4GRANT - DTICa dipeptide synthesized through the post-translational processing of 9-endorphin, acts ... (AChE) activity; Gly-Gln produces a comparable effect, suggesting that it may

AD-A242 587 _

GLYCYL-L-GLUTAMINE: A DIPEPTIDE NEUROTRANSMITTER

DERIVED FROM BETA-ENDORPHIN

D i MIDTERM REPORT

DTICS ELECTE WILLIAM R. MILLINGTON

U SEPTEMBER 1, 1991

Supported by

U.S. ARMY MEDICAL RESEARCH AND DEVELOPMENT COMMANDFort Detrick, Frederick, Maryland 21702-5012

4GRANT No. DAMD17-90-Z-0022

University of Missouri-Kansas City

2411 Holmes Street

Kansas City, Missouri 64108

Approved for public releasei distribution unlimited.

The findings in this report are not to be construed as anofficial Department of the Army position unless so designated

by other authorized documents

91-15749111 tIl! ll ii! 11111 Ii!llil 111 i 111

SECURITY CLASSIFICATION OF 7HiS PACE

REPORT DOCUMENTATION PAGE O i-AProved1 O0&4 No. 0 704.0 1,

14. REPORT SECURITY CLASSIFICATION 10 RESTRICT1vE MARKINGSunclassified

2a. SECURITY CLASSIFICATION AUTHORITY 3 0ISTRIBUTION/AVAILABLITY OF REPORT2Approved for public release;2b. OECLASSIFICATIONIOWNGRADING SCREDULE distribution unlimited

4. PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)

6.. NAME OF PERFORMING ORGANIZATION 1 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION

(If appficable)

Universit" of Missouri-Kansas City I6c. ADDRESS (Oty, State, and ZIP Code) 7b. ADORESS (Oty, State. and ZIP Cooe)School of Basic Life Sciences

2411 Holmes - Rocm M3-423Kansas City, MO 64108-2792

&a.. NAME OF FUNDING I SPONSORING 18b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION U.S. Army Medical (if applicable)

Research & Development Command DAD17-90-Z-O022

S8c. ADDRESS (City, State, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS

PROGRAM PROJECT TASK WORK UNIELEMENT NO. NO. 3M1- NO. CCESSIONFrederick, Maryland 21702-5012 61102A 61102BSi2 AC61102 6LI02BS1 AC DA34613

11. TITLE (Inude Security Clawuficarion)Glycyl-L-Glutamine: A Dipeptide Neurotransmitter Derived from Beta-endorphin

412. PERSONAL AUTHOR(S)

Millint on, Wi1i an R.1 3a. TYPE OF REPORT j13b. TIME COVERED 91114. DATE OF REPORT (YVear, Month, 3a) 7 . PAGE COUNTMid-Term FROM 90/3/1/ TO 8/31/91 1991 September 1

16. SUPPLEMENTARY NOTATION

17. COSATI CODES 18. SUBJECT TERMS (Continue on reverse if necessary and 4oentify by biocX numoer)FIELD GROUP SUB-GROUP Beta-Endorphi-n, Grcy1l---Glutarnine; Di Peotiden6 | O& | Propiarelanocortin; Brain; Pituitary; P A 1

19 A S7RAC" Continue on reve-se f necessary and oenrify oy olocx numoeri

The objective of this research is to test the hypothesis that glycyl-L-glutamine,a dipeptide synthesized through the post-translational processing of 9-endorphin, actsas a neurotransmitter in brain and a circulating hormone in the periphery. To test thishypothesis, we established three working objectives: (1) to evaluate the physiologicalresponses produced by glycyl-L-glutamine and test whether it modulates the opiateactions of 9-endorphin; (2) to examine its distribution in brain and pituitary anddetermine whether it is specifically co-localized with S-endorphin; (3) to characterizethe receptors which mediate glycyl-L-glutamine's pharmacologic effects. During thecurrent project period, we found that glycyl-L-glutamine produces three markedly dif-ferent pharmacologic responses: it stimulates a trophic response in cardiac myocytes,inducing expression of the asymmetric form of acetylcholinesterase, a classical markerfor synaptic innervation; it produces neuroimmune regulatory effects in T-lymphocytes,enhancing c-myc oncogene expression, a component of the T-lymphocyte proliferative

* 20. OISTRIsUriON/AVAILABILIY OF ABSTRACT 2. ABSTRACT SECURITY CLASSIFICATION

OUNCLASSIFIEDI JNLIMITED 0 SAME AS RPT C3 OTIC USERS Unclassified

22a. NAME OF RESPONSIBLE INDIVIDUAL 22b. TELEPHONE(InctUae Area Code) 22,. OFFICE SYMBOL

Mary Frances Bostian 301-663-7325 SGRD-RMI-S

00 Form 1473. JUN 86 Pvoseamonsareoosode. SECURITY CLASSIFICATION OF 'l-iS PAC

19. Abstract (continued):

response to antigens; and it inhibits 9-endorphin induced antinociception, S-endorphin'smost characteristic centrally mediated action. In addition, substantial progress hasbeen made toward establishing immunoassay and immunohistochemical methods for detectingendogenous glycyl-L-glutamine. Subsequent experiments will apply these methods to mapglycyl-L-glutamine's regional distribution in brain, determining whether it is selec-tively localized in 8-endorphin neurons, and to identify the factors which regulate itssynthesis and release. The present studies have thus demonstrated that glycyl-L-glut-amine produces pharmacologic effects in both brain and peripheral tissues, supportingthe concept that it acts as both a neurotransmitter and a circulating hormone, and havedeveloped the methods required to conclusively establish glycyl-L-glutamine's role inneural and endocrine communication.

4

I

Grant No. DAMDI7-90-Z-0022Page No. 11

FOREWORD

Opinions, interpretations, conclusions and recommendations are those of the author anaare not necessarily endorsed by the U.S. Army.

Where copyrighted material is quoted. permission has been obtained to use suchmaterial.

Where material from documents designated for limited distribution is quoted,permission has been obtained to use the material.

Citations of commercial organizations and trade names in this report do not constitutean official Department of the Army endorsement or approval of the products or services ofthese organizations.

"Z In conducting research using animals, the investigator(s) adhered to the "Guide forthe Care and Use of Laboratory Animals," prepared by the Committee on Care and Use ofLaboratory Animals of the Institute of Laboratory Animal Resources, National ResearchCouncil (NIH Publication No. 86-23, Revised 1985).

For the protection of human subjects, the investigator(s) have adhered to policies ofapplicable Federal Law 45 CFR 46.

In conducting research utilizing recombinant DNA technology, the investigator(s)adhered to current guidelines promulgated by the National Institutes of Health.

P1 Signature Date

CONUSJUN 89 -3-

TABLE OF CONTENTS

FOREWORD .............. .......................... 3

INTRODUCTION ............. ........................ 5

RESULTS AND DISCUSSION .......... ................... 6

I. Pharmacoloqic Effects ...... ................II. Analytical Methods ...... ................ 13

SUMMARY AND CONCLUSIONS ....... .................. 17

PUBLICATIONS ........... ........................ 30

REFERENCES ............ ......................... 32

FIGURES ............ .......................... 22

FIV

By

G; -, .t .:;

H 'I.. .. . ..

INTRODUCTION

Glycyl-L-glutamine is a dipeptide synthesized through thepost-translational processing of -endorphin. f-endorphin pro-cessing has been intensely studied in recent years because itprofoundly alters the peptide's analgetic activity, transforming-endorphin-l-31 from a highly potent opiate receptor agonist to

an antagonist, j-endorphin-l-27, and to opiate inactive forms, f3-endorphin-l-26 and the a-N-acetyl forms of all three peptides(Deakin et al., 1980; Akil et al., 1981; Nicolas and Li, 1985).Glycyl-L-glutamine, co-synthesized with P-endorphin-l-27 when f3-endorphin-l-31 is endoproteolytically cleaved, has not been asthoroughly evaluated as the larger G-endorphin forms, althoughexisting evidence indicates that it, too, may participate *insynaptic transmission. Indeed, many of its known effects counter-pose the opiate actions of P-endorphin-1-31 (Hirsch and O'Donohue,1986; McCain et al., 1986; Koelle et al., 1988; Lotwick et al.,1990). The objective of our research is to establish whetherglycyl-L-glutamine acts as a neurotransmitter in brain and acirculating hormone in the periphery.

Evidence that glycyl-L-glutamine functions in synaptictransmission first arose from electrophysiologic studies by Parishet al. (1983) showing that iontophoretic glycyl-L-glutamineapplication inhibited the firing frequencies of brainstem neurons.This activity was not reversed by naloxone, an opiate antagonist,or by stcychnine, which blocks receptors for glycine, one ofglycyl-L-glutamine's constituent amino acids. In addition, theseinvestigators isolated glycyl-L-glutamine from sheep brainstem,demonstrating that it is present in amounts equivalent to the sumof P-endorphin-l-27 and -1-26, as one would predict. Immunohisto-chemical studies also showed that glycyl-L-glutamine is localizedin the intermediate, but not in the anterior lobe of the pituitary(Plishka et al., 1985) where f-endorphin does not undergo C-termi-nal cleavage. (Eipper and Mains, 1980; O'Donohue and Dorsa, 1982).

Several additional lines of evidence further support theconcept that glycyl-L-glutamine acts as a neurotransmitter. Be-havioral studies revealed that glycyl-L-glutamine inhibits fl-endor-phin-l-31 induced grooming in rats, a response thought to reflectmechanisms of attention and arousal (Hirsch and O'Donohue, 1985).Glycyl-L-glutamine is also thought to function as a trophic agentat the neuromuscular junction (Lotwick et al., 1990) and in auto-nomic ganglia (Koelle et al., 1988). In both tissues, neuronalinnervation induces synaptic acetylcholinesterase (AChE) activity;Gly-Gln produces a comparable effect, suggesting that it may be theneurotrophic agent mediating the response. P-endorphin-1-31 (butnot -1-27) has the opposite effect, reducing AChE, and may beresponsible for the subsequent lowering of AChE activity observedduring synaptic reorganization.

Glycyl-L-glutamine may also play a role in the neuroendocrineregulation of the immune system. A large body of evidence supports

-5-

the concept that the central nervous system influences immuneresponse and much of this work focuses on the role of POMC peptidesas neuroimmune mediators (Weber and Pert, 1984; Morley et al.,1987). McCain et al. showed that very low concentrations ofglycyl-L-glutamine enhance phytohemagglutinin (PHA) induced T-lymphocyte proliferation (McCain et al., 1986; 1987). This was akey finding because the PHA response, which mimics antigen inducedlymphocyte activation, is often used as a measure of immunecompetence. Once again, -endorphin-l-31 has the opposite effect,suppressing PHA-induced proliferation (McCain et al., 1987). Theseobservations suggest that glycyl-L-glutamine release from theintermediate pituitary may partially counteract stress-inducedsuppression of immune function.

These intriguing findings support the concept that glycyl-L-glutamine, like other f-endorphin peptides, functions both as aneurotransmitter in brain and as a circulating hormone in theperiphery; however, the basic studies required to firmly establishsuch a role for glycyl-L-glutamine have not been performed. Severalcriteria must be fulfilled. First, glycyl-L-glutamine's pharmaco-logic spectrum of activity must be definitively established, empha-sizing its interactions with 6-endorphin-l-31 and opiate drugs.Second, it must be demonstrated that glycyl-L-glutamine is actuallypresent in, and only in, P-endorphin releasing neurons and endo-crine cells. And third, receptors for glycyl-L-glutamine must beunequivocally identified and thoroughly characterized. We predictthat glycyl-L-glutamine receptors exhibit 'synaptic specificity',meaning they are found only in fi-endorphin neuronal synapses. Theconcept of synaptic specificity is important because it means thatdrugs targeted on glycyl-L-glutamine receptors will act only at 0-endorphin neuronal synapses, unlike all existing opiate drugs whichinteract with different opioid receptor subtypes present in allthree opioid peptide systems. Thus, the longer term objective ofthis research is to establish the data base necessary to designtherapeutic agents targeted on glycyl-L-glutamine receptors toselectively modify the biological effects of both glycyl-L-glutamine and f-endorphin.

RESULTS AND DISCUSSION

I. Pharmacologic Effects:

The initial objective of our research was to further evaluatethe physiological responses produced by glycyl-L-glutamine. Wereasoned that knowledge about glycyl-L-glutamine's physiologicaleffects would guide subsequent efforts to localize glycyl-L-glutamine and its receptors, to determine whether it functionsalone in neural and endocrine communication or specificallymodulates /-endorphin's actions, and ultimately, to predict theeffects of pharmacologic agents targeted on glycyl-L-glutaminereceptors. To initiate these studies, we selected three

-6-

experimental paradigms, evaluating both peripheral and centralresponses to glycyl-L-glutamine, including; its trophic effect oncardiac myocytes, its neuroimmunomodulatory activity on T-lymphocyte proliferation and its central action on f-endorphin-induced antinociception.

A. Trophic Effects on Cardiac Myocytes: The initial objectiveof this project period was to test whether glycyl-L-glutamineproduces trophic effects on cardiac myocytes, extending our ongoinginterest in the cardiovascular effects of 8-endorphin peptides(Hirsch and Millington, 1991). The rational for the study wasbased on an earlier report that glycyl-L-glutamine induces theexpression of acetylcholinesterase (AChE) at the neuromuscularjunction (Lotwick wt al., 1990) and in sympathetic ganglia (Koelleet al., 1988). Multiple molecular forms of AChE exist in bothcardiac and skeletal muscle, as well as in ganglia and othertissues, but only one, the asymmetric A12 form is specificallyassociated with neuronal synapses (Rieger et al., 1980). A markedincrease in A12 AChE activity occurs during the developmental periodwhen skeletal muscle is innervated; however, the neuronal agentwhich induces A12 AChE has not been identified. Recently, Lotwicket al. reported that very low glycyl-L-glutamine concentrations (10nM - 10 AM) induce A12 AChE expression in rat and chick muscle cellsin vitro suggesting to the authors that it may be the trophicsubstance normally responsible for regulating the AChE response toinnervation (Lotwick et al., 1990). Although glycyl-L-glutaminehas not as yet been identified in motor neurons, it is known thatA-endorphin peptides, including O-endorphin-l-27, are transientlyexpressed during the critical developmental period (Haynes et al.,1984) inferring that glycyl-L-glutamine must also be present(Parish et al., 1983). Interestingly, P-endorphin-l-31 has theopposite effect on AChE, reducing expression of the A12 form (Hayneset al., 1984).

Our results demonstrate that glycyl-L-glutamine also producestrophic effects on cardiac myocytes (Battie et al., 1991). Myocytecultures were prepared by enzymatically dissociating ventriclesfrom 2-4 day old rats; the myocytes were confluent and beating forat least 48 h prior to experimentation (Hagler and Nyquist-Battie,1990). Separating AChE molecular forms by sucrose density gradientfractionation revealed that cultured myocytes, like skeletal musclecells, normally produce very little A12 AChE. Only about 3% oftotal AChE activity was attributable to the A12 form; globularforms, including monomeric, G, (56%), and tetrameric, G4 (42%),forms predominate (Fig. 1). Co-incubation with glycyl-L-glutamine(1 MM) for 72 h produced a dramatic increase in the A12 form,elevating it from 3% to 21% of total AChE activity. Correspondingdecreases occurred in the G, (36%) form, but there was no changein G4 AChE or in total AChE activity. Glycyl-L-glutamine did notchange the specific activity of total cellular acetylcholinesteraseor its rate of secretion. The response to glycyl-L-glutamineappeared to be relatively specific to the extent that neitherglycyl-L-glutamate nor glycyl-D-glutamine had any effect on AchE

-7-

expression. These results provide the first evidence that glycyl-L-glutamine acts as a trophic agent in the heart.

B-Endorphin Processing in Rat Heart: Yet to be determined,is whether glycyl-L-glutamine is normally present in heart althoughthe recent identification of immunoreactive -endorphin in ratheart (Forman et al., 1989) suggests that it may, indeed, bepresent. We plan to address this question directly, once we havedeveloped methods for quantifying tissue glycyl-L-glutamineconcentrations; however, the identification of C-terminallyshortened p-endorphin peptides in heart would provide compellinginferential evidence that glycyl-L-glutamine is also present andtechniques for isolating the molecular forms of P-endorphin areroutinely used in our laboratory.

We, therefore, initiated efforts to characterize the post-translational processing of f-endorphin in rat heart. Themethodologic approach consists of three steps: extraction usingSep-Pak C-18 cartridges; gel filtration high performance liquidchromatography (HPLC) to separate fl-lipotropin from P-endorphin-sized peptides (Eipper et al., 1983) and ion exchange HPLC toidentify the individual A-endorphin forms (Millington et al.,1987). Gel filtration HPLC demonstrated that fl-endorphin immuno-reactivity (ig-endorphin) in the heart is primarily attributableto /-endorphin sized molecules and not their immediate precursor,-lipotropin; /-lipotropin constituted only 22 % of total immuno-reactivity (Fig. 2) (Evans et al., 1991), similar to that of theintermediate pituitary and brain rather than the anterior pituitarywhere fl-lipotropin predominates (Eipper and Mains, 1980; O'Donohueand Dorsa, 1982). Further analysis of fl-endorphin sized peptidesby ion exchange HPLC revealed that P-endorphin-l-31 is the predomi-nant form expressed in heart, constituting 49.6% of total ifl-endor-phin but that a-N-acetylated, C-terminally shortened forms werealso produced, including P-endorphin-l-27 (14.3%), a-N-acetyl-p-endorphin-l-27 (8.4%), P-endorphin-l-26 (7.9%) and N-acetyl-o-en-dorphin-l-26 (4.7%) (Fig. 3). Thus, approximately 35% of the 0-endorphin-l-31 produced in heart is further cleaved to glycyl-L-glutamine and C-terminally shortened forms, P-endorphin-l-27 and1-endorphin-l-26.

P-Endorphin's precise cellular localization in heart tissueremains to be determined. But recently, however, our collaboratorshave found, using in situ hybridization histochemistry, that POMCmRNA is expressed by cardiac ventricular cells, suggesting thatPOMC peptides are synthesized by heart tissue, rather than auto-nomic neurons innervating the heart (personal communication, Dr.Lloyd Forman). Consistent with this finding, our studies alsorevealed that, in addition to R-endorphin, rat heart containsadditional products of POMC processing, including ACTH and a-melanocyte stimulating hormone (a-MSH). Thus, gel filtration HPLCshowed that, as in the case of /-lipotropin's conversion to fl-en-dorphin, ACTH was almost entirely cleaved to a-MSH; the a-MSH:ACTHratio was 6.1:1, similar to that for O-endorphin:g-lipotropin(4.6:1). Ongoing studies will use reverse phase HPLC to further

-8-

identify the specific a-MSH forms present in heart extracts (des-acetyl-a-MSH, a-MSH or di-acetyl-a-MSH). Identification of ACTHand a-MSH in rat heart further supports the conclusion that theimmunoreactive p-endorphin found in heart is an authentic productof POMC processing. Moreover, these studies are the first toidentify ACTH and a-MSH, as well as the individual molecular formsof f-endorphin, in rat heart and they provide strong evidence thatglycyl-L-glutamine is also expressed.

B. Neuroimmune Effects: A rapidly growing body of evidencedocuments the role of fl-endorphin peptides in mediating the effectsof stress on the immune system (Weber and Pert, 1984). f-endor-phin-l-31 produces a variety of actions on immune cell function,many of which are mediated by receptors recognizing the non-opioidforms of f-endorphin released from the intermediate lobe. Littleis known about the role of glycyl-L-glutamine, although McCain etal. have shown that glycyl-L-glutamine enhances phytohemagglutinin(PHA) induced T-lymphocyte proliferation (McCain et al., 1987), ameasure of the ability of lymphocytes to respond to antigenicstimuli which is often used to test immune competence. The effectof glycyl-L-glutamine on T-lymphocytes appears to be indirect, how-ever, resulting from inhibition of T-cell suppressor cell activity.-endorphin-l-31 produces the opposite response, enhancing suppres-

sor cell activity, which in turn inhibits T-lymphocyte prolifera-tion. Thus, anterior and intermediate lobe secretory productsproduce opposing actions on suppressor cells; 6-endorphin-l-31,released almost exclusively from the anterior lobe, activatessuppressor cells while glycyl-L-glutamine, released from theintermediate lobe, inhibits the response.

P-endorphin-l-31 also acts directly on T-lymphocytes, stimu-lating PHA-induced proliferation (Gilman et al., 1982; Gilmore andWeiner, 1988, 1989; Hemmick and Bidlack, 1990). The exact identityof the receptor mediating this, and other immune responses to P-endorphin remains somewhat controversial, however, and there isevidence for the involvement of both opioid (Madden et al., 1987;Avadia et al., 1989) and non-opioid (Hazum et al., 1979) f-endor-phin binding sites. Interestingly, in lymphocyte proliferationassays, O-endorphin-l-27 act as an agonist with a potency similarto P-endorphin-l-31, suggesting the involvement of a non-opioidreceptor, similar to our findings on central cardioregulation(Hirsch and Millington, 1991). Yet to be examined, is whetherglycyl-L-glutamine also directly stimulates T-lymphocyte prolifera-tion; if so, then one might hypothesize that C-terminal cleavageof O-endorphin-l-31 to P-endorphin-l-27 and glycyl-L-glutamineamplifies the response by producing two stimulatory peptides.

To test this hypothesis, we examined the effect of 8-endorphinand glycyl-L-glutamine on c-myc protooncogene expression in a humanT-lymphocyte cell line, Jurkat E6-1. Mitogenic stimulation inducesa rapid increase in c-myc mRNA levels, both in normal lymphocytes(Reed et al., 1986) and in Jurkat E6-1 cells (Hough et al., 1990);c-myc mRNA is thought to encode a transcriptional activating factor

-9-

which is a necessary intermediary in the proliferative response.Jurkat E6-1 cells (1 x 10 6) were grown in RPMI medium containing10% fetal calf serum, penicillin (100 Ag/ml) and streptomycin (100gg/ml) and mitogenesis was stimulated with concanavalin A (con A).Total RNA was isolated by centrifugation through 5.7 M CsCl fol-lowed by phenol/chloroform extraction and ethanol precipitation.RNA was then electrophoresed on formaldehyde agarose gels, trans-ferred to nitrocellulose by vacuum blotting and hybridized with a32P-labeled nick-translated 2.5 kb ecoRI fragment of human c-myccDNA. Quantitation was performed by densitometric analysis of theresulting autoradiograms and standardized against rRNA.

We found that incubating Jurkat E6-1 cells with both p-en-dorphin-l-31 (30 nM) and glycyl-L-glutamine (30 nM) produced a 2.5-fold increase in con A (50 ng/ml) stimulated c-myc mRNA expression;when added to the cell cultures alone, neither O-endorphin-l-31 norglycyl-L-glutamine had any significant effect at the dose tested(30 nM) (Fig 4). As expected, con A alone elevated c-myc mRNA to200% of control levels. Final interpretation of these resultsawaits full individual dose-response evaluations of both peptides.However, while yet preliminary, the results suggest that glycyl-L-glutamine potentiates O-endorphin-l-31 induced c-myc mRNA expres-sion. It will be of particular interest to test whether R-endor-phin-l-27 produces the same response and, if so, to evaluate thestructural determinants of the P-endorphin receptor involved todetermine whether glycyl-L-glutamine acts through the same or adifferent binding site.

O-Endorphin ProcessinQ in Human Pituitary: The finding thatglycyl-L-glutamine potentiates A-endorphin-induced c-myc geneexpression in a human lymphocyte cell line raised the question asto the source of glycyl-L-glutamine which normally regulates immunecell function. One likely source may be the lymphocytes them-selves, which are known to express 6-endorphin and other POMCderived peptides (Morley et al., 1987), a possibility which we arenow beginning to examine experimentally. A second source, ofcourse, is the intermediate pituitary which, at least in the rat,produces high levels of glycyl-L-glutamine. But adult humans lackan intact intermediate lobe questioning whether P-endorphin-l-31undergoes significant post-translational processing to glycyl-L-glutamine and C-terminally shortened p-endorphin peptides. Previousstudies have identified small amounts of N-acetyl-p-endorphin-l-27immunoreactivity in extracts of fetal (Facchinetti et al., 1989)and adult (Smith et al., 1985) human pituitaries, but N-acetyl-0-endorphin-l-26 was not produced, suggesting that species differ-ences in P-endorphin's primary sequence may prevent its synthesisin the human. These studies were limited in scope, however, anddid not examine non-acetylated R-endorphin forms, nor did theyidentify the cell type, whether corticotroph or melanotroph,containing acetylated P-endorphin peptides.

To address these questions, we analyzed the molecular formsof P-endorphin in human pituitary using both chromatographicanalysis and immunohistochemistry (Bernard et al., 1991). Our

-10-

previous studies showed that O-endorphin processing is quite stablein post-mortem rat brain, remaining essentially the same as controlvalues for up to 24 hours (Millington and Smith, 1991). Gel fil-tration HPLC analysis of human pituitaries revealed that the majorportion of total i-endcrphin is attributable to 1-lipotropin; P-endorphin sized peptides constituted only 17.6% of total ig-endor-phin (Fig. 5). Further ion exchange HPLC analysis showed that 3-endorphin-1-31 was the predominate f-endorphin peptide, composing85.0% of total ig-endorphin. P-endorphin-l-27 (4.5%) and fl-endor-phin-l-26 (6.6%) were also present along with very low levels ofa-N-acetyl-p-endorphin peptides, together comprising less than 4%of total io-endorphin (Fig. 6). Thus, approximately L6% of totalpituitary 1-endorphin-l-31 is further processed to glycyl-L-gluta-mine and C-terminally shiortened 8-endorphin peptides.

We also employed immunohistochemical experiments using anantiserum which specifically recognizes a-N-acetyl-fl-endorphinpeptides to map the distribution of cells which process P-endor-phin-l-31. The antiserum stained a small number of cells localizedalong the border between the anterior and neural lobes suggestingthat they may be derived from the fetal intermediate lobe; however,c-N-acetyl--endorphin immunoreactive cells were also dispersedthroughout the anterior lobe. a-MSH immunoreactive cells aresimilarly distributed, suggesting that cells resembling melano-trophs, biochemically, are distributed throughout the humanpituitary. In summary, these results demonstrate that, despiteimportant differences in primary sequence, O-endorphin is processedto both O-endorphin-l-27 and -1-26 the human pituitary.

C. Antinociception: The profound analgesia produced by /-endorphin-l-31 is, of course, its most characteristic physiologicaction. Thus, it seemed logical to us to begin our evaluation ofglycyl-L-glutamine's central effects by testing whether it modifiesO-endorphin-l-31-induced antinociception. Mithough glycyl-L-glutamine biosynthesis within endogenous pain control pathways hasnot, as yet, been demonstrated directly, it can be inferred fromthe fact that P-endorphin-l-27, co-synthesized with glycyl-L-glutamine when P-endorphin-l-31 is post-translationally processed,is present within the periaqueductal grey and other brain struc-tures known to involved in endogencus pain control (Berglund etal., 1989; Zakarian and Smyth, 1979; Basbaum and Fields, 1984).The C-terminal proteolysis of P-endorphin-l-31 within the paincontrol system is particularly intriguing because -endorphin-l-27has been shown to be a potent antagonist of -endorphin-l-31antinociception (Nicholas and Li, 1985). Previous studies havedemonstrated that glycyl-L-glutamine antagonizes certain behavioralresponses produced by )3-endorphin-l-31; specifically, glycyl-L-glutamine (22 nmol) inhibits P-endorphin-l-31 induced grooming andstretch-yawn syndrome, behaviors thought to reflect states ofattention and arousal while producing no overt behavioral responsewhen administered alone (Hirsch and O'Donohue, 1986). In light ofthese findings, we hypothesize that glycyl-L-glutamine may inhibitthe antinociceptive action of P-endorphin-l-31.

-11-

To test this hypothesis, we initiated studies of glycyl-L-glutamine's effect on O-endorphin-l-31 antinociception using thetail flick reflex as an antinociception test paradigm. The tailflick test measures the response time for a rat to remove its tailfrom a beam of light irradiating the distal tail surface (D'Amourand Smith, 1941). Rats are briefly restrained during the test, theheat intensity of the light is adjusted to produce a baselinereflex of 3.5 seconds, and for animals that do not respond, thetest is terminated after 8.5 seconds. Baseline values are calcu-lated from the mean of three trials and rats are tested once everyfifteen minutes for two hours. The data are expressed as percentof maximum possible effect (%MPE) using the conventional equation:%MPE = [test latency - baseline/cutoff - baseline] x 100. Intra-cerebroventricular (icv) injections are performed by implanting aguide cannula in the lateral ventricle (stereotaxic coordinates,from lambda: A=6.7 mm, L=I.5 mm, D=4.3 mm) under ketamine/xylazineanesthesia and held in place by dental acrylic. Rats are allowedto recover from surgery for one week prior to each experiment andpeptide injections are administered to conscious animals over aperiod of 60-90 seconds using a 10 jil syringe attached to aninjection cannula with PE-10 tubing. Cannula placements areverified by dye injection at the end of each experiment.

Our initial objectives were to establish this nociceptive testparadigm in our laboratory, ensuring that the method generatedreproducible data in our hands, and to produce initial dose-response data. We first tested the antinociceptive effect ofintraperitoneal morphine sulfate (3.5 mg/kg) and as expected, foundthat morphine produced a prolonged increase in nociceptive latency,approximately doubling the response time for up to 60 min;naltrexone (10 mg/kg, i.p.) completely blocked the response. Next,after implanting chronic intraventricular (icv) cannulae in rats,we further showed that central morphine administration (10 and 50Mg) produced a dose dependent increase in nociceptive latency. Theresponse followed a similar time-course, being maximal at 40 minand returning toward baseline values after 60 min. The responseto O-endorphin-l-31 administration (3, 5 or 8 Mg) was also dosedependent but of longer duration, persisting for up to two hours.

After establishing consistent baseline data, we then testedwhether glycyl-L-glutamine modulates the antinociceptive responseto P-endorphin-l-31. Our initial results indicate that, indeed,it does. O-Endorphin-l-31 (1.5 nmol; 5 gg, icv) alone prolongedthe response latency to a maximum of 53 %MPE sixty minutes aftericv injection (Fig. 7). Co-administration of glycyl-L-glutamine(15 nmol) completely abolished the antinociceptive response,returning the response latency to baseline values; the inhibitoryresponse was initially apparent 30 minutes after injection andremained significantly reduced for the remainder of the 120 minutetest session. Higher glycyl-L-glutamine doses (50, 500 and 1500nmol) also effectively inhibited the response to -endorphin-1-31.Glycyl-L-glutamine, administered alone (50 or 500 nmol, icv)produced no effect whatsoever, consistent with previous reports onglycyl-L-glutamine's behavioral effects (Hirsch and O'Donohue,

-12-

1986). While yet preliminary, these results support the conceptthat glycyl-L-glutamine antagonizes the central response to 6-endorphin-l-31. Thus, C-terminal proteolysis of O-endorphin-l-31produces two peptides, -endorphin-l-27 and glycyl-L-glutamine,both of which antagonize its antinociceptive potency.

II. Analytical Methods:

The results discussed thus far demonstrate that glycyl-L-glutamine produces pharmacological responses both in the periphery,on cardiac myocytes and immune cells, and in brain, modulating 0-endorphin-l-31 antinociception. These experiments provide clearevidence that glycyl-L-glutamine subserves a physiological role asboth an endocrine hormone and brain neurotransmitter; however, toconclusively substantiate such a role, it is also necessary todemonstrate that glycyl-L-glutamine is contained in and releasedfrom neural and endocrine cells. We have further hypothesized thatglycyl-L-glutamine is specifically synthesized through f-endorphin-1-31 processing, and therefore, is selectively localized within fl-endorphin neural and endocrine cells. This is important conceptbecause, if so, then one might predict that drugs targeted onglycyl-L-glutamine receptors will be quite specific in theiraction, selectively modulating neurotransmission at G-endorphinneuronal synapses, unlike existing drugs which act at opiatereceptors which mediate the effects of not only fi-endorphin, butother opioid peptides, as well. To test these hypotheses, however,it is first necessary to develop analytical techniques to isolateand quantify glycyl-L-glutamine and map its distribution in brain.Toward this objective, we have initiated efforts to develop highperformance liquid chromatography (HPLC) methods to isolate glycyl-L-glutamine and both immunoassay and immunohistochemical methodsto quantify and map its localization in brain.

A. HPLC Analysis: Our initial objective, was to establishmethods for separating glycyl-L-glutamine from related peptides;this is important for demonstrating the specificity of subsequentquantitation methods and for documenting the purity of radiolabeledglycyl-L-glutamine, to be used for radioimmunoassay (RIA) and/orreceptor binding studies. Both of our methods utilize reversephase HPLC; the first, using an acetonitrile gradient in 0.05%trifluoroacetic acid, produces baseline separation between glycyl-L-glutamine, glycyl-L-glutamate and glycyl-L-asparagine, as wellas structurally unrelated dipeptides, tripeptides, and the opioidpeptides, met-enkephalin and P-endorphin. The broad range ofpeptides separated by this method is ideal for demonstrating RIAspecificity; however, it is less optimum for purification purposesbecause glycyl-L-glutamine elutes relatively close to the solventfront (within 5 minutes). We therefore developed a second separa-tion method, using a sodium phosphate mobile phase containing ionpairing agents, which extends the dipeptide's retention time to 18minutes.

-13-

B. Glvcvl-L-Qlutamine Assay: We next initiated pilot studiesto establish a method for measuring glycyl-L-glutamine in brain.Peptides are typically measured by RIA, which requires both aspecific antisera and a radiolabeled peptide. But small molecules,such as glycyl-L-glutamine, present two potential difficulties.First, antisera raised against small molecules coupled to largerproteins, such as albumin, typically work well for immunohisto-chemistry, in which the antigen is fixed to tissue proteins, butdo not always recognize the free antigen by RIA. Secondly, manysmall peptides, including glycyl-L-glutamine, cannot be iodinated.Therefore, we began by identifying four potential alternatives formeasuring glycyl-L-glutamine (with increasing feasibility, as wellas difficulty): (i) RIA, labeling glycyl-L-glutamine with Bolton-Hunter reagent (an 125I-labeled molecule which can be coupled to thepeptide's N-terminal); (ii) RIA, using 3H-glycyl-L-glutamine; (iii)ELISA assay, in which glycyl-L-glutamine is first coupled to al-bumin; (iv) chromatographic isolation and quantitation by HPLC withfluorescent derivitization (Parish et al., 1983).

First, using an antisera which specifically recognizes glycyl-L-glutamine by immunohistochemistry (Plishka et al., 1985), wetested whether it would bind [1251] Bolton-Hunter labeled glycyl-L-glutamine in an RIA. We succeeded in labeling glycyl-L-glutaminewith Bolton-Hunter reagent, which we verified by HPLC, but foundthat the antisera did not recognize the Bolton-Hunter-glycyl-L-glutamine complex. This was not entirely unexpected; Bolton-Hunter reagent, a relatively large molecule, blocks the dipeptide'sN-terminal, a likely antigenic determinant. Initial attempts todevelop a RIA using H-glycyl-L-glutamine, were also unsuccessful,suggesting that, like many antisera raised against amino acids andsmall peptides (Buijs et al., 1989), the essential immunogenicepitope of our antiserum included, not only glycyl-L-glutamine, butalso a portion of the protein and/or coupling agent incorporatedin the immunoreactive complex used to immunize the host animals.Indeed, our earlier immunohistochemical studies established thatthe antiserum did recognize endogenous glycyl-L-glutamine inpituitary tissue sections in which, presumably, the dipeptideundergoes fixative-induce coupling to tissue proteins. Moreover,this immunohistochemical staining could be completely blocked bypreincubating the antiserum with free glycyl-L-glutamine. Theseresults suggested to us that perhaps a similar strategy could beused to quantitatively analyze endogenous glycyl-L-glutamine byELISA assay.

To test this idea, we first coupled glycyl-L-glutamine tobovine serum albumin (BSA), using either carbodiimide (l-ethyl-3 (3-dimethyl-aminopropyl hydrochloride) or glutaraldehyde as couplingagents, then removed the remaining coupling agent and free glycyl-L-glutamin: by dialysis. We found that the glycyl-L-glutamineantiserum did, indeed, recognize the glycyl-L-glutamine-BSA con-jugate. Initially, we demonstrated this by isolating the glycyl-L-glutamine-BSA conjugate using SDS polyacrylamide gel electro-phoresis (SDS PAGE) and Western blotting to ensure that no freeglycyl-L-glutamine or other interfering components were present,

-14-

then identified the immunogen using a second antiserum coupled toalkaline phosphatase. Once establishing that the antiserum speci-fically recognized the conjugate by SDS PAGE, subsequent studieswere conducted by applying the conjugate directly to nitrocellulosefilters. Both carbodiimide and glutaraldehyde were efficientcoupling agents although the sensitivity of the antiserum forglutaraldehyde conjugated glycyl-L-glutamine was nearly ten foldhigher than for the carbodiimide coupled dipeptide. Dilutionstudies indicated that glycyl-L-glutamine immunoreactivity wasconcentration dependent with respect to both antigen and antiserum;less than 1 nmol glycyl-L-glutamine could readily be detected atan antiserum dilution of 1:2,000. Consistent with our hypothesis,the antiserum did not recognize free glycyl-L-glutamine applied tothe nitrocellulose filter, although in control experiments, our 0-endorphin antiserum readily stained free P-endorphin similarlyapplied to the membrane. Hence, these studies demonstrate thatprotein coupling is required for the antiserum to recognize glycyl-L-glutamine. Moreover, they established the feasibility of usingBSA-glycyl-L-glutamine conjugation to quantitatively analyze theendogenous dipeptide.

This experimental strategy has also proven quite useful forcharacterizing the glycyl-L-glutamine antiserum specificity. Wefound that the antiserum did not recognize BSA-conjugated fl-endor-phin, conclusive evidence that the antiserum does not cross-reactwith the C-terminal of the intact R-endorphin molecule. Thus far,the antiserum appears to be specific for glycyl-L-glutamine; forexample, it does not recognize the inverse dipeptide sequence,glutamyl-L-glycine, glycyl-D-glutamine, or either of glycyl-L-glutamine's constituent amino acids.

Subsequent experiments demonstrated that pre-incubating theantiserum with free glycyl-L-glutamine inhibited antiserum bindingto the BSA-conjugated glycyl-L-glutamine in a concentration depen-dent manner. Thus, as shown in immunohistochemical studies, thefree dipeptide inhibits antiserum binding to conjugated glycyl-L-glutamine, even though the antiserum does not appear to recognizethe free, unconjugated dipeptide. Thus, using this strategy, itmay yet be feasible to develop an analytical method for measuringendogenous glycyl-L-glutamine. To test this possibility, weinitiated efforts to develop an ELISA method for detecting glycyl-L-glutamine in brain and pituitary extracts.

Actually developing an ELISA assay has proven to be moreproblematic than expected, however. But with perseverance, we haverecently succeeded in generating saturable binding curves for BSA-conjugated glycyl-L-glutamine; the limit sensitivity appears to beapproximately 1 pmole at an antiserum dilution of 1:10,000 (Fig.8). Ongoing experiments will now test whether pre-incubating theantiserum with free glycyl-L-glutamine inhibits binding to BSA-conjugated glycyl-L-glutamine, as it did in our preliminary studiesusing nitrocellulose filters. If so, then we may yet be successfulin developing an ELISA assay to quantitate endogenous glycyl-L-glutamine concentrations in brain.

-15-

C. Immunohistochemical Studies: One objective of thisresearch is to test the hypothesis that glycyl-L-glutamine isselectively localized within G-endorphin-releasing neurons andendocrine cells. This is important to establish because, if so,then drugs targeted on glycyl-L-glutamine may be quite selectivein their action. To test this hypothesis, we initiated immuno-histochemical studies to map the distribution of glycyl-L-glutamineand P-endorphin. The first, critical question to be addressed, asfor any immunohistochemical study, was whether the glycyl-L-gluta-mine antiserum specifically identifies glycyl-L-glutamine, and notf-endorphin from which it is derived, within neurons and endocrinecells. To determine this, we tested whether the antiserum wouldselectively stain melanotroph cells in the intermediate lobe of therat pituitary, which are known to contain glycyl-L-glutamine, butnot corticotrophs in the anterior lobe, which contain 1-endorphinbut not glycyl-L-glutamine.

Using a standard fluorescently-labeled second antibody method,we found that the antiserum (1:1000 dilution) produced intensestaining over virtually every cell in the intermediate lobe butproduced no specific staining whatsoever in the anterior lobe. Thefinding that melanotrophs, but not corticotrophs, are stainedindicates that the antiserum does not recognize the C-terminal ofO-endorphin-l-31 from which glycyl-L-glutamine is derived. Furthersupport for this conclusion is provided by control experimentsshowing that pre-incubating the antiserum with glycyl-L-glutamine(1 AM) inhibited melanotroph staining, but that O-endorphin-l-31(1 AM) did not. Thus, consistent with the results describedearlier for Western blotting experiments, these results indicatethat the antiserum selectively recognizes glycyl-L-glutamine andnot P-endorphin-l-31 or any other anterior pituitary peptide orprotein.

It remains to be determined whether the antiserum will stainglycyl-L-glutamine neurons in brain with equivalent specificity,but these experiments are now in progress. For our initialstudies, we used carbodiimide-fixed coronal sections cut at thelevel of the arcuate nucleus, the region containing the majorproportion of P-endorphin neuronal perikarya. We found that theantiserum did, in fact, stain neuronal processes, but only in themedian eminence. Control experiments confirmed the stainingspecificity by showing that it was completely blocked by pre-incubating the antiserum with 1 AM glycyl-L-glutamine. Neuronalcell bodies were not recognized by the antiserum, which was notunexpected because the animals had not received prior treatmentwith colchicine. These results may suggest that glycyl-L-glutamineis specifically localized within axons projecting to the medianeminence; however, a more likely explanation is that the sensi-tivity of our staining method is too low to detect the dipeptidein other neuronal processes within the hypothalamus. To test thispossibility, we plan to test more sensitive staining methods, suchas the peroxidase/antiperoxidase technique.

-16-

Testing the hypothesis that glycyl-L-glutamine is specificallylocalized in P-endorphin neurons and endocrine cells also requiresantisera against P-endorphin and, perhaps, other pro-opiomelanccortin-derived peptides. This has been problematic, up till now,because neither our -endorphin radioimmunoassay antisera nor othercommercially available ones adequately labeled P-endorphin con-taining cells. During this project period we have more aggres-sively tested a variety of antisera and have found both P-endorphinand a-MSH antisera which selectively label pituitary corticotrophand/or melanotroph cells with high sensitivity (Micevych and Elde,1982; Sherry et al., 1982) (Incstar Corp., Stillwater, MN).

We now have demonstrated; (a) that our glycyl-L-glutamineantiserum specifically recognizes the dipeptide in pituitary cells;(b) that it stains neuronal processes in brain; and (c) that thenecessary f-endorphin antiserum is available and operable in ourlaboratory. Our final objectives are to increase the stainingsensitivity for glycyl-L-glutamine in brain, where glycyl-L-glutamine levels are low relative to pituitary concentrations, andto confirm that our A-endorphin antiserum specifically stainsneuronal processes as previously reported (Micevych and Elde, 1982;Sherry et al., 1982). Once we have met these objectives, we willproceed with a detailed mapping study of glycyl-L-glutamineimmunoreactive cell bodies and processes in brain, comparing theirdistribution with that of P-endorphin immunoreactive neurons todetermine whether the two peptides are strictly co-localized or ifglycyl-L-glutamine is expressed in non-p-endorphin neurons as well.

SUMMARY AND CONCLUSIONS

A. Pharmacoloqic Responses: To date, we have demonstrated thatglycyl-L-glutamine produces three markedly different pharmacologicresponses: it induces a trophic response, stimulating A12 AChEexpression in cardiac myocytes; produces neuroimmune regulatoryeffects on T-lymphocytes, enhancing c-myc oncogene expression; andinhibits a centrally mediated response to P-endorphin-l-31, anti-nociception. Thus, glycyl-L-glutamine produces pharmacologiceffects in both brain and peripheral tissues, supporting theconcept that it acts as both a neurotransmitter and a circulatinghormone.

1. Trophic Effects on Cardiac Myocytes: Neonatal ventricularmyocytes loose the ability to synthesize A12 AChE when grown inculture. The mechanism is uncertain, although the fact that myo-cytes continue to produce regular contractions in culture suggeststhat it does not result from the lack of contractile activity butrather from the absence of either neuronal innervation or, perhaps,a trophic substance. Here, we have shown that glycyl-L-glutaminerestores A12 AChE expression to essentially the same level asobserved in vivo (Nyquist-Battie, 1991), suggesting the possibilitythat it may be the responsible regulatory agent. This observation

-17-

was predicated on similar results in skeletal muscle (Lotwick etal., 1990) and sympathetic ganglia (Koelle et al., 1988); together,these finding suggest that glycyl-L-glutamine's trophic action onAChE expression is widespread and not specific for cardiac tissue.The concurrent finding that heart tissue post-translationallyprocesses O-endorphin-1-31 to 6-endorphin-l-27, and presumablyglycyl-L-glutamine, supports the physiologic relevance of theobserved effects. Furthermore, evidence that ventricular cells,themselves, express POMC mRNA (personal communication, Dr. LloydForman) raises the intriguing possibility glycyl-L-glutamine, aswell as other POMC peptides, acts as an autocrine factor toregulate AChE expression and, perhaps, other aspects of cardiacfunction. This concept is supported by evidence that ventricularcells express other peptide precursors, including pro-atrialnatriuretic peptide (Gu, 1991) and pro-enkephalin (Springhorn andClaycomb, 1989), yet we have not, as yet, ruled out the possibilitythat glycyl-L-glutamine may also be released from autonomic neuronsinnervating the heart.

These studies are now complete and the resulting manuscriptis currently in preparation. However, our interest in the cardio-vascular effects of glycyl-L-glutamine continues. Studies are nowbeing initiated to test whether glycyl-L-glutamine participates inthe central regulation of cardiovascular function. This hypothesisarises from our prior studies showing that C-terminal cleavage ofP-endorphin-l-31 to P-endorphin-l-27 enhances its hypotensivepotency when centrally administered (Hirsch and Millington, 1991),in marked contrast to antinociception in which -endorphin-l-27acts as a highly potent antagonist of -endorphin-l-31. Thus, wehypothesize that glycyl-L-glutamine will also produce hypotension,based on the concept that C-terminal proteolysis of -endorphin-l-31 produces two peptides, glycyl-L-glutamine and -endorphin-l-27,with the same physiological action, thereby potentiating theresponse to neuronally released -endorphin peptides.

2. Neuroimmune Effects: An extensive literature now documentsthe role of P-endorphin and other POMC peptides in regulatingimmune cell function (Weber and Pert, 1984; Blalock et al., 1985;Morley et al., 1987). Consistent with these findings, glycyl-L-glutamine has been shown to inhibit T-cell suppressor cell activ-ity, thereby indirectly enhancing antigen induced T-lymphocyteproliferation (McCain et al., 1987). Our studies demonstrate that,in addition, glycyl-L-glutamine, produces direct effects on T-lymphocytes, enhancing conconavalin A (con A) induced c-myc onco-gene expression, a measure of T-lymphocyte proliferation. Together,these results indicate that glycyl-L-glutamine enhances T-lympho-cyte proliferation through both direct and indirect mechanisms.

Nevertheless, our initial finding, that glycyl-L-glutaminestimulates c-myc oncogene expression only in the presence ofequivalent P-endorphin concentrations, needs to be extended,testing complete dose response curves to determine whether theeffect can be duplicated by either peptide alone or if they act

-18-

only in concert. In addition to measuring c-myc expression, itwill also be necessary to demonstrate directly that glycyl-L-glutamine does, indeed, enhance T-lymphocyte proliferation. Weplan to accomplish this using 3H-thymidine incorporation, a wellestablished method for measuring the T-lymphocyte proliferativeresponse to immune modulators.

We also plan to continue investigating the source of glycyl-L-glutamine which modulates T-lymphocytes. As discussed pre-viously, one likely source is the intermediate pituitary whichsynthesizes high levels of glycyl-L-glutamine (Plishka et al.,1985), at least in the rat, and based on our studies, to a smallextent in the human pituitary as well (Bernard et al., 1991). Butit is also quite possible that macrophages or other immune cellssynthesize and secrete glycyl-L-glutamine to regulate T-lymphocytefunction locally. Indeed, there is considerable evidence thatother POMC peptides are so produced (Blalock, 1985). To evaluatethis question, we are initiating experiments, using Northern blotanalysis, to test whether macrophages, or other immune cell lines,express POMC mRNA. If so, we plan to further evaluate whetherimmune cells process P-endorphin-l-31 to glycyl-L-glutamine and toidentify the factors which regulate its synthesis and secretion.Thus, the overall objective of these studies is to investigateglycyl-L-glutamine synthesis by, and regulation of, T-lymphocytesand other immune cells.

3. O-Endorphin Antinociception: Perhaps our most excitingfinding, thus far, is the observation that glycyl-L-glutamineinhibits P-endorphin-l-31 antinociception. In light of priorevidence that P-endorphin-l-27 is also a potent antagonist of P-endorphin-l-31 (Nicolas and Li, 1985), this means that C-terminalcleavage of P-endorphin-l-31 produces two peptides which, when co-released, oppose its antinociceptive action. Our studies were alsopredicated on previous findings that glycyl-L-glutamine inhibitscertain behavioral responses to P-endorphin-l-31, as does P-endor-phin-l-27; again, as in our studies, glycyl-L-glutamine producedno effect when administered alone (Hirsch and O'Donohue, 1986).Thus, glycyl-L-glutamine appears to act as a neuromodulator of P-endorphin's behavioral and antinociceptive activities.

We now plan to expand our initial dose-response studies andfurther test whether glycyl-L-glutamine also modulates the responsetn icv morphine administration. If so, then future studies willexamine whether peripherally administered glycyl-L-glutamine andmorphine also interact. This would provide the added opportunityto evaluate glycyl-L-glutamine congeners, including glycyl-D-glutamine and cyclo-glycyl-L-glutamine, which are likely to beresistant to metabolism and, in the case of the cyclic dipeptide,should readily cross the blood-brain barrier (Hoffman et al.,1977). Thus, this line of investigation may be useful towardevaluating whether glycyl-L-glutamine, or its analogs, areeffective centrally, when administered peripherally. In the longerterm, the conclusive demonstration that glycyl-L-glutamine inhibitsmorphine antinociception would raise the intriguing possibility

-19-

that glycyl-L-glutamine receptor antagonists may be useful adjunctsto opiate analgesia.

B. Analytical Methods: The second broad objective of our researchis to develop analytical methods for measuring glycyl-L-glutamineand mapping its distribution in brain. These methods are essentialfor assessing whether glycyl-L-glutamine is selectively localizedin O-endorphin neurons and for studying the regulation of itssynthesis and release. Establishing these methods has been moreproblematic than expected, although we have now made considerableprogress toward developing both immunoassay and immunohistochemicalmethods for detecting glycyl-L-glutamine in biain and pituitarytissue.

The most promising approach to measuring glycyl-L-glutaminenow appears to be by ELISA assay, and thus far, we have generatedsensitive concentration dependent binding curves for BSA-conjugatedglycyl-L-glutamine. As an alternative method, we are now examiningthe feasibility of using HPLC, combined with orthophthalaldehydederivitization and fluorescent detection, to measure glycyl-L-glutamine in tissue extracts. Both analytical approaches requirepreliminary sample extraction, and for this purpose, we have foundthat small anion exchange columns effectively separate glycyl-L-glutamine from amino acids and other small peptides in crude tissueextracts. Thus we have developed all the necessary components foranalyzing endogenous glycyl-L-glutamine and should soon have anassay available.

Immunohistochemical studies have also generated promisingresults. Thus far, we have succeeded in selectively detectingglycyl-L-glutamine in the intermediate pituitary using a fluores-cently labeled second antibody, identifying the necessary perfusionand fixation components in the process, and have established thatthe antiserum does not recognize -endorphin or any other anteriorpituitary peptide or protein. We have also labeled glycyl-L-glutamine containing neuronal processes in brain tissue althoughfurther effort will be required to enhance the sensitivity of ourimmunohistochemical methods. But more sensitive detection tech-niques, such as peroxidase/antiperoxidase, are readily availableto us and we are now initiating experiments to evaluate theireffectiveness. When successful, subsequent studies will mapglycyl-L-glutamine's distribution in brain, comparing it to thatof P-endorphin, to establish whether the dipeptide is selectivelylocalized in -endorphin releasing neurons.

In summary, our results thus far have made considerableprogress toward completing our first objective, evaluating glycyl-L-glutamine's pharmacologic activity spectrum in selected experi-mental paradigms. We have also overcome several impediments towardaccomplishing our second objective, establishing the methods neces-sary to analyze glycyl-L-glutamine's tissue distribution. We planto continue these efforts, and additionally, to initiate studiesof the receptors which mediate glycyl-L-glutamine pharmacological

-20-

responses, our third and final research objective. Toward thatend, we are now establishing the requisite receptor bindingtechniques, testing as our initial goal, whether the dipeptideinhibits opiate receptor ligand binding. This, perhaps the mostexciting phase of our research, should ultimately resolve whetherglycyl-L-glutamine acts through receptors previously identified forother neurotransmitters or, perhaps, through its own, uniquereceptor.

-21-

'1

02 6

2-2A

ii

:1+

=_AC7'O!;S

Figure 1. Glycyl-L-glutamine stimulates expression of asymmetricacetylcholinesterase (AChE) forms in cultured fetal myocytes.Dissociated cells from ventricles of neonatal rats were preplatedin M199 medium and 0.5% fetal calf serum to remove non-musclecells, then transferred to laminin-coated culture dishes (1.5 x 105

cells/35 mm dish) in a defined medium and incubated with or without10-6 M glycyl-L-glutamine for 48 h. AChE molecular forms wereseparated by sucrose density centrifugation and peaks of AChEactivity were identified by calibrating the sucrose gradients withstandard proteins with known sedimentation coefficients (Nyquist-Battie et al., 1987). The upper panel illustrates the molecularforms of AChE in controls, and the lower panel glycyl-L-glutaminetreated myocytes. The peaks of AChE activity correspond, from leftto right, with the sedimentation coefficients of (lower panel): G,and G2, G4, A8 and Al2 .

2 2

Q)

Q-

40 ZOZO6 70

F- cC 1'0r

Figure 2. Gel filtration HPLC separation of 1-lipotropin and P-erdorphin-sized peptides in rat heart. P0-Lipotropin and 13-endor-phin-sized peptides were separated by gel filtration HPLC frompooled extracts of five rat hearts and P-endorphin immunoreactivitywas analyzed by RIA. The arrows mark the elution position ofporcine 1-lipotropin (I) and 1-endorphin-l-31 (II).

2 3

- I 111I1i IV V VI

CE 1.5-

0 .2

- I

C I) 0.5

0.2 1

20 20 40 SO 60

Figure 3. The molecular forms of P-endorphin in rat heart. fB-En-dorphin peptides were separated by cation exchange HPLC from pooledextracts of five rat hearts previously separated by gel filtrationHPLC (Fig. 2). The arrows mark the elution position of N-acetyl-p-endorphin-l-26 (I), N-acetyl-g-endorphin-l-27 (II), R-endorphin-1-26 (III), -endorphin-l-27 (IV), N-acetyl-R-endorphin-l-31 (V), P-endorphin-l-31 (VI).

24

600

z 00-'0J0U-

0 400'

()_ _ _ _ _ _ _ _ _ _ _

CONTROL CON A +PHA B-end GLG + B-end GLG

TREATM ENT

Figure 4. Glycyl-L-glutamine and /-endorphin-l-3. stimulate c-myconcogene expression in human Jurkat E6-1 T-lymphocyte cells. Jur-kat Eg-1 cells were grown in RPMI medium containing 10% fetal calfserum. M'itogenesis was stimulated with concanavalin A (con A) andthe cells were incubated for 24 h with glycyl-L-glutamine (GLG; 30nM) and/or G-endorphin-1-31 (B-end; 30 niM) ). C-myc mRNA wasisolated by Northern analysis and quantified by densitometricanalysis.

25

400

0

400 80

I'

40 80 120

FRACTIONS

Figure 5. Gel filtration separation of P-lipotropin and -endor-phin-sized peptides in human pituitary. g-Lipotropin (1-LPH) andP-endorphin-sized peptides (P-end) were separated by gel filtration(Sephadex G-50; 2.5 x 100 cm) a single human pituitary and 6-endorphin immunoreactivity was analyzed by RIA.

26

2LL

0 LJ

0r

JJ

30 45 60 75

PAC i I S

Figure 6. The molecular forms of A-endorphi' in the human pit-uitary. O-Endorphin (/3-E) peptides were separated by cationexchange chromatography from a single human pituitary previouslyanalyzed by gel filtration HPLC (Fig. 5) and 13-endorphin immuno-reactivity was measured by RIA.

27

80-

00

-201Ibaseline 5.0 15.0 35.0 50.0 65.0 80.0 95.0 110.0

time

Figure 7. G1I-cyl-L-glutamine inhibits 63-endorphin-1-31 inducedantinocicep-ion. Groups of 8-10 rats were injected intracerebro-ventricularly (icy) with either P-endorphin-l-31 alone (5 Mg; uppertrace) or O-endorphin-l-31 combined with glycyl-L-glutamine (3 Atg)and tail flick latencies were recorded at the indicated timepoints.

28

1.4-

1 .2

E 1/C

0.8

C

0.6

< 0.4-/

0.2

0.01 0.1 0.25 0.5 .5 1 10 100 1000 1000Carb. Conj. Glin Conc., uM

Figure 8. ELISA assay standard curve of BSA-conjugated glycyl-L-glutamine. Glycyl-L-glutamine was covalently linked to BSA bycarbodiimide conjugation, dialyzed and plated in 96 well microtiterplates. BSA-conjugated glycyl-L-glutamine was detected using aprimary antisera dilution of 1:10,000 followed by goat anti-rabbitIgG coupled to alkaline phosphatase. The carbodiimide conjugatedglycyl-L-glutamine concentration (carb. conj. Gly-Gln conc.) wasestimated from the glycyl-L-glutamine concentration used in theconjugation reaction. The limit sensitivity of the assay isapproximately 1 pmol.

29

PUBLICATIONS

Five manuscripts were submitted for publication during thisproject period and two are currently in preparation with submissionanticipated in November; in addition, eleven abstracts were sub-mitted and/or presented. Much of this work represents the com-pletion of studies initiated during previous USAMRDC funding(86PP6813) although it was supported, in part, by the currentgrant.

Manuscripts:

Hirsch, M.D. and Millington, W.R. Endoproteolytic conversion of-endorphin-l-31 to g-endorphin-l-27 potentiates its centralcardioregulatory activity. Brain Res. 550:61-68, 1991.

Millington, W.R. and Smith, D.L. The post-translational processingof -endorphin in human hypothalamus. J. Neurochem. 57:775-781,1991.

Millington, W.R., Dybdal, N.O., Mueller, G.P. and Chronwall, B.M.N-acetylation and C-terminal proteolysis of f-endorphin in theanterior lobe of the horse pituitary. Gen. Comp. Endocrinol. (InPress).

Lavigne, G.L., Millington, W.R. and Mueller, G.P. The CCK-A andCCK-B antagonists, devazepide and L-365,260, enhance morphineantinociception only in non-acclimated rats. Pain (Submitted).

Millington, W.R., Mueller, G.P. and Lavigne, G.L. Differentialeffects of cholecystokinin type A and B receptor antagonists oncholecystokinin stimulated pituitary P-endorphin secretion. J.Pharmacol. Exp. Ther. (Submitted).

Battie, C.N., Hagler, K. and Millington, W.R. Glycyl-L-glutamineregulates the expression of acetylcholinesterase asymmetric formsin cultured fetal cardiac myocytes. (In Preparation).

Evans, V.R., Forman, L.J. and Millington, W.R. Pro-opiomelano-cortin-derived peptides in rat heart. (In Preparation).

Abstracts:

Dybdal, N.O., Chronwall, B.M. and Millington, W.R. N-acetylationand C-terminal proteolysis of O-endorphin in the anterior lobe of-he horse pituitary. The Endocrine Society, 1990.

Lavigne, G.L. and Millington, W.R. Antinociceptive and pituitaryg-endorphin studies with a novel cholecystokinin-B (CCK-B) antago-nist, L-340-718. VIth World Congress cn Pain, 1990.

Chronwall, B.M., Farah, J.M., Morris, S.J., Sibley, D.R. andMillington, W.R. Temporal characteristics cf dopaminergicregulation of the rat intermediate pituitary: Secretion, POMC andD2 receptor gene expressions and cell proliferation. society forNeuroscience, 1990.

Hirsch, M.D., Villavicencio, A.E., McKenzie, J.E. and Millington,W.R. C-terminal proteolysis modifies cardioregulation by 9-en-dorphin. Society for Neuroscience, 1990.

Bernard, L.H., Evans, V.R., Chronwall, B.M. and Millington, W.R.Beta-endorphin processing in human hypothalamus and pituitary. TheEndocrine Society, 1991.

Bernard, L.H., Chronwall, B.M., Evans, V.R. and Millington, W.R.Post-translational processing of O-endorphin and ACTH in the humanpituitary. The Midwest Anesthesiology Residents Conference, 1991.

Lavigne, G.J. and Millington, W.R. The CCK-A and -B antagonists,devazepide and L-365,260, potentiate morphine antinociception, butonly in non-acclimated rats. Third IBRO World Congress of Neuro-science, 1991.

Evans, V.R., Forman, L.J. and Millington, W.R. Characterizationof pro-opiomelanocortin-derived peptides in rat heart. Societyfor Neuroscience, 1991.

Dickerson, D.S., Pratt, B.S., Millington, W.R. and Chronwall, B.M.DZ dopamine receptor regulation in the intermediate lobe of the ratpituitary. Society for Neuroscience, 1991

Battie, C.N., Hagler, K. and Millington, W.R. Glycyl-L-glutamineregulates the expression of acetylcholinesterase asymmetric formsin cultured fetal cardiac myocytes. Society for Neuroscience, 1991.

Bernard, L.H., Chronwall, B.M., Evans, Y.R. and Millington, W.R.Post-translational processing of O-endorphin and ACTH in the humanpituitary. American Society for Anesthesiology, 1991.

31

REFERENCES

Akil, H., Young, E., Watson, S.J. and Coy, D.H. Opiate bindingproperties of naturally occurring N- and C-terminal modified-endorphins. Peptides 2:289-292, 1981.

Avadia, H. Nitsan, P. and Abramsky, 0. Characterization of opiatebinding sites on membranes of rat lymphocytes. J. Neuroimnunol.21:93-102, 1989.

Basbaum, A.I. and Fields, H.L. Endogenous pain control systems:Brainstem spinal pathways and endorphin circuitry. Ann. Rev.Neurosci. 7:309-338, 1984.

Battie, C.N., Hagler, K. and Millington, W.R. Glycyl-L-glutamineregulates the expression of acetylcholinesterase asymmetricforms in cultured fetal cardiac myocytes. Society for Neuro-science, 1991.

Berglund, L.A., Millington, W.R. and Simpkins J.W. Gondadalsteroid and chronic morphine treatment do not change the post-translational processing of -endorphin in rat brain. Life Sci.44:591-601, 1989.

Bernard, L.H., Chronwall, B.M., Evans, V.R. and Millington, W.R.Post-translational processing of f-endorphin and ACTH in thehuman pituitary. American Society for Anesthesiology, 1991.

Blalock, J.E. Proopiomelanocortin-derived peptides in the immunesystem. Clin. Endocrinol. 22:823-827, 1985.

Buijs, R.M., Pool, C.W., Van Heerikhuize, J.J. et al. Antibodiesto small transmitter molecules and peptides. Biomed. Res. 10,Suppl. 3:213-221, 1989.

D'Amour, F.E. and Smith, D.L. A method for determining loss ofpain sensation. J. Pharmacol. Exp. Ther. 72:74-79, 1941.

Deakin, J.F., Dostrovsky, J.O. and Smyth, D.G. Influence of N-terminal acetylation and C-terminal proteolysis on the anal-gesic activity of O-endorphin. Biochem. J. 189:501-506, 1980.

Eipper, B.A., Glembotski, C.C. and Mains, R.E. Selective loss ofa-melanotropin-amidating activity in primary cultures of ratintermediate pituitary cells. J. Biol. Chem. 258, 7292-7298,1983.

Eipper, B.A. and Mains, R.E. Structure and biosynthesis of pro-adrenocortin/endorphin and related peptides. Endocrine Rev.1:1-27, 1980.

Evans, V.R., Forman, L.J. and Millington, W.R. Characterizationof POMC-derived peptides in rat heart. Soc. Neuroscience, 1991.

32

Facchinetti, F., Storchi, A.R., Petraglia, F. and Genazzani, A.R.Presence of acetylated and shortened endorphins in human fetalpituitary gland. Pediatric Res. 25:652-655, 1989.

Forman, L.J., Estilow, S. and Hock, C.E. Localization of 6-endorphin in the rat heart and modulation by testosterone.Proc. Soc. Exp. Biol. Med. 190: 240-245, 1989.

Gilman, S.C., Schwartz, J.M., Milner, R.J., Bloom, F.E. andFeldman, J.D. 3-Endorphin enhances lymphocyte proliferativeresponses. Proc. Natl. Acad. Sci. U.S.A. 79:4226-4230, 1982.

Gilmore, W. and Weiner, L.P. O-Endorphin enhances interleukin-2(IL-2) production in murine lymphocytes. J. Neuroimmunol.18:125-138, 1988.

Gilmore, W. and Weiner, L.P. The opioid specificity of beta-endorphin enhancement of murine lymphocyte proliferation.Immunopharmacol. 17:19-30, 1989.

Gu, J. Immunocytochemical detection of atrial natriuretic peptidein periarterial myocardiocytes of cardiac ventricles of adultnormotensive rat. J. Histochem. Cytochem. 39:829-834, 1991.

Hagler, K. and Nyquist-Battie, C. Acetylcholinesterase (AChE)expression in cultured perinatal cardiac myocytes. Soc.Neurosci., 1990.

Haynes, L.W., Smith, M.E. and Smyth, D.G. Evidence for the neuro-trophic regulation of collagen-tailed acetylcholinesterase inimmature skeletal muscle by 6-endorphin. J. Neurochem. 42:1542-1551, 1984.

Hazum, E., Chang, K.-J. and Cuatrecasas, P. Specific Non-opiatereceptors for P-endorphin. Science 205:1033-1035, 1979.

Hemmick, L.M. and Bidlack, J.M. O-endorphin stimulates rat Tlymphocyte proliferation. J. Neuroimmunol 29: 239-243, 1990.

Hirsch, M.D. and Millington, W.R. Endoproteolytic conversion ofg-endorphin-l-31 to ,3-endorphin-l-27 potentiates its centralcardioregulatory actiiity. Brain Res. 550:61-68, 1991.

Hirsch, M.D. and O'Donohue, T.L. Structural modifications ofproopiomelanocortin-derived peptides markedly alter theirbehavioral effects. J. Pharmacol. Exp. Ther. 237:378-385, 1986.

Hoffman, P.L., Walter, R. and Bulat, M. An enzymatically stablepeptide with activity in the central nervous system: Its penet-ration through the blood-CSF bdrrier. Brain Res. 122:87-94,1977.

Hough, C.J., Halperin, J.I., Mazorow, D.L. , Yeandle, S.L. andMillar, D.B. 1 -Endorphin modulates T-cell intracellular calciumflux and c-myc expression via a potassium channel. J. Neuro-immunol. 27:163-171, 1990.

Koelle, G.B., Massoulie, J., Eugene, D. and Melone, M.A.B. Effectsof glycyl-L-glutamine in vitro on the molecular forms of acetyl-cholinesterase in the preganglionically denervated superiorcervical ganglion of the cat. Proc. Natl. Acad. Sci. USA85:1686-1690, 1988.

Lotwick, H.S., Haynes, L.W. and Ham, J. Glycyl-L-glutamine stimu-lates the accumulation of A,2 acetylcholinesterase but not ofnicotinic acetylcholine receptors in quail embryonic myotubesby a cyclic AMP-independent mechanism. J. Neurochem. 54:1122-1129, 1990.

Madden, J.J., Donahoe, R.J., Zwemer-Collins, J., Shafer, D.A. andFalek, A. Binding of naloxone to human T lymphocytes. Biochem.Pharmacol. 36:4103-4109, 1987.

McCain, H.W., Lamster, I.B. and Bilotta, J. Modulation of humanT-cell suppressor activity by beta-endorphin and glycyl-L-glutamine. Int. J. Immunopharmacol. 8:443-446, 1986.

McCain, H.W., Billotta, J. and Lamster, I.B. Endorphinergicmodulation of immune function: Potent action of the diptptideglycyl-L-glutamine. Life Sci. 41:169-176, 1987.

Micevych, P.E. and Elde, R.P. Neurons containing a-melanocytestimulating and -endorphin immunoreactivity in the cathypothalamus. Peptides 3: 655-662, 1982.

Millington, W.R., O'Donohue, T.L. and Mueller, G.P. (1987).Dopaminergic agents selectively alter the post-translationalprocessing of O-endorphin in the intermediate pituitary of therat. J. Pharmacol. Exp. Ther. 243, 160-170.

Millington W.R. and Smith D.L. The post-translational processingof -endorphin in human hypothalamus. J Neurochem 57:775-781,1991.

Morley, J.E., Kay, N.E., Solomon, G.F. and Plotnikoff, N.P. Neuro-peptides: Conductors of the immune orchestra. Life Sci. 41:527-544, 1987.

Nicolas, P. and Li, C.H. P-Endorphin-(l-27) is a naturallyoccurring antagonist to etorphine-induced analgesia. Proc.Natl. Acad. Sci. USA 82:3178-3181, 1985.

Nyquist-Battie, C., Hodges-Savola, C. and Fernandez, H.L. Acetyl-cholinesterase molecular forms in rat heart. J. Mol. Cell.Cardiol. 19:935-943, 1987.

Nyquist-Battie, C. Changes in the expression of acetylcholin-esterase molecular forms during rat heart development. (InPress).

O'Donohue, T.L. and Dorsa, D.M. The opiomelanotropinergic neuronaland endocrine systems. Peptides 3:353-395, 1982.

Parish, D.C., Smyth, D.G., Normanton, J.R. and Wolstencroft, J.H.Glycyl glutamine, an inhibitory neuropeptide derived from 3-endorphin. Nature 306:267-270, 1983.

Plishka, R.J., Cangro, C.B. and Neale, J.H. Immunohistochemicallocalization of the pro-opiomelanocortin-gene product, glycyl-glutamine, in the intermediate pituitary. Brain Res.360:403-406, 1985.

Reed, J.C., Alpers, J.D., Nowell, P. and Hoover, R.G. Sequentialexpression of protooncogenes during lectin stimulated mito-genesis of normal human lymphocytes. Proc. Natl. Acad. Sci.USA 83:3982-2986, 1986.

Rieger, F., Koenig, J. and Vigny, M. Spontaneous contractile ac-tivity and the presence of the 16S form of acetylcholinesterasein rat muscle cells in culture. Reversible suppressive actionof tetrodotoxin. Dev. Biol. 76:358-365, 1980.

Sherry, S.H., Guay, A.T., Lee, A.K., Hedley-Whyte, E.T., Federman,M., Freidberg, S.R. and Woolfe, P.D. Concurrent production ofadrenocorticotropin and prolactin from two distinct cell linesin a single pituitary adenoma: A detailed immunohistochemicalanalysis. J. Clin. Endocrinology and Metabolism 55: 947-955,1982.

Smith, A.I., Cheng, M.C. and Funder, J.W. The identification andcharacterization of c-N-acetylated )3-endorphin in the humanpituitary gland. FEBS Lett. 185:109-111, 1985.

Spinghorn, J.1. and Claycomb, W.C. Proenkephalin mRNA expressionin developing rat heart and in cultured ventricular cardiacmuscle cells. Biochem. J. 258:73-78, 1989.

Weber, R.J. and Pert, C.B Opiatergic modulation of the immunesystem. In: Muller, E.E. and Genazzani, A.R., eds., Central andPeripheral Endorphins: Basic and Clinical Aspects, Raven Press,New York, pp. 35-42, 1984.

Zakarian, S. and Smyth, D.G. Distribution of O-endorphin relatedpeptides in rat pituitary and brain. Biochem. J. 202, 561-571,1982.

35