Differential activities of Sonic hedgehog mediated by Gli transcription factors define distinct neuronal subtypes in the dorsal thalamus

Differential activities of Sonic hedgehog mediated by Gli transcription

factors define distinct neuronal subtypes in the dorsal thalamus

Kazue Hashimoto-Toriia,b,c, Jun Motoyamad, Chi-Chung Huie, Atsushi Kuroiwab,Masato Nakafukuc, Kenji Shimamuraa,c,*

aDivision of Morphogenesis, Department of Embryogenesis, Institute of Molecular Embryology and Genetics,

Kumamoto University, Honjo 2-2-1, Kumamoto 860-0811, JapanbDivision of Biological Science Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan

cDepartment of Neurobiology, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, JapandMolecular Neuropathology Group, Brain Science Institute, RIKEN, Wako, Saitama 351-0198, Japan

eProgram in Developmental Biology and Division of Endocrinology, Research Institute, The Hospital for Sick Children,

555 University Avenue, Toronto, Ont., Canada M5G 1X8

Received 27 June 2003; received in revised form 27 August 2003; accepted 28 August 2003

Abstract

The dorsal thalamus (DT) is a pivotal region in the vertebrate brain that relays inputs from the peripheral sensory organs to higher cognitive

centers. It consists of clusters of neurons with relevant functions, called brain nuclei. However, the mechanisms underlying development of

the DT, including specification of the neuronal subtypes and morphogenesis of the nuclear structures, remain largely unknown. As a first step

to this end, we focused on two transcription factors Sox14 and Gbx2 that are expressed in the specific brain nuclei in the chick DT. The onset

of their expression was found in distinct populations of the postmitotic cells in the prosomere 2, which was regulated by the differential

activities of Sonic hedgehog (Shh) in a manner consistent with the action as a morphogen. Furthermore, both gain- and loss-of-function

results strongly suggest that such distinct inductive activities are mediated selectively by different Gli factors. These results suggest that

cooperation of the differential expression of Gli factors and the activity gradient of Shh signaling generates the distinct thalamic neurons at

the specific locations.

q 2003 Elsevier Ireland Ltd. All rights reserved.

Keywords: Dorsal thalamus; Shh; Sox14; Gbx2; Gli1; Gli2; Diencephalon; Patterning; Brain nuclei; Morphogen

1. Introduction

A number of studies have established that the

embryonic central nervous system (CNS) is regionalized

as Cartesian grids through the actions of anteroposterior

and dorsoventral patterning mechanisms (reviewed by

Lumsden and Krumlauf, 1996; Rubenstein et al., 1998).

Proliferative progenitor cells at the different locations of the

CNS produce distinct sets of neurons that constitute various

brain tissues. For instance, studies of the spinal cord and

telencephalon have revealed that molecularly distinct

domains of progenitor cells generate specific neuronal

subtypes that contribute to nearby as well as distant tissues

(reviewed by Jessell and Sanes, 2000; Corbin et al., 2001).

Patterning of the early neuroepithelial fields is in part

achieved by the actions of inductive signals emanating from

the localized sources, so that cells with distinct properties

arise in a spatially organized manner with respect to the

signaling centers (Agarwala et al., 2001).

There is evidence that an inductive signal regulates the

expression of distinct sets of transcription factors depending

on its concentration. For instance, a secreted glycoprotein

Sonic hedgehog (Shh) has been demonstrated to induce

floor plate properties at a high concentration and progres-

sively more dorsal molecular properties at lower concen-

trations in vitro (Roelink et al., 1995; Ericson et al., 1997).

0925-4773/$ - see front matter q 2003 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.mod.2003.09.001

Mechanisms of Development 120 (2003) 1097–1111

www.elsevier.com/locate/modo

* Corresponding author. Address: Division of Morphogenesis,

Department of Embryogenesis, Institute of Molecular Embryology and

Genetics, Kumamoto University, Honjo 2-2-1, Kumamoto 860-0811,

Japan. Tel.: þ81-96-373-6583; fax: þ81-96-373-6586.

E-mail address: [email protected] (K.

Shimamura).

Factors involved in the Hedgehog (Hh) signaling cascade

have been identified and turned out to be evolutionally

conserved among diverse organisms (reviewed by

Nybakken and Perrimon, 2002). However, the molecular

mechanism underlying the differential activities of Shh are

not yet fully understood. The Gli zinc finger transcription

factor is believed to mediate most of the transcriptional

outputs of Hh signaling (reviewed by Koebernick and

Pieler, 2002; Ruiz i Altaba et al., 2002). While a single Gli

factor Cubitus interruptus (Ci) is absolutely required for Hh

signaling in Drosophila (Methot and Basler, 2001; reviewed

by Aza-Blanc and Kornberg, 1999), vertebrates have three

Ci homologues, Gli1, Gli2 and Gli3, which exhibit different

expression patterns in the developing embryos and exert

distinct functions in some biological situations (Hui et al.,

1994; Ruiz i Altaba, 1998; Aza-Blanc et al., 2000; Persson

et al., 2002; Karlstrom et al., 2003). While models have

been proposed that the graded actions of Hh signaling are

attributed to the activities of several different states of Ci

(Wang and Holmgren, 2000; Methot and Basler, 1999),

exactly how these vertebrate Gli factors are involved in the

execution of the morphogen-like feature of Shh has not yet

been elucidated.

The dorsal thalamus (DT) is a pivotal forebrain structure

that functions as the major relay center connecting the

peripheral sensory organs and cerebral cortex (Jones, 1985,

1998). The DT is partitioned into many clusters of neurons

with relevant functions, called nuclei, which can be

distinguished histologically by features such as the cellular

morphology, differential cell density, and the trajectories of

axon bundles that often demarcate the outlines of the brain

nuclei. Each DT nucleus projects to a specific location

mainly in the telencephalon, and receives inputs from

specific targets. Therefore, the spatial organization of each

DT nucleus is crucial for establishing this precise topo-

graphical relationship to fulfill its functions as a relay

center. Developmentally, the DT is believed to be derived

from a single embryonic subdivision, the alar plate of

prosomere 2 (p2) (Puelles and Rubenstein, 1993). Since the

morphological characteristics that define each DT nucleus

become evident only late in development, the mechanisms

underlying the early stages of its development remain

largely unknown. Several attempts to address this issue have

been made recently (Redies et al., 2000; Yoon et al., 2000;

Nakagawa and O’Leary, 2001; Martınez-de-la-Torre et al.,

2002). For instance, Nakagawa and O’Leary have reported

the nested expression of transcription factors in the

developing mouse thalamus (Nakagawa and O’Leary,

2001). The distinct but overlapping expression patterns of

these regulatory genes often correlated with the histologi-

cally defined borders of the nuclei. Through ontogenic

studies of the expression patterns, they have suggested that

neurons constituting each thalamic nucleus are born with

distinct molecular properties early in development.

In order to elucidate the developmental mechanism of the

DT, we took a similar approach, finding molecular

differences in the DT nuclei and the DT rudiment of the

chick. We provide evidence that Shh emanating from the

basal plate of the p2 specifies two distinct neuronal subtypes

in the DT rudiment in a concentration-dependent manner.

Moreover, these graded actions of Shh are likely to be

mediated by Gli1 and Gli2 selectively, which has not been

reported in other biological situations in which Hh signaling

is involved. These results provide a significant insight into

the molecular mechanism underlying the morphogen-like

property of Shh signaling.

2. Results

2.1. Several DT nuclei are distinguished by expression

of transcription factors

In order to obtain molecular clues to study the DT

development, we have searched for molecules expressed in

subsets of DT nuclei. Among the list, we found that Sox14,

Gbx2 and Sox2 are expressed in restricted regions of the DT

at Hamburger and Hamilton’s stage 42 (HH42) when each

DT nucleus can be identified histologically (Fig. 1A,D).

Most of their expression correlated well with the topogra-

phical organization of the DT nuclei, in that the borders of

their expression domains often coincide with the boundaries

of the nuclei defined morphologically. Expression of Sox14

was restricted to the perirotundic area and the prospective

interstitial nucleus of the optic tract (arrowheads in Fig. 1B).

Gbx2 was expressed in many DT nuclei, as reported in mice

and chick (Miyashita-Lin et al., 1999; Nakagawa and

O’Leary, 2001; Martınez-de-la-Torre et al., 2002). For

instance, the nucleus rotundus, the largest nucleus in the

chick DT, was labeled by the prominent expression of Gbx2

(Fig. 1C), which was surrounded by the Sox14-positive

perirotundic area (Fig. 1B). The subrotundic nucleus and

subhabenular region also showed strong Gbx2 expression,

whereas the dorsal anterior nucleus and dorsolateral anterior

nucleus exhibited lower expression levels (Fig. 1C). Sox2

identified a different set of nuclei, such as nucleus ovoidalis

and the dorsointermediate posterior nucleus, both of which

also showed weak Gbx2 expression (arrowheads in Fig. 1E;

data not shown).

We next asked when such molecular heterogeneity

emerges during development. At HH22, when the thin

mantle layer of postmitotic neurons is formed in the DT (see

Fig. 1M), the expression patterns were already distinct in the

whole-mount specimens (Fig. 1F–H). Sox14 and Gbx2

expression is first detectable at about the same time (HH17)

in a region adjacent to the basal plate of p2 (Fig. 1I–K);

Sox14 expression started in a line of cells arrayed parallel to

the dorsal limit of the basal plate (Fig. 1I), whereas Gbx2-

positive cells emerged somewhat sporadically in a small

area with a extension dorsally (Fig. 1J). Double labeling

showed that there was no overlap between the expression of

these markers (Fig. 1K). As development proceeds,

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111098

the Gbx2 expression occupied a large quadrilateral area in

the DT rudiment whose anterior and ventral edges were

fringed by the hinge-shaped Sox14-positive domain

(Fig. 1L). Close examination of the sectioned material

revealed that they were expressed in a mutually exclusive

pattern, such that cells expressing either of them are

clustered right next to each other (Fig. 1N).

These observations provide evidence that some early

postmitotic neurons in the DT primordium are already

molecularly distinct, although lineage relationship between

these populations at the different stages remains unclear at

this moment.

2.2. Tissues adjacent to the DT rudiment are required

for the expression of Sox14 and Gbx2

To understand how this early molecular heterogeneity is

created in the rudiment of DT, we first asked whether the DT

Fig. 1. Expression of the transcription factors in the DT. (A–C) Coronal sections of HH42 embryos at the level of the rostral DT stained for thionin (A), Sox14

(B) and Gbx2 (C). (D,E) Coronal sections of HH42 embryos at the level of the caudal DT stained for thionin (D) and Sox2 (E). Sox14 is expressed in the cells

(arrowheads in B) that surround the nucleus rotundus where Gbx2-positive cells are evenly distributed (C). A high magnification of the Sox14-expressing

region is shown in the inset in (B). (F–H) HH22 dissected brain whole-mount in situ hybridized for Sox14 (F), Gbx2 (G) and Sox2(H). Note that Sox2 is

expressed strongly in the mantle layer of the DT anlagen (arrows in H), and weakly in the ventricular layer of the entire brain. Dark staining in the

telencephalon is background. For all the whole-mount specimens, the anterior is to the right and the dorsal is to the top. (I,J) HH17 DT rudiments stained for

Sox14 (I) and Gbx2 (J) showing the onset of their expression (arrowheads). Faint darkening in a broad area of p2 is background (J). (K,L) Two color in situ

hybridization for Sox14 (blue) and Gbx2(brown) of the whole-mount DT rudiments at HH17 (K) and HH22 (L). (M,N) Coronal sections of the DT anlagen at

HH20 stained for NeuN (M) and Sox14 (blue in N) and Gbx2 (orange in N). Note that Gbx2-expressing cells are adjacent to, but segregated from the small

domain of Sox14 expression in the layer of postmitotic neurons (arrowhead in K,L,N). For (M) and (N), the third ventricle is located at the right side of the

panels and the dorsal is to the top. Bars, 0.5 mm for (A)–(L); 0.05 mm for (M) and (N). ApR, perirotundic area; DA, dorsal anterior nucleus; DIP,

dorsointermediate posterior nucleus; DLA, dorsolateral anterior nucleus; DLL, dorsolateral lateral nucleus; DLP, Dorsolateral posterior nucleus; DMP,

dorsomedial posterior nucleus; (H), habenula; ITO, prospective interstitial nucleus of the optic tract; Me, mesencephalon; Ov, nucleus ovoidalis; PM, nucleus

paramedianus internus; p1, prosomere 1; p2, prosomere 2; Rt, rotundus nucleus; SH, subhabenular nucleus; SPC, Parvocellular superficial nucleus; SRt,

subrotundic nucleus; Te, telencephalon; VT, ventral thalamus; ZLI, zona limitans intrathalamica.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1099

rudiment acquires this spatial pattern intrinsically or

extrinsically. Previous study has shown that the molecular

regionality in the isolated mouse diencephalic explants

cultured in vitro was remarkably stable (Echevarria et al.,

2001). Fragments of the neural tube that contained the DT

rudiment and some flanking tissues that include the basal

plate, a posterior part of p3, and an anterior part of p1, were

isolated from HH8–12 embryos before the onset of Sox14

and Gbx2 expression (type A in Fig. 2A). After being

cultured in vitro for 72 h, the characteristic patterns of

Sox14 and Gbx2 expression were detected (Fig. 2B), which

were comparable to those in the normal embryo at the

corresponding stages (see Fig. 1L). By contrast, when those

flanking tissues, the boundaries of which are easily

discernible under a dissection microscope, were eliminated

prior to the culture (type B in Fig. 2A), neither Sox14 nor

Gbx2 expression was detected (Fig. 2C). Presence of the

flanking tissues was monitored in part by Shh which is

expressed in the basal plate and the boundary that

demarcates the anterior edge of DT (Fig. 2D,E). The results

of the explant studies are summarized in Table 1. Unlike

Sox14 or Gbx2, generation of Sox2-positive neurons did not

appear to require these tissues (n ¼ 14=14; Fig. 2F,G).

Expression of Tuj1 and MAP2, pan-neuronal markers, was

indistinguishable between these different preparations,

suggesting that neurogenesis in general was not affected

by the removal of the flanking tissues (Fig. 2H,I; data not

shown). No significant difference in cell death was detected

by TUNEL analysis (Fig. 2J,K).

These results suggest that the spatial information is

not intrinsic to the DT rudiment but dependent on the

environmental cues around it. Shh is expressed in the basal

plate of p2 at the time of their onset, and also in the anterior

boundary of the DT, zona limitans intrathalamica (ZLI),

after HH19 (data not shown; Martı et al., 1995; Larsen et al.,

2001; Zeltser et al., 2001). In fact, there is a tight correlation

between expression of Shh and that of Sox14 and Gbx2 in

the cultured explants (see Table 1). These results prompted

us to examine a role of Shh in the pattern formation of the

DT rudiment.

Fig. 2. Tissues adjacent to the DT rudiment are required for the expression

of Sox14 and Gbx2. (A) Experimental paradigms for dissection and culture

of the brain explants isolated from HH8–12 embryos. Cultured explants

which had been prepared as type A (B,D,F,H,J) or type B (C,E,G,I,K) were

fixed after 72 h of culture and stained for Sox14 (blue in B,C), Gbx2 (orange

in B,C), Shh (D,E), Sox2 (F,G) and Tuj1 (H,I), or examined for apoptosis

using TUNEL (J,K). The pineal body is developed in the dorsal tip of the

explant, which is indicative of p2 (asterisk in B–G,J,K). Bar, 0.25 mm.

Table 1

Summary of the explant studies

Age Shh þ Sox14 þ Gbx2 þ Gli1 þ Gli2 þ

Type A

HH12 6/6 20/20 20/20 15/15 15/15

HH11 32/32 28/28 24/24 2/2 2/2

HH10 18/18 14/14 13/13 ND ND

HH9 10/10 6/6 4/4 ND ND

HH8 23/23 13/13 16/16 4/4 4/4

Type B

HH12 0/8 0/20 0/20 12/12 12/12

HH11 0/28 0/26 0/16 14/14 14/14

HH10 0/13 0/15 0/14 ND ND

HH9 1/12 0/10 2/16 6/6 6/6

HH8 4/20 4/32 2/8 ND ND

The number of DT explants positive for the markers listed above which

were dissected from various stages of embryos shown in the left column.

Denominators represent the total number of explants examined. Type A and

B explants are depicted in Fig. 2A.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111100

2.3. Shh induces both Sox14 and Gbx2 in the DT

We first examined whether Shh is capable of inducing the

markers that we identified. Approximately same size of the

type B explants (Fig. 2A) dissected from HH12 embryos

were cultured with various concentrations of recombinant

Shh-N protein (Roelink et al., 1995). At 50 nM, only Gbx2

expression was detected in the explants (Fig. 3B). As the

concentration increased, Sox14 was induced progressively,

while Gbx2 expression appeared reduced (Fig. 3C,D). These

results are represented quantitatively in Fig. 3E, which is

based on 82 specimens that were single stained for either

Sox14 or Gbx2 with the same chromatic substrate. The

number of Tuj1-positive cells in the Shh-treated explants

was comparable to the control (data not shown), which

suggests that the augmentation of these markers expressed

in the postmitotic neurons was not simply due to promotion

of generic neurogenesis. We therefore concluded that Shh

altered the neuronal fates of the cells in the explants.

To address further whether the differential induction of

Sox14 and Gbx2 is indeed caused by a direct action of Shh

signal, we took advantage of the transmembrane protein

Smoothened which transduces the Hh signals (Alcedo et al.,

1996; van den Heuvel and Ingham, 1996). It has been shown

that a constitutively active form of Smoothened (SmoM2)

recapitulated the actions of Shh in a cell-autonomous

manner (Xie et al., 1998; Hynes et al., 2000). The type B

explants (see Fig. 2A) from HH12 embryos were micro-

electroporated with different concentrations of the SmoM2

plasmid using a fine glass needle (Ø ¼ 0.03 mm; Fig. 3H,I;

Haas et al., 2001), so that varying amounts of the

constitutively active Smoothened protein would be pro-

duced. The methodological validity was provided using a

Gfp construct (Fig. 3J,K). For strict comparison, approxi-

mately same locations within p2 were targeted for

electroporation (dorsal 3/4; Fig. 3J,K). As a result, only

Gbx2 was induced by the low level of SmoM2 expression

(2.5 mg/ml; n ¼ 17=21; Fig. 3N,O), whereas Sox14 was

predominantly induced instead by the high level of SmoM2

expression apparently in a cell-autonomous manner

(6.0 mg/ml; n ¼ 11=11; Fig. 3P,Q). Under this experimental

condition, any sign of Shh induction was detected by either

doses of SmoM2 (n ¼ 0=19; data not shown). A similar, but

less clear dose-dependency was also observed with the Shh

plasmid (data not shown). These results indicate that the

graded activities of Shh signaling can selectively induce

Sox14 and Gbx2 without a relay signal. On the other hand,

we did not see any significant change in Sox2 expression in

the DT upon Shh misexpression (n ¼ 12=12; data not

shown), suggesting that some neuronal populations in the

DT are independent of Shh signaling.

Next, we examined whether Shh is indeed required for

the induction of Sox14 and Gbx2 by using a function-

blocking antibody (Ericson et al., 1996; Gunhaga et al.,

2000). The DT rudiments with flanking tissues (type A

explant; Fig. 2A) in which both Sox14 and Gbx2 would be

expressed (see Fig. 2B) were dissected from HH12 embryos,

and cultured in the presence of anti-Shh monoclonal

antibody (5E1; 15 mg/ml). As a consequence, the expression

of Sox14 and Gbx2 was absent or severely reduced in those

explants (n ¼ 8=12; Fig. 4A,B). No significant change in

either the expression of Tuj1 (Fig. 4C,D) or TUNEL

staining (Fig. 4E,F) was observed, suggesting that the

concentration of the antibody used in this assay did not

affect generic neuronal differentiation or apoptosis. We also

estimated the time window for the requirement of Shh by

12 h-pulse-incubation with the antibody at the different time

point during the culture (Fig. 4G). The antibody treatment

during the first 24 h was sufficient to block Sox14 and Gbx2

expression, whereas the later treatment was much less

effective (Fig. 4G). Neuronal differentiation did not take

place markedly until after 24 h of culture (Fig. 4H–J), and

the onset of Sox14 and Gbx2 expression was about after 36 h

of culture (data not shown). Collectively, these results

suggest that Shh is required by the mitotic progenitor cells

in the DT, but to a much less extent by the postmitotic

neurons to maintain these molecular properties.

Thus, this series of experiments demonstrated that Shh

presumably produced by the adjacent tissues is necessary

and sufficient for the proper expression of Sox14 and Gbx2

in the DT.

2.4. The distinct activities of Shh signaling are mediated

by different Gli factors

As Shh signaling is mediated by Gli zinc finger proteins

(Lee et al., 1997; Hynes et al., 1997; Sasaki et al., 1997;

reviewed by Ingham and McMahon, 2001), we hypo-

thesized that these Gli factors may be differently involved

in the induction of Sox14 and Gbx2 in the DT. In support of

this idea, the expression of Gli1 and Gli2 appeared

complimentary in the DT rudiment at HH15 just before

the generation of Sox14- and Gbx2-expressing cells: Gli1

expression is restricted to a band just dorsal to the alar/basal

boundary, whereas Gli2 is expressed dorsal to the Gli1-

expressing domain (Fig. 5A). Soon after this stage (HH18),

the Gli1 expression expands dorsally retaining the highest

intensity towards the alar/basal boundary, and thus the

Gli1- and Gli2-expressing domains became overlapping

(Fig. 5B–E). Meanwhile, Sox14-positive postmitotic cells

emerged at the ventral edge of Gli1-expressing domain

(Fig. 5D), whereas Gbx2-positive cells were distributed over

the Gli2-expressing progenitor zone (Fig. 5E). The third

member, Gli3, was expressed in the most dorsal areas of the

DT, which fit neither the Sox14 nor Gbx2 territory at these

stages (Fig. 5F).

First, we tried to reveal potential differences in Gli1 and

Gli2 simply by misexpression. The pretectum or p1 was

chosen for this assay, because it turned out that the DT

rudiment or p2 did not respond to the change of Gli

expression for unknown reason (data not shown). We have

found that p1 similarly responded to activation of the Hh

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1101

Fig. 3. Shh is necessary and sufficient for the induction of Sox14 and Gbx2. (A–D,F,G) Explants of the DT rudiment without the flanking tissues isolated from

HH12 embryos were cultured in vitro for 72 h (A–D) or 48 h (F,G) with 0 (A,F), 50 (B), 300 (C) or 1250 nM (D,G) of recombinant Shh-N, and then hybridized

for Sox14 (dark blue in A–D) and Gbx2 (brown in A–D), or Gli1 (blue in F, brown in G) and Gli2 (brown in F, blue in G). (E) Quantitative representation of the

induction analysis. The same numbers explants for each treatment [n ¼ 8 (0 nM); n ¼ 11 (150 nM); n ¼ 10 (300 nM); n ¼ 12 (1250 nM)] were stained

individually for Sox14 or Gbx2 with the same chromatic substrate. Proportions of the Sox14- or Gbx2-expressing domains over the explants are represented by

the orange and green bars, respectively. (H) Schematic illustration depicting the microelectroporation for the type B explant. A fine glass capillary filled with

various concentrations of DNA was used for the cathode (I), and electroporation was made focally at the equivalent sites (dorsal 3/4 level) in the type B

explants. As examples, explants electroporated with 2.5 mg/ml (J) or 6.0 mg/ml (K) of the Gfp plasmid were shown. (L–Q0) Type B explants that had been

electroporated with control vector (L–M0), 2.5 mg/ml (N–O0), or 6.0 mg/ml (P–Q0) of constitutively active Smoothened (SmoM2) were cultured for 72 h and

then stained for Sox14 (L,N,P) and Gbx2 (M,O,Q). Locations of exogenous gene expression are visualized by GFP fluorescence (L0,M0,N0,O0,P0,Q0). Total

concentration of DNA was adjusted equally (6.0 mg/ml) for the all experiments. Bars, 0.25 mm for (A)–(D), (F), (G), (J), (K); 0.1 mm for (I), (L)–(Q0).

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111102

signaling (i.e. Shh, SmoM2) to express Sox14 and Gbx2 as

p2 did in the previous assays up to HH14 (Sox14 by Shh,

n ¼ 24=34; Gbx2 by Shh, n ¼ 35=38; Sox14 by SmoM2,

n ¼ 10=13; Gbx2 by SmoM2, n ¼ 13=13; data not shown),

consistent with the notion that the pretectum and DT

progressively segregate by HH16 in the chick (Larsen et al.,

2001). Again, the focal microelectroporation technique for

explants was employed to achieve precise control of the

location and the concentration of DNA. When Gli1 was

electroporated, Sox14 but not Gbx2 was induced, although

the location of this induction was restricted to the subregion

of the pretectm (n ¼ 2=9; Fig. 5I,J). Conversely, electro-

poration of Gli2 into the equivalent sites resulted in

induction of Gbx2 instead of Sox14 (n ¼ 8=12; Fig. 5K,L),

even when the concentration of the Gli2 plasmid was raised

up to four fold (data not shown). In the normal p1, however,

neither Gbx2 nor Sox14 are expressed in a similar way as in

the DT despite this competence and the presence of Shh in

the basal plate of p1 (see Fig. 1F,G). Yet, the potential

difference of Gli1 and Gli2 revealed here could be relevant

to the DT patterning, at least in the context of gene

regulation.

To address further whether Gli1 and Gli2 indeed play a

role in the expression of Sox14 and Gbx2, respectively, the

Gli mutant mice were analyzed. While single mutants for

Gli2 showed indistinguishable expression of Gbx2 and

Sox14 in the DT rudiment from the wild-type (Fig. 6A,B),

the Gli2; Gli3 double mutant mice exhibited a drastic

reduction of Gbx2 expression in the DT, but not of Sox14 at

11.5 days post coitum (dpc) (n ¼ 3=3; Fig. 6C). In these

animals, Shh expression in the ventral CNS was severely

down-regulated (compare Fig. 6D,E). Yet, Shh expression

in the presumptive ZLI was clearly detectable (Fig. 6E).

Moreover LacZ expression driven by the Patched1 promoter

was also detected in the prospective DT and ventral

thalamus (Fig. 6G), suggesting that Hh signaling was still

occurring in the absence of Gli2 and Gli3. While it has been

shown that Gli3 has an antagonizing activity against Hh

signaling (Ruiz i Altaba, 1999; Litingtung and Chiang,

2000), Gli3 has also been implicated in compensating for

loss of Gli2 activities (Mo et al., 1997; Motoyama et al.,

1998). Gli3 single mutants themselves had no defect in the

Gbx2 and Sox14 expression in the DT (data not shown).

Although the previous reports have identified no obvious

abnormality in the ventral patterning of the spinal cord in

the Gli1 mutants (Matise et al., 1998; Park et al., 2000; Bai

et al., 2002), the Sox14 expression in the DT, but not in the

pretectum, was absent or severely reduced at 11.5 dpc

(Fig. 7A,B). In contrast, the Gbx2 expression in the DT was

indistinguishable from the wild-type or heterozygotes

(Fig. 7C,D). No alteration in Shh or Patched1-LacZ

expression has been detected in the Gli1 mutants (data not

shown; Fig. 7E,F). We then examined whether Shh at any

dose can induce Sox14 in the absence of Gli1. The DT

explants equivalent to the type A in the chick were prepared

from 9.5 dpc Gli1 mutants and cultured in the presence of

various concentrations of recombinant Shh-N for 3 days. As

much as 1800 nM of Shh-N did not induce Sox14 in the

Gli1 2/2 explants (data not shown; Fig. 7G), whereas Sox14

was constantly induced in the wild type and Gli1 þ/2

explants with Shh-N above 600 nM (data not shown). When

Gli1 was electroporated to the Gli1-deficient explants,

Fig. 4. Explants of the DT rudiment with the flanking tissues dissected form

HH12 embryos cultured with normal mouse IgG (A,C,E) or anti-Shh

antibodies (B,D,F) for 48 h (C,D) and 72 h (A,B,E,F) were stained for

Sox14 (blue in A,B), Gbx2 (red in A,B) and Tuj1 (C,D), or examined for

apoptosis using TUNEL (E,F). Both Sox14 and Gbx2 expressions were

greatly diminished by anti-Shh antibodies without significant alterations in

neuronal differentiation or apoptosis. (G) Temporal change of the effect of

anti-Shh antibody on Sox14 and Gbx2 expression. The antibody was added

to the culture media for 12 h at different time points during the culture (72 h

in total) as indicated in the horizontal scale. Pixels of Sox14- (red for control

IgG, magenta for anti-Shh) and Gbx2-positive areas (turquoise for control

IgG, purple for anti-Shh) in the explants were calculated by NIH image

software. (H–J) HH12 type A explants at 12 (H), 18 (I), and 24 h (J) of

culture stained for Tuj1. Bar, 0.25 mm.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1103

Sox14 was induced where the exogenous gene was

expressed (Fig. 7I), although we found that a considerable

dose of Shh-N (.600 nM) was needed to for this induction

(Fig. 7G). Importantly, this was not achieved by exogenous

Gli2, except to a very low extent only in the presence of very

high dose of Shh-N (.1250 nM), indicating overt pre-

cedence of Gli1 for the Sox14 induction in the DT (Fig. 7G).

2.5. Mutual repression between Sox14 and Gbx2

While combination of the differential Gli expression and

the graded distribution of Shh protein could theoretically

define the precise positions of the distinct neuronal

subtypes, the complementary expression of Sox14 and

Gbx2 (see Fig. 1N) suggested that there may be mutual

repression between these transcription factors. In fact, when

Gbx2 was electroporated in the prospective Sox14 territory

at HH12, the Sox14 expression was suppressed (n ¼ 7=7;

Fig. 8B). Conversely, forced expression of Sox14 resulted in

suppression of Gbx2 apparently in a cell-autonomous

manner (n ¼ 8=8; Fig. 8D). Importantly, this cross regu-

lation is specific, as neither of them affected the expression

of Sox2 (for Sox14, n ¼ 8=8; Fig. 8F; data not shown for

Gbx2, n ¼ 12=12). On the other hand, we found that the

exogenous Gbx2 can induce endogenous Gbx2 (Fig. 8G). At

this stage (HH12 , ), exogenous Gbx2 did not induce Fgf8

(n ¼ 16=16; data not shown) as has been reported for the

earlier stages (Millet et al., 1999; Katahira et al., 2000).

Moreover, co-electroporation of Gbx2 and the dominant-

negative form of Fgfr3(Amaya et al., 1993; Kobayashi et al.,

2002) did not perturb this induction (n ¼ 4=4; data not

shown). Therefore it is unlikely that this auto-induction was

due to the cross-regulatory loop involved in the mid-

hindbrain development (Garda et al., 2001). No such auto-

induction, however, was observed for Sox14 (Fig. 8H).

These results suggest that the initial pattern created by

the diffusible signal is then consolidated by the transcrip-

tional regulations of the target genes in the DT, as reported

Fig. 5. Gli1 and Gli2 can selectively induce Sox14 and Gbx2, respectively. (A) A coronal section of the DT at HH15 double stained for Gli1 (blue) and Gli2

(orange). (B,C) HH18 chick brains whole-mount stained for Gli1 (B) and Gli2 (C). (D,E,F) Adjacent transverse sections of the DT at HH19 double stained for

Gli1 (blue) and Sox14 (orange, arrowhead) (D), for Gli2 (blue) and Gbx2 (orange) (E), or for Gli3 (purple) and Gbx2 (orange) (F). (F) is slightly more dorsal

than (D),(E). P1 explants dissected from HH12 embryos that had been electroporated with control vector (G–H0), mGli1 (I–J0) and mGli2 (K–L0) were cultured

for 72 h and stained for Sox14 (G,I,K) and Gbx2 (H,J,L). Locations of the exogenous gene expression are visualized by GFP fluorescence (G0,H0,I0,J0,K0,L0).

Bars, 0.1 mm for (A),(D)–(K0); 0.4 mm for (B),(C).

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111104

for many other developmental situations (reviewed by

Jessell, 2000).

3. Discussion

In this paper, we have shown that three regulatory genes,

Gbx2, Sox14 and Sox2 are expressed in the specific DT

nuclei (Fig. 1). Gbx2 encodes a homeodomain transcription

factor essential for the development of many DT nuclei

(Miyashita-Lin et al., 1999). Sox2 and Sox14, which belong

to subgroups B1 and B2 of the Sox family, respectively,

encode a group of proteins that carry a HMG DNA-binding

domain (Uchikawa et al., 1999; Hargrave et al., 2000;

reviewed by Kamachi et al., 2000). Detailed descriptions of

their expression in relation to the histogenesis of DT will be

published elsewhere (Hashimoto-Torii et al., in prepara-

tion). Here we will discuss the patterning mechanisms

operating in the early development of the DT with particular

regard to the Hh signaling pathway.

3.1. Differential involvement of the Gli factors in Hh

signaling

While Shh has been shown to act as a morphogen

inducing distinct genes in a concentration dependent

manner (Roelink et al., 1995; Ericson et al., 1997), the

underlying mechanisms are not yet fully understood. Our

present findings on the induction of Sox14 and Gbx2 by Shh

and Gli factors provide an important insight into this issue.

Namely, the high dose of Shh signaling and exogenous Gli1

led to Sox14 induction, whereas the low dose of Shh

signaling and Gli2 led to Gbx2 expression. This raises the

possibility that the high and low doses of Shh signals are

selectively mediated by Gli1 and Gli2, respectively. While

all Gli proteins recognize a consensus DNA sequence

(Kinzler and Vogelstein, 1990; Vortkamp et al., 1995), both

gain- and loss-of-function studies have revealed that Gli

factors differ in the property of Hh target gene regulation

(Ruiz i Altaba, 1998, 1999; Persson et al., 2002; Karlstrom

et al., 2003). For instance, it has been shown that Gli1 can

induce FoxA4 (Hnf3b), a floor plate marker, and Nkx2.1 in

the frog embryos, whereas Gli2 can induce motor neuron

marker HB9 instead (Ruiz i Altaba, 1998). Our in vitro

result that exogenous Gli2 was not able to induce Sox14

efficiently in the Gli1-deficient explants even in the

presence of very high dose of Shh-N would support this

possibility. Although it appears that Gli3 is also required for

inducing Gbx2 from the double mutant phenotype

(Fig. 6B,C), we think that it is unlikely that Gli3 plays a

role in the onset of Gbx2 expression in the normal situation,

simply because it is not expressed in the ventral but dorsal

domain of DT (Fig. 5F, Grove et al., 1998).

Curiously, however, no obvious defect has been reported

in the spinal cord of the Gli1-deficient animals (Matise et al.,

1998; Park et al., 2000; Bai et al., 2002). A conceivable

explanation is that other Gli factors, most likely to be Gli2,

may compensate for Gli1 functions in some situations. In

support of this notion, the Gli1; Gli2 double mutants exhibit

more severe abnormalities in the Hh-dependent processes

than either single mutant (Park et al., 2000). Alternatively,

for spatial patterning, the assigned functions to each Gli

factor may vary among the different regions of the CNS.

Even the morphogen-like action of Shh may also be variably

responsible for the spatial patterning among the different

regions of the CNS. For instance, Agarwala et al. (2001)

have recently provided in vivo evidence that an ectopic

focal source of Shh is solely sufficient for generating the

exquisite spatial pattern in the ventral midbrain as predicted

by the hypothesis that Shh acts as a morphogen. However,

Fig. 6. Gli2 and Gli3 are essential for Gbx2 expression, but not for Sox14.

(A–C) Lateral view of the DT of an 11.5 dpc wild-type A, Gli2 single (B)

and Gli2; Gli3 double mutant (C) double stained for Sox14 (blue) and Gbx2

(red). Note that Gbx2 expression is absent in the double mutant (C),

whereas Sox14 expression persists (arrow in C). (D,E) Head regions of a

9.5 dpc wild-type (D) and Gli2; Gli3 double homozygotes (E) in situ

hybridized for Shh. (F,G) Ptc1–LacZ expression revealed by X-gal staining

of the dissected brains from a 10.0 dpc Gli2 heterozygote (F) and Gli2; Gli3

double homozygote (G). Expression of Shh and Ptc1–LacZ persists around

the ZLI (arrow in D–G) in the double mutants. Bars, 0.5 mm.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1105

this has not been demonstrated in vivo in the spinal cord.

Regarding this, it may be noteworthy that the forebrain and

midbrain have a considerably larger alar plate area than the

spinal cord, which dramatically expands during this

patterning period. Consequently, the floor and basal plates

become located more distant from the roof plate, a tissue

also implicated in the dorsoventral patterning of the CNS.

Thus, the graded Shh signaling, together with differential

recruitment of Gli factors, may play more prominent roles in

establishing the dorsoventral patterning in these regions of

the CNS.

Gli genes are differentially expressed in the mitotic

progenitor cells of the chick DT (Fig. 5A–F), and therefore

could serve as a ‘pre-pattern’ for the action of Shh at the

neurogenic period when the Sox14- and Gbx2-positive

neurons are born. In fact, differential Gli expression

persisted in the type B explants from which the Shh

expressing regions had been excluded (see Fig. 3F), and the

induction of Sox14 by Shh-N appeared somewhat correlated

with the pattern of Gli1 expression (see Fig. 3C,F). This idea

would not necessarily be argued by the fact that Sox14 can

be induced by exogenous SmoM2 in the prospective Gbx2

territory (see Fig. 3D,P), because the expression of Gli

genes can also be regulated by Shh signaling. Transcrip-

tional activation of Gli1 by Shh has been shown previously

(Marigo et al., 1996; Grindley et al., 1997; Hynes et al.,

1997; Lee et al., 1997; Bai et al., 2002; Karlstrom et al.,

2003). We also found that a high dose of Shh or SmoM2

induced Gli1 and repressed Gli2, whereas a low dose

upregulated Gli2 expression (Fig. 3G; KH and KS,

unpublished).

Nevertheless, the Gli1-expressing domain appears con-

siderably broader than that of Sox14, even extending into the

Gbx2 domain (see Fig. 5D). Thus, the ‘Gli code’ does not

Fig. 7. Gli1 with high doses of Shh is required for the induction of Sox14. (A–D) Lateral views of the DT in an 11.5 dpc heterozygous (A,C) and homozygous

for Gli1 (B,D) stained for Sox14 (A,B) and Gbx2 (C,D). Insets are horizontal sections of the specimens as indicated by broken lines (A,B). The expression of

Sox14 is severely down-regulated in the DT of the Gli1 mutant (arrow in B), while its expression in p1 is unchanged (arrowhead in B). The Gbx2 expression is

not affected in the Gli1 mutant (D). (E,F) Ptc1–LacZ expression in a 10.5 dpc hetrozygote (E) and homozygote for Gli1 (F). (G) Quantification of the Sox14

induction in the Gli1-deficient DT rudiments by exogenous Gli1 and Gli2 in the presence of various concentrations of Shh-N. The percentage of the Sox14-

positive area per the GFP-expressing area was calculated. The diencephalic explants from 9.5 dpc Gli1 mutants were electroporated with a control vector (H)

and Gli1 (I), cultured in the presence of 1250 nM of Shh-N for 72 h, and then stained for Sox14 (H,I). Dorsal is to the top, and anterior is to the right. Sox14

expression in the Gli1-deficient DT is up regulated where the exogenous Gli1 was introduced in the dorsal portion of the explant (arrow in I). Location of the

exogenous gene expression is visualized by GFP fluorescence (H0,I0). Bars, 0.25 mm.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111106

seem to be solely responsible for defining the position of the

Sox14-positive neurons. Concerning this issue, it is inter-

esting that Gli1 and Gli2 may have different thresholds of

Shh dose to function. Although the previous studies have

reported that Gli1 activity is not post-translationally but

transcriptionally regulated by Hh signaling (Epstein et al.,

1996; Marigo et al., 1996; Dai et al., 1999), our in vitro

results suggest that Gli1 requires a high dose of Shh-N to

induces Sox14 (see Fig. 7G). In fact, Sox14 was not

expressed in the type B explants despite the persistent

expression of Gli1 (see Fig. 3A,F). This might explain why

the p2 alar plate was insensitive to the misexpression of

Gli1, and the Sox14 induction solely by exogenous Gli1 was

considerably inefficient in the p1 ðn ¼ 2=9Þ: Since Gbx2 was

induced by a low dose of Shh-N which is likely to be

mediated by Gli2, the lower dose of Shh may be sufficient to

activate Gli2 compared to Gli1. In Drosophila, Ci is the only

Gli factor that mediates Hh signaling (Orenic et al., 1990;

Von Ohlen and Hooper, 1997; Methot and Basler, 2001).

Previous studies have established that the Ci activities are

regulated at the multiple levels, such as proteolysis,

subcellular distribution, and activation, all of which are

dependent on Hh signaling (Aza-Blanc et al., 1997; Wang

and Holmgren, 2000; Methot and Basler, 1999, 2000). Thus

Gli1 and Gli2 may differ in the susceptibility to those Hh-

dependent regulations. Alternatively, the high dose of Shh

may indirectly allow Gli1 to induce Sox14, such as by

repressing antagonistic cues. For instance, it was shown that

Shh blocks Gli3 function to liberate Hh target genes from

repression (Lee et al, 1997; Ruiz i Altaba, 1998; Sasaki et al.,

1999; Von Mering and Basler, 1999; Aza-Blanc et al., 2000;

Litingtung and Chiang, 2000; Persson et al., 2002; Rallu

et al., 2002; Wijgerde et al., 2002). In the present case,

however, Gli3 does not appear to be the one, since Gli3 is

not expressed in the Sox14 territory, and there was in fact no

change in Sox14 expression in the Gli3 mutants. Members

of Bmp family expressed at the dorsal most portion of p2

(Furuta et al., 1997) could be good candidates, as it was

shown that Bmp signaling counteract with Shh signaling

(Liem et al., 2000).

Overall, the differential expression of Gli1 and Gli2

which possess different preference for the target genes, in

cooperation with the graded distribution of Shh protein,

perhaps play a role in assuring the precise birthplaces of the

distinct neuronal subtypes in the DT. The recruitment of

different Gli factors as an effector of Hh signaling must be a

key for further diversity and complexity of the patterns

generated by this signaling pathway during evolution as

discussed previously (Aza-Blanc et al., 2000; Persson et al.,

2002; Karlstrom et al., 2003).

3.2. ‘Ventral patterning’ in the alar plate

Our present finding that Gbx2 expression in the DT

rudiment is dependent on Shh signaling is somewhat

puzzling, as Gbx2 is expressed in and required for a large

region of the DT that is thought to be derived from the alar

plate (Fig. 4; Puelles and Rubenstein, 1993; Miyashita-Lin

et al., 1999). Studies in the spinal cord as well as other brain

territories have established that Shh is involved in the

ventral patterning of the CNS (the basal plate) and the alar

plate derivatives are thought to be negatively regulated by

Shh signaling (Ericson et al., 1996; Watanabe and

Nakamura, 2000). Furthermore, recent findings in the

developing telencephalon have led to the notion that

GABAergic and cholinergic interneurons are born in the

basal telencephalon presumably through the action of Shh,

whereas most glutamatergic neurons are not (reviewed by

Fig. 8. Cross-regulation between Gbx2 and Sox14. Oblique horizontal sections of HH21 embryos at the level of p2 which had been electroporated with a

control vector (A,C,E), mGbx2 (B), Sox14 (D,F) at HH12 were double stained for Sox14 (blue in A,B; brown in C,D) and cGbx2 (brown in A,B; blue in C,D)

which cross-hybridized with mGbx2 RNA, or single stained for Sox2 (E,F). Arrows indicate repression by the exogenous genes. (G,H) Dissected brains from

HH21 embryos electroporated at HH12 with mGbx2 (G) or Sox14 (H) hybridized with Gbx2- (G) or Sox14-30UTR probes (H). Exogenous Gbx2 induced

endogenous Gbx2 ectopically (arrowhead in G). Locations of exogenous gene expression are visualized by GFP fluorescence (F0,G0,H0). Bars, 0.4 mm.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1107

Wilson and Rubenstein, 2000; Corbin et al., 2001). Given

that the majority of DT neurons are glutamatergic

(Shigemoto et al., 1992), the dependence of Gbx2 on Shh

again does not make good sense. However, we found that

some Gbx2 expression, particularly at later stages, is

independent of Shh activity. For instance, DT rudiments

cultured in the presence of anti-Shh antibody did show

substantial Gbx2 expression when cultured for longer

periods (KH and KS, unpublished). Moreover, in Gli2;

Gli3 double mutant mice, Gbx2 expression was detectable

in the DT after 12.5 dpc (KH and KS, unpublished). These

findings raise the possibility that some Gbx2-positive cells

in the DT are independent of Shh activity, and these may

correspond to the glutamatergic DT neurons. Other

molecular markers that further distinguish subsets of cells

within the Gbx2-positive population (e.g. Sox2) will clarify

the precise requirement of Shh for the Gbx2-positive cells in

the DT. On the other hand, the notion for the generation of

GABAergic neurons in the telencephalon is consistent with

the fact that Sox14-positive cells in the DT, which are

dependent on Shh activity, are mostly GABAergic (Hashi-

moto-Torii et al., in preparation).

Thus, it appears that certain populations of DT neurons

are specified by Shh emanating from the basal plate, while

the DT itself is specified as the alar plate. This evokes an

idea that Shh expression, which spreads in the basal plate of

the anterior CNS (fore-midbrain region), may be respon-

sible for generating further cellular diversity within the alar

plate. In addition, it is conceivable that the ZLI where Shh

becomes expressed as development proceeds also contri-

butes to the further complexity in the DT as postulated

(Larsen et al., 2001; Zeltser et al., 2001). It has been shown

that the notochord plays a primary role in the ventral

specification of the spinal cord, which is then taken over by

the floor plate (Ericson et al., 1996). It could be that, in the

fore-midbrain region, signals including Shh emanating from

the axial mesoderm, the prechordal plate and notochord,

first specify the floor and basal plates, and subsequently Shh

released from the basal plate patterns the adjacent alar plate

cells. The initial Shh signaling from the axial mesoderm

might also set up the Gli patterns which serves as a pre-

pattern in the alar plate for the subsequent Shh signal from

the basal plate during the neurogenic periods as discussed

above.

4. Experimental procedures

4.1. Embryo manipulation

Embryological manipulations were performed in White

Leghorn chick embryos following standard protocols. Eggs

were incubated at 38 8C and the embryos were staged

according to Hamburger and Hamilton (Hamburger and

Hamilton, 1951).

4.2. Explant culture

In vitro culture of isolated brain fragments was carried

out as described previously (Shimamura and Takeichi,

1992), with slight modifications. Explants were isolated

from HH8-15 chick and 9.5–10 dpc mouse embryos and

cultured on the Nucleoporew filters (#110414; Costar)

floating on DMEM supplemented with 10% fetal bovine

serum (Nissui) in a CO2 incubator at 37 8C for 0.5–4 days.

No significant necrosis or obvious abnormalities in

morphogenesis were detected (see Figs. 2,4). For the Shh

titration experiments, recombinant mouse Shh-N (R and D

systems) was added to the culture media at 0–1800 nM. For

blocking of the Shh activity, monoclonal anti-Shh antibody

(5E1; DSHB, Iowa) was used (Ericson et al., 1996; Gunhaga

et al., 2000).

4.3. Electroporation

Full-length cDNA encoding mouse Gbx2 (a gift from Dr

G. Martin), chick Sox14 (a gift from Dr H. Kondoh), and a

constitutively active form of human Smoothened (SmoM2; a

gift from Dr A. Rosenthal) were inserted into the pCAGGS

vector (Tokui et al., 1997). A cDNA encoding the N-

terminal form of mouse Shh was inserted into the pEF-BOS

vector. Full-length mouse Gli1 and Gli2 cDNAs inserted

into the pCDNA3.1 vector were provided by Dr H. Sasaki

(Sasaki et al., 1999). Electroporation for HH8-14 chick

embryos was carried out in ovo as described (Funahashi

et al., 1999; Kobayashi et al., 2002). To achieve precise

control of DNA concentration and of transfection sites, a

method for microelectroporation using a microcapillary

electrode was employed for the explants (Haas et al., 2001).

Briefly, a pore of 0.03–4 mm diameter glass needle was

made from a glass capillary (Sutter, BF100-50-10) by a

Sutter P-97/IVF. A DNA solution was loaded into the tip of

glass needle and backfilled with mineral oil. A thin platinum

wire was inserted into the needle to function as a cathode.

For all the electroporation experiments, pCAGGS-GFP

plasmid (Momose et al., 1999) was mixed to 1 mg/ml, so

that the sites of exogenous gene expression can be

monitored under an epi-fluorescence dissecting microscope

(Leica MZFL3).

4.4. Mice

Mutant mouse strains, Ptc1 LacZ (Goodrich et al., 1997),

Gli1 zfd (Matise et al., 1998), Gli2 zfd (Mo et al., 1997; Ding

et al., 1998) and Gli3 XtJ (Johnson, 1967; Buscher et al.,

1997) were maintained in a mixed 129/Sv and CD1

background. Generation and analysis of Gli2; Gli3 double

mutants were performed as described (Mo et al., 1997).

To generate Gli2 zfd/zfd; Gli3 XtJ/XtJ; Ptc1 þ/LacZ mice, the

Gli2 þ/zfd; Gli3 þ/XtJ; Ptc1 þ/LacZ male mice were crossed

with Gli2 þ/zfd; Gli3 þ/XtJ female mice. Genotypes of mutant

embryos were determined by PCR using a standard

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111108

protocol. For genotyping the Gli1 mutant, GlidF

(50TTTGAAGGCTGTCGGAAGTCC30) and GlidR

(50TCATTGGAGTGGGTCCGATTC30) primers for ampli-

fication of the 165 bp fragment derived from the zinc finger

domain, as well as the standard primers for detection of Neo

cassette were used.

4.5. In situ hybridization

In situ hybridization of whole-mount specimens and

cryosectioned materials were performed as described

previously (Shimamura et al., 1995; Uchikawa et al.,

1999). Probes for mSox14 and cGli3 were obtained by

PCR. Probes for cSox2, cSox14, mGbx2, cGbx2, cShh,

cGli1, cGli2 and mShh were kind gifts from Drs H. Kondoh,

G. Martin, C. Tabin, and A. McMahon, respectively. Probes

for cSox14- and cGbx2-30 UTR were generated from the

Sac II–Apa I and HincII–Eco RI fragments, respectively.

For quantification of the labeled cells, NIH image software

was used as described previously (Ishihara et al., 2001).

Since cells expressing Sox14 or Gbx2 are distributed two-

dimensionally at the time of their emergence, pixels of the

labeled areas with an intensity above a given optical

thresholds were scored as the references for the number of

positive cells. When necessary, proportions of the labeled

areas to the defined fields of the explants were calculated as

an induction index.

4.6. Histostaining

Immunostaining of cryosections was performed as

described (Yoon et al., 2000). The antibodies used in this

study were monoclonal anti-Tuj1 (BABCO), anti-NeuN

(Chemicon), Texas Red-conjugated and HRP-conjugated

anti-mouse IgG (Jackson Lab.). Cell death was detected by

the TUNEL assay (Gavrieli et al., 1992; Wijsman et al.,

1993), using the Roche Kit as previously described (Teillet

et al., 1998). b-galactosidase staining for Ptc1 expression

was carried out as previously reported (Ding et al., 1998).

4.7. Nomenclature

The neuroanatomical nomenclature for chick brain used

in this study was adopted from previous literature (Redies

et al., 2000; Yoon et al., 2000; Puelles, 2001; Martınez-de-

la-Torre et al., 2002).

Acknowledgements

We are grateful to Drs Christoph Redies, Luis Puelles

and Loreta Medina for insightful discussions, Daisuke

Kobayashi for critical reading of the manuscript. We thank

Drs Alex Joyner, Jun Aruga, and Shigeru Makino for

providing the Gli1 KO mice, and Drs Masanori Uchikawa

and Hisato Kondoh for reagents and technical advice. This

work was supported by grants-in-aid from Ministry of

Education, Science, Sports and Culture of Japan. KH is the

recipient of a Research Fellowships for the Promotion of

Science for Young Scientists from the JSPS.

References

Agarwala, S., Sanders, T.A., Ragsdale, C.W., 2001. Sonic hedgehog control

of size and shape in midbrain pattern formation. Science 291,

2147–2150.

Alcedo, J., Ayzenzon, M., Von Ohlen, T., Noll, M., Hooper, J.E., 1996. The

Drosophila smoothened gene encodes a seven-pass membrane protein, a

putative receptor for the hedgehog signal. Cell 86, 221–232.

Amaya, E., Stein, P.A., Musci, T.J., Kirschner, M.W., 1993. FGF signalling

in the early specification of mesoderm in Xenopus. Development 118,

477–487.

Aza-Blanc, P., Ramirez-Weber, F.A., Laget, M.P., Schwartz, C., Kornberg,

T.B., 1997. Proteolysis that is inhibited by hedgehog targets Cubitus

interruptus protein to the nucleus and converts it to a repressor. Cell 89,

1043–1053.

Aza-Blanc, P., Kornberg, T.B., 1999. Ci: a complex transducer of the

hedgehog signal. Trends Genet. 15, 458–462.

Aza-Blanc, P., Lin, H.Y., Ruiz i Altaba, A., Kornberg, T.B., 2000.

Expression of the vertebrate Gli proteins in Drosophila reveals a

distribution of activator and repressor activities. Development 127,

4293–4301.

Bai, C.B., Auerbach, W., Lee, J.S., Stephen, D., Joyner, A.L., 2002. Gli2,

but not Gli1, is required for initial Shh signaling and ectopic activation

of the Shh pathway. Development 129, 4753–4761.

Buscher, D., Bosse, B., Heymer, J., Ruther, U., 1997. Evidence for genetic

control of Sonic hedgehog by Gli3 in mouse limb development. Mech.

Dev. 62, 175–182.

Corbin, J.G., Nery, S., Fishell, G., 2001. Telencephalic cells take a tangent:

non-radial migration in the mammalian forebrain. Nat. Neurosci. 1,

1177–1182.

Dai, P., Akimaru, H., Tanaka, Y., Maekawa, T., Nakafuku, M., Ishii, S.,

1999. Sonic Hedgehog-induced activation of the Gli1 promoter is

mediated by GLI3. J. Biol. Chem. 274, 8143–8152.

Ding, Q., Motoyama, J., Gasca, S., Mo, R., Sasaki, H., Rossant, J., Hui,

C.C., 1998. Diminished Sonic hedgehog signaling and lack of floor

plate differentiation in Gli2 mutant mice. Development 125,

2533–2543.

Echevarria, D., Vieira, C., Martinez, S., 2001. Mammalian neural tube

grafting experiments: an in vitro system for mouse experimental

embryology. Int. J. Dev. Biol. 45, 895–902.

Epstein, D.J., Martı, E., Scott, M.P., McMahon, A.P., 1996. Antagonizing

cAMP-dependent protein kinase A in the dorsal CNS activates a

conserved Sonic hedgehog signaling pathway. Development 122,

2885–2894.

Ericson, J., Morton, S., Kawakami, A., Roelink, H., Jessell, T.M., 1996.

Two critical periods of Sonic Hedgehog signaling required for the

specification of motor neuron identity. Cell 87, 661–673.

Ericson, J., Rashbass, P., Schedl, A., Brenner-Morton, S., Kawakami, A.,

van Heyningen, V., Jessell, T.M., Briscoe, J., 1997. Pax6 controls

progenitor cell identity and neuronal fate in response to graded Shh

signaling. Cell 90, 169–180.

Funahashi, J., Okafuji, T., Ohuchi, H., Noji, S., Tanaka, H., Nakamura, H.,

1999. Role of Pax-5 in the regulation of a mid-hindbrain organizer’s

activity. Dev. Growth Differ. 41, 59–72.

Furuta, Y., Piston, D.W., Hogan, B.L., 1997. Bone morphogenetic proteins

(BMPs) as regulators of dorsal forebrain development. Development

124, 2203–2212.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1109

Garda, A.L., Echevarrıa, D., Martınez, S., 2001. Neuroepithelial co-

expression of Gbx2 and Otx2 precedes Fgf8 expression in the isthmic

organizer. Mech. Dev. 101, 111–118.

Gavrieli, Y., Sherman, Y., Ben-Sasson, S.A., 1992. Identification of

programmed cell death in situ via specific labeling of nuclear DNA

fragmentation. J. Cell. Biol. 119, 493–501.

Goodrich, L.V., Milenkovic, L., Higgins, K.M., Scott, M.P., 1997. Altered

neural cell fates and medulloblastoma in mouse patched mutants.

Science 277, 1109–1113.

Grindley, J.C., Bellusci, S., Perkins, D., Hogan, B.L., 1997. Evidence for

the involvement of the Gli gene family in embryonic mouse lung

development. Dev. Biol. 188, 337–348.

Grove, E.A., Tole, S., Limon, J., Yip, L., Ragsdale, C.W., 1998. The hem of

the embryonic cerebral cortex is defined by the expression of multiple

Wnt genes and is compromised in Gli3-deficient mice. Development

125, 2315–2325.

Gunhaga, L., Jessell, T.M., Edlund, T., 2000. Sonic hedgehog signaling at

gastrula stages specifies ventral telencephalic cells in the chick embryo.

Development 127, 3283–3293.

Haas, K., Sin, W.C., Javaherian, A., Li, Z., Cline, H.T., 2001. Single-cell

electroporation for gene transfer in vivo. Neuron 29, 583–591.

Hamburger, V., Hamilton, H.J., 1951. A series of normal stages in the

development of the chick embryo. J. Morphol. 94, 257.

Hargrave, M., Karunaratne, A., Cox, L., Wood, S., Koopman, P., Yamada,

T., 2000. The HMG box transcription factor gene Sox14 marks a novel

subset of ventral interneurons and is regulated by sonic hedgehog. Dev.

Biol. 219, 142–153.

Hui, C.C., Slusarski, D., Platt, K.A., Holmgren, R., Joyner, A.L., 1994.

Expression of three mouse homologs of the Drosophila segment

polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3, in ectoderm- and

mesoderm-derived tissues suggests multiple roles during postimplanta-

tion development. Dev. Biol. 162, 402–413.

Hynes, M., Stone, D.M., Dowd, M., Pitts-Meek, S., Goddard, A., Gurney,

A., Rosenthal, A., 1997. Control of cell pattern in the neural tube by the

zinc finger transcription factor and oncogene Gli-1. Neuron 19, 15–26.

Hynes, M., Ye, W., Wang, K., Stone, D., Murone, M., Sauvage, F.,

Rosenthal, A., 2000. The seven-transmembrane receptor smoothened

cell-autonomously induces multiple ventral cell types. Nat. Neurosci. 3,

41–46.

Ingham, P.W., McMahon, A.P., 2001. Hedgehog signaling in animal

development: paradigms and principles. Genes Dev. 15, 3059–3087.

Ishihara, T., Araki, T., Sakuma, Y., 2001. Two distinct populations of

neurons expressing nitric oxide synthase mRNA in the female rat

preoptic area: site specific changes induced by sex steroids. J. Nippon

Med. Sch. 68, 328–334.

Jessell, T.M., 2000. Neuronal specification in the spinal cord: inductive

signals and transcriptional codes. Nat. Rev. Genet. 1, 20–29.

Jessell, T.M., Sanes, J.R., 2000. Development. The decade of the

developing brain. Curr. Opin. Neurobiol. 10, 599–611.

Johnson, D.R., 1967. Extra-toes: a new mutant gene causing multiple

abnormalities in the mouse. J. Embryol. Exp. Morphol. 175, 43–81.

Jones, E., 1985. The Thalamus, Plenum, New York.

Jones, E., 1998. A new view of specific and nonspecific thalamocortical

connections. Adv. Neurol. 77, 44–71.

Kamachi, Y., Uchikawa, M., Kondoh, H., 2000. Pairing SOX off: with

partners in the regulation of embryonic development. Trends Genet. 16,

182–187.

Karlstrom, R.O., Tyurina, O.V., Kawakami, A., Nishioka, N., Talbot, W.S.,

Sasaki, H., Schier, A.F., 2003. Genetic analysis of zebrafish gli1 and

gli2 reveals divergent requirements for gli genes in vertebrate

development. Development 130, 1549–1564.

Katahira, T., Sato, T., Sugiyama, S., Okafuji, T., Araki, I., Funahashi, J.,

Nakamura, H., 2000. Interaction between Otx2 and Gbx2 defines the

organizing center for the optic tectum. Mech.Dev. 91, 43–52.

Kinzler, K.W., Vogelstein, B., 1990. The GLI gene encodes a nuclear

protein which binds specific sequences in the human genome. Mol.

Cell. Biol. 10, 634–642.

Kobayashi, D., Kobayashi, M., Matsumoto, K., Ogura, T., Nakafuku, M.,

Shimamura, K., 2002. Early subdivisions in the neural plate define

distinct competence for inductive signals. Development 129, 83–93.

Koebernick, K., Pieler, T., 2002. Gli-type zinc finger proteins as bipotential

transducers of Hedgehog signaling. Differentiation 70, 69–76.

Larsen, C.W., Zeltser, L.M., Lumsden, A., 2001. Boundary formation and

compartition in the avian diencephalon. J. Neurosci. 21, 4699–4711.

Lee, J., Platt, K.A., Censullo, P., Ruiz i Altaba, A., 1997. Gli1 is a target of

Sonic hedgehog that induces ventral neural tube development.

Development 124, 2537–2552.

Liem, K.F. Jr., Jessell, T.M., Briscoe, J., 2000. Regulation of the neural

patterning activity of sonic hedgehog by secreted BMP inhibitors

expressed by notochord and somites. Development 127, 4855–4866.

Litingtung, Y., Chiang, C., 2000. Specification of ventral neuron types is

mediated by an antagonistic interaction between Shh and Gli3. Nat.

Neurosci. 3, 979–985.

Lumsden, A., Krumlauf, R., 1996. Patterning the vertebrate neuraxis.

Science 274, 1109–1115.

Marigo, V., Johnson, R.L., Vortkamp, A., Tabin, C.J., 1996. Sonic

hedgehog differentially regulates expression of GLI and GLI3 during

limb development. Dev. Biol. 180, 273–283.

Martı, E., Takada, R., Bumcrot, D.A., Sasaki, H., McMahon, A.P., 1995.

Distribution of Sonic hedgehog peptides in the developing chick and

mouse embryo. Development 121, 2537–2547.

Martınez-de-la-Torre, M., Garda, A.L., Puelles, E., Puelles, L., 2002. Gbx2

expression in the late embryonic chick dorsal thalamus. Brain Res. Bull.

57, 435–438.

Matise, M.P., Epstein, D.J., Park, H.L., Platt, K.A., Joyner, A.L., 1998. Gli2

is required for induction of floor plate and adjacent cells, but not most

ventral neurons in the mouse central nervous system. Development 125,

2759–2770.

Methot, N., Basler, K., 1999. Hedgehog controls limb development by

regulating the activities of distinct transcriptional activator and

repressor forms of Cubitus interruptus. Cell 96, 819–831.

Methot, N., Basler, K., 2001. An absolute requirement for Cubitus

interruptus in Hedgehog signaling. Development 128, 733–742.

Millet, S., Campbell, K., Epstein, D.J., Losos, K., Harris, E., Joyner, A.L.,

1999. A role for Gbx2 in repression of Otx2 and positioning the mid/

hindbrain organizer. Nature 401, 161–164.

Miyashita-Lin, E.M., Hevner, R., Wassarman, K.M., Martinez, S.,

Rubenstein, J.L., 1999. Early neocortical regionalization in the absence

of thalamic innervation. Science 285, 906–909.

Mo, R., Freer, A.M., Zinyk, D., Crackower, M.A., Heng, H.H.Q., Chik,

K.W., et al., 1997. Specific and redundant functions of Gli2 and Gli3

zinc finger genes in skeletal patterning and development. Development

124, 113–123.

Momose, T., Tonegawa, A., Takeuchi, J., Ogawa, H., Umesono, K.,

Yasuda, K., 1999. Efficient targeting of gene expression in chick

embryos by microelectroporation. Dev. Growth Differ. 41, 335–344.

Motoyama, J., Liu, J., Mo, R., Ding, Q., Post, M., Hui, C.C., 1998. Essential

function of Gli2 and Gli3 in the formation of lung, trachea and

oesophagus. Nat. Genet. 20, 54–57.

Nakagawa, Y., O’Leary, D.D., 2001. Combinatorial expression patterns of

LIM-homeodomain and other regulatory genes parcellate developing

thalamus. J. Neurosci. 21, 2711–2725.

Nybakken, K., Perrimon, N., 2002. Hedgehog signal transduction: recent

findings. Curr. Opin. Genet. Dev. 12, 503–511.

Orenic, T.V., Slusarski, D.C., Kroll, K.L., Holmgren, R.A., 1990. Cloning

and characterization of the segment polarity gene cubitus interruptus

Dominant of Drosophila. Genes Dev. 4, 1053–1067.

Park, H.L., Bai, C., Platt, K.A., Matıse, M.P., Beeghly, A., Hui, C.C.,

Nakashima, M., Joyner, A.L., 2000. Mouse Gli1 mutants are viable but

have defects in SHH signaling in combination with a Gli2 mutation.

Development 127, 1593–1605.

Persson, M., Stamataki, D., te Welscher, P., Andersson, E., Bose, J., Ruther,

U., Ericson, J., Briscoe, J., 2002. Dorsal-ventral patterning of the spinal

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–11111110

cord requires Gli3 transcriptional repressor activity. Genes Dev. 16,

2865–2878.

Puelles, L., Rubenstein, J.L., 1993. Expression patterns of homeobox and

other putative regulatory genes in the embryonic mouse forebrain

suggest a neuromeric organization. Trends Neurosci. 16, 472–479.

Puelles, L., 2001. Thoughts on the development, structure and evolution of

the mammalian and avian telencephalic pallium. Philos. Trans. R. Soc.

Lond. B Biol. Sci. 356, 1583–1598.

Rallu, M., Machold, R., Gaiano, N., Corbin, J.G., McMahon, A.P., Fishell,

G., 2002. Dorsoventral patterning is established in the telencephalon of

mutants lacking both Gli3 and Hedgehog signaling. Development 129,

4963–4974.

Redies, C., Ast, M., Nakagawa, S., Takeichi, M., Martınez-de-la-Torre, M.,

Puelles, L., 2000. Morphologic fate of diencephalic prosomeres and

their subdivisions revealed by mapping cadherin expression. J. Comp.

Neurol. 421, 481–514.

Roelink, H., Porter, J.A., Chiang, C., Tanabe, Y., Chang, D.T., Beachy,

P.A., Jessell, T.M., 1995. Floor plate and motor neuron induction by

different concentrations of the amino-terminal cleavage product of

Sonic hedgehog autoproteolysis. Cell 81, 445.

Rubenstein, J.L., Shimamura, K., Martinez, S., Puelles, L., 1998.

Regionalization of the prosencephalic neural plate. Annu. Rev.

Neurosci. 21, 445–477.

Ruiz i Altaba, A., 1998. Combinatorial Gli gene function in floor plate and

neuronal inductions by Sonic hedgehog. Development 125,

2203–2212.

Ruiz i Altaba, A., 1999. Gli proteins encode context-dependent positive and

negative functions: implications for development and disease. Devel-

opment 126, 3205–3216.

Ruiz i Altaba, A., Palma, V., Dahmane, N., 2002. Hedgehog-Gli signalling

and the growth of the brain. Nat. Rev. Neurosci. 3, 24–33.

Sasaki, H., Hui, C., Nakafuku, M., Kondoh, H., 1997. A binding site for Gli

proteins is essential for HNF-3beta floor plate enhancer activity in

transgenics and can respond to Shh in vitro. Development 124,

1313–1322.

Sasaki, H., Nishizaki, Y., Hui, C., Nakafuku, M., Kondoh, H., 1999.

Regulation of Gli2 and Gli3 activities by an amino-terminal repression

domain: implication of Gli2 and Gli3 as primary mediators of Shh

signaling. Development 126, 3915–3924.

Shigemoto, R., Nakanishi, S., Mizuno, N., 1992. Distribution of the mRNA

for a metabotropic glutamate receptor (mGluR1) in the central nervous

system: an in situ hybridization study in adult and developing rat.

J. Comp. Neurol. 322, 121–135.

Shimamura, K., Takeichi, M., 1992. Local and transient expression of E-

cadherin involved in mouse embryonic brain morphogenesis. Devel-

opment 116, 1011–1019.

Shimamura, K., Hartigan, D.J., Martinez, S., Puelles, L., Rubenstein, J.L.,

1995. Longitudinal organization of the anterior neural plate and neural

tube. Development 121, 3923–3933.

Teillet, M., Watanabe, Y., Jeffs, P., Duprez, D., Lapointe, F., Le Douarin,

N.M., 1998. Sonic hedgehog is required for survival of both myogenic

and chondrogenic somitic lineages. Development 125, 2019–2030.

Tokui, M., Takei, I., Tashiro, F., Shimada, A., Kasuga, A., Ishii, M., Ishii,

T., Takatsu, K., Saruta, T., Miyazaki, J., 1997. Intramuscular injection

of expression plasmid DNA is an effective means of long-term systemic

delivery of interleukin-5. Biochem. Biophys. Res. Commun. 233,

527–531.

Uchikawa, M., Kamachi, Y., Kondoh, H., 1999. Two distinct subgroups of

Group B Sox genes for transcriptional activators and repressors: their

expression during embryonic organogenesis of the chicken. Mech. Dev.

84, 103–120.

van den Heuvel, M., Ingham, P.W., 1996. smoothened encodes a receptor-

like serpentine protein required for hedgehog signalling. Nature 382,

547–551.

Von Mering, C., Basler, K., 1999. Distinct and regulated activities of

human Gli proteins in Drosophila. Curr. Biol. 9, 1319–1322.

Von Ohlen, T., Hooper, J.E., 1997. Hedgehog signaling regulates

transcription through Gli/Ci binding sites in the wingless enhancer.

Mech. Dev. 68, 149–156.

Vortkamp, A., Gessler, M., Grzeschik, K.H., 1995. Identification of

optimized target sequences for the GLI3 zinc finger protein. DNA Cell.

Biol. 14, 629–634.

Wang, Q.T., Holmgren, R.A., 2000. Nuclear import of cubitus interruptus is

regulated by hedgehog via a mechanism distinct from Ci stabilization

and Ci activation. Development 127, 3131–3139.

Watanabe, Y., Nakamura, H., 2000. Control of chick tectum territory along

dorsoventral axis by Sonic hedgehog. Development 127, 1131–1140.

Wijgerde, M., McMahon, J.A., Rule, M., McMahon, A.P., 2002. A direct

requirement for Hedgehog signaling for normal specification of all

ventral progenitor domains in the presumptive mammalian spinal cord.

Genes Dev. 16, 2849–2864.

Wijsman, J.H., Jonker, R.R., Keijzer, R., van de Velde, C.J., Cornelisse,

C.J., van Dierendonck, J.H., 1993. A new method to detect apoptosis in

paraffin sections: in situ end-labeling of fragmented DNA. J. Histochem.

Cytochem. 41, 7–12.

Wilson, S.W., Rubenstein, J.L., 2000. Induction and dorsoventral

patterning of the telencephalon. Neuron 28, 641–651.

Xie, J., Murone, M., Luoh, S.M., Ryan, A., Gu, Q., Zhang, C., Bonifas,

J.M., Lam, C.W., Hynes, M., Goddard, A., et al., 1998. Activating

Smoothened mutations in sporadic basal-cell carcinoma. Nature 391,

90–92.

Yoon, M.S., Puelles, L., Redies, C., 2000. Formation of cadherin-

expressing brain nuclei in diencephalic alar plate divisions. J. Comp.

Neurol. 427, 461–480.

Zeltser, L.M., Larsen, C.W., Lumsden, A., 2001. A new developmental

compartment in the forebrain regulated by Lunatic fringe. Nat.

Neurosci. 4, 683–684.

K. Hashimoto-Torii et al. / Mechanisms of Development 120 (2003) 1097–1111 1111