Indeed, most neurotransmitters thus far identified are amidated peptides in cnidarians Keywords peptidylglycine a-hydroxylating monooxygenase; cytochrome b561; planarian; neuroendocrine v
Trang 1a neuropeptide processing enzyme, colocalizes with
Akikazu Asada1, Hidefumi Orii1, Kenji Watanabe1and Motonari Tsubaki1,2,3
1 Department of Life Science, Graduate School of Life Science, University of Hyogo (formerly Himeji Institute of Technology), Hyogo, Japan
2 CREST, Japan Science and Technology Agency (JST), Saitama, Japan
3 Department of Molecular Science and Material Engineering, Graduate School of Science and Technology, Kobe University, Hyogo, Japan
Neuropeptides in the brain, in the nervous system, and
in various endocrine cells are synthesized in the rough
endoplasmic reticulum as large precursor proteins After
transit to vesicles and during axonal transportation
along axons, several processing enzymes residing in the
vesicles process the peptides to convert them to mature
forms C-terminal a-amidation of the peptides occurs in
the late stage [1] and is probably a rate-limiting step in
many instances [2] Over half of peptide hormones or neuropeptides are amidated in vertebrates; in insects, greater than 90% of such peptides show the presence of
a C-terminal amide moiety [3] This C-terminal amide is very important in their functions, as its absence often disrupts the activity or receptor-binding properties of the peptide ligands [4] Indeed, most neurotransmitters thus far identified are amidated peptides in cnidarians
Keywords
peptidylglycine a-hydroxylating
monooxygenase; cytochrome b561;
planarian; neuroendocrine vesicle;
neuropeptide amidation
Correspondence
M Tsubaki, Department of Molecular
Science and Material Engineering, Graduate
School of Science and Technology, Kobe
University, Rokkodai-cho, Nada-ku, Kobe,
Hyogo 657-8501, Japan
Fax: +81 78 803 6582
Tel: +81 78 803 6582
E-mail: mtsubaki@kobe-u.ac.jp
Note
The nucleotide sequence of planarian PHM
in this article has been submitted to the
DDBJ ⁄ EMBL ⁄ GenBank databases with
accession number AB195502.
(Received 27 September 2004, revised 2
December 2004, accepted 14 December
2004)
doi:10.1111/j.1742-4658.2004.04528.x
Planarians are one of the simplest animal groups with a central nervous system Their primitive central nervous system produces large quantities
of a variety of neuropeptides, of which many are amidated at their
C terminus In vertebrates, peptide amidation is catalyzed by two enzymes [peptidylglycine a-hydroxylating monooxygenase (PHM) and peptidyl-a-hydroxylglycine a-amidating lyase] acting sequentially In mammals, both enzymatic activities are contained within a single protein that is encoded
by a single gene By utilizing PCR with degenerate oligonucleotides derived from conserved regions of PHM, we succeeded in cloning a full-length cDNA encoding planarian PHM The deduced amino acid sequence showed full conservation of five His residues and one Met residue, which bind two Cu atoms that are essential for the activity of PHM Northern blot analysis confirmed the expression of a PHM mRNA of the expected size Distribution of the mRNA was analyzed by in situ hybridization, showing specific expression in neurons with two morphologically distinct structures, a pair of the ventral nerve cords and the brain The distribution
of PHM was very similar to that of cytochrome b561 This indicates that the ascorbate-related electron transfer system operates in the planarian cen-tral nervous system to support the PHM activity and that it predates the emergence of Plathelminthes in the evolutionary history
Abbreviations
AsA, ascorbic acid; CNS, central nervous system; DBH, dopamine b-hydroxylase; DBHL, dopamine b-hydroxylase-like protein; PAL, peptidyl-a-hydroxylglycine a-amidating lyase; PAM, peptidylglycine a-amidating monooxygenase; PHM, peptidylglycine a-hydroxylating
monooxygenase; MDA, monodehydroascorbic acid; TBH, tyramine b-hydroxylase; VNC, ventral nerve cords.
Trang 2(coral, jellyfish, sea anemones), which possess the most
primitive form of nervous system, suggesting that the
archetype nervous systems may have relied on amidated
peptides [5] The C-terminal a-amidation of
neuro-peptides is catalyzed by two enzymes: peptidylglycine
a-hydroxylating monooxygenase (PHM; EC 1.14.17.3)
and peptidyl-a-hydroxyglycine a-amidating lyase (PAL;
EC 4.3.2.5) consecutively [4,6,7] In vertebrates, PHM
and PAL are expressed together within one polypeptide
chain and denoted as peptidylglycine a-amidating
mono-oxygenase (PAM) [8–10] In some animals, such as
Drosophila[11] and Calliactis [12], PHM is encoded by a
distinct gene In the mollusc, Lymnaea, a multifunctional
PAM containing four different copies of PHM and a
single PAL has been found [13] Endogenously, this
zymogen was converted to yield a mixture of
mono-functional isoenzymes However, in the nematoda,
Caenorhabditis elegans, both types of enzyme are
pre-sent, based on analyses carried out using the SMART
web-based resource (http://smart.embl-heidelberg.de/)
[14]
Planarians – nonparasitic free-living flatworms of
the phylum Plathelminthes, class Turbellaria – are
phylogenetically very distant from vertebrates but have
acquired a central nervous system (CNS), mesodermal
tissues, and a bilateral structure during evolution
[15] Planarians use neuropeptides [16,17] as major
neurotransmitters Planarian neuropeptides (substance
P-, FMRFamide-related peptides) are extensively
a-amidated [16–19], as observed in other invertebrates
Therefore, it is of interest to examine which form
(PAM, PHM, or a different form) does exist in
planar-ians to characterize their primitive CNS as the
puta-tive ancestor of the sophisticated CNS of higher
animals
The monooxygenase reaction catalyzed by PHM (or
the PHM domain) requires copper, ascorbate, and
molecular oxygen, as found for its close family
mem-ber, dopamine b-hydroxylase (DBH) [2] Ascorbic acid
(AsA), which is accumulated in a relatively
concentra-ted manner inside the neuroendocrine secretory vesicles
[20], is converted to monodehydroascorbic acid
(MDA) radical after supplying the electron equivalent
to the catalytic copper center of the enzyme As there
is no transmembrane transport of AsA or MDA
rad-ical in the vesicles [21], it has been postulated that
a common transmembrane electron carrier,
cyto-chrome b561, conveys a reducing equivalent from the
extravesicular AsA to the intravesicular MDA radical
[21–27] Cytochrome b561, first observed in chromaffin
vesicles of bovine adrenal medullae [28,29] and later
found to be distributed in many neuroendocrine tissues
[30–33], is structurally and functionally very different
from other cytochromes [34] It spans the vesicle mem-brane [35] and has a favorable midpoint potential for interacting with AsA and the MDA radical [36–39] Very recently, we have succeeded in cloning the planarian cytochrome b561 cDNA [40] The cyto-chrome b561 gene product was specifically expressed in the planarian nervous system [40], suggesting that the AsA-related electron transfer system, which supports amidation of the neuropeptide, might predate or coin-cide with the emergence of the CNS in animals Thus,
as the second objective of our present study, we exam-ined the colocalization of PHM with cytochrome b561
in the planarian
In the present study, we successfully accomplished the two objectives raised above, namely we obtained
an entire planarian PHM cDNA and identified its spe-cific expression in the planarian CNS as being colocal-ized with cytochrome b561
Results
Cloning of planarian PHM
By using degenerate oligonucleotide primers (PAM-S4⁄ PAM-A3) (Fig 1) coding for the conserved region
of PHMs of various animals, we cloned a 265 bp PCR product from a cDNA library of the head portion of the planarian The deduced amino acid sequence of the PCR product (67 amino acid residues, excluding the primer regions) was highly homologous (containing 10 fully conserved amino acid residues) to other PHMs and therefore consistent with a partial sequence of the planarian PHM cDNA To obtain a longer nucleotide sequence, we carried out nested PCR on the first-strand planarian head cDNA by using new specific primers (spPAM-S2 and spPAM-S3) and SK20 vector primer and a universal primer A PCR product ( 560 bp) was cloned and sequenced The clone was found to encode the 3¢ end of the PHM cDNA, inclu-ding a stop codon (TAG), the 3¢ untranslated region (3¢-UTR), and the poly(A+) tail Stepwise dilution screening of the planarian head cDNA library by PCR using two specific primers (spPAM-S3 and spPAM-A3; PCR probe size, 144 bp) was then conducted One plaque was screened, in vivo excised, recloned, and sequenced completely This clone ( 1300 bp) was found to encode a major part of the planarian PHM gene, except for the NH2-terminal region Several cycles of PCR cloning, using specific primers (PPHM-A4, PPHM-A3, and PPHM-A5) and vector primers, followed by sequencing, gradually narrowed the uncov-ered 5¢ end region Finally, the most 5¢ end PCR prod-uct was obtained, cloned, and sequenced to identify an
Trang 3in-frame stop codon (TAG) at the most 5¢ end and the
putative initiation codon (ATG) followed by an open
reading frame To obtain the entire planarian PHM
cDNA, stepwise dilution screening of the planarian
head cDNA library with PCR was again conducted by
using two specific primers (PPHM-S4 and PPHM-A3)
One plaque, containing the entire planarian PHM
cDNA gene, was obtained, completely sequenced and
analyzed The strategy used to obtain the planarian
PHM cDNA sequence is summarized in Fig 1
The whole planarian PHM cDNA (1384 bp)
con-tained a single open reading frame (1149 bp), which
was preceded by the 5¢-UTR of 87 bp and followed by
the 3¢-UTR of 127 bp and the first 21 nucleotides of
the poly(A+) tail (Fig 2) This indicated clearly that
planarian PHM is coded in a gene separate from that
of PAL
Northern blot
Northern blot analysis of total RNA from the asexual
planarian was carried out The cDNA probe coding
for a major region of planarian PHM hybridized with
an mRNA giving a single band of 1.4 kb (Fig 3),
being consistent with the expected size based on the
cloned cDNA This result also confirmed that there is
no PAM-like gene (i.e a fused gene with PHM and
PAL) in the planarian that will generate an mRNA of
4 kb, as found for vertebrate PAMs
Deduced amino acid sequence of planarian PHM
The deduced amino acid sequence of planarian PHM
predicted a protein with a length of 382 residues
(Fig 2) and a theoretical molecular weight of 43701.34
A hydrophobic segment (13 amino acid residues) at the
N terminus satisfied the consensus rules for the signal sequences, with a cleavage site most likely to occur after amino acid 15, based on analyses conducted by using SignalP WWW and PSORT WWW servers An adjacent propeptide, present in vertebrate PAMs (inclu-ding human, rat, bovine and Xenopus) was absent, as observed for molluscan PHM [13]
The deduced planarian PHM amino acid sequence was compared with those of PHMs and PHM domains
of PAMs from several representative species (including Schistosoma, Drosophila, Calliactis, C elegans, rat, bovine and human) by a multiple sequence alignment (Fig 4) Five histidine and one methionine residue in the central portion of PHM were considered to bind two Cu atoms (CuA and CuB, also termed CuH and
CuM, respectively), which are essential for the catalytic reaction of PHM These six residues (indicated by #) were also conserved in planarian PHM (Fig 4) Fur-thermore, it is known that well-conserved Cys residues form putative disulfide bonds to fix the tertiary struc-ture of the PHM domain Eight of the 10 cysteine resi-dues were conserved in planarian-PHM, as found for other PHM sequences (except for C elegans-PAM) and are marked with + in Fig 4 Additionally, corres-ponding amino acid residues to the residues shown to interact with bound peptide–substrate, such as Arg240, Asn316 and Tyr318 in rat PHM [7,41] were also con-served in planarian PHM (marked with ! in Fig 4) Percentage amino acid identities between these repre-sentative PHMs and PAMs were calculated for their PHM core regions and then tabulated (Table 1) The planarian PHM sequence showed around 40% amino
Fig 1 Schematic representation of the cDNA coding for planarian peptidylglycine a-hydroxylating monooxygenase (PHM) and positions of degenerate and specific primers used in the present study The broad white bar and the two flanking small white bars represent the coding region and the noncoding regions of the planarian PHM cDNA, respectively; whereas the two flanking broad lines outside indicate the k ZAP
II vector The two gray boxes in the planarian PHM cDNA represent the conserved sequences [(P ⁄ V)FAFR(T ⁄ V)H(T ⁄ A)H for positions 231–
239 of human peptidylglycine a-amidating monooxygenase (PAM) and EMCN(F ⁄ L)Y(I ⁄ M ⁄ L) for positions 308–315 of human-PAM, respect-ively], for which two degenerate primers (PAM-S4 and PAM-A3) were designed The degenerate and specific primers used were indicated
by horizontal arrows above the planarian PHM cDNA The PCR clones and partial cDNA clones obtained were indicated below the planarian PHM cDNA in the order of nucleotide sequence determination from top to bottom Arrows indicate the direction of nucleotide sequence determination PO8 indicates the M13 reverse primer.
Trang 4acid identity with all the sequences listed in Table 1 The closest similarity was found to Schistosomas PHM (46.0%) and the most distant one was C elegans PAM (33.6%) To further understand the evolutionary rela-tionship, we conducted a detailed phylogenetic analysis
of planarian PHM in comparison with other PHMs
or PHM domains and DBHs [including tyramine b-hydroxylases (TBHs) and DBH-like proteins (DBHLs)] from various animals (28 sequences in total) The analysis showed that planarian PHM was very distant from DBH, TBH and DBHL (Fig 5) Pla-narian PHM was classified as a member of a subfamily consisting of PHMs from Drosophila, Heterodera,
C elegans and Schistosoma and was distinct from PHM domains of vertebrate PAMs and cnidarian PHMs (Fig 5)
Expression of PHM along the planarian CNS The planarian CNS is basically constituted by an anterior ‘brain’ and a pair of longitudinal ventral nerve cords (VNCs) that run through the length of the body The expression of the planarian PHM gene, as shown
by in situ hybridization, was specific in these neurons (Fig 6A,B) The planarian cytochrome b561 gene was also expressed to a similar extent in the brain and in a pair of VNCs (Fig 6C,D), in accordance with our pre-vious observation [40] The pattern of gene expression
Fig 2 Nucleotide and deduced amino acid sequences of planarian peptidylglycine a-hydroxylating monooxygenase (PHM) The nucleotides are numbered from the 5¢ to the 3¢ end, and the amino acid residues are numbered starting with the first ATG in the open reading frame The translation termination codon is indicated by an asterisk The sequence obtained by the first degenerate PCR is singly underlined, whereas the sequences corresponding to the two specific primers PPHM-S1 and PPHM-A1, used for production of the PCR probe for the Northern blot analysis, are underlined by broken lines.
28S
18S
Fig 3 Northern blot analysis of peptidylglycine a-hydroxylating
monooxygenase (PHM) mRNA in the planarian A PCR probe
pro-duced by two specific primers – PPHM-S1 and PPHM-A1 – was
used for Northern blot analysis Total RNA (10 lg) derived from the
whole planarian body was probed with the planarian PHM cDNA.
Trang 5Fig 4 Multiple alignment of peptidylglycine a-hydroxylating monooxygenase (PHM) or PHM-domain sequences of planarian, Schistosoma, Drosophila, Calliactis, Caenor-habditis elegans, rat, bovine and human ori-gin Amino acid residues common to all sequences are denoted by asterisks below the sequences, whereas conservative resi-dues are indicated by colons (:) or a full stop (.) The five His and one Met residue that form complexes with two Cu 2+ atoms are indicated by # in the uppermost line, whereas eight conserved Cys residues that form four putative Cys-Cys disulfide bridges and three conservative residues that might interact with the bound peptide substrate are marked with + and !, respectively The sequences were obtained from S mansoni (Schistosoma-PHM; AY172995), Drosophila (Drosophila-PHM; AY069103), Calliactis (Cal-liactis-PHM; AF036337), C elegans (C.ele-PHM; AC025726 and C.ele-PAM; U80438), rat (Rat-PAM; X59687 and M82845), bovine (Bovine-PAM; M18683), and human (Human-PAM; S75038) origin PAM, pept-idylglycine a-amidating monooxygenase.
Trang 6for these two proteins showed close similarities to
those of neuron-specific proteins, including a
synapto-tagmin homologue, DjSYT [42], a homologue of PC2
(prohormone convertase 2) [15], a synapsin
homo-logue, a carboxypeptidase E homologue [43] and a
homologue of the nicotinic acetylcholine receptor a7-1
subunit [44] All of these proteins are related to
neuro-transmission, including neurotransmitters, receptors⁄
channels and synaptic vesicles [44]
Expression of PHM in the planarian brain
The closer view of the head region by in situ
hybridiza-tion using the planarian PHM antisense probe showed
that numerous cells, which are considered to be
neu-rons, were clustered around the inverted U-shaped
axon bundles and that these cells were specifically
stained (Fig 7A; red) No appreciable segmental
ture was observed inside the inverted U-shaped
struc-ture, where the neuropil-containing axons and
dendrites locate (Fig 7A) Higher magnification views
showed that indeed only the neuronal cell bodies were
stained with the PHM probe (Fig 7D; brown)
Immu-nohistochemical staining of the head region, by using
antiserum against planarian cytochrome b561, showed
that both the inverted U-shaped axon bundles and the
surrounding neurons were stained (Fig 7B,C; red) A
higher magnification view (Fig 7E), and its merged
image with that stained for nuclei (Fig 7F; blue),
con-firmed this observation
Expression of PHM in the planarian eyes
The planarian eye is composed of two types of cells,
namely photoreceptor cells and pigmented cells [45]
The pigment cells form an eye cup and the
photorecep-tor cells project their microvilli into the inside of the cup, as indicated by immunohistochemical staining with antiplanarian arrestin (green) (Fig 7H,J) Planarian PHM was expressed only in the photoreceptor cells adjacent to the eye cup In particular, signals of the PHM were observable in the region surrounding the nuclei (brown in Fig 7G) The staining image with anti-planarian cytochrome b561was, however, slightly differ-ent from that of planarian arrestin; only the outside of the eye cup was stained (Fig 7J), suggesting that the expression level of cytochrome b561in the microvilli is not so significant
Discussion
This report describes the successful molecular cloning
of planarian PHM cDNA In mammals, PHM and PHL are fused as a bifunctional protein – PAM – that
is encoded by a single gene In some invertebrates, such as Drosophila [11] and Calliactis [12], PHM is encoded by a distinct gene In the mollusc, Lymnaea, a multifunctional PAM was found, which comprised four different copies of PHM and a single copy of PAL [13] Very recently it was reported that a human parasite S mansoni, which belongs to the same phylum (Plathelminthes) as planarian, but a different class (Trematoda), possessed a monofunctional PHM lack-ing a PAL domain [46] In the nematoda C elegans, both types of enzyme precursors, a monofunctional PHM and a bifunctional PAM, are present, based on analysis conducted using the SMART web-based resource (http://smart.embl-heidelberg.de/) (Figs 4 and 5) In our present study, the cDNA probe coding for the major region of planarian PHM hybridized with
an mRNA, giving a single band of 1.4 kb (Fig 2), being consistent with the expected size based on the
Table 1 Percentage identity of amino acid residues between peptidylglycine a-hydroxylating monooxygenases (PHMs) and PHM domains of peptidylglycine a-amidating monooxygenase (PAMs) from various animal species (including Schistosoma-PHM, Drosophila-PHM, Calliactis-PHM, C elegans-Calliactis-PHM, C elegans-PAM, rat-PAM, bovine-PAM, and human-PAM) were calculated by using the FASTA program of the DDBJ server (DNA Data Bank of Japan, Mishima, Japan) for the corresponding PHM core regions The PHM core regions were identical to those used in Fig 5.
Planaria-PHM
Schistosoma-PHM
Drosophila-PHM
Calliactis-PHM
C.ele-PAM
C.ele-PHM
Rat-PAM
Bovine-PAM
Human-PAM Planaria-PHM 100 46.0 41.5 43.3 33.6 38.9 40.0 40.0 41.6 Schistosoma-PHM 100 40.9 39.5 35.7 41.0 36.7 37.0 37.6 Drosophila-PHM 100 45.4 37.5 47.0 40.5 40.5 44.4 Calliactis-PHM 100 41.3 39.0 42.3 42.7 42.3
Trang 7cloned cDNA Furthermore, we were unable to find
any PCR clones longer than that shown in Fig 1 In
addition, there were four in-frame stop codons
down-stream of the terminator codon in the cloned PHM
cDNA (Fig 2) These results suggest that there is no PAM-like gene (i.e a fused gene containing both PHM and PAL) in the planarians, which would gener-ate a mRNA of 4 kb, as found for higher vertebrate PAMs Therefore, PHM and PAL are coded for by different genes in the planarians and only later in the evolution do these genes appear to have fused, forming
a single gene coding for the bifunctional PAM enzymes that we know from vertebrates
High-resolution (2.1 A˚) X-ray structures of rat PHM domain have been reported previously [41,47] The structure shows two copper-binding domains con-nected by a three-residue linker Domain 1 binds one copper (CuH also termed as CuA) with three histidine
Nd ligands (His107, His108 and His172), while domain
2 binds the other copper (CuM, also termed as CuB) with two histidine Neligands (His242 and His244) and
a methionine sulfur (Met314) [47] The two coppers are 11 A˚ apart and face the interdomain space in such
a way that the cleft between them is fully accessible to solvent and peptide substrates [47] As the peptide sub-strate, a-N-acetyl-diiodo-Tyr-Gly, binds in a pocket close to CuM, this leads to the proposal that dioxygen binds to the CuMsite, as previously suggested (on the basis of spectroscopic and kinetic data) for the related enzyme, DBH [48,49] Therefore, it was proposed that the bound peptide substrate mediates the electron transfer from CuA to CuB, where the activation of molecular oxygen takes place On the basis of FT-IR spectroscopy of carbon monoxide-bound PHM and EXAFS data, however, a superoxide-channeling mech-anism was proposed [50–52] This proposal is based on the finding that both copper sites are reactive towards
CO and, by inference, O2 [50] The substrate-free enzyme in the reduced state binds a single CO, assign-able to CuM–CO, with a frequency of 2092 cm)1 Binding of the peptidylglycine substrate caused the appearance of a second CO frequency at 2062 cm)1, assignable to CuH–CO As the binding of CO to the
CuHcenter suggests the potential binding of O2to the
CuH center, it was suggested that dioxygen is first reduced to superoxide at the CuH center and that the subsequent electron transfer is mediated by a super-oxide molecule, which channels from CuHto CuM[50] Tyr79 of rat PHM was considered to have an assisting role in the channeling of superoxide, formed at CuH, towards CuM The corresponding Tyr33 in planarian PHM probably has a similar role, if the latter scenario
is indeed valid
It was suggested that the evolution of the vertebrate CNS might have begun with free-living flatworms – planarians that evolved before the divergence of meta-zoans into invertebrate and chordate branches [53,54]
Fig 5 Phylogenetic relationships of peptidylglycine a-hydroxylating
monooxygenases (PHMs) or PHM domains from various animals.
Phylogenetic relationships of PHMs [or PHM domains of
peptidyl-glycine a-amidating monooxygenase (PAM)], dopamine
b-hydroxy-lases (DBHs), tyramine b-hydroxyb-hydroxy-lases (TBHs) and dopamine
b-hydroxylase-like proteins (DBHLs) were determined as described
in the text Amino acid sequences for Bovine-PAM (M18683),
Bovine-DBH (J05160), Equus-PAM (D29625), Equus-DBH
(AB029430), Human-PAM (S75038), Human-DBHL (BC018756),
Human-DBH (X13256), Mouse-PAM (U79523), Mouse-DBH
(S50200), Mouse-DBHL (AB065134), Mouse-MonX
(monooxyge-nase X) (AK081586), Rat-PAM (M82845 and X59687), Rat-DBH
(L12407), Ciona-PAM (AK113539), Xenopus-PAM (X62771),
C.ele-PHM (AC025726), C.ele-PAM (U80438), C.elegans-TBH (Z99942),
Chicken-DBHL (AF327450), planarian-PHM (this study),
Drosophila-PHM (AY069103), Drosphila-TBH (AE003442), Schistosoma-Drosophila-PHM
(AY172995), Lymnae-PHM4, Lymnae-PHM3, Lymnae-PHM2,
Lym-nae-PHM1 (AF109920), Calliactis-PHM (AF036337), Aplysia-PAM
(AF140271) and Heterodera-PHM (AY242521) were obtained from
the DNA Data Bank of Japan (Mishima, Japan) The phylogenetic
tree was created from the distance matrix by using the
neighbor-joining method.
Trang 8Planarians are among the simplest animals to develop
a body plan of bilateral symmetry and axes of growth
with gradients of genetic expression, enabling
cephali-zation, dorsal and ventral surfaces, medial and lateral
regions, and an aggregate of neural cells in the head
that form a bilobed brain The brain, a neural
struc-ture located in the head, differs from a ganglion by the
following characteristics, [53,54] namely (a) a brain
serves the entire body, not just restricted segments, (b)
it has functionally specialized parts, (c) it is bilobar
with commissures, (d) neurons form the surface with
axons in the central core, (e) interneurons are more
numerous than primary motor or primary sensory
neu-rons, and (f) multisynaptic rather than monosynaptic
circuits predominate The planarian brain fulfills all
the criteria listed above Neurons of the planarian
brain are known to more closely resemble those of
ver-tebrates than those of advanced inverver-tebrates,
exhibit-ing typical vertebrate features of multipolar shape,
extensive dendritic branching, the presence of dendritic
spines for synaptic contact, a single axon per neuron,
expression of certain vertebrate-like neural proteins,
and relatively slow, spontaneously generated electrical
activity [54] Thus, the planarian CNS appears to offer
insight into the origin of the vertebrate CNS by being
the simplest and most remote animal to exhibit these
vertebrate features [54]
Agata and coworkers described in detail the
struc-ture of the CNS of the freshwater planarian
Duge-sia japonica by using several molecular markers,
including prohormone convertase PC2 [15] and
synaptotagmin [42] By using both in situ hybridization
and immunohistochemical techniques, they found that
the CNS is composed of two morphologically distinct
structures, two longitudinal VNC and an anterior
brain The brain forms independent, inverted U-shaped
lobes located dorsally to the VNC The brain contains
nine branched structures (first-ninth) on each side of the body, and nine branches project away from each lobe towards the periphery of the head In spite of this unique structure of the CNS in the planarian in com-parison with those in vertebrates, the basic architecture
of the planarian CNS has undergone functional regio-nalization, based on the comparative genomic analysis
of planarian expressed sequence tags (ESTs) [44] and DNA microarray analysis [43] Furthermore, most of these nervous system-related genes were found to be shared among human, fruit fly and nematode These include genes for neurotransmission, such as the pro-duction of neurotransmitters, receptors⁄ channels for the neurotransmitters, and synaptic vesicles
In higher animals, such as human, neuropeptides are synthesized as preprohormones at the rough endoplas-mic reticulum During translocation across the rough endoplasmic reticulum membranes, the prepart is removed and the prohormone region is sorted into the Golgi apparatus and subsequently into neurosecretory dense-core vesicles [6] In the Golgi apparatus and neu-rosecretory vesicles, the prohormones are processed to yield the bioactive neuropeptides The following pro-cessing steps are involved, namely (a) prohormone convertase, which cleaves at the C-terminal sites of certain dibasic and monobasic residues and liberates immature propeptides, (b) carboxypeptidase specific for basic residues, which removes the C-terminal basic amino acid residues, and (c) PHM and PAL, two con-secutive enzymatic reactions where the C-terminal Gly residue is converted into an amide group [6] In pla-narians, a homolog of PC2 (prohormone convertase 2) [15] and a carboxypeptidase E homolog [43] were iden-tified and both genes were expressed in the planarian CNS [15,43], although a homolog of PAL has not been identified These observations, together with our find-ings on PHM (present study) and cytochrome b561
Fig 6 Whole mount in situ hybridization
with a planarian peptidylglycine
a-hydroxylat-ing monooxygenase (PHM) antisense RNA
probe (A, B) and a cytochrome b561
anti-sense RNA probe (C, D) (A, C) Dorsal view.
(B, D) Ventral view Anterior is left and
pos-terior is right Positive signals for the PHM
were detected in the brain (br), eyes (e),
and ventral nerve cords (vnc) The
distribu-tion was very similar to that of the planarian
cytochrome b561(C, D).
Trang 9[40], indicate that the amidated
neuropeptide-produ-cing neurons had already acquired the neuroendocrine
vesicles equivalent to those seen in the present-day
neurons of higher animals by the time of the
emer-gence of the Plathelminthes in the evolutionary
his-tory
The eye of the planarian is one of the most ancestral
and primitive types of visual systems in animals and is
of particular interest in light of the proposal that all eyes originate from a common evolutionary precursor [55] The planarian eyes consist of two cell types, namely pigment cells and photoreceptors [45] The pig-ment cells form a cup-shaped structure [56], while the photoreceptors are located outside the pigment eye cup and have two types of processes One type of process enters the eye cup and forms rhabdomeres, an
Trang 10assem-bly of microvilli which may be associated with
photo-pigments The other processes are optic nerve fibers
projecting onto the brain [15] As both PHM and a
homologue of PC2 (prohormone convertase), together
with cytochrome b561, are expressed in visual neurons
of the planarian, amidated peptides must have
import-ant mediators during the light perception in the
pla-narian It should be noted that cytochrome b561 was
present mainly in the region between the nucleus and
the rhabdomeres of the photoreceptor cells, but little
in the rhabdomeres The region where cytochrome b561
was present is termed ‘stalk’ and contains smooth
endoplasmic reticulum, Golgi apparatus, and smooth
surfaced and coated vesicles [57] Electron microscopic
studies have also shown that, in the stalk region,
effer-ent nerve fibers form synapses with dendritic spines
protruding from stalks of the photoreceptor cells The
structure appeared similar to that in the neuropil of
the brain [57] In the stalk region of the photoreceptor
cells, neuropeptides may be synthesized and work to
modulate the transmission of light stimulus To clarify
the role of the amidated neuropeptides in the planarian
eyes and visual neurons, more detailed morphological
and physiological studies, including identification of
the neuropeptides, will be required
Experimental procedures
Organisms
All planarians in this study were derived from a single
worm of Dugesia japonica collected in the Irima River in
Gifu, Japan and maintained as described (clonal strain: GI)
[58]
Cloning and sequencing of planarian PHM cDNA
A cDNA library of the planarian made with k ZAP II
vec-tor (Stratagene, La Jolla, CA, USA) [15] was used as a
tem-plate for PCR The degenerate primers used for PCR were
designed based on the conserved amino acid sequences in the PHM domain of human PAM, bovine PAM, rat PAM, Xenopus PAM, and C elegans PAM The primers were PAM-S4 (5¢-TTYGCITWYMGIGTICAYRCNCA-3¢)
(P⁄ V)FAFR(T ⁄ V)H(T ⁄ A)H; positions 231–238 of human PAM underlined in Fig 5] and PAM-A3 (5¢-RTACA TIADRTAIARRTTRCACATYTC-3¢) [coding for the con-served amino acid sequence of EMCN(F⁄ L)Y(I ⁄ M ⁄ L); positions 308–315 of human PAM underlined in Fig 5] The reaction mixture consisted of 1 lL each of 10· Taq buffer, 2.5 mm dNTPs, 10 lm forward primer, 10 lm reverse primer, and planarian head cDNA library mixture,
5 lL of autoclaved H2O, and 0.05 lL of AmpliTaq DNA polymerase (Applied Biosystems) in a final volume of
10 lL Thermal cycling was performed with 40 cycles of the following step program: 94C for 1 min, 43.5 C for 1 min (primer annealing), and 72C for 1 min (primer extension) Final primer extension was made at 72C for 5 min The PCR products were separated in a 6.0% polyacrylamide gel The band with expected size (265 bp) was isolated and used as a template for re-amplification with the same set of primers (PAM-S4⁄ PAM-A3) The PCR product was cloned into pT7Blue T vector (Novagen) by using TA cloning and sequenced with a DNA sequencer (Model DSQ-1000 L; Shimadzu Corp., Kyoto, Japan) Based on the nucleotide sequence obtained, two specific forward primers – spPAM-S2 (5¢-AAACTGCAGAGAAATTGGTAAGAAATCTCC-3¢)
TGCCAGATG-3¢) –were synthesized The former primer had a PstI site as an adaptor The 3¢ end of the cDNA was amplified by the 3¢-RACE method [59] using these primers and cDNA from the planarian head, cloned and sequenced This led to a clarification of the 3¢ half of the planarian PHM cDNA sequence including the terminal poly(A) sequence Based on the nucleotide sequence obtained, one specific reverse primer, spPAM-A3 (5¢-ATGAAGCTTAT CATTTCTTTGAACGCTTCG-3¢), was synthesized By using by using spPAM-S3 and spPAM-A3, the planarian cDNA library was screened by stepwise dilution, as previously described [60] A single phage plaque containing
Fig 7 Expression of the peptidylglycine a-hydroxylating monooxygenase (PHM) gene in the planarian brain and eyes A horizontal section of the planarian head was subjected to in situ hybridization (brown) with the planarian PHM antisense RNA probe (A) and to immunocytochemi-cal staining (red) with antibody against planarian cytochrome b561(B) (C) The image for nuclei staining (blue) with Hoechst 33342 was super-imposed on (B) (A), (B), and (C) are the same frame (D), (E), and (F) are a high magnification of (A), (B), and (C), respectively, and are the same frame White and yellow arrowheads point the same cell in (D), (E), and (F), respectively The expression signal for the PHM gene was observable in the cell body of neurons in the peripheral region of the brain Cytochrome b561was observable in neuropil (np), composed
of axons and dendrites of the neurons, as well as in their cell bodies (G) In situ hybridization of a horizontal section of eye with antisense PHM probe (brown) (H) Immunohistochemical staining with anti-arrestin (green) was superimposed on nuclei-staining (blue) with Hoechst
33342 The frame was the same as that in (G) The PHM gene was expressed in photoreceptor cells (I) Normarski image of a horizontal section of the eye was superimposed on the nuclei-staining (blue) with Hoechst 33342 (J) Double stained image with anti-arrestin immuno-globulin (green) and with anti-cytochrome b 561 immunoglobulin (red) The frame is the same as that in (I) Note that cytochrome b 561 present
in the ‘stalk’ region adjacent to the opening of the pigment eye cup showed a marked contrast to the even distribution of arrestin in photo-receptor cells Scale bars, 0.1 mm.