Brain Research, 582 (1992) 27-37 (~) 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.00
27
BRES 17769
Developmental regulation of insulin in the mammalian central nervous system Ruben Schechter a, Jennifer Whitmire a, Lynne Holtzclaw b, Mark George b, Robert Harlow a and Sherin U. Devaskar b aSt. Franscis Hospital of Tulsa Medical Research Institute, University of Oklahoma Health Science Center and the Department of Neonatology, St. Francis Hospital, Tulsa, OK (USA) and bDivision of Neonatology, Department of Pediatrics, St. Louis University School of Medicine, The Pediatric Research Institute and Cardinal Glennon Children's Hospital, St. Louis, MO (USA) (Accepted 7 January 1992) Key words." Development; Rabbit; Brain; Insulin; Cerebrospinal fluid
We delineated the ontogeny of rabbit brain insulin concentrations to define the regulatory role of development on this hormone in the central nervous system. Employing a sensitive ELISA, we observed higher concentrations in the late gestation fetal brain (-80-90 ng/g) and early neonatal brain (-195 ng/g) in comparison to the adult (-32 ng/g; P < 0.01). Further, we characterized this hormone to determine the identity of insulin (or an insulin-like substance) in brain. Employing porcine/bovine or rabbit insulin as standards, we observed that brain insulin mimicked authentic insulin in its migration on SDS-polyacrytamide and native gel electrophoresis, immunogenicity on Western blot analysis, and its elution profile on immunoaffinity column chromatographic, and high performance liquid chromatographic separation. We then examined the developmental effects on circulating and cerebrospinal fluid (CSF) radioimmunoassayable insulin levels. No statistically significant differences (ANOVA) existed through development in either the serum or CSF insulin levels. Employing multiple regression analysis, no correlation was evident between brain and either serum or CSF insulin concentration. A search for insulin mRNA by Northern blot analysis yielded minute amounts of atypical large sized transcripts. We conclude that the insulin peptide in the central nervous system closely resembles (or is identical to) circulating insulin in many properties and that there is a developmental increase in brain insulin concentrations, the maximal peak occuring in the late gestation fetus and early neonate. Insulin concentrations in brain demonstrate no conventional relationship to either the serum or CSF insulin levels, suggesting an additional source of peptide, which contributes beyond that which is available via the circulation. The amounts of insulin present within the central nervous system are minute (difficult to detect) but in the range (10-100 ng) where the hormone can interact with either insulin or insulin-like growth factor I (IGF-I) receptors that are abundantly present on developing brain cells, thereby executing the biological function of the hormone. INTRODUCTION There has been an ongoing controversy regarding the origin of insulin in the central nervous system (CNS) since the initial detection of the h o r m o n e in adult rat brain 29. While various investigations u n d e r t a k e n both in vitro 8'13'33'43'44and in vivo 47'51'52'57point towards the fact that an insulin-like peptide is synthesized de novo within brain cells, there are other studies which support the contention that insulin traverses the b l o o d - b r a i n barrier and gains access to the brain 4'53'54. Circulating insulin u n d e r conditions of hyperinsulinemia, has been observed to cross the b l o o d - b r a i n barrier and enter the cerebrospinal fluid (CSF), mainly in the adult h u m a n 53 and various other animal species 4'52, while access to the brain parenchyma has unequivocally been demonstrated in the neonatal rabbit alone 23. Further, there appear to be major functional differences exerted by insulin in the CNS based on the developmental stage of the brain. For ex-
ample, most of the biological effects relating to brain cellular growth and differentiation 1'2'4°, potentiation of brain glucose 11 and macromolecular metabolism 12 have been observed in the immature brain cells, whereas a hormonal regulatory effect on the whole animal's feeding behavior has been observed in the adult brain alone 26'55. Insulin's biological effect on brain cell growth promotion and differentiation is more global in comparison to its effect on feeding patterns. The former function involves most neurons and glial cells 1'2'40'41 that are widespread within all structures of the brain, whereas the latter constitutes a specialized function that is generally localized to discrete, perhaps small, areas of the brain 39. Previously it has been reported, in rats, mice and rabbits, that fetal concentrations of circulating insulin are higher than the early neonatal and adult values 5' ~5,~8. Although fetal hyperinsulinemia has not demonstrated a gross increase in fetal brain weight 32, insulin is a fetal tissue growth promoter 16, resulting in generalized
Correspondence: S.U. Devaskar, 1465 South Grand Blvd., St. Louis, MO 63104, USA.
28 organomegaly 32"5°. There have been no systematic studies quantitating brain insulin concentrations during these critical fetal and neonatal phases of CNS development 22' 31. Further, there have been no attempts to determine the relationship between brain insulin levels and the circulating or CSF insulin concentrations during normal development. We hypothesized that in the presence of higher circulating fetal insulin concentrations and an increased need for a global brain cell growth potentiating effect, increasing amounts of insulin will cross the bloodbrain barrier and gain access to the immature brain when compared to the adult. Thus we undertook the present investigation and delineated the ontogeny of brain insulin content, characterized this hormone (in the brain) and determined a similarity to authentic circulating insulin. Additionally, we measured the serum and CSF insulin concentrations and attempted to demonstrate a relationship with that of the brain.
MATERIALS AND METHODS
Animals New Zealand White rabbits at various developmental time points were employed (fetal: 23 d, n = 3; 25 d, n = 3; 27 d, n = 3; term being ~31 days; neonatal: 2 d, n = 4; 4 d, n = 4; 10 d, n = 4; and adult, n = 3). The NIH guidelines (as approved by the two institutions) for the care and use of all animals were followed carefully. Following phenobarbital administration, the brains were perfused, via the left ventricle and carotids, with chilled phosphate-buffered saline (PBS) (pH 7.4) containing one unit of heparin per milliliter to clear the brain of circulating blood, thereby minimizing contamination by insulin that is present within the cerebral vasculature both in the free and receptor-bound forms. These well perfused brains were subsequently removed, pooled from a single litter, and subjected to an acid-ethanol extraction as previously described 29. The percent recovery of the extraction procedure as determined with 1251-insulin (as a tracer) was 60-70%.
ELISA Brain insulin content was quantitated by a sensitive ELISA 44. Wells on microtest falcon plates were coated with varying concentrations, ranging from 500 fg to 10 ~g, of brain extracts and a porcine insulin (Eli Lilly, Indianapolis, IN) standard. These coated plates were left overnight at 4°C, in a pH of 9.6, prior to incubating successively with PBS/0.1% bovine serum albumin, mouse monoclonal anti-insulin antibody (Biogenex Laboratories, Dublin, CA) at a dilution of 1:10,000 for 1 h, rabbit anti-mouse IgG 1:50 for 1 h, mouse monoclonal peroxidase (ICN Biomedicals, Inc., Costa Mesa, CA) 1:1,000 for 1 h, and 3,3',5,5', tetramethyl benzidine (TMB) (ICN Biomedicals) -- a chromogen -- for 1 h. The plates were washed with 0.5% PBS-Tween 20 between each incubation and all antibody dilutions were done with PBS-0.1% BSA0.5% Tween 20. Addition of 2 N su!furic acid terminated the reaction, following which the optical density of each well was read at 450 nm using an ELISA reader. Controls included: (1) normal mouse serum, (2) PBS buffer alone, (3) antibody in the absence of an antigen, and (4) antibody pre-absorbed with saturating concentrations of porcine insulin. To study the cross-reactivity of the antiinsulin antibody, some of the wells were coated with the synthetic insulin-like growth factor I (IGF I; Amgen, Thousand Oaks, CA) and the mouse insulin-like growth factor II (Collaborative Research, Waltham, MA).
Characterization of brain insulin Based on our results, late gestation fetal or early neonatal brain extracts (with peak amounts of immunoassayable insulin) were used for the following characterization studies.
Protein gel electrophoresis The lyophilized fetal/neonatal brain extract (1 mg) was resuspended in Laemmli's buffer and subjected to conventional SDSpolyacrylamide gel electrophoresis, employing a 5-20% discontinuous gradient gel 35. Protein separation was accomplished under constant current (30 mA). The separated proteins were detected by either Coomassie blue 25 or the double staining silver techniques 24. Further, the brain extract (5-10 mg) was iodinated by the Chloramine T method as described previously4s. 1-2 mg of labeled brain extract was separated by SDS-PAGE under reducing conditions in an attempt to identify the brain insulin peptide. Very low molecular weight markers (Bethesda Research Laboratories, Gaithersburg, MD), porcine (Eli Lilly) or bovine insulin (Sigma Chemical Co., St. Louis, MO), and synthetic IGF-I (Amgen), were used as standards. To enhance the capability of detecting nanogram quantities of protein samples, we additionally employed an automated electrophoretic and developer PhastSystem (Pharmacia LKB Biotechnology, Inc., Piscataway, NJ) 37. An 8-25% native gradient gel was used for the initial electrophoretic separation of unlabeled brain extract proteins by charge and size. Bovine (Sigma Chemical Co.) and rabbit (Eli Lilly) insulins were used as standards. The conditions for electrophoretic protein separation were established and pre-programmed by Pharmacia 37 to be 400 V, 10 mA, 2.5 W, 10 Vh for the pre-run, 400 V, 1 mA, 2.5 W, 2 Vh for the application of protein samples and 400 V, 10 mA, 2.5 W, 135 Vh for the actual electrophoretic separation. All these steps were performed at 15°C. Following electrophoresis, the protein bands were visualized by silver staining24. Due to the minute concentrations of insulin in brain extracts, gels were overloaded with brain extract samples to obtain a visible insulin band that co-migrates with insulin standards. Rabbit brain extract proteins were also separated by the electrophoretic PhastSystem and transferred to 0.05/~m pore size nitrocellulose filters (Schleicher and Sehuell), employing the PhastTransfer semi-dry transfer kit (Pharmacia), a semi-dry transfer block and a discontinuous buffer system 3s. Efficient transfer was accomplished in 25 min at 1.2 mA at a temperature of 15°C. After electrophoretic transfer of proteins, the nitrocellulose filters were incubated with a blocking reagent (I Block) (Tropix Lab, Bedford, MA) in PBS, pH 7.4, containing 0.1% Tween 20 and 0.2% sodium azide for 30 rain, followed by a 1:100 dilution of a guinea pig anti-porcine insulin antibody (Linco, St. Louis, MO) for 2 h at room temperature. The primary antibody incubation was followed by a 1:20,000 dilution of an alkaline phosphatase conjugated goat anti-guinea pig IgG (Jackson ImmunoResearch Lab, Inc., West Grove, PA) for 15 min at room temperature. Insulin immunoreactivity on the filter was then detected by an ultrasensitive chemiluminescent technique described by Brownstein 7. The antigen-antibody complex was detected by reacting the alkaline phosphatase conjugated second antibody to adamantyl 1-2-dioxetane phosphatase (AMPPD), employing a chemiluminescent kit (Tropix Lab). This reaction caused the spontaneous emission of light which was visualized by exposure to X-ray film (Kodak XAR, Sigma) for 30 s. The specificity of the immunoreaction was controlled by employing the guinea pig anti-porcine insulin antibody that was pre-absorbed with saturating concentrations of porcine insulin.
lmmunoaffinity chromatography 4 mg of iodinated brain extract (4.6 × 106 cpm) was dissolved in 50 ml of PBS with 100 ~M phenylmethylsulphonyl fluoride and centrifuged at 10,000 rpm for 10 min at 4°C. The supernatant was cycled over the immunoabsorbent column, which was prepared by coupling anti-insulin monoclonal IgG to Affi-gel 10 (BioRad Laboratories, Richmond, CA) and equilibrated with PBS at 40 ml/h 56. The column bound -50% of the labeled brain extract. Following
29 lected from the immunoaffinity column were subjected to SDSPAGE and autoradiography in an attempt to characterize the eluates.
extensive washing of the column with 20 mM Tris-HCl, pH 7.4, 0.5 M NaCI and PBS, the immunoreactive brain insulin was eluted with 0.2 N acetic acid, 0.03% Triton X100, pH 3.0. The eluate was collected in 1 ml aliquots and the radioactivity assessed in a gamma counter. Iodinated porcine insulin (spec. act. 150/~Ci/~tg) was employed as a standard (60,000 cpm), of which 70% of the labeled insulin bound to the affinity column. The different aliquots col-
High performance liquid chromatography Brain extract from fetal and newborn rabbits were applied to the BioRad Hi-Pore 318 reverse phase high performance liquid chro-
~',~ ~, Control P. Insulin 27day ] 25day fetus 22day 2 day old ] / 3 day old | newborn 4 day old J = = = 10 day 01d_J
A
E ¢ O It) ~t m ¢ o a
2.0-
•e e o
1.5-
IGF I
a, IGF II
1.0-
m
_u 0
0.5-
I
lO/.,g
I
1
lOOng
lng
I
I
lOOpg 50pg
I
500fg
Brain ExtractlPeptide Concentrations E! 200ng 180ng-190ng 170rig-
i
160rig150ng-
|
140ng-.
15 110nglOOng90ng-
¢11 t,,, 80~g'70rig60ng50rig 40ng30ng20riglong23d 25d 270 2d 4
27
BRES 17769
Developmental regulation of insulin in the mammalian central nervous system Ruben Schechter a, Jennifer Whitmire a, Lynne Holtzclaw b, Mark George b, Robert Harlow a and Sherin U. Devaskar b aSt. Franscis Hospital of Tulsa Medical Research Institute, University of Oklahoma Health Science Center and the Department of Neonatology, St. Francis Hospital, Tulsa, OK (USA) and bDivision of Neonatology, Department of Pediatrics, St. Louis University School of Medicine, The Pediatric Research Institute and Cardinal Glennon Children's Hospital, St. Louis, MO (USA) (Accepted 7 January 1992) Key words." Development; Rabbit; Brain; Insulin; Cerebrospinal fluid
We delineated the ontogeny of rabbit brain insulin concentrations to define the regulatory role of development on this hormone in the central nervous system. Employing a sensitive ELISA, we observed higher concentrations in the late gestation fetal brain (-80-90 ng/g) and early neonatal brain (-195 ng/g) in comparison to the adult (-32 ng/g; P < 0.01). Further, we characterized this hormone to determine the identity of insulin (or an insulin-like substance) in brain. Employing porcine/bovine or rabbit insulin as standards, we observed that brain insulin mimicked authentic insulin in its migration on SDS-polyacrytamide and native gel electrophoresis, immunogenicity on Western blot analysis, and its elution profile on immunoaffinity column chromatographic, and high performance liquid chromatographic separation. We then examined the developmental effects on circulating and cerebrospinal fluid (CSF) radioimmunoassayable insulin levels. No statistically significant differences (ANOVA) existed through development in either the serum or CSF insulin levels. Employing multiple regression analysis, no correlation was evident between brain and either serum or CSF insulin concentration. A search for insulin mRNA by Northern blot analysis yielded minute amounts of atypical large sized transcripts. We conclude that the insulin peptide in the central nervous system closely resembles (or is identical to) circulating insulin in many properties and that there is a developmental increase in brain insulin concentrations, the maximal peak occuring in the late gestation fetus and early neonate. Insulin concentrations in brain demonstrate no conventional relationship to either the serum or CSF insulin levels, suggesting an additional source of peptide, which contributes beyond that which is available via the circulation. The amounts of insulin present within the central nervous system are minute (difficult to detect) but in the range (10-100 ng) where the hormone can interact with either insulin or insulin-like growth factor I (IGF-I) receptors that are abundantly present on developing brain cells, thereby executing the biological function of the hormone. INTRODUCTION There has been an ongoing controversy regarding the origin of insulin in the central nervous system (CNS) since the initial detection of the h o r m o n e in adult rat brain 29. While various investigations u n d e r t a k e n both in vitro 8'13'33'43'44and in vivo 47'51'52'57point towards the fact that an insulin-like peptide is synthesized de novo within brain cells, there are other studies which support the contention that insulin traverses the b l o o d - b r a i n barrier and gains access to the brain 4'53'54. Circulating insulin u n d e r conditions of hyperinsulinemia, has been observed to cross the b l o o d - b r a i n barrier and enter the cerebrospinal fluid (CSF), mainly in the adult h u m a n 53 and various other animal species 4'52, while access to the brain parenchyma has unequivocally been demonstrated in the neonatal rabbit alone 23. Further, there appear to be major functional differences exerted by insulin in the CNS based on the developmental stage of the brain. For ex-
ample, most of the biological effects relating to brain cellular growth and differentiation 1'2'4°, potentiation of brain glucose 11 and macromolecular metabolism 12 have been observed in the immature brain cells, whereas a hormonal regulatory effect on the whole animal's feeding behavior has been observed in the adult brain alone 26'55. Insulin's biological effect on brain cell growth promotion and differentiation is more global in comparison to its effect on feeding patterns. The former function involves most neurons and glial cells 1'2'40'41 that are widespread within all structures of the brain, whereas the latter constitutes a specialized function that is generally localized to discrete, perhaps small, areas of the brain 39. Previously it has been reported, in rats, mice and rabbits, that fetal concentrations of circulating insulin are higher than the early neonatal and adult values 5' ~5,~8. Although fetal hyperinsulinemia has not demonstrated a gross increase in fetal brain weight 32, insulin is a fetal tissue growth promoter 16, resulting in generalized
Correspondence: S.U. Devaskar, 1465 South Grand Blvd., St. Louis, MO 63104, USA.
28 organomegaly 32"5°. There have been no systematic studies quantitating brain insulin concentrations during these critical fetal and neonatal phases of CNS development 22' 31. Further, there have been no attempts to determine the relationship between brain insulin levels and the circulating or CSF insulin concentrations during normal development. We hypothesized that in the presence of higher circulating fetal insulin concentrations and an increased need for a global brain cell growth potentiating effect, increasing amounts of insulin will cross the bloodbrain barrier and gain access to the immature brain when compared to the adult. Thus we undertook the present investigation and delineated the ontogeny of brain insulin content, characterized this hormone (in the brain) and determined a similarity to authentic circulating insulin. Additionally, we measured the serum and CSF insulin concentrations and attempted to demonstrate a relationship with that of the brain.
MATERIALS AND METHODS
Animals New Zealand White rabbits at various developmental time points were employed (fetal: 23 d, n = 3; 25 d, n = 3; 27 d, n = 3; term being ~31 days; neonatal: 2 d, n = 4; 4 d, n = 4; 10 d, n = 4; and adult, n = 3). The NIH guidelines (as approved by the two institutions) for the care and use of all animals were followed carefully. Following phenobarbital administration, the brains were perfused, via the left ventricle and carotids, with chilled phosphate-buffered saline (PBS) (pH 7.4) containing one unit of heparin per milliliter to clear the brain of circulating blood, thereby minimizing contamination by insulin that is present within the cerebral vasculature both in the free and receptor-bound forms. These well perfused brains were subsequently removed, pooled from a single litter, and subjected to an acid-ethanol extraction as previously described 29. The percent recovery of the extraction procedure as determined with 1251-insulin (as a tracer) was 60-70%.
ELISA Brain insulin content was quantitated by a sensitive ELISA 44. Wells on microtest falcon plates were coated with varying concentrations, ranging from 500 fg to 10 ~g, of brain extracts and a porcine insulin (Eli Lilly, Indianapolis, IN) standard. These coated plates were left overnight at 4°C, in a pH of 9.6, prior to incubating successively with PBS/0.1% bovine serum albumin, mouse monoclonal anti-insulin antibody (Biogenex Laboratories, Dublin, CA) at a dilution of 1:10,000 for 1 h, rabbit anti-mouse IgG 1:50 for 1 h, mouse monoclonal peroxidase (ICN Biomedicals, Inc., Costa Mesa, CA) 1:1,000 for 1 h, and 3,3',5,5', tetramethyl benzidine (TMB) (ICN Biomedicals) -- a chromogen -- for 1 h. The plates were washed with 0.5% PBS-Tween 20 between each incubation and all antibody dilutions were done with PBS-0.1% BSA0.5% Tween 20. Addition of 2 N su!furic acid terminated the reaction, following which the optical density of each well was read at 450 nm using an ELISA reader. Controls included: (1) normal mouse serum, (2) PBS buffer alone, (3) antibody in the absence of an antigen, and (4) antibody pre-absorbed with saturating concentrations of porcine insulin. To study the cross-reactivity of the antiinsulin antibody, some of the wells were coated with the synthetic insulin-like growth factor I (IGF I; Amgen, Thousand Oaks, CA) and the mouse insulin-like growth factor II (Collaborative Research, Waltham, MA).
Characterization of brain insulin Based on our results, late gestation fetal or early neonatal brain extracts (with peak amounts of immunoassayable insulin) were used for the following characterization studies.
Protein gel electrophoresis The lyophilized fetal/neonatal brain extract (1 mg) was resuspended in Laemmli's buffer and subjected to conventional SDSpolyacrylamide gel electrophoresis, employing a 5-20% discontinuous gradient gel 35. Protein separation was accomplished under constant current (30 mA). The separated proteins were detected by either Coomassie blue 25 or the double staining silver techniques 24. Further, the brain extract (5-10 mg) was iodinated by the Chloramine T method as described previously4s. 1-2 mg of labeled brain extract was separated by SDS-PAGE under reducing conditions in an attempt to identify the brain insulin peptide. Very low molecular weight markers (Bethesda Research Laboratories, Gaithersburg, MD), porcine (Eli Lilly) or bovine insulin (Sigma Chemical Co., St. Louis, MO), and synthetic IGF-I (Amgen), were used as standards. To enhance the capability of detecting nanogram quantities of protein samples, we additionally employed an automated electrophoretic and developer PhastSystem (Pharmacia LKB Biotechnology, Inc., Piscataway, NJ) 37. An 8-25% native gradient gel was used for the initial electrophoretic separation of unlabeled brain extract proteins by charge and size. Bovine (Sigma Chemical Co.) and rabbit (Eli Lilly) insulins were used as standards. The conditions for electrophoretic protein separation were established and pre-programmed by Pharmacia 37 to be 400 V, 10 mA, 2.5 W, 10 Vh for the pre-run, 400 V, 1 mA, 2.5 W, 2 Vh for the application of protein samples and 400 V, 10 mA, 2.5 W, 135 Vh for the actual electrophoretic separation. All these steps were performed at 15°C. Following electrophoresis, the protein bands were visualized by silver staining24. Due to the minute concentrations of insulin in brain extracts, gels were overloaded with brain extract samples to obtain a visible insulin band that co-migrates with insulin standards. Rabbit brain extract proteins were also separated by the electrophoretic PhastSystem and transferred to 0.05/~m pore size nitrocellulose filters (Schleicher and Sehuell), employing the PhastTransfer semi-dry transfer kit (Pharmacia), a semi-dry transfer block and a discontinuous buffer system 3s. Efficient transfer was accomplished in 25 min at 1.2 mA at a temperature of 15°C. After electrophoretic transfer of proteins, the nitrocellulose filters were incubated with a blocking reagent (I Block) (Tropix Lab, Bedford, MA) in PBS, pH 7.4, containing 0.1% Tween 20 and 0.2% sodium azide for 30 rain, followed by a 1:100 dilution of a guinea pig anti-porcine insulin antibody (Linco, St. Louis, MO) for 2 h at room temperature. The primary antibody incubation was followed by a 1:20,000 dilution of an alkaline phosphatase conjugated goat anti-guinea pig IgG (Jackson ImmunoResearch Lab, Inc., West Grove, PA) for 15 min at room temperature. Insulin immunoreactivity on the filter was then detected by an ultrasensitive chemiluminescent technique described by Brownstein 7. The antigen-antibody complex was detected by reacting the alkaline phosphatase conjugated second antibody to adamantyl 1-2-dioxetane phosphatase (AMPPD), employing a chemiluminescent kit (Tropix Lab). This reaction caused the spontaneous emission of light which was visualized by exposure to X-ray film (Kodak XAR, Sigma) for 30 s. The specificity of the immunoreaction was controlled by employing the guinea pig anti-porcine insulin antibody that was pre-absorbed with saturating concentrations of porcine insulin.
lmmunoaffinity chromatography 4 mg of iodinated brain extract (4.6 × 106 cpm) was dissolved in 50 ml of PBS with 100 ~M phenylmethylsulphonyl fluoride and centrifuged at 10,000 rpm for 10 min at 4°C. The supernatant was cycled over the immunoabsorbent column, which was prepared by coupling anti-insulin monoclonal IgG to Affi-gel 10 (BioRad Laboratories, Richmond, CA) and equilibrated with PBS at 40 ml/h 56. The column bound -50% of the labeled brain extract. Following
29 lected from the immunoaffinity column were subjected to SDSPAGE and autoradiography in an attempt to characterize the eluates.
extensive washing of the column with 20 mM Tris-HCl, pH 7.4, 0.5 M NaCI and PBS, the immunoreactive brain insulin was eluted with 0.2 N acetic acid, 0.03% Triton X100, pH 3.0. The eluate was collected in 1 ml aliquots and the radioactivity assessed in a gamma counter. Iodinated porcine insulin (spec. act. 150/~Ci/~tg) was employed as a standard (60,000 cpm), of which 70% of the labeled insulin bound to the affinity column. The different aliquots col-
High performance liquid chromatography Brain extract from fetal and newborn rabbits were applied to the BioRad Hi-Pore 318 reverse phase high performance liquid chro-
~',~ ~, Control P. Insulin 27day ] 25day fetus 22day 2 day old ] / 3 day old | newborn 4 day old J = = = 10 day 01d_J
A
E ¢ O It) ~t m ¢ o a
2.0-
•e e o
1.5-
IGF I
a, IGF II
1.0-
m
_u 0
0.5-
I
lO/.,g
I
1
lOOng
lng
I
I
lOOpg 50pg
I
500fg
Brain ExtractlPeptide Concentrations E! 200ng 180ng-190ng 170rig-
i
160rig150ng-
|
140ng-.
15 110nglOOng90ng-
¢11 t,,, 80~g'70rig60ng50rig 40ng30ng20riglong23d 25d 270 2d 4
