autoradiographic and immunohistochemical localization of insulin-like growth factor-i receptor...

General and Comparative Endocrinology 141 (2005) 203–213

www.elsevier.com/locate/ygcen

Autoradiographic and immunohistochemical localization of insulin-like growth factor-I receptor binding sites in brain

of the brown trout, Salmo trutta �

Alastair Smith a, Shu Jin Chan b, Joaquim Gutiérrez c,¤

a Institute of Veterinary, Animal and Biomedical Sciences, Massey University, Private Bag 11222, Palmerston North, New Zealandb Departments of Biochemistry, and Molecular Biology and Medicine, The Howard Hughes Medical Institute, University of Chicago,

Chicago, IL 60637, USAc Departament de Fisiologia, Facultat de Biologia, Universitat de Barcelona, Avinguda Diagonal 645, 08028 Barcelona, Spain

Received 10 March 2004; revised 26 November 2004; accepted 16 December 2004

Abstract

Insulin-like growth factor-I (IGF-I), a peptide closely related to insulin, is known to play crucial roles in brain development.While the central sites of action of IGF-I in higher vertebrates are now well established, surprisingly little is known in the teleostmodel where the brain undergoes continual, indeterminate, growth. In this study, we have mapped the distribution of putative IGF-Ireceptor (IGF-IR) binding sites in the brain of the brown trout using both ligand binding in vitro autoradiography and immunohis-tochemistry. The presence of IGF binding proteins (IGFBPs) was further studied by competitive inhibition using unlabelled IGF-Iand des-(1-3)-IGF-I. In both juvenile and adult trout brain, [125I]IGF-I binding was highest in cerebellum and optic tectum, bothregions of the teleost brain known to grow the most actively throughout life. At the cellular level, IGF-IR immunoreactivity was con-Wrmed on cell bodies and dendrites, particularly of larger presumptive neurons including purkinje cells and dendritic Wbres through-out the molecular layer of the cerebellum. Abundant IGF-IR expression in hypothalamic regions may further be related to neurongrowth while a possible hypophysiotropic role will require further investigation. Competitive inhibition studies employing des-(1-3)-IGF-I also suggest IGFBPs are present in all regions exhibiting high [125I]IGF-I ligand binding and conWrms the presence of thisimportant regulatory component of the IGF-I system in the teleost brain. The importance of the IGF-I system in brain development,particularly in regions such as the cerebellum, together with the continual lifetime growth of the Wsh central nervous system, suggestthe teleost brain is an extremely useful site for studying the actions of IGF-I in relation to neuron proliferation, growth, and survivalin an adult brain. 2005 Elsevier Inc. All rights reserved.

Keywords: Teleost Wsh; Cerebellum; Optic tectum; IGF-I receptor; IGFBP

1. Introduction

Insulin-like growth factor-I (IGF-I) is a polypetidehormone closely related to insulin whose structure has

� Grant sponsor: European Union Marie Curie FellowshipGT973399 to A.S.

¤ Corresponding author. Fax: +34 934110358.E-mail address: [email protected] (J. Gutiérrez).

0016-6480/$ - see front matter 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.ygcen.2004.12.023

been highly conserved over vertebrate evolution (Chanet al., 1992; LeRoith et al., 1993; Upton et al., 1997,1998). In the central nervous system (CNS) of highervertebrates, it has been shown to play a variety of keyroles in the regulation of brain growth and development(Anlar et al., 1999; de Pablo and de la Rosa, 1995;D’Ercole et al., 1996). While IGF-I is a potent stimulatorof proliferation in various neuronal cell types, it mayalso be involved in neuronal migration, promote neurite

mailto: [email protected]

mailto: [email protected]

204 A. Smith et al. / General and Comparative Endocrinology 141 (2005) 203–213

outgrowth, and protect from apoptosis (Fernández et al.,1999; Leventhal and Feldman, 1997; Mozell andMcMorris, 1991; Torres-Aleman et al., 1992; Ye et al.,1996). Most of these diverse regulatory activities beginwhen the peptide binds to its speciWc cell surface recep-tor. Like the peptide itself, both the tetrameric structure,composed of two �- and two �-subunits, and the tyrosinekinase activity of the IGF-I receptor (IGF-IR), havebeen further highly conserved during vertebrate evolu-tionary history (Navarro et al., 1999). The peptide bindsto the extracellular �-subunits of the receptor whichbrings about tyrosine kinase activity in the intracellular�-subunits. This leads to further signalling cascades andthe ultimate physiological actions of IGF-I (LeRoithet al., 1995). In addition to the IGF-I peptide itself, thereare a number of IGF binding proteins (IGFBPs) whichregulate levels of freely available IGF-I able to bind withits receptor (Jones and Clemmens, 1995). The spatialexpression of both the IGF-I receptor and the IGFBPscan therefore provide good Wrst indicators as to thelikely sites and actions of this peptide within the centralnervous system.

The regional distribution of IGF-I receptor bindingsites in the mammal and chick brain is now well estab-lished (Bassas et al., 1989; Bohannon et al., 1988b; Karet al., 1993; Lesniak et al., 1988; Werther et al., 1987).During the rapid neural development of embryogenesis,these receptors are highly expressed in widespreadregions of the brain. Thereafter, IGF-IR expressiondeclines signiWcantly in many of these brain regions lead-ing to the more restricted distribution of binding sitesfound in adulthood. The presence of speciWc IGF-Ireceptors in the teleost brain has also recently been dem-onstrated (Drakenberg et al., 1993; Leibush et al., 1996).However, these studies have concentrated solely on char-acterization of receptor binding in whole brain homoge-nates and currently very little is known of the regionaldistribution of IGF-IR expression in the Wsh brain.

We have used a combination of both quantitative invitro receptor autoradiography and immunohistochem-istry to map putative IGF-I receptor binding sites in theearly juvenile and adult trout brain. The presence ofIGFBPs was also investigated using competitive inhibi-tion studies incorporating des-(1-3)-IGF-I which has areduced aYnity for the IGFBPs compared to the nor-mal, non-truncated, form of IGF-I. The aim of thesestudies has been to identify the likely central sites ofaction of IGF-I in an animal which is at an early stage ofevolution. Furthermore, in marked contrast to mammalsand birds, a unique feature of the teleost CNS is that itcontinues to grow throughout life, including much ofadulthood. In view of the importance of these peptidesto neuronal development, we particularly wished to con-sider these receptor binding proWles in relation to thisindeterminate growth and the likely sites of action in theadult brain.

2. Materials and methods

2.1. Animals

Juvenile (20 weeks post-hatching) (n D 10) and adult(3 years old) (n D 15) Brown trout (Salmo trutta) wereobtained from the Wsh farm Piscifactoria de Bagá, Barce-lona, Spain. Fish were anaesthetized in an overdose oftricaine methanesulfonate (100 mg/L) (MS222, Sigma,St. Louis, MO) and sacriWced by exsanguination. All ani-mals were treated in accordance with European Unionexperimental animal protection regulations.

2.2. In vitro receptor autoradiography

Brains were rapidly excised and frozen in isopentaneon ice and maintained at ¡80 °C until use. Cryostat sec-tions (20 �m) were cut in sagittal and coronal planes,thaw mounted onto polylysine coated slides, and storedovernight with dessicant at 4 °C prior to use. Sectionswere preincubated with PBS, 30 mM KCl, for 15 min atroom temperature to remove endogenous bound pep-tide. Incubations with radiolabelled hormone, with orwithout competing ligand, were performed in 30 mMHepes buVer containing 0.1% of BSA and 100 U/ml ofbacitracin (pH 7.4). Binding assays were performedusing [125I]IGF-I (human recombinant), speciWc activityof 2000 Ci/mmol, purchased from Amersham Life Sci(Little Chalfont, Buckinghamshire, UK). A similar aYn-ity of human recombinant and Wsh IGF-I for the IGF-IR in Wsh tissues has previously been demonstrated(Gutiérrez et al., 1995; Leibush et al., 1996). Commer-cially radiolabelled human recombinant IGF-I has alsobeen used for binding and proliferation studies in thegoldWsh retina (Boucher and Hitchcock, 1998a,b). Unla-belled IGF-I was a gift of Chiron Therapeutics, unla-belled IGF-I and all other chemicals were purchasedfrom Sigma Chemicals (St. Louis, MO, USA).

Optimization studies included timecourse binding of[125I]IGF-I at intervals of 1–24 h, and saturation studiesperformed with [125I]IGF-I in the range 25–250 pM. Pre-vious binding studies with trout brain membranes haveestablished greater speciWc binding at 4 °C (Leibushet al., 1996), all characterization and localization studieswere therefore performed at this temperature. For locali-zation incubations, sections were incubated in optimalconcentrations of radiolabelled ligand in the range 75–100 pM for 16–18 h. SpeciWcity of binding was demon-strated by parallel incubations of adjacent sections in thesame concentration of radioligand plus a 1000-foldexcess of unlabelled IGF-I. The presence of IGFBPs wasalso studied by competitive inhibition experiments usingdes-(1-3)-IGF-I (GroPep, Adelaide, SA, Australia).After incubation the sections were rinsed in three 5 minchanges of ice-cold PBS, rinsed brieXy in ice-cold dis-tilled water, and rapidly dried under a stream of cold air.

A. Smith et al. / General and Comparative Endocrinology 141 (2005) 203–213 205

Labelled and dried sections were then apposed to Hyper-Wlm 3H (Amersham) along with 125I-polymer standards(Amersham International, UK) in an X-ray cassette for10–14 days. Following exposure, the Wlm was developedin Agfa G150 (Agfa Gevaert N.V., Belgium) at roomtemperature for 2 min.

2.3. Imaging and quantiWcation

Films were scanned on a Agfa DUOSCAN T2000 XLusing Agfa FotoLook 32 Version 3. Images wereimported into Adobe Photoshop Version 5 (Adobe Sys-tems, San Jose, CA) and surrounding backgroundremoved, no other alterations were made. Optical densi-ties on developed Wlm were determined from the digi-tized images using SigmaScan Measurement Software(Jandel ScientiWc) and converted to fmol/mg proteinaccording to Nazarali et al. (1989). Values are means of6 £ 100 �m2 measurements (§SEM) per section on totalbinding autoradiograms with non-speciWc binding andbackground levels in corresponding regions subtracted.Replicate measurements were performed on three repre-sentative sections in midbrain per individual, a total offour brains were used with sections from each used intwo separate experiments. For analysis of comparativeinhibition of [125I]IGF-I binding by des-(1-3)-IGF-I andunlabelled IGF-I, the same optical density samplingstrategy was used and values retained in density format.Statistical comparisons were performed using ANOVAand post hoc two-sample Student t test’s on the means, Pvalues equal to or lower than 0.05 were considered statis-tically signiWcant. A graded tier system, comprising + to++++ (low to high, respectively) was also adopted forbinding levels to provide a simpliWed table for rapididentiWcation of relative binding intensities. Minimumand maximum optical density scores were placed intofour bins of equal range, raw optical densities did notadd detail to the Wnal analysis.

Following autoradiography, sections were stainedwith eosin and hematoxylin and anatomical regionsidentiWed with reference to a trout brain atlas (Meek andNieuwehuys, 1998), nomenclature used for brain nucleifurther follows that of Meek and Nieuwehuys (1998).

2.4. Immunohistochemistry

Immunohistochemistry for IGF-IR was performedby the avidin-biotin-immunoperoxidase technique, asdescribed previously (Cheng et al., 1998). Cryostat sec-tions were Wxed in 4% formaldehyde for 10 min. Afterquenching in 3% H2O2 for 10 min, tissues were blockedin 10% normal goat serum in 1£ PBS for 1 h, followedby incubation with primary antibodies at 4 °C overnight.The anti-IGF-IR polyclonal antibody (ZIR7) (corre-sponding to the C-terminal domain of ZebraWsh IGF-IRb) (Maures et al., 2002) was used at »2 �g/ml. The

eVectiveness of this antibody for immunoprecipitationand immunohistochemistry has previously been demon-strated (Maures et al., 2002). These same antibodies havealso been used eVectively for immunohistochemicalanalysis in Gilthead Seabream. Furthermore, alignmentof deduced amino acid sequences for the correspondingregions of the zebraWsh and trout IGFIR shows 90%homology (pairwise protein BLASTP, protein codes:AAM18901.1, ZebraWsh and AAC16493.1, Trout),Ayaso et al. (1999) also report a 90% predicted aminoacid sequence for the zebraWsh and trout IGFIR tyro-sine kinase domain. It is therefore reasonable that theantibody will also work in trout and salmon. After wash-ing, sections were incubated in biotinylated secondaryantibodies (1:400) for 30 min. The signal was detected

Fig. 1. Psuedocolour images of [125I]IGF-I binding on coronal sectionsof Brown trout (S. trutta) brain at the level of the mid-mesencephalicregion. (A) Juvenile (20 weeks post-fertilization) and (B) adult of 3years. Incubation conditions were 100 pM [125I]IGF-I for 18 h at 4 °C.Shown are total binding levels, adjacent sections incubated in [125I]IGF-I plus a 1000-fold excess of unlabelled IGF-I conWrm these sites repre-sent speciWc IGF-I binding. A high density of binding in both juvenileand adult is consistently seen in stratum griseum centrale of optic tec-tum (Otec), molecular layer of the cerebellum (CCm) (here at the levelof the valvula of the cerebellum) and granular layer of cerebellum(CCg). Moderate binding is present in lobus inferioris of the hypothala-mus (lih) and tori lateralis (ntlat) (both juvenile and adult). Only low,almost background, levels of binding are seen in midbrain tegmentalstructures such as torus semicircularis (ts). Scale bar, 2.5 mm.


and ampliWed using the ABC peroxidase method (Vec-tor, Burlingame, CA) and visualized with 3,3�-diam-inobenzidine. Controls for the immunohistochemistryprocedure were performed by omission of primary anti-body in the incubation and substitution with blockingsolution, and processed in parallel with the experimentalgroups. Sections were counterstained with 0.2% methylgreen for 5 min, brieXy washed in distilled water beforedehydration in graded alcohols, xylene and mountedwith permount and coverslips.

3. Results

3.1. Characterization studies

The characterization of [125I]IGF-I binding in thebrain of brown trout has previously been reported indetail indicating the presence of speciWc, biologicallyactive, IGF-I receptors in this tissue (Leibush et al.,1996). Our additional optimization studies also foundspeciWc binding of [125I]IGF-I to be localized, time-

dependent, saturable and inhibited in the presence ofexcess concentrations of unlabelled IGF-I. Time-depen-dent speciWc binding reached equilibrium after approxi-mately 19 h and thereafter remained stable, saturationwas achieved at approximately 150 pM. In regions ofhighest speciWc binding, mean values for theoreticalmaximum receptor number, R0, and equilibrium dissoci-ation constant, Kd, were 370 § 38 fmol/mg protein and0.26 § 0.14 nM, respectively.

3.2. In vitro autoradiography

In both the 20-week-old juvenile, and adult, highestbinding of [125I]IGF-I is found in optic tectum andmolecular layer of the cerebellum (Fig. 1). Binding of[125I]IGF-I is also found in nucleus tori lateralis (ntlat)and lobus inferioris of the hypothalamus (lih) of boththe 20 week old and adult brain. In the adult, speciWcityof [125I]IGF-I binding sites is observed at all levels of thebrain (Fig. 2), densitometry scorings are shown in Table1. Highest and most consistent binding of [125I]IGF-I isagain found in molecular layer of the cerebellum and

Fig. 2. Binding of [125I]IGF-I in adult trout (S. trutta) brain at the levels of olfactory bulb (A and B), preoptic region (C and D), midbrain tegmentum(E and F), caudal midbrain region (G and H), corpus of cerebellum (I and J), medulla oblongata (K and L). Levels of sections as shown (M). Shownare total binding ([125I]IGF-I alone) and non-speciWc binding ([125I]IGF-I plus a 1000-fold excess of unlabelled IGF-I). Only a low density of bindingcan be seen in olfactory bulb. A high density of binding can be seen in optic tectum (Otec) and molecular layer of cerebellum (CCm) at all levels, aswell as tractus opticus dorsalis (tod) and lobus caudalis cerebelli (lccb). Moderate binding is seen in granular layer of cerebellum (CCg), lobus inferi-oris of hypothalamus (lih), and nucleus tori lateralis (ntlat). Low to background binding is present in midbrain tegmental structures such as torussemicircularis (ts). Scale bar, 2.5 mm.


optic tectum (Figs. 2C–I). Moderately high binding isobserved in the stratum glomerulosum of the olfactorybulb (stgl) (Fig. 2A) and tractus opticus dorsalis (Fig.2D). Other regions displaying moderate [125I]IGF-Ibinding are lobus inferioris hypothalami (lih) (Fig. 2F),and torus lateralis (ntlat) (Fig. 2G). Binding in midbrain

Table 1Summary of the distribution of [125I]IGF-I binding sites in adult(3 years) brown trout (Salmo trutta)

Nomenclature follows Meek and Nieuwehuys (1997). Computer densi-tometry scoring of speciWc binding: ++++, high; +++, moderatelyhigh; ++, moderate; +, low; 0, not detectable; ?, unknown.

Structure [125I]IGF-I

TelencephalonOlfactory bulb

Stratum glomerulosum ++Stratum granulare 0/+

Dorsal telencephalon ++Ventral telencephalon 0/+

Preoptic areaN. parvocellularis pars anterior ++N. preopticus magnocellularis ++Chiasma opticum 0/+Optic nerve +Tractus opticus dorsalis +++N. parvocellularis pars posterior ++

Thalamus and pretectal regionCentral pretectal nucleus ++Thalamus dorsalis +++Thalamus ventralis +++

HypothalamusN. anterior tuberis +++N. recessus lateralis ++N. lateralis tuberis ++N. diVusus lobi inferioris hypothalami +++

PituitaryPars distalis ++++Pars intermedia 0

Tegmentum mesencephaliN. fasciculi longitudinalis medialis ?N. oculomotorius ?Torus semicircularis 0/+

Tectum mesencephaliStratum periventriculare ++Stratum album centrale +++Stratum griseum centrale +++Stratum griseum et album superWciale +++Stratum marginale ++

Valvula cerebelliGranular layer ++Molecular layer ++++Ganglionic layer ++++

Corpus cerebelliGranular layer ++Molecular layer ++++Ganglionic layer ++++

Lobus caudalis cerebelli ++++

Saccus vasculosus ++++

tegmentum, including torus semicircularis (ts) (Fig. 2E)was at low, almost background, levels. [125I]IGF-I bind-ing in optic chiasm appeared to be largely non-speciWc.

3.3. Presence of IGF binding proteins

Comparative displacement of [125I]IGF-I binding byunlabelled IGF-I and des-(1-3)-IGF-I in the major bind-ing sites of the trout brain is shown in Fig. 3. Displace-ment by des-(1-3)-IGF-I is slightly lower compared toIGF-I in all regions studied suggesting IGFBPs are pres-ent in all of these areas.

3.4. IGF-IR immunohistochemistry

Results of IGF-IR immunohistochemistry agree wellwith those of radioligand binding. In optic tectum,strongest immunoreactivity is found in cell bodies anddendrites of large presumptive neurons. Omission ofprimary antibody in control sections shows immunore-activity is not due to endogenous peroxidase activity(Fig. 4). At higher magniWcation immunoreactivity isseen in stratum griseum centrale (SGC) (Figs. 5A andB) as well as neuronal cell bodies and Wbres within thestratum Wbrosum et griseum superWciale (SFGS) (Fig.5A). In the stratum marginale (SM), however, glial cellsare largely devoid of immunostaining (Fig. 5C). In bothvalvula and corpus of the cerebellum, pronouncedimmunostaining is seen over purkinje cell bodies anddendritic Wbres extending into the molecular layerclearly (Fig. 6). A particularly discrete cytoplasmicimmunostaining is also seen in the saccus vasculosus(Fig. 7A), at higher magniWcation this appears to beimmunostaining over cerebrospinal Xuid (CSF)-con-tacting coronet cells (Fig. 7B).

Fig. 3. Competitive inhibition of [125I]IGF-I binding by unlabelledIGF-I and des-(1-3)-IGF-I in various regions of adult trout brain.Bars represent mean optical density measurements (arbitrary units)(§SEM) made from video digitized images of autoradiographs. Therelatively lower displacement by des-(1-3)-IGF-I compared to unla-belled IGF-I suggests the presence of IGF binding proteins in all brainregions studied.


4. Discussion

SpeciWc receptors for IGF-I have previously beenreported in membrane preparations of the trout brain(Leibush et al., 1996). In the present study, we have usedin vitro autoradiography and immunohistochemistry toprovide for the Wrst time a detailed mapping of the likelyregional distribution of these receptor binding sites inthe teleost brain. The results show a widespread distribu-tion of putative IGF-I receptors which would suggestfundamental roles for IGF-I in the vertebrate brain evenat an early stage of evolution. The localization of thesesites in particular discrete anatomical structures in bothjuvenile and adult further correlates well with the contin-ual indeterminate growth of the Wsh brain. Evidence of

Fig. 4. IGF-IR (ZIR-7) immunoreactivity (brown) in optic tectum ofadult brown trout against methyl green counterstain. (A) Viewthrough deeper most layers of optic tectum, (B) negative control withomission of primary antibody shows IGF-IR immunostaining is spe-ciWc. Vt, ventricle; Otec, optic tectum. Scale bar, 100 �m. (For interpre-tation of the references to color in this Wgure legend, the reader isreferred to the web version of this paper.)

IGF binding proteins in the teleost brain suggests thatthis important regulatory component of the IGF-I sys-tem also had early origins in this region of the vertebratecentral nervous system.

A signiWcant diVerence between the central nervoussystem of teleosts and that of higher vertebrates is thatthe Wsh brain continues to grow, albeit with possibleseasonal variations, throughout life, including much ofadulthood. Whereas in mammals and birds, postnatalneurogenesis appears to be kept at low rates, in Wsh thepropensity for producing new brain cells during adult-hood is very pronounced (Zupanc and Horschke,1995). In particular, the optic tectum and cerebellumare regions of the teleost brain which can grow mostprofoundly. The localization of putative IGF-I recep-tor binding sites in optic tectum could suggest a role invisual processing. However, there is increasing evidencethat the optic tectum in Wsh may also integrate inputsfrom a number of senses (Meek, 1983). In view of theknown actions of IGF-I, it seems more likely that thepresence of such sites in this region reXects the continu-ally growing nature of the optic tectum in Wsh. A con-tinual addition of new terminals and shifting of olderterminals to the tectum has been demonstrated in bothlarval and adult Wshes (Raymond and Easter, 1983). Inaddition, the observation of higher radioligand bindingand IGF-IR immunoreactivity in the ganglionic andperikarya rich layers of the optic tectum, may furthersuggest IGF-I plays a role in cell proliferation and tec-tal growth. Binding of [125I]IGF-I has also been foundin optic tectum of E16 chick which, along with retina,piriform cortex, and pineal, is a major site of IGF-Ireceptor mRNA expression during early rapid stages ofchick development (upto E20) (Holzenberger et al.,1996). Recent localization and organ culture studies(Boucher and Hitchcock, 1998a,b) have also showninsulin, IGF-1, and IGF-2 to stimulate proliferation ofretinal progenitor cells in the goldWsh. Although boththe optic tectum and eye of the Wsh continue to growthroughout life, the retinotectal map of neurons is rig-idly maintained through continual cell diVerentiation(Cook et al., 1983; Easter and Stuermer, 1984; Stuermerand Easter, 1984). It is therefore possible that IGF-Imay further play a role in such diVerentiation. It isinteresting to note that both tectum and retina canexert neurotrophic and guiding inXuences on eachother through various growth-modulating interactionsbetween the two structures (Cronly-Dillon and Sta-

Vord, 1986). This further raises intriguing questionsconcerning the coordination of IGF and IGF receptorexpression in both eye and tectum.

The molecular layer of the cerebellum, and atslightly lower levels in the granular layer, showed con-sistently high [125I]IGF-I binding throughout life (Figs.1A and B), our immunostaining results further supportthis (Fig. 6). Recent immunohistochemical localization


studies have further shown high IGF-I peptide immu-noreactivity over Purkinje perikarya (and/or dendrites)of the cerebellum of tilapia, Oreochromis mossambicus,(Reinecke et al., 1997) and sculpin Cottus scorpius(LoYng-Cueni et al., 1998). This region of the teleostbrain at least therefore appears to have an overlap ofpeptide immunoreactivity and putative receptor locali-zation. Interestingly, in this study of the adult troutbrain, a higher density of putative IGF-I receptors wasfound in the molecular layer of the cerebellum com-pared with the granular layer. This contrasts with mostadult rat studies where a higher prenatal density ofIGF-I receptors in the molecular layer declines signiW-

cantly over early development (Kar et al., 1993). In thegranular layer, IGF-I receptor density remains more

stable over rat development, Wnally leading to theapparent higher density of receptors in the granularlayer of adulthood (Baskin et al., 1986; Bohannonet al., 1988b; Kar et al., 1993; Lesniak et al., 1988). Therelative densities of IGF-I receptors in the teleost cere-bellum therefore appears much more similar to thatfound in the rat during early and more rapid stages ofdevelopment. Several studies in the rat brain haveshown IGF-I to be a critical factor in the developmentof the cerebellum. In vivo studies involving transgenicmice in whom an IGF-I transgene is highly expressedalso demonstrated signiWcant increases in cerebellargrowth (Ye et al., 1996). Recent radiolabelled thymi-dine incorporation studies have further shown themolecular layer to be a major proliferation zone within

Fig. 5. IGF-IR (ZIR-7) immunoreactivity (brown) in optic tectum of adult brown trout against methyl green counterstain. (A) View through wholeof tectum, 10£. (B) High power view (40£) of SGC shows distinct IGF-IR immunoreactivity on neuronal cell bodies (arrow) and ascending dentriticWbres (unWlled arrowhead). (C) Lack of IGF-IR immunoreactivity in glial cells (arrowheads) of the SM strata. Scale bar, 50 �m. SM, stratum margin-ale; SFGS, stratum Wbrosum et griseum superWciale; SGC, stratum griseum centrale, SAC, stratum album centrale; and SPV, stratum periventricu-lare. (For interpretation of the references to color in this Wgure legend, the reader is referred to the web version of this paper.)


the mormyrid Wsh cerebellum (Zupanc et al., 1996). Themaintenance of IGF-I receptor numbers in this regionof the brain would therefore appear to be consistentwith a role for IGF-I in the production of new neuronsin the adult teleost cerebellum. A role for IGF-I in neu-ron survival is also becoming established (Fernándezet al., 1999; Torres-Aleman et al., 1994). A greater sur-vival of neurons, combined with increased cell prolifer-ation would also help contribute to the continualgrowth of this region of the brain in Wsh. It is interest-ing to note that in members of the Gymnotiforms andMormyrids, approximately 75% of all new brain cellsare produced within the cerebellum (Zupanc, 1999).Furthermore, in the Mormyrids the cerebellumbecomes so hypertrophied and enlarged that it pushesaside the optic tectum and comes to dominate the mor-phology of the brain. Species of Wsh belonging to these

families would therefore appear to be ideal candidatemodels for studying the eVects of IGF-I in this regionof the vertebrate brain.

SpeciWc binding of IGF-I was also found in thelobus inferioris of the hypothalamus (lih). In the rathypothalamus, IGF-I receptors have been foundlargely conWned to median eminence. An up-regulationof IGF-I receptors has been found in this region infasted rats suggesting a possible role in regulatinghypophysiotropic hormone secretion from the medianeminance (Bohannon et al., 1988a). Although absent inmost Wsh species, a median eminance is still present inthe salmonids and eels but represents a less discreteregion bordering the hypothalamus and pituitary.While the pattern of binding within the lih closely cor-responds with the neuronal perikarya and Wbre tractsrunning through it, this may again be related to neuron

Fig. 6. IGF-IR (ZIR-7) immunoreactivity (brown) in cerebellum of adult brown trout, methyl green counterstain. (A) Valvula of cerebellum, lowmagniWcation (scale bar, 300 �m), note that in teleosts the cerebellum undergoes torsion such that in the valv. cerebelli the molecular layer (MOL)forms the inner layer and the granular layer (GRAN) occupying the outer. (B) Corpus of cerebellum, low magniWcation (scale bar, 300 �m), now withmolecular layer occupying the outermost layer. (C) High power view of valvula of cerebellum, asterisks indicate clearly immunopositive purkinje cellstaining, unWlled arrowheads indicate positive staining of neuronal Wbres running through the molecular layer of the valv. cerebelli, staining is largelyabsent from the granular layer, low magniWcation now with molecular layer on the outside. (D) High power view of IGF-IR immunostaining in cor-pus cerebelli (boxed region in B), black arrows indicate immunopositive purkinje cells, unWlled arrowheads indicate immunoreactive Wbres runningthrough molecular layer (now outer layer of cerebellum), positive immunoreactivity is largely absent from the granular layer, scale bar, 20 �m. (Forinterpretation of the references to color in this Wgure legend, the reader is referred to the web version of this paper.)


proliferation, diVerentiation, or survival. A possiblehypophysiotropic role for IGF-I in this region of theteleost brain cannot yet be rule out and requires further

Fig. 7. IGF-IR (ZIR-7) immunoreactivity (brown) in saccus vasculo-sus (SV) of adult trout brain, counterstained with methyl green. (A)Low power view (scale bar, 50 �m) showing strong immunopositiveIGF-IR staining in SV. (B) Detail of immunoreactivity in saccus vas-culosus with positive immunostaining in CSF-contacting cells (scalebar, 20 �m). (C) Negative control with omission of primary antibody,consecutive section, suggests immunostaining is speciWc and not due toendogenous peroxidase activity. L, lumen; CC, coronet cell. (For inter-pretation of the references to color in this Wgure legend, the reader isreferred to the web version of this paper.)

investigation. Putative IGF-I binding sites in olfactorybulb are similar to Wndings in birds (Bassas et al., 1989)and mammals (Bohannon et al., 1988b; Lesniak et al.,1988), where signiWcant receptor binding is consistentlylocalized in olfactory bulb. Consequently some work-ers have linked IGF-I with a role in olfaction and theimportance of this sense to these animals (Bohannonet al., 1988b; Lesniak et al., 1988). The high density ofputative IGF-I binding sites in the saccus vasculosus(SV) may possibly mirror that seen in the choroidplexus of the rat which exhibits high IGF-IR expres-sion (Lesniak et al., 1988). The discrete localization oflikely IGF-I binding sites in CSF-contacting coronetcells may represent an important site for sensingperipheral levels of IGFs in CSF.

We deliberately used in vitro autoradiographic ligandbinding for our studies so that we could simultaneouslystudy the presence of IGF binding proteins. The IGFBPsare an important component of the IGF system, playinga signiWcant role in regulating the availibility of IGF-Iwhich can bind to its receptor. IGFBPs are present inabundance in other Wsh tissues such as muscle (Degger etal., 2000) and retina (Boucher and Hitchcock, 1998a). Atruncated form of IGF-I, des-(1-3)-IGF-I, and naturallyoccuring in mammals, has a 40-fold lower aYnity forIGFBPs but retains full aYnity for the IGF-I receptor.Competitive binding studies using this form of the pep-tide have previously been used to show the presence ofIGFBPs in goldWsh retina (Boucher and Hitchcock,1998a). In this study, similar partial inhibition of IGF-Ibinding suggests the widespread presence of these pro-teins in Wsh brain.

In conclusion, the common factor of probable IGF-Ibinding sites in ganglionic layers and perikarya richregions of the trout brain argues strongly for a role forIGF-I in neuronal proliferation, growth and survival.Regions exhibiting highest speciWc [125I]IGF-I bindingand immunoreactivity such as cerebellum and optictectum are also highly stratiWed. The association oflikely IGF-I binding sites with complexly layered brainregions appears to be a primitive state which has per-sisted into the highest vertebrates, including the cere-bellum of man. Accompanying the continual growth ofthese regions in the teleost is a requirement for con-stant cell diVerentiation in which IGF-I may again playa signiWcant role. This would suggest diVering roles forIGF-I within the brain, even at an early stage of evolu-tion. A possible hypophysiotropic action in the hypo-thalamus cannot yet be ruled out and further studieswill be required to establish such a potential role. Theimportance of IGF-I to brain development, particu-larly in regions such as the cerebellum, together withthe continual indeterminate growth of Wshes, suggeststhat the teleost CNS may be an extremely useful modelfor studying the actions of IGF-I in this particularregion of the brain.


Acknowledgments

We are indebited to the Piscifactoria de Bagá, Depar-tament de Medi Natural de la Generalitat de Catalunya,and especially Antonino Clemente for providing Wsh.A.S. was a fellow of the European Community Postdoc-toral Program (ERB4001GT973399). We thank ChironTherapeutics for the gift of human recombinant IGF-I.This work was funded by grants; NATO (CRG.CRG974408), DGICY (AGL-2001-2903 ACU), and CIRIT(2001 SGR-00122) to J.G.

References

Anlar, B., Sullivan, K.A., Feldman, E.L., 1999. Insulin-like growth fac-tor-I and central nervous system development. Horm. Metab. Res.31, 120–125.

Ayaso, E., Nolan, K., Byrnes, L., 1999. ZebraWsh Insulin-like growthfactor I receptor gene (IGFIR): cDNA cloning and expressionstudies during development. Environment, development andgrowth of Wshes, FSBI Annual Intl. Symposium, University of St.Andrews, Scotland, UK, 91.

Baskin, D.G., Brewitt, B., Davidson, D.A., Corp, E., Paquette, T., Fig-lewicz, D.P., Lewellen, T.K., Graham, M.K., Woods, S.G., Dorsa,D.M., 1986. Quantitative autoradiographic evidence for insulinreceptors in the choroid plexus of the rat brain. Diabetes 35, 246–249.

Bassas, L., Girbau, M., Lesniak, M.A., Roth, J., de Pablo, F., 1989.Development of receptors for insulin and insulin-like growth fac-tor-I in head and brain of chick embryos: Autoradiographic locali-zation. Endocrinology 125, 2320–2327.

Bohannon, N.J., Corp, E.S., Wilcox, B.J., Figlewicz, D.P., Dorsa, D.M.,Baskin, D.G., 1988a. Characterization of insulin-like growth factorI receptors in the median eminence of the brain and their modula-tion by food restriction. Endocrinology 122, 1940–1947.

Bohannon, N.J., Corp, E.S., Wilcox, B.J., Figlewicz, D.P., Dorsa, D.M.,Baskin, D.G., 1988b. Localization of binding sites for insulin-likegrowth factor-I (IGF-I) in the rat brain by quantitative autoradiog-raphy. Brain Res. 444, 205–213.

Boucher, S.E., Hitchcock, P.F., 1998a. Insulin-like growth factor-Ibinds in the inner plexiform layer and circumferential germinalzone in the retina of the goldWsh. J. Comp. Neurol. 394, 395–401.

Boucher, S.E., Hitchcock, P.F., 1998b. Insulin-related growth factorsstimulate proliferation of retinal progenitors in the goldWsh. J.Comp. Neurol. 394, 386–394.

Chan, S.J., Nagamatsu, S., Cao, Q.P., Steiner, D.F., 1992. Structure andevolution of insulin and insulin-like growth factors in chordates.Prog. Brain Res. 92, 15–24.

Cheng, C.M., Joncas, G., Reinhardt, R.R., Farrer, R., Quarles, R., Jans-sen, J., McDonald, M.P., Crawley, J.N., Powell-Braxton, L., Bondy,C.A., 1998. Biochemical and morphometric analyses show thatmyelination in the insulin-like growth factor 1 null brain is propor-tionate to its neuronal composition. J. Neurosci. 18, 5673–5681.

Cook, J.E., Rankin, E.C., Stevens, H.P., 1983. A pattern of optic axons inthe normal goldWsh tectum consistent with the caudal migration ofoptic terminals during development. Exp. Brain Res. 52, 147–151.

Cronly-Dillon, J.R., StaVord, C.A., 1986. GoldWsh retina and tectuminXuence each other’s growth activity during regrowth of the reti-notectal projection. Brain Res. 395, 13–23.

de Pablo, F., de la Rosa, E., 1995. The developing CNS: A scenario forthe action of proinsulin, insulin and insulin-like growth factors.Trends Neurosci. 18, 143–150.

Degger, B., Upton, Z., Soole, K., Collet, C., Richardson, N., 2000. Com-parison of recombinant barramundi and human insulin-like growth

factor (IGF)-I in juvenile barramundi (Lates calcarifer): In vivo met-abolic eVects, association with circulating IGF-binding proteins, andtissue localisation. Gen. Comp. Endocrinol. 117, 395–403.

D’Ercole, A.J., Ye, P., Calikoglu, A.S., Gutierrez-Ospina, G., 1996. Therole of the insulin-like growth factors in the central nervous system.Mol. Neurobiol. 13, 227–255.

Drakenberg, K., Sara, V.R., Falkmer, S., Gammeltoft, S., Maake, C.,Reinecke, M., 1993. IdentiWcation of IGF-1 receptors in primitivevertebrates. Regul. Pept. 43, 73–81.

Easter Jr., S.S., Stuermer, C.A., 1984. An evaluation of the hypothesisof shifting terminals in goldWsh optic tectum. J. Neurosci. 4, 1052–1063.

Fernández, A.M., González delaVega, A.G., Planas, B., Torres-Aleman,I., 1999. Neuroprotective actions of peripherally administered insu-lin-like growth factor I in the injured olivo-cerebellar pathway. Eur.J. Neurosci. 11, 2019–2030.

Gutiérrez, J., Parrizas, M., Maestro, M.A., Navarro, I., Plisetskaya,E.M., 1995. Insulin and IGF-I binding and tyrosine kinase activityin Wsh heart. J. Endocrinol. 146, 35–44.

Holzenberger, M., Lapointe, F., Leibovici, M., Lievre, C.A., 1996. Theavian IGF type 1 receptor: cDNA analysis and in situ hybridiza-tion reveal conserved sequence elements and expression patternsrelevant for the development of the nervous system. Brain Res.Dev. Brain Res. 97, 76–87.

Jones, J., Clemmens, D., 1995. Insulin-like growth factors and theirbinding proteins: Biological actions. Endocr. Rev. 16, 3–34.

Kar, S., Chabot, J.G., Quirion, R., 1993. Quantitative autoradio-graphic localization of [125I]insulin-like growth factor I,[125I]insulin-like growth factor II, and [125I]insulin receptorbinding sites in developing and adult rat brain. J. Comp. Neurol.333, 375–397.

Leibush, B., Parrizas, M., Navarro, I., Lappova, Y., Maestro, M.A.,Encinas, M., Plisetskaya, E.M., Gutiérrez, J., 1996. Insulin and insu-lin-like growth factor-I receptors in Wsh brain. Regul. Pept. 61, 155–161.

LeRoith, D., Adamo, M., Werner, H., Roberts, C.J., 1995. Molecularand cellular biology of the insulin-like growth factors. In: Wein-traub, B.D. (Ed.), Molecular Endocrinology: Basic Concepts andClinical Correlations. Raven, New York.

LeRoith, D., Kavsan, V.M., Koval, A.P., and Roberts, C.T., Jr., 1993.Phylogeny of the insulin-like growth factors (IGFs) and receptors:A molecular approach. Mol. Reprod. Dev., 35, 332–336; discussion337–338.

Lesniak, M.A., Hill, J.M., Kiess, W., Rojeski, M., Pert, C.B., Roth, J.,1988. Receptors for insulin-like growth factors I and II: Autoradio-graphic localization in rat brain and comparison to receptors forinsulin. Endocrinology 123, 2089–2099.

Leventhal, P., Feldman, E., 1997. Insulin-like growth factors as regula-tors of cell motility: Signalling mechanisms. Trends Endocrinol.Metab. 8, 1–6.

LoYng-Cueni, D., Schmid, A.C., Graf, H., Reinecke, M., 1998. IGF-I inthe bony Wsh Cottus scorpius: cDNA, expression and diVerentiallocalization in brain and islets. Mol. Cell Endocrinol. 141, 187–194.

Maures, T., Chan, S.J., Xu, B., Sun, H., Ding, J., Duan, C., 2002. Struc-tural, biochemical, and expression analysis of two distinct insulin-like growth factor I receptors and their ligands in zebraWsh. Endo-crinology 143, 1858–1871.

Meek, J., 1983. Functional anatomy of the tectum mesencephali of thegoldWsh. An explorative analysis of the functional implications ofthe laminar structural organization of the tectum. Brain Res. 287,247–297.

Meek, J., Nieuwenhuys, R., 1997. In: Nieuwenhuys, R., ten Donkelaar,H.J., Nicholson, C., with contributions by J.L. (Eds.), The CentralNervous System of Vertebrates. Springer, Dubbeldam, Newyork,vol. 2, pp. 759–913.

Meek, J., Nieuwehuys, R., 1998. The Central Nervous System of Verte-brates, vol. 2, pp. 759–913.


Mozell, R.L., McMorris, F.A., 1991. Insulin-like growth factor I stimu-lates oligodendrocyte development and myelination in rat brainaggregate cultures. J. Neurosci. Res. 30, 382–390.

Navarro, I., Leibush, B., Moon, T.W., Plisetskaya, E.M., Banos, N.,Mendez, E., Planas, J.V., Gutiérrez, J., 1999. Insulin, insulin-likegrowth factor-I (IGF-I) and glucagon: The evolution of theirreceptors. Comp. Biochem. Physiol. B 122, 137–153.

Nazarali, A.J., Gutkind, J.S., Saavedra, J.M., 1989. Calibration of 125Ipolymer standards for use in quantitative receptor autoradiogra-phy. J. Neurosci. Methods 30, 247–253.

Raymond, P.A., Easter Jr., S.S., 1983. Postembryonic growth of theoptic tectum in goldWsh. I. Location of germinal cells and numbersof neurons produced. J. Neurosci. 3, 1077–1091.

Reinecke, M., Schmid, A., Ermatinger, R., LoYng-Cueni, D., 1997.Insulin-like growth factor I in the teleost Oreochromis mossambi-cus, the tilapia: Gene sequence, tissue expression, and cellular local-ization. Endocrinology 138, 3613–3619.

Stuermer, C.A., Easter Jr., S.S., 1984. A comparison of the normal and regen-erated retinotectal pathways of goldWsh. J. Comp. Neurol. 223, 57–76.

Torres-Aleman, I., Pons, S., Arevalo, M.A., 1994. The insulin-likegrowth factor I system in the rat cerebellum: Developmental regu-lation and role in neuronal survival and diVerentiation. J. Neurosci.Res. 39, 117–126.

Torres-Aleman, I., Pons, S., Santos-Benito, F.F., 1992. Survival of Pur-kinje cells in cerebellar cultures is increased by insulin-like growthfactor I. Eur. J. Neurosci. 4, 864–869.

Upton, Z., Francis, G.L., Chan, S.J., Steiner, D.F., Wallace, J.C., Bal-lard, F.J., 1997. Evolution of insulin-like growth factor (IGF) func-tion: Production and characterization of recombinant hagWsh IGF.Gen. Comp. Endocrinol. 105, 79–90.

Upton, Z., Yandell, C.A, Degger, B.G., Chan, S.J., Moriyama, S., Fran-cis, G.L., Ballard, F.J, 1998. Evolution of insulin-like growth factor-I (IGF-I) action: In vitro characterization of vertebrate IGF-I pro-teins. Comp. Biochem. Physiol. B 121, 35–41.

Werther, G.A., Hogg, A., OldWeld, B.J., McKinley, M.J., Figdor, R.,Allen, A.M., Mendelsohn, F.A., 1987. Localization and character-ization of insulin receptors in rat brain and pituitary gland using invitro autoradiography and computerized densitometry. Endocri-nology 121, 1562–1570.

Ye, P., Xing, Y., Dai, Z., D’Ercole, A.J., 1996. In vivo actions of insulin-like growth factor-I (IGF-I) on cerebellum development in trans-genic mice: Evidence that IGF-I increases proliferation of granulecell progenitors. Brain Res. Dev. Brain Res. 95, 44–54.

Zupanc, G.K., 1999. Neurogenesis, cell death and regeneration inthe adult gymnotiform brain. J. Exp. Biol. 202 (Pt.10), 1435–1446.

Zupanc, G.K., Horschke, I., 1995. Proliferation zones in the brain ofadult gymnotiform Wsh: A quantitative mapping study. J. Comp.Neurol. 353, 213–233.

Zupanc, G.K., Horschke, I., Ott, R., Rascher, G.B, 1996. Postembry-onic development of the cerebellum in gymnotiform Wsh. J. Comp.Neurol. 370, 443–464.

autoradiographic and immunohistochemical localization of insulin-like growth factor-i receptor...

Documents