Investigation of Proteins That Associate with Naadp-Gated Two-Pore Channels

INVESTIGATION OF PROTEINS THAT ASSOCIATE WITH NAADP-GATED TWO-PORE CHANNELS

A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF

THE UNIVERSITY OF MINNESOTA

Yaping Lin Moshier

IN PARTIAL FULTILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

Advisor: Dr. Jonathan S. Marchant

FEBRUARY, 2012

Acknowledgements

I would like to acknowledge my advisor Dr. Jonathan Marchant first and foremost, for his full-hearted support and encouragement in the past few years. The work presented here wouldn‟t have been accomplished without his intelligent insights and guidance. I would particularly like to appreciate efforts he put in to help me continue the PhD training, and encourage me to pursue higher achievements.

I would like to mention all the current and past members in the Marchant lab for their direct or indirect contributions to this work: Dr. Taisaku Nogi, Dan Zhang, Laramie Rapp, John Chan, Peter Sebastian, Xiaolong Liu, and Mike Keebler. I especially liked scientific discussions and sporadic social events with them. Everyone has made the lab a fun and inspiring working environment.

I would also like to thank my committee members- Dr. Stanley Thayer (chair), Dr. Esam El-Fakahany, and Dr. Timothy Walseth- for their great advice, suggestions, and technical assistance to refine this thesis.

Direct contributions from other collaborators have also been fascinating, and helped me to achieve my research goals more efficiently. Specifically, I would like to mention the following individuals: Dr. Sandip Patel from the University College London UK for all the TPC constructs and urchin homogenates to perform pharmacological analyses, live cell imaging and interaction studies; Dr. Sean Davidson from UCL for the TPC knockout samples; Dr. James Slama from the University of Toledo for the 5-azido nicotinic acid to make the photoprobe; Dr. Timothy Walseth from the department to synthesize the photoaffinity probe and all the PAL technical support and resources; Dr. Eugen Brailoiu from the Temple University School of Medicine for the calcium release assay in mammalian cells; Todd Markowski, LeeAnn Higgins, and Bruce Witthuhn from the Center for Mass Spectrometry and Proteomics at the U of M for the assistance of MS experimental design and technical support.

During the course of preparing this thesis, my work was mainly supported by the following funding agents: NIH grant GM088790 (Dr. Marchant), BBSRC grant BB/G013721/1 (Dr. Patel), and AHC seed grant from UMN 2101.95 (Dr. Walseth & Dr. Marchant). I was also supported by a Doctoral Dissertation Fellowship from the Graduate School at the University (2010-2011).

I would also like to grant special thanks to my family, friends and former co- workers and mentors in the US and Taiwan for their help, support and tremendous belief in me during the hard time in the past five and half years. Last not least, I would like to acknowledge my husband, Adam Moshier, for his understanding that I had to consistently concentrate on study and work, his appreciation and being supportive to my research, tolerating my eccentric scientific talk, and being extremely thoughtful and considerate before and during the graduate training. Without all these, I wouldn‟t have had enough strength to accomplish this work.

Abstract

All living organisms respond to environmental stimuli to coordinate biological activities by eliciting a sequence of signaling cascades, many of which converge to 2+ 2+ regulate cytosolic calcium ion concentration ([Ca ]cyt) via release of intracellular Ca stores. Three agonist-mediated Ca2+ mobilizing messengers have been identified, including inositol 1,4,5-trisphosphate (IP3), cyclic ADP ribose (cADPR), and nicotinic acid adenine dinucleotide phosphate (NAADP). NAADP is the most potent calcium 2+ mobilizer identified to date, and unlike IP3 and cADPR which target ER Ca stores via 2+ binding to IP3Rs and interaction with RyRs respectively, NAADP-mediated Ca release is evoked from acidic Ca2+ stores, such as endolysosomes and lysosome-like organelles. The endogenous target(s) for NAADP is however controversial, and several Ca2+ channels expressing in the endolysosomal system have been proposed to be gated by NAADP. Recently discovered two-pore channels (TPCs) have been identified as the long-sought NAADP targets via pharmacological, biochemical and genetic approaches.

In this thesis, I utilized a photoactivable NAADP analogue, 5-azido NAADP (5N3- NAADP), with the goal of identifying NAADP targets via a direct crosslinking approach coupled with mass spectrometry (MS). My results revealed that 5N3-NAADP was able to recapitulate the basic pharmacological properties of the native ligand NAADP; however, the labeled protein candidate(s) (22/23kDa doublet in mammals and 41kDa in sea urchin) was significantly smaller than the predicted size for TPC proteins. The photolabeled doublet was enriched in soluble fractions in several mammalian cells, even though higher affinity was conserved when membrane bound. The labeling pattern and intensity of the 22/23kDa doublet remained unchanged in either TPC-overexpressing or knockdown samples, suggesting NAADP binding and Ca2+ release are mediated by distinct protein identities. Combinatorial PAL-2D-MS analysis for the high affinity 22/23kDa doublet has identified multiple potential candidates, several of which have been shown to express in both cytosolic and membrane compartments, and are involved in vesicle fusion and trafficking in the endolysosomal system, a physiological mechanism regulated by NAADP and TPCs. Most conservatively, my studies demonstrate that there exists an iii accessory cellular binding partner for NAADP within the TPC signaling complex, which can then serve to indirectly regulate TPC activity. This is an important revision of current dogma, and crucial for rational design of drugs that may modulate NAADP activity. Such therapeutics may be important in disorders (diabetes, lysosomal storage disorders, and neuronal excitotoxicity) where NAADP signaling is pathologically perturbed.

The second part of my thesis utilized a parallel immunoprecipitation-MS (IP-MS) strategy in an attempt to characterize components of the TPC-signaling complex in mammalian cells. However, while the feasibility of IP-MS strategy to identify TPC proteins was demonstrated, the first candidate examined - CHERP (calcium homeostasis endoplasmic reticulum protein) - failed subsequent validation as a TPC accessory protein. CHERP has been proposed as an integral ER protein that interacts with ryanodine receptors, suggesting an intriguing model for coupling TPCs to RyRs. Remarkably, in contrast to all published data, my results demonstrated that CHERP is actually a nuclear protein. I characterized that the carboxyl terminal region is necessary and sufficient to target CHERP in the nucleus. I identified novel properties for CHERP and proposed its biological functions are to regulate mRNA synthesis, maturation and transport, which may account for some or all of CHERP‟s effect in mediating ER calcium regulation and cellular senescence. This illustrates the need for caution in interpretation of direct (one- step) immunoprecipitation-MS analysis.

Table of Contents Acknowledgements------i

Abstract------iii

Table of Contents------v

Summary of Tables ------vi

Summary of Figures ------vii

Summary of Manuscripts------viii

Abbreviations------ix

Chapter 1: Introduction------1

I. Intracellular Ca2+ stores- a growing community of acidic Ca2+ organelles------2 II. NAADP targets acidic Ca2+ stores------4 III. NAADP targets two-pore channels (TPCs) in the endolysosomal system------8 IV. Conclusion & Rationale for this study------10

Chapter 2: Photoaffinity labeling of nicotinic acid adenine dinucleotide phosphate (NAADP) targets in mammalian cells------11

I. Introduction------12 II. Materials and Methods------14 III. Results------17 IV. Discussion------35

Chapter 3: Elucidation of the TPC-interactome in mammalian cells by an immunoprecipitation-mass spectrometry strategy------40

I. Introduction------41 II. Material and Methods------43 III. Results------47 IV. Discussion------64

Bibliography------70 v

Summary of Tables

Table 2.1. Protein candidates for the 22/23kDa doublet identified by PAL-2D-MS------33

Table 2.2. Photolabeling characteristics of 22/23kDa doublet in three different mammalian samples------35

Table 3.1. Primer sequences for subcloning NH2- vs. C-terminal deletion constructs of CHERP------46

Table 3.2. Algorithms used for in silico prediction of domain structures and post- translational modifications of CHERP------59

Summary of Figures

32 32 Figure 2.1. Schematics of P-NAADP and P-5N3-NAADP synthesis------17

Figure 2.2. Validation of the photoaffinity labeling (PAL) method------19

Figure 2.3. Photoaffinity labeling of SKBR3 whole cell lysate------21

Figure 2.4. NADP and NAADP protection of PAL in SKBR3 whole cell lysate------23

Figure 2.5. Properties of the 22/23kDa doublet in different SKBR3 subcellular fractions------25

Figure 2.6. PAL in HEK293 cells and mouse pancreas------26

Figure 2.7. Lack of effect of overexpressed TPC isoforms on PAL profiles------29

Figure 2.8. PAL in TPC knockout mouse pancreas------30

Figure 2.9. Combinatorial PAL-2D-MS in SKBR3 membrane preparation------32

Figure 2.10. Combinatorial PAL-IP in SKBR3 WCL------34

Figure 3.1. Identification of TPC-interacting proteins in HEK293 cells------48

Figure 3.2. Non-specific interaction between TPCs and CHERP in HEK293 cells------50

Figure 3.3. Co-localization of overexpressed CHERP with TPC1/2 or organellar markers------52

Figure 3.4. Validation of anti-CHERP Ab followed by immunocytochemistry of endogenous CHERP in cell lines------56

Figure 3.5. Immunohistochemistry of endogenous CHERP in muscles------57

Figure 3.6. Schematic diagram of functional domains in CHERP------58

Figure 3.7. Subcellular distribution of wild-type vs. C-terminal truncation mutants of CHERP------60

Figure 3.8. C-terminal domains of CHERP is sufficient for nuclear localization------63

vii

Summary of Manuscripts

Yaping Lin-Moshier, Timothy F. Walseth, Dev Churamani, Sean M. Davidson, James T. Salma, Robert Hooper, Eugen Brailoiu, Sandip Patel, and Jonathan Marchant. Photoaffinity labeling of nicotinic acid adenine dinucleotide phosphate (NAADP) targets in mammalian cells. J. Biol. Chem. 2012 Jan. 287 (4): 2296-2307. **Paper of the week**

Timothy F. Walseth, Yaping Lin-Moshier, Pooja Jain, Margarida Ruas, John Parrington, Antony Galione, Jonathan S. Marchant, and James T. Slama. Photoaffinity labeling of high affinity nicotinic acid adenine dinucleotide 2‟-phosphate (NAADP) proteins in sea urchin egg. J. Biol. Chem. 2012 Jan. 287 (4): 2308-2315. **Paper of the week**

Yaping Lin-Moshier, Peter S. Sebastian, and Jonathan S. Marchant. Re-evaluation of the role of calcium homeostasis endoplasmic reticulum protein (CHERP) in cellular senescence signaling. 2012. (manuscript in preparation)

Yaping Lin-Moshier, and Jonathan S. Marchant. Ca2+ signaling in Xenopus oocytes. Cold Spring Harbor Laboratory Press. 2012 Apr. (manuscript in press)

viii

Abbreviations

2+ [Ca ]cyt = Free cytosolic calcium ion concentration ADP = Adenosine diphosphate

AMPA = α-amino-3-hydroxyl-5-methylisoxazole-4-propionic acid ATP = Adenosine triphosphate ANOVA = Analysis of variance cADPR = Cyclic adenosine diphosphate ribose

2+ Cav & VOCC = voltage-activated/operated Ca channel CD38 = cluster of differentiation 38 CHERP = Calcium homeostasis endoplasmic reticulum protein CICR = Calcium-inducted calcium release CREB = cAMP responsive element binding protein DAPI = 4‟,6-diamidino-2-phenylindole dihydrochloride DNA = Deoxynucleic acid DTT = Dithiothreitol EDTA = Ethylenediaminetetraacetic acid ER = Endoplasmic reticulum EST = Expressed sequence tag GFP = Green fluorescent protein

GPN = Glycyl-L-phynylalanine 2-naphthylamide HEPES = 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

EC50 = Half maximal effective concentration HPLC = High performance liquid chromatography

IC50 = Half maximal inhibitory concentration IP = Immunoprecipitation

IP3 = Inositol 1,4,5 trisphosphate ix

IP3R = Inositol 1,4,5 trisphosphate receptor

Kv = Voltage-activated potassium channel LAMP1 = Lysosomal-associated membrane protein 1 mRNA = Messenger ribonucleic acid MS = Mass spectrometry NAADP = Nicotinic acid adenine dinucleotide phosphate NAD+ = Nicotinamide adenine dinucleotide NADP = Nicotinamide adenine dinucleotide phosphate

+ Nav = Voltage-activated Na channel NMDA = N-methyl-D-aspartic acid PAL = Photoaffinity labeling PBS = Phosphate buffered saline PMSF = Phenylmethanesulfonylfluoride PTM = Post-translational modification PVDF = Polyvinylidene fluoride RFP = Red fluorescent protein RT-PCR = Reverse-transcription polymerase chain reaction RyR = Ryanodine receptor SDS-PAGE = Sodium dodecyl sulfate-polyacrylamide gel electrophoresis SERCA = Sarco/endoplasmic reticulum Ca2+-ATPase snRNP = Small nuclear ribonucleoprotein SOCE = Store-operated calcium entry STIM = Stromal interaction molecule TPCs = Two-pore channels TRPM2 = Transient receptor potential cation channel, subfamily M, member 2 TRPML1 = Transient receptor potential mucolipin 1

Chapter One - Introduction

All living organisms respond to environmental stimuli to coordinate biological activities by eliciting a sequence of signaling cascades. Many crucial signal transduction pathways converge in utilizing the calcium ion (Ca2+) as a pivotal second messenger that regulates important physiological functions (Berridge et al., 2000). The concentration of 2+ 2+ free Ca in the cytoplasm ([Ca ]cyt) is maintained at nanomolar range (typically 100nM) through cooperation of highly specialized proteins, such as pumps, exchangers and buffering macromolecules, localized on the plasma membrane and in many subcellular compartments (Berridge et al., 2000; Patel and Docampo, 2010; Galione et al., 2011). 2+ 2+ Elevation of [Ca ]cyt is mediated by fluxes from the extracellular space where [Ca ] is ~1mM, or release from intracellular Ca2+-containing organelles (luminal [Ca2+]= 3- 600μM), both mechanisms involve opening of Ca2+-permeable ion channels on the plasma membrane or organellar membranes, which then allow Ca2+ to move down its electrochemical gradient. Several Ca2+-permeable cation channels have been identified, whose gating is regulated by ligand binding (e.g. neurotransmitters, cellular metabolites of phospholipids or nucleotides, etc.), changes of membrane potential, temperature, mechanical stretch, or even Ca2+ itself. Interplay between different ligands and their calcium permeable receptors, localization and distribution of these channels and composition of their interacting proteins in the surrounding microenvironments, all contribute to the most fascinating spatiotemporal properties, complexity and specificity of calcium signaling (Berridge et al., 2000). While calcium influx through plasma membrane Ca2+ channels upon stimulation has been extensively studied, the understanding of intracellular Ca2+ organelles, Ca2+ channels expressed on these stores, mobilizing messengers that gate these channels, and signaling protein complex of these integral Ca2+-permeable ion channels is far more limited.

This thesis focuses on the most recently characterized Ca2+-containing organelles, the acidic Ca2+ stores, and a novel protein family of calcium-permeable ion channels expressed on these acidic organelles, two-pore channels (TPCs), in elucidating their 1 interacting protein complexes by structure-based crosslinking strategy (Chapter 2) and antibody-based purification (Chapter 3). A brief introduction of acidic calcium stores, the Ca2+-mobilizing messenger(s) targeting these organelles, and channels potentially gated by these mobilizers will be first reviewed.

I. Intracellular Ca2+ stores- a growing community of acidic Ca2+ organelles

2+ [Ca ]cyt is maintained at ~100nM, a concentration at least one order of magnitude lower than that found in most of the membranous organelles (Miyawaki et al., 1997; Pinton et al., 1998; Christensen et al., 2002; Parekh, 2003; Lloyd-Evans et 2+ al., 2008). Cells exploit several mechanisms to minimize fluctuation of [Ca ]cyt, so as to ensure high efficiency and full efficacy of Ca2+-mediated signal transduction upon agonist stimulation. Such organelles are therefore enriched with low-affinity and high-capacity Ca2+-buffering molecules, which are capable of sequestering excess free Ca2+ from the cytoplasm and/or releasing Ca2+ to the cytosol to initiate a sequence of signaling cascades (Berridge et al., 2000; Patel and Docampo, 2010; Galione et al., 2011). This “filling” or sequestering process normally requires energy generated from ATP or pyrophosphate hydrolysis to move Ca2+ against its electrochemical gradient. Such mechanisms are mediated by highly specialized integral protein complexes sitting on the membrane of these Ca2+ stores, including (i) the sarco/endoplasmic reticulum Ca2+-ATPase (SERCA) on the S/ER membrane and nuclear envelope (Downie et al., 1998; Guerrero-Hernandez et al., 2010), (ii) the secretory pathway Ca2+-ATPase (SPCA) on the Golgi apparatus (Missiaen et al., 2007), and (iii) the H+-ATPase coupled with the Ca2+/H+ exchanger (CAX), the Na+/Ca2+ exchanger (NCX) or other unknown mechanisms on acidic Ca2+ stores (Patel and Docampo, 2010; Galione et al., 2011). Further, Ca2+ flux into Ca2+ stores can be directly mediated by the Ca2+ uniporter and the NCX on the mitochondria (Palty et al., 2010; De Stefani et al., 2011), or indirectly by the store-operated calcium entry (SOCE) mechanism mediated by Ca2+ sensors (STIM1 & 2) on the ER or acidic stores, coupled with Ca2+ release-activated Ca2+ channels (CRACs, such as Orai) on the plasma membrane upon store depletion (Shen et al., 2011; Zbidi et al., 2011). 2

Although the ER is by far the largest and best-studied intracellular Ca2+ reservoir, recent studies in bacteria, protists, yeast, plants and higher organisms have highlighted the role of acidic Ca2+ stores in normal physiological processes and involvement in disease progression (Lloyd-Evans et al., 2010; Galione et al., 2011; Morgan et al., 2011). These acidic stores featured with low pH (high proton) and high Ca2+ content in their lumens include acidocalcisomes (prokaryotes), vacuoles (protists, slime molds, yeast and plants), endosomes/lysosomes (eukaryotes), lysosome-related organelles, secretory vesicles and Golgi apparatus (Patel and Docampo, 2010). Here, we focus on the endocytic pathways mediated by the endolysosomal system. Endolysosomes are traditionally referred to as the cellular defensive/digestive mechanisms to process pathogens and damaged macromolecules by hydrolase enzymes in their acidic lumen by means of endocytosis and dynamic membrane fusion and fission from divergent vesicles (Pillay et al., 2002). The luminal [Ca2+] in endosomes is initially similar to, if not the same as, that in the extracellular space (~1mM) and gradually decrease to ~60-400μM as they acidify by the action of H+- ATPase (Patel and Docampo, 2010). The acidic pH in the lysosomal lumen provides a great environment for digestive enzymes to process and reuse these macromolecules, or to release the endocytotic ligand-receptor complex and recycle receptors back to the plasma membrane. Further, the proton gradient serves as a driving force for exchangers (CAX or NAX and others) to take up Ca2+ from the cytoplasm, and luminal Ca2+ is further sequestered by Ca2+-buffering molecules, such as polyphosphate.

Defects of digestive enzymes in lysosomal lumen have profound effects on cellular metabolism, and diseases associated with dysregulation of lysosome functions are generally termed lysosomal storage disorders. Notably, several neurodegenerative lysosomal storage disorders, such as Nieman-Pick type C (NPC), mucolipidosis type IV (MLIV) and Chediak-Higashi syndrome, are featured with impaired Ca2+ homeostasis and defects in endosome/lysosome membrane fusion and fission, a biological process mediated also by Ca2+ (Styrt et al., 1988; Lloyd-Evans et

al., 2008; Bach et al., 2010). Further, acidic Ca2+ stores are targeted by bile acids and alcohol metabolites, and dysregulation of these stores is found to play an important role in the progression of acute pancreatitis (Gerasimenko et al., 2009). All these pathological findings attest to the significance of these acidic Ca2+-containing organelles in Ca2+ homeostasis and Ca2+-mediated biological processes. Therefore, elucidating mediators that target these acidic stores is of pivotal significance, and will help gain insight into the fundamental mechanisms for these diseases and assistance in rational drug design for treatment.

II. NAADP targets acidic Ca2+ stores Three intracellular second messengers capable of releasing Ca2+ have been identified since 1983 using permeabilized pancreatic acinar cells or sea urchin egg homogenates as model systems (Streb et al., 1983; Galione et al., 2011; Lee, 2011). 2+ The first Ca -mobilizing messenger is inositol 1,4,5-trisphosphate (IP3), a metabolite

of the plasma membrane phospholipid, phosphatidylinositol 4,5-bisphosphate (PIP2),

derived upon agonist-dependent activation of phospholipase C (PLC). IP3 was later 2+ found to target ER Ca stores by binding an ER membrane protein, the IP3 receptor

(IP3R), to elicit calcium release directly through opening of the channel or indirectly through a calcium-induced calcium release (CICR) mechanism. A series of seminal

experiments have characterized (i) the structural features of IP3R as a calcium- permeable cation channel on the ER, (ii) binding sites and spatial arrangements for

IP3 gating, (iii) pharmacological properties of ligand binding and affinity effected by free [Ca2+], (iv) residues for important protein-protein interactions and post-

translational modifications, and (v) biological activities related and relevant to IP3R- mediated Ca2+ signaling (Bezprozvanny et al., 1991; Mikoshiba et al., 1994; Choe and Ehrlich, 2006; Foskett et al., 2007). Using a sea urchin system, Lee and colleagues performed chemical screening 2+ with the attempt to identify potential Ca mobilizers other than IP3. Surprisingly, two pyridine nucleotide metabolites (NAD+ and NADP) that exhibited comparable Ca2+- releasing activities were identified. The chemical formula and structures of the

4 effective molecules in the commercial preparation were delineated later as cyclic adenosine diphosphate ribose (cADPR) and nicotinic acid adenine dinucleotide phosphate (NAADP) (Lee, 1991; Lee and Aarhus, 1995). cADPR is found to target ER Ca2+ stores via gating to the ryanodine receptor (RyR), another Ca2+-permeable cation channel on the ER, or increasing RyR sensitivity to Ca2+ (Mothet et al., 1998; Thomas et al., 2001). Recent studies in Xenopus oocytes also suggest that cADPR may target SERCA pump to indirectly mediate ER Ca2+ release (Yamasaki-Mann et al., 2009). Cellular levels of cADPR increase in response to cytokine treatment (interleukin 8, tumor necrosis factor α, interferon-, etc.), hormones (17-estradiol, cholecystokinin, angiotensin II, etc.) and nutrients (e.g. glucose), via activation of a protein family exhibiting ADP-ribosyl cyclase activity (e.g. CD38). Therefore, cADPR was confirmed as a native second messenger in vivo (Guse, 1999; Lee, 2011). 2+ Unlike IP3 and cADPR, which target ER Ca stores via binding IP3R and RyR, NAADP-mediated Ca2+ release possesses distinct pharmacological properties in that:

(i) homologous desensitization of IP3- and cADPR-mediated calcium responses does not affect NAADP-regulated Ca2+ release (Chini et al., 1995; Lee and Aarhus, 1995), (ii) the concentration-response relationship of NAADP-evoked Ca2+ release is a monotonic sigmoid or bell-shaped curve in different cell types with self-inactivation properties, which is absent in IP3- and cADPR-mediated mechanisms (Morgan and 2+ 2+ Galione, 2008), (iii) Ca -induced Ca release (CICR), a feature for IP3R and RyR- mediated Ca2+-release mechanism, is not shared by NAADP-targeted Ca2+ release (Chini and Dousa, 1996; Bak et al., 2002), and (iv) NAADP-regulated Ca2+ response is less sensitive to pH changes, which allows NAADP to target acidic stores more efficiently (Patel et al., 2000). Several lines of evidence have demonstrated that NAADP mobilizes Ca2+ from acidic stores, observations include: (i) in stratified sea urchin eggs, NAADP-targeted organelles migrate to the opposite pole where ER, nucleus and most of neutral organelles are resident, (ii) NAADP-targeted organelles can be stained by lysotracker, and NAADP-binding activity is co-localized with lysosomal marker, (iii) NAADP- mediated Ca2+ response is insensitive to thapsigargin treatment, which inhibits 5

SERCA activity and depletes ER Ca2+ stores; nonetheless, NAADP responsiveness is abolished when cells or purified microsomes are pretreated with protonophores or bafilomycin A1 to disrupt the H+ gradient in acidic Ca2+ stores, and (iv) NAADP- elicited Ca2+ response is totally abrogated in the presence of the lysosomotropic agent, glycyl-L-phynylalanine 2-naphthylamide (GPN) (Lee et al., 2000; Churchill et al., 2002; Galione and Churchill, 2002; Galione, 2006). That NAADP serves as a real second messenger in vivo, instead of simply a pharmacological agonist to an unknown protein in the acidic Ca2+ stores, was hinted from a study by Chini and colleague, where NAADP was detected after incubating its precursor, NADP+, with a variety of rat tissue homogenates (Chini and Dousa, 1995). Followed by similar studies, Dickey quantified the endogenous levels of NAADP with a range from 1.73 to 4.90 pmol/mg in mammalian tissues, suggesting NAADP may be a universal Ca2+-releasing messenger (Dickey, 1999). Taking the advantage of the unique pharmacological properties of NAADP binding in sea urchin, Churamani et al., have increased the sensitivity of detection methods, and carefully examined the endogenous levels of NAADP in several primary cells, including human red blood cells, rat hepatocytes, and E. coli, with a range of 81 to 142 fmol/mg protein (Churamani et al., 2004). Furthermore, results utilizing primary pancreatic acinar cells also revealed a basal NAADP level around 0.6 pmol/mg in this cell type, which was then increased to ~6 pmol/mg upon treatment of cholecystokinin (CCK) at its physiological concentration (10pM) (Yamasaki et al., 2005). These pioneer studies provide the first evidence for NAADP‟s role as an agonist-dependent Ca2+-mobilizing messenger in vivo. A multitude of evidence in other cell types and systems has substantiated NAADP as the most potent Ca2+ mobilizer identified to date. Moreover, NAADP-mediated Ca2+ signaling from acidic stores is universally pivotal to a variety of physiological activities, encompassing fertilization (Lim et al., 2001; Moccia et al., 2006; Morgan and Galione, 2007), cell growth and differentiation (Brailoiu et al., 2006; Aley et al., 2010), muscle contraction (Boittin, 2002), endothelium function (Gambara et al., 2008; Brailoiu et al., 2010b; Esposito et al., 2011), hormone secretion (Johnson and Misler, 2002; Kim et al., 2008), immune responses (Berg et al., 6

2000; Cordiglieri et al., 2010; Rah et al., 2010), synaptic activity (Pandey et al., 2009), neurotransmitter release (Brailoiu et al., 2001), and programmed cell death (Zhang et al., 2010; Gómez-Suaga et al., 2011; Pereira et al., 2011). In general, NAADP- mediated Ca2+ release is small and localized, which can then serve as a trigger Ca2+ signal to recruit larger Ca2+ stores, such as those within the ER, to propagate into a 2+ larger and global Ca wave by means of CICR mediated by RyRs or IP3Rs or both, hence the so-called two-pool mechanism or “channel chatter” phenomena (Brailoiu et al., 2001; Boittin, 2002; Patel and Brailoiu, 2012). Several studies have then focused on the identification of the endogenous NAADP-synthesizing enzyme(s), with the mammalian homologue of ADP-ribosyl cyclase, CD38, proposed to be the key enzyme for NAADP synthesis in response to cellular activities (Aarhus et al., 1995; Cosker et al., 2010). Although it is rather arbitrary and debatable, a recombinant ADP-ribosyl cyclase has been used widely in chemical reactions in vitro to synthesize NAADP and its structural homologues, which then facilitate determination of the essential functional groups for Ca2+ release activity and detailed pharmacological properties for NAADP receptor(s) in biological samples (Lee and Aarhus, 1997; Jain et al., 2010). The intriguing question remains: what is the endogenous receptor(s) for NAADP? Given that NAADP naturally targets acidic Ca2+ stores, several Ca2+-permeable ion channels expressing in the endolysosomal system, including TRPML1, TRPM2, and even RyR, have been examined and proposed as the endogenous NAADP targets, although conflicting results were obtained (Copello et al., 2001; Gerasimenko et al., 2003; Beck et al., 2006; Dong et al., 2008; Zhang et al., 2011). A recent breakthrough in this regard derives from three independent research groups converging on the identification of a novel protein family, two-pore channel(s), in the endolysosomal system as the long-sought NAADP target in vivo (Calcraft et al., 2009; Brailoiu et al., 2009; Zong et al., 2009).

III. NAADP targets two-pore channels (TPCs) in the endolysosomal system Two-pore channels (TPCs) are a novel protein family in the endolysosomal system that exhibit Ca2+ stimulation by NAADP. The TPC protein was first cloned in 2000 by Ishibashi et al., following an EST library screen against the α-subunits (the pore-forming subunit) of voltage-activated Na+ channels (Ishibashi et al., 2000). Sequence analysis of animal and plant TPCs has revealed their structural features as + 2+ “half” of the voltage-operated Na and Ca channels (Nav and Cav), minimally with two homologous domains (I & II), each of which contains six transmembrane segments (S1-S6) (Ishibashi et al., 2000; Furuichi et al., 2001). Interestingly, functional analyses of TPC protein overexpressing in Xenopus oocytes and CHO cells fail to detect a membrane current in response to voltage change, suggesting (i) TPCs have gating properties different from voltage-activated calcium and sodium channels, (ii) TPCs may express in different subcellular compartments other than the plasma membrane, and (iii) auxiliary subunits may be required for TPC function (Ishibashi et al., 2000). That plant TPC protein exhibits channel activities similar to the Ca2+-permeable slow vacuolar channel (a lysosome-like organelle in plants) has led researchers to speculate that TPC proteins in animals also function as Ca2+ channels in the endolysosomal system. Homology analysis of animal genome and EST database has identified another two TPC isoforms, namely TPC2 and TPC3, in addition to the TPC1 isoform characterized by Ishibashi et al. in 2000. Co-expression of fluorescent- tagged TPCs and organellar markers in human cell lines has revealed distinct subcellular distribution patterns of TPC protein isoforms, with TPC1 & TPC3 (in sea urchin and certain vertebrates; absent in primates) targeting both late endosomes and lysosomes, and TPC2 strictly residing in the lysosome (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009). The endolysosomal localization and domain/structure homologous to Ca2+ channels renders TPCs the potential targets for NAADP in acidic Ca2+ stores. Studies exploiting biochemical, biophysical, and pharmacological approaches for different TPC isoforms from different species (human, urchin & rodent) by different 8 laboratories have demonstrated that (i) NAADP-mediated Ca2+ release or current is enhanced in TPC overexpressing cells, and attenuated in TPC knockdown or knockout samples, (ii) NAADP-binding activity is enhanced in membranes or lysosomes purified from TPC2-overexpressing cells compared to native control, (iii) NAADP binding is preserved in TPC immunoprecipitates, and (iv) disruption of proton gradient by bafilomycin A1 or treatment of GPN to induce lysosomolysis in TPC-overexpressing cells abolishes NAADP responsiveness, suggesting NAADP and TPCs synergistically function on the same subcellular compartments- the endolysosomal acidic stores (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009; Ruas et al., 2010). Mutagenesis studies of TPC proteins have identified several important residues for their channel conductivity, localization and post-translational modifications. Interestingly, a single mutation on the putative pore-forming regions of TPCs (leucine 273 for TPC1 and leucine 265 for TPC2) has been demonstrated to abolish channel activity in response to high concentration of NAADP (Brailoiu et al., 2009, 2010a). Further, mutation on the lysosomal targeting sequence, the dilucine motif ((D/E)XXXL(L/I)) redirects TPC proteins from lysosomes to the plasma membrane, and the NAADP-mediated membrane current is now measurable by patch clamp techniques and is disengaged from the ER Ca2+ stores (Brailoiu et al., 2010a). Direct physiological relevance of TPCs in NAADP-mediated biological activities is obtained from loss-of-function studies, including (i) smooth muscle contraction upon NAADP treatment is abolished in TPC knockout mice, and agonist-mediated contraction no longer couples to the acidic stores, (ii) in skeletal muscle precursor cells, NAADP- potentiated muscle differentiation is attenuated when small interference RNA (siRNA) targeting endogenous TPC1 and TPC2 are introduced into the cells, and (iii) NAADP-elicited neuronal differentiation in pheochromocytoma cells (PC12) is also abolished in TPC2 knockdown cells (Brailoiu et al., 2006; Tugba Durlu-Kandilci et al., 2010; Aley et al., 2010). All these data establish TPCs as NAADP-targeted Ca2+ channels on endolysosomes.

How do TPCs then form a functional Ca2+ channel in the endolysosomal systems? Studies combining epitope tagging, protease protection, antibody detection, imaging analyses and chemical crosslinking methods have recently revealed several structural features and topology for TPC proteins on the membrane, including (i) both TPC1 and TPC2 contain cytosolic amino- and carboxyl-termini, and a cytosolic loop connecting the two homologous domains, (ii) essential amino acid residues and glycosylation sites on the P-loop between S5 and S6 have been identified, and the luminal localization and the pore-forming region is verified, and (iii) several positively charged residues on the S4 of both domains on TPCs may serve as voltage sensors (Churamani et al., 2011; Hooper et al., 2011; Patel et al., 2011). That TPC

protein structures resemble half of the pore-forming subunits for Nav and Cav raises the possibility that TPCs may dimerize in order to constitute a functional channel. Indeed, several researches employing colocalization and interaction strategies in cell lines have confirmed oligomerization of TPCs in vivo (Zong et al., 2009; Churamani et al., 2011; Rietdorf et al., 2011). Therefore, the role of TPCs as novel Ca2+- permeable ion channels on the acidic Ca2+ stores and their involvement in NAADP- mediated physiological activities have added a new chapter in the cellular Ca2+ regulation and broaden our knowledge in the complexity of Ca2+ signaling.

IV. Conclusion and rationale for this study The importance of NAADP as an essential agonist-generated Ca2+-mobilizing messenger targeting the acidic Ca2+ stores, and the significance of TPC function in the context of NAADP-mediated signaling are becoming clearer and appreciated. However, several intriguing questions remain: (i) where is the NAADP binding site(s)

on TPC proteins? (ii) Do TPCs, just like Nav and Cav, require auxiliary subunits to function? And if yes, what are the protein identities for these auxiliary subunits? (iii) Do TPCs interact with ER components and play a role in the context of NAADP- mediated “channel chatter” phenomena where NAADP triggers whole cell Ca2+ signals through Ca2+-mediated sensitization of ER Ca2+ release channels? This thesis attempts to address these fundamental questions. 10

Chapter Two

Photoaffinity labeling of nicotinic acid adenine dinucleotide phosphate (NAADP) targets in mammalian cells

Yaping Lin-Moshier, Timothy F. Walseth, Dev Churamani, Sean M. Davidson, James T. Salma, Robert Hooper, Eugen Brailoiu, Sandip Patel, Jonathan Marchant J Biol Chem. 2012 Jan; 287 (4): 2296-2307 *Paper of the week*

Introduction

NAADP is a potent Ca2+ releasing messenger that has been shown to regulate many physiological processes encompassing fertilization, secretion, neurite outgrowth and synaptic function (Lee, 2005; Galione et al., 2010, 2011; Hooper and Patel, 2011). While molecular identification of the Ca2+ channel(s) targeted by NAADP remained frustratingly elusive for many years, several lines of evidence have recently unmasked the two-pore channel (TPC) family of endolysosomal proteins as candidate NAADP receptors (Calcraft et al., 2009; Brailoiu et al., 2009; Zong et al., 2009). Supporting evidence derives from gain and loss of function approaches, electrophysiological analyses and radioligand binding. For example, overexpression of either of the two human TPC isoforms (TPC1 & TPC2) enhanced NAADP-evoked Ca2+ responses when assessed by Ca2+ imaging in multiple cell lines (Brailoiu et al., 2009, 2010; Calcraft et al., 2009; Dionisio et al., 2011; Ogunbayo et al., 2011; Zong et al., 2009). Reciprocally, knockdown of endogenous TPC1 ablated NAADP responsiveness (Brailoiu et al., 2009) and NAADP-activated Ca2+ currents were abolished in pancreatic β-cells isolated from a TPC2 knockout mouse (Calcraft et al., 2009). Electrophysiological insight has derived from different approaches in which TPC activity was recorded from individual lysosomes in vitro (Schieder et al, 2010), reconstituted channels in planar lipid bilayers (Pitt et al., 2010), or from channels rerouted to the cell surface via mutagenesis of a lysosomal targeting sequences (Brailoiu et al., 2010a; Yamaguchi et al., 2011). Each approach demonstrated addition of NAADP at nanomolar concentrations stimulated Ca2+ permeable currents and/or single channel activity. Finally, radioligand binding methods using membranes overexpressing TPC2, or endogenous TPC isoforms immunoprecipitated from sea urchin eggs, demonstrated enhanced 32P-NAADP binding relative to control samples (Calcraft et al., 2009; Ruas et al., 2010). Cumulatively, this growing dataset has established TPCs as NAADP-sensitive Ca2+ channels within the endolysosomal system. Despite this progress, little is currently known about the structural basis of NAADP interaction with the TPC proteins and the binding site(s) for the endogenous ligand remain unresolved.

Photoaffinity labeling (PAL) methods have proven a useful tool for pharmacological research with utility for first identifying targets of labeled ligands, and thereafter for probing the structural basis of drug-receptor interactions (Dormán and Prestwich, 2000). Photoactive probes can be generated by simple modification of native ligands to incorporate photoactivatible groups, such as azides, diazirines, diazocarbonyls or benzophenones (Weber and Beck-Sickinger, 1997), or by coupling the native ligand in its entirety to a more generic photoaffinity labeling module (Lamos et al., 2006). The former strategy maximizes the likelihood that the derivatized probe mimics the native ligand properties, whereas the latter approach provides further customizability through exploitation of additional tags to facilitate identification and further purification. In the context of NAADP signaling, recent structure-activity investigations have shown that the 5-position of the nicotinic ring of NAADP is tolerant to substitution (Jain et al., 2010).

Therefore, incorporation of an azide group at this position (5N3-NAADP) provides a simple strategy for derivatization of a photoactivatible NAADP probe (Jain et al., 2010). Such azido-based photoaffinity probes have previously been successfully applied to study interactions between agonists and different ion channels (Tanabe et al., 1999; Tomizawa et al., 2007, 2009). Here, we utilized the 5N3-NAADP photolabeling strategy with the aim of performing an unbiased characterization of NAADP binding partners within mammalian cells. Although 5N3-NAADP recapitulated the essential properties of NAADP as a Ca2+-mobilizing messenger, we were surprisingly unable to demonstrate direct labeling of either endogenous or overexpressed TPC proteins in several mammalian systems, or in the sea urchin egg homogenate preparation widely used for studying NAADP-evoked Ca2+ signaling. Consequently, we discuss the possibility that accessory components within a larger TPC complex may be responsible for binding NAADP rather than the TPC protein itself.

Materials and Methods

Chemicals and Reagents. NAADP was synthesized by incubating nicotinamide adenine dinucleotide phosphate (NADP, Sigma-Aldrich) with nicotinic acid in the presence of recombinant Aplysia ADP ribosyl cyclase (Aarhus et al., 1995) followed by high- performance liquid chromatography (HPLC) purification. Concentrations of NADP and NAADP were estimated using established methods (Aarhus et al., 1995). 32P-NAADP 32 32 32 and P-5N3-NAADP were prepared from P-nicotinamide adenine dinucleotide ( P- NAD, 800Ci/mmol, Perkin Elmer) using methods described elsewhere (Walseth et al., 2012). NADP was freshly purified by HPLC prior to experimentation to remove contaminating NAADP. Construction of TPC vectors tagged with GFP or myc have been described previously (Brailoiu et al., 2009; Yamaguchi et al., 2011). LAMP1-RFP was purchased from Addgene, and pCMV/myc/ER/GFP (pShooter-ER) was from Invitrogen.

Binding and Ca2+ release assays. Sea urchin (Lytechinus pictus) egg homogenates (25% v/v) were prepared using standard methods (Lee et al., 1993). Homogenates were diluted in intracellular-like medium (KGluIM: 20mM HEPES, 250mM potassium gluconate,

250mM N-methyl D-glucamine, 1mM MgCl2, pH 7.2) to 2.5% (v/v) for radioligand 32 32 binding assay. Homogenates were incubated with P-NAADP or P-5N3-NAADP at concentrations ranging from 3 to 7nM supplemented with indicated concentrations of unlabeled NAADP (90min on ice). Binding reactions (100μl) were terminated by dilution with 1ml HEPES buffer, followed by centrifugation (21,000g for 30 min, 4°C). Following aspiration of supernatant, pellets were washed and radioactivity was determined by liquid scintillation counting.

For the sea urchin Ca2+ release assay, homogenates were diluted to 2.5% (v/v) with KGluIM buffer supplemented with an ATP regeneration system (1mM MgATP, 10mM creatine phosphate, 10units/ml creatine phosphokinase), and 3μM fluo-3 for fluorimetric detection of free Ca2+, as described previously (Genazzani et al., 1996). Calcium imaging and microinjection of SKBR3 cells were performed using methods described in (Brailoiu et al., 2009).

Cell culture and sample preparation. SKBR3 (human breast carcinoma) and HEK293 (human embryonic kidney) cells were from the American Type Culture Collection (ATCC). SKBR3 cells were cultured in McCoy‟s 5A medium, and HEK293 cells in minimal essential medium (MEM). All media were supplemented with 10% fetal bovine serum (FBS), penicillin (100units/ml), streptomycin (100μg/ml), and L-glutamine (290μg/ml) (Invitrogen). Whole cell lysate (WCL) was prepared by resuspension of trypsinized cells in HEPES buffer (20mM HEPES, protease inhibitors (Roche), pH 7.3) prior to sonication on ice. The sonicated sample was then centrifuged (1,000g for 10 min, 4°C) and the pellet discarded. Subcellular fractionation of the WCL was performed by dual centrifugation steps. Supernatant from a second spin (10,000g for 20 min, 4°C) was collected and respun (100,000g for 1hr, 4°C) to yield supernatant (S100) and pellet (P100) fractions. The P100 pellet was washed and resuspended in HEPES buffer, prior to further sonication. For TPC overexpression experiments, SKBR3 cells were transfected (70-80% confluence) using LipofectamineTM 2000 reagent (Invitrogen) and harvested 48 hours later. Mouse pancreas (Pel-Freez Biologicals) was mixed with HEPES buffer (1:4, w/v) and homogenized, followed by centrifugation (4,000g for 10min, 4°C). The supernatant was collected after further centrifugation (17,000g for 30min, 4°C). Fractionation of sea urchin egg homogenates was followed by the same procedures for mammalian cell lines as described above. Protein concentrations were determined by bicinchoninic acid (BCA) protein assay (Thermo Scientific).

TPC1 knockout mice (B6; 129S5-Tpcn1Gt[OST359423]Lex) were generated by Lexicon Genetics using Gene Trap by retroviral insertion of the VICTR37 vector between coding exons 1 and 2 (NM_145853). Founder mice were backcrossed onto C57B16 mice for 3-4 generations before experiments. RT-PCR analysis revealed that the transcript was absent in the homozygous mutant mice. TPC2 knockout mice (B6; 129S5- Tpcn2tm1Lex) were generated by targeted gene disruption by homologous recombination targeting coding exons 18 through 20 (NM_146206). Disruption of the target gene was confirmed by Southern hybridization analysis. Founder mice were backcrossed onto

C57B16 mice for 3-4 generations before experiments. Sample preparation was performed the same way as described above.

32 Photoaffinity labeling. Samples were incubated with P-5N3-NAADP (4.8-9.5nM, final concentration) for 5-90 min on ice prior to photoactivation effected by exposure of samples to UV light on ice for 2 min. PAL samples were then incubated (<15 min) with SDS sample buffer supplemented with 2-mercaptoethanol (10%) to reduce free label. SDS-PAGE was then performed by separating samples on Criterion 12.5% Bis-Tris gels (Bio-Rad) subsequently stained with instant Coomassie Brilliant Blue reagent (Invitrogen), and air-dried over-night between cellophane sheets (Sigma). For resolution of the 32P-signal, air-dried gels were exposed on phosphor screens (Packard Instrument Co.) at room temperature or on extra-sensitive X-ray film (Thermo Scientific) for 24-96 hours at -80°C. Quantification of 32P-signals was performed densitometrically using OptiQuant software (Version 3.0, Packard Instrument Co.). Curve fitting and statistical analyses were performed using GraphPad Prism (GraphPad Software Inc.). Error bars are shown as meanS.E.M. for each experiment. For the combined PAL/Western blotting experiment, radioactive PAL samples were separated by SDS-PAGE (4-12% Bis-Tris NuPAGE gels), were then transferred to a nitrocellulose membrane (0.45μm, Invitrogen) and detected with an anti-myc antibody (1:1000, Santa Cruz) and an anti-mouse secondary antibody conjugated with horse radish peroxidase (1:5000, AbCam). The 32P- signal was exposed on X-ray film as described above.

Protein two-dimensional electrophoresis & immunoprecipitation. PAL reactions were done as described previously, and photolabeled samples were precleared, desalted and concentrated using Perfect-FOCUSTM kit (G-Biosciences). Samples were then separated on a gel strip (pH 3-10 or 4-7; GE Healthcare LifeSciences) overnight, followed by a regular 1D electrophoresis (Criterion 12.5%, IPG+1 Bis-Tris gel; Bio-Rad). Protein gels were stained with instant Coomassie reagent or silver staining kit (Invitrogen). 32P signal was obtained as described previously. Sample preparation and proteomic analysis was performed by the mass spectrometry core facility at the University of Minnesota.

For PAL-IP, photolabeled samples were solubilized with Triton X-100 (1%, final concentration) on ice for an hour. After centrifugation (16,000g for 30 min, 4°C), the supernatant (soluble PAL samples) was precleared by protein G agarose (Roche) and an anti-myc antibody was added to each sample to form a TPC-immunocomplex, which was then precipitated with protein G agarose (3-4 hr, 4°C). After stringent washing, the immunoprecipitated samples and supernatant (“output”) was separated by SDS-PAGE and 32P signal was detected by phosphorimaging. For IP-WB, TPC immunoreactivity was detected by Western blotting.

Results

The syntheses of both probes are illustrated in the schematics in Figure 2.1. The radioactive photoprobe contained an azido-group at the 5-carboxyl position of the 32 nicotinic ring (5N3-NAADP), and the radioactive phosphate moiety ( P) was incorporated at the alpha position to the ribose linked to adenine (Figure 2.2A). The basic principle underpinning the PAL technique is summarized in Figure 2.2B.

32 32 Figure 2.1. Schematics of P-NAADP and P-5N3- NAADP synthesis. (i) 32P-NAD was incubated at room temperature overnight with NAD kinase in the presence of ATP to convert NAD to NADP. (ii) 32P-NADP was purified by HPLC and a base-exchange reaction was carried out after adding in nicotinic acid and recombinant ADP ribosyl cyclase at pH 4.5. (iii) 32P-NADP was incubated with 5N3-nicotinic acid to synthesize photoprobe. 32 32 P-NAADP and P-5N3-NAADP were purified by HPLC, and photoactivation was confirmed by UV exposure followed by thin layer chromatography (TLC).

Azido groups are chemically inert until photoactivated by long wavelength UV light, which generates a nitrene moiety that reacts covalently with local residue(s) in the target proteins. In the absence of UV stimulation, or in the presence of excess non-derivatized

17 probe, labeling is prevented. These alternative outcomes are revealed by resolution of radioactive signal following electrophoretic separation of the PAL reaction (Figure 2.2B).

32 Prior to assessment of the labeling profile of P-5N3-NAADP, it was important to 32 ascertain the pharmacological properties of the derivatized probe ( P-5N3-NAADP). Using the prototypical model for studying NAADP signaling - egg homogenates from sea 32 32 urchin (Lytechinus pictus) - specific binding of P-5N3-NAADP and P-NAADP was displaced by NAADP in a dose-dependent manner (IC50=0.22±0.05nM and 0.42±0.08nM respectively, n=6-9; Figure 2.2C). Increasing concentrations of the cold photoprobe (5N3- NAADP) elicited progressively greater amounts of Ca2+ in a standard fluorimetric assay

(EC50=1.4μM vs. 32.5nM for NAADP; Figure 2.2D). These data demonstrate the retention of binding and Ca2+-release efficacy in the photoprobe. Consistent with data from another sea urchin species (Strongylocentrotus purpuratus, (Walseth et al., 2012)), homogenates subjected to the PAL reaction yielded labeling of a single, low molecular weight band (41±2kDa, n=3; Figure 2.2E). As shown in Figure 2.2E, labeling was completely displaced by NAADP (5μM), but not by NADP, NAD, or nicotinic acid adenine dinucleotide (NAAD) at the same concentration. Selectivity for NAADP versus NADP was high, with little observed displacement by NADP over a wide concentration range, while incubation with NAADP inhibited PAL of the 41kDa band with an IC50 of 0.66±0.23nM (Figure 2.2F&G, n=4). Finally, comparison of the distribution of this PAL labeled 41kDa band (Figure 2.2H) with the distribution of 32P-NAADP binding in different membrane fractions prepared from egg homogenates (Figure 1B in (Ruas et al., 2010)), showed remarkable congruence. For PAL, relative densitometry showed highest labeling in the S10P100 fraction („light membranes‟), more than 2-fold greater than P10 or P100 fractions which closely mimicked results from 32P-NAADP binding analyses (Ruas et al., 2010). Therefore, the labeling characteristics of the photoaffinity probe recapitulate the basic binding, activity and distribution characteristics of NAADP- receptors in sea urchin egg homogenates.

Figure 2.2. Validation of the photoaffinity labeling (PAL) method. (A) Chemical structure of 32 32 radioactive photoprobe ( P-5N3-NAADP, right) compared with P-NAADP (left). (B) Principle of photoaffinity labeling. In the absence of UV light, the probe reversibly binds target(s). UV exposure crosslinks the radioactive photoprobe to cellular binding partner(s), which are subsequently characterized by SDS-PAGE. NAADP specificity is demonstrated by protection from labeling in the presence of excess 32 32 unlabelled NAADP. (C) Displacement of P-NAADP (open) and P-5N3-NAADP (solid) binding in L. 2+ pictus egg homogenates by unlabeled NAADP. (D) Activation of Ca release by 5N3-NAADP (100nM~31.6μM) in L. pictus egg homogenates. (E) PAL in L. pictus egg homogenates revealed a single band at ~41kDa. Incubation of PAL reaction mixture with NAADP (5μM) protected labeling; whereas other analogues (NADP, NAD, and NAAD, 5μM) did not impact labeling. (F) Effect of indicated concentrations of NADP (top) and NAADP (bottom) on PAL labeling of the 41kDa band. (G) Densitometric quantification (n=4) of NAADP displacement data shown in (F). (H) PAL in subcellular fractions of L. pictus egg homogenates. Top, representative PAL image in crude homogenates (H) and membrane fractions (P10, S10P100, and P100). Bottom, densitometric quantification of similar experiments (n=3). 19

For PAL experiments using mammalian samples, we first employed SKBR3 cells. This cell line, derived from a mammary gland adenocarcinoma, has proved to be a useful cell type for studying and manipulating endogenous NAADP-evoked Ca2+ signals

(Brailoiu et al., 2009). Microinjection of 5N3-NAADP or NAADP (pipette concentration=200μM; estimated concentration in cells=2μM) into single SKBR3 cells elicited a cytoplasmic Ca2+ signal, whereas buffer injection was without effect. These data demonstrated the efficacy of this analog in mammalian systems (Figure 2.3A). To 32 identify NAADP binding partners, the radioactive analog ( P-5N3-NAADP) was incubated in SKBR3 whole cell lysate. After irradiation and electrophoretic separation, multiple labeled bands (Figure 2.3B), spanning a broad molecular weight range (20- 32 250kDa) were revealed. Not all bands labeled by P-5N3-NAADP in SKBR3 cells were displaced in the presence of excess NAADP during the cross-linking reaction. For example, several bands labeled between ~30-70kDa (Figure 2.3B, upper panel) displayed similar labeling intensity despite the presence of NAADP. Quantification of the extent of NAADP protection was performed by densitometry using gels from multiple different samples and PAL reactions, and these results were collated in Figure 2.3B (lower panel). To narrow down the list of candidates for further study, bands were prioritized that showed ≥2-fold reduction in 32P-signal intensity when the PAL reaction was performed in the presence of NAADP (10μM). From this analysis, it was evident that four bands - at ~250kDa (band 1), ~102kDa (band 2), ~27kDa (band 8) and a doublet at 22/23kDa („band‟ 9) - met this criterion. Candidates showing poor NAADP displacement (bands 3- 7) were not considered further.

Figure 2.3. Photoaffinity labeling of SKBR3 whole cell lysate. (A) Microinjection of 5N3-NAADP or NAADP (both ~2μM final) but not buffer alone (control) increases intracellular Ca2+ in SKBR3 cells. (B) Top, representative PAL image in WCL of SKBR3 cells. In the presence of UV light, the intensity of nine bands (~20-250kDa, bands 1-9) was quantified in the absence (lane 2) and presence (lane 3) of excess NAADP. Bottom, quantification of protection of labeling by NAADP, expressed as ratio of PAL intensities (absence/presence of NAADP). Four candidates (*) with estimated molecular weights ~250kDa (band „1‟), 102kDa („2‟), 27kDa („8‟), and a 22/23kDa doublet („9‟) displayed ≥2-fold displacement by NAADP (10μM, n=4). (C) Photolabeling kinetics of the 22/23kDa doublet („band 9‟) as assessed by densitometry. (D) Effects of other compounds (all at 10μM) on PAL in SKBR3. Bands 1, 2, 8, and 9 identified in (A) are highlighted. 21

As the most complete displacement was observed from the 22/23kDa doublet (>4- fold, „band‟ 9) and a kinetic analysis of labeling implied this candidate rapidly bound 32P-

5N3-NAADP (Figure 2.3C), we proceeded to assess selectivity of the 22/23kDa doublet for NAADP relative to other ligands. To achieve this, PAL was performed in SKBR3 whole cell lysates in the presence of a variety of analogs (all at 10μM final concentration; Figure 2.3D). For the 22/23kDa doublet, labeling was not protected by other nucleotide analogs (nicotinic acid; nicotinamide; NAD; NAAD; ATP or β-nicotinamide ribose 2+ monophosphate, β-NMN) or other Ca -mobilizing compounds (cADPR and IP3). However, PAL signal was decreased by the presence of high concentrations of NADP, as has been observed in prior 32P-NAADP binding studies in mammalian samples (Bak et al., 2001; Billington et al., 2006; Gambara et al., 2008; Calcraft et al., 2009) as well as a cell line overexpressing TPC2 (Calcraft et al., 2009). In contrast, pharmacological profiling of the other candidates (bands 1, 2 and 8) were suggestive of properties divergent from those reported for the NAADP receptor: encompassing poor selectivity of NADP versus NAADP (bands 1, 2 & 8), and/or displacement by other ligands (cADPR, band 8). To better substantiate this conclusion, selectivity for NAADP versus NADP was examined for each of the four bands at various ligand concentrations. Figure 2.4A shows a representative image demonstrating competition by NAADP or NADP that was used to fit competition curves. This is illustrated for the 22/23kDa doublet (Figure 2.4B) which displayed an IC50 of 137±17nM for NAADP and 1.5±0.3μM for NADP, respectively

(n=4; Figure 2.4B). Comparison of IC50 ratios (NADP/NAADP) for each of the four candidates in SKBR3 cells underscored that only the low molecular weight doublet („band 9‟) displayed clear selectivity for NAADP over NADP (~11-fold; Figure 2.3C).

Figure 2.4. NADP and NAADP protection of PAL in SKBR3 whole cell lysate. (A) Representative PAL image showing effects of increasing concentrations of NAADP (lanes 2-6) and NADP (lanes 8-12) on PAL. Candidate bands 1, 2, 8, and 9 are marked. (B) Densitometry of the 22/23kDa doublet (band „9‟, n=4) from (A) in the presence of NADP (open square) and NAADP (solid square). (C) Comparison of half maximal inhibitory concentrations (IC50) for NADP (open) and NAADP (solid) for bands 1, 2, 8, and 9 in SKBR3 WCL (n=4).

Cellular distribution of the 22/23kDa doublet was then assessed by subcellular fractionation. Centrifugation of whole cell lysate prior to PAL resulted in the majority (~75%) of PAL-tagged doublet („band 9‟) appearing in the supernatant (S100) compared with the membrane fraction (P100) when the same amount of proteins were used for PAL reactions (Figure 2.5A). Intriguingly, competition analyses revealed that the properties of doublet-associated PAL signal differed between the supernatant (S100) and pellet (P100) fractions. PAL was inhibited by significantly lower concentrations of NAADP in the membrane fraction, compared with either S100 or WCL samples (Figure 2.5B).

Quantification of these data revealed a ~20-fold lower IC50 in the membrane fraction (10±1nM, n=3, Figure 2.5C), compared with supernatant (198±40nM, n=3) used in the same set of experiments. Further, selectivity of the doublet for NAADP (over NADP) was even greater in the membrane fraction (P100, ~52-fold) compared with values measured in WCL (~11-fold) and S100 (~5-fold) samples in the same experiments (Figure 2.5C&D). Therefore, association of the doublet with membranes conferred a higher apparent NAADP binding affinity (~10nM; ~13-fold greater than WCL) and selectivity over NADP (~52-fold; 4.5-fold greater than WCL). In summary, these data suggest that the 22/23kDa doublet identified by PAL in SKBR3 cells displays high affinity and selectivity for NAADP (Figures 2.4 & 2.5), as well as a similar pharmacological displacement profile to that observed for NAADP-sensitive Ca2+ release (Figure 2.3D).

Next, we proceeded to apply the PAL approach in two other cell types, HEK293 cells and mouse pancreas. Both systems have been used as models for studying NAADP- evoked Ca2+ signals (Cancela et al., 1999, 2000; Calcraft et al., 2009; Zong et al., 2009; Brailoiu et al., 2010a; Ruas et al., 2010; Schieder et al., 2010). PAL samples from HEK WCL displayed a broadly similar labeling pattern to those observed from SKBR3 cells (Figure 2.6A). The same nomenclature (bands 1-9) was therefore used for similar sized bands, and additional PAL targets resolved in HEK293 cells were assigned in sequence (bands 10-12). Analysis of NAADP displacement yielded six candidate bands with

24 displacement 2-fold (Figure 2.6B). Four of these corresponded to those in SKBR3 cells (bands 1, 2, 8, 9) and two were novel (11 & 12).

Figure 2.5. Properties of the 22/23kDa doublet in different SKBR3 subcellular fractions. (A) Representative PAL image comparing SKBR3 WCL, soluble (S100) and membrane fractions (P100). (B) Competition PAL images of the 22/23kDa doublet (band „9‟) in indicated fractions. Higher amounts of protein were loaded for the P100 fraction to facilitate visualization of the 22/23kDa doublet. (C) Densitometry (n=4) of the 22/23kDa doublet (band „9‟) from WCL (square), S100 (open circle) and P100 (filled circle) fractions. (D) Comparison of IC50 values for NADP (open) and NAADP (solid) for band „9‟ in different fractions isolated from SKBR3 cells (n=4). 25

Figure 2.6. PAL in HEK293 cells and mouse pancreas. (A) Representative PAL gel in HEK293 WCL. Band numbering was kept consistent with similar sized bands observed in SKBR3 cells. (B) Ratio of densitometry of indicated bands („1‟-„12‟) from multiple independent experiments. Six candidates (*) with estimated molecular weights ~250kDa (band '1'), 102kDa ('2'), 27kDa ('8'), 23kDa doublet ('9'), 36kDa ('11'), and 35kDa ('12') displayed ≥2-fold displacement by NAADP (10μM, n=4). (C) Measurements of IC50 values for bands that displayed ≥2-fold protection by NAADP. (D-F) Similar processing for mouse pancreatic samples showing representative gel (D), NAADP displacement ratio (E), and IC50 estimation (F). 26

Competition curves comparing labeling in the presence of varying concentrations of NAADP and NADP were then performed for all these candidates and the quantified data (n=3 independent experiments) are shown in Figure 2.6C. These data revealed the highest affinity and highest selectivity for NAADP was again associated with band 9 (22/23kDa,

IC50 for NAADP = 52±2nM, versus 818±106nM for NADP using WCL, n=3). Moreover, similar to data from SKBR3 cells (Figure 2.5C), IC50 values in membrane (P100) fractions from HEK cells were indicative of a higher apparent affinity of the 22/23kDa candidate when membrane associated (IC50 for NAADP = 22±4nM vs. 134±23nM in P100 vs. S100 fractions, respectively; n=3).

As mouse pancreatic acinar or  cells have been widely used to study endogenous NAADP-evoked Ca2+ signaling (Cancela et al., 1999, 2000), we performed PAL in homogenized samples from mouse pancreas. NAADP displacement identified 6 candidates at low molecular weights (<50kDa) in the pancreatic samples with little obvious labeling of larger proteins (Figure 2.6D). As this labeling pattern was different from that observed in human cell lines, we employed a different nomenclature for the pancreatic candidates (bands A-F). Quantification of displacement by NAADP revealed that four PAL bands showed significant protection by NAADP, well in excess of the 2- fold criterion established in either human cell line (Bands C-F, ~ 6 to 30-fold displacement; Figure 2.6E). Assessment of the selectivity for NAADP over NADP revealed that band „F‟ (23kDa) and band „D‟ (27kDa) displayed the highest selectivity for NAADP over NADP (9.3±2.3-fold and 7.6±1.8-fold for „F‟ and „D‟ respectively, n=4; Figure 2.6F). Therefore in each of three different mammalian samples studied (SKBR3, HEK293 and mouse pancreas), the 22/23kDa doublet consistently displayed the highest apparent affinity and selectivity for NAADP.

The sizes of the NAADP-selective candidates in mammalian cells (22 & 23kDa) and in sea urchin (41kDa) were considerably lower than those expected for the two pore channel (TPC) proteins identified as NAADP-sensitive Ca2+ channels (Brailoiu et al., 2010a; Hooper et al., 2011; Yamaguchi et al., 2011). Two possible explanations for this discrepancy are (i) endogenous levels of TPCs are below the detection capability of the 27

PAL approach, or (ii) the low molecular weight candidates are actually degradation products of the full length TPC protein. These possibilities were discounted by examining the effects of TPC overexpression, which is a straightforward assay in the mammalian (but not sea urchin) system. Given the sensitivity of detection of Western blot methods is lower than that of 32P-based PAL methods (typically pico- vs. femto- gram sensitivity), immunological detection of overexpressed TPCs should obviate concerns that an insufficient amount of TPC existed in the PAL samples and whether the full length protein was present. Further, increased levels of TPC should cause an increase in PAL reactivity if NAADP is binding directly to these channels.

Therefore we overexpressed GFP-tagged human TPC1 and TPC2 in SKBR3 cells. Both constructs were observed to target to intracellular structures (Figure 2.7A). These data confirm previous reports that TPC constructs are faithfully targeted in the SKBR3 system (Brailoiu et al., 2009). Next we expressed myc-tagged TPC1 and TPC2 constructs with a view to performing Western blotting and phosphorimaging on the same set of PAL samples separated on a SDS-PAGE gel and transferred to a single membrane. The resulting PAL reaction (lower resolution owing to membrane transfer) and Western blot are shown in Figure 2.7B. These data revealed that TPC reactivity was localized as expected at sizes consistent with full length/glycosylated TPC1 (~93 & 130kDa) and TPC2 (~69, 107 & 229kDa), and control myc-GFP samples at <30kDa. However, overexpression of TPCs did not change the overall PAL labeling pattern and specifically the intensity of the 22/23kDa bands. Comparison of the PAL and Western blot data confirm that although full length TPC1 and TPC2 are clearly detectable in these samples, their migration does not correlate well with the 22/23kDa doublet.

PAL of the same TPC overexpressed samples was then performed at higher resolution to allow densitometric quantification (Figure 2.7C). Again, overexpression of myc-tagged TPC1 or TPC2 resulted in very similar PAL patterns, with no obvious appearance of novel candidates within the size range expected for these channels in either the presence or absence of NAADP. Further, PAL did not change the migration pattern of TPCs, as judged by comparison with non-irradiated samples (data not shown). 28

Densitometry of the 22/23kDa doublet in the overexpressed samples showed no increase in both TPC1-myc (103±4%, n=4) and TPC2-myc (104±6%, n=4) whole cell lysates compared to myc-GFP control (Figure 2.7D). Given the PAL approach has higher sensitivity than Western blotting, it is unlikely that the failure to demonstrate labeling of TPCs results from insufficient TPC levels in these samples. Rather, we conclude labeling 32 with P-5N3-NAADP does not correlate with TPC immunoreactivity in samples where the presence of full length TPC proteins is readily demonstrable.

Figure 2.7. Lack of effect of overexpressed TPC isoforms on PAL profiles. (A) Images of SKBR3 cells transfected with TPC1-GFP (upper left) or TPC2-GFP (upper right), and in live cells co-expressing LAMP1-mRFP (middle panel). Bottom, merged images for GFP and RFP. (B) Combined PAL/Western blot, allowing comparison of PAL reactivity (left) with immunological detection of TPC isoforms (right) on the same nitrocellulose membrane. Control cells were transfected with a myc-GFP construct. (C) Representative PAL gel comparing reactivity in control WCL (transfected with myc-GFP) with TPC1-myc (middle two lanes) and TPC2-myc (right two lanes) transfected cells. Overexpression of TPC isoforms did not impact the PAL labeling profile in the absence or presence of NAADP. (D) Quantification of levels of the 22/23kDa doublet in TPC1-myc and TPC2-myc transfected SKBR3 cells relative to controls (n=4 independent transfections). 29

Next, we examined PAL patterns in pancreatic samples derived from TPC1 and TPC2 knockout mice. The availability of these samples afforded the opportunity to examine PAL labeling patterns following ablation of endogenous TPC isoforms. Interestingly, the 22/23kDa doublet was present in all the different pancreatic samples (Figure 2.8), despite the targeting of TPC isoforms for ablation (TPC+/+ matched -/- +/+ -/- littermate versus TPC1 ; TPC2 versus TPC2 ). Further, the IC50s for displacement of

PAL reactivity by cold NAADP were similar for TPC1 (IC50 for NAADP of 483nM in +/+ -/- TPC versus 339 in TPC1 , Figure 2.8A&B), and TPC2 knockout samples (IC50 for NAADP of 268nM in TPC2+/+ versus 245nM in TPC2-/-, Figure 2.8C&D). These data demonstrate the 22/23kDa NAADP-sensitive doublet is present in TPC knockout samples, and must represent a unique protein identity distinct from either TPC isoform.

Figure 2.8. PAL in TPC knockout mouse pancreas. (A & C) Representative PAL profile for the 22/23kDa doublet compared between TPC1+/+ versus TPC1-/- and TPC2+/+ versus TPC2-/- WCL samples. (B & D) Quantification of PAL data to estimate IC50 in similar samples (n=3) shown in (A) and (C).

32 The crucial question remains: what are the protein identities of the P-5N3-NAADP labeled 22/23kDa doublet? We attempted two strategies to characterize the doublet using SKBR3 membranes or WCL. First, the PAL reaction mixture was precleared and protein samples were separated based on their isoelectric points and molecular weights. A representative protein two-dimensional (2D) gel is shown in Figure 2.9, and the 32P signals (lower panel) corresponding to the doublet are marked (asterisk). However, we could barely triangulate 32P signals with visible protein spots resolved by Coomassie or silver staining (dashed rectangles) suggesting that the amount of the doublet in the separated electrophoretic sample was too low for robust identification by mass spectrometry (<0.1ng). Nonetheless, the gel pieces underneath the 32P signals for the doublet were excised and peptide identification was attempted by LC/MS/MS. Multiple proteins were identified and raw data were screened based on three filtering criteria: (i) common contamination and reverse sequences were eliminated, (ii) MW of candidates ranged between 21 and 25kDa, and (iii) at least one unique peptide per protein was identified in both replicates, or at least three unique peptides per protein were identified in either replicate. Based on these criteria, twenty-two protein candidates were selected as listed in Table 2.1.

Second, as PAL coupled with immunoprecipitation (PAL-IP) strategy had been shown to be feasible to demonstrate association of photolabeled NAADP targets with TPCs in sea urchin (Walseth et al., 2012), we adopted similar methodology in SKBR3 cells overexpressing either TPC1/2-myc or control myc-GFP. The goal was to further enrich the doublet by immunoprecipitation for subsequent protein identification by MS. After PAL reaction, the photolabeled samples were solubilized (1% Triton X-100), and the TPC-immunocomplex was pulled down by anti-myc antibody. An IP-WB (omitting PAL) experiment of the same WCL samples was performed in parallel (Figure 2.10A) to estimate IP efficiency. As shown in Figure 2.10B, all the photolabeled protein bands were resolved in the supernatant (lanes 7-9, “output”) rather than in the IP pellets (lanes 4-6), whereas under similar condition, TPC1/2-myc and myc-GFP were exclusively detected in the immunoprecipitates (lanes 4-6, Figure 2.10C). Although this result again suggests that

TPCs are not direct targets for NAADP, the failure to demonstrate association of the doublet with TPCs by PAL-IP does not clearly indicate that the doublet is an accessory protein of TPCs, and this will be discussed later in the Discussion section.

Figure 2.9. Combinatorial PAL-2D-MS in SKBR3 membrane preparation. Representative gel image by Coomassie staining (upper panel) and phosphorimaging for 32P signals (lower panel) were shown. Protein spots on the 2D gel that corresponded to bands 1, 2, 8, and 9 on 1D gel were labeled. The 22/23kD spots (asterisk) were cut out and proteins identified by mass spectrometry were listed in Table 2.1. Dashed rectangle: cropped and enlarged gel areas; dashed circles: spots showing no visible overlapping of protein staining and 32P signal for the doublet. 32

Table 2.1. Protein candidates for the 22/23kDa doublet identified by PAL-2D-MS. Protein hits (22 in total) passed the screening criteria were listed including accession numbers from NCBI and theoretical molecular mass. Two independent replicates (Run1 & 2) and peptide numbers per protein identified in each repeat were shown.

# Identified Proteins (22) Acc. No. MW Run1 Run2 1 28 kDa heat- and acid-stable phosphoprotein Q13442 21 kDa 3 1 2 60S ribosomal protein L10 P27635 25 kDa 0 5 3 60S ribosomal protein L13a P40429 24 kDa 1 2 4 60S ribosomal protein L15 P61313 24 kDa 1 1 5 60S ribosomal protein L19 P84098 23 kDa 2 3 6 BAG family molecular chaperone regulator 2 O95816 24 kDa 0 9 7 COMM domain-containing protein 5 Q9GZQ3 25 kDa 1 5 8 Growth factor receptor-bound protein 2 P62993 25 kDa 0 3 9 GTP-binding nuclear protein Ran P62826 24 kDa 4 7 10 Heat shock protein beta-1 P04792 23 kDa 9 0 11 Hypoxanthine-guanine phosphoribosyltransferase P00492 25 kDa 1 9 12 Isoform 2 of Acyl-protein thioesterase 1 O75608 23 kDa 2 2 Isoform 2 of Protein-L-isoaspartate(D-aspartate) O- 13 methyltransferase P22061 25 kDa 5 8 14 Mps one binder kinase activator-like 1A Q7L9L4 25 kDa 2 3 15 Neighbor of COX4 O43402 24 kDa 1 2 16 Peroxiredoxin-6 P30041 25 kDa 1 4 17 Prefoldin subunit 3 P61758 23 kDa 2 4 18 Proteasome subunit beta type-3 P49720 23 kDa 1 6 19 Ras-related protein Rab-11A P62491 24 kDa 2 7 20 Ras-related protein Rab-14 P61106 24 kDa 1 6 21 Ras-related protein Rab-2A P61019 24 kDa 0 3 22 Ras-related protein Rab-7a P51149 23 kDa 0 4

Figure 2.10. Combinatorial PAL-IP in SKBR3 WCL. (A) Schematic diagram of the experimental design. (B) Phosphorimaging of the 32P signals showed that under this experimental condition, none of the photolabeled bands were co-immunoprecipitated with TPC-myc (lanes 4-6). (C) A parallel IP-WB revealed that under similar condition, TPC-myc and control (myc-GFP) were completely pulled down by anti-myc Ab. Green asterisk: myc-GFP; blue asterisk: TPC1-myc; red asterisk: TPC2-myc; HC: heavy chain; LC: light chain. 34

Discussion

The two-pore channels (TPC1 & TPC2 in humans) have recently been identified as a family of endolysosomal ion channels that release Ca2+ from acidic organelles in response to NAADP (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009). However, there is currently no insight as to how NAADP activates these channels at the structural level. Therefore, the original rationale for this work was to utilize a photolabeling method to identify which region(s) of the TPC proteins bind to NAADP, guided by similar studies that have successfully mapped regions/residues that coordinate ligand-ion channel interactions (Tanabe et al., 1999; Tomizawa et al., 2007, 2009). However, rather than labeling candidates with sizes equivalent to TPC1/2, results from all the systems examined in this study converged in identifying a lower molecular weight candidate (22/23kDa doublet in mammalian systems and a single 41kDa band in sea urchin) as an NAADP target(s) that displayed characteristics compatible with NAADP- evoked Ca2+ release. In each of the three mammalian systems examined, this labeled doublet displayed: (i) the highest affinity for NAADP, (ii) the greatest selectivity for NAADP over NADP, and (iii) a minority fraction of total PAL labeling as expected for a receptor (properties summarized in Table 2.2).

Table 2.2. Photolabeling characteristics of 22/23kDa doublet in three different mammalian samples. Comparison of PAL reactivity in WCL samples between the two human cell lines and mouse pancreas in terms of (i) fraction of total sample labeling (left), (ii) extent of protection by NAADP (10μM, middle) and (iii) selectivity of labeling (IC50(NADP) / IC50(NAADP), right). Doublet Labeling Inhibition Selectivity (% total) by NAADP over NAADP

HEK293 3 ± 0.2 % 2 ± 0.2-fold ~15-fold SKBR3 6 ± 0.4 % 4 ± 0.7-fold ~10-fold mouse pancreas 10 ± 0.6 % 30 ± 2.3 -fold ~8-fold

In SKBR3 cells, the pharmacology of this candidate mimicked known properties of NAADP-sensitive Ca2+ release and showed the highest degree of protection by NAADP 35 of any labeled proteins (Figure 2.3). Further selectivity for NAADP over NADP was ~52-fold in the SKBR3 membrane fraction (Figure 2.5), a ratio which compared favorably to published selectivity ratios from 32P-NAADP binding assays (e.g. 7 to 49- fold range (Bak et al., 2001; Billington et al., 2006; Gambara et al., 2008)), and was taken as representative of the NAADP-evoked Ca2+ release mechanism. Finally, comparison of PAL results between the three mammalian samples (Table 2.2) revealed the low- molecular weight doublet comprised a greater fraction of total labeling, and exhibited greater protection by NAADP in mouse pancreatic samples. This correlates with the frequent usage of this system for studying endogenous NAADP-evoked Ca2+ signals, compared with cell lines where responses to microinjected NAADP are less easily demonstrated. Therefore, the failure to demonstrate labeling of endogenous TPC proteins in different mammalian or sea urchin samples, even when TPC1 and TPC2 isoforms were overexpressed was surprising and merits critical reevaluation of: (i) the photoaffinity approach itself, and (ii) the issue of whether NAADP-evoked Ca2+ release occurs via TPC directly, or rather by binding accessory proteins within a larger TPC-containing complex.

First, while the photolabeling approach provided an unbiased sampling of cellular binding partners for NAADP, results must be interpreted cautiously. Methodological concerns include the specificity of the modified photoprobe relative to the endogenous ligand, the propensity for non-specific incorporation and preferential labeling of low- affinity, high capacity sites. In our hands, specificity of the photoprobe was demonstrated by high affinity binding and efficacy in sea urchin egg homogenate and SKBR3 cells. Further, the observed characteristics of doublet labeling (minority fraction of total incorporation, high affinity and extent of protection, appropriate pharmacology) are not consistent with those expected for a low affinity, non-specific site. Target abundance issues also seem unlikely given overexpression of TPC isoforms did not result in the appearance of a size-matched PAL band despite evident immunoreactivity.

How inconsistent are our data with published findings regarding NAADP binding to TPC proteins? Calcraft et al. reported that TPC2 overexpression in HEK293 cells 36 enhanced specific 32P-NAADP binding (Calcraft et al., 2009). However, it was noted that conditions that resulted in a >250-fold increase in TPC2 mRNA conferred only a ~3-fold increase in specific 32P-NAADP binding (Calcraft et al., 2009). Subsequent immunoprecipitation studies demonstrated that pull down of sea urchin TPCs recovered 32P-NAADP binding sites (Ruas et al., 2010) but neither the efficiency of recovery nor the purity of the preparation was assessed. Similarly, the potential association of accessory proteins in electrophysiological assessments of TPC function cannot be excluded (Brailoiu et al., 2010a; Pitt et al., 2010; Schieder et al., 2010). Therefore, in our opinion, currently published data do not exclude the possibility that the NAADP receptor and NAADP-gated channel (TPC) are separate molecular entities such that NAADP binds to an accessory protein within a larger TPC channel complex. Indeed, TPCs display sequence similarity to voltage-sensitive calcium and sodium channels that assemble as multi-protein complexes encompassing tightly associated subunits (Hooper and Patel, 2011). Discrimination of these alternatives would be facilitated by either direct demonstration of 32P-NAADP binding to purified TPC proteins (no mapping of NAADP binding site(s) on the TPC channels have yet been reported) or the identification of alternate NAADP binding partners (with subsequent proof of interaction with TPCs and conferral of NAADP binding ability).

In this regard, we note the compatibility of photoaffinity labeling tools with mass spectrometry methods as a logical strategy for characterizing NAADP targets. In this study, we combined PAL with either 2D-MS (Figure 2.9 & Table 2.1) or IP (Figure 2.10) to facilitate identification of the doublet. PAL-2D-MS yielded several proteins between 21 and 25kDa, even though visible protein spots were barely seen underneath the 32P signal. Laborious validation (NAADP binding, Ca2+ release assay, PAL, and interaction with TPCs) would be needed to carefully validate these PAL candidates. More sensibly, development of a bifunctional PAL probe (Lamos et al., 2006; Robinette et al., 2008) would be able to enrich NAADP targets by affinity chromatography prior to MS protein identification. Such a bifunctional probe perhaps should be first applied in the sea urchin system owing to its several advantages (single NAADP target, high abundance in egg

37 homogenates, and higher affinity to NAADP) over all the mammalian systems examined to date. The mammalian counterpart could then be identified by homology comparison, and validated by molecular biological and pharmacological studies.

The PAL-IP-MS strategy was attempted based on an assumption that a portion of the 22/23kDa doublet was associated with TPCs on the membrane regardless of NAADP binding. As commercial anti-TPC antibodies performed poorly in IP and Western blotting, and the endogenous levels for TPCs were generally below the detection limit, we took advantage of a well-established biochemical approach by overexpressing myc-tagged TPCs followed by myc IP, initially aimed for enriching the doublet in the TPC-myc immunocomplex for MS studies. In this regard, we have applied PAL-IP strategy on SKBR3 WCL (Figure 2.10) or membranes (in both SKBR3 and HEK293 cells; data not shown) with the initial protein amount ranging from 0.5-5 mg for PAL reaction. We have also tried different detergents (CHAPS, Triton X-100, and NP-40) with alternating combinations and concentrations (1-5%) to solubilized PAL protein samples prior to the IP reaction. However, these attempts all failed to identify the doublet within the TPC- immunocomplex. Obviously, the efficiency of co-immunoprecipitation in solubilized PAL samples and the percentage of TPC-myc recovered in the IP pellets needed to be assessed. Nonetheless, the failure to demonstrate association of TPC1/2-myc and the 22/23kDa doublet was possibly due to (i) the interaction of the doublet with TPCs may be very dynamic and regulated by NAADP, therefore, after 90 min of PAL incubation, the interaction may no longer be detectable, and (ii) the percentage of the doublet associated with TPCs is very low. We noticed that even in the sea urchin system, where the PAL labels were much higher, there were only ~5% of total photolabeled proteins being detected in the IP pellets (Walseth et al., 2012). Therefore, it will be necessary to improve or reevaluate this strategy for doublet identification in both urchin and mammalian systems. Finally, a direct IP-MS method (omitting PAL procedure; detailed in Chapter 3), affinity purification, or yeast two-hybrid system can be alternative approaches to identify TPC-interacting proteins (TPC-interactome) in solubilized WCL or membranes. These approaches have proven to be feasible to identify accessory proteins for a variety of

38 calcium channels - such as ryanodine receptors (Zissimopoulos et al., 2006), IP3R (Maximov et al., 2003), voltage-operated calcium channels (Müller et al., 2010), ligand- gated ionotropic glutamate receptors (Santos et al., 2010), nicotinic acetylcholine receptors (Paulo et al., 2010), transient receptor potential cation channel (TRPV6, (Schoeber et al., 2006)), and calcium release-activated calcium channel (Orai1, Krapivinsky et al., 2011). Further validation of TPC interaction and NAADP pharmacology will be necessary to narrow down protein candidates.

In conclusion, our data imply the existence of a NAADP binding partner displaying properties diagnostic of the NAADP-evoked Ca2+ release mechanism that is unique from the core TPC channel itself. Ongoing work is directed at identifying and validating these candidates with approaches proposed above.

Chapter Three

Elucidation of the TPC-interactome in mammalian cells by an immunoprecipitation-mass spectrometry strategy

Yaping Lin-Moshier, Peter S. Sebastian, Jonathan Marchant

Re-evaluation of the role of calcium homeostasis endoplasmic reticulum protein (CHERP) in cellular senescence signaling. (manuscript in preparation)

Introduction

Two-pore channels (TPCs) were recently identified as calcium-permeable ion channels localized on the membrane of the endolysosomal system (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009). A multitude of evidence has demonstrated that TPC-mediated calcium signals from acidic calcium stores are involved in a variety of physiological phenomena, including smooth muscle contraction, epithelial cell function, blood pressure control, muscle and neuronal differentiation, endolysosomal trafficking and autophagy (Brailoiu et al., 2006, 2010b; Aley et al., 2010; Ruas et al., 2010; Tugba Durlu-Kandilci et al., 2010; Esposito et al., 2011; Pereira et al., 2011). Further, manipulation of TPC protein levels by overexpression or knockdown, or altering TPC function by mutagenesis also changed cellular sensitivity to NAADP (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009; Dionisio et al., 2011; Pereira et al., 2011). All these findings support TPCs as putative NAADP-gated calcium channels in the endolysosomal system. Our previous studies utilizing a radioactive, photoactivable 32 NAADP homologue ( P-5N3-NAADP) have however suggested that TPCs were not direct NAADP targets; rather, an accessory protein(s) in the TPC signaling complex was responsible for NAADP binding in both sea urchin and mammalian systems ((Lin- Moshier et al., 2012; Walseth et al., 2012); also see Chapter 2). These studies suggest that the TPC protein functions in the context of a larger protein complex, just as with many other ion channels, including voltage-operated calcium channels (Arikkath and Campbell,

2003), Kv2.1 potassium channels (Peltola et al., 2011), ionotropic glutamate receptors (Yan and Tomita, 2011), Na+/K+-ATPase (Yoshimura et al., 2008), and voltage-gated Na+ channels (Yu et al., 2003). However, little is known about the larger TPC- interactome; therefore, in this chapter, we sought to apply a parallel approach to elucidate the TPC-interactome. We used antibody-based immunoprecitation coupled with mass spectrometry methods, aiming to identify TPC-interacting proteins in mammalian cells as a first step to characterize the TPC protein complex.

Expression of epitope-tagged TPCs has been shown to be feasible to study TPC- mediated calcium responses or Ca2+ currents in intact cells or isolated lysosomes in the 41 presence of NAADP (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009; Pitt et al., 2010; Schieder et al., 2010). These findings suggested that the epitope tagging on the carboxyl terminus of TPC proteins did not hinder TPC channel activity, therefore providing an opportunity for elucidating the TPC-signaling complex by antibody-based purification in cell lines overexpressing TPCs. Recently, co-expression and immunoprecipitation coupled with Western blotting (IP-WB) of epitope-tagged TPCs in mammalian cells have revealed a mechanism for oligomerization in TPC proteins (Zong et al., 2009; Churamani et al., 2011; Rietdorf et al., 2011), as well as their interaction with other calcium channels, TRPMLs, on the late endosomes/lysosomes (Yamaguchi et al., 2011). While these traditional reciprocal IP-WB methods can demonstrate interaction of TPCs with specific proteins, they are obviously inadequate to perform an unbiased and comprehensive characterization of the TPC-interactome. We therefore applied an IP-MS approach instead, to investigate the composition of the TPC-interactome. IP-MS methods are oftentimes referred to as the “gold standard” proteomic technique for examination of components in multiple protein complexes and have been successfully used to identify associated proteins for several calcium-permeable membrane receptors, including NMDARs, AMPARs, and VOCCs (Husi et al., 2000; Collins et al., 2006; Müller et al., 2010; Santos et al., 2010). Here, we utilized the same IP-MS strategy in mammalian cells overexpressing epitope-tagged TPCs with the goal of purification and characterization of TPC signaling complexes, and possibly identification of the putative NAADP receptor (the 22/23kDa doublet as revealed in the PAL study; see Chapter 2). Our results showed that: (i) TPCs were successfully identified in the immunoprecipitates by the IP-MS approach, and (ii) multiple protein candidates were detected in TPC immunoprecipitates that were absent from controls. However, careful validation of one of these candidates- CHERP (calcium homeostasis endoplasmic reticulum protein) - revealed that it was likely a false positive result. Much of the remaining chapter is concerned with reevaluation of the cellular role of CHERP, proposed to be an integral potassium channel subunit on the ER membrane that binds to the ryanodine receptor (LePlante et al., 2000; Ryan et al., 2011). While this would provide an exciting structural linkage between TPCs and RyRs in the context of triggering Ca2+ signals, our data suggest the role of CHERP as 42 proposed in the literature to date is incorrect. Consequently, we conclude by discussing the feasibility of one-step IP-MS strategy in investigation of TPC-interactome and propose other affinity purification approaches to better achieve this goal.

Materials and Methods

Constructs, reagents, antibodies & chemicals. GFP- and myc-tagged TPC1 and TPC2 were gifts from Dr. Sandip Patel (Department of Cellular & Developmental Biology, University College London, UK). Construction of these plasmids was described in (7). LAMP1-RFP was purchased from Addgene, and pCMV/myc/ER/GFP (pShooter-ER) was from Invitrogen. CHERP cDNA (NM_006387.5) was from OpenBiosystems. EGFP- C3, -N3, and mCherry vectors were from Clontech. In-fusion HD EcoDry PCR cloning kit was from Clontech. Protease inhibitor was from Roche Applied Science. Myc and GFP antibodies were from Santa Cruz. Anti-CHERP (ab15951) and anti-HSP90 were from Abcam. Anti-CREB was from Millipore. Secondary antibodies for Western blotting and immuno-histochemistry were from LI-COR Biosciences and Invitrogen, respectively. All chemicals were from Sigma-Aldrich.

Cell culture, transfection & sample preparation. HEK293, SKBR3 and Jurkat cells were purchased from ATCC. HEK293 were maintained in minimal essential medium (MEM), SKBR in McCoy‟s 5A medium, and Jurkat in RPMI-1640 medium. All media were supplemented with 10% FBS, penicillin (100units/ml), streptomycin (100μg/ml), and L-glutamine (290μg/ml) (Invitrogen).

For overexpression experiments, SKBR3 or HEK293 cells were transfected at 70% confluence using LipofectamineTM 2000 (Invitrogen) or TrueFect transfection reagents (United Biosystems) and harvested 48 hours later. Whole cell lysate (WCL) was prepared by resuspension of trypsinized cells in non-reducing co-immunoprecipitation buffer (137mM NaCl, 2mM EDTA, 20mM Tris-HCl, pH 8.0, 1% NP-40, 10% glycerol, 1Xprotease inhibitors) prior to solubilization for 1h at 4°C with mild agitation. The

43 solubilized sample was centrifuged at 100,000g for 1h at 4°C, and the supernatant was collected and stored at -80°C until use. For the preparation of soluble and membrane fractions, see Materials & Methods section in Chapter Two. For cytosolic and nuclear fractionations, a combination of Buffer I and Buffer IV was used (Buffer I: 10mM HEPES, pH 7.9, 10mM KCl, 0.1mM EDTA, 1mM DTT, 1mM PMSF, 10% glycerol, 0.2% NP-40, 0.15mM spermine, 0.5mM spermidine. Buffer IV: 20mM HEPES, pH 7.9, 0.8M NaCl, 1mM EDTA, 1mM DTT, 1mM PMSF, 10% glycerol. Both were supplemented with 1X protease inhibitor prior to sample preparation). Trypsinized cells were resuspended in 2X volumes of Buffer I and incubated on ice for 10min. The cytosolic fraction was separated by centrifugation at 4000rpm at 4°C for 5min. The pellet was resuspended in 1X volume of Buffer I and Buffer IV mixture (1:1, v/v), followed by incubation at 4°C for 30min on an end-over-end shaker. After centrifugation at 16,000rpm at 4°C for 15min, the nuclear fraction (supernatant) was collected and proteins from both fractions were quantified.

Immunoprecipitation, protein staining & mass spectrometry. Protein concentrations in soluble WCL were determined by BCA assay (Thermo Scientific). One milligram of soluble WCL was precleared with protein G agarose (Roche Applied Science), followed by incubation overnight with anti-myc antibody at 4°C on a shaker. The TPC- immunocomplex was purified by incubating the IP reaction mixture with protein G agarose, followed by washing with non-reducing co-IP buffer, 4X10min, at 4°C. Immunoprecipitated proteins were denatured in 1XLDS sample buffer (Invitrogen) and 100mM DTT at 50°C for 1h. Protein samples were separated on a 12.5% Criterion Bis- HCl gel (Bio-Rad) and protein bands resolved using the SilverQuest Silver Staining Kit (Invitrogen). TPC bands were identified and cut out from the gel, followed by in-gel digestion, purification of peptides and LC-MS-MS. Data analysis was performed using Scaffold software (Proteome Software, Inc).

RNA isolation, RT-PCR & Western blot. Trizol reagent was used to purify total RNA from HEK293 cells by following the manufacturer‟s protocol (Invitrogen). Reverse

44 transcription of 2μg total RNAs from native and transfected HEK293 was done by SuperScript III Reverse Transcriptase (Invitrogen). End-point PCR was performed using Advantage HD polymerase mix (Clontech) in the presence of gene-specific primers (7), followed by separation on a 2% agarose gel in the presence of SYBR DNA gel stain (Invitrogen) for visualization.

Western blotting was performed by separating protein samples on a 4-12% NuPAGE Bis-Tris gel followed by transferring onto a PVDF membrane (0.45μm, Invitrogen). Proteins of interest were detected by primary antibodies listed previously then by species specific secondary antibodies. An Odyssey Infrared Image System (LI- COR Biosciences) was used for resolving and analyzing fluorescence intensity of protein bands.

Construction of wild-type and truncation mutants, immunostaining & live cell imaging. In-fusion HD EcoDry PCR cloning kit was used to subclone CHERP (wild-type vs NH2 or C-terminal deletion constructs) into pEGFP or mCherry vectors following the manufacturer‟s instructions. Primer sequences were listed in Table 3.1. DNA sequences were confirmed by automatic sequencing at the BioMedical Genomics Center from the University of Minnesota.

Immunostaining of endogenous CHERP in HEK293 cells, primary cardiac myocytes, and skeletal muscle slices was preformed following normal procedures. Cells or tissue slices were fixed with 4% paraformaldehyde (in PBS) at room temperature for 30 min. Permeabilization was done by incubating fixed samples in 0.1% Triton X-100/PBS (PBST) for 10 min. Samples were then incubated in blocking solution (1% bovine serum albumin/PBST) at room temperature for 30min before hybridization with or without anti- CHERP antibody at 4°C overnight. Samples were washed the next day and incubated with a secondary antibody for 1hr at room temperature. After washing away excess secondary antibodies, samples were counterstained with DAPI (0.1ug/mL, Roche) for 1min before imaging.

For live cell imaging experiments, a confocal system (Olympus FluoView FV1000) was used to detect the fluorescence of CHERP-GFP. For comparison of the localization of truncation mutants, all images were captured at the same settings (fluorescence intensity, exposure time, and zoom), and the intensity values of GFP fluorescence within similar sized regions-of-interest in the cytoplasm and nucleus were measured using MetaMorph software (Molecular Devices, Inc.). Data was plotted using GraphPad Prism (GraphPad Software Inc.). Error bars are shown as meanS.E.M. for each experiment.

Table 3.1. Primer sequences for subcloning NH2- vs C-terminal deletion constructs of CHERP.

Constructs Forward primer (5’->3’) Reverse primer (5’->3’) Wild-type CHERP1-916 (pEGFP-C3) GAATTCTGCAGTCGACaATGGAGATGCCGCTGCCC GTGGATCCCGGGCCGCGGCTACTTACACTCGTCCCTG CHERP1-916 (pEGFP-N3) GTCAGATCCGCTAGCATGGAGATGCCGCTG CCGCGGTACCGTCGACCTTACACTCGTCCCTGGCC CHERP1-916 (pmCherry) GTCAGATCCGCTAGCaaATGGAGATGCCGCTGCCC CCGCGGTACCGTCGACaCTTACACTCGTCCCTGGCC

C-terminal deletion CHERP1-897 (pEGFP-N3) GTCAGATCCGCTAGCATGGAGATGCCGCTG CCGCGGTACCGTCGACGTAGTTCTCATAGGGGTC CHERP1-838 (pEGFP-N3) GTCAGATCCGCTAGCATGGAGATGCCGCTG CCGCGGTACCGTCGACCCTTGAGTCAGGGATGGG CHERP1-777 (pEGFP-N3) GTCAGATCCGCTAGCATGGAGATGCCGCTG CCGCGGTACCGTCGACGGAGTAGGAACGGGAGCA CHERP1-725 (pEGFP-N3) GTCAGATCCGCTAGCATGGAGATGCCGCTG CCGCGGTACCGTCGACCCGCCGGGCCCGCATTTT CHERP1-705 (pEGFP-N3) GTCAGATCCGCTAGCATGGAGATGCCGCTG CCGCGGTACCGTCGACACTGTTCCTGGGCCTGTC

N-terminal deletion CHERP342-916 (pEGFP-N3) GTCAGATCCGCTAGCATGCTCCAGATGCCGCAGATG GTGGATCCCGGGCCGCGGCTACTTACACTCGTCCCTG CHERP705-916 (pEGFP-N3) GTCAGATCCGCTAGCATGAGTGAAGGCTGGGAGCAG GTGGATCCCGGGCCGCGGCTACTTACACTCGTCCCTG

Italic: vector-specific sequences Underline: restriction sites Lower case: base for in frame adjustment Bold: gene-specific sequences

Results

Overexpression and co-immunoprecipitation coupled with Western blotting of epitope-tagged TPCs in human cell lines have been shown to be feasible to detect interaction of TPC proteins with other calcium channels, including TRPML1 and TPCs themselves (Zong et al., 2009; Rietdorf et al., 2011; Yamaguchi et al., 2011). Therefore, we applied similar strategy in HEK293 cells transiently transfected with TPC1 or TPC2- myc, initially attempting to purify, enrich, and identify TPC-protein complexes by mass spectrometry. After 48 hours of transfection, expression of TPC1 and TPC2 at both mRNA (Figure 3.1A) and protein levels (whole cell lysates, Figure 3.1B) were increased relative to control cells (transfected with a myc-GFP construct). The same samples were then subjected to co-immunoprecipitation procedures in the presence of an anti-myc antibody. Proteins that were co-precipitated with myc-GFP or TPC1/2-myc were separated by electrophoresis, and protein bands were resolved by the silver staining method. As shown in Figure 3.1C, multiple protein bands with molecular mass between 15- and 100kDa were revealed in all IP samples, hinting that the stringency of this IP condition was perhaps too low to differentiate specific TPC-interacting proteins from contaminants. Nonetheless, two protein bands (#1 & 2) in TPC1-myc lane, and one protein band (#3) in TPC2-myc lane were identified that appeared absent in the control sample. Two gel pieces in the control lane (#4 & 5) with similar mobility to bands #1-3 were also excised for MS analysis. For screening, we used the following criteria: (i) at least two unique peptides per protein were identified in bands #1-3, (ii) the same proteins were absent in control samples (bands #4-5), and (iii) were non-structural proteins or components of the transcription or translation machinery. Based on these criteria, TPC1 (in bands #1 & 2) and TPC2 (in band #3) were identified with high confidence. We therefore demonstrated the feasibility of the IP-MS strategy for identification and validation of TPC proteins overexpressing in biological samples as recently reported (Rietdorf et al., 2011). To our surprise, IP-MS from this pilot study also identified an ER- associated protein, CHERP (calcium homeostasis endoplasmic reticulum protein) that was co-precipitated with TPCs (in bands #2 & 3, but absent from bands #1, 4 & 5, Figure

3.1D), suggesting that CHERP might be a candidate component in TPC1/2-myc protein complexes.

Figure 3.1. Identification of TPC-interacting proteins in HEK293 cells. (A) RT-PCR results to confirm over-expression of TPC1-myc and TPC2-myc at mRNA levels. (B) Western blot detection to validate expression of TPC1 and TPC2-myc at protein levels. Blots were hybridized with an anti-myc antibody. (C) A protein gel was stained with the silver staining kit to reveal protein bands of interest. Bands #1 & 2 from TPC1-myc lane, #3 from TPC2-myc lane, and #4 & 5 from control lane were cut out for mass spectrometry analysis. (D) Representative MS/MS spectra and amino acid sequences for bands #1-3. 48

That CHERP was identified in the TPC immunocomplex was potentially exciting owing to the known role of NAADP-evoked Ca2+ responses as trigger calcium signals that are propagated into a larger Ca2+ wave via RyR-mediated Ca2+ release (calcium- induced calcium release; CICR) from the ER stores (Brailoiu et al., 2009; Galione et al., 2010, 2011). However, there is currently no molecular evidence as to how these two intracellular calcium channels structurally interact in the context of NAADP-mediated “channel chatter”. One hypothesis is that there exists a molecular “bridge” that brings these two subcellular compartments closer in space for CICR to occur. Such a molecular bridge may express in one or both organellar compartments. More recently, CHERP was identified as a RyR1 interacting protein by means of interaction proteomics approach (Ryan et al., 2011), hinting that CHERP could be a candidate.

CHERP was initially identified from a human leukemic cell line (HEL cells), although its mRNA is ubiquitously expressed in human tissues. In vitro transcription and translation of the same CHERP cDNA yielded two protein products with estimated molecular mass of 100- and 128kDa (LaPlante et al., 2000). Initial immunocytochemistry studies revealed that CHERP co-localized with the ER and was involved in ER calcium homeostasis potentially through regulation of potassium conductance in the ER (O‟Rourke et al., 1994; LePlante et al., 2000). In silico analysis of functional domains on CHERP predicted several protein-protein interaction modules, including polyglutamine track, proline-rich region, and serine/arginine-rich region at its carboxyl terminus (LaPlante et al., 2000; also see Table 3.2). Such properties rendered CHERP a possible candidate as a molecular scaffold for protein interaction. In our study, that CHERP was identified in TPC1/2-myc immunocomplex led us to hypothesize that CHERP could act as a “molecular bridge” between endolysosomal system and ER through its interaction with TPCs and RyRs.

To examine the validity of this hypothesis, we first examined the expression pattern of endogenous CHERP in subcellular fractionations of HEK293 cells overexpressing TPC1/2-myc. As shown in Figure 3.2A, TPC1/2-myc was detected in the membrane preparation (P100) and was totally absent in the soluble fraction (S100), indicating that 49 our fractionation method was able to recapitulate the localization of TPC proteins on the membranes as reported previously (Brailoiu et al., 2009; Calcraft et al., 2009; Zong et al., 2009; Hooper et al., 2011).

Figure 3.2. Non-specific interaction between TPCs and CHERP in HEK293 cells. (A) Subcellular distribution of TPC-myc (upper panel) and CHERP (middle panel) were characterized by Western blot in soluble (S100) and membrane (P100) preparations from HEK293 cells (mock control vs TPC-myc overexpression). CREB was used as a loading control (lower panel). (B) Whole cell lysates from HEK293 cells co-transfected with TPC1-myc and GFP-CHERP (lanes 1 & 2) or TPC2-myc and GFP-CHERP (lanes 3 & 4) were incubated with anti-myc Ab (lanes 2 & 4) or control IgG (lanes 1 & 3), followed by Western blot detection of CHERP by anti-CHERP Ab. The fact that CHERP was also detected in control IgG lanes revealed the interaction of TPC1/2 and CHERP was non-specific. (C) Reciprocal IP-WB was performed. The experimental design was similar to (B), except a myc-GFP construct was used as a control. Again, that myc-GFP was co-immunoprecipitated with CHERP (lane 1) suggested interaction between TPC1/2 and CHERP was non-specific. Open arrowhead: endogenous CHERP; closed arrowhead: GFP-CHERP. 50

The anti-CHERP antibody detected protein immunoreactivity with a molecular mass of 118.50.4kDa (n=15), about 15kDa larger than the theoretical MW (~103kDa). The majority of CHERP protein was detected in the soluble fraction (~80%, S100) and only ~20% was associated with membrane (P100), regardless of TPC overexpression. The specificity of the commercial CHERP antibody was validated by epitope-tagging and IP- MS methods (see Figure 3.4A-C).

Next, we applied a reciprocal IP-WB approach to attempt to validate the interaction of TPCs and CHERP. To avoid non-specific cross-reaction from the commercial antibodies, we constructed a GFP-tagged CHERP based on the nucleotide reference sequence published on NCBI (NM_006387.5) with the primer sequences as listed in Table 3.1. HEK293 cells were co-transfected with TPC1/2-myc and GFP-CHERP; for negative control, a myc-GFP vector was co-transfected with GFP-CHERP instead. Overexpression was carried out for 48 hours before IP experiments. Whole cell lysates from each transfection combination were incubated with anti-myc or anti-CHERP antibodies to pull down TPC-myc or CHERP protein complexes, respectively. The same WCL samples were incubated with a control IgG to estimate, if any, non-specific interaction from antibodies used for IP. As shown in Figure 3.2B, while TPC1/2-myc was immunoprecipitated by anti-myc Ab and not by the control IgG (lower blot), CHERP- both endogenous (open arrowhead) and exogenous (solid arrowhead) - was detected in all IP reactions (upper blot). Furthermore, in a parallel reciprocal IP-WB experiment, the commercial anti-CHERP antibody not only pulled down TPC1/2-myc, but also precipitated myc-GFP (Figure 3.2C). Therefore, we failed to confirm that the interaction between CHERP (endogenous vs. exogenous) and TPCs was specific using reciprocal IP- WB methods under these experimental conditions. Before we drew the conclusion that the CHERP interaction was a false positive result from the IP-MS experiment, we decided to take advantage of the availability of fluorescent-tagged CHERP, TPCs, and organellar markers in our laboratory to examine whether CHERP and TPCs were expressed in the same subcellular compartments in close vicinity. As CHERP was previously identified as a novel RyR1-interacting protein by metal affinity

51 chromatography-MS methods in HEK293 cells (Ryan et al., 2011), we used this cell line for live cell localization experiments (Figure 3.3).

Figure 3.3. Co-localization of overexpressed CHERP with TPC1/2 or organellar markers. HEK293 cells were seeded on MatTek dishes, and co-transfected with CHERP-mCherry and TPC1-GFP (A), TPC2- GFP (B), or myc-ER-GFP (C). In (D), cells were co-transfected with CHERP-GFP and LAMP1-RFP. All these results revealed that CHERP did not co-localize with TPC1, TPC2, lysosomal marker (LAMP1) or ER marker. (E) CHERP-GFP colocalized with DAPI staining suggesting CHERP is a nuclear protein.

HEK293 cells were co-transfected with CHERP-mCherry and TPC1-GFP (Figure 3.3A), TPC2-GFP (Figure 3.3B) or a GFP vector containing ER retention signals (Figure 52

3.3C). Live confocal imaging revealed that overexpressed CHERP did not co-localize with TPC1 and TPC2 (A & B); further, CHERP did not express in the ER or any subcellular compartment in the close vicinity to the ER (C). When a GFP-CHERP was co-expressed with RFP-LAMP1, a lysosome/late endosome marker, there was no colocalization of CHERP and LAMP1 either. These live cell imaging results all suggest that exogenous CHERP did not express in the endolysosomal system where TPCs resided, and it was unlikely an ER-associated or ER-integral protein.

Rather, the expression pattern of CHERP-mCherry or GFP-CHERP resembled nuclear staining; therefore, we counterstained cells with DAPI to verify CHERP‟s localization. As shown in Figure 3.3E, the expression of GFP-CHERP localized in DAPI- stained nuclei, suggesting CHERP is a nuclear protein in this cell type. Only rarely was GFP- or mCherry-tagged CHERP detected in the cytoplasm (asterisk in Figure 3.3C), which were possibly due to (i) overexpression has overwhelmed a nuclear transport mechanism, (ii) cells were under certain stress or undergoing cell cycle and therefore expelling CHERP from the nucleus or (iii) there was epitope break-down of the fusion protein. CHERP‟s nuclear expression pattern was also confirmed when we swapped GFP to the carboxyl terminus (see later in Figure 3.7B), suggesting the nuclear localization of the fusion protein was not an artifact from the position of epitope tagging.

Did epitope-tagged CHERP accurately represent the localization of endogenous CHERP? To rule out potential artifacts from overexpression experiments, we used immunocytochemistry in HEK293 cells to define the subcellular localization of endogenous CHERP. To verify the CHERP antibody, we first compared Western immunoreactivity of GFP-tagged CHERP, both amino- and carboxyl-terminal tagged, with endogenous CHERP by Western blotting using the commercial antibody (ab15951, Abcam), which was previously used to study endogenous CHERP in HEK293 cells and skeletal muscles (Ryan et al., 2011; also in Figure 3.2). Western blotting using anti- CHERP and anti-GFP antibodies revealed a protein band with a molecular mass of 1430.5kDa (n=9; Figure 3.4A, solid arrowhead), suggesting that the commercial anti- CHERP antibody specifically detected exogenous CHERP. 53

In HEK293 cells overexpressing CHERP-GFP, the same anti-CHERP antibody again detected immunoreactivity at around 118kDa (Figure 3.4A, open arrowhead), which was not recognizable by the anti-GFP Ab. The 118kDa protein band had a molecular mass almost equivalent to that of the GFP moiety (27kDa) subtracted from GFP-CHERP (143kDa), suggesting it was the endogenous CHERP and was able to be faithfully detected by the commercial antibody. To further verify the specificity, we used a direct IP-MS approach in HEK293 whole cell lysate in parallel. As shown in Figure 3.4B, CHERP is detected in precipitates from CHERP IP samples, but not from control IgG (left panel), indicating CHERP antibody recognizes endogenous CHERP by IP-WB, consistent with the data shown in Figure 3.2C. Furthermore, the CHERP protein band was detectable in CHERP IP samples and absent from the control IgG lane (Figure 3.4B, right panel). LC/MS/MS with high-stringency screening criteria confirmed CHERP peptides exclusively in CHERP immunoprecipitates. Therefore, the commercial antibody was confirmed to faithfully recognize CHERP. We noted several splicing factors or accessory proteins to spliceosomes were co-purified with CHERP (Figure 3.4C).

Next, we performed immunocytochemistry in HEK293 cells overexpressing GFP- CHERP to examine whether the antibody was able to recognize native CHERP in fixed cells. As shown in Figure 3.4D (left panel), the immunostaining pattern from CHERP Ab had overlapped perfectly with the GFP fluorescence, and the intensity was proportional to the expression level of GFP-CHERP. Therefore, as both Western blotting and immunocytochemistry studies demonstrated that the commercial CHERP antibody was able to detect exogenous CHERP in both denatured and native structures, we were more confident to use this antibody for Western blotting and immunostaining on endogenous CHERP. HEK293 cells were chosen as the first model system for characterizing the subcellular localization of endogenous CHERP. As shown in Figure 3.4D (right panel), the staining pattern of CHERP completely resembled that of DAPI, suggesting endogenous CHERP, just like its GFP-tagged counterpart, was a nuclear protein. Further, when Western blotting was performed in cytosolic and nuclear fractionations in HEK293 cells, the majority (~90%) of endogenous CHERP immunoreactivity was detected in the

54 nuclear fraction (Figure 3.4E). Our results conflict with published findings that CHERP is potentially an ER-integral protein in HEL cells and rat skeletal muscles (LePlante et al., 2000; O‟Rourke et al., 2003; Ryan et al., 2011).

To examine whether CHERP homologue in rat muscle tissues does have distinct distribution patterns to that of the human counterpart, we performed immunohistochemistry in rat skeletal muscle slices and primary cardiac myocytes (Figure 3.5). Cross-sections of rat gastrocnemius muscle were hybridized with the same commercial anti-CHERP antibody (ab15951, Abcam) as demonstrated previously in Ryan et al., 2011. Here, we included two negative controls- muscle slices omitting anti- CHERP Ab (“-primary Ab” control in Figure 3.5A, lower panel) or eliminating secondary Ab (data not shown) - to estimate non-specific cross-reactivity from primary or secondary antibodies. Tissue slices were counterstained with DAPI to label genomic DNA in the nuclei prior to confocal microscopy. Results were shown in Figure 3.5A. Consistent with the immunocytochemistry studies in human cell lines, most of the CHERP immunoreactivity was restricted in the DAPI-stained nuclei, suggesting CHERP was also a nuclear protein in rat skeletal muscle (Figure 3.5A, top two panels). Rarely, a cytosolic punctuate staining pattern of CHERP was observed that resembled muscle fibers (Figure 3.5A, “Cyt. puncta”); however, the same distribution pattern was also detected when anti-CHERP Ab was omitted (“-primary Ab”, Figure 3.5A), suggesting the cytosolic punctuate was non-specific retention of the secondary Ab. Furthermore, immunocytochemistry analysis in the other muscle type- primary cardiac myocytes- also revealed that CHERP resided in the nucleus (Figure 3.5B, top two panels). Western blotting of CHERP in cytosolic and nuclear fractionations of rat heart tissue confirmed a nuclear distribution pattern of CHERP in cardiac myocytes (Figure 3.5C).

Figure 3.4. Validation of anti-CHERP Ab followed by immunocytochemistry of endogenous CHERP in cell lines. (A) Western blotting of endogenous vs. GFP-tagged CHERP in HEK293 cells revealed a correspondence of the CHERP-GFP band (~1430.5kDa) detected by both anti-CHERP (left) & anti-GFP antibodies (middle). Protein bands were merged on the right panel. Open arrowhead: endogenous CHERP; closed arrowhead: CHERP-GFP; NS: non- specific detection; #: GFP. (B) A representative IP- WB result revealed that endogenous CHERP was detected in CHERP- immunoprecipitates but not pulled down by the control IgG (left). A representative protein gel stained by Coomassie staining method after IP reaction was shown (right). Protein bands (dashed rectangles) equivalent to CHERP electrophoretic mobility were excised for MS analysis. (C) MS results confirmed CHERP was recognized and detected by the antibody; further, multiple proteins associated with or in the U2 snRNP complex were co-purified with CHERP. (D) Colocalization of CHERP-GFP examined by GFP fluorescence (top left) and immunostaining (anti-CHERP Ab, middle left) indicated that the commercial anti-CHERP antibody specifically recognized CHERP. Immunocytochemistry of endogenous CHERP and DAPI revealed CHERP as a nuclear protein in HEK293 cells (right panel). (E) Western blot detection of endogenous CHERP in cytosolic vs. nuclear fractions in HEK293 cells confirmed nuclear distribution of CHERP in this cell type. HSP90 was used as a cytosolic marker, and CREB as a nuclear marker.

C Figure 3.5. Immunohistochemistry of endogenous CHERP in muscles. (A) Immunohistochemistry of endogenous CHERP in rat gastrocnemius muscle slices identified nuclear localization of CHERP in skeletal muscles. Rarely, a cytosolic punctuate staining of CHERP was detected (Cyt. puncta), but this resembled patterns observed when primary antibody (anti-CHERP) was omitted. C: cytosol; n: nucleus. (B) Immunocytochemistry in primary cardiac myocytes identified CHERP as a nuclear protein. (C) Western blot detection of endogenous CHERP in cytosolic vs. nuclear fractions in cardiac myocytes confirmed nuclear distribution of CHERP in this cell type.

So far, all our results demonstrated that CHERP is a nuclear protein, contrasting with published data implicating it as an ER-associated or ER-integral protein (LePlante et al., 2000; O‟Rourke et al., 2003; Ryan et al., 2011). If CHERP is targeted to the nucleus, 57 then it may harbor a nuclear importing sequence/domain, which would further evidence its localization within cells. Therefore, we performed in silico interrogation based on the protein reference sequence on the NCBI (NP_006378), attempting to define CHERP‟s functional domains and potential post-translational modifications. As shown in Table 3.2, most algorithms actually predicted CHERP as a nuclear resident based on features that (i) signal peptides were absent in its amino acid sequence, (ii) putative nuclear localization signals (NLS) were identified in its carboxyl terminal regions, and (iii) multiple mRNA processing motifs (SURP and G patch) and RNA polymerase II interacting domain (CID) were predicted in its N- or C-terminal regions, consistent with nuclear functionality. Other predicted features include: (i) an ambiguous prediction of 0-2 transmembrane domains between different software, and (ii) several types of post-translational modifications (PTMs) spanning the entire protein sequence, including glycosylation, phosphorylation, SUMOylation, and palmitoylation (Table 3.2).

A schematic diagram of functional domains in the CHERP sequence is illustrated in Figure 3.6, with the nuclear localization signals (NLS1, 2/3, & 4), transcription and spliceosome-related motifs (SURP, CID & G patch) and putative transmembrane domains (TM1 & TM2) highlighted. Next, we applied deletion mutagenesis approach to characterize the important domain(s) for nuclear localization of CHERP. The primer sequences used for deletion mutagenesis are listed in Table 3.1 in the Methods section. HEK293 cells were plated on MatTek dishes for confocal live cell imaging and regular T75 flasks for fractionation-Western blotting analysis. The results for the wild-type and C-terminal deletion mutants are illustrated in Figure 3.7.

Figure 3.6. Schematic diagram of functional domains in CHERP. SURP: suppressor-of-white-apricot and PRP21/SPP91; CID: CTD-interacting domain; G: poly G domain; TM: transmembrane domain; NLS: nuclear localization signal. 58

Table 3.2. Algorithms used for in silico prediction of domain structures and post-translational modifications of CHERP.

Function Algorithm Predicted results Transmembrane helices TopPred 2 (aa. 78-98; 232-252) HMMTOP 1 (aa. 641-659) LOCATE 1 (aa. 239-245) SOSUI None TMHMM None TMpred None

Signal peptide prediction SignalP None

Subcellular localization WoLF PSORT Nuclear Yloc Nuclear LOCATE Nuclear ESLPred Nuclear NucPred Nuclear

NLS prediction NLStradamus 1 (aa. 718-817) NLS-lit cleaned 2 (aa. 767-777; 774-783) mRNA processing PROSITE SURP motif (aa.15-57) CID domain (aa. 149-289) G_patch (aa. 841-891)

Glycosylation prediction NetNGlyc 1 (aa. 900) PROCITE 2 (aa. 456, 900) NetOGlyc 7

Phosphorylation prediction PROCITE Multiple sites(cAMP, cGMP, PKC, CK2, TYR) NetPhos Multiple sites at C-terminal

SUMOylation SUMOsp 2.0 3 (aa. 318, 660, 916)

Palmitoylation CSS-Palm 3.0 4 (aa. 69, 225, 772, 915)

Figure 3.7. Subcellular distribution of wild-type vs. C-terminal truncation mutants of CHERP. (A) Confirmation of CHERP-GFP expression by Western blotting (full length vs. C-terminal deletion constructs). (B) Live cell imaging in HEK293 cells overexpressing full-length (1-916) and serial C-terminal deleted CHERP (1-897, 1-838, 1-777, and 1-725). GFP was used as a control. All images were taken under the same settings. (C) Quantitative results from (B). Intensity values of GFP fluorescence was quantified and expressed as ratios (nucleus/cytosol). (D) WB detection of wild-type and CHERP1-725 mutant in cytosolic and nuclear fractions recapitulated the results from live cell imaging analysis. CREB was used as a nuclear marker.

We first performed a regular Western blotting in HEK293 WCL overexpressing GFP-tagged wild-type (1-916) CHERP vs. truncation mutants to ensure full-length proteins were synthesized (Figure 3.7A). Next, we transiently transfected these constructs into HEK293 cells and performed live cell imaging analysis using confocal microscopy. As shown in Figure 3.7B, while wild-type and the majority of the C-terminal truncation mutants recapitulated the nuclear localization pattern to the endogenous counterpart, the 60 distribution of CHERP1-725 mutant was enriched in the cytoplasm, resembling that of GFP protein alone (Figure 3.7B). The intensity values of GFP fluorescence of each CHERP construct in cytosolic and nuclear compartments was quantified, and data were processed and plotted as ratios of the intensity value in the nucleus to that in the cytoplasm (Figure 3.7C). An N/C ratio larger than 1 indicated that protein was enriched in the nuclear fraction; on the other hand, N/C ratios less than 1 indicated proteins were enriched in the cytosolic fraction. As shown in Figure 3.7C, wild-type CHERP and most of the C- terminal truncation mutants had N/C ratios above 20 (n=7-20), supporting strong nuclear localization. Statistical analysis revealed that there was no significant difference, in terms of subcellular distribution, among these “nuclear CHERP” constructs. For CHERP1-725 and GFP vector alone, the N/C ratios were between 0.5 and 2, and there was no significant difference between these two protein species. One-way ANOVA analysis demonstrated a significant difference (p<0.001) between the “nuclear CHERP” and CHERP1-725/GFP. These truncation studies suggested that the C-terminal region was essential for targeting CHERP in the nucleus. Fractionation-WB of wild-type and CHERP1-725 mutant also confirmed the differential subcellular distribution of these two CHERP constructs (Figure 3.7D).

Given that the N-terminal regions of CHERP have motifs that potentially interact with nuclear protein complexes (transcription and mRNA maturation machinery), we next asked whether the N-terminus was also important to transport CHERP into the nucleus. We again applied a deletion mutagenesis approach to remove all the transcription/spliceosome-related domains (CHERP342-916), and the primer sequences were listed in Table 3.1. This N-terminal truncation mutant was then overexpressed in HEK293 cells, along with the wild-type (CHERP1-916) and a GFP empty vector as controls. Live cell imaging results revealed that CHERP342-916 remained localized to the nucleus (lower left panel, Figure 3.8A); statistical analysis of the imaging data showed that there was no significant difference between the N-terminal deletion mutant and the wild-type CHERP (Figure 3.8B). The subcellular distribution pattern was further

61 confirmed by fractionation-WB (Figure 3.8C). All these data suggest that the N-terminal region was not required for CHERP‟s nuclear localization.

Is the C-terminus then sufficient for targeting CHERP to the nucleus? To answer this question, we first attempted to construct a CHERP725-916 mutant; however, the primer pairs were not compatible for PCR reaction. We therefore extended the forward primer further toward the 5‟ end and made a deletion construct based on amino acid 705-916 instead (CHERP705-916). A CHERP1-705 truncation mutant was made in parallel and its distribution pattern recapitulated to that of CHERP1-725 (upper middle and right panels, Figure 3.8A). The majority of the CHERP705-916 mutant was again localized in the nucleus (lower middle panel, Figure 3.8A), and the minor cytosolic distribution was possibly due to passive diffusion of this mutant from nucleus to cytoplasm since its molecular mass (50kDa) is close to the MW cutoff for the nuclear pore (40kDa). One-way ANOVA analysis revealed significant difference between CHERP705-916 mutant and other CHERP protein species (Figure 3.8B); further, the fractionation-WB data confirmed live cell imaging observations (Figure 3.8C). Therefore, all our data suggested that the C-terminal regions of CHERP were necessary and sufficient for residing CHERP in the nucleus.

To examine the universality of CHERP‟s nuclear identity, we applied fractionation- WB in other human cell lines, SKBR3 and Jurkat cells. Consistent with the findings in HEK293 and rat cardiac tissues, endogenous CHERP in SKBR3 and Jurkat was enriched in the nuclear preparation (Figure 3.8D). Live cell imaging and fractionation-WB of the GFP-tagged wild-type and N- or C-terminal deletion mutants in SKBR3 cell line also suggested that the C-terminal regions of CHERP were necessary and sufficient for CHERP‟s nuclear localization (Figure 3.8E & F).

Taken together, our studies initially attempted to identify TPC-protein complexes by immunoprecipitation-mass spectrometry methods. The feasibility of a one-step IP-MS strategy for identification of membrane protein complexes and alternative approaches will be discussed in the Discussion section. Even though TPCs were successfully characterized in the immunoprecipitates, we failed to validate the interaction of TPCs

62 with one promising protein candidate, CHERP. Nonetheless, the results from our careful validation experiments demonstrated that CHERP was a nuclear protein contrasting to its current published E/SR localization. We have therefore redefined properties of CHERP and will discuss the significance of the finding in the context of published literature also in the Discussion.

Figure 3.8. C-terminal domains of CHERP is sufficient for nuclear localization. (A) Live cell imaging in HEK293 cells over-expressing full-length (1-916), C-terminal (1-705 & 1-725) and N-terminal deleted CHERP (342-916 & 705-916). GFP was used as a control. All images were taken under the same settings. (B) Quantitative results from (A). Intensity values of GFP fluorescence was quantified and expressed as ratios (nucleus/cytosol). (C) WB detection of wild-type and deletion constructs of CHERP-GFP in cytosolic and nuclear fractions confirmed CHERP‟s subcellular localization as revealed by live cell images. (D) Fractionation-WB in Jurkat and SKBR3 cells also revealed that CHERP was a nuclear protein in these human cell lines. (E) Quantitative results from live cell imaging in SKBR3 cells. (F) Fractionation-WB confirmed observation from live cell imaging studies. Deletion mutagenesis in SKBR3 also confirmed that the C-terminal regions were essential and sufficient for CHERP‟s nuclear localization in this cell type. 63

Discussion

The calcium homeostasis endoplasmic reticulum protein (CHERP) has been described as a putative ER-integral protein (O‟Rourke et al., 1994, 2003; LaPlante et al., 2000; Ryan et al., 2011). Structure analysis of CHERP amino acid sequence reveals two potential transmembrane domains and a pore-forming feature similar to that of K+ channels (O‟Rourke et al., 2003). Inducible knockdown of endogenous CHERP in human leukemic cell lines stably transfected with an anti-sense CHERP abolishes thrombin- induced calcium release from the IP3-sensitive ER stores, responses similar to that of blockade of certain potassium channels, suggesting synergistic effects of CHERP function and ER K+ counter ion permeability (O‟Rourke et al., 1994, 2003; LaPlante et al., 2000). More recently, CHERP was found to co-purify with RyR1 via metal ion affinity chromatography of lysates from HEK293 cells overexpressing the cytosolic region of RyR1, and was demonstrated to co-localize with endogenous RyR in muscle tissues. siRNA-mediated knockdown of endogenous CHERP in HEK293 cells also attenuated RyR1-regulated Ca2+ release (Ryan et al., 2011). These published results were all interpreted to suggest that CHERP interacts with several calcium-permeable ER membrane proteins (IP3R and RyR) and mediates ER calcium homeostasis possibly by means of regulating potassium conductance in the ER. However, our studies in human cell lines (HEK293, SKBR3 & Jurkat cells) and rat muscle tissues (skeletal & cardiac myocytes) all revealed that CHERP is enriched in the nucleus (Figures 3.3, 3.4, 3.5, & 3.8D). Further, truncation mutagenesis and live cell imaging studies defined the C- terminal regions of CHERP as necessary and sufficient for targeting CHERP to the nucleus (Figures 3.7 & 3.8). Therefore, in my hands, the demonstration of novel nuclear localization of CHERP and a failure to observe CHERP‟s ER distribution pattern in several systems merits critical reevaluation of: (i) the specificity of the anti-CHERP antibodies used in prior studies, and (ii) the issue of whether CHERP is indeed an integral ER K+ channel, or rather, a nuclear protein that impacts ER calcium homeostasis via a different mechanism.

First, to validate the specificity of the anti-CHERP Ab (ab15951, Abcam) used here and in Ryan et al. 2011, we exploited both Western immunoreactivity and IP-MS based methods (Figure 3.4). These data show recognition of an 118kDa band in native cells (Figure 3.4A & B), a 143kDa band in GFP-CHERP overexpressing cells (~27kDa larger) (Figure 3.4A), and reveal multiple unique peptide hits to CHERP when used for IP-MS (Figure 3.4B & C). Therefore, the specificity of this anti-CHERP Ab was confirmed and data presented in our study were reliable. We noticed that several spliceosome-associated proteins and splicing factors were also co-precipitated with CHERP, suggesting CHERP may naturally form a protein complex with spliceosomes.

Is CHERP indeed a nuclear protein or have we somehow masked its ER properties proposed by most of the published findings? Taking a closer look at the colocalization data shown in Ryan et al. 2011 (CHERP vs. RyR1 in rat skeletal muscle; Figure 3.4A, upper panel), the same CHERP Ab did detect immunoreactivity in an oval structure right under the sarcolemma, a typical nuclear localization pattern in myocytes, from which RyR1 immunoreactivity was totally excluded (Ryan et al., 2011). Therefore, CHERP localization might also be in the nucleus in their studies and would require further validation using a DAPI staining method, and proper negative controls (omitting primary or secondary antibodies) ought to be included to further verify the cytosolic immunostaining pattern of CHERP.

CHERP was initially identified through an antibody (mAb213-21) screening method in IP3-sensitive membranes isolated from platelets (O‟Rourke et al., 1994), and later by cDNA and EST expression library screening using the same antibody (LaPlante et al., 2000). Perhaps the cytosolic staining in HEL and Jurkat cells demonstrated by the original anti-CHERP antibody, which was an IgM species, may simply result from inefficient entry into the nucleus given the huge size of IgM (~900kDa) quantitatively versus IgG (~150kDa) (LaPlante et al., 2000; O‟Rourke et al., 2003). Indeed, when cells were permeabilized thoroughly or when custom-made IgG antibodies were utilized instead for immunocytochemistry, the authors have identified strong nuclear/peri-nuclear and sporadic cytosolic staining for CHERP in a variety of cell types (personal 65 communication with Dr. Feinstein, the principal investigator of the early CHERP studies), converging with our major findings in this chapter.

The mobility of CHERP detected in this study is slower than theoretically predicted. The observed molecular mass of both endogenous and GFP-tagged CHERP is ~15kDa larger than the predicted sizes (118 vs. 103kDa, and 143 vs. 130kDa, respectively). One possibility for the size shift is that there may exist post-translational modifications (PTMs) on the CHERP protein that decrease its electrophoretic mobility (Dephoure et al., 2008; Choudhary et al., 2009; Rigbolt et al., 2011). In silico proteomics analysis of the CHERP sequence has identified several possible motifs for post-translational modification (Table 3.2). Several serine/threonine residues in the C-terminal regions of CHERP were found to be phosphorylated during cell-cycle regulation, T cell receptor (TCR)-mediated signal transduction, and embryonic stem cell differentiation (Dephoure et al., 2008; Mayya et al., 2009; Rigbolt et al., 2011). When HEK293 WCL were pretreated with alkaline phosphatase to remove the phosphate moiety, our Western blotting results also revealed that CHERP, both endogenous and exogenous, was phosphorylated in this cell type (data not shown). That dephosphorylation did not completely bring CHERP‟s molecular mass down to its theoretical size (~103kDa) suggest CHERP naturally harbors more than one type of PTMs. Several lines of evidence have revealed that phosphorylation, SUMOylation, glycosylation and palmitoylation are essential mechanisms for the translocation of nuclear proteins in different nuclear sub-domains or between nucleus and cytoplasm or even plasma membrane under certain stimuli (Zhang et al., 2000; Juang et al., 2002; Acconcia et al., 2005; Belanger et al., 2005; Du et al., 2008; Renner et al., 2010; Wang et al., 2010; Teng et al., 2011; Wilson et al., 2011). Therefore, the occasional cytosolic distribution of CHERP observed in our live cell imaging and immunocytochemistry studies might result from alternating combinations of PTMs in these particular cells. Verification of these potential PTMs could be examined by treatment of deglycosylation agents, site-directed mutagenesis, and mass spectrometry.

How then does a nuclear protein affect ER calcium homeostasis and cellular senescence? Domain analyses of CHERP in our studies have characterized several 66 protein-protein interaction (PPI) motifs, including a polyglutamine track, a proline-rich motif, several serine/arginine-rich motifs or RS domains, a SURP, a CID, and a G patch. Notably, most of these PPI modules are important features for proteins involved in the regulation of gene transcription, and mRNA maturation and translation (Sanford et al., 2005; Kuwasako et al., 2006; Guo et al., 2007; Zhong et al., 2009). That several splicing factors or spliceosome-interacting proteins (SR140, SF3B2 & SB3B3) are co-precipitated with CHERP (Figure 3.4C) suggests these proteins are naturally residing in the same macro-protein complex. Furthermore, previous studies guided by similar IP-MS strategy or tandem tagging of pre-mRNA followed by affinity chromatography also revealed that CHERP is an accessory protein in the 17S U2 snRNP complex (Neubauer et al., 1998; Hartmuth et al., 2002; Will et al., 2002; Ingham et al., 2005). All these findings substantiate CHERP‟s nuclear localization and suggest that CHERP may function as: (i) a molecular platform to facilitate assembly of snRNP for pre-mRNA splicing via SURP, G patch & RS domain, (ii) a scaffold to allow mRNA transcription and maturation to occur simultaneously and efficiently (SURP, CID, G patch & RS domain), and (iii) a nucleocytoplasmic shuttle that transports mature mRNA to translation-active compartments (G patch & RS domain). These hypotheses obviously need to be verified carefully; however, they do provide an alternative explanation for the dysregulation of ER calcium homeostasis and cellular senescence in CHERP knockdown cells reported previously (LaPlante et al., 2000; O‟Rourke et al., 2003; Ryan et al., 2011). In these studies, expression of endogenous CHERP was reduced by siRNA-mediated methods. If CHERP is the core for mRNA transcription/splicing machinery to assemble and function properly, knocking down CHERP may lead to malformation or disruption of the macromolecular machinery, therefore impede transcription functions. Further, if CHERP is also essential for mRNA export and translation, silencing CHERP may inhibit replenishment of important proteins that function to maintain normal cellular activities or synthesize nascent protein species in response to biological stimuli. As a result, cells may halt transcription/translation and metabolic activities, accumulate harmful chemicals in the cytoplasm, deregulate organellar functions (e.g. ER calcium homeostasis), undergo cell cycle arrest (senescence-like phenotype), and eventually die. 67

In summary, our data have defined novel nuclear properties for CHERP and the key regions for its nuclear targeting, and provided a possible explanation for functional defects in CHERP-knockdown cells. These data necessitate reevaluation of all published CHERP literature to date (O‟Rourke et al., 1994, 2003; Plante et al., 2000; Ryan et al., 2011).

Finally in light of this experience, I will focus on discussing the feasibility of the direct one-step IP-MS strategy in identification of the TPC complex. I propose alternative proteomics approaches to lower the false positive rate and better achieve the goal of characterizing the TPC-interactome. Direct co-immunoprecipitation of endogenous protein complexes is considered the gold standard for detecting protein-protein interactions (Selbach and Mann, 2006; Yang et al., 2008). This method isolates protein complexes from their most natural microenvironment at their physiological expression levels; however, it relies on the existence of good antibodies to perform immunoprecipitation, and successful identification depends on the abundance of the protein species. In our study, co-immunoprecipitation of endogenous TPC-protein complex is less favorable owing to (i) lack of good commercial TPC antibodies and (ii) low abundance of endogenous TPC expression in cells. We therefore utilized an epitope tagging and overexpression strategy attempting to facilitate characterization of the TPC- interactome. Several pilot studies have demonstrated the feasibility to immunoprecipitate TPC proteins followed by WB and MS confirmation. However, the major challenge we encountered as have many other research groups is that the composition of transmembrane protein complexes is very dynamic and highly sensitive to changes in detergent concentrations and the ionic strength in the IP buffer (Hall, 2005). Further, the expression level of signaling molecules is normally low and easily masked by non- specific false-positive interactions (Selbach and Mann, 2006; Yang et al., 2008). All these reasons may account for the failure to validate TPC interaction in our studies; therefore, optimizing IP conditions to better preserve the interactome and increasing the purity of immunoprecipitates prior to MS will be pivotal.

A multitude of proteomics techniques have been developed and applied in the identification of membrane protein interactomes. One should always keep in mind that all these methods are error-prone; therefore, validation of interaction remains necessary. Here, we propose several potential methods to facilitate identification of TPC-interactors as following: (i) tandem affinity purification (TAP): TAP-tagging strategy utilizes two (or more) distinct epitopes to perform dual purification to achieve higher purity. This approach has successfully identified novel interactors for membrane proteins, such as STIM1/Oria1 in human cells (Krapivinsky et al., 2011), nACh receptor in C. elegans (Gottschalk et al., 2005), Cf9 disease resistant protein in plants (Rivas et al., 2002), and so on. There are multiple choices for TAP tags with variable sizes & combinations of antibody-, metal ion-, Strepavidin/biotin-, and Ca2+/calmodulin-based affinity purifications (Yang et al., 2008; Xu et al., 2010), and (ii) Membrane Strep-protein interaction experiment (Membrane-SPINE): this method takes advantage of the intrinsic affinity between biotin and strepavidin (tagged on TPC proteins) to facilitate purification; further, the reversible fixative, formaldehyde, is used to increase crosslinking efficiency of the TPC-interactome in vivo prior to purification procedures, mainly to increase the chance for identification of low abundant but important interacting proteins (Herzberg et al., 2007; Müller et al., 2011). These methods are compatible with other enrichment strategies, such as membrane preparation, lysosomal isolation, and size exclusion chromatography. Future work will focus on utilizing and optimizing these approaches, and further, cross reference with protein identification by photoaffinity labeling strategy using bifunctional probe to get better insights for TPC-interactomes and their dynamics in the context of NAADP gating.

Bibliography

Aarhus R, Graeff RM, Dickey DM, Walseth TF, Lee HC (1995) ADP-ribosyl cyclase and CD38 catalyze the synthesis of a calcium-mobilizing metabolite from NADP. The Journal of Biological Chemistry 270:30327–30333.

Acconcia F, Ascenzi P, Bocedi A, Spisni E, Tomasi V, Trentalance A, Visca P, Marino M (2005) Palmitoylation-dependent Estrogen Receptor alpha Membrane Localization: Regulation by 17 beta-Estradiol. Molecular Biology of the Cell 16:231–237.

Aley PK, Mikolajczyk AM, Munz B, Churchill GC, Galione A, Berger F (2010) Nicotinic acid adenine dinucleotide phosphate regulates skeletal muscle differentiation via action at two-pore channels. Proceedings of the National Academy of Sciences of the United States of America 107:19927–19932.

Arikkath J, Campbell KP (2003) Auxiliary subunits: essential components of the voltage- gated calcium channel complex. Current Opinion in Neurobiology 13:298–307.

Bach G, Zeevi DA, Frumkin A, Kogot-Levin A (2010) Mucolipidosis type IV and the mucolipins. Biochemical Society Transactions 38:1432–1435.

Bak J, Billington RA, Timar G, Dutton AC, Genazzani AA (2001) NAADP receptors are present and functional in the heart. Current Biology 11:987–990.

Bak J, Billington RA, Genazzani AA (2002) Effect of luminal and extracellular Ca on NAADP binding and release properties. Biochemical and Biophysical Research Communications 295:806–811.

Beck A, Kolisek M, Bagley LA, Fleig A, Penner R (2006) Nicotinic acid adenine dinucleotide phosphate and cyclic ADP-ribose regulate TRPM2 channels in T lymphocytes. The FASEB Journal 20:962–964.

Belanger KD, Gupta A, MacDonald KM, Ott CM, Hodge CA, Cole CM, Davis LI (2005) Nuclear pore complex function in Saccharomyces cerevisiae is influenced by glycosylation of the transmembrane nucleoporin Pom152p. Genetics 171:935–947.

Berg I, Potter BV, Mayr GW, Guse AH (2000) Nicotinic acid adenine dinucleotide phosphate (NAADP+) is an essential regulator of T-lymphocyte Ca2+-signaling. The Journal of Cell Biology 150:581–588.

Berridge MJ, Lipp P, Bootman MD (2000) The versatility and universality of calcium signaling. Nature Reviews in Molecular Cell Biology 1:11–21.

Bezprozvanny I, Watras J, Ehrlich BE (1991) Bell-shaped calcium-response curves of Ins(1,4,5)P3- and calcium-gated channels from ER of cerebellum. Nature 351:751– 754.

Billington RA, Bellomo EA, Floriddia EM, Erriquez J, Distasi C, Genazzani AA (2006) A transport mechanism for NAADP in a rat basophilic cell line. The FASEB Journal 20:521–523.

Boittin F-X, Galion A, Evans AM (2002) Nicotinic acid adenine dinucleotide phosphate mediates Ca2+ signals and contraction in arterial smooth muscle via a two-pool mechanism. Circulation Research 91:1168–1175.

Brailoiu E, Churamani D, Cai X, Schrlau MG, Brailoiu GC, Gao X, Hooper R, Boulware MJ, Dun NJ, Marchant JS, Patel S (2009) Essential requirement for two-pore channel 1 in NAADP-mediated calcium signaling. The Journal of Cell Biology 186:201–209.

Brailoiu E, Churamani D, Pandey V, Brailoiu GC, Tuluc F, Patel S, Dun NJ (2006) Messenger-specific role for nicotinic acid adenine dinucleotide phosphate in neuronal differentiation. The Journal of Biological Chemistry 281:15923–15928.

Brailoiu E, Miyamoto MD, Dun NJ (2001) Nicotinic acid adenine dinucleotide phosphate enhances quantal neurosecretion at the frog neuromuscular junction: possible action on synaptic vesicles in the releasable pool. Molecular Pharmacology 60:718–724. aBrailoiu E, Rahman T, Churamani D, Prole DL, Brailoiu GC, Hooper R, Taylor CW, Patel S (2010) An NAADP-gated two-pore channel targeted to the plasma membrane uncouples triggering from amplifying Ca2+ signals. The Journal of Biological Chemistry 285:38511–38516. bBrailoiu GC, Gurzu B, Gao X, Parkesh R, Aley PK, Trifa DI, Galione A, Dun NJ, Madesh M, Patel S, Churchill GC, Brailoiu E (2010) Acidic NAADP-sensitive calcium stores in the endothelium: agonist-specific recruitment and role in regulating blood pressure. The Journal of Biological Chemistry 285:37133–37137.

Calcraft PJ, Ruas M, Pan Z, Cheng X, Arredouani A, Hao X, Tang J, Rietdorf K, Teboul L, Chuang KT, Lin P, Xiao R, Wang C, Zhu Y, Lin Y, Wyatt CN, Parrington J, Ma J, Evans AM, Galione A, Zhu MX (2009) NAADP mobilizes calcium from acidic organelles through two-pore channels. Nature 459:596–600.

Cancela JM, Churchill GC, Galione A (1999) Coordination of agonist- induced Ca2+- signalling patterns by NAADP in pancreatic acinar cells. Nature 398:74–76.

Cancela JM, Gerasimenko OV, Gerasimenko JV, Tepikin a V, Petersen OH (2000) Two different but converging messenger pathways to intracellular Ca2+ release: the roles of nicotinic acid adenine dinucleotide phosphate, cyclic ADP-ribose and inositol trisphosphate. The EMBO Journal 19:2549–2557.

Chini EN, Dousa TP (1995) Enzymatic synthesis and degradation of nicotinic acid adenine dinucleotide phosphate (NAADP), a Ca2+-releasing agonist, in rat tissues. Biochemical and Biophysical Research Communications 209:167–174.

Chini EN, Dousa TP (1996) NAADP does not behave as a CICR system. Biochemical Journal 711:709–711.

Chini EN, Beers KW, Dousa TP (1995) NAADP triggers a specific calcium release in sea urchin eggs. The Journal of Biological Chemistry 270:3216–3223.

Choe CU, Ehrlich BE (2006) The inositol 1,4,5-trisphosphate receptor (IP3R) and its regulators: sometimes good and sometimes bad teamwork. Science‟s STKE re15: 1- 12.

Choudhary C, Kumar C, Gnad F, Nielsen ML, Rehman M, Walther TC, Olsen JV, Mann M (2009) Lysine acetylation targets protein complexes and co-regulates major cellular functions. Science 325:834–840.

Christensen KA, Myers JT, Swanson JA (2002) pH-dependent regulation of lysosomal calcium in macrophages. Journal of Cell Science 115:599–607.

Churamani D, Carrey EA, Dickinson GD, Patel S (2004) Determination of cellular nicotinic acid-adenine dinucleotide phosphate (NAADP) levels. The Biochemical Journal 380:449–454.

Churamani D, Hooper R, Brailoiu E, Patel S (2011) Domain assembly of NAADP-gated two-pore channels. Biochemical Journal 441:317–323.

Churchill GC, Okada Y, Thomas JM, Genazzani AA, Patel S, Galione A (2002) NAADP mobilizes Ca2+ from reserve granules, lysosome-related organelles, in sea urchin eggs. Cell 111:703–708.

Collins MO, Husi H, Yu L, Brandon JM, Anderson CNG, Blackstock WP, Choudhary JS, Grant SGN (2006) Molecular characterization and comparison of the components and multiprotein complexes in the postsynaptic proteome. Journal of Neurochemistry 97:16–23.

Copello JA, Qi Y, Jeyakumar LH, Ogunbunmi E, Fleischer S (2001) Lack of effect of cADP-ribose and NAADP on the activity of skeletal muscle and heart ryanodine receptors. Cell Calcium 30:269–284.

Cordiglieri C, Odoardi F, Zhang B, Nebel M, Kawakami N, Klinkert WEF, Lodygin D, Lühder F, Breunig E, Schild D, Ulaganathan VK, Dornmair K, Dammermann W, Potter BVL, Guse AH, Flügel A (2010) Nicotinic acid adenine dinucleotide phosphate-mediated calcium signalling in effector T cells regulates autoimmunity of the central nervous system. Brain 133:1930–1943.

Cosker F, Cheviron N, Yamasaki M, Menteyne A, Lund FE, Moutin M-J, Galione A, Cancela J-M (2010) The ecto-enzyme CD38 is a nicotinic acid adenine dinucleotide phosphate (NAADP) synthase that couples receptor activation to Ca2+ mobilization from lysosomes in pancreatic acinar cells. The Journal of Biological Chemistry 285:38251–38259.

De Stefani D, Raffaello A, Teardo E, Szabò I, Rizzuto R (2011) A forty-kilodalton protein of the inner membrane is the mitochondrial calcium uniporter. Nature 476:336–340.

Dephoure N, Zhou C, Villén J, Beausoleil SA, Bakalarski CE, Elledge SJ, Gygi SP (2008) A quantitative atlas of mitotic phosphorylation. Proceedings of the National Academy of Sciences of the United States of America 105:10762–10767.

Dickey DM (1999) Characterization of the NAADP binding protein. Doctoral Thesis.

Dionisio N, Albarrán L, López JJ, Berna-Erro A, Salido GM, Bobe R, Rosado JA (2011) Acidic NAADP-releasable Ca2+ compartments in the megakaryoblastic cell line MEG01. Biochimica et Biophysica Acta 1813:1483–1494.

Dong XP, Cheng X, Mills E, Delling M, Wang F, Kurz T, Xu H (2008) The type IV mucolipidosis-associated protein TRPML1 is an endolysosomal iron release channel. Nature 455:992–996.

Dormán G, Prestwich GD (2000) Using photolabile ligands in drug discovery and development. Trends in Biotechnology 18:64–77.

Downie L, Priddle J, Hawes C, Evans DE (1998) A calcium pump at the higher plant nuclear envelope? FEBS Letters 429:44–48.

Du JX, Bialkowska AB, McConnell BB, Yang VW (2008) SUMOylation regulates nuclear localization of Krüppel-like factor 5. The Journal of Biological Chemistry 283:31991–32002.

Esposito B, Gambara G, Lewis AM, Palombi F, D‟Alessio A, Taylor LX, Genazzani AA, Ziparo E, Galione A, Churchill GC, Filippini A (2011) NAADP links histamine H1 receptors to secretion of von Willebrand factor in human endothelial cells. Blood 117:4968–4977.

Foskett JK, White C, Cheung K-H, Mak D-OD (2007) Inositol trisphosphate receptor Ca2+ release channels. Physiological Reviews 87:593–658.

Furuichi T, Cunningham KW, Muto S (2001) A putative two pore channel AtTPC1 mediates Ca2+ flux in Arabidopsis leaf cells. Plant & Cell Physiology 42:900–905.

Galione A (2006) NAADP, a new intracellular messenger that mobilizes Ca2+ from acidic stores. Biochemical Society Transactions 34:922–926.

Galione A, Churchill GC (2002) Interactions between calcium release pathways: multiple messengers and multiple stores. Cell Calcium 32:343–354.

Galione A, Morgan AJ, Arredouani A, Davis LC, Rietdorf K, Ruas M, Parrington J (2010) NAADP as an intracellular messenger regulating lysosomal calcium-release channels. Biochemical Society Transactions 38:1424–1431.

Galione A, Parrington J, Funnell T (2011) Physiological roles of NAADP-mediated Ca2+ signaling. Science China Life Sciences 54:725–732.

Gambara G, Billington RA, Debidda M, D‟Alessio A, Palombi F, Ziparo E, Genazzani AA, Filippini A (2008) NAADP-induced Ca2+ signaling in response to endothelin is via the receptor subtype B and requires the integrity of lipid rafts/caveolae. Journal of Cellular Physiology 216:396–404.

Genazzani AA, Empson RM, Galione A (1996) Unique inactivation properties of NAADP-sensitive Ca2+ release. The Journal of Biological Chemistry 271:11599– 11602.

Gerasimenko JV, Lur G, Sherwood MW, Ebisui E, Tepikin AV, Mikoshiba K, Gerasimenko OV, Petersen OH (2009) Pancreatic protease activation by alcohol 2+ metabolite depends on Ca release via acid store IP3 receptors. Proceedings of the National Academy of Sciences of the United States of America 106:10758–10763.

Gerasimenko JV, Maruyama Y, Yano K, Dolman NJ, Tepikin AV, Petersen OH, Gerasimenko OV (2003) NAADP mobilizes Ca2+ from a thapsigargin-sensitive store in the nuclear envelope by activating ryanodine receptors. The Journal of Cell Biology 163:271–282.

Gottschalk A, Almedom RB, Schedletzky T, Anderson SD, Yates JR, Schafer WR (2005) Identification and characterization of novel nicotinic receptor-associated proteins in Caenorhabditis elegans. The EMBO Journal 24:2566–2578.

Guerrero-Hernandez A, Dagnino-Acosta A, Verkhratsky A (2010) An intelligent sarco- endoplasmic reticulum Ca2+ store: release and leak channels have differential access to a concealed Ca2+ pool. Cell Calcium 48:143–149.

Guo L, Han A, Bates DL, Cao J, Chen L (2007) Crystal structure of a conserved N- terminal domain of histone deacetylase 4 reveals functional insights into glutamine- rich domains. Proceedings of the National Academy of Sciences of the United States of America 104:4297–4302.

Guse AH (1999) Cyclic ADP-ribose: a novel Ca2+-mobilising second messenger. Cellular Signalling 11:309–316.

Gómez-Suaga P, Luzón-Toro B, Churamani D, Zhang L, Bloor-Young D, Patel S, Woodman PG, Churchill GC, Hilfiker S (2011) Leucine-rich repeat kinase 2 regulates autophagy through a calcium-dependent pathway involving NAADP. Human Molecular Genetics 21:511–525.

Hall RA (2005) Co-immunoprecipitation as a strategy to evaluate receptor-receptor or receptor-protein interactions. G Protein-Coupled Receptor-Protein Interactions ISBN:978-0-471-23546-0:165–178.

Hartmuth K, Urlaub H, Vornlocher H-P, Will CL, Gentzel M, Wilm M, Lührmann R (2002) Protein composition of human prespliceosomes isolated by a tobramycin affinity-selection method. Proceedings of the National Academy of Sciences of the United States of America 99:16719–16724.

Herzberg C, Weidinger LAF, Dörrbecker B, Hübner S, Stülke J, Commichau FM (2007) SPINE: a method for the rapid detection and analysis of protein-protein interactions in vivo. Proteomics 7:4032–4035.

Hooper R, Churamani D, Brailoiu E, Taylor CW, Patel S (2011) Membrane topology of NAADP-sensitive two-pore channels and their regulation by N-linked glycosylation. The Journal of Biological Chemistry 286:9141–9149.

Hooper R, Patel S (2012) in Calcium Signaling, Advances in Experimental Medicine and Biology (Islam, M.S., ed), Vol. 740, Spring-Verlag New York Inc., New York, in press.

Husi H, Ward MA, Choudhary JS, Blackstock WP, Grant SG (2000) Proteomic analysis of NMDA receptor-adhesion protein signaling complexes. Nature Neuroscience 3:661–669.

Ingham RJ et al. (2005) WW Domains Provide a Platform for the Assembly of Multiprotein Networks. Molecular and Cellular Biology 25:7092–7106.

Ishibashi K, Suzuki M, Imai M (2000) Molecular cloning of a novel form (two-repeat) protein related to voltage-gated sodium and calcium channels. Biochemical and Biophysical Research Communications 270:370–376.

Jain P, Slama JT, Perez-Haddock LA, Walseth TF (2010) Nicotinic acid adenine dinucleotide phosphate analogues containing substituted nicotinic acid: effect of modification on Ca2+ release. Journal of Medicinal Chemistry 53:7599–7612.

Johnson JD, Misler S (2002) NAADP-sensitive calcium stores initiate insulin signaling in human beta cells. Proceedings of the National Academy of Sciences of the United States of America 99:14566–14571.

Juang Y-T, Solomou EE, Rellahan B, Tsokos GC (2002) Phosphorylation and O-linked glycosylation of Elf-1 leads to its translocation to the nucleus and binding to the promoter of the TCR zeta-chain. Journal of Immunology 168:2865–2871.

Kim BJ, Park KH, Yim CY, Takasawa S, Okamoto H, Im MJ, Kim UH (2008) Generation of nicotinic acid adenine dinucleotide and cyclic ADP-ribose by glucagon-like peptide-1eEvokes Ca2+ signal that is essential for insulin secretion in mouse pancreatic islets. Diabetes 57:868–878.

Krapivinsky G, Krapivinsky L, Stotz SC, Manasian Y, Clapham DE (2011) POST, partner of stromal interaction molecule 1 (STIM1), targets STIM1 to multiple transporters. Proceedings of the National Academy of Sciences of the United States of America 108:19234–19239.

Kuwasako K, He F, Inoue M, Tanaka A, Sugano S, Güntert P, Muto Y, Yokoyama S (2006) Solution structures of the SURP domains and the subunit-assembly mechanism within the splicing factor SF3a complex in 17S U2 snRNP. Structure 14:1677–1689.

Lamos SM, Krusemark CJ, McGee CJ, Scalf M, Smith LM, Belshaw PJ (2006) Mixed isotope photoaffinity reagents for identification of small-molecule targets by mass spectrometry. Angewandte Chemie 45:4329–4333.

LaPlante JM, O‟Rourke F, Lu X, Fein A, Olsen A, Feinstein MB (2000) Cloning of human CHERP: regulated expression of antisense cDNA depletes CHERP , inhibits

intracellular Ca2+ mobilization and decreases cell proliferation. Biological Journal 199:189–199.

Lee HC (1991) Specific binding of cyclic ADP-ribose to calcium-storing microsomes from sea urchin eggs. The Journal of Biological Chemistry 266:2276–2281.

Lee HC (2005) Nicotinic acid adenine dinucleotide phosphate (NAADP)-mediated calcium signaling. Journal of Biological Chemistry 280:33693–33696.

Lee HC (2011) Cyclic ADP-ribose and NAADP: fraternal twin messengers for calcium signaling. Science China Life Sciences 54:699–711.

Lee HC, Aarhus R (1997) Structural determinants of nicotinic acid adenine dinucleotide phosphate important for its calcium-mobilizing activity. The Journal of Biological Chemistry 272:20378–20383.

Lee HC, Aarhus R (1995) A derivative of NADP mobilizes calcium stores insensitive to IP3 and cADPR. The Journal of Biological Chemistry 270:2152–2157.

Lee HC, Aarhus R (2000) Functional visualization of the separate but interacting calcium stores sensitive to NAADP and cyclic ADP-ribose. Journal of cell science 113 Pt 24:4413–4420.

Lee Hc, Aarhus R, Walseth TF (1993) Calcium mobilization by dual receptors during fertilization of sea urchin. Science 261:352-335.

Lim D, Kyozuka K, Gragnaniello G, Carafoli E, Santella L (2001) NAADP+ initiates the Ca2+ response during fertilization of starfish oocytes. The FASEB Journal 15:2257– 2267.

Lin-Moshier Y, Walseth TF, Churamani D, Davidson SM, Slama JT, Hooper R, Brailoiu E, Patel S, Marchant JS (2012) Photoaffinity labeling of nicotinic acid adenine dinucleotide phosphate (NAADP) targets in mammalian cells. Journal of Biological Chemistry 287:2296–2307.

Lloyd-Evans E, Morgan AJ, He X, Smith D a, Elliot-Smith E, Sillence DJ, Churchill GC, Schuchman EH, Galione A, Platt FM (2008) Niemann-Pick disease type C1 is a sphingosine storage disease that causes deregulation of lysosomal calcium. Nature Medicine 14:1247–1255.

Lloyd-Evans E, Waller-Evans H, Peterneva K, Platt FM (2010) Endolysosomal calcium regulation and disease. Biochemical Society Transactions 38:1458–1464.

Maximov A, Tang T-S, Bezprozvanny I (2003) Association of the type 1 inositol (1,4,5)- trisphosphate receptor with 4.1N protein in neurons. Molecular and Cellular Neuroscience 22:271–283.

Mayya V, Lundgren DH, Hwang S-I, Rezaul K, Wu L, Eng JK, Rodionov V, Han DK (2009) Quantitative phosphoproteomic analysis of T cell receptor signaling reveals system-wide modulation of protein-protein interactions. Science Signaling 2:1-16.

Mikoshiba K, Furuichi T, Miyawaki A (1994) Structure and function of IP3R. Seminar in Cell Biology 5:273–281.

Missiaen L, Dode L, Vanoevelen J, Raeymaekers L, Wuytack F (2007) Calcium in the Golgi apparatus. Cell Calcium 41:405–416.

Miyawaki A, Llopis J, Heim R, Mccaffery JM, Adams JA, Ikurak M, Tsien RY (1997) Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887.

Moccia F, Nusco GA, Lim D, Kyozuka K, Santella L (2006) NAADP and InsP3 play distinct roles at fertilization in starfish oocytes. Developmental Biology 294:24–38.

Morgan AJ, Galione A (2007) Fertilization and nicotinic acid adenine dinucleotide phosphate induce pH changes in acidic Ca2+ stores in sea urchin eggs. The Journal of Biological Chemistry 282:37730–37737.

Morgan AJ, Galione A (2008) Investigating cADPR and NAADP in intact and broken cell preparations. Methods 46:194–203.

Morgan AJ, Platt FM, Lloyd-Evans E, Galione A (2011) Molecular mechanisms of endolysosomal Ca2+ signalling in health and disease. The Biochemical Journal 439:349–374.

Mothet JP, Fossier P, Meunier FM, Stinnakre J, Tauc L, Baux G (1998) Cyclic ADP- ribose and calcium-induced calcium release regulate neurotransmitter release at a cholinergic synapse of Aplysia. The Journal of Physiology 507 :405–414.

Müller CS, Haupt A, Bildl W, Schindler J, Knaus H-G, Meissner M, Rammner B, Striessnig J, Flockerzi V, Fakler B, Schulte U (2010) Quantitative proteomics of the Cav2 channel nano-environments in the mammalian brain. Proceedings of the National Academy of Sciences of the United States of America 107:14950–14957.

Müller VS, Jungblut PR, Meyer TF, Hunke S (2011) Membrane-SPINE: an improved method to identify protein-protein interaction partners of membrane proteins in vivo. Proteomics 11:2124–2128.

Neubauer G, King A, Rappsilber J, Calvio C, Watson M, Ajuh P, Sleeman J, Lamond A, Mann M (1998) Mass spectrometry and EST-database searching allows characterization of the multi-protein spliceosome complex. Nature Genetics 20:46– 50.

Ogunbayo OA, Zhu Y, Rossi D, Sorrentino V, Ma J, Zhu MX, Evans AM (2011) Cyclic adenosine diphosphate ribose activates ryanodine receptors, whereas NAADP activates two-pore domain channels. The Journal of Biological Chemistry 286:9136–9140.

O‟Rourke FA, LaPlante JM, Feinstein MB (2003) Antisense-mediated loss of calcium homoeostasis endoplasmic reticulum protein (CHERP; ERPROT213-21) impairs Ca2+ mobilization, nuclear factor of activated T-cells (NFAT) activation and cell proliferation in Jurkat T-lymphocytes. The Biochemical Journal 373:133–143.

O‟Rourke F, Soons K, Flaumenhauft R, Watras J, Baio-Larue C, Matthews E, Feinstein MB (1994) Ca2+ release by inositol 1,4,5-trisphosphate is blocked by the K+-channel blockers apamin and tetrapentylammonium ion, and a monoclonal antibody to a 63 kDa membrane protein: reversal of blockade by K+ ionophores nigericin and valinomycin and purificati. The Biochemical Journal 300 :673–683.

Palty R, Silverman WF, Hershfinkel M, Caporale T, Sensi SL, Parnis J, Nolte C, Fishman D, Shoshan-Barmatz V, Herrmann S, Khananshvili D, Sekler I (2010) NCLX is an essential component of mitochondrial Na+/Ca2+ exchange. Proceedings of the National Academy of Sciences of the United States of America 107:436–441.

Pandey V, Chuang CC, Lewis AM, Aley PK, Brailoiu E, Dun NJ, Churchill GC, Patel S (2009) Recruitment of NAADP-sensitive acidic Ca2+ stores by glutamate. The Biochemical Journal 422:503–512.

Parekh AB (2003) Mitochondrial regulation of intracellular Ca2+ signaling: more than just simple Ca2+ buffers. News in Physiological Sciences 18:252–256.

Patel S, Brailoiu E (2012) Triggering of Ca2+ signals by NAADP-gated two-pore channels: a role for membrane contact sites? Biochemical Society Transactions 40:153–157.

Patel S, Churchill GC, Galione A (2000) Unique kinetics of NAADP binding enhance the sensitivity of NAADP receptors for their ligand. Biochemical Journal 352:725–729.

Patel S, Docampo R (2010) Acidic calcium stores open for business: expanding the potential for intracellular Ca2+ signaling. Trends in Cell Biology 20:277–286.

Patel S, Ramakrishnan L, Rahman T, Hamdoun A, Marchant JS, Taylor CW, Brailoiu E (2011) The endo-lysosomal system as an NAADP-sensitive acidic Ca2+ store: role for the two-pore channels. Cell Calcium 50:157–167.

Paulo JA, Brucker WJ, Hawrot E (2010) Proteomic analysis of an α7 nicotinic acetylcholine receptor interactome. Proteome 8:1849–1858.

Peltola MA, Kuja-Panula J, Lauri SE, Taira T, Rauvala H (2011) AMIGO is an auxiliary subunit of the Kv2.1 potassium channel. EMBO reports 12:1293–1299.

Pereira GJS, Hirata H, Fimia GM, do Carmo LG, Bincoletto C, Han SW, Stilhano RS, Ureshino RP, Bloor-Young D, Churchill G, Piacentini M, Patel S, Smaili SS (2011) Nicotinic acid adenine dinucleotide phosphate (NAADP) regulates autophagy in cultured astrocytes. The Journal of Biological Chemistry 286:27875–27881.

Pillay CS, Elliott E, Dennison C (2002) Endolysosomal proteolysis and its regulation. The Biochemical Journal 363:417–429.

Pinton P, Pozzan T, Rizzuto R (1998) The Golgi apparatus is an inositol 1,4,5- trisphosphate-sensitive Ca2+ store, with functional properties distinct from those of the endoplasmic reticulum. The EMBO Journal 17:5298–5308.

Pitt SJ, Funnell TM, Sitsapesan M, Venturi E, Rietdorf K, Ruas M, Ganesan A, Gosain R, Churchill GC, Zhu MX, Parrington J, Galione A, Sitsapesan R (2010) TPC2 is a novel NAADP-sensitive Ca2+ release channel, operating as a dual sensor of luminal pH and Ca2+. The Journal of Biological Chemistry 285:35039–35046.

Rah S-Y, Mushtaq M, Nam T-S, Kim SH, Kim U-H (2010) Generation of cyclic ADP- ribose and nicotinic acid adenine dinucleotide phosphate by CD38 for Ca2+ signaling in interleukin-8-treated lymphokine-activated killer cells. The Journal of Biological Chemistry 285:21877–21887.

Renner F, Moreno R, Schmitz ML (2010) SUMOylation-dependent localization of IKK in PML nuclear bodies is essential for protection against DNA-damage-triggered cell death. Molecular Cell 37:503–515.

Rietdorf K, Funnell TM, Ruas M, Heinemann J, Parrington J, Galione A (2011) Two- pore channels form homo- and heterodimers. The Journal of Biological Chemistry 286:37058–37062.

Rigbolt KTG, Prokhorova TA, Akimov V, Henningsen J, Johansen PT, Kratchmarova I, Kassem M, Mann M, Olsen JV, Blagoev B (2011) System-wide temporal characterization of the proteome and phosphoproteome of human embryonic stem cell differentiation. Science Signaling 4:1–17.

Rivas S, Romeis T, Jones JDG (2002) The Cf-9 disease resistance protein is present in an ~420-kilodalton heteromultimeric membrane-associated complex at one molecule per complex. The Plant Cell 14:689–702.

Robinette D, Neamati N, Tomer KB and Borchers CH (2008) Photoaffinity labeling combined with mass spectrometric approaches as a tool for structural proteomics. Expert Review Proteomics 3:399–408.

Ruas M, Rietdorf K, Arredouani A, Davis LC, Lloyd-evans E, Koegel H, Funnell TM, Morgan AJ, Ward JA, Watanabe K, Cheng X, Churchill GC, Zhu MX, Platt FM, Wessel GM, Parrington J, Galione A (2010) Purified TPC isoforms form NAADP receptors with distinct roles for Ca2+ signaling and endolysosomal trafficking. Current Biology 20:1–7.

Ryan T, Sharma P, Ignatchenko A, MacLennan DH, Kislinger T, Gramolini AO (2011) Identification of novel ryanodine receptor 1 (RyR1) protein interaction with calcium homeostasis endoplasmic reticulum protein (CHERP). The Journal of Biological Chemistry 286:17060–17068.

Sanford JR, Ellis J, Cáceres JF (2005) Multiple roles of arginine/serine-rich splicing factors in RNA processing. Biochemical Society Transactions 33:443–446.

Santos SD, Manadas B, Duarte CB, Luı A (2010) Proteomic analysis of an interactome for long-form AMPA receptor subunits. Journal of Proteome Research 9:1670–1682.

Schieder M, Rötzer K, Brüggemann A, Biel M, Wahl-Schott CA (2010) Characterization of two-pore channel 2 (TPCN2)-mediated Ca2+ currents in isolated lysosomes. The Journal of Biological Chemistry 285:21219–21222.

Schoeber JP, Topala CN, Wang X, Diepens RJ, Lambers TT, Hoenderop JG, Bindels RJ (2006) RGS2 inhibits the epithelial Ca2+ channel TRPV6. The Journal of Biological Chemistry 281:29669–29674.

Selbach M, Mann M (2006) Protein interaction screening by quantitative immunoprecipitation combined with. Nature Methods 3:981–983.

Shen W-W, Frieden M, Demaurex N (2011) Remodelling of the endoplasmic reticulum during store-operated calcium entry. Biology of The Cell 103:365–380.

Streb H, Irvine RF, Berridge MJ, Schulz I (1983) Release of Ca2+ from a nonmitochondrial intracellular store in pancreatic acinar cells by IP3. Nature 306:67–69.

Styrt B, Pollack CR, Klempner MS (1988) An abnormal calcium uptake pump in Chediak-Higashi neutrophil lysosomes. Journal of Leukocyte Biology 44:130–135.

Tanabe K, Tucker SJ, Matsuo M, Proks P, Ashcroft FM, Seino S, Amachi T, Ueda K (1999) Direct photoaffinity labeling of the Kir6.2 subunit of the ATP-sensitive K+ channel by 8-azido-ATP. The Journal of Biological Chemistry 274:3931–3933.

Teng ACT, Al-Montashiri NAM, Cheng BLM, Lou P, Ozmizrak P, Chen H-H, Stewart AFR (2011) Identification of a phosphorylation-dependent nuclear localization motif in interferon regulatory factor 2 binding protein 2. PloS One 6:1–8.

Thomas JM, Masgrau R, Churchill GC, Galione A (2001) Pharmacological characterization of the putative cADP-ribose receptor. The Biochemical Journal 359:451–457.

Tomizawa M, Maltby D, Medzihradszky KF, Zhang N, Durkin KA, Presley J, Talley TT, Taylor P, Burlingame AL, Casida JE (2007) Defining nicotinic agonist binding surfaces through photoaffinity labeling. Biochemistry 46:8798–8806.

Tomizawa M, Talley TT, Park JF, Maltby D, Medzihradszky KF, Durkin KA, Cornejo- Bravo JM, Burlingame AL, Casida JE, Taylor P (2009) Nicotinic agonist binding site mapped by methionine- and tyrosine-scanning coupled with azidochloropyridinyl photoaffinity labeling. Journal of Medicinal Chemistry 52:3735–3741.

Tugba Durlu-Kandilci N, Ruas M, Chuang K-T, Brading A, Parrington J, Galione A (2010) TPC2 proteins mediate nicotinic acid adenine dinucleotide phosphate (NAADP)- and agonist-evoked contractions of smooth muscle. The Journal of Biological Chemistry 285:24925–24932.

Walseth TF, Lin-Moshier Y, Jain P, Ruas M, Parrington J, Galione A, Marchant JS, Slama JT (2012) Photoaffinity labeling of high affinity nicotinic acid adenine dinucleotide 2‟-phosphate (NAADP) proteins in sea urchin egg. The Journal of Biological Chemistry 287:2308–2315.

Wang L, Fujiwara S, Somers DE (2010) PRR5 regulates phosphorylation, nuclear import and subnuclear localization of TOC1 in the Arabidopsis circadian clock. The EMBO Journal 29:1903–1915.

Weber PJ, Beck-Sickinger AG (1997) Comparison of the photochemical behavior of four different photoactivatable probes. The Journal of Peptide Research 49:375–383.

Will CL, Urlaub H, Achsel T, Gentzel M, Wilm M, Lührmann R (2002) Characterization of novel SF3b and 17S U2 snRNP proteins, including a human Prp5p homologue and an SF3b DEAD-box protein. The EMBO Journal 21:4978–4988.

Wilson JP, Raghavan AS, Yang Y-Y, Charron G, Hang HC (2011) Proteomic analysis of fatty-acylated proteins in mammalian cells with chemical reporters reveals S- acylation of histone H3 variants. Molecular & Cellular Proteomics 10:1–16.

Xu X, Song Y, Li Y, Chang J, Zhang H, An L (2010) The tandem affinity purification method: an efficient system for protein complex purification and protein interaction identification. Protein Expression and Purification 72:149–156.

Yamaguchi S, Jha A, Li Q, Soyombo AA, Dickinson GD, Churamani D, Brailoiu E, Patel S, Muallem S (2011) Transient receptor potential mucolipin 1 (TRPML1) and two- pore channels are functionally independent organellar ion channels. The Journal of Biological Chemistry 286:22934–22942.

Yamasaki M, Thomas JM, Churchill GC, Garnham C, Lewis AM, Cancela J-M, Patel S, Galione A (2005) Role of NAADP and cADPR in the induction and maintenance of agonist-evoked Ca2+ spiking in mouse pancreatic acinar cells. Current Biology 15:874–878.

Yamasaki-Mann M, Demuro A, Parker I (2009) cADPR stimulates SERCA activity in Xenopus oocytes. Cell Calcium 45:293–299.

Yan D, Tomita S (2011) Defined criteria for auxiliary subunits of glutamate receptors. The Journal of Physiology 1:21–31.

Yang W, Steen H, Freeman MR (2008) Proteomic approaches to the analysis of multiprotein signaling complexes. Proteomics 8:832–851.

Yoshimura SH, Iwasaka S, Schwarz W, Takeyasu K (2008) Fast degradation of the auxiliary subunit of Na+/K+-ATPase in the plasma membrane of HeLa cells. Journal of Cell Science 121:2159–2168.

Yu FH, Westenbroek RE, Silos-Santiago I, Mccormick KA, Lawson D, Ge P, Ferriera H, Lilly J, Distefano PS, Catterall WA, Scheuer T, Curtis R (2003) Sodium Channel  4, a new disulfide-linked auxiliary subunit with similarity to 2. The Journal of Neuroscience 23:7577–7585.

Zbidi H, Jardin I, Woodard GE, Lopez JJ, Berna-Erro A, Salido GM, Rosado JA (2011) STIM1 and STIM2 are located in the acidic Ca2+ stores and associates with Orai1 upon depletion of the acidic stores in human platelets. The Journal of Biological Chemistry 286:12257–12270.

Zhang F, White RL, Neufeld KL (2000) Phosphorylation near nuclear localization signal regulates nuclear import of adenomatous polyposis coli protein. Proceedings of the National Academy of Sciences of the United States of America 97:12577–12582.

Zhang F, Xia M, Li PL (2010) Lysosome-dependent Ca2+ release response to Fas activation in coronary arterial myocytes through NAADP: evidence from CD38 gene knockouts. American journal of physiology Cell Physiology 298:1209–1216.

Zhang F, Xu M, Han W-Q, Li P-L (2011) Reconstitution of lysosomal NAADP-TRP- ML1 signaling pathway and its function in TRPML1 KO cells. American Journal of Physiology 301:421–430.

Zhong X-Y, Wang P, Han J, Rosenfeld MG, Fu X-D (2009) SR proteins in vertical integration of gene expression from transcription to RNA processing to translation. Molecular Cell 35:1–10.

Zissimopoulos S, West DJ, Williams AJ, Lai FA (2006) Ryanodine receptor interaction with the SNARE-associated protein snapin. Journal of Cell Science 119:2386–2397.

Zong X, Schieder M, Cuny H, Fenske S, Gruner C, Rötzer K, Griesbeck O, Harz H, Biel M, Wahl-Schott C (2009) The two-pore channel TPCN2 mediates NAADP- dependent Ca2+-release from lysosomal stores. Pflügers Archiv : European Journal of Physiology 458:891–899.