www.elsevier.com/locate/devbrainres

Developmental Brain Resear

Research report

Reelin expression in the retina and optic tectum

of developing common brown trout

Eva M. Candal, Hector J. Caruncho, Catalina Sueiro, Ramon Anadon, Isabel Rodrıguez-Moldes*

Department of Cell Biology and Ecology, Faculty of Biology, University of Santiago de Compostela, 15782-Santiago de Compostela, Spain

Accepted 12 October 2004

Available online 28 December 2004

Abstract

Reelin (RELN) is an extracellular matrix protein largely related with laminar organization in several brain areas. The development of

RELN immunoreactivity in the retina and the optic tectum of the brown trout are analyzed with a monoclonal (142) antibody against RELN

whose suitability has been ascertained by western blot. In the retina of embryos and alevins, RELN immunoreactivity is detected in cells of

the ganglion cell layer (GCL) and inner nuclear layer (INL), and in the inner plexiform layer (IPL), where it appears as bdiffuseQ material

confined to the ON-sublayer. In juveniles, RELN expression becomes restricted to a stripe of cells in the INL. RELN-immunoreactive

(RELN-ir) cells are absent from the outer nuclear layer (ONL) at any developmental stage. The developmental pattern of RELN expression in

the trout retina shows many similarities with that of amniotes: (a) RELN expression parallels the vitreal to scleral progression of

differentiation of the retina and, within each cell layer, RELN immunoreactivity appears confined to a subpopulation of postmitotic cells; (b)

at early stages RELN expression is exclusively observed in the central retina and as maturation progresses from the center to the periphery,

more RELN-ir cells are observed following the same spatial pattern. Differences with amniotes are noted regarding the absence of RELN

expression in the GCL and INL in adulthood, and in the ONL at any developmental stage. In the optic tectum (OT) of trout, as in amniotes,

RELN immunoreactivity increases within specific cell layers as lamination proceeds, and decreases when it is complete, except in the stratum

opticum (SO), where RELN-ir cells are observed throughout life. Time-course expression of RELN in the OT suggests a role in the early

modeling of synaptic contacts and the accommodation of new retinal arriving axons throughout life.

D 2004 Elsevier B.V. All rights reserved.

Theme: Development and regeneration

Topic: Visual system

Keywords: Immunohistochemistry; Postmitotic cells; Teleost; Visual system

0165-3806/$ - see front matter D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.devbrainres.2004.10.014

Abbreviations: INL, inner nuclear layer; INLi, inner part of the INL;

INLo, outer part of the INL; IPL, inner plexiform layer; IZ, intermediate

zone; L, lens; on, optic nerve; ON, on sublayer of the IPL; ONL, outer

nuclear layer; OPL, outer plexiform layer; re, retinal epithelium; SGC,

stratum griseum centrale; SGFS, stratum griseum et fibrosum superficiale;

SGP, stratum griseum periventriculare; SM, stratum marginale; SO, stratum

opticum; VZ, ventricular zone

* Corresponding author. Department of Fundamental Biology, Faculty

of Biology, University of Santiago de Compostela, 15782-Santiago de

Compostela, Spain. Fax: +34 81 596904.

E-mail address: bfmoldes@usc.es (I. Rodrıguez-Moldes).

1. Introduction

Regulation of neuronal positioning and establishment of

synaptic connections among related nervous centers is key

to organize appropriate brain architectonic patterns and

alterations in these processes can contribute to malignancy.

To date, several molecules have been documented to play

fundamental roles in these processes, among them a protein

named reelin (RELN) that was identified in 1995 as the

product of the breeler geneQ expression [9,19]. RELN-

deficient (breelerQ) mice show notable alterations in the

laminar organization of the cerebral and cerebellar cortices.

In the visual system, reeler mice show deficiencies such as

ch 154 (2005) 187–197

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197188

an attenuation of rod-driven retinal responses, and an

abnormal distribution of cell processes in the inner plexi-

form layer of the retina [46]. The superior colliculus of these

mutants presents anomalous distributions of axons despite

its normal cytoarchitecture [15]. RELN is involved in the

correct neuronal alignment in laminar brain structures, and

has also been implicated in axonal pathfinding, dendrite

arborization, maintenance of cytoskeletal stability, axon

remodeling and modulation of synaptic connections

[1,2,4,7,12–14,17,25,33,35–37,45,48,49,53,54].

AsRELN appears critical for correct cortical development,

many studies have focused on development of the RELN

immunoreactivity in the retina and brain of mammals. A

moderate RELN expression was observed in the retina, optic

tectum and other brain areas of the developing mouse [49],

where RELN is first expressed in pioneer neurons [44,45].

A faint RELN expression has also been observed in the

retina and optic tectum of adults [44,46,49]. In contrast with

mammals, studies of RELN expression in nonmammalian

vertebrates are very scarce: a few studies in sauropsids have

revealed a conserved pattern of RELN mRNA expression

during brain development [2,3,18,55], which includes a

moderate RELN mRNA expression in the retina and tectum

of embryos. However, the distribution of RELN has not

been studied in adult sauropsids. Regarding anamniotes,

recent studies have reported RELN expression in the

developing and adult optic tectum of the zebrafish [8] and in

the optic tectum of sea lamprey larvae, but not in adults

[38,39]. Although Costagli et al. [8] have reported RELN-ir

cells in the retina of zebrafish, detailed studies on the expre-

ssion of RELN in the anamniotan visual system are lacking.

The retina and optic tectum of teleosts exhibit a well-

organized laminar pattern, and the retinotectal projections

are patterned following a highly ordered retinotopic map.

The development of the retinotectal projections has been

well characterized in the trout: both the retina and optic

tectum expand actively by addition of new cells from

marginal regions but following different spatial patterns

[28,29,41]. Thus, to maintain retinotopy throughout devel-

opment, the connectional pattern must change accordingly.

Here we took advantage of the well-known patterns of

layering and synaptogenesis throughout the development of

the retina and optic tectum of trout, to assess the putative

roles of the RELN in the organization of these structures.

With such a purpose, we have studied by immunocytochem-

istry its developmental expression in both the retina and the

optic tectum using a monoclonal antibody (142) against

RELN. For characterization of this antibody, see Ref. [11].

2. Material and methods

2.1. Experimental animals

Thirty-three embryos (of total lengths comprised

between 5 and 14 mm), fourteen alevins (lengths between

15 and 26 mm), six juveniles (27, 30, and 35 mm, two of

each), and seven adults (23–29 cm length) of the brown

trout (Salmo trutta fario) were employed in this study.

Animals were supplied by a local hatchery (Centro

Ictioxenico de Sobrado dos Monxes, A Coruna, Spain)

and maintained in well-aerated freshwater tanks. Previously

to all experiments, animals were deeply anesthetized with a

0.05% solution of tricaine methane sulfonate (MS-222;

Sigma, St. Louis, MO) in fresh water. All experiments were

conducted in accordance with the European Community

guidelines on animal care and experimentation.

2.2. Western blot analysis

The specificity of the antiserum 142 for RELN has been

tested previously in mammals and in other vertebrates,

including fishes [2,11,18,20,31,32,38,40,47,55]. The spe-

cificity of the antibody in trout was checked by Western blot

of adult brain extracts. For comparative purpose, brain

extracts of rat were subjected to identical analyses.

Trouts and rats were killed by overdose of tricaine

methane sulfonate and chloroform, respectively. Brains were

quickly removed and mechanically homogenized in six-fold

volume of cold TRIS-saline buffer at pH 7.6 (50 mM)

containing EDTA (5 mM; Sigma), and the protease inhibitors

phenylmethylsulfonylfluoride (2 mM; Sigma) and N-ethyl-

maleimide (10 mM; Sigma). Samples were centrifuged at

20,000 � g at 4 8C for 20 min, and cold methanol was then

added to the supernatant and kept overnight at�20 8C. Aftera brief centrifugation, methanol was eliminated and the

protein concentration of the samples was determined by the

Bradford method. Proteins were separated by sodium

dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-

PAGE) on 6% acrylamide 80� 70� 0.75 mm slab gels, at a

constant 130 V for 80 min (Mini-Protean II PAGE System,

Bio-Rad, Richmond, CA). Samples of 50 Ag of protein were

applied to each lane, and SDS-PAGE molecular weight

standards were run in additional lanes. The separated proteins

were then electroblotted at 30 Vovernight at 4 8C onto a 0.45

Am pore size nitrocellulose membrane (Bio-Rad) using a

Mini-TransBlot system (Bio-Rad). Nonspecific binding sites

on the membrane were blocked by incubating in 5% milk

powder in 0.01 M TRIS-saline buffer (pH 8.0) containing

0.5% Tween-20 (TBST) for 2 h. The blots were then

incubated in agitation with the 142 monoclonal anti-RELN

antibody (a generous gift by Dr. Goffinet, University of

Louvain, Belgium) diluted 1:1000 in TBST containing 15%

normal sheep overnight at room temperature. This antibody

was raised against an epitope of the amino-terminal region of

the RELN [11]. After repeated washing in TBST, the blots

were incubated for 1 h in anti-mouse IgG, peroxidase-linked

species-specific whole antibody (from sheep) (Amersham,

Buckinghamshire, England, 1:5000 dilution). Staining was

visualized with a rapid electrochemoluminescent detection

system (ECL Western Blotting System; Amersham) and

exposed on to Hyperfilm-ECL (Amersham).

Fig. 1. (a) The specificity of the 142 antibody for RELN was verified by

immunoblotting of protein extracts of adult trout and rat brains. The

antiserum recognized three polypeptides of 380, 300 and 180 kDa in trout

(T). Rat extracts (R) ran in parallel reveal the same bands. Molecular weight

standards are indicated at the right. (b) RELN immunoreactivity can be

observed in some of the cells (arrows) and also as a diffuse material located

in the neuropile among the cell bodies (arrowheads) and in the IPL.

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197 189

2.3. RELN immunohistochemistry

Embryos, alevins and juveniles were fixed by immersion

in 4% paraformaldehyde in 0.1 M phosphate buffer (PB) at

pH 7.4. The chorion of the embryos was drilled before

immersion to gain better access of the fixative. Entire

embryos, heads of alevins and juveniles were immersed in

cool 30% sucrose in PB until they sank, then were

embedded in OCT Compound (Tissue Tek, Torrance, CA),

frozen with liquid-nitrogen-cooled isopentane, and serially

sectioned in transverse and sagittal planes on a cryostat. The

sections (18 Am thick) were mounted on chrome alum-

gelatinized slides. Sections were processed for RELN

immunohistochemistry following the peroxidase–anti-per-

oxidase (PAP) method. After pretreatment with 10% H2O2

in PB saline (PBS; pH 7.4) for 15–30 min to eliminate

endogenous peroxidase activity, the sections were preincu-

bated in 10% normal goat serum for 1 h, and incubated

overnight in humid chamber at room temperature with the

142 monoclonal anti-RELN antibody (1:3000 dilution). The

sections were rinsed in PBS (three rinses of 10 min each),

incubated in goat anti-mouse IgG (1:50, DAKO, Glostrup,

Denmark) for 1 h, rinsed in PBS and incubated in mouse

peroxidase–anti-peroxidase (PAP) complex (1:500, DAKO)

for 1 h. The antibodies and the PAP complex were diluted in

PBS containing 0.2% Triton X-100. After two PBS rinses,

the immune complex was developed with 0.5 mg/ml of 3,3V-diaminobenzidine tetrahydrochloride (DAB, Sigma) and

0.01% H2O2 in 0.05 M TRIS–HCl buffer (pH 7.6) for 5–

15 min. The sections were dehydrated and coverslipped. In

control series in which the first antibody was omitted or

replaced by nonimmune serum, no immunoreaction was

observed.

Serial transverse sections of embryos and larvae were

stained with haematoxylin to morphologically distinguish

the tectal and retinal layers.

2.4. Imaging

Microphotographs were made with an Olympus DP12

color digital camera and a Provis microscope (Olympus,

Tokyo, Japan). Contrast and brightness of microphoto-

graphs were adjusted using Adobe Photoshop (Adobe

Systems, San Jose, CA).

3. Results

3.1. Western blot analysis

Western blot analysis of protein extracts of adult trout

brains with the monoclonal antibody against RELN (142)

reveals three protein bands of about 380, 300 and 180 kDa

(Fig. 1a). The same bands are also observed in blots of rat

brain extracts running in parallel, confirming the specificity

of the antibody in the trout.

3.2. Immunohistochemistry

RELN immunoreactivity can be observed as a faintly to

strongly stained material located into some of cells (arrows in

Fig. 1b) and also as a diffuse material located among the cell

bodies (arrowheads in Fig. 1b) and in the plexiform layers.

We have considered this diffuse material as a cell-surface-

associated extracellular material on the basis of immunocy-

tochemical observations made in the brains of other

vertebrates by using the same antibody [38] or other RELN

antibodies [10,34]. Moreover, the cellular and secreted forms

of Reelin can be immunohistochemically discriminated as it

has been demonstrated by Kubasak et al. [24] using brefeldin

A to block RELN secretion in organotopic cultures.

The distribution of RELN immunoreactivity in the retina

of embryos, alevins, and juveniles of the common trout is

represented in Figs. 2 and 3. Figs. 4 and 5 show the

distribution of RELN immunoreactivity in the optic tectum

(OT) of embryos, alevins, juveniles and adults.

3.3. Reelin expression in developing trout retina

Differentiation in the retina of the brown trout (S. trutta

fario) follows a vitreal-to-scleral progression (each layer

containing specific cell types) and a central-to-peripheral

progression of maturation, with the central area being more

mature than the adjacent more peripheral areas. The

marginal retina remains as an undifferentiated neuroepithe-

lium throughout development.

In early embryos (11 to 24 days post-fertilization; from

4.5-mm to 10-mm embryos), retinal cells are arranged in

radial columns and no separation in layers is distinguish-

able. No RELN immunoreactivity is observed in this period.

In late embryos (25 to 34 days post-fertilization; 11-mm to

14-mm embryos), the central part of the retina becomes

layered: first, a thin inner plexiform layer (IPL) was formed

between the ganglion cell layer (GCL) and the inner nuclear

layer (INL); later, the INL becomes clearly distinguishable

from the outer nuclear layer (ONL) (see Fig. 2a). The first

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197190

Fig. 3. Schematic drawings of vertical sections of the retina of late embryos

(11 mm), hatchlings (15 mm) and early alevins (17 mm) at nasal,

intermediate and temporal levels to show the asymmetric distribution of

RELN immunoreactivity along the dorsoventral and naso-temporal axis.

Light gray areas in the retina are marking the laminated parts. Dark gray

areas represent diffuse RELN immunoreactivity in the ON-sublayer of the

IPL. Dots represent RELN-ir cells. Arrows indicate the limits of the areas

presenting an increasing number of RELN-ir cells with respect to previous

developmental stages. Broken lines in arrows indicate the areas in which the

relative number of RELN-ir cells decrease with respect to previous

developmental stages. Note that the pattern of RELN immunoreactivity

near the peripheral retina in larger stages (c) recapitulates that of central

parts of the retina in previous stages (b). Scale bar: 100 Am.

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197 191

RELN-immunoreactive (RELN-ir) cells are observed as

lamination begins: faint RELN-ir perikarya are observed in

the GCL, just in the border of the incipient IPL (Fig. 2b,

large arrow). RELN immunoreactivity is also observed as a

diffuse band in the IPL (Fig. 2b) and in a small number of

cells in the INL (Fig. 2b, arrowhead). In contrast, in

peripheral growth region the IPL has not yet formed, and no

RELN immunoreactivity is detected. At hatching (35 days

post-fertilization; 14.5-mm to 15-mm alevins), the outer

plexiform layer (OPL) is clearly distinguished in the central

part of the retina. The IPL presents a band of diffuse RELN-

ir material in its inner part (the presumptive ON-sublayer;

Fig. 2c). In the GCL RELN-ir cells (some of them located

just at the marginal border of the IPL) are very faintly

stained, and whether they represent ganglion cells or

amacrine cells could not be assessed (Fig. 2c, arrow).

RELN immunoreactivity in the GCL and INL extends from

the center to the periphery following the lateral extension of

the IPL. In the central retina, intensely stained cells are also

observed in the INL (Fig. 2c, arrowheads) whereas in more

peripheral regions the pattern of RELN immunoreactivity

resembles that observed in the central retina at earlier

developmental stages (Fig. 2c). At the level of the optic

papilla faint RELN immunoreactivity is also observed

associated with the emerging optic fibers (Fig. 2d). In early

alevins (39 to 58 days post-fertilization; 16-mm to 20-mm

alevins) RELN-ir INL cells show marked staining intensity,

these cells distributing in the inner part of the INL (Fig.

2e,f). These pear-shaped RELN-ir cells exhibit a process

coursing to the IPL, probably representing amacrine cells

(Fig. 2f, arrowheads). In these cells, the RELN-ir material

appears as small masses in both the perikarya and processes

(Fig. 2f). At the end of this period, RELN immunoreactivity

diminishes in retina following a central-to-peripheral pat-

tern. In late alevins (59 to 85 days post-fertilization; 21-mm

to 26-mm alevins), the relative number of RELN-ir cells and

the RELN immunoreactivity in the IPL strongly decreases

(Fig. 2g) and, in juveniles, only a few RELN-ir cells can be

seen in a single layer at the inner border of the INL (Fig.

2h). No RELN immunoreactivity is seen in the ONL and

photoreceptor layer at any stage studied.

The distribution of RELN immunoreactivity along the

retina is not homogeneous, as schematically is represented in

Fig. 3. Here, RELN-ir cells are shown as dots and RELN

immunoreactive diffuse material is shown in dark gray.

Fig. 2. (a) Schematic drawing of a transverse section of the retina of a 17-mm bro

(c)-peripheral (p) direction is represented. For abbreviations, see list. (b–h) Distrib

(c–g) and juveniles (h); (b) 11.5-mm embryo. A diffuse RELN-immunoreactive

perikarya are observed in the most central region of the GCL (large arrow) and

peripheral growth zone of the retina. (c) Faint RELN-ir cells are observed in the

RELN-ir material is confined to the inner part (ON-sublayer). Line indicates the

Oblique section of a 14.5-mm alevins retina showing RELN-ir perikarya in the m

immunoreactivity in the optic fiber layer. (e) RELN-ir cells are found in the GCL (

and INL. The perinuclear cytoplasm (arrow) and processes (arrowheads) of RELN-

A few RELN-ir amacrine cell perikarya (arrows) are seen at the inner border o

diminished in INL perikarya (arrowheads) and in the IPL. Scale bars: 25 Am (f,

RELN-ir cells are never observed in the peripheral growth

zone of the retina. In late embryos (Fig. 3a), RELN

immunoreactivity is lacking in the still neuroepithelial

regions of the retina. At hatching (Fig. 3b), RELN

immunoreactivity has extended following the appearance

wn trout alevins showing the different cell and plexiform layers. The cente

ution of RELN immunoreactivity in the retina of trout embryos (b), alevins

(RELN-ir) band is present in the IPL (short arrows), and faint RELN-i

the INL (arrowhead). Lines indicate the limit between the central and the

GCL (arrows) and the INL (arrowheads) of a 14.5-mm alevins. In the IPL

limit between the central (c) and the more peripheral adjacent retina. (d

arginal border of the IPL (arrows). The arrowhead indicates diffuse RELN

arrows) and the INL (arrowheads) of a 17-mm alevins. (f) Detail of the GCL

ir cells are stained. (g) Section through the central retina of a 24-mm alevins

f the INL. (h) In a 33-mm juvenile, RELN immunoreactivity has largely

g); 50 Am (a–e, h).

r

r

,

)

.

Fig. 4. (a–c) Drawings (a–c) and photomicrographs (c) of transverse sections of the optic tectum of early embryos (a), late embryos (b) and alevins (c) of the

brown trout showing the different cell and plexiform layers visible after haematoxylin staining. (d–k) Transverse sections through the optic tectum of embryos

(d), alevins (e–g), juveniles (h–j), and adult (k). (d) Optic tectum of a 12-mm embryo showing diffuse RELN immunoreactivity in the ventrolateral marginal

zone (arrows). (e) Optic tectum of a 14.5-mm alevins showing RELN-ir cells in the SGP just bordering VZ (arrows), itself being RELN-negative. (f) Optic

tectum of 16-mm alevins showing RELN-ir cells in the SGP and in the prospective SGC (arrows). RELN-ir cells are also observed in more superficial layers

(arrowheads). (g) Detail of faint RELN-ir cells of SGP of this 16-mm alevins. Note that RELN-ir material is perinuclear (arrows). (h) Section through the rostral

optic tectum of 17-mm alevins showing labeled RELN-ir cells in SO (arrowheads). Note very faint RELN immunoreactivity in SGP. (i, j) RELN

immunoreactivity is observed in some SO perikarya of 21-mm (arrowheads in i) and 24-mm alevins (j). (k) Optic tectum of an adult trout showing RELN-ir

cells in the SO, which is now located below a thick SM. For abbreviations, see list. Scale bars: 100 Am (a–c); 50 Am (d–f, h, i); 25 Am (g, j, k).

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197192

of differentiated areas, which is strongly asymmetrical in the

region of the choroid fissure. In the temporoventral retina the

RELN immunoreactivity appears in early alevins (Fig. 3c),

following the appearance of lamination. As described above,

the pattern of RELN immunoreactivity near the peripheral

region resembles that of the central part of the retina of

previous developmental stages (compare Fig. 3b and c).

3.4. Reelin expression in developing trout optic tectum

Descriptions below correspond to the rostral part of the

tectum whose development, for each developmental period,

is more advanced than more caudal parts.

In early embryos (Fig. 4a), the tectal neuroepithelium

consists of a ventricular zone (VZ) with closely packed

Fig. 5. Drawings of transverse sections of the optic tectum of alevins (15,

16 and 17-mm) summarizing the distribution of RELN immunoreactivity at

caudal, intermediate and rostral levels. Cell layers are represented as in Fig.

4. Dots represent RELN-ir cells.

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197 193

proliferating cells, a marginal zone (MZ) containing cell

processes, and postmitotic intermediate zone (IZ), located

between the VZ and the MZ. RELN expression is absent

from the OT of these embryos. Specific OT stratification

begins in late embryos (11 to 14 mm) when the IZ splits in

different cell strata. The cells that remain in deep tectal

layers give rise to the stratum griseum periventriculare

(SGP) and the stratum griseum centrale (SGC) (see Fig.

4b). RELN immunoreactivity is only observed as diffuse

stained material in the MZ (Fig. 4d, arrows). At hatching,

the stratum griseum et fibrosum superficiale (SGFS),

formed of cells and neuropile is appreciable externally to

the SGC throughout the tectum. Rostrally, a fiber layer (the

stratum album centrale, SAC) is outlined between the SGP

and the SGC, and the stratum opticum (SO) appears as the

most superficial layer (see Fig. 4c). Moderate RELN

expression is observed in SPG perikarya and neuropile just

bordering the VZ, which is RELN-negative (Fig. 4e). Some

RELN-ir cells also appear in the prospective superficial

layers (SGC, SGFS; not shown). In early alevins, the

number of cells has increased in all cell layers. The SAC

and the SO extend progressively to the caudal pole of the

tectum, and a special layer of neuropile of the teleost OT

(the stratum marginale; SM) becomes progressively thick-

ened (see Fig. 4c). In early alevins, the most intensely

RELN-ir cells are observed in the SO, settled in the subpial

surface (arrowheads in Fig. 4f, h). In the other layers, RELN

immunoreactivity rises following a ventricular-to-marginal

pattern: it increases first in the SGP and in the SGC (arrows

in Fig. 4f, g), and shortly after in the SGFS (not shown). As

development proceeds, RELN immunoreactivity decreases

following the same pattern: first in the SGP, and then in the

other cell layers (Fig. 4h), except in the SO, where intensely

RELN-ir perikarya are observed even in late alevins (Fig.

4i, j). As the trout grows, the SM becomes progressively

thickened, and the SO RELN-ir cells are observed at deeper

levels from the tectal surface (compare Fig. 4i–k). In

juveniles, the number of RELN-ir cells in the SO decreases

with respect to alevins, and they appear widely separated. In

adult trout, some RELN-ir cells are located in the SO (Fig.

4k), these cells being sparser than in alevins and juveniles.

As in the developing retina, the distribution of the RELN

immunoreactivity in the OT is not homogeneous rostrocau-

dally. In late embryos (not drawn), RELN-ir cells are only

observed in the MZ of the rostral region of the OT,

appearing in its ventrolateral but not in the dorsomedial

edge. The distribution of RELN-ir cells in alevins is

summarized in Fig. 5, where they are represented by black

dots. At hatching (Fig. 5a), RELN-ir cells are observed only

in the rostral tectum. In early alevins (Fig. 5b), the intense

RELN-ir cells extend ventrolaterally and dorsomedially in

the rostral tectum; at more caudal areas, the distribution of

RELN immunoreactivity is similar to that observed in the

rostral tectum at earlier stages. As development proceeds,

RELN immunoreactivity diminishes following also a

rostrocaudal pattern: RELN immunoreactivity diminishes

first rostrally (Fig. 5c) and then (in late alevins; not drawn)

in the intermediate and caudal tectum. An exception to this

rostral-to-caudal progression is found in the torus longi-

tudinalis (TL) of the OT, where RELN expression begins at

hatching in the caudal pole and extends from caudal to

rostral. In juveniles and adults, no RELN immunoreactivity

is progressively observed in the tectum, except in the SO.

4. Discussion

4.1. Reelin in trout

Present western blotting results evidence that three bands

of about 380, 300 and 180 kDa can be detected from the

brain extracts of trout and that this band pattern is similar to

that of rat brain extracts processed in parallel. In rodents, it

has been evidenced that 380 kDa is the molecular weight of

the full-length RELN while 300 and 180 kDa are those of

cleaved products [11] originated after processing by a

metalloproteinase [26]. Our results indicate that RELN, or a

protein closely related to RELN, is present in the trout brain.

Moreover, the cleaved products have similar molecular

weights in trout as in rodents, thus suggesting that the

cleavage points of RELN molecule are well conserved. The

142 antibody used in this study has been raised against the

N-terminus, an epitope that is present in the cleaved forms

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197194

of 300 and 180 kDa but not in the cleaved fragment that

seems to be the active form of RELN, containing repeats 3–

6 [23]. In spite that this antibody does not label this RELN

active form, it is useful to locate the cells that express RELN

and that contain cleaved products.

4.2. The temporal pattern of RELN distribution reflects the

maturation of the retina

The differentiation of the trout retina follows a similar

pattern to those reported in most vertebrates (see Ref.

[50]). The differentiation of ganglion and amacrine cells in

the GCL and inner part of the INL (INLi) is followed by

the sequential differentiation of bipolar and horizontal

cells in the outer part of the INL (INLo) and of

photoreceptor cells [56]. Our results reveal that the

expression of RELN parallels the vitreal-to scleral differ-

entiation of retinal cells revealed by the loss of expression

in these layers of the proliferating cell nuclear antigen

(PCNA), a proliferation marker [5,6]. Indeed, in the trout,

the first postmitotic cells (PCNA-immunonegative)

appeared in the central retina about 24 h before the

appearance of RELN in the central part of the GCL, IPL,

and INL. This finding supports the idea that RELN is

expressed by some postmitotic neurons (possibly pioneer

neurons) in the differentiating retina.

The teleost retina also grows by adding rings of new cells

from the peripheral growth zone to the adjacent central

retina [21,22,27,30] and, accordingly, the first postmitotic

cells are located in the most central part of the retina, while

the most immature cells lie closest to the peripheral growth

zone [5,6]. We demonstrate here that RELN immunoreac-

tivity in the retina extends following a center-to-periphery

gradient that reflects the pattern of cell maturation, being

always absent from the peripheral growing retina.

It is currently assumed that in the developing brain

RELN is secreted by pioneer neurons to act locally on

neighboring cells (via the extracellular matrix) instructing

them to assume their normal position [18,44]. However, in

the trout retina, RELN might be involved in functions other

than neuronal layering, since RELN expression in the INL

persists long time after the retina has completed its

stratification, a pattern that has also been reported in mice

[46]. Moreover, the overall cellular organization of the

neural retina of mouse proceeds almost normally in the

absence of a functional RELN pathway [44,46].

4.3. RELN immunoreactivity is specifically detected in the

ON-sublayer of the IPL

The formation of the optic fiber and the INL starts with the

differentiation of ganglion cells. In the retina of late trout

embryos, RELN immunoreactivity is first observed as a

diffuse-stained material in the IPL and in cells of the GCL just

bordering this layer. As the IPL expands peripherally, the

number of RELN-ir cells in the GCL increases, and some

cells in the central part of the INL become RELN-ir. It is

plausible that some pioneer GCL neurons secrete RELN to

promote the formation of the IPL. In fact, it has been

suggested that dendrites ramify preferentially in RELN-rich

zones [18].

Noteworthy, in trout alevins RELN distribution in the

IPL is not homogeneous but becomes rather confined to the

ON-sublayer, where the processes of ON-bipolar cells

(those that are depolarized when the light is turned on

[52]) contact dendrites of ON-ganglion cells. This finding

raises the possibility that RELN contributes to the formation

of the specific ON-circuitry of the trout IPL. This specificity

has also been observed in the postembryonic mouse [46].

Analysis of the neural retina in the reeler mouse has

revealed an important role for the reeler gene in the

organization of synaptic connections in the ON-sublayer:

the deficiency in the RELN pathway is associated with an

alteration of the number and distribution of synaptic endings

in this sublayer and a reduction in rod-bipolar cell density in

the INL [46].

4.4. Comparative analysis on the expression of reelin in the

retina among different phyla

The patterns of RELN expression in the developing trout

retina roughly coincides with that described in the retinas of

early embryo stages of amniotes [2,3,18,46,49,54,55]:

RELN expression parallels the patterns of differentiation

and, within each cell layer, RELN expression is confined to

a subpopulation of postmitotic cells; at early developmental

stages, RELN expression is observed in the central part of

the retina, whereas it is not observed in the immature

marginal retina, and, as maturation progresses from the

center to the periphery, more RELN-expressing cells are

observed following the same spatial pattern. Additionally,

the pattern of RELN expression in the ON-sublayer of the

IPL is similar to that reported in mouse [46]; studies in other

amniotes [2,3,18] have used in situ hybridization, which

labels RELN-expressing perikarya but not cell processes.

Accordingly, these studies have not referred to the presence

of RELN in plexiform layers.

Some differences between species have been noted

regarding the presence of RELN-ir cells in the ONL: while

they are present in lizard, crocodile and chick embryos

[3,18,55], in trout, as in turtle [4] and mouse [46], RELN-

expressing cells are absent from this layer at any devel-

opmental stage.

Differences have also been noted regarding postem-

bryonic development. While in trout juveniles RELN

immunoreactivity has disappeared in the GCL and is

restricted to sparse cells in the INLi, in mouse, RELN

immunoreactivity is present throughout life in cells in the

GCL, and also in subsets of amacrine and bipolar cells

located in the INLi and INLo, respectively [44,46]. The

absence of studies on RELN expression in postembryonic

stages of other vertebrates precludes further comparison.

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197 195

4.5. RELN immunoreactivity follows gradients of matura-

tion in the optic tectum

Here we observed that in the trout OT, as in other

vertebrates [2,3,8,16,38,55], RELN immunoreactivity

increases within specific cell layers at the time as the

general stratification begins and decreases when lamination

is complete (except in the SO, see below). Additionally,

RELN immunoreactivity increases in all cell layers follow-

ing the mediolateral and rostrocaudal pattern of cell differ-

entiation, being always absent from the proliferative zones

(PCNA-ir; see Refs. [5,6]).

In the OT of teleosts, the progression of both layering and

cell differentiation occurs simultaneously with the progres-

sive establishment of retinotectal projections [29,42,43]. In

the trout, RELN immunoreactivity was first detected in late

embryos at the time as the first optic fibers enter the

rostroventral tectum [28,29,41]. This expression follows

that of RELN in ganglion cells of the retina, when the optic

nerve emerges. In the trout, the earliest retinal afferents

contact their target cells as soon as they enter the tectum,

following a well-known pattern of arborization that coin-

cides with tectal maturation; accordingly, retinal afferents

are always absent from the areas of cell proliferation [28].

Here we demonstrate that, in the retinorecipient layers of the

brown trout OT, RELN-ir cells are detected at the time as

they receive retinal axons or their arborization, being always

absent from tectal regions lacking retinal afferents (non-

layered and PCNA-ir). These data are consistent with a role

for RELN in modeling or remodeling synaptic connections.

Such a modeling may be particularly important in the visual

system of teleosts. The OT grows by the addition of bands

of cells from a crescent-shaped proliferative zone that

extends from the caudal pole to the ventrolateral and

dorsolateral tectum [28]. Maintenance of a correct retino-

tectal map requires that retinal terminals became shifted to

more central positions of the tectum as it grows, either

matching to neighboring cells or with new postmitotic cells

[43]. In trout, the global organization of retinotectal

projections is completed in late alevins [29]. This may

account for the absence of RELN immunoreactivity

observed in most layers of the brown trout OT from late

alevins onwards (except in the SO, as discussed below).

4.6. Possible significance of RELN-ir cells found in the

stratum opticum in the adult trout

Some considerations need to be made on the presence of

RELN-ir cells in the SO of the brown trout. First, these cells

are intensely labeled, in contrast to the more diffuse labeling

of cells located in deeper layers of the tectum. Second, they

are first observed at hatching (the same as other RELN-ir

cells in the tectum), whereas optic fiber arborization is not

observed in this layer until late alevins [29]. Third, RELN-ir

cells persist in this layer throughout life, as has been also

reported in mouse [49]. Moreover, the superficial location of

these cells in the trout OT largely resembles that described in

mouse cortex for the Cajal-Retzius cells (CR). CR cells are

early-generated neurons that are located in the marginal zone

of the neocortex [34], and in a single layer in the outer

marginal zone of the hippocampus [12,51]. In the mouse, the

absence of extracellular RELN in the hippocampus leads to

reduce axonal branching and increasing number of misrouted

aberrant fibers [4]. A similar role of RELN in axonal routing

and branching could account for the presence of heavily

labeled cells of the trout SO during development and

adulthood. Since both the retina and tectum of the trout

continue to grow throughout life, the presence of RELN-ir

cells could contribute to facilitate the arrival of new axons to

their targets.

Acknowledgments

The authors wish to thank Dr. Goffinet of the University

of Namur (Belgium) for the gift of the 142 anti-Reelin

antibody.

Grant sponsors: Xunta de Galicia (PGIDT99BIO20002,

and PGIDT01PXI20007PR) and Spanish Science and

Technology Ministry (BXX2000-0453-C02).

References

[1] F. Aboitiz, Evolution of isocortical organization. A tentative scenario

including roles of reelin, p35/cdk5 and the subplate zone, Cereb.

Cortex 9 (1999) 655–661.

[2] B. Bernier, I. Bar, C. Pieau, C. Lambert de Rouvroit, A.M. Goffinet,

Reelin mRNA expression during embryonic brain development in the

turtle Emys orbicularis, J. Comp. Neurol. 413 (1999) 463–476.

[3] B. Bernier, I. Bar, G. D’Arcangelo, T. Curran, A.M. Goffinet, Reelin

mRNA expression during embryonic brain development in the chick,

J. Comp. Neurol. 422 (2000) 448–463.

[4] V. Borrell, J.A. del Rio, S. Alcantara, M. Derer, A. Martınez, G.

D’Arcangelo, K. Nakajima, K. Mikoshiba, P. Derer, T. Curran, E.

Soriano, Reelin regulates the development and synaptogenesis of the

layer-specific entorhino-hippocampal connections, J. Neurosci. 19

(1999) 1345–1358.

[5] E. Candal, R. Anadon, H.J. Caruncho, W.J. DeGrip, I. Rodrıguez-

Moldes, Proliferation, cell death and reelin expression during

development of a teleost, the common brown trout (Salmo trutta

fario), Int. J. Dev. Biol. 45 (S1) (2001) S69–S70.

[6] E. Candal, R. Anadon, I. Rodrıguez-Moldes, Patterns of cell

proliferation and cell death in the developing retina and optic tectum

of the brown trout, Dev. Brain Res. (in press).

[7] V.S. Caviness Jr., P. Rakic, Mechanism of cortical development. A

view frommutations in mice, Annu. Rev. Neurosci. 1 (1978) 297–326.

[8] A. Costagli, M. Kapsimali, S.W. Wilson, M. Mione, Conserved and

divergent patterns of Reelin expression in the zebrafish central

nervous system, J. Comp. Neurol. 450 (2002) 73–93.

[9] G. D’Arcangelo, G.G. Miao, S. Chen, H.D. Soares, J.I. Morgan, T.

Curran, A protein related to extracellular matrix proteins deleted in the

mouse mutant reeler, Nature 374 (1995) 719–723.

[10] G. D’Arcangelo, K. Nakajima, T. Miyata, M. Ogawa, K. Mikoshiba,

T. Curran, Reelin is a secreted glycoprotein recognized by the CR-50

monoclonal antibody, J. Neurosci. 17 (1997) 23–31.

[11] V. De Bergeyck, B. Naerhuyzen, A.M. Goffinet, C. Lambert de

Rouvroit, A panel of monoclonal antibodies against reelin, the

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197196

extracellular matrix protein defective in reeler mutant mice, J. Neurosci.

Methods 82 (1998) 17–24.

[12] J.A. Del Rio, B. Heimrich, V. Borrell, E. Ffrster, A. Drakew, S.Alcantara, K. Nakajima, T. Miyata, M. Ogawa, K. Mikoshiba, P.

Derer, M. Frotscher, E. Soriano, A role for Cajal-Retzius cells and

reelin in the development of hippocampal connections, Nature 385

(1997) 70–74.

[13] Y. Feng, C.A. Walsh, Protein–protein interactions, cytoskeletal

regulation and neuronal migration, Nat. Rev., Neurosci. 2 (2001)

408–416.

[14] E. Forster, C. Kaltschmidt, J. Deng, H. Cremer, T. Deller, M.

Frotscher, Lamina-specific cell adhesion on living slices of hippo-

campus, Development 125 (1998) 3399–3410.

[15] M. Frotscher, Cajal-Retzius cells, reelin and the formation of layers,

Curr. Opin. Neurobiol. 8 (1998) 570–575.

[16] A.M. Goffinet, The embryonic development of the cortical plate in

reptiles: a comparative study in Emys orbicularis and Lacerta agilis,

J. Comp. Neurol. 215 (1983) 437–452.

[17] A.M. Goffinet, Events governing organization of postmigratory

neurons: studies on brain development in normal and reeler mice,

Brain Res. Rev. 7 (1984) 261–296.

[18] A.M. Goffinet, I. Bar, B. Bernier, C. Trujillo, A. Raynaud, G. Meyer,

Reelin expression during embryonic development on lacertilian

lizards, J. Comp. Neurol. 414 (1999) 533–550.

[19] S. Hirotsune, T. Takahara, N. Sasaki, K. Hirose, A. Yoshiki, T.

Ohashi, M. Kusakabe, Y. Murakami, M. Muramatsu, S. Watanabe,

The reeler gene encodes a protein with an EGF-like motif expressed

by pioneer neurons, Nat. Genet. 10 (1995) 77–83.

[20] F. Impagnatiello, A.R. Guidotti, C. Pesold, Y. Dwivedi, H.J. Caruncho,

M.G. Pisu, D.P. Uzunov, N.R. Smalheiser, J.M. Davis, G.N. Pandey,

G.D. Pappas, P. Tueting, R.P. Sharma, E. Costa, A decrease of reelin

expression as a putative vulnerability factor in schizophrenia, Proc.

Natl. Acad. Sci. U. S. A. 95 (1998) 15718–15723.

[21] P.R. Johns, Growth of the adult goldfish eye: III. Source of the new

retinal cells, J. Comp. Neurol. 176 (1977) 343–358.

[22] P.R. Johns, S. Easter, Growth of the adult goldfish eye: II. Increase in

the retinal cell number, J. Comp. Neurol. 176 (1977) 331–342.

[23] Y. Jossin, N. Ignatova, T. Hiesberger, J. Herz, C. Lambert de Rouvroit,

A.M. Goffinet, The central fragment of Reelin, generated by

proteolytic processing in vivo, is critical to its function during cortical

plate development, J. Neurosci. 24 (2004) 514–5121.

[24] M.D. Kubasak, R. Brooks, S. Chen, S.A. Villeda, P.E. Phelps,

Developmental distribution of reelin-positive cells and their secreted

product in the rodent spinal cord, J. Comp. Neurol. 468 (2004)

165–178.

[25] C. Lambert de Rouvroit, A.M. Goffinet, Neuronal migration, Mech.

Dev. 105 (2001) 47–56.

[26] C. Lambert de Rouvroit, V. de Bergeyck, C. Cortvrindt, I. Bar, Y.

Eeckhout, A.M. Goffinet, Reelin, the extracellular matrix protein

deficient in reeler mutant mice, is processed by a metalloproteinase,

Exp. Neurol. 156 (1999) 214–227.

[27] A.H. Lyall, Cone arrangement in the teleost retina, Q. J. Microsc. Sci.

98 (1957) 189–201.

[28] S. Mansour-Robaey, G. Pinganaud, Quantitative and morphological

study of cell proliferation during morphogenesis in the trout visual

system, J. Hirnforsch. 31 (1990) 495–504.

[29] S. Mansour-Robaey, G. Pinganaud, Development of the retinotectal

arborizations in the trout, Anat. Embryol. 194 (1996) 279–287.

[30] R.L. Meyer, Evidence from thymidine labeling for continuing

growth of retina and tectum in juvenile goldfish, Exp. Neurol. 59

(1978) 99–111.

[31] G. Meyer, A.M. Goffinet, Prenatal development of reelin-immunor-

eactive neurons in the human neocortex, J. Comp. Neurol. 397 (1998)

29–40.

[32] G. Meyer, J.P. Schaaps, L. Moreau, A.M. Goffinet, Embryonic and

early fetal development of the human neocortex, J. Neurosci. 20

(2000) 1858–1868.

[33] T. Miyata, K. Nakajima, K. Mikoshiba, M. Ogawa, Regulation of

Purkinje cell alignment by reelin as revealed with CR-50 antibody,

J. Neurosci. 17 (1997) 3599–3609.

[34] M. Ogawa, T. Miyata, K. Nakajima, K. Yagyu, M. Seike, K. Ikenaka,

H. Yamamoto, K. Mikoshiba, The reeler gene-associated antigen of

Cajal-Retzius neurons is a crucial molecules for laminar organization

of cortical neurons, Neuron 14 (1995) 899–912.

[35] T. Ohshima, M. Ogawa, Veeranna, M. Hirasawa, G. Longenecker,

K. Ishiguro, H.C. Pant, R.O. Brady, A.B. Kulkarni, K. Mikoshiba,

Synergistic contributions of cyclin-dependent kinase 5/p35 and

Reelin/Dab 1 to the positioning of cortical neurons in the

developing mouse brain, Proc. Natl. Acad. Sci. U. S. A. 98

(2001) 2764–2769.

[36] A.L. Pearlman, A.M. Sheppard, Extracellular matrix in early cortical

development, Prog. Brain Res. 108 (1996) 117–134.

[37] A.L. Pearlman, P.L. Faust, M.E. Hatten, J.E. Brunstrom, New

directions for neuronal migration, Curr. Opin. Neurobiol. 8 (1998)

45–54.

[38] E. Perez-Costas, M. Melendez-Ferro, Y. Santos, R. Anadon, M.C.

Rodicio, H.J. Caruncho, Reelin immunoreactivity in the larval sea

lamprey brain, J. Chem. Neuroanat. 23 (2002) 211–221.

[39] E. Perez-Costas, M. Melendez-Ferro, C.G. Perez-Garcıa, H.J. Car-

uncho, M.C. Rodicio, Reelin immunoreactivity in the adult sea

lamprey brain, J. Chem. Neuroanat. 27 (2004) 7–21.

[40] G.C. Perez-Garcıa, F.J. Gonzalez-Delgado, M.L. Suarez-Sola, R.

Castro-Fuentes, J.M. Martın-Trujillo, R. Ferres-Torres, G. Meyer,

Reelin-immunoreactive neurons in the adult vertebrate pallium,

J. Chem. Neuroanat. 21 (2001) 41–51.

[41] G. Pinganaud, Le development du systeme visuel primaire de

Salmo irideus, Arch. Anat. Microsc. Morphol. Exp. 69 (1980)

215–231.

[42] P. Raymond, S.S. Easter Jr., Postembryonic growth of the optic tectum

in goldfish: I. Location of germinal cells and numbers of neurons

produced, J. Neurosci. 5 (1983) 1077–1091.

[43] P. Raymond, S.S. Easter Jr., J.A. Burham, M.K. Powers,

Postembryonic growth of the optic tectum in goldfish: II.

Modulation of cell proliferation by retinal fiber input, J. Neurosci.

5 (1983) 1092–1099.

[44] D.S. Rice, T. Curran, Disabled 1 is expressed in type AII amacrine

cells in the mouse retina, J. Comp. Neurol. 424 (2000) 327–338.

[45] D.S. Rice, T. Curran, Role of the reelin signaling pathway in central

nervous system development, Annu. Rev. Neurosci. 24 (2001)

1005–1039.

[46] D.S. Rice, S. Nusinowitz, A.M. Amizi, A. Martınez, E. Soriano, T.

Curran, The reelin pathway modulates the structure and function of

retinal synaptic circuitry, Neuron 31 (2001) 929–941.

[47] M.A. Rodrıguez-Dıaz, C. Pesold, W.S. Liu, V. Kriho, A. Guidotti,

G.D. Pappas, E. Costa, Colocalization of integrin receptors and reelin

in dendritic spine postsynaptic densities of adult nonhuman primate

cortex, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 3550–3555.

[48] J.R. Sanes, M. Yamagata, Formation of lamina-specific synaptic

connections, Curr. Opin. Neurobiol. 9 (1999) 79–87.

[49] S.N. Schiffmann, B. Bernier, A.M. Goffinet, Reelin mRNA expres-

sion during mouse brain development, Eur. J. Neurosci. 9 (1997)

1055–1071.

[50] B. Sivak, J. Sivak, Vertebrate eye development and refractive

function: an overview, in: M.E. Fini (Ed.), Vertebrate Eye Develop-

ment, Results and Problems in Cell Differentiation, vol. 31, Springer-

Verlag, Berlin, 2000, pp. 1–14.

[51] E. Soriano, J.A. del Rio, A. Martınez, H. Super, Organization of the

embryonic and early postnatal murine hippocampus: I. Immunocy-

tochemical characterization of neuronal populations in the subplate

and marginal zone, J. Comp. Neurol. 342 (1994) 571–595.

[52] W.K. Stell, A.T. Ishida, D.O. Lightfoot, Structural basis for on- and

off-center responses in retinal bipolar cells, Science 198 (1977)

1269–1271.

[53] H. Super, J.A. del Rio, A. Martınez, P. Perez-Sust, E. Soriano,

E.M. Candal et al. / Developmental Brain Research 154 (2005) 187–197 197

Disruption of neuronal migration and radial glia in the developing

cerebral cortex following ablation of Cajal-Retzius cells, Cereb.

Cortex 10 (2000) 602–613.

[54] F. Tissir, A.M. Goffinet, Reelin and brain development, Nat. Rev.,

Neurosci. 4 (2003) 496–505.

[55] F. Tissir, C. Lambert De Rouvroit, J.Y. Sire, G. Meyer, A.M. Goffinet,

Reelin expression during embryonic brain development in Crocodylus

niloticus, J. Comp. Neurol. 457 (2003) 250–262.

[56] E. Vecino, Spatiotemporal development of the fish retina: distribution

of calbindin D-28k, Semin. Cell Dev. Biol. 9 (1998) 271–277.