Brain Research, 506 (1990) 129-132 Elsevier
129
BRES 15095
Characterization of prolactin immunoreactivity in human cerebrospinal fluid Fred Nyberg l, Carl-A.ke Isacson 3, Erika Brostedt 2 and Paul
Roos 2
1Department of Pharmacology and 21nstitute of Biochemistry, University of Uppsala, Uppsala (Sweden) and -~AB KABI Diagnostica, NykOping (Sweden) (Accepted 6 June 1989)
Key words: Cerebrospinal fluid; Characterization; Human; Immunoreactive prolactin
Immunoreactive prolactin has been recovered from human cerebrospinal fluid (CSF) by hydrophobic interaction chromatography on Phenyl-Sepharose CL-4B and molecular sieving on Sephadex G-100. The hormone was obtained in a yield of about 1 ~g/liter CSF and was further analyzed by column electrophoresis on granular agarose at different pHs. All separation procedures were guided by radioimmunoassay. Upon chromatography on Sephadex G-100, the CSF prolactin coeluted with a pituitary reference preparation (molecular weight, 22.000 Da). A further relationship between the CSF prolactin and the pituitary hormone was demonstrated by the electrophoretic experiments. Thus, at alkaline pH, both preparations resolved into 3 or 4 active components, whereas only a single peak was observed in each preparation following runs performed at acidic pH. The results suggest that the prolactin activity present in the CSF specimens is identical with the hormone derived from the pituitary.
INTRODUCTION Evidence suggesting that prolactin acts directly on the brain in addition to its peripheral effects on structures, such as the gonads and m a m m a r y glands, has accumulated during the last decade. Many studies on the activity of prolactin in the central nervous system (CNS) have been probed by analysis of hormone content in the cerebrospinal fluid (CSF). Since this fluid is in direct contact with the CNS, its peptide levels may more adequately than e.g. those of plasma be functionally related to neurohormonal activity. Therefore, it is important to define how the CSF prolactin is related to its plasma analogue. In the first place, it is essential to establish the chemical nature of the h o r m o n e in the two compartments. The chemical characteristics of human prolactin in the pituitary and in various body fluids has been extensively investigated during the last decade. The major part of the hormone activity (present in the pituitary) has been characterized as a polypeptide of molecular weight 22,000 Da ( m o n o m e r prolactin) 5'9'19. It was found to be homogeneous in the ultracentrifuge ~9 but showed electrophoretic heterogenity when analyzed at alkaline pH 5'9'14'19. Similar properties have been reported for a m o n o m e r fraction of prolactin in amniotic fluid 3'1s, plasma 22 and pregnancy urine 15. The presence of prolactin in human cerebrospinal fluid
(CSF) has been confirmed in many laboratories ~'4' 6,8,~o,1~. The CSF levels of prolactin in normal subjects is about 1/~g/liter I. Patients with prolactin-secreting hypophyseal tumor have elevated concentrations of prolactin both in plasma and CSF 7'21, Under physiological conditions, the correlation between plasma and CSF prolactin has been reported to be linear I. The demonstration that prolactin levels in CSF rise, following intravenous prolactin administration in the rhesus monkeys 2, or when endogenous prolactin secretion is stimulated by dopamine antagonists ~2, clearly indicates that prolactin can enter CSF from the vascular compartment. It is thus likely that the h o r m o n e activity detected in the CSF originally derives from the pituitary. This study investigates the chemical nature of the prolactin immunoreactivity present in the fluid. The availability of large quantities of CSF allowed a partial purification of the h o r m o n e activity and subsequent analysis by a recently developed electrophoresis-radioimmunoassay procedure. MATERIALS AND METHODS The CSF was obtained by lumbar puncture from neurologic patients undergoing investigation for suspected intracranial pressure. All samples used in the study had essentially normal protein content. The chromatographic material (DEAE-Sepharose CL-6B, phenyl-Sepharose CL-4B, Sephadex G-100 and Agarose C) was
Correspondence: F. Nyberg, Department of Pharmacology, University of Uppsala, P.O. Box 591, S-751 24 Uppsala, Sweden.
130 purchased from Pharmacia Fine Chemicals, Uppsala, Sweden. All other chemicals and solvents were of reagent grade from the usual
0.2
SOUrCeS.
The CSF material (about 600 ml from 15 different subjects) was thawed and directly fractionated through a tandem-column system of DEAE-Sepharose CL 6B (3.2 × 8 cm) and phenyl-Sepharose CL-4B (2 × 10 cm), previously equilibrated with artificial CSF t6. After washing with two volumes of artificial CSF, the columns were dismantled and the prolactin activity, adsorbed to the phenylSepharose column, was eluted with redistilled water containing 40% acetonitrile. The eluate which was received in a volume of 30 ml was subsequently fractionated on a Sephadex G-100 column (5 × 90 cm) using 0.04 M NH4HCO 3 as eluent buffer. Fractions of 20 ml were collected at a flow rate of 80 ml/h and aliquots were withdrawn and assayed for prolactin activity. The active fractions were lyophilized and analyzed by electrophoresis. Electrophoresis was performed in columns of agarose suspension at pH 8.6 as described earlier ~4-15 and at pH 2.7 as in Nyberg and Tereniu¢ 3. After completed runs, fractions were withdrawn and the agarose removed by centrifugation. The distribution in the eluate of prolactin activity was determined by radioimmunoassay. Analysis of prolactin activity in individual CSF sample was carried out following a preseparation on disposable Sephadex G-25 columns (PD-10, Pharmacia Fine Chemicals). The prolactin fraction was lyophilized and reconstituted in 100 /~1 redistilled water prior to assay. The radioimmunoassay utilized kits for prolactin determinations (Radormon) supplied by KABI Diagnostica, Stockholm, Sweden. All samples were assayed in duplicate. The standard curve was linear from 5 to 100 ng/ml and 50% inhibition of tracer binding was achieved with 30 ng/ml.
®
33
0.1 In" I'¢,9
t
E
i
I
60
4O
20 c 0.2
o
®
®
0.1
I 0
J
~ , J
20 fraction
40
I 60
number
Fig. 2. Electrophoresis in agarose suspension (0.18%, w/v) at alkaline pH of CSF (upper panel) and pituitary (lower panel) prolactin. Buffer, 0.03 M sodium veronal (pH 8.6); voltage; 820 V; duration, 18 h. The prolactin activity was determined by radioimmunoassay.
RESULTS
The prolactin activity in the crude CSF material was present in the break-through volume from the ion exchanger and in similarity with the pituitary preparation 19 initially adsorbed to the phenyl-Sepharose column. In the subsequent Sephadex G-100 fractionation
0,4
®
®
0.2 VOID
NaCl
125 I-PRL
HSA
!
150
E
.E
0 0~ 1 0 0
m
20
50
I 60
0.4
o Q. I y-
o
I 40
(9
0.2
®
~. n.
129
BRES 15095
Characterization of prolactin immunoreactivity in human cerebrospinal fluid Fred Nyberg l, Carl-A.ke Isacson 3, Erika Brostedt 2 and Paul
Roos 2
1Department of Pharmacology and 21nstitute of Biochemistry, University of Uppsala, Uppsala (Sweden) and -~AB KABI Diagnostica, NykOping (Sweden) (Accepted 6 June 1989)
Key words: Cerebrospinal fluid; Characterization; Human; Immunoreactive prolactin
Immunoreactive prolactin has been recovered from human cerebrospinal fluid (CSF) by hydrophobic interaction chromatography on Phenyl-Sepharose CL-4B and molecular sieving on Sephadex G-100. The hormone was obtained in a yield of about 1 ~g/liter CSF and was further analyzed by column electrophoresis on granular agarose at different pHs. All separation procedures were guided by radioimmunoassay. Upon chromatography on Sephadex G-100, the CSF prolactin coeluted with a pituitary reference preparation (molecular weight, 22.000 Da). A further relationship between the CSF prolactin and the pituitary hormone was demonstrated by the electrophoretic experiments. Thus, at alkaline pH, both preparations resolved into 3 or 4 active components, whereas only a single peak was observed in each preparation following runs performed at acidic pH. The results suggest that the prolactin activity present in the CSF specimens is identical with the hormone derived from the pituitary.
INTRODUCTION Evidence suggesting that prolactin acts directly on the brain in addition to its peripheral effects on structures, such as the gonads and m a m m a r y glands, has accumulated during the last decade. Many studies on the activity of prolactin in the central nervous system (CNS) have been probed by analysis of hormone content in the cerebrospinal fluid (CSF). Since this fluid is in direct contact with the CNS, its peptide levels may more adequately than e.g. those of plasma be functionally related to neurohormonal activity. Therefore, it is important to define how the CSF prolactin is related to its plasma analogue. In the first place, it is essential to establish the chemical nature of the h o r m o n e in the two compartments. The chemical characteristics of human prolactin in the pituitary and in various body fluids has been extensively investigated during the last decade. The major part of the hormone activity (present in the pituitary) has been characterized as a polypeptide of molecular weight 22,000 Da ( m o n o m e r prolactin) 5'9'19. It was found to be homogeneous in the ultracentrifuge ~9 but showed electrophoretic heterogenity when analyzed at alkaline pH 5'9'14'19. Similar properties have been reported for a m o n o m e r fraction of prolactin in amniotic fluid 3'1s, plasma 22 and pregnancy urine 15. The presence of prolactin in human cerebrospinal fluid
(CSF) has been confirmed in many laboratories ~'4' 6,8,~o,1~. The CSF levels of prolactin in normal subjects is about 1/~g/liter I. Patients with prolactin-secreting hypophyseal tumor have elevated concentrations of prolactin both in plasma and CSF 7'21, Under physiological conditions, the correlation between plasma and CSF prolactin has been reported to be linear I. The demonstration that prolactin levels in CSF rise, following intravenous prolactin administration in the rhesus monkeys 2, or when endogenous prolactin secretion is stimulated by dopamine antagonists ~2, clearly indicates that prolactin can enter CSF from the vascular compartment. It is thus likely that the h o r m o n e activity detected in the CSF originally derives from the pituitary. This study investigates the chemical nature of the prolactin immunoreactivity present in the fluid. The availability of large quantities of CSF allowed a partial purification of the h o r m o n e activity and subsequent analysis by a recently developed electrophoresis-radioimmunoassay procedure. MATERIALS AND METHODS The CSF was obtained by lumbar puncture from neurologic patients undergoing investigation for suspected intracranial pressure. All samples used in the study had essentially normal protein content. The chromatographic material (DEAE-Sepharose CL-6B, phenyl-Sepharose CL-4B, Sephadex G-100 and Agarose C) was
Correspondence: F. Nyberg, Department of Pharmacology, University of Uppsala, P.O. Box 591, S-751 24 Uppsala, Sweden.
130 purchased from Pharmacia Fine Chemicals, Uppsala, Sweden. All other chemicals and solvents were of reagent grade from the usual
0.2
SOUrCeS.
The CSF material (about 600 ml from 15 different subjects) was thawed and directly fractionated through a tandem-column system of DEAE-Sepharose CL 6B (3.2 × 8 cm) and phenyl-Sepharose CL-4B (2 × 10 cm), previously equilibrated with artificial CSF t6. After washing with two volumes of artificial CSF, the columns were dismantled and the prolactin activity, adsorbed to the phenylSepharose column, was eluted with redistilled water containing 40% acetonitrile. The eluate which was received in a volume of 30 ml was subsequently fractionated on a Sephadex G-100 column (5 × 90 cm) using 0.04 M NH4HCO 3 as eluent buffer. Fractions of 20 ml were collected at a flow rate of 80 ml/h and aliquots were withdrawn and assayed for prolactin activity. The active fractions were lyophilized and analyzed by electrophoresis. Electrophoresis was performed in columns of agarose suspension at pH 8.6 as described earlier ~4-15 and at pH 2.7 as in Nyberg and Tereniu¢ 3. After completed runs, fractions were withdrawn and the agarose removed by centrifugation. The distribution in the eluate of prolactin activity was determined by radioimmunoassay. Analysis of prolactin activity in individual CSF sample was carried out following a preseparation on disposable Sephadex G-25 columns (PD-10, Pharmacia Fine Chemicals). The prolactin fraction was lyophilized and reconstituted in 100 /~1 redistilled water prior to assay. The radioimmunoassay utilized kits for prolactin determinations (Radormon) supplied by KABI Diagnostica, Stockholm, Sweden. All samples were assayed in duplicate. The standard curve was linear from 5 to 100 ng/ml and 50% inhibition of tracer binding was achieved with 30 ng/ml.
®
33
0.1 In" I'¢,9
t
E
i
I
60
4O
20 c 0.2
o
®
®
0.1
I 0
J
~ , J
20 fraction
40
I 60
number
Fig. 2. Electrophoresis in agarose suspension (0.18%, w/v) at alkaline pH of CSF (upper panel) and pituitary (lower panel) prolactin. Buffer, 0.03 M sodium veronal (pH 8.6); voltage; 820 V; duration, 18 h. The prolactin activity was determined by radioimmunoassay.
RESULTS
The prolactin activity in the crude CSF material was present in the break-through volume from the ion exchanger and in similarity with the pituitary preparation 19 initially adsorbed to the phenyl-Sepharose column. In the subsequent Sephadex G-100 fractionation
0,4
®
®
0.2 VOID
NaCl
125 I-PRL
HSA
!
150
E
.E
0 0~ 1 0 0
m
20
50
I 60
0.4
o Q. I y-
o
I 40
(9
0.2
®
~. n.
