Isolation of Gamma-Amino Butyric Acid from Pig Hypothalami and Demonstration of its Prolactin Release-Inhibiting (PIF) Activity in Vivo and in Vitro* ANDREW V. SCHALLY, TOMMIE W. REDDING, AKIRA ARIMURA, ANDRE DUPONT,** AND GEORGE L. LINTHICUM The Veterans Administration Louisiana 70112
Hospital and Tulane University School of Medicine, New Orleans,
ABSTRACT. A non-retarded fraction with prolactinrelease inhibiting factor (PIF) activity obtained by chromatography of a concentrate of porcine hypothalami on carboxymethyl-cellulose was chromatographically distinct from catecholamines. This fraction was purified further by six steps involving chromatography on Sephadex G-25, countercurrent distribution, free-flow electrophoresis, and chromatography on triethylaminoethyl cellulose. The PIF-active substance was isolated and identified as •y-aminobutyric acid (GABA) by: 1) amino acid analyses using sodium as well as lithium-based buffers for resolution of biological fluids, 2) thin-layer chromatography of underivatized material as well as phenylthiocarbamyl derivatives, and 3) mass spectroscopy. Natural and synthetic GABA inhibited prolactin, but not LH release in vitro from isolated rat pituitary halves at doses as low as 0.1 jug/ml. The inhibition was proportional to the dose; natural and synthetic GABA possessed identical PIF activity.
G
AMMA-AMINO butyric acid (GABA) has been identified as a constituent of brain tissue in a variety of mammals as well as amphibia and birds (1,2). It was suggested that GABA may act as a neurotransmitter in synapses of brain tissue (2). Recently, it was reported that GABA may possibly be involved in the control of discharge of LHreleasing hormone (3), MSH-release inhibiting factor (MIF) (4), and ACTH (5). In our search for a prolactin release inhibiting factor (PIF), we have previously purified dopamine and noradrenaline from pig
Received March 30, 1976. * Supported by the Veterans Administration and USPHS Grants AM 07467 and AM 09094. ** Present address: Laboratoire d'Endocrinologie Moleculaire, Centre Hospitalier de l'Universite Laval, Quebec, Canada.
Synthetic GABA also decreased prolactin release in monolayer cultures of rat pituitary cells and inhibited TRH-stimulated prolactin release. The inhibition of prolactin release in vitro by GABA could not be blocked by perphenazine, which inhibits PIF activity of catecholamines. GABA also suppressed prolactin release in vivo, although large doses were needed. Either rapid iv injection or infusion of GABA in doses of 1 to 100 mg in rats significantly decreased serum prolactin levels, which were previously elevated by pretreatment with monoiodotyrosine perphenazine, chlorpromazine, haloperidol, or sulpiride. /3-hydroxy GABA significantly depressed prolactin release, but /3-(p-chlorophenyl)-GABA (Lioresal®, CIBA) and 4 other analogs of GABA were not effective in vivo and/or in vitro. The results indicate that GABA can inhibit prolactin release by a direct action on the pituitary gland, but whether this effect is physiologically meaningful still remains to be determined. (Endocrinology 100: 681, 1977)
hypothalami and described their inhibitory action on prolactin release in vivo and in vitro (6,7). We have also obtained a neutral fraction from chromatography on carboxymethyl cellulose (CMC) of a concentrate of P I F activity prepared from pig hypothalamic extracts by gel filtration on Sephadex G-25 and phenol extraction. This fraction had a low catecholamine content and its P I F activity in vitro was not suppressed by perphenazine, an a- and /8adrenergic blocking agent. After six more purification steps, the P I F active substance was isolated and identified as GABA. We have also determined that synthetic GABA and some of its analogs inhibit prolactin release in vivo and in vitro under a variety of conditions. This paper describes our findings. A portion of these studies has b e e n previously reported in abstract form (8,9). 68.1
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
SCHALLY ET AL.
682
Materials and Methods Purification methods The methodology used in the extraction of 470,000 pig hypothalami was described previously (6,10,11). The purification techniques: gel filtration on Sephadex, chromatography on carboxymethyl cellulose (CMC), countercurrent distribution (CCD), free-flow electrophoresis, chromatography on triethylaminoethyl cellulose, were also described previously (6,10,11). The separation pattern was followed by the FolinLowry reaction (12) and/or by absorbency readings at 278 nm. All yields and doses are expressed as dry weight. Dopamine and norepinephrine were estimated by fluorometric techniques (6,13). Composition and structural methods Thin layer chromatography (TLC) of underivatized preparations was carried out on plates coated with 0.1 mm layer of microcrystalline cellulose (Avicell, Brinkmann Instruments) using l-butanol:acetic acid:water (4:1:5). The spots were visualized by spraying with chlorine-o-tolidine reagent (14). GABA (British Drug Houses); y-aminobutyryl-alanine-hydrobromide, y-aminobutyryl-histidine (homocarnosine) sulfate; •y-amino-)3-riydroxybutyric acid; y-aminobutyryl-leucine hydrobromide, and yguanidino-2-butyric acid (Nutritional Biochemical s Co.) were run as standards and were also used for biological experiments. Thin layer chromatography (TLC) of phenylthiocarbamyl (PTC) derivatives (15) was carried out on silica gel plates (Brinkmann Instruments) or cellulose using ethanol:acetic acid:water (85:10:5, v/v/v) and detected as above (14). Two dimensional TLC of 5-dimethylaminonaphthalene sulfonyl (Dansyl)-derivatives of natural and synthetic GABA was performed on polyamide sheets (Cheng-Chin, distributed by Pierce Co., Rockford, 111.) using water:formic acid (200:3 vol/vol) and benzene:acetic acid (9:1 vol/vol) (15). Amino acid analyses were carried out on the Beckman Spinco model 119 automatic amino acid analyzer equipped with System AA computing integrator using single column methodology and sodium-based buffers (16). Physiological fluid amino acid analyses were performed using both sodium and lithium-based buffers according to the methods of Lee (17).
Endo • 1977 Vol 100 • No 3
High and low resolution mass spectral data were determined as described by Nair and Serially (18). Infrared spectra were run as mineral oil mulls between sodium chloride plates using the Perkin Elmer model 137 infrared spectrophotometer. Biological assay methods The PIF activity was followed mainly by an in vitro assay based on inhibition of prolactin release from rat pituitary halves, incubated in Krebs-Ringer bicarbonate medium containing 200 mg glucose per 100 ml (KRBG) and 0.25% albumin (19). Four beakers per sample were run in each incubation. The prolactin released into the medium during the second hour i.e., after the sample was added, was compared with that released during the first hour in the absence of added sample (20). The results are expressed as the per cent change of medium prolactin concentration in the second hour from that found after the first hour (18). The PIF activity was also measured in vitro using monolayer cultures of rat pituitary cells (21). In this case, prolactin was expressed as ng released per ml. The PIF activity of natural and synthetic GABA was also tested in vivo in rats pretreated with the following substances to elevate basal blood prolactin levels. i. Monoiodotyrosine (I Tyr). a. Female rats were anesthetized with urethane, and 100-200 mg/kg of I Tyr (22) was injected ip. Thirty minutes later, GABA or saline (0.5 ml iv) was injected. Blood samples were removed at 0, 30, and 60 min for serum prolactin determinations. To calculate the per cent change in prolactin levels, the values at 60 min were compared to those at 30 min. b. A jugular vein of female rats, under urethane anesthesia, was cannulated with polyethylene tubing (PE-50) attached to an infusion pump. Saline or GABA was infused at a rate of 37 fi\/ min. Thirty minutes after the start of infusion a blood sample was removed and I Tyr, 100 mg/kg, was administered ip. The infusion was continued and additional blood samples removed at 60, 90 and 120 min. c. GABA was also given orally, by gavage, 30 min before I Tyr (ip). For injections, I Tyr was suspended in saline and brought into solution by raising the pH with sodium hydroxide.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
683
PIF ACTIVITY OF GABA 2.0rFolin Color Change in Prolactin Levels
-i 50 FIG. 1. Gel filtration of CMC concentrate of PIF (6.4 g) on Sephadex G-25 fine (column 7.5 x 170 cm). Solvent 1M acetic acid. Peptide analyses performed on 50 fi\ aliquots. PIF activity was expressed as indicated in the footnote to Table 1.
1.0
60
t
70
§ 0.5
80
I
90 0
50
100 150 200 250 300 350 400
100
.s
Fraction Number X 25ml ii. Chlorpromazine. Male, Sprague-Dawley strain rats were used. a. In experiment 1, chlorpromazine, 10 mg/kg, and pentobarbital sodium (Nembutal®), 3.5 mg/ kg, were given ip 30 min and 15 min, respectively, before GABA (20 mg/rat in 0.5 ml 0.9% NaCl) was injected into the jugular vein. The control group received saline in place of GABA. Blood was drawn before and 30 min after GABA or injection of saline. b. In experiment 2, rats were treated with chlorpromazine and Nembutal as described above and GABA, 10 mg/rat, was injected into the jugular vein, and then 20 mg/1.1 ml saline was infused iv over a period of 30 min. The control group received only 0.9% NaCl. Hi. Perphenazine. Female rats were given GABA, 5 or 20 mg iv, and 5 min later perphenazine, 10 mg/kg, was given ip. Blood samples were taken 30 min after GABA.
(23), was administered ip in doses of 10 mg/kg to male rats. Fifteen minutes later Nembutal (45 mg/kg) was given. Thirty min after sulpiride (= 0 time), GABA (10 mg/rat) was injected iv followed by iv infusion of 20 mg GABA over 30 min. Blood was collected at 0,30, and 60 min. /3-(p-chlorophenyl)-y-aminobutyric acid (Lioresal®, CIBA) (24) was also tested for PIF activity. Four-to-six rats per group were used in all experiments. Serum or medium prolactin levels were determined by radioimmunoassay (RIA) as described by Niswender et al. (25) using the NIAMDD rat prolactin kit. The results were processed by a computer program described by Duddleston et al. (26). The mean response was calculated for each treatment group and the differences between control and experimental samples were examined by using Duncan's new multiple range test (27) or Student's t test.
Results iv. Haloperidol. Haloperidol was given iv in the dose of 0.1 mg/kg 1 min after administration of GABA to urethane-anesthetized normal female rats pretreated or not with 50 /xg 17 j8-estradiol 72 and 48 h before the experiment. Blood was taken 15 and 30 min after GABA.
Isolation studies
Early stages of this purification were described previously (6). Fractions no. 12571800 (Fig. 1, Ref. 6) from Sephadex containing PIF activity with Rf = 0.39-0.24 and v. Sulpiride. Sulpiride, N-[(ethyl-l-pyrrolidinyl- Ve/Vt= 0.93-1.39 (retarded) were com2)-methyl] - methoxy - 2 - sulfamoyl -5-benzamide bined and lyophilized, yielding 400 g from
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
684
Endo i 1977 Vol 100 , No 3
SCHALLY ET AL. Folin Color Change in Prolactin Levels
_n
50 60 70 80 control
200 300 400 600 510 1383 546
90 100
3
422
470,000 hypothalami. These fractions were extracted with phenol (9,10), yielding 70 g which was used for the next purification step, which consisted of chromatography on carboxymethylcellulose (CMC) columns. Batches of 5-to-7 g of phenol extracts (dry weight) were applied on CMC columns (60 x 2.6 cm); this procedure being repeated 12 times (see Fig. 2, Ref. 6). The highest PIF activity emerged in tubes no. 120-145, and was due to noradrenaline (NE) and dopamine (DA) (6). PIF activity was also detected in non-retarded fractions no. 10-32 and 33-64 when they were tested at doses of 200 fig and 20 fig, respectively. Fraction no. 33-64 was low in catecholamines (6) and its PIF activity was not blocked by per-
phenazine. This fraction (total weight 6.4 g) was repurified by gel filtration on a preparative column of Sephadex G-25. Figure 1 shows the pattern of separation. Main PIF activity emerged in tubes no. 206-234 (Rf = 0.39 - 0.48) whose dried contents weighed 1.6 g. Its activity again was not blocked by perphenazine. Since in the same experiment perphenazine (5 /u-g/ml) completely blocked the effect of 100 ng/ml of dopamine or noradrenaline, the PIF activity of the fractions no. 206-234 was probably not due to catecholamines. Moreover, the Rf of catecholamines in this system is 0.34, corresponding to tube no. 300. This material was then processed further by countercurrent distribution (CCD) in a
Folin Color
FIG. 3. Free flow electrophoresis (FFE) of 1.5 g of PIF from
Change in Prolactin Levels
16
I
Transfer Number
Cell Number
12
I
FIG. 2. Countercurrent distribution I of Sephadex concentrate of PIF (1.6 g) in a system of 0.1% acetic acid:l-butanol:pyridine (11:5:3, v/v/v) by the single withdrawal method. The volume of lower phase was 3 ml and of upper phase was 5 ml. Sixhundred transfers were performed in 400 cell train. FolinLowry analyses were carried out using 50 /JL\ of lower phase for cells 0-399, and 50 fil of upper phase for fractions with transfer no. 400-600 removed from the CCD train.
20
24
previous CCD I in 0.37M py-
28
32
ridine acetate buffer, pH 6.3. Conditions: 1900 volts, 160 mA, at 4 C, 11 h. Peptide analyses were carried out on 100 /ul aliquots. "Dosing," marked by an arrow, indicates the point at which the sample was infused into the separation chamber.
36
Fraction Number X 10 ml
Control
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
685
PIF ACTIVITY OF GABA
0.2
FlG. 4. Countercurrent distribution II of 1.3 g of PIF from FFE in a system of 1-butanol:acetic acid:water (4:1:5, v/v/v). Lower phase was 3 ml and upper phase 5 ml. The numbers of transfers was 900. Folin-Lowry analyses were done on 50 /xl aliquots of lower phase.
1
Folin Color
50
Change in Prolactin Levels
60
1
I
70
0.1
80
1
control
90
OS
100
O
I
100 3>
200 300 400
Cell Number
system consisting of 0.1% acetic acid:lbutanol:pyridine (11:5:3). The pattern of separation is shown in Fig. 2. The PIF activity was in fractions no. 100-219 with a mean K = 0.2, but not much purification resulted in the step, since the recovery was 1.5 g. The material from CCD was then subjected to free-flow electrophoresis (FFE). At pH 6.3, PIF activity migrated toward the anode (Fig. 3). Fractions no. 33-44 had all the PIF activity and yielded 1.3 g. The PIF-rich area from FFE was then purified further by CCD in a system consisting of 1-butanol:acetic acid:water (4:1:5) (Fig. 4). Nine-hundred transfers were used. Most PIF activity appeared in cells no. 170-319 (mean K = 0.22), which yielded 0.898 g of material. 07
FIG. 5. Chromatography of PIF from CCD II (0.90 g) on TEAEcellulose. Column 1.8 x 100 cm,
^ 1 Ci/mmol; New
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
687
PIF ACTIVITY OF GABA England Nuclear) was mixed with 15 mg of unlabelled GABA and subjected to gel filtration on the same column of Sephadex G-25 used for the experiment shown in Fig. 6; it behaved identically to natural GABA. Biological activity of natural material, synthetic GABA, and its analogs i. In vitro. When natural GABA (Fraction no. 55-58, Fig. 6) was compared to synthetic GABA using rat isolated pituitary halves (Table 1, Exp. 1) the activity of the former was 98.6% of the latter. The doseresponse curves were parallel. In the next experiment, Lioresal® [/3-(p-chlorophenyl)GABA] and other analogs of GABA were TABLE 1. Effect of highly purified natural fractions with PIF activity (Fig. 6), synthetic GABA and its analogs on prolactin release from rat pituitaries in vitro, and lack of effect of perphenazine on the inhibition by GABA of prolactin release
Sample Experiment 1 Control Fractions no. 55-58 Fractions no. 55-58 Fractions no. 55-58 Synthetic GABA Synthetic GABA Synthetic GABA
Dose (/xg/ml)
Medium prolactin % of control (2nd Incubation/ 1st Incubation X 100)*
P os. Control
0.1 0.5 2.5 0.1 0.5 2.5
108.4 ± 5.0 91.7 ± 1.3 80.2 ± 0.7 73.7 ± 1.2 92.7 ± 3.4 79.0 ± 4.3 72.8 ± 6.3
Hospital and Tulane University School of Medicine, New Orleans,
ABSTRACT. A non-retarded fraction with prolactinrelease inhibiting factor (PIF) activity obtained by chromatography of a concentrate of porcine hypothalami on carboxymethyl-cellulose was chromatographically distinct from catecholamines. This fraction was purified further by six steps involving chromatography on Sephadex G-25, countercurrent distribution, free-flow electrophoresis, and chromatography on triethylaminoethyl cellulose. The PIF-active substance was isolated and identified as •y-aminobutyric acid (GABA) by: 1) amino acid analyses using sodium as well as lithium-based buffers for resolution of biological fluids, 2) thin-layer chromatography of underivatized material as well as phenylthiocarbamyl derivatives, and 3) mass spectroscopy. Natural and synthetic GABA inhibited prolactin, but not LH release in vitro from isolated rat pituitary halves at doses as low as 0.1 jug/ml. The inhibition was proportional to the dose; natural and synthetic GABA possessed identical PIF activity.
G
AMMA-AMINO butyric acid (GABA) has been identified as a constituent of brain tissue in a variety of mammals as well as amphibia and birds (1,2). It was suggested that GABA may act as a neurotransmitter in synapses of brain tissue (2). Recently, it was reported that GABA may possibly be involved in the control of discharge of LHreleasing hormone (3), MSH-release inhibiting factor (MIF) (4), and ACTH (5). In our search for a prolactin release inhibiting factor (PIF), we have previously purified dopamine and noradrenaline from pig
Received March 30, 1976. * Supported by the Veterans Administration and USPHS Grants AM 07467 and AM 09094. ** Present address: Laboratoire d'Endocrinologie Moleculaire, Centre Hospitalier de l'Universite Laval, Quebec, Canada.
Synthetic GABA also decreased prolactin release in monolayer cultures of rat pituitary cells and inhibited TRH-stimulated prolactin release. The inhibition of prolactin release in vitro by GABA could not be blocked by perphenazine, which inhibits PIF activity of catecholamines. GABA also suppressed prolactin release in vivo, although large doses were needed. Either rapid iv injection or infusion of GABA in doses of 1 to 100 mg in rats significantly decreased serum prolactin levels, which were previously elevated by pretreatment with monoiodotyrosine perphenazine, chlorpromazine, haloperidol, or sulpiride. /3-hydroxy GABA significantly depressed prolactin release, but /3-(p-chlorophenyl)-GABA (Lioresal®, CIBA) and 4 other analogs of GABA were not effective in vivo and/or in vitro. The results indicate that GABA can inhibit prolactin release by a direct action on the pituitary gland, but whether this effect is physiologically meaningful still remains to be determined. (Endocrinology 100: 681, 1977)
hypothalami and described their inhibitory action on prolactin release in vivo and in vitro (6,7). We have also obtained a neutral fraction from chromatography on carboxymethyl cellulose (CMC) of a concentrate of P I F activity prepared from pig hypothalamic extracts by gel filtration on Sephadex G-25 and phenol extraction. This fraction had a low catecholamine content and its P I F activity in vitro was not suppressed by perphenazine, an a- and /8adrenergic blocking agent. After six more purification steps, the P I F active substance was isolated and identified as GABA. We have also determined that synthetic GABA and some of its analogs inhibit prolactin release in vivo and in vitro under a variety of conditions. This paper describes our findings. A portion of these studies has b e e n previously reported in abstract form (8,9). 68.1
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
SCHALLY ET AL.
682
Materials and Methods Purification methods The methodology used in the extraction of 470,000 pig hypothalami was described previously (6,10,11). The purification techniques: gel filtration on Sephadex, chromatography on carboxymethyl cellulose (CMC), countercurrent distribution (CCD), free-flow electrophoresis, chromatography on triethylaminoethyl cellulose, were also described previously (6,10,11). The separation pattern was followed by the FolinLowry reaction (12) and/or by absorbency readings at 278 nm. All yields and doses are expressed as dry weight. Dopamine and norepinephrine were estimated by fluorometric techniques (6,13). Composition and structural methods Thin layer chromatography (TLC) of underivatized preparations was carried out on plates coated with 0.1 mm layer of microcrystalline cellulose (Avicell, Brinkmann Instruments) using l-butanol:acetic acid:water (4:1:5). The spots were visualized by spraying with chlorine-o-tolidine reagent (14). GABA (British Drug Houses); y-aminobutyryl-alanine-hydrobromide, y-aminobutyryl-histidine (homocarnosine) sulfate; •y-amino-)3-riydroxybutyric acid; y-aminobutyryl-leucine hydrobromide, and yguanidino-2-butyric acid (Nutritional Biochemical s Co.) were run as standards and were also used for biological experiments. Thin layer chromatography (TLC) of phenylthiocarbamyl (PTC) derivatives (15) was carried out on silica gel plates (Brinkmann Instruments) or cellulose using ethanol:acetic acid:water (85:10:5, v/v/v) and detected as above (14). Two dimensional TLC of 5-dimethylaminonaphthalene sulfonyl (Dansyl)-derivatives of natural and synthetic GABA was performed on polyamide sheets (Cheng-Chin, distributed by Pierce Co., Rockford, 111.) using water:formic acid (200:3 vol/vol) and benzene:acetic acid (9:1 vol/vol) (15). Amino acid analyses were carried out on the Beckman Spinco model 119 automatic amino acid analyzer equipped with System AA computing integrator using single column methodology and sodium-based buffers (16). Physiological fluid amino acid analyses were performed using both sodium and lithium-based buffers according to the methods of Lee (17).
Endo • 1977 Vol 100 • No 3
High and low resolution mass spectral data were determined as described by Nair and Serially (18). Infrared spectra were run as mineral oil mulls between sodium chloride plates using the Perkin Elmer model 137 infrared spectrophotometer. Biological assay methods The PIF activity was followed mainly by an in vitro assay based on inhibition of prolactin release from rat pituitary halves, incubated in Krebs-Ringer bicarbonate medium containing 200 mg glucose per 100 ml (KRBG) and 0.25% albumin (19). Four beakers per sample were run in each incubation. The prolactin released into the medium during the second hour i.e., after the sample was added, was compared with that released during the first hour in the absence of added sample (20). The results are expressed as the per cent change of medium prolactin concentration in the second hour from that found after the first hour (18). The PIF activity was also measured in vitro using monolayer cultures of rat pituitary cells (21). In this case, prolactin was expressed as ng released per ml. The PIF activity of natural and synthetic GABA was also tested in vivo in rats pretreated with the following substances to elevate basal blood prolactin levels. i. Monoiodotyrosine (I Tyr). a. Female rats were anesthetized with urethane, and 100-200 mg/kg of I Tyr (22) was injected ip. Thirty minutes later, GABA or saline (0.5 ml iv) was injected. Blood samples were removed at 0, 30, and 60 min for serum prolactin determinations. To calculate the per cent change in prolactin levels, the values at 60 min were compared to those at 30 min. b. A jugular vein of female rats, under urethane anesthesia, was cannulated with polyethylene tubing (PE-50) attached to an infusion pump. Saline or GABA was infused at a rate of 37 fi\/ min. Thirty minutes after the start of infusion a blood sample was removed and I Tyr, 100 mg/kg, was administered ip. The infusion was continued and additional blood samples removed at 60, 90 and 120 min. c. GABA was also given orally, by gavage, 30 min before I Tyr (ip). For injections, I Tyr was suspended in saline and brought into solution by raising the pH with sodium hydroxide.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
683
PIF ACTIVITY OF GABA 2.0rFolin Color Change in Prolactin Levels
-i 50 FIG. 1. Gel filtration of CMC concentrate of PIF (6.4 g) on Sephadex G-25 fine (column 7.5 x 170 cm). Solvent 1M acetic acid. Peptide analyses performed on 50 fi\ aliquots. PIF activity was expressed as indicated in the footnote to Table 1.
1.0
60
t
70
§ 0.5
80
I
90 0
50
100 150 200 250 300 350 400
100
.s
Fraction Number X 25ml ii. Chlorpromazine. Male, Sprague-Dawley strain rats were used. a. In experiment 1, chlorpromazine, 10 mg/kg, and pentobarbital sodium (Nembutal®), 3.5 mg/ kg, were given ip 30 min and 15 min, respectively, before GABA (20 mg/rat in 0.5 ml 0.9% NaCl) was injected into the jugular vein. The control group received saline in place of GABA. Blood was drawn before and 30 min after GABA or injection of saline. b. In experiment 2, rats were treated with chlorpromazine and Nembutal as described above and GABA, 10 mg/rat, was injected into the jugular vein, and then 20 mg/1.1 ml saline was infused iv over a period of 30 min. The control group received only 0.9% NaCl. Hi. Perphenazine. Female rats were given GABA, 5 or 20 mg iv, and 5 min later perphenazine, 10 mg/kg, was given ip. Blood samples were taken 30 min after GABA.
(23), was administered ip in doses of 10 mg/kg to male rats. Fifteen minutes later Nembutal (45 mg/kg) was given. Thirty min after sulpiride (= 0 time), GABA (10 mg/rat) was injected iv followed by iv infusion of 20 mg GABA over 30 min. Blood was collected at 0,30, and 60 min. /3-(p-chlorophenyl)-y-aminobutyric acid (Lioresal®, CIBA) (24) was also tested for PIF activity. Four-to-six rats per group were used in all experiments. Serum or medium prolactin levels were determined by radioimmunoassay (RIA) as described by Niswender et al. (25) using the NIAMDD rat prolactin kit. The results were processed by a computer program described by Duddleston et al. (26). The mean response was calculated for each treatment group and the differences between control and experimental samples were examined by using Duncan's new multiple range test (27) or Student's t test.
Results iv. Haloperidol. Haloperidol was given iv in the dose of 0.1 mg/kg 1 min after administration of GABA to urethane-anesthetized normal female rats pretreated or not with 50 /xg 17 j8-estradiol 72 and 48 h before the experiment. Blood was taken 15 and 30 min after GABA.
Isolation studies
Early stages of this purification were described previously (6). Fractions no. 12571800 (Fig. 1, Ref. 6) from Sephadex containing PIF activity with Rf = 0.39-0.24 and v. Sulpiride. Sulpiride, N-[(ethyl-l-pyrrolidinyl- Ve/Vt= 0.93-1.39 (retarded) were com2)-methyl] - methoxy - 2 - sulfamoyl -5-benzamide bined and lyophilized, yielding 400 g from
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
684
Endo i 1977 Vol 100 , No 3
SCHALLY ET AL. Folin Color Change in Prolactin Levels
_n
50 60 70 80 control
200 300 400 600 510 1383 546
90 100
3
422
470,000 hypothalami. These fractions were extracted with phenol (9,10), yielding 70 g which was used for the next purification step, which consisted of chromatography on carboxymethylcellulose (CMC) columns. Batches of 5-to-7 g of phenol extracts (dry weight) were applied on CMC columns (60 x 2.6 cm); this procedure being repeated 12 times (see Fig. 2, Ref. 6). The highest PIF activity emerged in tubes no. 120-145, and was due to noradrenaline (NE) and dopamine (DA) (6). PIF activity was also detected in non-retarded fractions no. 10-32 and 33-64 when they were tested at doses of 200 fig and 20 fig, respectively. Fraction no. 33-64 was low in catecholamines (6) and its PIF activity was not blocked by per-
phenazine. This fraction (total weight 6.4 g) was repurified by gel filtration on a preparative column of Sephadex G-25. Figure 1 shows the pattern of separation. Main PIF activity emerged in tubes no. 206-234 (Rf = 0.39 - 0.48) whose dried contents weighed 1.6 g. Its activity again was not blocked by perphenazine. Since in the same experiment perphenazine (5 /u-g/ml) completely blocked the effect of 100 ng/ml of dopamine or noradrenaline, the PIF activity of the fractions no. 206-234 was probably not due to catecholamines. Moreover, the Rf of catecholamines in this system is 0.34, corresponding to tube no. 300. This material was then processed further by countercurrent distribution (CCD) in a
Folin Color
FIG. 3. Free flow electrophoresis (FFE) of 1.5 g of PIF from
Change in Prolactin Levels
16
I
Transfer Number
Cell Number
12
I
FIG. 2. Countercurrent distribution I of Sephadex concentrate of PIF (1.6 g) in a system of 0.1% acetic acid:l-butanol:pyridine (11:5:3, v/v/v) by the single withdrawal method. The volume of lower phase was 3 ml and of upper phase was 5 ml. Sixhundred transfers were performed in 400 cell train. FolinLowry analyses were carried out using 50 /JL\ of lower phase for cells 0-399, and 50 fil of upper phase for fractions with transfer no. 400-600 removed from the CCD train.
20
24
previous CCD I in 0.37M py-
28
32
ridine acetate buffer, pH 6.3. Conditions: 1900 volts, 160 mA, at 4 C, 11 h. Peptide analyses were carried out on 100 /ul aliquots. "Dosing," marked by an arrow, indicates the point at which the sample was infused into the separation chamber.
36
Fraction Number X 10 ml
Control
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
685
PIF ACTIVITY OF GABA
0.2
FlG. 4. Countercurrent distribution II of 1.3 g of PIF from FFE in a system of 1-butanol:acetic acid:water (4:1:5, v/v/v). Lower phase was 3 ml and upper phase 5 ml. The numbers of transfers was 900. Folin-Lowry analyses were done on 50 /xl aliquots of lower phase.
1
Folin Color
50
Change in Prolactin Levels
60
1
I
70
0.1
80
1
control
90
OS
100
O
I
100 3>
200 300 400
Cell Number
system consisting of 0.1% acetic acid:lbutanol:pyridine (11:5:3). The pattern of separation is shown in Fig. 2. The PIF activity was in fractions no. 100-219 with a mean K = 0.2, but not much purification resulted in the step, since the recovery was 1.5 g. The material from CCD was then subjected to free-flow electrophoresis (FFE). At pH 6.3, PIF activity migrated toward the anode (Fig. 3). Fractions no. 33-44 had all the PIF activity and yielded 1.3 g. The PIF-rich area from FFE was then purified further by CCD in a system consisting of 1-butanol:acetic acid:water (4:1:5) (Fig. 4). Nine-hundred transfers were used. Most PIF activity appeared in cells no. 170-319 (mean K = 0.22), which yielded 0.898 g of material. 07
FIG. 5. Chromatography of PIF from CCD II (0.90 g) on TEAEcellulose. Column 1.8 x 100 cm,
^ 1 Ci/mmol; New
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:40 For personal use only. No other uses without permission. . All rights reserved.
687
PIF ACTIVITY OF GABA England Nuclear) was mixed with 15 mg of unlabelled GABA and subjected to gel filtration on the same column of Sephadex G-25 used for the experiment shown in Fig. 6; it behaved identically to natural GABA. Biological activity of natural material, synthetic GABA, and its analogs i. In vitro. When natural GABA (Fraction no. 55-58, Fig. 6) was compared to synthetic GABA using rat isolated pituitary halves (Table 1, Exp. 1) the activity of the former was 98.6% of the latter. The doseresponse curves were parallel. In the next experiment, Lioresal® [/3-(p-chlorophenyl)GABA] and other analogs of GABA were TABLE 1. Effect of highly purified natural fractions with PIF activity (Fig. 6), synthetic GABA and its analogs on prolactin release from rat pituitaries in vitro, and lack of effect of perphenazine on the inhibition by GABA of prolactin release
Sample Experiment 1 Control Fractions no. 55-58 Fractions no. 55-58 Fractions no. 55-58 Synthetic GABA Synthetic GABA Synthetic GABA
Dose (/xg/ml)
Medium prolactin % of control (2nd Incubation/ 1st Incubation X 100)*
P os. Control
0.1 0.5 2.5 0.1 0.5 2.5
108.4 ± 5.0 91.7 ± 1.3 80.2 ± 0.7 73.7 ± 1.2 92.7 ± 3.4 79.0 ± 4.3 72.8 ± 6.3
