Proteostasis Maintenance of Γ-Aminobutyric Acid Type a Receptors (Gabaars)

PROTEOSTASIS MAINTENANCE OF γ-AMINOBUTYRIC ACID TYPE A RECEPTORS (GABAARS)

YANLIN FU

Submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Thesis Advisor: Dr. Tingwei Mu

Department of Physiology and Biophysics

CASE WESTERN RESERVE UNIVERSITY

May, 2019

CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Yanlin Fu

candidate for the degree of PhD.

Committee Chair Corey Smith

Committee Member George Dubyak

Committee Member Andrea Romani

Committee Member Ben Strowbridge

Committee Member Martin Snider

Committee Member Tingwei Mu

Date of Defense March 13th 2019

*We also certify that written approval has been obtained for any proprietary material contained therein.

Table of Contents

List of Figures------vi

Acknowledgements------ix

List of Abbreviations------xi

Abstract------xv

Chapter 1 Introduction

1.1. General introduction of GABAARs and other cys-loop family receptors------1

1.2. Factors that affect Cys-loop receptors surface expression

1.2.1. Folding and assembly of Cys-loop receptors------5

1.2.2. ER associated degradation (ERAD) of the Cys-loop receptors------8

1.2.3. Factors that regulate surface trafficking ------9

1.2.4. Clustering ------11

1.2.5. Endocytosis------12

1.2.6. Other regulations of Cys-loop receptors

1.2.6.1. Lipid involvement in trafficking and clustering------14

1.2.6.2. Phosphorylation signaling in the biogenesis of the receptors------14

1.2.7. Mutations in Cys-loop receptors that lead to their reduced

surface expression level and diseases

1.2.7.1. Current mutations that cause reduced surface expression------16

1.2.7.2. Pathogenesis of A322D and D219N GABAARs------18

iii

1.3 Protein Export efficiency model and therapy strategies------21

Chapter 2 Modulating UPR ATF6 pathway and UPR sXBP1 pathway

promotes the surface trafficking of A322D GABAARs

2.1 Introduction------27

2.2 Materials and Methods------39

2.3 Results------45

2.4 Discussion------51

2.5 Figures------53

Chapter 3 The effect of BIX treatment on the trafficking of A322D and

D219N GABAARs

3.1 Introduction------67

3.2 Materials and Methods------71

3.3 Results------74

3.4 Discussion------79

3.5 Figures------83

Chapter 4 LMAN1 (aka ERGIC-53) promotes trafficking of Cys-loop

neuroreceptors

4.1 Introduction------95

4.2 Materials and Methods------99

4.3 Results------103

4.4 Discussion------107

4.5 Figures------110

Chapter 5 Conclusions and Future directions------122

Bibliography------126

List of Figures

Figure 1. Structural characteristics of the Cys-loop receptors------4

Figure 2. Protein biogenesis pathway of the Cys-loop receptors------26

Figure 3. A schematic cartoon figure of the unfolded protein response

(UPR) pathways------33

Figure 4. Molecular structures of GABAA receptors------54

Figure 5. Introduction of the A322D mutation in the α1 subunit does not activate

the UPR------56

Figure 6. ATF6 activation promotes the forward trafficking of α1(A322D)

subunit of GABAA receptors------58

Figure 7. ATF6 activation enhances the surface expression level of α1(A322D)

subunit of GABAA receptors without affecting their degradation rate------60

Figure 8. ATF6 activation modulates the proteostasis network of

Endoplsmic Reticulum------62

Figure 9. IRE1 activation increases the total and surface expression

α1(A322D) subunit of GABAA receptors without significantly

slowing their degradation rate------64

Figure 10. IRE1 activation modulates the proteostasis network of Endoplsmic

Reticulum and ATF6 or IRE1 pathways activation has different

effects on WT α1 subunit of GABAA receptors ------66

Figure 11. BIX, a potent BiP inducer, enhances total expression level of

α1(A322D) subunit of GABAA receptors in a dose and time

dependent manner------84

Figure 12. Influence of GABAA receptor protein expression levels on the effect of

BIX treatment------86

Figure 13. BIX, a potent BiP inducer, promotes the maturation and reduces

the degradation of α1(A322D) subunits------88

Figure 14. BIX enhances the functional surface expression of α1(A322D) subunit

of GABAA receptors------90

Figure 15. BIX enhances the functional surface expression of α1 subunit variants

of GABAA receptors------92

Figure 16. BIX’s enhancing of trafficking of α1(A322D)β2γ2 receptors effect

is through overexpression of BiP------94

Figure 17. A schematic cartoon figure of the LMAN1 trafficking between the ER

and the ER to Golgi Compartment (ERGIC) ------97

Figure 18. Transient knockdown of LMAN1 using LMAN1 siRNA-1 affects the

total and surface expression of endogenous GABAARs subunits------111

Figure 19. Transient knockdown of LMAN1 using LMAN1 siRNA-2 affects the

vii

total and surface expression of endogenous GABAARs subunits------113

Figure 20. Knockout of LMAN1 in mouse leads to decreased total expression

level of GABAARs subunits and other Cys-loop family receptor

subunits------115

Figure 21. Influence of LMAN1 knockout on the proteostasis network in the

central nervous system (1) ------117

Figure 22. Influence of LMAN1 knockout on the proteostasis network in the

central nervous system (2) ------119

Figure 23. LMAN1 interacts with GABAARs in HEK293T cells in a

glycan-independent manner------121

viii

Acknowledgement

I would like to first thank my thesis advisor Dr. Tingwei Mu. He is a young, pioneer and

talented principal investigator that studies regulators for WT and mutant GABAARs. After

I joined the lab, Tingwei provided me with many of his scientific ideas and gave me

instructions on experiments. Tingwei not only taught me how to perform good experiments

but also how to think scientifically, how to find breakthroughs in project. He is a very warm and supportive PI to me. I also learned a lot from two previous postdocs in the lab, Dr

Xiaojing Di and Dr Dongyun Han. They are always willing to help me. Every week’s lab

meeting also allows me to learn better about how to interpret data, how to report data and

pushes me to read a lot of cutting edge papers. Next, I would like to thank my committee,

Dr Corey Smith, Dr George Dubyak, Dr Andrea Romani, Dr Ben Strowbridge, Dr Martin

Snider. My committee always attend my committee meetings on time, provide me with precious feedbacks about my science projects, give both me and Tingwei a lot of encouragement and support when we need them very much. I am very grateful to them.

Special thanks to Dr Bill Schilling for providing me and Tingwei with many advices and suggestions. I also would like to thank the chair Dr Walter Boron and the Physiology and

Biophysics Department to gives me, a foreign student coming from China, this precious opportunity to study in the department to learn how to do better science and to give me such a supportive and warm environment. I would like to thank Jean Davis, Morley

Schwebel, Michael Little, Bart Jarmusch, Paul Zawolowycz to help me to organize and

make schedules for many of my study events. I would like to thank my father (Zhongliang

Fu) and my mother (Xiaoping Liu) for always supporting me unconditionally. I also would

like to thank my auntie (Min Liu), my uncle (Nanhao Li) and my cousin (Jiawen Yu) who

ix live in New York City for giving me a lot of emotional support and for accommodating me when I visit them. I would like to thank many of great friends that I met at Case who accept me as their friends.

List of Abbreviations

1NM-PP1 1-Tert-butyl-3-(naphthalen-1-ylmethyl)-1H-pyrazolo[3,4-d]pyrimidin-4-amine

5HT3Rs 5-hydroxytryptamine type-3 receptors

A322D alanine to aspartic acid

AAV adeno-associated virus

AD Alzheimer's disease

ATF6 Activating transcription factor 6

BDNF Brain-derived neurotrophic factor

BIG2 brefeldin A-inhibited GDP/GTP exchange factor 2 ()

BiP Immunoglobulin Heavy Chain-Binding Protein

BIX BiP protein inducer X

bZIP basic leucine zipper CaMKII Ca 2+ /calmodulin-dependent protein kinase II

CFTR cystic fibrosis transmembrane conductance regulator

cdc-42 cell division control protein 42 homolog

CREB cAMP response element-binding protein

CNS central nervous system

eIF2α translation initiation factor

ELIC the pentameric ligand-gated ion channel from Erwinia chrysanthemi

ER endoplasmic reticulum

ERGIC-53 endoplasmic reticulum-Golgi intermediate compartment protein-53

ERM ezrin, radixin, moesin

FKBP8 FK506 Binding Protein 8

GABAA γ-aminobutyric acid type A

GABARAP GABAAR associated protein

GRIF-1 GABAAR-interacting factor

GLIC Gloeobacter Ligand-gated Ion Channel

GC glucocorebrosidase

GlyRs glycine receptors

GODZ Golgi-specific DHHC (Asp-His-His-Cys) zinc finger protein

GRP78 glucose-regulated protein 78

GRP94 glucose-regulated protein 94

HAP1 Huntingtin-associated protein 1

HDAC7 histone deacetylase 7 hERG human ether-à-go-go related gene

HSP5a Heat Shock 70kDa Protein 5

ICD intracellular loop domain

IRE1 Inositol-requiring enzyme 1

KIF5 kinesin superfamily motor protein 5

LBD ligand-binding domain

LC light chain

LMAN1 Lectin, Mannose Binding 1

LQTS long QT syndrome

LSD lysosomal storage diseases mIPSC miniature inhibitory postsynaptic currents

MuSK muscle-specific tyrosine kinase

xii nAChRs\ nicotinic acetylcholine receptors

NBD nucleotide-binding domain

NLGN3 Neuroligin3

OS-9 Osteosarcoma Amplified 9, Endoplasmic Reticulum Lectin

PDIA4 Protein Disulfide Isomerase Family A Member 4

PE Phosphatidyl ethanolamine

PERK double-stranded RNA-activated protein kinase (PKR)–like ER kinase

PKA protein kinase A

PKC protein kinase C

PLIC1 ubiquitin-like protein 1

PLIC2 ubiquitin-like protein 2 pLGICs pentameric ligand-gated ion channels

PRIP Phospholipase C-related catalytically inactive protein

PTK protein tyrosine kinases

RACK-1 receptor for activated c-kinase

γ2L long splice variant of γ2 subunits ρ1 GABAC

RIC-3 resistance to inhibitors of acetylcholinesterase

ROS reactive oxygen species

SAHA suberoylanilide hydroxamic acid

SBD substrate-binding domain

SRP signal recognition particle

STT3A Dolichyl-Diphosphooligosaccharide--Protein Glycosyltransferase Subunit sXBP1 spliced x-box binding protein 1

xiii

TM transmembrane domain

TRAK1 Trafficking Kinesin Protein 1

TTR transthyretin

Unc-50 Unc-50 Inner Nuclear Membrane RNA Binding Protein

UPR Unfolded Protein Response

UPRE Unfolded Protein Response Element

VILIP-1 Visinin-like protein 1

XBP1 x-box binding protein 1

xiv

Proteostasis Maintenance of γ-aminobutyric Acid Type A Receptors (GABAARs) Abstract

YANLIN FU

Biogenesis of membrane proteins is controlled by the protein homeostasis

(proteostasis) network. Our lab has been focusing on protein quality control of γ-

aminobutyric acid type A receptors (GABAARs), the major inhibitory neurotransmitter- gated ion channels in the mammalian central nervous system (CNS). Proteostasis deficiency in GABAARs causes loss of their surface expression on the plasma membrane

and thus negatively affects GABAARs mediated inhibitory control in CNS, leading to

epilepsy and other neurological diseases. Two well-characterized examples are the A322D

and D219N mutation in the α1 subunit that both cause extensive misfolding and expedited

degradation of the mutant subunit in the endoplasmic reticulum (ER), resulting in familial

epilepsy. In this thesis, I show that modulating the ER proteostasis network regulates the

proteostasis of pathogenic A322D and D219N GABAARs and restores their functional

surface expression. Firstly, I demonstrate that modest activation of the unfolded protein

response (UPR)-ATF6 pathway or UPR-IRE1 pathway enhances the plasma membrane trafficking of the α1(A322D) protein in HEK293T cells. Secondly, I show that application of BIX, a specific potent ER resident HSP70 family protein BiP activator, significantly attenuates the degradation of α1(A322D) subunits, enhances their forward trafficking and increases the functional surface expression of the mutant A322D and D219N GABAARs

in human HEK293T cells and neuronal SH-SY5Y cells. In this thesis, I also show that modulating the ER proteostasis network regulates the proteostasis of wild type (WT)

GABAARs and affects their surface expression. I show that ER-Golgi intermediate compartment protein-53 (ERGIC-53, aka LMAN1), which cycles between the ER and

Golgi, is a trafficking factor for the Cys-loop superfamily of neuroreceptors including WT

GABAARs in the central nervous system. This is the first report of LMAN1 function in membrane protein trafficking as previously LMAN1 is a known cargo receptor for a number of soluble proteins.

xvi

Chapter 1:

Introduction

1.1 General introduction of GABAARs and other cys-loop family receptors

The Cys-loop receptors are activated by neurotransmitters, allowing ion flux through

neuronal cell membrane to maintain the electrical activity of the central nervous system

(CNS) [1]. They include γ-aminobutyric acid type A receptors (GABAARs), nicotinic

acetylcholine receptors (nAChRs), 5-hydroxytryptamine type-3 receptors (5HT3Rs), and

glycine receptors (GlyRs). The Cys-loop receptors are also called pentameric ligand-gated ion channels (pLGICs). The bacterial GLIC and ELIC and the Caenorhabditis elegans

GluCl are also part of this superfamily.

The Cys-loop receptors have prominent roles in the nervous system. As the most studied

member, nAChRs are cation channels, permeable to Na+, K+ and Ca2+ upon activation.

They are responsible for synaptic transmission in the CNS, in autonomic ganglia, and at

neuromuscular junctions and other peripheral synapses. The receptors are involved in

diseases such as tobacco dependence, Alzheimer's disease (AD), bipolar disease, and

myasthenia gravis[2,3]. nAChRs located at different locations are composed of different

sets of subunit subtypes. α1, β1, γ, and δ subunits or α1, β1, δ, and ε subunits form muscle-

type nAChRs at a 2:1:1:1 ratio, whereas (α4)3(β2)2, (α4)2(β2)3, or (α7)5 form 12 different

nicotinic receptor subtypes. α2−α10 and β2−β4 compose the most neuronal-type receptors

[4,5,6,7].

5-HT3Rs, the only inotropic receptor in serotonin receptor family, are also cation

channels. They are widely located at postsynaptic sites in hippocampus, cortex, substantia

nigra, and brain stem. They also exist in the presynaptic GABAergic nerve terminals in the

1 amygdala and CA1 region of the hippocampus, presynaptic glutamatergic synapses, and glial cell membranes in the medial nucleus of the solitary tract. They are involved in many clinical diseases such as drug addiction, cognitive function, schizophrenia, satiety control.

Its antagonists are used to treat post-infectious irritable bowel syndrome and severe diarrhea-predominant irritable bowel syndrome, chemotherapy-induced vomiting and radiotherapy-induced and post-operative nausea and vomiting [7]. The pentameric channels exist either as 5-HT3A homomeric receptors or 5-HT3A/3B heteromeric receptors with a stoichiometry of 3(5-HT3B):2(5-HT3A).

GABAARs are chloride channels. They are one of the main targets for anesthesia, epilepsy, anxiety disorders, mood disorders, and schizophrenia [8]. GABAARs are expressed post-synaptically, mediating phasic inhibition. They are also expressed at perisynaptic and extrasynaptic sites, mediating tonic inhibition [9]. There are abundant interchanges between the receptors locating at postsynaptic and extrasynaptic sites. To date, there are 19 GABAAR subunits belonging to eight classes based on sequence identity. They are α(1–6), β(1–3), γ(1–3), δ, ε, π, θ, and ρ(1–3) [10]. There are alternatively spliced variants of several of these subunits. For example, a short form (γ2S) and a long form (γ2L) of γ2 subunits exist, and their difference is that an eight-amino acid insert exists in the ICD of the γ2L subunit [11,12]. The majority of GABAAR subtypes expressed in the brain are composed of α1β2γ2, then α2β3γ2 and α3β3γ2, which form the stoichiometry of 2α:2β:1γ

[13].

Recently, high-resolution structures of the Cys-loop receptors including nAChR [14],

GluCl [15], GLIC [16], ELIC [17], 5-HT3R [18,19], GABAAR [20,21,22], and GlyR [23], have been elucidated. The common structural feature of this superfamily is that five

2 subunits form the receptor (Figure 1A). Each subunit has a large extracellular N-terminal domain, four transmembrane (TM) helices (M1-M4), and a large intracellular loop domain

(ICD) linking M3 and M4 (Figure 1B). The signature disulfide bond is formed by two cysteine residues, which are separated by 13 residues. This Cys-loop structure is important in the inter-subunit assembly because blocking its formation negatively affects the receptor assembly [24]. The N-terminal domains of the five subunits form the ligand-binding domain (LBD), which lies in the interfaces of adjacent subunits. The M2 transmembrane helices from five subunits form the channel pore, which allows the flux of specific ions.

M1 and M3 helices surround the M2, and M4 locates in the outermost area of the channel pore. The ICD between M3 and M4 is important for subunit clustering in the cell membrane.

The TM domains play an important role in channel folding, assembly, and gating.

FIGURE 1. Structural characteristics of the Cys-loop receptors. (A) The Cys-loop receptors are pentameric, forming a central ion pore. (B) Each subunit has a large extracellular domain, four transmembrane helices, and a large intracellular loop domain

(ICD) between TM3 and TM4. The two cysteines that form the signature disulfide bond are shown in sphere model. The cartoons are built from the crystal structures of GABAA receptors (4COF).

Figure 1

1.2 Factors that affect GABAARs surface expression

1.2.1 Folding and assembly of Cys-loop receptors

The correct synthesis and folding of individual subunits and the subunit assembly of specific forms are required for them to exit the ER for subsequent trafficking to the Golgi and plasma membrane. This is evidenced first by previous studies showing that only certain assembly of subunits can form functional surface receptor. Expression of α1, β2 or the long splice variant of γ2 subunits (γ2L) of GABAARs alone in the heterologous cells can lead to the formation of homomeric assemblies but they fail to exit the ER [25]. Co- expression of α and β but not α and γ or β and γ can lead to limited functional surface expression of the receptors [8]. When α, β, and γ subunits are coexpressed, the formation of 2 α and 2 β and 1γ subunit is strongly favored against other forms [8]. The preference of formation for certain assembly receptor subtypes may be due to the fact that forming the correct assembly structure hides the ER retention signal in the single receptor subunits. The

γ2L subunits containing an eight amino acid ER retention signal are retained in the ER when expressed alone, whereas the γ2s subunits without this retention signal are able to exit the ER and translocate onto cell surface even when expressed by themselves [26]. The

5-HT3B subunits cannot form a homopentamer since this subunit contains the ER retrieval signal, which can only be masked in the presence of the 5-HT3A subunits [27]. Mutation of a motif within a conserved transmembrane domain of nAChR subunits enables them to exit the ER, whereas insertion of this motif to proteins that originally successfully transported to cell surface makes them retained in ER. Assembly of native nAChR subunits into pentameric receptors covers this motif, leading to successful traffick from the ER to cell surface [28].

Although it is essential for the Cys-loop receptors to acquire their correct folding and

assembly status, these processes are difficult because each receptor, being a pentamer, has

a large molecular weight, which is about 250 kDa, and each subunit has multiple

transmembrane domains. As a result, the assembly process is generally inefficient and slow.

Only 25% of newly synthesized GABAARs are assembled into heteromeric receptors, and

30% of the translated α subunits of nAChRs are assembled [29]. The half-life of the

nAChR assembly is more than 90 mins, much longer than 7-10 mins, the half-life of

influenza hemagglutinin to form homotrimers [29]. The Green group has determined the

assembly models of nAChR by using pulse chase and coimmunoprecipitation assays with

subunit sequence-specific antibodies [29]. However, no folding and assembly models of

other Cys-loop receptors are available yet.

The assembly of Cys-loop receptors depends on the N-terminal signal. The N-terminal

extension and putative α-helix in the α1, β2 and γ 2 subunits of GABAARs are required for

the inter-subunit assembly and thus can affect the cell surface expression level of the receptors [30]. Also, N-terminal extension and α-helix of ρ1 GABAC receptors, which also

belong to Cys-loop receptor family, are also required for the normal assembly, trafficking

and cell surface expression of the receptors [31]. Previous studies determined the specific

amino acids located at the N terminus that are important for the subunit assembly for

GABAARs, nAChRs [32,33,34,35] and GlyRs [32,36]. However, the assembly of 5HT3Rs

[37], nAChRs [38], GlyRs [36] but not GABAARs [39] depends on N-glycosylation status.

In addition, a recent study showed that C-terminal motifs in nAChRs may also be important

for subunit assembly [40]. A highly conserved aspartate residue at the boundary of the M3-

M4 loop and the M4 domain is required for GABAAR surface expression [40]

Many chaperones play a critical role in the folding and assembly process of the Cys- loop receptors. The binding of a chaperone to the unassembled or unfolded proteins stabilizes the folding intermediates and increases their success rate of proper folding and assembly. BiP (also known as Grp78), an Hsp70 family protein in the ER, binds the hydrophobic patches of a protein. Overexpression of BiP promotes the maturation of WT

α1 subnits of GABAARs. BiP associates more strongly to misfolded mutant GABAA receptors harboring an A322D mutation in the α1 subunit compared to the wild type receptors [41], indicating that BiP acts early in the protein folding step by binding to the unfolded proteins. Consistently, BiP associates more strongly with unassembled nAChRs subunits [29]. Calnexin, an ER membrane bound L-type lectin protein, checks the protein folding status by recognizing the specific glycan structures on the polypeptide.

Overexpression of Calnexin promotes the maturation of WT α1 subnits of GABAARs

[41,42]. L-type calcium channel blockers promotes the trafficking of misfolding-prone mutant α1 subunit harboring the D219N mutation of GABAA receptors by increasing its interaction with calnexin as a result of increased calcium concentration in the ER after L- type calcium channel blockers application . ERp57, a protein disulfide isomerase, and calreticulin, an ER soluble homologue of calnexin, associate with nAChRs subunits and may promote the subunit stability [29,43]. RIC-3 (resistance to inhibitors of acetylcholinesterase 3) is an ER localized transmembrane protein and serves as a chaperone for 5HT3Rs. It enhances the folding, assembly and ER exit of 5HT3R [44,45]. However,

RIC-3’s effect on nAChR is relatively unclear yet. Overexpression of RIC-3 enhances the surface expression of α7-nAChRs but reduces that of α4β2-nAChRs by inhibiting the trafficking of the receptors onto the cell surface [46].

For WT cys-loop receptors, both agonists and antagonists can act as pharmacological

chaperones. For example, nicotine and its metabolite cotinine upregulate the surface

expression level of nAChRs by serving as pharmacological chaperones, promoting the

stabilization of the nAChRs in the ER (Fox, Moonschi, & Richards, 2015; Lester et al.,

2009). Similarly, GABAAR agonists and a competitive antagonist bicuculline enhance the

surface expression level of GABAARs by acting as pharmacological chaperones. The

application of brefeldin A, which inhibits the formation of COPI-mediated transport

vesicles from ER to Golgi, antagonizes this effect (Eshaq et al., 2010).

1.2.2 ER associated degradation (ERAD) of the Cys-loop receptors.

The folding and assembly process of the Cys-loop receptors is slow with a high failure rate. The subunits that fail to assemble or fold are degraded by ERAD [47,48,49]. Cells utilize this classical pathway to recognize and ubiquitinate unfolded proteins in the ER,

extract them to cytosol, and deliver them to the protein degradation complex in the cytosol called the proteasome. This whole process is accomplished with the synchronized action of a series of both soluble and ER membrane-bound chaperone proteins and cytosolic

chaperones, which can be collectively called ERAD machinery.

ERAD influences the trafficking and cell surface expression levels of the Cys-loop

receptors. The ubiquitin-like protein PLIC1 negatively regulates GABAAR degradation by

inhibiting ubiquitination (Tsetlin et al., 2011). PLIC1 and its paralog PLIC2 share an

ubiquitin-like proteasome-binding domain. The association of this domain with the ICD of

GABAAR subunits slows their ubiquitination and enhances their functional surface

expression (Bedford et al., 2001; Luscher et al., 2011; Wu, Wang, Zheleznyak, & Brown,

1999). Ring finger protein 34, an E3 ubiquitin ligase, interacts with the ICD of the γ2

subunits of GABAARs and reduces their expression by promoting the degradation of the

receptors through both lysosome and proteasome degradation pathways (Jin et al., 2014).

For nAChRs, blockage of the proteasome function increases their assembly in the ER,

leading to their enhanced surface expression in cultured myotubes (Christianson & Green,

2004; Wanamaker et al., 2003). Long-term inhibition of neuronal activity drastically enhances the ubiquitination level of GABAARs and decreases their cell surface stability,

whereas increasing the level of neuronal activity decreases the ubiquitination of GABAARs

and promotes their stability on the plasma membrane. Neuron activity itself can regulate

the potency of GABAAR-mediated effects through ubiquitination (Saliba, Michels, Jacob,

Pangalos, & Moss, 2007). It will be of great interest to elucidate the ERAD machinery,

such as critical E3 ligases and retrotranslocation channels, for the Cys-loop receptors. Cells

can also clear misfolded proteins through the autophagy pathway[50]. However, there is no report that cys-loop receptors can degraded through this pathway so far.

1.2.3 Factors that regulate surface trafficking

Golgi-specific DHHC (Asp-His-His-Cys) zinc finger protein (GODZ), which belongs

to DHHC family palmitoyl acyltransferases, specifically palmitoylates the γ2 subunits of

GABAARs. The palmitoylation is required for targeting the receptors to inhibitory

synapses. Knockdown of GODZ causes a selective loss of GABAARs, thus leading to

reduced GABAAR-medicated miniature inhibitory synaptic current amplitude and

frequency [51,52].

The brefeldin A-inhibited GDP/GTP exchange factor 2 (BIG2) interacts with the ICD

of β subunits of GABAARs. It enhances the trafficking of β3-containing GABAARs by promoting the membrane budding of vesicles from Golgi apparatus [53].

The GABAAR associated protein (GABARAP), which belongs to a ubiquitin-like

family protein in mammals and is enriched in Golgi and other somatodendritic membrane

compartments, facilitates the trafficking of GABAARs in hippocampus neurons through connecting the γ subunits with microtubules [54,55]. This GABARAP effect depends on

the interaction of phospholipids with GABARAP [56].

Phospholipase C-related catalytically inactive protein (PRIP) is an inositol 1,4,5-

trisphosphate binding protein. It may serve as a bridge protein which connects γ2-

containing GABAARs with GABARAP and promotes the trafficking of the receptors.

Interrupting the interaction of PRIP with γ2 subunits of GABAARs decreases the surface

expression level of the receptors in both cultured cell lines and neurons [57].

Visinin-like protein 1 (VILIP-1), a neuronal protein, enhances the surface expression

of α4β2-nAChRs in hippocampal neurons by promoting their exit from the trans-Golgi

network. This effect is activated by increasing intracellular Ca2+. As a result, it is an

important factor that mediates the neuron activity induced changes in surface expression

level of the receptors [58].

Protein Unc-50 Inner Nuclear Membrane RNA Binding Protein (Unc-50), which is

found in the nematode C elegans but is evolutionarily conserved, is needed for the transport

of specific types of nAChRs onto the cell surface by an unknown mechanism [59].

1.2.4 Clustering

Insertion of Cys-loop receptors into the plasma membrane is also tightly regulated. This

process is important for shaping the post-synaptic sites types and regulating the receptors-

mediated inhibitory or excitatory effects.

Gephyrin regulates the clustering of GlyRs and GABAARs. Gephryin is a scaffolding

protein that mainly accumulates in inhibitory GABAergic and glycinergic synapses in

various brain regions. Gephyrin-induced clustering of GABAARs is subunit-specific.

Gephyrin knockout in mice diminishes the number of α2, α3, β2/3, γ2 subunit-containing synaptic sites, whereas it does not affect the α1,α5 containing synaptic sites without

affecting the number of total inhibitory synaptic sites [60]. This could be due to the fact

that there are only certain types of GABAAR subunits that can associate with gephyrin.

Gephyrin E domain associates with a 10-amino acid hydrophobic motif within the

intracellular domain of the GABAAR α2, α3, γ2 and β3 subunits [61]. Gephyrin is also

important in regulating the neuron activity plasticity. Long term inhibitory potentiation of

neurons in the visual cortex increases GABAAR-mediated inhibitory post-synaptic currents

S383 by inducing the CaMKII phosphorylation of the GABAAR β3 residue and enhances

gephyrin clustering of β3-containing GABAARs. Collybistin, a guanidine exchange factor

activating cell division control protein 42 homolog (cdc-42), forms a binding complex with

gephyrin. Knockout of collybistin in mice does not affect glycinergic synaptic transmission

but decreases GABAergic synaptic transmission. Collybistin is not required for gephyrin-

mediated GlyR clustering but necessary for gephyrin-mediated clustering of certain

GABAARs at inhibitory postsynaptic sites [62,63,64].

GABAARs clustering is also mediated by gephyrin-independent pathways. Radixin,

which belongs to ERM (ezrin, radixin, moesin) family proteins, is known mediate the

clustering of α5-containing GABAARs. Depleting radixin or changing the radixin F-actin

binding motif in neurons disrupts the formation of α5 subunit-containing GABAAR

clustering [65].

The clustering of nAChR in the neuromuscular junction depends on agrin, a heparan

sulfate proteoglycan secreted by the presynaptic motor neuron, and rapsyn, an intracellular

scaffolding protein for Wnt signal. Agrin activates the muscle-specific tyrosine kinase

MuSK under the assistance of rapsyn, resulting in the phosphorylation of the β-subunit of

nAChRs and the local receptor clustering at the nerve terminus [66,67]. 14-3-3 proteins,

which, as mentioned above, assist the assembly of α4 subunit-containing nAChRs, could

also be involved in the clustering of α3-containing nAChRs at synapses on the surfaces of

ganglionic neurons [68].

1.2.5 Endocytosis

Surface receptors undergo constant recycling between the cell surface and intracellular

endosomes [26,69]. The internalized receptors are either recycled back onto the cell surface

through early and recycling endosomes or degraded through late endosomes in the

lysosomes. The regulation of the balance between the internalization and

recycling/degradation is also important in regulating the availability of the surface

expression of receptors and their roles in mediating neuronal excitatory or inhibitory effects.

For GABAARs, clathrin adaptor protein AP2 binds to the β and γ subunits, which in

turn interact with clathrin, the GTPase dynamin, and other binding partners and form the

GABAARs containing clathrin-coated pits [70].

Many important factors regulate the endocytosis and recycling process of Cys-loop

receptors. For GABAARs, Huntingtin-associated protein 1 (HAP1), which is an adaptor protein for the kinesin superfamily motor protein 5 (KIF5) [71], inhibits the degradation of endocytosed β1-3-containing GABAARs through the KIF5-dependent trafficking, favors the receptor recycling, and increases their surface expression and receptor-mediated inhibitory effect [72]. GABAAR-interacting factor, GRIF-1, and its paralog TRAK1, also

interact with KIF5. They could be involved in the KIF5 dependent trafficking of GABAARs

[8]. Inhibiting the lysosomal activity [72,73], preventing the trafficking of ubiquitinated γ2

subunit-containing GABAARs to lysosomes [73], or disrupting the ubiquitination at lysine

residues in the intracellular domain of the γ2 subunit [73] enhances the accumulation of

GABAARs at synapses.

Giant ankyrin-G (aka ANK3), an extended fibrous polypeptide with 2,600 residues, is

present in extrasynaptic microdomains on the somatodendritic surfaces of hippocampal and

cortical neurons and disrupts GABAAR endocytosis by interacting with the GABARAP

[74]. This process may be involved in the formation of GABAAR-mediated circuitry in the

cerebral cortex. Human mutations in the giant ankyrin exon are linked to autism and severe

cognitive dysfunction [75].

The internalization rate also depends on the extracellular conformation of the

GABAARs and the presence of GABAAR agonists or antagonists. GABAARs that contain

the R43Q mutant γ2 subunits have an increased clathrin-mediated and dynamin-dependent

endocytosis, which can be reduced by receptor antagonists. Furthermore, receptor agonists

enhance the endocytosis of both endogenous and recombinant wild type GABAARs in both

cultured neurons and COS-7 cells [76].

The nAChR agonist, antagonist α-bungarotoxin, and cross-linking anti-nAChR

antibodies promote the internalization of nAChRs [77,78,79]. This process depends on

actin activation, but it still happens without functional clathrin, caveolin, or dynamin [79].

Neuregulins 1β (NRG1β), which belongs to EGF family, induces the rapid internalization

of α7-nAChRs from the surface of these neurons. Its effect relies on tyrosine

phosphorylation and activation of actin cytoskeleton.

1.2.6 Other regulations of Cys-loop receptors

1.2.6.1 Lipid involvement in trafficking and clustering

Phosphatidyl ethanolamine (PE) is required for the surface expression of GABAARS

in cultured neurons under the assistance of GABARAP [80]. Membrane sphingolipids and other lipids promote the surface expression level of muscle-type nAChRs by affecting the biosynthesis process in the ER [81]. Decreasing the membrane cholesterol promotes the endocytosis of nAChRs and decreases their cell expression level [82]. The underlying mechanism is that membrane lipid serves as lipid rafts, which is required for the trafficking and membrane stabilization of the receptors.

1.2.6.2 Phosphorylation signaling in the biogenesis of the receptors

Phosphorylation affects the Cys-loop receptor channel properties [83] and modulates

the efficacy of receptor-mediated effect by influencing their trafficking, endocytosis, and

14 recycling process. Neuronal activities that lead to the change in the intracellular calcium signal regulate the activity of kinases and phosphatases, resulting in an altered the biogenesis process and thus the surface expression level of the receptors. For example, enhanced excitatory synaptic activities activates the phosphatase calcineurin through the

Ca2+/calmodulin upon an increase in intracellular Ca2+ concentration. Activated calcineurin dephosphorylates Ser327 in the GABAAR γ2 subunit, which leads to the enhanced lateral mobility of the receptors, decreased cluster size of GABAARs, and reduced GABAergic mIPSC [84]. Calcineurin is also involved in downregulation of the

α2-containing GABAAR membrane expression level in prolonged seizures activity linked to benzodiazepine pharmacoresistance [85]. PRIP, as mentioned above, modulates the

GABAAR surface expression level by affecting the phosphorylation of the receptors. PRIP inactivates the protein phosphatase 1α (PP1α), which dephosphorylates the GABAARs phosphorylated by protein kinase A (PKA). As a result, PRIP positively regulates the receptor surface expression and receptor-mediated inhibition effect in hippocampal neuron

[86,87,88].

Many neurosteroids or neurotrophic factors regulate the surface expression level of receptor by affecting the trafficking, endocytosis and recycling process. For example, neurosteroids promote the PKC phosphorylation of the α4 subunit Ser443 site, which enhances the insertion of the α4 subunit-containing GABAARs and leads to increased tonic inhibition [89]. However, the same neurosteroid does not have any effect on the α1 and α5- containing GABAARs, which mediate the phasic inhibition [89,90] [91]. Brain-derived neurotrophic factor (BDNF) induces an initial fast but short increases in GABAARs- induced mIPSC through the phosphorylation of β3 Ser408/409 by PKC and RACK-1

(receptor for activated c-kinase), which leads to decreased endocytosis of the receptors. A

following long lasting downregulation of GABAARs-induced mIPSC is due to increased

clathrin-mediated endocytosis of GABAARs by dephosphorylating β3 subunits of

GABAARs [92].

Phosphorylation also affects the trafficking, endocytosis and recycling process of

nAChRs and 5-HT3Rs. For example, inhibition of protein tyrosine kinases (PTKs)

enhances α7-nAChR-mediated responses to ACh both in oocytes and in hippocampal

neurons. The application of a protein tyrosine phosphatase inhibitor leads to the depression

of such responses. PTKs promote the exocytosis of α7-containing nAChRs [93]. Protein

tyrosine phosphatases enhance the turnover rate of nAChRs and they are required for

proper recycling of nAChRs onto the cell surface, whereas activation of the

serine/threonine protein kinase PKA slows the turnover of nAChRs [94,95,96,97]. PKC

enhances the trafficking of the 5HT3Rs onto the cell surface and this effect is mediated

through an actin-dependent pathway [98].

1.2.7 Mutations in GABAAR leads to their reduced surface expression level and

diseases.

1.2.7.1 Current mutations that cause reduced surface expression

Mutations in GABRA1, GABRB3, GABRG2, and GABRD genes that encode inhibitory

GABAARs lead to idiopathic epilepsy syndromes which represent about half of all the

epilepsies. Some of the mutations cause receptor dysfunction but others cause the epilepsy

phenotype by affecting the surface expression level of the GABAARs. Gallagher et al has

reviewed the mutations in GABAAR subunits that cause genetic epilepsies [99]. For

example, 975delC and S326fs328X mutations in GABRA1 lead to mRNA degradation and

reduce their protein products. P11S, S15F and G32R mutations in GABRB3 lead to abnormal N-linked glycosylation and reduced surface expression level of the mutant

receptors[99]. Other pathogenic mutations affect subunit folding or receptor assembly,

resulting in loss of functional surface expression of the Cys-loop receptors. For example, the R43Q mutation in the γ2 subunit of GABAARs affects its association with the αβ

subunit complex, leading to its retention in the ER [100]. GABAARs containing only αβ

subunits have reduced channel function, leading to childhood absence epilepsy and febrile

seizure. The D219N and A322D mutations in the α1 subunit of GABAARs are linked to

familial juvenile myoclonic epilepsy by affecting the folding and assembly of the subunit,

which leads to their enhanced ERAD and impaired surface expression [101,102]. The

R177G mutation in the γ2 subunits undermines subunit folding or assembly and leads to

epilepsy phenotype [103]. For nAChRs, β4R348C negatively affects the ER exit of

nAChRs and leads to reduced agonist-induced currents and amyotrophic lateral sclerosis

(ALS) [104]. The S143L, C128S, and R147L mutations located at N-terminal extracellular

domain of ϵ subunits for nAChRs influence the subunit assembly and are linked to

congenital myasthenic syndromes [105].

Expression of A322D but not D219N GABAARs can affect the formation of GABAergic

boutons and dendritic spines such as increasing the spine density and large mature

mushroom-like spines in cortical pyramidal cells and increasing boutons formed in

GABAergic cortical basket cells. A322D α1 subunits are believed to interact with and

inhibit the trafficking of other WT GABAAR subunits[106].

1.2.7.2 Pathogenesis of A322D and D219N GABAARs

Endoplasmic reticulum is a cisternae type organelle that exists in all eukaryotic cells

except red blood cells and spermatozoa. ER connects with the outer nuclear membrane and

may also directly connect with mitochondria. ER is separated into rough ER and smooth

ER. In rough ER, ribosomes are associated with the outer layer of the ER membrane which is responsible for the protein synthesis and newly synthesized polypeptide will be imported into ER through translocon Sec61 complex. Smooth ER does not have ribosome interactions and is responsible for lipid synthesis.

The newly synthesized polypeptides of secretory or membrane proteins expose their amino terminal of signal peptide at the tunnel exit of its-bound ribosome to bind to the signal recognition particle (SRP) which is a cytosolic ribonucleoprotein complex. SRP then associates with SRP receptor located on ER membrane and helps dock the ribosome- nascent chain complex to ER membrane. The newly synthesized polypeptide then enters the ER through the translocation pore comprised of Sec61 complex, the exposed polypeptide in ER quickly associates with important ER resident chaperones like BiP. BiP is localized at the ER luminal end of translocation pore and uses an ATP-dependent mechanism to seal the pore, to associate with newly synthesized polypeptide segment to prevent premature folding, and to serve as a driving force together with HSP40 chaperone

Sec63p to drag newly synthesized polypeptide through the translocation pore[107]. Sec63p is an ER membrane protein which interacts with BiP through its luminal J domain, helps

BiP to localize to the ER membrane and facilitates BiP’s interaction with substrate peptide through stimulating BiP ATP hydrolysis[107]. With the elongation of the newly

synthesized polypeptide, chaperone proteins will dissociate from their substrate which

allows it to fold into native conformation or assembles with other subunits if it is a part of

multi-subunit proteins[108]. Only correctly folded and assembled proteins are able to pass

the ER quality control system and are recruited into COPII-coated vesicles to be transported out of ER into Golgi. The proteins then get further processed in the Golgi and sent to their final designated working places to perform their functions. However, if the proteins are not folded correctly, ER has a series of checking systems to first identify those proteins, assist with folding through association and dissociation with chaperones. If the proteins fail to reach its correct conformation, then they are cleared through the ER associated degradation pathway or lysosome degradation pathway. The first type of the

ER quality control system include ER resident chaperone proteins such as HSP70 homologue BiP and its various HSP40 ER homologue co-chaperones, HSP90 homologue

GRP94, protein disulfide isomerase family proteins, all of which are reported to send their misfolded or unfolded substrates for degradation. The second type of ER quality control system includes the lectin chaperones: calnexin and calreticulin. N-linked glycosylation is important in the folding of glycoproteins. If the N-linked glycans are near critical cysteins, they can help recruit calnexin and calreticulin to block the cysteine residues and prevent premature formation of disulfide bonds. Calnexin and calreticulin are lectin chaperones

that recognize mono-glucosylated N-glycans on a glycoprotein and promote the folding of

their substrates together with protein-disulfide isomerases, such as ERp57 [109].

Terminally unfolded proteins are targeted for ER associated degradation.

A322D mutation was first identified in a family with autosomal dominant juvenile

myoclonic epilepsy. A322D mutant GABAARs subunits have a significantly lowered total

and surface expression level. The few remaining A322D mutant subunits are retained in the ER. Substitution of a charged residue to hydrophobic residue in the transmembrane domain increases the free energy cost that is required for M3 insertion. When the A322 is substituted with many other amino acids, the decreased hydrophobicity of the

transmembrane domain is correlated with an increased free energy required for membrane insertion. The failure of M3 insertion into the ER membrane leads to its exposure in the

ER lumen. A tandem mass spectrometry-based proteomics approach identified potential proteostasis network components for GABAA receptors, enabling follow-up studies on

their ERAD machinery [110]. GRP94 and osteosarcoma amplified 9 (OS-9), an ER lectin

protein, have been identified as degradation factors for A322D α1 subunits. OS-9 binds to

the misfolded subunits in a glycan-dependent manner and together with Grp94 directs

misfolded α1 subunits to Hrd1, an E3 ligase on the ER membrane, for ubiquitination. VCP

is a type II member of AAA ATPase. Its prominent function is to extract the ubiquitinated

misfolded proteins from the ER and route them to the cytosolic proteasome for degradation.

VCP facilitates the dislocation of the α1 subunits from the ER membrane to the cytosol,

targeting them to the 26S proteasome for degradation. A322D mutants are efficiently

degraded by the ER associated degradation pathway with a much faster degradation rate as

compared to WT alpha1 subunits. The half-life of A322D alpha1 subunits are about 30mins

compared to about 2hrs of WT alpha1 subunits.

The aspartic acid to asparagine mutation at position 219 in human α1 subunits was first

identified by Cossette et al [111]. The mutation negatively affects the interaction between

Asp219 and Lys 247 and leads to misfolding of the N-terminus domain. D219N GABAARs

have a reduced surface expression level and impaired inhibitory control of neuronal circuits

[42,111].

1.3 Protein Export efficiency model and therapy strategies

Balch et al has proposed a model of folding for export (FoldEx) which indicates that the

efficiency of protein export out of ER is dependent on folding and misfolding energetics

and the proteostasis network that regulates folding, degradation, and ER export[112]. The normal ER environment leads to the fast elimination of trafficking deficient mutations.

However, this model and many current results suggest that by improving the activities or

expression level of chaperones in the ER and by inhibiting ERAD pathways, the ER export efficiency can be significantly improved for trafficking deficient mutant proteins.

GABAARs have a slow efficiency in completing the folding and assembly process which

makes them more affected by ERAD compared to those fast folding proteins. Many studies

also have indicated that facilitating the mutant receptors to overcome the energy barrier

they encounter by modulating receptor specific chaperone and co-chaperone levels allows

them to acquire the native structure which they fail to reach under normal conditions. Since most of the mutant proteins retain normal function, enhanced ER export and membrane expression will alleviate symptoms drastically.

Slowing the degradation of trafficking deficient mutant: Inhibiting VCP using eeyarestatin I significantly enhances the trafficking of both wild type and mutant α1 subunits harboring the A322D mutation of GABAARs [113]. Knocking down GRP94 can significantly promote the surface trafficking of A322D GABAARS. Inhibiting HRD1, an

E3 ligase for A322D GABAARS also have a similar effect. Based on the above evidence,

modulating the ERAD rate is a promising way to enhance the surface trafficking of Cys- loop receptors.

Two classes of small molecules have been employed: proteostasis regulators and

pharmacological chaperones can promote the correct folding of trafficking deficient

proteins, promote their forward trafficking and relieve symptoms caused by these

mutations [114,115].

Proteostasis regulators operate on the proteostasis network components to correct

the folding and trafficking deficiency. For example, suberanilohydroxamic acid, acting as

a proteostasis regulator, enhances the functional cell surface expression of the A322D α1

subunit of GABAARs partially by increasing the BiP protein level and the interaction

between calnexin and the mutant α1 subunit in the ER [41]. SAHA has also been shown

to enhance the maturation and trafficking of the trafficking deficient mutant CFTR,

lysosomal glucocerebrosidase, and a1-antitrypsin. The effect is not due to BiP and calenxin

overexpression [116,117,118]. Verapamil, an L-type calcium channel blocker, acting as a

proteostasis regulator, enhances the function of the D219N α1 subunit of GABAARs by

promoting calnexin-assisted folding [42,102]. F805C mutation in human ether-à-go-go

related gene (hERG) potassium channel leads to its decreased surface expression.

Decreased hERG potassium channel expression on membranes of cardiomyocytes leads to

congenital long QT syndrome (LQTS) [119]. FK506 Binding Protein 8 (FKBP38) serves

as a chaperone protein to enhance surface trafficking of trafficking deficient F805C hERG

[120].

Pharmacological chaperones directly bind the receptors, stabilize the assembly intermediates, increase the successful rate of this process, and promote the surface expression level of the receptors. One allele of the Phe508 deletion in NBD1 domain of

cystic fibrosis transmembrane conductance regulator (CFTR) impairs the folding of the channels kinetically, prevents the channels from forming native structure and being incoportated into COPII vesicles and thus leads to their degradation through ERAD. As a result, the ΔF508 CFTR fails to reach the apical membrane of polarized epithelia lining many tissues, which leads to the development of cystic fibrosis [121,122,123]. Lumacaftor

(VX-809) is proved to modulate the conformation MSD1 domain and thus help correct the folding of ΔF508-CFTR and disease-related MSD1 mutants. Moreover, VX809 together with other small molecules that stabilize the NBD1 or NBD1:ICL4 interface can significantly increase the membrane trafficking of F508del-CFTR [124]. A series of

clinical trials show that small molecule CFTR correctors together with CFTR channel

potentiator can improve the lung function in patients with ΔF508-CFTR mutation

[125,126]. Another trafficking deficient mutation N470D in human ether-à-go-go related

gene (hERG) potassium channel leads to its misfolding, prolonged binding to calnexin and

ER retention. Decreased hERG potassium channel expression on membranes of

cardiomyocytes leads to congenital long QT syndrome (LQTS) [119]. E-4031 serve as a

chemical chaperone to promote the correct folding of the mutant channel and enhance its

surface expression onto cardiomyocytes [127].

Combining proteostasis regulators and pharmacological chaperones is expected to achieve better therapeutic effects. For example, coapplication of suberanilohydroxamic

acid, a proteostasis regulator, with eeyarestatin I additively promotes the forward

trafficking of misfolding prone α1 subunit harboring the A322D mutation of GABAARs

and enhances their functional cell surface expression [113].

Gaucher disease, which is the most common lysosomal storage diseases (LSD), is

caused by N370S or L444P mutation in glucocerebrosidase (GC) that are misfolded and extensively degraded through ER associated degradation pathway. As a result, mutant glucocerebrosidase fail to reach the lysosomes and leads to the accumulation of

glucosylceramide and neuropathic Gaucher disease. α subunit G269s mutation in β-

hexosaminidase A (HexA) leads to its misfolding and elimination by ER associated

degradation pathway, resulting in neuronal accumulation of GM2 gangliosides and causes

Tay-Sachs disease. Application of proteostasis regulators (PR) such as Celastrol or MG-

132 rescue the folding, trafficking of L444P GC by activation of UPR-PERK and UPR-

IRE1 pathway but not UPR-ATF6 pathway. Celastrol or MG-132 also rescues the forward

trafficking of αG269S HexA. Coapplication of Celastrol or MG-132 with pharmacological

chaperones act synergically to rescue the forward trafficking of N370S and L444P

GC[114].

Summary of thesis

It was previously reported that replacement of asparagine at the 322 position with

aspartic acid in the M3 domain of the α1 subunit of GABAARs decreases the membrane

insertion efficiency of the M3 domain, leading to its exposure into ER lumen. The aspartic

acid to asparagine at the 219 position causes misfolding of the N-terminus. As a result,

A322D and D219 GABAARs are degraded at a much faster rates compared to wild type

receptors through the ER associated degradation pathway which results in their decreased

surface expression level and reduced GABA induced currents. This thesis shows that

modulating the proteostasis network in endoplasmic reticulum by activating the unfolded

protein response ATF6 pathway or sXBP1 pathway, or by application of the small molecule BiX, a BiP specific regulator, helps promote the forward trafficking of the two

misfolding trafficking deficient mutant A322D and D219N GABAARs and increases their

functional surface level. In the last chapter, a new trafficking factor LMAN1 (aka ERGIC

53) in the ER is identified that positively affects the forward trafficking of WT GABAARs

and other Cys-loop family receptors.

FIGURE 2. Protein biogenesis pathway of the Cys-loop receptors. The receptor subunit proteins are co-translationally translocated onto the ER membrane. Molecular chaperones both in the ER and in the cytosol assist their folding. Properly folded subunits assemble into a pentamer, which is then transported from the ER to Golgi and to the plasma membrane. Misfolded proteins and unassembled subunits are degraded by the ER-

associated degradation pathway. The receptors on the plasma membrane undergo

endocytosis.

Figure 2

Chapter 2 Modulating UPR ATF6 pathway and sXBP1 pathway promote the

surface trafficking of A322D GABAARs

The content was published in Y.L. Fu, D.Y. Han, Y.J. Wang, X.J. Di, H.B. Yu, T.W. Mu, Remodeling the endoplasmic reticulum proteostasis network restores proteostasis of pathogenic GABAA receptors, PLoS One 13 (2018) e0207948 and was supported by the National Institue of Heealth (RO1NS105789 to TM)

2.1 Introduction:

Endoplasmic reticulum is the important organelle in the cell that is responsible for protein

synthesis, folding, and assembly. Correctly folded and assembled proteins are then

transported out of the ER into the Golgi through COPII-coated vesicles for further processing. ER is also responsible for Ca2+ storage and lipid and carbohydrate metabolism.

Many factors can disturb ER function including ageing, redox imbalance, calcium disturbance, mutations that cause proteins to misfold and aggregate, inhibition of protein glycosylation and other intrinsic or extrinsic environmental changes that disrupt the optimal physiological conditions for ER proteins to maintain their normal functions. As a result of increased misfolded and unfolded protein burden, ER will activate three pathways to cope with this stress, which is called unfolded protein response. The general effect of

UPR is to decrease the burden of ER by decreasing the new protein synthesis, increasing the expression level of factors that can promote folding, assembly, trafficking and facilitate degradation through ER associated degradation and the autophagy pathway. Early UPR activation is pro-survival. However, sustained UPR activation will leads to cell apoptosis in order to guarantee that malfunctioning cells will not disturb the whole body, as long term

UPR activation indicates that cells are incapable of adjusting the cell conditions back to physiological conditions. UPR executes these functions by manipulating its three branches:

IRE1 (inositol requiring enzyme 1), PERK [double-stranded RNA-activated protein kinase

(PKR)–like ER kinase], and ATF6 (activating transcription factor 6).

The IRE1 pathway is the most conserved UPR pathway among the three pathways and is the only pathway that exists in yeast, metazoans, to humans while ATF6 and PERK pathways only exist in higher eukaryotes like metazoans and humans. IRE1 is an ER transmembrane kinase/endoribonuclease in humans with its Kinase/RNase domain localized in cytoplasm. It has two isoforms: IRE1α and IRE1β. Under ER stress. Both

IRE1α and IRE1β will undergo lateral oligomerization which leads to their autophosphorylation and conformational change and, as a result, their endoribonuclease activity is activated. Activated IRE1α and IRE1β then cut two sites in the (x-box binding protein 1) XBP1 mRNA in cytoplasm [128,129]. IRE1α has a much higher XBP1 mRNA cleavage ability while IRE1β has a weaker one. A 26 bp intron region of XBP1 is cleaved and the two cut sites are then ligated by an unknown RNA ligase in cytoplasm to form a complete exon and to generate the spliced form of XBP1 mRNA [130]. The removal of the intron segment and the frame shift allows the spliced XBP1 to produce sXBP1 protein which is a leucine-zipper-containing transcription factor which binds to unfolded protein response element (UPRE) in the promoters of ER chaperone genes and ERAD factor genes to promote their expression. IRE1β’s RNase activity can also cleave 28S rRNA and leads to partial translational attenuation [131]. sXBP1 then translocates to nucleus which binds to UPRE in the promoter of downstream target proteins [132]. IRE1 activation can lead to the expression of ER chaperones, and factors that regulate ERAD proteins and lipid synthesis. IRE1 can also facilitate the degradation of mRNAs and inhibit protein synthesis by cleaving 28s rRNA to help alleviate the protein burden of the ER and maintain cell

homeostasis through regulated IRE1-dependent decay of mRNA (RIDD). Under strong

activation of the UPR, the IRE1 pathway activation will significantly enhance RIDD

pathway which leads to cell apoptosis by indirectly activating caspase-2 activity

[133,134,135]. The effect of IRE1 covers almost all the ER functions, consistent with its

role as the only available UPR pathway in yeast. Ire1p is the yeast homologue of human

IRE1. In yeast, its RNase activity is activated under ER stress and cleaves the unspliced

(U) HAC1 mRNA which is then ligated by tRNA ligase Rlg1p to form the spliced

HAC1 mRNA. The spliced HAC1 mRNA is then able to translate into Hac1p, which binds

to UPRE to activate downstream factors expression [136].

ATF6 belongs to the ATF-CREB basic leucine zipper (bZIP) DNA-binding protein family. Full length ATF6 is 670 amino acids long and its molecular weight is 90kDa. From

N-terminus to C-terminus, it is composed of a transcriptional activation domain (TAD), bZIP domain, transmembrane domain (TM), CD1 and CD2 domains conserved with cAMP

response element binding protein related protein (CREB-RP). It also contains S1P and S2P cleavage sites. Its luminal domain can sense ER stress and is indispensable for ER stress

induced ATF6 activation [137] [138], as the ATF6 forms dimers and oligomers through the disulfide bonds between its the cysteine residues in the luminal domain. ER stress reduces the ATF6 dimers and oligomers to monomers, thus exposing its ER exit signal and facilitates its incorporation into COPII coated vesicles and transferring onto Golgi membrane where it is cleaved by site 1 protease (S1P) at site 1 cleavage site and site 2 protease (S2P) at site 2 cleavage site[139]. The N-terminus ATF6 (1-373 amino acid), which is 50kDa in molecular weight, is then freed and translocated into the nucleus to serve as a transcriptional factor. The ER stress response element (ERSE) 1 in the promoter region

of GRP78, GRP94, Calreticulin ERP72 and GRP58 contain a consensus sequence of

CCAATN9CCACG. Under ER stress, ATF6 (1-373) binds to the CCACG region while

NF-Y/CBP continuously binds to CCAAT region. ATF6 (1-373) and NF-Y form an ER stress response factor (ERSF) in ERSE1. Human ATF6 is highly similar to yeast Hac1p in the DNA binding domain. ATF6 can also interact directly with and activate a consensus

DNA binding site TGACGTG(G). This consensus sequence is found in the promoter of the

ADP-ribosylation factor 1, apolipoprotein CII and human HMG14 genes. ATF6 has two

isoforms; ATF6α and ATF6β. ATF6α but not ATF6β is responsible for the inducing the

gene expression of ATF6 regulated downstream ER chaperones. Mouse embryonic

fibroblast (MEF) cells which lack ATF6α are more sensitive to ER stress while lack of

ATF6β does not change the ER stress response in those MEF cells. Both ATF6α

and ATF6β knockout mice develop normally but double knockout leads to embryonic lethality. ATF6α also heterodimerizes with XBP1 to activate the expression of specific ER-

associated degradation factors [140]. IRE1’s kinase and nuclease activities are needed for

ATF6 activation [141]. ATF6 is also important in maintaining the homeostasis

environment of the cells during tissue development[142].

Under ER stress, PERK (double-stranded RNA-activated protein kinase(PKR)-like ER

kinase) oligemerizes and undergoes autophosphorylation which activates PERK kinase

activity to phosphorylate and inhibit translation initiation factor eIF2α. Inhibited eIF2α

suppresses translation of most mRNAs. However, it enhances the translation of some

specific mRNAs such as transcription factor ATF4 with short open reading frames in their

5’-untranslated regions which are translated when eIF2α is inhibited. Transcription factor

ATF4 leads to the expression of two other transcription factors CHOP (transcription factor

C/EBP homologous protein) and GADD34 (growth arrest and DNA damage-inducible 34).

Transcription factors CHOP leads to the expression of downstream factors that promote apoptosis. GADD34 negatively regulates PERK pathway activity by producing PP1C which is a protein phosphatase that dephosphorylates eIF2α[136].

FIGURE 3. A schematic cartoon figure of the unfolded protein response (UPR) pathways.

UPR executes these functions through manipulating three branches: IRE1 (inositol

requiring enzyme 1), PERK [double-stranded RNA-activated protein kinase (PKR)–like

ER kinase], and ATF6 (activating transcription factor 6). Under ER stress. IRE1 will

undergo lateral oligomerization which leads to their autophosphorylation and

conformational change and, as a result, their endoribonuclease activity is activated.

Activated IRE1 cleaves a 26 bp intron region from the (x-box binding protein 1) XBP1

mRNA in cytoplasm to generate the spliced form of XBP1 mRNA. The removal of the

intron segment and the frame shift allows the spliced XBP1 to produce sXBP1 protein

which is a leucine-zipper-containing transcription factor which binds to unfolded protein

response element (UPRE) in the promoters of ER chaperone genes and ERAD factor genes

to promote their expression. ER stress reduces the ATF6 dimers and oligomers to

monomers which exposes its ER exit signal and facilitates its incorporation into COPII

coated vesicles and transfer onto Golgi membrane where it is cleaved by site 1 protease

(S1P) at site 1 cleavage site and site 2 protease (S2P) at site 2 cleavage site. The N-terminus

ATF6 (1-373 amino acid), which is 50kDa in molecular weight, is then freed and

translocated into the nucleus to serve as a transcriptional factor to induce the gene

expression of ATF6 regulated downstream ER chaperones. Under ER stress, PERK

(double-stranded RNA-activated protein kinase (PKR)-like ER kinase) oligemerizes and

undergoes autophosphorylation which activates PERK kinase activity to phosphorylate and inhibit translation initiation factor eIF2α. EIF2α activity inhibition enhances the translation of some specific mRNAs such as transcription factor ATF4. Transcription factor ATF4 leads to the expression of two other transcription factors CHOP (transcription factor C/EBP

32 homologous protein). CHOP leads to the expression of downstream factors that promote apoptosis.

Figure 3

Modulating UPR as a therapy for misfolding diseases

As mentioned in Chapter 1, enhancing the ER’s folding capacity is one of the ways to

increase the ER export efficiency of mutant proteins. Mild activation of UPR is used to

enhance the folding or degradation capacity of ER to rescue phenotypes. Alleviation and

attenuation of elongated UPR signals is used to induce autophagy and inhibit cell apoptosis.

Activation of ATF6 and XBP1 are known to induce the expression level of many ER

factors that regulate protein import, N-glycosylation, protein folding, retrograde trafficking, anterograde trafficking and ER associated degradation. The Wiseman group has listed most of the downstream targets of the two pathways using the method of chemical-genetic

activation of ATF6 or IRE1 pathway [143]. They attached a destabilized version of E.coli

dihydrofolate reductase (DHFR) to the N terminus of full length ATF6 through a Gly-Ser

linker. The destabilized DHFR domain leads to the degradation of the DHFR-ATF6 protein

but addition of trimethoprim (TMP) stabilizes the DHFR domain which causes the

overexpression of DHFR-ATF6 and the activation of ATF6 pathway. They also used

tetracyclin (tet)-repressor technology. The addition of doxycycline (dox) can lead to the

overexpression of spliced XBP1 protein which activates the XBP1s pathway. They believe

that both pathway activations mimic the physiological activation of ATF6 or sXBP1

pathway.

According to the Wiseman group’s results, activation of ATF6 pathway induces the

expression level of HSP70/90 factors like BiP, GRP94, and their co-factors disulfide redox

like PDIF4, ERO1l, lectin chaperone such as CRT, ER associated degradation factors

34 including HERPUD1, OS9, SEL1. Activation of XBP1s increases the factors that include import factors such as Sec61A1, Sec61B, N-glycosylation factors like Dolichyl-

Diphosphooligosaccharide--Protein Glycosyltransferase Subunit STT3A and STT3B, retrograde trafficking factors like COPA, COPB1, anterograde trafficking Sec23A, Sec23B,

Sec24A, ER associated degradation factors like EDEM2, EDEM3, DERL1, HRD1, disulfide redox like PDIA5, ERO1L and HSP70/90 co-factors ERDJ4 and ERDJ5. Both

XBP1s and ATF6 can activate HSP70/90 co-factor P58/IPK, HYOU1, disulfide redox

ERP57 and PDIA6, ER associated degradation factor VCP, N-glycosylation enzyme

UGGT1. Co-expression of XBP1s and ATF6 will activate HSP70/90 co-factor ERDJ3, disulfide redox PDIA10, ER associated degradation factor EDEM1, DERL2 and anterograde trafficking factor Sec13.

Since the UPR can regulate ER folding and degradation capacities, there are many studies on the effect of modulating UPR on protein misfolding diseases. There are in general three types of protein misfolding diseases. The first type is neurodegenerative diseases caused by abnormal aggregation of misfolded proteins, which leads to progressive loss of neuronal function in nervous system. This type of disease includes Alzheimer disease(AD),

Parkinson disease (PD) , amyotrophic lateral sclerosis (ALS), Huntington disease (HD), prion-related disorders, retinitis pigmentosa, demyelinating disorders, and frontotemporal dementia[144]. Ageing and mutations that cause proteins aggregates are risk factors for these diseases. The protein aggregates that accumulate in and fail to be cleared efficiently by ER will disrupt ER calcium homeostasis.[145,146], associate with and thus disrupt the normal function of important ER chaperones such as BiP and PDI and degradation factors such as E3 ligase, Derlin1[147,148,149,150,151,152,153,154,155], impair ER to Golgi

trafficking[156], and interfere with the activation of UPR[157] The upregulation of all

three branches of the UPR is detected in autopsy brain samples or human induced

pluripotent stem cells (iPSCs) from patients with various neurodegenerative diseases.

Prolonged UPR stress signals will in turn lead to cell death and loss of neurons and their

functions as a result[158].

Modulating the UPR pathway proved to be neuroprotective in various neurodegenerative

diseases. Hetz and Saxena have organized the current available studies on how modulating

PERK, IRE1/sXBP1 or ATF6 pathway may affect the progress of the neurodegenerative

disease models. PERK inhibitor (GSK2606414) and ATF4 KO have neuron protective

effects in Alzheimer’s disease, Parkinson disease, Amyotrophic lateral sclerosis. CHOP

KO has neuroprotective effects in Parkinson disease, Charcot-Marie-Tooth disease but

exacerbate Pelizaeus-Merzbacher disease. GADD34 ablation and eIF2α phosphatse

inhibitors such as salubrinal, guanabenz are neuroprotective to Amyotrophic lateral

sclerosis, Parkinson disease, Charcot-Marie-Tooth disease by maintaining low level of

eIF2α but have a negative effect on Prion-related diseases. Activating PERK pathway can decrease the protein level of WT, aggregation prone misfolded mutant rhodopsin levels

[159].

Blocking IRE1 signaling pathway by XBP1 conditional KO or IRE1 conditional KO is

neuroprotective in amyotrophic lateral sclerosis and Alzheimer disease through enhanced

autophagy[160,161]. Deletion of XBP1 will cause mild ER stress that is not toxic to the

cells but can help protect cells from PD through the activation of ER chaperones and

autophagy. However, delivery of XBP1s through adeno-associated virus (AAV) into CNX

is reported to have neuroprotective effects in Huntington disease and Parkinson

disease[144]. IRE1(I642G), a genetically modified human IRE1, of which RNase activity can be turned on by 1NM-PP1 without activating PERK, ATF6, RIDD, JNK and does not

induce cell death. IRE1 pathway activation promotes the degradation of class II mutant

P23H rhodopsin mutants through proteasomal and lysosomal degradation pathways.

Lack of ATF6α in mice renders them more sensitive to PD-inducing neurotoxin

[144,162,163]. In models of HD, ATF6 function is negatively affected which leads to

disease pathogenesis [144,164,165] Chemical-genetic activation of ATF6 by tetracycline-

induced expression of N-terminus ATF6 (1-373) prevent several class II mutant rhodopsins

from accumulating in cells without significantly affecting monomeric WT rhodopsin level.

[159].

The second type of the protein misfolding diseases is caused by secretion of destabilized,

amyloidogenic proteins to the extracellular space where they assemble into proteotoxic

aggregates and amyloid fibrils which results in organ malfunction or death. More than 13

diseases are caused by the secretion of unstable, aggregation–prone proteins. For example,

hepatic secretion of more than 100 destabilized mutant transthyretin (TTR) amyloidosis

leads to systemic amyloidosis. Clonally expanded cancerous plasma cells secrets a

destabilized amyloidogenic Ig light chain (LC) which forms proteotoxic soluble oligomers and

amyloid fibrils that can accumulate at and disrupt the normal function of kidney, heart, and

gastrointestinal tract [166]. Thapsigargin selectively reduces the secretion of amyloidogenic

LC and induces its degradation. The Wiseman group showed that activating ATF6 pathway

by overexpression of full length ATF6 reduces the secretion and extracellular concentration of

amyloidogenic LC, reducing soluble amyloidogenic LC aggregate levels. They also showed

that activating sXBP1 pathway has a similar effect but to a less extent. Lowering the amount

of amyloidogenic LC secreted from plasma cells into the blood by UPR activation should

be useful in patients with significant cardiac and/or renal involvement [167]. The Kelly

group used a cell-based high-throughput screen (HTS) to discover a small molecule 147

which can selectively activate UPR-ATF6 pathway modestly and increase the GRP94 and

BiP level without inducing XBP1 splicing and eIF2α phosphorylation. Application of this

147 molecule into liver derived HePG2 cells decreases amyloidogenic TTR mutant

secretion without affecting the secretion of WT TTR or the endogenous proteome.

Application of the 147 molecule also decreases the secretion and extracellular aggregation

of amyloidogenic immunoglobulin light chain from AL patient derived plasma cells [168].

The third type of the protein misfolding diseases was described in Chapter 1. Mutations

in proteins kinetically disfavor the normal folding of the proteins which leads to their

accelerated degradation and decreased surface expression level. These mutant proteins remain functional to various extents when expressed onto the cell surface or designated working locations. As already mentioned in Chapter 1, application of proteostasis regulators (PR) such as Celastrol or MG-132 rescues the folding, trafficking of L444P glucocorebrosidase GC by activation of the UPR-PERK and UPR-IRE1 pathways but not the UPR-ATF6 pathway. Celastrol or MG-132 also rescues the forward trafficking of

αG269S β-hexosaminidase A HexA. Coapplication of Celastrol or MG-132 with pharmacological chaperones act synergistically to rescue the forward trafficking of N370S and L444P glucocorebrosidase GC [114].

Expression of misfolded proteins sometimes leads to the activation of UPR. Prolonged

activation of the UPR can lead to cell apoptosis. As is mentioned earlier in this introduction,

aggregation of misfolded proteins in various neurodegenerative diseases leads to the

38 activation of all three UPR pathways. It is also reported that R451C Neuroligin3 (NLGN3) is misfolded, retained in the ER and has impaired trafficking ability onto cell surface.

Existence of this mutant protein activates UPR pathways and the phenotype of the mice carrying this mutation can be rescued by blocking the UPR-PERK pathway [169].

For A322D GABAARs, as is mentioned in Chapter 1, this misfolded mutant α1 subunits are retained in ER and degraded efficiently. The WT β2 and γ2 subunits fail to form pentamers with α1 subunits to be transported out of the ER onto the cell surface. The first question we addressed was that if the expression of the A322D GABAARs activates the unfolded protein response. Since mild activation of the ATF6 pathway and IRE1 pathway is known to upregulate many critical ER chaperones such as BiP and Calnexin without significantly disturbing the general homeostasis of cells, we further tested if ATF6 upregulation or sXBP1 upregulation can promote the forward trafficking of A322D

GABAARs.

2.2 Materials and Methods:

2.2.1 Chemicals, plasmids, and antibodies

Lactacystin was obtained from AdipoGen life science (#AG-CN2-0104). The pCMV6 plasmids containing human GABAA receptor α1 (Uniprot no. P14867-1), β2 (isoform 2,

Uniprot no. P47870-1), and γ2 (isoform 2, Uniprot no. P18507-2) subunits and the pCMV6 Entry Vector plasmid (pCMV6-EV) were purchased from Origene. The A322D mutation was introduced to the α1 subunit using QuikChange II site-directed mutagenesis

Kit (Agilent Genomics). DNA sequencing was used to confirm the cDNA sequences. N- terminal HA-tagged full length pCGN-ATF6α (human) plasmid came from Addgene

(#11974). The pEGFP-N1 plasmid was a kind gift from Dr. Fraser Moss (Case Western

Reserve University). The XBP1s plasmid was a kind gift from Dr Richard N. Sifers. The mouse monoclonal anti-α1 (clone BD24) antibodies came from Millipore (#MAB339).

The mouse monoclonal anti-β-actin antibody (#A1978) was obtained from Sigma. The rabbit polyclonal anti-BiP antibody was from Abgent (#AP50016). The mouse monoclonal ATF6 antibody (#73-505) was from BioAcademia. The rabbit monoclonal anti-Na, K-ATPase was from Abcam (#ab76020). The rabbit polyclonal anti-Matrin-3 antibody was from Bethyl Laboratories (#A300-591A). The mouse monoclonal anti-HA antibody (F-7) was from Santa Cruz (#SC-7392). The rabbit polyclonal anti-XBP1 antibody (M-186) was from Santa Cruz (#SC-7160).

2.2.2 Cell culture and transfection

HEK293T cells and SH-SY5Y cells were obtained from the ATCC. Cells were maintained at 37°C in 5% CO2 in Dulbecco’s Modified Eagle Medium (DMEM) (Corning Media) with

10% heat-inactivated fetal bovine serum (Sigma-Aldrich) and 1% Penicillin-Streptomycin

(Hyclone). Monolayers were passaged with Trypsin 0.05% (Hyclone). Cells were grown in 6-well plates or 10-cm dishes and allowed to reach ~60-80% confluency before transient transfection using TransIT-2020 (Mirus) according to the manufacturer’s instruction. Cell lines that stably expressing α1β2γ2 and α1(A322D)β2γ2 receptors were generated by transient transfection with α1:β2:γ2 (1:1:1) and α1(A322D):β2:γ2 (1:1:1) plasmids. Then cells were selected using 0.8 mg/mL G-418 (Enzo Life Sciences) and maintained in 0.5 mg/mL G-418. The ATF6 plasmid was transiently transfected at 1 µg/well in 6-well plates,

1.5 µg in 3-cm dishes, or 2 µg in 10-cm dishes using a 1:3 (µg plasmid: µl transfection reagent) ratio. Forty-eight hours post transfection, cells were collected.

To generate a monoclonal HEK293T cells stably expressing α1(A322D)β2γ2 receptors, the α1(A322D) sequence was subcloned into a pIRES2-EGFP bicistronic vector (Clontech)

using EcoRI and SacII restriction sites, which would allow the simultaneous expression of

α1 subunits and EGFP separately but from the same RNA transcript. Briefly, 6.5 µg of the

pIRES2-EGFP vector or 10 µg of the pCMV6-α1 plasmid was added to 4 µl of 10x NEB4

buffer (NEB #B7004S), 1 µl of EcoRI (NEB # R0101S) and 1µl of SacII (NEB #R0157S).

14 µl of RNase-free H2O was used to make up to a 40 µl of final volume. The reaction

mixture was incubated at 37 °C for 4.5 hrs. The reaction samples were then subject to 1%

agarose gels to confirm the success of the plasmid digestion. The WT α1 subunit insert

(1453 bp) and the digested pIRES2-EGFP vector (5285 bp) were then cut and extracted

from the agarose gels using Qiagen Gel Extraction Kit (Qiagen #28704) following the

protocol included. 70 ng of the digested pIRES2-EGFP vector and 235 ng of the WT α1

subunit insert (molar ratio of the digested vector and the insert was 1:10) were added with

1 µl of T4 ligase (NEB #M0202S) and 3 µl of T4 ligase buffer (NEB #B0202S), and

RNase-free H2O was used to make up to a 30 µl final volume. The reaction mixture was

incubated at room temperature for 3 hrs and transformed into the DH5α competent cells

(Invitrogen). The resulting pIRES2-EGFP-α1 plasmid was confirmed by DNA sequencing.

The A322D mutation in the α1 subunit was generated using QuikChange II site-directed mutagenesis Kit (Agilent Genomics). After transfection and G-418 treatment, cells that were GFP-positive were considered as those successfully transfected with the α1(A322D) subunit. GFP-positive cells were further diluted into 96-well plates, allowing a single cell distribution in each well. Cells with robust GFP signals were further selected to grow to population.

2.2.3 Quantitative RT-PCR

The relative expression levels of target genes were analyzed using quantitative RT-PCR

described previously [113,170] . Cells were either transfected with GFP, WT or α1(A322D)

GABAA receptors for 48 hrs or treated with thapsigargin (2 µM, 4hrs) as a positive control before total RNA was extracted from the cells using RNeasy Mini Kit (Qiagen #74104). cDNA was synthesized from 500 ng of total RNA using QuantiTect Reverse Transcription

Kit (Qiagen #205311). Quantitative PCR reactions (45 cycles of 15 s at 94°C, 30 s at 57°C, and 30 s at 72°C) were performed using cDNA, QuantiTect SYBR Green PCR Kit (Qiagen

#204143) and corresponding primers in the StepOnePlus system (Applied Biosystems) and analyzed using StepOne v2.2 software (Applied Biosystems). The forward and reverse primers for CHOP and GAPDH (housekeeping gene control) are: CHOP Forward (5' > 3')

GGAAACAGAGTGGTCATTCCC; Reverse (5' > 3') CTGCTTGAGCCGTTCATTCTC;

GAPDH Forward (5' > 3') GTCGGAGTCAACGGATT; Reverse (5' > 3')

AAGCTTCCCGTTCTCAG. Threshold cycle (CT) was extracted from the PCR

amplification plot, and the ∆CT value was defined as: ∆CT = CT (target gene) - CT

(housekeeping gene). The relative mRNA expression level of target genes of the chemical

compounds-treated cells was normalized to that of untreated cells: Relative mRNA

expression level = 2 exp [-(∆CT (treated cells) - ∆CT (untreated cells))].

2.2.4 Western blot analysis

Cells were harvested with Trypsin-EDTA (0.05%) (Hyclone). Cells were lysed in lysis

buffer (50 mM Tris, pH 7.5, 150 mM NaCl, and 1% Triton X-100) supplemented with

Roche complete protease inhibitor cocktail on ice for an hour and then subjected to

centrifugation (13,400 × g, 15 min, 4 °C) to remove cell debris and nucleus. The

supernatant was collected as the total cellular protein. Protein concentration was measured using MicroBCA assay (Pierce). Endoglycosidase H (endo H) digestion and Peptide-N-

Glycosidase F (PNGase F) (New England Biolabs) digestion were performed according to

the published procedure [171]. Loading samples were generated by mixing cell lysates

and 4x SDS sample loading buffer (Biorad) and separated in an 8% denaturing tris-glycine gel. Western blot analysis was performed using corresponding antibodies. Band intensity was quantified using Image J software from the NIH.

2.2.5 Nuclear extraction

The nuclear extraction was performed according to published procedure [172]. Cells were harvested with Trypsin-EDTA (0.05%) (Hyclone) and lysed on ice for 5 min with harvest buffer (10 mM HEPES pH 7.9, 50 mM NaCl, 0.5 M Sucrose, 0.1 mM EDTA, 0.5% Triton

X-100) supplemented with 1 mM DTT and the Roche complete protease inhibitor cocktail.

The lysates were then centrifuged at 1000 g for 10 mins at 4°C. The supernatant was collected as the cytoplasmic part. The pellet was re-suspended with buffer A (10 mM

HEPES pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA) supplemented with 1 mM

DTT and the Roche complete protease inhibitor cocktail and then subjected to another centrifugation at 1000 g for 5 mins at 4°C. The resulting pellet was re-suspended with buffer C (10 mM HEPES pH 7.9, 500 mM NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.1%

IGEPAL (NP40)) supplemented with 1 mM DTT and the Roche complete protease inhibitor cocktail, incubated on ice for 1 h, and then subject to centrifugation at 16000 g for 15 min at 4°C. The resulting supernatant was collected as the nuclear extract of the cells.

2.2.6 Biotinylation of cell surface proteins

HEK293T cells transfected with GABAA receptors were plated in 10-cm dishes for surface biotinylation experiments according to published procedure [173]. In brief, cells were either transfected with GFP or ATF6 or sXBP1 plasmids for 48 hrs. Then, intact cells were rinsed gently twice with ice-cold PBS and incubated with the membrane-impermeable biotinylation reagent Sulfo-NHS SS-Biotin (0.5 mg ⁄ mL; Pierce) in PBS for 30 min at 4 °C to label surface membrane proteins. Glycine (10 mM) in ice-cold PBS was added to cells for 5 min at 4 °C to quench the reaction. N-ethylmaleimide (NEM, 5 nM) in PBS was added for 15 min at room temperature to block the Sulfhydryl groups. Cells were then solubilized for 1 h at 4 °C in lysis buffer (Triton X-100, 1%; Tris–HCl, 50 mM; NaCl, 150 mM; and EDTA, 5 mM; pH 7.5) supplemented with Roche complete protease inhibitor cocktail and 5 mM NEM. The lysates were centrifuged (16,000 × g, 15 min at 4 °C), and the supernatant contained the biotinylated surface proteins. The concentration of the supernatant was measured using MicroBCA assay (Pierce). Biotinylated surface proteins were purified by incubating the above supernatant for 1 h at 4 °C with 30 μL of immobilized neutravidin-conjugated agarose bead slurry (Pierce), and being subjected to centrifugation (16000 ×g, 10 mins). The beads were washed three times with buffer (Triton

X-100, 0.5%; Tris–HCl, 50 mM; NaCl, 150 mM; and EDTA, 5 mM; pH 7.5). Surface proteins were eluted from beads by boiling for 5 mins with 60 μL of LSB ⁄ Urea buffer (2x

Laemmli sample buffer (LSB) with 100 mM DTT and 6 M urea; pH 6.8) before subjected to SDS-PAGE and Western blotting analysis.

2.2.7 Cycloheximide-chase assay

Single clone stable α1(A322D)β2γ2 GABAA receptors HEK293T cells were plated in to

6-well plates one or two days before transfection. Cells were then transfected with GFP or

treated with DMSO as control group and transfected with ATF6 or sXBP1 using

transfection protocol listed above. Cells were treated with 150 μg/mL cycloheximide

(Ameresco) to stop protein translation for the 2 hrs, 1 hr, and 0.5 hr before being collected

for Western blot analysis.

2.2.8 Statistical analysis

All data were presented as mean ± SEM. Statistical significance was evaluated using two-

tailed Student’s t-Test if two groups were compared and one-way ANOVA followed by post-hoc Tukey or Fisher test if more than two groups were compared. A p value of less than 0.05 was considered statistically significant.

2.3 Result:

2.3.1 A322D GABAARs does not induce the UPR.

Under ER stress, IRE1 dissociates from BiP and undergoes dimerization/oligomerization and trans-autophosphorylation. This activates the endoribonuclease activity of IRE1 and cleaves the mRNA of XBP1 (X-box binding protein

1) to produce the spliced, active form of a transcriptional factor, XBP1s. Under basal conditions, ATF6, a type II membrane protein, is located in the ER membrane with its N- terminus facing the cytosol [137,138]. Under ER stress, full-length ATF6 translocates from the ER to the Golgi, where its cytosolic domain is released by the sequential cleavage by site-1 protease (S1P) and site-2 protease (S2P). The cleaved N-terminal ATF6 domain

(ATF6-N) then moves into the nucleus, serving as a potent transcription factor [138]. sXBP1 and ATF6 are transcriptional factors. sXBP1 binds to UPRE and ATF6 binds to

ERSE which are located in the promoter region of targeted factors including degradation factors, chaperones. BiP are known to be upregulated both of sXBP1 or ATF6 upregulation.

We monitored the ATF6 pathway and sXBP1 pathway activation by quantifying the mRNA level of BiP. Quantitative RT-PCR analysis demonstrated that transient overexpression of α1 and α1(A322D) receptors in HEK293T cells for about 48hrs did not significantly enhance the BiP mRNA expression level at 48hrs after transfection (Fig. 5A), indicating that the ATF6 and IRE1 pathways are not activated with the expression of α1 and α1(A322D) receptors.

PERK is an ER transmembrane kinase. Under ER stress, PERK undergoes dimerization/oligomerization, and its cytosolic kinase domains become autophosphorylated. This phosphorylates eIF2α, which reduces the global protein synthesis by inhibiting mRNA translation and concomitantly increases the expression of the transcription factor ATF4. ATF4 enhances the transcription of the transcription factor

CHOP (C/EBP homologous protein). We monitored the PERK pathway activation by quantifying the mRNA level of CHOP. Quantitative RT-PCR analysis demonstrated that transient overexpression of α1 and α1(A322D) receptors in HEK293T cells for about 48hrs did not significantly enhance the CHOP mRNA expression level (Fig. 5B), indicating that the PERK pathway is not activated with the expression of α1 and α1(A322D) receptors at

48hrs after transfection. The above results indicate the introduction of the A322D mutation in the α1 subunit does not activate the three arms of the UPR in HEK293T cells possibly because the mutant receptors are rapidly disposed by the ERAD pathway.

2.3.2 Activation of the ATF6 pathway promotes the forward trafficking of the mutant

α1(A322D) subunit.

We want to determine the effect of activating the UPR on the maturation of the

α1(A322D) subunits. Among the three UPR arms, although all ATF6, IRE1 and PERK

pathways leads to cell apoptosis as discussed in the above introduction session, mild

activation of ATF6 pathway and IRE1 pathway is known to enhance the ER folding

capacity and survival of cells, whereas the PERK pathway activation often leads to

apoptosis and reduces the protein synthesis when the pathway is continuously turned on

[40]. Therefore, we focused on the ATF6 arm and IRE1 arm. We first examined how ATF6

influenced the maturation of α1(A322D) subunits. ATF6 has two homologs: ATF6α and

ATF6β. The role of ATF6α in the UPR has been well-defined, whereas ATF6β has not

been thoroughly studied. Here, we overexpressed full-length HA-tagged ATF6α in

HEK293T cells stably expressing α1 (A322D)β2γ2 GABAA receptors. Western blot

analysis showed that ATF6α overexpression generated the cleaved, activated N-terminal

form of ATF6 (N) in the nucleus (Fig 6A), confirming the activation of the ATF6 pathway in HEK293T cells. Furthermore, as expected [56], ATF6α overexpression increased total protein expression level of both full-length ATF6α and BiP, an ATF6 target gene (Fig 6B,

cf. lane 2 to 1, quantification of the BiP band intensity shown in Fig 6D). As a result, this

operation led to a substantial 1.74-fold increase of the total protein level of α1(A322D)

subunits (Fig 6B, quantification shown in Fig 6C). In order to evaluate whether ATF6

activation promotes the folding and ER-to-Golgi trafficking of the mutant α1(A322D)

subunits, we performed the endoglycosidase H (endo H) enzyme digestion experiment.

Endo H digestion produced a single α1(A322D) band at the molecule weight size of the

unglycosylated subunit (Fig 6B, lane 3), which is designated as endo H-sensitive, indicating that the majority of the α1(A322D) subunit is retained in the ER for fast degradation. ATF6 overexpression led to a clear visualization of a strong post-ER α1

(A322D) subunit band (Fig 6B, lane 4, top endo H-resistant band). Moreover, the ratio of the mature α1(A322D) subunit (endo H-sensitive band) to total α1(A322D) subunit (the sum of endo H-sensitive and endo H-resistant band) is increased (quantification shown in

Fig 6E). This result indicates that activation of the ATF6 pathway enhanced the folding and ER-toGolgi trafficking of the α1(A322D) subunits. Furthermore, surface biotinylation experiments showed that ATF6 overexpression substantially increased the surface protein level of the α1 (A322D) subunits (Fig 7A, quantification shown in Fig 7B). Interestingly, the cycloheximide chase experiment revealed that ATF6 activation did not significantly alter the degradation rate of α1(A322D) subunits (Fig 7C, quantification shown in Fig 7D).

As a control to evaluate whether ATF6 activation induced autophagy as a major way to degrade α1(A322D), we incubated lactacystin, a potent proteasome inhibitor, in HEK293T cells overexpressing ATF6. This operation further slowed down the degradation of

α1(A322D) (Fig 7C, quantification shown in Fig 7D), indicating that the cycloheximide chase experiments were mostly a result of ERAD. Possibly, because ATF6 activation has profound effect on ER proteostasis network, the ERAD pathway could be elevated in addition to the folding enhancement. Therefore, we evaluated how ATF6 activation regulated major ER chaperones in HEK293T cells expressing α1(A322D)β2γ2 receptors.

Genetic activation of ATF6 significantly increased the total protein level of Grp94, but not calreticulin or calnexin (Fig 8A, quantification in Fig 8B). Previously, we demonstrated that Grp94 targets α1(A322D) to the ERAD pathway [55]. As such, elevated Grp94

expression could lead to enhanced ERAD and contribute to the apparently unchanged

stability of α1(A322D) after ATF6 activation. In summary, the above experiments clearly

demonstrated that ATF6 activation promotes the folding of the α1(A322D) subunits and

their forward trafficking from the ER to Golgi and onward to the plasma membrane.

2.3.3 Activating the IRE1 pathway increases the surface level of the mutant α1

(A322D) subunit

Activating the IRE1 pathway increases the surface level of the mutant α1 (A322D) subunit IRE1, an ER transmembrane kinase/endoribonuclease, responds to misfolded proteins in the ER by oligomerization and autophosphorylation [40]. IRE1 activation cleaves the inactivated form of XBP1 (X-box binding protein 1) mRNA to produce spliced

XBP1 mRNA, which encodes the transcriptional factor XBP1s. Here, we evaluated how

IRE1 pathway activation affects the maturation of α1(A322D) subunits. In order to activate

the IRE1 pathway, we overexpressed active form of transcriptional factor XBP1s in

HEK293T cells expressing α1(A322D) β2γ2 GABAA receptors. Western blot analysis

showed that XBP1s overexpression increased total protein expression level of both XBP1s

and BiP (Fig 9A, quantification of the BiP band intensity shown in Fig 9C). XBP1s

overexpression led to a substantial 1.72-fold increase of the total protein level of α1(A322D)

subunits (Fig 9A, quantification shown in Fig 9B). Furthermore, surface biotinylation

experiments showed that XBP1s overexpression substantially increased the surface protein

level of the α1(A322D) subunits (Fig 9D, quantification shown in Fig 9E), indicating that

IRE1 activation promotes the surface expression of the α1 (A322D) subunit. The

cycloheximide chase experiment showed that although IRE1 pathway activation group has

48.5% α1(A322D) left compared to 26.5% of control group after 0.5 hr chase, and has 22.3%

α1(A322D) left compared to 15.4% of control group after 1 hr chase, there is no significant

difference in the remaining α1(A322D) at all the time points between GFP control group

and XBP1s overexpression group (Fig 9F, quantification shown in Fig 9G, n = 4). To determine whether XBP1s activation induced autophagy as an alternative to degrade

α1(A322D), we treated HEK293T cells overexpressing XBP1s with lactacystin, a potent proteasome inhibitor. Clearly, lactacystin incubation further reduced the degradation of

α1(A322D) (Fig 9F, quantification shown in Fig 9G), indicating that the cycloheximide chase experiments mostly resulted from ERAD. To explore the possible reasons that lead to enhanced surface trafficking without increased stability, we evaluated how IRE1 activation influenced major ER chaperones in HEK293T cells expressing α1(A322D)β2γ2 receptors. XBP1s overexpression significantly increased the total protein level of Grp94 and calreticulin, but not calnexin (Fig 10A, quantification in Fig 10B). Accordingly, the unaltered stability of α1(A322D) after IRE1 activation could result from the increased

Grp94 protein level, which promoted its ERAD targeting [55]. In addition, XBP1s overexpression significantly increased the total protein level of WT α1 subunits in

HEK293T cells expressing WT receptors, whereas ATF6 overexpression did not (Fig 10C, quantification shown in Fig 10D). The apparently unchanged WT α1 protein levels after

ATF6 activation could come from the competition between promoted folding, such as from

BiP upregulation, and enhanced ERAD, such as from Grp94 upregulation. The different effect between ATF6 and IRE1 activation could be because ATF6 and IRE1 have overlapping, but different downstream targets [50]. This result also suggests that ATF6 activation could have a more selective effect on the misfolding-prone mutant receptors over

WT receptors, and thus merits further development.

2.4 Discussions:

Previously, it was reported that the α1(A322D) mutant proteins have a very short half-life of about 30 mins [41,174]. In the ER, they are immediately recognized and eliminated by the ERAD pathway. It seems that the cells utilize the ERAD machinery efficiently to discard the mutants without the need to further activate the UPR. Therefore, our results demonstrated that introducing the A322D mutation into the α1 subunits does not significantly activate the UPR pathway. Interestingly, a previous study showed that overexpression of misfolded ∆F508 CFTR in Calu3∆FC5 cells induces the UPR [175]. In contrast, expression of endogenous level of mutant ΔF508 CFTR in primary airway cells does not activate the UPR [176,177]. Therefore, whether the UPR is activated might depend on the expression level of misfolded proteins. If the ERAD machinery is overloaded by their high expression levels, the cells need to further activate the UPR to handle misfolded proteins. Therefore, our expression of GABAA receptors in HEK293T cells likely mimic the physiologically relevant expression level in the central nervous systems.

We further demonstrated that activating either ATF6 pathway or IRE1 pathway increases the surface expression of pathogenic GABAA receptors carrying the A322D mutation in the α1 subunit. Both the IRE1 pathway and the ATF6 activation remodel the

ER proteostasis network to enhance the folding of the α1(A322D) subunit, enabling its successful transport to the plasma membrane. Interestingly, previous studies showed that

51 activating ATF6 reduces the secretion and extracellular aggregation of the amyloidogenic protein transthyretin by enhancing ER quality control stringency and thus ERAD [44]. Here, we showed that ATF6 activation enhances the forward trafficking of mutant α1(A322D) subunits. However, ATF6 activation does not influence the degradation rate of total mutant

α1(A322D) subunits. This may indicate that ATF6 activation favors the folding and forward trafficking of mutant α1(A322D) subunits. Previous study also showed that the specific activation of IRE1 pathway without affecting the two other UPR pathways by using a chemical control system promotes the degradation of misfolding mutant R21H rhodopsin protein in retinal cells [47]. However, our result showed XBP1s overexpression does not decrease the degradation rate of α1(A322D) subunits, but promotes their surface expression. It seems that ATF6 pathway or IRE1 pathway activation produces two competing downstream effects: folding enhancement and ERAD enhancement. The net effect on protein trafficking might depend on the specific proteins of interest with a goal to be cytoprotective. Operating ATF6 pathway or IRE1 pathway appears to be a promising strategy to ameliorate protein conformational diseases. ATF6 and IRE1 have selective targets among the ER proteostasis network. A comprehensive analysis of chaperones and folding enzymes that contribute to ATF6’s and IRE1’s effect needs to be carried out in the future by analyzing the α1(A322D) interactome before and after the activation of ATF6 pathway or IRE1 pathway

2.5 Figures:

FIGURE 4. Molecular structures of GABAA receptors. (A) A cartoon representation of the major pentameric GABAA receptor subtype in the central nervous system. It contains

two α1 subunits, two β2 subunits, and one γ2 subunit. This model was constructed from the cryo-EM structure (6D6U.pdb) [21] by using PYMOL. (B) Topology of the α1 subunit.

The large N-terminal domain resides in the ER lumen or extracellular space. Ala322, displayed as a space-filling model and indicated by an arrow, is located in the third transmembrane (TM3) helix. (C) Sequence alignment of TM3 residues of the α1, β2, β3, and γ2 subunits of GABAA receptors. The sequences are from the following Uniprot entries:

GBRA1, P14867; GBRB2, P47870-1; GBRB3, P28472-1; GBRG2, P18507-2. The A322

residue in the α1 subunit is highlighted in yellow. Hydrophobic residues are conserved in

this position.

(A) γ2 (B) β2 α1 β2 α1 ER or α1 extracellular

A322

Cytosolic

GBRA1 312 AMDWFIAVCY AFVFSALIEF ATVNYF 337 GBRB2 304 AIDMYLMGCF VFVFMALLEY ALVNYI 329 GBRB3 305 AIDMYLMGCF VFVFLALLEY AFVNYI 330 GBRG2 334 AMDLFVSVCF IFVFSALVEY GTLHYF 359

Figure 4

Figure 5. Introduction of the A322D mutation in the α1 subunit does not activate the

UPR. (A) HEK293T cells were transiently transfected with FLAG-tagged WT α1 or

α1(A322D) (together with β2 and γ2 subunits) of GABAA receptors. Forty-eight hrs post

transfection, mRNA was extracted. Quantitative RT-qPCR was performed to quantify the

mRNA expression level of HSP5A(A) and CHOP (B). The experiments were done using two biological replicates in duplicate each time. BLK: blank, untreated. Tg: thapsigargin treated. * p < 0.05.

Figure 5

Figure 6. ATF6 activation promotes the forward trafficking of α1(A322D) subunit of

GABAA receptors. (A) HEK293T cells expressing α1(A322D)β2γ2 receptors were transiently transfected with GFP or HA-tagged full-length ATF6α plasmids. Forty-eight hrs post transfection, the nuclear fractions were extracted and subject to SDS-PAGE. ATF6

(N) is the cleaved, activated N-terminal ATF6 in the nucleus. Matrin-3 serves as a nuclear protein loading control. (B) Cells were treated as in (A). Forty-eight hrs post transfection, cells were lysed, and total proteins were extracted. Total cellular proteins were incubated with or without endoglycosidase H enzyme (endo H) or peptide-N-glycosidase F (PNGase

F) for 1h at 37°C and then subjected to SDS-PAGE and Western blot analysis using corresponding antibodies. Endo H resistant α1 subunit bands (top arrow, lane 4) represent properly folded, post-ER α1 subunit glycoforms that traffic at least to the Golgi compartment, whereas endo H sensitive α1 subunit bands (bottom arrow, lanes 3 and 4) represent immature α1 subunit glycoforms that are retained in the ER. The PNGase F enzyme cleaves between the innermost N-acetyl-D-glucosamine and asparagine residues from N-linked glycoproteins, serving as a control for unglycosylated α1 subunits (lane 5).

Quantification of total cellular protein expression levels of α1 and BiP is shown in (C) and

(D) (n = 5 for α1 and n=4 for BiP, paired t-test). Quantification of the ratio of endo H resistant α1 / total α1 is shown in (E) (n=3, paired t-test).

Figure 6

Figure 7. ATF6 activation enhances the surface expression level of α1(A322D) subunit

of GABAA receptors without affecting the degradation rate of α1(A322D) subunit.

(A) HEK293T cells expressing α1(A322D)β2γ2 receptors were transiently transfected with GFP or HA-tagged full-length ATF6α plasmids. Forty-eight hrs post transfection, the cell surface proteins were tagged with biotin using membrane-impermeable biotinylation reagent sulfo-NHS SS-Biotin. Biotinylated surface proteins were affinity-purified using neutravidin-conjugated beads and then subjected to SDS-PAGE and Western blot analysis.

The Na+/K+-ATPase serves as a surface protein loading control. Quantification of

normalized surface α1(A322D) protein levels is shown in (B) (n = 6, paired t-test). (C)

HEK293T cells expressing α1(A322D)β2γ2 receptors were either transfected with GFP control, or ATF6, or transfected with ATF6 and treated with lactacystin (2.5μM for 24h).

Cycloheximide (150 μg/ml), a protein synthesis inhibitor, was added to different cell

groups for 0, 0.5 hr, 1 hr, and 2 hrs. Cells were then lysed and subjected to SDS-PAGE and

western blot analysis. The quantitation results are shown in (D) (n=5, one-way ANOVA followed by Fisher test, *, p<0.05 between GFP control and ATF6 + Lac group, #, p<0.05 between ATF6 group and ATF6 + Lac group).

Figure 7

Figure 8. ATF6 activation modulates the proteostasis network of Endoplsmic

Reticulum.

(A) HEK293T cells expressing α1(A322D)β2γ2 receptors were either transfected with

GFP or ATF6 for 48h. The cell lysates were then subjected to SDS-PAGE and Western blot analysis using corresponding antibodies. Quantification of total cellular chaperone protein expression levels is shown in (B) (n=4, paired t-test). *, p<0.05.

Figure 8

Figure 9. IRE1 activation increases the total and surface expression α1(A322D)

subunit of GABAA receptors without significantly slowing their degradation rate. (A)

HEK293T cells expressing α1(A322D)β2γ2 receptors were transiently transfected with

GFP or XBP1-s (spliced XBP1) plasmids. Forty-eight hrs post transfection, cells were lysed, and total proteins were extracted. The cell lysates are then subjected to SDS-PAGE and Western blot analysis using corresponding antibodies. Quantification of total cellular protein expression levels of α1 and BiP is shown in (B & C) (n = 5 for α1 and n=3 for BiP, paired t-test). (D) HEK293T cells were treated as in (A). Forty-eight hrs post transfection, the cell surface proteins were tagged with biotin using membrane-impermeable biotinylation reagent sulfo-NHS SS-Biotin. Biotinylated surface proteins were affinity- purified using neutravidin-conjugated beads and then subjected to SDS-PAGE and

Western blot analysis. The Na+/K+-ATPase serves as a surface protein loading control.

Quantification of normalized surface protein expression levels of α1 is shown in (E) (n =

5, paired t-test). (F) HEK293T cells expressing WT α1β2γ2 receptors were transfected with GFP, ATF6 or XBP1s plasmids. Forty-eight hrs post transfection, cells were lysed, and total proteins were extracted and subject to SDS-PAGE and Western blot analysis.

Quantification of total cellular protein expression levels of α1 is shown in (G) (n = 3, paired t-test using adjusted p values). *, p<0.05.

Figure 9

Figure 10. IRE1 activation modulates the proteostasis network of Endoplsmic

Reticulum and ATF6 or IRE1 pathways activation has different effects on WT α1 subunit of GABAA receptors. (A) HEK293T cells expressing α1(A322D)β2γ2 receptors were either transfected with GFP control, or XBP-s or transfected with XBP-s and treated with lactacystin (2.5 μM for 24h). Cycloheximide (150 μg/ml), a protein synthesis inhibitor, was added to different cell groups for 0, 0.5 hr, 1 hr, and 2 hrs. Cells were then lysed and subjected to SDS-PAGE and western blot analysis. The quantitation results are shown in

(B) (n=3, one-way ANOVA followed by Fisher test, *, p<0.05 between control group and

XBP1-s + Lac group, #, p<0.05 between XBP1-s group and XBP1-s + Lac group). (C and

D) HEK293T cells expressing α1(A322D)β2γ2 receptors were either transfected with GFP or XBP1s for 48h. The cell lysates were then subjected to SDS-PAGE and Western blot analysis using corresponding antibodies (C). Quantification of total cellular chaperone protein expression levels is shown in (D) (n=4, paired t-test). *, p<0.05.

Figure 10

Chapter 3: The effect of BIX treatment on the trafficking of A322D and D219N

GABAARs.

3.1 Introduction:

The immunoglobulin (Ig) heavy (H) chain Binding Protein (BiP) also known as 78-kDa

glucose-regulated protein GRP78 or Heat Shock 70kDa Protein 5 HSP5a is encoded by

HSPA5 and is an ER-resident HSP70 family protein. It has a conserved N-terminal nucleotide-binding domain (NBD) and a substrate-binding domain (SBD), a linker between this two domains and a KDEL ER-retention signal at its C-terminus. It recognizes and binds hydrophobic residues of the unfolded regions of the protein and interact with both glycosylated and non-glycosylated substrates [178,179]. BiP transiently binds to nascent polypeptide chain co-translationally to stabilize protein intermediates, promote protein folding and prevent protein aggregation or premature interaction. BiP’s binding to terminally misfolded proteins continues until the ER associated degradation of the misfolded protein substrates [179,180]. BiP also serve to prevent the leak of calcium ions

out of ER by functioning as a plug for Sec61 channels [181,182,183].

It is believed that the BiP expression level is kept at a relatively low level under basal

conditions[178]. BiP can be upregulated by the following transcription factors which bind

to the promoter of its gene and induce its upregulation: activating transcription factor 2, 4,

and 6 (ATF2, ATF4, ATF6), sXBP1, CBF/NF-Y. BiP mRNA can be preferentially

translated under conditions when general protein synthesis is inhibited. Increasing

PI3K/AKT activity can help stabilize BiP protein stability but the exact mechanism is

unclear yet. microRNA (miRNA) can also regulate BiP mRNA translation level by

interacting with its 3’-untranslated region. BiP also serves to regulate unfolded protein

response activation. Previously, it was believed that accumulation of misfolded protein in

ER activates the IRE1, PERK and ATF6 pathways through BiP. Under basal condition,

BiP directly binds to the luminal domains of IRE1α, IRE1β, PERK and ATF6.

Accumulated misfolded or unfolded proteins in ER will bind to and occupy BiP which

facilitates BiP dissociation from IRE1, PERK and ATF6. Freed Ire1 and PERK monomers

undergo dimerization and ATF6 are transported out of ER into Golgi to activate the UPR

[184,185,186]. However, more studies show that UPR activation is not dependent on BiP’s

dissociation from ER sensors. BiP serves to regulate the sensitivity of ER sensors to ER

stress and to prevent their being over sensitive to ER stress. It is shown that ATF6 binding

to BiP is stable with the presence of mutant misfolded proteins and under ER stress.

Abolishing the BiP binding ability of yeast Ire1p does not prevent UPR activation but

makes yeast more sensitive to environmental stimulations such as increased temperature

and ethanol[187]. Overexpression of BiP through UPR activation of genetic

overexpression actually serves to maintain the cell homeostasis by attenuating the

activation of UPR and protecting cells from cell death or apoptosis caused by activation of

UPR[178,184,188,189,190,191,192]. BiP is also able to protect cells from lipid

peroxidation and reactive oxygen species (ROS) generation [193].

Although BiP is mainly localized in ER as it contains the ER retaining KDEL signal,

Casas et al has summarized in a review that BiP can also be localized in mitochondria,

cytoplasm, cell surface and secreted into peripheral circulation and that it performs various

functions. An alternative splice form of GRP78, GRP78va, is located in the cytoplasm

[194]. This variant does not have the ER localization sequence and only exists in cancer

cells[195].

ER-localized DnaJ family proteins (ERdjs) from Erdj1 to ERdj7 and neucleotide

exchange factors serve as co-chapareones for BiP by delivering misfolded substrates to

BiP, modulating BiP’s conformation, promoting substrates’ degradation. Hendershot et al group have reviewed BiP’s co-chaperones. However, many of the functions of BiP’s cochaperones are still unknown [196].

Aging or mutations that disrupt BiP’s function will make cells more susceptible to damage caused by these factors. Aging is a risk factor for the onset of many neurodegenerative diseases. Aging can negatively affect BiP activity and expression level.

For example, it is reported that in old (20-24 months) mice, BiP ATPase activity is reduced by 20% and the carbonylation of GRP78 (an indicator of the oxidative damage to GRP78) is increased by 2 fold compared to young (3-5) month mice. GRP78 mRNA is reduced by

73% in (900 day old) compared to (21 day old) young Sprague-Dawley rats. Decreased ER chaperone activity and expression of BiP caused by aging makes the neurons more prone to extrinsic and intrinsic damage [197]. Mutant presenilin-1 leads to early onset familial

Alzehimer’s disease by inhibiting IRE1-pathway-mediated increase in BiP expression level and make neurons more susceptible to ER stress and leads to neuron death[198].

GRP78 has been proved to be neuroprotective in many neurodegenerative disease models and brain ischemia models. In many neurodegenerative diseases, UPR is activated and GRP78 levels increased as a result. Overexpression of BiP in CNS has a

neuroprotective effect. Introducing the AAV-BiP to substantia nigra of rats delayed the progression of PD that is cuased by overexpression of α-synuclein, leads to improved motor performance and improves the dopaminergic neuron survival [199]. Delivery of AAV-BiP into the retina decreases photoreceptor apoptosis, alleviates ER stress and improves visual function in transgenic rats with mutant rhodopsin [200]. BiP helps promote the folding of amyloid beta precursor protein (APP). Increasing BiP expression level increases its association with APP and decreases Aβ secretion by preventing the APP from entering

distal secretory compartment for processing into Aβ [201]. Upregulation of BiP caused by

ischemic preconditioning protects neurons from damage caused by following oxygen

glucose by activating autophagy. Inhibition of BiP disrupts the ischemic tolerance by preventing autophagic activation [202]. Overexpressionvof BiP helps to maintain mitochondria function in astrocytes after ischemia attack [203] .

Previously, Connolly et al. demonstrated that BiP interacts with α1 subunits of GABAA

receptors [27], Overexpression of BiP or calnexin improved the matured portion of A322D

α1 subunits in the cell [37]. SAHA, a FDA proved drug and a HDAC inhibitor, has been shown to promote the forward trafficking of A322D GABAARs and induce the expression

level of both BiP and Calnexin at the same time without inducing the IRE1 pathway

downstream targets probably through SAHA inhibiting histone deacetylase 7 (HDAC7)

[37]. As is noted in Chapter 1, SAHA has also enhances the maturation and trafficking of

trafficking deficient mutant CFTR, lysosomal glucocerebrosidase, and a1-antitrypsin. The effect is not due to BiP and Calnexin overexpression [116,117,118].

The small molecule BIX induces BiP expression through an ATF6-dependent

mechanism, but does not substantially induce expression of other ATF6 target genes, such

as Grp94 and PDIA4 in SK-N-SH neuroblastoma cells [51, 52]. Previous studies showed

that cerebral pretreatment with BIX in ischemia mice protects neurons from ER stress- induced cell death by enhancing BiP level [51]. Small molecules have advantages over gene delivery into the animal models and patients as small molecules do not require gene delivery system that generate serious side effects to human. Although the idea that BiP overexpression can rescue the trafficking of A322D GABAARs is not new, we now report

at the first time the effect of BiX in HEK293 cells and human neuroblastoma cell line SH-

SY5Y expressing A322D GABAARs for evaluating the BiX’s effect on restoring

GABAARs mediated currents in cells expressing with mutant receptors.

3.2 Materials and Methods

3.2.1 Chemicals, plasmids, and antibodies

BIX (BiP protein inducer X) was obtained from Sigma (#SML1073). Thapsigargin was

obtained from Enzo life science (BML-PE180-0001). Lactacystin was obtained from

AdipoGen life science (#AG-CN2-0104). Resazurin was obtained from MP biomedicals

(#0219548101). The pCMV6 plasmids containing human GABAA receptor α1 (Uniprot

no. P14867-1), β2 (isoform 2, Uniprot no. P47870-1), and γ2 (isoform 2, Uniprot no.

P18507-2) subunits and the pCMV6 Entry Vector plasmid (pCMV6-EV) were purchased

from Origene. The A322D or D219N mutation was established as previously mentioned

in Chapter 2.2.1. The mouse monoclonal anti-α1 (clone BD24) antibodies came from

Millipore (#MAB339). The mouse monoclonal anti-β-actin antibody (#A1978) was

obtained from Sigma. The rabbit polyclonal anti-BiP antibody is from Abgent

(#AP50016). The rabbit monoclonal anti-Na, K-ATPase was from Abcam (#ab76020).

3.2.2 Cell culture and transfection

HEK293T cells and SH-SY5Y cells were obtained from ATCC. Cells maintenance conditions were the same as previously mentioned in Chapter 2.2.2. Monoclonal

HEK293T cells stably expressing α1(A322D)β2γ2 receptors were used, the detailed establishment procedures of which were mentioned in Chapter 2.2.2.

3.2.3 Western blot analysis

Cells harvest procedures, Endoglycosidase H (endo H) digestion and Peptide-N-

Glycosidase F (PNGase F) (New England Biolabs) digestion procedures, and western blot procedures were all detailed in Chapter 2.2.4. Western blot analysis was performed using corresponding antibodies. Band intensity was quantified using Image J software from the

NIH.

3.2.4 Resazurin cell toxicity assay

HEK293T cells stably expressing α1(A322D)β2γ2 receptors were plated into a 96-well plate. The cells were separated into 7 groups which were treated with DMSO, BIX (1.2

μM, 6 μM, 12 μM, 24 μM, or 48 μM) for 24h, or thapsigargin (2 μM, 7h). Resazurin (0.15 mg / ml dissolved in DPBS) was added to cells for 1.5 h before plate reading. Fluorescence signal at 560 nm excitation / 590 nm emission was measured.

3.2.5 Biotinylation of cell surface proteins

HEK293T cells transfected with GABAA receptors were plated in 10-cm dishes for surface biotinylation experiments according to published procedure [173]. In brief, cells

were treated with BIX at 12 µM for 24 hrs. Then the procedures for surface biotinylation

of cell surface proteins are detailed in Chapter 2.2.6.

3.2.6 Cycloheximide-chase assay

Single clone HEK293T cells stably expressing α1(A322D)β2γ2 GABAA receptors were

plated in to 6-well plates one or two days before transfection. Cells were treated with

DMSO as control group or treated with BIX (12 μM) 24 hr using transfection protocol listed above. Cells were treated with 150 μg/mL cycloheximide (Ameresco) to stop protein translation for the 2 hrs, 1 hr, and 0.5 hr before being collected for Western blot analysis.

3.2.7 Whole-cell patch clamp electrophysiology recording

The whole-cell patch clamp procedure follows the protocol described in detail before

[204]. In brief, whole-cell currents were recorded 24 hrs post application of BIX (12µM) in monoclonal HEK293T cells stably expressing α1(A322D)β2γ2 receptors. The electrode internal solution contained 153 mM KCl, 1 mM MgCl2, 5 mM EGTA, 10 mM HEPES, and

2 mM MgATP (pH 7.3), whereas the external recording solution contains 142 mM NaCl,

8 mM KCl, 6 mM MgCl2, 1 mM CaCl2, 10 mM glucose, 10 mM HEPES, and 120 nM

Fenvalerate (pH 7.4). Coverslips containing HEK293T cells were placed in a RC-25

recording chamber (Warner Instruments) on the stage of an Olympus IX-71 inverted fluorescence microscope. Cells were perfused with external solution. A Quartz

MicroManifold with 100-µm inner diameter inlet tubes (ALA Scientific) was placed so that its tip was within 50 µm of the cell to be recorded. The whole-cell GABA-induced currents were recorded at a holding potential of -60 mV in voltage clamp mode using an

Axopatch 200B amplifier (Molecular Devices). The pClamp10 software was used for data acquisition and analysis.

3.2.8 Statistical analysis

All data were presented as mean ± SEM. Statistical significance was evaluated using two-tailed Student’s t-Test if two groups were compared and one-way ANOVA followed by post-hoc Tukey or Fisher test if more than two groups were compared. A p value of less than 0.05 was considered statistically significant.

3.3 Results

3.3.1 BIX treatment promotes the ER-to-Golgi trafficking and reduces the degradation of the mutant α1(A322D) subunit

Dose-response analysis showed that BIX (24 h treatment) increased the BiP protein levels significantly at concentrations of 12, 24, and 48 μM (Fig 11B, quantification of the BiP band intensity shown in Fig 11E). Moreover, BIX treatment increased the total α1(A322D) subunit significantly at 6 μM and this effect plateaued from 6 μM to 24 μM (Fig 11B, quantification of the α1 band intensity shown in Fig 11D). Time-course study demonstrated that BIX’s effect on increasing BiP and total α1(A322D) subunit level was achieved as early as 18 h and plateaued from 18 h to 30 h (Fig 11C, quantification of the α1 band intensity shown in Fig 11F and the BiP band intensity shown in Fig 11G). This time scale possibly reflects the involvement of protein maturation, such as protein folding and trafficking. Resazurin cell toxicity assay demonstrated that single-dose applications of BIX for 24 h did not induce significant toxicity to cells at concentrations no more than 12 μM

(Fig 11H). Higher concentrations (24 μM and 48 μM) of BIX led to reduced cell viability

(Fig 11H), which could account for the decreased total α1 protein levels at such

concentrations (Fig 11D). Therefore, we used the optimal incubation condition of BIX (12

μM, 24 h) for the following experiments. Furthermore, because the total expression level

of GABAA receptors could adjust the capacity of the ER proteostasis network prior to the

treatment, we evaluated whether BIX administration was effective in the context of

different protein synthesis load. HEK293T cells were transiently transfected with 0.15 μg

of α1(A322D), β2, and γ2 plasmids or 0.25 μg of them and treated with BIX. Western blot

analysis demonstrated that BIX treatment increased the total α1 (A322D) protein levels in

both cases (Fig 12A). It appeared that BIX treatment led to less increase of the total

α1(A322D) protein in the 0.25 μg transfection group compared to that in the 0.15 μg transfection group (Fig 12A, quantification of the α1 band and BiP band intensity shown in Fig 12B & 12C ), indicating that BIX’s efficacy partially depended on the protein synthesis load. Therefore, care must be taken not to overload the proteostasis network when testing the effect of added proteostasis regulators.

Next we determined whether BIX promoted the folding and the ER-to-Golgi trafficking of the α1(A322D)β2γ2 receptors. We first performed the endoglycosidase H (endo H) enzyme digestion experiment, which tracks the forward trafficking of a glycoprotein from the ER to the Golgi. Endo H cleaves the mannose-rich core glycans (the ER glycoform), which are attached to Asn residues in an Asn-X-Ser/Thr sequence (X can be any residue except proline). But endo H can’t eliminate the oligosaccharide chain after glycan remodeling in the Golgi. Therefore, this assay indirectly evaluates whether a glycoprotein is properly folded in the ER. The α1 subunit has two N-linked glycosylation sites at Asn38

and Asn138. The peptide-N-glycosidase F (PNGase F) enzyme cleaves between the

innermost N-acetyl-D-glucosamine and Asn residues from N-linked glycoproteins, serving

as a control for unglycosylated α1 subunits (Fig 13A, lane 5). Endo H digestion

experiments demonstrated that BIX treatment (12 μM, 24 h) clearly increased the endo H- resistant α1(A322D) band intensity (Fig 13A, cf. lanes 8 and 9 to lanes 6 and 7) as well as the ratio of endo H resistant / total α1(A322D) significantly (Fig 13A, quantification shown in Fig 13B), indicating that BIX treatment enhances the trafficking efficiency of the

α1(A322D) subunit from the ER to Golgi, and consequently, more properly folded

α1(A322D) proteins are able to reach at least to the Golgi. Previous study showed that

A322D mutation dramatically increases α1 degradation rate with a half-life as short as of

23 min compared to WT α1 subunits with a half-life of more than 90 min [13, 37]. We then performed cycloheximide-chase experiments to evaluate the degradation of α1(A322D) subunits by applying cycloheximide to inhibit protein synthesis in HEK293T cells stably expressing α1(A322D)β2γ2 receptors. BIX treatment significantly increased the remaining

α1(A322D) subunits from 26.7% to 51.3% after 0.5 h cycloheximide chase, and from 9% to 23% after 1 h cycloheximide chase (p < 0.05) (Fig 13C, quantification shown in Fig

13D). As a result, BIX treatment reduced the degradation of α1(A322D) subunits, consistent with the increased total α1 (A322D) protein levels. Moreover, addition of lactacystin, a potent proteasome inhibitor, with BIX treatment further attenuated the degradation of α1(A322D) (Fig 13C, quantification shown in Fig 13D), indicating that indeed the cycloheximide chase experiments reflected that the ERAD plays the most important role in the degradation of α1(A322D).

3.3.2 BIX treatment promotes the functional surface expression of α1(A322D) subunit

We next asked whether the upregulated α1(A322D) proteins afforded by BIX

administration reach the plasma membrane for their function. Surface biotinylation

experiments demonstrated that BIX treatment increased α1(A322D) subunit in the plasma

membrane significantly in HEK293T cells (Fig 14A & 14B). We also tested the effect of

BIX treatment on the surface trafficking of α1(A322D) subunits in human SH-SY5Y

neuroblastoma cells stably expressing α1(A322D)β2γ2 receptors. Surface biotinylation

experiments showed that BIX treatment significantly enhanced the surface expression of

α1 (A322D) subunits 2.2-fold (Fig 14C, quantification shown in Fig 14D). We then tested

whether the increased surface expression of α1(A322D) subunits is functional using whole-

cell voltage-clamping electrophysiology to record GABA-induced chloride currents. To

minimize the variation in the recording of GABA-induced currents among different cells, we generated monoclonal HEK293T cells stably expressing α1(A322D)β2γ2 GABAA receptors. To achieve that, we subcloned the α1(A322D) into a pIRES2-EGFP bicistronic

vector, which would allow the simultaneous expression of α1 subunits and EGFP

separately but from the same RNA transcript. This enabled us to select GFP-positive single cells for electrophysiology recording. The peak chloride current in response to GABA (3 mM) was only 6.0 pA in untreated HEK293T cells expressing α1(A322D)β2γ2 GABAA

receptors (Fig 14E), indicating that essentially no functional channels reside in the plasma

membrane. Strikingly, BIX treatment significantly increased this current to 30 pA (Fig 14E,

quantification shown in Fig 14F), indicating that BIX partially corrected the function of this pathogenic mutant GABAA receptors on the plasma membrane. Previously, we showed

that GABA-induced peak chloride current in HEK293T cells expressing WT GABAA

receptors was 138 pA [37]. Therefore, the peak current for BIX-rescued α1(A322D)β2γ2

77 receptors amounted to 22% of that for WT receptors, greater than that for SAHA-rescued mutant receptors [37]. We further tested the BIX effect on WT α1 subunit and another misfolding-prone mutant α1(D219N) subunit [53,54]. Surface biotinylation experiments showed that BIX treatment increased both the total and plasma membrane protein levels for WT α1 and α1(D219N) subunits significantly (Fig 15A, quantification shown in Fig

15B-15E), indicating that BIX treatment can enhance the surface trafficking of a variety of misfolding-prone subunits. The positive effect of BIX on WT α1 subunits was because these WT membrane proteins do not fold and assemble efficiently in the ER and part of them are degraded by the ERAD pathway [55]. A recent report revealed that despite the relatively modest peak current increase, SAHA treatment restored the receptor kinetics in heterosynaptic cultures harboring the α1(A322D) mutation that were indistinguishable from those harboring the WT receptors [38]. Therefore, although the physiological relevance of the BIX treatment remains to be established, since previous studies showed that BIX protects neurons from stress-induced cell death [51], BIX is promising to be further developed to correct GABAA receptor misfolding diseases. Next we determined the influence of BIX on the ER proteostasis network. BIX treatment (12 μM, 24 h) did not significantly change the expression levels of several major ER chaperones, including

Grp94, calreticulin and calnexin in HEK293T cells stably expressing α1 (A322D)β2γ2 receptors (Fig 16A, quantification shown in Fig 16B), indicating that BIX’s prominent role is to induce BiP expression. Therefore, we evaluated whether BIX’s effect on α1 (A322D) subunits depends on the BiP expression level. As expected, BIX treatment increased the

α1(A322D) protein levels in the non-targeting siRNA treatment groups in HEK293T cells stably expressing α1(A322D)β2γ2 receptors (Fig 16C, cf. lanes 3 and 4 to lanes 1 and 2)

(α1 (A322D) subunits quantification shown in Fig 16D and BiP quantification shown in

Fig 16E). After BiP expression was depleted with BiP siRNA treatment, BIX treatment

did not significantly change the α1(A322D) protein levels (Fig 16C, cf. lanes 5 and 6 to

lanes 1 and 2, quantification shown in Fig 16D). The data indicated that the effect of BIX

on α1(A322D) subunits is by modulating BiP expression level.

3.4 Discussion:

Here we show that BiX promotes the maturation and functional surface expression of

A322D GABAARs through the increased level of BiP expression. As BiX treatment alone

will enhance total intracellular level of A322D α1 subunits and BiP but knocking down

BiP level in BiX treatment group will counteract on the BiX’s effect on A322D α1 subunit

and BiP. BiX also enhances the surface expression of both WT and D219N GABAARs.

BiP, calnexin, GRP94 and disulfide isomerase are all proteins that can deliver terminal

misfolded proteins to ERAD. However, increasing the availability of BiP’s effect on

A322D α1 subunits act by promoting folding in contrast to GRP94’s degradation promoting effect. Under normal physiological condition, A322D α1 subunits seem to be directed to the ERAD pathway by GRP94 rather than being rescued by BiP, although both

GRP94 and BiP associate more strongerly with A322D α1 subunits compared to WTs.

Why under basal conditions, BiP is not able to rescue A322D α1 subunits from being degraded? Probably because there is little free BiP available but more free GRP94

available to A322D α1 subunits. When free soluble BiP’s availability is significantly

increased, then it can compete with GRP94 for misfolded substrates. So it is interesting to

know the difference between the BiP-substrate interaction and GRP94-substrate interaction.

For D219N mutant subunits, increased availability of ER soluble BiP increases its

interaction with N-terminus of the mutant subunits located in the ER. Another very

interesting question is that how BiP can promote the maturation of A322D mutant with a

misfolded M3 domain? It is previously shown that BiP is docked at the ER side of the

translocon Sec61 by its co-chaperone Sec63. When newly synthesized polypeptide exits

via the translocon, BiP will first associate with it and serve to stabilize its substrate and

prevent it from premature folding. It is reported by Spiess et al group that BiP binding to

the sequence preceding the transmembrane segment can promote the membrane integration

of the transmembrane domain. Without BiP binding, the membrane integration of the

transmembrane domain will need to overcome a higher energy barrier. It is possible that

increased availability of soluble BiP causes more BiP to interact with the N-terminus

domain of GABAARs subunits while they are still located at translocon Sec61 and their

synthesis and membrane integration is still going on as GABAAR monomer may have

exposed signal in the N-terminus for BiP binding. The binding of BiP to the N-terminus somehow reduces the energy barrier that is required for membrane insertion for not only

A322D mutant subunits but also WT subunits. The binding of BiP will then increase the probability that A322D mutant subunits can incorporate into the ER membrane and also increase the folding and membrane insertion efficiency of WT subunits. Replacement of alanine which has hydrophobic side chain with negatively charged aspartic acid increases the energy required for M3 domain insertion. A second possibility is that BiP recognizes

the hydrophobic M3 region which is exposed in the ER lumen. Is it possible that association

of BiP with M3 domain can promote the membrane insertion by itself. Alternatively, BiP

drags the protein close to its co-factors such as the translocon or other chaperones which

together with BiP promotes the membrane insertion. Genetic overexpression and our small

molecule treatment both show that increased availability of BiP itself is sufficient to

promote the folding, forwarding trafficking and the functional surface expression level of

mutant GABAARs.

Comparing BiP overexpression and UPR activation

ATF6 or sXBP1 pathway activation will also enhance the expression level of

degradation factor GRP94 which can compete with BiP for substrate and favor the

degradation of mutant receptors. The general effect of UPR activation to A322D

GABAARs is the balance between its trafficking and degradation promoting effects.

Compared to the effect of ATF6 or sXBP1 pathway activation on trafficking of A322D

GABAARs, application of BiX and upregulating BiP is sufficient enough to promote the

functional surface level of A322D GABAARs substantially without promoting their

degradation. It is known that overexpressing BiP is able to prevent the ER stress response

activation or alleviate ER stress induced cell apoptosis. Activating the ATF6 or sXBP1

pathways induces the expression of critical chaperones including BiP by a series of

intracellular processings. Overexpression of BiP may be more energy saving compared to

UPR activation as overexpression of BiP only decreases the likelihood and the scale of the cell homeostasis being disturbed compared to the UPR activation strategy. This advantage is the most prominent if BiP is the only critical factor that can serve to promote folding, increase degradation, stabilize protein intermediates in disease models. But in some disease models such as neurodegenerative diseases, BiP application may be effective in the very early stage while modulating UPR pathways may have a better effect on improving the maturation of aggregation prone proteins, enhancing their degradation and clearance.

Studies on comparison between BiP overexpression and UPR activation in those neurodegenerative disease models are lacking. Chemical genetic activation of ATF6 or sXBP1 and application of ATF6 activator are believed to mimic their physiological activation. Mild activation of ATF6 is believed to be safe in humans. However, BiP expression level is significantly upregulated in cancer cells. ATF6 activation also leads to cell apoptosis. Studies on the long term effects of BiX, UPR activators in animal models are necessary.

3.5 Figures:

Figure 11. BIX, a potent BiP inducer, enhances total expression level of α1(A322D) in a dose and time dependent manner. (A) Chemical structure of BIX. (B, D, E) Dose response of BIX treatment in regulating α1(A322D) total protein level. HEK293T cells

stably expressing α1(A322D)β2γ2 GABAA receptors were treated with BIX at the indicated concentrations or the vehicle control DMSO in the cell culture media for 24 h.

Cells were then lysed and subjected to SDS-PAGE and Western blot analysis (B).

Normalized band intensities for α1(A322D) subunits and BiP are shown in (D) and (E) (n

= 8). (C, F, G) Time course of BIX treatment in regulating α1(A322D) total protein level.

HEK293T cells stably expressing α1(A322D)β2γ2 GABAA receptors were treated with

BIX (12 µM) for the indicated time. Cells were then lysed and subjected to SDS-PAGE and Western blot analysis (C). Normalized band intensities for α1(A322D) subunits and

BiP are shown in (F) and (G) (n = 5). (H) HEK293T cells stably expressing

α1(A322D)β2γ2 GABAA receptors were plated into a 96-well plate on day 1. Cells were then treated with BIX at the indicated concentrations or the vehicle control DMSO in the cell culture media for 24 h. One group of HEK293T cells stably expressing

α1(A322D)β2γ2 GABAA receptors are treated with thapsigargin (2 μM, 7h) as cell toxicity positive control. Resazurin (0.15mg/ml dissolved in DPBS) is added to cells 1.5 h before plate reading. Fluorescence signal at 560 nm excitation / 590 nm emission was measured.

The ratios of fluorescence signal in the DMSO treatment group to treatment groups is shown in (H) (n=4, one-way ANOVA). Statistical significance was evaluated using one-

way ANOVA followed by post-hoc Tukey test in (D), (E), (F), and (G). *, p<0.05. Error bar = SEM.

Figure 11

Figure 12. Influence of GABAA receptor protein expression levels on the effect of BIX treatment.

(A) HEK293T cells were either transiently transfected with 0.15 μg α1(A322D) subunit,

0.15 μg β2 subunit, and 0.15 μg γ2 subunit of GABAA receptors or 0.25 μg α1(A322D) subunit, 0.25 μg β2 subunit, and 0.25 μg γ2 subunit of GABAA receptors. 0.15 μg per subunit group and 0.25 μg per subunit group were then treated with BIX (12 μM, 24h) or

DMSO as controls. Forty-eight hours post transfection, cells were lysed, and total proteins were extracted. The cell lysates were then subjected to SDS-PAGE and Western blot analysis using corresponding antibodies. Quantification of total cellular protein expression levels of α1 and BiP is shown in (B & C) (n = 4, paired t-test). *, p<0.05.

Figure 12

Figure 13. BIX, a potent BiP inducer, promotes the maturation and reduces the

degradation of α1(A322D) subunits. (A) HEK293T cells expressing α1(A322D)β2γ2

receptors were treated with BIX (12 µM, 24 h) or DMSO vehicle control. Then cells were

lysed, and total proteins were extracted. Total cellular proteins were incubated with or without endoglycosidase H enzyme (endo H) or peptide-N-glycosidase F (PNGase F) for

1h at 37°C and then subjected to SDS-PAGE and Western blot analysis. Endo H resistant

α1 subunit bands (top arrows, lanes 6-9) represent properly folded, post-ER α1 subunit glycoforms that traffic at least to the Golgi compartment, whereas endo H sensitive α1

subunit bands (bottom arrow, lanes 6-9) represent immature α1 subunit glycoforms that

are retained in the ER. The PNGase F enzyme cleaves between the innermost N-acetyl-D-

glucosamine and asparagine residues from N-linked glycoproteins, serving as a control for

unglycosylated α1 subunits (lane 5). The ratio of endo H resistant α1 / total α1, which was

calculated from endo H-resistant band intensity / (endo H-resistant + endo H-sensitive band

intensity), serves as a measure of trafficking efficiency of the α1(A322D) subunit.

Quantification of this ratio after endo H treatment (lanes 6-9) is shown in (B) (n = 3, paired

t-test). (C) HEK293T cells stably expressing α1(A322D)β2γ2 receptors were either treated

with DMSO vehicle control, or BIX (12 µM, 24 h) or BIX (12 µM, 24 h) and lactacystin

(2.5μM, 24h). Cycloheximide (150 μg/ml), a protein synthesis inhibitor, was added to

different cell groups for 0, 0.5 hr, 1 hr, and 2 hrs. Cells were then lysed and subjected to

SDS-PAGE and western blot analysis. The quantitation results are shown in (D) (n=5,

one-way ANOVA followed by Fisher test, *, p<0.05 between DMSO vehicle control and

treatment groups, #, p<0.05 between BIX group and BIX + Lac group).

Figure 13

Fig 14. BIX enhances the functional surface expression of A322D α1 subunit of

GABAA receptors. (A) HEK293T cells expressing α1(A322D)β2γ2 receptors were treated with BIX (12 µM, 24 h) or DMSO vehicle control. Then the cell surface proteins were

tagged with biotin using membrane-impermeable biotinylation reagent sulfo-NHS SS-

Biotin. Biotinylated surface proteins were affinity-purified using neutravidin-conjugated beads and then subjected to SDS-PAGE and Western blot analysis. The Na+/K+-ATPase

serves as a surface protein loading control. Quantification of normalized surface α1(A322D)

protein levels to the Na+/K+-ATPase controls is shown in (B) (n = 5, paired t-test). (C) SH-

SY5Y cells stably expressing α1(A322D)β2γ2 receptors were treated with BIX (12 µM, 24 h) or DMSO vehicle control. Then surface biotinylation assay was performed as in (A).

Quantification of normalized surface α1(A322D) protein levels is shown in (D) (n = 3, two tailed student t-test). * p < 0.05. (E) Representative whole-cell patch clamping recording traces in monoclonal HEK293T cells stably expressing α1(A322D)β2γ2 GABAA receptors.

Cells were treated with BIX (12 µM, 24h) or DMSO before voltage clamping. GABA

(3mM) was applied to induce chloride currents with a holding potential of -60 mV.

Quantification of the peak currents (Imax) is shown in (F). The number of patched cells in

each group is shown on the top of the bar. pA: picoampere.

Figure 14

Figure 15. BIX enhances the functional surface expression of α1 subunit variants of

GABAA receptors.

(A) HEK293T cells expressing α1β2γ2 receptors or α1(D219N)β2γ2 receptors were treated

with BIX (12 µM, 24 h) or DMSO vehicle control. Then the cell surface proteins were

tagged with biotin using membrane-impermeable biotinylation reagent sulfo-NHS SS-

Biotin. Biotinylated surface proteins were affinity-purified using neutravidin-conjugated beads and then subjected to SDS-PAGE and Western blot analysis. The Na+/K+-ATPase

serves as a surface protein loading control. Quantification of normalized total and surface

WT α1 protein levels is shown in (B & D) (n = 6 for total and n=5 for surface, paired t-

test). Quantification of normalized total and surface α1(D219N) protein levels is shown in

(C & E) (n = 6 for total and surface, paired t-test

Figure 15

Figure 16. BIX’s enhancing of trafficking of α1(A322D)β2γ2 receptors effect is

through overexpression of BiP. (A and B) HEK293T cells expressing α1(A322D)β2γ2

receptors were treated with BIX (12μM, 24h). The cell lysates were then subjected to SDS-

PAGE and Western blot analysis using corresponding antibodies (A). Quantification of normalized total cellular chaperone protein expression levels is shown in (B) (n=4, paired t-test). (C) HEK293T cells stably expressing α1(A322D)β2γ2 receptors were treated with

non-targeting (NT) or BiP siRNA for 48 hrs. Cells were then treated either with BIX (12

μM) or DMSO vehicle control for another 24 hrs. Cells were then lysed and subjected to

SDS-PAGE and western blot analysis. The quantitation results of α1(A322D) and BiP are

shown in (D&E) (n=3, one-way ANOVA). * p < 0.05; ** p < 0.01.

Figure 16

Chapter 4: LMAN1 (ERGIC-53) promotes trafficking of cys-loop neuroreceptors

This content was published on Y.L. Fu, B. Zhang, T.W. Mu, LMAN1 (ERGIC-53)

promotes trafficking of neuroreceptors, Biochem Biophys Res Commun 511 (2019) 356-

362 and was supported by the National Institute of Health (R01NS105789 to TM;

R01HL094505 and R03CA202131 to BZ).

4.1 introduction

The anterograde transport from the endoplasmic reticulum (ER) to the Golgi

apparatus is critical in guaranteeing properly folded and assembled proteins in eukaryotic

cells to reach their destinations in order to perform their functions [205]. The ER export

process of such cargo proteins (secretory and membrane proteins) begins with their

packaging into the COPII-coated vesicles. To bridge the soluble cargo proteins that are

processed in the ER lumen and COPII coats that are in the cytosolic side of the ER

membrane, cargo receptors that span the ER membrane are required to interact

simultaneously with cargo proteins and coat subunits. Although in principle integral

membrane proteins can directly bind COPII subunits if they contain COPII binding motifs,

cargo receptors for membrane proteins have been demonstrated, such as Erv14p in yeast

[206]. Probably the best-characterized cargo receptor in mammals is the ER-Golgi

intermediate compartment protein-53 (ERGIC-53, aka LMAN1). It cycles between the ER

and the Golgi through COPII and COPI dependent pathways [207]. LMAN1 is a known

cargo receptor for a number of soluble proteins, including blood clotting factor V and factor

VIII, cathepsin C, cathepsin Z, α1-antitrypsin, and matrix metalloproteinase-9 (MMP9)

[205,207,208,209,210]. Mutations in LMAN1 lead to the genetic bleeding disorder

combined deficiency of FV and FVIII [211,212]. As a type I transmembrane protein,

LMAN1 recognizes and binds to the high mannose structure on the glycoprotein substrates

through its ER luminal carbohydrate recognition domain (CRD) [213,214,215,216]. Its cytosolic C-terminus contains a diphenylalanine motif that binds to the COPII coat, which facilitates the anterograde transport, and a dilysine motif that binds to the COPI coat, which mediates the retrograde transport [211].

Figure 17. A schematic cartoon figure of the LMAN1 trafficking between the ER and

the ER to Golgi Compartment (ERGIC). LMAN1 monomers form hexamers in cells.

With the presence of high Ca2+ in the ER, LMAN1 hexamers bind to glycan structure on

its glycoprotein substrates. It then facilitate its substrate being incorporated into COPII-

coated vesicles by its FF motif interacting with COPII coats. When LMAN1-substrate is

transported into ERGIC, the drop in the Ca2+ concentration in ERGIC disrupt the its interaction. As a result, the substrate is freed and LMAN1 is incorporated into COPI-

coated vesicles through its KK motif interacting with COPI coats and retrotransloated back

to ER.

Figure 17

We focused on proteostasis maintenance of neuroreceptors, specifically the pentameric

Cys-loop receptors [217]. They play an essential role in the nervous system and include γ- aminobutyric acid type A receptors (GABAARs), nicotinic acetylcholine receptors

(nAChRs), 5-hydroxytryptamine type-3 receptors (5HT3Rs), and glycine receptors (GlyRs)

[218]. To reach their final destination, individual subunits undergo synthesis, folding and

assembly into pentameric receptors in the ER and then go through the ER-to-Golgi

transport, post Golgi transport and surface membrane insertion [29,219,220,221,222,223].

Disturbing any biogenesis step of these neuroreceptors could influence their surface

expression level and the functional synaptic transmission, and impairment of the

neurotransmission process can lead to related neurological diseases [217]. For example, for GABAARs, several proteins are known to be critical in their post-Golgi trafficking onto the plasma membrane, including GABARAP (GABAAR-associated protein), Hap1

(Huntington-associated protein 1), KIF5A, multidomain protein Muskelin [220,221].

Although the post-Golgi trafficking pathway of GABAARs is relatively well-understood

[51], little is known about their regulatory factors that control the ER-to-Golgi transport.

Our lab’s previous proteomics study identified LMAN1 as an interacting protein for GABAARs [110]. To better understand the role of LMAN1 in the ER-to-Golgi

trafficking of GABAARs and other Cys-loop receptors, we determined the influence of

genetic manipulation of LMAN1 on the endogenous protein levels of these neuroreceptors.

Furthermore, we reporte how LMAN1 knockout regulate the endogenous proteostasis network and the glycan-independent interaction between LMAN1 and GABAARs.

4.2 Materials and methods

4.2.1 Plasmids, siRNAs, and antibodies

The pCMV6 plasmids containing human GABAA receptor α1 (Uniprot no.

P14867-1), β2 (isoform 2, Uniprot no. P47870-1), and γ2 (isoform 2, Uniprot no. P18507-

2) subunits and the pCMV6 Entry Vector plasmid (pCMV6-EV) were obtained from

Origene. The Flag-tagged wild type (WT), ΔCRD, ΔHM, N156A, D181A, KKAA,

ΔHelix, Δβ1, Δβ2, Δβ3 and Δβ4 LMAN1 plasmids were constructed as previously described [224]. Mouse LMAN1 siRNA-1 (# J-050981-11-0005), Mouse LMAN1 siRNA-2 (# J-050981-09-0005) and non-targeting siRNA (D-001810-01-20) were obtained from Dharmacon. The mouse monoclonal anti-α1 (clone BD24) antibodies came from Millipore (#MAB339). The mouse monoclonal anti-β-actin antibody (#A1978) and anti-FLAG M2 peroxidase antibody (#A8592) were obtained from Sigma. The mouse anti-GABAA receptor β3 antibody was from Neuromab (#75149). The rabbit monoclonal anti-LMAN1 antibody was from Abcam (#Ab125006). The rabbit polyclonal anti-

GABAA receptor γ2 antibody is from Synaptic Systems (#224 003). The rabbit polyclonal anti-5HT3A receptor antibody was from Abcam (#Ab13897). The rabbit polyclonal anti- nAChR receptor α4 antibody was from Abcam (#Ab88239). The rabbit polyclonal anti-

BiP antibody was from Abgent (#AP50016). The rat monoclonal anti-GRP94 antibody was from Enzo Life Science (#ADI-SPA-850). The rabbit polyclonal anti-ERP44 antibody was from GeneTex (#106636). The rabbit polyclonal anti-Calnexin antibody was from Enzo Life Science (#ADI-SPA-860-F). The rabbit polyclonal anti-HSP70 antibody was from Enzo Life Science (#ADI-SPA-812-F). The rabbit polyclonal anti-

P4HB antibody was from Abgent (#AP2911B-EV20). The rabbit polyclonal anti-Sec13

antibody was from Abgent (#AP10738CS). The rabbit monoclonal anti-Na, K-ATPase

was from Abcam (#ab76020).

4.2.2 Cell culture and transfection Human HEK293T cells (ATCC) and mouse GT1-7 cells (Professor Pamela Mellon,

UCSD) were maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Hyclone) with

10% heat-inactivated fetal bovine serum (Sigma-Aldrich) and 1% Penicillin-Streptomycin

(Hyclone) at 37°C in 5% CO2. Monolayers were passaged upon reaching confluency with

Trypsin 0.05% (Hyclone). Cells were grown in 10-cm dishes and allowed to reach ~60-80% confluency before transient transfection using TransIT-2020 (Mirus) according to the manufacturer’s instruction. The HEK293T cell line that stably expressing α1β2γ2

GABAARs was generated by transient transfection with α1:β2:γ2 (1:1:1) plasmids. Then

cells were selected using 0.8 mg/mL G-418 (Enzo Life Sciences) and maintained in 0.5

mg/mL G-418. For siRNA transfections, cells were treated with 50 nM LMAN1 siRNA or

non-targeting (NT) siRNA using HiPerfect transfection reagent (Qiagen) for 68 hrs before

protein analysis.

4.2.3 Western blot analysis and immunoprecipitation Cells were rinsed with ice-cold Dulbecco’s Phosphate-Bufferd Saline (DPBS) twice before harvested with Co-IP buffer (50 mM Tris, pH 7.5, 150 mM NaCl, and 1%

Triton X-100) supplemented with Roche complete protease inhibitor cocktail. The lysates were then rotated for an hour at 4 °C and subject to centrifugation (16,000 × g, 15 min, 4

°C) to remove cell debris and nucleus. The supernatant was collected as the total cellular protein. Protein concentrations were determined by the MicroBCA assay (Pierce). Cell lysates were mixed with 4x SDS sample buffer in the presence of β-mercaptoethanol and separated in an 8% denaturing Tris-glycine gel. Western blot analysis was performed using

100

appropriate antibodies. Band intensities were quantified using the Image J software from the NIH.

Cell lysates (500 μg) were precleared with 30 μL of protein A/G Plus-agarose beads

(Santa Cruz Biotechnology) and 1.0 μg of normal mouse IgG for 1 h at 4 °C to remove nonspecific binding proteins. The precleared cell lysates were incubated with 30 µL mouse anti-Flag M2 magnetic beads (M8823, Sigma Aldrich) or normal protein A/G plus agarose beads (negative control for nonspecific binding) overnight at 4 °C. The magnetic beads were collected using a magnetic separation stand (Promega) and washed three times with the Co-IP buffer. The Flag-tagged proteins were eluted by incubation with 30 μL of SDS loading buffer in the presence of β-mercaptoethanol. The immunopurified eluents were separated in an 8% SDS-PAGE gel, and Western blot analysis was performed.

4.2.4 Biotinylation of cell surface proteins

GT1-7 cells were plated in 10-cm dishes for surface biotinylation experiments according to published procedure [111]. Then, intact cells were rinsed gently twice with ice-cold PBS and incubated with the membrane-impermeable biotinylation reagent Sulfo-

NHS SS-Biotin (0.5 mg/ mL; Pierce) in PBS containing 0.1 mM CaCl2 and 1 mM MgCl2

(PBS+CM) for 30 min at 4 °C to label surface membrane proteins. The reaction was

quenched by incubating the cells with 10 mM glycine in ice-cold PBS+CM for 5 min at

4 °C. Sulfhydryl groups were blocked by incubating the cells with 5 nM N-ethylmaleimide

(NEM) in PBS for 15 min at room temperature. Cells were solubilized for 1 h at 4 °C in lysis buffer (Triton X-100, 1%; SDS 0.1%, Tris–HCl, 50 mM; NaCl, 150 mM; and EDTA,

5 mM; pH 7.5) supplemented with Roche complete protease inhibitor cocktail and 5 mM

NEM. The lysates were cleared by centrifugation (16,000 × g, 15 min at 4 °C) to pellet

101

cellular debris. The supernatant contained the biotinylated surface proteins. MicroBCA

assay (Pierce) was then performed to measure the protein concentration in the supernatant.

Biotinylated surface proteins were affinity-purified from the above supernatant by incubating for 1 h at 4 °C with 30-50 μL of immobilized neutravidin-conjugated agarose bead slurry (Pierce). The beads were washed with Co-IP buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1% Triton X-100) twice followed by Co-IP buffer without Triton X-100 twice.

Surface proteins were eluted by boiling for 5 mins with 30-60 μL of LSB ⁄ Urea buffer (2x

Laemmli sample buffer (LSB) with 100 mM DTT and 6 M urea; pH 6.8) for SDS-PAGE and Western blotting analysis.

4.2.5 Mouse whole brain sample preparation Whole mouse brains were collected on ice and snap frozen in liquid nitrogen and stored at -80 °C. On the day of experiments, tissues were briefly thawed on ice and homogenized in homogenization buffer (25mM Tris-HCl pH 7.6, 150 mM NaCl, 1 mM

EDTA, 2% Triton-X-100 supplemented with Roche protease inhibitors) using a plastic micro tissue homogenizer. Homogenates were centrifuged at 800 g for 10 mins at 4 °C and supernatants were collected. Additional homogenization buffer was added to the pellet and homogenizing procedure was repeated. Supernatants were combined and rotated at 4 °C for 2-4 hrs. Debris in tissue lysates was removed by centrifugation at 13500 g for 20 mins and 18400 g for 30 mins at 4 °C. The animal studies followed the guidelines of the

Institutional Animal Care and Use Committees (IACUC) at Case Western Reserve

University and Cleveland Clinic Institutional Review Board.

4.2.6 Statistical analysis

102

All data were presented as mean ± SEM. Statistical significance was evaluated

using two-tailed Student’s t-Test. A p value of less than 0.05 was considered statistically

significant.

4.3 Results

4.3.1 LMAN1 positively regulates the surface expression of endogenous GABAAR

subunits.

Because our previous tandem mass spectrometry-based proteomics analysis

identified LMAN1 as an interactor for GABAARs [110] and LMAN1 is a trafficking factor from the ER to Golgi, we tested whether LMAN1 regulates the receptors’ surface

trafficking. To assess endogenous GABAARs, we used mouse GT1-7 hypothalamic GnRH

neuronal cells, which express endogenous α1 and β3 subunits of GABAARs [225,226]. If

LMAN1 plays a role in the anterograde transport of GABAARs from the ER to the Golgi,

knocking down LMAN1 would interfere with the ER-to-Golgi trafficking process, and as a result, surface level of endogenous GABAARs will be decreased. As expected, surface

biotinylation experiments demonstrated that the surface level of β3 subunits was reduced

significantly (Figs. 18A & 18B) after knocking down of LMAN1 expression level by

treating GT1-7 neurons with LMAN1 siRNA-1 (Figs. 18C & 18E). Such a reduction of the

surface β3 subunits was also significant (Figs. 19A & 19B) by using LMAN1 siRNA-2

(Figs. 19C & 19E). LMAN1 siRNA-2 treatment to GT1-7 neurons also decreased the total

intracellular level of β3 subunits significantly (Figs. 19C & 19D); LMAN1 siRNA-1

treatment to GT1-7 neurons decreased total intracellular levels of β3 subunits in 6 out of 7

103

groups, but not significantly (Figs. 18C & 18D). However, we did not find good anti-α1

antibody to detect the surface level of endogenous α1 subunits in GT1-7 neurons.

Nonetheless, the β3 subunit result indicated that reducing the endogenous LMAN1 level

attenuates the surface trafficking of endogenous GABAAR subunits.

4.3.2 LMAN1 has a more general role for the Cys-loop neuroreceptors.

Because GABAARs belong to the Cys-loop superfamily neuroreceptors [217,218],

we continued to determine whether LMAN1 has a general role within this superfamily,

which also includes nAChRs and 5-HT3Rs. We utilized the LMAN1 knockout mice, which were previously shown to result in combined deficiency of plasma factor V and factor VIII

[227]. Five whole mouse brains were collected to access endogenous neuroreceptors: two were wild type (WT) controls (Lman1 +/+), and three were LMAN1 knockouts (Lman1 -

/-). Knockout of LMAN1 in the brain was confirmed by Western Blot analysis (Fig. 20A,

cf. lanes 2, 4, and 5 to lanes 1 and 3). Depleting LMAN1 decreased the total protein level

of γ2 subunits of GABAARs significantly (Fig. 20B, cf. lanes 2, 4, and 5 to lanes 1 and 3;

quantification shown in Fig. 20C). Due to the unavailability of proper anti-GABAAR β2

or β3 antibodies for mouse brain tissues, we could not evaluate the β subunits. Interestingly,

LMAN1 knockout also significantly decreased the total protein level of 5-HT3A subunits

(Fig. 20D, cf. lanes 2, 4, and 5 to lanes 1 and 3; quantification shown in Fig. 20E), whereas

LMAN1 depletion did not seem to influence the protein level of the α4 subunits of nAChRs

significantly (Fig. 20F, cf. lanes 2, 4, and 5 to lanes 1 and 3; quantification shown in Fig.

20G). Collectively, these results indicated that LMAN1 has a more general role for endogenous Cys-loop receptor subunits.

104

4.3.3 Influence of LMAN1 knockout on the proteostasis network in the central nervous system

Because the proteostasis network (including chaperones, degradation factors, and trafficking factors) plays an essential role in controlling the biogenesis of membrane proteins, we evaluated how loss of LMAN1 affects the proteostasis network in the central nervous system [228,229]. It was previously reported that in Lman1-/- mouse liver, the total protein level of GRP78 increased substantially without significant induction of the unfolded protein response (UPR) genes, including Grp78, Grp94, Xbp1, Chop and Atf4

[227]. However, it is unknown how the proteostasis network was influenced in Lman1-/- mice brains. We evaluated the protein expression levels of major chaperones in the ER

(including GRP78, GRP94, and calnexin), folding enzymes in the ER (including P4HB and

ERP44), Hsp70 in the cytosol, and a COPII subunit in the cytosol (Sec13a) in Lman1+/+ and Lman1-/- mice brain homogenates (Figs. 21A-F & 22A-H). Western blotting analysis demonstrated that among these proteins, only the ERP44 level was increased significantly in LMAN1 knockout brain (Figs. 21E, F). These results indicated that loss of LMAN1 leads to limited changes in the proteostasis network in the brain.

4.3.4 LMAN1 interacts with GABAARs in a glycan-independent manner.

We next evaluated whether LMAN1 interacts with neuroreceptors to act as their

cargo receptors. We focused on GABAAR subunits because our previous proteomics study revealed that LMAN1 binds GABAARs [110]. Indeed, co-immunoprecipitation

experiments demonstrated that pulling down endogenous LMAN1 leads to the detection of

α1 subunits in HEK293T cells transiently expressing α1β2γ2 GABAARs (Fig. 23A, lane

3), indicating that LMAN1 acts as a cargo receptor for GABAARs. For its soluble cargo

105

glycoproteins, LMAN1 is known to interact with them through their mannose-rich glycans.

To evaluate whether LMAN1 also binds to the glycans installed on the α1 subunits, we

mutated the two N-glycosylation sites (Asn 38 and Asn 138) into glutamine to generate the

N38Q/N138Q α1 subunits, which cannot be glycosylated in the ER lumen. Intriguingly,

co-immunoprecipitation experiments demonstrated that such mutant α1 subunits interacted

with endogenous LMAN1 (Fig. 23A, lane 4), and the interaction in the mutant α1 form

was much stronger than that in the WT α1 form (Fig. 23A, cf. lane 4 to 3), indicating that

LMAN1 binds GABAAR α1 subunits independent of the glycan structure.

Furthermore, we evaluated which domain of LMAN1 plays an important role in its interaction with the α1 subunits. The LMAN1 mutations constructs were displayed in Fig.

23B. Co-immunoprecipitation experiments were used to determine how mutations in

LMAN1 affected its interaction with α1 subunits. It was previously reported that N156A or D181A mutation in LMAN1 disrupts its binding ability to mannose structure on its soluble substrates. Interestingly, the N156A or D181A mutation in LMAN1 did not affect the interaction between LMAN1 and α1 subunits (Fig. 23C, cf. lanes 5, 6 to lane 2),

consistent with their glycan-independent binding (Fig. 23A). The deletion of the complete

CRD domain disrupted the interaction between LMAN1 and α1 subunits (Fig. 23C, cf.

lane 3 to lane 2), possibly because deletion of such a domain induced global protein

conformational changes. In addition, neither four β-sheets in the CRD domain nor the helix domain were required for α1 subunits binding (Fig. 23C, cf. lane 8-12 to lane 2).

Oligomerization status of LMAN1 did not affect the α1 subunits binding because the ΔHM mutant, in which helix domain was deleted and both cysteine at 466 and 475 position were

106

mutated to disrupt the formation of the LMAN1 hexamer through disulfide bonds, also

interacted with α1 subunits (Fig. 23C, cf. lane 4 to lane 2). However, when the FF ER exit sorting signal was substituted for AA in LMAN1, this ER-retaining KKAA mutation interfered with the interaction between LMAN1 and α1 subunits (Fig. 23C, cf. lane 7 to lane 2). In conclusion, LMAN1 interacts with GABAARs independent of their

glycosylation status.

4.4 Discussion

In this study we demonstrated that LMAN1 interacts with membrane protein

GABAARs in a glycan-independent manner. Furthermore, LMAN1 positively affects the

surface level of GABAARs. These results indicated that LMAN1 also serves as a trafficking

factor for transmembrane proteins that are targeted to cell surface such as GABAARs, in addition to its known role of transporting soluble proteins. Since knockout of LMAN1 also affects the intracellular level of 5HT3A receptors, LMAN1 should have a more general role

in the anterograde transport of neuroreceptors.

Because we demonstrated that the glycan structure on the membrane proteins in the

ER lumen is not necessary for the LMAN1 interaction, LMAN1 must use other signals to

detect trafficking-competent GABAARs. Previously, it was shown that the interaction

between FVIII and LMAN1 is calcium dependent but is not dependent on the glycosylation

status of FVIII [230]. How LMAN1 interacts with transmembrane proteins like GABAARs

needs to be further elucidated. One intriguing possibility is that LMAN1 binds the

transmembrane domain of its cargo protein within the lipid bilayer. Erv14p, another cargo

107

receptor in yeast, was demonstrated to interact with its cargo through the intra-membrane interactions [231].

Knockout of LMAN1 in mice increases the intracellular level of an important ER chaperone ERP44. The upregulated ERP44 level upon LMAN1 depletion could have at least two possible effects. First, ERP44, also known as PDIA10, belongs to the protein disulfide isomerase (PDI) family, which catalyzes the formation of disulfide bond to assist protein folding and assembly. Therefore, ERP44 could be upregulated to counter the influence of loss of LMAN1 to handle the membrane proteins that were retained in the ER.

Such an effect is also consistent with the role of ERP44 as a downstream target of the UPR pathway [232] because the reduced trafficking from the ER to Golgi could cause the ER to sense an increased burden of proteins retaining in the ER, leading to the activation of the

UPR and its downstream chaperones. Future experiments are required to evaluate whether

ERP44 upregulation is through UPR activation. Second, ERP44 cycles between the ER and the cis-Golgi in a pH-dependent manner, and the pH gradient among ER, ERGIC, and cis-Golgi regulates the conformational changes of ERP44 for its binding with the clients

[233,234]. Therefore, ERP44 could transport the neuroreceptors from the ER to the Golgi independent of LMAN1. Collectively, these results indicated that loss of LMAN1 can influence the anterograde transport of neuroreceptors from the ER to the Golgi by regulating the proteostasis network.

It was previously shown that LMAN1 knockout mice do not show significant phenotypes with the two exceptions: 1. LMAN1 knockout mice on C57BL6/J background have a higher mortality rate compared to controls; 2. there are partially reduced FV, FVIII and α1-antitrypsin levels in plasma, and the FV level in platelets in LMAN1 knockout mice

108

[210,227]. This further indicated that there are other important trafficking pathways and factors available for ER exit sorting and that elevated chaperone levels may help maintain cell homeostasis.

109

4.5 Figures

Figure 18. Transient knockdown of LMAN1 using LMAN1 siRNA-1 affects the total

and surface expression of endogenous GABAARs subunits. LMAN1 siRNA and non-

targeting (NT) siRNA were applied to mouse hypothalamic GT1-7 neurons. Sixty-eight

hours post transfection, cells were harvested, and protein analysis was performed.

Knockdown of endogenous LMAN1 using LMAN1 siRNA-1 (A) reduces the surface level

of β3 subunits. Cell surface proteins were labeled with membrane-impermeable

biotinylation reagent sulfo-NHS SS-Biotin. Biotinylated surface proteins were affinity-

purified using neutravidin-conjugated beads and then subjected to SDS-PAGE and

Western blot analysis. The Na+/K+-ATPase serves as a surface protein loading control.

Quantification of normalized surface β3 protein levels to the Na+/K+-ATPase controls is shown in (B) (n = 8). Influence of knockdown of endogenous LMAN1 using LMAN1 siRNA-1 (C) on the total protein level of β3 subunits by using SDS-PAGE and Western blot analysis. Quantification of normalized total β3 protein levels to β-actin loading controls is shown in (D) (n = 7). Quantification of the LMAN1 knockdown efficiency is shown in (E) (n = 7). *, p < 0.05, paired student t-test.

110

Figure 18

111

Figure 19. Transient knockdown of LMAN1 using LMAN1 siRNA-2 affects the total

and surface expression of endogenous GABAARs subunits. LMAN1 siRNA and non-

targeting (NT) siRNA were applied to mouse hypothalamic GT1-7 neurons. Sixty-eight

hours post transfection, cells were harvested, and protein analysis was performed.

Knockdown of endogenous LMAN1 using LMAN1 siRNA-2 (A) reduces the surface level

of β3 subunits. Cell surface proteins were labeled with membrane-impermeable

biotinylation reagent sulfo-NHS SS-Biotin. Biotinylated surface proteins were affinity-

purified using neutravidin-conjugated beads and then subjected to SDS-PAGE and

Western blot analysis. The Na+/K+-ATPase serves as a surface protein loading control.

Quantification of normalized surface β3 protein levels to the Na+/K+-ATPase controls is shown in (B) (n = 8). Influence of knockdown of endogenous LMAN1 using LMAN1 siRNA-2 (C) on the total protein level of β3 subunits by using SDS-PAGE and Western blot analysis. Quantification of normalized total β3 protein levels to β-actin loading controls is shown in (D) (n = 7). Quantification of the LMAN1 knockdown efficiency is shown in (E) (n = 7). *, p < 0.05, paired student t-test.

112

Figure 19

113

Figure 20. Knockout of LMAN1 in mouse leads to decreased total expression level of

GABAARs subunits and other Cys-loop family receptor subunits. Whole brain lysates from two WT (LMAN1+/+) and three LMAN1 knockout (LMAN1-/-) mice were subject to

SDS-PAGE. Western blot results of LMAN1, γ2 subunits of GABAARs, 5HT3A subunits, and nAChR α4 subunits are shown in (A), (B), (D), and (F). Corresponding band intensity quantification results are shown in (C), (E), and (G) (n=3). *, p < 0.05.

114

Figure 20

115

Figure 21. Influence of LMAN1 knockout on the proteostasis network in the central nervous system (1). Western blot of whole brain lysates from two WT (LMAN1+/+) and three LMAN1 knockout (LMAN1-/-) mice. Results of GRP78, GRP94, ERP44, are shown in (A), (C), (E). Corresponding band intensity quantification results are shown in (B), (D),

(F) (n=3). *, p < 0.05.

116

Figure 21

117

Figure 22. Influence of LMAN1 knockout on the proteostasis network in the central nervous system (2). Western blot of whole brain lysates from two WT (LMAN1+/+) and three LMAN1 knockout (LMAN1-/-) mice. Results of Calnexin, HSP70, P4HB and Sec13a are shown in (A), (C), (E), (G). Corresponding band intensity quantification results are shown in (B), (D), (F), (H) (n=3). *, p < 0.05.

118

Figure 22

119

Figure 23. LMAN1 interacts with GABAARs in HEK293T cells in a glycan- independent manner. (A) WT or N38Q/N138Q α1 subunits together with β2 and γ2

subunits were overexpressed into HEK293T cells. Endogenous LMAN1 was co- immunoprecipitated using anti-LMAN1 antibody. The western blot result shows that both

WT and N38Q/N138Q α1 subunits are detected (lanes 3 and 4). IgG control results are shown in lanes 5 and 6. (B) Cartoon figure for LMAN1 WT and mutants. For Δβ1 LMAN1 mutant, H43-Q59 in CRD domains is deleted. For Δβ2 LMAN1 mutant, H43-N72 in CRD

domains is deleted. For Δβ3 LMAN1 mutant, H43-S76 in CRD domains is deleted. For

Δβ4 LMAN1 mutant, H43-A83 in CRD domains is deleted. (C) Flag-tagged WT or mutant

LMAN1 was overexpressed in HEK293T cells stably expressing WT GABAARs. WT or

mutant LMAN1 was co-immunoprecipitated using anti-Flag antibody. Only LMAN1 without CRD domain and LMAN1 without ER exit signal (diphenylalanine to di-alanine

mutation) abolish the interaction between LMAN1 and α1 subunits of GABAARs.

120

Figure 23

121

Chapter 5 Conclusions and Future directions:

Chapter 1 has described the all regulating factors that affect surface expression level of

cys-loop receptors with a focus on GABAARs. Not only the kinetics of the GABAARs, but

also affecting mRNA expression level, protein folding, glycosylation, assembly, ER

associated degradation, ER to Golgi trafficking, Golgi to cell surface trafficking, receptors

clustering on cell surface, and endocytosis can affect the surface level of GABAARs and

as a result, affect the strength of inhibitory neuron circuits strength in central nervous

system and the balance between the inhibitory and excitatory regulations on neuronal

activities. Two trafficking deficient mutations A322D and D219N located in the M3

domain or N-terminus domain of α1 subunits lead to their misfolding, exacerbate

degradation through ER, and significantly reduce surface expression level of the receptors.

These mutations lead to epilepsy both in humans carrying the mutations or mutations knocked in animal models. This thesis explores if modulating the proteostasis environment of ER by activating UPR-ATF6 pathway or UPR-sXBP1 pathway or application of a BiP specific inducer BiX can rescue these two trafficking deficient mutant receptors surface expression level. Chapter 2 showed that overexpression of A322D GABAARs in HEK293T

cells does not increase BiP or CHOP mRNA level compared to WT GABAARs 48hrs after

transfection of the receptors. However, detecting the BiP or CHOP mRNA level at 48hrs

post transfection is not enough to prove that UPR has never been activated. As UPR signals

rise and then drop under mild ER stress conditions, it is necessary to add more time points

post transfection to detect BiP and CHOP mRNA levels. BiP translation results from the activation of both ATF6 and sXBP1 pathways. CHOP translation is upregulated when

PERK pathway is activated. Detecting the BiP and CHOP mRNA levels should be

122

sufficient for detecting the UPR activation. However, XBP1 splicing experiment is a classic way of detecting the activation of IRE1 pathway. Detecting the nuclear expression of cleaved form of ATF6 is a more direct way to detect the activation of UPR-ATF6 pathway. Detection the level of BiP, GRP94, CHOP protein levels is also necessary for further testing if the expression of downstream protein targets of UPR is upregulated.

Chapter 2 also shows that activating the UPR-ATF6 pathway by overexpressing full length ATF6 can promote the forward trafficking and surface expression of A322D α1 subunits. As is mentioned in the introduction of Chapter 2, 147 is a mild specific ATF6 activator. It is necessary to detect its effect on the trafficking of A322D and D219N

GABAARs, as small molecules have advantages over gene delivery systems if they are

going to be tested in humans in the future. Chapter 2 also showed that activating UPR-

sXBP1 pathway by overexpressing sXBP1 can promote the forward trafficking and surface

expression of A322D α1 subunits. However, activating UPR-sXBP1 pathway seems to

increase the total protein level of WT α1 subunits while activating UPR-ATF6 pathway

does not. It is interesting to test further the underlying mechanism. Tere used a yellow

fluorescent protein-based essay to detect the different allosteric modulation effects on

α2β3γ2 type GABAARs. The study showed that after CHO-K1 cells are transiently transfected with halide sensing YFP-H148Q/I152L and α2β3γ2 GABAARs, there is

significant difference in quenching of YFP signal before and after the application of

GABAARs agonists such as Diazepam, Lorazepam, Clobazam, and Alpidem. However,

the study also showed that this yellow fluorescent protein-based essay is not sensitive

enough to show difference in YFP signal quenching after the application of mild positive

allosteric modulators of GABAARs such as TPA-023 and L-838417. TPA-023 and L-

123

838417 can increase the potentiation of GABA EC20 by 11% and 34% respectively, while chlordiazepoxide, which belongs to benzodiazepine family, can increase GABA EC20 by

100% [235,236] . EC20 is the concentration that elicits twenty percent maximal response.

Based on this study, I think that patch clamp method I used to test the effect of BiX on

functional surface expression of A322D GABAARs shown in chapter 3 is probably more

suitable compared to this yellow fluorescent protein-based essay to be used to detect the

effect UPR-ATF6 or UPR-sXBP1 pathways activation on the GABA induced currents in

cells expressing A322D or D219N GABAARs. It is also worthwhile to test if mutant

receptors can be degraded through the autophagy pathway. If after the autophagy pathway

is inhibited, the A322D α1 subunits degradation rate is not affected, this is a direct proof

that A322D or D219N α1 subunits are not degraded through autophagy pathway and as a

result, modulating autophagy pathway is not an ideal way to decrease the degradation rate

of mutant GABAAR subunits.

Chapter 3 has shown that application of a BiP inducer BiX can promote the trafficking and functional surface expression level of A322D and D219N GABAARs. As BiP has multi

neuro protection effect and small molecules have advantages over gene delivery systems,

it is necessary to test BiX’s effect on epilepsy mice models that carry A322D or D219N

mutations. It is also necessary to see if BiX can efficiently pass the blood brain barrier to

reach the central nervous system. The effect of BiX and molecule 147 on rescuing A322D

and D219N GABAAR should be compared. It is also necessary to test if there are long term

side effects of applying BiX and molecule 147. For example, if these molecular leads to

cell apoptosis or induce cancers. Although there are many controversies about applying

gene therapy method such as CRISPR/Cas9 to humans carrying various genetic mutations,

124

the cons and pros of using small molecule or gene therapy methods to treat trafficking

deficient mutant diseases such as A322D or D219N caused epilepsy should be compared

in the future. It is also interesting to figure out how GRP94 and BiP compete with the substrates and how BiP improves the membrane insertion efficiency of mutant membrane

domain.

Chapter 4 has shown that LMAN1 positively affects the surface level of GABAARs.

LMAN1 interacts with membrane proteins GABAARs in a glycan-independent manner.

These results indicated that LMAN1 serves as a trafficking factor not only for soluble,

secreted proteins but also for transmembrane proteins that are targeted to cell surface such

as GABAARs. Since knockout of LMAN1 also affects the intracellular level of 5HT3A

receptors, LMAN1 should have a more general role in the anterograde transport of

neuroreceptors. Knockout of LMAN1 in mice increases the intracellular level of an

important ER chaperone ERP44. More questions remain to be answered. How LMAN1

actually interacts with membrane proteins GABAARs? How does LMAN1 regulate ERP44

protein level? What is the function of the ERP44 upregulation? It was previously shown

that surviving LMAN1 knockout mice do not show significant phenotypes as reported

previously with the two exceptions: 1. LMAN1 knockout mice on C57BL6/J background

have a higher prenatal mortality rate compared to controls; 2. there are partially reduced

plasma FV and FVIII, and platelet FV levels in LMAN1 knockout mice. The fact that

LMAN1 knockout C57BL6/J mice have a higher prenatal mortality rate can indicate that

LMAN1 and ER-to-Golgi trafficking may play an important role in the early

developmental stage of C57BL6/J mice. Also, it will be of great interest to evaluate the

neurological behavior of LMAN1 knockout mice. For example, it will be interesting to

125 know if LMAN1 knockout mice will show a predisposition for epilepsy. If so, this can indicate that loss of LMAN1 leads to physiological dysfunctions by affecting the surface expression level of GABAARs in mice. However, since patients with homozygous loss-of- function LMAN1 mutations were only reported to have coagulation factor V and VIII deficiency but not neurological disorders which indicates that there are other ER-to-Golgi trafficking pathways available for neuroreceptors in humans.

126

Bibliography

[1] H.A. Lester, M.I. Dibas, D.S. Dahan, J.F. Leite, D.A. Dougherty, Cys-loop receptors: new twists and turns, Trends Neurosci. 27 (2004) 329-336. [2] S.D. Buckingham, A.K. Jones, L.A. Brown, D.B. Sattelle, Nicotinic acetylcholine receptor signalling: roles in Alzheimer's disease and amyloid neuroprotection, Pharmacol Rev 61 (2009) 39-61. [3] I. Posadas, B. Lopez-Hernandez, V. Cena, Nicotinic receptors in neurodegeneration, Curr Neuropharmacol 11 (2013) 298-314. [4] R.C. Hogg, M. Raggenbass, D. Bertrand, Nicotinic acetylcholine receptors: from structure to brain function, Rev Physiol Biochem Pharmacol 147 (2003) 1-46. [5] S. Mazzaferro, F. Gasparri, K. New, C. Alcaino, M. Faundez, P. Iturriaga Vasquez, R. Vijayan, P.C. Biggin, I. Bermudez, Non-equivalent ligand selectivity of agonist sites in (alpha4beta2)2alpha4 nicotinic acetylcholine receptors: a key determinant of agonist efficacy, J Biol Chem 289 (2014) 21795-21806. [6] E. Palma, S. Bertrand, T. Binzoni, D. Bertrand, Neuronal nicotinic alpha 7 receptor expressed in Xenopus oocytes presents five putative binding sites for methyllycaconitine, J Physiol 491 ( Pt 1) (1996) 151-161. [7] Z.S. Wu, H. Cheng, Y. Jiang, K. Melcher, H.E. Xu, Ion channels gated by acetylcholine and serotonin: structures, biology, and drug discovery, Acta Pharmacol Sin 36 (2015) 895- 907. [8] B. Luscher, T. Fuchs, C.L. Kilpatrick, GABAA receptor trafficking-mediated plasticity of inhibitory synapses, Neuron 70 (2011) 385-409. [9] Z. Nusser, N. Hajos, P. Somogyi, I. Mody, Increased number of synaptic GABA(A) receptors underlies potentiation at hippocampal inhibitory synapses, Nature 395 (1998) 172-177. [10] P.J. Whiting, T.P. Bonnert, R.M. McKernan, S. Farrar, B. Le Bourdelles, R.P. Heavens, D.W. Smith, L. Hewson, M.R. Rigby, D.J. Sirinathsinghji, S.A. Thompson, K.A. Wafford, Molecular and functional diversity of the expanding GABA-A receptor gene family, Ann N Y Acad Sci 868 (1999) 645-653. [11] P. Whiting, R.M. McKernan, L.L. Iversen, Another mechanism for creating diversity in gamma-aminobutyrate type A receptors: RNA splicing directs expression of two forms of gamma 2 phosphorylation site, Proc Natl Acad Sci U S A 87 (1990) 9966-9970. [12] P. Kofuji, J.B. Wang, S.J. Moss, R.L. Huganir, D.R. Burt, Generation of two forms of the gamma-aminobutyric acidA receptor gamma 2-subunit in mice by alternative splicing, J Neurochem 56 (1991) 713-715. [13] M. Vithlani, M. Terunuma, S.J. Moss, The dynamic modulation of GABA(A) receptor trafficking and its role in regulating the plasticity of inhibitory synapses, Physiol Rev 91 (2011) 1009-1022. [14] N. Unwin, Refined structure of the nicotinic acetylcholine receptor at 4A resolution, J Mol Biol 346 (2005) 967-989. [15] R.E. Hibbs, E. Gouaux, Principles of activation and permeation in an anion-selective Cys-loop receptor, Nature 474 (2011) 54-60. [16] N. Bocquet, H. Nury, M. Baaden, C. Le Poupon, J.P. Changeux, M. Delarue, P.J. Corringer, X- ray structure of a pentameric ligand-gated ion channel in an apparently open conformation, Nature 457 (2009) 111-114. [17] R.J. Hilf, R. Dutzler, X-ray structure of a prokaryotic pentameric ligand-gated ion channel, Nature 452 (2008) 375-379.

127

[18] G. Hassaine, C. Deluz, L. Grasso, R. Wyss, M.B. Tol, R. Hovius, A. Graff, H. Stahlberg, T. Tomizaki, A. Desmyter, C. Moreau, X.D. Li, F. Poitevin, H. Vogel, H. Nury, X-ray structure of the mouse serotonin 5-HT3 receptor, Nature 512 (2014) 276-281. [19] S. Basak, Y. Gicheru, S. Rao, M.S.P. Sansom, S. Chakrapani, Cryo-EM reveals two distinct serotonin-bound conformations of full-length 5-HT3A receptor, Nature 563 (2018) 270- 274. [20] P.S. Miller, A.R. Aricescu, Crystal structure of a human GABAA receptor, Nature 512 (2014) 270-275. [21] S. Zhu, C.M. Noviello, J. Teng, R.M. Walsh, Jr., J.J. Kim, R.E. Hibbs, Structure of a human synaptic GABAA receptor, Nature 559 (2018) 67-72. [22] S. Phulera, H. Zhu, J. Yu, D.P. Claxton, N. Yoder, C. Yoshioka, E. Gouaux, Cryo-EM structure of the benzodiazepine-sensitive alpha1beta1gamma2S tri-heteromeric GABAA receptor in complex with GABA, Elife 7 (2018) e39383. [23] J. Du, W. Lu, S. Wu, Y. Cheng, E. Gouaux, Glycine receptor mechanism elucidated by electron cryo-microscopy, Nature 526 (2015) 224-229. [24] W.N. Green, C.P. Wanamaker, The role of the cystine loop in acetylcholine receptor assembly, J Biol Chem 272 (1997) 20945-20953. [25] C.N. Connolly, B.J. Krishek, B.J. McDonald, T.G. Smart, S.J. Moss, Assembly and cell surface expression of heteromeric and homomeric gamma-aminobutyric acid type A receptors, J Biol Chem 271 (1996) 89-96. [26] C.N. Connolly, J.M. Uren, P. Thomas, G.H. Gorrie, A. Gibson, T.G. Smart, S.J. Moss, Subcellular localization and endocytosis of homomeric gamma2 subunit splice variants of gamma-aminobutyric acid type A receptors, Mol Cell Neurosci 13 (1999) 259-271. [27] G.W. Boyd, A.I. Doward, E.F. Kirkness, N.S. Millar, C.N. Connolly, Cell surface expression of 5-hydroxytryptamine type 3 receptors is controlled by an endoplasmic reticulum retention signal, J Biol Chem 278 (2003) 27681-27687. [28] J.M. Wang, L. Zhang, Y. Yao, N. Viroonchatapan, E. Rothe, Z.Z. Wang, A transmembrane motif governs the surface trafficking of nicotinic acetylcholine receptors, Nat Neurosci 5 (2002) 963-970. [29] C.P. Wanamaker, J.C. Christianson, W.N. Green, Regulation of nicotinic acetylcholine receptor assembly, Ann N Y Acad Sci 998 (2003) 66-80. [30] L.W. Wong, H.S. Tae, B.A. Cromer, Assembly, trafficking and function of alpha1beta2gamma2 GABAA receptors are regulated by N-terminal regions, in a subunit-specific manner, J Neurochem 134 (2015) 819-832. [31] L.W. Wong, H.S. Tae, B.A. Cromer, Role of the rho1 GABA(C) receptor N-terminus in assembly, trafficking and function, ACS Chem Neurosci 5 (2014) 1266-1277. [32] V. Tsetlin, D. Kuzmin, I. Kasheverov, Assembly of nicotinic and other Cys-loop receptors, J Neurochem 116 (2011) 734-741. [33] K. Sumikawa, Sequences on the N-terminus of ACh receptor subunits regulate their assembly, Brain Res Mol Brain Res 13 (1992) 349-353. [34] K. Sumikawa, T. Nishizaki, The amino acid residues 1-128 in the alpha subunit of the nicotinic acetylcholine receptor contain assembly signals, Brain Res Mol Brain Res 25 (1994) 257-264. [35] H.J. Kreienkamp, R.K. Maeda, S.M. Sine, P. Taylor, Intersubunit contacts governing assembly of the mammalian nicotinic acetylcholine receptor, Neuron 14 (1995) 635-644. [36] J. Kuhse, B. Laube, D. Magalei, H. Betz, Assembly of the inhibitory glycine receptor: identification of amino acid sequence motifs governing subunit stoichiometry, Neuron 11 (1993) 1049-1056.

128

[37] C.N. Connolly, K.A. Wafford, The Cys-loop superfamily of ligand-gated ion channels: the impact of receptor structure on function, Biochem Soc Trans 32 (2004) 529-534. [38] V. Avramopoulou, A. Mamalaki, S.J. Tzartos, Soluble, oligomeric, and ligand-binding extracellular domain of the human alpha7 acetylcholine receptor expressed in yeast: replacement of the hydrophobic cysteine loop by the hydrophilic loop of the ACh- binding protein enhances protein solubility, J Biol Chem 279 (2004) 38287-38293. [39] A.L. Buller, G.A. Hastings, E.F. Kirkness, C.M. Fraser, Site-directed mutagenesis of N-linked glycosylation sites on the gamma-aminobutyric acid type A receptor alpha 1 subunit, Mol Pharmacol 46 (1994) 858-865. [40] W.Y. Lo, E.J. Botzolakis, X. Tang, R.L. Macdonald, A conserved Cys-loop receptor aspartate residue in the M3-M4 cytoplasmic loop is required for GABAA receptor assembly, J Biol Chem 283 (2008) 29740-29752. [41] X.J. Di, D.Y. Han, Y.J. Wang, M.R. Chance, T.W. Mu, SAHA enhances Proteostasis of epilepsy- associated alpha1(A322D)beta2gamma2 GABA(A) receptors, Chem Biol 20 (2013) 1456- 1468. [42] D.Y. Han, B.J. Guan, Y.J. Wang, M. Hatzoglou, T.W. Mu, L-type Calcium Channel Blockers Enhance Trafficking and Function of Epilepsy-associated alpha1(D219N) Subunits of GABA(A) Receptors, ACS Chem Biol 10 (2015) 2135-2148. [43] C.P. Wanamaker, W.N. Green, Endoplasmic reticulum chaperones stabilize nicotinic receptor subunits and regulate receptor assembly, J Biol Chem 282 (2007) 31113-31123. [44] N.S. Millar, RIC-3: a nicotinic acetylcholine receptor chaperone, Br J Pharmacol 153 Suppl 1 (2008) S177-183. [45] M. Castillo, J. Mulet, L.M. Gutierrez, J.A. Ortiz, F. Castelan, S. Gerber, S. Sala, F. Sala, M. Criado, Role of the RIC-3 protein in trafficking of serotonin and nicotinic acetylcholine receptors, J Mol Neurosci 30 (2006) 153-156. [46] M. Castillo, J. Mulet, L.M. Gutierrez, J.A. Ortiz, F. Castelan, S. Gerber, S. Sala, F. Sala, M. Criado, Dual role of the RIC-3 protein in trafficking of serotonin and nicotinic acetylcholine receptors, J Biol Chem 280 (2005) 27062-27068. [47] J.A. Olzmann, R.R. Kopito, J.C. Christianson, The mammalian endoplasmic reticulum- associated degradation system, Cold Spring Harb Perspect Biol 5 (2013) a013185. [48] M.H. Smith, H.L. Ploegh, J.S. Weissman, Road to Ruin: Targeting Proteins for Degradation in the Endoplasmic Reticulum, Science 334 (2011) 1086-1090. [49] S.S. Vembar, J.L. Brodsky, One step at a time: endoplasmic reticulum-associated degradation, Nature Reviews Molecular Cell Biology 9 (2008) 944-957. [50] W.X. Ding, X.M. Yin, Sorting, recognition and activation of the misfolded protein degradation pathways through macroautophagy and the proteasome, Autophagy 4 (2008) 141-150. [51] C. Fang, L. Deng, C.A. Keller, M. Fukata, Y. Fukata, G. Chen, B. Luscher, GODZ-mediated palmitoylation of GABA(A) receptors is required for normal assembly and function of GABAergic inhibitory synapses, J Neurosci 26 (2006) 12758-12768. [52] C.A. Keller, X. Yuan, P. Panzanelli, M.L. Martin, M. Alldred, M. Sassoe-Pognetto, B. Luscher, The gamma2 subunit of GABA(A) receptors is a substrate for palmitoylation by GODZ, J Neurosci 24 (2004) 5881-5891. [53] H.W. Shin, N. Morinaga, M. Noda, K. Nakayama, BIG2, a guanine nucleotide exchange factor for ADP-ribosylation factors: its localization to recycling endosomes and implication in the endosome integrity, Mol Biol Cell 15 (2004) 5283-5294. [54] H. Wang, F.K. Bedford, N.J. Brandon, S.J. Moss, R.W. Olsen, GABA(A)-receptor-associated protein links GABA(A) receptors and the cytoskeleton, Nature 397 (1999) 69-72.

129

[55] J. Nymann-Andersen, H. Wang, L. Chen, J.T. Kittler, S.J. Moss, R.W. Olsen, Subunit specificity and interaction domain between GABA(A) receptor-associated protein (GABARAP) and GABA(A) receptors, J Neurochem 80 (2002) 815-823. [56] Z.W. Chen, C.S. Chang, T.A. Leil, R.W. Olsen, C-terminal modification is required for GABARAP-mediated GABA(A) receptor trafficking, J Neurosci 27 (2007) 6655-6663. [57] A. Mizokami, T. Kanematsu, H. Ishibashi, T. Yamaguchi, I. Tanida, K. Takenaka, K.I. Nakayama, K. Fukami, T. Takenawa, E. Kominami, S.J. Moss, T. Yamamoto, J. Nabekura, M. Hirata, Phospholipase C-related inactive protein is involved in trafficking of gamma2 subunit-containing GABA(A) receptors to the cell surface, J Neurosci 27 (2007) 1692- 1701. [58] C.J. Zhao, C. Noack, M. Brackmann, T. Gloveli, A. Maelicke, U. Heinemann, R. Anand, K.H. Braunewell, Neuronal Ca2+ sensor VILIP-1 leads to the upregulation of functional alpha4beta2 nicotinic acetylcholine receptors in hippocampal neurons, Mol Cell Neurosci 40 (2009) 280-292. [59] S. Eimer, A. Gottschalk, M. Hengartner, H.R. Horvitz, J. Richmond, W.R. Schafer, J.L. Bessereau, Regulation of nicotinic receptor trafficking by the transmembrane Golgi protein UNC-50, EMBO J 26 (2007) 4313-4323. [60] T.C. Jacob, S.J. Moss, R. Jurd, GABA(A) receptor trafficking and its role in the dynamic modulation of neuronal inhibition, Nat Rev Neurosci 9 (2008) 331-343. [61] M. Kneussel, J.H. Brandstatter, B. Gasnier, G. Feng, J.R. Sanes, H. Betz, Gephyrin- independent clustering of postsynaptic GABA(A) receptor subtypes, Mol Cell Neurosci 17 (2001) 973-982. [62] L. Saiepour, C. Fuchs, A. Patrizi, M. Sassoe-Pognetto, R.J. Harvey, K. Harvey, Complex role of collybistin and gephyrin in GABAA receptor clustering, J Biol Chem 285 (2010) 29623- 29631. [63] T.T. Chiou, B. Bonhomme, H. Jin, C.P. Miralles, H. Xiao, Z. Fu, R.J. Harvey, K. Harvey, S. Vicini, A.L. De Blas, Differential regulation of the postsynaptic clustering of gamma- aminobutyric acid type A (GABAA) receptors by collybistin isoforms, J Biol Chem 286 (2011) 22456-22468. [64] T. Papadopoulos, T. Soykan, The role of collybistin in gephyrin clustering at inhibitory synapses: facts and open questions, Front Cell Neurosci 5 (2011) 11. [65] S. Loebrich, R. Bahring, T. Katsuno, S. Tsukita, M. Kneussel, Activated radixin is essential for GABAA receptor alpha5 subunit anchoring at the actin cytoskeleton, EMBO J 25 (2006) 987-999. [66] Y. Lee, J. Rudell, S. Yechikhov, R. Taylor, S. Swope, M. Ferns, Rapsyn carboxyl terminal domains mediate muscle specific kinase-induced phosphorylation of the muscle acetylcholine receptor, Neuroscience 153 (2008) 997-1007. [67] J. Piguet, C. Schreiter, J.M. Segura, H. Vogel, R. Hovius, Acetylcholine receptor organization in membrane domains in muscle cells: evidence for rapsyn-independent and rapsyn- dependent mechanisms, J Biol Chem 286 (2011) 363-369. [68] M.M. Rosenberg, F. Yang, M. Giovanni, J.L. Mohn, M.K. Temburni, M.H. Jacob, Adenomatous polyposis coli plays a key role, in vivo, in coordinating assembly of the neuronal nicotinic postsynaptic complex, Mol Cell Neurosci 38 (2008) 138-152. [69] C.N. Connolly, J.T. Kittler, P. Thomas, J.M. Uren, N.J. Brandon, T.G. Smart, S.J. Moss, Cell surface stability of gamma-aminobutyric acid type A receptors. Dependence on protein kinase C activity and subunit composition, J Biol Chem 274 (1999) 36565-36572. [70] J.T. Kittler, P. Delmas, J.N. Jovanovic, D.A. Brown, T.G. Smart, S.J. Moss, Constitutive endocytosis of GABAA receptors by an association with the adaptin AP2 complex

130

modulates inhibitory synaptic currents in hippocampal neurons, J Neurosci 20 (2000) 7972-7977. [71] A.E. Twelvetrees, E.Y. Yuen, I.L. Arancibia-Carcamo, A.F. MacAskill, P. Rostaing, M.J. Lumb, S. Humbert, A. Triller, F. Saudou, Z. Yan, J.T. Kittler, Delivery of GABAARs to synapses is mediated by HAP1-KIF5 and disrupted by mutant huntingtin, Neuron 65 (2010) 53-65. [72] J.T. Kittler, P. Thomas, V. Tretter, Y.D. Bogdanov, V. Haucke, T.G. Smart, S.J. Moss, Huntingtin-associated protein 1 regulates inhibitory synaptic transmission by modulating gamma-aminobutyric acid type A receptor membrane trafficking, Proc Natl Acad Sci U S A 101 (2004) 12736-12741. [73] I.L. Arancibia-Carcamo, E.Y. Yuen, J. Muir, M.J. Lumb, G. Michels, R.S. Saliba, T.G. Smart, Z. Yan, J.T. Kittler, S.J. Moss, Ubiquitin-dependent lysosomal targeting of GABA(A) receptors regulates neuronal inhibition, Proc Natl Acad Sci U S A 106 (2009) 17552- 17557. [74] W.C. Tseng, P.M. Jenkins, M. Tanaka, R. Mooney, V. Bennett, Giant ankyrin-G stabilizes somatodendritic GABAergic synapses through opposing endocytosis of GABAA receptors, Proc Natl Acad Sci U S A 112 (2015) 1214-1219. [75] Z. Iqbal, G. Vandeweyer, M. van der Voet, A.M. Waryah, M.Y. Zahoor, J.A. Besseling, L.T. Roca, A.T. Vulto-van Silfhout, B. Nijhof, J.M. Kramer, N. Van der Aa, M. Ansar, H. Peeters, C. Helsmoortel, C. Gilissen, L.E. Vissers, J.A. Veltman, A.P. de Brouwer, R. Frank Kooy, S. Riazuddin, A. Schenck, H. van Bokhoven, L. Rooms, Homozygous and heterozygous disruptions of ANK3: at the crossroads of neurodevelopmental and psychiatric disorders, Hum Mol Genet 22 (2013) 1960-1970. [76] S. Chaumont, C. Andre, D. Perrais, E. Boue-Grabot, A. Taly, M. Garret, Agonist-dependent endocytosis of gamma-aminobutyric acid type A (GABAA) receptors revealed by a gamma2(R43Q) epilepsy mutation, J Biol Chem 288 (2013) 28254-28265. [77] P.A. St John, H. Gordon, Agonists cause endocytosis of nicotinic acetylcholine receptors on cultured myotubes, J Neurobiol 49 (2001) 212-223. [78] M. Akaaboune, S.M. Culican, S.G. Turney, J.W. Lichtman, Rapid and reversible effects of activity on acetylcholine receptor density at the neuromuscular junction in vivo, Science 286 (1999) 503-507. [79] P.A. St John, Cellular trafficking of nicotinic acetylcholine receptors, Acta Pharmacol Sin 30 (2009) 656-662. [80] Z.W. Chen, R.W. Olsen, GABAA receptor associated proteins: a key factor regulating GABAA receptor function, J Neurochem 100 (2007) 279-294. [81] C.J. Baier, F.J. Barrantes, Sphingolipids are necessary for nicotinic acetylcholine receptor export in the early secretory pathway, J Neurochem 101 (2007) 1072-1084. [82] V. Borroni, C.J. Baier, T. Lang, I. Bonini, M.M. White, I. Garbus, F.J. Barrantes, Cholesterol depletion activates rapid internalization of submicron-sized acetylcholine receptor domains at the cell membrane, Mol Membr Biol 24 (2007) 1-15. [83] S.L. Swope, S.J. Moss, L.A. Raymond, R.L. Huganir, Regulation of ligand-gated ion channels by protein phosphorylation, Adv Second Messenger Phosphoprotein Res 33 (1999) 49- 78. [84] H. Bannai, S. Levi, C. Schweizer, T. Inoue, T. Launey, V. Racine, J.B. Sibarita, K. Mikoshiba, A. Triller, Activity-dependent tuning of inhibitory neurotransmission based on GABAAR diffusion dynamics, Neuron 62 (2009) 670-682. [85] R. Eckel, B. Szulc, M.C. Walker, J.T. Kittler, Activation of calcineurin underlies altered trafficking of alpha2 subunit containing GABAA receptors during prolonged epileptiform activity, Neuropharmacology 88 (2015) 82-90.

131

[86] J.T. Kittler, S.J. Moss, Modulation of GABAA receptor activity by phosphorylation and receptor trafficking: implications for the efficacy of synaptic inhibition, Curr Opin Neurobiol 13 (2003) 341-347. [87] M. Terunuma, I.S. Jang, S.H. Ha, J.T. Kittler, T. Kanematsu, J.N. Jovanovic, K.I. Nakayama, N. Akaike, S.H. Ryu, S.J. Moss, M. Hirata, GABAA receptor phospho-dependent modulation is regulated by phospholipase C-related inactive protein type 1, a novel protein phosphatase 1 anchoring protein, J Neurosci 24 (2004) 7074-7084. [88] K. Yoshimura, H. Takeuchi, O. Sato, K. Hidaka, N. Doira, M. Terunuma, K. Harada, Y. Ogawa, Y. Ito, T. Kanematsu, M. Hirata, Interaction of p130 with, and consequent inhibition of, the catalytic subunit of protein phosphatase 1alpha, J Biol Chem 276 (2001) 17908- 17913. [89] A.M. Abramian, E. Comenencia-Ortiz, M. Vithlani, E.V. Tretter, W. Sieghart, P.A. Davies, S.J. Moss, Protein kinase C phosphorylation regulates membrane insertion of GABAA receptor subtypes that mediate tonic inhibition, J Biol Chem 285 (2010) 41795-41805. [90] A.M. Abramian, E. Comenencia-Ortiz, A. Modgil, T.N. Vien, Y. Nakamura, Y.E. Moore, J.L. Maguire, M. Terunuma, P.A. Davies, S.J. Moss, Neurosteroids promote phosphorylation and membrane insertion of extrasynaptic GABAA receptors, Proc Natl Acad Sci U S A 111 (2014) 7132-7137. [91] E. Comenencia-Ortiz, S.J. Moss, P.A. Davies, Phosphorylation of GABAA receptors influences receptor trafficking and neurosteroid actions, Psychopharmacology (Berl) 231 (2014) 3453-3465. [92] J.N. Jovanovic, P. Thomas, J.T. Kittler, T.G. Smart, S.J. Moss, Brain-derived neurotrophic factor modulates fast synaptic inhibition by regulating GABA(A) receptor phosphorylation, activity, and cell-surface stability, J Neurosci 24 (2004) 522-530. [93] C.H. Cho, W. Song, K. Leitzell, E. Teo, A.D. Meleth, M.W. Quick, R.A. Lester, Rapid upregulation of alpha7 nicotinic acetylcholine receptors by tyrosine dephosphorylation, J Neurosci 25 (2005) 3712-3723. [94] E.G. Bruneau, M. Akaaboune, The dynamics of recycled acetylcholine receptors at the neuromuscular junction in vivo, Development 133 (2006) 4485-4493. [95] Z.C. Qu, E. Moritz, R.L. Huganir, Regulation of tyrosine phosphorylation of the nicotinic acetylcholine receptor at the rat neuromuscular junction, Neuron 4 (1990) 367-378. [96] A. Sava, I. Barisone, D. Di Mauro, G. Fumagalli, C. Sala, Modulation of nicotinic acetylcholine receptor turnover by tyrosine phosphorylation in rat myotubes, Neurosci Lett 313 (2001) 37-40. [97] R. Xu, M.M. Salpeter, Protein kinase A regulates the degradation rate of Rs acetylcholine receptors, J Cell Physiol 165 (1995) 30-39. [98] H. Sun, X.Q. Hu, E.M. Moradel, F.F. Weight, L. Zhang, Modulation of 5-HT3 receptor- mediated response and trafficking by activation of protein kinase C, J Biol Chem 278 (2003) 34150-34157. [99] R.L. Macdonald, J.Q. Kang, M.J. Gallagher, GABAA Receptor Subunit Mutations and Genetic Epilepsies, in: th, J.L. Noebels, M. Avoli, M.A. Rogawski, R.W. Olsen, A.V. Delgado- Escueta (Eds.), Jasper's Basic Mechanisms of the Epilepsies, Bethesda (MD), 2012. [100] G. Frugier, F. Coussen, M.F. Giraud, M.F. Odessa, M.B. Emerit, E. Boue-Grabot, M. Garret, A gamma 2(R43Q) mutation, linked to epilepsy in humans, alters GABAA receptor assembly and modifies subunit composition on the cell surface, J Biol Chem 282 (2007) 3819-3828.

132

[101] M.J. Gallagher, L. Ding, A. Maheshwari, R.L. Macdonald, The GABA(A) receptor alpha 1 subunit epilepsy mutation A322D inhibits transmembrane helix formation and causes proteasomal degradation, Proc. Natl. Acad. Sci. USA 104 (2007) 12999-13004. [102] D.Y. Han, B.J. Guan, Y.J. Wang, M. Hatzoglou, T.W. Mu, L-type Calcium Channel Blockers Enhance Trafficking and Function of Epilepsy-associated alpha1(D219N) Subunits of GABA Receptors, ACS Chem Biol (2015). [103] E. Todd, K.N. Gurba, E.J. Botzolakis, A.K. Stanic, R.L. Macdonald, GABAA receptor biogenesis is impaired by the gamma2 subunit febrile seizure-associated mutation, GABRG2(R177G), Neurobiol Dis 69 (2014) 215-224. [104] C.I. Richards, R. Srinivasan, C. Xiao, E.D. Mackey, J.M. Miwa, H.A. Lester, Trafficking of alpha4* nicotinic receptors revealed by superecliptic phluorin: effects of a beta4 amyotrophic lateral sclerosis-associated mutation and chronic exposure to nicotine, J Biol Chem 286 (2011) 31241-31249. [105] A.G. Engel, K. Ohno, S.M. Sine, Congenital myasthenic syndromes: recent advances, Arch Neurol 56 (1999) 163-167. [106] P. Lachance-Touchette, M. Choudhury, A. Stoica, G. Di Cristo, P. Cossette, Single-cell genetic expression of mutant GABAA receptors causing Human genetic epilepsy alters dendritic spine and GABAergic bouton formation in a mutation-specific manner, Front Cell Neurosci 8 (2014) 317. [107] M. Schelhaas, J. Malmstrom, L. Pelkmans, J. Haugstetter, L. Ellgaard, K. Grunewald, A. Helenius, Simian Virus 40 depends on ER protein folding and quality control factors for entry into host cells, Cell 131 (2007) 516-529. [108] A. Shiber, K. Doring, U. Friedrich, K. Klann, D. Merker, M. Zedan, F. Tippmann, G. Kramer, B. Bukau, Cotranslational assembly of protein complexes in eukaryotes revealed by ribosome profiling, Nature 561 (2018) 268-272. [109] L. Ellgaard, E.M. Frickel, Calnexin, calreticulin, and ERp57: teammates in glycoprotein folding, Cell Biochem Biophys 39 (2003) 223-247. [110] Y.J. Wang, D.Y. Han, T. Tabib, J.R. Yates, 3rd, T.W. Mu, Identification of GABA(C) receptor protein homeostasis network components from three tandem mass spectrometry proteomics approaches, J Proteome Res 12 (2013) 5570-5586. [111] P. Lachance-Touchette, P. Brown, C. Meloche, P. Kinirons, L. Lapointe, H. Lacasse, A. Lortie, L. Carmant, F. Bedford, D. Bowie, P. Cossette, Novel alpha1 and gamma2 GABAA receptor subunit mutations in families with idiopathic generalized epilepsy, Eur J Neurosci 34 (2011) 237-249. [112] R.L. Wiseman, E.T. Powers, J.N. Buxbaum, J.W. Kelly, W.E. Balch, An adaptable standard for protein export from the endoplasmic reticulum, Cell 131 (2007) 809-821. [113] D.Y. Han, X.J. Di, Y.L. Fu, T.W. Mu, Combining valosin-containing protein (VCP) inhibition and suberanilohydroxamic acid (SAHA) treatment additively enhances the folding, trafficking, and function of epilepsy-associated gamma-aminobutyric acid, type A (GABAA) receptors, J Biol Chem 290 (2015) 325-337. [114] T.W. Mu, D.S. Ong, Y.J. Wang, W.E. Balch, J.R. Yates, 3rd, L. Segatori, J.W. Kelly, Chemical and biological approaches synergize to ameliorate protein-folding diseases, Cell 134 (2008) 769-781. [115] Y.J. Wang, X.J. Di, T.W. Mu, Using pharmacological chaperones to restore proteostasis, Pharmacol Res 83 (2014) 3-9. [116] D.M. Hutt, D. Herman, A.P. Rodrigues, S. Noel, J.M. Pilewski, J. Matteson, B. Hoch, W. Kellner, J.W. Kelly, A. Schmidt, P.J. Thomas, Y. Matsumura, W.R. Skach, M. Gentzsch, J.R. Riordan, E.J. Sorscher, T. Okiyoneda, J.R. Yates, 3rd, G.L. Lukacs, R.A. Frizzell, G.

133

Manning, J.M. Gottesfeld, W.E. Balch, Reduced histone deacetylase 7 activity restores function to misfolded CFTR in cystic fibrosis, Nat Chem Biol 6 (2010) 25-33. [117] M.Y. Chen, W.S. Liao, Z. Lu, W.G. Bornmann, V. Hennessey, M.N. Washington, G.L. Rosner, Y. Yu, A.A. Ahmed, R.C. Bast, Jr., Decitabine and suberoylanilide hydroxamic acid (SAHA) inhibit growth of ovarian cancer cell lines and xenografts while inducing expression of imprinted tumor suppressor genes, apoptosis, G2/M arrest, and autophagy, Cancer 117 (2011) 4424-4438. [118] M. Bouchecareilh, D.M. Hutt, P. Szajner, T.R. Flotte, W.E. Balch, Histone deacetylase inhibitor (HDACi) suberoylanilide hydroxamic acid (SAHA)-mediated correction of alpha1-antitrypsin deficiency, J Biol Chem 287 (2012) 38265-38278. [119] K.P. Zhang, B.F. Yang, B.X. Li, Translational toxicology and rescue strategies of the hERG channel dysfunction: biochemical and molecular mechanistic aspects, Acta Pharmacol Sin 35 (2014) 1473-1484. [120] V.E. Walker, R. Atanasiu, H. Lam, A. Shrier, Co-chaperone FKBP38 promotes HERG trafficking, J Biol Chem 282 (2007) 23509-23516. [121] F. Ran, N. Gadura, C.A. Michels, Hsp90 cochaperone Aha1 is a negative regulator of the Saccharomyces MAL activator and acts early in the chaperone activation pathway, J Biol Chem 285 (2010) 13850-13862. [122] S. Nishikawa, J.L. Brodsky, K. Nakatsukasa, Roles of molecular chaperones in endoplasmic reticulum (ER) quality control and ER-associated degradation (ERAD), J Biochem 137 (2005) 551-555. [123] B.H. Qu, E.H. Strickland, P.J. Thomas, Localization and suppression of a kinetic defect in cystic fibrosis transmembrane conductance regulator folding, J Biol Chem 272 (1997) 15739-15744. [124] H.Y. Ren, D.E. Grove, O. De La Rosa, S.A. Houck, P. Sopha, F. Van Goor, B.J. Hoffman, D.M. Cyr, VX-809 corrects folding defects in cystic fibrosis transmembrane conductance regulator protein through action on membrane-spanning domain 1, Mol Biol Cell 24 (2013) 3016-3024. [125] F. Holguin, Triple CFTR Modulator Therapy for Cystic Fibrosis, N Engl J Med 379 (2018) 1671-1672. [126] R.S. Pettit, C. Fellner, CFTR Modulators for the Treatment of Cystic Fibrosis, P T 39 (2014) 500-511. [127] Q. Gong, M.A. Jones, Z. Zhou, Mechanisms of pharmacological rescue of trafficking- defective hERG mutant channels in human long QT syndrome, J Biol Chem 281 (2006) 4069-4074. [128] M. Calfon, H. Zeng, F. Urano, J.H. Till, S.R. Hubbard, H.P. Harding, S.G. Clark, D. Ron, IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNA, Nature 415 (2002) 92-96. [129] K. Lee, W. Tirasophon, X. Shen, M. Michalak, R. Prywes, T. Okada, H. Yoshida, K. Mori, R.J. Kaufman, IRE1-mediated unconventional mRNA splicing and S2P-mediated ATF6 cleavage merge to regulate XBP1 in signaling the unfolded protein response, Genes Dev 16 (2002) 452-466. [130] A. Uemura, M. Oku, K. Mori, H. Yoshida, Unconventional splicing of XBP1 mRNA occurs in the cytoplasm during the mammalian unfolded protein response, J Cell Sci 122 (2009) 2877-2886. [131] Y. Imagawa, A. Hosoda, S. Sasaka, A. Tsuru, K. Kohno, RNase domains determine the functional difference between IRE1alpha and IRE1beta, FEBS Lett 582 (2008) 656-660.

134

[132] K. Yamamoto, H. Yoshida, K. Kokame, R.J. Kaufman, K. Mori, Differential contributions of ATF6 and XBP1 to the activation of endoplasmic reticulum stress-responsive cis-acting elements ERSE, UPRE and ERSE-II, J Biochem 136 (2004) 343-350. [133] D.S. Coelho, P.M. Domingos, Physiological roles of regulated Ire1 dependent decay, Front Genet 5 (2014) 76. [134] J.P. Upton, L. Wang, D. Han, E.S. Wang, N.E. Huskey, L. Lim, M. Truitt, M.T. McManus, D. Ruggero, A. Goga, F.R. Papa, S.A. Oakes, IRE1alpha cleaves select microRNAs during ER stress to derepress translation of proapoptotic Caspase-2, Science 338 (2012) 818-822. [135] M. Maurel, E. Chevet, J. Tavernier, S. Gerlo, Getting RIDD of RNA: IRE1 in cell fate regulation, Trends Biochem Sci 39 (2014) 245-254. [136] P. Walter, D. Ron, The unfolded protein response: from stress pathway to homeostatic regulation, Science 334 (2011) 1081-1086. [137] H. Yoshida, K. Haze, H. Yanagi, T. Yura, K. Mori, Identification of the cis-acting endoplasmic reticulum stress response element responsible for transcriptional induction of mammalian glucose-regulated proteins. Involvement of basic leucine zipper transcription factors, J Biol Chem 273 (1998) 33741-33749. [138] K. Haze, H. Yoshida, H. Yanagi, T. Yura, K. Mori, Mammalian transcription factor ATF6 is synthesized as a transmembrane protein and activated by proteolysis in response to endoplasmic reticulum stress, Mol Biol Cell 10 (1999) 3787-3799. [139] S. Nadanaka, T. Okada, H. Yoshida, K. Mori, Role of disulfide bridges formed in the luminal domain of ATF6 in sensing endoplasmic reticulum stress, Mol Cell Biol 27 (2007) 1027- 1043. [140] K. Yamamoto, T. Sato, T. Matsui, M. Sato, T. Okada, H. Yoshida, A. Harada, K. Mori, Transcriptional induction of mammalian ER quality control proteins is mediated by single or combined action of ATF6alpha and XBP1, Dev Cell 13 (2007) 365-376. [141] Y. Wang, J. Shen, N. Arenzana, W. Tirasophon, R.J. Kaufman, R. Prywes, Activation of ATF6 and an ATF6 DNA binding site by the endoplasmic reticulum stress response, J Biol Chem 275 (2000) 27013-27020. [142] R.F. Hillary, U. FitzGerald, A lifetime of stress: ATF6 in development and homeostasis, J Biomed Sci 25 (2018) 48. [143] M.D. Shoulders, L.M. Ryno, J.C. Genereux, J.J. Moresco, P.G. Tu, C. Wu, J.R. Yates, 3rd, A.I. Su, J.W. Kelly, R.L. Wiseman, Stress-independent activation of XBP1s and/or ATF6 reveals three functionally diverse ER proteostasis environments, Cell Rep 3 (2013) 1279- 1292. [144] C. Hetz, S. Saxena, ER stress and the unfolded protein response in neurodegeneration, Nat Rev Neurol 13 (2017) 477-491. [145] G.K. Gouras, J. Tsai, J. Naslund, B. Vincent, M. Edgar, F. Checler, J.P. Greenfield, V. Haroutunian, J.D. Buxbaum, H. Xu, P. Greengard, N.R. Relkin, Intraneuronal Abeta42 accumulation in human brain, Am J Pathol 156 (2000) 15-20. [146] P. Fernandez-Vizarra, A.P. Fernandez, S. Castro-Blanco, J. Serrano, M.L. Bentura, R. Martinez-Murillo, A. Martinez, J. Rodrigo, Intra- and extracellular Abeta and PHF in clinically evaluated cases of Alzheimer's disease, Histol Histopathol 19 (2004) 823-844. [147] S. Meier, M. Bell, D.N. Lyons, A. Ingram, J. Chen, J.C. Gensel, H. Zhu, P.T. Nelson, J.F. Abisambra, Identification of Novel Tau Interactions with Endoplasmic Reticulum Proteins in Alzheimer's Disease Brain, J Alzheimers Dis 48 (2015) 687-702. [148] J.F. Abisambra, U.K. Jinwal, L.J. Blair, J.C. O'Leary, 3rd, Q. Li, S. Brady, L. Wang, C.E. Guidi, B. Zhang, B.A. Nordhues, M. Cockman, A. Suntharalingham, P. Li, Y. Jin, C.A. Atkins, C.A.

135

Dickey, Tau accumulation activates the unfolded protein response by impairing endoplasmic reticulum-associated degradation, J Neurosci 33 (2013) 9498-9507. [149] A. Bellucci, L. Navarria, M. Zaltieri, E. Falarti, S. Bodei, S. Sigala, L. Battistin, M. Spillantini, C. Missale, P. Spano, Induction of the unfolded protein response by alpha-synuclein in experimental models of Parkinson's disease, J Neurochem 116 (2011) 588-605. [150] R. Takahashi, Y. Imai, N. Hattori, Y. Mizuno, Parkin and endoplasmic reticulum stress, Ann N Y Acad Sci 991 (2003) 101-106. [151] Y. Imai, M. Soda, H. Inoue, N. Hattori, Y. Mizuno, R. Takahashi, An unfolded putative transmembrane polypeptide, which can lead to endoplasmic reticulum stress, is a substrate of Parkin, Cell 105 (2001) 891-902. [152] H. Kikuchi, G. Almer, S. Yamashita, C. Guegan, M. Nagai, Z. Xu, A.A. Sosunov, G.M. McKhann, 2nd, S. Przedborski, Spinal cord endoplasmic reticulum stress associated with a microsomal accumulation of mutant superoxide dismutase-1 in an ALS model, Proc Natl Acad Sci U S A 103 (2006) 6025-6030. [153] H. Nishitoh, H. Kadowaki, A. Nagai, T. Maruyama, T. Yokota, H. Fukutomi, T. Noguchi, A. Matsuzawa, K. Takeda, H. Ichijo, ALS-linked mutant SOD1 induces ER stress- and ASK1- dependent motor neuron death by targeting Derlin-1, Genes Dev 22 (2008) 1451-1464. [154] M. Urushitani, S.A. Ezzi, A. Matsuo, I. Tooyama, J.P. Julien, The endoplasmic reticulum- Golgi pathway is a target for translocation and aggregation of mutant superoxide dismutase linked to ALS, FASEB J 22 (2008) 2476-2487. [155] J.D. Atkin, M.A. Farg, B.J. Turner, D. Tomas, J.A. Lysaght, J. Nunan, A. Rembach, P. Nagley, P.M. Beart, S.S. Cheema, M.K. Horne, Induction of the unfolded protein response in familial amyotrophic lateral sclerosis and association of protein-disulfide isomerase with superoxide dismutase 1, J Biol Chem 281 (2006) 30152-30165. [156] A.A. Cooper, A.D. Gitler, A. Cashikar, C.M. Haynes, K.J. Hill, B. Bhullar, K. Liu, K. Xu, K.E. Strathearn, F. Liu, S. Cao, K.A. Caldwell, G.A. Caldwell, G. Marsischky, R.D. Kolodner, J. Labaer, J.C. Rochet, N.M. Bonini, S. Lindquist, Alpha-synuclein blocks ER-Golgi traffic and Rab1 rescues neuron loss in Parkinson's models, Science 313 (2006) 324-328. [157] J.J. Credle, P.A. Forcelli, M. Delannoy, A.W. Oaks, E. Permaul, D.L. Berry, V. Duka, J. Wills, A. Sidhu, alpha-Synuclein-mediated inhibition of ATF6 processing into COPII vesicles disrupts UPR signaling in Parkinson's disease, Neurobiol Dis 76 (2015) 112-125. [158] R.J. Kaufman, Orchestrating the unfolded protein response in health and disease, J Clin Invest 110 (2002) 1389-1398. [159] W.C. Chiang, N. Hiramatsu, C. Messah, H. Kroeger, J.H. Lin, Selective activation of ATF6 and PERK endoplasmic reticulum stress signaling pathways prevent mutant rhodopsin accumulation, Invest Ophthalmol Vis Sci 53 (2012) 7159-7166. [160] C. Hetz, P. Thielen, S. Matus, M. Nassif, F. Court, R. Kiffin, G. Martinez, A.M. Cuervo, R.H. Brown, L.H. Glimcher, XBP-1 deficiency in the nervous system protects against amyotrophic lateral sclerosis by increasing autophagy, Genes Dev 23 (2009) 2294-2306. [161] R.L. Vidal, A. Figueroa, F.A. Court, P. Thielen, C. Molina, C. Wirth, B. Caballero, R. Kiffin, J. Segura-Aguilar, A.M. Cuervo, L.H. Glimcher, C. Hetz, Targeting the UPR transcription factor XBP1 protects against Huntington's disease through the regulation of FoxO1 and autophagy, Hum Mol Genet 21 (2012) 2245-2262. [162] N. Egawa, K. Yamamoto, H. Inoue, R. Hikawa, K. Nishi, K. Mori, R. Takahashi, The endoplasmic reticulum stress sensor, ATF6alpha, protects against neurotoxin-induced dopaminergic neuronal death, J Biol Chem 286 (2011) 7947-7957.

136

[163] K. Hashida, Y. Kitao, H. Sudo, Y. Awa, S. Maeda, K. Mori, R. Takahashi, M. Iinuma, O. Hori, ATF6alpha promotes astroglial activation and neuronal survival in a chronic mouse model of Parkinson's disease, PLoS One 7 (2012) e47950. [164] J.R. Naranjo, H. Zhang, D. Villar, P. Gonzalez, X.M. Dopazo, J. Moron-Oset, E. Higueras, J.C. Oliveros, M.D. Arrabal, A. Prieto, P. Cercos, T. Gonzalez, A. De la Cruz, J. Casado-Vela, A. Rabano, C. Valenzuela, M. Gutierrez-Rodriguez, J.Y. Li, B. Mellstrom, Activating transcription factor 6 derepression mediates neuroprotection in Huntington disease, J Clin Invest 126 (2016) 627-638. [165] M.R. Fernandez-Fernandez, I. Ferrer, J.J. Lucas, Impaired ATF6alpha processing, decreased Rheb and neuronal cell cycle re-entry in Huntington's disease, Neurobiol Dis 41 (2011) 23-32. [166] J.J. Chen, J.C. Genereux, S. Qu, J.D. Hulleman, M.D. Shoulders, R.L. Wiseman, ATF6 activation reduces the secretion and extracellular aggregation of destabilized variants of an amyloidogenic protein, Chem Biol 21 (2014) 1564-1574. [167] C.B. Cooley, L.M. Ryno, L. Plate, G.J. Morgan, J.D. Hulleman, J.W. Kelly, R.L. Wiseman, Unfolded protein response activation reduces secretion and extracellular aggregation of amyloidogenic immunoglobulin light chain, Proc Natl Acad Sci U S A 111 (2014) 13046- 13051. [168] L. Plate, C.B. Cooley, J.J. Chen, R.J. Paxman, C.M. Gallagher, F. Madoux, J.C. Genereux, W. Dobbs, D. Garza, T.P. Spicer, L. Scampavia, S.J. Brown, H. Rosen, E.T. Powers, P. Walter, P. Hodder, R.L. Wiseman, J.W. Kelly, Small molecule proteostasis regulators that reprogram the ER to reduce extracellular protein aggregation, Elife 5 (2016) e15550. [169] L. Ulbrich, F.L. Favaloro, L. Trobiani, V. Marchetti, V. Patel, T. Pascucci, D. Comoletti, S.J. Marciniak, A. De Jaco, Autism-associated R451C mutation in neuroligin3 leads to activation of the unfolded protein response in a PC12 Tet-On inducible system, Biochem J 473 (2016) 423-434. [170] T.W. Mu, D.S.T. Ong, Y.J. Wang, W.E. Balch, J.R. Yates, L. Segatori, J.W. Kelly, Chemical and biological approaches synergize to ameliorate protein-folding diseases, Cell 134 (2008) 769-781. [171] M.J. Gallagher, W.Z. Shen, L.Y. Song, R.L. Macdonald, Endoplasmic reticulum retention and associated degradation of a GABA(A) receptor epilepsy mutation that inserts an aspartate in the M3 transmembrane segment of the alpha 1 subunit, J. Biol. Chem. 280 (2005) 37995-38004. [172] P.D. Lu, C. Jousse, S.J. Marciniak, Y. Zhang, I. Novoa, D. Scheuner, R.J. Kaufman, D. Ron, H.P. Harding, Cytoprotection by pre-emptive conditional phosphorylation of translation initiation factor 2, EMBO J 23 (2004) 169-179. [173] P. Lachance-Touchette, P. Brown, C. Meloche, P. Kinirons, L. Lapointe, H. Lacasse, A. Lortie, L. Carmant, F. Bedford, D. Bowie, P. Cossette, Novel alpha 1 and gamma 2 GABA(A) receptor subunit mutations in families with idiopathic generalized epilepsy, Eur. J. Neurosci. 34 (2011) 237-249. [174] M.J. Gallagher, L. Ding, A. Maheshwari, R.L. Macdonald, The GABAA receptor alpha1 subunit epilepsy mutation A322D inhibits transmembrane helix formation and causes proteasomal degradation, Proc Natl Acad Sci U S A 104 (2007) 12999-13004. [175] R. Bartoszewski, A. Rab, A. Jurkuvenaite, M. Mazur, J. Wakefield, J.F. Collawn, Z. Bebok, Activation of the unfolded protein response by deltaF508 CFTR, Am J Respir Cell Mol Biol 39 (2008) 448-457.

137

[176] S. Nanua, U. Sajjan, S. Keshavjee, M.B. Hershenson, Absence of typical unfolded protein response in primary cultured cystic fibrosis airway epithelial cells, Biochem Biophys Res Commun 343 (2006) 135-143. [177] K. Hybiske, Z. Fu, C. Schwarzer, J. Tseng, J. Do, N. Huang, T.E. Machen, Effects of cystic fibrosis transmembrane conductance regulator and DeltaF508CFTR on inflammatory response, ER stress, and Ca2+ of airway epithelia, Am J Physiol Lung Cell Mol Physiol 293 (2007) L1250-1260. [178] C. Casas, GRP78 at the Centre of the Stage in Cancer and Neuroprotection, Front Neurosci 11 (2017) 177. [179] I.G. Haas, BiP (GRP78), an essential hsp70 resident protein in the endoplasmic reticulum, Experientia 50 (1994) 1012-1020. [180] K. Petrova, S. Oyadomari, L.M. Hendershot, D. Ron, Regulated association of misfolded endoplasmic reticulum lumenal proteins with P58/DNAJc3, EMBO J 27 (2008) 2862- 2872. [181] N.N. Alder, Y. Shen, J.L. Brodsky, L.M. Hendershot, A.E. Johnson, The molecular mechanisms underlying BiP-mediated gating of the Sec61 translocon of the endoplasmic reticulum, J Cell Biol 168 (2005) 389-399. [182] B.D. Hamman, L.M. Hendershot, A.E. Johnson, BiP maintains the permeability barrier of the ER membrane by sealing the lumenal end of the translocon pore before and early in translocation, Cell 92 (1998) 747-758. [183] N.G. Haigh, A.E. Johnson, A new role for BiP: closing the aqueous translocon pore during protein integration into the ER membrane, J Cell Biol 156 (2002) 261-270. [184] A. Bertolotti, Y. Zhang, L.M. Hendershot, H.P. Harding, D. Ron, Dynamic interaction of BiP and ER stress transducers in the unfolded-protein response, Nat Cell Biol 2 (2000) 326- 332. [185] J. Shen, X. Chen, L. Hendershot, R. Prywes, ER stress regulation of ATF6 localization by dissociation of BiP/GRP78 binding and unmasking of Golgi localization signals, Dev Cell 3 (2002) 99-111. [186] K. Okamura, Y. Kimata, H. Higashio, A. Tsuru, K. Kohno, Dissociation of Kar2p/BiP from an ER sensory molecule, Ire1p, triggers the unfolded protein response in yeast, Biochem Biophys Res Commun 279 (2000) 445-450. [187] Y. Kimata, D. Oikawa, Y. Shimizu, Y. Ishiwata-Kimata, K. Kohno, A role for BiP as an adjustor for the endoplasmic reticulum stress-sensing protein Ire1, J Cell Biol 167 (2004) 445-456. [188] B.M. Gardner, D. Pincus, K. Gotthardt, C.M. Gallagher, P. Walter, Endoplasmic reticulum stress sensing in the unfolded protein response, Cold Spring Harb Perspect Biol 5 (2013) a013169. [189] J.A. Morris, A.J. Dorner, C.A. Edwards, L.M. Hendershot, R.J. Kaufman, Immunoglobulin binding protein (BiP) function is required to protect cells from endoplasmic reticulum stress but is not required for the secretion of selective proteins, J Biol Chem 272 (1997) 4327-4334. [190] Z. Yu, H. Luo, W. Fu, M.P. Mattson, The endoplasmic reticulum stress-responsive protein GRP78 protects neurons against excitotoxicity and apoptosis: suppression of oxidative stress and stabilization of calcium homeostasis, Exp Neurol 155 (1999) 302-314. [191] M. Jeon, H. Choi, S.I. Lee, J.S. Kim, M. Park, K. Kim, S. Lee, S.J. Byun, GRP78 is required for cell proliferation and protection from apoptosis in chicken embryo fibroblast cells, Poult Sci 95 (2016) 1129-1136.

138

[192] D.R. Laybutt, A.M. Preston, M.C. Akerfeldt, J.G. Kench, A.K. Busch, A.V. Biankin, T.J. Biden, Endoplasmic reticulum stress contributes to beta cell apoptosis in type 2 diabetes, Diabetologia 50 (2007) 752-763. [193] K.L. Cook, D.R. Soto-Pantoja, P.A. Clarke, M.I. Cruz, A. Zwart, A. Warri, L. Hilakivi-Clarke, D.D. Roberts, R. Clarke, Endoplasmic Reticulum Stress Protein GRP78 Modulates Lipid Metabolism to Control Drug Sensitivity and Antitumor Immunity in Breast Cancer, Cancer Res 76 (2016) 5657-5670. [194] M. Ni, H. Zhou, S. Wey, P. Baumeister, A.S. Lee, Regulation of PERK signaling and leukemic cell survival by a novel cytosolic isoform of the UPR regulator GRP78/BiP, PLoS One 4 (2009) e6868. [195] M.S. Gorbatyuk, O.S. Gorbatyuk, The Molecular Chaperone GRP78/BiP as a Therapeutic Target for Neurodegenerative Disorders: A Mini Review, J Genet Syndr Gene Ther 4 (2013) 128. [196] K.F.R. Pobre, G.J. Poet, L.M. Hendershot, The endoplasmic reticulum (ER) chaperone BiP is a master regulator of ER functions: Getting by with a little help from ERdj friends, J Biol Chem 294 (2019) 2098-2108. [197] M.K. Brown, N. Naidoo, The endoplasmic reticulum stress response in aging and age- related diseases, Front Physiol 3 (2012) 263. [198] T. Katayama, K. Imaizumi, N. Sato, K. Miyoshi, T. Kudo, J. Hitomi, T. Morihara, T. Yoneda, F. Gomi, Y. Mori, Y. Nakano, J. Takeda, T. Tsuda, Y. Itoyama, O. Murayama, A. Takashima, P. St George-Hyslop, M. Takeda, M. Tohyama, Presenilin-1 mutations downregulate the signalling pathway of the unfolded-protein response, Nat Cell Biol 1 (1999) 479-485. [199] M.S. Gorbatyuk, A. Shabashvili, W. Chen, C. Meyers, L.F. Sullivan, M. Salganik, J.H. Lin, A.S. Lewin, N. Muzyczka, O.S. Gorbatyuk, Glucose regulated protein 78 diminishes alpha- synuclein neurotoxicity in a rat model of Parkinson disease, Mol Ther 20 (2012) 1327- 1337. [200] M.S. Gorbatyuk, T. Knox, M.M. LaVail, O.S. Gorbatyuk, S.M. Noorwez, W.W. Hauswirth, J.H. Lin, N. Muzyczka, A.S. Lewin, Restoration of visual function in P23H rhodopsin transgenic rats by gene delivery of BiP/Grp78, Proc Natl Acad Sci U S A 107 (2010) 5961- 5966. [201] T. Kudo, M. Okumura, K. Imaizumi, W. Araki, T. Morihara, H. Tanimukai, E. Kamagata, N. Tabuchi, R. Kimura, D. Kanayama, A. Fukumori, S. Tagami, M. Okochi, M. Kubo, H. Tanii, M. Tohyama, T. Tabira, M. Takeda, Altered localization of amyloid precursor protein under endoplasmic reticulum stress, Biochem Biophys Res Commun 344 (2006) 525-530. [202] X.Y. Zhang, T.T. Zhang, D.D. Song, J. Zhou, R. Han, Z.H. Qin, R. Sheng, Endoplasmic reticulum chaperone GRP78 is involved in autophagy activation induced by ischemic preconditioning in neural cells, Mol Brain 8 (2015) 20. [203] Y.B. Ouyang, L.J. Xu, J.F. Emery, A.S. Lee, R.G. Giffard, Overexpressing GRP78 influences Ca2+ handling and function of mitochondria in astrocytes after ischemia-like stress, Mitochondrion 11 (2011) 279-286. [204] X.J. Di, Y.J. Wang, D.Y. Han, Y.L. Fu, A.S. Duerfeldt, B.S. Blagg, T.W. Mu, Grp94 Protein Delivers gamma-Aminobutyric Acid Type A (GABAA) Receptors to Hrd1 Protein- mediated Endoplasmic Reticulum-associated Degradation, J Biol Chem 291 (2016) 9526- 9539. [205] N. Gomez-Navarro, E. Miller, Protein sorting at the ER-Golgi interface, J Cell Biol 215 (2016) 769-778. [206] H. Nakanishi, Y. Suda, A.M. Neiman, Erv14 family cargo receptors are necessary for ER exit during sporulation in Saccharomyces cerevisiae, J Cell Sci 120 (2007) 908-916.

139

[207] H.P. Hauri, F. Kappeler, H. Andersson, C. Appenzeller, ERGIC-53 and traffic in the secretory pathway, J Cell Sci 113 ( Pt 4) (2000) 587-596. [208] T. Duellman, J. Burnett, A. Shin, J. Yang, LMAN1 (ERGIC-53) is a potential carrier protein for matrix metalloproteinase-9 glycoprotein secretion, Biochem Biophys Res Commun 464 (2015) 685-691. [209] B. Zhang, M.A. Cunningham, W.C. Nichols, J.A. Bernat, U. Seligsohn, S.W. Pipe, J.H. McVey, U. Schulte-Overberg, N.B. de Bosch, A. Ruiz-Saez, G.C. White, E.G. Tuddenham, R.J. Kaufman, D. Ginsburg, Bleeding due to disruption of a cargo-specific ER-to-Golgi transport complex, Nat Genet 34 (2003) 220-225. [210] M. Zhu, C. Zheng, W. Wei, L. Everett, D. Ginsburg, B. Zhang, Analysis of MCFD2- and LMAN1-deficient mice demonstrates distinct functions in vivo, Blood Adv 2 (2018) 1014- 1021. [211] R. Khoriaty, M.P. Vasievich, D. Ginsburg, The COPII pathway and hematologic disease, Blood 120 (2012) 31-38. [212] Y. Ogawa, K. Yanagisawa, Y. Uchiyama, A. Matsumoto, M. Inoue, K. Toyama, Y. Miyazawa, N. Matsumoto, H. Handa, [Congenital factor V and factor VIII deficiency discovered in an elderly patient with abnormal bleeding after trauma], Rinsho Ketsueki 59 (2018) 383- 388. [213] T. Satoh, K. Suzuki, T. Yamaguchi, K. Kato, Structural basis for disparate sugar-binding specificities in the homologous cargo receptors ERGIC-53 and VIP36, PLoS One 9 (2014) e87963. [214] C. Zheng, H.H. Liu, S. Yuan, J. Zhou, B. Zhang, Molecular basis of LMAN1 in coordinating LMAN1-MCFD2 cargo receptor formation and ER-to-Golgi transport of FV/FVIII, Blood 116 (2010) 5698-5706. [215] C. Zheng, R.C. Page, V. Das, J.C. Nix, E. Wigren, S. Misra, B. Zhang, Structural characterization of carbohydrate binding by LMAN1 protein provides new insight into the endoplasmic reticulum export of factors V (FV) and VIII (FVIII), J Biol Chem 288 (2013) 20499-20509. [216] C. Zheng, B. Zhang, Combined deficiency of coagulation factors V and VIII: an update, Semin Thromb Hemost 39 (2013) 613-620. [217] Y.L. Fu, Y.J. Wang, T.W. Mu, Proteostasis Maintenance of Cys-Loop Receptors, Adv Protein Chem Struct Biol 103 (2016) 1-23. [218] A. Nemecz, M.S. Prevost, A. Menny, P.J. Corringer, Emerging Molecular Mechanisms of Signal Transduction in Pentameric Ligand-Gated Ion Channels, Neuron 90 (2016) 452- 470. [219] Y. Ge, Y. Kang, R.M. Cassidy, K.M. Moon, R. Lewis, R.O.L. Wong, L.J. Foster, A.M. Craig, Clptm1 Limits Forward Trafficking of GABAA Receptors to Scale Inhibitory Synaptic Strength, Neuron 97 (2018) 596-610 e598. [220] J.T. Kittler, K. McAinsh, S.J. Moss, Mechanisms of GABAA receptor assembly and trafficking: implications for the modulation of inhibitory neurotransmission, Mol Neurobiol 26 (2002) 251-268. [221] J.M. Lorenz-Guertin, T.C. Jacob, GABA type a receptor trafficking and the architecture of synaptic inhibition, Dev Neurobiol 78 (2018) 238-270. [222] J.S. Martenson, T. Yamasaki, N.H. Chaudhury, D. Albrecht, S. Tomita, Assembly rules for GABAA receptor complexes in the brain, Elife 6 (2017) e30826. [223] M. Mele, G. Leal, C.B. Duarte, Role of GABAA R trafficking in the plasticity of inhibitory synapses, J Neurochem 139 (2016) 997-1018.

140

[224] J.P. Klaus, P. Eisenhauer, J. Russo, A.B. Mason, D. Do, B. King, D. Taatjes, C. Cornillez-Ty, J.E. Boyson, M. Thali, C. Zheng, L. Liao, J.R. Yates, 3rd, B. Zhang, B.A. Ballif, J.W. Botten, The intracellular cargo receptor ERGIC-53 is required for the production of infectious arenavirus, coronavirus, and filovirus particles, Cell Host Microbe 14 (2013) 522-534. [225] T.G. Hales, H. Kim, B. Longoni, R.W. Olsen, A.J. Tobin, Immortalized hypothalamic GT1-7 neurons express functional gamma-aminobutyric acid type A receptors, Mol Pharmacol 42 (1992) 197-202. [226] P.L. Mellon, J.J. Windle, P.C. Goldsmith, C.A. Padula, J.L. Roberts, R.I. Weiner, Immortalization of hypothalamic GnRH neurons by genetically targeted tumorigenesis, Neuron 5 (1990) 1-10. [227] B. Zhang, C. Zheng, M. Zhu, J. Tao, M.P. Vasievich, A. Baines, J. Kim, R. Schekman, R.J. Kaufman, D. Ginsburg, Mice deficient in LMAN1 exhibit FV and FVIII deficiencies and liver accumulation of alpha1-antitrypsin, Blood 118 (2011) 3384-3391. [228] W.E. Balch, R.I. Morimoto, A. Dillin, J.W. Kelly, Adapting proteostasis for disease intervention, Science 319 (2008) 916-919. [229] G. Martinez, S. Khatiwada, M. Costa-Mattioli, C. Hetz, ER Proteostasis Control of Neuronal Physiology and Synaptic Function, Trends Neurosci 41 (2018) 610-624. [230] B. Zhang, R.J. Kaufman, D. Ginsburg, LMAN1 and MCFD2 form a cargo receptor complex and interact with coagulation factor VIII in the early secretory pathway, J Biol Chem 280 (2005) 25881-25886. [231] S. Pagant, A. Wu, S. Edwards, F. Diehl, E.A. Miller, Sec24 is a coincidence detector that simultaneously binds two signals to drive ER export, Curr Biol 25 (2015) 403-412. [232] T. Anelli, M. Alessio, A. Mezghrani, T. Simmen, F. Talamo, A. Bachi, R. Sitia, ERp44, a novel endoplasmic reticulum folding assistant of the thioredoxin family, EMBO J 21 (2002) 835-844. [233] S. Vavassori, M. Cortini, S. Masui, S. Sannino, T. Anelli, I.R. Caserta, C. Fagioli, M.F. Mossuto, A. Fornili, E. van Anken, M. Degano, K. Inaba, R. Sitia, A pH-regulated quality control cycle for surveillance of secretory protein assembly, Mol Cell 50 (2013) 783-792. [234] S. Watanabe, M. Harayama, S. Kanemura, R. Sitia, K. Inaba, Structural basis of pH- dependent client binding by ERp44, a key regulator of protein secretion at the ER-Golgi interface, Proc Natl Acad Sci U S A 114 (2017) E3224-E3232. [235] T. Johansson, T. Norris, H. Peilot-Sjogren, Yellow fluorescent protein-based assay to measure GABA(A) channel activation and allosteric modulation in CHO-K1 cells, PLoS One 8 (2013) e59429. [236] C. McCabe, D. Shaw, J.R. Atack, L.J. Street, K.A. Wafford, G.R. Dawson, D.S. Reynolds, J.C. Leslie, Subtype-selective GABAergic drugs facilitate extinction of mouse operant behaviour, Neuropharmacology 46 (2004) 171-178.

141