- Email: [email protected]
Significance of chloride channel activation in the gamma-aminobutyric acid induced growth hormone secretion in the neonatal rat pituitary
Life Sciences, Vol. Printed in the USA
52, pp.
1733-1739
Pergamon
Press
SIGNIFICANCE OF CHLORIDE CHANNEL ACTIVATION IN THE GAMMAAMINOBUTYRIC ACID INDUCED GROWTH HORMONE SECRETION IN THE NEONATAL RAT PITUITARY Zsuzsanna Acs, L. Zsom, Zsuzsanna Mergl and G.B. Makara Institute of Experimental Medicine Hungarian Academy of Sciences, Budapest, Hungary (Received
in final
form March
16,
1993)
Summary Growth hormone (GH) secretion of the neonatal pituitary is stimulated by raminobutyric acid (GABA) (1,2). Since in most cases GABA is known to act by increasing postsynaptic membrane permeability to chloride ions we tested the importance of chloride channel activation in the GH stimulatory effect of GABA in the neonatal pituitary. In the absence of chloride in the superfusion medium GABA was without effect on GH secretion of the neonatal pituitaries and its effect was attenuated by chloride channel inhibitors. The effect of growth hormone releasing hormone (GHRH) on GH secretion was attenuated in the chloride-free media, but it was not affected by simultaneous administration of chloride channel blockers. The present study indicates that GH stimulatory effect of GABA in the neonatal pituitaries might involve chloride channel activation probably resulting in secondary activation of calcium channels.
In the neonatal rat regulation of growth hormone (GH) secretion differs from that seen in the adult. In the first two weeks of life stimulatory influences seem to predominate and the known inhibitory neurotransmitter r-aminobutyric acid (GABA) also stimulates GH secretion (1,2). The GABA receptor involved in this GABA effect has not been characterized, but muscimol proved to be a good agonist and the GABAA receptor antagonists, bicuculline and SR 95103, are potent inhibitors (2). On the other hand the known chloride ionophore antagonist picrotoxin (3) effectively antagonizes the GH stimulatory effect of GABA in the neonatal pituitary (2). In the central nervous system GABAA receptors are coupled to chloride ion channels, opened by GABA to produce inhibition in the postsynaptic target cell. In some neonatal neurones, GABA depolarizes neuronal membranes, this depolarization is chloridedependent, and may be due to a modified chloride gradient that results from the reversed operation of the membrane chloride pump (for review see 4). In the present work we attempted to reveal if chloride channel activation was involved in the GH stimulatory effect of GABA in the neonatal rat pituitary by using either chloride-
Corresponding author: Zsuzsanna Acs, Institute of Experimental Medicine Hungarian Academy of Sciences H-1450 Budapest P.O.B. 67 Hungary
Copyright
0024-3205/93 $6.00 + .00 © 1993 Pergamon Press Ltd All rights reserved.
1734
GABA-induced
GH S e c r e t i o n
and C h l o r i d e
Channels
Vol.
52, No. 21,
1993
free media or blockers of the chloride-channel. Since in previous work we demonstrated that in the neonatal pituitary both calcium-flux (5) and intracellular calcium concentration (6) were increased by GABA or muscimol, we tested if there was an interplay between the activation of the chloride and of the calcium channels. Materials and methods Two-day-old Wistar rats of either sexes from our inbred colony were used in all experiments. The morning on which the pups were found in the cage of the mother was considered the first postnatal day. The neonates were left undisturbed with their mothers until the experiment. Solutions 1.) Krebs Ringer buffer (KRB, pH 7.4) 118.46 mM NaC1, 4.75 mM KC1, 25 mM NaHCO 3, 1.19 mM MgSO4, 1.19 mM KH2PO 4, 2.54 mM CaC12, 11.1 mM glucose, 0.02 mM bacitracin (Serva), 0.1 mM ascorbic acid, and 0.1% bovine serum albumin (BSA, Fraction V. Calbiochem) 2.) Chloride-free Krebs Ringer buffer NaC1 was omitted and substituted by either Na2SO 4 or by 4-chlor-N-(2-furylmethyl)-5sulphamoilantranyl acid sodium (Na-isethionate), CaCl2 and KCI were replaced by CaSO 4 and K2SO4, respectively. The concentration of other ions and the osmolarity of the buffer solution was unchanged. Drugs: T-butylbicyclophosphoro-thionate (TBPS; Research Biochemicals, Inc. Natick, MA) was dissolved in dimethylsulphoxid (DMSO) and further diluted with KRB buffer (the final DMSO concentration was 1 percent). GABA was from Serva (Heidelberg Germany), human growth hormone releasing hormone 1-44 (hGHRH) and Nifedipine were from Sigma (St Louis, Mo). Experimental protocol: 1.) Growth hormone secretion studies Five whole pituitaries from 2 day old rats were placed into the superfusion chamber and superfused with Krebs-Ringer buffer solution as described earlier (2). The chamber volume was 0.1 ml, the flow rate 0.1 ml/min, and 10 rain fractions were collected. After an initial equilibration period of 2 h, fractions were collected for 30 min to obtain GH secretory baseline value. Pituitaries were exposed to the test substances for 10 min. Introduction of chloride-free medium was performed either by abrupt change of the Krebs-Ringer to chloride-free medium or by stepwise reduction of the chloride content of the medium exposing the pituitaries for ten minutes to each chloride concentration reaching zero chloride concentration after 100 min (Fig.l). Test substances were administered after at least 60 min in chloride free media or after 30 min exposure to the chloride and calcium channel antagonists. Fractions were stored at -20°C until assayed for GH. Immunoreactive GH content of the fractions was determined by specific radioimmunoassay (7). The antiserum (aGH-S-5) and the hormones (rGH-RP-2 and rGH-I-6) were kindly provided by NIDDK. 2.) Calcium effiux studies As described earlier (5) rats were decapitated, the posterior pituitary lobes separated and discarded. Ten anterior pituitary lobes from the 2 day old rats were incubated at 37 ° C in 0.5
Vol.
52, No.
21,
1993
GABA-induced
GH S e c r e t i o n
and C h l o r i d e C h a n n e l s
1735
ml of calcium and BSA free KRB buffer containing 370 kBq 45Ca2+/ml (Institute of Isotopes, Budapest, Hungary; specific activity 110-130 GBq/mmol) for one hour under constant shaking (60 rpm) in a 95% 0 2, 5% CO 2 atmosphere. After rinsing the pituitaries 3 times with calcium free KRB they were placed in the superfusion chamber and superfused at flow rate 0.5 ml/min with either Krebs Ringer buffer (0.85 mM CaCI2) or isethionate substituted chloride-free buffer. Radioactivity of the 5 mL fractions was measured on an LKB RackBeta 1211 type counter.
16
'160
7
,ooi
IO
E
10
!
10
160
........i.-!..i..i..:
100
ii. _o 6O
-r
+
O
0l P 0
i
60
120
E
CS
r-iGAaAi
m c
180
240
i..~ [--IGABA 0
60
mlnutee
i
_*1
:1
q-
NIFEDIPINE
I
'O
o dg O
E
0
60
100
180
240
minute•
1l°oo
.......................
120
160
minutes
FIG. 1. Effect of chloride withdrawal on basal and on GABA (100 #M for 10 min) stimulated GH secretion of neonatal pituitaries. Upper left panel: rapid replacement with isethionate; upper right panel: stepwise isethionate replacement; lower left panel: rapid replacement in the presence of 1 IzM Nifedipine. Values are geometric means of four separate experiments with S.E.M.; * * : p<0.01 relative to pre-drug value.
The data were subjected to analysis of variance followed by Dunnett's multiple comparisons (8) the last predrug value serving as control. The increment in GH secretion or in calcium efflux was expressed as percent of increase relative to basal. The integrated response after different treatments was compared by analysis of variance followed by Dunn's paired comparisons (9). Number of experiments denote the number of exposed sets of five (GH secretion) or ten (calcium-effiux) pituitary glands. Results After a two-hour equilibration period baseline GH secretion and responsiveness of the pituitaries to GABA and G H R H was stable and reproducible over the time period
1736
GABA-induced
GH S e c r e t i o n and C h l o r i d e C h a n n e l s
Vol.
52, No. 21,
1993
studied. Rapid replacement of the superfusion buffer with either sodium sulphate or sodium isethionate substituted chloride-free media resulted in a transient rise in GH secretion; the increase seen after sodium isethionate substitution (Fig. 1) was smaller (201.8_+8.14%, n=4) than that after the sulphate substitution (440.0+40.1%, n=4; not shown). The increase in GH secretion after chloride replacement was attenuated in the presence of 1 ~M Nifedipine (Fig. 1). If, however chloride was replaced stepwise by isethionate, pituitary GH secretion was unchanged (Fig. 1). TABLE I. Integrated GH Secretory Response of Neonatal Pituitaries in Chloride-Free Media (percent of increase relative to basal secretion over 30 min)
Krebs Ringer
chloride-flee buffer isethionate introduction rapid stepwise
GABA 100 ~M
249.7± 14.0 (4) 219.1±5.4 (2)
95.4± 1.5"* (4)
GHRH 1 nM
160.9±21.3 (5) 243.6±25.6 (6)
101.2±2.9"* (4)
106.6±10.0"* (2)
60 min after replenishment with Krebs Ringer buffer 155.1 ±6.6"* (4) 153.0±9.5 (4)
149.9±12.3"* (6)
Values are means_+S.E.M, with number of separate experiments in parentheses. Test substances were administered after 60 or 30 min in chloride-free buffer after rapid or stepwise introduction of chloride-free buffer, respectively. **: p < 0.01 compared to increment in Krebs Ringer buffer
Pituitaries superfused with chloride-free buffer did not respond to GABA with increased GH secretion (Fig. 1). One hour after returning to the normal chloride KRB GABA again stimulated GH secretion, but its effect was smaller than in the control pituitaries (Table 1). The stimulatory effect of G H R H was abolished after sudden chloride withdrawal and was diminished when chloride was substituted stepwise with sodium isethionate (Table 1). GH secretion elicited by high potassium (25 mM) was comparable in the presence and in the absence of chloride (not shown). Equimolar concentration of the chloride channel blocker TBPS diminished the increase in GH secretion elicited by 10 tzM GABA (Fig. 2). TBPS had no effect on the response to G H R H (not shown). The potent GABA agonist muscimol stimulated calcium effiux from preloaded neonatal pituitaries only when they were superfused with chloride containing, but not with chloride-free (sodium isethionate substituted) KRB (Fig. 3).
Vol.
52, No. 21,
1993
GABA-induced
GH S e c r e t i o n
and C h l o r i d e
Channels
1737
FIG. 2. Effect of 10 # M GABA for 10 min on the GH secretion of neonatal pituitaries in the presence of 10 /~M TBPS. Values are geometric means of four separate experiments with S.E.M. *: p<0.05; ** :p<0.01; n.s.: not significant relative to pre-drug value.
TBPS
10 mm
_5 s
TT
o m
T ~ []
0
QABA
40
80
[] GABA 120
180
200
minutes
Discussion The marked increase in GH release seen when chloride was instantaneously replaced by impermeant anions (sulphate or isethionate) resembles the effect of these substances on the efflux of transmitters from superfused brain slices (10) or synaptosomas (11). The reduction of extracellular chloride by changing intracellular/extracellular chloride ratio could cause depolarization resulting in increased GH secretion. This increase was attenuated by the L-type calcium channel blocker Nifedipine (Fig. 1), indicating the presence and importance of chloride-gated voltage sensitive calcium channels in the GH secretory mechanism of the neonatal rat somatotrophs.
20"
|
20'
16'
is.
10'
t'
E IU C IM II OL
i C
o
L.~_
8'
8
MUIICIMOL 0
20
40
O0
O0
20
minutes
40
80
80
minutes
FIG. 3. Fractional calcium (45Ca2+) release from preloaded neonatal pituitaries. Effect of 10 min exposure to 10/~M muscimol in Krebs Ringer buffer (left) or in isethionate substituted chloride-free buffer (right). 45Ca2+ concentration expressed in the percent of the total 45Ca2+ in the preloaded pituitaries. Values are means with S.E.M of four separate experiments. * p<0.05 relative to pre-drug value.
Intracellular mechanisms mediating GABA effect in the neonatal pituitary somatotrophs are poorly understood. Pharmacological characteristics of the stimulatory
1738
GABA-induced
GH S e c r e t i o n
and C h l o r i d e C h a n n e l s
Vol.
52, No. 21,
1993
effect of GABA on G H secretion (2) argue for the involvement of GABA A receptors. The effectiveness of picrotoxin to inhibit GABA induced GH stimulation indicated that GH stimulatory effect of GABA in the neonatal pituitaries might involve chloride channel activation. The present results that GH stimulatory effect of GABA was abolished in the absence of chloride from the superfusion medium provided further evidence for the importance of chloride channel activation by GABA. On the other hand the chloride channel blocker TBPS was several magnitude less effective in our system (Fig. 2) than described for GABA evoked luteinizing hormone release in adult pituitaries (12). These findings indicate that in the neonatal pituitary either the GABA receptor and/or the intracellular mediator system function differs from that in the adult. The presence of different types or states of the GABA A receptor in the adult pituitary and its involvement in GABAergic regulation of pituitary hormone secretion has been well documented (12,13,14,15,16,17) although on some cell types the presence of GABA B receptor has been also demonstrated (16). On the other hand we are not aware of any direct evidence on the presence and function of GABA receptors in the neonatal rat pituitary. Our present results together with previous data (2) support the possibility that in the neonatal pituitary GABA acts at GABA A receptor sites. Neuronal GABAA-receptors are closely linked to the chloride channels of the membrane (3,18). GABA is predominantly inhibitory neurotransmitter in the central nervous system, but recent evidence indicate that in some cells GABA mediates bicuculline-sensitive depolarizing potentials which are mediated by GABA receptor activated increase in chloride conductance (19,20,21). In the adult pituitary Virmani et al. (12) demonstrated the involvement of the chloride and calcium channel activation in the luteinizing hormone stimulatory effect of GABA. Results on involvement of calcium channel activation by G H R H in the GH secretion of the neonatal pituitaries are controversial. G H R H failed to stimulate G H secretion in the absence of extracellular calcium (unpublished), but unlike in the adult (22) G H R H in concentrations stimulating GH secretion failed to affect calcium-efflux (5). Intracellular calcium concentration was not stimulated by G H R H in the neonatal pituitary cell suspension (unpublished) but its effect on somatotrophs might have been masked by the presence of other cell types. Our interpretation of these data is that calcium channel activation is involved in the GH releasing effect of G H R H in the neonatal pituitary, but its role is less pronounced then in the adult. Chloride channel antagonists failed to inhibit GH stimulation by G H R H (2, present paper), thus the diminished response to G H R H in sodium isethionate medium might be due to the impairment of calcium channel function in the absence of chloride. The present results indicate that the GABA-positive effect on GH secretion in the neonatal pituitary might occur through GABA receptors associated with chloride channels. These chloride channels seem to be linked to calcium channels, since in the absence of chloride muscimol failed to stimulate calcium efflux. A primary event in the releasing action of GABA could be an increase in chloride conductance bought about by the activation of GABA receptor and if the resting chloride potencial is less negative than resting membrane potencial, the resulting movement of chloride out of the cell will cause depolarization. On the other hand, depolarization could cause the direct activation of voltage-sensitive calcium channels. Thus both calcium and chloride channel activation might be involved in the GHreleasing effect of GABA in the neonatal pituitary.
Vol.
52, No.
21,
1993
GABA-induced
GH S e c r e t i o n
and C h l o r i d e C h a n n e l s
1739
Acknowledgements We thank Ms J. R~tcz for the expert technical help and the Radioimmunoassay Laboratory of our Institute for the GH measurements. We are grateful to the NIDDK for the generous gift of the rGH RIA material. This work was supported by the grant OTKA 2922. References 1. ZS. ACS, G.B. MAKARA and E. STARK, Life Sciences 34 1505-1511 (1984). 2. ZS. ,~CS, B. SZABO, G. KAPOCS and G.B.MAKARA, Endocrinology 120 1790-1798 (1987). 3. P.R. ANDREWS and G.A.R. JOHNSTON, Biochemical Pharmacology 28 2697-2702 (1979). 4. E. CHERUBINI, J.L. GAIARSA and Y. BEN-ARI, TINS 14 515-519 (1991). 5. ZS..,%CS, L. ZSOM and G.B. MAKARA, Life Sciences 50 273-279 (1992). 6. GY. HORVA.TH, ZS. ,&CS, ZS. MERGL, I. NAGY and G.B. MAKARA, Neuroendocrinology (in press). 7. C.A. BIRGE, G.T. PEAKE, I.K. MARIZ and W.H. DAUGHADAY, Endocrinology 81 195-204 (1967). 8. C.W. DUNNETI', J. Am. Stat. Assoc. 50 1096-1121 (1955). 9. O.J. DUNN, J. Am. Stat. Assoc. 56 52-61 (1961). 10. J.D. TURNER, R.J. BOAKES, J.A. HARDY and M.A. VIRMANI, J. Neurochemistry 48 1060-1068 (1987). 11. J.A. HARDY, R.J. BOAKES, D.J.E. THOMAS, A.M. KIDD, J.A. EDWARDSON, M. VIRMANI, J. TURNER and P.R. DODD, J. Neurochemistry 42 875-877 (1984). 12. M.A. VIRMANI, S.S. STOJILKOVIC and K.J. CAT1r, Endocrinology 126 2499-2505 (1990). 13. L. GRANDISON and A. GUIDOqTI, Endocrinology 105 754-759 (1979). 14. S.FISZER DE PLAZAS, D. BECU, A. MITRIDATE DE NOVARA, and C. LIBERTUN, Endocrinology 111 1974-1978 (1982). 15. F.V. DEFEUDIS, Neurochem, Int. 6 1-16 (1984). 16. R.A.ANDERSON and R. MITCHELL, J. Endocrinology 108 1-8 (1986). 17. L.T. ARANCIBIA, J.P. ROUSSEL and H. ASTIER, Endocrinology 121 980-986 (1987). 18. R.D. SCHWARTZ, Biochemical Pharmacology 37 3369-3375 (1988). 19. U. MISGELD, A. WAGNER and T. OHNO, Exp. Brain Res. 45 108-124 (1982). 20. P. CALABRESI, M. BENEDETFI, N.B. MERCURI, and G. BERNARDI, Neuroscience 30 663-670 (1989). 21. P. CALABRESI, N.B. MERCURI. and G. BERNARDI, J. Neurophysiol. 63 663-675 (1990). 22. L. CUTTLER, S.R. GAUM, B.A. COLLINS and R.J. MILLER, Endocrinology 130 945953 (1992).
52, pp.
1733-1739
Pergamon
Press
SIGNIFICANCE OF CHLORIDE CHANNEL ACTIVATION IN THE GAMMAAMINOBUTYRIC ACID INDUCED GROWTH HORMONE SECRETION IN THE NEONATAL RAT PITUITARY Zsuzsanna Acs, L. Zsom, Zsuzsanna Mergl and G.B. Makara Institute of Experimental Medicine Hungarian Academy of Sciences, Budapest, Hungary (Received
in final
form March
16,
1993)
Summary Growth hormone (GH) secretion of the neonatal pituitary is stimulated by raminobutyric acid (GABA) (1,2). Since in most cases GABA is known to act by increasing postsynaptic membrane permeability to chloride ions we tested the importance of chloride channel activation in the GH stimulatory effect of GABA in the neonatal pituitary. In the absence of chloride in the superfusion medium GABA was without effect on GH secretion of the neonatal pituitaries and its effect was attenuated by chloride channel inhibitors. The effect of growth hormone releasing hormone (GHRH) on GH secretion was attenuated in the chloride-free media, but it was not affected by simultaneous administration of chloride channel blockers. The present study indicates that GH stimulatory effect of GABA in the neonatal pituitaries might involve chloride channel activation probably resulting in secondary activation of calcium channels.
In the neonatal rat regulation of growth hormone (GH) secretion differs from that seen in the adult. In the first two weeks of life stimulatory influences seem to predominate and the known inhibitory neurotransmitter r-aminobutyric acid (GABA) also stimulates GH secretion (1,2). The GABA receptor involved in this GABA effect has not been characterized, but muscimol proved to be a good agonist and the GABAA receptor antagonists, bicuculline and SR 95103, are potent inhibitors (2). On the other hand the known chloride ionophore antagonist picrotoxin (3) effectively antagonizes the GH stimulatory effect of GABA in the neonatal pituitary (2). In the central nervous system GABAA receptors are coupled to chloride ion channels, opened by GABA to produce inhibition in the postsynaptic target cell. In some neonatal neurones, GABA depolarizes neuronal membranes, this depolarization is chloridedependent, and may be due to a modified chloride gradient that results from the reversed operation of the membrane chloride pump (for review see 4). In the present work we attempted to reveal if chloride channel activation was involved in the GH stimulatory effect of GABA in the neonatal rat pituitary by using either chloride-
Corresponding author: Zsuzsanna Acs, Institute of Experimental Medicine Hungarian Academy of Sciences H-1450 Budapest P.O.B. 67 Hungary
Copyright
0024-3205/93 $6.00 + .00 © 1993 Pergamon Press Ltd All rights reserved.
1734
GABA-induced
GH S e c r e t i o n
and C h l o r i d e
Channels
Vol.
52, No. 21,
1993
free media or blockers of the chloride-channel. Since in previous work we demonstrated that in the neonatal pituitary both calcium-flux (5) and intracellular calcium concentration (6) were increased by GABA or muscimol, we tested if there was an interplay between the activation of the chloride and of the calcium channels. Materials and methods Two-day-old Wistar rats of either sexes from our inbred colony were used in all experiments. The morning on which the pups were found in the cage of the mother was considered the first postnatal day. The neonates were left undisturbed with their mothers until the experiment. Solutions 1.) Krebs Ringer buffer (KRB, pH 7.4) 118.46 mM NaC1, 4.75 mM KC1, 25 mM NaHCO 3, 1.19 mM MgSO4, 1.19 mM KH2PO 4, 2.54 mM CaC12, 11.1 mM glucose, 0.02 mM bacitracin (Serva), 0.1 mM ascorbic acid, and 0.1% bovine serum albumin (BSA, Fraction V. Calbiochem) 2.) Chloride-free Krebs Ringer buffer NaC1 was omitted and substituted by either Na2SO 4 or by 4-chlor-N-(2-furylmethyl)-5sulphamoilantranyl acid sodium (Na-isethionate), CaCl2 and KCI were replaced by CaSO 4 and K2SO4, respectively. The concentration of other ions and the osmolarity of the buffer solution was unchanged. Drugs: T-butylbicyclophosphoro-thionate (TBPS; Research Biochemicals, Inc. Natick, MA) was dissolved in dimethylsulphoxid (DMSO) and further diluted with KRB buffer (the final DMSO concentration was 1 percent). GABA was from Serva (Heidelberg Germany), human growth hormone releasing hormone 1-44 (hGHRH) and Nifedipine were from Sigma (St Louis, Mo). Experimental protocol: 1.) Growth hormone secretion studies Five whole pituitaries from 2 day old rats were placed into the superfusion chamber and superfused with Krebs-Ringer buffer solution as described earlier (2). The chamber volume was 0.1 ml, the flow rate 0.1 ml/min, and 10 rain fractions were collected. After an initial equilibration period of 2 h, fractions were collected for 30 min to obtain GH secretory baseline value. Pituitaries were exposed to the test substances for 10 min. Introduction of chloride-free medium was performed either by abrupt change of the Krebs-Ringer to chloride-free medium or by stepwise reduction of the chloride content of the medium exposing the pituitaries for ten minutes to each chloride concentration reaching zero chloride concentration after 100 min (Fig.l). Test substances were administered after at least 60 min in chloride free media or after 30 min exposure to the chloride and calcium channel antagonists. Fractions were stored at -20°C until assayed for GH. Immunoreactive GH content of the fractions was determined by specific radioimmunoassay (7). The antiserum (aGH-S-5) and the hormones (rGH-RP-2 and rGH-I-6) were kindly provided by NIDDK. 2.) Calcium effiux studies As described earlier (5) rats were decapitated, the posterior pituitary lobes separated and discarded. Ten anterior pituitary lobes from the 2 day old rats were incubated at 37 ° C in 0.5
Vol.
52, No.
21,
1993
GABA-induced
GH S e c r e t i o n
and C h l o r i d e C h a n n e l s
1735
ml of calcium and BSA free KRB buffer containing 370 kBq 45Ca2+/ml (Institute of Isotopes, Budapest, Hungary; specific activity 110-130 GBq/mmol) for one hour under constant shaking (60 rpm) in a 95% 0 2, 5% CO 2 atmosphere. After rinsing the pituitaries 3 times with calcium free KRB they were placed in the superfusion chamber and superfused at flow rate 0.5 ml/min with either Krebs Ringer buffer (0.85 mM CaCI2) or isethionate substituted chloride-free buffer. Radioactivity of the 5 mL fractions was measured on an LKB RackBeta 1211 type counter.
16
'160
7
,ooi
IO
E
10
!
10
160
........i.-!..i..i..:
100
ii. _o 6O
-r
+
O
0l P 0
i
60
120
E
CS
r-iGAaAi
m c
180
240
i..~ [--IGABA 0
60
mlnutee
i
_*1
:1
q-
NIFEDIPINE
I
'O
o dg O
E
0
60
100
180
240
minute•
1l°oo
.......................
120
160
minutes
FIG. 1. Effect of chloride withdrawal on basal and on GABA (100 #M for 10 min) stimulated GH secretion of neonatal pituitaries. Upper left panel: rapid replacement with isethionate; upper right panel: stepwise isethionate replacement; lower left panel: rapid replacement in the presence of 1 IzM Nifedipine. Values are geometric means of four separate experiments with S.E.M.; * * : p<0.01 relative to pre-drug value.
The data were subjected to analysis of variance followed by Dunnett's multiple comparisons (8) the last predrug value serving as control. The increment in GH secretion or in calcium efflux was expressed as percent of increase relative to basal. The integrated response after different treatments was compared by analysis of variance followed by Dunn's paired comparisons (9). Number of experiments denote the number of exposed sets of five (GH secretion) or ten (calcium-effiux) pituitary glands. Results After a two-hour equilibration period baseline GH secretion and responsiveness of the pituitaries to GABA and G H R H was stable and reproducible over the time period
1736
GABA-induced
GH S e c r e t i o n and C h l o r i d e C h a n n e l s
Vol.
52, No. 21,
1993
studied. Rapid replacement of the superfusion buffer with either sodium sulphate or sodium isethionate substituted chloride-free media resulted in a transient rise in GH secretion; the increase seen after sodium isethionate substitution (Fig. 1) was smaller (201.8_+8.14%, n=4) than that after the sulphate substitution (440.0+40.1%, n=4; not shown). The increase in GH secretion after chloride replacement was attenuated in the presence of 1 ~M Nifedipine (Fig. 1). If, however chloride was replaced stepwise by isethionate, pituitary GH secretion was unchanged (Fig. 1). TABLE I. Integrated GH Secretory Response of Neonatal Pituitaries in Chloride-Free Media (percent of increase relative to basal secretion over 30 min)
Krebs Ringer
chloride-flee buffer isethionate introduction rapid stepwise
GABA 100 ~M
249.7± 14.0 (4) 219.1±5.4 (2)
95.4± 1.5"* (4)
GHRH 1 nM
160.9±21.3 (5) 243.6±25.6 (6)
101.2±2.9"* (4)
106.6±10.0"* (2)
60 min after replenishment with Krebs Ringer buffer 155.1 ±6.6"* (4) 153.0±9.5 (4)
149.9±12.3"* (6)
Values are means_+S.E.M, with number of separate experiments in parentheses. Test substances were administered after 60 or 30 min in chloride-free buffer after rapid or stepwise introduction of chloride-free buffer, respectively. **: p < 0.01 compared to increment in Krebs Ringer buffer
Pituitaries superfused with chloride-free buffer did not respond to GABA with increased GH secretion (Fig. 1). One hour after returning to the normal chloride KRB GABA again stimulated GH secretion, but its effect was smaller than in the control pituitaries (Table 1). The stimulatory effect of G H R H was abolished after sudden chloride withdrawal and was diminished when chloride was substituted stepwise with sodium isethionate (Table 1). GH secretion elicited by high potassium (25 mM) was comparable in the presence and in the absence of chloride (not shown). Equimolar concentration of the chloride channel blocker TBPS diminished the increase in GH secretion elicited by 10 tzM GABA (Fig. 2). TBPS had no effect on the response to G H R H (not shown). The potent GABA agonist muscimol stimulated calcium effiux from preloaded neonatal pituitaries only when they were superfused with chloride containing, but not with chloride-free (sodium isethionate substituted) KRB (Fig. 3).
Vol.
52, No. 21,
1993
GABA-induced
GH S e c r e t i o n
and C h l o r i d e
Channels
1737
FIG. 2. Effect of 10 # M GABA for 10 min on the GH secretion of neonatal pituitaries in the presence of 10 /~M TBPS. Values are geometric means of four separate experiments with S.E.M. *: p<0.05; ** :p<0.01; n.s.: not significant relative to pre-drug value.
TBPS
10 mm
_5 s
TT
o m
T ~ []
0
QABA
40
80
[] GABA 120
180
200
minutes
Discussion The marked increase in GH release seen when chloride was instantaneously replaced by impermeant anions (sulphate or isethionate) resembles the effect of these substances on the efflux of transmitters from superfused brain slices (10) or synaptosomas (11). The reduction of extracellular chloride by changing intracellular/extracellular chloride ratio could cause depolarization resulting in increased GH secretion. This increase was attenuated by the L-type calcium channel blocker Nifedipine (Fig. 1), indicating the presence and importance of chloride-gated voltage sensitive calcium channels in the GH secretory mechanism of the neonatal rat somatotrophs.
20"
|
20'
16'
is.
10'
t'
E IU C IM II OL
i C
o
L.~_
8'
8
MUIICIMOL 0
20
40
O0
O0
20
minutes
40
80
80
minutes
FIG. 3. Fractional calcium (45Ca2+) release from preloaded neonatal pituitaries. Effect of 10 min exposure to 10/~M muscimol in Krebs Ringer buffer (left) or in isethionate substituted chloride-free buffer (right). 45Ca2+ concentration expressed in the percent of the total 45Ca2+ in the preloaded pituitaries. Values are means with S.E.M of four separate experiments. * p<0.05 relative to pre-drug value.
Intracellular mechanisms mediating GABA effect in the neonatal pituitary somatotrophs are poorly understood. Pharmacological characteristics of the stimulatory
1738
GABA-induced
GH S e c r e t i o n
and C h l o r i d e C h a n n e l s
Vol.
52, No. 21,
1993
effect of GABA on G H secretion (2) argue for the involvement of GABA A receptors. The effectiveness of picrotoxin to inhibit GABA induced GH stimulation indicated that GH stimulatory effect of GABA in the neonatal pituitaries might involve chloride channel activation. The present results that GH stimulatory effect of GABA was abolished in the absence of chloride from the superfusion medium provided further evidence for the importance of chloride channel activation by GABA. On the other hand the chloride channel blocker TBPS was several magnitude less effective in our system (Fig. 2) than described for GABA evoked luteinizing hormone release in adult pituitaries (12). These findings indicate that in the neonatal pituitary either the GABA receptor and/or the intracellular mediator system function differs from that in the adult. The presence of different types or states of the GABA A receptor in the adult pituitary and its involvement in GABAergic regulation of pituitary hormone secretion has been well documented (12,13,14,15,16,17) although on some cell types the presence of GABA B receptor has been also demonstrated (16). On the other hand we are not aware of any direct evidence on the presence and function of GABA receptors in the neonatal rat pituitary. Our present results together with previous data (2) support the possibility that in the neonatal pituitary GABA acts at GABA A receptor sites. Neuronal GABAA-receptors are closely linked to the chloride channels of the membrane (3,18). GABA is predominantly inhibitory neurotransmitter in the central nervous system, but recent evidence indicate that in some cells GABA mediates bicuculline-sensitive depolarizing potentials which are mediated by GABA receptor activated increase in chloride conductance (19,20,21). In the adult pituitary Virmani et al. (12) demonstrated the involvement of the chloride and calcium channel activation in the luteinizing hormone stimulatory effect of GABA. Results on involvement of calcium channel activation by G H R H in the GH secretion of the neonatal pituitaries are controversial. G H R H failed to stimulate G H secretion in the absence of extracellular calcium (unpublished), but unlike in the adult (22) G H R H in concentrations stimulating GH secretion failed to affect calcium-efflux (5). Intracellular calcium concentration was not stimulated by G H R H in the neonatal pituitary cell suspension (unpublished) but its effect on somatotrophs might have been masked by the presence of other cell types. Our interpretation of these data is that calcium channel activation is involved in the GH releasing effect of G H R H in the neonatal pituitary, but its role is less pronounced then in the adult. Chloride channel antagonists failed to inhibit GH stimulation by G H R H (2, present paper), thus the diminished response to G H R H in sodium isethionate medium might be due to the impairment of calcium channel function in the absence of chloride. The present results indicate that the GABA-positive effect on GH secretion in the neonatal pituitary might occur through GABA receptors associated with chloride channels. These chloride channels seem to be linked to calcium channels, since in the absence of chloride muscimol failed to stimulate calcium efflux. A primary event in the releasing action of GABA could be an increase in chloride conductance bought about by the activation of GABA receptor and if the resting chloride potencial is less negative than resting membrane potencial, the resulting movement of chloride out of the cell will cause depolarization. On the other hand, depolarization could cause the direct activation of voltage-sensitive calcium channels. Thus both calcium and chloride channel activation might be involved in the GHreleasing effect of GABA in the neonatal pituitary.
Vol.
52, No.
21,
1993
GABA-induced
GH S e c r e t i o n
and C h l o r i d e C h a n n e l s
1739
Acknowledgements We thank Ms J. R~tcz for the expert technical help and the Radioimmunoassay Laboratory of our Institute for the GH measurements. We are grateful to the NIDDK for the generous gift of the rGH RIA material. This work was supported by the grant OTKA 2922. References 1. ZS. ACS, G.B. MAKARA and E. STARK, Life Sciences 34 1505-1511 (1984). 2. ZS. ,~CS, B. SZABO, G. KAPOCS and G.B.MAKARA, Endocrinology 120 1790-1798 (1987). 3. P.R. ANDREWS and G.A.R. JOHNSTON, Biochemical Pharmacology 28 2697-2702 (1979). 4. E. CHERUBINI, J.L. GAIARSA and Y. BEN-ARI, TINS 14 515-519 (1991). 5. ZS..,%CS, L. ZSOM and G.B. MAKARA, Life Sciences 50 273-279 (1992). 6. GY. HORVA.TH, ZS. ,&CS, ZS. MERGL, I. NAGY and G.B. MAKARA, Neuroendocrinology (in press). 7. C.A. BIRGE, G.T. PEAKE, I.K. MARIZ and W.H. DAUGHADAY, Endocrinology 81 195-204 (1967). 8. C.W. DUNNETI', J. Am. Stat. Assoc. 50 1096-1121 (1955). 9. O.J. DUNN, J. Am. Stat. Assoc. 56 52-61 (1961). 10. J.D. TURNER, R.J. BOAKES, J.A. HARDY and M.A. VIRMANI, J. Neurochemistry 48 1060-1068 (1987). 11. J.A. HARDY, R.J. BOAKES, D.J.E. THOMAS, A.M. KIDD, J.A. EDWARDSON, M. VIRMANI, J. TURNER and P.R. DODD, J. Neurochemistry 42 875-877 (1984). 12. M.A. VIRMANI, S.S. STOJILKOVIC and K.J. CAT1r, Endocrinology 126 2499-2505 (1990). 13. L. GRANDISON and A. GUIDOqTI, Endocrinology 105 754-759 (1979). 14. S.FISZER DE PLAZAS, D. BECU, A. MITRIDATE DE NOVARA, and C. LIBERTUN, Endocrinology 111 1974-1978 (1982). 15. F.V. DEFEUDIS, Neurochem, Int. 6 1-16 (1984). 16. R.A.ANDERSON and R. MITCHELL, J. Endocrinology 108 1-8 (1986). 17. L.T. ARANCIBIA, J.P. ROUSSEL and H. ASTIER, Endocrinology 121 980-986 (1987). 18. R.D. SCHWARTZ, Biochemical Pharmacology 37 3369-3375 (1988). 19. U. MISGELD, A. WAGNER and T. OHNO, Exp. Brain Res. 45 108-124 (1982). 20. P. CALABRESI, M. BENEDETFI, N.B. MERCURI, and G. BERNARDI, Neuroscience 30 663-670 (1989). 21. P. CALABRESI, N.B. MERCURI. and G. BERNARDI, J. Neurophysiol. 63 663-675 (1990). 22. L. CUTTLER, S.R. GAUM, B.A. COLLINS and R.J. MILLER, Endocrinology 130 945953 (1992).