Regulation of the Human Delta Opioid Receptor

Regulation of the human delta opioid receptor

Item Type text; Electronic Dissertation

Authors Navratilova, Edita

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 02/10/2021 07:59:10

Link to Item http://hdl.handle.net/10150/194173 REGULATION OF THE HUMAN DELTA OPIOID RECEPTOR

Edita Navratilova

______

A dissertation submitted to the Faculty of the

DEPARTMENT OF PHARMACOLOGY

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY WITH A MAJOR IN MEDICAL PHARMACOLOGY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2007 2

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the dissertation committee, we certify that we have read the dissertation prepared by Edita Navratilova entitled “Regulation of the human delta opioid receptor” and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy

______Date:02/27/2007 Henry I. Yamamura, Ph.D.

______Date:02/27/2007 William D. Stamer, Ph.D.

______Date:02/27/2007 Eva V. Varga, Ph.D.

______Date:02/27/2007 William R. Roeske, M.D.

______Date:02/27/2007 John W. Bloom, M.D.

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College. I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

______Date:02/27/2007 Dissertation Director: Henry I. Yamamura, Ph.D.

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for and advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: Edita Navratilova

ACKNOWLEDGEMENT

I thank my mentor, Henry I. Yamamura, Ph.D., for his excellent guidance and encouragements. Hank, my appreciation for what you have done for me and taught me on a professional and individual level grows stronger each day. I am confident that I will continue discovering and remembering your invaluable life lessons in the future. I feel privileged to have such a remarkable example to follow. Thank you.

I also thank Eva V. Varga, Ph.D., William R. Roeske, M.D., and all other members of Dr. Yamamura laboratory for their support and continuous help. 5

TABLE OF CONTENTS

LIST OF TABLES …………………………………………………………………… 8

LIST OF FIGURES ………………………………………………………………….. 9 LIST OF ABBREVIATIONS ……………………………………………………….. 11 ABSTRACT ………………………………………………………………………….. 12

1. INTRODUCTION ………………………………………………………………… 14 Opioid receptors ………………………………………………………………….. 14 Role of the delta opioid receptor in pain treatment ………………………………. 15 • DOR-mediated analgesia ………………………………………………… 16 • Role of the DOR in analgesic tolerance ………………………………….. 17 Regulation of DOR expression …………………………………………………… 19 • Delivery of the DOR to the plasma membrane …………………………… 19 • Agonist-mediated regulation of the DOR …………………………………. 21 Hypothesis ………………………………………………………………………… 24 Goals ………………………………………………………………………………. 25

2. MATERIALS AND METHODS …………………………………………………. 26 Delta opioid receptor ligands ……………………………………………………… 26 hDOR cDNA constructs and cell lines ……………………………………………. 27 • hDOR ……………………………………………………………………… 27 • hDOR(S363A) ………………………………………………………………28 • hDOR(S249A/S255A/S363A) ……………………………………………… 28 • TruncEt-hDOR …………………………………………………………….. 29 • hDOR-EGFP ………………………………………………………………. 29 Cell culture ………………………………………………………………………… 30 Western blot analysis ……………………………………………………………… 30 Confocal microscopy ……………………………………………………………… 31 [3H] naltrindole radioligand binding assay ……………………………………….. 32 GTPγ[35S] binding assay …………………………………………………………. 33 Inhibition of forskolin-stimulated cAMP formation assay ………………………. 34

3. AGONIST-MEDIATED HUMAN DOR PHOSPHORYLATION ……………. 36 Introduction ………………………………………………………………………. 36 Results ……………………………………………………………………………. 38 • S363 in the C-terminus of the hDOR is phosphorylated by morphine …….. 38 • Phosphorylation of S363 by morphine and deltorphin II is concentration- dependent ………………………………………………………………….. 39 6

TABLE OF CONTENTS – Continued

• The rates of hDOR(S363) phosphorylation by morphine and deltorphin II differ……………………………………………………………………….. 40 • Morphine- and deltorphin II-mediated hDOR(S363) phosphorylation displays a biphasic time-course ……………………………………………………… 41 • Morphine and deltorphin II differ in the rates of hDOR(S363) dephosphorylation ………………………………………………………… 42 Discussion …………………………………………………………………………. 43

4. RECRUITMENT OF β-ARRESTIN TO THE AGONIST-STIMULATED HUMAN DOR ……………………………………………………………………….. 46 Introduction ……………………………………………………………………….. 46 Results ……………………………………………………………………………. 49 • β-arrestin2 is recruited to the plasma membrane upon hDOR activation by deltorphin II, DPDPE, SNC80 but not morphine ………………………… 50 • β-arrestin2 interacts only transiently with the hDOR ……………………. 50 • S363A mutation does not eliminate agonist-mediated recruitment of β- arrestin2 …………………………………………………………………… 52 • Truncation of the C-terminus of the hDOR eliminates deltorphin II-mediated recruitment of β-arrestin2 ………………………………………………… 53 Discussion ………………………………………………………………………… 53

5. AGONIST-MEDIATED HUMAN DOR INTERNALIZATION ………………. 56 Introduction ……………………………………………………………………….. 56 Results …………………………………………………………………………….. 59 • Time-course of fluo-deltorphin internalization ………………………….. 59 • hDOR is internalized via a subset of clathrin-coated pits distinct from transferrin-containing CCPs ……………………………………………… 59 • hDOR endocytosis is agonist-specific …………………………………….. 61 • S363A mutation of the hDOR does not eliminate agonist-mediated receptor internalization …………………………………………………………….. 62 Discussion ………………………………………………………………………… 63

6. AGONIST-MEDIATED HUMAN DOR DOWN-REGULATION ……………. 67 Introduction ………………………………………………………………………. 67 Results ……………………………………………………………………………. 71 • hDOR is down-regulated in a time and dose dependent manner …………. 71 • hDOR down-regulation is agonist specific ……………………………….. 71 • hDOR endocytosis via clathrin-coated pits is necessary for receptor down- regulation ………………………………………………………………….. 73 • hDOR is not down-regulated in the proteasomes …………………………. 74 • hDOR down-regulation does not require ubiquitination …………………. 75 7

TABLE OF CONTENTS - Continued

• hDOR is down-regulated in the lysosome ………………………………… 76 • C-terminus of the hDOR contains the degradation targeting motif ………. 77 • S363 within the C-terminus of the hDOR is involved in DPDPE-mediated down-regulation ………………………………………………………….. 78 • The secondary down-regulation targeting motif in the hDOR exposed by SNC80 does not involve S249 and S255 ……………………………………. 79 Discussion ……………………………………………………………………. 80

7. AGONIST-MEDIATED DESENSITIZATION OF THE HUMAN DOR ……. 84 Introduction ………………………………………………………………………. 84 Results …………………………………………………………………………….. 88 • Deltorphin II-mediated hDOR desensitization ……………………………. 88 • Operational Model of Drug Action ……………………………………….. 89 • Modification of the Operational Model to calculate the kinetics of receptor desensitization …………………………………………………………….. 92 • Quantitative evaluation of the kinetics of hDOR desensitization using the modified Operational Model ………………………………………………. 93 • S363A mutation attenuates hDOR desensitization ………………………... 95 • Agonist-specific hDOR desensitization …………………………………… 99 • SNC80 desensitizes the hDOR more effectively than deltorphin II ………. 100 • S363A mutation of the hDOR has a differential effect on SNC80- and deltorphin II-mediated receptor desensitization ………………………….. 101 • Quantitative evaluation of agonist-specific hDOR or hDOR(S363A) desensitization using the Operational Model of Drug Action …………….. 102 Discussion ………………………………………………………………………… 105

8. SUMMARY AND CONCLUSION ………………………………………………. 112

TABLES ……………………………………………………………………………… 119

FIGURES …………………………………………………………………………….. 127

REFERENCES ……………………………………………………………………….. 161 8

LIST OF TABLES

35 Table 7.1. Emax and EC50 values of deltorphin II-mediated stimulation of GTPγ[ S] binding and inhibition of cAMP production in hDOR/CHO cells pretreated for indicated times with deltorphin II ……………………………………………………………….. 119

Table 7.2. Parameters of deltorphin II-mediated desensitization of hDOR calculated by equation (7.5) …………………………………………………………………………. 120

35 Table 7.3. EC50 values of deltorphin II-mediated stimulation of GTPγ[ S] binding and inhibition of cAMP production in naïve and deltorphin II pretreated hDOR/CHO or hDOR(S363A)/CHO cells ……………………………………………………………. 121

Table 7.4. Parameters of deltorphin II-mediated desensitization of hDOR and hDOR(S363A) calculated using the operational model (equation (7.1)) …………….. 122

35 Table 7.5. EC50 values of deltorphin II-mediated stimulation of GTPγ[ S] binding and inhibition of cAMP production in naïve and deltorphin II or SNC80 pretreated hDOR/CHO or hDOR(S363A)/CHO cells …………………………………………… 123

Table 7.6. Parameters of deltorphin II-mediated desensitization of hDOR and hDOR(S363A) calculated using the operational model (equation (7.1)) ……………... 124

Table 8.1. Maximal effects produced by activation of the wild-type hDOR by morphine, peptide agonists or SNC80 …………………………………………………………… 125

Table 8.2. Maximal effects produced by activation of the wild-type hDOR, the hDOR(S363A) or the truncET-hDOR by peptide agonists or SNC80 ………………... 126

LIST OF FIGURES

Fig. 1.1. Agonist-mediated regulation of the human delta opioid receptor……………. 127 Fig. 2.1. Molecular structures of the delta opioid receptor agonists……………………128 Fig. 2.2. Molecular structure of the human delta opioid receptor……………………... 129 Fig. 3.1. Agonist-mediated phosphorylation of the human δ-opioid receptor at S363…. 130 Fig. 3.2. Dose-response of morphine- and deltorphin II-mediated phosphorylation of the hDOR at S363……………………………………………………………………..... 131 Fig. 3.3. Time course of morphine- and deltorphin II-mediated phosphorylation of the hDOR at S363…………………………………………………………………………… 132 Fig. 3.4. Time course of hDOR(S363) dephosphorylation after morphine or deltorphin II treatment……………………………………………………………………………….. 133 Fig. 4.1. Deltorphin II, DPDPE and SNC80 but not morphine stimulate translocation of β-arrestin2-GFP in CHO cells expressing the hDOR………………………………. 134 Fig. 4.2. β-arrestin2-GFP does not internalize after agonist treatment into intracellular compartments………………………………………………………………………….. 135 Fig. 4.3. After agonist treatment β-arrestin2-GFP associates with endosomes only at or near the plasma membrane…………………………………………………………. 136 Fig. 4.4. Deltorphin II, DPDPE, SNC80 but not morphine stimulate translocation of β-arrestin2-GFP to the hDOR(S363A)………………………………………………… 137 Fig. 4.5. SNC80 but not deltorphin II promotes translocation of β-arrestin2-GFP to the truncated hDOR……………………………………………………………………. 138 Fig. 5.1. Fluo-deltorphin is internalized in CHO cells expressing the hDOR in a time- dependent manner……………………………………………………………………… 139 Fig. 5.2. hDOR is internalized via a subset of clathrin-coated pits distinct from transferrin-containing CCPs……………………………………………………………. 140 Fig. 5.3. hDOR-EGFP is internalized in CHO cells upon deltorphin II and SNC80 treatment……………………………………………………………………………….. 141 Fig. 5.4. [Gln4]deltorphin-rhodamine is internalized in CHO cells expressing either the wild-type or S363A mutant hDOR…………………………………………………….. 142 Fig. 6.1. SNC80 down-regulates the hDOR in a concentration and time dependent manner…………………………………………………………………………………. 143 Fig. 6.2. The human δ-opioid receptor down-regulation is agonist specific………….. 144 Fig. 6.3. Hypertonic sucrose prevents agonist-mediated hDOR down-regulation……. 145 Fig. 6.4. Lactacystin increases the number of human δ-opioid receptors in agonist- treated and untreated cells……………………………………………………………… 146 Fig. 6.5. Lactacystin has no effect on DPDPE and SNC80 mediated down-regulation of human δ-opioid receptors when new protein synthesis is blocked by cycloheximide… 147 Fig. 6.6. MG132 has no effect on DPDPE mediated down-regulation of human δ-opioid receptors after 4h of treatment……………………………………………….. 148 Fig. 6.7. Inactivation of the ubiquitin conjugating enzyme E1 in ts20 cells has no effect on DPDPE and SNC80 mediated down-regulation of the hDOR……………………… 149 10

LIST OF FIGURES - Continued

Fig. 6.8. Folimycin attenuates DPDPE mediated down-regulation of human δ-opioid receptors……………………………………………………………………………….. 150 Fig. 6.9. Truncation of the C-terminus of the human δ-opioid receptor specifically eliminates down-regulation by DPDPE but not by SNC80……………………………. 151 Fig. 6.10. Mutation of S363 to Ala in the human δ-opioid receptor specifically attenuates down-regulation by DPDPE…………………………………………………………… 152 Fig. 6.11. Mutation of S249 and S255 to Ala in the human δ-opioid receptor does not affect down-regulation by SNC80…………………………………………………….. 153 Fig. 7.1. Pretreatment of hDOR/CHO cells with deltorphin II leads to time-dependent desensitization of deltorphin II-stimulated GTPγ[35S] binding and inhibition of cAMP production……………………………………………………………………………… 154 Fig. 7.2. Mathematical simulation of the effect of agonist pre-treatment on concentration- response curves………………………………………………………………………… 155 Fig. 7.3. Time course of deltorphin II-stimulated desensitization of hDOR fitted by the modified operational model (equation 7.5)……………………………………………. 156 Fig. 7.4. S363A mutation attenuates deltorphin II-mediated hDOR desensitization….. 157 Fig. 7.5. Desensitization of the hDOR and hDOR(S363A) fitted by the operational model……………………………………………………………………………………158 Fig. 7.6. hDOR is desensitized more efficiently by SNC80 than by deltorphin II pretreatment……………………………………………………………………………. 159 Fig. 7.7. hDOR(S363A) is desensitized by SNC80 but not by deltorphin II pretreatment……………………………………………………………………………. 160

LIST OF ABBREVIATIONS

AP1 ……………….. Adaptor protein 1 ATP ……………….. Adenosine 5'-triphosphate β2AR ……………… β2 adrenergic receptor CaMK ……………… Calmodulin protein kinase cAMP ……………… 3'-5'-cyclic adenosine monophosphate CCK ……………… Cholicystokinin CREB ……………… cAMP response element binding protein DOR ……………….. Delta opioid receptor DRG ……………….. Dorsal root ganglion DRY motif ………… Aspartic acid, arginine, tyrosine DSLL motif ……….. Aspartic acid, serine, leucine, leucine EGFP ………………. Enhanced grean fluorescent protein ERK ……………….. Extracellular receptor activated kinase GFP ………………... Green fluorescent protein GPCR ……………… G protein-coupled receptor GRK ……………….. G protein-coupled receptor kinase hDOR ……………… Human delta opioid receptor IMDM ……………... Iscove’s Modified Dulbecco’s Medium KOR ……………….. Kappa opioid receptor LDCV ……………… Large dense core vesicle LENLEA motif …… Leucine, glutamic acid, asparagine, leucine, glutamic acid, alanine mAChR …………… Muscarinic acetylcholine receptor MAPK …………….. Mitogen-activated protein kinase MOR ………………. Mu opioid receptor PBS ……………….. Phosphate-buffered saline PD …………………. Ca2+, Mg2+ deficient phosphate-buffered saline PDZ domain ………. PDZ is an acronym combining the first letters of three proteins: post synaptic density protein (PSD95), Drosophila disc large tumor suppressor (DlgA), and zo-1 protein PKA ……………….. Protein kinase A PKC ……………….. Protein kinase C TRPV channel……… Transient receptor potential vanilloid type channel

ABSTRACT

Regulation of the human delta opioid receptor (hDOR) is implicated in the development

of tolerance to chronic morphine (Zhu et al., 1999). In addition, DORs are promising analgesic targets for the management of chronic pain states such as inflammatory or neuropathic pain (Cahill et al., 2007). Therefore, in this study, we investigated multiple aspects of hDOR regulation, including receptor phosphorylation, β-arrestin binding,

receptor internalization, down-regulation and desensitization, using recombinant Chinese

hamster ovary (CHO) cells expressing the wild-type or various mutant hDOR constructs.

We found that structurally diverse delta opioid agonists regulate the hDOR by different

mechanisms. We demonstrate that morphine is able to activate the initial step of the

regulatory events, phosphorylation of S363, but due to requirements for simultaneous

activation of multiple sites, morphine fails to promote β-arrestin binding, receptor

internalization and down-regulation. We also report that peptide delta opioid receptor

agonists and a non-peptide agonist SNC80 differ in their ability to down-regulate the

hDOR. Further differences in receptor phosphorylation, desensitization and β-arrestin

translocation between these two classes of full DOR agonists are reveled by truncation of

the receptor’s C-terminus or by mutation of the primary phosphorylation site, S363.

Studies using the mutant receptors identify the C-terminus as the important domain for hDOR phosphorylation, β-arrestin binding and down-regulation by both peptide and non- peptide agonists. S363 within the C-terminus is critically involved in receptor

phosphorylation, desensitization and down-regulation, but not in β-arrestin binding and 13

receptor internalization. In contrast to peptide agonists, SNC80 is able to phosphorylate and activate secondary intracellular domain(s), in addition to the C-terminus, which

participate in β-arrestin recruitment and receptor desensitization and down-regulation.

Therefore, agonist-specific differences were detected for multiple regulatory events

between morphine, peptide agonists and SNC80. Differential agonist-mediated regulation

of the human delta opioid receptor may be used to design pain therapy drugs with

improved analgesic properties and minimal side effects.

1. INTRODUCTION

Opioid receptors

Opioid receptors belong to the G protein-coupled receptor (GPCR) family, the largest family of membrane proteins in the human genome. Based on molecular cloning and pharmacological studies, three types of opioid receptors have been characterized: the μ- opioid receptor (MOR), the δ-opioid receptor (DOR) and the κ-opioid receptor (KOR)

(Kieffer and Gaveriaux-Ruff, 2002). Opioid receptors are expressed in the peripheral nervous system on pre- and post-synaptic sites in the spinal cord dorsal horn and on primary afferent neurons, as well as in the brain. The primary physiological function of the opioid receptors is the modulation of pain sensation. Alleviation of pain (analgesia) is achieved through activation of opioid receptors by binding of either endogenous or exogenous opioid agonists. Endogenous opioids (enkephalins, dynorphins, β-endorphin and endomorphins) (Przewlocki and Przewlocka, 2001) are neuropeptides synthesized in the body and released under pain conditions such as stress, injury and exercise (Drolet et al., 2001). Exogenously administered opioid agonists (morphine, levorphanol, oxymorphon, meperidine, oxycodone, fentanyl) (Inturrisi, 2002) represent the best analgesic drugs currently used for the treatment of moderate to severe pain. Regrettably, after prolonged treatment, these drugs produce serious side effects, including physical dependence and tolerance. Therefore, understanding the mechanism of sustained opioid- mediated effects is critical for improving the properties of clinically used analgesics by minimizing their undesirable effects. 15

Upon stimulation, opioid receptors produce their effect by activating the inhibitory Gi/o

proteins, and coupling to a number of signaling pathways involving ion channels (GIRK

type K+ channels and Ca2+ channels), protein kinases (PKA, PKC, MAPK) and

transcription factors (CREB, AP1). This acute signaling is extensively regulated by

receptor phosphorylation, desensitization, internalization, degradation and expression

(Borgland, 2001). Acute opioid signaling leads to hyperpolarization of the neuron and attenuation of neurotransmitter release, resulting in a decreased pain transmission. On the

other hand, prolonged opioid signaling leads to regulatory events, which are thought to

contribute to the development of opioid tolerance (von Zastrow, 2004a; von Zastrow,

2004b; von Zastrow et al., 2003). Successful long-term pain therapy must therefore

consider not only the analgesic effects of a drug but also its adverse effects upon

sustained treatment, such as opioid tolerance. The exact cellular mechanisms underlying

opioid tolerance, however, are not entirely understood. In addition, while considerable

attention has been given to the role of the MOR in pain therapy, much less is known

about the involvement of the delta opioid receptors.

Role of the delta opioid receptor in pain treatment

Studies on genetically altered mice have established that analgesics currently used in a

clinical setting, such as morphine, produce their effects primarily by activating the mu

opioid receptor (Kieffer, 1999). In spite of this, accumulating evidence points to the 16

usefulness of drugs acting on the delta opioid receptor in some clinical pain syndromes.

In addition, DORs are critically involved in the development of tolerance to morphine.

DOR-mediated analgesia. DOR-selective agonists are known to produce analgesia in

animal models of chronic pain including inflammatory (Cahill et al., 2003; Fraser et al.,

2000; Hurley and Hammond, 2000), neuropathic (Kabli and Cahill, 2007), and cancer

(Brainin-Mattos et al., 2006) pain. Animal studies indicate that delta opioid receptor selective drugs exhibit analgesic activity but produce less adverse side effects than MOR- acting drugs (Mika et al., 2001; Rapaka and Porreca, 1991). Furthermore, both DOR agonists (Porreca et al., 1992) and antagonists (Gomes et al., 2004) potentiate morphine- mediated analgesia. Interestingly, chronic inflammatory pain was shown to enhance plasma membrane expression of delta opioid receptors in the dorsal spinal cord neurons

(Cahill et al., 2003) and the dorsal root ganglion (DRG) neurons (Gendron et al., 2006) of

injured animals. Additionally, DORs are up-regulated in DRG neurons from neuropathic

pain model rats (Kabli and Cahill, 2007). Neuropathic pain, resulting from peripheral

nerve injury, is particularly difficult to treat with conventional MOR acting analgesics.

Therefore, delta-agonists are attractive alternatives to morphine for treatment of chronic

inflammatory and neuropathic pain. Indeed, increased expression of DORs in a model of

chronic inflammation enhances DOR-mediated spinal analgesia (Cahill et al., 2003;

Gendron et al., 2006). Furthermore, local peripheral administration of deltorphin II

reversed mechanical allodynia (perception of pain to normally innoccuous stimulus) in a

model of neuropathic pain (Kabli and Cahill, 2007). 17

Role of the DOR in analgesic tolerance. Chronic treatment with opioids leads to development of analgesic tolerance, defined as a decrease in pain relief. Therefore, higher doses of the drug are needed to maintain the same analgesic effect (King et al., 2005).

Although acute morphine administration produces analgesic effects by activating the

MORs, the DORs have been shown to play essential roles in the development and maintenance of analgesic tolerance to chronic morphine administration. Support for this conclusion comes from observations that morphine tolerance is abolished by administration of DOR specific antagonists (Abdelhamid et al., 1991) and in DOR knock out mice (Zhu et al., 1999).

The cooperativity of MOR and DOR receptors on the development of analgesic tolerance may be explained by the formation of MOR-DOR heterodimers (Gomes et al., 2004) in which the DOR seems to negatively regulate MOR function. Furthermore, plasma membrane expression of DORs was increased in the dorsal horn neurons of animals chronically treated with morphine (Morinville et al., 2004) or with a selective MOR agonist fentanyl (Cahill et al., 2001). The translocation of DORs from intracellular to plasma membrane compartments upon sustained morphine treatment was also observed in neurons from some brain regions such as the nucleus accumbens and the dorsal neostriatum but not from the frontal cortex (Lucido et al., 2005). Importantly, it was found that the enhanced DOR expression has functional consequences, because delta opioid agonists inhibited GABA neurotransmitter release in the brainstem neurons from 18

morphine-tolerant rats, but not from opioid naïve rats (Ma et al., 2006). Moreover, delta-

selective opioid agonists did not inhibit Ca2+ currents in DRG neurons under basal conditions, but up-regulation of cell surface DORs induced their coupling to Ca2+ channels (Walwyn et al., 2005). Therefore, chronic MOR agonist treatment induces up- regulation of functional delta opioid receptors. This increased DOR expression may negatively alter MOR function by formation of MOR-DOR heterodimers, leading to reduced analgesic effects of μ-agonists. On the other hand, up-regulation of the DOR may facilitate δ-opioid receptor-mediated analgesia. Indeed, chronic (48h) morphine treatment potentiated intrathecal [D-Ala2]deltorphin II-induced antinociception (Cahill et al., 2001). Similarly, activation of DORs in the brainstem has no effect in opioid naïve rats, but produced analgesia in morphine-tolerant rats and reduced morphine tolerance

(Ma et al., 2006).

To summarize, DOR agonists may provide a better alternative to currently used

analgesics for the treatment of chronic pain states such as inflammatory and neuropathic

pain or in patients tolerant to morphine. Increasing evidence suggests that the underlying

cause for the enhanced efficacy of DOR agonists under those pathological conditions is

the altered expression of the DOR receptors at the plasma membrane. Therefore, an

investigation of the mechanisms of DOR regulation is important for understanding the

delta opioid receptor function in the treatment of pathological pain states.

Regulation of DOR expression

The expression of the delta opioid receptor at the cell surface is regulated on multiple levels. It was demonstrated that the expression of the DOR gene is tightly controlled during development and differentiation by activation of PI-3 kinase/Akt/ NF-κB signaling pathway (Chen et al., 2006). Whether this pathway plays a role in transcriptional regulation of the DOR during chronic morphine treatment or in pathological pain states is not known. Newly synthesized DORs are delivered to the plasma membrane via a constitutive biosynthetic pathway and by a stimulus-regulated pathway. When plasma membrane localized receptors are stimulated by an agonist, they are removed from the cell surface and either degraded or recycled back to the membrane. Therefore, total expression of the DORs at the cell surface is determined by the balance between receptor synthesis, delivery, internalization, recycling and degradation.

Delivery of the DOR to the plasma membrane. Until recently, DORs were assumed to be delivered to the plasma membrane by the constitutive biosynthetic pathway. In this pathway, native protein is synthesized and assembled in the endoplasmic reticulum and delivered to the Golgi apparatus, where it undergoes posttranslational modifications such as glycosylation. Mature protein is then packaged into exocytic vesicles and constitutively inserted into the membrane. While most GPCRs, including the MOR, are processed primarily by this mechanism, a large pool of newly synthesized DORs is associated with large dense-core vesicles (LDCV) distributed in the cytoplasm of the neurons. These secretory granules are localized inside of the cell under basal conditions 20

and are transported to the plasma membrane upon various stimuli (Cahill et al., 2007).

Delta opioid ligands were shown to promote maturation of DORs and their trafficking to the plasma membrane, thereby acting as “chaperones” for DOR delivery (Petaja-Repo et al., 2002). Prolonged in vivo stimulation of the MOR results in targeting of the DOR to the plasma membrane. This up-regulation of plasma membrane DORs was not observed in MOR-/- mice, indicating that MORs play a crucial role in this process (Morinville et al., 2003). Finally, pronociceptive agents, such as capsaicin, bradykinin, and ATP, also promote trafficking of the DOR from intracellular stores to the plasma membrane (Zhang et al., 2006). Pronociceptive stimulus-induced delivery of DORs to the cell surface may explain why chronic inflammatory or neuropathic pain causes enhanced membrane expression of the DOR.

It was demonstrated recently, that in dorsal root ganglion neurons the DOR is sorted into

protachykinin-containing LDCVs (Guan et al., 2005). Direct interaction of the DOR with

the substance P domain of protachykinin is required for sorting of the receptor into

LDCVs, since it was abolished in protachykinin gene knock out mice. Moreover,

protachykinin is required for stimulus-induced insertion of DORs into the plasma

membrane. Importantly, protachykinin is also required for enhanced deltorphin I-

mediated spinal analgesia.

Based on these findings it is evident that plasma membrane expression of the DOR is a

dynamic process regulated by μ and δ opioid ligands and by pronociceptive agents. The 21

current model for plasma membrane delivery of the DOR involves packaging of the

DORs in the Golgi network into neuropeptide-containing secretory vesicles. Exocytosis

of these vesicles and insertion of DORs is stimulated by activation of some G protein-

coupled receptors (MOR, bradykinin receptor), ion channels (capsaicin-stimulated TRPV

channels, ATP channels) or increased intracellular Ca2+. Inserted DORs are functional and able to induce analgesic effects in vivo. However, at the plasma membrane, the

DORs are regulated by δ-opioid agonist activation. Therefore, the existing status of

DORs is determined by a balance between delivery (constitutive and stimulus-activated) of DORs to the plasma membrane and agonist-mediated regulation of plasma membrane- expressed DORs. The latter process involves receptor phosphorylation, β-arrestin binding, receptor internalization, recycling and down-regulation, as will be described in the following section.

Agonist-mediated regulation of the DOR. Many cellular events are triggered by binding

of opioid agonists to the delta opioid receptor. In addition to activating multiple second

messenger signaling pathways, the DOR itself undergoes a number of regulatory events

(Fig. 1.1) whose main purpose is to modulate responsiveness of the cell to further

extracellular stimuli (Daaka et al., 1997; Hall and Lefkowitz, 2002). Activation of the

DOR by an agonist facilitates the binding of G protein-coupled receptor kinases (GRK)

(Li et al., 2003) recruited to the plasma membrane by βγ subunits of the activated Gi/o protein (Quock et al., 1999). GRKs then phosphorylate the activated receptor at several residues within the C-terminal region of the DOR. Among these residues S363 has been 22

identified as the primary phosphorylation site (Kouhen et al., 2000). Agonist-bound and phosphorylated receptors display high affinity for the cytosolic adaptor proteins, β-

arrestins. Binding of β-arrestins assists in the targeting of the receptor to clathrin-coated

pits and receptor internalization via the endocytosis of the clathrin-coated vesicles. The internalization process may serve either to degrade the receptor in cellular digestive organelles, thereby terminating receptor function (down-regulation), or to reintroduce the receptor to the plasma membrane (resensitization).

While the nature of agonist-mediated regulation is conserved for most GPCRs, individual

differences exist in the mechanism or in the rate and extent of specific events. Therefore, even though most GPCRs are phosphorylated in response to agonist activation, there is no conserved sequence or region for this phosphorylation and even receptor kinases involved in receptor phosphorylation may differ. Affinities of β-arrestins to specific

GPCRs vary, which led to the classification of GPCRs into two classes: class A receptors preferably bind to β-arrestin2, while class B receptors bind with high affinities to both β- arrestin1 and 2 (Oakley et al., 2000). Association of β-arrestin with class A receptors is weak, and β-arrestin dissociates quickly after internalization. On the other hand, the complex between a class B receptor and β-arrestin remains stable after receptor internalization. The strength of the receptor-β-arrestin interaction is thought to have consequences on internalization of these receptors. Although most receptors endocytose via an arrestin- and clathrin coated pit-mediated mechanism, other pathways were described for example for muscarinic acetylcholine receptors mAchR (Lee et al., 1998) 23

and cholecystokinin CCK receptors (Roettger et al., 1995). Sorting of the receptors after

endocytosis into recycling and degradative pathways is presumably dependent on sorting

signals expressed selectively on different GPCRs. Thus, the μ-opioid and the β2

adrenergic receptors are preferably sorted into the recycling pathway due to specific

recycling signals present in their intracellular domains; LENLEA, (Tanowitz and von

Zastrow, 2003) and DSLL, (Gage et al., 2001), for the MORs and β2 adrenergic

receptors, respectively. On the other hand, degradation motifs present in the DOR (Cvejic

et al., 1996; Wang et al., 2003) are responsible for preferential trafficking of this receptor

into degradation pathways. Nevertheless, insertion of the LENLEA or DSLL recycling

signals into the DOR sequence is sufficient to redirect the receptor from degradation into the recycling pathway (Gage et al., 2001; Tanowitz and von Zastrow, 2003). These findings indicate that sorting of internalized receptors is regulated by a complex multi signal mechanism. Additional sorting may take place during trafficking of the receptors into separate digestive systems (lysosomes, proteasomes), as was demonstrated for down- regulation of the DOR (Chaturvedi et al., 2001) and the KOR (Li et al., 2000).

In addition to receptor-specific differences, the same G protein-coupled receptor may also

be differently regulated by various agonists. Using fluorescence lifetime analysis it was

demonstrated that depending on the bound agonist, the β2-adrenergic receptor-ligand

complex may stabilize diverse active receptor conformations (Ghanouni et al., 2001).

Mutational analyses demonstrate that the third extracellular loop of the δ-opioid receptor

selectively affects SNC80-mediated affinity and intrinsic activity while peptide agonist 24

(DPDPE) binding and activity were not changed (Hosohata et al., 2001). Therefore, agonist-selective receptor conformations could be responsible for some differences observed in agonist-mediated regulation of GPCRs. Probably the best characterized are the differences between morphine and other μ-opioid agonists in regulation of the MOR.

In HEK 293 cells (Keith et al., 1996) and in neurons (Keith et al., 1998; Sternini et al.,

1996), etorphine but not morphine was shown to induce μ-opioid receptor internalization.

Morphine, a partial agonist at the DOR, was also found very inefficient in phosphorylating the DOR (Zhang et al., 1999), recruiting β-arrestin to the receptor

(Whistler et al., 1999), and promoting DOR internalization (Keith et al., 1996; Kramer and Simon, 2000). Additionally, agonist-specific differences in the ability to induce hDOR internalization were also found in SK-N-BE cells between a peptide agonist

DPDPE and an alkaloid etorphine (Marie et al., 2003). These authors also demonstrate that even though prolonged treatment of SK-N-BE cells with peptide agonists targets the hDOR into lysosomes, treatment with etorphine results in recycling of the receptor.

Therefore, structurally different agonists may have differential effects on the mechanism or the extent of receptor regulation, which may result in different fates of the receptor.

Hypothesis

As reviewed above, cellular adaptations occurring in response to chronic morphine treatment and during pathological pain conditions include regulation of the plasma membrane expression of the DOR. Agonist-specific differences were identified in both the extent and the mechanism of DOR regulation among morphine, alkaloid agonists and 25 peptide agonists. However, the underlying molecular basis for such agonist-selective regulation is not completely understood. We hypothesized that structurally distinct delta- opioid agonists may activate agonist-selective receptor domains resulting in differential mechanisms of hDOR regulation. We anticipate that differential regulation of the hDOR by various opioid agonists will have functional consequences on the development of analgesic tolerance. Therefore, in the future, based on the regulatory profiles induced by specific agonists, opioid analgesics with minimal side-effects can be developed.

Goals

To corroborate our hypothesis, we examined regulation of the human delta-opioid receptor in a model system- Chinese hamster ovary (CHO) cells expressing recombinant hDOR constructs. We studied cellular mechanisms of several events, which contribute to receptor regulation: receptor phosphorylation (Chapter 3), β-arrestin recruitment (Chapter

4), receptor internalization (Chapter 5), receptor down-regulation (Chapter 6), and receptor desensitization (Chapter 7). For each regulatory event we investigated the molecular determinants of hDOR regulation using C-terminally truncated hDOR or several mutant hDOR constructs. Furthermore, we studied agonist-mediated differences in the cellular mechanisms of hDOR regulation and in molecular domains involved in this regulation, using three classes of DOR agonists: morphine, peptide agonists (DPDPE, deltorphin II), and a non-peptide agonist SNC80.

2. MATERIALS AND METHODS

Delta opioid receptor ligands

Cyclic[D-Pen2, D-Pen5]enkephalin (DPDPE) was synthesized in the laboratory of Dr.

Hruby, University of Arizona, Tucson, AZ; (D-Ala2)deltorphin II (deltorphin II) and (+)-

4-[(αR)-α-((2S,5R)-4-allyl-2,5-dimethyl-1-piperazinyl)-3-methoxybenzyl]N,N-

diethylbenzamide (SNC80) were from Tocris; and morphine sulphate (morphine) is from

Sigma-Aldrich. The chemical structures of these agonists are given in Fig. 2.1. In

GTPγ[35S] assay these agonists dose-dependently stimulated GTPγ[35S] binding with the

following Emax (as compared to maximum stimulation by SNC80) and EC50 values: deltorphin II (Emax = 100%, EC50 = 6.3 nM), DPDPE (Emax = 100%, EC50 = 26 nM),

SNC80 (Emax = 100%, EC50 = 11 nM), morphine (Emax = 80%, EC50 = 2.5 μM). In cAMP

assays these agonists dose-dependently inhibited forskolin-stimulated cAMP production

with the following Emax (as compared to maximum inhibition by SNC80) and EC50 values: deltorphin II (Emax = 100%, EC50 = 0.65 nM), SNC80 (Emax = 100%, EC50 = 0.64 nM), morphine (Emax = 89%, EC50 = 0.15 μM).

Other DOR ligands were obtained from the following sources: naltrindole hydrochloride

(naltrondole, Research Biochemicals International); naltrexone hydrochloride (naltrexon,

Tocris); [3H]naltrindole (PerkinElmer Life Sciences); [3H](D-Ala2)deltorphin II

(PerkinElmer Life Sciences); fluo-Deltorphin (Advanced Bioconcept Ltd.);

[Gln4]deltorphin-rhodamine was synthesized by solid phase synthesis using the Nα-Fmoc 27

strategy: [Gln4]deltorphin (Tyr-D-Ala-Phe-Gln-Val-Val-Gly) was conjugated to a

fluorescent label, tetramethylrhodamine isothiocyanate (Molecular Probes) via the Nε-

Lys residue present in the β-Ala-Gly-β-Ala-Gly-Lys spacer that was included at the C- terminus of [Gln4]deltorphin. In whole cell radioligand binding assay the rhodamine

4 labeled [Gln ]deltorphin exhibited a high affinity for the hDOR with an IC50 = 11 ± 2 nM.

hDOR cDNA constructs and cell lines

hDOR. The amino-acid sequence of the human delta opioid receptor (NCBI accession no:

NP_000902) is indicated in Fig. 2.2. Two Chinese Hamster Ovary cell lines expressing

the human δ-opioid receptor (subcloned into NheI and XhoI sites of pREP10 expression

vector (Invitrogen)) were used in our studies. These cell lines differ in the receptor

expression level: low expression cell line 1-209-19hDOR/CHO exhibited a BmaxB of 1.8 pmol/mg protein (Kd = 135 pM) and high expression cell line 2-209-19hDOR/CHO had a

3 BmaxB value of 6.5 pmol/mg protein (Kd = 130 pM) in [ H]naltrindole radioligand saturation binding assays. The previously characterized (Malatynska et al., 1996) high expression hDOR/CHO cell line was used in all experiments except in desensitization studies (Chapter 7) where the low expression cell line 1-209-19hDOR/CHO was used.

For mutagenesis experiments the hDOR was subcloned into EcoRI and ApaI sites of the pcDNA3.1 expression vector (Invitrogen).

hDOR(S363A). A mutation (Ser363 to Ala) was introduced into the hDORpcDNA3.1

vector using the Quick Change site-directed mutagenesis method (Stratagene) with

addition of 1M betaine (Sigma) using 5’-CCT GCA CCC CGG CCG ATG GTC C-3’ and

5’-GGA CCA TCG GCC GGG GTG CAG G-3’ primers. The presence of the Ser363 to

Ala mutation was verified by nucleotide sequencing at the University of Arizona

sequencing facility. A stable cell line expressing the human δ-opioid receptor (S363A)

mutant was generated by transfecting the hDOR(S363A) pcDNA3 construct into CHO

cells via FuGENE 6 (Roche Diagnostics) and G418 selection. In [3H]naltrindole

radioligand binding assays the selected recombinant CHO cell line expressing the human

δ-opioid receptor (S363A) mutant (hDOR(S363A)/CHO#13) displayed a BmaxB value of

2.0 pmol/mg (Kd=110 pM) (Navratilova et al., 2004). In functional assays the

hDOR(S363A)/CHO#13 cells did not show any alteration in activation of GTPγ[35S]

binding and inhibition of forskolin-stimulated cAMP formation as compared to the

hDOR/CHO cells.

hDOR(S249A/S255A/S363A). S249A and S255A mutations were introduced sequentially into the hDOR(S363A) pcDNA3.1 vector using the Quick Change site-directed mutagenesis method (Stratagene) with addition of 1M betaine (Sigma). 5’-TTC TCC

TTG GCG CCC GAC AGC A-3’ and 5’-TGC TGT CGG GCG CCA AGG AGA A-

3’primers were used for the S249A mutation and 5’-GCA GGG CGC GGT CCT TCT

CCT-3’ and 5’-GGT GTG CAT GCT CCA GTT CC-3’ primers were used for the S255A mutation. The presence of the S249A and S255A mutations was verified by nucleotide 29

sequencing at the University of Arizona sequencing facility. To ensure that no other mutations were accidentally created in the expression vector, a BstEII-NotI fragment of

the hDOR sequence, which includes the two mutated sites, was excised and ligated with

the complementary fragment of an undisturbed hDOR(S363A) pcDNA using T4 DNA

ligation kit (Roche Diagnostics). The accuracy of the ligation and the presence of all

three mutations were verified by nucleotide sequencing at the University of Arizona

sequencing facility. A stable cell line expressing the human δ-opioid receptor

(S249A/S255A/S363A) mutant was generated by transfecting the

hDOR(S249A/S255A/S363A) pcDNA3.1 construct into CHO cells via FuGENE 6

(Roche Diagnostics) and G418 selection. In [3H]naltrindole radioligand saturation

binding assay the recombinant CHO cell line expressing the mutant receptor

(hDOR(S249A/S255A/S363A)/CHO) bound with BmaxB =1.6 pmol/mg, Kd=38 pM.

TruncEt-hDOR. The Chinese hamster ovary cell line, expressing the hDOR truncated

after Gly338 and ligated in-frame into the pcDNA3.myc-His B+ expression vector

(Invitrogen), (truncEt-hDOR/CHO), has been described and characterized previously in

(Okura et al., 2003).

hDOR-EGFP. To prepare the hDOR with a fluorescent enhanced green fluorescent

protein (EGFP) epitope tag at the C-terminus, the hDOR open reading frame, containing

a mutated stop-codon (Okura et al., 2003) has been excised from the pcDNA3.myc-His

B+ shuttle vector, using HindIII and ApaI restriction enzymes and inserted in frame into 30

the HindIII and ApaI sites of the pEGFP-N3 vector (Invitrogen) using the T4 DNA ligation kit (Roche Diagnostics). The accuracy of the ligation was verified by nucleotide sequencing at the University of Arizona sequencing facility. A stable cell line expressing

the human δ-opioid receptor with the EGFP epitope tag was generated by transfecting the hDOR-EGFP construct into CHO cells via FuGENE 6 (Roche Diagnostics) and G418 selection. hDOR-EGFP/CHO#2 cell line was chosen based on the morphology of the cells and the hDOR-EGFP expression pattern using confocal fluorescent microscopy.

Cell culture

Recombinant Chinese hamster ovary (CHO) cells expressing the wild-type hDOR or different mutant hDOR constructs were grown and maintained in Ham’s-F12 medium containing 10% fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin and 800

µg/ml of hygromycin (hDOR/CHO) or 500 µg/ml G418 (all other cell lines) at 37°C in a

5% CO2 humidified atmosphere.

Western blot analysis

hDOR/CHO or hDOR(S363A)/CHO cells were treated for indicated time intervals with

delta opioid receptor agonists at 37°C in a serum-free IMDM medium. After agonist

treatment the cells were harvested in homogenization buffer (20 mM Tris, 4 mM EGTA,

2 mM EDTA, pH 7.5) containing 1% protease inhibitor cocktail and 0.1% phosphatase

inhibitor cocktails I and II (Sigma). Cell lysates were sonicated and boiled for 5 min with

reducing sample buffer (Invitrogen). Total protein concentration in each sample was 31

determined by the Bradford assay. Sample aliquots containing 5 μg total protein were

resolved on 10% polyacrylamide gels (NuPAGE, Invitrogen) and transferred to nitrocellulose membranes. Phosphorylation of the human δ-opioid receptor at S363 was measured by Western blots using a rabbit primary antibody specific for phospho-S363 of the δ-opioid receptor (1:1,000, Cell Signaling) and a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:100,000, Santa Cruz Biotechnology).

Immunoreactive bands were detected using SuperSignal West Dura chemiluminescent kit

(Pierce) and quantified using the Image J software (NIH). Statistical analyses were performed using the GraphPad Prism 4 software.

Confocal microscopy

For visualization of β-arrestin2 recruitment, CHO cells expressing the wild-type hDOR, the S363A mutant hDOR, or the truncEt-hDOR were grown on glass bottom chamber slides and transiently transfected with a green fluorescent protein-tagged β-arrestin2 construct (β-arrestin2-GFP, a gift from Dr. Lefkowitz) using the FuGENE 6 transfection protocol (Roche Diagnostics). 48h after transfection, the cells were left untreated or treated with delta opioid receptor agonists for indicated time intervals at 37°C. For visualization of receptor internalization, the hDOR/CHO or hDOR(S363A)/CHO cells were incubated in serum-free IMDM medium with 300 μM Fluo-deltorphin II (Fluo-

Peptides) or 1 μM [Gln4]deltorphin-rhodamine for indicated time intervals at 37°C.

Selective DOR antagonist naltrindole (1 μM, IC50 = 60 pM, Research Biochemicals

International) was added to some wells to determine DOR specificity. After treatment, 32

the cells were rinsed with ice-cold PBS, fixed with 4% paraformaldehyde and mounted in

a Vectashield Mounting Medium (Vector Laboratories). Slides were examined under a

Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective using excitation/emission filter sets 488/505 nm or 543/560 nm. Single optical sections were acquired through the trans-nuclear plane. The acquisition parameters were constant in all parallel experiments. The images were processed using the Adobe

Photoshop 7.0 software.

[3H] naltrindole radioligand binding assay

To measure receptor down-regulation the proportion of receptors remaining in the cells

after prolonged agonist treatment was measured by [3H] naltrindole radioligand binding.

CHO cells stably expressing the wild-type or various mutant human δ-opioid receptors were grown in 10 mm cell culture dishes to confluency and treated for the indicated time periods with delta opioid receptor agonists. After three washes with Ca2+, Mg2+ deficient

phosphate-buffered saline (PD), cells were harvested in PD buffer containing 0.02%

EDTA. After centrifugation at 1500 × g for 10 min, the cells were homogenized in ice-

cold 10 mM Tris-HCl, 1 mM EDTA, pH 7.4 buffer. A crude membrane fraction was collected by centrifugation at 40,000 × g for 15 min, and resuspended in binding assay buffer (50 mM Tris-HCl, 5 mM MgCl2, 0.1% bovine serum albumin, 50 μg/ml

bacitracin, 30 µM bestatin, 10 µM captopril and 0.1 mM phenylmethylsulfonyl fluoride,

pH 7.4). Approximately 50 µg of membrane preparation was incubated at 30°C for 90

min in 1 ml binding assay buffer, with 0.5 nM [3H] naltrindole (20 Ci/mmol, PerkinElmer 33

Life Sciences). 10 μM naltrexone was used to assess nonspecific binding. The reaction

was filtered using a Brandel Cell Harvester through Whatman GF/B glass fiber filters.

The filters were washed three times with 4 ml ice-cold saline (0.9% NaCl, pH 7.4). Filter- bound radioactivity was measured in EcoLite scintillation cocktail using a Beckman LS

6000SC liquid scintillation counter. Specific binding was calculated by subtracting from

the total binding the nonspecific binding measured in the presence of 10 μM naltrexone.

Down-regulation was defined as a decrease in specific binding per 1 μg membrane

protein after the agonist treatment and expressed as a percentage of the buffer-treated

control. Data were analyzed using the GraphPad Prizm 4 software.

GTPγ[35S] binding assay

Agonist-stimulated binding of GTPγ[35S] to crude membranes prepared from control and

agonist pretreated hDOR/CHO or hDOR(S363A)/CHO cells was determined as

previously described (Quock et al., 1997). Briefly, cell monolayers were washed with

Ca2+, Mg2+ deficient phosphate-buffered saline (PD) and harvested in PD buffer

containing 0.02% EDTA. After centrifugation at 1500 × g for 10 min, the cells were

homogenized in ice-cold 10 mM Tris-HCl, 1 mM EDTA, pH 7.4 buffer. A crude

membrane fraction was collected by centrifugation at 40,000 × g for 15 min, and

resuspended in GTPγS assay buffer (25 mM Tris-HCl, 150 mM NaCl, 2.5 mM MgCl2, 1

mM EDTA, 10 µM GDP, 30 µM bestatin, 10 µM captopril and 0.1 mM

phenylmethylsulfonyl fluoride, pH 7.4). Approximately 50 µg of membrane preparation

was incubated at 30°C for 90 min in 1 ml GTPγS assay buffer, with appropriate 34 concentrations of deltorphin II, in the presence of 0.1 nM guanosine 5’[γ-[35S] thio] triphosphate (GTPγ[35S], 1250 Ci/mmol, PerkinElmer Life Sciences). The reaction was filtered using a Brandel Cell Harvester through Whatman GF/B glass fiber filters. The filters were washed three times with 4 ml ice-cold 25 mM Tris-HCl, 120 mM NaCl, pH

7.4. Filter-bound radioactivity was measured in EcoLite scintillation cocktail using a

Beckman LS 6000SC liquid scintillation counter. The data were fitted and statistically analyzed using the GraphPad Prizm 4 software.

Inhibition of forskolin-stimulated cAMP formation assay

Cyclic AMP assays were performed according to a method modified from Gilman

(Gilman, 1970). CHO cells expressing the wild-type or S363A mutant hDOR were plated in 24-well plates and grown in Ham’s F12 media containing 10% fetal calf serum and antibiotics to ~ 200,000 cells/well confluency. On the day of the experiment, cells were rinsed with IMDM and incubated in the presence or absence of 100 nM deltorphin II.

After the appropriate time, the cells were washed three times (10 min each) with IMDM, to remove the agonist. Deltorphin II inhibition curves were measured by stimulating the cells with 100 µM water-soluble forskolin (7-deacetyl-7-[O-(N-methylpiperazino)-γ- butyryl]-dihydrochloride, Calbiochem) in the presence of various concentrations of the agonist and a non-selective phosphodiesterase inhibitor, 3-isobutyl-methylxanthine (5 mM), for 20 min, at 37°C, in 5% CO2 humidified atmosphere. The reaction was terminated by rinsing once with ice-cold IMDM, placing the plates on ice and replacing 35

the media with ice-cold TE buffer (50 mM Tris-HCl, 4 mM EDTA, pH 7.5). Cells were

dislodged from the wells using a Costar cell scraper, transferred to microcentrifuge tubes

and lysed by boiling for 10 min. Tubes were centrifuged in a bench top centrifuge for 3

min. 50 µl of each supernatant and 50 µl of a standard (0.125-128 pmol of cAMP) were

separately incubated with 50 µl (0.9 pmol/50 µl) of [3H]cAMP (PerkinElmer, Life

Sciences) and 100 µl of protein kinase A solution (6 µg PKA/100 μl in TE buffer with

0.1% BSA) on ice for 2h. Bound and free [3H]cAMP were separated by adding 100 µl of

activated charcoal in ice-cold TE buffer, containing 2% bovine serum albumin, to each

tube, and centrifuging at 5600 × g for 1 min. The radioactivity in 200 µl aliquots from each supernatant was counted in a Beckman LS 6000SC liquid scintillation counter. The amount of cAMP in the samples was determined by interpolating the cAMP standard

curve using GraphPad Prizm 4 software. 36

3. AGONIST-MEDIATED HUMAN DOR PHOSPHORYLATION

Taken in part from (Navratilova et al., 2005)

Introduction

Agonist-mediated phosphorylation of the hDOR represents an early step in a cascade of regulatory events outlined in Fig. 1.1. Phosphorylation of the δ-opioid receptor has been implicated in rapid desensitization (Pei et al., 1995) as well as in β-arrestin- and clathrin- mediated internalization (Law et al., 2000). Both second messenger-regulated protein kinases (Xiang et al., 2001) and G protein-coupled receptor kinases (GRKs) (Guo et al.,

2000; Hasbi et al., 1998; Li et al., 2003) can phosphorylate and regulate the delta opioid receptor. Intracellular domains of the DOR contain several consensus phosphorylation sites for second messenger-regulated S/T protein kinases. A distinct phosphorylation consensus motif for GRKs has not been identified, but it is expected that S/T residues present in an acidic environment at the C-terminus of the DOR may serve as substrates for GRK-mediated phosphorylation. GRK family of protein kinases consists of seven members, GRK1-7, which share high sequence homology (Reiter and Lefkowitz, 2006).

GRK2/3 are uniquely equipped to selectively regulate only activated receptors because they are recruited from the cytosol to the plasma membrane localized receptors by binding to βγ subunits of heterotrimeric G proteins (Pitcher et al., 1992). Phosphorylation of the DOR by GRK2 (Guo et al., 2000; Hasbi et al., 1998; Li et al., 2003) and heterologous phosphorylation by PKC (Xiang et al., 2001) has been demonstrated. While second-messenger kinases may be important in integrating cellular signals originating 37

from a number of different receptors, GRKs play a role in homologous phosphorylation

and regulation of GPCRs.

Interestingly, in contrast to other opioid agonists, it has been demonstrated that morphine

elicits minimal δ-opioid receptor phosphorylation and β-arrestin translocation (Zhang et

al., 1999), and does not stimulate rapid δ-opioid receptor internalization (Keith et al.,

1996). Overexpression of G protein-coupled receptor kinase 2 facilitated morphine-

mediated δ-opioid receptor phosphorylation (Zhang et al., 1999), suggesting that the morphine-activated δ-opioid receptor is a sub-optimal substrate for G protein-coupled receptor kinases (von Zastrow et al., 2003).

The peptide agonist, DADLE ([D-Ala2, D-Leu5]enkephalin) was shown to phosphorylate

several residues in the C-terminus of the rodent δ-opioid receptor, with S363 as the

primary phosphorylation site (Kouhen et al., 2000). It is presently not known whether

morphine is able to phosphorylate the same residues as full delta opioid agonists and

whether morphine-mediated phosphorylation follows the same kinetics as other delta

opioid agonists. To answer these questions, we used an antibody specific for S363- phosphorylated δ-opioid receptor and investigated phosphorylation of S363 in the human

δ-opioid receptor upon morphine treatment and correlated that with phosphorylation produced by full DOR agonists (SNC80, DPDPE, deltorphin II). In addition, we

compared the time courses of morphine- and deltorphin II-mediated phosphorylation as

well as the time courses of hDOR dephosphorylation after removal of the agonist. We 38

demonstrate that in Chinese hamster ovary cells (CHO) expressing the human δ-opioid receptor, morphine promotes phosphorylation of S363. Moreover, we show that the

kinetics of morphine- and deltorphin II-mediated hDOR(S363) phosphorylation and

dephosphorylation following morphine or deltorphin II treatment are different. Slower

kinetics of hDOR(S363) phosphorylation by morphine and faster kinetics of

dephosphorylation after removal of the drug indicates that complex regulatory

mechanisms are involved in agonist-specific hDOR phosphorylation.

Results

S363 in the C-terminus of the hDOR is phosphorylated by morphine. To study the ability

of different delta opioid agonists to stimulate phosphorylation of the hDOR at S363,

hDOR/CHO cells were treated with SNC80, DPDPE, deltorphin II or morphine for 30

min at 37°C in a serum-free medium. After cell harvesting, crude cell lysates were

prepared and sample aliquots containing equal amounts of total protein were resolved by

10% polyacrylamide gells. Phosphorylation of the receptor at S363 was measured by

Western blots using a rabbit primary antibody specific for phospho-S363 of the human δ- opioid receptor and quantified using the Image J software (NIH). As an internal control, we also quantified actin, a “housekeeping” protein, in each sample, and normalized the intensities of the phospho-S363 bands for actin immunoreactivity. We show that treatment

of hDOR/CHO cells with full agonists SNC80 (100 nM), DPDPE (100 nM) or deltorphin

II (100 nM) for 30 min at 37°C promotes phosphorylation of the human δ-opioid receptor

at S363. This phosphorylation was identified as an increased immunoreactivity of several 39

bands in the molecular weight range of 50-70 kDa (Fig. 3.1, lane 3), corresponding to

differently glycosylated (N-linked or O-linked) forms of the S363-phosphorylated human

δ-opioid receptor (data not shown). Surprisingly, treatment of hDOR/CHO cells with

morphine, a partial agonist at the hDOR, also led to receptor phosphorylation at S363 (Fig.

3.1, lane 2). Quantitative analyses of the 50-70 kDa bands show that morphine (30 μM)-

mediated S363 phosphorylation was 53 ± 8 % of maximal phosphorylation mediated by

deltorphin II (100 nM). Both morphine- and deltorphin II-mediated phosphorylation were

completely blocked by concomitant treatment with the selective δ-opioid receptor

antagonist, naltrindole (1μM) (Fig. 3.1, lanes 4, 5). The anti-phospho-S363 antibody did

not interact with proteins in untreated cell lysates (Fig. 3.1, lane 1), indicating that S363

phosphorylation is negligible under basal conditions. Finally, no specific

immunoreactivity was detected upon morphine or deltorphin II treatment of CHO cells

expressing a human δ-opioid receptor in which S363 was replaced by alanine

(hDOR(S363A)/hDOR) (Navratilova et al., 2004), confirming that the antibody

recognizes only the S363-phosphorylated human δ-opioid receptor (Fig. 3.1, lanes 6, 7).

Phosphorylation of S363 by morphine and deltorphin II is concentration-dependent. To determine the EC50 values for deltorphin II- or morphine-mediated phosphorylation of

S363, we treated the hDOR/CHO cells with increasing concentrations of the selected

agonists. After collecting the cell lysates, Western blots were performed (Fig. 3.2A,B)

and the intensities of bands corresponding to S363 phosphorylated hDOR were quantified

using the Image J software (NIH). The band intensities representing the proportion of the 40

phosphorylated receptors were fitted with a sigmoidal dose response curves with variable

Hill coefficients (Fig. 3.2C,D) and the EC50 values determined. The EC50 values for

morphine- and deltorphin II-mediated phosphorylation were 1.4±0.1 μM and 1.3±0.1 nM,

respectively. These values correlate well with the EC50 values for morphine- and deltorphin II-mediated GTPγS stimulation (2.5±0.7 μM and 6.3±0.4 nM, respectively).

The rates of hDOR(S363) phosphorylation by morphine and deltorphin II differ. Receptor

phosphorylation is a dynamic process dependent on the kinetics of phosphorylation and

dephosphorylation. Both processes in turn depend on the accessibility of the receptor phosphorylation site to protein kinases and phosphatases, a process highly regulated by the agonist and by receptor trafficking. Since morphine, in contrast to deltorphin II, does not promote intensive endocytosis of the receptor, we speculated that the kinetics of morphine- and deltorphin II-mediated phosphorylation of S363 would be different. To

investigate this hypothesis we measured a time course of hDOR phosphorylation at S363 after morphine and deltorphin II treatment (Fig. 3.3A,B). The intensities of hDOR(S363) phosphorylation were quantified and plotted as a function of the time of the agonist treatment. The obtained data for early time points (0-30 min) were fitted with a single exponential association curve (Fig. 3.3C). The half lives of morphine- and deltorphin II- mediated S363 phosphorylation determined from the regression analysis of the fitted curves were 4.2±0.8 min and 1.7±0.3 min, respectively. A statistical analysis using an F- test demonstrates that the two curves are significantly different (p=0.003). Therefore, 41

morphine and deltorphin II phosphorylate S363 within the C-terminus of the hDOR with

different rates.

Morphine- and deltorphin II-mediated hDOR(S363) phosphorylation displays a biphasic

time course. As evident from the time course of S363 phosphorylation (Fig. 3.3), receptor

phosphorylation reaches a maximum after 30-60 min and then receptor

dephosphorylation (still in the presence of the agonist) prevails. This biphasic time course

indicates that the rates of receptor phosphorylation and dephosphorylation are not

constant during the studied time period. Instead the kinetics of receptor phosphorylation

and dephosphorylation change over the time of agonist treatment. The mechanisms

governing this modulation of the agonist response are largely unknown, but may include

changes at the receptor level such as sequestration into cellular microdomains

(compartments) or recruitment of accessory molecules. Alternatively, a feedback mechanism regulating the activity of protein kinases and phosphatases may be present.

Due to the complexity of the kinetics of hDOR(S363) phosphorylation and

dephosphorylation at later time points, interpretation of the experimental data at later times becomes difficult. In order to study receptor dephosphorylation under better defined and simpler conditions, we investigated the time-course of hDOR dephosphorylation after removal of the agonist when new phosphorylation of the receptor does not occur.

Morphine and deltorphin II differ in the rates of hDOR(S363) dephosphorylation. In the

hDOR dephosphorylation studies, the cells were treated for 30 min with either morphine

(30 μM) or deltorphin II (100 nM), the conditions determined previously to produce

maximal phosphorylation of hDOR(S363). Then the agonist was thoroughly washed away

and the cells were left in IMDM medium for increasing time intervals before harvesting.

We hypothesized that because morphine and deltorphin II treatments lead to differential

trafficking of the receptor, the rates of dephosphorylation will depend on the status of

receptor sequestration (thus, on the type of agonist that phosphorylated the receptor). Fig.

3.4A,B shows the time-course of hDOR(S363) dephosphorylation after morphine or

deltorphin II pretreatment. The intensities of the phosphorylated hDOR(S363) were

quantified, plotted as the function of the time after agonist removal and fitted with a

single exponential decay curve (Fig. 3.4C). The half life of receptor dephosphorylation in

the absence of the agonist, as determined by regression analysis was 19±6 min and 64±22

min for morphine and deltorphin II pretreated cells, respectively. Comparison of the two

regression curves using the F-test shows that the rates of dephosphorylation after

morphine and deltorphin II pretreatment are significantly different (p=0.003). Therefore,

dephosphorylation following S363 phosphorylation by morphine or deltorphin II proceeds

with different rates.

These data demonstrate that morphine, similar to full agonists such as SNC80, DPDPE

and deltorphin II, promotes phosphorylation of hDOR at S363. However, the rates of

phosphorylation and dephosphorylation are significantly different between morphine and 43

deltorphin II. Specifically, morphine treatment promotes slower phosphorylation and

after removal of the agonist faster dephosphorylation than deltorphin II. Finally, our data

show that in the continuous presence of the agonist the time course of morphine and

deltorphin II phosphorylation/dephosphorylation is biphasic, which indicates that this

process involves multiple regulatory or feedback mechanisms.

Discussion

Our findings indicate that, like full δ-opioid receptor selective agonists deltorphin II,

DPDPE and SNC80, morphine is also able to promote phosphorylation of S363 in the C-

terminus of the human δ-opioid receptor. Phosphorylation of S363 is the initial step in δ-

opioid receptor regulation and is a prerequisite for phosphorylation of other residues in

the C-terminus including T358, T361 (Kouhen et al., 2000). Additional sites may also be

phosphorylated in other intracellular domains. Since other investigators have found that total morphine-mediated phosphorylation of the δ-opioid receptor is only 6% of total phosphorylation by ethorphine (Zhang et al., 1999), we hypothesize that morphine fails to phosphorylate residues other than S363. Phosphorylation of at least two to three

serine/threonine residues is necessary for binding of arrestin to rhodopsin (Gurevich and

Gurevich, 2004). Therefore, phosphorylation of S363 alone may not be sufficient to recruit

β-arrestin to morphine-bound DOR, impeding receptor internalization and down- regulation. Recently, phosphorylation of a homologous residue in the μ-opioid receptor,

S375, upon morphine treatment was demonstrated (Schulz et al., 2004). We hypothesize that phosphorylation of these homologous residues may therefore be involved in 44

responses mediated by morphine, including desensitization, while phosphorylation of

additional residues is necessary for recruitment of β-arrestin, receptor internalization and

down-regulation.

By comparing the time courses of hDOR(S363) phosphorylation and dephosphorylation

after removal of the agonist between morphine and deltorphin II, we demonstrate that the

rates of phosphorylation and dephosphorylation are significantly different between the

two agonists. Furthermore, both morphine and deltorphin II display a biphasic time-

course of hDOR(S363) phosphorylation in the continuous presence of the agonist, indicating that the rates of phosphorylation and dephosphorylation change during the time of agonist treatment. Based on these results we conclude that the process of phosphorylation/dephosphorylation is agonist specific, and involves multiple regulatory steps or feedback mechanisms that develop over the period of the agonist treatment. In the case of deltorphin II, these other regulatory mechanisms may include β-arrestin binding or hDOR internalization. However, morphine does not promote β-arrestin recruitment and hDOR internalization (discussed in Chapters 4 and 5). Therefore, we speculate that morphine stimulates some other regulatory or feedback mechanisms, which modulate the rates of receptor phosphorylation/dephosphorylation during the time of morphine treatment. In support of the feedback mechanism, it has been demonstrated that

ERK1/2 kinases, which are activated by morphine treatment (Asensio et al., 2006), modulate the activity of several DOR regulatory proteins. For example it was shown that

ERK1/2 kinase activity prevents opioid receptor desensitization and sequestration by 45

blocking β-arrestin interaction with activated DORs (Eisinger and Schulz, 2004). Another

known substrate of ERKs is Ser670 in the GRK2. Phosphorylation of this residue

attenuates βγ-mediated translocation of GRK2 to the plasma membrane and reduces its phosphorylating activity for the receptor (Elorza et al., 2000; Pitcher et al., 1999).

Whether sequestration of the hDOR into plasma membrane microdomains inaccessible to

phosphatases, or direct feedback inhibition of kinase activity are involved in morphine-

mediated regulation of hDOR phosphorylation remains to be determined.

In conclusion, we were able to demonstrate for the first time that activation of the human

δ-opioid receptor by morphine leads to phosphorylation of S363 in the C-terminus of the

receptor (Navratilova et al., 2005). Interestingly, we found that the rate of hDOR(S363)

phosphorylation is slower and the rate of dephosphorylation after the drug removal is

faster for morphine than for deltorphin II. The differential kinetics of the phosphorylation/dephosphorylation process is probably responsible for the lower maximal S363 phosphorylation observed after morphine treatment than after deltorphin II

treatment. Nevertheless, since receptor phosphorylation is a crucial step in G protein-

coupled receptor desensitization, we suggest that phosphorylation of S363 may have an

important role in the regulation of human δ-opioid receptor signaling in response to

morphine treatment. 46

4. RECRUITMENT OF β-ARRESTIN TO THE AGONIST-STIMULATED

HUMAN DOR

Introduction

Activated and phosphorylated GPCRs exhibit high affinity for the cytosolic scaffolding

proteins: arrestins. Arrestins play crucial roles in regulation of GPCRs including receptor

desensitization, internalization and signaling. Mammalian arrestin family consists of four

highly homologous proteins of about 400 amino acids (Zhang et al., 1997): two visual

arrestins (rod arrestin or arrestin1, and cone arrestin or arrestin4) are expressed almost exclusively in the retina; and two non-visual arrestins, called β-arrestins (β-arrestin1 or

arrestin2, and β-arrestin2 or arrestin3), are ubiquitously expressed throughout the body

including the nervous system. The main functions of arrestins were first elucidated for

rod arrestin and rhodopsin in the visual system and were later confirmed, with very few

modifications, for β-arrestins and other GPCRs.

Arrestins bind to multiple sites on the cytoplasmic surface of activated and

phosphorylated G protein-coupled receptors. To ensure high selectivity for activated

receptors yet high flexibility in recognizing virtually all GPCRs, β-arrestins have two

receptor recognition domains: a GPCR activation sensor domain and a GPCR

phosphorylation sensor domain (Gurevich and Gurevich, 2004). The activation sensor domain of the arrestins recognizes GPCRs in their active agonist-bound conformation.

Similar to G proteins and GRKs, this interaction is believed to involve the intracellular 47

part of the cavity formed by the seven transmembrane helices that opens upon agonist

activation of the receptor, and intracellular domains such as the DRY motif that also

participate in G protein binding (Gurevich and Gurevich, 2006). The phosphorylation

sensor, composed of a polar core in the center of the arrestin molecule, binds to clusters

of phosphorylated S/T residues (or, in some cases, negatively charged residues) in the

intracellular domains of the GPCRs. Ionic interaction between the receptor phosphates and the arrestin’s polar core disrupts a salt bridge, opening the polar core and exposing additional receptor binding sites. Therefore the multistep binding of β-arrestin to GPCR results in extensive conformational changes in β-arrestin structure and its high affinity to the receptor (Vishnivetskiy et al., 2002).

Several β-arrestin binding sites have been identified on the delta opioid receptor. Using

plasmon resonance spectroscopy, two low affinity binding sites for β-arrestin were

identified in the DOR C-terminus (S358, S361 and S363 residues) and in the third

intracellular loop (S249 and S255 residues) (Cen et al., 2001a; Cen et al., 2001b). These

sites may correspond to the proposed phosphorylation sensor binding domain. In

addition, a phosphorylation independent but activation dependent β-arrestin binding

region common to most GPCRs, including the DOR, was identified (Marion et al., 2006).

This domain consists of highly conserved proline and alanine residues in the second

intracellular loop, downstream to the DRY motif that plays a prominent role in the

formation of the activated state of the GPCRs. These findings indicate that β-arrestins

have the ability to bind to multiple sites on the activated and phosphorylated DOR. Some 48 of these sites may be to some extent dispensable; however, the overall affinity of the

DOR-β-arrestin complex will be determined by the strength of all available binding sites.

Due to the overlap of β-arrestin and G protein binding sites on the G protein-coupled receptor, β-arrestin is expected to uncouple (arrest) the receptor from G proteins (hence the name, arrestin) (Krupnick and Benovic, 1998). In addition to this original function, β- arrestins also serve as scaffolds for a number of regulatory and signaling molecules.

Among them are clathrin (Goodman et al., 1996), and the β2-adaptin subunit of the AP-2 complex (Laporte et al., 2000), all necessary for clathrin-coated pit-mediated endocytosis.

Indeed, a crucial role of β-arrestin in DOR internalization via clathrin-coated pits has been demonstrated (Laporte et al., 2000). Another important function of β-arrestins originates from the ability of β-arrestins to bind a number of signaling molecules and therefore assemble functional signaling complexes (signalosomes) (Shenoy et al., 2006).

The best studied GPCR-β-arrestin signaling complex involves the ERK kinase cascade

(Raf1-MEK1-ERK1/2) (Tohgo et al., 2003), but interaction with other partners have also been described (JNK3, p38 kinase, c-Src, ubiquitin ligase Mdm2) (Pierce et al., 2001).

Undoubtedly β-arrestins are fundamental in DOR regulation and signaling. Therefore, it is important to determine how different DOR agonists facilitate the hDOR- β-arrestin interactions and to determine what are the molecular determinants of such interactions.

In this chapter we will first study the ability of different classes of DOR agonists to

promote β-arrestin translocation to the plasma membrane in hDOR/CHO cells. Since receptor phosphorylation is essential for β-arrestin binding, we will next investigate

whether mutations, which affect hDOR phosphorylation, will also affect β-arrestin

recruitment. Our results demonstrate that treatment of hDOR/CHO cells with full delta

opioid agonists (deltorphin II, DPDPE and SNC80) leads to mobilization of β-arrestin2 to

the cell membrane. On the other hand, treatment with morphine does not promote the β-

arrestin2 translocation. Additionally, our data show that mutation of the primary

phosphorylation site, S363, to alanine, which completely prevented phosphorylation of the

DOR (Kouhen et al., 2000), has no effect on β-arrestin2 translocation by deltorphin,

DPDPE and SNC80. In contrast, truncation of the C-terminus after Gly338 blocks β-

arrestin2 translocation by deltorphin II, but SNC80-mediated translocation to the

truncated hDOR was still observed.

Results

To assess the role of β-arrestin in hDOR regulation, we studied the mobilization of β-

arrestin2 to the plasma membrane upon binding of different DOR agonists to the wild

type and phosphorylation deficient mutant hDOR constructs. For β-arrestin2

translocation experiments, hDOR/CHO cells were transfected with an EGFP-tagged β-

arrestin2 (β-arrestin2-GFP). 48h after transfection, the cells were treated with an

appropriate agonist and fixed in 4% paraformaldehyde. Microscopic slides were prepared 50

as described in “Materials and methods”. β-arrestin was visualized using a Zeiss laser

confocal fluorescent microscope.

β-arrestin2 is recruited to the plasma membrane upon hDOR activation by deltorphin II,

DPDPE, SNC80 but not morphine. First we investigated the ability of different DOR

agonists (deltorphin II, DPDPE, SNC80 and morphine) to promote β-arrestin2-GFP translocation to the wild-type hDOR. Fluorescence micrographs of hDOR/CHO cells expressing β-arrestin-GFP (Fig. 4.1) show that, in untreated cells, β-arrestin2-GFP is evenly distributed throughout the cytosol. Treatment of the cells for 5 min with deltorphin II (500 nM), DPDPE (500 nM) or SNC80 (500 nM) at 37ºC led to the translocation of β-arrestin2-GFP to the cell surface. In contrast, after 5-60 min treatment of the cells with morphine (10 μM) we did not detect any changes in the localization of

β-arrestin2-GFP in any of the investigated cells. Therefore, morphine treatment does not promote β-arrestin2-GFP translocation to the plasma membrane and binding of β- arrestin2-GFP to the hDOR although this translocation is readily observed after treatment with full delta opioid agonists, deltorphin II, DPDPE and SNC80.

β-arrestin2 interacts only transiently with the hDOR. The DOR is classified as a type B

GPCR in respect to β-arrestin binding (Lowe et al., 2002; Oakley et al., 2000). Therefore,

the hDOR is expected to form “stable” high affinity complexes with both β-arrestins and

the whole complex should internalize into intracellular endosomes. To investigate the 51

time course of β-arrestin2 translocation, we treated the cells expressing β-arrestin2-GFP with deltorphin II or SNC80 for 2-30 min and visualized location of the fluorescently labeled β-arrestin2. Confocal fluorescent microscopic images (Fig. 4.2) demonstrate that

β-arrestin2-GFP is mobilized to the plasma membrane as early as 2 min after agonist treatment. Contrary to the expected co-internalization of β-arrestin2 with the receptor into

intracellular compartments, we rarely observed individual β-arrestin2-GFP-labeled

endosomes inside the cell. Most of the compartments that contain β-arrestin2-GFP were

localized at the cell surface or in close proximity even after 30 min of agonist treatment;

consistent with the notion that β-arrestin2 associates with the hDOR only at the very

early stages of receptor endocytosis. This conclusion is further emphasized on images

taken though different sections of the same cell (Fig. 4.3). Confocal sections taken

through the basal or apical membrane show punctuated endosome-like compartments

over the entire plane of the membrane. In contrast, confocal sections taken through other

regions of the cell show staining only on the edges of the cell, where the confocal section

trans-sects the plasma membrane. Conversely, visualization of the receptor endocytosis

using EGFP-labeled hDOR (hDOR-EGFP, Chapter 5) demonstrates that after 5 min

agonist treatment already a significant number of receptors is internalized into late

endosomes and perinuclear compartments. Therefore, based on these results we conclude

that in our cellular system β-arrestin2 associates with the hDOR only transiently at early

stages of receptor endocytosis and quickly dissociates from the hDOR before the receptor

enters the intermediate and late stages of the endocytic pathway. For this reason, hDOR

should more appropriately be classified as a type A G protein-coupled receptor. 52

S363A mutation does not eliminate agonist-mediated recruitment of β-arrestin2. Next we investigated the role of receptor phosphorylation in β-arrestin binding to the hDOR. It

was suggested earlier that S363 is the primary phosphorylation site within the C-terminus

of the DOR, since mutation of this amino acid to alanine prevented murine DOR

phosphorylation in all other receptor phosphorylation sites (Kouhen et al., 2000). Because phosphorylation of multiple GPCR residues is required for high affinity binding of β- arrestin to the receptor (Gurevich and Gurevich, 2004), we hypothesized that mutation of

S363 to alanine in the hDOR would eliminate β-arrestin translocation. To study agonist-

mediated translocation of β-arrestin to the hDOR(S363A) mutant we transiently

transfected the hDOR(S363A)/CHO cells with β-arrestin2-GFP. 48 h after transfection,

the cells were treated for 5 min with saturating concentrations of deltorphin II (500 nM),

DPDPE (500 nM), SNC80 (500 nM) or morphine (10 μM). Images acquired on a Zeiss

laser confocal fluorescent microscope (Fig. 4.4) clearly show that deltorphin II, DPDPE

and SNC80 treatment promoted the translocation of β-arrestin2-GFP to the plasma

membrane in CHO cells expressing the hDOR(S363A) in a pattern similar to that

observed in cells expressing the wild-type receptor (Fig. 4.1). Once again morphine

treatment did not promote mobilization of β-arrestin2 to the plasma membrane. In

addition, a semi-quantitative analysis indicates no difference in the kinetics of β-arrestin2

translocation between the agonist-activated wild type hDOR and hDOR(S363A) mutant. 53

These results provide evidence that S363 in the hDOR is not required for agonist-mediated

β-arrestin2 translocation to the cell surface.

Truncation of the C-terminus of the hDOR eliminates deltorphin II-mediated recruitment of β-arrestin2. Since mutation of S363 to alanine in the hDOR C-terminus had no effect on the translocation of β-arrestin2, we next investigated whether truncation of the C- terminus, which contains additional six putative phosphorylation sites, would prevent β- arrestin binding. As shown in Fig. 4.5, treatment of the truncated hDOR/CHO cells with deltorphin II no longer leads to the plasma membrane recruitment of β-arrestin2-GFP.

These results demonstrate that the C terminus of the deltorphin II-occupied hDOR stimulates β-arrestin binding by activating its phosphorylation sensor. In contrast, a very limited number of SNC80 treated truncated hDOR/CHO cells showed β-arrestin2-GFP translocation (Fig. 4.5). This finding may indicate that additional sites in the intracellular loops of the SNC80-bound hDOR could compensate for the lack of the primary β-arrestin phosphorylation sensor-activating domain.

Discussion

Our data provide evidence that full agonists (deltorphin II, DPDPE, SNC80) promote translocation of β-arrestin to the plasma membrane. In agreement with previous studies

(Whistler and von Zastrow, 1998) morphine was completely unable to stimulate β- arrestin binding. It has been demonstrated that binding of β-arrestin to GPCRs requires 54

phosphorylation of several S/T residues in the intracellular domains of the receptors

(Gurevich and Gurevich, 2004). Accordingly, receptor phosphorylation was found to be

the rate-limiting step for β-arrestin and β2-adrenergic receptor interaction (Krasel et al.,

2004). Surprisingly, our phosphorylation studies (Chapter 3) show that morphine is able

to phosphorylate S363, yet morphine treatment had no effect on translocation of β-arrestin.

Furthermore, confocal microscopy images show that mutation of the primary

phosphorylation site, S363, in the hDOR did not eliminate the recruitment of β-arrestin2-

GFP to the agonist-activated hDOR. No visually apparent differences were observed in

the time course and the extent of β-arrestin2-GFP translocation to the hDOR or

hDOR(S363A). Therefore, phosphorylation of S363 does not support, and the lack of it

does not prevent β-arrestin2 recruitment. Conversely, truncation of the C-terminus after

G338 completely eliminated the ability of deltorphin II but not SNC80 to stimulate the translocation of β-arrestin2 from the cytosol to the plasma membrane.

According to the multi-step theory of β-arrestin binding to GPCRs (Gurevich and

Gurevich, 2006), activation of both β-arrestin’s sensors (the phosphorylation sensor and

the activation sensor) by receptor phosphates and an active receptor conformation is

necessary for high affinity binding. Our study, although not conclusive, provides some

explanations for the molecular mechanisms of the observed agonist differences in hDOR-

β-arrestin interaction. 1) The phosphorylation sensor is activated by residues present on

the C-terminus of the hDOR but this activation does not require S363. Therefore,

phosphorylation of other residues in the C-terminus, or negatively charged amino acids is 55 sufficient to activate the phosphorylation sensor. 2) Morphine may be unable to activate the phosphorylation sensor even though it promotes phosphorylation of S363 in the hDOR.

Alternatively, morphine may still activate the phosphorylation sensor, but fails to promote a conformational change in the second intracellular loop of the hDOR required for stimulation of the activation sensor. 3) Phosphorylation of the intracellular loops other than the C-terminus by SNC80 (Varga et al., 2004) may activate the phosphorylation sensor and thereby facilitate β-arrestin binding to the SNC80-occupied truncated hDOR.

To summarize, our results show striking differences between the abilities of morphine, peptide agonists and SNC80 to stimulate the multi-step binding of β-arrestin2 to the hDOR. Due to the requirement of multiple activation events to occur, morphine is unable to promote hDOR-β-arrestin interaction. Conversely, due to some redundancy in activating the phosphorylation sensor, full agonists similarly activate β-arrestin translocation to the wild type hDOR. However, the differences between peptide agonists and SNC80 are revealed by mutations that eliminate this redundancy. 56

5. AGONIST-MEDIATED HUMAN DOR INTERNALIZATION

Introduction

Binding of β-arrestins to GPCRs facilitates the targeting of the receptors to clathrin- coated pits and receptor internalization. Internalization of the GPCR into endosomes serves several purposes: 1) removal of the receptor from the plasma membrane precludes further activation of the receptor by agonists causing receptor desensitization; 2) assembly of signaling molecules on the receptor-arrestin complex assists in receptor signaling through diverse signaling pathways; 3) endosomal environment, which contains protein phosphatases, facilitates receptor dephosphorylation and allows recycling of a fully functional receptor back into the plasma membrane, resulting in receptor resensitization; 4) sorting of the receptor into degradation organelles provides the means for final destruction of the receptor (down-regulation). Delicate balance between these processes determines the ultimate functional consequences of receptor internalization

(Daaka et al., 1997; Hall and Lefkowitz, 2002). Because of the importance of GPCR endocytosis, this process is highly regulated. Internalization has been found to be receptor specific (Chu et al., 1997), and receptor-specific motifs such as the PDZ domain have been demonstrated to affect the rate of receptor endocytosis (Puthenveedu and von

Zastrow, 2006; Tsao and von Zastrow, 2000). Internalization of many GPCRs is dependent on phosphorylation. However, a phosphorylation independent pathway has been described for the DOR (Zhang et al., 2005). Mechanism of receptor internalization 57

may also be cell type specific. Finally, various agonists have been shown to differentially

affect internalization of the same receptor (Zhang et al., 1999). Several mechanisms of

internalization of plasma membrane associated receptors have been characterized in

mammalian cells: clathrin-coated pits (Trapaidze et al., 1996), and lipid rafts and

caveolae (Insel et al., 2005; Roettger et al., 1995) mediated endocytosis.

Clathrin is a cytosolic protein, which assembles at the plasma membrane into a lattice

structure creating basket-shaped pits. Clathrin-coated pits (CCPs) are recognized in

electron microscopy as high density formations at the plasma membrane (Ghinea et al.,

1992). Clathrin-coated pits are later pinched off by dynamin, a cytosolic GTPase (van der

Bliek et al., 1993) recruited to the neck of the CCP. The pinched off endocytic vesicles mature from early to late endosomes during trafficking through the endosomal

compartments. In addition to clathrin, other associated proteins and membrane lipids

participate in regulating vesicular scission, maturation and trafficking (Advani et al.,

1999; Prekeris et al., 1999). Membrane-bound proteins such as GPCRs and extracellular

molecules such as ligands can both be internalized via CCPs. β-arrestin plays a crucial

role in targeting the GPCR to CCPs and in the assembly of the clathrin coat. This is

accomplished by direct binding of β-arrestin to clathrin (Goodman et al., 1996) and β-

adaptin of the AP-2 complex (Laporte et al., 2000).

Caveolae and lipid rafts are plasma membrane microdomains enriched in cholesterol and

sphingomyelin (Razani et al., 2002). Caveolae also contain scaffolding proteins, 58

caveolin-1α, 1β, 2, or 3 (Kiss et al., 2002) that promote the flask-shaped invagination of this microdomain. Like for CCPs, dynamin machinery is also responsible for scission of caveolae from the plasma membrane. 2-Hydroxypropyl-β-cyclodextrin is frequently used to study lipid rafts/caveolae-mediated effects because it sequesters cholesterol from the

plasma membrane and, therefore, disrupts lipid rafts/caveolae (Neufeld et al., 1996).

However, cholesterol depletion also attenuates endocytosis via clathrin-coated pits. A

more selective inhibitor of lipid rafts, fillipin, intercalates into cholesterol-rich lipid rafts

without altering cholesterol composition (Monis et al., 2006), and therefore specifically

targets lipid rafts/caveolae without affecting other modes of endocytosis. Several

membrane proteins, including opioid receptors (Huang et al., 2007), have been shown to

localize to lipid rafts and may undergo caveolae-mediated endocytosis (Insel et al., 2005).

As reported above, DORs have been identified in two different endocytic plasma

membrane domains: CCPs and lipid rafts or caveolae. It is assumed that the mechanism

by which the receptor was endocytosed will affect the fate of the receptor. Therefore, by

manipulating endocytosis of the receptor, we may be able to modulate the extent of

receptor desensitization or resensitization. In this Chapter, we used fluorescently-labeled deltorphin II analogues or Green Fluorescent Protein-labeled hDOR (hDOR-EGFP) to study internalization of the delta opioid receptor in recombinant Chinese hamster ovary cells. The goal of the study was: 1) to determine the molecular mechanism of hDOR internalization; 2) to investigate agonist selectivity of hDOR endocytosis; 3) to study the role of S363 phosphorylation in hDOR internalization. 59

Results

Time course of fluo-deltorphin internalization. Since DOR agonist deltorphin II is internalized together with its receptor, fluorescently labeled deltorphin II (fluo- deltorphin, Fluo-Peptides) can be used to track the translocation of the hDOR in CHO cells after agonist treatment. hDOR/CHO cells were treated for increasing time intervals with fluo-deltorphin (300 nM) at 37ºC, cells were then fixed and examined under a Zeiss laser scanning confocal microscope. Fig. 5.1 illustrates that fluo-deltorphin is internalized in a time dependent manner into punctuated intracellular vesicles structurally consistent with endosomes. From our micrographs it is apparent that internalization of fluo- deltorphin occurs already after 3 min and is even more pronounced after 10 min of fluo- deltorphin treatment. After longer times (30-60 min) some fluo-deltorphin accumulates in the perinuclear compartments (Fig. 5.1C,D and F). Importantly, endocytosis of fluo- deltorphin is strictly mediated by the hDOR since it is completely blocked by a delta opioid receptor antagonist naltrindole (10 μM, Fig. 5.1E). Similar internalization pattern was also observed in CHO cells expressing an EGFP labeled hDOR (hDOR-

EGFP/CHO#2 cell line), confirming that the internalized agonist traces the endocytic pathway of the receptor (Fig. 5.3). Therefore, a fluorescently labeled DOR agonist accurately monitors internalization of the human delta opioid receptor in the CHO cells.

hDOR is internalized via a subset of clathrin-coated pits distinct from transferrin- containing CCPs. Most GPCRs are internalized via clathrin-coated pit mediated 60

endocytosis (Luttrell and Lefkowitz, 2002). However, other modes of receptor

endocytosis have also been described (Mueller et al., 2002; Okamoto et al., 2000). In

order to establish the mechanism of hDOR internalization in CHO cells we first

compared endocytosis of a fluorescently labeled analog of deltorphin II,

[Gln4]deltorphin-rhodamine, with endocytosis of a prototypical CCP ligand transferrin

(transferrin-AlexaFluor488, Molecular Probes). Secondly, we investigated deltorphin- rhodamine and transferrin internalization in the presence of inhibitors of different endocytic pathways.

Concomitant visualization of internalization of deltorphin-rhodamine (red) and

transferrin-AlexaFluor488 (green) revealed that most endosomes contain only one type of

ligand. Only a small portion of endosomes shows colocalization of both deltorphin- rhodamine and transferrin (yellow). This observation led to a conclusion that the two

ligands are not internalized through a homogeneous population of vesicles. Because of

the heterogeneity of the endocytic vesicles, we asked whether vesicles carrying

deltorphin II originate from CCPs or caveolae-mediated endocytosis. To address this

question we studied internalization in the presence of inhibitors of CCP endocytosis

(hypertonic sucrose) or caveolae-mediated endocytosis (2-hydroxypropyl-β-cyclodextrin

and fillipin). As seen in Fig. 5.2B hypertonic sucrose (0.5 M), an inhibitor of CCP

endocytosis, completely prevented internalization of both deltorphin-rhodamine and

transferrin. Interestingly, 2-hydroxypropyl-β-cyclodextrin (2%) selectively inhibited only

deltorphin-rhodamine endocytosis while transferrin was still endocytosed (Fig. 5.2C). In 61

contrast, fillipin had no effect on internalization of either ligand (Fig. 5.2D). Depletion of

cholesterol from the plasma membrane by 2-hydroxypropyl-β-cyclodextrin destroys lipid

rafts and caveolae formation. However, it was also demonstrated that cholesterol plays a

role in CCP morphology and that 2-hydroxypropyl-β-cyclodextrin disrupted CCP-

mediated endocytosis (Neufeld et al., 1996). Filipin, on the other hand, selectively

disrupts lipid rafts/caveolae by intercalating into cholesterol-rich domains without

altering cholesterol content (Monis et al., 2006). The obtained results are consistent with

the interpretation that the hDOR is internalized by a subset of clathrin-coated pits distinct

from transferrin-containing CCPs. Our results also implicate a possible higher sensitivity

of hDOR containing CCPs to cholesterol depletion.

hDOR endocytosis is agonist-specific. In order to study the ability of different delta

opioid agonists to promote receptor internalization, we constructed an hDOR with an

EGFP epitope tag at the C-terminus (as described in “Materials and Methods”). The

cDNA encoding the hDOR-EGFP construct was transfected into CHO cells and a stable

cell line (hDOR-EGFP/CHO#2) expressing the tagged receptor was generated.

Localization of the hDOR-EGFP was visualized on a Zeiss confocal fluorescent

microscope. In untreated cells the receptor was expressed throughout the cell, consistent

with previous reports of intracellular reserve pool of the DOR (Cahill et al., 2001; Wang

and Pickel, 2001). However, enhanced plasma membrane localization was clearly visible

(Fig. 5.3A). Upon treatment of the cells with deltorphin II (1μM) or SNC80 (1 μM), the receptor was internalized into punctuated intracellular endosomes (Fig. 5.3B and C, 62

respectively). A similar internalization pattern was also observed for DPDPE (not

shown). In contrast, morphine treatment (10 μM) did not promote receptor internalization

(Fig. 5.2D). These findings corroborate the results obtained for agonist-mediated translocation of β-arrestin2 (Fig. 4.1), which show that morphine does not promote translocation of β-arrestin2 to the plasma membrane. In conclusion, morphine, contrary

to deltorphin II, DPDPE and SNC80, does not support hDOR internalization, apparently

due to its inability to mobilize β-arrestin to the plasma membrane.

S363A mutation of the hDOR does not eliminate agonist-mediated hDOR internalization.

Phosphorylation of the receptor after agonist treatment was shown to be important for

internalization of some GPCRs, but may not be crucial for the DOR (Zhang et al., 2005).

Therefore, we investigated if a mutation of the primary phosphorylation site S363 in the C-

terminus of the hDOR will affect agonist-mediated endocytosis. Internalization of the

wild type and S363A mutant hDOR was determined using confocal fluorescent

microscopy by investigating the ability of the receptor to internalize a fluorescent

analogue of deltorphin II ([Gln4]deltorphin-rhodamine). In hDOR/CHO cells, after 30

min of agonist treatment, the fluorescent marker was concentrated in intracellular

punctuated compartments resembling endosomes (Fig. 5.4A). In cells expressing the

S363A mutant receptor we observed virtually indistinguishable internalization of

fluorescent deltorphin. These results coincide with our findings that S363A mutation does

not appreciably affect β-arrestin2 binding to the mutant receptor (Fig. 4.4) Therefore,

S363A mutation in the hDOR has a minimal effect on the ability of deltorphin II to 63

mobilize β-arrestin to the hDOR and to trigger receptor endocytosis. This finding implies that phosphorylation of S363 is not necessary for agonist-mediated hDOR internalization.

Discussion

The results presented in this Chapter demonstrate that treatment of the hDOR/CHO cells

with fluorescent deltorphin II analogs leads to receptor-mediated internalization of the

agonists. Endocytosis is already apparent 3 min after the treatment and continues for up

to 60 min. By comparing internalization of deltorphin-rhodamine with that of transferrin,

a marker of clathrin-coated pits-mediated endocytosis, we found that the mechanism of

GPCRs internalization is more complex than previously envisioned. Our confocal

micrographs provide evidence that in CHO cells the hDOR is internalized via

heterogeneous population of transferrin-containing CCPs and also via vesicles devoid of

transferrin. Therefore, a simple concept of homogeneous population of clathrin-coated

vesicles carrying random mixture of endocytosed receptors can no longer be sustained. In

addition to CCPs, some membrane associated receptors may be internalized via a caveolae-mediated mechanism (Ostrom and Insel, 2004). To investigate a possible role of lipid rafts or caveolae in hDOR endocytosis, we studied sensitivity of hDOR internalization to inhibitors of CCP-mediated and lipid rafts-mediated endocytosis. We found that, like transferrin, deltorphin II endocytosis is sensitive to an inhibitor of CCP endocytosis, hypertonic sucrose, but not to filipin, an inhibitor of caveolae. Interestingly, hDOR internalization is more sensitive to cholesterol depletion than transferrin. Since cholesterol also plays a role in the morphology of CCPs, these results indicate that the 64

hDOR is internalized via a cholesterol depletion-sensitive subset of CCPs different from

transferrin-containing CCPs. Indeed, it was reported recently that even within CCPs vesicles, multiple subpopulations of vesicles containing different assembly of receptors exists (Puthenveedu and von Zastrow, 2006). The composition of these vesicles is probably very dynamic and depends on the overall status of the cell. It has been well documented that the signaling of a GPCR extends beyond the plasma membrane and

continues in a different mode after receptor endocytosis (Luttrell et al., 1999; Miller et

al., 2000). Scaffolding proteins β-arrestins bound to the receptors assemble signaling

complexes consisting of regulatory and signaling molecules (Maudsley et al., 2000). In

addition, cholesterol rich lipid rafts have also been shown to serve as scaffolding

microdomains to which GPCRs, G-proteins, enzymes (adenylyl cyclase) and protein

kinases translocate during agonist-mediated signaling (Ostrom and Insel, 2004; Ostrom et

al., 2000). Therefore, endosomes are ideally suited to integrate signaling from multiple scaffolding complexes. In this respect it is tempting to speculate that depending on the

existing extracellular signals transduced at the moment by a number of different

receptors, diverse populations of intracellular vesicles are assembled. Each vesicle

selectively integrates a set of relevant incoming signals thereby creating a meaningful

signaling message, which determines the ultimate response of the cell. Therefore, one

message is constructed from a combination of multiple signals and, conversely, one type

of signal (transduced by one type of receptor) participates on formulation of several

messages.

In addition to the complexity of the mechanism of DOR internalization, it has also been

demonstrated that various DOR agonists have different effects on receptor

internalization. Thus, it was shown that the kinetics of hDOR internalization by alkaloids

(etorphine) and peptide agonists (DPDPE) differs in SK-N-BE cells (Marie et al., 2003).

Moreover, it has been demonstrated that morphine, a non-selective MOR and DOR

agonist, is unable to promote internalization of either receptor (Keith et al., 1998; Keith et

al., 1996; Sternini et al., 1996) In our studies, we used GFP-labeled hDOR stably

expressed in CHO cells to assess agonist specific differences in hDOR internalization.

We found that in our cellular system morphine was unable to bring about hDOR-EGFP internalization. However, no visually detectable differences were found between deltorphin II, DPDPE and SNC80-mediated internalization.

The C-terminus of the DOR plays a crucial role in receptor endocytosis. However, the

exact details of this role may be agonist and cell type specific. Truncation of the C-

terminus was shown to block DPDPE-mediated internalization in CHO cells (Trapaidze et al., 1996). In contrast, the truncated DOR was still internalized upon etorphine and

DADLE treatment in HEK293 cell (Murray et al., 1998). Surprisingly, in the same cells a

DOR with mutated C-terminal phosphorylation sites did not internalize (Whistler et al.,

2001). To explain these data, Whistler et al. suggested that the C-terminus functions as a

“brake” precluding the interaction of the receptor with β-arrestin (and therefore receptor internalization) in the absence of an agonist. Phosphorylation of the C-terminus is required to relieve this brake (Whistler et al., 2001). To investigate the role of C-terminal 66 phosphorylation in the hDOR, we studied receptor internalization in a mutant hDOR in which the primary phosphorylation site, S363, was mutated to alanine. We found that

S363A mutation in the hDOR did not eliminate the ability of the receptor to internalize

[Gln4]deltorphin-rhodamine. From these results we conclude that in CHO cells phosphorylation of S363 is not necessary for agonist-mediated hDOR internalization.

In summary, our results demonstrate that: 1) in hDOR/CHO cells agonist-mediated hDOR internalization occurs via a subset of clathrin-coated pits distinct from transferrin containing pits. Moreover, hDOR endocytosis differs from endocytosis of transferrin by its higher sensitivity to cholesterol depletion. 2) Internalization of hDOR is agonist selective, since morphine is completely inefficient in promoting receptor endocytosis. 3)

Elimination of the primary phosphorylation site in the C-terminus of the hDOR has no significant effect on receptor internalization. 67

6. AGONIST-MEDIATED HUMAN DOR DOWN-REGULATION

Taken in part from (Navratilova et al., 2004)

Introduction

Down-regulation of G protein-coupled receptors is different from receptor internalization because it represents total loss of cellular receptors rather than receptor redistribution from the cell surface into intracellular compartments. Down-regulation of the DOR was demonstrated by other investigators (Afify et al., 1998; Belcheva et al., 1992; Malatynska et al., 1996) but the exact mechanism is still not completely understood.

As with other GPCRs, down-regulation of the DOR requires receptor internalization via endocytosis and subsequent degradation inside the cell.

Two mechanisms of endocytosis have been proposed for GPCRs: clathrin-coated pits- mediated or caveolae-mediated endocytosis (as described in Chapter 5). In this Chapter we investigate whether inhibitors of endocytosis also inhibit hDOR down-regulation in hDOR/CHO cells. After the receptor is internalized, it is trafficked through a cascade of endosomes either to the recycling or to the degradation pathways. Recycling receptors are de-phosphorylated, re-sensitized and inserted into the plasma membrane, where they can continue their function. Receptors destined for degradation are targeted to intracellular digestive organelles, where they are destroyed. Two intracellular compartments may function as protein degradation machineries: lysosomes and proteasomes.

The lysosome was initially characterized as a membrane protected vacuolar structure that

contains various hydrolytic enzymes and has an acidic luminal pH. Lysosomal/vacuolar

system is now recognized as a heterologous digestive system of endosomes at different

stages of maturation (Ciechanover, 2005). At the initial stages of lysosomal proteolysis

are the early/intermediate/late endosomes and phagocytic vacuoles, which converge into

multivesicular bodies (MVB). MVBs later fuse with the digestive lysosome. The end products of the completed digestive process are the residual bodies. Many cellular as well as exogenous proteins targeted to the lysosome by receptor-mediated endocytosis, pinocytosis, and phagocytosis, are degraded by this system in a highly regulated manner.

Most of the lysosomal proteases function optimally at acidic pH, therefore low pH is essential for the ability of the lysosomes to degrade cargo proteins. Lysosomal degradation plays a major role in regulating exogenous as well as endogenous proteins.

The proteasome is a large, 26S, multicatalytic protease complex consisting of a barrel

shaped 20S core particle and two 19S regulatory units at both ends of the core particle

(Ciechanover, 2006). The core particle contains the catalytic activity of the proteasome,

while the role of the regulatory subunits is to recognize substrate proteins destined for

degradation by the proteasome. The recognition signal for the regulatory subunit is

composed of multiple ubiquitin molecules arranged in a polyubiquitin chain and

covalently attached to the substrate proteins. Ubiquitin, a small protein of 76 residues, is

conjugated to a lysine residue of the substrate protein in a cascade of enzymatic reactions, which involve the ubiquitin-activating enzyme, E1, the ubiquitin-carrier protein, E2, and 69

the ubiquitin-protein ligase, E3. The variety of ubiquitin-protein ligases ensures a high

selectivity of the ubiquitin/proteasome degradation mechanism. Among the diverse

processes regulated by the ubiquitin/proteasome degradation are quality control of the

protein synthetic pathway, cell cycle, development, differentiation and receptor-mediated down-regulation (Li et al., 2000; Petaja-Repo et al., 2001). In this study the contribution of each degradation mechanism to hDOR down-regulation was investigated using selective inhibitors.

Sequestration of proteins into different organelles is tightly regulated and governed by

consensus targeting motifs (sorting motifs). Targeting motifs for some GPCRs into

recycling or degradative pathways have been identified. In addition, post-translational

modifications (ubiquitination, phosphorylation) have been implicated as receptor

trafficking signals. In that respect, it was found that T353 located within the C-terminal tail

of the mouse DOR is involved in receptor down-regulation (Cvejic et al., 1996).

However, in the human DOR sequence this residue corresponds to an A, therefore a

different yet unidentified signal serves as a down-regulation motif in the human DOR.

As demonstrated previously in our laboratory (Okura et al., 2000), truncation of the

carboxyl-terminus of the hDOR at Gly338 completely abolishes DPDPE and attenuates

SNC80-mediated down-regulation. These results led us to a conclusion that the primary

down-regulation motif for peptide agonists is located within the C-terminus of the hDOR.

In addition, SNC80 may be able to expose a secondary down-regulation motif in other 70

intracellular domains. In this Chapter, after demonstrating that ubiquitination is not

involved in hDOR down-regulation, we investigated a putative role of phosphorylation as

a down-regulation signal.

In order to identify the primary down-regulation motif within the C-terminus we hypothesized that the primary phosphorylation site S363 (Kouhen et al., 2000) may also

serve as the degradation sorting signal. To validate this hypothesis we measured down-

regulation of the wild-type hDOR and hDOR(S363A) mutant after chronic treatment with

DOR agonists. In search for the secondary down-regulation motif used by SNC80 in

addition to the primary signal, we focused on the distal part of the third intracellular loop

shown to participate in β-arrestin binding to the DOR (Cen et al., 2001a; Cen et al.,

2001b). Phosphorylation is crucial for the activation of β-arrestins and a high affinity

binding between β-arrestin and a GPCR (Celver et al., 2002). Phosphorylation of the

truncated hDOR by SNC 80 but not DPDPE was demonstrated earlier in our laboratory

(Varga et al., 2004). Magnitude of phosphorylation of the truncated hDOR correlates

with the ability of SNC80 but not DPDPE to down-regulate the truncated hDOR. This led

us to select two serine residues, S249 and S255, localized in the β-arrestin binding domain

of the DOR as potential candidates for the secondary down-regulation motif used

exclusively by SNC80. The role of the putative down-regulation motifs (S363 and

S249/S255) was investigated by measuring down-regulation of mutant receptors in which

respective serines were replaced by alanines.

Results

hDOR is down-regulated in a time and dose dependent manner. Down-regulation of the

wild-type hDOR was studied using the 2-209-hDOR/CHO cell line characterized in

“Materials and Methods”. The cells were treated with six different concentrations of

SNC80 for 24h or with 500 nM SNC80 for 2-24h at 37ºC in IMDM medium. Cell

membranes were prepared and the amount of receptors remaining after the treatment was

determined by measuring the specific [3H] naltrindole binding. Down-regulation was

defined as a decrease in specific binding per 1 μg membrane protein after agonist

treatment as described in “Materials and Methods”. Fig. 6.1A demonstrates that the

hDOR is dose dependently down-regulated by SNC80. Using a sigmoidal dose response

curve fitting with the Hill coefficient nH = 1, the EC50 concentration was determined to be

21 ± 8 nM. Maximal down-regulation was achieved with 500 nM of SNC80. This

concentration was chosen for further experiments. Fig. 6.1B shows a time course of

SNC80 (500 nM)-mediated hDOR down-regulation. Data were fitted with a single

exponential decay curve and the half-life of down-regulation t½ = 2.3±0.4 h was determined. Maximal down-regulation achieved after 24h of 500 nM SNC80 treatment, calculated by subtracting the percent of remaining receptors from 100 %, was 90 %. 24h time of treatment was chosen for further experiments.

hDOR down-regulation is agonist specific. It has been demonstrated that a partial DOR

agonist, morphine, is very inefficient in down-regulating the DOR. As reported in

Chapters 4 and 5 in our cellular system, morphine displays differential efficacies in 72

activating G protein signaling pathways (GTPγS stimulation and cAMP inhibition) and

regulatory pathways (β-arrestin recruitment, hDOR internalization). Allouche and coworkers (1999) reported that additional differences in hDOR down-regulation exist between etorphine and peptide agonists (Allouche et al., 1999). In our previous report

(Okura et al., 2003) we also found that peptide (DPDPE) and non-peptide (SNC80) agonists exhibit differential efficacies in down-regulating the hDOR even though they both are full agonists in GTPγS stimulation and cAMP inhibition assays. To carefully characterize agonist-specific differences in hDOR down-regulation, in the present study we examined receptor down-regulation by all three classes of DOR agonists, as shown in

Fig. 6.2. Treatment of hDOR/CHO cells for 24h with morphine (5 μM) does not cause down-regulation of the hDOR. We did not detect any down-regulation even after 48h of morphine treatment (data not shown). In contrast, treatment with DPDPE (500 nM) or

SNC80 (500 nM) led to a reduction of hDOR numbers by 80±1 % and 90±1 %, respectively. The difference between the ability of DPDPE and SNC80 to down-regulate the hDOR is statistically significant (p<0.001) and was reproduced in many other down- regulation experiments involving inhibitors of down-regulation or mutations of the receptor. To summarize, we identified three classes of hDOR agonist: morphine, a peptide agonist (DPDPE) and a non-peptide agonist (SNC80), which have different efficacies to down-regulate the hDOR. These efficacies do not correspond to the efficacies to stimulate G protein-mediated signaling.

73 hDOR endocytosis via clathrin-coated pits is necessary for receptor down-regulation.

Hypertonic sucrose has been used to inhibit clathrin-coated pits-mediated endocytosis

(Law et al., 2000). Using confocal microscopy we have demonstrated (Chapter 5) that indeed, hypertonic sucrose (0.5 M) prevents hDOR-mediated internalization of rhodamine-labeled deltorphin II. In this Chapter we investigated whether internalization is necessary for hDOR down-regulation or whether an internalization-independent mechanism of hDOR degradation at the plasma membrane exists.

To answer this question, we measured down-regulation after 2h of DPDPE or SNC80 treatment in the absence and presence of hypertonic sucrose (0.5 M). Fig. 6.3 shows that hypertonic sucrose almost completely blocked hDOR down-regulation by DPDPE (39±1

% vs. 10±6 % in control and sucrose treated cells, respectively) and by SNC80 (51±6 % vs. 16±1 % in control and sucrose treated cells, respectively). We speculate that the residual reduction in receptor number in sucrose and agonist treated cells is due to either a slight toxicity of hypertonic sucrose treatment (as indicated by a small decrease in receptor number observed in sucrose treated controls) or incomplete inhibition of hDOR endocytosis (as indicated in Fig. 5.2B by the appearance of some internalized deltorphin- rhodamine in sucrose treated cells). Similarly, 2% 2-hydroxypropyl-β-cyclodextrin, which also blocked internalization of rhodamine-labeled deltorphin analog (Fig. 5.2C), prevented down-regulation of the hDOR (data not shown). These results demonstrate that hDOR endocytosis is required for agonist-mediated down-regulation of the receptor.

74 hDOR is not down-regulated in the proteasomes. What is the organelle responsible for hDOR down-regulation? It was reported that the delta opioid receptors are down- regulated by the proteasomes in HEK 293 cells (Chaturvedi et al., 2001; Yadav et al.,

2006). To investigate whether this mechanism also occurs for the human DOR in CHO cells, we measured the total number of receptors before and after agonist treatment in the presence or absence of proteasomal inhibitors. We confirmed that the decrease of receptor number after agonist treatment was reversed by a selective inhibitor of the proteasome, lactacystin (25 μM) (Fig. 6.4). However, the same treatment increased the basal levels of hDOR in the absence of the agonist by 2.1 fold. Therefore, even though the percent of receptors decreased by agonist treatment is lower in the presence of lactacystin, the absolute number of decreased receptors is approximately the same compared to the controls.

Proteasomes play a critical role in the quality control of newly synthesized proteins, including the DOR, by degrading misfolded proteins (Petaja-Repo et al., 2001).

Therefore, we hypothesized that the increase in the receptor number after lactacystin treatment is caused by accumulation of incorrectly folded receptor proteins. To investigate this possibility, we performed the same down-regulation experiment in the presence of an inhibitor of new protein synthesis, cycloheximide (5 μg/ml) (Fig. 6.5). 24h treatment with cycloheximide alone caused a small decrease in receptor number as compared to the control cells (26±16 %). However, as hypothesized, cycloheximide completely blocked the increase in hDOR observed after lactacystin treatment (24±16 % 75

decrease in receptor number by cycloheximide in the presence of lactacystin). Most importantly, cycloheximide also blocked any reduction in DPDPE- and SNC80-mediated hDOR down-regulation in the presence of lactacystin (90±2 % vs. 87±2 % down-

regulation by DPDPE in the absence and presence of lactacystin, respectively and 92±1

% vs. 91±1 % down-regulation by SNC80 in the absence and presence of lactacystin,

respectively). These data clearly indicate that the effect of lactacystin on hDOR

expression is caused by blockade of the degradation of newly synthesized receptors not

by blockade of degradation of agonist activated and internalized receptors targeted for down-regulation.

To obtain further support for this conclusion, we measured receptor down-regulation after

only 4h of agonist treatment in the presence or absence of another proteasome inhibitor,

MG132 (25 μM) (Fig. 6.6). After 4h of DPDPE treatment already about 54±6 % of the

receptors are down-regulated, but no significant increase in the basal levels of hDOR

expression by MG132 is observed. Importantly, down-regulation in the presence and

absence of MG132 are indistinguishable (54±6 % vs. 50±3 %, respectively), confirming

that proteasomes are not critically involved in agonist-mediated hDOR down-regulation.

hDOR down-regulation does not require ubiquitination. Ubiquitination is an important prerequisite for recognition of the target protein by the proteasome’s regulatory subunit and its subsequent degradation. Interestingly, ubiquitination was also demonstrated to be necessary to direct some GPCRs into lysosomes (Shenoy et al., 2001). To investigate the 76

role of ubiquitination in hDOR down-regulation we utilized E36ts20 cells which harbor a temperature sensitive ubiquitin-activating enzyme E1 (Handley-Gearhart et al., 1994).

The enzyme is fully functional at temperatures below 37ºC and inactivates at 41ºC. As apparent from Fig. 6.7, DPDPE treatment led to similar down-regulation of the human delta opioid receptors transfected into E36ts20 cells when the cells were incubated at permissive (33ºC) or non-permissive (41ºC) temperatures (39±9 % and 46±3 %,

respectively). Similar results were obtained for SNC80 treated cells (% of down-

regulated receptors was 41 % at 33ºC and 43 % at 41ºC) These results confirm our

previous finding that hDOR is not down-regulated by the proteasome. In addition, these

results demonstrate that in contrast to the β2AR (Shenoy et al., 2001) which requires

ubiquitination for its down-regulation in lysosomes, hDOR down-regulation proceeds

normally in the absence of ubiquitination. Later, these findings were reported by the

laboratory of M. von Zastrow (Tanowitz and von Zastrow, 2002). These authors find that

mutation of all lysine residues within the intracellular domains of the DOR , which

prevents receptor ubiquitination, had no effect on targeting the DOR to degradation in

lysosomes.

hDOR is down-regulated in the lysosome. Since we found that proteasomes are not

involved in hDOR down-regulation we investigated whether hDOR down-regulation

takes place in lysosomes. Lysosomal proteases are active at acidic pH, therefore low pH

is essential for the ability of the lysosomes to degrade the cargo proteins. Lysosomal

acidic environment is created and maintained by the function of vacuolar H+ ATPases 77

that actively transport protons into the lysosomes. Therefore blocking the H+ pump by a

selective inhibitor, folimycin, impairs the ability of the lysosomes to degrade proteins

(Mousavi et al., 2001).

In our control experiments we found (similar to the proteasomal inhibitors) that folimycin treatment for 24h increases the basal levels of hDOR, indicating that the lysosomes may also be involved in the quality control of the newly synthesized proteins. To eliminate the influence of the synthetic pathway we performed the down-regulation assays in the presence of a blocker of new protein synthesis, cycloheximide. As seen in Fig. 6.8, down-

regulation of hDOR is significantly less efficient in the presence of folimycin than in

control cells. This finding identifies the lysosomes as the primary organelle of hDOR degradation in CHO cells.

C-terminus of the hDOR contains the degradation targeting motif. In order to identify a

putative down-regulation motif in the amino acid sequence of the hDOR, we followed up

on the previous experiments from our laboratory (Okura et al., 2000) and confirmed that the C-terminus contains the primary degradation domain. Fig. 6.9 shows that in CHO cells expressing a truncated hDOR, down-regulation by DPDPE is completely eliminated

(6 % vs. 80 % for the truncEt-hDOR and wild type hDOR, respectively). In contrast,

SNC80 treatment leads only to a reduction in the number of down-regulated receptors

(47±6 % vs. 90±1 % for the truncEt-hDOR and wild type hDOR, respectively). This finding indicates that DPDPE and SNC80 utilize different molecular mechanisms to 78

down-regulate the hDOR. Specifically, these experiments identify the C-terminus as the

crucial domain for targeting the receptor to down-regulation after a peptide agonist

(DPDPE) stimulation, but SNC80 may be able to utilize other intracellular domains.

S363 within the C-terminus of the hDOR is involved in DPDPE-mediated down-

regulation. In order to find a specific down-regulation motif within the C-terminus, we

hypothesized that the primary phosphorylation site, S363, will also play a key role in

DPDPE-mediated down-regulation. To validate this hypothesis, we mutated S363 to A and created CHO cell line stably expressing the hDOR(S363A) mutant, as described in

“Materials and Methods” (hDOR(S363A)/CHO#13 cell line). The recombinant 2-209-19- hDOR/CHO and hDOR(S363A)/CHO#13 cells were treated with 500 nM DPDPE or 500 nM SNC80 for 24h and the extent of receptor down-regulation was measured using [3H]

naltrindole radioligand binding assay. The results indicate (Fig. 6.10) that after 24h

DPDPE treatment, down-regulation of the hDOR(S363A) mutant receptor was

significantly attenuated (45±9 %) compared to that of the wild-type human δ-opioid

receptor (80±1 %), p<0.01. In contrast, SNC80-mediated down-regulation was virtually unaffected by the mutation (87±1 % for the hDOR(S363A) mutant, and 90±1 % for the wild-type human δ-opioid receptor) (Navratilova et al., 2004). In our interpretation, these data demonstrate that Ser363 is part of the primary down-regulation targeting motif in the

C-terminal region of the hDOR.

The secondary down-regulation targeting motif in the hDOR exposed by SNC80 does not

involve S249 and S255. As demonstrated, in all experiments SNC80 is more efficacious

then DPDPE in down-regulating the hDOR, the truncated hDOR and the hDOR(S363A)

mutant. Notably, truncation of the C-terminus of the hDOR at Gly338 completely

abolishes down-regulation by DPDPE while SNC80 is still able to down-regulate about

50% of the receptors. In addition, mutation of S363 to A attenuates DPDPE but not

SNC80-mediated down-regulation. Based on these results, we hypothesized that SNC80

may activate a secondary down-regulation motif located in intracellular domains other

than the C-terminus. Since SNC80 is able to phosphorylate the truncated hDOR while

DPDPE is not (Varga et al., 2004), we also hypothesized that the putative secondary

degradation motif will include phospho S/T residues. Using the plasmon resonance

spectroscopy Cen and collaborators (Cen et al., 2001a) identified S249 and S255 as crucial

amino acids for binding of β-arrestins to the third intracellular loop of the DOR. β-

arrestin binding to a GPCR is considered necessary for subsequent receptor

internalization and down-regulation. Therefore, we tested whether these serine residues

may function as necessary signals for residual SNC80-mediated hDOR down-regulation.

In order to investigate the role of the S249/S255 motif in SNC80-mediated down- regulation we used the hDOR(S363A) mutant, which displays an amplified differential down-regulation between SNC80 and DPDPE. S249A and S255A mutation was introduced into the hDOR(S363A)pcDNA3.1 using the QuickChange mutagenesis kit generating an hDOR(S249A/S255A/S363A) triple mutant, as described in “Materials and 80

Methods”. hDOR(S249A/S255A/S363A) or hDOR(S363A) cDNA was transiently

transfected into CHO cells using FuGENE 6 transfection protocol and the ability of

DPDPE and SNC80 to down-regulate the two receptors was measured. Fig. 6.11 shows

that contrary to our expectation, SNC80 still down-regulates the triple mutant

hDOR(S249A/S255A/S363A) more effectively than DPDPE (69±7% for SNC80 and

57±4% for DPDPE). No change was detected compared to the hDOR(S363A) control

(74±12% for SNC80 and 55±11% for DPDPE). In conclusion, S249 and S255 do not

participate in the residual SNC80-mediated down-regulation. Instead, we hypothesize

that other S/T residues in the intracellular loops of the hDOR may be involved as secondary degradation motif.

Discussion

It has been demonstrated that the fate of the internalized G-protein coupled receptors is

both receptor-, and cell type-specific (Tsao and von Zastrow, 2000). Our studies on

receptor down-regulation provide ample evidence that hDOR down-regulation is also

agonist-specific. In support of this conclusion, chronic morphine treatment was

completely unable to down-regulate the hDOR (Fig. 6.2) even though morphine activated

G protein signal transduction pathways (as measured by stimulation of GTPγS binding

and inhibition of forskolin stimulated cAMP, “Materials and Methods”). In addition, a peptide agonist DPDPE and a non-peptide agonist SNC80, which both are full agonists in stimulating G protein signaling, display differential efficacies in their abilities to down-

regulate the receptor (Fig. 6.2). Other authors also demonstrated differences in regulation 81

of the hDOR by etorphine and peptide agonists in SK-N-BE cells (Allouche et al., 1999).

The differential efficacy observed in this study between SNC80 and DPDPE was detected in all experiments involving inhibitors of clathrin-coated pit-mediated endocytosis and inhibitors of proteasomes or lysosomes. Therefore, differential down-regulation is not caused by separate degradation machineries, but rather by enhanced targeting of SNC80-

bound receptors to the same degradation process. This more effective targeting of

SNC80-bound receptors can be explained by the ability of SNC80 to activate more

“down-regulation” domains in the hDOR than DPDPE. Support for this conclusion comes from our experiments using the C-terminally truncated hDOR and the S363A mutant hDOR.

Using the hDOR(S363A) receptor we identify Ser363 as a C-terminal residue critical for

DPDPE-mediated down-regulation. In contrast, this residue has no effect on down-

regulation by SNC80 (Fig 6.10 and (Navratilova et al., 2004)). Phosphorylation of the

COOH-terminus is important for receptor internalization (Whistler et al., 2001). In the

mouse δ-opioid receptor, receptor phosphorylation is hierarchical with Ser363 being the

primary phosphorylation site upon DPDPE treatment (Kouhen et al., 2000). The

carboxyl-terminal residues are also involved in post-endocytic sorting of the δ-opioid

receptor to degradation pathways upon chronic peptide agonist treatment (Trapaidze et

al., 2000). Significantly, in the present study we demonstrate that although Ser363 has an

important role in the human δ-opioid receptor down-regulation by DPDPE, it is not

essential for receptor down-regulation by SNC 80. 82

Using a truncated human δ-opioid receptor, we demonstrate that truncation of the

carboxyl-terminus at Gly338 completely abolished DPDPE-, but not SNC80-mediated down-regulation (Fig. 6.9) and (Okura et al., 2000). Therefore, in addition to the carboxyl-tail, other domains are involved in SNC80-mediated down-regulation (Okura et al., 2003). These findings corroborate the conclusion that structurally distinct agonists utilize specific receptor domains to regulate receptor trafficking; namely, DPDPE requires an intact carboxyl-terminus while SNC80 is able to activate other domains, in addition to the C-terminus, to down-regulate the human δ-opioid receptor. This additional mechanism is able to compensate for the lack of Ser363 phosphorylation in the (S363A)

mutant receptor and in the truncated hDOR.

In conclusion of this section, we found that in CHO cells hDOR is down-regulated after chronic (several hours) treatment with an agonist in a time- and dose-dependent manner.

Furthermore, different agonists have different propensities to down-regulate the hDOR with the rank order: morphine < DPDPE < SNC80. Internalization of the hDOR is necessary to allow receptor down-regulation. Once inside the cell, the receptor is

degraded by the lysosomes, not by the proteasomes, as earlier suggested (Chaturvedi et al., 2001). Importantly, ubiquitination of the hDOR, in contrast to other GPCRs, is not

necessary for its efficient down-regulation. By truncating the C-terminus we demonstrate

that this domain is absolutely required for peptide agonist (DPDPE) mediated down-

regulation. Within the C-terminus we identified S363 as a crucial amino acid for peptide 83 agonist mediated down-regulation. In contrast to DPDPE, SNC80-mediated down- regulation is not significantly affected by S363A mutation. Moreover, SNC80 is still able to down-regulate the truncated hDOR. These findings indicate that a domain other than the C-terminus is activated by SNC80 binding and used to target the receptor for degradation. In search of this putative down-regulation domain we found that β-arrestin binding site in the third intracellular loop, the S249/S255 motif, is not involved. Other intracellular domains, probably involving phospho-S/T residues, are therefore activated by SNC80 and function as a secondary down-regulation motif. 84

7. AGONIST-MEDIATED DESENSITIZATION OF THE HUMAN DOR

Taken in part from (Navratilova et al., 2007)

Introduction

Agonist-mediated desensitization of the opioid receptor is thought to function as a

protective mechanism against sustained opioid signaling and may prevent the

development of opioid tolerance. However, the exact molecular mechanism of opioid

receptor desensitization remains unresolved, due to difficulties in measuring and

interpreting receptor desensitization. Agonist-mediated receptor desensitization is

usually classified on the basis of the duration as rapid or prolonged, and on the basis of

the mechanism triggering desensitization as homologous (caused by activity of the same

receptor molecule) or heterologous (caused by activation of other receptors) (Lohse et al.,

1990; Tran et al., 2004). Long lasting opioid receptor desensitization may be responsible

for some aspects of opioid tolerance. Rapid agonist-mediated desensitization, on the other

hand, may protect the organism from sustained opioid signaling and prevent the

development of tolerance. In this study we examined the molecular mechanisms of the

rapid homologous desensitization of the human delta opioid receptor (hDOR).

Multiple regulatory mechanisms contribute to the rapid homologous desensitization of the DOR: GRK-mediated receptor phosphorylation, β-arrestin binding and receptor internalization. Since these regulatory events are all causally dependent on each other, it is not clear which is the key process directly responsible for DOR desensitization. In 85

earlier studies, phosphorylation of the endogenous human DOR in SK-N-BE cells was found to correlate with receptor desensitization (Hasbi et al., 1998) and DOR desensitization occurred under conditions impairing receptor internalization (Hasbi et al.,

2000; Willets and Kelly, 2001). In contrast, Law and coworkers demonstrated that both

phosphorylation and internalization contribute to DOR desensitization in HEK 293 cells

(Law et al., 2000). High affinity binding of β-arrestin to the receptor is expected to

sterically interfere with G protein coupling, thereby effectively attenuating G protein

signaling (Gurevich and Gurevich, 2004). In some studies β-arrestin recruitment has been

considered a hallmark of the homologous receptor desensitization (Barak et al., 2006).

However, the direct role of β-arrestin binding in DOR desensitization has not yet been

studied in detail.

As reviewed above, significant inconsistencies exist in the interpretation of the

relationship between different DOR regulatory events and DOR desensitization. In

addition, there is no clear agreement on the definition of the term “receptor

desensitization”. Usually, receptor desensitization is interpreted as molecular changes

occurring at the level of the receptor, but is generally measured at the level of the

receptor’s function. The functional effect, however, depends not only on the functional

status of the receptor but also on the signal transduction amplification between the

receptor and the effector (Trzeciakowski, 1999). We hypothesize that some of the

difficulties in interpreting desensitization experiments are due to the non-linearity in 86

receptor/effector coupling and could be prevented if the number of desensitized receptors

is correctly determined.

In this chapter we will: 1) characterize hDOR desensitization by measuring desensitization of the GTPγS stimulation and cAMP inhibition; 2) develop a mathematical analysis based on the Operational Model of Agonist Action (Black et al.,

1985) to quantitatively evaluate the proportion of hDOR molecules desensitized after agonist treatment; 3) discuss the role of delta opioid receptor phosphorylation at S363, β-

arrestin binding and receptor endocytosis in rapid hDOR desensitization; and 4)

investigate agonist-specific differences in hDOR desensitization. To perform detailed

analyses of hDOR functions, we used a well characterized cellular model system,

Chinese hamster ovary (CHO) cells stably expressing either the wild-type hDOR or a

phosphorylation deficient mutant, in which S363 was mutated to alanine [hDOR(S363A)].

Receptor desensitization was measured by two second messenger assays: stimulation of

GTPγ[35S] binding and inhibition of forskolin-stimulated cAMP production. The results

were analyzed using a modified version of the Operational Model of Agonist Action

(Black et al., 1985). Our data demonstrate that: 1) the developed mathematical model can be used to analyze a complex phenomenon, such as the time-course of receptor desensitization, and correctly calculate the proportion of desensitized receptors; 2) although a mutation of the primary phosphorylation site, S363, does not prevent β-

arrestin2 recruitment and receptor internalization, it significantly attenuates deltorphin II- 87 mediated hDOR desensitization; and 3) SNC80 is more effective in desensitizing the hDOR than deltorphin II because it does not depend solely on phosphorylation of S363. 88

Results

Deltorphin II-mediated hDOR desensitization. To study rapid desensitization of the

human delta opioid receptor, we used recombinant Chinese hamster ovary (CHO) cells

expressing the human delta opioid receptor (1-209-19hDOR/CHO, “Materials and

Methods”) and measured the effects of agonist pretreatment on two different cellular

functions. 1) We measured concentration-response curves for deltorphin II-stimulated

GTPγ[35S] binding in cell membranes isolated from untreated (non-desensitized) cells or cells that were pretreated (desensitized) for increasing time intervals with deltorphin II

(100 nM). 2) In separate sets of experiments, we measured deltorphin II dose-response curves for forskolin-stimulated cAMP production in untreated and deltorphin II pretreated cells.

As seen in Fig. 7.1A, the concentration curves of GTPγ[35S] binding in hDOR/CHO cells

pretreated with deltorphin II were shifted to the right, and at longer times of agonist pre-

treatment maximal GTPγ[35S] binding was also reduced compared to the control cells.

The deltorphin II-mediated shift in EC50, and the reduction of Emax were dependent on the time of pretreatment (Table 7.1). Similarly, Fig. 7.1B demonstrates that in agonist

pretreated cells, deltorphin II was less efficacious and less potent in inhibiting forskolin-

stimulated cAMP production as compared to the control cells. This effect was also

dependent on the time of deltorphin II pre-treatment (Table 7.1). The reduced cellular

response to deltorphin II, as measured by either GTPγ[35S] binding or inhibition of cAMP

accumulation, is indicative of a decreased ability of pretreated receptors to couple to G 89

proteins. Our results are in agreement with previous studies, which found that DOR signaling is desensitized upon agonist treatment. However, although qualitatively important, this observation does not provide any quantitative measure of the number of desensitized receptors.

Operational Model of Drug Action. Because, in our system, receptor desensitization leads

to a reduction in Emax as well as an increase in EC50, both of these changes need to be

considered in order to correctly evaluate the proportion of desensitized hDOR. In this

circumstance, the reduction of the maximal effect after desensitization is not proportional

to the number of desensitized receptors. Therefore, changes occurring at the receptor

level, like receptor phosphorylation or internalization, cannot be directly correlated with

the measured changes in the maximal effect. To obtain a quantitative estimate of the

number of desensitized receptors we employed the Operational Model of Drug Action

developed by Black and collaborators (Black et al., 1985), which describes the

correlation between a biological effect E and agonist concentration [A] as a function of

three parameters: Em, KA and τ:

E τ[A] E = m (7.1) A []AK ++ τ []A

Where Em, or the operational maximum, represents the maximum possible effect in the tissue, KA is the dissociation constant of the agonist and τ is the operational efficacy or

the transducer ratio. Parameter τ is defined as:

τ = [ ]T /[KR E ] (7.2) 90

Where [R]T is the total concentration of receptors and [KE] is the concentration of

occupied receptors required to produce half of the maximum Em. The operational model

describes how both Emax and EC50 change when either the total concentration of receptors

[R]T, or the concentration of receptors [KE] necessary to produce half of the operational

maximum (Em) changes. Therefore the Operational Model calculates the operational

efficacy τ (and thus the receptor number) based on changes in both Emax and EC50.

The dependence of the effect on [R]T was used earlier to estimate the proportion of β2

adrenergic (Lohse et al., 1990) or μ opioid receptors (Osborne and Williams, 1995) desensitized by agonist pretreatment. Using the approach of Lohse (Lohse et al., 1990), a paired set of dose-response curves obtained pre- and post-desensitization was fitted simultaneously using equation (7.1). The parameters Em and KA characterize the cellular

system and the agonist used, and therefore do not change with the number of desensitized

receptors. The only parameter that distinguishes the pre- and post-desensitization curves

is the operational efficacy (τ). Since τ is proportional to the number of functional (non-

desensitized) receptors, the fraction of the desensitized receptors D can be calculated as:

τ D 1−= eddesensitiz (7.3) τ control

In this equation, τcontrol is the initial operational efficacy in the non-desensitized state and

τdesensitized is the operational efficacy after a fraction of receptors have been desensitized

(Lohse et al., 1990). Two functional effects of hDOR stimulation by deltorphin II were measured in this study: stimulation of GTPγ[35S] binding and inhibition of forskolin- 91

stimulated cAMP formation. In order to quantitatively analyze the magnitude of the

effect, the measured levels of GTPγ[35S] binding or cAMP formation were normalized to

0% in the absence of the agonist and to 100% at saturating amounts of the agonist in

untreated (control) cells. In GTPγ[35S] binding experiments, deltorphin II-stimulated

35 levels of bound GTPγ[ S] (B) were first adjusted by subtracting the basal levels (B0B ) and

then normalized to the percent of adjusted maximum stimulation (BmaxB ) in control cells

according to the equation: Effect = (B - B0B )/ [BmaxB (con) - B0B (con)] × 100%.

In cAMP experiments, the residual level of cAMP production in control cells, which was

not inhibited by saturating concentrations of deltorphin II (Pmax(con)), was first subtracted

from all dose-response curves for deltorphin II-mediated inhibition of cAMP production

(P). Subsequently, each dose response curve was normalized to 100% forskolin- stimulated levels in the absence of the agonist (P0) and expressed as an increase in cAMP

inhibition, rather then decrease in cAMP production, in accordance with the equation:

Effect = (P0 – P)/[ P0 – Pmax(con)] × 100%. The average maximum level of forskolin-

stimulated cAMP in the absence of the agonist was 37 ± 12 and 30 ± 9 pmol/100,000

cells in hDOR/CHO and hDOR(S363A)/CHO cells, respectively. The residual level of

cAMP was 12 ± 5 and 14 ± 5 pmol/100,000 cells in hDOR/CHO and

hDOR(S363A)/CHO cells, respectively. During the assay time interval (0-60 min) the

levels of cAMP production were not affected by agonist-mediated superactivation of

adenylyl cyclase (Varga et al., 2003).

Modification of the Operational Model to calculate the kinetics of receptor

desensitization. To analyze data from experiments involving the time-course of receptor

desensitization, we have modified the operational model by adding restrictions that

should be fulfilled for the parameterτ. In the simplest model, we can assume that the

number of functional receptors [R]T, and consequently τ, decreases exponentially as a

function of the time of the pretreatment, according to:

τ t = τ 0 −kt)exp()( +τ ∞ (7.4)

where k is the rate constant of desensitization, τ∞ is the asymptotic value of τ approached

when the maximum number of receptors has been desensitized. (τ0 + τ∞) = τcontrol is the initial τ value in non-desensitized cells. We can now introduce this restriction on

parameters τ into equation (7.1) and obtain more meaningful estimates of the time course

of receptor desensitization, which will allow us to calculate the rate of receptor desensitization and the maximum level of desensitization at equilibrium. Substituting τ(t) from equation (7.4) into equation (7.1) yields:

E τ +− τ Akt ]}[)exp({ E = m 0 ∞ (7.5) A AK τ 0 kt +−++ τ ∞ })exp(1]{[

This equation is a five-parameter equation, which describes a whole series of dose-

response curves as a function of one variable, the time of deltorphin II pretreatment. Fig.

7.2 illustrates the theoretical dependence of the dose-response curves on the time of

pretreatment as predicted from equation (7.5). The theoretical curves were generated

-1 using arbitrary values Em = 100, τ0 = 20, τ∞ = 1, KA = 1 nM and k = 0.1 min , which were chosen to describe a realistic cellular model. The selected value of τ0 = 20 describes a 93

cellular system with significant receptor reserve. In this situation, Emax of the dose

response curve for non-desensitized receptors approaches the theoretical operational maximum Em and the EC50 is smaller than KA. As τ decreases with increasing time of

pretreatment, EC50 increases, initially with little change in the Emax. After an extended

pretreatment period, a decrease in Emax becomes apparent and is accompanied by further

increase in EC50 to the asymptotic value KA. In systems with no receptor reserve the

maximum effect (Emax) does not reach the theoretical maximum Em in the non-

desensitized state. Therefore, an attenuation of the Emax is already apparent when only a small proportion of the receptors have been desensitized. However, because the reduction in Emax is always accompanied by an increase in EC50, the reduction in Emax alone will

underestimate the proportion of desensitized receptors. This emphasizes the importance

of using changes in both Emax and EC50 when evaluating desensitization experiments. The

presented analysis expands the Operational Model of Agonism to calculate the kinetics of

receptor desensitization. A similar approach can also be applied to calculate

desensitization of any GPCR under various experimental conditions.

Quantitative evaluation of the kinetics of hDOR desensitization using the modified

Operational Model. The experimental data for the time-course of receptor desensitization were analyzed using equation (7.5), with the parameters Em and KA shared for all time

points (0-60 min). For curves corresponding to times (0-45 min), τ is defined by equation

(7.4) in which τ0 and τ∞ were fitted shared parameters; for the 60 min curve τ = τ60 was fitted independently since at this time long-term desensitization mechanisms may already 94 be present. In Figs. 7.3A and 7.3B the data obtained for the time course of hDOR desensitization (Figs. 7.1A and 7.1B, respectively) were fitted using equation (7.5). All five parameters (Em, τ0, τ∞, KA and k) were shared among all dose-response curves for

GTPγ[35S] binding and again in a separate calculation using inhibition of cAMP production. The values obtained for these parameters for the GTPγ[35S] and the cAMP assays are summarized in Table 7.2. The initial value of τcontrol = (τ0 + τ∞) is reported in

Table 7.2 instead of the fitted τ∞ value.

From Table 7.2 we can observe the fitted operational maximum Em is 120 ± 5% and 106

± 4% for the GTPγ[35S] and cAMP assays, respectively. In GTPγ[35S] measurements the operational efficacy in control cells τcontrol = (τ0 + τ∞) was 7.8 ± 2.4. Since by definition τ

= [R]T/[KE], we can conclude that the expression of the receptors in our cells is approximately 7.8 times higher than the number of receptors necessary to produce half of the operational maximum Em. Consequently, some receptor reserve exists in our cellular system for the GTPγ[35S] stimulation assay. After deltorphin II pretreatment, the operational efficacy τ is reduced and, at infinity, would reach the value 1.4 ± 0.5. The operational efficacy measured by the cAMP assay τcontrol = (τ0 + τ∞) = 36 ± 21 is higher than that measured by the GTPγ[35S] assay. The expression of the receptors in our cells is approximately 36 times higher than the number of receptors necessary to produce half of the operational maximum Em in the cAMP assay. This supports the conclusion that signal amplification measured by the cAMP assay is higher, as would be expected because it 95

reflects signaling events several steps downstream of G protein activation. Therefore,

even when significant portion of the receptors is desensitized upon agonist treatment,

only a small reduction in Emax is observed in cAMP assay and the predominant effect of

desensitization remains the rightward shift in EC50 value. Importantly, the desensitization

calculated using equation (7.3) from τ values before (τcontrol) and after (τ∞) the maximum

number of receptors have been desensitized correlates very well when comparing the two

assays: 82% for both GTPγ[35S] and cAMP assays, confirming that this parameter is not

dependent on which assay is chosen. The half-lives of receptor desensitization estimated

35 using t1/2 = 0.693/k from GTPγ[ S] and cAMP assays were 9.8 min and 8.4 min,

respectively. The estimated theoretical dissociation constants KA for deltorphin II

measured by the two assays were 81 ± 16 nM and 16 ± 8 nM, respectively. Values of KA are about 10 -50 fold larger than the EC50 values measured in untreated cells, again

supporting the presence of spare receptors.

S363A mutation attenuates hDOR desensitization. The molecular mechanisms underlying receptor desensitization remain controversial. It is currently accepted that several mechanisms may contribute to rapid receptor desensitization including receptor phosphorylation, binding of β-arrestin and receptor internalization. In HEK 293 cells phosphorylation of the mouse DOR is hierarchical with S363 in the C-terminal region

being the primary phosphorylation site (Kouhen et al., 2000). In addition, mutation of

S363 to Ala was shown to reduce agonist-mediated uncoupling of the receptor from

adenylyl cyclase signaling (Law et al., 2000). Since receptor phosphorylation leads to β- 96

arrestin recruitment and receptor internalization, it is not clear which of these

mechanisms is crucial for hDOR desensitization.

To study the role of S363 phosphorylation in the desensitization of the human DOR we

mutated S363 to alanine and created a CHO cell line stably expressing the mutant

hDOR(S363A) receptor (Navratilova et al., 2004). In Chapter 3 we demonstrated using

the Western blot analysis with a specific phospho-hDOR(S363) primary antibody (Cell

Signaling) that treatment of hDOR/CHO cells with deltorphin II (500 nM) leads to

phosphorylation of the hDOR at S363 (Fig. 3.1). The half-life of receptor phosphorylation

363 estimated from the time course of S phosphorylation is t1/2 = 1.7 ± 0.2 min (Fig. 3.3).

In addition, these experiments confirmed that S363 phosphorylation is eliminated by

mutating this amino acid to alanine (Fig. 3.1).

In this chapter we investigate the effect of S363A mutation on the hDOR desensitization.

Using the GTPγ[35S] binding assay, we showed that hDOR(S363A) mutant receptor was

desensitized by deltorphin II pretreatment to a much lesser extent than the wild-type

hDOR, as evident from a smaller shift of the dose-response curve (Fig. 7.4A). This

finding was reproduced by measuring the dose-response curves for inhibition of cAMP

production in hDOR/CHO and hDOR(S363A)/CHO cells (Fig. 7.4B). In order to

evaluate the statistical significance of these findings we fitted the observed data with

sigmoidal concentration-response curves using the Hill coefficient equal to 1, and

compared all curves using the extra sum-of-squares F test calculated with GraphPad 97

Prizm 4. The resulting Emax and EC50 values for the individual curves are shown in Table

7.3. We found that the control dose-response curves for untreated wild-type and mutant receptors were not significantly different in either GTPγ[35S] or cAMP assay, confirming that the S363A mutation had no effect on acute receptor signaling. Importantly, in hDOR/CHO cells the shift in dose-response curves upon deltorphin II pretreatment was

35 significant (EC50 increased 4.5-fold from 6.9 to 31 nM in GTPγ[ S] assay, p<0.0001 and

5.5-fold from 1.4 to 7.5 nM in cAMP assay, p<0.05). Conversely, in hDOR(S363A)/CHO cells the shift upon deltorphin II pre-treatment was much smaller

(2.3-fold from 5.8 to 13 nM and 1.4-fold from 2.0 to 2.8 nM in GTPγ[35S] and cAMP assay, respectively). This shift was statistically significant in GTPγ[35S] assay (p<0.001) but did not reach statistical significance in cAMP assay (p=0.48). Finally, the difference between the deltorphin II-mediated shift in hDOR/CHO cells and hDOR(S363A)/CHO cells was statistically significant for both assays (p<0.0001). These analyses clearly show that pretreatment of the hDOR/CHO cells for 60 min with 100 nM deltorphin II causes desensitization of the G protein signaling which is significantly attenuated in hDOR(S363A)/CHO cells. This finding identifies S363 as an important residue for hDOR desensitization.

To estimate the proportion of the wild-type and S363A mutant receptors desensitized by agonist pretreatment, we applied the Operational Model of Agonist Action (Black et al.,

1985) as explained in ”Materials and Methods”. For this analysis, we assumed that the dose-response curves for the hDOR and hDOR(S363A) under control conditions 98

(untreated) are described by the same equation and, therefore, share the parameter τ =

τcontrol. This assumption is justified by the results of the regression statistical comparison.

The fitted curves for GTPγ[35S] binding and inhibition of cAMP production experiments

in cells expressing the wild-type and S363A mutant receptors are presented in Figs. 7.5A

and 7.5B, respectively, and the calculated values for the fitted parameters are summarized

in Table 7.4. From the fitted parameters, τ, desensitization of the wild type and S363 mutant hDOR was computed using equation (7.3). Desensitization of the hDOR achieved after 60 min of deltorphin II treatment was 85% and 81% in GTPγ[35S] and cAMP

assays, respectively; values comparable to the maximal desensitization calculated from

the desensitization time course (82% in both assays). On the other hand, desensitization

of the S363A mutant was markedly lower, reaching only 20% and 18% in GTPγ[35S] and cAMP assays, respectively. Because the S363A mutation significantly reduces the number of receptors desensitized by deltorphin II, S363 must play a crucial role in receptor

desensitization.

One possibility for the role of S363 in hDOR desensitization may be that S363

phosphorylation hinders the interaction of the receptor with G proteins. Alternatively,

other events dependent on S363 phosphorylation may also be necessary for complete

receptor uncoupling. These events may include: phosphorylation of other residues;

binding of GRKs or β-arrestins to the receptor; receptor sequestration into clathrin-coated

pits, and receptor internalization into endosomes. Since deltorphin II-mediated

desensitization of the hDOR is significantly reduced by the mutation of S363 to alanine, 99

we hypothesized that every molecular event involved in receptor desensitization would

also be significantly attenuated by this mutation.

As described earlier in Chapter 4 the ability of the hDOR to recruit β-arrestin2-GFP to

the plasma membrane upon treatment of the cells with saturating concentrations of

deltorphin II (100 nM) was not altered by mutating the S363 to alanine (Fig. 4.4).

Conversely, in this chapter we show that S363A mutation almost completely eliminates rapid deltorphin II-mediated receptor desensitization. These results indicate that β-

arrestin2 binding to the receptor is not sufficient to uncouple the hDOR from its cognate

G proteins. Similarly, in Chapter 5 we demonstrated that S363A mutation does not

eliminate receptor-mediated internalization of a fluorescently labeled analog of

deltorphin II ([Gln4]deltorphin-rhodamine, Fig. 5.4). Therefore we can also deduce that

hDOR internalization is not sufficient to completely uncouple the receptor from G

proteins. Taken together, we conclude that neither β-arrestin2 binding nor internalization

of the receptor themselves are sufficient to desensitize the hDOR without

phosphorylation of S363.

Agonist-specific hDOR desensitization. We have demonstrated in Chapter 6 and in

(Varga et al., 2004) that peptide agonists and SNC80 differentially down-regulate the

hDOR (Fig. 6.2). Moreover, using truncated and S363A mutant hDOR constructs, we

showed that qualitative differences exist in the efficacies of peptide agonists and a non-

peptide agonist SNC80 to down-regulate the receptor. Namely, peptide agonists required 100

the C-terminus while SNC80 was able to use other intracellular domains to down-

regulate the hDOR (Fig. 6.9). In addition, phosphorylation of S363 seems to be crucial for

peptide agonists but is dispensable in SNC80-mediated down-regulation (Fig. 6.10)

(Navratilova et al., 2004).

In this chapter we extended our down-regulation studies and investigated whether hDOR

is also desensitized more effectively by SNC80 than by peptide agonist deltorphin II.

Furthermore, we studied the effect of S363A mutation on receptor desensitization by

SNC80. We hypothesized that due to putative additional phosphorylation sites activated

by SNC80, hDOR would be desensitized more effectively and desensitization would not

depend solely on S363.

SNC80 desensitizes the hDOR more effectively than deltorphin II. To study the

differences in the abilities of peptide agonists (deltorphin II) and SNC80 to desensitize the hDOR, we measured concentration-response curves of deltorphin II-stimulated

GTPγS binding in cell membranes isolated from non-desensitized (untreated) cells or cells that were pretreated for 60 min with either deltorphin II (100 nM) or SNC80 (100 nM). In a separate set of experiments we measured deltorphin II dose responses of forskolin-stimulated inhibition of cAMP production in untreated and deltorphin II or

SNC80 pretreated cells. Fig. 7.6A shows that in cells expressing the wild-type hDOR the

concentration curves of GTPγS binding are shifted to the right and the maximum GTPγS

binding is reduced by pretreatment of the cells with deltorphin II or SNC80 compared to 101

the control cells. Notably, the desensitization effect is larger with SNC80 than with

deltorphin II. A similar trend is observed in Fig. 7.6B where pretreatment of the cells

with both agonists reduces the efficacy of deltorphin II to inhibit forskolin-stimulated

cAMP production; however, the effect of SNC80 was again stronger. The experimental

data were fitted with sigmoidal dose-response curves using the GraphPad Prizm 4

software, the calculated Emax and EC50 values are summarized in Table 7.5. To determine

the statistical significance of the differences between the curves for control, deltorphin II-

pretreated and SNC80-pretreated cells we compared the fitted curves using extra sum-of-

square F test. In the GTPγS assay, deltorphin II and SNC80 shifted the dose-response curves for hDOR activation significantly to the right (EC50 increased 4.6 fold, p<0.0001

for deltorphin II and 3.4 fold, p<0.01 for SNC80). The difference between deltorphin II

and SNC80 is not statistically significant (p=0.6). SNC80 however, attenuated the

maximum effect Emax more than deltorphin II (81% and 72% of the untreated control for

deltorphin II and SNC80, respectively). In the cAMP assay again both deltorphin II and

SNC80 significantly shifted the dose-response curves to the right (EC50 increased 3.0

fold, p<0.01 for deltorphin II and 8.3 fold, p<0.0001 for SNC80). Although SNC80

produced a larger shift compared to deltorphin II, the difference between the two agonists

did not reach statistical significance (p=0.09).

S363A mutation of the hDOR has a differential effect on SNC80- and deltorphin II-

mediated receptor desensitization. As shown in Fig. 7.7A and Table 7.5, in cells

expressing the S363A mutant hDOR, deltorphin II pretreatment produced a much lower 102

desensitization of GTPγS binding compared to the wild-type receptor (EC50 increases 1.7

fold, p<0.01). In contrast, the non-peptide agonist SNC80 considerably desensitized the

deltorphin II dose response curve for GTPγS binding measured in hDOR(S363A)/CHO

cells (EC50 increases 4.4 fold, p<0.0001). This finding was reproduced by measuring cAMP inhibition in hDOR(S363A) cells pretreated with deltorphin II or SNC80. Fig.

7.7B and Table 7.5 indicates that deltorphin II pretreatment does not appreciably desensitize the hDOR(S363A) while SNC80 pretreatment does (EC50 increased 1.5 fold,

p=0.17 for deltorphin II and 2.5 fold, p<0.0001 for SNC80). The difference between

deltorphin II and SNC80 is statistically significant in GTPγS assay (p=0.013), but did not

reach statistical significance in cAMP assay (p=0.06).

Quantitative evaluation of agonist-specific hDOR or hDOR(S363A) desensitization using the Operational Model of Drug Action. In order to estimate the proportion of receptors desensitized by agonist pretreatment, we used the Operational model of agonist action

(Black et al., 1985). By fitting the measured data to the theoretical equation of the operational model (equation 7.1) we calculated the best fit of the operational maximum

Em and the dissociation constant KA for deltorphin II in our cellular model as well as the operational efficacy τ for untreated, and deltorphin II or SNC80 pretreated cells. The

fitted curves for the hDOR are plotted in Fig. 7.6C,D and for the hDOR(S363A) in Fig.

7.7C,D. The level of desensitization caused by different agonists was estimated from the

fitted τ values using equation (7.5). The calculated values of these parameters for GTPγS

binding and cAMP inhibition experiments in cells expressing the wild-type and (S363A) 103

mutant receptors are summarized in Table 7.6. The estimates of the operational efficacy τ

(16 and 15 for the hDOR and hDOR(S363A) in GTPγS assay, respectively; 18 and 30 for

the hDOR and hDOR(S363A) in cAMP assay, respectively) correlate very well with the

previously obtained values (Tables 7.3 and 7.5) and with the conclusion that more spare

receptors are available in cAMP measurements than in GTPγS measurements. According

to equation (7.2), agonist-mediated desensitization of the receptors results in a reduction

of the operational efficacy (Table 7.6). In both assays, SNC80 reduced the τ value for the

hDOR more than deltorphin II, signifying a larger desensitizing efficacy of SNC80.

Using equation (7.5) the proportion of desensitized hDOR was estimated to be 79% or

78% for deltorphin II and 84% or 90% for SNC80 in GTPγS or cAMP assays,

respectively. Mutation of S363 to alanine in the hDOR further enhanced the differential

desensitizing efficacies of deltorphin II and SNC80. The calculated proportion of

desensitized hDOR(S363A) was 29% or 14% for deltorphin II and 83% or 50% for

SNC80, in GTPγS or cAMP assays, respectively. Finally, the operational model provides an estimate of deltorphin II dissociation constants KA for stimulation of GTPγS binding

and inhibition of cAMP production. In GTPγS assay the calculated KA values were

116±22 nM for the hDOR and 112±26 nM for the hDOR(S363A). In cAMP assay the KA values are lower: 20±4 nM for the hDOR and 68±53 nM for the hDOR(S363A), which correlates to the previously obtained results (Table 7.4).

It can be concluded from these results that desensitization of hDOR shows agonist specificity revealed by the mutation of the primary phosphorylation site in the receptor’s 104

C-terminal domain. This finding corresponds to our previous study (Varga et al., 2004) which demonstrated the differential ability of a peptide agonist DPDPE and a non-peptide agonist SNC80 to induce delta opioid receptor down-regulation. Similarly to the present work, agonist specific down-regulation was further amplified by the mutation of S363 to A

(Navratilova et al., 2004). However, receptor down-regulation is not likely to be responsible for the observed rapid desensitization, since the time frame for receptor degradation (down-regulation) is several hours, rather than several minutes observed for hDOR desensitization. Notably, we have not detected any measurable loss of receptor binding sites after 1h pretreatment in our cell line (Fig. 6.1). 105

Discussion

A major obstacle in understanding the exact mechanism of rapid homologous DOR desensitization is the difficulty to correctly determine the fraction of desensitized receptors. In our present study we used a mathematical analysis based on the Operational

Model of Agonist Action (Black et al., 1985) to calculate the proportion of desensitized hDOR from measurements of deltorphin II-mediated desensitization of two functional effects: stimulation of GTPγ[35S] binding and inhibition of cAMP accumulation. We expanded the Operational Model by incorporating the time-course of receptor desensitization. To determine the molecular mechanism of rapid hDOR desensitization we investigated the role of deltorphin II-mediated phosphorylation of S363, translocation of β-arrestin2 to the plasma membrane and hDOR internalization. The results of these experiments demonstrate that: 1) the operational model allows one to correctly calculate the proportion of desensitized receptors; and 2) β-arrestin binding and receptor internalization are not sufficient to bring about hDOR desensitization without phosphorylation of S363.

As emphasized in “Materials and Methods”, receptor desensitization results not only in an attenuation of the maximum effect but also in a rightward shift of the concentration response curve (Fig. 7.2). Consequently, the measured reduction of a functional effect is not directly proportional to the number of desensitized receptors, particularly in the presence of spare receptors (Borgland et al., 2003; Connor et al., 2004). To minimize this disproportionality, many desensitization studies have been performed either in cell lines 106

which express low levels of the DOR, such as SK-N-BE (Hasbi et al., 1998), or in cells

where receptor level was experimentally manipulated (Law et al., 2000). In other studies,

a mathematical analysis was used to calculate the proportion of desensitized receptors.

Lohse adapted the Operational Model of Agonism (Black et al., 1985) and, assuming that

desensitization represents a loss of signal-transduction efficacy of the receptor/effector

system, derived an equation to calculate the number of desensitized receptors (Lohse et al., 1990). Similarly, Whaley and coworkers (Whaley et al., 1994) adapted models of

receptor/G protein activation of adenylyl cyclase to derive expressions that predict changes in EC50 and Emax as the receptor number varies. It can be demonstrated that after

transformation, the two methods yield identical equations. Since the Operational Model

(Black et al., 1985) is not limited to activation of adenylyl cyclase but describes generally any relationship between receptor activation and effector function, we used this method in our study. In addition, we expanded the method of Lohse (Lohse et al., 1990), and assuming that the time dependence of receptor desensitization can be approximated by an exponential function, we derived an equation which allows us to estimate the half-life of hDOR desensitization.

Results of our analysis indicate that in recombinant CHO cells the hDOR is desensitized

by deltorphin II treatment in a time-dependent manner with a half-life of approximately

10 min. Desensitization after 60 min of agonist treatment reached maximal levels that

corresponded to about 80% of receptors desensitized. The same treatment desensitized

only 20% of a mutant hDOR in which the primary phosphorylation site, S363, was 107

mutated to alanine. The obtained rate of DOR desensitization corresponds to results

reported by other investigators using cell lines with low expression levels of DOR

(Allouche et al., 1999; Law et al., 2000). These results suggest that the use of the

Operational Model enables us to correctly analyze receptor desensitization even in high

receptor expression systems containing spare receptors.

Agonist-mediated phosphorylation of the DOR by high affinity agonists was

demonstrated in our laboratory (Okura et al., 2000) as well as by other investigators

(Allouche et al., 1999; Eisinger et al., 2002; Li et al., 2003; Willets and Kelly, 2001).

Based on mutational analysis, S363 in the C-terminus of the mouse DOR was identified as

the primary phosphorylation residue (Kouhen et al., 2000). In Chapter 3, we found using

a phospho-hDOR(S363) primary antibody that S363 of the human DOR is phosphorylated

upon deltorphin II treatment in a time-dependent manner with a half-life of about 2 min.

This half-life is shorter than the half-life of receptor desensitization (~10 min), indicating

that the relationship between receptor phosphorylation and desensitization is not direct.

Indeed, elimination of S363 phosphorylation by mutation of this residue to alanine did not completely prevent hDOR desensitization produced by 60 min deltorphin II treatment, but only reduced the number of desensitized receptors from 80 to 20%. Therefore, the hDOR is desensitized by a S363 dependent and by a S363 independent mechanism.

Phosphorylation-dependent and independent mechanisms of desensitization were also

identified for rhodopsin (Xu et al., 1997) as well as for the DOR (Law et al., 2000). In

addition, it was reported that binding of a kinase-negative mutant (K220R) of GRK2 108

could desensitize the parathyroid hormone receptor (Dicker et al., 1999), and the

endothelin A and B receptors (Freedman et al., 1997) even in the absence of receptor

phosphorylation. However, the phosphorylation-independent mechanism of DOR desensitization likely does not involve GRK, since binding of GRK2 alone to a phosphorylation deficient mutant of the DOR did not cause desensitization of GTPγ[35S] binding in HEK 293 cells (Li et al., 2003).

Binding of β-arrestin to GPCRs requires phosphorylation of several S/T residues in the intracellular domains of the receptors (Gurevich and Gurevich, 2004). Accordingly, receptor phosphorylation was found to be the rate-limiting step for β-arrestin and β2-

adrenergic receptor interaction (Krasel et al., 2004). GPCR- β-arrestin interaction, in turn,

is considered to be a key step in the uncoupling of receptors from G proteins (Barak et

al., 2006). In agreement with this concept, cotransfection of β-arrestin1 or 2 with GRK3

was required for desensitization of DOR-coupled inwardly rectifying potassium channel

(Kir3) in Xenopus oocytes (Lowe et al., 2002). Surprisingly, our confocal microscopy

images (Chapter 4) show that mutation of the primary phosphorylation site S363 in the

hDOR that significantly impairs receptor desensitization did not prevent the recruitment of β-arrestin2-GFP to plasma membrane. No visually apparent differences were observed

in the time course and the extent of β-arrestin2-GFP translocation to the hDOR or

hDOR(S363A). Fluorescent microscopy does not allow quantification of the β-arrestin-

receptor interaction, however, β-arrestin2-GFP translocation clearly occurs after 5 min of

deltorphin II treatment in the hDOR(S363A) mutant receptor. Therefore, any undetected 109

changes in the interaction between β-arrestin2-GFP and the hDOR(S363A) compared to

the wild type hDOR cannot explain an almost complete blockade of receptor

desensitization observed after 60 min of agonist treatment. These results demonstrate that

β-arrestin2 binding to hDOR alone is not sufficient for receptor desensitization without

S363 phosphorylation. In support of this conclusion, it was reported recently using locus coeruleus neurons from β-arrestin2 knock-out mice that lack of β-arrestin2 expression

has no effect on the rate or the magnitude of the mu opioid receptor desensitization (Dang

and Christie, 2006). Nevertheless, in our system β-arrestin binding may still function as

an essential mechanism in the residual S363 phosphorylation-independent desensitization,

probably in conjunction with receptor internalization. Indeed, phosphorylation independent, β-arrestin2 dependent internalization of the DOR was reported in HEK 293

cells (Zhang et al., 2005). In addition, a study by Burns et al. found that visual arrestin is

able to quench nonphosphorylated rhodopsin (Burns et al., 2006). Recently, Marion and

coworkers (Marion et al., 2006) identified a common β-arrestin binding site formed by

ten residues of the second intracellular loop of most GPCRs which is independent of

GRK phosphorylation but dependent on agonist activation.

Desensitization of the DOR was reported in some studies to be independent of receptor

internalization (Hasbi et al., 2000; Willets and Kelly, 2001). In contrast, Law et al.

reported that both receptor phosphorylation and receptor endocytosis contribute to DOR

desensitization in HEK 293 cells (Law et al., 2000). Our results support the idea that the

primary mechanism of hDOR desensitization is S363 phosphorylation-dependent. In 110

addition, a small fraction of receptor desensitization is S363 phosphorylation-independent.

Further studies are needed to determine whether phosphorylation of other residues is responsible for the residual desensitization or whether it is indeed caused by

phosphorylation-independent and β-arrestin-dependent internalization.

By comparing desensitization produced by two structurally distinct agonists we were able

to demonstrate that SNC80 more effectively desensitizes the wild-type hDOR than

deltorphin II. This differential desensitization efficacy is caused by agonist-selective

activation of different receptor domains, not just by partial activation of the same

domains, since S363A mutation specifically affects deltorphin II-mediated

desensitization. The ability of SNC80 but not deltorphin II to activate non-C-terminal

receptor domains was also demonstrated in phosphorylation and down-regulation studies.

In summary, this study demonstrates that the Operational Model of Agonism provides an

accurate mathematical approach to quantify the number of receptors desensitized by

agonist treatment. By using this model on wild-type and phosphorylation-deficient

mutant hDOR we were able to correlate desensitization with receptor phosphorylation, β-

arrestin2 translocation and receptor internalization. We have demonstrated that in CHO

cells expressing the hDOR, deltorphin II treatment leads to phosphorylation of S363, translocation of β-arrestin2 to the plasma membrane, receptor internalization and uncoupling from G proteins. Interestingly, the S363A mutation completely eliminates phosphorylation of this residue, yet it has virtually no effect on β-arrestin2 translocation 111

and receptor internalization. On the other hand, S363A mutation significantly attenuates

receptor desensitization by deltorphin II. These results provide evidence that

phosphorylation of S363 is required to uncouple the receptor from G proteins. Recruitment

of β-arrestin2 and receptor internalization, on the other hand, are not sufficient to

desensitize the hDOR without S363 phosphorylation. Therefore, we conclude that phosphorylation of S363 represents the primary mechanism of the human delta opioid

receptor desensitization by deltorphin II. SNC80 desensitizes the hDOR more effectively

than deltorphin II. In addition, S363A mutation has little effect on SNC80-mediated

desensitization. Whether hDOR desensitization by SNC80 requires phosphorylation of

other residues or whether β-arrestin dependent mechanism is sufficient to desensitize

SNC80-bound receptors remains to be determined. 112

8. SUMMARY AND CONCLUSION

Enhanced plasma membrane expression of the DORs was observed after chronic

morphine treatment (Zhu et al., 1999) and in chronic inflammatory and neuropathic pain

states (Cahill et al., 2007). Importantly, upregulation of delta opioid receptors led to potentiation of deltorphin II-mediated analgesia in morphine tolerant animals (Cahill et al., 2001) and in animal models of chronic inflammation (Cahill et al., 2003). Delta

opioid receptor acting drugs may therefore, provide better alternatives to the currently

used analgesics for treatment of some pathological pain conditions. Agonist-mediated

regulation of the hDOR by non-selective agonists such as morphine, or delta selective

analgesics, will inevitably alter the effectiveness of the pain therapy. We hypothesize that

modifications of DOR regulation, either by use of pharmacological inhibitors of regulatory pathways or by use of DOR agonists with specific regulatory properties, may

improve analgesic properties and minimize the side effects of opioid treatment.

Therefore, in this study, we investigated multiple aspects of hDOR regulation using a

recombinant Chinese hamster ovary cellular system expressing the wild-type or various mutant hDOR constructs. We studied receptor phosphorylation using Western blot

analysis, recruitment of β-arrestin and receptor internalization using confocal laser

scanning microscopy, receptor down-regulation using radioligand binding, and receptor

desensitization using GTPγ[35S] binding and cAMP inhibition assays. To determine the

cellular mechanisms of these events we utilized a variety of selective inhibitors of the 113

regulatory pathways. Furthermore, to identify the signaling domains in the hDOR

responsible for receptor regulation, we generated several mutated and truncated receptor

constructs using molecular cloning techniques. We studied agonist differences in cellular

mechanisms of hDOR regulation and in molecular domains involved in this regulation

using three classes of DOR agonists: morphine, peptide agonists (DPDPE, deltorphin II)

and SNC80.

We confirmed that full delta agonists, such as deltorphin II, DPDPE and SNC80, phosphorylate the hDOR at S363, facilitate translocation of β-arrestin2 to the cell surface,

and promote receptor internalization and down-regulation. In contrast to reports that

DOR interacts strongly with both β-arrestins in Xenopus oocytes (Lowe et al., 2002), we

observed that in CHO cells, the interaction of β-arrestin2 with the hDOR is transient and

accordingly, β-arrestin2 dissociates soon after the receptor is internalized. The hDOR is

internalized via clathrin-coated pit-mediated endocytosis, as reported earlier (Trapaidze et

al., 1996). However, we found that the hDOR internalizes into vesicles separate from

transferrin-containing vesicles and the two populations of CCPs can be separated by their

differential selectivity to cholesterol depletion. In contrast to these acute effects, which

occur minutes after agonist treatment, receptor down-regulation takes place in response to

sustained (several hours) agonist treatment. We confirmed that internalization is required

for hDOR down-regulation. Furthermore, we clarified contradictory data in the literature

regarding the digestive system involved in DOR down-regulation (Chaturvedi et al.,

2001; Yadav et al., 2006). We demonstrate that in CHO cells the hDORs are not down- 114 regulated by the proteasome in response to agonist treatment, as initial results using proteasomal inhibitors may indicate, but rather by the lysosomes. However, proteasomes play a critical role in the synthetic pathway for the regulation of newly synthesized receptors (Petaja-Repo et al., 2001). Therefore, separating new protein synthesis and agonist-mediated receptor down-regulation is crucial for the correct interpretation of the effects of proteasomal inhibitors on the total receptor number. Next, by measuring changes in GTPγ[35S] stimulation and inhibition of cAMP production upon acute (2-60 min) agonist activation, we found that the regulatory events induced by the agonist lead to receptor desensitization. Further investigation led us to a conclusion that the primary event required for receptor desensitization is phosphorylation of S363 in the C-terminus of the hDOR. Surprisingly, our study indicates that β-arrestin2 binding and receptor internalization alone, without phosphorylation of S363, are not sufficient to fully desensitize the receptor (Navratilova et al., 2007).

Morphine is usually characterized as a partial agonist at the DOR. However, in our system, morphine is still able to stimulate GTPγ[35S] binding and inhibit cAMP to an approximately 80-90% level of the full agonists deltorphin II, DPDPE and SNC80 (Table

8.1). It is reported in the literature, that morphine is disproportionally less efficient in inducing receptor regulatory effects (Keith et al., 1996; Zhang et al., 1999). In this study we confirmed morphine is not able to mobilize β-arrestin2 to the cell surface, and promote receptor internalization and down-regulation. In contrast, we found that morphine treatment leads to phosphorylation of S363 (53% of the maximal 115

phosphorylation by deltorphin II, Table 8.1) (Navratilova et al., 2005). This finding is the

first to demonstrate that morphine effectively activates the initial step of the regulatory

pathway, phosphorylation of S363. Two possibilities exist to explain why morphine is not

able to activate the later regulatory events: 1) morphine may not be able to promote

phosphorylation of other residues within the C-terminus, which activate the

phosphorylation sensor in the two-step β-arrestin binding (Gurevich and Gurevich,

2006); 2) morphine may stabilize an active receptor conformation distinct from the

conformation imposed by full agonists, which is not recognized by the activation sensor

of β-arrestins. We conclude that even though some features of active receptor

conformation are preserved in morphine-bound state, other features are not. Therefore,

morphine-bound hDOR is not able to activate both sensors on the β-arrestin molecule.

Due to the requirement for simultaneous activation of both sensors, β-arrestin binding

and all regulatory events beyond β-arrestin binding are not supported by morphine.

Our studies using different mutated and truncated receptor constructs identified some

molecular domains involved in hDOR regulation (Table 8.2). The C-terminus was found absolutely necessary for peptide agonist-mediated receptor phosphorylation (Varga et al.,

2004), recruitment of β-arrestin and receptor down-regulation. In contrast, SNC80 is able

to activate other intracellular domain(s) to support receptor regulation, as demonstrated

by the ability of SNC80 to partially phosphorylate the truncET-hDOR (Varga et al.,

2004), promote β-arrestin translocation to the truncated receptor, and cause moderate receptor down-regulation. In search for this additional regulatory domain, we tested 116

whether S249/S255, a putative phosphorylation site and a β-arrestin binding site (Cen et al.,

2001a), may be involved. However, the mutation of these two serines to alanines had no

effect on the higher efficacy of SNC80 compared to peptide agonists to down-regulate

the receptor. Therefore, we hypothesize that other S/T residues in the intracellular loops

of the hDOR function as the secondary regulatory motif activated by SNC80.

To identify the motif in the C-terminus of the hDOR responsible for receptor

phosphorylation and regulation by peptide agonists and SNC80, we focused on the

primary phosphorylation site, S363 (El Kouhen et al., 2001). Our results indicate that the

mutation of S363 to alanine significantly, but not completely, attenuates peptide agonist-

mediated down-regulation and desensitization. Conversely, SNC80-mediated down-

regulation and desensitization were not affected by this mutation, indicating that a

putative secondary regulatory motif activated by SNC80 is able to compensate for the

lack of the S363 motif. Surprisingly, S363A mutation had virtually no effect on β-arrestin2

binding and receptor internalization. This is a novel finding in regulation of G protein-

coupled receptors, which indicates that β-arrestin binding and receptor internalization

may be regulated by intracellular domains distinct from those that regulate receptor

desensitization and down-regulation. Moreover, these data demonstrate that β-arrestin

binding and receptor internalization alone are not sufficient to desensitize the hDOR

without phosphorylation of S363. This conclusion contradicts the generally accepted notion that β-arrestin binding to the GPCR serves to uncouple the receptor from G proteins (Barak et al., 2006). 117

In most of our studies we observed noticeable differences between the abilities of

morphine, peptide agonists and SNC80 to stimulate the receptor regulatory mechanisms.

We conclude that due to the requirement of multiple activation events to occur for β-

arrestin binding, morphine is unable to promote hDOR-β-arrestin interaction and,

therefore, all downstream events including receptor internalization and down-regulation.

Conversely, SNC80 in contrast to peptide agonists is able to phosphorylate (Varga et al.,

2004) and activate additional intracellular domains. This additional activation is

responsible for an increased efficacy of SNC80 to down-regulate the hDOR, but may be

redundant in other regulatory events for the wild-type hDOR. However, the effect of the

supplementary SNC80 activation domain is revealed by mutations that eliminate this

redundancy. Therefore, S363A mutation and truncation of the C-terminus of the hDOR

both enhance the differences between SNC80 and peptide agonists in receptor down-

regulation and uncover additional differences in receptor desensitization and β-arrestin

binding (Table 8.2).

In conclusion, we found that structurally diverse delta opioid receptor agonists regulate

the hDOR by different mechanisms. We demonstrate that morphine is able to activate the first step of the regulatory events, phosphorylation of S363, but due to requirements for

simultaneous activation of multiple sites, morphine fails to promote β-arrestin binding,

receptor internalization and down-regulation. We also report that peptide DOR agonists

and a non-peptide SNC80 differ in their ability to down-regulate the hDOR. Further 118 differences in receptor phosphorylation, desensitization and β-arrestin translocation between these two classes of full delta opioid agonists are uncovered by truncation of the receptor’s C-terminus or by mutation of the primary phosphorylation site, S363. Studies using the mutant receptors identify the C-terminus as the important domain for hDOR phosphorylation, β-arrestin binding and down-regulation by both peptide and non-peptide agonists. S363 within the C-terminus is critically involved in receptor phosphorylation, desensitization and down-regulation by peptide agonists, but not in β-arrestin binding and receptor internalization. In contrast to peptide agonists, SNC80 is able to phosphorylate and activate secondary intracellular domain(s), in addition to the C-terminus, which participate in β-arrestin recruitment and receptor desensitization and down-regulation.

Therefore, agonist-specific differences were detected for multiple regulatory events between morphine, peptide agonists and SNC80. We propose that the differential agonist- mediated regulation of the hDOR may be utilized to design opioid drugs with improved analgesic properties and minimal side effects. In our future studies we first want to confirm that similar agonist-specific differences are also observed in rat dorsal root ganglion neurons. These primary sensory neurons are responsible for detection of a pain stimulus and its transmittion to the spinal cord. We hypothesize that agonist-mediated regulation of the DOR in DRG neurons may affect the development of tolerance.

Therefore, next we will investigate whether agonist-specific differences in DOR regulation will affect the development of opioid tolerance in vivo. 119

TABLES

35 Table 7.1. Emax and EC50 values of deltorphin II-mediated stimulation of GTPγ[ S] binding and inhibition of cAMP production in hDOR/CHO cells pretreated for indicated times with deltorphin II.

Time (min) 0 2 5 10 30 45 60

Emax 100 105 113 102 81 58 - (GTPγ[35S]) ±1 ±2 ±1 ±2 ±2 ±4

EC50 6.3 13 16 18 24 23 - (GTPγ[35S]) ±0.4 ±1 ±1 ±2 ±3 ±7

Emax 100 102 108 103 102 99 82

(cAMP) ±3 ±7 ±5 ±3 ±4 ±6 ±6

EC50 0.36 0.33 1.0 0.98 1.8 3.1 1.7

(cAMP) ±0.06 ±0.15 ±0.2 ±0.18 ±0.4 ±1.0 ±0.6

Deltorphin II concentration responses for GTPγ[35S] stimulation and cAMP inhibition in hDOR/CHO cells pretreated for different time intervals with deltorphin II were fitted with sigmoidal curves (Hill coefficient nH = 1). Calculated EC50 ± SE are presented in nM concentrations, Emax ± SE are in %. 120

Table 7.2. Parameters of deltorphin II-mediated desensitization of hDOR calculated by equation (7.5).

-1 Em (%) τcontrol τ∞ k (min ) KA (nM)

120 7.8 1.4 0.070 81 GTPγ[35S] ±5 ±2.4 ±0.5 ±0.018 ±16

106 36 6.4 0.082 16 cAMP ±4 ±21 ±3.4 0.034 ±8

Five parameters of equation (7.5) (Em = operational maximum, τcontrol = (τ0 + τ∞), and τ∞ the operational efficacies before and after desensitization, k = desensitization rate constant, KA = dissociation constant) were calculated by fitting the dose response curves of GTPγ[35S] or cAMP assays. 121

35 Table 7.3. EC50 values of deltorphin II-mediated stimulation of GTPγ[ S] binding and inhibition of cAMP production in naïve and deltorphin II pretreated hDOR/CHO or hDOR(S363A)/CHO cells.

hDOR(control) hDOR(deltorphin) S363A(control) S363A(deltorphin)

Emax 99 77 100 110

(GTPγ[35S]) ±2 ±5 ±2 ±3

EC50 6.9 31 5.8 13

(GTPγ[35S]) ±0.9 ±7 ±0.7 ±2

Emax 99 83 99 108

(cAMP) ±6 ±9 ±5 ±5

EC50 1.4 7.5 2.0 2.8

(cAMP) ±0.5 ±4.0 ±0.7 ±0.8

Deltorphin II concentration responses for GTPγ[35S] stimulation and cAMP inhibition in naïve and deltorphin II pretreated hDOR/CHO or hDOR(S363A)/CHO cells were fitted with sigmoidal curves (Hill coefficient nH = 1). Calculated EC50 ± SE are presented in nM concentrations, Emax ± SE are in %. 122

Table 7.4. Parameters of deltorphin II-mediated desensitization of hDOR and hDOR(S363A) calculated using the operational model (equation (7.1)).

Em (%) τcontrol τdel-wt τdel-S363A KA (nM)

110 13 1.9 10 97 GTPγ[35S] ±4 ±5 ±0.5 ±4 ±33

107 19 3.6 16 36 cAMP ±7 ±18 ±2.3 ±15 ±30

Parameters of the operational model (equation (7.1)) (Em = operational maximum; τcontrol ,

τdel-wt , τdel-S363A are operational efficacies for untreated cells and deltorphin II pretreated hDOR/CHO or hDOR(S363A)/CHO cells, respectively; KA = dissociation constant) were calculated by fitting the dose response curves of stimulation of GTPγ[35S] binding or inhibition of cAMP production. 123

hDOR hDOR hDOR S363A S363A S363A control deltorphin SNC80 control deltorphin SNC80

Emax 99 83 70 100 100 78

(GTPγ[35S]) ±2 ±5 ±10 ±2 ±3 ±7

EC50 6.8 31 23 5.7 9.5 25

(GTPγ[35S]) ±0.3 ±3 ±6 ±0.2 ±0.7 ±4

Emax 100 81 72 97 100 99

(cAMP) ±2 ±5 ±7 ±4 ±3 ±4

EC50 1.1 3.3 9.1 1.7 2.5 4.2

(cAMP) ±0.1 ±0.5 ±1.9 ±0.2 ±0.2 ±0.4

Deltorphin II concentration responses for GTPγ[35S] stimulation and cAMP inhibition in naïve and deltorphin II or SNC80 pretreated hDOR/CHO or hDOR(S363A)/CHO cells were fitted with sigmoidal curves (Hill coefficient nH = 1). Calculated EC50 ± SE are presented in nM concentrations, Emax ± SE are in %. 124

Table 7.6. Parameters of deltorphin II-mediated desensitization of hDOR and

hDOR(S363A) calculated using the operational model (equation (7.1)).

Em (%) τcontrol τdel τSNC80 KA (nM)

GTPγ[35S]

hDOR 105±6 16 3.3 2.5 116±22

hDOR(S363A) 112±6 15 11 2.7 112±26

cAMP

hDOR 105±6 18 3.9 1.9 20±4

hDOR(S363A) 106±9 30 25 15 68±53

Parameters of the operational model (equation (7.1)) (Em = operational maximum; τcontrol ,

τdet , τSNC80 are operational efficacies for untreated cells and deltorphin II or SNC80 pretreated hDOR/CHO or hDOR(S363A)/CHO cells, respectively; KA = dissociation

constant) were calculated by fitting the dose response curves of stimulation of GTPγ[35S] binding or inhibition of cAMP production. 125

Table 8.1. Maximal effects produced by activation of the wild-type hDOR by morphine, peptide agonists or SNC80.

Down- GTPγS cAMP p-S363 β-arrestin Internalization regulation (%) (%) (%) (%) Morphine 80 89 53 0 0 0

Peptide 100 100 100 ++ ++ 80

SNC80 100 100 100 ++ ++ 90

Summary of the effects of morphine, peptide agonists (DPDPE, deltorphin II), or SNC80 treatment on signaling and regulation of the wild-type hDOR. Maximum agonist- mediated GTPγS stimulation and inhibition of cAMP caused by saturating concentrations of the agonist are expressed as % of maximal effect produced by SNC80. Receptor phosphorylation at S363 is expressed as % maximal phosphorylation produced by SNC80.

Qualitative evaluation of β-arrestin2 recruitment and receptor internalization is indicated

by ++ and 0 if the effect is strong or not detectable. Down-regulation is expressed as % of

total receptor number down-regulated after 24h treatment with saturating concentrations

of the agonist. 126

Table 8.2. Maximal effects produced by activation of the wild-type hDOR, the hDOR(S363A) or the truncET-hDOR by peptide agonists or SNC80.

Phosphorylation Desensitization Down- β-arrestin (% max) (%) regulation (%) Peptide SNC80 Peptide SNC80 Peptide SNC80 Peptide SNC80

Wt hDOR 100 100 ++ ++ 81 89 80 90 hDOR(S363A) ++ ++ 18 65 45 87

Trunc-hDOR 0 19 0 + 6 47

Summary of the differential effects of peptide agonists (DPDPE, deltorphin II), or SNC80

treatment on regulation of the wild-type hDOR, the hDOR(S363A) mutant , and the

truncET-hDOR. Total receptor phosphorylation as reported in (Varga et al., 2004) is expressed as % maximal phosphorylation produced by SNC80. Qualitative evaluation of

β-arrestin2 recruitment is indicated by ++, + and 0 if the effect is strong, weak, or not

detectable, respectively. Desensitization represents the percent of desensitized receptors

by 60 min of agonist treatment calculated by averaging the results of GTPγS and cAMP

assays. Down-regulation is expressed as % of total receptor number down-regulated after

24h treatment with saturating concentrations of the agonist.

127

FIGURES Fig. 1.1 Agonist-mediated hDOR regulation

P P 3 β γ GRK 2 β-arrestin Giα AP2 Phosphorylation P P recruitment Recycling β-arrestin 1 4

Second messenger pathways

5 Lysosome Effect P P Internalization DOR Down-Regulation DOR agonist

Fig. 1.1. Agonist-mediated regulation of the human delta opioid receptor. Schematic diagram of regulatory events occurring after activation of the hDOR with an agonist: 1) agonist binding to the hDOR leads to activation of multiple second messenger signaling pathways producing cellular effects. 2) The activated receptor is then phosphorylated by

G protein-coupled receptor kinases (GRKs) at several residues. 3) The agonist-bound and phosphorylated receptor displays high affinity for the cytosolic adaptor protein, β- arrestin, which targets the receptor to clathrin-coated pits. 4) β-arrestin also assists in receptor internalization via the endocytosis of the clathrin-coated vesicles. 5) The internalization process may serve either to degrade the receptor in cellular digestive organelles, thereby terminating receptor function (down-regulation), or to reintroduce the receptor to the plasma membrane (recycling). 128

Fig. 2.1

AB OH HN+ CH3 + O H3N D-Pen S C N S H3C C C D-Pen Gly NH C2H5 H N C C2H5 O MeO O HO

CH CD3 NH+

Tyr-D-Ala-Phe-Glu-Val-Val-Gly-NH2

HO O OH

Fig. 2.1. Molecular structures of the delta opioid receptor agonists. Molecular structures of three classes of the DOR agonists used in this study are given: A) cyclic[D-

Pen2, D-Pen5]enkephalin (DPDPE); B) (+)-4-[(αR)-α-((2S,5R)-4-allyl-2,5-dimethyl-1- piperazinyl)-3-methoxybenzyl]N,N-diethylbenzamide (SNC80); C) (D-Ala2)deltorphin II

(deltorphin II); D) morphine. 129

Fig. 2.2

-NH A S P Y A D S A N A F L P P Q L E A G A S P A P E M 2 S P F G A A N A S G P P G A R S R A R D P P Q F S D S L P M I L S G E R S L P F W L C D P W D V A T L L V G Y V E V R V A M C A T W L A I V D L K A T W T A A A M V L Y V V V I H I K L M T V L A T A S S I K I F C L F Q I P C I I Y P D V V A S Y G F H L A V T L Y V L I G C G F P Y A S T M N S A A A V A A F W N G L T F L V C S L A S V V V S L G D I W P V L N A F I I F N V L T C L A P L V L A L I I G V N T N I V V L M F M I L T V Y F G I M K V L A Y S A C Y V I I V K G M L F V A L R N R D T D C P M I R R T Y R E L K A I T L * Q *

Y * R N T L R R P T M K A R L F K V F

F L * S K * C C C

* R D R H R D 255 G L K 338 P * S * A * E R V L V K L S K * R G S * P * F S S P D 249 S R * A R E A T A R E * * * R V HO- A A A G G G P G D S P T C A T 363

Fig. 2.2. Molecular structure of the human delta opioid receptor. The human DOR is a member of G protein-coupled receptor (GPCR) family, which is characterized by a conserved seven transmembrane structure with an extracellular N-terminal domain and an intracellular C-terminal domain. Amino-acid composition of the hDOR and location of extracellular, transmembrane, and intracellular domains is depicted. Possible intracellular phosphorylation serine and threonine sites are indicated by the asterixes.

Positions of mutated amino acids are indicated by the numbers. Truncation of the C- terminus after G338 is indicated by the arrow. 130

Fig. 3.1 Wild type hDOR hDOR (S363A) MW (kDa) Anti-hDOR(S363) 80 60 50

Anti-actin 40 Morphine -+-+-+- Deltorphin II - - +-+-+ Naltrindole - --++--

Fig. 3.1. Agonist-mediated phosphorylation of the human δ-opioid receptor at S363. Recombinant CHO cells expressing the hDOR or the S363A mutant hDOR were treated in serum free IMDM medium with opioid ligands for 30 min at 37°C and Western blot samples were prepared as described. Aliquots from each sample, corresponding to 5 μg total protein, were loaded onto 10% polyacrylamide gels, resolved and transferred to nitrocellulose membranes. Western blot analyses were performed using an anti-phospho- S363 antibody (Cell Signaling). Treatment of cells for 30 min with morphine (30 μM) or deltorphin II (100 nM) led to phosphorylation of the hDOR at S363 (lanes 2, 3). This phosphorylation was completely blocked by concomitant incubation of cells with 1μM naltrindole (lanes 4, 5). No immunoreactivity was detected in cells expressing the S363A mutant receptor after identical treatment with morphine or deltorphin II (lanes 6, 7). Intensities of the 50-70 kDa protein bands were quantified using the Image J software (NIH). For each sample the intensity of phospho-S363 bands was normalized to the intensity of actin immunoreactivity. The results were analyzed using GraphPad Prizm 4 and calculated as percent of the maximal phosphorylation produced by deltorphin II. In four independent experiments, morphine produced 53 ± 8% (mean ± S.E.M.) of the maximal deltorphin II-mediated phosphorylation, which was significantly greater than basal phosphorylation (5.4 ± 1.2%, p < 0.05). 131

Fig. 3.2 AB

0 0.3 1 3 10 30 100 0 0.3 1 3 10 30 100 [Morphine], (μM) [Deltorphin II], (nM) CD 120 120

100 100

80 80

60 60

40 40 Phosphorylation Phosphorylation 20 20 (%maximum phosphorylation) (%maximum (%maximum phosphorylation) 0 0 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -10.0 -9.5 -9.0 -8.5 -8.0 -7.5 -7.0 Log [Morphine], (M) Log [Deltorphin II], (M)

Fig. 3.2. Dose-response of morphine- and deltorphin II-mediated phosphorylation of the hDOR at S363. Recombinant CHO cells expressing the hDOR were treated in serum free IMDM medium with increasing concentrations of morphine or deltorphin II for 30 min at 37°C and Western blot samples were prepared as described. Aliquots from each sample were loaded onto 10% NuPAGE gels, resolved and transferred to nitrocellulose membranes. Western blot analyses were performed using an anti-phospho-S363 antibody. Treatment of cells for 30 min with morphine or deltorphin II led to a dose-dependent phosphorylation of the hDOR at S363. Representative immunoblots for morphine and deltorphin II are shown in (A) and (B), respectively. Intensities of the 50-70 kDa protein bands were quantified using the Image J software (NIH), the results were plotted as percent of the maximal phosphorylation produced by morphine (C) or deltorphin II (D) (mean ± S.E.M., n=2) and fitted with sigmoidal dose response curves with variable Hill coefficients. The EC50 values determined by regression analysis are 1.4±0.1 μM and 1.3±0.1 nM for morphine and deltorphin II, respectively. 132

Fig. 3.3 AB Morphine (30 μM) Deltorphin II (100 nM)

0 2 10 30 60 120 240 0 2 10 30 60 120 240 Time (min) Time (min)

C 120 100

40 Deltorphin II

Phosphorylation 20 Morphine (% phosphorylation @ 30 min) 30 @ phosphorylation (% 0 0 5 10 15 20 25 30 Time (min)

Fig. 3.3. Time course of morphine- and deltorphin II-mediated phosphorylation of the hDOR at S363. Recombinant CHO cells expressing the hDOR were treated in serum free IMDM medium with morphine (30 μM) or deltorphin II (100 nM) for increasing time intervals (0-240 min) at 37°C and Western blot samples were prepared as described. Aliquots from each sample were loaded onto 10% NuPAGE gels, resolved and transferred to nitrocellulose membranes. Western blot analyses were performed using an anti-phospho-S363 antibody. Treatment of cells with morphine or deltorphin II led to a time-dependent phosphorylation of the hDOR S363. Representative immunoblots for morphine and deltorphin II are shown in (A) and (B), respectively. Intensities of the 50- 70 kDa protein bands were quantified using the Image J software (NIH), the results were plotted as percent of the maximal phosphorylation produced by each drug after 30 min of treatment (mean ± S.E.M., n=3) and early time points (0-30 min) fitted with a single exponential association curve (C). The half-lives of hDOR(S363) phosphorylation determined by regression analysis are 4.2±0.8 min and 1.7±0.3 min, respectively. Statistical analysis demonstrates that the two curves are significantly different (p=0.003). 133

Fig. 3.4 ABMorphine (30 μM) Deltorphin II (100 nM)

0 10 30 60 120 240 0 10 30 60 120 240 Time (min) Time (min)

C 120 Deltorphin II 100 Morphine 80

Phosphorylation 20 (%maximum phosphorylation) (%maximum 0 0 30 60 90 120 150 180 210 240 Time (min) Fig. 3.4. Time course of hDOR(S363) dephosphorylation after morphine or deltorphin II treatment. Recombinant CHO cells expressing the hDOR were treated in serum free IMDM medium with morphine (30 μM) or deltorphin II (100 nM) for 30 min at 37°C, then the agonists were washed and the cells were left in fresh IMDM medium for indicated time intervals before harvesting. Western blot samples were prepared as described, aliquots from each sample were loaded onto 10% NuPAGE gels, resolved and transferred to nitrocellulose membranes. Western blot analyses were performed using an anti-phospho-S363 antibody. Removal of morphine or deltorphin II led to a time- dependent dephosphorylation of the hDOR. Representative immunoblots for morphine and deltorphin II are shown in (A) and (B), respectively. Intensities of the 50-70 kDa protein bands were quantified using the Image J software (NIH), the results were plotted as percent of the initial phosphorylation before removal of the agonist (mean ± S.E.M., n=4) and fitted with a single exponential dissociation curve (C). The half-lives of dephosphorylation determined by regression analysis are 19±6 min and 64±22 min for morphine and deltorphin II pretreated cells, respectively. Comparison of the two regression curves using the F-test shows that the rates of dephosphorylation after morphine and deltorphin II pretreatment are significantly different (p=0.003). 134

Fig. 4.1 ACB

D E

Fig. 4.1. Deltorphin II, DPDPE and SNC80 but not morphine stimulate translocation of β-arrestin2-GFP in CHO cells expressing the hDOR. CHO cells expressing the wild type hDOR were transfected with a green fluorescent protein tagged

β-arrestin2 construct (β-arrestin2-GFP, a gift from Dr. Lefkowitz). 48 h after transfection the cells were treated for 5 min at 37°C with vehicle (A), 500 nM deltorphin II (B), 500 nM DPDPE (C), 500 nM SNC80 (D) or 10 μM morphine (E). After agonist treatment, the cells were fixed with 4% paraformaldehyde and mounted in a Vectashield Mounting

Medium. Slides were examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective using a 488/505 nm excitation/emission filter set. Single optical sections were acquired through the trans- nuclear plane. The magnification bar represents 10 μm. Images were processed in Adobe

Photoshop 7.0. 135

Fig. 4.2

A B

C D

Fig. 4.2. β-arrestin2-GFP does not internalize after agonist treatment into intracellular compartments. hDOR/CHO cells transfected with a green fluorescent protein tagged β-arrestin2 construct were treated with 500 nM SNC80 at 37°C for 0 (A),

5 (B), 10 (C) and 30 (D) min time intervals as indicated. After agonist treatment, the cells were fixed and localization of β-arrestin2-GFP was examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective. Single optical sections were acquired through the trans-nuclear plane. The magnification bar represents 10 μm. Images were processed in Adobe Photoshop 7.0. 136

Fig. 4.3

ACB

basal trans-nuclear apical

Fig. 4.3. After agonist treatment β-arrestin2-GFP associates with endosomes only at or near the plasma membrane. hDOR/CHO cells transfected with a green fluorescent protein tagged β-arrestin2 construct were treated with 500 nM SNC80 for 5 min at 37°C.

After the treatment, the cells were fixed and examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective. Optical sections were acquired through the entire height of the cell. Optical sections taken through the basal membrane (A), trans-nuclear plane (B) and apical membrane (C) of the same cell are shown. The magnification bar represents 10 μm. Images were processed in Adobe

Photoshop 7.0. 137

Fig. 4.4

ACB

D E

Fig. 4.4. Deltorphin II, DPDPE, SNC80 but not morphine stimulate translocation of

β-arrestin2-GFP to the hDOR(S363A). CHO cells expressing the S363A mutant hDOR were transfected with a green fluorescent protein tagged β-arrestin2 construct. 48 h after transfection the cells were treated for 5 min at 37°C with vehicle (A), 500 nM deltorphin

II (B), 500 nM DPDPE (C), 500 nM SNC80 (D) or 10 μM morphine (E). After agonist treatment, the cells were fixed and examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective. Single optical sections were acquired through the trans-nuclear plane. The magnification bar represents 10 μm.

Images were processed in Adobe Photoshop 7.0. 138

Fig. 4.5

ACB

Fig. 4.5. SNC80 but not deltorphin II promotes translocation of β-arrestin2-GFP to the truncated hDOR. CHO cells expressing the truncated hDOR (truncEt-hDOR/CHO) were transfected with a green fluorescent protein tagged β-arrestin2 construct. 48 h after transfection the cells were treated for 5 min at 37°C with vehicle (A), 500 nM deltorphin

II (B) or 500 nM SNC80 (C). After agonist treatment, the cells were fixed and examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective. Single optical sections were acquired through the trans-nuclear plane. The magnification bar represents 10 μm. Images were processed in Adobe Photoshop 7.0. 139

Fig. 5.1

A B

C D

E F

Fig. 5.1. Fluo-deltorphin is internalized in CHO cells expressing the hDOR in a time-dependent manner. hDOR/CHO cells were incubated with fluo-deltorphin (300 nM) for 3 (A), 10 (B), 30 (C, E) or 60 (D,F) min at 37°C. In (E) selective DOR antagonist naltrindole (10 μM) was added together with the agonist. Panel (F) shows an individual cell at a higher magnification incubated with fluo-deltorphin for 60 min. After treatment, the cells were fixed and examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective. Single optical sections were acquired through the trans-nuclear plane. The acquisition parameters were constant in all parallel experiments. Images were processed in Adobe Photoshop 7.0. 140

Fig. 5.2

A B

C D

Fig. 5.2. hDOR is internalized via a subset of clathrin-coated pits distinct from transferrin-containing CCPs. hDOR/CHO cells were serum starved for 1h and pretreated with IMDM (A), 0.5M sucrose (B), 2% 2-hydroxypropyl-β-cyclodextrin (C) or 2

μg/ml fillipin (D) for an additional 1h. All cells were treated at 37°C simultaneously with both [Gln4]deltorphin-rhodamine (1μM) and transferrin-AlexaFluor488 (250 μg/ml). 30 min after the treatment, the cells were rinsed with ice cold PBS, fixed with 4% paraformaldehyde and mounted in a Vectashield Mounting Medium. Slides were examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a

100×/1.40 oil objective using excitation/emission filter sets 488/505 nm and 543/560 nm.

Single optical sections were acquired through the trans-nuclear plane. The acquisition parameters were constant in all parallel experiments. Images were processed in Adobe

Photoshop 7.0. The magnification bar represents 10 μm. 141

Fig. 5.3

A B

C D

Fig. 5.3. hDOR-EGFP is internalized in CHO cells upon deltorphin II and SNC80 treatment. hDOR-EGFP/CHO#2 cells were incubated with IMDM medium (A), 500 nM deltorphin II (B), 500 nM SNC80 (C) or 10 μM morphine (D) for 30 min at 37°C. After treatment, the cells were fixed and examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a 100×/1.40 oil objective. Single optical sections were acquired through the trans-nuclear plane. The acquisition parameters were constant in all parallel experiments. Images were processed in Adobe Photoshop 7.0. The magnification bar represents 10 μm. 142

Fig. 5.4

Fig. 5.4. [Gln4]deltorphin-rhodamine is internalized in CHO cells expressing either the wild-type or S363A mutant hDOR. CHO cells expressing the wild-type hDOR (A) or hDOR(S363A) mutant (B) were grown on glass bottom chamber slides, incubated in serum-free IMDM medium with 1 μM [Gln4]deltorphin-rhodamine for 30 min at 37°C.

After treatment, the cells were rinsed with ice-cold PBS, fixed with 4% paraformaldehyde and mounted in a Vectashield Mounting Medium. Slides were examined under a Zeiss LSM520 laser scanning confocal microscope equipped with a

100×/1.40 oil objective using a 543/560 nm excitation/emission filter set. Single optical sections were acquired through the trans-nuclear plane. The acquisition parameters were constant in all parallel experiments. The images were processed using the Adobe

Photoshop 7.0 software. The magnification bar represents 10 μm. 143

Fig. 6.1 AB

120 120

100 100

80 80 H]NTI bound H]NTI bound

3 60 3 60

(% control) 40 (% control) 40

20 20

speciffic [ 0 speciffic [ 0 -9 -8 -7 -6 2 4 6 12 18 24 Log [SNC80], (M) Time (h)

Fig. 6.1. SNC80 down-regulates the hDOR in a concentration and time dependent manner. CHO cells stably expressing the human δ-opioid receptors were treated for 24h with 1-500 nM SNC80 (A) or for 2-24h with 500 nM SNC80 (B). After three washes, cell membranes were prepared and [3H] naltrindole (0.5 nM) binding was measured.

Total number of receptors remaining after the agonist treatment was determined as specific binding per 1 μg membrane protein and expressed as a percentage of the buffer- treated control. Data shown represent means ± S.E.M. (A) Experimental data were fitted with a sigmoidal dose response curve using the Hill coefficient nH = 1. The curve fitting provides the EC50 = 21±8 nM. (B) Experimental data were fitted with a one phase exponential decay curve. The fitted value of the half life of hDOR down-regulation t½ =

2.3±0.4h and the maximum desensitization reached 12±5 %. 144

Fig. 6.2

103% 100 0 Down-regulation

80 20 control) (% 60 40 H]NTI Bound 3 ∗ 40 60 (% control) 20% 20 10% 80 Specific [ 0 100 DPDPE SNC80 Morphine (500 nM) (500 nM) (5 μM)

Fig. 6.2. The human δ-opioid receptor down-regulation is agonist specific. CHO cells stably expressing the hDOR were treated for 24h with 500 nM DPDPE, 500 nM SNC80 or 5 μM morphine. Total number of receptors remaining after the agonist treatment was determined by specific [3H] naltrindole binding per 1 μg membrane protein and expressed as a percentage of the buffer-treated control. Data shown represent means ±

S.E.M. The hDOR is more effectively down-regulated by SNC80 (90±1 %, n=5) than by

DPDPE (80±1 %, n=5). The difference is statistically significant * p<0.001. No hDOR down-regulation was observed upon morphine (5 μM) treatment. 145

Fig. 6.3

- Sucrose + 0.5 M Sucrose 100 0 Down-regulation 80 20 (% control) (%

60 40 H]NTI Bound H]NTI 3 40 60 (% control)

20 80 Speciffic [ 0 100 control DPDPE SNC80 2h treatment

Fig. 6.3. Hypertonic sucrose prevents agonist-mediated hDOR down-regulation.

CHO cells stably expressing the human δ-opioid receptors were treated for 2h with 500 nM DPDPE or 500 nM SNC80 in the absence or presence of 0.5 M sucrose. Total number of receptors remaining after the agonist treatment was determined by specific

[3H] naltrindole binding per 1 μg membrane protein and expressed as a percentage of the buffer-treated control. Data shown represent means ± S.E.M. The results demonstrate that hypertonic sucrose reduces hDOR down-regulation after 2h of agonist treatment from 39±1 % to 10±1 %, n=2 for DPDPE and from 51±6 % to 16±1 %, n=2 for SNC80. 146

Fig. 6.4

300 control 250 Lactacystin

200

150 H]NTI binding H]NTI 3

(%control) 100

50 speciffic [ 0 control SNC80 DPDPE

Fig. 6.4. Lactacystin increases the number of human δ-opioid receptors in agonist- treated and untreated cells. CHO cells stably expressing the human δ-opioid receptors were treated for 24h with 500 nM DPDPE or 500 nM SNC80 in the absence or presence of 25 μM lactacystin. Total number of receptors remaining after the agonist treatment was determined by specific [3H] naltrindole binding per 1 μg membrane protein and expressed as percent in untreated cells. Data shown represent means ± S.E.M of three independent experiments. In lactacystin untreated cells DPDPE and SNC80 treatment led to a reduction of total number of receptors by 84±2 % and 92±2 %, respectively.

Lactacystin treatment seems to prevent this reduction and even increases the number of hDORs by 21±18 % and 6±25 % after DPDPE and SNC80 treatment, respectively.

However, lactacystin also increases the basal levels of hDOR by 110±49 %. 147

Fig. 6.5

120 -20

100 0 Down-regulation

80 control 20 (% control) Chx 60 40

H]NTI binding Chx+Lactacystin 3

(%control) 40 60

20 80 speciffic [ 0 100 control SNC80 DPDPE

Fig. 6.5. Lactacystin has no effect on DPDPE and SNC80 mediated down-regulation of human δ-opioid receptors when new protein synthesis is blocked by cycloheximide. CHO cells stably expressing the hDOR were treated for 24h with 500 nM DPDPE or 500 nM SNC80 in the presence of 5 μg/ml cycloheximide alone (Chx) or

5 μg/ml cycloheximide plus 25 μM lactacystin (Chx+Lactacystin). Total number of receptors remaining after the agonist treatment was determined by specific [3H] naltrindole binding per 1 μg membrane protein and expressed as percent in untreated cells. Data shown represent means ± S.E.M of three independent experiments.

Cycloheximide alone causes a decrease in the total number of receptors by 26±16 % compared to untreated cells. Cycloheximide plus lactacystin show almost identical reduction (24±16 %). In cycloheximide treated cells DPDPE and SNC80 treatment leads to hDOR down-regulation by 90±2 % and 92±1 %, respectively. In cycloheximide plus lactacystin treated cells, DPDPE and SNC80 treatment causes similar hDOR down- regulation (87±2 % and 91±1 % after DPDPE and SNC80 treatment, respectively). 148

Fig. 6.6

140 control -40 120 MG132 -20 Down-regulation

100 0 control) (% 80 20 H]NTI binding

3 60 40 (% control) 40 60 20 80 speciffic [ 0 100 control DPDPE

Fig. 6.6. MG132 has no effect on DPDPE mediated down-regulation of human δ- opioid receptors after 4h of treatment. CHO cells stably expressing the human δ- opioid receptors were treated for 4h with 500 nM DPDPE in the absence or presence of

25 μM MG132. Total number of receptors remaining after the agonist treatment was determined by specific [3H] naltrindole binding per 1 μg membrane protein and expressed as percent in untreated cells. Data shown represent means ± S.E.M of two independent experiments. MG132 alone causes an insignificant increase in the total number of receptors by 12±15 % compared to untreated cells. 4h treatment with DPDPE leads to down-regulation of 54±6 % of receptors in the absence and 50±3 % of receptors in the presence of MG132, respectively. 149

Fig. 6.7

140 -40 ts20-33°C 120 -20 ts20-41°C Down-regulation

100 0 (% control) 80 20 H]NTI binding

3 60 40 (% control) 40 60 20 80 speciffic [ 0 100 control DPDPE SNC80

Fig. 6.7. Inactivation of the ubiquitin conjugating enzyme E1 in E36ts20 cells has no effect on DPDPE and SNC80 mediated down-regulation of the hDOR. E36ts20 cells were transiently transfected with the human δ-opioid receptors. 48h post transfection, the cells were treated for 4h with 500 nM DPDPE or 500 nM SNC80 at a permissive (33ºC) or a non-permissive (41ºC) temperature. Total number of receptors remaining after the agonist treatment was determined by specific [3H] naltrindole binding per 1 μg membrane protein and expressed as percent in untreated cells maintained at 33ºC. Data shown represent means ± S.E.M. At 41ºC the total number of hDORs is increased, compared to the total number of receptors expressed at 33ºC by 12±15 % in untreated cells. 4h DPDPE treatment leads to down-regulation of 39±9 % of receptors at the permissive and 46±3 % of receptors at the non-permissive temperature, respectively

(n=2). 4h SNC80 treatment leads to down-regulation of 41 % of receptors at the permissive and 43 % of receptors at the non-permissive temperature, respectively (n=1). 150

Fig. 6.8

control 150 Chx -50 Chx+Foli Down-regulation

100 0 (% control) H]NTI binding 3

(% control) (% 50 50 Speciffic [ 0 100 control DPDPE

Fig. 6.8. Folimycin attenuates DPDPE mediated down-regulation of human δ-opioid receptors. CHO cells stably expressing the hDORs were treated for 24h with 500 nM

DPDPE in the absence or presence of 5 μg/ml cycloheximide alone (Chx) or 5 μg/ml cycloheximide plus 25 μM folimycin (Chx+Foli). Total number of receptors remaining after the agonist treatment was determined by specific [3H] naltrindole binding per 1 μg membrane protein and expressed as percent in untreated cells. Data shown represent means ± S.E.M of three independent experiments. Cycloheximide alone causes a small decrease in the total number of receptors by 26±16 % while cycloheximide plus folimycin causes a small increase in the total number of receptors by 26±17 % compared to untreated cells. 24h treatment with DPDPE leads to down-regulation of 90±2 % of receptors in the presence of cycloheximide and only 11±10 % of receptors in the presence of both cycloheximide plus folimycin. 151

Fig. 6.9

100 94% 0

DPDPE Down-regulation 80 20 SNC80 (% control) 60 53% 40 H]NTI bound 3 40 60 (% control) (% 20% 20 10% 80 Specific [ 0 100 hDOR trunc-hDOR

Fig. 6.9. Truncation of the C terminus of the human δ-opioid receptor specifically eliminates down-regulation by DPDPE but not by SNC80. CHO cells stably expressing the wild-type hDOR or truncEt-hDOR were treated for 24h with 500 nM

DPDPE or 500 nM SNC80. After three washes, cell membranes were prepared and [3H] naltrindole (0.5 nM) binding was measured. Down-regulation was defined as a decrease in specific binding per 1 μg membrane protein after the agonist treatment and expressed as a percentage of the buffer-treated control. Data shown represent means ± S.E.M. of four independent experiments. Down-regulation by DPDPE in the truncEt-hDOR is almost eliminated (6 %) as compared to the hDOR (80±1 %). Down-regulation by

SNC80 in the truncEt-hDOR is significantly attenuated (47±6 %) as compared to the hDOR (90±1 %). 152

Fig. 6.10

100 0 DPDPE Down-regulation 80 SNC80 20 55% control) (% 60 40 H]NTI bound 3 40 60 (% control) 20% 13% 20 10 % 80 Specific [ 0 100 hDOR hDOR(S363A)

Fig. 6.10. Mutation of S363 to Ala in the human δ-opioid receptor specifically attenuates down-regulation by DPDPE. CHO cells stably expressing the wild-type hDOR or hDOR(S363A) mutant were treated for 24h with 500 nM DPDPE or 500 nM

SNC80. After three washes, cell membranes were prepared and [3H] naltrindole (0.5 nM) binding was measured. Down-regulation was defined as a decrease in specific binding per 1 μg membrane protein after the agonist treatment and expressed as a percentage of the buffer-treated control. Data shown represent means ± S.E.M. of at least four independent experiments. Down-regulation by DPDPE in the hDOR(S363A) mutant is significantly attenuated (45±9 %) as compared to the wild-type hDOR (80±1 %), p<0.01.

The hDOR is more effectively down-regulated by SNC80 (90±1 %) than by DPDPE

(80±1 %), p<0.001. The hDOR(S363A) mutant is more effectively down-regulated by

SNC80 (87±1 %) than by DPDPE (45±9 %), p<0.01. 153

Fig. 6.11

100 0 DPDPE Down-regulation 80 SNC80 20 (% control)

60 45% 40 43% H]NTI bound 3 40 26% 31% 60 (% control) (%

20 80 Speciffic [ 0 100 S363A S249A/S255A/S363A

Fig. 6.11. Mutation of S249 and S255 to Ala in the human δ-opioid receptor does not affect down-regulation by SNC80. CHO cells were transiently transfected with hDOR(S249A/S255A/S363A) triple mutant or hDOR(S363A) single mutant human δ- opioid receptors. 48h post-transfection the cells were treated for 24h with 500 nM

DPDPE or 500 nM SNC80. After three washes, cell membranes were prepared and [3H] naltrindole (0.5 nM) binding was measured. Down-regulation was defined as a decrease in specific binding per 1 μg membrane protein after the agonist treatment and expressed as a percentage of the buffer-treated control. Data shown represent means ± S.E.M. of at least three independent experiments. Down-regulation by SNC80 is significantly more effective than by DPDPE in the hDOR(S363A) single mutant (74±12 % and 55±11 %, respectively). Same down-regulation efficacies are also observed in the hDOR(S249A/S255A/S363A) triple mutant (69±7 % and 57±4 %, respectively). 154

Fig. 7.1 AB 120 120

100 100

80 80

60 60

GTP stimulation 40

γ 40 cAMP inhibition S]

35 20 20 [ (% max. in control cells) (%max. cells) control in 0 0 -11 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 -6 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

Fig. 7.1. Pretreatment of hDOR/CHO cells with deltorphin II leads to time- dependent desensitization of deltorphin II-stimulated GTPγ[35S] binding and inhibition of cAMP production. hDOR/CHO cells were pretreated for the indicated times with 100 nM deltorphin II in serum free IMDM medium at 37ºC. (A) Cell membranes were prepared and deltorphin II-stimulated GTPγ[35S] binding was measured as described in “Materials and Methods”. The measured deltorphin II concentration curves were normalized to percent of the maximal effect in control (untreated) cells. (B)

After pretreatment, the cells were washed and deltorphin II inhibition of forskolin (100

μM, 20 min)-stimulated cAMP accumulation was measured as described in “Materials and Methods”. Dose-response curves are plotted as percent increase of cAMP inhibition relative to control cells. The data points in (A) and (B) represent the averages and standard errors obtained from three independent experiments performed in duplicates.

Deltorphin II pretreatment: (●) 0 min, (○) 2 min, (■) 5 min, (□) 10 min, (▲) 30 min, (△)

45 min, (◆) 60 min. The dose-response curves for each time point of pretreatment in (A) and (B) were fitted to sigmoidal curves with a Hill coefficient equal to 1. 155

Fig. 7.2 AB 100 100

80 80

60 60

Effect 40 40

20 20 Desensitization (%)

0 0 -11 -10 -9 -8 -7 -6 0 10 20 30 40 50 60 Log [A], (M) Time (min)

Fig. 7.2. Mathematical simulation of the effect of agonist pre-treatment on concentration-response curves. (A) The concentration-response curves for an agonist

[A] were generated using equation (7.5), with parameters Em = 100, τ0 = 20, τ∞ = 1, KA =

1 nM and k = 0.1 min-1. The time of treatment for the individual curves was set to: 0, 5,

10, 30, and 60 min. (B) The time course of receptor desensitization was simulated with equation (4) using the same values of parameters τ0, τ∞ , and k as in graph (A). 156

Fig. 7.3 AB 120 120

100 100

80 80

60 60 S] stimulation 35

[ 40 40 γ cAMP inhibition 20

GTP 20 (%cells) control max. in (% max. cells) control in 0 0 -11 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 -6 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

Fig. 7.3. Time course of deltorphin II-stimulated desensitization of hDOR fitted by the modified operational model (equation 7.5). (A) The concentration-response curves for deltorphin II-stimulated GTPγ[35S] binding from graph 7.1A were fitted simultaneously using equation (7.5). (B) The concentration-response curves of deltorphin

II-mediated inhibition of forskolin (100 μM, 20 min)-stimulated cAMP accumulation plotted in graph 7.1B were fitted simultaneously using equation (7.5). Deltorphin II pretreatment: (●) 0 min, (○) 2 min, (■) 5 min, (□) 10 min, (▲) 30 min, (△) 45 min, (◆) 60 min. The parameters Em and KA were shared for all time points (0-60 min); for times 0-45 min the parameter τ is defined by equation (7.4) in which τ0 and τ∞ are fitted shared parameters; for the 60 min curve the τ was designated τ60 and was fitted independently.

The fitted parameters Em, KA, k, τ0, and τ∞ for (A) and (B) are summarized in Table 7.1. 157

Fig. 7.4 AB 120 120

100 100

80 80

60 60

S] stimulation 40 40 35 [ γ

20 inhibition cAMP 20 GTP (% max. in control cells) control in max. (% 0 in control(% max. cells) 0

-11 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 -6 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

Fig. 7.4. S363A mutation attenuates deltorphin II-mediated hDOR desensitization. hDOR/CHO (solid line) or hDOR(S363A)/CHO (dotted line) cells were pretreated with

IMDM medium or 100 nM deltorphin II for 60 min at 37ºC. (A) Cells were harvested, cell membranes prepared and deltorphin II-stimulated GTPγ[35S] binding measured, as described in “Materials and Methods”. Deltorphin II concentration curves were normalized to percent of maximum effect in control (untreated) cells. (B) Cells were washed three times to remove the agonist, and dose-response curves for deltorphin II- mediated inhibition of forskolin (100 μM, 20 min)-stimulated cAMP accumulation were measured as described in “Materials and Methods”. The inhibition curves were plotted as the percent of cAMP inhibition in control cells. The normalized data from three (A) and five (B) independent experiments (performed in duplicates) were combined and fitted by sigmoidal dose-response curves with a Hill coefficient set to 1. (●) hDOR/CHO cells, no pretreatment, (○) hDOR/CHO cells, deltorphin II pretreatment, (■) hDOR(S363A)/CHO cells, no pretreatment, (□) hDOR(S363A)/CHO cells, deltorphin II pretreatment. 158

Fig. 7.5 AB 120 120

100 100

80 80

60 60

S] stimulation 40 40 35 [ γ

20 cAMP inhibition 20 GTP % max.in controlcells

0 (% max. control in cells) 0

-11 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 -6 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

Fig. 7.5. Desensitization of the hDOR and hDOR(S363A) fitted by the operational model. The hDOR/CHO (solid line) or hDOR(S363A)/CHO (dotted line) cells were pretreated for 60 min with 100 nM deltorphin II in IMDM medium at 37ºC. (A) The concentration-response curves of deltorphin II-stimulated GTPγ[35S] binding from graph

7.4A were fitted simultaneously using equation (7.1). (B) The concentration-response curves of deltorphin II-mediated inhibition of forskolin (100 μM, 20 min)-stimulated cAMP accumulation from graph 7.4B were fitted simultaneously using equation (7.1).

The parameters Em and KA are shared for all curves. (●) hDOR/CHO cells, no pretreatment, (○) hDOR/CHO cells, deltorphin II pretreatment, (■) hDOR(S363A)/CHO cells, no pretreatment, (□) hDOR(S363A)/CHO cells, deltorphin II pretreatment. The fitted parameters Em, KA, and τ for (A) and (B) are summarized in Table 7.4. 159

Fig. 7.6 AB 120 120 hDOR-IMDM 100 hDOR-Del 100 80 hDOR-SNC 80 60 60 40 40 S stimulation S γ 20 20 cAMP inhibition

GTP 0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 -20 -20 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

CD 120 120 100 100 80 80 60 60 40 40 S stimulation γ 20 20 cAMP Inhibition

GTP 0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 -20 -20 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

Fig. 7.6. hDOR is desensitized more efficiently by SNC80 than by deltorphin II pretreatment. hDOR/CHO cells were pretreated with IMDM (solid line), 100 nM deltorphin II (dotted line) or 100 nM SNC80 (dashed line) in IMDM medium for 60 min at 37ºC. (A) Deltorphin II-stimulated GTPγ[35S] binding was measured, as described in “Materials and methods”. Deltorphin II concentration curves were normalized to percent of maximum effect in control cells. (B) Dose-response curves for deltorphin II-mediated inhibition of forskolin-stimulated cAMP accumulation were measured as described in “Materials and Methods”. The inhibition curves were plotted as the percent of cAMP inhibition in control cells. The normalized data from three (A) and five (B) independent experiments (performed in duplicates) were combined and fitted by sigmoidal dose- response curves with a Hill coefficient equal to 1. (C) The concentration-response curves of deltorphin II-stimulated GTPγ[35S] binding from graph A were fitted simultaneously using equation (7.1) with parameters Em and KA shared for all curves. (D) The concentration-response curves of deltorphin II-mediated inhibition of forskolin- stimulated cAMP accumulation from graph B were fitted simultaneously using equation

(7.1) with parameters Em and KA shared for all curves. (●) no pretreatment, (○) deltorphin II pretreatment, (■) SNC80 pretreatment. 160

Fig. 7.7 AB 120 120 S363A-IMDM 100 100 S363A-Del 80 S363A-SNC 80 60 60 40 40 S stimulation γ 20 20 cAMP inhibition

GTP 0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 -20 -20 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

CD 120 120 100 100 80 80 60 60 40 40 S stimulation γ 20 20 cAMP Inbibition

GTP 0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 -20 -20 Log [Deltorphin II], (M) Log [Deltorphin II], (M)

Fig. 7.7. hDOR(S363A) is desensitized by SNC80 but not by deltorphin II pretreatment. hDOR(S363A)/CHO cells were pretreated with IMDM (solid line), 100 nM deltorphin II (dotted line) or 100 nM SNC80 (dashed line) in IMDM medium for 60 min at 37ºC. (A) Deltorphin II-stimulated GTPγ[35S] binding was measured, as described in “Materials and Methods”. Deltorphin II concentration curves were normalized to percent of maximum effect in control cells. (B) Dose-response curves for deltorphin II- mediated inhibition of forskolin-stimulated cAMP accumulation were measured as described in “Materials and Methods”. The inhibition curves were plotted as the percent of cAMP inhibition in control cells. The normalized data from three (A) and five (B) independent experiments (performed in duplicates) were combined and fitted by sigmoidal dose-response curves with a Hill coefficient equal to 1. (C) The concentration- response curves of deltorphin II-stimulated GTPγ[35S] binding from graph A were fitted simultaneously using equation (7.1) with parameters Em and KA shared for all three curves. (D) The concentration-response curves of deltorphin II-mediated inhibition of forskolin-stimulated cAMP accumulation from graph B were fitted simultaneously using equation (7.1) with parameters Em and KA shared for all curves. (●) no pretreatment, (○) deltorphin II pretreatment, (■) SNC80 pretreatment. 161

REFERENCES

Abdelhamid EE, Sultana M, Portoghese PS and Takemori AE (1991) Selective blockage of delta opioid receptors prevents the development of morphine tolerance and dependence in mice. J Pharmacol Exp Ther 258(1):299-303. Advani RJ, Yang B, Prekeris R, Lee KC, Klumperman J and Scheller RH (1999) VAMP- 7 mediates vesicular transport from endosomes to lysosomes. J Cell Biol 146(4):765-776. Afify EA, Law PY, Riedl M, Elde R and Loh HH (1998) Role of carboxyl terminus of mu-and delta-opioid receptor in agonist-induced down-regulation. Brain Res Mol Brain Res 54(1):24-34. Allouche S, Roussel M, Marie N and Jauzac P (1999) Differential desensitization of human [delta]-opioid receptors by peptide and alkaloid agonists. Eur J Pharmacol 371(2-3):235-240. Asensio VJ, Miralles A and Garcia-Sevilla JA (2006) Stimulation of mitogen-activated protein kinase kinases (MEK1/2) by mu-, delta- and kappa-opioid receptor agonists in the rat brain: regulation by chronic morphine and opioid withdrawal. Eur J Pharmacol 539(1-2):49-56. Barak LS, Gilchrist J, Becker JM and Kim K-M (2006) Relationship between the G protein signaling and homologous desensitization of G protein-coupled receptors. Biochem Biophys Res Commun 339(2):695-700. Belcheva MM, Barg J, Gloeckner C, Gao XM, Chuang DM and Coscia CJ (1992) Antagonist-induced transient down-regulation of delta-opioid receptors in NG108-15 cells. Mol Pharmacol 42(3):445-452. Black JW, Leff P, Shankley NP and Wood J (1985) An operational model of pharmacological agonism: the effect of E/[A] curve shape on agonist dissociation constant estimation. Br J Pharmacol 84(2):561-571. Borgland SL (2001) Acute opioid receptor desensitization and tolerance: is there a link? Clin Exp Pharmacol Physiol 28(3):147-154. Borgland SL, Connor M, Osborne PB, Furness JB and Christie MJ (2003) Opioid agonists have different efficacy profiles for G protein activation, rapid desensitization, and endocytosis of mu-opioid receptors. J Biol Chem 278(21):18776-18784. Brainin-Mattos J, Smith ND, Malkmus S, Rew Y, Goodman M, Taulane J and Yaksh TL (2006) Cancer-related bone pain is attenuated by a systemically available delta- opioid receptor agonist. Pain 122(1-2):174-181. Burns ME, Mendez A, Chen CK, Almuete A, Quillinan N, Simon MI, Baylor DA and Chen J (2006) Deactivation of phosphorylated and nonphosphorylated rhodopsin by arrestin splice variants. J Neurosci 26(3):1036-1044. Cahill CM, Holdridge SV and Morinville A (2007) Trafficking of delta-opioid receptors and other G-protein-coupled receptors: implications for pain and analgesia. Trends Pharmacol Sci 28(1):23-31. 162

Cahill CM, Morinville A, Hoffert C, O'Donnell D and Beaudet A (2003) Up-regulation and trafficking of delta opioid receptor in a model of chronic inflammation: implications for pain control. Pain 101(1-2):199-208. Cahill CM, Morinville A, Lee MC, Vincent JP, Collier B and Beaudet A (2001) Prolonged morphine treatment targets delta opioid receptors to neuronal plasma membranes and enhances delta-mediated antinociception. J Neurosci 21(19):7598-7607. Celver J, Vishnivetskiy SA, Chavkin C and Gurevich VV (2002) Conservation of the phosphate-sensitive elements in the arrestin family of proteins. J Biol Chem 277(11):9043-9048. Cen B, Xiong Y, Ma L and Pei G (2001a) Direct and differential interaction of beta- arrestins with the intracellular domains of different opioid receptors. Mol Pharmacol 59(4):758-764. Cen B, Yu Q, Guo J, Wu Y, Ling K, Cheng Z, Ma L and Pei G (2001b) Direct binding of beta-arrestins to two distinct intracellular domains of the delta opioid receptor. J Neurochem 76(6):1887-1894. Chaturvedi K, Bandari P, Chinen N and Howells RD (2001) Proteasome involvement in agonist-induced down-regulation of mu and delta opioid receptors. J Biol Chem 276(15):12345-12355. Chen YL, Law PY and Loh HH (2006) Sustained activation of phosphatidylinositol 3- kinase/Akt/nuclear factor kappaB signaling mediates G protein-coupled delta- opioid receptor gene expression. J Biol Chem 281(6):3067-3074. Chu P, Murray S, Lissin D and von Zastrow M (1997) Delta and kappa opioid receptors are differentially regulated by dynamin-dependent endocytosis when activated by the same alkaloid agonist. J Biol Chem 272(43):27124-27130. Ciechanover A (2005) Intracellular protein degradation: from a vague idea, through the lysosome and the ubiquitin-proteasome system, and onto human diseases and drug targeting (Nobel lecture). Angew Chem Int Ed Engl 44(37):5944-5967. Ciechanover A (2006) The ubiquitin proteolytic system: from a vague idea, through basic mechanisms, and onto human diseases and drug targeting. Neurology 66(2 Suppl 1):S7-19. Connor M, Osborne PB and Christie MJ (2004) Mu-opioid receptor desensitization: is morphine different? Br J Pharmacol 143(6):685-696. Cvejic S, Trapaidze N, Cyr C and Devi LA (1996) Thr353, located within the COOH- terminal tail of the delta opiate receptor, is involved in receptor down-regulation. J Biol Chem 271(8):4073-4076. Daaka Y, Luttrell LM and Lefkowitz RJ (1997) Switching of the coupling of the beta2- adrenergic receptor to different G proteins by protein kinase A. Nature 390(6655):88-91. Dang V and Christie M (2006) Beta-arrestin-2 independent regulation of mu opioid receptor in locus coeruleus neurons, in International Narcotics Research Conference p 48, St. Paul, Minnesota. 163

Dicker F, Quitterer U, Winstel R, Honold K and Lohse MJ (1999) Phosphorylation- independent inhibition of parathyroid hormone receptor signaling by G protein- coupled receptor kinases. Proc Natl Acad Sci U S A 96(10):5476-5481. Drolet G, Dumont EC, Gosselin I, Kinkead R, Laforest S and Trottier JF (2001) Role of endogenous opioid system in the regulation of the stress response. Prog Neuropsychopharmacol Biol Psychiatry 25(4):729-741. Eisinger DA, Ammer H and Schulz R (2002) Chronic Morphine Treatment Inhibits Opioid Receptor Desensitization and Internalization. J Neurosci 22(23):10192- 10200. Eisinger DA and Schulz R (2004) Extracellular signal-regulated kinase/mitogen-activated protein kinases block internalization of delta-opioid receptors. J Pharmacol Exp Ther 309(2):776-785. El Kouhen R, Burd AL, Erickson-Herbrandson LJ, Chang CY, Law PY and Loh HH (2001) Phosphorylation of Ser363, Thr370, and Ser375 residues within the carboxyl tail differentially regulates mu-opioid receptor internalization. J Biol Chem 276(16):12774-12780. Elorza A, Sarnago S and Mayor F, Jr. (2000) Agonist-dependent modulation of G protein-coupled receptor kinase 2 by mitogen-activated protein kinases. Mol Pharmacol 57(4):778-783. Fraser GL, Gaudreau GA, Clarke PB, Menard DP and Perkins MN (2000) Antihyperalgesic effects of delta opioid agonists in a rat model of chronic inflammation. Br J Pharmacol 129(8):1668-1672. Freedman NJ, Ament AS, Oppermann M, Stoffel RH, Exum ST and Lefkowitz RJ (1997) Phosphorylation and Desensitization of Human Endothelin A and B Receptors. Evidence for G protein-coupled receptor kinase specificity. J Biol Chem 272(28):17734-17743. Gage RM, Kim KA, Cao TT and von Zastrow M (2001) A transplantable sorting signal that is sufficient to mediate rapid recycling of G protein-coupled receptors. J Biol Chem 276(48):44712-44720. Gendron L, Lucido AL, Mennicken F, O'Donnell D, Vincent JP, Stroh T and Beaudet A (2006) Morphine and pain-related stimuli enhance cell surface availability of somatic delta-opioid receptors in rat dorsal root ganglia. J Neurosci 26(3):953- 962. Ghanouni P, Gryczynski Z, Steenhuis JJ, Lee TW, Farrens DL, Lakowicz JR and Kobilka BK (2001) Functionally different agonists induce distinct conformations in the G protein coupling domain of the beta 2 adrenergic receptor. J Biol Chem 276(27):24433-24436. Ghinea N, Vu Hai MT, Groyer-Picard MT, Houllier A, Schoevaert D and Milgrom E (1992) Pathways of internalization of the hCG/LH receptor: immunoelectron microscopic studies in Leydig cells and transfected L-cells. J Cell Biol 118(6):1347-1358. Gilman AG (1970) A protein binding assay for adenosine 3':5'-cyclic monophosphate. Proc Natl Acad Sci U S A 67(1):305-312. 164

Gomes I, Gupta A, Filipovska J, Szeto HH, Pintar JE and Devi LA (2004) A role for heterodimerization of mu and delta opiate receptors in enhancing morphine analgesia. Proc Natl Acad Sci U S A 101(14):5135-5139. Goodman OB, Jr., Krupnick JG, Santini F, Gurevich VV, Penn RB, Gagnon AW, Keen JH and Benovic JL (1996) Beta-arrestin acts as a clathrin adaptor in endocytosis of the beta2-adrenergic receptor. Nature 383(6599):447-450. Guan JS, Xu ZZ, Gao H, He SQ, Ma GQ, Sun T, Wang LH, Zhang ZN, Lena I, Kitchen I, Elde R, Zimmer A, He C, Pei G, Bao L and Zhang X (2005) Interaction with vesicle luminal protachykinin regulates surface expression of delta-opioid receptors and opioid analgesia. Cell 122(4):619-631. Guo J, Wu Y, Zhang W, Zhao J, Devi LA, Pei G and Ma L (2000) Identification of G protein-coupled receptor kinase 2 phosphorylation sites responsible for agonist- stimulated delta-opioid receptor phosphorylation. Mol Pharmacol 58(5):1050- 1056. Gurevich VV and Gurevich EV (2004) The molecular acrobatics of arrestin activation. Trends Pharmacol Sci 25(2):105-111. Gurevich VV and Gurevich EV (2006) The structural basis of arrestin-mediated regulation of G-protein-coupled receptors. Pharmacol Ther 110(3):465-502. Hall RA and Lefkowitz RJ (2002) Regulation of G protein-coupled receptor signaling by scaffold proteins. Circ Res 91(8):672-680. Handley-Gearhart PM, Trausch-Azar JS, Ciechanover A and Schwartz AL (1994) Rescue of the complex temperature-sensitive phenotype of Chinese hamster ovary E36ts20 cells by expression of the human ubiquitin-activating enzyme cDNA. Biochem J 304 ( Pt 3):1015-1020. Hasbi A, Allouche S, Sichel F, Stanasila L, Massotte D, Landemore G, Polastron J and Jauzac P (2000) Internalization and Recycling of delta -Opioid Receptor Are Dependent on a Phosphorylation-Dephosphorylation Mechanism. J Pharmacol Exp Ther 293(1):237-247. Hasbi A, Polastron J, Allouche S, Stanasila L, Massotte D and Jauzac P (1998) Desensitization of the delta-opioid receptor correlates with its phosphorylation in SK-N-BE cells: involvement of a G protein-coupled receptor kinase. J Neurochem 70(5):2129-2138. Hosohata Y, Varga EV, Stropova D, Li X, Knapp RJ, Hruby VJ, Rice KC, Nagase H, Roeske WR and Yamamura HI (2001) Mutation W284L of the human delta opioid receptor reveals agonist specific receptor conformations for G protein activation. Life Sci 68(19-20):2233-2242. Huang P, Xu W, Yoon SI, Chen C, Chong PL and Liu-Chen LY (2007) Cholesterol reduction by methyl-beta-cyclodextrin attenuates the delta opioid receptor- mediated signaling in neuronal cells but enhances it in non-neuronal cells. Biochem Pharmacol 73(4):534-549. Hurley RW and Hammond DL (2000) The analgesic effects of supraspinal mu and delta opioid receptor agonists are potentiated during persistent inflammation. J Neurosci 20(3):1249-1259. 165

Insel PA, Head BP, Patel HH, Roth DM, Bundey RA and Swaney JS (2005) Compartmentation of G-protein-coupled receptors and their signalling components in lipid rafts and caveolae. Biochem Soc Trans 33(Pt 5):1131-1134. Inturrisi CE (2002) Clinical pharmacology of opioids for pain. Clin J Pain 18(4 Suppl):S3-13. Kabli N and Cahill CM (2007) Anti-allodynic effects of peripheral delta opioid receptors in neuropathic pain. Pain 127(1-2):84-93. Keith DE, Anton B, Murray SR, Zaki PA, Chu PC, Lissin DV, Monteillet-Agius G, Stewart PL, Evans CJ and von Zastrow M (1998) mu-Opioid receptor internalization: opiate drugs have differential effects on a conserved endocytic mechanism in vitro and in the mammalian brain. Mol Pharmacol 53(3):377-384. Keith DE, Murray SR, Zaki PA, Chu PC, Lissin DV, Kang L, Evans CJ and von Zastrow M (1996) Morphine activates opioid receptors without causing their rapid internalization. J Biol Chem 271(32):19021-19024. Kieffer BL (1999) Opioids: first lessons from knockout mice. Trends Pharmacol Sci 20(1):19-26. Kieffer BL and Gaveriaux-Ruff C (2002) Exploring the opioid system by gene knockout. Prog Neurobiol 66(5):285-306. King T, Ossipov MH, Vanderah TW, Porreca F and Lai J (2005) Is paradoxical pain induced by sustained opioid exposure an underlying mechanism of opioid antinociceptive tolerance? Neurosignals 14(4):194-205. Kiss AL, Turi A, Muller N, Kantor O and Botos E (2002) Caveolae and caveolin isoforms in rat peritoneal macrophages. Micron 33(1):75-93. Kouhen OM, Wang G, Solberg J, Erickson LJ, Law PY and Loh HH (2000) Hierarchical phosphorylation of delta-opioid receptor regulates agonist-induced receptor desensitization and internalization. J Biol Chem 275(47):36659-36664. Kramer HK and Simon EJ (2000) Mu and delta-opioid receptor agonists induce mitogen- activated protein kinase (MAPK) activation in the absence of receptor internalization. Neuropharmacol 39(10):1707-1719. Krasel C, Vilardaga JP, Bunemann M and Lohse MJ (2004) Kinetics of G-protein- coupled receptor signalling and desensitization. Biochem Soc Trans 32(Pt 6):1029-1031. Krupnick JG and Benovic JL (1998) The role of receptor kinases and arrestins in G protein-coupled receptor regulation. Annu Rev Pharmacol Toxicol 38:289-319. Laporte SA, Oakley RH, Holt JA, Barak LS and Caron MG (2000) The interaction of beta-arrestin with the AP-2 adaptor is required for the clustering of beta 2- adrenergic receptor into clathrin-coated pits. J Biol Chem 275(30):23120-23126. Law PY, Kouhen OM, Solberg J, Wang W, Erickson LJ and Loh HH (2000) Deltorphin II-induced rapid desensitization of delta-opioid receptor requires both phosphorylation and internalization of the receptor. J Biol Chem 275(41):32057- 32065. Lee KB, Pals-Rylaarsdam R, Benovic JL and Hosey MM (1998) Arrestin-independent internalization of the m1, m3, and m4 subtypes of muscarinic cholinergic receptors. J Biol Chem 273(21):12967-12972. 166

Li J, Xiang B, Su W, Zhang X, Huang Y and Ma L (2003) Agonist-induced formation of opioid receptor-G protein-coupled receptor kinase (GRK)-G beta gamma complex on membrane is required for GRK2 function in vivo. J Biol Chem 278(32):30219- 30226. Li JG, Benovic JL and Liu-Chen LY (2000) Mechanisms of agonist-induced down- regulation of the human kappa-opioid receptor: internalization is required for down-regulation. Mol Pharmacol 58(4):795-801. Lohse MJ, Benovic JL, Caron MG and Lefkowitz RJ (1990) Multiple pathways of rapid beta 2-adrenergic receptor desensitization. Delineation with specific inhibitors. J Biol Chem 265(6):3202-3211. Lowe JD, Celver JP, Gurevich VV and Chavkin C (2002) mu-Opioid receptors desensitize less rapidly than delta-opioid receptors due to less efficient activation of arrestin. J Biol Chem 277(18):15729-15735. Lucido AL, Morinville A, Gendron L, Stroh T and Beaudet A (2005) Prolonged morphine treatment selectively increases membrane recruitment of delta-opioid receptors in mouse basal ganglia. J Mol Neurosci 25(3):207-214. Luttrell LM, Ferguson SS, Daaka Y, Miller WE, Maudsley S, Della Rocca GJ, Lin F, Kawakatsu H, Owada K, Luttrell DK, Caron MG and Lefkowitz RJ (1999) Beta- arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science 283(5402):655-661. Luttrell LM and Lefkowitz RJ (2002) The role of beta-arrestins in the termination and transduction of G-protein-coupled receptor signals. J Cell Sci 115(Pt 3):455-465. Ma J, Zhang Y, Kalyuzhny AE and Pan ZZ (2006) Emergence of Functional {delta}- Opioid Receptors Induced by Long-Term Treatment with Morphine. Mol Pharmacol 69(4):1137-1145. Malatynska E, Wang Y, Knapp RJ, Waite S, Calderon S, Rice K, Hruby VJ, Yamamura HI and Roeske WR (1996) Human delta opioid receptor: functional studies on stably transfected Chinese hamster ovary cells after acute and chronic treatment with the selective nonpeptidic agonist SNC-80. J Pharmacol Exp Ther 278(3):1083-1089. Marie N, Lecoq I, Jauzac P and Allouche S (2003) Differential sorting of human delta- opioid receptors after internalization by peptide and alkaloid agonists. J Biol Chem 278(25):22795-22804. Marion S, Oakley RH, Kim K-M, Caron MG and Barak LS (2006) A beta-Arrestin Binding Determinant Common to the Second Intracellular Loops of Rhodopsin Family G Protein-coupled Receptors. J Biol Chem 281(5):2932-2938. Maudsley S, Pierce KL, Zamah AM, Miller WE, Ahn S, Daaka Y, Lefkowitz RJ and Luttrell LM (2000) The beta(2)-adrenergic receptor mediates extracellular signal- regulated kinase activation via assembly of a multi-receptor complex with the epidermal growth factor receptor. J Biol Chem 275(13):9572-9580. Mika J, Przewlocki R and Przewlocka B (2001) The role of delta-opioid receptor subtypes in neuropathic pain. Eur J Pharmacol 415(1):31-37. Miller WE, Maudsley S, Ahn S, Khan KD, Luttrell LM and Lefkowitz RJ (2000) beta- arrestin1 interacts with the catalytic domain of the tyrosine kinase c-SRC. Role of 167

beta-arrestin1-dependent targeting of c-SRC in receptor endocytosis. J Biol Chem 275(15):11312-11319. Monis GF, Schultz C, Ren R, Eberhard J, Costello C, Connors L, Skinner M and Trinkaus-Randall V (2006) Role of endocytic inhibitory drugs on internalization of amyloidogenic light chains by cardiac fibroblasts. Am J Pathol 169(6):1939- 1952. Morinville A, Cahill CM, Aibak H, Rymar VV, Pradhan A, Hoffert C, Mennicken F, Stroh T, Sadikot AF, O'Donnell D, Clarke PB, Collier B, Henry JL, Vincent JP and Beaudet A (2004) Morphine-induced changes in delta opioid receptor trafficking are linked to somatosensory processing in the rat spinal cord. J Neurosci 24(24):5549-5559. Morinville A, Cahill CM, Esdaile MJ, Aibak H, Collier B, Kieffer BL and Beaudet A (2003) Regulation of delta-opioid receptor trafficking via mu-opioid receptor stimulation: evidence from mu-opioid receptor knock-out mice. J Neurosci 23(12):4888-4898. Mousavi SA, Kjeken R, Berg TO, Seglen PO, Berg T and Brech A (2001) Effects of inhibitors of the vacuolar proton pump on hepatic heterophagy and autophagy. Biochim Biophys Acta 1510(1-2):243-257. Mueller A, Kelly E and Strange PG (2002) Pathways for internalization and recycling of the chemokine receptor CCR5. Blood 99(3):785-791. Murray SR, Evans CJ and von Zastrow M (1998) Phosphorylation is not required for dynamin-dependent endocytosis of a truncated mutant opioid receptor. J Biol Chem 273(39):24987-24991. Navratilova E, Eaton MC, Stropova D, Varga EV, Vanderah TW, Roeske WR and Yamamura HI (2005) Morphine promotes phosphorylation of the human delta- opioid receptor at serine 363. Eur J Pharmacol 519(3):212-214. Navratilova E, Varga EV, Stropova D, Jambrosic JC, Roeske WR and Yamamura HI (2004) Mutation S363A in the human delta-opioid receptor selectively reduces down-regulation by a peptide agonist. Eur J Pharmacol 485(1-3):341-343. Navratilova E, Waite S, Stropova D, Eaton MC, Alves ID, Hruby VJ, Roeske WR, Yamamura HI and Varga EV (2007) Quantitative evaluation of the human delta opioid receptor desensitization using the Operational Model of Drug Action. Mol Pharmacol submitted. Neufeld EB, Cooney AM, Pitha J, Dawidowicz EA, Dwyer NK, Pentchev PG and Blanchette-Mackie EJ (1996) Intracellular trafficking of cholesterol monitored with a cyclodextrin. J Biol Chem 271(35):21604-21613. Oakley RH, Laporte SA, Holt JA, Caron MG and Barak LS (2000) Differential affinities of visual arrestin, beta arrestin1, and beta arrestin2 for G protein-coupled receptors delineate two major classes of receptors. J Biol Chem 275(22):17201- 17210. Okamoto Y, Ninomiya H, Miwa S and Masaki T (2000) Cholesterol oxidation switches the internalization pathway of endothelin receptor type A from caveolae to clathrin-coated pits in Chinese hamster ovary cells. J Biol Chem 275(9):6439- 6446. 168

Okura T, Cowell SM, Varga E, Burkey TH, Roeske WR, Hruby VJ and Yamamura HI (2000) Differential down-regulation of the human delta-opioid receptor by SNC80 and [D-Pen(2),D-Pen(5)]enkephalin. Eur J Pharmacol 387(2):R11-13. Okura T, Varga EV, Hosohata Y, Navratilova E, Cowell SM, Rice K, Nagase H, Hruby VJ, Roeske WR and Yamamura HI (2003) Agonist-specific down-regulation of the human delta-opioid receptor. Eur J Pharmacol 459(1):9-16. Osborne PB and Williams JT (1995) Characterization of acute homologous desensitization of mu-opioid receptor-induced currents in locus coeruleus neurones. Br J Pharmacol 115(6):925-932. Ostrom RS and Insel PA (2004) The evolving role of lipid rafts and caveolae in G protein-coupled receptor signaling: implications for molecular pharmacology. Br J Pharmacol 143(2):235-245. Ostrom RS, Violin JD, Coleman S and Insel PA (2000) Selective enhancement of beta- adrenergic receptor signaling by overexpression of adenylyl cyclase type 6: colocalization of receptor and adenylyl cyclase in caveolae of cardiac myocytes. Mol Pharmacol 57(5):1075-1079. Pei G, Kieffer BL, Lefkowitz RJ and Freedman NJ (1995) Agonist-dependent phosphorylation of the mouse delta-opioid receptor: involvement of G protein- coupled receptor kinases but not protein kinase C. Mol Pharmacol 48(2):173-177. Petaja-Repo UE, Hogue M, Bhalla S, Laperriere A, Morello JP and Bouvier M (2002) Ligands act as pharmacological chaperones and increase the efficiency of delta opioid receptor maturation. Embo J 21(7):1628-1637. Petaja-Repo UE, Hogue M, Laperriere A, Bhalla S, Walker P and Bouvier M (2001) Newly Synthesized Human delta Opioid Receptors Retained in the Endoplasmic Reticulum Are Retrotranslocated to the Cytosol, Deglycosylated, Ubiquitinated, and Degraded by the Proteasome. J Biol Chem 276(6):4416-4423. Pierce KL, Luttrell LM and Lefkowitz RJ (2001) New mechanisms in heptahelical receptor signaling to mitogen activated protein kinase cascades. Oncogene 20(13):1532-1539. Pitcher JA, Inglese J, Higgins JB, Arriza JL, Casey PJ, Kim C, Benovic JL, Kwatra MM, Caron MG and Lefkowitz RJ (1992) Role of beta gamma subunits of G proteins in targeting the beta-adrenergic receptor kinase to membrane-bound receptors. Science 257(5074):1264-1267. Pitcher JA, Tesmer JJ, Freeman JL, Capel WD, Stone WC and Lefkowitz RJ (1999) Feedback inhibition of G protein-coupled receptor kinase 2 (GRK2) activity by extracellular signal-regulated kinases. J Biol Chem 274(49):34531-34534. Porreca F, Takemori AE, Sultana M, Portoghese PS, Bowen WD and Mosberg HI (1992) Modulation of mu-mediated antinociception in the mouse involves opioid delta-2 receptors. J Pharmacol Exp Ther 263(1):147-152. Prekeris R, Yang B, Oorschot V, Klumperman J and Scheller RH (1999) Differential roles of syntaxin 7 and syntaxin 8 in endosomal trafficking. Mol Biol Cell 10(11):3891-3908. Przewlocki R and Przewlocka B (2001) Opioids in chronic pain. Eur J Pharmacol 429(1- 3):79-91. 169

Puthenveedu MA and von Zastrow M (2006) Cargo Regulates Clathrin-Coated Pit Dynamics. Cell 127(1):113-124. Quock RM, Burkey TH, Varga E, Hosohata Y, Hosohata K, Cowell SM, Slate CA, Ehlert FJ, Roeske WR and Yamamura HI (1999) The delta-opioid receptor: molecular pharmacology, signal transduction, and the determination of drug efficacy. Pharmacol Rev 51(3):503-532. Quock RM, Hosohata Y, Knapp RJ, Burkey TH, Hosohata K, Zhang X, Rice KC, Nagase H, Hruby VJ, Porreca F, Roeske WR and Yamamura HI (1997) Relative efficacies of delta-opioid receptor agonists at the cloned human delta-opioid receptor. Eur J Pharmacol 326(1):101-104. Rapaka RS and Porreca F (1991) Development of delta opioid peptides as nonaddicting analgesics. Pharm Res 8(1):1-8. Razani B, Woodman SE and Lisanti MP (2002) Caveolae: from cell biology to animal physiology. Pharmacol Rev 54(3):431-467. Reiter E and Lefkowitz RJ (2006) GRKs and beta-arrestins: roles in receptor silencing, trafficking and signaling. Trends Endocrinol Metab 17(4):159-165. Roettger BF, Rentsch RU, Pinon D, Holicky E, Hadac E, Larkin JM and Miller LJ (1995) Dual pathways of internalization of the cholecystokinin receptor. J Cell Biol 128(6):1029-1041. Schulz S, Mayer D, Pfeiffer M, Stumm R, Koch T and Hollt V (2004) Morphine induces terminal micro-opioid receptor desensitization by sustained phosphorylation of serine-375. Embo J 23(16):3282-3289. Shenoy SK, Drake MT, Nelson CD, Houtz DA, Xiao K, Madabushi S, Reiter E, Premont RT, Lichtarge O and Lefkowitz RJ (2006) Beta-arrestin-dependent, G protein- independent ERK1/2 activation by the beta2 adrenergic receptor. J Biol Chem 281(2):1261-1273. Shenoy SK, McDonald PH, Kohout TA and Lefkowitz RJ (2001) Regulation of receptor fate by ubiquitination of activated beta 2-adrenergic receptor and beta-arrestin. Science 294(5545):1307-1313. Sternini C, Spann M, Anton B, Keith DE, Jr, Bunnett NW, von Zastrow M, Evans C and Brecha NC (1996) Agonist-selective endocytosis of mu opioid receptor by neurons in vivo. Proc Natl Acad Sci U S A 93(17):9241-9246. Tanowitz M and von Zastrow M (2002) Ubiquitination-independent Trafficking of G Protein-coupled Receptors to Lysosomes. J Biol Chem 277(52):50219-50222. Tanowitz M and von Zastrow M (2003) A Novel Endocytic Recycling Signal That Distinguishes the Membrane Trafficking of Naturally Occurring Opioid Receptors. J Biol Chem 278(46):45978-45986. Tohgo A, Choy EW, Gesty-Palmer D, Pierce KL, Laporte S, Oakley RH, Caron MG, Lefkowitz RJ and Luttrell LM (2003) The stability of the G protein-coupled receptor-beta-arrestin interaction determines the mechanism and functional consequence of ERK activation. J Biol Chem 278(8):6258-6267. Tran TM, Friedman J, Qunaibi E, Baameur F, Moore RH and Clark RB (2004) Characterization of agonist stimulation of cAMP-dependent protein kinase and G 170

protein-coupled receptor kinase phosphorylation of the beta2-adrenergic receptor using phosphoserine-specific antibodies. Mol Pharmacol 65(1):196-206. Trapaidze N, Cvejic S, Nivarthi RN, Abood M and Devi LA (2000) Role for C-tail residues in delta opioid receptor downregulation. DNA Cell Biol 19(2):93-101. Trapaidze N, Keith DE, Cvejic S, Evans CJ and Devi LA (1996) Sequestration of the delta opioid receptor. Role of the C terminus in agonist-mediated internalization. J Biol Chem 271(46):29279-29285. Trzeciakowski JP (1999) Stimulus amplification, efficacy, and the operational model. Part I--binary complex occupancy mechanisms. J Theor Biol 198(3):329-346. Tsao PI and von Zastrow M (2000) Type-specific sorting of G protein-coupled receptors after endocytosis. J Biol Chem 275(15):11130-11140. van der Bliek AM, Redelmeier TE, Damke H, Tisdale EJ, Meyerowitz EM and Schmid SL (1993) Mutations in human dynamin block an intermediate stage in coated vesicle formation. J Cell Biol 122(3):553-563. Varga EV, Navratilova E, Stropova D, Jambrosic J, Roeske WR and Yamamura HI (2004) Agonist-specific regulation of the delta-opioid receptor. Life Sciences 76(6):599-612. Varga EV, Rubenzik MK, Stropova D, Sugiyama M, Grife V, Hruby VJ, Rice KC, Roeske WR and Yamamura HI (2003) Converging protein kinase pathways mediate adenylyl cyclase superactivation upon chronic delta-opioid agonist treatment. J Pharmacol Exp Ther 306(1):109-115. Vishnivetskiy SA, Hirsch JA, Velez MG, Gurevich YV and Gurevich VV (2002) Transition of arrestin into the active receptor-binding state requires an extended interdomain hinge. J Biol Chem 277(46):43961-43967. von Zastrow M (2004a) A cell biologist's perspective on physiological adaptation to opiate drugs. Neuropharmacol 47(Supplement 1):286-292. von Zastrow M (2004b) Opioid receptor regulation. Neuromolecular Med 5(1):51-58. von Zastrow M, Svingos A, Haberstock-Debic H and Evans C (2003) Regulated endocytosis of opioid receptors: cellular mechanisms and proposed roles in physiological adaptation to opiate drugs. Curr Opin Neurobiol 13(3):348-353. Walwyn W, Maidment NT, Sanders M, Evans CJ, Kieffer BL and Hales TG (2005) Induction of delta Opioid Receptor Function by Up-Regulation of Membrane Receptors in Mouse Primary Afferent Neurons. Mol Pharmacol 68(6):1688-1698. Wang H and Pickel VM (2001) Preferential cytoplasmic localization of delta-opioid receptors in rat striatal patches: comparison with plasmalemmal mu-opioid receptors. J Neurosci 21(9):3242-3250. Wang W, Loh HH and Law PY (2003) The intracellular trafficking of opioid receptors directed by carboxyl tail and a di-leucine motif in Neuro2A cells. J Biol Chem 278(38):36848-36858. Whaley BS, Yuan N, Birnbaumer L, Clark RB and Barber R (1994) Differential expression of the beta-adrenergic receptor modifies agonist stimulation of adenylyl cyclase: a quantitative evaluation. Mol Pharmacol 45(3):481-489. 171

Whistler JL, Chuang HH, Chu P, Jan LY and von Zastrow M (1999) Functional dissociation of mu opioid receptor signaling and endocytosis: implications for the biology of opiate tolerance and addiction. Neuron 23(4):737-746. Whistler JL, Tsao P and von Zastrow M (2001) A phosphorylation-regulated brake mechanism controls the initial endocytosis of opioid receptors but is not required for post-endocytic sorting to lysosomes. J Biol Chem 276(36):34331-34338. Whistler JL and von Zastrow M (1998) Morphine-activated opioid receptors elude desensitization by beta-arrestin. Proc Natl Acad Sci U S A 95(17):9914-9919. Willets J and Kelly E (2001) Desensitization of endogenously expressed [delta]-opioid receptors: no evidence for involvement of G protein-coupled receptor kinase 2. Eur J Pharmacol 431(2):133-141. Xiang B, Yu GH, Guo J, Chen L, Hu W, Pei G and Ma L (2001) Heterologous activation of protein kinase C stimulates phosphorylation of delta-opioid receptor at serine 344, resulting in beta-arrestin- and clathrin-mediated receptor internalization. J Biol Chem 276(7):4709-4716. Xu J, Dodd RL, Makino CL, Simon MI, Baylor DA and Chen J (1997) Prolonged photoresponses in transgenic mouse rods lacking arrestin. Nature 389(6650):505- 509. Yadav PN, Chaturvedi K and Howells RD (2006) Inhibition of agonist-induced down regulation of the delta opioid receptor with a proteasome inhibitor attenuates opioid tolerance in human embryonic kidney 293 cells. J Pharmacol Exp Ther. Zhang J, Ferguson SS, Barak LS, Aber MJ, Giros B, Lefkowitz RJ and Caron MG (1997) Molecular mechanisms of G protein-coupled receptor signaling: role of G protein- coupled receptor kinases and arrestins in receptor desensitization and resensitization. Receptors Channels 5(3-4):193-199. Zhang J, Ferguson SS, Law PY, Barak LS and Caron MG (1999) Agonist-specific regulation of delta-opioid receptor trafficking by G protein-coupled receptor kinase and beta-arrestin. J Recept Signal Transduct Res 19(1-4):301-313. Zhang X, Bao L and Guan JS (2006) Role of delivery and trafficking of delta-opioid peptide receptors in opioid analgesia and tolerance. Trends Pharmacol Sci 27(6):324-329. Zhang X, Wang F, Chen X, Li J, Xiang B, Zhang YQ, Li BM and Ma L (2005) Beta- arrestin1 and beta-arrestin2 are differentially required for phosphorylation- dependent and -independent internalization of delta-opioid receptors. J Neurochem 95(1):169-178. Zhu Y, King MA, Schuller AG, Nitsche JF, Reidl M, Elde RP, Unterwald E, Pasternak GW and Pintar JE (1999) Retention of supraspinal delta-like analgesia and loss of morphine tolerance in delta opioid receptor knockout mice. Neuron 24(1):243- 252.