40 Biochimiea et Biophysica Acta, 675 (1981) 40-45 Elsevier/North-Holland Biomedical Press

BBA 29611

/3-ADRENERGIC RECEPTOR STIMULATED PEROXIDASE SECRETION FROM RAT LACRIMAL GLAND

ZEEV Y. FRIEDMAN, MARGALIT LOWE and ZVI SELINGER

Department of Biological Chemistry, The Hebrew University of Jerusalem, Institute of Life Sciences, Jerusalem {Israel)

(Received October 23rd, 1980) (Revised manuscript received February 1 lth, 1981)

Key words: Peroxidase secretion; f3-Adrenergic receptor; Dibutyryl cyclic AMP; {Lacrimal gland)

Incubation of rat extraorbital lacrimal gland slices with the 13-agonist isoproterenol caused peroxidase secretion but no K + release. The peroxidase secretion was inhibited by propranolol. Addition of dibutyryl cyclic AMP or adenosine 3',5'-cyclic phosphorothioate to lacrimal slices produced peroxidase secretion at a higher rate than that obtained with optimal concentration of isoproterenol. Methyl isobutylxanthine is also a strong stimulator of per- oxidase secretion. Peroxidase activity was determined by a modified sensitive guaiacol method. Membrane frac- tion of lacrimal cells was shown to contain an isoproterenol-stimulated adenylate cyclase activity. It is therefore suggested that there is a/3-adrenergic receptor in the rat lacrimal gland and that its stimulation causes activation of an adenylate cyclase which leads to peroxidase secretion.

Introduction

It has been established that cholinerglc stimulation causes secretion of peroxidase from the rat lacrimal gland similarly to the secretion of exportable proteins from other exocrine glands [1]. It was also shown that acetylcholine stimulates fluid secretion from the lacrimal gland in vivo [2,3], and that 4SCa2+ is taken up by the lacrimal cells upon cholinerglc stimulation [4]. Subsequently, it was shown that the stimulation of the muscarinic receptor in the rat lacrimal gland is associated with the release of K ÷ [5-7] .

Recently, it was established that there is also an a-adrenergic receptor in the rat lacrimal gland and that its stimulation causes K ÷ release [6] and perox- idase secretion [8]. Electrophysiological studies have shown that both acetylcholine and adrenalin act on the lacrimal acinar cell membrane by opening up channels that are mainly permeable to K ÷ [9]. Intro-

duction of Ca 2+ into lacrimal gland cells by means of the Ca 2+ ionophore A23187 causes K ÷ release [7].

In the present study we have used a sensitive assay for peroxidase and tested its secretion by lacrimal slices. It was found that isoproterenol as well as the cyclic AMP analogs dibutyryl cyclic AMP and adeno- sine 3',5'-cyclic phosphorothioate induces perox- idase secretion from the slices. The phosphodiesterase inhibitor N-methyl-3-isobutylxanthine also exhibited the same effect. These results, together with the find- ing of an isoproterenol induced adenylate cyclase in lacrimal cell membranes, point to the existence of a 13-adrenerglc receptor which is involved in peroxidase secretion from lacrimal ceils.

To complete the information about the Ca 2+ ionophore A23187 we show that the ionophore causes peroxidase secretion from lacrimal slices in addition to its known activity of K ~ release [7].

Materials and Methods

Abbreviations: Hepes, N-2-hydr oxyethylpiperazine-N'-2 - ethane sulfonic acid; Mops, 3-(N-morpholino)propanesulfonic acid;

Adenosine 3',5'-cyclic phosphorothioate was a gift from Prof. F. Eckstein from the Max-Planck Institut. 3-Isobutyl-l-methylxanthine was purchased from

0304-4165/81]0000-0000]$02.50 © Elsevier/North-Holland Biomedical Press

41

Aldrich, and guaiacol from Sigma. The sources of other chemicals were given in previous papers [11-13].

Lacrimal slice system Male albino rats, 180-220 g, fed ad libitum and

kept at 24 -+ I°C, were used. The rats, under ether anaesthesia, were killed by cutting through the heart. Both extraorbital lacrimal glands were removed and immediately cut into small slices with razor blades. The slices were first incubated in medium 1 (Krebs- Ringer solution buffered with Hepes: 118 mM NaC1, 4 mM KCI, 2.5 mM CaCI: (unless otherwise noted), 1.5 mM KH2PO4, 1.18 mM MgSO4, 25 mM Hepes buffer (pH 7.4) and 5 mM/3-hydroxybutyrate). The medium was bubbled with 100% O: before use and during each opening of the vessel.

The slices were preincubated for 10 min in a rotary shaking bath at 37°C operating at 180 rev./ rain. At the end of the preincubation period, which served to remove material from ruptured cells, the slices were poured over nylon stocking stretched over a beaker and washed with fresh medium. The pooled slices were divided into portions, each equivalent to about one lacrimal gland and placed in individual scintillation vials containing 2 ml medium. To initiate secretion a concentrated solution of secretagogue was added to the medium to give the desired final con- centration and the vessel was gassed and stoppered. At different time intervals, 20/A aliquots of the me- dium were removed for K ÷ and/or peroxidase deter- mination. Whenever additions were made or aliquots removed, the vessels were gassed with O~ and then tightly closed with rubber stoppers. At the end of the experiments, 2 ml of medium was added to each ves- sel and the tissue was ground with a polytron (Kine- matic GmbH) homogenizer. K* and peroxidase activ- ity was determined on aliquots of 20 tA of the homo- genate. Each experiment was repeated at least six- times.

Peroxidase determination The guaiacol method of peroxidase determination

[10] was used with some modifications. The compo- sition of the assay mixture was: 0.1 M potassium phosphate buffer, pH 8.1/1% guaiacol (88 mM)/ 0.0006% H202 (0.2 raM). The assay mixture was pre- pared freshly from stock solutions as follows: I M

potassium phosphate, pH 8.3 (10 ml), 90 ml H20, I ml guaiacol, 0.03% H202 fleshly prepared from a stock solution of 30% (2 ml).

To 20/A sample, 1.4 ml assay mixture was added. Absorption was read I min after mixing at 437 rim. The assay was performed at room temperature. This assay is 30-times more sensitive than the diamino- benzidine assay which was used by others [8].

K ÷ determination Samples of 20 pJ were diluted with 3 ml of distil-

led water and the concentration of K ÷ was deter- mined in an atomic absorption spectrophorometer.

Rat lacrimal membrane preparation Eight rats were killed and the lacimal glands were

removed and collected in 0.3 M sucrose, 0.05 mM EDTA, 1 mM mercaptoethanol, 0.1 f~g/ml diphenyl- phenylenediamine. The glands were homogenized in 40 ml collecting medium for 2 rain in a loose Teflon homogenizer and faltered through nylon mesh. The homogenate was centrifuged in a GLC centrifuge in the cold for 10 rain at 1200 rev./min (approx, 250 ×g), the supernatant was discarded, the pellet was resuspended in 16 no] collecting medium and recentrifuged as described above. The supernatant was discarded. The pellet was resuspended in 16 ml of Tris medium which contained: 50 mM Tris.HC1 buffer (pH 7.5)/0.05 mM EDTA/I mM mercaptoetha- nol/0.1 #g/ml diphenylphenylenediamine. The sus- pension was homogenized in a tight Teflon homo- genizer, centrifuged for 4 min at 1800 rev./min. The supernatant was carefully transferred with a Pasteur pipette into Sorval centrifugation tubes and centri- fuged for 15 rain at 14000 rev./min. The superna- tant was discarded and the pellet was resuspended in 16 ml Tris medium (for washing) and recentrifuged as described above. The supernatant was discarded and the pellet was suspended in preservation medium which contained: 20 mM Mops (pH 7.4)/2 mM MgC12/1 mM EDTA/2mM mercaptoethanol. The membrane preparation was quickly frozen in liquid nitrogen until use.

Adenylate cyelase activity determination Adenylate cyclase activity was determined as pre.

viously described [14] with the omission of bovine serum albumin from the reaction mixture.

42

R e s u l t s

The effect of isoproterenol The addition of isoproterenol 10 -s M to the slices

caused peroxidase secretion at a rate of 0.25%/min (Fig. 1), which is slower than the rate achieved by stimulation of the cholinergic or a-adrenergic recep- tors. Peroxidase secretion by isoproterenol was inhib- ited by 10-SM propranolol. Isoproterenol did not cause K ÷ release, whereas stimulation of the a-adre- nergic or cholinerglc receptors resulted in K+ release (Fig. 2).

Effect o f dibutyryl cyclic AMP and adenosine 3',5'- cyclic phosphoro thioate

The second messenger of the/3-adrenergic receptor is cyclic AMP. Two permeable derivatives of cyclic AMP, dibutyryl cyclic AMP and adenosine 3',5'- cyclic phosphorothioate [15] were studied. After the addition of 2 mM dibutyryl cyclic AMP there was a lag of about 20 min, followed by a linear peroxidase secretion at a rate of 1.2%/rain (Fig. 3). No K ÷ release was observed (Fig. 2).

40

g

I 1

Jo 20 30 40 50 60 TIME (rain)

Fig. 1. Time course of peroxidase secretion from lacrimal slices. Agonists and antagonists were added at time zero. • • , 10 -5 M carbachol; • - - • , 10 -4 M adrenalin+ 10-4 M propranolol+ 10 -5 M atropine; a % 5 ,ug/ml A23187; • • , 10 -5 M isoproterenol; u - Q, 10 -s M isoproterenol + 10 -s M propranolol; o 4 , no additions.

40

l O

*630

kt3

tO 20 30 40 50 60 TIME (min)

Fig. 2. K + release from lacrimal slices by various agonists. Q - - o, 10 -5 M carbachol; • •, 10 -4 M adrenaline + 10 -4 M propranolol+ 10 -s M atropine; . *, 5 /~g/rnl A23187; × X, 10 -5 M isoproterenol; o --% 2 mM dibutyryl cyclic AMP; a- A, 1 mM 34sobutyl-l-methyl- xanthine; o- ~>, basal release.

After addition of 2.9 mM adenosine 3',5'-cyclic phosphorothioate there was a lag of 10 rain followed by a gradual increase in the rate of peroxidase secre- tion to a rate of 1.75%/min (Fig. 3). It should be noted that while the rate of secretion by carbachol gradually diminished (Fig. 1), the rate of secretion by adenosine 3',5'-cyclic phosphorothioate increased with the time of incubation, It is also noteworthy that the secretion caused by the cyclic AMP deriva- tives is much higher than that produced by optimal concentration of isoproterenol.

Stimulation of adenylate cyclase activity in mem- brane preparations

A membrane fraction from lacrimal glands was prepared as described in the Materials and Methods section. The stimulation of adenylate cyclase activity by the addition of isoproterenol was measured. The results are summarized in Table I. A 4-fold increase in adenylate cyclase activity is caused by isoproterenol. A still higher activation is achieved in the presence of

80,

7O

o 60

"6

50

N Q 30 X i o ~ , w ~ 20

I0

o ,o 2o 3'0 ,io 5o TiME (rain)

Fig. 3. Peroxidase secretion in presence of cyclic AMP deriva- tives. All the reagents were added at time zero. o-- o, no addition; . - - - , 2 mM cyclic AMP; ~ ~, 10 -s M isoproterenol; • - - • , 2raM dibutyryl cyclic AMP; • •, 2.9 mM adenosine 3',5 '-cyclic phosphorothioate.

TABLE I

STIMULATION OF ADENYLATE CYCLASE ACTIVITY BY ISOPROTERENOL

Adenylate cyclase activity was determined as previously described [14]. Where indicated GTP 0.1 mM, isoproterenol 20 uM and NaF 10 mM were added to the adenylate cyclase assay. Also, guanosine 5'-(-,r-thio)triphosphate, a GTP analog which is not hydrolysed by GTPase [17] was added at a con- centration of 0.1 raM. The reaction was initiated by the addi- tion of membrane preparation and terminated after 10 rain incubation at 37°C.

Additions Adenylate cyclase activity pmoles/ mg per 10 rain

None 38 GTP 100 Isoproterenol + GTP 390 Isoproterenol + guanosine 5'-(7-thio)

triphosphate 1200 F- 880

43

the non-hydrolyzable GTP analog, guanosine 5'-(7- thio) tr iphosphate. These results resemble results of other systems which have adenyl cyclase [ 17].

Effect of 3-isobu tyl-l-methylxan thine After addit ion of the phosphodiesterase inh ib i tor

3 - i sobuty l - l -methy lxan th ine at a concen t ra t ion of

1 mM there was a lag of 10 rain followed by a l inear

peroxidase secretion at a rate of 1.1%/min (Fig. 4). No K ÷ release was observed (Fig. 2).

The effect was no t abolished by the presence of

propranolol , a tropine and phen toa lmine together or separately. A synergistic effect was observed with 10 -4 M 3- i sobuty l - l -methy lxan th ine and 10 -s M iso-

proterenol (Fig. 4.).

Peroxidase secretion by the Ca2÷-ionophore A23187 The second messenger of bo th the cholinergic and

e-adrenergic s t imulat ion is k n o w n to be Ca 2+ [16].

Accordingly, it has been shown that the Ca 2+ iono-

phore A23187 causes K ÷ release from lacrimal slices

4O

30

' °

I0 20 30 40 50 60

Time (min) Fig. 4.Time-course of peroxidase secretion in presence of methylisobutylxanthine. All reagents were added at time zero. o o, no addition; o - - ~ , 1 mM 3-isobutyl-1- methylxanthine; ~ ~, 1 0 ~ M 2-isobutyl-l-methylxan- thine; # #, IO-SM isoproterenol; • • , lO-4M 3-isobutyl-1 -methylxanthine + 10-S M isoproterenol.

44

[7]. If Ca 2+ is the second messenger of cholinergic and a-adrenerglc stimulation in the lacrimal system then its introduction into the cell by means of the Ca2+-ionophore A23187 should also cause peroxidase secretion. Indeed, we found that addition of 5 htg/ ml of the ionophore to the slices caused peroxidase secretion. The rate of secretion was 0.4%/min (Fig. 1).

Addition of the ionophore to slices incubated in a Ca2+-free medium did not cause any response, while subsequent addition of Ca 2+ produced the usual secre- tion (not shown).

Discussion

The present results sugest the existence of a 3-receptor in the lacrim~fl gland the stimulation of which leads to the activation of adenylate cyclase thereby causing an increase of cyclic AMP concentra- tion which in turn, mediates enzyme secretion but does not cause K ÷ release.

From the work of other investigators it is known that there are in the lacrimal gland, cholinergic and a-adrenerglc receptors, both mediating peroxidase secretion and K ÷ release [9]. The second messenger of the cholinergic and a-adrenergic receptors is known to be Ca :+ and, in accordance with this, it has also been shown that the Ca2*-ionophore A23187 causes K ÷ release from lacrimal slices [7]. To support the evidence that Ca 2÷ mediates the action of cholinerglc and a-adrenerglc receptors we have shown that iono- phore A23187 also causes peroxidase secretion in the presence of Ca 2÷.

These features of the lacrimal gland system are quite similar to those of the parotid gland system, having on the one hand a cholinergic and an a-adre- nergic receptor, the activation of which leads to the increase of cytosol Ca 2÷ concentration producing K ÷ release and enzyme secretion, and on the other hand a /3-adrenergic receptor, the stimulation of which causes activation of a membrane adenylate cyclase leading to the increase of cytosol cyclic AMP pro- ducing enzyme secretion (mainly amylase) but no K ÷ release [12]. There is however, one difference between the two systems. In the parotid gland sys- tem the main pathway for enzyme secretion is maxi- mal stimulation of the ~-receptor while the enzyme secretion caused by maximal stimulation of the

cholinergic and a-adrenergic receptors is minor [12]. In the lacrimal system however, we show that maxi- mal stimulation of the /~-adrenergic receptor causes only minor peroxidase secretion, while stimulation of the cholinergic or a-adrenergic receptors leads to a high rate of secretion (Fig. 1).

Surprisingly, dibutyryl cyclic AMP and adenosine 3',5'-cyclic phosphorothioate produce the highest rate and extent of peroxidase secretion. The reason for the difference between responses to/3-adrenergic stimulation and cyclic AMP analogs is not apparent.

In a publication by other investigators it was reported that neither isoproterenol nor dibutyryl cyclic AMP cause any peroxidase secretion [8]. The main reason for this is the way their experiment was done. They determined peroxidase activity after incu- bating the slices for 15 min in the presence of the agonists, but did not try longer periods of incubation. As can be seen from Fig. 1, the net peroxidase secre- tion (over basal)with 10 -s isoproterenol after 15 min is almost undetectable because of the low rate of secretion. In order to see a significant net secretion one should measure peroxidase activity after 40 min. In the case of dibutyryl cyclic AMP it can be seen from Fig. 3 that, even with a concentration of 2 raM, there is a lag of 10-20 min between the addition of the agonist and the appearance of peroxidase activity in the medium. Thus, measurement of peroxidase activity after only 15 min is not advisable.

Another group which measured secretion of 14C- labelled protein from lacrimal slices reported no effect of isoproterenol, although they performed a 40 min incubation with the agonist [18]. However, from the dose response curve in Fig. 1 of that work, one can see a doubling of protein discharge by 10-SM isoproterenol over the amount secreted by 10 ̀ 6 M of the agonist. It is true, however, that this discharge is very small relative to the discharge with noradrenaline, or carbachol, as is also reported by us. In Table I of the above work the authors report a secretion of about 3.7% of labelled protein with iso- proterenol compared to 2.9% of basal discharge. The true discharge by isoproterenol in their results is masked by the fact that they measure percent discharge of total glandular protein, whereas we are measuring the secretion of peroxidase which is a secretory protein. Also, leak of protein from non- secretory cells can mask the results when one mea-

sures percent of total protein. All these arguments

might be critical for the measurement of a small dis- charge of secretory enzymes as caused by isoprotere-

nol. In summary, in light of the present results, it is

conceivable that there is a/3-adrenergic receptor in the

lacrimal gland.

References

1 Herzog, V., Sies, H. and Miller, F. (1976) J. Cell Biol. 70,692-706

2 Botelho, S.Y., Goldstein, A.M. and Hisada, M. (1969) in The Exocrine Gland, pp. 227-239, University of Penn- sylvania Press, Philadelphia

3 Elsby, J.M. and Wilson, H. (1967) Br. J. Pha_rmacol. Chemother. 29, 1-7

4 Keryer, G. and Rossignol. B. (1976) Am. J. Physiol. 230, 99-104

5 Putney, J.W., Jr., Parod. R.J. and Marier, S.H. (1977) Life Sci. 20, 1905-1912

6 Parod, R.J. and Putney, J.W., Jr. (1978) J. Physiol. 281, 359-369

45

7 Parod, R.J. and Putney, J.W., Jr. (1978) J. Physiol. 281, 371-381.

8 Putney, J.W., Jr., van de Walle, C.M. and Leslie, B.A. (1978) Am. J. Physiol. 235, C188-C198

9 Iwatsuki, N. and Petersen, O.H. (1977) J. Physiol. 275, 507-520

10 Chance, B. and Maehly, A.C. (1955) in Methods in Enzy- mology (Colowick, S.P. and Kaplan, N.D., eds.), Vol. 2, p. 764, Academic Press, New York

11 Batzri, S. and Selinger, Z. (1973) J. Biol. Chem. 248, 354-360

12 Schramm, M. and Selinger, Z. (1975) J. Cyclic Nucl. Res. 1,181-192

13 Selinger, Z., Batzri, S., Eimerl, S. and Schramm, M. (1973) J. Biol. Chem. 248,369-372

14 Eckstein, F., Cassel, D., Levkovitz, H., Lowe, M. and Selinger, Z. (1979) J. Biol. Chem. 254,9829-9834

15 Eckstein, F., Eimerl, S. and Schramm, M. (1976) FEBS Lett. 64, 92-94

16 Selinger, Z., Eimerl, S. and Schramm, M. (1974) Proc. Natl. Acad. Sci. U.S.A. 71,128-131

17 Cassel, D. and Selinger, Z. (1977) Biochem. Biophys. Res. Commun. 77,868-873

18 Herman, G., Busson, S., Ovtracht, L., Maurs, C. and Ros- signol, B. (1978) Biol. Cell. 31,255-264