Discrimination between constitutive secretion and basal secretion from the regulated secretory pathway in GH3 cells

Biochi~ic~a et Biophysica AEta

ELSEVIER

Biochimica et Biophysica Acta 1313 (1996) 101-105

Discrimination between constitutive secretion and basal secretion from the regulated secretory pathway in GH3 cells Andrea Varro a.*, Joe N e m e t h ~ Chris J. Dickinson b, Tadataka Y a m a d a b G r a h a m J. Dockray a a Physiological Laboratory, Unirersi~" of Licerpool, Crown Street, P.O. Box 147, Licerpool L69 3BX, UK i, Department of Medicine, University of Michigan Medical School, Ann Arbor, MI, USA

Received 3 January 1996; revised 17 April 1996; accepted 3 May 1996

Abstract

The present experiments were undertaken to characterize basal release from vesicles of the regulated secretory pathway. In transfected GH3 cells, progastrin was released by the constitutive route, and mature, bioactive, amidated gastrin by the regulated secretory pathway. Studies using brefeldin A and bafilomycin A~ which inhibit progression through the Golgi complex suggested that both basal and stimulated release of amidated gastrin originated from mature secretory granules. Basal, but not stimulated, secretion of amidated gastrin was strongly inhibited at 22°C. Mature secretory vesicles therefore support both basal and evoked secretion although the mechanisms underlying the two processes differ in their temperature sensitivity. Keywords: Constitutive secretion; Regulated secretion; Basal secretion; Gastrin; Progastrin

1. Introduction

In neurons, exocrine and endocrine cells, the passage of secretory proteins from the trans-Golgi network (TGN) to the cell surface is mediated by either the constitutive or regulated secretory pathways [l]. The mechanisms responsible for evoked secretion from secretory vesicles of the regulated pathway have attracted considerable attention [2]. However, while it is recognized that there is also a resting or unstimulated secretion from the regulated pathway, this has received less attention even though unstimulated secretion of hormones or neurotransmitters in vivo may be of considerable physiological and pathophysiological importance [3]. In the case of insulin-secreting 13-cells of the islets of Langerhans, and the prolactin-secreting GH3 pituitary cell line, there is evidence for constitutive-like secretion of recently synthesized peptide that has passed from T G N to immature secretory granules, and from there via cartier vesicles direct to the cell surface; newly synthesized material therefore appears to be preferentially released [4-6]. This process would appear to be distinct from

* Corresponding author. Fax: +44 151 794 5315.

unstimulated secretion from mature secretory granules [7]. In the present study we have sought to distinguish between these mechanisms using treatments that arrest passage along the secretory pathway at defined points. We have made use of the observation that the mature product of post-translational processing of the precursor of the acidstimulating hormone gastrin, i.e., amidated gastrin, is generated from progastrin after delivery to secretory granules of the regulated pathway [8-10]; we selected for these studies GH3 cells, since they exhibit constitutive-like secretion from immature granules [6], and yet the total pool of regulated secretory granules turnovers relatively rapidly. The data suggest that mature secretory granules of the regulated pathway support both basal and evoked secretion, and that the former, unlike the latter, is inhibited at 22°C.

2. Materials and methods 2.1. Cells

GH3 cells were stably transfected with human gastrin cDNA using a retroviral vector as previously described [l l]. Cells were grown in Dulbecco's modified Eagle's

0167-4889/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. PH S0167-4889(96)0005 6-0

102

A. Varro et al. / Biochimica et Biophysica Acta 1313 (1996) 101-105

medium supplemented with 10% ( v / v ) horse serum and 5% ( v / v ) fetal bovine serum; they were plated out 48-72 h before experiments either on poly-L-lysine (0.1 mg" ml -~) coated multi-well plates (2 × 105. well 1), or for labelling experiments, on 9 cm poly-L-lysine coated Petri dishes (2 × l O 6 • dish-J).

I Secretion J T looo 500

150 O~ ~

°

'

"

°

"

..°."~"................ -A ° "_ .

~n

75 ~ 1D o 3

=

2.2. Pulse labelling experiments 0

o

Cells were incubated at 37°C for 30 rain in the presence of 1.2 mCi [35S]methionine (Amersham, 1000 C i / m m o l ) in 4 ml methionine-free media, then chased for up to 360 rain in media containing unlabelled Met. At intervals, cells and media were recovered by centrifuging; cells were extracted in boiling water and the extract centrifuged. Progastrin and its derivatives were concentrated on Waters C18 Sep-pak cartridges, the eluates immunoprecipitated sequentially with antibodies to progastrin, and amidated gastrin, and the products fractionated by reverse phase HPLC on a PLRP-S 8p~m column (Polymer Laboratories) with on-line liquid scintillation counting, as previously described [9,10].

2.3. Bafilomycin A I., brefeldin A and temperature-sensitive secretory respsonses Cells were incubated at 37°C or 22°C for up to 4 h in control media, or in the presence of 0.5p~M bafilomycin A~ (BAFI), or brefeldin A (BFA)(2 Ixg" m l - l ) . Media (basal secretion) was collected for estimation of progastrin and amidated gastrin as previously described [12]. The cellular content of progastrin-derived peptides was extracted in boiling water and recovered after centrifuging. Stimulated secretion was evoked by 50 mM KC1 for 15 minutes in media with appropriately reduced NaC1 concentrations; in control experiments, stimulation with KC1 was made in the absence of extracellular Ca z+, in media containing EGTA (! mM).

3. Results and discussion

3.1. Constitutive and regulated secretion of progastrin and amidated gastrins, respectively In pulse-chase experiments using [35S]Met, labelled progastrin and amidated gastrins appeared in the medium and in cell extracts with characteristically distinct kinetics (Fig. 1). The cellular content of labelled progastrin was highest at the start of the chase period and thereafter steadily declined to about 5% of maximum after 6 h. Experiments where the medium was sequentially replaced at intervals during the chase-phase, showed that the secretion of labelled progastrin reached a maximum after 60 min of chase and then declined in parallel with the decrease in cellular content. Approximately 90% of cellular

0

120

240

"~-

360

Chase time [rain}

100

Cell contend ~ , . , i-'" ""~""NHz-ge=

ID

~ 5o

0 u I1~ 0

0

.. oO.ooOO°'~ ~ °. -~'" ,

0

120

, 240

, 360

Chase time (min]

Fig. 1. Biosynthesis of progastrin and amidated (NH2-) gastrin in pulsechase experiments. Top panel, secretion into the medium of progastrin (continuous line; left axis) and amidated gastrin (dotted line, fight axis) during the chase period following pulse labelling with [35S]Met for 30 rain. Bottom panel, cell content of labelled progastrin and amidated gastrin during the chase period, expressed relative to the maximal counts in cell extracts (amidated gastrin: 498 cpm. 106 cells-t; progastrin: 8125 cpm. 106 cells-I). Samples were immunoprecipitated with antibodies to progastrin or amidated gastrin and fractionated by reversed phase HPLC. All progastrin-immunoprecipitates contained a single peak corresponding to the intact precursor; amidated gastrin immunoprecipitates contained predominantly G34 with a trace of G17. The labelling kinetics are compatible with release of progastrin from the constitutive secretory pathway and amidated gastrin from the regulated pathway; about 10% of progastrin is also converted to amidated gastrin.

progastrin subsequently appeared in the medium; taken together with the progastrin that is processed to amidated gastrin there would appear to be little or no loss of progastrin by degradation after secretion. In contrast to labelled progastrin, [35S]Met labelled amidated gastrin (mainly G34) was undetectable in cells and medium at the the start of the chase period. Labelled amidated gastrin in cell extracts was approximately 50% of maximum after 2 h chase and was maximal at 4 h (Fig. 1). The appearance of amidated gastrin in the medium followed that in cells with a maximal secretion rate at 4 h of chase. Taken together, the data suggest conversion of about 10% of labelled progastrin to amidated gastrin. However, the major fraction of labelled progastrin was secreted with a time-course compatible with release via the constitutive pathway, and more rapid than that reported for constitutive-like secretion in pancreatic 13-cells [5]. In contrast, the kinetics of appearance and secretion of amidated gastrin are compatible with formation in secretory vesicles of the regulated pathway as observed in native gastrin cells [9,10]. Functional evidence

A, Varro et al. / Biochimica et Biophysica Acta 1313 (1996) 101-105

I Pr°g"tri"

- medial

103

I Amideted gastrin - media, I

4

3O

=-

............

"¢

............... BFA

/.r

.~.;....._

=o

.......

..........~-~--~

.

........ . .................. . ....................

.

]Amidated Oeztrin -cell I

[ Progestrin - cell I

2OO

20O

~.~ ,It.

~ T

~.;~._.~

BAFL < . . . . . 2 .... BFA

. . . . ~ - - " " .....

ok . o u

.

~ ~.~--'~" :t: ......

~tlo o

T

=

;I;

I

Co.,to,

"1"

:

~"~:'=:L~-~--

T .

- -

~

N

Control

.......

.F.

C

- . - _ . ~ . ~ ........... BAFt

0

C

.

o

~.,.~

Q

I

I

i

|

i

i

0

2

4

0

2

4

Time (hour)

100 *-,

==

Time (hour)

Fig. 2. Secretion of progastrin and amidated gastrin in the presence of brefeldin A and bafilomycin A I- Cells were incubated in 0.5 ixM bafilomycin A t (BAF l) or 2 ixg. m l - t brefeldin A (BFA) for up to 4 h. Top panels, show concentrations of progastrin (left) and amidated gastrin (right) in media. Bottom panels, show corresponding data for progastrin (left) and amidated gastrin (right) in cell extracts normalized to the initial concentration (progastrin: 2.5-4.3 pmol. 10%ells-~; amidated gastrin: 11-30 fmol. 105cells - ]). Values are means + SE, n = 4; *, indicates values significantly different from control ( P < 0.05, ANOVA).

I

Evoked release in presence of BFA I

Evoked release in presence of BAFt I

20

12

-.-.

&--

KCI

I

!

.6

"6

qE .. c= 0

6

10

g I

4-,

E: m U

0 C

O

0

O 0

j

i

Durst(ion of BFA exposure {houri

i

i

i

0

Duration of BAFT exposure (houri

Fig. 3. Secretion of amidated gas(ran in response to 50 mM KC1 after incubation in either BAF I (fight) or BFA (left) for up to 4 h. Samples were incubated at 37°C in control media contai]aing BAF 1 and BAF for the times indicated; the medium was then replaced with either fresh medium containing 50 mM KC1 and either BAF t or BFA (or with control medium), and was collected 15 men later for assay. Values are means + SE, n = 4; *, indicates values significantly lower than response at time zero ( P < 0.05, ANOVA).

104

A. Varro et al. / Biochimica et Biophysica Acta 1313 (1996) 101-105

Table I Stimulated secretion of amidated gastrin, but not progastrin, in GH3 cells Peptide

Control release

Stimulated release

Stimulated release, Ca2+-free

Progastrin (nM) Amidated gastrin (pM)

0.41 + 0.09 3.33+0.48

0.49 + 0.08 10.87___0.9 *

0.38 + 0.02 2.1_+0.18

tion of progastrin as expected, and this was accompanied by a progressive increase in cellular content attributable to accumulation of material arrested in the ER (Fig. 2). Since BFA also inhibits delivery of precursor to regulated secretory granules [9], we hypothesized that it would have a delayed inhibitory effect on amidated gastrin release. We found that both basal and stimulated secretion of amidated gastrin were reduced in parallel after 2 h in BFA (Figs. 2 and 3). These effects coincided with a decrease in cellular content of amidated gastrin (Fig. 2). The parallel changes in basal and evoked release, and cell content, are consistent with the action of BFA in blocking progression of progastrin from ER to Golgi and thereafter delivery to newly forming vesicles of the regulated secretory pathway.

Mean+SE, n o 4 ; *, P < 0.05, compared with control, ANOVA. Medium was collected over a period of 15 rain; 50 mM KC1 was used to stimulate secretion. Calcium-free medium contained 1 mM EGTA.

for secretion of amidated gastrin from granules of the regulated pathway of secretion was provided by the observation that depolarizing concentrations of K + produced a four-fold increase in secretion of amidated gastrin that was inhibited in Ca2+-free medium (Table 1); in contrast, in the same samples there was no evoked secretion of progastrin. Amidated gastrin therefore provides a marker for the maturation of secretory vesicles of the regulated pathway, while progastrin marks the constitutive route.

3.3. Bafilomycin A 1 indicates role for cacuolar proton pump in long-term maintenance of basal release from regulated pathway The passage of secretory proteins through the Golgi complex involves translocation through an acidic compartment and is inhibited by blocking the vacuolar proton pump with BAF~ [6,15]. As expected, therefore, BAF~ inhibited progastrin secretion, and like the response to BFA this was accompanied by an increase in the cellular content of progastrin (Fig. 2). However, there was also inhibition of basal secretion of amidated gastrin, and of evoked secretion in the presence of 50 mM K+; both responses were delayed in onset and neither was signifi-

3.2. Brefeldin A: differential effects on basal release from regulated and constitutive pathways The fungal metabolite BFA disrupts passage from endoplasmic reticulum (ER) to Golgi complex and so rapidly inhibits constitutive and constitutive-like secretion [13,14]. In the present study, BFA promptly inhibited basal secre-

rAmi .,.d .,t,i, I

--

a.

4

37°C

20

2

""

10 "~ 22.C

0 o

"--

0

1

2

3

4

i ,~ _ _ _ 0

Time (hour) I Proleitrin

0.3

_i 11 i

°

0-

",

0

4

o~ o

I Amideted ga/trin - evoked secretion 15

2~'c

0.15-

~ 3

Time (hour)

evoked le©retion [

•

"~

';.

-

~, - - - - - - ~ , 1 2

37'c

_s

[ ] c,,.,~ol [ ] Kc=

ii

,

,.o

o

~' "

,0E__

s

i

° .0

Fig. 4. Differential temperature-sensitivity of basal and evoked secretion of amidated gastrin. Cells were incubated at either 37°C or 22°C. Top two panels, basal secretion of progastrin (left) and amidated gastrin (right) was promptly inhibited at 22°C; *, significantly different from 37°C (P < 0.05; ANOVA). Bottom panels, evoked release after preincubation for 1 h at either 22°C or 37°C, followed by replacement of medium with 50 mM KC1 at the same temperature for 15 rain (or control medium). Secretion of amidated gastrirl was similar at both temperatures (right); there was no stimulation of progastrin release at either temperature. The e×periment shown in the bottom panel gave similar results after preincubation at 37°C or 22°C for up to 4 h. Hatched bars, 50 mM KCI; open bars, control. * Significantly different to control ( P < 0.05, ANOVA).

A. Varro et aL / Biochimica et Biophysica Acta 1313 (1996) 101-105

cant until 4 h incubation with BAFI (Figs. 2 and 3). As with the response to BFA, the decline in both basal and stimulated release of amidated gastrin by BAFI were coincident with a decrease in its cellular content compatible with an inhibition of precursor delivery to secretory vesicles of the regulated pathway (Fig. 2). 3.4. Differential temperature sensitivity of basal and

105

these release processes, albeit by mechanisms differing in temperature sensitivity.

Acknowledgements Supported by the Medical Research Council and The Wellcome Trust.

euoked, secretion from regulated pathway Exit of secretory proteins from TGN to both constitutive and regulated secretor3 pathways is inhibited at 22°C; the temperature-sensitive step is distal to the site of action of BFA [9]. In keeping with this, incubation at 22°C promptly inhibited progastrin secretion. More surprisingly, however, incubation at 22 ° also immediately blocked basal secretion of amidated gastrin (Fig. 4). A temperature-sensitive step in the cycle of priming of secretory vesicles for exocytosis in adrenal chromaffin cells has previously been described [16]. The prompt effect of incubation at 22°C on basal release of amidated gastrin could suggest that there is only a small pool of secretory vesicles competent for basal secretion distal to this step in GH3 cells. This is unlikely however, since the stimulated secretory response is maintained even after several hours at 22°C (Fig. 4). An alternative possibility is that changes in membrane fluidity have more profound effects, on basal than stimulated secretion. Either way, the data s]how that the mechanisms which underlie basal and evoked release from the regulated secretory pathway can be distinguished by their temperature sensitivity. Taken together with the evidence presented above, it would appear that basal and evoked release from vesicles of the regulated pathway in GH3 cells derives from material delivered from TGN to vesicles at least 2 - 4 h earlier, and that a common pool of vesicles supports

References [1] Burgess, T.L. and Kelly, R.B. (1987) Ann. Rev. Cell. Biol. 3, 243-293. [2] Sudhof, T.C. (1995) Nature 375, 645-653. [3] Matsuuchi, L. and Kelly, R.B. (1991) J. Cell. Biochem. 112, 843852. [4] Arvan, P., Kuliawat, R., Prabakaran, D., Zavacki, A.M., Elahi, D. and Pilkey, D. (1991) J. Biol. Chem. 266, 14171-14174. [5] Kuliawat, R, and Aryan, P. (1992) J. Cell Biol. 118, 521-529. [6] Henomatsu, N., Yoshimori, T., Yamamoto, A., Moriyama, Y. and Tashiro, Y. (1993) Eur. J. Cell Biol. 62, 127-139. [7] Noel, G. and Mains, R.E. (1991) Mol. Endocrinol. 5, 787-794. [8] Rahier, J., Pauwels, S. and Dockray, G.J. (1987) Gastroenterology 92, 1146-1152. [9] Varro, A., Henry, J., Vaillant, C. and Dockray, G.J. (1994) J. Biol. Chem. 269, 20764-20770. [10] Van-o, A., Voronina, S. and Dockray, G.J. (1995) J. Clin. Invest. 95, 1642 - 1649. [11] Marino, L.R., Takeuchi, T., Dickinson, C.J. and Yamada, T. (1991) J. Biol. Chem. 266, 6133-6136. [12] Nemeth, J., Taylor, B., Pauwels, S., Varro, A. and Dockray, G.J. (1993) Gut 34, 90-95. [13] Lippincott-Schwartz, J., Donaldson, J.G., Schweizer, A., et al. (1990) Cell 60, 821-836. [ 14] Orci, L., Tagaya, M., Amherdt, M., et al. ( 1991 ) Cell 64, 1183-1195. [15] Yilla, M., Tan, A., Ito, K., Miwa, K. and Ploegh, H.L. (1993) J. Biol. Chem. 268, 19092-19100. [16] Bittner, M.A. and Holz, R.W. (1992) J. Biol. Chem. 267, 1622616229.