Immunoproteasome Function in Lymphocyte Activation

Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaft (Dr. rer. nat.)

vorgelegt von Christian Schmidt

an der

Mathematisch-Naturwissenschaftliche Sektion

Fachbereich Biologie

Konstanz, 2018

Konstanzer Online-Publikations-System (KOPS) URL: http://nbn-resolving.de/urn:nbn:de:bsz:352-2-3cw92n75tqfq9

Tag der mündlichen Prüfung: 11. Januar 2019

1. Referent: Herr Privatdozent Dr. Michael Basler 2. Referent: Herr Professor Dr. Martin Scheffner

TABLE OF CONTENTS

Zusammenfassung in deutscher Sprache……………………………………………………………..……..8 Abstract……………………………………………………………………………………………………..10 CHAPTER I - INTRODUCTION 1 Proteasomes and Immunoproteasomes ...... 12

1.1 The proteasome and the ubiquitin proteasome system ...... 12

1.2 Structure of standard proteasome and immunoproteasome ...... 13

1.2.1 Structure and dynamics of the 20S core particle ...... 13

1.2.2 Peptide hydrolysis in the 20S CP ...... 14

1.2.3 Assembly of proteasomes and immunoproteasomes ...... 15

1.2.4 Role of regulatory particles for proteasome and immunoproteasome function ...... 18

1.3 Tissue expression of immunoproteasomes ...... 20

1.4 Proteasomes and immunoproteasomes in MHC-I antigen presentation ...... 20

1.5 Proteasomes and immunoproteasomes in cellular homeostasis and signal transduction ...... 22

1.5.1 Role of proteasomes for nutrient availability and amino acid recycling ...... 22

1.5.2 Removal of misfolded, damaged and oxidized proteins ...... 23

1.5.3 Role of proteasomes in signal transduction ...... 24

1.6 Proteostasis stress, stress response pathways and apoptosis ...... 25

1.7 Proteasome and immunoproteasome inhibitors ...... 29

1.8 Proteasomes and immunoproteasomes in pathophysiology and aging ...... 32

1.8.1 Proteasomes and immunoproteasomes in neurodegenerative diseases ...... 32

1.8.2 Proteasomes and immunoproteasomes in infectious diseases ...... 34

1.8.3 Proteasomes and immunoproteasomes in malignant diseases and transplant rejection ...... 35

1.8.4 Proteasomes and immunoproteasomes in autoimmune diseases ...... 36

2 Lymphocytes and their role in autoimmune diseases ...... 37

2.1 B cells and humoral immunity ...... 37

2.1.1 B cell development and activation ...... 37 3

2.1.2 Plasma cells and autoantibodies in autoimmune diseases and transplant rejection ...... 38

2.2 T cells and T helper cell subsets ...... 39

2.2.1 Th1, Th2 and Th17 polarization ...... 41

2.2.2 Regulatory T cells ...... 43

2.3 Signal transduction via the T cell receptor and co-stimulatory molecules ...... 44

2.3.1 Signaling initiation and proximal signaling ...... 44

2.3.2 Co-stimulatory signaling ...... 46

2.3.3 Nuclear factor kappa B (NF-κB) signaling in T cells ...... 48

2.3.4 Calcium signaling and nuclear factor of activated T cells (NFAT) ...... 50

2.3.5 MAP-kinase signaling pathways ...... 51

2.3.6 Mechanistic target of rapamycin (mTOR) signaling and immunometabolism ...... 55

2.3.7 TCR signaling strength ...... 56

2.3.8 Functional outcome of T cell activation ...... 57

2.4 Dual specificity phosphatases and their role in T cell signaling ...... 58

2.4.1 Dual specificity phosphatase family overview ...... 59

2.4.2 Expression and potential functions of dual specificity phosphatases in T cells ...... 60

2.4.3 Regulation and functions of DUSP6 and DUSP5 in T cells ...... 62

3 Aim of this study ...... 64

CHAPTER II - MATERIALS AND METHODS

4 Materials ...... 66

4.1 Chemicals ...... 66

4.2 Disposables ...... 67

4.3 Kits and Reagents ...... 67

4.4 Buffers and solutions ...... 68

4.5 Oligonucleotides ...... 69

4.6 Antibodies for flow cytometry ...... 70

4.7 Antibodies for immunoblotting ...... 71 4

4.8 Antibodies for functional assays ...... 72

4.9 Mouse strains and animal protocols ...... 72

4.10 Lymphocytic choriomeningitis virus (LCMV) ...... 72

4.11 Cell lines ...... 72

4.12 Devices and Software ...... 73

5 Methods ...... 74

5.1 Cell culture and cell stimulation ...... 74

5.1.1 CTLL2 ...... 74

5.1.2 T1 cells ...... 74

5.1.3 Primary murine CD4+ T cells and CD19+ B cells ...... 74

5.1.4 Primary human CD4+ T cells and CD19+ B cells ...... 75

5.1.5 Mouse embryonic fibroblasts (MEFs) ...... 75

5.1.6 Ex vivo T cell expansion ...... 75

5.1.7 In vitro T helper cell polarization ...... 75

5.2 Determination of IL-2 activity in a CTLL-2 bioassay ...... 76

5.3 Cell viability assessment ...... 76

5.4 Inhibitor preparation ...... 76

5.5 Enzyme-linked immunosorbent assay (ELISA) ...... 77

5.6 CFSE proliferation assay ...... 77

5.7 Flow cytometry ...... 77

5.8 Preparation of 4% PFA solution in PBS ...... 78

5.9 Generation of cell lysates ...... 78

5.10 SDS-PAGE and Western blotting ...... 78

5.10.1 Enhanced chemiluminescence based immuno-detection ...... 79

5.10.2 Near-IR based immuno-detection ...... 79

5.11 Radioactive labeling and pulse-chase experiments ...... 80

5.12 RNA extraction and q-RT-PCR ...... 80 5

5.13 Mouse genotyping ...... 81

5.14 Agarose gel electrophoresis ...... 81

5.15 Statistical analysis and graphical data presentation ...... 81

CHAPTER III - RESULTS

6 ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition ...... 84

6.1 Activation of primary mouse T cells is ameliorated by ONX 0914 in an LMP7- and LMP2-co- dependent manner ...... 84

6.2 Naïve T and B cells contain almost only mixed and immunoproteasomes ...... 86

6.3 Impaired proliferation and impaired polarization of CD4+ T cells in vitro ...... 88

7 ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells ...... 92

7.1 ONX 0914 reduces IL-2 and CD69 expression at the mRNA level ...... 92

7.2 No influence of ONX 0914 on early canonical T cell activation pathways ...... 93

7.3 Analysis of ERK phosphorylation sustainment by quantitative near-IR immunoblotting and intracellular flow cytometry ...... 96

7.4 ONX 0914 causes mild protein homeostasis stress in activated T cells without induction of apoptosis ...... 98

7.5 CD4+ T cells overcome proteostasis stress likely via Nrf1-mediated standard proteasome up- regulation ...... 100

7.6 B cells are similarly affected by ONX 0914, but show higher susceptibility to apoptosis induction ...... 103

7.7 Induction of high immunoproteasome content in MEFs is sufficient to induce ubiquitin- conjugates after ONX 0914 treatment ...... 105

8 Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation ...... 106

8.1 Impaired ERK-phosphorylation after ONX 0914 treatment depends on de novo gene expression and correlates with DUSP6 accumulation ...... 106

8.2 DUSP6 degradation, but not expression is impaired by ONX 0914 treatment ...... 108

8.3 No evidence for a functional involvement of DUSP6 in ONX 0914-mediated amelioration of T cell activation ...... 110

9 Impaired in vivo activation of antigen-specific CD4+ T cells by ONX 0914 ...... 112 6

CHAPTER IV - DISCUSSION

10 Discussion ...... 114

10.1 Immunoproteasome inhibition and deficiency in T cell polarization ...... 118

10.2 Immunoproteasome inhibition in T cell activation signaling ...... 122

10.3 Immunoproteasome inhibition and its effects on proteostasis in lymphocyte activation ...... 128

10.4 Effects of mild proteostasis stress on T cell and B cell survival ...... 132

10.5 Possible signaling pathways involved in proteasome regulation after immunoproteasome inhibition ...... 136

10.6 Concluding remarks ...... 141

REFERENCES AND APPENDIX

11 References ...... 143 12 Appendix ...... 165

12.1 Appendix Figures ...... 165

12.2 Abbreviations ...... 168

12.3 Table of Figures ...... 171

12.4 Record of Contribution ...... 172

12.5 List of publications and oral presentations ...... 173

Acknowledgements - Danksagung 174

ZUSAMMENFASSUNG IN DEUTSCHER SPRACHE

Das Immunproteasom stellt eine spezialisierte Form des Proteasoms dar, eines multimeren Proteinkomplexes, der wichtige Funktionen in allen eukaryotischen Zellen übernimmt. Innerhalb eines 20S Proteasom Kernpartikels schneiden drei verschiedene Proteasen die zum Abbau bestimmte Polypeptidkette, wobei jede der drei Proteasen jeweils einmal in einem halben 20S Proteasom Partikel vorkommt. Die Positionen der in Standard-Proteasomen aktiven Untereinheiten β1c, β2c und β5c werden in Immunproteasomen von alternativen Untereinheiten besetzt: LMP2 wird anstelle von β1c eingebaut, MECL-1 anstelle von β2c und LMP7 anstelle von β5c. Das Immunproteasom wird in Zellen hämatopoetischen Ursprungs exprimiert sowie im Gewebe unter dem Einfluss pro-entzündlicher Zytokine, maßgeblich Interferon-γ. Das Immunproteasom hat eine gut charakterisierte Funktion im Rahmen der MHC-I Antigen-Prozessierung und erwies sich zudem als vielversprechendes Medikationsziel zur Behandlung von Autoimmunerkrankungen, da Immunproteasom-Inhibition in zahlreichen präklinischen Modellen für Autoimmunität krankheitshemmende Wirkung zeigte. Seit der Beschreibung des selektiven Immunproteasominhibitors ONX 0914 im Jahr 2009 war der genaue Wirkmechanismus im Rahmen solcher Autoimmunerkrankungen weitgehend unbekannt. Vorherige Studien haben ergeben, dass intrinsische Effekte in aktivierten T Zellen, die in vielen Autoimmunerkrankungen eine wichtige Rolle spielen, am zugrundeliegenden Mechanismus beteiligt zu sein scheinen. Aufbauend auf den Vorarbeiten für dieses Projekt, die im Rahmen meiner Masterarbeit 2013 durchgeführt wurden, war es das Ziel der hier vorgelegten Arbeit, den Effekt von ONX 0914 Behandlung auf molekularer Ebene näher zu charakterisieren, um somit den Wirkmechanismus der Immunproteasom-Inhibition besser zu verstehen.

In der vorliegenden Arbeit wurden primäre T Zellen und B Zellen aus Mäusen und Menschen funktional untersucht. Vorherige Ergebnisse aus der Masterarbeit wie eine verringerte CD69 Expression nach T Zell Aktivierung unter Behandlung mit ONX 0914 wurden erneut bestätigt und zudem auch in B Zellen, in humanen Zellen und in antigen-spezifisch aktivierten T Zellen in vivo nachgewiesen. Ferner wurden die Effekte nicht nur phänotypisch charakterisiert, sondern durch Untersuchung der mRNA-Ebene ergänzt. Eine ausgiebige Analyse kanonischer Signalwege der T Zell Aktivierung erbrachte Hinweise auf eine um etwa 20 % reduzierte Phosphorylierung der Kinase ERK nach Behandlung der T Zellen mit ONX 0914. Die Reduktion der ERK-Phosphorylierung wurde durch quantitatives Immunoblotting mit nah-Infrarot- Farbstoffen, durch durchflusszytometrische Quantifizierung und schließlich durch Konfokalmikroskopie bestätigt. Es zeigte sich jedoch, dass die im Signalweg unmittelbar vorangestellte Kinase nicht von einer Reduktion betroffen war. Aus diesem Grund wurden zahlreiche Phosphatasen untersucht, die als Kandidaten für eine funktionale Beteiligung infrage kamen. Zwei der untersuchten Phosphatasen waren durch Behandlung mit ONX 0914 beeinträchtigt. Die Dual Specificity Phosphatase DUSP5 wurde verringert exprimiert, während DUSP6 auf der Proteinebene akkumulierte. In kombinierten Cycloheximid

und radioaktiven Markierungsuntersuchungen konnte der durch ONX 0914 verringerte, aber nicht vollständig blockierte Abbau von DUSP6 während der T Zell Aktivierung nachgewiesen werden. Daher wurde eine mögliche Rolle für DUSP6 bei der reduzierten T Zell Aktivierung mithilfe DUSP6-defizienter Mäuse untersucht. Hierbei zeigte sich, dass DUSP6 nicht für die vorliegenden Effekte verantwortlich ist oder durch Redundanz von einer anderen Phosphatase ersetzt werden kann. Abschließend konnte der genaue Mechanismus der reduzierten ERK-Phosphorylierung daher nicht geklärt werden.

Neben der veränderten T Zell Signaltransduktion lag der zweite Fokus in dieser Arbeit auf der Analyse der Proteostase während der T Zell Aktivierung nach Behandlung mit ONX 0914. Im Unterschied zu bereits vorher untersuchten Zelllinien zeigte sich in primären T Zellen und B Zellen eine Anreicherung von poly- Ubiquitin-Konjugaten nach Aktivierung, wenn die Zellen mit ONX 0914 behandelt wurden. Als wahrscheinliche Ursache für diesen Effekt wurden zwei Parameter identifiziert: Zum einen wurde gefunden, dass nahezu alle Proteasome in primären T und B Zellen Immunproteasome oder gemischte Proteasome mit LMP7 darstellen. Zum anderen zeigte sich im Verlauf der Arbeit, dass der als LMP7- selektiv beschriebene Inhibitor auch LMP2 co-inhibiert. Da aktivierte Lymphozyten zudem eine ausgeprägte metabolische und proteomische Re-organisation erfahren, machen diese Eigenschaften die Zellen anfällig für Proteostase-Stress durch Immunproteasom-Inhibitoren. Anschließend wurde die Auswirkung des Proteostase-Stresses auf die Aktivierung und die Vitalität der Zellen näher untersucht. Es zeigte sich, dass T Zellen ohne Induktion der Apoptose und ohne signifikante Aktivierung der Integrativen Stressantwort die Ubiquitin-Konjugate innerhalb von 20 Stunden abbauen konnten. Dabei wurde eine erhöhte Neusynthese der Standard-Proteasom-Untereinheit β5c detektiert sowie eine Anreicherung von löslichem Nrf1 in Zelllysaten. Diese Ergebnisse deuten darauf hin, dass T Zellen durch Nrf1-vermittelte Proteasom-Expression bei mildem Proteostase-Stress überleben, wobei die funktionale Zell-Aktivierung gleichzeitig gehemmt ist. Ähnliche Effekte zeigten sich in B Zellen, jedoch wurde eine Induktion von Apoptose-Markern nach ONX 0914 Behandlung in B Zellen gefunden.

Zusammenfassend ergibt sich aus der vorliegenden Arbeit, dass die Behandlung mit ONX 0914, aber nicht Immunproteasom-Defizienz, milden Proteostase-Stress in aktivierten Lymphozyten verursacht, der ihre Funktionsfähigkeit hemmt. Dieser Effekt ist wahrscheinlich durch die besonders hohe Abhängigkeit primärer Lymphozyten von LMP7-beinhaltenden Proteasomen und durch die Co-Inhibition von LMP2 begründet. Der dargelegte Mechanismus erklärt somit sehr wahrscheinlich zumindest einen Teil der entzündungshemmenden Eigenschaften von Immunproteasom Inhibitoren und warum Immunproteasom- Inhibitoren vielversprechend für die klinische Anwendung sind mit einer voraussichtlich geringeren Toxizität im Vergleich zu nicht-selektiven Proteasom-Inhibitoren.

ABSTRACT

The immunoproteasome constitutes a specialized form of the proteasome, a multimeric protein complex with important functions in eukaryotic cells. Within a 20S core particle three different protease types cleave the polypeptide chains destined for degradation. Each of the three proteases constitutes one position per half-core particle. The standard proteasome active subunits β1c, β2c and β5c are replaced by alternative subunits in immunoproteasomes: LMP2 is incorporated at the position of β1c, MECL-1 substitutes for β2c and LMP7 substitutes for β5c. The immunoproteasome is expressed in cells of hematopoietic origin as well as in peripheral tissues under the influence of pro-inflammatory cytokines, mainly interferon-γ. Besides a well characterized role in MHC-I antigen processing, a potential role of the immunoproteasome as a drug target for treatment of autoimmune diseases was shown, as immunoproteasome inhibition proved to have disease ameliorating effects in several pre-clinical models for autoimmunity. Since the description of the immunoproteasome-selective inhibitor ONX 0914 in 2009, the underlying mechanism in the course of these autoimmune-diseases has remained largely elusive. Previous studies have shown that intrinsic effects in activated T cells, which are involved in many autoimmune disaeses, are likely to be important for the underlying mechanism. Based on previous work, which was performed during the course of my master’s thesis 2013, the aim of this study was to characterize the effects of ONX 0914 treatment at the molecular level in more detail in order to improve our understanding of the underlying mechanism of immunoproteasome inhibition.

In this work, primary T and B cells from mice and humans were functionally investigated. Previous results from my master’s thesis like ameliorated CD69 up-regulation upon activation after ONX 0914 treatment were corroborated and additionally the effect was shown in B cells, in human cells and in antigen- specifically activated T cells in vivo. Furthermore, the effects were not only phenotypically characterized but substantiated by analysis at the mRNA level. An extended investigation into canonical signaling pathways of T cell activation indicated a reduction in phosphorylation of the kinase ERK after treatment with ONX 0914. This reduction was corroborated using quantitative near-infrared-dye based immunoblotting, flow cytometry and finally by confocal microscopy. However, the direct up-stream kinase was not found to be affected. Therefore, dual specificity phosphatases were identified as potential candidates for a functional involvement and several dual specificity phosphatases were analyzed at the protein level. Two of the analyzed phosphatases were affected by ONX 0914 treatment. The dual specificity phosphatase DUSP5 was less expressed, while DUSP6 accumulated at protein level. In combined cycloheximid and radioactive labelling approaches it was shown that DUSP6 degradation was impaired, but not fully blocked by ONX 0914 in T cell activation. Therfore, the possible involvement of DUSP6 for impaired T cell activation was investigated using DUSP6-deficient mice. However, it was found that DUSP6 was not responsible for the observed effects in a non-redundant manner and

compensation by other phosphatases cannot be ruled out. Therefore, the mechanism leading to reduced ERK-phosphorylation could not be fully unraveled so far.

Apart from altered T cell signaling, the second focus of this work was set on proteostasis regulation during T cell activation after ONX 0914 treatment. Unlike previously investigated T cell lines, primary T cells and B cells showed ubiquitin-conjugate accumulation after activation when cells had been pre-treated with ONX 0914. Two factors were identified likely underlying this effect: First, it was found that almost all proteasomes in T cells and B cells constituted of LMP7-containing immunoproteasomes or mixed proteasomes. Second, it was found that the reportedly LMP7-selective inhibitor ONX 0914 co-inhibited LMP2 as well. As activated lymphocytes show marked metabolic and proteomic re-organization, these features render the cells susceptible to proteostasis stress after immunoproteasome inhibition. Consequently, the effect of enhanced proteostasis stress on activation and cell viability was characterized. It was found that T cells could alleviate the enhanced ubiquitin-conjugates within 20 hours of activation without significant induction of the integrated stress response or apoptosis. Accompanying, an enhanced neosynthesis of the standard proteasome subunit β5c and accumulation of soluble Nrf1 were detected in cell lysates. These results indicate that T cells survive mild proteostasis stress after ONX 0914 treatment likely via Nrf1-mediated proteasome up-regulation, while functional T cell activation is impaired at the same time. Similar effects were found in B cells, but induction of apoptosis-markers after ONX 0914 treatment was detected in B cells.

Taken together, this work provides evidence that ONX 0914 treatment, but not immunoproteasome- deficiency, results in mild proteostasis stress in activated lymphocytes, impeding their functional capacity. This effect is likely attributed to the particularly high dependency of primary lymphocytes on LMP7- containing proteasomes and co-inhibition of LMP2 by ONX 0914. The provided mechanism hence likely explains at least parts of the anti-inflammatory effects of immunoproteasome inhibitors and why immunoproteasome inhibition shows high clinical potential with less overall toxicity as compared to broad-spectrum proteasome inhibitors.

Chapter I – Introduction – Proteasomes and Immunoproteasomes

CHAPTER I: INTRODUCTION

1 PROTEASOMES AND IMMUNOPROTEASOMES

1.1 The proteasome and the ubiquitin proteasome system

Protein quality control is regulated by several protein quality control systems involving e.g. chaperones, endoplasmatic reticulum (ER) stress response, autophagy and, importantly, the ubiquitin-proteasome system [1]. As the main cytosolic and nuclear protein degradation machinery the proteasome is of vital importance for eukaryotic cells. An intensively studied fashion of protein degradation is the ATP- dependent and proteasome mediated degradation after poly-ubiquitination, i.e. the attachment of several moieties of the small protein ubiquitin to the target marked for degradation; hence collectively referred to as the ubiquitin proteasome system (UPS) [2–4]. A cascade of three enzymes orchestrates the ubiquitination of a target protein. First, a ubiquitin-activating enzyme (E1) primes ubiquitin by covalently binding its C-terminal glycine residue forming a reactive thioester bond with a cysteine in the E1 enzyme [5]. This process involves ATP-hydrolysis, as ubiquitin-adenylation precedes the thioester bond formation to be energetically favorable [6]. Second, the reactive thioester allows transfer of ubiquitin to one of a few E2 enzymes. Substrate specificity is then achieved in complex with an E3 ubiquitin-ligase assisting the transfer of ubiquitin from the E2 enzyme to its target protein directly in case of RING (really interesting new gene) E3 ligases or via additional intermediate thioester conjugation to the E3 ligase in case of HECT (homologous to E6AP carboxyl terminus) E3 ligases [7]. In the target protein ubiquitin is bound to a lysine residue via an isopeptide bond [8]. A chain of ubiquitin moieties can be formed as further ubiquitin can be attached to a lysine or the N-terminus of the already bound ubiquitin. Ubiquitin chain elongation via lysine 48 of ubiquitin (K48-linkage) is the main signal to facilitate protein degradation where a minimum of four ubiquitin molecules has been described to be needed [9]. However, seven lysine residues (K6, K11, K27, K29, K33, K48 and K63) and the N-terminus of ubiquitin allow for a high diversity of different linkage types as the chains can be made of homotypic linkages (i.e. chains of K48-binding ubiquitin) as well as heterotypic linkages where different lysine residues are used and chains can be branched when more than one lysine on the same ubiquitin is used for chain elongation [3, 10]. Branched K11/K48 ubiquitination as well as multiple mono-ubiquitination on one protein can also lead to proteasome binding and degradation [11, 12]. Ubiquitination is also involved in processes independent of protein degradation. Linear head-to- tail ubiquitin chains [13] as well as K63-linked poly-ubiquitin chains have for example been found to be involved in the regulation of the NF-κB (nuclear factor kappa B) signaling pathway ([14] and section 2.3.3), DNA repair [15] and receptor internalization [11, 16]. In contrast, proteins can be delivered to

Chapter I – Introduction – Proteasomes and Immunoproteasomes proteasomal degradation by ubiquitin-independent pathways [17–19] and ubiquitin-like modifiers contribute to protein regulation [20].

1.2 Structure of standard proteasome and immunoproteasome

The proteasome was initially described under several different names including the name “multicatalyical protease complex” introduced by Wilk and Orlowski [21]. It refers to the finding that this complex harbored activities similar to several previously known peptidases but lost these activities upon disassembly of the complex [21–23]. Today the term “proteasome”, originally proposed in 1988 [24], collectively refers to an intracellular protein complex that is composed of the 20S core particle (CP) alone or in combination with one or two of its cap structures. The 20S CP in complex with the 19S regulatory particle (19S RP, also called proteasome activator PA700) constitutes the 26S proteasome. Other protein complexes binding the 20S CP are 11S activators (PA28αβ, PA28γ), PA200 and PI31 [25, 26]. Different combinations of CPs in complex with one or two proteasome activators are possible, for example RP-CP (26S), RP-CP-RP (30S), RP-CP-PA200 and RP-CP-PA28αβ (hybrid proteasomes) or other combinations [27]. The 26S proteasome has been most intensively investigated and many functions of individual subunits in both the CP as well as the 19S RP could be elucidated. 19S RP subunits are responsible for example for recognition and binding of ubiquitinated proteins (regulatory particle non-ATPase, Rpn1, Rpn10 and Rpn13), ubiquitin removal (Rpn11 and proteasome-interacting proteins like Ubp6/Usp14) and protein unfolding, 20S CP gate opening and substrate translocation (regulatory particle AAA ATPase, Rpt1-6) [28, 29]. Further details are outlined in section 1.2.4.

1.2.1 Structure and dynamics of the 20S core particle

The 20S CP is a 730 kDa hetero-multimeric protein complex that forms a barrel-like structure consisting of four stacked rings [30, 31] (Figure 1A). The outer two rings contain seven α-subunits each, numbered α1-7. The α-ring forms a pore in its center, the size of which is regulated by conformational changes. The centrally protruding N-termini of the α-subunits function as a gate allowing the regulated entry of proteins for degradation [30]. Indeed, N-terminally truncated α3-containing mammalian proteasomes degrade fluorogenic substrates and poly-ubiquitinated proteins faster than wild-type (WT) proteasomes [32]. Physiologically, gate opening is regulated by interaction with proteasome activators, the mechanism of which was described originally for Trypanosoma brucei 11S activator co-crystallized with yeast proteasome: C-terminal extensions of the 11S subunits protrude into pockets of the α-subunits for allosteric regulation [33]. This was similarly found in other regulatory particles across different species [25, 34, 35]. Additionally, allosteric regulations between catalytic as well as non-catalytic β-subunits and gate opening and regulator binding were reported [36–38]. A large fraction of proteasomes in cells exists as free 20S

Chapter I – Introduction – Proteasomes and Immunoproteasomes

CPs, which are largely latent in vitro [39], but were recognized to facilitate ATP-independent degradation e.g. of oxidized proteins or proteins with intrinsically disordered regions ([27, 40–43] and section 1.5.2).

The inner two rings of the 20S CP consist of seven β-subunits each. Only three β-subunits in each ring have catalytic activity for peptide hydrolysis [44, 45] (Figure 1A-B). In standard proteasomes these are β1c, β2c and β5c (c-subunits for “constitutive”) [30, 46]. Apart from these standard or constitutive active β-subunits, four alternative catalytically active subunits can substitute at the position of their constitutive counterparts in the 20S CP. Two alternative subunits were first identified as major histocompatibility complex (MHC) locus encoded genes [47–49]. They were found to be expressed in lymphoid tissues as well as under the influence of the pro-inflammatory cytokine interferon-γ (IFN-γ) and are hence called i-subunits for “immuno-“ subunits [45, 50, 51] (section 1.3). As such the three immuno-subunits β1i (also called low molecular mass polypeptide (LMP)2), β2i (multicatalytic endopeptidase complex-like (MECL)-1) and β5i (LMP7) are incorporated into newly synthesized proteasomes building the so-called immunoproteasome [52–54]. Restricted to the thymus, cortical thymic epithelial cells (cTECs) and partially also thymic dendritic cells (DCs) furthermore express the fourth alternative subunit, β5t [55, 56]. Another tissue-specific proteasome isoform exists in testis and in particular in spermatids (reviewed in [57]). Besides standard and immunoproteasomes in which all β-subunits would be either c-subunits or i-subunits, respectively, the β-subunit diversity mathematically allows for 33 different β-subunit combinations in one 20S CP, but only 15 are possible given mutual incorporation dependencies of particular subunits [58]. Hence, mixed proteasomes exist in which c-subunits and i-subunits are together incorporated into one 20S CP (Figure 1C and further outlined in section 1.2.3).

1.2.2 Peptide hydrolysis in the 20S CP

The barrel-shaped structure of the 20S CP results in three interior chambers inside the proteasome that arise at the interfaces between the rings [30, 31] (Figure 1B). The inner chamber between the two β-rings is where peptide hydrolysis occurs. The active β-subunits show different activities: β1c, β2c and β5c have caspase-like, trypsin-like and chymotrypsin-like protease activities, respectively [21, 22, 30, 50]. I.e. β1c cleaves preferably after acidic amino acids, while β2c cleaves after basic residues and β5c preferably after hydrophobic amino acid side chains. The substrate preferences are based on peptide binding grooves near the catalytically active centers that favor binding of particular polypeptide side chain residues in the respective P1 to P3 and P1’ to P3’ positions relative to the cleavage site. Central to hydrolysis is the N-terminal threonine residue of the active β-subunits. A nucleophilic attack at the peptide bond carbonyl-C atom leads to a tetrahedral intermediate followed by an acyl-enzyme bound intermediate that is hydrolyzed to achieve the newly formed amino- and carboxyl-termini of the two peptide products [59, 60] (Figure 1D). Cleavage products generated by the 20S CP range in size from 3 to 22 amino acids in length (3-30 in archaea), the bulk of which is rapidly degraded by peptidases in the cytosol after release from the 20S CP 14

Chapter I – Introduction – Proteasomes and Immunoproteasomes

[61, 62]. The cleavage preferences between constitutive and immuno-subunits differ from each other [63, 64]. Even though both β5c and β5i have chymotrypsin-like activity, β5i accepts bulky amino acids better than β5c [31, 65]. LMP2 (β1i) has also chymotrypsin-like activity and hence replaces the caspase-like activity of β1c. It enhances cleavage after basic and hydrophobic residues and reduces cleavage after acidic residues [66, 67]. In contrast, the substitution of β2c by MECL-1 does not change the trypsin-like activity at the β2 position and it has remained elusive why this subunit substitution occurs [68]. The 2.9 Å resolution crystal structure of mouse immunoproteasome revealed details about the differences between standard and immunoproteasomes. Several amino acid exchanges between β5c and β5i result in an increased S1 pocket size of β5i and an overall higher hydrophilicity (Figure 1E) [31]. Noteworthy, the higher hydrophilicity and changed electron density around the active center of β5i indicate a kinetic advantage for peptide hydrolysis via attraction of water molecules and stabilization of the tetrahedral intermediate state [31]. These features make LMP7 (β5i) special as they indicate a faster peptide cleavage ability by LMP7-containing proteasomes. In line with this, a principal component analysis of available mammalian CP crystal structures revealed that while β5c undergoes a conformational change upon peptide ligand binding, the ligand-free structure of β5i already clusters with ligand-bound β5c structures [69]. Moreover, molecular dynamics simulations indicate a cluster transition upon removal of ligands in β5c, but not in β5i. Together this indicates that β5i is structurally pre-formed for optimal peptide binding and hydrolysis in contrast to β5c [69]. A comprehensive integration of modelling with experimental data by Liepe et al. further supports faster degradation of small peptides by immunoproteasomes compared to constitutive proteasomes taking allosteric positive and negative interactions of peptides with the CP into account [70]. These publications provide structure- and molecular dynamics-based models that might contribute to explain superior antigen processing and protein degradation capacity of immunoproteasomes reported in vitro or in vivo [71–74]. Nevertheless, the relevance of these observations remains controversial as i) ubiquitination and ATP-hydrolysis dependent protein unfolding and peptide entry into proteasomes before hydrolysis were reported as the rate-limiting step of proteasome degradation kinetics [28, 70, 75– 77], and ii) the findings were partially challenged by failure of independent reproduction [78] or contrasting results in other studies ([79, 80]), which will be further outlined in section 1.5.2.

1.2.3 Assembly of proteasomes and immunoproteasomes

Assembly of eukaryotic proteasome 20S CPs takes place after de novo synthesis of proteasome subunits [42] in an ordered fashion that is aided by the specialized chaperones PAC1, PAC2, PAC3, PAC4 and Ump1/POMP [81, 82]. The PAC1-PAC2 complex initiates proteasome core particle assembly to form the α-ring of one half-core-particle. The α-ring subsequently serves as the assembly site for the β-subunits, which is aided by the PAC3-PAC4 complex. Using siRNA against individual subunits and detecting the appearance of intermediate complexes, the order of β-subunit incorporation was first determined for

Chapter I – Introduction – Proteasomes and Immunoproteasomes

Figure 1: Structure of the 20S Core Particle and its peptidolytic activity A) Schematic representation of the barrel-shaped structure of the 20S CP with β-subunit positions and catalytic activities of the active β-subunits. B) Ribbon representation of the 20S CP with threonine residues of active β-subunits marked in blue, red and purple. Taken from Tanaka et al.2012 [45] C) Schematic representations of standard proteasomes, mixed proteasomes and immunoproteasomes. Immuno-subunits are marked in red, while standard or constitutive subunits are marked in blue. D) Biochemical reaction mechanism of peptide hydrolysis in the 20S CP. Adapted from Ruschak et al. 2011 [60] and modified. E) Schematic comparison between the substrate channels of β5c and β5i subunits. The active threonine is marked in red. Adapted from Cromm & Crews 2017 [65] 16

Chapter I – Introduction – Proteasomes and Immunoproteasomes standard proteasomes [83] and later for immuno- and thymoproteasome core particles [84]. Catalytically active β-subunits are first synthesized as precursors containing N-terminal propeptides that protect the active threonine residues in the catalytic β-subunits from Nα-acetylation [85]. Furthermore, the first incorporated subunit, β2, contains a C-terminal tail extending around the consecutively incorporated β3 subunit which is necessary for assembly [83]. The propeptides of β2 and β5 are also indispensable for mammalian proteasome maturation [83, 86]. Notably, in standard proteasomes β4 incorporation is then preceding β5 incorporation, whereas in immunoproteasomes β2i and β1i can simultaneously assemble on the α-ring and β5i incorporation is not dependent on β4 [83, 84]. Standard CP intermediates containing β2, β3, β4, β5, β6 and β1 typically incorporate β7 as the last subunit before assembly of the half-core-particles to the mature 20S CP, a step that requires both autocatalytic as well as trans-catalytic cleavage of the propeptide sequences [46, 87, 88]. When immuno- and standard proteasome subunits are both present in the same cell, immunoproteasome subunits are preferentially assembled into mature 20S CPs [89, 90]. Even though β5i/LMP7 is incorporated after β2i/MECL-1 and β1i/LMP2, formation of proteasomes containing MECL-1 and LMP2 largely depends on LMP7 incorporation as the maturation of the half-core particles to mature 20S CPs depends on LMP7. The preferential assembly of immunoproteasomes is also attributed to the higher affinity of the immunosubunit LMP7 to the IFN-γ inducible assembly chaperone POMP (proteasome maturation protein) [91]. After full assembly of mature 20S CPs POMP is degraded immediately and cells showing impaired CP maturation e.g. due to lack of the LMP7-propeptide accumulate POMP [92]. In spite of preferential assembly, mixed compositions of fully assembled 20S CPs are possible and were detected [93–95]. However, the composition variety is limited by the mutual incorporation dependencies and it was reported that 15 theoretically possible mature CP types could be identified with site-specific probes in Raji cells [58]. The authors interpreted their results assuming that LMP2 and MECL-1 are exclusively incorporated into LMP7-containing CPs. In contrast, the β5c subunit was reported to at least partially compensate for the loss of LMP7 in MECL-1 and LMP2 containing proteasomes [96], but reduced LMP2 and MECL-1 incorporation and precursor accumulation have been reported in LMP7-deficient lymphocyte blasts [86]. The interpretation of how strictly LMP2 and MECL-1 depend on LMP7 for incorporation thus differed between studies, indicating that LMP7-deficiency reduces, but not fully abrogates incorporation of LMP2 and MECL-1. Nevertheless, it is possible that particular mixed proteasome compositions only occur in gene-knockout models, but do not represent compositions as they would naturally occur in unmodified and healthy organisms, where the mixed compositions seem to be limited to β1c-β2c-β5i and β1i-β2c-β5i (reviewed in [97] and Figure 1C).

Once proteasomes are fully assembled, they compose relatively stable, long-lived protein complexes. When Heink et al. compared the half-lives of standard proteasomes with immunoproteasomes they found that independent of IFN-γ immunoproteasomes had a much shorter half-life of about 27 h in T2 cells as compared to standard proteasomes with about 133 h mean half-life [91]. Hence, they concluded that 17

Chapter I – Introduction – Proteasomes and Immunoproteasomes immunoproteasome expression is only a transient response. In contrast, Nandi et al. reported that the half- lives of proteasomes in IFN-γ treated H6 cells (containing immunoproteasomes) and in non-treated H6 cells (containing standard proteasomes) are not significantly different from each other [53]. Interestingly, a recent study using improved peptide-ion-based mass spectrometry for analysis of protein half-lives in non- dividing cells has unraveled the half-lives of proteasome subunits in primary B cells, monocytes, NK cells and hepatocytes. Not only do the authors find supporting evidence that protein-complex subunits show coherent turnover rates, but they also find much longer half-lives of the immunosubunits LMP2 and MECL-1 (PSMB9 and PSMB10) of more than 150 h as opposed to the standard subunits β1 (PSMB6) or β2 (PSMB7) of only about 20 to 30 h [98]. In contrast, hepatocytes showed significantly longer half-lives of standard proteasome subunits with mean half-lives between 109 and 210 h, while immunosubunits were not detected in hepatocytes [98]. Whether peptide fragments from specific subunit proteins detected by this method were in fact incorporated into mature CPs or not might have to be further taken into account. Nevertheless, it appears that the longevity of immuno- or standard proteasomes is a cell-type-dependent and possibly activation-status-dependent characteristic of cells.

1.2.4 Role of regulatory particles for proteasome and immunoproteasome function

The 19S regulator: The 19S RP (also PA700) consists of two sub-complexes. Rpn1, Rpn2, Rpn10, Rpn13 and Rpt1-6 form the so-called base while the subunits Rpn3, Rpn5-9, Rpn11-12 and Rpn15 form the lid of the 19S RP [29]. The 19S RP is best studied for its role in binding, unfolding and translocating ubiquitinylated-protein substrates for degradation by the proteasome 20S CP. Based on independent work by the Baumeister laboratory and the group of A. Martin an insight into the structure of the 19S RP containing proteasomes was achieved using cryo-electron microscopy [99–101]. In contrast, classical x-ray crystal structures of RPs could not be obtained before [45]. The six AAA-type ATPases Rpt1-6 form a hetero-multimeric ring that can be further subdivided into three dimers binding together via coiled-coil interactions. Rpt2, 3 and 5 were originally identified as regulators of 20S CP gate opening via conserved hydrophobic-tyrosine-X motif (HbYX) containing C-terminal protrusions that interact with pockets in the α-ring subunits of the 20S CP [102–104]. Recently, high resolution EM-reconstruction led to the discovery that also Rpt1 and Rpt6 take part in gate opening via non-conserved HbYX-independent motifs, while HbYX-motif interactions of the Rpt2, 3 and 5 subunits with the 20S CP alone are insufficient for gate opening [105, 106]. The ATPases of the base are vital for translocation of the substrates into the 20S CP and undergo a complex cycle of conformational changes as characterized using non-hydrolyzable nucleotide-derivatives [105]. Thus, conformational states of the 19S base could be assigned to substrate- accepting state (s1), commitment state (s2) and substrate-processing states (s3 and s4) [105, 107], which were similarly found even within intact cells [108, 109]. Rpn10 and Rpn13 are direct binders of ubiquitinated substrates while Rpn1 can contribute to substrate binding by indirect substrate delivery via

Chapter I – Introduction – Proteasomes and Immunoproteasomes ubiquitin-like and ubiquitin-associated domain protein receptors (UBL-UBA-receptors) that are not part of the 19S RP [101, 110–113]. Rpn2 was implicated in stabilizing Rpn1, Rpn10, Rpn13 and the deubiquitinating enzyme Rpn11, which facilitates ubiquitin-removal from the substrates before entry into the 20S CP [100, 110]. The PA28αβ and PA28γ regulators (11S regulator, REG): The mammalian 11S regulatory particle exists in two isoforms composed of either (proteasome activator of apparent molecular weight 28 kDa) (PA28)α and PA28β subunits [114, 115] or as a ring of PA28γ. Like immunoproteasome β-subunits, also PA28αβ is IFN-γ inducible and upregulated for example in dendritic cells (DCs) during maturation [116]. Interestingly PA28αβ locates primarily to the cytoplasm, while PA28γ has been identified as a primarily nuclear cap particle, which is not induced by IFN-γ, but rather reduced [117, 118]. Whether a preferential interaction between the PA28αβ regulator and immunoproteasomes exists, has remained controversial. Groettrup et al. originally observed a changed quality and quantity of peptides generated by PA28αβ- capped CPs, but did not detect preferential interactions with LMP2- and LMP7-containing proteasomes [119]. In contrast, combined liquid chromatography – mass spectrometry (LC-MS) analysis of the proteasome supramolecular complexes from different cell lines indicated that PA28αβ would preferentially interact with immunoproteasomes in intact cells [27]. A more recent in vitro analysis using thermophoresis revealed equal affinities of iCPs and cCPs towards PA28αβ and no difference in protein degradation between iCPs and cCPs capped with PA28αβ [120]. However, PA28αβ modifies peptide cleavage of CPs and hence peptide generation for antigen presentation [118, 119, 121, 122]. An overall enhanced ability to degrade short peptide substrates was reported [114, 115] as well as an enhanced ability to degrade oxidized proteins [123] when 20S CPs were bound to PA28αβ. However, IFN-γ inducible immunoproteasomes have major effects on MHC-I restricted CD8+ T cell epitope generation, while for PA28αβ this was only reported for some epitopes [124] and other studies did not observe a direct influence on MHC-I restricted antigen presentation or CD8+ T cell responses for PA28αβ [125, 126]. The PA200/Blm10 regulator: PA200 was identified as a proteasome activator in the nucleus that is involved in DNA repair [127]. It also stimulates the degradation of fluorogenic substrates by proteasomes in vitro [127] and facilitates gate opening of the 20S α-subunits [128]. Its ortholog bleomycin-sensitive 10 cap (Blm10) was studied in yeast and plays a role in proteasome assembly [34, 129–131]. The PI31 regulator: An opposing role compared to the abovementioned particles was described for PI31. This prolin-rich protein was found to impair immunoproteasome assembly and hence impairs MHC-I antigen processing [132]. Even though PI31 inhibits proteasome activity in vitro, an overall impact of PI31 on cellular proteostasis was not found [132, 133]. In contrast, studies using the Drosophila melanogaster DmPI31 indicated that PI31 is vital as its loss of function resulted in lethality (reviewed in [45]). Data by Fabre et al. suggests that PI31 preferentially interacts with standard proteasomes, but not immunoproteasomes [27].

Chapter I – Introduction – Proteasomes and Immunoproteasomes

1.3 Tissue expression of immunoproteasomes

The subunit composition of proteasomes varies between different tissues and in response to inflammatory cues. Specialized subunits of the immunoproteasome have been found in mammalian species including human, mouse and rat, whereas, in contrast, no evidence could be obtained for the existence of immunoproteasomes in chicken for example [134]. Most early studies on immunoproteasomes focus on their function in antigen processing (section 1.4) and thus tissue expression analysis rather focused on antigen presenting cells (APCs), target tissues for CD8+ T cells and on the IFN-γ induced changes of proteasome composition in the course of inflammation. However, two studies revealed already in 1993 and 1997 that lymphoid tissues express immunoproteasomes in the absence of inflammation and immunosubunits were detected in liver and lung of mice [135, 136]. The significance of steady state immunoproteasome expression in hematopoietic cells was increasingly investigated after it became apparent that antigen-processing independent cell-intrinsic functions of immunoproteasomes must exist in T cells and B cells [50, 137–139]. Basal expression of immunosubunits is independent of IFN-γ [140, 141]. However, upon treatment of cells with IFN-γ in vitro, immunoproteasomes and PA28αβ are induced [52, 119, 142] and infection of mice with viral or fungal pathogens led to a replacement of the vast majority of proteasomes by immunoproteasomes in inflamed tissues [140, 141, 143, 144]. Neuronal cells, in contrast, were found to be special with respect to immunoproteasome induction: Kremer et al. reported that only little amount of immunoproteasome was induced in the brain of mice compared to other organs even after intracranial infection with lymphocytic choriomeningitis virus (LCMV) [143]. Expression of immunosubunits was furthermore restricted to microglia and the nuclei of astrocytes [143]. Nevertheless, immunoproteasome expression in the brain might be involved in the pathogology of neurodegenerative diseases (outlined in section 1.8.1). Generally, the immunoproteasome has been studied quite extensively for its function in antigen processing. As the work presented here focusses primarily on antigen processing independent functions, the role of immunoproteasomes in antigen presentation is only briefly introduced in section 1.4. Section 1.5 will then contain a more extended introduction into functions of proteasomes and immunoproteasomes in cellular homeostasis and signal transduction.

1.4 Proteasomes and immunoproteasomes in MHC-I antigen presentation

Conventional αβ-T cells are activated upon engagement of their T cell receptor with a cognate peptide- MHC complex on an APC. CD8+ T cells (cytotoxic T lymphocytes, CTLs) recognize peptides on MHC class I complexes. The bulk of MHC-I peptides is generated by the proteasome in the cytosol and subsequently transferred into the ER by TAP-transporters to be loaded onto MHC-I molecules inside the secretory endomembrane compartment and is then transported to the cell surface [44, 50, 145, 146]. A large fraction of these peptides derives from proteins rapidly degraded after translation at the ribosome,

Chapter I – Introduction – Proteasomes and Immunoproteasomes so called defective ribosomal products (DRiPs), but peptides are also generated from the overall proteome of a cell [147–149]. Another fraction of antigenic peptides is derived from the endo-lysosomal pathway and delivered to the MHC-I system via a process called cross-presentation. This is essential when the source of antigen is not the APC’s own cytosol, but endocytosed material, e.g. from a pathogen that does not directly infect the APC or from tumor cells. Thus, cross-presentation plays a major role in CD8+ T cell priming by DCs [150, 151]. Peptides are also presented on MHC-I in the absence of inflammation, but self-peptides/self-antigens normally do not elicit an immune response under physiological conditions [152]. The C-terminal residues of antigenic peptides function as anchor residues to fit into MHC-I peptide binding grooves, which are allele-specific, best accept hydrophobic or basic C-terminal residues and which are stabilized by peptide binding [153]. Hence, the quality of proteasome cleavage products has an influence on their antigenic properties as the C-termini of peptides are determined by the proteasome [154]. The appropriate length of 8-10 amino acids for MHC-I loading is further achieved by cytosolic aminopeptidases and ER-residing peptidases, which process N-terminally extended peptides generated by the proteasome [154–157]. Several studies showed that the quality of peptides generated by immunoproteasome subunits is better suited for MHC-I loading. As mentioned above, LMP7 accepts bulky hydrophobic residues better than β5c and LMP7-deficient cells show even ~50% less MHC-I surface expression compared to WT cells [31, 50, 158]. The substitution of β1c by LMP2 changes the proteolytic activity from caspase-like to chymotrypsin-like activity, yielding more peptides appropriate for MHC-I loading. LMP2-deficient mice show altered peptide cleavage and CD8+ T cell responses [67]. However, immunoproteasomes are not always the “better” antigenic peptide producers as some antigens are destroyed by immunoproteasome subunits and are therefore rather produced by cells containing standard proteasomes. This was reported for particular Epstein-Barr-Virus epitopes and melanoma-derived antigens [159]. This mechanism also exists in cases where the immunoproteasome is needed for antigenic peptide generation, depending on structural features of LMP7 and not on the proteolytic activity [160, 161]. It was also demonstrated that β1c containing proteasomes destroy the UTY246-254 epitope generated in male HY- mice, which is preserved when LMP2 incorporates at the position of β1c showing that not the cleavage specificity, but the structural presence of LMP2 in the CP is important in this case [162].

Despite their different proteolytic activities, it seems that cCPs and iCPs do not show absolute qualitative differences in generating unique, non-overlapping peptide products. Instead, the quantities in which particular peptides are generated by iCPs or cCPs differ, changing the quantitative accessibility in the pool of suitable peptides available for MHC-I loading [163]. Nevertheless, a functional influence of this difference on antigen presentation-dependent immune functions and T cell repertoire generation has been well documented as described above [72, 93, 124, 159–162, 164–168].

Chapter I – Introduction – Proteasomes and Immunoproteasomes

It should be emphasized that the response of a CTL to peptide MHC-I on target cells depends on both, its pre-selection during thymic development as well as its initial priming by DCs. Cortical TECs were first described to lack immunoproteasome expression, which was only induced upon pathogenic infection or IFN-γ treatment [141]. In contrast, Tanaka and colleagues showed that cTECs express mainly the specialized form called thymoproteasome, in which the alternative subunit β5t is incorporated together with β2i and β1i [56, 169]. Instead, mTECs highly express immunoproteasomes (but also standard proteasomes), which is also true for DCs that can be involved in negative selection [141]. Furthermore, the initial CD8+ T cell priming in the periphery by DCs also depends on immunoproteasomes. Different peptide repertoires in positively and negatively selecting cells were hence proposed as a mechanism to avoid autoimmunity [51]. In 2016, Rock and colleagues provided evidence for the peptide switching model of antigenic selection in the thymus, as mice deficient for the four alternative subunits (β1i, β2i, β5i and β5t) show massive loss of thymocytes during negative selection [170]. This indicated that the peptides presented during positive selection and negative selection must normally be different from each other. Therefore, immuno- (and thymo-) proteasomes in the thymus play an important role in generating the T cell repertoire that must be suitable for recognizing foreign antigen without elicitation of an immune response against self-peptides.

1.5 Proteasomes and immunoproteasomes in cellular homeostasis and signal transduction

1.5.1 Role of proteasomes for nutrient availability and amino acid recycling

The ratio of protein synthesis and protein breakdown within a cell needs to be tightly regulated and adjusted according to intracellular and extracellular cues like shortage of amino acid supply for de novo synthesis or activation and growth signalling mediated via surface receptor stimulation. Besides uptake of amino acids via amino acid transporters or their synthesis via mostly citric acid cycle dependent metabolic intermediates, amino acids are also recycled after protein breakdown by the proteasome. Shortage of amino acid supply has been shown to be a factor driving cells into apoptosis under conditions of proteasome inhibition [171]. A central nutrient status sensor in cells is the mechanistic target of rapamycin complex (mTORC) pathway, which is inhibited upon nutrient deprivation. Zhao et al. reported that inhibition of mTORC with the synthetic inhibitor molecule Torin1 or pharmacologically with rapamycin increases the rate of K48-ubiquitination and degradation of long-lived proteins in cells [172]. In line with this Rousseau and Bertollotti find TORC1/mTORC1 (yeast/mammalian) inhibition to be inducing both regulatory particle assembly chaperones (RAC) and proteasome subunits and identify Mpk1/ERK5 to be involved in this regulation [173]. At the same time mTORC1 hyperactive signaling resulting from tuberous sclerosis

Chapter I – Introduction – Proteasomes and Immunoproteasomes complex (TSC)1 knockout increases proteasomal degradation by inducing Nrf1/Tcf11 expression, which in turn activates expression of standard proteasome subunits (section 1.6 and [174]). Interestingly, the mTOR pathway has been found to also modulate immunoproteasome assembly as PRAS40 tethers precursor proteins of β1i, β5i and β6 until PRAS40 gets phosphorylated by mTORC1. Consequently, immunoproteasome precursors are released from PRAS40 for assembly of functional iCPs [175]. Taken together, it has now become established that mTORC signaling not only regulates protein homeostasis and amino acid supply via autophagy induction, which for example occurs upon mTORC1 inhibition with rapamycin [176], but also via regulating essential components of the ubiquitin proteasome system.

1.5.2 Removal of misfolded, damaged and oxidized proteins

A bona fide function of proteasomes is the control of protein homeostasis in cells, i.e. the removal of proteins that are not properly folded, damaged by stress conditions like reactive oxygen species (ROS) formation or degraded on the basis of regular protein turnover. Misfolded proteins can be derived directly from protein translation in form of DRiPs (section 1.4). An inflammatory challenge like exposure to the pro-inflammatory cytokine IFN-γ enhances the formation of DRiPs, which are cleared by the proteasome [73]. Similarly, oxidative stress results in more proteasome substrates. While the formation of ROS can be triggered by environmental influence like certain drugs, toxins or ionizing radiation, ROS also plays a role as an intrinsically caused result of biochemical processes within cells. Normal mitochondrial ATP- production by oxidative phosphorylation produces superoxide anions as byproduct, which can be converted to hydrogen peroxide [177]. Macrophages are known to functionally use the production of ROS by NADPH-oxidase for killing of intracellular bacteria, a process called respiratory burst [178]. Family members of the NADPH-oxidase complex (Nox) are inducible in a wide range of cell types. Similarly to enhanced DRiP formation, enhanced oxidative stress is promoted by IFN-γ challenge as IFN-γ induces Nox isoforms like Nox1 resulting in increased ROS production [179, 180]. Nox1 is also highly expressed in lymphoid cells and was found to influence T helper cell polarization [181]. When the canonical antioxidant mechanisms of cells including ROS-alleviating enzymes like superoxide dismutase or ROS- scavenging molecules like glutathione cannot counteract the oxidizing effects of ROS, the resulting imbalance is what is called “oxidative stress”. It was found that degradation of oxidized proteins is mainly performed by uncapped 20S proteasomes instead of 26S proteasomes [182]. Also it was demonstrated that proteasome activity could be modulated by poly-ADP-ribosylation [183], which can be a consequence of oxidative stress induced DNA-damage. Interestingly, immunoproteasomes as well as PA28αβ were reported to be up-regulated by oxidative stress after exposure to H2O2 [184] or nitric oxide species [185]. It appeared that purified immunoproteasomes from mouse embryonic fibroblasts (MEFs) were to some extent superior in degrading oxidized proteins as compared to standard proteasomes [184]. Data pointing to an in vivo relevance of this finding came from Seifert et al. who showed that LMP7-deficient cells were

Chapter I – Introduction – Proteasomes and Immunoproteasomes less capable of degrading oxidized proteins and poly-ubiquitin conjugates [73]. As a consequence they report enhanced disease severity of experimental autoimmune encephalomyelitis (EAE), a mouse model for multiple sclerosis [73]. In line with this, St-Pierre et al. recently reported that LMP7/MECL-1 double- deficient mice show reduced mTEC survival and enhanced thymic involution explained by higher proteostasis stress burden resulting from promiscuous gene expression during negative selection, which cannot be sufficiently alleviated in cells lacking immunoproteasomes [186]. Also, Opitz et al. reported that LMP7-deficient mice show more severe Coxsackievirus B3 (CVB3) induced myocarditis partly because of impaired poly-ubiquitin-conjugate clearance [74]. Nevertheless, the requirement for immunoproteasomes to efficiently degrade poly-ubiquitin-conjugates in cells has remained controversial. Nathan et al. did not find the same results as reported by Seifert et al. with the same or very similar experiments [78] coming to the conclusion that immunoproteasomes and standard proteasomes do not differ in their ubiquitin- conjugate degrading ability. Similarly, Hewing et al. neither observed differences in poly-ubiquitin conjugate clearance after IFN-γ challenge nor after exposure to H2O2 in bone marrow (BM) derived macrophages from WT or LMP7-deficient mice [79]. De Verteuil et al. report to observe no differences in poly-ubiquitin conjugate clearance in DCs [167]. In LMP2-deficient B cells, Yewdell and colleagues did not observe differences in poly-ubiquitin conjugate clearance compared to WT cells [139]. Since the discrepancies between the findings of these studies remain unresolved, it is still not fully clarified, whether immunoproteasomes are superior over standard proteasomes in clearing poly-ubiquitin-conjugates and/or if such superiority might be strongly cell-type and stimulus dependent. Therefore, further work is required for contributing to clarification of this question.

1.5.3 Role of proteasomes in signal transduction

Engagement of signaling pathways as induced by extracellular or intracellular receptor stimulation often results in altered transcriptional activity of the cell [187]. Transcriptional regulation involves the tight control of the transcription factors that induce certain gene expression profiles. One means of controlling transcription factor activity is via their targeted degradation. The probably best characterized involvement of proteasomes in immune cell signaling and gene transcription is found in nuclear factor kappa B (NF-κB) signal transduction (details about this pathway are outlined in section 2.3.3 as part of the TCR signal transduction). Impairing NF-κB signaling is also part of the molecular mechanism of proteasome treatment in multiple myeloma ([188] and section 1.8.3). It has been controversially debated whether immunoproteasomes and standard proteasomes play different roles for the regulation of NF-κB signaling. Impaired NF-κB processing and IκBα degradation due to the lack of LMP2 were reported as molecular characteristics in NOD-mice, in LMP2-deficient mice and in T2 cells lacking LMP2 and LMP7 [189, 190], but the results were challenged by others [191–194]. Schmidt et al. reported impaired nuclear p65 accumulation in MEFs from LMP7-deficient mice after IFN-γ-pre-treatment and tumor necrosis factor

Chapter I – Introduction – Proteasomes and Immunoproteasomes

(TNF) stimulation [195]. Yewdell and colleagues showed a slightly delayed IκBα degradation in LMP2- deficient B cells [139]. Visekruna et al. found enhanced degradation of in vitro-translated IκBα by 20S CP purified from tissues of patients with inflammatory bowel disease (IBD) in which immunoproteasome expression was elevated [196]. They furthermore reported enhanced processing of p105 by immunoproteasomes [196] and another study also claims to have identified an influence of immunoproteasomes on the alternative pathway of NF-κB signaling [197]. However, recently Bitzer et al. showed that the lack of immunoproteasomes had no influence on IκBα phosphorylation or degradation as well as on p65/p50 nuclear translocation and DNA-binding in MEFs and peritoneal macrophages [194]. Moreover, several studies using selective immunoproteasome inhibitors came to the conclusion that immunoproteasome inhibition does not impair NF-κB signaling [198–200]. Another prominent example is the tumor suppressor p53, which plays a central cellular role in response to stress conditions. Under normal conditions p53 is mainly controlled by mouse double minute protein 2 (MDM2), which acts as an E3 ligase to target p53 for proteasomal degradation [201]. Multiple stress factors can release p53 from its repressors to allow for its activity as a transcription factor. Besides its role as a tumor suppressor p53 as well as other “classical” tumor suppressors were found to play important physiological roles in immune cells ranging from pathogen sensing over cytokine production to orchestration of inflammation [202]. Several further signaling pathways involved for example in oxidative stress (e.g. hypoxia-inducible factor, HIF1α), proteostasis stress (section 1.6) or cell cycle progression (e.g. cyclins and cyclin-dependent-kinase inhibitors) are regulated by the proteasome [203, 204]. Targeted degradation of transcription factors like p53 or STAT transcription factors is also a pathogenic mechanism shared by several viruses [205–207]. Hence, the proteasome plays a pivotal role for regulation of certain transcription factors and cell signaling pathways.

1.6 Proteostasis stress, stress response pathways and apoptosis

Dysregulation of normal protein homoeostasis can lead to an accumulation of poly-ubiquitinated proteins and oxidized or otherwise damaged proteins in the cell. Therefore, pathological alterations of the UPS can result in disease manifestations such as neurodegeneration. However, cells are to some extent capable of adjusting to stress conditions. Depending on the damage that is caused by stress conditions cells might eventually undergo apoptosis.

Keap1-Nrf2: Nuclear factor erythroid-derived 2 related factor (Nrf2) is broadly expressed at steady state, but rapidly degraded by the proteasome with a half-life of only about 12 minutes [208]. The degradation is promoted by Kelch-like ECH-associated protein 1 (Keap1) dimers, which sequester Nrf2 in the cytosol by interacting with two N-terminal domain regions of Nrf2. Thus, Keap1 promotes a conformation of Nrf2 in which several lysine residues between these domains are optimally exposed for poly-ubiquitination,

Chapter I – Introduction – Proteasomes and Immunoproteasomes thereby inducing Nrf2 degradation [209]. Activation of the Nrf2 pathway is induced in the presence of oxidative stress and upon proteasome dysfunction [210]. Under oxidative stress conditions the destabilizing conformational influence of Keap1 on Nrf2 is disrupted. Nrf2 can subsequently accumulate and translocate to the nucleus where it binds to antioxidant response elements (ARE) in the promotor regions of many genes including standard proteasome subunits and PA28αβ [211, 212]. In contrast, immunoproteasome subunits were not found to be regulated by Nrf2 even though Psmb8/β5i contains the Nrf2-responsive ARE consensus sequence upstream of the gene [211, 212]. Interaction of mutated p53 with Nrf2 was found to be a common mechanism of cancer cell resistance to proteasome inhibitor treatment [213].

Nrf1/TCF11: While Nrf2 also accumulates and translocates to the nucleus under conditions of proteasome inhibition [208], its main function seems to be the de novo proteasome subunit expression in response to oxidative stress. In contrast, it was found that the up-regulation of standard proteasome subunits, but not immunoproteasomes, in response to proteasome inhibition described by Meiners et al. [214] was rather attributed to Nrf1 (orthologue of human TCF11) [208, 215]. Nrf1 showed higher efficiency than Nrf2 to induce expression of PSMB6/β1c via the ARE promotor element, while Nrf1 and Nrf2 both relied on synergistic co-operation with the MafG transcription factor [208]. In MEFs deficient for Nrf1, but not in Nrf2-deficient MEFs, the up-regulation of representative members of PSMA-, PSMB-, PMSC-, and PSMD-subunits in response to proteasome inhibition was abrogated [215]. In line with the results of Meiners et al. [214] also bortezomib treatment rather suppressed immunoproteasome subunits instead of inducing them indicating that Nrf1 does not drive immunoproteasome expression [216]. In the physiological state, Nrf1/TCF11 is inserted into the ER-membrane via its N-terminal domain in a Sec61 dependent manner [208, 217, 218]. Furthermore, Nrf1 is glycosylated which was reported to negatively regulate Nrf1 stability and transcriptional activity [219]. Cytosolic de-glycosylation is required for transcriptional activity of Nrf1 [220]. Nrf1 has been shown to be short-lived and rapidly degraded by the proteasome [221]. The current model of Nrf1 turnover is that Nrf1 is inserted into the ER in Ncytosol/Clumen orientation, glycosylated within the ER, retro-translocated in a p97/VCP-dependent manner, then de- glycosylated, cleaved and rapidly degraded by the proteasome [220, 222]. It was first reported that the proteasome itself would be responsible for Nrf1 processing when it is partially inhibited, while Nrf1 processing would be abrogated when the proteasome is fully inhibited [216]. However, it later became clear that high-dose proteasome inhibition results in aggregation of Nrf1 in insoluble fractions obscuring that Nrf1 can be processed independently of proteasomes by DNA damage inducible 1 homologue 2 (DDI2) [223–225]. A role of Nrf1 in the absence of pharmacological proteasome inhibition was reported in mice conditionally deficient for Nrf1 in the liver, revealing that ARE-driven genes exist that are preferentially induced by Nrf1, but not by Nrf2 [226, 227]. Furthermore, Nrf1 was reported to be involved in glucose metabolism regulation [228] and most recently in ER-cholesterol homeostasis [229]. 26

Chapter I – Introduction – Proteasomes and Immunoproteasomes

Unfolded Protein Response / Integrated Stress response: The well-established term “unfolded protein response” (UPR) goes back to the discovery that misfolded proteins in the ER can induce reactions to consequently up-regulate ER-chaperone genes in mammalian cells and in yeast [230–232]. Thus, the UPR is conventionally seen as an ER-stress-response pathway. However, misfolding stress and accumulation of damaged proteins can also occur in other cellular organelles as well in the cytoplasm. Anne Bertolotti recently proposed to refer to different subcellular responses after misfolding stress as to the “ER-UPR” for the ER-stress-response associated UPR, the “mito-UPR” for mitochondria-associated induction of mitochondrial chaperones in response to misfolding stress and to the “cyto-UPR” for responses to protein misfolding in the cytoplasm as classically described in the heat shock response system [233].

The classical UPR as an ER-stress response is mediated by three major pathways [234]. Misfolding in the ER lumen leads to activation of the ER-membrane resident protein kinase RNA-like endoplasmic reticulum kinase (PERK) by oligomerization and auto-transphosphorylation. PERK phosphorylates eukaryotic initiation factor 2α (eIF2α), which results in dampening of total protein synthesis [235]. Phosphorylated eIF2α also induces activating transcription factor (ATF)4, which activates downstream transcription of metabolic pathways, antioxidant pathways, autophagy and apoptosis [234, 236]. PERK also activates the Nrf2 pathway (described above) by phosphorylation of Nrf2 [237]. As the phosphorylation of eIF2α can be induced by at least four kinases reacting to different and also ER- unrelated stress stimuli, eIF2α phosphorylation can be regarded as a more general marker for cellular stress response reactions and is hence a sign of the “integrated stress response” [236]. Another ER-membrane resident factor is endoribonuclease inositol-requiring enzyme 1-alpha (IRE1α), which is also activated via oligomerization and auto-transphosphorylation. It acts as an mRNA splicing enzyme that at low levels of ER-stress quite selectively splices the mRNA of X-box binding protein (XBP)1, resulting in an active protein product XBP1s from the spliced mRNA. XBP1s induces genes involved in lipid synthesis, ER- associated degradation (ERAD) and protein folding [234, 238]. IRE1 can also induce mitogen activated protein kinase (MAPK) signaling [239] (see also section 2.3.5). The third arm of the UPR is activated via another ER-membrane resident protein, ATF6. Upon stress, ATF6 is transported from the ER to the Golgi, where the cytoplasmic domain is cleaved off and translocates to the nucleus [240]. Active UPR promotes ERAD and enhances ER-chaperone levels like BiP/Grp78. Thus, the UPR helps cells to adapt to ER-stress, which is especially important in cells with an extended secretory system like B plasma cells [241, 242]. Noteworthy, also T cells depend on ER-function – especially upon activation – as cytokine production, cell blasting and increasing nutrient-uptake require the secretory system [243]. Another stress response pathway initiated is autophagy, which was shown to preserve cell viability when proteasomal degradation is insufficient and when ER-stress occurs [210, 244]. Prolonged, strong activation of the UPR does not promote cell survival, but induces apoptosis, a mechanism contributing to clinical efficacy of proteasome inhibitors (sections 1.7 and 1.8). A simplified overview of stress response pathways is shown in Figure 2. 27

Chapter I – Introduction – Proteasomes and Immunoproteasomes

Figure 2: Simplified schematic overview of selected cellular stress response pathways. A) Nrf2 is regulated in the cytosol by Keap1, which promotes Nrf2 ubiquitination and degradation by the proteasome. In response to oxidative stress, Nrf2 is released from the negative regulation by Keap1 and subsequently translocates to the nucleus for gene regulation. B) Nrf1 is co-translationally integrated into the ER membrane, where it is glycosylated. In a p97/VCP-dependent manner, Nrf1 is retro-translocated to the cyosol, de-glycosylated and cleaved by DDI2. Under physiological conditions, Nrf1 is rapidly degraded by the proteasome. In response to proteasome inhibition, cleaved Nrf1 accumulates and induces proteasome gene expression in the nucleus. C) The three-armed pathways of the unfolded protein response (and integrated stress response) are mediated via PERK-eIF2α, IRE1-XPB1 and/or ATF6. Phosphorylated eIF2α halts global translation, but promotes ATF4 induction. See text for details and references. Apoptosis: Apoptosis induction can be generally divided into the intrinsic and the extrinsic pathway of apoptosis. The latter plays a major role in target-cell killing by immune cells via Fas/FasL [245, 246]. Sustained pro-apoptotic UPR, however, initiates the intrinsic pathway, in which mitochondrial permeabilization and cytochrome c release into the cytoplasm initiate activation of caspases. One marker for pro-apoptotic activation of the UPR is the accumulation of CCAAT-enhancer-binding protein homologous protein (CHOP, also GADD153), which is induced upon prolonged PERK activation and inhibits anti-apoptotic Bcl-2 (B cell lymphoma 2) signaling [247–249]. Bcl-2 has been shown to be regulated by NF-κB [250] and normally counteracts the pro-apoptotic action of Bcl-2 associated X protein (BAX)-like superfamily and the action of BH3-only proteins at the ER-membrane [251]. Conventionally used molecular markers for apoptosis induction are cleavage of caspase-3 and cleavage of poly-ADP- ribose polymerase (PARP-1). The latter is commonly used as an apoptosis marker as the cleaved PARP N-terminal fragment actively inhibits overall PARP function in the cell, abrogating its DNA-repair capacity [252]. After their effector function phase, activated T cells are eliminated from the periphery by activation induced cell death (AICD), which involves the extrinsic pathway of apoptosis, but is also partially Fas/FasL-independent [253–256]. 28

Chapter I – Introduction – Proteasomes and Immunoproteasomes

1.7 Proteasome and immunoproteasome inhibitors

After the discovery of proteasomes natural products and chemicals were identified that modulate the activity of this complex in vitro [22]. Proteasome inhibitors were initially developed to study mechanisms of muscle atrophy and were then used as tools to investigate the function of the proteasome complex [257]. Peptide aldehydes had been described as calpain inhibitors before and were then identified as proteasome inhibitors that are still widely used for in vitro studies today (mainly MG-132, Z-LLL-al) [65, 258–261]. E.g. they were used to identify the peptide generating function of proteasomes for MHC-I mediated antigen presentation (experiments were performed with the inhibitor MG115) [262] and the proteasome- dependency of NF-κB signaling [263, 264]. However, it was the peptide aldehyde derived dipeptidyl boronic acid compound bortezomib (PS-341) that was the first proteasome inhibitor to be approved by the FDA in 2003 for the treatment of patients suffering from multiple myeloma (Velcade®, Millenium Pharmaceuticals) [265–267] (section 1.9.3). Today, Bortezomib is a first-line treatment option in multiple myeloma (MM) together with other chemotherapeutics like lenalidomide and dexamethasone and approved in at least 73 countries [268]. Bortezomib is also used as a scaffold molecule in searching for new site-specific proteasome inhibitors [269]. Carfilzomib (PR-171) is a second-generation proteasome inhibitor [270] used in MM patients after unsuccessful first-line treatment [271]. It belongs to the class of epoxyketones that irreversibly modify active site threonines. Interestingly, the inhibitory mechanism of expoxyketones reported to form a covalent morpholine-ring structure based on crystal structure of 2.6 to 2.9 Å resolution [31, 272–274] was challenged by Schrader et al. providing evidence for a 1,4 oxazepane seven-membered ring-structure based on 1.9 Å resolution structures of human proteasomes co-crystallized with the epoxyketone inhibitors oprozomib, epoxomicin or Dihydroeponemycin [275] (Figure 3D). A third proteasome inhibitor was approved in 2015: The orally available Ixazomibcitrate (MLN9708), which is hydrolyzed in water to generate the active form Ixazomib (MLN2238) [276, 277].

Natural proteasome inhibitors: A natural product that inhibits the 20S proteasome is the Streptomyces spec. metabolite lactacystin. It preferentially targets the chymotrypsin-like activity, but also inhibits the other 20S subunits [278]. Its mechanistic action depends on its derivatization in aqueous solution below pH 8.0 to form an active β-lactone, which forms an irreversible acyl-ester adduct at the active threonine [279]. Lactacystin-treatment of Jurkat cells and murine splenocytes led to reduced IL-2 and IFN-γ secretion, while IL-4 was reduced in splenocytes, but elevated in Jurkat cells [280]. Similar effects of lactacystin were also found in splenocytes from LMP7/MECL-1 double deficient mice [281]. Lactacystin also induced apoptosis in patient-derived B cell chronic lymphocytic leukemia cells (B-CLL) [282]. More potent than lactacystin was the Actinomycete-derived compound epoxomicin, which was found to also mainly target the chymotrypsin-like activity but also trypsin-like and caspase-like activities with lower potency [283,

Chapter I – Introduction – Proteasomes and Immunoproteasomes

284]. Epoxomicin later on gave rise to the development of carfilzomib and other epoxyketone inhibitors described below [285].

Subunit selective (immuno-) proteasome inhibitors: Apart from natural products and clinically approved proteasome inhibitors the list of newly developed synthetic 20S CP targeting drugs has expanded significantly within the past decade. The first immuno-subunit-selective compound reported was UK-101, which is structurally also based on epoxomicin [286]. In the prostate cancer cell line PC3, UK-101 seelctively modified LMP2 in a covalent fashion leading to shifts in electrophoretic mobility. While UK- 101 (in contrast to epoxomicin) did not impair NF-κB degradation, it did induce poly-ubiquitin conjugate accumulation and apoptosis [287].

The first compound described as an LMP7-selective inhibitor also belongs to the class of epoxyketones: ONX 0914 (formerly PR-957) [199]. It has at least 10-fold selectivity for LMP7 over β5c and was used in a wide range of pre-clinical studies (see section 1.8). ONX 0914-soaked crystals of mouse immuno- and standard proteasomes revealed the basis for its selectivity. LMP7-selectivity is mainly attributed to the S1 pocket, where the Met45 conformation allows ONX 0914 to bind to the larger S1 pocket of LMP7. While ONX 0914 can in principle also inhibit LMP2, steric hindrance by Phe31 of LMP2 counteracts its binding [31]. Only at high concentrations (above 300 nM) ONX 0914 broadly inhibited proteasome subunits [199]. Treatment of splenocytes with ONX 0914 reduced their IFN-γ and IL-2 secretion and Th17 polarization was impaired by ONX 0914 treatment [199, 288]. On the contrary, splenocytes lacking LMP7 were not affected by ONX 0914, substantiating LMP7-selectivity [199]. In contrast to UK-101, no poly-ubiquitin- conjugate accumulation was observed by ONX 0914-treatment up to 300 nM and ONX 0914 in Molt4 cells also did not impair NF-κB signaling in an NF-κB reporter cell line [199, 287]. An even more selective LMP7 inhibitor of the epoxyketone class is PR-924, which has 130-fold LMP7-selectivity over β5c in human proteasomes and showed pro-apoptotic efficacy in multiple myeloma cell lines [289].

The number of developed epoxyketone-based compounds has vastly expanded in recent years for both LMP7-selective as well as β5c-selective compounds ([290, 291] and Basler et al., 2018, see section 12.5). Other compound classes with highly subunit-selective inhibition profiles include oxathiazolones [292], N,C-capped dipeptidomimetics [293, 294], non-competetive reversible asparagine ethylenediamine-based compounds [295] and dipeptide boronates [162]. Of the immunosubunit-selective inhibitors, only the ONX 0914-related compound KZR-616 has reached clinical trial phase so far (Clinical trial identifier: NCT03393013). An overview of different inhibitors used in clinical and pre-clinical applications as well as their modes of Thr1-inhibtion is shown in Figure 3 (see Ettari et al. 2018 [277], Cromm & Crews 2017 [65] and Huber & Groll 2012 [272] for extended reviews).

Chapter I – Introduction – Proteasomes and Immunoproteasomes

Figure 3: Proteasome and immunoproteasome inhibitors and their molecular inhibition modes. A) Structures of the clinically approved proteasome inhibitors Bortezomib, Carfilzomib and Ixazomib. Occupancies of substrate binding pocket sites are indicated in blue (Cromm & Crews 2017 [65]). B) Structures of the experimentally used epoxyketone inhibitors ONX 0914 and PR-825 with substrate binding groove occupancies indicated in blue. Modified after Cromm & Crews 2017 [65] and Huber & Groll 2012 [272] C) Structures of the novel LMP7-specific inhibitor PRN1126 (Basler et al., 2018, section 13.5), the broad spectrum proteasome inhibitor MG-132 and the LMP2-selective boronate inhibitor ML604440 (Ettari et al. 2018 [277]) . D) Two proposed molecular modes of Thr1 inhibition by inhibitors with epoxyketone warhead as proposed by Huber & Groll 2012 [272] and Schrader et al. 2016 [275]. Brackets denote compounds which were used in crystal structures to elucidate the inhibitory modes. E-F) Molecular modes of Thr1 inhibition by inhibitors with aldehyde warhead (E) and boronate warhead (F). Modified after Huber & Groll 2012 [272]

Chapter I – Introduction – Proteasomes and Immunoproteasomes

1.8 Proteasomes and immunoproteasomes in pathophysiology and aging

Dysfunction of the proteasome or immunoproteasome can lead to a variety of disease phenotypes. In 2010, Torrelo et al. described a rather rare syndrome under the name CANDLE (Chronic Atypical Neutrophilic Dermatosis with Lipodystrophy and Elevated temperature) [296]. This and several independent reports of patients with similar disease syndromes could be linked to mutations in the LMP7 immunoproteasome subunit, the immunoproteasome assembly chaperone POMP and in some cases other proteasome subunits [297–299]. Patients suffer from lipodystrophy, wasting, skin lesions, recurrent fevers and overall autoinflammatory manifestations including elevated type-I interferon levels [300]. At the cellular level, this is accompanied by accumulation of ubiquitinated proteins in cells and hence cellular protein stress, which is elevated when stress-inducing environmental factors like virus-infections, cold or physical exercise occur [297, 301, 302]. However, whether the ubiquitin-conjugate-induced stress is the key feature leading to the inflammatory dysfunction or whether the loss of a selectively immunoproteasome-dependent molecular mechanism is rather responsible, remains unclear [301, 302]. Importantly, in spite of ubiquitin-conjugate accumulation, the NF-κB pathway was found to be non-impaired in the experiments of Arima et al., indicating that reduced overall capacity to degrade proteins does not necessarily affect particular degradation-dependent pathways controlled by the proteasome [302]. The mutations detected in patients were found to affect either the catalytic activity of mutated subunits alone or in combination with a structural disruption of individual subunits or the whole immunoproteasome particles [301, 302]. In contrast, a seemingly opposite role for the immunoproteasome was reported by Kimura et al, finding that LMP7-deficient mice were protected from high-fat-diet induced obesity and showed reduced adipose- tissue-associated inflammation [303]. In this case, however, LMP7 was genetically deleted and the cells did not express non- or malfunctioning variants of the protein. Nevertheless, LMP7-deficiency, LMP7- inhibition with ONX 0914 as well as one of the missense mutations in LMP7 were reported to impair adipocyte differentiation [301, 304]. It has remained a matter of debate how the different underlying details of impaired LMP7 function can influence the impact on inflammatory processes, which will be further outlined in Chapter IV Discussion.

1.8.1 Proteasomes and immunoproteasomes in neurodegenerative diseases

The UPS has long been appreciated to play a role in neurodegenerative diseases like Parkinson’s disease (PD), Alzheimer’s disease (AD), Huntington disease (HD) and Amyotrophic Lateral Sclerosis (ALS) [305, 306]. A shared feature of these pathologies is the inclusion of misfolded protein aggregates with toxic impact on the respective cell. While dysfunction of the UPS and induction of inducible proteasome subunits was reported in a mouse model for familial ALS [307–309], the mechanisms of how such dysfunctions are caused in the cell have remained largely unknown. Only recently, the use of cryo-electron

Chapter I – Introduction – Proteasomes and Immunoproteasomes tomography has revealed accumulation of proteasomes within poly-GA aggregates in a stalled substrate processing state possibly explaining how proteasomes are disturbed in this particular form of aggregates that similarly occur in some ALS patients [108]. In contrast, poly-Q fibrillary aggregates as found in neurons of HD patients did not sequester proteasomes or other large molecules, but seemed to disturb the ER membrane network, thereby exerting their toxic effects to the cell [108, 109]. This is in line with previous reports that N-terminal huntigtin (N-htt) fragments did not directly inhibit proteasome function [310] and that proteasome subunit activities were not reduced in brain extracts from a conditional mouse model for HD [311]. The observed induction of immunoproteasome subunits in those brain extracts was rather attributed to secondary inflammatory cues than to cell-intrinsic responses of neuronal cells after formation of inclusions [311, 312]. A potential role for immunoproteasomes in regulating neuroinflammation and the respective cytokine environment has also been suggested in AD, which is accompanied by amyloid plaque formations and tauopathy in CNS neurons and in immunoproteasome expressing brain regions. Plaque formation was found to be increased in aged humans, mice and rats compared to younger individuals [313–317]. However, mechanistic explanations for impaired proteostasis in AD are lacking and only limited conclusions can be drawn from in vitro studies showing proteasome impairment by amyloid-protein oligomers [318]. Interestingly, inhibition of the de-ubiquitinating enzyme USP14 showed that boosting proteasomal activity could enhance the cellular capacity to degrade proteins prone to aggregation in neurodegenerative diseases, including tau protein and TDP-43 [319, 320]. The efficiency of proteasomal degradation can also be regulated by posttranslational modification at the 19S regulatory cap. Cyclic-AMP activated protein kinase A (PKA) phosphorylates Rpn6 at Ser14 and enhances proteasomal degradation capacity [321]. Attempting enhancement of proteasome capacity has thus arisen as a potential treatment option in neuropathologies involving protein aggregates [322]. The relevance of impaired proteasome function in PD is not fully elucidated. It was found that α-synuclein, which aggregates in PD patient neurons, can be degraded by the proteasome in a ubiquitin-independent manner as it has features of an intrinsically disordered protein [323, 324]. Impairment of proteasome activity in α-synuclein overexpressing neuroblastoma cells was reported, but in vitro data did not support the idea of a direct inhibitory effect of α-synuclein on proteasome function [325]. On the other hand, mutations in Psmc4 (Rpt3) as well as a mutation in Psmb9 (LMP2) were found to be associated with increased risk to develop PD [326, 327]. Taken together, despite the several reports about immunoproteasome induction in neurodegenerative diseases, it seems that their contribution is rather confined to processes of neuroinflammation as also evidenced in models of intracranial virus-infection and inflammation after injury [143, 328, 329] instead of regulating the overall proteostasis capacities of cells in the brain differently than standard proteasomes in response to pathological aggregates.

Chapter I – Introduction – Proteasomes and Immunoproteasomes

1.8.2 Proteasomes and immunoproteasomes in infectious diseases

In spite of the relatively low and cell-type restricted immunoproteasome expression in the brain, immunoproteasome-deficiency influenced LCMV-induced meningitis as LMP7-deficient mice had a delayed onset of disease symptoms [143, 328]. This was likely attributed to less infiltration of the brain by epitope-specific CTLs, but altered presentation of peptides by brain resident and infiltrating APCs might contribute as well [328]. MECL-1-deficient mice showed about 20% reduced CD8+ T lymphocytes and a reduced generation of some LCMV-epitope specific CTLs [164]. Immunoproteasome inhibition with ONX 0914 also attenuated LCMV-induced meningitis [328]. Reduced generation of influenza nucleoprotein- specific CTLs and ~60% reduced CD8+ T lymphocytes were reported in LMP2-deficient mice [67]. Yet, clearance of LCMV from infected mice was not generally altered in mice lacking immunoproteasome activity [330].

Another virus-infection model, in which ONX 0914 showed beneficial effects is Cocksackie B3 virus (CB3V) induced myocarditis, where ONX 0914 dampened inflammation by reducing macrophage- mediated heart tissue-damage [331]. In contrast, immunoproteasome-deficiency enhanced susceptibility to CB3V-induced tissue damage and the authors report on enhanced poly-ubiquitin conjugates and increased oxidatively damaged proteins in LMP7-deficient splenocytes, but excluded altered virus-specific CTL responses as an underlying mechanism [74]. Later, the same group identified the cardio-protective protein pentraxin-3 (ptx-3) to depend on LMP7. They also showed that ONX 0914-treatment reduced ptx-3 expression in macrophages, likely via reducing ERK- and p38-phosphorylation [200]. Hence, ONX 0914 could have a negative outcome for disease-severity, but its beneficial effect via reduced macrophage- infiltration and cytokine secretion seems to be superior [200, 331]. Ptx-3 is a primarily macrophage- and neutrophil-derived secreted protein which is involved in pathogen clearance via interaction with the complement system [332]. Reduced Ptx-3 expression was confirmed by Kirschner et al. who showed that LMP7-deficient mice have an impaired ability to clear Streptococcus pneumoniae infection [333]. They reported that WT BM-derived macrophages express both standard and immunoproteasomes, but that LMP7-deficiency alters c-Jun and ATF2-signaling in response to LPS-stimulation [333]. Similarly, an influence on LPS-stimulated TLR4-signaling via signal transducer and activator of transcription (STAT)1, STAT3 and interferon regulatory factor (IRF)3 had been reported in thioglycollate-elicited macrophages from LMP7/MECL-1 double-deficient mice before [334]. Further studies have shown a role for immunoproteasomes in bacterial and protozoan infection: In contrast to WT mice, LMP7-deficient mice succumbed to infection with Toxoplasma gondii [169]. Upon infection with Listeria monocytogenes LMP7-deficient mice were less competent to clear the bacteria from the liver, but not from the spleen compared to WT mice [335]. The relatively reduced numbers of CD8+ T cells in MECL-1-deficient mice remained also during the course of L. monocytogenes infection, but this was attributed to T cell-intrinsic

Chapter I – Introduction – Proteasomes and Immunoproteasomes effects [138]. Likewise, altered peptide-processing and CD8+ T cell responses were reported to result in impaired immunity to the protozoan parasite Trypanosoma cruzi in mice deficient for all three immunoproteasome subunits [336]. As immunoproteasome inhibition with ONX 0914 was shown to impair Th17 polarization in the course of autoimmune disease models (section 1.8.4), the impact of ONX 0914 treatment on fungal infection with Candida albicans was tested, because Th17 cells play an important role in controlling fungal dissemination [337, 338]. ONX 0914 affected both T helper cells as well as antigen-presenting innate immune cells and mice treated with ONX 0914 showed exacerbation of candidiasis with enhanced neutrophil recruitment to the kidney, likely as a result of fungal dissemination [337]. In summary, immunoproteasomes were shown to play important roles in the course of infectious disease pathologies by both antigen-processing dependent and independent means, but insights into the mechanistic basis for these effects are limited, especially for adaptive immune cells.

1.8.3 Proteasomes and immunoproteasomes in malignant diseases and transplant rejection

Multiple Myeloma is a plasma cell cancer with an incidence of ~140,000 per year worldwide in which abnormal plasma cell proliferation and antibody production result in anemia, increased risk of infection, kidney damage and ultimately death of patients [268]. The development of bortezomib provided a treatment for this type of cancer and was used in patients since its approval in 2003 [266, 268], but relapses after initially successful treatment and development of bortezomib-resistance are still major issues [339]. The clinical introduction of carfilzomib has meanwhile added a second-line treatment option [340]. Bortezomib is also used in mantle cell myeloma [341]. Both inhibitors were also experimentally or clinically tested in other cancer types like acute myeloid leukemia [342] or non-Hodgkin-lymphoma [343] showing successful anti-tumor activity in these studies. Likewise, proteasome inhibition has also been implicated as a potentially successful treatment of Waldenström macroglobulinemia, another type of B cell lymphoma, in preclinical or clinical studies [344–346]. In contrast to blood cell cancers, the treatment of solid tumors with proteasome inhibitors was so far less successful. Nevertheless, phase I trials with Ixazomib in patients with solid tumors have been conducted [347] and it was reported that combination therapy with IFN-α might be efficacious in melanoma [348]. Of note, two groups have recently shown that ONX 0914 was effective in treating inflammation associated colon cancer in different colon cancer models in mice [349, 350]. These findings give rise to widening the clinical applicability possibly also to solid tumors. A hallmark of many cancer cells is an extraordinarily high metabolic activity required for neoplastic proliferation. As cancer cells hence need high amounts of metabolites for anabolic synthesis of new organelles and membranes, the tricarboxylic acid (TCA) cycle intermediates are exploited for anabolism, which results in enhanced anaplerosis and reduced oxidative phosphorylation in the mitochondria [351]. ATP-synthesis is then mainly driven by glycolysis instead of the respiratory chain. This anaerobic ATP production in spite of sufficient oxygen in cells is called “Warburg effect” [352]. As a

Chapter I – Introduction – Proteasomes and Immunoproteasomes consequence, enhanced protein synthesis in highly metabolically active and proliferating cells as well as mutations in protein coding sequences in tumor cells lead to more proteasome substrates and stronger dependency on protein homeostasis capacity [353, 354]. Indeed, enhanced capacity to degrade misfolded proteins was identified to promote tumorigenesis [355]. Plasma cells produce antibodies in large quantities and especially rely on a functioning ER-secretory system and UPR [241, 356]. Plasma-cell-derived proliferating lymphoma cells like multiple myeloma cells are therefore especially affected by proteasome inhibition, which leads to apoptosis by inducing unresolvable proteostasis stress, impairing proteasome- dependent signaling pathways and activating the terminal UPR [354, 357]. Of note, the approved treatments with proteasome inhibitors together with lenalidomide and dexamethasone are accompanied by several severe side effects like anemia, extremity pain, nausea, diarrhea, rashes, neuropathy and others [358, 359], which limits broader application.

Another pathological situation, where plasma cells cause damage to the body is the chronic antibody- mediated rejection after organ allograft transplantation. Bortezomib also showed efficacy to inhibit allograft transplant rejection [360] and recently Li et al. provided evidence that ONX 0914 is as efficient in preventing chronic rejection as bortezomib in a rat kidney transplantation model [361]. In contrast, skin allograft survival could not be prolonged with ONX 0914 [362].

1.8.4 Proteasomes and immunoproteasomes in autoimmune diseases

Since the initial description of ONX 0914 in 2009 and the concomitant demonstration of its anti-autoinflammatory effects in experimental rheumatoid arthritis (RA) and type I diabetes [199], several studies investigated possible beneficial roles of immunoproteasome inhibition in a variety of mouse models of auto-immunity. Ichikawa et al. showed that ONX 0914 treatment of lupus-prone mice reduced disease severity, decreased autoantibody production, decreased plasma cells in spleen and BM and reduced IFN-α production from plasmacytoid dendritic cells (pDCs) [363]. The effect on autoantibody production was milder in a mouse model for Hashimoto thyroiditis and only significant with ONX 0914 doses close to the maximum tolerated dose (MTD) of 30 mg/kg. Still, overall thyroiditis scores were already reduced at 2-10 mg/kg [364]. Basler and Mundt et al. found that ONX 0914-treatment reduced symptoms of EAE. Importantly, they showed that the effect of ONX 0914-treatment depended on BM-derived cells and demonstrated that both targeting LMP7 in WT mice as well as targeting β5c in LMP7-/- mice reduced disease severity [80]. Thus, the effect of ONX 0914 in this model was mediated by blocking the chymotrypsin-like activity at the β5-position, irrespective of its occupancy by β5c or LMP7. Mice lacking LMP7, LMP2 or MECL-1 were instead equally prone to EAE-induction as WT mice [80, 365], which was not in line with a previous report demonstrating higher EAE-susceptibility of LMP7-deficient mice [73]. In contrast, both immunoproteasome deficiency as well as inhibition protected mice from development of dextran sulfate sodium (DSS)-induced colitis [195, 366, 367]. Impaired Th1 and Th17 polarization and 36

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases effector cytokine secretion in vitro and in vivo were associated with the effects of ONX 0914 in these models [80, 288, 366] and an altered STAT-signaling in response to IL-6 and IL-12 was reported by Kalim et al. as potentially involved in the underlying mechanism [288]. While several studies showed an effect of immunoproteasome inhibition on T cell cytokine production [80, 199, 366], a comprehensive mechanistic explanation for the beneficial effects of ONX 0914 in these autoimmune pathologies is still lacking today.

2 LYMPHOCYTES AND THEIR ROLE IN AUTOIMMUNE DISEASES

Immunity against pathogens is provided by physical barriers, humoral factors like the complement system, innate immune cells as well as adaptive immune cells. While innate immune cells sense pathogens by pattern recognition receptors (PRRs) to detect pathogen-associated or damage-associated molecular patterns (PAMPs, DAMPs), adaptive immune cells, which are comprised of T and B lymphocytes, acquire antigen-specificity and provide adaptive immunological memory. Both arms of immunity can be dysregulated in autoimmune pathologies, where inflammation is induced in the absence of a foreign pathogen [152]. Of note, innate cells of lymphoid progenitor origin have been long known: NK cells are able to kill target cells in spite of having no antigen-specific receptors [368]. Conceptual understanding of “innate lymphocytes” has greatly expanded within the past decades as the innate lymphoid cells (ILCs) were discovered and characterized. These cells show remarkable similarities to CD4+ T helper cell polarization states and are now recognized to be involved in responses to infection, anti-tumor responses as well as in inflammatory diseases [369, 370]. Accumulating evidence suggests that ILCs and adaptive T helper cells are often synergistically involved in immune responses and, therefore, in referring to type-1, type-2 and type-3 immune responses recent research articles most often include both types of cells [371– 373]. However, ILCs are not further investigated in this work and experiments with respect to activation and polarization of lymphocytes were mainly based on in vitro purified cell populations. Hence, this introduction is restricted to T and B lymphocytes.

2.1 B cells and humoral immunity

2.1.1 B cell development and activation

B lymphocytes derive from common lymphoid progenitor cells in the BM. Immature B cells bear membrane bound immunoglobulin receptors of the IgM class (mIgM), the specificity of which is a result of V(D)J-gene rearrangement during maturation. Central tolerance mechanisms in the BM help to deplete auto-reactive B cells before final maturation and peripheral tolerance control in the spleen give rise to mature naïve IgM+ and IgD+ B cells [374]. Unless activated, B and T lymphocytes are small spherical cells and depend on oxidative phosphorylation-derived ATP for migratory mobility while circulating

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases through blood and secondary lymphoid tissues [375]. Most B cells are activated with the help of T helper cells in the germinal center of a lymph node [376]. However, T cell-independent B cell activation can occur in innate-like B1-cells and in marginal zone B cells [377]. B cell activation leads to clonal expansion and a transition to antibody-secreting states. In germinal center B cells this is accompanied by somatic hypermutation and affinity maturation, metabolic changes and an extension of cytoplasm and ER [241]. Besides short-lived extra-follicular plasmablasts, long-lived plasma cells and memory B cells are formed during and after infection. Memory B cells provide protection upon re-challenge with the specific pathogen [378]. However, B cells do not only produce antibodies, but also secrete cytokines and serve as professionals APCs, thereby being involved in inflammatory responses to pathogens by multiple means [379]. Both human and mouse B cells are marked by the CD19 lineage marker (B cell co-receptor) and are activated via the B cell receptor (BCR). B cell activation signaling resembles T cell receptor mediated activation signaling, but shows some distinct differences. As this work primarily focusses on T cells and T cell receptor signaling, BCR-induced signal transduction will only be introduced along with TCR- signaling where important differences are highlighted. A T-independent polyclonal stimulus to activate B cells is lipopolysaccharide (LPS) derived from gram-negative bacteria, which triggers toll-like receptor 4 (TLR4). Thus, dual BCR and TLR-signaling can also induce B cell responses independent of follicular T cell help [380]. A synergistic co-stimulatory signal for BCR-induced activation is provided via CD40, which engages the alternative pathway of NF-κB signaling [381–383] (section 2.3.3).

2.1.2 Plasma cells and autoantibodies in autoimmune diseases and transplant rejection

Despite the existence of autoantibodies also in healthy individuals, presence of autoantibodies (mainly anti-nuclear antibodies, ANAs) is a classical diagnostic element in autoimmunity [384]. Primarily-B-cell- driven autoimmune pathologies include systemic lupus erythematosus (SLE), Hashimoto thyroiditis, Sjögren’s syndrome and myasthenia gravis [385]. B cells also contribute to other autoimmune pathologies like multiple sclerosis, rheumatoid arthritis (driven by both T and B lymphocytes) and play major roles for rejection after organ transplantation [385, 386]. SLE is marked by the formation of spontaneous germinal centers and the secreted autoantibodies can lead to vascular damage and immune complex glomerulonephritis ultimately leading to kidney failure [387]. Clinical attempts to target B cell pathologies have included cytostatic drugs and monoclonal antibodies against CD19 and CD20 (Rituximab) [388, 389]. A major issue is that established strategies to deplete B cells are poorly successful to deplete long- lived non-proliferating plasma cells. While memory B cells are re-activated upon re-challenge, long-lived plasma cells in the BM produce antibodies, but do not respond to secondary challenge with antigen [378]. Hence, cytostatic and classical immunosuppressive drugs help to overcome acute rejection, but fail to deplete long-lived plasma cells [390, 391]. Similarly, while rituximab-treatment depletes peripheral blood memory B cells, sufficient cell amounts survive in lymphoid organs and inflamed tissues to mount a

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases memory response [385]. These also contribute to T cell activation as APCs in allograft rejection and autoimmunity [386]. Proteasome inhibition with bortezomib showed clinical efficacy for treatment of refractory SLE [392]. Similarly as observed in murine model studies, proteasome inhibition reduced type-I interferons, decreased autoantibody levels and decreased plasma cells, but left non-activated peripheral blood B cells largely unaffected [392, 393]. In contrast, Mulder et al. report an overall inhibitory effect of bortezomib and ONX 0914 also on naïve human B cell subsets in vitro [394]. The beneficial effects of bortezomib in SLE are also mediated via impairment of pDCs [395]. As mentioned above, effectiveness of ONX 0914 in a lupus mouse model and recently also for preventing chronic rejection in a rat allograft kidney transplant model has been shown [361, 363]. The high demand for functioning ER and secretory pathways in plasma cells likely results in high susceptibility to proteasome inhibition, but evidence at the molecular level remains to be demonstrated. Additionally, by which mechanism naïve B cells would be affected by immunoproteasome inhibition, has also not been shown yet.

2.2 T cells and T helper cell subsets

The T lymphocyte arm of adaptive immunity arises from BM-derived T cell precursors that migrate to the thymus for further development and maturation. Like the BCR, the T cell receptor (TCR) is a result of somatic gene recombination during development. First, a pre-TCR is established by β-chain rearrangement and subsequent to productive signaling, the α-locus is re-arranged to give rise to a TCR with almost random specificity, albeit some genetic bias [396]. The interaction of the CD4+CD8+ double positive (DP) thymocyte TCR with either MHC-I or MHC-II on thymic epithelial cells paves way for ultimate lineage fate decision as CD8+ T cells (MHC-I) or CD4+ T cell (MHC-II) [397, 398]. Thymocytes thus undergo positive selection by productive interaction with MHC and are subsequently depleted by negative selection in case of too strong interaction with self-peptide-MHC presented on either cTECs, mTECs or intrathymic DCs according to the affinity model of thymocyte selection [399]. Self-reactive CD4+ T cells can also give rise to natural regulatory T cell formation in the thymus [400, 401]. Self-peptide-MHCs include self- antigens derived from promiscuous gene expression, a hallmark and unique feature of mTECs, regulated by autoimmune regulator (Aire)-dependent and independent pathways [402]. Notably, immunoproteasome and thymoproteasome play important roles in shaping the CD8+ T cell repertoire via antigen processing and possibly also via alleviating proteostasis-stress in mTECs (sections 1.4 and 1.5.2.).

Mature, naïve T cells can thus be subdivided into the CD8+ cytotoxic lineage with important roles in immunity against intracellular bacteria, viruses as well as tumorigenic cells and the CD4+ arm of T helper cells. Both are primarily activated for the first time in the LN after DCs have migrated from inflamed tissues where they have taken up antigen to the draining LN. Of note, DCs themselves are a group of heterogeneous subtypes like dermal DCs, Langerhans cells, LN-resident DC subsets, monocyte-derived

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

DCs or plasmacytoid DCs [403]. Interestingly, some CD8+ T cell responses are not elicited by DCs that have themselves taken up antigen in the periphery, but instead by LN-resident DCs that have received antigens from immigrating DCs. Also, CD4+ T cells have been implicated to provide the additional activation cues to LN-resident DCs that are not themselves activated by local DAMPs and PAMPs [404]. Thus, in a process called dual recognition some but not all CD8+ T cell responses depend on CD4+ T cell help to fully license DCs for efficient CD8+ priming via CD40/CD40L interaction and via enhancing DC- intrinsic innate pathways [403, 405–408]. Once primed and activated by a DC, CD8+ T effector cells interact with somatic cells via the almost ubiquitously expressed MHC-I molecules. Effective recognition of the cognate p-MHC on a target cell induces target cell killing by the CTL via Fas/FasL-interaction and directed secretion of granzymes and perforins into the immunological synapse between the CTL and the target cell [409].

CD4+ T helper cells play major roles for orchestrating adaptive immune responses and cooperate also with CD8+ T cells and B cells. Thus, in most autoimmune diseases, helper T cells are involved, even if the pathology is mainly driven by autoantibody-producing B cells for example [152]. Therefore, a detailed understanding of CD4+ T cells and means to interfere with auto-inflammatory dysregulation of CD4+ T cells is of high interest. The classical Th1/Th2 paradigm of the late 1980s was based on the observed functional differences between Th cell subclasses secreting different hallmark cytokines [410]. While the number of distinguished subclasses has expanded since then, they are still mainly defined by their cytokine profiles and additionally via hallmark transcription factors. Helper T cell subsets now include Th1, Th2, Th17, Th9, Th22, regulatory T cells (Tregs) as well as follicular helper T cells (Tfh), the latter of which are also characterized by surface markers and their localization to lymph node follicles in vivo [411]. Th9 and Th22 cells have been connected to skin inflammatory diseases and/or to tumor immunity and autoimmunity, but their functions are still less well characterized and understood [412, 413]. T helper cell polarization depends on the nature of antigen recognition via the T cell receptor as well as on the local environment at the site of T cell activation which skews T cell polarization towards a particular type mainly via cytokine signals but also via co-stimulatory molecules [414]. Not all principles underlying T cell priming and polarization instructions by DCs are fully understood. Nevertheless, many key factors orchestrating T cell polarization and many functions of the polarized subclasses have been determined and are briefly introduced for the most extensively characterized helper cell polarization states. These sections include the notion of relevant signal transduction pathways involved in establishing or maintaining the polarization states. Details on the TCR-induced pathways will then be introduced in section 2.3. An overview of the Th cell polarization pathways, the cytokine environments promoting their differentiation and hallmark effector cytokines produces by polarized Th cells is shown in Figure 4.

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

Figure 4: Schematic overview of T helper cell polarization Besides TCR signaling strength the main signals to promote polarization of a naïve T cell towards a particular polarization state is the cytokine environment promoting the expression of hallmark transcription factors or receptors (AHR, aryl hydrocarbon reptor). Polarized T helper cells furthermore produce effector cytokines involved in their functions in the tissue. Furthermore, considerable plasticity allows partial re- polarization of helper T cells in changing environments (not shown). Image taken from Kaplan et al. 2015 [412] and modified after Mirshafiey et al. 2015 [413]

2.2.1 Th1, Th2 and Th17 polarization

Th1 cell polarization is triggered by IL-12 and IFN-γ via STAT4 and STAT1 signaling and induces the hallmark transcription factor T box transcription factor (T-bet) [415]. Intracellular bacteria like Listeria monocytogenes, intracellular parasites like Leishmania major as well as virus-infections drive DCs to promote Th1 polarization [416, 417]. However, several other cell types can produce IFN-γ during the course of an infection, thereby stabilizing Th1 states at sites of infection. These include NK cells, macrophages, Th1 cells themselves, type-1 ILCs, CTLs and others [368, 370, 417]. Th1 cells can promote CD8+ T cell responses as mentioned above. In the peripheral tissue, Th1 cells promote the inflammatory environment against the invading pathogen, but interestingly, Th1 cells are also producers of tissue- protective IL-10 in protozoa-infections [418, 419]. A well described role of Th1 cells is the so-called classical macrophage activation, in which Th1 cells amplify phagocytosis and oxidative burst responses in macrophages via both cytokines and direct cell contact as macrophages present antigens from phagocytosed material on MHC-II. The resulting M1 polarized macrophages also produce Th1-, but also Th17-stabilizing cytokines [420]. Excessive activation of M1 macrophages and Th1 cells contributes to tissue damage and fibrosis [421].

Th2 priming by DCs has become more characterized in recent years. Multiple mechanisms can induce Th2 responses depending on the origin of antigen or allergen. TCR strength was found to modulate Th2

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases polarization and DCs involved in Th2 polarization seem to have a semi-mature state in that they up- regulate MHC-II and co-stimulatory molecules, but do not secrete polarizing cytokines [417, 422, 423]. Furthermore, granulocytes like basophils and eosinophils play a role in modulating DC functions and providing IL-4 and IL-13 for Th2 polarization [417, 424, 425]. In vitro, CD4+ T cells are skewed towards Th2 by IL-4 via STAT6, but STAT6-independent pathways for Th2 polarization exist [426, 427]. IL-2 can for example contribute to Th2 polarization via STAT5 to induce IL-4R expression [428]. The hallmark transcription factor is GATA3 and Th2 cells secrete IL-4, IL-5, IL-13, IL25 and formerly IL-9, while IL-9 secreting cells are meanwhile independently characterized as Th9 cells [426, 429, 430]. Th2 cells are involved in helminth infections, asthmatic and allergic diseases, fungal infections and environmental cold [431–433].

The cytokine IL-12 is a heterodimer composed of the p35 and the p40 subunits. However, the p40 subunit is a shared subunit of IL-12 and IL-23 [434]. Before the description of Th17 cells, the shared subunit has led to confusing results as both IFN-γ-deficient mice as well as IL-12p35-deficient mice developed severe EAE, which was not in line with the view that Th1 cells would be pivotal effector cells contributing to autoimmune pathologies. In contrast, IL-12p40 deficient mice were resistant to EAE [435]. The reason became clear when the role of IL-23 for Th17 cells was identified, which were described as a distinct lineage in 2005 [436–438]. The primary signals for Th17 polarization are IL-6 and TGF-β while IL-23 stabilizes Th17 cells during terminal differentiation [439–441]. IL-6 induces STAT3-phosphorylation, which promotes expression of the transcription factor RAR-related orphan receptor (ROR)γt [442, 443]. Notably, IL-6 also inhibits the induction of forkhead box protein 3 (Foxp3)-expression by TGF-β, a cytokine which is shared by the polarization programs of Th17 and Treg cells [440]. However, STAT3- signaling can also limit Th17 cells. In high-density cultures of human memory CD4+CD45RO+ T cells re- activation under pro-Th17 conditions can lead to a transient STAT3-signaling peak resulting in up- regulation of the ATP-hydrolyzing ectonucleotidase CD39, which impairs Th17 polarization [444]. Th17 cells secrete pro-inflammatory cytokines IL-17A, IL-17F, IL-22 and GM-CSF (granulocyte macrophage colony stimulating factor), the latter of which was found to be a major driver of auto-immune pathology in EAE and RA [373, 445–447]. Thus, dysregulated Th17 responses contribute to autoimmunity. With respect to pathogens, Th17 cells play an important role in mucosal tissues in response to fungal infections like C. albicans infection, which stimulate C-type-lectins on APCs to promote secretion of pro-Th17 cytokine profiles [338, 448], but Th17 cells are also involved in responses against intracellular bacteria [449]. Th17 cells recruit and activate neutrophils [450] and contribute to epithelial barrier integrity via IL-22 [451]. Notably, γδT cells in mucosal tissues are also a source of early IL-17 production in response to infection [452].

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

2.2.2 Regulatory T cells

Tregs play a central role in controlling inflammatory responses and are distinguished in peripherally induced Tregs originating from naïve CD4+ CD25– T cells and thymus-derived natural Treg cells (CD4+ CD25+ nTregs or tTregs). In vitro induced Tregs (iTregs) can be obtained by TCR-activation in the presence of IL-2 and TGF-β, which are the factors likely also underlying Treg induction in the periphery in vivo (peripherally induced Tregs, pTregs). Both, TGF-β-induced SMAD3 signaling as well as IL-2Rβ- mediated STAT5 signaling were found to promote expression of Foxp3 that was identified as the hallmark transcription factor of Tregs [453–458]. At the same time, IL-2/STAT5 inhibits Th17 polarization [459]. Even though exogenous IL-2 promotes Treg formation, TGF-β1 signaling suppresses endogenous IL-2 production in T cells via SMAD3-dependent and independent means and functional activation of T cells is ameliorated by TGF-β1, but independently of SMAD3 [460]. Repression of IL-2 is furthermore governed by Foxp3 itself [461, 462]. In fact, consumption of external IL-2 is one mechanism, how Tregs exert their functional role in suppressing pro-inflammatory T cell activation as Tregs express the trimeric high affinity IL-2 receptor, internalize and degrade the IL-2-IL-2-receptor complex upon binding and recycle the IL-2Rα chain (CD25) [463, 464]. Furthermore, Treg cells can suppress IL-2 production in conventional T cells by direct cell-contact attenuating NF-κB nuclear accumulation in target T cells [465–467]. As the partially overlapping requirements of cytokine stimuli for Th17 and Treg polarization indicate, the respective transcriptional programs were described to be cross-regulated. While IL-6 signaling can inhibit Foxp3 expression, Foxp3 can likewise inhibit RORγt [468, 469] and the contributions of different stimuli and signaling pathways in shaping the polarization outcome are still being further investigated. Interestingly, polarization states of Th1, Th17 or Treg cells are appreciated to be no terminal cell fates, but are marked by dynamic plasticity. Th17 cells can lose their IL-17 expression and start to express Foxp3 to become functionally suppressive cells [470, 471]. Indeed, the suppressive function of Treg cells was found to be primarily governed by Foxp3 itself. Foxp3 interacts with TCR-induced transcription factors NFAT and RUNX1 and competition with AP-1 and NF-κB for NFAT binding contributes to IL-2 suppression [464]. Recently, Foxp3 was found to preferentially bind to active chromatin-regions, but it forms different complexes to either activate Foxp3-activated genes or to repress Foxp3-repressed genes by excluding other activating transcription factors from the respective enhancer sites [472]. Thus, Foxp3 acts as a passive repressor, while active repression in complex with the EZH2 cannot be fully ruled out [473]. Foxp3 mediated exclusion of ELF-1 and ETS-1 from enhancer region binding is in line with previous reports showing that reducing MEK/ERK-signaling, which acts upstream of ETS-1, promotes Treg formation and impairs Th17 formation, in part also by regulating Foxp3 expression ([474, 475] and section 2.3.5). Exposure to self-antigens during thymic development of T cells can trigger clonal deletion [399]. However, medullary development in particular is permissive for self-reactive T cells by inducing natural or

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases thymus-derived Tregs (tTregs). A high affinity TCR-stimulation in the medulla is required, but not sufficient to induce tTreg formation [476, 477]. The precise contributions of TCR-induced signaling pathways and concomitant accessory signals to tTreg formation are only partially explored and overlap with iTreg formation to some extent, but are also distinct. Thymus-derived Tregs express Foxp3 and are positive for the CD25/IL-2Rα already at the naïve state. When they emerge from the thymus tTregs are functionally mature, i.e. they do not require additional T cell priming in the periphery to exert suppressive functions [464].

2.3 Signal transduction via the T cell receptor and co-stimulatory molecules

Signal transduction via the TCR is engaged upon multiple occasions throughout the life of a T cell and can lead to a variety of different functional outcomes ranging from clonal deletion in the thymus, T cell activation and polarization in the periphery to target cell killing by a CTL. While emerging insight into the specific differences between different occasions of TCR-triggering has been documented, many canonical pathways are shared in TCR-induced signaling. As a detailed description of differences in T cell activation is beyond the scope of this work, TCR signaling will be introduced with a primary focus on T cell priming of naïve T cells and re-activation of primed cells in the periphery. Furthermore, many signaling pathways are similar between BCR and TCR signaling and selected differences are mentioned within the following sections, but details on BCR signaling will not be further outlined.

2.3.1 Signaling initiation and proximal signaling

TCR signaling is initiated by interaction of the TCR with its cognate peptide-MHC (pMHC) on the surface of an APC. In spite of several decades of TCR research, until today it is not fully understood how the signal is initially converted from the cell exterior to the cell interior and how the TCR can so delicately discriminate between self-peptides and agonist-peptides even though the difference in affinity is sometimes only as small as 10-fold [478]. The TCR complex is composed of the TCRα and TCRβ chains (except for γδ T cells) and the CD3 chains CD3γ, CD3δ, 2 x CD3ε and 2 x CD3ζ leading to an eight-component complex. None of the TCR complex components has intracellular kinase activity, but the CD3 chains contain in total ten cytosolic immunoreceptor tyrosine-based activation motifs (ITAMs) in one TCR- complex (CD3γ, CD3δ and CD3ε contain one ITAM each, CD3ζ chains contain three ITAMs each with each ITAM bearing two tyrosine residues for phosphorylation). Several not mutually exclusive models for initial TCR-mediated activation have been proposed based on experimental investigation as well as computational modeling [478, 479]. Details on initial TCR triggering are beyond the scope of this introduction, but an overview of some interesting aspects of this complex topic will be provided. TCR crosslinking by multiple TCR-pMHC interactions was proposed to induce TCR clustering and initiate

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases signaling and crosslinking is likely involved in the anti-CD3ε mediated stimulation widely used for in vitro T cell activation [480, 481]. As the TCRαβ dimer has only very short cytoplasmic tails, mechanical signal transduction does not occur via conventional conformational change of a receptor, but conformational change is transmitted to the CD3-chains in the TCR complex in an allosteric fashion [482]. This for example involves bifurcation of the CD3ζ chains when in complex with the TCR, but close alignment when bound to pMHC [483]. Accessibility of CD3ε ITAMs for tyrosine kinases like non-catalytic region of tyrosine kinase adaptor protein (Nck) was also reported to rely on a conformational change and TCR clustering upon activation [480] and the ITAM sequestration model might account for the fact that p56Lck (Lck) is found in active conformation already in resting cells [478, 484]. Cholesterol was shown to play a role in signaling via the TCR complex by promoting formation of TCR nanoclusters in the absence of pMHC interaction, however promoting the inactive TCR complex conformation [485–487]. Additionally, mechanical force also contributes to stability of single agonist-pMHC-TCR interactions (catch-bonds), while it destabilizes non-agonist-pMHC-TCR interactions [488]. Interestingly, it has been demonstrated that T cells are able to recognize even only one to a few cognate pMHC molecules on the APC surface to elicit T cell activation [489]. Even in the presence of only one cognate pMHC molecule TCR clustering follows as a downstream event, which might be explained by models like serial triggering of several TCRs by one p-MHC [490], endogenous signal amplification of primed complexes towards other TCR complexes that have not engaged a specific pMHC [491] or by changes in the membrane microenvironment leading to complex clustering in lipid rafts [478]. Signal propagation then depends on interaction with cytosolic kinases, with Lck being of particular importance [492, 493]. CD4 or CD8 co- receptor binding to MHC-II or MHC-I, respectively, brings Lck into proximity with the TCR complex as Lck interacts with CD4 or CD8 and is anchored to the plasma membrane via myristoylation and palmitylation [494, 495]. Upon TCR and co-receptor engagement, Lck phosphorylates the tyrosine- residues in the ITAMs of the TCR complex and in the subsequently recruited non-receptor tyrosine kinase ZAP70 (ζ-chain associated protein of 70kDa), which binds to phosphorylated ITAMs via SH2-domains (B cells lack ZAP70, but express Syk, which acts functionally largely equivalent to ZAP70 in T cells). Furthermore, Lck contributes also to recruitment of Nck, the role of which for TCR-mediated ERK activation and actin rearrangement has only become more characterized in recent years [495–498]. The digital nature imposed on particular signaling pathways by multimodal feedback enhancement is outlined in the following sections. Activity of Lck is regulated by the membrane spanning phosphatase CD45 in a dual manner. On the one hand, CD45 de-phosphorylates Tyr505 in Lck, which is an inhibitory phosphorylation resulting in an inactive conformation of Lck. Thus, CD45 would contribute to Lck activity. On the other hand, the activating Tyr394 phosphorylation in Lck can also be de-phosphorylated by CD45, thereby counteracting Lck activity [478]. Interestingly, another phosphorylation site in Lck has been identified only recently and mutations of this Tyr192 were found to impair CD45-mediated de-

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases phosphorylation of the inhibitory sites in Lck, thereby shedding new light on the mechanistic regulation of Lck [499]. Regulation by CD45 is central to the kinetic segregation model which is also based on the finding that CD45 can also de-phosphorylate TCR-complex components, but this model could not alone account for experimental findings [478, 479]. The proximal signal initiation at the level of the TCR complex already influences the distinct downstream signaling events. For example, analysis of signal propagation from TCR-complexes with several phosphorylation-incompetent mutant CD3 chains has shown that the multiplicity of ITAMs has an impact on the functional outcome of T cell activation. TCR- complexes with WT ITAMs can promote both, cytokine secretion as well as proliferation. In contrast, TCR-complexes with only two or four functional ITAMs could fully elicit ZAP70-phosphorylation, downstream ERK-signaling and IL-2, IFN-γ or TNFα secretion, but they were not sufficient to induce c-Myc expression and proliferation due to impaired clustering of TCR-complexes to central supramolecular activation clusters (cSMACs) [500] from which CD45 is excluded independent of ZAP70 [501]. The proximal signal propagates downstream via ZAP70, as it initiates assembly of the linker of activated T cells (LAT)-signalosome, which together with SLP76 forms a scaffold for binding of several effector molecules (an equivalent scaffolding function in B cells is exerted by SLP-65/BLNK) [502–504]. Phospholipase Cγ1 (PLCγ1, PLCγ2 in B cells) is recruited and activated in concert with IL-2 inducible tyrosine kinase (ITK). PLCγ1 cleaves phosphatidyl-inositol-4,5-bisphophate (PIP2) in the plasma membrane to generate inositol-triphosphate (IP3) in the cytosol and the membrane-embedded 2+ diacylglycerol (DAG) [505]. While IP3 is a potent initiator of Ca NFAT signaling (section 2.3.4), DAG provides a membrane site for recruitment of protein kinase C theta (PKCθ) via its C1 domain as well as for RasGRP. The DAG-mimetic chemical phorbol-12-myrisate-13-acetate (PMA) stimulates this pathway and is hence used for TCR-complex independent T cell activation [506, 507]. DAG thus induces another branching point in canonical T cell activation signaling as RasGRP is involved in MAP kinase signaling (section 2.3.5) and PKCθ acts as an upstream initiating molecule for NF-κB pathway engagement (section 2.3.3). The LAT-signalosome also initiates Vav1-mediated actin cytoskeleton re-organization and activation of the c-Jun N-terminal kinase (JNK, also stress-activated protein kinase, SAPK). ZAP-70 and Vav1 initiate the p38 pathway via Rac1 (section 2.3.5). Early, proximal signaling is engaged within seconds after in vitro stimulation [508]. For in vivo pMHC-mediated activation that can distinguish between small differences in antigen affinities more complex kinetic proofreading models for initial activation have been proposed [478, 479].

2.3.2 Co-stimulatory signaling

T cells that receive an activating TCR stimulus in the absence of co-stimulation become anergic, i.e. they do not respond to secondary TCR stimulation (even in the presence of co-stimulation) or die from AICD [509–511]. Several co-signaling molecules were identified to play roles in different T cell – APC

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases interactions. Co-signaling can be either stimulatory or inhibitory depending on the context of TCR- engagement [512]. The probably most intensively characterized co-stimulatory signaling is B7-CD28 interaction. The B7-1 (CD80) and B7-2 (CD86) receptors are either constitutively expressed at low level on the surface of an APC or are inducible upon inflammatory stimuli or recognition of DAMPs/PAMPs. B7-1/B7-2 interact with CD28 expressed on T cells during initial activation or upon re-stimulation in the periphery and this provides pro-proliferative, pro-survival and cytokine-production-enhancing stimulatory signals for T cell activation, most of which depend on enhancing TCR-dependent gene expression, essentially contributing to IL-2 expression [513–516]. CD28 co-stimulation was also shown to be essential for GM-CSF production by CD4+ T cells via regulating the expression of the upstream transcription factor Dec-1/BHLHE40, which promotes Csf-2 expression (the gene for GM-CSF) [517, 518]. Therefore, the mechanisms of TCR-stimulation with CD28 co-stimulation are of major importance in pro-inflammatory and auto-immune T cell activation. In contrast, B7-1/B7-2 interaction with the structurally similar cytotoxic T lymphocyte antigen (CTLA)-4 has an inhibitory function for effector T cell activation and contributes to the inhibitory functions of Tregs [519, 520].

The stimulatory signals provided by CD28 co-stimulation converge on TCR-induced pathways in multiple ways, promoting important events for downstream signal propagation. For example, the cytoplasmic tail of CD28 contains a YMNM-motif which acts as a target sequence for phosphorylation by Lck and Fyn. These kinases bind to proline-rich regions in CD28 via SH-3 domains [521, 522]. Phosphorylated YMNM-motifs enable binding of phosphoinositide-3-kinase (PI3K) via the p85 subunit (in B cells CD19 functions directly as a LAT-like molecule and PI3K as well as Vav bind to CD19 cytoplasmic tails, thus contributing CD28- like function to B cell activation [503]). PI3K increases the local concentration of phosphatidylinositol- 3,4,5-triphosphate in the plasma membrane which is a membrane binding hub for pleckstrin-homology (PH) domain containing proteins like AKT and phosphoinositide-dependent kinase 1 (PDK-1). Furthermore, the PI3K-PDK1 axis promotes protein kinase C theta (PKCθ) activation, thereby directly contributing to the TCR-induced activation of PKCθ via PLCγ1 and diacylglycerol (DAG). PLCγ1 activity is itself promoted by CD28-PIP3-recruited ITK. The role of these signaling molecules as well as other partners of CD28 including GADS, SLP76 and GRB2 will be outlined in the following sections about canonical T cell signaling pathways. A simplified schematic overview of proximal TCR and CD28 co- stimulatory signaling and the downstream pathways that are initiated is shown in Figure 5.

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

Figure 5: Simplified schematic overview of proximal TCR signaling events and CD28 co-stimulation (exemplified for a CD4+ T cell). Upon recognition of the cognate p-MHCII via TCR and CD4 co-receptor binding, Lck becomes active and phosphorylates ITAMs in the CD3 chains. ZAP70 binds to phosphorylated ITAMs and gets both, auto-/trans-phosphorylated and phosphorylated by Lck. ZAP70 can directly initiate the p38 pathway and induces assembly of the LAT signalosome which acts as a signaling hub for several effector molecules. ITK-activated PLCγ1 produces DAG and IP3 initiating NF-κB activation, MEK-ERK pathway and Ca2+ NFAT signaling. CD28 co-signaling via binding to co- stimulatory molecules CD80-CD86 converges on TCR-induced signaling e.g. via PI3K-PDK1 and contributes to mTOR complex signaling via AKT thereby amplifying and promoting TCR activation. Vav1 activates both cytoskeletal re-organization as well as the JNK pathway. Dotted lines marked with “+p” inidicate modification by phosphorylation. Small yellow circles indicate phosphorylation.

2.3.3 Nuclear factor kappa B (NF-κB) signaling in T cells

The NF-κB pathway in canonical T cell activation is under control of both, the TCR-induced signal as well as the CD28-mediated co-stimulatory signal, which leads to activating phosphorylation of PKCθ [522, 523]. However, TCR-mediated signaling not only provides the membrane-recruitment hub for PKCθ, but also induces an activating PKCθ-phosphorylation via GCK-like kinase (GLK) which is regulated by the LAT-signalosome component SLP76, that can also bind to the CD28 cytoplasmic tail [506]. In turn, active PKCθ signaling initiates the so-called CBM-complex assembly (CARMA-1/Bcl-10/MALT1). This complex, which is modified by K63 ubiquitination, activates the inhibitor of kappa B kinase (IKK) complex consisting of IKKα, IKKβ and NEMO (NF-κB essential modifier, also named IKKγ) [14, 506]. Cross-regulation with other pathways like Ca2+-calcineurin signaling contributes to CBM-complex activation (section 2.3.4). It is the IKK complex which finally acts on IκBα by phosphorylating it at two serine-residues which precedes its ubiquitination and degradation by the proteasome. Proteasome inhibition thus normally leads to an accumulation of phosphorylated IκBα because of its impaired degradation and hence prevents NF-κB activation. The NF-κB family contains five members in mammals: RelA/p65, RelB, c-Rel, p105/p50 (NF-κB1) and p100/p52 (NF-κB2) [524]. The family members functionally act as dimers, most prominently as p50:p65, p50:c-Rel or p65:c-Rel dimers [524, 525]. The

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

IκBα-sequestered NF-κB dimer cannot exert its functional activity as a transcription factor unless pathway engagement induces IκBα degradation. Therefore, the proteasome is a well-known regulator of this essential signaling pathway that is required for full T cell activation and function. An additional role of proteasomes for NF-κB signaling is the partial processing of NF-κB subunit pre-cursors. The canonical pathway as described above (Figure 6A) involves mainly p50:p65 dimers. In contrast, the non-conical pathway of NF-κB activation involves CBM-complex independent NF-κB inducing kinase (NIK) in concert with IKKα, which induce phosphorylation and ubiquitination of the precursor p100/NF-κB2 subunit [526]. As a consequence, the IκB-like domain of p100 is degraded by partial proteasomal processing allowing dimers containing the resulting p52 to enter the nucleus. The role of non-canonical NF-κB activation via NIK in T cells is less well characterized than the canonical pathway. However, PKCθ-mediated canonical signaling is rapidly engaged after TCR/CD28-stimulation, while NIK-mediated non-canonical signaling depends on de novo protein synthesis and consequent processing of p100 [523]. A role for NIK in Th17 and Tfh polarization has been found [527]. NIK itself is furthermore regulated via targeted degradation induced by TNF-receptor (TNFR) associated factor 3 (TRAF3) together with cIAP1

Figure 6: Schematic overview of canonical and non-canonical NF-κB signaling and the involvement of proteasomes exemplified for T cells A) Canoncial NF-κB signaling is rapidly engaged upon stimulation via both TCR and CD28 (compare also Figure 5). In brief, TCR signaling via LAT signalosome, SLP76-GLK and PLCγ1 converge with CD28-PI3K-PDK1 signaling to activate PKCθ. Subsequently the CBM-complex of CARMA1-MALT1-BCL10 is activated which in turn activated the IKK-complex. CBM-complex activity is additionally modulated via Ca2+- calcineurin. IKK-complex mediated phosphorylation and degradation of IκBα releases the p65:p50 dimer from cytosolic retention and allows nuclear translocation and active gene transcription. B) Non-canonical NF-κB pathway engagement has slower kinetics and depends on de novo gene expression. In the absence of the activating receptor ligation, TRAF3 in complex with cIAP destabilizes NIK to promote its degradation by the proteasome (dashed lines). Activation of the non-canonical pathway (exemplified for stimulation via ICOS in T cells or CD40 in B cells) requires TRAF2-cIAP-mediated disruption of the TRAF3 inhibitory NIK regulation by promoting TRAF3 degradation and subsequent NIK accumulation. NIK induces IKKα activation which leads to phosphorylation and ubiquitination of the p100 precursor of NF-κB2. By partial processing via the proteasome, p100 is converted into p52 by removal of the IκB-like portion of p100. Thus, the p52:RelB dimer can translocate to the nucleus for active gene transcription (see text for details).

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases and cIAP2. Induction of the non-canonical pathway involves degradation of TRAF3 as TRAF2 re-directs and the E3-ubiquitin ligases cIAP1 and cIAP2 via non-degrading ubiquitin-linkages to promote TRAF3 ubiquitination and degradation, thereby allowing NIK to accumulate [527–530]. Thus, particular co- stimulatory or cytokine signals can engage non-canonical NF-κB signaling in T cells with functional relevance as shown in the EAE mouse model [531]. Another IκBα-independent pathway of NF-κB activation in T cells is direct phosphorylation of p65 by several possible kinases [532–534]. Notably, alternative NF-κB signaling via CD40 can be induced in both B cells and T cells [381–383, 535]. Figure 6 provides an overview of canonical and non-canonical NF-κB activation in T cells highlighting the involvement of proteasomes in both pathways.

2.3.4 Calcium signaling and nuclear factor of activated T cells (NFAT)

Divalent calcium ions (Ca2+) act as second messengers in cells and take part in TCR-induced T cell activation (Figure 7). Numerus Ca2+ binding proteins in the cytoplasm, sarco-/endoplasmatic reticulum Ca2+ ATPase (SERCA) and controlled Ca2+ ion channels allow a tight temporal and spatial regulation of Ca2+ in the cytosol, which typically contains a ~10,000-fold lower Ca2+ concentration than the extracellular fluid and a ~4,000 fold lower concentration than the lumen of the ER [536, 537]. TCR activation induced

DAG and IP3 formation (section 2.3.3) precedes the opening of ER-resident calcium-release channels resulting in a peak of increased Ca2+ in the cytoplasm. Additional Ca2+ influx into the cytoplasm is consequently mediated via store-operated calcium entry (SOCE) as depletion of Ca2+ from the ER activates opening of calcium-release activated calcium channel (CRAC) in the plasmamembrane. SOCE is induced by the interaction of STIM1 with the CRAC component Orai1 via ER-plasmamembrane apposition at the immunological synapse [537, 538].

Cytosolic Ca2+ is rapidly bound by calcium-scavengers like calmodulin. Caldmodulin in turn activates the calcineurin phosphatase which de-phosphorylates nuclear factor of activated T cells (NFAT), a transcription factor retained in the cytosol by hyper phosphorylation before activation [539]. Two of the four NFAT homologues are of primary importance in T cells: NFAT1 (NFATc2) and NFAT2 (NFATc1) [536, 540, 541]. Upon de-phosphorylation of NFAT it translocates to the nucleus for binding to target DNA sequences. Hence, NFAT de-phosphorylation and nuclear translocation is a downstream readout of the calcium signaling pathway in T cells. Using the Ca2+ ionophore ionomycin circumvents the upstream TCR signaling to provide Ca2+ entry immediately through the plasma membrane and is experimentally used for T cell stimulation. In contrast, the immunosuppressive drugs cyclosporine A (Neoral®, Novartis) and FK506 (Tacrolimus, Prograf®, Astellas) inhibit calcineurin mediated NFAT activation [539, 542]. Furthermore, calcineurin inhibition by cyclosporine A also impairs NF-κB signaling as Bcl-10 is de- phosphorylated by calcineurin for sustained CBM-complex activity [543].

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

Figure 7: Ca2+-NFAT signaling in T cell activation (modified after Hogan et al. 2010 [536]). A) Ca2+ concentrations are low in resting T cells compared to extracellular fluid. ER-resident SERCA transports cytosolic Ca2+ into the ER. NFAT is highly phosphorylated in resting T cells (e.g. by kinases CK, GSK2 and DYRK) and is retained in the cytosol. B) TCR engagement triggers PLCγ1 mediated IP3 generation which binds to IP3R at the ER to release ER calcium stores. Upon depletion of ER Ca2+, store operated calcium entry (SOCE) is mediated via STIM-Orai1. Elevated cytosolic Ca2+ binds to Calmodulin (CaM) and activated calcineurin to de-phosphorylate NFAT allowing it to enter the nucleus. External Ca2+ entry can also be triggered pharmacologically by the Ca2+ ionophore ionomcyin.

2.3.5 MAP-kinase signaling pathways

Mitogen activated protein kinase (MAPK) signaling pathways are conserved mostly three-tiered cascades of enzymes that activate each other in a directed upstream-to-downstream fashion. Generally, pathway engagement involves a MAPK kinase kinase (MAP3K, e.g. Raf) that phosphorylates a MAPK kinase (MAP2K, e.g. MEK), which in turn phosphorylates a MAPK (e.g. ERK). Such cascades are thought to allow signal amplification and cross-branching with other pathways at different tiers. The three major MAPK pathways that are also involved in T cells are: The Raf-MEK-ERK pathway, the JNK pathway and the p38 signaling pathway.

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

Ras and extracellular-signal-regulated kinase (ERK) – the Ras-Raf-MEK-ERK pathway

The small GTPase Ras is tailored to the membrane via its C-terminal membrane targeting motif (CAAX). It is inactive in it GDP-bound form and gets activated upon exchange of GDP for GTP with the help of activating Ras guanine-nucleotide exchange factors (RasGEF). In TCR signaling, the RasGEF son of sevenless (SOS) which is recruited to the LAT signalosome via growth factor receptor-bound protein 2 (Grb2), and also the Ras guanosyl-releasing protein 1 (RasGRP1), which is activated by DAG as well as PKC, activate Ras [502, 507]. Interestingly, the combined activation of Ras by RasGRP and SOS are a critical component of the analog-to-digital conversion of Ras-MEK-ERK signaling modality in T cells. GTP-bound Ras allosterically activates SOS, thereby promoting its own activation by SOS. Thus, while a TCR stimulus of low strength activates ERK signaling via RasGRP, Ras-MEK-ERK signaling intensity does not correlate to the TCR stimulus strength once a certain threshold is reached. Instead, the positive feeback loop of Ras-SOS-Ras activation elicits maximal signaling engagement in an all-or-nothing response [507, 544]. In contrast, PMA-treatment-elicited Ras-MEK-ERK signaling correlates in its intensity with the PMA concentration which mimics DAG concentration in the plasma membrane [544]. Basal ERK pathway engagement via RasGRP1 has been reported to preserve tonic signaling induced TCRα expression (compare also 2.3.7). Active Ras recruits Raf, which in turn phosphorylates and activates the MAPK/ERK kinases MEK1 and MEK2 which are the primary activators of ERK [545, 546]. MEK1/2 phosphorylate ERK1/2 at both threonine and tyrosine residues in the Thr-Glu-Tyr (TEY) motif within the activation loop. This modification activates ERK kinase activity but also allows its shuttling from the cytosol to specific subcellular locations, mainly the nucleus via binding to importin 7 [547]. More than 250 substrates modified by active ERK1/2 have been identified, including important transcription factors involved in proliferation like c-Myc, Elk-1 and c-Fos [546, 547].

Alternative pathways of ERK phosphorylation were found in peripheral T cells. Independent of LAT and independent of Ras a complex consisting of B lymphocyte adaptor molecule of 32 kDa (Bam32), PLCγ1 and the serine/threonine kinase p21 activated kinase (PAK)1 can activate ERK. This complex leads to direct Raf phosphorylation with consecutive S217/S221 phosphorylation of MEK by Raf and to direct MEK phosphorylation at S298 via PAK1 [507, 548]. Furthermore, Lck and CD28 mediated PKCθ activation contributes to RasGRP activity inducing ERK signaling via Ras, independently of LAT [507].

Several negative feedback regulators of T cell signaling act on the Ras-MEK-ERK signaling pathway. While Ras shows slow intrinsic GTPase activity, the Ras-inactivating hydrolysis of GTP to GDP is promoted by Ras GTPase activity promoting proteins (RasGAP). For example, DOK1 and DOK2 are adaptor proteins associating with Grb2 and ZAP70 with known negative regulator function for this pathway via recruitment of RasGAP, counteracting Ras activity [549]. In contrast, members of the dual specificity phosphatases (DUSP) can directly interfere with ERK via de-phosphorylation (section 2.4).

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases c-Jun N-terminal kinase (JNK) pathway

The three mammalian JNK family members (JNK1, JNK2 and JNK3, the latter of which is not expressed in T cells) exist in 10 total isoforms between ~46 kDa and ~54 kDa resulting from alternative splicing [550]. JNK (also stress-activated protein kinase, SAPK) signaling is involved in T cell signaling in a complex manner. Mice expressing dominant negative (dn)JNK1 showed reduced negative selection [551], but loss of JNK2 in combination with dnJNK1 or JNK1 deficiency resulted in normal thymocyte development [552], implicating that the balance of JNK1/JNK2 plays a role for apoptosis during negative selection. In contrast to the readily activated MEK-ERK pathway, primary naïve CD4+ T cells express very low levels of Jnk1 and Jnk2 mRNA, which is inducible upon prolonged T cell stimulation along with their upstream kinases MKK4 and MKK7 [550, 553]. Naïve splenic murine T cells activated via either anti-CD3, anti-CD28 or the combination of both did not induce JNK activation as measured by c-Jun co- precipitation in an early study [554], but transient early JNK phosphorylation peaks (5-15 min) are detectable in purified splenic T cells upon anti-CD3/CD28 activation in vitro as reported by newer studies [555, 556]. However, JNK proteins were reported to be dispensable for early T cell activation and were not essential for IL-2 production, but influenced T cell differentiation and effector cytokine production [552, 557]. In contrast, antigen experienced effector T cells and Jurkat cells show relatively fast JNK activation [550, 553, 558] and JNK stabilized IL-2 mRNA in activated Jurkat cells [559]. In these cases JNK activation is also synergistically regulated by CD28 and TCR as Vav1 recruitment to CD28 via GRB2 and

PIP3 as well as the ZAP70 induced Vav1-phosphorylation act as upstream signal inducers of the small G- protein Rac (for which Vav1 serves as a GDP/GTP exchange factor) and the subsequent JNK phosphorylation [521, 558, 560]. Differential roles were reported in CD8+ T cells where loss of JNK1 decreased LCMV-specific CTL responses, decreased IL-2Rα expression and resulted in failure of JNK1-/- mice to resolve Leishmania major infection [561–563]. In contrast, JNK2-deficiency enhanced IL-2 production and increased LCMV-specific CTL responses, but impaired Th1 polarization and IFN-γ production [561, 562]. Interestingly, both JNK and ERK phosphorylation have been found as part of IRE1 mediated unfolded protein response influencing cell death / cell survival decisions of prolonged ER stress [239, 564].

JNK phosphorylates the transcription factor c-Jun which is a part of the AP-1 complex together with c-Fos induced by ERK [550, 558], but also by JNK mediated Elk1 phosphorylation under stress conditions [565]. It was reported that JNK2 preferentially associates with c-Jun in unstimulated cells and destabilizes c-Jun, while JNK1 phosphorylates and stabilizes c-Jun upon activation, possibly explaining the abovementioned differences also in T cells [566, 567]. Furthermore, synergistic action of c-Jun and NFAT1 promotes IL-2 expression [568] and JNK signaling is promoted by Ca2+ signaling but impaired by cyclosporine A [558], which is likely attributed to both synergistic NFAT1-c-Jun activity, but also to cross-regulation of

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases calcineurin with the CBM-complex (section 2.3.4), as JNK is also activated downstream of PKCθ via the CBM-complex, CYLD and TAK1 [569–571]. TAK1-JNK promotes survival of activated T cells [572]. p38 signaling pathway

The 38 kDa sized p38 family of MAP kinases consists of four members (p38α, p38β, p38γ and p38δ). MAPK p38 signaling is activated upon stimulation of naïve T cells and synergistically induced by TCR and CD28 stimulation [550, 554, 560]. While MKK4 can activate both, p38 and JNK, MKK3 and MKK6 specifically activate p38 [550]. Upstream Vav1 signaling as described above initiates MKK3, MKK6 and MKK4 to promote p38 phosphorylation in the activation loop, which can be selectively recognized and phosphorylated by different upstream kinases for different p38 isoforms [573]. In parallel, direct phosphorylation of p38 at Tyr323 by ZAP70 is an alternative p38 activation mechanism in T cells, which induces consecutive p38 auto-phosphorylation at Thr180 in the Thr-X-Tyr activation loop motif of p38α and p38β [574]. T cells lacking this alternative pathway activation show reduced auto-immunity in EAE and CIA [575]. Indeed, alternative and classical p38 activation have partially opposing functions, as the classical p38 pathway inhibits, while the alternative pathway promotes NFATc1 and IRF4 expression [576].

A schematic overview of the three major MAPK pathways in TCR-induced signaling is presented in Figure 8. The functional relevance of MAPK pathways in T cell activation, polarization, cross-regulation of cytokine signaling pathways, autoimmune pathologies, cancer and other pathophysiological settings is complex and a detailed description is beyond the scope of this introduction. However, some aspects related to this work are briefly summarized: Sustained ERK phosphorylation as opposed to transient ERK phosphorylation shapes the outcome of T cell activation and polarization [422]. TGF-β signaling can transiently impair ERK sustainment, while IL-6 signaling can promote ERK signaling. ERK-inhibition was associated with enhanced Treg formation and reduced Th17 polarization [474, 475]. B-Raf mediated sustained ERK-signaling is necessary for CD69 and IL-2 expression [577]. These are examples showing that T cell receptor signaling outcomes are among other ways shaped by fine-tuning of the ERK signaling pathway (see also section 2.3.8).

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

Figure 8: Schematic overview of MAPK pathways in TCR-induced signaling (compare also Figure 5) A) The Ras-Raf-MEK-ERK pathway plays a central role in T cell activation. Due to the SOS-Ras-positive feedback loop, TCR stimulation alone is sufficient to induce ERK signaling in a digital signaling modality, but CD28 co-signaling also activates Ras-MEK-ERK signaling. Activated Ras leads to Raf phosphorylation via recruitment to the plasma membrane and conformational change of Raf. Raf in turn directly phosphorylates MEK1/2 leading to subsequent ERK phosphorylation. Notably, Raf and MEK1/2 can be also activated LAT-independently in T cells via Bam32- PAK1. Downstream substrates of ERK exist both in the cytosol and in the nucleus, where a majority of substrates are transcription factors. Negative regulators of ERK phosphorylation are dual specificity phosphatases existing both in the cytosol (e.g. DUSP6) as well as in the nucleus. ERK activity promotes c-Fos induction, which forms the AP1 complex together with c-Jun. See text and references therein for details. B) JNK and p38 signaling are primarily activated via Rac (and Cdc42, not shown) upon GDP-GTP exchange by activated Vav1. ZAP70 can furthermore directly phosphorylate p38, which leads to additional auto-phosphorylation of p38 at different residues. Rac activated MEKK4 can induce both p38 via MKK3 and MKK6 as well as JNK via MKK4. JNK2 preferentially binds and destabilizes c-Jun in unactivated cells, while JNK1 phosphorylates and stabilizes c-Jun, leading to differential effects of JNK1 or JNK2-deficiency. Several pathways including the CBM- complex together with CYLD, TAK1 and MKK7 activate JNK signaling, which is also part of the ER stress response. Phosphorylated c-Jun forms the AP1 complex together with ERK-induced c-Fos, but also functions in concert with NFAT. However, JNK signaling is important in activation of antigen experienced, but not prominently in naïve T cells upon primary activation. Downstream effects of p38 are not further outlines in the cartoon. See text and references therein for details.

2.3.6 Mechanistic target of rapamycin (mTOR) signaling and immunometabolism

The mTOR pathway plays a central role for integration of metabolic cues and TCR-activation as well as

CD28-co-stimulatory signaling. PIP3 generated by CD28 co-stimulation recruits both AKT and PDK1 allowing for PDK1 mediated AKT phosphorylation at Thr-308 [578]. While this phosphorylation precedes mTOR complex 1 (mTORC1) activation via phosphorylation of tuberous sclerosis complex 2 (TSC2), full activation of AKT also requires Ser473 phosphorylation downstream of mTORC2 (PI3K-related kinase, PIKK in complex with Rictor) [579]. Rapamycin treatment primarily inhibits mTORC1 signaling, but has only minor effects on mTORC2 [352]. The TCR mediated activation of mTORC1 depends on Ras-MEK- ERK signaling, which is needed for sustained mTORC1 activation [580, 581]. Downstream effector molecules of mTORC1 are p70 S6 kinase and 4E-BP1 [582]. T cells deficient for mTOR show enhanced Treg polarization and it was found that mTORC1 promotes Th1 and Th17 differentiation while mTORC2

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases is critical for Th2 polarization [583, 584]. Signaling via mTOR plays also a central role in the metabolic reprogramming from resting T cells to the activated T cell state. Resting cells predominantly use oxidative phosphorylation to generate ATP for their migratory capacity [375, 585]. Upon activation, T cells switch to anaerobic glycolysis in spite of sufficient oxygen (“Warburg” effect, a hallmark of cancer cells, see [586] and section 1.8.3). This glycolytic switch is accompanied with enhanced glycolysis-derived ATP production, while TCA cycle derivatives are used for anabolic demands and de novo protein synthesis. Thus, T cells share metabolic features with many cancers, for which the Warburg effect was originally described [375]. Up-regulation of glycolysis and glutaminolysis depend on co-stimulatory signaling and IL-2, the absence of which lead to T cell anergy. In particular glutamine uptake is also dependent on ERK signaling [587, 588]. Furthermore, T cell activation depends on autophagy, which is negatively controlled by mTOR signaling when nutrient supply is sufficient [589, 590]. Importantly, it was shown that macroautophagy plays a role for degradation of cytosolic contents (i.e. including soluble proteins) in T cell activation which contributes to degradation of negative regulators of T cell signaling [591, 592]. However, early activation markers CD69 and CD25 after overnight activation were not affected by Atg5-deficiency in T cells, which renders them autophagy-incompetent [593].

2.3.7 TCR signaling strength

While the term “TCR signaling strength” is not a clearly defined modality (i.e. some articles only refer to the pMHC-TCR binding affinity as “strength”, while others refer to the total duration and magnitude of the activation stimulus including co-stimulatory signaling) it has become established, that signaling strength has an important impact on T cell activation and polarization outcomes [422]. Even T cells that are not activated by cognate peptide-MHC recognition depend on tonic signaling induced by self-peptide-MHC engagement in the periphery, which is necessary to preserve responsiveness towards pathogens [594]. The abovementioned dependency of TCRα expression on tonic basal ERK signaling is just one example of how T cells might adjust TCR signaling strength during their life-time, a concept described as “TCR tuning” [422, 507, 595, 596]. Weak and transient signaling will result in different activation outcomes as opposed to strong and sustained signaling [422]. As mentioned above, modulating factors are the affinity of TCR to p-MHC as well as the multiplicity of TCR-p-MHC interactions and co-stimulatory signals. In line with the concept of digital signaling in T cells, single T cells can be fully activated to secrete cytokines by single p-MHC-TCR interactions and show TCR clustering in the presence of only one cognate p-MHC [489]. Increasing numbers of p-MHC-TCR interactions on a single cell do not increase the amount of cytokine produced per cell, but increase the likeliness that cells respond and thus the number of responding T cells [489]. TCR strength and clustering are also important to drive Myc expression and proliferation [500, 597]. With respect to polarization it was demonstrated in vivo that the quality of the TCR stimulus can be the dominant determinant for Th1 versus Th2 fate decision over the cytokine milieu, because the TCR signal

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases quality influences cytokine receptor expression in T cells [598]. However, how TCR strength is qualitatively signaled at the molecular level in T cells is not fully understood. Modeling implied that TCR strength discrimination can be signaled via negative feedback regulators towards initially digital ERK signaling [599]. Interestingly, a recent report by Allison et al. shows that the concept of digital ERK activation is not sufficient to describe activation in CD4+ T cells. Indeed, the authors provide evidence that the ERK signaling sustainment, which they measure at about 3.5 h after activation, translates TCR signaling strength into AP-1 responsive genes [600]. Similarly, insight into the influence of strength on molecular events guiding polarization was recently obtained for Th2 cells: The transcription factor IRF4 is directly dependent on TCR signaling strength and exerts its function in complex with basic leucine-zipper transcription factor ATF-like (BATF) transcription factors [423]. Increasing IRF4 changes the gene expression landscape via the different affinities of the BATF-IRF4 complex to enhancer regions in BATF- IRF4-dependent genes. Thus, weak signaling will initiate only genes with high affinity enhancers to BATF-IRF4, whereas low affinity target genes are only induced upon increasing BATF-IRF4 occupancies. Among the highly sensitive genes responding to weak signaling in Th2 cells was DUSP6, suggesting that DUSP6 might be involved in fine-tuning TCR signaling strength [423]. ERK signaling was shown before as a signaling pathway involved in polarization fate decision (section 2.3.5). Taken together, alterations of T cell activation signaling strength including the Ras-MEK-ERK pathway are appreciated to have effects on the outcome of T cell activation and polarization.

2.3.8 Functional outcome of T cell activation

While engagement of intracellular signaling complexes after T cell stimulation is a matter of seconds to several minutes in primary peripheral T cells [601], functional consequences can be measured by up- regulation of cell surface markers and cytokine secretion within a few hours after activation. Two of the earliest markers for T cell activation are the up-regulation of the C-type lectin CD69 on the T cell surface and the secretion of IL-2, which are driven by NF-κB and MAPK pathways [602–606]. CD69 is a stimulation induced leukocyte activation marker, but its physiological role has only become more characterized within recent years. In two different models of arthritis CD69-/- mice showed opposing disease susceptibilities. Injection of anti-type II collagen antibodies into CD69-/- mice induced milder arthritis than in WT mice [607]. In contrast, immunization of CD69-/- with collagen II in complete Freund’s adjuvant (CFA) resulted in exacerbated arthritis compared to WT mice [608]. While having no direct influence on T cell proliferation in the LN [609], CD69 expression promotes lymphocyte retention in the LN by interacting with S1P1 in the surface and favoring its internalization [610]. In 2014, galectin-1 on the surface of DCs and macrophages was identified as a specific ligand for CD69, which modulated T cell polarization [611]. Both galectin-1 and CD69 restrain Th1 and Th17 responses, while promoting Treg and Th2 formation [611–613]. Moreover, in 2016, CD69 was found to regulate the L-type amino acid

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases transporter 1 (LAT1)-CD98 complex in the plasma membrane and CD69-deficiency attenuated skin inflammation in response to IL-23. However, CD69-deficiency did not alter mRNA levels of Slc7a5, which encodes LAT1 [614]. Hence, CD69 plays a role in metabolic regulation of immune cells as both S1P1 and LAT1-CD98 mediated glucose uptake regulate mTOR signaling [615]. Depending on the disease model and cell types investigated, CD69-deficiency thus shows different effects with respect to pro- or anti-inflammatory regulations and the role of CD69 needs to be further elucidated in the future. In vitro, however, CD69-deficiency showed no direct influence on lymphocyte activation and proliferation [616].

The secretion of IL-2 from activated T cells is another early activation hallmark. It has been first described as a major inducer of T cell proliferation [617] and is an important co-stimulatory signal for a fully functional T cell response [463]. CD4+ T cells produce higher amounts of IL-2 than CD8+ T cells upon primary activation. The interleukin-2 locus is a highly regulated transcriptional target of TCR and co- stimulatory stimulation driven T cell activation [618, 619]. Ras-MEK-ERK signaling contributes to IL-2 expression via Egr-1 and the AP-1 transcription factor [620–623]. Notably, for both expression of IL-2 and CD69 the sustainment of ERK signaling was reported to be a critical regulator. When ERK phosphorylation sustainment, but not initial ERK phosphorylation peak was only slightly reduced by shRNA directed against Kidins220, expression of IL-2 as well as CD69 were significantly reduced [577]. Likewise, impairment of ERK signaling sustainment, but not initial ERK activation, with siRNA against Sos-1 resulted in reduced IL-2 and CD69 up-regulation [624]. In general, IL-2 plays a double role for the immune system in that it both promotes T effector cell responses as well as in promoting Treg survival and function, which in turn suppresses T effector cell responses [625]. Autocrine IL-2 signaling and transient high peaks of IL-2 signaling are involved in effector T cell differentiation towards Th1, Th2 and Treg and in CTL expansion [428, 619, 625]. In contrast, IL-2 rather constrains Th17 differentiation [459].

2.4 Dual specificity phosphatases and their role in T cell signaling

Kinase mediated phosphorylation as one of the major post-translational modifications involved in immune cell signaling is controlled and negatively regulated by phosphatases. Similar to the specificity of kinases for particular target residues, protein phosphatases can be distinguished as ser/thr-phosphatases (e.g. protein phosphatase PP2A), protein tyrosine phosphatases (PTP) and dual specificity phosphatases (DUSP) which can de-phosphorylate both, ser/thr phosphorylation sites as well as tyr-phosphorylation sites. The balance of phosphorylation and de-phosphorylation is a pivotal determinant of immune cell activation status and phosphatase inhibition is sufficient to initiate T cell activation [626]. This introduction focuses on the family of DUSPs and those family members reported to be specifically involved in MAPK signaling, particularly MEK-ERK signaling.

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

2.4.1 Dual specificity phosphatase family overview

Common substrates for DUSPs (also: MAP kinase phosphatases, MKP) are the MAP kinases, in which both phosphorylations within the Thr-X-Tyr motif can be de-phosphorylated by DUSPs. Depending on classification criteria, 25 to 43 proteins are classified as DUSP [627, 628]. Within the 25 DUSPs sharing sequence homology of the phosphatase domain, 11 also share the kinase interaction motif (KIM) or MAP kinase binding domain (MKB) and these are referred to as “typical” DUSPs. In contrast, 14 atypical DUSPs lack the MKB/KIM of which two are non-active pseudophosphatases (DUSP24/STYX and DUSP27) [627]. Importantly, also pseudophosphatases can have a regulatory function for MAPK signaling by acting as anchoring proteins scavenging target molecules at particular locations within the cell [629]. DUSPs have been implicated in regulation of signal integration by different input stimuli [630, 631]. The peak and duration of ERK phosphorylation can be highly variable depending on the sources of input signals. For example, it was reported that altered ERK signaling was mediated by the induced degradation of DUSP1 and DUSP6 in a fibroblast cell line after growth factor stimulation in combination with lactoferrin co-stimulation; hence these DUSPs were interpreted as central signal integration molecules [630, 631]. This example points to a general feature appointed to many DUSPs in cellular signaling to function as signal integrators. Depending on basal state expression in comparison to activation induced expression and further depending on the kinetics of the latter (e.g. as immediate early genes or as delayed early genes), DUSPs can influence signaling pathways in multiple complex ways: i) autoregulation of a pathway by expression of its own negative regulator, ii) pathway memory regulation by expression of a negative regulator upon first signal engagement that influences the signaling pathway upon re-challenge with the same signal, and iii) cross-talk between pathways by induction of a negative regulator downstream of pathway A, that negatively regulates pathway B [631].

DUSPs differ with respect to their cellular localization as well as regarding their substrate specificities. For example, the first identified DUSP (DUSP1 /MKP-1) was initially described as an ERK-dephosphorylating phosphatase [632]. However, later studies using MEFs derived from MKP-1-deficient mice revealed that DUSP1 was likely more selective for p38 and JNK [633]. Substrate binding properties and substrate specificity also show distinct biochemical features between DUSPs. Some DUSPs are activated via conformational changes upon binding of their substrate, which is reported for example for DUSP1, DUSP2, DUSP4 and DUSP6, but not for DUSP5 or DUSP10 [631]. Table 1 provides an overview of selected DUSPs including their main substrate preferences and reported cellular localization. This introduction includes only DUSPs for which a role in T cells was reported. Details are only covered for DUSP5 and DUSP6, which are the most relevant DUSPs for the results of this work. For further details see review articles by Caunt & Keyse 2013 [631], Huang & Tan 2012 [627] and Low & Zhang 2016 [634].

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

Table 1: Overview of selected DUSPs, their main substrates, cellular localization and domain structure. “Typical” DUSPs share the kinase interacting motif (KIM) and show specificity towards different MAPK. This table was taken from [627] and modified according to [653, 654, 663, 751, 752]. Main substrate for DUSP8 according to *[653] or **[654].

2.4.2 Expression and potential functions of dual specificity phosphatases in T cells

A gene expression database meta-analysis in 2010 revealed that several MAPK specific DUSPs are globally expressed in immune cells including DUSP1, DUSP2, DUSP4, DUSP5, DUSP6, DUSP7, DUSP10 and DUSP16 [635]. DUSP1-deficiency resulted in impaired T cell activation and reduced severity of EAE. Mechanistically, DUSP1-deficiency was accompanied with enhanced JNK and p38 phosphorylation and reduced NFATc1 nuclear translocation [636]. Conversely, DUSP1 was found to function in an apoptosis-preventive manner in T acute lymphoblastic leukemia (T-ALL), hence promoting cancer progression [637]. DUSP2/PAC-1 was originally cloned from human T cells, characterized as an inducible nuclear phosphatase [638] and shown to de-phosphoryate ERK2 [639]. More recently, STAT3 was reported as a novel DUSP2 substrate in T cells and DUSP2-deficiency led to enhanced colitis severity in mice accompanied with exacerbated Th17 polarization [640]. DUSP3/VHR is an ERK and JNK specific phosphatase in T cells [641]. Its activity is regulated by ZAP70 dependent Tyr138 phosphorylation, which promotes its inhibition of the ERK2-Elk1 pathway [642]. DUSP4 has been characterized as an ERK and JNK specific phosphatase [643] and is itself both induced by ERK signaling as well as post-translationally modified [644]. Interestingly, phosphorylation of DUSP4 by ERK stabilizes DUSP4 whereas ERK inhibition results in enhanced degradation by the proteasome [645]. In contrast, not DUSP4 phosphorylation, but direct binding of DUSP4 to ERK or JNK enhances catalytic activity of DUSP4 [645, 646]. However, CD4+ T cells from DUSP4-deficient mice were hyper-proliferative in vitro due to enhanced CD25 up-regulation and increased STAT5-phosphorylation, but no influence on ERK, JNK or p38 was detected during T cell activation [647], which resulted in enhanced Treg formation and reduced Th17 polarization and concomitantly decreased susceptibility to EAE in DUSP4-/- mice [648, 649]. Details on DUSP5 and DUSP6 in T cells are outlined in section 2.4.3. DUSP7/Pyst2/MKPX was also reported as 60

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases an ERK-selective DUSP, which was highly expressed in cells derived from acute myeloid leukemia patients, but was detectable in T cells only after TCR-stimulation [650]. In contrast, another study reported that DUSP7 was not induced by T cell stimulation, but constitutively expressed and degraded after prolonged activation and that it de-phosphorylated ERK in vitro and reduced NFAT-AP1 promoter activity in reporter cell assays [651]. Only recently, it was reported that a reduced expression of DUSP7 was found in T cells derived from seropositive RA patients compared to healthy individuals [652]. DUSP8 was discrepantly described as ERK specific [653] or as JNK and p38 specific [654] and only implied to regulate JNK activity in response to oxidative stress in Jurkat cells in one study [655], but comprehensive data about a role in T cells is lacking. Data on DUSP9 is also quite limiting and it was reported to be expressed selectively in plasmacytoid DCs but not in unstimulated CD4+ or CD8+ T cells [656]. However, it shares sequence similarity to DUSP6 and was also reported to be ERK1/2 specific [657, 658]. DUSP10/MKP5 was described as a primarily JNK specific phosphatase [659]. A function in CD4+ T cells and DCs was demonstrated as two studies showed increased IFN-γ secretion from DUSP10-/- T cells, which was boosted by co-culture with DUSP10-/- DCs [660, 661]. Reduced survival rates of mice after LCMV-WE infection were accompanied by enhanced IFN-γ and TNFα levels in serum, but in vitro proliferation was reduced in T cells from DUSP10-deficient mice and a partial protection from EAE induction was observed [660]. In contrast to DUSP10, both proliferation and cytokine secretion were enhanced in T cells from DUSP14/MKP6-deficient mice [556]. DUSP14 expression was reported to be CD28-stimulus dependent and DUSP14 interacted with the CD28 cytoplasmic tail. Expression of a dnDUSP14 in T cells resulted in enhanced JNK and ERK phosphorylation [662]. Interestingly, while enhanced ERK and JNK phosphorylation were also observed in DUSP14-KO T cells after stimulation, a direct binding and de-phosphorylation of TAB1 at Ser438 was also detected [556]. DUSP16/MKP7 was found to be inducible by anti-CD3 stimulation in T cells and its expression peaked within 3 h of activation in naïve T cells whereas in vitro differentiated Th1, Th2 and Th17 cells showed a constitutive DUSP16 expression that was not further induced after TCR stimulation [663]. Deficiency of DUSP16 in T cells led to increased IL-2 secretion and to reduced Th17 cell differentiation. Interestingly, blockade of IL-2 with scavenging antibodies rescued Th17 differentiation of DUSP16-deficient cells, indicating that DUSP16 is needed to restrain IL-2 production, likely by dephosphorylating ERK and JNK, which were more strongly phosphorylated in activated DUSP16-deficient T cells compared to WT cells [663]. Taken together, several MAPK selective DUSPs have been identified as modulators of T cell activation, but most detailed characterizations have been performed for DUSP5 and DUSP6, which are separately introduced in the following section.

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases

2.4.3 Regulation and functions of DUSP6 and DUSP5 in T cells

Functions for DUSP6 in T cells have been reported by several groups. Bertin et al. found that DUSP6 was rapidly induced in CD4+ T cells with either anti-CD3/CD28 stimulation or LPS-treatment. In DUSP6-/- T cells, ERK phosphorylation was enhanced compared to WT cells within 1 h after anti-CD3/CD28 stimulation [664]. Furthermore, DUSP6-deficiency enhanced IFN-γ secretion, but did not affect IL-2 secretion from T cells. Combined IL-10- and DUSP6-deficiency enhanced colitis severity in the IL-10 colitis model. Taken together, DUSP6 appeared to be a suppressor of the pro-inflammatory CD4+ T cells response [664]. In line with these findings, Gonzales-Navajas et al. found that LPS-induction of DUSP6 in T cells decreased subsequent TCR-activation-induced IFN-γ secretion which was functionally controlled using siRNA against DUSP6 [665]. Also DUSP5 was shown to play a role for T cell functions. Overexpression of DUSP5 in splenic murine T cells reduces their proliferative capacity [666]. When mice were genetically modified to overexpress DUSP5, type-II collagen induced arthritis was strongly attenuated and less p-ERK positive and p-STAT3 positive CD4+ T cells were found as tissue infiltrates [667]. A recent study by Kidger et al. showed how DUSP5 can act as a promoter of cytosolic ERK signaling in spite of being a nuclear ERK anchoring and dephosphorylating enzyme. By anchoring ERK in the nucleus, DUSP5 impaired direct negative feedback from ERK to the upstream kinase Raf [668]. In contrast, in cells expressing feedback-insensitive mutant B-Raf, ERK signaling is generally enhanced and DUSP5 acted as a suppressor of ERK phosphorylation throughout the cell [668].

DUSPs are regulated by a complex combination of transcriptional regulation, regulation of mRNA stability via RNA binding proteins and micro-RNAs, as well as regulation at the protein level by phosphorylation, ubiquitination and degradation by the proteasome. These layers of regulation were described in more detail for DUSP6 and DUSP5. First, DUSP6 expression is induced by the MEK/ERK pathway itself, which was described in fibroblast growth factor stimulated NIH-3T3 cells [669]. In human T lymphoblasts, IL-2 contributes to DUSP6 expression as DUSP6 is down-regulated after IL-2 withdrawal [670]. Also DUSP5 is regulated by IL-2 signaling as identified in the CTLL-2 cell line [671] and serum stimulated NIH-3T3 cells express DUSP5 in an ERK-dependent manner [672]. In the MCF-7 breast cancer cell line both DUSP6 and DUSP5 were expressed by phorbol ester stimulation in a MEK/ERK-dependent and PKC- dependent manner. However, siRNA against the ERK-downstream transcription factor Ets-2 only abolished DUSP6 expression, but not DUSP5 expression, which in contrast was abolished by siRNA against c-Jun [673]. Interestingly, the MEK/ERK pathway not only induces DUSP6 and DUSP5 transcriptionally, but also influences them at the post-translational level. ERK2 directly interacts with DUSP6 to induce phosphorylation of DUSP6 at Ser159 and Ser197, which leads to their subsequent ubiquitination and degradation by the proteasome [674]. Ser197, but not Ser159 is furthermore phosphorylated by the mTOR pathway [675]. Also DUSP5 was shown to be regulated by ubiquitination

Chapter I – Introduction – Lymphocytes and their role in autoimmune diseases and degradation by the proteasome [672]. Of note, Marchetti and colleagues observed, that the phosphorylation of DUSP6 does not change its phosphatase activity, indicating that degradation-primed DUSP6 still retains its catalytic function [674]. In line with the findings that DUSP6 is induced and also degraded by ERK signaling it was observed in vascular smooth muscle cells that DUSP6 has only a rather short protein half-life of less than 30 min [676]. Xie et al. also show that the ubiquitination of DUSP6 is enhanced in the presence of SM22α in these cells [676]. More recently, Cheng et al. identified protein kinase PKN2 in colon cancer cells as an activator of DUSP6 via direct interaction with the DUSP6 linker region. ERK-regulation via PKN2-DUSP6 had a negative influence on cytokine secretion from colon cancer cells, thus suppressing M2 macrophage polarization, which is depending on IL-4 and IL-10 [677].

As outlined above, ERK signaling strength and duration vary depending on cell type and context of signal initiation [422, 678] and DUSPs have been discussed to play a role in the regulation of how context- specific signaling is regulated by a cell [631]. With respect to TCR signaling, DUSP6 and DUSP5 were implicated to mediate TCR strength responses, as both are among the several target genes regulated by miR-181a [679]. This micro-RNA is highly expressed in DP thymocytes, but also in CD4+ T cells, whereas it is relatively less expressed in polarized Th1 or Th2 cells. Increased miR-181a expression augments T cell responses by enhancing p-ERK and p-Lck levels and increasing IL-2 secretion. Exogenous overexpression of miR-181a insensitive DUSP6 reduced p-ERK levels back to or even below control levels, thus indicating that the enhanced phospho-ERK levels might be due to absence of DUSP6, but are at least counteracted by DUSP6-restoration in the cells [679]. Interestingly, prolonged in vitro activation of CD4+ T cells with anti-CD3/anti-CD28 antibodies leads to degradation of miR-181a [680], which in turn is likely to permit higher DUSP6 and DUSP5 expression. The role of the miR-181a regulation of DUSP6 in T cells is further supported by the finding that mice deficient for the precursor mir- 181ab1 show enhanced DUSP6 protein levels and reduced p-ERK levels in DP thymocytes as well as in CD4+ SP thymocytes [681]. Notably, human miR-181a is more strongly expressed in umbilical cord blood as compared to adult peripheral blood [682]. The amount of miR-181a in T cells is also higher in humans at 20-35 years of age and was found to be declined in 70-85 year old humans, whereas DUSP6 protein levels increase with age [683, 684]. Also, the peak of ERK phosphorylation after in vitro activation is lower in T cells from elderly donors compared to younger donors [684]. Most recently, the involvement of DUSP6 in regulating TCR strength was implicated by the finding that DUSP6 is sensitive to weak TCR signaling-mediated gene regulation as a target of high BATF-IRF4 transcription factor complex affinity in Th2 cells (section 2.3.7 and [423]). At the mRNA level, Dusp6 transcripts are further regulated via their 3’UTR, which can modulate mRNA stability due to RNA binding proteins [685]. Taken together, negative feedback towards MAP kinase signaling via DUSP6 and DUSP5 is a well-described means of signaling regulation in CD4+ T cells.

Chapter I – Introduction – Aim of this study

3 AIM OF THIS STUDY

Since the development of the prototype immunoproteasome selective inhibitor ONX 0914 in 2009 [199], the array of available subunit-selective proteasome inhibitors has increased greatly [65]. A multitude of studies has demonstrated the potential of these inhibitors for the treatment of autoimmune diseases and cancer and it was demonstrated that T cell proliferation and polarization were altered by immunoproteasome inhibition in vitro and in vivo (section 1.8). Apart from influencing MHC-I antigen presentation, it became clear, that beneficial effects of ONX 0914 must be in part antigen-presentation independent [68]. Given that many pre-clinical models showing treatment susceptibility with ONX 0914 involve CD4+ T cell responses a direct influence of ONX 0914 on T cells has been shown before. However, the molecular mechanism underlying this T cell intrinsic effect has remained elusive. Early work on the mechanistic action of ONX 0914 showed that effects classically associated with proteasome inhibition like ubiquitin-conjugate accumulation and impaired NF-κB signaling did not apply to immunoproteasome inhibition [199].

The main aim of this work was to identify novel mechanistic details about the influence of ONX 0914 on T cell activation at the molecular level. The focus of this work was set on two main issues: 1) The influence of immunoproteasome inhibition and deficiency on T cell activation signaling and 2) the role of immunoproteasomes for general protein homeostasis in activated lymphocytes. To study these questions, a combined approach using ex vivo expanded T cells, primary isolated murine and human T cells as well as the murine CTL-derived cell line T1 was used. Differentation and functional polarization of CD4+ T cells in the presence of ONX 0914 and in the absence of immunoproteaosmes were analyzed. ONX 0914 treatment led to impaired early T cell activation, impaired proliferation and impaired Th17 and Treg polarization. The experiments indicated that ONX 0914 treatment effects were depending on synergistic LMP2 and LMP7 inhibition, while lack of immunoproteasomes in T cells had no functional effect on T cell activation as compared to WT cells. Therefore, LMP7-deficient cells served as negative controls to further investigate the selectively immunoproteasome dependent effects of ONX 0914 treatment. To further investigate the underlying mechanistic basis of impaired T cell activation, differentiation and polarization, expanded T cells were used as a model system to identify signaling pathways affected by ONX 0914 treatment. Consequently, the identified candidate pathways should be tested directly in primary naïve human and murine T cells. Using this approach, a reduction in ERK signaling sustainment was identified. Therefore, additional experiments were aimed towards a characterization of the underlying mechanism leading to reduced ERK signaling. The upstream kinase as well as candidate phosphatases that might regulate ERK signaling were analyzed. In order to understand how immunoproteasome inhibition might alter ERK-regulating phosphatases, the identified candidate phosphatase (DUSP6) was characterized

Chapter I – Introduction – Aim of this study in more detail with respect to its regulation at the transcriptional and at the protein level in T cells employing biochemical assays for protein turnover characterization. Finally, it was aimed to test if a causative functional relationship between the observed accumulation of DUSP6 in ONX 0914 treated cells and impaired T cell activation and ERK-signaling existed. Using T cells from DUSP6-deficient mice, the view of a causative relationship was disconfirmed in this study.

Apart from the focus on signal transduction, an influence of immunoproteasome inhibition and deficiency on proteostasis regulation was investigated. A detailed characertization of the proteasome subunit content in naïve murine and human lymphocyte populations was performed and unraveled an almost exclusive dependency of naive lymphocytes on LMP7-containing proteasomes, while LMP7-deficient cells could fully compensate loss of immunoproteasomes with an almost exclusive standard proteasome composition. These results served as a basis to further investigate the impact of immunoproteasome inhibition on proteastasis during lymphocyte activation. It was therefore aimed to characterize if and how immunoproteasome inhibition (and deficiency) influenced clearance of poly-ubiquitin conjugates during T cell and B cell activation. The appearance of accumulating ubiquitin-conjugates in ONX 0914 treated cells gave rise to further investigate the functional consequences of this observation with respect to stress response pathways and apoptosis. The experiments showed an up-regulation of β5c after T cell and B cell activation, which was boosted by ONX 0914 treatement. Therefore, the possible involvement of Nrf1 in mediating up-regulation of β5c was investigated. Finally, it was aimed to test, if the observed amelioration of T cell activation applied to antigen-specifically activated CD4+ T cell as well. Using TCR-transgenic SMARTA mice ameliorated T cell activation after ONX 0914 treatment was corroborated in vivo.

Chapter II – Materials and Methods – Materials

CHAPTER II: MATERIALS AND METHODS

4 MATERIALS

4.1 Chemicals Chemical substance Supplier (Cat# / order number) 2-Mercaptoethanol for synthesis Merck (Cat#805740) 5(6)-Carboxyfluorescein diacetate N-succinimidyl ester Sigma Aldrich (Cat#21888) (CFSE) Acetic acid, 96% Roth (Cat#X895.2) Acetic acid, glacial VWR (Cat#20104.298) Agarose Premium SERVA (Cat#11381.02) Ammonium persulfate Sigma Aldrich (Cat#3678) Bovine serum albumin Sigma Aldrich (Cat#A9647) Brefeldin A Sigma Aldrich (Cat#B6542) Bromphenol blue AppliChem (Cat# A2331.0025) Cycloheximide Sigma Aldrich (Cat#D2650) Dodecylsulfate Na-salt in pellets (SDS) SERVA (Cat#20765.03) Dimethyl sufloxide (DMSO) Sigma Aldrich (D2660)

Di-sodium hydrogen phosphate dihydrate (Na2HPO4) Carl Roth (Cat#T877.1) Ethanol Carl Roth (Cat#9065.3) Ethidium bromide Sigma Aldrich (Cat#1510) Ethylenediamine tetraacetic acid disodium salt Carl Roth (Cat#8043.2) dehydrate (EDTA) Glycerol VWR (Cat#24386.298) Glycine PUFFERAN Roth (Cat#3908.2) Hydrochloric acid VWR (Cat#2052.244) Ionomycin calcium salt Sigma Aldrich (Cat#I0634) L-glutamine Sigma Aldrich (Cat#G3126) LU-001i see section 5.4 for origin and supplier Methanol, reinst Chemical depository, University of Konstanz ML-604440 see section 5.4 for origin and supplier Paraformaldehyde Sigma Aldrich (Cat#158127) Phorbol 12-myristate 13-acetate (PMA) Sigma Aldrich (Cat#P8139) Ponceau S Sigma Aldrich (Ca#5938.1) Potassium chloride (KCl) Carl Roth (Cat#6781.1)

Potassium dihydrogen phosphate (KH2PO4) Carl Roth (Cat#3904.1) PRN-1126 see section 5.4 for origin and supplier N,N,N′,N′-Tetramethylethylenediamine (TEMED) Sigma Aldrich (Cat#9281) Saponin Sigma Aldrich (Cat#S7900) Sodium chloride (NaCl) Sigma Aldrich (Cat#3957.2)

Chapter II – Materials and Methods – Materials

Sodium hydroxide (NaOH) Sigma Aldrich (Cat#30620)

Sodium orthovanadate (Na3VO4) Sigma Aldrich (Cat#S6508)

Sulfuric acid (H2SO4) Sigma Aldrich (Cat#30743) Triton X-100 Sigma Aldrich (Cat#X100) Trizma® base Sigma Aldrich (Cat#T1503) Tween® 20 Sigma Aldrich (Cat#1379) Z-Leu-Leu-Leu-al (MG-132) Sigma Aldrich (Cat#C2211)

4.2 Disposables Disposables Supplier (Cat# / order number) BD Vacutainer, CPT, 2 ml FICOLL BD Bioscience (Cat#362782) Cellstrainer 40 µm, nylon, sterile BD FalconTM (Cat#352340) Cellstar® 12-well cell culture plate Greiner Bio-One (Cat#665180) Cellstar® 96-well cell culture plate, F-bottom Greiner Bio-One (Cat#655180) Cellstar® 96-well cell culture plate, V-bottom Greiner Bio-One (Cat#651180) CRYO.S PP, with screw cap, sterile Greiner Bio-One (Cat#126279) Light Cycler Capillaries Roche (Cat#04929292001) MACS columns MS, LS Miltenyi (Cat#130-042-401 & 130-042-201) Microplate, 96-well, Microlon®, high binding Greiner Bio-One (Cat#655061) Multiply® µStrip Pro 8-strip SARSTEDT (Cat#72.991.002) Parafilm M Sigma Aldrich (Cat#P7793) Pre-Separation Filters Miltenyi (Cat#130-041-407) Protran 0.45 NC Premium nitrocellulose membrane GE Healthcare (Cat#10600003) Protran 0.45 NC nitrocellulose membrane GE Healthcare (Cat#10600002) RoboStrip® PP low profile white Analytik Jena (Cat#0501000602) SD100 Cellometer Cell Counting Chambers Nexcelom Bioscience (Cat#CHT4-SD100-002) Serological Pipettes SARSTEDT (Cat#86.1254.001) Spectra/PorTM 1 6-8kDa MWCO dialysis kit SpectrumTM Labs (Fisher Scientific) (Cat#132645T) Syringe Filter 0.22 µm TPP (Cat#99722) Tube 15 ml, 120 x 17 mm, PP, sterile SARSTEDT (Cat#62.554.502) Tube 50 ml, 114 x 28 mm, PP, sterile SARSTEDT (Cat#62.547.254) Whatman 3MM-CHR GE Healthcare (Cat#3030-917)

4.3 Kits and Reagents Kit / Reagent Supplier 2-Mercaptoethanol (50 mM) Gibco (Thermo Fisher Scientific) (Cat#313500-010) 35S-cysteine/methionine Hartmann Analytik (Cat#IS-103) Acrylamide/Bis Solution 37.5:1 SERVA (Cat#10688.01) AIM-V® Medium, GlutaMAXTM Gibco (Thermo Fisher Scientific) (Cat#12055-091) AO/PI Staining Solution Nexcelom (Cat#CS2-0106-1ML) Biozym cDNA synthesis kit Biozym (Cat#331470L) Captisol Ligand (Cat#RC-0C7)

Chapter II – Materials and Methods – Materials

Chameleon® Duo Pre-stained Ladder LI-COR (Cat#928-60000) cOmpleteTM EDTA-free protease inhibitor Roche (Cat#4693132001) DCTM protein assay Biorad (Cat#500-0113 / 500-0114 / 500-0115) DMEM Medium GlutaMAXTM Gibco (Thermo Fisher Scientific) (Cat#31966-021) DMEM/F12, no phenol red Gibco (Thermo Fisher Scientific) (Cat#21041-025) dNTP mix Genaxxon (Cat#: M3018, M3019, M3020, M3021) EZviewTM Red Protein A Affinity Gel Sigma Aldrich (Cat#P6486) FastStart DNA Master SYBR green-I kit Roche (Cat#12239264001) Fetal Bovine Serum Gibco (Thermo Fisher Scientific) (Cat#10270-106) Ficoll-PaqueTM PLUS GE Healthcare (Cat#17-1440-03) Fixation/Permeabilization concentrate eBioscience (Cat#00-5123-43) Fixation/Permeabilization diluent eBioscience (Cat#00-5223-56) GoTaq® G2 Flexi DNA Polymerase kit Promega (Cat#: M8291, M8295, M8296, M8297, M8298, M7806) Human serum, male AB plasma Sigma Aldrich (Cat#H4522) MACS® CD19 MicroBeads, mouse Miltenyi (Cat#130-052-201) MACS® CD4 (L3T4) MicroBeads, mouse Miltenyi (Cat#130-049-201) MACS® CD4+ T cell isolation kit, mouse Miltenyi (Cat#130-104-454) MACS® CD14 MicroBeads, human Miltenyi (Cat#130-050-201) MACS® CD4 MicroBeads, human Miltenyi (Cat#130-045-101) MACS® CD19 MicroBeads, human Miltenyi (Cat#130-050-301) Mouse IL-2 ELISA Ready-Set Go!® eBioscience (Cat# 88-7024-88) Mouse IL-6, recombinant protein eBioscience (Cat#14-8061-80) Murine IFN-γ, recombinant Peprotech (Cat#315-05) Odyssey® Blocking Buffer TBS LI-COR (Cat#927-50000) PageRuler Prestained Protein Ladder Thermo Fisher Scientific (Cat#26616) Penicillin-Streptomycin Thermo Fisher Scientific (Cat#1514122) PhosSTOPTM Roche (Cat#4906845001) Recombinant human TGF-β Peprotech (Cat#100-21) Reverse Transcription Kit Promega (Cat#A3500) RNAse AWAY Molecular BioProducts (Cat#7005-11) RNeasy Mini kit QIAGEN (Cat#74104) RPMI Medium 1640, GlutaMAXTM Gibco (Thermo Fisher Scientific) (Cat#61870-106) RPMI Medium 1640, without cysteine / Sigma Aldrich (Cat#R7513) methionine / glutamine Sensiscript RT Kit (50) QIAGEN (Cat#205211) SmartLadder Small Fragment (SF) Eurogentec (Cat#MW-1800-04) Sodium pyruvate solution (100 mM) Gibco (Thermo Fisher Scientific) (Cat#11360070) Vybrant® MTT Cell Proliferation Assay Kit ThermoFisher Scientific (Cat#V13154)

4.4 Buffers and solutions Buffer Name Recipe Blotting buffer 1 x running buffer, w/o SDS, 20% methanol

Chapter II – Materials and Methods – Materials

Cytosolic extraction buffer 5 mM HEPES, 75 mM NaCl, 2.5 mM KCl, 0.5 mM MgCl2, ph 7.4, 0.1% NP-40, 1x PhosSTOP Roche)

FACS buffer PBS, 2% FBS, 2 mM EDTA, 2 mM NaN3 NET-T NET-TON with 650 mM NaCl NET-TON 50 mM TrisHCL, pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% Triton X-100

PBS 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4 in

ddH2O, pH 7.4 PERM buffer FACS-buffer + 0.1% Saponin Ponceau Red staining solution 500 mg Ponceau S, 1 ml acetic acid (100 ml) Running buffer (10 x, 1 L) 30.6 g Trizma base, 144 g glycin SDS sample buffer (3 x, 26 ml) 8 ml glycerol, 0.4 ml Tris-pH 7, 6 ml 20% SDS TAE buffer (50 x, 1L) 242 g Trizma base, 57,1 ml glacial acetic acid, 100 ml 0.5 M EDTA TBS-T Tris-buffered saline (per L for 10x: 24.8 g Trizma base, 80 g NaCl pH 7.6), 0.15% Tween-20 Whole cell lysis buffer (& 1% NP-40, 137 mM NaCl, 1 mM EDTA, 20 mM Tris-HCl, pH 7.2 at nuclear lysis) 25°C, 2 mM Na3VO4, 0.15% SDS, 1x PhosStop (Roche), 1x complete protease inhibitors (Roche), 0.1% sodium deoxycholate

4.5 Oligonucleotides Name Sequence (5’-3’) PCR program (10 min at 95°C initial denaturation in all)

CS_mmu_LMP7_for GAA CAA AGT GAT CGA GAT TAA CCC 95°C 10 s CS_mmu_LMP7_rev GTC CTG GTC CCT TCT TGT C 62.5°C 10 s 72°C 10 s 38 cycles CS_mmu_β5c_for TGG CCT TCA AGT TTC TCC A 95°C 10 s CS_mmu_β5c_rev AGA TCA TGG TGC CCA TAG AC 57°C 8 s 72°C 12 s 38 cycles mmu_β1c_for TCG AGT GAC TGA CAA GCT GAC C 95°C 10 s mmu_β1c_rev GAA CAG AGT ACA CCT GCC CTC C 62°C 15 s 72°C 20s 40 cycles mmu_DUSP6_01_for CCC AAT AGT GCA ACG GAC TC 95°C 10 s mmu_DUSP6_01_rev GGG CTT CAT CTA TGA AAG AAA TGG 57°C 8s 72° C 13 s 40 cycles mmu_CD69_for CCT TGG GCT GTG TTA ATA GTG G 95°C 10 s mmu_CD69_rev GCT TCA GAA ACG TCA TGT CCT 59°C 8 s 72 °C 14 s 40 cycles

Chapter II – Materials and Methods – Materials mmu_IL2_for ACA TTG ACA CTT GTG CTC CT 95°C 10 s mmu_IL2_rev TTC CTG TAA TTC TCC ATC CTG C 58°C 8 s 72°C 8 s 40 cycles mmu_RPL13a_for TGA AGG CAT CAA CAT TTC TGG 95°C 10 s mmu_RPL13a_rev GGT AAG CAA ACT TTC TGG TAG G 57°C 6 s 72°C 15 s 40 cycles mmu_IPO8_for GTA CTT TAC AGA CAT GAT GCC TC 95°C 10 s mmu_IPO8_rev AAT GAA TAG CGG GAT ACA CTG 56°C 6 s 72°C 11 s 40 cycles DUSP6_geno_WT-REV CGA CTC GTA CAG CTC CTG TG 94°C 2 min 94 °C 20 s DUSP6_geno_MUTREV GCT CTA TGG CTT CTG AGG CG 65° C 15 s 10 cycles 68°C 10 s 94°C 15 s DUSP6_geno_COMFOR AAA CTG GGC ACC TTC ATT CA 60°C 15 s 28 cycles 72°C 10 s 72°C 2 min Oligonucleotides were synthesized by Mircosynth and delivered as lyophilized desalted Genomics Scale products that were re-solubilized to 100 µM stock concentrations in MilliQ H2O. Aliquots were stored at - 20°C with a concentration of 50 µM.

4.6 Antibodies for flow cytometry Epitope/target and fluorophore Supplier Catalogue# anti-mouse CD69-FITC eBioscience Cat#11-0691-85 anti-human CD69-FITC Biolegend Cat#310904 anti-mouse CD4-FITC BD Pharmingen Cat#553651 anti-mouse CD4-APC eBioscience Cat#17-0041-83 anti-human CD4-APC BD Pharmingen Cat#555349 anti-mouse CD25-PE Biolegend Cat#101904 anti-mouse CD19-APC Biolegend Cat#115512 anti-rabbit Alexa647 F(ab’)2 Invitrogen Cat#A21246 anti-rabbit Alexa488 F(ab’)2 Invitrogen Cat#A11070 anti-mouse TCR Vα2-PE BD Pharmingen Cat#561078 anti-Foxp3-PE (FJK-162) eBioscience Cat#12-5773 anti-IL-17A-APC eBioscience Cat#17-7177-81 anti-IL-17A-FITC eBioscience Cat#11-7177-81

Chapter II – Materials and Methods – Materials

4.7 Antibodies for immunoblotting Antibody / Epitope Supplier / Origin Catalogue# or reference anti-p-ERK1/2 Cell Signaling Cat#4370 anti-ERK1/2 Cell Signaling Cat#4695 anti-p-JNK (G9) Cell Signaling Cat#9255 anti-p-JNK (8E11) Cell Signaling Cat#4668 anti-JNK Cell Signaling Cat#9252 anti-p-STAT3 (Tyr705) Cell Signaling Cat#4113 anti-STAT3 Cell Signaling Cat#9139 anti-p-Akt(Ser473) Cell Signaling Cat#4060 anti-p-Akt(Thr308) Cell Signaling Cat#13038 anti-NFAT1 Cell Signaling Cat#5861 anti-phospho-p65 Cell Signaling Cat#3033 anti-NF-κB p65 Cell Signaling Cat#8242 anti-p-MEK1/2 Cell Signaling Cat#2338 anti-p-eIF2α Cell Signaling Cat#3398 anti-ATF4 Cell Signaling Cat#11815 anti-p53 Cell Signaling Cat#2524 anti-PSMB5 (β5c) Cell Signaling Cat#12919 anti-PSMB6 (β1c) Cell Signaling Cat#13267 anti-PSMB7 (β2c) Cell Signaling Cat#13207 anti-Lamin A/C Cell Signaling Cat#2032 anti-DUSP10 Cell Signaling Cat#3483 anti-phospho-p38 Cell Signaling Cat#9215 anti-p-S6 Cell Signaling Cat#4858 anti-p-IκBα Cell Signaling Cat#2859 anti-IκBα (N-term) Cell Signaling Cat#4814 anti-PARP Cell Signaling Cat#9532 anti-CHOP Cell Signaling Cat#2895 anti-DUSP16 Cell Signaling Cat#5523 anti-DUSP3 (VHR) Santa Cruz Biotechnology Cat#sc-274161 anti-DUSP9 (MKP-4) Santa Cruz Biotechnology Cat#sc-377106 anti-DUSP7 (PYST-2) Santa Cruz Biotechnology Cat#sc-137010 Mono-and-polyubiquitinated ENZO Lifesciences Cat#BML-PW8810 conjugates monoclonal Ab (FK2) anti-DUSP4 (MKP-2) BD Biosciences Cat#610850 anti-DUSP6 (MKP-3) Abcam Cat#ab220811 anti-DUSP5 Abcam Cat#ab200708 anti-TCF11/Nrf1 Cell Signaling Cat#8052 anti-Nrf2 Cell Signaling Cat#12721 anti-γ-Tubulin Sigma-Aldrich Cat#T6557 anti-α-Tubulin Sigma-Aldrich Cat#T5168 anti-LMP7 (mouse) Polyclonal rabbit antiserum, in house Khan 2001 [144]

Chapter II – Materials and Methods – Materials anti-LMP2 and anti-MECL-1 Polyclonal rabbit antiserum, (provided Guillaume 2010 [93] by Benoit J Van den Eynde, Ludwig Cancer Research, Brussels Branch, Brussels, Belgium) anti-IOTA Monoclonal, IB5, obtained from Klaus Kremer 2010 [143] Scherrer (Institute Jaques Monod, Paris, France) anti-LMP7 (human) Polyclonal rabbit antiserum, in house Macagno 1999 [116]

4.8 Antibodies for functional assays

Epitope/target Supplier Catalogue# anti-CD3 clone 17A2 Biologend (or eBioscience) Cat#100202 anti-CD28 clone 37.51 Biolegend (or eBioscience) Cat#102102 anti-CD40 clone 1C10 Biolegend Cat#102810 F(ab’)2 anti-mouse IgG FG Purified eBioscience Cat#16-5098-85 anti-IFN-γ (XMG1.2) eBioscience Cat#14-7311-85 anti-IL-4 (11B11) eBioscience Cat#16-7041-81

4.9 Mouse strains and animal protocols C57BL/6J (H-2b) mice were originally purchased from Charles River. LMP7-/-[158], and LMP2-/-[67] mice were kindly provided by John J. Monaco (Cincinnati Medical Center, Cincinnati, USA). SMARTA mice [686] (SM1-Ly5.1) were provided by the Swiss Immunological Mutant Mouse Repository. DUSP6-/- mice [687] were purchased from Charles River. Animals were kept in an SPF environment in the animal facility at the University of Konstanz. Animal experiments were approved by the review board of Regierungspräsidium Freiburg (G-16/154, T-16/15TFA and T-18/03TFA).

4.10 Lymphocytic choriomeningitis virus (LCMV) LCMV-WE was propagated in the L929 fibroblast line, and viral stocks were kept at -70°C [144]. Growth and titration on L929 cells was performed as described in [688] (kind gift from Ulrike Beck). Mice were infected i.v. with 4 x 105 pfu LCMV-WE.

4.11 Cell lines Cell line Origin and description Reference CTLL-2 Cytotoxic lymphoid line-2 cells originate from C57BL/6 mice immunized [689] with F4-5 Friends-virus induced leukemia obligatorily dependent on IL-2 for growth. Provided by Hajo Haase, TU Berlin, Germany X63-m-IL-2 A functional mouse IL-2 gene was inserted into the plasmacytoma cell line [690] X63-Ag8.653 resulting in a IL-2 producing cell line.

Chapter II – Materials and Methods – Materials

T1 Mouse CTL clone derived cell line with a TCR specific for a chemically [691–693] modified peptide used for photoaffinity labeling. Provided by W. Schamel, University of Freiburg, Germany (with permission from I. Luescher) Molt4 Human T-ALL patient derived T lymphoma cell line [694] L929 Provided by F. Lehmann-Grube, Heinrich Pette Institut, University of [688] Hamburg, Germany

4.12 Devices and Software Device Description / Method Supplier AccuriC6 Plus Flow Cytometer BD Biosciences Cellometer 2000 Auto Cell Counter and Viability Assessment Nexcelom Centrifuge 5471R Table top centrifuge Eppendorf ChemiDocTM XRS System Gel Imager BioRad DUOMAX 1030 Laboratory shaker Heidolph FACSCalibur Flow Cytometer BD Biosciences Hydro Speed Plate washer TECAN Infinite M200 Pro Plate reader for absorbance measurement TECAN Leica DMIL LED Light microscope Leica Light Cycler Quantitative real-time PCR Roche LSRFortessa™ Flow Cytometry BD Biosciences Multifuge 4KR, LH-4000 Centrifuge Heraeus buckets Modell 583 Gel Dryer Gel Dryer BioRad NanoVue Nanodrop liquid absorbance measurement GE Healthcare Odyssey Imager FC Near-infrared immunoblotting LI-COR Personal Molecule Imager Phosphoimager BioRad pH-meter 766 Calimatic pH-meter Knick Power PAC 3000 Power supply in gel-electrophoresis and blotting BioRad T3 Thermocycler Standard PCR Biometra Thermomixer C Reaction tube heating/mixing Eppendorf TProfessional Thermocycler Quantitative real-time PCR Biometra TopCount NXTTM Microplate Scintillation & Luminescence Counter Packard Vortex Genie 2 Fluid mixer Scientific Industries Software Supplier / Origin Version AccuriTMC6 Software BD Biosciences 1.0.264.21 GraphPad Prism® GraphPad Sofware, Inc. 6.04 FlowJo FLOWJO, LLC V9 and V10 Image Studio Lite LI-COR Ver 5.2 Inkscape freely available under GNU General Public License 0.92.2 PerlPrimer freely available under GNU General Public License v1.1.21 FACS DivaTM BD Bioscience 8.0.1 Quantitiy One BioRad 4.6.6

Chapter II – Materials and Methods – Methods

5 METHODS

5.1 Cell culture and cell stimulation

Cell viability was checked by trypan blue or AO/PI staining before experiments (Nexcelom Bioscience). Human T cells were isolated from PBMCs of healthy volunteers according to the Miltenyi human CD4+ T cell isolation protocol and cultured in AIM-V medium (2% human serum, 50 µM 2-mercaptoethanol, 10% FBS, 1% Pen-Strep). Cells were activated with the Human T cell activation and expansion kit (Miltenyi) according to the manufacturer’s protocol.

5.1.1 CTLL2

CTLL2 cells [695] were kindly provided by Prof. Hajo Haase, TU Berlin, and cultivated in RPMI 1640 (10% FBS, 1% Pen-Strep, 50 µM 2-mercaptoethanol, 1 mM sodium pyruvate) in the presence of 30 U/ml recombinant mouse IL-2. CTLL2 cells were used to assess activity of newly harvested recombinant IL-2 (produced with x63-m-IL2 cells [690], kindly provided by Ulrike Beck) in an MTT assay (section 5.2), which was afterwards used for cultivation.

5.1.2 T1 cells

T1 cells [691] were kindly provided by Prof. Wolfgang Schamel, University of Freiburg, with permission from Prof. Immanuel Lüscher, University of Lausanne, and cultured in RPMI 1640, supplemented with 10% FBS, 1% Pen-Strep and 50 µM 2-Mercaptoethanol. Activation was performed with plate-bound anti- CD3/CD28 (5 µg/ml in PBS each) (Biolegend/eBioscience).

5.1.3 Primary murine CD4+ T cells and CD19+ B cells

Different lymphocyte populatoins from spleen were isolated with CD19 beads, CD4+ T cell isolation kit or CD4 beads (Miltenyi) according to the manufacturer’s protocols and cultured in RPMI 1640 +supplements. T cell activation was performed with plate-bound anti-CD3/CD28 (5 µg/ml in PBS each) (Biolegend/eBioscience). Murine B cells were activated with 50 ng/ml PMA (pre-solved in DMSO) and 500 ng/ml ionomycin (pre-solved in DMSO) or 5 µg/ml anti-CD40 (Biolegend) and 10 μg/ml F(ab’)2 anti- mouse IgG (eBioscience). Bulk lymphocytes were enriched with Ficoll-Paque centrifugation using 3 ml Ficoll-Paque solution and 5 ml cell suspension. Centrifugation was performed in LH-4000 buckets using a Heraeus Multifuge 4KR at 2000 rpm for 15 min with acceleration preset 6 and deceleration preset 1.

Chapter II – Materials and Methods – Methods

5.1.4 Primary human CD4+ T cells and CD19+ B cells

Human PBMCs of healthy volunteers were pre-enriched using BD Vacutainer, CPT. Human CD4+ T cells were purified according to the Miltenyi human CD4+ T cell isolation protocol (CD14+ depletion and CD4+ MicroBeads isolation) and cultured in AIM-V medium (2% human serum, 50 µM 2-mercaptoethanol, 10% FBS, 1% Pen-Strep). Human CD19+ B cells were purified with CD19+ MicroBeads and cultures in AIM-V medium (2% human serum, 50 µM 2-mercaptoethanol, 10% FBS, 1% Pen-Strep). T cell activation was performed using the Human T cell activation/expansion kit according to the manufacturer’s protocol. B cells were activated with 50 ng/ml PMA and 500 ng/ml ionomycin. Blood donations were approved by the review board of Kanton Thurgau, Switzerland.

5.1.5 Mouse embryonic fibroblasts (MEFs)

Preparation of MEFs was performed from mouse embryos on gestation day 14. Head and liver were removed, embryos were minced and digested in trypsin/EDTA solution (Invitrogen) for 15 min at 37°C. Trypsin digestion was stopped with excessive culture medium followed by centrifugation. Cells were passed through a 100 µm filter and cultured for two days at 37°C, 5% CO2. Aliquots were kryo-conserved at -150°C (DMEM, 10% DMSO, 20% FBS).

5.1.6 Ex vivo T cell expansion

For ex vivo expansion 1x106 cells/ml were stimulated with 50 ng/ml Phorbol-12-myristate-13-acetate (PMA) and 500 ng/ml ionomycin for 25 h, followed by cultivation in the presence of 40 U/ml recombinant m-IL-2. IL-2 medium was renewed on day 4. Cells were either kryo-conserved on day 6 (RPMI 1640, 10% DMSO, 20% FBS) or used directly on day 7 for experiments. Kryo-conserved cells were thawed 16-20 h before use and kept overnight in medium containing 30 U/ml IL-2.

5.1.7 In vitro T helper cell polarization

CD4+ T cells were purified from mouse spleens with MACS CD4+ T cell isolation kit. 60.000 to 85.000 cells per 96-well cavity were activated with plate-bound anti-CD3/CD28 antibodies (5 µg/ml in PBS each) in the presence of polarizing conditions towards either Th17 lineage or Treg lineage or with medium only (Th0 conditions, non-polarizing). Polarization towards the Th17 lineage was performed with 30 ng/ml IL-6 (eBioscience), 2.5 ng/ml TGF-β (Peprotech), 10 µg/ml anti-IL4 (eBioscience) and 10 µg/ml anti-IFN-γ (eBioscience) in RPMI 1640 medium. After 3 days, cells were re-stimulated for 4 h with 50 ng/ml PMA and 500 ng/ml ionomycin in the presence of 10 mg/ml brefeldin A (pre-solved in DMSO at ), followed by intracellular staining against Foxp3 and IL-17A (both combinations, IL-17A-FITC with CD4-APC or CD4-FITC with IL-17A-APC were used (section 5.7). Polarization towards the Treg lineage was performed with 100 U/ml IL-2 (x63-mIL-2 derived, section 5.2), 5 ng/ml TGF-β (Peprotech) and anti- 75

Chapter II – Materials and Methods – Methods

IFN-γ (10 µg/ml, eBioscience). On day three, re-activation and intracellular staining were performed as denoted above.

5.2 Determination of IL-2 activity in a CTLL-2 bioassay Recombinant mouse IL-2 was used in several assays as cell culture supplement or for stimulation of cells via the IL-2 receptor. Here, the plasmacytoma cell line X63-m-IL-2 [690] was cultured in IMDM + 10% FBS and supernatants containing secreted recombinant IL-2 were collected and frozen at -80°C for long- term storage or at -20°C for short-term storage (kind gift from Ulrike Beck). The activity of IL-2 in the supernatants was determined via its potency to induce proliferation of CTLL-2 cells measured by MTT 5 assay (Invitrogen, Thermo Fisher Scientific). CTLL-2 cells were seeded at 2 * 10 cells/ml and 100 µl of cells per well (96-well plate) were cultured for 48 h in RPMI 1640 supplemented with 10 % FBS, 50 µM 2-mercaptoethanol, 1 % Pen-Strep, 1 mM sodium pyruvate and a dilution series of IL-2 containing supernatants (100 µl volume of dilution series pre-filled to 96-well cavities). Afterwards, 170 µl of the medium was discarded and 100 µl of DMEM/F12 without phenolred was added. 10 µl of 12 mM MTT stock solution was added as according to the manufacturer’s protocol (Vybrant® MTT Cell Proliferation

Assay) and the plates were incubated at 37°C 7% CO2 for 2-4 h. Then 100 µl of prepared SDS-HCl was added and the plates incubated for another 2-4 h. Up to three measurements of absorbance at 570 nm were performed in a TECAN infinite pro plate reader. The amount of IL-2 containing supernatant resulting in half-maximal proliferative response was determined as the activity of 1 unit. Thus, the activity could be measured in repeated assays using the mean of the determined values. For the IL-2 used in this work, the activity was 15 U/µl.

5.3 Cell viability assessment For cell viability assessment cells were stained using acridine orange and propidium iodide (AO/PI) staining solution in a Cellometer 2000 Auto Cell Counter instrument.

5.4 Inhibitor preparation PRN1126 was provided by Principia Biopharma. ONX 0914 was provided by Christopher J. Kirk (Kezar Life Science). LU-001i [290, 696] was provided by Herman S. Overkleeft. ML604440 [162] was provided by Millenium Pharmaceuticals. MG-132 was purchased from Sigma Aldrich. All compounds were dissolved in DMSO at 10 mM stock concentration and stored at -80°C. Before use, stocks were pre-diluted to 100 µM concentration in DMSO for use in experimental assays (except for MG-132, which was used from 10 mM stocks). DMSO therefore served as solvent-control. Each aliquot was only used after the first thawing of the stock aliquot to avoid repeated freeze-thaw cycles. For in vivo application, ONX 0914 was

Chapter II – Materials and Methods – Methods formulated in an aqueous solution of 10% (w/v) sulfobutylether-β-cyclodextrin (Captisol®) and 10 mM sodium citrate (pH 6). The formulation without ONX 0914 was used as vehicle control. Administration to mice was performed as an s.c. bolus dose of 10 mg/kg.

5.5 Enzyme-linked immunosorbent assay (ELISA) Mouse IL-2 ELISA Ready-Set Go! (eBioscience) was used according to the manufacturer’s protocol. In brief, purified naïve CD4+ T cells were activated by plate-bound anti-CD3/CD28 antibodies in 96-well flat-bottom plates in RPMI 1640 (50 µM 2-mercaptoethanol, 10% FBS, 1% Pen-Strep) for the time period of interest. Supernatants were collected and frozen at -80°C until further use. Upon thawing, samples were appropriately diluted to measure cytokine concentrations in the linear range as indicated by the cytokine standard dilution series. Absorbance was measured in a TECAN plate reader at 450 nm measurement wavelength and 570 nm reference wavelength. For comparison between groups, the mean measured absorbance was plotted in graphs with SD from triplicate experiments. Repeated experiments were analyzed as outlined in section 5.15 and in figure legends.

5.6 CFSE proliferation assay Up to 1x107 cells were stained with 1 µM CFSE in 1 ml PBS for 10 min at 37°C, washed twice with PBS and twice with medium. The initial staining intensity was measured by flow cytometry. Cells were activated with plate-bound antibodies in the presence of ONX 0914 or DMSO. CFSE dilution was measured by flow cytometry after 72 h.

5.7 Flow cytometry Surface staining was performed with antibodies diluted in FACS-buffer (20 min, 4°C) followed by 3 x washing in FACS buffer before analysis (primary antibody-dilution: 1:150 – 1:500). For intracellular staining against p-ERK, t-ERK and IκBα all compared samples were handled simultaneously using a multi-channel pipette to transfer samples between 96-well plates. 120 µl cell suspension from each 96-well cavity was transferred into 70 µl of 4% PFA. After centrifugation for 1.5 min, supernatants were discarded and cells fixed in pure 4% PFA for additional 5 min, followed by 1 x washing in FACS buffer. Upon collection of all time-points, surface staining was performed in FACS buffer for 20 min at 4°C. Afterwards, cells were incubated in 90% of 4°C-cold methanol for 45 min, washed in FACS buffer and primary antibody staining at 4°C in PERM buffer was performed overnight (anti-pERK: 1:800, anti-ERK: 1:600, anti-IкBα: 1:500). The next day, samples were washed 3 x in PERM buffer with ≥15 min incubation time in each step (4°C). Secondary antibody staining was then performed in PERM-buffer for 2 h at 4°C (anti-rabbit Alexa-647 or anti-rabbit Alexa-488: 1:1000), followed by

Chapter II – Materials and Methods – Methods additional 2 x washing in PERM buffer and re-suspension in FACS buffer for analysis. Flow cytometry was performed with FACSFortessa, FACSCalibur or AccuriC6 instruments (BD) analysis was performed using AccuriC6 or FlowJo V10 software. Note that in some experiments upon reproduction of the experiment different flow cytometers had to be used for technical reasons, resulting in different fluorescence intensities based on different devices. In order to subject all relevant data to analysis, but also account for this technical experiment to experiment variation, data was then subjected to normalization as outlined in section 5.15. Intracellular staining against IL-17A and Foxp3 was performed after surface staining against CD4, followed by fixation in Fix/Lysis solution (BD) over night. The next day, samples were stained against IL17A and Foxp3 in PERM buffer for 2-4 h at 4°C, washed and analyzed by flow cytometry using AccuriC6 or FACSCalibur.

5.8 Preparation of 4% PFA solution in PBS

1 L of 4% PFA was produced by heating 100 ml of 10 x PBS with ~700 ml de-ionized H2O under the ventilated hood with constant stirring and addition of 40 g PFA at ~60°C. The pH was slowly raised with

1 M NaOH until PFA was dissolved. Upon solvation the volume was filled up to 1 L with de-ionized H2O and the pH re-adjusted to pH 6.9 with small volumes of HCl solution.

5.9 Generation of cell lysates Whole cell lysates where generated by lysing cells either after harvesting and centrifugation (6000 g, 1.5 min, 4°C) or directly in plates with ice-cold whole cell lysis (WCL) buffer. Insoluble debris was discarded after 15 min fullspeed centrifugation (Eppendorf 5417R, 4°C) and supernatants transferred into new tubes. To preserve post-translational modifications on proteins, the whole lysates were immediately boiled at 95°C for 5 min in SDS sample buffer and then stored at -20°C. Nuclear and cytosolic extracts of proteins were fractionated biochemically as follows: Harvested cells were lysed in cytosolic extraction buffer for 20 – 30 min on ice. Supernatants were transferred into new reaction tubes and used as cytosolic fractions. Pellets were washed 3 times with 4° cold PBS and boiled in WCL buffer mixed with SDS sample buffer for 5 min at 95°C. Cytosolic and nuclear fractions were both kept at -20°C after boiling in SDS sample buffer to preserve post-translational modifications in the lysate. Lamin A/C was used as a nuclear marker.

5.10 SDS-PAGE and Western blotting

Discontinuous sodium dodecyl sulfate polyacrylamide electrophoresis (SDS-PAGE) protein separation was performed using either equal volumes of lysate (normalization on equal cell count in the respective

Chapter II – Materials and Methods – Methods experiment) or after protein concentration assessment using DCTM protein assay (Biorad) including reagent S for detergent-rich samples (normalization on total protein content). Equal volumes/protein amounts were separated by SDS-PAGE in 8–15% separation gels and 5% stacking gels with the following exemplified gel compositions:

5% stacking gel (6 ml) 10% separation gel (15 ml) 12% separation gel (10 ml) 4.1 ml MilliQ-H2O 5.9 ml MilliQ-H2O 4.9 ml MilliQ-H2O 1 ml acrylamide mix 5.0 ml acrylamide mix 6.0 ml acrylamide mix 0.75 ml 1.0 M Tris (pH 6.8) 3.8 ml 1.5 M Tris (pH 8.8) 3.8 ml 1.5 M Tris (pH 8.8) 0.03 ml 20 % SDS 0.075 ml 20 % SDS 0.075 ml 20 % SDS 0.06 ml APS 0.15 ml APS 0.15 ml APS 0.006 ml TEMED 0.009 ml TEMED 0.009 ml TEMED Protein separation was performed in SDS running buffer at constant voltage (55 V during stacking gel, up to 110 V during separation gel). After SDS-PAGE separation, proteins were blotted onto nitrocellulose membranes (GE Healthcare, Protran 0.45 NC nitrocellulose for ECL-based immuno-detection or Protran 0.45 NC Premium nitrocellulose membrane for near-IR based immuno-detection) by wet blot using Mini Trans-Blot® Cell (Biorad) in blotting buffer.

5.10.1 Enhanced chemiluminescence based immuno-detection

For ECL-based detection, membranes were either stained with Ponceau red staining solution before washing and blocking or were immediately blocked with 3% BSA in TBS-T for 1 h at RT and antibodies were diluted in 3% BSA in TBS-T (primary Ab overnight, 4°C, secondary for 1–3 h, RT). HRP-coupled anti-mouse/anti-rabbit secondary antibodies were purchased from Dako. Signal detection was performed using the ChemiDocTM Gel Imaging System (Biorad) with Super Signal West Pico and/or Femto kits (Thermo Scientific). Images were quantified by densitometry and processed to jpg-file format using Quantity One Analysis software (Biorad). PageRuler Prestained Protein Ladder (Thermo Fisher) was used as a molecular weight marker. Marker locations are indicated in immunoblots where necessary.

5.10.2 Near-IR based immuno-detection

Near-infrared detection was performed according to the LI-COR Odyssey® protocol for quantitative Western Blotting (LI-COR). Membranes were kept in clean, ponceau-S free and bromphenol-blue free plastic boxes for all incubation steps. After blotting, membranes were rinsed with H2O and dried at RT for at least 1 h to enhance protein retention. Afterwards, membranes were blocked with Odyssey® Blocking Buffer (TBS) (LI-COR). Secondary antibodies: IRDye800CW goat anti-rabbit or anti-mouse and IRDye680RD goat anti-mouse or anti-rabbit (1:15000). Signals were quantified and images processed to jpg-file format using the LI-COR Odyssey® Imager and Image Studio Lite Vers.5.2. Chameleon Duo Prestained Ladder (LI-COR) was used as a molecular weight marker. Marker locations are indicated in immunoblots where necessary. 79

Chapter II – Materials and Methods – Methods

5.11 Radioactive labeling and pulse-chase experiments

In preparation for radioactive labeling, dialysis tubes (SpectrumTM Labs) were sterilized by autoclave- sterilization in water, filled with sterile FBS and dialyzed 3 x against 5 L PBS overnight. T1 cells were 5 seeded at 0.5 – 1 * 10 /ml and cultured with 200 U/ml recombinant mouse IFN-y (Peprotech) for 3 days to induce higher immunoproteasome expression. Consequently 4-5x106 T1 cells per sample were pre-treated for 2 h with ONX 0914, DMSO or MG-132, followed by activation with plate bound antibodies for 2 h in the presence of MG-132 or DMSO. Medium was then exchanged to cysteine/methionine-free RPMI 1640 (Sigma Aldrich) supplemented with dialyzed FBS and L-glutamine (20 mg/L) for starvation for 1 h followed by a 15 min radioactive pulse using 250 µCi/ml (500 µl per sample) of 35S-labeled cysteine/methionine (Hartmann Analytic, IS-103) added to starvation medium. Cells were washed with normal RPMI 1640 and chased 0 min, 20 min and 40 min after the pulse. Medium was then discarded and cells lysed in the plate in ice-cold whole cell lysis buffer. Total 35S incorporation was assessed using a β-counter and lysate amounts were normalized on β-counts for immunoprecipitation against DUSP6. IP was performed with rabbit anti-DUSP6 (Abcam, ab220811) at 1 µg/sample with 30 µl Protein A beads (EZview Affinity Gel, Sigma) for 5-6 h at 4 °C. After washing (2x NET-TON (50 mM TrisHCL, pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% Triton X-100), 1 x NET-T (NET-TON with 650 nM NaCl)) the precipitate was boiled in 30 µl SDS sample buffer for 5’ at 95°C. 25 μl per lane were loaded onto SDS- gels, which were dried after protein separation and used for autoradiography in a phosphoimager (BioRad).

5.12 RNA extraction and q-RT-PCR

RNA was isolated from -80°C frozen cell pellets using the RNeasy Mini kit (QIAGEN) according to the manufacturer’s protocol. After RNA purity assessment and concentration determination with a NanoVue instrument (GE Healthcare) cDNA was synthesized using Oligo-dT-primers (Promega/Biozym) as follows:

Kit reagent Amount per sample

MgCl2 4 µl 10 x buffer 2 µl dNTP mix 2 µl RNasin RNAse inhibitor 0.5 µl AMV Polymerase 0.6 µl Oligo-dT Primers 1 µl test RNA 0.8 µg

PCR grade H2O Filled 20 µl total volume

For low amounts of RNA from naïve T cells the Sensiscript RT Kit (QIAGEN) was used. Quantitative RT- PCR was performed in a Biometra TProfessional Thermocycler and the Roche LightCycler Instrument (Roche) using the FastStart DNA SYBR green-I kit (Roche). Primers were designed with exon-exon-span 80

Chapter II – Materials and Methods – Methods using PerlPrimer. Amplicon sizes were re-calculated using the Primer-BLAST webtool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) and amplicons were checked via melting curves and/or agarose gels. Rpl13a and Ipo8 were chosen as housekeeping genes because of their reported stable expression in unstimulated and stimulated human T cells [697]. Data was analyzed with the 2-ΔΔCt method [698].

5.13 Mouse genotyping

Tail biopsies were boiled in 600 µl of 50 mM NaOH at 95°C for 45 min to extract crude DNA, followed by addition of 50 µl of 1 M Tris, pH 8. Crude DNA (2 µl per reaction) was used as template in a PCR reaction (25 µl total volume per reaction) using the GoTaq® G2 Flexi DNA Polymerase kit (Promega) with following conditions per 10 reactions:

5x green GoTaq® Flexi Buffer: 100 µl MgCl2 Solution, 25 mM 40 µl PCR Nucleotide Mix, 10 mM each: 10 µl Primer 1, Primer 2, Primer 3: each 5 µl GoTaq® G2 Flexi DNA Polymerase: 2.5 µl

The reaction was performed in a Biometra T3 Thermocycler with “DUSP6_geno” primers for DUSP6- gene targeted mice and reaction cycles as outlined in section 4.5. Amplicons were analyzed in agarose gels (1.5-1.8%).

SMARTA (SM1-Ly5.1) mice were tested via measuring the percentage of Vα2 TCR positive cells in the CD4+ T cell fraction from peripheral blood. Blood was sampled from mouse facial veins and analyzed by flow cytometry using anti-CD4-APC and anti-TCR Vα2-PE antibodies.

5.14 Agarose gel electrophoresis

Agarose gels were poured with 1.5% to 1.8% (w/v) agarose in TAE buffer. Per 50 ml liquid agarose solution 3 µl ethidium bromide was added during gel casting. 18-25 µl PCR products were loaded into gel cavities and run at 100 V (~ 30 min at RT). SmartLadder SF was used as a marker.

5.15 Statistical analysis and graphical data presentation

This section covers a description of reasons for data transformation and normalization, test choice and graphical data representation to assure transparency in statistical analysis. All individual figures indicate which type of graph and which statistical test is presented.

Chapter II – Materials and Methods – Methods

In this study, many experiments were designed to test relative differences between compound-treated groups and control groups in a biological read-out. For example, ONX 0914 treated and MG-132 treated cells are directly compared to DMSO solvent-treated control groups. As primary cells isolated from mice are limited in amount, no external normalization controls were used, but solvent-only controls. When experiments of compound-treated versus control-treated cells were performed repeatedly in independent experiments, pairwise analysis using either paired t tests or repeated measures ANOVA accounts for the fact that samples cannot be seen as fully independent from each other, thus violating test assumptions underlying unpaired tests. To assess within-sample variations, duplicates or triplicates were used and mean values were used for pooled comparison in repeated experiments. Furthermore, the direct comparison of groups within each experiment allows comparing relative effects even if technical reasons result in high non-biological data variation between experiments [699]. On the one hand, in the course of an experiment a priori unexpected complications can occur, that do not perturb the validity of data generated in the single experiment per se, but do not allow for data pooling with other reproductions of the same experimental setup for technical reasons. An example would be that the flow cytometer used in a first experiment to record data has an unexpected technical problem on the day of an independent reproduction for that experiment and therefore the data was recorded on another device. Relative differences of a measured parameter between two test groups can be assumed to still accurately reflect the investigated biological effect while the data spread indicated by pooled data in original measurement units is not reflecting biological variation. Moreover, measurement of an effect in spite of usage of another device also decreases the risk for unaccounted technical artifacts. As all relevant data should be reported and subjected to analysis, the paired study design with evaluation of relative differences thus allows for compliance with this principle in spite of technical variation.

Thus, statistical analysis was performed as follows: Flow cytometric analysis and ELISA measurements were regularly performed in duplicates or triplicates. To represent the within subject variation of individual experiments (and hence the precision of detection in one test) bar graphs with representative examples of quantification results show mean ± SD of duplicate or triplicate values where appropriate. If all experiments were performed on the same instrument with the same settings, the mean values of all independent reproductions expressed in measurement units were used for statistical evaluation of the observed effect. When technical differences between experiments occur while measuring the same effect of interest, normalized data is subjected to analysis. After normalization to the value of the DMSO treated control the biological variation inherent to repeated experiments is only expressed in the variance of the treated group. Hence, a comparison of two group means by one-tailed or two-tailed t-test would be incorrect as the data subjected to analysis is not treatment group data, but the ratio between two groups is the data subjected to analysis. The result is represented as mean difference with 95 % or 99 % confidence interval and significance is then evaluated by assessing whether the CI includes the value of the null 82

Chapter II – Materials and Methods – Methods hypothesis (One-sample t-test). I.e., if there is no influence of a treatment on the parameter compared to the normalization group, the relative difference between two means is 1.0. Therefore, statistical significance is analyzed compared to null hypothesis mean value of 1.0 using the t-statistic, which is appropriately represented by confidence intervals [700].

Immunoblot analysis from cell lysates is typically prone to technical data variation as efficiency in protein lysate yield, blotting efficiency, antibody-batches and room temperature can lead to different signal intensities and signal to-noise-ratios upon detection with the LI-COR Odyssey Imager. Therefore, raw intensity data is normalized to the internal loading control on the same membrane to account for within- experiment variation between lysates. Data is then normalized to the solvent-control sample for comparison of the measured effect accounting for technical experiment-to-experiment variation. As outlined above, mean-differences expressed as confidence intervals are calculated and p-values determined via one-sample t-test [700].

Measurements with more than one group were subjected to multiple comparison corrections, depending on the test, as calculated by the software. All statistical tests were computed using GraphPad Prism software. Tests are indicated in each figure. To most accurately represent data spread in repeated experiments, individual data points are shown. Data point pairs in paired t tests are indicated by connecting lines. Error bars are denoted in figure legends. The calculated p-values are indicated as exact values for highest possible transparency. Asterisks are further used as graphical indications for significance levels appointed by the p-values: *** p ≤ 0.001, ** p ≤ 0.01, * ≤ 0.05. Very low or very high p-values are indicated as p > or p < rounded values.

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition

CHAPTER III: RESULTS

6 ONX 0914 ATTENUATES T CELL ACTIVATION BY

SYNERGISTIC LMP7/LMP2 INHIBITION

Reduced up-regulation of CD69 as a read-out for ameliorated T cell activation after ONX 0914 treatment was originally desrcibed by Michael Basler and Anna Leidinger (unpublished and bachelor’s thesis [701]) and further characterized during my master’s thesis in 2013 [702]. LMP7 inhibition by ONX 0914 in WT T cells led to a similar attenuation of T cell activation as β5c inhibition with PR-825 in LMP7-deficient cells [702]. Moreover, Basler and Mundt et al. found PR-825 to be effective in the amelioration of EAE in LMP7-deficient mice and ONX 0914 impaired EAE in a BM-dependent manner [80]. Thus, it became apparent that disease attenuation was independent of blocking LMP7-specific functions but rather dependent on β5 activity in general, irrespective of the particular subunit at the β5 position. Hence, inhibition of the chymotrypsin-like activity could be the key determinant of disease-ameliorating effects of immunoproteasome inhibition observed in vitro and in vivo. LMP2-deficient cells intrinsically lack chymotrypsin-like activity at the β1 position. Therefore, if the reduction of the chymotrypsin-like activity was the key determinant of ONX 0914 efficacy, then it would be predicted that LMP2-deficient cells might show even stronger amelioration of T cell activation after ONX 0914 treatment. Given that EAE is recognized as a primarily T helper cell dependent disease model, this hypothesis was tested here in CD4+ T cells.

6.1 Activation of primary mouse T cells is ameliorated by ONX 0914 in an LMP7- and LMP2-co-dependent manner

To test the abovementioned hypothesis, CD4+ T cells were isolated from either WT, LMP7-deficient or LMP2-deficient mice and activated with plate-bound antibodies against CD3 and CD28 after a 2 h pulse- treatment with ONX 0914. After 5 h of stimulation, the up-regulation of CD69 was measured by flow cytometry and the secretion of IL-2 was measured in the supernatants by ELISA. In line with previous results in my master’s thesis [702], ONX 0914 treatment led to a marked reduction of both CD69 expression and IL-2 secretion within 5 h in WT, but not in LMP7-deficient T cells (Figure 9A, B). T cells from LMP2-deficient mice were also largely unaffected by ONX 0914 treatment and showed only a minor, although statistically significant reduction of CD69 expression and no apparent reduction of IL-2 secretion (Figure 9A, B). To corroborate that this result was not due to reduced LMP7-incorporation in LMP2- deficient proteasomes, lymphocytes from WT, LMP7-deficient and LMP2-deficient mice were enriched by 84

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition

Ficoll-Paque gradient centrifugation and analyzed for their proteasome subunit composition by immunoblotting. In line with the previously identified rules of cooperative incorporation, mature LMP7 was detected equally in lysates of LMP2-deficient cells and of WT cells. In contrast, LMP2 was markedly less incorporated into LMP7-deficient proteasomes as to be expected due to mutual incorporation interdependency [89]. In cells which were pulse-treated with ONX 0914 or DMSO for 2 h before lysis it became apparent that ONX 0914 treatment resulted in an electrophoretic mobility shift of LMP7 compared to samples from DMSO-treated cells (Figure 9C). Interestingly, in spite of its characterization as an LMP7-selective inhibitor, ONX 0914 also led to a robust electrophoretic mobility shift of LMP2 subunits with only little unmodified LMP2 detectable at the size LMP2 in DMSO-treated cells (Figure 9C, D). Therfore, in contrast to the initial hypothesis, these results indicated that ONX 0914 inhibited LMP7 as well as LMP2 in cells and that the effects of ONX 0914 treatment on T cell activation might be synergistically dependent on LMP7/LMP2 dual inhibition. Notably, in the immunoblots also β5c appeared to be modified by ONX 0914 to some extent, although electrophoretic shifts were less prominent (Figure 9C). Since LMP7-deficient cells, which prominently incorporated β5 instead of LMP7, were not functionally affected by ONX 0914 treatment in spite of some β5 modification, this observation contributed to the view that single subunit inhibition might be insufficient for the effects on T cell activation. The indicated relevance of LMP7/LMP2 dual inhibition was more extensively tested and confirmed both in vitro and in vivo by experiments performed by Michael Basler and colleagues in collaboration with Principia Biopharma, Takeda Pharmaceuticals and Leiden University (Basler et al., 2018, section 12.5). The observed electrophoretic mobility shift was likely attributed to the covalent modification caused by the irreversible inhibitory mechanism and reminiscent of the changed electrophoretic mobility of LMP2 after UK-101 treatment as observed by Wehenkel et al. [287]. However, novel LMP7 and LMP2-selective inhibitors were developed during the time this work was performed. Therefore, eventually the compounds LU-001i (an LMP2-selective inhibitor with epoxyketone warhead, [696]), PRN1126 (a reversible covalent LMP7-selective inhibitor, (Basler et al., 2018, section 12.5) and the boronate LMP2-selective inhibitor ML604440 [162] were tested for their ability to induce electrophoretic shifts as compared to ONX 0914. As expected, only the irreversible covalent inhibitors of the epoxyketone class induced electrophoretic shifts, i.e. ONX 0914 and LU-001i, which confirmed LMP2-selectivity of LU-001i also by immunoblotting. In contrast, PRN1126 and ML604440 did not change electrophoretic mobility of any subunit (Figure 9D). Importantly, proper inhibitory activity at the used concentrations of these compound batches was tested and confirmed elsewhere to control that absence of electrophoretic shifts was not due to compound inactivity (Basler et al., 2018, section 12.5 and [696]).

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition

Figure 9: ONX 0914 treatment leads to attenuated early T cell activation in a synergistically LMP7 and LMP2-dependent manner. A) MACS-enriched CD4+ T cells from WT, LMP7-deficient or LMP2-defient mice were pulse-treated with 0.3 % DMSO or 300 nM ONX 0914 for 2 h before activation with plate-bound anti-CD3/anti-CD28 antibodies for 5 h. Cells were harvested and stained for CD4-APC and CD69-FITC. Median fluorescence intensity of CD69 on CD4+ cells was measured using flow cytometry. Bar graphs on the left show one representative example of an experiment with mean±SD from triplicate measurements. Individual data plots on the right show the ratio of MFI measured for ONX 0914-treated cells over DMSO-treated cells from repeated experiments. Each dot represents one experiment, n ≥ 3. Statistical analysis was performed as one-sample t test for each genotype with the null hypothesis mean µ0 = 1 (i.e.: ratio of 1 = no difference), p-values are indicated in the figure, mean+95% CI. B) Supernatants from cells purified and treated as in A were used for quantification of relative IL-2 secretion by ELISA. Bar graphs on the left show one representative example an experiment plotted as mean±SD from triplicate measurements. Data plots on the right show the ratio of IL-2 secretion for ONX 0914 treated cells over DMSO treated cells. Each dot represents one experiment, n ≥ 3. Mean+95% CI. Statistical analysis as described in A. C) Ficoll-Paque enriched lymphocytes from spleens of WT, LMP7-deficient (L7) or LMP2-deficient (L2) mice were pulse-treated for 2 h with 0.3% DMSO (D) or 300 nM ONX 0914 (X) before cells were lysed and subjected to analysis of proteasome subunits by Western blotting. Tubulin was used as loading control for total lysate. IOTA was used as control for total proteasome content. One representative example of 3 independent experiments (ECL-based immunoblots). D) WT cells prepared as in C were pulse treated for 2 h with 300 nM of the indicated compound or 0.3% DMSO as control. Immunoblots were performed as in C. One example of two independent experiments (ECL-based immunoblots).

6.2 Naïve T and B cells contain almost only mixed and immunoproteasomes Ficoll enriched lymphocytes showed considerably high amounts of immuno- and mixed proteasomes, but markedly less β5c standard proteasome subunits in WT as compared to LMP7-deficient cells. It has been shown before that LMP7 is expressed in murine spleen [135, 136]. Also, T cells stimulated with concanavalin A [89] as well as human naïve B cells express immunoproteasomes [703]. However, the relative content of standard versus immuno- or mixed proteasomes in individual purified naive lymphocyte populations has not been fully addressed at the protein level. Therefore, a detailed characterization of subunit content in naïve T and B cells was performed. MACS-enriched naïve CD4+ T cells and CD19+ B cells were isolated from WT and LMP7-deficient mice (94-97% purity each). Using immunoblot

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition analysis, the relative abundance of all catalytically active standard and immunoproteasome β-subunits was determined. In line with the results obtained from total Ficoll-Paque enriched lymphocytes both CD4+ as well as CD19+ lymphocytes expressed all immuno-β-subunits. Cells lacking LMP7 incorporated β5c instead of LMP7 as expected (Figure 10A, B). Interestingly, while an absolute quantification of protein amount cannot be performed with this method, the comparison between LMP7-deficient and WT cells revealed that the relative abundance of β5c in WT cells compared to LMP7-deficient cells is diminished to hardly detectable levels (Figure 10A, B). In contrast, non-catalytic PMSA6 proteasome subunit (IOTA) signal intensity was comparable between LMP7-deficient and WT cells as well as the total loading control α-Tubulin, indicating that the overall proteasome content of naïve CD4+ and CD19+ lymphocytes is comparable in LMP7-deficient and WT cells. While MECL-1 and LMP2 were also markedly less incorporated into LMP7-deficient proteasomes, the signal intensities for β2c and β1c were only slightly elevated in LMP7-/- CD4+ T cells and not detectably different between WT and LMP7-/- CD19+ B cells (Figure 10A, B). These results indicated that at the naïve state proteasomes in T and B cells normally constitute of full immunoproteasomes as well as of mixed proteasomes (containing LMP7 and additionally β1c and/or β2c to an unknown extent). In contrast, the very low abundance of β5c in T cells and B cells from WT mice indicates that standard proteasomes are barely present in these cell types already at the naïve state. LMP7-deficient T cells and B cells, on the contrary, mostly contained standard proteasomes (Figure 10A, B), rendering primary T cells from WT and LMP7-deficient mice a good model to distinguish immunoproteasome-related effects as opposed to standard proteasome-related effects. Finally, the relative expression of β5c and LMP7 was also analyzed in human CD4+ T cells by comparing the signal intensities to murine CD4+ T cells from WT or LMP7-deficient mice. While such a trans-species comparison might be influenced by different antibody affinities to the mouse or human protein, respectively, it still offers a compromise to assess proteasome subunit abundance in the absence of LMP7- deficent human CD4+ T cells. The results indicated that low β5c and high LMP7 incorporation are a hallmark of naïve human CD4+ T cells as well (Figure 10C).

Figure 10: Naïve T cells and B cells contain predominantly mixed and immunoproteasomes, but not standard proteasomes A) MACS-enriched CD4+ T cells isolated from WT or LMP7-/- mice were lysed and subjected to analysis of proteasome subunits as indicated. IOTA was used as total proteasome control and α-Tubulin / γ-Tubulin were used as general loading control. One out of two independent experiments is shown. B) Experiment as described in A with MACS- enriched CD19+ B cells form WT or LMP7-/- mice. One out of two independent experiments in shown. C) MACS-enriched CD4+ T cells from voluntary healthy donors were lysed and subjected to analysis of proteasome subunit content as indicated. One example out of three independent reproductions is shown. 87

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition

Basal expression of LMP7 was shown previously to be IFN-γ independent in thymocytes, thymic stroma, macrophages and DCs [140, 141]. It was also reported before that immunoproteasomes have a much shorter half-life as compared to standard proteasomes [91]. Hence, it was possible that constant immunoproteasome de novo expression might occur in naïve T cells and B cells and that LMP7-deficient T cells would require a transcriptional up-regulation of the standard subunit β5c in order to compensate for the loss of LMP7. Therefore, it was asked, whether the enhanced β5c protein levels correlated with increased Psmb5 expression. First, using qRT-PCR, CD4+ T cells, CD19+ B cells and bulk liver homogenates were used here to assess relative mRNA expression levels of Psmb8 (LMP7), Psmb5 (β5c), Psmb9 (LMP2) and Psmb6 (β1c) normalized to housekeeping genes in unstimulated naïve cell populations compared to the levels in CD4+ T cells as a reference point. In line with early results by Stohwasser et al. [135], higher mean expression of standard subunits Psmb5 and Psmb6 were detected in liver lysates, albeit not with statistical significance in three independent experiments (Figure 11A). In contrast, expression levels of Psmb8 and Psmb9 appeared to be similar between B cell, T cells and liver homogenates, with only slightly reduced Psmb9 expression in liver and B cells compared to T cells (Figure 11A). Nevertheless, at comparable IOTA signal intensities and comparable β5c protein intensities between LMP7-deficient cells and liver homogenates in immunoblots, markedly reduced LMP7 protein was detectable in liver homogenates (Figure 11B). Notably, liver homogenates also contain blood cells; hence it is unclear, from which cell population LMP7 protein and mRNA are derived in bulk liver homogenates.

To address if β5c was up-regulated in LMP7-deficient cells at the transcriptional level at steady state, mRNA levels in CD4+ T cells, CD19+ B cells and liver homogenates were compared between WT and LMP7-deficient mice. While Psmb8 transcripts were absent from LMP7-deficient cells as expected, no relative difference between Psmb5, Psmb9 or Psmb6 transcripts compared to WT cells was observed for any of the analyzed mRNA extracts (primer pair and qPCR product quality controls for Psmb5 detection are provided in Appendix Figure 1). Merely, a non-significant minor tendency of enhanced standard subunit expression in CD4+ T cells was observed (Figure 11C). These results did not indicate that enhanced β5c protein incorporation in LMP7-deficient lymphocytes was a result of enhanced Psmb5 expression at steady state.

6.3 Impaired proliferation and impaired polarization of CD4+ T cells in vitro

As reported in my master’s thesis in 2013, CD4+ T cell from WT mice, but not from LMP7-deficient mice, showed impaired proliferation in the presence of ONX 0914 [702]. Given that ONX 0914 treatment reduced IL-2 secretion readily after activation and as IL-2 is a potent inducer of T cell proliferation, the question whether deprivation of autocrine IL-2 could be responsible for the impaired proliferation remained to be addressed. To investigate this, recombinant IL-2 was supplemented to the culture during

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition

Figure 11: No compensatory up-regulation of Psmb5 and Psmb6 at mRNA level in LMP7-deficient lymphocytes at steady state. A) MACS enriched CD4+ T cells, CD19+ B cells and bulk liver lysates were used to assess relative mRNA levels for the indicated genes via qRT- PCR. Each result was internally normalized to the Rpl13a and IPO8 from the same cDNA and data is presented as relative difference compared to CD4+ T cells (set as 1). For analysis, ratio-paired t test was performed to test expression levels compared to CD4+ T cells. B) MACS-enriched CD4+ T cells, CD19+ B cells and bulk liver homogenates from either WT or LMP7-deficient mice were used to assess indicated proteasome subunits at steady state by immunoblotting. Signals for IOTA were used as a loading control. One example out of three independent experiments with similar outcome. C) MACS-enriched CD4+ T cells, CD19+ B cells and bulk liver homogenates from either WT or LMP7-deficient mice were used to assess relative expression of indicated transcripts compared to WT expression levels for each cell type. Two-tailed ratio-paired t test, p-values as indicated, n= 3.

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition proliferation. However, this could not restore the proliferative capacity of CD4+ T cells in the presence of ONX 0914 (Figure 12A). The supplemented recombinant IL-2 was instead potent in inducing proliferation of CTLL-2 cells in the MTT-assay (section 5.2, data not shown) and induced STAT5- and ERK phosphorylation in expanded T cells, which was not directly affected by ONX 0914 treatment (Figure 12B). The recombinant IL-2 was also used as a supplement of a cytokine cocktail for in vitro polarization towards the regulatory T cell lineage. Kalim et al. had tested Th17 polarization and Treg polarization under different ONX 0914 treatment conditions of continuous treatment or pulse-treatment, respectively [288]. The overall effect of ONX 0914 on proliferation prompted us to test also Treg polarization under the same continuous treatment conditions like used for Th17 polarization. When CD4+ T cells were cultured under Treg polarising conditions in the presence of 200 nM ONX 0914, a reduced frequency of Foxp3+ CD4+ T cells was obtained after 3 days (Figure 13A, B), which is different from the observed promoting effect on Treg polarization of pulse-treatment with ONX 0914 as reported by Kalim et al. [288]. A reduction in Foxp+ CD4+ T cells after ONX 0914 treatment was also observed under non-skewing conditions that were used as polarization control (Th0, Figure 13B). Polarization of naïve CD4+ T cells towards the Th17 lineage was consistently reduced in the presence of 200 nM ONX 0914, albeit considerable experiment to experiment variation in the frequencies of IL-17A+ cells was observed (Figure 13D, E). Proliferation under non-skewing conditions (Th0, Figure 13C) was used as a control. These results are in line with the results by Muchamuel et al. [199] and Kalim et al. [288], except that no enhanced Foxp3+ Treg polarization in the presence of ONX 0914 was observed under Th17 polarizing conditions (Figure 13E). Thus, both bulk proliferation as well as T cell polarization under Th17 or Treg skewing conditions were impaired in the presence of ONX 0914 treatment.

Figure 12: Impaired proliferation in the presence of ONX 0914 is not caused by IL-2 deprivation alone. A) CFSE-dilution profile of primary murine MACS-enriched CD4+ T cells from WT mice that were activated with plate-bound anti- CD3/CD28 antibodies for 72 h in the presence or absence of ONX 0914 either with (right) or without (left) supplementation of recombinant murine IL-2. CFSE dilution was measured by flow cytometry. One example of two independent experiments. B) Expanded murine CD4+ T cells were pulse-treated with 0.3% DMSO or 300 nM ONX 0914 for 2 h before stimulation with 100 U/ml recombinant mouse IL-2 for indicated times. Immunoblots against phospho-STAT5 and phosphor-ERK with Tubulin as loading control. One example of two independent experiments.

Chapter III – Results – ONX 0914 attenuates T cell activation by synergistic LMP7/LMP2 inhibition

Figure 13: Impaired polarization of CD4+ T cells in the presence of ONX 0914. Naïve CD4+ T cells isolated from WT splenocytes were activated in vitro with plate-bound anti-CD3/CD28 antibodies under Treg polarization conditions (100 U/ml IL-2, 10 µg/ml anti-IFN-γ, 5 ng/ml TGF-β) (A, B) or Th17 polarizing conditions (2.5 ng/ml TGF-β, 30 ng/ml IL-6, 10µg/ml anti-IFN-γ and 10 µg/ml anti-IL-4) (C-E) for 3 d. On day three cells were re-stimulated with PMA/ionomycin in the presence of BFA for 4 h before intracellular staining for cytokines or transcription factors as indicated. A) Representative flow cytometry plots after cultivation under Treg polarizing conditions. Upper panel: Percentage of CD4+ in live and single cell gates (not shown). Lower panel: Foxp3 and CD4 co-staining to assess percentage of Foxp3+ CD4+ cells. Left: CD4+ only (FMO-control). Middle and right: Stained samples under polarization conditions in the presence of DMSO or ONX 0914 as indicated. Examples are shown from one out of four independent experiments. B) Pairwise analysis of mean percentages of Foxp3+CD4+ cells from four similar experiments as performed in A including the percentages of Foxp+ CD4+ cells from the Th0 control polarization (non-skewing conditions). Ratio-paired t test, p-value as indicated. C) Representative flow cytometry plots after polarization under Th0 conditions (without polarizing cytokines/antibodies). Shown are CD4+ only staining (FMO-control) and intracellular staining for IL-17A+ and Foxp3+ cells to asses their frequencies under non-polarizing conditions. D) Representative flow cytometry plots after polarization under Th17 polarizing conditions. Upper panel: CD4+ gates as described in A. Lower panel: Samples stained for intracellular IL-17A and Foxp3 on CD4+ cells after polarization in presence of ONX 0914 or DMSO as indicated. One example of five similar experiments is shown. E) Pairwise analysis of IL-17A+ and Foxp3+ positive fractions of CD4+ cells as the mean percentages from triplicates out of five independent experiments as described in D. Ratio-paired t test, p values as indicated. 91

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

7 ONX 0914 IMPAIRS ERK SIGNALING AND INDUCES MILD

PROTEOSTASIS STRESS IN T CELLS

The general impairment of T cell proliferation and polarization after continuous ONX 0914 treatment prompted us to refocus on the molecular events during the initial activation phase in order to obtain further mechanistic insight into the effects of ONX 0914 treatment. The observation that ONX 0914 treatment resulted in attenuated T cell activation already within the first 5 h of activation led to the question, whether ONX 0914 treatment might influence the degradation or processing of a factor involved in TCR-mediated signal transduction. This question was partially addressed also during my master’s thesis in 2013 [702]. Notably, several tested human and murine T cell lines showed markedly different responses to ONX 0914 treatment as compared to primary cells isolated from murine spleen with only the T1 cell line showing partially the same effects as compared to primary cells [702]. Therefore, it was further investigated whether or not evidence for altered signal transduction could be found directly in primary naïve or expanded T cells.

7.1 ONX 0914 reduces IL-2 and CD69 expression at the mRNA level

To further test the hypothesis of altered TCR-induced signaling in T cells it was investigated whether the effects of reduced CD69-up-regulation and IL-2 secretion (Figure 9) were altered at the transcriptional level or could potentially be only mediated by post-transcriptional mechanisms like an altered secretory pathway capacity of the cells after ONX 0914 treatment. Isolated naïve CD4+ T cells were pre-treated with ONX 0914 or DMSO for 2 h before activation with plate bound antibodies against CD3 and CD28. Immediately after pulse-treatment or 4 h after activation the cells were harvested and mRNA was extracted for retro-transcription to cDNA and quantitative real-time PCR analysis of CD69 and IL2 gene expression. Relative up-regulation of the expression levels compared to unstimulated controls showed considerable variation between the individual experiments (potentially owing to overall low mRNA yield from a limited amount of naïve T cells), but a consistent trend of reduction after ONX 0914 treatment was observed (Figure 14A and B). Quantitative RT-PCR products were checked for correct amplicon sizes and induction of gene transcription using agarose gels (Figure 14C) or primer melting curve assessment (not shown). Although a statistically significant reduction was only observed for the CD69 gene (Figure 14A), the results indicated that CD69 up-regulation and IL-2 secretion were likely already affected by ONX 0914 at the pre-transcriptional level; hence likely involved an influence on TCR-induced activation signaling.

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

Figure 14: Reduced expression of CD69 and IL2 mRNA transcripts after ONX 0914 treatment in primary activated CD4+ T cells. A and B) MACS-enriched naïve CD4+ T cells from WT mice were pre-treated with 0.3 % DMSO or 300 nM ONX 0914 for 2 h before activation with plate-bound anti-CD3/CD28 antibodies for 4 h. Total RNA was extracted and used for q-PCR after retrotanscription to cDNA. CD69 (A) and IL2 (B) gene expression relative to unstimulated DMSO controls are shown as matched pairs from four independent experiments each. Rpl13a was used as housekeeping gene for normalization. 2-ΔCt method, p-values as indicated, paired t-test. C) Amplicon products after q-PCR for indicated gene transcripts were checked via size assessment using agarose gels. RNA extracts from unstimulated cells and from cells after anti-CD3/CD28 stimulation were used to control for TCR-driven induction of CD69 and IL2 gene transcription, while Rpl13a was used as a housekeeping gene for normalization.

7.2 No influence of ONX 0914 on early canonical T cell activation pathways

To investigate into signaling pathways after TCR stimulation in primary T cells, ex vivo expanded T cells were used as an experimental setup [500, 702]. In brief, isolated CD4+ T cells were activated with PMA/ionomycin for 25 h and cultured in IL-2 containing medium for one week before re-activation with plate-bound anti-CD3/CD28 antibodies for analysis of signaling pathways by immunoblotting. LMP7 and β5c content were similar to naïve T cells (Figure 15A). Upon re-activation, expanded T cells up-regulated CD69 and showed reduced CD69 up-regulation after ONX 0914 treatment (Figure 15B), confirming previous observations [702]. Using 2 h pulse-treated expanded T cells several canonical signaling pathways were analyzed that are activated after stimulation by anti-CD3/CD28 antibodies. These included phosphorylation of MAPK p38, Akt(Thr308), Akt(Ser473), and ribosomal protein S6 (Figure 16A), in line with previous results from previous experiments, in which the focus had been set mainly on earlier time points (5 – 30 min) [702]. None of these investigated pathways were detectably influenced by ONX 0914 treatment. Translocation of NFAT1 into the nucleus was assessed by biochemical cellular fractionation after activation followed by immunoblotting against nuclear NFAT1. No influence on nuclear translocation was detected after ONX 0914 treatment (Figure 16B). Notably, NFAT1 was absent in nuclear lysates of unstimulated cells. As the proteasome is well known for its involvement in NF-κB signaling, components of the NF-κB signaling pathway were investigated at the level of IκBα-degradation, p65-phosphorylation as well as nuclear translocation of NF-κB p65. Immunoblot analysis in either antibody-stimulated or

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

Figure 15: Expanded T cells as experimental system for analysis of signaling pathways A) Comparison of β5c, LMP7 and β1c subunit composition in expanded and naïve T cells from WT or LMP7- deficeint mice by immunoblotting (ECL- based). One example of more than two similar experiments. B) CD69 up-regulation on WT expanded CD4+ T cells after 2 h pulse-treatment with DMSO or ONX 0914. One representative histogram and quantification of median fluorescence intensity showing mean±SD. The 95% CI in the right shows pooled data from four independent experiments (included are data from my master’s thesis [702] and this work). One sample t test, p-value as indicated. µ0=1

PMA/ionomycin-stimulated cells as well as flow cytometric measurement of IκBα degradation in primary naïve CD4+ T cells showed that ONX 0914 had no influence on any of the investigated NF-κB signaling components (Figure 16B, E-G).

Immunoblot analysis of JNK phosphorylation showed some experiment to experiment variation with two different available clones of anti-p-JNK antibody that were used (8E11 rabbit mononoclonal or G9 mouse monoclonal), but a transient peak of JNK phosphorylation at 30 min after activation and no marked differences between ONX 0914 or DMSO treated samples up to 2 h after activation were consistently observed (Figure 16C). As enhanced JNK activity can be a result of cellular stress, the analysis was extended to 5 h and additional MG-132 treatment as a control. At 3 h to 5 h after activation a slight enhancement of JNK phosphorylation was detectable in expanded T cells in both ONX 0914 and MG-132 treated cells (Figure 16D). However, in primary naïve T cells JNK phosphorylation after activation was not reliable detectable above background levels (data not shown). Previous studies reported that JNK signaling is less prevalent in primary naïve T cell activation than in effector T cells and not essential for IL-2 production (section 2.3.5). Therefore, further investigation of JNK was not pursued in this work, but focus was set to another MAPK signaling pathway, namely ERK1/2 signaling. In expanded T cells, indications pointing to an influence of ONX 0914 treatment on TCR- and CD28-induced ERK signaling were observed. At 3 h but not already within 1 h of activation a reduction of ERK phosphorylation after ONX 0914 treatment compared to DMSO treatment was observed in ECL-based immunoblots by densitometry quantification (Figure 16H). Moreover, this reduction was not detected in LMP7-deficient cells, highlighting sustained ERK signaling as a potential candidate pathway that might be selectively affected by immunoproteasome inhibition.

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

Figure 16: Analysis of canonical T cell activation signaling pathways in expanded and naïve CD4+ T cells. A) Expanded CD4+ T cells were pulse-treated for 2 h with ONX 0914 or DMSO and stimulated with plate-bound anti-CD3/CD28 antibodies for indicated time periods. Phospho-specific immunoblots against indicated target proteins were performed. One example of at least three experiments. B) Upper panel: Cells treated as in A were analysed for p65 phosphorylation and IκBα degradation. Lower panel: Nuclear extracts from cells treated as in A were analyzed for translocation of NFAT1 and NF-κB p65. Lamin A/C served as a loading control. One example of at least three experiments. C) Activated expanded T cell lysates were analyzed for JNK-phosphorylation. The upper panel shows one experiment using the 8E11 rabbit monoclonal antibody. The lower panel shows another experiment using the G9 mouse monoclonal antibody. D) Cells treated as in A, but with additional MG-132 control treated cells were analyzed for JNK phosphorylation (G9 antibody). One example of two independent experiments. Total JNK was used as a loading control. D = 0.3 % DMSO, X = 300 nM ONX 0914, MG = 10 µM MG-132 E-F) Degradation of IκBα measured by flow cytometry (median FI on CD4+ T cells for naïve primary T cells activated with antibodies (F) or PMA/ionomycin (G) and with additional MG-132 treated cells as control (G). G) Experiment as in B (upper panel) in cells stimulated with PMA/ionomycin. (E-G are individual experiments showing the same effect, hence considered as relevant data that has to be reported. Note, however, that they were not reproduced with identical conditions) H) Densitometric quantification of p-ERK signals normalized to α-Tubulin in either WT or LMP7-deficient cells treated as in A. Data is shown as relative to highest detected signal of pooled means ±SD from two (LMP7-/-) or three (WT) experiments. Two way ANOVA, p-value as indicated, n.d. = not determined. B and H: Vertical grey lines indicate signals from the same membrane and detection, but not originally juxtaposed. 95

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

7.3 Analysis of ERK phosphorylation sustainment by quantitative near-IR immunoblotting and intracellular flow cytometry

ECL-based immunoblotting is limited in its ability to reliably detect small quantitative differences [704]. Therefore, the LI-COR Odyssey Imaging system was implemented and used to re-investigate phospho- ERK signaling by near-infrared-dye-conjugated-secondary-antibody-based immunoblotting, which allows quantitative signal intensity measurements over several orders of magnitude. Using this method, the relative phospho-ERK signal intensity in ONX 0914 treated cells was analyzed by normalization to γ-Tubulin loading control signals from same signal acquisition. In five independent experiments, p-ERK levels were found to be reduced by 14.8% (±7.4%) in ONX 0914 treated and even more in MG-132 treated expanded T cells (Figure 17A,B). The detected phosphorylation sites Thr202/Tyr204 in the activation loop of ERK are primarily phosphorylated by MEK1 and MEK2. Therefore, Ser221 phosphorylation of MEK1/2, which is an activating phosphorylation in MEK1/2 [545] was also investigated in the same lysates. In contrast to ERK, no difference in MEK-phosphorylation intensity was detected in ONX 0914 treated cells, while MG-132 treated cells showed a significant reduction in MEK-phosphorylation compared to DMSO treated cells as well (Figure 17A, B). These results indicated that reduced ERK signaling after ONX 0914 treatment was likely not caused by an effect on the upstream signaling cascade, but rather via a direct negative regulation of ERK. As the observed difference in ERK-phosphorylation between ONX 0914 treated and DMSO treated cells were only of ~15% difference and also quantitative near-IR blotting showed considerable experiment-to-experiment variation, a third independent method was established to corroborate the validity of the observed effect. Therefore, primary naïve T cells were pulse- treated for 2 h with ONX 0914 or DMSO and activated for 3 h with anti-CD3/CD28 antibodies followed by intracellular staining for p-ERK and flow cytometry analysis at single cell resolution. In line with the results obtained by immunoblotting, phosphorylated ERK levels were found significantly reduced after ONX 0914 treatment in CD4+ T cells derived from WT mice (21.9% ±4.3%, Figure 17C, D). Notably, CD4+ T cells from LMP7-deficient mice did not show a significant reduction of ERK-phosphorylation after ONX 0914 treatment (3.1% ±3.1%, Figure 17C, D). Furthermore, a kinetic assessment of p-ERK levels in flow cytometry supported that sustained, but not early, ERK signaling was affected by ONX 0914 (Appendix Figure 3). Staining against total ERK revealed that total ERK levels were not altered in ONX 0914 treated compared to DMSO treated T cells from WT and LMP7-deficient mice (Figure 17E). Finally, primary human CD4+ T cells were isolated from PBMCs of healthy voluntary donors using CD14 depletion and CD4 microbeads magnetic sorting. Activation of these cells with anti-CD3/CD28/CD2 coated beads after pulse-treatment with ONX 0914 resulted in reduced up-regulation of CD69 (Figure 17F) as well as in reduced ERK-phosphorylation levels at 3 h and 5 h after stimulation, while total ERK levels were not affected (Figure 17G). MG-132 treatment even further reduced CD69 up-regulation,

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

Figure 17: ONX 0914 treatment reduces ERK signaling sustainment in expanded and naïve murine T cells and in primary human T cells. A) Expanded CD4+ T cells pre-treated with DMSO (D), 300 nM ONX 0914 (X) or continously with MG-132 (MG) were activated with plate- bound antibodies for 3 h or left unstimulated. Cells were lysed and used for near-IR immunoblotting against indicated proteins. Representative example from five experiments. B) Intensities of p-ERK and p-MEK relative to tubulin loading control at 3 h were quantified in five independent experiments as shown in A. Graphs show mean+95% CI. Ratios of ONX 0914 treated / DMSO treated signals were analyzed with one-sample t-test, µ0 = 1, p-values as indicated in the figure. C) MACS-enriched splenic CD4+ T cells from WT or LMP7-/- mice were 2 h pulse-treated with DMSO or ONX 0914 and activated for 3 h with plate-bound anti-CD3/CD28 antibodies or left unstimulated. Intracellular p-ERK1/2 levels were measured using flow cytometry. Histograms show one representative example of five independent experiments. One example of gating strategy and secondary-antibody-only controls is shown in Appendix Figure 2. D) Phospho-ERK+ median fluorescence intensity ratios of ONX 0914-treated/DMSO-treated cells from five independent experiments are shown as mean+95% CI. One-sample t-test with µ0 = 1, p-values as indicated. E) Experiment as in C, but intracellular staining was performed against total ERK. One example for two independent experiments. F) MACS-enriched human CD4+ T cells were pulse-treated with DMSO or ONX 0914 or continuously treated with MG-132 and activated with stimulating beads for 5 h. CD69 expression on CD4+ cells was measured by flow cytometry and median fluorescence intensities used for quantification. Representative histogram (left) and quantification with pooled data of three independent experiments (right) are shown (mean±SD); two-way repeated measures ANOVA, Sidak's post test, p-values as indicated. G) Cells treated as in F were lysed and used for near-IR detection of p-ERK1/2 intensities normalized to γ-Tubulin after 3 h and 5 h of activation. Top panel: Representative immunoblot, graph below: Ratio of ONX 0914 treated/DMSO treated signals from three independent experiments (mean with 95%CI) was analyzed with one-sample t-test, µ0 = 1, p-values are indicated.

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells but affected p-ERK signaling to a lesser extent than ONX 0914 treatment (Figure 17G). Taken together, these results confirmed that immunoproteasome inhibition reduced activation-induced ERK-signaling sustainment in primary human and murine CD4+ T cells.

7.4 ONX 0914 causes mild protein homeostasis stress in activated T cells without induction of apoptosis

Impaired proteasome capacity due to proteasome inhibition can result in proteostasis stress marked by accumulation of ubiquitin-conjugates. Muchamuel et al. tested the effect of ONX 0914 (formerly called PR-957) treatment on the accumulation of ubiquitin-conjugates in the human T cell derived cell line Molt4 [199]. Within the reportedly LMP7-selective concentration range ONX 0914 did not induce ubiquitin- conjugate accumulation and the same result was obtained in this work upon independent reproduction of the experiment as described by Muchamuel et al. [199] (Figure 18A). Due to the high relative abundance of LMP7-containing proteasomes compared to β5c-containing proteasomes in primary lymphocytes (section 6.2), it was re-addressed whether partial proteasome inhibition with ONX 0914 was sufficient to induce ubiquitin-conjugate accumulation in primary T cells during activation. When expanded CD4+ T cells were activated after pulse-treatment with 0.3% DMSO or 300 nM ONX 0914, an accumulation of ubiquitin-conjugates was detectable in ONX 0914 treated WT cells within 1 h of stimulation and more pronounced after 4 h (Figure 18B). Notably, this effect was LMP7-dependent as expanded cells from LMP7-deficient mice did not show an accumulation of ubiquitin-conjugates after ONX 0914 treatment (Figure 18B). MG-132 treatment induced a marked accumulation of ubiquitin-conjugates already within 30 min and affected both WT as well as LMP7-deficient cells (Figure 18B). MG-132 treatment also induced markers of the integrated stress-response within 3 h of stimulation in expanded T cells as detected by p53 accumulation, phosphorylation of eIF2α and ATF4 induction (Figure 18C). In contrast, no difference with respect to integrated stress response markers was detected between DMSO treated and ONX 0914 treated cells. An accumulation of ubiquitin-conjugates was similarly detected in freshly isolated naïve CD4+ T cells after pulse-treatment with ONX 0914 or continuous treatment with MG-132 in WT cells, while LMP7-defienct cells were only affected by MG-132 treatment and not by ONX 0914 (Figure 18D). Moreover, while MG-132 treatment induced PARP cleavage within 5 h in both LMP7-deficient and WT cells, ONX 0914 treatment did not (Figure 18D). Interestingly, if cells were left unstimulated for 5 to 6 h after ONX 0914 treatment ubiquitin-conjugate accumulation was less pronounced indicating that the bulk of ubiquitinated protein accumulation in ONX 0914 treated cells originated from T cell activation and was not due to an effect of immunoproteasome inhibition on steady-state protein homeostasis only (Figure 18E, F). LMP7-deficient cells showed no enhanced ubiquitin-conjugate formation compared to WT cells irrespective of stimulation or ONX 0914 treatment (Figure 18F). Finally, primary human CD4+

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

Figure 18: ONX 0914 treatment induces mild proteostasis stress in activated CD4+ T cells without inducing apoptosis A) Effect of ONX 0914 treatment on ubiquitin conjugate accumulation in Molt 4 cells. Experiment performed as detailed in Muchamuel et al. 2009, Nature Medicine, 15;7, 781-7, Supplementary Figure S1A. One example of two experiments (ECL-based immunoblot). B) Expanded CD4+ T cells from WT or LMP7-deficient mice were pulse-treated for 2 h with 0.3% DMSO (D) or 300 nM ONX 0914 (X) or continuously treated with 10 µM MG-132 (MG) and consequently activated with plate-bound anti-CD3/CD28 antibodies for the indicated times. Immunoblot analysis of ubiquitin conjugates with α-Tubulin as loading control is shown. A representative example of at least three experiments with similar outcome is displayed (ECL-based immunoblot). C) WT cells treated as in B were activated with stimulating antibodies and subjected to near-IR immunoblot analysis of stress response markers. ERK phosphorylation with normalized signal intensity is shown as activation control; γ-Tubulin was used as a loading control. One example of at least three experiments is shown. D) Primary mouse CD4+ T cells isolated from spleens of WT or LMP7-/- mice were treated and analyzed as in B, but with near-IR based immunoblots. One example of at least three experiments with similar outcome. E,F) Primary cells as in D were pulse-treated for 2 h with 0.3% DMSO (D) or 300nM ONX 0914 (X) and either stimulated with plate bound antibodies or left unstimulated for indicated time periods before subjection to ECL-based immunoblotting against indicated targets. Both are examples of more than three similar experiments with comparable outcome. G) Near-IR-based immunoblot analysis of primary human CD4+ T cells from healthy human donors, treated and analyzed as in B for indicated time periods. One example out of three independent experiments is shown.

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

T cells were analyzed for ubiquitin-conjugate accumulation after ONX 0914 treatment and activation with bead-coated antibodies against CD3/CD28/CD2. As observed in murine T cells, no enhanced or only minute PARP cleavage was detected after ONX 0914 treatment compared to DMSO as opposed to MG-132 treatment in human T cells (Figure 18G). However, a slightly enhanced phosphorylation of eIF2α was observed after ONX 0914 treatment as compared to DMSO treated cells. Taken together, these results emphasize that in primary mouse and human CD4+ T cells, in contrast to the previously tested Molt4 cell line, a mild form of proteostasis stress is induced by ONX 0914 treatment. However, ONX 0914 induced no or only weak signs of the integrated stress response and induced no apparent enhancement of apoptosis as measured by PARP cleavage.

7.5 CD4+ T cells overcome proteostasis stress likely via Nrf1-mediated standard proteasome up-regulation

Prolonged proteostasis stress as inducible with proteasome inhibitors like MG-132 eventually results in apoptotic cell death. As ONX 0914 treatment did not result in enhanced PARP cleavage in primary CD4+ T cells within 3 to 5 h after TCR-mediated activation, it was investigated how ONX 0914 treatment affects cell function and viability after prolonged activation. When naïve CD4+ T cells from WT or LMP7- deficient mice were pulse-treated with ONX 0914 or DMSO for 2 h followed by activation with plate- bound anti-CD3/CD28 antibodies for 20 h, the observed ubiquitin-conjugates detectable after 3 h and 6 h of activation appeared to be alleviated in ONX 0914 treated samples (Figure 19A). Moreover, assessment of cell viability as measured by AO/PI staining in a Cellometer 2000 cell counter showed that LMP7- deficient as well as WT cells and both either treated with ONX 0914 or DMSO contained comparable percentages of viable cells after 20 h (Figure 19B). Notably, the alleviation of ubiquitin-conjugates and the preserved cell viability could not be attributed to de novo expression of LMP7 containing proteasomes, as all detectable LMP7 protein remained electrophoretically shifted even 20 h after ONX 0914 pulse- treatment (Figure 19C). In contrast, de novo expressed LMP7 protein would not be modified by ONX 0914, which was not present in the medium during the period of activation. On the contrary, enhanced protein levels of the standard β5c subunit were detectable after 20 h, suggesting that a compensatory up- regulation of standard proteasome content might be induced upon immunoproteasome inhibition (Figure 19C). Given that context-specific EGF-receptor mediated STAT3-signaling, but not IL-6 induced STAT3- sginaling was reported to regulate standard and/or immunoproteasome expression [705, 706], it was assessed if autonomous STAT3 activation occurred during T cell activation in vitro and if alterations in autonomous STAT3-phosphorylation could be observed after ONX 0914 treatment. However, while STAT3 phosphorylation at Tyr705 was detected after in vitro activation with anti-CD3/CD28 antibodies at around 3 – 5 h, the results remained ambiguous. For example, some immunoblots showed slightly

100

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells enhanced STAT3-phosphorylation at 3 h after activation in ONX 0914 treated cells, but no alterations at 5 h after activation (Appendix Figure 4A), and some immunoblots showed enhanced p-STAT3 after 5 h of activation, but not after 3 h (Appendix Figure 4B). Likewise, some experiments indicated different signal levels for STAT3 at the level of both, total STAT3 and phospho-STAT3 detracting from the reliability of the results (Appendix Figure 4C). In contrast, in lysates from human CD4+ T cells activated with CD3/CD28/CD2-coated beads a robust induction of autonomous STAT3-phosphorylation at 3 h and 5 h after activation was observed. However, no differences in p-STAT3 levels between ONX 0914 treated and DMSO treated cells were detected as opposed to MG-132 treatment, which impaired autonomous STAT3-phosphorylation (Appendix Figure 4D). Therefore, characterization of β5c up-regulation was instead focused on the possible involvement of another reported Psmb5 inducer, namely Nrf1 (section 1.6).

To further assess whether standard proteasomes were up-regulated after ONX 0914 treatment in activated T cells, freshly isolated CD4+ T cells from WT mice were pulse-treated with DMSO or ONX 0914 or continuously treated with MG-132 before activation with plate-bound antibodies for 5 h and 9 h. Quantitative assessment of LMP7 and β5c protein levels by near-IR immunoblotting revealed that T cell activation induced slightly enhanced levels of mature β5c after 5 h and markedly after 9 h of activation (Figure 19D, E). In contrast, LMP7 was not induced over unstimulated control levels (Figure 19D, E). ONX 0914 treatment significantly boosted β5c up-regulation as compared to DMSO treated cells, while MG-132 treatment abrogated β5c up-regulation during activation (Figure 19E). Notably, also 9 h after activation, consistent results compared to the before mentioned experiments were obtained for PARP cleavage, phosphorylation of eIF2α and ATF4 induction, which were not induced by ONX 0914 treatment, but markedly induced by MG-132 treatment. Correlating with the enhanced β5c protein levels, soluble Nrf1 accumulation was detected in ONX 0914 treated cells (Figure 19D). In contrast, MG-132 treated cells did not show soluble Nrf1 accumulation (Figure 19D), which might be due to Nrf1 insolubility and aggresome formation at high degrees of proteasome inhibition [223, 224]. Full length Nrf1 has a reported molecular weight of 120 kDa, while different reports regarding the molecular weight of cleaved Nrf1 exist ranging from 90 to 110 kDa [222–224]. Hence, it is not unambiguously clear, which Nrf1 form is detected here as only one signal between 125 kDa and 90 kDa marker signals appears. Most likely the detected signal corresponds to the de-glycosylated cleaved form which was shown to accumulate in response to epoxomycin [222]. Nrf2 is also controlled by proteasomal degradation and promotes standard proteasome subunit expression (section 1.6). However, no reliable signals for Nrf2 could be obtained in the immunoblots (data not shown). Finally, to test if the observed β5c up-regulation contributed to preserving cell viability, PR-825 (which was reported as a β5c-selective inhibitor [199]) was added to a final concentration of 100 nM after 4 h of activation when the cells had been pulse-treated with ONX 0914 or DMSO for 2 h before stimulation. After a total of 20 h after activation cells were harvested and analyzed for PARP cleavage. Indeed, enhanced PARP cleavage was observed in cells pulse-treated with ONX 0914 101

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells and subsequently PR-825 treated indicating that β5c activity was necessary to prevent apoptosis induction after immunoproteasome inhibition (Figure 19F). Taken together, these results provided evidence that mild non-apoptotic proteostasis stress induced by ONX 0914 treatment in primary T cells could be overcome via standard proteasome up-regulation, but not immunoproteasome up-regulation, likely via Nrf1.

Figure 19: T cells up-regulate β5c containing standard proteasomes during activation and in response to immunoproteasome inhibition, likely via Nrf1 A) Naive T cells isolated from either WT mice were pulse-treated for 2 h with 0.3% DMSO (D) or 300 nM ONX 0914 (X) and activated with plate-bound antibodies against CD3/CD28 for 3 h, 6 h or 20 h. Immunoblot analysis against ubiquitin-conjugates is shown with -tubulin as loading control. One example of at least three experiments with similar outcome (ECL-based immunoblot). B) Naive T cells isolated from either WT or LMP7-deficient mice were pulse-treated for 2 h with 0.3% DMSO (D) or 300 nM ONX 0914 (X) and activated with plate-bound antibodies against CD3/CD28 for 20 h. Cells were harvested and cell viability was assessed using AO/PI staining in a Cellometer 2000. Pooled data from five independent experiments is shown as mean±SD. C) Cells treated as in B were lysed and subjected to immunoblot analysis of indicated proteins (near-IR immunoblot). D) Cells treated as in B (with continuous MG-132-treatment in addition (MG)) were lysed after indicated time periods. Immunoblots against indicated proteins with -tubulin as loading control are shown. One example out of three independent experiments is shown (near-IR immunoblot). E) Quantification of normalized β5c and LMP7 intensities from three independent experiments as shown in D is depicted as fold up-regulation over unstimulated DMSO control samples (mean+SD for each time point). Two-way ANOVA, Sidak’s multiple comparison. F) WT CD4+ T cells treated as in C were activated and PR-825 was added to 100 nM final concentration after 4 h. Cells were harvested 20 h after activation and immunoblots performed as indicated. One example out of three experiments is shown (near-IR immunoblot).

102

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

7.6 B cells are similarly affected by ONX 0914, but show higher susceptibility to apoptosis induction

The high content of LMP7 containing proteasomes in both CD4+ T cells as well as CD19+ B cells prompted us to investigate whether the observed effects of ONX 0914 treatment in T cells applied to B cells as well. Therefore, isolated murine CD19+ B cells were activated with either PMA/ionomycin treatment or with F(ab’)2 anti-mouse IgG and anti-CD40 after pulse-treatment with ONX 0914 or DMSO. Up-regulation of CD69 on B cells was similarly reduced by ONX 0914 treatment as compared to T cells (Figure 20A-C). Human CD19+ B cells also showed significantly reduced CD69 up-regulation after treatment with ONX 0914 or with continuous MG-132 treatment (Figure 20D, E). Furthermore, immunoblot analysis of murine B cells from WT and LMP7-/- mice after pulse-treatment with DMSO or ONX 014 and activation with PMA/ionomycin showed that ONX 0914 induced ubiquitin-conjugates in B cells in an LMP7-dependent manner (Figure 20F). In contrast to T cells, WT B cells treated with ONX 0914 showed enhanced PARP cleavage compared to DMSO treated cells (Figure 20F). Overall enhanced PARP cleavage in both WT and LMP7-deficient B cells was observed when the cells were left unstimulated in vitro for several hours (Figure 20F). To check if B cells are more susceptible to apoptosis induction after ONX 0914 treatment as compared to T cells or if the enhanced PARP cleavage was due to the rather strong chemical activation stimulus of PMA/ionomycin treatment, B cells activated with anti- CD40 and F(ab’)2 anti-mouse IgG for 3 h, 6 h and 10 h were analyzed by immunoblotting as well. Again, ONX 0914 treated cells showed enhanced PARP cleavage after 6 h to 10 h compared to DMSO treated cells (Figure 20G). Additionally, MG-132 treatment induced a more marked induction of PARP cleavage (Figure 20G). Nevertheless, similar to the results obtained in T cells, B cells treated with ONX 0914 did not show enhanced ATF4 induction, but showed accumulation of Nrf1 (Figure 20G). Likewise, β5c was induced in DMSO as well as ONX 0914 treated cells, while after 10 h β5c appeared more strongly induced in ONX 0914 treated cells (Figure 20G). These results indicate that similar mechanisms of mild proteostasis stress induction as found in T cells apply to B cells as well, but that B cells are more readily susceptible to induction of apoptosis upon immunoproteasome inhibition.

103

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

Figure 20: Effects of ONX 0914 on primary murine and human B cell activation, proteostasis and apoptosis induction A) CD19+ B cells were purified from spleen by MACS, pulse-treated with DMSO or ONX 0914 for 2 h or continuously treated with MG-132 before activation with F(ab')2 and anti-CD40 for 6 h. Up-regulation of CD69 on the surface of CD19+ cells was measured by flow cytometry (histogram, left) and quantified as shown in the bar graph on the right (mean +SD from duplicate measurements). Shown is one example of three independent experiments with similar outcome. B) CD69 up-regulation was quantified in three independent experiments as shown in A after 5 - 6 h of activation with F(ab')2 and anti-CD40. Ratio of compound-treated over DMSO control is shown from three experiments as mean+95% CI. One sample t-tests, p-values indicated in the figure. C) Murine CD19+ B cells as in A were activated with PMA/ionomycin for 5 h. CD69 up-regulation was measured by flow cytometry. One example of two independent experiments with similar outcome is shown (mean±SD of duplicate measurements). D) Experiment as in C performed with human CD19+ cells isolated from PBMCs of healthy voluntary donors. One example of three independent experiments shown as mean±SD from duplicate measurements of median CD69 FI. E) Relative difference of CD69 up-regulation of compound-treated over DMSO treated from three independent experiments as shown in D. One sample t test, p-values as indicated in the figure. F) Cells treated as in C and activated for indicated time periods were subjected to immunoblot analysis of indicated proteins. One example of three independent experiments with similar outcome is shown. (ECL-based immunoblot) G) Cells treated as in A and B, activated for indicated time periods were subjected to immunoblot analysis of indicated proteins. One example of two independent experiments with similar outcome is shown. (near-IR immunoblot) 104

Chapter III – Results – ONX 0914 impairs ERK signaling and induces mild proteostasis stress in T cells

7.7 Induction of high immunoproteasome content in MEFs is sufficient to induce ubiquitin-conjugates after ONX 0914 treatment

B cells and T cells were found to constitutively express high amounts of LMP7 containing proteasomes already at the naïve state (section 6.2). Furthermore, ONX 0914 targeted LMP7 and LMP2 as shown here and in Basler et al., 2018 (section 12.5). Hence, the data suggested that high immunoproteasome content in combination with activation-induced proteomic changes could be sufficient to explain why ONX 0914 treatment exerted the observed effects on T and B cells. To test, if high immunoproteasome content also rendered other primary cells susceptible to ubiquitin-conjugate formation after ONX 0914 treatment, MEFs were treated with 200 U/ml IFN-γ for 48 h in order to induce high immunoproteasome content. Then, cells were treated with 300 nM ONX 0914 or 0.3% DMSO for 2 h and were either left untreated or stimulated with PMA/ionomycin for 4 h. IFN-γ stimulation induced LMP7, which was not detectable in MEFs without IFN-γ stimulation, and reduced β5c content. ONX 0914 treatment resulted in electrophoretic shifts of LMP7, but also to partial modification of β5c subunits. Nevertheless, only in IFN-γ pre-treated MEFs ubiquitin-conjugates accumulated after ONX 0914 treatment irrespective of further stimulation with PMA/ionomycin (Figure 21). Thus, induction of high LMP7 content appeared sufficient to render MEFs susceptible to ubiquitin-conjugate accumulation after immunoproteasome inhibition.

Figure 21: Induction of high immunoproteasome content renders mouse embryonic fibroblasts susceptible to ubiquitin-conjugate formation after ONX 0914 treatment. MEFs isolated from WT embroys were cultured in the presence or absence of 200 U /ml IFN-γ for 48 h. Cells were pulse-treated with 300 nM ONX 0914 (ONX) or 0.3% DMSO (D) for 2 h, and lysed after consecutive 4 h in presence or absence of PMA and ionomycin in the medium. Immunoblot analysis against indicated proteins. One example of three experiments with similar outcome is shown. (ECL- based immunoblot)

105

Chapter III – Results – Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation

8 IMMUNOPROTEASOME INHIBITION IMPAIRS DUSP5

EXPRESSION AND DUSP6 DEGRADATION

8.1 Impaired ERK-phosphorylation after ONX 0914 treatment depends on de novo gene expression and correlates with DUSP6 accumulation

The intensities of ubiquitin-conjugate accumulation as described in section 7.4 indicated that the majority of accumulating conjugates was induced after TCR stimulation. Activation of T cells induces metabolic reprogramming that requires a re-organization of the cellular proteome and metabolome within a short period of time [585, 588]. To check, if ubiquitin-conjugate formation was mainly driven by TCR-induced de novo protein synthesis, expanded T cells were pulse-treated with ONX 0914 or DMSO or continuously with MG-132 and activated with plate-bound anti-CD3/CD28 antibodies in the presence or absence of cycloheximide to inhibit protein neo-synthesis. In line with the results presented in section 7.4, cycloheximide treatment abrogated ubiquitin-conjugate formation in ONX 0914 treated T cells. In contrast, MG-132 also induced ubiquitin-conjugate accumulation within 3 h in cycloheximide-treated T cells (Figure 22A). These results supported the view that protein neo-synthesis was the primary source of accumulating ubiquitin-conjugates in ONX 0914 treated T cells. In contrast, MG-132 treatment affected also steady-state protein turnover. Notably, cycloheximide treatment not only enhanced overall ERK- phosphorylation intensities at 3 h after T cell activation, but also abrogated the relative reduction of ERK- phosphorylation intensities in ONX 0914 treated cells as determined by near-IR immunoblot quantification (Figure 22A, B). This result indicated that the effect on ERK-signaling sustainment was either mediated by a process that depends on the appearance of ubiquitin-conjugates and/or was mediated by a de novo expressed negative feedback regulator. As mentioned above, MEK-phosphorylation remained unaffected by ONX 0914 treatment, supporting the hypothesis that an inducible negative feedback regulator of ERK might be functionally involved. Dual specificity phosphatases can dephosphorylate both, serine/threonine as well as tyrosine residues of their targets. Several DUSPs were tested for their potential involvement in the ONX 0914-mediated effects on T cell activation as particular DUSPs have been reported i) to be expressed or activation-induced in T cells, ii) to be specific for individual target molecules like ERK and iii) to be regulated by ubiquitination and proteasome-mediated degradation (section 2.4). Of the tested DUSPs, the ERK-specific DUSP6 prominently accumulated in expanded T cells as well as in activated naïve T cells after ONX 0914 treatment (Figure 22A, D). Interestingly, no DUSP6 accumulation was detectable in DMSO or ONX 0914 treated cells after cycloheximide treatment, while MG-132 treatment resulted in DUSP6 accumulation also in cycloheximide treated cells, albeit to a much lower extend compared to MG-132 treated cells in the absence of cyclohximide treatment (Figure 22A). Furthermore, 106

Chapter III – Results – Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation the accumulating DUSP6 showed an altered electrophoretic mobility and was shifted to higher molecular weight as compared to DUSP6 in unstimulated cells (Figure 22A, D). These results indicated that DUSP6 was modified after activation, likely by phosphorylation and ubiquitination as described for DUSP6 before (section 2.4.3). To check if DUSP6 might be transcriptionally up-regulated at mRNA level in ONX 0914 treated cells, mRNA from expanded T cells stimulated for 3 h was isolated and qRT-PCR was performed for DMSO treated, ONX 0914 treated and MG-132 treated cells. Clearly, Dusp6 was transcriptionally induced after activation (~17-fold), but no significant difference in Dusp6 up-regulation was detected between DMSO and ONX 0914 treated cells. In contrast, MG-132 treatment resulted in significantly reduced mRNA levels of Dusp6 in spite of prominent accumulation of DUSP6 at the protein level (Figure 22C). Thus, ONX 0914 treatment likely impaired degradation of DUSP6 protein, but did not alter DUSP6 expression. In contrast, DUSP5 protein levels were slightly reduced after 5 h in activated naïve T cells

Figure 22: Reduced ERK phosphorylation and accumulation of ubiquitin conjugates correlate with accumulation of DUSP6 A) Expanded murine CD4+ T cells were pulse treated with 0.3% DMSO (D) or 300 nM ONX 0914 (X) or continuously treated with 10 µM MG-132 (MG) before activation with plate- bound anti-CD3/anti-CD28 antibodies for 3 h in the presence or absence of cycloheximide (CHX). Shown are immunoblots for ubiquitin and DUSP6 with γ-tubulin as loading control. One example of three independent experiments is shown. B) Quantification of p-ERK intensities normalized to γ-tubulin from three independent experiments performed as in A. Paired t-test, p-values as indicated. C) Expanded murine CD4+ T cells treated as in D were used for RNA extraction and q-RT- PCR for Dusp6 after 3 h of activation. Fold change of Dusp6 mRNA (normalized to Rpl13 and Ipo8) over unstimulated DMSO control. Pooled data from three independent experiments. Two way ANOVA, Sidak’s post test, p-values as indicated. D) MACS-enriched naive murine CD4+ T cells from WT mice were pulse-treated with 0.3% DMSO (D) or 300 nM ONX 0914 (X) for 2 h before activation with anti-CD3/anti- CD28 antibodies. Cells were lysed after indicated times and immunoblots performed as indicated. One example of at least three similar experiments. E) Experiment as in D, but with primary human CD4+ T cells from healthy donors and continuous MG-132 treatment (10 µM) in addition. One example out of three independent experiments is shown. F) Expanded CD4+ T cells from WT or LMP7-/- mice were pulse-treated with 0.3% DMSO (D) or 300 nM ONX 0914 (X) for 2 h before activation with plate-bound antibodies for indicated time periods. Immunoblot analysis as indicated. One example of two experiments is shown.

107

Chapter III – Results – Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation after ONX 0914 treatment, while other DUSPs (e.g. DUSP10) were not influenced by ONX 0914 treatment (Figure 22D). In total, tested DUSPs included DUSP1, DUSP3, DUSP4, DUSP7, DUSP9 and DUSP10, which were not affected by immunoproteasome inhibition (Figure 22D and Appendix Figure 5, [707]). DUSP6 accumulation and reduced DUSP5 expression after ONX 0914 treatment and MG-132 treatment were also detected in human CD4+ T cells isolated from healthy donors (Figure 22E). Finally, LMP7-deficient cells were tested for an effect on DUSP5 and DUSP6. While expanded WT CD4+ T cells showed accumulating DUSP6 and reduced DUSP5 correlating with increased ubiquitin-conjugate formation after ONX 0914 treatment, no effects on DUSP6 and DUSP5 were found in LMP7-deficient cells, neither with nor without ONX 0914 treatment (Figure 22F).

8.2 DUSP6 degradation, but not expression is impaired by ONX 0914 treatment

As DUSP6 did not accumulate in cycloheximide treated cells after ONX 0914 treatment, while MG-132 treatment led to accumulation of the activation-independent DUSP6 protein in the absence of de novo protein synthesis, radioactive labeling of DUSP6 protein was performed in a pulse/chase approach in T1 cells to clarify if DUSP6 accumulation was truly caused by impaired degradation. T1 cells were identified as a cell line showing several similar effects after ONX 0914 treatment as compared to primary cells ([702] and Figure 23A). In this work, the results obtained in expanded or naïve T cells regarding ubiquitin- conjugate accumulation, DUSP6 accumulation, PARP cleavage, reduced ERK signaling sustainment and even more prominently reduced DUSP5 expression were all similarly detected in T1 cells (with delayed reduction in p-ERK as compared to primary cells, as reduced ERK signaling was detected after 5 h, but not after 3 h) pre-treated with IFN-γ (Figure 23A). IFN-γ pre-treatment slightly enhanced LMP7 protein content, but markedly reduced the relative content of β5c rendering the cells predominantly LMP7- dependent (Figure 23B). Furthermore Dusp6 mRNA expression was induced after activation with plate- bound anti-CD3/CD28 antibodies and the relative effects of ONX 0914 and MG-132 treatment were comparable with expanded T cells, even showing a trend towards reduced Dusp6 expression after ONX 0914 treatment (Figure 23C). Therefore, T1 cells were pulse-treated with ONX 0914 or DMSO or continuously treated with MG-132 and activated by plate-bound antibodies against CD3/CD28. After 1-2 h of activation, cells were starved in cysteine/methionine free medium for 1 h followed by a 15 min radioactive labeling with 35S-cys/met. Immunoprecipitation against DUSP6 (controlled by a beads-only IP, data not shown) revealed an incorporation of radioactively labelled amino acids into endogenous DUSP6 protein in DMSO and in ONX 0914 treated cells, while DUSP6 labeling was barely detectable in MG-132 treated cells as could be expected due to reduced DUSP6 expression after MG-132 treatment. Notably, radioactively labeled DUSP6 was markedly diminished within 40 min after the pulse in the DMSO treated

108

Chapter III – Results – Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation control samples, while it remained present in ONX 0914 treated samples (Figure 23D). These results validated that partial immunoproteasome inhibition by ONX 0914 impairs DUSP6 degradation. Thus, while MG-132 mediated broad-spectrum proteasome inhibition impaired DUSP6 regulation both at the transcriptional level as well as at the level of degradation, ONX 0914 mediated immunoproteasome inhibition still permitted Dusp6 expression, while impairing DUSP6 degradation. Taken together these results emphasize that the degree of proteasome inhibition by selective inhibitors as opposed to broad- spectrum inhibitors can determine the impact on a proteasome-regulated process in T cells.

Figure 23: Radioactive labeling in T1 cells shows that ONX 0914 impairs DUSP6 degradation. A, B) T1 cells were cultured in the presence of 200 U/ml IFN-γ for 3 d and were subjected to immunoblot against LMP7 and β5c (B, one example out of at least three experiments showing similar results). T1 cells were then pulse-treated with 300 nM ONX 0914 or 0.3% DMSO for 2 h before or continuously with 10 µM MG-132 and activated with plate-bound anti-CD3/CD28 antibodies for indicated time periods before lysis and subjection to immunoblotting against indicated proteins (ECL-based). One example of three experiments with similar outcome. C) T1 cells treated as in A, B were used for RNA extraction and q-RT-PCR after 3 h of activation. Fold up-regulation of Dusp6 transcripts over DMSO unstim control is shown. Data were pooled from three independent experiments (mean±SD). Repeated-measures ANOVA, Sidak’s post test, p-values as indicated. D) T1 cells were treated with 200 U/ml IFN-γ for 3 days to induce higher immunoproteasome content. Cells were then pulse-treated for 2 h with DMSO, ONX 0914 or MG-132 as in C. Consequently, cells were activated with plate-bound anti-CD3/CD28 antibodies in RPMI 1640 +supplements. After 1 - 2 h the cells were starved in methionine/cysteine-free RPMI 1640 for 1 h, followed by a 15 min radioactive pulse of 35S- cys/met in RPMI 1640 at 250 µCi/ml. Cells were washed and lysed directly after (chase 0) as well as 20 min and 40 min after the pulse. Lysates were loaded for an anti-DUSP6-IP according to β-count CPM. After 6 h of IP and washing, the radioactive signal of newly synthesized DUSP6 was detected with a phosphoimager. Total DUSP6 in the IP and the antibody light chain were used as loading controls. One example of three experiments with similar outcome is displayed.

109

Chapter III – Results – Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation

8.3 No evidence for a functional involvement of DUSP6 in ONX 0914- mediated amelioration of T cell activation

DUSP6 has been previously characterized as a regulator of ERK-signaling peak and sustainment during activation of mouse and human T cells [664, 665, 679, 684, 708, 709]. Furthermore, phosphorylated, degradation-primed DUSP6 was shown to retain its phosphatase activity [674]. Hence, the results obtained here for DUSP6 accumulation in activated T cells strongly suggested a functional involvement of DUSP6 in the effect of ONX 0914 on ERK either via direct de-phosphorylation or via cytosolic retention [710]. Using confocal laser scanning microscopy, it was investigated whether ONX 0914 treatment had an influence on nuclear translocation of total and phosphorylated ERK in expanded T cells. While image quantification confirmed a reduction of ERK-signaling after 3 h of activation also in the nuclei of expanded T cells, no difference in nuclear translocation of total ERK was detected in cells treated with ONX 0914 or MG-132 compared with DMSO treated cells (data not shown, see [707]). As DUSP6 has been characterized as a cytosolic phosphatase, these results disfavored the hypothesis that DUSP6 was responsible for reduced ERK-phosphorylation. Hence, in order to clarify if DUSP6 played a functional role for the effects of ONX 0914 on T cell activation, DUSP6-deficent mice were obtained to test if ONX 0914 had an altered effect on T cell activation and ERK signaling in the absence of accumulating DUSP6. T cells isolated from DUSP6-deficient mice alone or together with DUSP6-heterozygous mice were used and compared to cells from age- and sex-matched DUSP6-proficient mice or true littermates from heterozygous pairings. Immunoblots with the residual splenocytes after selective enrichment of CD4+ cells and with bulk liver homogenates isolated from DUSP6+/+, DUSP6+/- and DUSP6-/- mice confirmed the genotypes obtained in genotyping PCRs at the protein level (Figure 24H). ONX 0914 treatment in isolated T cells resulted in similar reduction of CD69 up-regulation (Figure 24A, B), reduced ERK- phosphorylation sustainment (Figure 24D, E) and impaired IL-2 secretion (Figure 24F, G) in all tested groups. No difference in relative LMP7 content was detected between CD4+ T cells from heterozygous or DUSP6-deficient mice (Figure 24C). Thus, in spite of the marked accumulation and the results of previous studies regarding DUSP6 function in T cells, these results showed that DUSP6 was fully dispensable for all tested effects of ONX 0914 treatment during in vitro T cell activation.

110

Chapter III – Results – Immunoproteasome Inhibition impairs DUSP5 expression and DUSP6 degradation

Figure 24: No evidence for a functional involvement of DUSP6 in the observed effects of ONX 0914 on T cell activation. A, B) MACS enriched CD4+ T cells from WT (DUSP6+/+) and DUSP6-/- mice were pulse-treated for 2 h with 300 nM ONX 0914 or 0.3% DMSO before activation with plate-bound anti-CD3/CD28 antibodies for 5 h. Cells were harvested and analyzed by flow cytometry. Representative plots for gating (upper panel) and histograms (lower panel) from one experiment are shown in A. Quantification of median fluorescence intensities as pooled data from three independent experiments is shown in B (mean±SD). Two-way ANOVA, Sidak’s post test, p-values as indicated. C) Immunoblot comparing LMP7 protein content between CD4+ T cells from DUSP6+/- and DUSP6-/- mice. One example of two experiments. D, E) MACS enriched CD4+ T cells from WT (DUSP6+/+), heterozygous (DUSP6+/-) and knockout (DUSP6-/-) mice treated as in A and activated for 3 h. Cells were harvested and stained for intracellular p-ERK1/2 for analysis by flow cytometry. One representative example of histograms is shown in D. Ratios of p-ERK1/2 MFI of p-ERK+ CD4+ cells between ONX 0914 treated over DMSO-treated cells is shown in E. Shown is pooled data from three independent experiments (mean with 95%CI). One sample t tests with µ0 = 1 (dashed line), p-values as indicated in the figure. F, G) Supernatants from cells treated as in A were used to quantify IL-2 secretion by ELISA. One representative example (mean±SD from triplicates) of three independent experiments is shown in F. Relative reduction of IL-2 secretion after ONX 0914 treatment is shown as ratio of ONX 0914-treated over DMSO-treated cells from three independent experiments in G. One-sample t test with null hypothesis mean value µ0 = 1 (dashed line), p-values as indicated, mean with 95%CI. Comparison between DUSP6+/+ and DUSP6-/- was performed as paired t test, p-values are indicated in the figure. H) Immunblots against indicated proteins from splenocytes after CD4+ sort and bulk liver homogenate from the three genotypes. 111 Shown is one example of two experiments. Chapter III – Results – Impaired in vivo activation of antigen-specific CD4+ T cells by ONX 0914

9 IMPAIRED IN VIVO ACTIVATION OF ANTIGEN-SPECIFIC

CD4+ T CELLS BY ONX 0914

The observed effects of ONX 0914 treatment on lymphocyte activation offer a mechanistic explanation for the beneficial effects of immunoproteasome inhibition in various pre-clinical models for autoimmune diseases. Given the complexity of TCR-mediated activation that is not covered by the use of invariable anti-CD3/CD28 stimulation for in vitro activation (section 2.3), it was aimed to check if impaired T cell activation was similarly detectable for in vivo activated antigen-specific T cells. Therefore, an LCMV- infection model was used. SMARTA mice [686] are transgenic for an MHC-II restricted T cell receptor recognizing the LCMV epitope GP61-80. By focusing on CD4+ T cell activation this model allows testing in vivo effects of ONX 0914 in antigen-specific T cells that are likely not affected by effects of ONX 0914 on MHC-I antigen-processing. SMARTA mice were treated subcutaneously with ONX 0914 at 10 mg/kg or with vehicle 2 h before intravenous infection with 4 * 105 pfu LCMV-WE. At 6 h after LCMV-infection, little or no T cell activation was detectable by CD69 up-regulation (data not shown). In contrast, almost all splenic CD4+ T cells were activated within 18 h after LCMV infection as measured by CD69 up- regulation. CD25 was also up-regulated, but to a lesser extent than CD69. Importantly, both CD69 and CD25 were ~45% less up-regulated on T cells from ONX 0914 treated mice. These results confirmed attenuated activation of antigen-specific T cells after ONX 0914 treatment in vivo.

Induction of ubiquitin-conjugates and analysis of covalent LMP7 modification in vivo was assessed using purified CD4+ T cells from infected and uninfected mice as outlined above. After 18 h of infection, enhanced ubiquitin-conjugate formation in CD4+ T cells from infected mice compared to uninfected mice was detected. Thus, in vivo activation in an LCMV-infection environment induced proteostasis stress in T cells. However, no relatively enhanced ubiquitin-conjugates were observed in cells from ONX 0914 treated mice as compared to vehicle treated. Interestingly, quantification of signal intensities from near-IR immunoblots normalized to γ-Tubulin indicated an up-regulation of the β5c subunit in LCMV infected mice, but an impaired up-regulation of β5c in ONX 0914 treated mice. The same was observed for IOTA as a marker of total proteasomes. In contrast, LMP7 protein levels were changed to a much lesser albeit statistically significant extent. However, it was not possible to detect signals for Nrf1 in the samples (data not shown) allowing no conclusion about an involvement of Nrf1. LMP7 subunits in ONX 0914 treated mice showed electrophoretic mobility shifts, but also signals at the molecular weight of unmodified LMP7, in both LCMV-infected and LCMV-uninfected cells. Taken together, while impaired T cell activation by ONX 0914 treatment was validated for in vivo activated T cells, the molecular alterations accompanied with impaired activation detected in vitro were not detectable in the in vivo activated T cells here. Rather, the immunoblots obtained at this particular time point after infection implicated a different molecular

112

Chapter III – Results – Impaired in vivo activation of antigen-specific CD4+ T cells by ONX 0914 characteristic in T cells within an LCMV infected environment as compared to in vitro activated T cells and further analysis of other time points is demanded.

Figure 25: Impaired activation of antigen-specific CD4+ T cells in vivo. SMARTA mice were treated with 10 mg/kg ONX 0914 or vehicle s.c. 2 h before i.v. infection with 4x105 pfu LCMV or left uninfected. 18 h after infection, splenocytes were subjected to flow cytometry to assess CD69 and CD25 up-regulation on CD4+ T cells. A, B) Histograms show one representative profile of a mouse from each group (A). Bar graphs (B) show pooled data from MFI quantifications (mean±SD). Two-way ANOVA, Sidak’s post test, p-values as indicated in the figure; n = 3 for uninfected controls, n = 4 or 5 mice for infected group. Note that one mouse was determined to be not properly infected and was therefore omitted from analysis (vehicle-treated, LCMV-infected, median CD69 FI: 3511, not shown in the graph). C) Near-IR based immunoblots of purified CD4+ T cells from 3 (uninfected) or 4 (LCMV-infected) mice of each treatment group with γ-Tubulin used as a loading control. D) Near-IR based quantification by normalization to γ-Tubulin signal intensity from the same membrane for the obtained signals as indicated on the y-axes. The signal quantification was performed with the immunoblot data shown in C. Two-way ANOVA, Sidak’s post test, p-values as indicated. Data points refer to 3 (uninfected) to 4 (LCMV-infected) individual samples from mice of each treatment group as shown in C.

113

Chapter IV– Discussion

CHAPTER IV: DISCUSSION

10 DISCUSSION

Immunoproteasome inhibition has been demonstrated to hold potential for the treatment of a variety of autoimmune diseases in pre-clinical models [80, 199, 363, 366]. The immunoproteasome inhibitor KZR- 616, a derivative of ONX 0914, has meanwhile entered clinical trial phase (clinical trial ID: NCT03393013). ONX 0914 treatment was described extensively to have anti-inflammatory effects by reducing cytokine secretion, but it has remained largely unknown by which molecular mechanism immunoproteasome inhibition alters cytokine secretion and disease outcome [80, 199, 288, 363, 366, 711]. The work presented in this thesis was aimed to shed new light on the molecular processes accompanied with immunoproteasome inhibition (and immunoproteasome-deficiency). Since in many autoimmune diseases lymphocytes play central roles and previous studies have indicated a T cell intrinsic function of immunoproteasomes, the focus of this study was set directly on primary lymphocytes from mouse and from humans to investigate how ONX 0914 treatment affects lymphocytes and their activation at the molecular level. In this respect, the presented work has addressed two major issues:

The involvement of immunoproteasomes in lymphocyte signal transduction The function of immunoproteasomes in lymphocyte protein homeostasis

Based on disease models in which both, LMP7-deficiency and ONX 0914 treatment resulted in disease amelioration as well as due the observation of a selective loss of LMP7-deficient CD8+ T cells when co- injected with WT cells into an LCMV-infected host [137], a functional segregation between immunoproteasome regulated processes and standard proteasome regulated processes within the same cell has been hypothesized. Thus, the existence of a selectively LMP7-regulated factor, which is affected by ONX 0914 treatment or LMP7-deficiency was proposed [68]. However, only in some disease models LMP7-deficiency showed the same effects as LMP7 inhibition by ONX 0914. While LMP7-deficiency and ONX 0914 affected LCMV-induced meningitis as well as DSS-induced and T-cell-transfer-induced colitis [195, 288, 328, 366], controversial data was reported with respect to EAE of either unaffected [80] or enhanced susceptibility to EAE in LMP7-deficient mice [73]. LMP7-germline-deficiency was connected to altered proteostasis capacity by some groups, but others have provided data disfavoring such a role [73, 74, 78, 79, 186]. Importantly, Basler & Mundt et al. showed that both inhibition of LMP7 in WT mice as well as inhibition of β5c in LMP7-deficient mice ameliorated symptoms of EAE [80]. Hence, in the EAE model, ONX 0914 efficacy rather depended on reduced proteasome capacity than on a selectively LMP7- processed factor. Therefore, the hypothesis was adapted in that a process seemed to be involved which

114

Chapter IV– Discussion either depended on the LMP7-specific cleavage activity or in general on the chymotrypsin-like activity of the proteasome β5 position. LMP2-deficient mice intrinsically lack chymotrypsin-like activity at the β1 position, but incorporate LMP7 at the β5 position (see also Figure 9). Therefore, it was predicted that LMP2-deficient cells would be even more affected by ONX 0914 treatment than WT cells, if indeed the chymotrypsin-like activity fulfilled the hypothesized function of processing a regulatory factor. However, experimental testing unexpectedly showed the opposite result: LMP2-deficient T cells were almost as little affected by ONX 0914 treatment as LMP7-deficient cells. This observation and the covalent modifications of both LMP7 and LMP2 subunits by ONX 0914 leading to apparently higher molecular weight in SDS- PAGE and immunoblotting revealed that in fact ONX 0914 inhibited both LMP7 and LMP2, which synergistically impaired T cells activation (Figure 9). Using the novel LMP7-selective inhibitor PRN1126, M. Basler and colleagues showed that LMP7-inhibition alone was neither sufficient to impair Th17 polarization, nor did PRN1126 alone ameliorate symptoms of EAE or DSS-induced colitis. However, combinations with additional LMP2-selective inhibitors were effective. Taken together, these data led to the finding that indeed, dual-inhibition of LMP7 and LMP2 was underlying the pre-clinical efficacy of ONX 0914 or combined LMP7- and LMP2-selective inhibitor treatments (Basler et al. 2018, section 12.5). However, LMP7-deficiency, LMP2-deficiency, as well as MECL-1-deficiency resulted in resistance to DSS-colitis and the affected cytokines reduced by ONX 0914 treatment were not intrinsically affected by LMP7-deficiency [195, 366, 367]. Therefore, even in the colitis model, disparate molecular mechanisms between ameliorating effects of either ONX 0914 treatment or immuno-subunit-deficiency seem to be at play. It appears somewhat surprising that the dual inhibition profile of ONX 0914 was overlooked in previous studies. However, at least three reasons are likely to account for the fact that LMP2 inhibition by ONX 0914 had remained underestimated: First, the originally used active-site ELISA in the first study describing ONX 0914 indicated much less LMP2 occupancy by the compound as compared to the more recently used assay with LMP2-selective fluorogenic substrates ([199] and Basler et al. 2018, section 12.5). Second, the crystal structure of ONX 0914 soaked mouse immunoproteasome showed that ONX 0914 could in principal bind to the LMP2 substrate pocket, but that its binding was impaired by steric hindrance of Phe31 [31]. Third, the observation that ONX 0914 had no effect on LMP7-deficient cells in contrast to WT cells [199, 366] was interpreted to support its selectivity. However, it had not been taken into account that the reportedly reduced LMP2-incorporation into LMP7-deficient proteasomes [86] (see also Figure 10) would obscure such a synergism. Since ONX 0914 had an effect on cytokine secretion in LMP2/MECL-1 double-deficient splenocytes [330], how additional MECL-1 germline deficiency seems to influence this effect of ONX 0914 remains fully elusive. Even more striking seems to be the observation of marked β5c modification by ONX 0914 at 300 nM in cell lysates, even though the reported IC50-value for β5c-binding is around 900 nM [31]. Indeed, the marked conformational re-arrangement of the β5c substrate binding channel in ONX 0914 soaked crystals of the immunoproteasome was interpreted to

115

Chapter IV– Discussion explain why β5c binding by ONX 0914 was unfavorable, hence explaining the selectivity for LMP7 together with the smaller S1 pocket size of β5c [31]. The immunoblot data obtained in this work indicate that ONX 0914 can more potently bind to β5c in the cellular context than previously determined in vitro (Figure 9, Figure 20 and Figure 21). It is interesting in this regard that a principal component analysis of eukaryotic 20S CP structures combined with molecular dynamics simulations has suggested that β5i is structurally pre-formed for optimal peptide ligand binding while β5c undergoes a conformational change largely affecting conformation on the surface of the CP, but also influencing overall β5c structure [69]. The authors write

[…] our findings suggest that assisting or impeding the backbone shift of the distal segments may have significant effects on affinity and kinetics of peptidic inhibitors. [69]

Their molecular dynamics simulation suggests that upon release of peptide ligands from β5c the structure re-arranges to the ligand-unbound conformation at a time scale of 400 ps – 2 ns [69]. Combined with the role of allosteric regulation reported to affect proteasome activity [38, 70], a possible explanation for the observed β5c modification after ONX 0914 treatment in cells is offered by the possibility that structural re- arrangements occurring in cells in the presence of proteasome substrates result in an enhanced ability of ONX 0914 to bind to β5c in cells as opposed to in vitro binding to purified 20S proteasomes.

LMP7-deficiency and immunoproteasome inhibition may have similar or disparate effects depending on the read-out of the experiment or investigated disease model. Importanly, as mentioned above, similar effect outcomes do not necessarily mean that the same mechanistic basis is underlying. It needs to be highlighted that conclusions drawn from studies using immunoproteasome-subunit inhibition or immunoproteasome-subunit deficiency are not suited for generalized conclusions about the function of a subunit like LMP7 for the investigated process. This is emphasized because previous studies did transfer conclusions from one approach to interpret results obtained in the other approach. A few examples may be noted: Opitz et al. demonstrated that LMP7-deficiency results in enhanced virus-induced myocarditis [74]. While Paeschke et al. later demonstrated that LMP7-deficiency and LMP7-inhibition both reduce expression of the cardio-protective protein PTX-3, the mechanistic conclusion that altered ERK- and p38- signaling in macrophages was involved was solely performed with ONX 0914 treatment [200]. Very recently the same group then showed that in contrast to LMP7-deficiency, ONX 0914 treatment did not increase susceptibility to virus-mediated myocarditis, but acted protective against exacerbated immune- mediated tissue damage. Hence they showed disparate mechanisms of action between immunoproteasome inhibition and immunoproteasome deficiency [331]. Similarly, Kalim et al. showed altered T helper cell polarization in LMP7-deficient as well as ONX 0914 treated cells, but mechanistic involvement of altered STAT1 and STAT3 signaling were solely investigated using ONX 0914 [288]. Joeris and colleagues observed that bortezomib-treatment showed dose-dependent adverse effects in ameliorating DSS-colitis

116

Chapter IV– Discussion while LMP7-deficient mice (which showed reduced or no DSS-colitis) did not show similar pathophysiology. They interpreted this observation as an indication that selective LMP7-inhibitors will be beneficial for the treatment of inflammatory bowel disease (IBD) [195]. However, while LMP7-selective inhibitors might indeed show less adverse effects and were indeed effective in mouse models of IBD [288, 366], the logic of the argument is based on the assumption that LMP7-deficiency and LMP7-inhibition will have the same effects, because Joeris and colleagues did not themselves test LMP7-selective inhibitors [195]. Recently, Vachharajani et al. as well as Körner et al. demonstrated independently that LMP7- deficiency and ONX 0914 treatment both impaired colon cancer progression in different models including colitis-associated cancer [349, 350]. Again, it is not demonstrated whether these effects are based on the same or similar molecular mechanisms. Clearly, it is tempting to assume that inhibition and deletion of a target protein might likely results in the same effects and this is a commonly used means to independently assess the function of the target of interest. In case of the immunoproteasome, however, it is established that things are more complicated. The current state of research is that proteasome subunits exert their functions when they are successfully incorporated into the 20S CP. While recent evidence showing the anchoring of pre-cursor proteins by regulators like PRAS40 [175] opens the question whether proteasome subunits could affect cellular processes also prior to incorporation, the vast majority of studies focusses on the function of proteasomes as assembled multi-protein complexes. Therefore, it is clear that inhibiting the enzymatic cleavage function of LMP7 in assembled 20S CPs as opposed to germline deficiency of LMP7, which results in incorporation of β5c, are distinct molecular characteristics [80]. Unraveling that ONX 0914 is not as selective as previously thought additionally demands distinguishing carefully between germline-deficiency and pharmacological inhibition. Only recently, Yang et al. reported that LMP7 plays a role in regulating proliferation and growth of glioma cells and linked their findings to reduced cyclin expressions and altered ERK and Akt signaling pathways [712]. However, their entire study was performed using RNA interference against Psmb8. Thus, the de novo expression of immunoproteasome subunits is impaired in an actively proliferating cell. Whether compensatory β5c incorporation occurs like in LMP7-germline-deficient cells was not investigated. The authors did not address if proteasomes were still functioning in these cells. This strongly limits their interpretation that their results are attributed to LMP7 functions [712], because they could also be attributed to disturbed proteasome function in general. There is also the possibility that LMP7-germline-deficiency might have a so far unidentified global impact during cellular or overall organismal ontogeny (which could speculatively involve different DNA- methylation patterns for examples). Indeed, two studies by De Verteuil et al. have reported globally altered gene expression patterns in LMP7/MECL-1-double-deficient DCs [167, 713]. Hence, germline-deficiency and short-term siRNA-mediated absence of LMP7-protein could again result in different molecular characteristics. Another example are studies focusing on mutated Psmb8-genes (like for example found in CANDLE syndrome), which might again lead to another underlying molecular characteristic in that it

117

Chapter IV– Discussion impairs function via mal-assembled 20S CPs [302]. Therefore, interpretation of any findings in published articles about the function of the immunoproteasome must carefully take into account different conditions of how the immunoproteasome was experimentally affected. This was also emphasized for the interpretations about the existence of intermediate proteasome species, where some subunit-combinations may only occur in individual gene-targeted mice, but are unlikely to contribute to the possible proteasome compositions in genetically unmodified organisms [97]. Experimental conditions used to interpret conclusions about the function of immunoproteasomes might hence be categorized mainly in four ways:

a. mutations leading to structural disruption of normally intact proteasome particles as found in CANDLE syndrome and similar diseases b. deletion of one or more immunoproteasome subunits from the genome which can be partly compensated by incorporation of standard subunits c. pharmacological inhibition of intact proteasome particles with broad-spectrum or subunit-selective inhibitors d. down-regulation of proteasome subunit expression in cells via RNA interference

In this work it was demonstrated that LMP7-deficiency did not affect early T cell activation (and T cell proliferation [702]) as compared to WT mice (Figure 9). In contrast, immunoproteasome inhibition with ONX 0914 impaired T cell activation markedly. It hence shows that these conditions are fundamentally different with respect to their consequences on early T cell activation, which was not addressed in more detail before.

10.1 Immunoproteasome inhibition and deficiency in T cell polarization

Impaired T cell activation as indicated by reduced CD69 surface expression and reduced IL-2 secretion is in line with previous results about ONX 0914 treatment in T cells [80, 199, 366]. An influence on cytokine secretion might also be connected to altered T cell proliferation and polarization, but impaired proliferation was not merely due to less IL-2 secretion, because proliferation could not be rescued by exogenous IL-2 (Figure 12). Impaired activation was detected already after 5 h of activation indicating an early cell- intrinsic effect of ONX 0914 treatment, which therefore gave rise to investigate into T cell signaling as discussed below. Notably, neither T cell activation markers nor T cell proliferation were affected by LMP7-germline-deficiency, which therefore was used as a control for immunoproteasome-selective (however, not necessarily LMP7-selective) effects of ONX 0914 [199, 366]. The overall impaired proliferation in the presence of ONX 0914 correlated with impaired functional polarization towards Treg cells and Th17 cells. However, some experimental results obtained in this work were in disagreement with results from similar experiments in previously published work. Impaired Th17 polarization by ONX 0914 treatment was consistently found here and in several other reports [199, 288, 337, 714, 715]. Conflicting 118

Chapter IV– Discussion data was found with respect to Th1 polarization and Treg polarization before [288, 714, 715]. In contrast to the results by Kalim et al. no enhanced Treg formation under Th17 skewing conditions was observed in ONX 0914 treated cells in this work. While it is unclear, how this partially conflicting data is explained, T cell polarization can delicately depend on the conditions of activation. For example optimal concentrations of anti-CD3/CD28 stimulating antibodies were shown to influence Foxp3 expression in activated T cells [716, 717]. As stressed by van Panhuys et al., quantitative differences in activation strength are often not considered in experimental design [598] and it has to be critically noted that only standard conditions of anti-CD3/CD28 stimulation have been used in this work as well. Experiments using lower and varying concentrations of anti-CD3/CD28 have not been systemically analyzed for comparison here. However, unappreciated differences in protocols regarding the nature of stimuli might influence experimental results significantly. This was recently exemplified with respect to opposing results that had been reported about the functional importance of BATF in Th2 polarization. In 2017, Iwata et al. found that loss of BATF is partially compensated by BATF3 in Th2 polarization, but BATF3 expression depended on the TCR stimulus strength and thus, compensation under different activation conditions depended on the differential sensitivities of genes in response to BATF3 levels. This indicated that different activation conditions have most likely caused the opposing results in several other studies before [423]. This example should emphasize that disparate results with previous reports might be due to delicate differences in stimulation conditions that could remain unidentified although standard protocols for cell activation are used, but stimulating antibody clones and commercial cytokines are not the same chemicals as used before (e.g. for technical reasons like supplier availability). While Th17 polarization was performed in a way aiming to reproduce the experiments by K. Kalim, Treg polarization was not performed with identical experimental conditions between this work and the work by Kalim et al. who analyzed Th17 and Th1 polarization in the continuous presence of ONX 0914, but Treg polarization after pulse-treatment before activation. In this work, the same conditions of ONX 0914 treatment were used for Th17 polarization and Treg polarization. Hence, it seems that the reported enhanced Foxp3+ Treg polarization under either Th17 or Treg skewing conditions is not a generalizable effect of ONX 0914 treatment, but depends on details in polarization conditions. The pulse-treatment was also used here for T cell activation and originally intended to ensure LMP7-selective binding without affecting β5c. Nevertheless, ONX 0914 resulted in LMP2 inhibition and some effect on β5c as well, which showed that the reported selectivity had been overestimated before (as discussed above). ONX 0914 had no effect on T cell activation or proliferation in LMP7-deficient cells (Figure 9 and [702]), showing that the observed partial off-target inhibition of β5c was not sufficient to impair T cell function. Instead, ONX 0914 nevertheless exerted its effect on T cell function in an immunoproteasome-dependent manner. It remains to be determined if pulse-treatment would result in differential effects on Th17 polarization as well. In the aftermath it was found that ONX 0914 treatment resulted in ubiquitin-conjugate accumulation after

119

Chapter IV– Discussion activation already after pulse-treatment (Figure 18), but the apparent enhanced ubiquitin-conjugates in ONX 0914 treated cells were cleared by T cells after 20 h of activation (Figure 19). While it was originally hypothesized that ONX 0914 might influence a factor that is selectively processed via the β5 position of the proteasome, the results in this work rather do not indicate the involvement of an ONX 0914 affected particular factor with a particular role in T cell polarization. That the extent of proteostasis stress in T cells might vary greatly depending on the strength of activation and the duration of ONX 0914 treatment appears more plausible instead. This might lead to different effects on T cell polarization in the continuous presence of ONX 0914 than by immunoproteasome inhibition only during the initial activation. Hence, the different protocols used by Kalim et al. might have misleadingly indicated a general immunomodulatory function of ONX 0914 treatment, which was not corroborated in this work. A more detailed characterization of T cell activation impairment by ONX 0914 under varying activating stimulus conditions in the future might also be demanded. Furthermore, the novel reportedly highly LMP7-specific inhibitor DPLG3 also impaired T cell proliferation [293] indicating that different requirements for stoichiometric subunit inhibition might underlie effects on early T cell activation and on proliferation after prolonged activation.

Although no further focus was set on polarization and regulatory T cells in this work, they may not only be relevant as directly affected cells by ONX 0914 treatment in vivo, but their functional action might be promoted via ONX 0914-mediated targeting of conventional T cells. Regulatory T cells suppress IL-2 expression in conventional T cells and impair their expansion. Likewise ONX 0914 treatment impairs IL-2 secretion during T cell activation and impairs their proliferation. I.e. by promoting suppressive impairment of T cell activation, ONX 0914 treatment might indirectly promote regulatory T cell function in vivo as well. Indeed, enhanced in vivo percentages of Treg cells in mesenteric lymph nodes of ONX 0914 treated mice after DSS-colitis induction were reported by Kalim et al. [288]. However, an increased percentage of Foxp3+ cells of LMP7-deficient cells (as opposed to WT cells) two weeks after adoptive transfer into RAG-/- mice was also found and further demonstrated a T cell-intrinsic role for immunoproteasomes. Nevertheless, these experiments do not explicitly allow the conclusion whether more Tregs were formed from naïve T cells or whether the relative enhancement of Tregs in vivo was due to selective loss or less generation of Th1 and Th17 cells in LMP7-deficient mice [288]. Loss of LMP7-deficient cells after adoptive transfer was reported previously [137] and might differentially affect CTLs, Th1 and Th17 cells as opposed to Tregs.

One novel aspect regarding T cell polarization that was uncovered in recent years might also be shortly discussed at this point: Evidence has now suggested that the classical concept of a predominantly cytokine- imposed T helper cell polarization needs to be revised for the fact that TCR signal strength and “TCR tuning” as a result of thymic selection (section 2.3.7) seem to have a greater impact on T cell polarization

120

Chapter IV– Discussion than previously appreciated [598, 718]. These novel findings emphasize that the selection in the thymus can already influence the T cell-intrinsic behavior upon re-activation in the periphery. While the immunoproteasome is known to be involved in MHC-I antigen-processing, an influence on CD4+ T cell selection in the thymus should not be ruled out in the light of some recent reports. Presentation of self- peptides on MHC-II for CD4+ T cell selection is less well understood as compared to MHC-I presentation, but it is known that autophagy and endocytosis contribute to the generation of MHC-II epitopes by lysosomal proteases like Cathepsin S [719] in mTECs and DCs. Furthermore, transfer of Aire-regulated tissue-specific auto-antigens (promiscuous gene expression) from mTECs to thymic DCs was shown to contribute to negative selection [399, 720, 721]. However, both an influence on mTEC function as well as an influence on global gene expression in DCs was reported for immunoproteasome-deficient mice [167, 186]. Additionally, cTECs present macroautophagy-derived self-epitopes which are hence dependent on both, lysosomal proteases and proteasomes [399]. Therefore, it cannot be ruled out that effects of immunoproteasome-deficiency might result in differences in selection in the thymus which subsequently influences the naïve CD4+ T cell response [718]. However, there seems to be not much evidence in favor of such an influence. The proposed mechanism underlying the thymic influence on naïve CD4+ T cell responses involves thymus-imposed differential CD5 expression and it could be tested if LMP7-deficient T cells in general show altered CD5 expression compared to WT T cells [718]. Importantly, and arguing against such thymus-imposed influence on CD4+ T cell activation, altered IL-2 secretion, ERK-signaling and proliferation were also reported to be influenced by thymus-imposed effects on CD4+ T cell activation and these parameters were not found to be differentially affected in polyclonal T cells from LMP7- deficient as compared to WT mice in this work (discussed below). Furthermore, DSS-colitis was less severe in irradiated recipient mice of either WT or LMP7-deficient background after reconstitution with LMP7-deficient bone marrow as opposed to WT bone marrow [195]. Although this model is not restricted to assessing the contribution of T cells it also argues against an indirect thymus-imposed effect in the DSS- model. Nevertheless, a possible experiment to test if the selection mechanism in the thymus of LMP7- deficient mice could pre-impose a different TCR tuning and polarization on CD4+ T cells in an in vivo environment like in the experiment by Kalim et al. mice could be performed as follows: Irradiated LMP7- deficient mice or WT mice both receive a bone marrow transfer from either LMP7-deficient mice or WT mice. After reconstitution and recovery of the peripheral T cell compartment, CD62L+ CD4+ T cells are isolated from these mice and transferred into RAG-deficient mice as performed by Kalim et al. [288]. If all LMP7-deficient T cells show the same altered polarization after transfer into RAG-deficient mice, it is clear that entirely T cell-intrinsic effects are at play irrespective of whether the cells were selected in a WT or LMP7-deficient thymus. In contrast, if all T cells selected in the LMP7-deficient thymus show altered polarization after transfer into RAG-deficient mice, the T cell-intrinsic effect of LMP7-deficiency was likely imposed on the cells during selection before re-activation in the periphery. Alternatively, the use of

121

Chapter IV– Discussion conditional LMP7-deficient mice by CD4-Cre driven LMP7fl/fl deletion would be of use to test this possibility. These latter aspects about possible differences between LMP7-deficient and WT cells regarding T cell polarization are highly theoretical at this point, and no further focus was set on T cell polarization in this work. In contrast, this work refocused on the initial events of T cell activation in order to obtain improved mechanistic insight into the effects of immunoproteasome inhibition or deficiency in T lymphocytes.

10.2 Immunoproteasome inhibition in T cell activation signaling

TCR-induced signaling is an integral part of T helper cell polarization as well. Therefore, molecular effects characterized at the level of initial T cell activation were considered to be important as a preceding analysis before further effort into underlying molecular details of more complex polarization signaling was pursued. Hence, the next step was to extend the analysis of TCR-induced signaling pathways under the influence of ONX 0914 as only limited data could be obtained in my master’s thesis in this respect [702]. Nevertheless, the previous establishment of expanded T cells during my master’s thesis was used as a basis to extend the investigation into TCR-signaling in this work. Importantly, characterization of CD69 and IL-2 up- regulation at the mRNA level provided an additional rationale that pre-transcriptional regulation might be involved, although a statistically significant difference at the mRNA level was only obtained for CD69 (Figure 14). The additional analysis of all catalytically active proteasome subunits in T cells and B cells from WT and LMP7-deficient mice indicated that naïve T cells and B cells were almost exclusively dependent on LMP7-containing proteasomes, which was therefore also characterized in expanded T cells (Figure 15). The similarities regarding reduced CD69 up-regulation after ONX 0914 treatment and the similar dependency on LMP7-containing proteasomes rendered expanded T cells as a model system to further investigate into signaling pathways by immunoblotting.

The proteasome is a well-known regulator of NF-κB signaling via degradation of IκBα (sections 1.5.3 and 2.3.3). It was previously reported that ONX 0914 treatment did not affect NF-κB signaling assessed by luciferase reporter assays or by IκBα degradation and p65 nuclear translocation in macrophages and cardiomyoctes [199, 200]. Also the combined treatment of the β1i inhibitor UK-101 and the β5i inhibitor LKS01 did not impair IκBα phosphorylation and degradation in PC-3 prostate cancer cells [198], which is functionally largely equivalent to ONX 0914 treatment as ONX 0914 is now recognized as an LMP7 and LMP2 inhibitor (Basler et al., 2018, section 12.5). However, a possible effect on NF-κB signaling was also tested for primary T cells in this work. Analyzing IκBα degradation, p65 phosphorylation as well as p65 nuclear translocation revealed that ONX 0914 treatment did not impair canonical NF-κB signaling in T cells in agreement with previous results (Figure 16). Notably, an influence of ONX 0914 treatment on non-canonical NF-κB activation via TRAF3 and NIK was not investigated here, but could be of interest for

122

Chapter IV– Discussion further investigation into Th17 polarization as this pathway was reported to play a role in Th17 polarization and is also controlled by the proteasome ([527] and section 2.3.3). The nuclear translocation of NFAT was also tested in expanded T cells in this work, but no difference in the Ca2+ NFAT signaling pathway between ONX 0914 treated and DMSO treated cells was indicated. Likewise, no influence of ONX 0914 on Akt(Thr308) signaling, mTORC1 (as indicated by S6 phosphorylation), mTORC2 (as indicated by Akt(Ser473) phosphorylation) or p38 was detected (Figure 16). Instead, a possible influence of ONX 0914 on the ERK signaling pathway was apparent in ECL-based immunoblots after optimization of experimental procedures and sample handling with respect to the work provided in my master’s thesis [702]. To assess whether ERK-signaling was altered by ONX 0914 treatment, densitometry analysis from ECL-based immunoblots was used. Indeed, this analysis indicated a reduction in ERK signaling sustainment in WT, but not LMP7-deficient cells. However, quantitative ECL-based immunoblotting has considerable limitations [704]. Therefore, the apparent, but rather small reduction in ERK-phosphorylation sustainment was more extensively analyzed using different approaches. Both, implementation of the newly obtained LI-COR Odyssey Imager and establishment of a protocol for intracellular detection by single- cell-based flow cytometry allowed corroborating the validity of the observed effect in both expanded T cells as well as in naïve T cells upon activation. Importantly, the effect was again absent in LMP7- deficient naïve T cells, showing that it was caused by immunoproteasome-selective inhibition and that it was no off-target effect of ONX 0914. Furthermore, reduced ERK-signaling was also observed in human T cells (Figure 17). Reduced ERK signaling sustainment can result in multiple downstream outcomes including reduced IL-2 cytokine secretion and reduced CD69 up-regulation as both were found to be dependent on ERK (section 2.3.8). The multitude of downstream ERK targets makes it difficult to identify the exact downstream alterations as a consequence of the observed reduction in ERK phosphorylation sustainment. A more detailed analysis of ERK downstream targets reported to be involved in CD69 and IL-2 regulation is a possible strategy to strengthen the so far only correlation-based assumption that reduced ERK-signaling might be responsible for reduced CD69 and IL-2 up-regulation after ONX 0914 treatment. However, this assumption is clearly in line with previous reports showing that impairment of sustained (without affecting early) ERK signaling in T cells led to reduced IL-2 secretion and CD69 up- regulation [577, 624]. Minor differences in ERK signaling peak and sustainment were reported before to have functional influences on TCR thresholds and to occur in pathophysiological settings [722]. Taken together, the reduction in ERK-phosphorylation sustainment of only about 20 % after ONX 0914 treatment as compared to DMSO-treated cells is nevertheless likely to be of functional relevance during T cell activation. Therefore, this work focused on investigating possible mechanisms leading to reduced ERK- signaling after ONX 0914 treatment. The importance of modulations in ERK signaling in T cells was recently further supported in a study by Allison et al. who not only provided evidence that the ERK pathway is involved in signaling TCR strength quantitatively, but also showed that scaled inhibition of

123

Chapter IV– Discussion p-ERK levels with MEK inhibitors showed correlated reduction in p-ERK levels and reduced activation of AP-1 responsive genes including CD69 [600]. Furthermore, ERK-signaling strength and duration have been previously linked to differential fate decisions in T cell polarization [422] (see also section 2.3.5).

In search for candidate pathways involved in reduced ERK-signaling sustainment it was observed that the upstream kinase MEK1/2 was unaffected by ONX 0914 in expanded T cells (Figure 17). Additionally, cycloheximide treatment experiments showed that reduced ERK signaling sustainment was depending on de novo protein synthesis (Figure 22). Together, this rendered direct negative feedback regulation by de novo expressed phosphatases the most likely candidate mechanism for further investigation. Of note, a possible Bam32-PLCγ1-PAK1-mediated phosphorylation of MEK at Ser298 has not been investigated here and cannot be ruled out to be affected (section 2.3.5). Nevertheless, based on the results presented above several DUSPs selective for ERK and/or other MAPKs were screened at protein level for potential effects by ONX 0914 (see also [707]). Most analyzed DUSPs were unaffected by ONX 0914 treatment. For example, the p38 and JNK selective DUSP10 showed consistent signal intensity over 5 h in activated naïve T cells with or without ONX 0914 treatment (Figure 22D), which is in line with the data by Zhang et al. that DUSP10 is constitutively expressed in naïve T cells and down-regulated after 24 h of activation [660]. In contrast, DUSP9 was not tested during T cell activation before and the results indicated its expression in expanded T cells and in T1 cells, but no differential regulation after proteasome inhibition (Appendix Figure 5, see also [707]). Other DUSPs like for example DUSP14 remain to be tested. DUSP14 was described as JNK and ERK specific and to also target TAB1. However, its localization in both nucleus and cytosol renders it a potential candidate for nuclear ERK de-phosphorylation. As reduced MAPK phosphorylation after ONX 0914 treatment was only detected for ERK in this study, it appears logical that a phosphatase would be involved which is highly specific for ERK as reported for DUSP6 and DUSP5. However, the characterization of substrate selectivity of different DUSPs was pre-dominantly performed with in vitro purified enzymes (see references in section 2.4.2). Phase separation, multi- component complex formation and spatial regulation might, however, have an influence on substrate selectivity in cellulo. I.e.: In spite of the principle ability to de-phosphorylate several MAPK substrates in vitro, additional regulation might result in more specific substrate binding in the cellular context. Such additional regulation of substrate selectivity could potentially explain contradictive results about substrate specificity as for example in the case of DUSP1 or DUSP8 (section 2.4.2). Therefore, also DUSPs that were not reported as exclusively ERK-specific were not omitted as potential candidate DUSPs to mediate negative feedback for ERK after ONX 0914 treatment. Nevertheless, the one candidate DUSP found to accumulate in ONX 0914 treated cells in this work was indeed the reportedly ERK-specific cytosolic DUSP6. In contrast, the nuclear DUSP5 was less expressed upon T cell activation in ONX 0914 treated cells.

124

Chapter IV– Discussion

Notably, an effect on JNK phosphorylation by ONX 0914 after prolonged activation was indicated in expanded T cells. This study mainly focused on pathways that could likely be responsible for the observed amelioration of naïve T cell activation within the first 5 h. Hence, the reportedly minor relevance of JNK in naïve T cell activation (section 2.3.5) and failure to reliably detect JNK phosphorylation in naïve T cells were the reasons not to pursue analysis of JNK signaling further in this work. Given that induction of mild proteostasis stress by ONX 0914 was detected consistently in naïve T cells, expanded T cells, T1 cells and B cells, it should be taken into account that JNK phosphorylation is described as a signaling event in response to stress, in particular the unfolded protein response (section 2.3.5). JNK was also found to have a role in autophagy initiation in activated T cells, which could therefore contribute to proteostasis recovery in T cells with impaired proteasome capacity ([239, 723] and section 2.3.6). Therefore, further characterization of the JNK pathway after ONX 0914 treatment might be considered for future investigations as it might be relevant in the activation of effector cells, but evidence for its relevance in the observed amelioration of naïve T cell activation here could not be obtained so far.

The complex spatio-temporal dynamics of DUSP expression and degradation has rendered the view of DUSPs to function as signaling integrators. I.e. DUSPs could enable cells to converge signaling input from diverse stimuli on particular activation pathways, thereby modulating signaling outcome depending on multiple input cues [631]. Therefore, the correlation of reduced ERK phosphorylation sustainment and accumulation of DUSP6 gave rise to the hypothesis that DUSP6 could function as a signal integrator of reduced proteasome capacity towards TCR induced activation signaling. Such a mechanism could also make sense in the light of the fast DUSP6 turnover that was reported before [676] and also found by radioactive labeling in this work (Figure 23). Why would a cell increase a protein transcriptionally after T cell activation, but at the same time constantly promote its rapid degradation? A putative explanation would be a sensory function of the protein once its degradation is impaired. I.e. the cell would be able to sense overwhelmed proteasomes and provide negative feedback signaling to the stimulatory input driving protein neo-synthesis. In case of DUSP6, this mechanism would putatively dampen ERK-induced protein synthesis as a large fraction of ERK substrates in the nucleus are transcription factors [546]. Reducing protein synthesis would thus also alleviate synthesis-derived proteasome substrates and hence allow the cell to recover from proteasome overload. Such a hypothesis appears plausible because reducing overall protein synthesis is a well-described means of the integrated stress response via phosphorylated eIF2α (section 1.6). Therefore, a direct regulation of individual pathways in response to impaired proteasome function would add an additional layer of stress response potentially active already under milder stress conditions as induced by immunoproteasome selective inhibition, which did not induce enhanced eIF2α phosphorylation or ATF4 induction as opposed to MG-132 treatment (Figure 18, 17 and 18). Hence, DUSP6 was interpreted here as a potential candidate sensor of reduced proteasome capacity to negatively mediate ERK-signaling. Indeed, it was also demonstrated in this work that dysregulation of DUSP6 after 125

Chapter IV– Discussion

ONX 0914 treatment occurred at the level of DUSP6 degradation, which was partially impaired, but not fully abrogated (Figure 23). Transcriptional activity of p53 has also been reported to induce Dusp6 expression during senescence of rat renal tubular epithelial cells [724]. Here, T cells treated with MG-132 showed pronounced p53 accumulation (Figure 18), but Dusp6 mRNA transcripts were found to be reduced in MG-132 treated expanded T cells (Figure 22) and T1 cells (Figure 23). Hence, p53 was not a potent inducer of Dusp6 expression under the conditions investigated here. However, p53 needs to be downregulated for proper CD4+ T cell function [725, 726]. Thus, some accumulation of p53 in ONX 0914 treated naïve CD4+ T cells (Figure 22D) might yet contribute to impaired activation. It would be possible that NF-κB signaling drives Dusp6 expression. As mentioned above, NF-κB was found unimpaired by ONX 0914, while MG-132 is known to impair NF-κB activation (compare also Figure 16). Also, DUSP6 has been shown to be induced by PMA treatment in MCF-7 cells and both MEK inhibition as well as PKC inhibition independently abrogated Dusp6 expression, which indicates that DUSP6 might be synergistically induced by both, ERK and NF-κB signaling [673]. Hence, the impaired Dusp6 expression in MG-132 treated cells, but not significantly altered Dusp6 expression in expanded CD4+ T cells would be in line with the observation that NF-κB signaling was not affected by ONX 0914. Taken together, based on the numerous reports (section 2.4.3) and the data obtained in this work, a direct feedback via sensing of impaired DUSP6 degradation was proposed as a hypothetical regulatory mechanism explaining the observed reduction in ERK phosphorylation. Importantly, however, upon testing this hypothesis with T cells from DUSP6-deficient mice, no evidence could be obtained further supporting a non-redundant role of DUSP6 to mediate negative feedback on ERK signaling after ONX 0914-mediated proteasome impairment. It has to be noted that high redundancy is a reported hallmark of DUSPs [627, 631]. Therefore, compensation of DUSP6 by another, unidentified phosphatase cannot be fully ruled out. DUSP5 was not found to be differently regulated in DUSP6-deficient mice as indicated by preliminary results (data not shown). Based on the data obtained in this work, however, the most appropriate conclusion is to rule out DUSP6 as the responsible phosphatase involved in reduced ERK signaling after ONX 0914 treatment, thus, disconfirming the abovementioned hypothesis. Hence, a detailed mechanism explaining reduced ERK-signaling on the basis of immunoproteasome inhibition could not be unraveled in this work, but demands further investigation.

The question of how immunoproteasome inhibition impairs ERK signaling sustainment remains elusive. ERK signaling underlies spatial regulation as ERK can be targeted to various different localizations in the cell [547]. Taking steady state localization and re-localizing dynamics into account is important to understand ERK signaling in different cellular contexts and with different activating stimuli [547]. For example, ERK signaling in T cells is influenced by phase separation of clustered up-stream molecules [727] and the outcomes of ERK signaling were reported to be context-specific depending on the presence or absence of ERK modulators like paxillin or DUSP6 [678]. Hence, taking ERK localization into account 126

Chapter IV– Discussion might likewise be useful to identify the upstream effector molecule responsible for reduced ERK signaling after ONX 0914 treatment. An analysis of nuclear translocation of total and phosphorylated ERK in ONX 0914 treated T cells was therefore performed (see [707], data not shown here). The results corroborated the reduced ERK phosphorylation after ONX 0914 treatment, which was detectable in nuclear areas of activated T cells. However, no altered translocation of total ERK was detected [707]. In combination with the data obtained with DUSP6-deficient T cells, these results indicate that a nuclear phosphatase would be a more likely candidate responsible for reduced ERK signaling sustainment after ONX 0914 treatment. So far, tested nuclear phosphatases (e.g. DUSP1) were either not affected by immunoproteasome inhibition or were not successfully detectable with the antibodies used. Clearly, a candidate to be tested is DUSP2/PAC-1 as it was identified as an immediate early gene in TCR-induced activation, shows nuclear localization and shows selectivity for ERK and p38 [631, 638, 643]. However, no reliable signals could be obtained in immunoblots so far ([707] and data not shown). A more unbiased approach of searching for accumulating ERK regulatory proteins e.g. by stable isotope labeling of amino acids in cell culture (SILAC) and mass spectrometry (MS) might be useful to identify the negative feedback pathway responsible for reduced ERK signaling under the influence of immunoproteasome inhibition in T cells in the future. Interestingly, we also observed that LMP7-specific inhibition with PRN1126 induced DUSP6 accumulation to some extent ([707], data not shown here). However, PRN1126 alone is not sufficient for therapeutic efficacy in pre-clinical models of colitis and EAE (Basler et al., 2018, section 12.5), which would also be in line with the results disconfirming a functional involvement of DUSP6 in reduced T cell activation, albeit only indirectly. Of note, the marked accumulation of DUSP6 might still be to some extent functionally relevant in ONX 0914 treated T cells, the outcome of which might not be identified in the assays used here. However, the results of this work indicate that such a functional role would likely be ERK-independent. Notably, interaction partners of DUSP6 other than ERK have been identified [676, 728]. Most recently, it was reported that DUSP6 can regulate intracellular adhesion molecule 1 (ICAM-1) expression in endothelial cells independently of ERK-signaling by promoting NF-κB signaling via an as yet undefined mechanism [729, 730]. However, if further functional consequences of DUSP6 accumulation upon ONX 0914 treatment exist in activated T cells remains speculative at this point and has to be addressed experimentally in the future. Immunoprecipitation of DUSP6 from ONX 0914 treated T cells and analysis of possible interaction partners by mass spectrometry might identify novel mechanistic hypotheses to investigate in the future.

127

Chapter IV– Discussion

10.3 Immunoproteasome inhibition and its effects on proteostasis in lymphocyte activation

Apart from the influence of ONX 0914 treatment on TCR-induced signal transduction as discussed above, this study also re-investigated the effect of ONX 0914 inhibition on overall cellular proteostasis. When ONX 0914 was originally described in 2009, the possibility that immunoproteasome-selective inhibition could result in ubiquitin-conjugate accumulation was tested experimentally in the Molt4 T cell-derived cell line [199]. However, no appearance of ubiquitin-conjugates and no p53 accumulation were detected with ONX 0914 concentrations in the reportedly selective range up to 300 nM and consistent data was obtained in this work (Figure 18). It was hence thought that LMP7-selective inhibition would not be sufficient to result in marked impairment of overall proteolytic function. Likewise, no effect on NF-κB signaling was detected by Muchamuel et al. in an NF-κB reporter assay [199]. While NF-κB signaling was also found to be unaffected in this work (as discussed above), re-investigation of primary naïve or expanded murine T cells as well as human T cells revealed that ONX 0914 potently induced accumulation of ubiquitin- conjugates in these cells upon activation (Figure 18). This observation might be explained by both the unexpectedly high immunoproteasome content detected in primary T cells (Figure 10) and the activation induced metabolic demands accompanying T cell activation [375, 585]. In line with this notion, Rouette et al. showed that THP-1 cells which almost exclusively expressed LMP7-containing, but not β5c-containing proteasomes, were susceptible to ubiquitin-conjugate accumulation and apoptosis induction after ONX 0914 treatment [731]. Sung Yun et al. recently reported enhanced formation of mature immunoproteasomes via PRAS40 in MEFs. Enhanced immunoproteasome content rendered the cells susceptible to poly-ubiquitin-conjugate accumulation after ONX 0914 treatment [175]. MEFs stimulated with IFN-γ in order to induce high immunoproteasome content in this work were also susceptible to ubiquitin-conjugate accumulation (Figure 21). Hence, the effect of ONX 0914 treatment seems to be predominantly depending on quantitative abundance of immunoproteasomes in cells. Furthermore, the finding that ONX 0914 inhibits both LMP7 and LMP2, which impairs T cell activation synergistically, (Figure 9) shapes the rationale to explain the influence of ONX 0914 on total proteostasis, also in naïve T and B cells.

When cells were left unstimulated after pulse-treatment with ONX 0914, less pronounced appearance of ubiquitin-conjugates was observed (Figure 18). This indicated that activation induced proteostasis was more affected than steady state proteostasis in resting expanded T cells, likely attributed to enhanced de novo protein synthesis after activation. Indeed, treatment of cells with cycloheximide abrogated accumulation of ubiquitin-conjugates in ONX 0914 treated cells, but still resulted in ubiquitin-conjugate accumulation in MG-132 treated cells (Figure 22). This supported the view that the bulk of proteostasis stress in ONX 0914 treated cells was derived from de novo synthesis and that ONX 0914 did not fully 128

Chapter IV– Discussion block protein degradation by the proteasome. The results obtained using cycloheximide also pointed to a possible link between impaired proteasome capacity and the reduction in ERK signaling sustainment and has therefore initially promoted the hypothesis of a functional involvement of DUSP6 as discussed above. While this hypothesis was disconfirmed in this work, the finding that immunoproteasome inhibition impairs proteostasis leading to ubiquitin-conjugate accumulation and impaired degradation of DUSP6 gives rise to shortly discuss the involvement of ubiquitination and proteasomal degradation in T cells in general and further strategies to investigate if a direct link between impaired proteasome capacity and ERK-signaling and/or general T cell activation exists. Afterwards, the characterized effects of ONX 0914 treatment on prolonged activation and cell viability including the apparent role of Nrf1 will be discussed.

Ubiquitination in T cell activation and polarization signaling involves both, modulatory ubiquitination, but also degradation-priming ubiquitination. For example, non-degrading ubiquitination of ZAP70 promotes its association with phosphatases negatively regulating ZAP70 function, which is counteracted by the deubiquitinating enzyme (DUB) Otud7b [732]. In contrast, the ubiquitin E3 ligase SLIM facilitates STAT1 and STAT4 ubiquitination and degradation and SLIM-deficiency resulted in enhanced IFN-γ production and increased STAT4-phosphorylation in response to IL-12 treatment [733]. Similarly, Smurf1 is an E3 ligase promoting K48-ubiquitination and degradation of STAT1 and Smurf1 is expressed by IFN-γ signaling itself, hence acting as a negative feedback protein [734]. These are only examples of the manifold involvements of E3 ligases in regulating transcription factors important for T cell signaling and polarization that have been identified within recent years [735]. Clearly, targeted degradation of selected signaling molecules is a highly important regulatory mechanism in T cell biology. The complexity of regulation beyond this posttranslational level is further enhanced by regulation at the posttranscriptional level, involving microRNAs and RNA binding proteins (RBPs), which are in part themselves regulated via ubiquitination and proteasomal degradation [680, 735, 736]. Therefore, the finding of ubiquitin-conjugate accumulation in activated T cells as a result of immunoproteasome inhibition has wider implications than only addressing possibly affected ubiquitin-stress response pathways. Even though no functional involvement of DUSP6 in reducing ERK phosphorylation sustainment could be substantiated in this work, the dysregulated degradation of DUSP6 shows that individual proteins regulated by the proteasome can be markedly affected, while at the same time no evidence for impaired NF-κB signaling could be obtained, which also depends on proteasomal degradation. Importantly, broad-spectrum proteasome inhibition and immunoproteasome inhibition affected DUSP6 regulation differently in that ONX 0914 predominantly impaired degradation, while MG-132 also markedly impaired DUSP6 expression. With respect to future investigations it has to be acknowledged that identifying altered pathways after immunoproteasome inhibition therefore must take into account the necessary focus towards the level of protein regulation. It is likely that DUSP6 would not have been identified as affected by ONX 0914 in a microarray testing differential mRNA expression after ONX 0914 treatment. Notably, it is very difficult to predict which 129

Chapter IV– Discussion proteins might be affected by immunoproteasome inhibition since an impact on targeted degradation depends on multiple factors influenced by signaling cues including i) the expression and activity of respective E3 ligases, ii) the induced expression and further modification of the target for degradation, iii) the expression and activity of counterbalancing DUBs and iv) the kinetic turnover rates of the degraded protein.

To further investigate how ONX 0914 treatment shapes T cell activation and polarization signaling it is therefore suggested to make use of methods to identify differentially regulated factors at a broader scale like mass spectrometry approaches. However, it has to be taken into account that different abundancies of proteins after ONX 0914 treatment might result from either altered degradation or altered expression. Hence, effects are ideally measured after short-term treatment. Problematically, if partial impairment of proteasome capacity after ONX 0914 treatment slowly induces accumulation of ubiquitin-conjugates while degradation is not fully blocked (as shown here for DUSP6), it is difficult to predict the optimal time window for analysis. Based on the ubiquitin-conjugate profiles in primary naïve T cells, 2 h to 6 h after ONX 0914 treatment and activation might be appropriate, which is already long enough to allow possible differential expression of protein-coding genes instead of reflecting their differential degradation. Moreover, protein quantification with mass-spectrometry based quantification methods like SILAC is difficult in non-dividing primary cells like CD4+ T cells isolated from spleen. Detection of differential protein degradation in freshly isolated T cells might require optimized peptide-ion intensity quantification as recently introduces by Mathieson et al. [98]. However, this method could still be unable to detect alterations of proteins that are long-lived in naïve T cells, but dynamically regulated upon activation, as short-term isotope-labeling will not cover long-lived proteins [98].

Expanded T cells and the T1 cell line might be suitable for a SILAC approach to identify different protein quantities in cells after ONX 0914 treatment. To overcome the abovementioned obstacles when using SILAC as an approach to identify differential degradation, enhancement of hits selectively affected by proteasomal degradation can be achieved using diGly-remnant-detecting antibodies after trypsin-digestions as shown by Kim et al. [737]. In HCT116 cells, the authors identified enriched peptides that were targets of ubiquitination in a SILAC approach. In order to distinguish between K63-linked non-degrading ubiquitination and K11 or K48 linked degrading ubiquitination, they compared heavy isotope-labeled cells after 2 h, 4 h or 8 h of bortezomib treatment with light isotope-labelled control cells without bortezomib treatment. Depending on the kinetics of peptide detection they grouped targets showing altered degradation into different categories. These included early strongly accumulating, early weakly accumulating, accumulating with intermediate kinetics, late accumulating, unchanged and downregulated peptides. Interestingly, among the hits of fast accumulating protein peptides were sequences derived from PSMB6, indicating a fast turnover of β1c (supplementary material to reference [737]). Among the DUSP family

130

Chapter IV– Discussion proteins bortezomib treatment resulted in intermediately fast accumulation of DUSP1 and late accumulation of DUSP2, while DUSP14 was unchanged. Interestingly, Nrf2 was shown to be a late accumulating target of low abundance (supplementary material to reference [737]), a notion that might theoretically apply to primary T cells as well as, but Nrf2 was not readily detectable in immunoblots in this work (data not shown). A similar approach to identify accumulating proteins in ONX 0914 treated cells after TCR stimulation might be useful as a future effort in immunoproteasome research. Especially, a combination of either bortezomib or MG-132 treated cells and ONX 0914 treated cells in such an analysis might reveal novel target proteins differentially regulated after broad-spectrum proteasome inhibition as compared to immunoproteasome inhibition. The results presented here render DUSP6 as a control target protein that should be expected as a hit in such an analysis. However, it cannot be ruled out that also ubiquitin-independent degradation occurs in activated T cells and could also be affected by immunoproteasome inhibition, but would not be covered by analysis using diGly-remnant-detecting antibodies.

One approach that has been designed before and tested by Sarah Mundt in J774 macrophages is terminal amine isotope labeling of substrates (TAILS) ([738], data unpublished). TAILS generates data showing quantitative differences in produced novel N-termini, e.g. as a result of proteasomal cleavage. However, different quantitative cleavage site occurrence does not allow concluding if the reduced generation of a particular N-terminus-containing peptide from a specific protein is due to impaired overall degradation of that protein or due to different cleavage site usage during the degradation, leading to alternative N-termini. Four out of 89 candidates to be differentially degraded after ONX 0914 treatment from the TAILS measurement were tested in immunoblots, but none could be confirmed to be differentially degraded or processed ([738], data unpublished). Notably, TAILS might allow detecting two different impacts of altered proteasomal cleavage, which is both impaired degradation of a protein as well as altered generation of the proteasome-produced peptides in general. Proteasome-dependent peptides are known to affect MHC-I peptide presentation, but it is unclear, how this could influence CD4+ T cell activation via invariant-CD3/CD28-targeting activating antibodies, unless an unknown mechanism exists, by which MHC-I abundance on the CD4+ T cell surface would affect CD4+ T cell activation upon anti-CD3/CD28 stimulation. The observed ameliorated T cell activation presented here cannot result merely from reduced MHC-I surface abundance, as LMP7-/- cells have intrinsically less MHC-I surface abundance [50, 158] and CD4+ T cells form LMP7-deficient mice did not show impaired activation as compared to ONX 0914 treated WT cells. The same notion also renders the hypothesis unlikely that a differently generated proteasomal peptide product in ONX 0914 treated or LMP7-/- cells might have a regulatory function within a signaling cascade. The principal possibility to interfere with signaling cascades like ERK or NF-κB with small peptides derived from endogenous protein sequences has been demonstrated [739, 740], but whether endogenous production of such peptides by the proteasome and subsequent action as inhibitory peptides 131

Chapter IV– Discussion occurs cell-intrinsically has not been shown to my knowledge. It seems unlikely, that in the presence of peptidases degrading the majority of proteasome-derived peptides within seconds [28], endogenous peptide-based auto-inhibition occurs; especially as sufficient amounts of these peptides would have to be derived from particular proteins like MEK or p65, which then would have to be degraded in great amounts. Hence, the hypothesis of regulatory peptides generated by proteasomes as proposed in my master’s thesis 2013 [702] is disfavored compared to the hypothesis that impaired degradation of an ERK-regulatory protein might underlie altered TCR-induced ERK signaling as implicated, but not ultimately corroborated by this study. Furthermore, the novel finding that therapeutic efficacy of ONX 0914 depends on dual LMP2/LMP7 inhibition further disfavors the former hypothesis of a selectively LMP7-processed factor as also discussed by Basler et al. (2018, section 12.5). Taken together, TAILS analysis to determine novel factors involved in ONX 0914 treated cells, which was conceived to identify altered peptide products, might be less useful to address differentially degraded proteins than quantifying peptide ion amounts from whole proteins in mass spectrometry.

10.4 Effects of mild proteostasis stress on T cell and B cell survival

The appearance of accumulating ubiquitin-conjugates in ONX 0914 treated cells in the absence of integrated stress response markers posed the question how prolonged activation of T cells after ONX 0914 treatment would affect T cell survival. Even after 20 h of activation no apparent reduction of cell viability and no enhanced PARP cleavage were found in ONX 0914 treated T cells as compared to DMSO treated T cells. While showing no difference between ONX 0914 treatment and DMSO control treatment, expanded T cells showed a generally enhanced PARP cleavage after activation as compared to naïve T cells (Figure 22F). This is likely attributed to the known sensitization of ex vivo expanded T cells towards AICD [256, 741]. A recently published study by Santos et al. reported that a novel LMP7 inhibitor (PKS21221) induced cell death in proliferating CD4+ T cells, but not in non-proliferating CD4+ T cells [295]. These data are consistent with our observation that pulse-treated naïve CD4+ T cells do not show enhanced cell death after 20 h of activation, a time window, in which naïve T cells do not yet proliferate. However, if reduced proliferation as observed here was caused by increased cell death during prolonged ONX 0914 treatment was not further tested. Furthermore, ubiquitin-conjugates were either alleviated or even reduced after 20 h in ONX 0914 treated cells compared to DMSO treated cells (Figure 19). To assess whether this alleviation was mediated by enhanced de novo expression of new immunoproteasomes it was made use of the observation that the irreversible covalent modification of LMP7 subunits was detectable via a shift in electrophoretic mobility. As ONX 0914 was only pulse-treated for 2 h before activation and was not present during activation, any newly synthesized and incorporated LMP7 would be expected to be detected at the same molecular weight as unmodified LMP7 proteins in DMSO treated samples. However,

132

Chapter IV– Discussion after 20 h of activation, almost all LMP7 protein remained electrophoretically shifted, indicating that no new LMP7 containing proteasomes were formed (Figure 19). Also the near-IR-immunoblot based quantification of total LMP7 (including the size for shifted and non-shifted LMP7) indicated no enhanced protein content of LMP7 after T cell activation in any of the treatment groups. This data is in clear contrast to a recent study by Sula Karreci et al. [293]. Using intracellular flow cytometry they reported a significant up-regulation of LMP7 content in both murine CD4+ and CD8+ T cells already at 3 h after in vitro activation with anti-CD3/CD28 antibodies [293]. Unfortunately, the authors neither presented secondary- antibody-only controls, isotype controls nor control stainings in LMP7-deficient cells to support the specificity of their antibody-staining. Hence, it remains unclear, if the reported antibody indeed specifically detected LMP7 protein [293]. In this work, in contrast to an up-regulation of LMP7 in activated T cells, an up-regulation of β5c was detected at protein level using quantitative immunoblotting. However, up- regulation of β5c seems to be only transient while later also expression of new LMP7-containing proteasomes occurs as indicated by appearance of unmodified LMP7 protein after 28 h – 30 h in immunoblots of ONX 0914 treated cells (data not shown). Also Griffin et al. used T cell blasts after 72 h ConA stimulation to investigate immunoproteasome assembly depending on the active de novo synthesis of LMP7 in thus activated T cells [89]. Furthermore, expanded T cells after one week of cultivation in IL-2 containing medium showed similar LMP7 and β5c content as compared to naïve cells, substantiating that β5c up-regulation was transient (Figure 15). Moreover, immunoblots with lysates of CD4+ T cells from LCMV-infected SMARTA mice (TCR-transgenic for an MHC-II-restricted LCMV epitope) were performed. These results also indicate an up-regulation of β5c at protein level in antigen-specific T cells after infection, but this did not seem to be promoted by ONX 0914. However, while the data obtained for in vivo activated T cells in principle confirms that ONX 0914 attenuates antigen-specific T cell activation also under in vivo conditions, the amount of data is so far insufficient to conclude about mechanistic similarities and differences of in vivo versus in vitro activated T cells. As one early time point (6 h after LCMV-infection; no or only minute T cell activation measurable yet, data not shown) and one later time point (18 after LCMV-infection; fully activated CD4+ T cell compartment, Figure 25) did not allow to address the question sufficiently at the mechanistic level, this issue needs to be further investigated in future experiments. So far, the insight into mechanisms of ONX 0914 treatment after in vivo activation therefore remains preliminary in contrast the more extensively characterized effect in vitro in this study. Nevertheless, taken together, the results in this work indicate that the precise regulation of LMP7- containing proteasomes or standard proteasomes in T cells should be investigated in greater detail. It might be interesting to test, if T cell activation in the presence of high IFN-γ concentrations in vitro shows different molecular characteristics compared to the data presented here. Potentially, IFN-γ treatment during T cell activation could promote earlier immunoproteaseome up-regulation with a possible influence on the recovery of ONX 0914 pulse-treated cells from the imposed proteostasis stress. The results could give

133

Chapter IV– Discussion hints to pathways that might differentially affect the consequence of ONX 0914 treatment on T cell activation under different pathophysiological settings.

In line with the observations in this study, Frisan et al. reported that human B cells contain high IP content and that activation did not correlate with further increase of IP content [703]. Importantly, B cells showed enhanced PARP cleavage after ONX 0914 treatment. In general, isolated murine B cells die rapidly in culture in the absence of a stimulus and the requirement of tonic BCR signaling is also relevant in vivo [742]. Accordingly, when B cells were left unstimulated for 5 h in culture, PARP cleavage was markedly enhanced as compared to PMA/ionomycin stimulated B cells. However, ONX 0914 treatment showed relatively enhanced PARP cleavage as compared to DMSO treatment in both PMA/ionomycin stimulated as well as anti-CD40 and F(ab’)2 anti-immunoglobulin stimulated B cells. Interestingly, in spite of modification of β5c in LMP7-deficient B cells as apparent in immunoblots, no effect of ONX 0914 on ubiquitin-conjugate accumulation and PARP cleavage were detected, which again emphasizes immunoproteasome dependency of the ONX 0914-mediated effects. The apparent higher susceptibility of B cells to apoptosis after ONX 0914 treatment might provide an explanation for the high efficacy of ONX 0914 to ameliorate autoimmunity in lupus-prone mice [363]. Furthermore, Li et al. recently showed that ONX 0914 treatment was as effective as bortezomib in preventing long-term allograft rejection after kidney transplantation [361]. As plasma cells are highly susceptible to proteasome inhibition [392, 393], it is likely that immunoproteasome inhibition therefore also imposes toxic proteostasis stress on plasma cells. However, if plasma cells are equally dependent on LMP7-containing proteasomes like naïve B cells remains to be determined.

It is an especially striking question why naïve T cells and B cells contain almost exclusively LMP7- containing proteasomes at steady state. And why would the cells then up-regulate standard proteasomes instead of immunoproteasomes during early T cell activation, an energy demanding process accompanied by higher proteostatic demands? This observation is in clear contrast to the concept that LMP7-containing proteasomes would be superior over standard proteaosmes in clearing ubiquitin-conjugates [73]. In line with data by Hewing et al. [79] and Nathan et al. [78] this work also contributes to the conclusion that immunoproteasomes are not superior over standard proteasomes in maintaining protein homeostasis, as LMP7-deficiency had no influence on T cell activation as compared to WT cells. It might also be noteworthy to emphasize that the hallmark studies identifying the Nrf1-mediated bounce-back induction of proteasome expression not only provided data showing that standard proteasomes are up-regulated, but also contain data showing that immunoproteasome subunits are instead down-regulated [214, 216]. While it is tempting to think that if immunoproteasomes are superior for ubiquitin-conjugate clearance it would not make sense that cells specifically up-regulate standard proteasome subunits instead of immunoproteasome when facing enhanced proteostasis stress, it has to be kept in mind that

134

Chapter IV– Discussion pharmacological immunoproteasome inhibition is an artificial situation not suitable to speculate about an evolutionary development of proteasome subunit regulation under stress conditions. However, the possibility that some selection pressure during evolution of the immune system has shaped the high dependency on LMP7-containing proteasomes in lymphocytes for an as yet to be unraveled reason, should be investigated in future research. Therefore, this paragraph will end with some speculation about possible reasons. It is interesting to note that lymphocytes were reported to be especially rich in MHC-I surface expression. In T cells this is driven by what has been referred to as the T cell enhanceosome including the RUNX1 transcription factor that is even boosted in regulatory T cells by Foxp3 [743, 744]. It is intriguing to think that a connection between this notion and the observations in this work might exist – in that the high expression of LMP7 is the pre-requisite to ensure high MHC-I on the surface of B cells and T cells. While β5c is barely detected at all in naïve WT T cells and B cells compared to LMP7-deficient cells, the c-subunits of β1 and β2 are similarly detectable between LMP7-deficient and WT lymphocytes. MHC-I is decreased by ~50% in LMP7-deficient, but not in LMP2- or MECL-1 deficient cells [50]. Also Basler and colleagues (2018, section 12.5) recently showed that LMP7-specific inhibition with PRN1126 alone is insufficient to reduce MHC-I expression on the surface of splenocytes as opposed to ONX 0914. Thus, immunoproteasome subunits together contribute to MHC-I surface expression levels. Therefore, it would be elegantly explained why naïve T and B cells almost exclusively contain LMP7-containing and almost no β5c-containing proteasomes, if the amount of MHC-I on the surface of T cells and B cells had an important function beyond antigen-presentation to CTLs that was not yet determined. Is it for example possible that high density of MHC-I on the surface might protect CD4+ T cells from NK-cell mediated bystander cell killing in an inflammatory environment? Importantly, however, the observation that MECL-1-deficient and LMP2-deficient CD8+ T cells also fail to survive in an LCMV-infected host [137] argues against the idea that the extent of MHC-I surface expression plays a major role for survival of T cells in an inflammatory environment. Seifert et al. [73] reported that LMP7 containing proteasomes were superior in clearing ubiquitin-conjugates and oxidatively damaged protein. While the former role of immunoproteasomes for ubiquitin-conjugate clearance was challenged by Nathan et al. the role of LMP7- containing proteasomes for degradation of oxidized proteins was not addressed [78]. However, if T cell activation in vivo in the presence of pro-inflammatory cytokines involved elevated levels of oxidatively damaged proteasomes, a high a priori expression of LMP7-containing proteasomes in lymphocytes could be beneficial. Clearly, however, in vitro activation of CD4+ T cells does not require LMP7 as shown in this work. An interesting aspect to consider might be the notion that activated lymphoctes were reported to undergo the Warburg effect, i.e. anaerobic ATP production via glycolysis instead of oxidative phosphorylation in spite of sufficient oxygen. Therefore, the question could be asked if maybe activated lymphocytes even depend to considerable parts on ubiquitin-independent degradation, as less ATP would be required. However, these ideas will need to be addressed experimentally in the future.

135

Chapter IV– Discussion

10.5 Possible signaling pathways involved in proteasome regulation after immunoproteasome inhibition

The observed dynamic up-regulation of β5c at protein level after activation raised the question if previously described pathways of Psmb5 regulation might be involved. Interestingly, our understanding of homeostatic proteasome gene regulation is still limited and has remained an ongoing research effort of recent years [172–174, 705]. STAT3 signaling has been implicated in Psmb5 regulation as well as in Psmb8 regulation before [705, 706]. Zhang et al. reported in 2016 that tight junction protein 1 (TJP1) confered bortezomib sensitivity to multiple myeloma cells, because presence of TJP1 impaired epidermal growth factor receptor (EGFR) signaling [706]. The authors showed that EGF-signaling in the absence of TJP1 induced stronger JAK1 and STAT3 phosphorylation and resulted in higher LMP7 and LMP2 expression. RNA interference-mediated STAT3 suppression resulted in reduced LMP7 and LMP2 expression, indicating that the immunoproteasomes were regulated by STAT3. Interestingly, however, IL-6 treatment, which is known to signal via STAT3, had no effect on LMP7 and LMP2 expression or total proteasome activity [706]. This indicates that other pathways contributed synergistically to LMP7 and LMP2 up-regulation or that differences in spatiotemporal STAT3 regulation might alter signaling outcome. Interestingly, even though Vangala et al. already reported in 2014 that STAT3 was required in EGF- mediated up-regulation of standard proteasome in different cells lines [705], an influence of TJP1 on standard proteasome up-regulation was not investigated by Zhang et al. [706]. However, STAT3- phosphorylation plays differential roles in regulating Th17 fate as IL-6 induces STAT3-phosphorylation, but early strong STAT3-signaling in T cell high density cultures can result in Th17 impairment (section 2.2.1). Furthermore, reduced STAT3 phosphorylation after ONX 0914 treatment was reported by Kalim et al. [288] and Liu et al. [715], but could not be reproduced in our hands (see [745]) and the observation has so far remained mechanistically elusive. Appearance of autonomous STAT3-phosphorylation in stimulated T cells in vitro was observed in this work. However, no consistent effect of ONX 0914 treatment on autonomous STAT3 signaling was detected in immunoblots of murine T cells and clearly no difference was observed in human T cells (Appendix Figure 4). The contribution of STAT3 to Psmb5 expression was therefore not further focused in this work. In contrast, soluble Nrf1 was found to accumulate in ONX 0914 treated T cells and B cells and correlated with a significant increase of β5c up-regulation in ONX 0914 treated T cells (Figure 19). Nrf1 has been widely appreciated as a regulator of proteasome expression. In Nrf1-deficient MEFs steady state proteasome gene expression was unaltered, while cells were lacking the ability to up-regulate proteasome subunits in response to MG-132 treatment [215]. Similar results were obtained independently using siRNA against TCF11/Nrf1 [208]. These data did not indicate that Nrf1 controls proteasome genes under physiological conditions, but only in response to proteasome inhibition. In contrast, Lee et al. demonstrated a role in both, basal proteasome subunit

136

Chapter IV– Discussion expression and stress-induced proteasome subunit expression in conditional Nrf1-KO models in the brain and in the liver [227, 746]. Recently, Widenmaier et al. showed that Nrf1 plays a role in cellular cholesterol homeostasis. While low cholesterol is sensed via SREBP2, excess cholesterol is sensed via inhibition of Nrf1 processing and transcriptional activity. High cholesterol stabilizes full length Nrf1 and even impairs Nrf1 cleavage and transcriptional activity in response to bortezomib treatment [229]. Interestingly, cholesterol-sensing also seems to play a role for the generation of cleaved Nrf1, as an Nrf1 mutant deficient for the cholesterol binding region (ΔCRAC-HA) showed enhanced Nrf1 cleavage and resulted in enhanced Psmb5 gene expression, that was not impaired in response to high cholesterol levels [229]. Thus, high cholesterol levels in the ER can repress Nrf1-mediated Psmb5 expression. It should be noted that this aspect offers a possible link to the investigation of Psmb5 regulation in LMP7-deficient cells, which is discussed below.

Indirect evidence for a role of Nrf1 in T cells came from at least two studies. Tomlin et al. showed that N-glycanase 1 (NGLY1)-deficiency abrogated accumulation of processed Nrf1 and proteasome subunit up-regulation in carfilzomib treated MEFs due to impaired Nrf1-deglycosylation. They also showed that NGLY1 inhibition in T-ALL (T cell acute lymphoblastic leukemia) cells promoted carfilzomib-induced cell death [220]. Second, a recent study by Moskowitz et al. found that CD8+ T cells from elderly humans showed decreased accessibility of Nrf1 to promoter regions throughout the genome in comparison to CD8+ T cells from younger donors, implicating a role for Nrf1 in age-related changes of immune cell homeostasis [747]. However, it is unclear if Nrf1 plays a physiological role in T cells in general. While we found that Nrf1 accumulation occurs during activation of naïve CD4+ T cells in response to immunoproteasome inhibition, interestingly, no accumulation of Nrf1 was detected in MG-132 treated cells. This is likely attributed to the reported tendency of Nrf1 aggregation upon high extents of proteasome inhibition [223]. Concomitantly no up-regulation of β5c protein levels was detected in MG- 132 treated cells. Hence, not only with respect to the integrated stress response, but also with respect to Nrf1 accumulation, the molecular characteristics of strong broad-spectrum proteasome inhibition and immunoproteasome selective inhibition segregate in spite of dual LMP7/LMP2 inhibition by ONX 0914. These data again show that immunoproteasome inhibition and broad-spectrum proteasome inhibition affect T cell activation differently, as was also demonstrated for DUSP6 dysregulation as discussed above. Notably, no direct evidence is provided in this work that Nrf1 is indeed driving enhanced β5c up- regulation, and while this assumption is fully in line with previous reports [174, 208, 215, 216, 222], it could be further corroborated experimentally in future experiments (e.g. in T cells from Nrf1-deficient mice). Using STAT3 inhibitors it could be tested if autonomous STAT3 signaling contributes to β5c up- regulation in both DMSO and ONX 0914 treated cells, while Nrf1 might promote the additional increase of β5c in ONX 0914 treated cells. Notably, a recent study by Rousseau and Bertolotti showed the involvement of the atypical MAPK ERK5 in regulating proteasome expression [173], a pathway that was 137

Chapter IV– Discussion not investigated in T cells in this work. The possibility that insufficient amino acid supply from endogenous degraded proteins [171] in activated T cells after ONX 0914 treatment contributes to impaired activation remains to be investigated.

Overall evidence in this work suggested that standard proteasomes could functionally fully compensate for loss of immunoproteasomes during T cell activation. It is yet an unambiguous molecular difference between WT and LMP7-deficient cells that this compensation for loss of LMP7 by β5c occurs in CD4+ T cells and CD19+ B cells as detectable in immunoblots. Interestingly, however, to my knowledge it has not been addressed before which underlying molecular mechanism is responsible for compensatory up- regulation of β5c and if the underlying mechanism might give hints to alterations between LMP7-deficient and WT cells which could be also relevant in disease progression. In this context, it is noteworthy that De Verteuil et al. did not observe enhanced protein level intensities of β5c in bone-marrow derived DCs from LMP7/MECL-1-double-deficient mice as compared to WT mice, indicating that in cells containing both standard and immunoproteasomes, lack of the immunoproteasomes does not necessarily induce higher standard proteasome content [713]. In contrast, CD4+ T cells and CD19+ B cells were found to almost exclusively contain LMP7-incorporating proteasomes (Figure 10). This difference gave rise to the question, whether a special molecular mechanism has to be at play in T cells and B cells to ensure sufficient standard proteasome incorporation in the absence of immunoproteasomes due to LMP7- germline-deficiency. The discrepancy in the controversial findings about the influence of LMP7-deficiency on overall cellular proteostasis could not be fully clarified yet (section 1.5.2). A study by Heink et al. reporting that immunoproteasomes had shorter half-lives than standard proteasomes [91] implied the possibility that naïve T and B cells might therefore rely on more constant de novo formation of immunoproteasomes already at steady state, hence requiring additional β5c up-regulation in cells lacking LMP7. A recent study by St-Pierre et al. also reported elevated proteostasis stress in mTECs of LMP7/MECL-1-double-deficient mice [186]. Therefore, it could not be fully ruled that loss of LMP7 could result in some degree of low-level proteostasis stress, maybe even responsible for compensatory up- regulation of β5c. However, standard proteasome subunits were not or only insignificantly up-regulated at the transcriptional level in naïve T and B cells (Figure 11). Furthermore, even 9 to 20 h after T cell activation no up-regulation of newly synthesized LMP7 was detectable in ONX 0914 treated cells. In contrast, the standard proteasome subunit β5c increased over time (Figure 19). Also, no indication of enhanced proteostasis stress in activated LMP7-deficient T cells and B cells was observed in the experiments performed in this study. We also tested if any signs of low-level proteostasis stress including Nrf1 accumulation might be detectable in resting lymphocytes. No difference in levels of the tested stress markers between LMP7-deficient and WT cells were observed and Nrf1 was not detectable in naïve lymphocytes, but in control samples from ONX 0914 treated, activated B cells (Appendix Figure 6). How can the strong appearance of compensating β5c at protein level in LMP7-deficient cells be explained even 138

Chapter IV– Discussion though the mRNA level appears to be unaltered? The most plausible explanation seems to be that peripheral naïve T and B cells are equipped with proteasomes during their delevopment before they are released into the periphery. In general, proteasomes are relatively long-lived protein complexes (section 1.2.3). The data in this work does not indicate a fast turnover of immunoproteasomes in T cells as could be implied based on the results by Heink et al. [91]. In contrast, a recent study by Mathieson et al. provided evidence that immunoproteasome are in fact rather long-lived in primary lymphocytes [98]. Hence, a constant high-level expression of the proteasome genes in resting T cells and B cells would most likely not be required.

An alternative theoretical mechanistic explanation could be based on the reported mechanisms of proteasome core particle assembly as outlined in section 1.2.3. LMP7 is preferentially incorporated over β5c, indicating that in the presence of both, β5c and LMP7, predominantly LMP7 containing proteasomes are assembled. This process is further assisted by POMP and potentially by PRAS40 phosphorylation state [91, 175]. It cannot be ruled out that additional undefined mechanisms might exist in lymphocytes promoting the predominant assembly of LMP7 containing proteasomes over standard proteasomes. The mRNA level of Psmb5 is lower in lymphatic tissues than in other tissues as shown before [135, 136], a trend which was also apparent in the present study, although not with statistical significance (Figure 11). Even though LMP7 is incorporated as the last subunit of β-ring assembly during half-core particle maturation, the full 20S CP maturation largely depends on LMP7, which is the reason why LMP7-deficient cells also show strongly diminished incorporation of LMP2 and MECL-1 (section 1.2.3). In that case, low- level abundant β5c precursor protein would be likely rapidly degraded by existing immunoproteasomes, thus not being incorporated into 20S CPs. In contrast, if cells are lacking LMP7, the full maturation to 20S core particles is impaired, which might result in an equilibrium shift towards assembly of standard proteasomes in spite of non-enhanced expression levels and potentially low β5c protein abundance. The appearance of reduced signals for incorporated MECL-1 and LMP2 with some precursor accumulation at the same time would be in line with such a mechanism. However, in that case it might be expected that constant failure to assemble iCPs with a subsequent equilibrium shift towards standard proteasomes could result in some detectable effect on proteostasis. As mentioned above, however, this was not detected in resting lymphocytes in this work (Appendix Figure 6). Notably, when hypothesizing that compensatory up-regulation of standard proteasomes in LMP7-deficient cells might occur at the transcriptional level due to low-level stress, it has to be considered which effect size could be reasonably expected for such an up- regulation at the mRNA level. The hallmark studies showing increased Psmb-gene up-regulation in response to proteasome inhibition consistently report up-regulations at mRNA level in the range of 1.5 to 4-fold [174, 208, 214–216, 220]. Another example is a recent study by Lee et al. showing that the Nrf2 pathway can be promoted with a small molecule compound KMS99220 that binds to Keap1. In response, proteasome genes including both immuno- and standard subunits were increased at mRNA level and at 139

Chapter IV– Discussion protein level. However, a 2.5 fold up-regulation of Psmb5 at RNA level correlated with ~7-fold up- regulation of mature β5c protein levels, while at the same time Psmb8 was also present and up-regulated [748]. Taken together, even proteasome inhibition with bortezomib, epoxomycin or carfilzomib, which are potent inducers of proteotoxic stress, did not induce proteasome gene expression above 4-fold compared to untreated cells. Hence, it does not seem appropriate to assume that low-level stress in LMP7-deficient cells due to inferior proteostasis capacity of standard proteasomes or assembly-failures would potently induce Psmb5 gene expression to more significant extent. The compensatory appearance of β5c protein could in principle be regulated predominantly at the protein level in the absence of LMP7. This interpretation is also in line with early reports showing that IFN-γ exposure to cells led to down-regulation of standard proteasomes at protein level without reducing the transcriptional level of standard proteasome expression [142, 749, 750].

While the absence of enhanced Psmb5 gene expression in resting LMP7-deficient cells compared to WT cells might simply be explained by longevity of proteasome complexes, it remains to be investigated if under strong inflammatory conditions, normally promoting immunoproteasome expression, additional regulatory mechanisms at the transcriptional level would be required to ensure proper proteasome function in LMP7-deficient cells. An insufficient compensatory capacity of β5c in LMP7-deficient cells under such conditions could provide a possible explanation for the observation that co-transferring CD8+ T cells from WT and LMP7-deficient mice into an LCMV-infected host leads to selective loss of LMP7-deficient cells [137]. In contrast, during T cell activation performed in this study, no indication for an insufficient compensatory capacity of β5c in LMP7-deficient cells was found. However, T cell activation in vitro did not induce up-regulation of immunoproteasomes. This indicates that no mechanistic pressure towards LMP7 expression occurs during in vitro T cell activation, which might be different under the influence of other stimuli. This possibility should be further investigated in the future. Notably, the novel findings that Nrf1 is involved in cholesterol regulation also opens new hypothetical links to proteasome gene regulation in LMP7-deficient mice under high-fat diet as LMP7-deficiency was reported to protect from obesity associated inflammation [303]. When high cholesterol levels impair Psmb5 expression in response to proteostasis stress as indicated recently by Widenmaier et al. [229], LMP7-deficient inflammatory cells might have a selective disadvantage due to their dependency on standard proteasomes. These aspects of Nrf1 and Psmb5 regulation in LMP7-deficient cells remain speculative so far, as in this study no influence of LMP7-deficiency on Nrf1 could be detected at steady state. Yet, further analysis of Nrf1 regulation and Psmb5 regulation in LMP7-deficient mice under inflammatory or high-fat conditions might be a perspective for future experiments.

140

Chapter IV– Discussion

10.6 Concluding remarks

This work provides a rationale to shape a novel view on how immunoproteasome inhibition influences pro- inflammatory immune cell activation. The stringent comparison of immunoproteasome inhibition with ONX 0914 as compared to LMP7/immunoproteasome deficiency in this work strongly emphasizes the important difference of these two conditions. The specific LMP7-cleavage functions as determined by its substrate binding groove and the reported kinetic advantage over β5c were dispensable for all tested functions of CD4+ T cells in this work. In contrast, the LMP7- and LMP2-co-dependent, but immunoproteasome selective inhibition by ONX 0914 has shown marked ameliorative effects on T cell activation. The hypothesis that these effects might be attributed to impairing a selectively LMP7-dependent factor in the cells has therefore become rather unlikely, even though the existence of such a mechanism in other processes that are affected by LMP7-deficiency (e.g. in the colitis model) cannot be ruled out. With respect to the immunosuppressive effects of ONX 0914 mediated immunoproteasome inhibition this work clearly points into a different direction: Induction of mild proteostasis stress by impairing protein degradation during a process demanding high proteasome capacity seems to be involved in the ameliorating effects of immunoproteasome inhibition in T cells. It is somehow remarkable that this seemingly straightforward effect of impairing protein homeostasis was not detected before. However, at least two factors have to be taken into account as to why an impact on proteostasis has remained undetected before. First, the dual inhibition profile of therapeutically effective doses of ONX 0914 was not identified before and misleadingly the selectivity of ONX 0914 for only LMP7 does not imply a major effect on overall protein turnover, which was even tested in a human T cell line and supported the view that overall ubiquitin conjugate clearance was not affected by ONX 0914 treatment [199]. The development of LMP7-specific inhibitory small molecules including the recently characterized PRN-1126 as well as the LMP2-specific compound LU-001i has now allowed for a more detailed analysis of how particular subunit inhibitions are shaping the disease outcome in pre-clinical models of autoimmune diseases (Basler et al. 2018, section 12.5). Second, while it was known that lymphocytes do express immunoproteasomes at steady state, it was not appreciated before that primary T and B lymphocytes at the naïve state barely contain β5c containing standard proteasome. Thus, all proteasome dependent processes in these cells depend solely on immuno- and mixed proteasomes. Interestingly, the fact that NF-κB signaling was found to be not impaired by ONX 0914 as found in this work as well as by others [199, 200], emphasizes the potent degradation capacity of the residual non-inhibited proteasome subunits as long as the proteasome is not overwhelmed by ubiquitin-conjugated degradation substrates. Also DUSP6, which accumulated in the presence of cycloheximide in MG-132 treated cells was only affected by ONX 0914 when de novo protein synthesis allowed DUSP6 expression and ubiquitin-conjugate formation. With more and more subunit-selective inhibitors being developed the results of this work implicate a potential use of

141

Chapter IV– Discussion different subunit selective inhibitors to fine tune proteostasis in cells depending on the pathophysiological involvement of different immune or non-immune cells and their respective proteasome subunit compositions. In line with this study, the different impact of ONX 0914 mediated immunoproteasome inhibition on ubiquitin-conjugate clearance in cells expressing high immunoproteasome content versus cells expressing low immunoproteasome content was recently demonstrated in cancer cell lines where ONX 0914 resulted in ubiquitin-conjugate accumulation in THP1 cells, but not NB4 cells [731]. The results provided in this work support the view that the degree of immunoproteasome content significantly shapes the effect of immunoproteasome inhibition in non-malignant primary T and B lymphocytes. It is therefore a perspective towards future research that characterization of proteasome subunit compositions in individual cell types (e.g. Th1, Th2, Th17 and Treg cells as well as effector and memory cells of both the CD4+ and CD8+ T cell compartments) and even depending on their activation and polarization states should be performed to understand how subunit-selective inhibitors could be used to impair particular cell subtypes involved in disease pathophysiology with the highest possible selectivity. Furthermore, it remains to be investigated which the precise mechanism is leading to the impairment of ERK-phosphorylation sustainment as reported in this work. Taken together, this work provides novel mechanistic insight into the effects of immunoproteasome inhibition in primary lymphocytes during their activation, which will hopefully contribute to a better understanding of immunoproteasome inhibitors as future therapeutic measures for application in human patients.

142

References

11 REFERENCES

1. Sontag, E. M., W. I. M. Vonk, and J. Frydman. 2014. Sorting out the trash: The spatial nature of eukaryotic protein quality control. Curr. Opin. Cell Biol. 26: 139–146. 2. Ciechanover, a. 1994. The ubiquitin-proteasome proteolytic pathway. Cell 79: 13–21. 3. Ciechanover, A., and A. Stanhill. 2013. The complexity of recognition of ubiquitinated substrates by the 26S proteasome. Biochim. Biophys. Acta . 4. Finley, D., and V. Chau. 1991. Ubiquitination. Annu. Rev. Cell Biol. 7: 25–69. 5. Hershko, A., A. Ciechanover, and I. A. Rose. 1981. Identification of the active amino acid residue of the polypeptide of ATP-dependent protein breakdown. J. Biol. Chem. 256: 1525–1528. 6. McDowell, G. S., and A. Philpott. 2013. Non-canonical ubiquitylation: Mechanisms and consequences. Int. J. Biochem. Cell Biol. 45: 1833– 1842. 7. Metzger, M. B., V. A. Hristova, and A. M. Weissman. 2012. HECT and RING finger families of E3 ubiquitin ligases at a glance. J. Cell Sci. 125: 531–537. 8. Ciechanover, A. 2013. Intracellular protein degradation: from a vague idea through the lysosome and the ubiquitin-proteasome system and onto human diseases and drug targeting. Bioorg. Med. Chem. 21: 3400–10. 9. Thrower, J. S., L. Hoffman, M. Rechsteiner, and C. M. Pickart. 2000. Recognition of the polyubiquitin proteolytic signal. EMBO J. 19: 94–102. 10. Komander, D., and M. Rape. 2012. The Ubiquitin Code. Annu. Rev. Biochem. 81: 203–229. 11. Akutsu, M., I. Dikic, and A. Bremm. 2016. Ubiquitin chain diversity at a glance. J. Cell Sci. 129: 875–880. 12. Grice, G. L., and J. A. Nathan. 2016. The recognition of ubiquitinated proteins by the proteasome. Cell. Mol. Life Sci. 73: 3497–3506. 13. Tokunaga, F., S. Sakata, Y. Saeki, Y. Satomi, T. Kirisako, K. Kamei, T. Nakagawa, M. Kato, S. Murata, S. Yamaoka, M. Yamamoto, S. Akira, T. Takao, K. Tanaka, and K. Iwai. 2009. Involvement of linear polyubiquitylation of NEMO in NF-κB activation. Nat. Cell Biol. 11: 123–132. 14. Chen, Z. J. 2005. Ubiquitin signalling in the NF-kappaB pathway. Nat. Cell Biol. 7: 758–65. 15. Hofmann, R. M., and C. M. Pickart. 1999. Noncanonical MMS2-encoded ubiquitin-conjugating enzyme functions in assembly of novel polyubiquitin chains for DNA repair. Cell 96: 645–53. 16. Galan, J. M., and R. Haguenauer-Tsapis. 1997. Ubiquitin Lys63 is involved in ubiquitination and endocytosis of a yeast plasma membrane protein. EMBO J. 16: 5847–5854. 17. Kahana, C., G. Asher, and Y. Shaul. 2005. Mechanisms of protein degradation: an odyssey with ODC. Cell Cycle 4: 1461–4. 18. Asher, G., P. Tsvetkov, C. Kahana, and Y. Shaul. 2005. A mechanism of ubiquitin-independent proteasomal degradation of the tumor suppressors p53 and p73. Genes Dev. 19: 316–21. 19. Baugh, J. M., E. G. Viktorova, and E. V. Pilipenko. 2009. Proteasomes Can Degrade a Significant Proportion of Cellular Proteins Independent of Ubiquitination. J. Mol. Biol. 386: 814–827. 20. Kerscher, O., R. Felberbaum, and M. Hochstrasser. 2006. Modification of proteins by ubiquitin and ubiquitin-like proteins. Annu. Rev. Cell Dev. Biol. 22: 159–80. 21. Orlowski, N., and S. Wilk. 1981. A multicatalytical protease complex from pituitary that forms enkephalin and enkephalin containing peptides. Biochem. Biophys. Res. Commun. 101: 814–822. 22. Rivett, A. J. 1989. The multicatalytic proteinase complex. Revis. Biol. Celular 20: 113–123. 23. Wilk, S., and M. Orlowski. 1980. CationǦSensitive Neutral Endopeptidase: Isolation and Specificity of the Bovine Pituitary Enzyme. J. Neurochem. 35: 1172–1182. 24. Arrigo, A.-P., K. Tanaka, A. L. Goldberg, and W. J. Welch. 1988. Identity of the 19S “prosome” particle with the large multifunctional protease complex of mammalian cells (the proteasome). Nature 331: 192–194. 25. Förster, A., E. I. Masters, F. G. Whitby, H. Robinson, and C. P. Hill. 2005. The 1.9 A structure of a proteasome-11S activator complex and implications for proteasome-PAN/PA700 interactions. Mol. Cell 18: 589–99. 26. Stadtmueller, B. M., and C. P. Hill. 2011. Proteasome activators. Mol. Cell 41: 8–19. 27. Fabre, B., T. Lambour, L. Garrigues, F. Amalric, N. Vigneron, T. Menneteau, A. Stella, B. Monsarrat, B. Van den Eynde, O. Burlet-Schiltz, and M.-P. Bousquet-Dubouch. 2015. Deciphering preferential interactions within supramolecular protein complexes: the proteasome case. Mol. Syst. Biol. 11: 771. 28. Collins, G. A., and A. L. Goldberg. 2017. The Logic of the 26S Proteasome. Cell 169: 792–806. 29. Tanaka, K. 2009. The proteasome: Overview of structure and functions. Proc. Japan Acad. Ser. B 85: 12–36. 30. Groll, M., L. Ditzel, J. Löwe, D. Stock, M. Bochtler, H. D. Bartunik, and R. Huber. 1997. Structure of 20S proteasome from yeast at 2.4 A resolution. Nature 386: 463–71. 31. Huber, E. M., M. Basler, R. Schwab, W. Heinemeyer, C. J. Kirk, M. Groettrup, and M. Groll. 2012. Immuno- and constitutive proteasome crystal structures reveal differences in substrate and inhibitor specificity. Cell 148: 727–38. 32. Choi, W. H., S. A. H. de Poot, J. H. Lee, J. H. Kim, D. H. Han, Y. K. Kim, D. Finley, and M. J. Lee. 2016. Open-gate mutants of the mammalian proteasome show enhanced ubiquitin-conjugate degradation. Nat. Commun. 7: 10963. 33. Whitby, F. G., E. I. Masters, L. Kramer, J. R. Knowlton, Y. Yao, C. C. Wang, and C. P. Hill. 2000. Structural basis for the activation of 20S proteasomes by 11S regulators. Nature 408: 115–120. 34. Sadre-Bazzaz, K., F. G. Whitby, H. Robinson, T. Formosa, and C. P. Hill. 2010. Structure of a Blm10 Complex Reveals Common Mechanisms for Proteasome Binding and Gate Opening. Mol. Cell 37: 728–735. 35. Yu, Y., D. M. Smith, H. M. Kim, V. Rodriguez, A. L. Goldberg, and Y. Cheng. 2010. Interactions of PAN’s C-termini with archaeal 20S proteasome and implications for the eukaryotic proteasome-ATPase interactions. EMBO J. 29: 692–702. 36. Osmulski, P. A., M. Hochstrasser, and M. Gaczynska. 2009. A Tetrahedral Transition State at the Active Sites of the 20S Proteasome Is Coupled to Opening of the α-Ring Channel. Structure 17: 1137–1147. 37. Ruschak, A. M., and L. E. Kay. 2012. Proteasome allostery as a population shift between interchanging conformers. Proc. Natl. Acad. Sci. 109: E3454–E3462. 38. Schmidtke, G., S. Emch, M. Groettrup, and H. G. Holzhütter. 2000. Evidence for the existence of a non-catalytic modifier site of peptide hydrolysis by the 20 S proteasome. J. Biol. Chem. 275: 22056–22063. 39. Coux, O., K. Tanaka, and A. L. Goldberg. 1996. Structure and Functions of the 20S and 26S Proteasomes. Annu. Rev. Biochem. 65: 801–847.

143

References

40. Fabre, B., T. Lambour, J. Delobel, F. Amalric, B. Monsarrat, O. Burlet-Schiltz, and M.-P. Bousquet-Dubouch. 2013. Subcellular Distribution and Dynamics of Active Proteasome Complexes Unraveled by a Workflow Combining in Vivo Complex Cross-Linking and Quantitative Proteomics. Mol. Cell. Proteomics 12: 687–699. 41. Raynes, R., L. C. D. Pomatto, and K. J. A. Davies. 2016. Degradation of oxidized proteins by the proteasome: Distinguishing between the 20S, 26S, and immunoproteasome proteolytic pathways. Mol. Aspects Med. 50: 41–55. 42. Yang, Y., K. Fruh, K. Ahn, and P. A. Peterson. 1995. In vivo assembly of the proteasomal complexes, implications for antigen processing. J. Biol. Chem. 270: 27687–27694. 43. Ben-Nissan, G., and M. Sharon. 2014. Regulating the 20S proteasome ubiquitin-independent degradation pathway. Biomolecules 4: 862–884. 44. York, I. A., A. L. Goldberg, X. Y. Mo, and K. L. Rock. 1999. Proteolysis and class I major histocompatibility complex antigen presentation. Immunol. Rev. 172: 49–66. 45. Tanaka, K., T. Mizushima, and Y. Saeki. 2012. The proteasome: molecular machinery and pathophysiological roles. Biol. Chem. 393: 217–34. 46. Baumeister, W., J. Walz, F. Zühl, and E. Seemüller. 1998. The proteasome: paradigm of a self-compartmentalizing protease. Cell 92: 367–80. 47. Brown, M. G., J. Driscoll, and J. J. Monaco. 1991. Structural and serological similarity of MHC-linked LMP and proteasome (multicatalytic proteinase) complexes. Nature 353: 355–357. 48. Glynne, R., S. H. Powis, S. Beck, A. Kelly, L. a Kerr, and J. Trowsdale. 1991. A proteasome-related gene between the two ABC transporter loci in the class II region of the human MHC. Nature 353: 357–360. 49. Kelly, A., S. H. Powis, R. Glynne, E. Radley, S. Beck, and J. Trowsdale. 1991. Second proteasome-related gene in the human MHC class II region. Nature 353: 667–668. 50. Groettrup, M., C. J. Kirk, and M. Basler. 2010. Proteasomes in immune cells: more than peptide producers? Nat. Rev. Immunol. 10: 73–8. 51. Murata, S., Y. Takahama, and K. Tanaka. 2008. Thymoproteasome: probable role in generating positively selecting peptides. Curr. Opin. Immunol. 20: 192–6. 52. Groettrup, M., R. Kraft, S. Kostka, S. Standera, R. Stohwasser, and P. M. Kloetzel. 1996. A third interferon-gamma-induced subunit exchange in the 20S proteasome. Eur. J. Immunol. 26: 863–9. 53. Nandi, D., H. Jiang, and J. J. Monaco. 1996. Identification of MECL-1 (LMP-10) as the third IFN-gamma-inducible proteasome subunit. J. Immunol. 156: 2361–4. 54. Kuckelkorn, U., S. Frentzel, R. Kraft, S. Kostka, M. Groettrup, and P. M. Kloetzel. 1995. Incorporation of major histocompatibility complex-- encoded subunits LMP2 and LMP7 changes the quality of the 20S proteasome polypeptide processing products independent of interferon-gamma. Eur. J. Immunol. 25: 2605–11. 55. Murata, S., K. Sasaki, T. Kishimoto, S.-I. Niwa, H. Hayashi, Y. Takahama, and K. Tanaka. 2007. Regulation of CD8+ T cell development by thymus-specific proteasomes. Science . 316: 1349–53. 56. Tomaru, U., A. Ishizu, S. Murata, Y. Miyatake, S. Suzuki, S. Takahashi, T. Kazamaki, J. Ohara, T. Baba, S. Iwasaki, K. Fugo, N. Otsuka, K. Tanaka, and M. Kasahara. 2009. Exclusive expression of proteasome subunit {beta}5t in the human thymic cortex. Blood 113: 5186–91. 57. Kniepert, A., and M. Groettrup. 2013. The unique functions of tissue-specific proteasomes. Trends Biochem. Sci. 1–8. 58. De Bruin, G., B. T. Xin, B. I. Florea, and H. S. Overkleeft. 2016. Proteasome Subunit Selective Activity-Based Probes Report on Proteasome Core Particle Composition in a Native Polyacrylamide Gel Electrophoresis Fluorescence-Resonance Energy Transfer Assay. J. Am. Chem. Soc. 138: 9874–9880. 59. Löwe, J., D. Stock, B. Jap, P. Zwickl, W. Baumeister, and R. Huber. 1995. Crystal structure of the 20S proteasome from the archaeon T. acidophilum at 3.4 A resolution. Science . 268: 533–9. 60. Ruschak, A. M., M. Slassi, L. E. Kay, and A. D. Schimmer. 2011. Novel proteasome inhibitors to overcome bortezomib resistance. J. Natl. Cancer Inst. 103: 1007–17. 61. Kisselev, A. F., T. N. Akopian, K. M. Woo, and A. L. Goldberg. 1999. The Sizes of Peptides Generated from Protein by Mammalian 26 and 20S Proteasomes. J. Biol. Chem. 274: 3363–3371. 62. Kisselev, A. F., T. N. Akopian, and A. L. Goldberg. 1998. Range of sizes of peptide products generated during degradation of different proteins by archaeal proteasomes. J. Biol. Chem. 273: 1982–1989. 63. Gaczynska, M., K. L. Rock, and A. L. Goldberg. 1993. Gamma-interferon and expression of MHC genes regulate peptide hydrolysis by proteasomes. Nature 365: 264–7. 64. Driscoll, J., M. G. Brown, D. Finley, and J. J. Monaco. 1993. MHC-linked LMP gene products specifically alter peptidase activities of the proteasome. Nature 365: 262–4. 65. Cromm, P. M., and C. M. Crews. 2017. The Proteasome in Modern Drug Discovery: Second Life of a Highly Valuable Drug Target. ACS Cent. Sci. 3: 830–838. 66. Gaczynska, M., K. L. Rock, T. Spies, and a L. Goldberg. 1994. Peptidase activities of proteasomes are differentially regulated by the major histocompatibility complex-encoded genes for LMP2 and LMP7. Proc. Natl. Acad. Sci. U. S. A. 91: 9213–7. 67. Van Kaer, L., P. G. Ashton-Rickardt, M. Eichelberger, M. Gaczynska, K. Nagashima, K. L. Rock, a L. Goldberg, P. C. Doherty, and S. Tonegawa. 1994. Altered peptidase and viral-specific T cell response in LMP2 mutant mice. Immunity 1: 533–41. 68. Basler, M., C. J. Kirk, and M. Groettrup. 2012. The immunoproteasome in antigen processing and other immunological functions. Curr. Opin. Immunol. 1–7. 69. Arciniega, M., P. Beck, O. F. Lange, M. Groll, and R. Huber. 2014. Differential global structural changes in the core particle of yeast and mouse proteasome induced by ligand binding. Proc. Natl. Acad. Sci. 111: 9479–9484. 70. Liepe, J., H. G. Holzhütter, E. Bellavista, P. M. Kloetzel, M. P. H. Stumpf, and M. Mishto. 2015. Quantitative time-resolved analysis reveals intricate, differential regulation of standard- and immuno-proteasomes. Elife 4: 1–23. 71. Sijts, A. J. A. M., S. Standera, R. E. M. Toes, T. Ruppert, N. J. C. M. Beekman, P. A. van Veelen, F. A. Ossendorp, C. J. M. Melief, and P. M. Kloetzel. 2000. MHC Class I Antigen Processing of an Adenovirus CTL Epitope Is Linked to the Levels of Immunoproteasomes in Infected Cells. J. Immunol. 164: 4500–4506. 72. Schultz, E. S., J. Chapiro, C. Lurquin, S. Claverol, O. Burlet-Schiltz, G. Warnier, V. Russo, S. Morel, F. Lévy, T. Boon, B. J. Van den Eynde, and P. van der Bruggen. 2002. The production of a new MAGE-3 peptide presented to cytolytic T lymphocytes by HLA-B40 requires the immunoproteasome. J. Exp. Med. 195: 391–9. 73. Seifert, U., L. P. Bialy, F. Ebstein, D. Bech-Otschir, A. Voigt, F. Schröter, T. Prozorovski, N. Lange, J. Steffen, M. Rieger, U. Kuckelkorn, O. Aktas, P.-M. Kloetzel, and E. Krüger. 2010. Immunoproteasomes preserve protein homeostasis upon interferon-induced oxidative stress. Cell 142: 613–24. 74. Opitz, E., A. Koch, K. Klingel, F. Schmidt, S. Prokop, A. Rahnefeld, M. Sauter, F. L. Heppner, U. Völker, R. Kandolf, U. Kuckelkorn, K. Stangl, E. Krüger, P. M. Kloetzel, and A. Voigt. 2011. Impairment of immunoproteasome function by β5i/lmp7 subunit deficiency results in severe enterovirus myocarditis. PLoS Pathog. 7. 144

References

75. Henderson, A., J. Erales, M. A. Hoyt, and P. Coffino. 2011. Dependence of proteasome processing rate on substrate unfolding. J. Biol. Chem. 286: 17495–17502. 76. Grant, E. P., M. T. Michalek, A. L. Goldberg, and K. L. Rock. 1995. Rate of antigen degradation by the ubiquitin-proteasome pathway influences MHC class I presentation. J. Immunol. 155: 3750–3758. 77. Peth, A., J. A. Nathan, and A. L. Goldberg. 2013. The ATP costs and time required to degrade ubiquitinated proteins by the 26 S proteasome. J. Biol. Chem. 288: 29215–29222. 78. Nathan, J. a, V. Spinnenhirn, G. Schmidtke, M. Basler, M. Groettrup, and A. L. Goldberg. 2013. Immuno- and constitutive proteasomes do not differ in their abilities to degrade ubiquitinated proteins. Cell 152: 1184–94. 79. Hewing, B., A. Ludwig, C. Dan, M. Pötzsch, C. Hannemann, A. Petry, D. Lauer, A. Görlach, E. Kaschina, D. N. Müller, G. Baumann, V. Stangl, K. Stangl, and N. Wilck. 2017. Immunoproteasome subunit ß5i/LMP7-deficiency in atherosclerosis. Sci. Rep. 7: 1–10. 80. Basler, M., S. Mundt, T. Muchamuel, C. Moll, J. Jiang, M. Groettrup, and C. J. Kirk. 2014. Inhibition of the immunoproteasome ameliorates experimental autoimmune encephalomyelitis. EMBO Mol. Med. 6: 226–238. 81. Hirano, Y., K. B. Hendil, H. Yashiroda, S. I. Iemura, R. Nagane, Y. Hioki, T. Natsume, K. Tanaka, and S. Murata. 2005. A heterodimeric complex that promotes the assembly of mammalian 20S proteasomes. Nature 437: 1381–1385. 82. Griffin, T. A., J. P. Slack, T. S. McCluskey, J. J. Monaco, and R. A. Colbert. 2000. Identification of proteassemblin, a mammalian homologue of the yeast protein, Ump1p, that is required for normal proteasome assembly. Mol. Cell Biol. Res. Commun. 3: 212–217. 83. Hirano, Y., T. Kaneko, K. Okamoto, M. Bai, H. Yashiroda, K. Furuyama, K. Kato, K. Tanaka, and S. Murata. 2008. Dissecting β-ring assembly pathway of the mammalian 20S proteasome. EMBO J. 27: 2204–2213. 84. Bai, M., X. Zhao, K. Sahara, Y. Ohte, Y. Hirano, T. Kaneko, H. Yashiroda, and S. Murata. 2014. Assembly Mechanisms of Specialized Core Particles of. Biomolecules 16: 662–677. 85. Arendt, C. S., and M. Hochstrasser. 1999. Eukaryotic 20S proteasome catalytic subunit propeptides prevent active site inactivation by N- terminal acetylation and promote particle assembly. EMBO J. 18: 3575–3585. 86. De, M., K. Jayarapu, L. Elenich, J. J. Monaco, R. a Colbert, and T. a Griffin. 2003. Beta 2 subunit propeptides influence cooperative proteasome assembly. J. Biol. Chem. 278: 6153–9. 87. Heinemeyer, W., M. Fischer, T. Krimmer, U. Stachon, and D. H. Wolf. 1997. The active sites of the eukaryotic 20 S proteasome and their involvement in subunit precursor processing. J. Biol. Chem. 272: 25200–25209. 88. Gu, Z. C., and C. Enenkel. 2014. Proteasome assembly. Cell. Mol. Life Sci. 71: 4729–4745. 89. Griffin, T. A., D. Nandi, M. Cruz, H. J. Fehling, L. V Kaer, J. J. Monaco, and R. A. Colbert. 1998. Immunoproteasome assembly: cooperative incorporation of interferon gamma (IFN-gamma)-inducible subunits. J. Exp. Med. 187: 97–104. 90. Groettrup, M., S. Standera, R. Stohwasser, and P. M. Kloetzel. 1997. The subunits MECL-1 and LMP2 are mutually required for incorporation into the 20S proteasome. Proc. Natl. Acad. Sci. U. S. A. 94: 8970–5. 91. Heink, S., D. Ludwig, P.-M. Kloetzel, and E. Krüger. 2005. IFN-gamma-induced immune adaptation of the proteasome system is an accelerated and transient response. Proc. Natl. Acad. Sci. U. S. A. 102: 9241–6. 92. Witt, E., D. Zantopf, M. Schmidt, R. Kraft, P. M. Kloetzel, and E. Krüger. 2000. Characterisation of the newly identified human Ump1 homologue POMP and analysis of LMP7(β5i) incorporation into 20 S proteasomes. J. Mol. Biol. 301: 1–9. 93. Guillaume, B., J. Chapiro, V. Stroobant, D. Colau, B. Van Holle, G. Parvizi, M.-P. Bousquet-Dubouch, I. Théate, N. Parmentier, and B. J. Van den Eynde. 2010. Two abundant proteasome subtypes that uniquely process some antigens presented by HLA class I molecules. Proc. Natl. Acad. Sci. U. S. A. 107: 18599–604. 94. Gohlke, S., A. Kloß, M. Tsokos, K. Textoris-Taube, C. Keller, P. M. Kloetzel, and B. Dahlmann. 2014. Adult human liver contains intermediate-type proteasomes with different enzymatic properties. Ann. Hepatol. 13: 429–438. 95. Drews, Wildgruber, Zong, Sukop, Nissum, Weber, Gomes, and Ping. 2007. Mammalian Proteasome Subpopulations with Distinct Molecular Compositions and Proteolytic Activities. Mol. Cell. Proteomics 6: 2021–2031. 96. Stohwasser, R., U. Kuckelkorn, R. Kraft, S. Kostka, and P. M. Kloetzel. 1996. 20S proteasome from LMP7 knock out mice reveals altered proteolytic activities and cleavage site preferences. FEBS Lett. 383: 109–113. 97. Dahlmann, B. 2016. Mammalian proteasome subtypes: Their diversity in structure and function. Arch. Biochem. Biophys. 591: 132–140. 98. Mathieson, T., H. Franken, J. Kosinski, N. Kurzawa, N. Zinn, G. Sweetman, D. Poeckel, V. S. Ratnu, M. Schramm, I. Becher, M. Steidel, K.- M. Noh, G. Bergamini, M. Beck, M. Bantscheff, and M. M. Savitski. 2018. Systematic analysis of protein turnover in primary cells. Nat. Commun. 9: 689. 99. Lasker, K., F. Forster, S. Bohn, T. Walzthoeni, E. Villa, P. Unverdorben, F. Beck, R. Aebersold, A. Sali, and W. Baumeister. 2012. Molecular architecture of the 26S proteasome holocomplex determined by an integrative approach. Proc. Natl. Acad. Sci. 109: 1380–1387. 100. Lander, G. C., E. Estrin, M. E. Matyskiela, C. Bashore, E. Nogales, and A. Martin. 2012. Complete subunit architecture of the proteasome regulatory particle. Nature 482: 186–191. 101. Sakata, E., S. Bohn, O. Mihalache, P. Kiss, F. Beck, I. Nagy, S. Nickell, K. Tanaka, Y. Saeki, F. Forster, and W. Baumeister. 2012. Localization of the proteasomal ubiquitin receptors Rpn10 and Rpn13 by electron cryomicroscopy. Proc. Natl. Acad. Sci. 109: 1479–1484. 102. Köhler, A., P. Cascio, D. S. Leggett, K. M. Woo, A. L. Goldberg, and D. Finley. 2001. The axial channel of the proteasome core particle is gated by the Rpt2 ATPase and controls both substrate entry and product release. Mol. Cell 7: 1143–1152. 103. Rabl, J., D. M. Smith, Y. Yu, S. C. Chang, A. L. Goldberg, and Y. Cheng. 2008. Mechanism of Gate Opening in the 20S Proteasome by the Proteasomal ATPases. Mol. Cell 30: 360–368. 104. Smith, D. M., S. C. Chang, S. Park, D. Finley, Y. Cheng, and A. L. Goldberg. 2007. Docking of the Proteasomal ATPases’ Carboxyl Termini in the 20S Proteasome’s α Ring Opens the Gate for Substrate Entry. Mol. Cell 27: 731–744. 105. Wehmer, M., T. Rudack, F. Beck, A. Aufderheide, G. Pfeifer, J. M. Plitzko, F. Förster, K. Schulten, W. Baumeister, and E. Sakata. 2017. Structural insights into the functional cycle of the ATPase module of the 26S proteasome. Proc. Natl. Acad. Sci. 114: 1305–1310. 106. Chen, S., J. Wu, Y. Lu, Y.-B. Ma, B.-H. Lee, Z. Yu, Q. Ouyang, D. J. Finley, M. W. Kirschner, and Y. Mao. 2016. Structural basis for dynamic regulation of the human 26S proteasome. Proc. Natl. Acad. Sci. 113: 12991–12996. 107. Unverdorben, P., F. Beck, P. led , A. Schweitzer, G. Pfeifer, J. M. Plitzko, W. Baumeister, and F. Forster. 2014. Deep classification of a large cryo-EM dataset defines the conformational landscape of the 26S proteasome. Proc. Natl. Acad. Sci. 111: 5544–5549. 108. Guo, Q., C. Lehmer, A. Martínez-Sánchez, T. Rudack, F. Beck, H. Hartmann, M. Pérez-Berlanga, F. Frottin, M. S. Hipp, F. U. Hartl, D. Edbauer, W. Baumeister, and R. Fernández-Busnadiego. 2018. In Situ Structure of Neuronal C9orf72 Poly-GA Aggregates Reveals Proteasome Recruitment. Cell 172: 696–705.e12. 109. Bäuerlein, F. J. B., I. Saha, A. Mishra, M. Kalemanov, A. Martínez-Sánchez, R. Klein, I. Dudanova, M. S. Hipp, F. U. Hartl, W. Baumeister, and R. Fernández-Busnadiego. 2017. In Situ Architecture and Cellular Interactions of PolyQ Inclusions. Cell 171: 179–187.e10. 110. Rosenzweig, R., V. Bronner, D. Zhang, D. Fushman, and M. H. Glickman. 2012. Rpn1 and Rpn2 coordinate ubiquitin processing factors at 145

References proteasome. J. Biol. Chem. 287: 14659–71. 111. Deveraux, Q., V. Ustrell, C. Pickart, and M. Rechsteiner. 1994. A 26 S protease subunit that binds ubiquitin conjugates. J. Biol. Chem. 269: 7059–7061. 112. Schreiner, P., X. Chen, K. Husnjak, L. Randles, N. Zhang, S. Elsasser, D. Finley, I. Dikic, K. J. Walters, and M. Groll. 2008. Ubiquitin docking at the proteasome through a novel pleckstrin-homology domain interaction. Nature 453: 548–552. 113. Husnjak, K., S. Elsasser, N. Zhang, X. Chen, L. Randles, Y. Shi, K. Hofmann, K. J. Walters, D. Finley, and I. Dikic. 2008. Proteasome subunit Rpn13 is a novel ubiquitin receptor. Nature 453: 481–8. 114. Dubiel, W., G. Pratt, K. Ferrell, and M. Rechsteiner. 1992. Purification of an 11 S regulator of the multicatalytic protease. J. Biol. Chem. 267: 22369–77. 115. Ma, C. P., C. a Slaughter, and G. N. DeMartino. 1992. Identification, purification, and characterization of a protein activator (PA28) of the 20 S proteasome (macropain). J. Biol. Chem. 267: 10515–23. 116. Macagno, A., M. Gilliet, F. Sallusto, A. Lanzavecchia, F. O. Nestle, and M. Groettrup. 1999. Dendritic cells up-regulate immunoproteasomes and the proteasome regulator PA28 during maturation. Eur. J. Immunol. 29: 4037–4042. 117. Wójcik, C., K. Tanaka, N. Paweletz, U. Naab, and S. Wilk. 1998. Proteasome activator (PA28) subunits, α, β and γ (Ki antigen) in NT2 neuronal precursor cells and HeLa S3 cells. Eur. J. Cell Biol. 77: 151–160. 118. Tanahashi, N., K. Yokota, J. Y. Ahn, C. H. Chung, T. Fujiwara, E. Takahashi, G. N. DeMartino, C. a Slaughter, T. Toyonaga, K. Yamamura, N. Shimbara, and K. Tanaka. 1997. Molecular properties of the proteasome activator PA28 family proteins and gamma-interferon regulation. Genes to Cells 2: 195–211. 119. Groettrup, M., T. Ruppert, L. Kuehn, M. Seeger, S. Standera, U. Koszinowski, and P. M. Kloetzel. 1995. The interferon-gamma-inducible 11 S regulator (PA28) and the LMP2/LMP7 subunits govern the peptide production by the 20 S proteasome in vitro. J. Biol. Chem. 270: 23808–15. 120. Schmidtke, G., R. Schregle, G. Alvarez, E. M. Huber, and M. Groettrup. 2017. The 20S immunoproteasome and constitutive proteasome bind with the same affinity to PA28αβ and equally degrade FAT10. Mol. Immunol. 0–1. 121. Groettrup, M., A. Soza, M. Eggers, L. Kuehn, T. P. Dick, H. Schild, H. G. Rammensee, U. H. Koszinowski, and P. M. Kloetzel. 1996. A role for the proteasome regulator PA28alpha in antigen presentation. Nature 381: 166–8. 122. Raule, M., F. Cerruti, N. Benaroudj, R. Migotti, J. Kikuchi, A. Bachi, A. Navon, G. Dittmar, and P. Cascio. 2014. PA28αβ reduces size and increases hydrophilicity of 20S immunoproteasome peptide products. Chem. Biol. 21: 470–480. 123. Pickering, A. M., and K. J. A. Davies. 2012. Differential roles of proteasome and immunoproteasome regulators Pa28αβ, Pa28γ and Pa200 in the degradation of oxidized proteins. Arch. Biochem. Biophys. 523: 181–190. 124. van Hall, T., a Sijts, M. Camps, R. Offringa, C. Melief, P.-M. Kloetzel, and F. Ossendorp. 2000. Differential Influence on Cytotoxic T Lymphocyte Epitope Presentation by Controlled Expression of Either Proteasome Immunosubunits or Pa28. J. Exp. Med. 192: 483–494. 125. de Graaf, N., M. J. G. van Helden, K. Textoris-Taube, T. Chiba, D. J. Topham, P. M. Kloetzel, D. M. W. Zaiss, and A. J. A. M. Sijts. 2011. PA28 and the proteasome immunosubunits play a central and independent role in the production of MHC class I-binding peptides in vivo. Eur. J. Immunol. 41: 926–935. 126. Schwarz, K., M. van Den Broek, S. Kostka, R. Kraft, A. Soza, G. Schmidtke, P. M. Kloetzel, and M. Groettrup. 2000. Overexpression of the proteasome subunits LMP2, LMP7, and MECL-1, but not PA28 alpha/beta, enhances the presentation of an immunodominant lymphocytic choriomeningitis virus T cell epitope. J. Immunol. 165: 768–78. 127. Ustrell, V., L. Hoffman, G. Pratt, and M. Rechsteiner. 2002. PA200, a nuclear proteasome activator involved in DNA repair. EMBO J. 21: 3516–25. 128. Ortega, J., J. B. Heymann, A. V Kajava, V. Ustrell, M. Rechsteiner, and A. C. Steven. 2005. The axial channel of the 20S proteasome opens upon binding of the PA200 activator. J. Mol. Biol. 346: 1221–7. 129. Marques, A. J., C. Glanemann, P. C. Ramos, and R. J. Dohmen. 2007. The C-terminal extension of the β7 subunit and activator complexes stabilize nascent 20 S proteasomes and promote their maturation. J. Biol. Chem. 282: 34869–34876. 130. Lehmann, A., K. Jechow, and C. Enenkel. 2008. Blm10 binds to pre-activated proteasome core particles with open gate conformation. EMBO Rep. 9: 1237–1243. 131. Iwanczyk, J., K. Sadre-Bazzaz, K. Ferrell, E. Kondrashkina, T. Formosa, C. P. Hill, and J. Ortega. 2006. Structure of the Blm10-20 S Proteasome Complex by Cryo-electron Microscopy. Insights into the Mechanism of Activation of Mature Yeast Proteasomes. J. Mol. Biol. 363: 648–659. 132. Zaiss, D. M. W., S. Standera, P.-M. Kloetzel, and A. J. A. M. Sijts. 2002. PI31 is a modulator of proteasome formation and antigen processing. Proc. Natl. Acad. Sci. 99: 14344–14349. 133. Li, X., D. Thompson, B. Kumar, and G. N. DeMartino. 2014. Molecular and cellular roles of PI31 (PSMF1) protein in regulation of proteasome function. J. Biol. Chem. 289: 17392–17405. 134. Erath, S., and M. Groettrup. 2015. No evidence for immunoproteasomes in chicken lymphoid organs and activated lymphocytes. Immunogenetics 67: 51–60. 135. Stohwasser, R., S. Standera, I. Peters, P. M. Kloetzel, and M. Groettrup. 1997. Molecular cloning of the mouse proteasome subunits MC14 and MECL-1: Reciprocally regulated tissue expression of interferon-γ-modulated proteasome subunits. Eur. J. Immunol. 27: 1182–1187. 136. Frentzel, S., I. Kuhn-Hartmann, M. Gernold, P. Gött, A. Seelig, and P. M. Kloetzel. 1993. The major-histocompatibility-complex-encoded beta-type proteasome subunits LMP2 and LMP7. Evidence that LMP2 and LMP7 are synthesized as proproteins and that cellular levels of both mRNA and LMP-containing 20S proteasomes are differentially regulated. Eur. J. Biochem. 216: 119–26. 137. Moebius, J., M. van den Broek, M. Groettrup, and M. Basler. 2010. Immunoproteasomes are essential for survival and expansion of T cells in virus-infected mice. Eur. J. Immunol. 40: 3439–49. 138. Zaiss, D. M. W., N. de Graaf, and A. J. a M. Sijts. 2008. The proteasome immunosubunit multicatalytic endopeptidase complex-like 1 is a T- cell-intrinsic factor influencing homeostatic expansion. Infect. Immun. 76: 1207–13. 139. Hensley, S. E., D. Zanker, B. P. Dolan, A. David, H. D. Hickman, A. C. Embry, C. N. Skon, K. M. Grebe, T. a Griffin, W. Chen, J. R. Bennink, and J. W. Yewdell. 2010. Unexpected role for the immunoproteasome subunit LMP2 in antiviral humoral and innate immune responses. J. Immunol. 184: 4115–22. 140. Barton, L. F., M. Cruz, R. Rangwala, G. S. Deepe, and J. J. Monaco. 2002. Regulation of immunoproteasome subunit expression in vivo following pathogenic fungal infection. J. Immunol. 169: 3046–52. 141. Nil, A., E. Firat, V. Sobek, K. Eichmann, and G. Niedermann. 2004. Expression of housekeeping and immunoproteasome subunit genes is differentially regulated in positively and negatively selecting thymic stroma subsets. Eur. J. Immunol. 34: 2681–9. 142. Akiyama, K., K. Yokota, S. Kagawa, N. Shimbara, T. Tamura, H. Akioka, H. G. Nothwang, C. Noda, K. Tanaka, and A. Ichihara. 1994. cDNA cloning and interferon gamma down-regulation of proteasomal subunits X and Y. Science . 265: 1231–4. 143. Kremer, M., A. Henn, C. Kolb, M. Basler, J. Moebius, B. Guillaume, M. Leist, B. J. Van den Eynde, and M. Groettrup. 2010. Reduced 146

References immunoproteasome formation and accumulation of immunoproteasomal precursors in the brains of lymphocytic choriomeningitis virus-infected mice. J. Immunol. 185: 5549–60. 144. Khan, S., M. van den Broek, K. Schwarz, R. de Giuli, P. a Diener, and M. Groettrup. 2001. Immunoproteasomes largely replace constitutive proteasomes during an antiviral and antibacterial immune response in the liver. J. Immunol. 167: 6859–68. 145. Goldberg, a L., and K. L. Rock. 1992. Proteolysis, proteasomes and antigen presentation. Nature 357: 375–379. 146. Niedermann, G., E. Geier, M. Lucchiari-Hartz, N. Hitziger, A. Ramsperger, and K. Eichmann. 1999. The specificity of proteasomes: impact on MHC class I processing and presentation of antigens. Immunol. Rev. 172: 29–48. 147. Rock, K. L., D. J. Farfán-Arribas, J. D. Colbert, and A. L. Goldberg. 2014. Re-examining class-I presentation and the DRiP hypothesis. Trends Immunol. 35: 144–152. 148. Anton, L. C., and J. W. Yewdell. 2014. Translating DRiPs: MHC class I immunosurveillance of pathogens and tumors. J. Leukoc. Biol. 95: 551–562. 149. Yewdell, J. W., L. C. Antón, and J. R. Bennink. 1996. Defective ribosomal products (DRiPs): a major source of antigenic peptides for MHC class I molecules? J. Immunol. 157: 1823–6. 150. Valečka, J., C. R. Almeida, B. Su, P. Pierre, and E. Gatti. 2018. Autophagy and MHC-restricted antigen presentation. Mol. Immunol. 99: 163– 170. 151. Grotzke, J. E., D. Sengupta, Q. Lu, and P. Cresswell. 2017. The ongoing saga of the mechanism(s) of MHC class I-restricted cross- presentation. Curr. Opin. Immunol. 46: 89–96. 152. Theofilopoulos, A. N., D. H. Kono, and R. Baccala. 2017. The multiple pathways to autoimmunity. Nat. Immunol. 18: 716–724. 153. Deres, K., W. Beck, S. Faath, G. Jung, and H.-G. Rammensee. 1993. MHC/Peptide Binding Studies indicate Hierarchy of Anchor Residues. Cell. Immunol. 151: 158–167. 154. Mo, X. Y., P. Cascio, K. Lemerise, A. L. Goldberg, and K. Rock. 1999. Distinct proteolytic processes generate the C and N termini of MHC class I-binding peptides. J. Immunol. 163: 5851–9. 155. York, I. A., S. C. Chang, T. Saric, J. A. Keys, J. M. Favreau, A. L. Goldberg, and K. L. Rock. 2002. The Er aminopeptidase ERAP I enhances or limits antigen presentation by trimming epitopes to 8-9 residues. Nat. Immunol. 3: 1177–1184. 156. Cascio, P., C. Hilton, a F. Kisselev, K. L. Rock, and a L. Goldberg. 2001. 26S proteasomes and immunoproteasomes produce mainly N- extended versions of an antigenic peptide. EMBO J. 20: 2357–66. 157. Saric, T., S. C. Chang, A. Hattori, I. A. York, S. Markant, K. L. Rock, M. Tsujimoto, and A. L. Goldberg. 2002. An IFN-y-induced aminopeptidase in the ER, ERAP I, trims precursors to MHC class I-presented peptides. Nat. Immunol. 3: 1169–1176. 158. Fehling, H. J., W. Swat, C. Laplace, R. Kühn, K. Rajewsky, U. Müller, and H. von Boehmer. 1994. MHC class I expression in mice lacking the proteasome subunit LMP-7. Science . 265: 1234–7. 159. Morel, S., F. Lévy, O. Burlet-Schiltz, F. Brasseur, M. Probst-Kepper, A.-L. Peitrequin, B. Monsarrat, R. Van Velthoven, J.-C. Cerottini, T. Boon, J. E. Gairin, and B. J. Van den Eynde. 2000. Processing of Some Antigens by the Standard Proteasome but Not by the Immunoproteasome Results in Poor Presentation by Dendritic Cells. Immunity 12: 107–117. 160. Gileadi, U., H. T. Moins-Teisserenc, I. Correa, B. L. Booth, P. R. Dunbar, a K. Sewell, J. Trowsdale, R. E. Phillips, and V. Cerundolo. 1999. Generation of an immunodominant CTL epitope is affected by proteasome subunit composition and stability of the antigenic protein. J. Immunol. 163: 6045–52. 161. Sijts, a J., T. Ruppert, B. Rehermann, M. Schmidt, U. Koszinowski, and P. M. Kloetzel. 2000. Efficient generation of a hepatitis B virus cytotoxic T lymphocyte epitope requires the structural features of immunoproteasomes. J. Exp. Med. 191: 503–14. 162. Basler, M., C. Lauer, J. Moebius, R. Weber, M. Przybylski, A. F. Kisselev, C. Tsu, and M. Groettrup. 2012. Why the structure but not the activity of the immunoproteasome subunit low molecular mass polypeptide 2 rescues antigen presentation. J. Immunol. 189: 1868–77. 163. Mishto, M., J. Liepe, K. Textoris-Taube, C. Keller, P. Henklein, M. Weberru??, B. Dahlmann, C. Enenkel, A. Voigt, U. Kuckelkorn, M. P. H. Stumpf, and P. M. Kloetzel. 2014. Proteasome isoforms exhibit only quantitative differences in cleavage and epitope generation. Eur. J. Immunol. 44: 3508–3521. 164. Basler, M., J. Moebius, L. Elenich, M. Groettrup, and J. J. Monaco. 2006. An altered T cell repertoire in MECL-1-deficient mice. J. Immunol. 176: 6665–72. 165. Chen, W., C. C. Norbury, Y. Cho, J. W. Yewdell, and J. R. Bennink. 2001. Immunoproteasomes shape immunodominance hierarchies of antiviral CD8(+) T cells at the levels of T cell repertoire and presentation of viral antigens. J. Exp. Med. 193: 1319–26. 166. Oh, I. S., K. Textoris-Taube, P. S. Sung, W. Kang, X. Gorny, T. Kähne, S. H. Hong, Y. J. Choi, C. Cammann, M. Naumann, J. H. Kim, S. H. Park, O. J. Yoo, P. M. Kloetzel, U. Seifert, and E. C. Shin. 2016. Immunoproteasome induction is suppressed in hepatitis C virus-infected cells in a protein kinase R-dependent manner. Exp. Mol. Med. 48: e270. 167. de Verteuil, D. a, A. Rouette, M.-P. Hardy, S. Lavallée, A. Trofimov, E. Gaucher, and C. Perreault. 2014. Immunoproteasomes Shape the Transcriptome and Regulate the Function of Dendritic Cells. J. Immunol. . 168. Basler, M., S. Mundt, and M. Groettrup. 2018. The immunoproteasome subunit LMP7 is required in the murine thymus for filling up a hole in the T cell repertoire. Eur. J. Immunol. 48: 419–429. 169. Tu, L., C. Moriya, T. Imai, H. Ishida, K. Tetsutani, X. Duan, S. Murata, K. Tanaka, C. Shimokawa, H. Hisaeda, and K. Himeno. 2009. Critical role for the immunoproteasome subunit LMP7 in the resistance of mice to Toxoplasma gondii infection. Eur. J. Immunol. 39: 3385–94. 170. Kincaid, E. Z., S. Murata, K. Tanaka, and K. L. Rock. 2016. Specialized proteasome subunits have an essential role in the thymic selection of CD8 + T cells. Nat. Immunol. 1–9. 171. Suraweera, A., C. Münch, A. Hanssum, and A. Bertolotti. 2012. Failure of amino acid homeostasis causes cell death following proteasome inhibition. Mol. Cell 48: 242–253. 172. Zhao, J., B. Zhai, S. P. Gygi, and A. L. Goldberg. 2015. mTOR inhibition activates overall protein degradation by the ubiquitin proteasome system as well as by autophagy. Proc. Natl. Acad. Sci. 112: 15790–15797. 173. Rousseau, A., and A. Bertolotti. 2016. An evolutionarily conserved pathway controls proteasome homeostasis. Nature 536: 184–189. 174. Zhang, Y., J. Nicholatos, J. R. Dreier, S. J. H. Ricoult, S. B. Widenmaier, G. S. Hotamisligil, D. J. Kwiatkowski, and B. D. Manning. 2014. Coordinated regulation of protein synthesis and degradation by mTORC1. Nature . 175. Yun, Y. S., K. H. Kim, B. Tschida, L. Chen, D. Largaespada, D. Kim, Y. S. Yun, K. H. Kim, B. Tschida, Z. Sachs, K. E. Noble-orcutt, and B. S. Moriarity. 2016. mTORC1 Coordinates Protein Synthesis and Immunoproteasome Formation via PRAS40 to Prevent Accumulation of Protein Stress. Mol. Cell 61: 625–639. 176. Laplante, M., and D. M. Sabatini. 2012. mTOR signaling in growth control and disease. Cell 149: 274–293. 177. Murphy, M. P. 2009. How mitochondria produce reactive oxygen species. Biochem. J. 417: 1–13. 178. Iles, K. E., and H. J. Forman. 2002. Macrophage Signaling and Respiratory Burst. Immunol. Res. 26: 095-106. 179. Rada, B., and T. L. Leto. 2008. Oxidative innate immune defenses by Nox / Duox family NADPH oxidases. Trends Innate Immun. 15: 164– 147

References

187. 180. Kuwano, Y., T. Kawahara, H. Yamamoto, S. Teshima-Kondo, K. Tominaga, K. Masuda, K. Kishi, K. Morita, and K. Rokutan. 2006. Interferon-gamma activates transcription of NADPH oxidase 1 gene and upregulates production of superoxide anion by human large intestinal epithelial cells. Am. J. Physiol. Cell Physiol. 290: C433-43. 181. Tse, H. M., T. C. Thayer, C. Steele, C. M. Cuda, L. Morel, J. D. Piganelli, and C. E. Mathews. 2010. NADPH Oxidase Deficiency Regulates Th Lineage Commitment and Modulates Autoimmunity. J. Immunol. 185: 5247–5258. 182. Pickering, A. M., and K. J. a Davies. 2012. Degradation of damaged proteins: the main function of the 20S proteasome. Prog. Mol. Biol. Transl. Sci. 109: 227–48. 183. Ullrich, O., T. Reinheckel, N. Sitte, R. Hass, T. Grune, and K. J. Davies. 1999. Poly-ADP ribose polymerase activates nuclear proteasome to degrade oxidatively damaged histones. Proc. Natl. Acad. Sci. U. S. A. 96: 6223–6228. 184. Pickering, A. M., A. L. Koop, C. Y. Teoh, G. Ermak, T. Grune, and K. J. a Davies. 2010. The immunoproteasome, the 20S proteasome and the PA28αβ proteasome regulator are oxidative-stress-adaptive proteolytic complexes. Biochem. J. 432: 585–594. 185. Kotamraju, S., S. Matalon, T. Matsunaga, T. Shang, J. M. Hickman-Davis, and B. Kalyanaraman. 2006. Upregulation of immunoproteasomes by nitric oxide: Potential antioxidative mechanism in endothelial cells. Free Radic. Biol. Med. 40: 1034–1044. 186. St-Pierre, C., E. Morgand, M. Benhammadi, A. Rouette, M. P. Hardy, L. Gaboury, and C. Perreault. 2017. Immunoproteasomes Control the Homeostasis of Medullary Thymic Epithelial Cells by Alleviating Proteotoxic Stress. Cell Rep. 21: 2558–2570. 187. Maniatis, T., S. Goodbourn, and J. A. Fischer. 1987. Regulation of inducible and tissue-specific gene expression. Science . 236: 1237–45. 188. Gandolfi, S., J. P. Laubach, T. Hideshima, D. Chauhan, K. C. Anderson, and P. G. Richardson. 2017. The proteasome and proteasome inhibitors in multiple myeloma. Cancer Metastasis Rev. 36: 561–584. 189. Hayashi, T. 2000. Essential Role of Human Leukocyte Antigen-encoded Proteasome Subunits in NF-kappa B Activation and Prevention of Tumor Necrosis Factor-alpha -induced Apoptosis. J. Biol. Chem. 275: 5238–5247. 190. Hayashi, T., and D. Faustman. 1999. NOD mice are defective in proteasome production and activation of NF-kappaB. Mol. Cell. Biol. 19: 8646–59. 191. Kessler, B. M., A. M. Lennon-Duménil, M. L. Shinohara, M. A. Lipes, and H. L. Ploegh. 2000. LMP2 expression and proteasome activity in NOD mice. Nat. Med. 6: 1064; author reply 1065-6. 192. Runnels, H. A., W. A. Watkins, and J. J. Monaco. 2000. LMP2 expression and proteasome activity in NOD mice. Nat. Med. 6: 2–3. 193. Hayashi, Kodama, and Faustman. 2000. Reply to “LMP2 expression and proteasome activity in NOD mice.” Nat. Med. 6: 1065–6. 194. Bitzer, A., M. Basler, D. Krappmann, and M. Groettrup. 2017. Immunoproteasome subunit deficiency has no influence on the canonical pathway of NF-κB activation. Mol. Immunol. 83: 147–153. 195. Schmidt, N., E. Gonzalez, A. Visekruna, A. a Kühl, C. Loddenkemper, H. Mollenkopf, S. H. E. Kaufmann, U. Steinhoff, and T. Joeris. 2010. Targeting the proteasome: partial inhibition of the proteasome by bortezomib or deletion of the immunosubunit LMP7 attenuates experimental colitis. Gut 59: 896–906. 196. Visekruna, A., T. Joeris, D. Seidel, A. Kroesen, C. Loddenkemper, M. Zeitz, S. H. E. Kaufmann, R. Schmidt-Ullrich, and U. Steinhoff. 2006. Proteasome-mediated degradation of IkappaBalpha and processing of p105 in Crohn disease and ulcerative colitis. J. Clin. Invest. 116: 3195–203. 197. Maldonado, M., R. J. Kapphahn, M. R. Terluk, N. D. Heuss, C. Yuan, D. S. Gregerson, and D. a. Ferrington. 2013. Immunoproteasome Deficiency Modifies the Alternative Pathway of NFκB Signaling. PLoS One 8: e56187. 198. Jang, E. R., N.-R. Lee, S. Han, Y. Wu, L. K. Sharma, K. C. Carmony, J. Marks, D.-M. Lee, J.-O. Ban, M. Wehenkel, J. T. Hong, K. B. Kim, and W. Lee. 2012. Revisiting the role of the immunoproteasome in the activation of the canonical NF-κB pathway. Mol. Biosyst. 8: 2295. 199. Muchamuel, T., M. Basler, M. a Aujay, E. Suzuki, K. W. Kalim, C. Lauer, C. Sylvain, E. R. Ring, J. Shields, J. Jiang, P. Shwonek, F. Parlati, S. D. Demo, M. K. Bennett, C. J. Kirk, and M. Groettrup. 2009. A selective inhibitor of the immunoproteasome subunit LMP7 blocks cytokine production and attenuates progression of experimental arthritis. Nat. Med. 15: 781–7. 200. Paeschke, A., A. Possehl, K. Klingel, M. Voss, K. Voss, M. Kespohl, M. Sauter, H. S. Overkleeft, N. Althof, C. Garlanda, and A. Voigt. 2016. The immunoproteasome controls the availability of the cardioprotective pattern recognition molecule Pentraxin3. Eur. J. Immunol. 46: 619–633. 201. Sullivan, K. D., M. D. Galbraith, Z. Andrysik, and J. M. Espinosa. 2017. Mechanisms of transcriptional regulation by p53. Cell Death Differ. 25: 133–143. 202. Muñoz-Fontela, C., A. Mandinova, S. A. Aaronson, and S. W. Lee. 2016. Emerging roles of p53 and other tumour-suppressor genes in immune regulation. Nat. Rev. Immunol. 16: 741–750. 203. Kallio, P. J., W. J. Wilson, O. Brien, Y. Makino, J. B. Chem, S. O. Brien, and L. Poellinger. 1999. Regulation of the Hypoxia-inducible Transcription Factor 1 α by the Ubiquitin-Proteasome Pathway. J. Biol. Chem. 274: 6519–6525. 204. Suh, K. S., T. Tanaka, S. Sarojini, G. Nightingale, R. Gharbaran, A. Pecora, and A. Goy. 2013. The role of the ubiquitin proteasome system in lymphoma. Crit. Rev. Oncol. Hematol. 87: 306–22. 205. Yang, L., R. Wang, Z. Ma, Y. Xiao, Y. Nan, Y. Wang, S. Lin, and Y. Zhang. 2017. Porcine Reproductive and Respiratory Syndrome Virus Antagonizes JAK/STAT3 Signaling via nsp5, Which Induces STAT3 Degradation. J. Virol. 91: 1–14. 206. Scheffner, M., J. M. Huibregtse, R. D. Vierstra, and P. M. Howley. 1993. The HPV-16 E6 and E6-AP complex functions as a ubiquitin- protein ligase in the ubiquitination of p53. Cell 75: 495–505. 207. Horvath, C. M. 2004. Weapons of STAT destruction: Interferon evasion by paramyxovirus V proteins. Eur. J. Biochem. 271: 4621–4628. 208. Steffen, J., M. Seeger, A. Koch, and E. Krüger. 2010. Proteasomal degradation is transcriptionally controlled by TCF11 via an ERAD- dependent feedback loop. Mol. Cell 40: 147–158. 209. Tong, K. I., Y. Katoh, H. Kusunoki, K. Itoh, T. Tanaka, and M. Yamamoto. 2006. Keap1 Recruits Neh2 through Binding to ETGE and DLG Motifs: Characterization of the Two-Site Molecular Recognition Model. Mol. Cell. Biol. 26: 2887–2900. 210. Kageyama, S., Y. S. Sou, T. Uemura, S. Kametaka, T. Saito, R. Ishimura, T. Kouno, L. Bedford, R. J. Mayer, M. S. Lee, M. Yamamoto, S. Waguri, K. Tanaka, and M. Komatsu. 2014. Proteasome dysfunction activates autophagy and the Keap1-Nrf2 pathway. J. Biol. Chem. 289: 24944– 24955. 211. Pickering, A. M., R. A. Linder, H. Zhang, H. J. Forman, and K. J. A. Davies. 2012. Nrf2-dependent induction of proteasome and Pa28αβ regulator are required for adaptation to oxidative stress. J. Biol. Chem. 287: 10021–10031. 212. Kwak, M.-K., N. Wakabayashi, J. L. Greenlaw, M. Yamamoto, and T. W. Kensler. 2003. Antioxidants Enhance Mammalian Proteasome Expression through the Keap1-Nrf2 Signaling Pathway. Mol. Cell. Biol. 23: 8786–8794. 213. Walerych, D., K. Lisek, R. Sommaggio, S. Piazza, Y. Ciani, E. Dalla, K. Rajkowska, K. Gaweda-Walerych, E. Ingallina, C. Tonelli, M. J. Morelli, A. Amato, V. Eterno, A. Zambelli, A. Rosato, B. Amati, J. R. Winiewski, and G. Del Sal. 2016. Proteasome machinery is instrumental in a common gain-of-function program of the p53 missense mutants in cancer. Nat. Cell Biol. 18: 897–909. 214. Meiners, S., D. Heyken, A. Weller, A. Ludwig, K. Stangl, P. M. Kloetzel, and E. Krüger. 2003. Inhibition of proteasome activity induces concerted expression of proteasome genes and de novo formation of mammalian proteasomes. J. Biol. Chem. 278: 21517–21525. 148

References

215. Radhakrishnan, S. K., C. S. Lee, P. Young, A. Beskow, J. Y. Chan, and R. J. Deshaies. 2010. Transcription Factor Nrf1 Mediates the Proteasome Recovery Pathway after Proteasome Inhibition in Mammalian Cells. Mol. Cell 38: 17–28. 216. Sha, Z., and A. L. Goldberg. 2014. Proteasome-mediated processing of Nrf1 is essential for coordinate induction of all proteasome subunits and p97. Curr. Biol. 24: 1573–1583. 217. Zhang, Y., D. H. Crouch, M. Yamamoto, and J. D. Hayes. 2006. Negative regulation of the Nrf1 transcription factor by its N-terminal domain is independent of Keap1: Nrf1, but not Nrf2, is targeted to the endoplasmic reticulum. Biochem. J. 399: 373–385. 218. Wang, W., and J. Y. Chan. 2006. Nrf1 is targeted to the endoplasmic reticulum membrane by an N-terminal transmembrane domain: Inhibition of nuclear translocation and transacting function. J. Biol. Chem. 281: 19676–19687. 219. Chen, J., X. Liu, F. Lü, X. Liu, Y. Ru, Y. Ren, L. Yao, and Y. Zhang. 2015. Transcription factor Nrf1 is negatively regulated by its O- GlcNAcylation status. FEBS Lett. 589: 2347–2358. 220. Tomlin, F. M., U. I. M. Gerling-Driessen, Y. C. Liu, R. A. Flynn, J. R. Vangala, C. S. Lentz, S. Clauder-Muenster, P. Jakob, W. F. Mueller, D. Ordoñez-Rueda, M. Paulsen, N. Matsui, D. Foley, A. Rafalko, T. Suzuki, M. Bogyo, L. M. Steinmetz, S. K. Radhakrishnan, and C. R. Bertozzi. 2017. Inhibition of NGLY1 Inactivates the Transcription Factor Nrf1 and Potentiates Proteasome Inhibitor Cytotoxicity. ACS Cent. Sci. 3: 1143– 1155. 221. Biswas, M., D. Phan, M. Watanabe, and J. Y. Chan. 2011. The Fbw7 tumor suppressor regulates nuclear factor E2-related dactor 1 transcription factor turnover through proteasome-mediated proteolysis. J. Biol. Chem. 286: 39282–39289. 222. Radhakrishnan, S. K., W. den Besten, and R. J. Deshaies. 2014. p97-dependent retrotranslocation and proteolytic processing govern formation of active Nrf1 upon proteasome inhibition. Elife 3: e01856. 223. Sha, Z., and A. L. Goldberg. 2016. Reply to Vangala et al.: Complete inhibition of the proteasome reduces new proteasome production by causing Nrf1 aggregation. Curr. Biol. 26: R836–R837. 224. Vangala, J. R., F. Sotzny, E. Krüger, R. J. Deshaies, and S. K. Radhakrishnan. 2016. Nrf1 can be processed and activated in a proteasome- independent manner. Curr. Biol. 26: R834–R835. 225. Koizumi, S., T. Irie, S. Hirayama, Y. Sakurai, H. Yashiroda, I. Naguro, H. Ichijo, J. Hamazaki, and S. Murata. 2016. The aspartyl protease DDI2 activates Nrf1 to compensate for proteasome dysfunction. Elife 5: 1–10. 226. Ohtsuji, M., F. Katsuoka, A. Kobayashi, H. Aburatani, J. D. Hayes, and M. Yamamoto. 2008. Nrf1 and Nrf2 play distinct roles in activation of antioxidant response element-dependent genes. J. Biol. Chem. 283: 33554–33562. 227. Lee, C. S., D. V. Ho, and J. Y. Chan. 2013. Nuclear factor-erythroid 2-related factor 1 regulates expression of proteasome genes in hepatocytes and protects against endoplasmic reticulum stress and steatosis in mice. FEBS J. 280: 3609–3620. 228. Hirotsu, Y., C. Higashi, T. Fukutomi, F. Katsuoka, T. Tsujita, Y. Yagishita, Y. Matsuyama, H. Motohashi, A. Uruno, and M. Yamamoto. 2014. Transcription factor NF-E2-related factor 1 impairs glucose metabolism in mice. Genes to Cells 19: 650–665. 229. Widenmaier, S. B., N. A. Snyder, T. B. Nguyen, A. Arduini, G. Y. Lee, A. P. Arruda, J. Saksi, A. Bartelt, and G. S. Hotamisligil. 2017. NRF1 Is an ER Membrane Sensor that Is Central to Cholesterol Homeostasis. Cell 171: 1094.e15-1101. 230. Kozutsumi, Y., M. Segal, K. Normington, M.-J. Gething, and J. Sambrook. 1988. The presence of malfolded proteins in the endoplasmic reticulum signals the induction of glucose-regulated proteins. Nature 332: 462–464. 231. Mori, K., W. Ma, M. J. Gething, and J. Sambrook. 1993. A transmembrane protein with a cdc2+CDC28-related kinase activity is required for signaling from the ER to the nucleus. Cell 74: 743–756. 232. Cox, J. S., C. E. Shamu, and P. Walter. 1993. Transcriptional Induction of Genes Encoding Endoplasmic-Reticulum Resident Proteins Requires a Transmembrane Protein-Kinase. Cell 73: 1197–1206. 233. Bertolotti, A. 2018. Importance of the subcellular location of protein deposits in neurodegenerative diseases. Curr. Opin. Neurobiol. 51: 127– 133. 234. Hetz, C., E. Chevet, and S. A. Oakes. 2015. Proteostasis control by the unfolded protein response. Nat. Cell Biol. 17: 829–838. 235. Harding, H. P., I. Novoa, Y. Zhang, H. Zeng, R. Wek, M. Schapira, and D. Ron. 2000. Regulated translation initiation controls stress-induced gene expression in mammalian cells. Mol. Cell 6: 1099–1108. 236. Donnelly, N., A. M. Gorman, S. Gupta, and A. Samali. 2013. The eIF2α kinases: Their structures and functions. Cell. Mol. Life Sci. 70: 3493– 3511. 237. Cullinan, S. B., D. Zhang, M. Hannink, E. Arvisais, R. J. Kaufman, and J. A. Diehl. 2003. Nrf2 Is a Direct PERK Substrate and Effector of PERK-Dependent Cell Survival. Mol. Cell. Biol. 23: 7198–7209. 238. Acosta-Alvear, D., Y. Zhou, A. Blais, M. Tsikitis, N. H. Lents, C. Arias, C. J. Lennon, Y. Kluger, and B. D. Dynlacht. 2007. XBP1 Controls Diverse Cell Type- and Condition-Specific Transcriptional Regulatory Networks. Mol. Cell 27: 53–66. 239. Darling, N. J., and S. J. Cook. 2014. The role of MAPK signalling pathways in the response to endoplasmic reticulum stress. Biochim. Biophys. Acta - Mol. Cell Res. 1843: 2150–2163. 240. Ye, J., R. B. Rawson, R. Komuro, X. Chen, U. P. Davé, R. Prywes, M. S. Brown, and J. L. Goldstein. 2000. ER stress induces cleavage of membrane-bound ATF6 by the same proteases that process SREBPs. Mol. Cell 6: 1355–1364. 241. Aronov, M., and B. Tirosh. 2016. Metabolic Control of Plasma Cell Differentiation- What We Know and What We Don’t Know. J. Clin. Immunol. 36: 12–17. 242. Ribatti, D. 2017. The discovery of plasma cells: An historical note. Immunol. Lett. 188: 64–67. 243. Smith, J. A. 2018. Regulation of cytokine production by the unfolded protein response; Implications for infection and autoimmunity. Front. Immunol. 9: 1–21. 244. Iwata, A., B. E. Riley, J. A. Johnston, and R. R. Kopito. 2005. HDAC6 and microtubules are required for autophagic degradation of aggregated Huntingtin. J. Biol. Chem. 280: 40282–40292. 245. Stalder, T., S. Hahn, and P. Erb. 1994. Fas antigen is the major target molecule for CD4+ T cell-mediated cytotoxicity. J. Immunol. 152: 1127–1133. 246. Ju, S. T., H. Cui, D. J. Panka, R. Ettinger, and A. Marshak-Rothstein. 1994. Participation of target Fas protein in apoptosis pathway induced by CD4+ Th1 and CD8+ cytotoxic T cells. Proc. Natl. Acad. Sci. 91: 4185–4189. 247. Ma, Y., J. W. Brewer, J. Alan Diehl, and L. M. Hendershot. 2002. Two distinct stress signaling pathways converge upon the CHOP promoter during the mammalian unfolded protein response. J. Mol. Biol. 318: 1351–1365. 248. Mccullough, K. D., J. L. Martindale, T. Aw, N. J. Holbrook, K. D. M. C. Cullough, and L. Klotz. 2001. Gadd153 Sensitizes Cells to Endoplasmic Reticulum Stress by Down-Regulating Bcl2 and Perturbing the Cellular Redox State Gadd153 Sensitizes Cells to Endoplasmic Reticulum Stress by Down-Regulating Bcl2 and Perturbing the Cellular Redox State. Mol. Cell. Biol. 21: 1249–1259. 249. Matsumoto, M., M. Minami, K. Takeda, Y. Sakao, and S. Akira. 1996. Ectopic expression of CHOP (GADD153) induces apoptosis in M1 myeloblastic leukemia cells. FEBS Lett. 395: 143–147. 250. Catz, S. D., and J. L. Johnson. 2001. Transcriptional regulation of bcl-2 by nuclear factor κB and its significance in prostate cancer. Oncogene 149

References

20: 7342–7351. 251. Taylor, R. C., S. P. Cullen, and S. J. Martin. 2008. Apoptosis: Controlled demolition at the cellular level. Nat. Rev. Mol. Cell Biol. 9: 231–241. 252. D’Amours, D., F. R. Sallmann, V. M. Dixit, and G. G. Poirier. 2001. Gain-of-function of poly(ADP-ribose) polymerase-1 upon cleavage by apoptotic proteases: implications for apoptosis. J. Cell Sci. 114: 3771–3778. 253. Green, D. R., N. Droin, and M. Pinkoski. 2003. Activation-induced cell death in T cells. Immunol. Rev. 193: 70–81. 254. Brunner, T., R. J. Mogil, D. LaFace, N. J. Yoo, A. Mahboubi, F. Echeverri, S. J. Martin, W. R. Force, D. H. Lynch, and C. F. Ware. 1995. Cell-autonomous Fas (CD95)/Fas-ligand interaction mediates activation-induced apoptosis in T-cell hybridomas. Nature 373: 441–4. 255. Ju, S. Te, D. J. Panka, H. Cui, R. Ettinger, M. Ei-Khatib, D. H. Sherrf, B. Z. Stanger, and A. Marshak-Rothstein. 1995. Fas(CD95)/FasL interactions required for programmed cell death after T-cell activation. Nature 373: 444–448. 256. Brenner, D., A. Golks, M. Becker, W. Müller, C. R. Frey, R. Novak, D. Melamed, F. Kiefer, P. H. Krammer, and R. Arnold. 2007. Caspase- cleaved HPK1 induces CD95L-independent activation-induced cell death in T and B lymphocytes. Blood 110: 3968–77. 257. Goldberg, A. L. 2012. Development of proteasome inhibitors as research tools and cancer drugs. J. Cell Biol. 199: 583–588. 258. Vinitsky, A., C. Michaud, J. C. Powers, and M. Orlowski. 1992. Inhibition of the chymotrypsin-like activity of the pituitary multicatalytic proteinase complex. Biochemistry 31: 9421–8. 259. Hayashi, M., Y. Saito, and S. Kawashima. 1992. Calpain activation is essential for membrane fusion of erythrocytes in the presence of exogenous Ca2+. Biochem. Biophys. Res. Commun. 182: 939–946. 260. Ito, A., R. Takahashi, C. Muira, and Y. Baba. 1975. Synthetic Study of Peptide Aldehydes. Chem. Pharm. Bull. 12: 3106–3113. 261. Tsubuki, S., H. Kawasaki, Y. Saito, N. Miyashita, M. Inomata, and S. Kawashima. 1993. Purification and characterization of a Z-Leu-Leu- Leu-MCA degrading protease expected to regulate neurite formation: A novel catalytic activity in proteasome. Biochem. Biophys. Res. Commun. 196: 1195–1201. 262. Rock, K. L., C. Gramm, L. Rothstein, K. Clark, R. Stein, L. Dick, D. Hwang, and A. L. Goldberg. 1994. Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class I molecules. Cell 78: 761–771. 263. Palombella, V. J., O. J. Rando, A. L. Goldberg, and T. Maniatis. 1994. The ubiquitinproteasome pathway is required for processing the NF- κB1 precursor protein and the activation of NF-κB. Cell 78: 773–785. 264. Traenckner, E. B., S. Wilk, and P. A. Baeuerle. 1994. A proteasome inhibitor prevents activation of NF-κB and stabilizes a newly phosphorylated form of IκBα that is still bound to NF-κB. EMBO J. 13: 5433–41. 265. Adams, J., M. Behnke, S. Chen, A. A. Cruickshank, L. R. Dick, L. Grenier, J. M. Klunder, Y. T. Ma, L. Plamondon, and R. L. Stein. 1998. Potent and selective inhibitors of the proteasome: Dipeptidyl boronic acids. Bioorganic Med. Chem. Lett. 8: 333–338. 266. Kane, R. C., A. T. Farrell, R. Sridhara, and R. Pazdur. 2006. United States Food and Drug Administration approval summary: Bortezomib for the treatment of progressive multiple myeloma after one prior therapy. Clin. Cancer Res. 12: 2955–2960. 267. Adams, J., V. J. Palombella, E. A. Sausville, J. Johnson, A. Destree, D. D. Lazarus, J. Maas, C. S. Pien, S. Prakash, and P. J. Elliott. 1999. Proteasome Inhibitors : A Novel Class of Potent and Effective Antitumor Agents. 2615–2622. 268. Cowan, A. J., C. Allen, A. Barac, H. Basaleem, I. Bensenor, M. P. Curado, K. Foreman, R. Gupta, J. Harvey, H. D. Hosgood, M. Jakovljevic, Y. Khader, S. Linn, D. Lad, L. Mantovani, V. M. Nong, A. Mokdad, M. Naghavi, M. Postma, G. Roshandel, K. Shackelford, M. Sisay, C. T. Nguyen, T. T. Tran, B. T. Xuan, K. N. Ukwaja, S. E. Vollset, E. Weiderpass, E. N. Libby, and C. Fitzmaurice. 2018. Global Burden of Multiple Myeloma: A Systematic Analysis for the Global Burden of Disease Study 2016. JAMA Oncol. 98121: 1–7. 269. Xu, Y., X. Yang, Y. Chen, H. Chen, H. Sun, W. Li, Q. Xie, L. Yu, and L. Shao. 2018. Discovery of novel 20S proteasome inhibitors by rational topology-based scaffold hopping of bortezomib. Bioorganic Med. Chem. Lett. 0–1. 270. Kuhn, D. J., Q. Chen, P. M. Voorhees, J. S. Strader, K. D. Shenk, C. M. Sun, S. D. Demo, M. K. Bennett, F. W. B. van Leeuwen, A. A. Chanan-Khan, and R. Z. Orlowski. 2007. Potent activity of carfilzomib, a novel, irreversible inhibitor of the ubiquitin-proteasome pathway, against preclinical models of multiple myeloma. Blood 110: 3281–90. 271. Shah, C., R. Bishnoi, Y. Wang, F. Zou, H. Bejjanki, S. Master, and J. S. Moreb. 2018. Efficacy and safety of carfilzomib in relapsed and/or refractory multiple myeloma: Systematic review and meta-analysis of 14 trials. Oncotarget 9: 23704–23717. 272. Huber, E. M., and M. Groll. 2012. Inhibitors for the immuno- and constitutive proteasome: Current and future trends in drug development. Angew. Chemie - Int. Ed. 51: 8708–8720. 273. Huber, E. M., W. Heinemeyer, and M. Groll. 2015. Bortezomib-resistant mutant proteasomes: Structural and biochemical evaluation with carfilzomib and ONX 0914. Structure 23: 407–417. 274. Harshbarger, W., C. Miller, C. Diedrich, and J. Sacchettini. 2015. Crystal structure of the human 20S proteasome in complex with carfilzomib. Structure 23: 418–424. 275. Schrader, J., F. Henneberg, R. A. Mata, K. Tittmann, T. R. Schneider, H. Stark, G. Bourenkov, and A. Chari. 2016. The inhibition mechanism of human 20S proteasomes enables next-generation inhibitor design. Science . 353: 594–8. 276. Shirley, M. 2016. Ixazomib: First global approval. Drugs 76: 405–411. 277. Ettari, R., M. Zappalà, S. Grasso, C. Musolino, V. Innao, and A. Allegra. 2018. Immunoproteasome-selective and non-selective inhibitors: A promising approach for the treatment of multiple myeloma. Pharmacol. Ther. 182: 176–192. 278. Fenteany, G., R. F. Standaert, W. S. Lane, S. Choi, E. J. Corey, and S. L. Schreiber. 1995. Inhibition of proteasome activities and subunit- specific amino-terminal threonine modification by lactacystin. Science . 268: 726–31. 279. Dick, L. R., A. A. Cruikshank, L. Grenier, F. D. Melandri, S. L. Nunes, and R. L. Stein. 1996. Mechanistic Studies on the Inactivation of the Proteasome by Lactacystin. J. Biol. Chem. 271: 7273–7276. 280. Rockwell, C. E., and N. Qureshi. 2010. Differential effects of lactacystin on cytokine production in activated Jurkat cells and murine splenocytes. Cytokine 51: 12–7. 281. Rockwell, C. E., J. J. Monaco, and N. Qureshi. 2012. A critical role for the inducible proteasomal subunits LMP7 and MECL1 in cytokine production by activated murine splenocytes. Pharmacology 89: 117–26. 282. Delic, J., P. Masdehors, S. Omura, J. M. Cosset, J. Dumont, J. L. Binet, and H. Magdelénat. 1998. The proteasome inhibitor lactacystin induces apoptosis and sensitizes chemo- and radioresistant human chronic lymphocytic leukaemia lymphocytes to TNF-alpha-initiated apoptosis. Br. J. Cancer 77: 1103–7. 283. Hanada, M., K. Sugawara, K. Kaneta, S. Toda, Y. Nishiyama, K. Tomita, H. Yamamoto, M. Konishi, and T. Oki. 1992. Epoxomicin, a new antitumor agent of microbial origin. J. Antibiot. (Tokyo). 45: 1746–1752. 284. Meng, L., R. Mohan, B. H. B. Kwok, M. Elofsson, N. Sin, and C. M. Crews. 1999. Epoxomicin, a potent and selective proteasome inhibitor, exhibits in vivo antiinflammatory activity. Proc. Natl. Acad. Sci. 96: 10403–10408. 285. Kim, K. B., and C. M. Crews. 2013. From epoxomicin to carfilzomib: chemistry, biology, and medical outcomes. Nat. Prod. Rep. 30: 600. 286. Ho, Y. K. (Abby), P. Bargagna-Mohan, M. Wehenkel, R. Mohan, and K. B. Kim. 2007. LMP2-Specific Inhibitors: Chemical Genetic Tools for Proteasome Biology. Chem. Biol. 14: 419–430. 150

References

287. Wehenkel, M., J.-O. Ban, Y.-K. Ho, K. C. Carmony, J. T. Hong, and K. B. Kim. 2012. A selective inhibitor of the immunoproteasome subunit LMP2 induces apoptosis in PC-3 cells and suppresses tumour growth in nude mice. Br. J. Cancer 107: 53–62. 288. Kalim, K. W., M. Basler, C. J. Kirk, and M. Groettrup. 2012. Immunoproteasome subunit LMP7 deficiency and inhibition suppresses Th1 and Th17 but enhances regulatory T cell differentiation. J. Immunol. 189: 4182–93. 289. Singh, A. V., M. Bandi, M. A. Aujay, C. J. Kirk, D. E. Hark, N. Raje, D. Chauhan, and K. C. Anderson. 2011. PR-924, a selective inhibitor of the immunoproteasome subunit LMP-7, blocks multiple myeloma cell growth both in vitro and in vivo. Br. J. Haematol. 152: 155–163. 290. de Bruin, G., E. M. Huber, B. Xin, E. J. Van Rooden, K. Al-ayed, K. Kim, A. F. Kisselev, C. Driessen, M. Van Der Stelt, G. A. Van Der Marel, M. Groll, and H. S. Overkleeft. 2014. Structure-Based Design of β 1i or β 5i Speci fi c Inhibitors of Human Immunoproteasomes. J. Med. Chem. 57: 6197–6209. 291. Xin, B. T., G. De Bruin, E. M. Huber, A. Besse, B. I. Florea, D. V. Filippov, G. A. Van Der Marel, A. F. Kisselev, M. Van Der Stelt, C. Driessen, M. Groll, and H. S. Overkleeft. 2016. Structure-Based Design of ß5c Selective Inhibitors of Human Constitutive Proteasomes. J. Med. Chem. 59: 7177–7187. 292. Fan, H., N. G. Angelo, J. D. Warren, C. F. Nathan, and G. Lin. 2014. Oxathiazolones selectively inhibit the human immunoproteasome over the constitutive proteasome. ACS Med. Chem. Lett. 5: 405–410. 293. Sula Karreci, E., H. Fan, M. Uehara, A. B. Mihali, P. K. Singh, A. T. Kurdi, Z. Solhjou, L. V. Riella, I. Ghobrial, T. Laragione, S. Routray, J. P. Assaker, R. Wang, G. Sukenick, L. Shi, F. J. Barrat, C. F. Nathan, G. Lin, and J. Azzi. 2016. Brief treatment with a highly selective immunoproteasome inhibitor promotes long-term cardiac allograft acceptance in mice. Proc. Natl. Acad. Sci. 201618548. 294. Singh, P. K., H. Fan, X. Jiang, L. Shi, C. F. Nathan, and G. Lin. 2016. Immunoproteasome β5i-Selective Dipeptidomimetic Inhibitors. ChemMedChem 2127–2131. 295. Santos, R. D. L. A., L. Bai, P. K. Singh, N. Murakami, H. Fan, W. Zhan, Y. Zhu, X. Jiang, K. Zhang, J. P. Assker, C. F. Nathan, H. Li, J. Azzi, and G. Lin. 2017. Structure of human immunoproteasome with a reversible and noncompetitive inhibitor that selectively inhibits activated lymphocytes. Nat. Commun. 8: 1–10. 296. Torrelo, A., S. Patel, I. Colmenero, D. Gurbindo, F. Lendínez, A. Hernández, J. C. López-Robledillo, A. Dadban, L. Requena, and A. S. Paller. 2010. Chronic atypical neutrophilic dermatosis with lipodystrophy and elevated temperature (CANDLE) syndrome. J. Am. Acad. Dermatol. 62: 489–95. 297. Torrelo, A. 2017. CANDLE Syndrome As a Paradigm of Proteasome-Related Autoinflammation. Front. Immunol. 8: 1–9. 298. Kunimoto, K., A. Kimura, K. Uede, M. Okuda, N. Aoyagi, F. Furukawa, and N. Kanazawa. 2013. A New Infant Case of Nakajo-Nishimura Syndrome with a Genetic Mutation in the Immunoproteasome Subunit: An Overlapping Entity with JMP and CANDLE Syndrome Related to PSMB8 Mutations. Dermatology 0012: 1–5. 299. Al-Mayouf, S. M., A. AlSaleem, N. AlMutairi, A. AlSonbul, T. Alzaid, A. M. Alazami, and H. Al-Mousa. 2018. Monogenic interferonopathies: Phenotypic and genotypic findings of CANDLE syndrome and its overlap with C1q deficient SLE. Int. J. Rheum. Dis. 21: 208– 213. 300. Liu, Y., Y. Ramot, A. Torrelo, A. S. Paller, N. Si, S. Babay, P. W. Kim, A. Sheikh, C. C. R. Lee, Y. Chen, A. Vera, X. Zhang, R. Goldbach- Mansky, and A. Zlotogorski. 2012. Mutations in proteasome subunit β type 8 cause chronic atypical neutrophilic dermatosis with lipodystrophy and elevated temperature with evidence of genetic and phenotypic heterogeneity. Arthritis Rheum. 64: 895–907. 301. Kitamura, A., Y. Maekawa, H. Uehara, K. Izumi, I. Kawachi, M. Nishizawa, Y. Toyoshima, H. Takahashi, D. M. Standley, K. Tanaka, J. Hamazaki, S. Murata, K. Obara, I. Toyoshima, and K. Yasutomo. 2011. A mutation in the immunoproteasome subunit PSMB8 causes autoinflammation and lipodystrophy in humans. J. Clin. Invest. 121: 4150–60. 302. Arima, K., A. Kinoshita, H. Mishima, N. Kanazawa, T. Kaneko, T. Mizushima, K. Ichinose, H. Nakamura, A. Tsujino, A. Kawakami, M. Matsunaka, S. Kasagi, S. Kawano, S. Kumagai, K. Ohmura, T. Mimori, M. Hirano, S. Ueno, K. Tanaka, M. Tanaka, I. Toyoshima, H. Sugino, A. Yamakawa, K. Tanaka, N. Niikawa, F. Furukawa, S. Murata, K. Eguchi, H. Ida, and K.-I. Yoshiura. 2011. Proteasome assembly defect due to a proteasome subunit beta type 8 (PSMB8) mutation causes the autoinflammatory disorder, Nakajo-Nishimura syndrome. Proc. Natl. Acad. Sci. U. S. A. 108: 14914–9. 303. Kimura, H., F. Usui, T. Karasawa, A. Kawashima, K. Shirasuna, Y. Inoue, T. Komada, M. Kobayashi, Y. Mizushina, T. Kasahara, K. Suzuki, Y. Iwasaki, T. Yada, P. Caturegli, and M. Takahashi. 2015. Immunoproteasome subunit LMP7 Deficiency Improves Obesity and Metabolic Disorders. Sci. Rep. 5: 15883. 304. Arimochi, H., Y. Sasaki, A. Kitamura, and K. Yasutomo. 2016. Differentiation of preadipocytes and mature adipocytes requires PSMB8. Sci. Rep. 6: 1–8. 305. Hipp, M. S., S. H. Park, and U. U. Hartl. 2014. Proteostasis impairment in protein-misfolding and -aggregation diseases. Trends Cell Biol. 24: 506–514. 306. Schmidt, M., and D. Finley. 2014. Regulation of proteasome activity in health and disease. Biochim. Biophys. Acta - Mol. Cell Res. 1843: 13– 25. 307. Cheroni, C., M. Marino, M. Tortarolo, P. Veglianese, S. De Biasi, E. Fontana, L. V. Zuccarello, C. J. Maynard, N. P. Dantuma, and C. Bendotti. 2009. Functional alterations of the ubiquitin-proteasome system in motor neurons of a mouse model of familial amyotrophic lateral sclerosis. Hum. Mol. Genet. 18: 82–96. 308. Kabashi, E., J. N. Agar, Y. Hong, D. M. Taylor, S. Minotti, D. A. Figlewicz, and H. D. Durham. 2008. Proteasomes remain intact, but show early focal alteration in their composition in a mouse model of amyotrophic lateral sclerosis. J. Neurochem. 105: 2353–2366. 309. Nardo, G., M. C. Trolese, and C. Bendotti. 2016. Major histocompatibility complex I expression by motor neurons and its implication in amyotrophic lateral sclerosis. Front. Neurol. 7: 1–15. 310. Hipp, M. S., C. N. Patel, K. Bersuker, B. E. Riley, S. E. Kaiser, T. A. Shaler, M. Brandeis, and R. R. Kopito. 2012. Indirect inhibition of 26S proteasome activity in a cellular model of Huntington’s disease. J. Cell Biol. 196: 573–587. 311. Díaz-Hernández, M., F. Hernández, E. Martín-Aparicio, P. Gómez-Ramos, M. a Morán, J. G. Castaño, I. Ferrer, J. Avila, and J. J. Lucas. 2003. Neuronal induction of the immunoproteasome in Huntington’s disease. J. Neurosci. 23: 11653–11661. 312. Díaz-Hernández, M., E. Martín-Aparicio, J. Avila, F. Hernández, and J. J. Lucas. 2004. Enhaced induction of the immunoproteasome by interferon gamma in neurons expressing mutant huntingtin. Neurotox. Res. 6: 463–468. 313. Mishto, M., E. Bellavista, A. Santoro, A. Stolzing, C. Ligorio, B. Nacmias, L. Spazzafumo, M. Chiappelli, F. Licastro, S. Sorbi, A. Pession, T. Ohm, T. Grune, and C. Franceschi. 2006. Immunoproteasome and LMP2 polymorphism in aged and Alzheimer’s disease brains. Neurobiol. Aging 27: 54–66. 314. Wagner, L. K., K. E. Gilling, E. Schormann, P. M. Kloetzel, F. L. Heppner, E. Krüger, and S. Prokop. 2017. Immunoproteasome deficiency alters microglial cytokine response and improves cognitive deficits in Alzheimer’s disease-like APPPS1 mice. Acta Neuropathol. Commun. 5: 52. 315. Orre, M., W. Kamphuis, S. Dooves, L. Kooijman, E. T. Chan, C. J. Kirk, V. Dimayuga Smith, S. Koot, C. Mamber, A. H. Jansen, H. Ovaa, and E. M. Hol. 2013. Reactive glia show increased immunoproteasome activity in Alzheimer’s disease. Brain 136: 1415–1431. 151

References

316. Ferrer, I., M. B. Rovira, M. L. S. Guerra, M. J. Rey, and F. Costa-Jussá. 2004. Neuropathology and Pathogenesis of Encephalitis following Amyloid β Immunization in Alzheimer’s Disease. Brain Pathol. 14: 11–20. 317. Giannini, C., A. Kloß, S. Gohlke, M. Mishto, T. P. Nicholson, P. W. Sheppard, P. M. Kloetzel, and B. Dahlmann. 2013. Poly-Ub-Substrate- Degradative Activity of 26S Proteasome Is Not Impaired in the Aging Rat Brain. PLoS One 8. 318. Thibaudeau, T. A., R. T. Anderson, and D. M. Smith. 2018. A common mechanism of proteasome impairment by neurodegenerative disease- associated oligomers. Nat. Commun. 9. 319. Lee, B. H., M. J. Lee, S. Park, D. C. Oh, S. Elsasser, P. C. Chen, C. Gartner, N. Dimova, J. Hanna, S. P. Gygi, S. M. Wilson, R. W. King, and D. Finley. 2010. Enhancement of proteasome activity by a small-molecule inhibitor of USP14. Nature 467: 179–184. 320. Boselli, M., B. H. Lee, J. Robert, M. A. Prado, S. W. Min, C. Cheng, M. Catarina Silva, C. Seong, S. Elsasser, K. M. Hatle, T. C. Gahman, S. P. Gygi, S. J. Haggarty, L. Gan, R. W. King, and D. Finley. 2017. An inhibitor of the proteasomal deubiquitinating enzyme USP14 induces tau elimination in cultured neurons. J. Biol. Chem. 292: 19209–19225. 321. Lokireddy, S., N. V. Kukushkin, and A. L. Goldberg. 2015. cAMP-induced phosphorylation of 26S proteasomes on Rpn6/PSMD11 enhances their activity and the degradation of misfolded proteins. Proc. Natl. Acad. Sci. 112: E7176–E7185. 322. Myeku, N., and K. E. Duff. 2018. Targeting the 26S Proteasome To Protect Against Proteotoxic Diseases. Trends Mol. Med. 24: 18–29. 323. Uversky, V. N., C. J. Oldfield, and A. K. Dunker. 2008. Intrinsically Disordered Proteins in Human Diseases: Introducing the D2 Concept. Annu. Rev. Biophys. 37: 215–246. 324. Tofaris, G. K., R. Layfield, and M. G. Spillantini. 2001. α-Synuclein metabolism and aggregation is linked to ubiquitin-independent degradation by the proteasome. FEBS Lett. 509: 22–26. 325. Zondler, L., M. Kostka, P. Garidel, U. Heinzelmann, B. Hengerer, B. Mayer, J. H. Weishaupt, F. Gillardon, and K. M. Danzer. 2017. Proteasome impairment by α-synuclein. PLoS One 12: e0184040. 326. Mo, M.-S., W. Huang, C.-C. Sun, L.-M. Zhang, L. Cen, Y.-S. Xiao, G.-F. Li, X.-L. Yang, S.-G. Qu, and P.-Y. Xu. 2016. Association Analysis of Proteasome Subunits and Transporter Associated with Antigen Processing on Chinese Patients with Parkinson’s Disease. Chin. Med. J. (Engl). 129: 1053–1058. 327. Wahl, C., S. Kautzmann, G. Krebiehl, K. Strauss, D. Woitalla, T. Müller, P. Bauer, O. Riess, and R. Krüger. 2008. A comprehensive genetic study of the proteasomal subunit S6 ATPase in German Parkinson’s disease patients. J. Neural Transm. 115: 1141–1148. 328. Mundt, S., B. Engelhardt, C. J. Kirk, M. Groettrup, and M. Basler. 2016. Inhibition and deficiency of the immunoproteasome subunit LMP7 attenuates LCMV-induced meningitis. Eur. J. Immunol. 46: 104–113. 329. Moritz, K. E., N. M. McCormack, M. B. Abera, C. Viollet, Y. J. Yauger, G. Sukumar, C. L. Dalgard, and B. G. Burnett. 2017. The role of the immunoproteasome in interferon-γ-mediated microglial activation. Sci. Rep. 7: 1–16. 330. Basler, M., U. Beck, C. J. Kirk, and M. Groettrup. 2011. The antiviral immune response in mice devoid of immunoproteasome activity. J. Immunol. 187: 5548–57. 331. Althof, N., C. C. Goetzke, M. Kespohl, K. Voss, A. Heuser, S. Pinkert, Z. Kaya, K. Klingel, and A. Beling. 2018. The immunoproteasome- specific inhibitor ONX 0914 reverses susceptibility to acute viral myocarditis. EMBO Mol. Med. 1–19. 332. Inforzato, A., A. Doni, I. Barajon, R. Leone, C. Garlanda, B. Bottazzi, and A. Mantovani. 2013. PTX3 as a paradigm for the interaction of pentraxins with the Complement system. Semin. Immunol. 25: 79–85. 333. Kirschner, F., K. Reppe, N. Andresen, M. Witzenrath, F. Ebstein, and P.-M. Kloetzel. 2016. Proteasome β5i Subunit Deficiency Affects Opsonin Synthesis and Aggravates Pneumococcal Pneumonia. PLoS One 11: e0153847. 334. Reis, J., F. Hassan, X. Q. Guan, J. Shen, J. J. Monaco, C. J. Papasian, A. a Qureshi, C. W. Van Way, S. N. Vogel, D. C. Morrison, and N. Qureshi. 2011. The immunoproteasomes regulate LPS-induced TRIF/TRAM signaling pathway in murine macrophages. Cell Biochem. Biophys. 60: 119–26. 335. Strehl, B., T. Joeris, M. Rieger, A. Visekruna, K. Textoris-Taube, S. H. E. Kaufmann, P.-M. Kloetzel, U. Kuckelkorn, and U. Steinhoff. 2006. Immunoproteasomes are essential for clearance of Listeria monocytogenes in nonlymphoid tissues but not for induction of bacteria-specific CD8+ T cells. J. Immunol. 177: 6238–44. 336. Ersching, J., J. R. Vasconcelos, C. P. Ferreira, B. C. Caetano, A. V. Machado, O. Bruna–Romero, M. A. Baron, L. R. P. Ferreira, E. Cunha- Neto, K. L. Rock, R. T. Gazzinelli, and M. M. Rodrigues. 2016. The Combined Deficiency of Immunoproteasome Subunits Affects Both the Magnitude and Quality of Pathogen- and Genetic Vaccination-Induced CD8+ T Cell Responses to the Human Protozoan Parasite Trypanosoma cruzi. PLOS Pathog. 12: e1005593. 337. Mundt, S., M. Basler, S. Buerger, H. Engler, and M. Groettrup. 2016. Inhibiting the immunoproteasome exacerbates the pathogenesis of systemic Candida albicans infection in mice. Sci. Rep. 6: 19434. 338. Richardson, J. P., and D. L. Moyes. 2015. Adaptive immune responses to Candida albicans infection. Virulence 6: 327–337. 339. Kumar, S. K., J. H. Lee, J. J. Lahuerta, G. Morgan, P. G. Richardson, J. Crowley, J. Haessler, J. Feather, A. Hoering, P. Moreau, X. Leleu, C. Hulin, S. K. Klein, P. Sonneveld, D. Siegel, J. Bladé, H. Goldschmidt, S. Jagannath, J. S. Miguel, R. Orlowski, A. Palumbo, O. Sezer, S. V. Rajkumar, and B. G. M. Durie. 2012. Risk of progression and survival in multiple myeloma relapsing after therapy with IMiDs and bortezomib: A multicenter international myeloma working group study. Leukemia 26: 149–157. 340. Thompson, J. L. 2013. Carfilzomib: A second-generation proteasome inhibitor for the treatment of relapsed and refractory multiple myeloma. Ann. Pharmacother. 47: 56–62. 341. Zinzani, P. L., C. Pellegrini, E. Merla, F. Ballerini, A. Fabbri, A. Guarini, V. Pavone, G. Quintini, B. Puccini, M. L. Vigliotti, V. Stefoni, E. Derenzini, A. Broccoli, L. Gandolfi, F. Quirini, B. Casadei, L. Argnani, and M. Baccarani. 2013. Bortezomib as salvage treatment for heavily pretreated relapsed lymphoma patients: a multicenter retrospective study. Hematol. Oncol. 31: 179–182. 342. Stapnes, C., A. P. Døskeland, K. Hatfield, E. Ersvær, A. Ryningen, J. B. Lorens, B. T. Gjertsen, and Ø. Bruserud. 2007. The proteasome inhibitors bortezomib and PR-171 have antiproliferative and proapoptotic effects on primary human acute myeloid leukaemia cells. Br. J. Haematol. 136: 814–828. 343. Demo, S. D., C. J. Kirk, M. A. Aujay, T. J. Buchholz, M. Dajee, M. N. Ho, J. Jiang, G. J. Laidig, E. R. Lewis, F. Parlati, K. D. Shenk, M. S. Smyth, C. M. Sun, M. K. Vallone, T. M. Woo, C. J. Molineaux, and M. K. Bennett. 2007. Antitumor activity of PR-171, a novel irreversible inhibitor of the proteasome. Cancer Res. 67: 6383–6391. 344. Simon, L., M. Baron, and V. Leblond. 2018. How we manage patients with Waldenström macroglobulinaemia. Br. J. Haematol. . 345. Roccaro, A. M., X. Leleu, A. Sacco, X. Jia, M. Melhem, A.-S. Moreau, H. T. Ngo, J. Runnels, A. Azab, F. Azab, N. Burwick, M. Farag, S. P. Treon, M. a Palladino, T. Hideshima, D. Chauhan, K. C. Anderson, and I. M. Ghobrial. 2008. Dual targeting of the proteasome regulates survival and homing in Waldenstrom macroglobulinemia. Blood 111: 4752–63. 346. Roccaro, A. M., A. Sacco, M. Aujay, H. T. Ngo, A. K. Azab, F. Azab, P. Quang, P. Maiso, J. Runnels, K. C. Anderson, S. Demo, and I. M. Ghobrial. 2010. Selective inhibition of chymotrypsin-like activity of the immunoproteasome and constitutive proteasome in Waldenstrom macroglobulinemia. Blood 115: 4051–60. 152

References

347. Gupta, N., S. Zhang, S. Pusalkar, M. Plesescu, S. Chowdhury, M. J. Hanley, B. Wang, C. Xia, X. Zhang, K. Venkatakrishnan, and D. R. Shepard. 2017. A phase I study to assess the mass balance, excretion, and pharmacokinetics of [14C]-ixazomib, an oral proteasome inhibitor, in patients with advanced solid tumors. Invest. New Drugs 1–9. 348. Suarez-Kelly, L. P., G. M. Kemper, M. C. Duggan, A. Stiff, T. C. Noel, J. Markowitz, E. A. Luedke, V. O. Yildiz, L. Yu, A. C. Jaime- Ramirez, V. Karpa, X. Zhang, and W. E. Carson. 2016. The combination of MLN2238 (ixazomib) with interferon-alpha results in enhanced cell death in melanoma. Oncotarget 7: 81172–81186. 349. Koerner, J., T. Brunner, and M. Groettrup. 2017. Inhibition and deficiency of the immunoproteasome subunit LMP7 suppress the development and progression of colorectal carcinoma in mice. Oncotarget 8: 50873–50888. 350. Vachharajani, N., T. Joeris, M. Luu, S. Hartmann, S. Pautz, E. Jenike, G. Pantazis, I. Prinz, M. J. Hofer, U. Steinhoff, and A. Visekruna. 2017. Prevention of colitis-associated cancer by selective targeting of immunoproteasome subunit LMP7. Oncotarget 8: 50447–50459. 351. Ochoa-Ruiz, E., and R. Diaz-Ruiz. 2012. Anaplerosis in cancer: Another step beyond the warburg effect. Am. J. Mol. Biol. 02: 291–303. 352. Zeng, H., and H. Chi. 2013. mTOR and lymphocyte metabolism. Curr. Opin. Immunol. 25: 347–355. 353. Deshaies, R. J. 2014. Proteotoxic crisis , the ubiquitin-proteasome system , and cancer therapy. BMC Biol. 11: 1–14. 354. Cenci, S., L. Oliva, F. Cerruti, E. Milan, and G. Bianchi. 2012. Pivotal Advance : Protein synthesis modulates responsiveness of differentiating and malignant plasma cells to proteasome inhibitors. J. Leukoc. Biol. 92: 921–931. 355. Chen, L., M. D. Brewer, L. Guo, R. Wang, P. Jiang, and X. Yang. 2017. Enhanced Degradation of Misfolded Proteins Promotes Tumorigenesis. Cell Rep. 18: 3143–3154. 356. Lam, W. Y., and D. Bhattacharya. 2018. Metabolic Links between Plasma Cell Survival, Secretion, and Stress. Trends Immunol. 39: 19–27. 357. Obeng, E. A., L. M. Carlson, D. M. Gutman, W. J. H. Jr, K. P. Lee, and L. H. Boise. 2006. Proteasome inhibitors induce a terminal unfolded protein response in multiple myeloma cells. Blood 107: 4907–4917. 358. Punke, A. P., J. A. Waddell, and D. A. Solimando. 2017. Lenalidomide, Bortezomib, and Dexamethasone (RVD) Regimen for Multiple Myeloma. Hosp. Pharm. 52: 27–32. 359. Windebank, A. J., and W. Grisold. 2008. Chemotherapy-induced neuropathy. J Peripher Nerv Syst 13: 27–46. 360. Ejaz, N. S., R. R. Alloway, F. Halleck, M. Dürr, K. Budde, and E. S. Woodle. 2014. Review of Bortezomib Treatment of Antibody-Mediated Rejection in Renal Transplantation. Antioxid. Redox Signal. 21: 2401–2418. 361. Li, J., M. Basler, G. Alvarez, T. Brunner, C. J. Kirk, and M. Groettrup. 2018. Immunoproteasome inhibition prevents chronic antibody- mediated allograft rejection in renal transplantation. Kidney Int. 93: 753–760. 362. Mundt, S., M. Basler, B. Sawitzki, and M. Groettrup. 2017. No prolongation of skin allograft survival by immunoproteasome inhibition in mice. Mol. Immunol. 88: 32–37. 363. Ichikawa, H. T., T. Conley, T. Muchamuel, J. Jiang, S. Lee, T. Owen, J. Barnard, S. Nevarez, B. I. Goldman, C. J. Kirk, R. J. Looney, and J. H. Anolik. 2012. Beneficial effect of novel proteasome inhibitors in murine lupus via dual inhibition of type I interferon and autoantibody- secreting cells. Arthritis Rheum. 64: 493–503. 364. Nagayama, Y., M. Nakahara, M. Shimamura, I. Horie, K. Arima, and N. Abiru. 2012. Prophylactic and therapeutic efficacies of a selective inhibitor of the immunoproteasome for Hashimoto’s thyroiditis, but not for Graves’ hyperthyroidism, in mice. Clin. Exp. Immunol. 168: 268–273. 365. Frausto, R. F., S. J. Crocker, B. Eam, J. K. Whitmire, and J. L. Whitton. 2007. Myelin oligodendrocyte glycoprotein peptide-induced experimental allergic encephalomyelitis and T cell responses are unaffected by immunoproteasome deficiency. J. Neuroimmunol. 192: 124–33. 366. Basler, M., M. Dajee, C. Moll, M. Groettrup, and C. J. Kirk. 2010. Prevention of experimental colitis by a selective inhibitor of the immunoproteasome. J. Immunol. 185: 634–41. 367. Fitzpatrick, L. R., V. Khare, J. S. Small, and W. A. Koltun. 2006. Dextran sulfate sodium-induced colitis is associated with enhanced low molecular mass polypeptide 2 (LMP2) expression and is attenuated in LMP2 knockout mice. Dig. Dis. Sci. 51: 1269–1276. 368. Hammer, Q., T. Rückert, and C. Romagnani. 2018. Natural killer cell specificity for viral infections. Nat. Immunol. 19. 369. Bernink, J., J. Mjösberg, and H. Spits. 2013. Th1- and Th2-like subsets of innate lymphoid cells. Immunol. Rev. 252: 133–8. 370. Cherrier, D. E., N. Serafini, and J. P. Di Santo. 2018. Innate Lymphoid Cell Development: A T Cell Perspective. Immunity 48: 1091–1103. 371. Odegaard, J. I., and A. Chawla. 2015. Type 2 responses at the interface between immunity and fat metabolism. Curr. Opin. Immunol. 36: 67– 72. 372. Artis, D., and H. Spits. 2015. The biology of innate lymphoid cells. Nature 517: 293–301. 373. Hirota, K., M. Hashimoto, Y. Ito, M. Matsuura, H. Ito, M. Tanaka, H. Watanabe, G. Kondoh, A. Tanaka, K. Yasuda, M. Kopf, A. J. Potocnik, B. Stockinger, N. Sakaguchi, and S. Sakaguchi. 2018. Autoimmune Th17 Cells Induced Synovial Stromal and Innate Lymphoid Cell Secretion of the Cytokine GM-CSF to Initiate and Augment Autoimmune Arthritis. Immunity 48: 1220–1232.e5. 374. Nemazee, D. 2017. Mechanisms of central tolerance for B cells. Nat. Rev. Immunol. 17: 281–294. 375. MacIver, N. J., R. D. Michalek, and J. C. Rathmell. 2013. Metabolic regulation of T lymphocytes. Annu. Rev. Immunol. 31: 259–83. 376. Mesin, L., J. Ersching, and G. D. Victora. 2016. Germinal Center B Cell Dynamics. Immunity 45: 471–482. 377. Puga, I., M. Cols, C. M. Barra, B. He, L. Cassis, M. Gentile, L. Comerma, A. Chorny, M. Shan, W. Xu, G. Magri, D. M. Knowles, W. Tam, A. Chiu, J. B. Bussel, S. Serrano, J. A. Lorente, B. Bellosillo, J. Lloreta, N. Juanpere, F. Alameda, T. Baró, C. D. de Heredia, N. Torán, A. Català, M. Torrebadell, C. Fortuny, V. Cusí, C. Carreras, G. a Diaz, J. M. Blander, C.-M. Farber, G. Silvestri, C. Cunningham-Rundles, M. Calvillo, C. Dufour, L. D. Notarangelo, V. Lougaris, A. Plebani, J.-L. Casanova, S. C. Ganal, A. Diefenbach, J. I. Aróstegui, M. Juan, J. Yagüe, N. Mahlaoui, J. Donadieu, K. Chen, and A. Cerutti. 2012. B cell-helper neutrophils stimulate the diversification and production of immunoglobulin in the marginal zone of the spleen. Nat. Immunol. 13: 170–80. 378. Weisel, F., and M. Shlomchik. 2017. Memory B Cells of Mice and Humans. Annu. Rev. Immunol. 35: 255–284. 379. Rawlings, D. J., M. A. Schwartz, S. W. Jackson, and A. Meyer-Bahlburg. 2012. Integration of B cell responses through Toll-like receptors and antigen receptors. Nat. Rev. Immunol. 12: 282–294. 380. Hua, Z., and B. Hou. 2013. TLR signaling in B-cell development and activation. Cell. Mol. Immunol. 10: 103–106. 381. Mizuno, T., and T. L. Rothstein. 2005. B cell receptor (BCR) cross-talk: CD40 engagement creates an alternate pathway for BCR signaling that activates I kappa B kinase/I kappa B alpha/NF-kappa B without the need for PI3K and phospholipase C gamma. J. Immunol. 174: 6062–70. 382. Haxhinasto, S. A., and G. A. Bishop. 2004. Synergistic B Cell Activation by CD40 and the B Cell Antigen Receptor. J. Biol. Chem. 279: 2575–2582. 383. Ying, H., Z. Li, L. Yang, and J. Zhang. 2011. Syk mediates BCR- and CD40-signaling integration during B cell activation. Immunobiology 216: 566–570. 384. Pisetsky, D. S. 2017. Antinuclear antibody testing - misunderstood or misbegotten? Nat. Rev. Rheumatol. 13: 495–502. 385. Hofmann, K., A.-K. Clauder, and R. A. Manz. 2018. Targeting B Cells and Plasma Cells in Autoimmune Diseases. Front. Immunol. 9: 835. 386. Karahan, G. E., F. H. J. Claas, and S. Heidt. 2017. B Cell Immunity in Solid Organ Transplantation. Front. Immunol. 7: 1–11. 387. Malkiel, S., A. N. Barlev, Y. Atisha-Fregoso, J. Suurmond, and B. Diamond. 2018. Plasma cell differentiation pathways in systemic lupus 153

References erythematosus. Front. Immunol. 9. 388. Faurschou, M., and D. R. W. Jayne. 2014. Anti–B Cell Antibody Therapies for Inflammatory Rheumatic Diseases. Annu. Rev. Med. 65: 263– 278. 389. Reddy, V., D. Jayne, D. Close, and D. Isenberg. 2013. B-cell depletion in SLE: Clinical and trial experience with rituximab and ocrelizumab and implications for study design. Arthritis Res. Ther. 15: 1–16. 390. Hoyer, B. F., K. Moser, A. E. Hauser, A. Peddinghaus, C. Voigt, D. Eilat, A. Radbruch, F. Hiepe, and R. A. Manz. 2004. Short-lived Plasmablasts and Long-lived Plasma Cells Contribute to Chronic Humoral Autoimmunity in NZB/W Mice. J. Exp. Med. 199: 1577–1584. 391. Starzl, T. E., C. G. Halgrimson, I. Penn, G. Martineau, G. Schroter, H. Amemiya, C. W. Putnam, and C. G. Groth. 1971. Cyclophosphamide and human organ transplantation. Lancet 2: 70–4. 392. Alexander, T., R. Sarfert, J. Klotsche, A. A. Kühl, A. Rubbert-Roth, H.-M. Lorenz, J. Rech, B. F. Hoyer, Q. Cheng, A. Waka, A. Taddeo, M. Wiesener, G. Schett, G.-R. Burmester, A. Radbruch, F. Hiepe, and R. E. Voll. 2015. The proteasome inhibitior bortezomib depletes plasma cells and ameliorates clinical manifestations of refractory systemic lupus erythematosus. Ann. Rheum. Dis. 74: 1474–8. 393. Seavey, M. M., L. D. Lu, K. L. Stump, N. H. Wallace, and B. A. Ruggeri. 2012. Novel, orally active, proteasome inhibitor, delanzomib (CEP- 18770), ameliorates disease symptoms and glomerulonephritis in two preclinical mouse models of SLE. Int. Immunopharmacol. 12: 257–270. 394. Mulder, A., S. Heidt, M. Vergunst, D. L. Roelen, and F. H. J. Claas. 2013. Proteasome inhibition profoundly affects activated human b cells. Transplantation 95: 1331–1337. 395. Hirai, M., N. Kadowaki, T. Kitawaki, H. Fujita, A. Takaori-kondo, W. Dc, and R. Fukui. 2011. Bortezomib suppresses function and survival of plasmacytoid dendritic cells by targeting intracellular trafficking of Toll-like receptors and endoplasmic reticulum homeostasis Bortezomib suppresses function and survival of plasmacytoid dendritic cells by . Hematology 117: 500–509. 396. Krangel, M. S. 2009. Mechanics of T cell receptor gene rearrangement. Curr. Opin. Immunol. 21: 133–139. 397. von Boehmer, H. 1994. Positive selection of lymphocytes. Cell 76: 219–28. 398. von Boehmer, H., P. Kisielow, H. Kishi, B. Scott, P. Borgulya, and H. S. Teh. 1989. The expression of CD4 and CD8 accessory molecules on mature T cells is not random but correlates with the specificity of the alpha beta receptor for antigen. Immunol. Rev. 109: 143–51. 399. Klein, L., B. Kyewski, P. M. Allen, and K. A. Hogquist. 2014. Positive and negative selection of the T cell repertoire: What thymocytes see (and don’t see). Nat. Rev. Immunol. 14: 377–391. 400. Pacholczyk, R., P. Kraj, and L. Ignatowicz. 2002. Peptide specificity of thymic selection of CD4+CD25+ T cells. J. Immunol. 168: 613–20. 401. Sakaguchi, S. 2005. Naturally arising Foxp3-expressing CD25+CD4+ regulatory T cells in immunological tolerance to self and non-self. Nat. Immunol. 6: 345–52. 402. Derbinski, J., J. Gäbler, B. Brors, S. Tierling, S. Jonnakuty, M. Hergenhahn, L. Peltonen, J. Walter, and B. Kyewski. 2005. Promiscuous gene expression in thymic epithelial cells is regulated at multiple levels. J. Exp. Med. 202: 33–45. 403. Greyer, M., P. G. Whitney, A. T. Stock, G. M. Davey, C. Tebartz, A. Bachem, J. D. Mintern, R. A. Strugnell, S. J. Turner, T. Gebhardt, M. O’Keeffe, W. R. Heath, and S. Bedoui. 2016. T Cell Help Amplifies Innate Signals in CD8+ DCs for Optimal CD8+ T Cell Priming. Cell Rep. 14: 586–597. 404. Bedoui, S., W. R. Heath, and S. N. Mueller. 2016. CD4(+) T-cell help amplifies innate signals for primary CD8(+) T-cell immunity. Immunol. Rev. 272: 52–64. 405. Bennett, S. R. M., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. A. P. Miller, and W. R. Heath. 1998. Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 393: 478–480. 406. Bennett, B. S. R. M., F. R. Carbone, F. Karamalis, J. F. A. P. Miller, and W. R. Heath. 1997. Induction of a CD8. 186: 4–9. 407. Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, and C. J. Melief. 1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393: 480–483. 408. Andreasen, S. O., J. E. Christensen, O. Marker, and A. R. Thomsen. 2000. Role of CD40 Ligand and CD28 in Induction and Maintenance of Antiviral CD8+ Effector T Cell Responses. J. Immunol. 164: 3689–3697. 409. Halle, S., O. Halle, and R. Förster. 2017. Mechanisms and Dynamics of T Cell-Mediated Cytotoxicity In Vivo. Trends Immunol. 38: 432–443. 410. Mosmann, T. R., and R. L. Coffman. 1989. TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7: 145–73. 411. Schmitt, N., and H. Ueno. 2015. Regulation of human helper T cell subset differentiation by cytokines. Curr. Opin. Immunol. 34: 130–136. 412. Kaplan, M. H., M. M. Hufford, and M. R. Olson. 2015. The development and in vivo function of T helper 9 cells. Nat. Rev. Immunol. 15: 295–307. 413. Mirshafiey, A., A. Simhag, N. M. M. El Rouby, and G. Azizi. 2015. T-helper 22 cells as a new player in chronic inflammatory skin disorders. Int J Dermatol 54: 880–8. 414. Coquet, J. M., L. Rausch, and J. Borst. 2015. The importance of co-stimulation in the orchestration of T helper cell differentiation. Immunol. Cell Biol. 93: 780–788. 415. Szabo, S. J., S. T. Kim, G. L. Costa, X. Zhang, C. G. Fathman, and L. H. Glimcher. 2000. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100: 655–69. 416. Szabo, S. J., B. M. Sullivan, S. L. Peng, and L. H. Glimcher. 2003. Molecular mechanisms regulating Th1 immune responses. Annu. Rev. Immunol. 21: 713–58. 417. Lutz, M. B. 2016. Induction of CD4(+) Regulatory and Polarized Effector/helper T Cells by Dendritic Cells. Immune Netw. 16: 13–25. 418. Anderson, C. F., M. Oukka, V. J. Kuchroo, and D. Sacks. 2007. CD4 + CD25 − Foxp3 − Th1 cells are the source of IL-10–mediated immune suppression in chronic cutaneous leishmaniasis. J. Exp. Med. 204: 285–297. 419. Jankovic, D., M. C. Kullberg, C. G. Feng, R. S. Goldszmid, C. M. Collazo, M. Wilson, T. A. Wynn, M. Kamanaka, R. A. Flavell, and A. Sher. 2007. Conventional T-bet + Foxp3 − Th1 cells are the major source of host-protective regulatory IL-10 during intracellular protozoan infection. J. Exp. Med. 204: 273–283. 420. Shapouri-Moghaddam, A., S. Mohammadian, H. Vazini, M. Taghadosi, S. A. Esmaeili, F. Mardani, B. Seifi, A. Mohammadi, J. T. Afshari, and A. Sahebkar. 2018. Macrophage plasticity, polarization, and function in health and disease. J. Cell. Physiol. 233: 6425–6440. 421. Zhou, D., K. Yang, L. Chen, Y. Wang, W. Zhang, Z. Xu, J. Zuo, H. Jiang, and J. Luan. 2017. Macrophage polarization and function: New prospects for fibrotic disease. Immunol. Cell Biol. 95: 864–869. 422. Yamane, H., and W. E. Paul. 2013. Early signaling events that underlie fate decisions of naive CD4(+) T cells toward distinct T-helper cell subsets. Immunol. Rev. 252: 12–23. 423. Iwata, A., V. Durai, R. Tussiwand, C. G. Briseño, X. Wu, G. E. Grajales-Reyes, T. Egawa, T. L. Murphy, and K. M. Murphy. 2017. Quality of TCR signaling determined by differential affinities of enhancers for the composite BATF-IRF4 transcription factor complex. Nat. Immunol. 18: 563–572. 424. Breedveld, A., T. Groot Kormelink, M. van Egmond, and E. C. de Jong. 2017. Granulocytes as modulators of dendritic cell function. J. 154

References

Leukoc. Biol. 102: jlb.4MR0217-048RR. 425. Otsuka, A., and K. Kabashima. 2015. Contribution of Basophils to Cutaneous Immune Reactions and Th2-Mediated Allergic Responses. Front. Immunol. 6: 1–6. 426. Maier, E., A. Duschl, and J. Horejs-Hoeck. 2012. STAT6-dependent and -independent mechanisms in Th2 polarization. Eur. J. Immunol. 42: 2827–33. 427. Ansel, K. M., I. Djuretic, B. Tanasa, and A. Rao. 2006. Regulation of Th2 differentiation and Il4 locus accessibility. Annu. Rev. Immunol. 24: 607–56. 428. Liao, W., D. E. Schones, J. Oh, Y. Cui, K. Cui, T. Y. Roh, K. Zhao, and W. J. Leonard. 2008. Priming for T helper type 2 differentiation by interleukin 2-mediated induction of interleukin 4 receptor α-chain expression. Nat. Immunol. 9: 1288–1296. 429. Paul, W. E., and J. Zhu. 2010. How are TH2-type immune responses initiated and amplified? Nat. Rev. Immunol. 10: 225–235. 430. Zheng, W., and R. a Flavell. 1997. The transcription factor GATA-3 is necessary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 89: 587–96. 431. Wills-Karp, M., J. Luyimbazi, X. Xu, B. Schofield, T. Y. Neben, C. L. Karp, and D. D. Donaldson. 1998. Interleukin-13: central mediator of allergic asthma. Science . 282: 2258–61. 432. Fort, M. M., J. Cheung, D. Yen, J. Li, S. M. Zurawski, S. Lo, S. Menon, T. Clifford, B. Hunte, R. Lesley, T. Muchamuel, S. D. Hurst, G. Zurawski, M. W. Leach, D. M. Gorman, and D. M. Rennick. 2001. IL-25 induces IL-4, IL-5, and IL-13 and Th2-associated pathologies in vivo. Immunity 15: 985–95. 433. McKenzie, G. J., a Bancroft, R. K. Grencis, and a N. McKenzie. 1998. A distinct role for interleukin-13 in Th2-cell-mediated immune responses. Curr. Biol. 8: 339–42. 434. Vignali, D. A. A., and V. K. Kuchroo. 2012. IL-12 family cytokines: immunological playmakers. Nat. Immunol. 13: 722–728. 435. Becher, B., B. G. Durell, and R. J. Noelle. 2002. Experimental autoimmune encephalitis and inflammation in the absence of interleukin-12. J. Clin. Invest. 110: 493–7. 436. Harrington, L. E., R. D. Hatton, P. R. Mangan, H. Turner, T. L. Murphy, K. M. Murphy, and C. T. Weaver. 2005. Interleukin 17-producing CD4+ effector T cells develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat. Immunol. 6: 1123–32. 437. Park, H., Z. Li, X. O. Yang, S. H. Chang, R. Nurieva, Y.-H. Wang, Y. Wang, L. Hood, Z. Zhu, Q. Tian, and C. Dong. 2005. A distinct lineage of CD4 T cells regulates tissue inflammation by producing interleukin 17. Nat. Immunol. 6: 1133–41. 438. Cua, D. J., J. Sherlock, Y. Chen, C. a Murphy, B. Joyce, B. Seymour, L. Lucian, W. To, S. Kwan, T. Churakova, S. Zurawski, M. Wiekowski, S. a Lira, D. Gorman, R. a Kastelein, and J. D. Sedgwick. 2003. Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature 421: 744–8. 439. McGeachy, M. J., Y. Chen, C. M. Tato, A. Laurence, B. Joyce-Shaikh, W. M. Blumenschein, T. K. McClanahan, J. J. O’Shea, and D. J. Cua. 2009. The interleukin 23 receptor is essential for the terminal differentiation of interleukin 17-producing effector T helper cells in vivo. Nat. Immunol. 10: 314–24. 440. Bettelli, E., Y. Carrier, W. Gao, T. Korn, T. B. Strom, M. Oukka, H. L. Weiner, and V. K. Kuchroo. 2006. Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature 441: 235–238. 441. Veldhoen, M., R. J. Hocking, C. J. Atkins, R. M. Locksley, and B. Stockinger. 2006. TGFβ in the context of an inflammatory cytokine milieu supports de novo differentiation of IL-17-producing T cells. Immunity 24: 179–189. 442. Ivanov, I. I., B. S. McKenzie, L. Zhou, C. E. Tadokoro, A. Lepelley, J. J. Lafaille, D. J. Cua, and D. R. Littman. 2006. The orphan nuclear receptor RORgammat directs the differentiation program of proinflammatory IL-17+ T helper cells. Cell 126: 1121–33. 443. Ruan, Q., V. Kameswaran, Y. Zhang, S. Zheng, J. Sun, J. Wang, J. DeVirgiliis, H.-C. Liou, A. a Beg, and Y. H. Chen. 2011. The Th17 immune response is controlled by the Rel-RORγ-RORγ T transcriptional axis. J. Exp. Med. 208: 2321–33. 444. Purvis, H. A., A. E. Anderson, D. A. Young, J. D. Isaacs, and C. M. U. Hilkens. 2014. A Negative Feedback Loop Mediated by STAT3 Limits Human Th17 Responses. J. Immunol. 193: 1142–1150. 445. Codarri, L., G. Gyülvészi, V. Tosevski, L. Hesske, A. Fontana, L. Magnenat, T. Suter, and B. Becher. 2011. RORγt drives production of the cytokine GM-CSF in helper T cells, which is essential for the effector phase of autoimmune neuroinflammation. Nat. Immunol. 12: 560–7. 446. McGeachy, M. J. 2011. GM-CSF: the secret weapon in the T(H)17 arsenal. Nat. Immunol. 12: 521–2. 447. El-Behi, M., B. Ciric, H. Dai, Y. Yan, M. Cullimore, F. Safavi, G.-X. Zhang, B. N. Dittel, and A. Rostami. 2011. The encephalitogenicity of T(H)17 cells is dependent on IL-1- and IL-23-induced production of the cytokine GM-CSF. Nat. Immunol. 12: 568–75. 448. Saijo, S., S. Ikeda, K. Yamabe, S. Kakuta, H. Ishigame, A. Akitsu, N. Fujikado, T. Kusaka, S. Kubo, S. hyun Chung, R. Komatsu, N. Miura, Y. Adachi, N. Ohno, K. Shibuya, N. Yamamoto, K. Kawakami, S. Yamasaki, T. Saito, S. Akira, and Y. Iwakura. 2010. Dectin-2 recognition of α- mannans and induction of Th17 cell differentiation is essential for host defense against candida albicans. Immunity 32: 681–691. 449. Li, Y., C. Wei, H. Xu, J. Jia, Z. Wei, R. Guo, Y. Jia, Y. Wu, Y. Li, X. Qi, Z. Li, and X. Gao. 2018. The Immunoregulation of Th17 in Host against Intracellular Bacterial Infection. Mediators Inflamm. 2018: 1–13. 450. Huang, W., L. Na, P. L. Fidel, and P. Schwarzenberger. 2004. Requirement of interleukin-17A for systemic anti-Candida albicans host defense in mice. J. Infect. Dis. 190: 624–31. 451. De Luca, A., T. Zelante, C. D’Angelo, S. Zagarella, F. Fallarino, A. Spreca, R. G. Iannitti, P. Bonifazi, J. C. Renauld, F. Bistoni, P. Puccetti, and L. Romani. 2010. IL-22 defines a novel immune pathway of antifungal resistance. Mucosal Immunol. 3: 361–373. 452. Lockhart, E., A. M. Green, and J. L. Flynn. 2006. IL-17 production is dominated by gammadelta T cells rather than CD4 T cells during Mycobacterium tuberculosis infection. J. Immunol. 177: 4662–9. 453. Schubert, L. a, E. Jeffery, Y. Zhang, F. Ramsdell, and S. F. Ziegler. 2001. Scurfin (FOXP3) acts as a repressor of transcription and regulates T cell activation. J. Biol. Chem. 276: 37672–9. 454. Fontenot, J. D., M. a Gavin, and A. Y. Rudensky. 2003. Foxp3 programs the development and function of CD4+CD25+ regulatory T cells. Nat. Immunol. 4: 330–6. 455. Burchill, M. A., J. Yang, C. Vogtenhuber, B. R. Blazar, and M. A. Farrar. 2007. IL-2 Receptor -Dependent STAT5 Activation Is Required for the Development of Foxp3+ Regulatory T Cells. J. Immunol. 178: 280–290. 456. Burchill, M. a, J. Yang, K. B. Vang, J. J. Moon, H. H. Chu, C.-W. J. Lio, A. L. Vegoe, C.-S. Hsieh, M. K. Jenkins, and M. a Farrar. 2008. Linked T cell receptor and cytokine signaling govern the development of the regulatory T cell repertoire. Immunity 28: 112–21. 457. Davidson, T. S., R. J. DiPaolo, J. Andersson, and E. M. Shevach. 2007. Cutting Edge: IL-2 Is Essential for TGF- -Mediated Induction of Foxp3+ T Regulatory Cells. J. Immunol. 178: 4022–4026. 458. Hori, S., T. Nomura, and S. Sakaguchi. 2003. Control of regulatory T cell development by the transcription factor Foxp3. Science . 299: 1057–61. 459. Laurence, A., C. M. Tato, T. S. Davidson, Y. Kanno, Z. Chen, Z. Yao, R. B. Blank, F. Meylan, R. Siegel, L. Hennighausen, E. M. Shevach, and J. J. O’shea. 2007. Interleukin-2 signaling via STAT5 constrains T helper 17 cell generation. Immunity 26: 371–81. 155

References

460. McKarns, S. C., R. H. Schwartz, and N. E. Kaminski. 2004. Smad3 Is Essential for TGF- 1 to Suppress IL-2 Production and TCR-Induced Proliferation, but Not IL-2-Induced Proliferation. J. Immunol. 172: 4275–4284. 461. Gavin, M. a, J. P. Rasmussen, J. D. Fontenot, V. Vasta, V. C. Manganiello, J. a Beavo, and A. Y. Rudensky. 2007. Foxp3-dependent programme of regulatory T-cell differentiation. Nature 445: 771–5. 462. Marson, A., K. Kretschmer, G. M. Frampton, E. S. Jacobsen, J. K. Polansky, K. D. MacIsaac, S. S. Levine, E. Fraenkel, H. von Boehmer, and R. a Young. 2007. Foxp3 occupancy and regulation of key target genes during T-cell stimulation. Nature 445: 931–5. 463. Boyman, O., and J. Sprent. 2012. The role of interleukin-2 during homeostasis and activation of the immune system. Nat. Rev. Immunol. 12: 180–90. 464. Sakaguchi, S., T. Yamaguchi, T. Nomura, and M. Ono. 2008. Regulatory T Cells and Immune Tolerance. Cell 133: 775–787. 465. Thornton, a M., and E. M. Shevach. 2000. Suppressor effector function of CD4+CD25+ immunoregulatory T cells is antigen nonspecific. J. Immunol. 164: 183–90. 466. Thornton, A. M., and E. M. Shevach. 1998. CD4+CD25+ immunoregulatory T cells suppress polyclonal T cell activation in vitro by inhibiting interleukin 2 production. J. Exp. Med. 188: 287–96. 467. Huang, Y.-H., D. K. Sojka, and D. J. Fowell. 2012. Cutting edge: Regulatory T cells selectively attenuate, not terminate, T cell signaling by disrupting NF-κB nuclear accumulation in CD4 T cells. J. Immunol. 188: 947–51. 468. Zhou, L., J. E. Lopes, M. M. W. Chong, I. I. Ivanov, R. Min, G. D. Victora, Y. Shen, J. Du, Y. P. Rubtsov, A. Y. Rudensky, S. F. Ziegler, and D. R. Littman. 2008. TGF-beta-induced Foxp3 inhibits T(H)17 cell differentiation by antagonizing RORgammat function. Nature 453: 236–40. 469. Ichiyama, K., H. Yoshida, Y. Wakabayashi, T. Chinen, K. Saeki, M. Nakaya, G. Takaesu, S. Hori, A. Yoshimura, and T. Kobayashi. 2008. Foxp3 inhibits RORγt-mediated IL-17A mRNA transcription through direct interaction with RORγt. J. Biol. Chem. 283: 17003–17008. 470. Gagliani, N., M. C. Amezcua Vesely, A. Iseppon, L. Brockmann, H. Xu, N. W. Palm, M. R. De Zoete, P. Licona-Limón, R. S. Paiva, T. Ching, C. Weaver, X. Zi, X. Pan, R. Fan, L. X. Garmire, M. J. Cotton, Y. Drier, B. Bernstein, J. Geginat, B. Stockinger, E. Esplugues, S. Huber, and R. A. Flavell. 2015. TH17 cells transdifferentiate into regulatory T cells uring resolution of inflammation. Nature 523: 221–225. 471. Obermajer, N., F. C. Popp, Y. Soeder, J. Haarer, E. K. Geissler, H. J. Schlitt, and M. H. Dahlke. 2014. Conversion of Th17 into IL-17A neg Regulatory T Cells: A Novel Mechanism in Prolonged Allograft Survival Promoted by Mesenchymal Stem Cell–Supported Minimized Immunosuppressive Therapy. J. Immunol. 193: 4988–4999. 472. Kwon, H. K., H. M. Chen, D. Mathis, and C. Benoist. 2017. Different molecular complexes that mediate transcriptional induction and repression by FoxP3. Nat. Immunol. 18: 1238–1248. 473. Bending, D., and M. Ono. 2017. FoxP3 partners up. Nat. Immunol. 18: 1181–1183. 474. Luo, X., Q. Zhang, V. Liu, Z. Xia, K. L. Pothoven, and C. Lee. 2008. Cutting edge: TGF-beta-induced expression of Foxp3 in T cells is mediated through inactivation of ERK. J. Immunol. 180: 2757–61. 475. Liu, H., S. Yao, S. M. Dann, H. Qin, C. O. Elson, and Y. Cong. 2013. ERK differentially regulates Th17- and Treg-cell development and contributes to the pathogenesis of colitis. Eur. J. Immunol. 1–11. 476. Li, M. O., and A. Y. Rudensky. 2016. T cell receptor signalling in the control of regulatory T cell differentiation and function. Nat. Rev. Immunol. 16: 220–233. 477. Cozzo Picca, C., D. M. Simons, S. Oh, M. Aitken, O. A. Perng, C. Mergenthaler, E. Kropf, J. Erikson, and A. J. Caton. 2011. CD4+CD25+Foxp3+ regulatory T cell formation requires more specific recognition of a self-peptide than thymocyte deletion. Proc. Natl. Acad. Sci. 108: 14890–14895. 478. Chakraborty, A. K., and A. Weiss. 2014. Insights into the initiation of TCR signaling. Nat. Immunol. 15: 798–807. 479. Courtney, A. H., W.-L. Lo, and A. Weiss. 2018. TCR Signaling: Mechanisms of Initiation and Propagation. Trends Biochem. Sci. 43: 108– 123. 480. Minguet, S., M. Swamy, B. Alarcón, I. F. Luescher, and W. W. A. Schamel. 2007. Full Activation of the T Cell Receptor Requires Both Clustering and Conformational Changes at CD3. Immunity 26: 43–54. 481. Cochran, J. R., D. Aivazian, T. O. Cameron, and L. J. Stern. 2001. Receptor clustering and transmembrane signaling in T cells. Trends Biochem. Sci. 26: 304–310. 482. Natarajan, K., J. Jiang, N. A. May, M. G. Mage, L. F. Boyd, A. C. McShan, N. G. Sgourakis, A. Bax, and D. H. Margulies. 2018. The Role of Molecular Flexibility in Antigen Presentation and T Cell Receptor-Mediated Signaling. Front. Immunol. 9: 1657. 483. Lee, M. S., C. R. Glassman, N. R. Deshpande, H. B. Badgandi, H. L. Parrish, C. Uttamapinant, P. S. Stawski, A. Y. Ting, and M. S. Kuhns. 2015. A Mechanical Switch Couples T Cell Receptor Triggering to the Cytoplasmic Juxtamembrane Regions of CD3ζζ. Immunity 43: 227–239. 484. Nika, K., C. Soldani, M. Salek, W. Paster, A. Gray, R. Etzensperger, L. Fugger, P. Polzella, V. Cerundolo, O. Dushek, T. Höfer, A. Viola, and O. Acuto. 2010. Constitutively active lck kinase in T cells drives antigen receptor signal transduction. Immunity 32: 766–777. 485. Swamy, M., K. Beck-Garcia, E. Beck-Garcia, F. A. Hartl, A. Morath, O. S. Yousefi, E. P. Dopfer, E. Molnár, A. K. Schulze, R. Blanco, A. Borroto, N. Martín-Blanco, B. Alarcon, T. Höfer, S. Minguet, and W. W. A. Schamel. 2016. A Cholesterol-Based Allostery Model of T Cell Receptor Phosphorylation. Immunity 44: 1091–1101. 486. Wang, F., K. Beck-García, C. Zorzin, W. W. A. Schamel, and M. M. Davis. 2016. Inhibition of T cell receptor signaling by cholesterol sulfate, a naturally occurring derivative of membrane cholesterol. Nat. Immunol. 17: 844–850. 487. Molnár, E., M. Swamy, M. Holzer, K. Beck-García, R. Worch, C. Thiele, G. Guigas, K. Boye, I. F. Luescher, P. Schwille, R. Schubert, and W. W. A. Schamel. 2012. Cholesterol and sphingomyelin drive ligand-independent T-cell antigen receptor nanoclustering. J. Biol. Chem. 287: 42664–42674. 488. Liu, B., W. Chen, B. D. Evavold, and C. Zhu. 2014. Accumulation of dynamic catch bonds between TCR and agonist peptide-MHC triggers T cell signaling. Cell 157: 357–368. 489. Huang, J., M. Brameshuber, X. Zeng, J. Xie, Q.-J. Li, Y.-H. Chien, S. Valitutti, and M. M. Davis. 2013. A Single Peptide-Major Histocompatibility Complex Ligand Triggers Digital Cytokine Secretion in CD4(+) T Cells. Immunity 39: 846–857. 490. Valitutti, S., S. Miller, M. Cella, E. Padovan, and A. Lanzavecchia. 1995. Serial triggering of many T-cell receptors by a few peptide-MHC complexes. Nature 375: 148–151. 491. Krogsgaard, M., Q.-J. Li, C. Sumen, J. B. Huppa, M. Huse, and M. M. Davis. 2005. Agonist/endogenous peptide-MHC heterodimers drive T cell activation and sensitivity. Nature 434: 238–43. 492. Veillette, A., M. A. Bookman, E. M. Horak, and J. B. Bolen. 1988. The CD4 and CD8 T cell surface antigens are associated with the internal membrane tyrosine-protein kinase p56lck. Cell 55: 301–8. 493. Barber, E. K., J. D. Dasgupta, S. F. Schlossman, J. M. Trevillyan, and C. E. Rudd. 1989. The CD4 and CD8 antigens are coupled to a protein- tyrosine kinase (p56lck) that phosphorylates the CD3 complex. Proc. Natl. Acad. Sci. U. S. A. 86: 3277–81. 494. Bijlmakers, M. J. 2009. Protein acylation and localization in T cell signaling (Review). Mol. Membr. Biol. 26: 93–103. 495. Ngoenkam, J., W. W. Schamel, and S. Pongcharoen. 2018. Selected signalling proteins recruited to the T-cell receptor–CD3 complex. 156

References

Immunology 153: 42–50. 496. Yiemwattana, I., J. Ngoenkam, P. Paensuwan, R. Kriangkrai, B. Chuenjitkuntaworn, and S. Pongcharoen. 2012. Essential role of the adaptor protein Nck1 in Jurkat T cell activation and function. Clin. Exp. Immunol. 167: 99–107. 497. Ngoenkam, J., P. Paensuwan, K. Preechanukul, B. Khamsri, I. Yiemwattana, E. Beck-García, S. Minguet, W. W. Schamel, and S. Pongcharoen. 2014. Non-overlapping functions of Nck1 and Nck2 adaptor proteins in T cell activation. Cell Commun. Signal. 12: 1–13. 498. Paensuwan, P., F. A. Hartl, O. S. Yousefi, J. Ngoenkam, P. Wipa, E. Beck-Garcia, E. P. Dopfer, B. Khamsri, D. Sanguansermsri, S. Minguet, W. W. Schamel, and S. Pongcharoen. 2016. Nck Binds to the T Cell Antigen Receptor Using Its SH3.1 and SH2 Domains in a Cooperative Manner, Promoting TCR Functioning. J. Immunol. 196: 448–58. 499. Courtney, A. H., J. F. Amacher, T. A. Kadlecek, M. N. Mollenauer, B. B. Au-Yeung, J. Kuriyan, and A. Weiss. 2017. A Phosphosite within the SH2 Domain of Lck Regulates Its Activation by CD45. Mol. Cell 67: 498–511.e6. 500. Guy, C. S., K. M. Vignali, J. Temirov, M. L. Bettini, A. E. Overacre, M. Smeltzer, H. Zhang, J. B. Huppa, Y.-H. Tsai, C. Lobry, J. Xie, P. J. Dempsey, H. C. Crawford, I. Aifantis, M. M. Davis, and D. a a Vignali. 2013. Distinct TCR signaling pathways drive proliferation and cytokine production in T cells. Nat. Immunol. 14: 262–70. 501. James, J. R., and R. D. Vale. 2012. Biophysical mechanism of T-cell receptor triggering in a reconstituted system. Nature 487: 64–69. 502. Brownlie, R. J., and R. Zamoyska. 2013. T cell receptor signalling networks: branched, diversified and bounded. Nat. Rev. Immunol. 13: 257– 269. 503. Harwood, N. E., and F. D. Batista. 2010. Early Events in B Cell Activation. Annu. Rev. Immunol. 28: 185–210. 504. Finco, T. S., T. Kadlecek, W. Zhang, L. E. Samelson, and A. Weiss. 1998. LAT is required for TCR-mediated activation of PLCgamma1 and the Ras pathway. Immunity 9: 617–26. 505. Navarro, M. N., and D. a Cantrell. 2014. Serine-threonine kinases in TCR signaling. Nat. Immunol. 15: 808–814. 506. Paul, S., and B. C. Schaefer. 2013. A new look at T cell receptor signaling to nuclear factor-κB. Trends Immunol. 34: 269–81. 507. Kortum, R. L., A. K. Rouquette-Jazdanian, and L. E. Samelson. 2013. Ras and extracellular signal-regulated kinase signaling in thymocytes and T cells. Trends Immunol. 34: 259–68. 508. Houtman, J. C. D., R. A. Houghtling, M. Barda-Saad, Y. Toda, and L. E. Samelson. 2005. Early phosphorylation kinetics of proteins involved in proximal TCR-mediated signaling pathways. J. Immunol. 175: 2449–58. 509. Quill, H., and R. H. Schwartz. 1987. Stimulation of normal inducer T cell clones with antigen presented by purified Ia molecules in planar lipid membranes: specific induction of a long-lived state of proliferative nonresponsiveness. J. Immunol. 138: 3704–3712. 510. Jenkins, M. K., D. M. Pardoll, J. Mizuguchi, T. M. Chused, and R. H. Schwartz. 1987. Molecular events in the induction of a nonresponsive state in interleukin 2-producing helper T-lymphocyte clones. Proc. Natl. Acad. Sci. 84: 5409–5413. 511. Fathman, C. G., and N. B. Lineberry. 2007. Molecular mechanisms of CD4+ T-cell anergy. Nat. Rev. Immunol. 7: 599–609. 512. Chen, L., and D. B. Flies. 2013. Molecular mechanisms of T cell co-stimulation and co-inhibition. Nat. Rev. Immunol. 13: 227–242. 513. Harding, F. A., J. G. McArthur, J. A. Gross, D. H. Raulet, and J. P. Allison. 1992. CD28-mediated signalling co-stimulates murine T cells and prevents induction of anergy in T-cell clones. Nature 356: 607–9. 514. Acuto, O., and F. Michel. 2003. CD28-mediated co-stimulation: a quantitative support for TCR signalling. Nat. Rev. Immunol. 3: 939–51. 515. Diehn, M., A. A. Alizadeh, O. J. Rando, C. L. Liu, K. Stankunas, D. Botstein, G. R. Crabtree, and P. O. Brown. 2002. Genomic expression programs and the integration of the CD28 costimulatory signal in T cell activation. Proc. Natl. Acad. Sci. U. S. A. 99: 11796–801. 516. Khoshnan, a, C. Tindell, I. Laux, D. Bae, B. Bennett, and a E. Nel. 2000. The NF-kappa B cascade is important in Bcl-xL expression and for the anti-apoptotic effects of the CD28 receptor in primary human CD4+ lymphocytes. J. Immunol. 165: 1743–54. 517. Martínez-Llordella, M., J. H. Esensten, S. L. Bailey-Bucktrout, R. H. Lipsky, A. Marini, J. Chen, M. Mughal, M. P. Mattson, D. D. Taub, and J. a Bluestone. 2013. CD28-inducible transcription factor DEC1 is required for efficient autoreactive CD4+ T cell response. J. Exp. Med. 210: 1603–19. 518. Lin, C. C., T. R. Bradstreet, E. A. Schwarzkopf, J. Sim, J. A. Carrero, C. Chou, L. E. Cook, T. Egawa, R. Taneja, T. L. Murphy, J. H. Russell, and B. T. Edelson. 2014. Bhlhe40 controls cytokine production by T cells and is essential for pathogenicity in autoimmune neuroinflammation. Nat. Commun. 5: 1–13. 519. Schildberg, F. A., S. R. Klein, G. J. Freeman, and A. H. Sharpe. 2016. Coinhibitory Pathways in the B7-CD28 Ligand-Receptor Family. Immunity 44: 955–972. 520. Evans, E. J., R. M. Esnouf, R. Manso-Sancho, R. J. C. Gilbert, J. R. James, C. Yu, J. A. Fennelly, C. Vowles, T. Hanke, B. Walse, T. Hünig, P. Sørensen, D. I. Stuart, and S. J. Davis. 2005. Crystal structure of a soluble CD28-Fab complex. Nat. Immunol. 6: 271–279. 521. Esensten, J. H., Y. A. Helou, G. Chopra, A. Weiss, and J. A. Bluestone. 2016. CD28 Costimulation: From Mechanism to Therapy. Immunity 44: 973–988. 522. Tuosto, L. 2011. NF-κB family of transcription factors: biochemical players of CD28 co-stimulation. Immunol. Lett. 135: 1–9. 523. Li, Y., C. E. Sedwick, J. Hu, and A. Altman. 2005. Role for protein kinase C-theta (PKC theta) in TCR/CD28-mediated signaling through the canonical but not the non-canonical pathway for NF-kB activation. J. Biol. Chem. 280: 1217–1223. 524. Oh, H., and S. Ghosh. 2013. NF-κB: roles and regulation in different CD4(+) T-cell subsets. Immunol. Rev. 252: 41–51. 525. Shi, J., and S.-C. Sun. 2015. TCR signaling to NF-κB and mTORC1: Expanding roles of the CARMA1 complex. Mol. Immunol. 1–12. 526. Senftleben, U., Y. Cao, G. Xiao, F. R. Greten, G. Krähn, G. Bonizzi, Y. Chen, Y. Hu, A. Fong, S. C. Sun, and M. Karin. 2001. Activation by IKKalpha of a second, evolutionary conserved, NF-kappa B signaling pathway. Science . 293: 1495–9. 527. Sun, S. C. 2012. The noncanonical NF-κB pathway. Immunol. Rev. 246: 125–140. 528. Liao, G., M. Zhang, E. W. Harhaj, and S. C. Sun. 2004. Regulation of the NF-κB-inducing kinase by tumor necrosis factor receptor-associated factor 3-induced degradation. J. Biol. Chem. 279: 26243–26250. 529. Giardino Torchia, M. L., D. B. Conze, D. Jankovic, and J. D. Ashwell. 2013. Balance between NF-κB p100 and p52 regulates T cell costimulation dependence. J. Immunol. 190: 549–55. 530. Schmukle, A. C., and H. Walczak. 2012. No one can whistle a symphony alone - how different ubiquitin linkages cooperate to orchestrate NF- B activity. J. Cell Sci. 125: 549–559. 531. Jin, W., X. F. Zhou, J. Yu, X. Cheng, and S. C. Sun. 2009. Regulation of Th17 cell differentiation and EAE induction by MAP3K NIK. Blood 113: 6603–6610. 532. Mattioli, I., A. Sebald, C. Bucher, R.-P. Charles, H. Nakano, T. Doi, M. Kracht, and M. L. Schmitz. 2004. Transient and selective NF-kappa B p65 serine 536 phosphorylation induced by T cell costimulation is mediated by I kappa B kinase beta and controls the kinetics of p65 nuclear import. J. Immunol. 172: 6336–44. 533. Sasaki, C. Y., T. J. Barberi, P. Ghosh, and D. L. Longo. 2005. Phosphorylation of RelA/p65 on serine 536 defines an I{kappa}B{alpha}- independent NF-{kappa}B pathway. J. Biol. Chem. 280: 34538–47. 534. Buss, H., A. Dörrie, M. L. Schmitz, E. Hoffmann, K. Resch, and M. Kracht. 2004. Constitutive and interleukin-1-inducible phosphorylation of 157

References p65 NF-{kappa}B at serine 536 is mediated by multiple protein kinases including I{kappa}B kinase (IKK)-{alpha}, IKK{beta}, IKK{epsilon}, TRAF family member-associated (TANK)-binding kinase 1 (TBK. J. Biol. Chem. 279: 55633–43. 535. Munroe, M. E., and G. A. Bishop. 2007. A Costimulatory Function for T Cell CD40. J. Immunol. 178: 671–682. 536. Hogan, P. G., R. S. Lewis, and A. Rao. 2010. Molecular Basis of Calcium Signaling in Lymphocytes: STIM and ORAI. Annu. Rev. Immunol. 28: 491–533. 537. Srikanth, S., and Y. Gwack. 2013. Orai1-NFAT signalling pathway triggered by T cell receptor stimulation. Mol. Cells 35: 182–94. 538. Lioudyno, M. I., J. A. Kozak, A. Penna, O. Safrina, S. L. Zhang, D. Sen, J. Roos, K. a Stauderman, and M. D. Cahalan. 2008. Orai1 and STIM1 move to the immunological synapse and are up-regulated during T cell activation. Proc. Natl. Acad. Sci. U. S. A. 105: 2011–6. 539. NA, C., and C. GR. 1992. Identification of calcineurin as a key signalling enzyme in T-lymphocyte activation. Nature 357: 695–7. 540. Serfling, E., A. Avots, S. Klein-Hessling, R. Rudolf, M. Vaeth, and F. Berberich-Siebelt. 2012. NFATc1/αA: The other Face of NFAT Factors in Lymphocytes. Cell Commun. Signal. 10: 16. 541. Hock, M., M. Vaeth, R. Rudolf, A. K. Patra, D. A. T. Pham, K. Muhammad, T. Pusch, T. Bopp, E. Schmitt, R. Rost, F. Berberich-Siebelt, D. Tyrsin, S. Chuvpilo, A. Avots, E. Serfling, and S. Klein-Hessling. 2013. NFATc1 induction in peripheral T and B lymphocytes. J. Immunol. 190: 2345–53. 542. Schreiber, S. L., and G. R. Crabtree. 1992. The mechanism of action of cyclosporin A and FK506. Immunol. Today 13: 136–42. 543. Frischbutter, S., C. Gabriel, H. Bendfeldt, A. Radbruch, and R. Baumgrass. 2011. Dephosphorylation of Bcl-10 by calcineurin is essential for canonical NF-kB activation in Th cells. Eur. J. Immunol. 41: 2349–2357. 544. Das, J., M. Ho, J. Zikherman, C. Govern, M. Yang, A. Weiss, A. K. Chakraborty, and J. P. Roose. 2009. Digital signaling and hysteresis characterize ras activation in lymphoid cells. Cell 136: 337–51. 545. Seger, R., N. G. Ahn, J. Posada, E. S. Munar, A. M. Jensen, J. A. Cooper, M. H. Cobb, and E. G. Krebs. 1992. Purification and characterization of mitogen-activated protein kinase activator(s) from epidermal growth factor-stimulated A431 cells. J. Biol. Chem. 267: 14373– 81. 546. Yoon, S., and R. Seger. 2006. The extracellular signal-regulated kinase: Multiple substrates regulate diverse cellular functions. Growth Factors 24: 21–44. 547. Wainstein, E., and R. Seger. 2016. The dynamic subcellular localization of ERK: Mechanisms of translocation and role in various organelles. Curr. Opin. Cell Biol. 39: 15–20. 548. Sommers, C. L., J. M. Gurson, R. Surana, M. Barda-Saad, J. Lee, A. Kishor, W. Li, A. J. Gasser, V. A. Barr, M. Miyaji, P. E. Love, and L. E. Samelson. 2008. Bam32: A novel mediator of Erk activation in T cells. Int. Immunol. 20: 811–818. 549. Acuto, O., V. Di Bartolo, and F. Michel. 2008. Tailoring T-cell receptor signals by proximal negative feedback mechanisms. Nat. Rev. Immunol. 8: 699–712. 550. Rincón, M., and G. Pedraza-Alva. 2003. JNK and p38 MAP kinases in CD4+ and CD8+ T cells. Immunol. Rev. 192: 131–42. 551. Rincón, M., A. Whitmarsh, D. D. Yang, L. Weiss, B. Dérijard, P. Jayaraj, R. J. Davis, and R. A. Flavell. 1998. The JNK Pathway Regulates the In Vivo Deletion of Immature CD4 + CD8 + Thymocytes. J. Exp. Med. 188: 1817–1830. 552. Dong, C., D. D. Yang, C. Tournier, A. J. Whitmarsh, J. Xu, R. J. Davis, and R. A. Flavell. 2000. JNK is required for effector T-cell\rfunction but not for T-cell activation. Nature 405: 91–94. 553. Weiss, L., A. J. Whitmarsh, D. D. Yang, M. Rincón, R. J. Davis, and R. A. Flavell. 2000. Regulation of c-Jun NH(2)-terminal kinase (Jnk) gene expression during T cell activation. J. Exp. Med. 191: 139–46. 554. Zhang, J., K. V Salojin, J. X. Gao, M. J. Cameron, I. Bergerot, and T. L. Delovitch. 1999. p38 mitogen-activated protein kinase mediates signal integration of TCR/CD28 costimulation in primary murine T cells. J. Immunol. 162: 3819–29. 555. Li, J. P., C. Y. Yang, H. C. Chuang, J. L. Lan, D. Y. Chen, Y. M. Chen, X. Wang, A. J. Chen, J. W. Belmont, and T. H. Tan. 2014. The phosphatase JKAP/DUSP22 inhibits T-cell receptor signalling and autoimmunity by inactivating Lck. Nat Commun 5: 3618. 556. Yang, C.-Y., J.-P. Li, L.-L. Chiu, J.-L. Lan, D.-Y. Chen, H.-C. Chuang, C.-Y. Huang, and T.-H. Tan. 2014. Dual-specificity phosphatase 14 (DUSP14/MKP6) negatively regulates TCR signaling by inhibiting TAB1 activation. J. Immunol. 192: 1547–57. 557. Tao, J., Y. Gao, M. O. Li, W. He, L. Chen, B. Harvev, R. J. Davis, R. A. Flavell, and Z. Yin. 2007. JNK2 negatively regulates CD8+ T cell effector function and anti-tumor immune response. Eur. J. Immunol. 37: 818–829. 558. Su, B., E. Jacinto, M. Hibi, T. Kallunki, M. Karin, and Y. Ben-Neriah. 1994. JNK is involved in signal integration during costimulation of T lymphocytes. Cell 77: 727–36. 559. Ching-Yi Chen, Fabienne Del Gatto–Konczak, Zhenguo Wu, M. K. 1998. Stabilization of Interleukin-2 mRNA by the c-Jun NH2-Terminal Kinase Pathway. Science . 280: 1945–1949. 560. Salojin, K. V, J. Zhang, and T. L. Delovitch. 1999. TCR and CD28 are coupled via ZAP-70 to the activation of the Vav/Rac-1-/PAK-1/p38 MAPK signaling pathway. J. Immunol. 163: 844–53. 561. Conze, D., T. Krahl, N. Kennedy, L. Weiss, J. Lumsden, P. Hess, R. a Flavell, G. Le Gros, R. J. Davis, and M. Rincón. 2002. c-Jun NH(2)- terminal kinase (JNK)1 and JNK2 have distinct roles in CD8(+) T cell activation. J. Exp. Med. 195: 811–23. 562. Arbour, N., D. Naniche, D. Homann, R. J. Davis, R. A. Flavell, and M. B. A. Oldstone. 2002. c-Jun NH 2 -Terminal Kinase ( JNK ) 1 and JNK2 Signaling Pathways Have Divergent Roles in CD8 ϩ T Cell – mediated Antiviral Immunity. 195: 801–810. 563. Constant, S. L., C. Dong, D. D. Yang, M. Wysk, R. J. Davis, and R. a Flavell. 2000. JNK1 is required for T cell-mediated immunity against Leishmania major infection. J. Immunol. 165: 2671–6. 564. Shah, S. Z. A., D. Zhao, T. Hussain, and L. Yang. 2017. The role of unfolded protein response and mitogen-activated protein kinase signaling in neurodegenerative diseases with special focus on prion diseases. Front. Aging Neurosci. 9: 1–14. 565. Cavigelli, M., F. Dolfi, F. X. Claret, and M. Karin. 1995. Induction of c-fos expression through JNK-mediated TCF/Elk-1 phosphorylation. EMBO J. 14: 5957–64. 566. Sabapathy, K., K. Hochedlinger, S. Y. Nam, A. Bauer, M. Karin, and E. F. Wagner. 2004. Distinct roles for JNK1 and JNK2 in regulating JNK activity and c-Jun-dependent cell proliferation. Mol. Cell 15: 713–725. 567. Jaeschke, A., M. Karasarides, J. J. Ventura, A. Ehrhardt, C. Zhang, R. A. Flavell, K. M. Shokat, and R. J. Davis. 2006. JNK2 Is a Positive Regulator of the cJun Transcription Factor. Mol. Cell 23: 899–911. 568. Walters, R. D., L. F. Drullinger, J. F. Kugel, and J. a Goodrich. 2013. NFATc2 recruits cJun homodimers to an NFAT site to synergistically activate interleukin-2 transcription. Mol. Immunol. 56: 48–56. 569. Blonska, M., B. P. Pappu, R. Matsumoto, H. Li, B. Su, D. Wang, and X. Lin. 2007. The CARMA1-Bcl10 Signaling Complex Selectively Regulates JNK2 Kinase in the T Cell Receptor-Signaling Pathway. Immunity 26: 55–66. 570. Staal, J., Y. Driege, T. Bekaert, A. Demeyer, D. Muyllaert, P. Van Damme, K. Gevaert, and R. Beyaert. 2011. T-cell receptor-induced JNK activation requires proteolytic inactivation of CYLD by MALT1. EMBO J. 30: 1742–1752. 571. Wan, Y. Y., H. Chi, M. Xie, M. D. Schneider, and R. A. Flavell. 2006. The kinase TAK1 integrates antigen and cytokine receptor signaling 158

References for T cell development, survival and function. Nat. Immunol. 7: 851–858. 572. Hirata, Y., A. Sugie, A. Matsuda, S. Matsuda, and S. Koyasu. 2013. TAK1-JNK axis mediates survival signal through Mcl1 stabilization in activated T cells. J. Immunol. 190: 4621–6. 573. Enslen, H., D. M. Brancho, and R. J. Davis. 2000. Molecular determinants that mediate selective activation of p38 MAP kinase isoforms. EMBO J. 19: 1301–11. 574. Salvador, J. M., P. R. Mittelstadt, T. Guszczynski, T. D. Copeland, H. Yamaguchi, E. Appella, A. J. Fornace, and J. D. Ashwell. 2005. Alternative p38 activation pathway mediated by T cell receptor-proximal tyrosine kinases. Nat. Immunol. 6: 390–5. 575. Jirmanova, L., M. L. G. Torchia, N. D. Sarma, P. R. Mittelstadt, and J. D. Ashwell. 2011. Lack of the T cell-specific alternative p38 activation pathway reduces autoimmunity and inflammation. Blood 118: 3280–3289. 576. Alam, M. S., M. M. Gaida, Y. Ogawa, A. G. A. Kolios, F. Lasitschka, and J. D. Ashwell. 2014. Counter-regulation of T cell effector function by differentially activated p38. J. Exp. Med. 211: 1257–1270. 577. Deswal, S., A. Meyer, G. J. Fiala, A. E. Eisenhardt, L. C. Schmitt, M. Salek, T. Brummer, O. Acuto, and W. W. a Schamel. 2013. Kidins220/ARMS associates with B-Raf and the TCR, promoting sustained Erk signaling in T cells. J. Immunol. 190: 1927–35. 578. Kane, L. P., P. G. Andres, K. C. Howland, a K. Abbas, and A. Weiss. 2001. Akt provides the CD28 costimulatory signal for up-regulation of IL-2 and IFN-gamma but not TH2 cytokines. Nat. Immunol. 2: 37–44. 579. Sarbassov, D. D., D. A. Guertin, S. M. Ali, and D. M. Sabatini. 2005. Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science . 307: 1098–101. 580. Gorentla, B. K., C.-K. Wan, and X.-P. Zhong. 2011. Negative regulation of mTOR activation by diacylglycerol kinases. Blood 117: 4022–31. 581. Le Borgne, M., E. L. Filbert, and A. S. Shaw. 2013. Kinase suppressor of Ras 1 is not required for the generation of regulatory and memory T cells. PLoS One 8: e57137. 582. BEUGNET, A., A. R. TEE, P. M. TAYLOR, and C. G. PROUD. 2003. Regulation of targets of mTOR (mammalian target of rapamycin) signalling by intracellular amino acid availability. Biochem. J. 372: 555–566. 583. Delgoffe, G. M., T. P. Kole, Y. Zheng, P. E. Zarek, K. L. Matthews, B. Xiao, P. F. Worley, S. C. Kozma, and J. D. Powell. 2009. The mTOR kinase differentially regulates effector and regulatory T cell lineage commitment. Immunity 30: 832–44. 584. Delgoffe, G. M., K. N. Pollizzi, A. T. Waickman, E. Heikamp, D. J. Meyers, M. R. Horton, B. Xiao, P. F. Worley, and J. D. Powell. 2011. The kinase mTOR regulates the differentiation of helper T cells through the selective activation of signaling by mTORC1 and mTORC2. Nat. Immunol. 12: 295–303. 585. Pearce, E. L., M. C. Poffenberger, C. Chang, and R. G. Jones. 2013. Fueling immunity: insights into metabolism and lymphocyte function. Science . 342: 1242454. 586. Otto, A. M. 2016. Warburg effect(s)—a biographical sketch of Otto Warburg and his impacts on tumor metabolism. Cancer Metab. 4: 5. 587. Carr, E. L., A. Kelman, G. S. Wu, R. Gopaul, E. Senkevitch, A. Aghvanyan, A. M. Turay, and K. a Frauwirth. 2010. Glutamine uptake and metabolism are coordinately regulated by ERK/MAPK during T lymphocyte activation. J. Immunol. 185: 1037–44. 588. Wang, R., and D. R. Green. 2012. Metabolic reprogramming and metabolic dependency in T cells. Immunol. Rev. 249: 14–26. 589. Bronietzki, A. W., M. Schuster, and I. Schmitz. 2015. Autophagy in T-cell development, activation and differentiation. Immunol. Cell Biol. 93: 25–34. 590. Jacquin, E., and L. Apetoh. 2018. Cell-Intrinsic Roles for Autophagy in Modulating CD4 T Cell Functions. Front. Immunol. 9: 1–9. 591. Hubbard, V. M., R. Valdor, B. Patel, R. Singh, A. M. Cuervo, and F. Macian. 2010. Macroautophagy Regulates Energy Metabolism during Effector T Cell Activation. J. Immunol. 185: 7349–7357. 592. Valdor, R., E. Mocholi, Y. Botbol, I. Guerrero-Ros, D. Chandra, H. Koga, C. Gravekamp, A. M. Cuervo, and F. Macian. 2014. Chaperone- mediated autophagy regulates T cell responses through targeted degradation of negative regulators of T cell activation. Nat. Immunol. 15: 1046– 1054. 593. Pua, H. H., I. Dzhagalov, M. Chuck, N. Mizushima, and Y.-W. He. 2007. A critical role for the autophagy gene Atg5 in T cell survival and proliferation. J. Exp. Med. 204: 25–31. 594. Štefanová, I., J. R. Dorfman, and R. N. Germain. 2002. Self-recognition promotes the foreign antigen sensitivity of naive T lymphocytes. Nature 420: 429–434. 595. Markegard, E., E. Trager, C. wen O. Yang, W. Zhang, A. Weiss, and J. P. Roose. 2011. Basal LAT-diacylglycerol-rasGRP1 signals in t cells maintain TCRα gene expression. PLoS One 6. 596. Cho, J., and J. Sprent. 2018. TCR tuning of T cell subsets. Immunol. Rev. 283: 129–137. 597. Preston, G. C., L. V Sinclair, A. Kaskar, J. L. Hukelmann, M. N. Navarro, I. Ferrero, H. R. MacDonald, V. H. Cowling, and D. A. Cantrell. 2015. Single cell tuning of Myc expression by antigen receptor signal strength and interleukin-2 in T lymphocytes. EMBO J. 34: 2008–2024. 598. VanPanhuys, N., F. Klauschen, and R. Germain. 2014. T-Cell-Receptor-Dependent Signal Intensity Dominantly Controls CD4+T Cell Polarization InVivo. Immunity 41: 63–74. 599. Altan-Bonnet, G., and R. N. Germain. 2005. Modeling T Cell Antigen Discrimination Based on Feedback Control of Digital ERK Responses. PLoS Biol. 3: e356. 600. Allison, K. A., E. Sajti, J. G. Collier, D. Gosselin, T. D. Troutman, E. L. Stone, S. M. Hedrick, and C. K. Glass. 2016. Affinity and dose of TCR engagement yield proportional enhancer and gene activity in CD4+ T cells. Elife 5: 921–31. 601. Caron, E., R. Roncagalli, T. Hase, W. E. Wolski, M. Choi, M. G. Menoita, S. Durand, A. García-Blesa, I. Fierro-Monti, T. Sajic, M. Heusel, T. Weiss, M. Malissen, R. Schlapbach, B. C. Collins, S. Ghosh, H. Kitano, R. Aebersold, B. Malissen, and M. Gstaiger. 2017. Precise Temporal Profiling of Signaling Complexes in Primary Cells Using SWATH Mass Spectrometry. Cell Rep. 18: 3219–3226. 602. López-Cabrera, M., E. Muñoz, M. V Blázquez, M. a Ursa, a G. Santis, and F. Sánchez-Madrid. 1995. Transcriptional regulation of the gene encoding the human C-type lectin leukocyte receptor AIM/CD69 and functional characterization of its tumor necrosis factor-alpha-responsive elements. J. Biol. Chem. 270: 21545–51. 603. Castellanos, M. C., C. Munoz, M. C. Montoya, E. Lara-Pezzi, M. Lopez-Cabrera, and M. O. de Landazuri. 1997. Expression of the leukocyte early activation antigen CD69 is regulated by the transcription factor AP-1. J. Immunol. 159: 5463–5473. 604. Risso, A., D. Smilovich, M. C. Capra, I. Baldissarro, G. Yan, A. Bargellesi, and M. E. Cosulich. 1991. CD69 in resting and activated T lymphocytes. Its association with a GTP binding protein and biochemical requirements for its expression. J. Immunol. 146: 4105–14. 605. Cosulich, M. E., A. Rubartelli, A. Risso, F. Cozzolino, and A. Bargellesi. 1987. Functional characterization of an antigen involved in an early step of T-cell activation. Proc. Natl. Acad. Sci. U. S. A. 84: 4205–9. 606. D’Ambrosio, D., D. a Cantrell, L. Frati, A. Santoni, and R. Testi. 1994. Involvement of p21ras activation in T cell CD69 expression. Eur. J. Immunol. 24: 616–20. 607. Murata, K., M. Inami, A. Hasegawa, S. Kubo, M. Kimura, M. Yamashita, H. Hosokawa, T. Nagao, K. Suzuki, K. Hashimoto, H. Shinkai, H. Koseki, M. Taniguchi, S. F. Ziegler, and T. Nakayama. 2003. CD69-null mice protected from arthritis induced with anti-type II collagen 159

References antibodies. Int. Immunol. 15: 987–92. 608. Sancho, D., M. Gómez, F. Viedma, E. Esplugues, M. Gordón-Alonso, M. Angeles García-López, H. de la Fuente, C. Martínez-A, P. Lauzurica, and F. Sánchez-Madrid. 2003. CD69 downregulates autoimmune reactivity through active transforming growth factor-β production in collagen-induced arthritis. J. Clin. Invest. 112: 872–882. 609. Alari-Pahissa, E., L. Notario, E. Lorente, J. Vega-Ramos, A. Justel, D. López, J. a Villadangos, and P. Lauzurica. 2012. CD69 does not affect the extent of T cell priming. PLoS One 7: e48593. 610. Shiow, L. R., D. B. Rosen, N. Brdicková, Y. Xu, J. An, L. L. Lanier, J. G. Cyster, and M. Matloubian. 2006. CD69 acts downstream of interferon-alpha/beta to inhibit S1P1 and lymphocyte egress from lymphoid organs. Nature 440: 540–4. 611. de la Fuente, H., A. Cruz-Adalia, G. Martinez del Hoyo, D. Cibrian-Vera, P. Bonay, D. Perez-Hernandez, J. Vazquez, P. Navarro, R. Gutierrez-Gallego, M. Ramirez-Huesca, P. Martin, and F. Sanchez-Madrid. 2014. The Leukocyte Activation Receptor CD69 Controls T Cell Differentiation through Its Interaction with Galectin-1. Mol. Cell. Biol. 34: 2479–2487. 612. Toscano, M. A., G. A. Bianco, J. M. Ilarregui, D. O. Croci, J. Correale, J. D. Hernandez, N. W. Zwirner, F. Poirier, E. M. Riley, L. G. Baum, and G. A. Rabinovich. 2007. Differential glycosylation of TH1, TH2 and TH-17 effector cells selectively regulates susceptibility to cell death. Nat. Immunol. 8: 825–834. 613. Ilarregui, J. M., D. O. Croci, G. A. Bianco, M. A. Toscano, M. Salatino, M. E. Vermeulen, J. R. Geffner, and G. A. Rabinovich. 2009. Tolerogenic signals delivered by dendritic cells to T cells through a galectin-1-driven immunoregulatory circuit involving interleukin 27 and interleukin 10. Nat. Immunol. 10: 981–991. 614. Cibrian, D., M. L. Saiz, H. De La Fuente, R. Sánchez-Díaz, O. Moreno-Gonzalo, I. Jorge, A. Ferrarini, J. Vázquez, C. Punzón, M. Fresno, M. Vicente-Manzanares, E. Daudén, P. M. Fernández-Salguero, P. Martín, and F. Sánchez-Madrid. 2016. CD69 controls the uptake of L-tryptophan through LAT1-CD98 and AhR-dependent secretion of IL-22 in psoriasis. Nat. Immunol. 17: 985–996. 615. Cibrián, D., and F. Sánchez-Madrid. 2017. CD69: from activation marker to metabolic gatekeeper. Eur. J. Immunol. 47: 946–953. 616. Lauzurica, P., D. Sancho, M. Torres, B. Albella, M. Marazuela, T. Merino, J. A. Bueren, C. Martínez-A, and F. Sánchez-Madrid. 2000. Phenotypic and functional characteristics of hematopoietic cell lineages in CD69-deficient mice. Blood 95: 2312–20. 617. Smith, K. A. 1988. Interleukin-2: inception, impact, and implications. Science . 240: 1169–76. 618. Jain, J., C. Loh, and A. Rao. 1995. Transcriptional regulation of the IL-2 gene. Curr. Opin. Immunol. 7: 333–42. 619. Liao, W., J.-X. Lin, and W. J. Leonard. 2013. Interleukin-2 at the crossroads of effector responses, tolerance, and immunotherapy. Immunity 38: 13–25. 620. Koike, T., H. Yamagishi, Y. Hatanaka, A. Fukushima, J. Chang, Y. Xia, M. Fields, P. Chandler, and M. Iwashima. 2003. A novel ERK- dependent signaling process that regulates interleukin-2 expression in a late phase of T cell activation. J. Biol. Chem. 278: 15685–92. 621. Whitehurst, C. E., and T. D. Geppert. 1996. MEK1 and the extracellular signal-regulated kinases are required for the stimulation of IL-2 gene transcription in T cells. J. Immunol. 156: 1020–9. 622. Decker, E. L., C. Skerka, and P. F. Zipfel. 1998. The Early Growth Response Protein (EGR-1) Regulates Interleukin-2 Transcription by Synergistic Interaction with the Nuclear Factor of Activated T Cells. J. Biol. Chem. 273: 26923–26930. 623. Collins, S., M. A. Lutz, P. E. Zarek, R. A. Anders, G. J. Kersh, and J. D. Powell. 2008. Opposing regulation of T cell function by Egr-1/NAB2 and Egr-2/Egr-3. Eur. J. Immunol. 38: 528–536. 624. Poltorak, M., I. Meinert, J. C. Stone, B. Schraven, and L. Simeoni. 2014. Sos1 regulates sustained TCR-mediated Erk activation. Eur. J. Immunol. 44: 1535–1540. 625. Abbas, A. K., E. Trotta, D. R. Simeonov, A. Marson, and J. A. Bluestone. 2018. Revisiting IL-2: Biology and therapeutic prospects. Sci. Immunol. 3. 626. Secrist, J. P., L. a Burns, L. Karnitz, G. a Koretzky, and R. T. Abraham. 1993. Stimulatory effects of the protein tyrosine phosphatase inhibitor, pervanadate, on T-cell activation events. J. Biol. Chem. 268: 5886–93. 627. Huang, C.-Y., and T.-H. Tan. 2012. DUSPs, to MAP kinases and beyond. Cell Biosci. 2: 24. 628. Jeffrey, K. L., M. Camps, C. Rommel, and C. R. Mackay. 2007. Targeting dual-specificity phosphatases: manipulating MAP kinase signalling and immune responses. Nat. Rev. Drug Discov. 6: 391–403. 629. Reiterer, V., D. Fey, W. Kolch, B. N. Kholodenko, and H. Farhan. 2013. Pseudophosphatase STYX modulates cell-fate decisions and cell migration by spatiotemporal regulation of ERK1/2. Proc. Natl. Acad. Sci. U. S. A. [epub-ah. 630. Geetha, N., J. Mihaly, A. Stockenhuber, F. Blasi, P. Uhrin, B. R. Binder, M. Freissmuth, and J. M. Breuss. 2011. Signal integration and coincidence detection in the mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) cascade: Concomitant activation of receptor tyrosine kinases and of LRP-1 leads to sustained ERK phosphorylation via down-regu. J. Biol. Chem. 286: 25663–25674. 631. Caunt, C. J., and S. M. Keyse. 2013. Dual-specificity MAP kinase phosphatases (MKPs): Shaping the outcome of MAP kinase signalling. FEBS J. 280: 489–504. 632. Sun, H., C. H. Charles, L. F. Lau, and N. K. Tonks. 1993. MKP-1 (3CH134), an immediate early gene product, is a dual specificity phosphatase that dephosphorylates MAP kinase in vivo. Cell 75: 487–493. 633. Wu, J. J., and A. M. Bennett. 2005. Essential role for mitogen-activated protein (MAP) kinase phosphatase-1 in stress-responsive MAP kinase and cell survival signaling. J. Biol. Chem. 280: 16461–16466. 634. Low, H. B., and Y. Zhang. 2016. Regulatory Roles of MAPK Phosphatases in Cancer. Immune Netw. 16: 85–98. 635. Arimura, Y., and J. Yagi. 2010. Comprehensive Expression Profiles of Genes for Protein Tyrosine Phosphatases in Immune Cells. Sci. Signal. 3: 1–11. 636. Zhang, Y., J. M. Reynolds, S. H. Chang, N. Martin-Orozco, Y. Chung, R. I. Nurieva, and C. Dong. 2009. MKP-1 is necessary for T cell activation and function. J. Biol. Chem. 284: 30815–30824. 637. Masiero, M., S. Minuzzo, I. Pusceddu, L. Moserle, L. Persano, V. Agnusdei, V. Tosello, G. Basso, A. Amadori, and S. Indraccolo. 2011. Notch3-mediated regulation of MKP-1 levels promotes survival of T acute lymphoblastic leukemia cells. Leukemia 25: 588–598. 638. Rohan, P. J., P. Davis, C. A. Moskaluk, M. Kearns, H. Krutzsch, U. Siebenlist, and K. Kelly. 1993. PAC-1: a mitogen-induced nuclear protein tyrosine phosphatase. Science . 259: 1763–6. 639. Ward, Y., S. Gupta, P. Jensen, M. Wartmann, R. J. Davis, and K. Kelly. 1994. Control of MAP kinase activation by the mitogen-induced threonine/tyrosine phosphatase PAC1. Nature 367: 651–4. 640. Lu, D., L. Liu, X. Ji, Y. Gao, X. Chen, Y. Liu, Y. Liu, X. Zhao, Y. Li, Y. Li, Y. Jin, Y. Zhang, M. A. McNutt, and Y. Yin. 2015. The phosphatase DUSP2 controls the activity of the transcription activator STAT3 and regulates T H 17 differentiation. Nat. Immunol. 16: 1263–1273. 641. Alonso, A., M. Saxena, S. Williams, and T. Mustelin. 2001. Inhibitory Role for Dual Specificity Phosphatase VHR in T Cell Antigen Receptor and CD28-induced Erk and Jnk Activation. J. Biol. Chem. 276: 4766–4771. 642. Alonso, A., S. Rahmouni, S. Williams, M. van Stipdonk, L. Jaroszewski, A. Godzik, R. T. Abraham, S. P. Schoenberger, and T. Mustelin. 2003. Tyrosine phosphorylation of VHR phosphatase by ZAP-70. Nat. Immunol. 4: 44–48. 160

References

643. Chu, Y., P. a Solski, R. Khosravi-Far, C. J. Der, and K. Kelly. 1996. The mitogen-activated protein kinase phosphatases PAC1,MKP-1, and MKP-2 have unique substrate specificities and reduced activity in vivo toward the ERK2 sevenmarker mutation. J. Biol. Chem. 271(11): 6497– 6501. 644. Brondello, J. M., A. Brunet, J. Pouysségur, and F. R. McKenzie. 1997. The dual specificity mitogen-activated protein kinase phosphatase-1 and -2 are induced by the p42/p44MAPK cascade. J. Biol. Chem. 272: 1368–76. 645. Peng, D. J., J. Y. Zhou, and G. S. Wu. 2010. Post-translational regulation of mitogen-activated protein kinase phosphatase-2 (MKP-2) by ERK. Cell Cycle 9: 4650–4655. 646. Chen, P., D. Hutter, X. Yang, M. Gorospe, R. J. Davis, and Y. Liu. 2001. Discordance between the Binding Affinity of Mitogen-activated Protein Kinase Subfamily Members for MAP Kinase Phosphatase-2 and Their Ability to Activate the Phosphatase Catalytically. J. Biol. Chem. 276: 29440–29449. 647. Huang, C. Y., Y. C. Lin, W. Y. Hsiao, F. H. Liao, P. Y. Huang, and T. H. Tan. 2012. DUSP4 deficiency enhances CD25 expression and CD4 + T-cell proliferation without impeding T-cell development. Eur. J. Immunol. 42: 476–488. 648. Hsiao, W. Y., Y. C. Lin, F. H. Liao, Y. C. Chan, and C. Y. Huang. 2015. Dual-specificity phosphatase 4 regulates STAT5 protein stability and helper T cell polarization. PLoS One 10: 1–24. 649. Barbour, M., R. Plevin, and H.-R. Jiang. 2016. MAP kinase phosphatase 2 deficient mice develop attenuated experimental autoimmune encephalomyelitis through regulating dendritic cells and T cells. Sci. Rep. 6: 38999. 650. Levy-Nissenbaum, O., O. Sagi-Assif, D. Kapon, S. Hantisteanu, T. Burg, P. Raanani, A. Avigdor, I. Ben-Bassat, and I. P. Witz. 2003. Dual- specificity phosphatase Pyst2-L is constitutively highly expressed in myeloid leukemia and other malignant cells. Oncogene 22: 7649–7660. 651. Alonso, A., J. J. Merlo, S. Na, N. Kholod, L. Jaroszewski, A. Kharitonenkov, S. Williams, A. Godzik, J. D. Posada, and T. Mustelin. 2002. Inhibition of T cell antigen receptor signaling by VHR-related MKPX (VHX), a new dual specificity phosphatase related to VH1 related (VHR). J. Biol. Chem. 277: 5524–5528. 652. Castro-Sánchez, P., R. Ramirez-Munoz, A. Lamana, A. Ortiz, I. González-Álvaro, and P. Roda-Navarro. 2017. mRNA profilin identifies low levels of phosphatases dual-specific phosphatase-7 (DUSP7) and cell division cycle-25B (CDC25B) in patients with early arthritis. Clin. Exp. Immunol. 189: 113–119. 653. Liu, R., J. H. van Berlo, A. J. York, R. J. Vagnozzi, M. Maillet, and J. D. Molkentin. 2016. DUSP8 Regulates Cardiac Ventricular Remodeling by Altering ERK1/2 Signaling. Circ. Res. 119: 249–60. 654. Muda, M., A. Theodosiou, N. Rodrigues, U. Boschert, M. Camps, C. Gillieron, K. Davies, A. Ashworth, and S. Arkinstall. 1996. The dual specificity phosphatases M3/6 and MKP-3 are highly selective for inactivation of distinct mitogen-activated protein kinases. J. Biol. Chem. 271: 27205–8. 655. Chen, Y. R., A. Shrivastava, and T. H. Tan. 2001. Down-regulation of the c-Jun N-terminal kinase (JNK) phosphatase M3/6 and activation of JNK by hydrogen peroxide and pyrrolidine. Oncogene 20: 367–374. 656. Niedzielska, M., F. A. M. Raffi, J. Tel, S. Muench, K. Jozefowski, N. Alati, J. Mages, U. Billmeier, M. Schiemann, U. K. Appelt, S. Wirtz, T. Sparwasser, C. G. Figdor, and S. M. Keyse. 2015. Selective Expression of the MAPK Phosphatase Dusp9/MKP-4 in Mouse Plasmacytoid Dendritic Cells and Regulation of IFN- β Production. J. Immunol. 195: 1753–1762. 657. Tanoue, T., T. Yamamoto, R. Maeda, and E. Nishida. 2001. A Novel MAPK Phosphatase MKP-7 Acts Preferentially on JNK/SAPK and p38α and β MAPKs. J. Biol. Chem. 276: 26629–26639. 658. Muda, M., U. Boschert, A. Smith, B. Antonsson, C. Gillieron, C. Chabert, M. Camps, I. Martinou, A. Ashworth, and S. Arkinstall. 1997. Molecular cloning and functional characterization of a novel mitogen-activated protein kinase phosphatase, MKP-4. J. Biol. Chem. 272: 5141– 5151. 659. Theodosiou, A., A. Smith, C. Gillieron, S. Arkinstall, and A. Ashworth. 1999. MKP5, a new member of the MAP kinase phosphatase family, which selectively dephosphorylates stress-activated kinases. Oncogene 18: 6981–6988. 660. Zhang, Y., J. N. Blattman, N. J. Kennedy, J. Duong, T. Nguyen, Y. Wang, R. J. Davis, P. D. Greenberg, R. A. Flavell, and C. Dong. 2004. Regulation of innate and adaptive immune responses by MAP kinase phosphatase 5. Nature 430: 793–7. 661. Cheng, Q., Q. Zhang, X. Xu, L. Yin, L. Sun, X. Lin, C. Dong, and W. Pan. 2014. MAPK Phosphotase 5 Deficiency Contributes to Protection against Blood-Stage Plasmodium yoelii 17XL Infection in Mice. J. Immunol. 192: 3686–3696. 662. Marti, F., A. Krause, N. H. Post, C. Lyddane, B. Dupont, M. Sadelain, and P. D. King. 2001. Negative-feedback regulation of CD28 costimulation by a novel mitogen-activated protein kinase phosphatase, MKP6. J. Immunol. 166: 197–206. 663. Zhang, Y., K. C. Nallaparaju, X. Liu, H. Jiao, J. M. Reynolds, Z.-X. Wang, and C. Dong. 2015. MAPK phosphatase 7 regulates T cell differentiation via inhibiting ERK-mediated IL-2 expression. J. Immunol. 194: 3088–95. 664. Bertin, S., B. Lozano-Ruiz, V. Bachiller, I. García-Martínez, S. Herdman, P. Zapater, R. Francés, J. Such, J. Lee, E. Raz, and J. M. González- Navajas. 2015. Dual-specificity phosphatase 6 regulates CD4+ T-cell functions and restrains spontaneous colitis in IL-10-deficient mice. Mucosal Immunol. 8: 505–15. 665. González-Navajas, J. M., S. Fine, J. Law, S. K. Datta, K. P. Nguyen, M. Yu, M. Corr, K. Katakura, L. Eckman, J. Lee, and E. Raz. 2010. TLR4 signaling in effector CD4+ T cells regulates TCR activation and experimental colitis in mice. J. Clin. Invest. 120: 570–581. 666. Kovanen, P. E., J. Bernard, A. Al-Shami, C. Liu, J. Bollenbacher-Reilley, L. Young, C. Pise-Masison, R. Spolski, and W. J. Leonard. 2008. T- cell development and function are modulated by dual specificity phosphatase DUSP5. J. Biol. Chem. 283: 17362–17369. 667. Moon, S. J., M. A. Lim, J. S. Park, J. K. Byun, S. M. Kim, M. K. Park, E. K. Kim, Y. M. Moon, J. K. Min, S. M. Ahn, S. H. Park, and M. La Cho. 2014. Dual-specificity phosphatase 5 attenuates autoimmune arthritis in mice via reciprocal regulation of the Th17/Treg cell balance and inhibition of osteoclastogenesis. Arthritis Rheumatol. 66: 3083–3095. 668. Kidger, A. M., L. K. Rushworth, J. Stellzig, J. Davidson, C. J. Bryant, C. Bayley, E. Caddye, T. Rogers, S. M. Keyse, and C. J. Caunt. 2017. Dual-specificity phosphatase 5 controls the localized inhibition, propagation, and transforming potential of ERK signaling. Proc. Natl. Acad. Sci. U. S. A. 114: E317–E326. 669. Ekerot, M., M. P. Stavridis, L. Delavaine, M. P. Mitchell, C. Staples, D. M. Owens, I. D. Keenan, R. J. Dickinson, K. G. Storey, and S. M. Keyse. 2008. Negative-feedback regulation of FGF signalling by DUSP6/MKP-3 is driven by ERK1/2 and mediated by Ets factor binding to a conserved site within the DUSP6/MKP-3 gene promoter. Biochem. J. 412: 287–98. 670. Chechlinska, M., J. K. Siwicki, M. Gos, M. Oczko-Wojciechowska, M. Jarzab, A. Pfeifer, B. Jarzab, and J. Steffen. 2009. Molecular signature of cell cycle exit induced in human T lymphoblasts by IL-2 withdrawal. BMC Genomics 10: 261. 671. Kovanen, P. E., A. Rosenwald, J. Fu, E. M. Hurt, L. T. Lam, J. M. Giltnane, G. Wright, L. M. Staudt, and W. J. Leonard. 2003. Analysis of γc-family cytokine target genes: Identification of dual-specificity phosphatase 5 (DUSP5) as a regulator of mitogen-activated protein kinase activity in interleukin-2 signaling. J. Biol. Chem. 278: 5205–5213. 672. Kucharska, A., L. K. Rushworth, C. Staples, N. A. Morrice, and S. M. Keyse. 2009. Regulation of the inducible nuclear dual-specificity phosphatase DUSP5 by ERK MAPK. Cell. Signal. 21: 1794–1805. 161

References

673. Nunes-Xavier, C. E., C. Tárrega, R. Cejudo-Marín, J. Frijhoff, Å. Sandin, A. Östman, and R. Pulido. 2010. Differential up-regulation of MAP kinase phosphatases MKP3/DUSP6 and DUSP5 by Ets2 and c-Jun converge in the control of the growth arrest versus proliferation response of MCF-7 breast cancer cells to phorbol ester. J. Biol. Chem. 285: 26417–26430. 674. Marchetti, S., C. Gimond, T. Touboul, D. Roux, and G. Pagès. 2005. Extracellular signal-regulated kinases phosphorylate mitogen-activated protein kinase phosphatase 3 / DUSP6 at serines 159 and 197 , two sites critical for its proteasomal degradation. Mol Cell Biol 25: 854–864. 675. Bermudez, O., S. Marchetti, G. Pagès, and C. Gimond. 2008. Post-translational regulation of the ERK phosphatase DUSP6/MKP3 by the mTOR pathway. Oncogene 27: 3685–3691. 676. Xie, X. L., X. Nie, J. Wu, F. Zhang, L. L. Zhao, Y. L. Lin, Y. J. Yin, H. Liu, Y. N. Shu, S. B. Miao, H. Li, P. Chen, and M. Han. 2015. Smooth muscle 22alpha facilitates angiotensin II-induced signaling and vascular contraction. J. Mol. Med. 93: 547–558. 677. Cheng, Y., Y. Zhu, J. Xu, M. Yang, P. Chen, W. Xu, J. Zhao, L. Geng, and S. Gong. 2018. PKN2 in colon cancer cells inhibits M2 phenotype polarization of tumor-associated macrophages via regulating DUSP6-Erk1/2 pathway. Mol. Cancer 17: 13. 678. Iwamoto, N., L. A. D’Alessandro, S. Depner, B. Hahn, B. A. Kramer, P. Lucarelli, A. Vlasov, M. Stepath, M. E. Böhm, D. Deharde, G. Damm, D. Seehofer, W. D. Lehmann, U. Klingmüller, and M. Schilling. 2016. Context-specific flow through the MEK/ERK module produces cell- and ligand-specific patterns of ERK single and double phosphorylation. Sci. Signal. 9: ra13. 679. Li, Q. J., J. Chau, P. J. R. Ebert, G. Sylvester, H. Min, G. Liu, R. Braich, M. Manoharan, J. Soutschek, P. Skare, L. O. Klein, M. M. Davis, and C. Z. Chen. 2007. miR-181a Is an Intrinsic Modulator of T Cell Sensitivity and Selection. Cell 129: 147–161. 680. Bronevetsky, Y., A. V Villarino, C. J. Eisley, R. Barbeau, A. J. Barczak, G. a Heinz, E. Kremmer, V. Heissmeyer, M. T. McManus, D. J. Erle, A. Rao, and K. M. Ansel. 2013. T cell activation induces proteasomal degradation of Argonaute and rapid remodeling of the microRNA repertoire. J. Exp. Med. 210: 417–32. 681. Schaffert, S. A., C. Loh, S. Wang, C. P. Arnold, R. C. Axtell, E. W. Newell, G. Nolan, K. M. Ansel, M. M. Davis, L. Steinman, and C.-Z. Chen. 2015. mir-181a-1/b-1 Modulates Tolerance through Opposing Activities in Selection and Peripheral T Cell Function. J. Immunol. 195: 1470–1479. 682. Palin, A. C., V. Ramachandran, S. Acharya, and D. B. Lewis. 2013. Human neonatal naive CD4+ T cells have enhanced activation-dependent signaling regulated by the microRNA miR-181a. J. Immunol. 190: 2682–91. 683. Huang, L., N. H. R. Litjens, N. M. Kannegieter, M. Klepper, C. C. Baan, and M. G. H. Betjes. 2017. pERK-dependent defective TCR- mediated activation of CD4+ T cells in end-stage renal disease patients. Immun. Ageing 14: 14. 684. Li, G., M. Yu, W.-W. Lee, M. Tsang, E. Krishnan, C. M. Weyand, and J. J. Goronzy. 2012. Decline in miR-181a expression with age impairs T cell receptor sensitivity by increasing DUSP6 activity. Nat. Med. 18: 1518–24. 685. Bermudez, O., P. Jouandin, J. Rottier, C. Bourcier, G. Pagès, and C. Gimond. 2011. Post-transcriptional regulation of the DUSP6/MKP-3 phosphatase by MEK/ERK signaling and hypoxia. J. Cell. Physiol. 226: 276–284. 686. Oxenius, a, M. F. Bachmann, R. M. Zinkernagel, and H. Hengartner. 1998. Virus-specific MHC-class II-restricted TCR-transgenic mice: effects on humoral and cellular immune responses after viral infection. Eur. J. Immunol. 28: 390–400. 687. Maillet, M., N. H. Purcell, M. A. Sargent, A. J. York, O. F. Bueno, and J. D. Molkentin. 2008. DUSP6 (MKP3) null mice show enhanced ERK1/2 phosphorylation at baseline and increased myocyte proliferation in the heart affecting disease susceptibility. J. Biol. Chem. 283: 31246– 31255. 688. Battegay, M., S. Cooper, A. Althage, J. Bänziger, H. Hengartner, and R. M. Zinkernagel. 1991. Quantification of lymphocytic choriomeningitis virus with an immunological focus assay in 24- or 96-well plates. J. Virol. Methods 33: 191–198. 689. Gillis, Steven and Smith, K. 1977. Long term culture of tumor-specific cytotoxic T cells. Nature 268: 154–156. 690. Bubeník, J., E. Lotzová, M. Indrová, J. Šímová, T. Jandlová, and D. Bubeníková. 1992. Use of IL-2 gene transfer in local immunotherapy of cancer. Cancer Lett. 62: 257–262. 691. Anjuère, F., D. Kuznetsov, P. Romero, J. C. Cerottini, C. V Jongeneel, and I. F. Luescher. 1997. Differential roles of T cell receptor alpha and beta chains in ligand binding among H-2Kd-restricted cytolytic T lymphocyte clones specific for a photoreactive Plasmodium berghei circumsporozoite peptide derivative. J. Biol. Chem. 272: 8505–14. 692. Kessler, B. 1999. T Cell Recognition of Hapten. ANATOMY OF T CELL RECEPTOR BINDING OF A H-2Kd-ASSOCIATED PHOTOREACTIVE PEPTIDE DERIVATIVE. J. Biol. Chem. 274: 3622–3631. 693. Doucey, M.-A., L. Goffin, D. Naeher, O. Michielin, P. Baumgärtner, P. Guillaume, E. Palmer, and I. F. Luescher. 2003. CD3 delta establishes a functional link between the T cell receptor and CD8. J. Bol. Chem. 278: 3257–64. 694. Minowada, J., T. Onuma, and G. E. Moore. 1972. Rosette-forming human lymphoid cell lines. I. Establishment and evidence for origin of thymus-derived lymphocytes. J. Natl. Cancer Inst. 49: 891–5. 695. Kaltenberg, J., L. M. Plum, J. L. Ober-Blöbaum, A. Hönscheid, L. Rink, and H. Haase. 2010. Zinc signals promote IL-2-dependent proliferation of T cells. Eur. J. Immunol. 40: 1496–1503. 696. Basler, M., E. Maurits, G. De Bruin, J. Koerner, H. S. Overkleeft, and M. Groettrup. 2018. Amelioration of autoimmunity with an inhibitor selectively targeting all active centres of the immunoproteasome. Br J Pharmacol. 2018 Jan;175(1):38-52. 697. Ledderose, C., J. Heyn, E. Limbeck, and S. Kreth. 2011. Selection of reliable reference genes for quantitative real-time PCR in human T cells and neutrophils. BMC Res. Notes 4: 427. 698. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25: 402–8. 699. Krzywinski, M., and N. Altman. 2014. Points of significance: Designing comparative experiments. Nat. Methods 11: 597–598. 700. Krzywinski, M., and N. Altman. 2013. Points of significance: Significance, P values and t-tests. Nat. Methods 10: 1041–1042. 701. Leidinger, A. 2012. The influence of an LMP7-selective Inhibitor on T cell signalling. Bachelor's thesis, University of Konstanz, Konstanz, Germany. 702. Schmidt, C. 2013. Consequence of Immunoproteasome Inhibition and Deficiency on Signal Transduction via the T Cell Receptor. Master's thesis, University of Konstanz, Konstanz, Germany. 703. Frisan, T., V. Levitsky, and M. G. Masucci. 2000. Variations in proteasome subunit composition and enzymatic activity in B-lymphoma lines and normal B cells. Int. J. Cancer 88: 881–888. 704. Janes, K. A. 2015. An analysis of critical factors for quantitative immunoblotting. Sci. Signal. 8: 1–12. 705. Vangala, J. R., S. Dudem, N. Jain, and S. V Kalivendi. 2014. Regulation of PSMB5 protein and β subunits of mammalian proteasome by constitutively activated signal transducer and activator of transcription 3 (STAT3): potential role in bortezomib-mediated anticancer therapy. J. Biol. Chem. 289: 12612–22. 706. Zhang, X., V. Baladandayuthapani, H. Lin, G. Mulligan, B. Li, D.-L. W. Esseltine, L. Qi, J. Xu, W. Hunziker, B. Barlogie, S. Z. Usmani, Q. Zhang, J. Crowley, A. Hoering, J. J. Shah, D. M. Weber, E. E. Manasanch, S. K. Thomas, B.-Z. Li, H.-H. Wang, J. Zhang, I. Kuiatse, J.-L. Tang, H. Wang, J. He, J. Yang, E. Milan, S. Cenci, W.-C. Ma, Z.-Q. Wang, R. E. Davis, L. Yang, and R. Z. Orlowski. 2016. Tight Junction Protein 1 162

References

Modulates Proteasome Capacity and Proteasome Inhibitor Sensitivity in Multiple Myeloma via EGFR/JAK1/STAT3 Signaling. Cancer Cell 29: 639–652. 707. Berger, T. 2018. Influence of Immunoproteasome Inhibition on the regulation of dual specificity phosphatases during T Cell activation. Master's thesis, University of Konstanz, Konstanz, Germany. 708. Rane, S., R. Das, V. Ranganathan, S. Prabhu, A. Das, H. Mattoo, J. Durdik, A. George, S. Rath, and V. Bal. 2014. Peripheral residence of naïve CD4 T cells induces MHC class II-dependent alterations in phenotype and function. BMC Biol. 12: 106. 709. Li, G. Y., Y. Zhou, R. S. Ying, L. Shi, Y. Q. Cheng, J. P. Ren, J. W. D. Griffin, Z. S. Jia, C. F. Li, J. P. Moorman, and Z. Q. Yao. 2015. Hepatitis C virus-induced reduction in miR-181a impairs CD4+ T-cell responses through overexpression of DUSP6. Hepatology 61: 1163–1173. 710. Karlsson, M., J. Mathers, R. J. Dickinson, M. Mandl, and S. M. Keyse. 2004. Both nuclear-cytoplasmic shuttling of the dual specificity phosphatase MKP-3 and its ability to anchor MAP kinase in the cytoplasm are mediated by a conserved nuclear export signal. J. Biol. Chem. 279: 41882–41891. 711. Guo, Y., X. Chen, D. Li, H. Liu, Y. Ding, R. Han, Y. Shi, and X. Ma. 2018. PR-957 mediates neuroprotection by inhibiting Th17 differentiation and modulating cytokine production in a mouse model of ischaemic stroke. Clin. Exp. Immunol. 193: 194–206. 712. Yang, B., J. Song, H. Sun, J. Xing, Z. Yang, C. Wei, T. Xu, Z. Yu, Y. Zhang, Y. Wang, H. Chang, Z. Xu, M. Hou, M. Ji, and Y. Zhang. 2018. PSMB8 regulates glioma cell migration, proliferation, and apoptosis through modulating ERK1/2 and PI3K/AKT signaling pathways. Biomed. Pharmacother. 100: 205–212. 713. de Verteuil, D., T. L. Muratore-Schroeder, D. P. Granados, M.-H. Fortier, M.-P. Hardy, A. Bramoullé, E. Caron, K. Vincent, S. Mader, S. Lemieux, P. Thibault, and C. Perreault. 2010. Deletion of immunoproteasome subunits imprints on the transcriptome and has a broad impact on peptides presented by major histocompatibility complex I molecules. Mol. Cell. Proteomics 9: 2034–47. 714. Liu, R., P. Zhang, C. Yang, Y. Pang, M. Zhang, N. Zhang, and L. Yue. 2017. ONX-0914, a selective inhibitor of immunoproteasome, ameliorates experimental autoimmune myasthenia gravis by modulating humoral response. J. Neuroimmunol. 311: 71–78. 715. Liu, H., C. Wan, Y. Ding, R. Han, Y. He, J. Xiao, and J. Hao. 2017. PR-957, a selective inhibitor of immunoproteasome subunit low-MW polypeptide 7, attenuates experimental autoimmune neuritis by suppressing Th17-cell differentiation and regulating cytokine production. FASEB J. 31: 1756–1766. 716. Li, C., P. J. R. Ebert, and Q.-J. Li. 2013. T cell receptor (TCR) and transforming growth factor beta (TGF-β) signaling converge on DNA (cytosine-5)-methyltransferase to control forkhead box protein 3 (foxp3) locus methylation and inducible regulatory T cell differentiation. J. Biol. Chem. 3. 717. Gabryšová, L., J. R. Christensen, X. Wu, A. Kissenpfennig, B. Malissen, and A. O’Garra. 2011. Integrated T-cell receptor and costimulatory signals determine TGF-β-dependent differentiation and maintenance of Foxp3+ regulatory T cells. Eur. J. Immunol. 41: 1242–8. 718. Tubo, N. J., and M. K. Jenkins. 2014. TCR signal quantity and quality in CD4+T cell differentiation. Trends Immunol. 35: 591–596. 719. Stoeckle, C., P. Quecke, T. Rückrich, T. Burster, M. Reich, E. Weber, H. Kalbacher, C. Driessen, A. Melms, and E. Tolosa. 2012. Cathepsin S dominates autoantigen processing in human thymic dendritic cells. J. Autoimmun. 38: 332–343. 720. Millet, V., P. Naquet, and R. R. Guinamard. 2008. Intercellular MHC transfer between thymic epithelial and dendritic cells. Eur. J. Immunol. 38: 1257–1263. 721. Gallegos, A. M., and M. J. Bevan. 2004. Central Tolerance to Tissue-specific Antigens Mediated by Direct and Indirect Antigen Presentation. J. Exp. Med. 200: 1039–1049. 722. Singh, K., P. Deshpande, S. Pryshchep, I. Colmegna, V. Liarski, C. M. Weyand, and J. J. Goronzy. 2009. ERK-Dependent T Cell Receptor Threshold Calibration in Rheumatoid Arthritis. J. Immunol. 183: 8258–8267. 723. Li, C., E. Capan, Y. Zhao, J. Zhao, D. Stolz, S. C. Watkins, S. Jin, and B. Lu. 2006. Autophagy Is Induced in CD4+ T Cells and Important for the Growth Factor-Withdrawal Cell Death. J. Immunol. 177: 5163–5168. 724. Zhang, H., Y. Chi, K. Gao, X. Zhang, and J. Yao. 2015. P53 protein-mediated Up-regulation of MAP kinase phosphatase 3 (MKP-3) contributes to the establishment of the cellular senescent phenotype through dephosphorylation of extracellular signal-regulated kinase 1/2 (ERK1/2). J. Biol. Chem. 290: 1129–1140. 725. Watanabe, M., K. D. Moon, M. S. Vacchio, K. S. Hathcock, and R. J. Hodes. 2014. Downmodulation of tumor suppressor p53 by T cell receptor signaling is critical for antigen-specific CD4(+) T cell responses. Immunity 40: 681–91. 726. Fray, M. a, and S. C. Bunnell. 2014. P53 Keeps Bystanders At the Gates. Immunity 40: 633–5. 727. Su, X., J. A. Ditlev, E. Hui, W. Xing, S. Banjade, J. Okrut, D. S. King, J. Taunton, M. K. Rosen, and R. D. Vale. 2016. Phase separation of signaling molecules promotes T cell receptor signal transduction. Science . 352: 595–599. 728. Wu, Z., P. Jiao, X. Huang, B. Feng, Y. Feng, S. Yang, P. Hwang, J. Du, Y. Nie, G. Xiao, and H. Xu. 2010. MAPK phosphatase-3 promotes hepatic gluconeogenesis through dephosphorylation of forkhead box O1 in mice. J. Clin. Invest. 120: 3901–3911. 729. Hsu, S.-F., Y.-B. Lee, Y.-C. Lee, A.-L. Chung, M. K. Apaya, L.-F. Shyur, C.-F. Cheng, F.-M. Ho, and T.-C. Meng. 2018. Dual-specificity Phosphatase DUSP6 Promotes Endothelial Inflammation through Inducible Expression of ICAM-1. FEBS J. . 730. Bennett, A. M. 2018. DUSPs, twists and turns in the Journey to Vascular Inflammation. FEBS J. 285: 1589–1592. 731. Rouette, A., A. Trofimov, D. Haberl, G. Boucher, V. P. Lavallée, G. D’Angelo, J. Hébert, G. Sauvageau, S. Lemieux, and C. Perreault. 2016. Expression of immunoproteasome genes is regulated by cell-intrinsic and -extrinsic factors in human cancers. Sci. Rep. 6: 1–14. 732. Hu, H., H. Wang, Y. Xiao, J. Jin, J.-H. Chang, Q. Zou, X. Xie, X. Cheng, and S.-C. Sun. 2016. Otud7b facilitates T cell activation and inflammatory responses by regulating Zap70 ubiquitination. J. Exp. Med. 213: 399–414. 733. Tanaka, T., M. A. Soriano, and M. J. Grusby. 2005. SLIM is a nuclear ubiquitin E3 ligase that negatively regulates STAT signaling. Immunity 22: 729–736. 734. Yuan, C., J. Qi, X. Zhao, and C. Gao. 2012. Smurf1 protein negatively regulates interferon-γ signaling through promoting STAT1 protein ubiquitination and degradation. J. Biol. Chem. 287: 17006–17015. 735. Hoefig, K. P., and V. Heissmeyer. 2018. Posttranscriptional regulation of T helper cell fate decisions. J. Cell Biol. jcb.201708075. 736. Jeker, L. T., and J. a Bluestone. 2013. MicroRNA regulation of T-cell differentiation and function. Immunol. Rev. 253: 65–81. 737. Kim, W., E. J. Bennett, E. L. Huttlin, A. Guo, J. Li, A. Possemato, M. E. Sowa, R. Rad, J. Rush, M. J. Comb, J. W. Harper, and S. P. Gygi. 2011. Systematic and quantitative assessment of the ubiquitin-modified proteome. Mol. Cell 44: 325–340. 738. Mundt, S. 2015. Targeting the Immunoproteasome in Health and Disease. PhD thesis, University of Konstanz, Konstanz, Germany . 739. Kelemen, B. R., K. Hsiao, and S. a Goueli. 2002. Selective in vivo inhibition of mitogen-activated protein kinase activation using cell- permeable peptides. J. Biol. Chem. 277: 8741–8. 740. Takada, Y., S. Singh, and B. B. Aggarwal. 2004. Identification of a p65 peptide that selectively inhibits NF-kappa B activation induced by various inflammatory stimuli and its role in down-regulation of NF-kappaB-mediated gene expression and up-regulation of apoptosis. J. Biol. Chem. 279: 15096–104. 741. Brenner, D., A. Golks, F. Kiefer, P. H. Krammer, and R. Arnold. 2005. Activation or suppression of NFkappaB by HPK1 determines 163

References sensitivity to activation-induced cell death. EMBO J. 24: 4279–90. 742. Lam, K., R. Ku, and K. Rajewsky. 1997. In Vivo Ablation of Surface Immunoglobulin on Mature B Cells by Inducible Gene Targeting Results in Rapid Cell Death. Cell 90: 1073–1083. 743. Mu, J., X. Tai, S. S. Iyer, J. D. Weissman, A. Singer, and D. S. Singer. 2014. Regulation of MHC Class I Expression by Foxp3 and Its Effect on Regulatory T Cell Function. J. Immunol. 192: 2892–2903. 744. Howcroft, T. K., J. D. Weissman, A. Gegonne, and D. S. Singer. 2005. A T lymphocyte-specific transcription complex containing RUNX1 activates MHC class I expression. J. Immunol. 174: 2106–2115. 745. Grimm, T. 2014. Influence of Immunoproteasome Inhibition and Deficiency on JAK / STAT Signaling in T cells and Macrophages. Bachelor's thesis, University of Konstanz, Konstanz, Germany. 746. Lee, C. S., C. Lee, T. Hu, J. M. Nguyen, J. Zhang, M. V. Martin, M. P. Vawter, E. J. Huang, and J. Y. Chan. 2011. Loss of nuclear factor E2- related factor 1 in the brain leads to dysregulation of proteasome gene expression and neurodegeneration. Proc. Natl. Acad. Sci. 108: 8408–8413. 747. Moskowitz, D. M., D. W. Zhang, B. Hu, S. Le Saux, R. E. Yanes, Z. Ye, J. D. Buenrostro, C. M. Weyand, W. J. Greenleaf, I. T. Program, V. Affairs, P. Alto, H. Care, and P. Alto. 2018. Epigenomics of human CD8 T cell differentiation and aging. Sci. Immunol. 2: 1–29. 748. Lee, J. A., H. J. Son, J. W. Choi, J. Kim, S. H. Han, N. Shin, J. H. Kim, S. J. Kim, J. Y. Heo, D. J. Kim, K. D. Park, and O. Hwang. 2018. Activation of the Nrf2 signaling pathway and neuroprotection of nigral dopaminergic neurons by a novel synthetic compound KMS99220. Neurochem. Int. 112: 96–107. 749. Akiyama, K., S. Kagawa, T. Tamura, N. Shimbara, M. Takashina, P. Kristensen, K. B. Hendil, K. Tanaka, and A. Ichihara. 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: 85–8. 750. Gaczynska, M., A. L. Goldberg, K. Tanaka, K. B. Hendil, and K. L. Rock. 1996. Proteasome subunits X and Y alter peptidase activities in opposite ways to the interferon-gamma-induced subunits LMP2 and LMP7. J. Biol. Chem. 271: 17275–17280. 751. Kidger, A. M., and S. M. Keyse. 2016. The regulation of oncogenic Ras/ERK signalling by dual-specificity mitogen activated protein kinase phosphatases (MKPs). Semin. Cell Dev. Biol. 50: 125–132. 752. Kumabe, S., M. Itsumi, H. Yamada, T. Yajima, T. Matsuguchi, and Y. Yoshikai. 2010. Dual specificity phosphatase16 is a negative regulator of c-Jun NH2-terminal kinase activity in T cells. Microbiol. Immunol. 54: 105–111.

164

Appendix

12 APPENDIX

12.1 Appendix Figures

Appendix Figure 1 (related to Figure 11): In silico and experimental control of Psmb5 primer pairs. A) Primer-BLAST search for the indicated primer pair (CS_mmu_beta5c), see Materials and Methods) to predict amplicon sizes and off- target amplicons (first two hits) showing the target amplicon and a predicted-sequence potential off-target at above 1 kbp (not detectable). B) Melting curve assessment after qPCR with duplicate samples from CD4+, CD19+ and liver homogenate samples of WT and LMP7- deficient mice. Two samples contain water-only controls (purple and yellow, no amplicons) C) Agarose-gel assessment of produced amplicons after qPCR. One example from each cell type origin is shown. SmartLadder SF was used as DNA marker.

165

Appendix

Appendix Figure 2 (related to Figure 17): Control stainings for intracellular p-ERK detection using flow cytometry. Naïve CD4+ T cells purified from WT mouse spleen were treated with either 0.3 % DMSO or with 300 nM ONX 0914 (not shown) for 2 h before activation with plate bound antibodies against CD3/CD28. Shown are plots for gating strategy from one experiment out of the pooled experiments in Figure 17D. The histogram below shows the measurement on CD4+ cells after staining against intracellular p-ERK1/2 (blue histograms) and the control staining for stimulated and unstimulated cells using the secondary antibody only.

Appendix Figure 3 (related to Figure 17): Kinetic of reduced ERK phosphorylation in activated naïve CD4+ T cells after ONX 0914 treatment Purified naïve CD4+ T cells from WT mice (upper panel) or LMP7-/- mice (lower panel) were pulse-treated for 2 h with 0.3 % DMSO or 300 nM ONX 0914, followed by activation with plate-bound antibodies against CD3/CD28 for indicated time periods. Intracellular staining against p- ERK1/2 was performed and cells were analyzed by flow cytometry. One example of two independent experiments with similar outcome. 1661

Appendix

Appendix Figure 4: Characterization of autonomous STAT3-phosphorylation in mouse and human T cells activated in vitro. A-C) Three examples of out of more than four independent experiments using murine CD4+ T cells pre-treated for 3 h with 0.3% DMSO (D) or 300 nM ONX 0914 (X) (or additionally with 10µM continuous MG-132, MG) and activated with plate bound anti-CD3/CD28 antibodies for indicated time periods before lysis and subjection to immunoblotting against p-STAT3(Tyr705). Total STAT3 and/or γ-Tubulin were used as control. D) One representative example out of three independent experiments using MACS-enriched human CD4+ T cells pre-treated as in A-C and activated with anti-CD3/CD28/CD2 coated beads for indicated time-periods.

Appendix Figure 5: Influence of ONX 0914 and MG-132 treatment on regulation of dual specificity phosphatases in expanded T cells and T1 cells A) Expanded CD4+ T cells were re-activated with plate-bound anti-CD3/CD28 antibodies after 2 h pre-treatment with 0.3% DMSO (D), 300 nM ONX 0914 (X) or 10 µM MG-132 (MG) continuously. Immunoblots against indicated proteins were performed after indicated time periods. One example of three or more independent reproductions is shown. B) Experiment as in A, but with IFN-γ pre-treated T1 cells. One example of two independent experiments is shown. Note that the selected immunoblots shown here were generated by Thilo Berger and also presented as part of his master’s thesis [707], which was supervised by the author of this work.

167

Appendix

Appendix Figure 6: No evidence for elevated steady-state proteostasis stress in LMP7-deficient primary cells compared to WT cells. A) MACS-enriched CD4+ T cells, CD19+ B cells and bulk liver homogenates from either WT or LMP7-deficient mice were used to assess integrated stress-response markers at steady state by immunoblotting. Indicated markers were addressed with γ-Tubulin used as a loading control. One example out of three independent experiments with similar outcome. B) Experiment as in A, but with an additional control sample from 6 h ONX 0914 treated WT B cells. One example out of three independent experiments with similar outcome is shown.

12.2 Abbreviations

AD Alzheimer's Disease AICD activation-induced cell death Aire Autoimmune regulator Akt Ak mouse thymoma oncogene ALS amyotrophic lateral sclerosis ANA antinuclear antibody ANOVA analysis of variance AP-1 activator protein 1 APC antigen presenting cell APS ammonium peroxosulfate ARE antioxidant response element ATF4 activating transcription factor 4 ATP adenosine triphosphate BATF basic leucine zipper ATF like transcription factor BAX Bcl-2 associated X protein Bcl-2 B cell lymphoma 2 BCR B cell receptor Blm10 bleomycin-sensitive 10 BM Bone marrow BSA bovine serum albumin CAAX C-terminal membrane targeting motif CANDLE Chronic Atypical Neutrophilic Dermatosis with Lipodystrophy and Elevated temperature CARMA-1 caspase recruitment domain containing MAGUK protein 1 CD Cluster of Differentiation CFSE 5(6)-Carboxyfluorescein diacetate N-succinimidyl ester CHOP C/EBP homologous protein CHX cycloheximide CIA collagen induced arthritis CP Core Particle CRAC calcium-release activated calcium channel cTEC cortical thymic epithelial cell CTL cytotoxic T lymphocyte CTLA4 cytotoxic T lymphocyte antigen 4 CTLL2 cytotoxic lymphoma line 2 CVB3 Cocksackie virus B3 CYLD cylindromatosis deubiquinating enzyme DAG diacylglycerol 168

Appendix

DAMP danger associated molecular pattern DC dendritic cell DDI2 DNA damage inducible 1 homologue 2 DMEM Dulbeccos Modified Eagle's medium DMSO dimethylsulfoxide dn dominant negative DNA desoxyribonucleic acid DRiP defective ribosomal particle DSS dextran sulfate sodium DUB deubiquitinating enzyme DUSP dual specificity phosphatase EAE experimental autoimmune encephalomyelitis ECL enhanced chemiluminescene EGFR epidermal growth factor receptor eIF eukaryotic initiation factor ELISA enzyme linked immunosorbent assay ER endoplasmic reticulum ERAD ER-associated degradation ERK extracellular signal-regulated kinase FACS fluorescence activated cell sorting FBS fetal bovine serum FMO Fluorescene minus one Foxp3 forkhead box protein 3 GATA3 GATA bining protein 3 GLK Germinal center kinase-like kinase Grb2 Growth factor receptor-bound protein 2 GTP Guanosine triphosphate HbYX hydrophobic-tyrosine-X HECT Homologous to E6AP carboxyl terminus IBD inflammatory bowel disease ICAM intracellular adhesion molecule IFN interferon IκB inhibitor of kappa B IKK inhibitor of kappa B kinase IL interleukin ILC innate lymphoid cell IP3 Inositol-triphosphate IOTA macropain subunit iota, alternative name for PSMA6 IRE1 inositol-requiring enzyme 1 IRF interferon regulated factor ITAM immunoreceptor tyrosine based activation motif ITK IL-2-inducible tyrosine kinase i.v. intravenously JNK c-Jun N-terminal kinase LC-MS Liquid chromatography – mass spectrometry Lck lymphocyte specific kinase LMP low molecular mass polypeptide LN lymph node LPS lipopolysaccharide MACS magnetic cell sorting MAPK mitogen activated protein kinase MECL multicatalyitc endopeptidase complex like MEF mouse embryonic fibroblast MEK MAPK ERK kinase MHC major histocompatibility complex MKK MAPK kinase kinase MM Multiple myeloma mTEC medullary thymic epithelial cell

169

Appendix mTORC mechanistic target of rapamycin complex MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide Nck non-catalytic region of tyrosine kinase adaptor protein NEMO NF-κB essential modifier MS mass spectrometry NFAT nuclear factor of activated T cells NF-κB nuclear factor kappa B NGLY N-glycanase NIK NF-κB inducing kinase NK natural killer (cell) Nox NADPH oxidase Nrf1 Nuclear factor erythroid-derived 2 related factor 1 Nrf2 Nuclear factor erythroid-derived 2 related factor 1 PA28 proteasome activator of apparent molecular weight 28 kDa PA700 proteasome activator of 700 kDa PAC proteasome assembly complex PAGE polyacrylamid gelectrophoresis PAK1 p21 activated kinase PAMP pathogen associated molecular pattern PARP poly-ADP-ribose polymerase PCR polymerase chain reaction pDC Parkinson's Disease pDC plasmacytoid dendritic cell PERK RNA-like endoplasmic reticulum kinase PFA paraformaldehyde PI3K phosphoinositide-3-kinase PIP2 phosphatidylinositol-4,5-bisphosphate PKC protein kinase C PKN2 protein kinase N2 PLC phospholipase C PMA phorbol-12-myrisate-13-acetate POMP proteasome maturation protein PRAS40 proline-rich Akt subtrate of 40 kDa PSMB proteasome subunit beta-type PTX-3 pentaxin-3 RA rheumatoid arthritis RAG recombination activating gene RasGRP Ras guanine nucleotide releasing protein RasGEF Ras guanine nucleotide exchange factor RING Really interesting new gene RNA ribonucleic acid RORγt RAR-related orphan receptor gamma t ROS reactive oxygen species RP regulatory particle Rpn regulatory particle non-ATPase Rpt regulatory particle AAA-ATPase RUNX1 Runt-related transcription factor 1 SAPK stress activated protein kinase s.c. subcutaneously SDS sodium dodecyl sulfate SILAC stable isotope labeling of amino acids in cell culture SLE systemic lupus erythematosus SLIM STAT-interacting LIM protein SLP76 SH2-domain containing leukocyte protein of 76kDa SMAC central supramolecular activation cluster SMARTA mouse strain description SOCE store operated calcium entry SOS son of sevenless

170

Appendix

SREBP2 sterol-regulatory element-binding protein 2 STAT signal transducer and activator of transcription TAB1 TGF-beta activated kinase 1 binding protein 1 TAILS terminal amine isotope labeling of substrates T-ALL T cell acute lymphoid leukemia TAP transporter associated with antigen presentation TBS tris-buffered saline TCA tricarboxylic acid TCR T cell receptor TGF transforming growth factor TJP tight junction protein TLR toll-like receptor TNF tumor necrosis factor TRAF TNF-receptor associated factor TSC tuberous sclerosis complex UBA ubiquitin associated domain UBL ubiquitin-like domain UPR unfolded protein response VCP valosin-containing protein WT Wild-type XBP1 X-Box binding protein ZAP70 zeta-chain associated protein of 70 kDa

12.3 Table of Figures

Figure 1: Structure of the 20S Core Particle and its peptidolytic activity 16 Figure 2: Simplified schematic overview of selected cellular stress response pathways. 28 Figure 3: Proteasome and immunoproteasome inhibitors and their molecular inhibition modes. 31 Figure 4: Schematic overview of T helper cell polarization 41 Figure 5: Simplified schematic overview of proximal TCR signaling events and CD28 co-stimulation 48 (exemplified for a CD4+ T cell). Figure 6: Schematic overview of canonical and non-canonical NF-κB signaling and the involvement of 49 proteasomes exemplified for T cells Figure 7: Ca2+-NFAT signaling in T cell activation 51 Figure 8: Schematic overview of MAPK pathways in TCR-induced signaling 55 Figure 9: ONX 0914 treatment leads to attenuated early T cell activation in a synergistically LMP7 and LMP2- 86 dependent manner. Figure 10: Naïve T cells and B cells contain predominantly mixed and immunoproteasomes, but not standard 87 proteasomes Figure 11: No compensatory up-regulation of Psmb5 and Psmb6 at mRNA level in LMP7-deficient lymphocytes 89 at steady state. Figure 12: Impaired proliferation in the presence of ONX 0914 is not caused by IL-2 deprivation alone. 90 Figure 13: Impaired polarization of CD4+ T cells in the presence of ONX 0914. 91 Figure 14: Reduced expression of CD69 and IL2 mRNA transcripts after ONX 0914 treatment in primary 93 activated CD4+ T cells. Figure 15: Expanded T cells as experimental system for analysis of signaling pathways 94 Figure 16: Analysis of canonical T cell activation signaling pathways in expanded and naïve CD4+ T cells. 95 Figure 17: ONX 0914 treatment reduces ERK signaling sustainment in expanded and naïve murine T cells and in 97 primary human T cells. Figure 18: ONX 0914 treatment induces mild proteostasis stress in activated CD4+ T cells without inducing 99 apoptosis Figure 19: T cells up-regulate β5c containing standard proteasomes during activation and in response to 102 immunoproteasome inhibition, likely via Nrf1 Figure 20: Effects of ONX 0914 on primary murine and human B cell activation, proteostasis and apoptosis 104 induction Figure 21: Induction of high immunoproteasome content renders mouse embryonic fibroblasts susceptible to 105 ubiquitin-conjugate formation after ONX 0914 treatment. Figure 22: Reduced ERK phosphorylation and accumulation of ubiquitin conjugates correlate with accumulation 107 of DUSP6 Figure 23: Radioactive labeling in T1 cells shows that ONX 0914 impairs DUSP6 degradation. 109

171

Appendix

Figure 24: No evidence for a functional involvement of DUSP6 in the observed effects of ONX 0914 on T cell 111 activation.

Figure 25: Impaired activation of antigen-specific CD4+ T cells in vivo. 113

Appendix Figure 1: In silico and experimental control of Psmb5 primer pairs. 165 Appendix Figure 2 (related to Figure 16): Control stainings for intracellular p-ERK detection using flow 166 cytometry. Appendix Figure 3 (related to Figure 15): Kinetic of reduced ERK phosphorylation in activated naïve CD4+ T 166 cells after ONX 0914 treatment. Appendix Figure 4: Characterization of autonomous STAT3-phosphorylation in mouse and human T cells 167 activated in vitro. Appendix Figure 5: Influence of ONX 0914 and MG-132 treatment on regulation of dual specificity phosphatases 167 in expanded T cells and T1 cells. Appendix Figure 6: No evidence for elevated steady-state proteostasis stress in LMP7-deficient primary cells 168 compared to WT cells.

12.4 Record of Contribution

All experimental results shown in this work were obtained by the author of this work, but contributions to the obtained results were made as follows: One of three independent experiments in Figure 19F was performed by Thilo Berger during a short-term employment at the Chair of Immunology. Appendix Figure 5 was performed by Thilo Berger during the course of his master’s thesis. The isolation of mouse embryonic fibroblasts in Figure 21 was performed by Annegret Bitzer. The technical injection of ONX 0914 or vehicle s.c. and i.v.-infection with LCMV presented in Figure 25 were performed by Michael Basler. LCMV-WE propagation as well as collection of recombinant IL-2 were performed by Ulrike Beck. Flow cytometry was partly assisted by staff members of the FlowKon Core Facility. Work involving radioactive material was supervised and assisted by Gunter Schmidtke. Mouse breeding and regular housing was performed by staff members of the Tierforschungsanlage (TFA) Konstanz.

172

Appendix

12.5 List of publications and oral presentations

Publications and unpublished manuscripts related to this thesis: Christian Schmidt, Thilo Berger, Marcus Groettrup and Michael Basler, 2018. Immunoproteasome inhibition impairs T and B cell activation by restraining ERK signaling and proteostasis. Manuscript status at the time of thesis-submission: unpublished, manuscript submitted to Frontiers in Immunology, in revision after independent review

Michael Basler, Michelle M. Lindstrom, Jacob J. LaStant, J. Michael Bradshaw, Timothy D. Owens, Christian Schmidt, Elmer Maurits, Christopher Tsu, Herman S. Overkleeft, Christopher J. Kirk, Claire L. Langrish & Marcus Groettrup, 2018. Co-inhibiting immunoproteasome subunits LMP2 and LMP7 is required to block autoimmunity.

Manuscript status at the time of thesis-submission: Manuscript accepted for publication in EMBO Reports, (in the press)

Michael Basler, Sarah Mundt, Annegret Bitzer, Christian Schmidt and Marcus Groettrup, 2015. The immunoproteasome: a novel drug target for autoimmune diseases. Clinical and Experimental Rheumatology 33, 74-9

Oral presentation related to this thesis: “Immunoproteasome function in T cell activation and autoimmunity” International Symposium Two Days of Proteostasis of the collaborative research center CRC969 (Sonderforschungsbereich, SFB 969), October 13th and 14th, 2016 in Konstanz

Publications not related to this thesis: René Hägerling, Cathrin Pollmann, Martin Andreas, Christian Schmidt, Harri Nurmi, Ralf H. Adams, Kari Alitalo, Volker Andresen, Stefan Schulte-Merker and Friedemann Kiefer, 2013. A novel multistep mechanism for initial lymphangiogenesis in mouse embryos based on ultramicroscopy, EMBO Journal 32, 629-644

173

Appendix

Acknowledgements – Danksagung

An erster Stelle dieser Danksagung ist es nicht nur angebracht, sondern mir auch ein persönliches Anliegen, Herrn PD Dr. Michael Basler meine tiefe Dankbarkeit auszusprechen. Die Bereitschaft, mich als ersten Doktoranden nach der Habilitation anzunehmen, ist von Beginn an etwas Besonderes für mich gewesen. Sich in die Signaltransduktionsprozesse primärer T Zellen zu stürzen war nicht nur im Hinblick auf die experimentellen Erfordernisse ein gutes Stück weit Neuland für uns, sondern benötigte auch viel Betreuer-Vertrauen in die erarbeitete Expertise des Doktoranden. Ohne die zahlreichen fachlichen Diskussionen, den Austausch über die Planung neuer Experimente, den persönlichen Rückhalt in schweren Zeiten und den Charme und Witz bei uns im Labor, hätte ich diese Arbeit niemals erstellen können. Ich durfte stets fortwährende Unterstützung erfahren und ich weiß, dass es nicht immer leicht mit mir war. Danke für alles, Michi! Ebenso gilt mein Dank Herrn Prof. Dr. Marcus Groettrup. Die berufliche und persönliche Unterstützung, die fachliche Unterstützung im Projekt von Beginn bis Ende, die Anerkennung und auch das Aushalten meines manchmal penetranten Gegenargumentierens sind ebenfalls nicht selbstverständlich. Ich freue mich, dass sich aus unseren wissenschaftlichen „Kabbeleien“ fruchtbare Entwicklungen für die wissenschaftliche Arbeit ergeben haben und möchte mich ganz herzlich bedanken. Positive Rückmeldungen und natürlich auch die Laborausflüge und der Retreat am Feldberg haben zur produktiven und freundschaftlichen Arbeitsatmosphäre sehr beigetragen. Danke, Marcus! Prof. Dr. Martin Scheffner und Prof. Dr. Wolfgang Schamel möchte ich für die Bereitschaft zur Begutachtung der Dissertation und für die Teilnahme in der Prüfungskommission für die mündliche Prüfung danken. Prof. Dr. Christof Hauck, Dr. Daniel Summerer und Dr. Thomas Böttcher haben mich im Thesis Committee der KoRS-CB betreut. Ihnen und der Graduiertenschule KoRS-CB möchte ich ebenfalls danken. Meine liebgewonnenen Teamkollegen von P11 haben besonderen Dank verdient. Julia Körner, Sarah Mundt und Annegret Bitzer haben sich mit mir durch die Tiefen der Immunproteasom-Forschung gekämpft und mich persönlich sehr unterstützt. Danke Euch! Anna, Tobias, Richi, Miki, Valerie, Nico, Jun, Faiz, Florian, Franzi H. und Franzi G., die BITg-Truppe und zahlreiche Student/inn/en haben die Atmosphäre auf P11 federführend gestaltet. Ich danke Euch! Danke an Annette, Ricarda und Stef für Eure Untersützung insbesondere in puncto Durchflusszytometrie. Martin Stöckl danke ich für die Betreuung am Mikroskop. Ein Extra-Dank geht an Annette für unsere Gespräche, die mir viel Mut gemacht haben. Danke an Gunter für die Zauberschule. Danke an Brigitte für Deinen Einsatz nicht nur im Sekretariat, sondern auch als IT-Managerin auf P11. Ulrike Beck und Heike Göbel im Besonderen und natürlich auch Gerardo und Tina haben wichtige Untersützung für meine Arbeit und auch persönlich geleistet. Danke! Die Arbeit in der Immunologie erfordert auch die Arbeit mit unseren Labormäusen. Ich habe Respekt davor, dass zahlreiche Mäuse ihr Leben für diese Forschung geben mussten – auch wenn sie eigens dafür gezüchtet wurden. Die Belastung für die Tiere so gering wie möglich zu halten und den Umgang mit den Tieren möglichst verantwortungsvoll und schonend zu gestalten, war mir wichtig. Hierin wurde ich sehr unterstützt von Dr. Gerald Mende, Dr. Margarethe Köberle, Dr. Dieter Schopper, Birgit Planitz, Andrea Westermann und Nicole Renner. Auch ihnen möchte ich danken. Tanja Grimm und Thilo Berger haben in ihrer Bachelorarbeit bzw. Masterarbeit tolle Arbeit geleistet und damit auch zu meinem PhD-Projekt viel beigetragen. Danke an Euch! Ich danke Franziska Biekert so sehr, dass ich es nicht in Worte fassen kann. Mit Dir ist alles schöner! Meine Freunde in Konstanz und überall haben mich so viel unterstüzt. Ich bin von Herzen dankbar. Danke an Patrick, Marit, Florian, Svenja, Ina, Janina, Nini, Bjarne, Joni, Shogo, Jannika, Alex, Anna, Stefan, Sinja, Christina, Vroni, Gerrit, Lukas, Patrick P., Wera, alle Pipetboys (Timo, Daniel, Philipp, Tim und Florian) und Swing in Konstanz. Der Schluss gebührt meiner Familie, meinen lieben Eltern und meinem Bruder. Ohne Euch wäre ich nichts. Ohne Euch hätte ich diesen Weg nicht gehen können. Ich danke Euch von Herzen.

174