The Role of PGAM5 in Regulating Viral Infection and the Pathogenesis of Intestinal Inflammation

The role of PGAM5 in regulating viral infection and the pathogenesis of intestinal inflammation

Der Naturwissenschaftlichen Fakultät

der Friedrich-Alexander-Universität Erlangen-Nürnberg

zur

Erlangung des Doktorgrades Dr. rer. nat.

vorgelegt von

Yuqiang Yu

Als Dissertation genehmigt von der Naturwissenschaftlichen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg

Tag der mündlichen Prüfung: 1st July 2021

Vorsitzender des Promotionsorgans: Prof. Dr. Wolfgang Achtziger

Gutachter/in: Prof. Dr. Falk Nimmerjahn

Prof. Dr. Christoph Becker

Table of contents

1 Abstract ...... 6 2 Zusammenfassung...... 7 3 Introduction ...... 9 3.1 PGAM5 ...... 9

Structure and expression ...... 9

PGAM5 acts as a regulator in cell death pathways...... 10

The role of PGAM5 in Wnt/β-catenin signaling pathway ...... 14

PGAM5 in the NLRP3 inflammasome ...... 14

PGAM5 activity is linked to multimerization ...... 15

3.2 Cellular defense against viruses ...... 16

Toll-like receptors ...... 16

RIG-I-like receptors ...... 19

MAVS is a key factor in the RLR pathway ...... 20

IFNs and antiviral responses ...... 23

Modeling RNA viral infection ...... 24

3.3 IBD ...... 25

3.4 Aims of the project ...... 26

4 Material and Methods ...... 27 4.1 Materials ...... 27

4.2 Animals ...... 31

4.3 Cell line ...... 31

4.4 Methods ...... 32

MEF isolation ...... 32

Generation of HeLa knock-out cells using CRISPR/Cas9 technology ...... 32

PGAM5 overexpression ...... 33

Cell stimulation and infection ...... 33

Genotyping ...... 33

Dextran sulfate sodium (DSS)-induced colitis ...... 35

Table of contents

Gene expression analysis ...... 35

Protein analysis ...... 36

Histological analysis ...... 40

Immunocytochemistry ...... 42

Statistical analysis ...... 43

5 Results ...... 44 5.1 PGAM5-MAVS interaction regulates TBK1/ IRF3 dependent antiviral responses 44

Intracellular poly(I:C) delivery induces the formation of PGAM5 multimers 44

PGAM5 deficiency attenuates intracellular poly(I:C)-induced IFNβ expression ...... 48

PGAM5 overexpression rescued IFNβ expression in PGAM5 deficient HeLa cells 52

PGAM5 functions upstream of TBK1 ...... 54

The formation of PGAM5 multimers and MAVS aggregates are independent of each other ...... 57

PGAM5 interacts with MAVS ...... 59

PGAM5 regulates VSV-induced IFNβ expression and inhibits VSV replication ...... 60

5.2 PGAM5 regulates IFN-stimulated genes (ISG) expression ...... 63

PGAM5 regulates extracellular poly(I:C)-induced STAT1 phosphorylation. 63

PGAM5 deficiency up-regulates extracellular poly(I:C)-induced SOCS3 expression...... 66

PGAM5 positively regulates IFN-stimulated STAT1 phosphorylation and downregulates SOCS3 expression...... 67

5.3 PGAM5 is dispensable for poly(I:C)-induced small intestinal inflammation...... 71

5.4 PGAM5 is dispensable for the development of experimental colitis...... 74

6 Discussion ...... 77 6.1 PGAM5 regulates IFNβ expression and antiviral response...... 77

Table of contents

6.2 The mechanisms of PGAM5-dependent antiviral pathways...... 79

6.3 PGAM5 is dispensable for multiple organ injury induced in vivo...... 83

6.4 Working model ...... 84

7 References ...... 85 8 List of abbreviations...... 100 9 List of Tables...... 103 10 List of figures ...... 104 11 Curriculum vitae...... Error! Bookmark not defined. 12 Conference Presentations ...... 106 13 Declaration ...... 107

Zusammenfassung

1 Abstract

Viral infections trigger host innate immune responses, characterized by the production of type-I interferons (IFN) including IFNβ. IFNβ induces cellular antiviral defense mechanisms by expressing IFN-stimulated genes (ISG), thereby contributing to pathogen clearance. Accumulating evidence suggests that mitochondria constitute a crucial platform for the induction of antiviral immunity and cell death. The mitochondrial protein phosphoglycerate mutase family member 5 (PGAM5) has been implicated in a broad range of biological processes including certain cell death pathways and NLRP3 inflammasome activation. The hypothesis of this thesis was that PGAM5 is involved in regulating cellular immune defense and cell death in the gut. Thus, this thesis aimed to investigate functional roles of PGAM5 via in vitro and in vivo models.

Initially, poly(I:C) was used to mimic RNA virus infection in HeLa cells and the presence of intracellular RNA leads to PGAM5 multimer formation and co-localization at aggregated mitochondria. Furthermore, this thesis showed that PGAM5 deficiency specifically attenuated IFNβ expression induced by intracellular poly(I:C) but not when poly(I:C) was added into the medium. Decreased phosphorylation levels of IRF3 and TBK1 in PGAM5 deficient cells further confirmed these finding. On the molecular level, a direct interaction of PGAM5 with the mitochondrial antiviral-signaling protein (MAVS) was demonstrated. Finally, this thesis verified the functional role of PGAM5 in the process of viral infection. PGAM5 deficient cells, upon infection with vesicular stomatitis virus (VSV), revealed diminished Ifnβ expression and increased VSV replication.

In addition, this thesis also demonstrated that PGAM5 is important for regulating ISG responses. PGAM5 deficient cells exhibited decreased phosphorylation levels of STAT1 and expression of ISG when challenged with three different types of IFNs. Mechanistically, PGAM5 deficiency significantly up-regulated SOCS3 expression.

These in vitro data supported the initial hypothesis about the functional role of PGAM5 in cellular immune defense. However, the hypothesis about cell death in gut could not be proved in vivo. PGAM5 is dispensable for poly(I:C)-induced small intestinal inflammation and DSS-induced colitis.

Collectively, this thesis identified PGAM5 as an important regulator in antiviral responses.

Zusammenfassung

2 Zusammenfassung

Virusinfektionen lösen eine angeborene Immunantwort des Wirts aus, die durch die Produktion von Typ-I-Interferonen (IFN), einschließlich IFNβ, gekennzeichnet ist. IFNβ induziert zelluläre antivirale Abwehrmechanismen durch die Produktion von IFN- stimulierten Genen (ISG) und trägt damit zur Krankheitsabwehr bei. Viele Hinweise legen nahe, dass Mitochondrien eine entscheidende Rolle während der Induktion der antiviralen Immunität und des Zelltods darstellen spielen. Das mitochondriale Protein Phosphoglyceratmutase 5 (PGAM5) spielt in einer Vielzahl biologischer Prozesse eine Rolle und reguliert z.B. den Vorgang spezieller Apoptosevorgänge und die Aktivierung des NLRP3-Inflammasoms. Daher war es das Ziel dieser Arbeit, den funktionellen Mechanismus von PGAM5, insbesondere während antiviraler Reaktionen, zu untersuchen.

Zunächst wurde Poly(I:C) dazu verwendet, um eine RNA-Virus Infektion in HeLa-Zellen zu simulieren. Die Anwesenheit intrazellulärer RNA führte zu einer Multimerisierung von PGAM5 sowie zu einer Kolokalisation dieser Multimere an aggregierten Mitochondrien. Darüber hinaus konnte in dieser Arbeit gezeigt werden, dass eine PGAM5-defiziente Zelllinie die Expression von IFNß spezifisch vermindert, jedoch nicht, wenn Poly(I:C) mit in das Kulturmedium gegeben wurde. Dies konnte durch den Nachweis der verminderten Phosphorylierung von IRF3 und TBK1 in PGAM5-defizienten Zelllinien bestätigt werden. Auf molekularer Ebene konnte eine direkte Interaktion von PGAM5 mit dem mitochondrial antiviral-signaling protein (MAVS) nachgewiesen werden. Schliesslich konnte in dieser Arbeit die funktionelle Rolle von PGAM5 im Prozess der Virusinfektion gezeigt werden, da PGAM5-defiziente Zellen nach der Infektion mit dem vesikulären Stomatitis-Virus (VSV) eine verminderte Ifnβ Expression und eine erhöhte VSV-Replikation zeigten.

Darüber hinaus konnte in dieser Arbeit nachgewiesen werden, dass PGAM5 wichtig für die Regulierung von ISG-Antworten ist. PGAM5-defiziente Zellen wiesen verminderte Phosphorylierung von STAT1 und eine geringere Expression von ISG auf, wenn sie mit drei verschiedenen Typen von Interferonen stimuliert wurden. Mechanistisch gesehen führte die PGAM5-Defizienz zu einer signifikanten Erhöhung der SOCS3-Expression.

Diese in vitro-Daten unterstützen die ursprüngliche Hypothese über die funktionelle Rolle von PGAM5 in der zellulären Immunabwehr. Die Hypothese über den Zelltod im Darm konnte jedoch in vivo nicht bewiesen werden. PGAM5 ist bei Poly(I:C)-induzierter Dünndarmentzündung und DSS-induzierter Kolitis entbehrlich.

Zusammenfassung

Zusammenfassend konnte in dieser Arbeit PGAM5 als wichtiger Regulator bei antiviralen Reaktionen identifiziert werden.

Introduction

3 Introduction

The innate immune system constitutes the first line of defense against an invasion of pathogens (Kawai and Akira, 2008). Several evolutionarily conserved pattern recognition receptors (PRRs) have been linked to the innate immune response, including Toll-like receptors (TLRs), RIG-I-like receptors (RLRs), Nod-like receptors (NLRs) and cytosolic DNA sensors. Numerous receptors interact with their respective ligands triggering various pathways to activate innate immune responses. RLRs, such as RIG-I and melanoma differentiation associated protein 5 (MDA5), are major cytoplasmic viral RNA sensors (Wu and Chen, 2014). The activation of RLRs further leads to the induction of type I interferons (IFN) and pro- inflammatory cytokines (Chiang et al., 2014). Subsequently, the secreted type I IFNs bind to their respective receptors and activate the signal transducers and activators of transcription (STATs) signaling pathway. The activation of STATs in turn induces the expression of IFN- stimulated genes (ISG) to protect cells from virus infection (Stark and Darnell, 2012).

The mammalian gastrointestinal tract is continuously exposed to numerous pathogens as well as food-derived antigens and environmental toxins, therefore the gastrointestinal tract is highly sensitive for pathogenic changes, such inflammatory bowel disease (IBD). IBD is a complex multifactorial disease that mainly refers to ulcerative colitis (UC) and Crohn’s disease (CD) (Sartor, 2008; Sartor and Muehlbauer, 2007; Xavier and Podolsky, 2007). Despite considerable efforts in the field of IBD research, the pathophysiology of UC and CD are still not entirely elucidated. Recent data showed that multiple pathways related to cell death and inflammation are involved in the pathogenesis and disease course of IBD (Schwarzer et al., 2020) (Rathinam and Chan, 2018).

Despite the fact that the mitochondrial protein PGAM5 has been identified as an important player in regulating multiple cell death pathways and inflammation, it is currently unknown whether PGAM5 regulates host defense in the steady state or in patients with IBD. This thesis is aimed to identify the functional contribution of PGAM5 in these processes.

3.1 PGAM5

Structure and expression

PGAM5 is a mitochondrial membrane protein that belongs to the PGAM family. The PGAM family contains a PGAM catalytic domain, which shares homologies to several metabolic

Introduction

enzymes. As consequence, some PGAM members, like PGAM1 and PGAM2, are well- characterized metabolic enzymes that contain mutase activity on small molecule substrates such as fructose-2,6-bisphosphate (PFKPB1–4) or 3-phosphoglycerate (Sadatomi et al., 2013). However, other members of this family, including PGAM5, lack a typical phosphotransferase or phosphohydrolase activity and instead show dephosphorylation activity specifically on Ser/Thr and, potentially, histidine residues (Panda et al., 2016; Takeda et al., 2009). Therefore, PGAM5 is characterized as an atypical mitochondrial Ser/Thr phosphatase, alongside suppressor of T cell receptor signaling 1 and 2 (STS1 and STS2) (Lo and Hannink, 2006; Sadatomi et al., 2013). Several PGAM5 targets have been identified, such as DRP1 (mediates mitochondrial fission), FUNDC1 (mediates mitophagy), ASK1 (mediates oxidative stress signaling), and BCL-XL (regulates cell survival) (Chen et al., 2014; Takeda et al., 2009; Wang et al., 2012; Wilkins et al., 2014).

In humans, the PGAM5 gene is located on chromosome 12 and is expressed in nearly all cells and tissues. It has two forms: a long form (PGAM5L) and a short form (PGAM5S). Both forms are identical up to amino acid 239 and have different C termini as the result of alternative splicing (Wang et al., 2012). The most abundant transcript is the PGAM5L, which contains six exons and encodes a protein of 289 amino acids (Lo and Hannink, 2006). Compared to PGAM5L, PGAM5S lacks 50 amino acids in C-terminal and instead possesses 16 unique amino acids via an alternative splicing, resulting in a protein that contains 255 amino acids (Lo and Hannink, 2006). Both isoforms share the N-terminal domain (Figure 1). As a member of the PGAM family, PGAM5 also possesses an active site with a histidine residue, H105 (Chaikuad et al., 2017). Mutation of the histidine residue H105 blocks PGAM5 catalytic activity completely (Wang et al., 2012).

Figure 1. Two isoforms of PGAM5. The two isoforms of PGAM5 are depicted with relevant regions of the proteins indicated. Transmembrane domain (Red); PGAM domain (Blue): Distinctive regions in PGAM5S (Green).

PGAM5 acts as a regulator in cell death pathways

As a mitochondrial phosphatase, PGAM5 has been implicated in multiple cell death pathways, for instance regulating TNF-induced necroptosis, promoting concanavalin A-induced hepatic

Introduction

necrosis, activating apoptosis and mitophagy induced by distinct ligands, and regulating ROS- induced oxeiptosis pathway (He et al., 2017; Holze et al., 2018; Park et al., 2018; Wang et al., 2012).

Necroptosis is a non-classical mode of programmed cell death, morphologically characterized by cell swelling and plasma membrane rupture (Patankar and Becker, 2020). Death receptors including the TNF receptor, normally associated with apoptosis, can also induce necroptosis when caspase-8 is inhibited (Fritsch et al., 2019). TNF receptor-mediated necroptosis is the best-characterized necroptotic pathway, involving the protein kinases RIPK1 and RIPK3, and the executioner protein MLKL as principal components (Sun et al., 2012). Necrosulfonamide, a MLKL inhibitor, has been shown to abolish the phosphorylation of PGAM5S (Wang et al., 2012). Furthermore, PGAM5 has been implicated in the recruitment of the necroptotic complex to mitochondria and the triggering of mitochondria fission through regulation of DRP1 (Wang et al., 2012) (Figure 2A). It has been shown that DRP1 binds to the mitochondrial membrane and promotes the mitochondrial fission process (Smirnova et al., 2001). The activity of DRP1 is regulated by its phosphorylation and dephosphorylation. Phosphorylation at Ser637 inhibits its GTPase activity and prevents mitochondrial fission (Chang and Blackstone, 2007; Cribbs and Strack, 2007). However, the role of the PGAM5/DRP1 axis on necroptosis is controversial, as several subsequent studies showed that PGAM5 and DRP1 deficiency or inhibition did not influence necroptosis (Murphy et al., 2013; Remijsen et al., 2014; Wang et al., 2014). It is possible that the role of PGAM5 in necroptosis is cell and organ-specific. For example, PGAM5 deficiency in mice surprisingly showed exacerbated rather than reduced necroptosis in the heart and brain in response to ischemia reperfusion injury (Lu et al., 2016). Besides, PGAM5 was dispensable for dispensable for necroptosis in mouse embryonic fibroblasts and bone marrow derived macrophages (Moriwaki et al., 2016). In the liver, it has been shown by our group that the PGAM5/DRP1 axis plays an essential role in Con A-induced hepatitis (He et al., 2017). The Con A treatment causes MLKL-dependent but RIPK3-independent programmed necrosis, which contributes to the pathogenesis of severe liver diseases (Gunther et al., 2016). Interestingly, compared to previous studies, which showed that PGAM5 dephosphorylated DRP1 at site S637 to process TNF receptor mediated necroptosis, it was indicated that PGAM5 sufficiency boosts DRP1 phosphorylation at S616 in the liver of Con A treated mice (He et al., 2017; Wang et al., 2012). It was also demonstrated that PGAM5 regulates mitochondrial fission via regulating DRP1 activity and PGAM5 deficiency or DRP1 inhibition protects hepatocytes from necroptosis (He et al., 2017). Recent studies lead to the suggestion that PGAM5 acts as an organ and cell specific protein during necroptosis and is involved in different mechanisms varying in the outcomes depending on its origin (Murphy et

Introduction

al., 2013; Remijsen et al., 2014; Wang et al., 2014) (Murphy et al., 2013; Remijsen et al., 2014; Wang et al., 2014).

Apoptosis is a form of programmed cell death, which is characterized by a series of typical morphological features including cell shrinkage, fragmentation into apoptotic bodies and phagocytosis by neighboring cells. The role of PGAM5 in apoptosis is still controversial. Several studies highlighted the important roles of PGAM5 in regulating apoptosis. Firstly, the insufficiency of PGAM5L, one of the two splice variants of PGAM5, inhibits intrinsic apoptosis caused by arenobufagin (Xu et al., 2015). Secondly, knockdown of PGAM5 inhibits Bax activity to decrease the suppressor of cytokine signaling 6 (SOCS6) induced intrinsic apoptosis (Lin et al., 2013). Thirdly, truncated PGAM5 antagonized its interaction with inhibitor of apoptotic proteins (IAPs) and induced apoptosis via a caspase-dependent pathway (Zhuang et al., 2013) (Figure 2B). On the other hand, it has been reported that PGAM5 knockdown does not affect apoptotic cell death induced by staurosporine (Lin et al., 2013; Wang et al., 2012). Altogether, it is possible that the role of PGAM5 in apoptosis is cell- and stimulus-specific.

Mitophagy is a highly selective and specialized form of autophagy, which often occurs to recycle defective mitochondria following damage or stress. It is crucial for mitochondrial quality control and essential for cell health. As a mitochondrial phosphatase, PGAM5 participates in mitophagy via targeting multiple molecules in the signaling pathway. Both PGAM5 isoforms can bind to apoptosis-inducing factor (AIF) to attenuate antioxidant responses, which in turn triggers mitophagic cell death (Lenhausen et al., 2016) (Figure 2C). While AIF insufficiency fails to block mitophagy induced by PGAM5 overexpression, X-linked inhibitor of apoptosis (XIAP) mediated ubiquitination is capable of blocking PGAM5L induced mitophagy (Lenhausen et al., 2016). Besides, PGAM5 has been shown to participate in PTEN-induced kinase 1 (PINK1) stabilization (Lu et al., 2014; Yan et al., 2020) (Figure 2C). PINK1 stabilization and subsequent processes, like parkin recruitment, trigger mitophagy to eliminate damaged or dysfunctional mitochondria (Springer and Kahle, 2011). PGAM5 deficient mice have a Parkinson’s-like movement disorder, indicating that mitophagy may contribute to neurodegeneration and movement disorders (Springer and Kahle, 2011). In damaged mitochondria, PGAM5 is processed by presenilins-associated rhomboid-like protein (PARL) and stabilizes PINK1 via prohibitin 2 (PHB2). Consequently, stabilized PINK1 phosphorylates and activates Parkin RBR E3 Ubiquitin Protein Ligase (PRKN) leading to mitophagy (Yan et al., 2020). PGAM5 is capable of dephosphorylating FUN14 Domain Containing 1 (FUNDC1), a key mitophagy receptor in mammals (Chen et al., 2014). In the steady state, FUNDC1 activity was inhibited via phosphorylation at Ser13 by casein kinase 2 (CK2) and at Tyr18 by Src kinase (Liu et al., 2012). Upon stress-induced mitochondrial membrane potential loss or

Introduction

under hypoxic conditions, FUNDC1 dissociates from CK2 and Src, but interacts with PGAM5, in turn PGAM5 dephosphorylates FUNDC1 and activates FUNDC1 to induce mitophagy.

Oxeiptosis is a novel caspase-independent cell-death pathway, which utilizes the reactive oxygen species (ROS)-sensing capabilities of Kelch-like ECH-associated protein 1 (KEAP1) and involves PGAM5 and the mitochondrial apoptosis inducing factor 1 (AIFM1), and is highly different from all previously described canonical cell death pathways (Holze et al., 2018) (Figure 2C). PGAM5 appears to be a crucial molecule in the oxeiptosis pathway. Under steady state (physiological ROS levels) PGAM5 physically interacts with KEAP1 and is segregated from AIFM1. Upon stress with high levels of ROS, PGAM5 dissociates from KEAP1 and translocate from the outer mitochondrial membrane into the mitochondrial lumen, and binds to AIFM1. As a phosphatase, PGAM5 can dephosphorylate AIFM1 at a highly conserved Ser116 and the dephosphorylation of AIFM1 induces spontaneous cell death. AIFM1 functions in cell death regulation through a nicotinamide adenine dinucleotide oxidoreductase (NADH) activity modulates complex I, a protein complex, that is indispensable for energy-generating respiratory chain (Bano and Prehn, 2018).

Figure 2. The roles of PGAM5 in regulating cell death. (A) PGAM5/IAPs axis regulates apoptosis. PGAM5 is cleaved by protease and rapidly released from the mitochondria during apoptosis. PGAM5 interacts with XIAP and cIAP1 and sensitizes cells to apoptosis. (B) PGAM5/AIF axis and PGAM5/PINK1 axis regulate mitophagy. PGAM5 regulates mitophagy via two different pathways. Firstly, both the short and long isoforms of PGAM5 binds to AIF to reduce the ability of controlling antioxidant responses. This interaction in turn triggers mitophagic cell death. Secondly, the interaction between PGAM5 and PINK1 stabilizes PINK1 in mitochondria and in turn triggers mitophagy. (C) PGAM5/KEAP1/AIFM1 axis regulates oxeiptosis. PGAM5 forms a complex with KEAP1 and AIFM1 in response to intracellular ROS exposure. This cell death program is named oxeiptosis.

Introduction

The role of PGAM5 in Wnt/β-catenin signaling pathway

Wnt/β-catenin signaling constitutes an evolutionary conserved signaling pathway, which is involved in embryonic patterning of body axes, regulation of stem cell fate and tissue homeostasis (Choi et al., 2013; Clevers and Nusse, 2012; Yang et al., 2015). It has been associated with various diseases including colorectal cancer. The cellular level of β-catenin is tightly regulated by a destruction complex consisting of the tumor suppressor adenomatous polyposis coli (APC), the scaffold protein AXIN1, casein kinase 1 α (CK1α) and glycogen synthase kinase 3 (GSK3) (Clevers and Nusse, 2012). In the absence of Wnt ligands, β- catenin (CTNNB1) is phosphorylated in the destruction complex leading to its subsequent ubiquitination and proteosomal degradation. Upon stimulation of Wnt receptors including frizzled family (FZD) and LDL receptor related protein 5/6 co-receptors (LRP5/6) with extracellular Wnt ligands, β-catenin degradation is blocked and thereby leading to protein accumulation and activation of β-catenin target genes.

Recently, a cell intrinsic pathway, which involves PGAM5 release from damaged mitochondria has been identified to regulate the Wnt/β-catenin pathway (Bernkopf et al., 2018). Under stressed conditions, mitochondria lose their membrane potential, which is critical for driving ATP synthesis, and this triggers cleavage of PGAM5 via the protease presenilin associated rhomboid like (PARL) (Sekine et al., 2012). PGAM5 has been shown to interact with AXIN1 and promotes dephosphorylation of β-catenin, which leads to the stabilization and robust activation of the Wnt/β-catenin pathway. PGAM5 was cleaved by PARL at its N-terminal membrane anchor and released from mitochondrial membranes to the cytosol. The cytosolic appearance, but not the mitochondrial appearance, of PGAM5 leads to β-catenin dephosphorylation which in turn inhibits β-catenin degradation, resulting in Wnt/β-catenin signaling activation. Of note, cytoplasmic PGAM5 expression also increases the numbers of mitochondrial (Bernkopf et al., 2018).

PGAM5 in the NLRP3 inflammasome

The NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome is a critical component of the innate immune system and responds to a broad range of microbial motifs, endogenous danger signals and environmental irritants (Swanson et al., 2019). Under most circumstances, NLRP3 inflammasome activation requires a priming step and an activation

Introduction

steps. As the first step, a priming signal is provided by inflammatory stimuli such as TLR4 agonists to promote NLRP3 and pro-IL1β expression. The activating signal then triggered by PAMPs (pathogen associated molecular patterns) or DAMPs (damage associated molecular patterns) and characterized by NLRP3 inflammasome formation and IL1β release (Hayward et al., 2018).

Several studies indicate that PGAM5 may influence the development of inflammation, especially the NLRP3 inflammasome activation (Kang et al., 2013; Moriwaki et al., 2016; Murphy et al., 2013). Accordingly, induction of IL-1β by LPS in PGAM5 deficient dendritic cells was strongly decreased. Moreover, PGAM5 drives IL1β secretion by LPS-primed BMDMs in response to multiple inflammasome agonists including ATP, nigericin, and poly (dA-dT), indicating its critical role in NLRP3 or Absent In Melanoma 2 (AIM2) inflammasome activation (Moriwaki et al., 2016). Although the physical interaction between PGAM5 and ASC is not detected in BMDMs, PGAM5 is found in the same detergent-insoluble compartment as the inflammasome, indicating that PGAM5 most likely controls inflammasome activation indirectly. Furthermore, PGAM5 accumulation and oligomerization has been shown in the detergent- insoluble compartment in LPS treated cells and it can be further enhanced by nigericin treatment (Moriwaki et al., 2016).

PGAM5 activity is linked to multimerization

Cellular functions are frequently linked to protein multimerization. A proteome-wide analyses in Escherichia coli addressed the abundance of multimerization (Taniguchi et al., 2010). They reported 719 proteins in monomeric form, 198 proteins presented as dimeric form, 16 proteins presented as trimeric form, 47 proteins presented as tetrameric form and other proteins presented as higher order complex. The high abundance of multimers in cells indicate that many proteins may be functional only in their multimeric form. Interestingly, PGAM5 multimeric complexes were observed in multiple cellular processes like NLRP3 inflammasome activation and CCCP-induced mitophagy (Moriwaki et al., 2016; Park et al., 2018; Ruiz et al., 2019). Multimerization may be a general form of PGAM5 appearance in cells and may also be an important process for regulating diverse signaling pathways.

PGAM5 dimers were identified in the in LPS treated BMDMs and the amount of dimers was significantly increased following nigericin treatment. The presence of PGAM5 dimers indicates that PGAM5 may function as a dimer or multimer and such structures maybe important for inflammation regulation.

Introduction

Recently, PGAM5 multimerization has been identified as a key process to regulate mitochondrial fission and mitophagy (Ma et al., 2020). Oxidative stress caused by selenite leads to the formation of PGAM5 multimers, which liberates PGAM5 from BCL-xL, leading to mitochondrial damage. This process can be inhibited by scavengers, which diminish mitochondrial ROS. PGAM5 multimers are also capable of eliminating damaged mitochondria via inducing mitochondrial fission through FUNDC1 dephosphorylation, resulting in cell survival. However, upon a combined treatment which induces mitochondrial ROS and mitotic arrest, cells were unable to recover from the lethal stress and underwent cell death instead of mitophagy (Ma et al., 2020).

Thus, the formation of PGAM5 multimers presents a distinct way of how cells deal with multiple stress conditions. This dramatic conformational change significantly promotes or inhibits the interaction with other molecules and is likely to regulate several cell signaling pathways.PGAM5 may regulates various pathways via forming multimers.

3.2 Cellular defense against viruses

Toll-like receptors

Toll-like receptors (TLRs) were characterized as pattern recognition receptors (PRRs) (Wada and Makino, 2016). Different species appear to have a distinct number of functional TLRs. E.g. in humans, 10 different TLRs (TLR1–10) and in mice 12 different TLRs (TLR1−9, 11−13) have been identified (Chen et al., 2019). TLRs are transmembrane proteins, which are located in the cell membrane and in endosomes. All TLRs share a similar domain organization, as each family member is characterized by extracellular domain containing leucine-rich repeats (LRRs), a transmembrane region and a cytoplasmic Toll/IL-1 receptor (TIR) domain (Lester and Li, 2014). They specifically recognize corresponding ligands (Table 1) (Dela Justina et al., 2020; Li et al., 2020).

Table 1. TLRs and their ligands

TLRs Ligands

TLR1, TLR2 and TLR6 Gram-positive and Gram-negative components such as lipoproteins. TLR4 LPS

TLR5 Flagellin TLR3 Double stranded RNA

Introduction

TLR7 and TLR8 Single stranded RNA

TLR9 Unmethylated CpG DNA

Upon ligand recognition, the TIR domain of TLRs interacts with respective cytosolic adapters and triggers downstream signaling pathways. Based on different adapters, TLR pathways can be classified as myeloid differentiation primary response gene 88 (MyD88)- or TIR-domain containing adaptor inducing IFN-β (TRIF)-dependent pathways. Both TLR3 and TLR4 utilize the adapter TRIF while other TLRs apart from TLR3 recruit also MyD88. These adapters further trigger the activation of NF-κB, activator protein 1 (AP-1) and IRF3 to initiate innate and adaptive immune responses against the invasion of various pathogens (Aziz et al., 2020; Li et al., 2020).

TLRs can recognize virus-derived PAMPs and induce the production of interferons (IFNs) and pro-inflammatory cytokines. Among these TLRs, TLR3, TLR7, TLR8, TLR9, TLR2 and TLR4 have been identified as important players in antiviral innate immunity (Lester and Li, 2014). These TLRs have been shown to induce different pathways to promote an antiviral response (Figure 3) (Lin et al., 2017; Moradi-Marjaneh et al., 2018).

TLR3 recognizes dsRNA, which constitutes a molecular signature of many viruses. Since the dsRNA can be a genetic information carried by many RNA viruses or an ephemeral replicative intermediate during virus replication, TLR3 recognizes viruses including dsRNA viruses, ssRNA viruses and DNA viruses. After the recognition of dsRNA, unlike other TLRs, TLR3 uses TRIF as a sole adaptor and induces the activation of NF-κB and production of IFNs (Saghazadeh and Rezaei, 2017). In in vitro experiments synthetic dsRNA analogues, like Polyriboinosinic:polyribocytidylic acid (poly(I:C)), were found to activate TLR3-dependent signaling pathways. Therefore, it has been widely used to mimic RNA virus infection.

TLR7 and TLR8 are mainly expressed in immune cells, indicating their important roles in innate immunity (Lester and Li, 2014). TLR7 and TLR8 are known as lysosomal sensors and bind to GC- and AU-rich ssRNA (Komura et al., 2019). TLR7 detects RNA degradation products rather than a ssRNA fragment itself. It is well known that lysosomes are important organelles for RNA degradation. Therefore, lysosomes serves as an original of RNA degradation products toTLR7. Upon viral ssRNA recognition, TLR7 and TLR8 use a MyD88- dependent pathway to activate NF-κB and produce IFN-α (Brennan and Gilmore, 2018).

TLR9 is another endocytic PRR that recognizes unmethylated CpG motifs from ssDNA fragments. Given the fact that unmethylated CpG often exists in microbial or viral genomes, TLR9 has a vital role in sensing bacterial infection or DNA viral infection (Ohto et al., 2015).

Introduction

Like TLR7 and TLR8, TLR9 is also expressed in numerous immune cells and utilizes MyD88- dependent pathways to activate NF-κB and produce IFN-α.

TLR2 and TLR4 differ from all TLRs mentioned above, as they don’t detect nucleic acids. They are located in the cell membrane and sense viral envelop proteins or viral proteins released to the extracellular milieu (Urcuqui-Inchima et al., 2017). After ligand binding, TLR2 and TLR4 follow different manners to induce downstream signaling pathways. TLR2 cooperates with other TLRs, like TLR1 or TLR6, to form heterodimers and initiate MyD88-dependent pathways to activate NF-κB. In contrast, TLR4 alone can induce MyD88- and TRIF- dependent pathways to activate NF-κB and produce IFNs, respectively (Lester and Li, 2014).

In summary, TLRs are crucial in antiviral pathways as they recognise different viral components to activate NF-κB and produce IFNs.

Figure 3. TLRs signaling pathways related to virus infection. TLRs are essential for containing virus infection. Several TLRs including TLR2, TLR3, TLR4, TLR7/8, and TLR9 are involved in antiviral responses by triggering the production of antiviral cytokines such as type I IFNs. Besides, some TLRs including TLR2, TLR3, TLR4 and TLR9 are able to active NF-κB pathway to further promote antiviral responses.

Introduction

RIG-I-like receptors

The RLRs family, encompassed of RIG-I, MDA5 and LGP2, recognizes dsRNA in the cytosol of most cell types (Chiang and Gack, 2017). All RLRs have a central helicase domain and a C-terminal domain which together are essential for viral RNA detection. Besides these two domains, RIG-I and MDA5, additionally have two similar N-terminal caspase activation and recruitment domains (CARDs), which mediate downstream signal transduction (Oshiumi et al., 2016). RIG-I recognizes short dsRNA possessing both panhandle structures and a 5′ triphosphate moiety, while MDA5 is thought to bind to long dsRNA moiety in a cooperative manner (Luo et al., 2011; Peisley et al., 2012; Zheng et al., 2015). LGP2, on the other hand, lacks the CARDs and is believed to be involved in regulating the function of RIG-I and MDA5.

As mentioned above, RLRs family receptors contribute to antiviral functions via recognizing viral RNA. Upon RNA ligand binding and oligomerization, RIG-I or MDA5 interacts with mitochondrial antiviral-signaling protein (MAVS) via CARD–CARD motifs. MAVS is an indispensable adaptor in the RLR signaling pathway and forms aggregates to transduce downstream signaling (Hou et al., 2011). It leads to the phosphorylation of TANK-binding kinase 1 (TBK1), IκB kinase-ε (IKKε) and IKKα/β/γ, which in turn activate several transcription factors, such as IRF3/7 and NF-κB. Such activation drives transcription of the genes encoding type I interferons and pro-inflammatory cytokines (Chiang et al., 2014; Zhao and Karijolich, 2019) (Figure 4). Subsequently, the secreted IFNα/IFNβ bind to their respective receptors and induce the expression of hundreds of ISG to protect cells from virus infection.

Although all RLRs exhibit a functional similarity, they still recognize different viruses. Influenza virus, Newcastle disease virus (NDV), Sendai virus (SeV), respiratory syncytial virus, measles virus, vesicular stomatitis virus (VSV), rabies virus and Japanese encephalitis virus are mainly recognized by RIG-I receptor (Foronjy et al., 2015; Jahan et al., 2020; Kato et al., 2006; Oh et al., 2016; Yang et al., 2019). Given the fact that AT-rich dsDNA can be a template for RNA polymerase III and transcribed into dsRNA containing a 5'-triphosphate moiety, RIG-I can also recognize some DNA viruses including herpes simplex virus type 1 which contain an AT-rich DNA genome (Zhao et al., 2016; Zhu and Zheng, 2020).

MDA5 is essential for innate immunity against various viruses including encephalomyocarditis virus (EMCV), mengovirus and murine norovirus (MacDuff et al., 2018; Sanchez David et al., 2019; Zhong et al., 2020). Of note, RIG-I and MDA5 sometimes work together against viral infection, e.g. during West Nile Virus infection (Zhang et al., 2017a). LGP2 is different to other RLR members, as it does not have CARD domain, which is mandatory for interacting with MAVS during antiviral signaling (Shi et al., 2017). Therefore, LGP2 was described as a

Introduction

regulator, but not an essential component in RIG-I and MDA5 signaling pathways (Uchikawa et al., 2016).

Figure 4. RLRs pathways related to virus infection. Despite the overall structural similarity between RIG-I and MDA5, they detect distinct viral species. RIG-I recognizes short dsRNA while MDA5 is thought to bind to long dsRNA. Their activation is tightly regulated by LGP2. RIG-I and MDA5 signal to MAVS, which induces the phosphorylation of TBK1 and IRF3. The activation of these two key molecules in turn initiates the production of IFNs.

MAVS is a key factor in the RLR pathway

MAVS, also known as CARD adapter inducing IFN beta (Cardif), virus-induced signaling adapter (VISA) and IFN-β promoter stimulator 1(IPS-1), is a key signaling protein activated by viral RNA and the sensors RIG-I and MDA5 (Kawai et al., 2005; Meylan et al., 2005; Seth et

Introduction

al., 2005; Xu et al., 2005). MAVS is a 540 amino acid protein, containing: a C-terminal transmembrane domain (TMD), a proline-rich domain and a N-terminal CARD domain. Different domains have different functions. Firstly, the CARD domain is important for MAVS interaction with upstream RIG-I and MDA5 (Rout et al., 2019). The interaction initiates the following series of MAVS-mediated responses. Secondly, the proline-rich domain is required for recruitment of other molecules, which are critical for activating downstream pathways. TRAF2, TRAF5 and other proteins are recruited to MAVS to form a MAVS signaling complex, which leads to the activation of NF-κB and IRF3 (Liu et al., 2013). These activation induce the production of type I IFNs and other pro-inflammatory cytokine. Thirdly, the TMD domain keeps MAVS located in the outer membrane of mitochondria and peroxisomes (Dixit et al., 2010). Upon viral infection, peroxisomal MAVS provides rapid but short-term protection via an IFN- independent pathway, whereas mitochondrial MAVS induces delayed, but a stable antiviral response via an IFN-dependent pathway (Dixit et al., 2010).

During viral infection, MAVS forms well-known aggregates, which are resistant to 2% sodium dodecyl sulfate (SDS), as shown by semi-denaturing detergent agarose gel electrophoresis (SDD-AGE), a commonly used assay to detect prion particles (Hou et al., 2011). Cell free assays revealed that a small amount of exogenous MAVS protein which contains CARD domain but lack the signaling domain was able to force endogenous full-length MAVS into aggregates (Hou et al., 2011). More importantly, these MAVS aggregates were functional in IRF3 activation and therefore triggered downstream signaling (Hou et al., 2011). Overall, the prion-like aggregates of MAVS have been shown to promote the activation of antiviral signaling pathways.

In order to fully understand the mechanisms mediating MAVS aggregation, it is key to figure out how molecules like RIG-I lead to MAVS prion conversion. Since yeast lack a RLR pathway, it has been widely used as a model to reveal several key molecules in this signaling pathway. Exogenous RIG-I expression in yeast is sufficient to induce prion formation of MAVS protein which contain CARD domain only, indicating a direct interaction between RIG-I and MAVS via CARD domains (Cai et al., 2014). A point mutation of MAVS CARD domain attenuated MAVS aggregation and MAVS-dependent antiviral signaling, further suggesting that RIG-I-dependent MAVS activation is necessary for signal transduction (Liu et al., 2013).

In innate immune signaling, it has been shown that posttranslational modifications (PTMs) play important roles via targeting distinct sensors or adaptors. There are several conventional types of PTMs such as phosphorylation and ubiquitination, as well as other unconventional modifications, including i.e. acetylation, methylation.

Introduction

Interestingly, phosphorylation and dephosphorylation have been shown to regulate MAVS function during viral infections (Oshiumi et al., 2016). The tyrosine kinase c-Abl was shown to phosphorylate MAVS, leading to MAVS activation and type I IFN production during VSV infection (Cheng et al., 2017; Song et al., 2010). Using tyrosine-scanning mutational analysis, it was revealed that VSV infection induces MAVS phosphorylation at Tyr9 to promote MAVS/TRAF3/TRAF6 complex formation and RLR signaling (Wen et al., 2012). Moreover, MAVS’s serine-rich clusters in the C-terminus, which contains Ser442, can be directly phosphorylated by TBK1 and IKK upon virus infection (Liu et al., 2015). The phosphorylation of MAVS leads to the recruitment of IRF3 and facilitates IRF3 phosphorylation (Figure 5). Finally, protein phosphatase magnesium-dependent 1A (PPM1A; also known as PP2Ca) has been shown to dephosphorylate several key molecules in the RLR pathway, including MAVS and TBK1/IKKε, leading to the inhibition of MAVS-mediated antiviral responses (Xiang et al., 2016).

Figure 5. MAVS’s aggregation and phosphorylation. Aggregation and phosphorylation of MAVS have been shown as two very important modifications during RLR pathway activation. After binding to viral RNA, RIG-I/MDA induces MAVS aggregation. Other molecules in the RIG-I pathway, including c-Abl, TBK1 and IKKε, are able to phosphorylate MAVS in transducing signaling. Both of these two modifications are crucial for inducing IFN expression.

Introduction

IFNs and antiviral responses

The IFNs were firstly identified in the early 1950s and later purified from the medium of cultured human white blood cells in 1981. There are several members belonging to a family of different IFNs (Cantell et al., 1981). Until now, more than 20 distinct IFNs have been identified in different species including human and mouse (Schneider et al., 2014). To induce an antiviral response, IFNs need to be released into extracellular milieu and recognized by respective receptors. Based on the specific recognition, IFNs were divided into three different types. IFNα, IFNβ, IFNδ, IFNε, IFNζ, IFNκ, IFNτ, and IFNω belong to type I IFNs, which bind to the IFNAR receptor complex (IFNAR1 and IFNAR2). Among type I IFNs, IFNα and IFNβ have received the most attention for the roles in fighting viral infection (Wang and Fish, 2019). Type I IFNs and IFNAR have a broad tissue distribution and can be produced by multiple types of immune cells, indicating their important role in antiviral innate immunity (Schneider et al., 2014). Viral infections trigger host innate immune responses, characterized by the production of type-I interferons (IFN) including IFNβ. IFNβ induces cellular antiviral defense mechanisms and thereby contributes to pathogen clearance.

IFN-γ (gamma) is the only IFN belonging to type II IFNs. IFN-γ is recognized by the type II IFN receptor which consists of two IFNγ receptor 1 (IFNGR1) and two IFNγ receptor 2 (IFNGR2) subunits. Unlike type I IFNs, IFNγ is mainly produced by immune cells and contributes to innate and adaptive immunity (Alspach et al., 2019).

Within the group of type III IFNs, the following molecules were identified: IFNL1, IFNL2, IFNL3 and IFNL4 (also known as IFN-λ1, IFN-λ2, IFN-λ3 and IFN-λ4) (Kotenko et al., 2019). These four members of the IFN-λ family can be recognized by the type III IFN receptor (IFNLR1) with a high binding affinity (Andreakos et al., 2019). Additionally, due to the structural similarity to the IL-10 cytokine family, type III IFNs can also bind to IL-10 receptor 2 with low affinity (Kotenko et al., 2019).

Although three types of IFNs were recognized by three different receptor complexes, they share a common JAKs-STATs signaling pathway to induce the expression of IFN-stimulated genes (ISG) (Raftery and Stevenson, 2017). For example, upon binding of type I IFNs to the IFNAR receptor, it recruits JAK1 and tyrosine kinase 2 (TYK2) to phosphorylate STATs. Following phosphorylation, STAT proteins undergo conformational change and form a homo- (type II IFN) or heterodimer (type I and type III IFN) (Wang et al., 2017). STAT1 and STAT2 heterodimers are released from IFN receptors and enter the nucleus, where they bind to IRF9

Introduction

to drive ISG expression (Au-Yeung and Horvath, 2018). The production of ISG contributes to antiviral response.

Modeling RNA viral infection

Poly(I:C), a synthetic analogue of viral dsRNA, has been used to mimic RNA virus infection and is a common tool for research of the viral immune responses. Via an intracellular or extracellular rout, poly(I:C) activates two distinct pathways crucial for antiviral responses. Firstly, the TLR3 signaling pathway senses extracellular poly(I:C) stimulation (Bianchi et al., 2017). Upon binding of naked poly(I:C), TLR3 recruits adaptor TRIF to activate NF-κB and IRF3/ IRF7, and subsequently induces the expression of type I IFNs and other pro- inﬂammatory cytokines. Secondly, the RLR signaling pathway is activated by intracellular poly(I:C) (Chen et al., 2017). Based on the molecular weight, low molecular weight poly(I:C) is mainly recognized by RIG-I, while high molecular weight poly(I:C) is recognized by MDA5 (Kato et al., 2008). Both RLRs transduce the signal through MAVS, which forms functional prion-like aggregates to activate downstream molecules, such as TBK1 (Hou et al., 2011; Kawai et al., 2005; Meylan et al., 2005; Seth et al., 2005; Xu et al., 2005). TBK1 in turn activates IRF3 and induces the production of type I IFNs and other pro-inflammatory cytokines.

Vesicular stomatitis virus (VSV) is a negative strand RNA virus, which belongs to the rhabdovirus family. It comprises a single negative strand of 11 kb genomic RNA encoding five major viral proteins: a nucleoprotein, a matrix protein (M), a phosphoprotein, an RNA- dependent RNA polymerase and a G protein (Obuchi et al., 2003). These proteins are crucial and support VSV to complete its life cycle in host cells, including host cells attachment, viral evasion, and viral replication. Unlike many other viruses, VSV can infect a large number of vertebrate and invertebrate hosts, and even some plants. However, its preferred hosts are: goats, horses, sheep, cattle, mules, and pigs. The most significant symptoms are vesicles on various organs including skin, tongue and teats. The VSV infection on susceptible livestock creates a burdensome socio-economic issue although it is not fatal to the animals. Only few VSV infection cases were reported on humans with self-limiting flu-like symptoms. Effective immune responses against VSV predominantly involve the robust production of IFNs and subsequently the generation of antiviral responses (Deng et al., 2020; Drokhlyansky et al., 2017). Given these facts, VSV has been widely used to study the host innate immune response against single strand RNA viruses.

Introduction

3.3 IBD

Type I IFNs is one of the most potent effectors for elimination of viruses and has been tightly connected to the pathogenesis of IBD. IFNAR1 deficient mice were reported to be hypersensitive to DSS-induced colitis and IFNβ administration protects mice from experimental colitis (Katakura et al., 2005). Besides, enteric viruses-induced IFNβ expression ameliorated gut inflammation (Yang et al., 2016).

The two most notable forms of IBD are UC and CD (Lee et al., 2020). Both diseases are chronic inflammatory intestinal disorders, but they differ in the location and nature of the inflammatory changes. CD is a multifactorial disease exhibiting loss of intestinal epithelial barrier integrity and dysregulated immune cell responses due to unknown environmental triggers in genetically predisposed individuals (Kim and Cheon, 2017). Patients with CD may develop inflammation in any part of the digestive tract but typically develop terminal ileitis. In contrast, UC is a chronic inflammation restricted to the colon and rectum which may present with a range of mild to severe symptoms including abdominal pain, diarrhea and hematochezia (Neurath, 2012). Both forms of the disease are thought to be caused by an abnormal immune response of the intestinal immune system against to the dysfunction of intestinal mucosal barrier (Zundler et al., 2019).

CD and UC patients show significant damage of the intestinal epithelium and manifest a notable increase in the death of intestinal epithelial cells (Patankar and Becker, 2020). The cell death in the intestinal epithelium has been shown to be tightly regulated by multiple cell death pathways including apoptosis and necroptosis (Gunther et al., 2013). The association between cell death and IBD has also been found in epithelial cell studies in both IBD patients and IBD mouse models. Furthermore, studies have identified several molecules in the apoptosis or necroptosis pathways as crucial players in regulating ileitis or colitis. For example, caspase-8 has been identified as an important molecule in regulating apoptosis and necroptosis and mice deficient for the caspase-8 specifically in IEC were shown to be sensitive to TNF induced cell death and spontaneously develop ileitis (Gunther et al., 2011; Schwarzer et al., 2020). Moreover, dependent on the microbial environment these mice were shown to suffer from spontaneous colitis as well (Stolzer et al., 2020).

Besides caspase-8, recent studies have highlighted the important role of the NLRP3 inflammasome in regulating gut homeostasis (Cui et al., 2020; Singh et al., 2019). However, the functional contribution of the NLRP3 inflammasome in IBD remains controversial. One of the first studies exploring the role of NLRP3 in colitis showed that NLRP3 deficiency protected mice from DSS-induced colitis and TNBS-induced colitis (Bauer et al., 2012; Bauer et al.,

Introduction

2010). In contrast, another study showed a totally opposing result. Mice with NLRP3 deficiency exhibited more severe DSS-induced colitis than WT mice (Zaki et al., 2010). In addition, mice lacking other components of the NLRP3 inflammasome such as ASC and caspase-1 were characterized by more severe inflammation in the DSS-model (Zaki et al., 2010).

3.4 Aims of the project

PGAM5 has been identified as a key molecule in a broad range of biological processes, for instance regulating certain cell death pathways and promoting the activation of the NLRP3 inflammasome. Here, this thesis hypothesized that PGAM5 might contribute to viral infections and to the pathogenesis of IBD.

Therefore, the following points were addressed in the thesis:

 Analysis of the dynamics and location of PGAM5 after viral challenge.

 Elucidation of a functional contribution of PGAM5 in regulating IFNs expression.

 Studying the signaling mechanism of PGAM5 in regulating IFNs expression.

 Characterization of the functional contribution of PGAM5 in regulating IFNs downstream pathways.

 Exploration of PGAM5 functions in murine disease models: poly(I:C)-induced small intestinal inflammation and DSS-induced colitis.

To verify the hypothesis, PGAM5 deficient HeLa cells were generated via using CRISPR/Cas9 technology and PGAM5 deficient murine embryonic fibroblasts (MEFs) were isolated from PGAM5−/− mice. Moreover, PGAM5 conditional and total knockout mice were used for in vivo studies.

Material and Methods

4 Material and Methods

4.1 Materials

The plastic and glass lab-wares were purchased from Corning Life Sciences (Tewksbury, USA), Greiner Bio-one (Frickenhausen, Germany), Becton-Dickinson GmBH (Heidelberg, Germany), Eppendorf (Hamburg, Germany), Nunc (Langenselbold, Germany), Simax (Czech Republic), Schott Duran (Germany) and Miltenyi Biotec (Bergisch Gladbach, Germany).

Table 2. General chemicals and reagents used in the study

Name Supplier Acetic acid Merck, Darmstadt, Germany Acetone Roth, Karlsruhe, Germany

Agar Roth, Karlsruhe, Germany

Agarose Roth, Karlsruhe, Germany

Ampicillin Roth, Karlsruhe, Germany

Ammonium hydroxide (NH4OH) Roth, Karlsruhe, Germany

Ammonium chloride (NH4Cl) Merck, Darmstadt, Germany BSA (Bovine serum Albumin, Miltenyi Biotec, Bergisch Fraction V ) Gladbach, Germany Bis-polyacrylamide Roth, Karlsruhe, Germany

Calcium Chloride (CaCl2) Merck, Darmstadt, Germany Citric acid Sigma-Aldrich, Steinheim, Germany Collagenase IV Sigma-Aldrich, Steinheim, Germany Dimethyl sulfoxide (DMSO) Roth, Karlsruhe, Germany

Material and Methods

Disodium hydrogen phosphate Roth, Karlsruhe, Germany

(Na2HPO4) DNase I Roche, Mannheim, Germany DMEM Thermofischer Scientific, Germany ECL western lightening plus PerkinElmer, Massachusetts, USA EDTA (Ethylenediaminetetraacetic Merck, Darmstadt, acid) Germany Eosin Merck, Darmstadt, Germany Entelan Merck, Darmstadt, Germany Ethanol Roth, Karlsruhe, Germany

Ethidium Bromide Fluka, Taufkirchen, Germany FCS (Fetal calf serum) PAA, Pasching, Austria

Fluorescent mounting medium Vector laboratories, Burlingame, USA Formaldehyde Merck, Darmstadt, Germany Formamide Merck, Darmstadt, Germany Glucose Sigma Aldrich, Taufkirchen, Germany Glycine Roth, Karlsruhe, Germany

Glycerol Merck, Darmstadt, Germany Haematoxylin Merck, Darmstadt, Germany Histofix Roth, Karlsruhe, Germany

Hoechst Invitrogen, Darmstadt, Germany

Material and Methods

Hydrogen peroxide (H2O2) 3% Roth, Karlsruhe, Germany

Hydrochloric acid (HCl) 37% Merck, Darmstadt, Germany Isofluran Florene® 100%, Abbott, Wiesbaden Isopropanol Roth, Karlsruhe, Germany

Magnesium chloride (MgCl2) Roth, Karlsruhe, Germany

β-mercaptoethanol Roth, Karlsruhe, Germany

Methanol Roth, Karlsruhe, Germany

Milk powder Sigma Aldrich, Saint Louis, USA Mini-PROTEAN® Precast Gels Bio-rad, Munich, Germany

NP40 lysis buffer Thermofischer Scientific, Germany Paraformaldehyde Merck, Darmstadt, Germany Penicillin/Streptomycin PAA, Pasching, Austria (Pen/Strep) PeqGold DNA ladder Peqlab, Erlangen, Germany PeqGold Protein Marker Peqlab, Erlangen, Germany Phosphate Buffered Saline (PBS) Sigma Aldrich, Steinheim, Germany Phosphatase inhibitor Roche, Mannheim, Germany Proteinase K Roth, Karlsruhe, Germany

Protein Block Dako, Hamburg, Germany

Protease Inhibitor Cocktail Tablets Roche, Mannheim, Germany RIPA Thermo Fisher Scientific, Germany SYBR green I Roche, Mannheim, Germany

Material and Methods

Sodium Chloride (NaCl) Roth, Karlsruhe, Germany

Sodium Dodecyl Sulphate (SDS) Roth, Karlsruhe, Germany

Sodium Hydroxide (NaOH) Roth, Karlsruhe, Germany

Trichloroacetic acid (TCA) Merck, Darmstadt, Germany Tris base Roth, Karlsruhe, Germany

Tris-HCl Roth, Karlsruhe, Germany

Triton-X 100 Merck, Darmstadt, Germany Trypan Blue Fluka, Taufkirchen, Germany Trypsin-EDTA PAA, Pasching, Austria

Tween 20 Roth, Karlsruhe, Germany

TSA (Tyramide Signal PerkinElmer, amplification) Massachusetts, USA

Table 3. Commonly used Instruments in the study (unless indicated in the text)

Name Supplier

Autoclave VX-150 Systec, Wettenberg, Germany

Centrifuge Heraeus Multifuge XIR Thermo Scientific, Osterode, Germany Cooling centrifuge 5430R Eppendorf, Hamburg, Germany

CO2 Cell Culture Incubator Heracell 240, Thermo Scientific, Germany Confocal Microscope Leica TCS SP8

Clean Bench Herasafe KS Thermo Scientific, Langenselbold, Germany Cryostat Leica

Infinite M200 microplate reader Tecan

Freezer -80 °C, Hera freeze Thermo Electron Corporation

Gentle MACS Dissociator Miltenyi Biotec, Bergisch Gladbach, Germany

Material and Methods

Light Cycler® 480 Roche Diagnostics, Mannheim, Germany Microscope DMI 400B Leica, Mannheim, Germany

ND-1000 Spectrophotometer NanoDrop, Thermo Scientific

pH-Meter inoLab, Klando, Czech Republic

Tissuelyser Retsch

Thermomixer Eppendorf

Vortexer Scientific Industries

Western developer Curix 60 AGFA

Gel image Peqlab

UV-Spectrophotometer Eppendorf Biophotometer

Water bath Memmert

4.2 Animals

C57BL/6 and Villin-Cre (B6.Cg-Tg(Vil1-cre)997Gum/J) mice were obtained from Jackson Laboratory (Bar Harbor, USA). PGAM5−/− (Pgam5tm1b(EUCOMM)Wtsi) mice were obtained from the International Knockout Mouse Consortium. PGAM5 fl/fl (B6(Cg)- Pgam5tm1c(EUCOMM)Wtsi/FkmcJ) mice (C57BL/6 background) were generated from PGAM5−/−mice by crossing them with FLP mice. PGAM5 fl/fl mice were crossed with Villin-Cre mice to generate conditional gut specific PGAM5 knockout mice. Genotyping was performed with PCR using DNA isolated from tissue pieces.

All mice used in the study were bred and maintained in individually ventilated cages under a 12h light/dark cycle. Food and tap water were provided ad libitum. All experimental procedures were performed using committee-approved protocols.

4.3 Cell line HeLa cells were obtained from the American Type Culture Collection (ATCC) and cultured in complete Dulbecco’s modified Eagle’s medium (DMEM) (Gibco) supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 μg/ml).

Material and Methods

4.4 Methods

MEF isolation

Breeding cages were set up by putting one adult male and two adult female mice in each cage. To obtain WT and PGAM5−/− MEFs from littermates, PGAM5+/− mice were used for breeding cages. Every morning after setting up mating, all females were checked for the presence of a copulation plug (vaginal plug) in the vagina and the age of the embryos was starting with as day E0.5 on the day when the vaginal plug was detected.

On day E13.5, MEFs were isolated from fetuses from pregnant mice. In brief, embryos were washed twice with phosphate buffered saline (PBS), and then the head and internal organs of each embryo were removed with scissors. Parts of the tissue from head was used for genotyping. The remaining tissues were dissected and digested in trypsin for 30- 45min at 37°C. Trypsin activity was neutralized via adding complete Dulbecco’s modified Eagle’s medium (DMEM) (Gibco) supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 μg/ml). After centrifugation at 300g for 5 min at 4°C, cell pellets were dispersed and cultured in complete DMEM. Cells were frozen for future usage when reaching 80-90% confluence and recorded as passage P0.

Generation of HeLa knock-out cells using CRISPR/Cas9 technology

For generation of PGAM5 and MAVS knockout cells from single-cell colonies, PGAM5 CRISPR/Cas9 KO Plasmid (Santa Cruz) or MAVS CRISPR/Cas9 KO Plasmid (Santa Cruz) was transfected into HeLa cells via using Lipofectamine 2000 (Invitrogen). Since these plasmids all contain a GFP marker, cells were checked under a fluorescence microscopy to visually confirm transfection. 24 h post transfection, cells which were successfully transfected with knockout plasmid showed a green fluorescence. To obtain GFP positive cells from cultured cells, cells were sorted (FACSAria II; BD) into 96-well plates (single cell/well). Sufficiency of knockout was verified by immunofluorescence and Western Blot (WB) analysis. Mycoplasma contamination was detected with the help of the Mycoplasma Detection Kit (Invivogen), which was used in accordance to the manufacturer’s instruction. In case of contamination, Plasmocin™ was added to the respective cell culture for two weeks and medium was replaced twice a week.

Material and Methods

PGAM5 overexpression

To reconstitute PGAM5 expression in PGAM5 knockout HeLa cells. Full-length PGAM5 vectors, phosphatase mutant PGAM5 vectors, long form PGAM5 vectors or short form PGAM5 vectors (Gifts from H. Ichijo and X. Wang; (Takeda et al., 2009; Wang et al., 2012)) were transfected into cultured cells using Lipofectamine 2000 in FBS free medium. 6 h post transfection, medium was replaced with culture medium containing FBS. 48 h post transfection, cells were used for poly(I:C) simulation. The overexpression was validated via WB or immunofluorescence.

Cell stimulation and infection

To stimulate cells with intracellular RNA, 1µg/ml low molecular weight poly(I:C) (Invivogen) or 0.5 µg/ml 5′pppdsRNA (Invivogen) was transfected into cells for 8h using Lyovec (Invivogen) or Lipofectamine 2000 (Invitrogen) according to the manufacturers instructions. To stimulate cells with intracellular DNA, 1 µg/ml Poly(dI:dC) (Invivogen) was transfected into cells for 8h by using Lipofectamine 2000 (Invitrogen). To active TLR3 or TLR4, cells were incubated with 50 µg/ml low molecular weight poly(I:C) (Invivogen) or 100 ng/ml LPS (Sigma).

For virus infection, VSV (Indiana strain; kindly provided by Prof. Peter Stäheli; Universitätsklinikum Freiburg) was propagated on BHK21 cells. MEFs were infected with VSV (0.1 MOI) in FBS-free medium in the incubator for 1h and then switched to complete medium containing 10% FBS for the indicated time points.

Genotyping

The genetic analysis for different mouse genotypes of 21-28 days old mice were performed from ear biopsies digested in Willi-buffer (150 μl) supplemented with Proteinase-K (20 mg/ml, 5 μl). For screening of MEFs, mouse embryonic tissues were also digested in the same buffer. The solution was kept under constant shaking at 1000 rpm at 55°C overnight. On the following day, activity of Proteinase-K was terminated by increasing the temperature to 95°C for 10 min. Subsequently, the solution was diluted with 1ml of distilled water for further analysis. For PCR, 2 μl of digested DNA sample was mixed with 2 μl of respective primer, 5 μl of Red Load Taq Master mix (Jena Biosciences) and 14 μl distilled water in PCR tube stripes (Peqlab).

Material and Methods

Genotyping of PGAM5−/−, PGAM5 flox/flox and Villin-Cre mice was performed using the following primers:

Table 4. Various primers used for genotyping in the study

Primer name Sequence PGAM5-F AGG CTG GAT CAC TAT AAG GC PGAM5-R CTG GAG ACA TTG TGA CCA TC PGAM5 flox-F CAG TAC AGT TCT AGG CTC CG PGAM5 flox-R GTT CTC GGT TTG ACT GAG AG Villin cre – 1 ACA GGC ACT AAG GGA GCC AAT G Villin cre – 2 ATT GCA GGT CAG AAA GAG GTC ACA G Villin cre – 3 GTT CTT GCG AAC CTC ATC ACT C

The following PCR program protocols were used:

Table 5. PCR programs used for genotyping in the study

PGAM5 PGAM5 flox Villin cre 1 95°C, 5´ 95°C, 5´ 94°C, 4´ 2 94°C, 40´´ 94°C, 40´´ 94°C, 45´´ 3 61°C, 45´´ 61°C, 45´´ 68°C, 1´30´´ 4 72°C, 1´30´´ 72°C, 1´30´´ 72°C, 5´ 5 Repeat steps 2-4; 35x Repeat steps 2-4, 35x Repeat steps 2-4; 35x

6 72°C, 5´ 72°C, 5´ 72°C, 5´

To analyze PCR products according to their size, agarose gel electrophoresis was performed. Gels were prepared by dissolving 1.5% agarose (Roth) in 1X TAE buffer in a microwave oven. The fluorescent dye Ethidium bromide was added to the agarose solution at a concentration of 5 μg/ml. The solution was placed into a cassette tray with spacer combs. After solidification, the gel was placed in the electrophoresis chamber (Peqlab), filled with 1X TAE buffer, loaded with the PCR products and run at 100V until desirable resolution of the bands was achieved. The DNA fragments were separated according to

Material and Methods

their size and was visualized under UV light. The images were recorded in a UV- transilluminator device (BioVision, Peqlab).

Subsequent buffers were prepared by following protocols:

Table 6. The composition of different buffers used in the procedure

Willi buffer 50X TAE

100 mM Tris 24.2 g Tris 200 mM NaCl 100 until desireable 10 mM EDTA resolution of the bands 0.2 % SDS was achieved ml 0.5 M 1 L distilled water EDTA 57.1 ml Acetic acid 842.9 ml distilled water

Dextran sulfate sodium (DSS)-induced colitis

Acute colitis was induced by addition of dextran sodium sulfate (DSS; molecular weight: 36 000–50 000; MP Biomedicals) at concentration of 2 % (wt/vol) to the drinking water ad libitum for 6 days and followed by 3 days of regular drinking water. A freshly prepared DSS solution was administered every third day throughout the experiment. Body weight of mice was recorded daily. Mice were sacrificed by cervical dislocation on day 9 and blood and colon samples were collected for histological and molecular analysis. Endoscopy was performed at day 9.

Gene expression analysis

4.4.7.1 RNA isolation and cDNA synthesis

Tissues were homogenized using a TissueLyser. Total RNA was extracted from tissues or cultured cells using peqGOLD Total RNA Kit (Peqlab, Erlangen, Germany) according to the manufacturer's protocol. The concentration of RNA was determined by photometric absorption at a wavelength of 260 nm and 292 nm in a ND-100 spectrophotometer (NanoDrop; Thermo Scientific). The ratio of 260/292 indicates the purity of samples. cDNA was generated from 1000ng RNA using the SCRIPT cDNA Synthesis Kit (Jena Bioscience) according to the manufacturer's protocol. cDNA samples were further diluted with 100 µl water before quantitative real-time PCR assay was performed.

Material and Methods

4.4.7.2 Quantitative PCR

Quantitative Polymerase chain reaction (PCR) assays on generated cDNA samples were performed by using specific primers targeting VSV or QuantiTect Primer assays (Qiagen) for other genes, along with SYBR Green (Roche) in a LightCycler® 480 System (Roche). 2 µl of diluted cDNA was used for analysis. A master mix composed of 1 μl of primer, 10 µl of SYBR green and 7 µl of distilled water (20 µl reaction per sample) was prepared. 96- well plates were used for qPCR. After sample addition, plates were sealed with transparent adhesive sealing film to avoid contamination. Hypoxanthine guanine phosphoribosyl transferase (HPRT) and β-actin were used as reference gene.

The following quantitative PCR primers targeting VSV-G or VSV-M (forward / reverse; 5’ to 3’) were used:

VSV-G: CAAGTCAAAATGCCCAAGAGTCACA/TTTCCTTGCATTGTTCTACAGATGG

VSV-M: TATGATCCGAATCAATTAAGATATG/GGACGTTTCCCTGCCATTCCGATG

The following program was used for quantitative PCR:

1. 95°C, 5 min

2. 95°C, 10 s

3. 55°C, 10 s

4. 72°C, 10 s + plate read

5. Repeat steps 2-5, 45 X

6. Melting curve: 65°C to 95°C

Increment of 0.5°C for 10 s

+ plate read

Protein analysis

Material and Methods

4.4.8.1 Protein extraction

To extract proteins under non-reducing conditions, whole-cell extracts from cultured cells were prepared in the presence of native lysis buffer [20 mM Tris-HCl (PH 7.5), 100 mM NaCl, 10% glycerol and 0.5% NP-40] supplemented with protease inhibitors (Complete Mini Protease Inhibitor Cocktail, Roche) and phosphatase inhibitors (PhosphoStop Phosphatase Inhibitor Cocktail, Roche). The lysed samples were incubated on ice for 30 min and centrifuged at 14000rpm for 20 min at 4°C. The supernatant containing the proteins was collected and stored at -80°C for further analyses.

To extract proteins under reducing conditions, whole-cell extracts from cultured cells or tissues were prepared in the presence of RIPA lysis buffer (Pierce) supplemented with protease inhibitors (Complete Mini Protease Inhibitor Cocktail, Roche) and phosphatase inhibitors (PhosphoStop Phosphatase Inhibitor Cocktail, Roche). The lysed samples were incubated on ice for 30 min with regular vortexing every 5 min and later centrifuged at 14000 rpm for 20 min at 4°C. The supernatant containing proteins was collected and stored at -80°C for further analyses.

4.4.8.2 Protein quantification

In order to measure protein concentrations, Bradford reagent (Roti-Quant®; Roth) and a spectrometric plate reader (Tecan) were employed. At first, a dilution series of BSA (along with 40 µl Bradford and 160 µl distilled water) were added into a 96-well plate to get a standard curve. Protein samples were diluted 1:10 (for tissues) or 1:5 (for cultured cells) and 2 μl of this mixture was used for analysis together with 40 µl Bradford and 160 µl distilled water. Samples were mixed thoroughly and measured using a spectrometric plate reader at 595 nm. The O.D was compared to the standard curve to obtain absolute protein concentrations for each sample.

4.4.8.3 Immunoprecipitation (IP) assay

For IP of heterologous-expressed Flag-tagged PGAM5 from HeLa cells, whole-cell extracts were prepared under non-reducing conditions and quantified by using Bradford assay. Anti-FLAG® M2 Magnetic Beads (Sigma) were pre-washed with TBS twice and

Material and Methods

adding into the supernatants, which were collected from cell lysates. All samples were agitated using a roller shaker at 4°C overnight. Then, beads were collected and washed with native lysis buffer five times. Co-precipitated proteins were eluted using SDS–PAGE loading buffer and heated.

4.4.8.4 Western blot

Western blot was performed with 20-50 μg of protein sample. For samples extracted under non-reducing conditions, native loading buffer (Bio-Rad) was added into protein samples in a 1:1 ratio and kept at room temperature for 7 min before loading into pre-cast gels (Mini-PROTEAN® TGX) side-by-side with a pre-stained protein ladder (peqGOLD IV, Peqlab). For samples extracted under reducing conditions, NuPAGE™ LDS Sample Buffer (Thermo Fisher) was added into the protein samples at a 1:3 ratio and denatured at 95°C for 5 min before loading into pre-cast gels. The unit was placed in a Mini-protean Tetra system (Bio-Rad) and filled with 1X running buffer (10X running buffer, 1:10). The protein was separated at 200 V until the dye reached the bottom of the gel.

After the electrophoresis, gels were visualized using a ChemiDoc Imaging Systems (Bio- Rad) to visualize protein loadings and then transferred to a nitrocellulose membrane (Whatmann) using a Trans-Blot Turbo Transfer System (Bio-Rad). Before transfer, all components (blotting papers, nitrocellulose membrane and gel) were soaked in blotting buffer for at least 1 min. The components were stacked on the chamber in the following order: one set of blotting papers, nitrocellulose membrane, protein gel, blotting papers. Air bubbles were removed by gentle rolling over the stack with a Pasteur pipette. The blotting chamber was closed, and transfer was performed via a program targeting the proper MW proteins.

In order to avoid unspecific binding, the membrane containing proteins was washed once with TBST and blocked with 0.1 % Tween-20 and 5 % skimmed milk for 1 h at room temperature on a tube rotator. The membrane was incubated with primary antibodies at 4°C overnight followed by incubation with the respective HRP-linked secondary antibody at room temperature for 1h. Subsequently, the membrane was washed 3 times in TBST and visualized by adding Western Lightening Plus- ECL (Perkin Elmer) solution following the manufacturer’s instructions. The membrane was developed on a ChemiDoc Imaging Systems (Bio-Rad) for a few minutes, depending on signal intensity. For detection of more than one antibody on the same membrane, antibodies were stripped from the membrane using stripping buffer at room temperature for 10 min two times and washed with TBST

Material and Methods

twice before proceeding with blocking and incubation of next primary antibody overnight. Antibody against β-Actin or GAPDH was used as loading controls.

Table 7. List of antibodies used

Primary antibodies Secondary antibodies (HRP conjugated) 1. Anti-Human PGAM5 (Sigma; HPA036978) Anti-rabbit

2. Anti-Murine PGAM5 (Santa Cruz; sc-515880) Anti-mouse

3. Anti-Human MAVS (Cell Signaling Technology; 3993) Anti-rabbit

4. Anti-Murine MAVS (Santa Cruz; sc-365334) Anti-mouse

5. Anti-Human/Murine TBK1 (Cell Signaling Technology; Anti-rabbit 3504) 6. Anti-Human/Murine pTBK1(Cell Signaling Technology; Anti-rabbit 5483) 7. Anti-Human IRF3 (Cell Signaling Technology; 11904) Anti-rabbit

8. Anti- Murine IRF3 (Abcam; ab68481) Anti-rabbit

9. Anti-Human/Murine pIRF3 (Cell Signaling Technology; Anti-rabbit 4947) 10. Anti-Human STAT1 (Cell Signaling Technology; 14995) Anti-rabbit

11. Anti-Human pSTAT1 (Cell Signaling Technology; 9167) Anti-rabbit

12. Anti-Human SOCS3 (Cell Signaling Technology; 52113) Anti-rabbit

13. Anti-Human/Murine β-Actin (Abcam; ab49900) directly labelled

Table 8. Composition of buffers used in Western blot

10X running Blotting buffer 1X stripping 10X TBS 1X TBST

30 g Tris 10ml blotting 15 g Glycine 24.2 g Tris 1 L 1 xTBS 144 g Glycine buffer (Bio- 500 μl Tween- 80 g NaCl 1 % Tween- 10 ml 10% Rad), 20 1 L distilled 20 SDS 10ml ethanol, water

Material and Methods

990 ml distilled 30ml distilled 1 L distilled pH 7.6 water water water (pH 2.5)

4.4.8.5 Semi-denaturing detergent agarose-gel electrophoresis (SDD-AGE) assay

SDD-AGE was performed as previously published (Hou et al., 2011). In brief, crude mitochondria were isolated from MEFs by using Mitochondria Isolation Kit for Cultured Cells (Thermo Fisher) following the manufacturer’s instructions. Suspended in sample buffer (0.5 x TBE, 10 % glycerol, 2 % SDS, and 0.0025 % bromophenol blue) and loaded onto a vertical 1.5 % agarose gel. After electrophoresis for 40 min in running buffer (1 x TBE and 0.1 % SDS) with a constant voltage of 100 V at 4 °C, western blotting was described.

Histological analysis

4.4.9.1 Tissue collection and preparation

For histological analyses freshly isolated tissue pieces were snap-frozen in liquid nitrogen and stored at -80 ° C or fixed in Histofix (4 % PFA) at room temperature for 24 h. Frozen tissues were embedded and fixed in Tissue-tek® and cut into 8 μm thin sections by a Cryostat on Superfrost microscopic glass slides. The sections were processed for staining immediately or stored at -80° C for later use. The samples from histo-fixed cassettes were dehydrated and embedded in liquid Paraffin (a routine procedure carried out by Dept. of Pathology, Erlangen). The waxed tissue was cut into 3 μm sections and collected on a water bath maintained at 40 °C. Thereafter, the slides were deparaffinized at 60 ° C and incubated in Roti Histol for three times, 5 min each. This was followed by rehydration in descending ethanol series (100 %, 96 %, 70 % v/v) 5 min each. The slides were then used for further histochemical staining.

4.4.9.2 H&E

Hematoxylin and eosin staining were used as the primary method to study basic morphology and tissue architecture. Rehydrated tissue sections were subjected to

Material and Methods

hematoxylin solution for 5 min followed by rinsing under tap water for 10 min. Then they were incubated in 1 % Eosin solution for 5 min followed by dehydration in ascending ethanol concentrations (70 %, 96 %, 100 %, 5 min each). After washing with Roti Histol, the sections were mounted with Entellan and coverslips.

4.4.9.3 Immunohistochemistry

Immunohistochemistry was performed by using the TSA kit (Perkin Elmer) which is based on Tyramide signal amplification technology. This assures amplification of fluorescent signals in immunochemistry, without losing resolution or increasing unspecific signals. Paraffin sections were prepared as in 4.4.9.1 and the samples were blocked for endogenous peroxidase activity by incubating in Methanol / H2O2 solution for 20 min. The slides were then subjected to antigen retrieval in either Tris / EDTA or Citrate buffer for 20 min at 500W in a microwave oven. After cooling the slides to room temperature, the tissue slices were circled by a hydrophobic pen to minimize reagent usage and prevent spillover of antibodies. The sections were incubated in Avidin / Biotin solution (Vector laboratories) for 15 min to block the endogenous cellular biotin. After washing in TBST, unspecific binding sites were blocked with protein block (Roti-block, Roth) in a 1:10 dilution with TBST / 2 % BSA for 15 min. The sections were incubated overnight with primary antibodies at 4°C in a humidity chamber. In order to detect the primary antibodies, on the following day, sections were incubated with biotinylated secondary antibodies diluted in TBS (1:1000) for 1 h at room temperature. Thereafter sections were washed and developed with the TSA cyanine / Fluroscein 3 system according to the manufacturer’s instructions or with streptavidin conjugated Dylight 488 / 555. Nuclei were counter stained with Hoechst (1:500 in TBS) for 8 min. The sections were dried and mounted with fluorescent mounting medium (Vectashield Laboratories).

Table 9. The composition of different solutions used in immunohistochemical staining

Methanol / H2O2 Tris / EDTA buffer Citrate buffer 11.5 ml distilled water, 1.21 g Tris base, 9 ml 0.1 M Citric acid,

1.5 ml of 30% H2O2, 0.37 g EDTA, 41 ml 0.1 M Sodium citrate 50 ml Methanol. 1 L distilled water, solution, 500 μl Tween 20, 450 ml distilled water,

Material and Methods

pH 9. pH 6.

Table 10. List of antibodies used in immunohistochemical staining

Antibody Catalog no. Dilution Supplier Anti-PGAM5 HPA036979 1:500 Sigma Anti-rabbit IgG (H+L) 7074 1:1000 Dianova Biotin conjugated

Immunocytochemistry

Immunohistochemistry and immunocytochemistry both utilize antibodies to provide visual details about protein abundance, distribution and localization. These terms are often confusing and are sometimes mistakenly used interchangeably. As their names imply, immunohistochemistry uses tissue sections, either paraffin embedded or frozen, whereas immunocytochemistry refers to the staining of isolated or cultured intact cells.

Cells grown on 8-well culture slides (Falcon) were cultured overnight and then fixed with 4% paraformaldehyde for 10 min at room temperature. Permeabilisation was further applied for nuclear protein staining via using 0.1% Triton X-100. After washing with PBS, cells were incubated in blocking buffer for 30 min at room temperature. Cells were incubated overnight with primary antibodies at 4°C in a humidity chamber. In order to detect the primary antibodies, on the following day, cells were incubated with fluorochrome-conjugated secondary antibody diluted in antibody dilution buffer for 1–2 h at room temperature in the dark. Nuclei were counter stained with Hoechst (1:500 in TBS) for 8 min. The slides were dried and mounted with fluorescent mounting medium (Vectashield Laboratories).

Table 11. The composition of different solutions used in immunocytochemistry staining

Blocking buffer Antibody dilution buffer PBS 0.4 g BSA, 0.8 g BSA, 80 g NaCl 4 ml FCS 4 ml FCS 2 g KCl, 36 ml PBS 36 ml PBS 14.4 g Na2HPO4 · 2H2O 2.4 g KH2PO4 pH 6.8

Material and Methods

1 L distilled water,

Table 12. List of antibodies used in immunocytochemistry staining

Primary Antibody/reagents Catalog no. Dilution Supplier Anti-TOMM20 ab56783 1:500 Abcam Anti-PGAM5 HPA036978 1:500 Sigma Alexa Fluor® 555 Donkey anti-rabbit 406412 1:500 Biolegend IgG (minimal x-reactivity) Antibody Alexa Fluor® 488 Goat anti-mouse 405319 1:500 Biolegend IgG (minimal x-reactivity) Antibody

4.4.10.1 TUNEL staining

Terminal deoxynucleotidyl transferase dUTP nick end-labeling (TUNEL) is a method for detecting DNA fragmentation by labeling the 3‘- hydroxyl termini in the double-strand DNA breaks. It has been widely used to identify and quantify apoptotic cells, or to detect excessive DNA breakage in individual cells.

Paraffin sections were prepared as in 4.4.9.1 and the samples were blocked for endogenous peroxidase activity by incubating in Methanol / H2O2 solution for 20 min. The slides were then subjected to antigen retrieval in Tris / EDTA for 20 min at 500 W in a microwave oven. After cooling of the slides to room temperature, they were circled by a hydrophobic pen and incubated with TUNEL reaction mixture (50 µl of Enzyme solution+ 450 µl Label solution) in 37 °C for 1 h. Thereafter sections were washed and nuclei were counter stained with Hoechst (1:500 in TBS) for 8 min. The sections were dried and mounted with fluorescent mounting medium (Vectashield Laboratories).

Statistical analysis

Significances were determined using the two-tailed student’s t-test. Differences were considered significant at * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001; N.S. = not significant (p>0.05). All data shown throughout the thesis are displayed with mean + standard deviation (S.D). Data are representative of at least three separately conducted experiments.

Results

5 Results

5.1 PGAM5-MAVS interaction regulates TBK1/ IRF3 dependent antiviral responses

Intracellular poly(I:C) delivery induces the formation of PGAM5 multimers

The innate immune system is the first line of protection from pathogen invasion (Kawai and Akira, 2008). Regarding to viral infections caused by RNA virus, Toll-like receptors (TLRs) and RIG-I-like receptors (RLRs) play crucial roles via activating various pathways to stimulate innate immune responses. RLRs such as RIG-I and melanoma differentiation associated protein 5 (MDA5) are major cytoplasmic viral RNA sensors (Wu and Chen, 2014). Viral RNA activates RIG-I or MDA5 and transduces the signal through the mitochondrial antiviral-signaling protein (MAVS) to induce the production of type I IFNs and other pro-inflammatory cytokines. The production of type I IFNs is a critical step in antiviral defense and adaptive immunity to clear pathogens (Cao, 2016). The aim of this thesis was to identify the functional role of PGAM5 in antiviral responses.

To analyze the functional roles of PGAM5 after viral challenges, poly(I:C) was initially used to mimic RNA virus infection in HeLa cells. Cells were challenged with poly(I:C) intracellularly for activating RLR signaling or extracellularly for activating TLR3 signaling. Study have demonstrated that intracellular poly(I:C) stimulation induces massive IFN expression in HeLa cells (Kuo et al., 2013), therefore capacity of poly(I:C) in activating TLR3 signaling was tested in HeLa cells. Cells were stimulated with poly(I:C) extracellularly and the protein levels of TLR3 were analyzed in Hela cells, HCT116 cells which express TLR3 were used as a positive control.

The protein level of TLR3 was detectable in HCT116 cells while not in HeLa cells (Figure 6A), raising the question whether extracellular poly(I:C) could activate the TLR3 pathway in HeLa cells. Interestingly, an up-regulation of TLR3 mRNA levels was detect in cells treated with extracellular poly(I:C) compared to mock treated cells (Figure 6B), suggesting that TLR3 is expressed in low levels in HeLa cells. In order to alternative test extracellular poly(I:C) can active TLR3 signaling in HeLa cells, cells were stimulated with poly(I:C) for 0- 24h and detected the phosphorylation of STAT1. A rapid induction of phospho-STAT1 was identified in HeLa cells following extracellular poly(I:C) stimulation while the amount of total STAT1 protein remains unchanged (Figure 6C), suggesting that extracellular poly(I:C) treatment was able to activate TLR3 signaling pathway in HeLa cells. Taken together, both extracellular and intracellular poly(I:C) treatment were capable of activating signaling pathways in HeLa cells.

Results

Figure 6. Extracellular poly(I:C) treatment activated the TLR3 pathway in HeLa cells in different time points. (A) Immunoblot analysis of TLR3 in lysates of extracellular poly(I:C) stimulated HeLa and HCT116 cells. β-actin served as a loading control. (B) mRNA expression of TLR3 in HeLa cells treated with vehicle (Mock) or 50 µg/ml extracellular poly(I:C) (p.IC-Ex). Experiments were performed three times and representative data are shown. Data are presented as mean + SD and student’s t-test was used for statistical calculation. **P<0.01. (C) Immunoblot analysis of STAT1, p-STAT1 and PGAM5 in lysates of extracellular poly(I:C) stimulated HeLa cells at different time points. β-actin served as a loading control.

To analyze the dynamics and location of PGAM5 after viral challenges, HeLa cells were challenged with intracellular or extracellular poly(I:C) and stained with PGAM5. The outer mitochondrial membrane protein Tomm20 was co-stained to locate the mitochondria. Confocal fluorescence microscopy revealed that PGAM5 staining overlapped with the mitochondrial marker Tomm20 in mock and poly(I:C) treated HeLa cells (Figure 7), suggesting that PGAM5 remains located in the mitochondria following stimulation. Interestingly, PGAM5 appeared to form aggregates in response to intracellular poly(I:C) stimulation (arrowhead, Figure 7). In contrast, extracellular poly(I:C) treatment did not

Results

induce visible PGAM5 aggregates in HeLa cells (Figure 7), suggesting that PGAM5 may forms aggregates in cells challenged with intracellular RNA.

Figure 7. Intracellular poly(I:C) treatment induced visible PGAM5 aggregates in HeLa cells. Confocal micrographs of HeLa cells stimulated with 50 µg/ml extracellular poly(I:C) (pIC-Ex) or 1 µg/ml intracellular poly(I:C) (pIC-In) for 8h and stained with the following antibodies: anti-Tomm20 and anti-PGAM5. Hoechst was used to stain the nucleus. Extracellular poly(I:C) was added directly into the medium. Intracellular poly(I:C) was transfected into the cells. Arrowheads indicate PGAM5 aggregates.

Several studies have indicated that PGAM5 forms multimers in response to different stresses (Ma et al., 2020; Tipton et al., 2018; Wilkins et al., 2014), therefore a hypothesis was made that the aggregates in cells challenged with intracellular RNA were made of PGAM5 multimers. To address the formation of PGAM5 multimers, intracellular poly(I:C) treated HeLa cells were lysed in non-reducing conditions and analyzed by SDS-PAGE

Results

(Figure 8A). Given the fact that the PGAM5 antibody used in here is a polyclonal antibody generated from rabbit, it would also detect some non-specific bands, which may interfere in PGAM5 multimer identification. To precisely identify PGAM5 multimers, cell lysates were used from PGAM5 CRISPR/Cas9 knockout cells as control to exclude non-specific bands. Several specific bands with a molecular weight of 32 kDa or multiples thereof were detected, indicating that PGAM5 exists not only as monomer, but also as dimers ((PGAM5)2) and PGAM5 multimers ((PGAM5) n). The levels of PGAM5 dimers and multimers in blots were used to indicate PGAM5 multimerization. Interestingly, PGAM5 dimers and multimers, although present at the steady state, were markedly increased after intracellular poly(I:C) treatment, indicating that intracellular RNA treatment induces PGAM5 multimerization. Of note, total PGAM5 protein levels of samples, which were lysed in RIPA buffer and separated by SDS-PAGE remained constant before and after intracellular poly(I:C) treatment, suggesting that intracellular RNA treatment did not impact PGAM5 total protein levels. β-Mercaptoethanol (BME) treatment partly reduced PGAM5 multimerization (Figure 8B), suggesting that PGAM5 multimer formation may partly rely on disulfide bonds. Collectively, data above indicated that the presence of intracellular RNA leads to PGAM5 multimer formation and localization at mitochondrial aggregates.

Figure 8. Intracellular poly(I:C) treatment induced PGAM5 multimers formation in HeLa cells. (A) SDS-PAGE analysis of PGAM5 multimers in lysates of HeLa cells under non-reducing or reducing conditions. HeLa cells were stimulated with 1 µg/ml intracellular poly(I:C) for 8h. Cell lysates of PGAM5 CRISPR/Cas9 knockout cells were used as control

Results

to exclude non-specific bands. N.S, non-specific band. (B) SDS-PAGE analysis of PGAM5 multimers in HeLa cell lysates treated with or without β-mercaptoethanol (BME at 35 mM), native lysis buffer was used for cell lysis. N.S, non-specific band.

PGAM5 deficiency attenuates intracellular poly(I:C)-induced IFNβ expression

As shown before, intracellular poly(I:C) treatment induced PGAM5 multimer formation in HeLa cells. It indicated that PGAM5 may have a functional role in antiviral responses induced by intracellular poly(I:C). Therefore, PGAM5 deficient HeLa cells which were generated by using the CRISPR/Cas9 technology were used. Cells overexpressing PGAM5 were used as positive controls to confirm the specificity of PGAM5 staining (Figure 9A). Confocal fluorescence microscopy revealed that PGAM5 staining overlapped with the mitochondrial marker Tomm20 in WT HeLa cells while PGAM5 knockout cells did not show PGAM5 expression. Expectedly, cells overexpressing PGAM5 showed very strong staining of PGAM5 (arrows, Figure 9A), suggesting that PGAM5 staining is highly specific and PGAM5 knockout HeLa cells had been successfully generated. Furthermore, PGAM5 deficiency in knockout cells was confirmed by western blot (Figure 9B).

Results

Figure 9. Generation of PGAM5 knockout HeLa cells. (A) Confocal microscopy imaging of WT and PGAM5 knockout HeLa cells stained with the following antibodies: anti- Tomm20 and anti-PGAM5. Hoechst was used to stain the nucleus. PGAM5 knockout HeLa cells were transfected with Flag tag PGAM5 vectors for 48h and used as positive control. Arrows indicate PGAM5 overexpressing cells. (B) Immunoblot analysis of PGAM5 in lysates of WT and PGAM5 CRISPR/Cas9 knockout HeLa cells. β-actin served as a loading control.

Together with WT cells, PGAM5 knockout HeLa cells were challenged with multiple ligands including intracellular/extracellular poly(I:C), poly(dI:dC) and lipopolysaccharide (LPS). These ligands were used to activate signaling pathways, which could be triggered by viral infection. Intracellular poly(I:C) was used to activate RLR signaling, extracellular poly(I:C) was used to active TLR3 signaling, poly(dI:dC) was used to activate cGAS– STING signaling and LPS was used to activate TLR2/TLR4 signaling. In this study, IFNB expression was used to indicate IFN response. As expected, IFNB expression was induced by these ligands in WT cells. Interestingly, PGAM5 deficiency specifically attenuated IFNB expression induced by intracellular poly(I:C) but not extracellular poly(I:C) (Figure 10A), suggesting that intracellular RNA sensors RLR rather than membrane-bound TLR3 require PGAM5. To further confirm the diminished IFNB expression in PGAM5 deficient HeLa cells, the expression of IFI27 and CXCL10, two downstream genes of IFNβ

Results

signaling, were also measured in PGAM5 sufficient and deficient HeLa cells. As expected, loss of PGAM5 also decreased the expression of these two target genes in HeLa cells (Figure 10B).

Figure 10. PGAM5 deficiency attenuates intracellular poly(I:C)- induced IFNB expression in HeLa cells. (A) mRNA expression of IFNB in WT and PGAM5 CRISPR/Cas9 knockout HeLa cells treated with vehicle (Mock), 1 µg/ml intracellular poly(I:C) (p.IC- In), 50 µg/ml extracellular poly(I:C) (p.IC-Ex), 1 µg/ml intracellular poly(dI:dC), or 100 ng/ml LPS.(B) mRNA expression of IFI27 (left) and CXCL10 (right) in WT and PGAM5 CRISPR/Cas9 knockout HeLa cells stimulated with 1 µg/ml intracellular poly(I:C) for 8h. Experiments were performed three times and representative data are shown. Data are presented as mean + SD and student’s t-test was used for statistical calculation. **P<0.01 and ***P<0.001. N.S., not significant.

In order to further investigate whether this observation is specific for PGAM5 in HeLa cells, MEFs were isolated from WT and PGAM5−/− mice. Breeding cages with PGAM5+/− male and females were created in order to get WT and PGAM5−/− embryos from littermates. DNA analysis of embryotic tissues was performed to identify the genotyping of isolated MEFs (Figure 11A). Furthermore, PGAM5 deficiency in MEFs was confirmed at the protein levels via western blot and the mRNA levels via qPCR (Figure 11B).

Results

Figure 11. Isolation of PGAM5−/− MEFs. (A) Genotyping of PGAM5 via using embryonic tissue. (B) Immunoblot analysis of PGAM5 in lysates of WT and PGAM5−/− MEFs. β-actin served as a loading control. (C) mRNA expression of Pgam5 in WT and PGAM5−/− MEFs. Experiments were performed three times and representative data are shown. Data are presented as mean + SD and student’s t-test was used for statistical calculation. ***P<0.001.

Isolated WT and PGAM5−/− MEFs were stimulated with intracellular poly(I:C) and IFN responses were measured by using qPCR analysis of IFNβ and IFNβ target genes. Similar to the results obtained from the experiments performed with HeLa cells, PGAM5 deficient MEFs showed a significantly decreased expression of Ifnb and IFNβ-induced genes following stimulation with intracellular poly(I:C) but not when poly(I:C) was simply added into the medium (Figure 12A-C). Intracellular poly(I:C) has been described to activate RIG- I like receptors (Kato et al., 2008). In line with this finding, the loss of PGAM5 also impaired Ifnb expression induced by RNA mimic 5´pppdsRNA (Figure 12D), a specific synthetic ligand that activates the RIG-I pathway (Marq et al., 2011). In summary, these data showed that the effects of intracellular poly(I:C) stimulation are not specific to HeLa cells, implicating that PGAM5 may generally needed to regulate IFN responses.

Results

Figure 12. PGAM5 deficiency attenuates intracellular poly(I:C)-induced Ifnβ expression in MEFs. (A and B) mRNA expression of Ifnb (A), Ifit1 and Il6 (B) in WT and PGAM5−/− MEFs stimulated with 1 µg/ml intracellular poly(I:C) for 8h. (C and D) mRNA expression of Ifnb in WT and PGAM5−/− MEFs stimulated with 50 µg/ml extracellular poly(I:C) (C) or 1 µg/ml 5pppRNA (D) for 8h. Experiments were performed three times and representative data are shown. Data are presented as mean + SD and student’s t-test was used for statistical calculation. *P<0.05, **P<0.01 and ***P<0.001. N.S., not significant.

PGAM5 overexpression rescued IFNβ expression in PGAM5 deficient HeLa cells

To assess the possibility that attenuated IFNβ expression in PGAM5 deficient HeLa cells might occur due to off-target effects of PGAM5 gRNAs, PGAM5 vectors were transfected into PGAM5 knockout HeLa cells to reconstitute PGAM5 expression. Western Blot analyses were used to verify PGAM5 expression (Figure 13A). Overexpression of PGAM5 restored IFNB expression in PGAM5 deficient HeLa cells treated with intracellular poly(I:C) (Figure 13B). Considering that PGAM5 has been identified as a phosphatase (Wang et al., 2012), functional role of phosphatase activity of PGAM5 in IFN responses was further

Results

tested. PGAM5 expression vectors coding phosphatase-dead PGAM5 (PGAM5-105A) were transfected into PGAM5 knockout cells and IFNB expression was measured via qPCR assay (Figure 13B). Interestingly, exogenous expression of phosphatase-dead PGAM5 restored IFNB expression in PGAM5 deficient HeLa cells treated with intracellular poly(I:C). Moreover, cells expressing PGAM5 or phosphatase-dead PGAM5 showed similar IFNB expression in response to intracellular poly(I:C) treatment, suggesting that phosphatase activity of PGAM5 was not required for IFN responses.

Figure 13. PGAM5 overexpression rescued intracellular poly(I:C)-induced IFNB expression and its function is independent of its phosphatase activity. (A) Immunoblot of PGAM5 expression in lysates from HeLa cells transfected with full-length PGAM5 (P) vectors or phosphatase mutant PGAM5 (P105A) vectors. Lysates from empty vector transfected WT or PGAM5 CRISPR/Cas9 knockout HeLa cells were used as control. β-actin served as a loading control. (B) mRNA expression of IFNB in HeLa cells stimulated with or without intracellular poly(I:C) for 8h. WT+Flag, WT cells transfected with empty vector. gPGAM5-KO+Flag, PGAM5 CRISPR/Cas9 knockout cells transfected with empty vector. gPGAM5-KO+Flag PGAM5, PGAM5 CRISPR/Cas9 knockout cells transfected with full-length PGAM5 vectors. gPGAM5-KO+Flag PGAM5-105A, PGAM5 CRISPR/Cas9 knockout cells transfected with phosphatase-mutated PGAM5 vectors. Experiments were performed three times and representative data are shown. Data are presented as mean + SD and student’s t-test was used for statistical calculation. **P<0.01 and ***P<0.001.

As demonstrated above, full length PGAM5 and phosphatase-dead PGAM5 expression were equally able to rescue IFNB expression in PGAM5 deficient situation. The expression in HeLa cells was achieved by transfection of Flag-tagged PGAM5 vectors. To rule out potential side effects of the Flag-tag, these experiments were repeated with myc-tagged PGAM5 vectors. Since PGAM5 has two forms: a long form (PGAM5L) and a short form (PGAM5S), which are generated from alternative splicing within the C domain (Wang et al., 2012). In order to study whether high level of IFNB expression relies on the C domain, both PGAM5 forms were expressed in PGAM5 knockout HeLa cells (Figure 14A). Similar restored IFNB expression was observed in cells under both conditions (Figure 14B),

Results

suggesting that the C domain is not essential for PGAM5-dependent IFNB expression. In summary, PGAM5 functions as a mediator of intracellular RNA induced IFN response and its function is independent of its alternative splicing in the C domain.

Figure 14. PGAM5 overexpression rescued intracellular poly(I:C)-induced IFNB expression and its function is independent of its alternative splicing in the C domain. (A) Immunoblot of PGAM5 expression in lysates of PGAM5 CRISPR/Cas9 knockout HeLa cells transfected with full-length PGAM5 (P) vectors, long form PGAM5 (PL) vectors or short form PGAM5 (PS) vectors. Lysates from WT and PGAM5 knockout HeLa cells were used as control. β-actin served as a loading control. (B) mRNA expression of IFNB in HeLa cells stimulated with or without intracellular poly(I:C) for 8h. gPGAM5-KO + Myc-PGAM5, PGAM5 CRISPR/Cas9 knockout cells transfected with full-length PGAM5 vectors. gPGAM5-KO + Myc-PGAM5L, PGAM5 CRISPR/Cas9 knockout cells transfected with long form PGAM5 vectors. gPGAM5-KO + Myc-PGAM5S, PGAM5 CRISPR/Cas9 knockout cells transfected with short form PGAM5 vectors. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. **P<0.01 and ***P<0.001.

PGAM5 functions upstream of TBK1

As shown above, PGAM5 regulates intracellular poly(I:C)-induced IFN expression, which depends on the activation of the RIG-I signaling pathway (Kato et al., 2008). Given the fact that RIG-I transduces signals through MAVS and induces phosphorylation of IRF3 and TBK1, the effects of PGAM5 on RIG-I signaling were then investigated by detecting these key molecules following intracellular poly(I:C) stimulation. Despite that the total protein levels of TBK1 and IRF3 remained constant following the stimulation, a rapid induction of phosphorylation of TBK1 and IRF3 was detected in WT MEFs treated with intracellular poly(I:C). In contrast, PGAM5 deficiency markedly impaired phosphorylation of IRF3 and TBK1 in MEFs (Figure 15A, B). Of note, intracellular poly(I:C) treatment didn’t alter protein levels of MAVS and PGAM5.

Results

Figure 15. PGAM5 deficiency attenuates intracellular poly(I:C)-induced IRF3 and TBK1 phosphorylation. (A) Immunoblot of phosphorylated or total proteins in lysates of WT and PGAM5-/- MEFs stimulated with intracellular poly(I:C) for different time points. β- actin served as a loading control. (B) Relative analysis of pTBK1 (upper panel) and pIRF3 (lower panel) intensity. The total protein level of IRF3 or TBK1 was used for normalization.

In order to functionally test whether PGAM5 regulates IFN response via a TBK1/IRF3 dependent pathway, BX795, a TBK1-specific inhibitor, was used in this thesis. MEFs were incubated with or without BX795 and Ifnb expression was measured after intracellular poly(I:C) treatment. As expected, intracellular poly(I:C) treatment induced rapid Ifnb expression and PGAM5 deficient MEFs showed impaired expression of Ifnb (Figure 16A). Interestingly, the presence of BX795 greatly diminished Ifnb expression and led to a similar expression level of Ifnb between WT and PGAM5 deficient MEFs, suggesting that PGAM5 acts via a TBK1 dependent pathway in MEFs. Similar results were obtained when the experiments were repeated in HeLa cells (Figure 16B).

Results

Figure 16. PGAM5 acts via a TBK1 dependent pathway. (A) mRNA expression of Ifnb in MEFs stimulated with intracellular poly(I:C) in the presence or absence of TBK1 inhibitor BX795. DMSO was used as control treatment. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. ***P<0.001. N.S., not significant. (B) mRNA expression of IFNB in HeLa cells stimulated with intracellular poly(I:C) in the presence or absence of TBK1 inhibitor BX795. DMSO was used as control treatment. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. ***P<0.001. N.S., not significant.

Since PGAM5 deficiency attenuates intracellular poly(I:C)-induced TBK1 phosphorylation and PGAM5 acts via a TBK1 dependent pathway, PGAM5 may function upstream of TBK1. To verify this hypothesis, the phosphorylation levels of TBK1 were analyzed in PGAM5-reconstituted HeLa cells by western blot (Figure 17A, B). As anticipated, PGAM5 deficiency impaired while PGAM5 overexpression restored phosphorylation of TBK1 in response to intracellular poly(I:C) treatment. Interestingly, with any treatment, HeLa cells overexpressed with PGAM5 showed rapid induction of phospho-TBK1. Similar results were obtained when the experiments were repeated in HeLa cells overexpressed phosphatase-dead PGAM5. Taken together, this thesis indicated that PGAM5 functions upstream of TBK1 and that high-level expression of PGAM5 can be sufficient to induce TBK1 phosphorylation independent of its phosphatase activity.

Results

Figure 17. PGAM5 acts via a TBK1 dependent pathway. (A) Immunoblot from lysates of PGAM5 CRISPR/Cas9 knockout HeLa cells transfected for 48h with WT PGAM5 vectors or phosphatase-mutant PGAM5 vectors. Lysates from empty vector transfected WT and PGAM5 CRISPR/Cas9 knockout HeLa cells were used as control. β-actin served as a loading control. (B) Relative analysis of pTBK1 intensity. Total protein levels of TBK1 were used for normalization and intensities were relative to pTBK1 intensity in WT cells.

The formation of PGAM5 multimers and MAVS aggregates are independent of each other

In order to better understand the role of PGAM5 in the signaling pathway upstream of TBK1/IFR3, this thesis then focused on MAVS. MAVS and PGAM5 are both located on the mitochondrial membrane and MAVS has been shown to be vital for activation of TBK1/IRF3 responses via forming functional aggregates (He et al., 2017; Hou et al., 2011; Liu et al., 2015; Seth et al., 2005). Despite fact that PGAM5 deficiency did not impair the protein levels of MAVS (Figure 15A), a semi-denaturing detergent agarose-gel electrophoresis (SDD-AGE) assay was performed to detect MAVS aggregates in cells stimulated with intracellular poly(I:C) (Figure 18). As expected, intracellular poly(I:C) treatment induced formation of MAVS aggregates. Importantly, equal amount of MAVS aggregates were detect in WT and PGAM5−/− cells, suggesting that PGAM5 promotes IFN expression without directly regulating MAVS aggregation.

Results

Figure 18. PGAM5 does not regulate MAVS aggregation. (A) Crude mitochondrial extracts were prepared from MEFs stimulated with intracellular poly(I:C) for 8h. Aliquots of the extracts were analyzed by SDD-AGE. MEF cell lysates were analyzed by using SDS-PAGE as control. β-actin served as a loading control.

To reciprocally test if there is a functional role of MAVS on regulating PGAM5 multimerization, MAVS knockout HeLa cells were generated via using CRISPR/Cas9 technology (Figure 19A, upper panel). As expected, MAVS deficiency in HeLa cells significantly diminished IFNB expression, confirming the key role of MAVS in the RIG-I pathway (Figure 19A, lower panel). To detect PGAM5 multimers in WT and MAVS knockout HeLa cells, cells were lysed in non-reducing conditions and analyzed by SDS- PAGE (Figure 19B). Consistent with previous results, intracellular poly(I:C) treatment induced significant increase of PGAM5 multimers in WT cells. Surprisingly, MAVS deficiency did not block PGAM5 multimerization, indicating that this step is independent of MAVS.

Results

Figure 19. MAVS does not regulate PGAM5 multimerization. (A) Upper panel: Immunoblot analysis of MAVS in lysates of WT and MAVS CRISPR/Cas9 knockout HeLa cells. β-actin served as a loading control. Lower panel: mRNA expression of IFNB in HeLa cells treated with vehicle (Mock) or 1 µg/ml intracellular poly(I:C) for 8h. ***P<0.001. (B) SDS-PAGE analysis of PGAM5 multimers in lysates of WT HeLa cells and MAVS knockout HeLa cells under non-reducing conditions. HeLa cells were stimulated with 1 µg/ml intracellular poly(I:C) for 8h. β-actin served as a loading control.

Taken together, these data from PGAM5 knockout cells and MAVS knockout cells indicated that the formation of PGAM5 multimers and MAVS aggregates are independent of each other.

PGAM5 interacts with MAVS

Given the above findings, we next studied whether PGAM5 and MAVS interact upon stimulation with intracellular poly(I:C). Flag-tagged PGAM5 vectors were transfected into PGAM5 deficiency HeLa cells and immunoprecipitation was conducted using anti-FLAG M2 Magnetic Beads. Although PGAM5 deficiency did not impact the protein levels of MAVS during intracellular poly(I:C) stimulation (Figure 15A), an interaction between PGAM5 and MAVS was identified (Figure 20). Interestingly, the interaction was detected in the steady state and was increased after intracellular poly(I:C) treatment, suggesting that the interaction between PGAM5 and MAVS may serve as an important factor in

Results

activating IFN pathway. Thus, our data demonstrated that PGAM5 interacts with MAVS and regulates TBK1/IRF3 dependent IFN expression during intracellular RNA treatment.

Figure 20. PGAM5 interacts with MAVS. Immunoprecipitation and immunoblotting of PGAM5 CRISPR/Cas9 knockout HeLa cells transfected for 48 h with (+) or without (-) full- length PGAM5 vectors followed by intracellular poly(I:C) (pIC-In) stimulation at indicated time points. β-actin served as a loading control.

Taken together, this thesis indicated that PGAM5 and MAVS are both necessary for efficient IFN induction and that their physical interaction occurs functionally independent of their oligomerization.

PGAM5 regulates VSV-induced IFNβ expression and inhibits VSV replication

Poly(I:C) is a well-established ligand and is widely used to mimic RNA virus infection and to induce IFN expression. As data shown above that PGAM5 is required for intracellular poly(I:C) induced IFN response, it is highly possible that PGAM5 may regulate anti-RNA viral responses. Vesicular stomatitis virus (VSV), a RNA virus that active RIG-1 signaling via intracellular viral RNA, was used in this thesis to test functional role of PGAM5 in antiviral responses. Following VSV infection, induction of IRF3 phosphorylation was measured to detect the activation of RIG-1 pathway. As expected, VSV infection induced pronounced phosphorylation of IRF3 in WT MEFs (Figure 21A and B). In contrast, PGAM5 deficient MEFs showed impaired levels of phospho-IRF3.

Results

Figure 21. PGAM5 deficiency attenuates VSV-induced IRF3 phosphorylation. (A) Immunoblot of phosphorylated or total proteins in lysates of WT and PGAM5-/- MEFs infected with VSV for the indicated hours. β-actin served as a loading control. (B) Relative analysis of pIRF3 intensity as shown in (A). The total protein level of IRF3 was used for normalization and intensities were relative to pIRF31 intensity in WT cells.

In line with findings above, VSV infection significantly induced expression of Ifnb, Ifit1 and Il6 in WT cells. In contrast, PGAM5 deficient MEFs showed decreased expression of these genes (Figure 22A and B). In order to study whether PGAM5 may have an impact on the replication of VSV, viral load was measured via using qPCR primers targeting VSV-G and VSV-M (Figure 22C), genes coding VSV G / M protein. As expected, the expression levels of these two genes were relative high in VSV-infected cells but were under-detection in non-infected cells, indicating high specify of primers. Interestingly, PGAM5 deficient MEFs showed a significantly higher virus load as compared to WT MEFs exposed to VSV, suggesting that PGAM5 is an important regulator of antiviral immunity.

Results

Figure 22. PGAM5 deficiency attenuates VSV-induced Ifnb expression. (A) and (B) mRNA expression of Ifnb (A) or Ifit1 and Il6 (B) in WT and PGAM5-/- MEFs infected with VSV. (C) qPCR analysis of VSV-G and VSV-M in WT and PGAM5-/- MEFs infected with VSV for 24 h. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. *P<0.05, **P<0.01.

Results

5.2 PGAM5 regulates IFN-stimulated genes (ISG) expression

PGAM5 regulates extracellular poly(I:C)-induced STAT1 phosphorylation.

The TLR3 and RLR pathways, which are activated by viral RNA are capable of inducing the expression of IFNs and subsequently a wide array of IFN-stimulated genes (ISG). The ISG have direct antiviral effects and are likely to influence the outcome of infection by many human viruses (Duncan et al., 2020). Upon viral detection and subsequent IFN production, IFN proteins bind to their special receptors and therefore initiate a signaling cascade through the JAK-STAT pathway (Stark and Darnell, 2012). The activation of this pathway leads to the phosphorylation of IFN receptor chains and reposition of STAT protein (Durham et al., 2019). As a consequence, STATs are phosphorylated and released from the receptors, phosphorylated STATs further form homodimers or heterodimers and translocate into nucleus (Durham et al., 2019). The nuclear localization of STATs drives ISG expression.

There are seven functional STAT proteins in mammals, STAT1, STAT2, STAT3, STAT4, STAT5a, STAT5b, and STAT6 (Alunno et al., 2019). Each STAT protein has its unique functions in multiple biological processes. STAT1 is a transcription factor predominately involved in the signal transduction by either type I, type II, or type III IFNs, suggesting that STAT1 is one of the most important STATs (Goropevsek et al., 2017; O'Shea et al., 2015; Villarino et al., 2015).

ISG is a group of genes whose expression is stimulated by IFNs. IFNs and signaling molecules such as JAKs, STATs which are present in steady state and upregulated upon IFN stimulation are also ISG (Schneider et al., 2014). The antiviral effects of ISG effects at different stages of the viral life cycle, from entry, replication, assembly to release of viral particle to cells. This functions make ISG a key player in antiviral defense.

As shown above, PGAM5 specifically regulates IFNB expression induced by intracellular poly(I:C) but not by extracellular poly(I:C). Not surprisingly, in line with the diminished IFNB expression seen in PGAM5 deficient cells, loss of PGAM5 also decreased the expression levels of IFI27 and CXCL10, two ISG. However, it is currently unknown whether PGAM5 might be similarly involved in regulating ISG expression induced by extracellular poly(I:C), or more specifically, whether PGAM5 is involved in regulating ISG expression induced by IFNs.

To elucidate a functional contribution of PGAM5 in response to extracellular poly(I:C) stimulation, WT and PGAM5 deficient HeLa cells generated via using the CRISPR/Cas9

Results

technology, were stimulated with different concentrations of poly(I:C) extracellularly. As expected, STAT1 phosphorylation levels were significantly induced by extracellular poly(I:C) stimulation in WT cells. Interestingly, PGAM5 deficiency attenuated STAT1 phosphorylation when different concentrations of poly(I:C) were added to medium (Figure 23A). Furthermore, to assess the possibility that attenuated levels of STAT1 phosphorylation might occur due to altered phosphorylation kinetics under a PGAM5 deficiency situation, HeLa cells were stimulated with extracellular poly(I:C) at various time points to measure the STAT1 phosphorylation levels during the entire stimulation process (Figure 23B). In both WT and PGAM5 knockout HeLa cells, which were treated with extracellular poly(I:C), phosphorylated STAT1 reached a peak at 4h post stimulation and went back to steady state levels at 24h post stimulation, indicating that PGAM5 deficiency does not affect the kinetics of STAT1 phosphorylation process. Interestingly, in line with the results shown in Figure 23A, PGAM5 deficient cells showed a significant decreased levels of STAT1 phosphorylation following stimulation with extracellular poly(I:C) at various time points.

Results

Figure 23. PGAM5 deficiency attenuates extracellular poly(I:C)-induced STAT1 phosphorylation. Immunoblot of phosphorylated or total proteins in lysates of WT and PGAM5 CRISPR/Cas9 knockout HeLa cells stimulated with extracellular poly(I:C) using different concentrations (A) and at different time points (B).

Consistent with the findings above, qPCR analysis of some ISG such as IFIT1, IFI27 and IL6 also supported the notion that PGAM5 regulates extracellular poly(I:C)-induced STAT1 phosphorylation (Figure 24). As expected, the mRNA levels of these molecules were significantly up-regulated after extracellular poly(I:C) treatment. Interestingly, PGAM5 deficiency led to a decreased IFIT1, IFI27 and IL6 expression as compared to WT HeLa cells.

Results

Figure 24. PGAM5 deficiency attenuates extracellular poly(I:C)-induced IFIT1, IFI27 and IL6 expression. mRNA expression of IFIT1(A), IFI27(B) and IL6 (C) in WT and PGAM5 CRISPR/Cas9 knockout HeLa cells stimulated with 50 µg/ml extracellular poly(I:C)) for 8h. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. **P<0.01.

Given the fact that STAT1, IFIT1, IFI27 and IL6 all belong to the ISG family and PGAM5 deficiency attenuates these transcripts upon extracellular poly(I:C) stimulation, a conclusion was made in this thesis that PGAM5 regulates extracellular poly(I:C)-induced ISG responses in HeLa cells.

PGAM5 deficiency up-regulates extracellular poly(I:C)-induced SOCS3 expression.

The suppressor of cytokine signaling 3 (SOCS3) which belongs to the SOCS family, is a key regulator of the JAK-STAT pathway, ensuring that activation of the pathway for essential cellular processes is spatially and temporally controlled to prevent pathology (Durham et al., 2019). It has been shown that SOCS3 regulates STAT1 phosphorylation and downstream gene expression (Tsai et al., 2011; Zhang et al., 2017b). To investigate whether the effect of PGAM5 on regulating the JAK-STAT pathway involves SOCS3, SOCS3 expression was measured under steady state conditions and upon extracellular poly(I:C) treatment (Figure 25A). As expected, extracellular poly(I:C) treatment induced SOCS3 expression in WT HeLa cells. In contrast, PGAM5 deficient HeLa cells showed increased SOCS3 expression. Interestingly, a significant increase of SOCS3 expression was observed in PGAM5 deficient HeLa cells, even in the steady state (Figure 25A).

Results

The latter finding was supported by performing western blot assays for SOCS3 protein (Figure 25B). While PGAM5 deficient HeLa cells showed a significant increase of SOCS3 protein levels following extracellular poly(I:C) treatment, protein levels in WT HeLa cells remained relatively constant (Figure 25B). Only a slight increase of SOCS3 protein levels was observed in WT HeLa cells. In PGAM5 deficient HeLa cells, the protein level of SOCS3 reached a peak at 2h post-treatment and lasted for more than 24h post-treatment. Compared to WT HeLa cells, PGAM5 deficiency increases the SOCS3 levels before and after extracellular poly(I:C) treatment (Figure 25B).

Taken together, PGAM5 deficiency up-regulates extracellular poly(I:C)-induced SOCS3 expression.

Figure 25. PGAM5 deficiency up- regulates SOCS3 gene expression and protein levels. (A) mRNA expression of SOCS3 in WT and PGAM5 CRISPR /Cas9 knockout HeLa cells stimulated with 50 µg/ml extracellular poly(I:C)) for 8h. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. ***P<0.001. (B) Immunoblot of SOCS3 in lysates of WT and PGAM5 CRISPR/Cas9 knockout HeLa cells stimulated with extracellular poly(I:C) for different time points.

PGAM5 positively regulates IFN-stimulated STAT1 phosphorylation and downregulates SOCS3 expression.

Results

Previous studies have clearly shown that ISG are induced by extracellular poly(I:C) due to the production of IFNs (Li et al., 2005). In order to further investigate the effect of PGAM5 in regulating ISG expression, three different types of IFNs (IFNß, IFNγ and IFNλ) were used to stimulate WT and PGAM5 knockout HeLa cells at different time points. For cells treated with IFNß, a significant induction of IFI27 and IFIT1 expression was observed in WT HeLa cells. In contrast, PGAM5 deficient HeLa cells showed impaired IFI27 and IFIT1 expression (Figure 26A). Interestingly, upon stimulation with IFNγ, PGAM5 deficiency significantly attenuated IFI27 and IFIT1 expression at an early time point (4h) but not at a late time point (Figure 26B). Notably, IFIT1 expression reached the peak at 4h post stimulation and almost decreased to steady state levels at 10h post stimulation, suggesting that IFNγ stimulation is attenuated at late time point (10h). This implies that PGAM5 deficiency impairs IFNγ-induced ISG expression. Regarding IFNλ, similar to the results obtained from the experiments performed with IFNß, PGAM5-deficient HeLa cells showed a significantly decreased expression of IFI27 and IFIT1 following stimulation with IFNλ (Figure 26C). In summary, PGAM5 functions as a mediator of different types of IFN- induced ISG responses.

Results

Figure 26. PGAM5 deficiency attenuates IFNs-induced IFI27 and IFIT1 expression. mRNA expression of IFI27 and IFIT1 in WT and PGAM5 CRISPR/Cas9 knockout HeLa cells (gPGAM5-KO) stimulated with IFNß (A), IFNγ (B), or IFNλ (C) for 4h or 10h. Experiments were performed three times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. **P<0.01, ***P<0.001.

As mentioned above, PGAM5 deficient cells showed a significant decrease of STAT1 phosphorylation following stimulation with extracellular poly(I:C). Studies were performed to further investigated whether PGAM5 could regulate IFN-induced STAT1 phosphorylation. Following a western blot assay, IFNß and IFNγ, but not IFNλ induced rapid phosphorylation of STAT1 (Figure 27A). Notably, PGAM5 deficiency significantly

Results

decreased the phosphorylation of STAT1 in response to the treatment with both IFNß and IFNγ. In agreement with the finding above, a strong up-regulation of SOCS3 was also detected in PGAM5 knockout HeLa cells compared to WT HeLa cells, under stimulation with IFNß, IFNγ and IFNλ (Figure 27B).

Figure 27. PGAM5 deficiency attenuates IFN-induced IFI27 and IFIT1 expression. (A) Immunoblot of phosphorylated or total proteins in lysates of WT and PGAM5 CRISPR/Cas9 knockout HeLa cells stimulated with IFNß, IFNγ or IFNλ. β actin was used as loading control. (B) mRNA expression of SOCS3 in WT and PGAM5 CRISPR/Cas9 knockout HeLa cells stimulated with IFNß, IFNγ or IFNλ for 4h and 10h. Experiments were performed 3 times and representative data are shown. Data are presented as mean +SD and student’s t-test was used for statistical calculation. **P<0.05, **P<0.01, ***P<0.001.

Collectively, this thesis demonstrates that the mitochondrial protein PGAM5 is important for expression of ISG genes. Following challenge of HeLa cells with the dsRNA-analog poly(I:C) extracellularly, PGAM5-deficient cells showed diminished phosphorylation of STAT1 and reduced expression of ISG as compared to WT cells. Moreover, PGAM5 deficient HeLa cells exhibited decreased phosphorylation levels of STAT1 and expression

Results

of ISG when challenged with three different types of IFNs. Mechanistically, PGAM5 deficiency significantly up-regulates SOCS3 expression.

5.3 PGAM5 is dispensable for poly(I:C)-induced small intestinal inflammation.

According to findings above from in vitro experiments, PGAM5 is an important regulator for IFNβ production and ISG expression in response to poly(I:C) stimulation. In a next step, this thesis further investigated the functional roles of PGAM5 under the treatment of poly(I:C) in vivo.

Intraperitoneal injection of poly(I:C) has been shown to induce pronounced small intestinal injury and inflammatory cytokine production in a time- and dose-dependent manner via activation of the TLR3 pathway (Gunther et al., 2015; Marafini et al., 2015; Zhou et al., 2007). Therefore, in order to test whether PGAM5 is involved in the phenotype induced by poly(I:C) in vivo, the established poly(I:C) model was employed in WT and PGAM5−/− mice.

Duodenum from WT and PGAM5−/− mice were collected and deficiency of PGAM5 was confirm in murine tissue (Figure 28). Duodenum sections were stained for PGAM5 to analyze its location and expression (Figure 28A). Fluorescence microscopy revealed that PGAM5 staining showed a strong signal in the WT tissue, whereas almost no signal was visible in the PGAM5−/− tissue, suggesting that the PGAM5 staining in the duodenum is highly specific. Interestingly, PGAM5 staining showed strong signals in the crypts instead in the villi, indicating that PGAM5 is mainly expressed in duodenal proliferating compartment. To further confirm these data, protein and mRNA levels of PGAM5 were analyzed via western blot and qPCR assays in duodenal tissues (Figure 28B and C). PGAM5 protein were detected in duodenal tissues from WT mice but not in the PGAM5−/− mice. Similar results were obtained by qPCR analysis. Taken together, data presented here demonstrated that PGAM5 is specially expressed in the epithelia of WT mice but not PGAM5−/− mice.

Results

Figure 28. Confirmation of PGAM5 deficiency in vivo. (A) Immunohistochemical staining for PGAM5 (red) in duodenal sections from WT and PGAM5−/− mice. (B) Immunoblot of PGAM5 proteins in lysates of duodenal tissues from WT and PGAM5−/− mice. β actin served as a loading control. (C) mRNA expression of Pgam5 in duodenum tissues from WT and PGAM5−/− mice. ***P<0.001.

To elucidate a functional contribution of PGAM5 in response to poly(I:C) in vivo, WT and PGAM5−/− mice were injected with poly(I:C) intraperitoneally. As expected, poly(I:C) administration induced massive tissue damage in the duodenum of WT mice (Figure 29A, upper left panel). Interestingly, similar damage was also observed in PGAM5−/− mice (Figure 29A, upper right panel). Furthermore, cell death in duodenal tissues was also detected by TUNEL staining (Figure 29A, lower panel). Similar to the results obtained from the histological staining, PGAM5 deficient duodena showed a similar level of TUNEL positive cells as compared to WT duodenum. In an agreement with the phenotype in vivo, Villin expression after poly(I:C) administration were comparable between WT and PGAM5−/− (Figure 29B, left panel). Given the in vitro data shown above, Ifnb expression

Results

was also detected after poly(I:C) treatment. Interestingly, PGAM5 deficiency did not impair Ifnb expression under these experimental conditions in vivo (Figure 29B, right panel). In summary, the data presented here demonstrated that PGAM5 is dispensable for poly(I:C)- induced small intestinal injury.

Figure 29. PGAM5 is dispensable for poly(I:C)-induced small intestinal injury. (A) Representative HE pictures (upper panel) and TUNEL staining (lower panel) of the duodenum of poly(I:C)-treated WT and PGAM5−/− mice. (B) mRNA expression of Villin and Ifnb in duodenum tissues from WT and PGAM5−/− mice after poly(I:C) injection.

Results

5.4 PGAM5 is dispensable for the development of experimental colitis.

Taking advantage of PGAM5 knockout mice model which was generated based on the knockout first strategy (Skarnes et al., 2011), PGAM5 fl/fl mice (PGAM5fl/fl) were generated from PGAM5−/− mice by crossing them to FLP mice. Then the PGAM5fl/fl mice were further crossed to Villin-cre mice to generate conditional intestinal epithelial cell (IEC)-specific PGAM5 knockout mice (PGAM5ΔIEC). Genotyping of PGAM5 flox and VillinCre was performed using ear biopsies (Figure 30A). To confirm the deficiency of PGAM5 in colonic epithelial cells, colon crypts were collected and analyzed via a western blot assay. PGAM5 protein was detected in PGAM5fl/fl mice while not in PGAM5ΔIEC mice (Figure 30B). Furthermore, immunohistochemically staining of PGAM5 was applied to analyze the dynamics and location of PGAM5 in colonic tissues (Figure 30C). In colon epithelial cells, fluorescence microscopy revealed that PGAM5 showed a strong staining in PGAM5fl/fl while almost no signal was visible in PGAM5ΔIEC tissue, suggesting that PGAM5 staining in the colon is highly specific. Despite the fact that PGAM5 signal was detected in the upper part of crypts, PGAM5 staining showed a stronger signal in lower parts of the crypts. Taken together, a conditional knockout mouse strain was generated which has no expression of PGAM5 in colon epithelial cells.

In the last few decades, dozens of different animal models of IBD have been developed and the dextran sodium sulfate (DSS)-induced colitis model is one of the most popular (Chassaing et al., 2014). DSS is a water-soluble, negatively charged sulfated polysaccharide with a highly variable molecular weight ranging from 5 to 1400 kDa. DSS was applied into drinking water to induce colitis in mice. Here, the DSS-induced colitis model was used to figure out whether PGAM5 plays a functional role in colitis development. It has been reported that the gut microbiota is a key factor in the pathogenesis of IBD (Carstens et al., 2019; Gao et al., 2018; Llewellyn et al., 2018; Paramsothy et al., 2017). To rule out the effects of different gut microbiota compositions in the experiments, littermates were used in all experiments.

Results

Figure 30. PGAM5 deficiency in colon epithelial cells from PGAM5 fl/fl VillinCre+ mice. (A) Genotyping of PGAM5 flox and VillinCre via using ear biopsies. (B) Immunoblot of PGAM5 proteins in lysates of colon crypt tissues from PGAM5 fl/fl VillinCre- (PGAM5fl/fl) and PGAM5 fl/fl VillinCre+ (PGAM5∆IEC) mice. β-actin served as a loading control. (C) Immunohistochemistry staining for PGAM5 in colon sections from PGAM5 fl/fl and PGAM5∆IEC mice.

To elucidate a functional contribution of PGAM5 in experimental colitis, DSS was applied into drinking water to induce acute colitis in PGAM5fl/fl and PGAM5ΔIEC mice. As expected, DSS administration induced massive bodyweight loss in PGAM5fl/fl control mice (Figure 31A). Of note, similar levels of bodyweight loss were also observed in PGAM5ΔIEC mice. Furthermore, colon damage was detected via endoscopic examination (Figure 31B) and histological staining of HE (Figure 31C). Consistent with the similar bodyweight loss observed between PGAM5fl/fl and PGAM5ΔIEC mice, PGAM5 deficient mice showed a

Results

similar level of tissue damage as compared to control mice (Figure 31B and C). Taken together, IEC-specific PGAM5 is dispensable for acute DSS-induced colitis.

Figure 31. PGAM5 is dispensable for DSS-induced colitis. (A) Bodyweight curve showing the body weight changes of indicated mice during DSS administration. 1.5% DSS was applied for 7 days, followed by drinking water. (B) Representative endoscopic pictures at day 7 of the colon from mice treated with 1.5% DSS for 7 days. (C) Representative histological pictures (HE) of colon from mice treated with 1.5% DSS for 7 days.

Discussion

6 Discussion

PGAM5 is an atypical protein phosphatase implicated in a number of functions within mitochondria, including mitochondrial homeostasis, mitophagy, and multiple cell death pathways. The functional role of PGAM5 has been studied in experiments which mitochondria were stressed. It has been shown that PGAM5 recruits and dephosphorylates DRP1 to promote mitochondrial fission and to overcome mitochondrial stress (Kang et al., 2015; Wang et al., 2012). At the same time, PGAM5 also stabilizes PINK1 or dephosphorylates FUNDC1 to positively regulate mitophagy during clearing damaged mitochondria (Liu et al., 2012; Lu et al., 2014; Sekine et al., 2016). Moreover, PGAM5 also regulates the anti-oxidative response by forming a tertiary complex with KEAP1 and NRF2 (Holze et al., 2018; Lo and Hannink, 2008). PGAM5 can be cleaved and released from mitochondria to activate Wnt signaling (Bernkopf et al., 2018). In summary, PGAM5 is a central player in regulating key pathways of cellular homeostasis.

Mitochondria are central hubs for cellular innate antiviral immunity, particularly in the RIG- I like pathway (Hou et al., 2011; Wen et al., 2012). This doctoral thesis mainly focused on the functional role of PGAM5 in regulating cellular responses to pathogenic challenges. In this thesis it could be shown that following a challenge of HeLa cells with the dsRNA- analog poly(I:C), PGAM5 oligomers and high levels of PGAM5 were found in mitochondrial aggregates. A direct interaction of PGAM5 with MAVS was demonstrated. In addition, PGAM5 deficient cells showed diminished expression of IFNβ and IFNβ target genes as compared to WT cells. Moreover, PGAM5 deficient MEFs exhibited decreased phosphorylation levels of IRF3 and TBK1 when challenged with poly(I:C) intracellularly. Furthermore, PGAM5-deficient MEFs revealed diminished IFNβ expression and increased VSV replication following infection with VSV. Besides, PGAM5 also regulated ISG expression via up-regulating SOCS3. Collectively, this study highlights PGAM5 as an important regulator for antiviral responses.

6.1 PGAM5 regulates IFNβ expression and antiviral response.

In this study, it was demonstrated that the mitochondrial protein PGAM5 is important for the antiviral cellular response. Following challenge of HeLa cells or MEFs with the double- stranded RNA-analog poly(I:C), PGAM5 deficiency specifically attenuated IFNβ expression induced by intracellular poly(I:C) but not when poly(I:C) was simply added to

Discussion

medium. Moreover, VSV was used to infect MEFs and the loss of PGAM5 decreased IFNβ responses and promoted VSV replication.

IFNβ expression can be induced by the activation of multiple intracellular pathways following viral or bacterial infection, and it is likely that the major driving forces for IFNβ production are viral infections (Nagarajan, 2011). IFNβ shares a common receptor IFNAR with IFNα, another type I IFN. Upon binding of the IFNβ to the IFNAR, the JAK-STAT signaling pathway is activated resulting in direct antiviral effects of IFNβ and expression of interferon-inducible genes (Fu et al., 1990; Qureshi et al., 1995; Stark and Darnell, 2012). Therefore, this thesis focused on IFNβ expression to identify the functional role of PGAM5 in antiviral responses.

Double-stranded RNA, which is present in many RNA viruses, induces robust immune responses via pattern recognition receptors, such as Toll-like receptor 3 (TLR3), and Rig- I. TLR3 is expressed either on the cell surface or in intracellular endosome compartments, while RIG-I is expressed in the cytosol. TLR3 preferentially recognizes viral RNA presented extracellularly while RIG-I expression in the cytoplasm allows the detection of actively replicating viruses (Blasius and Beutler, 2010; Schlee et al., 2007). In this thesis, experiments showed that PGAM5 promoted IFNβ expression only if poly(I:C) was delivered into the cells by transfection, suggesting that PGAM5 is involved in RIG-I mediated signaling. RIG-I has recently been demonstrated as an essential molecule in VSV-induced cytokine production (Crill et al., 2015). Indeed, when PGAM5-deficient cells were exposed to VSV, IFNβ-induction was diminished compared to WT cells, further supporting the conclusions above.

PGAM5 is a phosphatase and the phosphatase activity is crucial for signaling transduction (He et al., 2017; Wang et al., 2012). Besides, the alternative splicing of PGAM5 within the C domain generates two different forms of PGAM5, a long form (PGAM5L) and a short form (PGAM5S) (Wang et al., 2012). Interestingly, in cells overexpressing phosphatase- dead PGAM5, PGAM5L or PGAM5S led to restored IFNB expression, suggesting that neither the phosphatase activity nor the C domain are essential for PGAM5-dependent IFNβ expression. Taken together, this study indicates that PGAM5 functions as a mediator of intracellular RNA induced IFNβ response and its function is independent of its phosphatase activity. PGAM5 acts via a TBK1 dependent RIG-I pathway.

This thesis also showed that PGAM5 plays an important role in ISG induction. Upon poly(I:C) extracellular stimulation in HeLa cells, PGAM5 deficiency significantly decreased phosphorylation of STAT1 and expression of ISG. Moreover, under the stimulation of

Discussion

different types of IFNs, PGAM5 deficient HeLa cells exhibited decreased phosphorylation levels of STAT1 and expression of ISG as compared to WT cells.

STAT1 belongs to the STAT family and has been well documented as a key player in host defense against viral infection (Gao et al., 2012). STAT1 can be activated in response to IFNs by phosphorylation of specific tyrosine residues by JAKs (Raftery and Stevenson, 2017; Yang and Stark, 2008). Phosphorylated STAT1 further leads to a complex called IFN-stimulated gene factor 3 (ISGF3), a complex containing IRF9, phosphorylated STAT1 and STAT2. The ISGF3 complex then translocates from the cytoplasm into the nucleus to trigger transcription of ISG (Schneider et al., 2014). A significant induction of STAT1 phosphorylation and ISG expression was observed in poly(I:C) or IFN treated cells, indicating a strong activation of IFN pathways in HeLa cells. Moreover, a decrease of STAT1 phosphorylation and ISG expression was observe in PGAM5 deficient HeLa cells, suggesting that PGAM5 plays an important role in regulating ISG expression.

ISG are involved in a wide range of cellular activities, especially in antiviral responses (Liu et al., 2018). Under steady-state conditions, expression levels of ISG remain relatively low and can be rapidly boosted upon IFN stimulation (Schneider et al., 2014). ISG limit virus infection by directly targeting proteins or functions involved in the viral life cycle. For example, the IFN-induced protein with tetratricopeptide repeats (IFIT) family including IFIT1 inhibits viral replication and translation (Diamond and Farzan, 2013). Besides, IFI27 has also been shown to inhibit the viral replication (Itsui et al., 2006). This thesis clearly showed a diminished expression of IFIT1 and IFI27 in PGAM5 deficient HeLa cells, suggesting that PGAM5 may regulate antiviral response via ISG.

6.2 The mechanisms of PGAM5-dependent antiviral pathways.

This thesis used poly(I:C) to mimic viral infection and showed elevated PGAM5 multimer levels in cells treated with intracellular poly(I:C). The PGAM5 multimers were presented at steady-state but their numbers significantly increased after treatment. It has also been shown that PGAM5 multimers may only partially dependent on disulfide bonds. Data in this thesis linked PGAM5 multimerization with IFNβ expression and antiviral signaling pathways. PGAM5 multimeric complexes were previously described in processes like NLRP3 inflammasome activation and mitophagy, pathways in which PGAM5 were shown to play a role (Ma et al., 2020; Moriwaki et al., 2016; Park et al., 2018; Ruiz et al., 2019).It is currently unknown how PGAM5 multimerization regulates its function in IFN signaling

Discussion

pathway. One hypothesis is that the multimeric organization of PGAM5 might affect its interaction with its substrates. An example supporting this hypothesis is the formation of PGAM5 multimers during stimulation of selenite in HeLa cells. Selenite is a reagent that targets mitochondria and triggers ROS production to induce mitofission and mitophagy. The formation of PGAM5 multimers decreases its interaction with BCL-xL, which in turn promotes mitochondrial fission and mitophagy (Ma et al., 2020). Regarding the functional role of PGAM5 multimers in regulating IFNβ expression, one may speculate that PGAM5 multimerization increases its interaction with positive regulators or decreases its interaction with negative regulators in the RIG-I pathway. Another hypothesis is that the multimeric structure of PGAM5 is essential for PGAM5 function. For example, it has been shown that the dodecameric ring is an essential determinant of efficient catalytic activation of PGAM5 (Ruiz et al., 2019). The multimerization of PGAM5 is crucial for creating an interaction between a multimerization motif within the linker domain and the phosphatase domain. The interaction is important to keep the phosphatase active of PGAM5. Therefore it is possible that PGAM5 multimers serve as a molecular scaffold to support signal transduction.

Viral RNA activates RIG-I or MDA5 and transduces the signal through MAVS, which forms functional prion-like aggregates to activate downstream molecules such as TBK1 (Hou et al., 2011; Kawai et al., 2005; Meylan et al., 2005; Seth et al., 2005; Xu et al., 2005). TBK1 in turn activates IRF3 and induces the production of type I IFNs and other pro-inflammatory cytokines. This thesis showed that PGAM5 deficiency significantly impairs the phosphorylation levels of IRF3 and TBK1. Moreover, in the presence of a TBK1-specific inhibitor, no further IFNβ decrease was observed in PGAM5 deficient cells. Taken together, data in this thesis suggested that PGAM5 acts via a TBK1 dependent pathway. Interestingly, this thesis also observed that PGAM5 overexpression can be sufficient to induce TBK1 phosphorylation. This finding suggests that PGAM5 functions upstream of TBK1 and that PGAM5 levels within the cell can affect the activation level of TBK1. It has been shown that PGAM5 forms multimers when challenged with intracellular poly(I:C) and that MAVS, a key regulator upstream of TBK1, forms functional aggregates to activate antiviral innate immune responses (Hou et al., 2011; Seth et al., 2005; Yu et al., 2020). Here, this thesis further detected a direct interaction of PGAM5 and MAVS. Interestingly, data showed that the formation of PGAM5 multimers and MAVS aggregates are independent of each other. The aggregation of MAVS is a key process in inducing IFNβ expression and MAVS aggregation is undetectable in the steady-state (Hou et al., 2011). Nevertheless, it is not the case for PGAM5 multimers. An increase of PGAM5 multimers was found after intracellular poly(I:C) challenge, but PGAM5 multimers were also identified

Discussion

in the steady-state, indicating that PGAM5 multimers may not only function under the stimulation of intracellular poly(I:C) but also under steady-state.

It has been shown that several molecules like SIRT5, Ube2D3 and Ube2N are essential for IFN production via regulating MAVS aggregation (Liu et al., 2020; Shi et al., 2017). Given the important role of PGAM5 in regulating IFNβ expression, the role of PGAM5 in MAVS aggregation was investigated. Unexpectedly, the thesis showed that the formation of PGAM5 multimers and MAVS aggregates are independent of each other, suggesting that PGAM5 may modify MAVS independently of the formation of aggregates. Proteomic analyses have proved that MAVS signaling can also be regulated via a direct protein– protein interaction (Mehta and Trinkle-Mulcahy, 2016; Refolo et al., 2020). For example, the mitochondrial protein Interferon Induced Protein With Tetratricopeptide Repeats 3 (IFIT3) has been shown to function as a scaffold to facilitate the interaction of MAVS with TBK1 (Liu et al., 2011). Indeed, this thesis further uncovered that PGAM5 and MAVS physically interact and that this interaction was significantly enhanced by intracellular poly(I:C) stimulation. Given the crucial role of MAVS in this pathway, data here therefore suggested that PGAM5 promotes the antiviral response by direct effects on MAVS. It is currently unknown how the interaction between PGAM5 and MAVS regulates IFNβ expression. One may speculate that this interaction may impair the post-translational modifications (PTMs) on MAVS, especially given the fact that PGAM5 was identified as a phosphatase (He et al., 2017; Wang et al., 2012). Phosphorylation of serine residues 442, 444, and 44 of MAVS which are induced by TBK1 and IKKβ activation, activing downstream signaling pathways to promote IFNs expression (Liu et al., 2015). Conversely, phosphorylation of threonine 234 and serine 233 of MAVS was shown to inhibit antiviral responses (Vitour et al., 2009). However, it is currently unknown whether PGAM5 regulates any types of PTMs on MAVS. Hence, in future studies it would be interesting to elucidate the mechanism of PGAM5 modification of MAVS. Collectively, we demonstrate that the mitochondrial protein PGAM5 is important for the antiviral cellular response. Following challenge of HeLa cells with the dsRNA-analog poly(I:C) or viral RNA, PGAM5 oligomers and high levels of PGAM5 were found in mitochondrial aggregates. Importantly, a direct interaction of PGAM5 with MAVS was demonstrated. Furthermore, PGAM5 regulates the phosphorylation of TBK1/IRF3 to induce IFNβ expression. In summary, this thesis suggests that PGAM5 interacts with MAVS and regulates TBK1/IRF3 dependent IFNβ expression during intracellular RNA treatment (Figure 32).

Discussion

Figure 32. A proposed antiviral signaling pathway related to PGAM5. Following challenge of intracellular poly(I:C) or viral infection, PGAM5 formed oligomers and high levels of PGAM5 were found in mitochondrial aggregates. PGAM5 was found to directly interact with MAVS and regulate the phosphorylation levels of IRF3 and TBK1 to contribute to the expression of IFNβ and IFNβ target genes.

Besides, this study also demonstrated that PGAM5 regulates ISG induction and upregulates SOCS3 expression. Upon stimulation with extracellular poly(I:C) in HeLa cells, PGAM5 deficiency significantly increased SOCS3 expression. Moreover, under the stimulation of different types of IFNs, PGAM5 deficient HeLa cells exhibited up-regulated SOCS3 expression compared to WT cells. SOCS3 belongs to the SOCS family and has been shown to have an amino-terminal kinase-inhibitory region (KIR) that inhibits both cytokine receptors and JAK activity (Hansen et al., 1999; Nicholson et al., 2000; Sasaki et al., 2000; Yoshimura et al., 2018). During this process, SOCS3 binds to receptors to regulate JAK functions. The binding between SOCS3 and receptors instead of interfering with STAT recruitment, it inhibits catalytic activity of JAK in a manner that is analogous to SOCS1 (Bjorbak et al., 2000; Cohney et al., 1999; Sasaki et al., 2000). The inhibition of

Discussion

JAK catalytic activity results in decreased phosphorylation of STATs and expression of ISG (Tsai et al., 2019). Constitutive expression of SOCS3 inhibits IFN-induced STAT1 phosphorylation and ISG expression (Sakai et al., 2002). Thus, this study indeed suggested that STAT1 phosphorylation was regulated by increasing SOCS3 expression. In PGAM5 deficient HeLa cells, this thesis showed decreased phosphorylation levels of STAT1 and increased expression levels of SOCS3. This is in accordance with the finding that PGAM5 regulates ISG expression.

6.3 PGAM5 is dispensable for multiple organ injury induced in vivo.

In vitro data in this thesis showed that PGAM5 is an important regulator of IFNβ production and ISG expression in response to poly(I:C) stimulation. The thesis further investigated the functional roles of PGAM5 in an independent experiments in vivo. Studies indicated that PGAM5 is dispensable for poly(I:C)-induced small intestinal inflammation and DSS- induced colitis.

Poly(I:C) has been shown to be able to induce massive small intestinal injury via a TLR3 dependent pathway (Gunther et al., 2015; Marafini et al., 2015; Zhou et al., 2007). The consequence of abnormal TLR3 signaling is induction of cytokines produce in IECs. In vitro data indicated that PGAM5 regulates poly(I:C) induced STAT1 phosphorylation and ISG expression. It is currently unknown whether PGAM5 is involved in small intestinal injury. Interestingly, this thesis showed that PGAM5 is dispensable for poly(I:C)-induced small intestine tissue damage. A previous report has shown that the induction of TLR3 leads to the activation of NF-κB and production of IFNs (Kawai and Akira, 2007). However, neutralization of IFNα, IFNβ, and IFNλ did not protect mice from poly(I:C)-induced small intestine tissue damage, suggesting that IFNs are not crucial for poly(I:C)-induced damage (Zhou et al., 2007). Nevertheless, IL15 was shown to play an important role in poly(I:C)-induced small intestinal injury. Therefore, it service an explanation that PGAM5 is dispensable for poly(I:C)-induced small intestine damage.

DSS-induced colitis has been wildly used to research IBD. Studies have shown that multiple key molecules involved in cell death or NLRP3 inflammasome are involved in regulating DSS-induced colitis (Moriwaki et al., 2014; Zaki et al., 2010; Zhang et al., 2019). PGAM5 is a mitochondrial protein which has been implicated in a broad range of biological processes, for instance regulating cell death pathways and promoting activation of the NLRP3 inflammasome (He et al., 2017; Moriwaki et al., 2016; Wang et al., 2012). It is

Discussion

currently unknown whether PGAM5 regulates colitis. Interestingly, this thesis showed that PGAM5 is dispensable for acute DSS-induced colitis. Therefore, PGAM5 may function in a cell and organ-specific pathway under a variety of insults. Besides, DSS-induced colitis involves complicated pathways and multiple types of cells. Therefore, it may explain that PGAM5 is dispensable for DSS-induced colitis.

6.4 Working model

In conclusion, this thesis could show for the first time that PGAM5 is a key molecule in antiviral responses:

1. Intracellular poly(I:C) induces the formation of PGAM5 multimers.

2. PGAM5-MAVS interaction regulates TBK1/ IRF3 dependent antiviral responses.

3. PGAM5 regulates STAT1 phosphorylation and ISG expression.

4. PGAM5 is dispensable for poly(I:C)-induced small intestinal inflammation and DSS-induced colitis.

References

7 References

Alspach, E., Lussier, D.M., and Schreiber, R.D. (2019). Interferon gamma and Its Important Roles in Promoting and Inhibiting Spontaneous and Therapeutic Cancer Immunity. Cold Spring Harb Perspect Biol 11. Alunno, A., Padjen, I., Fanouriakis, A., and Boumpas, D.T. (2019). Pathogenic and Therapeutic Relevance of JAK/STAT Signaling in Systemic Lupus Erythematosus: Integration of Distinct Inflammatory Pathways and the Prospect of Their Inhibition with an Oral Agent. Cells 8. Andreakos, E., Zanoni, I., and Galani, I.E. (2019). Lambda interferons come to light: dual function cytokines mediating antiviral immunity and damage control. Curr Opin Immunol 56, 67-75. Au-Yeung, N., and Horvath, C.M. (2018). Transcriptional and chromatin regulation in interferon and innate antiviral gene expression. Cytokine Growth Factor Rev 44, 11-17. Aziz, N., Kang, Y.G., Kim, Y.J., Park, W.S., Jeong, D., Lee, J., Kim, D., and Cho, J.Y. (2020). Regulation of 8-Hydroxydaidzein in IRF3-Mediated Gene Expression in LPS- Stimulated Murine Macrophages. Biomolecules 10. Bano, D., and Prehn, J.H.M. (2018). Apoptosis-Inducing Factor (AIF) in Physiology and Disease: The Tale of a Repented Natural Born Killer. EBioMedicine 30, 29-37. Bauer, C., Duewell, P., Lehr, H.A., Endres, S., and Schnurr, M. (2012). Protective and aggravating effects of Nlrp3 inflammasome activation in IBD models: influence of genetic and environmental factors. Dig Dis 30 Suppl 1, 82-90. Bauer, C., Duewell, P., Mayer, C., Lehr, H.A., Fitzgerald, K.A., Dauer, M., Tschopp, J., Endres, S., Latz, E., and Schnurr, M. (2010). Colitis induced in mice with dextran sulfate sodium (DSS) is mediated by the NLRP3 inflammasome. Gut 59, 1192-1199. Bernkopf, D.B., Jalal, K., Bruckner, M., Knaup, K.X., Gentzel, M., Schambony, A., and Behrens, J. (2018). Pgam5 released from damaged mitochondria induces mitochondrial biogenesis via Wnt signaling. J Cell Biol 217, 1383-1394. Bianchi, F., Pretto, S., Tagliabue, E., Balsari, A., and Sfondrini, L. (2017). Exploiting poly(I:C) to induce cancer cell apoptosis. Cancer Biol Ther 18, 747-756. Bjorbak, C., Lavery, H.J., Bates, S.H., Olson, R.K., Davis, S.M., Flier, J.S., and Myers, M.G., Jr. (2000). SOCS3 mediates feedback inhibition of the leptin receptor via Tyr985. J Biol Chem 275, 40649-40657. Blasius, A.L., and Beutler, B. (2010). Intracellular toll-like receptors. Immunity 32, 305-315.

References

Brennan, J.J., and Gilmore, T.D. (2018). Evolutionary Origins of Toll-like Receptor Signaling. Mol Biol Evol 35, 1576-1587. Cai, X., Chen, J., Xu, H., Liu, S., Jiang, Q.X., Halfmann, R., and Chen, Z.J. (2014). Prion- like polymerization underlies signal transduction in antiviral immune defense and inflammasome activation. Cell 156, 1207-1222. Cantell, K., Hirvonen, S., Kauppinen, H.L., and Myllyla, G. (1981). Production of interferon in human leukocytes from normal donors with the use of Sendai virus. Methods Enzymol 78, 29-38. Cao, X. (2016). Self-regulation and cross-regulation of pattern-recognition receptor signalling in health and disease. Nat Rev Immunol 16, 35-50. Carstens, A., Dicksved, J., Nelson, R., Lindqvist, M., Andreasson, A., Bohr, J., Tysk, C., Talley, N.J., Agreus, L., Engstrand, L., et al. (2019). The Gut Microbiota in Collagenous Colitis Shares Characteristics With Inflammatory Bowel Disease-Associated Dysbiosis. Clin Transl Gastroenterol 10, e00065. Chaikuad, A., Filippakopoulos, P., Marcsisin, S.R., Picaud, S., Schroder, M., Sekine, S., Ichijo, H., Engen, J.R., Takeda, K., and Knapp, S. (2017). Structures of PGAM5 Provide Insight into Active Site Plasticity and Multimeric Assembly. Structure 25, 1089-1099 e1083. Chang, C.R., and Blackstone, C. (2007). Cyclic AMP-dependent protein kinase phosphorylation of Drp1 regulates its GTPase activity and mitochondrial morphology. J Biol Chem 282, 21583-21587. Chassaing, B., Aitken, J.D., Malleshappa, M., and Vijay-Kumar, M. (2014). Dextran sulfate sodium (DSS)-induced colitis in mice. Curr Protoc Immunol 104, 15 25 11-15 25 14. Chen, C.Y., Shih, Y.C., Hung, Y.F., and Hsueh, Y.P. (2019). Beyond defense: regulation of neuronal morphogenesis and brain functions via Toll-like receptors. J Biomed Sci 26, 90. Chen, G., Han, Z., Feng, D., Chen, Y., Chen, L., Wu, H., Huang, L., Zhou, C., Cai, X., Fu, C., et al. (2014). A regulatory signaling loop comprising the PGAM5 phosphatase and CK2 controls receptor-mediated mitophagy. Mol Cell 54, 362-377. Chen, S.N., Zou, P.F., and Nie, P. (2017). Retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) in fish: current knowledge and future perspectives. Immunology 151, 16- 25. Cheng, J., Liao, Y., Xiao, L., Wu, R., Zhao, S., Chen, H., Hou, B., Zhang, X., Liang, C., Xu, Y., et al. (2017). Autophagy regulates MAVS signaling activation in a phosphorylation- dependent manner in microglia. Cell Death Differ 24, 276-287. Chiang, C., and Gack, M.U. (2017). Post-translational Control of Intracellular Pathogen Sensing Pathways. Trends Immunol 38, 39-52.

References

Chiang, J.J., Davis, M.E., and Gack, M.U. (2014). Regulation of RIG-I-like receptor signaling by host and viral proteins. Cytokine Growth Factor Rev 25, 491-505. Choi, Y.S., Zhang, Y., Xu, M., Yang, Y., Ito, M., Peng, T., Cui, Z., Nagy, A., Hadjantonakis, A.K., Lang, R.A., et al. (2013). Distinct functions for Wnt/beta-catenin in hair follicle stem cell proliferation and survival and interfollicular epidermal homeostasis. Cell Stem Cell 13, 720-733. Clevers, H., and Nusse, R. (2012). Wnt/beta-catenin signaling and disease. Cell 149, 1192-1205. Cohney, S.J., Sanden, D., Cacalano, N.A., Yoshimura, A., Mui, A., Migone, T.S., and Johnston, J.A. (1999). SOCS-3 is tyrosine phosphorylated in response to interleukin-2 and suppresses STAT5 phosphorylation and lymphocyte proliferation. Mol Cell Biol 19, 4980- 4988. Cribbs, J.T., and Strack, S. (2007). Reversible phosphorylation of Drp1 by cyclic AMP- dependent protein kinase and calcineurin regulates mitochondrial fission and cell death. EMBO Rep 8, 939-944. Crill, E.K., Furr-Rogers, S.R., and Marriott, I. (2015). RIG-I is required for VSV-induced cytokine production by murine glia and acts in combination with DAI to initiate responses to HSV-1. Glia 63, 2168-2180. Cui, S., Wang, C., Bai, W., Li, J., Pan, Y., Huang, X., Yang, H., Feng, Z., Xiang, Q., Fei, L., et al. (2020). CD1d1 intrinsic signaling in macrophages controls NLRP3 inflammasome expression during inflammation. Sci Adv 6. Dela Justina, V., Giachini, F.R., Priviero, F., and Webb, R.C. (2020). Double-stranded RNA and Toll-like receptor activation: a novel mechanism for blood pressure regulation. Clin Sci (Lond) 134, 303-313. Deng, M., Tam, J.W., Wang, L., Liang, K., Li, S., Zhang, L., Guo, H., Luo, X., Zhang, Y., Petrucelli, A., et al. (2020). TRAF3IP3 negatively regulates cytosolic RNA induced antiviral signaling by promoting TBK1 K48 ubiquitination. Nat Commun 11, 2193. Diamond, M.S., and Farzan, M. (2013). The broad-spectrum antiviral functions of IFIT and IFITM proteins. Nat Rev Immunol 13, 46-57. Dixit, E., Boulant, S., Zhang, Y., Lee, A.S., Odendall, C., Shum, B., Hacohen, N., Chen, Z.J., Whelan, S.P., Fransen, M., et al. (2010). Peroxisomes are signaling platforms for antiviral innate immunity. Cell 141, 668-681. Drokhlyansky, E., Goz Ayturk, D., Soh, T.K., Chrenek, R., O'Loughlin, E., Madore, C., Butovsky, O., and Cepko, C.L. (2017). The brain parenchyma has a type I interferon response that can limit virus spread. Proc Natl Acad Sci U S A 114, E95-E104. Duncan, C.J.A., Randall, R.E., and Hambleton, S. (2020). Genetic Lesions of Type I Interferon Signalling in Human Antiviral Immunity. Trends Genet.

References

Durham, G.A., Williams, J.J.L., Nasim, M.T., and Palmer, T.M. (2019). Targeting SOCS Proteins to Control JAK-STAT Signalling in Disease. Trends Pharmacol Sci 40, 298-308. Foronjy, R.F., Taggart, C.C., Dabo, A.J., Weldon, S., Cummins, N., and Geraghty, P. (2015). Type-I interferons induce lung protease responses following respiratory syncytial virus infection via RIG-I-like receptors. Mucosal Immunol 8, 161-175. Fritsch, M., Gunther, S.D., Schwarzer, R., Albert, M.C., Schorn, F., Werthenbach, J.P., Schiffmann, L.M., Stair, N., Stocks, H., Seeger, J.M., et al. (2019). Caspase-8 is the molecular switch for apoptosis, necroptosis and pyroptosis. Nature 575, 683-687. Fu, X.Y., Kessler, D.S., Veals, S.A., Levy, D.E., and Darnell, J.E., Jr. (1990). ISGF3, the transcriptional activator induced by interferon alpha, consists of multiple interacting polypeptide chains. Proc Natl Acad Sci U S A 87, 8555-8559. Gao, B., Wang, H., Lafdil, F., and Feng, D. (2012). STAT proteins - key regulators of antiviral responses, inflammation, and tumorigenesis in the liver. J Hepatol 57, 430-441. Gao, X., Cao, Q., Cheng, Y., Zhao, D., Wang, Z., Yang, H., Wu, Q., You, L., Wang, Y., Lin, Y., et al. (2018). Chronic stress promotes colitis by disturbing the gut microbiota and triggering immune system response. Proc Natl Acad Sci U S A 115, E2960-E2969. Goropevsek, A., Holcar, M., and Avcin, T. (2017). The Role of STAT Signaling Pathways in the Pathogenesis of Systemic Lupus Erythematosus. Clin Rev Allergy Immunol 52, 164- 181. Gunther, C., Buchen, B., He, G.W., Hornef, M., Torow, N., Neumann, H., Wittkopf, N., Martini, E., Basic, M., Bleich, A., et al. (2015). Caspase-8 controls the gut response to microbial challenges by Tnf-alpha-dependent and independent pathways. Gut 64, 601- 610. Gunther, C., He, G.W., Kremer, A.E., Murphy, J.M., Petrie, E.J., Amann, K., Vandenabeele, P., Linkermann, A., Poremba, C., Schleicher, U., et al. (2016). The pseudokinase MLKL mediates programmed hepatocellular necrosis independently of RIPK3 during hepatitis. J Clin Invest 126, 4346-4360. Gunther, C., Martini, E., Wittkopf, N., Amann, K., Weigmann, B., Neumann, H., Waldner, M.J., Hedrick, S.M., Tenzer, S., Neurath, M.F., et al. (2011). Caspase-8 regulates TNF- alpha-induced epithelial necroptosis and terminal ileitis. Nature 477, 335-339. Gunther, C., Neumann, H., Neurath, M.F., and Becker, C. (2013). Apoptosis, necrosis and necroptosis: cell death regulation in the intestinal epithelium. Gut 62, 1062-1071. Hansen, J.A., Lindberg, K., Hilton, D.J., Nielsen, J.H., and Billestrup, N. (1999). Mechanism of inhibition of growth hormone receptor signaling by suppressor of cytokine signaling proteins. Mol Endocrinol 13, 1832-1843. Hayward, J.A., Mathur, A., Ngo, C., and Man, S.M. (2018). Cytosolic Recognition of Microbes and Pathogens: Inflammasomes in Action. Microbiol Mol Biol Rev 82.

References

He, G.W., Gunther, C., Kremer, A.E., Thonn, V., Amann, K., Poremba, C., Neurath, M.F., Wirtz, S., and Becker, C. (2017). PGAM5-mediated programmed necrosis of hepatocytes drives acute liver injury. Gut 66, 716-723. Holze, C., Michaudel, C., Mackowiak, C., Haas, D.A., Benda, C., Hubel, P., Pennemann, F.L., Schnepf, D., Wettmarshausen, J., Braun, M., et al. (2018). Oxeiptosis, a ROS- induced caspase-independent apoptosis-like cell-death pathway. Nat Immunol 19, 130- 140. Hou, F., Sun, L., Zheng, H., Skaug, B., Jiang, Q.X., and Chen, Z.J. (2011). MAVS forms functional prion-like aggregates to activate and propagate antiviral innate immune response. Cell 146, 448-461. Itsui, Y., Sakamoto, N., Kurosaki, M., Kanazawa, N., Tanabe, Y., Koyama, T., Takeda, Y., Nakagawa, M., Kakinuma, S., Sekine, Y., et al. (2006). Expressional screening of interferon-stimulated genes for antiviral activity against hepatitis C virus replication. J Viral Hepat 13, 690-700. Jahan, A.S., Biquand, E., Munoz-Moreno, R., Le Quang, A., Mok, C.K., Wong, H.H., Teo, Q.W., Valkenburg, S.A., Chin, A.W.H., Man Poon, L.L., et al. (2020). OTUB1 Is a Key Regulator of RIG-I-Dependent Immune Signaling and Is Targeted for Proteasomal Degradation by Influenza A NS1. Cell Rep 30, 1570-1584 e1576. Kang, T.B., Yang, S.H., Toth, B., Kovalenko, A., and Wallach, D. (2013). Caspase-8 blocks kinase RIPK3-mediated activation of the NLRP3 inflammasome. Immunity 38, 27-40. Kang, Y.J., Bang, B.R., Han, K.H., Hong, L., Shim, E.J., Ma, J., Lerner, R.A., and Otsuka, M. (2015). Regulation of NKT cell-mediated immune responses to tumours and liver inflammation by mitochondrial PGAM5-Drp1 signalling. Nat Commun 6, 8371. Katakura, K., Lee, J., Rachmilewitz, D., Li, G., Eckmann, L., and Raz, E. (2005). Toll-like receptor 9-induced type I IFN protects mice from experimental colitis. J Clin Invest 115, 695-702. Kato, H., Takeuchi, O., Mikamo-Satoh, E., Hirai, R., Kawai, T., Matsushita, K., Hiiragi, A., Dermody, T.S., Fujita, T., and Akira, S. (2008). Length-dependent recognition of double- stranded ribonucleic acids by retinoic acid-inducible gene-I and melanoma differentiation- associated gene 5. J Exp Med 205, 1601-1610. Kato, H., Takeuchi, O., Sato, S., Yoneyama, M., Yamamoto, M., Matsui, K., Uematsu, S., Jung, A., Kawai, T., Ishii, K.J., et al. (2006). Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature 441, 101-105. Kawai, T., and Akira, S. (2007). Signaling to NF-kappaB by Toll-like receptors. Trends Mol Med 13, 460-469. Kawai, T., and Akira, S. (2008). Toll-like receptor and RIG-I-like receptor signaling. Ann N Y Acad Sci 1143, 1-20.

References

Kawai, T., Takahashi, K., Sato, S., Coban, C., Kumar, H., Kato, H., Ishii, K.J., Takeuchi, O., and Akira, S. (2005). IPS-1, an adaptor triggering RIG-I- and Mda5-mediated type I interferon induction. Nat Immunol 6, 981-988. Kim, D.H., and Cheon, J.H. (2017). Pathogenesis of Inflammatory Bowel Disease and Recent Advances in Biologic Therapies. Immune Netw 17, 25-40. Komura, F., Takahashi, Y., Inoue, T., Takakura, Y., and Nishikawa, M. (2019). Development of a Nanostructured RNA/DNA Assembly as an Adjuvant Targeting Toll-Like Receptor 7/8. Nucleic Acid Ther 29, 335-342. Kotenko, S.V., Rivera, A., Parker, D., and Durbin, J.E. (2019). Type III IFNs: Beyond antiviral protection. Semin Immunol 43, 101303. Kuo, R.L., Kao, L.T., Lin, S.J., Wang, R.Y., and Shih, S.R. (2013). MDA5 plays a crucial role in enterovirus 71 RNA-mediated IRF3 activation. PLoS One 8, e63431. Lee, H.S., Lobbestael, E., Vermeire, S., Sabino, J., and Cleynen, I. (2020). Inflammatory bowel disease and Parkinson's disease: common pathophysiological links. Gut. Lenhausen, A.M., Wilkinson, A.S., Lewis, E.M., Dailey, K.M., Scott, A.J., Khan, S., and Wilkinson, J.C. (2016). Apoptosis Inducing Factor Binding Protein PGAM5 Triggers Mitophagic Cell Death That Is Inhibited by the Ubiquitin Ligase Activity of X-Linked Inhibitor of Apoptosis. Biochemistry 55, 3285-3302. Lester, S.N., and Li, K. (2014). Toll-like receptors in antiviral innate immunity. J Mol Biol 426, 1246-1264. Li, B., Xia, Y., and Hu, B. (2020). Infection and atherosclerosis: TLR-dependent pathways. Cell Mol Life Sci 77, 2751-2769. Li, K., Chen, Z., Kato, N., Gale, M., Jr., and Lemon, S.M. (2005). Distinct poly(I-C) and virus-activated signaling pathways leading to interferon-beta production in hepatocytes. J Biol Chem 280, 16739-16747. Lin, H.Y., Lai, R.H., Lin, S.T., Lin, R.C., Wang, M.J., Lin, C.C., Lee, H.C., Wang, F.F., and Chen, J.Y. (2013). Suppressor of cytokine signaling 6 (SOCS6) promotes mitochondrial fission via regulating DRP1 translocation. Cell Death Differ 20, 139-153. Lin, T.H., Su, H.H., Kang, H.Y., and Chang, T.H. (2017). The Interactive Roles of Lipopolysaccharides and dsRNA/Viruses on Respiratory Epithelial Cells and Dendritic Cells in Allergic Respiratory Disorders: The Hygiene Hypothesis. Int J Mol Sci 18. Liu, B.C., Sarhan, J., and Poltorak, A. (2018). Host-Intrinsic Interferon Status in Infection and Immunity. Trends Mol Med 24, 658-668. Liu, L., Feng, D., Chen, G., Chen, M., Zheng, Q., Song, P., Ma, Q., Zhu, C., Wang, R., Qi, W., et al. (2012). Mitochondrial outer-membrane protein FUNDC1 mediates hypoxia- induced mitophagy in mammalian cells. Nat Cell Biol 14, 177-185.

References

Liu, S., Cai, X., Wu, J., Cong, Q., Chen, X., Li, T., Du, F., Ren, J., Wu, Y.T., Grishin, N.V., et al. (2015). Phosphorylation of innate immune adaptor proteins MAVS, STING, and TRIF induces IRF3 activation. Science 347, aaa2630. Liu, S., Chen, J., Cai, X., Wu, J., Chen, X., Wu, Y.T., Sun, L., and Chen, Z.J. (2013). MAVS recruits multiple ubiquitin E3 ligases to activate antiviral signaling cascades. Elife 2, e00785. Liu, X., Zhu, C., Zha, H., Tang, J., Rong, F., Chen, X., Fan, S., Xu, C., Du, J., Zhu, J., et al. (2020). SIRT5 impairs aggregation and activation of the signaling adaptor MAVS through catalyzing lysine desuccinylation. EMBO J 39, e103285. Liu, X.Y., Chen, W., Wei, B., Shan, Y.F., and Wang, C. (2011). IFN-induced TPR protein IFIT3 potentiates antiviral signaling by bridging MAVS and TBK1. J Immunol 187, 2559- 2568. Llewellyn, S.R., Britton, G.J., Contijoch, E.J., Vennaro, O.H., Mortha, A., Colombel, J.F., Grinspan, A., Clemente, J.C., Merad, M., and Faith, J.J. (2018). Interactions Between Diet and the Intestinal Microbiota Alter Intestinal Permeability and Colitis Severity in Mice. Gastroenterology 154, 1037-1046 e1032. Lo, S.C., and Hannink, M. (2006). PGAM5, a Bcl-XL-interacting protein, is a novel substrate for the redox-regulated Keap1-dependent ubiquitin ligase complex. J Biol Chem 281, 37893-37903. Lo, S.C., and Hannink, M. (2008). PGAM5 tethers a ternary complex containing Keap1 and Nrf2 to mitochondria. Exp Cell Res 314, 1789-1803. Lu, W., Karuppagounder, S.S., Springer, D.A., Allen, M.D., Zheng, L., Chao, B., Zhang, Y., Dawson, V.L., Dawson, T.M., and Lenardo, M. (2014). Genetic deficiency of the mitochondrial protein PGAM5 causes a Parkinson's-like movement disorder. Nat Commun 5, 4930. Lu, W., Sun, J., Yoon, J.S., Zhang, Y., Zheng, L., Murphy, E., Mattson, M.P., and Lenardo, M.J. (2016). Mitochondrial Protein PGAM5 Regulates Mitophagic Protection against Cell Necroptosis. PLoS One 11, e0147792. Luo, D., Ding, S.C., Vela, A., Kohlway, A., Lindenbach, B.D., and Pyle, A.M. (2011). Structural insights into RNA recognition by RIG-I. Cell 147, 409-422. Ma, K., Zhang, Z., Chang, R., Cheng, H., Mu, C., Zhao, T., Chen, L., Zhang, C., Luo, Q., Lin, J., et al. (2020). Dynamic PGAM5 multimers dephosphorylate BCL-xL or FUNDC1 to regulate mitochondrial and cellular fate. Cell Death Differ 27, 1036-1051. MacDuff, D.A., Baldridge, M.T., Qaqish, A.M., Nice, T.J., Darbandi, A.D., Hartley, V.L., Peterson, S.T., Miner, J.J., Iwai, K., and Virgin, H.W. (2018). HOIL1 Is Essential for the Induction of Type I and III Interferons by MDA5 and Regulates Persistent Murine Norovirus Infection. J Virol 92.

References

Marafini, I., Monteleone, I., Di Fusco, D., Cupi, M.L., Paoluzi, O.A., Colantoni, A., Ortenzi, A., Izzo, R., Vita, S., De Luca, E., et al. (2015). TNF-alpha Producing Innate Lymphoid Cells (ILCs) Are Increased in Active Celiac Disease and Contribute to Promote Intestinal Atrophy in Mice. PLoS One 10, e0126291. Marq, J.B., Hausmann, S., Veillard, N., Kolakofsky, D., and Garcin, D. (2011). Short double-stranded RNAs with an overhanging 5' ppp-nucleotide, as found in arenavirus genomes, act as RIG-I decoys. J Biol Chem 286, 6108-6116. Mehta, V., and Trinkle-Mulcahy, L. (2016). Recent advances in large-scale protein interactome mapping. F1000Res 5. Meylan, E., Curran, J., Hofmann, K., Moradpour, D., Binder, M., Bartenschlager, R., and Tschopp, J. (2005). Cardif is an adaptor protein in the RIG-I antiviral pathway and is targeted by hepatitis C virus. Nature 437, 1167-1172. Moradi-Marjaneh, R., Hassanian, S.M., Fiuji, H., Soleimanpour, S., Ferns, G.A., Avan, A., and Khazaei, M. (2018). Toll like receptor signaling pathway as a potential therapeutic target in colorectal cancer. J Cell Physiol 233, 5613-5622. Moriwaki, K., Balaji, S., McQuade, T., Malhotra, N., Kang, J., and Chan, F.K. (2014). The necroptosis adaptor RIPK3 promotes injury-induced cytokine expression and tissue repair. Immunity 41, 567-578. Moriwaki, K., Farias Luz, N., Balaji, S., De Rosa, M.J., O'Donnell, C.L., Gough, P.J., Bertin, J., Welsh, R.M., and Chan, F.K. (2016). The Mitochondrial Phosphatase PGAM5 Is Dispensable for Necroptosis but Promotes Inflammasome Activation in Macrophages. J Immunol 196, 407-415. Murphy, J.M., Czabotar, P.E., Hildebrand, J.M., Lucet, I.S., Zhang, J.G., Alvarez-Diaz, S., Lewis, R., Lalaoui, N., Metcalf, D., Webb, A.I., et al. (2013). The pseudokinase MLKL mediates necroptosis via a molecular switch mechanism. Immunity 39, 443-453. Nagarajan, U. (2011). Induction and function of IFNbeta during viral and bacterial infection. Crit Rev Immunol 31, 459-474. Neurath, M.F. (2012). Animal models of inflammatory bowel diseases: illuminating the pathogenesis of colitis, ileitis and cancer. Digestive diseases 30 Suppl 1, 91-94. Nicholson, S.E., De Souza, D., Fabri, L.J., Corbin, J., Willson, T.A., Zhang, J.G., Silva, A., Asimakis, M., Farley, A., Nash, A.D., et al. (2000). Suppressor of cytokine signaling-3 preferentially binds to the SHP-2-binding site on the shared cytokine receptor subunit gp130. Proc Natl Acad Sci U S A 97, 6493-6498. O'Shea, J.J., Schwartz, D.M., Villarino, A.V., Gadina, M., McInnes, I.B., and Laurence, A. (2015). The JAK-STAT pathway: impact on human disease and therapeutic intervention. Annu Rev Med 66, 311-328.

References

Obuchi, M., Fernandez, M., and Barber, G.N. (2003). Development of recombinant vesicular stomatitis viruses that exploit defects in host defense to augment specific oncolytic activity. J Virol 77, 8843-8856. Oh, S.W., Onomoto, K., Wakimoto, M., Onoguchi, K., Ishidate, F., Fujiwara, T., Yoneyama, M., Kato, H., and Fujita, T. (2016). Leader-Containing Uncapped Viral Transcript Activates RIG-I in Antiviral Stress Granules. PLoS Pathog 12, e1005444. Ohto, U., Shibata, T., Tanji, H., Ishida, H., Krayukhina, E., Uchiyama, S., Miyake, K., and Shimizu, T. (2015). Structural basis of CpG and inhibitory DNA recognition by Toll-like receptor 9. Nature 520, 702-705. Oshiumi, H., Kouwaki, T., and Seya, T. (2016). Accessory Factors of Cytoplasmic Viral RNA Sensors Required for Antiviral Innate Immune Response. Front Immunol 7, 200. Panda, S., Srivastava, S., Li, Z., Vaeth, M., Fuhs, S.R., Hunter, T., and Skolnik, E.Y. (2016). Identification of PGAM5 as a Mammalian Protein Histidine Phosphatase that Plays a Central Role to Negatively Regulate CD4(+) T Cells. Mol Cell 63, 457-469. Paramsothy, S., Kamm, M.A., Kaakoush, N.O., Walsh, A.J., van den Bogaerde, J., Samuel, D., Leong, R.W.L., Connor, S., Ng, W., Paramsothy, R., et al. (2017). Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo- controlled trial. Lancet 389, 1218-1228. Park, Y.S., Choi, S.E., and Koh, H.C. (2018). PGAM5 regulates PINK1/Parkin-mediated mitophagy via DRP1 in CCCP-induced mitochondrial dysfunction. Toxicol Lett 284, 120- 128. Patankar, J.V., and Becker, C. (2020). Cell death in the gut epithelium and implications for chronic inflammation. Nat Rev Gastroenterol Hepatol 17, 543-556. Peisley, A., Jo, M.H., Lin, C., Wu, B., Orme-Johnson, M., Walz, T., Hohng, S., and Hur, S. (2012). Kinetic mechanism for viral dsRNA length discrimination by MDA5 filaments. Proc Natl Acad Sci U S A 109, E3340-3349. Qureshi, S.A., Salditt-Georgieff, M., and Darnell, J.E., Jr. (1995). Tyrosine-phosphorylated Stat1 and Stat2 plus a 48-kDa protein all contact DNA in forming interferon-stimulated- gene factor 3. Proc Natl Acad Sci U S A 92, 3829-3833. Raftery, N., and Stevenson, N.J. (2017). Advances in anti-viral immune defence: revealing the importance of the IFN JAK/STAT pathway. Cell Mol Life Sci 74, 2525-2535. Rathinam, V.A.K., and Chan, F.K. (2018). Inflammasome, Inflammation, and Tissue Homeostasis. Trends Mol Med 24, 304-318. Refolo, G., Vescovo, T., Piacentini, M., Fimia, G.M., and Ciccosanti, F. (2020). Mitochondrial Interactome: A Focus on Antiviral Signaling Pathways. Front Cell Dev Biol 8, 8.

References

Remijsen, Q., Goossens, V., Grootjans, S., Van den Haute, C., Vanlangenakker, N., Dondelinger, Y., Roelandt, R., Bruggeman, I., Goncalves, A., Bertrand, M.J., et al. (2014). Depletion of RIPK3 or MLKL blocks TNF-driven necroptosis and switches towards a delayed RIPK1 kinase-dependent apoptosis. Cell Death Dis 5, e1004. Rout, A.K., Udgata, S.R., Dehury, B., Pradhan, S.P., Swain, H.S., Behera, B.K., and Das, B.K. (2019). Structural bioinformatics insights into the CARD-CARD interaction mediated by the mitochondrial antiviral-signaling protein of black carp. J Cell Biochem 120, 12534- 12543. Ruiz, K., Thaker, T.M., Agnew, C., Miller-Vedam, L., Trenker, R., Herrera, C., Ingaramo, M., Toso, D., Frost, A., and Jura, N. (2019). Functional role of PGAM5 multimeric assemblies and their polymerization into filaments. Nat Commun 10, 531. Sadatomi, D., Tanimura, S., Ozaki, K., and Takeda, K. (2013). Atypical protein phosphatases: emerging players in cellular signaling. Int J Mol Sci 14, 4596-4612. Saghazadeh, A., and Rezaei, N. (2017). Implications of Toll-like receptors in Ebola infection. Expert Opin Ther Targets 21, 415-425. Sakai, I., Takeuchi, K., Yamauchi, H., Narumi, H., and Fujita, S. (2002). Constitutive expression of SOCS3 confers resistance to IFN-alpha in chronic myelogenous leukemia cells. Blood 100, 2926-2931. Sanchez David, R.Y., Combredet, C., Najburg, V., Millot, G.A., Beauclair, G., Schwikowski, B., Leger, T., Camadro, J.M., Jacob, Y., Bellalou, J., et al. (2019). LGP2 binds to PACT to regulate RIG-I- and MDA5-mediated antiviral responses. Sci Signal 12. Sartor, R.B. (2008). Microbial influences in inflammatory bowel diseases. Gastroenterology 134, 577-594. Sartor, R.B., and Muehlbauer, M. (2007). Microbial host interactions in IBD: implications for pathogenesis and therapy. Curr Gastroenterol Rep 9, 497-507. Sasaki, A., Yasukawa, H., Shouda, T., Kitamura, T., Dikic, I., and Yoshimura, A. (2000). CIS3/SOCS-3 suppresses erythropoietin (EPO) signaling by binding the EPO receptor and JAK2. J Biol Chem 275, 29338-29347. Schlee, M., Barchet, W., Hornung, V., and Hartmann, G. (2007). Beyond double-stranded RNA-type I IFN induction by 3pRNA and other viral nucleic acids. Curr Top Microbiol Immunol 316, 207-230. Schneider, W.M., Chevillotte, M.D., and Rice, C.M. (2014). Interferon-stimulated genes: a complex web of host defenses. Annu Rev Immunol 32, 513-545. Schwarzer, R., Jiao, H., Wachsmuth, L., Tresch, A., and Pasparakis, M. (2020). FADD and Caspase-8 Regulate Gut Homeostasis and Inflammation by Controlling MLKL- and GSDMD-Mediated Death of Intestinal Epithelial Cells. Immunity 52, 978-993 e976.

References

Sekine, S., Kanamaru, Y., Koike, M., Nishihara, A., Okada, M., Kinoshita, H., Kamiyama, M., Maruyama, J., Uchiyama, Y., Ishihara, N., et al. (2012). Rhomboid protease PARL mediates the mitochondrial membrane potential loss-induced cleavage of PGAM5. J Biol Chem 287, 34635-34645. Sekine, S., Yao, A., Hattori, K., Sugawara, S., Naguro, I., Koike, M., Uchiyama, Y., Takeda, K., and Ichijo, H. (2016). The Ablation of Mitochondrial Protein Phosphatase Pgam5 Confers Resistance Against Metabolic Stress. EBioMedicine 5, 82-92. Seth, R.B., Sun, L., Ea, C.K., and Chen, Z.J. (2005). Identification and characterization of MAVS, a mitochondrial antiviral signaling protein that activates NF-kappaB and IRF 3. Cell 122, 669-682. Shi, Y., Yuan, B., Zhu, W., Zhang, R., Li, L., Hao, X., Chen, S., and Hou, F. (2017). Ube2D3 and Ube2N are essential for RIG-I-mediated MAVS aggregation in antiviral innate immunity. Nat Commun 8, 15138. Singh, V., Yeoh, B.S., Walker, R.E., Xiao, X., Saha, P., Golonka, R.M., Cai, J., Bretin, A.C.A., Cheng, X., Liu, Q., et al. (2019). Microbiota fermentation-NLRP3 axis shapes the impact of dietary fibres on intestinal inflammation. Gut 68, 1801-1812. Skarnes, W.C., Rosen, B., West, A.P., Koutsourakis, M., Bushell, W., Iyer, V., Mujica, A.O., Thomas, M., Harrow, J., Cox, T., et al. (2011). A conditional knockout resource for the genome-wide study of mouse gene function. Nature 474, 337-342. Smirnova, E., Griparic, L., Shurland, D.L., and van der Bliek, A.M. (2001). Dynamin-related protein Drp1 is required for mitochondrial division in mammalian cells. Mol Biol Cell 12, 2245-2256. Song, T., Wei, C., Zheng, Z., Xu, Y., Cheng, X., Yuan, Y., Guan, K., Zhang, Y., Ma, Q., Shi, W., et al. (2010). c-Abl tyrosine kinase interacts with MAVS and regulates innate immune response. FEBS Lett 584, 33-38. Springer, W., and Kahle, P.J. (2011). Regulation of PINK1-Parkin-mediated mitophagy. Autophagy 7, 266-278. Stark, G.R., and Darnell, J.E., Jr. (2012). The JAK-STAT pathway at twenty. Immunity 36, 503-514. Stolzer, I., Kaden-Volynets, V., Ruder, B., Letizia, M., Bittel, M., Rausch, P., Basic, M., Bleich, A., Baines, J.F., Neurath, M.F., et al. (2020). Environmental Microbial Factors Determine the Pattern of Inflammatory Lesions in a Murine Model of Crohn's Disease-Like Inflammation. Inflamm Bowel Dis 26, 66-79. Sun, L., Wang, H., Wang, Z., He, S., Chen, S., Liao, D., Wang, L., Yan, J., Liu, W., Lei, X., et al. (2012). Mixed lineage kinase domain-like protein mediates necrosis signaling downstream of RIP3 kinase. Cell 148, 213-227.

References

Swanson, K.V., Deng, M., and Ting, J.P. (2019). The NLRP3 inflammasome: molecular activation and regulation to therapeutics. Nat Rev Immunol 19, 477-489. Takeda, K., Komuro, Y., Hayakawa, T., Oguchi, H., Ishida, Y., Murakami, S., Noguchi, T., Kinoshita, H., Sekine, Y., Iemura, S., et al. (2009). Mitochondrial phosphoglycerate mutase 5 uses alternate catalytic activity as a protein serine/threonine phosphatase to activate ASK1. Proc Natl Acad Sci U S A 106, 12301-12305. Taniguchi, Y., Choi, P.J., Li, G.W., Chen, H., Babu, M., Hearn, J., Emili, A., and Xie, X.S. (2010). Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science 329, 533-538. Tipton, P., Su, T., and Hannink, M. (2018). Assembly of PGAM5 into Multimeric Complexes Provides a Mechanism for Allosteric Regulation of Phosphatase Activity. Methods Enzymol 607, 353-372. Tsai, M.C., Chen, W.J., Tsai, M.S., Ching, C.H., and Chuang, J.I. (2011). Melatonin attenuates brain contusion-induced oxidative insult, inactivation of signal transducers and activators of transcription 1, and upregulation of suppressor of cytokine signaling-3 in rats. J Pineal Res 51, 233-245. Tsai, M.H., Pai, L.M., and Lee, C.K. (2019). Fine-Tuning of Type I Interferon Response by STAT3. Front Immunol 10, 1448. Uchikawa, E., Lethier, M., Malet, H., Brunel, J., Gerlier, D., and Cusack, S. (2016). Structural Analysis of dsRNA Binding to Anti-viral Pattern Recognition Receptors LGP2 and MDA5. Mol Cell 62, 586-602. Urcuqui-Inchima, S., Cabrera, J., and Haenni, A.L. (2017). Interplay between dengue virus and Toll-like receptors, RIG-I/MDA5 and microRNAs: Implications for pathogenesis. Antiviral Res 147, 47-57. Villarino, A.V., Kanno, Y., Ferdinand, J.R., and O'Shea, J.J. (2015). Mechanisms of Jak/STAT signaling in immunity and disease. J Immunol 194, 21-27. Vitour, D., Dabo, S., Ahmadi Pour, M., Vilasco, M., Vidalain, P.O., Jacob, Y., Mezel- Lemoine, M., Paz, S., Arguello, M., Lin, R., et al. (2009). Polo-like kinase 1 (PLK1) regulates interferon (IFN) induction by MAVS. J Biol Chem 284, 21797-21809. Wada, J., and Makino, H. (2016). Innate immunity in diabetes and diabetic nephropathy. Nat Rev Nephrol 12, 13-26. Wang, B.X., and Fish, E.N. (2019). Global virus outbreaks: Interferons as 1st responders. Semin Immunol 43, 101300. Wang, W., Xu, L., Su, J., Peppelenbosch, M.P., and Pan, Q. (2017). Transcriptional Regulation of Antiviral Interferon-Stimulated Genes. Trends Microbiol 25, 573-584.

References

Wang, X., Jiang, W., Yan, Y., Gong, T., Han, J., Tian, Z., and Zhou, R. (2014). RNA viruses promote activation of the NLRP3 inflammasome through a RIP1-RIP3-DRP1 signaling pathway. Nat Immunol 15, 1126-1133. Wang, Z., Jiang, H., Chen, S., Du, F., and Wang, X. (2012). The mitochondrial phosphatase PGAM5 functions at the convergence point of multiple necrotic death pathways. Cell 148, 228-243. Wen, C., Yan, Z., Yang, X., Guan, K., Xu, C., Song, T., Zheng, Z., Wang, W., Wang, Y., Zhao, M., et al. (2012). Identification of tyrosine-9 of MAVS as critical target for inducible phosphorylation that determines activation. PLoS One 7, e41687. Wilkins, J.M., McConnell, C., Tipton, P.A., and Hannink, M. (2014). A conserved motif mediates both multimer formation and allosteric activation of phosphoglycerate mutase 5. J Biol Chem 289, 25137-25148. Wu, J., and Chen, Z.J. (2014). Innate immune sensing and signaling of cytosolic nucleic acids. Annu Rev Immunol 32, 461-488. Xavier, R.J., and Podolsky, D.K. (2007). Unravelling the pathogenesis of inflammatory bowel disease. Nature 448, 427-434. Xiang, W., Zhang, Q., Lin, X., Wu, S., Zhou, Y., Meng, F., Fan, Y., Shen, T., Xiao, M., Xia, Z., et al. (2016). PPM1A silences cytosolic RNA sensing and antiviral defense through direct dephosphorylation of MAVS and TBK1. Sci Adv 2, e1501889. Xu, L.G., Wang, Y.Y., Han, K.J., Li, L.Y., Zhai, Z., and Shu, H.B. (2005). VISA is an adapter protein required for virus-triggered IFN-beta signaling. Mol Cell 19, 727-740. Xu, W., Jing, L., Wang, Q., Lin, C.C., Chen, X., Diao, J., Liu, Y., and Sun, X. (2015). Bax- PGAM5L-Drp1 complex is required for intrinsic apoptosis execution. Oncotarget 6, 30017- 30034. Yan, C., Gong, L., Chen, L., Xu, M., Abou-Hamdan, H., Tang, M., Desaubry, L., and Song, Z. (2020). PHB2 (prohibitin 2) promotes PINK1-PRKN/Parkin-dependent mitophagy by the PARL-PGAM5-PINK1 axis. Autophagy 16, 419-434. Yang, J., and Stark, G.R. (2008). Roles of unphosphorylated STATs in signaling. Cell Res 18, 443-451. Yang, J.Y., Kim, M.S., Kim, E., Cheon, J.H., Lee, Y.S., Kim, Y., Lee, S.H., Seo, S.U., Shin, S.H., Choi, S.S., et al. (2016). Enteric Viruses Ameliorate Gut Inflammation via Toll-like Receptor 3 and Toll-like Receptor 7-Mediated Interferon-beta Production. Immunity 44, 889-900. Yang, W., Ru, Y., Ren, J., Bai, J., Wei, J., Fu, S., Liu, X., Li, D., and Zheng, H. (2019). G3BP1 inhibits RNA virus replication by positively regulating RIG-I-mediated cellular antiviral response. Cell Death Dis 10, 946.

References

Yang, Z., Balic, A., Michon, F., Juuri, E., and Thesleff, I. (2015). Mesenchymal Wnt/beta- Catenin Signaling Controls Epithelial Stem Cell Homeostasis in Teeth by Inhibiting the Antiapoptotic Effect of Fgf10. Stem Cells 33, 1670-1681. Yoshimura, A., Ito, M., Chikuma, S., Akanuma, T., and Nakatsukasa, H. (2018). Negative Regulation of Cytokine Signaling in Immunity. Cold Spring Harb Perspect Biol 10. Yu, Y.Q., Zielinska, M., Li, W., Bernkopf, D.B., Heilingloh, C.S., Neurath, M.F., and Becker, C. (2020). PGAM5-MAVS interaction regulates TBK1/ IRF3 dependent antiviral responses. Sci Rep 10, 8323. Zaki, M.H., Boyd, K.L., Vogel, P., Kastan, M.B., Lamkanfi, M., and Kanneganti, T.D. (2010). The NLRP3 inflammasome protects against loss of epithelial integrity and mortality during experimental colitis. Immunity 32, 379-391. Zhang, H.L., Ye, H.Q., Liu, S.Q., Deng, C.L., Li, X.D., Shi, P.Y., and Zhang, B. (2017a). West Nile Virus NS1 Antagonizes Interferon Beta Production by Targeting RIG-I and MDA5. J Virol 91. Zhang, J., Qin, D., Yang, Y.J., Hu, G.Q., Qin, X.X., Du, C.T., and Chen, W. (2019). MLKL deficiency inhibits DSS-induced colitis independent of intestinal microbiota. Mol Immunol 107, 132-141. Zhang, Y., Ma, C.A., Lawrence, M.G., Break, T.J., O'Connell, M.P., Lyons, J.J., Lopez, D.B., Barber, J.S., Zhao, Y., Barber, D.L., et al. (2017b). PD-L1 up-regulation restrains Th17 cell differentiation in STAT3 loss- and STAT1 gain-of-function patients. J Exp Med 214, 2523-2533. Zhao, J., Zeng, Y., Xu, S., Chen, J., Shen, G., Yu, C., Knipe, D., Yuan, W., Peng, J., Xu, W., et al. (2016). A Viral Deamidase Targets the Helicase Domain of RIG-I to Block RNA- Induced Activation. Cell Host Microbe 20, 770-784. Zhao, Y., and Karijolich, J. (2019). Know Thyself: RIG-I-Like Receptor Sensing of DNA Virus Infection. J Virol 93. Zheng, J., Yong, H.Y., Panutdaporn, N., Liu, C., Tang, K., and Luo, D. (2015). High- resolution HDX-MS reveals distinct mechanisms of RNA recognition and activation by RIG-I and MDA5. Nucleic Acids Res 43, 1216-1230. Zhong, X., Feng, L., Zang, R., Lei, C.Q., Yang, Q., and Shu, H.B. (2020). ZFYVE1 negatively regulates MDA5- but not RIG-I-mediated innate antiviral response. PLoS Pathog 16, e1008457. Zhou, R., Wei, H., Sun, R., and Tian, Z. (2007). Recognition of double-stranded RNA by TLR3 induces severe small intestinal injury in mice. J Immunol 178, 4548-4556. Zhu, H., and Zheng, C. (2020). The Race between Host Antiviral Innate Immunity and the Immune Evasion Strategies of Herpes Simplex Virus 1. Microbiol Mol Biol Rev 84.

References

Zhuang, M., Guan, S., Wang, H., Burlingame, A.L., and Wells, J.A. (2013). Substrates of IAP ubiquitin ligases identified with a designed orthogonal E3 ligase, the NEDDylator. Mol Cell 49, 273-282. Zundler, S., Becker, E., Schulze, L.L., and Neurath, M.F. (2019). Immune cell trafficking and retention in inflammatory bowel disease: mechanistic insights and therapeutic advances. Gut 68, 1688-1700.

List of abbreviations

8 List of abbreviations.

°C Degree Celsius µg microgram µg/ml microgram per millilitre µl microliter µm micrometer PRRs pattern recognition receptors TLRs Toll-like receptors RLRs RIG-I-like receptors NLRs Nod-like receptors MDA5 melanoma differentiation associated protein 5 VSV vesicular stomatitis Indiana virus LRRs leucine-rich repeats TIR Toll/IL-1 receptor LPS lipopolysaccharide MyD88 myeloid differentiation primary response gene (88) TRIF TIR-domain containing adaptor inducing IFN-β AP-1 activator protein 1 PAMPs pathogen-associated molecular patterns CARDs recruitment domains TBK1 TANK-binding kinase 1 IKKε IκB kinase-ε IRF3 IFN-regulatory factor 3 NDV Newcastle disease virus SeV Sendai virus VSV vesicular stomatitis virus EMCV encephalomyocarditis virus MAVS mitochondrial antiviral-signaling protein Cardif CARD adapter inducing IFN beta VISA virus-induced signaling adapter IPS-1 IFN-β promoter stimulator 1 TMD transmembrane domain SDS sodium dodecyl sulfate

100

List of abbreviations

SDD-AGE semi-denaturing detergent agarose gel electrophoresis PTMs posttranslational modifications PPM1A protein phosphatase magnesium-dependent 1A IFN interferons IFNΓR1 IFNγ receptor 1 IFNΓR2 IFNγ receptor 2 JAKs janus kinases STATs signal transducers and activators of transcription ISG IFN-stimulated genes JAK1 janus kinase 1 TYK2 tyrosine kinase 2 PGAM5 Phosphoglycerate mutase family member 5 DRP1 dynamin related protein 1 SOCS6 suppressor of cytokine signaling 6 IAPs inhibitors of apoptotic protein RIP receptor-interacting protein XIAP X-linked inhibitor of apoptosis PINK1 PTEN-induced kinase 1 PARL presenilins-associated rhomboid-like protein PHB2 prohibitin 2 PRKN Parkin RBR E3 Ubiquitin Protein Ligase FUNDC1 FUN14 Domain Containing 1 CK2 casein kinase 2 ROS reactive oxygen species KEAP1 Kelch-like ECH-associated protein 1 AIFM1 apoptosis inducing factor 1 NADH Nicotinamide adenine dinucleotide APC adenomatous polyposis coli CK1α casein kinase 1 α GSK3 glycogen synthase kinase 3 FZD frizzled family LRP5/6 LDL receptor related protein 5/6 co-receptors PARL protease presenilin associated rhomboid like NLRP3 NOD-, LRR- and pyrin domain-containing protein 3 MEFs murine embryonic fibroblasts ATCC American Type Culture Collection

101

List of abbreviations

DMEM Dulbecco’s modified Eagle’s medium PBS phosphate buffered saline WB Western Blot DSS dextran sulfate sodium PCR Polymerase chain reaction HPRT hypoxanthine guanine phosphoribosyl transferase IP Immunoprecipitation TUNEL Terminal deoxynucleotidyl transferase dUTP nick end labeling S.D standard deviation BME β-mercaptoethanol WT wild type ISG IFN-stimulated genes IBD Inflammatory bowel disease UC ulcerative colitis CD Crohn’s disease IFIT3 Interferon Induced Protein With Tetratricopeptide Repeats 3 KIR kinase-inhibitory region

102

List of Tables

9 List of Tables.

Table 1. TLRs and their ligands ...... 16 Table 2. General chemicals and reagents used in the study ...... 27 Table 3. Commonly used Instruments in the study (unless indicated in the text) ...... 30 Table 4. Various primers used for genotyping in the study ...... 34 Table 5. PCR programs used for genotyping in the study ...... 34 Table 6. The composition of different buffers used in the procedure ...... 35 Table 7. List of antibodies used ...... 39 Table 8. Composition of buffers used in Western blot...... 39 Table 9. The composition of different solutions used in immunohistochemical staining . 41 Table 10. List of antibodies used in immunohistochemical staining ...... 42 Table 11. The composition of different solutions used in immunocytochemistry staining 42 Table 12. List of antibodies used in immunocytochemistry staining ...... 43

103

List of Figures

10 List of figures

Figure 1. Two isoforms of PGAM5...... 10 Figure 2. The roles of PGAM5 in regulating cell death...... 13 Figure 3. TLRs signaling pathways related to virus infection...... 18 Figure 4. RLRs pathways related to virus infection...... 20 Figure 5. MAVS’s aggregation and phosphorylation...... 22 Figure 6. Extracellular poly(I:C) treatment activated the TLR3 pathway in HeLa cells in different time points...... 45 Figure 7. Intracellular poly(I:C) treatment induced visible PGAM5 aggregates in HeLa cells...... 46 Figure 8. Intracellular poly(I:C) treatment induced PGAM5 multimers formation in HeLa cells...... 47 Figure 9. Generation of PGAM5 knockout HeLa cells...... 49 Figure 10. PGAM5 deficiency attenuates intracellular poly(I:C)-induced IFNB expression in HeLa cells...... 50 Figure 11. Isolation of PGAM5−/− MEFs...... 51 Figure 12. PGAM5 deficiency attenuates intracellular poly(I:C)-induced Ifnβ expression in MEFs...... 52 Figure 13. PGAM5 overexpression rescued intracellular poly(I:C)-induced IFNB expression and its function is independent of its phosphatase activity...... 53 Figure 14. PGAM5 overexpression rescued intracellular poly(I:C)-induced IFNB expression and its function is independent of its alternative splicing in the C domain. ... 54 Figure 15. PGAM5 deficiency attenuates intracellular poly(I:C)-induced IRF3 and TBK1 phosphorylation...... 55 Figure 16. PGAM5 acts via a TBK1 dependent pathway...... 56 Figure 17. PGAM5 acts via a TBK1 dependent pathway...... 57 Figure 18. PGAM5 does not regulate MAVS aggregation...... 58 Figure 19. MAVS does not regulate PGAM5 multimerization...... 59 Figure 20. PGAM5 interacts with MAVS...... 60 Figure 21. PGAM5 deficiency attenuates VSV-induced IRF3 phosphorylation...... 61 Figure 22. PGAM5 deficiency attenuates VSV-induced Ifnb expression...... 62 Figure 23. PGAM5 deficiency attenuates extracellular poly(I:C)-induced STAT1 phosphorylation...... 65 Figure 24. PGAM5 deficiency attenuates extracellular poly(I:C)-induced IFIT1, IFI27 and IL6 expression...... 66

104

List of Figures

Figure 25. PGAM5 deficiency up-regulates SOCS3 gene expression and protein levels...... 67 Figure 26. PGAM5 deficiency attenuates IFNs-induced IFI27 and IFIT1 expression. .... 69 Figure 27. PGAM5 deficiency attenuates IFN-induced IFI27 and IFIT1 expression...... 70 Figure 28. Confirmation of PGAM5 deficiency in vivo...... 72 Figure 29. PGAM5 is dispensable for poly(I:C)-induced small intestinal injury...... 73 Figure 30. PGAM5 deficiency in colon epithelial cells from PGAM5 fl/fl VillinCre+ mice. 75 Figure 31. PGAM5 is dispensable for DSS-induced colitis...... 76 Figure 32. A proposed antiviral signaling pathway related to PGAM5...... 82

105

Conference Presentations

11 Conference Presentations

1. 06/2016 Poster Presentation, Seeon Conference 2. 06/2019 Poster Presentation, IZKF symposium, Bad Staffelstein Declaration

12 Declaration

I hereby declare that this thesis is a result of my own independent work and investigation. I state that this dissertation has not been previously accepted for any degree in any other university. Parts of the results from this thesis have been accepted for publication at the journal Scientific Reports with the title `` PGAM5-MAVS interaction regulates TBK1/ IRF3 dependent antiviral responses´´ on April 2020.

Date and place:

Signature:

107