Search for the Evolutionary Origin of a Brain: Planarian Brain Characterized
by Microarray
Masumi Nakazawa,* Francesc Cebrià, ৠKatsuhiko Mineta,* Kazuho Ikeo,*
Kiyokazu Agata, à and Takashi Gojobori*
*Center for Information Biology and DNA Data Bank of Japan, National Institute of Genetics, Mishima, Japan; RIKEN, Center
for Developmental Biology, Kobe, Japan; àDepartment of Biology, Faculty of Science, Okayama University, Okayama, Japan;
§Department of Cell and Structural Biology, University of Illinois, Urbana-Champaign
The origin of the brain remains a challenging problem in evolutionary studies. To understand when and how the
structural brain emerged, we analyzed the central nervous system (CNS) of a lower invertebrate, planarian. We
conducted a large-scale screening of the head part–specific genes in the planarian by constructing a cDNA microarray.
Competitive hybridization of cDNAs between a head portion and the other body portion of planarians revealed 205 genes
with head part–specific spikes, including essential genes in the vertebrate nervous system. The expression patterns of the
top 30 genes showing the strongest spikes implied that the planarian brain has undergone functional regionalization. We
demonstrate the complex cytoarchitecture of the planarian brain, despite its simple superficiality of the morphology.
Introduction
The brain is thought to have originally emerged as the
aggregation of neural cells on the anterior side, and then
functional differentiation may have occurred among these
cells during evolution. The evolutionary origin and
processes of the brain, however, remain elusive. Recent
studies of the homeodomain-containing proteins Otx and
orthodenticle gave strong evidences for the common
origin for the central nervous system (CNS) of vertebrates
and invertebrates; these homologs have a conserved role in
the formation of the head and/or brain in mice as well as in
Drosophila (Finkelstein et al. 1990; Matsuo et al. 1995).
The identification of brain features in lower organisms
would prove invaluable for an understanding of the
evolution of brain structure. To understand when and
how the structural and functional brain evolutionarily
emerged, we have conducted expression sequence tag
(EST) sequencing and DNA microarray experiments in
planarian.
We chose a freshwater flatworm, planarian (Platyhelminthes; Dugesia japonica) because this organism
shows ‘‘cephalization’’ that has been considered as one
of the simplest brains (Keenan, Coss, and Koopowitz
1981). Although the exact phylogenetic position of
Platyhelminthes is still under debate, their simple body
plan mirror biological features of the groups emerged at
the base of Bilateria (Valentine 1994). The planarian CNS
is composed of two morphologically distinct structures,
the ventral nerve cords (VNCs) and the cephalic ganglion.
The latter forms an independent bilobed inverted U-shaped
structure located dorsally to the VNCs (Agata et al. 1998).
Interestingly, the planarian cephalic ganglion exhibits
many morphological features characteristic of the vertebrate CNS, such as multipolar nerve cells and knoblike
protrusions along dendrites resembling dendritic spines
(Sarnat and Netsky 1985). Three genes homologous to otx
are specifically expressed in the planarian cephalic
ganglion (Agata et al. 1998; Umesono, Watanabe, and
Agata 1997, 1999). These facts categorize the planarian
cephalic ganglion as a brain. Thus, it is of particular interest to know whether the contemporary planarian
maintains features of an evolutionarily primitive brain that
were derived from the common ancestor of the Bilateria.
We identified genes specifically expressed in a planarian brain utilizing a planarian cDNA microarray.
Microarrays are powerful tools to examine the expression
profile of multiple genes simultaneously, leading to
prediction of gene functions and relationships (Duggan
et al. 1999). In our EST project, we have sequenced more
than 9,000 redundant clones derived from a planarian
cDNA library. We randomly selected 1,640 nonredundant
genes from these sequenced clones, the upper limits of
our spotting capability. To create the cDNA microarray,
we overcame several difficulties such as ambiguous reproducibility, inefficient hybridization, and unclearness
of quantitative spike criteria by establishing our own
experimentation.
We then conducted competitive hybridization experiments of cDNAs between a head portion and the other
body portion of planarians in order to screen genes
specifically expressed in a planarian brain. In this study,
we show extensive evidence that a planarian brain has
complex cytoarchitecture characterized by gene expression
profiles, implying that the functional regionalization took
place in the planarian brain in spite of superficial
simplicity of morphology.
Key words: central nervous system, origin of brain, planarian,
microarray.
E-mail: [email protected].
Probe Preparation
Mol. Biol. Evol. 20(5):784–791. 2003
DOI: 10.1093/molbev/msg086
Ó 2003 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038
784
Materials and Methods
Planarian
A clonal strain of planarian, Dugesia japonica, was
obtained from the Irima River, Gifu prefecture, Japan, and
established by Kenji Watanabe at the Himeji Institute of
Technology. Intact worms were cultured in autoclaved tap
water at 228C.
After homology search in our planarian EST project
(Mineta et al. in preparation), 1,640 nonredundant clones
from planarian ESTs were spotted. We amplified the
clones by PCR in a 100-ll system using vector-specific,
Evolutionary Origin of a Brain 785
FIG. 1.—Planarian cDNA microarray. (A) Flow chart demonstrating the strategy for large-scale screening of planarian brain-related genes. An
image of the planarian cDNA chip is included. Planarians were divided into a head portion and the other body portion at the doted line. Cy-3/Cy-5
fluoro-linked cDNAs derived from the two parts were competitively hybridized on a planarian cDNA chip. The fluorescent signals were separately
scanned. Here, a superimposed view of the two images is represented in pseudocolors (head, green; body, red). (B) Histogram of the averaged H/B ratio
from five experiments. Data divisions are decided by the equations displayed at X-axis. ‘‘d’’ represents the standard deviation of the data set. The spot
numbers included in each division are indicated inside the graph.
amino-linked primers in a 96-well format. After
checking by electrophoresis, PCR products were precipitated by centrifugation (4,000 rpm for 1 h with TS38LB rotor [TOMY]) after incubation with sodium
acetate and 2-propanol at 208C overnight. The
luciferase gene (pGL2-Basic Vector [Promega]) was
used as an external control.
Fabrication of cDNA Microarray
Microscopic slides coated with poly-L lysine were
used as the base of the DNA chip (Matsunami). PCR
amplified samples were solved in spotting buffer (DMSO:
Polymer [FujiFilm]: TE ¼ 2:1:1). After denaturation at
958C for 3 min, the samples were spotted onto slide
glasses with 32 solid pins at the pitch of 0.5 mm by an
SPBIO 2000 arrayer (Hitachi Software Engineering).
Spotted glasses were burnt at 808C for 1 h in an oven,
followed by exposure to 60 mJ of UV to immobilize the
DNAs. We blocked glass surfaces with 175 mM succinic
anhydride in 1-methyl-2-pyrrolidinone and 44 mM sodium
borate (pH 8.0).
Target Preparation
As the brain occupies the anterior part of the
planarian, we designed experiments to screen brain-related
genes as shown in figure 1A. We extracted poly-A tailed
messenger RNA (mRNA) separately from the planarian
head portion and the body portions. Planarians were
starved for at least 1 week before use. Approximately 600
worms were divided into two parts as shown in figure 1A
and collected for one experiment. Total RNAs derived
from each part were extracted using Isogene (Nippon
Gene), and mRNAs were then isolated using OligoTexdT30 ,Super. (Takara).
Labeling of Target
Fluoro-linked nucleotides were incorporated during
the reverse transcription of mRNAs to cDNAs. Two
micrograms of mRNA were incubated with 0.5 lg random
6-mer (Takara) at 708C for 5 min. Samples were then
incubated with 0.5 mM dGTP, dATP and 0.2 mM dTTP,
dCTP; 0.1 mM fluorolink Cy3 or Cy5-dUTP; dCTP
(Pharmacia); 1 unit RNase inhibitor; 10 mM dithiothreitol;
first strand buffer; and 400 units of Super Script II
(GIBCO BRL) at 258C for 10 min followed by at 428C for
1 h. The reaction was stopped with EDTA, and mRNAs
were hydrolyzed by incubation with NaOH at 658C for
1 h. After neutralization by adding Tris-HCl (pH 7.4), the
labeled cDNAs were washed with TE and concentrated by
using Microcon-30 (Millipore).
Hybridization and Scanning
Fluoro-labeled targets were adjusted to 5 3 SSC and
denatured at 958C for 2 min. After the addition of SDS to
a 0.25% final concentration, the target was placed on the
microarray under a glass cover slip measuring 22 mm 3
45 mm. Hybridization was performed in a humid chamber
at 658C for 24 to 48 h. The slide was then washed in 2 3
SSC containing 0.1% SDS, followed by four incubations
in 0.2 3 SSC and 0.1% SDS at room temperature (RT), at
408C, and two additional times at RT. The slide was
immediately scanned by ScanArray 4000 Version 1.2
786 Nakazawa et al.
software (GSI Lumonics). Analyses were performed using
QuantArray Version 1.0.0.0 software (GSI Lumonics).
Table 1
The Top 205 Clones Showing the Highest H/B
Whole Mount in Situ Hybridization
Categorya
For whole mount in situ hybridization, we utilized
planarians starved for 1 week. Worms were treated with
2% HCl for 5 min at 48C and then fixed in Carnoy’s
solution (ethanol: chloroform: acetic acid in proportions
6:3:1, respectively) for 2 h at 48C. Hybridization was
performed using 20 ng/ml of DIG-labeled riboprobes, as
previously described (Umesono, Watanabe, and Agata
1997; Agata et al. 1998).
No match
Unknown
Hypothetical
Known
Analyses of the Data from cDNA Chip Experiments
After the addition of the control samples and the
duplicated spots, the planarian cDNA chip comprised a total
of 1,728 spots. We used 1,689 spots in the following
analysis after excluding those spots with bad forms. The
fluorescent intensities of each dye were separately scanned.
We calculated the ratio of fluorescent intensities between
two dyes on each spot (H/B ratio) according to the following
equation: H/B ratio ¼ (fluorescent intensity of a head-part
cDNA pool)/(fluorescent intensity of a body-part cDNA
pool).
We conducted five independent experiments. As the
hybridization efficiency differs among experiments, each
H/B ratio was normalized to the median of the data set of
each experiment. The average of H/B ratio among five
experiments was calculated for each spot. Finally, all the
spots were ordered according to the averaged H/B ratio
from the highest value to the lowest value.
Results
Brain-Related Genes Discovered by Using Planarian
cDNA Chip
We conducted five independent experiments. Four
experiments were done using the composition of head
cDNA labeled by Cy-5 and body cDNA labeled by Cy-3,
and the fifth experiment was done using inverted
composition. The average of Pearson’s correlation coefficient (PCC) among the five experiments was 0.63.
However, when the PCC of the fifth experiment was
excluded, it gave a higher value of 0.80. This might be
because different the fluorescent dyes, Cy-3 and Cy-5,
were not integrated at the same amount, even in the same
cDNA pools.
We calculated an H/B ratio for each spot. An H/B
value actually reflects the relative abundance of the spotted
gene within the head portion to the abundance of the
spotted gene within the other body portion. The histogram
of H/B ratios (fig. 1B) demonstrates a tremendous
difference between gene groupings, with larger and
smaller H/B ratios at approximately 1.45. Thus, we
designated an H/B ratio of more than 1.45 as the criterion
for head part–specific genes. We then found that 205 genes
showed H/B values higher than this criterion.
The 205 genes are separated into four categories as
a result of homology search (table 1). Eighty-two genes are
homologous to functionally known genes (designated as
Total
Number of
Spots (%)
94
26
3
82
(45.9)
(12.7)
(1.40)
(40.0)
Clones Similar to Nervous
System–Related Gene (%)b
33 (16.1)
205 (100)
NOTE.—The 205 clones showing the highest H/B ratio were classified
according to the results of homology search using BlastX/BlastN against to
international database.
a
Criteria of category are as follows. No match: the clones that do not match
any genes in database with less than 104 E-value by Blast search; Unknown: the
clones that match to EST/ORF/hypothetical genes in database with less than
104 E-value by Blast search; Hypothetical: the clones that match to hypothetical/
predicted proteins in database with less than 104 E-value by Blast search;
Known: the clones that are the exception to the No match, Unknown, and Hypothetical categories.
b
Among the Known, the clones homologous to nervous system–related
genes of other species are counted.
‘‘Known’’ in table 1). In particular, 33 genes are
homologous to the neural-related genes in other organisms, such as synapsin, synaptotagmin, prohormone
convertase 2, and nicotinic acetylcholine receptor. These
genes are functionally essential and well conserved
throughout CNS of higher organisms. Because of the high
similarity at the amino acid level, these homologous genes
in the planarian should be basic components in planarian
CNS as well as in other organisms. The remainder of the
123 genes, categorized into three additional groups, are
functionally unknown (designated as ‘‘No match,’’ ‘‘Unknown,’’ and ‘‘Hypothetical’’ in table 1). Therefore, we
uncovered four times as many unknown head part–specific
genes through cDNA microarray analysis as known
neural-related genes. The planarian brain contains a large
number of functionally unknown genes that are currently
beyond our fundamental understanding of the basic neural
genes.
Conservation of the Nervous System–Related Gene
Homologues Supports the Unique Origin of the Brain
Two examples of the neural-related gene homologs
among 205 candidates imply that the planarian brain is
derived from the same origin as the invertebrate and
vertebrate. The first example is neural cell adhesion
molecule (NCAM). Planarian EST 0944_HH shows a high
amino acid sequence similarity to NCAM of Xenopus
laevis at the fibronectin type III region (E-value of Blast
search: 4.00e-6) (fig. 2A). This gene product, showing
a high H/B value (it ranks as the 21st H/B value), is
expressed throughout the entirety of the planarian CNS,
including the cephalic ganglion, longitudinal VNCs, and
peripheral nervous plexus. In vertebrates, NCAM is
involved in neuron-neuron adhesion, neurite fasciculation,
and the outgrowth of neurites; this molecule is expressed
in neurons and presumptive neural tissues (Edelman
1986). In insects, fasciclin is a functional homolog of
NCAM (Grenningloh et al. 1990). It is worth noting that
the planarian NCAM homolog is more similar to the
Evolutionary Origin of a Brain 787
FIG. 2.—Nervous system-related genes in planarian. The inferred amino acid sequences of planarian EST are aligned with homologous genes in
other species. (A) Planarian EST 0944_HH (NCAM), (B) 0251_HH (arrestin). Asterisks indicate conserved amino acids among the aligned species.
vertebrate homologous gene than to that of an insect. It
might suggest an ancient emergence of the NCAM gene before the planarian divergence from Deuterostomia lineage.
The second example of the neural-related gene
homologs is arrestin. EST 0251_HH (7th H/B value)
shows a high sequence similarity to beta arrestin of Homo
sapiens (E-value: 1.00e-32) (fig. 2B). This EST is expressed highly specifically in the visual cells, suggesting
a function in planarian phototransduction (see fig. 3).
Arrestin participates in the termination of phototransduction in both vertebrates and invertebrates (Shinohara et al.
1992). High conservation of the arrestin-mediated phototransduction from the planarian to the higher organisms
indicates the functional importance of this pathway.
These two genes mirror the overall pattern of neuralrelated genes found in our search, demonstrating that the
planarian nervous system shares common features with the
nervous system of other invertebrates and vertebrates.
These observations strongly support the common origin of
the CNS, even though they have independently evolved in
each lineage after diversification. Common CNS features
from planarian to mammals should reflect the evolutionarily ancestral brain structure.
Functional Regionalization in Planarian Brain Inferred
from the Expression Patterns of Head Part–Specific Genes
We then performed whole mount in situ hybridization
for the head part–specific genes, regardless of whether
they were known or unknown, in order to know the
molecular features of the planarian brain. We detail the top
30 genes having the highest H/B values from the total
205 candidates. We observed clearly distinct expression patterns that are separated into seven categories, designated types A to G (fig. 3).
Type A genes are expressed throughout the CNS.
These genes are likely to have basic and fundamental
functions in the nervous system of most organisms. Most
of the genes in type A show sequence similarities to neural
specific genes in other organisms, including 0944_HH/
NCAM (21st), 4307_HH/nicotinic acetylcholine receptor
(23rd), and 2814_HH/synapsin (26th). The planarian ESTs
0180_E (10th) and 6698_HH (25th) are also contained
within this group, although they are functionally unknown.
These genes are indispensable in keeping the general
functions of the CNS, as they are ubiquitously expressed.
Compared with type A genes, type B and C genes
exhibit the expression patterns in specific regions of the
cephalic ganglion. Type B genes are expressed particularly
in the leaflike regions on both sides in the bilobed cephalic
ganglion. The planarian cephalic ganglion actually contains two leaflike structures connected by a commissure,
where neuronal cell bodies accumulate. Type B genes may
have a role in processing and transporting the biological
signals to the body, as this region is connected to VNCs.
One of the type B genes, EST 0107_E (24th), which is an
unknown gene, has high sequence similarity with human
EST clones, KIAA0513 (E-value: 7.00e-8), derived from
human brain (Nagase et al. 1998). This should be a novel
brain-related gene conserved from the planarian to the
human.
Contrary to type B genes, type C gene expressions are
restricted to the branch regions in the cephalic ganglion.
These genes are likely to be involved in signal transduction. For example, 1791_HH (11th) homologous to the
G-protein alpha subunit (E-value: 5.00e-32) is mainly
expressed in the distal part of each branch with reduced
levels in other portions. Due to the sequence similarity
with G-protein, relating to the signal transduction
throughout the species (Simon, Strathmann, and Gautam
1991), 1791_HH may mediate the transduction of signals
received in sensory cells locating on the head periphery.
These differing expression patterns imply that
functional regionalization occurs within the planarian
CNS. Since the branch and leaflike regions are morphologically distinguished, these structurally different regions
should possess specialized functions.
The expression patterns of the type E genes, however,
demonstrate that the regionalization occurs even inside the
leaflike structure, which appears to be morphologically
homogenous. EST 1020_HH (8th), 0053_E (18th), and
0639_E (29th) in the type E genes are functionally
unknown genes. These expression patterns are characterized by the following three points. First, they show
788 Nakazawa et al.
Evolutionary Origin of a Brain 789
FIG. 4.—Cytoarchitecture map of planarian brain. The represented
patterns of each type (A to G) are schematically drawn using different
colors at a half side.
gradient expression patterns within the leaflike region. The
posterior portion of the brain stained more strongly than
the anterior side, and the interior of the two leaves stained
with a greater intensity than the exterior. This gradient
pattern suggests that the different areas within the leaflike
structure possess different functions, although they appear
morphologically homogeneous. Second, 0053_E and
0639_HH are expressed around the eyes, as observed
with the type D 0251_HH/beta arrestin homologous gene.
Moreover, 0053_E is expressed at the tip of the head.
These stained cells may be a kind of a photoreceptor or
mechanoreceptor, judging from the cell location. Third,
1020_HH and 0053_E are expressed in dispersed cells
along the VNCs. These cells are closely associated to
VNCs and could have some kind of supporting role for
neurons from the nerve cords.
Type E expression patterns indicate the regionalization in the planarian cephalic ganglion, similar to that
observed in the human brain. In fact, it is reported that
each branch mediates distinct functions. For example, two
eyes are located on the dorsal side of the third branch, with
the axon of each eye extending towards the third branch.
The sixth to ninth branches extend to the surface of the
head region, forming auricles that make a putative sensory
organ of taste (Umesono, Watanabe, and Agata 1999).
FIG. 3.—Expression patterns of the top 30 head part-specific genes.
The top 30 genes showing the highest H/B values are categorized by the
expression patterns into type A to type G (see text). The entire planarian
and the anterior part (ventral views) of the whole mount in situ
hybridization utilizing gene probes are pictured. Homologous sequence
names are shown with E-value. H/B value and order are given in blue.
The clones without the sequence names do not have homologous genes in
international database.
These different sensory neurons should innervate to the
specific region of the brain through specific branches,
indicating that different signals are processed in distinct
areas of the cephalic ganglion. Judging from the expression patterns of type E genes, photosignals may be
processed in the posterior/interior region in the planarian
cephalic ganglion.
The genes categorized in type F show particular
patterns in which expression signals appear not only in
cephalic ganglion but also in the entire head region. Based
on the results of RNA interference experiments of these
genes that are expressed in the anterior side of the
planarian body, we presume that they are strongly related
to cephalization in planarian (Cebrià et al. 2002).
The last group, type G genes exhibit specific
expressions in sensory cells. ESTs 1681_HH (13th) and
0821_HN (14th) are expressed on the lateral ends of the
brain branches where sensory organs such as the auricles
are located (Umesono, Watanabe, and Agata 1999).
1681_HH, similar to the PKD2 gene that confers
autosomal dominant polycystic kidney disease in Homo
sapiens (Pennekamp et al. 1998) (E-value: 2e-11), is expressed in two distinct curved regions around the brain.
Another gene, unknown 0821_HN, is expressed in a single
and inside curved region along periphery of the head. The
stained cells by these genes are located in the extended
region to the tips of the brain branches, suggesting these
cells are related to the CNS.
Collectively, the variety of the expression patterns of
the top 30 head part–specific genes demonstrates the
highly organized planarian CNS. We designate these 30
genes as the planarian brain-specific genes. Their expressions indicate that planarian CNS is functionally
regionalized.
Discussion
Use of cDNA microarray revealed 205 planarian
genes strongly expressed in the head part, regardless of
known or unknown functions. Of the 205 head part–
specific genes, we discovered 33 genes homologous to the
nervous system–related genes of both vertebrates and invertebrates. Several genes, including NCAM and arrestin,
exhibited similar expression patterns in the planarian to
previously characterized higher organisms, suggesting that
the essential gene set encoding the fundamentals of the
nervous system have remained unchanged since the
divergence between Deuterostomia and Lophotrochozoa
lineages. Our results indicate that the CNS of the Bilateria
must have derived from a single origin.
We observed functional regionalization in the
planarian CNS. Previous studies suggested that the
planarian brain is divided into four regions according
to otx homologous gene expressions. The first region
characterized by DjotxA shows expression in photoreceptor cells. DjotxB is expressed in the intermediate regions
connecting to the VNCs. Djotp is expressed in the brain
branches (Umesono, Watanabe, and Agata 1997, 1999).
The fourth region is lateral to branches where otx/otp is
undetectable. Moreover, we identified at least seven
different expression patterns in the planarian CNS,
790 Nakazawa et al.
indicating complex cytoarchitecture in the planarian brain
as shown in figure 4. The functional regionalization of the
planarian brain allows the performance of complex
processes, as observed in higher organisms. These data
suggest that the planarian brain can be utilized as a model
of the CNS in higher organisms.
We cannot, however, rule out the possibility that
functional regionalization of the CNS in the planarian have
occurred independently of the other species. However, this
study clearly showed common character that the brain is
functionally regionalized even in lower invertebrates. The
primitive brain that emerged at the beginning of the
Bilateria is supposed to have been not so simple. At the
very beginning of the metazoan, nerve cells may have just
accumulated at the particular region as observed in the
modern cnidarian. By the time of Bilaterian diversification,
the aggregation of neurons may have then produced
functional differentiation, creating hierarchies to perform
higher-order processes. Assuming that emergence of the
plathyhelminthes in the evolutionary history was before
the Cambrian explosion (Valentine 1994), the primitive
brain has likely emerged more than 560 MYA. Many
autonomic functions of the brain are intact from flat worms
to human, and new high-ordered functions may have been
added in different regions. Moreover, in some areas of the
brain, evolution might have given modifications of
important functions specific to the species (Gerhart and
Kirschner 1997).
We have discovered brain-specific genes in the
planarian, including the genes with unknown functions.
It may be because orthologous genes have diverged to
a great extent among the species. The possibility remains, however, that these are novel genes not yet identified in other species. Further analysis of these planarian
genes may reveal novel brain-specific genes in mammals, as in the case of planarian EST 0107_E, which is
homologous to human brain EST. We suggest that evolutionary studies play an important role as new gene finder
for higher organisms. Moreover, we would conduct the
functional analyses of the candidates obtained as
planarian brain-specific genes using knockout methods
such as planarian RNA interference (Sánchez Alvarado
and Newmark 1999). Competitive hybridization of the
expressed genes between knocked out and intact worms
on a cDNA chip may reveal the transcriptional pathways of brain-specific genes in the planarian, giving hints
for the analyses of genetic networks in mammalian
CNS. We suggested that the evolutionary study of brains
give profound insights into the understanding of the
complex nervous systems in higher organisms such as
mammals.
Acknowledgments
We would like to thank Ms. Yumi Takezawa, Ms.
Akemi Mizuguchi, Ms. Hiroko Oizumi, and Ms. Chie
Iwamoto in NIG for experimental help. We greatly
appreciate the help of Dr. Atsumi Tsujimoto, Mr. Hiroaki
Ono, Ms. Noriko Yurino in DNA Chip Research Inc., and
Hitachi Software Engineering Co., Ltd. for their material
and technical support. This study was supported by JSPS
research fellowships to M.N., a Grant-in-Aid for Scientific
Research on Priority Areas (C) and a Grant-in-Aid for
Creative Basic Research to K.A. and T.G. from the
Ministry of Education, Culture, Sports, Science and
Technology, Japan.
Literature Cited
Agata, K., Y. Soejima, K. Kato, C. Kobayashi, Y. Umesono, and
K. Watanabe. 1998. Structure of the planarian central nervous
system (CNS) revealed by neuronal cell markers. Zool. Sci.
15:433–440.
Cebrià, F., C. Kobayashi, Y. Umesono et al. (11 co-authors).
2002. FGFR-related gene nou-darake restricts brain tissues to
the head region of planarians. Nature 419:620–624.
Duggan, D. J., M. Bittner, Y. Chen, P. Meltzer, and J. M. Trent.
1999. Expression profiling using cDNA microarrays. Nat.
Genet. 21(Suppl.):10–14.
Edelman, G. M. 1986. Cell adhesion molecules in the regulation
of animal form and tissue pattern. Annu. Rev. Cell. Dev. Biol.
2:81–116.
Finkelstein, R., D. Smouse, T, M. Capaci, A. C. Spradling, and
N. Perrimon. 1990. The orthodenticle gene encodes a novel
homeo domain protein involved in the development of the
Drosophila nervous system and ocellar visual structures.
Genes Dev. 4:1516–1527.
Gerhart, J., and M. Kirschner. 1997. Cells, embryos, and
evolution. Blackwell Science, Malden, Mass.
Grenningloh, G., A. J. Bieber, E. J. Rehm, P. M. Snow, Z. R.
Traquina, M. Hortsch, N. H. Patel, and C. S. Goodman. 1990.
Molecular genetics of neuronal recognition in Drosophila:
evolution and function of immunoglobulin superfamily cell
adhesion molecules. Cold Spring Harbor Symp. Quant. Biol.
55:327–340.
Keenan, L. C., R. Coss, and H. Koopowitz. 1981. Cytoarchitecture of primitive brains: golgi studies in flatworms. J
Comp. Neurol. 195:697–716.
Matsuo, I., S. Kuratani, C. Kimura, N. Takeda, and S. Aizawa.
1995. Mouse otx2 functions in the formation and patterning of
rostral head. Genes Dev. 9:2646–2658.
Nagase, T., N. Ishikawa, A. Tanaka, H. Kotani, N. Nomura, and
O. Ohara. 1998. Prediction of the coding sequences of unidentified human genes. IX. The complete sequences of 100
new cDNA clones from brain which can code for large
proteins in vitro. DNA Res. 5:31–39.
Pennekamp, P., N. Bogdanova, M. Wilda, A. Markoff, H.
Hameister, J. Horst, and B. Dworniczak. 1998. Characterization of the murine polycystic kidney disease (Pkd2) gene.
Mamm. Genome 9:749–752.
Sánchez Alvarado, A., and P. A. Newmark. 1999. Doublestranded RNA specifically disrupts gene expression during
planarian regeneration. Proc. Natl. Acad. Sci. USA 96:5049–
5054.
Sarnat, H. B., and M. G. Netsky. 1985. The brain of the planarian
as the ancestor of the human brain. Can. J. Neurol. Sci. 12:
296–302.
Shinohara, T., T. Kikuchi, M. Tsuda, and K. Yamaki. 1992. A
family of retinal s-antigens (arrestins) and their genes:
comparative analysis of human, mouse, rat, bovine and
Drosophila. Comp. Biochem. Physiol. 103B:505–509.
Simon, M. I., M. P. Strathmann, and N. Gautam. 1991. Diversity of G protein in signal transduction. Science 252:
802–808.
Evolutionary Origin of a Brain 791
Umesono, Y., K. Watanabe, and K. Agata. 1997. A planarian
orthopedia homolog is specifically expressed in the branch
region of both the mature and regenerating brain. Dev.
Growth Differ. 39:723–727.
———. 1999. Distinct structural domains in the planarian brain
defined by the expression of evolutionarily conserved
homeobox genes. Dev. Genes Evol. 209:31–39.
Valentine, J. W. 1994. Late Precambrian bilaterians: grades and
clades. Proc. Natl. Acad. Sci. USA 91:6751–6757.
Dan Graur, Associate Editor
Accepted January 9, 2003