Journal of Neuroendocrinology, 2015, 27, 166–176 © 2014 British Society for Neuroendocrinology

ORIGINAL ARTICLE

Multiple Ca2+ Channel-Dependent Components in Growth Hormone Secretion from Rat Anterior Pituitary Somatotrophs E. Sosial and I. Nussinovitch Department of Medical Neurobiology, Institute for Medical Research-Israel-Canada, The Hebrew University Faculty of Medicine, Jerusalem, Israel.

Journal of Neuroendocrinology

Correspondence to: Itzhak Nussinovitch, Department of Medical Neurobiology, Institute for Medical Research-Israel-Canada, The Hebrew University Faculty of Medicine, PO Box 12272, Jerusalem 91120, Israel (e-mail: [email protected]).

The involvement of L-type Ca2+ channels in both ‘basal’ and ‘stimulated’ growth hormone (GH) secretion is well established; however, knowledge regarding the involvement of non-L-type Ca2+ channels is lacking. We investigated whether non-L-type Ca2+ channels regulate GH secretion from anterior pituitary (AP) cells. To this end, GH secretion was monitored from dissociated AP cells, which were incubated for 15 min with 2 mM K+ (‘basal’ secretion) or 60 mM K+ (‘stimulated’ secretion). The role of non-L-type Ca2+ influx was investigated using specific channel blockers, including x-agatoxin-IVA, x-conotoxin GVIA or SNX-482, to block P/Q-, N- or R-type Ca2+ channels, respectively. Our results demonstrate that P/Q-, N- and R-type Ca2+ channels contributed 21.2  1.9%, 20.2  7.6% and 11.4  1.8%, respectively, to ‘basal’ GH secretion and 18.3  1.0%, 24.4  5.4% and 14.2  4.8%, respectively, to ‘stimulated’ GH secretion. After treatment with a ‘cocktail’ that comprised the previously described non-L-type blockers, non-L-type Ca2+ channels contributed 50.9  0.4% and 45.5  2.0% to ‘basal’ and ‘stimulated’ GH secretion, respectively. Similarly, based on the effects of nifedipine (10 lM), L-type Ca2+ channels contributed 34.2  3.7% and 54.7  4.1% to ‘basal’ and ‘stimulated’ GH secretion, respectively. Interestingly, the relative contributions of L-type/non-L-type Ca2+ channels to ‘stimulated’ GH secretion were well correlated with the relative contributions of L-type/non-Ltype Ca2+ channels to voltage-gated Ca2+ influx in AP cells. Finally, we demonstrated that compartmentalisation of Ca2+ channels is important for GH secretion. Lipid raft disruption (methyl-b-cyclodextrin, 10 mM) abrogated the compartmentalisation of Ca2+ channels and substantially reduced ‘basal’ and ‘stimulated’ GH secretion by 43.2  3.4% and 58.4  4.0%, respectively. In summary, we have demonstrated that multiple Ca2+ channel-dependent pathways regulate GH secretion. The proper function of these pathways depends on their compartmentalisation within AP cell membranes. Key words: Ca2+ channels, anterior pituitary, somatotrophs, growth hormone, lipid rafts

Growth hormone (GH) secretion from the anterior pituitary (AP) gland is primarily regulated by two hypothalamic hormones: GH-releasing hormone (GHRH) and somatostatin (1), as well as the gastrointestinal tract hormone ghrelin (2). At the cellular level, both ‘basal’ and ‘stimulated’ GH secretion have been observed. ‘Basal’ GH secretion is most likely generated by spontaneous rhythmic electrical activity and voltage-gated Ca2+ influx (VGCI), which is coupled to this electrical activity (3,4). This notion is supported by the spontaneous oscillations in intracellular calcium concentrations [Ca2+]i observed in somatotrophs (3,5). ‘Stimulated’ GH secretion can be induced by K+-depolarisations (6) or by hormones that act on somatotrophs. For example, GHRH depolarises somatotrophs (7–9)

doi: 10.1111/jne.12240

and stimulates GH secretion by augmenting both VGCI and [Ca2+]i (10–13). The involvement of L-type Ca2+ channels in both ‘basal’ and ‘K+-stimulated’ GH secretion from somatotrophs has been well documented (3,6,14–16). However, substantially less is known regarding the involvement of P/Q-type, N-type and R-type Ca2+ channels in GH secretion. We recently identified multiple Ca2+ channel a1 subunits in the membranes of AP cells, including L-type (Cav1.2, Cav1.3), P/Q-type (Cav2.1), N-type (Cav2.2) and R-type (Cav2.3) channels. We also demonstrated that a substantial fraction (approximately 46%) of the VGCI in AP cells is carried through non-L-type Ca2+ channels (P/Q-, N- and R-type channels) (17). Additionally, via flotation

Non-L-type Ca2+ channels and GH secretion

experiments, we demonstrated that these multiple pathways for VGCI in AP cells are segregated among raft and nonraft membrane compartments (17). In the present study, we examined the relative contribution of L-type and non-L-type VGCI to GH secretion. We demonstrate that multiple Ca2+ channel-dependent components are involved in both ‘basal’ and ‘stimulated’ GH secretion, which depend on VGCI through L-type and non-L-type Ca2+ channels. Moreover, we demonstrate that these multiple components in ‘stimulated’ GH secretion are considerably correlated with the multiple components of VGCI. Furthermore, we demonstrate that the lateral segregation of Ca2+ channels among raft and nonraft membrane domains in the membrane of pituitary cells is important for both ‘basal’ and ‘stimulated’ GH secretion.

167

membrane is highly permeable to K+, an increase in extracellular K+ concentration [K]e from 2 to 40 or 60 mM K+ is expected to depolarise pituitary cells at most by 76 or 86 mV, respectively (58 log([K]e2/[K]e1), where [K]e1 = 2 mM K+ and [K]e2 = 40 or 60 mM K+. Under the same assumption, a decrease in [K]e from 5 mM K+ (physiological concentration) to 2 mM K+ (our ‘control’ solution) is expected to hyperpolarise pituitary cells at the most by 23 mV (which increases the availability of Ca2+ channels). Thus, an increase in [K]e from 2 to 40 or 60 mM K+ is expected to depolarise pituitary cells, with respect to their physiological resting potential, at most by 53 or 63 mV, respectively. The resting membrane potential in pituitary cells was reported to oscillate between 60 and 50 mV (20). Therefore, it is reasonable to assume that ‘stimulated’ GH secretion in our experiments was monitored at membrane potentials near or below 0 mV. Depolarisations to these membrane potentials are expected to produce a maximal VGCI based on our previous whole-cell experiments in AP cells (17).

Materials and methods Primary cell cultures Animals were sacrificed in accordance with the guidelines of the Authority for Animal Facilities-Ethics Committee, Hebrew University (Jerusalem, Israel). AP glands were dissected from three male rats (Sabra strain, 250–300 g, 7–8 weeks old) for each single experiment (n = 1) and a total of 9–18 rats (n = 3–6) were used for each one of the experimental points in the present study. The isolated anterior lobes (after the removal of intermediate and posterior lobes) were subsequently subjected to approximately 35 min of enzymatic dissociation at 37 °C as described previously (18,19). In brief, the dissociation medium (F-12; Biological Industries, Beit Haemek, Israel) contained: 1.5 mg/ml protease (type XIV, 5.6 U/mg), 0.6 mg/ml dispase (type II), 1 mg/ml collagenase (type I), 40 U/ml DNase (type II), 2.5 mg/ml bovine serum albumin (fraction V) and the antibiotic kanamycin sulphate (2.5 mM). All enzymes and chemicals were obtained from Sigma (St Louis, MO, USA). After dissociation, the cells were re-suspended in F-12, centrifuged and washed (twice) to terminate enzymatic activity, mesh filtered (100 lm) and seeded into 48-well culture plates (50 ll per well, which contained approximately 50 000 cells). The culture plates were then transferred to the incubator (37 °C, 5% CO2). After 2 h, during which the cells were allowed to adhere, 400 ll of incubation medium was added to each well. The pituitary cells were maintained in the incubator for 48 h before the start of the GH secretion experiments. The ‘incubation’ medium (Dulbecco’s modified Eagle’s medium) was supplemented with 10% foetal calf serum, 1% glutamine and penicillin/streptomycin (medium and supplements from Biological Industries, Kibbutz Beit Haemek, Israel).

Experimental solutions GH secretion was assayed after incubation with three distinct physiological solutions: one ‘control’ (2 mM K+) solution, which was used to monitor ‘basal’ GH secretion, and two ‘high-K+’ solutions (40 mM K+ and 60 mM K+), which were used to monitor depolarisation-stimulated GH secretion. The ‘control’ physiological solution contained (mM): 140 NaCl, 2 KCl, 1 MgCl2, 5 CaCl2, 10 D-glucose, 10 HEPES and 0.1% bovine serum albumin (pH 7.2). The ‘high-K+’ solutions contained the same constituents, with the exception that the NaCl was replaced with an equal concentration of KCl; 40 mM K+ and 60 mM K+ solutions contained 102 and 82 mM NaCl, respectively.

K+-depolarisations In accordance with the assumptions that the intracellular K+ concentration [K]i in pituitary cells remains constant and that the pituitary cell

Journal of Neuroendocrinology, 2015, 27, 166–176

GH secretion After 48 h in the incubator, the tissue culture plates were transferred to the hood (approximately 24 °C). The ‘incubation’ medium was gently removed; the pituitary cells were washed three times with the ‘control’ physiological solution and then equilibrated for 15 min with fresh ‘control’ solution. At the end of equilibration period, the ‘control’ solution was replaced with fresh ‘control’ (2 mM K+), 40 mM K+ or 60 mM K+ solutions for 15 min of incubation. Duplicate samples (approximately 400 ll) for each experimental condition were then collected and stored at 20 °C for the subsequent GH assay. To examine the involvement of Ca2+ channels in GH secretion, Ca2+ channel blockers were added to the ‘control’ (2 mM K+), 40 mM K+ and 60 mM K+ solutions for 15 min of incubation. In each experiment, ‘basal’ (2 K+) and ‘stimulated’ (40 K+, 60 K+) GH secretion, in the absence and presence of Ca2+ channel blockers, were assayed and compared. The effects of each channel blocker were repeatedly examined during independent experiments (n ≥ 3). Each single experimental point (n = 1) represents the averaged GH secretion from approximately 100 000 AP cells (two samples, each of which included approximately 50 000 AP cells) that were obtained from three male rats. Thus, our minimal n value (n = 3) represents the averaged GH secretion from 300 000 AP cells cultured from nine male rats, whereas our n values (n = 3–6) represent experiments performed on nine to 18 rats. The GH concentration (ng/ml) in samples was assayed using a Rat GH enzyme-linked immunsorbent assay kit, EZRMGH-45K (Millipore, Billerica, MA, USA). All samples from a single experiment were assayed using the same assay kit.

Calcium channel blockers The involvement of Ca2+ channels in GH secretion was investigated using specific channel blockers, including nifedipine to block L-type currents and peptide toxins to block non-L-type, P/Q-type, N-type and R-type channels. P/Q-type (Cav2.1) channels were blocked using x-agatoxin-IVA (250 nM). This concentration is expected to effectively block both P-type and Q-type currents (IC50 values of approximately 1 and 100 nM, respectively) (21,22). N-type (Cav2.2) channels were blocked by x-conotoxin GVIA (2 lM). This concentration is expected to effectively block N-type currents (23). R-type (Cav2.3) channels were blocked by SNX-482 (30 nM). This concentration is expected to effectively block R-type currents (IC50 values of approximately 15–30 nM) (24). At higher concentrations, SNX-482 may block L-type (25), N-type (24) and P/Q-type (26) currents. Importantly, the same blockers at the same concentrations have been previously used to block high-voltage activated (HVA) Ca2+ channel currents (with Ba2+ as a charge carrier) in AP cells (17). Nifedipine was purchased from Sigma and the peptide toxins © 2014 British Society for Neuroendocrinology

168

E. Sosial and I. Nussinovitch

were obtained from Alomone Labs (Jerusalem, Israel). Nifedipine was dissolved in dimethylsulphoxide (DMSO; final concentration of 0.05%), and the peptide toxins were dissolved in the ‘control’ physiological solution. DMSO 0.05% had no effects on GH secretion (see Supporting information, Fig. S1).

Identification of Ca2+ channels in lipid microdomains Membrane extraction A 100-mg sample from the AP glands was homogenised in 800 ll ice-cold hypo-osmotic buffer [25 mM Tris HCl, pH 7.5, 5 mM ethylenediaminetetraacetic acid (EDTA)], which contained a protease inhibitor (Roche Diagnostic, Indianapolis, IN, USA), for approximately 1 min. The homogenate was then centrifuged (15 min, 1000 g at 4 °C), the pellet discarded, and the supernatant centrifuged again (30 min, 42500 g at 4 °C) to obtain a pellet of extracted membranes. To increase the extraction yield, this pellet was resuspended in ice-cold hypo-osmotic buffer for approximately 1 min and centrifuged (30 min, 42500 g at 4 °C). The pellet (which contained the extracted membranes) was then resuspended in 1 ml phosphate-buffered saline (PBS).

Cholesterol depletion The cholesterol membrane content was manipulated using methyl-b-cyclodextrin (MbCD) or the less effective drug 2-hydroxypropyl-b-cyclodextrin (OHpMbCD). These drugs are commonly used to deplete or enrich cholesterol content in biological membranes (28). Fresh stocks of MbCD or OHpMbCD (100 mM) were prepared before each flotation or GH secretion experiment. During the flotation experiments, extracted pituitary cell membranes (see Membrane extraction) were incubated with 10 mM MbCD for 30 min at 37 °C (55 ll of 100 mM MbCD added to 495 ll of extracted membranes in PBS). The incubation was terminated by cooling samples to 4 °C (placing them on ice for 5–10 min). DRMs were then extracted by incubating with 1% Triton for 15 min (27.5 ll of 20% Triton buffer added to a 550 ll sample) as described previously. During the GH secretion experiments and after the 15-min equilibration period in the ‘control’ solution (see GH secretion), the pituitary cells were incubated for an additional 15 min with ‘control’ solution that contained MbCD (5 or 10 mM) or OHpMbCD (10 mM). At the end of this period, the MbCD or OHpMbCD solution was replaced with the experimental solution (2, 40 or 60 mM K+) for an additional 15 min.

Statistical analysis Flotation assays Detergent-resistant membranes (DRMs) were isolated from extracted membranes (see membrane extraction) using Nycodenz gradients as described previously (17,27). Briefly, 500 ll of extracted membranes were added to 50 ll of ice-cold Triton buffer (25 mM Tris HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA and 10% Triton 9100) for 15 min at an ice-cold temperature (the final concentration of Triton was approximately 1%). After incubation, the extracted membranes were adjusted to 35% Nycodenz by adding an equal volume (550 ll) of ice-cold 70% Nycodenz prepared in TNE (25 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM EDTA) and loaded into the bottom of an ultracentrifuge tube (TLS-55; Beckman Coulter, Fullerton, CA, USA). An 8–35% Nycodenz linear step gradient in TNE was then overlaid above the extracted membranes (200 ll for each step; 25%, 22.5%, 20%, 18%, 15%, 12% and 8% Nycodenz) and spun at 260 000 g for 4 h at 4 °C. Samples of 180 ll were then collected from the top to the bottom of the tube and subjected to immunoblotting using sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis.

Western blot analysis The distribution of a1 subunits among the light and heavy Nycodenz gradient fractions was resolved using 6% (for Ca2+ channels) and 13% (for caveolin-1) polyacrylamide gels in Tris/glycine running buffer (25 mM Tris-base, pH 8.3, 190 mM glycine and 0.1% SDS). Gel electrophoresis was performed at 100 V for 30 min (stacking) and at 150 V for 1 h (running). The proteins were subsequently electrotransferred (1 h, 1000 mA) to polyvinylidene difluoride membranes (Millipore) using a Tris/glycine transfer buffer that contained sarkosyl (20 mM Tris-base, pH 8.3, 150 mM glycine, 20% methanol and 0.1% sarkosyl). The membranes were blocked with 5% low-fat milk in TBST (10 mM Tris HCl, pH 7.4, 150 mM NaCl, and 0.05% Tween-20) before incubation with primary antibodies (overnight at 4 °C). The primary antibodies used were rabbit anti-Cav1.2 (dilution 1 : 500; Alomone Labs), antiCav1.3 (dilution 1 : 200; Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Cav2.1 (dilution 1 : 500; Alomone Labs) and anti-caveolin-1 (dilution 1 : 200; Santa Cruz Biotechnology). The secondary antibodies used were goat anti-rabbit IgG-HRP (dilution 1 : 5000; Santa Cruz Biotechnology). Ca2+ channel a1 subunits at various gradient fractions were identified using enhanced chemiluminescence. © 2014 British Society for Neuroendocrinology

The results are reported as the mean  SE. Significant differences between two groups of tested parameters were examined using a paired Student’s t-test. Significant differences among many groups of tested parameters were examined using one-way ANOVA followed by Holm–Sidak post-hoc pairwise multiple comparison tests (SIGMA PLOT, version 11; Systat Software Inc., Chicago, IL, USA). P < 0.05 was considered statistically significant.

Results Multiple Ca2+ channel-dependent components in GH secretion To determine the appropriate experimental conditions, GH secretion was first assayed after 5-, 10- and 15-min incubations with increasing extracellular K+ concentrations [K]e (2–100 mM). Figure 1 shows that definitive GH responses were observed only after 15 min of incubation. An increase in [K]e caused a gradual increase in GH secretion. However, at 100 mM K+ (and sometimes at 80 mM K+, not shown), stimulated GH secretion was inhibited. This inhibition at high [K]e may result from Ca2+-dependent inactivation (CDI) of Ca2+ channels (29) or Ca2+-dependent internalisation of Ca2+ channels (30,31). A similar dependence of GH secretion on [K]e was observed in additional standardisation experiments (n = 3); therefore, we monitored GH secretion after 15-min incubations with 2 mM K+ (‘basal’ secretion) and 60 mM K+ (‘stimulated’ secretion). GH secretion after 15-min incubations with 40 mM K+ was also monitored to provide an additional reference point. As expected, Fig. 2(A,B) shows that blocking L-type Ca2+ channels with a saturating concentration of nifedipine (10 lM) significantly reduced both ‘basal’ and ‘stimulated’ GH secretion to 65.8  3.7% (n = 5) and 45.3  4.1% (n = 5) of control values, respectively. However, a significant component of GH secretion persisted during this nifedipine block, suggesting the involvement of non-L-type Ca2+ channels. Figure 2(C,D) shows that a ‘cocktail’ of non-L-type Ca2+ channel blockers significantly reduced both ‘basal’ and Journal of Neuroendocrinology, 2015, 27, 166–176

Non-L-type Ca2+ channels and GH secretion

GH secretion 120

GH (ng/ml)

100

Stimulated Basal

80 60 40 5 min 10 min 15 min

20 0

2 mM

20 mM 40 mM 60 mM 80 mM 100 mM [K+]

Fig. 1. K+- stimulated growth hormone (GH) secretion: standardisation. Anterior pituitary cells were incubated with increasing extracellular K+ concentrations [K+]e for 5, 10 or 15 min (same experiment). At the end of incubation periods, samples were collected, and the GH content was determined using enzyme-linked immunsorbent assay kits. The results demonstrate a clear dependence of GH secretion on [K+]e after 15-min incubations (but not after 5- or 10-min incubations). An increase in [K+]e caused a gradual increase in GH secretion (2–80 mM K+). Similar results were obtained during repeat experiments (n = 3). We therefore continued to monitor GH secretion at 2, 40 and 60 mM [K+]e (arrows). GH secretion at 2 and 60 mM [K+]e represents ‘basal’ and ‘stimulated’ secretion, respectively. GH secretion at 40 mM [K+]e was measured as an additional reference point between ‘basal’ and ‘stimulated’ secretion.

‘stimulated’ GH secretion to 49.1  0.4% (n = 3) and 54.5  2.0% (n = 3) of control values, respectively. The ‘cocktail’ of toxins contained 250 nM x-agatoxin-IVA, 2 lM x-conotoxin GVIA and 30 nM SNX-482 to block P/Q-type, N-type and R-type Ca2+ currents, respectively. An identical ‘cocktail’ was recently used to block nonL-type Ca2+ channel currents in AP cells (17). These toxin concentrations are expected to specifically and effectively block P/Q-type, N-type, and R-type Ca2+ channels (see Materials and methods) (17). The contribution of individual non-L-type Ca2+ channels to GH secretion was further investigated using specific Ca2+ channel toxins. Figure 3(A) demonstrates an experiment in which the effects of the ‘cocktail’ on GH secretion were compared with the effects of its toxin constituents. It is evident from this experiment that the specific block of N-type, P/Q-type or R-type Ca2+ channels resulted in the inhibition of GH secretion and that the combined effects of the ‘cocktail’ were larger than the effects of the individual toxins on GH secretion. Figure 3(B) illustrates a different experiment in which the effects of the ‘cocktail’ on GH secretion were compared with the effects of individual toxins, SNX-482 and x-conotoxin MVIIC, as well as the effects of cadmium (150 lM). As expected, the effects of individual toxins were smaller than the effects of the ‘cocktail’ and the effects of cadmium were larger than those of the ‘cocktail’. These experiments were repeated and the effects are summarised and compared in Fig. 4 and Table 1. Statistical analysis indicated that the effects of the ‘cocktail’ were significantly larger than the effects of single Ca2+ channels blockers and the effects of cadmium were significantly larger than the effects of nifedipine and the ‘cocktail’ (Fig. 4). The potent inhibition of both ‘basal’ and ‘stimulated’ GH secretion by cadmium (150 lM) confirms that these Journal of Neuroendocrinology, 2015, 27, 166–176

169

components of GH secretion largely depend on the VGCI. The residual GH secretion after cadmium block suggests that a fraction of Ca2+ channels were not blocked at this concentration. Additionally, based on the comparison between the effects of nifedipine and ‘cocktail’, the relative contribution of non-L-type Ca2+ channels to ‘basal’ GH secretion (50.9  0.4%; n = 3) was significantly greater than the contribution of L-type channels (34.2  3.7%; n = 5). By contrast, the relative contributions of non-L-type Ca2+ channels to ‘stimulated’ GH secretion (45.5  2.0%; n = 3) was not significantly different from that of L-type Ca2+ channels (54.7  4.1%; n = 5) (Fig. 4). In summary, our results clearly demonstrate multiple components in both ‘basal’ and ‘stimulated’ GH secretion, both of which depend on the VGCI through L-type, P/Q-type, N-type and R-type Ca2+ channels.

Multiple components in GH secretion and Ba2+ influx are correlated In our previous study, we demonstrated that HVA Ba2+ influx in AP cells is carried through L-type, P/Q-type, N-type and R-type Ca2+ channels (17). We therefore examined whether the contributions of these Ca2+ channel types to GH secretion and HVA Ba2+ influx are similar. Table 1 (see also Supporting information, Fig. S2) compares the effects of Ca2+ channel blockers on ‘stimulated’ GH secretion and the HVA Ba2+ influx. The effects of nifedipine, x-agatoxin-IVA and SNX-482 on ‘stimulated’ GH secretion were not significantly different from their effects on the HVA Ba2+ influx. By contrast, the effects of x-conotoxin-GVIA on ‘stimulated’ GH secretion were significantly larger than on the HVA Ba2+ influx. These findings may indicate that the relative contribution of N-type channels to the HVA Ba2+ influx in our previous study was underestimated.

Ca2+channel localisation and GH secretion Using flotation assays, we recently demonstrated that HVA Ca2+ channels in AP cells are segregated among various membrane compartments (17); Cav2.1 channels (and the raft marker caveolin-1) were primarily localised in light gradient fractions, Cav1.2 and Cav1.3 channels were distributed among light and heavy gradient fractions, and Cav2.2 and Cav2.3 were primarily localised in heavy gradient fractions. The localisation of Cav2.1, Cav1.2 and Cav1.3 in light gradient fractions (i.e. in DRMs) indicates their localisation in cholesterol-rich lipid microdomains (32,33). We therefore examined whether cholesterol depletion, which is expected to disrupt lipid microdomains, shifts the localisation of these channels from light to heavy gradient fractions (i.e. from raft to nonraft membrane domains) (34,35). Thus, flotation experiments were repeated after membrane cholesterol depletion using MbCD (see Materials and methods section). Figure 5 demonstrates that pre-treatment with MbCD (10 mM for 30 min) to deplete membrane cholesterol and disrupt lipid rafts shifted the Cav channels (Cav1.2, Cav1.3 and Cav2.1) and caveolin-1 from light to heavy gradient fractions, which strengthens the evidence for their localisation in cholesterol-rich membrane domains. As shown in the Supporting information © 2014 British Society for Neuroendocrinology

170

E. Sosial and I. Nussinovitch

(A)

( C)

90

90 Cocktail of non-L-type blockers * *

Nifedipine * *

30

% of control GH secretion

100

Nifedipine

80

n=5

40 mM [K+]

Control, n = 3 Cocktail, n = 3

0 2 mM

60 mM

( D)

40 mM [K+]

60 mM

100 Cocktail of non-L-type blockers % of control GH secretion

2 mM

(B)

30

Control, n = 5 Nifedipine, n = 5,4,5

0

*

60

GH (ng/ml)

GH (ng/ml)

60

n=4

60

n=5

40 20

80 60

n=3

n=3

n=3

40 20 0

0 2 mM

40 mM [K+]

60 mM

2 mM

40 mM [K+]

60 mM

Fig. 2. Growth hormone (GH) secretion depends on both L-type and non-L-type Ca2+channels. (A) Effects of nifedipine (10 lM) on GH secretion at 2, 40 and 60 mM K+. Nifedipine reduced GH secretion. The effects at 2 and 60 mM K+ but not at 40 mM K+ were significant compared to the control. (B) Same experiments as in (A), illustrating the effects of nifedipine on the percentage of control GH release at 2, 40 and 60 mM K+. One-way ANOVA identified significant differences among the mean values. Post-hoc pairwise multiple comparison tests identified significant differences between the effects at 2 and 60 mM K+ and those at 40 and 60 mM K+ but not between those at 2 and 40 mM K+. (C) Effects of the ‘cocktail’ on GH secretion at 2, 40 and 60 mM K+. All effects were significant compared to the control. (D) Same experiments as in (C), illustrating the effects of the cocktail as the percentage of control GH release at 2, 40 and 60 mM K+. One-way ANOVA and post-hoc pairwise multiple comparison tests indicated that there were no differences among the mean values. The ‘cocktail’ of non-L-type channel blockers contained 250 nM x-agatoxin-IVA, 2 lM x-conotoxin GVIA and 30 nM SNX-482 to block P/Q-, N- and R-type Ca2+ channels, respectively.

(B) GH secretion (15 min)

90

Control

90

Control SNX-482

60

SNX-482 Aga-IVA Cono-GVIA

60

Cono-MVIIC

Cocktail 30

0

2 mM

40 mM 60 mM [K+]

Gh (ng/ml)

Gh (ng/ml)

(A) GH secretion (15 min)

Cocktail 30

0

Cd

2 mM

40 mM 60 mM [K+]

Fig. 3. Dependence of growth hormone (GH) secretion on P/Q-type, N-type, and R-type Ca2+ channels. Two single experiments are illustrated. (A) In the first experiment, the effects of the ‘cocktail’ on GH secretion were compared with the effects of its toxin constituents. The ‘cocktail’ constituents included 300 nM x-agatoxin-IVA (Aga-IVA) (P/Q-type blocker), 2 lM x-conotoxin GVIA (Cono-GVIA) (N-type blocker) and 30 nM SNX-482 (R-type blocker). (B) In the second experiment, the effects of the ‘cocktail’ on GH secretion were compared with the effects of 30 nM SNX-482 (R-type blocker) and 1 lM x-conotoxin MVIIC (Cono-MVIIC) (P/Q- and N-type blocker). The effects of SNX-482 in this experiment were smaller than the effects in the first experiment (A). This may be explained by differences in the relative proportions of R-type channels in the two experiments combined with any experimental error. Additionally, the effects of cadmium (Cd) on GH secretion were examined. Cd (150 lM) potently reduced both ‘basal’ and ‘stimulated’ GH release to approximately 17% and 27% of the control secretions, respectively. © 2014 British Society for Neuroendocrinology

Journal of Neuroendocrinology, 2015, 27, 166–176

Non-L-type Ca2+ channels and GH secretion

(A) 100

171

Basal secretion (2 mM K+)

% Decrease in GH secretion

n=3 80 60 40

n=3 n=5 n=4

n=4 20

n=4 n=3

0 Nifed L-type

Aga-IVA Cono-GVIA SNX Cono-MVIIC Cocktail Non-L-type P/Q-type N-type R-type P/Q&N-type

Cd

(B) % Decrease in GH secretion

100

Stimulated secretion (60 mM K+) n=3

80 60

n=5 n=4

n=3 40 n=4 n=4

20

n=3

0 Nifed L-type

Aga-IVA Cono-GVIA SNX Cono-MVIIC Cocktail Non-L-type P/Q-type N-type R-type P/Q&N-type

Cd

Fig. 4. Multiple Ca2+ channel-dependent components of growth hormone (GH) secretion. Histograms summarising the effects of Ca2+ channel blockers on GH secretion. (A) Effects on ‘basal’ GH secretion ([K+]e = 2 mM K+). One-way ANOVA identified significant differences among the mean values of the cocktail of AgaIVA, Cono-GVIA and SNX. Post-hoc pairwise multiple comparison tests identified significant differences between the pairs: cocktail-(Aga-IVA); cocktail-(ConoGVIA); and cocktail-SNX. Additionally, one-way ANOVA identified significant differences among the mean values of nifedipine, cocktail and cadmium. Post-hoc pairwise multiple comparison tests identified significant differences between all pairs: nifedipine-cocktail; cadmium-nifedipine; and cadmium-cocktail. (B) Effects on ‘stimulated’ GH secretion ([K+]e = 60 mM K+). One-way ANOVA identified significant differences between the mean values of cocktail; Aga-IVA; ConoGVIA and SNX. Post-hoc pairwise multiple comparison tests identified significant differences between the pairs: cocktail-(Aga-IVA); cocktail-(Cono-GVIA) and cocktail-SNX. Additionally, one-way ANOVA identified significant differences among the mean values of nifedipine, cocktail and cadmium. Post-hoc pairwise multiple comparison tests identified significant differences between the pairs: cadmium-nifedipine and cadmium-cocktail, although a significant difference was not found for nifedipine-cocktail.

(Fig. S3), this shift is demonstrated further by comparing the distribution of Cav1.2, Cav1.3, Cav2.1 and caveolin-1 among light and heavy gradient fractions under control conditions (17) and after MbCD treatment (present study). We continued these experiments to examine whether cholesterol depletion with MbCD alters ‘basal’ and ‘stimulated’ GH secretion. Figure 6(A) shows the results of a single experiment that compared GH secretion after pre-incubation with MbCD (5 and 10 mM, 15 min) and the less potent dextrin OHpMbCD (10 mM, 15 min) (28). MbCD reduced both ‘basal’ and ‘stimulated’ GH secretion in a concentration-dependent manner. The less potent dextrin OHpMbCD had, as expected, smaller effects on both ‘basal’ and ‘stimulated’ GH secretion. Similar results were obtained in additional experiments and are summarised in Fig. 6(B,C) and Table 2. The statistical analysis identified differences between the effects of MbCD (5 mM) and MbCD (10 mM) but not between the effects of MbCD Journal of Neuroendocrinology, 2015, 27, 166–176

(5 mM) and OHpMbCD (10 mM) (Fig. 6). Thus, it can be concluded that membrane cholesterol depletion (i.e. lipid microdomain disruption) alters both the compartmentalisation of Ca2+ channels on the cell surface and Ca2+-dependent GH secretion.

Discussion In the present study, we demonstrate multiple components of both ‘basal’ and ‘stimulated’ GH secretion, which depend on VGCI through L-type and non-L-type Ca2+ channels. Additionally, we demonstrate that these multiple components of ‘stimulated’ GH secretion are well correlated with the multiple components of Ba2+ influx through L-type, P/Q-type and R-type Ca2+ channels. Furthermore, we demonstrate that the compartmentalisation of Ca2+ channels among raft and nonraft membrane domains in AP cells is important for both ‘basal’ and ‘stimulated’ GH secretion. © 2014 British Society for Neuroendocrinology

172

E. Sosial and I. Nussinovitch

Table 1. Effects of Ca2+ Channel Blockers on Growth Hormone (GH) Secretion. Percentage of decrease in cumulative GH secretion Ca

2+

channel blockers

Nifedipine (10 lM) Cocktail; non-L-type blockers x-Agatoxin-IVA (250 nM) x-Conotoxin-GVIA (2 lM) SNX-482 (30 nM) x-Conotoxin-MVIIC (1 lM) Cadmium (150 lM)

Channels blocked

‘Basal’ secretion

L-type P/Q-type, N-type, R-type P/Q-type N-type R-type P/Q-type, N-type All VGCCs

34.2 50.9 21.2 20.2 11.4 21.1 85.6

      

3.7 0.4 1.9 7.6 1.8 3.2 1.3

(n (n (n (n (n (n (n

= = = = = = =

‘Stimulated’ secretion 5) 3) 4) 4) 3) 4) 3)

54.7 45.5 18.3 24.4 14.2 40.9 78.0

      

4.1 2.0 1.0 5.4 4.8 7.3 2.9

(n (n (n (n (n (n (n

= = = = = = =

5) 3) 4) 4) 3) 4) 3)

Percentage of decrease in peak Ba2+ current obtained from Tzour et al. (17) 45.4 30.0 17.3 9.7 12.1 – –

    

2.6 4.0 2.5 1.0 0.5

(n (n (n (n (n

= = = = =

9) 8) 16) 17) 9)

One-way ANOVA followed by post-hoc analysis of Ca2+ channel blocker effects indicated that the effects of the ‘cocktail’ were significantly greater than the effects of the single Ca2+ channel blockers, and the effects of cadmium were significantly larger than the effects of nifedipine and the ‘cocktail’ for both ‘basal’ and ‘stimulated; GH secretion. Additionally, the effects of the cocktail were significantly greater than the effects of nifedipine on ‘basal’ GH secretion but not ‘stimulated’ GH secretion (Fig. 4).

Multiple Ca2+ channel-dependent components of GH secretion Evidence for the dependence of GH secretion on Ca2+ influx through non-L-type channels first originated from the combined effects of non-L-type Ca2+ channel blockers (the ‘cocktail’) on GH secretion (Fig. 2). Additional evidence for multiple Ca2+ channeldependent components of GH secretion originated from the separate effects of P/Q-type, N-type and R-type channel blockers on GH secretion (Figs 3 and 4). Interestingly, under basal conditions ([K+]e = 2 mM K+), the relative contribution of non-L-type channels on GH secretion was significantly greater than the contribution of L-type channels (Table 1). By contrast, in response to stronger depolarisation ([K+]e = 60 mM K+), the relative contributions of L-type and non-L-type channels to GH secretion were similar. These findings suggest that, under basal conditions, Ca2+ influx, which triggers GH secretion, is carried largely through non-L-type Ca2+ channels. This finding may be explained by differences in the biophysical properties (e.g., V0.5 for activation and inactivation) and regulation of non-L-type Ca2+ channels (P/Q-, N- and R-types) compared to L-type Ca2+ channels (21,22,36). The similarities between the relative contributions of various Ca2+ channel types to K+ ‘stimulated’ GH secretion and HVA Ba2+ influx (Table 1; see also Supporting information, Fig. S2) suggest that the K+-depolarisations used to stimulate GH secretion (in response to an increase in [K+]e from 2 to 60 mM K+) were similar to the step depolarisations used to activate maximal HVA Ba2+ influx in AP cells (typically, voltage steps from 80 to 0 mV) (17). Given the assumption that pituitary cell membranes are highly permeable to K+, we estimated that an increase in [K]e from 2 to 40 or 60 mM K+ is expected to depolarise pituitary cells to membrane potentials near or < 0 mV (see Materials and methods section). This similarity exists despite the difference in the charge carrier (Ba2+ versus Ca2+) and the Ca2+-dependent inactivation (29) and internalisation (30,31) of Ca2+ channels, which is expected to substantially reduce Ca2+ currents during 15 min of K+-depolarisation. © 2014 British Society for Neuroendocrinology

Both the ‘basal’ and ‘stimulated’ GH secretions in the present study represent cumulative secretion during a 15-min period. During this time, K+-depolarisations are expected to cause substantial inactivation of Ca2+ channels. Therefore, cumulative ‘stimulated’ GH secretion during 15 min depends both on the initial relative proportion of HVA channels in the membrane and their inactivation time constants. Because HVA Ca2+ channels inactivate at various rates (i.e. L-type channels at a slow rate, P/Q-type and N-type channels at intermediate rates, and R-type channels at a fast rate) (22), their relative contributions to instantaneous ‘stimulated’ GH secretion will be underestimated depending on their inactivation time constants. During 15 min of depolarisation, this underestimation is expected to be significant not only for the fast inactivating R-type channels, but also for the more slowly inactivating Ca2+ channels. This finding may further be investigated by simultaneously monitoring ‘K+-stimulated’ GH secretion and changes in [Ca2+]i at the same time as applying various Ca2+ channel blockers. By contrast, the interpretation of the relative contribution of various HVA Ca2+ channels to ‘basal’ GH secretion may be simpler. ‘Basal’ GH secretion reflects VGCI, which is activated by spontaneous electrical activity reported in somatotrophs (3). This model is supported in the present study by the cadmium block of ‘basal’ GH secretion (illustrated in Figs 3 and 4). Thus, our results may reflect the real contribution of L-, P/Q-, N-, and R-type channels to basal GH secretion under our experimental conditions. Additionally, the relative contribution of various Ca2+ channel types to GH secretion observed in the present study may differ from their relative contribution under normal physiological conditions. First, the present study was performed on dissociated male rat AP cells at room temperature (approximately 24 °C). Under physiological conditions (approximately 37 °C) in intact AP glands, GH secretion is determined by the coordinated action of somatotrophic networks (37). Inter-connected somatotrophs in networks may express a range of Ca2+ channel proportions, which may change during or in between GH surges. Second, in the present study, GH secretion was triggered by a nonspecific stimulus Journal of Neuroendocrinology, 2015, 27, 166–176

Non-L-type Ca2+ channels and GH secretion

(A) MβCD

1

2

3

4

5

6

7

8

9

10

11

12

13

kDa

Cav1.2

200

Cav1.3

200

Cav2.1

170

Caveolin1

25 110 80 50 20 –10

% of maximum density

(B)

173

Cav1.2

110 80 50 20 –10

Cav1.3

110 80 50 20 –10

Cav2.1

110 80 50 20 –10

Caveolin-1

0

2

4

6

8

10

12

14

Fraction # Fig. 5. Cholesterol depletion shifted Ca2+ channels from raft to nonraft membrane domains. (A) Flotation assay after cholesterol depletion with MbCD (10 mM for 30 min). Immunoblots of pituitary Ca2+ channel a1 subunits, Cav1.2, Cav1.3 and Cav2.1, and the raft marker caveolin-1 distributed among various fractions of Nycodenz gradients. Each lane represents a different fraction of the gradient, from the lightest (#1) to heaviest (#13) fractions. (B) Quantitative analysis of the immunoblots in (A), which demonstrated that Cav1.2, Cav1.3 Cav2.1 and caveolin-1 predominantly localise in the heavy gradient fractions (fraction #≥6). Under the control conditions, Cav1.2, Cav1.3, Cav2.1 and caveolin-1 were predominately localised in light gradient fractions (fraction #≤6) (17) (see also Supporting information, Fig. S3). Thus, cholesterol depletion shifted the Ca2+ channel localisation from cholesterol-rich raft membrane domains to nonraft membrane domains.

(K+-depolarisation) from mixed populations of AP cells. This nonspecific stimulus is also expected to trigger the secretion of other pituitary hormones, which may in turn indirectly affect GH secretion (38). Thus, additional studies are needed to quantitatively determine the relative contribution of the various Ca2+ channels to GH secretion under a variety of physiological conditions.

Ca2+ channel compartmentalisation and GH secretion We previously reported that large fractions of L-type channels (Cav1.2, Cav1.3) and most P/Q-type channels (Cav2.1) reside in DRMs (i.e. in either caveolar or noncaveolar lipid microdomains) (17). In the present study, we provide further evidence for the localisation of these channel types in cholesterol-rich membrane domains; lipid raft disruption, using cholesterol depletion (MbCD 10 mM), shifted these Ca2+ channels from raft to nonraft membrane domains (Fig. 5). At least 50% of the VGCI in AP cells is Journal of Neuroendocrinology, 2015, 27, 166–176

carried through L-type and P/Q-type channels (17). Similarly, at least 50% of the K+-stimulated GH secretion is carried through L-type and P/Q-type channels (see Supporting information, Fig. S2) (17). Therefore, it is of substantial interest that lipid raft disruption, using the same methodology (MbCD 10 mM), reduced ‘basal’ and ‘stimulated’ GH secretion by approximately 43% and 58%, respectively (Fig. 6 and Table 2). The exact mechanism that underlies this suppression in GH secretion is unknown. Membrane cholesterol depletion with MbCD is expected to strip cholesterol from the whole membrane and to affect various cellular processes, such as clathrin-dependent endocytosis (39). An MbCD-induced block of endocytosis in AP cells may perturb the balance between endocytosis and exocytosis. This effect may result in a reduction in stimulated exocytosis of GH, which will be reflected as a decrease in GH secretion, as observed in the present study (Fig. 6). Cholesterol depletion is also expected to dramatically alter the function of ion channels (40,41). Previous © 2014 British Society for Neuroendocrinology

174

E. Sosial and I. Nussinovitch

(A)

GH secretion

100

GH (ng/ml)

80

60

40 Control MβCD 5 mM

20

OHpβCD 10 mM

Fig. 6. Cholesterol depletion suppressed both ‘basal’ and ‘stimulated’ growth hormone (GH) secretion. (A) Comparison of the effects of methyl-bcyclodextrin (MbCD) (5, 10 mM) and 2-hydroxypropyl-b-cyclodextrin (OHpMbCD) (10 mM) on GH secretion within a single experiment. It should be noted that the effects of MbCD were concentration-dependent and that the less potent dextrin, OHpMbCD, was less effective than MbCD in the suppression of GH secretion. (B) Histogram summarising the effects of MbCD (5, 10 mM) and OHpMbCD (10 mM) on ‘basal’ GH secretion. One-way ANOVA identified significant differences among the mean values. Post-hoc pairwise multiple comparison tests identified significant differences between all pairs, with the exception of MbCD (5 mM) and OHpMbCD (10 mM). (C) Histogram summarising the effects of MbCD (5, 10 mM) and OHpMbCD on ‘stimulated’ GH secretion. One-way ANOVA identified significant differences among the mean values. Post-hoc pairwise multiple comparison tests identified significant differences between all pairs, with the exception of MbCD (5 mM) and OHpMbCD (10 mM).

MβCD 10 mM 0 2 mM

(B) 120

40 mM [K+]

60 mM

Table 2. Effects of Cholesterol Depletion on Growth Hormone (GH) Secretion. Percentage of control GH secretion

Basal secretion Drugs

‘Basal’ secretion

‘Stimulated’ secretion

MbCD (5 mM) MbCD (10 mM) OHpMbCD (10 mM)

79.4  2.7 (n = 5) 43.2  3.4 (n = 6) 80.8  6.8 (n = 3)

72.7  4.8 (n = 5) 58.4  4.0 (n = 6) 74.4  2.2 (n = 3)

n=6 % of control GH secretion

100 n=5

n=3

80 60 n=6 40 20 0 Control

(C) 120

MβCD 5 mM

OHpβCD 10 mM

MβCD 10 mM

Stimulated secretion n=6

% of control GH secretion

100 n=5

80

n=3 n=6

60 40

One-way ANOVA followed by post-hoc analysis identified significant differences between the effects of methyl-b-cyclodextrin (MbCD) (5 mM) and MbCD (10 mM) but not between the effects of MbCD (5 mM) and 2-hydroxypropyl-b-cyclodextrin (OHpMbCD) (10 mM) in both ‘basal’ and ‘stimulated’ GH secretion (Fig. 6).

studies have demonstrated that membrane cholesterol depletion affected the functional properties of L-type (42) and P/Q-type (43) Ca2+ channels. We recently demonstrated substantial changes in the functional properties of Ca2+ channels in AP cells after alterations in their lipid microenvironment (44). Thus, it is plausible that the MbCD-induced shift in L-type and P/Q-type Ca2+ channels from raft to nonraft membrane domains (Fig. 5) affects their functional properties, thereby decreasing Ca2+ influx and GH secretion. Finally, based on our results, P/Q-type (Cav2.1) channels in AP cells are of special interest. The P/Q-type (Cav2.1) channels are the major constituent in non-L-type Ca2+ influx (17) and in non-L-type-dependent GH secretion (Fig. 4), and predominantly reside in lipid rafts (17), implying that these channels are amenable to specific regulation and play an important role in GH secretion.

Physiological relevance

20 0 Control

MβCD 5 mM

© 2014 British Society for Neuroendocrinology

OHpβCD 10 mM

MβCD 10 mM

The existence of multiple pathways for VGCI in somatotrophs that are able to regulate multiple components of GH secretion may be utilised for the selective regulation or fine-tuning of GH secretion under a variety of physiological conditions. The preferential localisation of Cav1.2 and Cav2.1 in lipid microdomains makes them amenable to spatial and temporal regulation by specific signalling pathways. Although changes in membrane potential are expected

Journal of Neuroendocrinology, 2015, 27, 166–176

Non-L-type Ca2+ channels and GH secretion

to simultaneously regulate all Ca2+ channels, the activation of specific signalling pathways that are able to co-localise with specific Ca2+ channels in lipid microdomains are expected to specifically regulate these Ca2+ channels (e.g. the regulation of cardiac Ca2+ channels by b-adrenergic receptors) (45). Moreover, shifts of these Ca2+ channels from raft to nonraft domains, and vice versa, under various physiological/pathophysiological conditions is expected to affect their function and consequently GH secretion. GH secretion is under complex regulation and modulation by hypothalamic, peripheral and intrapituitary hormones (1,2,14,46,47). It will be of interest to determine whether some of these hormones exert their regulatory (or modulatory) action via selective activation (or inhibition) of specific Ca2+ channel types. Furthermore, long-term modulatory changes in GH secretion may result from changes in the relative proportion of L-type/non-L-type Ca2+ channels in somatotrophs. This finding may be relevant to age-related (48) or metabolic (46) changes in GH secretion. Indeed, age, sex and hormonal statedependent changes in the relative proportion of L-type/non-L-type Ca2+ channels have been reported previously (49,50).

Acknowledgements

10

11

12

13

14

15 16 17

This research was supported by the Israel Science Foundation (ISF) grant number 1325/08 to I.N. 18

Received 24 July 2014, revised 25 November 2014, accepted 25 November 2014

References 1 Muller EE, Locatelli V, Cocchi D. Neuroendocrine control of growth hormone secretion. Physiol Rev 1999; 79: 511–607. 2 Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999; 402: 656–660. 3 Van Goor F, Zivadinovic D, Martinez-Fuentes AJ, Stojilkovic SS. Dependence of pituitary hormone secretion on the pattern of spontaneous voltage-gated calcium influx. Cell type-specific action potential secretion coupling. J Biol Chem 2001; 276: 33840–33846. 4 Stojilkovic SS, Zemkova H, Van Goor F. Biophysical basis of pituitary cell type-specific Ca2+ signaling-secretion coupling. Trends Endocrinol Metab 2005; 16: 152–159. 5 Holl RW, Thorner MO, Mandell GL, Sullivan JA, Sinha YN, Leong DA. Spontaneous oscillations of intracellular calcium and growth hormone secretion. J Biol Chem 1988; 263: 9682–9685. 6 Stojilkovic SS, Izumi S, Catt KJ. Participation of voltage-sensitive calcium channels in pituitary hormone release. J Biol Chem 1988; 263: 13054–13061. 7 Kato M, Suzuki M. Growth hormone releasing factor depolarizes rat pituitary cells in Na+-dependent mechanism. Brain Res 1989; 476: 145–148. 8 Naumov AP, Herrington J, Hille B. Actions of growth-hormone-releasing hormone on rat pituitary cells: intracellular calcium and ionic currents. Pflugers Arch 1994; 427: 414–421. 9 Yang SK, Wang K, Parkington H, Chen C. Involvement of tetrodotoxinresistant Na+ current and protein kinase C in the action of growth hormone (GH)-releasing hormone on primary cultured somatotropes from

Journal of Neuroendocrinology, 2015, 27, 166–176

19

20

21

22 23

24

25

26

27

28

175

GH-green fluorescent protein transgenic mice. Endocrinology 2008; 149: 4726–4735. Nussinovitch I. Growth hormone releasing factor evokes rhythmic hyperpolarizing currents in rat anterior pituitary cells. J Physiol 1988; 395: 303–318. Holl RW, Thorner MO, Leong DA. Intracellular calcium concentration and growth hormone secretion in individual somatotropes: effects of growth hormone-releasing factor and somatostatin. Endocrinology 1988; 122: 2927–2932. Lussier BT, French MB, Moor BC, Kraicer J. Free intracellular Ca2+ concentration and growth hormone (GH) release from purified rat somatotrophs. III. Mechanism of action of GH-releasing factor and somatostatin. Endocrinology 1991; 128: 592–603. Chen C, Clarke IJ. Modulation of Ca2+ influx in the ovine somatotroph by growth hormone-releasing factor. Am J Physiol 1995; 268: E204–E212. Yang SK, Steyn F, Chen C. Influence of membrane ion channel in pituitary somatotrophs by hypothalamic regulators. Cell Calcium 2012; 51: 231–239. Stojilkovic SS, Tabak J, Bertram R. Ion channels and signaling in the pituitary gland. Endocr Rev 2010; 31: 845–915. Stojilkovic SS. Molecular mechanisms of pituitary endocrine cell calcium handling. Cell Calcium 2012; 51: 212–221. Tzour A, Sosial E, Meir T, Canello T, Naveh-Many T, Gabizon R, Nussinovitch I. Multiple pathways for high voltage-activated Ca2+ influx in anterior pituitary lactotrophs and somatotrophs. J Neuroendocrinol 2013; 25: 76–86. Ben-Tabou De-Leon S, Blotnick E, Nussinovitch I. Effects of osmotic swelling on voltage-gated calcium channel currents in rat anterior pituitary cells. Am J Physiol Cell Physiol 2003; 285: C840–C852. Ben-Tabou De-Leon S, Ben-Zeev G, Nussinovitch I. Effects of osmotic shrinkage on voltage-gated Ca2+ channel currents in rat anterior pituitary cells. Am J Physiol Cell Physiol 2006; 290: C222–C232. Bonnefont X, Mollard P. Electrical activity in endocrine pituitary cells in situ: a support for a multiple-function coding. FEBS Lett 2003; 548: 49–52. Sather WA, Tanabe T, Zhang JF, Mori Y, Adams ME, Tsien RW. Distinctive biophysical and pharmacological properties of class A (BI) calcium channel alpha 1 subunits. Neuron 1993; 11: 291–303. Randall AD. The molecular basis of voltage-gated Ca2+ channel diversity: is it time for T? J Membr Biol 1998; 161: 207–213. Plummer MR, Logothetis DE, Hess P. Elementary properties and pharmacological sensitivities of calcium channels in mammalian peripheral neurons. Neuron 1989; 2: 1453–1463. Newcomb R, Szoke B, Palma A, Wang G, Chen X, Hopkins W, Cong R, Miller J, Urge L, Tarczy-Hornoch K, Loo JA, Dooley DJ, Nadasdi L, Tsien RW, Lemos J, Miljanich G. Selective peptide antagonist of the class E calcium channel from the venom of the tarantula Hysterocrates gigas. Biochemistry 1998; 37: 15353–15362. Bourinet E, Stotz SC, Spaetgens RL, Dayanithi G, Lemos J, Nargeot J, Zamponi GW. Interaction of SNX482 with domains III and IV inhibits activation gating of alpha(1E) (Ca(V)2.3) calcium channels. Biophys J 2001; 81: 79–88. Arroyo G, Aldea M, Fuentealba J, Albillos A, Garcia AG. SNX482 selectively blocks P/Q Ca2+ channels and delays the inactivation of Na+ channels of chromaffin cells. Eur J Pharmacol 2003; 475: 11–18. Naslavsky N, Stein R, Yanai A, Friedlander G, Taraboulos A. Characterization of detergent-insoluble complexes containing the cellular prion protein and its scrapie isoform. J Biol Chem 1997; 272: 6324–6331. Kilsdonk EP, Yancey PG, Stoudt GW, Bangerter FW, Johnson WJ, Phillips MC, Rothblat GH. Cellular cholesterol efflux mediated by cyclodextrins. J Biol Chem 1995; 270: 17250–17256.

© 2014 British Society for Neuroendocrinology

176

E. Sosial and I. Nussinovitch

29 Imredy JP, Yue DT. Mechanism of Ca2+-sensitive inactivation of L-type Ca2+ channels. Neuron 1994; 12: 1301–1318. 30 Liu J, Bangalore R, Rutledge A, Triggle DJ. Modulation of L-type Ca2+ channels in clonal rat pituitary cells by membrane depolarization. Mol Pharmacol 1994; 45: 1198–1206. 31 Hall DD, Dai S, Tseng PY, Malik Z, Nguyen M, Matt L, Schnizler K, Shephard A, Mohapatra DP, Tsuruta F, Dolmetsch RE, Christel CJ, Lee A, Burette A, Weinberg RJ, Hell JW. Competition between alpha-actinin and Ca2+-calmodulin controls surface retention of the L-type Ca2+ channel Ca(V)1.2. Neuron 2013; 78: 483–497. 32 Brown DA. Lipid rafts, detergent-resistant membranes, and raft targeting signals. Physiology (Bethesda) 2006; 21: 430–439. 33 Lingwood D, Simons K. Lipid rafts as a membrane-organizing principle. Science 2010; 327: 46–50. 34 Taverna E, Saba E, Rowe J, Francolini M, Clementi F, Rosa P. Role of lipid microdomains in P/Q-type calcium channel (Cav2.1) clustering and function in presynaptic membranes. J Biol Chem 2004; 279: 5127–5134. 35 Xia F, Leung YM, Gaisano G, Gao X, Chen Y, Fox JE, Bhattacharjee A, Wheeler MB, Gaisano HY, Tsushima RG. Targeting of voltage-gated K+ and Ca2+ channels and soluble N-ethylmaleimide-sensitive factor attachment protein receptor proteins to cholesterol-rich lipid rafts in pancreatic alpha-cells: effects on glucagon stimulus-secretion coupling. Endocrinology 2007; 148: 2157–2167. 36 Catterall WA. Structure and regulation of voltage-gated Ca2+ channels. Annu Rev Cell Dev Biol 2000; 16: 521–555. 37 Bonnefont X, Lacampagne A, Sanchez-Hormigo A, Fino E, Creff A, Mathieu MN, Alonso G, Courtois-Coutry N, Pincus SM, Robinson ICAF, Mollard P. Revealing the large-scale network organization of growth hormonesecreting cells. Proc Natl Acad Sci USA 2005; 102: 16880–16885. 38 Denef C. Paracrinicity: the story of 30 years of cellular pituitary crosstalk. J Neuroendocrinol 2008; 20: 1–70. 39 Rodal SK, Skretting G, Garred Ø, Vilhardt F, van Deurs B, Sandvig K. Extraction of cholesterol with methyl-b-cyclodextrin perturbs formation of clathrin-coated endocytic vesicles. Mol Biol Cell 1999; 10: 961–974. 40 Dart C. Lipid microdomains and the regulation of ion channel function. J Physiol 2010; 588: 3169–3178. 41 Levitan I, Fang Y, Rosenhouse-Dantsker A, Romanenko V. Cholesterol and ion channels. Subcell Biochem 2010; 51: 509–549. 42 Pouvreau S, Berthier C, Blaineau S, Amsellem J, Coronado R, Strube C. Membrane cholesterol modulates dihydropyridine receptor function in mice fetal skeletal muscle cells. J Physiol 2004; 555: 365–381.

© 2014 British Society for Neuroendocrinology

43 Davies A, Douglas L, Hendrich J, Wratten J, Van Tran MA, Foucault I, Koch D, Pratt WS, Saibil HR, Dolphin AC. The calcium channel alpha2delta-2 subunit partitions with CaV2.1 into lipid rafts in cerebellum: implications for localization and function. J Neurosci 2006; 26: 8748–8757. 44 Ben-Zeev G, Telias M, Nussinovitch I. Lysophospholipids modulate voltage-gated calcium channel currents in pituitary cells; effects of lipid stress. Cell Calcium 2010; 47: 514–524. 45 Balijepalli RC, Foell JD, Hall DD, Hell JW, Kamp TJ. Localization of cardiac L-type Ca2+ channels to a caveolar macromolecular signaling complex is required for beta(2)-adrenergic regulation. Proc Natl Acad Sci USA 2006; 103: 7500–7505. 46 Luque RM, Gahete MD, Cordoba-Chacon J, Childs GV, Kineman RD. Does the pituitary somatotrope play a primary role in regulating GH output in metabolic extremes? Ann NY Acad Sci 2011; 1220: 82–92. 47 Perez-Castro C, Renner U, Haedo MR, Stalla GK, Arzt E. Cellular and molecular specificity of pituitary gland physiology. Physiol Rev 2012; 92: 1–38. 48 Melmed S. Idiopathic adult growth hormone deficiency. J Clin Endocrinol Metab 2013; 98: 2187–2197. 49 Sedej S, Tsujimoto T, Zorec R, Rupnik M. Voltage-activated Ca(2+) channels and their role in the endocrine function of the pituitary gland in newborn and adult mice. J Physiol 2004; 555: 769–782. 50 Fiordelisio T, Jimenez N, Baba S, Shiba K, Hernandez-Cruz A. Immunoreactivity to neurofilaments in the rodent anterior pituitary is associated with the expression of alpha 1A protein subunits of voltage-gated Ca2+ channels. J Neuroendocrinol 2007; 19: 870–881.

Supporting Information The following supplementary material is available: Fig. S1. Dimethylsulphoxide 0.05% does not affect growth hormone secretion. Fig. S2. Similar effects of Ca2+ channel blockers on ‘stimulated’ growth hormone secretion and on high-voltage activated Ba2+ influx. Fig. S3. The distribution Ca2+ channels among light and heavy gradient fractions after cholesterol depletion.

Journal of Neuroendocrinology, 2015, 27, 166–176