Pepndes. Vol 11, pp 5-12 ©PergamonPress plc, 1990 Pnnted m the U S A

0196-9781/90 $3 00 + 00

Metabolic Clearance of Insulin From the Cerebrospinal Fluid in the Anesthetized Rat M. M A N I N , 2 Y B R O E R , * M. B A L A G E , W . R O S T E N E * A N D J. G R I Z A R D

Laboratoire d'Ftude du MFtabolisme AzotF, URA CNRS 164, INRA Thetx 63122 Ceyrat, France *UnttF de Recherche sur les Peptides Neurodtgestifs et le Diabbte, U 55 INSERM, Hopital Saint Antoine 75571 Paris cFdex 12 R e c e i v e d 15 June 1989

MANIN, M , Y BROER, M BALAGE, W ROSTENE AND J GRIZARD Metabohc clearance of msuhn from the cerebrospmal flutd m the anesthetlzed rat PEPTIDES 11(1) 5-12, 1990 --Infusion of 125I-(Tyr A14)-Insuhn at tracer doses into the cerebrospmal fluid (CSF) resulted in a slow rate of increase m the CSF-labeled lnsuhn during the first 2 hours with a plateau thereafter Labeled msuhn was cleared from the CSF at a higher rate than 3H-muhn, a marker of CSF bulk flow The labeled insulin was mainly &stnbuted m all the ventncular and penventrlcular brain regions Small amounts of degraded msuhn appeared in the CSF Colnfusmn with an excess of unlabeled lnsuhn impaired the clearance and degradatmn of labeled insulin It also inhibited the labehng m medial hypothalamus, olfactory bulbs and bram stem In contrast, comfuslon of nbonuclease B (used to test the specificity of uptake) was without any effect It was concluded that there is an active insulin intake from CSF into brain specific compartments that is presumably essential for the effects of msuhn on brain function Brain

CSF

Insuhn receptors

INSULIN receptors and lmmunoreactlve lnsuhn have been shown to be widely &stributed throughout the central nervous system (CNS) (16,17). In addition, there is evidence, from both In VlVO and in wtro studies, that lnsuhn is able to act on the CNS [see recent reviews (6) and (28)]; however, the biological role of insulin m the brain remains unclear. Its role as a CNS neuromodulator in food intake regulation (35), glucose homeostasis (19) or as growth factor during brain development (11,21) has been suggested; the source and the mode of access of insulin to its brain binding sites remain to be determined. For example, CSF insulin has been proposed as a feedback signal in some of the central acuons of insulin, hke long-term regulation of energy balance (35). Though only hmited reformation is available on the source and the mode of access of insuhn mto the brain, it was suggested that blood insulin is able to penetrate into the CSF (34) and from there to reach brain insulin targets (4). In addition, the questmn whether or not insulin within the CSF could provide a circulatory signal to the brain is a matter of debate" firstly, intracerebroventrlcular infusions of lnsuhn gave conflicting results, especmlly as far as food retake was concerned [see Discussion m (25)]; secondly, there were some uncertainties about correlations between plasma insulin and CSF lnsuhn (24). Alternatively, the role

of CSF in ehmlnatlng insulin as a waste product might be important. Studtes on the metabolism of neuroactlve substances, especially with neurohypophyslal peptldes present within the CSF, have been shown to clarify those possibilities (20). Such xnvestigations have not yet been c a m e d out for insulin In the present work, we have stu&ed the clearance from CSF of 125I-(Tyr A14)-insulin after lntracerebroventricular Infusion to anesthetized rats. We have also investigated the compartmental distribution of insulin in brain and the anatomical distribution of its binding sites. METHOD

Hormones and Chemicals Actrapid monocomponent porcine insulin (5 IU/ml) was obtained from NOVO (Bagsvaerd, Denmark). Goat antigulnea pig IgG, rabbit y-globuhn (Pentex, fraction II) normal guinea pig serum and the gumea pig antiporcme insulin antibody (GP 30) were obtained from Miles Scientific (Elkart, IN). Human [125I] monolodo tyrosme A14 insulin (IM 166), 3H-inulin and 3HHyperfilm were purchased from Amersham, Bucks, U.K The msuhn radiolmmunoassay (RIA) kit (SB INSI 1) was from the Commissariat ~ l'Energie Atomique, France. Bovine serum albu-

~Thls work was supported by grants from Aide aux Jeunes Dlab&Nues et Instltut National de la Recherche Agronomlque. A portion of this work was presented at the 13th European Symposium on Hormones and Cell Regulation, September 26-29, 1988, Mont Samte Odde, France 2Requests for repnnts should be addressed to M Manta, Laboratolre d'Etude du M6tabohsme Azotd, INRA Clermont-TheIx, 63122 Ceyrat, France

6

MANIN, BROER, BALAGE, ROSTENE AND GRIZARD

min (BSA, fraction V) and nbonuclease B were obtained from Sigma Chemicals Co., St. Louis. MO.

(3500 cpm/mm, 2 ng/mm) for 160 man and measured in CSF samples.

Ammals and Experiments

lntracerebroventncular Infusion Procedure

Adult male Wlstar rats (250-350 g), fasted overnight, were anesthetized with nembutal (60 mg/kg IP) and maintained under anesthesia for the duration of the experiment by periodic subcutaneous injections of nembutal. The animals were placed in a hot bed to maintain adequate temperature They were stereotaxically amplanted with a stainless steel guide cannula into the right lateral cerebral venmcle as previously described (25) Another cannula was placed into the magna ClSterna for CSF samphng; the latter was constructed with polyethylene tubing (inner diameter = 0 76 mm; outer diameter = 1 22 mm) and silicone tubing and closed wath a mandrel Cannulae were fixed to the skull by anchoring screws and dental cement Four separate expenments (I-IV) were performed Experiment I was carried out to study the appearance of labeled materials in the CSF and brain areas following intracerebroventrlcular infusions of ~25I-msulin at tracer dose (about 0 0009 ng/mm in most cases and 0.003 ng/mln when autoradiographlc studies were performed) with or without unlabeled Insulin (10--45 ng/mln) or ribonuclease B (10 ng/min); the latter peptide, a 124 amino acid peptlde with 4 disulfide cross-hnkings, was used to assess the specificity of insulin uptake (4). The experiments were always performed in parallel with two animals, one receiving the tracer alone and the other the tracer plus the unlabeled substances. Thirty p~l CSF samples were withdrawn every 20 mm for a 180-220 mm period and 125I-Insulin concentraUons were determined. At the end of the infusion, the excess of tracer was removed by a subsequent infusion of saline for 12 min at 20 ~l/min, CSF was simultaneously sampled from the magna clsterna at the same rate. We showed that this process eliminated more than 90% of the labeled insulin level within the CSF. Then, the animals were killed by decapitation and brains were quickly removed. In most cases, brains were microdlssected to obtain cerebral cortex, medial and lateral hypothalaml, olfactory bulbs, brain stem, cerebellum and the remaining tissue. In some cases, brains were frozen on dry ice and submitted to autoradiographlc studies described by Broer et al (10) with the following modifications brain sections (200 ~ m thick) cut at various brain levels were exposed to 3H-Hyperfilm after preincubation for 1 hr at 4°C in Krebs Ringer phosphate buffer (KRP) to eliminate unbound radioactivity. Study of specificity of insulin binding was assessed on adjacent sections at the level of the hypothalamus; those latter were first preincubated in KRP buffer without insulin and then incubated for 15 hr at 4°C m the absence or in the presence of an excess of unlabeled insulin (1 p,M) (10). Experiment II was performed to study thoroughly the effect of unlabeled insulin added to the labeled insuhn infused. Each animal underwent the 125I-Insuhn lnfusmn at tracer dose (about 0.0009 ng/mln) for 180-220 min and then the mixture of tracer (0.0009 ng/mln) plus unlabeled insulin (0 9-30 ng/min) or rthonuclease B (10 ng/mln) for an additional 140-mm period. The CSF 125Iinsuhn and the radioactivity of microdissected brain areas were determined as described above. Experiment III was performed m order to test the accuracy of the use of iodinated insulin Unlabeled lnsuhn alone was intracerebroventricularly infused (1-3 ng/min) for a 180-min period. Immunoreactlve insulin in brain compartments was determined. The results were compared to those using labeled insulin In Experiment IV, we compared the importance of CSF bulk flow to CSF insulin clearance by using 3H-inulin, a marker of CSF bulk flow (20). 3H-Inulin was intracerebroventricularly infused

lntracerebroventricular infusions of substances were performed with a miniature syringe pump (Braun infuser) at a rate of 1.5 ~l/mln This infusion rate was lower than the CSF bulk flow Solutions were maintained into ice for the durauon of the experiment in order to avoid hormone degradation. Moreover, to diminish the aggregation of insulin (9) the tracer and the unlabeled insulin were dassolved in a saline solution containing 0.05-0.1% BSA and glutamic acid (3 5 mg/ml). The infusion started with the introduction of anfuslon cannula into the guide cannula. Samples of cisternal CSF were withdrawn by means of a peristaltic pump (Gilson, 20 ptl/min) into a polyethylene tubing, calibrated for two subsequent samplings of 30 ~1. Indeed, to prevent adsorption of hormone on tubing wall, 30 p,l of 2 5% BSA solution were withdrawn before the 30 ~tl CSF sampling and the whole withdrawal was then collected into the assay tube. The infusion rate allowed wathdrawals of 30 ~tl CSF every 20 man. Our procedure represented an additional turnover rate of only 10% of the CSF pool for a 20-rain period. It seems unhkely that it could have an artefactual effect since it has been shown that about 50-150 ~,1 CSF can be collected without any sign of disturbing the rat (8). In addlUon, the magna clsterna IS the only convenient sampling site an small animals Although the existence of unstirred pools of CSF, or of disturbances m the normal production or removal of CSF might distort our findings, those phenomena would affect substances in the same way Thus, the comparisons between thear clearances would not be compromised In each experiment, the actual infusion rate of msuhn was measured three times from Immunoreactive msuhn and TCA precipitable radioactive material refused for 5 min in test tubes

Insuhn Assay Intact ~251-1nsulIn in CSF or infusion solutions was measured by TCA preclpltaUon Immediately after samphng, 440 jxl of cold 0.17% BSA (to obtain with BSA already present in the CSF samples a final concentration of 0.3% BSA) and 500 jxl of cold 10% TCA were added to the CSF After centrlfugatlon m a microfuge, the supernatant was collected and the radioactivity was measured both in the supernatant and precipitate To validate thas method, several samples of CSF and infusion solutions were submitted to lmmunopreclpltatlon. Samples (50 Ixl) were incubated 2 hr at room temperature with 50 txl of guinea p~g antlporclne insulin antiserum at 1/100 final dilution. NonspeclfiC binding was determined by incubating CSF samples in a similar way except that anti-insulin antiserum was substituted by normal guinea pig serum The complex was then precipitated by 20 Ixl of a second antibody (goat antxgulnea pig IgG). The mixture was incubated again 16 hr at 4°C After addition of 250 ~xl buffer (containing 10 g/1 sodium barbital, 2 5 g/l BSA, 100 mg/l pentex rabbit ~-globulin), samples were cenmfuged 15 mln at 1500 x g. Precipitates and the supernatants were counted for radioactivity Comparisons of both methods were made under conditions where degradation of insulin was relatively high (I.e., infusion of labeled insulin alone) and under conditions where its degradation was low (when an excess of unlabeled insulin was added), the ratios of the ~mmunopreclpltated radioactivity to TCA-preclpitated radioactivity (0.93 + 0 03 under high degradation, m e a n _+ SE, n = 3, 0.95 _+0.02 under low degradation, n = 2) indicated that both methods gave similar results. This was also the case for infusion solutions (0 960_+0.001, n = 3) So the TCA precipitation method, which

CSF INSULIN CLEARANCE AND BRAIN DISTRIBUTION

7

INFUSION RATE.S ] 29 hE/ram 2 ng/mbt 11'09 ng/mm

TRACER +

•-

20

40

UNLABELED INSULIN

A



t

p,.

7

Metabolic clearance of insulin from the cerebrospinal fluid in the anesthetized rat.

Infusion of 125I-(Tyr A14)-insulin at tracer doses into the cerebrospinal fluid (CSF) resulted in a slow rate of increase in the CSF-labeled insulin d...
730KB Sizes 0 Downloads 0 Views