MODULATION OF SEROUS SALIVARY GLAND FUNCTION BY THE SYMPATHETIC NERVOUS SYSTEM

A biochemical and ultrastructural study with special reference to ^-adrenoceptor subtypes

av

ROGER HENRIKSSON

AKADEMISK AVHANDLING

som med vederbörligt tillstånd av rektorsämbetet vid Umeå universitet

för avläggande av medicine doktorsexamen kommer att offentligen försvaras

i Histologiska institutionens föreläsningssal onsdagen den 9 december 1981, kl. 09.15.

CENTRALTRYCKERIETBRÖDERNA LARSSON I UMEA AB 1981

MODULATION OF SEROUS SALIVARY GLAND FUNCTION BY THE SYMPATHETIC NERVOUS SYSTEM.A Biochemical and U1trastructural Study With Special Reference to 3- Adrenoceptor Subtypes.Abstract: The aim of the present investigation was to study the influ­ence of the sympathetic nervous system and of various adrenoceptor agents on enzyme secretion and morphology in rat parotid and guinea-pig submandibular glands. Biochemical methods were combined with electron microscopical techniques.

Two different in vitro systems were employed, batch-incubation and microperifusion, to characterize the sympathetically evoked amylase re­lease and its correlation to cyclic AMP. By using various selective 3- adrenoceptor agonists and antagonists a dominance of the 3^-adrenoceptor over the 3^ - in regulating amylase release - was establ ished. Continuous noradrenaline perifusion caused a rapid initial amylase discharge, clo­sely correlated to tissue levels of cyclic AMP; no correlation between the two was observed during the later phase. Prenalterol (a 3^-agonist) failed to elevate glandular cyclic AMP. This was in contrast to its po­tent secretagogic effect. On the other hand, terbutaline (a 32-agonist) was a weak secretagogue but markedly raised the levels of cyclic AMP. Thus, 3-adrenoceptor activation may lead to release of large amounts of amylase despite minimal or no increase in cyclic AMP. Moreover, these effects seemed to be dissociated in salivary glands with regard to the 3-adrenoceptor subtypes. This was further substantiated by the findings that repeated injections of prenalterol induced qualitative changes in the granule populations, similar to those caused by the non-selective 3-agonist isoprénaline. Terbutaline was without effect. However, acinar cells size was increased following both prenalterol and terbutaline treatment. The data suggest that the 3-adrenergic effects on acinar cell size and granule population may be independently regulated.

A decreased sympathetic activity of long duration was induced by neonatal or adult extirpation of the superior cervical ganlion on one side. Acinar cell size, as well as granule and amylase content was re­duced 9 weeks after neonatal denervation. Ganglionectomy performed in adult animals was without significant effects.

The secretory behaviour of neonatally denervated glands was cha­racterized by an increased postjunctional sensitivity to 3-adrenoceptor agonists. Of special interest was the finding that neonatal denervation seemed to transform terbutaline from a partial to a full secretory ago­nist, thus changing its effects in the direction of those of prenalte­rol and noradrenaline. Moreover, increased levels of cyclic AMP as well as an enhanced response to DBcAMP were noted in the denervated glands as were intracellular changes. The denervation supersensitivity after neonatal denervation seems to differ from that observed in adult dener­vated glands. The results of the studies on denervated glands suggest that the sympathetic nervous system plays a fundamental role in the early maturation of the rat parotid gland as well as for the develop­ment of the 3-adrenoceptor subtypes.Key words: Rat parotid gland, Guinea-pig submandibular gland, 3-Adreno­ceptor subtypes, Sympathetic denervation, Sympathetic superstimulation, Amylase secretion, Cyclic AMP, U1trastructural stereology.

Roger Henriksson, Dept, of Histology and Cell Biology, University of Umeå, 901 87 Umeå, Sweden.

ISSN 0346-6612

UMEÅ UNIVERSITY MEDICAL DISSERTATIONSNew Series No 73 - ISSN 0346-6612

From the Department of Histology and Cell Biology University of Umeå, Umeå, Sweden

Modulation of Serous Salivary Gland Function by the

Sympathetic Nervous System

A Biochemical and Ultrastructural

Study with Special Reference to

ß-Adrenoceptor Subtypes

ByROGER HENRIKSSON

Umeå Universitet 1981

Ord

Tala tyst min munviskar jag till mig3forma varsamtmina tankar till ord>strö ej meningar lättvindigtsom sårar de svaga.Oeh allt vad jag säger har redan sagts förut i tusen versioner om gott ooh ont strunt ooh sanning.Så förbli varsam min mun.M Holmkvist Kiruna

3

CONTENTS

REPORTS CONSTITUTING THE THESIS ................................................. 4

INTRODUCTIONGeneral aspects ..................................................................... 5Salivary Glands ..................................................................... 9

AIM OF THE INVESTIGATION .............................................................. 12

MATERIALS AND METHODS ....................................................................13

RESULTS AND DISCUSSIONCharacterization of 3-adrenoceptor induced amylase release and its correlation to glandular cyclic AMP ...15Effects of chronic variation in sympathetic activity on secretory behaviour and morphology ..............................22

GENERAL SUMMARY ........................................................... 27

ACKNOWLEDGEMENTS .............................................................................29.

REFERENCES .......................................................................................30

PAPER IDynamics of ß-adrenoceptor induced amylase releaseand cyclic AMP accumulation ................................................35

PAPER IICharacterization of the rat parotid ß-adrenoceptor____43

PAPER IIIDissociation of ß-adrenoceptor-induced effects on amylase secretion and cyclic AMP accumulation ............... 49

PAPER IVßj-and ßp-adrenoceptor agonists have different effects on rat parotid acinar cells ............................................... 69

PAPER VEffects of neonatal sympathetic denervation on amylase secretion in the adult rat parotid gland: Difference in ß-j-and ßp-adrenoceptor response ................................... 87

PAPER VITrophic effect of the sympathetic nervous system on the early development of the rat parotid gland. A Quantitative U1trastructural Study ..................................105

4

REPORTS CONSTITUTING THE THESIS

This thesis is based upon the following publicationsreference to which will be made by citation of the appropriateRoman numerals.

I Dynamics of 3-adrenoceptor induced amylase release and cyclic AMP accumulation in the guinea-pig submandibular gland.Acta Physiologica Scandinavia 106:281-287 (1979). (In collabora- tion with B. Carlsöö, Å. Danielsson and L.-Ä. Idahl.)

II Characterization of the rat parotid 3-adrenoceptor. British Journal of Pharmacology 72:271-276 (1981). (In collaboration with B. Carlsöö, Ä. Danielsson and L.-Å. Idahl.)

III Dissociation of 3-adrenoceptor-induced effects on amylase secretion and cyclic AMP accumulation. Submitted for publication. (In collaboration with B. Carlsöö,,Â. Danielsson and L.-Â. Idahl.)

IV 3*- and 3?-adrenoceptor agonists have different effects on rat parotid acinar cells. Submitted for publication.

V Effects of neonatal sympathetic denervation on amylase secretion in the adult rat parotid gland: Difference in 3.- and 3o-adreno- ceptor response. European Journal of Pharmacology. In press.(In collaboration with B. Carlsöö,''Ä. üanielsson, S. Hellström and L.-Å. Idahl. )

VI Trophic effect of the sympathetic nervous system on the early development of the rat parotid gland. A Quantitative Ultra- structural Study. The Anatomical Record. In press. (Ir collabora­tion with G.D. Bloom, B. Carlsöö, A. Danielsson and S. Hellström).

5

INTRODUCTION

GENERAL ASPECTS

The integrity of the two major divisions of the autonomic nervous system, the sympathetic and parasympathetic, is of vital importance for maintaining homeostasis. The two systems consist of neurons and ganglia which innervate glands, heart, blood vessels, and visceral smooth muscle. In general these two systems have antagonistic effects. The parasympathetic part serves to conserve or restore the organism whereas the sympathetic part prepares the organism for "fight and flight". In salivary glands stimulation by the parasympathetic system induces a copious salivary flow which is poor in protein, whereas sympathetic stimuli induce secretion of much smaller amounts of saliva which on the other hand are rich in protein.

Pharmacologically, the two systems are classified as cholinergic (parasympathetic) and adrenergic (sympathetic) respectively. Choli­nergic is a term describing those nerve fibers which utilize acetyl­choline as neurotransmitter substance and the adrenergic division mainly secretes noradrenaline as its transmitter at the neuroeffector junction.

The sympathetic division of the autonomic nervous system origi­nates from cells located in the lateral horns of the spinal cord and extends from the base of the cranium to the coccyx. The preganglionic fibers leave the cord and establish contact with postganglionic nerve cells, located in ganglia on each side of the spinal cord or in the abdominal cavity. The nerve terminals contain the presynaptic machinery for synthesis, storage, release and metabolism of the neurotransmitters. The postsynaptic events, resulting from transmitter release or from blood borne substances, are initiated through interaction of trans­mitter and specific structures, receptors, on the effector cell.

The sympathetic neurons synthesize and store catecholamines in small quanta which are released in response to an appropriate stimulus. The transmitter substance then interacts with a specific receptor for catecholamines, the adrenergic receptor or adrenoceptor. Catecholamines released from e.g. the adrenal medulla can via the blood stream also interact with adrenoceptors. The transmitter-receptor interaction is the initial step in a complicated and incompletely understood series of events (schematically depicted in Fig 1).

The agonist is bound to the specific receptor with subsequent in­duction of an initial stimulus. The binding is influenced by several factors, among these the sterical configuration of the drugs. After transduction, modulation and amplification processes, the initial stimulus is further transmitted to the effector system and the physio­logical effect is produced (see Ariens and Simonis, 1976).

6

DRUG

INITIALSTIMULUS AMPLIFIER

SYSTEM

INPUTSIGNAL

•transduction

SIGNAL , AMPLIFICATION{MODULATION

PHYSIOLOGICALEFFECT

Fig 1

The adrenoceptor conceptFor a better understanding of modern concepts regarding sympathe­

tic influence on adrenoceptors in salivary glands it is necessary to outline the development of the adrenoceptor concept.

The general idea of receptive mechanisms for biologically active substances in effector cells traces its historical roots to the pioneer work of Langley (1905), Dale (1906) and Ehrlich (1909). Langley intro­duced the term "receptive substance" to designate the specific hypo­thetical component with which chemical substances i.e. nicotine or adrenaline must interact to cause their characteristic response in the effector organ. Langley also discussed the occurrence of different types of "receptive substances" within the same tissue and that such "substances" determined whether the organ would be stimulated or inhi­bited by a drug. Somewhat later Ehrlich proposed the term "receptor" for a hypothetical chemical structure with which a specific potential chemical agent would react.

Further support for the concept of receptive sites of interaction with catecholamines and the existence of multiple types of adrenergic receptors was suggested by the results of Dale. He demonstrated that ergot alkaloids inhibited some of the effects of adrenaline or of sympathetic nerve stimulation while other responses were unaffected.

The nature of chemical transmission was established in 1923 by Otto Loewe although the possibility of neuronal chemical transmission was suggested as early as 1877 by Du Bois Raymond (according to Ahlquist, 1979). Cannon and Rosenblueth (1933) forwarded the concept of two hypothetical adrenergic transmitters, an excitatory (sympathin E) and an inhibitory transmitter (sympathin I), to explain the varied

7

effects of adrenergic stimulation. This concept, however, did not unveil mysteries of the adrenergic neuroeffector junction. In the 1940's von Euler showed that noradrenaline was the adrenergic neurotransmitter.

In 1948 Ahlquist presented his dual receptor theory which is still the basis for our present understanding of adrenergic responses (Schematically summarized in Fig 2). Ahlquist suggested that the varied effects of sympathetic activity correspond to two different receptors at the effector cell and are not due to different neurotransmitters.The presence of two groups of adrenergic receptors could be demonstra­ted by applying pharmacologically closely related compounds with diffe­rent orders of potency in different tissues. The first class of recep­tors, called alpha (a), responded to the agonists in the following potency order: adrenaline > noradrenaline > a-methylnoradrenaline > a-methyladrenal ine > isoprénaline. The second type of adrenergic recep­tor, called beta (3), was stimulated by agonists in the potency order of isoprénaline > adrenaline > a-methyladrenal ine > a-methylnoradrena­line > noradrenaline. Stimulation of a-receptors caused contraction of the nictitating membrane, vasoconstriction and contraction of smooth muscle in the uterus and ureter whereas 3-stimulation caused positive cardiac inotropy and chronotropy, vasodilatation and relaxation of the uterus. Furthermore, Ahlquist showed that both a- and 3-receptors could occur in the same organ.

This classification of adrenergic receptors has been extended to include most effects mediated by catecholamines. However, it was the introduction of the 3-adrenoceptor blocking agent dichloroisoprenaline that made the dual adrenoceptor concept fully acceptable (see Furchgott, 1972). The a-receptor responses were early shown to be blocked by ergot alkaloids (Dale, 1906) and by the sympathetic antagonist dibenamine (Furchgott, 1972).

In 1967 Lands and co-workers showed that the 3-adrenoceptor popu­lation was not homogeneous and divided it, by absolute organ separation, into 3>|- and 32-adrenoceptors. This classification was based on the finding that the order of potency of several 3-agonists could be divi­ded into two main groups when tested on isolated organs. Albeit some controversial studies exist regarding this strict subdivision (Bristow et al., 1970; Levy, 1973; Ahlquist, 1976), this grouping is now welT^accepted. However, the 3-adrenoceptor subtypes co-exist in varying proportions depending on the tissue examined (Carlsson et aK> 1972, 1977; Daly and Levy, 1979; Minneman et al_., 1981; Wagner et_ al_., 1981). 3i-receptors in cardiac and adipose tissue have high and essentially equal affinity for the naturally occurring catecholamines adrenaline and noradrenaline, whereas 32"receptors in e.g. smooth muscle cells have a much higher affinity for adrenaline than noradrenaline.

A somewhat different theory regarding the subdivision of the 3- adrenoceptors has been discussed (Ariens, 1981). According to this view the 3-adrenoceptor is divided into a neurotransmitter receptor for nor­adrenaline (3-|) and a hormone receptor for adrenaline (32)* The con­centration of the 3-adrenoceptor subtypes has been shown to be indepen­dently regulated (Minneman et al., 1979) and the molecular substructure seems to be clearly different "(Venter et al., 1981). Recent observa-

THE D

EVEL

OPM

ENT O

F TH

E MO

DER

N AD

RENO

CEPT

OR C

ONC

EPT

8

03£

i •§CO H 03

£3 0 0

03 ?s

£ 03 O CO

•r3 4303 £

. £ te 43 4-3 ^co Q)

£ Ss *

43CO•. *£

43 £ CO O

*£ te £ 03 o 43 te £ 03 0303 03£ £

.. .- . •£ •£S) O Ü 43>* “ 03 03

03 03r-3 r-303 03CO Co

<0 03 O Co CO I t2te

Sitotote

te8

§

£ O -SO *£ COV +3 ®

£ 43 03 43O <0 43 £

*£ 03 03 V43 Ss r-3 ,<0 -£ *^ te,v £ £3 o5s O4 o'* Ss O

03 03 43 £ ^ *£V ^ rQ!s *£ *£

£ o <0 ’ oco 03 ter£o s s£»■ te Pl4 te

0340*£40

<0

£<teO

co £V oCO *<4* Co r-3 40 Q)03 <0 £ 43 03 03 CO £ ?-)V 40 r£ Ss £ s CÜ O ® Stete S

ft; 8 vFLi r-3 gte o 8te £

a £ 03 03 te

^ ^ V te ste£3A I ..

CO^ 40CO

*• ^te £<0 Q £ te Q) 03 40 40 O £PO

I8

03*££4O

CO £"03 4-3£ O —s

03 03 £ 4-3£ V-3 O 03

•O P33 £ ter-3 03 r£ s »O

03 40 £3 03 'K INO£ « Ss 0303 r-3 03 "03 Î--3 03 &

£3te £ £ O r-3 03 £ ps£ 03 03 *£ 03 £ v-A O V *r-3 H

« 03 £ £ £ r-i 03 O r-3 PX. \-3Ss £3 0 -o •£ 03 03 4-3 f-3 O 03O 03 V O CO *£ r-3 Q> Ss Ss 43 te£ 40 v g: FXi 03 43 Û4 43 £ C\]

«Kte, Ö £ *£ £ O 4-3 £ 03 O OO Ö3 O te O £ 03 03 03 4-3 £ Ss te

O 03 r-3 O *£ 4A .03 £ £ 03 O 03 te£ Ss 03 £3 ~£ O 4~» Û4 £4 S !s 43 P£

£ O 033 03 £ «K r£O ‘O 033 £ *£ 03 03

*0 40 03 03 cô S CO40 03 ‘O 4-3 03 44i 43*0 CO te 03 CO CO £

A 1CO £ CO

te 03 ’O 43 •• *£ O •• *£ O •£‘O 03 r£ £3 43 £ 03 ’£ 43 £ £ •£ A 4-3 £<-£ te £ 03 CO O S 43 CO O O 4-3 CO O£ 03 ‘O 0 •£ 033 £3 03 •g S1 •£ 03 •£ te

‘O SO te £ 03 M r-3 A 4^ r-3 £ A £ 03S Q 4-3 £ £ O 43 03 £ O O 43

03 g §> £ 03 O33 £ +3 K •£ te te £03 ‘O 03 ’£ 03 03 03 *£ 43 te 03 03

*0 40 43 -£ H r£ V-3 4-3 03 H40 03 03 CO CO 03 03 v> CO 43 03 03

•» £ £ ’£ £ £ "03 03 £ £03 V *£ co 03 £3*£ V 03 O 03 ï-3 £i*£ •£

£3 CO 43 43 £3 03 03 4-3 43 r£ 03 *£ 03 43 43te CO 03 03 03 r-3 *£ £ 03 03 Cs 03 V 03 £ 03 03CO 40 r-3 Q3 03 O "03 03 03 03 co £ "0 O 03 03 0303 CO 03 r-3 r-3 Cb Ss 43t4^ Oi C3 CO 4-3 r-3 r-3Ss 0 £3 03 03 *£ 03 0 03 03 r-S Ss .03 O 03 03te PO 03 OO 00 <o P^ CO CO *> 0Q Oj ,£* te Orj te

SIto§

3

<s

CCL

03£ 03

*£ £ 03 v-3 £ 03 te£ £ 03

03 “r-3g

g g &^3 O co te H

te ^ PO ^ te te

•3>PO

9

tions have indicated that 3-adrenoceptors may be present presynapti- cally and that they may potentiate the transmitter release (Dahlöf, 1981).

The a-adrenergic receptors have also been divided into o^- and a2-adrenoceptor subtypes (see Exton, 1981) due to differences in affinities for various selective a-agonists and antagonists. Originally, oc-j-adrenergic receptors were considered as postsynaptic and o^-receptors as presynaptic receptors which inhibit the transmitter release. However, this strict anatomical division has been shown to be invalid, since a2-adrenoceptors have been demonstrated postsynaptically.

Over the last years it has become apparent that several and varied factors such as aging, temperature, changes in thyroid function, adre- nalactomy, myocardial ischemia, and skeletal muscle activity, modify the properties of a- and 3-receptors. (Kunos, 1978; Pointon and Banerjee, 1979; Minneman et aj., 1981). Thus, adrenoceptors do not seem to be static entities in Tiïe plasma membrane but they are subject to dynamic regulation by a wide variety of physiological and pathophysio­logical factors.3-adrenoceptor activation and the cyclic AMP system

3-Adrenoceptor stimulation seems to involve activation of the. membrane bound enzyme adenylate cyclase with a subsequent production of the second messenger cyclic AMP (Robinson et al_., 1971). The 3- adrenoceptor and the adenylate cyclase proteins are products of diffe­rent genes (Insel et al., 1976) and they are suggested to float freely in the phosphol ipiT~mTTieu of the cell membrane. 3-adrenoceptor acti­vation causes an enhanced turnover rate of phospholipids with subse­quent changes in fluidity within the membrane. This may enhance the ability of a receptor to couple to adenylate cyclase (Strittmatter et al., 1981). The activation of the 3-adrenoceptor by agonists and the süïï- sequent transmission of the signal to adenylate cyclase is dependent on guanyl-triphosphate (GTP). These effects of GTP are controlled by a guanine nucleotide regulatory protein (see Lefkowitz and Hoffman, 1980). Moreover, it seems that the 3-adrenoceptor determines the affinity but a component distal to the receptor is of importance for the relative drug intrinsic activities (Pike et aK, 1979).

With respect to a-adrenoceptor activation this will in some cases induce calcium fluxes and hydrolysis of phosphatidyl inositol. In other cases inhibition of the enzyme adenylate cyclase is brought about (see Exton, 1981). Recent studies have shown that the c^-adreno- ceptor subtype mediates those metabolic effects which involve phos­phatidyl inositol turnover and the release of intracellular calcium as well as a calcium influx. Activation of the a^-subtype decreases the adenylate cyclase activity by a mechanism which is independent of calcium but probably dependent on GTP (Exton, 1981).

SALIVARY GLANDS

The character of adrenoceptors and the subsequent intracel1ular events that take place in salivary glands upon stimulation of these receptors are similar to those found in other tissues and organs. Over

10

the years an ever increasing number of reports have been published on various aspects of salivary gland secretion. Therefore, for the sake of clarity and simplicity, the remaining part of this chapter will deal mainly with modern concepts of sympathetic influence on parotid glands with special respect to the 3-adrenoceptor.

Salivary glands consist in general of epithelial secretory end- pieces, ducts and connective tissue containing blood vessels and nerves. Myoepithelial cells are found around the acini and ducts. The acinar cells produce the primary secretory fluid consisting mainly of water, inorganic ions and macromolecules. The duct system modifies the ionic composition and transports the secretory fluid to the oral cavity.

In salivary glands, macromolecules, (e.g. amylase) synthesized by the classical route common to most exportable proteins, are sequestered intracellularly in membrane-bounded secretory granules. Upon stimula­tion, the secretory granules are transferred to the luminal surface of the cells where, after membrane fusion, the granule content is dischar­ged into the acinar lumen by the process of exocytosis (Amsterdam et al. 1969; Butcher and Putney, 1980; KanagaSuntheram and Lim, 1981). However, an alternative mechanism may exist for amylase discharge. Thus some enzyme molecules may reach the saliva directly from cisternae of the rough reticulum without being prepackaged into secretory granules (Garrett et aj_., 1978).Adrenoceptors

Since the first discoveries of the autonomic control of salivary glands, substantial information has accumulated regarding chemical transmission and sympathetic influence as well as adrenoceptor struc­tures in salivary glands (see e.g. Emme!in, 1967; Schneyer and Hall, 1967; Schramm and Sel inger, 1975; Butcher and Putney, 1980). Salivary gland function depends on the integrity of both parasympathetic and sympathetic innervation. The innervation pattern differs markedly bétween different animals and glands (Garrett, 1974). The glands used in our studies, rat parotid and guinea-pig submandibular glands, are well supplied with adrenergic nerves in intimate association with the acinar cells (Garrett, 1974; Aim et al., 1973; Bloom et a]_., 1977).These sympathetic nerves, the celTlxxïies of which lie in the superior cervical ganglion, innervate the gland parenchyma via adrenoceptors.The adrenergic innervation of the blood vessels seems to be of diffe­rent origin (Aim and Ekström, 1976).

Numerous studies have established the existence of both a- and 3-adrenoceptors as well as cholinergic receptors in the studied glands (see Schramm and Sel inger, 1975; Butcher and Putney, 1980). Activation of the a-adrenoceptor as well as the cholinergic receptor mainly re­gulates the water and inorganic ion content of the primary secretory fluid. However, some release of enzymes is also seen upon activation of these receptors. The a-adrenoceptor primarily involved in the rat sal ivary glands seems to be of the a>|-subtype (Bylund and Martinez, 1980; Danielsson, unpublished).

More recently interest has been focused on the role on undecapep- tides (e.g. Substance P) in salivary glands. These agonists seem to

11

have an effect both on salivation and amylase release (Butcher and Putney, 1980; Brown and Hanley, 1981), and seem to engage the same intracellular pathways as a-adrenergic and cholinergic processes (see Butcher and Putney, 1980). a-Adrenergic and cholinergic agonists as well as isoprénaline increase cyclic GMP accumulation in salivary glands but the physiological significance of this is not fully understood (Wojcik et al_., 1975).

Activation of the 3-adrenoceptors in parotid and submandibular glands leads predominantly to exocytosis with release of enzymes e.g. amylase and peroxidase (Amsterdam et al., 1969; Carl sob et a]_., 1974a; Schramm and Selinger, 1975; Butcher ancf Putney, 1980). Previous studies have given some indication that the 3-adrenoceptor is mainly of the 31- subtype (Thulin 1972; Butcher et_ al_., 1975; Au et ä]_. , 1977; Carlsöö et a]_., 1978). The 3-adrenoceptor-induced depolarisation is closely related to protein secretion and is blocked by 3i-receptor antagonists (Petersen and Pedersen, 1974; Emme!in et al., 1980).

The 3-adrenoceptor-induced amylase release is believed to be asso­ciated with the activation of membrane bound adenylate cyclase, and cyclic AMP has been accepted as the classical second messenger in the 3-adrenergic regulation of enzyme release (Carlsöö et å]_., 1974b; Schramm and Selinger, 1975; Butcher and Putney, 198117 Yu et aK, 1979). Furthermore, activation of cyclic AMP-dependent protein kinase and phosphorylation of microsomal proteins, seem to be involved in 3- adrenoceptor activation (Dowd et £]_., 1981). However, the role of cyclic AMP as the exclusive mecTTator of the 3-adrenoceptor-induced sec­retory process has been questioned (e.g. Harfield and Tenenhouse, 1973; Butcher et al_., 1975).

It is now established that calcium plays an important role in adrenoceptor-mediated effects in salivary gland secretion (Butcher and Putney, 1980; Berridge, 1980). a-Receptor, 1 ike cholinergic, stimula­tion requires the presence of extracellular calcium whereas 3-adreno­ceptor-mediated effects on cyclic AMP accumulation and amylase release are dependent on intracellular pools of calcium. An increased level of intracellular calcium seems to be of importance for the secretory process and is brought about by either a phospholipid dependent in­creased influx (a-activation and cholinergic receptor stimulation) or by a sodium dependent release of intracellular stores of calcium in the 3-adrenoceptor mediated secretory process (Butcher and Putney, 1980; Watson et, a]_., 1980).Receptor effect interaction

a- and 3-adrenoceptors as well as cholinergic receptors often seem to interact with each others effects. If, for example, sympathetic nerves are stimulated at the same time that a low level of parasympat­hetic stimulation is applied, a much greater secretion of saliva and amylase is brought about than the sum of the two separate effects. This would seem to indicate that the enzyme secretion caused by sympathetic stimulation is dependent on a parasympathetic evoked fluid secretion (Asking et aj_., 1979; Templeton, 1980). It has also been observed that the 3-adrenoceptor stimulated amylase release is enhanced by a-agonists (Petersen and Ueda, 1975). Furthermore, carbachol and a-adrenoceptor

12

Stimulation inhibits the stimulatory effect of 3-adrenoceptor activa­tion on adenylate cyclase activity and cyclic AMP accumulation (Butcher et aK , 1976; Harper and Brooker, 1977; Oron et al_., 1978a, b).

The role of myoepithelial cells in the secretory process has been outlined in a recent review (Garrett and Emmelin, 1981). The myoepithe­lial cells receive nerve fibers from both branches of the autonomic nervous system. Cholinergic receptor and sympathetic activation as well as circulating catecholamines cause contraction of the cells via a- adrenoceptor mechanisms whereas 3-adrenoceptors do not seem to be in­volved in this cellular activity.

Apart from the direct effect on the release process, 3-agonists also seem to stimulate amylase synthesis (Byrt, 1966; Grand et al., 1969), whereas cholinergic and a-agonists inhibit this process TMcPher- son and Hales, 1978).Variation in sympathetic activity

The effect of variation in sympathétic activity on salivary gland function has attracted considerable interest over a long period of time. Intensive stimulation of the 3-adrenoceptor with isoprénaline or via sympathetic nerves, has been shown to induce hypertrophic and hyperplastic changes in parotid and submandibular acinar cells (Selye et al., 1961; Brown-Grant, 1961; Schneyer, 1969; Barka, 1970; Muir et äT., 1 975; Bloom et aj_., 1979). However, morphological changes after a period of decreasecTsympathetic activity, are less pronounced (Schneyer and Hall, 1967; Klein, 1979).

The sensitivity or responsiveness of the effector cells in sali­vary glands is to a major degree influenced by their innervation. When functional nerve activity is changed the sensitivity of the receptors on distal effector cells is slowly altered in a direction which com­pensates for the change in neural input (Emmelin, 1965; Stefano and Perec, 1981). Salivary glands that have been subjected to pretreatment with adrenoceptor agonists, show a desensitization to further stimula­tion and also a decrease in receptor density (Strittmatter et al., 1977; Barka and Burke, 1977; Burke and Barka, 1978; Bloom et al_., 1979"; Roscher et al., 1981). On the other hand, a decreasecFexposure of the effector ceTTs to nerve stimuli, effected e.g. by sympathetic denerva­tion, in parotid and submandibular glands leads to supersensitivity and an increase in adrenoceptor density (e.g. Arnett and Davis, 1979; Pointon and Banerjee, 1979b;; Ludford and Talamo, 1980; Stefano and Perec, 1981). Hence, the general law of denervation can be applied to salivary glands (Cannon, 1939).

AIM OF INVESTIGATION

Against the background outlined above the aim of the present thesis can be briefly summarized as:

—To characterize the 3-adrenoceptor induced enzyme release in serous salivary glands with regard to secretory dynamics and related changes in cyclic AMP

13

—To follow up initial studies on the effects of chronic 3-adreno­ceptor stimulation on gland structure and enzyme content and

— To further study the effects of sympathetic denervation on salivary gland structure and function and to focus main interest on the effects of neonatal denervation.

MATERIALS AND METHODS

Animals and tissue preparation

Female Sprague-Dawley rats, 9-12 weeks old (II - VI) and male guinea-pigs, 10 - 15 weeks old (I, III) were deprived of food for 18 h before use, but had free access to water. The glands studied, rat paro­tid and guinea-pig submandibular glands are both almost entirely serous and contain large amounts of e.g. amylase. Furthermore, they are homo­geneous and surrounded by a connective tissue capsule; they can thus easily be dissected free. Experiments were as a rule started between 8 pm and 9 pm to avoid diurnal variations. The glands were rapidly re­moved under either ether or sodium pentobarbital anesthesia and care­fully freed from extraglandular tissue. Pieces of uniform size were excised from the glands under a stereomicroscope and were subjected to the various analyses outlined below.Surgical sympathectomy

The superior cervical ganglion on one side was removed from new­born rats (within 4 hours after birth) or from adult animals (6 weeks of age) (V, VI). Nine weeks later the sympathectomized and the contra­lateral parotid glands were removed and various parameters of secretion were studied as well as the ultrastructural stereology of the glands (V, VI). The success of the sympathetic denervation was verified in electron- and fluorescence microscopical studies (V, VI). A further control was the lack of secretory response in gland tissue when stimu­lated with 5-hydroxydopamine (V). Finally, the rats as a rule exhibited ptosis on the operated side but not on the contralateral control side.Chronically increased sympathetic activity

This was mimicked by treating adult rats (8 weeks of age) twice daily for 10 days with the non-selective 3-adrenoceptor agonist isopré­naline, the 3i-agonist prenalterol or terbutaline, a 32_selective agonist. Doses were 2.3 nmol/g animal weight. The last injection was given 24 h before the experiments. After 18 h starvation the parotid and submandibular glands were excised and specimens were subjected to stereological analyses or used for measuring total amylase content (IV).Studies on secretion and cyclic AMP accumulation

Gland specimens were transferred to a Krebs-Henseleit bicarbonate buffer (pH 7.4). The buffer was supplemented with pyruvate, glutamate and fumarate as well as albumin (1 mg/ml) and glucose (0*6 mg/ml). For both the batch-incubations and perifusion experiments the same basal medium as above was used. Before incubation a piece of tissue was

14

usually taken for analyses of total amylase and cyclic AMP contents.

The in vitro secretory studies were performed with a multichannel microperifusion system (I, II, III, V) or a batch-incubation technique (II, V). The former, a micromodification of an earlier developed system (Idahl, 1972), makes it possible to obtain concomitant measurements of amylase secretion and glandular cyclic AMP contents over short periods of stimulation; details are given in paper I. The batch incubation technique is described in paper II.

In vivo secretion was studied after single injections of the 3- adrenoceptor agonists mentioned above, drug concentration was 2.3 nmol/g animal weight (IV).Amylase and cyclic AMP assay

Amylase was assayed by a micromodification of the 3,5-dinitrosal i- cylate (DNS) method (Danielsson, 1974). One unit of amylase is defined as the activity liberating reducing groups which corresponds to 1 ymol of maltose monohydrate per min at 25° C. Amylase release in the batch- incubation experiments is expressed as the percentage of amylase in medium in relation to the total amylase content in media plus homoge­nate. In the perifusion studies amylase release is expressed as U/min. Cyclic AMP was extracted from the tissue and assayed radioimmunologi- cally (Heilman et aj_., 1974). Cyclic AMP is expressed as pmol/mg tissue dry weight. ForTetails see respective paper.Electron microscopy

The techniques employed for the ultrastructural studies including the stereological methods are described in papers IV and VI. Tissue specimens were fixed in glutaraldehyde and postfixed in osmium tet- roxide. After a cold buffer rinse the specimens were dehydrated in graded ethanol solutions and embedded in Epon 812. Thick (1 ym) as well as thin (70 nm) sections were cut and poststained with uranyl acetate and lead citrate. The parotid gland specimens from the experi­mental and control animals were fixed and embedded in parallel.Stereological methods

The stereological analyses were carried out on randomly taken electron micrographs of parotid acinar cells and using to the point counting method of Weibel (1979) employing a multipurpose grid. The diameters of the secretory granules in the electron micrographs were measured in a Zeiss TGZ-3 particle size analyzer. The stereological calculations were performed according to Helander (1978) and are de­scribed in detail in papers IV and VI. It should be pointed out that values obtained by stereological methods are subject to various in­herent errors mainly due to tissue preparative procedures including sectioning. In comparative studies, however, these errors diminish in importance (IV, VI).Drugs

The compounds metoprolol, H 35/25 and IPS 339 which have been used in our most recent work were kindly supplied by Hassle AB, Mölndal, Sweden. All other information concerning drugs is found in the separate papers.

15

RESULTS AND DISCUSSION

CHARACTERIZATION OF (3-ADRENOCEPTOR INDUCED AMYLASE RELEASE AND ITS CORRELATION TO GLANDULAR CYCLIC AMP

Experimental proceduresThe present secretion studies were mainly performed with two diffe­

rent in vitro techniques: a microperifusion- and a batch-incubation one. The former is very useful for studying the dynamics of secretion as se­veral parameters can be measured simultaneously and they can also be closely correlated even to short time factors. Batch-incubations are more suitable for the study of e.g. agonist-antagonist interactions.

In vitro techniques reduce the effects of certain factors which influence and modify adrenoceptor-induced amylase release in vivo. Rem­nant nerve terminals and catechol-0-methyl transferase (CÖMT) can, how­ever, to some extent complicate this kind of investigation. Differences in inactivation of the agonists used, as has been shown for noradrena­line and isoprénaline (see Furchgott, 1972) may also modify the influ­ence of different agonists on the adrenoceptors. The employed drugs could further act indirectly by releasing endogenous stores of noradre­naline from the nerve endings. However, pretreatment with reserpine did not modify the secretory effect of the tested agonists with the sole exception of 5-hydroxydopamine (II). Moreover,the synthetic 3-subtype selective agonists used do not contain a catechol moiety and are con­sequently persistent to degradation by COMT.

According to Furchgott (1972) a proper pharmacological analysis of the 3-adrenoceptor requires blocking of the a-receptor. However, in our system, a-antagonists such as phenoxybenzamine and phentolamine were found to initiate amylase release; this discharge was further found to be blocked by the 3-adrenoceptor antagonist propranolol (Table 1). A potentiating effect of a-antagonists on 3-adrenoceptor induced amylase release as well as on cyclic AMP accumulation has been reported in the literature (Butcher et , 1975; Jirakulsomchok and Schneyer, 1979;Yu et aj_., 1979). On the other hand, the synthetic 3-selective agonists used+exert no detectable effect on a-receptor mediated processes such as K -release from salivary glands (Danielsson, unpublished).Amylase secretion

Amylase discharge was detectable as early as 15 sec after stimula­tion was begun. Continued exposure of the slices to noradrenaline re­sulted in an initial rapid secretory peak followed by a slow declining second phase (I). Amylase release was still clearly detectable after 60 min of perifusion. Discontinuation of noradrenaline stimulation rapidly reduced amylase secretion to basal levels, and a second lower peak was obtained after repeating the stimulus (I). This may indicate a rapid desensitization phenomenon (Harper and Brooker, 1977) or may be due to depletion of preexisting amylase stores.

16

Table 1.

Amylase secretion with an a-adrenoceptor antagonist alone and in com­bination with a 3“antagonist.

n % amylase release

Non-stimulated control 6 18.1 - 3.2Ihenoxybenzamine (10 **M) 6 28.8 -2.9Ehenoxybenzamine (10^M) + propranolol (2 x 10 °M)

6 18.8 - 3.5

Experimental design as described in paper II for the batch-incubation technique. Mean values - SEM.

The properties of adrenoceptors can be described in terms of the relative potencies of a series of adrenergic agonists and antagonists eliciting or inhibiting a particular tissue response (Fig 2). In other words the two measurable parameters of interest are the potency, i.e. the concentration range over which a drug can produce a graded physio­logical response,and the intrinsic activity, i.e. the biological re­sponse which can be obtained by a maximally effective concentration of the drug (Furchgott, 1972).

The development of high affinity selective 3-j- and 32-adreno­ceptor agonists and antagonists has made feasible more specific analy­ses of 3-adrenoceptor-mediated effects. However, the selectivity of these drugs is not absolute and varies for each individual drug. By studying the concentration-response relations of different adrenocep­tor agonists inducing amylase secretion from the parotid gland we re­corded the potency order of prenalterol > isoprénaline > noradrenaline > adrenaline > salbutamol > terbutaline (II). The 3i-agonist prenal­terol (Carlsson et al., 1977; Minneman et a]_. , 1981) was much more potent and showeTVTTigher intrinsic activity than the 32_selective agonist terbutaline. The relatively strong effects of the 32-adreno­ceptor agonists salbutamol and terbutaline, within higher concentration ranges (II, III), may be due to a non-selective interaction with the 3i-adrenoceptors (Hedberg et al_. , 1981).

In the experiments in which selective 3-adrenoceptor antagonists were used, we could further substantiate the dominance of 3^-receptors in the amylase release process (Fig 3). The non-selective 3-antagonist propranolol was the most effective blocker of amylase secretion. The 3i-selective antagonist metoprolol was also an effective blocker, whereas H 35/25, a 32_selective antagonist, (Minneman ejt aj^., 1981) was virtually without effect on maximal noradrenaline- and prenalterol- induced amylase release.

INH

IBIT

ION

OF NO

RA

DR

ENA

LIN

E (10'6

M) IND

UC

ED AM

YLA

SE REL

EASE

17

X 100-

ANTAGONIST ( M )

A B

Hg 3

Effects of the non-seleative ß-adrenoceptor antagonist pro­pranolol (*——»), the èj-seleotive antagonist metoprololÙ.-----A, and the antagonist H 35/25 (o------o)on maximal amylase release from rat parotid gland induced by A) noradrenaline and B) prenalterolf respectively. The batch- incubation technique, as described in paper II, was used. The points represent mean values of 5+separate experiments. The maximal amylase release was, 39. 3 - 3.3 % for prenalterol and 32.5 * 3.1 % for noradrenaline. Bon-stimulated control re­lease was 15.4 - 2.9 %.

The greater capacity of prenalterol, as compared to terbutaline, to elicit amylase release was also evident in the in vivo experiments (IV). A single injection of prenalterol (2*3 nmol/g animal weight) caused a pronounced decrease in parotid amylase as well as an exocytotic discharge of secretory granules (Fig 4). Terbutaline at the same con­centration was far less effective (IV).

18

Fig 4

Electron micrograph illustrating the general appearance of acinar cells 60 min after injection of prenalterol. The lumen (L) has increased in size, its irregular horders in­dicating areas of fusion with secretory granules (SG). Unsecreted granules preserve their original density. Some flocculent secretory material can be seen in the lumen (12,200X).

Recently developed techniques utilizing radiolabelled 3-adrenergic ligands, make it possible to identify and study receptor binding sites. Ligand studies on parotid glands have shown that the binding sites in the glands are of the 31~type (Au et al_. , 1977; Ludford and Talamo, 1980). Preliminary studies in our laboratories using ^H-dihydroalpre- nolol (a 3-antagonist) binding to crude parotid membrane preparations also indicate a predominance of 3^-adrenoceptors in this gland. There are, however, some shortcomings with radioligand binding studies which may complicate the interpretation of results obtained. There.is an evident risk that radioligand binding sites occur, which are not uni­quely linked to the biochemical machinery responsible for producing the characteristic physiological response of the tissue. Another problem is that the pharmacological and physiological selectivity of

19

several 3-j- and 32_active agents seems to be lost in binding studies (Minneman et a]_. , 1981).

Our experiments including both 3-subtype selective agonists and antagonists clearly indicate that the sympathetically evoked amylase release in parotid glands is mainly mediated by the 3^-adrenoceptor. However, the present results do not exclude the possibility that 32- adrenoceptors also play a role in the amylase release process. As shown in Fig 3 the 32-antagonist H 35/25 was without effect in blocking maximal noradrenaline- and prenalterol-induced amylase release. More­over, H 35/25 and another 32_antagonist, IPS 339 (Minneman et al_., 1981), could completely inhibit the amylase release induced by a suFmaximal concentration of terbutaline (Fig 5). The metoprolol inhibition curve displayed a two-phase character (Fig 3 A) suggesting the presence of more than one single receptor population involved in the noradrenaline- induced amylase discharge. The increased maximal secretory response to terbutaline recorded in neonatally denervated glands also lends support to such an assumption (V).

The co-existence of both 3-adrenoceptor subtypes (3-| and 32) in one and the same tissue mediating similar effects, has been shown earlier. Smooth muscle relaxation (trachea, blood vessels) and chro- notropy (heart muscle) can be induced by activation of either or both of the 3-adrenoceptor subtypes (see Minneman £t al_., 1981; Wagner et al_. , 1981). The biological significance of this phenomenon is un- cTear. It has been proposed that the 3>|-receptor acts as a neuronal receptor and as such is more sensi vite to noradrenaline whereas the 32- receptor is a hormonal one, more sensitive to circulating adrenaline (Ariens, 1981). Omini and collaborators (1979) have forwarded the in­teresting hypothesis that when one subtype is inhibited or destroyed the other may take over on a functional level.The correlation between amylase secretion and cyclic AMP accumulation in the 3-adrenoceptor-mediated secretory process

After exposure to 3-adrenoceptor agonists a rapid accumulation of cyclic AMP has been found to take place in salivary glands. This accu­mulation is closely correlated to amylase release during the initial phase of the secretory process (I). However, after maximal amylase re­lease was achieved the cyclic AMP levels continued to rise; this was especially the case in the guinea-pig submandibular gland (I, III).There is thus a lack of correlation between cyclic AMP and amylase release during the later phases of the secretory process.

A deviation between cyclic AMP levels and amylase release at high agonist concentrations has also been reported by Durham et aK, (1974). Butcher and co-workers (1975) noted that low concentrations of adreno­ceptor agonists, although sufficient to cause amylase release, were without detectable effects on total cyclic AMP content in rat parotid gland. Those authors also found that (+)-propranolol markedly blocked isoprenaline-induced cyclic AMP accumulation without causing a parallel decrease in amylase release. Furthermore, it has been demonstrated that the presence of EGTA inhibits cyclic AMP accumulation but is without effect on protein release (Harfield and Tenenhouse, 1973).

20

ANTAGONIST (H

Fig 5

Effect of the ^2~se^eo^ve antagonists H 35/25 (0----- 0)and IPS 339 (X-----X) on a submaximat terbutaline-inducedamylase release. The batch-incubation technique, described in detail in II, was used. The points represent mean values of 4 separate experiments. The amylase release was 20.4 ± 2.1 %.

It has earlier been shown that if either carbachol or an a- adrenergic agonist are present together with a 3-agonist in the in­cubation-bath this will cause an augmented amylase release - above that expected by each agonist separately (Petersen and Ueda, 1975; Harper and Brooker, 1978; Asking and Gjörstrup, 1980). On the other hand, the cyclic AMP increase after 3-adrenoceptor stimulation can be inhibited by either carbachol or a-adrenergic agonists (Butcher et al., 1976a, b; Harper and Brooker, 1977; Oron ejt ajl^., 1978a, b). Thus, although it is accepted that cyclic AMP is of great importance for 3-adrenoceptor

21

stimulated amylase release, it is at present not possible to explain why inhibition of cyclic AMP accumulation would lead to enhanced amy­lase secretion.

Terbutaline caused a significant accumulation of cyclic AMP in both rat parotid and guinea-pig submandibular glands, whereas prenal- terol was without significant effects both in vitro and in vivo (III, IV). The effects observed in salivary glands are to some degree similar to those noted in cat myocardium and soleus muscle. Prenalterol and another 3i-selective agonist, tazolol, fail to significantly activate myocardial adenylate cyclase and elevate cyclic AMP in spite of a mar­ked physiological muscle response (Vaquelin et al_., 1976; Hedberg et al_., 1981). In the cat soleus muscle, whicïï mainly contains 32“ adrenoceptors (Minneman et aj_., 1981), procaterol, a 32"selective agonist, stimulates adenylate cyclase activity whereas prenalterol is without effect (Hedberg and Mattsson, 1981). Furthermore, prenalterol inhibits isoprenaline-stimulated adenylate cyclase activity in both the soleus muscle as well as in the myocardium, whereas selective 32“ agonists are without significant effect (Hedberg and Mattsson, 1981).

Since prenalterol seems to be a full agonist in eliciting amylase release, the lack of significant effects of the drug on cyclic AMP accumulation both in vitro (III) and in vivo (IV) supports the assump­tion that 3i-adrenoceptor-induced amÿTase secretion under certain con­ditions could be mediated by mechanisms not directly involving cyclic AMP. On the other hand, the rise in cyclic AMP levels after stimulation with noradrenaline or terbutaline appears essential for 3?“adrenoceptor- induced amylase release. It is possible though that prenalterol may induce high levels of cyclic AMP locally in small compartments of the acinar cell whereas the total content of cyclic AMP in the gland tissue would remain low. On the other hand, the high concentration of cyclic AMP noted after noradrenaline and terbutaline stimulation could be due to activation of several compartments as has been suggested by Hedberg and Mattsson (1981) for cardiac tissue. This theory is difficult to accept, since the majority of 3-adrenoceptors seems to be of the 3-|- subtype. Another explanation could be the spare receptor theory (see Hedberg and Mattsson, 1980). However, for the relevance of this theory, a direct one-to-one coupling between the 3-adrenoceptor and adenylate cyclase has to be proven. This theory appears less probable if one considers that: a) prenalterol seems to be a full agonist with respect to amylase release and is without significant effect on cyclic AMP, b) terbutaline is a weak secretagogue but its effect on cyclic AMP is pronounced.

The results obtained with 3-adrenoceptor subtype selective ago­nists raises doubts concerning the role of cyclic AMP as an essential mediator in 3-adrenoceptor induced amylase release.

22

EFFECTS OF CHRONIC VARIATION IN SYMPATHETIC ACTIVITY ON SECRETORY BEHAVIOUR AND MORPHOLOGY

Decreased sympathetic activityIn our studies a state of decreased sympathetic activity was indu­

ced by neonatal (within 4 hours after birth) or adult (6 weeks of age) extirpation of the superior cervical ganglion on one side (V, VI). The neonatal sympathetic denervation was performed in animals in which the parotid glands were still under development, whereas the adult denerva­tion was carried out when gland development was completed (VI). Nine weeks after the operation the denervated glands were used for the experiments and the contralateral glands served as controls. The effi­cacy of the denervation technique was proven by applying electron- and fluorenscence microscopical techniques (V, VI), as well as by the de­monstration of a decreased vn vitro secretory effect of 5-hydroxydopa- mine (V). This drug liberates endogenous stores of noradrenaline at sympathetic nerve endings (Thoenen and Tranzer, 1971).

Neonatal sympathetic denervation caused hypotrophie alterations in the adult animals (VI). The stereological and morphological analyses showed that acinar cell size was decreased and the number of granules as well as the mean size of the individual granules was reduced. On the other hand ganglionectomy performed in adult animals was without signi­ficant effects (VI).

Gland fine structure and especially cellular granule content was well correlated to the tissue amounts of amylase. Neonatal ganglionec­tomy caused a significant decrease in amylase (V), whereas amylase levels in the adult denervated glands were not altered as compared to controls (unpublished).

The sympathetic nervous system thus seems to play a fundamental role in the early maturation of the parotid gland. Support for this conclusion is found in a study on rat submandibular gland by Srini- vasaan and Chang (1977). The trophic role of noradrenaline-containing neurons on postsynaptic cells of various organs has also previously been noted (Bevan, 1975; Jonsson et aj_., 1979).

A decreased transfer of stimulatory agonists to effector cells - induced by e.g. interruption of the pathways of the autonomic nervous system - will cause a state of supersensitivity in such cells (Cannon, 1939). In the sympathetic nervous system,two phenomena can be identi­fied which contribute to the development of an increased responsiveness after sympathetic denervation. Initially a strong response to submaxi­mal stimulatory concentrations of agonists is seen. This early pheno­menon corresponds to the loss of nerve terminals containing mechanisms for inactivating catecholamines and it is supposed to be mainly an acute presynaptic feature unrelated to a supersensitivity of the effec­tor cell. This prejunctional event covers only those amines which are substrates for neuronal uptake mechanisms. A second, slowly developing postsynaptic event appears to involve changes at or beyond the site of which receptor interaction takes place; this is recorded as an enhanced maximal response to different agonists. This postsynaptic component is

23

further characterized by its lack of specificity (Trendelenburg, 1963, 1966; Kebabian and Zatz, 1980; Stefano and Perec, 1981).

These effects of sympathetic denervation on the secretory process in salivary glands have been extensively studied in adult animals (see e.g. Emmelin, 1967; Stefano and Perec, 1981). However, in the present study, main interest was focused on the effects of neonatal denervation.

Neonatal sympathetic denervation caused an increased maximal sec­retory response, not only to the 3-adrenoceptor agonists isoprénaline and noradrenaline, but also to dopamine, carbamylcholine and the dibu- tyryl analogue of cyclic AMP (DBcAMP) (V). The enhanced maximal amylase release caused by noradrenaline and isoprénaline in combination with the unspecific supersensitivity to non-adrenergic drugs points towards postjunctional changes in the effector cells (Trendelenburg, 1963,1966; Kebabian and Zatz, 1980; Stefano and Perec, 1981). An increase in 3-adrenoceptor density has also been described in salivary glands subjected to sympathetic denervation (Pointon and Banerjee, 1979;Arnett and Davis, 1979; Ludford and Talamo, 1980).

The augmented maximal response to the 3?-agonist terbutaline was especially notable after neonatal ganglionectomy, whereas on the other hand, maximal amylase release following stimulation with the 31-agonist prenalterol was not significantly altered (V). Furthermore, a relati­vely small increase in maximal response was observed after stimulation with noradrenal ine, which has a preference for 3>|-receptors, while the maximal response to the non-selective 3-agonist isoprénaline was mar­kedly enhanced (V).

Prenalterol has been shown to be a potent enzyme secretagogue, whereas terbutaline has only minor effects on amylase release in normal glands (II). Sympathectomy appears to transform terbutaline from a par­tial to a full secretory agonist, well in the class of prenalterol and noradrenaline. These results could reflect an altered proportion of the 3-adrenoceptor subtypes. In fact, preliminary studies demonstrate an equalization of the effects of selective 3i~ and 32-adrenoceptor anta­gonists in inhibiting the binding of a radiolabel led 3-adrenoceptor ligand (^H-dihydroalprenolol) from denervated parotid crude membrane preparations (unpublished). An intact sympathetic nervous system thus seems to be of great importance for the differentiation of the 3- adrenoceptor into its subtypes, just as it seems to be for the matura­tion of the parotid gland as a whole (VI). Non-parallel changes in 3^- and 32"adrenoceptor densities under different experimental conditions have also been reported for other tissues (Minneman et al_., 1979;Barney eit aK » 1980).

The perifusion experiments also demonstrated that the secretory dynamics were altered in the neonatally denervated glands (V). Nor­adrenaline is known to activate both a- and 3-adrenoceptors on acinar cells as well as a-adrenoceptors on myoepithelial cells. Myoepithelial cells are of importance for raising the speed of the initial flow of saliva and maintaining a high flow rate of secretion (Emmelin and Gjörstrup, 1973). A state of supersensitivity is also apparent in these

24

CONCENTRATION (H )

Fig 6

Concentration-response curve for noradrenaline. Amylase release is expressed as a percentage of amylase released into the medium in relation to total amylase activity in medium plus homogenate. Each point represents mean values - SEM for indicated number of experiments and animals.0 ---- 0 represents glands from animals (9 weeks of age)in which sympathetic denervation was performed two weeks earlier by avulsion of the superior cervical ganglion.9----- 9 contralateral control glands.

* p < 0.05

25

cells after sympathetic denervation (Emmelin and Thulin, 1973; Peusner et aj_., 1979a). In addition to the alterations in 3-adrenoceptor re­sponse, the initial high amylase release in the neonatally denervated glands could be related to a supersensitivity of the a-adrenoceptors of the myoepithelial cells (V).

Other factors besides quantitative and qualitative changes of the 3-adrenoceptor population may also play a role in the altered sensiti­vity of the neonatally denervated acinar cells. This is suggested by the increased levels of non-stimulated and stimulated cyclic AMP as well as the enhanced secretory response to DBcAMP (V). Similar results have been obtained in studies on rat submandibular and pineal gland (Arnett and Davis, 1979; Kebabian and Zatz, 1980). The pineal gland is also supplied with sympathetic nerves from the superior cervical gang­lion. An increase in 3-adrenoceptor density and adenylate cyclase as well as a decreased phosphodiesterase activity have been reported in this gland after sympathetic denervation (Kebabian and Zatz, 1980). Thus, denervation-induced supersentitivity may either be caused by modifying the number of 3-adrenoceptors and/or by altering the meta­bolism of intracellular compounds.

The supersensitivity observed in the neonatally denervated glands seems to differ from that noted in the adult denervated tissues two weeks after ganglionectomy. The most evident effect in the latter appears to be a leftward shift of the noradrenaline concentration- response curve, without significant change in the maximal response (Fig 6) (Peusner et , 1979b).Increased sympathetic activity

In the present study a state of increased sympathetic activity was mimicked by daily injections of various 3-agonists (IV). Repeated activation of the rat parotid 3-adrenoceptor with isoprénaline gives rise to both hypertrophy and hyperplasia of the acinar cells, with concomitant alterations in the secretory granule population (Selye et_ aj_., 1961; Barka and Burke, 1977; Bloom et 1979).

Stimulation with the 3-subtype selective agonists revealed that prenalterol (3-j) induced almost similar effects on the granule polu- lation as did isoprenaline. This was expressed as changes in granule size and substructure as well as decrease of amylase contents (IV).The 32-agonist terbutaline was virtually without effect on the granule population (IV). However, acinar cell size was significantly increased after repeated stimulation with both prenalterol and terbutaline, but less pronounced than after isoprénaline treatment.

These results suggest that the effects of the non-selective ß- agonist isoprénaline on acinar cell size and granule population may be independently regulated. The effects on the granule population could be mediated by ß^-adrenoceptors (II, III, IV); the increase in acinar cell size is possibly brought about by activation of other mechanisms.

26

20 (mini

Fig 7

Dynamics of amylase release in rat parotid gland in response to continuous exposure to noradrenaline (10~^M). Each point represents the mean of two different experiments and animals. Amylase release is expressed as units/min □----□ corre­sponds to perifusion experiments performed on glands from animals (9 weeks of age) subjected to ten days treatmentwith isoprénaline (2.3 nmol/g animal weight). £---- Hrepresents glands from control animals peri fused in paral­lel. The perifusion technique is described in detail in paper I. The gland content of amylase in animals subjected to isoprénaline treatments was 14,150 units/g wet weight, and 20,945 units/g wet weight in controls.

27

A dissociation of the pathways for stimulus-secretion coupling and stimulus-growth coupling has also been proposed by other investi­gators (Sans and Galanti, 1979). A discrepancy exists between the op­timal doses of isoprénaline producing on the one hand DNA synthesis and on the other hand maximal secretion (Brenner and Stanton, 1970; Johnson and Sreebny, 1973). By using a series of isoprénaline analogues, Kirby and co-workers (1969) were able to show that amylase secretion could be initiated without any increase in DNA synthesis. However, extended analyses are necessary to elucidate if the increase in acinar cell size is a specific phenomenon or is secondary to the changes in secretory activity.

The secretory response in salivary glands subjected to chronic 3-adrenoceptor stimulation, is altered in an opposite manner compared to that seen in denervated glands. The former glands are subsensitive to further stimulation with 3-agonists both in vivo (Byrt, 1966;Barka, 1970) and in vitro (Bloom et al., 197"97 Barka and Burke, 1977), and this subsensitivity is clearly demonstrated with the perifusion system (Fig 7). The decrease in sensitivity could be due to both a diminished 3-adrenoceptor density and/or alterations in the cyclic AMP system (Burke and Barka, 1978; Bloom et aK , 1979; Revis and Durham, 1979; Roscher et al., 1981).

GENERAL SUMMARY

1. Two different in vitro techniques, a batch-incubation system and a multi-channel microperifusion system as well as in vivo in­jections, have been employed in order to characterize sympatheti­cally induced amylase release and its correlation to cyclic AMP levels in rat parotid and guinea-pig submandibular glands.

a) A dominance of 3^- over 32"recePt°rs was established with the aid of different 3-adrenoceptor agonists and antagonists.

b) Continuous noradrenaline stimulation caused a rapid initial enzyme discharge (within 15 seconds), which was closely correlated to a rise in cyclic AMP levels.

c) In later stages of the noradrenaline-stimulated enzyme release there was a lack of correlation between enzyme release and tissue levels of cyclic AMP. Amylase release gradually declined, whereas tissue cyclic AMP continued to increase. In contrast to its potent secretagogic effect, the 3i-agonist prenalterol failed to signifi­cantly stimulate tissue cyclic AMP, whereas terbutaline, a weak stimulator of amylase release, caused a clear increase in cyclic AMP accumulation.

It is concluded that amylase release is mainly mediated by 3'1- adrenoceptors , whereas the 32-adrenoceptors are of minor impor­tance in the amylase secretory process. Furthermore, it is evident that a 3-adrenoceptor-mediated amylase release can take place with no or only slight increase in glandular cyclic AMP.

28

2. Superstimulation of the rat parotid ß-adrenoceptor was induced by long-term treatment with the non-selective 3-agonist isoprénaline, the 3^-agonist prenalterol and the 32-agonist terbutaline.

-a) Isoprénaline caused a dramatic increase in acinar cell size and granule content as well as alterations of granular substructure and a decrease in amylase content.

b) The ß-j-agonist prenalterol exhibited almost similar effects on the granule population as did isoprénaline, whereas terbutaline was without effect.

c) Both of the ß-subtype selective agonists induced an increase in acinar cell size although it was less pronounced compared to the effect of isoprénaline.

The results indicate that chronic ß-adrenoceptor-induced effects on acinar cell size and granule population could be due to acti­vation of different 3-receptor subtypes.

3. Sympathetic denervation was performed neonatally (within 4 hours after birth) or in adult animals (6 weeks of age) on one side.The contralateral side was used as a control. Nine weeks later the glands were used for experiments.

a) U1 trastruc.tural stereological analyses demonstrated hypotrophie changes in acinar cell size and granule content as well as a decline in amylase content after neonatal denervation. Adult denervation was without effect on these parameters.

b) The neonatally denervated glands displayed an enhanced postjunc­tional secretory response to 3-adrenoceptor agonists. Especially notable was the increase in maximal response to the ß2-agonist terbutaline whereas the ß-j-agonist prenalterol was without signi­ficant effect.

c) Increased levels of non-stimulated and stimulated cyclic AMP as well as an enhanced response to dibuturyl cyclic AMP were recorded in the neonatally denervated glands but not in contralateral controls.

It is concluded that neonatal sympathetic denervation alters the ß-adrenoceptor-mediated secretory response in adult glands. This change shows both quantitative and qualitative features, viz an enhanced maximal response and an increased importance of ß2~ adrenoceptors. The "supersensitivity" observed in neonatally denervated glands seems to differ in character from that found in adult denervated glands. Finally, an intact sympathetic nervous system seems to be of principal importance for the overall matu­ration of the salivary gland and for the differentation of the ß-adrenoceptor.

29

ACKNOWLEDGEMENTSThe present thesis work was carried out at the Department of Histology and Cell Biology, University of Umeå. These years have been a valuable experience in many ways, both professionally and otherwise. Many Thanks to all of you who have contributed to the stimulating atmosphere in which I have 'had thé privilege to work.

I especially wish to express my sincere gratitude to the following persons:

Gunnar D Bloom, former head of the department, for his genuine interest in this work, his willingness to help out at all times, and especially his view that no problem is too small for a professor,

Bengt Carlsöö andÅke Danielsson, my teachers and supervisors who introduced me into this scientific field, and through their enthusiasm, gave me energy and encouraged me to go on in soite of many difficulties,

Marianne Borg,Siw Domeij,Kristina Forsgren,Ann-Louise Grehn and Berit Lundström, for excellent technical

^stance and invaluable help with all the experiments based on rough ideas and \incomplete instructions,

|Lars-Åke Idahl, for contributing his expert knowledge in our stimu­lating collaborative work both in a

practical and theoretical sense,

'Sten Hellström, for teaching me the surgical techniques and for stimu­

lating discussions,

Inge-Bert Täljedal, present head of the departmen and a very stimulating 01 for placing working facilities at my dispos and for showing me kin interest over the years,

Herbert Helander, who introduce me into the field of ultrastructur. stereology,

Ann-Christin Edlund,Lena Jokela andMona Britt Jonsson, for skilful secreterial assistance throughout the years, and Gerd Karlsson and Birgitta Karlsson for final typing of the manuscript,

Anders Andersson, Holger Bäckström Torsten Edström, Pekka Fredriksson,Yngve Lindström and Erik öhlund for valua^ technical assistance

This investigation was supported by grants the Swedish Medical Research Council(19X-04492 Svenska Medicinska Sällskapet, the Mångberg foundation, and the faculty of Medicine, University of Umeå, Sweden.

30

REFERENCES

Ahlquist, R.P. (1948) Am. J. Physiol. 153:586-600Ahlquist, R.P. (1976) Prog. Drug Res. 20:27-42Ahlquist, R.P. (1979) Trends in Pharmacol. Sci. 1:16-17Alm, P., Bloom, G.D. and Carlsöö, B. (1973) t. Zellforsch 138: 407-420Alm, P. and Ekström, J. (1976) Experientia 33:523-52'4Amsterdam, A., Ohad, I. and Schramm, M. (1969) J. Cell Biol. 41:753-773Ariens, E.J. (1981) Trends Pharmacol. Sci. 2:170-172' ~~Ariëns, E.J. and Simonis, A.M. (1976) In Beta-adrenoceptor Blocking

Agents (Saxena, P.R. and Forsyth, R.P., eds) pp 3-27, Elsevier/ North-Holi and, Amsterdam

Arnett, C.D. and Davis, J.N. (1979) J. Pharmacol Exp. Ther. 211:394- 400 _______Asking, B., Delfs, U., Emmelin, N. and Gjörstrup, P. (1979)

Experientia 35:1336-1337Asking, B. and Gjörstrup, P. (1980) Acta Phys. Scand. 109:415-420 Au, D.K., Malbon, C.C. and Butcher, F.R. (1977) Biochim. Biophys. Acta

500:361-371Barka, T. (1970) Lab. Invest. 22:73-80.Barka, T. and Burke, G.T. (1977) Histochem. J. 9:453-466 Barney, C.C., Katowich, M.J., Fregiy, M. J. and Tyler, P.E. (1980)

Can. J. Physiol. Pharmacol. 58:1170-1173 Berridge, M. (198Û) Trends. Pharmacol. Sci. 1:419-424 Bevan, R.D. (1975) Cire.' Rës. 37:14-19Bloom, G.D., Carlsöö, B., Danielsson, Å., Gustafsson, H. and

Henriksson, R. (1979) Med Biol 57:224-233 Bloom, G.D., Carlsöö, B. ancf Gustafsson, H. (1977) Acta Anat. 97:

291-299 ” —Brenner, G.M. and Stanton, H.C. (1970) J. Pharmacol. Exp. Ther. 173:

166-175 'Bristow, M., Sherrod, T.T. and Green, R.D. (1970) J. Pharmacol. Exp.

Ther. 171:52-61 'BrowrTïïrant, K. (1961) Nature 191:1076-1078Brown, C.L. and Hanley, M.R. (1981) Br. J. Pharmacol. 73:517-523 Burke, G.T. and Barka, T. (1978) Biochim. "Biophys. Acta 539:54-61 Butcher, F.R., Goldman, J.A. and Nemerovski, M. (1975) Biochim.

Biophys. Acta 392:82-94Butcher, F.R., McBride, P.A. and Rudich, L. (1976a) Mol. Cell Endocrin.

5:243-254 ~ 'Butcher, F.R., Rudich, L., Emler, C. and Nemerovski, M. (1976b)

Mol. Pharmacol. 12:862-870Butcher, F.R. and Jr. Putney, J.W. (1980) Adv. Cyclic Nucleotide Res.

13:215-247 ~ —Bylund, D.B. and Martinez, J.R. (1980) Nature 285:229-230 Byrt, P. (1966) Nature 212:1212-1215 Cannon, W.B. (19l9T~Àm. J. Med. Sci. 198:737-750 Cannon, W.B. and Rosenölueth, A. (1933) Am. J. Physiol. 104:557-574 Carlsson, E., Äblad, B., Brändström, A. and Carlsson, B. (1972) Li fe

Sci. 11:953-958 ~Carlsson, E., Dahlöf, C.-G., Hedberg, A., Persson, H. and Tångstrand,

B. (1977) Naunyn Schmiedeberg1s Arch. Pharmacol. 300:101-105

31

Carlsöö, B., Danielsson, A., Marklund, S. and Stigbrand, T. (1974a) Acta Physiol. Scand. 92:263-271

Carlsöö, B., Danielsson, Ä., Marklund, S. and Stigbrand, T. (1974b) Acta Physiol. Scand. 91:203-210

Carlsoö, B., Danielsson, Â. and Henriksson, R. 1978) Br. J. Pharmacol.62:364-366 '

Danielsson, Â. (1974) Analyt. Biochem. 59:220-234 Dahlöf, C.D. (1981) Acta Physiol. Scand. suppl. 500 Dale, H.H. (1906) J. Physiol. "34:163-206Daly, M.J. and Levy, 'fi.P. (1979) In Trends in Autonomic Pharmacology

(Kalsner, S., ed.), vol. 1, pp. 347-38”S. Urban and Schwarzenberg, Baltimore and Munich.

Dowd, F.J., Watson, W.L., Horio, B., Lau, Y.-S. and Park, K. (1981) Biochem. Biophys. Res. Comm. 101:281-288

Durham," J.P.", Baserga, F. and Butcher, F.R. (1974). In Control ofProfileration in Animal Cells (Clarkson, B. and Baserga, R., eds.) "pp7 5913-607." Cold Spring Harbor Laboratory, N.Y.

Ekström, J. (1980) Acta Physiol. Scand. 108:253-261 Ehrlich, P. (1909) Ber. Dtch. Chem. Ges. 42:17-32 Emmelin, N. (1965) Experi'entia 21:57-65Emmelin, N. (1967) In Handbook of Physiology. Alimentary Canal II.

Chap. 37:595-632 ' 'Emmelin, N. and Gjörstrup, P. (1973) J. Physiol. 230:185-198 Emmelin, N. and Thulin, A. (1973) Br. J. "Pharmacol. 48:73-79 Emmelin, N., Grammp, W. and Thesleff, P. (1980) J. Physiol. 302:183-

195 —Exton, J.H. (1981) Mol. Cell. Endocrinol. 23:233-264 Euler, U.S.v. (1946) Acta Physiol. Scand. 12:73-97 Furchgott, R.F. (1972)' Tn Handbook of experimental pharmacology

(Blascko, H. and Muscholl, E7,'eds.), Voi. 33, Catecholamines, pp. 283-335, Berlin: Springer-Verlag

Garrett, J.R. (1974) In Secretory Mechanisms of Exocrine Glands. (Thorn, N.A. and Petersen, O.M., eds.) pp 17-27. Munksgaard, Copenhagen.

Garrett, J.R. and Emmelin, N. (1981) Med Biol. 57:1-28 Garrett, J.R., Thulin, A. and Kidd, A. (19787. Cell. Tiss. Res,

188:235-250Grand, R.J. and Gross, P.R. (1969) J. Biol. Chem. 244:5608-5615 Harfield, B. and Tenenhouse, A. (1973) Can. J. Phys. Pharmacol. 51:

997-1001Harper, J.F. and Brooker, G. (1977) Mol. Pharmacol. 13:1048-1059 Harper, J.F. and Brooker, G. (1978) Mol‘7 Pharmacol. 14:1031-1045 Hedberg, A., Carlsson, E., Fellenius, E. and Lundgren, B. (1981)

Naunyn-Schmiedeberg1s Arch. Pharmacol. In press Hedberg, A. and Mattsson, H. (1981). J. Pharmacol. Exp. Ther. In press Helander, H.F. (1978) Cell Tiss. Res. 189:287-303 Heilman, B., Idahl, L.-Â.rLernmark, A. and Täljedal, I.-B. (1974)

Proc. Natl. Acad. Sci. 71:3405-3409 Idah!, X.-A. (1972). AnäTyt. Biochem. 50:386-398 Insel, P.A., Maguire, M.E., Gilman, A.G., Boure, H.R., Coffino, P.

and Melmon, K.L. (1976) Mol. Pharmacol. 12:1062-1067 Jirakulsomchok, D. and Schneyer, C.A. (1979). Am. J. Physiol. E 371 -

E 385Johnson, D.A. and Sreebny, L.M. (1973) Lab Invest. 28:263-269

32

Jonsson, G., Hallman, H. and Felten, D.L. (1979) Neurosci. Lett. suppl 3. p. 366

Kanagasuntheram, P. and Lim. S.C. (1981) J. Physiol. 312:445-454 Kebabian, J.W. and Zatz, M. (1980) In The Cell Surface and Neuronal

Function (Cotman, C.W., Poste, G. and Nicolson, G.L., eds.), pp 303-349. Elsevier/North-Holland, Biomedical Press

Kirby, K.C., Swern, D. and Baserga, R. (1969) Mol. Pharmacol. 5:572- 579

Klein, R.M. (1979) Cell Tissue Kinet. 12:411-423 Kunos, G. (1978) Ann. Rev. Pharmacol. Toxicol. 18:291-311 Lands, A.M., Arnold, A., McAüliff, J.P., Ludüena, F.P. and Brown, T.G.

(1967a) Nature 214:597-598Lands, A.M., Lu'duena, F.P. and Buzzo, H.J. (1967b) Life Sci. 6:

2241-2249Langer, S.Z. (1974) Biochem. Pharmacol. 23:1793-1800 Langley, J.N. ( 1905)~~J7 PhysToT. 33:374-413Lefkowitz, R.J. and Hoffman,B.B. (1980) Trends Pharmacol. Sci. 1:

314-318Levy, B. (1973) J. Pharmacol. Exp. Ther. 186:134-142 Ludford, J.M. ancT Talamo, ^B'.R. ( 1980)" J7 Biol. Chem. 255:4619-4627 McPherson, M.A. and Hales, C.N. (1978) Biochem. J. 176:855-863 Minneman, K.P., Dibner, M.D., Wolfe, B.B. and Molinoff, P.B. (1979)

Science 204:866-868Minneman', K.P., Pittman, R.N. and Molinoff, P.B. (1981). Ann. Rev.

Neurosci. 4:419-461Muir,' T.'C.Y Poll oc h, D. and Turner, C.J. (1975) J. Pharmacol. Exp.

Ther. 195:372-381Omini, C., Somerville, A.R., Folco, G.C., Rossoni, G. and Puglisi, L.

(1979) Pharmacol. Res. Comm. 11:529-535 Oron, Y., Kellogg, J. and Larner, J. (1978a) Mol. Pharmacol. 14:

1018-1030Oron, Y., Kellogg, J. and Larner, J. (1978b) FEBS LETTERS 94:331-334 Petersen, 0. and Pedersen, G.L. (1974) J. Membrane Biol .~~16:353-362 Petersen, 0. H. and Ueda, N. (1975) J, Physiol. 260:1p-2p Peusner de, I.C.W., Stefano, F.J.E. and Perec, C.J. (1979a)

Naunyn-Schmiedeberg1s Arch. Pharmacol. 308:211-216 Peusner de,_I.C.W., Perec, C.J. and Stefano, F.J.E. (1979b)

Naunyn-Schmiedeberg1s Arch. Pharmacol. 308:217-221 Pike, L.J., Limbird, i.E1. and Lefkowitz, R.J. (1979). Nature 280:

502-504Pointon, S.E. and Banerjee, S.P. (1979a) Biochim. Biophys. Acta

583:129-132Pointon, S.E. and Banerjee, S.P. (1979b) Biochim. Biophys. Acta 584:

231-241Revis, N.W. and Durham, J.P. (1979) J. Histochem. Cytochem. 27:1317-

1321Robison, G.A., Butcher, R.W. and Sutherland, E.W. (1971) Cyclic AMP.

New York. Academic Press.Roscher, A.A., Wiesman, U.N. and Honegger, U.E. (1981) J. Pharmacol.

Exp. Ther. 216:419-424Sans, J. and Galanti, N. (1979) Cell. Molecul. Biol. 25:107-112 Schneyer, C.A. (1969) Proc. Soc. Exp.' Biol. Med. 131:71 -75 Schneyer, C.A. and Hall, H.D. (1967) In Secretory Mechanisms of

Salivary Glands. (Schneyer, L.H. and Schneyer, C.A., eds.7 pp. 155-177. Academic Press, London

33

Schramm, M. and Selinger, Z. (1975) In Stimulus-SecretionCoupling in the Gastrointestinal Tract (Case, R.M. and Goebell,H., eds.), pp.” 49-64.' MTR Press Ltd, London

Selye, H., Cantin, M. and Veilleux, R. (1961) Growth 25:243-248 Srinivasan, R. and Chang, W.W.L. (1977) Cell Ti ss. Res. 180:99-109 Stefano, F.J.E. and Perec, C.J. (1981) Trend? Pharmacol. Sci. 2:107-109 Strittmatter, W.J., Davis, J.N., and Lefkowitz," R.J? (1977) J. Biol.

Chem. 252:5478-5482Strittmatter, W.J., Hirata, F. and Axelrod, J. (1981) Adv. Cyclic

Nucleotide Res. 14:83-91 Templeton, D. (1980) Pflügers Arch. 384:287-289 Thoenen, H. and Tranzer, J.P. (1971) In Progress in Brain Research.

(Eränkö, 0., ed.). Vol. 34, pp. 223-‘236. Elsevier, Amsterdam Thulin, A. (1972) Experientia 28:420 Trendelenburg, U. (1963) Pharmacol. Rev. 15:225-276 Trendelenburg, U. (1966) Pharmacol. Rev. 18:629-640 Vaquelin, G., Lacombe, M. L.,' Guel laen, G., Strosberg, A.D. and

Hanoune, J. (1976) Biochem. Pharmacol. 25:2604-2608 Venter, J.C., Fraser, C.M., Soiefer, A.I., Jeffrey, D.R., Straqss,

W.L., Charlton, R.R. and Greguski, R (1981) Adv. Cyclic Nucleotide Res^ 14:135-143 ~ ”

Watson, E.L., Friedman, J. and Siegel, I.A. (1980) Life Sci. 26: 1919-1926

Wagner, J., Nick, B., Rohm, N. and Schümann, H.J. (1981) Arch, int. Pharmacodyn. 249:26-38

Weibel, E.R. (1979) Stereological Methods. Vol. 1 Academic Press,London

Wojcik, J.D., Grand, R.J. and Kimberg, D.V. (1975) Biochim. Biophys. Acta 41 1:250-262

Yu, JT^Tf., Burns, S., Jirakulsomchok, D. and Schneyer, C.A. (1979)Ala. J. Med. Sci. 16:166-169

Now KARIN..