| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Medicine (H.J.W., A.L.), University of Birmingham, Birmingham B15 2TT, United Kingdom; Department of Anatomy and Cell Biology (M.B.), UMDS, Guy’s Hospital, London SE1 9RT, United Kingdom; Department of Medicine, Physiology and Paediatrics (D.J.H.), Lawson Research Institute, London, Ontario N6A 4V2, Canada
Address all correspondence and requests for reprints to: Ann Logan, Department of Medicine, The University of Birmingham, Wolfson Research Laboratories, Queen Elizabeth Medical Centre, Edgbaston, Birmingham, B15 2TH, United Kingdom. E-mail: a.logan{at}bham.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
IGF-I is widely distributed in the fetal and neonatal CNS but restricted in the adult (4, 5, 6). Both the action and bioavailability of IGFs are regulated by a family of six high affinity binding proteins (IGFBP-1–6; see review by Jones and Clemmons (7), which are complexed to IGFs in the circulation and the extracellular space. IGFBP-2, -4, -5, and -6 are all reported to be expressed in the intact mature CNS, although IGFBP-2 predominates, particularly in the meninges and choroid plexus (8, 9, 10, 11, 12). IGF-1R messenger RNA (mRNA) is heterogeneously distributed in the CNS from early stages of development, but levels decline postnatally (13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23).
A precise role for IGF-I in the injured adult CNS remains to be defined, although a neurotrophic activity may be particularly relevant and IGF-I, together with IGFBPs and signalling receptors, is suggested to play an important role in the recovery of neural tissue from injury [see review by Logan et al. (1)]. We hypothesize that, if the IGF-I axis regulates traumatic CNS wound responses by an autocrine/paracrine action, then the expression of IGF-I, IGFBP, and receptors will alter in injury responsive cells at the site of a penetrating wound. Specifically, in the early proliferative phase of the injury response, we would expect IGF-I bioactivity to increase (either via increased expression/transport of ligand/receptor or by potentiation of IGF-I activity by locally expressed IGFBPs), whereas in the later phases as cellular responses decline, the expression of ligand/receptor/potentiating IGFBPs is reduced, perhaps with a concomitant increase in expression of inhibitory IGFBPs. Therefore, the aims of this study were to localize and quantify the levels of IGF-I, IGFBP-1 through -6, and IGF-1R mRNAs and proteins within CNS wounds by ribonuclease protection assay (RPA), in situ hybridization and immunohistochemistry over the course of the cellular response using a rat model of penetrating brain injury.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals and surgery
Surgery was performed aseptically under a British Government
Home Office Licence. Groups of adult, female, 250 g Sprague-Dawley
rats were anesthetized ip with a mixture of Medetomidine (SmithKline
Beecham, Welwyn Garden City, Hertfordshire, UK) at 100 µg/kg body
weight and Ketamine hydrochloride (Parke Davis Veterinary, Eastleigh,
Hampshire, UK) at 10 mg/kg body weight or Hypnorm (fentanyl nitrate
0.135 mg/liter and fluanisone 10 mg/ml; Janssen Pharmaceuticals,
Oxford, UK)/Hypnovel (midazolam 1 ml in 10 ml water; Roche, Welwyn
Garden City, Hertfordshire, UK) anesthesia at 8 ml/kg body weight.
Buprenorphine (Sterling Health, Guildford, Surrey, UK) was administered
postoperatively as an analgesic. After craniotomy, to expose the
cerebral cortex, an incisional wound was placed in the mediolateral
right cerebral cortex using a David Kopf (Charles River, Margate, Kent,
UK) stereotactic instrument. The lesion was made precisely to a depth
of 4 mm along a 4.5 mm line parallel and 3 mm lateral to the midline
and spanning the coronal suture. Animals were allowed to recover post
surgery for periods of 0, 2, 5, 7, and 15 dpl.
Histology
Groups of four animals were deeply anesthetized (as for surgery)
and perfused transcardially with 250 ml of saline followed by 250 ml of
4% paraformaldehyde (PFA) in saline. After excision, brains were
postfixed overnight in 4% PFA (wt/vol, in 0.1 M sodium
tetraborate) at 4 C, dehydrated in graded alcohols, embedded in low
melting point polyester wax (24) and stored at 4 C. Sections, 7 µm
thick, were cut through the lesion site using a microtome (Bright,
Huntingdon, Cambridgeshire, UK) fitted with a cooled chuck, and mounted
on slides coated with either a 1% gelatin solution (for
immunohistochemistry), or with Biobond (British Biocell Int., Cardiff,
Glamorgan, UK, for in situ hybridization). Once mounted,
sections were air dried overnight and stored with desiccant at -70 C
(in situ hybridization) or 4 C (immunohistochemistry) until
use.
RNA extraction
Groups of three rats were killed with an anesthetic overdose,
the brains removed, rapidly dissected on ice, and stored at -70 C
until extraction. Total cellular RNA was extracted from the lesioned
and unlesioned hemispheres of the brain, using the RNAzol B method
(Biogenesis, Bournemouth, Dorset, UK), a modification of the
single-step method (25). Briefly, frozen tissue was weighed and
homogenized in 2 ml RNAzol per 100 mg tissue. Chloroform was added (0.2
ml/2 ml) and shaken for 15 sec before placing on ice for 5 min. Samples
were centrifuged at 5000 x g for 20 min at 4 C (Dupont
Sorvall RC5C, SS34 rotor, High Wycombe, Buckinghamshire, UK). The
aqueous phase was removed and precipitated with an equal volume of
isopropanol for 15 min on ice. Samples were centrifuged as above,
washed with 75% ethanol, and redissolved in 200 µl ultrapure water.
RNA was quantified by optical density at 260 nm and 280 nm and checked
for integrity on a 1% agarose gel.
Complementary RNA (cRNA) probe synthesis
The 500-bp coding region of rat IGF-I (L. Murphy, Winnipeg,
Canada) is contained within the pGEM blue plasmid (Promega).
HindIII and PvuII were used to linearize the
plasmid, and T7 and SP6 were used to generate the transcript for
antisense and sense templates, respectively. The rat IGFBP-1–6 coding
regions (from S. Shimasaki, San Diego, CA) were contained within the
plasmid pBluescriptSK+ (Stratagene). The plasmids contained
407, 397, 699, 444, 300, and 246 bp fragments, respectively.
EcoRI was used to generate antisense fragments for IGFBP-1
and -2, whereas ApaI was used for IGFBP-3; SmaI
for IGFBP-4 and -6; and SacII for IGFBP-5 antisense
templates. HindIII was used to generate all sense templates
for the IGFBPs except for IGFBP-3 and IGFBP-6, where BamHI
and EcoRI were used respectively. Riboprobe transcripts were
made with T7 and T3, for the antisense and sense templates, for all
IGFBPs, with the exception of IGFBP-3, where T3 was used for antisense
and T7 for sense. The 265 bp IGF-1R fragment (D. LeRoith, Bethesda, MD)
was cloned within the pGEM-3 plasmid (Promega). EcoRI and
BamHI were used to linearize the riboprobe, with SP6 and T7
being used to generate the respective antisense and sense templates.
Cyclophilin (J. Douglass, Portland, OR) was used as an internal control
in all ribonuclease protection assays to ensure equal loading on gels.
The plasmid contains a 680 bp coding region of rat cyclophilin
complementary DNA cloned within pSP65 (Promega). Cyclophilin has a
ubiquitous tissue and phylogenetic distribution and represents
0.1–0.4% of total cytosolic protein in many mammalian tissues. It
possesses strong homology to the enzyme peptidyl-prolyl cis-trans
isomerase, which catalyses the slow cis-trans isomerization of proline
peptide bonds in oligopeptides and accelerates slow, rate-limiting
steps in the folding of several proteins in the cell.
HindIII was used to linearize the plasmid for the antisense
template. SP6 polymerase was used to generate the antisense cRNA
probe.
Preparation of cRNA probes for ribonuclease assay. Transcription buffer (40 mM Tris pH 7.5, 6 mM MgCl2, 2 mM spermidine and 10 mM sodium chloride), 20 U rRNasin, 10 mM DTT, 0.5 mM ATP, GTP, uridine (U)TP, 12 µM cytidine (C)TP (150 µM for cyclophilin), 0.5 µg linearized antisense plasmid, 50 µCi 32P CTP (5 µCi for cyclophilin), and 15U RNA polymerase were incubated together at 37 C for 1 h. DNase I (1U) was added and the reaction left at 37 C for a further 15 min. Yeast transfer (t)RNA (20 µg) was added before a phenol:chloroform/chloroform extraction and ethanol precipitation. The cRNA probe was washed with 75% ethanol, briefly dried and resuspended in ultrapure water. The probe, 1 µl in 5 ml Ecolite+ scintillation fluid (ICN Flow, High Wycombe, Buckinghamshire, UK), was counted (using a Pharmacia counter) for 60 sec. 32P-labeled cRNA probes were stored at -20 C for no longer than 1 week.
Preparation of cRNA probes for in situ hybridization. Transcription buffer (40 mM Tris, pH 7.5, 6 mM MgCl2, 2 mM spermidine and 10 mM sodium chloride), 20 U rRNasin, 10 mM DTT, 2 mM ATP, GTP, CTP, 0.5 µg linearized antisense or sense plasmid, 200 µCi 35S UTP, and 15U RNA polymerase were incubated together at 37 C for 2 h. DNase I (1U) was added and the reaction left at 37 C for a further 15 min. To the reaction mix, 60 mM EDTA pH8.0 was added to a final volume of 50 µl. This was applied to a Sephadex G50 Quick Spin column (Boehringer Mannheim, Lewes, E. Sussex, UK) that was centrifuged at 1100 x g (Heraeus Sepatech Varifuge 3.2RS, Brentwood, Essex, UK) for 4 min. DTT was added to a final concentration of 150 mM. The probe, 1 µl in 5 ml Ecolite+ scintillation fluid, was counted for 60 sec. 35S-labeled cRNA probes were stored at -20 C and were kept for approximately 2 weeks. Before use, the 35S-labeled cRNA probes were recounted and checked on a 4% polyacrylamide/8 M urea gel to check integrity.
Ribonuclease protection assay
Ribonuclease protection assay (RPA) was performed on total RNA
extracted from the cerebral hemispheres of three rats from each
treatment group. Total RNA (20 µg) was dissolved in 30 µl of
hybridization solution (80% formamide, 40 mM PIPES pH 6.4,
400 mM sodium chloride, and 1 mM EDTA pH 8.0)
containing 60,000 cpm and 20,000 cpm of an IGF-related and cyclophilin
32P-labeled cRNA probe. After being heated to 85 C for 5
min, the cRNA probe was allowed to anneal the endogenous RNA at 45 C
overnight. At the end of the hybridization, the solution was diluted
with 350 µl of RNase digestion buffer (300 mM sodium
chloride, 10 mM Tris pH 7.4, and EDTA pH 7.5), containing
40 µg/ml of RNase A and 500U/ml of RNase T1, and incubated for 1
h at 30 C. Proteinase K (100 µg) in 10% SDS was added to the sample
and the mixture incubated at 37 C for an additional 20 min. After a
phenol/chloroform extraction and ethanol precipitation, the pellet
containing the RNA:RNA hybrid was briefly dried and resuspended in
loading buffer (80% formamide, 0.1% xylene cyanol, 0.1% bromophenol
blue and 2 mM EDTA, pH 8.0). The samples were boiled at 90
C for 5 min and separated on a 4% polyacrylamide/8 M urea
gel. 32P end-labeled (DNA polymerase 1) HinfI
digested pBR322 fragments were used as molecular markers. The mRNA
protected fragments were visualized by autoradiography against Amersham
Hyperfilm MP at -70 C.
Statistical analysis of RPA
Autoradiographic films were scanned using a ScanJet IIc and
Deskscan II software (both from Hewlett Packard, Geneva, Switzerland)
into a Macintosh LC475 computer (Apple Computer Inc., Cupertino, CA).
Bands corresponding to the protected RNA fragments of both the IGF and
cyclophilin were densitometrically quantified using NIH Image Analysis
software (NIH). The IGF-related bands of interest were normalized by
dividing IGF-related mRNA pixel values by cyclophilin values. Both the
mean and SEM were calculated and plotted. Significance
differences in relation to data from either the 0 dpl control or the
contralateral unlesioned hemisphere were examined using single-factor,
ANOVA with a significance level (
) of 5%.
In situ hybridization
Mounted sections were dewaxed in ethanol, rehydrated, washed,
and digested with 10 µg/ml proteinase K in 0.1 M Tris
containing 50 mM EDTA at 37 C for 30 min. Sections were
rinsed in deionized water followed by an incubation in 0.1
M triethanolamine (TEA), pH 8.0, for 3 min. The sections
were then acetylated for 10 min with 0.25% acetic anhydride in 0.1
M TEA for 10 min, rinsed in 2x standard sodium citrate
(SSC), dehydrated through a graded series of ethanol washes, and air
dried under vacuum for 2 h before hybridization. Hybridization
with the 35S-labeled cRNA probe (1 x 107
cpm/ml) was performed at 55 C overnight in 10 mM Tris pH
8.0 containing 50% formamide, 0.3 M NaCl, 1 mM
EDTA pH 8.0, 0.02% polyvinylpyrrolidone, 0.02% Ficoll, 0.02% BSA, 10
mM DTT, 0.05 mg/ml torula yeast RNA (C. P. Laboratories,
Bishops Stortford, Hertfordshire, UK), 0.5 µg/ml tRNA and 10%
dextran sulphate (wt/vol). After hybridization, sections were rinsed
for 1 h in 4x SSC and treated with 25 µg/ml RNase A in 10
mM Tris pH 8.0 containing 0.5 M NaCl and 1
mM EDTA, pH 8.0, at 37 C for 30 min. This was followed by
increasing high-stringency washes of SSC containing 1 mM
DTT, finishing with 0.1x SSC at 65 C for 30 min. Slides were then
dehydrated through a graded series of ethanol, dried under vacuum and
then exposed to Hyperfilm ß-max film for 10 days at 4 C to examine
gross changes in mRNA. For microscopic analysis, slides were exposed to
Ilford K5 liquid autoradiographic emulsion (Ilford Ltd., Basildon,
Essex, UK) for 2 weeks at 4 C, processed with Kodak D19 developer,
rinsed and fixed with Kodak rapid fixer. The slides were rinsed for 30
min in tap water, counterstained with Mayer’s haemalum, examined by
dark-field and bright-field microscopy using a Zeiss Axioscope
microscope, and photographed using Ilford PanF ISO 50 black and white
film.
Immunohistochemistry
All antibodies used were IgG fractions of rabbit polyclonals,
except for IGFBP-5 (IgG fraction of an IGFBP-5 guinea pig polyclonal).
Antihuman IGF-I antibody (TCS Biologicals Ltd., Botolph Claydon,
Buckinghamshire, UK) cross-reacted 10% with IGF-II, and was used at
1:200. Antirat IGFBP-1 antibody (S. Shimasaki) had no known
cross-reaction with other binding proteins and was used at 1:3500.
Antibovine IGFBP-2 antibody (Amano Biologicals, Troy, CA) cross-reacted
0.1% with IGFBP-1, -3, -4, and -5, and was used at 1:3000. Antirat
IGFBP-3 antibody (S. Shimasaki) cross-reacted with no other binding
proteins, and was used at 1:3000. Anti-human IGFBP-4 antibody (TCS
Biologicals) had a cross-reactivity with IGFBP-1, -3, and -5 of
0.1–1% and up to 50% with IGFBP-2, and was used at 1:3000. Antihuman
IGFBP-5 antibody (TCS Biologicals) had a cross-reactivity with IGFBP-2
and IGFBP-3 of less than 1%, and was used at 1:250. Antirat IGFBP-6
antibody (S. Shimasaki) cross-reacted with no other known binding
proteins, and was used at 1:3500. Antihuman IGF-1R antibody (TCS
Biologicals Ltd.) cross-reacted with neither IGF receptor-2 (IGF-2R)
nor the insulin receptor, and was used at 1:3000. Recombinant human
IGF-I, IGFBP-1, and IGFBP-3 were from TCS Biologicals Ltd. Recombinant
IGFBP-2, -4, -5, and -6 were purchased from Amano Biologicals. All
peptides and antibodies were usually stored at -20 C. If in frequent
use, aliquots were stored at 4 C to prevent repetitive freeze/thawing.
All recombinant proteins were stored at -70 C.
Immunoperoxidase staining was performed using the ABC Vectastain Elite kit (Vector Laboratories, Peterborough, Cambridgeshire, UK). The tissue sections were dewaxed in 100% ethanol for 5 min and then rehydrated in 5 min steps in descending concentrations of ethanol to ultra-pure water. Subsequent to a 5 min equilibration in PBS, the endogenous peroxidase was quenched by incubating with 0.01% hydrogen peroxide (H2O2) in PBS for 30 min. The sections were rinsed in PBS and incubated in 1.5% goat serum (vol/vol, Vector Laboratories) diluted in PBS containing 0.1% BSA for 30 min to block nonspecific staining.
After an overnight incubation at 4 C, with the appropriate concentration of growth factor, binding protein or receptor protein-A-purified primary antibody diluted in PBS supplemented with 5% BSA, the sections were treated with a 1:200 dilution of biotinylated goat antirabbit IgG for 1 h, except for the IGFBP-5 antibody treated sections where biotinylated anti-guinea pig IgG was used. This was followed by a 1-h incubation with the Vectastain Elite ABC reagent (Vector Laboratories), a biotin-avidin-peroxidase complex. Finally, the sections were treated for 2–7 min with 0.5 mg/ml 3'3'-diaminobenzidine in PBS containing 0.01% H2O2. All steps were separated by PBS buffer washes. The sections were washed in PBS, counterstained with Mayer’s haemalum, dehydrated, cleared, and mounted. Sections were examined by bright-field microscopy at high and low power magnification, under differential interference contrast (DIC) optics, on a Zeiss Axioscope microscope and photographed using Fujicolor Super G plus ISO 200 color film.
The specificity of the antibodies was assessed by preincubating the primary antibody with excess (>1 µg) of the appropriate homologous or heterologous antigenic growth factor/binding protein. All binding protein antibodies were preincubated with each recombinant binding protein to determine specificity of staining. Sections were also processed with the primary or secondary antibody omitted. All of these controls yielded no visible staining of the sections.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. All data mentioned, but not
presented herein, are available for inspection on written request to
the corresponding author. In every section studied by
immunohistochemistry, all of the IGFBPs localized to the white matter
tracts of the brain, probably associated with myelin. Paradoxically,
only the mRNA of IGFBP-5 was detected in oligodendrocytes at
significant levels. The localization of IGFBP-2 to myelinated tracts
has been catalogued by us in detail elsewhere (9). The significance of
this depot of binding proteins remains to be established but clearly
indicates storage at sites distal to those of synthesis. It seems that
the choroid plexus is a likely source of binding proteins.
Sequestration to specific anatomical sites might occur after transport
via cerebrospinal fluid after secretion from the plexus.
|
). IGF-I mRNA expression was
increased significantly (F>Fcrit at 5% level; Table 2)
in lesioned hemispheres at 5 dpl and 7 dpl, when compared with the
unlesioned hemisphere (0 dpl). There was also a significant difference
(F>Fcrit at 5% level; Table 2) between levels in lesioned, compared
with contralateral unlesioned hemispheres at 5 dpl and 7 dpl.
Expression of IGF-I mRNA peaked at 7 dpl in the lesioned hemisphere. At
10 dpl, levels of IGF-I mRNA in the lesioned cerebrum were not
significantly different from those both of the 0 dpl lesioned and the
contralateral unlesioned hemisphere, in which there was a uniform
pattern of low level mRNA expression. The injury-related changes in
IGF-I mRNA expression within the wounds were visualized by in
situ hybridization and, using morphological criteria, signal was
localized to cellular components of the damaged neuropil, which are
predominantly astrocytes and neurons (Figs. 2 and 3). Endothelial cells of the wound microvasculature also
expressed IGF-I mRNA.
|
|
|
|
).
|
). The
same complementary DNA probe was previously shown by us to detect
IGFBP-1 mRNA in adult rat liver (26).
In contrast to the mRNA studies, IGFBP-1 peptide was observed in the
ependyma, choroid plexus, meninges, the white matter tracts and also in
the cerebral microvasculature of the intact and lesioned brain. After
injury, numerous immunopositive macrophages, neurons, astrocytes and,
in particular, microvasculature endothelial cells were observed in the
wound from 2 dpl, with the number of positive cells becoming maximal
between 5–7 dpl (Table 1).
IGFBP-2
A 397-bp protected mRNA species was identified by RPA
corresponding to IGFBP-2 RNA in all samples of unlesioned and lesioned
hemispheres. Even by this relatively insensitive technique of analysis,
changes in expression within the whole hemisphere were significant
(F>Fcrit at 5% level; Table 2), with IGFBP-2 mRNA increasing at 7 dpl
when compared with the 0 dpl hemispheres (Fig. 5). As
previously reported by us (9), levels of IGFBP-2 mRNA were high in the
meninges and choroid plexus of the intact brain, detected by in
situ hybridization. Some mRNA was also expressed in layers V and
VI of the neocortex. Macroscopic and microscopic increases in IGFBP-2
mRNA were identified in the damaged neural parenchyma, predominately
astrocytes and neurons, identified by morphological criteria. The
increased local levels of IGFBP-2 mRNA were apparent, by the more
sensitive in situ hybridization method, at 2 dpl and were
sustained thereafter until at least 15 dpl. At 2 dpl, IGFBP-2 mRNA
expression was extensive within the damaged neuropil (Fig. 6).
|
|
). Thereafter, the level of
immunoreactivity declined towards that seen in controls.
IGFBP-3
A 699-bp protected mRNA species corresponding to IGFBP-3 was
identified by RPA both in unlesioned and contralateral lesioned
cerebral hemispheres. By RPA, no significant difference in IGFBP-3 mRNA
signal was detected in the lesioned hemisphere compared with either the
contralateral unlesioned hemisphere at any time point or the 0 dpl
animals (data not shown). The more sensitive in situ
hybridization technique demonstrated low levels of IGFBP-3 mRNA
expression in the meninges and choroid plexus in intact brains. Local
increases in IGFBP-3 mRNA expression occurred within cerebral wounds,
with especially prominent levels associated with blood vessels (Fig. 7). The wound response was sustained between 2–10 dpl,
and peaked at 7 dpl. In the intact brain, the spatial and temporal
pattern of IGFBP-3 immunopositive staining in the intact and lesioned
brain was very similar to that of IGFBP-1, with particularly high
levels of protein associated with the cerebral microvasculature, the
response also peaking at 5 dpl (Table 1).
|
) in lesioned
cerebral hemispheres when compared with the contralateral unlesioned
hemispheres (data not shown). In situ hybridization of
sections of the intact brain showed strikingly high levels IGFBP-4
mRNA in the tenia tecta, indiseum griseum, choroid plexus, and layers
II, V and VI of the neocortex (not shown). No macroscopic or
microscopic changes in IGFBP-4 mRNA were identified within the neuropil
of the injured cerebrum (Table 1).
In the intact brain, the distribution of IGFBP-4 peptide was very
similar to that of other binding proteins, with the ependyma, choroid
plexus, meninges, and myelinated nerve tracts staining. Similarly,
after injury IGFBP-4 immunopositive astrocytes, neurons, microglia and
macrophages appeared in the damaged neuropil; this response becoming
maximal between 7–10 dpl (Table 1).
IGFBP-5
Rat IGFBP-5 cRNA probed RPAs demonstrated a 300-bp protected mRNA
species corresponding to IGFBP-5 mRNA in all RNA samples of unlesioned
and lesioned cerebral hemispheres. RPA revealed no significant
differences in the expression of IGFBP-5 mRNA in any of the animals
studied (data not shown). In situ hybridization of the
intact brain localized IGFBP-5 mRNA to blood vessels, meninges, choroid
plexus, subependymal plate and, unique for IGFBP-5, myelinated nerve
tracts. By this more sensitive method, a focal increase in IGFBP-5 mRNA
expression was strikingly prominent in the developing glia limitans of
the wound after 7 dpl (Figs. 8 and 9).
|
|
).
IGFBP-6
A 246 bp protected mRNA species was identified by RPA
corresponding to IGFBP-6 mRNA, and no significant differences in levels
were observed between unlesioned and lesioned cerebral hemispheres (not
shown). In situ hybridization of the intact brain localized
low levels of IGFBP-6 mRNA to the meninges and choroid plexus; however,
there were no macroscopic or microscopic changes apparent in IGFBP-6
mRNA within the damaged cerebral neuropil at any time point (Table 1).
In contrast to the mRNA, there were significant levels of IGFBP-6
peptide in the meninges, choroid plexus, ependyma, and subependymal
layers of the intact hemisphere. Immunopositive macrophages, neurons,
astrocytes, and microglia were also observed at all time points in the
immediate area of neuropil around the lesion, with labeled neurons,
astrocytes and microglia being most abundant at 5 dpl (Table 1).
IGF-1 receptor
A 265-bp protected mRNA species corresponding to IGF-1R was
present in all RPA samples. This expression was seen to be ubiquitous
through the hemisphere by in situ hybridization, with the
highest levels found in areas where high levels of IGF-I mRNA also
occurred. There were no significant increases in IGF-1R mRNA in any of
the lesioned animals observed by either RPA or in situ
hybridization (Table 1). In intact and lesioned cerebral hemispheres
there was a pattern of IGF-1R peptide localization similar to that of
IGF-I peptide. Immunoreactivity was visible in the ependyma, meninges,
choroid plexus and, after injury, in cortical neurons, glia and
macrophages, with a strong reaction also observed in the endothelial
cells of the brain microvasculature (Table 1).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Modulation of IGF-I action is complex, occurring at the levels of transcription, translation, and ligand activation. IGF availability can also be regulated by IGFBPs. Because no significant changes were observed in the expression of IGF-1R mRNA after injury, altered receptor transcription may not be a prime determinant of IGF-I bioactivity within CNS wounds. Presumably, receptor turnover on target cells is sufficient to meet the needs of both physiological and pathological responses. Discrete focal and transient increases in transcription and localization of IGF-I and IGFBPs within injury-responsive cells indicate ligand expression and bioavailability as prime determinants of IGF-I bioactivity.
IGF-I and IGF-1R expression in cortical wounds
IGF-I mRNA transcripts in the same size range were identified in
intact and lesioned brains by RPA (450, 500, and 520 bp). The
expression of these multiple transcripts has been demonstrated
previously (35, 36, 37, 38), the variations being contributed to by the
differing use of polyadenylation sites in exon 6 (39). In agreement
with previous reports (6, 21, 40), the intact rat brain showed
ubiquitous low levels of expression of the IGF-I mRNA signal by
in situ hybridization, with some discrete anatomical foci of
higher expression. The pattern of expression accurately reflected that
seen for IGF-1R [confirming the work of others (6, 41)], with both
ligand and receptor synthesis colocalized to specific sets of neurons,
suggesting both a paracrine and autocrine activity of IGF-I in the
intact brain. In general, neurons and astrocytes of the intact cortical
neuropil showed little or no IGF-I or IGF-1R peptide.
IGF-I mRNA and peptide increased coordinately within the same injury-responsive cells after wounding. Although the choroid plexus does contain immunoreactive IGF-I, little or no IGF-I is measurable in the cerebrospinal fluid by RIA after injury (unpublished observations). Hence, this distal site does not seem to be a primary source of IGF-I for injury-responsive cells in the wound. The discrete IGF-I mRNA and peptide response is acute, substantial, restricted to the lesioned area, and its peak coincides with the time of both maximal astrocyte reactivity and active axonal sprouting. This suggests that IGF-I acts in wounds in both an autocrine and/or paracrine manner. Because IGF-I is reported to stimulate both gliosis and neurite outgrowth in vivo and in vitro, acute phase gliogenic and neurotrophic roles within CNS wounds are implicated. The transient nature of IGF-I activation is evidenced by the declining levels of both mRNA and protein observed within wounds as the cellular responses proceed towards scar maturation and cessation of axonal regeneration. The rapid decline in IGF-I levels may reflect restoration of tissue homeostasis, which may be enhanced by changed patterns of expression of specific IGFBPs.
IGF-I and IGF-1R in the microvasculature
The strong localization of IGF-I and IGF-1R mRNAs and peptides to
endothelial cells of the brain microvasculature may be associated with
transport across the blood brain barrier (20) and colocalization to
highly vascularized sites (42). High concentrations of IGF-I are found
in the serum and, because CNS wounds are the focus of an angiogenic
response, the presence of IGF-I in capillary endothelial cells suggests
a possible route of entry from the blood into the lesion. However, the
expression of IGF-I mRNA by blood vessels in the wound suggests that
de novo synthesis by endothelial cells, as well as import
from the blood, might contribute to the focal increase of IGF-I in the
wound microvasculature. Interestingly, an angiogenic activity of IGF-I
has also been demonstrated (43).
IGFBP expression in cortical wounds
The bioavailability of IGF-I within wounds is probably modulated
by IGFBPs and documentation of the spatial and temporal response of
IGFBPs within wounds offers a first step in the elucidation of their
function in brain injury responses. Functional interaction between
IGFBPs and IGF-I is suggested by both their colocalization and
coordinate modulation of expression. Previous studies with other models
of CNS damage suggest that IGFBP-1, -2, -4, -5, and -6 may all
contribute to injury responses (2, 44, 45, 46, 47).
In most cases, our observations of the anatomical distribution of IGFBP mRNA expression in the intact adult rat brain correlated with previous studies. Whereas the mRNA for IGFBP-2 through -6 were all expressed in the choroid plexus, perhaps indicative of their secretion into the CSF, individual binding proteins showed distinct anatomical patterns of gene expression and peptide localization elsewhere within the brain, suggestive of distinct physiological roles. For example, IGFBP-3 and IGFBP-5 mRNA expression was localized to blood vessels, having perhaps a role in modulating IGF efflux into the brain from the circulation. In contrast, IGFBP-4 mRNA showed an extremely precise pattern of high expression in specific clusters of neurons (such as the tenia tecta and indiseum griseum) indicating possible roles in neuronal function. However, at present, any physiological function for the IGFs and their binding proteins in the normal adult brain remains speculative.
Our study demonstrated many instances of the presence of IGFBP peptide in the absence of significant levels of mRNA, e.g. in myelinated tracts throughout the brain, where high levels of alI IGFBPs were found colocalized with expression of the IGFBP-5 mRNA only. Similarly, little or no IGFBP-1, IGFBP-4, or IGFBP-6 mRNAs, but significant levels of immunoreactive IGFBP proteins were seen in injury-responsive cells within wounds. As IGFBP-2 through -6 mRNAs are expressed in the choroid plexus epithelium, from where they can be secreted into the CSF, transport through ventricles and subarachnoid spaces to sites of sequestration and activity in brain parenchyma is a possibility. Whether localization at sites distal to synthesis supports this transport hypothesis or simply reflects protein accumulation in areas of very low local mRNA expression is the subject of a subsequent study to be reported by us elsewhere. Certainly, after a penetrating CNS injury, IGFBPs may concentrate in the wound from distal as well as local sources.
IGFBPs in the microvasculature
The colocalization of IGFBP-1, -3, and -5 with IGF-I to
endothelial cells of the brain microvasculature suggest their possible
involvement in the modulation of IGF-I efflux across the blood/brain
barrier. Because IGFBP-1 may not be synthesized at all within the brain
and is capable of crossing the endothelial cell border (48), it may
function to transport IGF-I from the blood across the blood/brain
barrier. Interestingly, two isoforms of IGFBP-1 have been identified,
one which associates with the cell membrane, and a soluble species
which can inhibit IGF-I bioavailability in vitro, suggesting
alternative roles for IGFBP-1 peptide. IGFBP-1 is known to be involved
in adhesion cascades, by binding via its RGD sequence to
5ß1-integrins (49). Integrins have been implicated in the wounding
cascade in the same model as used in these experiments (M. Berry,
unpublished observations). Hence, the IGFBP-1/IGF-I complex may be
transported across the endothelial cell barrier into the wound, where
it may subsequently bind to the extracellular matrix. In this
reservoir, IGF-I may be cleaved from IGFBP-1 by locally produced
proteases, allowing its interaction with other IGFBPs or IGF-1R
both on the endothelium and on other injury-responsive cells in the
wound. As increased levels of matrix are deposited in the maturing
wound and proteolysis subsides, locally bound IGFBP-1 may serve to
sequester and thereby limit IGF-I availability within wounds. The
appearance of IGFBP-1 protein at the wound site over the 2–5 dpl
period suggests rapid mobilization and its persistence in the lesion is
consistent with the above arguments. In contrast, the striking, rapid
and transient (2–5 dpl) increase in IGFBP-3 mRNA expression observed
within blood vessels of the injured cerebrum, is suggestive of
regulated expression, perhaps consolidating the early IGF-I
potentiating activity of IGFBP-1 and mediating an acute phase,
injury-related efflux of IGF-I from the circulation. Hence, in the
injured brain, IGFBP-3 may also play an important role in regulating
IGF-I transport to, and localization within, wounds via the
extracellular fluid. The observed late (5–10 dpl) localization of
IGFBP-5 protein to endothelial cells of the wound microvasculature,
together with its reported inhibitory effects on IGF action in other
in vivo and in vitro systems indicates a role in
chronic IGF-I sequestration, thereby aiding homeostasis recovery by
down-regulating IGF-I efflux and bioactivity.
IGFBPs within the damaged neuropil
All six IGFBPs accumulate in penetrant CNS wounds, but IGFBP-2
predominates. Whereas IGFBP-1 and IGFBP-3 seem to be associated
predominately with the responding wound microvasculature, IGFBP-2, -4,
-5, and -6 proteins localize strongly to astrocytes and neurons. Our
observations of the spatial and temporal changes in IGFBPs suggest that
whilst IGFBP-2 and IGFBP-6 predominantly promote IGF-I activity in the
acute phase of the injury response, IGFBP-4 and IGFBP-5 may be
inhibitory, facilatating IGF-I down-regulation as the cellular wounding
response subsides.
In cerebral wounds there was a significant, rapid, and precise focal increase in IGFBP-2 protein and mRNA expression, which correlated spatially and temporally with IGF-I expression. Because astrocytes, neurons, and macrophages all coexpressed the peptides, a functional relationship between IGF-I bioactivity and this binding protein is plausible. IGFBP-2 potentiates IGF activity (7, 49) and thus could enhance the acute phase effects of IGF-I on target cells, either by altering the half-life or facilitating ligand presentation to the receptor. Although IGFBP-6 gene expression did not change detectably after injury, macrophages, neurons and astrocytes also became strongly immunopositive in the acute phase of the response, suggesting a potential role in enhancing IGF-I bioactivity in the early phases of the wounding response.
Current evidence suggests that IGFBP-4 acts predominantly to suppress IGF activity (7, 49). In these cerebral lesions, the levels of IGFBP-4 peptide expression increased chronically, peaking between 7–10 dpl, and we suggest that the late activation of this inhibitory IGFBP may aid IGF-I down-regulation to regain tissue homeostasis. The increase of IGFBP-5 expression also occurred relatively late in the wounding response (only apparent after 7 days) and was extremely focal, with IGFBP-5 mRNA and peptide tightly colocalized to the astrocytes comprising the reforming glia limitans. This region is matrix-rich and, because IGFBP-5 is known to interact with the extracellular matrix (particularly heparan sulphates) and is inhibitory to IGF action (7, 49), this binding protein may also help to sequestrate and down-regulate IGF-I activity within wounds as the cellular response subsides.
We conclude from these studies on IGF-I, IGFBPs and IGF-1R that cellular responses to brain injury may be regulated by IGF-I under the influence of several IGFBPs, because both are colocalized within injury-responsive cells. It is possible that IGF-I functions as a local autocrine/paracrine factor contributing to the acute phase injury responsiveness of astrocytes and neurons. The temporal and spatial availability of the ligand, and therefore its bioactivity, may be regulated by the differential presence of stimulatory vs. inhibitory IGFBPs.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received December 20, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
5ß1 integrin by means of its
Arg-Gly-Asp sequence. Proc Natl Acad Sci USA 90:10553–10557This article has been cited by other articles:
 |
|
 |
|
  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]
|
|
|
|
 |
|
 C. Lopez-Lopez, D. LeRoith, and I. Torres-Aleman Insulin-like growth factor I is required for vessel remodeling in the adult brain PNAS, June 29, 2004; 101(26): 9833 - 9838. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. J. Walter, M. Berry, D. J. Hill, S. Cwyfan-Hughes, J. M. P. Holly, and A. Logan Distinct Sites of Insulin-Like Growth Factor (IGF)-II Expression and Localization in Lesioned Rat Brain: Possible Roles of IGF Binding Proteins (IGFBPs) in the Mediation of IGF-II Activity Endocrinology, January 1, 1999; 140(1): 520 - 532. [Abstract] [Full Text]
|
|
| |||||||||||||||
, unlesioned hemisphere; •, lesioned hemisphere; *,
P < 0.05 with respect to 0 dpl sample.
