Midline Signaling and Evolution of the Forebrain in Chordates: A

SYMPOSIUM

Midline Signaling and Evolution of the Forebrain in Chordates:A Focus on the Lamprey Hedgehog CaseSylvie Retaux1 and Shungo Kano

NeD-UPR3294, CNRS, Institut Alfred Fessard, avenue de la Terrasse, 91198 Gif-sur-Yvette, France

From the symposium ‘‘Insights of Early Chordate Genomics: Endocrinology and Development in Amphioxus, Tunicates

and Lampreys’’ presented at the annual meeting of the Society for Integrative and Comparative Biology, January 3–7,

2010, at Seattle, Washington.

1E-mail: [email protected]

Synopsis Lampreys are agnathans (vertebrates without jaws). They occupy a key phylogenetic position in the emergence

of novelties and in the diversification of morphology at the dawn of vertebrates. We have used lampreys to investigate the

possibility that embryonic midline signaling systems have been a driving force for the evolution of the forebrain in

vertebrates. We have focused on Sonic Hedgehog/Hedgehog (Shh/Hh) signaling. In this article, we first review and sum-

marize our recent work on the comparative analysis of embryonic expression patterns for Shh/Hh, together with Fgf8

(fibroblast growth factor 8) and Wnt (wingless-Int) pathway components, in the embryonic lamprey forebrain.

Comparison with nonvertebrate chordates on one hand, and jawed vertebrates on the other hand, shows that these

morphogens/growth factors acquired new expression domains in the most rostral part of the neural tube in lampreys

compared to nonvertebrate chordates, and in jawed vertebrates compared to lampreys. These data are consistent with the

idea that changes in Shh, Fgf8 or Wnt signaling in the course of evolution have been instrumental for the emergence and

diversification of the telencephalon, a part of the forebrain that is unique to vertebrates. We have then used comparative

genomics on Shh/Hh loci to identify commonalities and differences in noncoding regulatory sequences across species and

phyla. Conserved noncoding elements (CNEs) can be detected in lamprey Hh introns, even though they display unique

structural features and need adjustments of parameters used for in silico alignments to be detected, because of

lamprey-specific properties of the genome. The data also show conservation of a ventral midline enhancer located in

Shh/Hh intron 2 of all chordates, the very species which possess a notochord and a floor plate, but not in earlier emerged

deuterostomes or protostomes. These findings exemplify how the Shh/Hh locus is one of the best loci to study genome

evolution with regards to developmental events.

The telencephalon is a vertebratenovelty

The vertebrate forebrain is an amazingly sophisticat-

ed part of the nervous system. It is composed of the

telencephalon (including the pallium or cortex and

the basal ganglia) and the diencephalon (including

the thalamus and hypothalamus). During embryo-

genesis, the forebrain is generated from the

anterior-most part of the neural plate and neural

tube through complex morphogenetic events

(Fig. 1). Meanwhile, waves of tightly controlled pro-

liferation, neurogenesis, specification, migration, and

axonal outgrowth generate thousands of types of

neurones that are cytoarchitectonically organized

and topographically interconnected (for reviews see

Wilson and Rubenstein, 2000; Guillemot, 2005).

Among vertebrates, the forebrain is also the part of

the brain that has undergone the most dramatic

morphological and anatomical diversification, this

being particularly true when considering the most

rostral and dorsal part, the telencephalon. Would

anyone have thought, for example, that the mamma-

lian cerebral cortex, with its well-known organization

into six layers, could be homologous with the everted

dorsal pallium of teleost fishes?

Ten to fifteen years of ‘‘evo-devo’’ studies using

comparative expression-pattern analysis for develop-

mental genes have shown that the forebrain, includ-

ing the telencephalon of various vertebrate species,

Integrative and Comparative Biology, volume 50, number 1, pp. 98–109

doi:10.1093/icb/icq032

Advanced Access publication May 11, 2010

� The Author 2010. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved.

For permissions please email: [email protected] from https://academic.oup.com/icb/article-abstract/50/1/98/733550by gueston 09 April 2018

Fig. 1 Development and evolution of signaling systems in the chordate forebrain. (A) Summary of the development of the forebrain in

a generic jawed vertebrate. Left: at neural plate stage, a fate map of the anterior neural plate shows a region destined to become the

telencephalon (tel, pink and green domains) and diencephalon (di, blue domains). The anterior neural ridge (ANR) and notochord (red)

are indicated. Middle: after closure of the neural tube, the forebrain develops under the influence of signaling molecules secreted from

the ventral side (Shh, red), the dorsal midline (Wnts and Bmps, brown) and the rostral pole (Fgf, blue). They generate a field of

organization in the adjacent neuroepithelium and control growth and morphogenesis. Right: regionalization of the forebrain as a

consequence of the concerted action of signaling centers. The telencephalon (tel) is further subdivided into the subpallial regions

(pink shades) and pallial areas (green shades) and the diencephalon (di) is subdivided into several tranverse domains including the

thalamus (thal) and the hypothalamus (hyp), mes, mesencephalon; is, isthmus; cb, cerebellum; met, metencephalon. (B) Evolution of the

expression of Hedgehog, Fgf8 and SFPRs in chordates. A simplified phylogenetic tree of chordates shows drawings of embryonic

expression of Shh/Hh (red), Fgf8 (blue), and SFRP1/5 (green) in representative species. The drawing of amphioxus is compiled from data

from Lin et al. (2009), Onai et al. (2009), Shimeld (1999), and Shimeld et al. (2007). The drawing of Ciona is compiled from data from

the aniseed database (Imai et al. 2002; Takatori et al. 2002). The drawing of the lamprey is from Guerin et al. (2009) and Osorio et al.

(2005). The typical gnathostome is represented by a zebrafish brain and is compiled from the ZFIN database. fp, floorplate; hb,

hindbrain; hypoth, hypothalamus; mhb, mid-hindbrain boundary; no, notochord; p, pineal gland; pcp, prechordal plate; tel, telencephalon;

zli, zona limitans intrathalamica.

Hedgehog and chordate brain evolution 99

Downloaded from https://academic.oup.com/icb/article-abstract/50/1/98/733550by gueston 09 April 2018

are specified and built according to shared genetic

mechanisms (e.g., Puelles et al. 2000; Bachy et al.

2002). In contrast, the sister group of vertebrates,

the urochordates, represented by ascidians (Delsuc

et al. 2006), together with the earlier-emerged

group of cephalochordates, represented by amphiox-

us, do not possess a telencephalon. The anterior tip

of their central nervous system is composed of a

sensory vesicle (ascidians) or a cerebral vesicle

(amphioxus). When comparing the patterning of

the anterior neural tube among the three groups

of chordates, one observes similarities that are

likely inherited from their common ancestor. For

example Otx expression is a hallmark of the anterior

neural tube in all chordates (Wada et al. 1998). In

the ascidian, Ciona intestinalis, detailed patterning

analysis has led to the proposal that the anterior

ventral neural tube corresponds to the vertebrate

presumptive hypothalamus (Moret et al. 2005), i.e.,

a diencephalic, but not telencephalic, region of the

brain.

The telencephalon appears to be a vertebrate

novelty (or synapomorphy). In the course of our

studies, we have used the evolutionary developmental

approach to examine the possible genetic mecha-

nisms that may lie at the origin of the vertebrate

forebrain. To this end, we have used lampreys as

model animals, because they occupy a very crucial

phylogenetic position in the chordate tree, i.e., they

belong to agnathans or cyclostomes, the sister group

of jawed vertebrates (Osorio and Retaux 2008).

Any character that is shared by lampreys and other

vertebrates but not by urochordates or cephalochor-

dates is therefore potentially relevant for the descrip-

tion of the common ancestor of all vertebrates.

Hence, we may be able to at least have a better pic-

ture of what has happened �500-million years ago

(mya), when this common ancestor experienced the

emergence and swelling of a part of the nervous

system at the tip of its dorsal (alar) neural tube: a

telencephalon.

Midline signaling as a driver offorebrain evolution

During embryogenesis, the brain develops under the

influence of several signaling centers, sometimes

called secondary organizers (Fig. 1A). These centers

secrete diffusible molecules with morphogen proper-

ties, which control growth and patterning and gen-

erate a field of organization in the adjacent

neuroepithelium [recently reviewed by Vieira et al.

(2010)]. The morphogen gradients are later translat-

ed into discrete neuroepithelial domains or divisions

that correspond to the future functional units of the

mature brain. Therefore, any change in the time of

appearance, in the strength, or in the exact location

of such secondary organizers has the potential of

significantly influencing the size or shape or pattern-

ing, in other words the neuroanatomy, of the brain

region they ‘‘organize’’.

A number of the signaling centers mentioned

above are located at a midline position respective

to the neural tube. Ventrally, the notochord and pre-

chordal plate (of mesodermal origin) and the in-

duced floor plate of the neural tube secrete Sonic

Hedgehog (Shh) (Echelard et al. 1993). At

mid-embryogenesis, new areas of Shh expression

appear in the forebrain, always emanating from the

ventral midline: the zona limitans intrathalamica (zli)

and the ventral telencephalon (Kiecker and Lumsden

2004; Scholpp et al. 2006; Vieira and Martinez 2006).

Dorsally, the roof plate of the neural tube secretes

molecules of the wingless-int (Wnt) (Lee and Jessell

1999; Muroyama et al. 2002) and bone morphoge-

netic protein (Bmp) families (Liem et al. 1997). Wnt

ligands are also secreted from the mid-diencephalon.

Finally, at the rostral tip of the neural tube, the an-

terior neural ridge (ANR) and later the rostral telen-

cephalon produce fibroblast growth factor 8 (Fgf8,

also expressed at the mid-hindbrain organizer)

(Shimamura and Rubenstein 1997; Rubenstein et al.

1998).

In this article, we first review and summarize our

recent work on the comparative analysis of embry-

onic expression patterns for Shh/Hh, together with

Fgf8 and Wnt pathway components, in the embryon-

ic lamprey forebrain. We then use comparative ge-

nomics on Shh/Hh loci to investigate whether

changes in noncoding regulatory sequences can ac-

count for the observed changes in Shh/Hh expression

across species. Our findings exemplify how the

Shh/Hh locus is an ideal locus to study the evolution

of the genome with regards to developmental events.

Hedgehog, Fgf8, and Wnt pathwayexpression patterns in lampreys

We have carried out a medium-scale in situ hybrid-

ization screen for genes expressed in the forebrain of

lamprey embryos (Guerin et al. 2009). Genes in-

volved in proliferation, stemcellness, neurogenesis,

and regional patterning showed globally highly sim-

ilar expression when compared to their gnathostome

counterparts, suggesting that the basic mechanisms

governing these events are shared among all verte-

brates. Yet strikingly, we observed major differences

100 S. Retaux and S. Kano


when considering the midline signaling systems

(summarized in Fig. 1B):

(1) The two lamprey Hedgehogs (Hh) (Osorio et al.

2005; J.-H. Xiao and Sylvie Retaux, unpublished

data) were expressed in the notochord and floor

plate, in the zli (for only one of the lamprey’s

Hhs), and in the hypothalamus. None of them

was expressed, even at a late amnocoete larval

stage, in the ventral telencephalon.

(2) Lamprey Fgf8 was classically expressed at the

mid-hindbrain boundary together with some ad-

ditional expression areas in the diencephalon,

also present in jawed vertebrates. Yet it was

not expressed at the telencephalic tip before a

late larval stage.

(3) Lamprey Wnt pathway components such as Wnt

ligands, their frizzled receptors, and their SFRP

inhibitors were also studied. Major differences

were observed for the SFRPs: lamprey SFRP1/5

expression was restricted to the hypothalamus,

whereas gnathostome orthologs expression also

expands into the telencephalon.

The differences in expression of the Hh, Fgf8 and

SFRP signaling pathways components in lampreys

and jawed vertebrates summarized above concern

mainly the telencephalon, whereas other expression

domains (diencephalic or mesencephalic) are con-

served (Guerin et al. 2009) (Fig. 1B). These differ-

ences are interesting to discuss in regard to the

peculiar anatomy of the lamprey telencephalon and

to the evolution of the telencephalon in vertebrates.

The lack of Hh signal in the telencephalon, togeth-

er with the absence of expression of Nkx2.1 in the

same area (Murakami et al. 2001; Osorio et al. 2005),

correlates with the absence of a pallidal division in

the ventral telencephalon of lampreys (Weigle and

Northcutt 1999). In a sense, the mouse Nkx2.1�/�

(knock-out) phenotype mimics the lampreys’ condi-

tion, with a ventral telencephalon that is almost ex-

clusively composed of a striatal division (Sussel et al.

1999).

There is heterochrony of Fgf8 expression in lam-

preys and it is tempting to relate the lack of Fgf8

signal at the rostral tip of the telencephalon during

the first 10 days of development to the extremely

slow growth of the telencephalon in embryonic lam-

preys. Indeed, whereas the gnathostome telencephalic

vesicles undergo considerable growth and swelling

during the early phases of embryogenesis, lamprey

telencephalon remains very tiny and starts growing

significantly only in the larval stages (Villar-Cheda

et al. 2006). It is probable that other Fgfs, such as

Fgf3, which has a complementary role to Fgf8 in fish

(Walshe and Mason 2003), are expressed in the lam-

prey telencephalon. Moreover, as important and re-

ciprocal interactions occur between the ventral (Shh)

and rostral (Fgf8) signaling centers and shape the

gnathostome telencephalon (e.g., Kuschel et al.

2003), it is probable that the absence of the former

and the late appearance of the latter influence the

patterning and morphogenesis of this region in lam-

preys considerably. These differences are also intrigu-

ing and raise questions on how the reciprocal

transcriptional regulations these pathways exert on

each other are controlled in lampreys.

Inhibition of the Wnt pathway by SFRP1/5 does

not occur in the lamprey telencephalon. This is also

surprising because in jawed vertebrates SFRP1 and

SFRP5 (and the fish-specific Tlc) (Tendeng and

Houart 2006) are in major part responsible for the

inhibition of Wnt signals emanating from the dien-

cephalon, and serve in determining the fate of the

anterior telencephalon (Houart et al. 2002). As lam-

preys do have a telencephalon, albeit very small, it is

highly probable that other factors involved in inhi-

bition of the Wnt signal are present in the embryonic

lamprey’s forebrain. They remain to be discovered,

and, once again, will likely highlight the distinctive

manner in which lampreys regulate and control the

specification and growth of their anterior neural

tube.

The hypotheses driven by the comparative ap-

proach await functional validation. Within this

framework, and in an attempt to mimic the gnathos-

tome situation in the lamprey telencephalon, we have

injected mammalian Shh protein into the basal

telencephalon of stage-24 lamprey embryos.

Unfortunately, this did not result in the ectopic in-

duction of Nkx2.1 in the subpallium (J. Osorio and

S.R., unpublished data), nor in any particular phe-

notype. More experiments interfering with the mid-

line signaling pathways in lampreys are needed to

understand the interactions between these pathways

and the effects they produce.

The data reported above would also strongly ben-

efit from a similar analysis in hagfish, the only other

extant agnathan/cyclostome representative. The diffi-

culty in obtaining hagfish embryos has long ham-

pered the study of embryology in this species, but

the recent success reported in breeding these animals

has generated hope for future years (Ota et al. 2007).

It will be crucial to know whether hagfish embryos

‘‘behave’’ like lamprey embryos in terms of telence-

phalic transcriptome and associated gene regulatory

networks. For example, if hagfish do not express Shh

in their ventral telencephalon, it will further reinforce



the idea that Hh/Shh signaling has been recruited to

the telencephalon in the ancestor of jawed verte-

brates. Alternatively, if hagfish embryos do express

Shh in their telencephalon, it will substantiate a

loss of Hh expression during lamprey evolution,

and will suggest that the recruitment of Hh in the

basal telencephalon happened in the common ances-

tor of all vertebrates.

Comparing midline signaling in basal chordates

A larger and general view can be obtained by further

comparing conditions between vertebrate and non-

vertebrate chordates. A search of the literature and ex-

pression databases (http://crfb.univ-mrs.fr/aniseed/)

for ascidians and amphioxus for reported midline

components resulted in the schematic summary in

Fig. 1B. In these two nonvertebrate chordate repre-

sentatives, Hh expression is certainly ventral (in the

ventral nerve cord and notochord in amphioxus

only) but it is restricted to the posterior nervous

system, in regions that are comparable to the hind-

brain/spinal cord patterning domains of vertebrates

(Shimeld 1999; Takatori et al. 2002). In addition, in

Ciona, Hh1-expressing cells are found at the tip of

the sensory vesicle. Later on, Hh2 is partially

co-localized with Nkx2.1 (Ristoratore et al. 1999;

Islam et al. 2010). In amphioxus the expression of

Fgf8 spreads throughout the anterior ectoderm, like

Otx (Meulemans and Bronner-Fraser 2007; Holland

2009). In Ciona, two or three Fgf8-expressing cells

were reported in a position corresponding to the

junction between the sensory vesicle and the posteri-

or nerve cord (Ikuta and Saiga 2007), thus

pre-figuring a possible anterior–posterior brain

boundary, which is exemplified by the vertebrate

mid-hindbrain boundary (Imai et al. 2009). Finally,

SFRP1/5 in Ciona is not expressed in the anterior

sensory vesicle, while amphioxus SFRP1/5 has not

been identified to date.

When comparing these organizers across chor-

dates, it appears that they become localized more

and more anteriorly in the neural tube as the subdi-

visions of the forebrain become larger and more nu-

merous, and as the telencephalon emerges and

expands (Fig. 1B). This is particularly true for

Shh/Hh and Fgf8 which have complex expression

patterns composed of multiple expression areas.

This tendency correlates well with the two main

roles of these signaling centers and the molecules

they produce: they organize and control the growth

of the neuroepithelium. As such, they are ideally

placed to function as drivers of brain evolution,

both for the emergence of novelties and for the di-

versification of structures.

Transcriptional regulation of Shh/Hhmidline signaling

As a next step in the understanding of the molecular

mechanisms of the evolution of the forebrain, we

studied transcriptional regulation of the genes

coding for morphogen factors. Indeed, subtle differ-

ences in expression domains or in the timing of tran-

script expression mostly originate in noncoding

regulatory sequences. As a case study, we choose

the Shh/Hh gene/locus.

The regulatory logics of Shh transcription has

been the focus of many studies in gnathostomes.

Using phylogenetic footprinting, it is possible to

align and compare Shh loci, and to detect stretches

of highly conserved sequences between species

(Goode et al. 2003, 2005; Jeong et al. 2006; Ertzer

et al. 2007; see also Strahle and Rastegar, 2008 for

review). Highly conserved sequences often corre-

spond to exon-coding fragments, but also reside in

noncoding intronic sequences or in 50- or 30-gene

regions (Fig. 2A). The fact that these noncoding se-

quences are highly conserved—sometimes even

better than coding sequences themselves—suggests

that they are under strong selective pressure, and

that they contain crucial regulatory elements, includ-

ing binding sites for transcription factor (Vavouri

and Lehner, 2009).

In the case of gnathostome Shh genes, the align-

ment and VISTA visualization of conserved noncod-

ing elements (CNEs) can be easily performed

(Fig. 2A, lines 1–4; see also Appendix 1). A

number of CNEs have been previously identified by

this method, and have been functionally tested for

enhancer activity in the mouse or the zebrafish

(Jeong et al. 2006; Ertzer et al. 2007). Hence, the

regulation of Shh expression appears highly modular,

with specific enhancers governing expression in the

notochord, the floor plate, the zli, the hypothalamus,

and the telencephalon. Of note, this feature of the

Shh locus is highly favorable in the context of our

hypothesis of Hh signaling as a driver of brain evo-

lution: the modular nature of cis-regulatory elements

and the pleiotropy of gene products are among the

‘‘evolutionary toolkit’’ that allows for selective

spatio-temporal changes of expression patterns, and

thus of morphological changes (Carroll, 2008).

Interestingly, recent studies have shown that this

modular nature of enhancers also applies to the

Fgf8 locus (Komisarczuk et al. 2009).



Fig. 2 Comparison of Shh/Hh loci across vertebrates. (A) VISTA alignment and search for CNEs in several vertebrates (mouse, chick,

zebrafish, medaka, and lamprey) using the Fugu Shh locus as baseline. The zebrafish intronic elements previously described (ar-A and

ar-B in intron 1 and ar-C in intron 2 (Muller et al. 1999; Ertzer et al. 2007) were added to identify the conserved peaks. In this and the

following plots, gray shading indicates more than 75% sequence identity. The exons (E1–E3) appear as rectangles (gray peaks on the

plots), and non-coding conserved sequences appear as ovals (black peaks on the plots). ar-X stands for ‘activating region’ X. (B) VISTA

alignment and search for CNEs in zebrafish and lamprey. When using the Fugu Shh locus as baseline, the ar-C element in intron 2

appears as a single peak of conservation (black arrowhead). (C) VISTA alignment and search for CNEs in zebrafish and Fugu, using the

lamprey Hhb locus as baseline. The ar-C element in intron 2 appears dispersed into four peaks (black arrows). (D) Local alignment

showing lamprey, zebrafish and Fugu ar-C nucleotide sequences, illustrating the dispersed nature of lamprey ar-C. The four

sub-elements ar-C1 to ar-C4 identified in zebrafish by Hadzhiev et al. (2007) are dispersed along intron 2 in a manner that corresponds

to the four conservation peaks detected in C (first peak: beginning of ar-C; second peak: C1; third peak: C2; fourth peak: end of

C2 þ C3 þ C4). Note that ar-C4 in Fugu also contains an insertion. The full alignment can be seen in Supplementary Fig. 1.



Hh regulatory logics in lampreys

Through in silico searches in the sea lamprey

Petromyzon marinus (Pm) preliminary genome assem-

bly (http://pre.ensembl.org/Petromyzon_marinus/

Info/Index), and through cosmid library screening

and sequencing in the river lamprey Lampetra fluvia-

tilis (Lf), we have obtained genomic sequences for

two distinct Hh genes in the two lamprey species.

We have called them Hha and Hhb (S. Kano et al.

unpublished data). They are both Shh-like in terms of

molecular phylogeny and expression patterns, and

therefore allowed us to perform a comparative geno-

mic analysis with their gnathostome orthologous

Shh locus. We have recently analyzed and reported

CNEs in lamprey Hhs by comparisons based on the

coelacanth Shh locus (S. Kano et al. unpublished

data). Here we present the same approach for

Lampetra Hhb (LfHhb), and use the Fugu genome

as baseline for comparison, i.e., complete genome

that allows more systematic comparisons and that

has proven very helpful in the past (Goode et al.

2003, 2005).

Using the Fugu Shh locus as baseline, previously

identified intronic motifs ar-A and ar-C were easily

identified among gnathostomes (Fig. 2A). Other

motifs, ar-B in intron 1 and ar-D in the 50-upstream

region, were less conserved, although a few conserved

blocks can be detected among teleost species for the

ar-B motif, originally identified in zebrafish (Muller

et al. 1999; Ertzer et al. 2007). When compared with

gnathostome loci, almost no significant CNEs are

found at first glance in the Lampetra fluviatilis

LfHhb locus, besides the three exons (Fig. 2A, fifth

line). However, the use of a more limited number of

species’ loci for comparison allowed the detection of

conserved blocks and significant CNEs in lamprey

(Fig. 3B). More specifically, these include ar-C [the

‘‘midline’’ enhancer described in mice and zebrafish

(Jeong et al. 2006; Ertzer et al. 2007; Hadzhiev et al.

2007)] and ar-D (an additional floorplate enhancer

also active both in mice and zebrafish) (Jeong et al.

2006; Ertzer et al. 2007).

When mapped onto the Fugu Shh locus, the

ar-C element emerges as a single conservation peak

Fig. 3 comparison of Shh/Hh loci among deuterostomes. (A) Each Shh/Hh locus was independently compared with Fugu and zebrafish

because global alignments performed together disturb proper alignment, as mentioned above (data not shown). The presentation is the

same as in Fig. 2. (B) Local alignment of the ar-C motif shared among chordates. The conserved sequence shared among chordates

corresponds to the ar-C1 sub-element. The position of a consensus FoxA2 binding site is indicated.



(black arrowhead in Fig. 2B). However, when the

reverse in silico experiment is done, i.e., mapping

the Fugu or the zebrafish loci onto the lamprey

LfHhb used as baseline, ar-C emerges as four indi-

vidual peaks in intron 2 (arrows in Fig. 2C). This

demonstrates that ar-C is split into several conserved

subelements or blocks in the lamprey locus. These

conserved blocks can be readily observed on local

alignments (Fig. 2D and Supplementary Fig. 1). In

particular, a �1 kb-long insertion occurs between

ar-C1 and the rest of ar-C. The lamprey C1

sub-element was further validated as a midline en-

hancer responsible for notochord and floor-plate ex-

pression after injections in zebrafish embryos

(S. Kano et al. unpublished data). These findings

show and confirm that functional CNEs are found

in the lamprey genome, as first reported for Hox

genes (Carr et al. 1998; Irvine et al. 2002) and

more recently in a survey of 13 gene loci (McEwen

et al. 2009).

These data also raise two issues. (1) The first is an

alignment issue when comparing the diverged lam-

prey genome with other vertebrate genomes. When

carefully checking local alignments, the lamprey

genome appears to contain lamprey-specific inser-

tions in the introns. In fact, it was difficult to tune

alignments when both Hha and Hhb loci were in-

cluded together with vertebrates because the two

lamprey paralogs evolved independently, and the

paralog-specific insertions disturb alignments

(better results are also obtained on the Hha locus

when it is analyzed independently from Hhb; not

shown). This may well explain why fewer CNEs

were found in lamprey genes than in other verte-

brates (McEwen et al. 2009). (2) The second is a

window-size issue for MLAGAN and VISTA analysis.

We have used a 20-bp window size and a 70% iden-

tity cut-off in our analysis whereas McEwen et al.

(2009) have used a window size of 40 bp and a

cut-off of 65% identity. Using these parameters, we

were able to dissect out a �30 bp lamprey ar-C1.

Thus, a narrow window size is advisable when sur-

veying CNEs in the lamprey genome.

Survey of CNEs shared amongchordates and deuterostomes inShh/Hh loci

The lamprey CNEs are likely to be essential for Hh

expression. As CNEs frequently contain regulatory

elements (Vavouri et al. 2007; Vavouri and Lehner,

2009), and as Hh genes are expressed at least in part

in similar domains in vertebrate and nonvertebrate

chordates, we have pursued a similar comparative

genomics approach on other chordates and deutero-

stomes. Of note, some CNEs were recently identified

in amphioxus and showed enhancer activity in zeb-

rafish (Hufton et al. 2009).

We compared Hh loci between fugu and three

other gnathostomes (zebrafish and two amniotes

species: chick and mouse), a jawless vertebrate

(lamprey), two nonvertebrate chordates (ascidians

and amphioxus), a hemichordate (acorn worm), an

echinoderm (sea urchin), and a protostome (fly)

(Fig. 3A). Conservation peaks were detected even

though the degree of conservation was moderate in

most cases (see also Fig. 4). Interestingly, the amphi-

oxus Hh locus displayed highest conservation with

fish genomes (all the intronic peaks are colored in

pink). Putative amphioxus ar-A, ar-B, and ar-C

motifs showed significant conservation but were

shorter than vertebrate CNEs. It is reasonable to

think that the important conservation found between

amphioxus and vertebrates reflects the fact that am-

phioxus’ body plan and Hh expression pattern re-

semble those of vertebrates (see also Holland et al.

2008; Holland, 2009).

We found that the intron 2 ar-C motif is well

shared only among chordates (Fig. 3A). Further, it

was possible to identify a ‘‘core’’ ar-C1 motif, pro-

posed by Hadzhiev et al. (2007) to be the major

functional sub-element of ar-C, in all the chordate

Hh loci examined (Fig. 3B). This raises the

Fig. 4 Hh CNEs in bilaterians. A simplified tree of bilaterians

highlights the bilaterian origin of Hh CNEs (dotted line), and the

probable emergence and fixation of a midline ar-C motif in

chordates. The shaded grey points to the interesting phylogenetic

position of lampreys which possess a telencephalon and whose

genome has undergone whole-genome duplications (WGD;

Kuraku et al., 2008).



interesting possibility that ar-C is a key motif for the

chordate lineage (see also Fig. 4). Indeed, ar-C is

notably responsible for notochord and floorplate ex-

pression in these animals which all possess a noto-

chord and a floor plate as a defining feature, e.g., in

mice (Jeong et al. 2006), in zebrafish (Hadzhiev et al.

2007), and in lamprey (S. Kano et al. unpublished

data). In addition, it is worth mentioning that the

ar-C1 element contains a consensus FoxA2 transcrip-

tion factor binding site (Fig. 3B). This is interesting

in regards of the described implication of this factor,

expressed in the floor plate, in the regulation of Shh

expression at the ventral midline (for review, Placzek

and Briscoe, 2005).

Surprisingly, our analysis also revealed conserved

peaks in the introns of Saccoglossus, Strongylocentrus

and Drosophila, although Hh expression in these an-

imals is not comparable to chordates. The evolution

of Hh genes has been the focus of much interest, and

their origins have been traced to the first bilaterians

(Adamska et al. 2007). Our data on ‘‘signatures’’ re-

maining on every Shh/Hh intron suggest that the Hh

introns have been maintained since bilaterians ap-

peared, and further highlight this evolutionary sce-

nario. They are also in line with the idea that intron

gain is an extremely rare event in vertebrate evolu-

tion (Loh et al. 2008), so that most introns are of

ancestral origin (Fig. 4).

The primary function of ar-C is up-regulation at

the midline. This element was apparently recruited

into the midline of the tail (i.e., notochord and

floorplate) in early chordates. Additional regulatory

elements must have further emerged to drive Shh/Hh

expression in new domains—such as the forebrain

Shh expression domains observed in jawed verte-

brates. Of note, repressor elements also appeared,

such as zebrafish ar-C2 and ar-C4 which act as

floor-plate repressors (Hadzhiev et al. 2007). This

hypothesis needs further investigation on CNEs, sim-

ilar to those presented here.

Conclusions

Shh/Hh is known as one of the most powerful mor-

phogens during embryonic development. As it is par-

ticularly involved in controlling morphogenesis,

patterning and growth of the neural tube, we have

hypothesized that variations of its expression pattern

during embryogenesis may be an evolutionary driv-

ing force, causing changes in size and/or morpholo-

gy. Using lamprey as a key phylogenetically-placed

model animal, we have examined this possibility at

two levels: comparative expression patterns and com-

parative genomics.

A recent survey of 13 lamprey genes for CNEs by

McEwen et al. (2009) demonstrated that some CNEs

were indeed identified and validated; however, their

number was much less than expected. While a

large-scale analysis is now feasible and required in

the genome era, deep insight from some particular

cases are still essential. The Hedgehog locus is one of

best models to associate genomic evolutionary events

with developmental mechanisms, because it has been

deeply studied in several model animals, broadly

distributed phylogenetically. Here, we add a new

case study of lamprey and other deuterostome

Hedgehog loci.

Supplementary Data

Supplementary Data are available at ICB online.

Acknowledgments

We thank our colleagues, collaborators and friends

Jean-Stephane Joly, Sylvie Mazan, and Didier Casane,

and the other group members previously involved in

lamprey work Joana Osorio, Adele Guerin, and Jin-

Hua Xiao, for long-term and fertile scientific

interactions.

Funding

Work supported by a Groupement d’Interet

Scientifique (GIS) Genomique Marine and Agence

Nationale de la Recherche ANR-Neuro (MIDLINE)

(to S.R.). S.K. was supported by an ANR postdoc-

toral fellowship.

References

Adamska M, Matus DQ, Adamski M, Green K, Rokhsar DS,

Martindale MQ, Degnan BM. 2007. The evolutionary origin

of hedgehog proteins. Curr Biol 17:R836–7.

Bachy I, Berthon J, Retaux S. 2002. Defining pallial and sub-

pallial divisions in the developing Xenopus forebrain. Mech

Dev 117:163–72.

Brudno M, Do CB, Cooper GM, Kim MF, Davydov E,

Program NCS, Green ED, Sidow A, Batzoglou S. 2003.

LAGAN and Multi-LAGAN: Efficient tools for large-scale

multiple alignment of genomic DNA. Genome Res 13:

721–31.

Carr JL, Shashikant CS, Bailey WJ, Ruddle FH. 1998.

Molecular evolution of Hox gene regulation: Cloning and

transgenic analysis of the lamprey HoxQ8 gene. J Exp Zool

280:73–85.

Carroll SB. 2008. Evo-devo and an expanding evolutionary

synthesis: A genetic theory of morphological evolution.

Cell 134:25–36.

Delsuc F, Brinkmann H, Chourrout D, Philippe H. 2006.

Tunicates and not cephalochordates are the closest living

relatives of vertebrates. Nature 439:965–8.



Echelard Y, Epstein DJ, St-Jacques B, Shen L, Mohler J,

McMahon JA, McMahon AP. 1993. Sonic hedgehog, a

member of a family of putative signaling molecules, is

implicated in the regulation of CNS polarity. Cell 75:

1417–30.

Ertzer R, Muller F, Hadzhiev Y, Rathnam S, Fischer N,

Rastegar S, Strahle U. 2007. Cooperation of sonic hedgehog

enhancers in midline expression. Dev Biol 301:578–89.

Frazer KA, Pachter L, Poliakov A, Rubin EM, Dubchak I.

2004. VISTA: computational tools for comparative geno-

mics. Nucleic Acids Res 32:W273–9.

Goode DK, Snell P, Elgar G. 2003. Comparative analysis of

vertebrate Shh genes identifies novel conserved non-coding

sequence. Mamm Genome 14:192–201.

Goode DK, Snell P, Smith SF, Cooke JE, Elgar G. 2005.

Highly conserved regulatory elements around the SHH

gene may contribute to the maintenance of conserved syn-

teny across human chromosome 7q36.3. Genomics 86:

172–81.

Guerin A, d’Aubenton-Carafa Y, Marrakchi E, Da Silva C,

Wincker P, Mazan S, Retaux S. 2009. Neurodevelopment

genes in lampreys reveal trends for forebrain evolution in

craniates. PLoS One 4:e5374.

Guillemot F. 2005. Cellular and molecular control of neuro-

genesis in the mammalian telencephalon. Curr Opin Cell

Biol 17:639–47.

Hadzhiev Y, Lang M, Ertzer R, Meyer A, Strahle U, Muller F.

2007. Functional diversification of sonic hedgehog paralog

enhancers identified by phylogenomic reconstruction.

Genome Biol 8:R106.

Holland LZ. 2009. Chordate roots of the vertebrate nervous

system: Expanding the molecular toolkit. Nat Rev Neurosci

10:736–46.

Holland LZ, et al. 2008. The amphioxus genome illuminates

vertebrate origins and cephalochordate biology. Genome

Res 18:1100–11.

Houart C, Caneparo L, Heisenberg C, Barth K, Take-Uchi M,

Wilson S. 2002. Establishment of the telencephalon during

gastrulation by local antagonism of Wnt signaling. Neuron

35:255–65.

Hufton AL, Mathia S, Braun H, Georgi U, Lehrach H,

Vingron M, Poustka AJ, Panopoulou G. 2009. Deeply con-

served chordate noncoding sequences preserve genome syn-

teny but do not drive gene duplicate retention. Genome

Res 19:2036–51.

Ikuta T, Saiga H. 2007. Dynamic change in the expression of

developmental genes in the ascidian central nervous system:

revisit to the tripartite model and the origin of the mid-

brain-hindbrain boundary region. Dev Biol 312: 631–43.

Imai KS, Satoh N, Satou Y. 2002. Region specific gene expres-

sions in the central nervous system of the ascidian embryo.

Gene Expr Patterns 2:319–21.

Imai KS, Stolfi A, Levine M, Satou Y. 2009. Gene regulatory

networks underlying the compartmentalization of the

Ciona central nervous system. Development 136:285–93.

Irvine SQ, Carr JL, Bailey WJ, Kawasaki K, Shimizu N,

Amemiya CT, Ruddle FH. 2002. Genomic analysis of Hox

clusters in the sea lamprey Petromyzon marinus. J Exp

Zool 294:47–62.

Islam A, Moly P, Miyamot Y, Kusakabe T. 2010. Distinctive

expression patterns of Hedgehog pathway genes in the

ciona intestinalis larva: implications for a role of

Hedgehog signaling in postembryonic development and

chordate evolution. Zoolog Sci 27:84–90.

Jeong Y, El-Jaick K, Roessler E, Muenke M, Epstein DJ. 2006.

A functional screen for sonic hedgehog regulatory elements

across a 1 Mb interval identifies long-range ventral fore-

brain enhancers. Development 133:761–72.

Kiecker C, Lumsden A. 2004. Hedgehog signaling from the

ZLI regulates diencephalic regional identity. Nat Neurosci

7:1242–9.

Komisarczuk AZ, Kawakami K, Becker TS. 2009. Cis-regula-

tion and chromosomal rearrangement of the fgf8 locus after

the teleost/tetrapod split. Dev Biol 336:301–12.

Kuraku S, Meyer A, Kuratani S. 2008. Timing of genome

duplications relative to the origin of the vertebrates:

did cyclostomes diverge before, or after? Mol Biol Evol

26:47–59.

Kuschel S, Ruther U, Theil T. 2003. A disrupted balance

between Bmp/Wnt and Fgf signaling underlies the ventra-

lization of the Gli3 mutant telencephalon. Dev Biol 260:

484–95.

Lee KJ, Jessell TM. 1999. The specification of dorsal cell fates

in the vertebrate central nervous system. Annu Rev

Neurosci 22:261–94.

Liem KF Jr, Tremml G, Jessell TM. 1997. A role for the roof

plate and its resident TGFbeta-related proteins in neuronal

patterning in the dorsal spinal cord. Cell 91:127–38.

Lin Y, Cai Z, Huang S, Yang L, Wang C, Liu Z, Cao J, An Y,

Zhang H. 2009. Ptc, Smo, Sufu, and the Hedgehog signal-

ing pathway in amphioxus. Evol Dev 11:710–8.

Loh YH, Brenner S, Venkatesh B. 2008. Investigation

of loss and gain of introns in the compact genomes of

pufferfishes (Fugu and Tetraodon). Mol Biol Evol 25:

526–35.

Mayor C, Brudno M, Schwartz JR, Poliakov A, Rubin E M,

Frazer KA, Pachter LS, Dubchak I. 2000. VISTA:

Visualizing global DNA sequence alignments of arbitrary

length. Bioinformatics 16:1046–7.

McEwen GK, Goode DK, Parker HJ, Woolfe A, Callaway H,

Elgar G. 2009. Early evolution of conserved regulatory

sequences associated with development in vertebrates.

PLoS Genet 5:e1000762.

Meulemans D, Bronner-Fraser M. 2007. Insights from

amphioxus into the evolution of vertebrate cartilage.

PLoS One 2:e787.

Moret F, Christiaen L, Deyts C, Blin M, Vernier P, Joly JS.

2005. Regulatory gene expressions in the ascidian ventral

sensory vesicle: Evolutionary relationships with the verte-

brate hypothalamus. Dev Biol 277:567–79.

Muller F, Chang B, Albert S, Fischer N, Tora L, Strahle U.

1999. Intronic enhancers control expression of zebrafish

sonic hedgehog in floor plate and notochord.

Development 126:2103–16.



Murakami Y, Ogasawara M, Sugahara F, Hirano S, Satoh N,

Kuratani S. 2001. Identification and expression of the lam-

prey Pax6 gene: Evolutionary origin of the segmented brain

of vertebrates. Development 128:3521–31.

Muroyama Y, Fujihara M, Ikeya M, Kondoh H, Takada S.

2002. Wnt signaling plays an essential role in neuronal

specification of the dorsal spinal cord. Genes Dev 16:

548–53.

Onai T, Lin HC, Schubert M, Koop D, Osborne PW,

Alvarez S, Alvarez R, Holland ND, Holland LZ. 2009.

Retinoic acid and Wnt/beta-catenin have complementary

roles in anterior/posterior patterning embryos of the basal

chordate amphioxus. Dev Biol 332:223–33.

Osorio J, Mazan S, Retaux S. 2005. Organisation of the lam-

prey (Lampetra fluviatilis) embryonic brain: Insights from

LIM-homeodomain, Pax and hedgehog genes. Dev Biol

288:100–12.

Osorio J, Retaux S. 2008. The lamprey in evolutionary studies.

Dev Genes Evol 218:221–35.

Ota KG, Kuraku S, Kuratani S. 2007. Hagfish embryology

with reference to the evolution of the neural crest. Nature

446:672–5.

Placzek M, Briscoe J. 2005. The floor plate: multiple cells,

multiple signals. Nat Rev Neurosci 6:230–40.

Puelles L, Kuwana E, Puelles E, Bulfone A, Shimamura K,

Keleher J, Smiga S, Rubenstein JL. 2000. Pallial and sub-

pallial derivatives in the embryonic chick and mouse tele-

ncephalon, traced by the expression of the genes Dlx-2,

Emx-1, Nkx-2.1, Pax-6, and Tbr-1. J Comp Neurol 424:

409–38.

Ristoratore F, Spagnuolo A, Aniello F, Branno M, Fabbrini F,

Di Lauro R. 1999. Expression and functional analysis of

Cititf1, an ascidian NK-2 class gene, suggest its role in

endoderm development. Development 126:5149–59.

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

Regionalization of the prosencephalic neural plate. Annu

Rev Neurosci 21:445–77.

Scholpp S, Wolf O, Brand M, Lumsden A. 2006. Hedgehog

signalling from the zona limitans intrathalamica orches-

trates patterning of the zebrafish diencephalon.


Shimamura K, Rubenstein JL. 1997. Inductive interactions

direct early regionalization of the mouse forebrain.


Shimeld SM. 1999. The evolution of the hedgehog gene family

in chordates: insights from amphioxus hedgehog. Dev

Genes Evol 209:40–7.

Shimeld SM, van den Heuvel M, Dawber R, Briscoe J. 2007.

An amphioxus Gli gene reveals conservation of midline

patterning and the evolution of hedgehog signalling diver-

sity in chordates. PLoS One 2:e864.

Strahle U, Rastegar S. 2008. Conserved non-coding

sequences and transcriptional regulation. Brain Res Bull

75:225–30.

Sussel L, Marin O, Kimura S, Rubenstein JL. 1999. Loss of

Nkx2.1 homeobox gene function results in a ventral to

dorsal molecular respecification within the basal

telencephalon: evidence for a transformation of the palli-

dum into the striatum. Development 126:3359–70.

Takatori N, Satou Y, Satoh N. 2002. Expression of hedgehog

genes in Ciona intestinalis embryos. Mech Dev 116:235–8.

Tendeng C, Houart C. 2006. Cloning and embryonic expres-

sion of five distinct sfrp genes in the zebrafish Danio rerio.

Gene Expr Patterns 6:761–71.

Vavouri T, Lehner B. 2009. Conserved noncoding elements

and the evolution of animal body plans. Bioessays 31:

727–35.

Vavouri T, Walter K, Gilks W, Lehner B, Elgar G. 2007.

Parallel evolution of conserved non-coding elements that

target a common set of developmental regulatory genes

from worms to humans. Genome Biol 8:R15.

Vieira C, Martinez S. 2006. Sonic hedgehog from the basal

plate and the zona limitans intrathalamica exhibits differ-

ential activity on diencephalic molecular regionalization

and nuclear structure. Neuroscience 143:129–40.

Vieira C, Pombero A, Garcia-Lopez R, Gimeno L,

Echevarria D, Martinez S. 2010. Molecular mechanisms

controlling brain development: an overview of neuroepithe-

lial secondary organizers. Int J Dev Biol 54:7–20.

Villar-Cheda B, Perez-Costas E, Melendez-Ferro M,

Abalo XM, Rodriguez-Munoz R, Anadon R, Rodicio MC.

2006. Cell proliferation in the forebrain and midbrain of

the sea lamprey. J Comp Neurol 494:986–1006.

Wada H, Saiga H, Satoh N, Holland PW. 1998. Tripartite

organization of the ancestral chordate brain and the anti-

quity of placodes: insights from ascidian Pax-2/5/8, Hox

and Otx genes. Development 125:1113–22.

Walshe J, Mason I. 2003. Unique and combinatorial functions

of Fgf3 and Fgf8 during zebrafish forebrain development.


Weigle C, Northcutt RG. 1999. The chemoarchitecture of the

forebrain of lampreys: evolutionary implications by com-

parisons with gnathostomes. Eur J Morphol 37:122–5.

Wilson SW, Rubenstein JL. 2000. Induction and dorsoventral

patterning of the telencephalon. Neuron 28:641–51.

Appendix 1: Materials and methods

Genomic sequences for Shh/Hhloci comparison

MLAGAN (Brudno et al. 2003) and VISTA plots

(Mayor et al. 2000; Frazer et al. 2004) were applied

on genomic sequences with 2 kb of upstream and

3 kb of downstream sequences around Shh/hh loci

(except for the lamprey, 5 kb upstream). Most geno-

mic sequences for Shh/Hh loci were obtained from

the UCSC genome browser, except the acorn worm

(see below) and the lamprey sequence obtained by

ourselves [LfHhb accession number: FP929027]. Each

position of sequence and its assemble version is indi-

cated as follows: mouse (chr5: 28,780,380–28,795,641,



mm9), chicken (chr2: 8,021,868–8,036,924, galGal3),

zebrafish (chr7: 36,654,711–36,666,126, danRer5),

fugu (chrUn: 211,116,380–211,127,115, fr2), medaka

(chr20: 17,733,043–17,744,774, oryLat2), ascidian

(chr05q: 4,583,360–4,593,934, ci2), lancelet (chrUn:

399879050–399909584, braFlo1), sea urchin (scaf-

fold68424: 22314–56576, strPur2), fly (chr3R:

18950425–18970881, dm3). The genomic sequence,

scaffold44409, for Saccoglossus kowalevskii hedgehog

locus was obtained, using BLAST search with partial

cDNA sequence (DQ431035.1), from BCM-HGSC

web site, which is generated by Acorn Worm

Genome Sequencing Consortium (Dr R. Gibbs, per-

sonal communication). The genomic sequence was

manually annotated with aids of BlastX.

[Note: position of the Shh sequences, version

of assembly.] [Note #2: in the newest zebrafish

assembly (danRer6), two zebrafish Shha loci were

merged into a single one. The sequence used in

the present study is slightly different from the

newest one (in ver danRer6) with 99.4% of

similarity.]



Documents

Midline Signaling and Evolution of the Forebrain in Chordates: A