Origin and Diversification of a Salamander Sex
Pheromone System
Sunita Janssenswillen,y,1 Wim Vandebergh,y,1 Dag Treer,1 Bert Willaert,1 Margo Maex,1
Ines Van Bocxlaer,1 and Franky Bossuyt*,1
1
Amphibian Evolution Lab, Biology Department, Vrije Universiteit Brussel (VUB), Brussels, Belgium
These authors contributed equally to this work.
*Corresponding author: E-mail: [email protected].
Associate editor: Emma Teeling
y
Abstract
Sex pheromones form an important facet of reproductive strategies in many organisms throughout the Animal Kingdom.
One of the oldest known sex pheromones in vertebrates are proteins of the Sodefrin Precursor-like Factor (SPF) system,
which already had a courtship function in early salamanders. The subsequent evolution of salamanders is characterized
by a diversification in courtship and reproduction, but little is known on how the SPF pheromone system diversified in
relation to changing courtship strategies. Here, we combined transcriptomic, genomic, and phylogenetic analyses to
investigate the evolution of the SPF pheromone system in nine salamandrid species with distinct courtship displays. First,
we show that SPF originated from vertebrate three-finger proteins and diversified through multiple gene duplications in
salamanders, while remaining a single copy in frogs. Next, we demonstrate that tail-fanning newts have retained a high
phylogenetic diversity of SPFs, whereas loss of tail-fanning has been associated with a reduced importance or loss of SPF
expression in the cloacal region. Finally, we show that the attractant decapeptide sodefrin is cleaved from larger SPF
precursors that originated by a 62 bp insertion and consequent frameshift in an ancestral Cynops lineage. This led to the
birth of a new decapeptide that rapidly evolved a pheromone function independently from uncleaved proteins.
Key words: amphibians, evolution, SPF pheromone system.
Introduction
Article
Sex pheromones, molecules that are secreted by an individual
to elicit a reaction in an individual of the other sex, form an
important element of courtship and reproduction in the
Animal Kingdom. One of the oldest known vertebrate sex
pheromone systems is the Sodefrin Precursor-like Factor (SPF)
system, a family of proteins that already had a courtship
function early in salamander evolution (Van Bocxlaer I,
Treer D, Maex M, Vandebergh W, Janssenswillen S, Stegen
G, Kok PJR, Willaert B, Matthijs S, Martens E, Mortier A, De
Greve H, Proost P, Bossuyt F, downloaded from http://dx.doi.
org/10.7287/peerj.preprints.457, last accessed August 26,
2014). In advanced salamanders (Salamandroidea), which reproduce by internal fertilization without copulation (i.e.,
males deposit a spermatophore that females pick up with
their cloaca), male SPF proteins have been demonstrated to
accelerate or induce spermatophore pickup in two families. In
terrestrially courting Plethodontidae (lungless salamanders),
these molecules are widely used and are either slapped on the
nose of the female with the male’s mental gland or scratched
into the female’s skin by his premaxillary teeth (Arnold 1977;
Houck and Reagan 1990; Houck and Sever 1994; Houck and
Arnold 2003; Houck et al. 2008). In the family Salamandridae,
the use of SPF proteins was shown in the aquatically courting
newt species Lissotriton helveticus, where pheromones are
produced in the male abdominal gland and tail-fanned toward the female’s nostrils (Wambreuse and Bels 1984).
These proteins persuade females to follow the male’s path,
which culminates in spermatophore pick-up without substantial physical contact between both sexes. In both families,
SPF courtship pheromones are effective as full-length proteins, that is, after removal of an N-terminal signal peptide,
they are not further cleaved to obtain functional proteins.
In addition to full-length courtship pheromones, males of
the salamandrid genus Cynops use a small peptide (sodefrin,
or one of its variants) that is cleaved posttranslationally from
a larger SPF-precursor protein (Kikuyama et al. 1995; Iwata
et al. 1999). This decapeptide was shown to be able to attract
females from a short distance (i.e., an attractant pheromone)
and to be species-specific (Iwata et al. 1999; Yamamoto et al.
2000). The SPF-family thus has evolved in multiple mature
forms (uncleaved proteins and cleaved peptides) that can
influence behavior in salamanders in different ways and can
therefore be considered important components of salamander courtship.
Salamandridae form an ideal family to study the evolution
of pheromones in relation to courtship diversification. The
proposed ancestral courtship mode of salamandrids is the use
of amplexus, a form of contact also known in anurans, in
which the male embraces the female as part of the mating
process (Houck and Arnold 2003). Several species of salamandrids display modified forms of amplexus, often accompanied
by tail-fanning or by male use of specific courtship glands
(Arnold 1977; Halliday 1977, 1990; Houck and Arnold 2003;
ß The Author 2014. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please
e-mail: [email protected]
472
Mol. Biol. Evol. 32(2):472–480 doi:10.1093/molbev/msu316 Advance Access publication November 20, 2014
MBE
Salamander Sex Pheromone System . doi:10.1093/molbev/msu316
Sever 2003). In contrast, most aquatically courting Eurasian
salamandrids lack amplexus or another form of strong physical contact, and largely rely on the use of pheromones to lead
females over a spermatophore. Little is known on how the use
of SPF pheromones has diversified in relation to such evolutionary transitions in courtship behavior.
Here, we combined transcriptomics, genomics, and phylogenetics to study the evolution of SPF pheromones in nine
salamandrid species with diverse courtship strategies. First, we
combined our transcriptomic data with relevant sequences in
the genome of other vertebrates to reconstruct the evolutionary origin of the SPF pheromone system. Next, we studied
the evolutionary dynamics of this pheromone system by investigating how SPF precursor diversity in the cloacal region in
salamandrids has evolved in species with distinct mating
strategies. Finally, we combined our transcriptome data
with known relationships of salamandrids to demonstrate
how and when a new decapeptide pheromone originated
from full-length SPF precursors in the genus Cynops.
Results and Discussion
Table 1. Number of Cloacal SPF Contigs Assembled at 99% Identity
per Species and Sex.
Species
Modern Asian newts
Pachytriton granulosus
Cynops pyrrhogaster
Modern European newts
Ichthyosaura alpestris
Lissotriton vulgaris
L. helveticus
Gender Alpha Beta
Beta
Total
SPF
SPF Sodefrin
Male
Male
Male
Female
Male
Female
Male
Female
Male
Euproctus platycephalus
New World newts
Notophthalmus viridescens Male
Female
Taricha granulosa
Male
Primitive newts
Pleurodeles waltl
Male
Female
2
19
7
12
0
3
9
34
2
0
6
0
2
0
0
8
1
13
3
19
0
0
0
0
0
0
0
0
0
10
1
19
3
21
0
0
21
0
3
26
0
2
0
0
0
47
0
5
2
0
10
0
0
0
12
0
SPF Transcript Diversity in Salamandridae
The cloacal tissue of 14 sexually active animals (nine males
and five females, of nine species) was sampled and 96 cDNA
transcripts per tissue were amplified and sequenced with a
diverse set of primers (supplementary table S1,
Supplementary Material online). When grouping into contigs
of 99% identity, 158 different SPF transcripts (GenBank numbers KM463769–KM463825, KM463827–KM463845, KM46
3847–KM463884, KM463886–KM463889, KM463891–
KM463933) were obtained from 10 of the 14 cloacal tissues
investigated (table 1). Our cDNA sequences indicate a high
intra and interspecific SPF transcript diversity, with coding
cDNA sequences ranging from 21 to 229 amino acids. Part
of this observed variation is caused by multiple duplications in
SPF genes during the evolution of Caudata (salamanders),
which has expanded the extant arsenal of expressed SPF proteins (Van Bocxlaer I, Treer D, Maex M, Vandebergh W,
Janssenswillen S, Stegen G, Kok PJR, Willaert B, Matthijs S,
Martens E, Mortier A, De Greve H, Proost P, Bossuyt F, downloaded from http://dx.doi.org/10.7287/peerj.preprints.457, last
accessed August 26, 2014). However, our results indicate that
additional mechanisms underlie the transcript diversity in
Salamandridae. First, we observe large deletions (up to 141
AAs) coinciding with one or multiple exons in our SPF transcripts, indicating that transcript diversity is partly caused by
exon-skipping (supplementary fig. S1, Supplementary
Material online). The exons were previously predicted by
aligning our cDNA transcripts with SPF-like genes of
Silurana tropicalis, and by sequencing SPF exon boundaries
in the DNA of one salamandrid species (see Materials and
Methods). Second, we observe nucleotide insertions that are
not a multiple of three, and cause frameshift mutations (supplementary fig. S1, Supplementary Material online). These
transcripts are characterized by the introduction of a premature or delayed stop codon, and may be deleterious (Ohno
1970; Hillman et al. 2004) or generate entirely different
proteins and peptides (Tokunaga et al. 1998; Neu-Yilik et al.
2004; Taylor and Raes 2004; Raes and Van de Peer 2005).
Third, we find transcripts in which the first part is identical
to one transcript while the remaining part corresponds to
another transcript, suggesting recombination at the gene or
transcript level (supplementary fig. S1, Supplementary
Material online). The three kinds of transcripts are a result
of multiple genetic mechanisms in which genomic and/or
transcriptional levels have played a role. These combined observations thus identify SPF as a multiple-gene pheromone
protein family encoded by hypervariable transcripts.
Origin of the SPF Protein Pheromone System
The often large genome size of salamanders has entailed that
no complete caudate genome sequences are available so far,
and the genomic position and origin of SPF pheromoneencoding genes have remained elusive. We therefore identified homologous genes in other vertebrates (see Materials
and Methods) and reconstructed phylogenetic relationships
based on their translated sequences. Comparison of our SPF
sequences with domain databases identifies salamander pheromone precursors as members of the Three-Finger Protein
(TFP) superfamily. These proteins contain one or several domains of a three-finger motif, which is defined by a distinct
disulfide-bonding pattern between cysteines (Petranka et al.
1992; Fleming et al. 1993; Casey et al. 1994; Noel et al. 1998;
Palmer et al. 2007). The full length of most caudate SPF precursors encodes for a two-domain TFP (2D-TFP), with an
anterior 10-cysteine motif followed by a posterior 8-cysteine
motif, or with an anterior 8-cysteine motif followed by a
posterior 6-cysteine motif) (supplementary fig. S2,
Supplementary Material online). Screening of the S. tropicalis
frog genome for 2D-TFP genes resulted in 26 sequences on 11
incomplete genomic scaffolds, likely representing fragments
473
Janssenswillen et al. . doi:10.1093/molbev/msu316
of one or more larger gene clusters. The Anolis carolinensis
lizard genome revealed 19 sequences in a single cluster (17 of
which have a complete 2D-TFP; supplementary table S2,
Supplementary Material online). In contrast with the large
amount of genes in frog and anole, we retrieved nine (of
which four with a complete 2D-TFP) genes in the zebrafish,
three in sheep, two in fugu, and a single 2D-TFP gene in
coelacanth, several mammals (human, mouse, gibbon, elephant), and birds (chicken and zebrafinch) (supplementary
table S2, Supplementary Material online). Synteny of the 2DTFP-surrounding genes was found in the anole lizard, human,
mouse, and one of the 11 S. tropicalis genomic scaffolds
(fig. 1A and supplementary table S2, Supplementary
Material online).
To infer phylogenetic relationships among vertebrate
2D-TFP genes, and to identify the origin of the SPF protein
pheromone system, we combined a representative set of
salamandrid SPF-precursors with relevant vertebrate translated genes and transcripts in a data matrix of 95 sequences.
Sequences of teleost fishes were chosen as outgroup sequences, a choice that is supported by phylogenetic analyses
of TFP-domains (see Materials and Methods, and supplementary fig. S3, Supplementary Material online). Bayesian and
maximum likelihood (ML) analyses, using 12 alternative
alignment strategies, consistently recovered two amphibianspecific 2D-TFP clades, indicating that this result is independent of alignment ambiguity (supplementary table S3,
Supplementary Material online). Importantly, all pheromone
precursors of Plethodontidae and Salamandridae, including
that of the decapeptide sodefrin in Cynops, together with a
single frog sequence fall in a distinct clade that is strongly
supported (posterior probability = 0.98; fig. 1b, henceforth referred to as the SPF clade). The position of the frog sequence
within the SPF clade remains unclear, and low support values
could indicate a basal position (all salamander SPFs form a
clade) or a position more closely related to part of the salamander SPFs. Expressed sequence tags (ESTs) of this SPF gene
in the frog S. tropicalis were detected in the gastrula stage of a
Silurana embryo (supplementary table S2, Supplementary
Material online), where it is not expected to have a pheromone function, but its presence in sexually active frogs is
unknown. Similarly, little is known about the function of
2D-three-finger motif (TFM) proteins outside the salamander
SPF clade in both amniotes and amphibians. In snakes, several
2D-TFM proteins have been identified as phospholipase inhibitors (PLIs). They are present in the snake’s blood and have
been postulated to be endogenous venom neutralizers
against their phospholipases A2 (Dunn and Broady 2001).
In anole lizards, EST counts of 2D-TFM proteins indicate
that they are mainly expressed in male testes and dewlap,
and in female ovaries, suggesting a function in reproduction
(supplementary table S2, Supplementary Material online).
However, at present, no pheromone function has been
shown for any of these proteins.
Phylogenetic relationships of sequences in the SPF clade
indicate several salamander-specific gene duplications
(fig. 1b), which gave rise to multiple copies of SPF genes
from which the pheromone repertoire could expand.
474
MBE
Interestingly, each of the species for which we found SPF
precursors has retained copies that represent the basal split
in SPF diversification (fig. 1b). This confirms that these species
produce proteins that diverged already in the Late Paleozoic
(Van Bocxlaer I, Treer D, Maex M, Vandebergh W,
Janssenswillen S, Stegen G, Kok PJR, Willaert B, Matthijs S,
Martens E, Mortier A, De Greve H, Proost P, Bossuyt F, downloaded from http://dx.doi.org/10.7287/peerj.preprints.457, last
accessed August 26, 2014). To recognize this fact, we further
refer to these proteins as alpha SPFs (fig. 1b, red clade) and
beta SPFs (fig. 1b, green clade).
Multiple Pheromone Precursors Are Present in TailFanning Newts
Given that the pheromone function of SPF likely dates back
to the earliest salamanders, all extant salamanders potentially
could make use of these pheromones during courtship.
However, since courtship strategies vary enormously between
salamanders, it is equally possible that some of these pheromones have become less important, were lost, or shifted expression to other glands during caudate evolution. To predict
whether cloacal SPF maintained a pheromone function in
aquatically courting salamandrids, we compared the
number of precursors expressed in the cloaca and/or abdominal gland of multiple species. Screening of cDNA often shows
a high diversity of precursors in the sexually dimorphic glands
(Houck and Arnold 2003; Palmer et al. 2007) and high precursor abundance can therefore be used as an indication of
pheromone use (Palmer et al. 2007).
We found important differences in the number of transcripts in the cloaca, ranging from zero to 47. As could be
expected, the male cloaca on average contains much more
SPF precursors than that of females (table 1). Additionally, the
two species in our study that both independently lost tailfanning display during courtship (Houck and Arnold 2003)
had the least amount of SPF precursors in their cloacal region
(fig. 2). First, in Taricha granulosa, only five SPF transcripts
were detected using a large set of primers. Males of this species rub their submandibular glands over the female’s nostrils
during dorsal amplexus, and it has been suggested that pheromones are transferred via that gland (Smith 1941; Davis and
Twitty 1964; Sever 2003). Second, in Euproctus platycephalus,
we could not detect SPF precursors in the cloaca, despite
intensive screening with multiple primer combinations. The
male of this species transfers spermatophores from his hooklike extended cloaca to the female cloaca (Alcher 1981; Wells
2010). This direct transmission, sometimes assisted by the
spurs on the male hind legs, resembles copulation and may
have obviated pheromone use from the cloacal region. Our
screening suggests that, in both lineages independently, loss
of tail-fanning behavior has been accompanied with a reduced importance or loss of SPF in their cloaca. In contrast,
our study predicts that SPF is likely to have a pheromone
function in the majority of tail-fanning, aquatically reproducing salamandrids. Although these species display essential
variations in courtship behavior, they all use some form of
tail-fanning to induce the typical female responses that lead
MBE
Salamander Sex Pheromone System . doi:10.1093/molbev/msu316
(a)
PLI
XRCC1 ZnF ETHE1
01
PHLDB LYPD3 SPF
02
03 04
PLI
05
CD177
06
S. tropicalis
A. carolinensis*
H. sapiens
M. musculus
01
to 15
01
17
16
*
01
(b)
D. rerio 01
D. rerio 02
D. rerio 03
T. rubripes 02
T. rubripes 01
A. japonica 02
A. japonica 01
D. rerio 04
L. chalumnae 01
Fishes
Coelacanth
C. pyrrhogaster 04
C. pyrrhogaster 03
A. mexicanum 01
A. mexicanum 02
S. tropicalis 25
Amphibians
Alpha SPF
S. tropicalis 03
S. tropicalis 05
S. tropicalis 06
S. tropicalis 04
S. tropicalis 21
S. tropicalis 26
S. tropicalis 22
S. tropicalis 01
S. tropicalis 10
S. tropicalis 09
S. tropicalis 08
S. tropicalis 14
S. tropicalis 24
S. tropicalis 23
S. tropicalis 13
S. tropicalis 17
S. tropicalis 18
S. tropicalis 19
S. tropicalis 12
S. tropicalis 11
S. tropicalis 16
S. tropicalis 20
S. tropicalis 15
S. tropicalis 07
S. tropicalis 02
D. ocoee alpha SPF
P. stormi
A. ferreus
E. Guttolineata
L. helveticus alpha SPF 3
L. helveticus 20
L. vulgaris 18
I. alpestris 10
P. granulosus 08
C. pyrrhogaster 28
P. waltl 09
T. granulosa 04
N. viridescens 36
A. mexicanum 04
A. tigrinum 01
A. mexicanum 03
L. helveticus beta SPF 1
T. granulosa 02
C. ensicauda silefrin
C. pyrrhogaster sodefrin
C. pyrrhogaster 02
I. alpestris 04
L. helveticus 18
P. granulosus 06
P. waltl 07
N. viridescens 03
L. vulgaris 02
M. musculus 01
L. africana 01
H. sapiens 01
N. leucogenys 01
0. aries 03
0. aries 01
0. aries 02
A. carolinensis 16
A. carolinensis 15
A. carolinensis 06
A. carolinensis 07
A. carolinensis 10
A. carolinensis 11
A. carolinensis 05
A. carolinensis 12
A. carolinensis 03
A. carolinensis 08
A. carolinensis 01
A. carolinensis 04
A. carolinensis 02
T. guttata 01
G. gallus 01
P. textilis PLI
E. climacophora PLI
B. jararacussu PLI
N. ater PLI
A. carolinensis 14
Beta SPF
SPF clade
Mammals
Birds & Reptiles
A. carolinensis 13
A. carolinensis 17
0.3
FIG. 1. Evolutionary origin of caudate SPF pheromones. (a) Gene organization and direction of the 2D-TFP genes PLI (orange) and SPF (green) show
synteny in the four vertebrate genomes. Arrows indicate the direction of the genes as organized in the genome. The genes PLI 17 to LYPD3 of Anolis
carolinensis (indicated with a dashed line) have undergone an inversion and are here shown in the opposite direction to illustrate synteny. XRCC1, X-ray
repair crosscomplementing protein 1; ZnF, zinc finger protein; ETHE1, ethylmalonic encephalopathy 1; PHLDB, pleckstrin homology-like domain, family
B; LYPD3, Lymphocyte antigen 6/plasminogen activator urokinase receptor 3; CD177, CD177 antigen. (b) Bayesian consensus tree for SPF origin and
diversification in vertebrates. Black squares correspond to Bayesian posterior probabilities greater than 0.95. Consistently supported clades of fishes,
mammals, and reptiles and birds are marked in purple, dark blue, and orange, respectively. Amphibian sequences form two distinct clades: One clade is
indicated in light blue and the other clade is referred to as the SPF clade (indicated with a gray square). The SPF clade contains caudate alpha (red) and
beta (green) SPF sequences and a single Silurana tropicalis sequence. Red and green circles indicate precursors of peptides and proteins with a
demonstrated pheromone function.
475
Janssenswillen et al. . doi:10.1093/molbev/msu316
MBE
FIG. 2. Pheromone precursor diversity during salamandrid evolution. The phylogenetic tree follows Zhang et al. (2008). The presence of amplexus and
tail-fanning behavior during courtship is indicated per species as present (black circle), absent, (white circle), or facultative (black and white circle). The
number of alpha and beta SPF transcripts (contigs at 99% identity) in males is indicated at the right (alpha SPF in red, beta SPF in green, beta SPF
sodefrin in orange). The red and green squares indicate the inferred presence of alpha and beta SPF precursors, respectively. The red and green triangles
denote the inferred loss of alpha and beta SPF, respectively. The orange triangle represents the inferred origin of beta SPF sodefrin. The origin of tailfanning (+TF), the loss of amplexus (-A), and the independent loss of tail-fanning (TF) in Taricha and Euproctus are shown. The figures illustrate the
courtship strategies (male in gray, female in white).
to spermatophore pickup (Arnold 1977; Malacarne and
Giacoma 1986; Halliday 1990; Kikuyama and Toyoda 1999;
Houck and Arnold 2003). Although this study does not
cover possible expression of SPF in other glands (i.e., outside
the cloacal region, of special interest in species that are not
tail-fanning), our results suggest that alpha and beta SPF proteins both may be used as sex pheromones in tail-fanning
salamandrids (fig. 1b). In the clade of Eurasian modern newts
(fig. 2, clade indicated with -A), this has resulted in female
following behavior with no, or a limited amount of physical
contact between both sexes.
The Birth of a New Pheromone
The first characterized salamandrid pheromone was the decapeptide sodefrin, which was shown to function as an attractant in the red-bellied newt, Cynops pyrroghaster (Kikuyama
et al. 1995). Our phylogenetic analyses indicate that the precursors of these pheromones are beta SPF’s (fig. 1b), but their
cDNA sequences essentially differ from other SPF precursors
by the presence of a 62-bp insert (fig. 3). As a consequence,
the last part of the sodefrin precursor has an amino acid
sequence that results from a frameshift and cannot be aligned
with any of the other SPF sequences. Interestingly, it is exactly
the frameshifted sequence that contains the sodefrin amino
acid sequence SIPSKDALLK (fig. 3). The 62-bp insert and the
resulting frameshifted sequence contain several lysine–
arginine (KR), arginine–arginine (RR), or arginine-x-x(-x-x)-arginine positions (RXX(XX)R) that typically represent cleavage
sites in secretory peptide precursors (fig. 3). The latter flanks
476
the sodefrin decapeptide and allows the posttranslational
cleavage from its precursor to SIPSKDALLK-ISA, followed by
further degradation or cleavage of the last three amino acids.
To reconstruct the origin of this decapeptide, we investigated
whether other species also expressed these deviating beta SPF
precursors that would allow expression of such a peptide.
Despite intense screening with multiple pairs of primers specifically designed to detect precursors of sodefrin and its variants (supplementary table S1, Supplementary Material
online, see also Materials and Methods), we could not
detect such precursors in any of the other salamanders.
One of the SPF precursors of L. montandoni shows the
same decapeptide when a frameshift is introduced in silico
(Osikowski et al. 2008), but no precursors that actually could
result in cleavage of that decapeptide were detected. This
strongly indicates that the sodefrin precursor originated
after the split of Pachytriton from Cynops, which has been
estimated at about 34.3 Ma (Zhang et al. 2008). Furthermore,
the presence of a variant in the sword-tailed newt, C. ensicauda (Yamamoto et al. 2000) indicates that the decapeptide
originated before the divergence of C. pyrrhogaster from C.
ensicauda, an event that has been estimated at 26.5 Ma
(Zhang, et al. 2008). As a consequence, the most parsimonious interpretation of our analyses indicates that the decapeptide pheromone originated approximately between 34.3 and
26.5 Ma, and obtained a pheromone function in this ancestral
Cynops lineage.
Our analyses throw light on how a random male signal can
become a pheromone for females, even though the new
decapeptide shows no similarity to existing salamander
Salamander Sex Pheromone System . doi:10.1093/molbev/msu316
Beta SPF
Sodefrin
Beta SPF
Sodefrin
Beta SPF
Sodefrin
Beta SPF
Sodefrin
Beta SPF
Sodefrin
MBE
***************************** **************************************************** *********** ****** ****** ******* ** **
ctattatgcgagcagtgtttcgctctccacgccagcagctgctcgggtatcttcacgcagtgctctcctgacgtcactcactgtgtcgcaggcctcgagaacagcacactggggactgatatt
L L C E Q C F A L H A S S C S G I F T Q C S P D V T H C V A G L E N S T L G T D I
ctattatgcgagcagtgtttcgctctccaaaccagcagctgctcgggtatcttcacgcagtgctctcctgacgtcactcactgcgtcgcaggcctagagaactgcacaccggggactcatgtt
L L C E Q C F A L Q T S S C S G I F T Q C S P D V T H C V A G L E N C T P G T H V
*********** ************************ ***** ******* * ****** ****** ** * * *** *
*****
*
* *** * ******************
attctaactgcattcaaggactgtctggatccttcccaaaaatcagcctgtagcagagagttctcctccagatcctcagtgttttctttccagctcaacagaatttgctgtgactctgatttc
I L T A F K D C L D P S Q K S A C S R E F S S R S S V F S F Q L N R I C C D S D F
attctaactgcgttcaaggactgtctggatccttccgaaaaagcagcctgcggtagagaggtctccttcacagctccagcggcctctttatggacaagcaggacatgctgtgactctgatttc
I L T A F K D C L D P S E K A A C G R E V S F T A P A A S L W T S R T C C D S D F
***** ******************** ** ***** **** **** **
**** ** *******
*********
tgcaatggcggggatgtgcaggtgccttctgcagacaacacccccaacggatacatatgtgaagattgtttcaatgaccagtccacggacccctgcaca-----------------------C N G G D V Q V P S A D N T P N G Y I C E D C F N D Q S T D P C T
tgcaacagcggggatgtgcaggtgcctcctccagacgacactcccagtgg---------------ttgtggcagtgaccag---------ccctgcacagcgccagaacacctaagggaaaca
C N S G D V Q V P P P D D T P S G
C G S D Q
P C T A P E H L R E T
******
* **** ** ***
** ***** * * * ** ************* ********* * * * *****
-------------------------------------- gcgactggagtagttcagtgcaccgggaagcagaacacttgtctcagcttctatggaactagctccaggcctggtgaagctggg
A T G V V Q C T G K Q N T C L S F Y G T S S R P G E A G
gtgcgctctacaagatcgattagagagaagagaagaga gcgacttttttggttccatgttccgatcagaagaacgcatctggcaccttctatggaactgcctccaggccggctaagactgggg
V R S T R S I R E K R R E
R L F W F H V P I R R T H L A P S M E L P P G R L R L G
***** *
**************** *** ******* *
agaccgtatagcgggaaaggatgcactactcgagacttctgcaaacttggaatttttaacctggcaggaacacaagccaatgatagtggtttcaagtgttcccctgccctgaaactttga
R P Y S G K G C T T R D F C K L G I F N L A G T Q A N D S G F K C S P A L K L aggagtataccttcaaaggatgcactactcaagatttctgcatag
R S I P S K D A L L K I S A -
FIG. 3. The origin of a decapeptide pheromone. Comparison of the nucleotide and amino acid sequences of a beta SPF precursor with a sodefrin
precursor. The 62-bp insert (turquoise) has caused a frameshift that leads to the origin of the decapeptide sequence (orange rectangle) and a premature
stop codon. Asterisks show identical nucleotides.
signals. First, the new peptide could easily originate, because it
is derived from a family of proteins that are abundantly expressed. All SPF precursors likely share the same biosynthetic
pathways and thus form an ideal starting point for the origin
of new signals. Organisms often secrete chemicals as by-products and waste products (Steiger et al. 2010) that initially do
not contain intended information (Schaefer and Ruxton
2012). Second, the signal could easily be maintained without
having a pheromone function yet. Indeed, our phylogenetic
reconstructions indicate that both alpha and beta SPF proteins were expressed in the Cynops ancestor at the timing of
the origin of sodefrin (fig. 2), and that the new signal originated in the shadow of existing uncleaved pheromones. The
decapeptide could thus be retained before it had a biological
function, because the other transcripts continued to fulfill
their pheromone function (Raes and Van de Peer 2005).
Finally, we hypothesize that this association with functional
pheromones has predestined the decapeptide to become a
sex pheromone. Because both the place and timing of expression are identical, the new cue not only was more likely to
provide a starting point for a relevant signal, but was also
biased toward a function in sexual communication. Most
animal olfactory systems are equipped with a broad enough
range of relatively nonspecific receptors to allow such new
chemicals to be detected and evolve into a pheromone
(Wyatt 2003).
Materials and Methods
and Pleurodeles waltl (primitive newt with amplexus). We also
sampled two species that do not show tail-fanning behavior
during courtship, E. platycephalus (a newt with amplexus
from Sardinia) and T. granulosa (a New World newt with
amplexus). Additionally, we analysed the cloaca of the females
of five tail-fanning species (L. vulgaris, L. helveticus, I. alpestris,
N. viridescens, P. waltl) that show a variety of courtship strategies. Animals were collected 1) from a pond in Haacht,
Belgium (L. vulgaris, L. helveticus and I. alpestris) with permission of Agentschap voor Natuur en Bos (permit ANB/BL-FF/
V12-00050); 2) from the pet trade Squama, Herent, Belgium
(N. viridescens, Pa. granulosus and C. pyrrhogaster); 3) or from
hobby-breeders (P. waltl and E. platycephalus). Animals were
brought into courtship mood by species-specific temperature/daylight/moisture manipulation. The courtship mood
was assessed by 1) a swollen belly in females, and 2) species-specific morphological changes in males: Swollen cloaca
(all species), tail crest (L. helveticus, L. vulgaris, I. alpestris),
bright white spots at the end of the tail (Pa. granulosus),
nuptial pads and/or spurs on limbs (E. platycephalus, N. viridescens, T. granulosa, P. waltl), overall skin roughening (T.
granulosa), and neck glands (C. pyrrhogaster). Additionally,
if both sexes were available, opposite genders were brought
together for mating, and only females that showed responsive
courtship behavior were used in the study. Animals were
euthanized in accordance with EU and Belgian regulations
concerning animal welfare, and the cloacal region was sampled in 1 ml of RNA-later (Life Technologies).
Species and Tissue Sampling
We sampled the cloaca of the following broad range of tailfanning salamandrid representatives, based on their phylogenetic position (Zhang et al. 2008) and variability in mating
behavior strategies: Pachytriton granulosus and C. pyrrhogaster (modern Asian newts), L. vulgaris, L. helveticus, and
Ichthyosaura alpestris (modern European newts),
Notophthalmus viridescens (New World newt with amplexus),
Amplification and Cloning of Salamander SPF
Precursors
For each tissue, 1 mg total RNA (A260/A280 ratio’s between 1.7
and 1.8) was isolated with Tri Reagent T9424 (Sigma Aldrich).
First strand cDNA was generated with the SMARTer RACE
cDNA Amplification Kit of Clontech. Precursors were
identified using both polymerase chain reaction (PCR) and
477
MBE
Janssenswillen et al. . doi:10.1093/molbev/msu316
RACE-PCR. For PCR, we screened with two degenerate SPF
primers (SPF1 and SPF2, see supplementary table S1,
Supplementary Material online; [von Reis 2007]). The PCR
reaction was executed with Advantage High Fidelity 2 PCR
kit (15 ml PCR-grade water; 2.5 ml 5 HF 2PCR buffer; 1 ml
dimethyl sulfoxide (DMSO); 2.5 ml 50 HF 2 dNTP mix; 1 ml
SPF1F primer (10 mM); 1 ml SPF2R primer (10 mM); 1 ml cDNA
(9 ng/ml); 1 ml 50 Advantage HF 2 polymerase mix) with
cycle 94 C for 3 min/4 (94 C for 45 s/53 C for 45 s/
72 C for 90 s)/30 (94 C for 30 s, 53 C for 30 s/72 C for
1 min)/72 C for 7 min. To enhance our chance of obtaining a
broad range of SPF precursors, we designed 14 new RACEprimers (supplementary table S1, Supplementary Material
online). For RACE-PCR, we used the Advantage High
Fidelity 2 PCR kit (13.5 ml PCR-grade water; 2.5 ml 5 HF
2PCR buffer; 1 ml DMSO; 2.5 ml 50 HF 2 dNTP mix; 2.5 ml
Universal RACE Primer Mix (UPM mix; 2 mM UPM1 + 0.4 mM
UPM2); 1 ml SPF primer (10 mM); 1 ml cDNA (9 ng/ ml); 1 ml
50 Advantage HF 2 polymerase mix) in the cycle reaction
5 (94 C for 30 s/72 C for 3 min), 5 (94 C for 30 s/70 C
for 30 s/72 C for 3 min), 25 (94 C for 30 s/68 C for 30 s/
72 C for 3 min). After checking on a 1% electrophorese gel,
PCR products were purified with the QiaQuick PCR purification kit (Qiagen) and 2 ml was used for cloning into a pGEM-T
Easy vector (PROMEGA). Two microliters of the ligation
products were inserted into 50 ml TOP10 chemically competent cells (Invitrogen) that were grown overnight on LBagar
(25 ml) plates containing Ampicillin (50 ml/ml) and X-Gal
(40 ml 20 mg/ml). Sequencing was done using a Big Dye
Terminator Cycle Sequencing Kit v3.1 (Applied Biosystems,
PE; identical protocol as in manual) on a GeneScan 3100
automated sequencer (Applied BioSystems). For each tissue,
we sequenced one side of 96 colonies that were randomly
selected from amplified products obtained with the largest
primer diversity. Electropherograms were read using the
CodonCode Aligner 3.7.1.1 software package (CodonCode
Corporation) and nucleotide sequences of the coding
region were compiled using a 99% similarity threshold, after
quality trimming to eliminate differences due to PCR error.
Transcripts with premature stop codons greater than 50 nt
upstream of the final exon are often targets of nonsense
mediated decay (Nagy and Maquat 1998), but since many
contraexamples exist (Tokunaga et al. 1998; Neu-Yilik et al.
2004), all transcripts were withheld for comparing interspecific transcript diversity (fig. 2). To understand transcript diversity, we checked for conservation of the exon boundaries
in the DNA of P. waltl. DNA [A260/A280 ratio of 1.9] was
isolated from limb muscle tissue using the DNeasy Blood &
Tissue Kit (Qiagen). DNA amplification was retrieved with
an additional primer set (supplementary table S1,
Supplementary Material online) in a PCR executed with
JumpStart RED Accu Taq LA DNA polymerase (18 ml PCRgrade water; 1.25 ml dNTP, 0.5 ml of each primer (SPF17F
and SPF18R); 2.75 ml buffer; 1 ml Taq polymerase; 1 ml DNA
(50 ng/ml)) in a temperature cycle reaction of 96 C for 30 s/
30 (94 C for 15 s/62 C for 30 s/68 C for 20 min)/68 C for
30 min. After checking on a 1% electrophorese gel, PCR products were purified with the QiaQuick PCR purification kit
478
(Qiagen) and sequencing was done using a Big Dye
Terminator Cycle Sequencing Kit v3.1 (Applied Biosystems,
PE; identical protocol as in manual) on a GeneScan 3100
automated sequencer (Applied BioSystems).
Identification of Related Genes in Vertebrates
Because our salamandrid SPF protein sequences have a cysteine-rich domain, this domain was first identified by comparisons with the NCBI domain CDD v3.11 (Marchler-Bauer
et al. 2011), Swiss-Model v8.05 (Arnold et al. 2006; Bordoli
et al. 2008) and InterProScan 5 (Mulder et al. 2003; Quevillon
et al. 2005) Databases. Next, we used our salamandrid SPF
contigs to search for homologous vertebrate sequences using
the BLAST algorithms (January 2014) at NCBI on translated
nucleotide databases of fishes, coelacanth, amphibians, turtles, crocodiles, birds, reptiles, monotremes, marsupials,
afrotherians, xenarthrans, primates, glires, and laurasiatherians. Sequences with the same domain as SPF (i.e., two domains of a three-finger-motif) and a significant BLAST hit (E
value <1, coverage 4 70%) were retained for phylogenetic
analyses (supplementary table S2, Supplementary Material
online). Genomic information was inferred from BLAT
(January 2014) and BLAST (January 2014) searches on the
genomes of the zebrafish (Danio rerio; Zv9/DanRer7 July
2010), fugu (Takifugu rubripes; v05/fr3 October 2011), coelacanth (Latimeria chalumnae; latCha1 August 2011), the
Western clawed frog (S. tropicalis; JGI4.2/xenTro3
November 2009), Carolina Anole (A. carolinensis; anoCar2.0
May 2010), chicken (Gallus gallus; ICGSC galGal4 November
2011), Zebra Finch (Taeniopygia guttata; taeGut324 February
2013), mouse (Mus musculus; GRCm38/mm10 December
2011), and human (Homo sapiens; GRCh37/hg19 February
2009) on UCSC, NCBI, and Ensembl genome browsers (supplementary table S3, Supplementary Material online). Gene
predictions were executed on neighboring sequences using
PredictProtein release 1.0.88 (Rost et al. 2004) and UniProt
release 2013_12. To have the broadest evolutionary representation of the examined salamandrid cDNA sequences, we
included the two most divergent sequences per species (as
calculated by uncorrected pairwise distances in PAUP* v. 4.2)
(Swofford 1998), together with the precursors of sodefrin in C.
pyrrhogaster and silefrin in C. ensicauda (NCBI nrs.: CAB53093
and CAB53094). As plethodontid representatives, we choose
one SPF precursor transcript per species for Eurycea guttolineata, Aneides ferreus, Plethodon stormi, and Desmognathus
ocoee (NCBI nrs.: AAZ06338, AAZ06335, AAZ06331,
AAZ06329, respectively).
Alignment and Phylogenetic Analyses
Retrieved homologous mRNA sequences were translated into
protein sequences with ExPASy (Gasteiger et al. 2003).
Because our sequences contain two domains of a threefinger motif, we first checked that these domains stem
from the same ancestral TFM-duplication that happened
before vertebrate diversification (i.e., that none of the sequences are the result of a more recent TFM-duplication).
To do that, we created an alignment in which TFM1 and
MBE
Salamander Sex Pheromone System . doi:10.1093/molbev/msu316
TFM2 domains were treated as independent sequences. We
used the local-alignment algorithm with generalized affine
gap costs and iterative refinement (E-INS-I) with standard
parameters implemented in Mafft v7 (Katoh and Standley
2013), and inferred phylogenetic relationships using
MrBayes 3.2.2. (Ronquist et al. 2012) (supplementary fig. S3,
Supplementary Material online). We found a single sequence
with a recent TFM duplication, A. carolinensis 09 (supplementary fig. S3, Supplementary Material online), which was excluded from further analyses. The remaining 95 complete
amino acid sequences (from the start codon Methionine to
the stop codon, thus including TFM1 and TFM2) were aligned
using the local-alignment algorithm with generalized affine
gap costs and iterative refinement (E-INS-I) with standard
parameters implemented in Mafft v7 (Katoh and Standley
2013). To evaluate the robustness of our results against alignment ambiguities, five alternative alignment strategies covering a broad range of alternative gap penalties and offset values
were applied. In addition, six alignments by L-INS-i (an alternative local-alignment algorithm implemented in Mafft) covering the same range of gap penalties were implemented
(supplementary table S3, Supplementary Material online).
Branch support was assessed by Bayesian phylogeny inference
and ML bootstrapping, using MrBayes 3.2.2. (Ronquist et al.
2012) and RAxML 7.0.4 (Stamatakis et al. 2008), respectively.
MrBayes analyses were executed with a mixed prior for the
AA substitution model. Two parallel runs of four Markov
chain Monte Carlo chains were performed, with a length of
10,000,000 generations, a sampling frequency of one per 1,000
generations, and a burn-in corresponding to the first
2,000,000 generations. Convergence of the parallel runs was
confirmed by split frequency standard deviations (<0.01) and
potential scale reduction factors (approximating 1.0) for all
model parameters. Adequate posterior sampling was verified
using Tracer 1.5 (Rambaut and Drummond 2007), by checking if the runs had reached effective sampling sizes greater
than 200 for all model parameters. Bootstrap supports were
obtained by performing 500 “rapid” bootstrap replicates,
using the WAG + G + I model (choice of model based on
the tree with the best likelihood of the EINSI alignment
with standard MAFFT gap opening penalty and offset value
after 500 bootstrap replicates out of three tested AA substitution models [JTT, WAG and LG]).
Supplementary Material
Supplementary tables S1–S3 and figures S1–S3 are available at
Molecular Biology and Evolution online (http://www.mbe.
oxfordjournals.org/).
Acknowledgments
The authors thank Kim Roelants for drawings of animals and
Frank Pasmans for E. platycephalus. The authors also thank
the Agentschap voor Natuur en Bos for giving permits. This
research was supported by a European Research Council
starting grant [ERC 204509, project TAPAS] and the Fonds
voor Wetenschappelijk Onderzoek (FWO) Vlaanderen [grant
no. G.0133.08 and G026715N]. M.M. and I.V.B. received a
fellowship from FWO-Vlaanderen. S.J. is supported by IWTVlaanderen.
References
Alcher M. 1981. Reproduction en e levage de Euproctus platycephalus
(Urodela, Salamandridae). Amphib Reptil. 2:97–105.
Arnold K, Bordoli L, Kopp J, Schwede T. 2006. The SWISS-MODEL
Workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22:195–201.
Arnold SJ. 1977. The evolution of courtship behavior in New World
salamanders with some comments on Old World salamandrids.
In: Taylor DH, Guttman SI, editors. The reproductive biology of
amphibians. New York: Springer. p. 141–183.
Bordoli L, Kiefer F, Arnold K, Benkert P, Battey J, Schwede T. 2008.
Protein structure homology modeling using SWISS-MODEL workspace. Nat Protoc. 4:1–13.
Casey JR, Petranka JG, Kottra J, Fleenor DE, Rosse W. 1994. The structure
of the urokinase-type plasminogen activator receptor gene. Blood
84:1151–1156.
Davis WC, Twitty VC. 1964. Courtship behavior and reproductive isolation in the species of Taricha (Amphibia, Caudata). Copeia 1964:
601–610.
Dunn RD, Broady KW. 2001. Snake inhibitors of phospholipase A2 enzymes. Biochim Biophys Acta. 1533:29–37.
Fleming TJ, O’huigin C, Malek T. 1993. Characterization of two novel Ly6 genes. Protein sequence and potential structural similarity to
alpha-bungarotoxin and other neurotoxins. J Immunol. 150:
5379–5390.
Gasteiger E, Gattiker A, Hoogland C, Ivanyi I, Appel RD, Bairoch A. 2003.
ExPASy: the proteomics server for in-depth protein knowledge and
analysis. Nucleic Acids Res. 31:3784–3788.
Halliday T. 1977. The courtship of European newts: an evolutionary
perspective. In: Taylor DH, Guttman SI, editors. The reproductive
biology of amphibians. New York: Springer. p. 185–232.
Halliday T. 1990. The evolution of courtship behavior in newts and
salamanders. Adv Study Behav. 19:137–169.
Hillman RT, Green RE, Brenner SE. 2004. An unappreciated role for RNA
surveillance. Genome Biol. 5:R8.
Houck LD, Arnold S. 2003. Courtship and mating behavior. In: Sever DM,
editor. Reproductive biology and phylogeny of Urodela. Enfield
(NH): Science Publishers, Inc.. p. 383–424.
Houck LD, Reagan NL. 1990. Male courtship pheromones increase
female receptivity in a plethodontid salamander. Anim Behav. 39:
729–734.
Houck LD, Sever DM. 1994. The role of the skin in reproduction and
behavior. In: Heatwole H, Barthalamus G, editors. Amphibian biology. Chipping Norton: Surrey Beatty and Sons. p. 351–381.
Houck LD, Watts RA, Mead LM, Palmer CA, Arnold SJ, Feldhoff PW,
Feldhoff RC. 2008. A candidate vertebrate pheromone, SPF, increases
female receptivity in a salamander. In: Hurst J, Beynon R, Roberts SC,
Wyatt T, editors. Chemical signals in vertebrates 11. Springer.
New York, USA. p. 213–221.
Iwata T, Umezawa K, Toyoda F, Takahashi N, Matsukawa H, Yamamoto
K, Miura S, Hayashi H, Kikuyama S. 1999. Molecular cloning of newt
sex pheromone precursor cDNAs: evidence for the existence of
species-specific forms of pheromones. FEBS Lett. 457:400–404.
Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol
Evol. 30:772–780.
Kikuyama S, Toyoda F. 1999. Sodefrin: a novel sex pheromone in a newt.
Rev Reprod. 4:1–4.
Kikuyama S, Toyoda F, Ohmiya Y, Matsuda K, Tanaka S, Hayashi H. 1995.
Sodefrin: a female-attracting peptide pheromone in newt cloacal
glands. Science 267:1643–1645.
Malacarne G, Giacoma C. 1986. Chemical signals in European newt
courtship. Ital J Zool. 53:79–83.
Marchler-Bauer A, Lu S, Anderson JB, Chitsaz F, Derbyshire MK,
DeWeese-Scott C, Fong JH, Geer LY, Geer RC, Gonzales NR. 2011.
479
Janssenswillen et al. . doi:10.1093/molbev/msu316
CDD: a conserved domain database for the functional annotation of
proteins. Nucleic Acids Res. 39:D225–D229.
Mulder NJ, Apweiler R, Attwood TK, Bairoch A, Barrell D, Bateman A,
Binns D, Biswas M, Bradley P, Bork P. 2003. The InterPro Database,
2003 brings increased coverage and new features. Nucleic Acids Res.
31:315–318.
Nagy E, Maquat LE. 1998. A rule for termination-codon position within
intron-containing genes: when nonsense affects RNA abundance.
Trends Biochem Sci. 23:198–199.
Neu-Yilik G, Gehring NH, Hentze MW, Kulozik AE. 2004. Nonsensemediated mRNA decay: from vacuum cleaner to Swiss army knife.
Genome Biol. 5:218.
Noel LS, Champion BR, Holley CL, Simmons CJ, Morris DC, Payne JA,
Lean JM, Chambers TJ, Zaman G, Lanyon LE. 1998. RoBo-1, a novel
member of the urokinase plasminogen activator receptor/CD59/Ly6/snake toxin family selectively expressed in rat bone and growth
plate cartilage. J Biol Chem. 273:3878–3883.
Ohno S. 1970. Evolution by gene duplication. London: George Alien &
Unwin Ltd.
Osikowski A, Babik W, Grzmil P, Szymura JM. 2008. Multiple sex pheromone genes are expressed in the abdominal glands of the Smooth
Newt (Lissotriton vulgaris) and Montandon’s Newt (L.
montandoni)(Salamandridae). Zool Sci. 25:587–592.
Palmer CA, Watts RA, Houck LD, Picard AL, Arnold SJ. 2007.
Evolutionary replacement of components in a salamander pheromone signaling complex: more evidence for phenotypic-molecular
decoupling. Evolution 61:202–215.
Petranka JG, Fleenor DE, Sykes K, Kaufman RE, Rosse WF. 1992. Structure
of the CD59-encoding gene: further evidence of a relationship to
murine lymphocyte antigen Ly-6 protein. Proc Natl Acad Sci U S A.
89:7876–7879.
Quevillon E, Silventoinen V, Pillai S, Harte N, Mulder N, Apweiler R,
Lopez R. 2005. InterProScan: protein domains identifier. Nucleic
Acids Res. 33:W116–W120.
Raes J, Van de Peer Y. 2005. Functional divergence of proteins through
frameshift mutations. Trends Genet. 21:428–431.
Rambaut A, Drummond A. 2007. Tracer v1. 5. [cited 2014 Jan 31].
Available from: http://beast.bio.ed.ac.uk/Tracer.
Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, H€ohna S,
Larget B, Liu L, Suchard MA, Huelsenbeck JP. 2012. MrBayes 3.2:
480
MBE
efficient Bayesian phylogenetic inference and model choice across
a large model space. Syst Biol. 61:539–542.
Rost B, Yachdav G, Liu J. 2004. The predictprotein server. Nucleic Acids
Res. 32:W321–W326.
Schaefer MH, Ruxton G. 2012. By–product information can stabilize the
reliability of communication. J Evol Biol. 25:2412–2421.
Sever D. 2003. Courtship and mating glands. In: Sever D, editor.
Reproductive biology and phylogeny of Urodela (Amphibia).
Enfield (NH): Science Publishers, Inc.. p. 323–381.
Smith R. 1941. Mating behavior in Triturus torosa and related newts.
Copeia 255–262.
Stamatakis A, Hoover P, Rougemont J. 2008. A rapid bootstrap algorithm for the RAxML web servers. Syst Biol. 57:758–771.
Steiger S, Schmitt T, Schaefer HM. 2010. The origin and dynamic evolution of chemical information transfer. Proc R Soc B Biol Sci. 970–979.
Swofford D. 1998. Paup*, Phylogenetic analysis using parsimony (* and
other methods), version 4. Sunderland (MA): Sinauer Associates.
Taylor JS, Raes J. 2004. Duplication and divergence: the evolution of new
genes and old ideas. Annu Rev Genet. 38:615–643.
Tokunaga C, Tatematsu K, Kuroda Si, Nakagawa N, Kikkawa U. 1998.
Molecular cloning and characterization of RBCK2, a splicing variant
of a RBCC family protein, RBCK1. FEBS Lett. 435:11–15.
von Reis M. 2007. Evolution of sodefrin precursor pheromones in salamandrid newts. [M.Sc. thesis]. Oregon State University, Corvallis, OR
97330, USA.
Wambreuse P, Bels V. 1984. Analyse qualitative et quantitative de la
parade sexuelle du triton palme Triturus helveticus (Razoumowsky
1798). Cah Ethol Appl. 4:193–218.
Wells KD. 2010. The ecology and behavior of amphibians. Chicago:
University of Chicago Press.
Wyatt TD. 2003. Pheromones and animal behaviour: communication by
smell and taste. Cambridge University Press.
Yamamoto K, Kawai Y, Hayashi T, Ohe Y, Hayashi H, Toyoda F,
Kawahara G, Iwata T, Kikuyama S. 2000. Silefrin, a sodefrin-like pheromone in the abdominal gland of the sword-tailed newt, Cynops
ensicauda. FEBS Lett. 472:267–270.
Zhang P, Papenfuss TJ, Wake MH, Qu L, Wake DB. 2008. Phylogeny and
biogeography of the family Salamandridae (Amphibia: Caudata) inferred from complete mitochondrial genomes. Mol Phylogenet Evol.
49:586–597.