Nonopioid actions of dynorphin and neuropathic pain

Item Type text; Dissertation-Reproduction (electronic)

Authors Tang, Qingbo

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 27/09/2021 14:34:07

Link to Item http://hdl.handle.net/10150/289125 INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may t>e from any type of computer printer.

The quality of this reproduction is dependent upon the quality of ttie copy submitted. Broken or indistinct print, colored or poor quality illustrations and photograp>hs, print bleedthrough, substandard margins, and improper alignment can adversely aff^ reproduction. in the unlikely event that the author dki not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g.. maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overiaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6' x 9' k>lack and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning 300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA 800-521-0600

NONOPIOID ACTIONS OF DYNORPHIN AND NEUROPATHIC PAIN

by Qingbo Tang

A Dissertation Submitted to the Faculty of the

COMMITTEE ON PHARMACOLOGY AND TOXICOLOGY (GRADUATE)

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College THE UNIVERSITY OF ARIZONA

2000 UMI Number. 9965931

UMI^

UMI Microfomn9965931 Copyright 2000 by Bell & Howell Infomiation and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Art)or, Ml 48106-1346 2

THE UNIVERSITY OF ARIZONA ® GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have

read the dissertacion prepared by Oinabo Tang entitled Nonopioid Actions of Dynorphin and Neuropathic Pain

and recommend that it be accepted as fulfilling the dissertation reqiiijc^ment for the Degree of Doctor of Philosophy

Fh.D. Date Vt./oo tsemara futscner, Tn.HT Date

day (jptfiaoirx, fn.u. Dite/ ^f7/

Date

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy 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.

Dissertation Date 3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an 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 acknowledgment 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. 4

ACKNOWLEDGEMENTS

As I start to prepare this acknowledgment, many names come into my mind. First is my Ph.D. advisor Josephine Lai. Josephine has been a passionate advisor to me, whenever I have problem, either academically or in personal maters, I can turn to her and count on her help. I have always been taken care o^ whether it was something I needed for experiments, or a meeting that I wanted attending or an absence from the lab when my children got sick. She has always encouraged me and provided timely guidance when I started new experiments, and spent many extra hours in helping me in analyzing data and writing papers. For all of these and many others I have forgotten to mention, I sincerely thank Josephine. I would also like to thank Frank Porreca. His passion for science and the wonderful work done in his lab have been the driving force for this project. My interaction with him has always been stimulating and beneficial to me.

The other faculty members on my dissertation committee also deserve special recognition. I thank them for all their advice and comments on my research. Ronald Lynch has been a close collaborator on the calcium imaging project. He not only opened his lab to me, but also helped me in designing experiments. He has been a good fiiend of mine and whenever I approach him, I feel his Idndness. I knew Jay Gandolfi before I entered into this program. He was my wife's advisor during her graduate study here and has been very nice to my whole family since then. I have been enjoyed a very good relationship with the whole Gandolfi lab and some experiments described in this dissertation was done in his lab. I sincerely thank Jay. I also want to thank Bernard Futscher for agreeing to be in my committee. I admire his expertise in molecular biology and wish I had made better use of it.

There have been many post-docs, graduate students and staff in this department that have helped me along the way. Dr. Ma has taught me radioligand binding assay and other laboratory techniques. He can always find things when I need them for my experiments. I have enjoyed a good friendship with Di Bian and Chengming Zhong. whenever I need animals or need somebody doing animal surgery for me, I can always count on them. Sandy Werget is one of the nicest people I know and I have enjoyed a good relationship with her. She and Ruizhong Wang have also introduced me into immunohistochemistry. Allan Parrish, a former postdoc in Dr. Gandolfi's lab, now an assistant professor in Texas A & M, has taught me gel shift assay. Todd Vanderah has always been nice to me, his presence is a blessing to my daily experience at this department. Mike Ossipov is my counsel on computer solftwares. Whenever I have a problem with my computer or a program, I run into his office. C.J. Kovelowski provided me tissues of opioid tolerant rats and behavoral data on the tolerant animals. Ronnie Gandhoke produced binding datat on transient expressed NMDA receptors in HEK293 cells. I also thank the rest of the Lai lab, especially Connie Kim, Xiaofang Wu, and Shannon Burgess. Finally I would also like to thank Cherri Anderson, Melissa, Terri Vorholzer for their help with administrative chores. 5

TABLE OF CONTENTS

LIST OF FIGURES 8 LIST OF TABLES 11 ABSTRACT 12

L INTRODUCTION 14 Dynorphin: History, Expression and Pharmacology 14 History 14 Structure of the pro

TABLE OF CONTENTS - Continued

Preparation of cortical neuronal cultures 78 Intrathecal catheter placements and inducement of opioid tolerance 78 Nociceptive assess 79 Tissue dissection 79 Immunoassay of dynorphin in spinal tissue 79 Electrophoresis mobility shift assay 80 Results 81 Studies with cortical cultures 81 Assessment of tolerance 81 Immunoassay of (fynorphin 83 Gel shift experiments 84 Discussion 86

IV. BINDING EXPERIMENTS WITH lODINATED DYNORPHIN A(2-I7) 90 Introduction 90 Materials and Methods 94 Chemicals 94 Radiochemicals 94 Synthesis of (des-Tyrosyl)(fynorphin A(2-17..) 94 Preparation of membranes from rat brain cortex 95 Radioligand binding assays in rat brain membranes 95 Transient expression of NMDAR in HEK293 cells 96 Radioligand binding assays in transected HEK293 cells 98 Data analysis 98

Results 100 Saturation analysis 100 Modulation with NMDA receptor ligands 101 Studies with transient expression ofNMDAla/2A 104 Discussion 106

V. EFFECT OF DYNORPHIN ON THE INTRACELLULAR CALCIUM LEVEL 112 Introduction 112 Materials and Methods 114 Materials 114 Preparation of cortical neuron cultures 114 Fluorescence microscopy 114 Radioligand binding. 116 7

TABLE OF CONTENTS — Continued

Results 117 Discussion 126

VI. STUDIES OF GENE EXPRESSION IN L5/L6 LIGATED RATS.130 Introduction 130 Materials and Methods 131 Preparation of total RNA 131 Preparation of cDNA probe 131 Hybridization of cDNA to the Atlas array 132 ^antitative analysis 132 Results 134 Profile of genes in the Atlas arrc^ 134 Profile of gene expression in the dorsal horn of rat spiruil cord. 134 Changes in gene expression after L5/L6 ligation 135 Discussion 138 VII. SUMMARY STATEMENTS 144

APPENDIX 1. ABBREVIATIONS AND DEFINITIONS 154 APPENDIX 2. LIST OF PUBLICATIONS 156 REFERENCES 157 8

LIST OF FIGURES

2.1. AP-1 and NF-icB activity in the dorsal hom of the lumbar spinal cord 52

2.2. Time course for changes in the DNA binding activity of AP-1 and NFKB following nerve L5/L5 ligation 53

2.3. Time course for spinal prodynorphin-ir increases following L5/L6 nerve ligation. .54

2.4. DAB immunostaining for NF-KB-ir in spinal cords of sham-operated and L5/L6 ligated rat at day 7 55

2.5. NF-icB-ir in the dorsal hom of the lumbar spinal cord of rats with sham peration or L5/L6 nerve ligation 56

2.6. Immunoreactivity of phospho-ERKl/2 in different layers of a normal spinal cord...58

2.7. Time course for increase of p-ERKl/ERK2-ir in the dorsal hom of the lumbar spinal cord following nerve injury 59

2.8. Colocalization of p-ERKl/ERK2-ir and prodynorphin-ir in the dorsal hom of the lumbar spinal cord 60

2.9. p-CREB-ir in the dorsal hom of the lumbar spinal cord of a rat at day 7 following L5/L6 nerve ligation 61

2.10. Double labeling of PDYN-ir and p-CREB-ir in the dorsal hom of the lumbar spinal cord 62

2.11. Immunoreactivity of p-p38 MAP kinase in the lumbar spinal cord of a rat at day 7 following L5/L6 nerve ligation 64

2.12. Time course for the increase of p-p38 MAP kinase staining in the lumbar spinal cord following L5/L6 nerve ligation 65

2.13. Double immunolabeling of PKCy and PDYN in the dorsal hom of the lumbar spinal cord in rats with sham operation or L5/L6 nerve ligation 66 9

LIST OF FIGURES - Continued

2.14. C-Fos immunostaining in the sham-operated and L5/L6 ligated rats 69

2.15. Colocalization between c-Fos-ir and PDYN-ir 70

3 .1. NF-KB and AP-1 like activity in neonatal rat cortical cultures at day 4 82

3 .2. Effect of 10 NM DAMGO on the DNA binding activity of AP-1 and NF-KB 82

3 .3. Effect of chronic treatment with DAMGO on spinal dynorphin levels 83

3.4. Effect of chronic treatment with i.th. DAMGO on the DNA binding activity of AP-1 and NF-KB 85

4.1. Saturation and Rothenthal analysis of [^^'l]dynorphin A(2-17) binding to rat brain cortical membranes 100

4.2. Effects of NMDA receptor ligands on the binding of [^^'l]dynorphin A(2-17) to the brain cortical membranes 102

4.3. Dose effect of 7-chlorokynurenic acid and MK801 on the binding of 1 nM [^^'l]dynorphin A(2-17) to rat brain cortical membranes 103

5.1. Fluorescent image of representative cell that was treated with 300 ^M NMDA or 30 foM dynorphin A(2-17) 118

5.2. Time course for intracellular calcium increase induced by 100 ^M dynorphin A(2-17) or 300 ^iM NMDA 119

5.3. Dose-effect of dynorphin A(2-17) and NMDA on increase in [Ca^^i 120

5.4. Effect of 100 ^M naloxone on the increase in [Ca^^i induced by 100 ^M dynorphin A(2-17) 121

5.5. The effect of 30 nM MK801 on the increase in [Ca^^i induced by 300 nM NMDA, or by 30 jiM dynorphin A(2-17) 123

5.6. The effect of sequential application of 300 ^M NMDA, 300 ^M MK801 and 30 ^M dynorphin A(2-17) on [Ca^'^iin cortical cells 124 10

LIST OF FIGURES -Continued

5.7. Increase in [Ca^^i mediated by NMDA alone or by NMDA in the presence of 100 nM djmorphin A(2-17) in cortical cells 124

6.1. Changes in gene expression following L5/L6 ligation 137

7 .1. Regulation of dynorphin gene expression 148

7.2 Speculations on possible effects of dynorphin on neurotransmission in the dorsal horn of the spinal cord 153 11

LIST OF TABLES

2.1. C-Fos expression in different layers of the dorsal horn of the lumbar spinal cord in sham-operat^ rats or rats with spinal nerve ligation 68

4.1. Transient expression ofNRla/NR2A receptor complexes 105

5.1. Affinity of dynorphin A analogs for the human K, 5, n opioid receptors 125

6.1. Genes whose expression was increased following L5/L6 nerve ligation 135

6.2. Genes whose expression was decreased following L5/L6 nerve ligation 136 12

ABSTRACT

The hypothesis of this dissertation is that peripheral nerve injury will elevate dynorphin in the spinal cord, which has a facilitating effect on the nociceptive pathway via a non-opioid mechanism. Two major questions are addressed in this dissertation: how dynorphin is up regulated, and how dynorphin produces a facilitating effect on the nociceptive pathway. The first question was approached by examining the activation of

NF-KB, AP-1, CREB, and MAP kinase pathways following nerve injury. These experiments indicate that there was an increase in the activation of NF-KB, AP-1 and

CREB. While it was not known whether the increase in AP-1 and NF-KB activity was occuring in dynorphin-expressing cells, phospho-CREB-ir did not colocalize with prodynorphin. Thus CREB may not play a direct role in dynorphin gene expression.

Following nerve injury, there was an increase in phospho-ERKl/2, which was also seen in dynorphin-expressing cells. It is possible that the activation of ERKl/2 may contribute to the up-regulation of dynorphin in some dynorphin-expressing cells. There was also a dramatic increase in phosopho-p38-ir. However, it did not colocalize with prodynorphin.

In summary, these experiments indicated that there are profound changes in the cellular signaling pathways following nerve injury, which may play an important role in initiating neuroplastic changes in the spinal cord, including the up-regulation of dynorphin.

There is considerable evidence suggesting the involvement of NMDA receptors in the excitatory effect of dynorphin in vivo and possible interaction between dynorphin and 13

NMDA receptors. The interaction between dynorphin and NMDA receptors was further investigated by radioligand binding assay using iodinated dynorphin A(2-17). These experiments suggest dynorphin binds to NMDA receptors with relatively high affinity.

Dynorphin appears to interact preferentially with the closed state of NMDA receptors, and may be inhibitory on NMDA receptors. To look for mechanisms which may underlie the excitatory effect of dynorphin, I examined its effect on intracellular calcium level

([Ca^^]i). These experiments suggest that dynorphin A(2-17) elicited an increase in

[Ca^^i in the cortical neurons via a non-NMDA, non-opioid mechanism. This novel action of dynorphin could provide a potential mechanism for the excitatory actions of dynorphin in vivo. 14

INTRODUCTION

Dynorphin: History, Gene structure. Expression, Piiannacology

History

Dynorphin was first isolated fi'om pituitary extracts and characterized by

Goldstein and colleagues in 1979 (Goldstein et al., 1979). The potency of the peptide is

so high that they named it dynorphin, from the Greek prefix dyn-, which means strength

or power. This peptide contains the amino acid sequence for leu-enkaphalin at its amino-

terminus and was found to be 17 amino acids in length when sequencing was completed.

Shortly, Kangawa and colleagues (Kangawa and Matsuo, 1979) reported the partial

sequence of a-neo-endorphin, another highly potent, leu-enkaphalin containing opioid

peptide from hypothalamic extracts. Subsequently, several other leu-enkephalin-extended

peptides were isolated from brain, including dynorphin A 1-8 (Minamino et al., 1980), P-

neo-endorphin (Minamino et al., 1981), dynorphin B (Fischli et al., 1982; Kilpatrick et al., 1982). Since these peptides are colocalized as shown by immunohistochemical studies, it was proposed that they are derived from a common precursor molecule

(Watson et al., 1982). This proposal was confirmed by the subsequent cloning of the prodynorphin cDNA (Kakidani et al., 1982).

Structure of the prodynoqfhin gene.

The human prodynorphin (PDYN) gene (Horikawa et al., 1983) contains four exons separated by three introns (Fig): exon 1 (1.4 kb), intron A (1.2 kb), exon 2 (60b), 15 intron B (9.9 kb), exon 3 (145bp), intron C (1.7 kb), and exon 4 (2.2kb). Exons 1 and 4 contain untranslated sequences and are very large when compared to other opioid peptide genes. Rat PDYN has a similar structure, with exons 1 and 2 encoding the majority of the

5'-untranslated sequence while exons 3 and 4 contain the translated regions (Douglass et al., 1989). Southern and Northern blotting experiments show that there is only one PDYN existing in human, rat and pig. The size of the rat PDYN mRNA in brain is about 2400 nucleotides (Civelli et al., 1985). The mRNA species are 100 nucleotides smaller in adrenal cortex and in Sertoli cells of the rats (Garrett and Douglass, 1989). The difference is due to the deletion of exon 2.

Eiqtression of tfynorphin in the CNS

Hypothalamus. PDYN mRNA has been localized in the magnocellular divisions of supraoptic and paraventricular nuclei of the hypothalamus (Sherman et al., 1986).

Chronic osmotic challenge has been shown to increase PDYN mRNA levels in these areas. An increase in the levels of PDYN mRNA in the hypothalamus has also been found in rats treated with nicotine (Hollt and Horn, 1989).

Striatum. PDYN mRNA has been detected in the striatum, the nucleus accumbens, and the olfactory tubercle by in situ hybridization (Young et al., 1986). The mesostriatal DA system does not appear to exert major effect on PDYN level in this brain region, since destruction of the substantia nigra with 6-hydroxy-DA or chronic treatment with the DA antagonist haloperidol did not change the PDYN mRNA in the striatum 16

(Morris et al., 1989). In contrast, destruction of the dorsal raphe nucleus by injection of S-

7-dihydroxytryptamine caused significant reductions in PDYN mRNA levels in the medial nucleus accumbens and the caudomedial striatum ^orris et al., 1988c), suggesting the raphe-striatal serotonergic pathway tonically enhances PDYN expression in the striatal targeted neurons. Striatal level of PDYN was increased by chronic treatment of GAB A transaminase inhibitors (Reimer and HoIIt, 1991). In addition, inactivation of striatal opioid receptors by P-chlomaltrexamine caused an increase in the

PDYN expression at the site of injection (Morris et al., 1988b), suggesting an activation of the opioid receptors by endogenous opioids inhibits striatal PDYN biosynthesis. This effect seems to be mediated via the 5- and K-receptors.

Hippocampus. Dynorphins are synthesized in the granule cells of the dentate gyrus and transported along the mossy fiber pathways that innervate the pyramidal cells

(Morris et al., 1986). A brief train of high-frequency electrical stimulation to the dentate gyrus decreased the levels of PDYN mRNA on the stimulated, but not the unstimulated side (Morris et al., 1988a). Chronic electrical stimulation of the dentate gyrus also caused a marked decrease in PDYN mRNA levels (Morris et al., 1987). Electrical stimulation of the perforant pathway also induced a marked reduction in PDYN levels. Furthermore, the perforant path stimulation-induced decrease in PDYN mRNA could be blocked by the glutamate receptor antagonist, indicating the endogenous glutamate in this pathway negatively regulates the expression of PDYN gene in dentate gyrus cells (Xie et al.,

1991). Electrophysiological studies suggest an inhibitory action of dynorphin on 17 hippocampal pyramidal cells (Henriksen et al., 1982), thus the decrease in PDYN expression might contribute to the hyperexcitability found in the kindling state. In aged rats, there is an increase in hippocampal levels of PDYN expression, which may affect

LTP associated with learning and memory.

Spinal cord. PDYN mRNA in the cord increases in response to acute or chronic inflammatory processes (ladarola et al., 1986; Weihe et al., 1989). PDYN mRNA levels were elevated as early as 4h afler the injection of the inflammatory agent into the paw, peaking after 3 days and returning to normal after about 2 weeks (ladarola et al., 1988).

In situ hybridization showed that the increase in the PDYN mRNA occurs in the superficial dorsal-hom laminae I and n and Laminae V and VI (Ruda et al., 1988). The changes only occur in those spinal cord segments that received sensory inputs from the affected area. Other stimuli that increase the activity of pain-transmitting primary afferents, such as heat and nerve injury also induce the increase in PDYN mRNA

(Noguchi et al., 1991; Przewlocki et al., 1988). Since increased prodynorphin expression is preceded by an early induction of c-fos and prodynorphin gene contains a functional

AP-I element, it is conceivable to suggest that AP-1 mediates the activation of the PDYN under these conditions (Naranjo et al., 1991). However, there has been no report that addresses whether there is an increased activity of AP-1 in spinal cord in rats subjected to such stimuli.

The pharmacological actions of ^norphin 18

Analgesic effects. A classic property of opioid agonists is their ability to induce analgesia. The analgesic efficacy of dynorphin, however, is not certain. Dynorphin does not produce analgesia when injected into the mammalian brain, a finding that has been confirmed in many laboratories (Chavldn et al., 1982; Friedman et al., 1981; Walker et al., 1982b). In those few cases where analgesia has been reported, it is generally observed only with very high doses (18) or only on certain kinds of tests ^layes et al., 1983;

Nakazawa et al., 198S). In contrast, dynorphin antagonizes morphine or other opioids induced analgesia in naive animals. Adding to the confusion is the observation that this antagonist has also been observed with the des-Tyr fragments of dynorphin, which are likely metabolites of dynorphin (Walker et al., 1982c). On the other hand, dynorphin also potentiates morphine induced analgesia in morphine tolerant animals (Friedman et al.,

1981; Tulunay et al., 1981), suggesting that while not analgesic itself, the peptide can nevertheless substitute for morphine in tolerant/dependent animals. Moreover, termination of dynorphin results in no withdrawal signs (Khazan et al., 1983).

While evidence does not support a direct role of dynorphin in analgesia in the brain, the evidence for the analgesic effect of dynorphin in the spinal cord is somewhat better, yet it is still controversial. Early studies have shown that dynorphin A(l-13) is analgesic when injected intrathecally, with a potency equal to or greater than that of morphine (Goldstein et al., 1979; Han and Xie, 1984; Herman and Goldstein, 1985;

Nakazawa et al., 198S; Spampinato and Candeletti, 1985). This conclusion is compromised by the finding that high dose of i.th. dynorphin produced flaccid paralysis 19

((Stevens and Yaksh, 1986), also noted by others). When doses of dynorphin that do not induce paralysis are used, no antinociceptive activity could be detected using three analgesic tests: tail flick, hot-plate, and writhing. In fact, carefiil examination of early reports indicates that it is very difficult to dissociate the analgesic and motor effects of dynorphin in these commonly used tests. However, i.th. dynorphin, like other kappa selective compound such as U-SO, 488H does induce analgesic effect as shown by vocalization test (Spampinato and Candeletti, 1985), which presumably is not affected by motor activity.

Motor effects. Dynorphin has pronounced effects on the mammalian motor systems, as indicated in some of the earliest studies of this peptide. High doses of dynorphin induced barrel rotation in rats when injected into brain (Herrera-Marschitz et al., 1984; Kaneko et al., 1983; Petrie et al., 1982; Walker et al., 1982b). On the other hand, high doses of i.th dynorphin induce flaccid paralysis (Faden and Jacobs, 1984; Han and Xie, 1984; Herman and Goldstein, 1985). In both brain and spind cord, these effects are not blocked by naloxone and can be produced by des-Tyr fragments of dynorphin. Thus, they are produced via non-opioid sites.

Other effects. Like other opioids, dynorphin lowers blood pressure and heart rate

(Feuerstein and Faden, 1984; Gautret and Schmitt, 1985; Laurent and Schmitt, 1983) when administered into hypothalamic nuclei or other brain regions. Dynorphin also has peripheral effects on the circulatory system. It relaxes rat superior mesenteric arteries. which is blocked by naloxone (Kiang and Wei, 1984). It also inhibits neurogenic plasma

extravasation (Woo et al., 1983). Dynorphin itself has no effect on respiration, but it

enhances morphine's depression of respiration in morphine naive animals, while

attenuates morphine's action in morphine-tolerant animals (Woo et al., 1983). Dynorphin

alone has a slight hyperthermic effect, when given together with morphine, however, it

potentiates hypothermia induced by morphine (Lee, 1984). The effect of dynorphin on feeding is not certain. Several investigators have reported that dynorphin stimulates food

intake (Gosnell et al., 1986; Morley and Levine, 1983; Morley et al., 1984; Morley et al.,

1982), while others have found that dynorphin has no effect on food intake in either

normal or genetically obese rats (Hoskins and Ho, 1986). Dynorphin given i.cv or intravenously increases prolactin levels and decreases luteinizing hormone levels

(Gilbeau et al., 1986; Kinoshita et al., 1982; Matsushita et al., 1982). It also suppresses oxytocin release in lactating rats (Wright et al., 1983). These effects can be reversed by naloxone. Dynorphin is also implicated in brain injury, in particular, stroke. Baskin et al

(Baskin et al., 198S) found that treatment of stroked cats with dynorphin prolonged survival. Studies have also shown that dynorphin has anticonvulsant activity. Przwocka et al (Przewlocka et al., 1983) reported that dynorphin antagonized the convulsant effects of pentylenetetrazol. Finally, dynorphin may have modulatory effects on the immune system.

Effects on neuronal activity. The effects of dynorphin on the neuronal activity have been best characterized in the hippocampus, which can be examined in vitro as well 21 as in vivo. Both inhibitory and excitatory effects of dynorphin on spontaneous and evoked activity have been demonstrated (Chavkin et al., 1985; Henriksen et al., 1982;

Moises and Walker, 1985; Vidal et al., 1984). In most cases, the excitatory effects are similar to other opioids that act on opioid ^ and a receptor and can be blocked by naloxone. The excitatory effects of opioid on hippocampus may result from a disinhibition via inhibitory actions of opioids on inhibitory intemeurons. In contrast, the inhibitory effects appear to result from a direct action on the postsynaptic receptors and can not be blocked by naloxone. In addition, inhibitory actions are also produced by des-

Tyr fragments of dynorphin, suggesting dynorphin produces these effects via non-opioid mechanism. Very likely, they are mediated via NMDA receptors (see below).

The effects of dynorphin on other CNS areas have not been extensively studied.

Like other opioids, dynorphin inhibits the spontaneous activity in hypothalamic arcute neurons in vitro (MacMillan and Clarke, 1983). Sutor and Zieglgansberger (Sutor and

Zieglgansberger, 1984) found both depolarizing and hyperpolarizing effects, and increases and decreases in EPSPs in neocortical neurons in vitro. In this case the mechanisms by which dynorphin produces these effects were not investigated, since application of another opioid, D-Ala3-D-enkephalin (DADL), also produces hyperpolarizing effects and decreases EPSP amplitude in a naloxone dependent way, the inhibitory effects seem to be mediated via opioid receptors.

The Non-opioid Actions of Dynorphin 22

As seen above, the in vivo pharmacology of dynorphin not only encompasses the predicted set of kappa opioid efifects, but also suggests a wide range of non-opiate effects.

These include a number of the efifects of dynorphin observed in rats and mice which are completely unaffected by extremely high doses of naloxone. Most prominent among these are its effect on motor activity and grooming behavior (Herman et al., 1980; Walker et al., 1982a), changes in the electroencephalogram and its depressant effects on hippocampal cell firing (Walker et al., 1982b). The non-opioid nature of these effects is further supported by the finding that at least this subset of dynorphin effects are equally well produced by des-Tyr dynorphin, a fragment having very low affinity on the opioid receptors in the CNS (Walker et al., 1982a). Other effects of dynorphin, such as the inhibition of morphine analgesia (Walker et al., 1982a) and motor paralysis (Faden and

Jacobs, 1984) are also mimicked by des-Tyr dynorphin. Taken together, these findings suggest that in vivo pharmacology of dynorphin has two components, one which is opiate in nature and mediated through opioid receptor mechanisms and another which is non- opioid and mediated through non-opioid sites.

Interaction of tfynorphin with NMD A receptors

Several converging lines of evidence suggest the involvement of NMDA receptors in the non-opioid actions of dynorphin. Caudle and Isaac (Caudle and Isaac,

1987; Caudle and Isaac, 1988) have found that dynorphin temporarily inhibits reflexes mediated by Group I and m afferents (AP and Aa fibers), which correlates with the reversible inhibition of tail-shock vocalization, reversible hindlimb paralysis. In contrast. 23 high doses of dynorphin produced an initial potentiation of group IV reflex (C-fiber), followed by a loss of the C-fiber reflex in a time course that paralleled closely the time course of the loss of the tail-flick reflex. Since the tail-flick reflex pathway is a polysynaptic spinal reflex utilizing NMDA receptors, a simple explanation for the initial potentiation of dynorphin on the C-fiber reflex is that dynorphin activates NMDA receptors. Indeed, a NMDA receptor antagonist had just the opposite effects of dynorphin on the spinal reflexes; potentiated Group I and m reflexes while inhibited the C-fiber reflex. Furthermore, APV inhibits the dynorphin-induced behavioral effects (Caudle and

Isaac, 1988). Subsequent studies showed that other competitive and noncompetitive

NMDA receptor antagonists also blocked hindlimb paralysis induced by both dynorphin and des-Tyr dynorphin fragments (Bakshi and Faden, 1990; Shukia et al., 1992; Shukla and Lemaire, 1992). The interaction of dynorphin with NMDA receptors are further supported by the finding that dynrophin modulated the binding of several NMDA receptor ligands. In competition analysis using rat brain membranes, dynorphin and its related fragments attenuate the binding of [^H]glutamate (Massardier and Hunt, 1989) and [^H]MK801 (Shukla et al., 1992) but potentiate that of ['H]CGP39653 (Dumont and

Lemaire, 1994). This binding profile suggests that dynorphin is inhibitory on the

NMDARs. Electrophysiological studies of NMDA-mediated currents in isolated trigeminal neurons (Chen et al., 199S), or in Xenopus oocytes that expressed NMDAR subunit complexes (Brauneis et al., 1996), also indicated that dynorphin attenuates

NMDAR conductance. This inhibitory action of dynorphin on NMDA receptor is not entirely consistent with the excitatory actions of dynorphin observed in vivo. A solution 24 to this apparent paradox is the proposition that dynorphin may produce excitatory effects through a disinhibition mechanism (Chen et al., 199S). Dynorphin has also been shown to potentiate NMDAR mediated current under extremely low glycine concentration in the assay buffer (Zhang et al., 1997). However, these data were obtained under nonphysiological conditions and may not reflect the action of dynorphin in vivo.

The activation of NMDA receptors by dynorphin may also be mediated indirectly through excitatory amino acids (EAAs). I.th. injection of dynorphin was shown to cause depletion of tissue levels of glutamate and aspartate (Bakshi et al., 1990) presumably due to excess release of EAAs. Moreover, microdialysis analysis indicated that administration of dynorphin into the hippocampus (Faden, 1992) and the spinal cord (Skilling et al.,

1992) dose-dependently increased the release of EAAs. How dynorphin induces the release of EAAs is not clear. Skilling et al (Skilling et al., 1992) found that the NMDA receptor channel blocker MK801 reduced the release of EAAs induced by dynorphin, indicating the activation of NMDA receptor itself is involved in the release of EAAs.

Alternatively, the release of EAAs may be secondary to ischemia induced by dynorphin.

It was found that paralytic doses of dynorphin caused large dose-dependent reductions in spinal cord blood flow that were accompanied by 3 fold increases in the concentrations of lactic acid (Long et al., 1987) and EAAs (Long et al., 1994) in spinal cord CSF.

However, the release of EAAs does not correlate well with hindlimb paralysis and neurotoxicity in these studies. For example, naloxone reduced the release of glutamate as well as aspartate induced by dynorphin but did not block the associated development of paralysis. Similarly, a kappa opioid agonist US0488 also induced release of aspartate but 25 did not produce either reversible paralysis or irreversible paralysis due to excitotoxcity.

Moreover, doses of dynorphin that produced reversible paralysis did not induce detectable increase in the release of EAAs, indicating that EAAs are not involved in this effect of dynorphin (Skilling et al., 1992). Taken together. There is no direct evidence for the involvement of EAAs in dynorphin-induced neurotoxicity.

Other noHopioid sites bound by tfynotphin

Dynorphin was reported to displace the binding of neuropetide Y and peptide YY

&om membrane preparation of neuroblastoma cell lines SK-N-MC and SMS-MSN and may function as an antagonist on the Y1 and Y2 receptors for NPY/PYY (Miura et al.,

1994). Dynorphin also induced ACTH release in sheep and in AtT 20 cells, the latter does not express opioid receptors, thus serves as system of choice for investigation of the non-opioid effects of dynorphin. Dynoprhin A(l-13) was found to associate specifically with AtT 20 cells, through an uptake process and a binding site. It seems to bind to a fraction containing secretary vesicles, with a Kd of about 100 nM. The binding of dynorphin A (1-3) was competed by dynorphin A(l-17), des-Tyr dynorphin fragments as well as ACTH with ICSOs ranging from 200 nM to 2 jiM (Yarygin et al., 1998).

Likewise, dynorphin was demonstrated to bind to a low affinity, high capacity site in heart tissue (Dumont and Lemaire, 1993). Further studies found this binding site is oubain sensitive, suggesting that it binds to a Na+/K+ -ATPase (Dumont and Lemaire,

1996). This finding, together with early observations by Yamasaki and Way that dynorphin is inhibitory on Ca^^ pump and decrease the Ca^^ efflux in rat erythrocyte 26

Ghosts (Yamasaki and Way, 1983; Yamasaki and Way, 1985), suggests dynorphin interacts with ion pumps. Dynorphin has also been shown to acts as an antagonist on a tachykine peptide receptor (Donaldson et al., 1996). Finally, dynorphin has also found to binds to nociceptin receptor with high affinity (Makman et al., 1997). Taken together, dynorphin seems to bind to a variety of targets. In a few attempts where dynorphin is labeled directly and used as radioligand, high capacity, multiple binding sites are revealed

(Dumont and Lemaire, 1993; Yarygin et al., 1998). The Bm«x obtained usually were several hundreds of pmol per mg protein, which may not represent a single receptor.

Although dynorphin has been demonstrated to bind to these sites, the physiological significance of these binding sites have not been fiilly explored.

Mechanisms of Neuropathic Pain

Peripheral nerve injury may produce abnormal painful conditions that last long after the healing of the apparent injury. Such neuropathic pains are associated with many diseases such as diabetes, cancer, nerve compression, viral and bacterial infections.

Symptoms of neuropathic pain include spontaneous pain, allodynia and hyperalgesia.

Spontaneous pain is produced in the absence of any obvious stimuli, allodynia is the pain resulting from a non-painfiil stimulus, while hyperalgesia is characterized by increased sensitivity to noxious stimuli. Clinically, neuropathic pain has burning, lancinating and electric shock qualities.

PeriphenU changes 27

Both peripheral and central neural mechanisms contribute to neuropathic pain.

Following nerve injury, peripheral input to the central nervous system is altered.

Damaged nerves may form peripheral collateral sprouts or microneuromas, generating ectopic discharges that are perceived as nociceptive (Devor, 1991; Kajander and Bennett,

1992). Ectopic impulses can also be generated in the neuroma from DRG cells of damaged nerves (Kajander et al., 1992). The production of ectopic discharges is probably due to accumulation of sodium channels at the injured site and cell body of DRG (Devor et al., 1993). In addition, injured fibers are also more sensitive to mechano/chemical stimuli, such as pressure and noradrenaline (Devor, 1991; Sato and Perl, 1991). In the rat, ectopic discharges reach their maximum in myelinated fibers in one to two weeks, and in unmyelinated fibers in about one month. Another possibility is that the normal isolation between adjacent axons in a nerve trunk may break down, permitting low-threshold fibers to activate high threshold fibers that normally transmit noxious signals (Devor and Wall,

1990). Finally, sensory neurons in the dorsal root ganglion may be innervated by sympathetic terminals and driven by sympathetic tone, producing sympathic dependent pain (Chung et al., 1996; Roberts, 1986).

Neuropathic pain is not always associated with increased afferent input. For example, deafiferentation may result in significant increases in the spontaneous activity of dorsal horn neurons, presumably due to a loss of a tonic large fiber-mediated, indirect, inhibitory input (Basbaum and Wall, 1976; Millar et al., 1976). This may be the mechanism underlying the so called "deafferentation pain". 28

Central changes

Peripheral nerve injury also induces changes in the spinal cord. These changes include central sensitization, disinhibition, structural reorganization and alterations in glial fijnctions. There are also changes in the expression of neurotransmitters, neurotransmitter receptors and cytokines.

Central sensitization. Central sensitization is elicited by small caliber, slowly conducting c-afTerent fibers (Woolf and Wall, 1986). Low frequency C-fiber conditioning stimulus lasting tens of seconds can produce a central facilitation that lasts for minutes and hours. The facilitation is manifested as an increase in the size of the cutaneous receptive fields and responsiveness of dorsal horn neurons, and a drop in their threshold

(Cook et al., 1986; Woolf and Thompson, 1991).

The cellular mechanisms responsible for central sensitization are NMDA receptor dependent. NMDA receptor is normally blocked at resting potentials by magnesium.

When C-fiber is activated, it releases both excitatory amino acids and tachykinins.

Excitatory amino acids and tachykinins then depolarize dorsal horn neurons by the activation of non-NMDA excitatory amino acid receptors and tachykinin receptors respectively (Thompson et al, 1990; Nagy et al, 1993), removing the magnesium block and generating a large inward current. The NMDA ion channel also permits Ca^^ to enter cells, which further activates intracellular enzymes such as protein kinase C and nitric oxide synthase. PKC activation then leads to the phosphorylation of several proteins. 29

including NMD A receptor (Chen and Huang, 1992; Coderre and Melzack, 1992). The

phosphorylation of NMDA receptor can reduce the Mg^*^ block, leading to increased

glutamate sensitivity. Nitric acid has also been implicated in central hyperalgesia (Meller

et al., 1992), although where and how it acts is not clear.

Central sensitization can be produced by any C-fiber activity sufficient to produce

a summation of their slow synaptic potential such that the Mg^*^ block of NMDA receptor

can be removed. At the time of injury to a nerve, short and intense injury discharges

occur (Grubb et al, 1993), application of analgesia or NMDA receptor blockers at this time can reduce the development of spontaneous pain or local wound sensitivity

(Richmond et al, 1993). At longer time, neuroma or microneuromas may form as result of axonal regeneration. As discussed above, ectopic impulses can be generated in the

neuroma as well as from the injured DRG. In addition, injured fibers are also more sensitive to mechano/chemical stimuli, such as pressure and noradrenaline. This increased sensory inflow is sufficient to induce and maintain central sensitization.

Dishinhibiton. An increase in the excitability may also occur as the result of the removal of a tonic or phasic inhibition, i.e. disinhibition. The administration of the

GABA or glycine receptor antagonists has been shown to produce touch evoked allodynia (Yaksh, 1989) and a laige increase in the response to A3 fiber-inputs (Sivilotti and Wool^ 1994). As a resuk of the removal of tonic inhibition, dorsal horn neurons that are exclusively responsive to high threshold input can also response to low-threshold- 30 inputs. A number of observations may explain how disinhibition may occur. It was observed there was a reduction in the GABA/glycine receptors on the central terminals of injured nerves (Bhisitkul et al., 1990), Similarly, a reduction in glycine or GABA content in inhibitory intemeurons was also found (Castero-Lopes et al, 1993). Finally, associated with nerve injury, there is degeneration of the inhibitory intemeurons (Sugimoto et al.,

1989). In agreement with the disinhibition proposal, it was found that chronic peripheral nerve section diminishes the primary afferent A-fiber-mediated inhibition of dorsal hom neurons (Woolf and Doubell, 1994). Although there are morphological changes in the dorsal hom after never injury, whether they represent degenerating inhibitory intemeurons is not clear (Dubner and Ruda, 1992). If degeneration of inhibitory intemeurons does contribute to the hyperexcitability of dorsal hom neurons, then the fact that neuropathic pain in some animals dissolves spontaneously 6 weeks after nerve injury would imply either a subsequent decreased afferent input or a down-regulation of excitatory neurons in these animals.

Structural reorganization. Neuropathic pain may also be the consequence of stmctural reorganization in the dorsal hom. Peripheral nerve injury induces both atrophic and regenerative changes in primary sensory neurons that together promote a rewiring of their central synaptic connections. The atrophic changes in the DRG may include the actual degeneration of their central terminals or terminal withdrawal from synaptic sites

(Woolf et al., 1992). On the other hand, nerve injury induces regenerative changes including the expression of developmentally regulated growth-associated proteins such as 31

GAP-43 (Woolf et al., 1990). The combination of vacant synaptic sites as result of the atrophic changes and the presence of regenerative machinery may lead to sprouting of terminal branches from their nomnal sites of termination into new areas. For example, it has been demonstrated that AP afiferents that normally synapse with non-nociceptive secondary neurons in lamina m, extend to laminae I and n, and make novel synapses with nociceptive terminals (Shortland and Woolf^ 1993). The redirection of low-threshold mechanoreceptor myelinated afferent terminals into areas associated with nociception could lead to abnormal signal generation in response to innocuous stimuli. This possibility has been shown by the observation that AP-fiber stimulation begins to induce c-Fos expression in the superficial dorsal horn, which normally is only induced by C- fiber stimulation (Molander et al., 1992).

Immune activation. While the role of peripheral inflammation in producing hyperalgesia and allodynia has been well documented, the role of inflammation or immune activation in the neuropathic pain produced by peripheral nerve injury is not certain. That the chemical toxicity of chromic gut is required for the development of abnormal behaviors observed following chronic constriction injury (CCI) suggests a local, inflammatory neuritis in the etiology of neuropathic pain in this model (Maves et al., 1993). In addition, abnormal pain can be produced by endoneural injection of cytokines. It is suggested persistent existence of inflammation at the nerve injury site is one of the mechanisms underlying the neuropathic pain demonstrated by these animals.

In other animal models of neuropathic pain such as those produced by LS/L6 nerve tight 32 ligation or sciatic or L5 nerve cryoneurolysis, whether there is inflammation at the nerve injury site is less clear. Interestingly, in all models mentioned above, there is an alteration of immune activation in the spinal cord. Both nerve cryoneurolysis and CCI induce glial activation, and an increase in the production of IL-13, TNF-a, TGF-P, IL-6 bilaterally

(DeLeo et al., 1997). Recently, it was also shown that IL-6 mRNA was significantly elevated in the spinal neurons at 3 days and 7 days post spinal nerve tight ligation

(Amida et al., 1998), suggesting the existence of immune activation in this model.

The importance of immune activation in the spinal cord in neuropathic pain is further supported by the fact that inducement of inflammation in the spinal cord by i.th. injection of LPS also produces hyperalgesia. Both IL-I and TNF-a have been implicated in mediating the effect of LPS since it is abolished by the systemic administration of IL-1 antagonist IL-lra, or TNF-a antagonist, TNF-binding protein (Meller et al., 1994;

Watldns et al., 1995a; Watkins et al., 1995b). Furthermore, these abnormal pain sensations can also be produced by a single, i.th. injection of cytokines such as IL-1, IL-6

(DeLeo et al., 1996).

It remains to be established how a peripheral nerve injury or inflammation activates immune system in the CNS. It is possible that cytokines released from the injury site may be transported by retrograde axonal or nonaxonal mechanisms into the CNS and initiate immune reaaion there. A second possibility is that activation of dorsal horn neurons may induce the expression of cytokines in thes« neurons, where they are released 33 and activate glia subsequently. Finally, increased afferent input may activate glia indirectly by releasing neurotransmitters from afferent terminal or potassium ions from depolarized dorsal horn neurons. Extracellular glutamate and have been shown to activate glia.

Immune activation is also implicated in acute hyperalgesia induced by subcutaneous formalin (Watkins et al., 1997). Thus destruction of glia metabolism by glia specific metabolic specific inhibitor or specific inhibition of cytokine and NO production in glia blocks formalin-induced hyperalgesia. Formalin-induced hyperalgesia is also blocked by IL-IR antagonist or NGF antiserum. A role for IL-1 in producing hyperalgesia is also suggested by observations that neurons express IL-1 receptors, and

IL-ip has been shown to produce hyperalgesia via selective enhancement of nociceptive neuronal response to afferent pain signals (Oka et al., 1994). Similarly, NGF has been shown to potentiate NMDA receptor (Lewin et al., 1994). A role of NGF in pain modulation is further supported by the fact that primary afferents expressing high affinity

NGF receptors terminate exclusively in lamina I and the outer region of lamina n (Yip and Johnson, 1987).

Nonopioid Actions of Dynoiphin and Neuropahtic pain

Several lines of evidence suggest nonopioid actions of dynorphin may contribute to abnormal pain behaviors associated with inflammation and nerve injury. The up- regulation of dynorphin in the spinal cord has been extensively studied in rats with 34 peripheral inflammation produced by chemical methods. The injection of complete

Freud's adjuvant (CFA), or carrageenan into the hindpaw of the rat can produce an intense inflammation characterized by erythema, edema and hyperalgesia that is limited to the injected paw (Hargreaves et al., 1988; ladarola et al., 1988). In this model, the hyperalgesia peaks vsrithin 2-6 h after inflammation is induced and persists for 10-14 days. Accompanying hyperalgesia is an up-regulation of dynorphin expression in the dorsal horn of the spinal cord. RNA blot analysis revealed an inflammation-induced increase in the prodynorphin mRNA after only 4 h, with a maximal eightfold increase occurring between two and five days and returning to control levels by 10-14 days

(ladarola et al., 1988; ladarola et al., 1986). The changes in dynorphin gene expression parallel the development and time course of behavioral hyperalgesia induced by CFA. An increase in the level of dynorphin peptide itself occurs later. It is apparent by two days, and a three fold increase can be detected by 4 days after the induction of inflammation

(Millan et al., 1988; Millan et al., 1986). The increase is mostly concentrated in the medial part of the superficial laminae of the dorsal horn and the neck of the dorsal horn

(laminae V and VI). Dynorphin is found in both local circuit neurons and projection neurons (Nahin et al., 1989). It has been shown that dynorphin immunoreactivity is increased in the superficial dorsal horn following high-frequency activation of unmyelinated primary afferent fibers (Hutchison et al., 1990), and dynorphin- immunoreactive neurons receive direct monosynaptic input from unmyelinated afferents

(Morgan and Curran, 1989), thus one trigger for dynorphin expression may be the activation of C-fibers. The up-regulation of dynorphin gene expression is preceded by an 35 increase in c-fos expression in the dorsal horn that peaks within 30 min after injection of an inflammatory agent into the rat hindpaw (Draisci and ladarola, 1989). Furthermore, double-labeled studies have demonstrated the co-localization of PDYN mRNA and Fos-

IR in both the superficial laminae and neck of the dorsal horn on the side receiving input from the inflamed hindpaw (Noguchi et al., 1991). Over 80% of the neurons that express

PDYN mRNA are also c-fos-ir positive, while the number of neurons expressing c-fos is much greater than that of double-labeled neurons. Thus c-Fos signaling pathway may be coupled to dynorphin gene expression. Suppoiting this concept is the finding that an AP­

I-like binding site in the promoter region of the rat PDYN gene binds to c-Fos and c-Jun proteins and transfection experiments indicate that this site is a target of Fos/Jun trans- activation (Naranjo et al., 1991). However, it must be pointed that protein levels of c-Fos and c-Jun do not necessarily reflect their activities. For example, phosphorylation of c-

Jun proteins by c-Jun N-terminal kinase (JNK) has been established as an important pathway for the activation of AP-1. The activation of this pathway will be examined in this dissertation.

Considering the excitatory, nonopioid action of exogenous dynorphin in the spinal cord, and the parallel between PDYN mRNA expression and development of hyperalgesia after inflammation, it is tempting to assume a role of dynorphin in the production of hyperalgesia after peripheral inflammation. However, hyperalgesia seems to develop faster (peak within 2-6 h after inflammation) than up-regulation of dynorphin

(apparent increase seen by two days after inflammation). Furthermore, in rats that have 36 been treated with capsaicin neonatally, a treatment that presumably destroys almost all the unmyelinated primary afferent fibers, there is attenuated response of c-fos and dynorphin expression afler inflammation, but the amplitude of hyperalgesia is not affected (Hylden et al., 1992). Thus dynorphin expression does not appear to correlate with hyperalgesia very well in this model, probably because inflammation-induced hyperalgesia also has peripheral components.

Spinal dynorphin is also elevated in response to peripheral nerve injury (Draisci et al., 1991; Kajander et al., 1990). In a rat model of neuropathic pain produced by LS/L6 ligation, there is dramatic increase in the spinal dynorphin expression that correlates with the abnormal pain behaviors in time and extent (Malan et al, 2000). The significance of the increased dynorphin expression in spinal cord is demonstrated by the finding that i.th. antisera to dynorphin have the same profile of actions as the NMDAR antagonist, MK-

801, in blocking thermal hyperalgesia and restoring the efficacy of morphine against ailodynia brought upon by nerve-ligation injury (Nichols et al., 1997). These findings suggest that the elevated level of dynorphin may contribute to the hyperesthetic states, possibly through modulating NMDAR function in the maintenance of neuropathic pain.

These findings are consistent with the nonopioid actions of dynorphin described above, and provide the first direct evidence suggesting a role of endogenous dynorphin in facilitating nociception or producing central sensitization. Taken together. There is good correlation between spinal dynorphin level and facilitation of the pain sensory pathway, though how dynorphin produces this effect has not been established definitively. 37

Summary of Research Goals

As reviewed above, although dynorphin was isolated as an endogenous opioid peptide, its behavior is very different from other opioid peptides. Opioid peptides act on opioid receptors and usually are inhibitory on neurotransmission or neuronal excitability, but dynorphin tends to have excitatory effects in vivo which are not mediated through opioid receptors. The physiological implications of these so-called nonopioid actions of dynorphin are particularly relevant to abnormal pain behaviors associated with peripheral nerve injury and opioid tolerance. In both conditions there is an elevated level of spinal dynorphin, and dynorphin antisera alleviate associated abnormal pain syndromes. These fmdings suggest dynorphin may play an important role in the etiology of abnormal pain syndromes in these animal models, yet the mechanism of this action is still a mystery.

The fact that NMDA receptor antagonists block the effect of dynorphin suggests the involvement of the excitatory amino acid system in mediating the effect of dynorphin.

While there are a few reports showing dynorphin enhances NMDA response, most electrophysiological experiments with native NMDA receptors in isolated neurons or cloned NMDA receptors expressed in oocytes suggest a direct inhibitory action of dynorphin on NMDA receptors. This inhibitory action of dynorphin on NMDA receptors is further supported by radioligand experiments in which dynorphin or des-tyr dynorphin has been shown to attenuate the binding of agonists but enhance that of antagonists to

NMDA receptors. Taken together, it appears there is contradiction between results derived from in vivo and those from in vitro experiments. To resolve this issue, this 38 dissertation addresses two major questions; How dynorphin is up regulated and how it may produce excitatory effects in vivo. Chapter II characterizes the molecular mechanism that lead to up-regulation of dsmorphin in rats with nerve injury by studying the DNA binding activity of transcriptional factors AP-I, NF-K-B, CREB, as well as activation of

MAP kinase pathways. Since neuropathic pain and opioid (morphine) tolerance appear to share many common aspects and dynorphin has been implicated both in neuropathic pain and opioid tolerance (see chapter M), the activation of AP-1 and NF-KB is also studied in animals chronically treated with an opioid agonist DAMGO. Chapter IV and V address potential mechanisms by which dynorphin produces excitatory effects. In chapter IV, the binding site(s) for the des-tyr fragment of dynorphin A(l-17) (Dynorphin A(2-17)) is characterized with radioiodinated dynorphin A(2-17) with emphasis on its interaction with the NMDA receptors. In chapter V, the effect of dynorphin on the intracellular calcium level is examined, as elevated intracellular calcium usually accompanies neuronal excitation. Chapter VI looks into gene expression in the spinal cord of nerve- ligated rats using cDNA arrays. 39

II. REGULATION OF DYNORPHIN EXPRESSION IN RATS WITH

NERVE INJURY

Introduction

Peripheral nerve injuries have been shown to induce dramatic central changes that

may exacerbate pain. Some of these changes may occur at the level of gene expression.

After peripheral nerve injury, there is rapid increase in the expression of cytokines

(DeLeo et al., 1997), dynorphin (Draisci et al., 1991; Kajander et al., 1990), and

substance P receptors (Maimberg et al., 1997). As reviewed above, both cytokines and

dynorphin may play important roles in initiating and maintaining the painful states due to

peripheral nerve injury. Despite the significance of these changes, the molecular

mechanisms that lead to these changes have not been studied extensively.

One trigger for these CNS changes may be the increased afferent input brought

about by nerve injury. The excitatory amino acids and other neurotransmitters released from nerve tenninals may then regulate gene expression in the dorsal horn neurons by

activating intracellular signal transduction pathways or altering intracellular ions,

particularly, calcium. Additionally, increased release of EAAs from nerve terminals or

ions from depolarized dorsal horn neurons may also activate and afifect gene expression

in glial cells. The activation of these pathways then leads to changes in gene expression 40 by regulating the activity of transcription factors. Here, I will focus on AP-1, NF-KB and

CREB, since PDYN gene contains regulatory elements that bind to AP-1, NF-KB and

CREB, and stimuli that induce PDYN expression also tend to activate AP-1, NF-KB and

CREB (see below).

AP-1 is a family of transcription factors consisting of homodimers and heterodimers of Jun (v-Jun, c-Jun, JunB, JunD), Fos (v-Fos, c-Fos, FosB, Fral, Fra2) or activating transcription factor (ATF2, ATF3, B-ATF) bZIP (basic region leucine zipper) proteins. The regulation of AP-1 activity can be regulated at two major levels: abundance and activity.

A variety of stimuli induce c-Fos expression in the nervous system, among them are generalized seizures, tactile stimulation, cortical stimulation, light stimulation, dehydration in the mammalian CNS, peripheral nerve damage and inflammation, chronic drug treatment. In the CNS, c-Fos expression has been widely accepted as the index of cellular activation. Proximal to the c-fos TATA box is a CRE (cyclic AMP responsive elements) that is likely to be occupied by CREB or ATF proteins, which can mediate c- fos induction via cAMP- and Ca^^-dependent pathway (Sheng et al., 1991). Another element that regulates c-fos transcription is a sis inducible enhancer (SIE) which is recognized by the STAT (signal transducer and activator of transcription) group of transcription factors. The STATs are further activated and translocated to the nucleus in response to signals that activate JAK group of tyrosine kinases (Darnell et al., 1994). A 41 third cis element in the c-fos promoter region is the serum-responsive element (SRE), which is the target of Elk-1. Elk-1 can be further phosphorylated by ERKs (extracellular signal regulated kinase) (Treisman, 1994), the JNKs (Jun amino-terminal kinases) and p38 (Treisman, 1992).

C-Jun is a central component of all AP-1 complexes. The c-jun expression is also elevated in response to many stimuli, including growth factors, cytokines and UV irradiation (Karin, 199S). Induction is mediated through the c-jun TRE, which is recognized by c-Jun-ATF2 heterodimers. The ability of c-Jun to activate transcription is enhanced by its phosphorylation at serine 73 and serine 63, which is carried out by JNK and p38. JNK and p38 in turn is activated by the exposure of cells to proinflammatory cytokines, growth faaors and UV irradiation (Karin et al., 1997).

In rats with peripheral inflammation, there is an increase in the level of spinal dynorphin that is preceded by increased expression of c-Fos. It is proposed that up- regulation of c-Fos is responsible for the up-regulation of dynorphin induced by inflammation (Dubner and Ruda, 1992). Since the activity of AP-1 is not only regulated through abundance, but also phosphorylation, a better support for this proposal should be the correlation between AP-1 activity and dynorphin expression. In this study, the expression of dynorphin and DNA binding activity of AP-1 will be compared in rats with nerve injury. 42

NF-KB is a nuclear factor that, when activated, binds to a 10 bp sequence in the

enhancer region of the gene encoding K light chain of antibody molecules in B cells (Sen

and Baltimore, 1986). It was later found this factor is widely expressed and regulates the

expression of a variety of genes, most of which encode proteins important in immunity

and inflammation (Baeuerle and Henkel, 1994). In unstimulated cells, NF-tcB exists in a

latent form, as a complex with an inhibitory protein IKB. Upon activation, IKB is

phosphorylated and then ubiquitinated and degraded by the proteasome. This allows NF-

KB to translocate to the nucleus and binds to the corresponding enhancer sequence,

leading to an increase in the expression of the target gene.

A range of stimuli also activate NF-KB in glial and neuronal cells of the CNS, including nerve growth factor (NGF), glutamate, depolarization, cytokines ILl and TNF, phorbol ester, re-oxygenation, neurotoxic peptide (A3) and oxidative stress. Genes in the

CNS induced by NF-KB activation include cytokines (IL6, ILl and TNF), enzymes

(COX-2 and inducible nitrogen oxide synthetase), adhesion molecules (VCAM-1 and

ICAM-1, MHC class I), and interferon. Therefore, as is the case in the periphery, NF-KB could play a role in immune and inflammatory responses in the CNS (O'Neill and

Kaltschmidt, 1997).

NF-KB activation is associated with CNS injury, certain viral infection and neurodegenerative diseases such as multiple sclerosis and Alzheimer's disease. For 43

example, an increase in NF-KB was observed in the ischemic cortex (Clemens et al.,

1997; Salminen et al., 1995). In a study with cortical impact model of traumatic brain

injury, a long lasting increase in NF-KB was observed in the cerebral cortex ipsilateral to

the site of injury (Yang et al., 199S). Likewise, following traumatic spinal cord injury,

there was significant increase in the NF-KB activity, and activated NF-KB was seen in

macrophage/microglia as well as neurons (Bethea et al., 1998). An increase in NF-KB

was also observed in the cortex and hippocampus following seizure induced by kainate,

pilocarpine (Unlap and Jope, 199S). NF-KB was also found to be activated in regions around early plaque stages in Alzheimer's disease (Kaltschmidt et al., 1997).

Comparison of activation of NF-KB, AP-1 and the up-regulation of dynorphin expression suggests that the three events occur under some stressful conditions such as injury, inflammation and neuronal excitotoxicity. This chapter tests whether there is a link between the activation of AP-1 and NF-KB and the expression of PDYN. In addition, since the PDYN gene also contains regulatory elements that bind to CREB, the activation of CREB and PDYN expression were also compared. The working hypothesis is that increased afferent input induces hyper-excitation of dorsal horn neurons and subsequently the activation of a series of intracellular signal pathways, particularly MAP kinase pathways, which further leads to the activation of transcription factors such as AP-1, NF-

KB, CREB and uhimately increased expression of PDYN. To test this hypothesis, AP-L and NF-KB activation was measured by electrophoresis mobility assay and the time courses of their activation is compared with that of dynorphin expression following 44

L5/L6 nerve ligation, phosphorylation of CREB was measured with immunohistochemistry. The activation of MAP kinase pathways such as the activation of

ERK1/ERK2, p38, c-Jun-N terminal kinase was studied by immunohistochemistry with antibodies against their phosphorylated, activated forms. 45

Materials and Methods

Antibodies

Antibodies against phospho-MAPK (ERKl/2), phospho-p38 MAPK, phospho-

MAPK/JNK, phospho-CREB were obtained from New England Biolabs (Beverly, MA).

Anti-c-Fos antibody was from Calbiochem (La Jolla, California). Anti-NF-icB was obtained from Boehringer Mannheim (Indianapolis, IN), which recognizes an epitope overlapping the nuclear location signal of the p65 subunit, thus stains selectively the released, activated form of NF-KB. All these antibodies have been characterized extensively by their manufacturers and other researchers.

Animals and surgeries

All experiments were approved by the Animal Care and Use Committee of the

University of Arizona. Male Sprague-Dawley rats (200-300 g) were used in all experiments. L5/L6 ligated rats were prepared according to the method described by Kim and Chung (Kim and Chung, 1992). Rats were anesthetized with halothane administered

'to effect'. After surgical preparation and exposure of the dorsal vertebral column, the L5 and L6 spinal nerves were exposed and tightly ligated with 6-0 silk suture distal to the dorsal root ganglion. For sham surgeries, the nerves were exposed but not ligated. The incisions were closed and the animals were allowed to recover. Tactile allodynia was determined by measuring paw withdrawal in response to probing with a series of calibrated fine filaments (von Frey filaments). The strength of the von Frey stimuli range 46

from 0.4 to 15 g. Withdrawal threshold was determined by increasing and decreasing

stimulus strength eliciting paw withdrawal. All nerve-ligated rats were verified to be

allodynic (responding to a stimulus of less than S g) prior to harvesting tissue. To harvest

tissues, spinal cords were removed by applying hydraulic pressure via an 18-gauge

needle and SO-cc syringe filled with saline. Dorsal quadrants of the lumber region of each

spinal cord were isolated and frozen on dry ice. Tissues were stored at -80°C until use.

Electrophoresis mobility shift experiments

Preparation of rmclear extracts: Dorsal horn quadrants were homogenized 20 to

30 strokes using a glass homogenizer in 0.5 ml of ice cold HEGD buffer (25 mM HEPES,

1.5 mM EDTA, 10% (v/v) glycerol, 1 mM DTT, 0.1 mg/ml PMSF). Homogenates were

then centrifuged at 12000 rpm for 15 min in a bench top centrifuger. The supernatant was

then removed completely and the pellet was resuspended in 40 ^1 of HEGDK buffer

(HEGD buffer plus 0.5 M KCl). The suspension that contains nuclei was then incubated on ice for 1 hr and centrifuged at 12000 rpm for 15 min. The protein content of nuclear extract in the supernatant was then determined by the method of Bradford (Bradford,

1976).

Labeling of probes. The double-stranded oligonucleotides containing the NF-KB consensus sequence (5'-AGT TGA GGG GAC TTT CCC AGG C-3'), or AP-1 consensus sequence (5'-CGC TTG ATG AGT CAG CCG GAA-3) were from Promega

(Madison, WI). The oligonucleotides were end-labeled with y-[^^P]ATP (6000 Ci/mmol, 47

NEN) and T4 polynucleotide kinase (Promega) and purified on a G-25 column

(Boehringer Mannheim, Indianapolis, IN)-

Gel shift assay. Nuclear extracts (3-6 ^g protein in S ^1 HEGDK buffer) were incubated for 30 min at RT with 10 fmol of probe in 20 ^1 of O.OS ^g/^1 dIdC, 10% glycerol, 1 mM DTT. DNA-protein complexes were then resolved on a 5% nondenaturing polyacrylamide mini-gel at 30 mA for about 30 min in IX TBE. Gels were vacuum-dried and exposed to Kodax x-ray films overnight. For quantitative analysis, gels were visualized and the radioactivity of each band was counted using an Instant Imager

(Hewlet Parkard). The radioactivity of each band was normalized to that of control sample using the following equation: (sample radioactivity - background) / (control radioactivity - background).

Tissue fixation

Rats were deeply anesthetized with halothane and initially perfused with phosphate buffered saline (PBS) for 5 min, followed by perfusion with 4% paraformaldehyde in phosphate buffer for 20 min. Spinal cords were removed and post- fixed in PBS containing 4% formaldehyde overnight at 4°C. Tissues were then immersed in 30% sucrose at 4°C. Cross- or longitudinal sections of 40 jim were then obtained using a cryostat.

DAB immunostcdning 48

Sections were washed 2X with PBS and treated with 1% H2O2 for 30 min. After blocking in PBS containing S% BSA for 1 hrs, sections were incubated with primary antibody in PBS containing 2.5 % BSA, OA % triton X-100 overnight at room temperature. Antibody for prodynorphin (a gift from DR. Robert Elde, University of

Minnesota, Minneapolis) was raised in guinea pig (GP1821) and used at a dilution of

1 ;20,000. Antibody for NF-KB was a monoclonal antibody that recognizes an epitope overlapping the nuclear localization signal of the P6S subunit and therefore selectively stains the activated form of NF-K-B. It was used at a dilution of 1:1000. Sections were then washed 2X 15 min with PBS. Specific antibody binding was visualized by a commercial kit (Vector Elite ABC kit. Vector Laboratories, Burlingame, CA) until a brown product was observed. Sections were mounted on glass slide, dried using a series of concentrations of alcohol and xylene, and covered with Permount and coverslips.

Double fluorescent immunostaining

Floating sections were blocked in TBST (50 mM Tris-Cl, pH7.4, 150 mM NaCl,

01% Triton X-100) containing 0.1 % BSA at room temperature for 1 hr, then incubated with primary antibodies at appropriate dilutions in TBST/3 % BSA overnight at For control experiments, the primary antibodies were omitted at this step. After 3X 15 min washes with TBST/0.1 % BSA, sections were incubated with appropriate secondary antibodies at room temperature for at least 2 hrs. For prodynorphin (PDYN) staining, the secondary antibody was Texa-Red labeled anti-guinea pig IgG (Vector Laboratories,

Burlington, CA) diluted to 1:2000. The other secondary antibody was a biotinylated anti- 49

mouse or rabbit IgG used at a dilution of 1:1000 and was further visualized with fluorescein-Iabeied streptavidin (diluted to 1:2000) (Vector Laboratories, Burlington,

CA). After washing with TBST/0.1% BSA, sections were mounted on gelatin-coated slides and covered with Vectoshield (Vector Laboratories, Burlington, CA) and coverslips. The slides were observed with a Nikon Eclipse E800 microscope. For colocalization of two antigens, pictures of the same field were taken under appropriate filter cubes, then superimposed using software Metamorph (Universal Imaging

Corporation).

Controls for immuHohistochemistry

Since all antibodies used in this section have been extensively characterized by their manufacturers and other researchers, we have not attempted to characterize their specificity. However, in most experiments, we included a group of sections that were treated in the same way as experimental groups except that the primary antibodies were omitted. In these control sections, there was no apparent, specific fluorescent staining.

Statistics

Statistical significance is determined by unpaired t-test at 95% confidence level. 50

Results

Gel shift experiments

There were three protein complexes that bind to NF-KB oligonucleotides, all of which could be competed by excess cold NF-KB oligonucleotides. Only the bottom band seemed to be specific, since the upper two bands could also be competed by excess cold

OCT oligonucleotides, which are different from NF-KB oligonucleotides in sequence.

Similarly, there were two protein complexes that bind to AP-1 oligonucleotides in the spinal cord, both of which could be competed by excess cold AP-1 oligonucleotides. The upper band appeared to be AP-1 specific, since it was not competed by OCT oligonucleotides, while the bottom band was competed by OCT oligonucleotides (Figure

2.1). L5/L6 ligation significantly increased the NF-KB and AP-1 binding activity in the ipsilateral dorsal quadrants. At day 3 following nerve ligation, the earliest time at which allodynia can be reliably tested, there was about a 2 fold increase in both NF-KB and AP-

1 binding activity. The peak increase was detected at day 7. By day 11, both NF-KB and

AP-1 activity returned to control level (Figure 2.2).

DAB immunostaining studies

PDYN and activated form of NF-KB. TO correlate dynorphin expression with the activation of NF-KB, the immunostaining of PDYN and that of activated NF-KB were compared. PDYN inununoreactivity was detected in laminae I/II, laminae IV/V as well as 51 the dorsal commissure of central canal of the spinal cord. After L5/L6 ligation, increased staining was seen in the corresponding areas of the ipsilateral side of spinal cord at day 4, with peak at day 10 and then returned to control level gradually (Figure 2.3). Using an antibody that selectively stains activated NF-K-B, motor neurons appeared to have intense staining for activated NF-KB while the superficial layers and laminae IV and V stained lightly. There was no staining in control sections when the anti-NF-ICB antibody was omitted. After nerve ligation, no apparent change was observed in the motor neurons

(Figure 2.4). Since the staining in the dorsal horn was very weak, it was difficult to determine whether it had similar pattern of distribution as PDYN or there was an increase in the immunostaining for activated form of NF-KB (Figure 2.5). Thus our attempt to correlate dynorphin expression with the activation of NF-KB spatially by immunostaining was not successful. 52

AP-l NF-icB u

Z j n o ?5 o. 3 A. On

Figure 2.1. AP-l and NF-KB activity in the nuclear extracts of dorsal horn of the spinal cord. Arrows indicate specific bands that can be competed by excess of homologous but not heterologous cold oligonucleotides (OCT). 53

•NF-KB 20 400 , ^ AP-1 0) 15- C 300. o O p o O ^ 1 200. tf c 0) |i 2 Q^ SQ. S»• CD 100. 0.

10 20 30 40 50 60 70 3d 7d 11 d Days after ligation

NF-zcB

AP-l

Sham 3d 7d lid

Figure 2.2. Time course for changes in the DNA binding activity of AP-1 and NF-KB (left) following L5/L6 nerve ligation. Data were expressed as percentage of control level in animals that received sham surgery and represented means of 5 to 7 samples. The right panel showed changes in spinal dynorphin content measured by immunoassay following L5/L6 nerve ligation. 54

naive

30(ltpm

Figure 2.3. Time course of spinal prodynorphin-ir increases following L5/L6 nerve ligation. The increase was seen in the ipsilateral side (left) of the spinal cord within 3 days following ligation, peaked at day 10 and returned to normal level by day 30. Staining was mostly seen in laminae I, n, IV, V, as well as areas around the central canal. At day 20, while the staining in neuronal bodies in deeper lamina returned to normal levels, dense staining remained in fibers (Courtesy of Ruizhong Wang). 55

Figure 2.4. DAB immunostaining demonstrated NF-jcB-ir in the spinal cords of sham operated (A), and L5/L6 ligated rat (B) at day 7 following surgery. The strongest staining was seen in motor neurons in the ventral part of the spinal cord. Panel C showed a micrograph of the spinal cord when the primary antibody was omitted. By visional observations, no apparent changes were observed between sham and ligated rats as well as between the ipsilateral (right) and contralateral (left) sides of the spinal cord of nerve ligated rat. 56

Figure 2.5. Photomicrographs showing NF-KB-ir immunostaining in the dorsal horn of the lumbar spinal cord of rats with sham operation or L5/L6 nerve ligation. Staining was much weaker compared with staining in motor neurons. There was no apparent difference between sham operated and nerve ligated rats as well as between the ipsilateral (right) and contralateral sides (left) of the spinal cord of the nerve ligated rats. 57

Colocalization studies

PDYN and phospho-ERKl/ERK2. Phospho-ERKl/ERK2 immunoreativity was

detected with both DAB staining and fluorescent staining. It was detected in discrete cells

of laminae I, II, laminae IV and V as well as large cells around the central canal of the

spinal cord. Both cell bodies as well as projections were stained (Figure 2.6). Upon nerve

injury, there were more p-ERKl/2-ir positive cells in the ipsilateral side than in the

contralateral side (Figure 2.7). The increase is mostly seen in laminae I and n. It was apparent at day 4 and day 7 following nerve ligation, and appeared to return to normal level by day 11. A significant percent of cells in laminae I-II were stained with both

PDYN immunoreactivity and phospho-ERKl/ERK2 immunoreactivity (Figure 2.8). It appeared that these cells are one subset of PDYN-ir positive cells that have smaller size than those stained with PDYN-ir only.

Phospho-CREB and PDYN. Phospho-CREB immunoreactivity was detected throughout all the layers of gray matter as well as white matter of the spinal cord. The immunoreactivity appeared to be located in the nuclei, which should be further confirmed with a fluorescent dye that stains only the nuclei. Two types of nuclei were observed, one was smaller and widely distributed in the white as well gray matters, while the other was larger and seen only in the gray matter. After nerve ligation, there was increased staining in the ipsilateral side of the spinal cord, this increase was seen for both small and large

nuclei (Figure 2.9). Although phospho-CREB staining was seen throughout the spinal 58

cord, there was no significant colocalization of phospho-CREB staining and PDYN

staining in cells (Figure 2.10). Thus CREB may not be involved in the regulation of

PDYN expression.

Figure 2.6. Immunoreactivity of phospho-ERKI/2 in different layers of a normal spinal cord. The staining was seen in cells of laminae I, II (A), FV, V (B), as well as areas around the central canal (C). Cells in the deeper layers appeared to be larger than those in the superficial layers. The staining appears to be in both nuclei and cytosol. 59

sham

Figure 2.7. Time course for the increase of ERKl/ERK2-ir in the dorsal horn of the lumbar spinal cord following nerve injury. At day 4 following nerve ligation, there was an increase in the number of phospho-ERKI/2-immunoreactive neurons in the superficial dorsal horn of the ipsilateral side (left column) in comparison with sham-operated control or the contralateral side (right colunm) of the same animals. 60

PDYN

PDYN^p ERK1 2

Figure 2.8. Double immunostaining demonstrated colocalization of p-ERKl/ERK2-ir and prodynorphin-ir in the dorsal horn of the lumbar spinal cord of a rat at day 7 following L5/L6 nerve ligation. Arrows in the left colunm show 4 small cells that have both p- ERKl/2-ir and PDYN-ir while arrowhead shows a large neuron that has only PDYN-ir in the superficial dorsal horn. The right column shows a small neuron with staining of both antigens (arrow) and a large neuron with staining of only p-ERKl/2-ir (arrowhead). 61

Figure 2.9. Micrographs showing p-CREB immunoreactivity in the dorsal horn of the lumbar spinal cord of a rat at day 7 following L5/L6 nerve ligation. The nuclear staining is seen in both white and gray matters. While small nuclei are seen throughout the section, big nuclei are only seen in the gray matter. There is an increase in the number of p-CREB-ir cells in the ipsilateral side. Figure 2.10. Micrographs demonstrating double labeling of PDYN-ir and p-CREB-ir in the dorsal horn of the lumbar spinal cord of a rat at day 7 following L5/L6 nerve ligation. There was no colocalization of these two antigens. 63

Phospho-p38 and PDYN. The immunostaining of activated, phosphorylated form of p38 was detected throughout the spinal cord. The staining was homogenous and was observed in both white and gray mater, thus may represent glial cells. Nevertheless, the identity of phospho-p38-ir positive cells needs to be further determined using markers specific for neurons or glial cells. After L5/L6 nerve ligation, there was a dramatic increase in both the intensity as well as the numbers of immunostained cells in the ipsilateral side in comparison with the contralateral side of the same spinal cord, or the spinal cord of sham-operated rat (Figure 2.11). This increase started to attenuate by day

11, the longest time-point tested in this experiment (Figure 2.12). There was no colocalization between phospho-p38 and PDYN staining (data not shown).

PDYN and PKCy. Intense, nuclear PKCy staining was detected in inner laminae n. By visional observation, there was no apparent change in the intensity as well as numbers of immunoreactive cells after nerve ligation (Figure 2.13). In addition, there was no colocalization between PKCy and PDYN staining. In the superficial layers of the dorsal horn of the spinal cord, PDYN neuronal staining appeared to localize in laminae I rather than laminae n (Figure 2.13). 64

Figure 2.11 Inununoreactivity of p-p38 MAP kinase in the lumbar spinal cord of a rat at day 7 following L5/L6 nerve ligation. The staining was seen throughout in both gray and white maters. There was an increase in the number of labeled cells in the ipsilateral side (B, D) compared with the contralateral side (A, C). This ipsilateral increase was seen in the dorsal horn (A, B), ventral horn (C, D) as well as areas around the central canal (E). Panel F shows two labeled cells at high magnification. 65

sham

IB

Figure 2.12. Time course for the increase of p-p38 MAP kinase staining in the lumbar spinal cord folloAYing L5/L6 nerve ligation. There was a dramatic increase in the number of labeled cells in the ipsilateral side (right column) compared with the contralateral side (left column) at day 4, 7, as well as day 11. The peak appeared to be on day 4. 66

Figure 2.13. Micrographs showing double immunolabeling of PKCy (green) and PDYN (red) in the dorsal horn of the lumbar spinal cord. Intense staining for PKCy was seen in inner laminae n, with a few labeled cells in laminae HI. Following nerve ligation, there was no apparent increase in the PKCy-ir in the ipsilateral side (D) of the spinal cord in comparison with the contralateral side (C) or sham>operated rats (A, B). There was also absolutely no colocaiization between these two antigens. 67

PDYN and c-Fos. C-Fos has been suggested to play a role in the up-regulation of

dynorphin and one way to increase the AP-1 activity is the up-regulation of c-fos gene

expression. We wondered if c-Fos is elevated following nerve injury. Nuclear c-Fos

staining was observed in laminae I, II, m, IV, V and occasionally in deeper layers. By

counting cells labeled with c-Fos staining from the digital images, we could not detect a

significant difference in the number of cells labeled with c-Fos staining between the

ipsilateral and contralateral sides (Table 2.1 and Figure 2.14). This result should be

further confirmed using more quantitative approaches such as Western blot. However, c-

Fos did colocalize with prodynorphin. Colocalization was mostly observed in lamina IV

or V (Figure 2.15). While most PDYN cells in the superficial layer were not stained with

c-Fos (data not shown). Since a significant number of dynorphin-expressing cells did not

have visible c-Fos staining, and there was no apparent increase in the c-Fos staining, the

coupling of c-Fos signaling pathway to dynorphin gene expression is questionable in this

model.

Phospho-c-Jun andphospho-ATF-2. There was no obvious, specific staining for

antibodies against these two antigens. Since no positive controls were used in these

experiments, it is not known whether it was due to technical problems or they were not activated in the spinal cord following nerve injury. 68

Phospho-JNK. There was no obvious, specific staining for this antibody with

trials using dilutions from 1:250 to 1:40, OOO. This part of experiment is inconclusive.

Table 2.1. C-Fos expression in differem layers of the dorsal horn of the lumbar spinal cord in sham-operat^ rats or rats with spinal nerve ligation (SNL). Data represent mean ± S.E.M. of 4 to 6 sections from 2 to 3 rats. Positive cells were determined by the staining intensity (usually more than 2 fold increase over background) and morphology (discrete nuclear staining). Layers were estimated approximately.

Sham' 4dSNL 7dSNL lid SNL contralateral Laminae I/U 4.811.1 5.8 ±2.0 5.5 ±0.6 1.5 ±0.5 Laminae m 7.0 ±0.8 5 .5 ±0.9 6.4 ± 1.9 3.7 ± 1.7 Laminae TVfW 4.5 ±1.1 5.0 ±1.3 5.8 ± 1.4 4.3 ±0.8

ipsilateral Laminae I/II 4.8 ±1.1 5.5 ±0.6 3.0 ±0.9 2.5 ±0.5 Laminae HE 7.0 ±0.8 7.5 ± 1.6 8.8 ±0.7 3.5 ±0.5 Laminae IVA/^ 4.5 ±1.1 4.4 ±0.6 4.4 ±0.6 4.5 ±0.6

# Since there is no ipsi or contralateral in sham-operated rats, the numbers for sham- operated rats represent the mean number of cells from the right and left sides of the spinal cords. 69

sham

11 d

2 0 0 m

Figure 2.14. Fluorescent micrographs illustrating c-Fos immunostaining in the dorsal hom of sham-operated and L5/L6 Iigated rats. There is no significant difference in the number of cells stained with c-Fos-ir between the ipsilateral side (right column) and contralateral side (left column) in nerve Iigated rats. PDYN

t Fos * PDYN

2 00 in

Figure 2.15, Fluorescent double immunostaining illustrating the colocalization between c-Fos-ir and prodynorphin-ir. Shown is laminae IV or V of the ipsilateral side of dorsal hom of a rat at day 7 following nerve ligation. More cells were stained with c-Fos-ir than with prodynorphin-ir. 71

Discussion

One major finding of experiments in this section is that upon peripheral nerve

injury, there is a significant increase in the DNA binding activity of transcriptional

factors AP-1 and NF-K-B in the corresponding spinal cord segments. The increase is

obvious at day 4 and peaks at day 7, then returns to normal level by day 11. Since

activation of both transcription factors is associated with immune activation, cellular

stress, thus peripheral nerve injury may induce these changes in the spinal cord. This

proposition is consistent with some earlier findings that cytokines are elevated in the

spinal cord upon nerve injury (Arruda et al., 1998; DeLeo et al., 1997). How peripheral

nerve injury induce these changes is not clear. Although it has been proposed that

cytokines or other factors involved in inflammation may diffuse fi'om the injured site to

the spinal cord (DeLeo et al., 1996), thus initiate AP-1 and NF-KB activation in the spinal cord, there is no experimental evidence for this. Alternatively, the activation of AP-1 and

NF-KB is due to increased afferent input associated with nerve injury. In this regard, the activation of AP-1 and NF-KB in the nervous system has been linked to neuronal excitation and activation of excitatory amino acid receptors (Guerrini et al., 199S).

Dynorphin level is also increased upon nerve injury or dorsal rhizhotomy. In rats with LS/L6 ligation, the increase is apparent at day 3 and peaks at day 10, then returns to

normal level at day 20. It appears that activation of AP-1 and NF-KB precedes the increased expression of dynorphin, which is consistent with our hypothesis that activation 72 of AP-1 and NF-kB may contribute to the up-regulation of dynorphin. If AP-1 and NF-

KB do contribute directly to the up-regulation of dynorphin, then they should be localized in dynorphin-expressing neurons. To confirm this, we examined the distribution of dynorphin with those of AP-1 and NF-KB components. Since these two transcriptional factors are rather ubiquitous, and their amount does not necessarily correspond to their activity, we are particularly interested in the distribution of their activated forms. C-Jun and ATF2 are two potential components that have been shown to bind to AP-1 element, and both are activated as result of phosphorylation (Karin et al., 1997). However, when the spinal cord sections were stained with antibodies that recognize the phosphorylated form of c-Jun or ATF-2, no obvious immunoreactivity was observed in sections from either sham-operated or LS/L6 ligated rats. It is not clear whether this is due to the technical difHculty (i.e. these antibodies do not work for inununohistochemistry in our hands) or this suggests that these two transcription factors are not activated in the spinal cord as a result of nerve injury. Likewise, an antibody that recognizes the activated form of NF-KB also only gave weak immunostaining in the dorsal horn of the spinal cord.

Thus our attempt to extend the results of our gel shift experiments was not successfiil.

In the promoter region of dynorphin gene, there are also several cAMP responsive elements (CRE). One of the proteins that bind to CRE is CREB (CRE binding protein). In response to elevated cAMP as well as intracellular calcium (Gonzalez and Montminy,

1989; Sheng et al., 1991), CREB is phosphorylated and then activates transcription. The ability of CREB to respond to intracellular calcium makes it a potential mediator between 73 neural activity and gene expression. Indeed, CREB is activated in many cells throughout the spinal cord section and there is an elevated level of the activated form of CREB in the ipsilateral side of the spinal cord. The identities of cells stained with phospho-CREB are not certain at this time and remain to be determined. However, based on their distribution, it appears that CREB is activated in both neurons and glial cells. While its activation in neurons may be coupled to neuronal activity, its activation in glial cells is supposed to be indirect and probably due to the sustained activation of neurons. In this regard, glial cells can be depolarized by excess neurotransmitters and/or increased extracellular potassium released by depolarized neurons. Double-staining with phospho-

CREB and PDYN shows there is no apparent colocalization between these two antigens, thus activation of CREB may not account for the up-regulation of dynorphin following nerve injury.

Earlier studies indicate c-Fos may be involved in the up-regulation of dynorphin associated with inflammation (Dubner and Ruda, 1992). This conclusion is based on the observation that c-Fos expression is elevated in the spinal cord upon inflammation and the time course of c-Fos elevation precedes that of dynorphin. Furthermore, 90% of dynorphin cells also express c-Fos (Naranjo et al., 1991; Koguchi et al., 1991). In the

L5/L6 nerve ligation model, it is not clear whether c-Fos expression is also elevated. In our experiments, there was no difference in the number of cells stained with c-Fos antibody in the dorsal horn between the ipsilateral side and contralateral side. In addition, while some cells in laminae V or deep layers were labeled by both c-Fos and 74 prodynorphin antibodies, a significant portion of cells expressing prodynorphin in the superficial layers were not stained with the c-Fos antibody. Therefore, the mechanisms that lead to dynorphin up-regulation due to nerve injury and those due to inflammation may be different.

To examine the signal transduction pathways that lead to the activation of NF-KB,

AP-1 and CREB, we also immunostained spinal cord sections with antibodies that recognize the phosphorylated, activated form of several MAP kinases such as ERKl/2, p38, and INK. Discrete, cellular, staining for phospho-ERKl/ERK2 was observed in laminae I, II, laminae IV, V as well as the area around the central canal. Following nerve injury, there were more phospho-ERKl/ERK2-immunoreactive cells in laminae I and n of the ipsi lateral side than in the contralateral side of the lumbar spinal cord. Considering the distribution of the activated form of ERKl/2 and the fact that more cells are stained in the ipsilateral side, it is reasonable to suggest that ERKl/2 activation is nociceptive specific. However, whether this is true remains to be established. Since many prodynorphin-containing cells do not have phospho-ERKl/2 immunoreactivity, ERKl/2 activation does not seem to contribute directly to the up-regulation of prodynorphin in all neurons. Alternatively, it is known that the activation of ERKl/2 in neurons is transient as suggested by studies in other system (English and Sweatt, 1996), thus what we observed may only reflect the ongoing activation of ERKl/2. It is possible that ERKl/2 was also activated in those dynorphin-expressing cells that are not stained with phospho-

ERKl/2 at this moment. If this is true, then the role of ERKl/2 activation in the 75 regulation of dynorphin gene expression could be underestimated. It remains interesting to investigate the roles of ERXl/2 pathway in the regulation of dynorphin gene expression using specific inhibitors of this pathway.

MAP kinase p-38 is preferentially activated by stress signals and cytokines, and is implicated in neuronal apoptosis pCia et al., 1995). In the spinal cord of sham-operated rats, cells are also stained with phospho-p-38 antibody, indicating constitutive activation of this enzyme in the spinal cord. The fact that more cells are stained in the ipsilateral side than in the contralateral side of the spinal cord indicates nerve injury produces stress or elevated cytokines in the spinal cord. This result is consistent with those reported by

DeLeo et al (DeLeo et al., 1997) who demonstrated an increase in the lumbar spinal IL- ip, TNF-a and TGF-P-like immunoreactivity following chronic constriction or cryoneurolysis of sciatic nerve. It is possible that the activation of p38 observed here results from actions of these cytokines. On the other hand, p38 pathway also plays an important role in mediating expression of genes whose products are involved in immune activation and apoptosis (Graves et al., 1996; Xia et al., 1995), thus p38 may play a role in the up-regulation of these cytokines. These possibilities can be differentiated using specific inhibitors for p-38. 76

III. REGULATION OF DYNORPHIN EXPRESSION IN OPIOID-

TOLERANCE

INTRODUCTION

Opioid tolerance describes the states associated with sustained administration of opioids in which the ability of opioids to relieve pain is diminished. Numerous theories have been proposed to explain this phenomenon, including the down-regulation of opioid receptors, uncoupling of opioid receptors to the intracellular signal transduction, rd)ound of enzymatic activity such as adenyiyl cyclase that normally is inhibited by opioid receptors etc. These hypotheses have been tested in various in vitro and in vivo opioid tolerant models, but have not been established definitely. Recent studies suggest significant similarity between opioid tolerance and neuropathic pain. First, both processes are dependent on NMD A receptor activity, as blockade of the NMDA receptors has been shown to prevent the development o^ and reverse established, opioid tolerance as well as reverse abnormal pain behaviors associated with nerve injury (Mao et al., 1994; Mao et al., 1995a; Mao et al., 199Sb). Second, opioid activity is diminished in both states in terms of potency and efficacy. Third, sustained spinal administration of opioids induces allodynia and hyperalgesia that are conmionly associated with nerve injury (Porreca et al, unpublished observations). Fourth, in both states, there is up-regulation of djmorphin, and administration of dynorphin antisera reverses abnormal pain behaviors ((Nichols et al.,

1997); Porreca et al, unpublished observations). Thus the elevated levels of spinal 77 dynorphin are believed to play an important role in the development and maintenance of pathological states associated with opioid tolerance and neuropathic pain. While this hypothesis remains to be established, a question associated with it is how administration of exogenous opioids induces the up-regulation of dynorphin, an endogenous opioid peptide.

Opioids have been shown to activate AP-1 and NF-icB in rat embryonic cortical cultures (Hou et al., 1996). Since gene encoding prodynorphin contains AP-1 and NF-TCB motifs in its regulatory r^ion, it is possible that opioid-induced elevation of spinal dynorphin results from activation of opioid receptors, which subsequently activates AP-1 and NF-KB. In parallel to studies with nerve injury model, this chapter was aimed to explore the relationship between AP-1 and NF-KB activation and dynorphin expression in opioid tolerant rats induced by chronic DAMGO treatment. 78

Materials and Method

Preparation of cortical neuron cultures.

4 day-old Sprague Dawley rats were anesthetized with ether and decapitated.

Brains were taken out and immersed in ice-cold dissection buffer (137 mM NaCl, S mM

KCl, 0.2 mM Na2HP04, 0.2 mM KH2PO4, 33 mM glucose, 44 mM sucrose and 10 mM

HEPES, pH 7.3). Brain cortice were then excised and minced. Tissues were dispersed

sequentially by flame-polished Pasture pipettes of decreasing sizes. Cells and cell

aggregates were then plated on L-polylysine-coated coverslips placed in 6-well titer

plates (tissues from one rat were dispersed into 6 wells). Cultures were maintained in

minimal essential medium (MEM) supplemented with 0.4% glucose, 10% fetal calf serum, 10% horse serum, 10 mM KCl. Medium was changed on day 3 to maintenance medium that has similar compositions as plating medium except fetal calf serum was omitted.

Intrathecal catheter placements and inducement of opioid tolerance

Male Sprague-Dawley rats (200-300 g) were used in opioid tolerant experiments.

Under anesthesia, the rats were implanted with i.th. catheters for administration of drugs into the region of the lumbar cord according to the method described by Yaksh and Rudy

(Yaksh et al., 1976). A PE-10 polyethylene tube (8 cm) was inserted through the atlanto- occipital membrane and to the lumbar enlargement and secured to the musculature at the 79

incision site. Rats were then received 4.4 mg/kg, i.m. of gentamycin and allowed to

recover.

Tolerance regimen. Tolerance to DAMGO was induced by daily i.th.

administration of 1 ^g DAMGO (ED90) for 7 days and tested on day 8. A separate group

received 1 |ig DAMGO at day I and day 8, and saline at days 2-7, which served as

control in addition to rats received saline for 8 days.

Nociceptive assays

Nociceptive testing was performed using the warm water tail-flick test. The tail

was dipped in a water bath maintained at SS^C and the tail withdrawal latency was

measured before and after DAMGO treatment. A cut-off latency of 15 s was employed to avoid tissue damage.

Tissue dissection

Rat dorsal quadrants of the lumbar enlargement corresponding to the LI to L6 spinal segments were dissected as described in chapter n. Tissue samples were weighed, immediately frozen on dry ice and stored at -70°C until assayed. Left quadrants were used for immunoassay for dynorphin, while right quadrants were used for gel shift assay for AP-1 and NF-KB activity.

Immunoassay of dynorphin in spinal tissue 80

For extraction of dynorphin, spinal cord quadrants were placed in IN acetic acid,

disrupted with a polytron homogenizer, and incubated at 95 °C for 20 min. Samples were

centrifuged at 10,000x g for 20 min at 4°C. Supematants were then lyophilized and stored at -70°C until assayed. Protein concentrations of extracts were determined using the bicinchoninic acid method, with bovine serum albumin as standard. Immunoassay for dynorphin was performed with a commercial enzyme immunoassy system (Pennisula

Laboratories, Belmont, CA). Briefly, to each well precoated with anti-rabbit antiserum antiserum, were added dynorphin standard or the sample to be assayed, plus anti- dynorphin A antiserum and biotinylated dynorphin A All assays were performed in duplicates. Plates were incubated 2 hours and washed 5 times. Streptavidin-linked horse radish peroxidase was added and plates were incubated at room temperature for 1 hour.

Tetramethylbenzidine solution was then added and incubated at room temperature for 30 min. Reactions were stopped with 2 N HCl. The reaction product was analyzed by absorbance spectroscopy at 450 nM. The ICsowas calculated by non-linear least squares fitting to standard formulas (GraphPad PRISM).

Electrophoretic mobility shift assety

Procedures for nuclear extraction and electrophoretic mobility shift assay and analysis of shifted bands were described in Chapter n. 81

Results

Studies with cortical cultures

First, the effect of DAMGO treatment on the binding activity of NF-KB and AP-1

was studied with cortical cultures (3 to S days in culture). There was one major band in

the cortical culture for both NF-KB and AP-1 oligonucleotides. They were specific, since

they were competed by lOX or SOX cold homologous oligonucleotides, but not by

heterologous oligonucleotides OCT. After treatment with 10 ^ M DAMGO, there was a rapid increase in the binding activity of both NF-KB and AP-1, with 2 fold increase at 4 hr, and 3 to 4 fold increase at 24 h and 48 hr, respectively. This effect was blocked by 10

|iM naloxone, suggesting it was mediated through an opioid receptor. Naloxone alone did not have significant effect. This result is consistent with that of an earlier study with rat cortical cultures (Hou et al., 1996). In that study, embryonic cultures were used and a much higher increase was observed, presumably, because embryonic cultures contain more neurons. It is thus assumed this effect of DAMGO is primarily on neurons.

Assessment of tolerance

1 ^g i.th. DAMGO was able to significantly increase the tail-withdrawal latency in naive rats, achieving a maximal %MP£ of 67.6 ± 11.2. When tested on day 8, the same dose of DAMGO lost its ability to produce antinociception (%MP£ = lO.S ± 2.5) in rats 82

NF-KB AP-l

. I

CoMNF-KB - lOX SOX - - - lOX SOX - - Cold AP-l

Cold OCT - - - 1 OX SOX - lOX SOX Cold OCT

Figure 3.1. Basal NF-icB and AP-l like activity in neonatal rat cortical cultures at day 4.

400

•£ 300 ou X 200. o 0) 100. a. 1

4 h 24 h 48 h 48 h + nai nal

NF-/

AP-l

Ctr 4h 24h 48h +nal nal

Figure 3.2. Effect of 10 |U.M DAMGO on the DNA binding activity of AP-l (hatched bar, n = 1) and NF-KB (open bar, n = 2). Nal, abbreviation for naloxone. 83 treated with 1 fig i.th. DAMGO per day for 7 days, but not in rats that received 1 jig

DAMGO on day 1 for antinociceptive test and saline on subsequent 6 days. This indicated that daily i.th. injection of 1 |ig DAMGO for 7 days induced tolerance reliably.

Immunoassay of dynorphin

There was a significant, although moderate increase in spinal dynorphin level in tolerant rats (mean ± S.E.M. = 1567±112 pg/mg protein) in comparison with saline- treated rats (mean ± S.E.M. = 1150±110 pg/mg protein) (Figure 3.3).

2000-, 5

Saline DAMGO

Figure 3.3. Spinal dynorphin levels in the dorsal horn of spinal cords of rats injected with saline (open bar) or DAMGO (hatched bar). * P < 0.05 as compared with saline treated rats (Courtesy of Ibrahim et. al) 84

Electrophoresis mobility shift assc^

Electrophoresis mobility shift assay was performed with radiolabeled double- stranded oligonucleotides containing NF-KB or AP-I consensus sequence and nuclear extracts from the dorsal quadrants of the lumbar segments of the spinal cord. As shown in chapter n, there were three protein complexes that bound to NF-KB oligonucleotides, all of which were competed by excess cold NF-KB oligonucleotides. Only the bottom band seemed to be specific, since the upper two bands were competed by excess cold OCT oligonucleotides. Treatment of rats with 1 ^g i.th. DAMGO for 8 days resulted in a significant, although moderate increase in NF-KB binding activity, while injection of

DAMGO at day 1 and 8 only did not. Similarly, there were two protein complexes that bound to AP-1 oligonucleotides from nuclear extracts from dorsal quadrants, which were competed by excess cold AP-1 oligonucleotides. The upper band seemed to be AP-1 specific, since it was not competed by OCT oligonucleotides, while the bottom band was competed by OCT oligonucleotides. Chronic treatment with DAMGO also moderately increased the binding activity of AP-1, while animals received injection at day 1 and 8 only did not show any increase in the binding activity of AP-1. 85

•NF-KB 200, ^ AP-1 2 c o O 2 ^00- c CD

0 CL

D1.8 D1-8

NT-'B

AP-1

5 5 1: 'cc rji ='

Figure 3.4. Effect of chronic treatment with i.th. DAMGO on the DNA binding activity of AP-1 and NF-icB. Data are normalized to those obtained with saline-injected rats. * P < 0.05 compared with 100 %. D1.8, animals that received 1 jig DAMGO at day 1 and day 8, servicing as control to determine whether acute injection of DAMGO has any effect; Dl-8, animals that received I |j.g DAMGO injection from day 1 to day 8. 86

Discussion

A major finding in this section is that chronic DAMGO produced an increase in the DNA binding activity of AP-1 and NF-KB in both rat cortical cultures and in the spinal cord of DAMGO tolerant rats. This increase seems to be mediated through opioid receptors in the cortical culture, since it can be blocked by naloxone. It is not known whether in our cortical culture, cells that have opioid receptors also express dynorphin, thus the significance of activation of AP-1 and NF-KB related to dynorphin expression remains to be established. Opioid receptors are known to produce inhibitory actions on neurons by coupling to ion chaimels, either enhancing potassium outflow or inhibiting calcium inflow (Moises et al., 1994; Moises and Walker, 198S; Vidal et al., 1984). The net effect is the shortening of action potential and decrease in calcium influx into the cells. Since increase in the intracellular calcium level has been linked to the activation of

AP-1 and NF-KB, this inhibitory action of opioids on neuronal activity would have a negative effect on, and can not account for the activation of AP-1 ad NF-KB. Another possibility is that activation of opioid receptor leads to a decrease in the cellular cAMP level, which then causes the activation of AP-1 and NF-KB. In regarding this, maintained cellular levels of a cAMP analogue have been shown to prevent the effects of G-protein- coupled receptors on the activation of p44 MAP kinase (Burt et al., 1996). Thus it is possible that opioid activation decreases cAMP level, which then activates AP-1 and NF-

KB through MAP kinase pathways. Alternatively, opioid receptors may activate MAP kinase pathways via the 3y subunits of Gi protein (Burt et al., 1996; Li and Chang, 1996), 87 which then activate AP-1 and NF-k-B. In these experiments the activation of MAP kinase pathways by opioid receptors peaks at S to IS min after exposure to opioids, then desensitizes. It takes about 1 hr to recover after discontinuation of opioids. The effects of chronic opioid treatment on the MAP kinase pathway remain unknown. If desensitization of MAP kinase pathways continues with chronic treatment of opioids, then the activation of AP-1 and NF-KB seen after chronic opioid exposure could not be due to the activation of MAP kinase pathway via the Py subunits of G-, protein. On the other hand, there is no evidence that the activation of opioid receptors can lead to an activation of MAP kinase in the spinal cord, or this process desensitizes upon chronic opioid treatment in vivo. The interaction between MAP kinase pathways and opioid system in vivo needs to be further studied.

To directly relate opioid receptor activation to dynorphin expression in vivo, there is another requirement that must be satisfied, that is, the same neuron expresses both opioid ^ receptor and dynorphin. Dynorphin is expressed in the superficial Laminae I and n, as well as in laminae IV and V. Although some dynorphin containing neurons project to supraspinal cord sites (Leah et al., 1988; Nahin, 1987; Nahin et al., 1989), most neurons that express dynorphin in laminae I and n are believed to be intemeurons (Nahin et al., 1989) and innervate projection neurons heavily (Nahin et al., 1992). Opioid receptors are also expressed at high level in the Laminae I and n, and about 70% of opioid receptors are expressed presynaptically in the afferent terminals, which account for the most staining in laminae I, and can be knocked down by dorsal rhizhotomy. About 88

30% of opioid ^ receptors are expressed in neuronal bodies in laminae n (Arvidsson et al., 1995; Ding et ai., 1996), which are also believed to be intemeurons. Thus it is possible that intemeurons in laminae II express both opioid ^ receptors and dynorphin, and activation of opioid receptor may regulate dynorphin gene expression in these cells.

Alternatively, chronic opioid treatment may also induce changes in the neural circuit that is responsible for the alteration in dynorphin gene expression. For example, much evidence suggests activation of NMD A receptors plays a role in the initiation as well as maintenance of opioid tolerance, in particular, morphine tolerance (Manning et al., 1996; Trujillo and AJdl, 1991). Thus activation of AP-1 and NF-KB and subsequent inducement of dynorphin expression may result from an activation of NMDA receptors.

Again, the activation of NMDA receptors may affect the expression of dynorphin in the same cell or cells downstream in the neural circuit. These are possibilities that can be differentiated by double inununolabeling studies.

Activation of MAP kinase pathway and subsequent activation of AP-1 and NF-KB may also play a role in the hypersensitivity of the nociceptive pathway in the spinal cord occurring in opioid tolerant rats. It has been noticed that there are many similarities between the hypersensitivity of the nociceptive pathway in the spinal cord and long term potentiation (LTP) associated with memory in brain. Since MAP kinase pathway has 89 been implicated in learning and memory (English and Sweatt, 1996), it is likely that this pathway is also involved in opioid-induced abnormal sensation and tolerance. 90

IV. BINDING EXPERIMENTS WITH LODINATED

DYNORPHIN A(2-17)

INTRODUCTION

Application of exogenous dynorphin to CNS produces a number of non-opioid receptor mediated effects, including hind-limb paralysis, motor activity impairment and loss of tail-flick response (Caudle and Isaac, 1987; Przewlocki et al., 1983; Walker et al.,

1982a). These neurotoxic effects of dynorphin may be mediated by the N-methyl-D- aspartate receptors (NMDAR) based on the observation that they could be prevented by pretreatment with NMDAR antagonists (Bakshi and Faden, 1990; Long et al., 1989). The relationship between dynorphin neurotoxicity and NMDAR activity has not been clearly established. It has been proposed that dynorphin could contribute to excitotoxicity indirectly by enhancing the release of excitatory amino acids such as aspartate and glutamate (Long et al., 1994; Skilling et al., 1992). The increase in release of these neurotransmitters, however, does not correlate well with excitotoxicity. Dynorphin has also been proposed to mediate its effiects through directly interacting with the NMDAR based on radioligand binding analysis. In competition analysis using rat brain membranes, dynorphin and its related fragments attenuate the binding of [^HJglutamate

(Massardier and Hunt, 1989) and [^H]MK801 (Shukla et al., 1992) but potentiate that of

[^H]CGP396S3 (Dumont and Lemaire, 1994). This binding profile suggests that dynorphin is rather inhibitory on the NMDARs. Electrophysiological studies of NMD A- 91 mediated currents in isolated trigeminal neurons (Chen et al., 199S), or in Xenopus oocytes that expressed NMDAR subunit complexes (Brauneis et al., 1996), also indicated that dynorphin attenuates NMDAR conductance. This inhibitory action of dynorphin on

NMDA receptor is not entirely consistent with the excitatory actions of dynorphin observed in vivo. A solution to this apparent paradox is the proposition that dynorphin may produce excitatory effects through a disinhibition mechanism (Chen et al., 1995).

Dynorphin has also been shown to potentiate NMDAR mediated current under extremely low glycine concentration in the assay buffer (Zhang et al., 1997). However, these data were obtained under nonphysiological conditions and may not reflect the action of dynorphin in vivo.

The non-opioid actions of endogenous dynorphin have also been implicated in several pathological processes in spinal cord. In response to spinal cord trauma, there was significant increase in dynorphin immunoreactivity (dyn-ir) at the injury site, but not distant from the lesion. Dyn-ir was found elevated as early as 2 h and as late as 2 weeks after trauma, and was significantly correlated with the degree of injury (Faden et al.,

198S). These data are consistent with the hypothesis that dynorphin systems may be involved in the secondary injury that follows spinal trauma. This conclusion was fiirther substantiated by the finding that treatment with dynorphin antiserum significantly improved outcome after trauma as compared to control treatment with normal rabbit serum or leucine-enkephalin antiserum. Likewise, spinal dynorphin is elevated in response to peripheral nerve injury or inflammation (Draisci et al., 1991; Kajander et al.. 92

1990), and its significance in the abnormal pain produced by peripheral nerve injury has also been demonstrated. In a rat model of neuropathic pain, morphine did not alter tactile allodynia and showed reduced potency and efficacy to block the tail-flick reflex in nerve- injured animals (Ossipov et al., 199S). While intrathecal administration of MK-801 or antisera to dynorphin A (1-13) did not alter the tactile allodynia associated with nerve- ligation injury or the baseline tail-flick latency in either sham- operated or nerve-injured animals, co- administration of i.t. morphine with MK-801, or i.t. antisera to dynorphin A

(1-13) given prior to morphine elicited both a full antiallodynic response and a complete block of the tail-flick reflex in nerve-injured animals (Nichols et al., 1997). These findings suggest that the elevated level of dynorphin is involved in the attenuation of the effectiveness of spinal morphine in a model of neuropathic pain, possibly through modulating NMDAR function in the maintenance of neuropathic pain. This proposed action of dynorphin may not involve neurotoxicity as that seen with application of exogenous dynorphin since it can be reversed by antisera to dynorphin. The simplest hypothesis that would link the action of elevated dynorphin and NMDAR function is that dynorphin activates NMDAR through a direct interaction with the receptor complex. To test this hypothesis, we have charaaerized the interaction of dynorphin with the NMDAR using [^^^I]dynorphin A2.17 in a direct radioligand saturation binding analysis on membranes prepared from rat brain cortices as well as that prepared from HEK 293 cells which have been transfected with the cDNAs for the NMDARla and NMDAR2A subunits. Our data demonstrate clearly that dynorphin binds directly to the NMDAR complex, and defmes a novel high affinity site which appears to be associated with the 93 closed/desensitized state, rather than the active state, of the receptor. Hence, the high affinity interaction of dynorphin with the NMDAR complex defined by this direct radioligand binding analysis does not support an excitatory action of dynorphin on

NMDAR. 94

MATERIALS AND METHODS

Chemicals

MK801, AP-5, 7-chlorokynurenic acid (CKA), ifenprodil, (+)HA-966 and glutamic acid were from Research Biochemcal International (Natick, MA). Glycine, bestatin, captopril, bacitracin, phenylmethylsulfonyl fluoride (PMSF), bovine serum albumin (BSA) and other buffering reagents were from Sigma (St. Louis, MO).

Radiochemicals

{des-Tyro^r^ynoTphATi A2.17 (sequence: GGFLRRIRPKLKWDNQ; See Methods below for synthesis) was radioiodinated by the Bolton & Hunter procedure (Bolton and

Hunter, 1973) and purified using reverse phase HPLC on a Vydac CI8 column (Custom

Labeling Service, Amersham Life Sciences, Arlington Heights, IL). The [^^1](des-

7>roirv/)dynorphin A2-17 was supplied lyophilized with a specific activity of about 2000

Ci/mmol. When reconstituted in 1 ml water, it contains 5% lactose, 0.25% bovine serum albumin and 0.3 TlU/ml aprotinin in 50 mM sodium phosphate buffer, pH 7.4. [^MK-

801 was from NEN (Boston, MA).

Synthesis of (des-Tyrosyl)

Dynorphin A(2-17)-OH was synthesized by standard solid phase Fmoc chemistry protocols on an Applied Biosystems 431A peptide synthesizer using Wang resin presubstituted with Fmoc-Lys(Boc)-OH (Advanced ChemTech, Louisville, KY, 0.68 95 mmol/g substitution level) on a 0.25 mmol scale. Purification was effected by RP-HPLC

(CI8, Vydac Protein and Peptide column, 5.0 x 25.0 cm), gradient elution from 0 - 70 % acetonitrile (0.1% aq. TFA makeup solvent) over 70 min. with a flow rate of 40 ml/min.

The purity of the peptide was assessed as being >95% by analytical RP-HPLC (two gradient systems) and TLC (four solvent systems). Amino acid analysis was carried out by mass spectrometry (Department of Chemistry, Mass Spectrometry Facility, University of Arizona). Some of the peptide used in this study was also kindly provided by NIDA

(the peptide was synthesized by Multiple System Peptides Inc.; purity, 98%; protein content 74%).

Preparation of membranes from rat brain cortex

Sprague-Dawley rats (male, 250-300 g) were anesthetized by ether and decapitated. Brain cortices were rapidly removed and homogenized in ice-cold 5 mM

Tris-Cl, pH 8.0. The homogenate was centrifuged at 48,000 X g for 25 min at 4°C. The pellet was resuspended in 5 mM Tris-Cl /lO mM EDTA pH 8.0 and incubated at 37°C for 10 min. The suspension was then centrifuged at 48,000 X g for 25 min at 4°C. The membrane pellet was washed with 5 mM Tris-Cl, pH 8.0 and centrifuged as above.

Membrane pellets were stored at -80°C until use. Protein content was determined by the method of Lowry (Lowry et al., 1951). Aliquots of membranes were thawed and washed twice in ice-cold 5 mM Tris-Cl, pH 7.5, prior to each experiment

Radioligand binding assess in rat brain membranes 96

Assays using \}^^lL\(des-Tyrosyl}dynoT^hxTi Aj.n were carried out in a shaking water bath at 25°C for 2 hr in 0.5 ml of binding buffer (5 mM Tris-Cl buffer, pH 7.5, containing 0.2 % BSA, 100 nM PMSF, 30 nM bestatin, 50 ^g/ml bacitracin and lO^M captopril). 10 naloxone was included in the assay buffer to prevent potential binding of the peptide to opioid receptors. For saturation analysis, 300 ^g of rat brain cortical membranes were incubated with 0.6-25 nM [^^'l](ffej-2>TOJv/)dynorphin A2.17 (['^^I](

73To^/)dynorphin A2.17 was used as a tracer and unlabeled (d!e5-2^o^/)dynorphin A2.17 was added to yield the desired concentrations of radioligand). Non-specific binding was defined by the binding of radioligand in the presence of 200 (des-Tyrosy/)dynorphin

A2.17. To examine the modulation of \^^^\](des-Tyrosyt}d,ynovpYdn A2-17 binding by

NMDA receptor ligands, binding of [^^'l](ife5-2)To^/)dynorphin A2-17 at a final concentration of 1 nM or 1 nM was conducted in the presence of a) AP-V, b) CKA, c) ifenprodil, d) (+)HA-966, e) MK801 and f) CNQX. Binding was terminated by rapid filtration through Whatman GF/B filters (pretreated with 0.2% BSA at 4 ®C for at least 2 hrs) on a Brendel cell harvester and the filters rinsed with 6 X 3 ml of cold saline. Fibers were transferred to polypropylene tubes and the amount of radioactivity in each filter was determined by gamma counting (Beckman Gamma 5500).

Transient expression of NMDAR in HEK293 cells

HEK 293 cells (ATCC CRL 1573) were maintained in 5% fetal calf serum / 5% newborn calf serum / 90% DMEM / 100 U per ml penicillin and 100 ^g per ml streptomycin in a humidified atmosphere containing 5% CO2 (culture reagents were from 97

Life Technologies, Gaithersburg, MD). Prior to transfection, cells were seeded into 175 cm^ flasks at approximately 3 X 10^ cells per flask and harvested at approximately 50% confluency (while they were still in logarithmic growth). The cells were resuspended by trypsinization and pelleted by low speed centrifugation at room temperature. The cells were then washed once in Hepes buffered saline (HBS) and resuspended in the same buffer at a density of 20 X 10^ cells/ml. For electroporation of these cells, 0.7 ml of cell suspension was mixed with the following; 30(ig of pRc/CMV containing the cDNA for

NMDARla, 30i/g of pCDNAl containing the cDNA for NMDAR2A and 440wg of salmon testes DNA (Sigma, St. Louis, MO), to a final volume of 0.8ml with 1 X HBS.

For non-transfected control cells, 0.7 mi of cell suspension and 440ug of salmon testes

DNA were made up to 0.8 ml with IX HBS. Each 0.8 ml batch of cells was eiectroporated separately in disposable 0.4 cm^ cuvettes in a Bio-Rad Gene Pulser n

Unit fitted with a Capacitance Extender n Module (Hercules, CA). Electroporation was carried out under constant 200 V and 900 Ohms. The cells were left undisturbed for 10 minutes prior to plating into two 75cm^ flasks containing culture media as described above. The cell cultures were supplemented with a final concentration of 200 ^M

MK801 six hours after plating to inhibit any potential excitotoxicity that might occur with over- expression of the NMDAR complexes in transfected cells. Cells were harvested 72 hr after plating. Prior to cell harvest, the media were aspirated and cells incubated in serum-free medium containing 500 ^M glutamic acid and 500 fiM glycine for 1 hr. The cells were then washed in serum-free medium and resuspended in ice-cold

5 mM Tris-Cl, pH. 8.0 and pelleted. Crude membranes were prepared from these cells in 98 a similar mamier as described above for brain cortical membranes. Protein content was determined by the method of Lowry. Membranes were stored at -SO^C until use.

Radioligand binding assess in transected HEK293 cells

For each experiment, aliquots of membranes were thawed and washed twice in ice-cold 5 mM Tris-Cl, pH 7.5. For \}^^^{des-Tyrosyt)dynoTp\an A2.17 binding, membranes were resuspended in the binding buffer as described above for brain membranes to a concentration of SOO |ig/ml. Non-specific binding was defined as that in the presence of 50 |iM (des-TyrosytjAyx\ovp\nxi A2-17. For ['H]MK801 and

[^H]CGP39,653 binding, membranes were resuspended to a concentration of 500 fig/ml of 5 mMTris-Cl, pH 7.5 containing 100 jiM PMSF. Binding reactions using [^MK-801 were carried out in the presence of 100 glutamic acid, 100 fiM glycine and 50 (jM spermidine. Non-specific binding for [^H]MK-801 and [^H]CGP39,653was defined as that in the presence of 10 MK801 and 100|iM glutamate, respectively. Incubation and filtration of the samples were as described for brain membranes. Radioactivity of

['^'l](rosy/)dynorphin A2.17 was determined by ganuna counting; that of

[^H]MK801 and ['H]CGP39,653was determined by liquid scintillation counting.

Data analysis

Data from saturation binding were analyzed by non-linear regression analysis and

Rosenthal transformation. Statistical analysis was carried out using the student's t-test. 99

Statistical significance is defined at 95% confidence level. Both binding and statistical analyses were done using the program GraphPad PRISM. 100

RESULTS

Saturation analysis

Initial experiments showed that ['^^^(i/e^-TyrosyOdynorphin A2-17 rapidly associated with rat brain cortical membranes; the binding reached an equilibrium after 30 min at 25°C and remained stable for at least 4 hours (data not shown). Hence, all subsequent binding experiments were incubated for 2 hr at 25 °C. Under these

Q. 0.3 O) c c 0.2L o o0) Q. 0.0, CO SB(pmoi/mg) 10 15 20 25 [Free ligand], nM

Figure 4.1. Saturation and Rosenthal (inset) analysis of ['^^I]dynorphin A2-17 binding to rat brain cortical membranes. Each ligand concentration was done in duplicate. The data are representative of 3 independent experiments. Kj, 9.4 ±1.6 (mean±S.E.M., n=3); Bmax, 2.4 ± 0.6 (mean ± S.E.M., n=3). 101 conditions, [^^^I](rasy/)dynorphin A2.17 exhibited a saturable, specific binding in the concentration range of 0.6 nM and 25 nM (Figure 4.1). The specific binding typically constituted approximately 2S%-30% of the total binding. Rosenthal transformation of the binding data yielded a Kd value of 9.4 ±1.6 nM and a Bmu value of 2.4 ± 0.6 pmol/mg protein. At high concentrations of [^^'l](d!ej-2>ro^/)dynorphin A(2-17) (ranging from 1

to 10 piM), the specific binding increased further in a dose-dependent manner, but this increase was not saturable (data not shown), suggesting that cortical brain membranes contain a hgh capacity of low affinity binding for (cks-Tyro^l)dynoTp\un

A(2-17).

Modulation with NMD A receptor ligands

To test whether the observed specific binding of ]}^^^(^ks-TyrosylydynoTpYan Aj.

17 in the brain membranes was associated with the NMDAR, the binding of I nM

\}^^^(des-Tyrosyr)dyxioxpYan A2.17 to brain cortical membranes was carried out in the presence of 100 nM of a number of ligands specific for the NMDAR (Figure4.2A). AP-

V, an antagonist at the glutamate binding site of the NMDAR, 7-chlorokynurenic acid

(CKA) and (+)HA-966, antagonist and partial agonist, respectively, at the glycine binding site, and ifenprodil, an antagonist at the spermidine binding site of NMDAR, significantly potentiated the specific binding of I nM [*^'l](ro^/)dynorphin A(2-17). In contrast, MK-801, a NMDAR-specific ion channel blocker, but not the agonists glutamate and glycine, significantly attenuated the specific binding of 1 nM[^^^I](/)dynorphin A(2-17). Specific binding of ['~^l](i/e5-rvro5v/)dynorphin A(2-17) at high concentration of the peptide (1 nM), presumably occupying also the low affinity

i AP-5 Chlorokynurenic acid 200r CO) (+)HA-966 .SC 150. •• Ifenprodil X [=3 MK801 CO 100. * m. 50.

CO CO i o ? £CO -50. o -LOOL

B 0> 200. ] AP-5 C chlorokynurenic acid 5 150^ CZD (+)HA-966 •• Ifenprodil 5 100 E!ZI3 MK801

CO 50 CD <3)O) C CO Ox: T

Figure 4.2. Effect of NMDA receptor ligands on the binding of ['^^I]dynorphin A2-17 to the brain cortical membranes. The changes are expressed % of specific binding of ['"^Ijdynorphin A2-17 in the absence of NMDA receptor ligands and represent the means ± S.E.M. of at least three experiments conducted in triplicates. Statistical significance was determined by 95% confidence interval, * P<0.05 as compared with zero. (A), with 1 nM ['^^I]dynorphin Az-n- (B), with l(iM ['^^I]dynorphin A2-17. 103

-10 -8 -6 -5 Log[7-Chlorokynurenic acid], M

-10 -8 -6 -5 -4 -3 Log[MK801]. M

Figure 4.3. Dose effect of (A) 7-chlorokynuremc acid and (B) MK801 on the binding of 1 nM ['"^I]dynorphin2-i7 to rat brain cortical membranes. The results represent means ± S.E.M. of three experiments conducted in triplicates. *P<0.05 as compared with zero. 104

binding sites for the peptide, was not modulated by the NMDAR ligands (Figure 4.2B).

Further analysis of the dose-effect of CKA and MK-801 on the specific binding of 1 nM

['^^(d!ef-7>ro^/)dynorphin Aj.n showed that the potentiating effect of CKA had a half-

maximal concentration of 18 and a maximal 2 fold potentiation, while the inhibitory

effect of MK-801 was significantly only at high concentration of the ligand (>100 fiM),

showing a maximal inhibition of S0% (Figure 4.3).

Studies with transient expression ofNMDARla/R2A

HEK 293 cells which were transfected with the cDNA for the NMDARla and

NMDAR2A subunits expressed a saturable, specific binding of [^H]MK-801 with a Kd

value of 18 nM (data not shown). Transient expression produced on average 202 ± 20 fmol/mg protein (n = 8) of specific binding of [^H]MK-801 (at 40 nM) when compared with S4.6 ±21.4 fmol/mg protein in non-transfected cells. The net saturable binding of

[^H]MK-801 in transfected cells was 148 ±18 fmol/mg protein. Specific binding of

[^H]CGP39,653 (at 50nM) was 362 ± 61 fmol/mg protein (n = 3) in trransfected cells and

undetectable in non-transfected cells. On the other hand, a significant level of specific

binding was observed in no-transfected cells for ^^^^{des-Tyrosyl)dynorp\i\n A(2-17) at concentration of 10 nM (1.1 pmol/mg protein, n = 2). Parallel assays in transfected cells showed that the specific binding of this peptide was 2.6 pmol/mg protein, thus a net specific binding of 1.4Sn pmol/mg may be attributed to the expression of the

NRla/NR2A complexes in the transfected cells (Table 4.1). 105

Table 4.1. Transient expression of NRla/NR2A receptor complexes results in co­ operation of [^MK801, [^CGP39,653 and ['^'l](des-Tyrosyl)dynorphin A(2-17)

Ligand* Net Specific Binding n (fmol/mg)*' ['H]MK801 143 ± 17.8 4 [^H]CGP39.653 362 ±61 3 ['"l]dynorphin A(2-17) 1456 ±34 2

*The concentration of the radioligand used was 40 nM for [^MK801, 50 nM for ^H]CGP39,653 and 10 nM for (des-Tyrosl)dynorphin A(2-17). ^et specific binding of a ligand was determined by substracting the specific binding of the ligand in non-transfected cells from that in transfected cells in parallel binding assays using the same amount of membrane protein per assay. 106

DISCUSSION

Dynorphm2.i7binds to NKfDA receptor with high affinity. Dynorphin A has been shown to regulate the binding of NMDAR ligands such as Glutamate (Massardier and

Hunt, 1989), MK-801 (Shukla et al., 1992) and CGP-39653 (Dumont and Lemaire,

1994), arguing for the existing of a dynorphin binding site associated with NMDAR.

However, the affinity of the peptide for the NMDAR could not be accurately determined from these experiments because both NMDAR ligands and dynorphin have modulatory effects on NMDAR. With the radiolabled dynorphin A2.17 which does not interact with opioid receptors, we are able to determine directly the affinity of this peptide to NMDAR in the absence of NMDAR ligands. We detect a high affinity binding site of dynorphin in brain cortical membranes with a density that is similar to the density of NMDAR in rat brain defmed by [^H]MK-801 binding (Wong et al., 1988). In addition, this high affinity binding site can be modulated by several antagonists at different sites of the NMDAR, suggesting that this binding site is associated with the NMDAR complex. The high affinity binding we report here appears to be different from the low apparent affinity

(ICjo are in the micromolar range) obtained from indirect binding experiments as cited above. Most likely, this is due to different experimental conditions and modulatory interactions between dynorphin and NMDAR ligands. For example, the binding of

[^H]MK-801 (enhanced by glycine, glutamate and spermidine that are usually included in the binding assay) to NMDAR may promote a conformation of receptor that has lower 107 affinity for dynorphin, and this may explain why the IC50 value of dynorphin against

[^H]MK-801 is in the micromolar range in ligand competition analysis.

The co-expression of [^MK-801, [^CGP39,653 and [^"l]dynorphin Aj-i? specific binding in HEK 293 cells which have been transfected with the NMDAR la and

2A subunits provides further support for the existence of a dynorphin binding site on the

NMDAR. Based on the saturation analysis of [^H]MK-801, these transiently transfected cells express an appreciable density of NMDAR complexes which exhibit a high affinity for the antagonist, suggesting that the in vitro expression system using the NMDAR la and 2A subunits gives rise to fiinctional NMDAR complexes. Interestingly, in these membrane preparations, the binding capacity for [^H]CGP39,6S3 was roughly 2 fold of that for [^H]MK-801, which may reflect the stoichiometric ratio of the two ligands on the functional NMDAR complexes. At 10 nM of [^^^Ijdynorphin A2.17 (which approximates the dissociation constant determined from cortical membranes), non-transfected HEK 293 cells exhibit a substantial capacity of specific binding sites for the peptide, which is further increased by the expression of NMDAR complexes after transfection. This net increase in the specific binding of the peptide in transfected cells, however, is consistently higher than that for MK-801 and CGP39,653. This disparity is not likely to be due to differences in the affinity of the ligands for NMDAR, because the binding analysis used saturating concentration of both [^CGP39,6S3 and [^H]MK-801. Rather, these findings suggest that each NMDAR complex may possibly interact with several molecules of dynorphin. Alternatively, because the in vitro expression of NMDAR 108

subunits may generate a heterogeneous configuration of NMDAR complexes in the same

or in different cells, it is also possible that while MK-801 and CGP39,6S3 may

selectively bind to the "functional" NMDAR complexes, the binding of dynorphin may

not be as conformation-specific.

Dynorphin behaves like an antagonist on NMDA receptor. MK-801 has been

proposed to bind to the open state of NMDAR because its inhibitory action is dependent on the opening of the channel pore of NMDAR (Wong et al., 1988). Thus, agonists like glycine, glutamate and spermidine promote the binding of MK-801 (Ransom and Stec,

1988). In contrast, ifenprodil, an antagonist at the spermidine site (Reynolds and Miller,

1989), kynurenic acid and HA-966, antagonists at the glycine site (Reynolds et al., 1989), attenuate the binding of MK-801, presumably by shifting the receptor to the closed state.

Dynorphin, in many aspects, behaves like those antagonists. It attenuates the binding of

MK-801, but enhances the binding of [^H]CGP-39653, an antagonist at the glutamate binding site of NMDAR (Dumont and Lemaire, 1994), probably also by promoting the closed state of NMDAR. This effect is not due to a direct competition against endogenous glutamate, glycine or spermidine since dynorphin does not bind to these sites directly. In our experiments, all tested antagonists enhance the binding of [^^^I]dynorphin

A2-17, while MK-801 significantly attenuates the binding of [^^'l]dynorphin Aj.n. One explanation is that dynorphin has high affinity for the closed state of NMDAR, which is promoted by NMDAR antagonists such as AP-S, 7-chlorokynurenic acid and ifenprodil.

The attenuation of the binding of ['^^IJdynorphin Aj.n by MK-801 is consistent with this 109

postulation because MK-801 stablizes the open state of the channel. Since dynoiphin

promotes and prefers to bind to the closed state of NMDAR, it behaves like an antagonist

on NMDAR. These data thus predicts an inhibitory action of dynorphin on NMDAR

activity. An inhibitory action of dynorphin on NMDAR function is consistent with earlier

electrophysiological studies in isolated trigeminal neurons (Chen et.al.^ 199S). In their

studies, micromolar concentration of dynorphin has been shown to inhibit NMDA

mediated currents dose dependently and shortens the mean open time of channels. Other

evidence, however, appears to argue for an excitatory action of dynorphin on NMDAR

activity. Caudle and colleagues (Caudle et al., 1994) showed that in hippocampal slices,

low concentrations of dynorphin enhanced NMDA mediated currents, while high

concentrations of dynorphin inhibited NMDA mediated currents through kappa opioid

receptor. In this case, it is not clear whether the potentiating effect of dynorphin on the

NMDAR is through a direct or indirect mechanism, Dynorphin also potentiates NMDA

mediated currents under low glycine concentrations in Xenopus oocytes that express

NMDAR (Zhang et al., 1997), but the physiological relevance of this effect of dynorphin is not clear. Taken together, radioligand binding analysis continues to support an inhibitory action of dynorphin on NMDAR, and its physiological role remains contentious. However, since different subtypes of NMDA receptors coexist in tissues, thus we were investigating the action of dynorphin on population of mixed NMDA receptors and it is impossible to exclude the possibility that dynorphin may have excitatory effect on some species of NMDA receptors. Furthermore, whether this high 110 affinity binding site for dynorphin on the NMDAR can be excitatory under certain physiological conditions is not known.

If the dynorphin binding site on the NMDAR does not account for the excitotoxicity of dynorphin, it may be surmised that some non-opioid, non-NMDAR site of action of dynorphin may be responsible for inducing NMDAR excitotoxcity.

Pharmacologically, dynorphin produces several non-opioid neuropathological effects like reversible hindlimb paralysis, persistent loss of tail-flick reflex, and persistent hindlimb paralysis (Caudle and Isaac, 1987; Stalling et al., 1992). High doses of dynorphin have been shown to cause vasoconstriction and an increase in the release of excitatory amino acids (EAA) (Long et al., 1994), which may contribute to the progression of neurotoxicity of dynorphin. The involvement of EAAs in the neurotoxicity of dynorphin has been further supported by observations that competitive and non-competitive NMDA antagonists prevent the effects of dynorphin (Bakshi and Faden, 1990). Dynorphin may thus have a pre-synaptic action that potentiates EAA release.

It has been proposed that the excitotoxicity of djrnorphin may be mediated through the inhibition of GABAergic neurons by inhibiting presynaptic NMDAR (Chen et al., 1995). This proposed disinhibition mechanism of dynorphin assumes the existence of multiple subtypes of NMDAR, and that these subtypes have differential sensitivity to dynorphin. In support of this, studies with different subtypes of cloned NMDA receptors expressed in Xenopus Oocytes show that different subunit combinations exhibit Ill differential sensitivity to inhibition by dynorphin, in the order of e 1/^1 > 62/ (^1 > e4/ ^1

> E3/ ^1 (Brauneis et al., 1996). On the other hand, the neuroanatomical and molecular characterization of NMDAR subtypes in the spinal cord neurons remains to be established. In summary, the mechanisms by which dynrophin causes non-opioid effects are still not clear. At higher doses, there is release of EAAs and usually there is neuronal death. Whether endogenous dynorphin acts on non-opioid sites in normal animals or only under pathological circumstances in which level of dynorphin is elevated is not known.

The inhibitory action of dynorphin on NMDAR in vitro based on radioligand binding and electrophysiological studies is well supported; yet its physiological relevance, particularly in relation to dynorphin's excitotoxic effect, remains elusive. 112

V. EFFECT OF DYNORPHIN ON THE INTRACELLULAR

CALCIUM LEVEL

Introduction

The effects produced by exogenously applied dynorphin vary from neurotoxicity

(Caudle and Isaac, 1987) to inhibition of calcium potentials (Werz and Macdonald,

1985). Some of these effects, including loss of tail-flick and persistent hindlimb paralysis, are equally well produced with des-tyrosine dynorphins that have very low affinity for opioid receptors sites (Walker et al., 1982a), suggesting they are mediated through nonopioid sites. The nonopioid effects of dynorphin most likely involve NMDA receptor

(Bakshi and Faden, 1990; Long et al., 1989), and indeed, dynorphin may interact with

NMDA receptor directly, since dynorphin modulates the binding of a number of NMDA receptor ligands like glutamate (Massardier and Hunt, 1989), CGP-396S3 (Dumont and

Lemaire, 1994) and MK801(Schukla et at, 1992). However, results with binding experiments rather suggest hat dynorphin behaves like an antagonist at NMDA receptors.

This proposition is further supported by electrophysiological studies which show that dynorphin inhibits the NMDA receptor mediated currents in isolated trigeminal neurons

(Chen et al, 1995) and cloned NMDA receptor expressed in Xenopus oocytes (Braunes et. al, 1996). This inhibitory action of dynorphin on NMDA receptors is not entirely consistent with the excitatory actions of dynorphin observed in vivo. A solution to this apparent paradox is the proposition that dynorphin may produce excitatory effects through a disinhibition mechanism (Chen et al, 1995). Interestingly, these groups also 113 showed that at low glycine concentrations, dynorphin potentiates NMDA mediated currents in oocytes expressing cloned NMDA receptors (Zhang et al, 1997). Recently, a moderate stimulatory action of djmorphin on NMDA mediated currents in isolated periaqueductal gray neurons was also reported (Lai et al, 1998). In this case, the potentiating effect of dynorphin seems to be different from that seen on recombinant

NMDA receptors in that it is not dependent on external glycine concentrations. In summary, dynorphin may either inhibit or potentiate NMDA receptors depending on conditions and subtype combinations and more studies are needed. The present studies was initially designed to study the effea of dynorphin on NMDA receptors in cortical neurons, unexpectedly, we observed dynorphin alone produces an increase in the intracellular calcium. This action is not mediated by opioid receptors, nor NMDA receptors. This results would suggest a new mechanism via which dynorphin may produce an excitatory effect on neurons. 114

Materials and Methods

Materials

(ifej-2)T05y/)dynorphin A(2-17) (sequence; GFLRRIRPKLKWDNQ) was kindly provided by NIDA (the peptide was synthesized by Multiple System Peptides Inc.; purity,

98%; protein content 74%). Tissue culture media were from Life Technologies

(Gaithersburg, MD). MK-801, N-methyl-D-aspartate (NMDA), and other bufTering reagents were from Sigma (St. Louis, MO). Fluo 3/AM was from Molecular Probes

(Eugene, OR). Radioligands were from New England Nuclear (Boston, MA).

Preparation of cortical neuron cultures

4 day-old Sprague Dawley rats were anesthetized with ether and decapitated.

Brains were taken out and immersed in ice-cold dissection buffer (for 1 liter; 137 mM

NaCl, 5 mM KCl, 0.2 mM NaiHPO^.THiO, 0.2 mM KH2PO4, 33 mM glucose, 44 mM sucrose, 10 mM HEPES, pH 7.3). Brain cortex was excised and minced. Tissue was dispersed sequentially by flame-polished Pasteur pipettes of decreasing sizes. Cells and cell aggregates were then plated on poly-lysine-coated cover slips and maintained in

MEM supplemented with 0.4% glucose, 10% fetal calf serum, 10% horse serum, 10 mM

KCI. Medium was changed on day 3 to maintenance medium that had similar composition as plating medium except that the fetal calf serum was omitted.

Fluorescence microscope 115

Cortical cells were loaded with 6 (xM Fluo-3/AM in MEM (diluted from 1 mM

stock in DMSO ) for 60 min at 37 °C, then rinsed and incubated for an additional 30 min.

Each cover slip was then mounted onto a temperature-controlled chamber (37 ''C),

covered with 1 ml of bathing buffer (10 mM HEPES, 25 mM glucose, 137 mM NaCl, 10

mM KCI, 3 mM CaCh, pH 7.4) and placed on the microscope stage. Fluorescence images

were acquired using a Photometries liquid-cooled CCD camera attached to an Olympus

IMT-2 inverted microscope equipped with an Olympus 60X 1.4NA objective and 6.7X eyepiece. This camera has a linear response up to 5 x 10^ counts/image element. An electronic shutter under computer control was utilized to regulate exposure time.

Standard optics (Omega Optical, Brattleboro, VT) for Fluo-3 included a 10-nm band pass excitation filter centered at 480 nm and a 15-nm band pass emission filter centered at 530

nm. The digitized output of the camera was stored on a microcomputer and analyzed using customized software on a Silicon Graphics IRIS 10/900. The fluorescence intensity within a single cell was quantified by acquiring sequential images with a constant exposure time. This was set at 300 msec, which provided images with fluorescence intensities well above background [>500 integrated optical density] without causing saturation of imaging elements. All measurements were well within the linear response range of the CCD camera. For each cell, images were taken before (basal fluorescence) and after each drug administration (sample fluorescence). Drugs were administered directly to the bathing buffer in small volumes of concentrated stock to yield the desired final concentration. The maximal fluorescence for each cell was determined by the addition of 40 mM KCI at the end of the experiment, which was used to depolarize cells 116

and cause a calcium influx. The fluorescence intensity within a cell body is expressed as

the mean optical density averaged over the area of the cell body. Changes in intracellular

calcium concentration is expressed as a percent of maximal fluorescence induced by 40

mM KCl by the following equation; (sample - basal) / (maximal - basal) X 100%.

Statistical significance is determined by unpaired t-test.

Radioligand binding

The affinity of dynorphin A(l-17) and dynorphin A(2-17) for the three cloned

opioid receptor types from human was determined using clonal HN9.10 cells (Lee et al.,

1990) that express 5, k opioid receptors. The cDNAs for these receptors were kindly

provided by Dr. Lei Yu (Raynor et al., 1995), Dr. Brigitte Kieffer (Simonin et al., 1995),

and Dr. Erik Manssonr (Mansson et al., 1994), respectively. Transfection, cell culture

conditions and membrane preparation were carried out as described previously (Lai et al.,

1994). Ligand competition analysis was carried out using 10 concentrations of the

dynorphin A peptides against [^H]U69,593 labeled k receptors, [^H]DAMGO labeled ^

receptors, or [^H]pCl-DPDPE labeled 5 receptors. Assay conditions were as that described previously (Lai et al., 1994). Non-specific binding was defined as the amount of radioligand bound in the presence of 10 ^M naloxone. Reactions were terminated by

rapid filtration through Whatman GF/B filters and washed with 3X4 mis of ice-cold saline. Radioactivity was determined by liquid scintillation counting. Data were analyzed by GraphPad Prism. The ICso values were converted to Ki values by the Cheng and

Prusoff equation. 117

RESULTS

After 3 to 4 days in culture, many cortical cells were fluorescent after the cell culture was incubated with the cell permeate form of the calcium sensitive dye, Fluo-

3/AM. These cells were clearly distinguishable from other cells by the difference in their basal fluorescence, presumably due to a difference in the efficiency of dye uptake. These cells also displayed a neuronal-like morphology; most of these cells consisted of two or three long projections, and exhibited a steady basal level of fluorescence upon uptake of

FIuo-3. These cells are believed to be neurons since neurons load Fluo-3/AM better than glia cells. Image analysis was performed on cells that exhibited these morphological charaaeristics and could be visualized as discrete cells. All images were captured with a

60 X oil immersion objective as described above. Under this magnification, typically only a small number of cells were visible within the field of view and could be monitored simultaneously. Of all the cells sampled in this study, 45% (n = 179) responded to dynorphin A(2-17); 76% (n = 30) responded to NMD A. Figure 5.1 shows representative cells that responded to NMD A and dynorphin A(2-17). Those cells which did not exhibit an increase in [Ca^^i in response to drug were excluded from further analysis. In responsive cells, 100 dynorphin A(2-17) or 300 nM NMDA produced an increase in

[Ca^^]i time-dependently, reaching a plateau within 1 minute and remaining stable up to 5 minutes (Figure 5.2). The response induced by dynorphin A(2-17) and NMDA was dose- dependent (Figure 5.3). The EC50 value of NMDA mediated increase in [Ca^^i was 125

|iM. On the other hand, the highest concentration of dynorphin A(2-17) used in the assay 118

subtraction B

Figure 5.1. Fluorescence image of representative cells that was treated with 300 UM NMD A (A, B, C), 30 |J.M dynorphin A(2-17) (D, E, F). The intensity was shown in pseudocolor, with blue representing low, green, intermediate, and red, high intensity. C and F showed the net increase in [Ca~'^]i induced by drugs. 119

50 •B o 40

CM o CO 3 o -g 20 VU ii. g' o m Q-

Time (min)

60

^CO 3 O -D 30

O) 20 CL

Time (min)

Figure 5.2. Time course for intracellular calcium increase induced by 100 jiM dynorphin A(2-17) (A) or 300 NMDA (B). Data represents means of 9 to 10 cells. D (TQ p B n •-1fD re •a"-1 -2+, fDt/l U) Change in [Ca ]i as % of Change in [Ca^^]j as % of m response induced by KCI 3i-f O response induced by KCI ot/> 3a CO oi cn M CO 4k U1 a n1 o O O O O o 0 O O O O O o -1— 1— ~1 1 1 1 1 1 1 5 ES nSf> o o vO o o> o D b o n o £L •3 5" > (8 K» §Q. _• 2 ^ C (O o fO, s > O) 5o.

O) b o

N) OI

o Iv) o 121 did not reach a maximal response. Pretreatment with 10 ^M naloxone for 3 minutes prior to the addition of 100 ^M dynorphin A(2-17) did not block the increase in [Ca"^]i in response to dynorphin A(2-17) (Figiu-e 5.4), suggesting that the stimulatory effect of dynorphin A(2-17) on [Ca"^; even at high concentration was not mediated by opioid receptors. Radioligand competition analysis using dynorphin A(2-17) against opioid receptors K, fo. and 6 from stable transfected cells confirmed the low affinity of this peptide for the three cloned opioid receptor types (Table 5.1).

50r o o 40. c/} CO — T3 30. CD CM O CO 3 20.

O -10-

Figure 5.4. Effect of 10 nM naloxone on the increase in [Ca* ], induced by 100 (j.M dynorphin A(2-17). Data are from 8 cells for dynorphin A(2-17) alone (open bar), 4 cells for naloxone alone (solid bar), and 4 cells for naloxone plus 100 fiM dynorphin A(2-17) (hatched bar) 122

To determine whether the increase in [Ca^^i induced by dynorphin A(2-17) was

mediated through NMDAR activation, the effect of MK-801 on the dynorphin A(2-17)

induced response was investigated. First, the effective concentration of MK-801 in

blocking NMDAR mediated [Ca^^i was determined for our experimental system. Cells

were observed for their response to 300 nM NMDA, upon which 30 (jM MK-801was

added at 1 minute after the application of NMD A. MK-801 decreased the elevated

[Ca^^i gradually, reaching basal level by 2 min after the addition of MK-801 (Figure

S.5A). When the experiment was repeated using 30 |iM dynorphin A(2-17), the

subsequent application of 30 ^M MK-801 had no effect on the elevated [Ca^^i produced

by dynorphin A(2-17) (Figure 5.5B). To show that the lack of effect of MK-801 on dynorphin A(2-17) mediated [Ca^^i was not simply due to the absence of NMDAR in the sampled cells, we took the following alternative approach (Figure 5.6). Cells were first determined for their responsiveness to NMD A by the addition of 300 ^M NMD A. This response was blocked by the addition of 30 ^M MK-801 with NMDA still present in the bathing buffer. It was found that a subsequent application of 30 ^M dynorphin A(2-17) to the bathing buffer produced an increase in [Ca^i in 6 among 7 tested cells without any apparent delay in the time course or a decrease in the amplitude of the signal to dynorphin A(2-17) when compared with that observed with dynorphin A(2-17) alone (c./

Figure S.SB). When cells were tested for responsiveness to NMDA and dynorphin A(2-

17), of the 21 cells which responded to NMDA, IS of these cells also responded to dynorphin 123

A

50 o o ^ ic: 40 i 30MMMK801 .-=•8 M CM U <0 3 of 20 •i sC 8. 0)CO w

Time after addition of NMDA (min) B 80 o o

CM O "g 40

8. 20

Time after addition of dynorphin f\2A7) (mm)

Figure 5.5. The effect of 30 ^iM MK801 on the increase in [Ca"^]i induced by 300 i^M NMDA (A, n=8), or by 30 (iM dynorphinA(2-17) (B, n=5). * indicates significant decrease from response induced by NMDA at 1 min. 124

40,-

30 |iM MK801 30 pM dynorphin

1 2 3 4 5 6 7 Time after addition of NtJDA (nin)

Figure 5.6. The effect of sequential application of 300 fiM NMD A, 300 jiM MK801 and 30 (J.M dynorphin A(2-17) on [Ca"'], in cortical cells (n = 6). * indicates it is significantly different from the response induced by NMD A at I min.

_ 100. o o

CO >» 80- CD .Q

CM 60- CO 3

•E oj 40. (u <2 cn c Q. CD 20 O ^^5 ^4:0 -3.0 -2.5 Log[NMDA]. M

Figure 5.7. Increase in [Ca^^i mediated by NMDA alone (triangle, n=4), or by NMDA in the presence of 100 jiM dynorphin A(2-17) (rectangle, n=6) in cortical cells. The fluorescence was sampled 1 min after the addition of each dose of NMDA 125

A(2-17). Thus, a statistically significant number of sampled cells exhibited responsiveness to both compounds. Furthermore, blockade of NMDAR by MK801 did not alter the prevalence of cells which responded to subsequent challenge with dynorphin

A(2-17).

To study whether dynorphin A(2-17) is inhibitory on NMD A receptors, the effect of dynorphin A(2-17) on the dose-effect of NMD A mediated increase in [Ca^^i was studied. Administration of 100 ^M dynorphin A(2-17) elicited an increase of 20 ± 5% of maximal [Ca^^i response after 1 min, and this increase was further enhanced by the subsequent addition of increasing concentration of NMD A. The mean difference in the two dose response curves is 34%, which is comparable to the change induced by 100 ^M dynorphin A(2-17) alone (see Figure 3 A). Thus, these data suggest that dynorphin A(2-

17) and NMD A have an additive effect on [Ca^^i and an inhibitory action of dynorphin

A(2-17) on NMDA receptors can not be detected in this system.

Table I. Affinity of dynorphin A analogs for the human K, 5, ^ opioid receptors (OR).

icOR^ 5 OR* uOR* Ki(nM) Ki(nM) Ki(nM) Dynorphin A(l-17) 5.3 nM 182 6.9 Dynorphin A(2-17) > 10,000 > 10,000 > 10,000

•KOR were labeled with 1.8 nM ['H]U69,593 (Kd =1.4 nM); 50R were labeled with 1.5 nM [^H]pCl-DPDPE (Kd = 1.3 nM); NOR were labeled with 450 pM ['H]DAMG0 (Kd = 370 p}^. Data were mean values from two independent experiments 126

Discussion

The major goal of this study is to determine whether dynorphin may potentiate

neuronal excitability through an opioid independent mechanism, and whether these

effects of dynorphin are mediated through a direct activation of the NMDA receptors.

Our findings demonstrate that dynorphin A(2-17) produces an increase in [Ca^^i in

cultured cortical cells and thus may have a direct effect on neuronal cell excitability. This

excitatory effect of dynorphin A(2-17) is dose-dependent and time-dependent, and the

magnitude of this response is comparable to that elicited by NMDA in these cultured cells. Both the dynorphin A(2-17) and NMDA induced increase in [Ca^^i remained stable for the duration of exposure to the drugs (up to 9 minutes), showing little evidence of desensitization of the signal transduction mechanism. We found a statistically significant coincidence of responsiveness to dynorphin A(2-17) and NMDA in the cultured cortical cells; since NMDAR are primarily expressed in neuronal cells, this fmding suggests that the observed responsiveness to dynorphin A(2-17) is most likely to be associated with neuronal cells in these cultures. This excitatory effect of dynorphin

A(2-17) is not opioid receptor-mediated because this peptide exhibits very low affinity for opioid receptors (>10,000 nM) due to the absence of a N-terminal tyrosyl moiety, and its effect on [Ca^^i is resistant to the opioid antagonist naloxone. That dynorphin A(2-17) does not directly activate NMDA receptors in the cortical neurons is based on the fact that blockade of NMDA receptors has no effect on the excitatory action of dynorphin

A(2-17). On the other hand, we cannot evaluate whether d3morphin A(2-17) may 127

modulate the effect of NMDA by interacting with the NMDA receptor because we cannot

pre-block the dynorphin responsive site for not knowing what it is. This non-opioid, non-

NMDA action of dynorphin A is consistent with both a potentiation of excitatory amino

acids release as well as a direct postsynaptic potentiation. In both cases, dynorphin could

then promote NMDA receptor activation indirectly. For the latter, an increase in

intracellular calcium can potentiate the NMDA response through the activation of protein

kinase C, which has been shown to increase the probability of channel openings and by

reducing the voltage-dependent Mg^ block of NMDA-receptor channels (Chen and

Huang, 1992). Alternatively, it is possible that dynorphin-mediated increase in [Ca^i

may modulate the excitability of neural circuits involving NMDA receptors, rendering a sensitivity of the effect of dynorphin to MK-801.In this regard, the mechanism of dynorphin A(2-17) mediated increase in [Ca^^i is clearly distinct from opioid receptor

mediated effects of dynorphin A Much of the evidence that supports a modulatory role of dynorphin A on calcium conductance in neuronal cells shows that its effect is primarily inhibitory and is mediated by opioid receptors. For instance, dynorphin A inhibits high threshold, voltage dependent calcium currents in sensory neurons via 6 and ; receptors (Moises et al., 1994). In the hippocampal CA3 region, dynorphin A attenuates calcium channel-dependent synaptic transmission at mossy fibers (Castillo et al., 1996).

Dynorphin also reduces calcium currents in acutely dissociated nodose ganglion neurons via a Gi/Go coupled system in a naloxone sensitive manner (Gross et al., 1990). Such effects of dynorphin A are consistent with the inhibitory role of opioid receptors in neurotransmission and neurotransmitter release ((Zachariou and Goldstein, 1997);(Tiseo 128 et al., 1989)). The inhibitory action of dynorphin is consistent with some in vivo observations, such as the protection of dynorphin against focal cerebral ischemia in a feline model (Baskin et al., 1984), enhancement of the antinociceptive effect of morphine in morphine tolerant animals (Takemori et al., 1993), and protection against NMDA- induced neurotoxicity in cultured neurons (Choi et al., 1989).

Increases in [Ca^^i in the cortical cells by dynorphin A(2-17) may result from i) calcium influx via calcium channels, ii) a release of calcium from intracellular organelles or iii) inhibiting calcium efflux. Preliminary data suggest that the effect of dynorphin may be due to the activation of a calcium influx, because the effect can be blocked in calcium free medium (data not shown). Further experiment will be needed to substantiate this finding. This sensitivity to extracellular calcium argues against, but does not preclude, the possible action of dynorphin A(2-17) in coupling to IP3 receptor mediated intracellular calcium mobilization. Regarding the possibility of dynorphin A on calcium efflux, it was found that dynorphin inhibited calcium efflux from red blood cell ghost and Ca^ -

ATPase (Yamasaki and Way, 1983; Yamasaki and Way, 1985). However, this effect appears to be an opioid receptor mediated effect, which may not be relevant to the effect of dynorphin A(2-17).

In conclusion, dynorphin A(2-17) produces an increase in [Ca^i in cortical neurons which is independent of opioid or NMDA receptors. This effect of dynorphin is excitatory and is mediated by a novel mechanism. This action of dynorphin may underlie 129 its modulatory effect on excitatory neurotransmitter release in vivo; postsynaptically, this mechanism could also potentiate NMD A receptor function through a calcium dependent phosphorylation of the receptors. Thus, this novel mechanism of dynorphin may be the link to its excitatory actions in vivo. 130

VI. STUDIES OF GENE EXPRESSION IN L5/L6 LIGATED RATS

Introduction

Peripheral nerve injury has been shown to induce an increase in the expression of

dynorphin, substance P receptor, cytokines and neuropeptide Y in the spinal cord. This

presumably due to an increase in the transcription of these genes, although the expression

of some of these genes has not been studied at the transcription level. With the development of new techniques such as DNA arrays, now it is possible to study the

expression of hundreds or thousands of genes in one experiment. The challenge now is to

make sense of the observed alteration in gene expression and relate them to functional changes. In attempt to systematically study alterations of gene expression after LS/L6 ligation, we have tried a DNA array made by Clontech (Palo Alto, CA). One of the advantages of this array is that it only includes genes of known fiinction and does not

require expensive equipment. The disadvantage is that the number of genes in the array is limited and the genes of interest may not be in the array. 131

Materials and Methods

Preparation of total RNA

L5/L6 ligated rats were prepared according to the method described in Chapter n.

All nerve-ligated rats were verified to be allodynic (responding to a stimulus of less than

5 g) prior to harvesting tissue. To harvest tissue, the spinal cord was removed by applying

hydraulic pressure via an 18-gauge neddle and SO-cc syringe filled with saline. Dorsal

quadrants of the lumber segments of each spinal cord were dissected and frozen in liquid

nitrogen. Tissues were stored at -SO^'C until use. Total RNA was prepared using Atlas

Pure RNA Isolation Kit (Clontech, Palo Alto, CA). Briefly, 6 quadrants from each group

were homogenized in l.S ml denaturation buffer (2.7 M Guanidine thiocyanate, 1.3 M

amonium thiocyanate , 0.1 M NaOAc (pH 4.0)). Suspensions were incubated on ice for 5

to 10 min and centrifiiged at 15,000 x g for 5 min to remove cellular debris. Supematants were then extracted with saturated phenol (saturated with 19 % glycerol, 0.25 M NaOAc

(pH 4.0)) / chloroform 3 times and precipitated with isopropanol. RNA pellets were washed with 80 % ethanol once and dried and dissolved in water. DNA was removed by incubating samples in 200 ^1 of reactions containing 50 U DNase I. Reactions were then terminated by the addition of 20 ^1 of 0.1 M EDTA (pH 8.0) containing 1 mg/ml glycogen. Samples were extracted with phenol / chloroform twice and precipitated with ethanol, air-dried and dissolved in nuclease-free H2O at a concentration of 2 ^g/^1.

Preparation of cDNA probes 132

5 total RNA for each sample was radiolabeled in 10 |il or 20 ^1 of reaction by a commercial kit from Clontech as instructed. The kit uses a mixture of primers only for genes that are in the array, thus only these genes are synthesized to cDNA probes.

Reactions were carried out at 48 "C for 25 min and stopped by adding 1 |il of 10 X termination mix. Labeled probes were then purified using CHROMA SPIN-200 DEPC-

H2O (Clontech) as instructed by the manufacturer. Radioactivity of each fraction was counted using a scintillation counter and fractions that comprise the first peak of counts were pooled and stored at -20 °C until used.

Hybridization of cDNA to the Atlas array

Membranes containing the Atlas array were prehybridized at 68 °C for 30 min with 7.5 ml of ExpressHyb (provided by Clontech) containing 100 ^g/ml heat-denatured sheared salmon. Hybidization of memebranes was carried out at 68 °C overnight in 7.5 ml of ExpressHyb containing 100 (ig/ml heat-denatured sheared salmon testis DNA, 1

Hg/ml Cot-1 DNA. The final probe concentration is approximately 1 x 10® cpm/ml.

Membranes were then washed 4 x 30 min with 2 X SSC, 1% SDS, followed by 2 X 30 min washes with 0.1 X SSC, 0.5 % SDS at 68°C. The Atlas Arrays were then exposed to a phosphorimaging screen for 48 hrs or to x-ray films at -70 °C for 2 to 4 weeks.

Quantitative Analysis

Each gene is represented in duplicate in the Array, and in most cases, levels of signals from duplicate are very similar, indicating they are specific. To analyze the 133 expression of genes on the array quantitatively, a grid for each region was made to match the array. The intensity in each rectangle (contains one spot for a particular gene) was integrated to yield total signal in the area, this value was then normalized with the average of hybridization signals of genes from the same region and normalized values were then compared between two membranes to yield normalized ratio. It is necessary to use normalized values, since there is difference in the overall strength of signal as well as background between membranes. 134

Results

Profile of genes in the AtlasIMRat cDNA expression array

There are 588 rat cDNAs, 9 housekeeping control cDNAs, and negative controls

in duplicate dots on each membrane. These genes represent several different functional

classes, and are grouped to six regions. Region A includes oncogenes, tumor suppressor

genes, cell cycle regulators and genes involved in stress response. Region B includes

intracellular signal transduction modulators and effectors. Region C includes genes

involved in apoptosis or expressed in nervous system. Region D includes genes involved

in lipid and steroid metabolism and transport, translation, and genes encoding cell-surface

antigens and cell adhesion molecules. Region E includes genes encoding receptors for

growth factors, interleukins and hormones as well as channels and transporters. Region F

includes genes involved in cell-cell communication and protein turnover.

Profile ofgene expression in the dorsal ham of rat spinal cord

Among 588 gene in the array, about 80 genes were detected by cDNA probes

prepared from dorsal quadrants of rat spinal cords, including transcription factors such as

jun-B, jun-D, N-myc, celll cycle regulators cyclin D3 (Gl/S-specific) and cyclin-

dependent kinase 5 (i.e. tau protein kinase II). Several kinases, including c-ra^ casein

kinase 11, MEK2, MEK5, MAPKl, protein kinase C pi, P2, C wd mast/stem cell growth factor tyrosine kinase, are also expressed at high level. Nervous system-related proteins

detected include myelin proteolipid protein DM-20, peripheral myelin protein 22, neuron- 135 specific enolase, nerve growth factor inducible protein, synapsins 2A and 2B, plasma membrane calcium-transporting ATPase, glycine receptor a-1 chain precursor, secretogranin II, glycine transporter. Among cell-surface antigens and cell adhesion molecules detected are P-selectin precursor, neuronal acetylcholine receptor 3-2 chain precursor. A few growth factors and hormones are detectable, including fibroblast growth factor, insulin like growth factor n, somatostatin, testicular luteinizing hormone 3 subunit and macrophage migration inhibitory factor. Others are involved in translation and protein degradation.

Table 6.1. Genes whose expression was increased following SNL

Label Poaitjon G«n«Bank i Name Nonnalizad ratio % changa

1 B3k 031873 UM domain sarifM/thraonina Mnaaa 1 2.50 ISO

2 D5f D10831 P-aatoctin pracureor 1.8S 8S

3 B7c U094S7 DP0E3; cAMP-ptwaphodiaataraaa 40 1.64 64

4 83h M61177 MAPK1; ERK1; P44/MAPK 1.35 35

Changes in gene expression after L5/L6 ligation

Changes in gene expression were first identified by quantitative analysis as described in the method, then confirmed by films. Preliminary analysis indicated that there is 2.5 fold increase in the expression of LIM domain serine/threonine kinase 1, 1.9 fold increase in P-selectin precursor, 1.6 fold in cAMP-phosphodiesterase 4D, 1.4 fold in 136

MAPKl. A few genes are down-regulated after L5/L6 ligation, including glutathione S- transferase, adenosine A1 receptor, peripheral myelin protein 22, GyS, 9, synapsin 2A and 2B, and glycine transporter.

Table 6.2. Genes whose expression was decreased following SNL

Label Position GancBank* Nam* Normalizad ratio %changa

1 Clj J03752 gkjtatMona S^nsfaraaa (ECZS.1.18) 0.35 65

2 C2a M64299 adanoaina A1 racaptor 0.45 55

3 C3h M69139 SR13 myain proiain. PMP22; CD2S 0.23 77

4 BSh L3Se21 GgammaS, GgammaS 0.37 63

5 CSk M2792S aynapiin 2AA2B 0.53 47

6 E6a M88596 glycina tranaportar 1 0.50 50 137

sham

k 2

Figure 6.1. X-ray autoradiography demonstrated changes in gene expression following SNL. Sham and SNL indicated membranes probed with P^^-labeled cDNA prepared from the dorsal horn quandrants of sham-operated and SNL animals respectively. Arrows in sham indicated genes whose expression was decreased following SNL, identities of these genes were listed in table 2. Arrows in SNL indicated genes whose expression was increased following SNL, and identities of these genes were listed in table 1. 138

Discussion

The idea to study the differential expression of hundreds and thousands of genes between two states of tissue (normal vs diseased) is gaining more and more popular.

Combination of cDNA array printed on filter paper with standard hybridization does not require special equipment, thus might be suitable for small research labs. However, practical, usually only about 20% of genes in the array can be easily detected with X-ray film or phosphoimaging, thus a lot of genes are precluded from analysis. Among 588 genes, only 80 genes showed up in our experiment. These include transcription factors, enzymes in MAP kinase pathways, growth factors and hormones. MAP kinase cascade is typically studied in the context of mitotic regulation, but its components have been shown to most abundantly expressed in postmitotic neurons of the developed nervous system.

Our experiment also indicates they are expressed at high level in the spinal cord, although their physiological roles there are not clear. Recent studies suggest possible involvement of the MAP kinase cascade in the activity-dependent modulation of synaptic connections between neurons, a putative mechanism for the neuronal basis of learning and memory, in particular, hippocampal long term potentiation. It has been noted that there is parallel between hippocampal long term potentiation and central sensitization (increased response to an input after persistent stimulation) in the spinal cord, thus MAP kinase cascade may be involved in the central sensitization in the spinal cord. 139

Changes of gene expression observed here after L5/L6 ligation must be interpreted cautiously. First, changes at mRNA levels do not necessarily mean changes in protein levels or function, although in this discussion, I assume changes at protein levels follow changes at mRNA levels. Second, it is difficult to determine whether the observed changes reflect compensatory reactions of the system or induce the diseased states.

Keeping these in mind, I will try to discuss the meanings of changes observed in this experiment.

UM kinase. LIM kinase is a serine/threonine kinase that contains 2 amino- temiinal LIM domians, a central PDZ domain, and a carboxyl-terminal kinase domain. It is mainly expressed in neurons and hemizygous deletion of the 7ql 1.23 region containing

LIM kinase is implicated in the visuospatial constructive cognition defect in Williams syndrome (Frangiskakis et al., 1996). Cofilin, actin-depolymerizing factor/cofilamentous protein, has been identified as a functionally important substrate for LIM kinase.

Posphorylation of cofilin inhibits its actin filament severing activity thus decreases the actin dynamics (Arber et al., 1998; Yang et al., 1998). There is about 2.5 fold increase in the LIM kinase expression after L5/L6 ligation, it is possible that an increased expression of LIM kinase may lead to changes in the cytoskeleton or morphology of cells in the spinal cord.

P-selectin. P-selectin belongs to the family of extracellular molecules that includes E-, P-, and L-selectins. They are transmembrane molecules and homologous to 140

CRl and CR2 complement receptors. Selectin expression is one of the earliest phases of

inflammatory response leading to leukocyte activation and extravasation. P-seletin is

rapidly expressed on vascular endothelium in response to stimuli such as thrombin,

histamine and oxygen radicals. It mediates early neutrophil-endothelial cell interaction

that leads to the rolling of neutrophils on the endothelium and facilitates subsequent

cellular adhesion, migration. The role of P-selectin in spinal cord injury is supported by

fmdings that administration of anti-P-selectin antibody or knockout of P-selectin has been

shown to improve recovery from spinal cord injury (Farooque et al., 1999; Taoka et al.,

1997). The elevated level of P-selectin precursor after LS/L6 ligation may indicate an

inflammatory process in the spinal cord, a finding that is consistent with reports that

expression of cytokines is increased in the spinal cord after peripheral nerve injury

(DeLeo et al., 1996; DeLeo et al., 1997).

ERK1/ERK2 (extracellular signal regulated kinase) are two mitogen-activated

protein kinase (MAP kinase) that are usually activated in response to growth factor

(Seger and Krebs, 1995). Although this MAP kinase cascade is typically studied in the

context of mitotic ceil regulation, its components are actually most abundantly expressed

in postmitotic neurons of the developed nervous system (Boulton et al., 1991; Fiore et al.,

1993). The physiological roles of this cascade in mature neurons are not known, although

this cascade can be activated by neurotrophins (Davis, 1993), neurotransmitters that

activate G-protein coupled receptor (Gardner et al., 1993), as well as neuronal depolarization (Rosen et al., 1994). Recent studies imply a role of this cascade in the 141 hippocampal long term potentiation (Blum et a!., 1999; English and Sweatt, 1996) as well as drug-induced adaptations of mesolimbic dopamine system by chronic exposure to morphine or cocaine (Berhow et al., 1996). Our experiments described in Chapter n have shown that there is an increase in the activation of ERKl/2 in laminae I and II of the lumbar spinal cord following L5/L6 nerve ligation. Here we also observed an increase in the ERK1/ERK2 expression. The relationship between the activation of ERKl/2 and the up-regulation of ERKl/2 expression is not clear, yet these two observations suggest a profound alteration in the ERKl/2 pathway. It is possible that this pathway may be a major signal transduction pathway that leads to neuroplastic changes following nerve injury. The significance of this pathway in neuropathic pain needs further studies.

PMP22. The peripheral myelin protein has been implicated in type lA Charcot-

Marie-Tooth disease(CMTlA) that is characterized by an onset in childhood and resuks in widespread demyelination and late axonal loss. CMTIA is caused by overexpression or point mutations in the gene for peripheral myelin protein 22 (Harding, 1995). PMP22 makes up -2-5% of the total myelin protein and its role in myelin is not known. While overexpression of PMP22 is associated with CMTIA null mutations in the PMP22 gene give the human disease hereditary neuropathy with liability to pressure palsies (HNPP)

(Nicholson et al., 1994). HNPP is characterized by a less severe reduction in nerve conduction velocity and by focal myelin thickenings termed tomacula. After nerve ligation, we found a reduction in the expression of PMP22. Since PMP22 is expressed in the astral cells, the reduction probably reflects the loss of astral cells surrounding afferent 142

terminals in the spinal cord after nerve ligation as a result of injury induced

neurodegeneration.

Glycine transporter l(GlyT-l). GLYTs are capable of removing glycine from

neurons, thus play a crucial role in terminating the inhibitory actions of glycine at

strychnine-sensitive sites and as a co-activator of the NMD A receptor. Two GLYTs,

GLYTl and GLYT2, have been identified. In situ hybridization revealed colocalization

of GLYTl and the NMDA receptor (Smith et al., 1992), which is consistent with the

proposition that GLYTl plays a role in the regulation of the NMDA receptor. A decrease

in the expression of GLYTl may lead to a deficiency in removal of glycine from the

NMDA receptor, thus promotes sustained activation of the NMDA receptor in the spinal

cord after LS/L6 nerve ligation.

Adenosine Al receptor. Adenosine is an endogenous compound with various

modulatory effects in the central nervous system. It is believed that release of

neurotransmitters is partly controlled by the activation of adenosine receptors (Li and

Perl, 1994; Sah, 1990). In the spinal cord, adenosine Al and A2 receptor are primarily

located on intrinsic neurons in the substantia gelatinosa of the dorsal horn (Choca et al.,

1988). Adenosine and its analogues have been demonstrated to have antinociceptive

effects in acute (Sawynok and Sweeney, 1989) as well as chronic (Cui et al., 1997) pain

models at doses lower than those producing motor effects. Antinociception of adenosine

and its analogues results from an inhibition of intrinsic neurons by an increase in 143 conductance and presynaptic inhibition of sensory nerve terminals to inhibit the release of substance P and perhaps glutamate. Endogenous adenosine systems are also believed to contribute to antinociceptive properties of caffeine, opioids, noradrenaline, S- hydroxytryptamine, tricyclic antidepressants and transcutaneous electrical nerve stimulation (Sawynok, 1998). The downregulation of adenosine A1 receptor after nerve injury may imply a reduced antinociceptive action of drugs acting on this receptor. 144

Summary Statements

The central hypothesis of this dissertation is that peripheral nerve injury will

induce elevated level of dynorphin in the spinal cord, which will have a facilitating effect

on the nociceptive pathway via a non-opioid mechanism. To advance this hypothesis, two

major questions are addressed in this dissertation: how dynorphin is up regulated in the spinal cord of rats with peripheral nerve injury, and how dynorphin, an endogenous opioid peptide, produces a facilitating or excitatory effect on the nociceptive pathway.

Some results of this dissertation are summarized below along with future research opportunities.

1. Regulation of djmorphin expression (Chapter 2 and 3)

Previous studies indicate that dynorphin expression is up-regulated in the spinal cord following trauma, peripheral inflammation. More recently. Studies from our group suggest that dynorphin level is also elevated after peripheral nerve injury or chronic opioid treatment. However, literature on how dynorphin is up regulated in vivo is quite limited. Therefore, I attempted to address this issue by examining the role of some common signaling pathways in dynorphin gene expression. I reasoned that if any of these pathways is directly responsible for the up regulation of dynorphin gene expression following nerve injury, it must be activated in response to peripheral nerve injury, its activation must occur in dynorphin-expressing cells, and its activation should precede or 145

correlate the time course for the up-regulation of dynorphin. The studies in chapter 3

suggest that there are similar changes occurring in the morphine-induced opioid

tolerance. But studies with tolerance were not as complete as studies with neuropathic

pain. Therefore, this summary will focus on neuropathic pain. It is possible that findings

with neuropathic pain can be extended to morphine-induced tolerance.

Figure 7.1 summarizes the results in regarding how dynorphin is up regulated. In this summary, we assume that the activation of these pathways and transcriptional factors is due to an increase in the intracellular calcium, which in turn is due to the response of dorsal horn cells to a sustained, increased afferent input. Although this assumption is reasonable, there is no evidence to support it at this time. An increase in the intracellular calcium concentration then may activate a variety of intracellular targets, including the

MAP kinase pathways.

Following peripheral nerve injury, there was dramatic increase in the phospho-38-ir, which appeared to occur in glial cells as judged by morphology, and did not colocalize with PDYN. There was also an increase in the activation of CREB following nerve injury, but phospho-CREB did not colocalize with PDYN. Thus the pathway from p-38 to CREB does not play a direct role in the regulation of dynorphin gene expression. It is not known whether the activation of p-38 is due to the elevated level of cytokines in the spinal cord or vice visa. This issue could be resolved using p-38 specific inhibitors or cytokine antagonists. Profound activation of p-38 following nerve injury suggests an 146 activation of glial cells in the spinal cord. Since glial cells are also suggested to play a role in the acute hyperalgesia induced by peripheral inflammation, it is possible that glial cells may also play a role in neuropathic pain. Whether p-38 activation plays a role in neuropathic pain and inflammation-induced hyperalgesia can be studied using p-38 inhibitor SB202190. An assumption made here is that p-38 is also activated in the spinal cord by peripheral inflammation, which can be easily tested.

While AP-1 and NF-KB are activated by nerve injury, whether the increased activation occurs in dynorphin-expressing cells is not clear. Thus the role of the pathway from SAPK/JNK kinases to AP-1 and NF-KB in dynorphin gene expression is not known at this moment. This pathway usually plays a role in apoptosis and is activated in response to stress (Xia et al., 1995), thus this finding is provocative and its significance in neuropathic pain is worthy of further investigation. Before pursuing further along this line of research, it is important to determine which components of these transcriptional factors contributes to the increased DNA binding activity of AP-1 or NF-KB. This could be accomplished by supershift experiments using antibodies that recognize components of these transcriptional factors.

Following nerve injury, there was an increase in the phospho-ERKl/2-ir in laminae I, n of the dorsal horn, which was also located in some dynorphin-expressing cells. It is possible that the activation of ERKl/2 may play a role in the regulation of dynorphin gene expression in some cells. This possibility can be tested with PD980S9, which 147 specifically inhibits the activation of ERKl/2. It has also been proposed that the activation of ERKl/2 is dependent on the neuronal activity or NMDA receptor activation

(Ji et, al, 1999), thus the activation of ERKl/2 following nerve injury may reflect ongoing neuronal activation or activation of NMDA receptors in the dorsal horn.

Whether this is true remains to be tested. If this proposal can be established, then it would be interesting to see whether the activation of ERKl/2 can be used as marker to visualize the transmission of pain sensation in vivo. For example, allodynia may be transmitted along the pathway of touch sensation, so we can ask whether a touch can elicit an activation of ERKl/2 in neurons along this pathway. Finding a marker for neuronal activity was another goal for this research, and the activation of ERKl/2 may potentially be such a marker.

In summary, these experiments indicated that there are profound changes in the cellular signaling pathways following nerve injury, which may play an important role in initiating neuroplastic changes in the spinal cord, including the up-regulation of dynorphin. These changes affect not only neurons, but also glial cells. This conclusion is also supported by our initial experiment with cDNA array described in Chapter 6. That study suggests that nerve injury alters the mRNA levels of genes including adenosine receptor and ERKl that are expressed in neurons, as well as genes such as PMP22 and glycine transporter that are expressed in glial cells. The significance of these changes was speculated in Chapter 6. 148

Free radical scavengers PD98059-H i NO, H2O2, PKC, PKA, PAF, PYKj

MEK1/2 MKK3, MKK6 MKK4, MKK7

(g) (g) Erk1/Erk2 4 P38 MAPK SAPK/JNK IXI SB202190 /I

TC^EIk TCF/Elkl ^CREJTFEl'^ AP-'i U NF-icB

Figure 7.1. Regulation of dynorphin gene expression; relationship with MAP kinase pathways. The bottom line represents a schematic view of sites that can be bound by transcriptional factors in prodynorphin gene. The arrows beside transcription factors and enzymes indicate an increased activity in response to nerve injury. 149

2. Mechanisms of non-opioid actions of dynorphin (Chapter 4 and 5)

Previous studies implicate the NMDA system in the non-opioid actions of dynorphin

In addition, the fact that dynorphin modulates the binding of NMDA receptor ligands suggests a direct interaction between dynorphin and NMDA receptors. By direct binding analysis using ['^^I]dynorphin A(2-17) on membranes prepared from rat brain cortices as well as from HEK 293 cells expressing NMDARla/2A we have demonstrated that dynorphin binds directly to the NMDAR complex with relative high affinity. Dynorphin appears to bind preferentially to the closed state of the receptor. This finding prompted me to reexamine the indirect binding experiments using NMDAR ligands as radioligands from literature, which appear to suggest that dynorphin and its fragments shift the

NMDAR to the closed state. Based on these observations, I proposed that the direct effect of dynorphin on the NMDAR is inhibitory. This proposal is consistent with subsequent electrophysiological studies from other groups using isolated neurons or Xenopus oocytes that expressed NMDARs. Taken together, these in vitro experiments have profoundly changed our view on how dynorphin acts on the NMDA system and suggest that a direct activation of dynorphin on the NMDAR is less likely. The physiological relevance of this inhibitory action of dynorphin on the NMDAR is rather difiHcult to determine at this moment. 150

The calcium imaging studies revealed an unexpected action of dynorphin.

Dynorphin A (2-17) can produce an increase in the intracellular calcium independent of the opioid or NMDA receptors. This result would suggest a new mechanism via which dynorphin may produce an excitatory efifect on neurons. Figure 7.2 sununarizes some possible effects of an increase in the intracellular calcium on neurotransmission in the dorsal horn of the spinal cord. Dynorphin is predominantly located in intemeurons. Under normal conditions, it may produce an inhibitory effect on the neurotransmission by acting on the opioid receptors (not shown) or NMDA receptors on the afferent nerve terminals or project neurons. There is no definitive evidence for this putative action of dynorphin at this time. Under abnormal conditions, dynorphin may produce an increase in [Ca^^i via an unknown site (Rx) on the afferent nerve terminal. An increase in [Ca^i would be expected to enhance the release of neurotransmitters such as glutamate, aspartate (EAAs), or substance P, CGRP, depending on the types of the afferent neural terminals. EAAs then activate NMDARs on project neurons and afferent nerve terminals. This prediction is consistent with evidence from literature which suggest that dynorphin increases the release of EAAs in the spinal cord (Stalling et al, 1992), and dynorphin and dynorphin(2-

17) potentiates capsaicin-evoked release of CGRP from perfused spinal cord tissues

(Claude et. al, 1999). Since CGRP is contained in a subset of C-fibers, this result does suggest dynorphin may have a facilitatory effect on the release of neurotransmitters from high threshold evoked primary afferent terminals in the spinal cord. Nevertheless, this should be confirmed in a more defined system such as dorsal root ganglion (DRG) cultures. In addition, if we suggest that dynorphin facilitates the release of 151 neurortransmitters from DRG by increasing [Ca^i, then whether dynorphin produces an increase in [Ca^i in DRG cells needs to be tested experimentally. Furthermore, microdialysis studies suggest NMDA receptor antagonist can block the release of EAAs produced by i.th. injection of dynorphin (Stalling et al, 1992), this would imply that the release of EAAs from the sensory terminals is also dependent on NMDARs, which has to be tested with DRG cultures. If this is true, this may also partially explain the sensitivity of non-opioid actions of dynorphin to NMDAR antagonists. It is also possible that an increase in [Ca^^i produced by dynorphin may also potentiate NMDARs in the afiTerent nerve terminal via PKC pathway as described below.

An increase in [Ca^i may also modulate the activities of ion chaimels or receptors in the projecting neurons in the dorsal horn. For example, it may activate PKC, and phosphorylation of NMDARs by PKC has been showed to potentiate NMDARs. To pursue along this direction further, it is necessary to extend our finding with cortical cultures to spinal cord cultures. Then the role of PKC in dynorphin mediated effects can be examined using PKC inhibitors.

Finally, an important direction for future studies is to determine how dynorphin produces an increase in [Ca^i. Whether its action is dependent on the extracellular calcium, or it acts on a calcium channel, or it acts through a Gq coupled receptor, or it causes depolarization of cells by acting other targets. In this regard, the interaction 152 between dynorphin and Ca^*ATPase as suggested by a study (Yamasald and Way. 1985) is worthy of mention. 153

Afferent term ins

Dynorphin

Project neuron intemeuron;

Figure 7.2. Speculations on possible effects of dynorphin on neurotransmission in the dorsal horn of the spinal cord. indicates a stimulatory action, while indicates an inhibitory action. 154

APPENDIX 1. ABBREVIATIONS AND DEFINITIONS

AP-1, Activated protein

ATF, Activating transcription &ctor

BSA Bovine serum albumin

[Ca^^i, Intracellular calcium concentration

CaMKn, Calmodulin dependent kinase n

CCI, chronic constriction injury

CGP-39653, 2-amino-4-propyl-5-phospho-3-pentanoic acid

CKA, 7-chlorolQiiurenic acid, an antagonist at the glycine site of the NMDA receptor

CNS, Central nervous system

CRE, cyclic AMP responsive elements

CREB, CRE binding proteins

DAMGO, [D-Ala^,MePhe^,Gly-ol']enkephalin

EC50, Half maximal effective concentration

EDTA Ethylenediaminetetraacetic acid

ERKs, extracellular signal regulated kinase

HBS, HEPES buffered saline

JNKs, Jun amino-terminal kinases

ICso, Half maximal inhibition concentration

ILs, Interleukines if, Immunoreactivity 155

Kd, Dissociation constant

Ki, Inhibition constant

LTP, Long term potentiation

MAP kinase, Nfitogen Activated Phosphorylation Kinase

MK801, (+)-5-methyI-10,11-dihydro-5H-dibenzo[a,d]-cyclohepten-5,10-imine

NF-KB, a Nuclear Factor binds to a 10 bp sequence in the enhancer of the gene encoding tc light chain of antibody in B cells

NGF, Nerve growth factor

NMDA, N-methyl-D-aspartate

NO, Nitric oxide

PBS, Phosphate buffered saline pCl-DPDPE, {d-Pen^, pCl-Phe*, D-Pen']enkephalin, a selective delta opioid receptor agonist

PDYN, Prodynorphin

PKA, cAMP dependent protein kinase

PKC, Protein kinase C

PKG, cGMP dependent protein kinase

PMSF, Phenylmethylsulfonylfluoride, an protease inhibitor

SNL, Spinal nerve ligation

TGF-P, Transforming growth factor, a member of cytokine family

TNF, Tumor necrosis factor, a member of cytokine family

U69,593, (+)-(5<*,7a,83)-N-methyI-N-[7-(l-pyrrolidinyl-l-oxaspiro[4,5]dec-8- yl]benzeneacetamide), a opioid K receptor agonist 156

APPENDIX 2. LIST OF PUBLICATIONS

F.Poireca, Q.Tang, D^ian, M. Riedl, R. Elde, J. Lai. Spinal opioid mu receptor expression in lumbar spinal cord of rats following nerve injury. Brain Research. 795 (1998); 197-203

Tang, Q., R_ Gandhoke, A. Burritt, V. J. Hruby, F. Porreca, and J. Lai. High-Affinity Interaction of (des-Tyrosyl)Dynorphin A(2-17) with NMD A Receptors. J Pharmacol Exp Ther 291: 760-765, 1999.

Q. Tang, R.M. Lynch, F. Porreca, J. Lai. Dynorphin A elicits an increase in intracellular calcium in cultured neurons via a non-opioid, non-NMDA mechanism. J. Neurophysiol. (submitted). 157

REFERENCES

Arber, S., F.A. Barbayannis, H. Hanser, C. Schneider, C.A. Stanyon, O. Bernard, and P. Caroni. 1998. Regulation of actin dynamics through phosphorylation of cofilin by LIM- kinase [see comments]. Nature. 393:805-9.

Amida, J.L., R.W. Colbum, A.J. Rickman, M.D. Rutkowski, and J.A. DeLeo. 1998. Increase of interleukin-6 mRNA in the spinal cord following peripheral nerve injury in the rat: potential role of lL-6 in neuropathic pain. Brain ResMolBrain Res. 62:228-35.

Arvidsson, U., M. Riedl, S. Chakrabarti, J.H. Lee, A.H. Nakano, R.J. Dado, H.H. Loh, P.Y. Law, M.W. Wessendor^ and R. Elde. 1995. Distribution and targeting of a mu- opioid receptor (MORI) in brain and spinal cord. JNeurosci. 15:3328-41.

Baeuerie, P.A., and T. Henkel. 1994. Function and activation of NF-kappa B in the immune system. Arum Rev Immunol. 12:141-79.

Bakshi, R., and A.L Faden. 1990. Competitive and non-competitive NMD A antagonists limit dynorphin A- induced rat hindlimb paralysis. Brain Res. 507:1-5.

Bakshi, R., A.H. Newman, and A.L Faden. 1990. Dynorphin A-(l-17) induces alterations in free fatty acids, excitatory amino acids, and motor function through an opiate-receptor- mediated mechanism. JNeurosci. 10:3793-800.

Baron, C., C. Benes, H. Van Tan, R. Fagard, and M.P. Roisin. 1996. Potassium chloride pulse enhances mitogen-activated protein kinase activity in rat hippocampal slices. J Neurochem. 66:1005-10.

Baron, R., and M. Saguer. 1993. Postherpetic neuralgia. Are C-nociceptors involved in signalling and maintenance of tactile allodynia? Brain. 116:1477-96.

Basbaum, A.I., and P.D. Wall. 1976. Chronic changes in the response of cells in adult cat dorsal horn following partial deafferentation: the appearance of responding cells in a previously non-responsive region. Brain Res. 116:181-204.

Baskin, D.S., Y. Hosobuchi, H.H. Loh, and N.M. Lee. 1984. Dynorphin(l-13) improves survival in cats with focal cerebral ischaemia. Nature. 312:551-2.

Baskin, D.S., H. Kuroda, Y. Hosobuchi, and N.M. Lee. 1985. Treatment of stroke with opiate antagonists—effects of exogenous antagonists and dynorphin 1-13. Neuropeptides. 5:307-10.

Berhow, M.T., N. Ifiroi, and E.J. Nestler. 1996. Regulation of ERK (extracellular signal regulated kinase), part of the neurotrophin signal transduction cascade, in the rat 158 mesolimbic dopamine system by chronic exposure to morphine or cocaine. J Nettrosci. 16:4707-15.

Bethea, J.R., M. Castro, R.W. Keane, T.T. Lee, W.D. Dietrich, and R.P. Yezierski. 1998. Traumatic spinal cord injury induces nuclear factor-kappaB activation. J Nettrosci. 18:3251-60.

Bhisitkul, R.B., J.D. Kocsis, T.R. Gordon, and S.G. Waxman. 1990. Trophic influence of the distal nerve segment on GABAA receptor expression in axotomiz^ adult sensory neurons. Exp Neurol. 109:273-8.

Blum, S., A.N. Moore, F. Adams, and PJC. Dash. 1999. A mitogen-activated protein kinase cascade in the CA1/CA2 subfield of the dorsal hippocampus is essential for long- term spatial memory. J Nettrosci. 19:3535-44.

Boulton, T.G., S.H. Nye, D.J. Robbins, N.Y. Ip, E. Radziejewska, SJ). Morgenbesser, R.A. DePinho, N. Panayotatos, MJl Cobb, and G.D. Yancopoulos. 1991. ERKs: a family of protein-serine/threonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell. 65:663-75.

Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 72:248- 54.

Brambilla, R., N. Gnesutta, L. Minichiello, G. White, A.J. Roylance, C.E. Herron, M. Ramsey, D.P. Wolfer, V. Cestari, C. RossL-Amaud, S.G. Grant, P.F. Chapman, HJ. Lipp, E. Sturani, and R. Klein. 1997. A role for the Ras signalling pathway in synaptic transmission and long- term memory. Nature. 390:281-6.

Brauneis, U., M. Oz, R.W. Peoples, F.F. Weight, and L. Zhang. 1996. Differential sensitivity of recombinant N-methyl-D-aspartate receptor subunits to inhibition by dynorphin. J Pharmacol Exp Ther. 279:1063-8.

Burt, A.R., LC. Carr, I. Mullaney, N.G. Anderson, and G. Milligan. 1996. Agonist activation of p42 and p44 mitogen-activated protein kinases following expression of the mouse delta opioid receptor in Rat-1 fibroblasts: effects of receptor expression levels and comparisons with G-protein activation. Biochem J. 320:227-35.

Castillo, P.E., P.A. Salin, M.G. Weisskop^ and R.A. Nicoll. 1996. Characterizing the site and mode of action of dynorphin at hippocampal mossy fiber synapses in the guinea pig. J Neurosci. 16:5942-50. 159

Caudle, R.M., C. Chavkin, and R. Dubner. 1994. Kappa 2 opioid receptors inhibit NMDA receptor-mediated synaptic currents in guinea pig CA3 pyramidal cells. J Neurosci. 14:5580-9.

Caudle, R.M., and L. Isaac. 1987. Intrathecal dynorphin(l-13) results in an irreversible loss of the tail- flick reflex in rats. Brain Res. 435:1-6.

Caudle, R.M., and L. Isaac. 1988. Influence of dynorphin (1-13) on spinal reflexes in the rat. J Pharmacol Exp Ther. 246:508-13.

Chavldn, C., S.J. Henriksen, G.R. Siggins, and F.E. Bloom. 1985. Selective inactivation of opioid receptors in rat hippocampus demonstrates that dynorphin-A and -B may act on mu-receptors in the CAl region. Brain Res. 331:366-70.

Chavldn, C., I.F. James, and A. Goldstein. 1982. Dynorphin is a specific endogenous Hgand of the kappa opioid receptor. Science. 215:413-5.

Chen, L., Y. Gu, and L.Y. Huang. 1995. The mechanism of action for the block of NMDA receptor channels by the opioid peptide dynorphin. J Neurosci. 15:4602-11.

Chen, L., and L.Y. Huang. 1992. Protein kinase C reduces Mg2+ block of NMDA- receptor channels as a mechanism of modulation. Nature. 356:521-3.

Choca, J.I., R.D. Green, and H.K. Proudfit. 1988. Adenosine A1 and A2 receptors of the substantia gelatinosa are located predominantly on intrinsic neurons: an autoradiography study. J Pharmacol Exp Ther. 24T.1S1-64.

Chung, K., B.H. Lee, Y.W. Yoon, and J.M. Chung. 1996. Sympathetic sprouting in the dorsal root ganglia of the injured peripheral nerve in a rat neuropathic pain model. J Camp Neurol. 376:241-52.

Civelli, O., J. Douglass, A. Goldstein, and E. Herbert. 1985. Sequence and expression of the rat prodynorphin gene. Proc Natl Acad Sci USA. 82:4291-5.

Claude, P., Gracia, N., Wagner, L. and Hargreaves, K.M. 1999. Effect of dynorphin on ICGRP release from capsaicin-sensitive fibers. World Congress on pain (9th: 1999: Vienna, Austria) Abstracts, p262, abs#8, lASP press, 1999

Clemens, J.A., D.T. Stephenson, E.P. Dixon, E.B. Smalstig, R.E. Mincy, K.S. Rash, and S.P. Little. 1997. Global cerebral ischemia activates nuclear factor-kappa B prior to evidence of DNA Cementation. Brain Res Mol Brain Res. 48:187-96. 160

Coderre, T.J., and R. Melzack. 1992. The contribution of excitatory amino acids to central sensitization and persistent nociception after formalin-induced tissue injury. J Neurosci. 12:3665-70.

Coogan, A.N., D.M. 0*Leary, and J.J. O'Connor. 1999. P42/44 MAP kinase inhibitor PD980S9 attenuates multiple forms of synaptic plasticity in rat dentate gyrus in vitro. J Neurophysiol. 81:103-10.

Cook, A.J., C.J. Woolf^ and P.D. Wall. 1986. Prolonged C-flbre mediated facilitation of the flexion reflex in the rat is not due to changes in afferent terminal or motoneurone excitability. Neurosci Lett. 70:91-6.

Cui, J.G., A. Sollevi, B. Linderoth, and B.A. Meyerson. 1997. Adenosine receptor activation suppresses tactile hypersensitivity and potentiates spinal cord stimulation in mononeuropathic rats. Neurosci Lett. 223:173-6.

Darnell, J.E., Jr., I.M. Kerr, and G.R. Stark. 1994. Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science. 264:1415-21.

Davis, R.J. 1993. The mitogen-activated protein kinase signal transduction pathway. J BiolChem. 268:14553-6.

DeLeo, J.A, R.W. Colbum, M. Nichols, and A. Malhotra. 1996. Interleukin-6-mediated hyperalgesia/allodynia and increased spinal IL-6 expression in a rat mononeuropathy model. J Interferon Cytokine Res. 16:695-700.

DeLeo, J.A, R.W. Colbum, and A J. Rickman. 1997. Cytokine and growth factor immunohistochemical spinal profiles in two animal models of mononeuropathy. Brain Res. 759:50-7.

Devor, M. 1991. Neuropathic pain and injured nerve; peripheral mechanisms. Br Med Bull. 47:619-30.

Devor, M., R. Govrin-Lippmann, and K. Angelides. 1993. Na+ channel immunolocalization in peripheral mammalian axons and changes following nerve injury and neuroma formation. J Neurosci. 13:1976-92.

Devor, M., and P.D. Wall. 1990. Cross-excitation in dorsal root ganglia of nerve-injured and intact rats. JNeuropf^siol. 64:1733-46.

Ding, Y.Q., T. Kaneko, S. Nomura, and N. Mizuno. 1996. Immunohistochemical localization of mu-opioid receptors in the central nervous system of the rat. J Comp Neurol. 367:375-402. 161

Donaldson, L.F., C.A. Haskell, and M.R. Hanley. 1996. Functional characterization by heterologous expression of a novel cloned tachykinin peptide receptor. Biochem J. 320:1- 5.

Douglass, J., C.T. McMurray, J.E. Garrett, J.P. Adelman, and L. Calavetta. 1989. Characterization of the rat prodynorphin gene. Kfol Endocrinol. 3:2070-8.

Draisci, G., and M.J. ladarola. 1989. Temporal analysis of increases in c-fos, preprodynorphin and preproenkephalin mRNAs in rat spinal cord. Brain Res Mol Brain Res. 6:31-7.

Draisci, G., K.C. Kajander, R. Dubner, G.J. Bennett, and M.J. ladarola. 1991. Up- regulation of opioid gene expression in spinal cord evoked by experimental nerve injuries and inflammation, ^roin560:186-92.

Dubner, R., and M.A. Ruda. 1992. Activity-dependent neuronal plasticity following tissue injury and inflammation. Trends Neurosci. 15:96-103.

Dumont, M., and S. Lemaire. 1993. Characterization of non-opioid [3H]d)morphin A-(l- 13) binding sites in the rat heart. J Mol Cell Cardiol. 25:983-91.

Dumont, M., and S. Lemaire. 1994. Dynorphin potentiation of [3H]CGP-39653 binding to rat brain membranes. Eur J Pharmacol. 271:241-4.

Dumont, M., and S. Lemaire. 1996. Interactions of dynorphin A and related peptides with cardiac ouabain binding sites. J Mol Cell Cardiol. 28:615-21.

English, J.D., and J.D. Sweatt. 1996. Activation of p42 mitogen-activated protein kinase in hippocampal long term potentiation. J Biol Chem. 271:24329-32.

Faden, A.I. 1992. Dynorphin increases extracellular levels of excitatory amino acids in the brain through a non-opioid mechanism. J Neurosci. 12:425-9.

Faden, A.I., and T.P. Jacobs. 1984. Dynorphin-related peptides cause motor dysfunction in the rat through a non-opiate action. Br J Pharmacol. 81:271-6.

Faden, A.I., C.J. Molineaux, J.G. Rosenberger, T.P. Jacobs, and B.M. Cox. 1985. Increased dynorphin immunoreactivity in spinal cord after traumatic injury. Regul Pept. 11:35-41.

Farooque, M., J. Isaksson, and Y. Olsson. 1999. Improved recovery after spinal cord trauma in ICAM-1 and P-selectin knockout mice. Neuroreport. 10:131-4. 162

Feuerstein, G., and A.I. Faden. 1984. Cardiovascular effects of dynorphin A-(I-8), dynorphin A-(l-13) and dynorphin A-(l-17) microinjected into the preoptic medialis nucleus of the rat. Neuropeptides. 5:295-8.

Fiore, R.S., V.E. Bayer, SX. Pelech, J. Posada, J.A. Cooper, and JAl. Baraban. 1993. p42 mitogen-activated protein kinase in brain; prominent Ionization in neuronal cell bodies and dendrites. Neuroscience. 55:463-72.

Fischli, W., A. Goldstein, M.W. Hunkapiller, and L.£. Hood. 1982. Isolation and amino acid sequence analysis of a 4,000-dalton dynorphin from porcine pituitary. Proc Natl AcadSci USA. 79:5435-7.

Frangiskakis, J.M., AK. Ewart, C.A Morris, C.B. Mervis, J. Bertrand, B.F. Robinson, B.P. Klein, G.J. Ensing, L.A Everett, E.D. Green, C. Proschel, N.J. Gutowski, M. Noble, D.L. Atkinson, S.J. Odelberg, and M.T. Keating. 1996. LlM-kinasel hemizygosity implicated in impaired visuospatial constructive cognition. Cell. 86:59-69.

Friedman, H.J., M.F. Jei^ J.K. Chang, N.M. Lee, and HJI. Loh. 1981. Dynorphin: a possible modulatory peptide on morphine or beta-endorphin analgesia in mouse. Eur J Pharmacol. 69:357-60.

Gardner, AM., R.R. Vaillancourt, and G.L. Johnson. 1993. Activation of mitogen- activated protein Idnase/extracellular signal- regulated kinase kinase by G protein and tyrosine kinase oncoproteins. J Biol Chem. 268:17896-901.

Garrett, J.E., and J.O. Douglass. 1989. Human chorionic gonadotropin regulates expression of the proenkephalin gene in adult rat Leydig cells. Mol Endocrinol. 3:2093- ICQ.

Gautret, B., and H. Schmitt. 1985. Central and peripheral sites for cardiovascular actions of dynorphin-(l- 13) in rats. Eur J Pharmacol. Ill :263-6.

Gilbeau, P.M., Y. Hosobuchi, and N.M. Lee. 1986. Dynorphin effects on plasma concentrations of anterior pituitary hormones in the nonhuman primate. J Pharmacol Exp Ther. 238:974-7.

Goldstein, A, S. Tachibana, L.L Lowney, M. Hunkapiller, and L. Hood. 1979. Dynorphin-(l-13), an extraordinarily potent opioid peptide. Proc Natl Acad Sci USA. 76:6666-70.

Gonzalez, G.A., and M.R. Montminy. 1989. Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at serine 133. Cell. 59:675-80. 163

Ckisnell, B.A., J.E. Moriey, and A.S. Levine. 1986. Opioid-induced feeding: localization of sensitive brain sites. Brain Res. 369:177-84.

Graves, J.D., K.E. Draves, A. Craxton, J. Saklatvala, E.G. Krebs, and E.A. Clark. 1996. Involvement of stress-activated protein kinase and p38 mitogen- activated protein kinase in mIgM-induced apoptosis of human B lymphocytes. Proc Natl Acad Sci USA. 93:13814-8.

Gross, R,A., H.C. Moises, M.D. Uhler, and R.L. Macdonald. 1990. Dynorphin A and cAMP-dependent protein kinase independently regulate neuronal calcium currents. Proc Natl Acad Sci USA. 87:7025-9.

Guerrini, L., F. Blasi, and S. Denis-Donini. 1995. Synaptic activation of NF-kappa B by glutamate in cerebellar granule neurons in vitro. Proc Natl Acad Sci USA. 92:9077-81.

Han, J.S., and C.W. Xie. 1984. Dynorphin: potent analgesic effect in spinal cord of the rat. Sci Sin [B]. 27:169-77.

Harding, A.E. 1995. From the syndrome of Charcot, Marie and Tooth to disorders of peripheral myelin proteins. Brain. 118:809-18.

Hargreaves, K., R. Dubner, F. Brown, C. Flores, and J. Joris. 1988. A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain. 32:77-88.

Hawes, B.E., S. Fried, X. Yao, B. Weig, and M.P. Graziano. 1998. Nociceptin (ORL-1) and mu-opioid receptors mediate mitogen-activated protein kinase activation in CHO cells through a Gi-coupled signaling pathway: evidence for distinct mechanisms of agonist-mediated desensitization. JNeurochem. 71:1024-33.

Hayes, A.G., M. Skingle, and M3. Tyers. 1983. Antinociceptive profile of dynorphin in the rat. Life Sci. 33:657-60.

Henriksen, S.J., G. Chouvet, and F.E. Bloom. 1982. In vivo cellular responses to electrophoretically applied dynorphin in the rat hippocampus. Life Sci. 31:1785-8.

Herman, B.H., and A Goldstein. 1985. Antinociception and paralysis induced by intrathecal dynorphin A. J Pharmacol Exp Ther. 232:27-32.

Herman, B.H., F. Leslie, and A. Goldstein. 1980. Behavioral effects and in vivo degradation of intraventricularly administered dynorphin-(l-13) and D-Ala2-dynorphin- (1-11) in rats. Life Sci. 27:883-92. 164

Herrera-Marschitz, M., T. Hokfelt, U. Ungerstedt, L. Terenius, and M. Goldstein. 1984. Effect of intranigral injections of dynoiphin, dynorphin fragments and alpha- neoendorphin on rotational behaviour in the rat. Eur J Pharmacol. 102:213-27.

Hollt, v., and G. Horn. 1989. Nicotine and opioid peptides. Prog Brain Res. 79:187-93.

Horikawa, S., T. Takai, M. Toyosato, H. Takahashi, M. Noda, H. Kakidani, T. Kubo, T. Hirose, S. Inayama, H. Hayashida, and et al. 1983. Isolation and structural organization of the human preproenkephalin B gene. Nature. 306:611-4.

Hoskins, B., and I.K. Ho. 1986. Lack of effect of dynorphin on consummatory behaviors in obese and normal rats. Life Sci. 39:589-93.

Hou, Y.N., M. Vlaskovska, G. Cd)ers, L. Kasakov, S. Liljequist, and L. Terenius. 1996. A mu-receptor opioid agonist induces AP-1 and NF-kappa B transcription factor activity in primary cultures of rat cortical neurons. Neurosci Lett. 212:159-62.

Hutchison, W.D., C.R. Morton, and L. Terenius. 1990. Dynorphin A: in vivo release in the spinal cord of the cat. Brain Res. 532:299-306.

Hylden, J.L., K. Noguchi, and M.A. Ruda. 1992. Neonatal capsaicin treatment attenuates spinal Fos activation and dynorphin gene expression following peripheral tissue inflammation and hyperalgesia. J Neurosci. 12:1716-25. ladarola, M.J., L.S. Brady, G. Draisci, and R. Dubner. 1988. Enhancement of dynorphin gene expression in spinal cord following experimental inflammation: stimulus specificity, behavioral parameters and opioid receptor binding. Pain. 35:313-26. ladarola, M.J., J. Douglass, O. Civelli, and J.R. Naranjo. 1986. Increased spinal cord dynorphin mRNA during peripheral inflammation. NIDA Res Kionogr. 75:406-9.

Kajander, K.C., and G.J. Bennett. 1992. Onset of a painful peripheral neuropathy in rat: a partial and differential deafferentation and spontaneous discharge in A beta and A delta primary afferent neurons. JNeurophysioL 68:734-44.

Kajander, K.C., Y. Sahara, M.J. ladarola, and G.J. Bennett. 1990. Dynorphin increases in the dorsal spinal cord in rats with a painful peripheral neuropathy. Peptides. 11:719-28.

Kajander, K.C., S. Wakisaka, and G.J. Bennett. 1992. Spontaneous discharge originates in the dorsal root ganglion at the onset of a painful peripheral neuropathy in the rat. Neurosci Lett. 138:225-8. 165

Kakidani, H., Y. Funitani, H. Takahashi, M. Noda, Y Morimoto, T. Hirose, M. Asai, S. Inayama, S. Nakanishi, and S. Numa. 1982. Cloning and sequence analysis of cDNA for porcine beta-neo- endorphin/dynorphin precursor. Nature. 298:245-9.

Kaltschmidt, B., M. Uherek, B. Volk, P.A. Baeuerle, and C. Kaltschmidt. 1997. Transcription factor NF-kappaB is activated in primary neurons by amyloid beta peptides and in neurons surrounding early plaques from patients with ^zheimer disease. Proc Natl Acad Sci USA. 94:2642-7.

Kaneko, T., T. Nakazawa, M. Ikeda, K. Yamatsu, T. Iwama, T. Wada, M. Satoh, and H. Takagi. 1983. Sites of analgesic action of dynorphin. Life Sci. 33:661-4.

Kangawa, K., and H. Matsuo. 1979. alpha-Neo-endorphin : a "big" Leu-enkephalin with potent opiate activity from porcine hypothalami. Biochem Biopf^s Res Comtnun. 86:153- 60.

Karin, M. 1995. The regulation of AP-1 activity by mitogen-activated protein kinases. J Biol Chem. 270:16483-6.

Karin, M., Z. Liu, and E. Zandi. 1997. AP-1 function and regulation. Curr Opin Cell Biol. 9:240-6.

Khazan, N., G.A. Young, and D. Calligaro. 1983. Self-administration of dynorphin-[l- 13] and D-ala2-dynorphin-[l-ll] (kappa opioid agonists) in morphine (mu opioid agonist)-dependent rats. Life Sci. 33:559-62.

Kiang, J.G., and E.T. Wei. 1984. Sensitivity to morphine-evoked bradycardia in rats is modified by dynorphin (1-13), Leu- and Met-enkephalin. J Pharmacol Exp Ther. 229:469-73.

Kilpatrick, D.L., A. Wahlstrom, H.W. Lahm, R. Blacher, and S. Udenfriend. 1982. Rimorphin, a unique, naturally occurring [Leu]enkephalin-containing peptide found in association with dynorphin and alpha-neo-endorphin. Proc Natl Acad Sci USA. 79:6480- 3.

Kim, S.H., and J.M. Chung. 1992. An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain. 50:355-63.

Kinoshita, F., Y. Nakai, H. Katakami, and H. Imura. 1982. Suppressive effect of dynorphin-(l-13) on luteinizing hormone release in conscious castrated rats. Life Sci. 30:1915-9. 166

Lai, }., S.W. Ma, R.H. Zhu, R.B. Rothman, K.U. Lentes, and F. Porreca. 1994. Pharmacological characterization of the cloned Icappa opioid receptor as a kappa lb subtype. Neuroreport. 5:2161-4.

Lai, S.L., Y. Gu, and L.Y. Huang. 1998. Dynorphin uses a non-opioid mechanism to potentiate N-methyl-D- aspartate currents in single rat periaqueductal gray neurons. Neurosci Lett. 247:115-8.

Laurent, S., and H. Schmitt. 1983. Central cardiovascular effects of kappa agonists dynorphin-(l-13) and ethylketocyclazocine in the anaesthetized rat. Eur J Pharmacol. 96:165-9.

Leah, J., D. Menetrey, and J. de Pommery. 1988. neuropeptides in long ascending spinal tract cells in the rat: evidence for parallel processing of ascending information. Neuroscience. 24:195-207.

Lee, H.J., D.N. Hammond, T.H. Large, J.D. Roback, J.A. Sim, D.A. Brown, U.H Otten, and B.H. Wainer. 1990. Neuronal properties and trophic activities of immortalized hippocampal cells from embryonic and young adult mice. J Neurosci. 10:1779-87.

Lee, N.M. 1984. The role of dynorphin in narcotic tolerance mechanisms. NIDA Res Monogr. 54:162-7.

Lewin, G.R., A. RuefT, and L.M. Mendell. 1994. Peripheral and central mechanisms of NGF-induced hyperalgesia. Eur J Neurosci. 6:1903-12.

Li, J., and E.R. Perl. 1994. Adenosine inhibition of synaptic transmission in the substantia gelatinosa. JNeurophysiol. 72:1611-21.

Li, L.Y., and K.J. Chang. 1996. The stimulatory effect of opioids on mitogen-activated protein kinase in Chinese hamster ovary cells transfected to express mu-opioid receptors. Mol Pharmacol. 50:599-602.

Long, J.B., R.C. Kinney, D.S. Malcolm, G.M. Graeber, and J.W. Holaday. 1987. Intrathecal dynorphin Al-13 and dynorphin A3-13 reduce rat spinal cord blood flow by non-opioid mechanisms. Brain Res. 436:374-9.

Long, J.B., D.D. Rigamonti, B. de Costa, K.C. Rice, and A Martinez-Arizala. 1989. Dynorphin A-induced rat hindlimb paralysis and spinal cord injury are not altered by the kappa opioid antagonist nor-binaltorphimine. Brain Res. 497:155-62.

Long, J.B., D.D. Rigamonti, M.A Oleshansky, C.P. Wingfield, and A Martinez-Arizala. 1994. Dynorphin A-induced rat spinal cord injury: evidence for excitatory amino acid 167

involvement in a pharmacological model of ischemic spinal cord injury. J Pharmacol Exp Ther. 269:358-66.

MacMillan, S.J., and G. Clarke. 1983. Opioid peptides have differential actions on sub- populations of arcuate neurones. Life Sci. 33:529-32.

Makman, M.H., W.D. Lyman, and B. Dvoridn. 1997. Presence and characterization of nociceptin (orphanin FQ) receptor binding in adult rat and human fetal hypothalamus. Brain Res. 762:247-50.

Maimberg, A.B., C. Chen, S. Tonegawa, and A.I. Basbaum. 1997. Preserved acute pain and reduced neuropathic pain in mice lacking PKCgamma [see comments]. Science. 278:279-83.

Manning, B.H., J. Mao, H. Frenk, D.D. Price, and D.J. Mayer. 1996. Continuous co­ administration of dextromethorphan or MK-801 with morphine: attenuation of morphine dependence and naloxone-reversible attenuation of morphine tolerance. Pain. 67:79-88.

Mansson, £., L. Bare, and D. Yang. 1994. Isolation of a human kappa opioid receptor cDNA from placenta. Biochem Biophys Res Commun. 202:1431-7. Mantyh, P.W., S.D. Rogers, P. Honore, B.J. Allen, J.R. Ghilardi, J. Li, R.S. Daughters,

D.A. Lappi, R.G. Wiley, and D.A Simone. 1997. Inhibition of hyperalgesia by ablation of lamina I spinal neurons expressing the substance P receptor [see conmients]. Science. 278:275-9.

Mao, J., D.D. Price, and D.J. Mayer. 1994. Thermal hyperalgesia in association with the development of morphine tolerance in rats: roles of excitatory amino acid receptors and protein kinase C. JNeurosci. 14:2301-12.

Mao, J., D.D. Price, and D.J. Mayer. 1995a. Experimental mononeuropathy reduces the antinociceptive effects of morphine: implications for conunon intracellular mechanisms involved in morphine tolerance and neuropathic pain [see conunents]. Pain. 61:353-64.

Mao, J., D.D. Price, and D.J. Mayer. 1995b. Mechanisms of hyperalgesia and morphine tolerance: a current view of their possible interactions. Pain. 62:259-74.

Massardier, D., and P.F. Hunt. 1989. A direct non-opiate interaction of dynorphin-(l-13) with the N-methyl-D- aspartate (NMDA) receptor. Eur J Pharmacol. 170:125-6.

Matsushita, N., Y. Kato, A Shimatsu, H. Katakami, M. Fujino, H. Matsuo, and H. Imura. 1982. Stimulation of prolactin secretion in the rat by alpha-neo-endorphin, beta-neo- endorphin and dynorphin. Biochem Biophys Res Commun. 107:735-41. 168

Maves, T.J., P.S. Pechman, G.F. Gebhait, and S.T. Meller. 1993. Possible chemical contribution from chromic gut sutures produces disorders of pain sensation like those seen in man [published erratum appears in Pain 1993 Oct;SS(l);131-4] [see conmients]. Pain. 54:57-69.

McEachem, J.C., and C.A. Shaw. 1996. An alternative to the LTP orthodoxy: a plasticity-pathology continuum model. Brain Res Brain Res Rev. 22:51-92.

Meller, S.T., C. Dykstra, and G.F. Gebhart. 1992. Production of endogenous nitric oxide and activation of soluble guanylate cyclase are required for N-methyl-D-aspartate- produced facilitation of the nociceptive tail-flick reflex. Eur J Pharmacol. 214:93-6.

Meller, S.T., C. Dykstra, D. Grzybycki, S. Murphy, and G.F. Gebhart. 1994. The possible role of glia in nociceptive processing and hyperalgesia in the spinal cord of the rat. Neuropharmacology. 33:1471-8.

Millan, M.J., A. Czlonkowski, B. Morris, C. Stein, R. Arendt, A. Huber, V. Hollt, and A Herz. 1988. Inflammation of the hind limb as a model of unilateral, localized pain: influence on multiple opioid systems in the spinal cord of the rat. Pain. 35:299-312.

Millan, M.J., M.H. Millan, L.D. Reid, and A. Herz. 1986. The role of the mediobasal arcuate hypothalamus in relation to opioid systems in the control of ingestive behaviour in the rat. Brain Res. 381:29-42.

Millar, J., A.I. Basbaum, and P.D. Wall. 1976. Restructuring of the somatotopic map and appearance of abnormal neuronal activity in the gracile nucleus after partial deafferentation. Exp Neurol. 50:658-72.

Minamino, N., K. Kangawa, N. Chino, S. Sakakibara, and H. Matsuo. 1981. Beta-neo- endorphin, a new hypothalamic "big" Leu-enkephalin of porcine origin: its purification and the complete amino acid sequence. Biochem Biophys Res Commun. 99:864-70.

Minamino, N., K. Kangawa, A. Fukuda, H. Matsuo, and M. lagarashi. 1980. A new opioid octapeptide related to dynorphin from porcine hypothalamus. Biochem Biophys Res Commun. 95:1475-81.

Miura, M., A. Inui, K. Sano, N. Ueno, A. Teranishi, Y. Hirosue, M. Nakajima, M. Okita, J. Togami, K. Koshiya, and et al. 1994. Dynorphin binds to neuropeptide Y and peptide YY receptors in human neuroblastoma cell lines. Am J Physiol. 267:E702-9.

Moises, H.C., K.I. Rusin, and R.L. Macdonald. 1994. Mu- and kappa-opioid receptors selectively reduce the same transient components of high-threshold calcium current in rat dorsal root ganglion sensory neurons. JNeurosci. 14:5903-16. 169

Moises, H.C., and J.M. Walker. 198S. Electrophysiological effects of dynorphin peptides on hippocampal pyramidal cells in rat. Eur JPhcamacol. 108:85-98.

Molander, C., J. Hongpaisan, and G. Grant. 1992. Changing pattern of c-FOS expression in spinal cord neurons after electrical stimulation of the chronically injured sciatic nerve in the rat. Neuroscience. 50:223-36.

Morgan, J.L, and T. Curran. 1989. Stimulus-transcription coupling in neurons: role of cellular immediate-early genes. Trends Neurosci. 12:459-62.

Morley, J.E., and A.S. Levine. 1983. Involvement of dynorphin and the kappa opioid receptor in feeding. Peptides. 4:797-800.

Morley, J.E., A-S. Levine, B.A. Gosnell, and C.J. Billington. 1984. Which opioid receptor mechanism modulates feeding? Appetite. 5:61-8.

Morley, J.E., A-S. Levine, M. Grace, and J. Kneip. 1982. Dynorphin-(l-13), dopamine and feeding in rats. Pharmacol Biochem Behav. 16:701-5.

Morris, B.J., K.J. Feasey, G. ten Bruggencate, A. Herz, and V. Hollt. 1988a. Electrical stimulation in vivo increases the expression of proenkephalin mRNA and decreases the expression of prodynorphin mRNA in rat hippocampal granule cells. Proceedings of the National Accuiemy of Sciences of the United States of America. 85:3226-30.

Morris, B.J., I. Haarmann, B. Kempter, V. Hollt, and A. Herz. 1986. Localization of prodynorphin messenger ma in rat brain by in situ hybridization using a synthetic oligonucleotide probe. Neuroscience Letters. 69:104-8.

Morris, B.J., A. Herz, and V. Hollt. 1989. Localization of striatal opioid gene expression, and its modulation by the mesostriatal dopamine pathway: an in situ hybridization study. Journal of Molecular Neuroscience. 1:9-18.

Morris, B.J., V. Hollt, and A. Herz. 1988b. Opioid gene expression in rat striatum is modulated via opioid receptors: evidence from localized receptor inactivation. Neuroscience Letters. 89:80-4.

Morris, B.J., M.E. Moneta, G. ten Bruggencate, and V. Hollt. 1987. Levels of prodynorphin mRNA in rat dentate gyrus are decreased during hippocampal kindling. Neuroscience Letters. 80:298-302.

Morris, B.J., S. Reimer, V. Hollt, and A. Herz. 1988c. Regulation of striatal prodynorphin mRNA levels by the raphe-striatal pathway. Brain Research. 464:15-22. 170

Nahin, R.L. 1987. Immunocytochemical identification of long ascending peptidergic neurons contributing to the spinoreticular tract in the rat. Neuroscience. 23:859-69.

Nahin, R.L., J.L. Hylden, and E. Humphrey. 1992. Demonstration of dynorphin A 1-8 immunoreactive axons contacting spinal cord projection neurons in a rat model of peripheral inflammation and hyperalgesia. Pain. 51:135-43.

Nahin, R.L., J.L. Hylden, M.J. ladarola, and R. Dubner. 1989. Peripheral inflammation is associated with increased dynorphin immunoreactivity in both projection and local circuit neurons in the superficial dor^ horn of the rat lumbar spinal cord. Neurosci Lett. 96:247-52.

Nakazawa, T., M. Ikeda, T. Kaneko, and K. Yamatsu. 1985. Analgesic effects of dynorphin-A and morphine in mice. Peptides. 6:75-8.

Naranjo, J.R., B. Mellstrom, M. Achaval, and P. Sassone-Corsi. 1991. Molecular pathways of pain: Fos/Jun-mediated activation of a noncanonical AP-1 site in the prodynorphin gene. Neuron. 6:607-17.

Nichols, M.L., Y. Lopez, M.H. Ossipov, D. Bian, and F. Porreca. 1997. Enhancement of the antiallodynic and antinociceptive efficacy of spinal morphine by antisera to dynorphin A (1-13) orMK-801 in a nerve- ligation model of peripheral neuropathy. Pain. 69:317-22.

Nicholson, G.A., L.J. Valentijn, A.K. Cherryson, M.L. Kennerson, T.L. Bragg, R.M. DeKroon, D.A. Ross, J.D. Pollard, J.G. McLrod, P.A. Bolhuis, and et al. 1994. A frame shift mutation in the PMP22 gene in hereditary neuropathy with liability to pressure palsies [published erratum appears in Nat Genet 1994 May;7(l):113]. Nat Genet. 6:263- 6.

Noguchi, K., K. Kowalski, R. Traub, A. Solodkin, M.J. ladarola, and M.A. Ruda. 1991. Dynorphin expression and Fos-like immunoreactivity following inflammation induced hyperalgesia are colocalized in spinal cord neurons. Brain ResMolBrain Res. 10:227-33.

Oka, T., S. Aou, and T. Hori. 1994. Intracerebroventricular injection of interleukin-1 beta enhances nociceptive neuronal responses of the trigeminal nucleus caudalis in rats. Brain Res. 656:236-44.

O'Neill, L.A., and C. Kaltschmidt. 1997. NF-kappa B: a crucial transcription factor for glial and neuronal cell function [see comments]. Trends Neurosci. 20:252-8.

Ossipov, M.H., Y. Lopez, M.L. Nichols, D. Bian, and F. Porreca. 1995. Inhibition by spin^ morphine of the tail-flick response is attenuated in rats with nerve ligation injury. Neurosci Lett. 199:83-6. 171

Petrie, E.C., S.T. Tiffany, T.B. Baker, and J.L. Dahl. 1982. Dynorphin (1-13); analgesia, hypothermia, cross-tolerance with morphine and beta-endorphin. Peptides. 3:41-7.

Przewlocka, B., L. Stala, W. Lason, and R. Przewlocki. 1983. The effect of various opiate receptor agonists on the seizure threshold in the rat. Is dynorphin an endogenous anticonvulsant? Life Sci. 33:595-8.

Przewlocki, R., I. Haarmann, K. Nikolarakis, A. Herz, and V. Hollt. 1988. Prodynorphin gene expression in spinal cord is enhanced after traumatic injury in the rat. Brcan Res. 464:37-41.

Przewlocki, R., L. Stala, M. Greczek, G.T. Shearman, B. Przewlocka, and A. Herz. 1983. Analgesic effects of mu-, delta- and kappa-opiate agonists and, in particular, dynorphin at the spinal level. Life Sci. 33:649-52.

Ransom, R.W., and N.L. Stec. 1988. Cooperative modulation of [3H]MK-801 binding to the N-methyl-D- aspartate receptor-ion channel complex by L-glutamate, glycine, and polyamines. JNeurochem. 51:830-6.

Raynor, K., H Kong, A Mestek, L.S. Bye, M. Tian, J. Liu, L. Yu, and T. Reisine. 1995. Characterization of the cloned human mu opioid receptor. J Pharmacol Exp Ther. 272:423-8.

Reimer, S., and V. Hollt. 1991. GABAergic regulation of striatal opioid gene expression. Brain ResMolBrain Res. 10:49-54.

Reynolds, I.J., K.M. Harris, and RJ. Miller. 1989. NMDA receptor antagonists that bind to the strychnine-insensitive glycine site and inhibit NMDA-induced Ca2+ fluxes and [3H]GABA release. Eur J Pharmacol. 172:9-17.

Reynolds, I.J., and R.J. Miller. 1989. Ifenprodil is a novel type of N-methyl-D-aspartate receptor antagonist: interaction with polyamines. Mol Pharmacol. 36:758-65.

Roberts, W.J. 1986. A hypothesis on the physiological basis for causalgia and related pains. Pain. 24:297-311.

Rosen, L.B., D.D. Ginty, M.J. Weber, and M.E. Greenberg. 1994. Membrane depolarization and calcium influx stimulate MEK and MAP kinase via activation of Ras. Neuron. 12:1207-21.

Ruda, M.A, M.J. ladarola, L.V. Cohen, and W.S.d. Young. 1988. In situ hybridization histochemistry and immunocytochemistry reveal an increase in spinal dynorphin 172 biosynthesis in a rat model of peripheral inflammation and hyperalgesia. Proc Natl Acad Sci USA. 85:622-6.

Sah, D.W. 1990. Neurotransmitter modulation of calcium current in rat spinal cord neurons. y^Afeurascf. 10:136-41.

Salminen, A., P.K. Liu, and C.Y. Hsu. 199S. Alteration of transcription factor binding activities in the ischemic rat brain. Biochem BiophysRes Commun. 212:939-44.

Sato, J., and E.R. Perl. 1991. Adrenergic excitation of cutaneous pain receptors induced by peripheral nerve injury. Science. 251:1608-10.

Sawynok, J. 1998. Adenosine receptor activation and nociception. Eur J Pharmacol. 347:1-11.

Sawynok, J., and M.I. Sweeney. 1989. The role of purines in nociception. Neuroscience. 32:557-69.

Seger, R., and E.G. Krebs. 1995. The MAPK signaling cascade. Faseb J. 9:726-35.

Sen, R., and D. Baltimore. 1986. Inducibility of kappa immunoglobulin enhancer-binding protein Nf-kappaB by a posttranslational mechanism. Cell. 47:921-8.

Sheng, M., M.A. Thompson, and M.E. Greenberg. 1991. CREB: a Ca(2+)-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science. 252:1427- 30.

Sherman, T.G., O. Civelli, J. Douglass, E. Herbert, S. Burke, and S.J. Watson. 1986. Hypothalamic dynorphin and vasopressin mRNA expression in normal and Brattleboro rats. Fed Proc. 45:2323-7.

Shortland, P., and C.J. Woolf 1993. Chronic peripheral nerve section results in a rearrangement of the central axonal arborizations of axotomized A beta primary afferent neurons in the rat spinal cord. JComp Neurol. 330:65-82.

Shukia, V.K., M. Bansinath, M. Dumont, and S. Lemaire. 1992. Selective involvement of kappa opioid and phencyclidine receptors in the analgesic and motor effects of dynorphin-A-(l-13)-Tyr-Leu-Phe-Asn- Gly-Pro. Brain Res. 591:176-80.

Shukia, V.K., and S. Lemaire. 1992. Central non-opioid physiological and pathophysiological effects of dynorphin A and related peptides. J Psychiatry Neurosci. 17:106-19. 173

Simonin, F., C. Gaveriaux-Ruff, K. Befort, H. Matthes, B. Lannes, G. Micheletti, M.G. Mattei, G. Charron, B. Bloch, and B. Kieffer. 1995. kappa-Opioid receptor in humans: cDNA and genomic cloning, chromosomal assignment, functional expression, pharmacology, and expression pattern in the central nervous system. Proc Natl Acad Sci USA. 92:7006-10.

Sivilotti, L., and C.J. Woolf. 1994. The contribution of GABAA and glycine receptors to central sensitization: disinhibition and touch-evoked allodynia in the spinal cord. J Neurophysiol. 72:169-79.

Skilling, S.R., X. Sun, H.J. Kurtz, and A.A. Larson. 1992. Selective potentiation of NMDA-induced activity and release of excitatory amino acids by dynorphin: possible roles in paralysis and neurotoxicity. Brain Res. 575:272-8.

Smith, K.E., L.A. Borden, P.R. Hartig, T. Branchek, and R.L. Weinshank. 1992. Cloning and expression of a glycine transporter reveal colocalization with NMDA receptors. Neuron. 8:927-35.

Spampinato, S., and S. Candeletti. 1985. Characterization of dynorphin A-induced antinociception at spinal level. Eur J Pharmacol. 110:21-30.

Stevens, C.W., and T.L. Yaksh. 1986. Dynorphin A and related peptides administered intrathecally in the rat: a search for putative kappa opiate receptor activity. J Pharmacol Exp Ther. 238:833-8.

Sugimoto, T., G.J. Bennett, and K.C. Kajander. 1989. Strychnine-enhanced transsynaptic degeneration of dorsal horn neurons in rats with an experimental painful peripheral neuropathy. Neurosci Lett. 98:139-43.

Sutor, B., and W. Zieglgansberger. 1984. Actions of D-Ala2-D-Leu5-enkephalin and dynorphin A (1-17) on neocortical neurons in vitro. Neuropeptides. 5:241-4.

Takemori, A.E., H.H. Loh, and N.M. Lee. 1993. Suppression by dynorphin A and [des- Tyrl]dynorphin A peptides of the expression of opiate withdrawal and tolerance in morphine-dependent mice. J Pharmacol Exp Ther. 266:121-4.

Taoka, Y., K. Okajima, M. Uchiba, K. Murakami, S. Kushimoto, M. Johno, M. Naruo, H. Okabe, and K. Takatsuki. 1997. Role of neutrophils in spinal cord injury in the rat. Neuroscience. 79:1177-82.

Tiseo, P.J., M.W. Adler, and L.Y. Liu-Chen. 1989. Differential release of substance P and somatostatin in the rat spinal cord in response to noxious cold and heat: effect of dynorphin A(l-17). NIDA ResMonogr. 95:297. 174

Treisman, R. 1992. The serum response element. Trends Biochem Sci. 17:423-6.

Treisman, R. 1994. Ternary complex factors; growth factor regulated transcriptional activators. Curr Opin Genet Dev. 4:96-101.

Trujilio, K.A., and H. Akil. 1991. Inhibition of morphine tolerance and dependence by the NMDA receptor antagonist MK-801. Science. 251:8S-7.

Tulunay, F.C., M-F. Jen, J.K. Chang, H.H. Loh, and N.M. Lee. 1981. Possible regulatory role of dynorphin on morphine- and beta-endorphin- induced analgesia. J Pharmacol Exp Ther. 219:296-8.

Unlap, T., and R.S. Jope. 1995. Diurnal variation in kainate-induced AP-1 activation in rat brain; influence of glucocorticoids. Brain ResMol Brain Res. 28;193-200.

Vidal, C., R. Maier, and W. Zieglgansberger. 1984. Effects of dynorphin A (1-17), dynorphin A (1-13) and D-ala2-D-leu5- enkephalin on the excitability of pyramidal cells in CAl and CA2 of the rat hippocampus in vitro. Neuropeptides. 5:237-40.

Walker, J.M., H.C. Moises, D.H. Coy, G. Baldrighi, and H. Akil. 1982a. Nonopiate effects of dynorphin and des-Tyr-dynorphin. Science. 218:1136-8.

Walker, J.M., H.C. Moises, D.H. Coy, E.A. Young, S.J. Watson, and H. Akil. 1982b. Dynorphin (1-17): lack of analgesia but evidence for non-opiate electrophysiological and motor effects. Life Sci. 31;1821-4.

Walker, J.M., D.E. Tucker, D.H. Coy, B.B. Walker, and H. Akil. 1982c. Des-tyrosine- dynorphin antagonizes morphine analgesia. Eur J Pharmacol. 85:121-2.

Watkins, L.R., L.E. Goehler, J. Reiton, M.T. Brewer, and S.F. Maier. 1995a. Mechanisms of tumor necrosis factor-alpha (TNF-alpha) hyperalgesia. Brain Res. 692:244-50.

Watkins, L.R., S.F. Maier, and L.£. Goehler. 1995b. Immune activation: the role of pro­ inflammatory cytokines in inflammation, illness responses and pathological pain states. Pain. 63:289-302.

Watkins, L.R., D. Martin, P. Ulrich, K.J. Tracey, and S.F. Maier. 1997. Evidence for the involvement of spinal cord glia in subcutaneous formalin induced hyperalgesia in the rat. Pain. 71:225-35.

Watson, S.J., H. Khachaturian, H. Akil, D.H. Coy, and A. Goldstein. 1982. Comparison of the distribution of dynorphin systems and enkephalin systems in brain. Science. 218:1134-6. 175

Weihe, E., M.J. Millan, V. Hollt, D. Nohr, and A. Herz. 1989. Induction of the gene encoding pro-dynorphin by experimentally induced arthritis enhances staining for dynorphin in the spinal cord of rats. Neuroscience. 31;77-95.

Werz, M.A., and R.L. Macdonald. 1985. Dynorphin and neoendorphin peptides decrease dorsal root ganglion neuron calcium-dependent action potential duration. J Pharmacol Exp Ther. 234:49-56.

Willis, W.D., Jr. 1997. Is central sensitization of nociceptive transmission in the spinal cord a variety of long-term potentiation? [comment]. Neuroreport. 8:iii.

Wong, E.H., A.R. Knight, and G.N. Woodruff 1988. [3H]MK-801 labels a site on the N- methyl-D-aspartate receptor channel complex in rat brain membranes. J Netirochem. 50:274-81.

Woo, S.K., F.C. Tulunay, H.H. Loh, and N.M. Lee. 1983. Effect of dynorphin-(l-13) and related peptides on respiratory rate and morphine-induced respiratory rate depression. Eur J Pharmacol. 96:117-22.

Woolf, C.J., and T.P. Doubell. 1994. The pathophysiology of chronic pain—increased sensitivity to low threshold A beta-fibre inputs. Curr Opin Neurobiol. 4:525-34.

Woolf, C.J., M.L. Reynolds, C. Molander, C. O'Brien, R.M. Lindsay, and L.I. Benowitz. 1990. The growth-associated protein GAP-43 appears in dorsal root ganglion cells and in the dorsal horn of the rat spinal cord following peripheral nerve injury. Neuroscience. 34:465-78.

Woolf, C.J., P. Shortland, and R.E. Coggeshall. 1992. Peripheral nerve injury triggers central sprouting of myelinated afferents. Nature. 355:75-8.

Woolf, C.J., and S.W. Thompson. 1991. The induction and maintenance of central sensitization is dependent on N-methyl-D-aspartic acid receptor activation; implications for the treatment of post-injury pain hypersensitivity states. Pain. 44:293-9.

Woolf, C.J., and P.D. Wall. 1986. Relative effectiveness of C primary afferent fibers of different origins in evoking a prolonged facilitation of the flexor reflex in the rat. J Neurosci. 6:1433-42.

Wright, D.M., C.E. Pill, and G. Clarke. 1983. Effect of ACTH on opiate inhibition of oxytocin release. Life Sci. 33:495-8.

Xia, Z., M. Dickens, J. Raingeaud, R.J. Davis, and M.E. Greenberg. 1995. Opposing effects of ERK and Jl^-p38 MAP Idnases on apoptosis. Science. 270:1326-31. 176

Xie, C.W., J.F. McGinty, PJI. Lee, C.L. Mitchell, and J.S. Hong. 1991. A glutamate antagonist blocks perforant path stimulation-induced reduction of dynorphin peptide and prodynorphin mRNA levels in rat hippocampus. Brcrin Res. 562:243-50.

Yaksh, T.L. 1989. Behavioral and autonomic correlates of the tactile evoked allodynia produced by spinal glycine inhibition: effects of modulatory receptor systems and excitatory amino acid antagonists. Pain. 37:111-23.

Yaksh, T.L., J.C. Yeung, and T.A. Rudy. 1976. An inability to antagonize with naloxone the elevated nocicq;)tive thresholds resulting from electrical stimulation of the mesencephalic central gray. Life Sci. 18:1193-8.

Yamasaki, Y., and E.L. Way. 1983. Possible inhibition of Ca+-»- pump of rat erythrocyte ghosts by opioid k agonists. Life Sci. 33:723-6.

Yamasaki, Y., and E.L. Way. 1985. Inhibition of Ca-H--ATPase of rat erythrocyte membranes by k-opioid agonists. Neuropeptides. 5:359-62.

Yang, K., X.S. Mu, and RX. Hayes. 1995. Increased cortical nuclear factor-kappa B (NF- kappa B) DNA binding activity after traumatic brain injury in rats. Neurosci Lett. 197:101-4.

Yang, N., O. Higuchi, K. Ohashi, K. Nagata, A Wada, K. Kangawa, E. Nishida, and K. Mizuno. 1998. Cofilin phosphorylation by LIM-kinase 1 and its role in Rac-mediated actin reorganization [see comments]. Nature. 393:809-12.

Yarygin, K.N., X.H. Zhang, and N.M. Lee. 1998. Non-opioid dynorphin binding site on secretory vesicles of a pituitary- derived cell line. Breun Res. 791:99-107.

Yip, H.K., and E.M. Johnson, Jr. 1987. Nerve growth factor receptors in rat spinal cord: an autoradiographic and immunohistochemical study. Neuroscience. 22:267-79.

Young, W.S.d., T.I. Bonner, and M.R. Brann. 1986. Mesencephalic dopamine neurons regulate the expression of neuropeptide mRNAs in the rat forebrain. Proc Natl Acad Sci USA. 83:9827-31.

Zachariou, V., and B.D. Goldstein. 1997. Dynorphin-(l-8) inhibits the release of substance P-like immunoreactivity in the spinal cord of rats following a noxious mechanical stimulus. Eur JPharmacoL 323:159-65.

Zhang, L., R.W. Peoples, M. Oz, J. Harvey-White, F.F. Weight, and U. Brauneis. 1997. Potentiation of NMDA receptor-mediated responses by dynorphin at low extracellular glycine concentrations. JNeurophysioL 78:582-90.