| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Physiology and Biochemistry and Molecular Biology, Colorado State University, Fort Collins, Colorado 80523
Address all correspondence and requests for reprints to: Dr. Douglas N. Ishii, Physiology Department, Colorado State University, Fort Collins, Colorado 80523. E-mail: dnishii{at}aurogen.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Despite these advances, the prospect for the use of IGFs to treat central nervous system (CNS) diseases and disorders is dampened considerably by the widespread belief that polypeptides of the size of IGFs (Mr, 7.5 kDa) are unlikely to cross the blood-CNS barrier (B-CNS-B). Treatment, then, might require drilling of an access hole through the skull for IGF delivery. This would introduce risks such as potential CNS infection or surgical mishap, and a less risky procedure would be welcome.
Several recent lines of evidence show that IGFs may overcome the B-CNS-B. Experimental and clinical brain injury can be prevented or ameliorated by the administration of IGF by the sc or iv route (21, 22, 23, 24). After sc administration, IGF-I of molecular size indistinguishable from authentic IGF-I can be recovered from cerebrospinal fluid (CSF), suggesting that circulating IGF may enter and directly support the CNS (25).
There is a need to more fully understand the properties of IGF uptake into CSF. The purpose of this study was to test the hypotheses that uptake of circulating IGF into CSF is not mediated by type I or type II IGF receptors, and that sequestration of IGF to circulating IGF-binding protein-1 (IGFBP-1), -3, -4, or -5 is not required for uptake of IGF into CSF.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals and surgical procedures
All work involving rats was performed in accordance with the
principles set forth in the NIH Guide for the Care and Use of
Laboratory Animals. Adult, male rats (12 weeks old; 300–350 g;
Harlan Sprague Dawley, Inc., Indianapolis, IN) were
randomly assigned to treatment groups. Rats were injected sc in the
midback with a single bolus dose of vehicle, hIGF-I, or the indicated
factor. After 90 min, rats were anesthetized with an ip injection of 90
mg/kg ketamine and 8 mg/kg xylazine. An incision was made
longitudinally beginning at the base of the skull and continuing
caudad. Each animal was positioned on a bent platform to flex the
atlanto-occipital joint approximately 90°, as described by Hudson
et al. (26). The underlying muscles were
separated to expose the atlanto-occipital membrane. The exposed surface
of the atlanto-occipital membrane was thoroughly cleaned using saline
and cotton swabs. Using a disposable 1-cc syringe with a 26-gauge
needle secured in a micromanipulator, approximately 100 µl CSF were
extracted from the subarachnoid space of the cisterna magna through the
atlanto-occipital membrane. If any blood contamination was seen, the
sample was discarded. Tail blood samples were collected in tubes
containing heparin to prevent coagulation and centrifuged at 10,000 rpm
for 3 min in a microfuge. The CSF and plasma were stored at -70 C
until assayed.
CSF and plasma analysis by ELISA
Following the protocol recommended by Diagnostics Systems Laboratories, Inc. (Webster, TX), samples were subjected
to an acid-ethanol extraction to dissociate IGFBPs from IGFs in CSF and
plasma. Modifications were made in the extraction procedure to maximize
the sensitivity for small volumes of CSF (25). Extracted
CSF (30 µl), extracted and diluted plasma (30 µl), IGF and IGF
analog standards (varying concentrations), and BSA buffer (10 µg/ml
BSA in PBS, pH 7.4) were assayed in triplicate. The wells were
precoated with mouse anti-hIGF-I monoclonal antibody. In a sandwich
assay, mouse monoclonal anti-hIGF-I antibody conjugated to horseradish
peroxidase (100 µl) was added to each well, and the microwell plate
was incubated for 2 h at room temperature at 400 rpm on an orbital
shaker. The wells were aspirated and washed five times using normal
saline with a nonionic detergent. Tetramethylbenzidine chromogen
solution was added (100 µl/well), and samples were incubated for 10
min at room temperature on the orbital shaker. The reaction was
terminated with a stopping solution containing 0.2 M
sulfuric acid. The absorbance was measured on an MRX microplate reader
(Dynatech Corp., Chantilly, VA) set for dual wavelengths
at 450 nm and a background wavelength correction set at 600 nm.
The mean absorbance of standards was plotted against the hIGF-I or
analog concentration using a CSS:Statistica (StatSoft, Tulsa, OK)
software package. The BSA buffer blank was subtracted from the values,
and the best-fit regression curve was generated by computer (Fig. 1
). The hIGF-I concentrations of the
unknowns were determined from their standard curves. Any sample with
hIGF-I absorbance higher than that of all of the standards was diluted
and reassayed. Differences between group means were determined by
Newman-Keul post-hoc test in the case of three or more
groups or by t test in comparisons involving only two
groups. Significance was accepted at P < 0.05.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. The graphs
demonstrate both the linear relationship between concentration and
absorbance and the ability of the ELISA to detect picogram quantities
of immunoreactive hIGF-I (r = 0.997), des(1, 2, 3)hIGF-I (r =
0.996), [Leu24]hIGF-I (r = 0.990), and
[Leu60]hIGF-I (r = 0.988). The ELISA
detected des(1, 2, 3)hIGF-I with 2-fold and
[Leu60]hIGF-I with 3-fold lower sensitivity.
Peptide concentration values in subsequent figures were corrected for
the loss in sensitivity by using the appropriate standard curve. The
ELISA did not detect hIGF-II or insulin (Table 1). Neither untreated rat CSF nor plasma
reacted with the ELISA (Table 1 and Fig. 1), showing that this test is
specific for hIGF-I and its analogs, and the presence of rat
substances, such as rat insulin or IGFBP, have no influence on the
ELISA. Previous tests showed that 100 pg rat IGF-I, rat IGF-II, and
proinsulin are not detected by this assay (25).
|
). On the other hand, immunoreactive
hIGF-I uptake into CSF reached saturation at doses above 100 µg/rat
(Fig. 2A). Saturation is consistent with the possibility that uptake of
hIGF-I into CSF is carrier mediated.
|
). Rats were injected sc with either 150 µg
hIGF-I alone (n = 8) or with a combination consisting of 150 µg
hIGF-I plus 400 µg hIGF-II (n = 6). Plasma and CSF were
withdrawn for ELISA 90 min later. Immunoreactive hIGF-I uptake into CSF
was significantly reduced (P < 0.02) in rats treated
with the combination vs. hIGF-I alone (Fig. 3A). The administration of hIGF-II also
reduced (P < 0.05) plasma levels of hIGF-I from 600 to
300 ng/ml (Fig. 3B).
|
Rats were injected sc with equivalent amounts (200 µg) of hIGF-I or
des(1, 2, 3)hIGF-I, and CSF and plasma were withdrawn for ELISA 90 min
later. Des(1, 2, 3)hIGF-I was clearly detected in CSF, and sequestration
to IGFBP-3, IGFBP-4, and IGFBP-5 did not appear essential for hIGF-I
uptake (Fig. 4A). Plasma levels of
immunoreactive des(1, 2, 3)hIGF-I were lower than those of hIGF-I (Fig. 4B), possibly because des(1, 2, 3)hIGF-I has a shorter half-life due to
reduced binding to IGFBP.
|
Rats were injected sc with [Leu24]hIGF-I (200
µg; n = 3), [Leu60]hIGF-I (100 µg;
n = 4), or hIGF-I (200 µg; n = 3), and CSF as well as
plasma were withdrawn 90 min later. A lower concentration of
[Leu60]hIGF-I was used due to its high cost.
Both [Leu24]hIGF-I and
[Leu60]hIGF-I were readily detected in the CSF
samples (Fig. 5A). Taken together, these
data suggest that the uptake carrier is unlikely to be the type I IGF
receptor.
|
and Fig. 5B).
In the case of des(1, 2, 3)hIGF-I, this may be due in part to reduced
binding to IGFBP, resulting in a larger fraction of the free forms and
a shorter half-life. Additionally, it is possible that these analogs
are more rapidly cleared from plasma than IGF-I due to faster
metabolism or excretion.
Effect of treatment on plasma glucose levels
High doses of hIGF-I and its analogs may cross-occupy the
insulin receptor and reduce glucose levels in these acute experiments.
Glucose was measured in stored plasma samples at the conclusion of
tests, and no significant differences were found between the untreated
group and groups treated with hIGF-I (100, 200, and 300 µg),
des(1, 2, 3)hIGF-I, [Leu24]hIGF-I, or
[Leu60]hIGF-I.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
hIGF-II reduced hIGF-I uptake into CSF as well as reduced hIGF-I levels in plasma. The latter most likely resulted from the displacement by hIGF-II of hIGF-I from serum IGFBP, thereby causing the rapid elimination of free hIGF-I. The present results cannot distinguish between the possibilities that reduced hIGF-I uptake into CSF was a consequence of diminished circulating levels of hIGF-I, competition for a common carrier for CSF uptake, or both. Radioactivity is detected in brain parenchyma by autoradiography after the intracarotid administration of [125I]IGF-II (33).
The hIGF-I detected in CSF is unlikely to be due to contamination
of samples with blood
Increasing doses of hIGF-I resulted in a linear increase in serum
hIGF-I. If CSF samples were systematically contaminated with blood or
due to a slow leakage of IGF from blood into CSF, the distribution of
hIGF-I in CSF should linearly increase together with blood hIGF-I
levels. By contrast, hIGF-I uptake into CSF was observed to saturate,
and this departure from linearity showed that systematic contamination
with blood or slow leakage from blood was unlikely. The well shaped
saturation curve with the absence of substantial deviation of values
showed that sporadic contamination was also unlikely. The plasma level
of [Leu24]hIGF-I was higher than that of
hIGF-I. In systematic contamination, it would be expected that the CSF
[Leu24]hIGF-I level should likewise be higher
than that of CSF hIGF-I, but the reverse was observed. In other studies
the plasma hIGF-I levels were higher in nondiabetic vs.
diabetic rats, whereas the CSF hIGF-I levels were the same
(25). This result also is inconsistent with
contamination.
The rapid appearance of hIGF-I within 90 min in CSF would probably exclude retrograde axonal transport. We are unaware of any evidence that proteins can be transferred out of neurons into CSF after retrograde transport.
The uptake of circulating IGFs into CSF may not require
IGFBP
IGFBPs can prolong the half-lives of IGFs and may regulate their
biological activity. There was no reduction in uptake of
des(1, 2, 3)hIGF-I vs. hIGF-I, showing that uptake of hIGF-I
into CSF is independent of sequestration to IGFBP-3, -4, and -5 and
possibly all IGFBPs. It is likely that hIGFs are taken up in the
unbound form, a suggestion supported by the observation that
des(1, 2, 3)hIGF-I and hIGF-I uptakes into CSF were similar in amount
despite the lower total plasma level of the former. Although its total
plasma concentration was lower, a predominant fraction of
des(1, 2, 3)hIGF-I might be expected to be in the free form. However, this
decrease in plasma des (1, 2, 3)hIGF-I is not as great as might be
expected from reduced binding to IGFBP. This is undoubtedly due to the
acute nature of these experiments, as 90 min might be insufficient time
for substantial des(1, 2, 3)hIGF-I clearance to occur.
Des(1, 2, 3)hIGF-I is a naturally occurring truncated form of hIGF-I found in fetal and adult brain (36, 37) and seems to be the predominant form of hIGF-I in CSF. It is more potent than hIGF-I in vitro (38) and in vivo (39), and this is not due to enhanced affinity for the type I receptor (31). It is possible that the increased potency of endogenous des(1, 2, 3)hIGF-I compared with hIGF-I is due to reduced sequestration to IGFBPs resulting in higher concentrations of unbound des(1, 2, 3)hIGF-I.
The IGF carrier may differ from the type I IGF receptor, type II
IGF receptor, and insulin receptor
hIGF-I, hIGF-II, and des(1, 2, 3)hIGF-I all bind to the type I IGF
receptor, and uptake of these ligands might at first glance seem
consistent with the involvement of the type I IGF receptor in CSF
uptake. This possibility was refuted, however, by uptake from the
circulation into the CSF of [Leu24]hIGF-I and
[Leu60]hIGF-I.
Binding of [125I]IGFs has been studied in isolated brain microvessels (40, 41, 42). The internalization of [125I]IGF-I into isolated endothelial cells is believed to occur through interaction with the type I IGF receptor (40). For example, phosphorylation is activated by IGF binding to the type I receptor. Whether IGF-I binding is related to CSF uptake is uncertain, because this type of tissue preparation does not readily permit correlation of binding with IGF uptake into CSF, and the present studies show that uptake does not require the type I IGF receptor. If type I IGF receptors significantly outnumber carrier sites on endothelial cells, radioactive ligand binding studies of isolated microvessels may be unable to discriminate the carrier sites. New methods to identify carrier sites will need to be devised.
Furthermore, these data show that the IGF carrier is not the type II IGF receptor. hIGF-I and des(1, 2, 3)hIGF-I do not bind appreciably to the type II IGF receptor, yet these ligands are readily taken up into CSF. The type II IGF receptor is comprised of a single polypeptide, does not appear to contain intrinsic tyrosine kinase activity, binds mannose-6-phosphate in addition to IGF-II, and is believed to target IGF-II for degradation in lysosomes (43). Taken together, these results suggest that the uptake carrier is neither the type I nor the type II IGF receptor.
The insulin and type I IGF receptors are structurally similar, each
comprised of two 
- and two ß-subunits. Both receptors contain
intrinsic tyrosine kinase activity associated with the intracellular
domain of the ß-subunits. hIGF-I has approximately 100-fold reduced
affinity for the insulin receptor relative to insulin, and
[Leu60]hIGF-I, even at doses as high as 10
µM, does not bind to the insulin receptor
(32). The uptake of hIGF-I and
[Leu60]hIGF-I into CSF indicates that their
uptake is unlikely to require binding to the insulin receptor.
Insulin uptake from the circulation into the CSF has been studied (44, 45, 46, 47, 48). Uptake is saturable, consistent with an uptake carrier (47). It is also regulated, as insulin uptake into the CSF is reduced when marmots go into hibernation (49). The combined data suggest that insulin and IGFs, members of a common gene family, have the capacity for uptake into CSF. Insulin does not gate glucose entry into neurons. Insulin and IGFs additionally share neurotropic properties, including the capacity to induce neurites, support neuron survival (50), and increase expression of tubulin and neurofilament genes (7, 8).
Uptake of IGF into CSF from the circulation is substantial
Uptake of IGF from the circulation may contribute a substantial
fraction to the total IGF in CSF. After acute sc injection, hIGF-I
levels of 1.5 ng/ml were attained in CSF. This may be considered a
minimum, because endogenous rat IGF-I most likely competed against
exogenous hIGF-I for uptake. Human CSF contains approximately 3 ng/ml
IGF-I (51, 52); thus, the level of 1.5 ng/ml hIGF-I
attained in the present studies shows that acute administration of
hIGF-I can contribute substantially to total IGF-I content in CSF.
Similar conclusions were reached after constant hIGF-I infusion from sc
implanted osmotic minipumps (25).
Plasma levels of IGF-II range from about 440-1000 ng/ml in healthy 25- to 30-yr-old humans, and IGF-I ranges from about 60–280 ng/ml. After constant sc infusion, where steady state conditions of plasma hIGF-I are attained, CSF uptake saturates at approximately 150 ng/ml plasma hIGF-I in rats (25). This indicates that under normal conditions, a significant uptake of IGF-I into CSF occurs in rats and possibly also in humans.
IGF administration by routes that increase circulating IGF levels
can alter gene expression, preserve neural circuitry, and protect
against loss of function in the CNS
The data presented herein may help explain recent observations
showing that IGF administered sc or iv can support the CNS. IGF-II gene
expression is reduced in the brain of diabetic rats (53),
and continuous sc infusion of IGF-I returns IGF-II messenger RNA
content toward normal in diabetic rats, although the low IGF-I dose did
not reduce hyperglycemia (25). Noradrenergic axons arising
from the locus ceruleus descend the spinal cord and regulate the
hindlimb reflex. 6-Hydroxydopamine, a structural congener of
noradrenaline, when injected into the CSF at the cisterna magna, is
picked up by the descending noradrenergic axons. This results in the
loss of these axons and the hindlimb reflex in rats. However, sc
infusion of IGF-I, begun 20 min after 6-hydroxydopamine treatment,
spares both the noradrenergic spinal cord axons as well as the reflex
(24). Likewise, 3-acylpyridine produces ataxia by
lesioning inferior olive neurons in the cerebellum, and peripheral
IGF-I administration can prevent or reverse ataxia and preserve the
inferior olive neurons (23).
The management of CNS trauma is difficult and may benefit from systemic IGF administration. Learning and neuromotor function are spared after lateral fluid percussion brain injury in rats sc treated with IGF-I (21). Intravenous IGF treatment improves the clinical outcome after moderate to severe head injury in a phase II clinical trial (22). These observations are consistent with the known capacity of IGFs to support a variety of different types of CNS neurons. A better understanding of the properties of IGF uptake into the CSF may stimulate further research into the use of systemically administered IGF to treat experimental and clinical CNS diseases and disorders.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 17, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. L. Chen, N. A. Kassem, M. Sadeghi, and J. E. Preston Insulin-Like Growth Factor-II Uptake Into Choroid Plexus and Brain of Young and Old Sheep J. Gerontol. A Biol. Sci. Med. Sci., February 1, 2008; 63(2): 141 - 148. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. S. McEwen Physiology and Neurobiology of Stress and Adaptation: Central Role of the Brain Physiol Rev, July 1, 2007; 87(3): 873 - 904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. E. Sonntag, C. Bennett, R. Ingram, A. Donahue, J. Ingraham, H. Chen, T. Moore, J. K. Brunso-Bechtold, and D. Riddle Growth hormone and IGF-I modulate local cerebral glucose utilization and ATP levels in a model of adult-onset growth hormone deficiency Am J Physiol Endocrinol Metab, September 1, 2006; 291(3): E604 - E610. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. C. Russo, P. D. Gluckman, E. L. Feldman, and G. A. Werther The Insulin-Like Growth Factor System and Its Pleiotropic Functions in Brain Endocr. Rev., December 1, 2005; 26(7): 916 - 943. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Carro, C. Spuch, J. L. Trejo, D. Antequera, and I. Torres-Aleman Choroid Plexus Megalin Is Involved in Neuroprotection by Serum Insulin-Like Growth Factor I J. Neurosci., November 23, 2005; 25(47): 10884 - 10893. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. K. Dalsgaard, P. Ott, F. Dela, A. Juul, B. K. Pedersen, J. Warberg, J. Fahrenkrug, and N. H. Secher The CSF and arterial to internal jugular venous hormonal differences during exercise in humans Exp Physiol, May 1, 2004; 89(3): 271 - 277. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
