and Kb, and to see how their surface levels are recovered. Is Kb faster than Db or do they come back at the same time?

In human cells, murine MHC class I molecules are not transported to the cell surface as long as no extra

β2m is supplied (Figure 5-2). Most molecules probably remain as inside the cell as free heavy chains.

Does gp40 bind to these? Our results indicate that gp40 binds to the α1/α2 superdomain of MHC class I molecules, i.e., the part of the protein that only folds in the presence of β2m (see Figure 3-3 C). Therefore, it is unlikely that gp40 can bind free heavy chains.

The Db free heavy chain has been shown to travel to the surface to a limited extent in the absence of

β2m (Allen et al., 1986; Machold et al., 1995; Williams et al., 1989), and even to present antigen (Bix and b Raulet, 1992). If β2m is required for class I recognition by gp40, then in β2m-deficient cells, where D and gp40 are expressed, Db free heavy chains should be able to make it to the surface and be detected there

b with MAb 28-14-8s. But alternatively, D might be able to fold the α1/α2 superdomain to a sufficiently stable conformation even without β2m (perhaps with the help of peptide) such that gp40 can recognize it. The biochemical differences between class I allotypes are not nearly sufficiently investigated.

133 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40 5.1.6.3 Overexpression by 293T cells

We realized in the experiment with the HEK293T cells that expression by transfection leads to much higher amounts of protein in this cell line than expression by transduction. This is probably mostly due to the presence of the SV40 large T antigen in these cells, which leads to an amplification of the transfected plasmid and thus to a massive overexpression of transfected constructs. Anyways, also in other cells, a difference in expression levels between transduction and transfection might be the case. It needs to be quantified to what levels transfected or transduced constructs are expressed in a certain cell line. This only works in selected cells, because in an unselected population, the amount of protein would depend more on the transfection/transduction efficiency than on the expression levels.

The massive overexpression probably causes for the HEK293T cells all sorts of problems, since the most highly expressing cells regularly die after two or three days. Most likely, a strong stress response occurs that which probably also has an influence on the data generated from transfected HEK cells.

6 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40

6.1 Establishment of Db/Kb domain swaps in gp40-expressing L cells

The murine MHC class I allotypes Db and Kb are found in BL/6 mice. They share a high homology (84% identity, Figure 6-1). The most striking difference is their export rate: Db has an export halftime of around 60 minutes, Kb only needs 20 minutes (Degen and Williams, 1991; Fritzsche et al., 2015). Kb also folds faster and is more conformationally stable than Db (Fritzsche et al., 2015).

H-2Db MGAMAPRTLLLLLAAALAPTQTRAGPHSMRYFETAVSRPGLEEPRYISVGYVDNKEFVRF 60 H-2Kb ---MVPCTLLLLLAAALAPTQTRAGPHSLRYFVTAVSRPGLGEPRYMEVGYVDDTEFVRF 57 *.* *********************:*** ******** ****:.*****:.*****

H-2Db DSDAENPRYEPRAPWMEQEGPEYWERETQKAKGQEQWFRVSLRNLLGYYNQSAGGSHTLQ 120 H-2Kb DSDAENPRYEPRARWMEQEGPEYWERETQKAKGNEQSFRVDLRTLLGYYNQSKGGSHTIQ 117 ************* *******************:** ***.**.******** *****:*

H-2Db QMSGCDLGSDWRLLRGYLQFAYEGRDYIALNEDLKTWTAADMAAQITRRKWEQSGAAEHY 180 H-2Kb VISGCEVGSDGRLLRGYQQYAYDGCDYIALNEDLKTWTAADMAALITKHKWEQAGEAERL 177 :***::*** ****** *:**:* ******************* **::****:* **:

134 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40

H-2Db KAYLEGECVEWLHRYLKNGNATLLRTDSPKAHVTHHPRSKGEVTLRCWALGFYPADITLT 240 H-2Kb RAYLEGTCVEWLRRYLKNGNATLLRTDSPKAHVTHHSRPEDKVTLRCWALGFYPADITLT 237 :***** *****:***********************.*.:.:******************

H-2Db WQLNGEELTQDMELVETRPAGDGTFQKWASVVVPLGKEQNYTCRVYHEGLPEPLTLRWEP 300 H-2Kb WQLNGEELIQDMELVETRPAGDGTFQKWASVVVPLGKEQYYTCHVYHQGLPEPLTLRWEP 297 ******** ****************************** ***:***:************

H-2Db PPSTDSYMVIVAVLGVLGAMAIIGAVVAFVMK-RRRNTGGKGGDYALAPGSQSSEMSLRD 359 H-2Kb PPSTVSNMATVAVLVVLGAAIVTGAVVAFVMKMRRRNTGGKGGDYALAPGSQTSDLSLPD 357 **** * *. **** **** : ********* *******************:*::** *

H-2Db CKA------362 H-2Kb CKVMVHDPHSLA 369 **.

Figure 6-1 Sequence alignment of H-2Db and H-2Kb shows 84% identity An ‘*’ (asterisk) indicates positions which have a single, fully conserved residue. A ‘:’ (colon) indicates conservation between groups of strongly similar properties - scoring > 0.5 in the Gonnet PAM 250 matrix. A ‘.’ (period) indicates conservation between groups of weakly similar properties - scoring =< 0.5 in the Gonnet PAM 250 matrix. Alignment done by the author with ClustalW (http://www.ebi.ac.uk/Tools/msa/clustalw2/).

A difference between the allotypes is also observed in the context of gp40: Db is affected more strongly by gp40 than Kb, as visible from the surface levels, subcellular localization, and export dynamics of the allotypes (compare allotypes in Figure 3-1).

Which domain of the MHC class I molecules is responsible for this effect? It was likely to be the ER- lumenal domain, since gp40 most likely binds to it, but we could not exclude the possibility that the transmembrane or cytosolic domain play a role as well.

Figure 6-2 Kb and Db domain swap constructs The nomenclature of the constructs is K for Kb sequences, D for Db sequences. Denoted are the signal sequences with a small letter, the lumenal domain (α1, α2, α3), the transmembrane domain and the cytosolic tail with a capital letter, respectively. All constructs have the signal sequence for Kb (SS), to ensure similar ER import. The lumenal domains of MHC (α1, α2, α3) are swapped as one domain, the transmembrane domain (TMD) and cytosolic tail also. The kKKK (Y256N) construct has an additional glycosylation site at position 256, as Db has (orange circles = glycosylation sites). The color coding is green for Kb, red for Db. Drawn by the author.

135 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40

We therefore used L cells, which were transfected with Db and Kb domain swaps (Figure 6-2) and sorted, and we determined their susceptibility towards gp40 function. Cells were then transduced with gp40- IRES-GFP, surface stained for Db and Kb, and analyzed on a flow cytometer.

Figure 6-3 The lumenal domain of MHC class I determines the effectivity of gp40 L cells were transfected with Db and Kb domain swaps and sorted. They were then transduced with gp40-IRES-GFP and surface stained for Db and Kb. The GFP MFI was divided into different gates, and for each, an APM MFI was taken and depicted as a curve. The nomenclature is as in Figure 6-2. N = 8, DKK N = 7. Data by I. Huppertz, figure drawn by the author.

The presence of a Kb lumenal domain renders the constructs relatively resistant against gp40 (see Figure 6-3, blue curves), whereas a Db lumenal domain renders them highly sensitive (see Figure 6-3, pink curves). Introducing a third glycosylation site into Kb, just as it is found in Db, still retains the Kb phenotype (see Figure 6-3, blue curve, lime triangle). The lumenal domain clearly determines whether the construct has a Kb or Db phenotype. This result correlates with the export speed of these domain swaps, where constructs with Kb lumenal domain are faster (Figure 5A-B in (Fritzsche et al., 2015)).

Therefore, the observed sensitivity towards gp40 might be a secondary effect of the export rate of the particular construct. Looking very closely, there seems to be additionally a small effect of the cytoplasmic tail and transmembrane domain; for example, DKK has higher surface levels than DDD, and KDD has lower surface levels than KKK. These tiny effects are also observed in the export rate, so the correlation still exists (Fritzsche et al., 2015).

136 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40

6.1.1 Are the endogenous molecules Dk and Kk in L cells retained by gp40?

L cells are murine cells and express the class I allotypes Dk and Kk. Like Db and Kb, these two allotypes show differences in their export rate (Williams et al., 1985). Kk has an export half-time of 15 minutes, whereas Dk needs 240 minutes. The difference between the export rates is thus much higher than that between Db and Kb. We wondered whether gp40 retains these two allotypes, and we transduced L cells with gp40-IRES-GFP. These cells were then surface stained for Dk and Kk, and analyzed in a flow cytometer.

Figure 6-4 Dk is more affected by gp40 than Kk L cell were transduced with gp40-IRES-GFP and surface stained for Dk and Kk. The GFP MFI was divided into different gates, and for each an APM MFI was taken and depicted as a curve. Blue curve = Kk; pink curve = Dk. Kk N = 3, Dk N = 1. Data by I. Huppertz, figure drawn by the author.

The curve for Kk looks very similar to that observed for Kb (see Figure 6-4). Dk, on the other hand, looks like Db. So far21, the data support the interpretation that sensitivity towards gp40 is inversely correlated with the export speed. Still, if gp40 binds the variable α1/α2 superdomain, there probably is some effect of the binding affinity.

Dk trafficks much more slowly than Db, and therefore one might expect that it is even more strongly retained by gp40 than Db, but this we do not see in this experiment. Perhaps the Db effect is already maximal, and the MHC class I surface levels cannot tell us if Dk is more sensitive towards gp40 or not. Alternatively, gp40 does not bind to Dk as tightly as it does to Db, therefore even though Dk trafficks more slowly, it is not more retained by gp40 than Db.

21 This experiment was only done once.

137 MHC class I domain swaps: the ER-lumenal domain of class I is responsible for its interaction with gp40 6.1.2 Discussion

It is clear from our domain swap experiments (see 6.1) and those of Gutermann (Gutermann, 2001) that the ER-lumenal domain of class I is most important in determining susceptibility to gp40, with a possible small effect of the transmembrane and cytosolic domains (see below).

Db and Kb are retained differently by gp40 (compare Figure 3-1), and the results of the domain swap experiment suggest the simplest interpretation, namely that differential retention is caused by the different export rate of the constructs (Fritzsche, 2014; Fritzsche et al., 2015).

An alternative, or perhaps additional, explanation is that differential retention by gp40 is caused by differential binding strength of gp40 to the different allotypes. This hypothesis is not yet substantiated. Until we know from a crystal what the gp40-class I interaction really looks like, we cannot say exactly which residues are involved in the interaction (see 5.1.6.1). Based on our own data, we think that gp40 binds to the MHC class I molecules via the binding groove, which is the most variable domain of the molecule (see Figure 3-3C). Allotypic variation in the binding interface would lead to the binding affinity between gp40 and MHC class I being different for different MHC class I allotypes. Indeed, we observe a stronger interaction signal with Db in immunoprecipitation (compare Figure 3-4A and B), but this might be because Db spends more time in the ER and is therefore more likely to be bound by gp40, which leads us back to the export speed. It is therefore necessary to perform the immunoprecipitations from lysates of cells that were treated with BFA to retain both MHC class I molecules and gp40 in the ER for an extended time. In such a BFA block experiment, we found that the interaction signal for class I is indeed increased (V. Ramnarayan, unpublished observations). This binding increase is consistent with both hypotheses but clearly indicates that in any case dwelling time and export rate of class I are very important.

The data from other allotypes, which could shed more light on this issue, are not yet conclusive. Db and Dk, although they have different export rates, appear to have a similar gp40 inhibition phenotype (compare Figure 6-3 and Figure 6-4). However, in this experiment, these two allotypes cannot be compared directly. They, and other interesting allotypes such as Kk and Ld, need to be expressed in some standardized cell line system, maybe a human cell line, and then investigated in the presence and absence of gp40. An allotype whose export speed lies between Kb and Db should also show a retention phenotype that lies in between.

138 Outlook

An alternative line of experiment to investigate differential affinities of gp40 to class I are in vitro binding experiments with purified proteins, for example by surface plasmon resonance (SPR, Biacore).

In the domain swap experiments, we found a small but reproducible effect of the transmembrane and cytosolic domains, with the Kb-derived domains promoting greater resistance to gp40. This could be investigated further by testing another murine MHC I lumenal domain (for example, Dk) with Db and/or Kb transmembrane and cytosolic domains and determining whether the effect remains. Again, resistance to gp40 might be caused by greater export rates.

7 Outlook

This study has tried to understand how the MCMV immunoevasin gp40 prevents the export of MHC class I molecules to the surface of the cell. We showed that gp40 does not use the MHC class I quality control to retain MHC class I inside the cell, and that the folding, peptide loading, and PLC incorporation of affected MHC class I molecules is basically unchanged. gp40 is interacting with MHC class I, most likely by directly binding the α1/α2 superdomain, as it was suggested for RAE-1, the other target of gp40. Another finding was that the so-called linker, a flexible sequence in the gp40 ER-lumenal domain, plays a role in retaining gp40 in the early secretory pathway.

There are still many unanswered question, but the most urgent one is: which protein is responsible for the retention of gp40? We believe that an unidentified retention factor binds to the linker sequence of gp40 and results in its retention. It is probably advisable to follow this up with two strategies that complement each other.

One is to characterize the linker sequence further. For example, one might try to narrow down the number of amino acids required for its function by exchanging parts of the linker sequence for a non- sense sequence. If the whole length linker is necessary, it possibly forms a three-dimensional structure, which might be shown by synthesizing the linker as a peptide and measure its hydrodynamic radius under various conditions. Still, the most important experiment to prove that this sequence is sufficient for retention is to transplant it to an unrelated protein, such as CD4. The read-out in all these experiments are either cell surface levels of MHC class I or CD4, as determined by flow cytometry, or the maturation of CD4, as determined by radiolabelling and pulse-chase.

139 Outlook

At the same time, one needs to find the protein that interacts with the linker of gp40 to hold it back in the early secretory pathway. Possible approaches here are isolating the interactome of gp40-wt and gp40-LDel, which lacks the linker sequence, by immunoprecipitation, and identifying the binding partners by mass spectrometry. If the linker is functional after transplantation onto CD4, this construct should be used as well. Alternatively, one might use the linker sequence (or the part of it which turned out to be necessary) and perform a tandem affinity purification (Puig et al., 2001). A functional approach would be a genetic screen using gene-trap mutagenesis in haploid mammalian cells (Timms et al., 2013).

As soon as some putative interaction partners are identified, their function can be investigated more closely by siRNA knock-down and overexpression of the respective gene.

Less urgent but still interesting is the interaction interface between gp40 and MHC class I molecules, for example to understand why gp40 binds to murine but not human MHC class I. This is important in order to be able to create a non-class I-binding mutant of gp40. Such a mutant could help segregate the class I-specific effects of gp40 from effects on other proteins such as RAE-1. A co-crystal of gp40 and MHC class I would be ideal, but previous attempts by other groups of crystallizing the complex have failed. An alternative would be a careful docking of the gp40 and MHC class I crystals and verifying the interaction partners with point mutants. It might be useful to know if gp40 is able to bind to monkey or hamster MHC class I.

Unrelated to gp40 but interesting to investigate is the role of β2m in the folding, export and peptide loading (and specificity) of MHC class I (see 5.1.6) as well as the influence of cell confluency on MHC class I surface levels and APM transcription (see 4.4.7).

Finally, in this study, we have several times observed an influence of the rate of ER-to-surface transport of a protein on its availability for intracellular interactions. One example are the interactions of Db and Kb with the PLC, which are more pronounced when the class I molecules are held by gp40 in the ER-ERGIC cycle (3.9.2). In another example, restricting Kb to the ER with the help of BFA increases its interaction with gp40 (6.1.2). These observations suggest that in cells, the interactions of a protein determine its rate of trafficking, but also its rate of trafficking determines its interactions.

140 Additional publications

8 Additional publications

During the work for my PhD, I became a co-author on two additional publications:

1. Hein Z, Uchtenhagen H, Abualrous ET, Saini SK, Janßen L, Van Hateren A, Wiek C, Hanenberg H, Momburg F, Achour A, Elliott T, Springer S, Boulanger D (2014). Peptide-independent stabilization of MHC class I molecules breaches cellular quality control. J Cell Sci. 127, 2885-97. doi: 10.1242/jcs.145334. My contribution was support in the lentiviral expression system used in the experimental work.

2. Hsu HT*, Janßen L*22, Lawand M, Kim J, Perez-Arroyo A, Culina S, Gdoura A, Burgevin A, Cumenal D, Fourneau Y, Moser A, Kratzer R, Wong FS, Springer S, van Endert P (2014). Endoplasmic reticulum targeting alters regulation of expression and antigen presentation of proinsulin. J Immunol. 192, 4957-66. doi: 10.4049/jimmunol.1300631. My contribution was Figure 4B-E.

22 The asterisk indicates joint first authorship.

141 References

9 References

Abarca-Heidemann, K., Friederichs, S., Klamp, T., Boehm, U., Guethlein, L.A., and Ortmann, B. (2002). Regulation of the expression of mouse TAP-associated glycoprotein (tapasin) by cytokines. Immunol. Lett. 83, 197–207.

Abe, F., Van Prooyen, N., Ladasky, J.J., and Edidin, M. (2009). Interaction of Bap31 and MHC class I molecules and their traffic out of the endoplasmic reticulum. J. Immunol. Baltim. Md 1950 182, 4776– 4783.

Abele, R., and Tampé, R. (2006). Modulation of the antigen transport machinery TAP by friends and enemies. FEBS Lett. 580, 1156–1163.

Achour, A., Michaëlsson, J., Harris, R.A., Ljunggren, H.-G., Kärre, K., Schneider, G., and Sandalova, T. (2006). Structural basis of the differential stability and receptor specificity of H-2Db in complex with murine versus human beta2-microglobulin. J. Mol. Biol. 356, 382–396.

Acosta-Alvear, D., Zhou, Y., Blais, A., Tsikitis, M., Lents, N.H., Arias, C., Lennon, C.J., Kluger, Y., and Dynlacht, B.D. (2007). XBP1 controls diverse cell type- and condition-specific transcriptional regulatory networks. Mol. Cell 27, 53–66.

Ahn, K., Gruhler, A., Galocha, B., Jones, T.R., Wiertz, E.J., Ploegh, H.L., Peterson, P.A., Yang, Y., and Früh, K. (1997). The ER-luminal domain of the HCMV glycoprotein US6 inhibits peptide translocation by TAP. Immunity 6, 613–621.

Akiyama, K., Kagawa, S., Tamura, T., Shimbara, N., Takashina, M., Kristensen, P., Hendil, K.B., Tanaka, K., and Ichihara, A. (1994). Replacement of proteasome subunits X and Y by LMP7 and LMP2 induced by interferon-gamma for acquirement of the functional diversity responsible for antigen processing. FEBS Lett. 343, 85–88.

Allen, H., Fraser, J., Flyer, D., Calvin, S., and Flavell, R. (1986). Beta 2-microglobulin is not required for cell surface expression of the murine class I histocompatibility antigen H-2Db or of a truncated H-2Db. Proc. Natl. Acad. Sci. U. S. A. 83, 7447–7451.

Alzhanova, D., Edwards, D.M., Hammarlund, E., Scholz, I.G., Horst, D., Wagner, M.J., Upton, C., Wiertz, E.J., Slifka, M.K., and Früh, K. (2009). Cowpox virus inhibits the transporter associated with antigen processing to evade T cell recognition. Cell Host Microbe 6, 433–445.

Ameri, K., and Harris, A.L. (2008). Activating transcription factor 4. Int. J. Biochem. Cell Biol. 40, 14–21.

Andrejeva, J., Childs, K.S., Young, D.F., Carlos, T.S., Stock, N., Goodbourn, S., and Randall, R.E. (2004). The V proteins of paramyxoviruses bind the IFN-inducible RNA helicase, mda-5, and inhibit its activation of the IFN-beta promoter. Proc. Natl. Acad. Sci. U. S. A. 101, 17264–17269.

142 References

Antoniou, A.N., Ford, S., Alphey, M., Osborne, A., Elliott, T., and Powis, S.J. (2002). The oxidoreductase ERp57 efficiently reduces partially folded in preference to fully folded MHC class I molecules. EMBO J. 21, 2655–2663.

Arapovid, J., Lenac Rovis, T., Reddy, A.B., Krmpotid, A., and Jonjid, S. (2009). Promiscuity of MCMV immunoevasin of NKG2D: m138/fcr-1 down-modulates RAE-1epsilon in addition to MULT-1 and H60. Mol. Immunol. 47, 114–122.

Arapovic, J., Lenac, T., Antulov, R., Polic, B., Ruzsics, Z., Carayannopoulos, L.N., Koszinowski, U.H., Krmpotic, A., and Jonjic, S. (2009). Differential susceptibility of RAE-1 isoforms to mouse cytomegalovirus. J. Virol. 83, 8198–8207.

Arase, H., Mocarski, E.S., Campbell, A.E., Hill, A.B., and Lanier, L.L. (2002). Direct recognition of cytomegalovirus by activating and inhibitory NK cell receptors. Science 296, 1323–1326.

Arnold, B., Burgert, H.G., Hamann, U., Hämmerling, G., Kees, U., and Kvist, S. (1984). Cytolytic T cells recognize the two amino-terminal domains of H-2 K antigens in tandem in influenza A infected cells. Cell 38, 79–87.

Ashrafi, G.H., Tsirimonaki, E., Marchetti, B., O’Brien, P.M., Sibbet, G.J., Andrew, L., and Campo, M.S. (2002). Down-regulation of MHC class I by bovine papillomavirus E5 oncoproteins. Oncogene 21, 248– 259.

Babid, M., Pyzik, M., Zafirova, B., Mitrovid, M., Butorac, V., Lanier, L.L., Krmpotid, A., Vidal, S.M., and Jonjid, S. (2010). Cytomegalovirus immunoevasin reveals the physiological role of “missing self” recognition in natural killer cell dependent virus control in vivo. J. Exp. Med. 207, 2663–2673.

Balachandran, S., Roberts, P.C., Brown, L.E., Truong, H., Pattnaik, A.K., Archer, D.R., and Barber, G.N. (2000). Essential Role for the dsRNA-Dependent Protein Kinase PKR in Innate Immunity to Viral Infection. Immunity 13, 129–141.

Barnstable, C.J., Bodmer, W.F., Brown, G., Galfre, G., Milstein, C., Williams, A.F., and Ziegler, A. (1978). Production of monoclonal antibodies to group A erythrocytes, HLA and other human cell surface antigens-new tools for genetic analysis. Cell 14, 9–20.

Bauer, S., Groh, V., Wu, J., Steinle, A., Phillips, J.H., Lanier, L.L., and Spies, T. (1999). Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science 285, 727–729.

Beck, S., Kelly, A., Radley, E., Khurshid, F., Alderton, R.P., and Trowsdale, J. (1992). DNA sequence analysis of 66 kb of the human MHC class II region encoding a cluster of genes for antigen processing. J. Mol. Biol. 228, 433–441.

Bedard, K., Szabo, E., Michalak, M., and Opas, M. (2005). Cellular functions of endoplasmic reticulum chaperones calreticulin, calnexin, and ERp57. Int. Rev. Cytol. 245, 91–121.

Berhane, S., Aresté, C., Ablack, J.N., Ryan, G.B., Blackbourn, D.J., Mymryk, J.S., Turnell, A.S., Steele, J.C., and Grand, R.J.A. (2011). Adenovirus E1A interacts directly with, and regulates the level of expression of, the immunoproteasome component MECL1. Virology 421, 149–158.

143 References

Bernales, S., McDonald, K.L., and Walter, P. (2006). Autophagy counterbalances endoplasmic reticulum expansion during the unfolded protein response. PLoS Biol. 4, e423.

Bij, W. van der, and Speich, R. (2001). Management of Cytomegalovirus Infection and Disease after Solid-Organ Transplantation. Clin. Infect. Dis. 33, S32–S37.

Biswas, A., Meissner, T.B., Kawai, T., and Kobayashi, K.S. (2012). Cutting edge: impaired MHC class I expression in mice deficient for Nlrc5/class I transactivator. J. Immunol. Baltim. Md 1950 189, 516–520.

Bix, M., and Raulet, D. (1992). Functionally conformed free class I heavy chains exist on the surface of beta 2 microglobulin negative cells. J. Exp. Med. 176, 829–834.

Boname, J.M., de Lima, B.D., Lehner, P.J., and Stevenson, P.G. (2004). Viral Degradation of the MHC Class I Peptide Loading Complex. Immunity 20, 305–317. van Boxel-Dezaire, A.H.H., Rani, M.R.S., and Stark, G.R. (2006). Complex modulation of cell type-specific signaling in response to type I interferons. Immunity 25, 361–372.

Browne, H., Smith, G., Beck, S., and Minson, T. (1990). A complex between the MHC class I homologue encoded by human cytomegalovirus and beta 2 microglobulin. Nature 347, 770–772.

Brucet, M., Marqués, L., Sebastián, C., Lloberas, J., and Celada, A. (2004). Regulation of murine Tap1 and Lmp2 genes in macrophages by interferon gamma is mediated by STAT1 and IRF-1. Genes Immun. 5, 26– 35.

Calò, V., Migliavacca, M., Bazan, V., Macaluso, M., Buscemi, M., Gebbia, N., and Russo, A. (2003). STAT proteins: from normal control of cellular events to tumorigenesis. J. Cell. Physiol. 197, 157–168.

Calorini, L., Simile, M., Hauser, S., and Gattonicelli, S. (1994). Reexpression of the major histocompatibility complex (MHC) class-I antigen h-2k(b) by m1 (b16-f10) murine melanoma-cells. Int. J. Oncol. 5, 741–748.

Campbell, A.E., Slater, J.S., and Futch, W.S. (1989). Murine cytomegalovirus-induced suppression of antigen-specific cytotoxic T lymphocyte maturation. Virology 173, 268–275.

Cao, S.S., and Kaufman, R.J. (2012). Unfolded protein response. Curr. Biol. CB 22, R622-626.

Capps, G.G., and Zúñiga, M.C. (1993). The cytoplasmic domain of the H-2Ld class I major histocompatibility complex molecule is differentially accessible to immunological and biochemical probes during transport to the cell surface. J. Biol. Chem. 268, 21263–21270.

Caramelo, J.J., Castro, O.A., Alonso, L.G., De Prat-Gay, G., and Parodi, A.J. (2003). UDP-Glc:glycoprotein glucosyltransferase recognizes structured and solvent accessible hydrophobic patches in molten globule- like folding intermediates. Proc. Natl. Acad. Sci. U. S. A. 100, 86–91.

Carroll, I.R., Wang, J., Howcroft, T.K., and Singer, D.S. (1998). HIV Tat represses transcription of the beta 2-microglobulin promoter. Mol. Immunol. 35, 1171–1178.

144 References

Caruso, C., Buffa, S., Candore, G., Colonna-Romano, G., Dunn-Walters, D., Kipling, D., and Pawelec, G. (2009). Mechanisms of immunosenescence. Immun. Ageing A 6, 10.

Cascio, P., Hilton, C., Kisselev, A.F., Rock, K.L., and Goldberg, A.L. (2001). 26S proteasomes and immunoproteasomes produce mainly N-extended versions of an antigenic peptide. EMBO J. 20, 2357– 2366.

Cerundolo, V., Alexander, J., Anderson, K., Lamb, C., Cresswell, P., McMichael, A., Gotch, F., and Townsend, A. (1990). Presentation of viral antigen controlled by a gene in the major histocompatibility complex. Nature 345, 449–452.

Cerwenka, A., Bakker, A.B., McClanahan, T., Wagner, J., Wu, J., Phillips, J.H., and Lanier, L.L. (2000). Retinoic acid early inducible genes define a ligand family for the activating NKG2D receptor in mice. Immunity 12, 721–727.

Charni, S., Aguilo, J.I., Garaude, J., de Bettignies, G., Jacquet, C., Hipskind, R.A., Singer, D., Anel, A., and Villalba, M. (2009). ERK5 knockdown generates mouse leukemia cells with low MHC class I levels that activate NK cells and block tumorigenesis. J. Immunol. Baltim. Md 1950 182, 3398–3405.

Charni, S., de Bettignies, G., Rathore, M.G., Aguiló, J.I., van den Elsen, P.J., Haouzi, D., Hipskind, R.A., Enriquez, J.A., Sanchez-Beato, M., Pardo, J., et al. (2010). Oxidative phosphorylation induces de novo expression of the MHC class I in tumor cells through the ERK5 pathway. J. Immunol. Baltim. Md 1950 185, 3498–3503.

Chatterjee-Kishore, M., Kishore, R., Hicklin, D.J., Marincola, F.M., and Ferrone, S. (1998). Different requirements for signal transducer and activator of transcription 1alpha and interferon regulatory factor 1 in the regulation of low molecular mass polypeptide 2 and transporter associated with antigen processing 1 gene expression. J. Biol. Chem. 273, 16177–16183.

Cloutier, N., and Flamand, L. (2010). Kaposi sarcoma-associated herpesvirus latency-associated nuclear antigen inhibits interferon (IFN) beta expression by competing with IFN regulatory factor-3 for binding to IFNB promoter. J. Biol. Chem. 285, 7208–7221.

Cobbs, C.S., Harkins, L., Samanta, M., Gillespie, G.Y., Bharara, S., King, P.H., Nabors, L.B., Cobbs, C.G., and Britt, W.J. (2002). Human cytomegalovirus infection and expression in human malignant glioma. Cancer Res. 62, 3347–3350.

Cretney, E., Degli-Esposti, M.A., Densley, E.H., Farrell, H.E., Davis-Poynter, N.J., and Smyth, M.J. (1999). m144, a murine cytomegalovirus (MCMV)-encoded major histocompatibility complex class I homologue, confers tumor resistance to natural killer cell-mediated rejection. J. Exp. Med. 190, 435–444.

Decker, T., Kovarik, P., and Meinke, A. (1997). GAS elements: a few nucleotides with a major impact on cytokine-induced gene expression. J. Interferon Cytokine Res. Off. J. Int. Soc. Interferon Cytokine Res. 17, 121–134.

Degen, E., and Williams, D.B. (1991). Participation of a novel 88-kD protein in the biogenesis of murine class I histocompatibility molecules. J. Cell Biol. 112, 1099–1115.

145 References

Degen, E., Cohen-Doyle, M.F., and Williams, D.B. (1992). Efficient dissociation of the p88 chaperone from major histocompatibility complex class I molecules requires both beta 2-microglobulin and peptide. J. Exp. Med. 175, 1653–1661.

Del Val, M., Münch, K., Reddehase, M.J., and Koszinowski, U.H. (1989). Presentation of CMV immediate- early antigen to cytolytic T lymphocytes is selectively prevented by viral genes expressed in the early phase. Cell 58, 305–315.

Der, S.D., Zhou, A., Williams, B.R., and Silverman, R.H. (1998). Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc. Natl. Acad. Sci. U. S. A. 95, 15623–15628.

Dick, T.P., Bangia, N., Peaper, D.R., and Cresswell, P. (2002). Disulfide bond isomerization and the assembly of MHC class I-peptide complexes. Immunity 16, 87–98.

Dolan, A., Cunningham, C., Hector, R.D., Hassan-Walker, A.F., Lee, L., Addison, C., Dargan, D.J., McGeoch, D.J., Gatherer, D., Emery, V.C., et al. (2004). Genetic content of wild-type human cytomegalovirus. J. Gen. Virol. 85, 1301–1312.

Dölken, L., Pfeffer, S., and Koszinowski, U.H. (2009). Cytomegalovirus microRNAs. Virus Genes 38, 355– 364.

Dölken, L., Krmpotic, A., Kothe, S., Tuddenham, L., Tanguy, M., Marcinowski, L., Ruzsics, Z., Elefant, N., Altuvia, Y., Margalit, H., et al. (2010). Cytomegalovirus microRNAs Facilitate Persistent Virus Infection in Salivary Glands. PLoS Pathog 6, e1001150.

Dong, G., Wearsch, P.A., Peaper, D.R., Cresswell, P., and Reinisch, K.M. (2009). Insights into MHC class I peptide loading from the structure of the tapasin-ERp57 thiol oxidoreductase heterodimer. Immunity 30, 21–32.

DuBridge, R.B., Tang, P., Hsia, H.C., Leong, P.M., Miller, J.H., and Calos, M.P. (1987). Analysis of mutation in human cells by using an Epstein-Barr virus shuttle system. Mol. Cell. Biol. 7, 379–387.

Dunn, C., Chalupny, N.J., Sutherland, C.L., Dosch, S., Sivakumar, P.V., Johnson, D.C., and Cosman, D. (2003). Human cytomegalovirus glycoprotein UL16 causes intracellular sequestration of NKG2D ligands, protecting against natural killer cell cytotoxicity. J. Exp. Med. 197, 1427–1439.

Eichner, T., and Radford, S.E. (2011). Understanding the complex mechanisms of β2-microglobulin amyloid assembly. FEBS J. 278, 3868–3883.

Erusalimsky, J.D., Gilmore, D., and Milstein, C. (1987). The induction of class I HLA by interferon-alpha is independent of the cell cycle, but the expression is enhanced by a G1/S block. Eur. J. Immunol. 17, 623– 628.

Everett, H., and McFadden, G. (1999). Apoptosis: an innate immune response to virus infection. Trends Microbiol. 7, 160–165.

Farfán-Arribas, D.J., Stern, L.J., and Rock, K.L. (2012). Using intein catalysis to probe the origin of major histocompatibility complex class I-presented peptides. Proc. Natl. Acad. Sci. U. S. A. 109, 16998–17003.

146 References

Frickel, E.-M., Frei, P., Bouvier, M., Stafford, W.F., Helenius, A., Glockshuber, R., and Ellgaard, L. (2004). ERp57 is a multifunctional thiol-disulfide oxidoreductase. J. Biol. Chem. 279, 18277–18287.

Friedman, D.J., and Ricciardi, R.P. (1988). Adenovirus type 12 E1A gene represses accumulation of MHC class I mRNAs at the level of transcription. Virology 165, 303–305.

Fritzsche, S. (2014). Differential Transport of Murine MHC Class I Molecules from the Endoplasmic Reticulum to the Golgi Apparatus. PhD Thesis. Jacobs University Bremen.

Fritzsche, S., Abualrous, E.T., Borchert, B., Momburg, F., and Springer, S. (2015). Release from endoplasmic reticulum matrix proteins controls cell surface transport of MHC class I molecules. Traffic Cph. Den. 16, 591–603.

Früh, K., Ahn, K., Djaballah, H., Sempé, P., van Endert, P.M., Tampé, R., Peterson, P.A., and Yang, Y. (1995). A viral inhibitor of peptide transporters for antigen presentation. Nature 375, 415–418.

Garbi, N., Tanaka, S., Momburg, F., and Hämmerling, G.J. (2006). Impaired assembly of the major histocompatibility complex class I peptide-loading complex in mice deficient in the oxidoreductase ERp57. Nat. Immunol. 7, 93–102.

García-García, M., Argilés, À., Gouin-Charnet, A., Durfort, M., García-Valero, J., and Mourad, G. (1999). Impaired lysosomal processing of β2-microglobulin by infiltrating macrophages in dialysis amyloidosis. Kidney Int. 55, 899–906.

García-Sastre, A., and Biron, C.A. (2006). Type 1 Interferons and the Virus-Host Relationship: A Lesson in Détente. Science 312, 879–882.

Garstka, M., Borchert, B., Al-Balushi, M., Praveen, P.V.K., Kühl, N., Majoul, I., Duden, R., and Springer, S. (2007). Peptide-receptive major histocompatibility complex class I molecules cycle between endoplasmic reticulum and cis-Golgi in wild-type lymphocytes. J. Biol. Chem. 282, 30680–30690.

Gasser, D.L., Klein, K.A., Choi, E., and Seidman, J.G. (1985). A new beta-2 microglobulin allele in mice defined by DNA sequencing. Immunogenetics 22, 413–416.

Gavioli, R., Gallerani, E., Fortini, C., Fabris, M., Bottoni, A., Canella, A., Bonaccorsi, A., Marastoni, M., Micheletti, F., Cafaro, A., et al. (2004). HIV-1 Tat Protein Modulates the Generation of Cytotoxic T Cell Epitopes by Modifying Proteasome Composition and Enzymatic Activity. J. Immunol. 173, 3838–3843.

Ghazizadeh, S., Carroll, J.M., and Taichman, L.B. (1997). Repression of retrovirus-mediated transgene expression by interferons: implications for gene therapy. J. Virol. 71, 9163–9169.

147 References

Gilbert, M.J., Riddell, S.R., Plachter, B., and Greenberg, P.D. (1996). Cytomegalovirus selectively blocks antigen processing and presentation of its immediate-early gene product. Nature 383, 720–722.

Gold, M.C., Munks, M.W., Wagner, M., Koszinowski, U.H., Hill, A.B., and Fling, S.P. (2002). The Murine Cytomegalovirus Immunomodulatory Gene m152 Prevents Recognition of Infected Cells by M45-Specific CTL But Does Not Alter the Immunodominance of the M45-Specific CD8 T Cell Response In Vivo. J. Immunol. 169, 359–365.

Gold, M.C., Munks, M.W., Wagner, M., McMahon, C.W., Kelly, A., Kavanagh, D.G., Slifka, M.K., Koszinowski, U.H., Raulet, D.H., and Hill, A.B. (2004). Murine cytomegalovirus interference with antigen presentation has little effect on the size or the effector memory phenotype of the CD8 T cell response. J. Immunol. Baltim. Md 1950 172, 6944–6953.

González-Navajas, J.M., Lee, J., David, M., and Raz, E. (2012). Immunomodulatory functions of type I interferons. Nat. Rev. Immunol. 12, 125–135.

Goudsmit, J., Back, N.K., and Nara, P.L. (1991). Genomic diversity and antigenic variation of HIV-1: links between pathogenesis, epidemiology and vaccine development. FASEB J. Off. Publ. Fed. Am. Soc. Exp. Biol. 5, 2427–2436.

Goujon, M., McWilliam, H., Li, W., Valentin, F., Squizzato, S., Paern, J., and Lopez, R. (2010). A new bioinformatics analysis tools framework at EMBL–EBI. Nucleic Acids Res. 38, W695–W699.

Grahame-Clarke, C. (2005). Human cytomegalovirus, endothelial function and atherosclerosis. Herpes J. IHMF 12, 42–45.

Grey, F., Antoniewicz, A., Allen, E., Saugstad, J., McShea, A., Carrington, J.C., and Nelson, J. (2005). Identification and characterization of human cytomegalovirus-encoded microRNAs. J. Virol. 79, 12095– 12099.

Groettrup, M., van den Broek, M., Schwarz, K., Macagno, A., Khan, S., de Giuli, R., and Schmidtke, G. (2001). Structural plasticity of the proteasome and its function in antigen processing. Crit. Rev. Immunol. 21, 339–358.

Gruhler, A., Peterson, P.A., and Früh, K. (2000). Human Cytomegalovirus Immediate Early Glycoprotein US3 Retains MHC Class I Molecules by Transient Association. Traffic 1, 318–325.

Guan, H., Jiao, J., and Ricciardi, R.P. (2008). Tumorigenic Adenovirus Type 12 E1A Inhibits Phosphorylation of NF-κB by PKAc, Causing Loss of DNA Binding and Transactivation. J. Virol. 82, 40–48.

Gupta, S., Cuffe, L., Szegezdi, E., Logue, S.E., Neary, C., Healy, S., and Samali, A. (2010). Mechanisms of ER Stress-Mediated Mitochondrial Membrane Permeabilization. Int. J. Cell Biol. 2010, 170215.

148 References

Gutermann, A. (2001). MCMV interferiert mit der MHC-Klasse-I-Antigenpräsentation: Untersuchungen zur Speziesspezifität und zur Interaktion der viralen Effektormoleküle während der Virusinfektion. Ludwig-Maximilians-Universität München.

Hamilton, S.T., van Zuylen, W., Shand, A., Scott, G.M., Naing, Z., Hall, B., Craig, M.E., and Rawlinson, W.D. (2014). Prevention of congenital cytomegalovirus complications by maternal and neonatal treatments: a systematic review. Rev. Med. Virol. 24, 420–433.

Hämmerling, G.J., Rüsch, E., Tada, N., Kimura, S., and Hämmerling, U. (1982). Localization of allodeterminants on H-2Kb antigens determined with monoclonal antibodies and H-2 mutant mice. Proc. Natl. Acad. Sci. U. S. A. 79, 4737–4741.

Hammond, C., Braakman, I., and Helenius, A. (1994). Role of N-linked oligosaccharide recognition, glucose trimming, and calnexin in glycoprotein folding and quality control. Proc. Natl. Acad. Sci. U. S. A. 91, 913–917.

Han, J., Back, S.H., Hur, J., Lin, Y.-H., Gildersleeve, R., Shan, J., Yuan, C.L., Krokowski, D., Wang, S., Hatzoglou, M., et al. (2013). ER-stress-induced transcriptional regulation increases protein synthesis leading to cell death. Nat. Cell Biol. 15, 481–490.

Hansen, T.H., Myers, N.B., and Lee, D.R. (1988). Studies of two antigenic forms of Ld with disparate beta 2-microglobulin (beta 2m) associations suggest that beta 2m facilitate the folding of the alpha 1 and alpha 2 domains during de novo synthesis. J. Immunol. 140, 3522–3527.

Harms, J.S., and Splitter, G.A. (1995). Interferon-gamma inhibits transgene expression driven by SV40 or CMV promoters but augments expression driven by the mammalian MHC I promoter. Hum. Gene Ther. 6, 1291–1297.

Harton, J.A., and Ting, J.P. (2000). Class II transactivator: mastering the art of major histocompatibility complex expression. Mol. Cell. Biol. 20, 6185–6194.

Haze, K., Yoshida, H., Yanagi, H., Yura, T., and Mori, K. (1999). Mammalian transcription factor ATF6 is synthesized as a transmembrane protein and activated by proteolysis in response to endoplasmic reticulum stress. Mol. Biol. Cell 10, 3787–3799.

Hein, Z., Uchtenhagen, H., Abualrous, E.T., Saini, S.K., Janßen, L., Van Hateren, A., Wiek, C., Hanenberg, H., Momburg, F., Achour, A., et al. (2014). Peptide-independent stabilization of MHC class I molecules breaches cellular quality control. J. Cell Sci. 127, 2885–2897.

Hengel, H., Koopmann, J.-O., Flohr, T., Muranyi, W., Goulmy, E., Hämmerling, G.J., Koszinowski, U.H., and Momburg, F. (1997). A Viral ER-Resident Glycoprotein Inactivates the MHC-Encoded Peptide Transporter. Immunity 6, 623–632.

Hengel, H., Koszinowski, U.H., and Conzelmann, K.-K. (2005). Viruses know it all: new insights into IFN networks. Trends Immunol. 26, 396–401.

Hermel, E., Robinson, P.J., She, J.X., and Lindahl, K.F. (1993). Sequence divergence of B2m alleles of wild Mus musculus and Mus spretus implies positive selection. Immunogenetics 38, 106–116.

149 References

Herrmann, F., Trowsdale, J., Huber, C., and Seliger, B. (2003). Cloning and functional analyses of the mouse tapasin promoter. Immunogenetics 55, 379–388.

Hetz, C. (2012). The unfolded protein response: controlling cell fate decisions under ER stress and beyond. Nat. Rev. Mol. Cell Biol. 13, 89–102.

Hill, A., Jugovic, P., York, I., Russ, G., Bennink, J., Yewdell, J., Ploegh, H., and Johnson, D. (1995). Herpes simplex virus turns off the TAP to evade host immunity. Nature 375, 411–415.

Hislop, A.D., Ressing, M.E., Leeuwen, D. van, Pudney, V.A., Horst, D., Koppers-Lalic, D., Croft, N.P., Neefjes, J.J., Rickinson, A.B., and Wiertz, E.J.H.J. (2007). A CD8+ T cell immune evasion protein specific to Epstein-Barr virus and its close relatives in Old World primates. J. Exp. Med. 204, 1863–1873.

Hoeben, R.C., Migchielsen, A.A., van der Jagt, R.C., van Ormondt, H., and van der Eb, A.J. (1991). Inactivation of the Moloney murine leukemia virus long terminal repeat in murine fibroblast cell lines is associated with methylation and dependent on its chromosomal position. J. Virol. 65, 904–912.

Holtappels, R., Podlech, J., Geginat, G., Steffens, H.P., Thomas, D., and Reddehase, M.J. (1998). Control of murine cytomegalovirus in the lungs: relative but not absolute immunodominance of the immediate- early 1 nonapeptide during the antiviral cytolytic T-lymphocyte response in pulmonary infiltrates. J. Virol. 72, 7201–7212.

Holtappels, R., Podlech, J., Pahl-Seibert, M.-F., Jülch, M., Thomas, D., Simon, C.O., Wagner, M., and Reddehase, M.J. (2004). Cytomegalovirus misleads its host by priming of CD8 T cells specific for an epitope not presented in infected tissues. J. Exp. Med. 199, 131–136.

Hosokawa, N., Wada, I., Hasegawa, K., Yorihuzi, T., Tremblay, L.O., Herscovics, A., and Nagata, K. (2001). A novel ER alpha-mannosidase-like protein accelerates ER-associated degradation. EMBO Rep. 2, 415– 422.

Howcroft, T.K., Strebel, K., Martin, M.A., and Singer, D.S. (1993). Repression of MHC class I gene promoter activity by two-exon Tat of HIV. Science 260, 1320–1322.

Howcroft, T.K., Raval, A., Weissman, J.D., Gegonne, A., and Singer, D.S. (2003). Distinct transcriptional pathways regulate basal and activated major histocompatibility complex class I expression. Mol. Cell. Biol. 23, 3377–3391.

Hsu, V.W., Yuan, L.C., Nuchtern, J.G., Lippincott-Schwartz, J., Hammerling, G.J., and Klausner, R.D. (1991). A recycling pathway between the endoplasmic reticulum and the Golgi apparatus for retention of unassembled MHC class I molecules. Nature 352, 441–444.

Hughes, E.A., and Cresswell (1998). The thiol oxidoreductase ERp57 is a component of the MHC class I peptide-loading complex. Curr. Biol. CB 8, 709–712.

150 References

Hughes, E.A., Hammond, C., and Cresswell, P. (1997). Misfolded major histocompatibility complex class I heavy chains are translocated into the cytoplasm and degraded by the proteasome. Proc. Natl. Acad. Sci. U. S. A. 94, 1896–1901.

Huizen, R. van, Martindale, J.L., Gorospe, M., and Holbrook, N.J. (2003). P58IPK, a Novel Endoplasmic Reticulum Stress-inducible Protein and Potential Negative Regulator of eIF2α Signaling. J. Biol. Chem. 278, 15558–15564.

Hulpke, S., and Tampé, R. (2013). The MHC I loading complex: a multitasking machinery in adaptive immunity. Trends Biochem. Sci. 38, 412–420.

Hung, J.-H., Su, I.-J., Lei, H.-Y., Wang, H.-C., Lin, W.-C., Chang, W.-T., Huang, W., Chang, W.-C., Chang, Y.- S., Chen, C.-C., et al. (2004). Endoplasmic Reticulum Stress Stimulates the Expression of Cyclooxygenase- 2 through Activation of NF-κB and pp38 Mitogen-activated Protein Kinase. J. Biol. Chem. 279, 46384– 46392.

Inoue, M., Mimura, K., Izawa, S., Shiraishi, K., Inoue, A., Shiba, S., Watanabe, M., Maruyama, T., Kawaguchi, Y., Inoue, S., et al. (2012). Expression of MHC Class I on breast cancer cells correlates inversely with HER2 expression. Oncoimmunology 1, 1104–1110.

Jackson, M.R., Cohen-Doyle, M.F., Peterson, P.A., and Williams, D.B. (1994). Regulation of MHC class I transport by the molecular chaperone, calnexin (p88, IP90). Science 263, 384–387.

Jackson, S.E., Mason, G.M., and Wills, M.R. (2011). Human cytomegalovirus immunity and immune evasion. Virus Res. 157, 151–160.

Jamieson, A.M., Diefenbach, A., McMahon, C.W., Xiong, N., Carlyle, J.R., and Raulet, D.H. (2002). The role of the NKG2D immunoreceptor in immune cell activation and natural killing. Immunity 17, 19–29.

Janeway, C.A., Travers, P., Walport, M., Shlomchik, M.J., Jr, C.A.J., Travers, P., Walport, M., and Shlomchik, M.J. (2001a). Immunobiology (Garland Science).

Janeway, J., Travers, P., Walport, M., and Shlomchik, M.J. (2001b). Induced innate responses to infection.

Janßen, L., Ramnarayan, V.R., Aboelmagd, M., Iliopoulou, M., Hein, Z., Majoul, I., Fritzsche, S., Halenius, A., and Springer, S. (2016). The murine cytomegalovirus immunoevasin gp40 binds MHC class I molecules to retain them in the early secretory pathway. J. Cell Sci. 129, 219–227.

Johnson, D.R., and Pober, J.S. (1990). Tumor necrosis factor and immune interferon synergistically increase transcription of HLA class I heavy- and light-chain genes in vascular endothelium. Proc. Natl. Acad. Sci. 87, 5183–5187.

Johnson, K.E., Song, B., and Knipe, D.M. (2008). Role for herpes simplex virus 1 ICP27 in the inhibition of type I interferon signaling. Virology 374, 487–494.

Jonjid, S., Mutter, W., Weiland, F., Reddehase, M.J., and Koszinowski, U.H. (1989). Site-restricted persistent cytomegalovirus infection after selective long-term depletion of CD4+ T lymphocytes. J. Exp. Med. 169, 1199–1212.

151 References

Jonjid, S., Pavid, I., Lucin, P., Rukavina, D., and Koszinowski, U.H. (1990). Efficacious control of cytomegalovirus infection after long-term depletion of CD8+ T lymphocytes. J. Virol. 64, 5457–5464.

Joyner, A., Yamamoto, Y., and Bernstein, A. (1982). Retrovirus long terminal repeats activate expression of coding sequences for the herpes simplex virus thymidine kinase gene. Proc. Natl. Acad. Sci. U. S. A. 79, 1573–1577.

Kärre, K., Ljunggren, H.G., Piontek, G., and Kiessling, R. (1986). Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. Nature 319, 675–678.

Kasahara, M., Hayashi, M., Tanaka, K., Inoko, H., Sugaya, K., Ikemura, T., and Ishibashi, T. (1996). Chromosomal localization of the proteasome Z subunit gene reveals an ancient chromosomal duplication involving the major histocompatibility complex. Proc. Natl. Acad. Sci. U. S. A. 93, 9096–9101.

Kim, S., Lee, S., Shin, J., Kim, Y., Evnouchidou, I., Kim, D., Kim, Y.-K., Kim, Y.-E., Ahn, J.-H., Riddell, S.R., et al. (2011). Human cytomegalovirus microRNA miR-US4-1 inhibits CD8+ T cell responses by targeting the aminopeptidase ERAP1. Nat. Immunol. 12, 984–991.

Kincaid, R.P., and Sullivan, C.S. (2012). Virus-Encoded microRNAs: An Overview and a Look to the Future. PLoS Pathog 8, e1003018.

Kit, S., Dubbs, D.R., Piekarski, L.J., and Hsu, T.C. (1963). DELETION OF THYMIDINE KINASE ACTIVITY FROM L CELLS RESISTANT TO BROMODEOXYURIDINE. Exp. Cell Res. 31, 297–312.

Kleijnen, M.F., Huppa, J.B., Lucin, P., Mukherjee, S., Farrell, H., Campbell, A.E., Koszinowski, U.H., Hill, A.B., and Ploegh, H.L. (1997). A mouse cytomegalovirus glycoprotein, gp34, forms a complex with folded class I MHC molecules in the ER which is not retained but is transported to the cell surface. EMBO J. 16, 685–694.

Koppers-Lalic, D., Reits, E.A.J., Ressing, M.E., Lipinska, A.D., Abele, R., Koch, J., Rezende, M.M., Admiraal, P., Leeuwen, D. van, Bienkowska-Szewczyk, K., et al. (2005). Varicelloviruses avoid T cell recognition by UL49.5-mediated inactivation of the transporter associated with antigen processing. Proc. Natl. Acad. Sci. U. S. A. 102, 5144–5149.

Kraus, A., Groenendyk, J., Bedard, K., Baldwin, T.A., Krause, K.-H., Dubois-Dauphin, M., Dyck, J., Rosenbaum, E.E., Korngut, L., Colley, N.J., et al. (2010). Calnexin deficiency leads to dysmyelination. J. Biol. Chem. 285, 18928–18938.

Krmpotic, A., Messerle, M., Crnkovic-Mertens, I., Polic, B., Jonjic, S., and Koszinowski, U.H. (1999). The immunoevasive function encoded by the mouse cytomegalovirus gene m152 protects the virus against T cell control in vivo. J. Exp. Med. 190, 1285–1296.

Krmpotic, A., Hasan, M., Loewendorf, A., Saulig, T., Halenius, A., Lenac, T., Polic, B., Bubic, I., Kriegeskorte, A., Pernjak-Pugel, E., et al. (2005). NK cell activation through the NKG2D ligand MULT-1 is

152 References selectively prevented by the glycoprotein encoded by mouse cytomegalovirus gene m145. J. Exp. Med. 201, 211–220.

Ladasky, J.J., Boyle, S., Seth, M., Li, H., Pentcheva, T., Abe, F., Steinberg, S.J., and Edidin, M. (2006). Bap31 enhances the endoplasmic reticulum export and quality control of human class I MHC molecules. J. Immunol. Baltim. Md 1950 177, 6172–6181.

Landais, I., and Nelson, J.A. (2013). Functional genomics approaches to understand cytomegalovirus replication, latency and pathogenesis. Curr. Opin. Virol. 3, 408–415.

Landais, I., Pelton, C., Streblow, D., DeFilippis, V., McWeeney, S., and Nelson, J.A. (2015). Human Cytomegalovirus miR-UL112-3p Targets TLR2 and Modulates the TLR2/IRAK1/NFκB Signaling Pathway. PLoS Pathog. 11, e1004881.

Lankat-Buttgereit, B., and Tampé, R. (1999). The transporter associated with antigen processing TAP: structure and function. FEBS Lett. 464, 108–112.

Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., et al. (2007). Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947–2948.

Lee, A.-H., Iwakoshi, N.N., and Glimcher, L.H. (2003a). XBP-1 regulates a subset of endoplasmic reticulum resident chaperone genes in the unfolded protein response. Mol. Cell. Biol. 23, 7448–7459.

Lee, N., Iyer, S.S., Mu, J., Weissman, J.D., Ohali, A., Howcroft, T.K., Lewis, B.A., and Singer, D.S. (2010). Three novel downstream promoter elements regulate MHC class I promoter activity in mammalian cells. PloS One 5, e15278.

Lee, S., Park, B., and Ahn, K. (2003b). Determinant for endoplasmic reticulum retention in the luminal domain of the human cytomegalovirus US3 glycoprotein. J. Virol. 77, 2147–2156.

Lehner, P.J., Karttunen, J.T., Wilkinson, G.W., and Cresswell, P. (1997). The human cytomegalovirus US6 glycoprotein inhibits transporter associated with antigen processing-dependent peptide translocation. Proc. Natl. Acad. Sci. U. S. A. 94, 6904–6909.

Lehner, P.J., Surman, M.J., and Cresswell, P. (1998). Soluble tapasin restores MHC class I expression and function in the tapasin-negative cell line .220. Immunity 8, 221–231.

Lemke, H., Hämmerling, G.J., and Hämmerling, U. (1979). Fine specificity analysis with monoclonal antibodies of antigens controlled by the major histocompatibility complex and by the Qa/TL region in mice. Immunol. Rev. 47, 175–206.

Leone, P., Shin, E.-C., Perosa, F., Vacca, A., Dammacco, F., and Racanelli, V. (2013). MHC Class I Antigen Processing and Presenting Machinery: Organization, Function, and Defects in Tumor Cells. J. Natl. Cancer Inst.

Levitskaya, J., Coram, M., Levitsky, V., Imreh, S., Steigerwald-Mullen, P.M., Klein, G., Kurilla, M.G., and Masucci, M.G. (1995). Inhibition of antigen processing by the internal repeat region of the Epstein-Barr virus nuclear antigen-1. Nature 375, 685–688.

153 References

Li, S., Paulsson, K.M., Sjögren, H.O., and Wang, P. (1999). Peptide-bound major histocompatibility complex class I molecules associate with tapasin before dissociation from transporter associated with antigen processing. J. Biol. Chem. 274, 8649–8654.

Lindquist, J.A., Hämmerling, G.J., and Trowsdale, J. (2001). ER60/ERp57 forms disulfide-bonded intermediates with MHC class I heavy chain. FASEB J. Off. Publ. Fed. Am. Soc. Exp. Biol. 15, 1448–1450.

Lippincott-Schwartz, J., Donaldson, J.G., Schweizer, A., Berger, E.G., Hauri, H.-P., Yuan, L.C., and Klausner, R.D. (1990). Microtubule-dependent retrograde transport of proteins into the ER in the presence of brefeldin a suggests an ER recycling pathway. Cell 60, 821–836.

Little, A.M., Nössner, E., and Parham, P. (1995). Dissociation of beta 2-microglobulin from HLA class I heavy chains correlates with acquisition of epitopes in the cytoplasmic tail. J. Immunol. Baltim. Md 1950 154, 5205–5215.

Ljunggren, H.G., and Kärre, K. (1990). In search of the “missing self”: MHC molecules and NK cell recognition. Immunol. Today 11, 237–244.

Ljunggren, H.G., Stam, N.J., Ohlén, C., Neefjes, J.J., Höglund, P., Heemels, M.T., Bastin, J., Schumacher, T.N., Townsend, A., and Kärre, K. (1990). Empty MHC class I molecules come out in the cold. Nature 346, 476–480.

Lodoen, M., Ogasawara, K., Hamerman, J.A., Arase, H., Houchins, J.P., Mocarski, E.S., and Lanier, L.L. (2003). NKG2D-mediated natural killer cell protection against cytomegalovirus is impaired by viral gp40 modulation of retinoic acid early inducible 1 gene molecules. J. Exp. Med. 197, 1245–1253.

Lodoen, M.B., Abenes, G., Umamoto, S., Houchins, J.P., Liu, F., and Lanier, L.L. (2004). The cytomegalovirus m155 gene product subverts natural killer cell antiviral protection by disruption of H60- NKG2D interactions. J. Exp. Med. 200, 1075–1081.

Ma, W., Lehner, P.J., Cresswell, P., Pober, J.S., and Johnson, D.R. (1997). Interferon-γ Rapidly Increases Peptide Transporter (TAP) Subunit Expression and Peptide Transport Capacity in Endothelial Cells. J. Biol. Chem. 272, 16585–16590.

Machold, R.P., Andrée, S., Van Kaer, L., Ljunggren, H.G., and Ploegh, H.L. (1995). Peptide influences the folding and intracellular transport of free major histocompatibility complex class I heavy chains. J. Exp. Med. 181, 1111–1122.

Maley, F., Trimble, R.B., Tarentino, A.L., and Plummer, T.H. (1989). Characterization of glycoproteins and their associated oligosaccharides through the use of endoglycosidases. Anal. Biochem. 180, 195–204.

Marchetti, B., Ashrafi, G.H., Tsirimonaki, E., O’Brien, P.M., and Campo, M.S. (2002). The bovine papillomavirus oncoprotein E5 retains MHC class I molecules in the Golgi apparatus and prevents their transport to the cell surface. Oncogene 21, 7808–7816.

Marchetti, B., Gault, E.A., Cortese, M.S., Yuan, Z., Ellis, S.A., Nasir, L., and Campo, M.S. (2009). Bovine papillomavirus type 1 oncoprotein E5 inhibits equine MHC class I and interacts with equine MHC I heavy chain. J. Gen. Virol. 90, 2865–2870.

154 References

Martin, B.K., Chin, K.C., Olsen, J.C., Skinner, C.A., Dey, A., Ozato, K., and Ting, J.P. (1997). Induction of MHC class I expression by the MHC class II transactivator CIITA. Immunity 6, 591–600.

McCoy, W.H., 4th, Wang, X., Yokoyama, W.M., Hansen, T.H., and Fremont, D.H. (2012). Structural Mechanism of ER Retrieval of MHC Class I by Cowpox. PLoS Biol. 10, e1001432.

Meissner, T.B., Li, A., Biswas, A., Lee, K.-H., Liu, Y.-J., Bayir, E., Iliopoulos, D., van den Elsen, P.J., and Kobayashi, K.S. (2010). NLR family member NLRC5 is a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. U. S. A. 107, 13794–13799.

Messerle, M., Crnkovic, I., Hammerschmidt, W., Ziegler, H., and Koszinowski, U.H. (1997). Cloning and mutagenesis of a herpesvirus genome as an infectious bacterial artificial chromosome. Proc. Natl. Acad. Sci. U. S. A. 94, 14759–14763.

Meurs, E., Chong, K., Galabru, J., Thomas, N.S.B., Kerr, I.M., Williams, B.R.G., and Hovanessian, A.G. (1990). Molecular cloning and characterization of the human double-stranded RNA-activated protein kinase induced by interferon. Cell 62, 379–390.

Meylan, E., Curran, J., Hofmann, K., Moradpour, D., Binder, M., Bartenschlager, R., and Tschopp, J. (2005). Cardif is an adaptor protein in the RIG-I antiviral pathway and is targeted by hepatitis C virus. Nature 437, 1167–1172.

Miller, D.M., Zhang, Y., Rahill, B.M., Waldman, W.J., and Sedmak, D.D. (1999). Human cytomegalovirus inhibits IFN-alpha-stimulated antiviral and immunoregulatory responses by blocking multiple levels of IFN-alpha signal transduction. J. Immunol. Baltim. Md 1950 162, 6107–6113.

Mitra, A.K., Célia, H., Ren, G., Luz, J.G., Wilson, I.A., and Teyton, L. (2004). Supine orientation of a murine MHC class I molecule on the membrane bilayer. Curr. Biol. CB 14, 718–724.

Mocarski Jr., E.S. (2007). Betaherpes viral genes and their functions. In Human Herpesviruses: Biology, Therapy, and Immunoprophylaxis, A. Arvin, G. Campadelli-Fiume, E. Mocarski, P.S. Moore, B. Roizman, R. Whitley, and K. Yamanishi, eds. (Cambridge: Cambridge University Press), p.

Montealegre, S., Venugopalan, V., Fritzsche, S., Kulicke, C., Hein, Z., and Springer, S. (2015). Dissociation of β2-microglobulin determines the surface quality control of major histocompatibility complex class I molecules. FASEB J. Off. Publ. Fed. Am. Soc. Exp. Biol. 29, 2780–2788.

Munks, M.W., Pinto, A.K., Doom, C.M., and Hill, A.B. (2007). Viral interference with antigen presentation does not alter acute or chronic CD8 T cell immunodominance in murine cytomegalovirus infection. J. Immunol. Baltim. Md 1950 178, 7235–7241.

Myers, N.B., Lie, W.R., Nett, M., Rubocki, R.J., and Hansen, T.H. (1989). The conformation of Ld induced by beta 2-microglobulin is fixed during de novo synthesis and irreversible by exchange or dissociation. J. Immunol. Baltim. Md 1950 142, 2751–2758.

Nakamura, K., Zuppini, A., Arnaudeau, S., Lynch, J., Ahsan, I., Krause, R., Papp, S., De Smedt, H., Parys, J.B., Muller-Esterl, W., et al. (2001). Functional specialization of calreticulin domains. J. Cell Biol. 154, 961–972.

155 References

Neefjes, J.J., Momburg, F., and Hämmerling, G.J. (1993). Selective and ATP-dependent translocation of peptides by the MHC-encoded transporter. Science 261, 769–771.

Nelson, M.I., and Holmes, E.C. (2007). The evolution of epidemic influenza. Nat. Rev. Genet. 8, 196–205.

Niedermann, G. (2002). Immunological functions of the proteasome. Curr. Top. Microbiol. Immunol. 268, 91–136.

Nössner, E., and Parham, P. (1995). Species-specific differences in chaperone interaction of human and mouse major histocompatibility complex class I molecules. J. Exp. Med. 181, 327–337.

Offermann, M.K., and Faller, D.V. (1989). Autocrine induction of major histocompatibility complex class I antigen expression results from induction of beta interferon in oncogene-transformed BALB/c-3T3 cells. Mol. Cell. Biol. 9, 1969–1977.

Offermann, M.K., and Faller, D.V. (1990). Effect of cellular density and viral oncogenes on the major histocompatibility complex class I antigen response to gamma-interferon in BALB-c/3T3 cells. Cancer Res. 50, 601–605.

Ortmann, B., Androlewicz, M.J., and Cresswell, P. (1994). MHC class I/beta 2-microglobulin complexes associate with TAP transporters before peptide binding. Nature 368, 864–867.

Ou, W.J., Cameron, P.H., Thomas, D.Y., and Bergeron, J.J. (1993). Association of folding intermediates of glycoproteins with calnexin during protein maturation. Nature 364, 771–776.

Ozato, K., Hansen, T.H., and Sachs, D.H. (1980). Monoclonal antibodies to mouse MHC antigens. II. Antibodies to the H-2Ld antigen, the products of a third polymorphic locus of the mouse major histocompatibility complex. J. Immunol. Baltim. Md 1950 125, 2473–2477.

Pääbo, S., Bhat, B.M., Wold, W.S., and Peterson, P.A. (1987). A short sequence in the COOH-terminus makes an adenovirus membrane glycoprotein a resident of the endoplasmic reticulum. Cell 50, 311–317.

Pagny, S., Cabanes-Macheteau, M., Gillikin, J.W., Leborgne-Castel, N., Lerouge, P., Boston, R.S., Faye, L., and Gomord, V. (2000). Protein recycling from the Golgi apparatus to the endoplasmic reticulum in plants and its minor contribution to calreticulin retention. Plant Cell 12, 739–756.

Paladino, P., Collins, S.E., and Mossman, K.L. (2010). Cellular localization of the herpes simplex virus ICP0 protein dictates its ability to block IRF3-mediated innate immune responses. PloS One 5, e10428.

Pan, Y., Trojan, J., Guo, Y., and Anthony, D.D. (2013). Rescue of MHC-1 antigen processing machinery by down-regulation in expression of IGF-1 in human glioblastoma cells. PloS One 8, e58428.

Panniers, R., and Henshaw, E.C. (1983). A GDP/GTP exchange factor essential for eukaryotic initiation factor 2 cycling in Ehrlich ascites tumor cells and its regulation by eukaryotic initiation factor 2 phosphorylation. J. Biol. Chem. 258, 7928–7934.

156 References

Paquet, M.-E., Cohen-Doyle, M., Shore, G.C., and Williams, D.B. (2004). Bap29/31 influences the intracellular traffic of MHC class I molecules. J. Immunol. Baltim. Md 1950 172, 7548–7555.

Park, J.H., Chang, S.H., Kim, M.C., Shin, S.H., Youn, H.J., Kim, J.K., Jang, Y.S., and Kim, C.W. (1998). Up- regulation of the expression of major histocompatibility complex class I antigens by plasmid DNA transfection in non-hematopoietic cells. FEBS Lett. 436, 55–60.

Parker, K.C., DiBrino, M., Hull, L., and Coligan, J.E. (1992). The beta 2-microglobulin dissociation rate is an accurate measure of the stability of MHC class I heterotrimers and depends on which peptide is bound. J. Immunol. Baltim. Md 1950 149, 1896–1904.

Paulsson, K.M., Jevon, M., Wang, J.W., Li, S., and Wang, P. (2006). The double lysine motif of tapasin is a retrieval signal for retention of unstable MHC class I molecules in the endoplasmic reticulum. J. Immunol. Baltim. Md 1950 176, 7482–7488.

Peaper, D.R., Wearsch, P.A., and Cresswell, P. (2005). Tapasin and ERp57 form a stable disulfide-linked dimer within the MHC class I peptide-loading complex. EMBO J. 24, 3613–3623.

Pedersen, L.Ø., Stryhn, A., Holtet, T.L., Etzerodt, M., Gerwien, J., Nissen, M.H., Thøgersen, H.C., and Buus, S. (1995). The interaction of beta 2-microglobulin (β2m) with mouse class I major histocompatibility antigens and its ability to support peptide binding. A comparison of human and mouse β2m. Eur. J. Immunol. 25, 1609–1616.

Perez-Galarza, J., Carlotti, F., Rabelink, M.J., Cramer, S., Hoeben, R.C., Fibbe, W.E., and van Pel, M. (2014). Optimizing reporter constructs for in vivo bioluminescence imaging of interferon-γ stimulated mesenchymal stromal cells. Exp. Hematol. 42, 793–803.e1.

Pfeffer, S., Zavolan, M., Grässer, F.A., Chien, M., Russo, J.J., Ju, J., John, B., Enright, A.J., Marks, D., Sander, C., et al. (2004). Identification of virus-encoded microRNAs. Science 304, 734–736.

Pfeffer, S., Sewer, A., Lagos-Quintana, M., Sheridan, R., Sander, C., Grässer, F.A., van Dyk, L.F., Ho, C.K., Shuman, S., Chien, M., et al. (2005). Identification of microRNAs of the herpesvirus family. Nat. Methods 2, 269–276.

Platanias, L.C. (2005). Mechanisms of type-I- and type-II-interferon-mediated signalling. Nat. Rev. Immunol. 5, 375–386.

Polid, B., Hengel, H., Krmpotid, A., Trgovcich, J., Pavid, I., Luccaronin, P., Jonjid, S., and Koszinowski, U.H. (1998). Hierarchical and redundant lymphocyte subset control precludes cytomegalovirus replication during latent infection. J. Exp. Med. 188, 1047–1054.

Powers, C.J., and Früh, K. (2008). Signal Peptide-Dependent Inhibition of MHC Class I Heavy Chain Translation by Rhesus Cytomegalovirus. PLoS Pathog 4, e1000150.

Praveen, P.V.K., Yaneva, R., Kalbacher, H., and Springer, S. (2010). Tapasin edits peptides on MHC class I molecules by accelerating peptide exchange. Eur. J. Immunol. 40, 214–224.

Proffitt, J.A., and Blair, G.E. (1997). The MHC-encoded TAP1/LMP2 bidirectional promoter is downregulated in highly oncogenic adenovirus type 12 transformed cells. FEBS Lett. 400, 141–144.

157 References

Proffitt, J.L., Sharma, E., and Blair, G.E. (1994). Adenovirus 12-mediated down-regulation of the major histocompatibility complex (MHC) class I promoter: identification of a negative regulatory element responsive to Ad12 E1A. Nucleic Acids Res. 22, 4779–4788.

Puig, O., Caspary, F., Rigaut, G., Rutz, B., Bouveret, E., Bragado-Nilsson, E., Wilm, M., and Séraphin, B. (2001). The tandem affinity purification (TAP) method: a general procedure of protein complex purification. Methods San Diego Calif 24, 218–229.

Pyzik, M., Charbonneau, B., Gendron-Pontbriand, E.-M., Babid, M., Krmpotid, A., Jonjid, S., and Vidal, S.M. (2011). Distinct MHC class I-dependent NK cell-activating receptors control cytomegalovirus infection in different mouse strains. J. Exp. Med. 208, 1105–1117.

Qin, L., Ding, Y., Pahud, D.R., Chang, E., Imperiale, M.J., and Bromberg, J.S. (1997). Promoter attenuation in gene therapy: interferon-gamma and tumor necrosis factor-alpha inhibit transgene expression. Hum. Gene Ther. 8, 2019–2029.

Raulet, D.H. (2003). Roles of the NKG2D immunoreceptor and its ligands. Nat. Rev. Immunol. 3, 781–790.

Rawlinson, W.D., Farrell, H.E., and Barrell, B.G. (1996). Analysis of the complete DNA sequence of murine cytomegalovirus. J. Virol. 70, 8833–8849.

Reddehase, M.J., Keil, G.M., and Koszinowski, U.H. (1984). The cytolytic T lymphocyte response to the murine cytomegalovirus. II. Detection of virus replication stage-specific antigens by separate populations of in vivo active cytolytic T lymphocyte precursors. Eur. J. Immunol. 14, 56–61.

Remoli, A.L., Marsili, G., Perrotti, E., Gallerani, E., Ilari, R., Nappi, F., Cafaro, A., Ensoli, B., Gavioli, R., and Battistini, A. (2006). Intracellular HIV-1 Tat protein represses constitutive LMP2 transcription increasing proteasome activity by interfering with the binding of IRF-1 to STAT1. Biochem. J. 396, 371–380.

Reusch, U., Muranyi, W., Lucin, P., Burgert, H.G., Hengel, H., and Koszinowski, U.H. (1999). A cytomegalovirus glycoprotein re-routes MHC class I complexes to lysosomes for degradation. EMBO J. 18, 1081–1091.

Revilleza, M.J., Wang, R., Mans, J., Hong, M., Natarajan, K., and Margulies, D.H. (2011). How the virus outsmarts the host: function and structure of cytomegalovirus MHC-I-like molecules in the evasion of natural killer cell surveillance. J. Biomed. Biotechnol. 2011, 724607.

Richards, R., Scholz, I., Powers, C., Skach, W.R., and Früh, K. (2011). The Cytoplasmic Domain of Rhesus Cytomegalovirus Rh178 Interrupts Translation of Major Histocompatibility Class I Leader Peptide- Containing Proteins prior to Translocation. J. Virol. 85, 8766–8776.

Rizvi, S.M., and Raghavan, M. (2006). Direct peptide-regulatable interactions between MHC class I molecules and tapasin. Proc. Natl. Acad. Sci. U. S. A. 103, 18220–18225.

Romieu-Mourez, R., François, M., Boivin, M.-N., Stagg, J., and Galipeau, J. (2007). Regulation of MHC Class II Expression and Antigen Processing in Murine and Human Mesenchymal Stromal Cells by IFN-γ, TGF-β, and Cell Density. J. Immunol. 179, 1549–1558.

158 References

Roulston, A., Marcellus, R.C., and Branton, P.E. (1999). Viruses and Apoptosis. Annu. Rev. Microbiol. 53, 577–628.

Samali, A., Fitzgerald, U., Deegan, S., and Gupta, S. (2010). Methods for monitoring endoplasmic reticulum stress and the unfolded protein response. Int. J. Cell Biol. 2010, 830307.

Sansoni, P., Vescovini, R., Fagnoni, F.F., Akbar, A., Arens, R., Chiu, Y.-L., Čičin-Šain, L., Dechanet-Merville, J., Derhovanessian, E., Ferrando-Martinez, S., et al. (2014). New advances in CMV and immunosenescence. Exp. Gerontol. 55, 54–62.

Saraste, J., and Svensson, K. (1991). Distribution of the intermediate elements operating in ER to Golgi transport. J. Cell Sci. 100 ( Pt 3), 415–430.

Saric, T., Chang, S.-C., Hattori, A., York, I.A., Markant, S., Rock, K.L., Tsujimoto, M., and Goldberg, A.L. (2002). An IFN-gamma-induced aminopeptidase in the ER, ERAP1, trims precursors to MHC class I- presented peptides. Nat. Immunol. 3, 1169–1176.

Scalzo, A.A., Fitzgerald, N.A., Simmons, A., La Vista, A.B., and Shellam, G.R. (1990). Cmv-1, a genetic locus that controls murine cytomegalovirus replication in the spleen. J. Exp. Med. 171, 1469–1483.

Schoenborn, J.R., and Wilson, C.B. (2007). Regulation of interferon-gamma during innate and adaptive immune responses. Adv. Immunol. 96, 41–101.

Schoenhals, G.J., Krishna, R.M., Grandea, A.G., Spies, T., Peterson, P.A., Yang, Y., and Früh, K. (1999). Retention of empty MHC class I molecules by tapasin is essential to reconstitute antigen presentation in invertebrate cells. EMBO J. 18, 743–753.

Schuessler, A., Walker, D.G., and Khanna, R. (2014). Cytomegalovirus as a novel target for immunotherapy of glioblastoma multiforme. Front. Oncol. 4, 275.

Schust, D.J., Tortorella, D., Seebach, J., Phan, C., and Ploegh, H.L. (1998). Trophoblast class I major histocompatibility complex (MHC) products are resistant to rapid degradation imposed by the human cytomegalovirus (HCMV) gene products US2 and US11. J. Exp. Med. 188, 497–503.

Schwartz, O., Maréchal, V., Le Gall, S., Lemonnier, F., and Heard, J.M. (1996). Endocytosis of major histocompatibility complex class I molecules is induced by the HIV-1 Nef protein. Nat. Med. 2, 338–342.

Schwarz, K., Eggers, M., Soza, A., Koszinowski, U.H., Kloetzel, P.M., and Groettrup, M. (2000). The proteasome regulator PA28alpha/beta can enhance antigen presentation without affecting 20S proteasome subunit composition. Eur. J. Immunol. 30, 3672–3679.

Serwold, T., Gonzalez, F., Kim, J., Jacob, R., and Shastri, N. (2002). ERAAP customizes peptides for MHC class I molecules in the endoplasmic reticulum. Nature 419, 480–483.

Shental-Bechor, D., and Levy, Y. (2008). Effect of glycosylation on protein folding: A close look at thermodynamic stabilization. Proc. Natl. Acad. Sci. 105, 8256–8261.

Shields, M.J., Moffat, L.E., and Ribaudo, R.K. (1998). Functional comparison of bovine, murine, and human β2-microglobulin: interactions with murine MHC i molecules. Mol. Immunol. 35, 919–928.

159 References

Silva, C.A., Oliveira, A.C.P. de, Vilas-Boas, L., Fink, M.C.D.S., Pannuti, C.S., and Vidal, J.E. (2010). Neurologic cytomegalovirus complications in patients with AIDS: retrospective review of 13 cases and review of the literature. Rev. Inst. Med. Trop. São Paulo 52, 305–310.

Sinclair, J. (2008). Human cytomegalovirus: Latency and reactivation in the myeloid lineage. J. Clin. Virol. Off. Publ. Pan Am. Soc. Clin. Virol. 41, 180–185.

Singer, D.S., Mozes, E., Kirshner, S., and Kohn, L.D. (1997). Role of MHC class I molecules in autoimmune disease. Crit. Rev. Immunol. 17, 463–468.

Sissons, J.G.P., Bain, M., and Wills, M.R. (2002). Latency and Reactivation of Human Cytomegalovirus. J. Infect. 44, 73–77.

Smith, H.R.C., Heusel, J.W., Mehta, I.K., Kim, S., Dorner, B.G., Naidenko, O.V., Iizuka, K., Furukawa, H., Beckman, D.L., Pingel, J.T., et al. (2002). Recognition of a virus-encoded ligand by a natural killer cell activation receptor. Proc. Natl. Acad. Sci. U. S. A. 99, 8826–8831.

Smith, L.M., Shellam, G.R., and Redwood, A.J. (2006). Genes of murine cytomegalovirus exist as a number of distinct genotypes. Virology 352, 450–465.

Smith, M.H., Parker, J.M., Hodges, R.S., and Barber, B.H. (1986). The preparation and characterization of anti-peptide heteroantisera recognizing subregions of the intracytoplasmic domain of class I H-2 antigens. Mol. Immunol. 23, 1077–1092.

Smith, M.J., Basora, T., Kieran, J.E., and Nieto, M.C. (1993). Alteration within a discrete region of the H- 2Ld alpha 1 helix upon association with human beta 2 microglobulin. Immunol. Cell Biol. 71 ( Pt 2), 145– 149.

Soldà, T., Galli, C., Kaufman, R.J., and Molinari, M. (2007). Substrate-specific requirements for UGT1- dependent release from calnexin. Mol. Cell 27, 238–249.

Sousa, M., and Parodi, A.J. (1995). The molecular basis for the recognition of misfolded glycoproteins by the UDP-Glc:glycoprotein glucosyltransferase. EMBO J. 14, 4196–4203.

Spiliotis, E.T., Manley, H., Osorio, M., Zúñiga, M.C., and Edidin, M. (2000). Selective export of MHC class I molecules from the ER after their dissociation from TAP. Immunity 13, 841–851.

Stern-Ginossar, N., Elefant, N., Zimmermann, A., Wolf, D.G., Saleh, N., Biton, M., Horwitz, E., Prokocimer, Z., Prichard, M., Hahn, G., et al. (2007). Host immune system gene targeting by a viral miRNA. Science 317, 376–381.

Su, R.C., and Miller, R.G. (2001). Stability of surface H-2K(b), H-2D(b), and peptide-receptive H-2K(b) on splenocytes. J. Immunol. Baltim. Md 1950 167, 4869–4877.

Suh, W.K., Cohen-Doyle, M.F., Fruh, K., Wang, K., Peterson, P.A., and Williams, D.B. (1994). Interaction of MHC class I molecules with the transporter associated with antigen processing. Science 264, 1322–1326.

Symons, J.A., Alcamí, A., and Smith, G.L. (1995). Vaccinia virus encodes a soluble type I interferon receptor of novel structure and broad species specificity. Cell 81, 551–560.

160 References

Tabeta, K., Georgel, P., Janssen, E., Du, X., Hoebe, K., Crozat, K., Mudd, S., Shamel, L., Sovath, S., Goode, J., et al. (2004). Toll-like receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection. Proc. Natl. Acad. Sci. U. S. A. 101, 3516–3521.

Tan, P., Kropshofer, H., Mandelboim, O., Bulbuc, N., Hämmerling, G.J., and Momburg, F. (2002). Recruitment of MHC class I molecules by tapasin into the transporter associated with antigen processing-associated complex is essential for optimal peptide loading. J. Immunol. Baltim. Md 1950 168, 1950–1960.

Tector, M., Zhang, Q., and Salter, R.D. (1997). Beta 2-microglobulin and calnexin can independently promote folding and disulfide bond formation in class I histocompatibility proteins. Mol. Immunol. 34, 401–408.

Timms, R.T., Duncan, L.M., Tchasovnikarova, I.A., Antrobus, R., Smith, D.L., Dougan, G., Weekes, M.P., and Lehner, P.J. (2013). Haploid Genetic Screens Identify an Essential Role for PLP2 in the Downregulation of Novel Plasma Membrane Targets by Viral E3 Ubiquitin Ligases. PLoS Pathog 9, e1003772.

Tomasec, P., Braud, V.M., Rickards, C., Powell, M.B., McSharry, B.P., Gadola, S., Cerundolo, V., Borysiewicz, L.K., McMichael, A.J., and Wilkinson, G.W. (2000). Surface expression of HLA-E, an inhibitor of natural killer cells, enhanced by human cytomegalovirus gpUL40. Science 287, 1031.

Tortorella, D., Gewurz, B.E., Furman, M.H., Schust, D.J., and Ploegh, H.L. (2000). Viral Subversion of the Immune System. Annu. Rev. Immunol. 18, 861–926.

Townsend, A., Elliott, T., Cerundolo, V., Foster, L., Barber, B., and Tse, A. (1990). Assembly of MHC class I molecules analyzed in vitro. Cell 62, 285–295.

Trimble, R.B., and Tarentino, A.L. (1991). Identification of distinct endoglycosidase (endo) activities in Flavobacterium meningosepticum: endo F1, endo F2, and endo F3. Endo F1 and endo H hydrolyze only high mannose and hybrid glycans. J. Biol. Chem. 266, 1646–1651.

Tsai, B., Ye, Y., and Rapoport, T.A. (2002). Retro-translocation of proteins from the endoplasmic reticulum into the cytosol. Nat. Rev. Mol. Cell Biol. 3, 246–255.

Turnquist, H.R., Petersen, J.L., Vargas, S.E., McIlhaney, M.M., Bedows, E., Mayer, W.E., Grandea, A.G., Van Kaer, L., and Solheim, J.C. (2004). The Ig-like domain of tapasin influences intermolecular interactions. J. Immunol. Baltim. Md 1950 172, 2976–2984.

Umbach, J.L., Kramer, M.F., Jurak, I., Karnowski, H.W., Coen, D.M., and Cullen, B.R. (2008). MicroRNAs expressed by herpes simplex virus 1 during latent infection regulate viral mRNAs. Nature 454, 780–783. del Val, M., Hengel, H., Häcker, H., Hartlaub, U., Ruppert, T., Lucin, P., and Koszinowski, U.H. (1992). Cytomegalovirus prevents antigen presentation by blocking the transport of peptide-loaded major histocompatibility complex class I molecules into the medial-Golgi compartment. J. Exp. Med. 176, 729– 738.

Valente, G., Ozmen, L., Novelli, F., Geuna, M., Palestro, G., Forni, G., and Garotta, G. (1992). Distribution of interferon-gamma receptor in human tissues. Eur. J. Immunol. 22, 2403–2412.

161 References

Vassilakos, A., Cohen-Doyle, M.F., Peterson, P.A., Jackson, M.R., and Williams, D.B. (1996). The molecular chaperone calnexin facilitates folding and assembly of class I histocompatibility molecules. EMBO J. 15, 1495–1506.

Vembar, S.S., and Brodsky, J.L. (2008). One step at a time: endoplasmic reticulum-associated degradation. Nat. Rev. Mol. Cell Biol. 9, 944–957.

Vertegaal, A.C.O., Kuiperij, H.B., Houweling, A., Verlaan, M., Eb, A.J. van der, and Zantema, A. (2003). Differential Expression of Tapasin and Immunoproteasome Subunits in Adenovirus Type 5- Versus Type 12-transformed Cells. J. Biol. Chem. 278, 139–146.

Vigneron, N., and Van den Eynde, B.J. (2014). Proteasome Subtypes and Regulators in the Processing of Antigenic Peptides Presented by Class I Molecules of the Major Histocompatibility Complex. Biomolecules 4, 994–1025.

Vlodavsky, I., Fielding, P.E., Fielding, C.J., and Gospodarowicz, D. (1978). Role of contact inhibition in the regulation of receptor-mediated uptake of low density lipoprotein in cultured vascular endothelial cells. Proc. Natl. Acad. Sci. U. S. A. 75, 356–360.

Vogel, R., Al-Daccak, R., Drews, O., Alonzeau, J., Mester, G., Charron, D., Stevanovic, S., and Mallet, J. (2013). Mass spectrometry reveals changes in MHC I antigen presentation after lentivector expression of a gene regulation system. Mol. Ther. Nucleic Acids 2, e75.

Wagner, M., Gutermann, A., Podlech, J., Reddehase, M.J., and Koszinowski, U.H. (2002). Major histocompatibility complex class I allele-specific cooperative and competitive interactions between immune evasion proteins of cytomegalovirus. J. Exp. Med. 196, 805–816.

Wang, R., Natarajan, K., Revilleza, M.J.R., Boyd, L.F., Zhi, L., Zhao, H., Robinson, H., and Margulies, D.H. (2012). Structural basis of mouse cytomegalovirus m152/gp40 interaction with RAE1γ reveals a paradigm for MHC/MHC interaction in immune evasion. Proc. Natl. Acad. Sci. U. S. A. 109, E3578-3587.

Wearsch, P.A., and Cresswell, P. (2007). Selective loading of high-affinity peptides onto major histocompatibility complex class I molecules by the tapasin-ERp57 heterodimer. Nat. Immunol. 8, 873– 881.

Wearsch, P.A., Peaper, D.R., and Cresswell, P. (2011). Essential glycan-dependent interactions optimize MHC class I peptide loading. Proc. Natl. Acad. Sci. U. S. A. 108, 4950–4955. van de Weijer, M.L., Luteijn, R.D., and Wiertz, E.J.H.J. (2015). Viral immune evasion: Lessons in MHC class I antigen presentation. Semin. Immunol. 27, 125–137.

Wier, M.L., and Edidin, M. (1986). Effects of cell density and extracellular matrix on the lateral diffusion of major histocompatibility antigens in cultured fibroblasts. J. Cell Biol. 103, 215–222.

Wiertz, E.J., Tortorella, D., Bogyo, M., Yu, J., Mothes, W., Jones, T.R., Rapoport, T.A., and Ploegh, H.L. (1996a). Sec61-mediated transfer of a membrane protein from the endoplasmic reticulum to the proteasome for destruction. Nature 384, 432–438.

162 References

Wiertz, E.J., Jones, T.R., Sun, L., Bogyo, M., Geuze, H.J., and Ploegh, H.L. (1996b). The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84, 769–779.

Williams, A.P., Peh, C.A., Purcell, A.W., McCluskey, J., and Elliott, T. (2002). Optimization of the MHC class I peptide cargo is dependent on tapasin. Immunity 16, 509–520.

Williams, D.B., Swiedler, S.J., and Hart, G.W. (1985). Intracellular transport of membrane glycoproteins: two closely related histocompatibility antigens differ in their rates of transit to the cell surface. J. Cell Biol. 101, 725–734.

Williams, D.B., Barber, B.H., Flavell, R.A., and Allen, H. (1989). Role of beta 2-microglobulin in the intracellular transport and surface expression of murine class I histocompatibility molecules. J. Immunol. Baltim. Md 1950 142, 2796–2806.

Wright, K.L., White, L.C., Kelly, A., Beck, S., Trowsdale, J., and Ting, J.P. (1995). Coordinate regulation of the human TAP1 and LMP2 genes from a shared bidirectional promoter. J. Exp. Med. 181, 1459–1471.

Xaus, J., Comalada, M., Barrachina, M., Herrero, C., Goñalons, E., Soler, C., Lloberas, J., and Celada, A. (2000). The expression of MHC class II genes in macrophages is cell cycle dependent. J. Immunol. Baltim. Md 1950 165, 6364–6371.

Xiao, C., and Rajewsky, K. (2009). MicroRNA Control in the Immune System: Basic Principles. Cell 136, 26–36.

Yamamoto, K., Sato, T., Matsui, T., Sato, M., Okada, T., Yoshida, H., Harada, A., and Mori, K. (2007). Transcriptional induction of mammalian ER quality control proteins is mediated by single or combined action of ATF6alpha and XBP1. Dev. Cell 13, 365–376.

Yewdell, J.W., and Nicchitta, C.V. (2006). The DRiP hypothesis decennial: support, controversy, refinement and extension. Trends Immunol. 27, 368–373.

Yoon, S.I., Jones, B.C., Logsdon, N.J., Harris, B.D., Kuruganti, S., and Walter, M.R. (2012). Epstein-Barr virus IL-10 engages IL-10R1 by a two-step mechanism leading to altered signaling properties. J. Biol. Chem. 287, 26586–26595.

York, I.A., Roop, C., Andrews, D.W., Riddell, S.R., Graham, F.L., and Johnson, D.C. (1994). A cytosolic herpes simplex virus protein inhibits antigen presentation to CD8+ T lymphocytes. Cell 77, 525–535.

Yu, Y.Y.L., Turnquist, H.R., Myers, N.B., Balendiran, G.K., Hansen, T.H., and Solheim, J.C. (1999). An Extensive Region of an MHC Class I α2 Domain Loop Influences Interaction with the Assembly Complex. J. Immunol. 163, 4427–4433.

Yu, Y.Y.L., Harris, M.R., Lybarger, L., Kimpler, L.A., Myers, N.B., Virgin, H.W., and Hansen, T.H. (2002). Physical association of the K3 protein of gamma-2 herpesvirus 68 with major histocompatibility complex class I molecules with impaired peptide and beta(2)-microglobulin assembly. J. Virol. 76, 2796–2803.

Zagorac, G.B., Mahmutefendid, H., Tomaš, M.I., Kučid, N., Le Bouteiller, P., and Lučin, P. (2012). Early endosomal rerouting of major histocompatibility class I conformers. J. Cell. Physiol. 227, 2953–2964.

163 References

Zanker, D., Waithman, J., Yewdell, J.W., and Chen, W. (2013). Mixed Proteasomes Function To Increase Viral Peptide Diversity and Broaden Antiviral CD8+ T Cell Responses. J. Immunol. 191, 52–59.

Zeidler, R., Eissner, G., Meissner, P., Uebel, S., Tampé, R., Lazis, S., and Hammerschmidt, W. (1997). Downregulation of TAP1 in B lymphocytes by cellular and Epstein-Barr virus-encoded interleukin-10. Blood 90, 2390–2397.

Zernich, D., Purcell, A.W., Macdonald, W.A., Kjer-Nielsen, L., Ely, L.K., Laham, N., Crockford, T., Mifsud, N.A., Bharadwaj, M., Chang, L., et al. (2004). Natural HLA class I polymorphism controls the pathway of antigen presentation and susceptibility to viral evasion. J. Exp. Med. 200, 13–24.

Zijlstra, M., Bix, M., Simister, N.E., Loring, J.M., Raulet, D.H., and Jaenisch, R. (1990). Beta 2- microglobulin deficient mice lack CD4-8+ cytolytic T cells. Nature 344, 742–746.

164